diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,90846 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 12972,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 7.70891150169596e-05,
+      "grad_norm": 8.23936653137207,
+      "learning_rate": 2.5641025641025643e-08,
+      "loss": 1.5914,
+      "step": 1
+    },
+    {
+      "epoch": 0.0001541782300339192,
+      "grad_norm": 8.8346586227417,
+      "learning_rate": 5.1282051282051286e-08,
+      "loss": 1.5383,
+      "step": 2
+    },
+    {
+      "epoch": 0.00023126734505087883,
+      "grad_norm": 10.278841972351074,
+      "learning_rate": 7.692307692307694e-08,
+      "loss": 1.6608,
+      "step": 3
+    },
+    {
+      "epoch": 0.0003083564600678384,
+      "grad_norm": 11.908944129943848,
+      "learning_rate": 1.0256410256410257e-07,
+      "loss": 1.6491,
+      "step": 4
+    },
+    {
+      "epoch": 0.000385445575084798,
+      "grad_norm": 8.718999862670898,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 1.5346,
+      "step": 5
+    },
+    {
+      "epoch": 0.00046253469010175765,
+      "grad_norm": 8.059788703918457,
+      "learning_rate": 1.5384615384615387e-07,
+      "loss": 1.53,
+      "step": 6
+    },
+    {
+      "epoch": 0.0005396238051187173,
+      "grad_norm": 9.216470718383789,
+      "learning_rate": 1.7948717948717948e-07,
+      "loss": 1.5931,
+      "step": 7
+    },
+    {
+      "epoch": 0.0006167129201356768,
+      "grad_norm": 9.295415878295898,
+      "learning_rate": 2.0512820512820514e-07,
+      "loss": 1.5061,
+      "step": 8
+    },
+    {
+      "epoch": 0.0006938020351526364,
+      "grad_norm": 7.587814807891846,
+      "learning_rate": 2.307692307692308e-07,
+      "loss": 1.5034,
+      "step": 9
+    },
+    {
+      "epoch": 0.000770891150169596,
+      "grad_norm": 10.28200912475586,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 1.5683,
+      "step": 10
+    },
+    {
+      "epoch": 0.0008479802651865557,
+      "grad_norm": 10.186943054199219,
+      "learning_rate": 2.820512820512821e-07,
+      "loss": 1.612,
+      "step": 11
+    },
+    {
+      "epoch": 0.0009250693802035153,
+      "grad_norm": 8.96688175201416,
+      "learning_rate": 3.0769230769230774e-07,
+      "loss": 1.5478,
+      "step": 12
+    },
+    {
+      "epoch": 0.0010021584952204748,
+      "grad_norm": 8.598114013671875,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 1.77,
+      "step": 13
+    },
+    {
+      "epoch": 0.0010792476102374346,
+      "grad_norm": 8.468049049377441,
+      "learning_rate": 3.5897435897435896e-07,
+      "loss": 1.5732,
+      "step": 14
+    },
+    {
+      "epoch": 0.001156336725254394,
+      "grad_norm": 8.353447914123535,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 1.6234,
+      "step": 15
+    },
+    {
+      "epoch": 0.0012334258402713536,
+      "grad_norm": 8.753851890563965,
+      "learning_rate": 4.102564102564103e-07,
+      "loss": 1.6226,
+      "step": 16
+    },
+    {
+      "epoch": 0.0013105149552883133,
+      "grad_norm": 7.985566139221191,
+      "learning_rate": 4.358974358974359e-07,
+      "loss": 1.5986,
+      "step": 17
+    },
+    {
+      "epoch": 0.0013876040703052729,
+      "grad_norm": 9.336605072021484,
+      "learning_rate": 4.615384615384616e-07,
+      "loss": 1.4641,
+      "step": 18
+    },
+    {
+      "epoch": 0.0014646931853222326,
+      "grad_norm": 8.00399398803711,
+      "learning_rate": 4.871794871794872e-07,
+      "loss": 1.5511,
+      "step": 19
+    },
+    {
+      "epoch": 0.001541782300339192,
+      "grad_norm": 8.389213562011719,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 1.5059,
+      "step": 20
+    },
+    {
+      "epoch": 0.0016188714153561516,
+      "grad_norm": 14.020890235900879,
+      "learning_rate": 5.384615384615386e-07,
+      "loss": 1.7145,
+      "step": 21
+    },
+    {
+      "epoch": 0.0016959605303731114,
+      "grad_norm": 6.704775810241699,
+      "learning_rate": 5.641025641025642e-07,
+      "loss": 1.5278,
+      "step": 22
+    },
+    {
+      "epoch": 0.0017730496453900709,
+      "grad_norm": 6.921572208404541,
+      "learning_rate": 5.897435897435898e-07,
+      "loss": 1.502,
+      "step": 23
+    },
+    {
+      "epoch": 0.0018501387604070306,
+      "grad_norm": 11.839808464050293,
+      "learning_rate": 6.153846153846155e-07,
+      "loss": 1.4523,
+      "step": 24
+    },
+    {
+      "epoch": 0.0019272278754239901,
+      "grad_norm": 7.50053071975708,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 1.4585,
+      "step": 25
+    },
+    {
+      "epoch": 0.0020043169904409497,
+      "grad_norm": 8.217403411865234,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.5492,
+      "step": 26
+    },
+    {
+      "epoch": 0.0020814061054579094,
+      "grad_norm": 7.217784881591797,
+      "learning_rate": 6.923076923076924e-07,
+      "loss": 1.424,
+      "step": 27
+    },
+    {
+      "epoch": 0.002158495220474869,
+      "grad_norm": 7.2611799240112305,
+      "learning_rate": 7.179487179487179e-07,
+      "loss": 1.4658,
+      "step": 28
+    },
+    {
+      "epoch": 0.0022355843354918284,
+      "grad_norm": 6.959589958190918,
+      "learning_rate": 7.435897435897436e-07,
+      "loss": 1.4955,
+      "step": 29
+    },
+    {
+      "epoch": 0.002312673450508788,
+      "grad_norm": 8.893731117248535,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 1.3793,
+      "step": 30
+    },
+    {
+      "epoch": 0.002389762565525748,
+      "grad_norm": 6.077394485473633,
+      "learning_rate": 7.948717948717949e-07,
+      "loss": 1.4154,
+      "step": 31
+    },
+    {
+      "epoch": 0.002466851680542707,
+      "grad_norm": 6.705117225646973,
+      "learning_rate": 8.205128205128206e-07,
+      "loss": 1.4288,
+      "step": 32
+    },
+    {
+      "epoch": 0.002543940795559667,
+      "grad_norm": 6.546511173248291,
+      "learning_rate": 8.461538461538463e-07,
+      "loss": 1.3884,
+      "step": 33
+    },
+    {
+      "epoch": 0.0026210299105766267,
+      "grad_norm": 6.519981384277344,
+      "learning_rate": 8.717948717948718e-07,
+      "loss": 1.3881,
+      "step": 34
+    },
+    {
+      "epoch": 0.002698119025593586,
+      "grad_norm": 7.000728130340576,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 1.4235,
+      "step": 35
+    },
+    {
+      "epoch": 0.0027752081406105457,
+      "grad_norm": 6.453318119049072,
+      "learning_rate": 9.230769230769232e-07,
+      "loss": 1.4284,
+      "step": 36
+    },
+    {
+      "epoch": 0.0028522972556275054,
+      "grad_norm": 5.742129802703857,
+      "learning_rate": 9.487179487179487e-07,
+      "loss": 1.4163,
+      "step": 37
+    },
+    {
+      "epoch": 0.002929386370644465,
+      "grad_norm": 5.8094916343688965,
+      "learning_rate": 9.743589743589745e-07,
+      "loss": 1.293,
+      "step": 38
+    },
+    {
+      "epoch": 0.0030064754856614245,
+      "grad_norm": 6.008228302001953,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.4182,
+      "step": 39
+    },
+    {
+      "epoch": 0.003083564600678384,
+      "grad_norm": 5.845043659210205,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.444,
+      "step": 40
+    },
+    {
+      "epoch": 0.003160653715695344,
+      "grad_norm": 5.4758806228637695,
+      "learning_rate": 1.0512820512820514e-06,
+      "loss": 1.396,
+      "step": 41
+    },
+    {
+      "epoch": 0.0032377428307123032,
+      "grad_norm": 5.759223937988281,
+      "learning_rate": 1.076923076923077e-06,
+      "loss": 1.3574,
+      "step": 42
+    },
+    {
+      "epoch": 0.003314831945729263,
+      "grad_norm": 6.016756534576416,
+      "learning_rate": 1.1025641025641026e-06,
+      "loss": 1.3853,
+      "step": 43
+    },
+    {
+      "epoch": 0.0033919210607462227,
+      "grad_norm": 6.545979976654053,
+      "learning_rate": 1.1282051282051283e-06,
+      "loss": 1.492,
+      "step": 44
+    },
+    {
+      "epoch": 0.003469010175763182,
+      "grad_norm": 5.281697750091553,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.3644,
+      "step": 45
+    },
+    {
+      "epoch": 0.0035460992907801418,
+      "grad_norm": 5.430404186248779,
+      "learning_rate": 1.1794871794871795e-06,
+      "loss": 1.3613,
+      "step": 46
+    },
+    {
+      "epoch": 0.0036231884057971015,
+      "grad_norm": 5.730790138244629,
+      "learning_rate": 1.2051282051282053e-06,
+      "loss": 1.349,
+      "step": 47
+    },
+    {
+      "epoch": 0.0037002775208140612,
+      "grad_norm": 7.0034098625183105,
+      "learning_rate": 1.230769230769231e-06,
+      "loss": 1.4167,
+      "step": 48
+    },
+    {
+      "epoch": 0.0037773666358310205,
+      "grad_norm": 5.198720932006836,
+      "learning_rate": 1.2564102564102565e-06,
+      "loss": 1.2908,
+      "step": 49
+    },
+    {
+      "epoch": 0.0038544557508479803,
+      "grad_norm": 5.8838605880737305,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 1.4535,
+      "step": 50
+    },
+    {
+      "epoch": 0.00393154486586494,
+      "grad_norm": 5.737644672393799,
+      "learning_rate": 1.307692307692308e-06,
+      "loss": 1.2674,
+      "step": 51
+    },
+    {
+      "epoch": 0.004008633980881899,
+      "grad_norm": 5.439259052276611,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.327,
+      "step": 52
+    },
+    {
+      "epoch": 0.0040857230958988595,
+      "grad_norm": 5.283633708953857,
+      "learning_rate": 1.358974358974359e-06,
+      "loss": 1.2557,
+      "step": 53
+    },
+    {
+      "epoch": 0.004162812210915819,
+      "grad_norm": 5.148982524871826,
+      "learning_rate": 1.3846153846153848e-06,
+      "loss": 1.3224,
+      "step": 54
+    },
+    {
+      "epoch": 0.004239901325932778,
+      "grad_norm": 5.781235694885254,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 1.3519,
+      "step": 55
+    },
+    {
+      "epoch": 0.004316990440949738,
+      "grad_norm": 5.436124801635742,
+      "learning_rate": 1.4358974358974359e-06,
+      "loss": 1.2495,
+      "step": 56
+    },
+    {
+      "epoch": 0.0043940795559666975,
+      "grad_norm": 5.363502025604248,
+      "learning_rate": 1.4615384615384618e-06,
+      "loss": 1.304,
+      "step": 57
+    },
+    {
+      "epoch": 0.004471168670983657,
+      "grad_norm": 5.214340686798096,
+      "learning_rate": 1.4871794871794873e-06,
+      "loss": 1.1908,
+      "step": 58
+    },
+    {
+      "epoch": 0.004548257786000617,
+      "grad_norm": 5.095510959625244,
+      "learning_rate": 1.5128205128205128e-06,
+      "loss": 1.3102,
+      "step": 59
+    },
+    {
+      "epoch": 0.004625346901017576,
+      "grad_norm": 4.8631157875061035,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 1.3529,
+      "step": 60
+    },
+    {
+      "epoch": 0.004702436016034536,
+      "grad_norm": 4.890917778015137,
+      "learning_rate": 1.5641025641025642e-06,
+      "loss": 1.1276,
+      "step": 61
+    },
+    {
+      "epoch": 0.004779525131051496,
+      "grad_norm": 5.280998706817627,
+      "learning_rate": 1.5897435897435897e-06,
+      "loss": 1.2194,
+      "step": 62
+    },
+    {
+      "epoch": 0.004856614246068455,
+      "grad_norm": 5.293563365936279,
+      "learning_rate": 1.6153846153846157e-06,
+      "loss": 1.319,
+      "step": 63
+    },
+    {
+      "epoch": 0.004933703361085414,
+      "grad_norm": 5.016965866088867,
+      "learning_rate": 1.6410256410256412e-06,
+      "loss": 1.3868,
+      "step": 64
+    },
+    {
+      "epoch": 0.0050107924761023746,
+      "grad_norm": 4.839664936065674,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.2705,
+      "step": 65
+    },
+    {
+      "epoch": 0.005087881591119334,
+      "grad_norm": 4.863608360290527,
+      "learning_rate": 1.6923076923076926e-06,
+      "loss": 1.2731,
+      "step": 66
+    },
+    {
+      "epoch": 0.005164970706136293,
+      "grad_norm": 5.9279656410217285,
+      "learning_rate": 1.717948717948718e-06,
+      "loss": 1.324,
+      "step": 67
+    },
+    {
+      "epoch": 0.005242059821153253,
+      "grad_norm": 4.750378608703613,
+      "learning_rate": 1.7435897435897436e-06,
+      "loss": 1.2538,
+      "step": 68
+    },
+    {
+      "epoch": 0.005319148936170213,
+      "grad_norm": 5.045099258422852,
+      "learning_rate": 1.7692307692307695e-06,
+      "loss": 1.2533,
+      "step": 69
+    },
+    {
+      "epoch": 0.005396238051187172,
+      "grad_norm": 4.8463006019592285,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 1.3231,
+      "step": 70
+    },
+    {
+      "epoch": 0.005473327166204132,
+      "grad_norm": 4.522879600524902,
+      "learning_rate": 1.8205128205128205e-06,
+      "loss": 1.3319,
+      "step": 71
+    },
+    {
+      "epoch": 0.005550416281221091,
+      "grad_norm": 5.500392913818359,
+      "learning_rate": 1.8461538461538465e-06,
+      "loss": 1.2909,
+      "step": 72
+    },
+    {
+      "epoch": 0.005627505396238052,
+      "grad_norm": 4.754296779632568,
+      "learning_rate": 1.871794871794872e-06,
+      "loss": 1.2207,
+      "step": 73
+    },
+    {
+      "epoch": 0.005704594511255011,
+      "grad_norm": 4.711799144744873,
+      "learning_rate": 1.8974358974358975e-06,
+      "loss": 1.2839,
+      "step": 74
+    },
+    {
+      "epoch": 0.00578168362627197,
+      "grad_norm": 5.301671504974365,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 1.3236,
+      "step": 75
+    },
+    {
+      "epoch": 0.00585877274128893,
+      "grad_norm": 4.835858345031738,
+      "learning_rate": 1.948717948717949e-06,
+      "loss": 1.2828,
+      "step": 76
+    },
+    {
+      "epoch": 0.00593586185630589,
+      "grad_norm": 5.167266368865967,
+      "learning_rate": 1.9743589743589744e-06,
+      "loss": 1.3363,
+      "step": 77
+    },
+    {
+      "epoch": 0.006012950971322849,
+      "grad_norm": 5.193398952484131,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.2662,
+      "step": 78
+    },
+    {
+      "epoch": 0.006090040086339809,
+      "grad_norm": 4.928783416748047,
+      "learning_rate": 2.025641025641026e-06,
+      "loss": 1.2551,
+      "step": 79
+    },
+    {
+      "epoch": 0.006167129201356768,
+      "grad_norm": 5.207440376281738,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.4568,
+      "step": 80
+    },
+    {
+      "epoch": 0.006244218316373728,
+      "grad_norm": 5.189006805419922,
+      "learning_rate": 2.0769230769230773e-06,
+      "loss": 1.1864,
+      "step": 81
+    },
+    {
+      "epoch": 0.006321307431390688,
+      "grad_norm": 5.419140338897705,
+      "learning_rate": 2.1025641025641028e-06,
+      "loss": 1.3153,
+      "step": 82
+    },
+    {
+      "epoch": 0.006398396546407647,
+      "grad_norm": 4.961038589477539,
+      "learning_rate": 2.1282051282051283e-06,
+      "loss": 1.2585,
+      "step": 83
+    },
+    {
+      "epoch": 0.0064754856614246065,
+      "grad_norm": 4.714632034301758,
+      "learning_rate": 2.153846153846154e-06,
+      "loss": 1.2694,
+      "step": 84
+    },
+    {
+      "epoch": 0.006552574776441567,
+      "grad_norm": 5.331617832183838,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 1.3765,
+      "step": 85
+    },
+    {
+      "epoch": 0.006629663891458526,
+      "grad_norm": 4.9930644035339355,
+      "learning_rate": 2.2051282051282052e-06,
+      "loss": 1.3456,
+      "step": 86
+    },
+    {
+      "epoch": 0.006706753006475485,
+      "grad_norm": 4.928962230682373,
+      "learning_rate": 2.230769230769231e-06,
+      "loss": 1.2938,
+      "step": 87
+    },
+    {
+      "epoch": 0.0067838421214924454,
+      "grad_norm": 5.059289932250977,
+      "learning_rate": 2.2564102564102566e-06,
+      "loss": 1.2987,
+      "step": 88
+    },
+    {
+      "epoch": 0.006860931236509405,
+      "grad_norm": 4.943025588989258,
+      "learning_rate": 2.282051282051282e-06,
+      "loss": 1.2336,
+      "step": 89
+    },
+    {
+      "epoch": 0.006938020351526364,
+      "grad_norm": 5.614814758300781,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.1704,
+      "step": 90
+    },
+    {
+      "epoch": 0.007015109466543324,
+      "grad_norm": 4.728719234466553,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 1.301,
+      "step": 91
+    },
+    {
+      "epoch": 0.0070921985815602835,
+      "grad_norm": 5.109025478363037,
+      "learning_rate": 2.358974358974359e-06,
+      "loss": 1.3068,
+      "step": 92
+    },
+    {
+      "epoch": 0.007169287696577244,
+      "grad_norm": 5.220505237579346,
+      "learning_rate": 2.384615384615385e-06,
+      "loss": 1.3854,
+      "step": 93
+    },
+    {
+      "epoch": 0.007246376811594203,
+      "grad_norm": 4.759350776672363,
+      "learning_rate": 2.4102564102564105e-06,
+      "loss": 1.2805,
+      "step": 94
+    },
+    {
+      "epoch": 0.007323465926611162,
+      "grad_norm": 5.338301181793213,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.2815,
+      "step": 95
+    },
+    {
+      "epoch": 0.0074005550416281225,
+      "grad_norm": 5.214847087860107,
+      "learning_rate": 2.461538461538462e-06,
+      "loss": 1.1316,
+      "step": 96
+    },
+    {
+      "epoch": 0.007477644156645082,
+      "grad_norm": 4.639960289001465,
+      "learning_rate": 2.4871794871794875e-06,
+      "loss": 1.2396,
+      "step": 97
+    },
+    {
+      "epoch": 0.007554733271662041,
+      "grad_norm": 4.69920015335083,
+      "learning_rate": 2.512820512820513e-06,
+      "loss": 1.2721,
+      "step": 98
+    },
+    {
+      "epoch": 0.007631822386679001,
+      "grad_norm": 4.799991607666016,
+      "learning_rate": 2.5384615384615385e-06,
+      "loss": 1.3245,
+      "step": 99
+    },
+    {
+      "epoch": 0.0077089115016959605,
+      "grad_norm": 4.986448764801025,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 1.2549,
+      "step": 100
+    },
+    {
+      "epoch": 0.00778600061671292,
+      "grad_norm": 4.512473106384277,
+      "learning_rate": 2.5897435897435903e-06,
+      "loss": 1.2412,
+      "step": 101
+    },
+    {
+      "epoch": 0.00786308973172988,
+      "grad_norm": 5.038204669952393,
+      "learning_rate": 2.615384615384616e-06,
+      "loss": 1.3146,
+      "step": 102
+    },
+    {
+      "epoch": 0.00794017884674684,
+      "grad_norm": 5.079225540161133,
+      "learning_rate": 2.6410256410256413e-06,
+      "loss": 1.3502,
+      "step": 103
+    },
+    {
+      "epoch": 0.008017267961763799,
+      "grad_norm": 4.596055507659912,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.2562,
+      "step": 104
+    },
+    {
+      "epoch": 0.008094357076780759,
+      "grad_norm": 4.764355659484863,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 1.2504,
+      "step": 105
+    },
+    {
+      "epoch": 0.008171446191797719,
+      "grad_norm": 4.672957897186279,
+      "learning_rate": 2.717948717948718e-06,
+      "loss": 1.3247,
+      "step": 106
+    },
+    {
+      "epoch": 0.008248535306814677,
+      "grad_norm": 5.352134704589844,
+      "learning_rate": 2.743589743589744e-06,
+      "loss": 1.2011,
+      "step": 107
+    },
+    {
+      "epoch": 0.008325624421831638,
+      "grad_norm": 5.508421897888184,
+      "learning_rate": 2.7692307692307697e-06,
+      "loss": 1.275,
+      "step": 108
+    },
+    {
+      "epoch": 0.008402713536848598,
+      "grad_norm": 4.892576694488525,
+      "learning_rate": 2.794871794871795e-06,
+      "loss": 1.2882,
+      "step": 109
+    },
+    {
+      "epoch": 0.008479802651865556,
+      "grad_norm": 5.162463665008545,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 1.2715,
+      "step": 110
+    },
+    {
+      "epoch": 0.008556891766882516,
+      "grad_norm": 4.8386549949646,
+      "learning_rate": 2.846153846153846e-06,
+      "loss": 1.2729,
+      "step": 111
+    },
+    {
+      "epoch": 0.008633980881899476,
+      "grad_norm": 4.82133150100708,
+      "learning_rate": 2.8717948717948717e-06,
+      "loss": 1.2241,
+      "step": 112
+    },
+    {
+      "epoch": 0.008711069996916435,
+      "grad_norm": 4.835055828094482,
+      "learning_rate": 2.897435897435898e-06,
+      "loss": 1.1711,
+      "step": 113
+    },
+    {
+      "epoch": 0.008788159111933395,
+      "grad_norm": 5.417538642883301,
+      "learning_rate": 2.9230769230769236e-06,
+      "loss": 1.1772,
+      "step": 114
+    },
+    {
+      "epoch": 0.008865248226950355,
+      "grad_norm": 5.100035190582275,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 1.2341,
+      "step": 115
+    },
+    {
+      "epoch": 0.008942337341967314,
+      "grad_norm": 4.800838470458984,
+      "learning_rate": 2.9743589743589746e-06,
+      "loss": 1.2045,
+      "step": 116
+    },
+    {
+      "epoch": 0.009019426456984274,
+      "grad_norm": 4.5708184242248535,
+      "learning_rate": 3e-06,
+      "loss": 1.1864,
+      "step": 117
+    },
+    {
+      "epoch": 0.009096515572001234,
+      "grad_norm": 5.127934455871582,
+      "learning_rate": 3.0256410256410256e-06,
+      "loss": 1.293,
+      "step": 118
+    },
+    {
+      "epoch": 0.009173604687018192,
+      "grad_norm": 4.894773960113525,
+      "learning_rate": 3.051282051282052e-06,
+      "loss": 1.2704,
+      "step": 119
+    },
+    {
+      "epoch": 0.009250693802035153,
+      "grad_norm": 5.011133193969727,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.3439,
+      "step": 120
+    },
+    {
+      "epoch": 0.009327782917052113,
+      "grad_norm": 4.952587127685547,
+      "learning_rate": 3.102564102564103e-06,
+      "loss": 1.2353,
+      "step": 121
+    },
+    {
+      "epoch": 0.009404872032069071,
+      "grad_norm": 4.612563610076904,
+      "learning_rate": 3.1282051282051284e-06,
+      "loss": 1.2477,
+      "step": 122
+    },
+    {
+      "epoch": 0.009481961147086031,
+      "grad_norm": 4.67457389831543,
+      "learning_rate": 3.153846153846154e-06,
+      "loss": 1.3014,
+      "step": 123
+    },
+    {
+      "epoch": 0.009559050262102992,
+      "grad_norm": 4.838489532470703,
+      "learning_rate": 3.1794871794871795e-06,
+      "loss": 1.2135,
+      "step": 124
+    },
+    {
+      "epoch": 0.00963613937711995,
+      "grad_norm": 5.4575042724609375,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 1.3339,
+      "step": 125
+    },
+    {
+      "epoch": 0.00971322849213691,
+      "grad_norm": 4.730915069580078,
+      "learning_rate": 3.2307692307692313e-06,
+      "loss": 1.2298,
+      "step": 126
+    },
+    {
+      "epoch": 0.00979031760715387,
+      "grad_norm": 5.235102653503418,
+      "learning_rate": 3.256410256410257e-06,
+      "loss": 1.195,
+      "step": 127
+    },
+    {
+      "epoch": 0.009867406722170829,
+      "grad_norm": 4.515792369842529,
+      "learning_rate": 3.2820512820512823e-06,
+      "loss": 1.2009,
+      "step": 128
+    },
+    {
+      "epoch": 0.009944495837187789,
+      "grad_norm": 4.882987022399902,
+      "learning_rate": 3.307692307692308e-06,
+      "loss": 1.3124,
+      "step": 129
+    },
+    {
+      "epoch": 0.010021584952204749,
+      "grad_norm": 4.703249931335449,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.2973,
+      "step": 130
+    },
+    {
+      "epoch": 0.010098674067221708,
+      "grad_norm": 5.116337776184082,
+      "learning_rate": 3.358974358974359e-06,
+      "loss": 1.3368,
+      "step": 131
+    },
+    {
+      "epoch": 0.010175763182238668,
+      "grad_norm": 4.505654335021973,
+      "learning_rate": 3.384615384615385e-06,
+      "loss": 1.12,
+      "step": 132
+    },
+    {
+      "epoch": 0.010252852297255628,
+      "grad_norm": 4.808957576751709,
+      "learning_rate": 3.4102564102564107e-06,
+      "loss": 1.3359,
+      "step": 133
+    },
+    {
+      "epoch": 0.010329941412272586,
+      "grad_norm": 5.045513153076172,
+      "learning_rate": 3.435897435897436e-06,
+      "loss": 1.2173,
+      "step": 134
+    },
+    {
+      "epoch": 0.010407030527289547,
+      "grad_norm": 4.800675868988037,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 1.3637,
+      "step": 135
+    },
+    {
+      "epoch": 0.010484119642306507,
+      "grad_norm": 4.922788619995117,
+      "learning_rate": 3.487179487179487e-06,
+      "loss": 1.2737,
+      "step": 136
+    },
+    {
+      "epoch": 0.010561208757323465,
+      "grad_norm": 5.645283222198486,
+      "learning_rate": 3.5128205128205127e-06,
+      "loss": 1.3652,
+      "step": 137
+    },
+    {
+      "epoch": 0.010638297872340425,
+      "grad_norm": 5.270390510559082,
+      "learning_rate": 3.538461538461539e-06,
+      "loss": 1.2676,
+      "step": 138
+    },
+    {
+      "epoch": 0.010715386987357385,
+      "grad_norm": 5.082615375518799,
+      "learning_rate": 3.5641025641025646e-06,
+      "loss": 1.2817,
+      "step": 139
+    },
+    {
+      "epoch": 0.010792476102374344,
+      "grad_norm": 4.738155841827393,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 1.2964,
+      "step": 140
+    },
+    {
+      "epoch": 0.010869565217391304,
+      "grad_norm": 4.902276992797852,
+      "learning_rate": 3.6153846153846156e-06,
+      "loss": 1.143,
+      "step": 141
+    },
+    {
+      "epoch": 0.010946654332408264,
+      "grad_norm": 4.572447776794434,
+      "learning_rate": 3.641025641025641e-06,
+      "loss": 1.2367,
+      "step": 142
+    },
+    {
+      "epoch": 0.011023743447425224,
+      "grad_norm": 5.859162330627441,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 1.2027,
+      "step": 143
+    },
+    {
+      "epoch": 0.011100832562442183,
+      "grad_norm": 5.429625034332275,
+      "learning_rate": 3.692307692307693e-06,
+      "loss": 1.3681,
+      "step": 144
+    },
+    {
+      "epoch": 0.011177921677459143,
+      "grad_norm": 4.913431167602539,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.1966,
+      "step": 145
+    },
+    {
+      "epoch": 0.011255010792476103,
+      "grad_norm": 5.548055648803711,
+      "learning_rate": 3.743589743589744e-06,
+      "loss": 1.1941,
+      "step": 146
+    },
+    {
+      "epoch": 0.011332099907493062,
+      "grad_norm": 4.827651500701904,
+      "learning_rate": 3.7692307692307694e-06,
+      "loss": 1.27,
+      "step": 147
+    },
+    {
+      "epoch": 0.011409189022510022,
+      "grad_norm": 4.984038829803467,
+      "learning_rate": 3.794871794871795e-06,
+      "loss": 1.2527,
+      "step": 148
+    },
+    {
+      "epoch": 0.011486278137526982,
+      "grad_norm": 4.7654032707214355,
+      "learning_rate": 3.8205128205128204e-06,
+      "loss": 1.1242,
+      "step": 149
+    },
+    {
+      "epoch": 0.01156336725254394,
+      "grad_norm": 4.45728874206543,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 1.2219,
+      "step": 150
+    },
+    {
+      "epoch": 0.0116404563675609,
+      "grad_norm": 4.937605381011963,
+      "learning_rate": 3.871794871794872e-06,
+      "loss": 1.2036,
+      "step": 151
+    },
+    {
+      "epoch": 0.01171754548257786,
+      "grad_norm": 5.271725654602051,
+      "learning_rate": 3.897435897435898e-06,
+      "loss": 1.3552,
+      "step": 152
+    },
+    {
+      "epoch": 0.01179463459759482,
+      "grad_norm": 4.775473117828369,
+      "learning_rate": 3.923076923076923e-06,
+      "loss": 1.1872,
+      "step": 153
+    },
+    {
+      "epoch": 0.01187172371261178,
+      "grad_norm": 5.3986358642578125,
+      "learning_rate": 3.948717948717949e-06,
+      "loss": 1.3389,
+      "step": 154
+    },
+    {
+      "epoch": 0.01194881282762874,
+      "grad_norm": 4.677184581756592,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 1.2231,
+      "step": 155
+    },
+    {
+      "epoch": 0.012025901942645698,
+      "grad_norm": 5.348373889923096,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.2328,
+      "step": 156
+    },
+    {
+      "epoch": 0.012102991057662658,
+      "grad_norm": 4.47370719909668,
+      "learning_rate": 4.025641025641026e-06,
+      "loss": 1.215,
+      "step": 157
+    },
+    {
+      "epoch": 0.012180080172679618,
+      "grad_norm": 5.143750190734863,
+      "learning_rate": 4.051282051282052e-06,
+      "loss": 1.2161,
+      "step": 158
+    },
+    {
+      "epoch": 0.012257169287696577,
+      "grad_norm": 5.946848392486572,
+      "learning_rate": 4.076923076923077e-06,
+      "loss": 1.1238,
+      "step": 159
+    },
+    {
+      "epoch": 0.012334258402713537,
+      "grad_norm": 4.650530815124512,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.2754,
+      "step": 160
+    },
+    {
+      "epoch": 0.012411347517730497,
+      "grad_norm": 4.534331798553467,
+      "learning_rate": 4.128205128205128e-06,
+      "loss": 1.2325,
+      "step": 161
+    },
+    {
+      "epoch": 0.012488436632747455,
+      "grad_norm": 5.198877334594727,
+      "learning_rate": 4.1538461538461545e-06,
+      "loss": 1.2318,
+      "step": 162
+    },
+    {
+      "epoch": 0.012565525747764416,
+      "grad_norm": 4.763823986053467,
+      "learning_rate": 4.17948717948718e-06,
+      "loss": 1.2049,
+      "step": 163
+    },
+    {
+      "epoch": 0.012642614862781376,
+      "grad_norm": 5.088624000549316,
+      "learning_rate": 4.2051282051282055e-06,
+      "loss": 1.2649,
+      "step": 164
+    },
+    {
+      "epoch": 0.012719703977798334,
+      "grad_norm": 5.234319686889648,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 1.308,
+      "step": 165
+    },
+    {
+      "epoch": 0.012796793092815294,
+      "grad_norm": 4.7018537521362305,
+      "learning_rate": 4.2564102564102566e-06,
+      "loss": 1.2356,
+      "step": 166
+    },
+    {
+      "epoch": 0.012873882207832255,
+      "grad_norm": 5.618417263031006,
+      "learning_rate": 4.282051282051282e-06,
+      "loss": 1.1963,
+      "step": 167
+    },
+    {
+      "epoch": 0.012950971322849213,
+      "grad_norm": 5.07379674911499,
+      "learning_rate": 4.307692307692308e-06,
+      "loss": 1.2074,
+      "step": 168
+    },
+    {
+      "epoch": 0.013028060437866173,
+      "grad_norm": 4.746245861053467,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 1.302,
+      "step": 169
+    },
+    {
+      "epoch": 0.013105149552883133,
+      "grad_norm": 4.660124778747559,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 1.2835,
+      "step": 170
+    },
+    {
+      "epoch": 0.013182238667900092,
+      "grad_norm": 4.880448818206787,
+      "learning_rate": 4.384615384615385e-06,
+      "loss": 1.3333,
+      "step": 171
+    },
+    {
+      "epoch": 0.013259327782917052,
+      "grad_norm": 5.0500359535217285,
+      "learning_rate": 4.4102564102564104e-06,
+      "loss": 1.2318,
+      "step": 172
+    },
+    {
+      "epoch": 0.013336416897934012,
+      "grad_norm": 4.640452861785889,
+      "learning_rate": 4.435897435897436e-06,
+      "loss": 1.1425,
+      "step": 173
+    },
+    {
+      "epoch": 0.01341350601295097,
+      "grad_norm": 4.400023460388184,
+      "learning_rate": 4.461538461538462e-06,
+      "loss": 1.319,
+      "step": 174
+    },
+    {
+      "epoch": 0.01349059512796793,
+      "grad_norm": 4.9543046951293945,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 1.1423,
+      "step": 175
+    },
+    {
+      "epoch": 0.013567684242984891,
+      "grad_norm": 5.030699729919434,
+      "learning_rate": 4.512820512820513e-06,
+      "loss": 1.2126,
+      "step": 176
+    },
+    {
+      "epoch": 0.01364477335800185,
+      "grad_norm": 5.696242809295654,
+      "learning_rate": 4.538461538461539e-06,
+      "loss": 1.4041,
+      "step": 177
+    },
+    {
+      "epoch": 0.01372186247301881,
+      "grad_norm": 5.10263729095459,
+      "learning_rate": 4.564102564102564e-06,
+      "loss": 1.2331,
+      "step": 178
+    },
+    {
+      "epoch": 0.01379895158803577,
+      "grad_norm": 4.404088497161865,
+      "learning_rate": 4.58974358974359e-06,
+      "loss": 1.2334,
+      "step": 179
+    },
+    {
+      "epoch": 0.013876040703052728,
+      "grad_norm": 4.599002361297607,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 1.0995,
+      "step": 180
+    },
+    {
+      "epoch": 0.013953129818069688,
+      "grad_norm": 5.312249660491943,
+      "learning_rate": 4.641025641025642e-06,
+      "loss": 1.3024,
+      "step": 181
+    },
+    {
+      "epoch": 0.014030218933086648,
+      "grad_norm": 4.873595714569092,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.3503,
+      "step": 182
+    },
+    {
+      "epoch": 0.014107308048103609,
+      "grad_norm": 4.9252400398254395,
+      "learning_rate": 4.692307692307693e-06,
+      "loss": 1.2383,
+      "step": 183
+    },
+    {
+      "epoch": 0.014184397163120567,
+      "grad_norm": 5.004896640777588,
+      "learning_rate": 4.717948717948718e-06,
+      "loss": 1.1963,
+      "step": 184
+    },
+    {
+      "epoch": 0.014261486278137527,
+      "grad_norm": 5.2026801109313965,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 1.4058,
+      "step": 185
+    },
+    {
+      "epoch": 0.014338575393154487,
+      "grad_norm": 4.740382194519043,
+      "learning_rate": 4.76923076923077e-06,
+      "loss": 1.2259,
+      "step": 186
+    },
+    {
+      "epoch": 0.014415664508171446,
+      "grad_norm": 4.338659763336182,
+      "learning_rate": 4.7948717948717955e-06,
+      "loss": 1.0815,
+      "step": 187
+    },
+    {
+      "epoch": 0.014492753623188406,
+      "grad_norm": 5.065215110778809,
+      "learning_rate": 4.820512820512821e-06,
+      "loss": 1.2062,
+      "step": 188
+    },
+    {
+      "epoch": 0.014569842738205366,
+      "grad_norm": 4.5627264976501465,
+      "learning_rate": 4.8461538461538465e-06,
+      "loss": 1.2032,
+      "step": 189
+    },
+    {
+      "epoch": 0.014646931853222325,
+      "grad_norm": 5.128222465515137,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 1.3514,
+      "step": 190
+    },
+    {
+      "epoch": 0.014724020968239285,
+      "grad_norm": 4.774002552032471,
+      "learning_rate": 4.8974358974358975e-06,
+      "loss": 1.2028,
+      "step": 191
+    },
+    {
+      "epoch": 0.014801110083256245,
+      "grad_norm": 5.117179870605469,
+      "learning_rate": 4.923076923076924e-06,
+      "loss": 1.19,
+      "step": 192
+    },
+    {
+      "epoch": 0.014878199198273203,
+      "grad_norm": 4.972648620605469,
+      "learning_rate": 4.948717948717949e-06,
+      "loss": 1.2141,
+      "step": 193
+    },
+    {
+      "epoch": 0.014955288313290164,
+      "grad_norm": 4.6805219650268555,
+      "learning_rate": 4.974358974358975e-06,
+      "loss": 1.1642,
+      "step": 194
+    },
+    {
+      "epoch": 0.015032377428307124,
+      "grad_norm": 4.577054023742676,
+      "learning_rate": 5e-06,
+      "loss": 1.28,
+      "step": 195
+    },
+    {
+      "epoch": 0.015109466543324082,
+      "grad_norm": 4.717634677886963,
+      "learning_rate": 5.025641025641026e-06,
+      "loss": 1.2222,
+      "step": 196
+    },
+    {
+      "epoch": 0.015186555658341042,
+      "grad_norm": 5.113860607147217,
+      "learning_rate": 5.051282051282051e-06,
+      "loss": 1.3246,
+      "step": 197
+    },
+    {
+      "epoch": 0.015263644773358002,
+      "grad_norm": 5.178426742553711,
+      "learning_rate": 5.076923076923077e-06,
+      "loss": 1.2102,
+      "step": 198
+    },
+    {
+      "epoch": 0.015340733888374961,
+      "grad_norm": 5.071588039398193,
+      "learning_rate": 5.1025641025641024e-06,
+      "loss": 1.2794,
+      "step": 199
+    },
+    {
+      "epoch": 0.015417823003391921,
+      "grad_norm": 5.163846015930176,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 1.2664,
+      "step": 200
+    },
+    {
+      "epoch": 0.015494912118408881,
+      "grad_norm": 4.933408260345459,
+      "learning_rate": 5.1538461538461534e-06,
+      "loss": 1.3126,
+      "step": 201
+    },
+    {
+      "epoch": 0.01557200123342584,
+      "grad_norm": 5.551726341247559,
+      "learning_rate": 5.179487179487181e-06,
+      "loss": 1.4489,
+      "step": 202
+    },
+    {
+      "epoch": 0.0156490903484428,
+      "grad_norm": 4.847557067871094,
+      "learning_rate": 5.205128205128206e-06,
+      "loss": 1.2887,
+      "step": 203
+    },
+    {
+      "epoch": 0.01572617946345976,
+      "grad_norm": 5.323663234710693,
+      "learning_rate": 5.230769230769232e-06,
+      "loss": 1.271,
+      "step": 204
+    },
+    {
+      "epoch": 0.01580326857847672,
+      "grad_norm": 4.9948201179504395,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 1.1802,
+      "step": 205
+    },
+    {
+      "epoch": 0.01588035769349368,
+      "grad_norm": 5.172011375427246,
+      "learning_rate": 5.282051282051283e-06,
+      "loss": 1.1817,
+      "step": 206
+    },
+    {
+      "epoch": 0.015957446808510637,
+      "grad_norm": 5.043161869049072,
+      "learning_rate": 5.307692307692308e-06,
+      "loss": 1.2308,
+      "step": 207
+    },
+    {
+      "epoch": 0.016034535923527597,
+      "grad_norm": 4.964840888977051,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.314,
+      "step": 208
+    },
+    {
+      "epoch": 0.016111625038544557,
+      "grad_norm": 5.201618194580078,
+      "learning_rate": 5.358974358974359e-06,
+      "loss": 1.2482,
+      "step": 209
+    },
+    {
+      "epoch": 0.016188714153561518,
+      "grad_norm": 5.381608963012695,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 1.2502,
+      "step": 210
+    },
+    {
+      "epoch": 0.016265803268578478,
+      "grad_norm": 5.2073187828063965,
+      "learning_rate": 5.41025641025641e-06,
+      "loss": 1.3514,
+      "step": 211
+    },
+    {
+      "epoch": 0.016342892383595438,
+      "grad_norm": 5.278835773468018,
+      "learning_rate": 5.435897435897436e-06,
+      "loss": 1.313,
+      "step": 212
+    },
+    {
+      "epoch": 0.016419981498612395,
+      "grad_norm": 5.034635543823242,
+      "learning_rate": 5.461538461538461e-06,
+      "loss": 1.1963,
+      "step": 213
+    },
+    {
+      "epoch": 0.016497070613629355,
+      "grad_norm": 4.991930961608887,
+      "learning_rate": 5.487179487179488e-06,
+      "loss": 1.3247,
+      "step": 214
+    },
+    {
+      "epoch": 0.016574159728646315,
+      "grad_norm": 5.150703430175781,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 1.2084,
+      "step": 215
+    },
+    {
+      "epoch": 0.016651248843663275,
+      "grad_norm": 4.838485240936279,
+      "learning_rate": 5.538461538461539e-06,
+      "loss": 1.248,
+      "step": 216
+    },
+    {
+      "epoch": 0.016728337958680235,
+      "grad_norm": 4.620452404022217,
+      "learning_rate": 5.564102564102565e-06,
+      "loss": 1.1368,
+      "step": 217
+    },
+    {
+      "epoch": 0.016805427073697195,
+      "grad_norm": 4.712934494018555,
+      "learning_rate": 5.58974358974359e-06,
+      "loss": 1.2305,
+      "step": 218
+    },
+    {
+      "epoch": 0.016882516188714152,
+      "grad_norm": 4.66661262512207,
+      "learning_rate": 5.615384615384616e-06,
+      "loss": 1.1054,
+      "step": 219
+    },
+    {
+      "epoch": 0.016959605303731112,
+      "grad_norm": 5.062489986419678,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 1.2248,
+      "step": 220
+    },
+    {
+      "epoch": 0.017036694418748072,
+      "grad_norm": 5.344682216644287,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 1.3622,
+      "step": 221
+    },
+    {
+      "epoch": 0.017113783533765033,
+      "grad_norm": 5.120663166046143,
+      "learning_rate": 5.692307692307692e-06,
+      "loss": 1.1208,
+      "step": 222
+    },
+    {
+      "epoch": 0.017190872648781993,
+      "grad_norm": 4.629307746887207,
+      "learning_rate": 5.717948717948718e-06,
+      "loss": 1.204,
+      "step": 223
+    },
+    {
+      "epoch": 0.017267961763798953,
+      "grad_norm": 5.405148029327393,
+      "learning_rate": 5.743589743589743e-06,
+      "loss": 1.2831,
+      "step": 224
+    },
+    {
+      "epoch": 0.01734505087881591,
+      "grad_norm": 4.422632217407227,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 1.2251,
+      "step": 225
+    },
+    {
+      "epoch": 0.01742213999383287,
+      "grad_norm": 4.856597900390625,
+      "learning_rate": 5.794871794871796e-06,
+      "loss": 1.3051,
+      "step": 226
+    },
+    {
+      "epoch": 0.01749922910884983,
+      "grad_norm": 4.129710674285889,
+      "learning_rate": 5.820512820512822e-06,
+      "loss": 1.0233,
+      "step": 227
+    },
+    {
+      "epoch": 0.01757631822386679,
+      "grad_norm": 4.654583930969238,
+      "learning_rate": 5.846153846153847e-06,
+      "loss": 1.2445,
+      "step": 228
+    },
+    {
+      "epoch": 0.01765340733888375,
+      "grad_norm": 4.510683536529541,
+      "learning_rate": 5.871794871794873e-06,
+      "loss": 1.1158,
+      "step": 229
+    },
+    {
+      "epoch": 0.01773049645390071,
+      "grad_norm": 4.56011438369751,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 1.2008,
+      "step": 230
+    },
+    {
+      "epoch": 0.017807585568917667,
+      "grad_norm": 4.707639694213867,
+      "learning_rate": 5.923076923076924e-06,
+      "loss": 1.2206,
+      "step": 231
+    },
+    {
+      "epoch": 0.017884674683934627,
+      "grad_norm": 4.402471542358398,
+      "learning_rate": 5.948717948717949e-06,
+      "loss": 1.193,
+      "step": 232
+    },
+    {
+      "epoch": 0.017961763798951588,
+      "grad_norm": 4.763029098510742,
+      "learning_rate": 5.974358974358975e-06,
+      "loss": 1.2553,
+      "step": 233
+    },
+    {
+      "epoch": 0.018038852913968548,
+      "grad_norm": 4.92691707611084,
+      "learning_rate": 6e-06,
+      "loss": 1.2258,
+      "step": 234
+    },
+    {
+      "epoch": 0.018115942028985508,
+      "grad_norm": 4.550500392913818,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 1.2513,
+      "step": 235
+    },
+    {
+      "epoch": 0.018193031144002468,
+      "grad_norm": 4.7478251457214355,
+      "learning_rate": 6.051282051282051e-06,
+      "loss": 1.239,
+      "step": 236
+    },
+    {
+      "epoch": 0.018270120259019425,
+      "grad_norm": 4.82733678817749,
+      "learning_rate": 6.076923076923077e-06,
+      "loss": 1.2731,
+      "step": 237
+    },
+    {
+      "epoch": 0.018347209374036385,
+      "grad_norm": 4.546608924865723,
+      "learning_rate": 6.102564102564104e-06,
+      "loss": 1.1815,
+      "step": 238
+    },
+    {
+      "epoch": 0.018424298489053345,
+      "grad_norm": 5.261005878448486,
+      "learning_rate": 6.128205128205129e-06,
+      "loss": 1.2074,
+      "step": 239
+    },
+    {
+      "epoch": 0.018501387604070305,
+      "grad_norm": 4.9964213371276855,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 1.3077,
+      "step": 240
+    },
+    {
+      "epoch": 0.018578476719087265,
+      "grad_norm": 4.793504238128662,
+      "learning_rate": 6.17948717948718e-06,
+      "loss": 1.1757,
+      "step": 241
+    },
+    {
+      "epoch": 0.018655565834104226,
+      "grad_norm": 5.3314642906188965,
+      "learning_rate": 6.205128205128206e-06,
+      "loss": 1.3881,
+      "step": 242
+    },
+    {
+      "epoch": 0.018732654949121186,
+      "grad_norm": 4.648246765136719,
+      "learning_rate": 6.230769230769231e-06,
+      "loss": 1.2445,
+      "step": 243
+    },
+    {
+      "epoch": 0.018809744064138142,
+      "grad_norm": 4.99802827835083,
+      "learning_rate": 6.256410256410257e-06,
+      "loss": 1.27,
+      "step": 244
+    },
+    {
+      "epoch": 0.018886833179155103,
+      "grad_norm": 4.553841590881348,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.1916,
+      "step": 245
+    },
+    {
+      "epoch": 0.018963922294172063,
+      "grad_norm": 4.673896312713623,
+      "learning_rate": 6.307692307692308e-06,
+      "loss": 1.2046,
+      "step": 246
+    },
+    {
+      "epoch": 0.019041011409189023,
+      "grad_norm": 5.352630615234375,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 1.3785,
+      "step": 247
+    },
+    {
+      "epoch": 0.019118100524205983,
+      "grad_norm": 4.59555721282959,
+      "learning_rate": 6.358974358974359e-06,
+      "loss": 1.0682,
+      "step": 248
+    },
+    {
+      "epoch": 0.019195189639222943,
+      "grad_norm": 5.211584568023682,
+      "learning_rate": 6.384615384615384e-06,
+      "loss": 1.3154,
+      "step": 249
+    },
+    {
+      "epoch": 0.0192722787542399,
+      "grad_norm": 4.7507243156433105,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 1.1679,
+      "step": 250
+    },
+    {
+      "epoch": 0.01934936786925686,
+      "grad_norm": 4.9127516746521,
+      "learning_rate": 6.435897435897437e-06,
+      "loss": 1.1803,
+      "step": 251
+    },
+    {
+      "epoch": 0.01942645698427382,
+      "grad_norm": 5.001187801361084,
+      "learning_rate": 6.461538461538463e-06,
+      "loss": 1.2204,
+      "step": 252
+    },
+    {
+      "epoch": 0.01950354609929078,
+      "grad_norm": 5.021228313446045,
+      "learning_rate": 6.487179487179488e-06,
+      "loss": 1.2126,
+      "step": 253
+    },
+    {
+      "epoch": 0.01958063521430774,
+      "grad_norm": 4.73190975189209,
+      "learning_rate": 6.512820512820514e-06,
+      "loss": 1.1656,
+      "step": 254
+    },
+    {
+      "epoch": 0.0196577243293247,
+      "grad_norm": 4.766290187835693,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 1.2573,
+      "step": 255
+    },
+    {
+      "epoch": 0.019734813444341658,
+      "grad_norm": 4.61043643951416,
+      "learning_rate": 6.564102564102565e-06,
+      "loss": 1.2092,
+      "step": 256
+    },
+    {
+      "epoch": 0.019811902559358618,
+      "grad_norm": 4.742149353027344,
+      "learning_rate": 6.58974358974359e-06,
+      "loss": 1.2197,
+      "step": 257
+    },
+    {
+      "epoch": 0.019888991674375578,
+      "grad_norm": 4.5568389892578125,
+      "learning_rate": 6.615384615384616e-06,
+      "loss": 1.2677,
+      "step": 258
+    },
+    {
+      "epoch": 0.019966080789392538,
+      "grad_norm": 4.9256367683410645,
+      "learning_rate": 6.641025641025641e-06,
+      "loss": 1.1879,
+      "step": 259
+    },
+    {
+      "epoch": 0.020043169904409498,
+      "grad_norm": 5.019280433654785,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.1391,
+      "step": 260
+    },
+    {
+      "epoch": 0.02012025901942646,
+      "grad_norm": 5.337982654571533,
+      "learning_rate": 6.692307692307692e-06,
+      "loss": 1.1885,
+      "step": 261
+    },
+    {
+      "epoch": 0.020197348134443415,
+      "grad_norm": 5.246166229248047,
+      "learning_rate": 6.717948717948718e-06,
+      "loss": 1.2816,
+      "step": 262
+    },
+    {
+      "epoch": 0.020274437249460375,
+      "grad_norm": 5.1329731941223145,
+      "learning_rate": 6.743589743589745e-06,
+      "loss": 1.0985,
+      "step": 263
+    },
+    {
+      "epoch": 0.020351526364477335,
+      "grad_norm": 5.237858772277832,
+      "learning_rate": 6.76923076923077e-06,
+      "loss": 1.3882,
+      "step": 264
+    },
+    {
+      "epoch": 0.020428615479494296,
+      "grad_norm": 5.131351470947266,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 1.1944,
+      "step": 265
+    },
+    {
+      "epoch": 0.020505704594511256,
+      "grad_norm": 5.077263832092285,
+      "learning_rate": 6.820512820512821e-06,
+      "loss": 1.2041,
+      "step": 266
+    },
+    {
+      "epoch": 0.020582793709528216,
+      "grad_norm": 5.028573036193848,
+      "learning_rate": 6.846153846153847e-06,
+      "loss": 1.248,
+      "step": 267
+    },
+    {
+      "epoch": 0.020659882824545173,
+      "grad_norm": 5.382236480712891,
+      "learning_rate": 6.871794871794872e-06,
+      "loss": 1.3058,
+      "step": 268
+    },
+    {
+      "epoch": 0.020736971939562133,
+      "grad_norm": 4.82020902633667,
+      "learning_rate": 6.897435897435898e-06,
+      "loss": 1.2581,
+      "step": 269
+    },
+    {
+      "epoch": 0.020814061054579093,
+      "grad_norm": 4.485026836395264,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 1.1932,
+      "step": 270
+    },
+    {
+      "epoch": 0.020891150169596053,
+      "grad_norm": 4.891682147979736,
+      "learning_rate": 6.948717948717949e-06,
+      "loss": 1.2035,
+      "step": 271
+    },
+    {
+      "epoch": 0.020968239284613013,
+      "grad_norm": 4.9292683601379395,
+      "learning_rate": 6.974358974358974e-06,
+      "loss": 1.2209,
+      "step": 272
+    },
+    {
+      "epoch": 0.021045328399629974,
+      "grad_norm": 4.799893379211426,
+      "learning_rate": 7e-06,
+      "loss": 1.3824,
+      "step": 273
+    },
+    {
+      "epoch": 0.02112241751464693,
+      "grad_norm": 4.358574390411377,
+      "learning_rate": 7.025641025641025e-06,
+      "loss": 1.1656,
+      "step": 274
+    },
+    {
+      "epoch": 0.02119950662966389,
+      "grad_norm": 4.752989768981934,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 1.1773,
+      "step": 275
+    },
+    {
+      "epoch": 0.02127659574468085,
+      "grad_norm": 4.714949607849121,
+      "learning_rate": 7.076923076923078e-06,
+      "loss": 1.1622,
+      "step": 276
+    },
+    {
+      "epoch": 0.02135368485969781,
+      "grad_norm": 4.488297939300537,
+      "learning_rate": 7.102564102564104e-06,
+      "loss": 1.1412,
+      "step": 277
+    },
+    {
+      "epoch": 0.02143077397471477,
+      "grad_norm": 4.944022178649902,
+      "learning_rate": 7.128205128205129e-06,
+      "loss": 1.1468,
+      "step": 278
+    },
+    {
+      "epoch": 0.02150786308973173,
+      "grad_norm": 4.799287796020508,
+      "learning_rate": 7.153846153846155e-06,
+      "loss": 1.1958,
+      "step": 279
+    },
+    {
+      "epoch": 0.021584952204748688,
+      "grad_norm": 4.447544574737549,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 1.3105,
+      "step": 280
+    },
+    {
+      "epoch": 0.021662041319765648,
+      "grad_norm": 4.781645774841309,
+      "learning_rate": 7.205128205128206e-06,
+      "loss": 1.3335,
+      "step": 281
+    },
+    {
+      "epoch": 0.021739130434782608,
+      "grad_norm": 5.0389018058776855,
+      "learning_rate": 7.230769230769231e-06,
+      "loss": 1.2225,
+      "step": 282
+    },
+    {
+      "epoch": 0.021816219549799568,
+      "grad_norm": 5.132277011871338,
+      "learning_rate": 7.256410256410257e-06,
+      "loss": 1.2914,
+      "step": 283
+    },
+    {
+      "epoch": 0.02189330866481653,
+      "grad_norm": 5.015186309814453,
+      "learning_rate": 7.282051282051282e-06,
+      "loss": 1.1996,
+      "step": 284
+    },
+    {
+      "epoch": 0.02197039777983349,
+      "grad_norm": 5.795614242553711,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 1.251,
+      "step": 285
+    },
+    {
+      "epoch": 0.02204748689485045,
+      "grad_norm": 4.816180229187012,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 1.2001,
+      "step": 286
+    },
+    {
+      "epoch": 0.022124576009867405,
+      "grad_norm": 4.811607837677002,
+      "learning_rate": 7.35897435897436e-06,
+      "loss": 1.2148,
+      "step": 287
+    },
+    {
+      "epoch": 0.022201665124884366,
+      "grad_norm": 4.706379413604736,
+      "learning_rate": 7.384615384615386e-06,
+      "loss": 1.249,
+      "step": 288
+    },
+    {
+      "epoch": 0.022278754239901326,
+      "grad_norm": 5.7372894287109375,
+      "learning_rate": 7.410256410256411e-06,
+      "loss": 1.3184,
+      "step": 289
+    },
+    {
+      "epoch": 0.022355843354918286,
+      "grad_norm": 4.672527313232422,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 1.1606,
+      "step": 290
+    },
+    {
+      "epoch": 0.022432932469935246,
+      "grad_norm": 4.6642560958862305,
+      "learning_rate": 7.461538461538462e-06,
+      "loss": 1.2362,
+      "step": 291
+    },
+    {
+      "epoch": 0.022510021584952206,
+      "grad_norm": 4.640800952911377,
+      "learning_rate": 7.487179487179488e-06,
+      "loss": 1.237,
+      "step": 292
+    },
+    {
+      "epoch": 0.022587110699969163,
+      "grad_norm": 5.178220272064209,
+      "learning_rate": 7.512820512820513e-06,
+      "loss": 1.1799,
+      "step": 293
+    },
+    {
+      "epoch": 0.022664199814986123,
+      "grad_norm": 4.580568790435791,
+      "learning_rate": 7.538461538461539e-06,
+      "loss": 1.2442,
+      "step": 294
+    },
+    {
+      "epoch": 0.022741288930003083,
+      "grad_norm": 4.631171226501465,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.2399,
+      "step": 295
+    },
+    {
+      "epoch": 0.022818378045020044,
+      "grad_norm": 4.3523101806640625,
+      "learning_rate": 7.58974358974359e-06,
+      "loss": 1.1657,
+      "step": 296
+    },
+    {
+      "epoch": 0.022895467160037004,
+      "grad_norm": 5.684840202331543,
+      "learning_rate": 7.615384615384615e-06,
+      "loss": 1.2356,
+      "step": 297
+    },
+    {
+      "epoch": 0.022972556275053964,
+      "grad_norm": 4.739688873291016,
+      "learning_rate": 7.641025641025641e-06,
+      "loss": 1.2422,
+      "step": 298
+    },
+    {
+      "epoch": 0.02304964539007092,
+      "grad_norm": 4.905622482299805,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 1.3017,
+      "step": 299
+    },
+    {
+      "epoch": 0.02312673450508788,
+      "grad_norm": 4.725374221801758,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 1.3432,
+      "step": 300
+    },
+    {
+      "epoch": 0.02320382362010484,
+      "grad_norm": 4.97802209854126,
+      "learning_rate": 7.717948717948718e-06,
+      "loss": 1.2064,
+      "step": 301
+    },
+    {
+      "epoch": 0.0232809127351218,
+      "grad_norm": 4.46768856048584,
+      "learning_rate": 7.743589743589745e-06,
+      "loss": 1.1462,
+      "step": 302
+    },
+    {
+      "epoch": 0.02335800185013876,
+      "grad_norm": 4.416783809661865,
+      "learning_rate": 7.76923076923077e-06,
+      "loss": 1.2325,
+      "step": 303
+    },
+    {
+      "epoch": 0.02343509096515572,
+      "grad_norm": 5.031586647033691,
+      "learning_rate": 7.794871794871796e-06,
+      "loss": 1.2013,
+      "step": 304
+    },
+    {
+      "epoch": 0.023512180080172678,
+      "grad_norm": 5.045072555541992,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 1.1312,
+      "step": 305
+    },
+    {
+      "epoch": 0.02358926919518964,
+      "grad_norm": 4.986116886138916,
+      "learning_rate": 7.846153846153847e-06,
+      "loss": 1.1599,
+      "step": 306
+    },
+    {
+      "epoch": 0.0236663583102066,
+      "grad_norm": 5.043438911437988,
+      "learning_rate": 7.871794871794873e-06,
+      "loss": 1.2043,
+      "step": 307
+    },
+    {
+      "epoch": 0.02374344742522356,
+      "grad_norm": 4.92976188659668,
+      "learning_rate": 7.897435897435898e-06,
+      "loss": 1.2303,
+      "step": 308
+    },
+    {
+      "epoch": 0.02382053654024052,
+      "grad_norm": 5.960122585296631,
+      "learning_rate": 7.923076923076924e-06,
+      "loss": 1.203,
+      "step": 309
+    },
+    {
+      "epoch": 0.02389762565525748,
+      "grad_norm": 5.050799369812012,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 1.3918,
+      "step": 310
+    },
+    {
+      "epoch": 0.023974714770274436,
+      "grad_norm": 5.031795978546143,
+      "learning_rate": 7.974358974358975e-06,
+      "loss": 1.2533,
+      "step": 311
+    },
+    {
+      "epoch": 0.024051803885291396,
+      "grad_norm": 4.596309185028076,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.1893,
+      "step": 312
+    },
+    {
+      "epoch": 0.024128893000308356,
+      "grad_norm": 4.877758979797363,
+      "learning_rate": 8.025641025641026e-06,
+      "loss": 1.2398,
+      "step": 313
+    },
+    {
+      "epoch": 0.024205982115325316,
+      "grad_norm": 4.588964462280273,
+      "learning_rate": 8.051282051282052e-06,
+      "loss": 1.1656,
+      "step": 314
+    },
+    {
+      "epoch": 0.024283071230342276,
+      "grad_norm": 4.996614456176758,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 1.2567,
+      "step": 315
+    },
+    {
+      "epoch": 0.024360160345359236,
+      "grad_norm": 4.830722332000732,
+      "learning_rate": 8.102564102564103e-06,
+      "loss": 1.2854,
+      "step": 316
+    },
+    {
+      "epoch": 0.024437249460376193,
+      "grad_norm": 5.052509784698486,
+      "learning_rate": 8.12820512820513e-06,
+      "loss": 1.1399,
+      "step": 317
+    },
+    {
+      "epoch": 0.024514338575393153,
+      "grad_norm": 5.15505838394165,
+      "learning_rate": 8.153846153846154e-06,
+      "loss": 1.1843,
+      "step": 318
+    },
+    {
+      "epoch": 0.024591427690410114,
+      "grad_norm": 4.3718671798706055,
+      "learning_rate": 8.17948717948718e-06,
+      "loss": 1.2297,
+      "step": 319
+    },
+    {
+      "epoch": 0.024668516805427074,
+      "grad_norm": 5.402135372161865,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 1.1474,
+      "step": 320
+    },
+    {
+      "epoch": 0.024745605920444034,
+      "grad_norm": 4.940793991088867,
+      "learning_rate": 8.230769230769232e-06,
+      "loss": 1.2823,
+      "step": 321
+    },
+    {
+      "epoch": 0.024822695035460994,
+      "grad_norm": 4.902515888214111,
+      "learning_rate": 8.256410256410256e-06,
+      "loss": 1.3237,
+      "step": 322
+    },
+    {
+      "epoch": 0.024899784150477954,
+      "grad_norm": 4.852400779724121,
+      "learning_rate": 8.282051282051283e-06,
+      "loss": 1.1324,
+      "step": 323
+    },
+    {
+      "epoch": 0.02497687326549491,
+      "grad_norm": 4.758856296539307,
+      "learning_rate": 8.307692307692309e-06,
+      "loss": 1.2108,
+      "step": 324
+    },
+    {
+      "epoch": 0.02505396238051187,
+      "grad_norm": 4.814808368682861,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.2657,
+      "step": 325
+    },
+    {
+      "epoch": 0.02513105149552883,
+      "grad_norm": 4.526566028594971,
+      "learning_rate": 8.35897435897436e-06,
+      "loss": 1.2241,
+      "step": 326
+    },
+    {
+      "epoch": 0.02520814061054579,
+      "grad_norm": 5.531708717346191,
+      "learning_rate": 8.384615384615385e-06,
+      "loss": 1.3275,
+      "step": 327
+    },
+    {
+      "epoch": 0.02528522972556275,
+      "grad_norm": 4.945657253265381,
+      "learning_rate": 8.410256410256411e-06,
+      "loss": 1.274,
+      "step": 328
+    },
+    {
+      "epoch": 0.025362318840579712,
+      "grad_norm": 4.540581703186035,
+      "learning_rate": 8.435897435897436e-06,
+      "loss": 1.266,
+      "step": 329
+    },
+    {
+      "epoch": 0.02543940795559667,
+      "grad_norm": 4.201421737670898,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 1.2355,
+      "step": 330
+    },
+    {
+      "epoch": 0.02551649707061363,
+      "grad_norm": 4.603540420532227,
+      "learning_rate": 8.487179487179488e-06,
+      "loss": 1.2529,
+      "step": 331
+    },
+    {
+      "epoch": 0.02559358618563059,
+      "grad_norm": 4.829327583312988,
+      "learning_rate": 8.512820512820513e-06,
+      "loss": 1.3069,
+      "step": 332
+    },
+    {
+      "epoch": 0.02567067530064755,
+      "grad_norm": 4.978155612945557,
+      "learning_rate": 8.53846153846154e-06,
+      "loss": 1.1573,
+      "step": 333
+    },
+    {
+      "epoch": 0.02574776441566451,
+      "grad_norm": 4.506424427032471,
+      "learning_rate": 8.564102564102564e-06,
+      "loss": 1.1211,
+      "step": 334
+    },
+    {
+      "epoch": 0.02582485353068147,
+      "grad_norm": 5.001166343688965,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 1.1474,
+      "step": 335
+    },
+    {
+      "epoch": 0.025901942645698426,
+      "grad_norm": 4.721814155578613,
+      "learning_rate": 8.615384615384617e-06,
+      "loss": 1.3049,
+      "step": 336
+    },
+    {
+      "epoch": 0.025979031760715386,
+      "grad_norm": 4.9357008934021,
+      "learning_rate": 8.641025641025641e-06,
+      "loss": 1.2877,
+      "step": 337
+    },
+    {
+      "epoch": 0.026056120875732346,
+      "grad_norm": 4.828952789306641,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 1.1093,
+      "step": 338
+    },
+    {
+      "epoch": 0.026133209990749307,
+      "grad_norm": 4.813793659210205,
+      "learning_rate": 8.692307692307692e-06,
+      "loss": 1.2452,
+      "step": 339
+    },
+    {
+      "epoch": 0.026210299105766267,
+      "grad_norm": 4.303406715393066,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 1.0951,
+      "step": 340
+    },
+    {
+      "epoch": 0.026287388220783227,
+      "grad_norm": 4.791975498199463,
+      "learning_rate": 8.743589743589743e-06,
+      "loss": 1.2548,
+      "step": 341
+    },
+    {
+      "epoch": 0.026364477335800184,
+      "grad_norm": 4.471169948577881,
+      "learning_rate": 8.76923076923077e-06,
+      "loss": 1.112,
+      "step": 342
+    },
+    {
+      "epoch": 0.026441566450817144,
+      "grad_norm": 4.681336879730225,
+      "learning_rate": 8.794871794871796e-06,
+      "loss": 1.2183,
+      "step": 343
+    },
+    {
+      "epoch": 0.026518655565834104,
+      "grad_norm": 5.055361270904541,
+      "learning_rate": 8.820512820512821e-06,
+      "loss": 1.1957,
+      "step": 344
+    },
+    {
+      "epoch": 0.026595744680851064,
+      "grad_norm": 4.525293827056885,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 1.1316,
+      "step": 345
+    },
+    {
+      "epoch": 0.026672833795868024,
+      "grad_norm": 4.73473596572876,
+      "learning_rate": 8.871794871794872e-06,
+      "loss": 1.2229,
+      "step": 346
+    },
+    {
+      "epoch": 0.026749922910884984,
+      "grad_norm": 4.338306427001953,
+      "learning_rate": 8.897435897435898e-06,
+      "loss": 1.2271,
+      "step": 347
+    },
+    {
+      "epoch": 0.02682701202590194,
+      "grad_norm": 4.502996444702148,
+      "learning_rate": 8.923076923076925e-06,
+      "loss": 1.2868,
+      "step": 348
+    },
+    {
+      "epoch": 0.0269041011409189,
+      "grad_norm": 4.5346198081970215,
+      "learning_rate": 8.94871794871795e-06,
+      "loss": 1.0935,
+      "step": 349
+    },
+    {
+      "epoch": 0.02698119025593586,
+      "grad_norm": 4.440368175506592,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 1.1267,
+      "step": 350
+    },
+    {
+      "epoch": 0.02705827937095282,
+      "grad_norm": 4.624384880065918,
+      "learning_rate": 9e-06,
+      "loss": 1.215,
+      "step": 351
+    },
+    {
+      "epoch": 0.027135368485969782,
+      "grad_norm": 4.52625846862793,
+      "learning_rate": 9.025641025641027e-06,
+      "loss": 1.2398,
+      "step": 352
+    },
+    {
+      "epoch": 0.027212457600986742,
+      "grad_norm": 4.4209818840026855,
+      "learning_rate": 9.051282051282051e-06,
+      "loss": 1.1526,
+      "step": 353
+    },
+    {
+      "epoch": 0.0272895467160037,
+      "grad_norm": 4.647578239440918,
+      "learning_rate": 9.076923076923078e-06,
+      "loss": 1.2131,
+      "step": 354
+    },
+    {
+      "epoch": 0.02736663583102066,
+      "grad_norm": 3.973903179168701,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 1.0613,
+      "step": 355
+    },
+    {
+      "epoch": 0.02744372494603762,
+      "grad_norm": 5.1517252922058105,
+      "learning_rate": 9.128205128205129e-06,
+      "loss": 1.1746,
+      "step": 356
+    },
+    {
+      "epoch": 0.02752081406105458,
+      "grad_norm": 4.37183952331543,
+      "learning_rate": 9.153846153846155e-06,
+      "loss": 1.2774,
+      "step": 357
+    },
+    {
+      "epoch": 0.02759790317607154,
+      "grad_norm": 4.5695295333862305,
+      "learning_rate": 9.17948717948718e-06,
+      "loss": 1.1766,
+      "step": 358
+    },
+    {
+      "epoch": 0.0276749922910885,
+      "grad_norm": 4.399728775024414,
+      "learning_rate": 9.205128205128206e-06,
+      "loss": 1.2096,
+      "step": 359
+    },
+    {
+      "epoch": 0.027752081406105456,
+      "grad_norm": 4.5362043380737305,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.1005,
+      "step": 360
+    },
+    {
+      "epoch": 0.027829170521122416,
+      "grad_norm": 4.581974983215332,
+      "learning_rate": 9.256410256410257e-06,
+      "loss": 1.1755,
+      "step": 361
+    },
+    {
+      "epoch": 0.027906259636139377,
+      "grad_norm": 5.120265960693359,
+      "learning_rate": 9.282051282051283e-06,
+      "loss": 1.1285,
+      "step": 362
+    },
+    {
+      "epoch": 0.027983348751156337,
+      "grad_norm": 4.422695636749268,
+      "learning_rate": 9.307692307692308e-06,
+      "loss": 1.2343,
+      "step": 363
+    },
+    {
+      "epoch": 0.028060437866173297,
+      "grad_norm": 4.639062881469727,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.2733,
+      "step": 364
+    },
+    {
+      "epoch": 0.028137526981190257,
+      "grad_norm": 5.054222106933594,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 1.1177,
+      "step": 365
+    },
+    {
+      "epoch": 0.028214616096207217,
+      "grad_norm": 5.064762592315674,
+      "learning_rate": 9.384615384615385e-06,
+      "loss": 1.1147,
+      "step": 366
+    },
+    {
+      "epoch": 0.028291705211224174,
+      "grad_norm": 4.730871200561523,
+      "learning_rate": 9.410256410256412e-06,
+      "loss": 1.3437,
+      "step": 367
+    },
+    {
+      "epoch": 0.028368794326241134,
+      "grad_norm": 4.744129180908203,
+      "learning_rate": 9.435897435897436e-06,
+      "loss": 1.3698,
+      "step": 368
+    },
+    {
+      "epoch": 0.028445883441258094,
+      "grad_norm": 4.724253177642822,
+      "learning_rate": 9.461538461538463e-06,
+      "loss": 1.1685,
+      "step": 369
+    },
+    {
+      "epoch": 0.028522972556275054,
+      "grad_norm": 4.356321334838867,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 1.154,
+      "step": 370
+    },
+    {
+      "epoch": 0.028600061671292015,
+      "grad_norm": 4.506900310516357,
+      "learning_rate": 9.512820512820514e-06,
+      "loss": 1.233,
+      "step": 371
+    },
+    {
+      "epoch": 0.028677150786308975,
+      "grad_norm": 4.505119800567627,
+      "learning_rate": 9.53846153846154e-06,
+      "loss": 1.3049,
+      "step": 372
+    },
+    {
+      "epoch": 0.02875423990132593,
+      "grad_norm": 4.503683090209961,
+      "learning_rate": 9.564102564102565e-06,
+      "loss": 1.1666,
+      "step": 373
+    },
+    {
+      "epoch": 0.02883132901634289,
+      "grad_norm": 4.283431529998779,
+      "learning_rate": 9.589743589743591e-06,
+      "loss": 1.1497,
+      "step": 374
+    },
+    {
+      "epoch": 0.028908418131359852,
+      "grad_norm": 5.00119686126709,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 1.2709,
+      "step": 375
+    },
+    {
+      "epoch": 0.028985507246376812,
+      "grad_norm": 4.850398540496826,
+      "learning_rate": 9.641025641025642e-06,
+      "loss": 1.2135,
+      "step": 376
+    },
+    {
+      "epoch": 0.029062596361393772,
+      "grad_norm": 5.15523624420166,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 1.1788,
+      "step": 377
+    },
+    {
+      "epoch": 0.029139685476410732,
+      "grad_norm": 4.818391799926758,
+      "learning_rate": 9.692307692307693e-06,
+      "loss": 1.1626,
+      "step": 378
+    },
+    {
+      "epoch": 0.02921677459142769,
+      "grad_norm": 5.061548233032227,
+      "learning_rate": 9.71794871794872e-06,
+      "loss": 1.2423,
+      "step": 379
+    },
+    {
+      "epoch": 0.02929386370644465,
+      "grad_norm": 4.615644931793213,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 1.1704,
+      "step": 380
+    },
+    {
+      "epoch": 0.02937095282146161,
+      "grad_norm": 4.75278902053833,
+      "learning_rate": 9.76923076923077e-06,
+      "loss": 1.1854,
+      "step": 381
+    },
+    {
+      "epoch": 0.02944804193647857,
+      "grad_norm": 4.710444450378418,
+      "learning_rate": 9.794871794871795e-06,
+      "loss": 1.2158,
+      "step": 382
+    },
+    {
+      "epoch": 0.02952513105149553,
+      "grad_norm": 4.5607709884643555,
+      "learning_rate": 9.820512820512821e-06,
+      "loss": 1.1627,
+      "step": 383
+    },
+    {
+      "epoch": 0.02960222016651249,
+      "grad_norm": 4.754578113555908,
+      "learning_rate": 9.846153846153848e-06,
+      "loss": 1.3299,
+      "step": 384
+    },
+    {
+      "epoch": 0.029679309281529447,
+      "grad_norm": 4.9476542472839355,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 1.1626,
+      "step": 385
+    },
+    {
+      "epoch": 0.029756398396546407,
+      "grad_norm": 4.63112211227417,
+      "learning_rate": 9.897435897435899e-06,
+      "loss": 1.2714,
+      "step": 386
+    },
+    {
+      "epoch": 0.029833487511563367,
+      "grad_norm": 4.136993885040283,
+      "learning_rate": 9.923076923076923e-06,
+      "loss": 1.1761,
+      "step": 387
+    },
+    {
+      "epoch": 0.029910576626580327,
+      "grad_norm": 4.531764984130859,
+      "learning_rate": 9.94871794871795e-06,
+      "loss": 1.1452,
+      "step": 388
+    },
+    {
+      "epoch": 0.029987665741597287,
+      "grad_norm": 5.1062774658203125,
+      "learning_rate": 9.974358974358974e-06,
+      "loss": 1.2642,
+      "step": 389
+    },
+    {
+      "epoch": 0.030064754856614247,
+      "grad_norm": 4.920534610748291,
+      "learning_rate": 1e-05,
+      "loss": 1.2648,
+      "step": 390
+    },
+    {
+      "epoch": 0.030141843971631204,
+      "grad_norm": 5.01254415512085,
+      "learning_rate": 9.99999984413795e-06,
+      "loss": 1.1768,
+      "step": 391
+    },
+    {
+      "epoch": 0.030218933086648164,
+      "grad_norm": 4.516940116882324,
+      "learning_rate": 9.999999376551802e-06,
+      "loss": 1.303,
+      "step": 392
+    },
+    {
+      "epoch": 0.030296022201665124,
+      "grad_norm": 4.425420761108398,
+      "learning_rate": 9.99999859724159e-06,
+      "loss": 1.3071,
+      "step": 393
+    },
+    {
+      "epoch": 0.030373111316682085,
+      "grad_norm": 5.187442302703857,
+      "learning_rate": 9.999997506207361e-06,
+      "loss": 1.3249,
+      "step": 394
+    },
+    {
+      "epoch": 0.030450200431699045,
+      "grad_norm": 4.589512348175049,
+      "learning_rate": 9.999996103449184e-06,
+      "loss": 1.2048,
+      "step": 395
+    },
+    {
+      "epoch": 0.030527289546716005,
+      "grad_norm": 4.254446983337402,
+      "learning_rate": 9.999994388967143e-06,
+      "loss": 1.2693,
+      "step": 396
+    },
+    {
+      "epoch": 0.03060437866173296,
+      "grad_norm": 4.697052478790283,
+      "learning_rate": 9.999992362761349e-06,
+      "loss": 1.2305,
+      "step": 397
+    },
+    {
+      "epoch": 0.030681467776749922,
+      "grad_norm": 4.343505382537842,
+      "learning_rate": 9.999990024831926e-06,
+      "loss": 1.2191,
+      "step": 398
+    },
+    {
+      "epoch": 0.030758556891766882,
+      "grad_norm": 4.3586015701293945,
+      "learning_rate": 9.999987375179023e-06,
+      "loss": 1.197,
+      "step": 399
+    },
+    {
+      "epoch": 0.030835646006783842,
+      "grad_norm": 4.417754173278809,
+      "learning_rate": 9.999984413802802e-06,
+      "loss": 1.2121,
+      "step": 400
+    },
+    {
+      "epoch": 0.030912735121800802,
+      "grad_norm": 4.340965747833252,
+      "learning_rate": 9.999981140703447e-06,
+      "loss": 1.1745,
+      "step": 401
+    },
+    {
+      "epoch": 0.030989824236817762,
+      "grad_norm": 4.833497047424316,
+      "learning_rate": 9.999977555881163e-06,
+      "loss": 1.2646,
+      "step": 402
+    },
+    {
+      "epoch": 0.031066913351834723,
+      "grad_norm": 4.495340824127197,
+      "learning_rate": 9.999973659336176e-06,
+      "loss": 1.1996,
+      "step": 403
+    },
+    {
+      "epoch": 0.03114400246685168,
+      "grad_norm": 4.7980475425720215,
+      "learning_rate": 9.999969451068725e-06,
+      "loss": 1.2813,
+      "step": 404
+    },
+    {
+      "epoch": 0.03122109158186864,
+      "grad_norm": 4.268666744232178,
+      "learning_rate": 9.999964931079076e-06,
+      "loss": 1.2938,
+      "step": 405
+    },
+    {
+      "epoch": 0.0312981806968856,
+      "grad_norm": 4.3926777839660645,
+      "learning_rate": 9.999960099367507e-06,
+      "loss": 1.236,
+      "step": 406
+    },
+    {
+      "epoch": 0.031375269811902556,
+      "grad_norm": 4.350759029388428,
+      "learning_rate": 9.999954955934323e-06,
+      "loss": 1.1252,
+      "step": 407
+    },
+    {
+      "epoch": 0.03145235892691952,
+      "grad_norm": 4.4295220375061035,
+      "learning_rate": 9.999949500779842e-06,
+      "loss": 1.2229,
+      "step": 408
+    },
+    {
+      "epoch": 0.03152944804193648,
+      "grad_norm": 4.337591171264648,
+      "learning_rate": 9.999943733904404e-06,
+      "loss": 1.1971,
+      "step": 409
+    },
+    {
+      "epoch": 0.03160653715695344,
+      "grad_norm": 4.42830753326416,
+      "learning_rate": 9.999937655308373e-06,
+      "loss": 1.1197,
+      "step": 410
+    },
+    {
+      "epoch": 0.0316836262719704,
+      "grad_norm": 5.042124271392822,
+      "learning_rate": 9.999931264992122e-06,
+      "loss": 1.1953,
+      "step": 411
+    },
+    {
+      "epoch": 0.03176071538698736,
+      "grad_norm": 4.555182933807373,
+      "learning_rate": 9.999924562956052e-06,
+      "loss": 1.2661,
+      "step": 412
+    },
+    {
+      "epoch": 0.03183780450200432,
+      "grad_norm": 4.939544677734375,
+      "learning_rate": 9.999917549200581e-06,
+      "loss": 1.2496,
+      "step": 413
+    },
+    {
+      "epoch": 0.031914893617021274,
+      "grad_norm": 4.4207539558410645,
+      "learning_rate": 9.999910223726146e-06,
+      "loss": 1.2442,
+      "step": 414
+    },
+    {
+      "epoch": 0.03199198273203824,
+      "grad_norm": 4.292929649353027,
+      "learning_rate": 9.999902586533205e-06,
+      "loss": 1.1871,
+      "step": 415
+    },
+    {
+      "epoch": 0.032069071847055194,
+      "grad_norm": 4.947376251220703,
+      "learning_rate": 9.999894637622231e-06,
+      "loss": 1.1849,
+      "step": 416
+    },
+    {
+      "epoch": 0.03214616096207216,
+      "grad_norm": 4.807240009307861,
+      "learning_rate": 9.999886376993723e-06,
+      "loss": 1.3515,
+      "step": 417
+    },
+    {
+      "epoch": 0.032223250077089115,
+      "grad_norm": 4.25323486328125,
+      "learning_rate": 9.999877804648194e-06,
+      "loss": 1.1788,
+      "step": 418
+    },
+    {
+      "epoch": 0.03230033919210607,
+      "grad_norm": 4.476768970489502,
+      "learning_rate": 9.999868920586178e-06,
+      "loss": 1.1857,
+      "step": 419
+    },
+    {
+      "epoch": 0.032377428307123035,
+      "grad_norm": 4.515340805053711,
+      "learning_rate": 9.999859724808231e-06,
+      "loss": 1.1074,
+      "step": 420
+    },
+    {
+      "epoch": 0.03245451742213999,
+      "grad_norm": 4.648842811584473,
+      "learning_rate": 9.999850217314924e-06,
+      "loss": 1.2364,
+      "step": 421
+    },
+    {
+      "epoch": 0.032531606537156955,
+      "grad_norm": 4.573676109313965,
+      "learning_rate": 9.999840398106852e-06,
+      "loss": 1.3302,
+      "step": 422
+    },
+    {
+      "epoch": 0.03260869565217391,
+      "grad_norm": 4.733384132385254,
+      "learning_rate": 9.999830267184625e-06,
+      "loss": 1.2797,
+      "step": 423
+    },
+    {
+      "epoch": 0.032685784767190876,
+      "grad_norm": 4.6709303855896,
+      "learning_rate": 9.999819824548875e-06,
+      "loss": 1.2243,
+      "step": 424
+    },
+    {
+      "epoch": 0.03276287388220783,
+      "grad_norm": 4.113539695739746,
+      "learning_rate": 9.999809070200256e-06,
+      "loss": 1.0434,
+      "step": 425
+    },
+    {
+      "epoch": 0.03283996299722479,
+      "grad_norm": 4.282526969909668,
+      "learning_rate": 9.999798004139435e-06,
+      "loss": 1.2439,
+      "step": 426
+    },
+    {
+      "epoch": 0.03291705211224175,
+      "grad_norm": 4.742609977722168,
+      "learning_rate": 9.999786626367102e-06,
+      "loss": 1.3161,
+      "step": 427
+    },
+    {
+      "epoch": 0.03299414122725871,
+      "grad_norm": 4.848706245422363,
+      "learning_rate": 9.999774936883969e-06,
+      "loss": 1.2556,
+      "step": 428
+    },
+    {
+      "epoch": 0.03307123034227567,
+      "grad_norm": 5.125500202178955,
+      "learning_rate": 9.999762935690762e-06,
+      "loss": 1.2509,
+      "step": 429
+    },
+    {
+      "epoch": 0.03314831945729263,
+      "grad_norm": 4.103264808654785,
+      "learning_rate": 9.999750622788232e-06,
+      "loss": 1.3005,
+      "step": 430
+    },
+    {
+      "epoch": 0.03322540857230959,
+      "grad_norm": 4.143542289733887,
+      "learning_rate": 9.999737998177144e-06,
+      "loss": 1.1695,
+      "step": 431
+    },
+    {
+      "epoch": 0.03330249768732655,
+      "grad_norm": 4.100920677185059,
+      "learning_rate": 9.999725061858286e-06,
+      "loss": 1.23,
+      "step": 432
+    },
+    {
+      "epoch": 0.03337958680234351,
+      "grad_norm": 4.6039299964904785,
+      "learning_rate": 9.999711813832465e-06,
+      "loss": 1.1966,
+      "step": 433
+    },
+    {
+      "epoch": 0.03345667591736047,
+      "grad_norm": 5.09745979309082,
+      "learning_rate": 9.999698254100506e-06,
+      "loss": 1.197,
+      "step": 434
+    },
+    {
+      "epoch": 0.03353376503237743,
+      "grad_norm": 4.1532769203186035,
+      "learning_rate": 9.999684382663254e-06,
+      "loss": 1.0883,
+      "step": 435
+    },
+    {
+      "epoch": 0.03361085414739439,
+      "grad_norm": 4.5222697257995605,
+      "learning_rate": 9.999670199521577e-06,
+      "loss": 1.2341,
+      "step": 436
+    },
+    {
+      "epoch": 0.03368794326241135,
+      "grad_norm": 4.2074294090271,
+      "learning_rate": 9.999655704676357e-06,
+      "loss": 1.3028,
+      "step": 437
+    },
+    {
+      "epoch": 0.033765032377428304,
+      "grad_norm": 4.485882759094238,
+      "learning_rate": 9.999640898128495e-06,
+      "loss": 1.2003,
+      "step": 438
+    },
+    {
+      "epoch": 0.03384212149244527,
+      "grad_norm": 4.506419658660889,
+      "learning_rate": 9.999625779878918e-06,
+      "loss": 1.2515,
+      "step": 439
+    },
+    {
+      "epoch": 0.033919210607462225,
+      "grad_norm": 4.3678436279296875,
+      "learning_rate": 9.99961034992857e-06,
+      "loss": 1.2829,
+      "step": 440
+    },
+    {
+      "epoch": 0.03399629972247919,
+      "grad_norm": 4.121408462524414,
+      "learning_rate": 9.999594608278407e-06,
+      "loss": 1.1691,
+      "step": 441
+    },
+    {
+      "epoch": 0.034073388837496145,
+      "grad_norm": 4.504795074462891,
+      "learning_rate": 9.999578554929415e-06,
+      "loss": 1.1358,
+      "step": 442
+    },
+    {
+      "epoch": 0.03415047795251311,
+      "grad_norm": 4.258967876434326,
+      "learning_rate": 9.999562189882594e-06,
+      "loss": 1.1833,
+      "step": 443
+    },
+    {
+      "epoch": 0.034227567067530065,
+      "grad_norm": 4.424483776092529,
+      "learning_rate": 9.999545513138964e-06,
+      "loss": 1.2847,
+      "step": 444
+    },
+    {
+      "epoch": 0.03430465618254702,
+      "grad_norm": 4.2883687019348145,
+      "learning_rate": 9.999528524699563e-06,
+      "loss": 1.1569,
+      "step": 445
+    },
+    {
+      "epoch": 0.034381745297563986,
+      "grad_norm": 4.397137641906738,
+      "learning_rate": 9.999511224565453e-06,
+      "loss": 1.1686,
+      "step": 446
+    },
+    {
+      "epoch": 0.03445883441258094,
+      "grad_norm": 4.637729644775391,
+      "learning_rate": 9.999493612737712e-06,
+      "loss": 1.2365,
+      "step": 447
+    },
+    {
+      "epoch": 0.034535923527597906,
+      "grad_norm": 4.562082290649414,
+      "learning_rate": 9.999475689217434e-06,
+      "loss": 1.1553,
+      "step": 448
+    },
+    {
+      "epoch": 0.03461301264261486,
+      "grad_norm": 4.052862644195557,
+      "learning_rate": 9.999457454005743e-06,
+      "loss": 1.1823,
+      "step": 449
+    },
+    {
+      "epoch": 0.03469010175763182,
+      "grad_norm": 4.691180229187012,
+      "learning_rate": 9.999438907103772e-06,
+      "loss": 1.2576,
+      "step": 450
+    },
+    {
+      "epoch": 0.03476719087264878,
+      "grad_norm": 4.515274524688721,
+      "learning_rate": 9.999420048512678e-06,
+      "loss": 1.1499,
+      "step": 451
+    },
+    {
+      "epoch": 0.03484427998766574,
+      "grad_norm": 4.499693393707275,
+      "learning_rate": 9.999400878233636e-06,
+      "loss": 1.1628,
+      "step": 452
+    },
+    {
+      "epoch": 0.0349213691026827,
+      "grad_norm": 4.56060266494751,
+      "learning_rate": 9.999381396267841e-06,
+      "loss": 1.2407,
+      "step": 453
+    },
+    {
+      "epoch": 0.03499845821769966,
+      "grad_norm": 4.329947471618652,
+      "learning_rate": 9.99936160261651e-06,
+      "loss": 1.1197,
+      "step": 454
+    },
+    {
+      "epoch": 0.035075547332716624,
+      "grad_norm": 4.710033416748047,
+      "learning_rate": 9.999341497280875e-06,
+      "loss": 1.2235,
+      "step": 455
+    },
+    {
+      "epoch": 0.03515263644773358,
+      "grad_norm": 4.5141167640686035,
+      "learning_rate": 9.99932108026219e-06,
+      "loss": 1.2179,
+      "step": 456
+    },
+    {
+      "epoch": 0.03522972556275054,
+      "grad_norm": 4.853623390197754,
+      "learning_rate": 9.999300351561727e-06,
+      "loss": 1.2093,
+      "step": 457
+    },
+    {
+      "epoch": 0.0353068146777675,
+      "grad_norm": 4.331170558929443,
+      "learning_rate": 9.999279311180779e-06,
+      "loss": 1.3841,
+      "step": 458
+    },
+    {
+      "epoch": 0.03538390379278446,
+      "grad_norm": 4.377639293670654,
+      "learning_rate": 9.999257959120658e-06,
+      "loss": 1.2103,
+      "step": 459
+    },
+    {
+      "epoch": 0.03546099290780142,
+      "grad_norm": 4.253114700317383,
+      "learning_rate": 9.999236295382695e-06,
+      "loss": 1.1429,
+      "step": 460
+    },
+    {
+      "epoch": 0.03553808202281838,
+      "grad_norm": 4.613873481750488,
+      "learning_rate": 9.999214319968242e-06,
+      "loss": 1.2951,
+      "step": 461
+    },
+    {
+      "epoch": 0.035615171137835334,
+      "grad_norm": 4.3931498527526855,
+      "learning_rate": 9.999192032878667e-06,
+      "loss": 1.223,
+      "step": 462
+    },
+    {
+      "epoch": 0.0356922602528523,
+      "grad_norm": 4.192042827606201,
+      "learning_rate": 9.99916943411536e-06,
+      "loss": 1.1564,
+      "step": 463
+    },
+    {
+      "epoch": 0.035769349367869255,
+      "grad_norm": 4.241518974304199,
+      "learning_rate": 9.99914652367973e-06,
+      "loss": 1.1852,
+      "step": 464
+    },
+    {
+      "epoch": 0.03584643848288622,
+      "grad_norm": 4.183101177215576,
+      "learning_rate": 9.999123301573208e-06,
+      "loss": 1.2052,
+      "step": 465
+    },
+    {
+      "epoch": 0.035923527597903175,
+      "grad_norm": 4.752673149108887,
+      "learning_rate": 9.999099767797236e-06,
+      "loss": 1.1693,
+      "step": 466
+    },
+    {
+      "epoch": 0.03600061671292014,
+      "grad_norm": 4.796976089477539,
+      "learning_rate": 9.999075922353286e-06,
+      "loss": 1.3416,
+      "step": 467
+    },
+    {
+      "epoch": 0.036077705827937095,
+      "grad_norm": 4.705587863922119,
+      "learning_rate": 9.999051765242843e-06,
+      "loss": 1.2008,
+      "step": 468
+    },
+    {
+      "epoch": 0.03615479494295405,
+      "grad_norm": 4.491495609283447,
+      "learning_rate": 9.999027296467412e-06,
+      "loss": 1.1668,
+      "step": 469
+    },
+    {
+      "epoch": 0.036231884057971016,
+      "grad_norm": 4.496662139892578,
+      "learning_rate": 9.99900251602852e-06,
+      "loss": 1.2627,
+      "step": 470
+    },
+    {
+      "epoch": 0.03630897317298797,
+      "grad_norm": 4.168951511383057,
+      "learning_rate": 9.998977423927714e-06,
+      "loss": 1.1625,
+      "step": 471
+    },
+    {
+      "epoch": 0.036386062288004936,
+      "grad_norm": 4.014923095703125,
+      "learning_rate": 9.998952020166554e-06,
+      "loss": 1.1659,
+      "step": 472
+    },
+    {
+      "epoch": 0.03646315140302189,
+      "grad_norm": 4.847691059112549,
+      "learning_rate": 9.998926304746626e-06,
+      "loss": 1.1743,
+      "step": 473
+    },
+    {
+      "epoch": 0.03654024051803885,
+      "grad_norm": 4.13588809967041,
+      "learning_rate": 9.998900277669531e-06,
+      "loss": 1.1655,
+      "step": 474
+    },
+    {
+      "epoch": 0.03661732963305581,
+      "grad_norm": 4.349898815155029,
+      "learning_rate": 9.998873938936897e-06,
+      "loss": 1.2163,
+      "step": 475
+    },
+    {
+      "epoch": 0.03669441874807277,
+      "grad_norm": 4.191585540771484,
+      "learning_rate": 9.998847288550363e-06,
+      "loss": 1.1745,
+      "step": 476
+    },
+    {
+      "epoch": 0.036771507863089734,
+      "grad_norm": 4.346571922302246,
+      "learning_rate": 9.998820326511587e-06,
+      "loss": 1.1258,
+      "step": 477
+    },
+    {
+      "epoch": 0.03684859697810669,
+      "grad_norm": 4.2652459144592285,
+      "learning_rate": 9.998793052822256e-06,
+      "loss": 1.2362,
+      "step": 478
+    },
+    {
+      "epoch": 0.036925686093123654,
+      "grad_norm": 4.184598445892334,
+      "learning_rate": 9.998765467484066e-06,
+      "loss": 1.1744,
+      "step": 479
+    },
+    {
+      "epoch": 0.03700277520814061,
+      "grad_norm": 4.4486260414123535,
+      "learning_rate": 9.998737570498737e-06,
+      "loss": 1.1812,
+      "step": 480
+    },
+    {
+      "epoch": 0.03707986432315757,
+      "grad_norm": 4.7114410400390625,
+      "learning_rate": 9.998709361868012e-06,
+      "loss": 1.2171,
+      "step": 481
+    },
+    {
+      "epoch": 0.03715695343817453,
+      "grad_norm": 4.759378910064697,
+      "learning_rate": 9.998680841593647e-06,
+      "loss": 1.2212,
+      "step": 482
+    },
+    {
+      "epoch": 0.03723404255319149,
+      "grad_norm": 4.3762922286987305,
+      "learning_rate": 9.998652009677421e-06,
+      "loss": 1.1983,
+      "step": 483
+    },
+    {
+      "epoch": 0.03731113166820845,
+      "grad_norm": 4.322701454162598,
+      "learning_rate": 9.998622866121129e-06,
+      "loss": 1.2557,
+      "step": 484
+    },
+    {
+      "epoch": 0.03738822078322541,
+      "grad_norm": 4.573087692260742,
+      "learning_rate": 9.998593410926591e-06,
+      "loss": 1.2969,
+      "step": 485
+    },
+    {
+      "epoch": 0.03746530989824237,
+      "grad_norm": 4.229235649108887,
+      "learning_rate": 9.998563644095642e-06,
+      "loss": 1.2178,
+      "step": 486
+    },
+    {
+      "epoch": 0.03754239901325933,
+      "grad_norm": 4.301479339599609,
+      "learning_rate": 9.998533565630138e-06,
+      "loss": 1.2294,
+      "step": 487
+    },
+    {
+      "epoch": 0.037619488128276285,
+      "grad_norm": 4.079572677612305,
+      "learning_rate": 9.998503175531955e-06,
+      "loss": 1.2311,
+      "step": 488
+    },
+    {
+      "epoch": 0.03769657724329325,
+      "grad_norm": 5.241011619567871,
+      "learning_rate": 9.998472473802984e-06,
+      "loss": 1.2433,
+      "step": 489
+    },
+    {
+      "epoch": 0.037773666358310205,
+      "grad_norm": 4.18917989730835,
+      "learning_rate": 9.998441460445145e-06,
+      "loss": 1.3185,
+      "step": 490
+    },
+    {
+      "epoch": 0.03785075547332717,
+      "grad_norm": 4.833195209503174,
+      "learning_rate": 9.998410135460368e-06,
+      "loss": 1.2041,
+      "step": 491
+    },
+    {
+      "epoch": 0.037927844588344126,
+      "grad_norm": 4.499256610870361,
+      "learning_rate": 9.998378498850605e-06,
+      "loss": 1.2148,
+      "step": 492
+    },
+    {
+      "epoch": 0.03800493370336108,
+      "grad_norm": 4.2141947746276855,
+      "learning_rate": 9.998346550617833e-06,
+      "loss": 1.1423,
+      "step": 493
+    },
+    {
+      "epoch": 0.038082022818378046,
+      "grad_norm": 4.486852169036865,
+      "learning_rate": 9.998314290764038e-06,
+      "loss": 1.2306,
+      "step": 494
+    },
+    {
+      "epoch": 0.038159111933395,
+      "grad_norm": 4.096106052398682,
+      "learning_rate": 9.998281719291234e-06,
+      "loss": 1.113,
+      "step": 495
+    },
+    {
+      "epoch": 0.038236201048411966,
+      "grad_norm": 4.3826985359191895,
+      "learning_rate": 9.998248836201452e-06,
+      "loss": 1.1976,
+      "step": 496
+    },
+    {
+      "epoch": 0.03831329016342892,
+      "grad_norm": 4.423654079437256,
+      "learning_rate": 9.998215641496743e-06,
+      "loss": 1.1227,
+      "step": 497
+    },
+    {
+      "epoch": 0.03839037927844589,
+      "grad_norm": 3.936204671859741,
+      "learning_rate": 9.998182135179173e-06,
+      "loss": 1.1393,
+      "step": 498
+    },
+    {
+      "epoch": 0.03846746839346284,
+      "grad_norm": 4.782196521759033,
+      "learning_rate": 9.998148317250835e-06,
+      "loss": 1.1804,
+      "step": 499
+    },
+    {
+      "epoch": 0.0385445575084798,
+      "grad_norm": 4.143787384033203,
+      "learning_rate": 9.998114187713834e-06,
+      "loss": 1.1194,
+      "step": 500
+    },
+    {
+      "epoch": 0.038621646623496764,
+      "grad_norm": 4.448570728302002,
+      "learning_rate": 9.998079746570299e-06,
+      "loss": 1.2344,
+      "step": 501
+    },
+    {
+      "epoch": 0.03869873573851372,
+      "grad_norm": 4.438679218292236,
+      "learning_rate": 9.99804499382238e-06,
+      "loss": 1.3456,
+      "step": 502
+    },
+    {
+      "epoch": 0.038775824853530684,
+      "grad_norm": 4.5678606033325195,
+      "learning_rate": 9.99800992947224e-06,
+      "loss": 1.196,
+      "step": 503
+    },
+    {
+      "epoch": 0.03885291396854764,
+      "grad_norm": 4.333302021026611,
+      "learning_rate": 9.997974553522066e-06,
+      "loss": 1.2677,
+      "step": 504
+    },
+    {
+      "epoch": 0.0389300030835646,
+      "grad_norm": 4.986594200134277,
+      "learning_rate": 9.997938865974063e-06,
+      "loss": 1.2973,
+      "step": 505
+    },
+    {
+      "epoch": 0.03900709219858156,
+      "grad_norm": 4.153099536895752,
+      "learning_rate": 9.997902866830459e-06,
+      "loss": 1.1358,
+      "step": 506
+    },
+    {
+      "epoch": 0.03908418131359852,
+      "grad_norm": 4.446223735809326,
+      "learning_rate": 9.997866556093491e-06,
+      "loss": 1.2034,
+      "step": 507
+    },
+    {
+      "epoch": 0.03916127042861548,
+      "grad_norm": 4.689149379730225,
+      "learning_rate": 9.997829933765433e-06,
+      "loss": 1.242,
+      "step": 508
+    },
+    {
+      "epoch": 0.03923835954363244,
+      "grad_norm": 4.689703941345215,
+      "learning_rate": 9.997792999848561e-06,
+      "loss": 1.2383,
+      "step": 509
+    },
+    {
+      "epoch": 0.0393154486586494,
+      "grad_norm": 4.7472429275512695,
+      "learning_rate": 9.997755754345179e-06,
+      "loss": 1.3211,
+      "step": 510
+    },
+    {
+      "epoch": 0.03939253777366636,
+      "grad_norm": 4.735124111175537,
+      "learning_rate": 9.997718197257612e-06,
+      "loss": 1.2133,
+      "step": 511
+    },
+    {
+      "epoch": 0.039469626888683315,
+      "grad_norm": 4.425952911376953,
+      "learning_rate": 9.997680328588198e-06,
+      "loss": 1.1441,
+      "step": 512
+    },
+    {
+      "epoch": 0.03954671600370028,
+      "grad_norm": 4.210488319396973,
+      "learning_rate": 9.9976421483393e-06,
+      "loss": 1.2533,
+      "step": 513
+    },
+    {
+      "epoch": 0.039623805118717235,
+      "grad_norm": 4.078352451324463,
+      "learning_rate": 9.997603656513295e-06,
+      "loss": 1.179,
+      "step": 514
+    },
+    {
+      "epoch": 0.0397008942337342,
+      "grad_norm": 4.232064247131348,
+      "learning_rate": 9.997564853112586e-06,
+      "loss": 1.2968,
+      "step": 515
+    },
+    {
+      "epoch": 0.039777983348751156,
+      "grad_norm": 4.38081169128418,
+      "learning_rate": 9.997525738139595e-06,
+      "loss": 1.1834,
+      "step": 516
+    },
+    {
+      "epoch": 0.03985507246376811,
+      "grad_norm": 4.6666388511657715,
+      "learning_rate": 9.997486311596754e-06,
+      "loss": 1.16,
+      "step": 517
+    },
+    {
+      "epoch": 0.039932161578785076,
+      "grad_norm": 4.461724758148193,
+      "learning_rate": 9.997446573486524e-06,
+      "loss": 1.2314,
+      "step": 518
+    },
+    {
+      "epoch": 0.04000925069380203,
+      "grad_norm": 4.663398265838623,
+      "learning_rate": 9.997406523811383e-06,
+      "loss": 1.1742,
+      "step": 519
+    },
+    {
+      "epoch": 0.040086339808818996,
+      "grad_norm": 4.144864559173584,
+      "learning_rate": 9.997366162573828e-06,
+      "loss": 1.1362,
+      "step": 520
+    },
+    {
+      "epoch": 0.04016342892383595,
+      "grad_norm": 4.398372650146484,
+      "learning_rate": 9.997325489776375e-06,
+      "loss": 1.1925,
+      "step": 521
+    },
+    {
+      "epoch": 0.04024051803885292,
+      "grad_norm": 4.672181606292725,
+      "learning_rate": 9.997284505421559e-06,
+      "loss": 1.2667,
+      "step": 522
+    },
+    {
+      "epoch": 0.040317607153869874,
+      "grad_norm": 4.4285569190979,
+      "learning_rate": 9.997243209511935e-06,
+      "loss": 1.3158,
+      "step": 523
+    },
+    {
+      "epoch": 0.04039469626888683,
+      "grad_norm": 4.436081886291504,
+      "learning_rate": 9.99720160205008e-06,
+      "loss": 1.1041,
+      "step": 524
+    },
+    {
+      "epoch": 0.040471785383903794,
+      "grad_norm": 4.347367286682129,
+      "learning_rate": 9.997159683038584e-06,
+      "loss": 1.1595,
+      "step": 525
+    },
+    {
+      "epoch": 0.04054887449892075,
+      "grad_norm": 4.27985143661499,
+      "learning_rate": 9.997117452480063e-06,
+      "loss": 1.283,
+      "step": 526
+    },
+    {
+      "epoch": 0.040625963613937714,
+      "grad_norm": 4.468918800354004,
+      "learning_rate": 9.997074910377151e-06,
+      "loss": 1.1987,
+      "step": 527
+    },
+    {
+      "epoch": 0.04070305272895467,
+      "grad_norm": 4.384061813354492,
+      "learning_rate": 9.997032056732498e-06,
+      "loss": 1.1361,
+      "step": 528
+    },
+    {
+      "epoch": 0.040780141843971635,
+      "grad_norm": 4.80853271484375,
+      "learning_rate": 9.996988891548777e-06,
+      "loss": 1.2749,
+      "step": 529
+    },
+    {
+      "epoch": 0.04085723095898859,
+      "grad_norm": 4.914130210876465,
+      "learning_rate": 9.99694541482868e-06,
+      "loss": 1.4303,
+      "step": 530
+    },
+    {
+      "epoch": 0.04093432007400555,
+      "grad_norm": 4.093295097351074,
+      "learning_rate": 9.996901626574915e-06,
+      "loss": 1.155,
+      "step": 531
+    },
+    {
+      "epoch": 0.04101140918902251,
+      "grad_norm": 4.31951379776001,
+      "learning_rate": 9.996857526790212e-06,
+      "loss": 1.206,
+      "step": 532
+    },
+    {
+      "epoch": 0.04108849830403947,
+      "grad_norm": 6.6202521324157715,
+      "learning_rate": 9.996813115477323e-06,
+      "loss": 1.1679,
+      "step": 533
+    },
+    {
+      "epoch": 0.04116558741905643,
+      "grad_norm": 4.6812334060668945,
+      "learning_rate": 9.996768392639015e-06,
+      "loss": 1.205,
+      "step": 534
+    },
+    {
+      "epoch": 0.04124267653407339,
+      "grad_norm": 4.417919158935547,
+      "learning_rate": 9.996723358278077e-06,
+      "loss": 1.2173,
+      "step": 535
+    },
+    {
+      "epoch": 0.041319765649090345,
+      "grad_norm": 4.809331893920898,
+      "learning_rate": 9.996678012397317e-06,
+      "loss": 1.2465,
+      "step": 536
+    },
+    {
+      "epoch": 0.04139685476410731,
+      "grad_norm": 4.317751884460449,
+      "learning_rate": 9.99663235499956e-06,
+      "loss": 1.1406,
+      "step": 537
+    },
+    {
+      "epoch": 0.041473943879124266,
+      "grad_norm": 4.429257392883301,
+      "learning_rate": 9.996586386087653e-06,
+      "loss": 1.2509,
+      "step": 538
+    },
+    {
+      "epoch": 0.04155103299414123,
+      "grad_norm": 4.199450969696045,
+      "learning_rate": 9.996540105664464e-06,
+      "loss": 1.1988,
+      "step": 539
+    },
+    {
+      "epoch": 0.041628122109158186,
+      "grad_norm": 4.068055152893066,
+      "learning_rate": 9.996493513732878e-06,
+      "loss": 1.212,
+      "step": 540
+    },
+    {
+      "epoch": 0.04170521122417515,
+      "grad_norm": 4.471511363983154,
+      "learning_rate": 9.996446610295797e-06,
+      "loss": 1.2122,
+      "step": 541
+    },
+    {
+      "epoch": 0.041782300339192106,
+      "grad_norm": 4.128568172454834,
+      "learning_rate": 9.996399395356149e-06,
+      "loss": 1.2008,
+      "step": 542
+    },
+    {
+      "epoch": 0.04185938945420906,
+      "grad_norm": 4.495331764221191,
+      "learning_rate": 9.996351868916874e-06,
+      "loss": 1.2308,
+      "step": 543
+    },
+    {
+      "epoch": 0.04193647856922603,
+      "grad_norm": 4.763047218322754,
+      "learning_rate": 9.996304030980939e-06,
+      "loss": 1.2247,
+      "step": 544
+    },
+    {
+      "epoch": 0.04201356768424298,
+      "grad_norm": 4.249602317810059,
+      "learning_rate": 9.996255881551322e-06,
+      "loss": 1.2335,
+      "step": 545
+    },
+    {
+      "epoch": 0.04209065679925995,
+      "grad_norm": 4.013154983520508,
+      "learning_rate": 9.996207420631029e-06,
+      "loss": 1.0859,
+      "step": 546
+    },
+    {
+      "epoch": 0.042167745914276904,
+      "grad_norm": 4.579375743865967,
+      "learning_rate": 9.996158648223077e-06,
+      "loss": 1.1609,
+      "step": 547
+    },
+    {
+      "epoch": 0.04224483502929386,
+      "grad_norm": 4.4072747230529785,
+      "learning_rate": 9.996109564330513e-06,
+      "loss": 1.2142,
+      "step": 548
+    },
+    {
+      "epoch": 0.042321924144310824,
+      "grad_norm": 5.151735782623291,
+      "learning_rate": 9.99606016895639e-06,
+      "loss": 1.2514,
+      "step": 549
+    },
+    {
+      "epoch": 0.04239901325932778,
+      "grad_norm": 4.274787902832031,
+      "learning_rate": 9.996010462103791e-06,
+      "loss": 1.087,
+      "step": 550
+    },
+    {
+      "epoch": 0.042476102374344744,
+      "grad_norm": 4.291683197021484,
+      "learning_rate": 9.995960443775814e-06,
+      "loss": 1.0762,
+      "step": 551
+    },
+    {
+      "epoch": 0.0425531914893617,
+      "grad_norm": 4.732009410858154,
+      "learning_rate": 9.99591011397558e-06,
+      "loss": 1.2677,
+      "step": 552
+    },
+    {
+      "epoch": 0.042630280604378665,
+      "grad_norm": 4.424540996551514,
+      "learning_rate": 9.995859472706223e-06,
+      "loss": 1.3005,
+      "step": 553
+    },
+    {
+      "epoch": 0.04270736971939562,
+      "grad_norm": 4.752233505249023,
+      "learning_rate": 9.995808519970902e-06,
+      "loss": 1.278,
+      "step": 554
+    },
+    {
+      "epoch": 0.04278445883441258,
+      "grad_norm": 3.9832558631896973,
+      "learning_rate": 9.995757255772795e-06,
+      "loss": 1.1136,
+      "step": 555
+    },
+    {
+      "epoch": 0.04286154794942954,
+      "grad_norm": 4.465639114379883,
+      "learning_rate": 9.995705680115098e-06,
+      "loss": 1.1282,
+      "step": 556
+    },
+    {
+      "epoch": 0.0429386370644465,
+      "grad_norm": 4.6259446144104,
+      "learning_rate": 9.995653793001023e-06,
+      "loss": 1.2864,
+      "step": 557
+    },
+    {
+      "epoch": 0.04301572617946346,
+      "grad_norm": 3.95756459236145,
+      "learning_rate": 9.995601594433808e-06,
+      "loss": 1.1661,
+      "step": 558
+    },
+    {
+      "epoch": 0.04309281529448042,
+      "grad_norm": 4.663955211639404,
+      "learning_rate": 9.995549084416706e-06,
+      "loss": 1.1438,
+      "step": 559
+    },
+    {
+      "epoch": 0.043169904409497376,
+      "grad_norm": 4.4860615730285645,
+      "learning_rate": 9.995496262952993e-06,
+      "loss": 1.3304,
+      "step": 560
+    },
+    {
+      "epoch": 0.04324699352451434,
+      "grad_norm": 4.189501762390137,
+      "learning_rate": 9.995443130045958e-06,
+      "loss": 1.1138,
+      "step": 561
+    },
+    {
+      "epoch": 0.043324082639531296,
+      "grad_norm": 4.266587734222412,
+      "learning_rate": 9.995389685698918e-06,
+      "loss": 1.1725,
+      "step": 562
+    },
+    {
+      "epoch": 0.04340117175454826,
+      "grad_norm": 4.076508045196533,
+      "learning_rate": 9.995335929915202e-06,
+      "loss": 1.2068,
+      "step": 563
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 4.271369934082031,
+      "learning_rate": 9.995281862698161e-06,
+      "loss": 1.2824,
+      "step": 564
+    },
+    {
+      "epoch": 0.04355534998458218,
+      "grad_norm": 4.289413928985596,
+      "learning_rate": 9.995227484051168e-06,
+      "loss": 1.1504,
+      "step": 565
+    },
+    {
+      "epoch": 0.043632439099599137,
+      "grad_norm": 4.292287826538086,
+      "learning_rate": 9.995172793977613e-06,
+      "loss": 1.1755,
+      "step": 566
+    },
+    {
+      "epoch": 0.04370952821461609,
+      "grad_norm": 4.293979167938232,
+      "learning_rate": 9.995117792480903e-06,
+      "loss": 1.1852,
+      "step": 567
+    },
+    {
+      "epoch": 0.04378661732963306,
+      "grad_norm": 4.438719272613525,
+      "learning_rate": 9.995062479564472e-06,
+      "loss": 1.2362,
+      "step": 568
+    },
+    {
+      "epoch": 0.043863706444650014,
+      "grad_norm": 4.045646667480469,
+      "learning_rate": 9.995006855231763e-06,
+      "loss": 1.1894,
+      "step": 569
+    },
+    {
+      "epoch": 0.04394079555966698,
+      "grad_norm": 4.521982669830322,
+      "learning_rate": 9.994950919486248e-06,
+      "loss": 1.2075,
+      "step": 570
+    },
+    {
+      "epoch": 0.044017884674683934,
+      "grad_norm": 4.346794605255127,
+      "learning_rate": 9.994894672331413e-06,
+      "loss": 1.0808,
+      "step": 571
+    },
+    {
+      "epoch": 0.0440949737897009,
+      "grad_norm": 3.817455291748047,
+      "learning_rate": 9.994838113770762e-06,
+      "loss": 1.1479,
+      "step": 572
+    },
+    {
+      "epoch": 0.044172062904717854,
+      "grad_norm": 4.4810566902160645,
+      "learning_rate": 9.994781243807825e-06,
+      "loss": 1.1589,
+      "step": 573
+    },
+    {
+      "epoch": 0.04424915201973481,
+      "grad_norm": 4.639770030975342,
+      "learning_rate": 9.994724062446145e-06,
+      "loss": 1.2405,
+      "step": 574
+    },
+    {
+      "epoch": 0.044326241134751775,
+      "grad_norm": 3.9520983695983887,
+      "learning_rate": 9.994666569689291e-06,
+      "loss": 1.106,
+      "step": 575
+    },
+    {
+      "epoch": 0.04440333024976873,
+      "grad_norm": 4.4173407554626465,
+      "learning_rate": 9.994608765540842e-06,
+      "loss": 1.1598,
+      "step": 576
+    },
+    {
+      "epoch": 0.044480419364785695,
+      "grad_norm": 3.9496123790740967,
+      "learning_rate": 9.994550650004404e-06,
+      "loss": 1.2553,
+      "step": 577
+    },
+    {
+      "epoch": 0.04455750847980265,
+      "grad_norm": 4.36904239654541,
+      "learning_rate": 9.994492223083603e-06,
+      "loss": 1.2274,
+      "step": 578
+    },
+    {
+      "epoch": 0.04463459759481961,
+      "grad_norm": 4.925714015960693,
+      "learning_rate": 9.994433484782076e-06,
+      "loss": 1.3003,
+      "step": 579
+    },
+    {
+      "epoch": 0.04471168670983657,
+      "grad_norm": 4.471517086029053,
+      "learning_rate": 9.994374435103489e-06,
+      "loss": 1.3164,
+      "step": 580
+    },
+    {
+      "epoch": 0.04478877582485353,
+      "grad_norm": 4.279041290283203,
+      "learning_rate": 9.994315074051525e-06,
+      "loss": 1.1954,
+      "step": 581
+    },
+    {
+      "epoch": 0.04486586493987049,
+      "grad_norm": 4.745578289031982,
+      "learning_rate": 9.994255401629878e-06,
+      "loss": 1.2623,
+      "step": 582
+    },
+    {
+      "epoch": 0.04494295405488745,
+      "grad_norm": 4.342554569244385,
+      "learning_rate": 9.994195417842276e-06,
+      "loss": 1.3354,
+      "step": 583
+    },
+    {
+      "epoch": 0.04502004316990441,
+      "grad_norm": 4.38990592956543,
+      "learning_rate": 9.994135122692454e-06,
+      "loss": 1.2993,
+      "step": 584
+    },
+    {
+      "epoch": 0.04509713228492137,
+      "grad_norm": 4.334774971008301,
+      "learning_rate": 9.99407451618417e-06,
+      "loss": 1.2366,
+      "step": 585
+    },
+    {
+      "epoch": 0.045174221399938326,
+      "grad_norm": 4.226545333862305,
+      "learning_rate": 9.994013598321208e-06,
+      "loss": 1.2286,
+      "step": 586
+    },
+    {
+      "epoch": 0.04525131051495529,
+      "grad_norm": 4.3245391845703125,
+      "learning_rate": 9.993952369107363e-06,
+      "loss": 1.2417,
+      "step": 587
+    },
+    {
+      "epoch": 0.045328399629972246,
+      "grad_norm": 4.576898097991943,
+      "learning_rate": 9.99389082854645e-06,
+      "loss": 1.3929,
+      "step": 588
+    },
+    {
+      "epoch": 0.04540548874498921,
+      "grad_norm": 4.47149658203125,
+      "learning_rate": 9.993828976642311e-06,
+      "loss": 1.2802,
+      "step": 589
+    },
+    {
+      "epoch": 0.04548257786000617,
+      "grad_norm": 4.014610290527344,
+      "learning_rate": 9.993766813398796e-06,
+      "loss": 1.1964,
+      "step": 590
+    },
+    {
+      "epoch": 0.04555966697502312,
+      "grad_norm": 4.17547607421875,
+      "learning_rate": 9.993704338819786e-06,
+      "loss": 1.2604,
+      "step": 591
+    },
+    {
+      "epoch": 0.04563675609004009,
+      "grad_norm": 4.3998823165893555,
+      "learning_rate": 9.993641552909172e-06,
+      "loss": 1.2443,
+      "step": 592
+    },
+    {
+      "epoch": 0.045713845205057044,
+      "grad_norm": 4.851663589477539,
+      "learning_rate": 9.99357845567087e-06,
+      "loss": 1.3904,
+      "step": 593
+    },
+    {
+      "epoch": 0.04579093432007401,
+      "grad_norm": 4.441982746124268,
+      "learning_rate": 9.993515047108813e-06,
+      "loss": 1.2615,
+      "step": 594
+    },
+    {
+      "epoch": 0.045868023435090964,
+      "grad_norm": 4.005769729614258,
+      "learning_rate": 9.993451327226955e-06,
+      "loss": 1.1988,
+      "step": 595
+    },
+    {
+      "epoch": 0.04594511255010793,
+      "grad_norm": 4.6932454109191895,
+      "learning_rate": 9.99338729602927e-06,
+      "loss": 1.1599,
+      "step": 596
+    },
+    {
+      "epoch": 0.046022201665124884,
+      "grad_norm": 4.53994083404541,
+      "learning_rate": 9.993322953519749e-06,
+      "loss": 1.1581,
+      "step": 597
+    },
+    {
+      "epoch": 0.04609929078014184,
+      "grad_norm": 4.505131721496582,
+      "learning_rate": 9.9932582997024e-06,
+      "loss": 1.2383,
+      "step": 598
+    },
+    {
+      "epoch": 0.046176379895158805,
+      "grad_norm": 4.228984355926514,
+      "learning_rate": 9.993193334581259e-06,
+      "loss": 1.3035,
+      "step": 599
+    },
+    {
+      "epoch": 0.04625346901017576,
+      "grad_norm": 4.068345546722412,
+      "learning_rate": 9.993128058160373e-06,
+      "loss": 1.1668,
+      "step": 600
+    },
+    {
+      "epoch": 0.046330558125192725,
+      "grad_norm": 4.136154651641846,
+      "learning_rate": 9.993062470443814e-06,
+      "loss": 1.2611,
+      "step": 601
+    },
+    {
+      "epoch": 0.04640764724020968,
+      "grad_norm": 3.893854856491089,
+      "learning_rate": 9.992996571435668e-06,
+      "loss": 1.0874,
+      "step": 602
+    },
+    {
+      "epoch": 0.046484736355226645,
+      "grad_norm": 3.9865612983703613,
+      "learning_rate": 9.992930361140045e-06,
+      "loss": 1.1812,
+      "step": 603
+    },
+    {
+      "epoch": 0.0465618254702436,
+      "grad_norm": 4.678125381469727,
+      "learning_rate": 9.992863839561076e-06,
+      "loss": 1.2949,
+      "step": 604
+    },
+    {
+      "epoch": 0.04663891458526056,
+      "grad_norm": 4.224588871002197,
+      "learning_rate": 9.992797006702902e-06,
+      "loss": 1.1542,
+      "step": 605
+    },
+    {
+      "epoch": 0.04671600370027752,
+      "grad_norm": 4.713046073913574,
+      "learning_rate": 9.992729862569694e-06,
+      "loss": 1.2428,
+      "step": 606
+    },
+    {
+      "epoch": 0.04679309281529448,
+      "grad_norm": 4.193148612976074,
+      "learning_rate": 9.992662407165637e-06,
+      "loss": 1.2829,
+      "step": 607
+    },
+    {
+      "epoch": 0.04687018193031144,
+      "grad_norm": 4.588696479797363,
+      "learning_rate": 9.992594640494937e-06,
+      "loss": 1.3168,
+      "step": 608
+    },
+    {
+      "epoch": 0.0469472710453284,
+      "grad_norm": 3.988802671432495,
+      "learning_rate": 9.99252656256182e-06,
+      "loss": 1.1058,
+      "step": 609
+    },
+    {
+      "epoch": 0.047024360160345356,
+      "grad_norm": 4.400096893310547,
+      "learning_rate": 9.992458173370525e-06,
+      "loss": 1.3421,
+      "step": 610
+    },
+    {
+      "epoch": 0.04710144927536232,
+      "grad_norm": 4.6701483726501465,
+      "learning_rate": 9.992389472925323e-06,
+      "loss": 1.2688,
+      "step": 611
+    },
+    {
+      "epoch": 0.04717853839037928,
+      "grad_norm": 4.002415180206299,
+      "learning_rate": 9.99232046123049e-06,
+      "loss": 1.163,
+      "step": 612
+    },
+    {
+      "epoch": 0.04725562750539624,
+      "grad_norm": 4.165858268737793,
+      "learning_rate": 9.992251138290336e-06,
+      "loss": 1.157,
+      "step": 613
+    },
+    {
+      "epoch": 0.0473327166204132,
+      "grad_norm": 4.390151023864746,
+      "learning_rate": 9.992181504109177e-06,
+      "loss": 1.147,
+      "step": 614
+    },
+    {
+      "epoch": 0.04740980573543016,
+      "grad_norm": 4.475891590118408,
+      "learning_rate": 9.992111558691357e-06,
+      "loss": 1.1861,
+      "step": 615
+    },
+    {
+      "epoch": 0.04748689485044712,
+      "grad_norm": 4.247600555419922,
+      "learning_rate": 9.992041302041238e-06,
+      "loss": 1.0589,
+      "step": 616
+    },
+    {
+      "epoch": 0.047563983965464074,
+      "grad_norm": 4.422875881195068,
+      "learning_rate": 9.991970734163195e-06,
+      "loss": 1.1923,
+      "step": 617
+    },
+    {
+      "epoch": 0.04764107308048104,
+      "grad_norm": 4.901252746582031,
+      "learning_rate": 9.991899855061633e-06,
+      "loss": 1.1754,
+      "step": 618
+    },
+    {
+      "epoch": 0.047718162195497994,
+      "grad_norm": 4.342195510864258,
+      "learning_rate": 9.991828664740969e-06,
+      "loss": 1.1996,
+      "step": 619
+    },
+    {
+      "epoch": 0.04779525131051496,
+      "grad_norm": 4.116771697998047,
+      "learning_rate": 9.991757163205638e-06,
+      "loss": 1.1596,
+      "step": 620
+    },
+    {
+      "epoch": 0.047872340425531915,
+      "grad_norm": 4.246543884277344,
+      "learning_rate": 9.991685350460102e-06,
+      "loss": 1.2953,
+      "step": 621
+    },
+    {
+      "epoch": 0.04794942954054887,
+      "grad_norm": 4.528024673461914,
+      "learning_rate": 9.991613226508838e-06,
+      "loss": 1.1836,
+      "step": 622
+    },
+    {
+      "epoch": 0.048026518655565835,
+      "grad_norm": 4.674347877502441,
+      "learning_rate": 9.991540791356342e-06,
+      "loss": 1.2399,
+      "step": 623
+    },
+    {
+      "epoch": 0.04810360777058279,
+      "grad_norm": 3.99076247215271,
+      "learning_rate": 9.99146804500713e-06,
+      "loss": 1.1949,
+      "step": 624
+    },
+    {
+      "epoch": 0.048180696885599755,
+      "grad_norm": 4.242091178894043,
+      "learning_rate": 9.991394987465734e-06,
+      "loss": 1.2284,
+      "step": 625
+    },
+    {
+      "epoch": 0.04825778600061671,
+      "grad_norm": 4.19851016998291,
+      "learning_rate": 9.991321618736713e-06,
+      "loss": 1.2451,
+      "step": 626
+    },
+    {
+      "epoch": 0.048334875115633676,
+      "grad_norm": 4.685157299041748,
+      "learning_rate": 9.991247938824641e-06,
+      "loss": 1.1317,
+      "step": 627
+    },
+    {
+      "epoch": 0.04841196423065063,
+      "grad_norm": 3.7727530002593994,
+      "learning_rate": 9.991173947734109e-06,
+      "loss": 1.1234,
+      "step": 628
+    },
+    {
+      "epoch": 0.04848905334566759,
+      "grad_norm": 4.488282203674316,
+      "learning_rate": 9.991099645469733e-06,
+      "loss": 1.1491,
+      "step": 629
+    },
+    {
+      "epoch": 0.04856614246068455,
+      "grad_norm": 4.348680019378662,
+      "learning_rate": 9.991025032036141e-06,
+      "loss": 1.273,
+      "step": 630
+    },
+    {
+      "epoch": 0.04864323157570151,
+      "grad_norm": 4.021731853485107,
+      "learning_rate": 9.990950107437989e-06,
+      "loss": 1.092,
+      "step": 631
+    },
+    {
+      "epoch": 0.04872032069071847,
+      "grad_norm": 3.9224579334259033,
+      "learning_rate": 9.990874871679948e-06,
+      "loss": 1.1298,
+      "step": 632
+    },
+    {
+      "epoch": 0.04879740980573543,
+      "grad_norm": 4.388615608215332,
+      "learning_rate": 9.990799324766704e-06,
+      "loss": 1.3351,
+      "step": 633
+    },
+    {
+      "epoch": 0.048874498920752386,
+      "grad_norm": 4.323206901550293,
+      "learning_rate": 9.990723466702972e-06,
+      "loss": 1.3192,
+      "step": 634
+    },
+    {
+      "epoch": 0.04895158803576935,
+      "grad_norm": 4.259064674377441,
+      "learning_rate": 9.99064729749348e-06,
+      "loss": 1.2824,
+      "step": 635
+    },
+    {
+      "epoch": 0.04902867715078631,
+      "grad_norm": 4.554323673248291,
+      "learning_rate": 9.990570817142974e-06,
+      "loss": 1.302,
+      "step": 636
+    },
+    {
+      "epoch": 0.04910576626580327,
+      "grad_norm": 4.136668682098389,
+      "learning_rate": 9.990494025656227e-06,
+      "loss": 1.2416,
+      "step": 637
+    },
+    {
+      "epoch": 0.04918285538082023,
+      "grad_norm": 4.030932426452637,
+      "learning_rate": 9.990416923038022e-06,
+      "loss": 1.2393,
+      "step": 638
+    },
+    {
+      "epoch": 0.04925994449583719,
+      "grad_norm": 4.34704065322876,
+      "learning_rate": 9.990339509293169e-06,
+      "loss": 1.175,
+      "step": 639
+    },
+    {
+      "epoch": 0.04933703361085415,
+      "grad_norm": 4.10532808303833,
+      "learning_rate": 9.990261784426494e-06,
+      "loss": 1.0991,
+      "step": 640
+    },
+    {
+      "epoch": 0.049414122725871104,
+      "grad_norm": 4.257048606872559,
+      "learning_rate": 9.99018374844284e-06,
+      "loss": 1.2731,
+      "step": 641
+    },
+    {
+      "epoch": 0.04949121184088807,
+      "grad_norm": 4.365378379821777,
+      "learning_rate": 9.990105401347075e-06,
+      "loss": 1.1691,
+      "step": 642
+    },
+    {
+      "epoch": 0.049568300955905024,
+      "grad_norm": 5.146248817443848,
+      "learning_rate": 9.990026743144084e-06,
+      "loss": 1.1753,
+      "step": 643
+    },
+    {
+      "epoch": 0.04964539007092199,
+      "grad_norm": 4.770695686340332,
+      "learning_rate": 9.989947773838766e-06,
+      "loss": 1.3062,
+      "step": 644
+    },
+    {
+      "epoch": 0.049722479185938945,
+      "grad_norm": 4.065369606018066,
+      "learning_rate": 9.989868493436052e-06,
+      "loss": 1.1505,
+      "step": 645
+    },
+    {
+      "epoch": 0.04979956830095591,
+      "grad_norm": 3.940561294555664,
+      "learning_rate": 9.989788901940878e-06,
+      "loss": 1.1703,
+      "step": 646
+    },
+    {
+      "epoch": 0.049876657415972865,
+      "grad_norm": 4.689897537231445,
+      "learning_rate": 9.98970899935821e-06,
+      "loss": 1.2645,
+      "step": 647
+    },
+    {
+      "epoch": 0.04995374653098982,
+      "grad_norm": 3.9842276573181152,
+      "learning_rate": 9.989628785693027e-06,
+      "loss": 1.2856,
+      "step": 648
+    },
+    {
+      "epoch": 0.050030835646006785,
+      "grad_norm": 4.0935773849487305,
+      "learning_rate": 9.989548260950332e-06,
+      "loss": 1.2049,
+      "step": 649
+    },
+    {
+      "epoch": 0.05010792476102374,
+      "grad_norm": 4.526483058929443,
+      "learning_rate": 9.989467425135145e-06,
+      "loss": 1.2788,
+      "step": 650
+    },
+    {
+      "epoch": 0.050185013876040706,
+      "grad_norm": 4.23220682144165,
+      "learning_rate": 9.989386278252505e-06,
+      "loss": 1.2215,
+      "step": 651
+    },
+    {
+      "epoch": 0.05026210299105766,
+      "grad_norm": 4.511578559875488,
+      "learning_rate": 9.989304820307469e-06,
+      "loss": 1.2454,
+      "step": 652
+    },
+    {
+      "epoch": 0.05033919210607462,
+      "grad_norm": 3.9522716999053955,
+      "learning_rate": 9.98922305130512e-06,
+      "loss": 1.1437,
+      "step": 653
+    },
+    {
+      "epoch": 0.05041628122109158,
+      "grad_norm": 4.068902492523193,
+      "learning_rate": 9.989140971250553e-06,
+      "loss": 1.1963,
+      "step": 654
+    },
+    {
+      "epoch": 0.05049337033610854,
+      "grad_norm": 4.930382251739502,
+      "learning_rate": 9.989058580148885e-06,
+      "loss": 1.2457,
+      "step": 655
+    },
+    {
+      "epoch": 0.0505704594511255,
+      "grad_norm": 3.9273664951324463,
+      "learning_rate": 9.988975878005256e-06,
+      "loss": 1.1408,
+      "step": 656
+    },
+    {
+      "epoch": 0.05064754856614246,
+      "grad_norm": 4.123290061950684,
+      "learning_rate": 9.988892864824817e-06,
+      "loss": 1.2368,
+      "step": 657
+    },
+    {
+      "epoch": 0.050724637681159424,
+      "grad_norm": 3.8407251834869385,
+      "learning_rate": 9.988809540612747e-06,
+      "loss": 1.2847,
+      "step": 658
+    },
+    {
+      "epoch": 0.05080172679617638,
+      "grad_norm": 4.222978115081787,
+      "learning_rate": 9.988725905374242e-06,
+      "loss": 1.1422,
+      "step": 659
+    },
+    {
+      "epoch": 0.05087881591119334,
+      "grad_norm": 4.184953689575195,
+      "learning_rate": 9.988641959114512e-06,
+      "loss": 1.1177,
+      "step": 660
+    },
+    {
+      "epoch": 0.0509559050262103,
+      "grad_norm": 4.532317161560059,
+      "learning_rate": 9.988557701838791e-06,
+      "loss": 1.2804,
+      "step": 661
+    },
+    {
+      "epoch": 0.05103299414122726,
+      "grad_norm": 4.091014385223389,
+      "learning_rate": 9.988473133552336e-06,
+      "loss": 1.1743,
+      "step": 662
+    },
+    {
+      "epoch": 0.05111008325624422,
+      "grad_norm": 4.535579204559326,
+      "learning_rate": 9.988388254260417e-06,
+      "loss": 1.161,
+      "step": 663
+    },
+    {
+      "epoch": 0.05118717237126118,
+      "grad_norm": 4.38757848739624,
+      "learning_rate": 9.988303063968325e-06,
+      "loss": 1.2166,
+      "step": 664
+    },
+    {
+      "epoch": 0.051264261486278134,
+      "grad_norm": 4.727634906768799,
+      "learning_rate": 9.988217562681373e-06,
+      "loss": 1.1831,
+      "step": 665
+    },
+    {
+      "epoch": 0.0513413506012951,
+      "grad_norm": 4.283156871795654,
+      "learning_rate": 9.988131750404889e-06,
+      "loss": 1.2051,
+      "step": 666
+    },
+    {
+      "epoch": 0.051418439716312055,
+      "grad_norm": 4.735593318939209,
+      "learning_rate": 9.988045627144226e-06,
+      "loss": 1.2458,
+      "step": 667
+    },
+    {
+      "epoch": 0.05149552883132902,
+      "grad_norm": 3.9615490436553955,
+      "learning_rate": 9.98795919290475e-06,
+      "loss": 1.3039,
+      "step": 668
+    },
+    {
+      "epoch": 0.051572617946345975,
+      "grad_norm": 4.121702671051025,
+      "learning_rate": 9.987872447691853e-06,
+      "loss": 1.1451,
+      "step": 669
+    },
+    {
+      "epoch": 0.05164970706136294,
+      "grad_norm": 4.138181209564209,
+      "learning_rate": 9.987785391510943e-06,
+      "loss": 1.2745,
+      "step": 670
+    },
+    {
+      "epoch": 0.051726796176379895,
+      "grad_norm": 4.431112289428711,
+      "learning_rate": 9.987698024367444e-06,
+      "loss": 1.2789,
+      "step": 671
+    },
+    {
+      "epoch": 0.05180388529139685,
+      "grad_norm": 4.268251895904541,
+      "learning_rate": 9.987610346266808e-06,
+      "loss": 1.128,
+      "step": 672
+    },
+    {
+      "epoch": 0.051880974406413816,
+      "grad_norm": 4.024346351623535,
+      "learning_rate": 9.987522357214496e-06,
+      "loss": 1.1851,
+      "step": 673
+    },
+    {
+      "epoch": 0.05195806352143077,
+      "grad_norm": 4.691232681274414,
+      "learning_rate": 9.987434057215996e-06,
+      "loss": 1.1774,
+      "step": 674
+    },
+    {
+      "epoch": 0.052035152636447736,
+      "grad_norm": 4.2596635818481445,
+      "learning_rate": 9.987345446276815e-06,
+      "loss": 1.1782,
+      "step": 675
+    },
+    {
+      "epoch": 0.05211224175146469,
+      "grad_norm": 4.4914231300354,
+      "learning_rate": 9.987256524402473e-06,
+      "loss": 1.3086,
+      "step": 676
+    },
+    {
+      "epoch": 0.05218933086648165,
+      "grad_norm": 4.257656097412109,
+      "learning_rate": 9.987167291598518e-06,
+      "loss": 1.2465,
+      "step": 677
+    },
+    {
+      "epoch": 0.05226641998149861,
+      "grad_norm": 4.148374080657959,
+      "learning_rate": 9.987077747870512e-06,
+      "loss": 1.2106,
+      "step": 678
+    },
+    {
+      "epoch": 0.05234350909651557,
+      "grad_norm": 4.15828275680542,
+      "learning_rate": 9.986987893224036e-06,
+      "loss": 1.2535,
+      "step": 679
+    },
+    {
+      "epoch": 0.05242059821153253,
+      "grad_norm": 4.436988353729248,
+      "learning_rate": 9.986897727664693e-06,
+      "loss": 1.2644,
+      "step": 680
+    },
+    {
+      "epoch": 0.05249768732654949,
+      "grad_norm": 4.664417266845703,
+      "learning_rate": 9.986807251198105e-06,
+      "loss": 1.3725,
+      "step": 681
+    },
+    {
+      "epoch": 0.052574776441566454,
+      "grad_norm": 4.027350425720215,
+      "learning_rate": 9.986716463829913e-06,
+      "loss": 1.2051,
+      "step": 682
+    },
+    {
+      "epoch": 0.05265186555658341,
+      "grad_norm": 4.5459818840026855,
+      "learning_rate": 9.986625365565776e-06,
+      "loss": 1.3047,
+      "step": 683
+    },
+    {
+      "epoch": 0.05272895467160037,
+      "grad_norm": 4.167656898498535,
+      "learning_rate": 9.986533956411373e-06,
+      "loss": 1.1959,
+      "step": 684
+    },
+    {
+      "epoch": 0.05280604378661733,
+      "grad_norm": 4.064361095428467,
+      "learning_rate": 9.986442236372404e-06,
+      "loss": 1.1627,
+      "step": 685
+    },
+    {
+      "epoch": 0.05288313290163429,
+      "grad_norm": 4.358853816986084,
+      "learning_rate": 9.986350205454587e-06,
+      "loss": 1.1343,
+      "step": 686
+    },
+    {
+      "epoch": 0.05296022201665125,
+      "grad_norm": 3.9868597984313965,
+      "learning_rate": 9.98625786366366e-06,
+      "loss": 1.1268,
+      "step": 687
+    },
+    {
+      "epoch": 0.05303731113166821,
+      "grad_norm": 3.7771265506744385,
+      "learning_rate": 9.98616521100538e-06,
+      "loss": 1.1284,
+      "step": 688
+    },
+    {
+      "epoch": 0.05311440024668517,
+      "grad_norm": 4.3471574783325195,
+      "learning_rate": 9.986072247485523e-06,
+      "loss": 1.2269,
+      "step": 689
+    },
+    {
+      "epoch": 0.05319148936170213,
+      "grad_norm": 3.956973075866699,
+      "learning_rate": 9.985978973109885e-06,
+      "loss": 1.2157,
+      "step": 690
+    },
+    {
+      "epoch": 0.053268578476719085,
+      "grad_norm": 4.183494567871094,
+      "learning_rate": 9.98588538788428e-06,
+      "loss": 1.2887,
+      "step": 691
+    },
+    {
+      "epoch": 0.05334566759173605,
+      "grad_norm": 4.123586654663086,
+      "learning_rate": 9.985791491814545e-06,
+      "loss": 1.1641,
+      "step": 692
+    },
+    {
+      "epoch": 0.053422756706753005,
+      "grad_norm": 4.9690680503845215,
+      "learning_rate": 9.985697284906532e-06,
+      "loss": 1.221,
+      "step": 693
+    },
+    {
+      "epoch": 0.05349984582176997,
+      "grad_norm": 3.8592402935028076,
+      "learning_rate": 9.985602767166114e-06,
+      "loss": 1.2379,
+      "step": 694
+    },
+    {
+      "epoch": 0.053576934936786925,
+      "grad_norm": 4.504653453826904,
+      "learning_rate": 9.985507938599186e-06,
+      "loss": 1.2235,
+      "step": 695
+    },
+    {
+      "epoch": 0.05365402405180388,
+      "grad_norm": 4.732753753662109,
+      "learning_rate": 9.985412799211658e-06,
+      "loss": 1.2707,
+      "step": 696
+    },
+    {
+      "epoch": 0.053731113166820846,
+      "grad_norm": 3.8671162128448486,
+      "learning_rate": 9.98531734900946e-06,
+      "loss": 1.282,
+      "step": 697
+    },
+    {
+      "epoch": 0.0538082022818378,
+      "grad_norm": 4.004355430603027,
+      "learning_rate": 9.985221587998549e-06,
+      "loss": 1.1439,
+      "step": 698
+    },
+    {
+      "epoch": 0.053885291396854766,
+      "grad_norm": 4.045041084289551,
+      "learning_rate": 9.985125516184889e-06,
+      "loss": 1.1859,
+      "step": 699
+    },
+    {
+      "epoch": 0.05396238051187172,
+      "grad_norm": 4.078817367553711,
+      "learning_rate": 9.985029133574471e-06,
+      "loss": 1.231,
+      "step": 700
+    },
+    {
+      "epoch": 0.054039469626888686,
+      "grad_norm": 4.425305366516113,
+      "learning_rate": 9.984932440173304e-06,
+      "loss": 1.2619,
+      "step": 701
+    },
+    {
+      "epoch": 0.05411655874190564,
+      "grad_norm": 4.202983856201172,
+      "learning_rate": 9.98483543598742e-06,
+      "loss": 1.1445,
+      "step": 702
+    },
+    {
+      "epoch": 0.0541936478569226,
+      "grad_norm": 4.361485958099365,
+      "learning_rate": 9.984738121022861e-06,
+      "loss": 1.2118,
+      "step": 703
+    },
+    {
+      "epoch": 0.054270736971939564,
+      "grad_norm": 4.436924934387207,
+      "learning_rate": 9.9846404952857e-06,
+      "loss": 1.3065,
+      "step": 704
+    },
+    {
+      "epoch": 0.05434782608695652,
+      "grad_norm": 4.535494804382324,
+      "learning_rate": 9.984542558782018e-06,
+      "loss": 1.2161,
+      "step": 705
+    },
+    {
+      "epoch": 0.054424915201973484,
+      "grad_norm": 4.041503429412842,
+      "learning_rate": 9.984444311517923e-06,
+      "loss": 1.2332,
+      "step": 706
+    },
+    {
+      "epoch": 0.05450200431699044,
+      "grad_norm": 4.140787601470947,
+      "learning_rate": 9.98434575349954e-06,
+      "loss": 1.2384,
+      "step": 707
+    },
+    {
+      "epoch": 0.0545790934320074,
+      "grad_norm": 4.458608627319336,
+      "learning_rate": 9.984246884733014e-06,
+      "loss": 1.1531,
+      "step": 708
+    },
+    {
+      "epoch": 0.05465618254702436,
+      "grad_norm": 4.457377910614014,
+      "learning_rate": 9.984147705224509e-06,
+      "loss": 1.1315,
+      "step": 709
+    },
+    {
+      "epoch": 0.05473327166204132,
+      "grad_norm": 4.368088722229004,
+      "learning_rate": 9.984048214980209e-06,
+      "loss": 1.2896,
+      "step": 710
+    },
+    {
+      "epoch": 0.05481036077705828,
+      "grad_norm": 4.405154228210449,
+      "learning_rate": 9.983948414006315e-06,
+      "loss": 1.2627,
+      "step": 711
+    },
+    {
+      "epoch": 0.05488744989207524,
+      "grad_norm": 4.391892433166504,
+      "learning_rate": 9.98384830230905e-06,
+      "loss": 1.2523,
+      "step": 712
+    },
+    {
+      "epoch": 0.0549645390070922,
+      "grad_norm": 4.066841125488281,
+      "learning_rate": 9.983747879894655e-06,
+      "loss": 1.3027,
+      "step": 713
+    },
+    {
+      "epoch": 0.05504162812210916,
+      "grad_norm": 3.8345208168029785,
+      "learning_rate": 9.98364714676939e-06,
+      "loss": 1.2225,
+      "step": 714
+    },
+    {
+      "epoch": 0.055118717237126115,
+      "grad_norm": 3.9913508892059326,
+      "learning_rate": 9.983546102939539e-06,
+      "loss": 1.2512,
+      "step": 715
+    },
+    {
+      "epoch": 0.05519580635214308,
+      "grad_norm": 4.175078868865967,
+      "learning_rate": 9.983444748411399e-06,
+      "loss": 1.092,
+      "step": 716
+    },
+    {
+      "epoch": 0.055272895467160035,
+      "grad_norm": 4.498879909515381,
+      "learning_rate": 9.983343083191287e-06,
+      "loss": 1.2141,
+      "step": 717
+    },
+    {
+      "epoch": 0.055349984582177,
+      "grad_norm": 4.295947551727295,
+      "learning_rate": 9.983241107285544e-06,
+      "loss": 1.2172,
+      "step": 718
+    },
+    {
+      "epoch": 0.055427073697193956,
+      "grad_norm": 4.796656131744385,
+      "learning_rate": 9.983138820700526e-06,
+      "loss": 1.0474,
+      "step": 719
+    },
+    {
+      "epoch": 0.05550416281221091,
+      "grad_norm": 4.575197219848633,
+      "learning_rate": 9.983036223442612e-06,
+      "loss": 1.2008,
+      "step": 720
+    },
+    {
+      "epoch": 0.055581251927227876,
+      "grad_norm": 4.701834201812744,
+      "learning_rate": 9.982933315518197e-06,
+      "loss": 1.3672,
+      "step": 721
+    },
+    {
+      "epoch": 0.05565834104224483,
+      "grad_norm": 4.1152753829956055,
+      "learning_rate": 9.982830096933697e-06,
+      "loss": 1.1963,
+      "step": 722
+    },
+    {
+      "epoch": 0.055735430157261796,
+      "grad_norm": 3.986194372177124,
+      "learning_rate": 9.982726567695547e-06,
+      "loss": 1.2208,
+      "step": 723
+    },
+    {
+      "epoch": 0.05581251927227875,
+      "grad_norm": 4.519706726074219,
+      "learning_rate": 9.982622727810202e-06,
+      "loss": 1.2464,
+      "step": 724
+    },
+    {
+      "epoch": 0.05588960838729572,
+      "grad_norm": 4.6598992347717285,
+      "learning_rate": 9.982518577284135e-06,
+      "loss": 1.2623,
+      "step": 725
+    },
+    {
+      "epoch": 0.05596669750231267,
+      "grad_norm": 4.343133926391602,
+      "learning_rate": 9.98241411612384e-06,
+      "loss": 1.302,
+      "step": 726
+    },
+    {
+      "epoch": 0.05604378661732963,
+      "grad_norm": 4.320357322692871,
+      "learning_rate": 9.98230934433583e-06,
+      "loss": 1.2826,
+      "step": 727
+    },
+    {
+      "epoch": 0.056120875732346594,
+      "grad_norm": 3.914271116256714,
+      "learning_rate": 9.982204261926636e-06,
+      "loss": 1.0489,
+      "step": 728
+    },
+    {
+      "epoch": 0.05619796484736355,
+      "grad_norm": 4.293032169342041,
+      "learning_rate": 9.98209886890281e-06,
+      "loss": 1.0239,
+      "step": 729
+    },
+    {
+      "epoch": 0.056275053962380514,
+      "grad_norm": 4.971100807189941,
+      "learning_rate": 9.981993165270922e-06,
+      "loss": 1.3301,
+      "step": 730
+    },
+    {
+      "epoch": 0.05635214307739747,
+      "grad_norm": 4.393875598907471,
+      "learning_rate": 9.981887151037563e-06,
+      "loss": 1.1637,
+      "step": 731
+    },
+    {
+      "epoch": 0.056429232192414434,
+      "grad_norm": 4.458146095275879,
+      "learning_rate": 9.981780826209342e-06,
+      "loss": 1.2018,
+      "step": 732
+    },
+    {
+      "epoch": 0.05650632130743139,
+      "grad_norm": 4.671881198883057,
+      "learning_rate": 9.981674190792887e-06,
+      "loss": 1.2186,
+      "step": 733
+    },
+    {
+      "epoch": 0.05658341042244835,
+      "grad_norm": 4.1135029792785645,
+      "learning_rate": 9.98156724479485e-06,
+      "loss": 1.2273,
+      "step": 734
+    },
+    {
+      "epoch": 0.05666049953746531,
+      "grad_norm": 4.136716842651367,
+      "learning_rate": 9.981459988221891e-06,
+      "loss": 1.2256,
+      "step": 735
+    },
+    {
+      "epoch": 0.05673758865248227,
+      "grad_norm": 4.991126537322998,
+      "learning_rate": 9.981352421080704e-06,
+      "loss": 1.1702,
+      "step": 736
+    },
+    {
+      "epoch": 0.05681467776749923,
+      "grad_norm": 3.9331271648406982,
+      "learning_rate": 9.981244543377992e-06,
+      "loss": 1.0802,
+      "step": 737
+    },
+    {
+      "epoch": 0.05689176688251619,
+      "grad_norm": 4.251241207122803,
+      "learning_rate": 9.981136355120483e-06,
+      "loss": 1.2407,
+      "step": 738
+    },
+    {
+      "epoch": 0.056968855997533145,
+      "grad_norm": 4.525993347167969,
+      "learning_rate": 9.981027856314918e-06,
+      "loss": 1.1908,
+      "step": 739
+    },
+    {
+      "epoch": 0.05704594511255011,
+      "grad_norm": 4.329502105712891,
+      "learning_rate": 9.980919046968063e-06,
+      "loss": 1.3154,
+      "step": 740
+    },
+    {
+      "epoch": 0.057123034227567066,
+      "grad_norm": 4.49136209487915,
+      "learning_rate": 9.980809927086704e-06,
+      "loss": 1.3037,
+      "step": 741
+    },
+    {
+      "epoch": 0.05720012334258403,
+      "grad_norm": 4.0258941650390625,
+      "learning_rate": 9.980700496677643e-06,
+      "loss": 1.1127,
+      "step": 742
+    },
+    {
+      "epoch": 0.057277212457600986,
+      "grad_norm": 4.217193126678467,
+      "learning_rate": 9.980590755747698e-06,
+      "loss": 1.1301,
+      "step": 743
+    },
+    {
+      "epoch": 0.05735430157261795,
+      "grad_norm": 4.262319087982178,
+      "learning_rate": 9.980480704303719e-06,
+      "loss": 1.2313,
+      "step": 744
+    },
+    {
+      "epoch": 0.057431390687634906,
+      "grad_norm": 4.561182975769043,
+      "learning_rate": 9.980370342352562e-06,
+      "loss": 1.2237,
+      "step": 745
+    },
+    {
+      "epoch": 0.05750847980265186,
+      "grad_norm": 4.379204750061035,
+      "learning_rate": 9.980259669901105e-06,
+      "loss": 1.2885,
+      "step": 746
+    },
+    {
+      "epoch": 0.057585568917668827,
+      "grad_norm": 4.246951103210449,
+      "learning_rate": 9.980148686956252e-06,
+      "loss": 1.2487,
+      "step": 747
+    },
+    {
+      "epoch": 0.05766265803268578,
+      "grad_norm": 4.429943084716797,
+      "learning_rate": 9.980037393524925e-06,
+      "loss": 1.2902,
+      "step": 748
+    },
+    {
+      "epoch": 0.05773974714770275,
+      "grad_norm": 4.189937114715576,
+      "learning_rate": 9.979925789614053e-06,
+      "loss": 1.1855,
+      "step": 749
+    },
+    {
+      "epoch": 0.057816836262719704,
+      "grad_norm": 4.239128589630127,
+      "learning_rate": 9.979813875230604e-06,
+      "loss": 1.3016,
+      "step": 750
+    },
+    {
+      "epoch": 0.05789392537773666,
+      "grad_norm": 4.345081329345703,
+      "learning_rate": 9.97970165038155e-06,
+      "loss": 1.3184,
+      "step": 751
+    },
+    {
+      "epoch": 0.057971014492753624,
+      "grad_norm": 4.280864238739014,
+      "learning_rate": 9.979589115073888e-06,
+      "loss": 1.3088,
+      "step": 752
+    },
+    {
+      "epoch": 0.05804810360777058,
+      "grad_norm": 4.612633228302002,
+      "learning_rate": 9.979476269314635e-06,
+      "loss": 1.125,
+      "step": 753
+    },
+    {
+      "epoch": 0.058125192722787544,
+      "grad_norm": 4.349769115447998,
+      "learning_rate": 9.979363113110825e-06,
+      "loss": 1.2171,
+      "step": 754
+    },
+    {
+      "epoch": 0.0582022818378045,
+      "grad_norm": 4.421850681304932,
+      "learning_rate": 9.979249646469516e-06,
+      "loss": 1.239,
+      "step": 755
+    },
+    {
+      "epoch": 0.058279370952821465,
+      "grad_norm": 3.841928482055664,
+      "learning_rate": 9.979135869397776e-06,
+      "loss": 1.1576,
+      "step": 756
+    },
+    {
+      "epoch": 0.05835646006783842,
+      "grad_norm": 4.535758018493652,
+      "learning_rate": 9.979021781902705e-06,
+      "loss": 1.2456,
+      "step": 757
+    },
+    {
+      "epoch": 0.05843354918285538,
+      "grad_norm": 4.633460998535156,
+      "learning_rate": 9.978907383991412e-06,
+      "loss": 1.1629,
+      "step": 758
+    },
+    {
+      "epoch": 0.05851063829787234,
+      "grad_norm": 4.196053504943848,
+      "learning_rate": 9.978792675671031e-06,
+      "loss": 1.1678,
+      "step": 759
+    },
+    {
+      "epoch": 0.0585877274128893,
+      "grad_norm": 4.235719203948975,
+      "learning_rate": 9.978677656948712e-06,
+      "loss": 1.135,
+      "step": 760
+    },
+    {
+      "epoch": 0.05866481652790626,
+      "grad_norm": 3.9202029705047607,
+      "learning_rate": 9.978562327831626e-06,
+      "loss": 1.1177,
+      "step": 761
+    },
+    {
+      "epoch": 0.05874190564292322,
+      "grad_norm": 4.615420341491699,
+      "learning_rate": 9.978446688326964e-06,
+      "loss": 1.2012,
+      "step": 762
+    },
+    {
+      "epoch": 0.05881899475794018,
+      "grad_norm": 4.768507480621338,
+      "learning_rate": 9.978330738441935e-06,
+      "loss": 1.3307,
+      "step": 763
+    },
+    {
+      "epoch": 0.05889608387295714,
+      "grad_norm": 4.350339412689209,
+      "learning_rate": 9.978214478183767e-06,
+      "loss": 1.127,
+      "step": 764
+    },
+    {
+      "epoch": 0.058973172987974096,
+      "grad_norm": 4.4393134117126465,
+      "learning_rate": 9.97809790755971e-06,
+      "loss": 1.2217,
+      "step": 765
+    },
+    {
+      "epoch": 0.05905026210299106,
+      "grad_norm": 4.164086818695068,
+      "learning_rate": 9.97798102657703e-06,
+      "loss": 1.1628,
+      "step": 766
+    },
+    {
+      "epoch": 0.059127351218008016,
+      "grad_norm": 4.689718723297119,
+      "learning_rate": 9.977863835243016e-06,
+      "loss": 1.198,
+      "step": 767
+    },
+    {
+      "epoch": 0.05920444033302498,
+      "grad_norm": 4.145715236663818,
+      "learning_rate": 9.97774633356497e-06,
+      "loss": 1.2088,
+      "step": 768
+    },
+    {
+      "epoch": 0.059281529448041936,
+      "grad_norm": 4.122262477874756,
+      "learning_rate": 9.977628521550223e-06,
+      "loss": 1.1543,
+      "step": 769
+    },
+    {
+      "epoch": 0.05935861856305889,
+      "grad_norm": 4.569461822509766,
+      "learning_rate": 9.977510399206118e-06,
+      "loss": 1.1734,
+      "step": 770
+    },
+    {
+      "epoch": 0.05943570767807586,
+      "grad_norm": 4.216443061828613,
+      "learning_rate": 9.977391966540017e-06,
+      "loss": 1.2426,
+      "step": 771
+    },
+    {
+      "epoch": 0.05951279679309281,
+      "grad_norm": 4.43640661239624,
+      "learning_rate": 9.977273223559306e-06,
+      "loss": 1.1938,
+      "step": 772
+    },
+    {
+      "epoch": 0.05958988590810978,
+      "grad_norm": 3.9541783332824707,
+      "learning_rate": 9.977154170271389e-06,
+      "loss": 1.149,
+      "step": 773
+    },
+    {
+      "epoch": 0.059666975023126734,
+      "grad_norm": 4.332737445831299,
+      "learning_rate": 9.977034806683685e-06,
+      "loss": 1.2612,
+      "step": 774
+    },
+    {
+      "epoch": 0.0597440641381437,
+      "grad_norm": 4.155104637145996,
+      "learning_rate": 9.976915132803638e-06,
+      "loss": 1.0815,
+      "step": 775
+    },
+    {
+      "epoch": 0.059821153253160654,
+      "grad_norm": 4.293869495391846,
+      "learning_rate": 9.976795148638707e-06,
+      "loss": 1.2542,
+      "step": 776
+    },
+    {
+      "epoch": 0.05989824236817761,
+      "grad_norm": 4.076062202453613,
+      "learning_rate": 9.976674854196377e-06,
+      "loss": 1.1219,
+      "step": 777
+    },
+    {
+      "epoch": 0.059975331483194574,
+      "grad_norm": 4.0625152587890625,
+      "learning_rate": 9.976554249484144e-06,
+      "loss": 1.1951,
+      "step": 778
+    },
+    {
+      "epoch": 0.06005242059821153,
+      "grad_norm": 4.384305477142334,
+      "learning_rate": 9.976433334509525e-06,
+      "loss": 1.1677,
+      "step": 779
+    },
+    {
+      "epoch": 0.060129509713228495,
+      "grad_norm": 4.4182448387146,
+      "learning_rate": 9.976312109280063e-06,
+      "loss": 1.2527,
+      "step": 780
+    },
+    {
+      "epoch": 0.06020659882824545,
+      "grad_norm": 3.99910044670105,
+      "learning_rate": 9.976190573803314e-06,
+      "loss": 1.1124,
+      "step": 781
+    },
+    {
+      "epoch": 0.06028368794326241,
+      "grad_norm": 4.698251247406006,
+      "learning_rate": 9.976068728086857e-06,
+      "loss": 1.2172,
+      "step": 782
+    },
+    {
+      "epoch": 0.06036077705827937,
+      "grad_norm": 4.009255409240723,
+      "learning_rate": 9.975946572138284e-06,
+      "loss": 1.1481,
+      "step": 783
+    },
+    {
+      "epoch": 0.06043786617329633,
+      "grad_norm": 4.688887596130371,
+      "learning_rate": 9.975824105965215e-06,
+      "loss": 1.1883,
+      "step": 784
+    },
+    {
+      "epoch": 0.06051495528831329,
+      "grad_norm": 4.354454517364502,
+      "learning_rate": 9.975701329575283e-06,
+      "loss": 1.1541,
+      "step": 785
+    },
+    {
+      "epoch": 0.06059204440333025,
+      "grad_norm": 4.520771026611328,
+      "learning_rate": 9.97557824297614e-06,
+      "loss": 1.2542,
+      "step": 786
+    },
+    {
+      "epoch": 0.06066913351834721,
+      "grad_norm": 4.378393650054932,
+      "learning_rate": 9.975454846175466e-06,
+      "loss": 1.2588,
+      "step": 787
+    },
+    {
+      "epoch": 0.06074622263336417,
+      "grad_norm": 3.970188856124878,
+      "learning_rate": 9.975331139180951e-06,
+      "loss": 1.1532,
+      "step": 788
+    },
+    {
+      "epoch": 0.060823311748381126,
+      "grad_norm": 4.00541877746582,
+      "learning_rate": 9.975207122000305e-06,
+      "loss": 1.2816,
+      "step": 789
+    },
+    {
+      "epoch": 0.06090040086339809,
+      "grad_norm": 4.243439674377441,
+      "learning_rate": 9.975082794641264e-06,
+      "loss": 1.1577,
+      "step": 790
+    },
+    {
+      "epoch": 0.060977489978415046,
+      "grad_norm": 4.396401405334473,
+      "learning_rate": 9.974958157111578e-06,
+      "loss": 1.2457,
+      "step": 791
+    },
+    {
+      "epoch": 0.06105457909343201,
+      "grad_norm": 4.4037017822265625,
+      "learning_rate": 9.974833209419016e-06,
+      "loss": 1.2012,
+      "step": 792
+    },
+    {
+      "epoch": 0.06113166820844897,
+      "grad_norm": 4.3235907554626465,
+      "learning_rate": 9.974707951571369e-06,
+      "loss": 1.0912,
+      "step": 793
+    },
+    {
+      "epoch": 0.06120875732346592,
+      "grad_norm": 4.980129718780518,
+      "learning_rate": 9.974582383576446e-06,
+      "loss": 1.1828,
+      "step": 794
+    },
+    {
+      "epoch": 0.06128584643848289,
+      "grad_norm": 4.494161128997803,
+      "learning_rate": 9.974456505442073e-06,
+      "loss": 1.3294,
+      "step": 795
+    },
+    {
+      "epoch": 0.061362935553499844,
+      "grad_norm": 3.907623767852783,
+      "learning_rate": 9.974330317176103e-06,
+      "loss": 1.1584,
+      "step": 796
+    },
+    {
+      "epoch": 0.06144002466851681,
+      "grad_norm": 4.022689342498779,
+      "learning_rate": 9.9742038187864e-06,
+      "loss": 1.1379,
+      "step": 797
+    },
+    {
+      "epoch": 0.061517113783533764,
+      "grad_norm": 4.5095977783203125,
+      "learning_rate": 9.974077010280851e-06,
+      "loss": 1.2845,
+      "step": 798
+    },
+    {
+      "epoch": 0.06159420289855073,
+      "grad_norm": 4.331615447998047,
+      "learning_rate": 9.97394989166736e-06,
+      "loss": 1.1422,
+      "step": 799
+    },
+    {
+      "epoch": 0.061671292013567684,
+      "grad_norm": 4.0961527824401855,
+      "learning_rate": 9.973822462953856e-06,
+      "loss": 1.248,
+      "step": 800
+    },
+    {
+      "epoch": 0.06174838112858464,
+      "grad_norm": 4.432220935821533,
+      "learning_rate": 9.973694724148281e-06,
+      "loss": 1.1686,
+      "step": 801
+    },
+    {
+      "epoch": 0.061825470243601605,
+      "grad_norm": 4.118536949157715,
+      "learning_rate": 9.973566675258598e-06,
+      "loss": 1.1243,
+      "step": 802
+    },
+    {
+      "epoch": 0.06190255935861856,
+      "grad_norm": 3.9236130714416504,
+      "learning_rate": 9.973438316292794e-06,
+      "loss": 1.1906,
+      "step": 803
+    },
+    {
+      "epoch": 0.061979648473635525,
+      "grad_norm": 4.201266288757324,
+      "learning_rate": 9.97330964725887e-06,
+      "loss": 1.109,
+      "step": 804
+    },
+    {
+      "epoch": 0.06205673758865248,
+      "grad_norm": 4.015562534332275,
+      "learning_rate": 9.973180668164844e-06,
+      "loss": 1.1616,
+      "step": 805
+    },
+    {
+      "epoch": 0.062133826703669445,
+      "grad_norm": 5.223426818847656,
+      "learning_rate": 9.97305137901876e-06,
+      "loss": 1.382,
+      "step": 806
+    },
+    {
+      "epoch": 0.0622109158186864,
+      "grad_norm": 4.341038227081299,
+      "learning_rate": 9.972921779828679e-06,
+      "loss": 1.1932,
+      "step": 807
+    },
+    {
+      "epoch": 0.06228800493370336,
+      "grad_norm": 4.124241352081299,
+      "learning_rate": 9.972791870602682e-06,
+      "loss": 1.2678,
+      "step": 808
+    },
+    {
+      "epoch": 0.06236509404872032,
+      "grad_norm": 4.108554363250732,
+      "learning_rate": 9.972661651348865e-06,
+      "loss": 1.2885,
+      "step": 809
+    },
+    {
+      "epoch": 0.06244218316373728,
+      "grad_norm": 4.48992919921875,
+      "learning_rate": 9.972531122075349e-06,
+      "loss": 1.2948,
+      "step": 810
+    },
+    {
+      "epoch": 0.06251927227875424,
+      "grad_norm": 4.17953634262085,
+      "learning_rate": 9.97240028279027e-06,
+      "loss": 1.0968,
+      "step": 811
+    },
+    {
+      "epoch": 0.0625963613937712,
+      "grad_norm": 4.093837738037109,
+      "learning_rate": 9.972269133501787e-06,
+      "loss": 1.1572,
+      "step": 812
+    },
+    {
+      "epoch": 0.06267345050878816,
+      "grad_norm": 4.524899482727051,
+      "learning_rate": 9.972137674218077e-06,
+      "loss": 1.2688,
+      "step": 813
+    },
+    {
+      "epoch": 0.06275053962380511,
+      "grad_norm": 4.406780242919922,
+      "learning_rate": 9.972005904947332e-06,
+      "loss": 1.3075,
+      "step": 814
+    },
+    {
+      "epoch": 0.06282762873882208,
+      "grad_norm": 4.174956798553467,
+      "learning_rate": 9.97187382569777e-06,
+      "loss": 1.1822,
+      "step": 815
+    },
+    {
+      "epoch": 0.06290471785383904,
+      "grad_norm": 3.8772571086883545,
+      "learning_rate": 9.971741436477625e-06,
+      "loss": 1.1562,
+      "step": 816
+    },
+    {
+      "epoch": 0.062981806968856,
+      "grad_norm": 3.9671881198883057,
+      "learning_rate": 9.971608737295151e-06,
+      "loss": 1.2294,
+      "step": 817
+    },
+    {
+      "epoch": 0.06305889608387295,
+      "grad_norm": 4.8057098388671875,
+      "learning_rate": 9.971475728158622e-06,
+      "loss": 1.2987,
+      "step": 818
+    },
+    {
+      "epoch": 0.06313598519888991,
+      "grad_norm": 4.444809913635254,
+      "learning_rate": 9.971342409076328e-06,
+      "loss": 1.2099,
+      "step": 819
+    },
+    {
+      "epoch": 0.06321307431390688,
+      "grad_norm": 4.493791103363037,
+      "learning_rate": 9.971208780056582e-06,
+      "loss": 1.1324,
+      "step": 820
+    },
+    {
+      "epoch": 0.06329016342892384,
+      "grad_norm": 4.225183963775635,
+      "learning_rate": 9.971074841107715e-06,
+      "loss": 1.1474,
+      "step": 821
+    },
+    {
+      "epoch": 0.0633672525439408,
+      "grad_norm": 4.375453948974609,
+      "learning_rate": 9.970940592238077e-06,
+      "loss": 1.0902,
+      "step": 822
+    },
+    {
+      "epoch": 0.06344434165895775,
+      "grad_norm": 4.046267509460449,
+      "learning_rate": 9.97080603345604e-06,
+      "loss": 1.3007,
+      "step": 823
+    },
+    {
+      "epoch": 0.06352143077397472,
+      "grad_norm": 4.696292400360107,
+      "learning_rate": 9.970671164769989e-06,
+      "loss": 1.0967,
+      "step": 824
+    },
+    {
+      "epoch": 0.06359851988899168,
+      "grad_norm": 3.9629440307617188,
+      "learning_rate": 9.970535986188337e-06,
+      "loss": 1.0908,
+      "step": 825
+    },
+    {
+      "epoch": 0.06367560900400863,
+      "grad_norm": 4.310845851898193,
+      "learning_rate": 9.970400497719508e-06,
+      "loss": 1.1777,
+      "step": 826
+    },
+    {
+      "epoch": 0.06375269811902559,
+      "grad_norm": 4.32405948638916,
+      "learning_rate": 9.97026469937195e-06,
+      "loss": 1.2285,
+      "step": 827
+    },
+    {
+      "epoch": 0.06382978723404255,
+      "grad_norm": 4.7154130935668945,
+      "learning_rate": 9.970128591154131e-06,
+      "loss": 1.3087,
+      "step": 828
+    },
+    {
+      "epoch": 0.06390687634905952,
+      "grad_norm": 4.456521034240723,
+      "learning_rate": 9.969992173074534e-06,
+      "loss": 1.2192,
+      "step": 829
+    },
+    {
+      "epoch": 0.06398396546407648,
+      "grad_norm": 3.9812963008880615,
+      "learning_rate": 9.969855445141666e-06,
+      "loss": 1.1708,
+      "step": 830
+    },
+    {
+      "epoch": 0.06406105457909343,
+      "grad_norm": 4.128241539001465,
+      "learning_rate": 9.969718407364051e-06,
+      "loss": 1.0954,
+      "step": 831
+    },
+    {
+      "epoch": 0.06413814369411039,
+      "grad_norm": 4.4097089767456055,
+      "learning_rate": 9.969581059750231e-06,
+      "loss": 1.1696,
+      "step": 832
+    },
+    {
+      "epoch": 0.06421523280912735,
+      "grad_norm": 4.542466163635254,
+      "learning_rate": 9.96944340230877e-06,
+      "loss": 1.3143,
+      "step": 833
+    },
+    {
+      "epoch": 0.06429232192414432,
+      "grad_norm": 3.86757755279541,
+      "learning_rate": 9.969305435048251e-06,
+      "loss": 1.2229,
+      "step": 834
+    },
+    {
+      "epoch": 0.06436941103916127,
+      "grad_norm": 4.428218841552734,
+      "learning_rate": 9.969167157977275e-06,
+      "loss": 1.1942,
+      "step": 835
+    },
+    {
+      "epoch": 0.06444650015417823,
+      "grad_norm": 4.047502040863037,
+      "learning_rate": 9.969028571104462e-06,
+      "loss": 1.1897,
+      "step": 836
+    },
+    {
+      "epoch": 0.06452358926919519,
+      "grad_norm": 4.053440570831299,
+      "learning_rate": 9.968889674438453e-06,
+      "loss": 1.1565,
+      "step": 837
+    },
+    {
+      "epoch": 0.06460067838421214,
+      "grad_norm": 4.556980609893799,
+      "learning_rate": 9.968750467987907e-06,
+      "loss": 1.2299,
+      "step": 838
+    },
+    {
+      "epoch": 0.06467776749922911,
+      "grad_norm": 4.244764804840088,
+      "learning_rate": 9.968610951761504e-06,
+      "loss": 1.2834,
+      "step": 839
+    },
+    {
+      "epoch": 0.06475485661424607,
+      "grad_norm": 4.483509540557861,
+      "learning_rate": 9.968471125767942e-06,
+      "loss": 1.2776,
+      "step": 840
+    },
+    {
+      "epoch": 0.06483194572926303,
+      "grad_norm": 4.05586051940918,
+      "learning_rate": 9.968330990015935e-06,
+      "loss": 1.3225,
+      "step": 841
+    },
+    {
+      "epoch": 0.06490903484427998,
+      "grad_norm": 4.234732627868652,
+      "learning_rate": 9.968190544514225e-06,
+      "loss": 1.3191,
+      "step": 842
+    },
+    {
+      "epoch": 0.06498612395929695,
+      "grad_norm": 3.936809778213501,
+      "learning_rate": 9.968049789271564e-06,
+      "loss": 1.1848,
+      "step": 843
+    },
+    {
+      "epoch": 0.06506321307431391,
+      "grad_norm": 4.090641498565674,
+      "learning_rate": 9.96790872429673e-06,
+      "loss": 1.1973,
+      "step": 844
+    },
+    {
+      "epoch": 0.06514030218933087,
+      "grad_norm": 4.085000038146973,
+      "learning_rate": 9.967767349598517e-06,
+      "loss": 1.1307,
+      "step": 845
+    },
+    {
+      "epoch": 0.06521739130434782,
+      "grad_norm": 4.048746585845947,
+      "learning_rate": 9.967625665185737e-06,
+      "loss": 1.2981,
+      "step": 846
+    },
+    {
+      "epoch": 0.06529448041936478,
+      "grad_norm": 4.098531246185303,
+      "learning_rate": 9.967483671067224e-06,
+      "loss": 1.2129,
+      "step": 847
+    },
+    {
+      "epoch": 0.06537156953438175,
+      "grad_norm": 4.320443630218506,
+      "learning_rate": 9.967341367251833e-06,
+      "loss": 1.1298,
+      "step": 848
+    },
+    {
+      "epoch": 0.06544865864939871,
+      "grad_norm": 4.260039806365967,
+      "learning_rate": 9.967198753748432e-06,
+      "loss": 1.2355,
+      "step": 849
+    },
+    {
+      "epoch": 0.06552574776441566,
+      "grad_norm": 4.163205623626709,
+      "learning_rate": 9.967055830565917e-06,
+      "loss": 1.2541,
+      "step": 850
+    },
+    {
+      "epoch": 0.06560283687943262,
+      "grad_norm": 3.9784324169158936,
+      "learning_rate": 9.966912597713196e-06,
+      "loss": 1.1646,
+      "step": 851
+    },
+    {
+      "epoch": 0.06567992599444958,
+      "grad_norm": 4.5116682052612305,
+      "learning_rate": 9.966769055199197e-06,
+      "loss": 1.2124,
+      "step": 852
+    },
+    {
+      "epoch": 0.06575701510946655,
+      "grad_norm": 3.7753348350524902,
+      "learning_rate": 9.966625203032871e-06,
+      "loss": 1.1988,
+      "step": 853
+    },
+    {
+      "epoch": 0.0658341042244835,
+      "grad_norm": 4.4301323890686035,
+      "learning_rate": 9.966481041223188e-06,
+      "loss": 1.146,
+      "step": 854
+    },
+    {
+      "epoch": 0.06591119333950046,
+      "grad_norm": 4.574663162231445,
+      "learning_rate": 9.966336569779133e-06,
+      "loss": 1.1577,
+      "step": 855
+    },
+    {
+      "epoch": 0.06598828245451742,
+      "grad_norm": 3.872276544570923,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 1.0697,
+      "step": 856
+    },
+    {
+      "epoch": 0.06606537156953438,
+      "grad_norm": 3.989187240600586,
+      "learning_rate": 9.96604669802396e-06,
+      "loss": 1.1815,
+      "step": 857
+    },
+    {
+      "epoch": 0.06614246068455135,
+      "grad_norm": 4.288700580596924,
+      "learning_rate": 9.965901297730914e-06,
+      "loss": 1.0518,
+      "step": 858
+    },
+    {
+      "epoch": 0.0662195497995683,
+      "grad_norm": 4.061006546020508,
+      "learning_rate": 9.965755587839638e-06,
+      "loss": 1.2016,
+      "step": 859
+    },
+    {
+      "epoch": 0.06629663891458526,
+      "grad_norm": 4.13923454284668,
+      "learning_rate": 9.96560956835922e-06,
+      "loss": 1.1651,
+      "step": 860
+    },
+    {
+      "epoch": 0.06637372802960222,
+      "grad_norm": 4.172252655029297,
+      "learning_rate": 9.965463239298764e-06,
+      "loss": 1.1611,
+      "step": 861
+    },
+    {
+      "epoch": 0.06645081714461917,
+      "grad_norm": 4.616154670715332,
+      "learning_rate": 9.965316600667394e-06,
+      "loss": 1.2395,
+      "step": 862
+    },
+    {
+      "epoch": 0.06652790625963614,
+      "grad_norm": 4.157519340515137,
+      "learning_rate": 9.965169652474247e-06,
+      "loss": 1.271,
+      "step": 863
+    },
+    {
+      "epoch": 0.0666049953746531,
+      "grad_norm": 4.29591178894043,
+      "learning_rate": 9.965022394728487e-06,
+      "loss": 1.252,
+      "step": 864
+    },
+    {
+      "epoch": 0.06668208448967006,
+      "grad_norm": 4.058745861053467,
+      "learning_rate": 9.964874827439297e-06,
+      "loss": 1.2369,
+      "step": 865
+    },
+    {
+      "epoch": 0.06675917360468701,
+      "grad_norm": 4.279184818267822,
+      "learning_rate": 9.964726950615875e-06,
+      "loss": 1.3584,
+      "step": 866
+    },
+    {
+      "epoch": 0.06683626271970398,
+      "grad_norm": 4.2356672286987305,
+      "learning_rate": 9.964578764267441e-06,
+      "loss": 1.1432,
+      "step": 867
+    },
+    {
+      "epoch": 0.06691335183472094,
+      "grad_norm": 5.137538909912109,
+      "learning_rate": 9.96443026840323e-06,
+      "loss": 1.1755,
+      "step": 868
+    },
+    {
+      "epoch": 0.0669904409497379,
+      "grad_norm": 4.083669185638428,
+      "learning_rate": 9.964281463032507e-06,
+      "loss": 1.285,
+      "step": 869
+    },
+    {
+      "epoch": 0.06706753006475485,
+      "grad_norm": 4.169551372528076,
+      "learning_rate": 9.964132348164544e-06,
+      "loss": 1.1206,
+      "step": 870
+    },
+    {
+      "epoch": 0.06714461917977181,
+      "grad_norm": 4.000573635101318,
+      "learning_rate": 9.96398292380864e-06,
+      "loss": 1.1801,
+      "step": 871
+    },
+    {
+      "epoch": 0.06722170829478878,
+      "grad_norm": 4.286387920379639,
+      "learning_rate": 9.96383318997411e-06,
+      "loss": 1.0842,
+      "step": 872
+    },
+    {
+      "epoch": 0.06729879740980574,
+      "grad_norm": 3.734882354736328,
+      "learning_rate": 9.963683146670286e-06,
+      "loss": 1.1005,
+      "step": 873
+    },
+    {
+      "epoch": 0.0673758865248227,
+      "grad_norm": 4.232124328613281,
+      "learning_rate": 9.963532793906529e-06,
+      "loss": 1.2437,
+      "step": 874
+    },
+    {
+      "epoch": 0.06745297563983965,
+      "grad_norm": 4.824527740478516,
+      "learning_rate": 9.963382131692208e-06,
+      "loss": 1.2967,
+      "step": 875
+    },
+    {
+      "epoch": 0.06753006475485661,
+      "grad_norm": 4.362327575683594,
+      "learning_rate": 9.963231160036716e-06,
+      "loss": 1.23,
+      "step": 876
+    },
+    {
+      "epoch": 0.06760715386987358,
+      "grad_norm": 4.00999641418457,
+      "learning_rate": 9.963079878949466e-06,
+      "loss": 1.2246,
+      "step": 877
+    },
+    {
+      "epoch": 0.06768424298489054,
+      "grad_norm": 4.2067131996154785,
+      "learning_rate": 9.962928288439891e-06,
+      "loss": 1.2255,
+      "step": 878
+    },
+    {
+      "epoch": 0.06776133209990749,
+      "grad_norm": 4.093700408935547,
+      "learning_rate": 9.962776388517441e-06,
+      "loss": 1.199,
+      "step": 879
+    },
+    {
+      "epoch": 0.06783842121492445,
+      "grad_norm": 4.058950424194336,
+      "learning_rate": 9.962624179191586e-06,
+      "loss": 1.1883,
+      "step": 880
+    },
+    {
+      "epoch": 0.0679155103299414,
+      "grad_norm": 4.307193756103516,
+      "learning_rate": 9.962471660471815e-06,
+      "loss": 1.1255,
+      "step": 881
+    },
+    {
+      "epoch": 0.06799259944495838,
+      "grad_norm": 3.914050579071045,
+      "learning_rate": 9.96231883236764e-06,
+      "loss": 1.1739,
+      "step": 882
+    },
+    {
+      "epoch": 0.06806968855997533,
+      "grad_norm": 3.9826557636260986,
+      "learning_rate": 9.962165694888583e-06,
+      "loss": 1.1386,
+      "step": 883
+    },
+    {
+      "epoch": 0.06814677767499229,
+      "grad_norm": 3.8433432579040527,
+      "learning_rate": 9.962012248044195e-06,
+      "loss": 1.0846,
+      "step": 884
+    },
+    {
+      "epoch": 0.06822386679000925,
+      "grad_norm": 3.975701093673706,
+      "learning_rate": 9.961858491844044e-06,
+      "loss": 1.1565,
+      "step": 885
+    },
+    {
+      "epoch": 0.06830095590502622,
+      "grad_norm": 4.125971794128418,
+      "learning_rate": 9.961704426297712e-06,
+      "loss": 1.1605,
+      "step": 886
+    },
+    {
+      "epoch": 0.06837804502004317,
+      "grad_norm": 4.062132358551025,
+      "learning_rate": 9.961550051414808e-06,
+      "loss": 1.2098,
+      "step": 887
+    },
+    {
+      "epoch": 0.06845513413506013,
+      "grad_norm": 4.446691036224365,
+      "learning_rate": 9.961395367204953e-06,
+      "loss": 1.1216,
+      "step": 888
+    },
+    {
+      "epoch": 0.06853222325007709,
+      "grad_norm": 4.357199668884277,
+      "learning_rate": 9.961240373677793e-06,
+      "loss": 1.0838,
+      "step": 889
+    },
+    {
+      "epoch": 0.06860931236509404,
+      "grad_norm": 4.313502311706543,
+      "learning_rate": 9.961085070842992e-06,
+      "loss": 1.1937,
+      "step": 890
+    },
+    {
+      "epoch": 0.06868640148011101,
+      "grad_norm": 3.93827223777771,
+      "learning_rate": 9.960929458710231e-06,
+      "loss": 1.1371,
+      "step": 891
+    },
+    {
+      "epoch": 0.06876349059512797,
+      "grad_norm": 3.7555925846099854,
+      "learning_rate": 9.960773537289213e-06,
+      "loss": 1.1783,
+      "step": 892
+    },
+    {
+      "epoch": 0.06884057971014493,
+      "grad_norm": 4.327067852020264,
+      "learning_rate": 9.960617306589654e-06,
+      "loss": 1.2371,
+      "step": 893
+    },
+    {
+      "epoch": 0.06891766882516188,
+      "grad_norm": 3.748533248901367,
+      "learning_rate": 9.960460766621299e-06,
+      "loss": 1.1646,
+      "step": 894
+    },
+    {
+      "epoch": 0.06899475794017884,
+      "grad_norm": 4.586325168609619,
+      "learning_rate": 9.960303917393906e-06,
+      "loss": 1.2139,
+      "step": 895
+    },
+    {
+      "epoch": 0.06907184705519581,
+      "grad_norm": 4.419130802154541,
+      "learning_rate": 9.960146758917254e-06,
+      "loss": 1.4065,
+      "step": 896
+    },
+    {
+      "epoch": 0.06914893617021277,
+      "grad_norm": 4.031462669372559,
+      "learning_rate": 9.959989291201141e-06,
+      "loss": 1.2925,
+      "step": 897
+    },
+    {
+      "epoch": 0.06922602528522973,
+      "grad_norm": 4.59142541885376,
+      "learning_rate": 9.959831514255383e-06,
+      "loss": 1.2006,
+      "step": 898
+    },
+    {
+      "epoch": 0.06930311440024668,
+      "grad_norm": 3.835858106613159,
+      "learning_rate": 9.95967342808982e-06,
+      "loss": 1.2036,
+      "step": 899
+    },
+    {
+      "epoch": 0.06938020351526364,
+      "grad_norm": 4.229325771331787,
+      "learning_rate": 9.959515032714304e-06,
+      "loss": 1.1446,
+      "step": 900
+    },
+    {
+      "epoch": 0.06945729263028061,
+      "grad_norm": 4.047120094299316,
+      "learning_rate": 9.959356328138712e-06,
+      "loss": 1.1789,
+      "step": 901
+    },
+    {
+      "epoch": 0.06953438174529757,
+      "grad_norm": 3.6105780601501465,
+      "learning_rate": 9.959197314372937e-06,
+      "loss": 1.192,
+      "step": 902
+    },
+    {
+      "epoch": 0.06961147086031452,
+      "grad_norm": 4.577813625335693,
+      "learning_rate": 9.959037991426896e-06,
+      "loss": 1.2333,
+      "step": 903
+    },
+    {
+      "epoch": 0.06968855997533148,
+      "grad_norm": 3.890207052230835,
+      "learning_rate": 9.958878359310518e-06,
+      "loss": 1.1167,
+      "step": 904
+    },
+    {
+      "epoch": 0.06976564909034844,
+      "grad_norm": 4.432231426239014,
+      "learning_rate": 9.958718418033757e-06,
+      "loss": 1.196,
+      "step": 905
+    },
+    {
+      "epoch": 0.0698427382053654,
+      "grad_norm": 3.846874475479126,
+      "learning_rate": 9.958558167606585e-06,
+      "loss": 1.1761,
+      "step": 906
+    },
+    {
+      "epoch": 0.06991982732038236,
+      "grad_norm": 4.190489292144775,
+      "learning_rate": 9.958397608038994e-06,
+      "loss": 1.129,
+      "step": 907
+    },
+    {
+      "epoch": 0.06999691643539932,
+      "grad_norm": 3.783397912979126,
+      "learning_rate": 9.95823673934099e-06,
+      "loss": 1.2228,
+      "step": 908
+    },
+    {
+      "epoch": 0.07007400555041628,
+      "grad_norm": 3.776473045349121,
+      "learning_rate": 9.958075561522605e-06,
+      "loss": 1.1348,
+      "step": 909
+    },
+    {
+      "epoch": 0.07015109466543325,
+      "grad_norm": 4.117181777954102,
+      "learning_rate": 9.957914074593889e-06,
+      "loss": 1.1504,
+      "step": 910
+    },
+    {
+      "epoch": 0.0702281837804502,
+      "grad_norm": 4.3804545402526855,
+      "learning_rate": 9.957752278564905e-06,
+      "loss": 1.2356,
+      "step": 911
+    },
+    {
+      "epoch": 0.07030527289546716,
+      "grad_norm": 4.283797264099121,
+      "learning_rate": 9.957590173445746e-06,
+      "loss": 1.2674,
+      "step": 912
+    },
+    {
+      "epoch": 0.07038236201048412,
+      "grad_norm": 4.3605852127075195,
+      "learning_rate": 9.957427759246515e-06,
+      "loss": 1.2091,
+      "step": 913
+    },
+    {
+      "epoch": 0.07045945112550107,
+      "grad_norm": 4.470867156982422,
+      "learning_rate": 9.957265035977338e-06,
+      "loss": 1.2004,
+      "step": 914
+    },
+    {
+      "epoch": 0.07053654024051804,
+      "grad_norm": 4.192192077636719,
+      "learning_rate": 9.95710200364836e-06,
+      "loss": 1.2206,
+      "step": 915
+    },
+    {
+      "epoch": 0.070613629355535,
+      "grad_norm": 4.35572624206543,
+      "learning_rate": 9.956938662269745e-06,
+      "loss": 1.2407,
+      "step": 916
+    },
+    {
+      "epoch": 0.07069071847055196,
+      "grad_norm": 3.7318854331970215,
+      "learning_rate": 9.956775011851679e-06,
+      "loss": 1.1075,
+      "step": 917
+    },
+    {
+      "epoch": 0.07076780758556891,
+      "grad_norm": 4.309021949768066,
+      "learning_rate": 9.956611052404362e-06,
+      "loss": 1.1874,
+      "step": 918
+    },
+    {
+      "epoch": 0.07084489670058587,
+      "grad_norm": 4.061330318450928,
+      "learning_rate": 9.956446783938016e-06,
+      "loss": 1.2302,
+      "step": 919
+    },
+    {
+      "epoch": 0.07092198581560284,
+      "grad_norm": 4.194582462310791,
+      "learning_rate": 9.956282206462886e-06,
+      "loss": 1.2344,
+      "step": 920
+    },
+    {
+      "epoch": 0.0709990749306198,
+      "grad_norm": 3.940514326095581,
+      "learning_rate": 9.956117319989226e-06,
+      "loss": 1.1443,
+      "step": 921
+    },
+    {
+      "epoch": 0.07107616404563676,
+      "grad_norm": 4.2155046463012695,
+      "learning_rate": 9.95595212452732e-06,
+      "loss": 1.1655,
+      "step": 922
+    },
+    {
+      "epoch": 0.07115325316065371,
+      "grad_norm": 4.344760417938232,
+      "learning_rate": 9.95578662008747e-06,
+      "loss": 1.3106,
+      "step": 923
+    },
+    {
+      "epoch": 0.07123034227567067,
+      "grad_norm": 3.8140082359313965,
+      "learning_rate": 9.955620806679987e-06,
+      "loss": 1.188,
+      "step": 924
+    },
+    {
+      "epoch": 0.07130743139068764,
+      "grad_norm": 3.9605846405029297,
+      "learning_rate": 9.955454684315215e-06,
+      "loss": 1.21,
+      "step": 925
+    },
+    {
+      "epoch": 0.0713845205057046,
+      "grad_norm": 4.253360271453857,
+      "learning_rate": 9.955288253003508e-06,
+      "loss": 1.283,
+      "step": 926
+    },
+    {
+      "epoch": 0.07146160962072155,
+      "grad_norm": 4.852066516876221,
+      "learning_rate": 9.955121512755242e-06,
+      "loss": 1.2646,
+      "step": 927
+    },
+    {
+      "epoch": 0.07153869873573851,
+      "grad_norm": 4.225571155548096,
+      "learning_rate": 9.954954463580813e-06,
+      "loss": 1.1779,
+      "step": 928
+    },
+    {
+      "epoch": 0.07161578785075548,
+      "grad_norm": 4.433430194854736,
+      "learning_rate": 9.954787105490635e-06,
+      "loss": 1.3087,
+      "step": 929
+    },
+    {
+      "epoch": 0.07169287696577244,
+      "grad_norm": 4.311334609985352,
+      "learning_rate": 9.954619438495142e-06,
+      "loss": 1.1685,
+      "step": 930
+    },
+    {
+      "epoch": 0.0717699660807894,
+      "grad_norm": 4.312464237213135,
+      "learning_rate": 9.954451462604788e-06,
+      "loss": 1.2989,
+      "step": 931
+    },
+    {
+      "epoch": 0.07184705519580635,
+      "grad_norm": 4.1521477699279785,
+      "learning_rate": 9.954283177830047e-06,
+      "loss": 1.2642,
+      "step": 932
+    },
+    {
+      "epoch": 0.07192414431082331,
+      "grad_norm": 4.0158820152282715,
+      "learning_rate": 9.954114584181407e-06,
+      "loss": 1.1011,
+      "step": 933
+    },
+    {
+      "epoch": 0.07200123342584028,
+      "grad_norm": 3.9613866806030273,
+      "learning_rate": 9.953945681669381e-06,
+      "loss": 1.174,
+      "step": 934
+    },
+    {
+      "epoch": 0.07207832254085723,
+      "grad_norm": 4.178623199462891,
+      "learning_rate": 9.953776470304499e-06,
+      "loss": 1.174,
+      "step": 935
+    },
+    {
+      "epoch": 0.07215541165587419,
+      "grad_norm": 4.1417131423950195,
+      "learning_rate": 9.95360695009731e-06,
+      "loss": 1.2549,
+      "step": 936
+    },
+    {
+      "epoch": 0.07223250077089115,
+      "grad_norm": 4.326009750366211,
+      "learning_rate": 9.953437121058384e-06,
+      "loss": 1.1613,
+      "step": 937
+    },
+    {
+      "epoch": 0.0723095898859081,
+      "grad_norm": 3.9384677410125732,
+      "learning_rate": 9.953266983198307e-06,
+      "loss": 1.1953,
+      "step": 938
+    },
+    {
+      "epoch": 0.07238667900092507,
+      "grad_norm": 4.292700290679932,
+      "learning_rate": 9.953096536527688e-06,
+      "loss": 1.2971,
+      "step": 939
+    },
+    {
+      "epoch": 0.07246376811594203,
+      "grad_norm": 3.9370415210723877,
+      "learning_rate": 9.952925781057152e-06,
+      "loss": 1.2132,
+      "step": 940
+    },
+    {
+      "epoch": 0.07254085723095899,
+      "grad_norm": 4.184019088745117,
+      "learning_rate": 9.952754716797345e-06,
+      "loss": 1.1628,
+      "step": 941
+    },
+    {
+      "epoch": 0.07261794634597594,
+      "grad_norm": 3.967825412750244,
+      "learning_rate": 9.952583343758934e-06,
+      "loss": 1.242,
+      "step": 942
+    },
+    {
+      "epoch": 0.0726950354609929,
+      "grad_norm": 4.628615379333496,
+      "learning_rate": 9.9524116619526e-06,
+      "loss": 1.2041,
+      "step": 943
+    },
+    {
+      "epoch": 0.07277212457600987,
+      "grad_norm": 3.8717031478881836,
+      "learning_rate": 9.952239671389049e-06,
+      "loss": 1.0256,
+      "step": 944
+    },
+    {
+      "epoch": 0.07284921369102683,
+      "grad_norm": 3.9724347591400146,
+      "learning_rate": 9.952067372079003e-06,
+      "loss": 1.1419,
+      "step": 945
+    },
+    {
+      "epoch": 0.07292630280604379,
+      "grad_norm": 4.037529468536377,
+      "learning_rate": 9.951894764033202e-06,
+      "loss": 1.2491,
+      "step": 946
+    },
+    {
+      "epoch": 0.07300339192106074,
+      "grad_norm": 4.202883720397949,
+      "learning_rate": 9.951721847262413e-06,
+      "loss": 1.2303,
+      "step": 947
+    },
+    {
+      "epoch": 0.0730804810360777,
+      "grad_norm": 4.007006645202637,
+      "learning_rate": 9.951548621777409e-06,
+      "loss": 1.1551,
+      "step": 948
+    },
+    {
+      "epoch": 0.07315757015109467,
+      "grad_norm": 4.146481513977051,
+      "learning_rate": 9.951375087588993e-06,
+      "loss": 1.2194,
+      "step": 949
+    },
+    {
+      "epoch": 0.07323465926611163,
+      "grad_norm": 4.004715919494629,
+      "learning_rate": 9.951201244707986e-06,
+      "loss": 1.1161,
+      "step": 950
+    },
+    {
+      "epoch": 0.07331174838112858,
+      "grad_norm": 4.251381874084473,
+      "learning_rate": 9.951027093145222e-06,
+      "loss": 1.2399,
+      "step": 951
+    },
+    {
+      "epoch": 0.07338883749614554,
+      "grad_norm": 4.143471717834473,
+      "learning_rate": 9.950852632911563e-06,
+      "loss": 1.1627,
+      "step": 952
+    },
+    {
+      "epoch": 0.07346592661116251,
+      "grad_norm": 3.974905252456665,
+      "learning_rate": 9.950677864017882e-06,
+      "loss": 1.1123,
+      "step": 953
+    },
+    {
+      "epoch": 0.07354301572617947,
+      "grad_norm": 4.213186740875244,
+      "learning_rate": 9.950502786475078e-06,
+      "loss": 1.224,
+      "step": 954
+    },
+    {
+      "epoch": 0.07362010484119642,
+      "grad_norm": 4.501376628875732,
+      "learning_rate": 9.950327400294063e-06,
+      "loss": 1.254,
+      "step": 955
+    },
+    {
+      "epoch": 0.07369719395621338,
+      "grad_norm": 3.8753960132598877,
+      "learning_rate": 9.950151705485774e-06,
+      "loss": 1.1435,
+      "step": 956
+    },
+    {
+      "epoch": 0.07377428307123034,
+      "grad_norm": 4.606151580810547,
+      "learning_rate": 9.949975702061162e-06,
+      "loss": 1.2086,
+      "step": 957
+    },
+    {
+      "epoch": 0.07385137218624731,
+      "grad_norm": 3.9693284034729004,
+      "learning_rate": 9.949799390031203e-06,
+      "loss": 1.2506,
+      "step": 958
+    },
+    {
+      "epoch": 0.07392846130126426,
+      "grad_norm": 4.064651966094971,
+      "learning_rate": 9.949622769406888e-06,
+      "loss": 1.1616,
+      "step": 959
+    },
+    {
+      "epoch": 0.07400555041628122,
+      "grad_norm": 4.274019241333008,
+      "learning_rate": 9.949445840199227e-06,
+      "loss": 1.2061,
+      "step": 960
+    },
+    {
+      "epoch": 0.07408263953129818,
+      "grad_norm": 4.298183917999268,
+      "learning_rate": 9.949268602419253e-06,
+      "loss": 1.2899,
+      "step": 961
+    },
+    {
+      "epoch": 0.07415972864631513,
+      "grad_norm": 4.174452781677246,
+      "learning_rate": 9.949091056078012e-06,
+      "loss": 1.2072,
+      "step": 962
+    },
+    {
+      "epoch": 0.0742368177613321,
+      "grad_norm": 4.208117485046387,
+      "learning_rate": 9.948913201186579e-06,
+      "loss": 1.2707,
+      "step": 963
+    },
+    {
+      "epoch": 0.07431390687634906,
+      "grad_norm": 4.637174606323242,
+      "learning_rate": 9.948735037756037e-06,
+      "loss": 1.2792,
+      "step": 964
+    },
+    {
+      "epoch": 0.07439099599136602,
+      "grad_norm": 3.6787731647491455,
+      "learning_rate": 9.948556565797497e-06,
+      "loss": 1.1813,
+      "step": 965
+    },
+    {
+      "epoch": 0.07446808510638298,
+      "grad_norm": 4.200822830200195,
+      "learning_rate": 9.948377785322082e-06,
+      "loss": 1.2527,
+      "step": 966
+    },
+    {
+      "epoch": 0.07454517422139993,
+      "grad_norm": 3.7479918003082275,
+      "learning_rate": 9.948198696340943e-06,
+      "loss": 1.0949,
+      "step": 967
+    },
+    {
+      "epoch": 0.0746222633364169,
+      "grad_norm": 3.8182127475738525,
+      "learning_rate": 9.94801929886524e-06,
+      "loss": 1.1945,
+      "step": 968
+    },
+    {
+      "epoch": 0.07469935245143386,
+      "grad_norm": 4.381171703338623,
+      "learning_rate": 9.947839592906163e-06,
+      "loss": 1.302,
+      "step": 969
+    },
+    {
+      "epoch": 0.07477644156645082,
+      "grad_norm": 4.537683963775635,
+      "learning_rate": 9.947659578474911e-06,
+      "loss": 1.3543,
+      "step": 970
+    },
+    {
+      "epoch": 0.07485353068146777,
+      "grad_norm": 4.541914939880371,
+      "learning_rate": 9.94747925558271e-06,
+      "loss": 1.213,
+      "step": 971
+    },
+    {
+      "epoch": 0.07493061979648474,
+      "grad_norm": 4.564115524291992,
+      "learning_rate": 9.9472986242408e-06,
+      "loss": 1.2574,
+      "step": 972
+    },
+    {
+      "epoch": 0.0750077089115017,
+      "grad_norm": 4.06321907043457,
+      "learning_rate": 9.947117684460443e-06,
+      "loss": 1.227,
+      "step": 973
+    },
+    {
+      "epoch": 0.07508479802651866,
+      "grad_norm": 4.070244312286377,
+      "learning_rate": 9.946936436252923e-06,
+      "loss": 1.2062,
+      "step": 974
+    },
+    {
+      "epoch": 0.07516188714153561,
+      "grad_norm": 3.8233132362365723,
+      "learning_rate": 9.946754879629535e-06,
+      "loss": 1.1082,
+      "step": 975
+    },
+    {
+      "epoch": 0.07523897625655257,
+      "grad_norm": 4.254786014556885,
+      "learning_rate": 9.9465730146016e-06,
+      "loss": 1.2617,
+      "step": 976
+    },
+    {
+      "epoch": 0.07531606537156954,
+      "grad_norm": 5.448151111602783,
+      "learning_rate": 9.946390841180457e-06,
+      "loss": 1.3154,
+      "step": 977
+    },
+    {
+      "epoch": 0.0753931544865865,
+      "grad_norm": 3.762033462524414,
+      "learning_rate": 9.946208359377463e-06,
+      "loss": 1.1022,
+      "step": 978
+    },
+    {
+      "epoch": 0.07547024360160345,
+      "grad_norm": 4.151965618133545,
+      "learning_rate": 9.946025569203994e-06,
+      "loss": 1.1938,
+      "step": 979
+    },
+    {
+      "epoch": 0.07554733271662041,
+      "grad_norm": 3.941340684890747,
+      "learning_rate": 9.945842470671447e-06,
+      "loss": 1.2629,
+      "step": 980
+    },
+    {
+      "epoch": 0.07562442183163737,
+      "grad_norm": 3.871173620223999,
+      "learning_rate": 9.945659063791239e-06,
+      "loss": 1.0924,
+      "step": 981
+    },
+    {
+      "epoch": 0.07570151094665434,
+      "grad_norm": 3.7768266201019287,
+      "learning_rate": 9.9454753485748e-06,
+      "loss": 1.0993,
+      "step": 982
+    },
+    {
+      "epoch": 0.0757786000616713,
+      "grad_norm": 3.877859354019165,
+      "learning_rate": 9.945291325033587e-06,
+      "loss": 1.1623,
+      "step": 983
+    },
+    {
+      "epoch": 0.07585568917668825,
+      "grad_norm": 3.9639294147491455,
+      "learning_rate": 9.945106993179074e-06,
+      "loss": 1.0856,
+      "step": 984
+    },
+    {
+      "epoch": 0.07593277829170521,
+      "grad_norm": 4.09315824508667,
+      "learning_rate": 9.94492235302275e-06,
+      "loss": 1.2126,
+      "step": 985
+    },
+    {
+      "epoch": 0.07600986740672216,
+      "grad_norm": 4.234272480010986,
+      "learning_rate": 9.944737404576129e-06,
+      "loss": 1.1984,
+      "step": 986
+    },
+    {
+      "epoch": 0.07608695652173914,
+      "grad_norm": 3.7921226024627686,
+      "learning_rate": 9.94455214785074e-06,
+      "loss": 1.158,
+      "step": 987
+    },
+    {
+      "epoch": 0.07616404563675609,
+      "grad_norm": 4.407901763916016,
+      "learning_rate": 9.944366582858131e-06,
+      "loss": 1.1847,
+      "step": 988
+    },
+    {
+      "epoch": 0.07624113475177305,
+      "grad_norm": 4.234422206878662,
+      "learning_rate": 9.944180709609874e-06,
+      "loss": 1.1911,
+      "step": 989
+    },
+    {
+      "epoch": 0.07631822386679,
+      "grad_norm": 3.9739015102386475,
+      "learning_rate": 9.943994528117557e-06,
+      "loss": 1.1974,
+      "step": 990
+    },
+    {
+      "epoch": 0.07639531298180696,
+      "grad_norm": 4.158312797546387,
+      "learning_rate": 9.943808038392786e-06,
+      "loss": 1.1854,
+      "step": 991
+    },
+    {
+      "epoch": 0.07647240209682393,
+      "grad_norm": 3.7783594131469727,
+      "learning_rate": 9.943621240447188e-06,
+      "loss": 1.1394,
+      "step": 992
+    },
+    {
+      "epoch": 0.07654949121184089,
+      "grad_norm": 4.1893157958984375,
+      "learning_rate": 9.943434134292412e-06,
+      "loss": 1.2535,
+      "step": 993
+    },
+    {
+      "epoch": 0.07662658032685785,
+      "grad_norm": 4.0416388511657715,
+      "learning_rate": 9.943246719940118e-06,
+      "loss": 1.14,
+      "step": 994
+    },
+    {
+      "epoch": 0.0767036694418748,
+      "grad_norm": 5.405656337738037,
+      "learning_rate": 9.943058997401993e-06,
+      "loss": 1.2001,
+      "step": 995
+    },
+    {
+      "epoch": 0.07678075855689177,
+      "grad_norm": 3.7127041816711426,
+      "learning_rate": 9.942870966689742e-06,
+      "loss": 1.1386,
+      "step": 996
+    },
+    {
+      "epoch": 0.07685784767190873,
+      "grad_norm": 4.761865615844727,
+      "learning_rate": 9.942682627815084e-06,
+      "loss": 1.1629,
+      "step": 997
+    },
+    {
+      "epoch": 0.07693493678692569,
+      "grad_norm": 4.404749393463135,
+      "learning_rate": 9.942493980789762e-06,
+      "loss": 1.1531,
+      "step": 998
+    },
+    {
+      "epoch": 0.07701202590194264,
+      "grad_norm": 4.3769402503967285,
+      "learning_rate": 9.94230502562554e-06,
+      "loss": 1.2155,
+      "step": 999
+    },
+    {
+      "epoch": 0.0770891150169596,
+      "grad_norm": 4.384872913360596,
+      "learning_rate": 9.942115762334196e-06,
+      "loss": 1.287,
+      "step": 1000
+    },
+    {
+      "epoch": 0.07716620413197657,
+      "grad_norm": 3.899238109588623,
+      "learning_rate": 9.941926190927532e-06,
+      "loss": 1.1246,
+      "step": 1001
+    },
+    {
+      "epoch": 0.07724329324699353,
+      "grad_norm": 3.7073872089385986,
+      "learning_rate": 9.941736311417362e-06,
+      "loss": 1.2122,
+      "step": 1002
+    },
+    {
+      "epoch": 0.07732038236201048,
+      "grad_norm": 3.755033254623413,
+      "learning_rate": 9.94154612381553e-06,
+      "loss": 1.1603,
+      "step": 1003
+    },
+    {
+      "epoch": 0.07739747147702744,
+      "grad_norm": 4.051525115966797,
+      "learning_rate": 9.941355628133887e-06,
+      "loss": 1.121,
+      "step": 1004
+    },
+    {
+      "epoch": 0.0774745605920444,
+      "grad_norm": 3.882408380508423,
+      "learning_rate": 9.941164824384313e-06,
+      "loss": 1.2453,
+      "step": 1005
+    },
+    {
+      "epoch": 0.07755164970706137,
+      "grad_norm": 4.029034614562988,
+      "learning_rate": 9.940973712578706e-06,
+      "loss": 1.1899,
+      "step": 1006
+    },
+    {
+      "epoch": 0.07762873882207832,
+      "grad_norm": 4.212346076965332,
+      "learning_rate": 9.940782292728975e-06,
+      "loss": 1.2007,
+      "step": 1007
+    },
+    {
+      "epoch": 0.07770582793709528,
+      "grad_norm": 4.014603137969971,
+      "learning_rate": 9.940590564847059e-06,
+      "loss": 1.2438,
+      "step": 1008
+    },
+    {
+      "epoch": 0.07778291705211224,
+      "grad_norm": 4.43209171295166,
+      "learning_rate": 9.940398528944906e-06,
+      "loss": 1.2154,
+      "step": 1009
+    },
+    {
+      "epoch": 0.0778600061671292,
+      "grad_norm": 4.110109329223633,
+      "learning_rate": 9.940206185034496e-06,
+      "loss": 1.0843,
+      "step": 1010
+    },
+    {
+      "epoch": 0.07793709528214617,
+      "grad_norm": 4.09305477142334,
+      "learning_rate": 9.940013533127813e-06,
+      "loss": 1.1252,
+      "step": 1011
+    },
+    {
+      "epoch": 0.07801418439716312,
+      "grad_norm": 3.9109747409820557,
+      "learning_rate": 9.939820573236873e-06,
+      "loss": 1.3243,
+      "step": 1012
+    },
+    {
+      "epoch": 0.07809127351218008,
+      "grad_norm": 4.246638774871826,
+      "learning_rate": 9.939627305373703e-06,
+      "loss": 1.0503,
+      "step": 1013
+    },
+    {
+      "epoch": 0.07816836262719704,
+      "grad_norm": 4.346315383911133,
+      "learning_rate": 9.939433729550354e-06,
+      "loss": 1.2939,
+      "step": 1014
+    },
+    {
+      "epoch": 0.078245451742214,
+      "grad_norm": 4.1435322761535645,
+      "learning_rate": 9.939239845778894e-06,
+      "loss": 1.2417,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07832254085723096,
+      "grad_norm": 4.223635673522949,
+      "learning_rate": 9.93904565407141e-06,
+      "loss": 1.2014,
+      "step": 1016
+    },
+    {
+      "epoch": 0.07839962997224792,
+      "grad_norm": 3.9585742950439453,
+      "learning_rate": 9.938851154440012e-06,
+      "loss": 1.0748,
+      "step": 1017
+    },
+    {
+      "epoch": 0.07847671908726488,
+      "grad_norm": 4.183574199676514,
+      "learning_rate": 9.93865634689682e-06,
+      "loss": 1.1727,
+      "step": 1018
+    },
+    {
+      "epoch": 0.07855380820228183,
+      "grad_norm": 3.769174098968506,
+      "learning_rate": 9.938461231453985e-06,
+      "loss": 1.2277,
+      "step": 1019
+    },
+    {
+      "epoch": 0.0786308973172988,
+      "grad_norm": 3.8185031414031982,
+      "learning_rate": 9.938265808123667e-06,
+      "loss": 1.1434,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07870798643231576,
+      "grad_norm": 4.209797382354736,
+      "learning_rate": 9.938070076918056e-06,
+      "loss": 1.1186,
+      "step": 1021
+    },
+    {
+      "epoch": 0.07878507554733272,
+      "grad_norm": 4.088057518005371,
+      "learning_rate": 9.937874037849346e-06,
+      "loss": 1.2485,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07886216466234967,
+      "grad_norm": 4.1362504959106445,
+      "learning_rate": 9.937677690929766e-06,
+      "loss": 1.1779,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07893925377736663,
+      "grad_norm": 4.136228084564209,
+      "learning_rate": 9.937481036171555e-06,
+      "loss": 1.1269,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0790163428923836,
+      "grad_norm": 4.334633827209473,
+      "learning_rate": 9.937284073586972e-06,
+      "loss": 1.191,
+      "step": 1025
+    },
+    {
+      "epoch": 0.07909343200740056,
+      "grad_norm": 4.252748012542725,
+      "learning_rate": 9.9370868031883e-06,
+      "loss": 1.2082,
+      "step": 1026
+    },
+    {
+      "epoch": 0.07917052112241751,
+      "grad_norm": 4.121849060058594,
+      "learning_rate": 9.936889224987834e-06,
+      "loss": 1.2799,
+      "step": 1027
+    },
+    {
+      "epoch": 0.07924761023743447,
+      "grad_norm": 4.112194061279297,
+      "learning_rate": 9.936691338997894e-06,
+      "loss": 1.2097,
+      "step": 1028
+    },
+    {
+      "epoch": 0.07932469935245143,
+      "grad_norm": 3.915144920349121,
+      "learning_rate": 9.936493145230817e-06,
+      "loss": 1.0446,
+      "step": 1029
+    },
+    {
+      "epoch": 0.0794017884674684,
+      "grad_norm": 4.162792682647705,
+      "learning_rate": 9.936294643698958e-06,
+      "loss": 1.1855,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07947887758248535,
+      "grad_norm": 4.035287380218506,
+      "learning_rate": 9.936095834414693e-06,
+      "loss": 1.2207,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07955596669750231,
+      "grad_norm": 4.192042827606201,
+      "learning_rate": 9.935896717390421e-06,
+      "loss": 1.2242,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07963305581251927,
+      "grad_norm": 4.014597415924072,
+      "learning_rate": 9.93569729263855e-06,
+      "loss": 1.1331,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07971014492753623,
+      "grad_norm": 3.836013078689575,
+      "learning_rate": 9.935497560171516e-06,
+      "loss": 1.1603,
+      "step": 1034
+    },
+    {
+      "epoch": 0.0797872340425532,
+      "grad_norm": 4.218697547912598,
+      "learning_rate": 9.93529752000177e-06,
+      "loss": 1.2531,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07986432315757015,
+      "grad_norm": 3.811758518218994,
+      "learning_rate": 9.935097172141785e-06,
+      "loss": 1.1419,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07994141227258711,
+      "grad_norm": 4.130055904388428,
+      "learning_rate": 9.93489651660405e-06,
+      "loss": 1.1776,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08001850138760407,
+      "grad_norm": 4.011743068695068,
+      "learning_rate": 9.934695553401076e-06,
+      "loss": 1.2096,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08009559050262104,
+      "grad_norm": 4.1407928466796875,
+      "learning_rate": 9.934494282545393e-06,
+      "loss": 1.1341,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08017267961763799,
+      "grad_norm": 3.8697617053985596,
+      "learning_rate": 9.934292704049546e-06,
+      "loss": 1.1262,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08024976873265495,
+      "grad_norm": 5.077935695648193,
+      "learning_rate": 9.934090817926105e-06,
+      "loss": 1.1413,
+      "step": 1041
+    },
+    {
+      "epoch": 0.0803268578476719,
+      "grad_norm": 4.142470359802246,
+      "learning_rate": 9.933888624187656e-06,
+      "loss": 1.3075,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08040394696268886,
+      "grad_norm": 4.138065814971924,
+      "learning_rate": 9.933686122846804e-06,
+      "loss": 1.1422,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08048103607770583,
+      "grad_norm": 4.101687908172607,
+      "learning_rate": 9.933483313916174e-06,
+      "loss": 1.2461,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08055812519272279,
+      "grad_norm": 4.085910320281982,
+      "learning_rate": 9.933280197408413e-06,
+      "loss": 1.11,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08063521430773975,
+      "grad_norm": 3.8039958477020264,
+      "learning_rate": 9.933076773336179e-06,
+      "loss": 1.1203,
+      "step": 1046
+    },
+    {
+      "epoch": 0.0807123034227567,
+      "grad_norm": 4.0165886878967285,
+      "learning_rate": 9.932873041712158e-06,
+      "loss": 1.1276,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08078939253777366,
+      "grad_norm": 3.982520818710327,
+      "learning_rate": 9.932669002549052e-06,
+      "loss": 1.1715,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08086648165279063,
+      "grad_norm": 4.542549133300781,
+      "learning_rate": 9.93246465585958e-06,
+      "loss": 1.2555,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08094357076780759,
+      "grad_norm": 3.6177902221679688,
+      "learning_rate": 9.932260001656482e-06,
+      "loss": 0.995,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08102065988282454,
+      "grad_norm": 4.135311603546143,
+      "learning_rate": 9.932055039952518e-06,
+      "loss": 1.2257,
+      "step": 1051
+    },
+    {
+      "epoch": 0.0810977489978415,
+      "grad_norm": 4.676593780517578,
+      "learning_rate": 9.931849770760467e-06,
+      "loss": 1.214,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08117483811285846,
+      "grad_norm": 3.9388997554779053,
+      "learning_rate": 9.931644194093124e-06,
+      "loss": 1.1846,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08125192722787543,
+      "grad_norm": 4.0811896324157715,
+      "learning_rate": 9.931438309963308e-06,
+      "loss": 1.1154,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08132901634289239,
+      "grad_norm": 4.8191118240356445,
+      "learning_rate": 9.931232118383854e-06,
+      "loss": 1.2899,
+      "step": 1055
+    },
+    {
+      "epoch": 0.08140610545790934,
+      "grad_norm": 4.420556545257568,
+      "learning_rate": 9.931025619367617e-06,
+      "loss": 1.2523,
+      "step": 1056
+    },
+    {
+      "epoch": 0.0814831945729263,
+      "grad_norm": 4.228475093841553,
+      "learning_rate": 9.930818812927471e-06,
+      "loss": 1.2077,
+      "step": 1057
+    },
+    {
+      "epoch": 0.08156028368794327,
+      "grad_norm": 4.304680347442627,
+      "learning_rate": 9.93061169907631e-06,
+      "loss": 1.1991,
+      "step": 1058
+    },
+    {
+      "epoch": 0.08163737280296023,
+      "grad_norm": 4.341451168060303,
+      "learning_rate": 9.930404277827044e-06,
+      "loss": 1.1833,
+      "step": 1059
+    },
+    {
+      "epoch": 0.08171446191797718,
+      "grad_norm": 3.8857333660125732,
+      "learning_rate": 9.930196549192608e-06,
+      "loss": 1.1814,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08179155103299414,
+      "grad_norm": 3.722275972366333,
+      "learning_rate": 9.929988513185952e-06,
+      "loss": 1.199,
+      "step": 1061
+    },
+    {
+      "epoch": 0.0818686401480111,
+      "grad_norm": 3.972513437271118,
+      "learning_rate": 9.929780169820043e-06,
+      "loss": 1.0716,
+      "step": 1062
+    },
+    {
+      "epoch": 0.08194572926302807,
+      "grad_norm": 3.907376289367676,
+      "learning_rate": 9.929571519107873e-06,
+      "loss": 1.0642,
+      "step": 1063
+    },
+    {
+      "epoch": 0.08202281837804502,
+      "grad_norm": 3.913541793823242,
+      "learning_rate": 9.92936256106245e-06,
+      "loss": 1.1794,
+      "step": 1064
+    },
+    {
+      "epoch": 0.08209990749306198,
+      "grad_norm": 4.19666862487793,
+      "learning_rate": 9.929153295696803e-06,
+      "loss": 1.2815,
+      "step": 1065
+    },
+    {
+      "epoch": 0.08217699660807894,
+      "grad_norm": 4.4634175300598145,
+      "learning_rate": 9.928943723023973e-06,
+      "loss": 1.2451,
+      "step": 1066
+    },
+    {
+      "epoch": 0.0822540857230959,
+      "grad_norm": 3.9684886932373047,
+      "learning_rate": 9.928733843057033e-06,
+      "loss": 1.1221,
+      "step": 1067
+    },
+    {
+      "epoch": 0.08233117483811286,
+      "grad_norm": 3.8272829055786133,
+      "learning_rate": 9.928523655809062e-06,
+      "loss": 1.1193,
+      "step": 1068
+    },
+    {
+      "epoch": 0.08240826395312982,
+      "grad_norm": 4.0728325843811035,
+      "learning_rate": 9.928313161293168e-06,
+      "loss": 1.1795,
+      "step": 1069
+    },
+    {
+      "epoch": 0.08248535306814678,
+      "grad_norm": 3.951913595199585,
+      "learning_rate": 9.928102359522473e-06,
+      "loss": 1.1846,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08256244218316373,
+      "grad_norm": 3.7784245014190674,
+      "learning_rate": 9.927891250510118e-06,
+      "loss": 1.1188,
+      "step": 1071
+    },
+    {
+      "epoch": 0.08263953129818069,
+      "grad_norm": 4.366214752197266,
+      "learning_rate": 9.927679834269266e-06,
+      "loss": 1.1808,
+      "step": 1072
+    },
+    {
+      "epoch": 0.08271662041319766,
+      "grad_norm": 4.112607002258301,
+      "learning_rate": 9.927468110813099e-06,
+      "loss": 1.1893,
+      "step": 1073
+    },
+    {
+      "epoch": 0.08279370952821462,
+      "grad_norm": 3.9396910667419434,
+      "learning_rate": 9.927256080154813e-06,
+      "loss": 1.2093,
+      "step": 1074
+    },
+    {
+      "epoch": 0.08287079864323157,
+      "grad_norm": 4.674061298370361,
+      "learning_rate": 9.927043742307631e-06,
+      "loss": 1.2488,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08294788775824853,
+      "grad_norm": 5.152288436889648,
+      "learning_rate": 9.926831097284788e-06,
+      "loss": 1.2201,
+      "step": 1076
+    },
+    {
+      "epoch": 0.08302497687326549,
+      "grad_norm": 3.9294207096099854,
+      "learning_rate": 9.926618145099544e-06,
+      "loss": 1.2437,
+      "step": 1077
+    },
+    {
+      "epoch": 0.08310206598828246,
+      "grad_norm": 4.059928894042969,
+      "learning_rate": 9.926404885765175e-06,
+      "loss": 1.1745,
+      "step": 1078
+    },
+    {
+      "epoch": 0.08317915510329942,
+      "grad_norm": 4.130738735198975,
+      "learning_rate": 9.926191319294974e-06,
+      "loss": 1.2372,
+      "step": 1079
+    },
+    {
+      "epoch": 0.08325624421831637,
+      "grad_norm": 4.510885238647461,
+      "learning_rate": 9.92597744570226e-06,
+      "loss": 1.3444,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 3.76959490776062,
+      "learning_rate": 9.92576326500036e-06,
+      "loss": 1.0194,
+      "step": 1081
+    },
+    {
+      "epoch": 0.0834104224483503,
+      "grad_norm": 4.379603862762451,
+      "learning_rate": 9.925548777202636e-06,
+      "loss": 1.2125,
+      "step": 1082
+    },
+    {
+      "epoch": 0.08348751156336726,
+      "grad_norm": 4.4609375,
+      "learning_rate": 9.925333982322456e-06,
+      "loss": 1.2694,
+      "step": 1083
+    },
+    {
+      "epoch": 0.08356460067838421,
+      "grad_norm": 4.134787082672119,
+      "learning_rate": 9.925118880373208e-06,
+      "loss": 1.188,
+      "step": 1084
+    },
+    {
+      "epoch": 0.08364168979340117,
+      "grad_norm": 3.97082781791687,
+      "learning_rate": 9.924903471368308e-06,
+      "loss": 1.0874,
+      "step": 1085
+    },
+    {
+      "epoch": 0.08371877890841813,
+      "grad_norm": 3.723400592803955,
+      "learning_rate": 9.924687755321183e-06,
+      "loss": 1.2485,
+      "step": 1086
+    },
+    {
+      "epoch": 0.0837958680234351,
+      "grad_norm": 4.063663005828857,
+      "learning_rate": 9.92447173224528e-06,
+      "loss": 1.1585,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08387295713845205,
+      "grad_norm": 4.590901851654053,
+      "learning_rate": 9.924255402154072e-06,
+      "loss": 1.2752,
+      "step": 1088
+    },
+    {
+      "epoch": 0.08395004625346901,
+      "grad_norm": 3.818544626235962,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 1.0903,
+      "step": 1089
+    },
+    {
+      "epoch": 0.08402713536848597,
+      "grad_norm": 4.31704044342041,
+      "learning_rate": 9.923821820979695e-06,
+      "loss": 1.2734,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08410422448350292,
+      "grad_norm": 4.051934719085693,
+      "learning_rate": 9.923604569923562e-06,
+      "loss": 1.304,
+      "step": 1091
+    },
+    {
+      "epoch": 0.0841813135985199,
+      "grad_norm": 3.77911376953125,
+      "learning_rate": 9.923387011906183e-06,
+      "loss": 1.1781,
+      "step": 1092
+    },
+    {
+      "epoch": 0.08425840271353685,
+      "grad_norm": 4.27567720413208,
+      "learning_rate": 9.923169146941121e-06,
+      "loss": 1.2181,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08433549182855381,
+      "grad_norm": 4.199497699737549,
+      "learning_rate": 9.922950975041963e-06,
+      "loss": 1.2001,
+      "step": 1094
+    },
+    {
+      "epoch": 0.08441258094357076,
+      "grad_norm": 4.348501682281494,
+      "learning_rate": 9.922732496222306e-06,
+      "loss": 1.3706,
+      "step": 1095
+    },
+    {
+      "epoch": 0.08448967005858772,
+      "grad_norm": 3.888430118560791,
+      "learning_rate": 9.922513710495774e-06,
+      "loss": 1.2277,
+      "step": 1096
+    },
+    {
+      "epoch": 0.08456675917360469,
+      "grad_norm": 4.729867935180664,
+      "learning_rate": 9.922294617876007e-06,
+      "loss": 1.3851,
+      "step": 1097
+    },
+    {
+      "epoch": 0.08464384828862165,
+      "grad_norm": 4.120565891265869,
+      "learning_rate": 9.922075218376664e-06,
+      "loss": 1.1222,
+      "step": 1098
+    },
+    {
+      "epoch": 0.0847209374036386,
+      "grad_norm": 4.8957839012146,
+      "learning_rate": 9.921855512011422e-06,
+      "loss": 1.2947,
+      "step": 1099
+    },
+    {
+      "epoch": 0.08479802651865556,
+      "grad_norm": 4.164621829986572,
+      "learning_rate": 9.921635498793983e-06,
+      "loss": 1.0555,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08487511563367253,
+      "grad_norm": 4.292223930358887,
+      "learning_rate": 9.921415178738056e-06,
+      "loss": 1.2633,
+      "step": 1101
+    },
+    {
+      "epoch": 0.08495220474868949,
+      "grad_norm": 4.393791198730469,
+      "learning_rate": 9.921194551857384e-06,
+      "loss": 1.2356,
+      "step": 1102
+    },
+    {
+      "epoch": 0.08502929386370645,
+      "grad_norm": 4.085275650024414,
+      "learning_rate": 9.920973618165719e-06,
+      "loss": 1.072,
+      "step": 1103
+    },
+    {
+      "epoch": 0.0851063829787234,
+      "grad_norm": 4.277721881866455,
+      "learning_rate": 9.920752377676835e-06,
+      "loss": 1.1315,
+      "step": 1104
+    },
+    {
+      "epoch": 0.08518347209374036,
+      "grad_norm": 4.160389423370361,
+      "learning_rate": 9.920530830404525e-06,
+      "loss": 1.2079,
+      "step": 1105
+    },
+    {
+      "epoch": 0.08526056120875733,
+      "grad_norm": 4.0644917488098145,
+      "learning_rate": 9.920308976362602e-06,
+      "loss": 1.1534,
+      "step": 1106
+    },
+    {
+      "epoch": 0.08533765032377429,
+      "grad_norm": 4.176297187805176,
+      "learning_rate": 9.920086815564898e-06,
+      "loss": 1.3915,
+      "step": 1107
+    },
+    {
+      "epoch": 0.08541473943879124,
+      "grad_norm": 4.304412841796875,
+      "learning_rate": 9.919864348025261e-06,
+      "loss": 1.275,
+      "step": 1108
+    },
+    {
+      "epoch": 0.0854918285538082,
+      "grad_norm": 4.072107791900635,
+      "learning_rate": 9.919641573757563e-06,
+      "loss": 1.1173,
+      "step": 1109
+    },
+    {
+      "epoch": 0.08556891766882516,
+      "grad_norm": 4.24763822555542,
+      "learning_rate": 9.919418492775694e-06,
+      "loss": 1.1268,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08564600678384213,
+      "grad_norm": 4.156879901885986,
+      "learning_rate": 9.91919510509356e-06,
+      "loss": 1.1882,
+      "step": 1111
+    },
+    {
+      "epoch": 0.08572309589885908,
+      "grad_norm": 4.036397933959961,
+      "learning_rate": 9.918971410725089e-06,
+      "loss": 1.2513,
+      "step": 1112
+    },
+    {
+      "epoch": 0.08580018501387604,
+      "grad_norm": 4.336670398712158,
+      "learning_rate": 9.918747409684225e-06,
+      "loss": 1.0309,
+      "step": 1113
+    },
+    {
+      "epoch": 0.085877274128893,
+      "grad_norm": 4.337256908416748,
+      "learning_rate": 9.918523101984933e-06,
+      "loss": 1.1959,
+      "step": 1114
+    },
+    {
+      "epoch": 0.08595436324390995,
+      "grad_norm": 3.9590811729431152,
+      "learning_rate": 9.918298487641202e-06,
+      "loss": 1.15,
+      "step": 1115
+    },
+    {
+      "epoch": 0.08603145235892692,
+      "grad_norm": 4.2955522537231445,
+      "learning_rate": 9.918073566667033e-06,
+      "loss": 1.1446,
+      "step": 1116
+    },
+    {
+      "epoch": 0.08610854147394388,
+      "grad_norm": 4.317296028137207,
+      "learning_rate": 9.917848339076448e-06,
+      "loss": 1.2451,
+      "step": 1117
+    },
+    {
+      "epoch": 0.08618563058896084,
+      "grad_norm": 4.0374755859375,
+      "learning_rate": 9.917622804883488e-06,
+      "loss": 1.1549,
+      "step": 1118
+    },
+    {
+      "epoch": 0.0862627197039778,
+      "grad_norm": 3.970003366470337,
+      "learning_rate": 9.917396964102218e-06,
+      "loss": 1.123,
+      "step": 1119
+    },
+    {
+      "epoch": 0.08633980881899475,
+      "grad_norm": 3.7095046043395996,
+      "learning_rate": 9.917170816746713e-06,
+      "loss": 1.1372,
+      "step": 1120
+    },
+    {
+      "epoch": 0.08641689793401172,
+      "grad_norm": 4.35107946395874,
+      "learning_rate": 9.916944362831073e-06,
+      "loss": 1.2134,
+      "step": 1121
+    },
+    {
+      "epoch": 0.08649398704902868,
+      "grad_norm": 4.03241491317749,
+      "learning_rate": 9.91671760236942e-06,
+      "loss": 1.119,
+      "step": 1122
+    },
+    {
+      "epoch": 0.08657107616404563,
+      "grad_norm": 3.650089979171753,
+      "learning_rate": 9.916490535375889e-06,
+      "loss": 1.1689,
+      "step": 1123
+    },
+    {
+      "epoch": 0.08664816527906259,
+      "grad_norm": 4.5127787590026855,
+      "learning_rate": 9.916263161864634e-06,
+      "loss": 1.2254,
+      "step": 1124
+    },
+    {
+      "epoch": 0.08672525439407956,
+      "grad_norm": 4.502647399902344,
+      "learning_rate": 9.916035481849834e-06,
+      "loss": 1.3046,
+      "step": 1125
+    },
+    {
+      "epoch": 0.08680234350909652,
+      "grad_norm": 4.246219635009766,
+      "learning_rate": 9.915807495345682e-06,
+      "loss": 1.1592,
+      "step": 1126
+    },
+    {
+      "epoch": 0.08687943262411348,
+      "grad_norm": 3.7990305423736572,
+      "learning_rate": 9.915579202366393e-06,
+      "loss": 1.2351,
+      "step": 1127
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 3.929375648498535,
+      "learning_rate": 9.915350602926198e-06,
+      "loss": 1.1652,
+      "step": 1128
+    },
+    {
+      "epoch": 0.08703361085414739,
+      "grad_norm": 3.9628517627716064,
+      "learning_rate": 9.915121697039352e-06,
+      "loss": 1.0718,
+      "step": 1129
+    },
+    {
+      "epoch": 0.08711069996916436,
+      "grad_norm": 4.691619396209717,
+      "learning_rate": 9.914892484720124e-06,
+      "loss": 1.2074,
+      "step": 1130
+    },
+    {
+      "epoch": 0.08718778908418132,
+      "grad_norm": 3.8248648643493652,
+      "learning_rate": 9.914662965982803e-06,
+      "loss": 1.0582,
+      "step": 1131
+    },
+    {
+      "epoch": 0.08726487819919827,
+      "grad_norm": 4.224147319793701,
+      "learning_rate": 9.914433140841702e-06,
+      "loss": 1.1567,
+      "step": 1132
+    },
+    {
+      "epoch": 0.08734196731421523,
+      "grad_norm": 4.157731533050537,
+      "learning_rate": 9.914203009311146e-06,
+      "loss": 1.1007,
+      "step": 1133
+    },
+    {
+      "epoch": 0.08741905642923219,
+      "grad_norm": 4.48982048034668,
+      "learning_rate": 9.913972571405482e-06,
+      "loss": 1.3227,
+      "step": 1134
+    },
+    {
+      "epoch": 0.08749614554424916,
+      "grad_norm": 3.823336362838745,
+      "learning_rate": 9.913741827139081e-06,
+      "loss": 1.2045,
+      "step": 1135
+    },
+    {
+      "epoch": 0.08757323465926611,
+      "grad_norm": 4.46881103515625,
+      "learning_rate": 9.913510776526324e-06,
+      "loss": 1.2047,
+      "step": 1136
+    },
+    {
+      "epoch": 0.08765032377428307,
+      "grad_norm": 4.13620662689209,
+      "learning_rate": 9.913279419581619e-06,
+      "loss": 1.2527,
+      "step": 1137
+    },
+    {
+      "epoch": 0.08772741288930003,
+      "grad_norm": 3.984325408935547,
+      "learning_rate": 9.913047756319388e-06,
+      "loss": 1.1912,
+      "step": 1138
+    },
+    {
+      "epoch": 0.08780450200431698,
+      "grad_norm": 4.111431121826172,
+      "learning_rate": 9.912815786754075e-06,
+      "loss": 1.2131,
+      "step": 1139
+    },
+    {
+      "epoch": 0.08788159111933395,
+      "grad_norm": 3.925600290298462,
+      "learning_rate": 9.912583510900142e-06,
+      "loss": 1.0794,
+      "step": 1140
+    },
+    {
+      "epoch": 0.08795868023435091,
+      "grad_norm": 3.7339839935302734,
+      "learning_rate": 9.91235092877207e-06,
+      "loss": 1.1662,
+      "step": 1141
+    },
+    {
+      "epoch": 0.08803576934936787,
+      "grad_norm": 3.6147632598876953,
+      "learning_rate": 9.912118040384358e-06,
+      "loss": 1.0365,
+      "step": 1142
+    },
+    {
+      "epoch": 0.08811285846438482,
+      "grad_norm": 4.1122355461120605,
+      "learning_rate": 9.911884845751529e-06,
+      "loss": 1.2303,
+      "step": 1143
+    },
+    {
+      "epoch": 0.0881899475794018,
+      "grad_norm": 4.269367218017578,
+      "learning_rate": 9.911651344888117e-06,
+      "loss": 1.217,
+      "step": 1144
+    },
+    {
+      "epoch": 0.08826703669441875,
+      "grad_norm": 3.916430711746216,
+      "learning_rate": 9.911417537808684e-06,
+      "loss": 1.1381,
+      "step": 1145
+    },
+    {
+      "epoch": 0.08834412580943571,
+      "grad_norm": 3.4542276859283447,
+      "learning_rate": 9.911183424527802e-06,
+      "loss": 1.0333,
+      "step": 1146
+    },
+    {
+      "epoch": 0.08842121492445267,
+      "grad_norm": 3.716357707977295,
+      "learning_rate": 9.91094900506007e-06,
+      "loss": 1.228,
+      "step": 1147
+    },
+    {
+      "epoch": 0.08849830403946962,
+      "grad_norm": 4.024401664733887,
+      "learning_rate": 9.910714279420103e-06,
+      "loss": 1.1965,
+      "step": 1148
+    },
+    {
+      "epoch": 0.08857539315448659,
+      "grad_norm": 3.9577510356903076,
+      "learning_rate": 9.910479247622534e-06,
+      "loss": 1.1488,
+      "step": 1149
+    },
+    {
+      "epoch": 0.08865248226950355,
+      "grad_norm": 4.212368011474609,
+      "learning_rate": 9.910243909682014e-06,
+      "loss": 1.2137,
+      "step": 1150
+    },
+    {
+      "epoch": 0.0887295713845205,
+      "grad_norm": 3.935518503189087,
+      "learning_rate": 9.910008265613219e-06,
+      "loss": 1.0983,
+      "step": 1151
+    },
+    {
+      "epoch": 0.08880666049953746,
+      "grad_norm": 4.040730953216553,
+      "learning_rate": 9.909772315430837e-06,
+      "loss": 1.1936,
+      "step": 1152
+    },
+    {
+      "epoch": 0.08888374961455442,
+      "grad_norm": 3.9304912090301514,
+      "learning_rate": 9.90953605914958e-06,
+      "loss": 1.2163,
+      "step": 1153
+    },
+    {
+      "epoch": 0.08896083872957139,
+      "grad_norm": 3.890301465988159,
+      "learning_rate": 9.909299496784177e-06,
+      "loss": 1.2424,
+      "step": 1154
+    },
+    {
+      "epoch": 0.08903792784458835,
+      "grad_norm": 3.953073263168335,
+      "learning_rate": 9.909062628349375e-06,
+      "loss": 1.2468,
+      "step": 1155
+    },
+    {
+      "epoch": 0.0891150169596053,
+      "grad_norm": 4.048395156860352,
+      "learning_rate": 9.908825453859944e-06,
+      "loss": 1.2238,
+      "step": 1156
+    },
+    {
+      "epoch": 0.08919210607462226,
+      "grad_norm": 3.6528916358947754,
+      "learning_rate": 9.90858797333067e-06,
+      "loss": 1.1587,
+      "step": 1157
+    },
+    {
+      "epoch": 0.08926919518963922,
+      "grad_norm": 3.6531436443328857,
+      "learning_rate": 9.908350186776357e-06,
+      "loss": 1.1732,
+      "step": 1158
+    },
+    {
+      "epoch": 0.08934628430465619,
+      "grad_norm": 4.267482280731201,
+      "learning_rate": 9.908112094211831e-06,
+      "loss": 1.259,
+      "step": 1159
+    },
+    {
+      "epoch": 0.08942337341967314,
+      "grad_norm": 4.037517547607422,
+      "learning_rate": 9.907873695651935e-06,
+      "loss": 1.1787,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0895004625346901,
+      "grad_norm": 3.696422815322876,
+      "learning_rate": 9.907634991111535e-06,
+      "loss": 1.0961,
+      "step": 1161
+    },
+    {
+      "epoch": 0.08957755164970706,
+      "grad_norm": 4.135885238647461,
+      "learning_rate": 9.907395980605509e-06,
+      "loss": 1.2558,
+      "step": 1162
+    },
+    {
+      "epoch": 0.08965464076472403,
+      "grad_norm": 4.450069904327393,
+      "learning_rate": 9.907156664148761e-06,
+      "loss": 1.2009,
+      "step": 1163
+    },
+    {
+      "epoch": 0.08973172987974098,
+      "grad_norm": 3.922731876373291,
+      "learning_rate": 9.906917041756208e-06,
+      "loss": 1.1875,
+      "step": 1164
+    },
+    {
+      "epoch": 0.08980881899475794,
+      "grad_norm": 4.55860710144043,
+      "learning_rate": 9.90667711344279e-06,
+      "loss": 1.2364,
+      "step": 1165
+    },
+    {
+      "epoch": 0.0898859081097749,
+      "grad_norm": 4.525522232055664,
+      "learning_rate": 9.90643687922347e-06,
+      "loss": 1.2342,
+      "step": 1166
+    },
+    {
+      "epoch": 0.08996299722479185,
+      "grad_norm": 4.051623344421387,
+      "learning_rate": 9.90619633911322e-06,
+      "loss": 1.2236,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09004008633980883,
+      "grad_norm": 3.817911148071289,
+      "learning_rate": 9.905955493127037e-06,
+      "loss": 1.0846,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09011717545482578,
+      "grad_norm": 3.949352979660034,
+      "learning_rate": 9.905714341279938e-06,
+      "loss": 1.1237,
+      "step": 1169
+    },
+    {
+      "epoch": 0.09019426456984274,
+      "grad_norm": 3.8742692470550537,
+      "learning_rate": 9.905472883586958e-06,
+      "loss": 1.0737,
+      "step": 1170
+    },
+    {
+      "epoch": 0.0902713536848597,
+      "grad_norm": 4.435111999511719,
+      "learning_rate": 9.905231120063149e-06,
+      "loss": 1.3749,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09034844279987665,
+      "grad_norm": 3.793482780456543,
+      "learning_rate": 9.904989050723583e-06,
+      "loss": 1.1484,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09042553191489362,
+      "grad_norm": 4.095993995666504,
+      "learning_rate": 9.904746675583356e-06,
+      "loss": 1.1579,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09050262102991058,
+      "grad_norm": 4.051393032073975,
+      "learning_rate": 9.904503994657574e-06,
+      "loss": 1.1387,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09057971014492754,
+      "grad_norm": 3.8025002479553223,
+      "learning_rate": 9.90426100796137e-06,
+      "loss": 1.0579,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09065679925994449,
+      "grad_norm": 4.1807708740234375,
+      "learning_rate": 9.904017715509893e-06,
+      "loss": 1.2583,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09073388837496145,
+      "grad_norm": 3.890801191329956,
+      "learning_rate": 9.90377411731831e-06,
+      "loss": 1.1381,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09081097748997842,
+      "grad_norm": 3.741750955581665,
+      "learning_rate": 9.903530213401806e-06,
+      "loss": 1.1599,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09088806660499538,
+      "grad_norm": 4.385606288909912,
+      "learning_rate": 9.903286003775592e-06,
+      "loss": 1.2269,
+      "step": 1179
+    },
+    {
+      "epoch": 0.09096515572001233,
+      "grad_norm": 4.160192012786865,
+      "learning_rate": 9.903041488454888e-06,
+      "loss": 1.0707,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09104224483502929,
+      "grad_norm": 4.092873573303223,
+      "learning_rate": 9.90279666745494e-06,
+      "loss": 1.2221,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09111933395004625,
+      "grad_norm": 4.053955078125,
+      "learning_rate": 9.902551540791016e-06,
+      "loss": 1.1687,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09119642306506322,
+      "grad_norm": 4.5065789222717285,
+      "learning_rate": 9.902306108478393e-06,
+      "loss": 1.004,
+      "step": 1183
+    },
+    {
+      "epoch": 0.09127351218008017,
+      "grad_norm": 4.21845006942749,
+      "learning_rate": 9.902060370532371e-06,
+      "loss": 1.2317,
+      "step": 1184
+    },
+    {
+      "epoch": 0.09135060129509713,
+      "grad_norm": 3.860323190689087,
+      "learning_rate": 9.901814326968276e-06,
+      "loss": 1.1142,
+      "step": 1185
+    },
+    {
+      "epoch": 0.09142769041011409,
+      "grad_norm": 4.450876712799072,
+      "learning_rate": 9.901567977801444e-06,
+      "loss": 1.1097,
+      "step": 1186
+    },
+    {
+      "epoch": 0.09150477952513106,
+      "grad_norm": 3.953404664993286,
+      "learning_rate": 9.901321323047235e-06,
+      "loss": 1.0921,
+      "step": 1187
+    },
+    {
+      "epoch": 0.09158186864014801,
+      "grad_norm": 3.6998918056488037,
+      "learning_rate": 9.901074362721024e-06,
+      "loss": 1.2101,
+      "step": 1188
+    },
+    {
+      "epoch": 0.09165895775516497,
+      "grad_norm": 3.8323581218719482,
+      "learning_rate": 9.900827096838213e-06,
+      "loss": 1.1592,
+      "step": 1189
+    },
+    {
+      "epoch": 0.09173604687018193,
+      "grad_norm": 3.705106496810913,
+      "learning_rate": 9.900579525414213e-06,
+      "loss": 1.1281,
+      "step": 1190
+    },
+    {
+      "epoch": 0.09181313598519888,
+      "grad_norm": 3.9382426738739014,
+      "learning_rate": 9.900331648464459e-06,
+      "loss": 1.1911,
+      "step": 1191
+    },
+    {
+      "epoch": 0.09189022510021586,
+      "grad_norm": 4.3277740478515625,
+      "learning_rate": 9.900083466004409e-06,
+      "loss": 1.2685,
+      "step": 1192
+    },
+    {
+      "epoch": 0.09196731421523281,
+      "grad_norm": 3.8461456298828125,
+      "learning_rate": 9.899834978049531e-06,
+      "loss": 1.2131,
+      "step": 1193
+    },
+    {
+      "epoch": 0.09204440333024977,
+      "grad_norm": 4.075659275054932,
+      "learning_rate": 9.89958618461532e-06,
+      "loss": 1.171,
+      "step": 1194
+    },
+    {
+      "epoch": 0.09212149244526673,
+      "grad_norm": 3.7878012657165527,
+      "learning_rate": 9.899337085717284e-06,
+      "loss": 1.1431,
+      "step": 1195
+    },
+    {
+      "epoch": 0.09219858156028368,
+      "grad_norm": 3.931734800338745,
+      "learning_rate": 9.899087681370958e-06,
+      "loss": 1.2347,
+      "step": 1196
+    },
+    {
+      "epoch": 0.09227567067530065,
+      "grad_norm": 4.06639289855957,
+      "learning_rate": 9.898837971591885e-06,
+      "loss": 1.2163,
+      "step": 1197
+    },
+    {
+      "epoch": 0.09235275979031761,
+      "grad_norm": 3.754260301589966,
+      "learning_rate": 9.89858795639564e-06,
+      "loss": 1.1521,
+      "step": 1198
+    },
+    {
+      "epoch": 0.09242984890533457,
+      "grad_norm": 4.012813568115234,
+      "learning_rate": 9.898337635797803e-06,
+      "loss": 1.1976,
+      "step": 1199
+    },
+    {
+      "epoch": 0.09250693802035152,
+      "grad_norm": 3.8789308071136475,
+      "learning_rate": 9.898087009813985e-06,
+      "loss": 1.1611,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09258402713536848,
+      "grad_norm": 3.8602101802825928,
+      "learning_rate": 9.89783607845981e-06,
+      "loss": 1.2561,
+      "step": 1201
+    },
+    {
+      "epoch": 0.09266111625038545,
+      "grad_norm": 3.982560157775879,
+      "learning_rate": 9.897584841750922e-06,
+      "loss": 1.2404,
+      "step": 1202
+    },
+    {
+      "epoch": 0.0927382053654024,
+      "grad_norm": 4.090329170227051,
+      "learning_rate": 9.897333299702982e-06,
+      "loss": 1.1678,
+      "step": 1203
+    },
+    {
+      "epoch": 0.09281529448041936,
+      "grad_norm": 4.530256271362305,
+      "learning_rate": 9.897081452331677e-06,
+      "loss": 1.2437,
+      "step": 1204
+    },
+    {
+      "epoch": 0.09289238359543632,
+      "grad_norm": 4.283809185028076,
+      "learning_rate": 9.896829299652705e-06,
+      "loss": 1.1977,
+      "step": 1205
+    },
+    {
+      "epoch": 0.09296947271045329,
+      "grad_norm": 3.9148154258728027,
+      "learning_rate": 9.896576841681792e-06,
+      "loss": 1.2863,
+      "step": 1206
+    },
+    {
+      "epoch": 0.09304656182547025,
+      "grad_norm": 3.6620705127716064,
+      "learning_rate": 9.896324078434668e-06,
+      "loss": 1.1627,
+      "step": 1207
+    },
+    {
+      "epoch": 0.0931236509404872,
+      "grad_norm": 4.204239368438721,
+      "learning_rate": 9.896071009927098e-06,
+      "loss": 1.166,
+      "step": 1208
+    },
+    {
+      "epoch": 0.09320074005550416,
+      "grad_norm": 4.137142658233643,
+      "learning_rate": 9.895817636174857e-06,
+      "loss": 1.2225,
+      "step": 1209
+    },
+    {
+      "epoch": 0.09327782917052112,
+      "grad_norm": 4.27466344833374,
+      "learning_rate": 9.895563957193744e-06,
+      "loss": 1.1555,
+      "step": 1210
+    },
+    {
+      "epoch": 0.09335491828553809,
+      "grad_norm": 4.381038188934326,
+      "learning_rate": 9.89530997299957e-06,
+      "loss": 1.2687,
+      "step": 1211
+    },
+    {
+      "epoch": 0.09343200740055504,
+      "grad_norm": 3.764836549758911,
+      "learning_rate": 9.895055683608175e-06,
+      "loss": 1.0278,
+      "step": 1212
+    },
+    {
+      "epoch": 0.093509096515572,
+      "grad_norm": 4.309133529663086,
+      "learning_rate": 9.89480108903541e-06,
+      "loss": 1.217,
+      "step": 1213
+    },
+    {
+      "epoch": 0.09358618563058896,
+      "grad_norm": 4.008470058441162,
+      "learning_rate": 9.894546189297148e-06,
+      "loss": 1.1429,
+      "step": 1214
+    },
+    {
+      "epoch": 0.09366327474560592,
+      "grad_norm": 3.7469325065612793,
+      "learning_rate": 9.894290984409281e-06,
+      "loss": 1.1331,
+      "step": 1215
+    },
+    {
+      "epoch": 0.09374036386062289,
+      "grad_norm": 4.368958473205566,
+      "learning_rate": 9.894035474387719e-06,
+      "loss": 1.2149,
+      "step": 1216
+    },
+    {
+      "epoch": 0.09381745297563984,
+      "grad_norm": 4.350530624389648,
+      "learning_rate": 9.893779659248393e-06,
+      "loss": 1.2107,
+      "step": 1217
+    },
+    {
+      "epoch": 0.0938945420906568,
+      "grad_norm": 3.9404361248016357,
+      "learning_rate": 9.893523539007248e-06,
+      "loss": 1.1363,
+      "step": 1218
+    },
+    {
+      "epoch": 0.09397163120567376,
+      "grad_norm": 4.558434009552002,
+      "learning_rate": 9.893267113680257e-06,
+      "loss": 1.1958,
+      "step": 1219
+    },
+    {
+      "epoch": 0.09404872032069071,
+      "grad_norm": 4.263239860534668,
+      "learning_rate": 9.893010383283404e-06,
+      "loss": 1.2427,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09412580943570768,
+      "grad_norm": 4.081235408782959,
+      "learning_rate": 9.892753347832695e-06,
+      "loss": 1.1318,
+      "step": 1221
+    },
+    {
+      "epoch": 0.09420289855072464,
+      "grad_norm": 3.902259111404419,
+      "learning_rate": 9.892496007344155e-06,
+      "loss": 1.225,
+      "step": 1222
+    },
+    {
+      "epoch": 0.0942799876657416,
+      "grad_norm": 4.030986785888672,
+      "learning_rate": 9.892238361833826e-06,
+      "loss": 1.1857,
+      "step": 1223
+    },
+    {
+      "epoch": 0.09435707678075855,
+      "grad_norm": 4.28933048248291,
+      "learning_rate": 9.891980411317774e-06,
+      "loss": 1.2686,
+      "step": 1224
+    },
+    {
+      "epoch": 0.09443416589577551,
+      "grad_norm": 3.7828292846679688,
+      "learning_rate": 9.89172215581208e-06,
+      "loss": 1.1151,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09451125501079248,
+      "grad_norm": 3.7171690464019775,
+      "learning_rate": 9.891463595332844e-06,
+      "loss": 1.18,
+      "step": 1226
+    },
+    {
+      "epoch": 0.09458834412580944,
+      "grad_norm": 4.4071478843688965,
+      "learning_rate": 9.891204729896187e-06,
+      "loss": 1.1709,
+      "step": 1227
+    },
+    {
+      "epoch": 0.0946654332408264,
+      "grad_norm": 3.7841458320617676,
+      "learning_rate": 9.890945559518247e-06,
+      "loss": 1.234,
+      "step": 1228
+    },
+    {
+      "epoch": 0.09474252235584335,
+      "grad_norm": 4.163311004638672,
+      "learning_rate": 9.890686084215182e-06,
+      "loss": 1.1784,
+      "step": 1229
+    },
+    {
+      "epoch": 0.09481961147086032,
+      "grad_norm": 4.381045341491699,
+      "learning_rate": 9.89042630400317e-06,
+      "loss": 1.1428,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09489670058587728,
+      "grad_norm": 4.111490726470947,
+      "learning_rate": 9.890166218898405e-06,
+      "loss": 1.2638,
+      "step": 1231
+    },
+    {
+      "epoch": 0.09497378970089423,
+      "grad_norm": 4.06538200378418,
+      "learning_rate": 9.889905828917103e-06,
+      "loss": 1.2292,
+      "step": 1232
+    },
+    {
+      "epoch": 0.09505087881591119,
+      "grad_norm": 3.760375499725342,
+      "learning_rate": 9.8896451340755e-06,
+      "loss": 1.1367,
+      "step": 1233
+    },
+    {
+      "epoch": 0.09512796793092815,
+      "grad_norm": 3.769625186920166,
+      "learning_rate": 9.889384134389844e-06,
+      "loss": 1.2137,
+      "step": 1234
+    },
+    {
+      "epoch": 0.09520505704594512,
+      "grad_norm": 4.986996650695801,
+      "learning_rate": 9.889122829876412e-06,
+      "loss": 1.2135,
+      "step": 1235
+    },
+    {
+      "epoch": 0.09528214616096208,
+      "grad_norm": 3.983733654022217,
+      "learning_rate": 9.888861220551494e-06,
+      "loss": 1.1738,
+      "step": 1236
+    },
+    {
+      "epoch": 0.09535923527597903,
+      "grad_norm": 3.9600632190704346,
+      "learning_rate": 9.888599306431397e-06,
+      "loss": 1.0924,
+      "step": 1237
+    },
+    {
+      "epoch": 0.09543632439099599,
+      "grad_norm": 3.82342791557312,
+      "learning_rate": 9.888337087532452e-06,
+      "loss": 1.2129,
+      "step": 1238
+    },
+    {
+      "epoch": 0.09551341350601295,
+      "grad_norm": 4.06931734085083,
+      "learning_rate": 9.888074563871007e-06,
+      "loss": 1.0888,
+      "step": 1239
+    },
+    {
+      "epoch": 0.09559050262102992,
+      "grad_norm": 3.582798719406128,
+      "learning_rate": 9.88781173546343e-06,
+      "loss": 1.0759,
+      "step": 1240
+    },
+    {
+      "epoch": 0.09566759173604687,
+      "grad_norm": 4.151041507720947,
+      "learning_rate": 9.887548602326104e-06,
+      "loss": 1.09,
+      "step": 1241
+    },
+    {
+      "epoch": 0.09574468085106383,
+      "grad_norm": 4.339078426361084,
+      "learning_rate": 9.887285164475438e-06,
+      "loss": 1.1669,
+      "step": 1242
+    },
+    {
+      "epoch": 0.09582176996608079,
+      "grad_norm": 4.104028701782227,
+      "learning_rate": 9.887021421927853e-06,
+      "loss": 1.1578,
+      "step": 1243
+    },
+    {
+      "epoch": 0.09589885908109774,
+      "grad_norm": 4.224393844604492,
+      "learning_rate": 9.886757374699792e-06,
+      "loss": 1.1235,
+      "step": 1244
+    },
+    {
+      "epoch": 0.09597594819611471,
+      "grad_norm": 3.995770215988159,
+      "learning_rate": 9.88649302280772e-06,
+      "loss": 1.1039,
+      "step": 1245
+    },
+    {
+      "epoch": 0.09605303731113167,
+      "grad_norm": 4.252878665924072,
+      "learning_rate": 9.886228366268114e-06,
+      "loss": 1.1712,
+      "step": 1246
+    },
+    {
+      "epoch": 0.09613012642614863,
+      "grad_norm": 4.067622661590576,
+      "learning_rate": 9.885963405097477e-06,
+      "loss": 1.2701,
+      "step": 1247
+    },
+    {
+      "epoch": 0.09620721554116558,
+      "grad_norm": 4.1838765144348145,
+      "learning_rate": 9.885698139312326e-06,
+      "loss": 1.189,
+      "step": 1248
+    },
+    {
+      "epoch": 0.09628430465618255,
+      "grad_norm": 3.9121227264404297,
+      "learning_rate": 9.8854325689292e-06,
+      "loss": 1.1588,
+      "step": 1249
+    },
+    {
+      "epoch": 0.09636139377119951,
+      "grad_norm": 4.973738193511963,
+      "learning_rate": 9.885166693964654e-06,
+      "loss": 1.3419,
+      "step": 1250
+    },
+    {
+      "epoch": 0.09643848288621647,
+      "grad_norm": 3.7892560958862305,
+      "learning_rate": 9.884900514435266e-06,
+      "loss": 1.1613,
+      "step": 1251
+    },
+    {
+      "epoch": 0.09651557200123342,
+      "grad_norm": 3.8883039951324463,
+      "learning_rate": 9.884634030357634e-06,
+      "loss": 1.0897,
+      "step": 1252
+    },
+    {
+      "epoch": 0.09659266111625038,
+      "grad_norm": 3.8611433506011963,
+      "learning_rate": 9.884367241748364e-06,
+      "loss": 1.1545,
+      "step": 1253
+    },
+    {
+      "epoch": 0.09666975023126735,
+      "grad_norm": 3.8657636642456055,
+      "learning_rate": 9.884100148624096e-06,
+      "loss": 1.2268,
+      "step": 1254
+    },
+    {
+      "epoch": 0.09674683934628431,
+      "grad_norm": 3.610783576965332,
+      "learning_rate": 9.883832751001479e-06,
+      "loss": 1.0718,
+      "step": 1255
+    },
+    {
+      "epoch": 0.09682392846130126,
+      "grad_norm": 3.7750887870788574,
+      "learning_rate": 9.883565048897183e-06,
+      "loss": 1.1832,
+      "step": 1256
+    },
+    {
+      "epoch": 0.09690101757631822,
+      "grad_norm": 4.035364627838135,
+      "learning_rate": 9.883297042327899e-06,
+      "loss": 1.2015,
+      "step": 1257
+    },
+    {
+      "epoch": 0.09697810669133518,
+      "grad_norm": 3.7171030044555664,
+      "learning_rate": 9.883028731310335e-06,
+      "loss": 1.1283,
+      "step": 1258
+    },
+    {
+      "epoch": 0.09705519580635215,
+      "grad_norm": 3.9022812843322754,
+      "learning_rate": 9.88276011586122e-06,
+      "loss": 1.18,
+      "step": 1259
+    },
+    {
+      "epoch": 0.0971322849213691,
+      "grad_norm": 4.099169731140137,
+      "learning_rate": 9.882491195997301e-06,
+      "loss": 1.1186,
+      "step": 1260
+    },
+    {
+      "epoch": 0.09720937403638606,
+      "grad_norm": 4.0367326736450195,
+      "learning_rate": 9.882221971735343e-06,
+      "loss": 1.1546,
+      "step": 1261
+    },
+    {
+      "epoch": 0.09728646315140302,
+      "grad_norm": 3.8437790870666504,
+      "learning_rate": 9.88195244309213e-06,
+      "loss": 1.1853,
+      "step": 1262
+    },
+    {
+      "epoch": 0.09736355226641998,
+      "grad_norm": 3.9312708377838135,
+      "learning_rate": 9.881682610084467e-06,
+      "loss": 1.2333,
+      "step": 1263
+    },
+    {
+      "epoch": 0.09744064138143695,
+      "grad_norm": 4.007068157196045,
+      "learning_rate": 9.881412472729175e-06,
+      "loss": 1.1373,
+      "step": 1264
+    },
+    {
+      "epoch": 0.0975177304964539,
+      "grad_norm": 4.4400763511657715,
+      "learning_rate": 9.881142031043098e-06,
+      "loss": 1.3182,
+      "step": 1265
+    },
+    {
+      "epoch": 0.09759481961147086,
+      "grad_norm": 4.121272563934326,
+      "learning_rate": 9.880871285043095e-06,
+      "loss": 1.301,
+      "step": 1266
+    },
+    {
+      "epoch": 0.09767190872648782,
+      "grad_norm": 3.7698986530303955,
+      "learning_rate": 9.880600234746047e-06,
+      "loss": 1.132,
+      "step": 1267
+    },
+    {
+      "epoch": 0.09774899784150477,
+      "grad_norm": 4.014803886413574,
+      "learning_rate": 9.88032888016885e-06,
+      "loss": 1.1003,
+      "step": 1268
+    },
+    {
+      "epoch": 0.09782608695652174,
+      "grad_norm": 3.873784303665161,
+      "learning_rate": 9.880057221328425e-06,
+      "loss": 1.1009,
+      "step": 1269
+    },
+    {
+      "epoch": 0.0979031760715387,
+      "grad_norm": 4.115602970123291,
+      "learning_rate": 9.879785258241705e-06,
+      "loss": 1.3411,
+      "step": 1270
+    },
+    {
+      "epoch": 0.09798026518655566,
+      "grad_norm": 3.987370491027832,
+      "learning_rate": 9.879512990925648e-06,
+      "loss": 1.1694,
+      "step": 1271
+    },
+    {
+      "epoch": 0.09805735430157261,
+      "grad_norm": 3.914703607559204,
+      "learning_rate": 9.879240419397227e-06,
+      "loss": 1.2862,
+      "step": 1272
+    },
+    {
+      "epoch": 0.09813444341658958,
+      "grad_norm": 4.038669586181641,
+      "learning_rate": 9.878967543673436e-06,
+      "loss": 1.2318,
+      "step": 1273
+    },
+    {
+      "epoch": 0.09821153253160654,
+      "grad_norm": 3.7776105403900146,
+      "learning_rate": 9.878694363771289e-06,
+      "loss": 1.0864,
+      "step": 1274
+    },
+    {
+      "epoch": 0.0982886216466235,
+      "grad_norm": 3.926527976989746,
+      "learning_rate": 9.878420879707816e-06,
+      "loss": 1.1557,
+      "step": 1275
+    },
+    {
+      "epoch": 0.09836571076164045,
+      "grad_norm": 4.13185977935791,
+      "learning_rate": 9.878147091500065e-06,
+      "loss": 1.2331,
+      "step": 1276
+    },
+    {
+      "epoch": 0.09844279987665741,
+      "grad_norm": 4.489853858947754,
+      "learning_rate": 9.877872999165109e-06,
+      "loss": 1.2494,
+      "step": 1277
+    },
+    {
+      "epoch": 0.09851988899167438,
+      "grad_norm": 4.077756404876709,
+      "learning_rate": 9.877598602720034e-06,
+      "loss": 1.1705,
+      "step": 1278
+    },
+    {
+      "epoch": 0.09859697810669134,
+      "grad_norm": 4.43337345123291,
+      "learning_rate": 9.877323902181949e-06,
+      "loss": 1.1507,
+      "step": 1279
+    },
+    {
+      "epoch": 0.0986740672217083,
+      "grad_norm": 4.278952598571777,
+      "learning_rate": 9.877048897567977e-06,
+      "loss": 1.1005,
+      "step": 1280
+    },
+    {
+      "epoch": 0.09875115633672525,
+      "grad_norm": 3.6082239151000977,
+      "learning_rate": 9.876773588895265e-06,
+      "loss": 1.1287,
+      "step": 1281
+    },
+    {
+      "epoch": 0.09882824545174221,
+      "grad_norm": 3.5894458293914795,
+      "learning_rate": 9.876497976180978e-06,
+      "loss": 1.1771,
+      "step": 1282
+    },
+    {
+      "epoch": 0.09890533456675918,
+      "grad_norm": 4.010733127593994,
+      "learning_rate": 9.8762220594423e-06,
+      "loss": 1.0699,
+      "step": 1283
+    },
+    {
+      "epoch": 0.09898242368177614,
+      "grad_norm": 3.891744613647461,
+      "learning_rate": 9.87594583869643e-06,
+      "loss": 1.1908,
+      "step": 1284
+    },
+    {
+      "epoch": 0.09905951279679309,
+      "grad_norm": 4.345311164855957,
+      "learning_rate": 9.875669313960588e-06,
+      "loss": 1.2161,
+      "step": 1285
+    },
+    {
+      "epoch": 0.09913660191181005,
+      "grad_norm": 3.8743138313293457,
+      "learning_rate": 9.87539248525202e-06,
+      "loss": 1.2969,
+      "step": 1286
+    },
+    {
+      "epoch": 0.099213691026827,
+      "grad_norm": 4.233188152313232,
+      "learning_rate": 9.875115352587977e-06,
+      "loss": 1.1773,
+      "step": 1287
+    },
+    {
+      "epoch": 0.09929078014184398,
+      "grad_norm": 3.835361957550049,
+      "learning_rate": 9.874837915985743e-06,
+      "loss": 1.2376,
+      "step": 1288
+    },
+    {
+      "epoch": 0.09936786925686093,
+      "grad_norm": 3.9066238403320312,
+      "learning_rate": 9.874560175462612e-06,
+      "loss": 1.2814,
+      "step": 1289
+    },
+    {
+      "epoch": 0.09944495837187789,
+      "grad_norm": 3.9215028285980225,
+      "learning_rate": 9.874282131035899e-06,
+      "loss": 1.1192,
+      "step": 1290
+    },
+    {
+      "epoch": 0.09952204748689485,
+      "grad_norm": 3.68405818939209,
+      "learning_rate": 9.87400378272294e-06,
+      "loss": 1.2173,
+      "step": 1291
+    },
+    {
+      "epoch": 0.09959913660191182,
+      "grad_norm": 3.8745553493499756,
+      "learning_rate": 9.87372513054109e-06,
+      "loss": 1.1877,
+      "step": 1292
+    },
+    {
+      "epoch": 0.09967622571692877,
+      "grad_norm": 4.057330131530762,
+      "learning_rate": 9.873446174507719e-06,
+      "loss": 1.2889,
+      "step": 1293
+    },
+    {
+      "epoch": 0.09975331483194573,
+      "grad_norm": 4.208259105682373,
+      "learning_rate": 9.873166914640218e-06,
+      "loss": 1.1993,
+      "step": 1294
+    },
+    {
+      "epoch": 0.09983040394696269,
+      "grad_norm": 3.965298652648926,
+      "learning_rate": 9.872887350956e-06,
+      "loss": 1.1844,
+      "step": 1295
+    },
+    {
+      "epoch": 0.09990749306197964,
+      "grad_norm": 4.23174524307251,
+      "learning_rate": 9.872607483472491e-06,
+      "loss": 1.1607,
+      "step": 1296
+    },
+    {
+      "epoch": 0.09998458217699661,
+      "grad_norm": 4.803770542144775,
+      "learning_rate": 9.872327312207145e-06,
+      "loss": 1.2843,
+      "step": 1297
+    },
+    {
+      "epoch": 0.10006167129201357,
+      "grad_norm": 4.673795223236084,
+      "learning_rate": 9.872046837177421e-06,
+      "loss": 1.2134,
+      "step": 1298
+    },
+    {
+      "epoch": 0.10013876040703053,
+      "grad_norm": 3.8976125717163086,
+      "learning_rate": 9.871766058400812e-06,
+      "loss": 1.1635,
+      "step": 1299
+    },
+    {
+      "epoch": 0.10021584952204748,
+      "grad_norm": 4.471933841705322,
+      "learning_rate": 9.87148497589482e-06,
+      "loss": 1.1844,
+      "step": 1300
+    },
+    {
+      "epoch": 0.10029293863706444,
+      "grad_norm": 4.1576457023620605,
+      "learning_rate": 9.871203589676971e-06,
+      "loss": 1.1461,
+      "step": 1301
+    },
+    {
+      "epoch": 0.10037002775208141,
+      "grad_norm": 3.978757381439209,
+      "learning_rate": 9.870921899764807e-06,
+      "loss": 1.2684,
+      "step": 1302
+    },
+    {
+      "epoch": 0.10044711686709837,
+      "grad_norm": 3.5319221019744873,
+      "learning_rate": 9.87063990617589e-06,
+      "loss": 1.1454,
+      "step": 1303
+    },
+    {
+      "epoch": 0.10052420598211532,
+      "grad_norm": 4.05499267578125,
+      "learning_rate": 9.870357608927798e-06,
+      "loss": 1.1962,
+      "step": 1304
+    },
+    {
+      "epoch": 0.10060129509713228,
+      "grad_norm": 4.118753433227539,
+      "learning_rate": 9.870075008038137e-06,
+      "loss": 1.1875,
+      "step": 1305
+    },
+    {
+      "epoch": 0.10067838421214924,
+      "grad_norm": 3.8087265491485596,
+      "learning_rate": 9.869792103524517e-06,
+      "loss": 1.1726,
+      "step": 1306
+    },
+    {
+      "epoch": 0.10075547332716621,
+      "grad_norm": 4.719503879547119,
+      "learning_rate": 9.869508895404584e-06,
+      "loss": 1.1398,
+      "step": 1307
+    },
+    {
+      "epoch": 0.10083256244218317,
+      "grad_norm": 4.060819625854492,
+      "learning_rate": 9.86922538369599e-06,
+      "loss": 1.2485,
+      "step": 1308
+    },
+    {
+      "epoch": 0.10090965155720012,
+      "grad_norm": 4.301494121551514,
+      "learning_rate": 9.868941568416413e-06,
+      "loss": 1.1975,
+      "step": 1309
+    },
+    {
+      "epoch": 0.10098674067221708,
+      "grad_norm": 3.7967967987060547,
+      "learning_rate": 9.868657449583547e-06,
+      "loss": 1.1155,
+      "step": 1310
+    },
+    {
+      "epoch": 0.10106382978723404,
+      "grad_norm": 5.051084995269775,
+      "learning_rate": 9.8683730272151e-06,
+      "loss": 1.2833,
+      "step": 1311
+    },
+    {
+      "epoch": 0.101140918902251,
+      "grad_norm": 5.0358381271362305,
+      "learning_rate": 9.868088301328813e-06,
+      "loss": 1.358,
+      "step": 1312
+    },
+    {
+      "epoch": 0.10121800801726796,
+      "grad_norm": 4.280052185058594,
+      "learning_rate": 9.867803271942432e-06,
+      "loss": 1.2046,
+      "step": 1313
+    },
+    {
+      "epoch": 0.10129509713228492,
+      "grad_norm": 3.971712350845337,
+      "learning_rate": 9.867517939073727e-06,
+      "loss": 1.1826,
+      "step": 1314
+    },
+    {
+      "epoch": 0.10137218624730188,
+      "grad_norm": 4.396337985992432,
+      "learning_rate": 9.867232302740489e-06,
+      "loss": 1.2291,
+      "step": 1315
+    },
+    {
+      "epoch": 0.10144927536231885,
+      "grad_norm": 3.7291693687438965,
+      "learning_rate": 9.866946362960526e-06,
+      "loss": 1.2395,
+      "step": 1316
+    },
+    {
+      "epoch": 0.1015263644773358,
+      "grad_norm": 3.65598726272583,
+      "learning_rate": 9.86666011975166e-06,
+      "loss": 1.2252,
+      "step": 1317
+    },
+    {
+      "epoch": 0.10160345359235276,
+      "grad_norm": 3.8488965034484863,
+      "learning_rate": 9.866373573131744e-06,
+      "loss": 1.132,
+      "step": 1318
+    },
+    {
+      "epoch": 0.10168054270736972,
+      "grad_norm": 4.217064380645752,
+      "learning_rate": 9.86608672311864e-06,
+      "loss": 1.2116,
+      "step": 1319
+    },
+    {
+      "epoch": 0.10175763182238667,
+      "grad_norm": 3.518402576446533,
+      "learning_rate": 9.865799569730227e-06,
+      "loss": 1.0713,
+      "step": 1320
+    },
+    {
+      "epoch": 0.10183472093740364,
+      "grad_norm": 4.360300064086914,
+      "learning_rate": 9.865512112984414e-06,
+      "loss": 1.2738,
+      "step": 1321
+    },
+    {
+      "epoch": 0.1019118100524206,
+      "grad_norm": 4.2740068435668945,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 1.3994,
+      "step": 1322
+    },
+    {
+      "epoch": 0.10198889916743756,
+      "grad_norm": 4.366950511932373,
+      "learning_rate": 9.864936289492285e-06,
+      "loss": 1.2464,
+      "step": 1323
+    },
+    {
+      "epoch": 0.10206598828245451,
+      "grad_norm": 4.6967997550964355,
+      "learning_rate": 9.864647922781868e-06,
+      "loss": 1.2328,
+      "step": 1324
+    },
+    {
+      "epoch": 0.10214307739747147,
+      "grad_norm": 3.914935827255249,
+      "learning_rate": 9.864359252785847e-06,
+      "loss": 1.2629,
+      "step": 1325
+    },
+    {
+      "epoch": 0.10222016651248844,
+      "grad_norm": 3.91147780418396,
+      "learning_rate": 9.864070279522222e-06,
+      "loss": 1.1347,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1022972556275054,
+      "grad_norm": 3.9424116611480713,
+      "learning_rate": 9.863781003009005e-06,
+      "loss": 1.1586,
+      "step": 1327
+    },
+    {
+      "epoch": 0.10237434474252236,
+      "grad_norm": 4.317541122436523,
+      "learning_rate": 9.863491423264232e-06,
+      "loss": 1.2135,
+      "step": 1328
+    },
+    {
+      "epoch": 0.10245143385753931,
+      "grad_norm": 3.8650455474853516,
+      "learning_rate": 9.86320154030596e-06,
+      "loss": 1.1562,
+      "step": 1329
+    },
+    {
+      "epoch": 0.10252852297255627,
+      "grad_norm": 4.095459461212158,
+      "learning_rate": 9.862911354152258e-06,
+      "loss": 1.0685,
+      "step": 1330
+    },
+    {
+      "epoch": 0.10260561208757324,
+      "grad_norm": 4.160573482513428,
+      "learning_rate": 9.862620864821218e-06,
+      "loss": 1.2398,
+      "step": 1331
+    },
+    {
+      "epoch": 0.1026827012025902,
+      "grad_norm": 4.309733867645264,
+      "learning_rate": 9.862330072330953e-06,
+      "loss": 1.1783,
+      "step": 1332
+    },
+    {
+      "epoch": 0.10275979031760715,
+      "grad_norm": 4.311089038848877,
+      "learning_rate": 9.862038976699589e-06,
+      "loss": 1.2346,
+      "step": 1333
+    },
+    {
+      "epoch": 0.10283687943262411,
+      "grad_norm": 4.145449638366699,
+      "learning_rate": 9.861747577945275e-06,
+      "loss": 1.241,
+      "step": 1334
+    },
+    {
+      "epoch": 0.10291396854764108,
+      "grad_norm": 4.2260236740112305,
+      "learning_rate": 9.861455876086181e-06,
+      "loss": 1.2695,
+      "step": 1335
+    },
+    {
+      "epoch": 0.10299105766265804,
+      "grad_norm": 4.495214462280273,
+      "learning_rate": 9.86116387114049e-06,
+      "loss": 1.2767,
+      "step": 1336
+    },
+    {
+      "epoch": 0.103068146777675,
+      "grad_norm": 4.408441066741943,
+      "learning_rate": 9.860871563126409e-06,
+      "loss": 1.1875,
+      "step": 1337
+    },
+    {
+      "epoch": 0.10314523589269195,
+      "grad_norm": 3.792508363723755,
+      "learning_rate": 9.860578952062161e-06,
+      "loss": 1.1621,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1032223250077089,
+      "grad_norm": 4.099066734313965,
+      "learning_rate": 9.86028603796599e-06,
+      "loss": 1.244,
+      "step": 1339
+    },
+    {
+      "epoch": 0.10329941412272588,
+      "grad_norm": 3.596609354019165,
+      "learning_rate": 9.859992820856155e-06,
+      "loss": 1.151,
+      "step": 1340
+    },
+    {
+      "epoch": 0.10337650323774283,
+      "grad_norm": 4.01649808883667,
+      "learning_rate": 9.859699300750937e-06,
+      "loss": 1.1552,
+      "step": 1341
+    },
+    {
+      "epoch": 0.10345359235275979,
+      "grad_norm": 3.891861915588379,
+      "learning_rate": 9.85940547766864e-06,
+      "loss": 1.1511,
+      "step": 1342
+    },
+    {
+      "epoch": 0.10353068146777675,
+      "grad_norm": 3.784027576446533,
+      "learning_rate": 9.859111351627576e-06,
+      "loss": 1.0498,
+      "step": 1343
+    },
+    {
+      "epoch": 0.1036077705827937,
+      "grad_norm": 3.873615026473999,
+      "learning_rate": 9.858816922646088e-06,
+      "loss": 1.1498,
+      "step": 1344
+    },
+    {
+      "epoch": 0.10368485969781067,
+      "grad_norm": 4.082927227020264,
+      "learning_rate": 9.858522190742529e-06,
+      "loss": 1.2154,
+      "step": 1345
+    },
+    {
+      "epoch": 0.10376194881282763,
+      "grad_norm": 3.9876515865325928,
+      "learning_rate": 9.858227155935271e-06,
+      "loss": 1.1871,
+      "step": 1346
+    },
+    {
+      "epoch": 0.10383903792784459,
+      "grad_norm": 3.943120241165161,
+      "learning_rate": 9.857931818242715e-06,
+      "loss": 1.0969,
+      "step": 1347
+    },
+    {
+      "epoch": 0.10391612704286154,
+      "grad_norm": 4.118403434753418,
+      "learning_rate": 9.857636177683267e-06,
+      "loss": 1.0838,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1039932161578785,
+      "grad_norm": 3.8626832962036133,
+      "learning_rate": 9.857340234275363e-06,
+      "loss": 1.1041,
+      "step": 1349
+    },
+    {
+      "epoch": 0.10407030527289547,
+      "grad_norm": 3.960684299468994,
+      "learning_rate": 9.857043988037453e-06,
+      "loss": 1.1214,
+      "step": 1350
+    },
+    {
+      "epoch": 0.10414739438791243,
+      "grad_norm": 3.8273346424102783,
+      "learning_rate": 9.856747438988005e-06,
+      "loss": 1.1456,
+      "step": 1351
+    },
+    {
+      "epoch": 0.10422448350292939,
+      "grad_norm": 4.401994705200195,
+      "learning_rate": 9.856450587145507e-06,
+      "loss": 1.1389,
+      "step": 1352
+    },
+    {
+      "epoch": 0.10430157261794634,
+      "grad_norm": 4.069916248321533,
+      "learning_rate": 9.856153432528467e-06,
+      "loss": 1.2821,
+      "step": 1353
+    },
+    {
+      "epoch": 0.1043786617329633,
+      "grad_norm": 3.9717392921447754,
+      "learning_rate": 9.855855975155414e-06,
+      "loss": 1.1286,
+      "step": 1354
+    },
+    {
+      "epoch": 0.10445575084798027,
+      "grad_norm": 3.949397087097168,
+      "learning_rate": 9.855558215044887e-06,
+      "loss": 1.1939,
+      "step": 1355
+    },
+    {
+      "epoch": 0.10453283996299723,
+      "grad_norm": 3.813549757003784,
+      "learning_rate": 9.855260152215455e-06,
+      "loss": 1.069,
+      "step": 1356
+    },
+    {
+      "epoch": 0.10460992907801418,
+      "grad_norm": 4.000561237335205,
+      "learning_rate": 9.854961786685697e-06,
+      "loss": 1.1664,
+      "step": 1357
+    },
+    {
+      "epoch": 0.10468701819303114,
+      "grad_norm": 4.400665283203125,
+      "learning_rate": 9.854663118474217e-06,
+      "loss": 1.1951,
+      "step": 1358
+    },
+    {
+      "epoch": 0.10476410730804811,
+      "grad_norm": 4.009283542633057,
+      "learning_rate": 9.854364147599635e-06,
+      "loss": 1.1911,
+      "step": 1359
+    },
+    {
+      "epoch": 0.10484119642306507,
+      "grad_norm": 4.075754642486572,
+      "learning_rate": 9.854064874080589e-06,
+      "loss": 1.2182,
+      "step": 1360
+    },
+    {
+      "epoch": 0.10491828553808202,
+      "grad_norm": 3.848576784133911,
+      "learning_rate": 9.853765297935738e-06,
+      "loss": 1.1876,
+      "step": 1361
+    },
+    {
+      "epoch": 0.10499537465309898,
+      "grad_norm": 4.087264060974121,
+      "learning_rate": 9.853465419183759e-06,
+      "loss": 1.2082,
+      "step": 1362
+    },
+    {
+      "epoch": 0.10507246376811594,
+      "grad_norm": 4.3004913330078125,
+      "learning_rate": 9.853165237843347e-06,
+      "loss": 1.1629,
+      "step": 1363
+    },
+    {
+      "epoch": 0.10514955288313291,
+      "grad_norm": 3.889831781387329,
+      "learning_rate": 9.852864753933218e-06,
+      "loss": 1.1833,
+      "step": 1364
+    },
+    {
+      "epoch": 0.10522664199814986,
+      "grad_norm": 3.902057409286499,
+      "learning_rate": 9.852563967472106e-06,
+      "loss": 1.2362,
+      "step": 1365
+    },
+    {
+      "epoch": 0.10530373111316682,
+      "grad_norm": 4.216920852661133,
+      "learning_rate": 9.852262878478762e-06,
+      "loss": 1.2058,
+      "step": 1366
+    },
+    {
+      "epoch": 0.10538082022818378,
+      "grad_norm": 3.6932597160339355,
+      "learning_rate": 9.851961486971959e-06,
+      "loss": 1.133,
+      "step": 1367
+    },
+    {
+      "epoch": 0.10545790934320073,
+      "grad_norm": 3.947735071182251,
+      "learning_rate": 9.851659792970485e-06,
+      "loss": 1.2308,
+      "step": 1368
+    },
+    {
+      "epoch": 0.1055349984582177,
+      "grad_norm": 3.9944536685943604,
+      "learning_rate": 9.85135779649315e-06,
+      "loss": 1.1327,
+      "step": 1369
+    },
+    {
+      "epoch": 0.10561208757323466,
+      "grad_norm": 3.8380393981933594,
+      "learning_rate": 9.851055497558783e-06,
+      "loss": 1.2182,
+      "step": 1370
+    },
+    {
+      "epoch": 0.10568917668825162,
+      "grad_norm": 3.927506446838379,
+      "learning_rate": 9.850752896186231e-06,
+      "loss": 1.1246,
+      "step": 1371
+    },
+    {
+      "epoch": 0.10576626580326857,
+      "grad_norm": 3.877490520477295,
+      "learning_rate": 9.850449992394357e-06,
+      "loss": 1.1259,
+      "step": 1372
+    },
+    {
+      "epoch": 0.10584335491828553,
+      "grad_norm": 3.9563748836517334,
+      "learning_rate": 9.850146786202048e-06,
+      "loss": 1.2217,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1059204440333025,
+      "grad_norm": 3.84112811088562,
+      "learning_rate": 9.849843277628206e-06,
+      "loss": 1.1055,
+      "step": 1374
+    },
+    {
+      "epoch": 0.10599753314831946,
+      "grad_norm": 4.415043830871582,
+      "learning_rate": 9.849539466691755e-06,
+      "loss": 1.1178,
+      "step": 1375
+    },
+    {
+      "epoch": 0.10607462226333642,
+      "grad_norm": 3.7989046573638916,
+      "learning_rate": 9.849235353411632e-06,
+      "loss": 1.2218,
+      "step": 1376
+    },
+    {
+      "epoch": 0.10615171137835337,
+      "grad_norm": 4.221613883972168,
+      "learning_rate": 9.848930937806802e-06,
+      "loss": 1.3153,
+      "step": 1377
+    },
+    {
+      "epoch": 0.10622880049337034,
+      "grad_norm": 3.621276378631592,
+      "learning_rate": 9.84862621989624e-06,
+      "loss": 1.18,
+      "step": 1378
+    },
+    {
+      "epoch": 0.1063058896083873,
+      "grad_norm": 4.181283473968506,
+      "learning_rate": 9.848321199698945e-06,
+      "loss": 1.1736,
+      "step": 1379
+    },
+    {
+      "epoch": 0.10638297872340426,
+      "grad_norm": 4.222801685333252,
+      "learning_rate": 9.848015877233935e-06,
+      "loss": 1.1519,
+      "step": 1380
+    },
+    {
+      "epoch": 0.10646006783842121,
+      "grad_norm": 4.010768890380859,
+      "learning_rate": 9.847710252520242e-06,
+      "loss": 1.1414,
+      "step": 1381
+    },
+    {
+      "epoch": 0.10653715695343817,
+      "grad_norm": 4.021345138549805,
+      "learning_rate": 9.847404325576921e-06,
+      "loss": 1.2202,
+      "step": 1382
+    },
+    {
+      "epoch": 0.10661424606845514,
+      "grad_norm": 4.335690498352051,
+      "learning_rate": 9.847098096423046e-06,
+      "loss": 1.1836,
+      "step": 1383
+    },
+    {
+      "epoch": 0.1066913351834721,
+      "grad_norm": 4.25598669052124,
+      "learning_rate": 9.84679156507771e-06,
+      "loss": 1.2209,
+      "step": 1384
+    },
+    {
+      "epoch": 0.10676842429848905,
+      "grad_norm": 3.9418535232543945,
+      "learning_rate": 9.84648473156002e-06,
+      "loss": 1.2035,
+      "step": 1385
+    },
+    {
+      "epoch": 0.10684551341350601,
+      "grad_norm": 4.38209867477417,
+      "learning_rate": 9.84617759588911e-06,
+      "loss": 1.2745,
+      "step": 1386
+    },
+    {
+      "epoch": 0.10692260252852297,
+      "grad_norm": 3.9577906131744385,
+      "learning_rate": 9.845870158084123e-06,
+      "loss": 1.2592,
+      "step": 1387
+    },
+    {
+      "epoch": 0.10699969164353994,
+      "grad_norm": 3.4693827629089355,
+      "learning_rate": 9.845562418164232e-06,
+      "loss": 1.1585,
+      "step": 1388
+    },
+    {
+      "epoch": 0.1070767807585569,
+      "grad_norm": 3.7138640880584717,
+      "learning_rate": 9.845254376148617e-06,
+      "loss": 1.1364,
+      "step": 1389
+    },
+    {
+      "epoch": 0.10715386987357385,
+      "grad_norm": 3.9818434715270996,
+      "learning_rate": 9.844946032056487e-06,
+      "loss": 1.0948,
+      "step": 1390
+    },
+    {
+      "epoch": 0.10723095898859081,
+      "grad_norm": 3.7591545581817627,
+      "learning_rate": 9.844637385907066e-06,
+      "loss": 1.0582,
+      "step": 1391
+    },
+    {
+      "epoch": 0.10730804810360776,
+      "grad_norm": 3.9569385051727295,
+      "learning_rate": 9.844328437719595e-06,
+      "loss": 1.1992,
+      "step": 1392
+    },
+    {
+      "epoch": 0.10738513721862473,
+      "grad_norm": 3.6534078121185303,
+      "learning_rate": 9.844019187513335e-06,
+      "loss": 1.1582,
+      "step": 1393
+    },
+    {
+      "epoch": 0.10746222633364169,
+      "grad_norm": 3.8868963718414307,
+      "learning_rate": 9.843709635307563e-06,
+      "loss": 1.2037,
+      "step": 1394
+    },
+    {
+      "epoch": 0.10753931544865865,
+      "grad_norm": 4.207004070281982,
+      "learning_rate": 9.843399781121585e-06,
+      "loss": 1.2641,
+      "step": 1395
+    },
+    {
+      "epoch": 0.1076164045636756,
+      "grad_norm": 3.9298391342163086,
+      "learning_rate": 9.843089624974716e-06,
+      "loss": 1.1404,
+      "step": 1396
+    },
+    {
+      "epoch": 0.10769349367869256,
+      "grad_norm": 3.9258861541748047,
+      "learning_rate": 9.84277916688629e-06,
+      "loss": 1.184,
+      "step": 1397
+    },
+    {
+      "epoch": 0.10777058279370953,
+      "grad_norm": 3.8614659309387207,
+      "learning_rate": 9.842468406875665e-06,
+      "loss": 1.2738,
+      "step": 1398
+    },
+    {
+      "epoch": 0.10784767190872649,
+      "grad_norm": 4.084595680236816,
+      "learning_rate": 9.842157344962214e-06,
+      "loss": 1.2203,
+      "step": 1399
+    },
+    {
+      "epoch": 0.10792476102374345,
+      "grad_norm": 3.8629918098449707,
+      "learning_rate": 9.84184598116533e-06,
+      "loss": 1.1434,
+      "step": 1400
+    },
+    {
+      "epoch": 0.1080018501387604,
+      "grad_norm": 3.8280506134033203,
+      "learning_rate": 9.841534315504427e-06,
+      "loss": 1.1438,
+      "step": 1401
+    },
+    {
+      "epoch": 0.10807893925377737,
+      "grad_norm": 4.113556861877441,
+      "learning_rate": 9.841222347998933e-06,
+      "loss": 1.2824,
+      "step": 1402
+    },
+    {
+      "epoch": 0.10815602836879433,
+      "grad_norm": 4.399749279022217,
+      "learning_rate": 9.8409100786683e-06,
+      "loss": 1.178,
+      "step": 1403
+    },
+    {
+      "epoch": 0.10823311748381129,
+      "grad_norm": 4.358874320983887,
+      "learning_rate": 9.840597507531997e-06,
+      "loss": 1.2722,
+      "step": 1404
+    },
+    {
+      "epoch": 0.10831020659882824,
+      "grad_norm": 3.7543954849243164,
+      "learning_rate": 9.840284634609508e-06,
+      "loss": 1.1464,
+      "step": 1405
+    },
+    {
+      "epoch": 0.1083872957138452,
+      "grad_norm": 4.325103282928467,
+      "learning_rate": 9.839971459920338e-06,
+      "loss": 1.1934,
+      "step": 1406
+    },
+    {
+      "epoch": 0.10846438482886217,
+      "grad_norm": 4.050751209259033,
+      "learning_rate": 9.839657983484018e-06,
+      "loss": 1.3296,
+      "step": 1407
+    },
+    {
+      "epoch": 0.10854147394387913,
+      "grad_norm": 4.21981954574585,
+      "learning_rate": 9.839344205320088e-06,
+      "loss": 1.2409,
+      "step": 1408
+    },
+    {
+      "epoch": 0.10861856305889608,
+      "grad_norm": 4.275995254516602,
+      "learning_rate": 9.839030125448108e-06,
+      "loss": 1.3003,
+      "step": 1409
+    },
+    {
+      "epoch": 0.10869565217391304,
+      "grad_norm": 4.038300514221191,
+      "learning_rate": 9.838715743887662e-06,
+      "loss": 1.1667,
+      "step": 1410
+    },
+    {
+      "epoch": 0.10877274128893,
+      "grad_norm": 3.92463755607605,
+      "learning_rate": 9.838401060658352e-06,
+      "loss": 1.1469,
+      "step": 1411
+    },
+    {
+      "epoch": 0.10884983040394697,
+      "grad_norm": 4.041322231292725,
+      "learning_rate": 9.83808607577979e-06,
+      "loss": 1.2825,
+      "step": 1412
+    },
+    {
+      "epoch": 0.10892691951896392,
+      "grad_norm": 4.066822528839111,
+      "learning_rate": 9.837770789271623e-06,
+      "loss": 1.1624,
+      "step": 1413
+    },
+    {
+      "epoch": 0.10900400863398088,
+      "grad_norm": 3.8004565238952637,
+      "learning_rate": 9.8374552011535e-06,
+      "loss": 1.1102,
+      "step": 1414
+    },
+    {
+      "epoch": 0.10908109774899784,
+      "grad_norm": 3.726020336151123,
+      "learning_rate": 9.837139311445102e-06,
+      "loss": 1.1266,
+      "step": 1415
+    },
+    {
+      "epoch": 0.1091581868640148,
+      "grad_norm": 4.26671028137207,
+      "learning_rate": 9.836823120166116e-06,
+      "loss": 1.2413,
+      "step": 1416
+    },
+    {
+      "epoch": 0.10923527597903177,
+      "grad_norm": 3.6223459243774414,
+      "learning_rate": 9.836506627336261e-06,
+      "loss": 1.1612,
+      "step": 1417
+    },
+    {
+      "epoch": 0.10931236509404872,
+      "grad_norm": 3.5984158515930176,
+      "learning_rate": 9.836189832975267e-06,
+      "loss": 1.0419,
+      "step": 1418
+    },
+    {
+      "epoch": 0.10938945420906568,
+      "grad_norm": 3.670827627182007,
+      "learning_rate": 9.835872737102886e-06,
+      "loss": 1.2345,
+      "step": 1419
+    },
+    {
+      "epoch": 0.10946654332408264,
+      "grad_norm": 3.62565541267395,
+      "learning_rate": 9.835555339738882e-06,
+      "loss": 1.0729,
+      "step": 1420
+    },
+    {
+      "epoch": 0.1095436324390996,
+      "grad_norm": 4.1612958908081055,
+      "learning_rate": 9.83523764090305e-06,
+      "loss": 1.2735,
+      "step": 1421
+    },
+    {
+      "epoch": 0.10962072155411656,
+      "grad_norm": 3.878818988800049,
+      "learning_rate": 9.83491964061519e-06,
+      "loss": 1.2431,
+      "step": 1422
+    },
+    {
+      "epoch": 0.10969781066913352,
+      "grad_norm": 3.963610887527466,
+      "learning_rate": 9.834601338895133e-06,
+      "loss": 1.2172,
+      "step": 1423
+    },
+    {
+      "epoch": 0.10977489978415048,
+      "grad_norm": 4.701980113983154,
+      "learning_rate": 9.83428273576272e-06,
+      "loss": 1.2959,
+      "step": 1424
+    },
+    {
+      "epoch": 0.10985198889916743,
+      "grad_norm": 4.165730953216553,
+      "learning_rate": 9.833963831237819e-06,
+      "loss": 1.1758,
+      "step": 1425
+    },
+    {
+      "epoch": 0.1099290780141844,
+      "grad_norm": 4.231865882873535,
+      "learning_rate": 9.833644625340305e-06,
+      "loss": 1.1345,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11000616712920136,
+      "grad_norm": 3.8427863121032715,
+      "learning_rate": 9.833325118090086e-06,
+      "loss": 1.1748,
+      "step": 1427
+    },
+    {
+      "epoch": 0.11008325624421832,
+      "grad_norm": 4.339493274688721,
+      "learning_rate": 9.833005309507076e-06,
+      "loss": 1.2364,
+      "step": 1428
+    },
+    {
+      "epoch": 0.11016034535923527,
+      "grad_norm": 3.940251350402832,
+      "learning_rate": 9.832685199611217e-06,
+      "loss": 1.1587,
+      "step": 1429
+    },
+    {
+      "epoch": 0.11023743447425223,
+      "grad_norm": 4.076071739196777,
+      "learning_rate": 9.832364788422464e-06,
+      "loss": 1.1959,
+      "step": 1430
+    },
+    {
+      "epoch": 0.1103145235892692,
+      "grad_norm": 3.910961389541626,
+      "learning_rate": 9.832044075960795e-06,
+      "loss": 1.0885,
+      "step": 1431
+    },
+    {
+      "epoch": 0.11039161270428616,
+      "grad_norm": 4.00604772567749,
+      "learning_rate": 9.831723062246204e-06,
+      "loss": 1.2461,
+      "step": 1432
+    },
+    {
+      "epoch": 0.11046870181930311,
+      "grad_norm": 4.331727981567383,
+      "learning_rate": 9.831401747298702e-06,
+      "loss": 1.2664,
+      "step": 1433
+    },
+    {
+      "epoch": 0.11054579093432007,
+      "grad_norm": 4.105038642883301,
+      "learning_rate": 9.831080131138325e-06,
+      "loss": 1.1604,
+      "step": 1434
+    },
+    {
+      "epoch": 0.11062288004933703,
+      "grad_norm": 4.079939365386963,
+      "learning_rate": 9.830758213785123e-06,
+      "loss": 1.1103,
+      "step": 1435
+    },
+    {
+      "epoch": 0.110699969164354,
+      "grad_norm": 3.9948172569274902,
+      "learning_rate": 9.830435995259165e-06,
+      "loss": 1.1889,
+      "step": 1436
+    },
+    {
+      "epoch": 0.11077705827937095,
+      "grad_norm": 4.091009616851807,
+      "learning_rate": 9.830113475580541e-06,
+      "loss": 1.1374,
+      "step": 1437
+    },
+    {
+      "epoch": 0.11085414739438791,
+      "grad_norm": 4.397194862365723,
+      "learning_rate": 9.829790654769356e-06,
+      "loss": 1.1226,
+      "step": 1438
+    },
+    {
+      "epoch": 0.11093123650940487,
+      "grad_norm": 4.208889961242676,
+      "learning_rate": 9.829467532845738e-06,
+      "loss": 1.1658,
+      "step": 1439
+    },
+    {
+      "epoch": 0.11100832562442182,
+      "grad_norm": 4.262278079986572,
+      "learning_rate": 9.829144109829832e-06,
+      "loss": 1.3317,
+      "step": 1440
+    },
+    {
+      "epoch": 0.1110854147394388,
+      "grad_norm": 4.373659133911133,
+      "learning_rate": 9.828820385741802e-06,
+      "loss": 1.1994,
+      "step": 1441
+    },
+    {
+      "epoch": 0.11116250385445575,
+      "grad_norm": 3.869906425476074,
+      "learning_rate": 9.82849636060183e-06,
+      "loss": 1.2039,
+      "step": 1442
+    },
+    {
+      "epoch": 0.11123959296947271,
+      "grad_norm": 3.81036114692688,
+      "learning_rate": 9.828172034430118e-06,
+      "loss": 1.0779,
+      "step": 1443
+    },
+    {
+      "epoch": 0.11131668208448967,
+      "grad_norm": 4.158944129943848,
+      "learning_rate": 9.827847407246885e-06,
+      "loss": 1.1907,
+      "step": 1444
+    },
+    {
+      "epoch": 0.11139377119950664,
+      "grad_norm": 4.358891487121582,
+      "learning_rate": 9.827522479072369e-06,
+      "loss": 1.1776,
+      "step": 1445
+    },
+    {
+      "epoch": 0.11147086031452359,
+      "grad_norm": 4.275304794311523,
+      "learning_rate": 9.82719724992683e-06,
+      "loss": 1.2849,
+      "step": 1446
+    },
+    {
+      "epoch": 0.11154794942954055,
+      "grad_norm": 4.336565971374512,
+      "learning_rate": 9.826871719830542e-06,
+      "loss": 1.1965,
+      "step": 1447
+    },
+    {
+      "epoch": 0.1116250385445575,
+      "grad_norm": 4.239962577819824,
+      "learning_rate": 9.826545888803802e-06,
+      "loss": 1.229,
+      "step": 1448
+    },
+    {
+      "epoch": 0.11170212765957446,
+      "grad_norm": 4.397287368774414,
+      "learning_rate": 9.826219756866923e-06,
+      "loss": 1.2252,
+      "step": 1449
+    },
+    {
+      "epoch": 0.11177921677459143,
+      "grad_norm": 4.538805961608887,
+      "learning_rate": 9.82589332404024e-06,
+      "loss": 1.2506,
+      "step": 1450
+    },
+    {
+      "epoch": 0.11185630588960839,
+      "grad_norm": 3.814425468444824,
+      "learning_rate": 9.825566590344098e-06,
+      "loss": 1.1343,
+      "step": 1451
+    },
+    {
+      "epoch": 0.11193339500462535,
+      "grad_norm": 3.6596169471740723,
+      "learning_rate": 9.825239555798875e-06,
+      "loss": 1.2412,
+      "step": 1452
+    },
+    {
+      "epoch": 0.1120104841196423,
+      "grad_norm": 4.104322910308838,
+      "learning_rate": 9.824912220424953e-06,
+      "loss": 1.1035,
+      "step": 1453
+    },
+    {
+      "epoch": 0.11208757323465926,
+      "grad_norm": 3.8343546390533447,
+      "learning_rate": 9.824584584242746e-06,
+      "loss": 1.1575,
+      "step": 1454
+    },
+    {
+      "epoch": 0.11216466234967623,
+      "grad_norm": 3.6301510334014893,
+      "learning_rate": 9.824256647272676e-06,
+      "loss": 1.2006,
+      "step": 1455
+    },
+    {
+      "epoch": 0.11224175146469319,
+      "grad_norm": 4.543994903564453,
+      "learning_rate": 9.823928409535191e-06,
+      "loss": 1.2539,
+      "step": 1456
+    },
+    {
+      "epoch": 0.11231884057971014,
+      "grad_norm": 4.104208469390869,
+      "learning_rate": 9.82359987105075e-06,
+      "loss": 1.2179,
+      "step": 1457
+    },
+    {
+      "epoch": 0.1123959296947271,
+      "grad_norm": 4.09733772277832,
+      "learning_rate": 9.823271031839843e-06,
+      "loss": 1.1579,
+      "step": 1458
+    },
+    {
+      "epoch": 0.11247301880974406,
+      "grad_norm": 4.35043478012085,
+      "learning_rate": 9.822941891922965e-06,
+      "loss": 1.2598,
+      "step": 1459
+    },
+    {
+      "epoch": 0.11255010792476103,
+      "grad_norm": 4.023888111114502,
+      "learning_rate": 9.82261245132064e-06,
+      "loss": 1.1077,
+      "step": 1460
+    },
+    {
+      "epoch": 0.11262719703977798,
+      "grad_norm": 3.594707489013672,
+      "learning_rate": 9.822282710053403e-06,
+      "loss": 1.1578,
+      "step": 1461
+    },
+    {
+      "epoch": 0.11270428615479494,
+      "grad_norm": 4.166836261749268,
+      "learning_rate": 9.821952668141817e-06,
+      "loss": 1.1718,
+      "step": 1462
+    },
+    {
+      "epoch": 0.1127813752698119,
+      "grad_norm": 3.9749162197113037,
+      "learning_rate": 9.821622325606454e-06,
+      "loss": 1.1749,
+      "step": 1463
+    },
+    {
+      "epoch": 0.11285846438482887,
+      "grad_norm": 4.0859150886535645,
+      "learning_rate": 9.821291682467912e-06,
+      "loss": 1.1667,
+      "step": 1464
+    },
+    {
+      "epoch": 0.11293555349984583,
+      "grad_norm": 5.696316242218018,
+      "learning_rate": 9.820960738746805e-06,
+      "loss": 1.0963,
+      "step": 1465
+    },
+    {
+      "epoch": 0.11301264261486278,
+      "grad_norm": 4.3305559158325195,
+      "learning_rate": 9.820629494463763e-06,
+      "loss": 1.1439,
+      "step": 1466
+    },
+    {
+      "epoch": 0.11308973172987974,
+      "grad_norm": 4.143655300140381,
+      "learning_rate": 9.820297949639439e-06,
+      "loss": 1.0771,
+      "step": 1467
+    },
+    {
+      "epoch": 0.1131668208448967,
+      "grad_norm": 4.505462169647217,
+      "learning_rate": 9.8199661042945e-06,
+      "loss": 1.2824,
+      "step": 1468
+    },
+    {
+      "epoch": 0.11324390995991367,
+      "grad_norm": 3.8366472721099854,
+      "learning_rate": 9.819633958449642e-06,
+      "loss": 1.1783,
+      "step": 1469
+    },
+    {
+      "epoch": 0.11332099907493062,
+      "grad_norm": 4.8427910804748535,
+      "learning_rate": 9.819301512125565e-06,
+      "loss": 1.3027,
+      "step": 1470
+    },
+    {
+      "epoch": 0.11339808818994758,
+      "grad_norm": 4.196298599243164,
+      "learning_rate": 9.818968765343e-06,
+      "loss": 1.1545,
+      "step": 1471
+    },
+    {
+      "epoch": 0.11347517730496454,
+      "grad_norm": 4.095616817474365,
+      "learning_rate": 9.818635718122692e-06,
+      "loss": 1.1564,
+      "step": 1472
+    },
+    {
+      "epoch": 0.11355226641998149,
+      "grad_norm": 4.081837177276611,
+      "learning_rate": 9.8183023704854e-06,
+      "loss": 1.141,
+      "step": 1473
+    },
+    {
+      "epoch": 0.11362935553499846,
+      "grad_norm": 4.244540214538574,
+      "learning_rate": 9.817968722451911e-06,
+      "loss": 1.2587,
+      "step": 1474
+    },
+    {
+      "epoch": 0.11370644465001542,
+      "grad_norm": 4.162115097045898,
+      "learning_rate": 9.817634774043026e-06,
+      "loss": 1.2686,
+      "step": 1475
+    },
+    {
+      "epoch": 0.11378353376503238,
+      "grad_norm": 4.111065864562988,
+      "learning_rate": 9.817300525279562e-06,
+      "loss": 1.1118,
+      "step": 1476
+    },
+    {
+      "epoch": 0.11386062288004933,
+      "grad_norm": 4.0052714347839355,
+      "learning_rate": 9.816965976182362e-06,
+      "loss": 1.145,
+      "step": 1477
+    },
+    {
+      "epoch": 0.11393771199506629,
+      "grad_norm": 3.8451459407806396,
+      "learning_rate": 9.81663112677228e-06,
+      "loss": 1.0982,
+      "step": 1478
+    },
+    {
+      "epoch": 0.11401480111008326,
+      "grad_norm": 4.380990982055664,
+      "learning_rate": 9.816295977070193e-06,
+      "loss": 1.2695,
+      "step": 1479
+    },
+    {
+      "epoch": 0.11409189022510022,
+      "grad_norm": 4.092302322387695,
+      "learning_rate": 9.815960527096996e-06,
+      "loss": 1.2088,
+      "step": 1480
+    },
+    {
+      "epoch": 0.11416897934011717,
+      "grad_norm": 3.8186841011047363,
+      "learning_rate": 9.815624776873605e-06,
+      "loss": 1.1231,
+      "step": 1481
+    },
+    {
+      "epoch": 0.11424606845513413,
+      "grad_norm": 4.1122894287109375,
+      "learning_rate": 9.815288726420949e-06,
+      "loss": 1.1516,
+      "step": 1482
+    },
+    {
+      "epoch": 0.1143231575701511,
+      "grad_norm": 4.764173984527588,
+      "learning_rate": 9.814952375759979e-06,
+      "loss": 1.2084,
+      "step": 1483
+    },
+    {
+      "epoch": 0.11440024668516806,
+      "grad_norm": 4.217214584350586,
+      "learning_rate": 9.814615724911664e-06,
+      "loss": 1.0574,
+      "step": 1484
+    },
+    {
+      "epoch": 0.11447733580018501,
+      "grad_norm": 3.7358455657958984,
+      "learning_rate": 9.814278773896997e-06,
+      "loss": 1.1282,
+      "step": 1485
+    },
+    {
+      "epoch": 0.11455442491520197,
+      "grad_norm": 3.7960875034332275,
+      "learning_rate": 9.813941522736981e-06,
+      "loss": 1.1453,
+      "step": 1486
+    },
+    {
+      "epoch": 0.11463151403021893,
+      "grad_norm": 4.146248817443848,
+      "learning_rate": 9.813603971452643e-06,
+      "loss": 1.2265,
+      "step": 1487
+    },
+    {
+      "epoch": 0.1147086031452359,
+      "grad_norm": 4.388052463531494,
+      "learning_rate": 9.813266120065028e-06,
+      "loss": 1.1369,
+      "step": 1488
+    },
+    {
+      "epoch": 0.11478569226025286,
+      "grad_norm": 4.071401119232178,
+      "learning_rate": 9.812927968595199e-06,
+      "loss": 1.1689,
+      "step": 1489
+    },
+    {
+      "epoch": 0.11486278137526981,
+      "grad_norm": 4.517001628875732,
+      "learning_rate": 9.812589517064237e-06,
+      "loss": 1.1537,
+      "step": 1490
+    },
+    {
+      "epoch": 0.11493987049028677,
+      "grad_norm": 4.021905899047852,
+      "learning_rate": 9.812250765493243e-06,
+      "loss": 1.0266,
+      "step": 1491
+    },
+    {
+      "epoch": 0.11501695960530373,
+      "grad_norm": 4.433926105499268,
+      "learning_rate": 9.811911713903339e-06,
+      "loss": 1.108,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1150940487203207,
+      "grad_norm": 4.291812419891357,
+      "learning_rate": 9.811572362315661e-06,
+      "loss": 1.2626,
+      "step": 1493
+    },
+    {
+      "epoch": 0.11517113783533765,
+      "grad_norm": 4.35188627243042,
+      "learning_rate": 9.811232710751366e-06,
+      "loss": 1.1318,
+      "step": 1494
+    },
+    {
+      "epoch": 0.11524822695035461,
+      "grad_norm": 4.012880325317383,
+      "learning_rate": 9.810892759231629e-06,
+      "loss": 1.1044,
+      "step": 1495
+    },
+    {
+      "epoch": 0.11532531606537157,
+      "grad_norm": 4.010427474975586,
+      "learning_rate": 9.810552507777643e-06,
+      "loss": 1.2086,
+      "step": 1496
+    },
+    {
+      "epoch": 0.11540240518038852,
+      "grad_norm": 3.74871563911438,
+      "learning_rate": 9.810211956410625e-06,
+      "loss": 1.1279,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1154794942954055,
+      "grad_norm": 3.983808755874634,
+      "learning_rate": 9.809871105151805e-06,
+      "loss": 1.1586,
+      "step": 1498
+    },
+    {
+      "epoch": 0.11555658341042245,
+      "grad_norm": 4.453268051147461,
+      "learning_rate": 9.80952995402243e-06,
+      "loss": 1.156,
+      "step": 1499
+    },
+    {
+      "epoch": 0.11563367252543941,
+      "grad_norm": 4.205508708953857,
+      "learning_rate": 9.809188503043773e-06,
+      "loss": 1.1945,
+      "step": 1500
+    },
+    {
+      "epoch": 0.11571076164045636,
+      "grad_norm": 4.098243713378906,
+      "learning_rate": 9.80884675223712e-06,
+      "loss": 1.1781,
+      "step": 1501
+    },
+    {
+      "epoch": 0.11578785075547332,
+      "grad_norm": 4.6129937171936035,
+      "learning_rate": 9.808504701623777e-06,
+      "loss": 1.1673,
+      "step": 1502
+    },
+    {
+      "epoch": 0.11586493987049029,
+      "grad_norm": 4.1249494552612305,
+      "learning_rate": 9.808162351225073e-06,
+      "loss": 1.1907,
+      "step": 1503
+    },
+    {
+      "epoch": 0.11594202898550725,
+      "grad_norm": 3.684288740158081,
+      "learning_rate": 9.807819701062345e-06,
+      "loss": 1.1824,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1160191181005242,
+      "grad_norm": 4.17221212387085,
+      "learning_rate": 9.80747675115696e-06,
+      "loss": 1.1458,
+      "step": 1505
+    },
+    {
+      "epoch": 0.11609620721554116,
+      "grad_norm": 3.946536064147949,
+      "learning_rate": 9.807133501530297e-06,
+      "loss": 1.1267,
+      "step": 1506
+    },
+    {
+      "epoch": 0.11617329633055813,
+      "grad_norm": 4.113000392913818,
+      "learning_rate": 9.806789952203759e-06,
+      "loss": 1.1736,
+      "step": 1507
+    },
+    {
+      "epoch": 0.11625038544557509,
+      "grad_norm": 4.112078666687012,
+      "learning_rate": 9.806446103198761e-06,
+      "loss": 1.2042,
+      "step": 1508
+    },
+    {
+      "epoch": 0.11632747456059205,
+      "grad_norm": 3.994236946105957,
+      "learning_rate": 9.806101954536741e-06,
+      "loss": 1.2522,
+      "step": 1509
+    },
+    {
+      "epoch": 0.116404563675609,
+      "grad_norm": 4.026162624359131,
+      "learning_rate": 9.805757506239157e-06,
+      "loss": 1.1353,
+      "step": 1510
+    },
+    {
+      "epoch": 0.11648165279062596,
+      "grad_norm": 4.128935813903809,
+      "learning_rate": 9.805412758327483e-06,
+      "loss": 1.2068,
+      "step": 1511
+    },
+    {
+      "epoch": 0.11655874190564293,
+      "grad_norm": 3.9604172706604004,
+      "learning_rate": 9.80506771082321e-06,
+      "loss": 1.1924,
+      "step": 1512
+    },
+    {
+      "epoch": 0.11663583102065989,
+      "grad_norm": 6.198178291320801,
+      "learning_rate": 9.804722363747852e-06,
+      "loss": 1.165,
+      "step": 1513
+    },
+    {
+      "epoch": 0.11671292013567684,
+      "grad_norm": 4.178301811218262,
+      "learning_rate": 9.804376717122938e-06,
+      "loss": 1.1974,
+      "step": 1514
+    },
+    {
+      "epoch": 0.1167900092506938,
+      "grad_norm": 4.095193386077881,
+      "learning_rate": 9.804030770970019e-06,
+      "loss": 1.1372,
+      "step": 1515
+    },
+    {
+      "epoch": 0.11686709836571076,
+      "grad_norm": 3.5960633754730225,
+      "learning_rate": 9.803684525310662e-06,
+      "loss": 1.1578,
+      "step": 1516
+    },
+    {
+      "epoch": 0.11694418748072773,
+      "grad_norm": 3.994582414627075,
+      "learning_rate": 9.803337980166455e-06,
+      "loss": 1.1687,
+      "step": 1517
+    },
+    {
+      "epoch": 0.11702127659574468,
+      "grad_norm": 4.156383991241455,
+      "learning_rate": 9.802991135558998e-06,
+      "loss": 1.1485,
+      "step": 1518
+    },
+    {
+      "epoch": 0.11709836571076164,
+      "grad_norm": 3.9778475761413574,
+      "learning_rate": 9.802643991509923e-06,
+      "loss": 1.2049,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1171754548257786,
+      "grad_norm": 4.528823375701904,
+      "learning_rate": 9.802296548040868e-06,
+      "loss": 1.3013,
+      "step": 1520
+    },
+    {
+      "epoch": 0.11725254394079555,
+      "grad_norm": 4.259623050689697,
+      "learning_rate": 9.801948805173494e-06,
+      "loss": 1.1648,
+      "step": 1521
+    },
+    {
+      "epoch": 0.11732963305581252,
+      "grad_norm": 4.2941460609436035,
+      "learning_rate": 9.801600762929482e-06,
+      "loss": 1.218,
+      "step": 1522
+    },
+    {
+      "epoch": 0.11740672217082948,
+      "grad_norm": 4.13802433013916,
+      "learning_rate": 9.801252421330531e-06,
+      "loss": 1.2677,
+      "step": 1523
+    },
+    {
+      "epoch": 0.11748381128584644,
+      "grad_norm": 3.9115195274353027,
+      "learning_rate": 9.800903780398357e-06,
+      "loss": 1.0531,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1175609004008634,
+      "grad_norm": 4.135284423828125,
+      "learning_rate": 9.800554840154697e-06,
+      "loss": 1.1867,
+      "step": 1525
+    },
+    {
+      "epoch": 0.11763798951588036,
+      "grad_norm": 3.7419018745422363,
+      "learning_rate": 9.800205600621307e-06,
+      "loss": 1.1606,
+      "step": 1526
+    },
+    {
+      "epoch": 0.11771507863089732,
+      "grad_norm": 4.145817279815674,
+      "learning_rate": 9.799856061819958e-06,
+      "loss": 1.1399,
+      "step": 1527
+    },
+    {
+      "epoch": 0.11779216774591428,
+      "grad_norm": 3.8120839595794678,
+      "learning_rate": 9.799506223772442e-06,
+      "loss": 1.1803,
+      "step": 1528
+    },
+    {
+      "epoch": 0.11786925686093123,
+      "grad_norm": 4.319042682647705,
+      "learning_rate": 9.79915608650057e-06,
+      "loss": 1.0973,
+      "step": 1529
+    },
+    {
+      "epoch": 0.11794634597594819,
+      "grad_norm": 4.340944766998291,
+      "learning_rate": 9.798805650026173e-06,
+      "loss": 1.3635,
+      "step": 1530
+    },
+    {
+      "epoch": 0.11802343509096516,
+      "grad_norm": 4.413815975189209,
+      "learning_rate": 9.798454914371096e-06,
+      "loss": 1.1947,
+      "step": 1531
+    },
+    {
+      "epoch": 0.11810052420598212,
+      "grad_norm": 3.837949514389038,
+      "learning_rate": 9.798103879557207e-06,
+      "loss": 1.216,
+      "step": 1532
+    },
+    {
+      "epoch": 0.11817761332099908,
+      "grad_norm": 4.424009323120117,
+      "learning_rate": 9.797752545606391e-06,
+      "loss": 1.2992,
+      "step": 1533
+    },
+    {
+      "epoch": 0.11825470243601603,
+      "grad_norm": 3.9235024452209473,
+      "learning_rate": 9.797400912540553e-06,
+      "loss": 1.1753,
+      "step": 1534
+    },
+    {
+      "epoch": 0.11833179155103299,
+      "grad_norm": 3.9856998920440674,
+      "learning_rate": 9.797048980381614e-06,
+      "loss": 1.2046,
+      "step": 1535
+    },
+    {
+      "epoch": 0.11840888066604996,
+      "grad_norm": 3.952277898788452,
+      "learning_rate": 9.796696749151515e-06,
+      "loss": 1.1669,
+      "step": 1536
+    },
+    {
+      "epoch": 0.11848596978106692,
+      "grad_norm": 3.826277732849121,
+      "learning_rate": 9.796344218872218e-06,
+      "loss": 1.2616,
+      "step": 1537
+    },
+    {
+      "epoch": 0.11856305889608387,
+      "grad_norm": 4.182483196258545,
+      "learning_rate": 9.795991389565698e-06,
+      "loss": 1.0979,
+      "step": 1538
+    },
+    {
+      "epoch": 0.11864014801110083,
+      "grad_norm": 4.177645206451416,
+      "learning_rate": 9.795638261253955e-06,
+      "loss": 1.2607,
+      "step": 1539
+    },
+    {
+      "epoch": 0.11871723712611779,
+      "grad_norm": 3.5428824424743652,
+      "learning_rate": 9.795284833959003e-06,
+      "loss": 1.011,
+      "step": 1540
+    },
+    {
+      "epoch": 0.11879432624113476,
+      "grad_norm": 4.085594654083252,
+      "learning_rate": 9.794931107702877e-06,
+      "loss": 1.1675,
+      "step": 1541
+    },
+    {
+      "epoch": 0.11887141535615171,
+      "grad_norm": 4.093392848968506,
+      "learning_rate": 9.794577082507631e-06,
+      "loss": 1.2673,
+      "step": 1542
+    },
+    {
+      "epoch": 0.11894850447116867,
+      "grad_norm": 4.299323081970215,
+      "learning_rate": 9.794222758395336e-06,
+      "loss": 1.2598,
+      "step": 1543
+    },
+    {
+      "epoch": 0.11902559358618563,
+      "grad_norm": 3.8711843490600586,
+      "learning_rate": 9.79386813538808e-06,
+      "loss": 1.1941,
+      "step": 1544
+    },
+    {
+      "epoch": 0.11910268270120258,
+      "grad_norm": 4.2434983253479,
+      "learning_rate": 9.793513213507974e-06,
+      "loss": 1.1717,
+      "step": 1545
+    },
+    {
+      "epoch": 0.11917977181621955,
+      "grad_norm": 4.333231449127197,
+      "learning_rate": 9.793157992777147e-06,
+      "loss": 1.2443,
+      "step": 1546
+    },
+    {
+      "epoch": 0.11925686093123651,
+      "grad_norm": 3.959233045578003,
+      "learning_rate": 9.792802473217742e-06,
+      "loss": 1.191,
+      "step": 1547
+    },
+    {
+      "epoch": 0.11933395004625347,
+      "grad_norm": 3.931117534637451,
+      "learning_rate": 9.792446654851928e-06,
+      "loss": 1.1582,
+      "step": 1548
+    },
+    {
+      "epoch": 0.11941103916127042,
+      "grad_norm": 4.2051777839660645,
+      "learning_rate": 9.792090537701883e-06,
+      "loss": 1.1121,
+      "step": 1549
+    },
+    {
+      "epoch": 0.1194881282762874,
+      "grad_norm": 4.133641242980957,
+      "learning_rate": 9.791734121789814e-06,
+      "loss": 1.214,
+      "step": 1550
+    },
+    {
+      "epoch": 0.11956521739130435,
+      "grad_norm": 3.797180652618408,
+      "learning_rate": 9.791377407137936e-06,
+      "loss": 1.0736,
+      "step": 1551
+    },
+    {
+      "epoch": 0.11964230650632131,
+      "grad_norm": 4.1591477394104,
+      "learning_rate": 9.791020393768495e-06,
+      "loss": 1.2181,
+      "step": 1552
+    },
+    {
+      "epoch": 0.11971939562133826,
+      "grad_norm": 4.558741569519043,
+      "learning_rate": 9.790663081703746e-06,
+      "loss": 1.1797,
+      "step": 1553
+    },
+    {
+      "epoch": 0.11979648473635522,
+      "grad_norm": 4.007091522216797,
+      "learning_rate": 9.790305470965964e-06,
+      "loss": 1.2147,
+      "step": 1554
+    },
+    {
+      "epoch": 0.11987357385137219,
+      "grad_norm": 4.392326831817627,
+      "learning_rate": 9.789947561577445e-06,
+      "loss": 1.2772,
+      "step": 1555
+    },
+    {
+      "epoch": 0.11995066296638915,
+      "grad_norm": 3.7429401874542236,
+      "learning_rate": 9.789589353560505e-06,
+      "loss": 1.2085,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1200277520814061,
+      "grad_norm": 4.087551116943359,
+      "learning_rate": 9.789230846937473e-06,
+      "loss": 1.1084,
+      "step": 1557
+    },
+    {
+      "epoch": 0.12010484119642306,
+      "grad_norm": 4.172464370727539,
+      "learning_rate": 9.788872041730703e-06,
+      "loss": 1.2449,
+      "step": 1558
+    },
+    {
+      "epoch": 0.12018193031144002,
+      "grad_norm": 4.280933856964111,
+      "learning_rate": 9.788512937962562e-06,
+      "loss": 1.1364,
+      "step": 1559
+    },
+    {
+      "epoch": 0.12025901942645699,
+      "grad_norm": 3.9372894763946533,
+      "learning_rate": 9.788153535655442e-06,
+      "loss": 1.1644,
+      "step": 1560
+    },
+    {
+      "epoch": 0.12033610854147395,
+      "grad_norm": 3.8368165493011475,
+      "learning_rate": 9.787793834831745e-06,
+      "loss": 1.138,
+      "step": 1561
+    },
+    {
+      "epoch": 0.1204131976564909,
+      "grad_norm": 3.9816088676452637,
+      "learning_rate": 9.787433835513901e-06,
+      "loss": 1.0688,
+      "step": 1562
+    },
+    {
+      "epoch": 0.12049028677150786,
+      "grad_norm": 3.9207653999328613,
+      "learning_rate": 9.787073537724351e-06,
+      "loss": 1.1746,
+      "step": 1563
+    },
+    {
+      "epoch": 0.12056737588652482,
+      "grad_norm": 4.689396858215332,
+      "learning_rate": 9.78671294148556e-06,
+      "loss": 1.2326,
+      "step": 1564
+    },
+    {
+      "epoch": 0.12064446500154179,
+      "grad_norm": 4.082599639892578,
+      "learning_rate": 9.786352046820007e-06,
+      "loss": 1.1073,
+      "step": 1565
+    },
+    {
+      "epoch": 0.12072155411655874,
+      "grad_norm": 4.284193515777588,
+      "learning_rate": 9.785990853750193e-06,
+      "loss": 1.2251,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1207986432315757,
+      "grad_norm": 3.9901416301727295,
+      "learning_rate": 9.785629362298637e-06,
+      "loss": 1.2174,
+      "step": 1567
+    },
+    {
+      "epoch": 0.12087573234659266,
+      "grad_norm": 4.175442695617676,
+      "learning_rate": 9.785267572487876e-06,
+      "loss": 1.1415,
+      "step": 1568
+    },
+    {
+      "epoch": 0.12095282146160963,
+      "grad_norm": 4.132479190826416,
+      "learning_rate": 9.784905484340463e-06,
+      "loss": 1.2834,
+      "step": 1569
+    },
+    {
+      "epoch": 0.12102991057662658,
+      "grad_norm": 3.794235944747925,
+      "learning_rate": 9.784543097878977e-06,
+      "loss": 1.0903,
+      "step": 1570
+    },
+    {
+      "epoch": 0.12110699969164354,
+      "grad_norm": 3.9358041286468506,
+      "learning_rate": 9.784180413126009e-06,
+      "loss": 1.0461,
+      "step": 1571
+    },
+    {
+      "epoch": 0.1211840888066605,
+      "grad_norm": 3.783879041671753,
+      "learning_rate": 9.78381743010417e-06,
+      "loss": 1.2404,
+      "step": 1572
+    },
+    {
+      "epoch": 0.12126117792167745,
+      "grad_norm": 4.348257064819336,
+      "learning_rate": 9.783454148836089e-06,
+      "loss": 1.2484,
+      "step": 1573
+    },
+    {
+      "epoch": 0.12133826703669442,
+      "grad_norm": 4.235533237457275,
+      "learning_rate": 9.783090569344418e-06,
+      "loss": 1.0986,
+      "step": 1574
+    },
+    {
+      "epoch": 0.12141535615171138,
+      "grad_norm": 3.815244197845459,
+      "learning_rate": 9.782726691651819e-06,
+      "loss": 0.9998,
+      "step": 1575
+    },
+    {
+      "epoch": 0.12149244526672834,
+      "grad_norm": 3.847841739654541,
+      "learning_rate": 9.782362515780983e-06,
+      "loss": 1.1051,
+      "step": 1576
+    },
+    {
+      "epoch": 0.1215695343817453,
+      "grad_norm": 4.120035648345947,
+      "learning_rate": 9.781998041754613e-06,
+      "loss": 1.1412,
+      "step": 1577
+    },
+    {
+      "epoch": 0.12164662349676225,
+      "grad_norm": 4.0854105949401855,
+      "learning_rate": 9.781633269595432e-06,
+      "loss": 1.1599,
+      "step": 1578
+    },
+    {
+      "epoch": 0.12172371261177922,
+      "grad_norm": 4.094209671020508,
+      "learning_rate": 9.78126819932618e-06,
+      "loss": 1.2077,
+      "step": 1579
+    },
+    {
+      "epoch": 0.12180080172679618,
+      "grad_norm": 4.1311516761779785,
+      "learning_rate": 9.78090283096962e-06,
+      "loss": 1.1408,
+      "step": 1580
+    },
+    {
+      "epoch": 0.12187789084181314,
+      "grad_norm": 3.8271164894104004,
+      "learning_rate": 9.780537164548529e-06,
+      "loss": 1.2063,
+      "step": 1581
+    },
+    {
+      "epoch": 0.12195497995683009,
+      "grad_norm": 3.9459309577941895,
+      "learning_rate": 9.780171200085703e-06,
+      "loss": 1.1439,
+      "step": 1582
+    },
+    {
+      "epoch": 0.12203206907184705,
+      "grad_norm": 4.513016223907471,
+      "learning_rate": 9.779804937603963e-06,
+      "loss": 1.2068,
+      "step": 1583
+    },
+    {
+      "epoch": 0.12210915818686402,
+      "grad_norm": 4.303330421447754,
+      "learning_rate": 9.779438377126138e-06,
+      "loss": 1.2965,
+      "step": 1584
+    },
+    {
+      "epoch": 0.12218624730188098,
+      "grad_norm": 4.000280380249023,
+      "learning_rate": 9.779071518675086e-06,
+      "loss": 1.1383,
+      "step": 1585
+    },
+    {
+      "epoch": 0.12226333641689793,
+      "grad_norm": 3.8880882263183594,
+      "learning_rate": 9.778704362273673e-06,
+      "loss": 1.0986,
+      "step": 1586
+    },
+    {
+      "epoch": 0.12234042553191489,
+      "grad_norm": 4.341008186340332,
+      "learning_rate": 9.778336907944793e-06,
+      "loss": 1.1779,
+      "step": 1587
+    },
+    {
+      "epoch": 0.12241751464693185,
+      "grad_norm": 4.198411464691162,
+      "learning_rate": 9.777969155711356e-06,
+      "loss": 1.1633,
+      "step": 1588
+    },
+    {
+      "epoch": 0.12249460376194882,
+      "grad_norm": 3.935933828353882,
+      "learning_rate": 9.777601105596288e-06,
+      "loss": 1.1661,
+      "step": 1589
+    },
+    {
+      "epoch": 0.12257169287696577,
+      "grad_norm": 3.575408935546875,
+      "learning_rate": 9.777232757622534e-06,
+      "loss": 1.037,
+      "step": 1590
+    },
+    {
+      "epoch": 0.12264878199198273,
+      "grad_norm": 4.11090087890625,
+      "learning_rate": 9.77686411181306e-06,
+      "loss": 1.2121,
+      "step": 1591
+    },
+    {
+      "epoch": 0.12272587110699969,
+      "grad_norm": 4.116640090942383,
+      "learning_rate": 9.77649516819085e-06,
+      "loss": 1.0669,
+      "step": 1592
+    },
+    {
+      "epoch": 0.12280296022201666,
+      "grad_norm": 3.7206943035125732,
+      "learning_rate": 9.776125926778902e-06,
+      "loss": 1.0977,
+      "step": 1593
+    },
+    {
+      "epoch": 0.12288004933703361,
+      "grad_norm": 3.60390305519104,
+      "learning_rate": 9.775756387600239e-06,
+      "loss": 1.0635,
+      "step": 1594
+    },
+    {
+      "epoch": 0.12295713845205057,
+      "grad_norm": 4.16797399520874,
+      "learning_rate": 9.7753865506779e-06,
+      "loss": 1.1748,
+      "step": 1595
+    },
+    {
+      "epoch": 0.12303422756706753,
+      "grad_norm": 3.8636176586151123,
+      "learning_rate": 9.77501641603494e-06,
+      "loss": 1.1328,
+      "step": 1596
+    },
+    {
+      "epoch": 0.12311131668208448,
+      "grad_norm": 4.086244583129883,
+      "learning_rate": 9.77464598369444e-06,
+      "loss": 1.2368,
+      "step": 1597
+    },
+    {
+      "epoch": 0.12318840579710146,
+      "grad_norm": 3.9947853088378906,
+      "learning_rate": 9.77427525367949e-06,
+      "loss": 1.0748,
+      "step": 1598
+    },
+    {
+      "epoch": 0.12326549491211841,
+      "grad_norm": 3.5423076152801514,
+      "learning_rate": 9.773904226013207e-06,
+      "loss": 1.1288,
+      "step": 1599
+    },
+    {
+      "epoch": 0.12334258402713537,
+      "grad_norm": 4.289074897766113,
+      "learning_rate": 9.773532900718717e-06,
+      "loss": 1.2039,
+      "step": 1600
+    },
+    {
+      "epoch": 0.12341967314215233,
+      "grad_norm": 4.612265586853027,
+      "learning_rate": 9.773161277819175e-06,
+      "loss": 1.2929,
+      "step": 1601
+    },
+    {
+      "epoch": 0.12349676225716928,
+      "grad_norm": 4.081154823303223,
+      "learning_rate": 9.772789357337746e-06,
+      "loss": 1.221,
+      "step": 1602
+    },
+    {
+      "epoch": 0.12357385137218625,
+      "grad_norm": 4.336784839630127,
+      "learning_rate": 9.772417139297622e-06,
+      "loss": 1.2045,
+      "step": 1603
+    },
+    {
+      "epoch": 0.12365094048720321,
+      "grad_norm": 4.129827499389648,
+      "learning_rate": 9.772044623722005e-06,
+      "loss": 1.1447,
+      "step": 1604
+    },
+    {
+      "epoch": 0.12372802960222017,
+      "grad_norm": 4.177978515625,
+      "learning_rate": 9.771671810634123e-06,
+      "loss": 1.1951,
+      "step": 1605
+    },
+    {
+      "epoch": 0.12380511871723712,
+      "grad_norm": 3.896273374557495,
+      "learning_rate": 9.771298700057214e-06,
+      "loss": 1.1928,
+      "step": 1606
+    },
+    {
+      "epoch": 0.12388220783225408,
+      "grad_norm": 3.946012258529663,
+      "learning_rate": 9.770925292014542e-06,
+      "loss": 1.1168,
+      "step": 1607
+    },
+    {
+      "epoch": 0.12395929694727105,
+      "grad_norm": 4.167520046234131,
+      "learning_rate": 9.77055158652939e-06,
+      "loss": 1.3042,
+      "step": 1608
+    },
+    {
+      "epoch": 0.124036386062288,
+      "grad_norm": 4.672422885894775,
+      "learning_rate": 9.77017758362505e-06,
+      "loss": 1.2575,
+      "step": 1609
+    },
+    {
+      "epoch": 0.12411347517730496,
+      "grad_norm": 4.5983123779296875,
+      "learning_rate": 9.769803283324847e-06,
+      "loss": 1.1992,
+      "step": 1610
+    },
+    {
+      "epoch": 0.12419056429232192,
+      "grad_norm": 4.06178092956543,
+      "learning_rate": 9.769428685652112e-06,
+      "loss": 1.2428,
+      "step": 1611
+    },
+    {
+      "epoch": 0.12426765340733889,
+      "grad_norm": 4.024518013000488,
+      "learning_rate": 9.769053790630198e-06,
+      "loss": 1.1464,
+      "step": 1612
+    },
+    {
+      "epoch": 0.12434474252235585,
+      "grad_norm": 4.302224636077881,
+      "learning_rate": 9.768678598282481e-06,
+      "loss": 1.0665,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1244218316373728,
+      "grad_norm": 4.330562114715576,
+      "learning_rate": 9.76830310863235e-06,
+      "loss": 1.1544,
+      "step": 1614
+    },
+    {
+      "epoch": 0.12449892075238976,
+      "grad_norm": 4.35994291305542,
+      "learning_rate": 9.767927321703217e-06,
+      "loss": 1.1948,
+      "step": 1615
+    },
+    {
+      "epoch": 0.12457600986740672,
+      "grad_norm": 3.8011748790740967,
+      "learning_rate": 9.767551237518508e-06,
+      "loss": 1.155,
+      "step": 1616
+    },
+    {
+      "epoch": 0.12465309898242369,
+      "grad_norm": 3.746467113494873,
+      "learning_rate": 9.767174856101672e-06,
+      "loss": 1.1757,
+      "step": 1617
+    },
+    {
+      "epoch": 0.12473018809744064,
+      "grad_norm": 3.849261522293091,
+      "learning_rate": 9.766798177476175e-06,
+      "loss": 0.9983,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1248072772124576,
+      "grad_norm": 4.078786849975586,
+      "learning_rate": 9.766421201665498e-06,
+      "loss": 1.2175,
+      "step": 1619
+    },
+    {
+      "epoch": 0.12488436632747456,
+      "grad_norm": 4.006483554840088,
+      "learning_rate": 9.766043928693146e-06,
+      "loss": 1.2068,
+      "step": 1620
+    },
+    {
+      "epoch": 0.12496145544249151,
+      "grad_norm": 4.358697891235352,
+      "learning_rate": 9.765666358582637e-06,
+      "loss": 1.1717,
+      "step": 1621
+    },
+    {
+      "epoch": 0.12503854455750849,
+      "grad_norm": 4.190584182739258,
+      "learning_rate": 9.765288491357514e-06,
+      "loss": 1.1927,
+      "step": 1622
+    },
+    {
+      "epoch": 0.12511563367252543,
+      "grad_norm": 4.142174243927002,
+      "learning_rate": 9.764910327041333e-06,
+      "loss": 1.1538,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1251927227875424,
+      "grad_norm": 3.6901378631591797,
+      "learning_rate": 9.764531865657671e-06,
+      "loss": 1.1287,
+      "step": 1624
+    },
+    {
+      "epoch": 0.12526981190255937,
+      "grad_norm": 3.891272783279419,
+      "learning_rate": 9.764153107230124e-06,
+      "loss": 1.084,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1253469010175763,
+      "grad_norm": 4.083333969116211,
+      "learning_rate": 9.763774051782306e-06,
+      "loss": 1.1683,
+      "step": 1626
+    },
+    {
+      "epoch": 0.12542399013259328,
+      "grad_norm": 3.742480516433716,
+      "learning_rate": 9.763394699337846e-06,
+      "loss": 1.1251,
+      "step": 1627
+    },
+    {
+      "epoch": 0.12550107924761023,
+      "grad_norm": 4.092676639556885,
+      "learning_rate": 9.763015049920397e-06,
+      "loss": 1.1928,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1255781683626272,
+      "grad_norm": 3.6800365447998047,
+      "learning_rate": 9.76263510355363e-06,
+      "loss": 1.1753,
+      "step": 1629
+    },
+    {
+      "epoch": 0.12565525747764417,
+      "grad_norm": 4.610281944274902,
+      "learning_rate": 9.762254860261229e-06,
+      "loss": 1.2049,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1257323465926611,
+      "grad_norm": 3.89145827293396,
+      "learning_rate": 9.761874320066903e-06,
+      "loss": 0.9942,
+      "step": 1631
+    },
+    {
+      "epoch": 0.12580943570767808,
+      "grad_norm": 3.8079841136932373,
+      "learning_rate": 9.761493482994374e-06,
+      "loss": 1.2347,
+      "step": 1632
+    },
+    {
+      "epoch": 0.12588652482269502,
+      "grad_norm": 4.090712547302246,
+      "learning_rate": 9.761112349067387e-06,
+      "loss": 1.1586,
+      "step": 1633
+    },
+    {
+      "epoch": 0.125963613937712,
+      "grad_norm": 3.8824753761291504,
+      "learning_rate": 9.760730918309702e-06,
+      "loss": 1.1221,
+      "step": 1634
+    },
+    {
+      "epoch": 0.12604070305272896,
+      "grad_norm": 3.8684356212615967,
+      "learning_rate": 9.760349190745104e-06,
+      "loss": 1.0676,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1261177921677459,
+      "grad_norm": 4.076964378356934,
+      "learning_rate": 9.759967166397386e-06,
+      "loss": 1.1901,
+      "step": 1636
+    },
+    {
+      "epoch": 0.12619488128276288,
+      "grad_norm": 3.883690357208252,
+      "learning_rate": 9.759584845290368e-06,
+      "loss": 1.179,
+      "step": 1637
+    },
+    {
+      "epoch": 0.12627197039777982,
+      "grad_norm": 3.9495391845703125,
+      "learning_rate": 9.759202227447888e-06,
+      "loss": 1.2468,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1263490595127968,
+      "grad_norm": 4.153524398803711,
+      "learning_rate": 9.758819312893795e-06,
+      "loss": 1.0873,
+      "step": 1639
+    },
+    {
+      "epoch": 0.12642614862781376,
+      "grad_norm": 3.8686201572418213,
+      "learning_rate": 9.758436101651965e-06,
+      "loss": 1.0967,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1265032377428307,
+      "grad_norm": 4.673964977264404,
+      "learning_rate": 9.758052593746287e-06,
+      "loss": 1.2438,
+      "step": 1641
+    },
+    {
+      "epoch": 0.12658032685784767,
+      "grad_norm": 4.521989345550537,
+      "learning_rate": 9.757668789200676e-06,
+      "loss": 1.2298,
+      "step": 1642
+    },
+    {
+      "epoch": 0.12665741597286465,
+      "grad_norm": 4.26433801651001,
+      "learning_rate": 9.757284688039054e-06,
+      "loss": 1.1951,
+      "step": 1643
+    },
+    {
+      "epoch": 0.1267345050878816,
+      "grad_norm": 3.7774715423583984,
+      "learning_rate": 9.75690029028537e-06,
+      "loss": 1.2376,
+      "step": 1644
+    },
+    {
+      "epoch": 0.12681159420289856,
+      "grad_norm": 4.080611228942871,
+      "learning_rate": 9.756515595963591e-06,
+      "loss": 1.1003,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1268886833179155,
+      "grad_norm": 4.495629787445068,
+      "learning_rate": 9.7561306050977e-06,
+      "loss": 1.1129,
+      "step": 1646
+    },
+    {
+      "epoch": 0.12696577243293247,
+      "grad_norm": 4.433581352233887,
+      "learning_rate": 9.755745317711696e-06,
+      "loss": 1.2694,
+      "step": 1647
+    },
+    {
+      "epoch": 0.12704286154794944,
+      "grad_norm": 4.003334045410156,
+      "learning_rate": 9.755359733829604e-06,
+      "loss": 1.1851,
+      "step": 1648
+    },
+    {
+      "epoch": 0.12711995066296639,
+      "grad_norm": 4.140914440155029,
+      "learning_rate": 9.75497385347546e-06,
+      "loss": 1.1482,
+      "step": 1649
+    },
+    {
+      "epoch": 0.12719703977798336,
+      "grad_norm": 4.328746795654297,
+      "learning_rate": 9.754587676673323e-06,
+      "loss": 1.2779,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1272741288930003,
+      "grad_norm": 4.75211238861084,
+      "learning_rate": 9.754201203447268e-06,
+      "loss": 1.2699,
+      "step": 1651
+    },
+    {
+      "epoch": 0.12735121800801727,
+      "grad_norm": 3.97086763381958,
+      "learning_rate": 9.753814433821393e-06,
+      "loss": 1.189,
+      "step": 1652
+    },
+    {
+      "epoch": 0.12742830712303424,
+      "grad_norm": 4.095003604888916,
+      "learning_rate": 9.753427367819808e-06,
+      "loss": 1.1897,
+      "step": 1653
+    },
+    {
+      "epoch": 0.12750539623805118,
+      "grad_norm": 4.253332614898682,
+      "learning_rate": 9.753040005466645e-06,
+      "loss": 1.1711,
+      "step": 1654
+    },
+    {
+      "epoch": 0.12758248535306815,
+      "grad_norm": 3.9506523609161377,
+      "learning_rate": 9.752652346786054e-06,
+      "loss": 1.2071,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1276595744680851,
+      "grad_norm": 4.006979465484619,
+      "learning_rate": 9.752264391802203e-06,
+      "loss": 1.1794,
+      "step": 1656
+    },
+    {
+      "epoch": 0.12773666358310207,
+      "grad_norm": 4.240119457244873,
+      "learning_rate": 9.75187614053928e-06,
+      "loss": 1.1955,
+      "step": 1657
+    },
+    {
+      "epoch": 0.12781375269811904,
+      "grad_norm": 4.19885778427124,
+      "learning_rate": 9.751487593021491e-06,
+      "loss": 1.1181,
+      "step": 1658
+    },
+    {
+      "epoch": 0.12789084181313598,
+      "grad_norm": 3.8711414337158203,
+      "learning_rate": 9.75109874927306e-06,
+      "loss": 1.0409,
+      "step": 1659
+    },
+    {
+      "epoch": 0.12796793092815295,
+      "grad_norm": 4.2193756103515625,
+      "learning_rate": 9.750709609318227e-06,
+      "loss": 1.255,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1280450200431699,
+      "grad_norm": 4.001354694366455,
+      "learning_rate": 9.750320173181256e-06,
+      "loss": 1.0984,
+      "step": 1661
+    },
+    {
+      "epoch": 0.12812210915818686,
+      "grad_norm": 3.9252028465270996,
+      "learning_rate": 9.749930440886424e-06,
+      "loss": 1.0892,
+      "step": 1662
+    },
+    {
+      "epoch": 0.12819919827320383,
+      "grad_norm": 4.056916236877441,
+      "learning_rate": 9.749540412458028e-06,
+      "loss": 1.177,
+      "step": 1663
+    },
+    {
+      "epoch": 0.12827628738822078,
+      "grad_norm": 4.366199970245361,
+      "learning_rate": 9.749150087920386e-06,
+      "loss": 1.1389,
+      "step": 1664
+    },
+    {
+      "epoch": 0.12835337650323775,
+      "grad_norm": 3.8969857692718506,
+      "learning_rate": 9.748759467297835e-06,
+      "loss": 1.1479,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1284304656182547,
+      "grad_norm": 4.55438756942749,
+      "learning_rate": 9.748368550614723e-06,
+      "loss": 1.3332,
+      "step": 1666
+    },
+    {
+      "epoch": 0.12850755473327166,
+      "grad_norm": 4.168787002563477,
+      "learning_rate": 9.747977337895426e-06,
+      "loss": 1.1065,
+      "step": 1667
+    },
+    {
+      "epoch": 0.12858464384828863,
+      "grad_norm": 4.430618762969971,
+      "learning_rate": 9.747585829164332e-06,
+      "loss": 1.176,
+      "step": 1668
+    },
+    {
+      "epoch": 0.12866173296330558,
+      "grad_norm": 4.035445690155029,
+      "learning_rate": 9.747194024445851e-06,
+      "loss": 1.2115,
+      "step": 1669
+    },
+    {
+      "epoch": 0.12873882207832255,
+      "grad_norm": 3.8563239574432373,
+      "learning_rate": 9.746801923764409e-06,
+      "loss": 1.1879,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1288159111933395,
+      "grad_norm": 4.492901802062988,
+      "learning_rate": 9.74640952714445e-06,
+      "loss": 1.2354,
+      "step": 1671
+    },
+    {
+      "epoch": 0.12889300030835646,
+      "grad_norm": 3.8549299240112305,
+      "learning_rate": 9.746016834610438e-06,
+      "loss": 1.1373,
+      "step": 1672
+    },
+    {
+      "epoch": 0.12897008942337343,
+      "grad_norm": 3.6520705223083496,
+      "learning_rate": 9.745623846186858e-06,
+      "loss": 1.0816,
+      "step": 1673
+    },
+    {
+      "epoch": 0.12904717853839037,
+      "grad_norm": 4.377813816070557,
+      "learning_rate": 9.74523056189821e-06,
+      "loss": 1.1365,
+      "step": 1674
+    },
+    {
+      "epoch": 0.12912426765340734,
+      "grad_norm": 3.965122699737549,
+      "learning_rate": 9.744836981769013e-06,
+      "loss": 1.0827,
+      "step": 1675
+    },
+    {
+      "epoch": 0.12920135676842429,
+      "grad_norm": 4.009072303771973,
+      "learning_rate": 9.744443105823802e-06,
+      "loss": 1.1322,
+      "step": 1676
+    },
+    {
+      "epoch": 0.12927844588344126,
+      "grad_norm": 4.5139994621276855,
+      "learning_rate": 9.744048934087138e-06,
+      "loss": 1.2437,
+      "step": 1677
+    },
+    {
+      "epoch": 0.12935553499845823,
+      "grad_norm": 3.827956199645996,
+      "learning_rate": 9.743654466583591e-06,
+      "loss": 1.1468,
+      "step": 1678
+    },
+    {
+      "epoch": 0.12943262411347517,
+      "grad_norm": 3.9646661281585693,
+      "learning_rate": 9.743259703337758e-06,
+      "loss": 1.1022,
+      "step": 1679
+    },
+    {
+      "epoch": 0.12950971322849214,
+      "grad_norm": 3.9882500171661377,
+      "learning_rate": 9.742864644374248e-06,
+      "loss": 1.1333,
+      "step": 1680
+    },
+    {
+      "epoch": 0.12958680234350908,
+      "grad_norm": 3.8956985473632812,
+      "learning_rate": 9.74246928971769e-06,
+      "loss": 1.2208,
+      "step": 1681
+    },
+    {
+      "epoch": 0.12966389145852605,
+      "grad_norm": 4.2606000900268555,
+      "learning_rate": 9.742073639392735e-06,
+      "loss": 1.2494,
+      "step": 1682
+    },
+    {
+      "epoch": 0.12974098057354302,
+      "grad_norm": 4.1300530433654785,
+      "learning_rate": 9.741677693424048e-06,
+      "loss": 1.111,
+      "step": 1683
+    },
+    {
+      "epoch": 0.12981806968855997,
+      "grad_norm": 3.594024896621704,
+      "learning_rate": 9.741281451836313e-06,
+      "loss": 1.0742,
+      "step": 1684
+    },
+    {
+      "epoch": 0.12989515880357694,
+      "grad_norm": 4.158742427825928,
+      "learning_rate": 9.740884914654237e-06,
+      "loss": 1.0348,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1299722479185939,
+      "grad_norm": 3.8110404014587402,
+      "learning_rate": 9.74048808190254e-06,
+      "loss": 1.1757,
+      "step": 1686
+    },
+    {
+      "epoch": 0.13004933703361085,
+      "grad_norm": 3.863544464111328,
+      "learning_rate": 9.740090953605963e-06,
+      "loss": 1.0151,
+      "step": 1687
+    },
+    {
+      "epoch": 0.13012642614862782,
+      "grad_norm": 4.125679016113281,
+      "learning_rate": 9.739693529789264e-06,
+      "loss": 1.1922,
+      "step": 1688
+    },
+    {
+      "epoch": 0.13020351526364476,
+      "grad_norm": 4.0904083251953125,
+      "learning_rate": 9.73929581047722e-06,
+      "loss": 1.0156,
+      "step": 1689
+    },
+    {
+      "epoch": 0.13028060437866174,
+      "grad_norm": 4.615362644195557,
+      "learning_rate": 9.73889779569463e-06,
+      "loss": 1.2547,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1303576934936787,
+      "grad_norm": 3.9230780601501465,
+      "learning_rate": 9.738499485466304e-06,
+      "loss": 1.1324,
+      "step": 1691
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 4.175113201141357,
+      "learning_rate": 9.738100879817077e-06,
+      "loss": 1.1832,
+      "step": 1692
+    },
+    {
+      "epoch": 0.13051187172371262,
+      "grad_norm": 3.8793387413024902,
+      "learning_rate": 9.7377019787718e-06,
+      "loss": 1.0902,
+      "step": 1693
+    },
+    {
+      "epoch": 0.13058896083872956,
+      "grad_norm": 4.145679950714111,
+      "learning_rate": 9.73730278235534e-06,
+      "loss": 1.1542,
+      "step": 1694
+    },
+    {
+      "epoch": 0.13066604995374653,
+      "grad_norm": 4.0476179122924805,
+      "learning_rate": 9.736903290592589e-06,
+      "loss": 1.2334,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1307431390687635,
+      "grad_norm": 3.9774255752563477,
+      "learning_rate": 9.73650350350845e-06,
+      "loss": 1.1498,
+      "step": 1696
+    },
+    {
+      "epoch": 0.13082022818378045,
+      "grad_norm": 3.953183174133301,
+      "learning_rate": 9.736103421127847e-06,
+      "loss": 1.1513,
+      "step": 1697
+    },
+    {
+      "epoch": 0.13089731729879742,
+      "grad_norm": 4.064455986022949,
+      "learning_rate": 9.735703043475727e-06,
+      "loss": 1.2054,
+      "step": 1698
+    },
+    {
+      "epoch": 0.13097440641381436,
+      "grad_norm": 4.00160026550293,
+      "learning_rate": 9.735302370577049e-06,
+      "loss": 1.13,
+      "step": 1699
+    },
+    {
+      "epoch": 0.13105149552883133,
+      "grad_norm": 4.219839096069336,
+      "learning_rate": 9.734901402456792e-06,
+      "loss": 1.1448,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1311285846438483,
+      "grad_norm": 4.194035053253174,
+      "learning_rate": 9.734500139139952e-06,
+      "loss": 1.1548,
+      "step": 1701
+    },
+    {
+      "epoch": 0.13120567375886524,
+      "grad_norm": 3.898303985595703,
+      "learning_rate": 9.734098580651555e-06,
+      "loss": 1.119,
+      "step": 1702
+    },
+    {
+      "epoch": 0.13128276287388221,
+      "grad_norm": 4.194339752197266,
+      "learning_rate": 9.733696727016626e-06,
+      "loss": 1.1971,
+      "step": 1703
+    },
+    {
+      "epoch": 0.13135985198889916,
+      "grad_norm": 4.266892433166504,
+      "learning_rate": 9.733294578260224e-06,
+      "loss": 1.1156,
+      "step": 1704
+    },
+    {
+      "epoch": 0.13143694110391613,
+      "grad_norm": 3.6440038681030273,
+      "learning_rate": 9.73289213440742e-06,
+      "loss": 1.1899,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1315140302189331,
+      "grad_norm": 4.273855686187744,
+      "learning_rate": 9.7324893954833e-06,
+      "loss": 1.1433,
+      "step": 1706
+    },
+    {
+      "epoch": 0.13159111933395004,
+      "grad_norm": 3.3972716331481934,
+      "learning_rate": 9.73208636151298e-06,
+      "loss": 1.0053,
+      "step": 1707
+    },
+    {
+      "epoch": 0.131668208448967,
+      "grad_norm": 3.7814152240753174,
+      "learning_rate": 9.731683032521583e-06,
+      "loss": 1.16,
+      "step": 1708
+    },
+    {
+      "epoch": 0.13174529756398395,
+      "grad_norm": 4.149326324462891,
+      "learning_rate": 9.731279408534255e-06,
+      "loss": 1.1135,
+      "step": 1709
+    },
+    {
+      "epoch": 0.13182238667900092,
+      "grad_norm": 4.2798051834106445,
+      "learning_rate": 9.73087548957616e-06,
+      "loss": 1.2493,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1318994757940179,
+      "grad_norm": 3.8137776851654053,
+      "learning_rate": 9.730471275672478e-06,
+      "loss": 1.1796,
+      "step": 1711
+    },
+    {
+      "epoch": 0.13197656490903484,
+      "grad_norm": 4.015539646148682,
+      "learning_rate": 9.730066766848415e-06,
+      "loss": 1.1601,
+      "step": 1712
+    },
+    {
+      "epoch": 0.1320536540240518,
+      "grad_norm": 3.887502431869507,
+      "learning_rate": 9.729661963129183e-06,
+      "loss": 1.1712,
+      "step": 1713
+    },
+    {
+      "epoch": 0.13213074313906875,
+      "grad_norm": 4.104816436767578,
+      "learning_rate": 9.729256864540025e-06,
+      "loss": 1.1473,
+      "step": 1714
+    },
+    {
+      "epoch": 0.13220783225408572,
+      "grad_norm": 3.9987316131591797,
+      "learning_rate": 9.728851471106195e-06,
+      "loss": 1.223,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1322849213691027,
+      "grad_norm": 4.665685176849365,
+      "learning_rate": 9.728445782852967e-06,
+      "loss": 1.1008,
+      "step": 1716
+    },
+    {
+      "epoch": 0.13236201048411964,
+      "grad_norm": 4.16191291809082,
+      "learning_rate": 9.728039799805635e-06,
+      "loss": 1.0905,
+      "step": 1717
+    },
+    {
+      "epoch": 0.1324390995991366,
+      "grad_norm": 4.354928016662598,
+      "learning_rate": 9.727633521989505e-06,
+      "loss": 1.1924,
+      "step": 1718
+    },
+    {
+      "epoch": 0.13251618871415355,
+      "grad_norm": 4.347275733947754,
+      "learning_rate": 9.727226949429913e-06,
+      "loss": 1.169,
+      "step": 1719
+    },
+    {
+      "epoch": 0.13259327782917052,
+      "grad_norm": 4.061772346496582,
+      "learning_rate": 9.726820082152204e-06,
+      "loss": 1.2223,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1326703669441875,
+      "grad_norm": 4.027084827423096,
+      "learning_rate": 9.726412920181742e-06,
+      "loss": 1.1924,
+      "step": 1721
+    },
+    {
+      "epoch": 0.13274745605920443,
+      "grad_norm": 4.302231788635254,
+      "learning_rate": 9.726005463543913e-06,
+      "loss": 1.1269,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1328245451742214,
+      "grad_norm": 3.7703239917755127,
+      "learning_rate": 9.725597712264123e-06,
+      "loss": 1.1173,
+      "step": 1723
+    },
+    {
+      "epoch": 0.13290163428923835,
+      "grad_norm": 3.777994155883789,
+      "learning_rate": 9.725189666367787e-06,
+      "loss": 1.0722,
+      "step": 1724
+    },
+    {
+      "epoch": 0.13297872340425532,
+      "grad_norm": 3.945777177810669,
+      "learning_rate": 9.724781325880348e-06,
+      "loss": 1.0914,
+      "step": 1725
+    },
+    {
+      "epoch": 0.1330558125192723,
+      "grad_norm": 4.009458065032959,
+      "learning_rate": 9.724372690827264e-06,
+      "loss": 1.0405,
+      "step": 1726
+    },
+    {
+      "epoch": 0.13313290163428923,
+      "grad_norm": 4.944936752319336,
+      "learning_rate": 9.723963761234013e-06,
+      "loss": 1.2138,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1332099907493062,
+      "grad_norm": 3.801499605178833,
+      "learning_rate": 9.723554537126084e-06,
+      "loss": 1.1508,
+      "step": 1728
+    },
+    {
+      "epoch": 0.13328707986432317,
+      "grad_norm": 3.6166648864746094,
+      "learning_rate": 9.723145018528995e-06,
+      "loss": 1.1144,
+      "step": 1729
+    },
+    {
+      "epoch": 0.13336416897934011,
+      "grad_norm": 3.8566949367523193,
+      "learning_rate": 9.722735205468277e-06,
+      "loss": 1.0411,
+      "step": 1730
+    },
+    {
+      "epoch": 0.13344125809435708,
+      "grad_norm": 3.865793466567993,
+      "learning_rate": 9.722325097969477e-06,
+      "loss": 1.153,
+      "step": 1731
+    },
+    {
+      "epoch": 0.13351834720937403,
+      "grad_norm": 4.346949577331543,
+      "learning_rate": 9.721914696058165e-06,
+      "loss": 1.2115,
+      "step": 1732
+    },
+    {
+      "epoch": 0.133595436324391,
+      "grad_norm": 3.8929085731506348,
+      "learning_rate": 9.721503999759926e-06,
+      "loss": 1.1708,
+      "step": 1733
+    },
+    {
+      "epoch": 0.13367252543940797,
+      "grad_norm": 4.4507575035095215,
+      "learning_rate": 9.721093009100368e-06,
+      "loss": 1.2546,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1337496145544249,
+      "grad_norm": 4.165073394775391,
+      "learning_rate": 9.720681724105112e-06,
+      "loss": 1.1604,
+      "step": 1735
+    },
+    {
+      "epoch": 0.13382670366944188,
+      "grad_norm": 3.6533870697021484,
+      "learning_rate": 9.7202701447998e-06,
+      "loss": 1.0856,
+      "step": 1736
+    },
+    {
+      "epoch": 0.13390379278445882,
+      "grad_norm": 3.7656784057617188,
+      "learning_rate": 9.71985827121009e-06,
+      "loss": 1.1357,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1339808818994758,
+      "grad_norm": 4.213956356048584,
+      "learning_rate": 9.719446103361662e-06,
+      "loss": 1.1676,
+      "step": 1738
+    },
+    {
+      "epoch": 0.13405797101449277,
+      "grad_norm": 4.172744274139404,
+      "learning_rate": 9.719033641280211e-06,
+      "loss": 1.2312,
+      "step": 1739
+    },
+    {
+      "epoch": 0.1341350601295097,
+      "grad_norm": 4.124050140380859,
+      "learning_rate": 9.718620884991455e-06,
+      "loss": 1.1368,
+      "step": 1740
+    },
+    {
+      "epoch": 0.13421214924452668,
+      "grad_norm": 4.168154716491699,
+      "learning_rate": 9.718207834521124e-06,
+      "loss": 1.0489,
+      "step": 1741
+    },
+    {
+      "epoch": 0.13428923835954362,
+      "grad_norm": 4.139477252960205,
+      "learning_rate": 9.71779448989497e-06,
+      "loss": 1.0745,
+      "step": 1742
+    },
+    {
+      "epoch": 0.1343663274745606,
+      "grad_norm": 4.274909496307373,
+      "learning_rate": 9.717380851138765e-06,
+      "loss": 1.0202,
+      "step": 1743
+    },
+    {
+      "epoch": 0.13444341658957756,
+      "grad_norm": 3.696258306503296,
+      "learning_rate": 9.716966918278295e-06,
+      "loss": 1.1022,
+      "step": 1744
+    },
+    {
+      "epoch": 0.1345205057045945,
+      "grad_norm": 3.8992645740509033,
+      "learning_rate": 9.716552691339369e-06,
+      "loss": 1.0238,
+      "step": 1745
+    },
+    {
+      "epoch": 0.13459759481961148,
+      "grad_norm": 4.183124542236328,
+      "learning_rate": 9.716138170347808e-06,
+      "loss": 1.1277,
+      "step": 1746
+    },
+    {
+      "epoch": 0.13467468393462842,
+      "grad_norm": 3.98201584815979,
+      "learning_rate": 9.71572335532946e-06,
+      "loss": 1.2144,
+      "step": 1747
+    },
+    {
+      "epoch": 0.1347517730496454,
+      "grad_norm": 4.045650005340576,
+      "learning_rate": 9.715308246310181e-06,
+      "loss": 1.1166,
+      "step": 1748
+    },
+    {
+      "epoch": 0.13482886216466236,
+      "grad_norm": 4.4853973388671875,
+      "learning_rate": 9.714892843315857e-06,
+      "loss": 1.1546,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1349059512796793,
+      "grad_norm": 4.153515815734863,
+      "learning_rate": 9.714477146372383e-06,
+      "loss": 1.0463,
+      "step": 1750
+    },
+    {
+      "epoch": 0.13498304039469627,
+      "grad_norm": 4.120546817779541,
+      "learning_rate": 9.714061155505673e-06,
+      "loss": 1.1675,
+      "step": 1751
+    },
+    {
+      "epoch": 0.13506012950971322,
+      "grad_norm": 4.298232555389404,
+      "learning_rate": 9.713644870741669e-06,
+      "loss": 1.1224,
+      "step": 1752
+    },
+    {
+      "epoch": 0.1351372186247302,
+      "grad_norm": 3.9219980239868164,
+      "learning_rate": 9.713228292106319e-06,
+      "loss": 1.1257,
+      "step": 1753
+    },
+    {
+      "epoch": 0.13521430773974716,
+      "grad_norm": 3.6702487468719482,
+      "learning_rate": 9.712811419625592e-06,
+      "loss": 1.1176,
+      "step": 1754
+    },
+    {
+      "epoch": 0.1352913968547641,
+      "grad_norm": 4.173069953918457,
+      "learning_rate": 9.712394253325483e-06,
+      "loss": 1.07,
+      "step": 1755
+    },
+    {
+      "epoch": 0.13536848596978107,
+      "grad_norm": 4.047333240509033,
+      "learning_rate": 9.711976793232e-06,
+      "loss": 1.1868,
+      "step": 1756
+    },
+    {
+      "epoch": 0.13544557508479801,
+      "grad_norm": 4.153733730316162,
+      "learning_rate": 9.711559039371165e-06,
+      "loss": 1.1111,
+      "step": 1757
+    },
+    {
+      "epoch": 0.13552266419981499,
+      "grad_norm": 4.564488410949707,
+      "learning_rate": 9.711140991769028e-06,
+      "loss": 1.1292,
+      "step": 1758
+    },
+    {
+      "epoch": 0.13559975331483196,
+      "grad_norm": 4.647412300109863,
+      "learning_rate": 9.710722650451649e-06,
+      "loss": 1.1204,
+      "step": 1759
+    },
+    {
+      "epoch": 0.1356768424298489,
+      "grad_norm": 4.287388324737549,
+      "learning_rate": 9.71030401544511e-06,
+      "loss": 1.1814,
+      "step": 1760
+    },
+    {
+      "epoch": 0.13575393154486587,
+      "grad_norm": 4.7093024253845215,
+      "learning_rate": 9.709885086775512e-06,
+      "loss": 1.2857,
+      "step": 1761
+    },
+    {
+      "epoch": 0.1358310206598828,
+      "grad_norm": 4.541007041931152,
+      "learning_rate": 9.709465864468971e-06,
+      "loss": 1.2471,
+      "step": 1762
+    },
+    {
+      "epoch": 0.13590810977489978,
+      "grad_norm": 4.468155860900879,
+      "learning_rate": 9.709046348551626e-06,
+      "loss": 1.1188,
+      "step": 1763
+    },
+    {
+      "epoch": 0.13598519888991675,
+      "grad_norm": 4.554619312286377,
+      "learning_rate": 9.708626539049628e-06,
+      "loss": 1.2205,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1360622880049337,
+      "grad_norm": 3.8080332279205322,
+      "learning_rate": 9.708206435989152e-06,
+      "loss": 1.0702,
+      "step": 1765
+    },
+    {
+      "epoch": 0.13613937711995067,
+      "grad_norm": 4.381828308105469,
+      "learning_rate": 9.707786039396389e-06,
+      "loss": 0.9736,
+      "step": 1766
+    },
+    {
+      "epoch": 0.1362164662349676,
+      "grad_norm": 4.196829319000244,
+      "learning_rate": 9.70736534929755e-06,
+      "loss": 1.1242,
+      "step": 1767
+    },
+    {
+      "epoch": 0.13629355534998458,
+      "grad_norm": 4.237636566162109,
+      "learning_rate": 9.70694436571886e-06,
+      "loss": 1.1547,
+      "step": 1768
+    },
+    {
+      "epoch": 0.13637064446500155,
+      "grad_norm": 4.274960517883301,
+      "learning_rate": 9.706523088686568e-06,
+      "loss": 1.1723,
+      "step": 1769
+    },
+    {
+      "epoch": 0.1364477335800185,
+      "grad_norm": 4.287873268127441,
+      "learning_rate": 9.706101518226939e-06,
+      "loss": 1.1283,
+      "step": 1770
+    },
+    {
+      "epoch": 0.13652482269503546,
+      "grad_norm": 4.142662525177002,
+      "learning_rate": 9.705679654366249e-06,
+      "loss": 1.1434,
+      "step": 1771
+    },
+    {
+      "epoch": 0.13660191181005243,
+      "grad_norm": 4.222457408905029,
+      "learning_rate": 9.705257497130807e-06,
+      "loss": 1.2198,
+      "step": 1772
+    },
+    {
+      "epoch": 0.13667900092506938,
+      "grad_norm": 4.041723251342773,
+      "learning_rate": 9.704835046546928e-06,
+      "loss": 1.0801,
+      "step": 1773
+    },
+    {
+      "epoch": 0.13675609004008635,
+      "grad_norm": 3.751447916030884,
+      "learning_rate": 9.704412302640951e-06,
+      "loss": 1.1612,
+      "step": 1774
+    },
+    {
+      "epoch": 0.1368331791551033,
+      "grad_norm": 4.4663615226745605,
+      "learning_rate": 9.703989265439233e-06,
+      "loss": 1.0822,
+      "step": 1775
+    },
+    {
+      "epoch": 0.13691026827012026,
+      "grad_norm": 3.6068341732025146,
+      "learning_rate": 9.703565934968146e-06,
+      "loss": 1.1349,
+      "step": 1776
+    },
+    {
+      "epoch": 0.13698735738513723,
+      "grad_norm": 4.411525726318359,
+      "learning_rate": 9.703142311254083e-06,
+      "loss": 1.1509,
+      "step": 1777
+    },
+    {
+      "epoch": 0.13706444650015417,
+      "grad_norm": 4.782855033874512,
+      "learning_rate": 9.702718394323456e-06,
+      "loss": 1.1925,
+      "step": 1778
+    },
+    {
+      "epoch": 0.13714153561517115,
+      "grad_norm": 3.9642584323883057,
+      "learning_rate": 9.702294184202692e-06,
+      "loss": 1.1764,
+      "step": 1779
+    },
+    {
+      "epoch": 0.1372186247301881,
+      "grad_norm": 3.878993272781372,
+      "learning_rate": 9.70186968091824e-06,
+      "loss": 1.0982,
+      "step": 1780
+    },
+    {
+      "epoch": 0.13729571384520506,
+      "grad_norm": 4.065914630889893,
+      "learning_rate": 9.701444884496564e-06,
+      "loss": 1.0681,
+      "step": 1781
+    },
+    {
+      "epoch": 0.13737280296022203,
+      "grad_norm": 4.171785354614258,
+      "learning_rate": 9.701019794964151e-06,
+      "loss": 1.0463,
+      "step": 1782
+    },
+    {
+      "epoch": 0.13744989207523897,
+      "grad_norm": 4.044963359832764,
+      "learning_rate": 9.700594412347499e-06,
+      "loss": 1.1055,
+      "step": 1783
+    },
+    {
+      "epoch": 0.13752698119025594,
+      "grad_norm": 4.143056869506836,
+      "learning_rate": 9.700168736673133e-06,
+      "loss": 1.0994,
+      "step": 1784
+    },
+    {
+      "epoch": 0.13760407030527289,
+      "grad_norm": 4.355381965637207,
+      "learning_rate": 9.699742767967586e-06,
+      "loss": 1.2089,
+      "step": 1785
+    },
+    {
+      "epoch": 0.13768115942028986,
+      "grad_norm": 3.8300368785858154,
+      "learning_rate": 9.699316506257421e-06,
+      "loss": 1.0759,
+      "step": 1786
+    },
+    {
+      "epoch": 0.13775824853530683,
+      "grad_norm": 3.7723491191864014,
+      "learning_rate": 9.698889951569208e-06,
+      "loss": 1.0988,
+      "step": 1787
+    },
+    {
+      "epoch": 0.13783533765032377,
+      "grad_norm": 3.426234245300293,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 1.129,
+      "step": 1788
+    },
+    {
+      "epoch": 0.13791242676534074,
+      "grad_norm": 3.9809820652008057,
+      "learning_rate": 9.698035963365038e-06,
+      "loss": 1.0853,
+      "step": 1789
+    },
+    {
+      "epoch": 0.13798951588035768,
+      "grad_norm": 4.098197937011719,
+      "learning_rate": 9.697608529902321e-06,
+      "loss": 1.2014,
+      "step": 1790
+    },
+    {
+      "epoch": 0.13806660499537465,
+      "grad_norm": 4.355566501617432,
+      "learning_rate": 9.697180803568042e-06,
+      "loss": 1.091,
+      "step": 1791
+    },
+    {
+      "epoch": 0.13814369411039162,
+      "grad_norm": 3.870091199874878,
+      "learning_rate": 9.69675278438887e-06,
+      "loss": 1.0783,
+      "step": 1792
+    },
+    {
+      "epoch": 0.13822078322540857,
+      "grad_norm": 3.829244375228882,
+      "learning_rate": 9.696324472391486e-06,
+      "loss": 1.0885,
+      "step": 1793
+    },
+    {
+      "epoch": 0.13829787234042554,
+      "grad_norm": 3.7947609424591064,
+      "learning_rate": 9.695895867602591e-06,
+      "loss": 0.9867,
+      "step": 1794
+    },
+    {
+      "epoch": 0.13837496145544248,
+      "grad_norm": 4.241781234741211,
+      "learning_rate": 9.695466970048912e-06,
+      "loss": 1.1875,
+      "step": 1795
+    },
+    {
+      "epoch": 0.13845205057045945,
+      "grad_norm": 4.397819995880127,
+      "learning_rate": 9.695037779757185e-06,
+      "loss": 1.2111,
+      "step": 1796
+    },
+    {
+      "epoch": 0.13852913968547642,
+      "grad_norm": 4.159852504730225,
+      "learning_rate": 9.694608296754168e-06,
+      "loss": 1.1907,
+      "step": 1797
+    },
+    {
+      "epoch": 0.13860622880049336,
+      "grad_norm": 3.971301317214966,
+      "learning_rate": 9.69417852106664e-06,
+      "loss": 1.1413,
+      "step": 1798
+    },
+    {
+      "epoch": 0.13868331791551033,
+      "grad_norm": 4.702109336853027,
+      "learning_rate": 9.693748452721392e-06,
+      "loss": 1.046,
+      "step": 1799
+    },
+    {
+      "epoch": 0.13876040703052728,
+      "grad_norm": 4.870896339416504,
+      "learning_rate": 9.693318091745237e-06,
+      "loss": 1.2241,
+      "step": 1800
+    },
+    {
+      "epoch": 0.13883749614554425,
+      "grad_norm": 3.7580301761627197,
+      "learning_rate": 9.692887438165007e-06,
+      "loss": 1.1777,
+      "step": 1801
+    },
+    {
+      "epoch": 0.13891458526056122,
+      "grad_norm": 3.787914276123047,
+      "learning_rate": 9.692456492007548e-06,
+      "loss": 1.1804,
+      "step": 1802
+    },
+    {
+      "epoch": 0.13899167437557816,
+      "grad_norm": 3.9445436000823975,
+      "learning_rate": 9.692025253299732e-06,
+      "loss": 1.135,
+      "step": 1803
+    },
+    {
+      "epoch": 0.13906876349059513,
+      "grad_norm": 4.596111297607422,
+      "learning_rate": 9.691593722068442e-06,
+      "loss": 1.1305,
+      "step": 1804
+    },
+    {
+      "epoch": 0.13914585260561207,
+      "grad_norm": 4.179446220397949,
+      "learning_rate": 9.69116189834058e-06,
+      "loss": 1.1967,
+      "step": 1805
+    },
+    {
+      "epoch": 0.13922294172062905,
+      "grad_norm": 3.8127710819244385,
+      "learning_rate": 9.69072978214307e-06,
+      "loss": 1.1141,
+      "step": 1806
+    },
+    {
+      "epoch": 0.13930003083564602,
+      "grad_norm": 4.053515434265137,
+      "learning_rate": 9.690297373502855e-06,
+      "loss": 1.2279,
+      "step": 1807
+    },
+    {
+      "epoch": 0.13937711995066296,
+      "grad_norm": 3.7114830017089844,
+      "learning_rate": 9.689864672446887e-06,
+      "loss": 1.0809,
+      "step": 1808
+    },
+    {
+      "epoch": 0.13945420906567993,
+      "grad_norm": 3.6122548580169678,
+      "learning_rate": 9.689431679002148e-06,
+      "loss": 1.0588,
+      "step": 1809
+    },
+    {
+      "epoch": 0.13953129818069687,
+      "grad_norm": 4.3173604011535645,
+      "learning_rate": 9.68899839319563e-06,
+      "loss": 1.1858,
+      "step": 1810
+    },
+    {
+      "epoch": 0.13960838729571384,
+      "grad_norm": 3.7968335151672363,
+      "learning_rate": 9.688564815054349e-06,
+      "loss": 1.1411,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1396854764107308,
+      "grad_norm": 4.016263961791992,
+      "learning_rate": 9.688130944605332e-06,
+      "loss": 1.0934,
+      "step": 1812
+    },
+    {
+      "epoch": 0.13976256552574776,
+      "grad_norm": 3.948855400085449,
+      "learning_rate": 9.687696781875634e-06,
+      "loss": 1.1486,
+      "step": 1813
+    },
+    {
+      "epoch": 0.13983965464076473,
+      "grad_norm": 4.118218898773193,
+      "learning_rate": 9.687262326892317e-06,
+      "loss": 1.1233,
+      "step": 1814
+    },
+    {
+      "epoch": 0.1399167437557817,
+      "grad_norm": 4.039265155792236,
+      "learning_rate": 9.68682757968247e-06,
+      "loss": 1.0337,
+      "step": 1815
+    },
+    {
+      "epoch": 0.13999383287079864,
+      "grad_norm": 3.6836297512054443,
+      "learning_rate": 9.686392540273198e-06,
+      "loss": 1.1196,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1400709219858156,
+      "grad_norm": 4.069727420806885,
+      "learning_rate": 9.685957208691623e-06,
+      "loss": 1.1305,
+      "step": 1817
+    },
+    {
+      "epoch": 0.14014801110083255,
+      "grad_norm": 3.9679081439971924,
+      "learning_rate": 9.685521584964885e-06,
+      "loss": 1.1529,
+      "step": 1818
+    },
+    {
+      "epoch": 0.14022510021584952,
+      "grad_norm": 3.8460636138916016,
+      "learning_rate": 9.685085669120142e-06,
+      "loss": 1.0946,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1403021893308665,
+      "grad_norm": 4.128662586212158,
+      "learning_rate": 9.684649461184574e-06,
+      "loss": 1.2238,
+      "step": 1820
+    },
+    {
+      "epoch": 0.14037927844588344,
+      "grad_norm": 4.167182922363281,
+      "learning_rate": 9.684212961185374e-06,
+      "loss": 0.9351,
+      "step": 1821
+    },
+    {
+      "epoch": 0.1404563675609004,
+      "grad_norm": 4.262301921844482,
+      "learning_rate": 9.683776169149755e-06,
+      "loss": 1.1101,
+      "step": 1822
+    },
+    {
+      "epoch": 0.14053345667591735,
+      "grad_norm": 4.976638317108154,
+      "learning_rate": 9.683339085104952e-06,
+      "loss": 1.1853,
+      "step": 1823
+    },
+    {
+      "epoch": 0.14061054579093432,
+      "grad_norm": 4.163305282592773,
+      "learning_rate": 9.68290170907821e-06,
+      "loss": 1.1477,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1406876349059513,
+      "grad_norm": 4.624197483062744,
+      "learning_rate": 9.682464041096801e-06,
+      "loss": 1.2981,
+      "step": 1825
+    },
+    {
+      "epoch": 0.14076472402096823,
+      "grad_norm": 4.289773464202881,
+      "learning_rate": 9.682026081188009e-06,
+      "loss": 1.1704,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1408418131359852,
+      "grad_norm": 5.031392574310303,
+      "learning_rate": 9.681587829379143e-06,
+      "loss": 1.3166,
+      "step": 1827
+    },
+    {
+      "epoch": 0.14091890225100215,
+      "grad_norm": 4.640775203704834,
+      "learning_rate": 9.68114928569752e-06,
+      "loss": 1.0848,
+      "step": 1828
+    },
+    {
+      "epoch": 0.14099599136601912,
+      "grad_norm": 3.823460817337036,
+      "learning_rate": 9.680710450170482e-06,
+      "loss": 1.1245,
+      "step": 1829
+    },
+    {
+      "epoch": 0.1410730804810361,
+      "grad_norm": 4.686859130859375,
+      "learning_rate": 9.680271322825392e-06,
+      "loss": 1.1221,
+      "step": 1830
+    },
+    {
+      "epoch": 0.14115016959605303,
+      "grad_norm": 4.298460960388184,
+      "learning_rate": 9.679831903689624e-06,
+      "loss": 1.1457,
+      "step": 1831
+    },
+    {
+      "epoch": 0.14122725871107,
+      "grad_norm": 4.24713659286499,
+      "learning_rate": 9.679392192790573e-06,
+      "loss": 1.1776,
+      "step": 1832
+    },
+    {
+      "epoch": 0.14130434782608695,
+      "grad_norm": 4.055549621582031,
+      "learning_rate": 9.678952190155655e-06,
+      "loss": 1.1615,
+      "step": 1833
+    },
+    {
+      "epoch": 0.14138143694110392,
+      "grad_norm": 4.140902042388916,
+      "learning_rate": 9.678511895812301e-06,
+      "loss": 1.0056,
+      "step": 1834
+    },
+    {
+      "epoch": 0.1414585260561209,
+      "grad_norm": 4.143174648284912,
+      "learning_rate": 9.678071309787962e-06,
+      "loss": 1.1289,
+      "step": 1835
+    },
+    {
+      "epoch": 0.14153561517113783,
+      "grad_norm": 3.9411823749542236,
+      "learning_rate": 9.677630432110103e-06,
+      "loss": 1.1361,
+      "step": 1836
+    },
+    {
+      "epoch": 0.1416127042861548,
+      "grad_norm": 4.089599132537842,
+      "learning_rate": 9.677189262806213e-06,
+      "loss": 1.1156,
+      "step": 1837
+    },
+    {
+      "epoch": 0.14168979340117174,
+      "grad_norm": 3.4913694858551025,
+      "learning_rate": 9.676747801903798e-06,
+      "loss": 1.0064,
+      "step": 1838
+    },
+    {
+      "epoch": 0.1417668825161887,
+      "grad_norm": 3.990382432937622,
+      "learning_rate": 9.676306049430377e-06,
+      "loss": 1.1155,
+      "step": 1839
+    },
+    {
+      "epoch": 0.14184397163120568,
+      "grad_norm": 4.364477634429932,
+      "learning_rate": 9.675864005413494e-06,
+      "loss": 1.1699,
+      "step": 1840
+    },
+    {
+      "epoch": 0.14192106074622263,
+      "grad_norm": 4.722569942474365,
+      "learning_rate": 9.675421669880707e-06,
+      "loss": 1.1372,
+      "step": 1841
+    },
+    {
+      "epoch": 0.1419981498612396,
+      "grad_norm": 4.089046478271484,
+      "learning_rate": 9.674979042859593e-06,
+      "loss": 1.1623,
+      "step": 1842
+    },
+    {
+      "epoch": 0.14207523897625654,
+      "grad_norm": 3.667163610458374,
+      "learning_rate": 9.67453612437775e-06,
+      "loss": 1.0287,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1421523280912735,
+      "grad_norm": 4.879021167755127,
+      "learning_rate": 9.674092914462788e-06,
+      "loss": 1.2368,
+      "step": 1844
+    },
+    {
+      "epoch": 0.14222941720629048,
+      "grad_norm": 3.920139789581299,
+      "learning_rate": 9.67364941314234e-06,
+      "loss": 1.049,
+      "step": 1845
+    },
+    {
+      "epoch": 0.14230650632130742,
+      "grad_norm": 3.814567804336548,
+      "learning_rate": 9.673205620444057e-06,
+      "loss": 1.1007,
+      "step": 1846
+    },
+    {
+      "epoch": 0.1423835954363244,
+      "grad_norm": 3.913419008255005,
+      "learning_rate": 9.672761536395608e-06,
+      "loss": 1.1205,
+      "step": 1847
+    },
+    {
+      "epoch": 0.14246068455134134,
+      "grad_norm": 3.8160576820373535,
+      "learning_rate": 9.672317161024679e-06,
+      "loss": 1.155,
+      "step": 1848
+    },
+    {
+      "epoch": 0.1425377736663583,
+      "grad_norm": 3.9836699962615967,
+      "learning_rate": 9.67187249435897e-06,
+      "loss": 1.0829,
+      "step": 1849
+    },
+    {
+      "epoch": 0.14261486278137528,
+      "grad_norm": 3.8983733654022217,
+      "learning_rate": 9.67142753642621e-06,
+      "loss": 1.0859,
+      "step": 1850
+    },
+    {
+      "epoch": 0.14269195189639222,
+      "grad_norm": 4.181762218475342,
+      "learning_rate": 9.670982287254136e-06,
+      "loss": 1.0088,
+      "step": 1851
+    },
+    {
+      "epoch": 0.1427690410114092,
+      "grad_norm": 4.015240669250488,
+      "learning_rate": 9.670536746870507e-06,
+      "loss": 1.1885,
+      "step": 1852
+    },
+    {
+      "epoch": 0.14284613012642614,
+      "grad_norm": 4.08405876159668,
+      "learning_rate": 9.670090915303103e-06,
+      "loss": 1.0023,
+      "step": 1853
+    },
+    {
+      "epoch": 0.1429232192414431,
+      "grad_norm": 4.5213189125061035,
+      "learning_rate": 9.669644792579717e-06,
+      "loss": 1.2307,
+      "step": 1854
+    },
+    {
+      "epoch": 0.14300030835646008,
+      "grad_norm": 4.091536521911621,
+      "learning_rate": 9.669198378728162e-06,
+      "loss": 1.1163,
+      "step": 1855
+    },
+    {
+      "epoch": 0.14307739747147702,
+      "grad_norm": 4.354223728179932,
+      "learning_rate": 9.668751673776272e-06,
+      "loss": 1.0047,
+      "step": 1856
+    },
+    {
+      "epoch": 0.143154486586494,
+      "grad_norm": 3.7144696712493896,
+      "learning_rate": 9.668304677751894e-06,
+      "loss": 1.1268,
+      "step": 1857
+    },
+    {
+      "epoch": 0.14323157570151096,
+      "grad_norm": 4.206509590148926,
+      "learning_rate": 9.667857390682897e-06,
+      "loss": 1.0878,
+      "step": 1858
+    },
+    {
+      "epoch": 0.1433086648165279,
+      "grad_norm": 4.0310492515563965,
+      "learning_rate": 9.667409812597168e-06,
+      "loss": 1.135,
+      "step": 1859
+    },
+    {
+      "epoch": 0.14338575393154487,
+      "grad_norm": 4.384598731994629,
+      "learning_rate": 9.66696194352261e-06,
+      "loss": 1.0957,
+      "step": 1860
+    },
+    {
+      "epoch": 0.14346284304656182,
+      "grad_norm": 3.8433432579040527,
+      "learning_rate": 9.666513783487145e-06,
+      "loss": 0.9749,
+      "step": 1861
+    },
+    {
+      "epoch": 0.1435399321615788,
+      "grad_norm": 3.990506887435913,
+      "learning_rate": 9.666065332518714e-06,
+      "loss": 1.086,
+      "step": 1862
+    },
+    {
+      "epoch": 0.14361702127659576,
+      "grad_norm": 3.817073106765747,
+      "learning_rate": 9.665616590645278e-06,
+      "loss": 1.0804,
+      "step": 1863
+    },
+    {
+      "epoch": 0.1436941103916127,
+      "grad_norm": 3.8277182579040527,
+      "learning_rate": 9.665167557894808e-06,
+      "loss": 1.0075,
+      "step": 1864
+    },
+    {
+      "epoch": 0.14377119950662967,
+      "grad_norm": 3.6581485271453857,
+      "learning_rate": 9.664718234295303e-06,
+      "loss": 1.122,
+      "step": 1865
+    },
+    {
+      "epoch": 0.14384828862164661,
+      "grad_norm": 3.7937848567962646,
+      "learning_rate": 9.664268619874776e-06,
+      "loss": 1.0867,
+      "step": 1866
+    },
+    {
+      "epoch": 0.14392537773666358,
+      "grad_norm": 4.170339584350586,
+      "learning_rate": 9.663818714661259e-06,
+      "loss": 1.0888,
+      "step": 1867
+    },
+    {
+      "epoch": 0.14400246685168056,
+      "grad_norm": 3.6754279136657715,
+      "learning_rate": 9.6633685186828e-06,
+      "loss": 1.0507,
+      "step": 1868
+    },
+    {
+      "epoch": 0.1440795559666975,
+      "grad_norm": 3.820255994796753,
+      "learning_rate": 9.662918031967463e-06,
+      "loss": 1.01,
+      "step": 1869
+    },
+    {
+      "epoch": 0.14415664508171447,
+      "grad_norm": 4.000062942504883,
+      "learning_rate": 9.662467254543337e-06,
+      "loss": 1.1658,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1442337341967314,
+      "grad_norm": 3.684593439102173,
+      "learning_rate": 9.662016186438527e-06,
+      "loss": 1.0732,
+      "step": 1871
+    },
+    {
+      "epoch": 0.14431082331174838,
+      "grad_norm": 4.005534648895264,
+      "learning_rate": 9.661564827681152e-06,
+      "loss": 1.0009,
+      "step": 1872
+    },
+    {
+      "epoch": 0.14438791242676535,
+      "grad_norm": 3.8382081985473633,
+      "learning_rate": 9.661113178299353e-06,
+      "loss": 1.1442,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1444650015417823,
+      "grad_norm": 4.236837863922119,
+      "learning_rate": 9.660661238321289e-06,
+      "loss": 1.0885,
+      "step": 1874
+    },
+    {
+      "epoch": 0.14454209065679927,
+      "grad_norm": 3.7837772369384766,
+      "learning_rate": 9.660209007775133e-06,
+      "loss": 1.0377,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1446191797718162,
+      "grad_norm": 3.9629247188568115,
+      "learning_rate": 9.659756486689082e-06,
+      "loss": 1.0251,
+      "step": 1876
+    },
+    {
+      "epoch": 0.14469626888683318,
+      "grad_norm": 3.9419615268707275,
+      "learning_rate": 9.659303675091348e-06,
+      "loss": 1.181,
+      "step": 1877
+    },
+    {
+      "epoch": 0.14477335800185015,
+      "grad_norm": 3.833371162414551,
+      "learning_rate": 9.658850573010162e-06,
+      "loss": 0.9706,
+      "step": 1878
+    },
+    {
+      "epoch": 0.1448504471168671,
+      "grad_norm": 4.696406364440918,
+      "learning_rate": 9.658397180473768e-06,
+      "loss": 1.1338,
+      "step": 1879
+    },
+    {
+      "epoch": 0.14492753623188406,
+      "grad_norm": 3.9524292945861816,
+      "learning_rate": 9.65794349751044e-06,
+      "loss": 1.032,
+      "step": 1880
+    },
+    {
+      "epoch": 0.145004625346901,
+      "grad_norm": 4.391104221343994,
+      "learning_rate": 9.657489524148459e-06,
+      "loss": 1.1298,
+      "step": 1881
+    },
+    {
+      "epoch": 0.14508171446191798,
+      "grad_norm": 3.786226749420166,
+      "learning_rate": 9.657035260416126e-06,
+      "loss": 1.0417,
+      "step": 1882
+    },
+    {
+      "epoch": 0.14515880357693495,
+      "grad_norm": 3.7811176776885986,
+      "learning_rate": 9.656580706341763e-06,
+      "loss": 1.1276,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1452358926919519,
+      "grad_norm": 4.132617473602295,
+      "learning_rate": 9.656125861953711e-06,
+      "loss": 1.0482,
+      "step": 1884
+    },
+    {
+      "epoch": 0.14531298180696886,
+      "grad_norm": 4.651538848876953,
+      "learning_rate": 9.655670727280326e-06,
+      "loss": 1.2372,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1453900709219858,
+      "grad_norm": 3.871948719024658,
+      "learning_rate": 9.655215302349986e-06,
+      "loss": 1.0957,
+      "step": 1886
+    },
+    {
+      "epoch": 0.14546716003700277,
+      "grad_norm": 5.455825328826904,
+      "learning_rate": 9.65475958719108e-06,
+      "loss": 1.0278,
+      "step": 1887
+    },
+    {
+      "epoch": 0.14554424915201974,
+      "grad_norm": 3.909919261932373,
+      "learning_rate": 9.65430358183202e-06,
+      "loss": 1.1265,
+      "step": 1888
+    },
+    {
+      "epoch": 0.1456213382670367,
+      "grad_norm": 4.387859344482422,
+      "learning_rate": 9.653847286301238e-06,
+      "loss": 1.1409,
+      "step": 1889
+    },
+    {
+      "epoch": 0.14569842738205366,
+      "grad_norm": 4.335275650024414,
+      "learning_rate": 9.65339070062718e-06,
+      "loss": 1.0492,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1457755164970706,
+      "grad_norm": 3.7810895442962646,
+      "learning_rate": 9.652933824838315e-06,
+      "loss": 0.9817,
+      "step": 1891
+    },
+    {
+      "epoch": 0.14585260561208757,
+      "grad_norm": 4.186948299407959,
+      "learning_rate": 9.652476658963122e-06,
+      "loss": 1.1084,
+      "step": 1892
+    },
+    {
+      "epoch": 0.14592969472710454,
+      "grad_norm": 4.576048374176025,
+      "learning_rate": 9.652019203030105e-06,
+      "loss": 1.1037,
+      "step": 1893
+    },
+    {
+      "epoch": 0.14600678384212148,
+      "grad_norm": 3.583650827407837,
+      "learning_rate": 9.651561457067785e-06,
+      "loss": 1.045,
+      "step": 1894
+    },
+    {
+      "epoch": 0.14608387295713846,
+      "grad_norm": 3.9985146522521973,
+      "learning_rate": 9.6511034211047e-06,
+      "loss": 1.0088,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1461609620721554,
+      "grad_norm": 4.34122371673584,
+      "learning_rate": 9.650645095169403e-06,
+      "loss": 1.0772,
+      "step": 1896
+    },
+    {
+      "epoch": 0.14623805118717237,
+      "grad_norm": 4.4444122314453125,
+      "learning_rate": 9.650186479290472e-06,
+      "loss": 1.1531,
+      "step": 1897
+    },
+    {
+      "epoch": 0.14631514030218934,
+      "grad_norm": 3.7410149574279785,
+      "learning_rate": 9.649727573496499e-06,
+      "loss": 0.9934,
+      "step": 1898
+    },
+    {
+      "epoch": 0.14639222941720628,
+      "grad_norm": 3.8100745677948,
+      "learning_rate": 9.649268377816092e-06,
+      "loss": 1.0822,
+      "step": 1899
+    },
+    {
+      "epoch": 0.14646931853222325,
+      "grad_norm": 4.091446876525879,
+      "learning_rate": 9.64880889227788e-06,
+      "loss": 1.1054,
+      "step": 1900
+    },
+    {
+      "epoch": 0.14654640764724022,
+      "grad_norm": 4.245041847229004,
+      "learning_rate": 9.64834911691051e-06,
+      "loss": 1.1384,
+      "step": 1901
+    },
+    {
+      "epoch": 0.14662349676225717,
+      "grad_norm": 4.033145427703857,
+      "learning_rate": 9.647889051742649e-06,
+      "loss": 1.0701,
+      "step": 1902
+    },
+    {
+      "epoch": 0.14670058587727414,
+      "grad_norm": 4.123029708862305,
+      "learning_rate": 9.647428696802979e-06,
+      "loss": 1.0922,
+      "step": 1903
+    },
+    {
+      "epoch": 0.14677767499229108,
+      "grad_norm": 4.149605751037598,
+      "learning_rate": 9.646968052120196e-06,
+      "loss": 1.074,
+      "step": 1904
+    },
+    {
+      "epoch": 0.14685476410730805,
+      "grad_norm": 3.9435675144195557,
+      "learning_rate": 9.646507117723023e-06,
+      "loss": 1.0313,
+      "step": 1905
+    },
+    {
+      "epoch": 0.14693185322232502,
+      "grad_norm": 4.518271446228027,
+      "learning_rate": 9.646045893640197e-06,
+      "loss": 1.1398,
+      "step": 1906
+    },
+    {
+      "epoch": 0.14700894233734196,
+      "grad_norm": 3.978560447692871,
+      "learning_rate": 9.645584379900473e-06,
+      "loss": 1.1128,
+      "step": 1907
+    },
+    {
+      "epoch": 0.14708603145235893,
+      "grad_norm": 4.344027042388916,
+      "learning_rate": 9.64512257653262e-06,
+      "loss": 1.0262,
+      "step": 1908
+    },
+    {
+      "epoch": 0.14716312056737588,
+      "grad_norm": 4.042947769165039,
+      "learning_rate": 9.644660483565434e-06,
+      "loss": 1.1174,
+      "step": 1909
+    },
+    {
+      "epoch": 0.14724020968239285,
+      "grad_norm": 4.025622844696045,
+      "learning_rate": 9.644198101027721e-06,
+      "loss": 1.0611,
+      "step": 1910
+    },
+    {
+      "epoch": 0.14731729879740982,
+      "grad_norm": 4.026158809661865,
+      "learning_rate": 9.64373542894831e-06,
+      "loss": 1.083,
+      "step": 1911
+    },
+    {
+      "epoch": 0.14739438791242676,
+      "grad_norm": 4.082542419433594,
+      "learning_rate": 9.643272467356048e-06,
+      "loss": 1.132,
+      "step": 1912
+    },
+    {
+      "epoch": 0.14747147702744373,
+      "grad_norm": 3.592789888381958,
+      "learning_rate": 9.642809216279793e-06,
+      "loss": 1.019,
+      "step": 1913
+    },
+    {
+      "epoch": 0.14754856614246067,
+      "grad_norm": 3.6364972591400146,
+      "learning_rate": 9.64234567574843e-06,
+      "loss": 1.0443,
+      "step": 1914
+    },
+    {
+      "epoch": 0.14762565525747764,
+      "grad_norm": 4.323209285736084,
+      "learning_rate": 9.641881845790858e-06,
+      "loss": 1.0813,
+      "step": 1915
+    },
+    {
+      "epoch": 0.14770274437249462,
+      "grad_norm": 3.8409078121185303,
+      "learning_rate": 9.641417726435991e-06,
+      "loss": 1.0383,
+      "step": 1916
+    },
+    {
+      "epoch": 0.14777983348751156,
+      "grad_norm": 3.611025810241699,
+      "learning_rate": 9.64095331771277e-06,
+      "loss": 1.0272,
+      "step": 1917
+    },
+    {
+      "epoch": 0.14785692260252853,
+      "grad_norm": 4.067416667938232,
+      "learning_rate": 9.640488619650145e-06,
+      "loss": 1.1394,
+      "step": 1918
+    },
+    {
+      "epoch": 0.14793401171754547,
+      "grad_norm": 3.7739274501800537,
+      "learning_rate": 9.640023632277088e-06,
+      "loss": 1.0281,
+      "step": 1919
+    },
+    {
+      "epoch": 0.14801110083256244,
+      "grad_norm": 4.0226263999938965,
+      "learning_rate": 9.639558355622589e-06,
+      "loss": 1.0501,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1480881899475794,
+      "grad_norm": 4.34377908706665,
+      "learning_rate": 9.639092789715656e-06,
+      "loss": 1.0817,
+      "step": 1921
+    },
+    {
+      "epoch": 0.14816527906259636,
+      "grad_norm": 3.9317402839660645,
+      "learning_rate": 9.638626934585314e-06,
+      "loss": 1.0944,
+      "step": 1922
+    },
+    {
+      "epoch": 0.14824236817761333,
+      "grad_norm": 3.948791265487671,
+      "learning_rate": 9.638160790260606e-06,
+      "loss": 1.0707,
+      "step": 1923
+    },
+    {
+      "epoch": 0.14831945729263027,
+      "grad_norm": 4.429360389709473,
+      "learning_rate": 9.637694356770595e-06,
+      "loss": 1.1624,
+      "step": 1924
+    },
+    {
+      "epoch": 0.14839654640764724,
+      "grad_norm": 4.858312129974365,
+      "learning_rate": 9.637227634144359e-06,
+      "loss": 0.995,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1484736355226642,
+      "grad_norm": 3.9074039459228516,
+      "learning_rate": 9.636760622410997e-06,
+      "loss": 1.1135,
+      "step": 1926
+    },
+    {
+      "epoch": 0.14855072463768115,
+      "grad_norm": 3.9309797286987305,
+      "learning_rate": 9.636293321599625e-06,
+      "loss": 1.0696,
+      "step": 1927
+    },
+    {
+      "epoch": 0.14862781375269812,
+      "grad_norm": 3.8177499771118164,
+      "learning_rate": 9.635825731739376e-06,
+      "loss": 1.0943,
+      "step": 1928
+    },
+    {
+      "epoch": 0.14870490286771507,
+      "grad_norm": 4.232987880706787,
+      "learning_rate": 9.635357852859404e-06,
+      "loss": 1.0883,
+      "step": 1929
+    },
+    {
+      "epoch": 0.14878199198273204,
+      "grad_norm": 4.08406400680542,
+      "learning_rate": 9.634889684988875e-06,
+      "loss": 1.1446,
+      "step": 1930
+    },
+    {
+      "epoch": 0.148859081097749,
+      "grad_norm": 4.512781620025635,
+      "learning_rate": 9.63442122815698e-06,
+      "loss": 1.1857,
+      "step": 1931
+    },
+    {
+      "epoch": 0.14893617021276595,
+      "grad_norm": 4.197699069976807,
+      "learning_rate": 9.633952482392924e-06,
+      "loss": 1.0965,
+      "step": 1932
+    },
+    {
+      "epoch": 0.14901325932778292,
+      "grad_norm": 3.925628185272217,
+      "learning_rate": 9.63348344772593e-06,
+      "loss": 1.1148,
+      "step": 1933
+    },
+    {
+      "epoch": 0.14909034844279986,
+      "grad_norm": 4.278095245361328,
+      "learning_rate": 9.63301412418524e-06,
+      "loss": 1.0734,
+      "step": 1934
+    },
+    {
+      "epoch": 0.14916743755781683,
+      "grad_norm": 4.0269269943237305,
+      "learning_rate": 9.632544511800114e-06,
+      "loss": 1.1376,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1492445266728338,
+      "grad_norm": 3.731536626815796,
+      "learning_rate": 9.63207461059983e-06,
+      "loss": 1.0154,
+      "step": 1936
+    },
+    {
+      "epoch": 0.14932161578785075,
+      "grad_norm": 3.8483800888061523,
+      "learning_rate": 9.631604420613685e-06,
+      "loss": 1.0973,
+      "step": 1937
+    },
+    {
+      "epoch": 0.14939870490286772,
+      "grad_norm": 3.9684271812438965,
+      "learning_rate": 9.631133941870993e-06,
+      "loss": 1.0039,
+      "step": 1938
+    },
+    {
+      "epoch": 0.14947579401788466,
+      "grad_norm": 4.196876049041748,
+      "learning_rate": 9.630663174401085e-06,
+      "loss": 1.2552,
+      "step": 1939
+    },
+    {
+      "epoch": 0.14955288313290163,
+      "grad_norm": 4.365589141845703,
+      "learning_rate": 9.630192118233309e-06,
+      "loss": 1.0885,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1496299722479186,
+      "grad_norm": 3.9205033779144287,
+      "learning_rate": 9.629720773397036e-06,
+      "loss": 1.0257,
+      "step": 1941
+    },
+    {
+      "epoch": 0.14970706136293555,
+      "grad_norm": 3.8212361335754395,
+      "learning_rate": 9.62924913992165e-06,
+      "loss": 1.0147,
+      "step": 1942
+    },
+    {
+      "epoch": 0.14978415047795252,
+      "grad_norm": 4.525760650634766,
+      "learning_rate": 9.628777217836558e-06,
+      "loss": 1.1392,
+      "step": 1943
+    },
+    {
+      "epoch": 0.1498612395929695,
+      "grad_norm": 4.177234172821045,
+      "learning_rate": 9.628305007171177e-06,
+      "loss": 1.2419,
+      "step": 1944
+    },
+    {
+      "epoch": 0.14993832870798643,
+      "grad_norm": 4.351520538330078,
+      "learning_rate": 9.627832507954949e-06,
+      "loss": 1.1474,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1500154178230034,
+      "grad_norm": 4.03515625,
+      "learning_rate": 9.627359720217334e-06,
+      "loss": 1.0986,
+      "step": 1946
+    },
+    {
+      "epoch": 0.15009250693802034,
+      "grad_norm": 3.882413387298584,
+      "learning_rate": 9.626886643987806e-06,
+      "loss": 1.1141,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1501695960530373,
+      "grad_norm": 4.117105007171631,
+      "learning_rate": 9.626413279295859e-06,
+      "loss": 1.1688,
+      "step": 1948
+    },
+    {
+      "epoch": 0.15024668516805428,
+      "grad_norm": 3.7657737731933594,
+      "learning_rate": 9.625939626171004e-06,
+      "loss": 1.0316,
+      "step": 1949
+    },
+    {
+      "epoch": 0.15032377428307123,
+      "grad_norm": 4.3290019035339355,
+      "learning_rate": 9.625465684642773e-06,
+      "loss": 1.1699,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1504008633980882,
+      "grad_norm": 4.567865371704102,
+      "learning_rate": 9.624991454740709e-06,
+      "loss": 1.1349,
+      "step": 1951
+    },
+    {
+      "epoch": 0.15047795251310514,
+      "grad_norm": 3.6431381702423096,
+      "learning_rate": 9.624516936494385e-06,
+      "loss": 0.9967,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1505550416281221,
+      "grad_norm": 4.145571231842041,
+      "learning_rate": 9.62404212993338e-06,
+      "loss": 1.1052,
+      "step": 1953
+    },
+    {
+      "epoch": 0.15063213074313908,
+      "grad_norm": 3.933396339416504,
+      "learning_rate": 9.623567035087295e-06,
+      "loss": 1.0349,
+      "step": 1954
+    },
+    {
+      "epoch": 0.15070921985815602,
+      "grad_norm": 4.341703414916992,
+      "learning_rate": 9.623091651985754e-06,
+      "loss": 1.164,
+      "step": 1955
+    },
+    {
+      "epoch": 0.150786308973173,
+      "grad_norm": 3.696288585662842,
+      "learning_rate": 9.622615980658391e-06,
+      "loss": 1.0052,
+      "step": 1956
+    },
+    {
+      "epoch": 0.15086339808818994,
+      "grad_norm": 4.249678611755371,
+      "learning_rate": 9.622140021134863e-06,
+      "loss": 1.1709,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1509404872032069,
+      "grad_norm": 4.261125087738037,
+      "learning_rate": 9.621663773444843e-06,
+      "loss": 1.0577,
+      "step": 1958
+    },
+    {
+      "epoch": 0.15101757631822388,
+      "grad_norm": 3.759453296661377,
+      "learning_rate": 9.621187237618024e-06,
+      "loss": 1.0644,
+      "step": 1959
+    },
+    {
+      "epoch": 0.15109466543324082,
+      "grad_norm": 4.062593460083008,
+      "learning_rate": 9.620710413684112e-06,
+      "loss": 1.1486,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1511717545482578,
+      "grad_norm": 4.127292156219482,
+      "learning_rate": 9.62023330167284e-06,
+      "loss": 1.0529,
+      "step": 1961
+    },
+    {
+      "epoch": 0.15124884366327473,
+      "grad_norm": 4.187131881713867,
+      "learning_rate": 9.619755901613947e-06,
+      "loss": 1.0464,
+      "step": 1962
+    },
+    {
+      "epoch": 0.1513259327782917,
+      "grad_norm": 3.996009111404419,
+      "learning_rate": 9.619278213537202e-06,
+      "loss": 1.1509,
+      "step": 1963
+    },
+    {
+      "epoch": 0.15140302189330868,
+      "grad_norm": 4.108388423919678,
+      "learning_rate": 9.618800237472385e-06,
+      "loss": 1.0326,
+      "step": 1964
+    },
+    {
+      "epoch": 0.15148011100832562,
+      "grad_norm": 3.868997573852539,
+      "learning_rate": 9.618321973449294e-06,
+      "loss": 1.0271,
+      "step": 1965
+    },
+    {
+      "epoch": 0.1515572001233426,
+      "grad_norm": 4.561322212219238,
+      "learning_rate": 9.617843421497746e-06,
+      "loss": 1.0995,
+      "step": 1966
+    },
+    {
+      "epoch": 0.15163428923835953,
+      "grad_norm": 3.980226755142212,
+      "learning_rate": 9.617364581647578e-06,
+      "loss": 0.983,
+      "step": 1967
+    },
+    {
+      "epoch": 0.1517113783533765,
+      "grad_norm": 4.084739685058594,
+      "learning_rate": 9.616885453928641e-06,
+      "loss": 1.257,
+      "step": 1968
+    },
+    {
+      "epoch": 0.15178846746839347,
+      "grad_norm": 4.914328575134277,
+      "learning_rate": 9.616406038370809e-06,
+      "loss": 1.1405,
+      "step": 1969
+    },
+    {
+      "epoch": 0.15186555658341042,
+      "grad_norm": 3.754298210144043,
+      "learning_rate": 9.615926335003968e-06,
+      "loss": 1.0765,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1519426456984274,
+      "grad_norm": 3.9111487865448,
+      "learning_rate": 9.615446343858028e-06,
+      "loss": 1.0048,
+      "step": 1971
+    },
+    {
+      "epoch": 0.15201973481344433,
+      "grad_norm": 4.179695129394531,
+      "learning_rate": 9.614966064962911e-06,
+      "loss": 1.0866,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1520968239284613,
+      "grad_norm": 3.932072877883911,
+      "learning_rate": 9.614485498348563e-06,
+      "loss": 1.0682,
+      "step": 1973
+    },
+    {
+      "epoch": 0.15217391304347827,
+      "grad_norm": 4.217230796813965,
+      "learning_rate": 9.614004644044943e-06,
+      "loss": 1.1172,
+      "step": 1974
+    },
+    {
+      "epoch": 0.1522510021584952,
+      "grad_norm": 4.254328727722168,
+      "learning_rate": 9.613523502082029e-06,
+      "loss": 1.1907,
+      "step": 1975
+    },
+    {
+      "epoch": 0.15232809127351218,
+      "grad_norm": 4.149200916290283,
+      "learning_rate": 9.613042072489819e-06,
+      "loss": 1.1092,
+      "step": 1976
+    },
+    {
+      "epoch": 0.15240518038852913,
+      "grad_norm": 4.106222152709961,
+      "learning_rate": 9.612560355298328e-06,
+      "loss": 1.103,
+      "step": 1977
+    },
+    {
+      "epoch": 0.1524822695035461,
+      "grad_norm": 3.983444929122925,
+      "learning_rate": 9.612078350537586e-06,
+      "loss": 1.0695,
+      "step": 1978
+    },
+    {
+      "epoch": 0.15255935861856307,
+      "grad_norm": 4.1216959953308105,
+      "learning_rate": 9.611596058237647e-06,
+      "loss": 1.1196,
+      "step": 1979
+    },
+    {
+      "epoch": 0.15263644773358,
+      "grad_norm": 3.9716129302978516,
+      "learning_rate": 9.611113478428577e-06,
+      "loss": 1.2409,
+      "step": 1980
+    },
+    {
+      "epoch": 0.15271353684859698,
+      "grad_norm": 3.9019081592559814,
+      "learning_rate": 9.610630611140464e-06,
+      "loss": 0.9612,
+      "step": 1981
+    },
+    {
+      "epoch": 0.15279062596361392,
+      "grad_norm": 4.146617889404297,
+      "learning_rate": 9.610147456403412e-06,
+      "loss": 1.0516,
+      "step": 1982
+    },
+    {
+      "epoch": 0.1528677150786309,
+      "grad_norm": 3.6473166942596436,
+      "learning_rate": 9.609664014247542e-06,
+      "loss": 1.0253,
+      "step": 1983
+    },
+    {
+      "epoch": 0.15294480419364787,
+      "grad_norm": 4.137308120727539,
+      "learning_rate": 9.609180284702994e-06,
+      "loss": 1.0064,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1530218933086648,
+      "grad_norm": 3.682206869125366,
+      "learning_rate": 9.608696267799928e-06,
+      "loss": 1.1105,
+      "step": 1985
+    },
+    {
+      "epoch": 0.15309898242368178,
+      "grad_norm": 3.6087026596069336,
+      "learning_rate": 9.608211963568518e-06,
+      "loss": 0.973,
+      "step": 1986
+    },
+    {
+      "epoch": 0.15317607153869875,
+      "grad_norm": 4.12633752822876,
+      "learning_rate": 9.60772737203896e-06,
+      "loss": 1.1194,
+      "step": 1987
+    },
+    {
+      "epoch": 0.1532531606537157,
+      "grad_norm": 3.913084030151367,
+      "learning_rate": 9.607242493241463e-06,
+      "loss": 1.1383,
+      "step": 1988
+    },
+    {
+      "epoch": 0.15333024976873266,
+      "grad_norm": 4.2011003494262695,
+      "learning_rate": 9.606757327206258e-06,
+      "loss": 1.1338,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1534073388837496,
+      "grad_norm": 4.860283851623535,
+      "learning_rate": 9.606271873963591e-06,
+      "loss": 0.9991,
+      "step": 1990
+    },
+    {
+      "epoch": 0.15348442799876658,
+      "grad_norm": 3.819664716720581,
+      "learning_rate": 9.605786133543732e-06,
+      "loss": 1.1397,
+      "step": 1991
+    },
+    {
+      "epoch": 0.15356151711378355,
+      "grad_norm": 4.044528961181641,
+      "learning_rate": 9.60530010597696e-06,
+      "loss": 1.0723,
+      "step": 1992
+    },
+    {
+      "epoch": 0.1536386062288005,
+      "grad_norm": 3.8798742294311523,
+      "learning_rate": 9.604813791293579e-06,
+      "loss": 1.0981,
+      "step": 1993
+    },
+    {
+      "epoch": 0.15371569534381746,
+      "grad_norm": 4.1378703117370605,
+      "learning_rate": 9.604327189523906e-06,
+      "loss": 1.1575,
+      "step": 1994
+    },
+    {
+      "epoch": 0.1537927844588344,
+      "grad_norm": 5.460372447967529,
+      "learning_rate": 9.60384030069828e-06,
+      "loss": 1.1,
+      "step": 1995
+    },
+    {
+      "epoch": 0.15386987357385137,
+      "grad_norm": 4.201817035675049,
+      "learning_rate": 9.603353124847054e-06,
+      "loss": 1.0258,
+      "step": 1996
+    },
+    {
+      "epoch": 0.15394696268886834,
+      "grad_norm": 3.7332353591918945,
+      "learning_rate": 9.602865662000604e-06,
+      "loss": 1.1201,
+      "step": 1997
+    },
+    {
+      "epoch": 0.1540240518038853,
+      "grad_norm": 3.7747762203216553,
+      "learning_rate": 9.602377912189319e-06,
+      "loss": 1.0083,
+      "step": 1998
+    },
+    {
+      "epoch": 0.15410114091890226,
+      "grad_norm": 4.299161434173584,
+      "learning_rate": 9.601889875443605e-06,
+      "loss": 0.969,
+      "step": 1999
+    },
+    {
+      "epoch": 0.1541782300339192,
+      "grad_norm": 4.0587263107299805,
+      "learning_rate": 9.601401551793891e-06,
+      "loss": 1.0928,
+      "step": 2000
+    },
+    {
+      "epoch": 0.15425531914893617,
+      "grad_norm": 3.8370776176452637,
+      "learning_rate": 9.600912941270624e-06,
+      "loss": 1.1446,
+      "step": 2001
+    },
+    {
+      "epoch": 0.15433240826395314,
+      "grad_norm": 3.871699333190918,
+      "learning_rate": 9.600424043904263e-06,
+      "loss": 1.1099,
+      "step": 2002
+    },
+    {
+      "epoch": 0.15440949737897008,
+      "grad_norm": 4.0479230880737305,
+      "learning_rate": 9.599934859725288e-06,
+      "loss": 1.1099,
+      "step": 2003
+    },
+    {
+      "epoch": 0.15448658649398705,
+      "grad_norm": 4.1467485427856445,
+      "learning_rate": 9.599445388764199e-06,
+      "loss": 1.0611,
+      "step": 2004
+    },
+    {
+      "epoch": 0.154563675609004,
+      "grad_norm": 3.658740997314453,
+      "learning_rate": 9.598955631051512e-06,
+      "loss": 1.0018,
+      "step": 2005
+    },
+    {
+      "epoch": 0.15464076472402097,
+      "grad_norm": 4.59723424911499,
+      "learning_rate": 9.598465586617757e-06,
+      "loss": 1.1884,
+      "step": 2006
+    },
+    {
+      "epoch": 0.15471785383903794,
+      "grad_norm": 3.7733352184295654,
+      "learning_rate": 9.597975255493492e-06,
+      "loss": 1.0512,
+      "step": 2007
+    },
+    {
+      "epoch": 0.15479494295405488,
+      "grad_norm": 4.303390026092529,
+      "learning_rate": 9.597484637709282e-06,
+      "loss": 1.0479,
+      "step": 2008
+    },
+    {
+      "epoch": 0.15487203206907185,
+      "grad_norm": 3.650953769683838,
+      "learning_rate": 9.596993733295717e-06,
+      "loss": 0.9842,
+      "step": 2009
+    },
+    {
+      "epoch": 0.1549491211840888,
+      "grad_norm": 3.5866963863372803,
+      "learning_rate": 9.596502542283399e-06,
+      "loss": 1.1231,
+      "step": 2010
+    },
+    {
+      "epoch": 0.15502621029910577,
+      "grad_norm": 3.8881676197052,
+      "learning_rate": 9.596011064702954e-06,
+      "loss": 1.1773,
+      "step": 2011
+    },
+    {
+      "epoch": 0.15510329941412274,
+      "grad_norm": 3.849320650100708,
+      "learning_rate": 9.595519300585024e-06,
+      "loss": 1.0562,
+      "step": 2012
+    },
+    {
+      "epoch": 0.15518038852913968,
+      "grad_norm": 3.8360087871551514,
+      "learning_rate": 9.595027249960264e-06,
+      "loss": 1.17,
+      "step": 2013
+    },
+    {
+      "epoch": 0.15525747764415665,
+      "grad_norm": 3.9184608459472656,
+      "learning_rate": 9.594534912859356e-06,
+      "loss": 1.0578,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1553345667591736,
+      "grad_norm": 4.464005470275879,
+      "learning_rate": 9.59404228931299e-06,
+      "loss": 1.0775,
+      "step": 2015
+    },
+    {
+      "epoch": 0.15541165587419056,
+      "grad_norm": 3.8075225353240967,
+      "learning_rate": 9.59354937935188e-06,
+      "loss": 0.9951,
+      "step": 2016
+    },
+    {
+      "epoch": 0.15548874498920753,
+      "grad_norm": 3.766676902770996,
+      "learning_rate": 9.59305618300676e-06,
+      "loss": 1.1488,
+      "step": 2017
+    },
+    {
+      "epoch": 0.15556583410422448,
+      "grad_norm": 3.7311699390411377,
+      "learning_rate": 9.592562700308372e-06,
+      "loss": 1.1167,
+      "step": 2018
+    },
+    {
+      "epoch": 0.15564292321924145,
+      "grad_norm": 4.18145751953125,
+      "learning_rate": 9.592068931287486e-06,
+      "loss": 1.1209,
+      "step": 2019
+    },
+    {
+      "epoch": 0.1557200123342584,
+      "grad_norm": 3.6943624019622803,
+      "learning_rate": 9.591574875974884e-06,
+      "loss": 1.0483,
+      "step": 2020
+    },
+    {
+      "epoch": 0.15579710144927536,
+      "grad_norm": 3.553483009338379,
+      "learning_rate": 9.591080534401371e-06,
+      "loss": 1.0577,
+      "step": 2021
+    },
+    {
+      "epoch": 0.15587419056429233,
+      "grad_norm": 3.7024497985839844,
+      "learning_rate": 9.590585906597764e-06,
+      "loss": 1.0646,
+      "step": 2022
+    },
+    {
+      "epoch": 0.15595127967930927,
+      "grad_norm": 4.6477837562561035,
+      "learning_rate": 9.590090992594901e-06,
+      "loss": 1.0384,
+      "step": 2023
+    },
+    {
+      "epoch": 0.15602836879432624,
+      "grad_norm": 3.604308843612671,
+      "learning_rate": 9.589595792423636e-06,
+      "loss": 0.9234,
+      "step": 2024
+    },
+    {
+      "epoch": 0.1561054579093432,
+      "grad_norm": 4.46380090713501,
+      "learning_rate": 9.589100306114842e-06,
+      "loss": 1.1194,
+      "step": 2025
+    },
+    {
+      "epoch": 0.15618254702436016,
+      "grad_norm": 4.057259559631348,
+      "learning_rate": 9.588604533699415e-06,
+      "loss": 1.1086,
+      "step": 2026
+    },
+    {
+      "epoch": 0.15625963613937713,
+      "grad_norm": 3.9798672199249268,
+      "learning_rate": 9.58810847520826e-06,
+      "loss": 1.0903,
+      "step": 2027
+    },
+    {
+      "epoch": 0.15633672525439407,
+      "grad_norm": 4.816097259521484,
+      "learning_rate": 9.587612130672302e-06,
+      "loss": 1.107,
+      "step": 2028
+    },
+    {
+      "epoch": 0.15641381436941104,
+      "grad_norm": 4.069070339202881,
+      "learning_rate": 9.587115500122489e-06,
+      "loss": 1.0659,
+      "step": 2029
+    },
+    {
+      "epoch": 0.156490903484428,
+      "grad_norm": 4.3908796310424805,
+      "learning_rate": 9.58661858358978e-06,
+      "loss": 0.9479,
+      "step": 2030
+    },
+    {
+      "epoch": 0.15656799259944496,
+      "grad_norm": 3.6012563705444336,
+      "learning_rate": 9.586121381105158e-06,
+      "loss": 1.0304,
+      "step": 2031
+    },
+    {
+      "epoch": 0.15664508171446193,
+      "grad_norm": 4.042422771453857,
+      "learning_rate": 9.58562389269962e-06,
+      "loss": 1.08,
+      "step": 2032
+    },
+    {
+      "epoch": 0.15672217082947887,
+      "grad_norm": 4.079514980316162,
+      "learning_rate": 9.585126118404183e-06,
+      "loss": 1.1668,
+      "step": 2033
+    },
+    {
+      "epoch": 0.15679925994449584,
+      "grad_norm": 3.817542314529419,
+      "learning_rate": 9.584628058249878e-06,
+      "loss": 1.0957,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1568763490595128,
+      "grad_norm": 4.3651604652404785,
+      "learning_rate": 9.584129712267759e-06,
+      "loss": 1.0863,
+      "step": 2035
+    },
+    {
+      "epoch": 0.15695343817452975,
+      "grad_norm": 4.226108074188232,
+      "learning_rate": 9.583631080488893e-06,
+      "loss": 1.0109,
+      "step": 2036
+    },
+    {
+      "epoch": 0.15703052728954672,
+      "grad_norm": 3.7006771564483643,
+      "learning_rate": 9.58313216294437e-06,
+      "loss": 0.9356,
+      "step": 2037
+    },
+    {
+      "epoch": 0.15710761640456367,
+      "grad_norm": 4.0176100730896,
+      "learning_rate": 9.582632959665293e-06,
+      "loss": 1.0112,
+      "step": 2038
+    },
+    {
+      "epoch": 0.15718470551958064,
+      "grad_norm": 4.495166301727295,
+      "learning_rate": 9.582133470682785e-06,
+      "loss": 1.149,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1572617946345976,
+      "grad_norm": 4.30600643157959,
+      "learning_rate": 9.581633696027986e-06,
+      "loss": 1.0351,
+      "step": 2040
+    },
+    {
+      "epoch": 0.15733888374961455,
+      "grad_norm": 4.271859645843506,
+      "learning_rate": 9.581133635732053e-06,
+      "loss": 1.2107,
+      "step": 2041
+    },
+    {
+      "epoch": 0.15741597286463152,
+      "grad_norm": 4.35610818862915,
+      "learning_rate": 9.580633289826166e-06,
+      "loss": 1.2124,
+      "step": 2042
+    },
+    {
+      "epoch": 0.15749306197964846,
+      "grad_norm": 3.801640272140503,
+      "learning_rate": 9.580132658341519e-06,
+      "loss": 1.0812,
+      "step": 2043
+    },
+    {
+      "epoch": 0.15757015109466543,
+      "grad_norm": 3.8817687034606934,
+      "learning_rate": 9.579631741309319e-06,
+      "loss": 1.1933,
+      "step": 2044
+    },
+    {
+      "epoch": 0.1576472402096824,
+      "grad_norm": 3.9286177158355713,
+      "learning_rate": 9.5791305387608e-06,
+      "loss": 1.0202,
+      "step": 2045
+    },
+    {
+      "epoch": 0.15772432932469935,
+      "grad_norm": 4.174068450927734,
+      "learning_rate": 9.578629050727208e-06,
+      "loss": 1.1846,
+      "step": 2046
+    },
+    {
+      "epoch": 0.15780141843971632,
+      "grad_norm": 4.3177409172058105,
+      "learning_rate": 9.578127277239807e-06,
+      "loss": 1.0551,
+      "step": 2047
+    },
+    {
+      "epoch": 0.15787850755473326,
+      "grad_norm": 4.210177898406982,
+      "learning_rate": 9.577625218329882e-06,
+      "loss": 1.0957,
+      "step": 2048
+    },
+    {
+      "epoch": 0.15795559666975023,
+      "grad_norm": 4.341221332550049,
+      "learning_rate": 9.577122874028733e-06,
+      "loss": 1.133,
+      "step": 2049
+    },
+    {
+      "epoch": 0.1580326857847672,
+      "grad_norm": 4.481259822845459,
+      "learning_rate": 9.576620244367676e-06,
+      "loss": 1.1467,
+      "step": 2050
+    },
+    {
+      "epoch": 0.15810977489978414,
+      "grad_norm": 4.296179294586182,
+      "learning_rate": 9.576117329378051e-06,
+      "loss": 1.0854,
+      "step": 2051
+    },
+    {
+      "epoch": 0.15818686401480112,
+      "grad_norm": 4.161882400512695,
+      "learning_rate": 9.575614129091211e-06,
+      "loss": 1.099,
+      "step": 2052
+    },
+    {
+      "epoch": 0.15826395312981806,
+      "grad_norm": 4.072216033935547,
+      "learning_rate": 9.575110643538528e-06,
+      "loss": 1.1539,
+      "step": 2053
+    },
+    {
+      "epoch": 0.15834104224483503,
+      "grad_norm": 3.7138614654541016,
+      "learning_rate": 9.574606872751391e-06,
+      "loss": 1.1287,
+      "step": 2054
+    },
+    {
+      "epoch": 0.158418131359852,
+      "grad_norm": 4.442841529846191,
+      "learning_rate": 9.574102816761209e-06,
+      "loss": 1.12,
+      "step": 2055
+    },
+    {
+      "epoch": 0.15849522047486894,
+      "grad_norm": 4.6196393966674805,
+      "learning_rate": 9.573598475599405e-06,
+      "loss": 1.0678,
+      "step": 2056
+    },
+    {
+      "epoch": 0.1585723095898859,
+      "grad_norm": 3.877990961074829,
+      "learning_rate": 9.573093849297423e-06,
+      "loss": 0.9528,
+      "step": 2057
+    },
+    {
+      "epoch": 0.15864939870490286,
+      "grad_norm": 4.252166748046875,
+      "learning_rate": 9.572588937886727e-06,
+      "loss": 1.0336,
+      "step": 2058
+    },
+    {
+      "epoch": 0.15872648781991983,
+      "grad_norm": 3.8500077724456787,
+      "learning_rate": 9.57208374139879e-06,
+      "loss": 1.1263,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1588035769349368,
+      "grad_norm": 3.8318703174591064,
+      "learning_rate": 9.571578259865112e-06,
+      "loss": 1.0754,
+      "step": 2060
+    },
+    {
+      "epoch": 0.15888066604995374,
+      "grad_norm": 4.030394554138184,
+      "learning_rate": 9.571072493317207e-06,
+      "loss": 1.0642,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1589577551649707,
+      "grad_norm": 4.103557109832764,
+      "learning_rate": 9.570566441786605e-06,
+      "loss": 1.0528,
+      "step": 2062
+    },
+    {
+      "epoch": 0.15903484427998765,
+      "grad_norm": 3.8708770275115967,
+      "learning_rate": 9.570060105304856e-06,
+      "loss": 1.0407,
+      "step": 2063
+    },
+    {
+      "epoch": 0.15911193339500462,
+      "grad_norm": 4.086610794067383,
+      "learning_rate": 9.569553483903531e-06,
+      "loss": 1.1052,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1591890225100216,
+      "grad_norm": 3.6814303398132324,
+      "learning_rate": 9.569046577614212e-06,
+      "loss": 1.0038,
+      "step": 2065
+    },
+    {
+      "epoch": 0.15926611162503854,
+      "grad_norm": 4.457508087158203,
+      "learning_rate": 9.568539386468501e-06,
+      "loss": 1.1407,
+      "step": 2066
+    },
+    {
+      "epoch": 0.1593432007400555,
+      "grad_norm": 3.961879253387451,
+      "learning_rate": 9.568031910498021e-06,
+      "loss": 1.18,
+      "step": 2067
+    },
+    {
+      "epoch": 0.15942028985507245,
+      "grad_norm": 4.2065253257751465,
+      "learning_rate": 9.56752414973441e-06,
+      "loss": 0.9892,
+      "step": 2068
+    },
+    {
+      "epoch": 0.15949737897008942,
+      "grad_norm": 3.6328790187835693,
+      "learning_rate": 9.567016104209326e-06,
+      "loss": 1.0919,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1595744680851064,
+      "grad_norm": 3.8659377098083496,
+      "learning_rate": 9.56650777395444e-06,
+      "loss": 1.0576,
+      "step": 2070
+    },
+    {
+      "epoch": 0.15965155720012333,
+      "grad_norm": 4.030930995941162,
+      "learning_rate": 9.565999159001442e-06,
+      "loss": 0.9824,
+      "step": 2071
+    },
+    {
+      "epoch": 0.1597286463151403,
+      "grad_norm": 4.241566181182861,
+      "learning_rate": 9.565490259382047e-06,
+      "loss": 1.0389,
+      "step": 2072
+    },
+    {
+      "epoch": 0.15980573543015728,
+      "grad_norm": 4.0185089111328125,
+      "learning_rate": 9.564981075127979e-06,
+      "loss": 1.0236,
+      "step": 2073
+    },
+    {
+      "epoch": 0.15988282454517422,
+      "grad_norm": 4.9292826652526855,
+      "learning_rate": 9.564471606270985e-06,
+      "loss": 1.0604,
+      "step": 2074
+    },
+    {
+      "epoch": 0.1599599136601912,
+      "grad_norm": 4.441895008087158,
+      "learning_rate": 9.563961852842824e-06,
+      "loss": 1.0312,
+      "step": 2075
+    },
+    {
+      "epoch": 0.16003700277520813,
+      "grad_norm": 3.789350748062134,
+      "learning_rate": 9.56345181487528e-06,
+      "loss": 1.073,
+      "step": 2076
+    },
+    {
+      "epoch": 0.1601140918902251,
+      "grad_norm": 4.255967617034912,
+      "learning_rate": 9.562941492400149e-06,
+      "loss": 1.039,
+      "step": 2077
+    },
+    {
+      "epoch": 0.16019118100524207,
+      "grad_norm": 3.9734389781951904,
+      "learning_rate": 9.56243088544925e-06,
+      "loss": 1.0608,
+      "step": 2078
+    },
+    {
+      "epoch": 0.16026827012025902,
+      "grad_norm": 4.481680870056152,
+      "learning_rate": 9.561919994054414e-06,
+      "loss": 1.1822,
+      "step": 2079
+    },
+    {
+      "epoch": 0.16034535923527599,
+      "grad_norm": 4.583976745605469,
+      "learning_rate": 9.561408818247493e-06,
+      "loss": 1.1088,
+      "step": 2080
+    },
+    {
+      "epoch": 0.16042244835029293,
+      "grad_norm": 3.912966012954712,
+      "learning_rate": 9.560897358060355e-06,
+      "loss": 1.0697,
+      "step": 2081
+    },
+    {
+      "epoch": 0.1604995374653099,
+      "grad_norm": 3.9257588386535645,
+      "learning_rate": 9.56038561352489e-06,
+      "loss": 1.0764,
+      "step": 2082
+    },
+    {
+      "epoch": 0.16057662658032687,
+      "grad_norm": 3.7236075401306152,
+      "learning_rate": 9.559873584673e-06,
+      "loss": 1.0295,
+      "step": 2083
+    },
+    {
+      "epoch": 0.1606537156953438,
+      "grad_norm": 4.096950531005859,
+      "learning_rate": 9.55936127153661e-06,
+      "loss": 1.0696,
+      "step": 2084
+    },
+    {
+      "epoch": 0.16073080481036078,
+      "grad_norm": 3.90035080909729,
+      "learning_rate": 9.558848674147657e-06,
+      "loss": 1.1163,
+      "step": 2085
+    },
+    {
+      "epoch": 0.16080789392537773,
+      "grad_norm": 3.8980672359466553,
+      "learning_rate": 9.558335792538099e-06,
+      "loss": 1.0165,
+      "step": 2086
+    },
+    {
+      "epoch": 0.1608849830403947,
+      "grad_norm": 4.154537200927734,
+      "learning_rate": 9.557822626739912e-06,
+      "loss": 1.0332,
+      "step": 2087
+    },
+    {
+      "epoch": 0.16096207215541167,
+      "grad_norm": 3.6484971046447754,
+      "learning_rate": 9.557309176785092e-06,
+      "loss": 1.011,
+      "step": 2088
+    },
+    {
+      "epoch": 0.1610391612704286,
+      "grad_norm": 4.025908470153809,
+      "learning_rate": 9.556795442705647e-06,
+      "loss": 1.0423,
+      "step": 2089
+    },
+    {
+      "epoch": 0.16111625038544558,
+      "grad_norm": 4.023404598236084,
+      "learning_rate": 9.556281424533606e-06,
+      "loss": 1.0278,
+      "step": 2090
+    },
+    {
+      "epoch": 0.16119333950046252,
+      "grad_norm": 4.211247444152832,
+      "learning_rate": 9.555767122301016e-06,
+      "loss": 1.0354,
+      "step": 2091
+    },
+    {
+      "epoch": 0.1612704286154795,
+      "grad_norm": 4.087972640991211,
+      "learning_rate": 9.55525253603994e-06,
+      "loss": 1.0494,
+      "step": 2092
+    },
+    {
+      "epoch": 0.16134751773049646,
+      "grad_norm": 3.722517728805542,
+      "learning_rate": 9.554737665782464e-06,
+      "loss": 1.0458,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1614246068455134,
+      "grad_norm": 3.7930004596710205,
+      "learning_rate": 9.55422251156068e-06,
+      "loss": 1.0477,
+      "step": 2094
+    },
+    {
+      "epoch": 0.16150169596053038,
+      "grad_norm": 3.9096274375915527,
+      "learning_rate": 9.55370707340671e-06,
+      "loss": 1.1077,
+      "step": 2095
+    },
+    {
+      "epoch": 0.16157878507554732,
+      "grad_norm": 3.9330289363861084,
+      "learning_rate": 9.553191351352691e-06,
+      "loss": 0.9962,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1616558741905643,
+      "grad_norm": 4.029851913452148,
+      "learning_rate": 9.55267534543077e-06,
+      "loss": 1.138,
+      "step": 2097
+    },
+    {
+      "epoch": 0.16173296330558126,
+      "grad_norm": 4.2497334480285645,
+      "learning_rate": 9.552159055673122e-06,
+      "loss": 1.0651,
+      "step": 2098
+    },
+    {
+      "epoch": 0.1618100524205982,
+      "grad_norm": 3.8570566177368164,
+      "learning_rate": 9.551642482111931e-06,
+      "loss": 1.035,
+      "step": 2099
+    },
+    {
+      "epoch": 0.16188714153561518,
+      "grad_norm": 4.435809135437012,
+      "learning_rate": 9.551125624779407e-06,
+      "loss": 1.1124,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16196423065063212,
+      "grad_norm": 4.071799278259277,
+      "learning_rate": 9.55060848370777e-06,
+      "loss": 1.1761,
+      "step": 2101
+    },
+    {
+      "epoch": 0.1620413197656491,
+      "grad_norm": 3.8368101119995117,
+      "learning_rate": 9.550091058929264e-06,
+      "loss": 1.1273,
+      "step": 2102
+    },
+    {
+      "epoch": 0.16211840888066606,
+      "grad_norm": 4.056320667266846,
+      "learning_rate": 9.549573350476142e-06,
+      "loss": 1.1069,
+      "step": 2103
+    },
+    {
+      "epoch": 0.162195497995683,
+      "grad_norm": 4.4123148918151855,
+      "learning_rate": 9.549055358380687e-06,
+      "loss": 1.0627,
+      "step": 2104
+    },
+    {
+      "epoch": 0.16227258711069997,
+      "grad_norm": 3.741130828857422,
+      "learning_rate": 9.548537082675191e-06,
+      "loss": 1.1405,
+      "step": 2105
+    },
+    {
+      "epoch": 0.16234967622571692,
+      "grad_norm": 3.8030614852905273,
+      "learning_rate": 9.548018523391965e-06,
+      "loss": 1.1287,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1624267653407339,
+      "grad_norm": 3.879117727279663,
+      "learning_rate": 9.54749968056334e-06,
+      "loss": 1.035,
+      "step": 2107
+    },
+    {
+      "epoch": 0.16250385445575086,
+      "grad_norm": 3.7147738933563232,
+      "learning_rate": 9.546980554221659e-06,
+      "loss": 0.9949,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1625809435707678,
+      "grad_norm": 4.359158515930176,
+      "learning_rate": 9.546461144399293e-06,
+      "loss": 1.069,
+      "step": 2109
+    },
+    {
+      "epoch": 0.16265803268578477,
+      "grad_norm": 3.803569793701172,
+      "learning_rate": 9.54594145112862e-06,
+      "loss": 1.0991,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1627351218008017,
+      "grad_norm": 3.9304962158203125,
+      "learning_rate": 9.545421474442042e-06,
+      "loss": 1.0607,
+      "step": 2111
+    },
+    {
+      "epoch": 0.16281221091581868,
+      "grad_norm": 4.038872718811035,
+      "learning_rate": 9.544901214371976e-06,
+      "loss": 1.0551,
+      "step": 2112
+    },
+    {
+      "epoch": 0.16288930003083565,
+      "grad_norm": 4.135643482208252,
+      "learning_rate": 9.544380670950857e-06,
+      "loss": 1.0892,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1629663891458526,
+      "grad_norm": 4.121045112609863,
+      "learning_rate": 9.543859844211141e-06,
+      "loss": 1.0992,
+      "step": 2114
+    },
+    {
+      "epoch": 0.16304347826086957,
+      "grad_norm": 3.6090307235717773,
+      "learning_rate": 9.543338734185295e-06,
+      "loss": 1.0014,
+      "step": 2115
+    },
+    {
+      "epoch": 0.16312056737588654,
+      "grad_norm": 3.578199625015259,
+      "learning_rate": 9.54281734090581e-06,
+      "loss": 1.0165,
+      "step": 2116
+    },
+    {
+      "epoch": 0.16319765649090348,
+      "grad_norm": 4.004699230194092,
+      "learning_rate": 9.542295664405193e-06,
+      "loss": 1.1468,
+      "step": 2117
+    },
+    {
+      "epoch": 0.16327474560592045,
+      "grad_norm": 3.9040887355804443,
+      "learning_rate": 9.541773704715966e-06,
+      "loss": 1.0745,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1633518347209374,
+      "grad_norm": 3.8351752758026123,
+      "learning_rate": 9.541251461870672e-06,
+      "loss": 1.1056,
+      "step": 2119
+    },
+    {
+      "epoch": 0.16342892383595437,
+      "grad_norm": 3.704251527786255,
+      "learning_rate": 9.540728935901867e-06,
+      "loss": 0.9895,
+      "step": 2120
+    },
+    {
+      "epoch": 0.16350601295097134,
+      "grad_norm": 4.368956565856934,
+      "learning_rate": 9.540206126842129e-06,
+      "loss": 1.0922,
+      "step": 2121
+    },
+    {
+      "epoch": 0.16358310206598828,
+      "grad_norm": 3.9333720207214355,
+      "learning_rate": 9.539683034724054e-06,
+      "loss": 1.1807,
+      "step": 2122
+    },
+    {
+      "epoch": 0.16366019118100525,
+      "grad_norm": 3.7927298545837402,
+      "learning_rate": 9.539159659580254e-06,
+      "loss": 1.0409,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1637372802960222,
+      "grad_norm": 4.061433792114258,
+      "learning_rate": 9.53863600144336e-06,
+      "loss": 1.1204,
+      "step": 2124
+    },
+    {
+      "epoch": 0.16381436941103916,
+      "grad_norm": 3.5982680320739746,
+      "learning_rate": 9.538112060346016e-06,
+      "loss": 1.157,
+      "step": 2125
+    },
+    {
+      "epoch": 0.16389145852605613,
+      "grad_norm": 4.211023330688477,
+      "learning_rate": 9.537587836320887e-06,
+      "loss": 1.0758,
+      "step": 2126
+    },
+    {
+      "epoch": 0.16396854764107308,
+      "grad_norm": 3.861417055130005,
+      "learning_rate": 9.53706332940066e-06,
+      "loss": 1.1376,
+      "step": 2127
+    },
+    {
+      "epoch": 0.16404563675609005,
+      "grad_norm": 3.5378191471099854,
+      "learning_rate": 9.536538539618031e-06,
+      "loss": 1.0481,
+      "step": 2128
+    },
+    {
+      "epoch": 0.164122725871107,
+      "grad_norm": 3.9547457695007324,
+      "learning_rate": 9.53601346700572e-06,
+      "loss": 1.0943,
+      "step": 2129
+    },
+    {
+      "epoch": 0.16419981498612396,
+      "grad_norm": 3.9615726470947266,
+      "learning_rate": 9.53548811159646e-06,
+      "loss": 1.0865,
+      "step": 2130
+    },
+    {
+      "epoch": 0.16427690410114093,
+      "grad_norm": 3.7515220642089844,
+      "learning_rate": 9.534962473423008e-06,
+      "loss": 1.1495,
+      "step": 2131
+    },
+    {
+      "epoch": 0.16435399321615787,
+      "grad_norm": 4.2690534591674805,
+      "learning_rate": 9.534436552518134e-06,
+      "loss": 1.0762,
+      "step": 2132
+    },
+    {
+      "epoch": 0.16443108233117484,
+      "grad_norm": 4.034238338470459,
+      "learning_rate": 9.533910348914624e-06,
+      "loss": 1.2103,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1645081714461918,
+      "grad_norm": 3.70621395111084,
+      "learning_rate": 9.533383862645285e-06,
+      "loss": 1.0361,
+      "step": 2134
+    },
+    {
+      "epoch": 0.16458526056120876,
+      "grad_norm": 4.004809856414795,
+      "learning_rate": 9.532857093742941e-06,
+      "loss": 1.1349,
+      "step": 2135
+    },
+    {
+      "epoch": 0.16466234967622573,
+      "grad_norm": 4.048299312591553,
+      "learning_rate": 9.532330042240434e-06,
+      "loss": 1.1853,
+      "step": 2136
+    },
+    {
+      "epoch": 0.16473943879124267,
+      "grad_norm": 4.68448543548584,
+      "learning_rate": 9.531802708170623e-06,
+      "loss": 1.0591,
+      "step": 2137
+    },
+    {
+      "epoch": 0.16481652790625964,
+      "grad_norm": 3.5135207176208496,
+      "learning_rate": 9.531275091566384e-06,
+      "loss": 0.9884,
+      "step": 2138
+    },
+    {
+      "epoch": 0.16489361702127658,
+      "grad_norm": 4.260873317718506,
+      "learning_rate": 9.530747192460609e-06,
+      "loss": 1.1469,
+      "step": 2139
+    },
+    {
+      "epoch": 0.16497070613629355,
+      "grad_norm": 3.82432222366333,
+      "learning_rate": 9.530219010886214e-06,
+      "loss": 1.1344,
+      "step": 2140
+    },
+    {
+      "epoch": 0.16504779525131053,
+      "grad_norm": 4.045070171356201,
+      "learning_rate": 9.529690546876125e-06,
+      "loss": 1.0296,
+      "step": 2141
+    },
+    {
+      "epoch": 0.16512488436632747,
+      "grad_norm": 4.116434574127197,
+      "learning_rate": 9.529161800463291e-06,
+      "loss": 1.1479,
+      "step": 2142
+    },
+    {
+      "epoch": 0.16520197348134444,
+      "grad_norm": 3.7906370162963867,
+      "learning_rate": 9.528632771680677e-06,
+      "loss": 1.0819,
+      "step": 2143
+    },
+    {
+      "epoch": 0.16527906259636138,
+      "grad_norm": 4.043308734893799,
+      "learning_rate": 9.528103460561262e-06,
+      "loss": 1.1089,
+      "step": 2144
+    },
+    {
+      "epoch": 0.16535615171137835,
+      "grad_norm": 4.255034446716309,
+      "learning_rate": 9.52757386713805e-06,
+      "loss": 1.0368,
+      "step": 2145
+    },
+    {
+      "epoch": 0.16543324082639532,
+      "grad_norm": 3.905506134033203,
+      "learning_rate": 9.527043991444053e-06,
+      "loss": 0.9421,
+      "step": 2146
+    },
+    {
+      "epoch": 0.16551032994141227,
+      "grad_norm": 4.145698070526123,
+      "learning_rate": 9.526513833512312e-06,
+      "loss": 1.1444,
+      "step": 2147
+    },
+    {
+      "epoch": 0.16558741905642924,
+      "grad_norm": 3.9307868480682373,
+      "learning_rate": 9.525983393375877e-06,
+      "loss": 1.1171,
+      "step": 2148
+    },
+    {
+      "epoch": 0.16566450817144618,
+      "grad_norm": 3.6001784801483154,
+      "learning_rate": 9.525452671067816e-06,
+      "loss": 0.9414,
+      "step": 2149
+    },
+    {
+      "epoch": 0.16574159728646315,
+      "grad_norm": 4.438070774078369,
+      "learning_rate": 9.524921666621221e-06,
+      "loss": 1.1429,
+      "step": 2150
+    },
+    {
+      "epoch": 0.16581868640148012,
+      "grad_norm": 3.5969676971435547,
+      "learning_rate": 9.524390380069195e-06,
+      "loss": 0.97,
+      "step": 2151
+    },
+    {
+      "epoch": 0.16589577551649706,
+      "grad_norm": 4.561358451843262,
+      "learning_rate": 9.52385881144486e-06,
+      "loss": 1.0779,
+      "step": 2152
+    },
+    {
+      "epoch": 0.16597286463151403,
+      "grad_norm": 3.637179136276245,
+      "learning_rate": 9.52332696078136e-06,
+      "loss": 0.9986,
+      "step": 2153
+    },
+    {
+      "epoch": 0.16604995374653098,
+      "grad_norm": 3.7705628871917725,
+      "learning_rate": 9.522794828111849e-06,
+      "loss": 1.0871,
+      "step": 2154
+    },
+    {
+      "epoch": 0.16612704286154795,
+      "grad_norm": 3.748274803161621,
+      "learning_rate": 9.522262413469505e-06,
+      "loss": 1.0919,
+      "step": 2155
+    },
+    {
+      "epoch": 0.16620413197656492,
+      "grad_norm": 3.7525219917297363,
+      "learning_rate": 9.52172971688752e-06,
+      "loss": 1.043,
+      "step": 2156
+    },
+    {
+      "epoch": 0.16628122109158186,
+      "grad_norm": 4.160288333892822,
+      "learning_rate": 9.521196738399107e-06,
+      "loss": 1.0777,
+      "step": 2157
+    },
+    {
+      "epoch": 0.16635831020659883,
+      "grad_norm": 4.212968826293945,
+      "learning_rate": 9.520663478037493e-06,
+      "loss": 0.9929,
+      "step": 2158
+    },
+    {
+      "epoch": 0.1664353993216158,
+      "grad_norm": 4.7758588790893555,
+      "learning_rate": 9.520129935835924e-06,
+      "loss": 1.0723,
+      "step": 2159
+    },
+    {
+      "epoch": 0.16651248843663274,
+      "grad_norm": 4.326691150665283,
+      "learning_rate": 9.519596111827665e-06,
+      "loss": 0.9742,
+      "step": 2160
+    },
+    {
+      "epoch": 0.16658957755164971,
+      "grad_norm": 3.9504857063293457,
+      "learning_rate": 9.519062006045995e-06,
+      "loss": 1.0658,
+      "step": 2161
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 3.9129183292388916,
+      "learning_rate": 9.518527618524213e-06,
+      "loss": 1.1364,
+      "step": 2162
+    },
+    {
+      "epoch": 0.16674375578168363,
+      "grad_norm": 3.5338165760040283,
+      "learning_rate": 9.517992949295637e-06,
+      "loss": 1.0188,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1668208448967006,
+      "grad_norm": 3.9790713787078857,
+      "learning_rate": 9.5174579983936e-06,
+      "loss": 1.0366,
+      "step": 2164
+    },
+    {
+      "epoch": 0.16689793401171754,
+      "grad_norm": 3.752887725830078,
+      "learning_rate": 9.516922765851453e-06,
+      "loss": 1.0697,
+      "step": 2165
+    },
+    {
+      "epoch": 0.1669750231267345,
+      "grad_norm": 4.03433895111084,
+      "learning_rate": 9.516387251702566e-06,
+      "loss": 1.1041,
+      "step": 2166
+    },
+    {
+      "epoch": 0.16705211224175145,
+      "grad_norm": 4.188197612762451,
+      "learning_rate": 9.515851455980325e-06,
+      "loss": 1.0372,
+      "step": 2167
+    },
+    {
+      "epoch": 0.16712920135676843,
+      "grad_norm": 3.5528385639190674,
+      "learning_rate": 9.515315378718132e-06,
+      "loss": 1.021,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1672062904717854,
+      "grad_norm": 4.173407077789307,
+      "learning_rate": 9.514779019949411e-06,
+      "loss": 1.0054,
+      "step": 2169
+    },
+    {
+      "epoch": 0.16728337958680234,
+      "grad_norm": 3.9701037406921387,
+      "learning_rate": 9.514242379707604e-06,
+      "loss": 1.0418,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1673604687018193,
+      "grad_norm": 4.371380805969238,
+      "learning_rate": 9.513705458026162e-06,
+      "loss": 1.1153,
+      "step": 2171
+    },
+    {
+      "epoch": 0.16743755781683625,
+      "grad_norm": 4.344913482666016,
+      "learning_rate": 9.51316825493856e-06,
+      "loss": 1.0405,
+      "step": 2172
+    },
+    {
+      "epoch": 0.16751464693185322,
+      "grad_norm": 4.681027412414551,
+      "learning_rate": 9.512630770478292e-06,
+      "loss": 1.0118,
+      "step": 2173
+    },
+    {
+      "epoch": 0.1675917360468702,
+      "grad_norm": 4.137608528137207,
+      "learning_rate": 9.51209300467887e-06,
+      "loss": 1.1043,
+      "step": 2174
+    },
+    {
+      "epoch": 0.16766882516188714,
+      "grad_norm": 3.731459379196167,
+      "learning_rate": 9.511554957573816e-06,
+      "loss": 0.901,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1677459142769041,
+      "grad_norm": 3.322524070739746,
+      "learning_rate": 9.511016629196676e-06,
+      "loss": 0.9328,
+      "step": 2176
+    },
+    {
+      "epoch": 0.16782300339192105,
+      "grad_norm": 3.5919430255889893,
+      "learning_rate": 9.510478019581013e-06,
+      "loss": 0.9859,
+      "step": 2177
+    },
+    {
+      "epoch": 0.16790009250693802,
+      "grad_norm": 3.575169086456299,
+      "learning_rate": 9.509939128760406e-06,
+      "loss": 1.0423,
+      "step": 2178
+    },
+    {
+      "epoch": 0.167977181621955,
+      "grad_norm": 4.213058948516846,
+      "learning_rate": 9.509399956768452e-06,
+      "loss": 1.0613,
+      "step": 2179
+    },
+    {
+      "epoch": 0.16805427073697193,
+      "grad_norm": 3.759490966796875,
+      "learning_rate": 9.508860503638765e-06,
+      "loss": 1.0447,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1681313598519889,
+      "grad_norm": 4.783963203430176,
+      "learning_rate": 9.508320769404978e-06,
+      "loss": 1.0838,
+      "step": 2181
+    },
+    {
+      "epoch": 0.16820844896700585,
+      "grad_norm": 3.997779369354248,
+      "learning_rate": 9.507780754100739e-06,
+      "loss": 1.156,
+      "step": 2182
+    },
+    {
+      "epoch": 0.16828553808202282,
+      "grad_norm": 3.6696715354919434,
+      "learning_rate": 9.507240457759717e-06,
+      "loss": 1.0282,
+      "step": 2183
+    },
+    {
+      "epoch": 0.1683626271970398,
+      "grad_norm": 3.8719160556793213,
+      "learning_rate": 9.506699880415597e-06,
+      "loss": 1.0963,
+      "step": 2184
+    },
+    {
+      "epoch": 0.16843971631205673,
+      "grad_norm": 3.7864603996276855,
+      "learning_rate": 9.50615902210208e-06,
+      "loss": 1.0538,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1685168054270737,
+      "grad_norm": 4.357522964477539,
+      "learning_rate": 9.505617882852884e-06,
+      "loss": 1.0002,
+      "step": 2186
+    },
+    {
+      "epoch": 0.16859389454209064,
+      "grad_norm": 4.590842247009277,
+      "learning_rate": 9.505076462701752e-06,
+      "loss": 1.0613,
+      "step": 2187
+    },
+    {
+      "epoch": 0.16867098365710761,
+      "grad_norm": 5.570279121398926,
+      "learning_rate": 9.504534761682431e-06,
+      "loss": 1.0677,
+      "step": 2188
+    },
+    {
+      "epoch": 0.16874807277212459,
+      "grad_norm": 3.9890170097351074,
+      "learning_rate": 9.503992779828698e-06,
+      "loss": 0.9587,
+      "step": 2189
+    },
+    {
+      "epoch": 0.16882516188714153,
+      "grad_norm": 3.743421792984009,
+      "learning_rate": 9.503450517174344e-06,
+      "loss": 1.1186,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1689022510021585,
+      "grad_norm": 4.146291255950928,
+      "learning_rate": 9.502907973753173e-06,
+      "loss": 1.1364,
+      "step": 2191
+    },
+    {
+      "epoch": 0.16897934011717544,
+      "grad_norm": 4.190448760986328,
+      "learning_rate": 9.50236514959901e-06,
+      "loss": 1.1372,
+      "step": 2192
+    },
+    {
+      "epoch": 0.1690564292321924,
+      "grad_norm": 4.3127641677856445,
+      "learning_rate": 9.501822044745701e-06,
+      "loss": 1.1888,
+      "step": 2193
+    },
+    {
+      "epoch": 0.16913351834720938,
+      "grad_norm": 3.6845946311950684,
+      "learning_rate": 9.501278659227101e-06,
+      "loss": 0.9711,
+      "step": 2194
+    },
+    {
+      "epoch": 0.16921060746222633,
+      "grad_norm": 3.9466843605041504,
+      "learning_rate": 9.500734993077089e-06,
+      "loss": 1.0328,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1692876965772433,
+      "grad_norm": 4.372476577758789,
+      "learning_rate": 9.500191046329561e-06,
+      "loss": 1.0785,
+      "step": 2196
+    },
+    {
+      "epoch": 0.16936478569226024,
+      "grad_norm": 3.853376865386963,
+      "learning_rate": 9.499646819018429e-06,
+      "loss": 1.0748,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1694418748072772,
+      "grad_norm": 3.7514140605926514,
+      "learning_rate": 9.499102311177622e-06,
+      "loss": 1.1136,
+      "step": 2198
+    },
+    {
+      "epoch": 0.16951896392229418,
+      "grad_norm": 3.941514730453491,
+      "learning_rate": 9.498557522841085e-06,
+      "loss": 1.2299,
+      "step": 2199
+    },
+    {
+      "epoch": 0.16959605303731112,
+      "grad_norm": 4.029524326324463,
+      "learning_rate": 9.498012454042786e-06,
+      "loss": 1.0973,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1696731421523281,
+      "grad_norm": 4.247512340545654,
+      "learning_rate": 9.497467104816709e-06,
+      "loss": 1.1146,
+      "step": 2201
+    },
+    {
+      "epoch": 0.16975023126734506,
+      "grad_norm": 3.7475624084472656,
+      "learning_rate": 9.496921475196847e-06,
+      "loss": 1.0355,
+      "step": 2202
+    },
+    {
+      "epoch": 0.169827320382362,
+      "grad_norm": 3.81801438331604,
+      "learning_rate": 9.496375565217225e-06,
+      "loss": 1.1312,
+      "step": 2203
+    },
+    {
+      "epoch": 0.16990440949737898,
+      "grad_norm": 4.464087963104248,
+      "learning_rate": 9.49582937491187e-06,
+      "loss": 1.1643,
+      "step": 2204
+    },
+    {
+      "epoch": 0.16998149861239592,
+      "grad_norm": 3.895153522491455,
+      "learning_rate": 9.49528290431484e-06,
+      "loss": 1.1044,
+      "step": 2205
+    },
+    {
+      "epoch": 0.1700585877274129,
+      "grad_norm": 4.139657974243164,
+      "learning_rate": 9.494736153460204e-06,
+      "loss": 1.1443,
+      "step": 2206
+    },
+    {
+      "epoch": 0.17013567684242986,
+      "grad_norm": 3.7436201572418213,
+      "learning_rate": 9.494189122382046e-06,
+      "loss": 1.0922,
+      "step": 2207
+    },
+    {
+      "epoch": 0.1702127659574468,
+      "grad_norm": 4.397914409637451,
+      "learning_rate": 9.493641811114472e-06,
+      "loss": 1.1507,
+      "step": 2208
+    },
+    {
+      "epoch": 0.17028985507246377,
+      "grad_norm": 3.714843511581421,
+      "learning_rate": 9.493094219691606e-06,
+      "loss": 1.0862,
+      "step": 2209
+    },
+    {
+      "epoch": 0.17036694418748072,
+      "grad_norm": 3.666947603225708,
+      "learning_rate": 9.492546348147584e-06,
+      "loss": 1.0753,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1704440333024977,
+      "grad_norm": 4.164958953857422,
+      "learning_rate": 9.491998196516564e-06,
+      "loss": 1.133,
+      "step": 2211
+    },
+    {
+      "epoch": 0.17052112241751466,
+      "grad_norm": 4.119793891906738,
+      "learning_rate": 9.491449764832722e-06,
+      "loss": 1.0534,
+      "step": 2212
+    },
+    {
+      "epoch": 0.1705982115325316,
+      "grad_norm": 3.694411277770996,
+      "learning_rate": 9.49090105313025e-06,
+      "loss": 1.0354,
+      "step": 2213
+    },
+    {
+      "epoch": 0.17067530064754857,
+      "grad_norm": 4.022437572479248,
+      "learning_rate": 9.490352061443355e-06,
+      "loss": 1.0719,
+      "step": 2214
+    },
+    {
+      "epoch": 0.17075238976256552,
+      "grad_norm": 4.365479946136475,
+      "learning_rate": 9.489802789806265e-06,
+      "loss": 1.2675,
+      "step": 2215
+    },
+    {
+      "epoch": 0.17082947887758249,
+      "grad_norm": 4.730959892272949,
+      "learning_rate": 9.489253238253225e-06,
+      "loss": 1.1868,
+      "step": 2216
+    },
+    {
+      "epoch": 0.17090656799259946,
+      "grad_norm": 3.6116209030151367,
+      "learning_rate": 9.488703406818496e-06,
+      "loss": 1.0207,
+      "step": 2217
+    },
+    {
+      "epoch": 0.1709836571076164,
+      "grad_norm": 4.13337516784668,
+      "learning_rate": 9.488153295536358e-06,
+      "loss": 1.2437,
+      "step": 2218
+    },
+    {
+      "epoch": 0.17106074622263337,
+      "grad_norm": 4.369235038757324,
+      "learning_rate": 9.487602904441106e-06,
+      "loss": 1.1526,
+      "step": 2219
+    },
+    {
+      "epoch": 0.1711378353376503,
+      "grad_norm": 3.9253830909729004,
+      "learning_rate": 9.487052233567055e-06,
+      "loss": 1.0983,
+      "step": 2220
+    },
+    {
+      "epoch": 0.17121492445266728,
+      "grad_norm": 3.9150092601776123,
+      "learning_rate": 9.486501282948535e-06,
+      "loss": 1.0501,
+      "step": 2221
+    },
+    {
+      "epoch": 0.17129201356768425,
+      "grad_norm": 4.1889424324035645,
+      "learning_rate": 9.485950052619898e-06,
+      "loss": 1.0532,
+      "step": 2222
+    },
+    {
+      "epoch": 0.1713691026827012,
+      "grad_norm": 4.013206481933594,
+      "learning_rate": 9.485398542615507e-06,
+      "loss": 1.0343,
+      "step": 2223
+    },
+    {
+      "epoch": 0.17144619179771817,
+      "grad_norm": 3.747413158416748,
+      "learning_rate": 9.484846752969747e-06,
+      "loss": 0.9813,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1715232809127351,
+      "grad_norm": 4.145966053009033,
+      "learning_rate": 9.484294683717021e-06,
+      "loss": 1.047,
+      "step": 2225
+    },
+    {
+      "epoch": 0.17160037002775208,
+      "grad_norm": 3.959221124649048,
+      "learning_rate": 9.483742334891747e-06,
+      "loss": 0.988,
+      "step": 2226
+    },
+    {
+      "epoch": 0.17167745914276905,
+      "grad_norm": 4.141855239868164,
+      "learning_rate": 9.483189706528358e-06,
+      "loss": 1.0136,
+      "step": 2227
+    },
+    {
+      "epoch": 0.171754548257786,
+      "grad_norm": 4.037559509277344,
+      "learning_rate": 9.482636798661311e-06,
+      "loss": 1.003,
+      "step": 2228
+    },
+    {
+      "epoch": 0.17183163737280296,
+      "grad_norm": 3.8605642318725586,
+      "learning_rate": 9.482083611325076e-06,
+      "loss": 1.105,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1719087264878199,
+      "grad_norm": 4.074838161468506,
+      "learning_rate": 9.48153014455414e-06,
+      "loss": 1.073,
+      "step": 2230
+    },
+    {
+      "epoch": 0.17198581560283688,
+      "grad_norm": 3.978645086288452,
+      "learning_rate": 9.480976398383013e-06,
+      "loss": 1.1489,
+      "step": 2231
+    },
+    {
+      "epoch": 0.17206290471785385,
+      "grad_norm": 4.13065242767334,
+      "learning_rate": 9.480422372846212e-06,
+      "loss": 1.1217,
+      "step": 2232
+    },
+    {
+      "epoch": 0.1721399938328708,
+      "grad_norm": 4.274265766143799,
+      "learning_rate": 9.479868067978282e-06,
+      "loss": 1.0354,
+      "step": 2233
+    },
+    {
+      "epoch": 0.17221708294788776,
+      "grad_norm": 3.908271312713623,
+      "learning_rate": 9.47931348381378e-06,
+      "loss": 1.0812,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1722941720629047,
+      "grad_norm": 4.128848075866699,
+      "learning_rate": 9.478758620387281e-06,
+      "loss": 1.0114,
+      "step": 2235
+    },
+    {
+      "epoch": 0.17237126117792168,
+      "grad_norm": 4.402393341064453,
+      "learning_rate": 9.478203477733377e-06,
+      "loss": 1.1512,
+      "step": 2236
+    },
+    {
+      "epoch": 0.17244835029293865,
+      "grad_norm": 4.182422161102295,
+      "learning_rate": 9.477648055886682e-06,
+      "loss": 1.0967,
+      "step": 2237
+    },
+    {
+      "epoch": 0.1725254394079556,
+      "grad_norm": 3.684023857116699,
+      "learning_rate": 9.477092354881818e-06,
+      "loss": 1.0843,
+      "step": 2238
+    },
+    {
+      "epoch": 0.17260252852297256,
+      "grad_norm": 4.2031660079956055,
+      "learning_rate": 9.476536374753434e-06,
+      "loss": 1.1126,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1726796176379895,
+      "grad_norm": 3.8642640113830566,
+      "learning_rate": 9.475980115536193e-06,
+      "loss": 1.0219,
+      "step": 2240
+    },
+    {
+      "epoch": 0.17275670675300647,
+      "grad_norm": 4.426207542419434,
+      "learning_rate": 9.475423577264772e-06,
+      "loss": 1.1832,
+      "step": 2241
+    },
+    {
+      "epoch": 0.17283379586802344,
+      "grad_norm": 3.9104762077331543,
+      "learning_rate": 9.474866759973871e-06,
+      "loss": 1.0344,
+      "step": 2242
+    },
+    {
+      "epoch": 0.17291088498304039,
+      "grad_norm": 4.044737815856934,
+      "learning_rate": 9.474309663698202e-06,
+      "loss": 1.0671,
+      "step": 2243
+    },
+    {
+      "epoch": 0.17298797409805736,
+      "grad_norm": 4.002004623413086,
+      "learning_rate": 9.473752288472499e-06,
+      "loss": 1.0373,
+      "step": 2244
+    },
+    {
+      "epoch": 0.17306506321307433,
+      "grad_norm": 3.7214102745056152,
+      "learning_rate": 9.473194634331512e-06,
+      "loss": 1.0278,
+      "step": 2245
+    },
+    {
+      "epoch": 0.17314215232809127,
+      "grad_norm": 3.870584487915039,
+      "learning_rate": 9.472636701310005e-06,
+      "loss": 0.9826,
+      "step": 2246
+    },
+    {
+      "epoch": 0.17321924144310824,
+      "grad_norm": 3.6483523845672607,
+      "learning_rate": 9.472078489442766e-06,
+      "loss": 1.086,
+      "step": 2247
+    },
+    {
+      "epoch": 0.17329633055812518,
+      "grad_norm": 3.894209384918213,
+      "learning_rate": 9.471519998764593e-06,
+      "loss": 1.0286,
+      "step": 2248
+    },
+    {
+      "epoch": 0.17337341967314215,
+      "grad_norm": 3.856490135192871,
+      "learning_rate": 9.470961229310307e-06,
+      "loss": 1.1207,
+      "step": 2249
+    },
+    {
+      "epoch": 0.17345050878815912,
+      "grad_norm": 5.136345386505127,
+      "learning_rate": 9.470402181114747e-06,
+      "loss": 1.0207,
+      "step": 2250
+    },
+    {
+      "epoch": 0.17352759790317607,
+      "grad_norm": 4.148111343383789,
+      "learning_rate": 9.46984285421276e-06,
+      "loss": 1.1127,
+      "step": 2251
+    },
+    {
+      "epoch": 0.17360468701819304,
+      "grad_norm": 3.8354268074035645,
+      "learning_rate": 9.469283248639223e-06,
+      "loss": 1.0683,
+      "step": 2252
+    },
+    {
+      "epoch": 0.17368177613320998,
+      "grad_norm": 3.8218307495117188,
+      "learning_rate": 9.468723364429021e-06,
+      "loss": 1.0874,
+      "step": 2253
+    },
+    {
+      "epoch": 0.17375886524822695,
+      "grad_norm": 4.011178016662598,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 1.0788,
+      "step": 2254
+    },
+    {
+      "epoch": 0.17383595436324392,
+      "grad_norm": 3.5386111736297607,
+      "learning_rate": 9.467602760238268e-06,
+      "loss": 0.8796,
+      "step": 2255
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 3.82024884223938,
+      "learning_rate": 9.467042040327582e-06,
+      "loss": 1.0407,
+      "step": 2256
+    },
+    {
+      "epoch": 0.17399013259327784,
+      "grad_norm": 4.042757987976074,
+      "learning_rate": 9.46648104191996e-06,
+      "loss": 1.1627,
+      "step": 2257
+    },
+    {
+      "epoch": 0.17406722170829478,
+      "grad_norm": 3.925922393798828,
+      "learning_rate": 9.465919765050375e-06,
+      "loss": 1.1055,
+      "step": 2258
+    },
+    {
+      "epoch": 0.17414431082331175,
+      "grad_norm": 3.940410852432251,
+      "learning_rate": 9.465358209753824e-06,
+      "loss": 0.9985,
+      "step": 2259
+    },
+    {
+      "epoch": 0.17422139993832872,
+      "grad_norm": 3.904409646987915,
+      "learning_rate": 9.464796376065316e-06,
+      "loss": 1.0892,
+      "step": 2260
+    },
+    {
+      "epoch": 0.17429848905334566,
+      "grad_norm": 3.91276216506958,
+      "learning_rate": 9.464234264019876e-06,
+      "loss": 1.0863,
+      "step": 2261
+    },
+    {
+      "epoch": 0.17437557816836263,
+      "grad_norm": 3.8074424266815186,
+      "learning_rate": 9.463671873652551e-06,
+      "loss": 1.0709,
+      "step": 2262
+    },
+    {
+      "epoch": 0.17445266728337958,
+      "grad_norm": 4.295366287231445,
+      "learning_rate": 9.463109204998405e-06,
+      "loss": 1.0679,
+      "step": 2263
+    },
+    {
+      "epoch": 0.17452975639839655,
+      "grad_norm": 3.8966548442840576,
+      "learning_rate": 9.462546258092512e-06,
+      "loss": 1.0652,
+      "step": 2264
+    },
+    {
+      "epoch": 0.17460684551341352,
+      "grad_norm": 3.9106335639953613,
+      "learning_rate": 9.461983032969972e-06,
+      "loss": 1.0031,
+      "step": 2265
+    },
+    {
+      "epoch": 0.17468393462843046,
+      "grad_norm": 4.223884582519531,
+      "learning_rate": 9.4614195296659e-06,
+      "loss": 1.1193,
+      "step": 2266
+    },
+    {
+      "epoch": 0.17476102374344743,
+      "grad_norm": 3.972520351409912,
+      "learning_rate": 9.460855748215427e-06,
+      "loss": 1.1208,
+      "step": 2267
+    },
+    {
+      "epoch": 0.17483811285846437,
+      "grad_norm": 3.772442102432251,
+      "learning_rate": 9.460291688653702e-06,
+      "loss": 1.0707,
+      "step": 2268
+    },
+    {
+      "epoch": 0.17491520197348134,
+      "grad_norm": 4.243513584136963,
+      "learning_rate": 9.45972735101589e-06,
+      "loss": 1.0806,
+      "step": 2269
+    },
+    {
+      "epoch": 0.17499229108849831,
+      "grad_norm": 3.9305193424224854,
+      "learning_rate": 9.459162735337176e-06,
+      "loss": 1.0473,
+      "step": 2270
+    },
+    {
+      "epoch": 0.17506938020351526,
+      "grad_norm": 3.5326061248779297,
+      "learning_rate": 9.458597841652758e-06,
+      "loss": 1.0894,
+      "step": 2271
+    },
+    {
+      "epoch": 0.17514646931853223,
+      "grad_norm": 4.2558746337890625,
+      "learning_rate": 9.458032669997858e-06,
+      "loss": 1.1245,
+      "step": 2272
+    },
+    {
+      "epoch": 0.17522355843354917,
+      "grad_norm": 3.9435577392578125,
+      "learning_rate": 9.457467220407707e-06,
+      "loss": 1.0961,
+      "step": 2273
+    },
+    {
+      "epoch": 0.17530064754856614,
+      "grad_norm": 4.034180641174316,
+      "learning_rate": 9.456901492917564e-06,
+      "loss": 1.0136,
+      "step": 2274
+    },
+    {
+      "epoch": 0.1753777366635831,
+      "grad_norm": 4.0210676193237305,
+      "learning_rate": 9.456335487562695e-06,
+      "loss": 1.058,
+      "step": 2275
+    },
+    {
+      "epoch": 0.17545482577860005,
+      "grad_norm": 3.803864002227783,
+      "learning_rate": 9.455769204378388e-06,
+      "loss": 0.9875,
+      "step": 2276
+    },
+    {
+      "epoch": 0.17553191489361702,
+      "grad_norm": 3.870218515396118,
+      "learning_rate": 9.455202643399948e-06,
+      "loss": 1.0243,
+      "step": 2277
+    },
+    {
+      "epoch": 0.17560900400863397,
+      "grad_norm": 3.751405954360962,
+      "learning_rate": 9.454635804662697e-06,
+      "loss": 1.0363,
+      "step": 2278
+    },
+    {
+      "epoch": 0.17568609312365094,
+      "grad_norm": 3.922459125518799,
+      "learning_rate": 9.454068688201975e-06,
+      "loss": 1.0977,
+      "step": 2279
+    },
+    {
+      "epoch": 0.1757631822386679,
+      "grad_norm": 4.521456241607666,
+      "learning_rate": 9.453501294053139e-06,
+      "loss": 1.1405,
+      "step": 2280
+    },
+    {
+      "epoch": 0.17584027135368485,
+      "grad_norm": 4.3583478927612305,
+      "learning_rate": 9.452933622251561e-06,
+      "loss": 1.1479,
+      "step": 2281
+    },
+    {
+      "epoch": 0.17591736046870182,
+      "grad_norm": 3.9341979026794434,
+      "learning_rate": 9.452365672832635e-06,
+      "loss": 1.0467,
+      "step": 2282
+    },
+    {
+      "epoch": 0.1759944495837188,
+      "grad_norm": 4.151937484741211,
+      "learning_rate": 9.45179744583177e-06,
+      "loss": 1.0061,
+      "step": 2283
+    },
+    {
+      "epoch": 0.17607153869873574,
+      "grad_norm": 3.7211530208587646,
+      "learning_rate": 9.451228941284389e-06,
+      "loss": 1.026,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1761486278137527,
+      "grad_norm": 4.126251697540283,
+      "learning_rate": 9.450660159225938e-06,
+      "loss": 1.055,
+      "step": 2285
+    },
+    {
+      "epoch": 0.17622571692876965,
+      "grad_norm": 3.7324936389923096,
+      "learning_rate": 9.450091099691876e-06,
+      "loss": 1.0071,
+      "step": 2286
+    },
+    {
+      "epoch": 0.17630280604378662,
+      "grad_norm": 3.911670446395874,
+      "learning_rate": 9.449521762717682e-06,
+      "loss": 1.0975,
+      "step": 2287
+    },
+    {
+      "epoch": 0.1763798951588036,
+      "grad_norm": 3.6570870876312256,
+      "learning_rate": 9.44895214833885e-06,
+      "loss": 0.911,
+      "step": 2288
+    },
+    {
+      "epoch": 0.17645698427382053,
+      "grad_norm": 3.5680623054504395,
+      "learning_rate": 9.448382256590893e-06,
+      "loss": 0.9641,
+      "step": 2289
+    },
+    {
+      "epoch": 0.1765340733888375,
+      "grad_norm": 3.8727200031280518,
+      "learning_rate": 9.447812087509343e-06,
+      "loss": 1.1022,
+      "step": 2290
+    },
+    {
+      "epoch": 0.17661116250385445,
+      "grad_norm": 3.9284706115722656,
+      "learning_rate": 9.447241641129743e-06,
+      "loss": 1.1042,
+      "step": 2291
+    },
+    {
+      "epoch": 0.17668825161887142,
+      "grad_norm": 4.463204383850098,
+      "learning_rate": 9.446670917487662e-06,
+      "loss": 1.1087,
+      "step": 2292
+    },
+    {
+      "epoch": 0.1767653407338884,
+      "grad_norm": 3.742551803588867,
+      "learning_rate": 9.446099916618676e-06,
+      "loss": 1.143,
+      "step": 2293
+    },
+    {
+      "epoch": 0.17684242984890533,
+      "grad_norm": 3.764596462249756,
+      "learning_rate": 9.445528638558389e-06,
+      "loss": 1.0978,
+      "step": 2294
+    },
+    {
+      "epoch": 0.1769195189639223,
+      "grad_norm": 3.68927001953125,
+      "learning_rate": 9.444957083342415e-06,
+      "loss": 0.8989,
+      "step": 2295
+    },
+    {
+      "epoch": 0.17699660807893924,
+      "grad_norm": 3.6712417602539062,
+      "learning_rate": 9.444385251006389e-06,
+      "loss": 0.9342,
+      "step": 2296
+    },
+    {
+      "epoch": 0.17707369719395621,
+      "grad_norm": 4.3771467208862305,
+      "learning_rate": 9.44381314158596e-06,
+      "loss": 1.1587,
+      "step": 2297
+    },
+    {
+      "epoch": 0.17715078630897318,
+      "grad_norm": 3.71481990814209,
+      "learning_rate": 9.443240755116797e-06,
+      "loss": 1.0263,
+      "step": 2298
+    },
+    {
+      "epoch": 0.17722787542399013,
+      "grad_norm": 3.5364553928375244,
+      "learning_rate": 9.442668091634583e-06,
+      "loss": 0.9791,
+      "step": 2299
+    },
+    {
+      "epoch": 0.1773049645390071,
+      "grad_norm": 4.014560699462891,
+      "learning_rate": 9.442095151175024e-06,
+      "loss": 0.9919,
+      "step": 2300
+    },
+    {
+      "epoch": 0.17738205365402404,
+      "grad_norm": 4.145420074462891,
+      "learning_rate": 9.441521933773839e-06,
+      "loss": 1.047,
+      "step": 2301
+    },
+    {
+      "epoch": 0.177459142769041,
+      "grad_norm": 4.094470977783203,
+      "learning_rate": 9.440948439466764e-06,
+      "loss": 1.095,
+      "step": 2302
+    },
+    {
+      "epoch": 0.17753623188405798,
+      "grad_norm": 3.7132604122161865,
+      "learning_rate": 9.440374668289553e-06,
+      "loss": 1.0473,
+      "step": 2303
+    },
+    {
+      "epoch": 0.17761332099907493,
+      "grad_norm": 3.507946491241455,
+      "learning_rate": 9.439800620277981e-06,
+      "loss": 1.0164,
+      "step": 2304
+    },
+    {
+      "epoch": 0.1776904101140919,
+      "grad_norm": 3.750944137573242,
+      "learning_rate": 9.439226295467834e-06,
+      "loss": 1.048,
+      "step": 2305
+    },
+    {
+      "epoch": 0.17776749922910884,
+      "grad_norm": 3.6753499507904053,
+      "learning_rate": 9.438651693894919e-06,
+      "loss": 1.0394,
+      "step": 2306
+    },
+    {
+      "epoch": 0.1778445883441258,
+      "grad_norm": 4.113043785095215,
+      "learning_rate": 9.438076815595058e-06,
+      "loss": 0.9472,
+      "step": 2307
+    },
+    {
+      "epoch": 0.17792167745914278,
+      "grad_norm": 4.260312080383301,
+      "learning_rate": 9.437501660604094e-06,
+      "loss": 1.1437,
+      "step": 2308
+    },
+    {
+      "epoch": 0.17799876657415972,
+      "grad_norm": 3.7821967601776123,
+      "learning_rate": 9.436926228957883e-06,
+      "loss": 1.0623,
+      "step": 2309
+    },
+    {
+      "epoch": 0.1780758556891767,
+      "grad_norm": 3.920973062515259,
+      "learning_rate": 9.436350520692303e-06,
+      "loss": 1.0917,
+      "step": 2310
+    },
+    {
+      "epoch": 0.17815294480419364,
+      "grad_norm": 3.868009567260742,
+      "learning_rate": 9.435774535843243e-06,
+      "loss": 1.0319,
+      "step": 2311
+    },
+    {
+      "epoch": 0.1782300339192106,
+      "grad_norm": 3.780529022216797,
+      "learning_rate": 9.435198274446614e-06,
+      "loss": 0.9861,
+      "step": 2312
+    },
+    {
+      "epoch": 0.17830712303422758,
+      "grad_norm": 3.820247173309326,
+      "learning_rate": 9.434621736538343e-06,
+      "loss": 1.1113,
+      "step": 2313
+    },
+    {
+      "epoch": 0.17838421214924452,
+      "grad_norm": 4.4522705078125,
+      "learning_rate": 9.434044922154375e-06,
+      "loss": 1.1872,
+      "step": 2314
+    },
+    {
+      "epoch": 0.1784613012642615,
+      "grad_norm": 4.132708549499512,
+      "learning_rate": 9.433467831330671e-06,
+      "loss": 1.0276,
+      "step": 2315
+    },
+    {
+      "epoch": 0.17853839037927843,
+      "grad_norm": 3.8560283184051514,
+      "learning_rate": 9.432890464103208e-06,
+      "loss": 1.077,
+      "step": 2316
+    },
+    {
+      "epoch": 0.1786154794942954,
+      "grad_norm": 3.8112268447875977,
+      "learning_rate": 9.432312820507982e-06,
+      "loss": 1.059,
+      "step": 2317
+    },
+    {
+      "epoch": 0.17869256860931237,
+      "grad_norm": 4.3246259689331055,
+      "learning_rate": 9.431734900581011e-06,
+      "loss": 1.2361,
+      "step": 2318
+    },
+    {
+      "epoch": 0.17876965772432932,
+      "grad_norm": 4.359192371368408,
+      "learning_rate": 9.431156704358318e-06,
+      "loss": 1.0469,
+      "step": 2319
+    },
+    {
+      "epoch": 0.1788467468393463,
+      "grad_norm": 3.730458974838257,
+      "learning_rate": 9.430578231875956e-06,
+      "loss": 0.9637,
+      "step": 2320
+    },
+    {
+      "epoch": 0.17892383595436323,
+      "grad_norm": 4.066617965698242,
+      "learning_rate": 9.429999483169987e-06,
+      "loss": 0.956,
+      "step": 2321
+    },
+    {
+      "epoch": 0.1790009250693802,
+      "grad_norm": 3.660848379135132,
+      "learning_rate": 9.429420458276495e-06,
+      "loss": 1.0173,
+      "step": 2322
+    },
+    {
+      "epoch": 0.17907801418439717,
+      "grad_norm": 4.498156547546387,
+      "learning_rate": 9.428841157231576e-06,
+      "loss": 1.0781,
+      "step": 2323
+    },
+    {
+      "epoch": 0.17915510329941411,
+      "grad_norm": 3.933738946914673,
+      "learning_rate": 9.42826158007135e-06,
+      "loss": 1.1102,
+      "step": 2324
+    },
+    {
+      "epoch": 0.17923219241443109,
+      "grad_norm": 3.9478960037231445,
+      "learning_rate": 9.427681726831948e-06,
+      "loss": 1.0797,
+      "step": 2325
+    },
+    {
+      "epoch": 0.17930928152944806,
+      "grad_norm": 4.165237903594971,
+      "learning_rate": 9.427101597549522e-06,
+      "loss": 1.0487,
+      "step": 2326
+    },
+    {
+      "epoch": 0.179386370644465,
+      "grad_norm": 3.704395294189453,
+      "learning_rate": 9.426521192260239e-06,
+      "loss": 1.0585,
+      "step": 2327
+    },
+    {
+      "epoch": 0.17946345975948197,
+      "grad_norm": 4.005001544952393,
+      "learning_rate": 9.425940511000286e-06,
+      "loss": 1.0563,
+      "step": 2328
+    },
+    {
+      "epoch": 0.1795405488744989,
+      "grad_norm": 3.9118220806121826,
+      "learning_rate": 9.425359553805866e-06,
+      "loss": 1.1235,
+      "step": 2329
+    },
+    {
+      "epoch": 0.17961763798951588,
+      "grad_norm": 4.479408264160156,
+      "learning_rate": 9.424778320713196e-06,
+      "loss": 1.1773,
+      "step": 2330
+    },
+    {
+      "epoch": 0.17969472710453285,
+      "grad_norm": 3.661409854888916,
+      "learning_rate": 9.424196811758515e-06,
+      "loss": 0.9807,
+      "step": 2331
+    },
+    {
+      "epoch": 0.1797718162195498,
+      "grad_norm": 3.719728946685791,
+      "learning_rate": 9.423615026978076e-06,
+      "loss": 1.1761,
+      "step": 2332
+    },
+    {
+      "epoch": 0.17984890533456677,
+      "grad_norm": 4.085165023803711,
+      "learning_rate": 9.42303296640815e-06,
+      "loss": 1.0771,
+      "step": 2333
+    },
+    {
+      "epoch": 0.1799259944495837,
+      "grad_norm": 3.7032485008239746,
+      "learning_rate": 9.422450630085026e-06,
+      "loss": 1.1194,
+      "step": 2334
+    },
+    {
+      "epoch": 0.18000308356460068,
+      "grad_norm": 3.613755941390991,
+      "learning_rate": 9.42186801804501e-06,
+      "loss": 1.009,
+      "step": 2335
+    },
+    {
+      "epoch": 0.18008017267961765,
+      "grad_norm": 3.7317869663238525,
+      "learning_rate": 9.421285130324425e-06,
+      "loss": 1.0114,
+      "step": 2336
+    },
+    {
+      "epoch": 0.1801572617946346,
+      "grad_norm": 3.6703238487243652,
+      "learning_rate": 9.42070196695961e-06,
+      "loss": 0.9959,
+      "step": 2337
+    },
+    {
+      "epoch": 0.18023435090965156,
+      "grad_norm": 3.9614789485931396,
+      "learning_rate": 9.420118527986923e-06,
+      "loss": 0.935,
+      "step": 2338
+    },
+    {
+      "epoch": 0.1803114400246685,
+      "grad_norm": 4.142663478851318,
+      "learning_rate": 9.41953481344274e-06,
+      "loss": 1.1967,
+      "step": 2339
+    },
+    {
+      "epoch": 0.18038852913968548,
+      "grad_norm": 3.7918946743011475,
+      "learning_rate": 9.418950823363446e-06,
+      "loss": 1.0058,
+      "step": 2340
+    },
+    {
+      "epoch": 0.18046561825470245,
+      "grad_norm": 4.344407558441162,
+      "learning_rate": 9.41836655778546e-06,
+      "loss": 0.9986,
+      "step": 2341
+    },
+    {
+      "epoch": 0.1805427073697194,
+      "grad_norm": 3.8436927795410156,
+      "learning_rate": 9.417782016745198e-06,
+      "loss": 1.1001,
+      "step": 2342
+    },
+    {
+      "epoch": 0.18061979648473636,
+      "grad_norm": 4.281997203826904,
+      "learning_rate": 9.41719720027911e-06,
+      "loss": 1.1118,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1806968855997533,
+      "grad_norm": 4.253276824951172,
+      "learning_rate": 9.416612108423653e-06,
+      "loss": 1.0024,
+      "step": 2344
+    },
+    {
+      "epoch": 0.18077397471477027,
+      "grad_norm": 3.944852590560913,
+      "learning_rate": 9.416026741215305e-06,
+      "loss": 1.0365,
+      "step": 2345
+    },
+    {
+      "epoch": 0.18085106382978725,
+      "grad_norm": 3.773387908935547,
+      "learning_rate": 9.415441098690562e-06,
+      "loss": 0.9963,
+      "step": 2346
+    },
+    {
+      "epoch": 0.1809281529448042,
+      "grad_norm": 3.8287198543548584,
+      "learning_rate": 9.414855180885933e-06,
+      "loss": 1.0586,
+      "step": 2347
+    },
+    {
+      "epoch": 0.18100524205982116,
+      "grad_norm": 4.154395580291748,
+      "learning_rate": 9.41426898783795e-06,
+      "loss": 0.9962,
+      "step": 2348
+    },
+    {
+      "epoch": 0.1810823311748381,
+      "grad_norm": 4.041691780090332,
+      "learning_rate": 9.413682519583156e-06,
+      "loss": 1.1823,
+      "step": 2349
+    },
+    {
+      "epoch": 0.18115942028985507,
+      "grad_norm": 4.263957977294922,
+      "learning_rate": 9.413095776158117e-06,
+      "loss": 0.9897,
+      "step": 2350
+    },
+    {
+      "epoch": 0.18123650940487204,
+      "grad_norm": 3.6769165992736816,
+      "learning_rate": 9.412508757599413e-06,
+      "loss": 1.0331,
+      "step": 2351
+    },
+    {
+      "epoch": 0.18131359851988899,
+      "grad_norm": 3.703239679336548,
+      "learning_rate": 9.411921463943641e-06,
+      "loss": 0.9509,
+      "step": 2352
+    },
+    {
+      "epoch": 0.18139068763490596,
+      "grad_norm": 3.7787482738494873,
+      "learning_rate": 9.411333895227415e-06,
+      "loss": 0.9942,
+      "step": 2353
+    },
+    {
+      "epoch": 0.1814677767499229,
+      "grad_norm": 3.9517102241516113,
+      "learning_rate": 9.410746051487367e-06,
+      "loss": 1.0088,
+      "step": 2354
+    },
+    {
+      "epoch": 0.18154486586493987,
+      "grad_norm": 3.5368359088897705,
+      "learning_rate": 9.410157932760148e-06,
+      "loss": 1.0106,
+      "step": 2355
+    },
+    {
+      "epoch": 0.18162195497995684,
+      "grad_norm": 3.456052780151367,
+      "learning_rate": 9.40956953908242e-06,
+      "loss": 1.0061,
+      "step": 2356
+    },
+    {
+      "epoch": 0.18169904409497378,
+      "grad_norm": 3.8283514976501465,
+      "learning_rate": 9.408980870490872e-06,
+      "loss": 1.0836,
+      "step": 2357
+    },
+    {
+      "epoch": 0.18177613320999075,
+      "grad_norm": 3.708845853805542,
+      "learning_rate": 9.4083919270222e-06,
+      "loss": 0.9974,
+      "step": 2358
+    },
+    {
+      "epoch": 0.1818532223250077,
+      "grad_norm": 3.6090118885040283,
+      "learning_rate": 9.407802708713123e-06,
+      "loss": 0.906,
+      "step": 2359
+    },
+    {
+      "epoch": 0.18193031144002467,
+      "grad_norm": 4.260048866271973,
+      "learning_rate": 9.407213215600377e-06,
+      "loss": 1.1036,
+      "step": 2360
+    },
+    {
+      "epoch": 0.18200740055504164,
+      "grad_norm": 4.0702056884765625,
+      "learning_rate": 9.406623447720711e-06,
+      "loss": 1.072,
+      "step": 2361
+    },
+    {
+      "epoch": 0.18208448967005858,
+      "grad_norm": 4.009487152099609,
+      "learning_rate": 9.406033405110896e-06,
+      "loss": 1.0236,
+      "step": 2362
+    },
+    {
+      "epoch": 0.18216157878507555,
+      "grad_norm": 4.346181869506836,
+      "learning_rate": 9.405443087807717e-06,
+      "loss": 0.978,
+      "step": 2363
+    },
+    {
+      "epoch": 0.1822386679000925,
+      "grad_norm": 4.057311534881592,
+      "learning_rate": 9.404852495847979e-06,
+      "loss": 1.13,
+      "step": 2364
+    },
+    {
+      "epoch": 0.18231575701510946,
+      "grad_norm": 4.186524391174316,
+      "learning_rate": 9.4042616292685e-06,
+      "loss": 1.0866,
+      "step": 2365
+    },
+    {
+      "epoch": 0.18239284613012643,
+      "grad_norm": 3.631807565689087,
+      "learning_rate": 9.40367048810612e-06,
+      "loss": 0.9123,
+      "step": 2366
+    },
+    {
+      "epoch": 0.18246993524514338,
+      "grad_norm": 3.841275930404663,
+      "learning_rate": 9.403079072397692e-06,
+      "loss": 1.19,
+      "step": 2367
+    },
+    {
+      "epoch": 0.18254702436016035,
+      "grad_norm": 3.969886541366577,
+      "learning_rate": 9.402487382180088e-06,
+      "loss": 0.9525,
+      "step": 2368
+    },
+    {
+      "epoch": 0.18262411347517732,
+      "grad_norm": 3.8512344360351562,
+      "learning_rate": 9.401895417490197e-06,
+      "loss": 0.9227,
+      "step": 2369
+    },
+    {
+      "epoch": 0.18270120259019426,
+      "grad_norm": 3.7888450622558594,
+      "learning_rate": 9.401303178364923e-06,
+      "loss": 1.0298,
+      "step": 2370
+    },
+    {
+      "epoch": 0.18277829170521123,
+      "grad_norm": 3.4874343872070312,
+      "learning_rate": 9.400710664841192e-06,
+      "loss": 0.997,
+      "step": 2371
+    },
+    {
+      "epoch": 0.18285538082022817,
+      "grad_norm": 3.7443392276763916,
+      "learning_rate": 9.400117876955943e-06,
+      "loss": 1.0684,
+      "step": 2372
+    },
+    {
+      "epoch": 0.18293246993524515,
+      "grad_norm": 4.328792572021484,
+      "learning_rate": 9.399524814746133e-06,
+      "loss": 1.1274,
+      "step": 2373
+    },
+    {
+      "epoch": 0.18300955905026212,
+      "grad_norm": 3.9465620517730713,
+      "learning_rate": 9.398931478248736e-06,
+      "loss": 1.1876,
+      "step": 2374
+    },
+    {
+      "epoch": 0.18308664816527906,
+      "grad_norm": 3.4830267429351807,
+      "learning_rate": 9.398337867500744e-06,
+      "loss": 0.9418,
+      "step": 2375
+    },
+    {
+      "epoch": 0.18316373728029603,
+      "grad_norm": 3.6330952644348145,
+      "learning_rate": 9.397743982539166e-06,
+      "loss": 1.0558,
+      "step": 2376
+    },
+    {
+      "epoch": 0.18324082639531297,
+      "grad_norm": 4.2686076164245605,
+      "learning_rate": 9.397149823401029e-06,
+      "loss": 1.0845,
+      "step": 2377
+    },
+    {
+      "epoch": 0.18331791551032994,
+      "grad_norm": 3.876206636428833,
+      "learning_rate": 9.396555390123371e-06,
+      "loss": 1.0598,
+      "step": 2378
+    },
+    {
+      "epoch": 0.1833950046253469,
+      "grad_norm": 4.2616472244262695,
+      "learning_rate": 9.395960682743255e-06,
+      "loss": 1.1524,
+      "step": 2379
+    },
+    {
+      "epoch": 0.18347209374036386,
+      "grad_norm": 4.4118475914001465,
+      "learning_rate": 9.39536570129776e-06,
+      "loss": 1.2057,
+      "step": 2380
+    },
+    {
+      "epoch": 0.18354918285538083,
+      "grad_norm": 3.960028648376465,
+      "learning_rate": 9.394770445823976e-06,
+      "loss": 1.0039,
+      "step": 2381
+    },
+    {
+      "epoch": 0.18362627197039777,
+      "grad_norm": 3.7380380630493164,
+      "learning_rate": 9.394174916359016e-06,
+      "loss": 1.0101,
+      "step": 2382
+    },
+    {
+      "epoch": 0.18370336108541474,
+      "grad_norm": 4.420310020446777,
+      "learning_rate": 9.393579112940007e-06,
+      "loss": 0.9947,
+      "step": 2383
+    },
+    {
+      "epoch": 0.1837804502004317,
+      "grad_norm": 3.7579898834228516,
+      "learning_rate": 9.392983035604098e-06,
+      "loss": 0.9812,
+      "step": 2384
+    },
+    {
+      "epoch": 0.18385753931544865,
+      "grad_norm": 4.156567573547363,
+      "learning_rate": 9.392386684388446e-06,
+      "loss": 1.1538,
+      "step": 2385
+    },
+    {
+      "epoch": 0.18393462843046562,
+      "grad_norm": 3.7567379474639893,
+      "learning_rate": 9.391790059330234e-06,
+      "loss": 1.0734,
+      "step": 2386
+    },
+    {
+      "epoch": 0.18401171754548257,
+      "grad_norm": 4.441596031188965,
+      "learning_rate": 9.391193160466658e-06,
+      "loss": 1.009,
+      "step": 2387
+    },
+    {
+      "epoch": 0.18408880666049954,
+      "grad_norm": 4.164433479309082,
+      "learning_rate": 9.39059598783493e-06,
+      "loss": 1.1177,
+      "step": 2388
+    },
+    {
+      "epoch": 0.1841658957755165,
+      "grad_norm": 4.159829139709473,
+      "learning_rate": 9.389998541472282e-06,
+      "loss": 1.0572,
+      "step": 2389
+    },
+    {
+      "epoch": 0.18424298489053345,
+      "grad_norm": 4.0909743309021,
+      "learning_rate": 9.389400821415962e-06,
+      "loss": 1.0564,
+      "step": 2390
+    },
+    {
+      "epoch": 0.18432007400555042,
+      "grad_norm": 4.253016471862793,
+      "learning_rate": 9.388802827703231e-06,
+      "loss": 1.162,
+      "step": 2391
+    },
+    {
+      "epoch": 0.18439716312056736,
+      "grad_norm": 4.08793306350708,
+      "learning_rate": 9.388204560371377e-06,
+      "loss": 1.1314,
+      "step": 2392
+    },
+    {
+      "epoch": 0.18447425223558434,
+      "grad_norm": 3.5922398567199707,
+      "learning_rate": 9.387606019457696e-06,
+      "loss": 1.0718,
+      "step": 2393
+    },
+    {
+      "epoch": 0.1845513413506013,
+      "grad_norm": 3.838916063308716,
+      "learning_rate": 9.387007204999503e-06,
+      "loss": 1.1038,
+      "step": 2394
+    },
+    {
+      "epoch": 0.18462843046561825,
+      "grad_norm": 3.8841969966888428,
+      "learning_rate": 9.386408117034131e-06,
+      "loss": 0.9366,
+      "step": 2395
+    },
+    {
+      "epoch": 0.18470551958063522,
+      "grad_norm": 4.599263668060303,
+      "learning_rate": 9.385808755598932e-06,
+      "loss": 1.0969,
+      "step": 2396
+    },
+    {
+      "epoch": 0.18478260869565216,
+      "grad_norm": 3.7928740978240967,
+      "learning_rate": 9.38520912073127e-06,
+      "loss": 1.0684,
+      "step": 2397
+    },
+    {
+      "epoch": 0.18485969781066913,
+      "grad_norm": 4.084198474884033,
+      "learning_rate": 9.384609212468531e-06,
+      "loss": 1.1246,
+      "step": 2398
+    },
+    {
+      "epoch": 0.1849367869256861,
+      "grad_norm": 4.296706199645996,
+      "learning_rate": 9.384009030848118e-06,
+      "loss": 1.0905,
+      "step": 2399
+    },
+    {
+      "epoch": 0.18501387604070305,
+      "grad_norm": 4.311187267303467,
+      "learning_rate": 9.383408575907446e-06,
+      "loss": 1.1047,
+      "step": 2400
+    },
+    {
+      "epoch": 0.18509096515572002,
+      "grad_norm": 3.62276554107666,
+      "learning_rate": 9.382807847683952e-06,
+      "loss": 1.1747,
+      "step": 2401
+    },
+    {
+      "epoch": 0.18516805427073696,
+      "grad_norm": 3.8191843032836914,
+      "learning_rate": 9.382206846215087e-06,
+      "loss": 1.052,
+      "step": 2402
+    },
+    {
+      "epoch": 0.18524514338575393,
+      "grad_norm": 4.004209041595459,
+      "learning_rate": 9.381605571538321e-06,
+      "loss": 0.9742,
+      "step": 2403
+    },
+    {
+      "epoch": 0.1853222325007709,
+      "grad_norm": 3.64969539642334,
+      "learning_rate": 9.381004023691142e-06,
+      "loss": 1.0338,
+      "step": 2404
+    },
+    {
+      "epoch": 0.18539932161578784,
+      "grad_norm": 4.0959978103637695,
+      "learning_rate": 9.38040220271105e-06,
+      "loss": 1.0634,
+      "step": 2405
+    },
+    {
+      "epoch": 0.1854764107308048,
+      "grad_norm": 4.057779788970947,
+      "learning_rate": 9.37980010863557e-06,
+      "loss": 1.0364,
+      "step": 2406
+    },
+    {
+      "epoch": 0.18555349984582176,
+      "grad_norm": 3.817060708999634,
+      "learning_rate": 9.379197741502236e-06,
+      "loss": 1.0794,
+      "step": 2407
+    },
+    {
+      "epoch": 0.18563058896083873,
+      "grad_norm": 4.1733880043029785,
+      "learning_rate": 9.378595101348602e-06,
+      "loss": 1.0477,
+      "step": 2408
+    },
+    {
+      "epoch": 0.1857076780758557,
+      "grad_norm": 3.8468310832977295,
+      "learning_rate": 9.377992188212241e-06,
+      "loss": 0.9913,
+      "step": 2409
+    },
+    {
+      "epoch": 0.18578476719087264,
+      "grad_norm": 4.041358470916748,
+      "learning_rate": 9.377389002130741e-06,
+      "loss": 1.1645,
+      "step": 2410
+    },
+    {
+      "epoch": 0.1858618563058896,
+      "grad_norm": 3.796046733856201,
+      "learning_rate": 9.376785543141709e-06,
+      "loss": 1.1405,
+      "step": 2411
+    },
+    {
+      "epoch": 0.18593894542090658,
+      "grad_norm": 3.9358856678009033,
+      "learning_rate": 9.376181811282764e-06,
+      "loss": 1.073,
+      "step": 2412
+    },
+    {
+      "epoch": 0.18601603453592352,
+      "grad_norm": 4.8310956954956055,
+      "learning_rate": 9.375577806591552e-06,
+      "loss": 1.1654,
+      "step": 2413
+    },
+    {
+      "epoch": 0.1860931236509405,
+      "grad_norm": 4.1156792640686035,
+      "learning_rate": 9.374973529105722e-06,
+      "loss": 0.9819,
+      "step": 2414
+    },
+    {
+      "epoch": 0.18617021276595744,
+      "grad_norm": 4.034856796264648,
+      "learning_rate": 9.374368978862952e-06,
+      "loss": 1.0477,
+      "step": 2415
+    },
+    {
+      "epoch": 0.1862473018809744,
+      "grad_norm": 4.216608047485352,
+      "learning_rate": 9.373764155900932e-06,
+      "loss": 1.1058,
+      "step": 2416
+    },
+    {
+      "epoch": 0.18632439099599138,
+      "grad_norm": 3.979484796524048,
+      "learning_rate": 9.373159060257368e-06,
+      "loss": 0.9858,
+      "step": 2417
+    },
+    {
+      "epoch": 0.18640148011100832,
+      "grad_norm": 3.889035940170288,
+      "learning_rate": 9.372553691969987e-06,
+      "loss": 0.9886,
+      "step": 2418
+    },
+    {
+      "epoch": 0.1864785692260253,
+      "grad_norm": 3.900179862976074,
+      "learning_rate": 9.37194805107653e-06,
+      "loss": 0.919,
+      "step": 2419
+    },
+    {
+      "epoch": 0.18655565834104224,
+      "grad_norm": 4.21008825302124,
+      "learning_rate": 9.371342137614754e-06,
+      "loss": 1.086,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1866327474560592,
+      "grad_norm": 3.7535738945007324,
+      "learning_rate": 9.370735951622435e-06,
+      "loss": 0.9342,
+      "step": 2421
+    },
+    {
+      "epoch": 0.18670983657107618,
+      "grad_norm": 4.046445369720459,
+      "learning_rate": 9.370129493137367e-06,
+      "loss": 1.1538,
+      "step": 2422
+    },
+    {
+      "epoch": 0.18678692568609312,
+      "grad_norm": 3.904557228088379,
+      "learning_rate": 9.369522762197357e-06,
+      "loss": 1.0304,
+      "step": 2423
+    },
+    {
+      "epoch": 0.1868640148011101,
+      "grad_norm": 3.9920108318328857,
+      "learning_rate": 9.368915758840235e-06,
+      "loss": 1.0235,
+      "step": 2424
+    },
+    {
+      "epoch": 0.18694110391612703,
+      "grad_norm": 4.578022480010986,
+      "learning_rate": 9.36830848310384e-06,
+      "loss": 1.0139,
+      "step": 2425
+    },
+    {
+      "epoch": 0.187018193031144,
+      "grad_norm": 4.43154764175415,
+      "learning_rate": 9.367700935026038e-06,
+      "loss": 1.1063,
+      "step": 2426
+    },
+    {
+      "epoch": 0.18709528214616097,
+      "grad_norm": 3.994178533554077,
+      "learning_rate": 9.367093114644703e-06,
+      "loss": 1.0885,
+      "step": 2427
+    },
+    {
+      "epoch": 0.18717237126117792,
+      "grad_norm": 4.0110015869140625,
+      "learning_rate": 9.366485021997728e-06,
+      "loss": 1.05,
+      "step": 2428
+    },
+    {
+      "epoch": 0.1872494603761949,
+      "grad_norm": 3.9679675102233887,
+      "learning_rate": 9.36587665712303e-06,
+      "loss": 1.0291,
+      "step": 2429
+    },
+    {
+      "epoch": 0.18732654949121183,
+      "grad_norm": 4.2142558097839355,
+      "learning_rate": 9.36526802005853e-06,
+      "loss": 1.0634,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1874036386062288,
+      "grad_norm": 3.840346574783325,
+      "learning_rate": 9.36465911084218e-06,
+      "loss": 1.0937,
+      "step": 2431
+    },
+    {
+      "epoch": 0.18748072772124577,
+      "grad_norm": 3.9532737731933594,
+      "learning_rate": 9.36404992951194e-06,
+      "loss": 1.1248,
+      "step": 2432
+    },
+    {
+      "epoch": 0.18755781683626271,
+      "grad_norm": 3.815847873687744,
+      "learning_rate": 9.363440476105787e-06,
+      "loss": 0.9521,
+      "step": 2433
+    },
+    {
+      "epoch": 0.18763490595127968,
+      "grad_norm": 4.12700080871582,
+      "learning_rate": 9.36283075066172e-06,
+      "loss": 1.0596,
+      "step": 2434
+    },
+    {
+      "epoch": 0.18771199506629663,
+      "grad_norm": 4.737533092498779,
+      "learning_rate": 9.36222075321775e-06,
+      "loss": 1.1245,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1877890841813136,
+      "grad_norm": 3.9905295372009277,
+      "learning_rate": 9.36161048381191e-06,
+      "loss": 1.1027,
+      "step": 2436
+    },
+    {
+      "epoch": 0.18786617329633057,
+      "grad_norm": 4.189011096954346,
+      "learning_rate": 9.360999942482247e-06,
+      "loss": 1.0894,
+      "step": 2437
+    },
+    {
+      "epoch": 0.1879432624113475,
+      "grad_norm": 3.8845021724700928,
+      "learning_rate": 9.360389129266822e-06,
+      "loss": 1.0176,
+      "step": 2438
+    },
+    {
+      "epoch": 0.18802035152636448,
+      "grad_norm": 4.049330234527588,
+      "learning_rate": 9.359778044203718e-06,
+      "loss": 0.8773,
+      "step": 2439
+    },
+    {
+      "epoch": 0.18809744064138142,
+      "grad_norm": 3.8811943531036377,
+      "learning_rate": 9.359166687331032e-06,
+      "loss": 1.0447,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1881745297563984,
+      "grad_norm": 3.700079917907715,
+      "learning_rate": 9.35855505868688e-06,
+      "loss": 0.9359,
+      "step": 2441
+    },
+    {
+      "epoch": 0.18825161887141537,
+      "grad_norm": 4.782196521759033,
+      "learning_rate": 9.357943158309396e-06,
+      "loss": 1.1038,
+      "step": 2442
+    },
+    {
+      "epoch": 0.1883287079864323,
+      "grad_norm": 3.7085964679718018,
+      "learning_rate": 9.357330986236723e-06,
+      "loss": 1.0557,
+      "step": 2443
+    },
+    {
+      "epoch": 0.18840579710144928,
+      "grad_norm": 4.303460597991943,
+      "learning_rate": 9.356718542507032e-06,
+      "loss": 1.0125,
+      "step": 2444
+    },
+    {
+      "epoch": 0.18848288621646622,
+      "grad_norm": 3.9245944023132324,
+      "learning_rate": 9.356105827158505e-06,
+      "loss": 0.9875,
+      "step": 2445
+    },
+    {
+      "epoch": 0.1885599753314832,
+      "grad_norm": 4.061418533325195,
+      "learning_rate": 9.355492840229338e-06,
+      "loss": 1.0002,
+      "step": 2446
+    },
+    {
+      "epoch": 0.18863706444650016,
+      "grad_norm": 4.799796104431152,
+      "learning_rate": 9.354879581757753e-06,
+      "loss": 1.0555,
+      "step": 2447
+    },
+    {
+      "epoch": 0.1887141535615171,
+      "grad_norm": 3.832697629928589,
+      "learning_rate": 9.354266051781978e-06,
+      "loss": 1.1255,
+      "step": 2448
+    },
+    {
+      "epoch": 0.18879124267653408,
+      "grad_norm": 4.201719284057617,
+      "learning_rate": 9.353652250340268e-06,
+      "loss": 0.9662,
+      "step": 2449
+    },
+    {
+      "epoch": 0.18886833179155102,
+      "grad_norm": 3.8492846488952637,
+      "learning_rate": 9.353038177470886e-06,
+      "loss": 1.0201,
+      "step": 2450
+    },
+    {
+      "epoch": 0.188945420906568,
+      "grad_norm": 4.532588005065918,
+      "learning_rate": 9.35242383321212e-06,
+      "loss": 1.0201,
+      "step": 2451
+    },
+    {
+      "epoch": 0.18902251002158496,
+      "grad_norm": 4.5830206871032715,
+      "learning_rate": 9.35180921760227e-06,
+      "loss": 1.0741,
+      "step": 2452
+    },
+    {
+      "epoch": 0.1890995991366019,
+      "grad_norm": 4.1921868324279785,
+      "learning_rate": 9.351194330679653e-06,
+      "loss": 1.0388,
+      "step": 2453
+    },
+    {
+      "epoch": 0.18917668825161887,
+      "grad_norm": 3.923304796218872,
+      "learning_rate": 9.350579172482607e-06,
+      "loss": 1.0602,
+      "step": 2454
+    },
+    {
+      "epoch": 0.18925377736663584,
+      "grad_norm": 4.064269065856934,
+      "learning_rate": 9.349963743049479e-06,
+      "loss": 1.1611,
+      "step": 2455
+    },
+    {
+      "epoch": 0.1893308664816528,
+      "grad_norm": 4.29310417175293,
+      "learning_rate": 9.349348042418643e-06,
+      "loss": 0.9577,
+      "step": 2456
+    },
+    {
+      "epoch": 0.18940795559666976,
+      "grad_norm": 3.83111572265625,
+      "learning_rate": 9.348732070628482e-06,
+      "loss": 1.0647,
+      "step": 2457
+    },
+    {
+      "epoch": 0.1894850447116867,
+      "grad_norm": 4.015953063964844,
+      "learning_rate": 9.348115827717398e-06,
+      "loss": 1.0998,
+      "step": 2458
+    },
+    {
+      "epoch": 0.18956213382670367,
+      "grad_norm": 5.061834335327148,
+      "learning_rate": 9.347499313723811e-06,
+      "loss": 1.0665,
+      "step": 2459
+    },
+    {
+      "epoch": 0.18963922294172064,
+      "grad_norm": 3.875699043273926,
+      "learning_rate": 9.346882528686159e-06,
+      "loss": 1.0707,
+      "step": 2460
+    },
+    {
+      "epoch": 0.18971631205673758,
+      "grad_norm": 3.674454927444458,
+      "learning_rate": 9.346265472642895e-06,
+      "loss": 1.031,
+      "step": 2461
+    },
+    {
+      "epoch": 0.18979340117175456,
+      "grad_norm": 3.886066198348999,
+      "learning_rate": 9.34564814563249e-06,
+      "loss": 1.0683,
+      "step": 2462
+    },
+    {
+      "epoch": 0.1898704902867715,
+      "grad_norm": 3.8345463275909424,
+      "learning_rate": 9.345030547693426e-06,
+      "loss": 1.0606,
+      "step": 2463
+    },
+    {
+      "epoch": 0.18994757940178847,
+      "grad_norm": 4.000880718231201,
+      "learning_rate": 9.344412678864214e-06,
+      "loss": 1.1364,
+      "step": 2464
+    },
+    {
+      "epoch": 0.19002466851680544,
+      "grad_norm": 4.1079607009887695,
+      "learning_rate": 9.34379453918337e-06,
+      "loss": 1.1008,
+      "step": 2465
+    },
+    {
+      "epoch": 0.19010175763182238,
+      "grad_norm": 3.764862537384033,
+      "learning_rate": 9.343176128689434e-06,
+      "loss": 0.9816,
+      "step": 2466
+    },
+    {
+      "epoch": 0.19017884674683935,
+      "grad_norm": 4.217895984649658,
+      "learning_rate": 9.342557447420961e-06,
+      "loss": 0.9649,
+      "step": 2467
+    },
+    {
+      "epoch": 0.1902559358618563,
+      "grad_norm": 4.198402404785156,
+      "learning_rate": 9.34193849541652e-06,
+      "loss": 1.0241,
+      "step": 2468
+    },
+    {
+      "epoch": 0.19033302497687327,
+      "grad_norm": 3.5659427642822266,
+      "learning_rate": 9.341319272714704e-06,
+      "loss": 1.0963,
+      "step": 2469
+    },
+    {
+      "epoch": 0.19041011409189024,
+      "grad_norm": 4.269014358520508,
+      "learning_rate": 9.340699779354114e-06,
+      "loss": 1.1575,
+      "step": 2470
+    },
+    {
+      "epoch": 0.19048720320690718,
+      "grad_norm": 3.8241798877716064,
+      "learning_rate": 9.340080015373374e-06,
+      "loss": 1.0775,
+      "step": 2471
+    },
+    {
+      "epoch": 0.19056429232192415,
+      "grad_norm": 4.083940505981445,
+      "learning_rate": 9.339459980811122e-06,
+      "loss": 1.0798,
+      "step": 2472
+    },
+    {
+      "epoch": 0.1906413814369411,
+      "grad_norm": 3.9632468223571777,
+      "learning_rate": 9.338839675706017e-06,
+      "loss": 1.0339,
+      "step": 2473
+    },
+    {
+      "epoch": 0.19071847055195806,
+      "grad_norm": 3.6100375652313232,
+      "learning_rate": 9.338219100096728e-06,
+      "loss": 1.0676,
+      "step": 2474
+    },
+    {
+      "epoch": 0.19079555966697503,
+      "grad_norm": 3.7622032165527344,
+      "learning_rate": 9.337598254021947e-06,
+      "loss": 1.0779,
+      "step": 2475
+    },
+    {
+      "epoch": 0.19087264878199198,
+      "grad_norm": 3.8564085960388184,
+      "learning_rate": 9.33697713752038e-06,
+      "loss": 1.0088,
+      "step": 2476
+    },
+    {
+      "epoch": 0.19094973789700895,
+      "grad_norm": 4.029922962188721,
+      "learning_rate": 9.33635575063075e-06,
+      "loss": 1.1112,
+      "step": 2477
+    },
+    {
+      "epoch": 0.1910268270120259,
+      "grad_norm": 4.0595502853393555,
+      "learning_rate": 9.335734093391797e-06,
+      "loss": 0.9914,
+      "step": 2478
+    },
+    {
+      "epoch": 0.19110391612704286,
+      "grad_norm": 3.9760942459106445,
+      "learning_rate": 9.335112165842277e-06,
+      "loss": 1.0767,
+      "step": 2479
+    },
+    {
+      "epoch": 0.19118100524205983,
+      "grad_norm": 4.477503776550293,
+      "learning_rate": 9.334489968020968e-06,
+      "loss": 1.0201,
+      "step": 2480
+    },
+    {
+      "epoch": 0.19125809435707677,
+      "grad_norm": 4.229430198669434,
+      "learning_rate": 9.333867499966659e-06,
+      "loss": 1.0431,
+      "step": 2481
+    },
+    {
+      "epoch": 0.19133518347209374,
+      "grad_norm": 3.821248769760132,
+      "learning_rate": 9.333244761718157e-06,
+      "loss": 1.0613,
+      "step": 2482
+    },
+    {
+      "epoch": 0.1914122725871107,
+      "grad_norm": 3.8744101524353027,
+      "learning_rate": 9.332621753314285e-06,
+      "loss": 0.8911,
+      "step": 2483
+    },
+    {
+      "epoch": 0.19148936170212766,
+      "grad_norm": 3.607072114944458,
+      "learning_rate": 9.331998474793886e-06,
+      "loss": 0.9713,
+      "step": 2484
+    },
+    {
+      "epoch": 0.19156645081714463,
+      "grad_norm": 4.2475810050964355,
+      "learning_rate": 9.331374926195819e-06,
+      "loss": 1.076,
+      "step": 2485
+    },
+    {
+      "epoch": 0.19164353993216157,
+      "grad_norm": 4.018882751464844,
+      "learning_rate": 9.330751107558957e-06,
+      "loss": 1.0054,
+      "step": 2486
+    },
+    {
+      "epoch": 0.19172062904717854,
+      "grad_norm": 3.6863954067230225,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.9955,
+      "step": 2487
+    },
+    {
+      "epoch": 0.19179771816219549,
+      "grad_norm": 3.837610960006714,
+      "learning_rate": 9.329502660324437e-06,
+      "loss": 1.0766,
+      "step": 2488
+    },
+    {
+      "epoch": 0.19187480727721246,
+      "grad_norm": 3.6833620071411133,
+      "learning_rate": 9.328878031804613e-06,
+      "loss": 1.0074,
+      "step": 2489
+    },
+    {
+      "epoch": 0.19195189639222943,
+      "grad_norm": 3.849368095397949,
+      "learning_rate": 9.328253133401663e-06,
+      "loss": 1.0552,
+      "step": 2490
+    },
+    {
+      "epoch": 0.19202898550724637,
+      "grad_norm": 4.261739730834961,
+      "learning_rate": 9.327627965154547e-06,
+      "loss": 1.1149,
+      "step": 2491
+    },
+    {
+      "epoch": 0.19210607462226334,
+      "grad_norm": 5.0863938331604,
+      "learning_rate": 9.32700252710224e-06,
+      "loss": 1.1051,
+      "step": 2492
+    },
+    {
+      "epoch": 0.19218316373728028,
+      "grad_norm": 3.6857082843780518,
+      "learning_rate": 9.326376819283737e-06,
+      "loss": 1.0603,
+      "step": 2493
+    },
+    {
+      "epoch": 0.19226025285229725,
+      "grad_norm": 4.171791076660156,
+      "learning_rate": 9.325750841738046e-06,
+      "loss": 0.9976,
+      "step": 2494
+    },
+    {
+      "epoch": 0.19233734196731422,
+      "grad_norm": 3.735818386077881,
+      "learning_rate": 9.325124594504191e-06,
+      "loss": 1.0204,
+      "step": 2495
+    },
+    {
+      "epoch": 0.19241443108233117,
+      "grad_norm": 3.711371660232544,
+      "learning_rate": 9.32449807762122e-06,
+      "loss": 1.0821,
+      "step": 2496
+    },
+    {
+      "epoch": 0.19249152019734814,
+      "grad_norm": 3.6877596378326416,
+      "learning_rate": 9.323871291128192e-06,
+      "loss": 0.9228,
+      "step": 2497
+    },
+    {
+      "epoch": 0.1925686093123651,
+      "grad_norm": 4.088908672332764,
+      "learning_rate": 9.323244235064182e-06,
+      "loss": 1.0792,
+      "step": 2498
+    },
+    {
+      "epoch": 0.19264569842738205,
+      "grad_norm": 3.9644157886505127,
+      "learning_rate": 9.322616909468284e-06,
+      "loss": 1.0787,
+      "step": 2499
+    },
+    {
+      "epoch": 0.19272278754239902,
+      "grad_norm": 3.950051784515381,
+      "learning_rate": 9.321989314379609e-06,
+      "loss": 1.1276,
+      "step": 2500
+    },
+    {
+      "epoch": 0.19279987665741596,
+      "grad_norm": 4.185604095458984,
+      "learning_rate": 9.321361449837286e-06,
+      "loss": 1.0614,
+      "step": 2501
+    },
+    {
+      "epoch": 0.19287696577243293,
+      "grad_norm": 4.12734842300415,
+      "learning_rate": 9.320733315880455e-06,
+      "loss": 1.0167,
+      "step": 2502
+    },
+    {
+      "epoch": 0.1929540548874499,
+      "grad_norm": 3.2868363857269287,
+      "learning_rate": 9.320104912548281e-06,
+      "loss": 0.9841,
+      "step": 2503
+    },
+    {
+      "epoch": 0.19303114400246685,
+      "grad_norm": 3.8465094566345215,
+      "learning_rate": 9.319476239879941e-06,
+      "loss": 0.949,
+      "step": 2504
+    },
+    {
+      "epoch": 0.19310823311748382,
+      "grad_norm": 3.994518995285034,
+      "learning_rate": 9.318847297914627e-06,
+      "loss": 1.0177,
+      "step": 2505
+    },
+    {
+      "epoch": 0.19318532223250076,
+      "grad_norm": 4.2856268882751465,
+      "learning_rate": 9.318218086691553e-06,
+      "loss": 1.1933,
+      "step": 2506
+    },
+    {
+      "epoch": 0.19326241134751773,
+      "grad_norm": 3.8407466411590576,
+      "learning_rate": 9.317588606249945e-06,
+      "loss": 1.0947,
+      "step": 2507
+    },
+    {
+      "epoch": 0.1933395004625347,
+      "grad_norm": 3.695981979370117,
+      "learning_rate": 9.316958856629049e-06,
+      "loss": 1.0531,
+      "step": 2508
+    },
+    {
+      "epoch": 0.19341658957755165,
+      "grad_norm": 4.270875453948975,
+      "learning_rate": 9.316328837868126e-06,
+      "loss": 0.8991,
+      "step": 2509
+    },
+    {
+      "epoch": 0.19349367869256862,
+      "grad_norm": 4.000987529754639,
+      "learning_rate": 9.315698550006456e-06,
+      "loss": 1.0195,
+      "step": 2510
+    },
+    {
+      "epoch": 0.19357076780758556,
+      "grad_norm": 3.681183099746704,
+      "learning_rate": 9.315067993083331e-06,
+      "loss": 1.0773,
+      "step": 2511
+    },
+    {
+      "epoch": 0.19364785692260253,
+      "grad_norm": 5.025501728057861,
+      "learning_rate": 9.314437167138065e-06,
+      "loss": 1.1425,
+      "step": 2512
+    },
+    {
+      "epoch": 0.1937249460376195,
+      "grad_norm": 3.879424810409546,
+      "learning_rate": 9.31380607220999e-06,
+      "loss": 1.0285,
+      "step": 2513
+    },
+    {
+      "epoch": 0.19380203515263644,
+      "grad_norm": 4.076712608337402,
+      "learning_rate": 9.313174708338446e-06,
+      "loss": 1.0066,
+      "step": 2514
+    },
+    {
+      "epoch": 0.1938791242676534,
+      "grad_norm": 3.7475578784942627,
+      "learning_rate": 9.312543075562797e-06,
+      "loss": 1.019,
+      "step": 2515
+    },
+    {
+      "epoch": 0.19395621338267036,
+      "grad_norm": 3.9618778228759766,
+      "learning_rate": 9.311911173922423e-06,
+      "loss": 1.0797,
+      "step": 2516
+    },
+    {
+      "epoch": 0.19403330249768733,
+      "grad_norm": 4.08378267288208,
+      "learning_rate": 9.311279003456719e-06,
+      "loss": 1.0427,
+      "step": 2517
+    },
+    {
+      "epoch": 0.1941103916127043,
+      "grad_norm": 3.681499719619751,
+      "learning_rate": 9.3106465642051e-06,
+      "loss": 1.0812,
+      "step": 2518
+    },
+    {
+      "epoch": 0.19418748072772124,
+      "grad_norm": 3.9023890495300293,
+      "learning_rate": 9.31001385620699e-06,
+      "loss": 1.0172,
+      "step": 2519
+    },
+    {
+      "epoch": 0.1942645698427382,
+      "grad_norm": 3.5285422801971436,
+      "learning_rate": 9.309380879501839e-06,
+      "loss": 1.0496,
+      "step": 2520
+    },
+    {
+      "epoch": 0.19434165895775515,
+      "grad_norm": 3.923809051513672,
+      "learning_rate": 9.30874763412911e-06,
+      "loss": 1.0218,
+      "step": 2521
+    },
+    {
+      "epoch": 0.19441874807277212,
+      "grad_norm": 3.9072353839874268,
+      "learning_rate": 9.30811412012828e-06,
+      "loss": 1.05,
+      "step": 2522
+    },
+    {
+      "epoch": 0.1944958371877891,
+      "grad_norm": 3.4843831062316895,
+      "learning_rate": 9.307480337538847e-06,
+      "loss": 1.0038,
+      "step": 2523
+    },
+    {
+      "epoch": 0.19457292630280604,
+      "grad_norm": 3.578922748565674,
+      "learning_rate": 9.306846286400326e-06,
+      "loss": 1.0023,
+      "step": 2524
+    },
+    {
+      "epoch": 0.194650015417823,
+      "grad_norm": 3.625309705734253,
+      "learning_rate": 9.306211966752243e-06,
+      "loss": 1.0317,
+      "step": 2525
+    },
+    {
+      "epoch": 0.19472710453283995,
+      "grad_norm": 3.810683012008667,
+      "learning_rate": 9.305577378634148e-06,
+      "loss": 1.0276,
+      "step": 2526
+    },
+    {
+      "epoch": 0.19480419364785692,
+      "grad_norm": 4.362732887268066,
+      "learning_rate": 9.3049425220856e-06,
+      "loss": 1.0026,
+      "step": 2527
+    },
+    {
+      "epoch": 0.1948812827628739,
+      "grad_norm": 3.8706610202789307,
+      "learning_rate": 9.304307397146184e-06,
+      "loss": 0.9928,
+      "step": 2528
+    },
+    {
+      "epoch": 0.19495837187789083,
+      "grad_norm": 3.865088701248169,
+      "learning_rate": 9.303672003855495e-06,
+      "loss": 1.0645,
+      "step": 2529
+    },
+    {
+      "epoch": 0.1950354609929078,
+      "grad_norm": 3.7114272117614746,
+      "learning_rate": 9.303036342253143e-06,
+      "loss": 1.003,
+      "step": 2530
+    },
+    {
+      "epoch": 0.19511255010792475,
+      "grad_norm": 3.746708631515503,
+      "learning_rate": 9.302400412378762e-06,
+      "loss": 0.9651,
+      "step": 2531
+    },
+    {
+      "epoch": 0.19518963922294172,
+      "grad_norm": 4.280477046966553,
+      "learning_rate": 9.301764214272e-06,
+      "loss": 0.9485,
+      "step": 2532
+    },
+    {
+      "epoch": 0.1952667283379587,
+      "grad_norm": 3.7413394451141357,
+      "learning_rate": 9.301127747972516e-06,
+      "loss": 1.0762,
+      "step": 2533
+    },
+    {
+      "epoch": 0.19534381745297563,
+      "grad_norm": 4.122095584869385,
+      "learning_rate": 9.300491013519995e-06,
+      "loss": 1.0784,
+      "step": 2534
+    },
+    {
+      "epoch": 0.1954209065679926,
+      "grad_norm": 3.7376046180725098,
+      "learning_rate": 9.29985401095413e-06,
+      "loss": 1.0385,
+      "step": 2535
+    },
+    {
+      "epoch": 0.19549799568300955,
+      "grad_norm": 4.303701877593994,
+      "learning_rate": 9.299216740314639e-06,
+      "loss": 1.0439,
+      "step": 2536
+    },
+    {
+      "epoch": 0.19557508479802652,
+      "grad_norm": 4.109951019287109,
+      "learning_rate": 9.298579201641248e-06,
+      "loss": 1.0927,
+      "step": 2537
+    },
+    {
+      "epoch": 0.1956521739130435,
+      "grad_norm": 4.093001365661621,
+      "learning_rate": 9.29794139497371e-06,
+      "loss": 1.0977,
+      "step": 2538
+    },
+    {
+      "epoch": 0.19572926302806043,
+      "grad_norm": 3.4555115699768066,
+      "learning_rate": 9.297303320351784e-06,
+      "loss": 0.9951,
+      "step": 2539
+    },
+    {
+      "epoch": 0.1958063521430774,
+      "grad_norm": 3.6072354316711426,
+      "learning_rate": 9.296664977815251e-06,
+      "loss": 0.9743,
+      "step": 2540
+    },
+    {
+      "epoch": 0.19588344125809437,
+      "grad_norm": 3.506932020187378,
+      "learning_rate": 9.296026367403912e-06,
+      "loss": 0.9706,
+      "step": 2541
+    },
+    {
+      "epoch": 0.1959605303731113,
+      "grad_norm": 4.099440097808838,
+      "learning_rate": 9.295387489157577e-06,
+      "loss": 1.0866,
+      "step": 2542
+    },
+    {
+      "epoch": 0.19603761948812828,
+      "grad_norm": 3.744800090789795,
+      "learning_rate": 9.29474834311608e-06,
+      "loss": 0.9554,
+      "step": 2543
+    },
+    {
+      "epoch": 0.19611470860314523,
+      "grad_norm": 3.71545672416687,
+      "learning_rate": 9.294108929319266e-06,
+      "loss": 1.0344,
+      "step": 2544
+    },
+    {
+      "epoch": 0.1961917977181622,
+      "grad_norm": 4.267796516418457,
+      "learning_rate": 9.293469247806999e-06,
+      "loss": 1.0125,
+      "step": 2545
+    },
+    {
+      "epoch": 0.19626888683317917,
+      "grad_norm": 3.7793099880218506,
+      "learning_rate": 9.292829298619161e-06,
+      "loss": 1.0101,
+      "step": 2546
+    },
+    {
+      "epoch": 0.1963459759481961,
+      "grad_norm": 3.366549015045166,
+      "learning_rate": 9.292189081795651e-06,
+      "loss": 0.9683,
+      "step": 2547
+    },
+    {
+      "epoch": 0.19642306506321308,
+      "grad_norm": 3.89920973777771,
+      "learning_rate": 9.291548597376382e-06,
+      "loss": 1.0093,
+      "step": 2548
+    },
+    {
+      "epoch": 0.19650015417823002,
+      "grad_norm": 3.8635799884796143,
+      "learning_rate": 9.290907845401283e-06,
+      "loss": 1.0173,
+      "step": 2549
+    },
+    {
+      "epoch": 0.196577243293247,
+      "grad_norm": 3.8625593185424805,
+      "learning_rate": 9.2902668259103e-06,
+      "loss": 1.0594,
+      "step": 2550
+    },
+    {
+      "epoch": 0.19665433240826397,
+      "grad_norm": 3.8342230319976807,
+      "learning_rate": 9.289625538943405e-06,
+      "loss": 0.9789,
+      "step": 2551
+    },
+    {
+      "epoch": 0.1967314215232809,
+      "grad_norm": 3.684582233428955,
+      "learning_rate": 9.288983984540573e-06,
+      "loss": 0.8711,
+      "step": 2552
+    },
+    {
+      "epoch": 0.19680851063829788,
+      "grad_norm": 4.02480936050415,
+      "learning_rate": 9.288342162741803e-06,
+      "loss": 0.9812,
+      "step": 2553
+    },
+    {
+      "epoch": 0.19688559975331482,
+      "grad_norm": 3.966688871383667,
+      "learning_rate": 9.287700073587106e-06,
+      "loss": 0.9101,
+      "step": 2554
+    },
+    {
+      "epoch": 0.1969626888683318,
+      "grad_norm": 3.733137607574463,
+      "learning_rate": 9.28705771711652e-06,
+      "loss": 1.0469,
+      "step": 2555
+    },
+    {
+      "epoch": 0.19703977798334876,
+      "grad_norm": 4.436910629272461,
+      "learning_rate": 9.286415093370087e-06,
+      "loss": 1.0353,
+      "step": 2556
+    },
+    {
+      "epoch": 0.1971168670983657,
+      "grad_norm": 4.32921028137207,
+      "learning_rate": 9.285772202387872e-06,
+      "loss": 1.2546,
+      "step": 2557
+    },
+    {
+      "epoch": 0.19719395621338268,
+      "grad_norm": 4.157625675201416,
+      "learning_rate": 9.285129044209958e-06,
+      "loss": 1.144,
+      "step": 2558
+    },
+    {
+      "epoch": 0.19727104532839962,
+      "grad_norm": 4.597263813018799,
+      "learning_rate": 9.284485618876442e-06,
+      "loss": 1.1378,
+      "step": 2559
+    },
+    {
+      "epoch": 0.1973481344434166,
+      "grad_norm": 3.756883382797241,
+      "learning_rate": 9.283841926427435e-06,
+      "loss": 0.993,
+      "step": 2560
+    },
+    {
+      "epoch": 0.19742522355843356,
+      "grad_norm": 3.802865982055664,
+      "learning_rate": 9.283197966903073e-06,
+      "loss": 1.0976,
+      "step": 2561
+    },
+    {
+      "epoch": 0.1975023126734505,
+      "grad_norm": 3.97096848487854,
+      "learning_rate": 9.282553740343501e-06,
+      "loss": 1.0863,
+      "step": 2562
+    },
+    {
+      "epoch": 0.19757940178846747,
+      "grad_norm": 3.7261781692504883,
+      "learning_rate": 9.28190924678888e-06,
+      "loss": 1.1109,
+      "step": 2563
+    },
+    {
+      "epoch": 0.19765649090348442,
+      "grad_norm": 3.8818700313568115,
+      "learning_rate": 9.281264486279398e-06,
+      "loss": 1.1454,
+      "step": 2564
+    },
+    {
+      "epoch": 0.1977335800185014,
+      "grad_norm": 3.9310739040374756,
+      "learning_rate": 9.280619458855246e-06,
+      "loss": 1.0731,
+      "step": 2565
+    },
+    {
+      "epoch": 0.19781066913351836,
+      "grad_norm": 4.061331272125244,
+      "learning_rate": 9.279974164556643e-06,
+      "loss": 0.9986,
+      "step": 2566
+    },
+    {
+      "epoch": 0.1978877582485353,
+      "grad_norm": 3.422016143798828,
+      "learning_rate": 9.279328603423815e-06,
+      "loss": 1.0443,
+      "step": 2567
+    },
+    {
+      "epoch": 0.19796484736355227,
+      "grad_norm": 4.034252643585205,
+      "learning_rate": 9.278682775497012e-06,
+      "loss": 1.0879,
+      "step": 2568
+    },
+    {
+      "epoch": 0.1980419364785692,
+      "grad_norm": 4.055642127990723,
+      "learning_rate": 9.278036680816497e-06,
+      "loss": 1.1014,
+      "step": 2569
+    },
+    {
+      "epoch": 0.19811902559358618,
+      "grad_norm": 4.593109607696533,
+      "learning_rate": 9.277390319422555e-06,
+      "loss": 1.0524,
+      "step": 2570
+    },
+    {
+      "epoch": 0.19819611470860315,
+      "grad_norm": 4.03125524520874,
+      "learning_rate": 9.276743691355476e-06,
+      "loss": 1.0198,
+      "step": 2571
+    },
+    {
+      "epoch": 0.1982732038236201,
+      "grad_norm": 3.723781108856201,
+      "learning_rate": 9.27609679665558e-06,
+      "loss": 0.9915,
+      "step": 2572
+    },
+    {
+      "epoch": 0.19835029293863707,
+      "grad_norm": 3.6599931716918945,
+      "learning_rate": 9.275449635363195e-06,
+      "loss": 1.0793,
+      "step": 2573
+    },
+    {
+      "epoch": 0.198427382053654,
+      "grad_norm": 3.8307621479034424,
+      "learning_rate": 9.274802207518668e-06,
+      "loss": 1.0124,
+      "step": 2574
+    },
+    {
+      "epoch": 0.19850447116867098,
+      "grad_norm": 3.984130620956421,
+      "learning_rate": 9.274154513162364e-06,
+      "loss": 0.9356,
+      "step": 2575
+    },
+    {
+      "epoch": 0.19858156028368795,
+      "grad_norm": 3.825328826904297,
+      "learning_rate": 9.27350655233466e-06,
+      "loss": 1.0427,
+      "step": 2576
+    },
+    {
+      "epoch": 0.1986586493987049,
+      "grad_norm": 4.283047199249268,
+      "learning_rate": 9.272858325075957e-06,
+      "loss": 1.0344,
+      "step": 2577
+    },
+    {
+      "epoch": 0.19873573851372187,
+      "grad_norm": 3.8551671504974365,
+      "learning_rate": 9.272209831426667e-06,
+      "loss": 1.022,
+      "step": 2578
+    },
+    {
+      "epoch": 0.1988128276287388,
+      "grad_norm": 3.6993603706359863,
+      "learning_rate": 9.271561071427222e-06,
+      "loss": 0.9627,
+      "step": 2579
+    },
+    {
+      "epoch": 0.19888991674375578,
+      "grad_norm": 3.686549186706543,
+      "learning_rate": 9.270912045118065e-06,
+      "loss": 1.1547,
+      "step": 2580
+    },
+    {
+      "epoch": 0.19896700585877275,
+      "grad_norm": 3.775413990020752,
+      "learning_rate": 9.270262752539662e-06,
+      "loss": 1.1395,
+      "step": 2581
+    },
+    {
+      "epoch": 0.1990440949737897,
+      "grad_norm": 3.9551045894622803,
+      "learning_rate": 9.269613193732492e-06,
+      "loss": 1.05,
+      "step": 2582
+    },
+    {
+      "epoch": 0.19912118408880666,
+      "grad_norm": 3.7017507553100586,
+      "learning_rate": 9.268963368737053e-06,
+      "loss": 1.0631,
+      "step": 2583
+    },
+    {
+      "epoch": 0.19919827320382363,
+      "grad_norm": 4.082302093505859,
+      "learning_rate": 9.268313277593859e-06,
+      "loss": 1.177,
+      "step": 2584
+    },
+    {
+      "epoch": 0.19927536231884058,
+      "grad_norm": 3.7110402584075928,
+      "learning_rate": 9.267662920343436e-06,
+      "loss": 1.0281,
+      "step": 2585
+    },
+    {
+      "epoch": 0.19935245143385755,
+      "grad_norm": 4.061951160430908,
+      "learning_rate": 9.267012297026334e-06,
+      "loss": 1.0825,
+      "step": 2586
+    },
+    {
+      "epoch": 0.1994295405488745,
+      "grad_norm": 3.88759708404541,
+      "learning_rate": 9.266361407683115e-06,
+      "loss": 1.0694,
+      "step": 2587
+    },
+    {
+      "epoch": 0.19950662966389146,
+      "grad_norm": 4.016077518463135,
+      "learning_rate": 9.265710252354358e-06,
+      "loss": 1.0718,
+      "step": 2588
+    },
+    {
+      "epoch": 0.19958371877890843,
+      "grad_norm": 3.7587389945983887,
+      "learning_rate": 9.26505883108066e-06,
+      "loss": 1.0342,
+      "step": 2589
+    },
+    {
+      "epoch": 0.19966080789392537,
+      "grad_norm": 3.6690351963043213,
+      "learning_rate": 9.264407143902632e-06,
+      "loss": 1.057,
+      "step": 2590
+    },
+    {
+      "epoch": 0.19973789700894234,
+      "grad_norm": 3.946889638900757,
+      "learning_rate": 9.263755190860904e-06,
+      "loss": 1.0364,
+      "step": 2591
+    },
+    {
+      "epoch": 0.1998149861239593,
+      "grad_norm": 3.688429832458496,
+      "learning_rate": 9.263102971996123e-06,
+      "loss": 1.1029,
+      "step": 2592
+    },
+    {
+      "epoch": 0.19989207523897626,
+      "grad_norm": 3.6999905109405518,
+      "learning_rate": 9.262450487348952e-06,
+      "loss": 1.0277,
+      "step": 2593
+    },
+    {
+      "epoch": 0.19996916435399323,
+      "grad_norm": 3.4279160499572754,
+      "learning_rate": 9.261797736960067e-06,
+      "loss": 0.9504,
+      "step": 2594
+    },
+    {
+      "epoch": 0.20004625346901017,
+      "grad_norm": 3.8050403594970703,
+      "learning_rate": 9.261144720870167e-06,
+      "loss": 0.9475,
+      "step": 2595
+    },
+    {
+      "epoch": 0.20012334258402714,
+      "grad_norm": 4.1190009117126465,
+      "learning_rate": 9.260491439119963e-06,
+      "loss": 1.1579,
+      "step": 2596
+    },
+    {
+      "epoch": 0.20020043169904408,
+      "grad_norm": 3.654827117919922,
+      "learning_rate": 9.259837891750183e-06,
+      "loss": 0.9909,
+      "step": 2597
+    },
+    {
+      "epoch": 0.20027752081406106,
+      "grad_norm": 3.997933864593506,
+      "learning_rate": 9.259184078801571e-06,
+      "loss": 1.1038,
+      "step": 2598
+    },
+    {
+      "epoch": 0.20035460992907803,
+      "grad_norm": 3.865818738937378,
+      "learning_rate": 9.258530000314893e-06,
+      "loss": 0.9885,
+      "step": 2599
+    },
+    {
+      "epoch": 0.20043169904409497,
+      "grad_norm": 3.9201407432556152,
+      "learning_rate": 9.257875656330923e-06,
+      "loss": 1.1399,
+      "step": 2600
+    },
+    {
+      "epoch": 0.20050878815911194,
+      "grad_norm": 3.548663377761841,
+      "learning_rate": 9.257221046890459e-06,
+      "loss": 1.0436,
+      "step": 2601
+    },
+    {
+      "epoch": 0.20058587727412888,
+      "grad_norm": 4.565245628356934,
+      "learning_rate": 9.256566172034312e-06,
+      "loss": 1.0835,
+      "step": 2602
+    },
+    {
+      "epoch": 0.20066296638914585,
+      "grad_norm": 3.7430996894836426,
+      "learning_rate": 9.255911031803308e-06,
+      "loss": 1.1998,
+      "step": 2603
+    },
+    {
+      "epoch": 0.20074005550416282,
+      "grad_norm": 3.844200611114502,
+      "learning_rate": 9.255255626238295e-06,
+      "loss": 1.0305,
+      "step": 2604
+    },
+    {
+      "epoch": 0.20081714461917977,
+      "grad_norm": 3.844482898712158,
+      "learning_rate": 9.25459995538013e-06,
+      "loss": 1.1108,
+      "step": 2605
+    },
+    {
+      "epoch": 0.20089423373419674,
+      "grad_norm": 4.113783836364746,
+      "learning_rate": 9.253944019269695e-06,
+      "loss": 0.9527,
+      "step": 2606
+    },
+    {
+      "epoch": 0.20097132284921368,
+      "grad_norm": 3.4817748069763184,
+      "learning_rate": 9.25328781794788e-06,
+      "loss": 0.9548,
+      "step": 2607
+    },
+    {
+      "epoch": 0.20104841196423065,
+      "grad_norm": 3.778887987136841,
+      "learning_rate": 9.2526313514556e-06,
+      "loss": 1.0612,
+      "step": 2608
+    },
+    {
+      "epoch": 0.20112550107924762,
+      "grad_norm": 4.771206855773926,
+      "learning_rate": 9.251974619833779e-06,
+      "loss": 1.1353,
+      "step": 2609
+    },
+    {
+      "epoch": 0.20120259019426456,
+      "grad_norm": 3.8691132068634033,
+      "learning_rate": 9.251317623123363e-06,
+      "loss": 1.0599,
+      "step": 2610
+    },
+    {
+      "epoch": 0.20127967930928153,
+      "grad_norm": 4.0417070388793945,
+      "learning_rate": 9.25066036136531e-06,
+      "loss": 1.0021,
+      "step": 2611
+    },
+    {
+      "epoch": 0.20135676842429848,
+      "grad_norm": 4.274367809295654,
+      "learning_rate": 9.2500028346006e-06,
+      "loss": 1.0906,
+      "step": 2612
+    },
+    {
+      "epoch": 0.20143385753931545,
+      "grad_norm": 3.9163665771484375,
+      "learning_rate": 9.249345042870222e-06,
+      "loss": 0.9961,
+      "step": 2613
+    },
+    {
+      "epoch": 0.20151094665433242,
+      "grad_norm": 4.183355808258057,
+      "learning_rate": 9.248686986215189e-06,
+      "loss": 1.1059,
+      "step": 2614
+    },
+    {
+      "epoch": 0.20158803576934936,
+      "grad_norm": 3.863225221633911,
+      "learning_rate": 9.248028664676529e-06,
+      "loss": 1.0913,
+      "step": 2615
+    },
+    {
+      "epoch": 0.20166512488436633,
+      "grad_norm": 4.189373016357422,
+      "learning_rate": 9.24737007829528e-06,
+      "loss": 1.094,
+      "step": 2616
+    },
+    {
+      "epoch": 0.20174221399938327,
+      "grad_norm": 3.7555601596832275,
+      "learning_rate": 9.246711227112509e-06,
+      "loss": 1.0754,
+      "step": 2617
+    },
+    {
+      "epoch": 0.20181930311440024,
+      "grad_norm": 3.700612783432007,
+      "learning_rate": 9.246052111169283e-06,
+      "loss": 0.989,
+      "step": 2618
+    },
+    {
+      "epoch": 0.20189639222941722,
+      "grad_norm": 4.023221015930176,
+      "learning_rate": 9.2453927305067e-06,
+      "loss": 1.0806,
+      "step": 2619
+    },
+    {
+      "epoch": 0.20197348134443416,
+      "grad_norm": 3.949838161468506,
+      "learning_rate": 9.244733085165868e-06,
+      "loss": 1.0044,
+      "step": 2620
+    },
+    {
+      "epoch": 0.20205057045945113,
+      "grad_norm": 4.30818510055542,
+      "learning_rate": 9.244073175187912e-06,
+      "loss": 1.0475,
+      "step": 2621
+    },
+    {
+      "epoch": 0.20212765957446807,
+      "grad_norm": 4.544628620147705,
+      "learning_rate": 9.243413000613974e-06,
+      "loss": 1.0492,
+      "step": 2622
+    },
+    {
+      "epoch": 0.20220474868948504,
+      "grad_norm": 3.771693706512451,
+      "learning_rate": 9.242752561485213e-06,
+      "loss": 1.0835,
+      "step": 2623
+    },
+    {
+      "epoch": 0.202281837804502,
+      "grad_norm": 3.973714590072632,
+      "learning_rate": 9.242091857842803e-06,
+      "loss": 1.0409,
+      "step": 2624
+    },
+    {
+      "epoch": 0.20235892691951896,
+      "grad_norm": 4.235926628112793,
+      "learning_rate": 9.241430889727936e-06,
+      "loss": 1.1188,
+      "step": 2625
+    },
+    {
+      "epoch": 0.20243601603453593,
+      "grad_norm": 4.404877185821533,
+      "learning_rate": 9.24076965718182e-06,
+      "loss": 0.9393,
+      "step": 2626
+    },
+    {
+      "epoch": 0.2025131051495529,
+      "grad_norm": 3.5260396003723145,
+      "learning_rate": 9.24010816024568e-06,
+      "loss": 0.9963,
+      "step": 2627
+    },
+    {
+      "epoch": 0.20259019426456984,
+      "grad_norm": 3.725531816482544,
+      "learning_rate": 9.239446398960756e-06,
+      "loss": 0.9743,
+      "step": 2628
+    },
+    {
+      "epoch": 0.2026672833795868,
+      "grad_norm": 3.668250799179077,
+      "learning_rate": 9.238784373368306e-06,
+      "loss": 1.0292,
+      "step": 2629
+    },
+    {
+      "epoch": 0.20274437249460375,
+      "grad_norm": 4.116795539855957,
+      "learning_rate": 9.238122083509602e-06,
+      "loss": 1.1924,
+      "step": 2630
+    },
+    {
+      "epoch": 0.20282146160962072,
+      "grad_norm": 3.9489383697509766,
+      "learning_rate": 9.237459529425938e-06,
+      "loss": 1.0364,
+      "step": 2631
+    },
+    {
+      "epoch": 0.2028985507246377,
+      "grad_norm": 3.3951849937438965,
+      "learning_rate": 9.236796711158617e-06,
+      "loss": 0.9725,
+      "step": 2632
+    },
+    {
+      "epoch": 0.20297563983965464,
+      "grad_norm": 3.6997361183166504,
+      "learning_rate": 9.236133628748965e-06,
+      "loss": 1.0287,
+      "step": 2633
+    },
+    {
+      "epoch": 0.2030527289546716,
+      "grad_norm": 3.9400417804718018,
+      "learning_rate": 9.235470282238322e-06,
+      "loss": 1.1658,
+      "step": 2634
+    },
+    {
+      "epoch": 0.20312981806968855,
+      "grad_norm": 4.31356954574585,
+      "learning_rate": 9.23480667166804e-06,
+      "loss": 1.0993,
+      "step": 2635
+    },
+    {
+      "epoch": 0.20320690718470552,
+      "grad_norm": 3.9322855472564697,
+      "learning_rate": 9.234142797079496e-06,
+      "loss": 0.9019,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2032839962997225,
+      "grad_norm": 4.056495666503906,
+      "learning_rate": 9.233478658514078e-06,
+      "loss": 1.0006,
+      "step": 2637
+    },
+    {
+      "epoch": 0.20336108541473943,
+      "grad_norm": 3.7447257041931152,
+      "learning_rate": 9.232814256013192e-06,
+      "loss": 1.009,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2034381745297564,
+      "grad_norm": 3.8734138011932373,
+      "learning_rate": 9.232149589618257e-06,
+      "loss": 0.9408,
+      "step": 2639
+    },
+    {
+      "epoch": 0.20351526364477335,
+      "grad_norm": 3.8473434448242188,
+      "learning_rate": 9.231484659370717e-06,
+      "loss": 1.0765,
+      "step": 2640
+    },
+    {
+      "epoch": 0.20359235275979032,
+      "grad_norm": 3.7293701171875,
+      "learning_rate": 9.230819465312022e-06,
+      "loss": 1.1238,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2036694418748073,
+      "grad_norm": 3.6699728965759277,
+      "learning_rate": 9.230154007483646e-06,
+      "loss": 0.968,
+      "step": 2642
+    },
+    {
+      "epoch": 0.20374653098982423,
+      "grad_norm": 3.8388984203338623,
+      "learning_rate": 9.229488285927077e-06,
+      "loss": 0.9983,
+      "step": 2643
+    },
+    {
+      "epoch": 0.2038236201048412,
+      "grad_norm": 4.062560558319092,
+      "learning_rate": 9.228822300683817e-06,
+      "loss": 1.0297,
+      "step": 2644
+    },
+    {
+      "epoch": 0.20390070921985815,
+      "grad_norm": 3.8813700675964355,
+      "learning_rate": 9.228156051795388e-06,
+      "loss": 1.1485,
+      "step": 2645
+    },
+    {
+      "epoch": 0.20397779833487512,
+      "grad_norm": 3.9767556190490723,
+      "learning_rate": 9.227489539303329e-06,
+      "loss": 0.9742,
+      "step": 2646
+    },
+    {
+      "epoch": 0.2040548874498921,
+      "grad_norm": 4.134043216705322,
+      "learning_rate": 9.22682276324919e-06,
+      "loss": 1.0414,
+      "step": 2647
+    },
+    {
+      "epoch": 0.20413197656490903,
+      "grad_norm": 3.7752439975738525,
+      "learning_rate": 9.226155723674543e-06,
+      "loss": 0.9868,
+      "step": 2648
+    },
+    {
+      "epoch": 0.204209065679926,
+      "grad_norm": 3.694692373275757,
+      "learning_rate": 9.225488420620977e-06,
+      "loss": 1.116,
+      "step": 2649
+    },
+    {
+      "epoch": 0.20428615479494294,
+      "grad_norm": 3.801795721054077,
+      "learning_rate": 9.22482085413009e-06,
+      "loss": 1.0275,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2043632439099599,
+      "grad_norm": 3.7427256107330322,
+      "learning_rate": 9.224153024243505e-06,
+      "loss": 1.1208,
+      "step": 2651
+    },
+    {
+      "epoch": 0.20444033302497688,
+      "grad_norm": 4.123345375061035,
+      "learning_rate": 9.223484931002856e-06,
+      "loss": 1.0278,
+      "step": 2652
+    },
+    {
+      "epoch": 0.20451742213999383,
+      "grad_norm": 3.854743003845215,
+      "learning_rate": 9.222816574449796e-06,
+      "loss": 1.1139,
+      "step": 2653
+    },
+    {
+      "epoch": 0.2045945112550108,
+      "grad_norm": 3.620626449584961,
+      "learning_rate": 9.222147954625992e-06,
+      "loss": 1.03,
+      "step": 2654
+    },
+    {
+      "epoch": 0.20467160037002774,
+      "grad_norm": 3.9541471004486084,
+      "learning_rate": 9.22147907157313e-06,
+      "loss": 1.0944,
+      "step": 2655
+    },
+    {
+      "epoch": 0.2047486894850447,
+      "grad_norm": 4.048750400543213,
+      "learning_rate": 9.220809925332911e-06,
+      "loss": 1.08,
+      "step": 2656
+    },
+    {
+      "epoch": 0.20482577860006168,
+      "grad_norm": 3.7369632720947266,
+      "learning_rate": 9.220140515947056e-06,
+      "loss": 0.989,
+      "step": 2657
+    },
+    {
+      "epoch": 0.20490286771507862,
+      "grad_norm": 4.2197489738464355,
+      "learning_rate": 9.219470843457294e-06,
+      "loss": 1.1959,
+      "step": 2658
+    },
+    {
+      "epoch": 0.2049799568300956,
+      "grad_norm": 3.825822591781616,
+      "learning_rate": 9.218800907905379e-06,
+      "loss": 1.026,
+      "step": 2659
+    },
+    {
+      "epoch": 0.20505704594511254,
+      "grad_norm": 3.7242703437805176,
+      "learning_rate": 9.218130709333077e-06,
+      "loss": 1.0186,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2051341350601295,
+      "grad_norm": 3.669600248336792,
+      "learning_rate": 9.217460247782174e-06,
+      "loss": 1.0269,
+      "step": 2661
+    },
+    {
+      "epoch": 0.20521122417514648,
+      "grad_norm": 4.220530986785889,
+      "learning_rate": 9.216789523294462e-06,
+      "loss": 1.1483,
+      "step": 2662
+    },
+    {
+      "epoch": 0.20528831329016342,
+      "grad_norm": 3.4208221435546875,
+      "learning_rate": 9.216118535911766e-06,
+      "loss": 0.8748,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2053654024051804,
+      "grad_norm": 4.243244647979736,
+      "learning_rate": 9.215447285675916e-06,
+      "loss": 1.18,
+      "step": 2664
+    },
+    {
+      "epoch": 0.20544249152019733,
+      "grad_norm": 3.644279718399048,
+      "learning_rate": 9.214775772628759e-06,
+      "loss": 1.042,
+      "step": 2665
+    },
+    {
+      "epoch": 0.2055195806352143,
+      "grad_norm": 4.032573223114014,
+      "learning_rate": 9.21410399681216e-06,
+      "loss": 1.1426,
+      "step": 2666
+    },
+    {
+      "epoch": 0.20559666975023128,
+      "grad_norm": 3.6652822494506836,
+      "learning_rate": 9.213431958268004e-06,
+      "loss": 0.9583,
+      "step": 2667
+    },
+    {
+      "epoch": 0.20567375886524822,
+      "grad_norm": 3.9682374000549316,
+      "learning_rate": 9.212759657038186e-06,
+      "loss": 0.9783,
+      "step": 2668
+    },
+    {
+      "epoch": 0.2057508479802652,
+      "grad_norm": 3.6445670127868652,
+      "learning_rate": 9.212087093164624e-06,
+      "loss": 0.9417,
+      "step": 2669
+    },
+    {
+      "epoch": 0.20582793709528216,
+      "grad_norm": 3.758375406265259,
+      "learning_rate": 9.211414266689245e-06,
+      "loss": 1.0221,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2059050262102991,
+      "grad_norm": 4.022796154022217,
+      "learning_rate": 9.210741177653997e-06,
+      "loss": 1.107,
+      "step": 2671
+    },
+    {
+      "epoch": 0.20598211532531607,
+      "grad_norm": 3.7436466217041016,
+      "learning_rate": 9.210067826100845e-06,
+      "loss": 1.0894,
+      "step": 2672
+    },
+    {
+      "epoch": 0.20605920444033302,
+      "grad_norm": 4.940155982971191,
+      "learning_rate": 9.20939421207177e-06,
+      "loss": 1.0682,
+      "step": 2673
+    },
+    {
+      "epoch": 0.20613629355535,
+      "grad_norm": 3.8470208644866943,
+      "learning_rate": 9.208720335608767e-06,
+      "loss": 1.0735,
+      "step": 2674
+    },
+    {
+      "epoch": 0.20621338267036696,
+      "grad_norm": 3.703685998916626,
+      "learning_rate": 9.208046196753848e-06,
+      "loss": 1.0533,
+      "step": 2675
+    },
+    {
+      "epoch": 0.2062904717853839,
+      "grad_norm": 4.3979291915893555,
+      "learning_rate": 9.207371795549043e-06,
+      "loss": 1.2048,
+      "step": 2676
+    },
+    {
+      "epoch": 0.20636756090040087,
+      "grad_norm": 3.9959652423858643,
+      "learning_rate": 9.206697132036395e-06,
+      "loss": 1.01,
+      "step": 2677
+    },
+    {
+      "epoch": 0.2064446500154178,
+      "grad_norm": 3.741591453552246,
+      "learning_rate": 9.206022206257969e-06,
+      "loss": 0.971,
+      "step": 2678
+    },
+    {
+      "epoch": 0.20652173913043478,
+      "grad_norm": 3.790801763534546,
+      "learning_rate": 9.205347018255844e-06,
+      "loss": 1.2011,
+      "step": 2679
+    },
+    {
+      "epoch": 0.20659882824545175,
+      "grad_norm": 4.108438491821289,
+      "learning_rate": 9.20467156807211e-06,
+      "loss": 1.0247,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2066759173604687,
+      "grad_norm": 3.6132218837738037,
+      "learning_rate": 9.203995855748882e-06,
+      "loss": 1.0329,
+      "step": 2681
+    },
+    {
+      "epoch": 0.20675300647548567,
+      "grad_norm": 3.6847808361053467,
+      "learning_rate": 9.203319881328284e-06,
+      "loss": 1.0839,
+      "step": 2682
+    },
+    {
+      "epoch": 0.2068300955905026,
+      "grad_norm": 3.7520806789398193,
+      "learning_rate": 9.202643644852461e-06,
+      "loss": 1.0236,
+      "step": 2683
+    },
+    {
+      "epoch": 0.20690718470551958,
+      "grad_norm": 3.8004398345947266,
+      "learning_rate": 9.201967146363572e-06,
+      "loss": 1.0043,
+      "step": 2684
+    },
+    {
+      "epoch": 0.20698427382053655,
+      "grad_norm": 3.612555503845215,
+      "learning_rate": 9.201290385903796e-06,
+      "loss": 1.0641,
+      "step": 2685
+    },
+    {
+      "epoch": 0.2070613629355535,
+      "grad_norm": 3.6692962646484375,
+      "learning_rate": 9.200613363515325e-06,
+      "loss": 0.9243,
+      "step": 2686
+    },
+    {
+      "epoch": 0.20713845205057047,
+      "grad_norm": 3.6443002223968506,
+      "learning_rate": 9.199936079240364e-06,
+      "loss": 1.0226,
+      "step": 2687
+    },
+    {
+      "epoch": 0.2072155411655874,
+      "grad_norm": 4.001389503479004,
+      "learning_rate": 9.19925853312114e-06,
+      "loss": 1.092,
+      "step": 2688
+    },
+    {
+      "epoch": 0.20729263028060438,
+      "grad_norm": 4.218007564544678,
+      "learning_rate": 9.198580725199894e-06,
+      "loss": 1.0681,
+      "step": 2689
+    },
+    {
+      "epoch": 0.20736971939562135,
+      "grad_norm": 4.421291351318359,
+      "learning_rate": 9.197902655518887e-06,
+      "loss": 1.092,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2074468085106383,
+      "grad_norm": 3.6270670890808105,
+      "learning_rate": 9.197224324120391e-06,
+      "loss": 0.9676,
+      "step": 2691
+    },
+    {
+      "epoch": 0.20752389762565526,
+      "grad_norm": 3.7732226848602295,
+      "learning_rate": 9.196545731046695e-06,
+      "loss": 0.9676,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2076009867406722,
+      "grad_norm": 4.031553268432617,
+      "learning_rate": 9.195866876340107e-06,
+      "loss": 1.0286,
+      "step": 2693
+    },
+    {
+      "epoch": 0.20767807585568918,
+      "grad_norm": 3.7905490398406982,
+      "learning_rate": 9.195187760042952e-06,
+      "loss": 1.0421,
+      "step": 2694
+    },
+    {
+      "epoch": 0.20775516497070615,
+      "grad_norm": 3.822108745574951,
+      "learning_rate": 9.194508382197566e-06,
+      "loss": 1.0981,
+      "step": 2695
+    },
+    {
+      "epoch": 0.2078322540857231,
+      "grad_norm": 3.774367332458496,
+      "learning_rate": 9.193828742846307e-06,
+      "loss": 1.0733,
+      "step": 2696
+    },
+    {
+      "epoch": 0.20790934320074006,
+      "grad_norm": 4.311622619628906,
+      "learning_rate": 9.193148842031548e-06,
+      "loss": 0.9735,
+      "step": 2697
+    },
+    {
+      "epoch": 0.207986432315757,
+      "grad_norm": 3.389493465423584,
+      "learning_rate": 9.192468679795671e-06,
+      "loss": 0.9759,
+      "step": 2698
+    },
+    {
+      "epoch": 0.20806352143077397,
+      "grad_norm": 4.131453990936279,
+      "learning_rate": 9.19178825618109e-06,
+      "loss": 1.011,
+      "step": 2699
+    },
+    {
+      "epoch": 0.20814061054579094,
+      "grad_norm": 3.75632643699646,
+      "learning_rate": 9.191107571230217e-06,
+      "loss": 0.9215,
+      "step": 2700
+    },
+    {
+      "epoch": 0.2082176996608079,
+      "grad_norm": 3.789395332336426,
+      "learning_rate": 9.190426624985497e-06,
+      "loss": 1.1335,
+      "step": 2701
+    },
+    {
+      "epoch": 0.20829478877582486,
+      "grad_norm": 4.15241003036499,
+      "learning_rate": 9.189745417489378e-06,
+      "loss": 1.1109,
+      "step": 2702
+    },
+    {
+      "epoch": 0.2083718778908418,
+      "grad_norm": 3.8731484413146973,
+      "learning_rate": 9.189063948784332e-06,
+      "loss": 0.9326,
+      "step": 2703
+    },
+    {
+      "epoch": 0.20844896700585877,
+      "grad_norm": 3.67352294921875,
+      "learning_rate": 9.188382218912844e-06,
+      "loss": 1.058,
+      "step": 2704
+    },
+    {
+      "epoch": 0.20852605612087574,
+      "grad_norm": 3.72332501411438,
+      "learning_rate": 9.187700227917415e-06,
+      "loss": 1.0898,
+      "step": 2705
+    },
+    {
+      "epoch": 0.20860314523589268,
+      "grad_norm": 3.6140379905700684,
+      "learning_rate": 9.187017975840568e-06,
+      "loss": 1.0351,
+      "step": 2706
+    },
+    {
+      "epoch": 0.20868023435090965,
+      "grad_norm": 3.6925249099731445,
+      "learning_rate": 9.186335462724834e-06,
+      "loss": 1.0802,
+      "step": 2707
+    },
+    {
+      "epoch": 0.2087573234659266,
+      "grad_norm": 3.8140275478363037,
+      "learning_rate": 9.185652688612766e-06,
+      "loss": 1.0149,
+      "step": 2708
+    },
+    {
+      "epoch": 0.20883441258094357,
+      "grad_norm": 3.920686960220337,
+      "learning_rate": 9.184969653546932e-06,
+      "loss": 1.1498,
+      "step": 2709
+    },
+    {
+      "epoch": 0.20891150169596054,
+      "grad_norm": 3.6589865684509277,
+      "learning_rate": 9.184286357569913e-06,
+      "loss": 0.9951,
+      "step": 2710
+    },
+    {
+      "epoch": 0.20898859081097748,
+      "grad_norm": 3.9114480018615723,
+      "learning_rate": 9.183602800724311e-06,
+      "loss": 1.1772,
+      "step": 2711
+    },
+    {
+      "epoch": 0.20906567992599445,
+      "grad_norm": 3.792837381362915,
+      "learning_rate": 9.182918983052743e-06,
+      "loss": 0.9794,
+      "step": 2712
+    },
+    {
+      "epoch": 0.20914276904101142,
+      "grad_norm": 3.940286636352539,
+      "learning_rate": 9.182234904597838e-06,
+      "loss": 1.0731,
+      "step": 2713
+    },
+    {
+      "epoch": 0.20921985815602837,
+      "grad_norm": 3.8480448722839355,
+      "learning_rate": 9.181550565402248e-06,
+      "loss": 1.1191,
+      "step": 2714
+    },
+    {
+      "epoch": 0.20929694727104534,
+      "grad_norm": 3.821777820587158,
+      "learning_rate": 9.180865965508638e-06,
+      "loss": 0.9818,
+      "step": 2715
+    },
+    {
+      "epoch": 0.20937403638606228,
+      "grad_norm": 4.0068511962890625,
+      "learning_rate": 9.180181104959686e-06,
+      "loss": 1.0953,
+      "step": 2716
+    },
+    {
+      "epoch": 0.20945112550107925,
+      "grad_norm": 3.940507650375366,
+      "learning_rate": 9.179495983798094e-06,
+      "loss": 1.07,
+      "step": 2717
+    },
+    {
+      "epoch": 0.20952821461609622,
+      "grad_norm": 4.213906288146973,
+      "learning_rate": 9.178810602066575e-06,
+      "loss": 1.0311,
+      "step": 2718
+    },
+    {
+      "epoch": 0.20960530373111316,
+      "grad_norm": 3.6800248622894287,
+      "learning_rate": 9.178124959807854e-06,
+      "loss": 0.9295,
+      "step": 2719
+    },
+    {
+      "epoch": 0.20968239284613013,
+      "grad_norm": 3.5638949871063232,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 0.9917,
+      "step": 2720
+    },
+    {
+      "epoch": 0.20975948196114708,
+      "grad_norm": 4.166075229644775,
+      "learning_rate": 9.17675289387982e-06,
+      "loss": 1.0279,
+      "step": 2721
+    },
+    {
+      "epoch": 0.20983657107616405,
+      "grad_norm": 3.7445201873779297,
+      "learning_rate": 9.17606647029605e-06,
+      "loss": 1.0408,
+      "step": 2722
+    },
+    {
+      "epoch": 0.20991366019118102,
+      "grad_norm": 3.632784128189087,
+      "learning_rate": 9.175379786356162e-06,
+      "loss": 1.0679,
+      "step": 2723
+    },
+    {
+      "epoch": 0.20999074930619796,
+      "grad_norm": 3.969790458679199,
+      "learning_rate": 9.174692842102968e-06,
+      "loss": 1.1247,
+      "step": 2724
+    },
+    {
+      "epoch": 0.21006783842121493,
+      "grad_norm": 3.7024993896484375,
+      "learning_rate": 9.174005637579297e-06,
+      "loss": 1.0884,
+      "step": 2725
+    },
+    {
+      "epoch": 0.21014492753623187,
+      "grad_norm": 4.002373695373535,
+      "learning_rate": 9.173318172827994e-06,
+      "loss": 1.1234,
+      "step": 2726
+    },
+    {
+      "epoch": 0.21022201665124884,
+      "grad_norm": 3.7659733295440674,
+      "learning_rate": 9.172630447891915e-06,
+      "loss": 1.0724,
+      "step": 2727
+    },
+    {
+      "epoch": 0.21029910576626581,
+      "grad_norm": 4.004013538360596,
+      "learning_rate": 9.171942462813938e-06,
+      "loss": 1.0574,
+      "step": 2728
+    },
+    {
+      "epoch": 0.21037619488128276,
+      "grad_norm": 3.9406075477600098,
+      "learning_rate": 9.171254217636958e-06,
+      "loss": 1.1462,
+      "step": 2729
+    },
+    {
+      "epoch": 0.21045328399629973,
+      "grad_norm": 4.120555877685547,
+      "learning_rate": 9.17056571240388e-06,
+      "loss": 1.0335,
+      "step": 2730
+    },
+    {
+      "epoch": 0.21053037311131667,
+      "grad_norm": 3.955146312713623,
+      "learning_rate": 9.169876947157628e-06,
+      "loss": 1.0698,
+      "step": 2731
+    },
+    {
+      "epoch": 0.21060746222633364,
+      "grad_norm": 4.344998836517334,
+      "learning_rate": 9.169187921941148e-06,
+      "loss": 1.1751,
+      "step": 2732
+    },
+    {
+      "epoch": 0.2106845513413506,
+      "grad_norm": 4.642431259155273,
+      "learning_rate": 9.16849863679739e-06,
+      "loss": 1.1332,
+      "step": 2733
+    },
+    {
+      "epoch": 0.21076164045636755,
+      "grad_norm": 3.788193464279175,
+      "learning_rate": 9.167809091769334e-06,
+      "loss": 1.0431,
+      "step": 2734
+    },
+    {
+      "epoch": 0.21083872957138453,
+      "grad_norm": 3.8150315284729004,
+      "learning_rate": 9.167119286899965e-06,
+      "loss": 1.1172,
+      "step": 2735
+    },
+    {
+      "epoch": 0.21091581868640147,
+      "grad_norm": 4.12971305847168,
+      "learning_rate": 9.16642922223229e-06,
+      "loss": 1.0524,
+      "step": 2736
+    },
+    {
+      "epoch": 0.21099290780141844,
+      "grad_norm": 3.7752163410186768,
+      "learning_rate": 9.165738897809335e-06,
+      "loss": 1.0584,
+      "step": 2737
+    },
+    {
+      "epoch": 0.2110699969164354,
+      "grad_norm": 4.05067253112793,
+      "learning_rate": 9.165048313674131e-06,
+      "loss": 1.1037,
+      "step": 2738
+    },
+    {
+      "epoch": 0.21114708603145235,
+      "grad_norm": 4.372398376464844,
+      "learning_rate": 9.164357469869737e-06,
+      "loss": 1.0896,
+      "step": 2739
+    },
+    {
+      "epoch": 0.21122417514646932,
+      "grad_norm": 3.3885154724121094,
+      "learning_rate": 9.163666366439223e-06,
+      "loss": 0.8929,
+      "step": 2740
+    },
+    {
+      "epoch": 0.21130126426148627,
+      "grad_norm": 3.9432756900787354,
+      "learning_rate": 9.162975003425676e-06,
+      "loss": 1.0253,
+      "step": 2741
+    },
+    {
+      "epoch": 0.21137835337650324,
+      "grad_norm": 3.852996587753296,
+      "learning_rate": 9.162283380872197e-06,
+      "loss": 0.9896,
+      "step": 2742
+    },
+    {
+      "epoch": 0.2114554424915202,
+      "grad_norm": 3.6689140796661377,
+      "learning_rate": 9.161591498821907e-06,
+      "loss": 1.1041,
+      "step": 2743
+    },
+    {
+      "epoch": 0.21153253160653715,
+      "grad_norm": 4.0420403480529785,
+      "learning_rate": 9.16089935731794e-06,
+      "loss": 1.0409,
+      "step": 2744
+    },
+    {
+      "epoch": 0.21160962072155412,
+      "grad_norm": 3.9778621196746826,
+      "learning_rate": 9.160206956403448e-06,
+      "loss": 1.0178,
+      "step": 2745
+    },
+    {
+      "epoch": 0.21168670983657106,
+      "grad_norm": 4.711827278137207,
+      "learning_rate": 9.1595142961216e-06,
+      "loss": 0.9995,
+      "step": 2746
+    },
+    {
+      "epoch": 0.21176379895158803,
+      "grad_norm": 3.6355984210968018,
+      "learning_rate": 9.158821376515574e-06,
+      "loss": 1.0407,
+      "step": 2747
+    },
+    {
+      "epoch": 0.211840888066605,
+      "grad_norm": 4.1716203689575195,
+      "learning_rate": 9.158128197628578e-06,
+      "loss": 0.9243,
+      "step": 2748
+    },
+    {
+      "epoch": 0.21191797718162195,
+      "grad_norm": 3.8837966918945312,
+      "learning_rate": 9.157434759503823e-06,
+      "loss": 1.0692,
+      "step": 2749
+    },
+    {
+      "epoch": 0.21199506629663892,
+      "grad_norm": 4.1938700675964355,
+      "learning_rate": 9.156741062184543e-06,
+      "loss": 1.0286,
+      "step": 2750
+    },
+    {
+      "epoch": 0.21207215541165586,
+      "grad_norm": 3.5614426136016846,
+      "learning_rate": 9.156047105713986e-06,
+      "loss": 1.0052,
+      "step": 2751
+    },
+    {
+      "epoch": 0.21214924452667283,
+      "grad_norm": 3.653048276901245,
+      "learning_rate": 9.155352890135417e-06,
+      "loss": 1.0898,
+      "step": 2752
+    },
+    {
+      "epoch": 0.2122263336416898,
+      "grad_norm": 3.959932327270508,
+      "learning_rate": 9.154658415492116e-06,
+      "loss": 1.1145,
+      "step": 2753
+    },
+    {
+      "epoch": 0.21230342275670674,
+      "grad_norm": 3.750103235244751,
+      "learning_rate": 9.153963681827382e-06,
+      "loss": 1.018,
+      "step": 2754
+    },
+    {
+      "epoch": 0.21238051187172372,
+      "grad_norm": 4.643415451049805,
+      "learning_rate": 9.153268689184524e-06,
+      "loss": 1.0096,
+      "step": 2755
+    },
+    {
+      "epoch": 0.21245760098674069,
+      "grad_norm": 4.112432956695557,
+      "learning_rate": 9.152573437606874e-06,
+      "loss": 1.1197,
+      "step": 2756
+    },
+    {
+      "epoch": 0.21253469010175763,
+      "grad_norm": 3.4916539192199707,
+      "learning_rate": 9.151877927137778e-06,
+      "loss": 1.0737,
+      "step": 2757
+    },
+    {
+      "epoch": 0.2126117792167746,
+      "grad_norm": 3.6449830532073975,
+      "learning_rate": 9.151182157820595e-06,
+      "loss": 1.031,
+      "step": 2758
+    },
+    {
+      "epoch": 0.21268886833179154,
+      "grad_norm": 3.4803693294525146,
+      "learning_rate": 9.150486129698706e-06,
+      "loss": 1.0208,
+      "step": 2759
+    },
+    {
+      "epoch": 0.2127659574468085,
+      "grad_norm": 4.427255153656006,
+      "learning_rate": 9.1497898428155e-06,
+      "loss": 1.1607,
+      "step": 2760
+    },
+    {
+      "epoch": 0.21284304656182548,
+      "grad_norm": 3.903338670730591,
+      "learning_rate": 9.149093297214392e-06,
+      "loss": 1.0678,
+      "step": 2761
+    },
+    {
+      "epoch": 0.21292013567684243,
+      "grad_norm": 3.8415660858154297,
+      "learning_rate": 9.148396492938806e-06,
+      "loss": 1.0445,
+      "step": 2762
+    },
+    {
+      "epoch": 0.2129972247918594,
+      "grad_norm": 4.833527565002441,
+      "learning_rate": 9.14769943003218e-06,
+      "loss": 1.2442,
+      "step": 2763
+    },
+    {
+      "epoch": 0.21307431390687634,
+      "grad_norm": 3.9709668159484863,
+      "learning_rate": 9.14700210853798e-06,
+      "loss": 1.0287,
+      "step": 2764
+    },
+    {
+      "epoch": 0.2131514030218933,
+      "grad_norm": 4.086493015289307,
+      "learning_rate": 9.146304528499674e-06,
+      "loss": 0.9555,
+      "step": 2765
+    },
+    {
+      "epoch": 0.21322849213691028,
+      "grad_norm": 3.9784085750579834,
+      "learning_rate": 9.145606689960756e-06,
+      "loss": 1.0339,
+      "step": 2766
+    },
+    {
+      "epoch": 0.21330558125192722,
+      "grad_norm": 4.03312873840332,
+      "learning_rate": 9.144908592964732e-06,
+      "loss": 1.0068,
+      "step": 2767
+    },
+    {
+      "epoch": 0.2133826703669442,
+      "grad_norm": 4.233692646026611,
+      "learning_rate": 9.144210237555124e-06,
+      "loss": 1.1303,
+      "step": 2768
+    },
+    {
+      "epoch": 0.21345975948196114,
+      "grad_norm": 4.059010982513428,
+      "learning_rate": 9.143511623775469e-06,
+      "loss": 0.9926,
+      "step": 2769
+    },
+    {
+      "epoch": 0.2135368485969781,
+      "grad_norm": 3.532331705093384,
+      "learning_rate": 9.142812751669327e-06,
+      "loss": 0.9392,
+      "step": 2770
+    },
+    {
+      "epoch": 0.21361393771199508,
+      "grad_norm": 3.3265960216522217,
+      "learning_rate": 9.142113621280265e-06,
+      "loss": 0.9405,
+      "step": 2771
+    },
+    {
+      "epoch": 0.21369102682701202,
+      "grad_norm": 3.408034324645996,
+      "learning_rate": 9.14141423265187e-06,
+      "loss": 0.9525,
+      "step": 2772
+    },
+    {
+      "epoch": 0.213768115942029,
+      "grad_norm": 3.924441337585449,
+      "learning_rate": 9.140714585827748e-06,
+      "loss": 1.0772,
+      "step": 2773
+    },
+    {
+      "epoch": 0.21384520505704593,
+      "grad_norm": 3.8719756603240967,
+      "learning_rate": 9.140014680851516e-06,
+      "loss": 0.9904,
+      "step": 2774
+    },
+    {
+      "epoch": 0.2139222941720629,
+      "grad_norm": 3.7581069469451904,
+      "learning_rate": 9.139314517766811e-06,
+      "loss": 1.0082,
+      "step": 2775
+    },
+    {
+      "epoch": 0.21399938328707988,
+      "grad_norm": 3.611703872680664,
+      "learning_rate": 9.138614096617285e-06,
+      "loss": 0.9756,
+      "step": 2776
+    },
+    {
+      "epoch": 0.21407647240209682,
+      "grad_norm": 3.581341505050659,
+      "learning_rate": 9.137913417446603e-06,
+      "loss": 1.0092,
+      "step": 2777
+    },
+    {
+      "epoch": 0.2141535615171138,
+      "grad_norm": 3.8926100730895996,
+      "learning_rate": 9.137212480298451e-06,
+      "loss": 1.0584,
+      "step": 2778
+    },
+    {
+      "epoch": 0.21423065063213073,
+      "grad_norm": 3.6705634593963623,
+      "learning_rate": 9.136511285216527e-06,
+      "loss": 1.0492,
+      "step": 2779
+    },
+    {
+      "epoch": 0.2143077397471477,
+      "grad_norm": 3.6244938373565674,
+      "learning_rate": 9.13580983224455e-06,
+      "loss": 0.9325,
+      "step": 2780
+    },
+    {
+      "epoch": 0.21438482886216467,
+      "grad_norm": 3.5282609462738037,
+      "learning_rate": 9.135108121426247e-06,
+      "loss": 0.9749,
+      "step": 2781
+    },
+    {
+      "epoch": 0.21446191797718162,
+      "grad_norm": 4.567509651184082,
+      "learning_rate": 9.13440615280537e-06,
+      "loss": 1.0813,
+      "step": 2782
+    },
+    {
+      "epoch": 0.21453900709219859,
+      "grad_norm": 3.8714842796325684,
+      "learning_rate": 9.133703926425683e-06,
+      "loss": 1.0343,
+      "step": 2783
+    },
+    {
+      "epoch": 0.21461609620721553,
+      "grad_norm": 3.845766305923462,
+      "learning_rate": 9.133001442330964e-06,
+      "loss": 1.0322,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2146931853222325,
+      "grad_norm": 3.7517282962799072,
+      "learning_rate": 9.132298700565011e-06,
+      "loss": 1.0577,
+      "step": 2785
+    },
+    {
+      "epoch": 0.21477027443724947,
+      "grad_norm": 3.8736910820007324,
+      "learning_rate": 9.131595701171636e-06,
+      "loss": 1.0193,
+      "step": 2786
+    },
+    {
+      "epoch": 0.2148473635522664,
+      "grad_norm": 3.8819291591644287,
+      "learning_rate": 9.130892444194666e-06,
+      "loss": 1.0163,
+      "step": 2787
+    },
+    {
+      "epoch": 0.21492445266728338,
+      "grad_norm": 3.802039623260498,
+      "learning_rate": 9.130188929677948e-06,
+      "loss": 1.0257,
+      "step": 2788
+    },
+    {
+      "epoch": 0.21500154178230033,
+      "grad_norm": 4.080732822418213,
+      "learning_rate": 9.129485157665339e-06,
+      "loss": 0.9556,
+      "step": 2789
+    },
+    {
+      "epoch": 0.2150786308973173,
+      "grad_norm": 3.748866081237793,
+      "learning_rate": 9.12878112820072e-06,
+      "loss": 0.9364,
+      "step": 2790
+    },
+    {
+      "epoch": 0.21515572001233427,
+      "grad_norm": 3.8058340549468994,
+      "learning_rate": 9.12807684132798e-06,
+      "loss": 0.9971,
+      "step": 2791
+    },
+    {
+      "epoch": 0.2152328091273512,
+      "grad_norm": 4.344945430755615,
+      "learning_rate": 9.127372297091028e-06,
+      "loss": 1.1081,
+      "step": 2792
+    },
+    {
+      "epoch": 0.21530989824236818,
+      "grad_norm": 4.168917655944824,
+      "learning_rate": 9.126667495533791e-06,
+      "loss": 1.0641,
+      "step": 2793
+    },
+    {
+      "epoch": 0.21538698735738512,
+      "grad_norm": 3.46258282661438,
+      "learning_rate": 9.125962436700207e-06,
+      "loss": 0.9893,
+      "step": 2794
+    },
+    {
+      "epoch": 0.2154640764724021,
+      "grad_norm": 3.5184786319732666,
+      "learning_rate": 9.125257120634234e-06,
+      "loss": 1.0043,
+      "step": 2795
+    },
+    {
+      "epoch": 0.21554116558741906,
+      "grad_norm": 3.7548696994781494,
+      "learning_rate": 9.124551547379846e-06,
+      "loss": 0.9768,
+      "step": 2796
+    },
+    {
+      "epoch": 0.215618254702436,
+      "grad_norm": 3.801421642303467,
+      "learning_rate": 9.12384571698103e-06,
+      "loss": 0.9912,
+      "step": 2797
+    },
+    {
+      "epoch": 0.21569534381745298,
+      "grad_norm": 4.047616004943848,
+      "learning_rate": 9.123139629481792e-06,
+      "loss": 1.0559,
+      "step": 2798
+    },
+    {
+      "epoch": 0.21577243293246995,
+      "grad_norm": 3.8471899032592773,
+      "learning_rate": 9.12243328492615e-06,
+      "loss": 1.099,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2158495220474869,
+      "grad_norm": 3.7033841609954834,
+      "learning_rate": 9.121726683358146e-06,
+      "loss": 1.0111,
+      "step": 2800
+    },
+    {
+      "epoch": 0.21592661116250386,
+      "grad_norm": 4.303826808929443,
+      "learning_rate": 9.12101982482183e-06,
+      "loss": 1.025,
+      "step": 2801
+    },
+    {
+      "epoch": 0.2160037002775208,
+      "grad_norm": 3.404069662094116,
+      "learning_rate": 9.120312709361271e-06,
+      "loss": 0.9853,
+      "step": 2802
+    },
+    {
+      "epoch": 0.21608078939253778,
+      "grad_norm": 3.566990613937378,
+      "learning_rate": 9.119605337020554e-06,
+      "loss": 1.0534,
+      "step": 2803
+    },
+    {
+      "epoch": 0.21615787850755475,
+      "grad_norm": 4.058454990386963,
+      "learning_rate": 9.118897707843779e-06,
+      "loss": 0.9761,
+      "step": 2804
+    },
+    {
+      "epoch": 0.2162349676225717,
+      "grad_norm": 3.7274980545043945,
+      "learning_rate": 9.118189821875066e-06,
+      "loss": 0.9726,
+      "step": 2805
+    },
+    {
+      "epoch": 0.21631205673758866,
+      "grad_norm": 3.776718854904175,
+      "learning_rate": 9.117481679158546e-06,
+      "loss": 1.0489,
+      "step": 2806
+    },
+    {
+      "epoch": 0.2163891458526056,
+      "grad_norm": 3.99607253074646,
+      "learning_rate": 9.116773279738367e-06,
+      "loss": 0.9758,
+      "step": 2807
+    },
+    {
+      "epoch": 0.21646623496762257,
+      "grad_norm": 3.691988945007324,
+      "learning_rate": 9.116064623658696e-06,
+      "loss": 1.0635,
+      "step": 2808
+    },
+    {
+      "epoch": 0.21654332408263954,
+      "grad_norm": 4.369837284088135,
+      "learning_rate": 9.115355710963714e-06,
+      "loss": 1.1056,
+      "step": 2809
+    },
+    {
+      "epoch": 0.2166204131976565,
+      "grad_norm": 3.6378722190856934,
+      "learning_rate": 9.114646541697617e-06,
+      "loss": 0.8665,
+      "step": 2810
+    },
+    {
+      "epoch": 0.21669750231267346,
+      "grad_norm": 3.6210851669311523,
+      "learning_rate": 9.113937115904618e-06,
+      "loss": 0.9098,
+      "step": 2811
+    },
+    {
+      "epoch": 0.2167745914276904,
+      "grad_norm": 3.999722480773926,
+      "learning_rate": 9.113227433628948e-06,
+      "loss": 0.9969,
+      "step": 2812
+    },
+    {
+      "epoch": 0.21685168054270737,
+      "grad_norm": 3.984334707260132,
+      "learning_rate": 9.112517494914847e-06,
+      "loss": 1.0185,
+      "step": 2813
+    },
+    {
+      "epoch": 0.21692876965772434,
+      "grad_norm": 3.8469057083129883,
+      "learning_rate": 9.111807299806584e-06,
+      "loss": 1.0389,
+      "step": 2814
+    },
+    {
+      "epoch": 0.21700585877274128,
+      "grad_norm": 4.206716537475586,
+      "learning_rate": 9.111096848348429e-06,
+      "loss": 1.0314,
+      "step": 2815
+    },
+    {
+      "epoch": 0.21708294788775825,
+      "grad_norm": 4.628408908843994,
+      "learning_rate": 9.110386140584677e-06,
+      "loss": 1.0994,
+      "step": 2816
+    },
+    {
+      "epoch": 0.2171600370027752,
+      "grad_norm": 3.524751663208008,
+      "learning_rate": 9.109675176559639e-06,
+      "loss": 0.9583,
+      "step": 2817
+    },
+    {
+      "epoch": 0.21723712611779217,
+      "grad_norm": 3.7516028881073,
+      "learning_rate": 9.108963956317635e-06,
+      "loss": 1.0813,
+      "step": 2818
+    },
+    {
+      "epoch": 0.21731421523280914,
+      "grad_norm": 4.167080879211426,
+      "learning_rate": 9.108252479903012e-06,
+      "loss": 1.0686,
+      "step": 2819
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 4.301987648010254,
+      "learning_rate": 9.107540747360124e-06,
+      "loss": 1.0793,
+      "step": 2820
+    },
+    {
+      "epoch": 0.21746839346284305,
+      "grad_norm": 4.356029987335205,
+      "learning_rate": 9.106828758733344e-06,
+      "loss": 1.0524,
+      "step": 2821
+    },
+    {
+      "epoch": 0.21754548257786,
+      "grad_norm": 3.900385618209839,
+      "learning_rate": 9.10611651406706e-06,
+      "loss": 1.0539,
+      "step": 2822
+    },
+    {
+      "epoch": 0.21762257169287696,
+      "grad_norm": 3.5353188514709473,
+      "learning_rate": 9.105404013405677e-06,
+      "loss": 1.0595,
+      "step": 2823
+    },
+    {
+      "epoch": 0.21769966080789394,
+      "grad_norm": 3.6671149730682373,
+      "learning_rate": 9.104691256793618e-06,
+      "loss": 1.1179,
+      "step": 2824
+    },
+    {
+      "epoch": 0.21777674992291088,
+      "grad_norm": 3.817493200302124,
+      "learning_rate": 9.103978244275316e-06,
+      "loss": 0.9962,
+      "step": 2825
+    },
+    {
+      "epoch": 0.21785383903792785,
+      "grad_norm": 3.781205654144287,
+      "learning_rate": 9.103264975895227e-06,
+      "loss": 1.0873,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2179309281529448,
+      "grad_norm": 3.8384957313537598,
+      "learning_rate": 9.102551451697816e-06,
+      "loss": 1.0985,
+      "step": 2827
+    },
+    {
+      "epoch": 0.21800801726796176,
+      "grad_norm": 3.9920754432678223,
+      "learning_rate": 9.10183767172757e-06,
+      "loss": 1.1307,
+      "step": 2828
+    },
+    {
+      "epoch": 0.21808510638297873,
+      "grad_norm": 3.888017416000366,
+      "learning_rate": 9.101123636028993e-06,
+      "loss": 1.0162,
+      "step": 2829
+    },
+    {
+      "epoch": 0.21816219549799568,
+      "grad_norm": 3.8230643272399902,
+      "learning_rate": 9.100409344646593e-06,
+      "loss": 1.0684,
+      "step": 2830
+    },
+    {
+      "epoch": 0.21823928461301265,
+      "grad_norm": 3.7765209674835205,
+      "learning_rate": 9.09969479762491e-06,
+      "loss": 0.9862,
+      "step": 2831
+    },
+    {
+      "epoch": 0.2183163737280296,
+      "grad_norm": 3.8649239540100098,
+      "learning_rate": 9.098979995008486e-06,
+      "loss": 1.058,
+      "step": 2832
+    },
+    {
+      "epoch": 0.21839346284304656,
+      "grad_norm": 3.9235241413116455,
+      "learning_rate": 9.098264936841891e-06,
+      "loss": 1.0378,
+      "step": 2833
+    },
+    {
+      "epoch": 0.21847055195806353,
+      "grad_norm": 4.208407878875732,
+      "learning_rate": 9.097549623169701e-06,
+      "loss": 1.0499,
+      "step": 2834
+    },
+    {
+      "epoch": 0.21854764107308047,
+      "grad_norm": 3.320283889770508,
+      "learning_rate": 9.096834054036516e-06,
+      "loss": 0.9359,
+      "step": 2835
+    },
+    {
+      "epoch": 0.21862473018809744,
+      "grad_norm": 4.07008695602417,
+      "learning_rate": 9.096118229486945e-06,
+      "loss": 1.193,
+      "step": 2836
+    },
+    {
+      "epoch": 0.2187018193031144,
+      "grad_norm": 3.860478162765503,
+      "learning_rate": 9.095402149565615e-06,
+      "loss": 0.9158,
+      "step": 2837
+    },
+    {
+      "epoch": 0.21877890841813136,
+      "grad_norm": 4.054520606994629,
+      "learning_rate": 9.094685814317174e-06,
+      "loss": 1.0763,
+      "step": 2838
+    },
+    {
+      "epoch": 0.21885599753314833,
+      "grad_norm": 3.9474873542785645,
+      "learning_rate": 9.093969223786277e-06,
+      "loss": 1.0329,
+      "step": 2839
+    },
+    {
+      "epoch": 0.21893308664816527,
+      "grad_norm": 3.73187255859375,
+      "learning_rate": 9.093252378017604e-06,
+      "loss": 1.0298,
+      "step": 2840
+    },
+    {
+      "epoch": 0.21901017576318224,
+      "grad_norm": 4.097843647003174,
+      "learning_rate": 9.092535277055845e-06,
+      "loss": 1.1312,
+      "step": 2841
+    },
+    {
+      "epoch": 0.2190872648781992,
+      "grad_norm": 3.9080405235290527,
+      "learning_rate": 9.091817920945704e-06,
+      "loss": 0.9503,
+      "step": 2842
+    },
+    {
+      "epoch": 0.21916435399321615,
+      "grad_norm": 3.772008180618286,
+      "learning_rate": 9.09110030973191e-06,
+      "loss": 0.9951,
+      "step": 2843
+    },
+    {
+      "epoch": 0.21924144310823312,
+      "grad_norm": 3.8021061420440674,
+      "learning_rate": 9.090382443459201e-06,
+      "loss": 1.068,
+      "step": 2844
+    },
+    {
+      "epoch": 0.21931853222325007,
+      "grad_norm": 3.884033203125,
+      "learning_rate": 9.089664322172331e-06,
+      "loss": 0.9915,
+      "step": 2845
+    },
+    {
+      "epoch": 0.21939562133826704,
+      "grad_norm": 3.7119226455688477,
+      "learning_rate": 9.08894594591607e-06,
+      "loss": 0.9613,
+      "step": 2846
+    },
+    {
+      "epoch": 0.219472710453284,
+      "grad_norm": 3.494401454925537,
+      "learning_rate": 9.088227314735208e-06,
+      "loss": 0.912,
+      "step": 2847
+    },
+    {
+      "epoch": 0.21954979956830095,
+      "grad_norm": 4.098977088928223,
+      "learning_rate": 9.087508428674546e-06,
+      "loss": 0.9818,
+      "step": 2848
+    },
+    {
+      "epoch": 0.21962688868331792,
+      "grad_norm": 3.923501491546631,
+      "learning_rate": 9.086789287778902e-06,
+      "loss": 0.9673,
+      "step": 2849
+    },
+    {
+      "epoch": 0.21970397779833487,
+      "grad_norm": 3.787045955657959,
+      "learning_rate": 9.086069892093114e-06,
+      "loss": 1.0714,
+      "step": 2850
+    },
+    {
+      "epoch": 0.21978106691335184,
+      "grad_norm": 4.722983360290527,
+      "learning_rate": 9.085350241662028e-06,
+      "loss": 1.0963,
+      "step": 2851
+    },
+    {
+      "epoch": 0.2198581560283688,
+      "grad_norm": 3.732903242111206,
+      "learning_rate": 9.084630336530515e-06,
+      "loss": 1.0903,
+      "step": 2852
+    },
+    {
+      "epoch": 0.21993524514338575,
+      "grad_norm": 4.350831508636475,
+      "learning_rate": 9.083910176743455e-06,
+      "loss": 1.1121,
+      "step": 2853
+    },
+    {
+      "epoch": 0.22001233425840272,
+      "grad_norm": 3.895369291305542,
+      "learning_rate": 9.083189762345746e-06,
+      "loss": 0.9028,
+      "step": 2854
+    },
+    {
+      "epoch": 0.22008942337341966,
+      "grad_norm": 4.156214237213135,
+      "learning_rate": 9.082469093382303e-06,
+      "loss": 1.0989,
+      "step": 2855
+    },
+    {
+      "epoch": 0.22016651248843663,
+      "grad_norm": 3.565054416656494,
+      "learning_rate": 9.081748169898054e-06,
+      "loss": 1.0293,
+      "step": 2856
+    },
+    {
+      "epoch": 0.2202436016034536,
+      "grad_norm": 4.331825256347656,
+      "learning_rate": 9.08102699193795e-06,
+      "loss": 1.1025,
+      "step": 2857
+    },
+    {
+      "epoch": 0.22032069071847055,
+      "grad_norm": 5.1282734870910645,
+      "learning_rate": 9.080305559546947e-06,
+      "loss": 0.9675,
+      "step": 2858
+    },
+    {
+      "epoch": 0.22039777983348752,
+      "grad_norm": 4.076694011688232,
+      "learning_rate": 9.079583872770025e-06,
+      "loss": 1.0403,
+      "step": 2859
+    },
+    {
+      "epoch": 0.22047486894850446,
+      "grad_norm": 3.9768247604370117,
+      "learning_rate": 9.078861931652178e-06,
+      "loss": 1.0653,
+      "step": 2860
+    },
+    {
+      "epoch": 0.22055195806352143,
+      "grad_norm": 3.875684976577759,
+      "learning_rate": 9.078139736238414e-06,
+      "loss": 0.8698,
+      "step": 2861
+    },
+    {
+      "epoch": 0.2206290471785384,
+      "grad_norm": 3.808002471923828,
+      "learning_rate": 9.077417286573759e-06,
+      "loss": 1.0765,
+      "step": 2862
+    },
+    {
+      "epoch": 0.22070613629355534,
+      "grad_norm": 3.7562291622161865,
+      "learning_rate": 9.076694582703252e-06,
+      "loss": 0.9728,
+      "step": 2863
+    },
+    {
+      "epoch": 0.22078322540857231,
+      "grad_norm": 3.816787004470825,
+      "learning_rate": 9.075971624671953e-06,
+      "loss": 1.1007,
+      "step": 2864
+    },
+    {
+      "epoch": 0.22086031452358926,
+      "grad_norm": 3.6228833198547363,
+      "learning_rate": 9.075248412524932e-06,
+      "loss": 0.9763,
+      "step": 2865
+    },
+    {
+      "epoch": 0.22093740363860623,
+      "grad_norm": 3.993579864501953,
+      "learning_rate": 9.074524946307282e-06,
+      "loss": 0.9982,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2210144927536232,
+      "grad_norm": 3.9125168323516846,
+      "learning_rate": 9.0738012260641e-06,
+      "loss": 1.015,
+      "step": 2867
+    },
+    {
+      "epoch": 0.22109158186864014,
+      "grad_norm": 3.7729082107543945,
+      "learning_rate": 9.073077251840514e-06,
+      "loss": 0.8864,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2211686709836571,
+      "grad_norm": 3.5397660732269287,
+      "learning_rate": 9.072353023681653e-06,
+      "loss": 0.9667,
+      "step": 2869
+    },
+    {
+      "epoch": 0.22124576009867405,
+      "grad_norm": 4.714186668395996,
+      "learning_rate": 9.071628541632675e-06,
+      "loss": 0.8968,
+      "step": 2870
+    },
+    {
+      "epoch": 0.22132284921369103,
+      "grad_norm": 4.182320594787598,
+      "learning_rate": 9.070903805738744e-06,
+      "loss": 1.1579,
+      "step": 2871
+    },
+    {
+      "epoch": 0.221399938328708,
+      "grad_norm": 3.612896203994751,
+      "learning_rate": 9.070178816045043e-06,
+      "loss": 1.0278,
+      "step": 2872
+    },
+    {
+      "epoch": 0.22147702744372494,
+      "grad_norm": 3.8569085597991943,
+      "learning_rate": 9.069453572596774e-06,
+      "loss": 1.0517,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2215541165587419,
+      "grad_norm": 3.7985715866088867,
+      "learning_rate": 9.068728075439153e-06,
+      "loss": 1.112,
+      "step": 2874
+    },
+    {
+      "epoch": 0.22163120567375885,
+      "grad_norm": 3.7914130687713623,
+      "learning_rate": 9.068002324617407e-06,
+      "loss": 1.1309,
+      "step": 2875
+    },
+    {
+      "epoch": 0.22170829478877582,
+      "grad_norm": 3.6983115673065186,
+      "learning_rate": 9.067276320176783e-06,
+      "loss": 1.0458,
+      "step": 2876
+    },
+    {
+      "epoch": 0.2217853839037928,
+      "grad_norm": 3.6084189414978027,
+      "learning_rate": 9.066550062162547e-06,
+      "loss": 0.9881,
+      "step": 2877
+    },
+    {
+      "epoch": 0.22186247301880974,
+      "grad_norm": 3.6229233741760254,
+      "learning_rate": 9.065823550619976e-06,
+      "loss": 0.9464,
+      "step": 2878
+    },
+    {
+      "epoch": 0.2219395621338267,
+      "grad_norm": 3.949673652648926,
+      "learning_rate": 9.065096785594364e-06,
+      "loss": 1.0257,
+      "step": 2879
+    },
+    {
+      "epoch": 0.22201665124884365,
+      "grad_norm": 3.5906119346618652,
+      "learning_rate": 9.064369767131021e-06,
+      "loss": 0.9943,
+      "step": 2880
+    },
+    {
+      "epoch": 0.22209374036386062,
+      "grad_norm": 3.8634934425354004,
+      "learning_rate": 9.063642495275273e-06,
+      "loss": 1.015,
+      "step": 2881
+    },
+    {
+      "epoch": 0.2221708294788776,
+      "grad_norm": 4.381479740142822,
+      "learning_rate": 9.062914970072463e-06,
+      "loss": 1.1069,
+      "step": 2882
+    },
+    {
+      "epoch": 0.22224791859389453,
+      "grad_norm": 3.475149393081665,
+      "learning_rate": 9.062187191567947e-06,
+      "loss": 1.0361,
+      "step": 2883
+    },
+    {
+      "epoch": 0.2223250077089115,
+      "grad_norm": 3.652864694595337,
+      "learning_rate": 9.061459159807096e-06,
+      "loss": 0.9466,
+      "step": 2884
+    },
+    {
+      "epoch": 0.22240209682392847,
+      "grad_norm": 3.7307932376861572,
+      "learning_rate": 9.060730874835301e-06,
+      "loss": 1.0076,
+      "step": 2885
+    },
+    {
+      "epoch": 0.22247918593894542,
+      "grad_norm": 4.080811977386475,
+      "learning_rate": 9.060002336697968e-06,
+      "loss": 1.0577,
+      "step": 2886
+    },
+    {
+      "epoch": 0.2225562750539624,
+      "grad_norm": 3.7193477153778076,
+      "learning_rate": 9.059273545440516e-06,
+      "loss": 1.1106,
+      "step": 2887
+    },
+    {
+      "epoch": 0.22263336416897933,
+      "grad_norm": 3.971679925918579,
+      "learning_rate": 9.058544501108383e-06,
+      "loss": 1.0908,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2227104532839963,
+      "grad_norm": 3.5091354846954346,
+      "learning_rate": 9.05781520374702e-06,
+      "loss": 0.8429,
+      "step": 2889
+    },
+    {
+      "epoch": 0.22278754239901327,
+      "grad_norm": 3.8452577590942383,
+      "learning_rate": 9.057085653401896e-06,
+      "loss": 1.0097,
+      "step": 2890
+    },
+    {
+      "epoch": 0.22286463151403021,
+      "grad_norm": 3.8790271282196045,
+      "learning_rate": 9.056355850118492e-06,
+      "loss": 1.1066,
+      "step": 2891
+    },
+    {
+      "epoch": 0.22294172062904719,
+      "grad_norm": 4.002335071563721,
+      "learning_rate": 9.055625793942308e-06,
+      "loss": 1.0349,
+      "step": 2892
+    },
+    {
+      "epoch": 0.22301880974406413,
+      "grad_norm": 3.5299620628356934,
+      "learning_rate": 9.054895484918863e-06,
+      "loss": 0.9838,
+      "step": 2893
+    },
+    {
+      "epoch": 0.2230958988590811,
+      "grad_norm": 4.129887104034424,
+      "learning_rate": 9.054164923093684e-06,
+      "loss": 1.1289,
+      "step": 2894
+    },
+    {
+      "epoch": 0.22317298797409807,
+      "grad_norm": 3.8371896743774414,
+      "learning_rate": 9.05343410851232e-06,
+      "loss": 0.9709,
+      "step": 2895
+    },
+    {
+      "epoch": 0.223250077089115,
+      "grad_norm": 3.9022574424743652,
+      "learning_rate": 9.052703041220332e-06,
+      "loss": 1.043,
+      "step": 2896
+    },
+    {
+      "epoch": 0.22332716620413198,
+      "grad_norm": 3.414318323135376,
+      "learning_rate": 9.0519717212633e-06,
+      "loss": 1.0308,
+      "step": 2897
+    },
+    {
+      "epoch": 0.22340425531914893,
+      "grad_norm": 3.9684841632843018,
+      "learning_rate": 9.051240148686815e-06,
+      "loss": 0.9831,
+      "step": 2898
+    },
+    {
+      "epoch": 0.2234813444341659,
+      "grad_norm": 3.743062973022461,
+      "learning_rate": 9.050508323536488e-06,
+      "loss": 1.1446,
+      "step": 2899
+    },
+    {
+      "epoch": 0.22355843354918287,
+      "grad_norm": 4.060634613037109,
+      "learning_rate": 9.049776245857947e-06,
+      "loss": 1.1005,
+      "step": 2900
+    },
+    {
+      "epoch": 0.2236355226641998,
+      "grad_norm": 3.9923877716064453,
+      "learning_rate": 9.049043915696831e-06,
+      "loss": 0.9606,
+      "step": 2901
+    },
+    {
+      "epoch": 0.22371261177921678,
+      "grad_norm": 3.815098285675049,
+      "learning_rate": 9.048311333098798e-06,
+      "loss": 1.1327,
+      "step": 2902
+    },
+    {
+      "epoch": 0.22378970089423372,
+      "grad_norm": 3.9265224933624268,
+      "learning_rate": 9.04757849810952e-06,
+      "loss": 1.0604,
+      "step": 2903
+    },
+    {
+      "epoch": 0.2238667900092507,
+      "grad_norm": 3.8014273643493652,
+      "learning_rate": 9.046845410774685e-06,
+      "loss": 0.9963,
+      "step": 2904
+    },
+    {
+      "epoch": 0.22394387912426766,
+      "grad_norm": 4.092421531677246,
+      "learning_rate": 9.04611207114e-06,
+      "loss": 1.1604,
+      "step": 2905
+    },
+    {
+      "epoch": 0.2240209682392846,
+      "grad_norm": 4.311123847961426,
+      "learning_rate": 9.045378479251179e-06,
+      "loss": 1.1204,
+      "step": 2906
+    },
+    {
+      "epoch": 0.22409805735430158,
+      "grad_norm": 3.77968430519104,
+      "learning_rate": 9.044644635153963e-06,
+      "loss": 1.0621,
+      "step": 2907
+    },
+    {
+      "epoch": 0.22417514646931852,
+      "grad_norm": 3.6823370456695557,
+      "learning_rate": 9.0439105388941e-06,
+      "loss": 0.9396,
+      "step": 2908
+    },
+    {
+      "epoch": 0.2242522355843355,
+      "grad_norm": 3.8284711837768555,
+      "learning_rate": 9.043176190517362e-06,
+      "loss": 1.1167,
+      "step": 2909
+    },
+    {
+      "epoch": 0.22432932469935246,
+      "grad_norm": 3.750458002090454,
+      "learning_rate": 9.042441590069526e-06,
+      "loss": 0.9483,
+      "step": 2910
+    },
+    {
+      "epoch": 0.2244064138143694,
+      "grad_norm": 3.8216614723205566,
+      "learning_rate": 9.041706737596397e-06,
+      "loss": 1.0972,
+      "step": 2911
+    },
+    {
+      "epoch": 0.22448350292938637,
+      "grad_norm": 4.093216896057129,
+      "learning_rate": 9.040971633143782e-06,
+      "loss": 1.2209,
+      "step": 2912
+    },
+    {
+      "epoch": 0.22456059204440332,
+      "grad_norm": 3.4413156509399414,
+      "learning_rate": 9.040236276757514e-06,
+      "loss": 0.8431,
+      "step": 2913
+    },
+    {
+      "epoch": 0.2246376811594203,
+      "grad_norm": 3.6368567943573,
+      "learning_rate": 9.03950066848344e-06,
+      "loss": 1.0675,
+      "step": 2914
+    },
+    {
+      "epoch": 0.22471477027443726,
+      "grad_norm": 3.8164429664611816,
+      "learning_rate": 9.038764808367422e-06,
+      "loss": 1.1033,
+      "step": 2915
+    },
+    {
+      "epoch": 0.2247918593894542,
+      "grad_norm": 3.761277198791504,
+      "learning_rate": 9.038028696455335e-06,
+      "loss": 1.0389,
+      "step": 2916
+    },
+    {
+      "epoch": 0.22486894850447117,
+      "grad_norm": 3.7666642665863037,
+      "learning_rate": 9.03729233279307e-06,
+      "loss": 1.0024,
+      "step": 2917
+    },
+    {
+      "epoch": 0.22494603761948812,
+      "grad_norm": 3.848273992538452,
+      "learning_rate": 9.03655571742654e-06,
+      "loss": 1.0972,
+      "step": 2918
+    },
+    {
+      "epoch": 0.22502312673450509,
+      "grad_norm": 3.224196672439575,
+      "learning_rate": 9.035818850401667e-06,
+      "loss": 0.9211,
+      "step": 2919
+    },
+    {
+      "epoch": 0.22510021584952206,
+      "grad_norm": 3.469214916229248,
+      "learning_rate": 9.03508173176439e-06,
+      "loss": 1.0197,
+      "step": 2920
+    },
+    {
+      "epoch": 0.225177304964539,
+      "grad_norm": 3.8008034229278564,
+      "learning_rate": 9.034344361560663e-06,
+      "loss": 1.0002,
+      "step": 2921
+    },
+    {
+      "epoch": 0.22525439407955597,
+      "grad_norm": 3.6283414363861084,
+      "learning_rate": 9.033606739836463e-06,
+      "loss": 1.0574,
+      "step": 2922
+    },
+    {
+      "epoch": 0.22533148319457294,
+      "grad_norm": 4.280021667480469,
+      "learning_rate": 9.032868866637772e-06,
+      "loss": 1.0326,
+      "step": 2923
+    },
+    {
+      "epoch": 0.22540857230958988,
+      "grad_norm": 3.6488280296325684,
+      "learning_rate": 9.032130742010594e-06,
+      "loss": 1.0502,
+      "step": 2924
+    },
+    {
+      "epoch": 0.22548566142460685,
+      "grad_norm": 3.933264970779419,
+      "learning_rate": 9.031392366000945e-06,
+      "loss": 1.1104,
+      "step": 2925
+    },
+    {
+      "epoch": 0.2255627505396238,
+      "grad_norm": 3.7967422008514404,
+      "learning_rate": 9.030653738654864e-06,
+      "loss": 1.0144,
+      "step": 2926
+    },
+    {
+      "epoch": 0.22563983965464077,
+      "grad_norm": 3.846534490585327,
+      "learning_rate": 9.029914860018399e-06,
+      "loss": 1.0428,
+      "step": 2927
+    },
+    {
+      "epoch": 0.22571692876965774,
+      "grad_norm": 3.627790689468384,
+      "learning_rate": 9.029175730137611e-06,
+      "loss": 0.996,
+      "step": 2928
+    },
+    {
+      "epoch": 0.22579401788467468,
+      "grad_norm": 3.8956503868103027,
+      "learning_rate": 9.028436349058584e-06,
+      "loss": 1.1075,
+      "step": 2929
+    },
+    {
+      "epoch": 0.22587110699969165,
+      "grad_norm": 4.157092571258545,
+      "learning_rate": 9.027696716827416e-06,
+      "loss": 1.0656,
+      "step": 2930
+    },
+    {
+      "epoch": 0.2259481961147086,
+      "grad_norm": 3.8785336017608643,
+      "learning_rate": 9.026956833490217e-06,
+      "loss": 0.9546,
+      "step": 2931
+    },
+    {
+      "epoch": 0.22602528522972556,
+      "grad_norm": 3.8939132690429688,
+      "learning_rate": 9.026216699093114e-06,
+      "loss": 1.0963,
+      "step": 2932
+    },
+    {
+      "epoch": 0.22610237434474253,
+      "grad_norm": 3.9576644897460938,
+      "learning_rate": 9.025476313682253e-06,
+      "loss": 1.0504,
+      "step": 2933
+    },
+    {
+      "epoch": 0.22617946345975948,
+      "grad_norm": 3.753843069076538,
+      "learning_rate": 9.024735677303793e-06,
+      "loss": 0.9969,
+      "step": 2934
+    },
+    {
+      "epoch": 0.22625655257477645,
+      "grad_norm": 3.669050693511963,
+      "learning_rate": 9.023994790003908e-06,
+      "loss": 0.9765,
+      "step": 2935
+    },
+    {
+      "epoch": 0.2263336416897934,
+      "grad_norm": 3.700514078140259,
+      "learning_rate": 9.023253651828789e-06,
+      "loss": 1.1164,
+      "step": 2936
+    },
+    {
+      "epoch": 0.22641073080481036,
+      "grad_norm": 3.998882532119751,
+      "learning_rate": 9.022512262824642e-06,
+      "loss": 1.0073,
+      "step": 2937
+    },
+    {
+      "epoch": 0.22648781991982733,
+      "grad_norm": 4.040007591247559,
+      "learning_rate": 9.021770623037688e-06,
+      "loss": 1.0537,
+      "step": 2938
+    },
+    {
+      "epoch": 0.22656490903484428,
+      "grad_norm": 4.1160173416137695,
+      "learning_rate": 9.021028732514166e-06,
+      "loss": 1.1404,
+      "step": 2939
+    },
+    {
+      "epoch": 0.22664199814986125,
+      "grad_norm": 3.754762887954712,
+      "learning_rate": 9.020286591300325e-06,
+      "loss": 1.045,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2267190872648782,
+      "grad_norm": 3.404191017150879,
+      "learning_rate": 9.019544199442438e-06,
+      "loss": 0.9894,
+      "step": 2941
+    },
+    {
+      "epoch": 0.22679617637989516,
+      "grad_norm": 3.9192099571228027,
+      "learning_rate": 9.01880155698679e-06,
+      "loss": 0.9169,
+      "step": 2942
+    },
+    {
+      "epoch": 0.22687326549491213,
+      "grad_norm": 3.7557027339935303,
+      "learning_rate": 9.018058663979676e-06,
+      "loss": 1.0611,
+      "step": 2943
+    },
+    {
+      "epoch": 0.22695035460992907,
+      "grad_norm": 3.523135185241699,
+      "learning_rate": 9.017315520467416e-06,
+      "loss": 1.019,
+      "step": 2944
+    },
+    {
+      "epoch": 0.22702744372494604,
+      "grad_norm": 3.6484501361846924,
+      "learning_rate": 9.016572126496339e-06,
+      "loss": 0.9376,
+      "step": 2945
+    },
+    {
+      "epoch": 0.22710453283996299,
+      "grad_norm": 4.25970983505249,
+      "learning_rate": 9.015828482112793e-06,
+      "loss": 0.9697,
+      "step": 2946
+    },
+    {
+      "epoch": 0.22718162195497996,
+      "grad_norm": 3.3932673931121826,
+      "learning_rate": 9.01508458736314e-06,
+      "loss": 0.9625,
+      "step": 2947
+    },
+    {
+      "epoch": 0.22725871106999693,
+      "grad_norm": 3.8393819332122803,
+      "learning_rate": 9.014340442293755e-06,
+      "loss": 1.0045,
+      "step": 2948
+    },
+    {
+      "epoch": 0.22733580018501387,
+      "grad_norm": 3.575310230255127,
+      "learning_rate": 9.013596046951037e-06,
+      "loss": 1.0233,
+      "step": 2949
+    },
+    {
+      "epoch": 0.22741288930003084,
+      "grad_norm": 3.7742843627929688,
+      "learning_rate": 9.012851401381391e-06,
+      "loss": 0.992,
+      "step": 2950
+    },
+    {
+      "epoch": 0.22748997841504778,
+      "grad_norm": 4.013235092163086,
+      "learning_rate": 9.012106505631244e-06,
+      "loss": 0.931,
+      "step": 2951
+    },
+    {
+      "epoch": 0.22756706753006475,
+      "grad_norm": 3.5595037937164307,
+      "learning_rate": 9.011361359747034e-06,
+      "loss": 1.0418,
+      "step": 2952
+    },
+    {
+      "epoch": 0.22764415664508172,
+      "grad_norm": 4.25527811050415,
+      "learning_rate": 9.01061596377522e-06,
+      "loss": 1.0631,
+      "step": 2953
+    },
+    {
+      "epoch": 0.22772124576009867,
+      "grad_norm": 4.163780689239502,
+      "learning_rate": 9.009870317762273e-06,
+      "loss": 1.0558,
+      "step": 2954
+    },
+    {
+      "epoch": 0.22779833487511564,
+      "grad_norm": 3.863081216812134,
+      "learning_rate": 9.009124421754676e-06,
+      "loss": 0.8664,
+      "step": 2955
+    },
+    {
+      "epoch": 0.22787542399013258,
+      "grad_norm": 3.8147034645080566,
+      "learning_rate": 9.008378275798939e-06,
+      "loss": 1.0287,
+      "step": 2956
+    },
+    {
+      "epoch": 0.22795251310514955,
+      "grad_norm": 4.305243968963623,
+      "learning_rate": 9.007631879941575e-06,
+      "loss": 1.1372,
+      "step": 2957
+    },
+    {
+      "epoch": 0.22802960222016652,
+      "grad_norm": 3.846221685409546,
+      "learning_rate": 9.006885234229119e-06,
+      "loss": 0.9999,
+      "step": 2958
+    },
+    {
+      "epoch": 0.22810669133518346,
+      "grad_norm": 3.8825647830963135,
+      "learning_rate": 9.00613833870812e-06,
+      "loss": 1.0099,
+      "step": 2959
+    },
+    {
+      "epoch": 0.22818378045020044,
+      "grad_norm": 3.7892098426818848,
+      "learning_rate": 9.005391193425144e-06,
+      "loss": 0.9672,
+      "step": 2960
+    },
+    {
+      "epoch": 0.22826086956521738,
+      "grad_norm": 4.183690547943115,
+      "learning_rate": 9.004643798426772e-06,
+      "loss": 0.9231,
+      "step": 2961
+    },
+    {
+      "epoch": 0.22833795868023435,
+      "grad_norm": 3.8365590572357178,
+      "learning_rate": 9.003896153759602e-06,
+      "loss": 1.1283,
+      "step": 2962
+    },
+    {
+      "epoch": 0.22841504779525132,
+      "grad_norm": 3.8340327739715576,
+      "learning_rate": 9.00314825947024e-06,
+      "loss": 0.957,
+      "step": 2963
+    },
+    {
+      "epoch": 0.22849213691026826,
+      "grad_norm": 4.034872531890869,
+      "learning_rate": 9.002400115605319e-06,
+      "loss": 1.0195,
+      "step": 2964
+    },
+    {
+      "epoch": 0.22856922602528523,
+      "grad_norm": 3.9130313396453857,
+      "learning_rate": 9.00165172221148e-06,
+      "loss": 1.1405,
+      "step": 2965
+    },
+    {
+      "epoch": 0.2286463151403022,
+      "grad_norm": 3.705179214477539,
+      "learning_rate": 9.00090307933538e-06,
+      "loss": 1.0191,
+      "step": 2966
+    },
+    {
+      "epoch": 0.22872340425531915,
+      "grad_norm": 3.985041618347168,
+      "learning_rate": 9.000154187023696e-06,
+      "loss": 1.0271,
+      "step": 2967
+    },
+    {
+      "epoch": 0.22880049337033612,
+      "grad_norm": 3.9260165691375732,
+      "learning_rate": 8.999405045323113e-06,
+      "loss": 1.0308,
+      "step": 2968
+    },
+    {
+      "epoch": 0.22887758248535306,
+      "grad_norm": 3.8115878105163574,
+      "learning_rate": 8.99865565428034e-06,
+      "loss": 0.9711,
+      "step": 2969
+    },
+    {
+      "epoch": 0.22895467160037003,
+      "grad_norm": 3.6064395904541016,
+      "learning_rate": 8.997906013942097e-06,
+      "loss": 1.0217,
+      "step": 2970
+    },
+    {
+      "epoch": 0.229031760715387,
+      "grad_norm": 3.8164539337158203,
+      "learning_rate": 8.99715612435512e-06,
+      "loss": 1.0237,
+      "step": 2971
+    },
+    {
+      "epoch": 0.22910884983040394,
+      "grad_norm": 11.672921180725098,
+      "learning_rate": 8.99640598556616e-06,
+      "loss": 1.0373,
+      "step": 2972
+    },
+    {
+      "epoch": 0.22918593894542091,
+      "grad_norm": 3.9608118534088135,
+      "learning_rate": 8.995655597621985e-06,
+      "loss": 1.0344,
+      "step": 2973
+    },
+    {
+      "epoch": 0.22926302806043786,
+      "grad_norm": 3.9548451900482178,
+      "learning_rate": 8.994904960569378e-06,
+      "loss": 1.0988,
+      "step": 2974
+    },
+    {
+      "epoch": 0.22934011717545483,
+      "grad_norm": 3.6159064769744873,
+      "learning_rate": 8.994154074455135e-06,
+      "loss": 1.0542,
+      "step": 2975
+    },
+    {
+      "epoch": 0.2294172062904718,
+      "grad_norm": 3.7155611515045166,
+      "learning_rate": 8.993402939326072e-06,
+      "loss": 1.0449,
+      "step": 2976
+    },
+    {
+      "epoch": 0.22949429540548874,
+      "grad_norm": 3.6567227840423584,
+      "learning_rate": 8.99265155522902e-06,
+      "loss": 0.9836,
+      "step": 2977
+    },
+    {
+      "epoch": 0.2295713845205057,
+      "grad_norm": 3.778001546859741,
+      "learning_rate": 8.991899922210821e-06,
+      "loss": 1.0483,
+      "step": 2978
+    },
+    {
+      "epoch": 0.22964847363552265,
+      "grad_norm": 4.267853260040283,
+      "learning_rate": 8.991148040318335e-06,
+      "loss": 0.9866,
+      "step": 2979
+    },
+    {
+      "epoch": 0.22972556275053962,
+      "grad_norm": 3.8425397872924805,
+      "learning_rate": 8.990395909598442e-06,
+      "loss": 1.0059,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2298026518655566,
+      "grad_norm": 4.225255489349365,
+      "learning_rate": 8.989643530098028e-06,
+      "loss": 1.0807,
+      "step": 2981
+    },
+    {
+      "epoch": 0.22987974098057354,
+      "grad_norm": 3.793910026550293,
+      "learning_rate": 8.988890901864006e-06,
+      "loss": 1.0763,
+      "step": 2982
+    },
+    {
+      "epoch": 0.2299568300955905,
+      "grad_norm": 4.9717020988464355,
+      "learning_rate": 8.988138024943294e-06,
+      "loss": 1.0412,
+      "step": 2983
+    },
+    {
+      "epoch": 0.23003391921060745,
+      "grad_norm": 3.8835487365722656,
+      "learning_rate": 8.98738489938283e-06,
+      "loss": 1.0469,
+      "step": 2984
+    },
+    {
+      "epoch": 0.23011100832562442,
+      "grad_norm": 3.7455644607543945,
+      "learning_rate": 8.98663152522957e-06,
+      "loss": 0.8876,
+      "step": 2985
+    },
+    {
+      "epoch": 0.2301880974406414,
+      "grad_norm": 4.213279724121094,
+      "learning_rate": 8.985877902530482e-06,
+      "loss": 1.0481,
+      "step": 2986
+    },
+    {
+      "epoch": 0.23026518655565834,
+      "grad_norm": 3.730454921722412,
+      "learning_rate": 8.985124031332549e-06,
+      "loss": 1.1566,
+      "step": 2987
+    },
+    {
+      "epoch": 0.2303422756706753,
+      "grad_norm": 3.9859375953674316,
+      "learning_rate": 8.984369911682773e-06,
+      "loss": 1.0923,
+      "step": 2988
+    },
+    {
+      "epoch": 0.23041936478569225,
+      "grad_norm": 3.603957176208496,
+      "learning_rate": 8.983615543628167e-06,
+      "loss": 1.0712,
+      "step": 2989
+    },
+    {
+      "epoch": 0.23049645390070922,
+      "grad_norm": 4.084186553955078,
+      "learning_rate": 8.982860927215765e-06,
+      "loss": 1.0741,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2305735430157262,
+      "grad_norm": 3.7829389572143555,
+      "learning_rate": 8.982106062492613e-06,
+      "loss": 1.1044,
+      "step": 2991
+    },
+    {
+      "epoch": 0.23065063213074313,
+      "grad_norm": 4.031806468963623,
+      "learning_rate": 8.98135094950577e-06,
+      "loss": 1.0782,
+      "step": 2992
+    },
+    {
+      "epoch": 0.2307277212457601,
+      "grad_norm": 4.211963653564453,
+      "learning_rate": 8.980595588302315e-06,
+      "loss": 1.0644,
+      "step": 2993
+    },
+    {
+      "epoch": 0.23080481036077705,
+      "grad_norm": 3.739516258239746,
+      "learning_rate": 8.979839978929342e-06,
+      "loss": 1.0677,
+      "step": 2994
+    },
+    {
+      "epoch": 0.23088189947579402,
+      "grad_norm": 3.528867483139038,
+      "learning_rate": 8.979084121433958e-06,
+      "loss": 0.9519,
+      "step": 2995
+    },
+    {
+      "epoch": 0.230958988590811,
+      "grad_norm": 3.507763147354126,
+      "learning_rate": 8.978328015863287e-06,
+      "loss": 0.9619,
+      "step": 2996
+    },
+    {
+      "epoch": 0.23103607770582793,
+      "grad_norm": 4.500532150268555,
+      "learning_rate": 8.977571662264471e-06,
+      "loss": 1.2321,
+      "step": 2997
+    },
+    {
+      "epoch": 0.2311131668208449,
+      "grad_norm": 3.3748843669891357,
+      "learning_rate": 8.976815060684659e-06,
+      "loss": 0.999,
+      "step": 2998
+    },
+    {
+      "epoch": 0.23119025593586184,
+      "grad_norm": 3.790412425994873,
+      "learning_rate": 8.976058211171027e-06,
+      "loss": 1.1291,
+      "step": 2999
+    },
+    {
+      "epoch": 0.23126734505087881,
+      "grad_norm": 3.6807661056518555,
+      "learning_rate": 8.975301113770756e-06,
+      "loss": 0.9742,
+      "step": 3000
+    },
+    {
+      "epoch": 0.23134443416589578,
+      "grad_norm": 3.902876138687134,
+      "learning_rate": 8.97454376853105e-06,
+      "loss": 1.1042,
+      "step": 3001
+    },
+    {
+      "epoch": 0.23142152328091273,
+      "grad_norm": 3.766279935836792,
+      "learning_rate": 8.973786175499123e-06,
+      "loss": 1.112,
+      "step": 3002
+    },
+    {
+      "epoch": 0.2314986123959297,
+      "grad_norm": 3.676820993423462,
+      "learning_rate": 8.973028334722212e-06,
+      "loss": 0.9986,
+      "step": 3003
+    },
+    {
+      "epoch": 0.23157570151094664,
+      "grad_norm": 3.577911138534546,
+      "learning_rate": 8.972270246247558e-06,
+      "loss": 0.9868,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2316527906259636,
+      "grad_norm": 3.7404353618621826,
+      "learning_rate": 8.97151191012243e-06,
+      "loss": 1.1051,
+      "step": 3005
+    },
+    {
+      "epoch": 0.23172987974098058,
+      "grad_norm": 3.7032761573791504,
+      "learning_rate": 8.9707533263941e-06,
+      "loss": 1.0054,
+      "step": 3006
+    },
+    {
+      "epoch": 0.23180696885599752,
+      "grad_norm": 3.8245747089385986,
+      "learning_rate": 8.969994495109868e-06,
+      "loss": 1.0588,
+      "step": 3007
+    },
+    {
+      "epoch": 0.2318840579710145,
+      "grad_norm": 3.9548802375793457,
+      "learning_rate": 8.96923541631704e-06,
+      "loss": 1.1028,
+      "step": 3008
+    },
+    {
+      "epoch": 0.23196114708603147,
+      "grad_norm": 3.6920297145843506,
+      "learning_rate": 8.968476090062941e-06,
+      "loss": 0.9935,
+      "step": 3009
+    },
+    {
+      "epoch": 0.2320382362010484,
+      "grad_norm": 3.5321335792541504,
+      "learning_rate": 8.967716516394911e-06,
+      "loss": 0.9922,
+      "step": 3010
+    },
+    {
+      "epoch": 0.23211532531606538,
+      "grad_norm": 3.3093225955963135,
+      "learning_rate": 8.966956695360304e-06,
+      "loss": 0.9767,
+      "step": 3011
+    },
+    {
+      "epoch": 0.23219241443108232,
+      "grad_norm": 3.7494637966156006,
+      "learning_rate": 8.966196627006493e-06,
+      "loss": 1.068,
+      "step": 3012
+    },
+    {
+      "epoch": 0.2322695035460993,
+      "grad_norm": 3.80411696434021,
+      "learning_rate": 8.965436311380866e-06,
+      "loss": 1.0047,
+      "step": 3013
+    },
+    {
+      "epoch": 0.23234659266111626,
+      "grad_norm": 3.954469919204712,
+      "learning_rate": 8.96467574853082e-06,
+      "loss": 1.0272,
+      "step": 3014
+    },
+    {
+      "epoch": 0.2324236817761332,
+      "grad_norm": 4.207308292388916,
+      "learning_rate": 8.963914938503777e-06,
+      "loss": 1.0145,
+      "step": 3015
+    },
+    {
+      "epoch": 0.23250077089115018,
+      "grad_norm": 3.938061237335205,
+      "learning_rate": 8.963153881347164e-06,
+      "loss": 0.9881,
+      "step": 3016
+    },
+    {
+      "epoch": 0.23257786000616712,
+      "grad_norm": 3.8759334087371826,
+      "learning_rate": 8.962392577108433e-06,
+      "loss": 1.0464,
+      "step": 3017
+    },
+    {
+      "epoch": 0.2326549491211841,
+      "grad_norm": 4.220122814178467,
+      "learning_rate": 8.96163102583505e-06,
+      "loss": 0.9949,
+      "step": 3018
+    },
+    {
+      "epoch": 0.23273203823620106,
+      "grad_norm": 3.3575594425201416,
+      "learning_rate": 8.960869227574486e-06,
+      "loss": 1.0295,
+      "step": 3019
+    },
+    {
+      "epoch": 0.232809127351218,
+      "grad_norm": 3.665529489517212,
+      "learning_rate": 8.960107182374242e-06,
+      "loss": 1.1279,
+      "step": 3020
+    },
+    {
+      "epoch": 0.23288621646623497,
+      "grad_norm": 4.267427921295166,
+      "learning_rate": 8.959344890281826e-06,
+      "loss": 1.0118,
+      "step": 3021
+    },
+    {
+      "epoch": 0.23296330558125192,
+      "grad_norm": 4.257380962371826,
+      "learning_rate": 8.958582351344759e-06,
+      "loss": 1.0997,
+      "step": 3022
+    },
+    {
+      "epoch": 0.2330403946962689,
+      "grad_norm": 4.008502006530762,
+      "learning_rate": 8.957819565610585e-06,
+      "loss": 1.0658,
+      "step": 3023
+    },
+    {
+      "epoch": 0.23311748381128586,
+      "grad_norm": 3.9444122314453125,
+      "learning_rate": 8.957056533126861e-06,
+      "loss": 1.0731,
+      "step": 3024
+    },
+    {
+      "epoch": 0.2331945729263028,
+      "grad_norm": 4.073714256286621,
+      "learning_rate": 8.956293253941155e-06,
+      "loss": 1.0287,
+      "step": 3025
+    },
+    {
+      "epoch": 0.23327166204131977,
+      "grad_norm": 4.01447057723999,
+      "learning_rate": 8.955529728101055e-06,
+      "loss": 1.1299,
+      "step": 3026
+    },
+    {
+      "epoch": 0.23334875115633671,
+      "grad_norm": 4.0629706382751465,
+      "learning_rate": 8.954765955654164e-06,
+      "loss": 1.0278,
+      "step": 3027
+    },
+    {
+      "epoch": 0.23342584027135369,
+      "grad_norm": 3.837249279022217,
+      "learning_rate": 8.954001936648095e-06,
+      "loss": 1.0456,
+      "step": 3028
+    },
+    {
+      "epoch": 0.23350292938637066,
+      "grad_norm": 3.798231363296509,
+      "learning_rate": 8.953237671130486e-06,
+      "loss": 0.9937,
+      "step": 3029
+    },
+    {
+      "epoch": 0.2335800185013876,
+      "grad_norm": 3.616384267807007,
+      "learning_rate": 8.952473159148982e-06,
+      "loss": 0.9586,
+      "step": 3030
+    },
+    {
+      "epoch": 0.23365710761640457,
+      "grad_norm": 3.4887986183166504,
+      "learning_rate": 8.951708400751246e-06,
+      "loss": 0.997,
+      "step": 3031
+    },
+    {
+      "epoch": 0.2337341967314215,
+      "grad_norm": 3.35778546333313,
+      "learning_rate": 8.950943395984959e-06,
+      "loss": 0.922,
+      "step": 3032
+    },
+    {
+      "epoch": 0.23381128584643848,
+      "grad_norm": 4.111606121063232,
+      "learning_rate": 8.950178144897814e-06,
+      "loss": 1.1236,
+      "step": 3033
+    },
+    {
+      "epoch": 0.23388837496145545,
+      "grad_norm": 3.4538538455963135,
+      "learning_rate": 8.949412647537519e-06,
+      "loss": 0.8232,
+      "step": 3034
+    },
+    {
+      "epoch": 0.2339654640764724,
+      "grad_norm": 3.7339892387390137,
+      "learning_rate": 8.948646903951801e-06,
+      "loss": 0.9423,
+      "step": 3035
+    },
+    {
+      "epoch": 0.23404255319148937,
+      "grad_norm": 4.029750347137451,
+      "learning_rate": 8.947880914188397e-06,
+      "loss": 1.1581,
+      "step": 3036
+    },
+    {
+      "epoch": 0.2341196423065063,
+      "grad_norm": 3.9381988048553467,
+      "learning_rate": 8.947114678295066e-06,
+      "loss": 1.0831,
+      "step": 3037
+    },
+    {
+      "epoch": 0.23419673142152328,
+      "grad_norm": 3.86995530128479,
+      "learning_rate": 8.946348196319578e-06,
+      "loss": 1.0168,
+      "step": 3038
+    },
+    {
+      "epoch": 0.23427382053654025,
+      "grad_norm": 3.5996012687683105,
+      "learning_rate": 8.945581468309718e-06,
+      "loss": 0.9506,
+      "step": 3039
+    },
+    {
+      "epoch": 0.2343509096515572,
+      "grad_norm": 4.0110368728637695,
+      "learning_rate": 8.94481449431329e-06,
+      "loss": 1.0381,
+      "step": 3040
+    },
+    {
+      "epoch": 0.23442799876657416,
+      "grad_norm": 4.450444221496582,
+      "learning_rate": 8.944047274378105e-06,
+      "loss": 1.0377,
+      "step": 3041
+    },
+    {
+      "epoch": 0.2345050878815911,
+      "grad_norm": 4.031339645385742,
+      "learning_rate": 8.943279808552e-06,
+      "loss": 1.059,
+      "step": 3042
+    },
+    {
+      "epoch": 0.23458217699660808,
+      "grad_norm": 3.6864678859710693,
+      "learning_rate": 8.942512096882825e-06,
+      "loss": 1.034,
+      "step": 3043
+    },
+    {
+      "epoch": 0.23465926611162505,
+      "grad_norm": 4.179306507110596,
+      "learning_rate": 8.941744139418438e-06,
+      "loss": 1.007,
+      "step": 3044
+    },
+    {
+      "epoch": 0.234736355226642,
+      "grad_norm": 3.602816104888916,
+      "learning_rate": 8.940975936206717e-06,
+      "loss": 0.9409,
+      "step": 3045
+    },
+    {
+      "epoch": 0.23481344434165896,
+      "grad_norm": 3.6022086143493652,
+      "learning_rate": 8.94020748729556e-06,
+      "loss": 1.0962,
+      "step": 3046
+    },
+    {
+      "epoch": 0.2348905334566759,
+      "grad_norm": 3.6227774620056152,
+      "learning_rate": 8.93943879273287e-06,
+      "loss": 1.0189,
+      "step": 3047
+    },
+    {
+      "epoch": 0.23496762257169287,
+      "grad_norm": 3.8395297527313232,
+      "learning_rate": 8.938669852566578e-06,
+      "loss": 0.9859,
+      "step": 3048
+    },
+    {
+      "epoch": 0.23504471168670985,
+      "grad_norm": 3.932426929473877,
+      "learning_rate": 8.937900666844616e-06,
+      "loss": 1.0921,
+      "step": 3049
+    },
+    {
+      "epoch": 0.2351218008017268,
+      "grad_norm": 3.409926414489746,
+      "learning_rate": 8.937131235614945e-06,
+      "loss": 0.9903,
+      "step": 3050
+    },
+    {
+      "epoch": 0.23519888991674376,
+      "grad_norm": 3.986931800842285,
+      "learning_rate": 8.93636155892553e-06,
+      "loss": 1.0701,
+      "step": 3051
+    },
+    {
+      "epoch": 0.23527597903176073,
+      "grad_norm": 3.698172092437744,
+      "learning_rate": 8.935591636824359e-06,
+      "loss": 1.0048,
+      "step": 3052
+    },
+    {
+      "epoch": 0.23535306814677767,
+      "grad_norm": 3.6447818279266357,
+      "learning_rate": 8.934821469359434e-06,
+      "loss": 1.0963,
+      "step": 3053
+    },
+    {
+      "epoch": 0.23543015726179464,
+      "grad_norm": 3.609434127807617,
+      "learning_rate": 8.934051056578768e-06,
+      "loss": 0.9999,
+      "step": 3054
+    },
+    {
+      "epoch": 0.23550724637681159,
+      "grad_norm": 3.5842690467834473,
+      "learning_rate": 8.933280398530394e-06,
+      "loss": 1.1416,
+      "step": 3055
+    },
+    {
+      "epoch": 0.23558433549182856,
+      "grad_norm": 3.902055501937866,
+      "learning_rate": 8.932509495262357e-06,
+      "loss": 0.9909,
+      "step": 3056
+    },
+    {
+      "epoch": 0.23566142460684553,
+      "grad_norm": 4.051197528839111,
+      "learning_rate": 8.931738346822723e-06,
+      "loss": 1.0003,
+      "step": 3057
+    },
+    {
+      "epoch": 0.23573851372186247,
+      "grad_norm": 3.764064311981201,
+      "learning_rate": 8.930966953259563e-06,
+      "loss": 1.0474,
+      "step": 3058
+    },
+    {
+      "epoch": 0.23581560283687944,
+      "grad_norm": 3.6911094188690186,
+      "learning_rate": 8.930195314620975e-06,
+      "loss": 1.0557,
+      "step": 3059
+    },
+    {
+      "epoch": 0.23589269195189638,
+      "grad_norm": 3.983623743057251,
+      "learning_rate": 8.929423430955062e-06,
+      "loss": 1.0087,
+      "step": 3060
+    },
+    {
+      "epoch": 0.23596978106691335,
+      "grad_norm": 3.7596616744995117,
+      "learning_rate": 8.92865130230995e-06,
+      "loss": 1.0966,
+      "step": 3061
+    },
+    {
+      "epoch": 0.23604687018193032,
+      "grad_norm": 3.731916904449463,
+      "learning_rate": 8.927878928733777e-06,
+      "loss": 1.0216,
+      "step": 3062
+    },
+    {
+      "epoch": 0.23612395929694727,
+      "grad_norm": 3.540071487426758,
+      "learning_rate": 8.927106310274695e-06,
+      "loss": 0.9634,
+      "step": 3063
+    },
+    {
+      "epoch": 0.23620104841196424,
+      "grad_norm": 3.379556655883789,
+      "learning_rate": 8.926333446980873e-06,
+      "loss": 0.9989,
+      "step": 3064
+    },
+    {
+      "epoch": 0.23627813752698118,
+      "grad_norm": 3.504272937774658,
+      "learning_rate": 8.925560338900496e-06,
+      "loss": 0.9902,
+      "step": 3065
+    },
+    {
+      "epoch": 0.23635522664199815,
+      "grad_norm": 3.81402325630188,
+      "learning_rate": 8.924786986081764e-06,
+      "loss": 1.0649,
+      "step": 3066
+    },
+    {
+      "epoch": 0.23643231575701512,
+      "grad_norm": 3.7274930477142334,
+      "learning_rate": 8.92401338857289e-06,
+      "loss": 1.0921,
+      "step": 3067
+    },
+    {
+      "epoch": 0.23650940487203206,
+      "grad_norm": 3.597395658493042,
+      "learning_rate": 8.923239546422104e-06,
+      "loss": 0.9825,
+      "step": 3068
+    },
+    {
+      "epoch": 0.23658649398704903,
+      "grad_norm": 3.9207396507263184,
+      "learning_rate": 8.922465459677649e-06,
+      "loss": 0.974,
+      "step": 3069
+    },
+    {
+      "epoch": 0.23666358310206598,
+      "grad_norm": 3.931241035461426,
+      "learning_rate": 8.92169112838779e-06,
+      "loss": 1.0103,
+      "step": 3070
+    },
+    {
+      "epoch": 0.23674067221708295,
+      "grad_norm": 4.226929187774658,
+      "learning_rate": 8.9209165526008e-06,
+      "loss": 1.0043,
+      "step": 3071
+    },
+    {
+      "epoch": 0.23681776133209992,
+      "grad_norm": 3.8953144550323486,
+      "learning_rate": 8.920141732364971e-06,
+      "loss": 1.0027,
+      "step": 3072
+    },
+    {
+      "epoch": 0.23689485044711686,
+      "grad_norm": 3.6181445121765137,
+      "learning_rate": 8.919366667728607e-06,
+      "loss": 1.0596,
+      "step": 3073
+    },
+    {
+      "epoch": 0.23697193956213383,
+      "grad_norm": 4.111177921295166,
+      "learning_rate": 8.918591358740027e-06,
+      "loss": 1.0282,
+      "step": 3074
+    },
+    {
+      "epoch": 0.23704902867715077,
+      "grad_norm": 3.385059118270874,
+      "learning_rate": 8.917815805447574e-06,
+      "loss": 0.9483,
+      "step": 3075
+    },
+    {
+      "epoch": 0.23712611779216775,
+      "grad_norm": 3.6011502742767334,
+      "learning_rate": 8.917040007899596e-06,
+      "loss": 1.0622,
+      "step": 3076
+    },
+    {
+      "epoch": 0.23720320690718472,
+      "grad_norm": 3.875084161758423,
+      "learning_rate": 8.91626396614446e-06,
+      "loss": 0.9209,
+      "step": 3077
+    },
+    {
+      "epoch": 0.23728029602220166,
+      "grad_norm": 3.934360980987549,
+      "learning_rate": 8.91548768023055e-06,
+      "loss": 1.0676,
+      "step": 3078
+    },
+    {
+      "epoch": 0.23735738513721863,
+      "grad_norm": 3.5961403846740723,
+      "learning_rate": 8.91471115020626e-06,
+      "loss": 0.9208,
+      "step": 3079
+    },
+    {
+      "epoch": 0.23743447425223557,
+      "grad_norm": 3.771339178085327,
+      "learning_rate": 8.913934376120006e-06,
+      "loss": 1.0828,
+      "step": 3080
+    },
+    {
+      "epoch": 0.23751156336725254,
+      "grad_norm": 3.8882675170898438,
+      "learning_rate": 8.913157358020212e-06,
+      "loss": 1.0836,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2375886524822695,
+      "grad_norm": 4.413726329803467,
+      "learning_rate": 8.912380095955326e-06,
+      "loss": 1.0002,
+      "step": 3082
+    },
+    {
+      "epoch": 0.23766574159728646,
+      "grad_norm": 4.3546953201293945,
+      "learning_rate": 8.911602589973803e-06,
+      "loss": 1.0706,
+      "step": 3083
+    },
+    {
+      "epoch": 0.23774283071230343,
+      "grad_norm": 4.231818199157715,
+      "learning_rate": 8.910824840124116e-06,
+      "loss": 1.104,
+      "step": 3084
+    },
+    {
+      "epoch": 0.23781991982732037,
+      "grad_norm": 3.49495267868042,
+      "learning_rate": 8.910046846454756e-06,
+      "loss": 0.9728,
+      "step": 3085
+    },
+    {
+      "epoch": 0.23789700894233734,
+      "grad_norm": 3.724919080734253,
+      "learning_rate": 8.909268609014228e-06,
+      "loss": 1.0903,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2379740980573543,
+      "grad_norm": 4.132431983947754,
+      "learning_rate": 8.908490127851047e-06,
+      "loss": 1.0781,
+      "step": 3087
+    },
+    {
+      "epoch": 0.23805118717237125,
+      "grad_norm": 3.652280330657959,
+      "learning_rate": 8.907711403013748e-06,
+      "loss": 1.0157,
+      "step": 3088
+    },
+    {
+      "epoch": 0.23812827628738822,
+      "grad_norm": 3.8804502487182617,
+      "learning_rate": 8.90693243455088e-06,
+      "loss": 1.0803,
+      "step": 3089
+    },
+    {
+      "epoch": 0.23820536540240517,
+      "grad_norm": 3.9575884342193604,
+      "learning_rate": 8.906153222511014e-06,
+      "loss": 1.0079,
+      "step": 3090
+    },
+    {
+      "epoch": 0.23828245451742214,
+      "grad_norm": 3.4292240142822266,
+      "learning_rate": 8.905373766942722e-06,
+      "loss": 1.0367,
+      "step": 3091
+    },
+    {
+      "epoch": 0.2383595436324391,
+      "grad_norm": 3.641467332839966,
+      "learning_rate": 8.904594067894603e-06,
+      "loss": 1.0492,
+      "step": 3092
+    },
+    {
+      "epoch": 0.23843663274745605,
+      "grad_norm": 3.656708240509033,
+      "learning_rate": 8.903814125415267e-06,
+      "loss": 1.0476,
+      "step": 3093
+    },
+    {
+      "epoch": 0.23851372186247302,
+      "grad_norm": 3.6211674213409424,
+      "learning_rate": 8.903033939553336e-06,
+      "loss": 1.0771,
+      "step": 3094
+    },
+    {
+      "epoch": 0.23859081097749,
+      "grad_norm": 3.7082033157348633,
+      "learning_rate": 8.902253510357455e-06,
+      "loss": 1.0533,
+      "step": 3095
+    },
+    {
+      "epoch": 0.23866790009250693,
+      "grad_norm": 3.7945261001586914,
+      "learning_rate": 8.90147283787628e-06,
+      "loss": 0.9977,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2387449892075239,
+      "grad_norm": 3.3629956245422363,
+      "learning_rate": 8.900691922158476e-06,
+      "loss": 0.8274,
+      "step": 3097
+    },
+    {
+      "epoch": 0.23882207832254085,
+      "grad_norm": 3.971607208251953,
+      "learning_rate": 8.899910763252735e-06,
+      "loss": 1.1203,
+      "step": 3098
+    },
+    {
+      "epoch": 0.23889916743755782,
+      "grad_norm": 3.6717000007629395,
+      "learning_rate": 8.899129361207754e-06,
+      "loss": 0.9675,
+      "step": 3099
+    },
+    {
+      "epoch": 0.2389762565525748,
+      "grad_norm": 3.761371374130249,
+      "learning_rate": 8.898347716072254e-06,
+      "loss": 0.9512,
+      "step": 3100
+    },
+    {
+      "epoch": 0.23905334566759173,
+      "grad_norm": 4.109459400177002,
+      "learning_rate": 8.897565827894963e-06,
+      "loss": 1.1068,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2391304347826087,
+      "grad_norm": 3.7636146545410156,
+      "learning_rate": 8.89678369672463e-06,
+      "loss": 0.9598,
+      "step": 3102
+    },
+    {
+      "epoch": 0.23920752389762565,
+      "grad_norm": 3.6051504611968994,
+      "learning_rate": 8.896001322610013e-06,
+      "loss": 1.0457,
+      "step": 3103
+    },
+    {
+      "epoch": 0.23928461301264262,
+      "grad_norm": 3.4127960205078125,
+      "learning_rate": 8.895218705599894e-06,
+      "loss": 0.8803,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2393617021276596,
+      "grad_norm": 3.451889991760254,
+      "learning_rate": 8.89443584574306e-06,
+      "loss": 0.8208,
+      "step": 3105
+    },
+    {
+      "epoch": 0.23943879124267653,
+      "grad_norm": 3.950897455215454,
+      "learning_rate": 8.893652743088321e-06,
+      "loss": 1.0732,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2395158803576935,
+      "grad_norm": 3.8806371688842773,
+      "learning_rate": 8.8928693976845e-06,
+      "loss": 1.0772,
+      "step": 3107
+    },
+    {
+      "epoch": 0.23959296947271044,
+      "grad_norm": 3.3952512741088867,
+      "learning_rate": 8.892085809580435e-06,
+      "loss": 0.998,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2396700585877274,
+      "grad_norm": 4.141806602478027,
+      "learning_rate": 8.891301978824975e-06,
+      "loss": 1.0228,
+      "step": 3109
+    },
+    {
+      "epoch": 0.23974714770274438,
+      "grad_norm": 4.028316974639893,
+      "learning_rate": 8.890517905466991e-06,
+      "loss": 0.9701,
+      "step": 3110
+    },
+    {
+      "epoch": 0.23982423681776133,
+      "grad_norm": 3.646756172180176,
+      "learning_rate": 8.889733589555368e-06,
+      "loss": 1.0763,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2399013259327783,
+      "grad_norm": 3.909493923187256,
+      "learning_rate": 8.888949031138997e-06,
+      "loss": 1.0528,
+      "step": 3112
+    },
+    {
+      "epoch": 0.23997841504779524,
+      "grad_norm": 3.6650424003601074,
+      "learning_rate": 8.888164230266798e-06,
+      "loss": 1.015,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2400555041628122,
+      "grad_norm": 3.6798250675201416,
+      "learning_rate": 8.887379186987695e-06,
+      "loss": 0.9954,
+      "step": 3114
+    },
+    {
+      "epoch": 0.24013259327782918,
+      "grad_norm": 4.122318744659424,
+      "learning_rate": 8.886593901350632e-06,
+      "loss": 1.1222,
+      "step": 3115
+    },
+    {
+      "epoch": 0.24020968239284612,
+      "grad_norm": 3.966783046722412,
+      "learning_rate": 8.885808373404572e-06,
+      "loss": 0.9446,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2402867715078631,
+      "grad_norm": 3.8579766750335693,
+      "learning_rate": 8.885022603198481e-06,
+      "loss": 1.0659,
+      "step": 3117
+    },
+    {
+      "epoch": 0.24036386062288004,
+      "grad_norm": 3.9546401500701904,
+      "learning_rate": 8.884236590781354e-06,
+      "loss": 1.054,
+      "step": 3118
+    },
+    {
+      "epoch": 0.240440949737897,
+      "grad_norm": 3.887936592102051,
+      "learning_rate": 8.883450336202194e-06,
+      "loss": 1.0296,
+      "step": 3119
+    },
+    {
+      "epoch": 0.24051803885291398,
+      "grad_norm": 3.620002031326294,
+      "learning_rate": 8.882663839510017e-06,
+      "loss": 0.8848,
+      "step": 3120
+    },
+    {
+      "epoch": 0.24059512796793092,
+      "grad_norm": 4.082600116729736,
+      "learning_rate": 8.881877100753857e-06,
+      "loss": 1.0739,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2406722170829479,
+      "grad_norm": 4.3715500831604,
+      "learning_rate": 8.881090119982766e-06,
+      "loss": 1.1382,
+      "step": 3122
+    },
+    {
+      "epoch": 0.24074930619796484,
+      "grad_norm": 3.6317408084869385,
+      "learning_rate": 8.880302897245805e-06,
+      "loss": 1.0448,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2408263953129818,
+      "grad_norm": 3.779918909072876,
+      "learning_rate": 8.879515432592057e-06,
+      "loss": 1.0362,
+      "step": 3124
+    },
+    {
+      "epoch": 0.24090348442799878,
+      "grad_norm": 3.709547519683838,
+      "learning_rate": 8.878727726070614e-06,
+      "loss": 0.8987,
+      "step": 3125
+    },
+    {
+      "epoch": 0.24098057354301572,
+      "grad_norm": 3.8204429149627686,
+      "learning_rate": 8.877939777730585e-06,
+      "loss": 1.0463,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2410576626580327,
+      "grad_norm": 3.505138874053955,
+      "learning_rate": 8.877151587621096e-06,
+      "loss": 1.0174,
+      "step": 3127
+    },
+    {
+      "epoch": 0.24113475177304963,
+      "grad_norm": 3.879589557647705,
+      "learning_rate": 8.876363155791286e-06,
+      "loss": 1.0325,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2412118408880666,
+      "grad_norm": 3.7704927921295166,
+      "learning_rate": 8.875574482290308e-06,
+      "loss": 1.0305,
+      "step": 3129
+    },
+    {
+      "epoch": 0.24128893000308357,
+      "grad_norm": 3.636652946472168,
+      "learning_rate": 8.874785567167335e-06,
+      "loss": 0.9645,
+      "step": 3130
+    },
+    {
+      "epoch": 0.24136601911810052,
+      "grad_norm": 3.62375545501709,
+      "learning_rate": 8.873996410471549e-06,
+      "loss": 1.071,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2414431082331175,
+      "grad_norm": 3.3911542892456055,
+      "learning_rate": 8.87320701225215e-06,
+      "loss": 1.0456,
+      "step": 3132
+    },
+    {
+      "epoch": 0.24152019734813443,
+      "grad_norm": 3.846412181854248,
+      "learning_rate": 8.872417372558353e-06,
+      "loss": 0.9716,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2415972864631514,
+      "grad_norm": 3.8464903831481934,
+      "learning_rate": 8.871627491439389e-06,
+      "loss": 0.9607,
+      "step": 3134
+    },
+    {
+      "epoch": 0.24167437557816837,
+      "grad_norm": 3.595994234085083,
+      "learning_rate": 8.870837368944505e-06,
+      "loss": 1.0328,
+      "step": 3135
+    },
+    {
+      "epoch": 0.24175146469318531,
+      "grad_norm": 3.919416904449463,
+      "learning_rate": 8.870047005122957e-06,
+      "loss": 1.1371,
+      "step": 3136
+    },
+    {
+      "epoch": 0.24182855380820228,
+      "grad_norm": 4.039884567260742,
+      "learning_rate": 8.869256400024023e-06,
+      "loss": 1.0813,
+      "step": 3137
+    },
+    {
+      "epoch": 0.24190564292321926,
+      "grad_norm": 3.456634044647217,
+      "learning_rate": 8.86846555369699e-06,
+      "loss": 1.0678,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2419827320382362,
+      "grad_norm": 3.9659950733184814,
+      "learning_rate": 8.867674466191164e-06,
+      "loss": 1.0229,
+      "step": 3139
+    },
+    {
+      "epoch": 0.24205982115325317,
+      "grad_norm": 3.6779067516326904,
+      "learning_rate": 8.866883137555869e-06,
+      "loss": 0.9999,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2421369102682701,
+      "grad_norm": 3.332307815551758,
+      "learning_rate": 8.866091567840436e-06,
+      "loss": 1.0288,
+      "step": 3141
+    },
+    {
+      "epoch": 0.24221399938328708,
+      "grad_norm": 3.756145477294922,
+      "learning_rate": 8.865299757094217e-06,
+      "loss": 1.0806,
+      "step": 3142
+    },
+    {
+      "epoch": 0.24229108849830405,
+      "grad_norm": 3.585340976715088,
+      "learning_rate": 8.864507705366577e-06,
+      "loss": 0.9627,
+      "step": 3143
+    },
+    {
+      "epoch": 0.242368177613321,
+      "grad_norm": 3.905465841293335,
+      "learning_rate": 8.863715412706897e-06,
+      "loss": 1.0693,
+      "step": 3144
+    },
+    {
+      "epoch": 0.24244526672833797,
+      "grad_norm": 4.029237270355225,
+      "learning_rate": 8.862922879164571e-06,
+      "loss": 1.0433,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2425223558433549,
+      "grad_norm": 3.680603265762329,
+      "learning_rate": 8.86213010478901e-06,
+      "loss": 0.9543,
+      "step": 3146
+    },
+    {
+      "epoch": 0.24259944495837188,
+      "grad_norm": 3.9981207847595215,
+      "learning_rate": 8.86133708962964e-06,
+      "loss": 0.9039,
+      "step": 3147
+    },
+    {
+      "epoch": 0.24267653407338885,
+      "grad_norm": 4.111728191375732,
+      "learning_rate": 8.860543833735902e-06,
+      "loss": 0.9712,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2427536231884058,
+      "grad_norm": 4.376039028167725,
+      "learning_rate": 8.859750337157247e-06,
+      "loss": 0.9934,
+      "step": 3149
+    },
+    {
+      "epoch": 0.24283071230342276,
+      "grad_norm": 3.6740853786468506,
+      "learning_rate": 8.858956599943151e-06,
+      "loss": 1.117,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2429078014184397,
+      "grad_norm": 3.9488065242767334,
+      "learning_rate": 8.858162622143096e-06,
+      "loss": 0.9975,
+      "step": 3151
+    },
+    {
+      "epoch": 0.24298489053345668,
+      "grad_norm": 3.6119964122772217,
+      "learning_rate": 8.857368403806586e-06,
+      "loss": 1.0049,
+      "step": 3152
+    },
+    {
+      "epoch": 0.24306197964847365,
+      "grad_norm": 3.641690731048584,
+      "learning_rate": 8.856573944983131e-06,
+      "loss": 1.03,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2431390687634906,
+      "grad_norm": 4.013689041137695,
+      "learning_rate": 8.855779245722264e-06,
+      "loss": 1.0664,
+      "step": 3154
+    },
+    {
+      "epoch": 0.24321615787850756,
+      "grad_norm": 3.818681478500366,
+      "learning_rate": 8.854984306073533e-06,
+      "loss": 0.9579,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2432932469935245,
+      "grad_norm": 4.519876003265381,
+      "learning_rate": 8.854189126086494e-06,
+      "loss": 1.01,
+      "step": 3156
+    },
+    {
+      "epoch": 0.24337033610854147,
+      "grad_norm": 4.186661243438721,
+      "learning_rate": 8.853393705810725e-06,
+      "loss": 0.9834,
+      "step": 3157
+    },
+    {
+      "epoch": 0.24344742522355844,
+      "grad_norm": 3.6581454277038574,
+      "learning_rate": 8.852598045295815e-06,
+      "loss": 1.0094,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2435245143385754,
+      "grad_norm": 3.646156072616577,
+      "learning_rate": 8.851802144591371e-06,
+      "loss": 1.016,
+      "step": 3159
+    },
+    {
+      "epoch": 0.24360160345359236,
+      "grad_norm": 3.637533187866211,
+      "learning_rate": 8.851006003747013e-06,
+      "loss": 1.0252,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2436786925686093,
+      "grad_norm": 3.3992221355438232,
+      "learning_rate": 8.850209622812374e-06,
+      "loss": 0.9184,
+      "step": 3161
+    },
+    {
+      "epoch": 0.24375578168362627,
+      "grad_norm": 3.7890117168426514,
+      "learning_rate": 8.849413001837105e-06,
+      "loss": 1.0386,
+      "step": 3162
+    },
+    {
+      "epoch": 0.24383287079864324,
+      "grad_norm": 3.5062248706817627,
+      "learning_rate": 8.848616140870875e-06,
+      "loss": 1.0384,
+      "step": 3163
+    },
+    {
+      "epoch": 0.24390995991366018,
+      "grad_norm": 3.830472469329834,
+      "learning_rate": 8.847819039963359e-06,
+      "loss": 1.0629,
+      "step": 3164
+    },
+    {
+      "epoch": 0.24398704902867716,
+      "grad_norm": 3.8684401512145996,
+      "learning_rate": 8.847021699164255e-06,
+      "loss": 1.0184,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2440641381436941,
+      "grad_norm": 3.6413917541503906,
+      "learning_rate": 8.846224118523271e-06,
+      "loss": 1.0626,
+      "step": 3166
+    },
+    {
+      "epoch": 0.24414122725871107,
+      "grad_norm": 3.6060991287231445,
+      "learning_rate": 8.845426298090133e-06,
+      "loss": 1.1064,
+      "step": 3167
+    },
+    {
+      "epoch": 0.24421831637372804,
+      "grad_norm": 3.7429842948913574,
+      "learning_rate": 8.844628237914584e-06,
+      "loss": 0.9361,
+      "step": 3168
+    },
+    {
+      "epoch": 0.24429540548874498,
+      "grad_norm": 3.62557053565979,
+      "learning_rate": 8.843829938046376e-06,
+      "loss": 0.9631,
+      "step": 3169
+    },
+    {
+      "epoch": 0.24437249460376195,
+      "grad_norm": 3.561077833175659,
+      "learning_rate": 8.843031398535276e-06,
+      "loss": 1.0517,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2444495837187789,
+      "grad_norm": 3.694340467453003,
+      "learning_rate": 8.842232619431075e-06,
+      "loss": 0.9603,
+      "step": 3171
+    },
+    {
+      "epoch": 0.24452667283379587,
+      "grad_norm": 3.8666601181030273,
+      "learning_rate": 8.841433600783568e-06,
+      "loss": 1.0499,
+      "step": 3172
+    },
+    {
+      "epoch": 0.24460376194881284,
+      "grad_norm": 3.9767086505889893,
+      "learning_rate": 8.840634342642572e-06,
+      "loss": 1.0151,
+      "step": 3173
+    },
+    {
+      "epoch": 0.24468085106382978,
+      "grad_norm": 3.6622021198272705,
+      "learning_rate": 8.839834845057915e-06,
+      "loss": 0.9359,
+      "step": 3174
+    },
+    {
+      "epoch": 0.24475794017884675,
+      "grad_norm": 4.173803329467773,
+      "learning_rate": 8.839035108079443e-06,
+      "loss": 1.0276,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2448350292938637,
+      "grad_norm": 3.558023691177368,
+      "learning_rate": 8.838235131757014e-06,
+      "loss": 0.9519,
+      "step": 3176
+    },
+    {
+      "epoch": 0.24491211840888066,
+      "grad_norm": 3.633117914199829,
+      "learning_rate": 8.837434916140504e-06,
+      "loss": 0.9188,
+      "step": 3177
+    },
+    {
+      "epoch": 0.24498920752389763,
+      "grad_norm": 3.513375997543335,
+      "learning_rate": 8.8366344612798e-06,
+      "loss": 1.0165,
+      "step": 3178
+    },
+    {
+      "epoch": 0.24506629663891458,
+      "grad_norm": 3.9629836082458496,
+      "learning_rate": 8.835833767224811e-06,
+      "loss": 1.0958,
+      "step": 3179
+    },
+    {
+      "epoch": 0.24514338575393155,
+      "grad_norm": 3.611243724822998,
+      "learning_rate": 8.83503283402545e-06,
+      "loss": 1.1016,
+      "step": 3180
+    },
+    {
+      "epoch": 0.24522047486894852,
+      "grad_norm": 3.9855403900146484,
+      "learning_rate": 8.834231661731656e-06,
+      "loss": 1.1526,
+      "step": 3181
+    },
+    {
+      "epoch": 0.24529756398396546,
+      "grad_norm": 3.8271896839141846,
+      "learning_rate": 8.833430250393376e-06,
+      "loss": 0.8604,
+      "step": 3182
+    },
+    {
+      "epoch": 0.24537465309898243,
+      "grad_norm": 4.085109710693359,
+      "learning_rate": 8.832628600060572e-06,
+      "loss": 1.0036,
+      "step": 3183
+    },
+    {
+      "epoch": 0.24545174221399937,
+      "grad_norm": 3.691535711288452,
+      "learning_rate": 8.831826710783226e-06,
+      "loss": 1.1316,
+      "step": 3184
+    },
+    {
+      "epoch": 0.24552883132901634,
+      "grad_norm": 3.610386371612549,
+      "learning_rate": 8.83102458261133e-06,
+      "loss": 1.0648,
+      "step": 3185
+    },
+    {
+      "epoch": 0.24560592044403332,
+      "grad_norm": 3.5450527667999268,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 1.0361,
+      "step": 3186
+    },
+    {
+      "epoch": 0.24568300955905026,
+      "grad_norm": 3.520220994949341,
+      "learning_rate": 8.829419609783936e-06,
+      "loss": 0.8683,
+      "step": 3187
+    },
+    {
+      "epoch": 0.24576009867406723,
+      "grad_norm": 3.9032230377197266,
+      "learning_rate": 8.8286167652285e-06,
+      "loss": 1.0631,
+      "step": 3188
+    },
+    {
+      "epoch": 0.24583718778908417,
+      "grad_norm": 3.8549489974975586,
+      "learning_rate": 8.82781368197864e-06,
+      "loss": 1.0082,
+      "step": 3189
+    },
+    {
+      "epoch": 0.24591427690410114,
+      "grad_norm": 3.810291051864624,
+      "learning_rate": 8.827010360084419e-06,
+      "loss": 0.9334,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2459913660191181,
+      "grad_norm": 3.9244349002838135,
+      "learning_rate": 8.826206799595922e-06,
+      "loss": 1.0369,
+      "step": 3191
+    },
+    {
+      "epoch": 0.24606845513413506,
+      "grad_norm": 3.6987247467041016,
+      "learning_rate": 8.825403000563248e-06,
+      "loss": 0.963,
+      "step": 3192
+    },
+    {
+      "epoch": 0.24614554424915203,
+      "grad_norm": 3.6557397842407227,
+      "learning_rate": 8.824598963036511e-06,
+      "loss": 0.9969,
+      "step": 3193
+    },
+    {
+      "epoch": 0.24622263336416897,
+      "grad_norm": 3.776524066925049,
+      "learning_rate": 8.823794687065836e-06,
+      "loss": 1.0117,
+      "step": 3194
+    },
+    {
+      "epoch": 0.24629972247918594,
+      "grad_norm": 3.713355779647827,
+      "learning_rate": 8.822990172701367e-06,
+      "loss": 0.9675,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2463768115942029,
+      "grad_norm": 3.96952748298645,
+      "learning_rate": 8.82218541999326e-06,
+      "loss": 0.9109,
+      "step": 3196
+    },
+    {
+      "epoch": 0.24645390070921985,
+      "grad_norm": 3.8969883918762207,
+      "learning_rate": 8.821380428991686e-06,
+      "loss": 1.07,
+      "step": 3197
+    },
+    {
+      "epoch": 0.24653098982423682,
+      "grad_norm": 3.5697784423828125,
+      "learning_rate": 8.820575199746835e-06,
+      "loss": 0.9745,
+      "step": 3198
+    },
+    {
+      "epoch": 0.24660807893925377,
+      "grad_norm": 3.8984580039978027,
+      "learning_rate": 8.819769732308908e-06,
+      "loss": 1.0264,
+      "step": 3199
+    },
+    {
+      "epoch": 0.24668516805427074,
+      "grad_norm": 4.095597267150879,
+      "learning_rate": 8.818964026728121e-06,
+      "loss": 1.1447,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2467622571692877,
+      "grad_norm": 3.7202227115631104,
+      "learning_rate": 8.818158083054706e-06,
+      "loss": 0.9318,
+      "step": 3201
+    },
+    {
+      "epoch": 0.24683934628430465,
+      "grad_norm": 3.9623115062713623,
+      "learning_rate": 8.817351901338908e-06,
+      "loss": 1.1674,
+      "step": 3202
+    },
+    {
+      "epoch": 0.24691643539932162,
+      "grad_norm": 4.0379862785339355,
+      "learning_rate": 8.816545481630992e-06,
+      "loss": 0.9857,
+      "step": 3203
+    },
+    {
+      "epoch": 0.24699352451433856,
+      "grad_norm": 3.6850240230560303,
+      "learning_rate": 8.815738823981228e-06,
+      "loss": 0.9235,
+      "step": 3204
+    },
+    {
+      "epoch": 0.24707061362935553,
+      "grad_norm": 3.994870185852051,
+      "learning_rate": 8.814931928439914e-06,
+      "loss": 1.0827,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2471477027443725,
+      "grad_norm": 3.5296335220336914,
+      "learning_rate": 8.81412479505735e-06,
+      "loss": 1.06,
+      "step": 3206
+    },
+    {
+      "epoch": 0.24722479185938945,
+      "grad_norm": 3.7254252433776855,
+      "learning_rate": 8.81331742388386e-06,
+      "loss": 1.0456,
+      "step": 3207
+    },
+    {
+      "epoch": 0.24730188097440642,
+      "grad_norm": 4.138040542602539,
+      "learning_rate": 8.81250981496978e-06,
+      "loss": 1.0802,
+      "step": 3208
+    },
+    {
+      "epoch": 0.24737897008942336,
+      "grad_norm": 4.0760884284973145,
+      "learning_rate": 8.811701968365455e-06,
+      "loss": 1.0602,
+      "step": 3209
+    },
+    {
+      "epoch": 0.24745605920444033,
+      "grad_norm": 3.428788185119629,
+      "learning_rate": 8.810893884121255e-06,
+      "loss": 0.9812,
+      "step": 3210
+    },
+    {
+      "epoch": 0.2475331483194573,
+      "grad_norm": 3.9063422679901123,
+      "learning_rate": 8.81008556228756e-06,
+      "loss": 0.9296,
+      "step": 3211
+    },
+    {
+      "epoch": 0.24761023743447425,
+      "grad_norm": 3.583909749984741,
+      "learning_rate": 8.809277002914762e-06,
+      "loss": 1.0541,
+      "step": 3212
+    },
+    {
+      "epoch": 0.24768732654949122,
+      "grad_norm": 4.865204811096191,
+      "learning_rate": 8.808468206053271e-06,
+      "loss": 1.006,
+      "step": 3213
+    },
+    {
+      "epoch": 0.24776441566450816,
+      "grad_norm": 3.553562641143799,
+      "learning_rate": 8.807659171753513e-06,
+      "loss": 0.9913,
+      "step": 3214
+    },
+    {
+      "epoch": 0.24784150477952513,
+      "grad_norm": 3.982586145401001,
+      "learning_rate": 8.806849900065925e-06,
+      "loss": 1.0069,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2479185938945421,
+      "grad_norm": 4.025597095489502,
+      "learning_rate": 8.806040391040962e-06,
+      "loss": 0.9886,
+      "step": 3216
+    },
+    {
+      "epoch": 0.24799568300955904,
+      "grad_norm": 3.7594804763793945,
+      "learning_rate": 8.805230644729093e-06,
+      "loss": 0.9497,
+      "step": 3217
+    },
+    {
+      "epoch": 0.248072772124576,
+      "grad_norm": 3.5578064918518066,
+      "learning_rate": 8.804420661180801e-06,
+      "loss": 0.9121,
+      "step": 3218
+    },
+    {
+      "epoch": 0.24814986123959296,
+      "grad_norm": 3.8382201194763184,
+      "learning_rate": 8.803610440446584e-06,
+      "loss": 1.0003,
+      "step": 3219
+    },
+    {
+      "epoch": 0.24822695035460993,
+      "grad_norm": 4.217833995819092,
+      "learning_rate": 8.802799982576956e-06,
+      "loss": 1.1424,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2483040394696269,
+      "grad_norm": 3.7423839569091797,
+      "learning_rate": 8.801989287622443e-06,
+      "loss": 1.0629,
+      "step": 3221
+    },
+    {
+      "epoch": 0.24838112858464384,
+      "grad_norm": 3.7119791507720947,
+      "learning_rate": 8.801178355633591e-06,
+      "loss": 1.1239,
+      "step": 3222
+    },
+    {
+      "epoch": 0.2484582176996608,
+      "grad_norm": 3.545254707336426,
+      "learning_rate": 8.800367186660955e-06,
+      "loss": 1.0791,
+      "step": 3223
+    },
+    {
+      "epoch": 0.24853530681467778,
+      "grad_norm": 3.572874069213867,
+      "learning_rate": 8.799555780755108e-06,
+      "loss": 1.0352,
+      "step": 3224
+    },
+    {
+      "epoch": 0.24861239592969472,
+      "grad_norm": 4.043218612670898,
+      "learning_rate": 8.798744137966634e-06,
+      "loss": 1.0846,
+      "step": 3225
+    },
+    {
+      "epoch": 0.2486894850447117,
+      "grad_norm": 3.8153271675109863,
+      "learning_rate": 8.797932258346138e-06,
+      "loss": 1.0246,
+      "step": 3226
+    },
+    {
+      "epoch": 0.24876657415972864,
+      "grad_norm": 3.6751506328582764,
+      "learning_rate": 8.797120141944237e-06,
+      "loss": 0.9848,
+      "step": 3227
+    },
+    {
+      "epoch": 0.2488436632747456,
+      "grad_norm": 3.716115713119507,
+      "learning_rate": 8.79630778881156e-06,
+      "loss": 0.899,
+      "step": 3228
+    },
+    {
+      "epoch": 0.24892075238976258,
+      "grad_norm": 3.8847246170043945,
+      "learning_rate": 8.795495198998753e-06,
+      "loss": 1.0613,
+      "step": 3229
+    },
+    {
+      "epoch": 0.24899784150477952,
+      "grad_norm": 3.672245502471924,
+      "learning_rate": 8.79468237255648e-06,
+      "loss": 1.0189,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2490749306197965,
+      "grad_norm": 4.07139253616333,
+      "learning_rate": 8.793869309535411e-06,
+      "loss": 1.1237,
+      "step": 3231
+    },
+    {
+      "epoch": 0.24915201973481343,
+      "grad_norm": 3.6188600063323975,
+      "learning_rate": 8.793056009986243e-06,
+      "loss": 0.9832,
+      "step": 3232
+    },
+    {
+      "epoch": 0.2492291088498304,
+      "grad_norm": 3.553046703338623,
+      "learning_rate": 8.792242473959676e-06,
+      "loss": 1.0778,
+      "step": 3233
+    },
+    {
+      "epoch": 0.24930619796484738,
+      "grad_norm": 3.8086423873901367,
+      "learning_rate": 8.791428701506433e-06,
+      "loss": 1.0775,
+      "step": 3234
+    },
+    {
+      "epoch": 0.24938328707986432,
+      "grad_norm": 3.5978622436523438,
+      "learning_rate": 8.790614692677244e-06,
+      "loss": 1.0341,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2494603761948813,
+      "grad_norm": 3.933943271636963,
+      "learning_rate": 8.789800447522863e-06,
+      "loss": 1.0398,
+      "step": 3236
+    },
+    {
+      "epoch": 0.24953746530989823,
+      "grad_norm": 3.6028807163238525,
+      "learning_rate": 8.78898596609405e-06,
+      "loss": 0.9908,
+      "step": 3237
+    },
+    {
+      "epoch": 0.2496145544249152,
+      "grad_norm": 3.7010693550109863,
+      "learning_rate": 8.788171248441588e-06,
+      "loss": 0.9977,
+      "step": 3238
+    },
+    {
+      "epoch": 0.24969164353993217,
+      "grad_norm": 3.41683292388916,
+      "learning_rate": 8.787356294616266e-06,
+      "loss": 0.9295,
+      "step": 3239
+    },
+    {
+      "epoch": 0.24976873265494912,
+      "grad_norm": 3.393082618713379,
+      "learning_rate": 8.786541104668895e-06,
+      "loss": 0.9624,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2498458217699661,
+      "grad_norm": 3.504213333129883,
+      "learning_rate": 8.785725678650298e-06,
+      "loss": 1.0289,
+      "step": 3241
+    },
+    {
+      "epoch": 0.24992291088498303,
+      "grad_norm": 3.6844801902770996,
+      "learning_rate": 8.78491001661131e-06,
+      "loss": 0.9146,
+      "step": 3242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.710338830947876,
+      "learning_rate": 8.784094118602788e-06,
+      "loss": 1.0341,
+      "step": 3243
+    },
+    {
+      "epoch": 0.25007708911501697,
+      "grad_norm": 4.066130638122559,
+      "learning_rate": 8.783277984675593e-06,
+      "loss": 1.0393,
+      "step": 3244
+    },
+    {
+      "epoch": 0.25015417823003394,
+      "grad_norm": 3.7111411094665527,
+      "learning_rate": 8.782461614880611e-06,
+      "loss": 1.0909,
+      "step": 3245
+    },
+    {
+      "epoch": 0.25023126734505086,
+      "grad_norm": 3.810129404067993,
+      "learning_rate": 8.781645009268738e-06,
+      "loss": 1.0154,
+      "step": 3246
+    },
+    {
+      "epoch": 0.2503083564600678,
+      "grad_norm": 3.654636859893799,
+      "learning_rate": 8.780828167890882e-06,
+      "loss": 0.9366,
+      "step": 3247
+    },
+    {
+      "epoch": 0.2503854455750848,
+      "grad_norm": 3.8680331707000732,
+      "learning_rate": 8.780011090797974e-06,
+      "loss": 1.0582,
+      "step": 3248
+    },
+    {
+      "epoch": 0.25046253469010177,
+      "grad_norm": 3.812549591064453,
+      "learning_rate": 8.779193778040948e-06,
+      "loss": 0.9273,
+      "step": 3249
+    },
+    {
+      "epoch": 0.25053962380511874,
+      "grad_norm": 3.7380712032318115,
+      "learning_rate": 8.778376229670766e-06,
+      "loss": 0.9623,
+      "step": 3250
+    },
+    {
+      "epoch": 0.25061671292013565,
+      "grad_norm": 4.139057159423828,
+      "learning_rate": 8.777558445738394e-06,
+      "loss": 1.1061,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2506938020351526,
+      "grad_norm": 3.806485891342163,
+      "learning_rate": 8.776740426294818e-06,
+      "loss": 1.0968,
+      "step": 3252
+    },
+    {
+      "epoch": 0.2507708911501696,
+      "grad_norm": 3.924656867980957,
+      "learning_rate": 8.775922171391035e-06,
+      "loss": 1.0496,
+      "step": 3253
+    },
+    {
+      "epoch": 0.25084798026518657,
+      "grad_norm": 3.7743098735809326,
+      "learning_rate": 8.775103681078063e-06,
+      "loss": 1.1391,
+      "step": 3254
+    },
+    {
+      "epoch": 0.25092506938020354,
+      "grad_norm": 3.758241653442383,
+      "learning_rate": 8.774284955406925e-06,
+      "loss": 0.9986,
+      "step": 3255
+    },
+    {
+      "epoch": 0.25100215849522045,
+      "grad_norm": 3.812593460083008,
+      "learning_rate": 8.77346599442867e-06,
+      "loss": 0.9571,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2510792476102374,
+      "grad_norm": 3.9296791553497314,
+      "learning_rate": 8.772646798194353e-06,
+      "loss": 1.0788,
+      "step": 3257
+    },
+    {
+      "epoch": 0.2511563367252544,
+      "grad_norm": 3.9175896644592285,
+      "learning_rate": 8.771827366755046e-06,
+      "loss": 1.1493,
+      "step": 3258
+    },
+    {
+      "epoch": 0.25123342584027136,
+      "grad_norm": 3.4963676929473877,
+      "learning_rate": 8.771007700161839e-06,
+      "loss": 1.0635,
+      "step": 3259
+    },
+    {
+      "epoch": 0.25131051495528833,
+      "grad_norm": 3.776111602783203,
+      "learning_rate": 8.770187798465832e-06,
+      "loss": 1.0474,
+      "step": 3260
+    },
+    {
+      "epoch": 0.25138760407030525,
+      "grad_norm": 3.916110038757324,
+      "learning_rate": 8.769367661718143e-06,
+      "loss": 1.0697,
+      "step": 3261
+    },
+    {
+      "epoch": 0.2514646931853222,
+      "grad_norm": 3.6865622997283936,
+      "learning_rate": 8.7685472899699e-06,
+      "loss": 0.9469,
+      "step": 3262
+    },
+    {
+      "epoch": 0.2515417823003392,
+      "grad_norm": 4.2629594802856445,
+      "learning_rate": 8.767726683272253e-06,
+      "loss": 1.1568,
+      "step": 3263
+    },
+    {
+      "epoch": 0.25161887141535616,
+      "grad_norm": 3.736191511154175,
+      "learning_rate": 8.766905841676361e-06,
+      "loss": 1.0092,
+      "step": 3264
+    },
+    {
+      "epoch": 0.25169596053037313,
+      "grad_norm": 3.6174426078796387,
+      "learning_rate": 8.766084765233399e-06,
+      "loss": 0.9476,
+      "step": 3265
+    },
+    {
+      "epoch": 0.25177304964539005,
+      "grad_norm": 3.8853113651275635,
+      "learning_rate": 8.765263453994556e-06,
+      "loss": 0.9572,
+      "step": 3266
+    },
+    {
+      "epoch": 0.251850138760407,
+      "grad_norm": 5.05413293838501,
+      "learning_rate": 8.764441908011038e-06,
+      "loss": 1.0719,
+      "step": 3267
+    },
+    {
+      "epoch": 0.251927227875424,
+      "grad_norm": 3.739021062850952,
+      "learning_rate": 8.763620127334063e-06,
+      "loss": 0.9814,
+      "step": 3268
+    },
+    {
+      "epoch": 0.25200431699044096,
+      "grad_norm": 3.654526948928833,
+      "learning_rate": 8.762798112014867e-06,
+      "loss": 1.0566,
+      "step": 3269
+    },
+    {
+      "epoch": 0.25208140610545793,
+      "grad_norm": 4.029781818389893,
+      "learning_rate": 8.761975862104694e-06,
+      "loss": 1.1511,
+      "step": 3270
+    },
+    {
+      "epoch": 0.25215849522047484,
+      "grad_norm": 4.01661491394043,
+      "learning_rate": 8.761153377654811e-06,
+      "loss": 1.0468,
+      "step": 3271
+    },
+    {
+      "epoch": 0.2522355843354918,
+      "grad_norm": 3.568437337875366,
+      "learning_rate": 8.760330658716497e-06,
+      "loss": 0.9401,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2523126734505088,
+      "grad_norm": 3.7211058139801025,
+      "learning_rate": 8.75950770534104e-06,
+      "loss": 1.0863,
+      "step": 3273
+    },
+    {
+      "epoch": 0.25238976256552575,
+      "grad_norm": 3.9265799522399902,
+      "learning_rate": 8.758684517579746e-06,
+      "loss": 1.0554,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2524668516805427,
+      "grad_norm": 3.979198694229126,
+      "learning_rate": 8.757861095483942e-06,
+      "loss": 1.0998,
+      "step": 3275
+    },
+    {
+      "epoch": 0.25254394079555964,
+      "grad_norm": 3.5346436500549316,
+      "learning_rate": 8.75703743910496e-06,
+      "loss": 0.9852,
+      "step": 3276
+    },
+    {
+      "epoch": 0.2526210299105766,
+      "grad_norm": 3.676072835922241,
+      "learning_rate": 8.756213548494152e-06,
+      "loss": 0.9912,
+      "step": 3277
+    },
+    {
+      "epoch": 0.2526981190255936,
+      "grad_norm": 3.5911636352539062,
+      "learning_rate": 8.755389423702884e-06,
+      "loss": 0.9803,
+      "step": 3278
+    },
+    {
+      "epoch": 0.25277520814061055,
+      "grad_norm": 3.92370343208313,
+      "learning_rate": 8.754565064782533e-06,
+      "loss": 0.9514,
+      "step": 3279
+    },
+    {
+      "epoch": 0.2528522972556275,
+      "grad_norm": 3.5190653800964355,
+      "learning_rate": 8.753740471784497e-06,
+      "loss": 1.0362,
+      "step": 3280
+    },
+    {
+      "epoch": 0.25292938637064444,
+      "grad_norm": 4.347536087036133,
+      "learning_rate": 8.752915644760185e-06,
+      "loss": 1.0841,
+      "step": 3281
+    },
+    {
+      "epoch": 0.2530064754856614,
+      "grad_norm": 3.641761064529419,
+      "learning_rate": 8.752090583761017e-06,
+      "loss": 1.0113,
+      "step": 3282
+    },
+    {
+      "epoch": 0.2530835646006784,
+      "grad_norm": 3.5918169021606445,
+      "learning_rate": 8.751265288838435e-06,
+      "loss": 0.9651,
+      "step": 3283
+    },
+    {
+      "epoch": 0.25316065371569535,
+      "grad_norm": 3.788519859313965,
+      "learning_rate": 8.750439760043892e-06,
+      "loss": 1.0641,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2532377428307123,
+      "grad_norm": 3.539388656616211,
+      "learning_rate": 8.749613997428852e-06,
+      "loss": 1.1261,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2533148319457293,
+      "grad_norm": 4.3505167961120605,
+      "learning_rate": 8.748788001044799e-06,
+      "loss": 1.1336,
+      "step": 3286
+    },
+    {
+      "epoch": 0.2533919210607462,
+      "grad_norm": 3.823928117752075,
+      "learning_rate": 8.74796177094323e-06,
+      "loss": 1.0041,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2534690101757632,
+      "grad_norm": 3.804438829421997,
+      "learning_rate": 8.747135307175657e-06,
+      "loss": 1.0816,
+      "step": 3288
+    },
+    {
+      "epoch": 0.25354609929078015,
+      "grad_norm": 3.7979674339294434,
+      "learning_rate": 8.746308609793601e-06,
+      "loss": 1.0633,
+      "step": 3289
+    },
+    {
+      "epoch": 0.2536231884057971,
+      "grad_norm": 3.7804579734802246,
+      "learning_rate": 8.745481678848609e-06,
+      "loss": 1.09,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2537002775208141,
+      "grad_norm": 4.1087470054626465,
+      "learning_rate": 8.744654514392232e-06,
+      "loss": 1.1016,
+      "step": 3291
+    },
+    {
+      "epoch": 0.253777366635831,
+      "grad_norm": 3.6745924949645996,
+      "learning_rate": 8.743827116476039e-06,
+      "loss": 1.0273,
+      "step": 3292
+    },
+    {
+      "epoch": 0.253854455750848,
+      "grad_norm": 3.827929735183716,
+      "learning_rate": 8.742999485151617e-06,
+      "loss": 1.1064,
+      "step": 3293
+    },
+    {
+      "epoch": 0.25393154486586494,
+      "grad_norm": 3.5309934616088867,
+      "learning_rate": 8.742171620470561e-06,
+      "loss": 0.9478,
+      "step": 3294
+    },
+    {
+      "epoch": 0.2540086339808819,
+      "grad_norm": 3.4990084171295166,
+      "learning_rate": 8.741343522484486e-06,
+      "loss": 0.9285,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2540857230958989,
+      "grad_norm": 3.40860915184021,
+      "learning_rate": 8.74051519124502e-06,
+      "loss": 0.8972,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2541628122109158,
+      "grad_norm": 3.403424024581909,
+      "learning_rate": 8.739686626803802e-06,
+      "loss": 1.059,
+      "step": 3297
+    },
+    {
+      "epoch": 0.25423990132593277,
+      "grad_norm": 4.193357944488525,
+      "learning_rate": 8.738857829212495e-06,
+      "loss": 0.9641,
+      "step": 3298
+    },
+    {
+      "epoch": 0.25431699044094974,
+      "grad_norm": 4.222657203674316,
+      "learning_rate": 8.738028798522762e-06,
+      "loss": 1.0695,
+      "step": 3299
+    },
+    {
+      "epoch": 0.2543940795559667,
+      "grad_norm": 3.387126922607422,
+      "learning_rate": 8.737199534786297e-06,
+      "loss": 1.1003,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2544711686709837,
+      "grad_norm": 3.9024600982666016,
+      "learning_rate": 8.736370038054796e-06,
+      "loss": 1.0785,
+      "step": 3301
+    },
+    {
+      "epoch": 0.2545482577860006,
+      "grad_norm": 3.78934383392334,
+      "learning_rate": 8.735540308379974e-06,
+      "loss": 1.0066,
+      "step": 3302
+    },
+    {
+      "epoch": 0.25462534690101757,
+      "grad_norm": 3.661975383758545,
+      "learning_rate": 8.73471034581356e-06,
+      "loss": 1.0309,
+      "step": 3303
+    },
+    {
+      "epoch": 0.25470243601603454,
+      "grad_norm": 3.687753915786743,
+      "learning_rate": 8.733880150407296e-06,
+      "loss": 1.0173,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2547795251310515,
+      "grad_norm": 3.935116767883301,
+      "learning_rate": 8.733049722212946e-06,
+      "loss": 1.1019,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2548566142460685,
+      "grad_norm": 3.9937877655029297,
+      "learning_rate": 8.73221906128228e-06,
+      "loss": 1.1622,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2549337033610854,
+      "grad_norm": 3.838475227355957,
+      "learning_rate": 8.731388167667083e-06,
+      "loss": 0.977,
+      "step": 3307
+    },
+    {
+      "epoch": 0.25501079247610237,
+      "grad_norm": 3.6274874210357666,
+      "learning_rate": 8.73055704141916e-06,
+      "loss": 1.098,
+      "step": 3308
+    },
+    {
+      "epoch": 0.25508788159111934,
+      "grad_norm": 3.570650577545166,
+      "learning_rate": 8.729725682590329e-06,
+      "loss": 0.9655,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2551649707061363,
+      "grad_norm": 3.7613351345062256,
+      "learning_rate": 8.728894091232417e-06,
+      "loss": 0.9714,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2552420598211533,
+      "grad_norm": 3.4735591411590576,
+      "learning_rate": 8.728062267397268e-06,
+      "loss": 0.8959,
+      "step": 3311
+    },
+    {
+      "epoch": 0.2553191489361702,
+      "grad_norm": 3.8954594135284424,
+      "learning_rate": 8.727230211136747e-06,
+      "loss": 1.0228,
+      "step": 3312
+    },
+    {
+      "epoch": 0.25539623805118716,
+      "grad_norm": 4.0239949226379395,
+      "learning_rate": 8.726397922502727e-06,
+      "loss": 1.0255,
+      "step": 3313
+    },
+    {
+      "epoch": 0.25547332716620413,
+      "grad_norm": 3.6205320358276367,
+      "learning_rate": 8.725565401547096e-06,
+      "loss": 0.8827,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2555504162812211,
+      "grad_norm": 3.7505264282226562,
+      "learning_rate": 8.724732648321756e-06,
+      "loss": 1.0148,
+      "step": 3315
+    },
+    {
+      "epoch": 0.2556275053962381,
+      "grad_norm": 3.5387229919433594,
+      "learning_rate": 8.723899662878627e-06,
+      "loss": 0.9492,
+      "step": 3316
+    },
+    {
+      "epoch": 0.255704594511255,
+      "grad_norm": 3.82112193107605,
+      "learning_rate": 8.72306644526964e-06,
+      "loss": 0.9834,
+      "step": 3317
+    },
+    {
+      "epoch": 0.25578168362627196,
+      "grad_norm": 3.7389678955078125,
+      "learning_rate": 8.722232995546742e-06,
+      "loss": 0.9932,
+      "step": 3318
+    },
+    {
+      "epoch": 0.25585877274128893,
+      "grad_norm": 3.6923506259918213,
+      "learning_rate": 8.721399313761896e-06,
+      "loss": 0.9353,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2559358618563059,
+      "grad_norm": 3.9490480422973633,
+      "learning_rate": 8.720565399967076e-06,
+      "loss": 1.0178,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2560129509713229,
+      "grad_norm": 3.4159951210021973,
+      "learning_rate": 8.719731254214271e-06,
+      "loss": 0.967,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2560900400863398,
+      "grad_norm": 3.7180635929107666,
+      "learning_rate": 8.71889687655549e-06,
+      "loss": 0.902,
+      "step": 3322
+    },
+    {
+      "epoch": 0.25616712920135676,
+      "grad_norm": 4.072498321533203,
+      "learning_rate": 8.718062267042749e-06,
+      "loss": 1.0151,
+      "step": 3323
+    },
+    {
+      "epoch": 0.25624421831637373,
+      "grad_norm": 4.77862548828125,
+      "learning_rate": 8.71722742572808e-06,
+      "loss": 1.0888,
+      "step": 3324
+    },
+    {
+      "epoch": 0.2563213074313907,
+      "grad_norm": 3.858112096786499,
+      "learning_rate": 8.716392352663535e-06,
+      "loss": 1.015,
+      "step": 3325
+    },
+    {
+      "epoch": 0.25639839654640767,
+      "grad_norm": 3.6426641941070557,
+      "learning_rate": 8.715557047901174e-06,
+      "loss": 1.0053,
+      "step": 3326
+    },
+    {
+      "epoch": 0.2564754856614246,
+      "grad_norm": 3.9015743732452393,
+      "learning_rate": 8.714721511493074e-06,
+      "loss": 1.2689,
+      "step": 3327
+    },
+    {
+      "epoch": 0.25655257477644156,
+      "grad_norm": 4.314653396606445,
+      "learning_rate": 8.713885743491327e-06,
+      "loss": 1.0598,
+      "step": 3328
+    },
+    {
+      "epoch": 0.2566296638914585,
+      "grad_norm": 4.168483257293701,
+      "learning_rate": 8.713049743948038e-06,
+      "loss": 1.1855,
+      "step": 3329
+    },
+    {
+      "epoch": 0.2567067530064755,
+      "grad_norm": 4.1147356033325195,
+      "learning_rate": 8.71221351291533e-06,
+      "loss": 1.0772,
+      "step": 3330
+    },
+    {
+      "epoch": 0.25678384212149247,
+      "grad_norm": 4.284569263458252,
+      "learning_rate": 8.711377050445333e-06,
+      "loss": 0.9965,
+      "step": 3331
+    },
+    {
+      "epoch": 0.2568609312365094,
+      "grad_norm": 4.178953647613525,
+      "learning_rate": 8.710540356590198e-06,
+      "loss": 1.056,
+      "step": 3332
+    },
+    {
+      "epoch": 0.25693802035152635,
+      "grad_norm": 3.9649317264556885,
+      "learning_rate": 8.709703431402092e-06,
+      "loss": 0.9488,
+      "step": 3333
+    },
+    {
+      "epoch": 0.2570151094665433,
+      "grad_norm": 3.9904897212982178,
+      "learning_rate": 8.70886627493319e-06,
+      "loss": 1.033,
+      "step": 3334
+    },
+    {
+      "epoch": 0.2570921985815603,
+      "grad_norm": 3.681057929992676,
+      "learning_rate": 8.708028887235682e-06,
+      "loss": 0.9747,
+      "step": 3335
+    },
+    {
+      "epoch": 0.25716928769657726,
+      "grad_norm": 3.687772750854492,
+      "learning_rate": 8.707191268361778e-06,
+      "loss": 0.9872,
+      "step": 3336
+    },
+    {
+      "epoch": 0.2572463768115942,
+      "grad_norm": 3.913752794265747,
+      "learning_rate": 8.7063534183637e-06,
+      "loss": 1.085,
+      "step": 3337
+    },
+    {
+      "epoch": 0.25732346592661115,
+      "grad_norm": 4.11881685256958,
+      "learning_rate": 8.705515337293682e-06,
+      "loss": 1.0011,
+      "step": 3338
+    },
+    {
+      "epoch": 0.2574005550416281,
+      "grad_norm": 3.4855170249938965,
+      "learning_rate": 8.704677025203972e-06,
+      "loss": 0.9699,
+      "step": 3339
+    },
+    {
+      "epoch": 0.2574776441566451,
+      "grad_norm": 3.873643636703491,
+      "learning_rate": 8.703838482146837e-06,
+      "loss": 1.0217,
+      "step": 3340
+    },
+    {
+      "epoch": 0.25755473327166206,
+      "grad_norm": 3.451205015182495,
+      "learning_rate": 8.702999708174557e-06,
+      "loss": 0.9768,
+      "step": 3341
+    },
+    {
+      "epoch": 0.257631822386679,
+      "grad_norm": 3.673360586166382,
+      "learning_rate": 8.702160703339422e-06,
+      "loss": 1.0151,
+      "step": 3342
+    },
+    {
+      "epoch": 0.25770891150169595,
+      "grad_norm": 3.864973306655884,
+      "learning_rate": 8.701321467693741e-06,
+      "loss": 1.0275,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2577860006167129,
+      "grad_norm": 3.8561880588531494,
+      "learning_rate": 8.700482001289838e-06,
+      "loss": 1.0697,
+      "step": 3344
+    },
+    {
+      "epoch": 0.2578630897317299,
+      "grad_norm": 4.72223424911499,
+      "learning_rate": 8.699642304180045e-06,
+      "loss": 1.1577,
+      "step": 3345
+    },
+    {
+      "epoch": 0.25794017884674686,
+      "grad_norm": 3.793133020401001,
+      "learning_rate": 8.698802376416718e-06,
+      "loss": 0.9992,
+      "step": 3346
+    },
+    {
+      "epoch": 0.2580172679617638,
+      "grad_norm": 3.360287666320801,
+      "learning_rate": 8.697962218052217e-06,
+      "loss": 0.9239,
+      "step": 3347
+    },
+    {
+      "epoch": 0.25809435707678074,
+      "grad_norm": 3.907996416091919,
+      "learning_rate": 8.697121829138925e-06,
+      "loss": 0.9704,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2581714461917977,
+      "grad_norm": 3.603130578994751,
+      "learning_rate": 8.696281209729234e-06,
+      "loss": 0.9822,
+      "step": 3349
+    },
+    {
+      "epoch": 0.2582485353068147,
+      "grad_norm": 3.7168142795562744,
+      "learning_rate": 8.695440359875554e-06,
+      "loss": 1.0306,
+      "step": 3350
+    },
+    {
+      "epoch": 0.25832562442183166,
+      "grad_norm": 3.7132720947265625,
+      "learning_rate": 8.694599279630306e-06,
+      "loss": 1.0643,
+      "step": 3351
+    },
+    {
+      "epoch": 0.25840271353684857,
+      "grad_norm": 3.9718921184539795,
+      "learning_rate": 8.693757969045928e-06,
+      "loss": 0.9594,
+      "step": 3352
+    },
+    {
+      "epoch": 0.25847980265186554,
+      "grad_norm": 3.7555110454559326,
+      "learning_rate": 8.692916428174872e-06,
+      "loss": 1.0797,
+      "step": 3353
+    },
+    {
+      "epoch": 0.2585568917668825,
+      "grad_norm": 3.6487925052642822,
+      "learning_rate": 8.692074657069602e-06,
+      "loss": 0.9086,
+      "step": 3354
+    },
+    {
+      "epoch": 0.2586339808818995,
+      "grad_norm": 3.890230178833008,
+      "learning_rate": 8.6912326557826e-06,
+      "loss": 1.0957,
+      "step": 3355
+    },
+    {
+      "epoch": 0.25871106999691645,
+      "grad_norm": 3.4611878395080566,
+      "learning_rate": 8.690390424366358e-06,
+      "loss": 0.9665,
+      "step": 3356
+    },
+    {
+      "epoch": 0.25878815911193337,
+      "grad_norm": 3.949563980102539,
+      "learning_rate": 8.689547962873386e-06,
+      "loss": 0.9853,
+      "step": 3357
+    },
+    {
+      "epoch": 0.25886524822695034,
+      "grad_norm": 3.7639212608337402,
+      "learning_rate": 8.688705271356208e-06,
+      "loss": 1.0879,
+      "step": 3358
+    },
+    {
+      "epoch": 0.2589423373419673,
+      "grad_norm": 3.696392297744751,
+      "learning_rate": 8.68786234986736e-06,
+      "loss": 0.9405,
+      "step": 3359
+    },
+    {
+      "epoch": 0.2590194264569843,
+      "grad_norm": 4.061792850494385,
+      "learning_rate": 8.687019198459395e-06,
+      "loss": 1.0346,
+      "step": 3360
+    },
+    {
+      "epoch": 0.25909651557200125,
+      "grad_norm": 3.8702781200408936,
+      "learning_rate": 8.686175817184878e-06,
+      "loss": 1.0629,
+      "step": 3361
+    },
+    {
+      "epoch": 0.25917360468701817,
+      "grad_norm": 3.8583784103393555,
+      "learning_rate": 8.685332206096391e-06,
+      "loss": 1.0525,
+      "step": 3362
+    },
+    {
+      "epoch": 0.25925069380203514,
+      "grad_norm": 3.554065465927124,
+      "learning_rate": 8.684488365246526e-06,
+      "loss": 0.9169,
+      "step": 3363
+    },
+    {
+      "epoch": 0.2593277829170521,
+      "grad_norm": 3.972473382949829,
+      "learning_rate": 8.683644294687893e-06,
+      "loss": 1.012,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2594048720320691,
+      "grad_norm": 4.070230484008789,
+      "learning_rate": 8.68279999447312e-06,
+      "loss": 0.9605,
+      "step": 3365
+    },
+    {
+      "epoch": 0.25948196114708605,
+      "grad_norm": 3.5376861095428467,
+      "learning_rate": 8.681955464654839e-06,
+      "loss": 0.8759,
+      "step": 3366
+    },
+    {
+      "epoch": 0.25955905026210296,
+      "grad_norm": 3.8391125202178955,
+      "learning_rate": 8.681110705285705e-06,
+      "loss": 0.989,
+      "step": 3367
+    },
+    {
+      "epoch": 0.25963613937711993,
+      "grad_norm": 3.567213296890259,
+      "learning_rate": 8.680265716418381e-06,
+      "loss": 1.0166,
+      "step": 3368
+    },
+    {
+      "epoch": 0.2597132284921369,
+      "grad_norm": 3.484469413757324,
+      "learning_rate": 8.679420498105553e-06,
+      "loss": 0.9996,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2597903176071539,
+      "grad_norm": 3.5866034030914307,
+      "learning_rate": 8.678575050399912e-06,
+      "loss": 0.9088,
+      "step": 3370
+    },
+    {
+      "epoch": 0.25986740672217085,
+      "grad_norm": 3.686366319656372,
+      "learning_rate": 8.677729373354169e-06,
+      "loss": 1.0459,
+      "step": 3371
+    },
+    {
+      "epoch": 0.2599444958371878,
+      "grad_norm": 3.622049570083618,
+      "learning_rate": 8.676883467021046e-06,
+      "loss": 1.1149,
+      "step": 3372
+    },
+    {
+      "epoch": 0.26002158495220473,
+      "grad_norm": 3.5664947032928467,
+      "learning_rate": 8.676037331453283e-06,
+      "loss": 1.0691,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2600986740672217,
+      "grad_norm": 3.5930142402648926,
+      "learning_rate": 8.675190966703631e-06,
+      "loss": 1.0553,
+      "step": 3374
+    },
+    {
+      "epoch": 0.2601757631822387,
+      "grad_norm": 3.7880308628082275,
+      "learning_rate": 8.674344372824855e-06,
+      "loss": 1.024,
+      "step": 3375
+    },
+    {
+      "epoch": 0.26025285229725564,
+      "grad_norm": 4.235076427459717,
+      "learning_rate": 8.673497549869738e-06,
+      "loss": 1.0325,
+      "step": 3376
+    },
+    {
+      "epoch": 0.2603299414122726,
+      "grad_norm": 3.9409871101379395,
+      "learning_rate": 8.672650497891075e-06,
+      "loss": 1.0099,
+      "step": 3377
+    },
+    {
+      "epoch": 0.26040703052728953,
+      "grad_norm": 3.7281947135925293,
+      "learning_rate": 8.671803216941674e-06,
+      "loss": 1.0297,
+      "step": 3378
+    },
+    {
+      "epoch": 0.2604841196423065,
+      "grad_norm": 4.101794719696045,
+      "learning_rate": 8.67095570707436e-06,
+      "loss": 1.0915,
+      "step": 3379
+    },
+    {
+      "epoch": 0.26056120875732347,
+      "grad_norm": 3.782160997390747,
+      "learning_rate": 8.67010796834197e-06,
+      "loss": 1.0381,
+      "step": 3380
+    },
+    {
+      "epoch": 0.26063829787234044,
+      "grad_norm": 3.6193161010742188,
+      "learning_rate": 8.669260000797355e-06,
+      "loss": 0.9774,
+      "step": 3381
+    },
+    {
+      "epoch": 0.2607153869873574,
+      "grad_norm": 3.8134310245513916,
+      "learning_rate": 8.668411804493384e-06,
+      "loss": 1.0437,
+      "step": 3382
+    },
+    {
+      "epoch": 0.2607924761023743,
+      "grad_norm": 3.7592546939849854,
+      "learning_rate": 8.667563379482934e-06,
+      "loss": 0.922,
+      "step": 3383
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 3.698714256286621,
+      "learning_rate": 8.666714725818903e-06,
+      "loss": 0.9594,
+      "step": 3384
+    },
+    {
+      "epoch": 0.26094665433240827,
+      "grad_norm": 3.6574625968933105,
+      "learning_rate": 8.6658658435542e-06,
+      "loss": 1.0836,
+      "step": 3385
+    },
+    {
+      "epoch": 0.26102374344742524,
+      "grad_norm": 3.509253740310669,
+      "learning_rate": 8.665016732741748e-06,
+      "loss": 1.0096,
+      "step": 3386
+    },
+    {
+      "epoch": 0.2611008325624422,
+      "grad_norm": 3.7333405017852783,
+      "learning_rate": 8.664167393434484e-06,
+      "loss": 1.1459,
+      "step": 3387
+    },
+    {
+      "epoch": 0.2611779216774591,
+      "grad_norm": 3.782824993133545,
+      "learning_rate": 8.66331782568536e-06,
+      "loss": 1.0318,
+      "step": 3388
+    },
+    {
+      "epoch": 0.2612550107924761,
+      "grad_norm": 3.6097073554992676,
+      "learning_rate": 8.662468029547341e-06,
+      "loss": 0.8997,
+      "step": 3389
+    },
+    {
+      "epoch": 0.26133209990749307,
+      "grad_norm": 3.526141405105591,
+      "learning_rate": 8.661618005073412e-06,
+      "loss": 0.9869,
+      "step": 3390
+    },
+    {
+      "epoch": 0.26140918902251004,
+      "grad_norm": 3.5635085105895996,
+      "learning_rate": 8.66076775231656e-06,
+      "loss": 0.9146,
+      "step": 3391
+    },
+    {
+      "epoch": 0.261486278137527,
+      "grad_norm": 3.9402894973754883,
+      "learning_rate": 8.659917271329801e-06,
+      "loss": 1.1514,
+      "step": 3392
+    },
+    {
+      "epoch": 0.2615633672525439,
+      "grad_norm": 4.007856845855713,
+      "learning_rate": 8.659066562166157e-06,
+      "loss": 1.0842,
+      "step": 3393
+    },
+    {
+      "epoch": 0.2616404563675609,
+      "grad_norm": 3.9499897956848145,
+      "learning_rate": 8.65821562487866e-06,
+      "loss": 1.0736,
+      "step": 3394
+    },
+    {
+      "epoch": 0.26171754548257786,
+      "grad_norm": 4.446572303771973,
+      "learning_rate": 8.657364459520367e-06,
+      "loss": 1.0391,
+      "step": 3395
+    },
+    {
+      "epoch": 0.26179463459759483,
+      "grad_norm": 3.946072816848755,
+      "learning_rate": 8.656513066144342e-06,
+      "loss": 1.1018,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2618717237126118,
+      "grad_norm": 3.971186637878418,
+      "learning_rate": 8.655661444803664e-06,
+      "loss": 1.0861,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2619488128276287,
+      "grad_norm": 3.788757801055908,
+      "learning_rate": 8.654809595551429e-06,
+      "loss": 1.1189,
+      "step": 3398
+    },
+    {
+      "epoch": 0.2620259019426457,
+      "grad_norm": 3.536916732788086,
+      "learning_rate": 8.653957518440743e-06,
+      "loss": 0.9544,
+      "step": 3399
+    },
+    {
+      "epoch": 0.26210299105766266,
+      "grad_norm": 3.843003749847412,
+      "learning_rate": 8.653105213524733e-06,
+      "loss": 0.9992,
+      "step": 3400
+    },
+    {
+      "epoch": 0.26218008017267963,
+      "grad_norm": 3.6438448429107666,
+      "learning_rate": 8.65225268085653e-06,
+      "loss": 1.0176,
+      "step": 3401
+    },
+    {
+      "epoch": 0.2622571692876966,
+      "grad_norm": 4.137069225311279,
+      "learning_rate": 8.65139992048929e-06,
+      "loss": 1.0994,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2623342584027135,
+      "grad_norm": 3.9459714889526367,
+      "learning_rate": 8.650546932476173e-06,
+      "loss": 1.0303,
+      "step": 3403
+    },
+    {
+      "epoch": 0.2624113475177305,
+      "grad_norm": 3.64534068107605,
+      "learning_rate": 8.649693716870364e-06,
+      "loss": 0.9538,
+      "step": 3404
+    },
+    {
+      "epoch": 0.26248843663274746,
+      "grad_norm": 3.5429468154907227,
+      "learning_rate": 8.648840273725055e-06,
+      "loss": 0.9813,
+      "step": 3405
+    },
+    {
+      "epoch": 0.26256552574776443,
+      "grad_norm": 3.9356422424316406,
+      "learning_rate": 8.64798660309345e-06,
+      "loss": 1.0192,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2626426148627814,
+      "grad_norm": 3.7008917331695557,
+      "learning_rate": 8.647132705028776e-06,
+      "loss": 1.0213,
+      "step": 3407
+    },
+    {
+      "epoch": 0.2627197039777983,
+      "grad_norm": 3.5144479274749756,
+      "learning_rate": 8.646278579584265e-06,
+      "loss": 0.9657,
+      "step": 3408
+    },
+    {
+      "epoch": 0.2627967930928153,
+      "grad_norm": 3.8584909439086914,
+      "learning_rate": 8.64542422681317e-06,
+      "loss": 1.1071,
+      "step": 3409
+    },
+    {
+      "epoch": 0.26287388220783225,
+      "grad_norm": 3.93780779838562,
+      "learning_rate": 8.644569646768755e-06,
+      "loss": 1.0209,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2629509713228492,
+      "grad_norm": 3.6556358337402344,
+      "learning_rate": 8.643714839504296e-06,
+      "loss": 1.0174,
+      "step": 3411
+    },
+    {
+      "epoch": 0.2630280604378662,
+      "grad_norm": 4.030715465545654,
+      "learning_rate": 8.64285980507309e-06,
+      "loss": 1.0071,
+      "step": 3412
+    },
+    {
+      "epoch": 0.2631051495528831,
+      "grad_norm": 3.5671634674072266,
+      "learning_rate": 8.642004543528442e-06,
+      "loss": 1.0613,
+      "step": 3413
+    },
+    {
+      "epoch": 0.2631822386679001,
+      "grad_norm": 3.487668991088867,
+      "learning_rate": 8.641149054923673e-06,
+      "loss": 0.9665,
+      "step": 3414
+    },
+    {
+      "epoch": 0.26325932778291705,
+      "grad_norm": 3.8126344680786133,
+      "learning_rate": 8.640293339312119e-06,
+      "loss": 0.9724,
+      "step": 3415
+    },
+    {
+      "epoch": 0.263336416897934,
+      "grad_norm": 3.6022071838378906,
+      "learning_rate": 8.639437396747127e-06,
+      "loss": 1.0247,
+      "step": 3416
+    },
+    {
+      "epoch": 0.263413506012951,
+      "grad_norm": 3.6988253593444824,
+      "learning_rate": 8.638581227282064e-06,
+      "loss": 1.0596,
+      "step": 3417
+    },
+    {
+      "epoch": 0.2634905951279679,
+      "grad_norm": 3.569603204727173,
+      "learning_rate": 8.637724830970307e-06,
+      "loss": 1.0096,
+      "step": 3418
+    },
+    {
+      "epoch": 0.2635676842429849,
+      "grad_norm": 3.479275941848755,
+      "learning_rate": 8.636868207865244e-06,
+      "loss": 1.0446,
+      "step": 3419
+    },
+    {
+      "epoch": 0.26364477335800185,
+      "grad_norm": 3.994907855987549,
+      "learning_rate": 8.636011358020286e-06,
+      "loss": 0.9947,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2637218624730188,
+      "grad_norm": 3.892772912979126,
+      "learning_rate": 8.635154281488851e-06,
+      "loss": 1.1448,
+      "step": 3421
+    },
+    {
+      "epoch": 0.2637989515880358,
+      "grad_norm": 3.7717888355255127,
+      "learning_rate": 8.634296978324374e-06,
+      "loss": 1.0227,
+      "step": 3422
+    },
+    {
+      "epoch": 0.2638760407030527,
+      "grad_norm": 3.6846694946289062,
+      "learning_rate": 8.633439448580302e-06,
+      "loss": 0.9988,
+      "step": 3423
+    },
+    {
+      "epoch": 0.2639531298180697,
+      "grad_norm": 3.5329177379608154,
+      "learning_rate": 8.6325816923101e-06,
+      "loss": 0.9628,
+      "step": 3424
+    },
+    {
+      "epoch": 0.26403021893308665,
+      "grad_norm": 3.6999776363372803,
+      "learning_rate": 8.631723709567242e-06,
+      "loss": 1.0493,
+      "step": 3425
+    },
+    {
+      "epoch": 0.2641073080481036,
+      "grad_norm": 3.73701810836792,
+      "learning_rate": 8.630865500405218e-06,
+      "loss": 1.0119,
+      "step": 3426
+    },
+    {
+      "epoch": 0.2641843971631206,
+      "grad_norm": 3.6966707706451416,
+      "learning_rate": 8.630007064877538e-06,
+      "loss": 1.0202,
+      "step": 3427
+    },
+    {
+      "epoch": 0.2642614862781375,
+      "grad_norm": 3.5695645809173584,
+      "learning_rate": 8.629148403037715e-06,
+      "loss": 0.9439,
+      "step": 3428
+    },
+    {
+      "epoch": 0.2643385753931545,
+      "grad_norm": 3.6183085441589355,
+      "learning_rate": 8.628289514939287e-06,
+      "loss": 1.0232,
+      "step": 3429
+    },
+    {
+      "epoch": 0.26441566450817144,
+      "grad_norm": 3.7045369148254395,
+      "learning_rate": 8.6274304006358e-06,
+      "loss": 0.932,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2644927536231884,
+      "grad_norm": 3.988922119140625,
+      "learning_rate": 8.626571060180812e-06,
+      "loss": 1.024,
+      "step": 3431
+    },
+    {
+      "epoch": 0.2645698427382054,
+      "grad_norm": 3.669790029525757,
+      "learning_rate": 8.625711493627902e-06,
+      "loss": 1.0481,
+      "step": 3432
+    },
+    {
+      "epoch": 0.2646469318532223,
+      "grad_norm": 3.9124808311462402,
+      "learning_rate": 8.62485170103066e-06,
+      "loss": 1.0879,
+      "step": 3433
+    },
+    {
+      "epoch": 0.26472402096823927,
+      "grad_norm": 3.7043819427490234,
+      "learning_rate": 8.623991682442685e-06,
+      "loss": 1.0116,
+      "step": 3434
+    },
+    {
+      "epoch": 0.26480111008325624,
+      "grad_norm": 3.8101680278778076,
+      "learning_rate": 8.623131437917598e-06,
+      "loss": 1.0168,
+      "step": 3435
+    },
+    {
+      "epoch": 0.2648781991982732,
+      "grad_norm": 3.652815103530884,
+      "learning_rate": 8.622270967509032e-06,
+      "loss": 1.1259,
+      "step": 3436
+    },
+    {
+      "epoch": 0.2649552883132902,
+      "grad_norm": 3.873318910598755,
+      "learning_rate": 8.621410271270632e-06,
+      "loss": 1.1589,
+      "step": 3437
+    },
+    {
+      "epoch": 0.2650323774283071,
+      "grad_norm": 3.489452362060547,
+      "learning_rate": 8.620549349256056e-06,
+      "loss": 0.9174,
+      "step": 3438
+    },
+    {
+      "epoch": 0.26510946654332407,
+      "grad_norm": 3.7744300365448,
+      "learning_rate": 8.61968820151898e-06,
+      "loss": 1.0291,
+      "step": 3439
+    },
+    {
+      "epoch": 0.26518655565834104,
+      "grad_norm": 3.8608028888702393,
+      "learning_rate": 8.61882682811309e-06,
+      "loss": 1.0441,
+      "step": 3440
+    },
+    {
+      "epoch": 0.265263644773358,
+      "grad_norm": 3.826913595199585,
+      "learning_rate": 8.61796522909209e-06,
+      "loss": 1.058,
+      "step": 3441
+    },
+    {
+      "epoch": 0.265340733888375,
+      "grad_norm": 4.063562393188477,
+      "learning_rate": 8.617103404509699e-06,
+      "loss": 0.9656,
+      "step": 3442
+    },
+    {
+      "epoch": 0.2654178230033919,
+      "grad_norm": 4.010993957519531,
+      "learning_rate": 8.616241354419642e-06,
+      "loss": 1.1192,
+      "step": 3443
+    },
+    {
+      "epoch": 0.26549491211840887,
+      "grad_norm": 4.178236961364746,
+      "learning_rate": 8.615379078875664e-06,
+      "loss": 0.9984,
+      "step": 3444
+    },
+    {
+      "epoch": 0.26557200123342584,
+      "grad_norm": 4.172964572906494,
+      "learning_rate": 8.614516577931526e-06,
+      "loss": 1.0439,
+      "step": 3445
+    },
+    {
+      "epoch": 0.2656490903484428,
+      "grad_norm": 3.271082878112793,
+      "learning_rate": 8.613653851641001e-06,
+      "loss": 0.9694,
+      "step": 3446
+    },
+    {
+      "epoch": 0.2657261794634598,
+      "grad_norm": 3.594104766845703,
+      "learning_rate": 8.612790900057873e-06,
+      "loss": 1.0078,
+      "step": 3447
+    },
+    {
+      "epoch": 0.2658032685784767,
+      "grad_norm": 3.8845951557159424,
+      "learning_rate": 8.611927723235943e-06,
+      "loss": 1.1074,
+      "step": 3448
+    },
+    {
+      "epoch": 0.26588035769349366,
+      "grad_norm": 3.545318365097046,
+      "learning_rate": 8.611064321229027e-06,
+      "loss": 1.0609,
+      "step": 3449
+    },
+    {
+      "epoch": 0.26595744680851063,
+      "grad_norm": 3.784323215484619,
+      "learning_rate": 8.610200694090951e-06,
+      "loss": 0.9997,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2660345359235276,
+      "grad_norm": 3.6356170177459717,
+      "learning_rate": 8.609336841875561e-06,
+      "loss": 0.9148,
+      "step": 3451
+    },
+    {
+      "epoch": 0.2661116250385446,
+      "grad_norm": 4.235599040985107,
+      "learning_rate": 8.608472764636714e-06,
+      "loss": 1.1123,
+      "step": 3452
+    },
+    {
+      "epoch": 0.2661887141535615,
+      "grad_norm": 3.789675712585449,
+      "learning_rate": 8.607608462428273e-06,
+      "loss": 1.0678,
+      "step": 3453
+    },
+    {
+      "epoch": 0.26626580326857846,
+      "grad_norm": 3.547661542892456,
+      "learning_rate": 8.606743935304134e-06,
+      "loss": 1.0874,
+      "step": 3454
+    },
+    {
+      "epoch": 0.26634289238359543,
+      "grad_norm": 3.812304735183716,
+      "learning_rate": 8.605879183318188e-06,
+      "loss": 0.9825,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2664199814986124,
+      "grad_norm": 3.6646666526794434,
+      "learning_rate": 8.605014206524352e-06,
+      "loss": 1.097,
+      "step": 3456
+    },
+    {
+      "epoch": 0.26649707061362937,
+      "grad_norm": 3.943058490753174,
+      "learning_rate": 8.60414900497655e-06,
+      "loss": 0.9482,
+      "step": 3457
+    },
+    {
+      "epoch": 0.26657415972864634,
+      "grad_norm": 3.732710838317871,
+      "learning_rate": 8.603283578728723e-06,
+      "loss": 1.0008,
+      "step": 3458
+    },
+    {
+      "epoch": 0.26665124884366326,
+      "grad_norm": 3.5974366664886475,
+      "learning_rate": 8.60241792783483e-06,
+      "loss": 1.0185,
+      "step": 3459
+    },
+    {
+      "epoch": 0.26672833795868023,
+      "grad_norm": 3.4493045806884766,
+      "learning_rate": 8.601552052348833e-06,
+      "loss": 1.0396,
+      "step": 3460
+    },
+    {
+      "epoch": 0.2668054270736972,
+      "grad_norm": 3.9309725761413574,
+      "learning_rate": 8.60068595232472e-06,
+      "loss": 0.9397,
+      "step": 3461
+    },
+    {
+      "epoch": 0.26688251618871417,
+      "grad_norm": 3.614811420440674,
+      "learning_rate": 8.599819627816486e-06,
+      "loss": 1.0507,
+      "step": 3462
+    },
+    {
+      "epoch": 0.26695960530373114,
+      "grad_norm": 3.847226619720459,
+      "learning_rate": 8.598953078878142e-06,
+      "loss": 1.0411,
+      "step": 3463
+    },
+    {
+      "epoch": 0.26703669441874806,
+      "grad_norm": 4.487602233886719,
+      "learning_rate": 8.598086305563714e-06,
+      "loss": 1.1693,
+      "step": 3464
+    },
+    {
+      "epoch": 0.267113783533765,
+      "grad_norm": 3.617213487625122,
+      "learning_rate": 8.597219307927239e-06,
+      "loss": 1.0972,
+      "step": 3465
+    },
+    {
+      "epoch": 0.267190872648782,
+      "grad_norm": 4.097979545593262,
+      "learning_rate": 8.59635208602277e-06,
+      "loss": 1.0642,
+      "step": 3466
+    },
+    {
+      "epoch": 0.26726796176379897,
+      "grad_norm": 3.6943976879119873,
+      "learning_rate": 8.595484639904375e-06,
+      "loss": 1.0495,
+      "step": 3467
+    },
+    {
+      "epoch": 0.26734505087881594,
+      "grad_norm": 4.322422981262207,
+      "learning_rate": 8.594616969626134e-06,
+      "loss": 1.0781,
+      "step": 3468
+    },
+    {
+      "epoch": 0.26742213999383285,
+      "grad_norm": 3.314831495285034,
+      "learning_rate": 8.593749075242143e-06,
+      "loss": 0.8829,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2674992291088498,
+      "grad_norm": 3.7989742755889893,
+      "learning_rate": 8.592880956806509e-06,
+      "loss": 0.9417,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2675763182238668,
+      "grad_norm": 3.5251259803771973,
+      "learning_rate": 8.592012614373355e-06,
+      "loss": 0.9525,
+      "step": 3471
+    },
+    {
+      "epoch": 0.26765340733888376,
+      "grad_norm": 3.975076198577881,
+      "learning_rate": 8.591144047996817e-06,
+      "loss": 1.1792,
+      "step": 3472
+    },
+    {
+      "epoch": 0.26773049645390073,
+      "grad_norm": 3.9003539085388184,
+      "learning_rate": 8.59027525773105e-06,
+      "loss": 1.0923,
+      "step": 3473
+    },
+    {
+      "epoch": 0.26780758556891765,
+      "grad_norm": 3.6139657497406006,
+      "learning_rate": 8.589406243630212e-06,
+      "loss": 0.9957,
+      "step": 3474
+    },
+    {
+      "epoch": 0.2678846746839346,
+      "grad_norm": 4.016913414001465,
+      "learning_rate": 8.588537005748484e-06,
+      "loss": 1.1282,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2679617637989516,
+      "grad_norm": 3.5826406478881836,
+      "learning_rate": 8.587667544140063e-06,
+      "loss": 1.0285,
+      "step": 3476
+    },
+    {
+      "epoch": 0.26803885291396856,
+      "grad_norm": 3.5834274291992188,
+      "learning_rate": 8.586797858859149e-06,
+      "loss": 0.941,
+      "step": 3477
+    },
+    {
+      "epoch": 0.26811594202898553,
+      "grad_norm": 3.8702480792999268,
+      "learning_rate": 8.585927949959965e-06,
+      "loss": 1.0519,
+      "step": 3478
+    },
+    {
+      "epoch": 0.26819303114400245,
+      "grad_norm": 3.2615435123443604,
+      "learning_rate": 8.585057817496747e-06,
+      "loss": 0.9371,
+      "step": 3479
+    },
+    {
+      "epoch": 0.2682701202590194,
+      "grad_norm": 3.683424949645996,
+      "learning_rate": 8.584187461523741e-06,
+      "loss": 0.9695,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2683472093740364,
+      "grad_norm": 3.533581018447876,
+      "learning_rate": 8.583316882095209e-06,
+      "loss": 1.0256,
+      "step": 3481
+    },
+    {
+      "epoch": 0.26842429848905336,
+      "grad_norm": 3.866887331008911,
+      "learning_rate": 8.582446079265428e-06,
+      "loss": 1.115,
+      "step": 3482
+    },
+    {
+      "epoch": 0.26850138760407033,
+      "grad_norm": 3.444634437561035,
+      "learning_rate": 8.581575053088687e-06,
+      "loss": 0.9583,
+      "step": 3483
+    },
+    {
+      "epoch": 0.26857847671908724,
+      "grad_norm": 3.9610986709594727,
+      "learning_rate": 8.580703803619292e-06,
+      "loss": 0.9974,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2686555658341042,
+      "grad_norm": 3.6639208793640137,
+      "learning_rate": 8.57983233091156e-06,
+      "loss": 1.0442,
+      "step": 3485
+    },
+    {
+      "epoch": 0.2687326549491212,
+      "grad_norm": 3.690537929534912,
+      "learning_rate": 8.578960635019822e-06,
+      "loss": 1.0666,
+      "step": 3486
+    },
+    {
+      "epoch": 0.26880974406413816,
+      "grad_norm": 3.4930202960968018,
+      "learning_rate": 8.578088715998425e-06,
+      "loss": 0.9999,
+      "step": 3487
+    },
+    {
+      "epoch": 0.2688868331791551,
+      "grad_norm": 3.7361273765563965,
+      "learning_rate": 8.577216573901727e-06,
+      "loss": 0.9615,
+      "step": 3488
+    },
+    {
+      "epoch": 0.26896392229417204,
+      "grad_norm": 3.7041215896606445,
+      "learning_rate": 8.576344208784104e-06,
+      "loss": 1.1125,
+      "step": 3489
+    },
+    {
+      "epoch": 0.269041011409189,
+      "grad_norm": 3.4133498668670654,
+      "learning_rate": 8.575471620699942e-06,
+      "loss": 0.9674,
+      "step": 3490
+    },
+    {
+      "epoch": 0.269118100524206,
+      "grad_norm": 3.8262181282043457,
+      "learning_rate": 8.57459880970364e-06,
+      "loss": 1.1,
+      "step": 3491
+    },
+    {
+      "epoch": 0.26919518963922295,
+      "grad_norm": 3.9435431957244873,
+      "learning_rate": 8.573725775849617e-06,
+      "loss": 1.1601,
+      "step": 3492
+    },
+    {
+      "epoch": 0.2692722787542399,
+      "grad_norm": 4.27089786529541,
+      "learning_rate": 8.5728525191923e-06,
+      "loss": 1.085,
+      "step": 3493
+    },
+    {
+      "epoch": 0.26934936786925684,
+      "grad_norm": 3.4078094959259033,
+      "learning_rate": 8.571979039786135e-06,
+      "loss": 0.9794,
+      "step": 3494
+    },
+    {
+      "epoch": 0.2694264569842738,
+      "grad_norm": 3.4169962406158447,
+      "learning_rate": 8.571105337685575e-06,
+      "loss": 0.9074,
+      "step": 3495
+    },
+    {
+      "epoch": 0.2695035460992908,
+      "grad_norm": 3.312422752380371,
+      "learning_rate": 8.570231412945092e-06,
+      "loss": 0.9965,
+      "step": 3496
+    },
+    {
+      "epoch": 0.26958063521430775,
+      "grad_norm": 4.496835231781006,
+      "learning_rate": 8.569357265619172e-06,
+      "loss": 1.0674,
+      "step": 3497
+    },
+    {
+      "epoch": 0.2696577243293247,
+      "grad_norm": 3.7908670902252197,
+      "learning_rate": 8.56848289576231e-06,
+      "loss": 0.9237,
+      "step": 3498
+    },
+    {
+      "epoch": 0.26973481344434164,
+      "grad_norm": 3.9191906452178955,
+      "learning_rate": 8.567608303429024e-06,
+      "loss": 0.9648,
+      "step": 3499
+    },
+    {
+      "epoch": 0.2698119025593586,
+      "grad_norm": 3.607534408569336,
+      "learning_rate": 8.566733488673837e-06,
+      "loss": 1.0898,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2698889916743756,
+      "grad_norm": 4.031132698059082,
+      "learning_rate": 8.56585845155129e-06,
+      "loss": 0.9965,
+      "step": 3501
+    },
+    {
+      "epoch": 0.26996608078939255,
+      "grad_norm": 3.5480751991271973,
+      "learning_rate": 8.564983192115934e-06,
+      "loss": 0.9761,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2700431699044095,
+      "grad_norm": 4.214536666870117,
+      "learning_rate": 8.564107710422343e-06,
+      "loss": 1.0331,
+      "step": 3503
+    },
+    {
+      "epoch": 0.27012025901942643,
+      "grad_norm": 3.7173125743865967,
+      "learning_rate": 8.563232006525093e-06,
+      "loss": 1.0236,
+      "step": 3504
+    },
+    {
+      "epoch": 0.2701973481344434,
+      "grad_norm": 4.071146011352539,
+      "learning_rate": 8.562356080478781e-06,
+      "loss": 1.0041,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2702744372494604,
+      "grad_norm": 3.6947145462036133,
+      "learning_rate": 8.56147993233802e-06,
+      "loss": 1.0646,
+      "step": 3506
+    },
+    {
+      "epoch": 0.27035152636447735,
+      "grad_norm": 3.899477243423462,
+      "learning_rate": 8.560603562157428e-06,
+      "loss": 0.9962,
+      "step": 3507
+    },
+    {
+      "epoch": 0.2704286154794943,
+      "grad_norm": 3.6149401664733887,
+      "learning_rate": 8.559726969991646e-06,
+      "loss": 0.9407,
+      "step": 3508
+    },
+    {
+      "epoch": 0.27050570459451123,
+      "grad_norm": 3.5481820106506348,
+      "learning_rate": 8.558850155895325e-06,
+      "loss": 0.9833,
+      "step": 3509
+    },
+    {
+      "epoch": 0.2705827937095282,
+      "grad_norm": 3.8647804260253906,
+      "learning_rate": 8.557973119923126e-06,
+      "loss": 0.9107,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2706598828245452,
+      "grad_norm": 4.967718601226807,
+      "learning_rate": 8.557095862129732e-06,
+      "loss": 1.1314,
+      "step": 3511
+    },
+    {
+      "epoch": 0.27073697193956214,
+      "grad_norm": 3.775641918182373,
+      "learning_rate": 8.556218382569832e-06,
+      "loss": 0.9328,
+      "step": 3512
+    },
+    {
+      "epoch": 0.2708140610545791,
+      "grad_norm": 3.7387449741363525,
+      "learning_rate": 8.555340681298136e-06,
+      "loss": 1.074,
+      "step": 3513
+    },
+    {
+      "epoch": 0.27089115016959603,
+      "grad_norm": 3.9792206287384033,
+      "learning_rate": 8.554462758369362e-06,
+      "loss": 1.0373,
+      "step": 3514
+    },
+    {
+      "epoch": 0.270968239284613,
+      "grad_norm": 4.240675926208496,
+      "learning_rate": 8.553584613838243e-06,
+      "loss": 1.0737,
+      "step": 3515
+    },
+    {
+      "epoch": 0.27104532839962997,
+      "grad_norm": 3.890157699584961,
+      "learning_rate": 8.552706247759527e-06,
+      "loss": 0.9872,
+      "step": 3516
+    },
+    {
+      "epoch": 0.27112241751464694,
+      "grad_norm": 3.4992425441741943,
+      "learning_rate": 8.551827660187978e-06,
+      "loss": 0.9768,
+      "step": 3517
+    },
+    {
+      "epoch": 0.2711995066296639,
+      "grad_norm": 3.600581407546997,
+      "learning_rate": 8.550948851178368e-06,
+      "loss": 0.9799,
+      "step": 3518
+    },
+    {
+      "epoch": 0.2712765957446808,
+      "grad_norm": 4.245118141174316,
+      "learning_rate": 8.55006982078549e-06,
+      "loss": 1.1133,
+      "step": 3519
+    },
+    {
+      "epoch": 0.2713536848596978,
+      "grad_norm": 3.7988290786743164,
+      "learning_rate": 8.549190569064144e-06,
+      "loss": 0.9969,
+      "step": 3520
+    },
+    {
+      "epoch": 0.27143077397471477,
+      "grad_norm": 4.244655132293701,
+      "learning_rate": 8.548311096069148e-06,
+      "loss": 1.1576,
+      "step": 3521
+    },
+    {
+      "epoch": 0.27150786308973174,
+      "grad_norm": 3.901003360748291,
+      "learning_rate": 8.547431401855333e-06,
+      "loss": 1.0142,
+      "step": 3522
+    },
+    {
+      "epoch": 0.2715849522047487,
+      "grad_norm": 3.8016793727874756,
+      "learning_rate": 8.546551486477542e-06,
+      "loss": 1.0619,
+      "step": 3523
+    },
+    {
+      "epoch": 0.2716620413197656,
+      "grad_norm": 3.773573637008667,
+      "learning_rate": 8.545671349990633e-06,
+      "loss": 1.0336,
+      "step": 3524
+    },
+    {
+      "epoch": 0.2717391304347826,
+      "grad_norm": 3.533284902572632,
+      "learning_rate": 8.544790992449479e-06,
+      "loss": 1.1181,
+      "step": 3525
+    },
+    {
+      "epoch": 0.27181621954979956,
+      "grad_norm": 3.5303707122802734,
+      "learning_rate": 8.543910413908967e-06,
+      "loss": 0.9646,
+      "step": 3526
+    },
+    {
+      "epoch": 0.27189330866481654,
+      "grad_norm": 3.93528413772583,
+      "learning_rate": 8.543029614423994e-06,
+      "loss": 1.0863,
+      "step": 3527
+    },
+    {
+      "epoch": 0.2719703977798335,
+      "grad_norm": 3.518523931503296,
+      "learning_rate": 8.542148594049475e-06,
+      "loss": 1.0908,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2720474868948504,
+      "grad_norm": 3.6154558658599854,
+      "learning_rate": 8.541267352840336e-06,
+      "loss": 1.0267,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2721245760098674,
+      "grad_norm": 3.5488035678863525,
+      "learning_rate": 8.54038589085152e-06,
+      "loss": 1.0351,
+      "step": 3530
+    },
+    {
+      "epoch": 0.27220166512488436,
+      "grad_norm": 3.697274923324585,
+      "learning_rate": 8.539504208137977e-06,
+      "loss": 1.0168,
+      "step": 3531
+    },
+    {
+      "epoch": 0.27227875423990133,
+      "grad_norm": 3.5824666023254395,
+      "learning_rate": 8.538622304754679e-06,
+      "loss": 1.0174,
+      "step": 3532
+    },
+    {
+      "epoch": 0.2723558433549183,
+      "grad_norm": 3.9082562923431396,
+      "learning_rate": 8.537740180756608e-06,
+      "loss": 1.1437,
+      "step": 3533
+    },
+    {
+      "epoch": 0.2724329324699352,
+      "grad_norm": 3.7677783966064453,
+      "learning_rate": 8.536857836198759e-06,
+      "loss": 0.9499,
+      "step": 3534
+    },
+    {
+      "epoch": 0.2725100215849522,
+      "grad_norm": 3.8008973598480225,
+      "learning_rate": 8.535975271136142e-06,
+      "loss": 0.9116,
+      "step": 3535
+    },
+    {
+      "epoch": 0.27258711069996916,
+      "grad_norm": 3.5162312984466553,
+      "learning_rate": 8.53509248562378e-06,
+      "loss": 0.9755,
+      "step": 3536
+    },
+    {
+      "epoch": 0.27266419981498613,
+      "grad_norm": 3.512768030166626,
+      "learning_rate": 8.534209479716708e-06,
+      "loss": 0.9291,
+      "step": 3537
+    },
+    {
+      "epoch": 0.2727412889300031,
+      "grad_norm": 3.6765835285186768,
+      "learning_rate": 8.533326253469983e-06,
+      "loss": 0.9966,
+      "step": 3538
+    },
+    {
+      "epoch": 0.27281837804502,
+      "grad_norm": 3.5264675617218018,
+      "learning_rate": 8.532442806938663e-06,
+      "loss": 0.985,
+      "step": 3539
+    },
+    {
+      "epoch": 0.272895467160037,
+      "grad_norm": 3.555009365081787,
+      "learning_rate": 8.531559140177828e-06,
+      "loss": 1.0493,
+      "step": 3540
+    },
+    {
+      "epoch": 0.27297255627505396,
+      "grad_norm": 3.558711528778076,
+      "learning_rate": 8.530675253242573e-06,
+      "loss": 0.8952,
+      "step": 3541
+    },
+    {
+      "epoch": 0.2730496453900709,
+      "grad_norm": 3.6628801822662354,
+      "learning_rate": 8.529791146188003e-06,
+      "loss": 1.0614,
+      "step": 3542
+    },
+    {
+      "epoch": 0.2731267345050879,
+      "grad_norm": 3.692908525466919,
+      "learning_rate": 8.528906819069234e-06,
+      "loss": 1.0244,
+      "step": 3543
+    },
+    {
+      "epoch": 0.27320382362010487,
+      "grad_norm": 3.419480085372925,
+      "learning_rate": 8.528022271941404e-06,
+      "loss": 0.9742,
+      "step": 3544
+    },
+    {
+      "epoch": 0.2732809127351218,
+      "grad_norm": 3.549691915512085,
+      "learning_rate": 8.527137504859654e-06,
+      "loss": 0.9707,
+      "step": 3545
+    },
+    {
+      "epoch": 0.27335800185013875,
+      "grad_norm": 3.792904853820801,
+      "learning_rate": 8.52625251787915e-06,
+      "loss": 1.008,
+      "step": 3546
+    },
+    {
+      "epoch": 0.2734350909651557,
+      "grad_norm": 3.560368776321411,
+      "learning_rate": 8.525367311055063e-06,
+      "loss": 0.961,
+      "step": 3547
+    },
+    {
+      "epoch": 0.2735121800801727,
+      "grad_norm": 3.6131019592285156,
+      "learning_rate": 8.524481884442583e-06,
+      "loss": 0.9666,
+      "step": 3548
+    },
+    {
+      "epoch": 0.27358926919518967,
+      "grad_norm": 3.8992390632629395,
+      "learning_rate": 8.523596238096913e-06,
+      "loss": 1.0684,
+      "step": 3549
+    },
+    {
+      "epoch": 0.2736663583102066,
+      "grad_norm": 3.5087990760803223,
+      "learning_rate": 8.522710372073265e-06,
+      "loss": 0.9931,
+      "step": 3550
+    },
+    {
+      "epoch": 0.27374344742522355,
+      "grad_norm": 4.000070095062256,
+      "learning_rate": 8.521824286426872e-06,
+      "loss": 1.0817,
+      "step": 3551
+    },
+    {
+      "epoch": 0.2738205365402405,
+      "grad_norm": 3.854365825653076,
+      "learning_rate": 8.520937981212973e-06,
+      "loss": 0.9886,
+      "step": 3552
+    },
+    {
+      "epoch": 0.2738976256552575,
+      "grad_norm": 3.266382932662964,
+      "learning_rate": 8.520051456486827e-06,
+      "loss": 0.9731,
+      "step": 3553
+    },
+    {
+      "epoch": 0.27397471477027446,
+      "grad_norm": 3.62668514251709,
+      "learning_rate": 8.519164712303703e-06,
+      "loss": 1.0892,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2740518038852914,
+      "grad_norm": 3.941357135772705,
+      "learning_rate": 8.518277748718887e-06,
+      "loss": 1.0447,
+      "step": 3555
+    },
+    {
+      "epoch": 0.27412889300030835,
+      "grad_norm": 3.726266860961914,
+      "learning_rate": 8.517390565787672e-06,
+      "loss": 1.0465,
+      "step": 3556
+    },
+    {
+      "epoch": 0.2742059821153253,
+      "grad_norm": 3.8830811977386475,
+      "learning_rate": 8.516503163565374e-06,
+      "loss": 1.0176,
+      "step": 3557
+    },
+    {
+      "epoch": 0.2742830712303423,
+      "grad_norm": 3.832172393798828,
+      "learning_rate": 8.515615542107317e-06,
+      "loss": 1.066,
+      "step": 3558
+    },
+    {
+      "epoch": 0.27436016034535926,
+      "grad_norm": 3.768930196762085,
+      "learning_rate": 8.514727701468837e-06,
+      "loss": 1.2196,
+      "step": 3559
+    },
+    {
+      "epoch": 0.2744372494603762,
+      "grad_norm": 3.7359375953674316,
+      "learning_rate": 8.513839641705288e-06,
+      "loss": 1.0866,
+      "step": 3560
+    },
+    {
+      "epoch": 0.27451433857539315,
+      "grad_norm": 3.686213970184326,
+      "learning_rate": 8.512951362872037e-06,
+      "loss": 1.072,
+      "step": 3561
+    },
+    {
+      "epoch": 0.2745914276904101,
+      "grad_norm": 3.735572338104248,
+      "learning_rate": 8.512062865024463e-06,
+      "loss": 0.9229,
+      "step": 3562
+    },
+    {
+      "epoch": 0.2746685168054271,
+      "grad_norm": 3.594026803970337,
+      "learning_rate": 8.511174148217958e-06,
+      "loss": 1.0068,
+      "step": 3563
+    },
+    {
+      "epoch": 0.27474560592044406,
+      "grad_norm": 3.5880775451660156,
+      "learning_rate": 8.51028521250793e-06,
+      "loss": 1.0385,
+      "step": 3564
+    },
+    {
+      "epoch": 0.274822695035461,
+      "grad_norm": 3.8719637393951416,
+      "learning_rate": 8.509396057949799e-06,
+      "loss": 1.0011,
+      "step": 3565
+    },
+    {
+      "epoch": 0.27489978415047794,
+      "grad_norm": 3.555837631225586,
+      "learning_rate": 8.508506684598999e-06,
+      "loss": 0.9061,
+      "step": 3566
+    },
+    {
+      "epoch": 0.2749768732654949,
+      "grad_norm": 3.370540142059326,
+      "learning_rate": 8.50761709251098e-06,
+      "loss": 0.9201,
+      "step": 3567
+    },
+    {
+      "epoch": 0.2750539623805119,
+      "grad_norm": 3.50242280960083,
+      "learning_rate": 8.5067272817412e-06,
+      "loss": 0.9497,
+      "step": 3568
+    },
+    {
+      "epoch": 0.27513105149552886,
+      "grad_norm": 3.7081899642944336,
+      "learning_rate": 8.505837252345135e-06,
+      "loss": 0.9182,
+      "step": 3569
+    },
+    {
+      "epoch": 0.27520814061054577,
+      "grad_norm": 3.7657527923583984,
+      "learning_rate": 8.504947004378277e-06,
+      "loss": 1.0016,
+      "step": 3570
+    },
+    {
+      "epoch": 0.27528522972556274,
+      "grad_norm": 3.5126376152038574,
+      "learning_rate": 8.504056537896123e-06,
+      "loss": 0.93,
+      "step": 3571
+    },
+    {
+      "epoch": 0.2753623188405797,
+      "grad_norm": 3.6629865169525146,
+      "learning_rate": 8.503165852954193e-06,
+      "loss": 1.0295,
+      "step": 3572
+    },
+    {
+      "epoch": 0.2754394079555967,
+      "grad_norm": 3.717280149459839,
+      "learning_rate": 8.502274949608018e-06,
+      "loss": 1.0295,
+      "step": 3573
+    },
+    {
+      "epoch": 0.27551649707061365,
+      "grad_norm": 4.10526180267334,
+      "learning_rate": 8.501383827913137e-06,
+      "loss": 1.1103,
+      "step": 3574
+    },
+    {
+      "epoch": 0.27559358618563057,
+      "grad_norm": 3.4711949825286865,
+      "learning_rate": 8.500492487925107e-06,
+      "loss": 0.9718,
+      "step": 3575
+    },
+    {
+      "epoch": 0.27567067530064754,
+      "grad_norm": 3.5179035663604736,
+      "learning_rate": 8.499600929699501e-06,
+      "loss": 1.0066,
+      "step": 3576
+    },
+    {
+      "epoch": 0.2757477644156645,
+      "grad_norm": 3.881817579269409,
+      "learning_rate": 8.498709153291901e-06,
+      "loss": 1.0456,
+      "step": 3577
+    },
+    {
+      "epoch": 0.2758248535306815,
+      "grad_norm": 4.031117916107178,
+      "learning_rate": 8.497817158757906e-06,
+      "loss": 1.0358,
+      "step": 3578
+    },
+    {
+      "epoch": 0.27590194264569845,
+      "grad_norm": 4.566905498504639,
+      "learning_rate": 8.49692494615313e-06,
+      "loss": 1.0547,
+      "step": 3579
+    },
+    {
+      "epoch": 0.27597903176071537,
+      "grad_norm": 3.682335138320923,
+      "learning_rate": 8.49603251553319e-06,
+      "loss": 1.0357,
+      "step": 3580
+    },
+    {
+      "epoch": 0.27605612087573234,
+      "grad_norm": 3.5989487171173096,
+      "learning_rate": 8.495139866953732e-06,
+      "loss": 0.8995,
+      "step": 3581
+    },
+    {
+      "epoch": 0.2761332099907493,
+      "grad_norm": 3.775329351425171,
+      "learning_rate": 8.494247000470404e-06,
+      "loss": 1.0149,
+      "step": 3582
+    },
+    {
+      "epoch": 0.2762102991057663,
+      "grad_norm": 3.7653584480285645,
+      "learning_rate": 8.493353916138873e-06,
+      "loss": 1.0037,
+      "step": 3583
+    },
+    {
+      "epoch": 0.27628738822078325,
+      "grad_norm": 4.354711532592773,
+      "learning_rate": 8.492460614014816e-06,
+      "loss": 1.0801,
+      "step": 3584
+    },
+    {
+      "epoch": 0.27636447733580016,
+      "grad_norm": 3.7385101318359375,
+      "learning_rate": 8.49156709415393e-06,
+      "loss": 0.9805,
+      "step": 3585
+    },
+    {
+      "epoch": 0.27644156645081713,
+      "grad_norm": 3.78240704536438,
+      "learning_rate": 8.490673356611919e-06,
+      "loss": 1.0614,
+      "step": 3586
+    },
+    {
+      "epoch": 0.2765186555658341,
+      "grad_norm": 4.034103870391846,
+      "learning_rate": 8.489779401444503e-06,
+      "loss": 1.0574,
+      "step": 3587
+    },
+    {
+      "epoch": 0.2765957446808511,
+      "grad_norm": 3.6883256435394287,
+      "learning_rate": 8.488885228707413e-06,
+      "loss": 1.165,
+      "step": 3588
+    },
+    {
+      "epoch": 0.27667283379586805,
+      "grad_norm": 3.4828081130981445,
+      "learning_rate": 8.4879908384564e-06,
+      "loss": 0.924,
+      "step": 3589
+    },
+    {
+      "epoch": 0.27674992291088496,
+      "grad_norm": 3.475923776626587,
+      "learning_rate": 8.487096230747223e-06,
+      "loss": 0.9768,
+      "step": 3590
+    },
+    {
+      "epoch": 0.27682701202590193,
+      "grad_norm": 3.595651626586914,
+      "learning_rate": 8.486201405635655e-06,
+      "loss": 0.8966,
+      "step": 3591
+    },
+    {
+      "epoch": 0.2769041011409189,
+      "grad_norm": 4.367059707641602,
+      "learning_rate": 8.485306363177485e-06,
+      "loss": 0.9618,
+      "step": 3592
+    },
+    {
+      "epoch": 0.27698119025593587,
+      "grad_norm": 3.2522506713867188,
+      "learning_rate": 8.484411103428516e-06,
+      "loss": 0.9475,
+      "step": 3593
+    },
+    {
+      "epoch": 0.27705827937095284,
+      "grad_norm": 3.634666919708252,
+      "learning_rate": 8.48351562644456e-06,
+      "loss": 1.0517,
+      "step": 3594
+    },
+    {
+      "epoch": 0.27713536848596976,
+      "grad_norm": 3.7717621326446533,
+      "learning_rate": 8.482619932281446e-06,
+      "loss": 1.017,
+      "step": 3595
+    },
+    {
+      "epoch": 0.27721245760098673,
+      "grad_norm": 3.5509636402130127,
+      "learning_rate": 8.481724020995017e-06,
+      "loss": 0.9433,
+      "step": 3596
+    },
+    {
+      "epoch": 0.2772895467160037,
+      "grad_norm": 3.6704728603363037,
+      "learning_rate": 8.480827892641125e-06,
+      "loss": 0.991,
+      "step": 3597
+    },
+    {
+      "epoch": 0.27736663583102067,
+      "grad_norm": 3.710653305053711,
+      "learning_rate": 8.479931547275644e-06,
+      "loss": 1.1041,
+      "step": 3598
+    },
+    {
+      "epoch": 0.27744372494603764,
+      "grad_norm": 3.6124982833862305,
+      "learning_rate": 8.479034984954454e-06,
+      "loss": 0.9931,
+      "step": 3599
+    },
+    {
+      "epoch": 0.27752081406105455,
+      "grad_norm": 3.4628562927246094,
+      "learning_rate": 8.47813820573345e-06,
+      "loss": 1.1006,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2775979031760715,
+      "grad_norm": 3.8903744220733643,
+      "learning_rate": 8.477241209668543e-06,
+      "loss": 1.072,
+      "step": 3601
+    },
+    {
+      "epoch": 0.2776749922910885,
+      "grad_norm": 3.902669906616211,
+      "learning_rate": 8.476343996815657e-06,
+      "loss": 0.9923,
+      "step": 3602
+    },
+    {
+      "epoch": 0.27775208140610547,
+      "grad_norm": 3.763310670852661,
+      "learning_rate": 8.475446567230727e-06,
+      "loss": 0.9965,
+      "step": 3603
+    },
+    {
+      "epoch": 0.27782917052112244,
+      "grad_norm": 3.7086451053619385,
+      "learning_rate": 8.4745489209697e-06,
+      "loss": 1.0033,
+      "step": 3604
+    },
+    {
+      "epoch": 0.27790625963613935,
+      "grad_norm": 3.9299726486206055,
+      "learning_rate": 8.473651058088548e-06,
+      "loss": 1.1179,
+      "step": 3605
+    },
+    {
+      "epoch": 0.2779833487511563,
+      "grad_norm": 3.552938222885132,
+      "learning_rate": 8.47275297864324e-06,
+      "loss": 1.0242,
+      "step": 3606
+    },
+    {
+      "epoch": 0.2780604378661733,
+      "grad_norm": 3.6222376823425293,
+      "learning_rate": 8.47185468268977e-06,
+      "loss": 1.0164,
+      "step": 3607
+    },
+    {
+      "epoch": 0.27813752698119026,
+      "grad_norm": 4.27018928527832,
+      "learning_rate": 8.470956170284141e-06,
+      "loss": 1.0589,
+      "step": 3608
+    },
+    {
+      "epoch": 0.27821461609620723,
+      "grad_norm": 3.718721628189087,
+      "learning_rate": 8.470057441482374e-06,
+      "loss": 0.9983,
+      "step": 3609
+    },
+    {
+      "epoch": 0.27829170521122415,
+      "grad_norm": 3.726409673690796,
+      "learning_rate": 8.469158496340496e-06,
+      "loss": 1.028,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2783687943262411,
+      "grad_norm": 3.4664266109466553,
+      "learning_rate": 8.46825933491455e-06,
+      "loss": 0.9796,
+      "step": 3611
+    },
+    {
+      "epoch": 0.2784458834412581,
+      "grad_norm": 3.7428462505340576,
+      "learning_rate": 8.4673599572606e-06,
+      "loss": 1.0337,
+      "step": 3612
+    },
+    {
+      "epoch": 0.27852297255627506,
+      "grad_norm": 3.687169313430786,
+      "learning_rate": 8.466460363434714e-06,
+      "loss": 1.0132,
+      "step": 3613
+    },
+    {
+      "epoch": 0.27860006167129203,
+      "grad_norm": 4.0733113288879395,
+      "learning_rate": 8.465560553492978e-06,
+      "loss": 0.9581,
+      "step": 3614
+    },
+    {
+      "epoch": 0.27867715078630895,
+      "grad_norm": 3.739532709121704,
+      "learning_rate": 8.46466052749149e-06,
+      "loss": 0.9367,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2787542399013259,
+      "grad_norm": 3.9746146202087402,
+      "learning_rate": 8.463760285486362e-06,
+      "loss": 0.9157,
+      "step": 3616
+    },
+    {
+      "epoch": 0.2788313290163429,
+      "grad_norm": 3.625314235687256,
+      "learning_rate": 8.462859827533718e-06,
+      "loss": 0.9241,
+      "step": 3617
+    },
+    {
+      "epoch": 0.27890841813135986,
+      "grad_norm": 3.546565532684326,
+      "learning_rate": 8.4619591536897e-06,
+      "loss": 0.9287,
+      "step": 3618
+    },
+    {
+      "epoch": 0.27898550724637683,
+      "grad_norm": 3.9717471599578857,
+      "learning_rate": 8.46105826401046e-06,
+      "loss": 1.0355,
+      "step": 3619
+    },
+    {
+      "epoch": 0.27906259636139374,
+      "grad_norm": 3.866204261779785,
+      "learning_rate": 8.46015715855216e-06,
+      "loss": 1.0423,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2791396854764107,
+      "grad_norm": 3.634885549545288,
+      "learning_rate": 8.459255837370984e-06,
+      "loss": 1.0229,
+      "step": 3621
+    },
+    {
+      "epoch": 0.2792167745914277,
+      "grad_norm": 3.85805344581604,
+      "learning_rate": 8.45835430052312e-06,
+      "loss": 0.9998,
+      "step": 3622
+    },
+    {
+      "epoch": 0.27929386370644466,
+      "grad_norm": 3.9002537727355957,
+      "learning_rate": 8.457452548064778e-06,
+      "loss": 0.9979,
+      "step": 3623
+    },
+    {
+      "epoch": 0.2793709528214616,
+      "grad_norm": 3.3640360832214355,
+      "learning_rate": 8.456550580052177e-06,
+      "loss": 0.8977,
+      "step": 3624
+    },
+    {
+      "epoch": 0.27944804193647854,
+      "grad_norm": 3.474541664123535,
+      "learning_rate": 8.455648396541548e-06,
+      "loss": 1.0045,
+      "step": 3625
+    },
+    {
+      "epoch": 0.2795251310514955,
+      "grad_norm": 3.5307717323303223,
+      "learning_rate": 8.454745997589139e-06,
+      "loss": 0.9458,
+      "step": 3626
+    },
+    {
+      "epoch": 0.2796022201665125,
+      "grad_norm": 3.8024139404296875,
+      "learning_rate": 8.45384338325121e-06,
+      "loss": 1.0375,
+      "step": 3627
+    },
+    {
+      "epoch": 0.27967930928152945,
+      "grad_norm": 3.663609027862549,
+      "learning_rate": 8.452940553584032e-06,
+      "loss": 0.8606,
+      "step": 3628
+    },
+    {
+      "epoch": 0.2797563983965464,
+      "grad_norm": 3.6172802448272705,
+      "learning_rate": 8.452037508643897e-06,
+      "loss": 0.9782,
+      "step": 3629
+    },
+    {
+      "epoch": 0.2798334875115634,
+      "grad_norm": 3.8547043800354004,
+      "learning_rate": 8.4511342484871e-06,
+      "loss": 0.9364,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2799105766265803,
+      "grad_norm": 3.862354040145874,
+      "learning_rate": 8.450230773169958e-06,
+      "loss": 1.0593,
+      "step": 3631
+    },
+    {
+      "epoch": 0.2799876657415973,
+      "grad_norm": 3.5931904315948486,
+      "learning_rate": 8.449327082748794e-06,
+      "loss": 1.0177,
+      "step": 3632
+    },
+    {
+      "epoch": 0.28006475485661425,
+      "grad_norm": 3.79412579536438,
+      "learning_rate": 8.448423177279954e-06,
+      "loss": 1.0407,
+      "step": 3633
+    },
+    {
+      "epoch": 0.2801418439716312,
+      "grad_norm": 3.9364240169525146,
+      "learning_rate": 8.447519056819787e-06,
+      "loss": 1.0459,
+      "step": 3634
+    },
+    {
+      "epoch": 0.2802189330866482,
+      "grad_norm": 3.5306715965270996,
+      "learning_rate": 8.446614721424661e-06,
+      "loss": 1.0045,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2802960222016651,
+      "grad_norm": 4.1420111656188965,
+      "learning_rate": 8.445710171150957e-06,
+      "loss": 1.1036,
+      "step": 3636
+    },
+    {
+      "epoch": 0.2803731113166821,
+      "grad_norm": 3.604893684387207,
+      "learning_rate": 8.444805406055072e-06,
+      "loss": 0.9265,
+      "step": 3637
+    },
+    {
+      "epoch": 0.28045020043169905,
+      "grad_norm": 3.910055637359619,
+      "learning_rate": 8.44390042619341e-06,
+      "loss": 1.0674,
+      "step": 3638
+    },
+    {
+      "epoch": 0.280527289546716,
+      "grad_norm": 4.729837894439697,
+      "learning_rate": 8.442995231622393e-06,
+      "loss": 1.0476,
+      "step": 3639
+    },
+    {
+      "epoch": 0.280604378661733,
+      "grad_norm": 3.926619529724121,
+      "learning_rate": 8.442089822398456e-06,
+      "loss": 0.9571,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2806814677767499,
+      "grad_norm": 3.6743786334991455,
+      "learning_rate": 8.441184198578044e-06,
+      "loss": 1.0757,
+      "step": 3641
+    },
+    {
+      "epoch": 0.2807585568917669,
+      "grad_norm": 3.5217857360839844,
+      "learning_rate": 8.44027836021762e-06,
+      "loss": 1.0539,
+      "step": 3642
+    },
+    {
+      "epoch": 0.28083564600678385,
+      "grad_norm": 3.6347432136535645,
+      "learning_rate": 8.439372307373658e-06,
+      "loss": 0.9301,
+      "step": 3643
+    },
+    {
+      "epoch": 0.2809127351218008,
+      "grad_norm": 4.677338123321533,
+      "learning_rate": 8.438466040102647e-06,
+      "loss": 0.9828,
+      "step": 3644
+    },
+    {
+      "epoch": 0.2809898242368178,
+      "grad_norm": 3.5795114040374756,
+      "learning_rate": 8.437559558461085e-06,
+      "loss": 0.8683,
+      "step": 3645
+    },
+    {
+      "epoch": 0.2810669133518347,
+      "grad_norm": 3.54679799079895,
+      "learning_rate": 8.436652862505488e-06,
+      "loss": 1.0562,
+      "step": 3646
+    },
+    {
+      "epoch": 0.2811440024668517,
+      "grad_norm": 4.047598838806152,
+      "learning_rate": 8.435745952292384e-06,
+      "loss": 0.981,
+      "step": 3647
+    },
+    {
+      "epoch": 0.28122109158186864,
+      "grad_norm": 3.662014961242676,
+      "learning_rate": 8.434838827878315e-06,
+      "loss": 0.9972,
+      "step": 3648
+    },
+    {
+      "epoch": 0.2812981806968856,
+      "grad_norm": 3.410169839859009,
+      "learning_rate": 8.433931489319835e-06,
+      "loss": 0.9462,
+      "step": 3649
+    },
+    {
+      "epoch": 0.2813752698119026,
+      "grad_norm": 3.725191354751587,
+      "learning_rate": 8.43302393667351e-06,
+      "loss": 0.9657,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2814523589269195,
+      "grad_norm": 3.510698080062866,
+      "learning_rate": 8.432116169995923e-06,
+      "loss": 1.0496,
+      "step": 3651
+    },
+    {
+      "epoch": 0.28152944804193647,
+      "grad_norm": 3.6888866424560547,
+      "learning_rate": 8.43120818934367e-06,
+      "loss": 0.9789,
+      "step": 3652
+    },
+    {
+      "epoch": 0.28160653715695344,
+      "grad_norm": 3.4661200046539307,
+      "learning_rate": 8.430299994773354e-06,
+      "loss": 0.9091,
+      "step": 3653
+    },
+    {
+      "epoch": 0.2816836262719704,
+      "grad_norm": 3.9719278812408447,
+      "learning_rate": 8.429391586341602e-06,
+      "loss": 1.0466,
+      "step": 3654
+    },
+    {
+      "epoch": 0.2817607153869874,
+      "grad_norm": 3.836193323135376,
+      "learning_rate": 8.428482964105043e-06,
+      "loss": 1.0741,
+      "step": 3655
+    },
+    {
+      "epoch": 0.2818378045020043,
+      "grad_norm": 3.6839404106140137,
+      "learning_rate": 8.427574128120331e-06,
+      "loss": 0.9394,
+      "step": 3656
+    },
+    {
+      "epoch": 0.28191489361702127,
+      "grad_norm": 3.4464738368988037,
+      "learning_rate": 8.42666507844412e-06,
+      "loss": 1.0155,
+      "step": 3657
+    },
+    {
+      "epoch": 0.28199198273203824,
+      "grad_norm": 4.588211536407471,
+      "learning_rate": 8.425755815133092e-06,
+      "loss": 1.1744,
+      "step": 3658
+    },
+    {
+      "epoch": 0.2820690718470552,
+      "grad_norm": 4.032186508178711,
+      "learning_rate": 8.42484633824393e-06,
+      "loss": 1.0072,
+      "step": 3659
+    },
+    {
+      "epoch": 0.2821461609620722,
+      "grad_norm": 3.7464022636413574,
+      "learning_rate": 8.423936647833337e-06,
+      "loss": 0.9378,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2822232500770891,
+      "grad_norm": 3.7981278896331787,
+      "learning_rate": 8.423026743958028e-06,
+      "loss": 1.0546,
+      "step": 3661
+    },
+    {
+      "epoch": 0.28230033919210606,
+      "grad_norm": 4.554448127746582,
+      "learning_rate": 8.422116626674728e-06,
+      "loss": 1.037,
+      "step": 3662
+    },
+    {
+      "epoch": 0.28237742830712304,
+      "grad_norm": 3.7701575756073,
+      "learning_rate": 8.42120629604018e-06,
+      "loss": 1.0272,
+      "step": 3663
+    },
+    {
+      "epoch": 0.28245451742214,
+      "grad_norm": 3.8832554817199707,
+      "learning_rate": 8.420295752111138e-06,
+      "loss": 1.0422,
+      "step": 3664
+    },
+    {
+      "epoch": 0.282531606537157,
+      "grad_norm": 4.044317245483398,
+      "learning_rate": 8.41938499494437e-06,
+      "loss": 1.0457,
+      "step": 3665
+    },
+    {
+      "epoch": 0.2826086956521739,
+      "grad_norm": 3.7039005756378174,
+      "learning_rate": 8.418474024596659e-06,
+      "loss": 0.9673,
+      "step": 3666
+    },
+    {
+      "epoch": 0.28268578476719086,
+      "grad_norm": 3.4527359008789062,
+      "learning_rate": 8.417562841124797e-06,
+      "loss": 0.9998,
+      "step": 3667
+    },
+    {
+      "epoch": 0.28276287388220783,
+      "grad_norm": 4.247575759887695,
+      "learning_rate": 8.416651444585591e-06,
+      "loss": 1.0611,
+      "step": 3668
+    },
+    {
+      "epoch": 0.2828399629972248,
+      "grad_norm": 3.4777634143829346,
+      "learning_rate": 8.415739835035864e-06,
+      "loss": 0.9773,
+      "step": 3669
+    },
+    {
+      "epoch": 0.2829170521122418,
+      "grad_norm": 4.208752155303955,
+      "learning_rate": 8.414828012532446e-06,
+      "loss": 0.9631,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2829941412272587,
+      "grad_norm": 3.4110782146453857,
+      "learning_rate": 8.41391597713219e-06,
+      "loss": 1.0187,
+      "step": 3671
+    },
+    {
+      "epoch": 0.28307123034227566,
+      "grad_norm": 3.7919259071350098,
+      "learning_rate": 8.413003728891953e-06,
+      "loss": 1.1607,
+      "step": 3672
+    },
+    {
+      "epoch": 0.28314831945729263,
+      "grad_norm": 3.560163736343384,
+      "learning_rate": 8.41209126786861e-06,
+      "loss": 0.9773,
+      "step": 3673
+    },
+    {
+      "epoch": 0.2832254085723096,
+      "grad_norm": 3.5800511837005615,
+      "learning_rate": 8.411178594119046e-06,
+      "loss": 0.8893,
+      "step": 3674
+    },
+    {
+      "epoch": 0.28330249768732657,
+      "grad_norm": 4.5871901512146,
+      "learning_rate": 8.410265707700167e-06,
+      "loss": 1.1218,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2833795868023435,
+      "grad_norm": 3.5251963138580322,
+      "learning_rate": 8.409352608668882e-06,
+      "loss": 1.1023,
+      "step": 3676
+    },
+    {
+      "epoch": 0.28345667591736046,
+      "grad_norm": 3.715285539627075,
+      "learning_rate": 8.408439297082118e-06,
+      "loss": 0.9962,
+      "step": 3677
+    },
+    {
+      "epoch": 0.2835337650323774,
+      "grad_norm": 3.63982892036438,
+      "learning_rate": 8.407525772996818e-06,
+      "loss": 1.043,
+      "step": 3678
+    },
+    {
+      "epoch": 0.2836108541473944,
+      "grad_norm": 4.022475242614746,
+      "learning_rate": 8.406612036469935e-06,
+      "loss": 1.2085,
+      "step": 3679
+    },
+    {
+      "epoch": 0.28368794326241137,
+      "grad_norm": 3.487618923187256,
+      "learning_rate": 8.405698087558432e-06,
+      "loss": 1.0158,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2837650323774283,
+      "grad_norm": 3.937352180480957,
+      "learning_rate": 8.404783926319295e-06,
+      "loss": 1.0598,
+      "step": 3681
+    },
+    {
+      "epoch": 0.28384212149244525,
+      "grad_norm": 3.6437175273895264,
+      "learning_rate": 8.403869552809512e-06,
+      "loss": 0.9678,
+      "step": 3682
+    },
+    {
+      "epoch": 0.2839192106074622,
+      "grad_norm": 4.383546829223633,
+      "learning_rate": 8.402954967086093e-06,
+      "loss": 1.051,
+      "step": 3683
+    },
+    {
+      "epoch": 0.2839962997224792,
+      "grad_norm": 4.169625282287598,
+      "learning_rate": 8.402040169206054e-06,
+      "loss": 1.0238,
+      "step": 3684
+    },
+    {
+      "epoch": 0.28407338883749617,
+      "grad_norm": 3.8336832523345947,
+      "learning_rate": 8.40112515922643e-06,
+      "loss": 0.9547,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2841504779525131,
+      "grad_norm": 3.8428425788879395,
+      "learning_rate": 8.40020993720427e-06,
+      "loss": 0.9351,
+      "step": 3686
+    },
+    {
+      "epoch": 0.28422756706753005,
+      "grad_norm": 3.674729108810425,
+      "learning_rate": 8.399294503196629e-06,
+      "loss": 1.0025,
+      "step": 3687
+    },
+    {
+      "epoch": 0.284304656182547,
+      "grad_norm": 3.847797155380249,
+      "learning_rate": 8.398378857260581e-06,
+      "loss": 1.0309,
+      "step": 3688
+    },
+    {
+      "epoch": 0.284381745297564,
+      "grad_norm": 3.6674485206604004,
+      "learning_rate": 8.397462999453212e-06,
+      "loss": 0.9968,
+      "step": 3689
+    },
+    {
+      "epoch": 0.28445883441258096,
+      "grad_norm": 4.155699253082275,
+      "learning_rate": 8.39654692983162e-06,
+      "loss": 1.1423,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2845359235275979,
+      "grad_norm": 3.99698543548584,
+      "learning_rate": 8.395630648452919e-06,
+      "loss": 1.0867,
+      "step": 3691
+    },
+    {
+      "epoch": 0.28461301264261485,
+      "grad_norm": 3.8566694259643555,
+      "learning_rate": 8.394714155374234e-06,
+      "loss": 0.9227,
+      "step": 3692
+    },
+    {
+      "epoch": 0.2846901017576318,
+      "grad_norm": 3.5863726139068604,
+      "learning_rate": 8.393797450652701e-06,
+      "loss": 0.9605,
+      "step": 3693
+    },
+    {
+      "epoch": 0.2847671908726488,
+      "grad_norm": 3.88511323928833,
+      "learning_rate": 8.392880534345477e-06,
+      "loss": 0.9894,
+      "step": 3694
+    },
+    {
+      "epoch": 0.28484427998766576,
+      "grad_norm": 3.3645994663238525,
+      "learning_rate": 8.391963406509721e-06,
+      "loss": 1.0001,
+      "step": 3695
+    },
+    {
+      "epoch": 0.2849213691026827,
+      "grad_norm": 3.5766923427581787,
+      "learning_rate": 8.391046067202617e-06,
+      "loss": 0.9889,
+      "step": 3696
+    },
+    {
+      "epoch": 0.28499845821769965,
+      "grad_norm": 4.069863796234131,
+      "learning_rate": 8.390128516481351e-06,
+      "loss": 0.9495,
+      "step": 3697
+    },
+    {
+      "epoch": 0.2850755473327166,
+      "grad_norm": 3.9088964462280273,
+      "learning_rate": 8.389210754403132e-06,
+      "loss": 1.0693,
+      "step": 3698
+    },
+    {
+      "epoch": 0.2851526364477336,
+      "grad_norm": 4.604135990142822,
+      "learning_rate": 8.388292781025173e-06,
+      "loss": 1.1569,
+      "step": 3699
+    },
+    {
+      "epoch": 0.28522972556275056,
+      "grad_norm": 3.65946364402771,
+      "learning_rate": 8.38737459640471e-06,
+      "loss": 0.9248,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2853068146777675,
+      "grad_norm": 3.6877028942108154,
+      "learning_rate": 8.386456200598982e-06,
+      "loss": 1.1605,
+      "step": 3701
+    },
+    {
+      "epoch": 0.28538390379278444,
+      "grad_norm": 3.4884543418884277,
+      "learning_rate": 8.38553759366525e-06,
+      "loss": 1.0238,
+      "step": 3702
+    },
+    {
+      "epoch": 0.2854609929078014,
+      "grad_norm": 3.4332854747772217,
+      "learning_rate": 8.384618775660784e-06,
+      "loss": 1.0155,
+      "step": 3703
+    },
+    {
+      "epoch": 0.2855380820228184,
+      "grad_norm": 4.001807689666748,
+      "learning_rate": 8.383699746642866e-06,
+      "loss": 1.1088,
+      "step": 3704
+    },
+    {
+      "epoch": 0.28561517113783536,
+      "grad_norm": 3.441525459289551,
+      "learning_rate": 8.382780506668792e-06,
+      "loss": 0.9096,
+      "step": 3705
+    },
+    {
+      "epoch": 0.28569226025285227,
+      "grad_norm": 3.314866542816162,
+      "learning_rate": 8.381861055795876e-06,
+      "loss": 1.0886,
+      "step": 3706
+    },
+    {
+      "epoch": 0.28576934936786924,
+      "grad_norm": 4.307245254516602,
+      "learning_rate": 8.380941394081437e-06,
+      "loss": 1.0525,
+      "step": 3707
+    },
+    {
+      "epoch": 0.2858464384828862,
+      "grad_norm": 3.5689730644226074,
+      "learning_rate": 8.380021521582813e-06,
+      "loss": 1.0542,
+      "step": 3708
+    },
+    {
+      "epoch": 0.2859235275979032,
+      "grad_norm": 3.881000518798828,
+      "learning_rate": 8.379101438357352e-06,
+      "loss": 0.9946,
+      "step": 3709
+    },
+    {
+      "epoch": 0.28600061671292015,
+      "grad_norm": 4.110035419464111,
+      "learning_rate": 8.378181144462418e-06,
+      "loss": 1.0302,
+      "step": 3710
+    },
+    {
+      "epoch": 0.28607770582793707,
+      "grad_norm": 3.9843645095825195,
+      "learning_rate": 8.377260639955385e-06,
+      "loss": 1.026,
+      "step": 3711
+    },
+    {
+      "epoch": 0.28615479494295404,
+      "grad_norm": 3.5178589820861816,
+      "learning_rate": 8.376339924893642e-06,
+      "loss": 0.9419,
+      "step": 3712
+    },
+    {
+      "epoch": 0.286231884057971,
+      "grad_norm": 3.800537109375,
+      "learning_rate": 8.375418999334591e-06,
+      "loss": 0.9474,
+      "step": 3713
+    },
+    {
+      "epoch": 0.286308973172988,
+      "grad_norm": 3.798588991165161,
+      "learning_rate": 8.374497863335649e-06,
+      "loss": 1.0826,
+      "step": 3714
+    },
+    {
+      "epoch": 0.28638606228800495,
+      "grad_norm": 3.9119489192962646,
+      "learning_rate": 8.37357651695424e-06,
+      "loss": 1.0957,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2864631514030219,
+      "grad_norm": 3.5828166007995605,
+      "learning_rate": 8.37265496024781e-06,
+      "loss": 0.951,
+      "step": 3716
+    },
+    {
+      "epoch": 0.28654024051803884,
+      "grad_norm": 3.7555758953094482,
+      "learning_rate": 8.371733193273808e-06,
+      "loss": 0.9873,
+      "step": 3717
+    },
+    {
+      "epoch": 0.2866173296330558,
+      "grad_norm": 3.638547897338867,
+      "learning_rate": 8.370811216089705e-06,
+      "loss": 1.1217,
+      "step": 3718
+    },
+    {
+      "epoch": 0.2866944187480728,
+      "grad_norm": 3.6661555767059326,
+      "learning_rate": 8.36988902875298e-06,
+      "loss": 1.0319,
+      "step": 3719
+    },
+    {
+      "epoch": 0.28677150786308975,
+      "grad_norm": 3.355048179626465,
+      "learning_rate": 8.36896663132113e-06,
+      "loss": 0.9512,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2868485969781067,
+      "grad_norm": 3.8450980186462402,
+      "learning_rate": 8.368044023851656e-06,
+      "loss": 1.0523,
+      "step": 3721
+    },
+    {
+      "epoch": 0.28692568609312363,
+      "grad_norm": 3.896583318710327,
+      "learning_rate": 8.36712120640208e-06,
+      "loss": 1.0813,
+      "step": 3722
+    },
+    {
+      "epoch": 0.2870027752081406,
+      "grad_norm": 3.890866279602051,
+      "learning_rate": 8.366198179029937e-06,
+      "loss": 1.0645,
+      "step": 3723
+    },
+    {
+      "epoch": 0.2870798643231576,
+      "grad_norm": 3.7487337589263916,
+      "learning_rate": 8.365274941792771e-06,
+      "loss": 0.9914,
+      "step": 3724
+    },
+    {
+      "epoch": 0.28715695343817454,
+      "grad_norm": 3.695530652999878,
+      "learning_rate": 8.364351494748141e-06,
+      "loss": 1.011,
+      "step": 3725
+    },
+    {
+      "epoch": 0.2872340425531915,
+      "grad_norm": 3.758422613143921,
+      "learning_rate": 8.363427837953622e-06,
+      "loss": 1.0366,
+      "step": 3726
+    },
+    {
+      "epoch": 0.28731113166820843,
+      "grad_norm": 4.224013328552246,
+      "learning_rate": 8.362503971466795e-06,
+      "loss": 1.0637,
+      "step": 3727
+    },
+    {
+      "epoch": 0.2873882207832254,
+      "grad_norm": 4.0081305503845215,
+      "learning_rate": 8.361579895345263e-06,
+      "loss": 1.133,
+      "step": 3728
+    },
+    {
+      "epoch": 0.28746530989824237,
+      "grad_norm": 3.7337818145751953,
+      "learning_rate": 8.36065560964663e-06,
+      "loss": 0.9504,
+      "step": 3729
+    },
+    {
+      "epoch": 0.28754239901325934,
+      "grad_norm": 3.62839412689209,
+      "learning_rate": 8.359731114428529e-06,
+      "loss": 1.093,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2876194881282763,
+      "grad_norm": 3.5839309692382812,
+      "learning_rate": 8.358806409748592e-06,
+      "loss": 1.0897,
+      "step": 3731
+    },
+    {
+      "epoch": 0.28769657724329323,
+      "grad_norm": 3.644613265991211,
+      "learning_rate": 8.35788149566447e-06,
+      "loss": 1.0344,
+      "step": 3732
+    },
+    {
+      "epoch": 0.2877736663583102,
+      "grad_norm": 3.906480073928833,
+      "learning_rate": 8.356956372233829e-06,
+      "loss": 1.092,
+      "step": 3733
+    },
+    {
+      "epoch": 0.28785075547332717,
+      "grad_norm": 3.4734041690826416,
+      "learning_rate": 8.356031039514344e-06,
+      "loss": 1.0349,
+      "step": 3734
+    },
+    {
+      "epoch": 0.28792784458834414,
+      "grad_norm": 3.4001238346099854,
+      "learning_rate": 8.355105497563705e-06,
+      "loss": 0.9855,
+      "step": 3735
+    },
+    {
+      "epoch": 0.2880049337033611,
+      "grad_norm": 3.7522032260894775,
+      "learning_rate": 8.354179746439614e-06,
+      "loss": 1.0141,
+      "step": 3736
+    },
+    {
+      "epoch": 0.288082022818378,
+      "grad_norm": 3.7555811405181885,
+      "learning_rate": 8.353253786199788e-06,
+      "loss": 1.0243,
+      "step": 3737
+    },
+    {
+      "epoch": 0.288159111933395,
+      "grad_norm": 3.771195411682129,
+      "learning_rate": 8.352327616901956e-06,
+      "loss": 0.9578,
+      "step": 3738
+    },
+    {
+      "epoch": 0.28823620104841197,
+      "grad_norm": 3.630093574523926,
+      "learning_rate": 8.35140123860386e-06,
+      "loss": 1.0626,
+      "step": 3739
+    },
+    {
+      "epoch": 0.28831329016342894,
+      "grad_norm": 3.6648457050323486,
+      "learning_rate": 8.350474651363254e-06,
+      "loss": 1.1257,
+      "step": 3740
+    },
+    {
+      "epoch": 0.2883903792784459,
+      "grad_norm": 3.7269270420074463,
+      "learning_rate": 8.349547855237904e-06,
+      "loss": 1.0953,
+      "step": 3741
+    },
+    {
+      "epoch": 0.2884674683934628,
+      "grad_norm": 3.8183059692382812,
+      "learning_rate": 8.348620850285594e-06,
+      "loss": 1.1094,
+      "step": 3742
+    },
+    {
+      "epoch": 0.2885445575084798,
+      "grad_norm": 3.764166831970215,
+      "learning_rate": 8.347693636564119e-06,
+      "loss": 0.9972,
+      "step": 3743
+    },
+    {
+      "epoch": 0.28862164662349676,
+      "grad_norm": 3.2361299991607666,
+      "learning_rate": 8.34676621413128e-06,
+      "loss": 0.8723,
+      "step": 3744
+    },
+    {
+      "epoch": 0.28869873573851373,
+      "grad_norm": 4.178921699523926,
+      "learning_rate": 8.345838583044903e-06,
+      "loss": 0.8852,
+      "step": 3745
+    },
+    {
+      "epoch": 0.2887758248535307,
+      "grad_norm": 4.049960136413574,
+      "learning_rate": 8.344910743362819e-06,
+      "loss": 1.1337,
+      "step": 3746
+    },
+    {
+      "epoch": 0.2888529139685476,
+      "grad_norm": 4.307690143585205,
+      "learning_rate": 8.343982695142873e-06,
+      "loss": 1.0333,
+      "step": 3747
+    },
+    {
+      "epoch": 0.2889300030835646,
+      "grad_norm": 3.709958791732788,
+      "learning_rate": 8.343054438442926e-06,
+      "loss": 1.0488,
+      "step": 3748
+    },
+    {
+      "epoch": 0.28900709219858156,
+      "grad_norm": 3.61205792427063,
+      "learning_rate": 8.342125973320848e-06,
+      "loss": 0.9665,
+      "step": 3749
+    },
+    {
+      "epoch": 0.28908418131359853,
+      "grad_norm": 3.7218470573425293,
+      "learning_rate": 8.341197299834524e-06,
+      "loss": 1.0409,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2891612704286155,
+      "grad_norm": 3.4968018531799316,
+      "learning_rate": 8.340268418041855e-06,
+      "loss": 0.9222,
+      "step": 3751
+    },
+    {
+      "epoch": 0.2892383595436324,
+      "grad_norm": 3.451004981994629,
+      "learning_rate": 8.339339328000749e-06,
+      "loss": 1.0035,
+      "step": 3752
+    },
+    {
+      "epoch": 0.2893154486586494,
+      "grad_norm": 3.4331042766571045,
+      "learning_rate": 8.338410029769133e-06,
+      "loss": 0.9606,
+      "step": 3753
+    },
+    {
+      "epoch": 0.28939253777366636,
+      "grad_norm": 3.4850504398345947,
+      "learning_rate": 8.337480523404938e-06,
+      "loss": 0.9549,
+      "step": 3754
+    },
+    {
+      "epoch": 0.28946962688868333,
+      "grad_norm": 3.5593273639678955,
+      "learning_rate": 8.33655080896612e-06,
+      "loss": 1.0124,
+      "step": 3755
+    },
+    {
+      "epoch": 0.2895467160037003,
+      "grad_norm": 3.5251691341400146,
+      "learning_rate": 8.335620886510637e-06,
+      "loss": 0.9167,
+      "step": 3756
+    },
+    {
+      "epoch": 0.2896238051187172,
+      "grad_norm": 3.627453088760376,
+      "learning_rate": 8.33469075609647e-06,
+      "loss": 0.9891,
+      "step": 3757
+    },
+    {
+      "epoch": 0.2897008942337342,
+      "grad_norm": 3.599273204803467,
+      "learning_rate": 8.333760417781605e-06,
+      "loss": 1.0972,
+      "step": 3758
+    },
+    {
+      "epoch": 0.28977798334875116,
+      "grad_norm": 3.6854488849639893,
+      "learning_rate": 8.332829871624042e-06,
+      "loss": 1.1368,
+      "step": 3759
+    },
+    {
+      "epoch": 0.2898550724637681,
+      "grad_norm": 3.5526113510131836,
+      "learning_rate": 8.331899117681799e-06,
+      "loss": 0.9347,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2899321615787851,
+      "grad_norm": 3.6977930068969727,
+      "learning_rate": 8.330968156012902e-06,
+      "loss": 0.9981,
+      "step": 3761
+    },
+    {
+      "epoch": 0.290009250693802,
+      "grad_norm": 3.659409761428833,
+      "learning_rate": 8.330036986675392e-06,
+      "loss": 1.023,
+      "step": 3762
+    },
+    {
+      "epoch": 0.290086339808819,
+      "grad_norm": 4.15533971786499,
+      "learning_rate": 8.329105609727323e-06,
+      "loss": 1.018,
+      "step": 3763
+    },
+    {
+      "epoch": 0.29016342892383595,
+      "grad_norm": 3.824834108352661,
+      "learning_rate": 8.32817402522676e-06,
+      "loss": 0.8737,
+      "step": 3764
+    },
+    {
+      "epoch": 0.2902405180388529,
+      "grad_norm": 3.59409499168396,
+      "learning_rate": 8.327242233231784e-06,
+      "loss": 1.0568,
+      "step": 3765
+    },
+    {
+      "epoch": 0.2903176071538699,
+      "grad_norm": 3.5583786964416504,
+      "learning_rate": 8.326310233800488e-06,
+      "loss": 0.9326,
+      "step": 3766
+    },
+    {
+      "epoch": 0.2903946962688868,
+      "grad_norm": 3.7277417182922363,
+      "learning_rate": 8.325378026990976e-06,
+      "loss": 0.972,
+      "step": 3767
+    },
+    {
+      "epoch": 0.2904717853839038,
+      "grad_norm": 3.8586301803588867,
+      "learning_rate": 8.324445612861367e-06,
+      "loss": 1.0652,
+      "step": 3768
+    },
+    {
+      "epoch": 0.29054887449892075,
+      "grad_norm": 3.91520357131958,
+      "learning_rate": 8.32351299146979e-06,
+      "loss": 0.9839,
+      "step": 3769
+    },
+    {
+      "epoch": 0.2906259636139377,
+      "grad_norm": 3.715453863143921,
+      "learning_rate": 8.322580162874393e-06,
+      "loss": 1.1271,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2907030527289547,
+      "grad_norm": 3.917518138885498,
+      "learning_rate": 8.321647127133327e-06,
+      "loss": 1.0822,
+      "step": 3771
+    },
+    {
+      "epoch": 0.2907801418439716,
+      "grad_norm": 4.15562629699707,
+      "learning_rate": 8.320713884304769e-06,
+      "loss": 0.9178,
+      "step": 3772
+    },
+    {
+      "epoch": 0.2908572309589886,
+      "grad_norm": 3.6164376735687256,
+      "learning_rate": 8.3197804344469e-06,
+      "loss": 0.9087,
+      "step": 3773
+    },
+    {
+      "epoch": 0.29093432007400555,
+      "grad_norm": 3.697296380996704,
+      "learning_rate": 8.318846777617913e-06,
+      "loss": 1.0643,
+      "step": 3774
+    },
+    {
+      "epoch": 0.2910114091890225,
+      "grad_norm": 3.5980827808380127,
+      "learning_rate": 8.317912913876019e-06,
+      "loss": 1.009,
+      "step": 3775
+    },
+    {
+      "epoch": 0.2910884983040395,
+      "grad_norm": 3.787522077560425,
+      "learning_rate": 8.316978843279438e-06,
+      "loss": 1.124,
+      "step": 3776
+    },
+    {
+      "epoch": 0.2911655874190564,
+      "grad_norm": 3.599940299987793,
+      "learning_rate": 8.316044565886405e-06,
+      "loss": 1.0432,
+      "step": 3777
+    },
+    {
+      "epoch": 0.2912426765340734,
+      "grad_norm": 3.8612630367279053,
+      "learning_rate": 8.315110081755166e-06,
+      "loss": 0.9916,
+      "step": 3778
+    },
+    {
+      "epoch": 0.29131976564909035,
+      "grad_norm": 3.687645673751831,
+      "learning_rate": 8.314175390943987e-06,
+      "loss": 1.0088,
+      "step": 3779
+    },
+    {
+      "epoch": 0.2913968547641073,
+      "grad_norm": 3.6906330585479736,
+      "learning_rate": 8.313240493511132e-06,
+      "loss": 0.9998,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2914739438791243,
+      "grad_norm": 3.6079602241516113,
+      "learning_rate": 8.312305389514894e-06,
+      "loss": 1.0303,
+      "step": 3781
+    },
+    {
+      "epoch": 0.2915510329941412,
+      "grad_norm": 3.642810583114624,
+      "learning_rate": 8.31137007901357e-06,
+      "loss": 0.9044,
+      "step": 3782
+    },
+    {
+      "epoch": 0.29162812210915817,
+      "grad_norm": 3.4344608783721924,
+      "learning_rate": 8.310434562065472e-06,
+      "loss": 0.9379,
+      "step": 3783
+    },
+    {
+      "epoch": 0.29170521122417514,
+      "grad_norm": 3.8744451999664307,
+      "learning_rate": 8.309498838728924e-06,
+      "loss": 1.1047,
+      "step": 3784
+    },
+    {
+      "epoch": 0.2917823003391921,
+      "grad_norm": 3.5419445037841797,
+      "learning_rate": 8.308562909062264e-06,
+      "loss": 1.0705,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2918593894542091,
+      "grad_norm": 3.666670560836792,
+      "learning_rate": 8.307626773123842e-06,
+      "loss": 1.0101,
+      "step": 3786
+    },
+    {
+      "epoch": 0.291936478569226,
+      "grad_norm": 3.46775221824646,
+      "learning_rate": 8.306690430972023e-06,
+      "loss": 0.983,
+      "step": 3787
+    },
+    {
+      "epoch": 0.29201356768424297,
+      "grad_norm": 3.711484909057617,
+      "learning_rate": 8.305753882665178e-06,
+      "loss": 0.97,
+      "step": 3788
+    },
+    {
+      "epoch": 0.29209065679925994,
+      "grad_norm": 3.968447685241699,
+      "learning_rate": 8.304817128261702e-06,
+      "loss": 1.0543,
+      "step": 3789
+    },
+    {
+      "epoch": 0.2921677459142769,
+      "grad_norm": 3.9683480262756348,
+      "learning_rate": 8.303880167819994e-06,
+      "loss": 0.97,
+      "step": 3790
+    },
+    {
+      "epoch": 0.2922448350292939,
+      "grad_norm": 3.354612112045288,
+      "learning_rate": 8.302943001398466e-06,
+      "loss": 1.0434,
+      "step": 3791
+    },
+    {
+      "epoch": 0.2923219241443108,
+      "grad_norm": 4.114500999450684,
+      "learning_rate": 8.302005629055549e-06,
+      "loss": 1.0238,
+      "step": 3792
+    },
+    {
+      "epoch": 0.29239901325932777,
+      "grad_norm": 3.511687994003296,
+      "learning_rate": 8.301068050849685e-06,
+      "loss": 0.9847,
+      "step": 3793
+    },
+    {
+      "epoch": 0.29247610237434474,
+      "grad_norm": 3.513735294342041,
+      "learning_rate": 8.300130266839323e-06,
+      "loss": 0.9874,
+      "step": 3794
+    },
+    {
+      "epoch": 0.2925531914893617,
+      "grad_norm": 3.9487833976745605,
+      "learning_rate": 8.29919227708293e-06,
+      "loss": 1.0424,
+      "step": 3795
+    },
+    {
+      "epoch": 0.2926302806043787,
+      "grad_norm": 3.702522039413452,
+      "learning_rate": 8.298254081638988e-06,
+      "loss": 0.9582,
+      "step": 3796
+    },
+    {
+      "epoch": 0.2927073697193956,
+      "grad_norm": 4.400388240814209,
+      "learning_rate": 8.297315680565984e-06,
+      "loss": 1.0504,
+      "step": 3797
+    },
+    {
+      "epoch": 0.29278445883441256,
+      "grad_norm": 3.7499802112579346,
+      "learning_rate": 8.296377073922427e-06,
+      "loss": 1.0402,
+      "step": 3798
+    },
+    {
+      "epoch": 0.29286154794942953,
+      "grad_norm": 3.343437433242798,
+      "learning_rate": 8.295438261766829e-06,
+      "loss": 0.8718,
+      "step": 3799
+    },
+    {
+      "epoch": 0.2929386370644465,
+      "grad_norm": 3.8932011127471924,
+      "learning_rate": 8.294499244157724e-06,
+      "loss": 1.0511,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2930157261794635,
+      "grad_norm": 3.75187087059021,
+      "learning_rate": 8.293560021153652e-06,
+      "loss": 0.9994,
+      "step": 3801
+    },
+    {
+      "epoch": 0.29309281529448045,
+      "grad_norm": 4.034709930419922,
+      "learning_rate": 8.292620592813173e-06,
+      "loss": 1.0376,
+      "step": 3802
+    },
+    {
+      "epoch": 0.29316990440949736,
+      "grad_norm": 3.411324977874756,
+      "learning_rate": 8.291680959194852e-06,
+      "loss": 0.9321,
+      "step": 3803
+    },
+    {
+      "epoch": 0.29324699352451433,
+      "grad_norm": 3.779232978820801,
+      "learning_rate": 8.29074112035727e-06,
+      "loss": 1.0499,
+      "step": 3804
+    },
+    {
+      "epoch": 0.2933240826395313,
+      "grad_norm": 4.236926555633545,
+      "learning_rate": 8.289801076359025e-06,
+      "loss": 1.0988,
+      "step": 3805
+    },
+    {
+      "epoch": 0.2934011717545483,
+      "grad_norm": 3.3468666076660156,
+      "learning_rate": 8.28886082725872e-06,
+      "loss": 0.9899,
+      "step": 3806
+    },
+    {
+      "epoch": 0.29347826086956524,
+      "grad_norm": 3.675314426422119,
+      "learning_rate": 8.287920373114976e-06,
+      "loss": 0.9013,
+      "step": 3807
+    },
+    {
+      "epoch": 0.29355534998458216,
+      "grad_norm": 4.145412445068359,
+      "learning_rate": 8.286979713986426e-06,
+      "loss": 1.0496,
+      "step": 3808
+    },
+    {
+      "epoch": 0.29363243909959913,
+      "grad_norm": 4.000509738922119,
+      "learning_rate": 8.286038849931713e-06,
+      "loss": 1.0509,
+      "step": 3809
+    },
+    {
+      "epoch": 0.2937095282146161,
+      "grad_norm": 3.838998794555664,
+      "learning_rate": 8.285097781009497e-06,
+      "loss": 0.9234,
+      "step": 3810
+    },
+    {
+      "epoch": 0.29378661732963307,
+      "grad_norm": 3.82969069480896,
+      "learning_rate": 8.284156507278448e-06,
+      "loss": 1.0399,
+      "step": 3811
+    },
+    {
+      "epoch": 0.29386370644465004,
+      "grad_norm": 3.689297914505005,
+      "learning_rate": 8.283215028797252e-06,
+      "loss": 0.886,
+      "step": 3812
+    },
+    {
+      "epoch": 0.29394079555966696,
+      "grad_norm": 3.5808119773864746,
+      "learning_rate": 8.2822733456246e-06,
+      "loss": 1.0377,
+      "step": 3813
+    },
+    {
+      "epoch": 0.2940178846746839,
+      "grad_norm": 4.0584821701049805,
+      "learning_rate": 8.281331457819204e-06,
+      "loss": 1.1038,
+      "step": 3814
+    },
+    {
+      "epoch": 0.2940949737897009,
+      "grad_norm": 4.134930610656738,
+      "learning_rate": 8.28038936543979e-06,
+      "loss": 0.9839,
+      "step": 3815
+    },
+    {
+      "epoch": 0.29417206290471787,
+      "grad_norm": 3.6218936443328857,
+      "learning_rate": 8.279447068545085e-06,
+      "loss": 1.027,
+      "step": 3816
+    },
+    {
+      "epoch": 0.29424915201973484,
+      "grad_norm": 4.330170154571533,
+      "learning_rate": 8.27850456719384e-06,
+      "loss": 1.103,
+      "step": 3817
+    },
+    {
+      "epoch": 0.29432624113475175,
+      "grad_norm": 3.6379551887512207,
+      "learning_rate": 8.277561861444818e-06,
+      "loss": 1.0382,
+      "step": 3818
+    },
+    {
+      "epoch": 0.2944033302497687,
+      "grad_norm": 3.776359796524048,
+      "learning_rate": 8.276618951356787e-06,
+      "loss": 1.0924,
+      "step": 3819
+    },
+    {
+      "epoch": 0.2944804193647857,
+      "grad_norm": 3.7215919494628906,
+      "learning_rate": 8.275675836988535e-06,
+      "loss": 1.1014,
+      "step": 3820
+    },
+    {
+      "epoch": 0.29455750847980267,
+      "grad_norm": 3.9651782512664795,
+      "learning_rate": 8.27473251839886e-06,
+      "loss": 1.0527,
+      "step": 3821
+    },
+    {
+      "epoch": 0.29463459759481964,
+      "grad_norm": 3.5446691513061523,
+      "learning_rate": 8.273788995646571e-06,
+      "loss": 0.9641,
+      "step": 3822
+    },
+    {
+      "epoch": 0.29471168670983655,
+      "grad_norm": 3.6998002529144287,
+      "learning_rate": 8.272845268790494e-06,
+      "loss": 1.1796,
+      "step": 3823
+    },
+    {
+      "epoch": 0.2947887758248535,
+      "grad_norm": 3.6286091804504395,
+      "learning_rate": 8.271901337889468e-06,
+      "loss": 1.0408,
+      "step": 3824
+    },
+    {
+      "epoch": 0.2948658649398705,
+      "grad_norm": 3.6755175590515137,
+      "learning_rate": 8.270957203002337e-06,
+      "loss": 0.9034,
+      "step": 3825
+    },
+    {
+      "epoch": 0.29494295405488746,
+      "grad_norm": 3.6150128841400146,
+      "learning_rate": 8.270012864187965e-06,
+      "loss": 0.9529,
+      "step": 3826
+    },
+    {
+      "epoch": 0.29502004316990443,
+      "grad_norm": 3.8087635040283203,
+      "learning_rate": 8.269068321505225e-06,
+      "loss": 0.9891,
+      "step": 3827
+    },
+    {
+      "epoch": 0.29509713228492135,
+      "grad_norm": 3.8580355644226074,
+      "learning_rate": 8.268123575013008e-06,
+      "loss": 0.9276,
+      "step": 3828
+    },
+    {
+      "epoch": 0.2951742213999383,
+      "grad_norm": 3.485034227371216,
+      "learning_rate": 8.267178624770212e-06,
+      "loss": 1.0299,
+      "step": 3829
+    },
+    {
+      "epoch": 0.2952513105149553,
+      "grad_norm": 3.557379961013794,
+      "learning_rate": 8.26623347083575e-06,
+      "loss": 0.9734,
+      "step": 3830
+    },
+    {
+      "epoch": 0.29532839962997226,
+      "grad_norm": 3.9309725761413574,
+      "learning_rate": 8.265288113268548e-06,
+      "loss": 1.0063,
+      "step": 3831
+    },
+    {
+      "epoch": 0.29540548874498923,
+      "grad_norm": 3.5980284214019775,
+      "learning_rate": 8.264342552127542e-06,
+      "loss": 0.8885,
+      "step": 3832
+    },
+    {
+      "epoch": 0.29548257786000615,
+      "grad_norm": 3.4705045223236084,
+      "learning_rate": 8.263396787471685e-06,
+      "loss": 0.9333,
+      "step": 3833
+    },
+    {
+      "epoch": 0.2955596669750231,
+      "grad_norm": 3.6276614665985107,
+      "learning_rate": 8.26245081935994e-06,
+      "loss": 0.9775,
+      "step": 3834
+    },
+    {
+      "epoch": 0.2956367560900401,
+      "grad_norm": 3.9132349491119385,
+      "learning_rate": 8.261504647851283e-06,
+      "loss": 0.9186,
+      "step": 3835
+    },
+    {
+      "epoch": 0.29571384520505706,
+      "grad_norm": 3.6847267150878906,
+      "learning_rate": 8.260558273004703e-06,
+      "loss": 0.934,
+      "step": 3836
+    },
+    {
+      "epoch": 0.29579093432007403,
+      "grad_norm": 3.9422383308410645,
+      "learning_rate": 8.259611694879202e-06,
+      "loss": 0.9443,
+      "step": 3837
+    },
+    {
+      "epoch": 0.29586802343509094,
+      "grad_norm": 3.9089691638946533,
+      "learning_rate": 8.258664913533791e-06,
+      "loss": 1.0795,
+      "step": 3838
+    },
+    {
+      "epoch": 0.2959451125501079,
+      "grad_norm": 3.8518786430358887,
+      "learning_rate": 8.257717929027504e-06,
+      "loss": 1.1894,
+      "step": 3839
+    },
+    {
+      "epoch": 0.2960222016651249,
+      "grad_norm": 3.669287919998169,
+      "learning_rate": 8.256770741419374e-06,
+      "loss": 0.9979,
+      "step": 3840
+    },
+    {
+      "epoch": 0.29609929078014185,
+      "grad_norm": 4.07830286026001,
+      "learning_rate": 8.255823350768455e-06,
+      "loss": 1.0086,
+      "step": 3841
+    },
+    {
+      "epoch": 0.2961763798951588,
+      "grad_norm": 3.9762747287750244,
+      "learning_rate": 8.254875757133813e-06,
+      "loss": 1.0408,
+      "step": 3842
+    },
+    {
+      "epoch": 0.29625346901017574,
+      "grad_norm": 3.584479331970215,
+      "learning_rate": 8.253927960574525e-06,
+      "loss": 1.0119,
+      "step": 3843
+    },
+    {
+      "epoch": 0.2963305581251927,
+      "grad_norm": 3.4980578422546387,
+      "learning_rate": 8.252979961149683e-06,
+      "loss": 0.997,
+      "step": 3844
+    },
+    {
+      "epoch": 0.2964076472402097,
+      "grad_norm": 3.391723155975342,
+      "learning_rate": 8.252031758918386e-06,
+      "loss": 0.9273,
+      "step": 3845
+    },
+    {
+      "epoch": 0.29648473635522665,
+      "grad_norm": 3.5890274047851562,
+      "learning_rate": 8.251083353939752e-06,
+      "loss": 1.0653,
+      "step": 3846
+    },
+    {
+      "epoch": 0.2965618254702436,
+      "grad_norm": 3.8049674034118652,
+      "learning_rate": 8.250134746272909e-06,
+      "loss": 1.0471,
+      "step": 3847
+    },
+    {
+      "epoch": 0.29663891458526054,
+      "grad_norm": 3.6828417778015137,
+      "learning_rate": 8.249185935976998e-06,
+      "loss": 1.0122,
+      "step": 3848
+    },
+    {
+      "epoch": 0.2967160037002775,
+      "grad_norm": 3.8039426803588867,
+      "learning_rate": 8.24823692311117e-06,
+      "loss": 1.0566,
+      "step": 3849
+    },
+    {
+      "epoch": 0.2967930928152945,
+      "grad_norm": 4.135082721710205,
+      "learning_rate": 8.247287707734594e-06,
+      "loss": 1.0346,
+      "step": 3850
+    },
+    {
+      "epoch": 0.29687018193031145,
+      "grad_norm": 3.7773916721343994,
+      "learning_rate": 8.246338289906447e-06,
+      "loss": 0.9774,
+      "step": 3851
+    },
+    {
+      "epoch": 0.2969472710453284,
+      "grad_norm": 3.629441499710083,
+      "learning_rate": 8.245388669685922e-06,
+      "loss": 1.0698,
+      "step": 3852
+    },
+    {
+      "epoch": 0.29702436016034534,
+      "grad_norm": 3.7880666255950928,
+      "learning_rate": 8.24443884713222e-06,
+      "loss": 1.0878,
+      "step": 3853
+    },
+    {
+      "epoch": 0.2971014492753623,
+      "grad_norm": 3.7315597534179688,
+      "learning_rate": 8.243488822304561e-06,
+      "loss": 1.0335,
+      "step": 3854
+    },
+    {
+      "epoch": 0.2971785383903793,
+      "grad_norm": 4.079202651977539,
+      "learning_rate": 8.24253859526217e-06,
+      "loss": 1.0879,
+      "step": 3855
+    },
+    {
+      "epoch": 0.29725562750539625,
+      "grad_norm": 4.179590702056885,
+      "learning_rate": 8.241588166064294e-06,
+      "loss": 1.0974,
+      "step": 3856
+    },
+    {
+      "epoch": 0.2973327166204132,
+      "grad_norm": 3.6889772415161133,
+      "learning_rate": 8.240637534770182e-06,
+      "loss": 1.0824,
+      "step": 3857
+    },
+    {
+      "epoch": 0.29740980573543013,
+      "grad_norm": 3.830030679702759,
+      "learning_rate": 8.239686701439105e-06,
+      "loss": 1.0321,
+      "step": 3858
+    },
+    {
+      "epoch": 0.2974868948504471,
+      "grad_norm": 4.01475191116333,
+      "learning_rate": 8.23873566613034e-06,
+      "loss": 1.0373,
+      "step": 3859
+    },
+    {
+      "epoch": 0.2975639839654641,
+      "grad_norm": 3.6347076892852783,
+      "learning_rate": 8.237784428903182e-06,
+      "loss": 0.9952,
+      "step": 3860
+    },
+    {
+      "epoch": 0.29764107308048104,
+      "grad_norm": 3.510388135910034,
+      "learning_rate": 8.236832989816932e-06,
+      "loss": 0.9956,
+      "step": 3861
+    },
+    {
+      "epoch": 0.297718162195498,
+      "grad_norm": 3.1566152572631836,
+      "learning_rate": 8.23588134893091e-06,
+      "loss": 0.9462,
+      "step": 3862
+    },
+    {
+      "epoch": 0.29779525131051493,
+      "grad_norm": 4.099442481994629,
+      "learning_rate": 8.234929506304443e-06,
+      "loss": 0.8942,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2978723404255319,
+      "grad_norm": 3.81709885597229,
+      "learning_rate": 8.233977461996879e-06,
+      "loss": 0.9744,
+      "step": 3864
+    },
+    {
+      "epoch": 0.29794942954054887,
+      "grad_norm": 3.949967861175537,
+      "learning_rate": 8.233025216067567e-06,
+      "loss": 1.0687,
+      "step": 3865
+    },
+    {
+      "epoch": 0.29802651865556584,
+      "grad_norm": 3.714970111846924,
+      "learning_rate": 8.232072768575875e-06,
+      "loss": 1.1127,
+      "step": 3866
+    },
+    {
+      "epoch": 0.2981036077705828,
+      "grad_norm": 3.600409984588623,
+      "learning_rate": 8.231120119581189e-06,
+      "loss": 0.9565,
+      "step": 3867
+    },
+    {
+      "epoch": 0.2981806968855997,
+      "grad_norm": 3.3856570720672607,
+      "learning_rate": 8.230167269142897e-06,
+      "loss": 0.9503,
+      "step": 3868
+    },
+    {
+      "epoch": 0.2982577860006167,
+      "grad_norm": 4.353588581085205,
+      "learning_rate": 8.229214217320405e-06,
+      "loss": 1.0665,
+      "step": 3869
+    },
+    {
+      "epoch": 0.29833487511563367,
+      "grad_norm": 3.9014816284179688,
+      "learning_rate": 8.22826096417313e-06,
+      "loss": 1.003,
+      "step": 3870
+    },
+    {
+      "epoch": 0.29841196423065064,
+      "grad_norm": 4.135391712188721,
+      "learning_rate": 8.227307509760505e-06,
+      "loss": 1.0497,
+      "step": 3871
+    },
+    {
+      "epoch": 0.2984890533456676,
+      "grad_norm": 3.6809937953948975,
+      "learning_rate": 8.22635385414197e-06,
+      "loss": 1.09,
+      "step": 3872
+    },
+    {
+      "epoch": 0.2985661424606845,
+      "grad_norm": 3.6578569412231445,
+      "learning_rate": 8.225399997376984e-06,
+      "loss": 1.0836,
+      "step": 3873
+    },
+    {
+      "epoch": 0.2986432315757015,
+      "grad_norm": 3.6818604469299316,
+      "learning_rate": 8.22444593952501e-06,
+      "loss": 1.0283,
+      "step": 3874
+    },
+    {
+      "epoch": 0.29872032069071847,
+      "grad_norm": 3.6182854175567627,
+      "learning_rate": 8.223491680645533e-06,
+      "loss": 0.9956,
+      "step": 3875
+    },
+    {
+      "epoch": 0.29879740980573544,
+      "grad_norm": 3.5641016960144043,
+      "learning_rate": 8.222537220798046e-06,
+      "loss": 1.107,
+      "step": 3876
+    },
+    {
+      "epoch": 0.2988744989207524,
+      "grad_norm": 3.3794901371002197,
+      "learning_rate": 8.22158256004205e-06,
+      "loss": 0.9449,
+      "step": 3877
+    },
+    {
+      "epoch": 0.2989515880357693,
+      "grad_norm": 3.4423718452453613,
+      "learning_rate": 8.220627698437069e-06,
+      "loss": 1.0393,
+      "step": 3878
+    },
+    {
+      "epoch": 0.2990286771507863,
+      "grad_norm": 4.050296306610107,
+      "learning_rate": 8.21967263604263e-06,
+      "loss": 1.0819,
+      "step": 3879
+    },
+    {
+      "epoch": 0.29910576626580326,
+      "grad_norm": 3.7262563705444336,
+      "learning_rate": 8.218717372918277e-06,
+      "loss": 1.0081,
+      "step": 3880
+    },
+    {
+      "epoch": 0.29918285538082023,
+      "grad_norm": 3.7771074771881104,
+      "learning_rate": 8.217761909123567e-06,
+      "loss": 0.932,
+      "step": 3881
+    },
+    {
+      "epoch": 0.2992599444958372,
+      "grad_norm": 3.4512460231781006,
+      "learning_rate": 8.216806244718068e-06,
+      "loss": 0.9152,
+      "step": 3882
+    },
+    {
+      "epoch": 0.2993370336108541,
+      "grad_norm": 3.5177505016326904,
+      "learning_rate": 8.215850379761357e-06,
+      "loss": 0.9113,
+      "step": 3883
+    },
+    {
+      "epoch": 0.2994141227258711,
+      "grad_norm": 3.6450557708740234,
+      "learning_rate": 8.214894314313034e-06,
+      "loss": 0.964,
+      "step": 3884
+    },
+    {
+      "epoch": 0.29949121184088806,
+      "grad_norm": 3.773829698562622,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 1.1156,
+      "step": 3885
+    },
+    {
+      "epoch": 0.29956830095590503,
+      "grad_norm": 3.8969192504882812,
+      "learning_rate": 8.21298158217997e-06,
+      "loss": 1.045,
+      "step": 3886
+    },
+    {
+      "epoch": 0.299645390070922,
+      "grad_norm": 3.853782892227173,
+      "learning_rate": 8.212024915614482e-06,
+      "loss": 1.0426,
+      "step": 3887
+    },
+    {
+      "epoch": 0.299722479185939,
+      "grad_norm": 3.673476219177246,
+      "learning_rate": 8.211068048795877e-06,
+      "loss": 1.0286,
+      "step": 3888
+    },
+    {
+      "epoch": 0.2997995683009559,
+      "grad_norm": 3.4116404056549072,
+      "learning_rate": 8.210110981783807e-06,
+      "loss": 0.8559,
+      "step": 3889
+    },
+    {
+      "epoch": 0.29987665741597286,
+      "grad_norm": 3.4718472957611084,
+      "learning_rate": 8.209153714637943e-06,
+      "loss": 0.9776,
+      "step": 3890
+    },
+    {
+      "epoch": 0.29995374653098983,
+      "grad_norm": 3.6767029762268066,
+      "learning_rate": 8.208196247417968e-06,
+      "loss": 1.1066,
+      "step": 3891
+    },
+    {
+      "epoch": 0.3000308356460068,
+      "grad_norm": 3.493272066116333,
+      "learning_rate": 8.207238580183571e-06,
+      "loss": 1.0995,
+      "step": 3892
+    },
+    {
+      "epoch": 0.30010792476102377,
+      "grad_norm": 3.817072868347168,
+      "learning_rate": 8.206280712994459e-06,
+      "loss": 1.033,
+      "step": 3893
+    },
+    {
+      "epoch": 0.3001850138760407,
+      "grad_norm": 4.376103401184082,
+      "learning_rate": 8.205322645910352e-06,
+      "loss": 0.8971,
+      "step": 3894
+    },
+    {
+      "epoch": 0.30026210299105766,
+      "grad_norm": 3.5054662227630615,
+      "learning_rate": 8.204364378990976e-06,
+      "loss": 1.0437,
+      "step": 3895
+    },
+    {
+      "epoch": 0.3003391921060746,
+      "grad_norm": 3.6018145084381104,
+      "learning_rate": 8.203405912296079e-06,
+      "loss": 0.9484,
+      "step": 3896
+    },
+    {
+      "epoch": 0.3004162812210916,
+      "grad_norm": 3.9170470237731934,
+      "learning_rate": 8.202447245885414e-06,
+      "loss": 0.9874,
+      "step": 3897
+    },
+    {
+      "epoch": 0.30049337033610857,
+      "grad_norm": 3.545153856277466,
+      "learning_rate": 8.20148837981875e-06,
+      "loss": 0.9289,
+      "step": 3898
+    },
+    {
+      "epoch": 0.3005704594511255,
+      "grad_norm": 4.0050482749938965,
+      "learning_rate": 8.200529314155865e-06,
+      "loss": 1.1052,
+      "step": 3899
+    },
+    {
+      "epoch": 0.30064754856614245,
+      "grad_norm": 3.5742580890655518,
+      "learning_rate": 8.199570048956553e-06,
+      "loss": 0.9474,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3007246376811594,
+      "grad_norm": 3.7283453941345215,
+      "learning_rate": 8.19861058428062e-06,
+      "loss": 0.9697,
+      "step": 3901
+    },
+    {
+      "epoch": 0.3008017267961764,
+      "grad_norm": 3.4887049198150635,
+      "learning_rate": 8.197650920187882e-06,
+      "loss": 0.9355,
+      "step": 3902
+    },
+    {
+      "epoch": 0.30087881591119336,
+      "grad_norm": 4.035374641418457,
+      "learning_rate": 8.196691056738173e-06,
+      "loss": 1.0705,
+      "step": 3903
+    },
+    {
+      "epoch": 0.3009559050262103,
+      "grad_norm": 3.5932343006134033,
+      "learning_rate": 8.19573099399133e-06,
+      "loss": 1.0442,
+      "step": 3904
+    },
+    {
+      "epoch": 0.30103299414122725,
+      "grad_norm": 3.732898712158203,
+      "learning_rate": 8.19477073200721e-06,
+      "loss": 0.9415,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3011100832562442,
+      "grad_norm": 3.413482666015625,
+      "learning_rate": 8.193810270845683e-06,
+      "loss": 0.8275,
+      "step": 3906
+    },
+    {
+      "epoch": 0.3011871723712612,
+      "grad_norm": 4.12077522277832,
+      "learning_rate": 8.192849610566627e-06,
+      "loss": 0.9177,
+      "step": 3907
+    },
+    {
+      "epoch": 0.30126426148627816,
+      "grad_norm": 3.9081289768218994,
+      "learning_rate": 8.191888751229934e-06,
+      "loss": 0.9281,
+      "step": 3908
+    },
+    {
+      "epoch": 0.3013413506012951,
+      "grad_norm": 3.6880979537963867,
+      "learning_rate": 8.190927692895508e-06,
+      "loss": 1.0174,
+      "step": 3909
+    },
+    {
+      "epoch": 0.30141843971631205,
+      "grad_norm": 3.402858018875122,
+      "learning_rate": 8.189966435623266e-06,
+      "loss": 1.0482,
+      "step": 3910
+    },
+    {
+      "epoch": 0.301495528831329,
+      "grad_norm": 3.7268900871276855,
+      "learning_rate": 8.189004979473138e-06,
+      "loss": 1.1489,
+      "step": 3911
+    },
+    {
+      "epoch": 0.301572617946346,
+      "grad_norm": 3.7350738048553467,
+      "learning_rate": 8.188043324505067e-06,
+      "loss": 1.0077,
+      "step": 3912
+    },
+    {
+      "epoch": 0.30164970706136296,
+      "grad_norm": 3.5450897216796875,
+      "learning_rate": 8.187081470779006e-06,
+      "loss": 0.9558,
+      "step": 3913
+    },
+    {
+      "epoch": 0.3017267961763799,
+      "grad_norm": 3.7310359477996826,
+      "learning_rate": 8.18611941835492e-06,
+      "loss": 0.9612,
+      "step": 3914
+    },
+    {
+      "epoch": 0.30180388529139685,
+      "grad_norm": 3.8756463527679443,
+      "learning_rate": 8.185157167292791e-06,
+      "loss": 0.9793,
+      "step": 3915
+    },
+    {
+      "epoch": 0.3018809744064138,
+      "grad_norm": 4.031418323516846,
+      "learning_rate": 8.184194717652609e-06,
+      "loss": 1.0657,
+      "step": 3916
+    },
+    {
+      "epoch": 0.3019580635214308,
+      "grad_norm": 3.99357008934021,
+      "learning_rate": 8.183232069494378e-06,
+      "loss": 0.984,
+      "step": 3917
+    },
+    {
+      "epoch": 0.30203515263644776,
+      "grad_norm": 3.870607852935791,
+      "learning_rate": 8.182269222878112e-06,
+      "loss": 0.9012,
+      "step": 3918
+    },
+    {
+      "epoch": 0.30211224175146467,
+      "grad_norm": 3.4795637130737305,
+      "learning_rate": 8.181306177863843e-06,
+      "loss": 0.8978,
+      "step": 3919
+    },
+    {
+      "epoch": 0.30218933086648164,
+      "grad_norm": 3.652648687362671,
+      "learning_rate": 8.18034293451161e-06,
+      "loss": 1.0422,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3022664199814986,
+      "grad_norm": 3.868264675140381,
+      "learning_rate": 8.179379492881465e-06,
+      "loss": 1.0272,
+      "step": 3921
+    },
+    {
+      "epoch": 0.3023435090965156,
+      "grad_norm": 3.5765531063079834,
+      "learning_rate": 8.178415853033477e-06,
+      "loss": 0.9447,
+      "step": 3922
+    },
+    {
+      "epoch": 0.30242059821153255,
+      "grad_norm": 4.241714000701904,
+      "learning_rate": 8.177452015027721e-06,
+      "loss": 0.9008,
+      "step": 3923
+    },
+    {
+      "epoch": 0.30249768732654947,
+      "grad_norm": 3.668001413345337,
+      "learning_rate": 8.176487978924288e-06,
+      "loss": 0.9762,
+      "step": 3924
+    },
+    {
+      "epoch": 0.30257477644156644,
+      "grad_norm": 3.9289772510528564,
+      "learning_rate": 8.175523744783281e-06,
+      "loss": 1.0743,
+      "step": 3925
+    },
+    {
+      "epoch": 0.3026518655565834,
+      "grad_norm": 3.804590940475464,
+      "learning_rate": 8.174559312664815e-06,
+      "loss": 0.9993,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3027289546716004,
+      "grad_norm": 3.694150924682617,
+      "learning_rate": 8.173594682629018e-06,
+      "loss": 0.9693,
+      "step": 3927
+    },
+    {
+      "epoch": 0.30280604378661735,
+      "grad_norm": 3.8130977153778076,
+      "learning_rate": 8.172629854736029e-06,
+      "loss": 1.0525,
+      "step": 3928
+    },
+    {
+      "epoch": 0.30288313290163427,
+      "grad_norm": 4.602511405944824,
+      "learning_rate": 8.171664829046e-06,
+      "loss": 1.1879,
+      "step": 3929
+    },
+    {
+      "epoch": 0.30296022201665124,
+      "grad_norm": 3.7024033069610596,
+      "learning_rate": 8.170699605619096e-06,
+      "loss": 0.9487,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3030373111316682,
+      "grad_norm": 3.632319688796997,
+      "learning_rate": 8.169734184515493e-06,
+      "loss": 1.1081,
+      "step": 3931
+    },
+    {
+      "epoch": 0.3031144002466852,
+      "grad_norm": 3.9944117069244385,
+      "learning_rate": 8.168768565795377e-06,
+      "loss": 0.9995,
+      "step": 3932
+    },
+    {
+      "epoch": 0.30319148936170215,
+      "grad_norm": 3.9244701862335205,
+      "learning_rate": 8.167802749518956e-06,
+      "loss": 1.1025,
+      "step": 3933
+    },
+    {
+      "epoch": 0.30326857847671906,
+      "grad_norm": 3.9141428470611572,
+      "learning_rate": 8.166836735746438e-06,
+      "loss": 1.0337,
+      "step": 3934
+    },
+    {
+      "epoch": 0.30334566759173603,
+      "grad_norm": 3.821218967437744,
+      "learning_rate": 8.165870524538052e-06,
+      "loss": 1.0971,
+      "step": 3935
+    },
+    {
+      "epoch": 0.303422756706753,
+      "grad_norm": 4.062706470489502,
+      "learning_rate": 8.164904115954036e-06,
+      "loss": 1.1178,
+      "step": 3936
+    },
+    {
+      "epoch": 0.30349984582177,
+      "grad_norm": 3.5247349739074707,
+      "learning_rate": 8.163937510054638e-06,
+      "loss": 0.9598,
+      "step": 3937
+    },
+    {
+      "epoch": 0.30357693493678695,
+      "grad_norm": 3.914620876312256,
+      "learning_rate": 8.162970706900124e-06,
+      "loss": 1.0434,
+      "step": 3938
+    },
+    {
+      "epoch": 0.30365402405180386,
+      "grad_norm": 3.9191715717315674,
+      "learning_rate": 8.162003706550767e-06,
+      "loss": 1.163,
+      "step": 3939
+    },
+    {
+      "epoch": 0.30373111316682083,
+      "grad_norm": 3.459897994995117,
+      "learning_rate": 8.161036509066856e-06,
+      "loss": 1.0139,
+      "step": 3940
+    },
+    {
+      "epoch": 0.3038082022818378,
+      "grad_norm": 3.6575376987457275,
+      "learning_rate": 8.16006911450869e-06,
+      "loss": 0.9167,
+      "step": 3941
+    },
+    {
+      "epoch": 0.3038852913968548,
+      "grad_norm": 3.691288948059082,
+      "learning_rate": 8.159101522936582e-06,
+      "loss": 0.9456,
+      "step": 3942
+    },
+    {
+      "epoch": 0.30396238051187174,
+      "grad_norm": 4.538893222808838,
+      "learning_rate": 8.158133734410853e-06,
+      "loss": 1.0096,
+      "step": 3943
+    },
+    {
+      "epoch": 0.30403946962688866,
+      "grad_norm": 3.614013671875,
+      "learning_rate": 8.157165748991845e-06,
+      "loss": 1.0055,
+      "step": 3944
+    },
+    {
+      "epoch": 0.30411655874190563,
+      "grad_norm": 3.9665722846984863,
+      "learning_rate": 8.156197566739901e-06,
+      "loss": 1.1126,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3041936478569226,
+      "grad_norm": 3.547726631164551,
+      "learning_rate": 8.155229187715385e-06,
+      "loss": 1.0368,
+      "step": 3946
+    },
+    {
+      "epoch": 0.30427073697193957,
+      "grad_norm": 3.3927464485168457,
+      "learning_rate": 8.154260611978673e-06,
+      "loss": 1.0033,
+      "step": 3947
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 3.441749334335327,
+      "learning_rate": 8.153291839590147e-06,
+      "loss": 0.9378,
+      "step": 3948
+    },
+    {
+      "epoch": 0.30442491520197346,
+      "grad_norm": 3.7463624477386475,
+      "learning_rate": 8.152322870610206e-06,
+      "loss": 1.0562,
+      "step": 3949
+    },
+    {
+      "epoch": 0.3045020043169904,
+      "grad_norm": 3.6333110332489014,
+      "learning_rate": 8.151353705099261e-06,
+      "loss": 1.0003,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3045790934320074,
+      "grad_norm": 3.9135475158691406,
+      "learning_rate": 8.150384343117733e-06,
+      "loss": 1.1028,
+      "step": 3951
+    },
+    {
+      "epoch": 0.30465618254702437,
+      "grad_norm": 3.670753240585327,
+      "learning_rate": 8.149414784726058e-06,
+      "loss": 0.9757,
+      "step": 3952
+    },
+    {
+      "epoch": 0.30473327166204134,
+      "grad_norm": 3.691105365753174,
+      "learning_rate": 8.148445029984683e-06,
+      "loss": 1.0124,
+      "step": 3953
+    },
+    {
+      "epoch": 0.30481036077705825,
+      "grad_norm": 3.77278995513916,
+      "learning_rate": 8.147475078954067e-06,
+      "loss": 0.9402,
+      "step": 3954
+    },
+    {
+      "epoch": 0.3048874498920752,
+      "grad_norm": 3.847795009613037,
+      "learning_rate": 8.146504931694678e-06,
+      "loss": 1.0041,
+      "step": 3955
+    },
+    {
+      "epoch": 0.3049645390070922,
+      "grad_norm": 3.791046380996704,
+      "learning_rate": 8.145534588267006e-06,
+      "loss": 1.023,
+      "step": 3956
+    },
+    {
+      "epoch": 0.30504162812210917,
+      "grad_norm": 3.8239328861236572,
+      "learning_rate": 8.144564048731542e-06,
+      "loss": 0.9401,
+      "step": 3957
+    },
+    {
+      "epoch": 0.30511871723712614,
+      "grad_norm": 3.5764706134796143,
+      "learning_rate": 8.143593313148794e-06,
+      "loss": 0.9707,
+      "step": 3958
+    },
+    {
+      "epoch": 0.30519580635214305,
+      "grad_norm": 4.011223793029785,
+      "learning_rate": 8.142622381579285e-06,
+      "loss": 1.0103,
+      "step": 3959
+    },
+    {
+      "epoch": 0.30527289546716,
+      "grad_norm": 4.034478664398193,
+      "learning_rate": 8.14165125408355e-06,
+      "loss": 0.8388,
+      "step": 3960
+    },
+    {
+      "epoch": 0.305349984582177,
+      "grad_norm": 3.6760923862457275,
+      "learning_rate": 8.140679930722126e-06,
+      "loss": 1.0874,
+      "step": 3961
+    },
+    {
+      "epoch": 0.30542707369719396,
+      "grad_norm": 3.658113718032837,
+      "learning_rate": 8.139708411555575e-06,
+      "loss": 0.9909,
+      "step": 3962
+    },
+    {
+      "epoch": 0.30550416281221093,
+      "grad_norm": 4.3825364112854,
+      "learning_rate": 8.138736696644467e-06,
+      "loss": 1.0232,
+      "step": 3963
+    },
+    {
+      "epoch": 0.30558125192722785,
+      "grad_norm": 3.9493494033813477,
+      "learning_rate": 8.137764786049382e-06,
+      "loss": 0.9977,
+      "step": 3964
+    },
+    {
+      "epoch": 0.3056583410422448,
+      "grad_norm": 3.7453453540802,
+      "learning_rate": 8.136792679830913e-06,
+      "loss": 0.9792,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3057354301572618,
+      "grad_norm": 3.767953634262085,
+      "learning_rate": 8.135820378049667e-06,
+      "loss": 1.0847,
+      "step": 3966
+    },
+    {
+      "epoch": 0.30581251927227876,
+      "grad_norm": 3.4496097564697266,
+      "learning_rate": 8.13484788076626e-06,
+      "loss": 0.9574,
+      "step": 3967
+    },
+    {
+      "epoch": 0.30588960838729573,
+      "grad_norm": 3.7329258918762207,
+      "learning_rate": 8.133875188041323e-06,
+      "loss": 0.9645,
+      "step": 3968
+    },
+    {
+      "epoch": 0.3059666975023127,
+      "grad_norm": 3.522775888442993,
+      "learning_rate": 8.1329022999355e-06,
+      "loss": 1.0296,
+      "step": 3969
+    },
+    {
+      "epoch": 0.3060437866173296,
+      "grad_norm": 3.753998041152954,
+      "learning_rate": 8.131929216509445e-06,
+      "loss": 1.0741,
+      "step": 3970
+    },
+    {
+      "epoch": 0.3061208757323466,
+      "grad_norm": 3.6926846504211426,
+      "learning_rate": 8.130955937823821e-06,
+      "loss": 0.963,
+      "step": 3971
+    },
+    {
+      "epoch": 0.30619796484736356,
+      "grad_norm": 3.8235857486724854,
+      "learning_rate": 8.129982463939313e-06,
+      "loss": 1.0803,
+      "step": 3972
+    },
+    {
+      "epoch": 0.30627505396238053,
+      "grad_norm": 3.8400890827178955,
+      "learning_rate": 8.129008794916609e-06,
+      "loss": 1.074,
+      "step": 3973
+    },
+    {
+      "epoch": 0.3063521430773975,
+      "grad_norm": 3.7315080165863037,
+      "learning_rate": 8.12803493081641e-06,
+      "loss": 1.1322,
+      "step": 3974
+    },
+    {
+      "epoch": 0.3064292321924144,
+      "grad_norm": 3.864604949951172,
+      "learning_rate": 8.127060871699435e-06,
+      "loss": 1.0681,
+      "step": 3975
+    },
+    {
+      "epoch": 0.3065063213074314,
+      "grad_norm": 4.137643337249756,
+      "learning_rate": 8.12608661762641e-06,
+      "loss": 1.0562,
+      "step": 3976
+    },
+    {
+      "epoch": 0.30658341042244835,
+      "grad_norm": 3.801588296890259,
+      "learning_rate": 8.125112168658074e-06,
+      "loss": 1.0874,
+      "step": 3977
+    },
+    {
+      "epoch": 0.3066604995374653,
+      "grad_norm": 3.4541311264038086,
+      "learning_rate": 8.12413752485518e-06,
+      "loss": 0.9565,
+      "step": 3978
+    },
+    {
+      "epoch": 0.3067375886524823,
+      "grad_norm": 3.6811790466308594,
+      "learning_rate": 8.123162686278493e-06,
+      "loss": 1.0322,
+      "step": 3979
+    },
+    {
+      "epoch": 0.3068146777674992,
+      "grad_norm": 4.22686243057251,
+      "learning_rate": 8.122187652988786e-06,
+      "loss": 0.9425,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3068917668825162,
+      "grad_norm": 3.42470383644104,
+      "learning_rate": 8.12121242504685e-06,
+      "loss": 0.9987,
+      "step": 3981
+    },
+    {
+      "epoch": 0.30696885599753315,
+      "grad_norm": 3.6424388885498047,
+      "learning_rate": 8.120237002513484e-06,
+      "loss": 0.979,
+      "step": 3982
+    },
+    {
+      "epoch": 0.3070459451125501,
+      "grad_norm": 3.6647961139678955,
+      "learning_rate": 8.119261385449502e-06,
+      "loss": 1.0419,
+      "step": 3983
+    },
+    {
+      "epoch": 0.3071230342275671,
+      "grad_norm": 3.583043098449707,
+      "learning_rate": 8.118285573915726e-06,
+      "loss": 0.9044,
+      "step": 3984
+    },
+    {
+      "epoch": 0.307200123342584,
+      "grad_norm": 3.9360716342926025,
+      "learning_rate": 8.117309567972995e-06,
+      "loss": 1.0924,
+      "step": 3985
+    },
+    {
+      "epoch": 0.307277212457601,
+      "grad_norm": 3.999488592147827,
+      "learning_rate": 8.116333367682158e-06,
+      "loss": 1.0132,
+      "step": 3986
+    },
+    {
+      "epoch": 0.30735430157261795,
+      "grad_norm": 3.6380109786987305,
+      "learning_rate": 8.115356973104076e-06,
+      "loss": 1.0445,
+      "step": 3987
+    },
+    {
+      "epoch": 0.3074313906876349,
+      "grad_norm": 3.977320909500122,
+      "learning_rate": 8.11438038429962e-06,
+      "loss": 1.0968,
+      "step": 3988
+    },
+    {
+      "epoch": 0.3075084798026519,
+      "grad_norm": 3.90857195854187,
+      "learning_rate": 8.113403601329678e-06,
+      "loss": 1.1308,
+      "step": 3989
+    },
+    {
+      "epoch": 0.3075855689176688,
+      "grad_norm": 3.865142345428467,
+      "learning_rate": 8.112426624255145e-06,
+      "loss": 0.9358,
+      "step": 3990
+    },
+    {
+      "epoch": 0.3076626580326858,
+      "grad_norm": 3.489595890045166,
+      "learning_rate": 8.111449453136932e-06,
+      "loss": 0.9765,
+      "step": 3991
+    },
+    {
+      "epoch": 0.30773974714770275,
+      "grad_norm": 4.063758850097656,
+      "learning_rate": 8.110472088035961e-06,
+      "loss": 0.9821,
+      "step": 3992
+    },
+    {
+      "epoch": 0.3078168362627197,
+      "grad_norm": 3.65039324760437,
+      "learning_rate": 8.109494529013165e-06,
+      "loss": 1.0804,
+      "step": 3993
+    },
+    {
+      "epoch": 0.3078939253777367,
+      "grad_norm": 3.800696611404419,
+      "learning_rate": 8.108516776129489e-06,
+      "loss": 0.8819,
+      "step": 3994
+    },
+    {
+      "epoch": 0.3079710144927536,
+      "grad_norm": 3.8089771270751953,
+      "learning_rate": 8.107538829445891e-06,
+      "loss": 1.0294,
+      "step": 3995
+    },
+    {
+      "epoch": 0.3080481036077706,
+      "grad_norm": 3.4781205654144287,
+      "learning_rate": 8.106560689023342e-06,
+      "loss": 0.9528,
+      "step": 3996
+    },
+    {
+      "epoch": 0.30812519272278754,
+      "grad_norm": 3.5084052085876465,
+      "learning_rate": 8.105582354922822e-06,
+      "loss": 0.9852,
+      "step": 3997
+    },
+    {
+      "epoch": 0.3082022818378045,
+      "grad_norm": 3.660351037979126,
+      "learning_rate": 8.104603827205329e-06,
+      "loss": 1.032,
+      "step": 3998
+    },
+    {
+      "epoch": 0.3082793709528215,
+      "grad_norm": 3.5761704444885254,
+      "learning_rate": 8.103625105931865e-06,
+      "loss": 0.9191,
+      "step": 3999
+    },
+    {
+      "epoch": 0.3083564600678384,
+      "grad_norm": 3.8501265048980713,
+      "learning_rate": 8.10264619116345e-06,
+      "loss": 1.0343,
+      "step": 4000
+    },
+    {
+      "epoch": 0.30843354918285537,
+      "grad_norm": 3.4951345920562744,
+      "learning_rate": 8.101667082961114e-06,
+      "loss": 0.9921,
+      "step": 4001
+    },
+    {
+      "epoch": 0.30851063829787234,
+      "grad_norm": 3.844111204147339,
+      "learning_rate": 8.1006877813859e-06,
+      "loss": 0.979,
+      "step": 4002
+    },
+    {
+      "epoch": 0.3085877274128893,
+      "grad_norm": 3.7992372512817383,
+      "learning_rate": 8.09970828649886e-06,
+      "loss": 1.0304,
+      "step": 4003
+    },
+    {
+      "epoch": 0.3086648165279063,
+      "grad_norm": 4.003841876983643,
+      "learning_rate": 8.098728598361063e-06,
+      "loss": 1.0203,
+      "step": 4004
+    },
+    {
+      "epoch": 0.3087419056429232,
+      "grad_norm": 3.7176103591918945,
+      "learning_rate": 8.097748717033587e-06,
+      "loss": 0.9943,
+      "step": 4005
+    },
+    {
+      "epoch": 0.30881899475794017,
+      "grad_norm": 3.827420234680176,
+      "learning_rate": 8.096768642577521e-06,
+      "loss": 0.9806,
+      "step": 4006
+    },
+    {
+      "epoch": 0.30889608387295714,
+      "grad_norm": 3.841545820236206,
+      "learning_rate": 8.09578837505397e-06,
+      "loss": 1.0092,
+      "step": 4007
+    },
+    {
+      "epoch": 0.3089731729879741,
+      "grad_norm": 3.845114231109619,
+      "learning_rate": 8.094807914524048e-06,
+      "loss": 1.0792,
+      "step": 4008
+    },
+    {
+      "epoch": 0.3090502621029911,
+      "grad_norm": 3.8567841053009033,
+      "learning_rate": 8.093827261048879e-06,
+      "loss": 1.0728,
+      "step": 4009
+    },
+    {
+      "epoch": 0.309127351218008,
+      "grad_norm": 3.801530361175537,
+      "learning_rate": 8.092846414689605e-06,
+      "loss": 0.9051,
+      "step": 4010
+    },
+    {
+      "epoch": 0.30920444033302497,
+      "grad_norm": 3.59698748588562,
+      "learning_rate": 8.091865375507375e-06,
+      "loss": 1.0773,
+      "step": 4011
+    },
+    {
+      "epoch": 0.30928152944804194,
+      "grad_norm": 3.8643345832824707,
+      "learning_rate": 8.090884143563352e-06,
+      "loss": 0.9275,
+      "step": 4012
+    },
+    {
+      "epoch": 0.3093586185630589,
+      "grad_norm": 3.8212454319000244,
+      "learning_rate": 8.08990271891871e-06,
+      "loss": 1.0513,
+      "step": 4013
+    },
+    {
+      "epoch": 0.3094357076780759,
+      "grad_norm": 3.5950045585632324,
+      "learning_rate": 8.088921101634637e-06,
+      "loss": 1.0344,
+      "step": 4014
+    },
+    {
+      "epoch": 0.3095127967930928,
+      "grad_norm": 3.403311252593994,
+      "learning_rate": 8.087939291772331e-06,
+      "loss": 0.9926,
+      "step": 4015
+    },
+    {
+      "epoch": 0.30958988590810976,
+      "grad_norm": 4.000549793243408,
+      "learning_rate": 8.086957289393002e-06,
+      "loss": 1.0657,
+      "step": 4016
+    },
+    {
+      "epoch": 0.30966697502312673,
+      "grad_norm": 3.72603702545166,
+      "learning_rate": 8.085975094557876e-06,
+      "loss": 1.061,
+      "step": 4017
+    },
+    {
+      "epoch": 0.3097440641381437,
+      "grad_norm": 3.4024152755737305,
+      "learning_rate": 8.084992707328184e-06,
+      "loss": 0.9515,
+      "step": 4018
+    },
+    {
+      "epoch": 0.3098211532531607,
+      "grad_norm": 3.527376174926758,
+      "learning_rate": 8.084010127765174e-06,
+      "loss": 0.9454,
+      "step": 4019
+    },
+    {
+      "epoch": 0.3098982423681776,
+      "grad_norm": 3.594805955886841,
+      "learning_rate": 8.083027355930106e-06,
+      "loss": 0.9794,
+      "step": 4020
+    },
+    {
+      "epoch": 0.30997533148319456,
+      "grad_norm": 3.747175693511963,
+      "learning_rate": 8.08204439188425e-06,
+      "loss": 0.9221,
+      "step": 4021
+    },
+    {
+      "epoch": 0.31005242059821153,
+      "grad_norm": 3.6072657108306885,
+      "learning_rate": 8.081061235688889e-06,
+      "loss": 1.0778,
+      "step": 4022
+    },
+    {
+      "epoch": 0.3101295097132285,
+      "grad_norm": 3.919334888458252,
+      "learning_rate": 8.080077887405315e-06,
+      "loss": 0.9001,
+      "step": 4023
+    },
+    {
+      "epoch": 0.3102065988282455,
+      "grad_norm": 4.0752973556518555,
+      "learning_rate": 8.079094347094839e-06,
+      "loss": 1.0525,
+      "step": 4024
+    },
+    {
+      "epoch": 0.3102836879432624,
+      "grad_norm": 3.7410547733306885,
+      "learning_rate": 8.078110614818777e-06,
+      "loss": 0.9558,
+      "step": 4025
+    },
+    {
+      "epoch": 0.31036077705827936,
+      "grad_norm": 3.7749099731445312,
+      "learning_rate": 8.07712669063846e-06,
+      "loss": 0.9536,
+      "step": 4026
+    },
+    {
+      "epoch": 0.31043786617329633,
+      "grad_norm": 3.8618111610412598,
+      "learning_rate": 8.07614257461523e-06,
+      "loss": 0.9301,
+      "step": 4027
+    },
+    {
+      "epoch": 0.3105149552883133,
+      "grad_norm": 3.772061586380005,
+      "learning_rate": 8.075158266810442e-06,
+      "loss": 1.0186,
+      "step": 4028
+    },
+    {
+      "epoch": 0.31059204440333027,
+      "grad_norm": 4.043780326843262,
+      "learning_rate": 8.074173767285465e-06,
+      "loss": 1.0044,
+      "step": 4029
+    },
+    {
+      "epoch": 0.3106691335183472,
+      "grad_norm": 4.081550598144531,
+      "learning_rate": 8.073189076101673e-06,
+      "loss": 1.1462,
+      "step": 4030
+    },
+    {
+      "epoch": 0.31074622263336416,
+      "grad_norm": 3.861663818359375,
+      "learning_rate": 8.072204193320459e-06,
+      "loss": 1.1358,
+      "step": 4031
+    },
+    {
+      "epoch": 0.3108233117483811,
+      "grad_norm": 3.9835574626922607,
+      "learning_rate": 8.071219119003223e-06,
+      "loss": 1.1444,
+      "step": 4032
+    },
+    {
+      "epoch": 0.3109004008633981,
+      "grad_norm": 3.3269333839416504,
+      "learning_rate": 8.070233853211385e-06,
+      "loss": 0.8577,
+      "step": 4033
+    },
+    {
+      "epoch": 0.31097748997841507,
+      "grad_norm": 4.0250091552734375,
+      "learning_rate": 8.069248396006365e-06,
+      "loss": 0.9865,
+      "step": 4034
+    },
+    {
+      "epoch": 0.311054579093432,
+      "grad_norm": 3.576920986175537,
+      "learning_rate": 8.068262747449604e-06,
+      "loss": 1.0196,
+      "step": 4035
+    },
+    {
+      "epoch": 0.31113166820844895,
+      "grad_norm": 3.694272994995117,
+      "learning_rate": 8.067276907602551e-06,
+      "loss": 1.0175,
+      "step": 4036
+    },
+    {
+      "epoch": 0.3112087573234659,
+      "grad_norm": 3.5465962886810303,
+      "learning_rate": 8.06629087652667e-06,
+      "loss": 1.0006,
+      "step": 4037
+    },
+    {
+      "epoch": 0.3112858464384829,
+      "grad_norm": 3.7290923595428467,
+      "learning_rate": 8.065304654283434e-06,
+      "loss": 0.9712,
+      "step": 4038
+    },
+    {
+      "epoch": 0.31136293555349986,
+      "grad_norm": 3.9318015575408936,
+      "learning_rate": 8.064318240934327e-06,
+      "loss": 0.9798,
+      "step": 4039
+    },
+    {
+      "epoch": 0.3114400246685168,
+      "grad_norm": 3.745164394378662,
+      "learning_rate": 8.063331636540848e-06,
+      "loss": 0.9533,
+      "step": 4040
+    },
+    {
+      "epoch": 0.31151711378353375,
+      "grad_norm": 4.017617225646973,
+      "learning_rate": 8.062344841164508e-06,
+      "loss": 1.0166,
+      "step": 4041
+    },
+    {
+      "epoch": 0.3115942028985507,
+      "grad_norm": 3.4828991889953613,
+      "learning_rate": 8.061357854866827e-06,
+      "loss": 0.9534,
+      "step": 4042
+    },
+    {
+      "epoch": 0.3116712920135677,
+      "grad_norm": 3.8771891593933105,
+      "learning_rate": 8.060370677709338e-06,
+      "loss": 0.9229,
+      "step": 4043
+    },
+    {
+      "epoch": 0.31174838112858466,
+      "grad_norm": 4.284106731414795,
+      "learning_rate": 8.059383309753587e-06,
+      "loss": 0.8905,
+      "step": 4044
+    },
+    {
+      "epoch": 0.3118254702436016,
+      "grad_norm": 3.635425567626953,
+      "learning_rate": 8.058395751061135e-06,
+      "loss": 1.0268,
+      "step": 4045
+    },
+    {
+      "epoch": 0.31190255935861855,
+      "grad_norm": 3.7544236183166504,
+      "learning_rate": 8.057408001693544e-06,
+      "loss": 0.9735,
+      "step": 4046
+    },
+    {
+      "epoch": 0.3119796484736355,
+      "grad_norm": 3.6741573810577393,
+      "learning_rate": 8.0564200617124e-06,
+      "loss": 1.0407,
+      "step": 4047
+    },
+    {
+      "epoch": 0.3120567375886525,
+      "grad_norm": 3.6029105186462402,
+      "learning_rate": 8.055431931179296e-06,
+      "loss": 0.9976,
+      "step": 4048
+    },
+    {
+      "epoch": 0.31213382670366946,
+      "grad_norm": 3.3681766986846924,
+      "learning_rate": 8.054443610155836e-06,
+      "loss": 0.9445,
+      "step": 4049
+    },
+    {
+      "epoch": 0.3122109158186864,
+      "grad_norm": 3.6307098865509033,
+      "learning_rate": 8.053455098703635e-06,
+      "loss": 0.9859,
+      "step": 4050
+    },
+    {
+      "epoch": 0.31228800493370334,
+      "grad_norm": 3.4906904697418213,
+      "learning_rate": 8.052466396884323e-06,
+      "loss": 1.0344,
+      "step": 4051
+    },
+    {
+      "epoch": 0.3123650940487203,
+      "grad_norm": 3.574552297592163,
+      "learning_rate": 8.05147750475954e-06,
+      "loss": 0.9103,
+      "step": 4052
+    },
+    {
+      "epoch": 0.3124421831637373,
+      "grad_norm": 4.1314377784729,
+      "learning_rate": 8.050488422390939e-06,
+      "loss": 1.0707,
+      "step": 4053
+    },
+    {
+      "epoch": 0.31251927227875426,
+      "grad_norm": 4.234683990478516,
+      "learning_rate": 8.049499149840183e-06,
+      "loss": 1.0068,
+      "step": 4054
+    },
+    {
+      "epoch": 0.3125963613937712,
+      "grad_norm": 3.7397048473358154,
+      "learning_rate": 8.048509687168949e-06,
+      "loss": 0.9691,
+      "step": 4055
+    },
+    {
+      "epoch": 0.31267345050878814,
+      "grad_norm": 3.67864727973938,
+      "learning_rate": 8.047520034438925e-06,
+      "loss": 0.9445,
+      "step": 4056
+    },
+    {
+      "epoch": 0.3127505396238051,
+      "grad_norm": 4.083065032958984,
+      "learning_rate": 8.046530191711808e-06,
+      "loss": 1.1873,
+      "step": 4057
+    },
+    {
+      "epoch": 0.3128276287388221,
+      "grad_norm": 3.7557568550109863,
+      "learning_rate": 8.045540159049315e-06,
+      "loss": 1.0247,
+      "step": 4058
+    },
+    {
+      "epoch": 0.31290471785383905,
+      "grad_norm": 3.5163233280181885,
+      "learning_rate": 8.044549936513165e-06,
+      "loss": 1.039,
+      "step": 4059
+    },
+    {
+      "epoch": 0.312981806968856,
+      "grad_norm": 7.071890354156494,
+      "learning_rate": 8.043559524165096e-06,
+      "loss": 1.0503,
+      "step": 4060
+    },
+    {
+      "epoch": 0.31305889608387294,
+      "grad_norm": 3.8393161296844482,
+      "learning_rate": 8.042568922066852e-06,
+      "loss": 1.083,
+      "step": 4061
+    },
+    {
+      "epoch": 0.3131359851988899,
+      "grad_norm": 4.509868144989014,
+      "learning_rate": 8.041578130280194e-06,
+      "loss": 1.0581,
+      "step": 4062
+    },
+    {
+      "epoch": 0.3132130743139069,
+      "grad_norm": 4.245510578155518,
+      "learning_rate": 8.040587148866893e-06,
+      "loss": 0.9752,
+      "step": 4063
+    },
+    {
+      "epoch": 0.31329016342892385,
+      "grad_norm": 4.26503849029541,
+      "learning_rate": 8.03959597788873e-06,
+      "loss": 0.9107,
+      "step": 4064
+    },
+    {
+      "epoch": 0.3133672525439408,
+      "grad_norm": 3.7438037395477295,
+      "learning_rate": 8.038604617407501e-06,
+      "loss": 0.9565,
+      "step": 4065
+    },
+    {
+      "epoch": 0.31344434165895774,
+      "grad_norm": 5.089138507843018,
+      "learning_rate": 8.037613067485012e-06,
+      "loss": 1.0168,
+      "step": 4066
+    },
+    {
+      "epoch": 0.3135214307739747,
+      "grad_norm": 4.153073310852051,
+      "learning_rate": 8.036621328183079e-06,
+      "loss": 1.0259,
+      "step": 4067
+    },
+    {
+      "epoch": 0.3135985198889917,
+      "grad_norm": 3.4369683265686035,
+      "learning_rate": 8.035629399563533e-06,
+      "loss": 0.9818,
+      "step": 4068
+    },
+    {
+      "epoch": 0.31367560900400865,
+      "grad_norm": 3.7807652950286865,
+      "learning_rate": 8.034637281688219e-06,
+      "loss": 1.0017,
+      "step": 4069
+    },
+    {
+      "epoch": 0.3137526981190256,
+      "grad_norm": 3.68034291267395,
+      "learning_rate": 8.033644974618983e-06,
+      "loss": 0.9743,
+      "step": 4070
+    },
+    {
+      "epoch": 0.31382978723404253,
+      "grad_norm": 3.763885021209717,
+      "learning_rate": 8.032652478417697e-06,
+      "loss": 0.9948,
+      "step": 4071
+    },
+    {
+      "epoch": 0.3139068763490595,
+      "grad_norm": 3.9806506633758545,
+      "learning_rate": 8.031659793146237e-06,
+      "loss": 1.1496,
+      "step": 4072
+    },
+    {
+      "epoch": 0.3139839654640765,
+      "grad_norm": 3.6926496028900146,
+      "learning_rate": 8.030666918866487e-06,
+      "loss": 1.0143,
+      "step": 4073
+    },
+    {
+      "epoch": 0.31406105457909345,
+      "grad_norm": 3.7122464179992676,
+      "learning_rate": 8.029673855640352e-06,
+      "loss": 1.0272,
+      "step": 4074
+    },
+    {
+      "epoch": 0.3141381436941104,
+      "grad_norm": 3.5680923461914062,
+      "learning_rate": 8.028680603529742e-06,
+      "loss": 0.8766,
+      "step": 4075
+    },
+    {
+      "epoch": 0.31421523280912733,
+      "grad_norm": 3.5684890747070312,
+      "learning_rate": 8.027687162596584e-06,
+      "loss": 0.9874,
+      "step": 4076
+    },
+    {
+      "epoch": 0.3142923219241443,
+      "grad_norm": 3.833163022994995,
+      "learning_rate": 8.026693532902811e-06,
+      "loss": 1.0566,
+      "step": 4077
+    },
+    {
+      "epoch": 0.3143694110391613,
+      "grad_norm": 4.123239040374756,
+      "learning_rate": 8.025699714510374e-06,
+      "loss": 1.093,
+      "step": 4078
+    },
+    {
+      "epoch": 0.31444650015417824,
+      "grad_norm": 3.9509053230285645,
+      "learning_rate": 8.024705707481228e-06,
+      "loss": 0.9761,
+      "step": 4079
+    },
+    {
+      "epoch": 0.3145235892691952,
+      "grad_norm": 3.3474016189575195,
+      "learning_rate": 8.023711511877347e-06,
+      "loss": 0.9236,
+      "step": 4080
+    },
+    {
+      "epoch": 0.31460067838421213,
+      "grad_norm": 4.015679359436035,
+      "learning_rate": 8.022717127760715e-06,
+      "loss": 1.0176,
+      "step": 4081
+    },
+    {
+      "epoch": 0.3146777674992291,
+      "grad_norm": 3.6416444778442383,
+      "learning_rate": 8.021722555193324e-06,
+      "loss": 1.0104,
+      "step": 4082
+    },
+    {
+      "epoch": 0.31475485661424607,
+      "grad_norm": 3.4460465908050537,
+      "learning_rate": 8.020727794237182e-06,
+      "loss": 0.8351,
+      "step": 4083
+    },
+    {
+      "epoch": 0.31483194572926304,
+      "grad_norm": 3.7392327785491943,
+      "learning_rate": 8.019732844954306e-06,
+      "loss": 1.059,
+      "step": 4084
+    },
+    {
+      "epoch": 0.31490903484428,
+      "grad_norm": 3.5918447971343994,
+      "learning_rate": 8.018737707406728e-06,
+      "loss": 1.0905,
+      "step": 4085
+    },
+    {
+      "epoch": 0.3149861239592969,
+      "grad_norm": 3.957096576690674,
+      "learning_rate": 8.017742381656486e-06,
+      "loss": 0.9457,
+      "step": 4086
+    },
+    {
+      "epoch": 0.3150632130743139,
+      "grad_norm": 3.9353511333465576,
+      "learning_rate": 8.016746867765639e-06,
+      "loss": 0.9994,
+      "step": 4087
+    },
+    {
+      "epoch": 0.31514030218933087,
+      "grad_norm": 3.582047700881958,
+      "learning_rate": 8.015751165796247e-06,
+      "loss": 1.0187,
+      "step": 4088
+    },
+    {
+      "epoch": 0.31521739130434784,
+      "grad_norm": 3.996525287628174,
+      "learning_rate": 8.014755275810389e-06,
+      "loss": 1.0468,
+      "step": 4089
+    },
+    {
+      "epoch": 0.3152944804193648,
+      "grad_norm": 3.70849871635437,
+      "learning_rate": 8.013759197870153e-06,
+      "loss": 1.1356,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3153715695343817,
+      "grad_norm": 3.3762667179107666,
+      "learning_rate": 8.012762932037638e-06,
+      "loss": 0.873,
+      "step": 4091
+    },
+    {
+      "epoch": 0.3154486586493987,
+      "grad_norm": 3.614732503890991,
+      "learning_rate": 8.011766478374961e-06,
+      "loss": 0.9919,
+      "step": 4092
+    },
+    {
+      "epoch": 0.31552574776441566,
+      "grad_norm": 3.7229347229003906,
+      "learning_rate": 8.010769836944241e-06,
+      "loss": 0.9781,
+      "step": 4093
+    },
+    {
+      "epoch": 0.31560283687943264,
+      "grad_norm": 3.661196231842041,
+      "learning_rate": 8.009773007807615e-06,
+      "loss": 1.0271,
+      "step": 4094
+    },
+    {
+      "epoch": 0.3156799259944496,
+      "grad_norm": 3.394929885864258,
+      "learning_rate": 8.00877599102723e-06,
+      "loss": 1.0682,
+      "step": 4095
+    },
+    {
+      "epoch": 0.3157570151094665,
+      "grad_norm": 3.5974974632263184,
+      "learning_rate": 8.007778786665246e-06,
+      "loss": 0.9009,
+      "step": 4096
+    },
+    {
+      "epoch": 0.3158341042244835,
+      "grad_norm": 3.423668622970581,
+      "learning_rate": 8.006781394783831e-06,
+      "loss": 0.9942,
+      "step": 4097
+    },
+    {
+      "epoch": 0.31591119333950046,
+      "grad_norm": 3.99906849861145,
+      "learning_rate": 8.005783815445168e-06,
+      "loss": 1.1061,
+      "step": 4098
+    },
+    {
+      "epoch": 0.31598828245451743,
+      "grad_norm": 3.8605785369873047,
+      "learning_rate": 8.004786048711452e-06,
+      "loss": 1.059,
+      "step": 4099
+    },
+    {
+      "epoch": 0.3160653715695344,
+      "grad_norm": 3.41093373298645,
+      "learning_rate": 8.003788094644888e-06,
+      "loss": 0.9386,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3161424606845513,
+      "grad_norm": 4.836550235748291,
+      "learning_rate": 8.002789953307692e-06,
+      "loss": 0.9721,
+      "step": 4101
+    },
+    {
+      "epoch": 0.3162195497995683,
+      "grad_norm": 3.4783308506011963,
+      "learning_rate": 8.001791624762097e-06,
+      "loss": 0.8727,
+      "step": 4102
+    },
+    {
+      "epoch": 0.31629663891458526,
+      "grad_norm": 3.9497110843658447,
+      "learning_rate": 8.00079310907034e-06,
+      "loss": 1.0713,
+      "step": 4103
+    },
+    {
+      "epoch": 0.31637372802960223,
+      "grad_norm": 3.943293809890747,
+      "learning_rate": 7.999794406294674e-06,
+      "loss": 0.9778,
+      "step": 4104
+    },
+    {
+      "epoch": 0.3164508171446192,
+      "grad_norm": 4.106851100921631,
+      "learning_rate": 7.998795516497362e-06,
+      "loss": 1.0506,
+      "step": 4105
+    },
+    {
+      "epoch": 0.3165279062596361,
+      "grad_norm": 4.133145332336426,
+      "learning_rate": 7.997796439740682e-06,
+      "loss": 1.1453,
+      "step": 4106
+    },
+    {
+      "epoch": 0.3166049953746531,
+      "grad_norm": 3.896491289138794,
+      "learning_rate": 7.99679717608692e-06,
+      "loss": 1.0087,
+      "step": 4107
+    },
+    {
+      "epoch": 0.31668208448967006,
+      "grad_norm": 3.7162492275238037,
+      "learning_rate": 7.995797725598373e-06,
+      "loss": 1.0709,
+      "step": 4108
+    },
+    {
+      "epoch": 0.316759173604687,
+      "grad_norm": 4.147136688232422,
+      "learning_rate": 7.994798088337357e-06,
+      "loss": 1.0759,
+      "step": 4109
+    },
+    {
+      "epoch": 0.316836262719704,
+      "grad_norm": 3.5119404792785645,
+      "learning_rate": 7.993798264366189e-06,
+      "loss": 0.9334,
+      "step": 4110
+    },
+    {
+      "epoch": 0.3169133518347209,
+      "grad_norm": 3.72426438331604,
+      "learning_rate": 7.992798253747202e-06,
+      "loss": 1.002,
+      "step": 4111
+    },
+    {
+      "epoch": 0.3169904409497379,
+      "grad_norm": 3.564457654953003,
+      "learning_rate": 7.991798056542747e-06,
+      "loss": 1.0832,
+      "step": 4112
+    },
+    {
+      "epoch": 0.31706753006475485,
+      "grad_norm": 3.688892126083374,
+      "learning_rate": 7.990797672815177e-06,
+      "loss": 1.0163,
+      "step": 4113
+    },
+    {
+      "epoch": 0.3171446191797718,
+      "grad_norm": 3.9110841751098633,
+      "learning_rate": 7.989797102626862e-06,
+      "loss": 1.015,
+      "step": 4114
+    },
+    {
+      "epoch": 0.3172217082947888,
+      "grad_norm": 3.6422502994537354,
+      "learning_rate": 7.988796346040182e-06,
+      "loss": 1.0146,
+      "step": 4115
+    },
+    {
+      "epoch": 0.3172987974098057,
+      "grad_norm": 3.7644057273864746,
+      "learning_rate": 7.987795403117528e-06,
+      "loss": 1.0263,
+      "step": 4116
+    },
+    {
+      "epoch": 0.3173758865248227,
+      "grad_norm": 3.885549306869507,
+      "learning_rate": 7.986794273921309e-06,
+      "loss": 1.0733,
+      "step": 4117
+    },
+    {
+      "epoch": 0.31745297563983965,
+      "grad_norm": 3.661712169647217,
+      "learning_rate": 7.985792958513932e-06,
+      "loss": 1.0752,
+      "step": 4118
+    },
+    {
+      "epoch": 0.3175300647548566,
+      "grad_norm": 3.5156235694885254,
+      "learning_rate": 7.98479145695783e-06,
+      "loss": 0.9757,
+      "step": 4119
+    },
+    {
+      "epoch": 0.3176071538698736,
+      "grad_norm": 3.432779312133789,
+      "learning_rate": 7.983789769315438e-06,
+      "loss": 1.0079,
+      "step": 4120
+    },
+    {
+      "epoch": 0.3176842429848905,
+      "grad_norm": 3.5986807346343994,
+      "learning_rate": 7.982787895649207e-06,
+      "loss": 1.0681,
+      "step": 4121
+    },
+    {
+      "epoch": 0.3177613320999075,
+      "grad_norm": 4.053248405456543,
+      "learning_rate": 7.981785836021601e-06,
+      "loss": 0.9929,
+      "step": 4122
+    },
+    {
+      "epoch": 0.31783842121492445,
+      "grad_norm": 4.073676109313965,
+      "learning_rate": 7.980783590495089e-06,
+      "loss": 0.9481,
+      "step": 4123
+    },
+    {
+      "epoch": 0.3179155103299414,
+      "grad_norm": 4.069306373596191,
+      "learning_rate": 7.979781159132157e-06,
+      "loss": 1.1544,
+      "step": 4124
+    },
+    {
+      "epoch": 0.3179925994449584,
+      "grad_norm": 3.5793960094451904,
+      "learning_rate": 7.978778541995304e-06,
+      "loss": 1.0315,
+      "step": 4125
+    },
+    {
+      "epoch": 0.3180696885599753,
+      "grad_norm": 3.4428999423980713,
+      "learning_rate": 7.977775739147038e-06,
+      "loss": 0.9997,
+      "step": 4126
+    },
+    {
+      "epoch": 0.3181467776749923,
+      "grad_norm": 3.8016507625579834,
+      "learning_rate": 7.976772750649874e-06,
+      "loss": 0.9788,
+      "step": 4127
+    },
+    {
+      "epoch": 0.31822386679000925,
+      "grad_norm": 3.3368539810180664,
+      "learning_rate": 7.975769576566348e-06,
+      "loss": 0.9229,
+      "step": 4128
+    },
+    {
+      "epoch": 0.3183009559050262,
+      "grad_norm": 3.78816556930542,
+      "learning_rate": 7.974766216959001e-06,
+      "loss": 1.0265,
+      "step": 4129
+    },
+    {
+      "epoch": 0.3183780450200432,
+      "grad_norm": 3.6755471229553223,
+      "learning_rate": 7.973762671890387e-06,
+      "loss": 1.0043,
+      "step": 4130
+    },
+    {
+      "epoch": 0.3184551341350601,
+      "grad_norm": 3.3952746391296387,
+      "learning_rate": 7.972758941423071e-06,
+      "loss": 0.895,
+      "step": 4131
+    },
+    {
+      "epoch": 0.3185322232500771,
+      "grad_norm": 3.5413053035736084,
+      "learning_rate": 7.971755025619632e-06,
+      "loss": 0.9635,
+      "step": 4132
+    },
+    {
+      "epoch": 0.31860931236509404,
+      "grad_norm": 3.619885206222534,
+      "learning_rate": 7.970750924542659e-06,
+      "loss": 1.0574,
+      "step": 4133
+    },
+    {
+      "epoch": 0.318686401480111,
+      "grad_norm": 3.1986031532287598,
+      "learning_rate": 7.96974663825475e-06,
+      "loss": 0.9205,
+      "step": 4134
+    },
+    {
+      "epoch": 0.318763490595128,
+      "grad_norm": 3.69963002204895,
+      "learning_rate": 7.968742166818521e-06,
+      "loss": 1.0614,
+      "step": 4135
+    },
+    {
+      "epoch": 0.3188405797101449,
+      "grad_norm": 3.788604736328125,
+      "learning_rate": 7.967737510296591e-06,
+      "loss": 1.0176,
+      "step": 4136
+    },
+    {
+      "epoch": 0.31891766882516187,
+      "grad_norm": 3.8696343898773193,
+      "learning_rate": 7.9667326687516e-06,
+      "loss": 1.0235,
+      "step": 4137
+    },
+    {
+      "epoch": 0.31899475794017884,
+      "grad_norm": 3.4079203605651855,
+      "learning_rate": 7.965727642246191e-06,
+      "loss": 0.9388,
+      "step": 4138
+    },
+    {
+      "epoch": 0.3190718470551958,
+      "grad_norm": 4.057232856750488,
+      "learning_rate": 7.964722430843021e-06,
+      "loss": 1.059,
+      "step": 4139
+    },
+    {
+      "epoch": 0.3191489361702128,
+      "grad_norm": 3.7363650798797607,
+      "learning_rate": 7.963717034604765e-06,
+      "loss": 1.0545,
+      "step": 4140
+    },
+    {
+      "epoch": 0.31922602528522975,
+      "grad_norm": 3.5568385124206543,
+      "learning_rate": 7.962711453594101e-06,
+      "loss": 0.9876,
+      "step": 4141
+    },
+    {
+      "epoch": 0.31930311440024667,
+      "grad_norm": 3.632174491882324,
+      "learning_rate": 7.961705687873722e-06,
+      "loss": 0.9774,
+      "step": 4142
+    },
+    {
+      "epoch": 0.31938020351526364,
+      "grad_norm": 4.495326042175293,
+      "learning_rate": 7.960699737506333e-06,
+      "loss": 1.1305,
+      "step": 4143
+    },
+    {
+      "epoch": 0.3194572926302806,
+      "grad_norm": 3.538480281829834,
+      "learning_rate": 7.959693602554648e-06,
+      "loss": 1.0082,
+      "step": 4144
+    },
+    {
+      "epoch": 0.3195343817452976,
+      "grad_norm": 3.430492877960205,
+      "learning_rate": 7.958687283081394e-06,
+      "loss": 1.1023,
+      "step": 4145
+    },
+    {
+      "epoch": 0.31961147086031455,
+      "grad_norm": 3.7447872161865234,
+      "learning_rate": 7.957680779149315e-06,
+      "loss": 1.0244,
+      "step": 4146
+    },
+    {
+      "epoch": 0.31968855997533147,
+      "grad_norm": 3.6742663383483887,
+      "learning_rate": 7.956674090821156e-06,
+      "loss": 0.9341,
+      "step": 4147
+    },
+    {
+      "epoch": 0.31976564909034844,
+      "grad_norm": 4.139322280883789,
+      "learning_rate": 7.955667218159679e-06,
+      "loss": 0.9498,
+      "step": 4148
+    },
+    {
+      "epoch": 0.3198427382053654,
+      "grad_norm": 3.798419952392578,
+      "learning_rate": 7.95466016122766e-06,
+      "loss": 1.0443,
+      "step": 4149
+    },
+    {
+      "epoch": 0.3199198273203824,
+      "grad_norm": 3.6184287071228027,
+      "learning_rate": 7.953652920087884e-06,
+      "loss": 0.9629,
+      "step": 4150
+    },
+    {
+      "epoch": 0.31999691643539935,
+      "grad_norm": 3.602494478225708,
+      "learning_rate": 7.952645494803145e-06,
+      "loss": 1.0702,
+      "step": 4151
+    },
+    {
+      "epoch": 0.32007400555041626,
+      "grad_norm": 4.118215084075928,
+      "learning_rate": 7.95163788543625e-06,
+      "loss": 0.9996,
+      "step": 4152
+    },
+    {
+      "epoch": 0.32015109466543323,
+      "grad_norm": 3.864248037338257,
+      "learning_rate": 7.950630092050022e-06,
+      "loss": 1.0801,
+      "step": 4153
+    },
+    {
+      "epoch": 0.3202281837804502,
+      "grad_norm": 3.5450005531311035,
+      "learning_rate": 7.949622114707288e-06,
+      "loss": 0.9193,
+      "step": 4154
+    },
+    {
+      "epoch": 0.3203052728954672,
+      "grad_norm": 3.6530864238739014,
+      "learning_rate": 7.948613953470892e-06,
+      "loss": 0.8586,
+      "step": 4155
+    },
+    {
+      "epoch": 0.32038236201048415,
+      "grad_norm": 3.8235666751861572,
+      "learning_rate": 7.947605608403688e-06,
+      "loss": 0.9473,
+      "step": 4156
+    },
+    {
+      "epoch": 0.32045945112550106,
+      "grad_norm": 4.016197681427002,
+      "learning_rate": 7.946597079568538e-06,
+      "loss": 1.0146,
+      "step": 4157
+    },
+    {
+      "epoch": 0.32053654024051803,
+      "grad_norm": 3.6410627365112305,
+      "learning_rate": 7.945588367028324e-06,
+      "loss": 0.9481,
+      "step": 4158
+    },
+    {
+      "epoch": 0.320613629355535,
+      "grad_norm": 5.766059875488281,
+      "learning_rate": 7.94457947084593e-06,
+      "loss": 0.9786,
+      "step": 4159
+    },
+    {
+      "epoch": 0.32069071847055197,
+      "grad_norm": 3.8257815837860107,
+      "learning_rate": 7.943570391084254e-06,
+      "loss": 1.1402,
+      "step": 4160
+    },
+    {
+      "epoch": 0.32076780758556894,
+      "grad_norm": 4.60271692276001,
+      "learning_rate": 7.942561127806212e-06,
+      "loss": 1.0461,
+      "step": 4161
+    },
+    {
+      "epoch": 0.32084489670058586,
+      "grad_norm": 3.0530004501342773,
+      "learning_rate": 7.941551681074723e-06,
+      "loss": 0.8957,
+      "step": 4162
+    },
+    {
+      "epoch": 0.32092198581560283,
+      "grad_norm": 3.847745180130005,
+      "learning_rate": 7.94054205095272e-06,
+      "loss": 0.9406,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3209990749306198,
+      "grad_norm": 3.877429962158203,
+      "learning_rate": 7.939532237503151e-06,
+      "loss": 0.9846,
+      "step": 4164
+    },
+    {
+      "epoch": 0.32107616404563677,
+      "grad_norm": 3.6080520153045654,
+      "learning_rate": 7.93852224078897e-06,
+      "loss": 1.0752,
+      "step": 4165
+    },
+    {
+      "epoch": 0.32115325316065374,
+      "grad_norm": 3.4427030086517334,
+      "learning_rate": 7.937512060873149e-06,
+      "loss": 0.9899,
+      "step": 4166
+    },
+    {
+      "epoch": 0.32123034227567066,
+      "grad_norm": 3.60064697265625,
+      "learning_rate": 7.936501697818661e-06,
+      "loss": 0.9592,
+      "step": 4167
+    },
+    {
+      "epoch": 0.3213074313906876,
+      "grad_norm": 3.334470510482788,
+      "learning_rate": 7.935491151688504e-06,
+      "loss": 1.0138,
+      "step": 4168
+    },
+    {
+      "epoch": 0.3213845205057046,
+      "grad_norm": 3.6496033668518066,
+      "learning_rate": 7.934480422545674e-06,
+      "loss": 1.0665,
+      "step": 4169
+    },
+    {
+      "epoch": 0.32146160962072157,
+      "grad_norm": 3.650097131729126,
+      "learning_rate": 7.933469510453189e-06,
+      "loss": 1.085,
+      "step": 4170
+    },
+    {
+      "epoch": 0.32153869873573854,
+      "grad_norm": 3.6096320152282715,
+      "learning_rate": 7.932458415474073e-06,
+      "loss": 0.9372,
+      "step": 4171
+    },
+    {
+      "epoch": 0.32161578785075545,
+      "grad_norm": 4.109353065490723,
+      "learning_rate": 7.931447137671364e-06,
+      "loss": 1.0568,
+      "step": 4172
+    },
+    {
+      "epoch": 0.3216928769657724,
+      "grad_norm": 3.8872742652893066,
+      "learning_rate": 7.930435677108106e-06,
+      "loss": 1.0137,
+      "step": 4173
+    },
+    {
+      "epoch": 0.3217699660807894,
+      "grad_norm": 4.055763244628906,
+      "learning_rate": 7.929424033847362e-06,
+      "loss": 1.0044,
+      "step": 4174
+    },
+    {
+      "epoch": 0.32184705519580636,
+      "grad_norm": 3.5979840755462646,
+      "learning_rate": 7.9284122079522e-06,
+      "loss": 0.982,
+      "step": 4175
+    },
+    {
+      "epoch": 0.32192414431082333,
+      "grad_norm": 3.5764944553375244,
+      "learning_rate": 7.927400199485705e-06,
+      "loss": 0.9204,
+      "step": 4176
+    },
+    {
+      "epoch": 0.32200123342584025,
+      "grad_norm": 3.7426390647888184,
+      "learning_rate": 7.926388008510968e-06,
+      "loss": 0.9989,
+      "step": 4177
+    },
+    {
+      "epoch": 0.3220783225408572,
+      "grad_norm": 3.4947009086608887,
+      "learning_rate": 7.925375635091094e-06,
+      "loss": 0.959,
+      "step": 4178
+    },
+    {
+      "epoch": 0.3221554116558742,
+      "grad_norm": 4.316697597503662,
+      "learning_rate": 7.924363079289203e-06,
+      "loss": 1.0835,
+      "step": 4179
+    },
+    {
+      "epoch": 0.32223250077089116,
+      "grad_norm": 4.12495756149292,
+      "learning_rate": 7.923350341168416e-06,
+      "loss": 1.0846,
+      "step": 4180
+    },
+    {
+      "epoch": 0.32230958988590813,
+      "grad_norm": 3.6122055053710938,
+      "learning_rate": 7.922337420791879e-06,
+      "loss": 0.8974,
+      "step": 4181
+    },
+    {
+      "epoch": 0.32238667900092505,
+      "grad_norm": 3.608186960220337,
+      "learning_rate": 7.921324318222737e-06,
+      "loss": 1.066,
+      "step": 4182
+    },
+    {
+      "epoch": 0.322463768115942,
+      "grad_norm": 3.8409266471862793,
+      "learning_rate": 7.920311033524156e-06,
+      "loss": 1.0132,
+      "step": 4183
+    },
+    {
+      "epoch": 0.322540857230959,
+      "grad_norm": 4.176751136779785,
+      "learning_rate": 7.919297566759304e-06,
+      "loss": 1.0244,
+      "step": 4184
+    },
+    {
+      "epoch": 0.32261794634597596,
+      "grad_norm": 3.635542869567871,
+      "learning_rate": 7.918283917991367e-06,
+      "loss": 1.0494,
+      "step": 4185
+    },
+    {
+      "epoch": 0.32269503546099293,
+      "grad_norm": 3.756164073944092,
+      "learning_rate": 7.917270087283544e-06,
+      "loss": 1.0204,
+      "step": 4186
+    },
+    {
+      "epoch": 0.32277212457600984,
+      "grad_norm": 3.905221939086914,
+      "learning_rate": 7.91625607469904e-06,
+      "loss": 0.9397,
+      "step": 4187
+    },
+    {
+      "epoch": 0.3228492136910268,
+      "grad_norm": 3.6469991207122803,
+      "learning_rate": 7.915241880301075e-06,
+      "loss": 1.0839,
+      "step": 4188
+    },
+    {
+      "epoch": 0.3229263028060438,
+      "grad_norm": 3.7983455657958984,
+      "learning_rate": 7.914227504152874e-06,
+      "loss": 0.9868,
+      "step": 4189
+    },
+    {
+      "epoch": 0.32300339192106076,
+      "grad_norm": 3.3824074268341064,
+      "learning_rate": 7.913212946317684e-06,
+      "loss": 0.8911,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3230804810360777,
+      "grad_norm": 3.6762890815734863,
+      "learning_rate": 7.912198206858752e-06,
+      "loss": 1.0045,
+      "step": 4191
+    },
+    {
+      "epoch": 0.32315757015109464,
+      "grad_norm": 4.075868129730225,
+      "learning_rate": 7.911183285839347e-06,
+      "loss": 1.0257,
+      "step": 4192
+    },
+    {
+      "epoch": 0.3232346592661116,
+      "grad_norm": 3.5539023876190186,
+      "learning_rate": 7.91016818332274e-06,
+      "loss": 1.0092,
+      "step": 4193
+    },
+    {
+      "epoch": 0.3233117483811286,
+      "grad_norm": 3.818380355834961,
+      "learning_rate": 7.909152899372218e-06,
+      "loss": 1.097,
+      "step": 4194
+    },
+    {
+      "epoch": 0.32338883749614555,
+      "grad_norm": 3.38431453704834,
+      "learning_rate": 7.908137434051083e-06,
+      "loss": 0.896,
+      "step": 4195
+    },
+    {
+      "epoch": 0.3234659266111625,
+      "grad_norm": 3.587066888809204,
+      "learning_rate": 7.907121787422638e-06,
+      "loss": 0.9987,
+      "step": 4196
+    },
+    {
+      "epoch": 0.32354301572617944,
+      "grad_norm": 4.137657165527344,
+      "learning_rate": 7.906105959550206e-06,
+      "loss": 0.9355,
+      "step": 4197
+    },
+    {
+      "epoch": 0.3236201048411964,
+      "grad_norm": 3.3995325565338135,
+      "learning_rate": 7.90508995049712e-06,
+      "loss": 0.9642,
+      "step": 4198
+    },
+    {
+      "epoch": 0.3236971939562134,
+      "grad_norm": 3.7008519172668457,
+      "learning_rate": 7.90407376032672e-06,
+      "loss": 1.0123,
+      "step": 4199
+    },
+    {
+      "epoch": 0.32377428307123035,
+      "grad_norm": 3.6094677448272705,
+      "learning_rate": 7.903057389102361e-06,
+      "loss": 0.9955,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3238513721862473,
+      "grad_norm": 3.729969024658203,
+      "learning_rate": 7.902040836887413e-06,
+      "loss": 0.9154,
+      "step": 4201
+    },
+    {
+      "epoch": 0.32392846130126424,
+      "grad_norm": 3.7781825065612793,
+      "learning_rate": 7.901024103745244e-06,
+      "loss": 1.0113,
+      "step": 4202
+    },
+    {
+      "epoch": 0.3240055504162812,
+      "grad_norm": 3.7726991176605225,
+      "learning_rate": 7.90000718973925e-06,
+      "loss": 1.0299,
+      "step": 4203
+    },
+    {
+      "epoch": 0.3240826395312982,
+      "grad_norm": 3.447260618209839,
+      "learning_rate": 7.898990094932826e-06,
+      "loss": 1.0302,
+      "step": 4204
+    },
+    {
+      "epoch": 0.32415972864631515,
+      "grad_norm": 3.4710397720336914,
+      "learning_rate": 7.897972819389385e-06,
+      "loss": 0.8961,
+      "step": 4205
+    },
+    {
+      "epoch": 0.3242368177613321,
+      "grad_norm": 3.3661441802978516,
+      "learning_rate": 7.896955363172347e-06,
+      "loss": 0.9486,
+      "step": 4206
+    },
+    {
+      "epoch": 0.32431390687634903,
+      "grad_norm": 3.684475898742676,
+      "learning_rate": 7.895937726345145e-06,
+      "loss": 0.9692,
+      "step": 4207
+    },
+    {
+      "epoch": 0.324390995991366,
+      "grad_norm": 3.5266542434692383,
+      "learning_rate": 7.894919908971225e-06,
+      "loss": 1.0262,
+      "step": 4208
+    },
+    {
+      "epoch": 0.324468085106383,
+      "grad_norm": 3.6420974731445312,
+      "learning_rate": 7.893901911114041e-06,
+      "loss": 1.0388,
+      "step": 4209
+    },
+    {
+      "epoch": 0.32454517422139995,
+      "grad_norm": 3.7208940982818604,
+      "learning_rate": 7.892883732837062e-06,
+      "loss": 0.9965,
+      "step": 4210
+    },
+    {
+      "epoch": 0.3246222633364169,
+      "grad_norm": 3.839670419692993,
+      "learning_rate": 7.891865374203765e-06,
+      "loss": 1.0087,
+      "step": 4211
+    },
+    {
+      "epoch": 0.32469935245143383,
+      "grad_norm": 3.901395559310913,
+      "learning_rate": 7.890846835277638e-06,
+      "loss": 0.9884,
+      "step": 4212
+    },
+    {
+      "epoch": 0.3247764415664508,
+      "grad_norm": 3.6374828815460205,
+      "learning_rate": 7.889828116122183e-06,
+      "loss": 1.0152,
+      "step": 4213
+    },
+    {
+      "epoch": 0.3248535306814678,
+      "grad_norm": 4.1036763191223145,
+      "learning_rate": 7.888809216800913e-06,
+      "loss": 1.0062,
+      "step": 4214
+    },
+    {
+      "epoch": 0.32493061979648474,
+      "grad_norm": 3.640856981277466,
+      "learning_rate": 7.887790137377348e-06,
+      "loss": 0.9845,
+      "step": 4215
+    },
+    {
+      "epoch": 0.3250077089115017,
+      "grad_norm": 3.1671056747436523,
+      "learning_rate": 7.886770877915027e-06,
+      "loss": 0.892,
+      "step": 4216
+    },
+    {
+      "epoch": 0.32508479802651863,
+      "grad_norm": 3.5967583656311035,
+      "learning_rate": 7.885751438477489e-06,
+      "loss": 1.0032,
+      "step": 4217
+    },
+    {
+      "epoch": 0.3251618871415356,
+      "grad_norm": 3.60170316696167,
+      "learning_rate": 7.884731819128298e-06,
+      "loss": 0.9933,
+      "step": 4218
+    },
+    {
+      "epoch": 0.32523897625655257,
+      "grad_norm": 3.8272790908813477,
+      "learning_rate": 7.883712019931017e-06,
+      "loss": 1.0296,
+      "step": 4219
+    },
+    {
+      "epoch": 0.32531606537156954,
+      "grad_norm": 4.294167995452881,
+      "learning_rate": 7.882692040949226e-06,
+      "loss": 1.1175,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3253931544865865,
+      "grad_norm": 3.4372246265411377,
+      "learning_rate": 7.881671882246518e-06,
+      "loss": 1.0131,
+      "step": 4221
+    },
+    {
+      "epoch": 0.3254702436016034,
+      "grad_norm": 4.079759120941162,
+      "learning_rate": 7.880651543886491e-06,
+      "loss": 0.9886,
+      "step": 4222
+    },
+    {
+      "epoch": 0.3255473327166204,
+      "grad_norm": 5.348727226257324,
+      "learning_rate": 7.87963102593276e-06,
+      "loss": 0.9807,
+      "step": 4223
+    },
+    {
+      "epoch": 0.32562442183163737,
+      "grad_norm": 3.5309953689575195,
+      "learning_rate": 7.878610328448948e-06,
+      "loss": 0.9316,
+      "step": 4224
+    },
+    {
+      "epoch": 0.32570151094665434,
+      "grad_norm": 3.289616346359253,
+      "learning_rate": 7.877589451498692e-06,
+      "loss": 0.9425,
+      "step": 4225
+    },
+    {
+      "epoch": 0.3257786000616713,
+      "grad_norm": 3.5728421211242676,
+      "learning_rate": 7.876568395145636e-06,
+      "loss": 0.8591,
+      "step": 4226
+    },
+    {
+      "epoch": 0.3258556891766883,
+      "grad_norm": 3.5462687015533447,
+      "learning_rate": 7.87554715945344e-06,
+      "loss": 0.875,
+      "step": 4227
+    },
+    {
+      "epoch": 0.3259327782917052,
+      "grad_norm": 3.9166672229766846,
+      "learning_rate": 7.87452574448577e-06,
+      "loss": 1.0411,
+      "step": 4228
+    },
+    {
+      "epoch": 0.32600986740672216,
+      "grad_norm": 3.8413209915161133,
+      "learning_rate": 7.873504150306308e-06,
+      "loss": 0.9692,
+      "step": 4229
+    },
+    {
+      "epoch": 0.32608695652173914,
+      "grad_norm": 4.090485095977783,
+      "learning_rate": 7.872482376978746e-06,
+      "loss": 1.0688,
+      "step": 4230
+    },
+    {
+      "epoch": 0.3261640456367561,
+      "grad_norm": 3.474355697631836,
+      "learning_rate": 7.87146042456678e-06,
+      "loss": 1.0016,
+      "step": 4231
+    },
+    {
+      "epoch": 0.3262411347517731,
+      "grad_norm": 3.8390350341796875,
+      "learning_rate": 7.870438293134133e-06,
+      "loss": 1.0577,
+      "step": 4232
+    },
+    {
+      "epoch": 0.32631822386679,
+      "grad_norm": 4.005039691925049,
+      "learning_rate": 7.869415982744524e-06,
+      "loss": 0.8633,
+      "step": 4233
+    },
+    {
+      "epoch": 0.32639531298180696,
+      "grad_norm": 3.2557785511016846,
+      "learning_rate": 7.868393493461688e-06,
+      "loss": 0.8518,
+      "step": 4234
+    },
+    {
+      "epoch": 0.32647240209682393,
+      "grad_norm": 3.4779703617095947,
+      "learning_rate": 7.867370825349375e-06,
+      "loss": 1.043,
+      "step": 4235
+    },
+    {
+      "epoch": 0.3265494912118409,
+      "grad_norm": 3.7263035774230957,
+      "learning_rate": 7.86634797847134e-06,
+      "loss": 1.0086,
+      "step": 4236
+    },
+    {
+      "epoch": 0.3266265803268579,
+      "grad_norm": 3.4993364810943604,
+      "learning_rate": 7.865324952891354e-06,
+      "loss": 0.9986,
+      "step": 4237
+    },
+    {
+      "epoch": 0.3267036694418748,
+      "grad_norm": 31.114559173583984,
+      "learning_rate": 7.864301748673197e-06,
+      "loss": 1.041,
+      "step": 4238
+    },
+    {
+      "epoch": 0.32678075855689176,
+      "grad_norm": 3.9260265827178955,
+      "learning_rate": 7.863278365880662e-06,
+      "loss": 1.1013,
+      "step": 4239
+    },
+    {
+      "epoch": 0.32685784767190873,
+      "grad_norm": 3.4418816566467285,
+      "learning_rate": 7.862254804577549e-06,
+      "loss": 0.9764,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3269349367869257,
+      "grad_norm": 3.3301095962524414,
+      "learning_rate": 7.861231064827673e-06,
+      "loss": 1.0296,
+      "step": 4241
+    },
+    {
+      "epoch": 0.32701202590194267,
+      "grad_norm": 3.416653633117676,
+      "learning_rate": 7.86020714669486e-06,
+      "loss": 1.0018,
+      "step": 4242
+    },
+    {
+      "epoch": 0.3270891150169596,
+      "grad_norm": 3.7522192001342773,
+      "learning_rate": 7.859183050242945e-06,
+      "loss": 0.8415,
+      "step": 4243
+    },
+    {
+      "epoch": 0.32716620413197656,
+      "grad_norm": 3.878190279006958,
+      "learning_rate": 7.858158775535773e-06,
+      "loss": 1.0617,
+      "step": 4244
+    },
+    {
+      "epoch": 0.3272432932469935,
+      "grad_norm": 3.7095284461975098,
+      "learning_rate": 7.857134322637205e-06,
+      "loss": 1.0556,
+      "step": 4245
+    },
+    {
+      "epoch": 0.3273203823620105,
+      "grad_norm": 3.8554065227508545,
+      "learning_rate": 7.85610969161111e-06,
+      "loss": 1.0377,
+      "step": 4246
+    },
+    {
+      "epoch": 0.32739747147702747,
+      "grad_norm": 3.8045499324798584,
+      "learning_rate": 7.855084882521366e-06,
+      "loss": 1.073,
+      "step": 4247
+    },
+    {
+      "epoch": 0.3274745605920444,
+      "grad_norm": 4.000002384185791,
+      "learning_rate": 7.854059895431869e-06,
+      "loss": 1.0777,
+      "step": 4248
+    },
+    {
+      "epoch": 0.32755164970706135,
+      "grad_norm": 4.595569610595703,
+      "learning_rate": 7.853034730406516e-06,
+      "loss": 0.9705,
+      "step": 4249
+    },
+    {
+      "epoch": 0.3276287388220783,
+      "grad_norm": 3.743229627609253,
+      "learning_rate": 7.852009387509227e-06,
+      "loss": 1.0785,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3277058279370953,
+      "grad_norm": 3.717169761657715,
+      "learning_rate": 7.850983866803923e-06,
+      "loss": 1.0172,
+      "step": 4251
+    },
+    {
+      "epoch": 0.32778291705211227,
+      "grad_norm": 3.679903984069824,
+      "learning_rate": 7.849958168354538e-06,
+      "loss": 0.9691,
+      "step": 4252
+    },
+    {
+      "epoch": 0.3278600061671292,
+      "grad_norm": 3.682389497756958,
+      "learning_rate": 7.848932292225025e-06,
+      "loss": 1.0234,
+      "step": 4253
+    },
+    {
+      "epoch": 0.32793709528214615,
+      "grad_norm": 4.109451770782471,
+      "learning_rate": 7.847906238479337e-06,
+      "loss": 1.0486,
+      "step": 4254
+    },
+    {
+      "epoch": 0.3280141843971631,
+      "grad_norm": 4.467040061950684,
+      "learning_rate": 7.846880007181443e-06,
+      "loss": 1.0206,
+      "step": 4255
+    },
+    {
+      "epoch": 0.3280912735121801,
+      "grad_norm": 3.9203736782073975,
+      "learning_rate": 7.845853598395327e-06,
+      "loss": 0.9341,
+      "step": 4256
+    },
+    {
+      "epoch": 0.32816836262719706,
+      "grad_norm": 3.5691542625427246,
+      "learning_rate": 7.844827012184978e-06,
+      "loss": 0.9578,
+      "step": 4257
+    },
+    {
+      "epoch": 0.328245451742214,
+      "grad_norm": 3.7378454208374023,
+      "learning_rate": 7.843800248614396e-06,
+      "loss": 1.04,
+      "step": 4258
+    },
+    {
+      "epoch": 0.32832254085723095,
+      "grad_norm": 3.8135650157928467,
+      "learning_rate": 7.8427733077476e-06,
+      "loss": 1.021,
+      "step": 4259
+    },
+    {
+      "epoch": 0.3283996299722479,
+      "grad_norm": 3.978492021560669,
+      "learning_rate": 7.84174618964861e-06,
+      "loss": 1.0348,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3284767190872649,
+      "grad_norm": 3.6482443809509277,
+      "learning_rate": 7.840718894381464e-06,
+      "loss": 0.9996,
+      "step": 4261
+    },
+    {
+      "epoch": 0.32855380820228186,
+      "grad_norm": 3.369802474975586,
+      "learning_rate": 7.839691422010208e-06,
+      "loss": 0.9249,
+      "step": 4262
+    },
+    {
+      "epoch": 0.3286308973172988,
+      "grad_norm": 3.726884603500366,
+      "learning_rate": 7.838663772598897e-06,
+      "loss": 0.9568,
+      "step": 4263
+    },
+    {
+      "epoch": 0.32870798643231575,
+      "grad_norm": 3.5390725135803223,
+      "learning_rate": 7.837635946211603e-06,
+      "loss": 1.0817,
+      "step": 4264
+    },
+    {
+      "epoch": 0.3287850755473327,
+      "grad_norm": 4.038122177124023,
+      "learning_rate": 7.836607942912403e-06,
+      "loss": 1.0059,
+      "step": 4265
+    },
+    {
+      "epoch": 0.3288621646623497,
+      "grad_norm": 3.1766395568847656,
+      "learning_rate": 7.83557976276539e-06,
+      "loss": 0.8842,
+      "step": 4266
+    },
+    {
+      "epoch": 0.32893925377736666,
+      "grad_norm": 3.975059986114502,
+      "learning_rate": 7.834551405834665e-06,
+      "loss": 1.1503,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3290163428923836,
+      "grad_norm": 4.144785404205322,
+      "learning_rate": 7.833522872184338e-06,
+      "loss": 0.974,
+      "step": 4268
+    },
+    {
+      "epoch": 0.32909343200740054,
+      "grad_norm": 3.740612268447876,
+      "learning_rate": 7.832494161878537e-06,
+      "loss": 0.956,
+      "step": 4269
+    },
+    {
+      "epoch": 0.3291705211224175,
+      "grad_norm": 3.4083251953125,
+      "learning_rate": 7.831465274981395e-06,
+      "loss": 0.9164,
+      "step": 4270
+    },
+    {
+      "epoch": 0.3292476102374345,
+      "grad_norm": 3.277799367904663,
+      "learning_rate": 7.830436211557057e-06,
+      "loss": 0.9708,
+      "step": 4271
+    },
+    {
+      "epoch": 0.32932469935245146,
+      "grad_norm": 3.846964120864868,
+      "learning_rate": 7.82940697166968e-06,
+      "loss": 0.9867,
+      "step": 4272
+    },
+    {
+      "epoch": 0.32940178846746837,
+      "grad_norm": 3.877805233001709,
+      "learning_rate": 7.828377555383433e-06,
+      "loss": 1.0738,
+      "step": 4273
+    },
+    {
+      "epoch": 0.32947887758248534,
+      "grad_norm": 3.754934310913086,
+      "learning_rate": 7.827347962762495e-06,
+      "loss": 1.0322,
+      "step": 4274
+    },
+    {
+      "epoch": 0.3295559666975023,
+      "grad_norm": 4.1259284019470215,
+      "learning_rate": 7.826318193871052e-06,
+      "loss": 1.0407,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3296330558125193,
+      "grad_norm": 4.191363334655762,
+      "learning_rate": 7.825288248773309e-06,
+      "loss": 1.0295,
+      "step": 4276
+    },
+    {
+      "epoch": 0.32971014492753625,
+      "grad_norm": 3.45347261428833,
+      "learning_rate": 7.824258127533477e-06,
+      "loss": 0.9255,
+      "step": 4277
+    },
+    {
+      "epoch": 0.32978723404255317,
+      "grad_norm": 3.499629497528076,
+      "learning_rate": 7.823227830215776e-06,
+      "loss": 1.0538,
+      "step": 4278
+    },
+    {
+      "epoch": 0.32986432315757014,
+      "grad_norm": 4.018231391906738,
+      "learning_rate": 7.822197356884442e-06,
+      "loss": 1.0472,
+      "step": 4279
+    },
+    {
+      "epoch": 0.3299414122725871,
+      "grad_norm": 3.717026472091675,
+      "learning_rate": 7.82116670760372e-06,
+      "loss": 0.9186,
+      "step": 4280
+    },
+    {
+      "epoch": 0.3300185013876041,
+      "grad_norm": 4.538501262664795,
+      "learning_rate": 7.820135882437866e-06,
+      "loss": 1.1079,
+      "step": 4281
+    },
+    {
+      "epoch": 0.33009559050262105,
+      "grad_norm": 3.6810877323150635,
+      "learning_rate": 7.819104881451145e-06,
+      "loss": 0.9466,
+      "step": 4282
+    },
+    {
+      "epoch": 0.33017267961763797,
+      "grad_norm": 3.926638603210449,
+      "learning_rate": 7.818073704707834e-06,
+      "loss": 0.9839,
+      "step": 4283
+    },
+    {
+      "epoch": 0.33024976873265494,
+      "grad_norm": 3.823995590209961,
+      "learning_rate": 7.817042352272224e-06,
+      "loss": 1.0753,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3303268578476719,
+      "grad_norm": 3.6535205841064453,
+      "learning_rate": 7.816010824208613e-06,
+      "loss": 1.0285,
+      "step": 4285
+    },
+    {
+      "epoch": 0.3304039469626889,
+      "grad_norm": 3.4686245918273926,
+      "learning_rate": 7.814979120581311e-06,
+      "loss": 0.9721,
+      "step": 4286
+    },
+    {
+      "epoch": 0.33048103607770585,
+      "grad_norm": 3.8238701820373535,
+      "learning_rate": 7.81394724145464e-06,
+      "loss": 1.0513,
+      "step": 4287
+    },
+    {
+      "epoch": 0.33055812519272276,
+      "grad_norm": 3.5260727405548096,
+      "learning_rate": 7.812915186892933e-06,
+      "loss": 0.9242,
+      "step": 4288
+    },
+    {
+      "epoch": 0.33063521430773973,
+      "grad_norm": 3.374612331390381,
+      "learning_rate": 7.811882956960532e-06,
+      "loss": 0.984,
+      "step": 4289
+    },
+    {
+      "epoch": 0.3307123034227567,
+      "grad_norm": 3.7520065307617188,
+      "learning_rate": 7.810850551721793e-06,
+      "loss": 0.9958,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3307893925377737,
+      "grad_norm": 3.351409435272217,
+      "learning_rate": 7.809817971241079e-06,
+      "loss": 0.9151,
+      "step": 4291
+    },
+    {
+      "epoch": 0.33086648165279064,
+      "grad_norm": 3.3581125736236572,
+      "learning_rate": 7.808785215582766e-06,
+      "loss": 0.9481,
+      "step": 4292
+    },
+    {
+      "epoch": 0.33094357076780756,
+      "grad_norm": 3.93084716796875,
+      "learning_rate": 7.807752284811243e-06,
+      "loss": 1.1273,
+      "step": 4293
+    },
+    {
+      "epoch": 0.33102065988282453,
+      "grad_norm": 4.633861064910889,
+      "learning_rate": 7.806719178990906e-06,
+      "loss": 1.0091,
+      "step": 4294
+    },
+    {
+      "epoch": 0.3310977489978415,
+      "grad_norm": 3.382524013519287,
+      "learning_rate": 7.805685898186164e-06,
+      "loss": 0.958,
+      "step": 4295
+    },
+    {
+      "epoch": 0.33117483811285847,
+      "grad_norm": 3.2215330600738525,
+      "learning_rate": 7.804652442461438e-06,
+      "loss": 0.9924,
+      "step": 4296
+    },
+    {
+      "epoch": 0.33125192722787544,
+      "grad_norm": 3.757098913192749,
+      "learning_rate": 7.80361881188116e-06,
+      "loss": 1.0064,
+      "step": 4297
+    },
+    {
+      "epoch": 0.33132901634289236,
+      "grad_norm": 3.6601150035858154,
+      "learning_rate": 7.802585006509766e-06,
+      "loss": 0.9142,
+      "step": 4298
+    },
+    {
+      "epoch": 0.33140610545790933,
+      "grad_norm": 4.117657661437988,
+      "learning_rate": 7.801551026411715e-06,
+      "loss": 1.1443,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3314831945729263,
+      "grad_norm": 3.925748109817505,
+      "learning_rate": 7.800516871651465e-06,
+      "loss": 0.9226,
+      "step": 4300
+    },
+    {
+      "epoch": 0.33156028368794327,
+      "grad_norm": 3.6940126419067383,
+      "learning_rate": 7.799482542293491e-06,
+      "loss": 1.1027,
+      "step": 4301
+    },
+    {
+      "epoch": 0.33163737280296024,
+      "grad_norm": 3.6253130435943604,
+      "learning_rate": 7.798448038402283e-06,
+      "loss": 0.9479,
+      "step": 4302
+    },
+    {
+      "epoch": 0.33171446191797715,
+      "grad_norm": 3.360285758972168,
+      "learning_rate": 7.79741336004233e-06,
+      "loss": 0.9113,
+      "step": 4303
+    },
+    {
+      "epoch": 0.3317915510329941,
+      "grad_norm": 3.5105605125427246,
+      "learning_rate": 7.796378507278144e-06,
+      "loss": 0.9038,
+      "step": 4304
+    },
+    {
+      "epoch": 0.3318686401480111,
+      "grad_norm": 4.047402858734131,
+      "learning_rate": 7.79534348017424e-06,
+      "loss": 0.9321,
+      "step": 4305
+    },
+    {
+      "epoch": 0.33194572926302807,
+      "grad_norm": 3.2786002159118652,
+      "learning_rate": 7.794308278795148e-06,
+      "loss": 0.9316,
+      "step": 4306
+    },
+    {
+      "epoch": 0.33202281837804504,
+      "grad_norm": 3.5382752418518066,
+      "learning_rate": 7.793272903205406e-06,
+      "loss": 1.0196,
+      "step": 4307
+    },
+    {
+      "epoch": 0.33209990749306195,
+      "grad_norm": 3.999952793121338,
+      "learning_rate": 7.792237353469567e-06,
+      "loss": 0.9702,
+      "step": 4308
+    },
+    {
+      "epoch": 0.3321769966080789,
+      "grad_norm": 3.6367757320404053,
+      "learning_rate": 7.791201629652189e-06,
+      "loss": 0.9877,
+      "step": 4309
+    },
+    {
+      "epoch": 0.3322540857230959,
+      "grad_norm": 5.327296733856201,
+      "learning_rate": 7.790165731817847e-06,
+      "loss": 0.8918,
+      "step": 4310
+    },
+    {
+      "epoch": 0.33233117483811286,
+      "grad_norm": 3.5317442417144775,
+      "learning_rate": 7.789129660031121e-06,
+      "loss": 0.9834,
+      "step": 4311
+    },
+    {
+      "epoch": 0.33240826395312983,
+      "grad_norm": 3.6593918800354004,
+      "learning_rate": 7.788093414356605e-06,
+      "loss": 1.0577,
+      "step": 4312
+    },
+    {
+      "epoch": 0.3324853530681468,
+      "grad_norm": 3.4909262657165527,
+      "learning_rate": 7.787056994858906e-06,
+      "loss": 0.9607,
+      "step": 4313
+    },
+    {
+      "epoch": 0.3325624421831637,
+      "grad_norm": 3.5621049404144287,
+      "learning_rate": 7.786020401602638e-06,
+      "loss": 0.9476,
+      "step": 4314
+    },
+    {
+      "epoch": 0.3326395312981807,
+      "grad_norm": 3.4073727130889893,
+      "learning_rate": 7.784983634652425e-06,
+      "loss": 1.0184,
+      "step": 4315
+    },
+    {
+      "epoch": 0.33271662041319766,
+      "grad_norm": 4.091675281524658,
+      "learning_rate": 7.783946694072908e-06,
+      "loss": 1.0435,
+      "step": 4316
+    },
+    {
+      "epoch": 0.33279370952821463,
+      "grad_norm": 3.9657843112945557,
+      "learning_rate": 7.782909579928733e-06,
+      "loss": 1.0276,
+      "step": 4317
+    },
+    {
+      "epoch": 0.3328707986432316,
+      "grad_norm": 3.8811726570129395,
+      "learning_rate": 7.78187229228456e-06,
+      "loss": 1.022,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3329478877582485,
+      "grad_norm": 4.559733867645264,
+      "learning_rate": 7.780834831205056e-06,
+      "loss": 1.0362,
+      "step": 4319
+    },
+    {
+      "epoch": 0.3330249768732655,
+      "grad_norm": 3.447314500808716,
+      "learning_rate": 7.779797196754901e-06,
+      "loss": 0.9834,
+      "step": 4320
+    },
+    {
+      "epoch": 0.33310206598828246,
+      "grad_norm": 3.546065330505371,
+      "learning_rate": 7.77875938899879e-06,
+      "loss": 0.9406,
+      "step": 4321
+    },
+    {
+      "epoch": 0.33317915510329943,
+      "grad_norm": 3.527249336242676,
+      "learning_rate": 7.777721408001421e-06,
+      "loss": 1.0047,
+      "step": 4322
+    },
+    {
+      "epoch": 0.3332562442183164,
+      "grad_norm": 3.692897319793701,
+      "learning_rate": 7.77668325382751e-06,
+      "loss": 1.1071,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 3.6986513137817383,
+      "learning_rate": 7.775644926541778e-06,
+      "loss": 0.9281,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3334104224483503,
+      "grad_norm": 3.8168718814849854,
+      "learning_rate": 7.77460642620896e-06,
+      "loss": 1.0708,
+      "step": 4325
+    },
+    {
+      "epoch": 0.33348751156336726,
+      "grad_norm": 3.7896907329559326,
+      "learning_rate": 7.773567752893803e-06,
+      "loss": 0.9714,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3335646006783842,
+      "grad_norm": 3.77909255027771,
+      "learning_rate": 7.772528906661059e-06,
+      "loss": 0.9923,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3336416897934012,
+      "grad_norm": 3.9389283657073975,
+      "learning_rate": 7.771489887575498e-06,
+      "loss": 1.0688,
+      "step": 4328
+    },
+    {
+      "epoch": 0.3337187789084181,
+      "grad_norm": 3.601837635040283,
+      "learning_rate": 7.770450695701896e-06,
+      "loss": 1.0515,
+      "step": 4329
+    },
+    {
+      "epoch": 0.3337958680234351,
+      "grad_norm": 4.004146099090576,
+      "learning_rate": 7.769411331105044e-06,
+      "loss": 1.0052,
+      "step": 4330
+    },
+    {
+      "epoch": 0.33387295713845205,
+      "grad_norm": 3.7516090869903564,
+      "learning_rate": 7.768371793849736e-06,
+      "loss": 1.0585,
+      "step": 4331
+    },
+    {
+      "epoch": 0.333950046253469,
+      "grad_norm": 3.6058356761932373,
+      "learning_rate": 7.767332084000784e-06,
+      "loss": 1.0234,
+      "step": 4332
+    },
+    {
+      "epoch": 0.334027135368486,
+      "grad_norm": 3.795377016067505,
+      "learning_rate": 7.766292201623012e-06,
+      "loss": 1.0903,
+      "step": 4333
+    },
+    {
+      "epoch": 0.3341042244835029,
+      "grad_norm": 3.492107391357422,
+      "learning_rate": 7.765252146781245e-06,
+      "loss": 0.8837,
+      "step": 4334
+    },
+    {
+      "epoch": 0.3341813135985199,
+      "grad_norm": 3.525789976119995,
+      "learning_rate": 7.76421191954033e-06,
+      "loss": 0.8463,
+      "step": 4335
+    },
+    {
+      "epoch": 0.33425840271353685,
+      "grad_norm": 4.0740509033203125,
+      "learning_rate": 7.763171519965118e-06,
+      "loss": 1.152,
+      "step": 4336
+    },
+    {
+      "epoch": 0.3343354918285538,
+      "grad_norm": 3.6718363761901855,
+      "learning_rate": 7.762130948120472e-06,
+      "loss": 0.9143,
+      "step": 4337
+    },
+    {
+      "epoch": 0.3344125809435708,
+      "grad_norm": 3.7916324138641357,
+      "learning_rate": 7.761090204071267e-06,
+      "loss": 0.9976,
+      "step": 4338
+    },
+    {
+      "epoch": 0.3344896700585877,
+      "grad_norm": 3.5439155101776123,
+      "learning_rate": 7.760049287882388e-06,
+      "loss": 0.9737,
+      "step": 4339
+    },
+    {
+      "epoch": 0.3345667591736047,
+      "grad_norm": 3.6218690872192383,
+      "learning_rate": 7.75900819961873e-06,
+      "loss": 0.9956,
+      "step": 4340
+    },
+    {
+      "epoch": 0.33464384828862165,
+      "grad_norm": 3.703711986541748,
+      "learning_rate": 7.757966939345201e-06,
+      "loss": 0.854,
+      "step": 4341
+    },
+    {
+      "epoch": 0.3347209374036386,
+      "grad_norm": 3.890967845916748,
+      "learning_rate": 7.756925507126717e-06,
+      "loss": 1.0805,
+      "step": 4342
+    },
+    {
+      "epoch": 0.3347980265186556,
+      "grad_norm": 3.6906652450561523,
+      "learning_rate": 7.755883903028205e-06,
+      "loss": 1.0733,
+      "step": 4343
+    },
+    {
+      "epoch": 0.3348751156336725,
+      "grad_norm": 3.4065797328948975,
+      "learning_rate": 7.754842127114606e-06,
+      "loss": 0.9758,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3349522047486895,
+      "grad_norm": 3.7064595222473145,
+      "learning_rate": 7.753800179450867e-06,
+      "loss": 1.0737,
+      "step": 4345
+    },
+    {
+      "epoch": 0.33502929386370645,
+      "grad_norm": 3.8116419315338135,
+      "learning_rate": 7.752758060101951e-06,
+      "loss": 1.0783,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3351063829787234,
+      "grad_norm": 3.957881450653076,
+      "learning_rate": 7.751715769132823e-06,
+      "loss": 1.0667,
+      "step": 4347
+    },
+    {
+      "epoch": 0.3351834720937404,
+      "grad_norm": 3.905587911605835,
+      "learning_rate": 7.750673306608473e-06,
+      "loss": 1.0612,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3352605612087573,
+      "grad_norm": 4.058816432952881,
+      "learning_rate": 7.749630672593886e-06,
+      "loss": 1.2289,
+      "step": 4349
+    },
+    {
+      "epoch": 0.3353376503237743,
+      "grad_norm": 3.449509382247925,
+      "learning_rate": 7.748587867154068e-06,
+      "loss": 0.9291,
+      "step": 4350
+    },
+    {
+      "epoch": 0.33541473943879124,
+      "grad_norm": 3.368961811065674,
+      "learning_rate": 7.747544890354031e-06,
+      "loss": 0.9528,
+      "step": 4351
+    },
+    {
+      "epoch": 0.3354918285538082,
+      "grad_norm": 3.8779165744781494,
+      "learning_rate": 7.7465017422588e-06,
+      "loss": 1.075,
+      "step": 4352
+    },
+    {
+      "epoch": 0.3355689176688252,
+      "grad_norm": 3.9211997985839844,
+      "learning_rate": 7.745458422933409e-06,
+      "loss": 0.9959,
+      "step": 4353
+    },
+    {
+      "epoch": 0.3356460067838421,
+      "grad_norm": 3.9596736431121826,
+      "learning_rate": 7.744414932442908e-06,
+      "loss": 1.0908,
+      "step": 4354
+    },
+    {
+      "epoch": 0.33572309589885907,
+      "grad_norm": 4.090301036834717,
+      "learning_rate": 7.743371270852346e-06,
+      "loss": 1.2187,
+      "step": 4355
+    },
+    {
+      "epoch": 0.33580018501387604,
+      "grad_norm": 3.7745120525360107,
+      "learning_rate": 7.742327438226796e-06,
+      "loss": 0.9202,
+      "step": 4356
+    },
+    {
+      "epoch": 0.335877274128893,
+      "grad_norm": 3.6797232627868652,
+      "learning_rate": 7.74128343463133e-06,
+      "loss": 0.8468,
+      "step": 4357
+    },
+    {
+      "epoch": 0.33595436324391,
+      "grad_norm": 3.7442944049835205,
+      "learning_rate": 7.740239260131042e-06,
+      "loss": 0.974,
+      "step": 4358
+    },
+    {
+      "epoch": 0.3360314523589269,
+      "grad_norm": 3.405045509338379,
+      "learning_rate": 7.739194914791028e-06,
+      "loss": 0.9661,
+      "step": 4359
+    },
+    {
+      "epoch": 0.33610854147394387,
+      "grad_norm": 3.6131653785705566,
+      "learning_rate": 7.738150398676397e-06,
+      "loss": 1.014,
+      "step": 4360
+    },
+    {
+      "epoch": 0.33618563058896084,
+      "grad_norm": 4.105853080749512,
+      "learning_rate": 7.73710571185227e-06,
+      "loss": 1.2013,
+      "step": 4361
+    },
+    {
+      "epoch": 0.3362627197039778,
+      "grad_norm": 3.5816614627838135,
+      "learning_rate": 7.736060854383778e-06,
+      "loss": 0.8909,
+      "step": 4362
+    },
+    {
+      "epoch": 0.3363398088189948,
+      "grad_norm": 3.8828015327453613,
+      "learning_rate": 7.735015826336064e-06,
+      "loss": 1.0176,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3364168979340117,
+      "grad_norm": 3.9706432819366455,
+      "learning_rate": 7.733970627774275e-06,
+      "loss": 1.0422,
+      "step": 4364
+    },
+    {
+      "epoch": 0.33649398704902866,
+      "grad_norm": 4.061374664306641,
+      "learning_rate": 7.732925258763577e-06,
+      "loss": 0.9859,
+      "step": 4365
+    },
+    {
+      "epoch": 0.33657107616404563,
+      "grad_norm": 3.8068432807922363,
+      "learning_rate": 7.731879719369145e-06,
+      "loss": 1.1827,
+      "step": 4366
+    },
+    {
+      "epoch": 0.3366481652790626,
+      "grad_norm": 3.7083537578582764,
+      "learning_rate": 7.73083400965616e-06,
+      "loss": 0.994,
+      "step": 4367
+    },
+    {
+      "epoch": 0.3367252543940796,
+      "grad_norm": 3.8531298637390137,
+      "learning_rate": 7.72978812968982e-06,
+      "loss": 0.8946,
+      "step": 4368
+    },
+    {
+      "epoch": 0.3368023435090965,
+      "grad_norm": 3.4262845516204834,
+      "learning_rate": 7.728742079535326e-06,
+      "loss": 0.9598,
+      "step": 4369
+    },
+    {
+      "epoch": 0.33687943262411346,
+      "grad_norm": 3.567720413208008,
+      "learning_rate": 7.727695859257896e-06,
+      "loss": 0.9415,
+      "step": 4370
+    },
+    {
+      "epoch": 0.33695652173913043,
+      "grad_norm": 3.5674967765808105,
+      "learning_rate": 7.726649468922756e-06,
+      "loss": 0.9562,
+      "step": 4371
+    },
+    {
+      "epoch": 0.3370336108541474,
+      "grad_norm": 4.217917442321777,
+      "learning_rate": 7.725602908595144e-06,
+      "loss": 1.0205,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3371106999691644,
+      "grad_norm": 3.5196731090545654,
+      "learning_rate": 7.724556178340307e-06,
+      "loss": 1.0654,
+      "step": 4373
+    },
+    {
+      "epoch": 0.3371877890841813,
+      "grad_norm": 4.0760369300842285,
+      "learning_rate": 7.723509278223504e-06,
+      "loss": 0.9772,
+      "step": 4374
+    },
+    {
+      "epoch": 0.33726487819919826,
+      "grad_norm": 3.8422629833221436,
+      "learning_rate": 7.722462208310002e-06,
+      "loss": 0.9518,
+      "step": 4375
+    },
+    {
+      "epoch": 0.33734196731421523,
+      "grad_norm": 3.41591739654541,
+      "learning_rate": 7.721414968665081e-06,
+      "loss": 1.0063,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3374190564292322,
+      "grad_norm": 3.6463801860809326,
+      "learning_rate": 7.720367559354032e-06,
+      "loss": 1.0955,
+      "step": 4377
+    },
+    {
+      "epoch": 0.33749614554424917,
+      "grad_norm": 4.359663963317871,
+      "learning_rate": 7.719319980442154e-06,
+      "loss": 0.9804,
+      "step": 4378
+    },
+    {
+      "epoch": 0.3375732346592661,
+      "grad_norm": 3.783148765563965,
+      "learning_rate": 7.718272231994759e-06,
+      "loss": 0.9911,
+      "step": 4379
+    },
+    {
+      "epoch": 0.33765032377428306,
+      "grad_norm": 4.130781173706055,
+      "learning_rate": 7.71722431407717e-06,
+      "loss": 1.0291,
+      "step": 4380
+    },
+    {
+      "epoch": 0.3377274128893,
+      "grad_norm": 3.954774856567383,
+      "learning_rate": 7.716176226754716e-06,
+      "loss": 0.9907,
+      "step": 4381
+    },
+    {
+      "epoch": 0.337804502004317,
+      "grad_norm": 3.640958547592163,
+      "learning_rate": 7.715127970092742e-06,
+      "loss": 0.9913,
+      "step": 4382
+    },
+    {
+      "epoch": 0.33788159111933397,
+      "grad_norm": 3.703672170639038,
+      "learning_rate": 7.714079544156602e-06,
+      "loss": 0.9763,
+      "step": 4383
+    },
+    {
+      "epoch": 0.3379586802343509,
+      "grad_norm": 3.6405487060546875,
+      "learning_rate": 7.71303094901166e-06,
+      "loss": 0.8953,
+      "step": 4384
+    },
+    {
+      "epoch": 0.33803576934936785,
+      "grad_norm": 3.752634286880493,
+      "learning_rate": 7.711982184723287e-06,
+      "loss": 0.8807,
+      "step": 4385
+    },
+    {
+      "epoch": 0.3381128584643848,
+      "grad_norm": 3.402266025543213,
+      "learning_rate": 7.71093325135687e-06,
+      "loss": 0.9654,
+      "step": 4386
+    },
+    {
+      "epoch": 0.3381899475794018,
+      "grad_norm": 3.3396193981170654,
+      "learning_rate": 7.709884148977808e-06,
+      "loss": 0.8923,
+      "step": 4387
+    },
+    {
+      "epoch": 0.33826703669441877,
+      "grad_norm": 4.080255031585693,
+      "learning_rate": 7.708834877651504e-06,
+      "loss": 0.9685,
+      "step": 4388
+    },
+    {
+      "epoch": 0.3383441258094357,
+      "grad_norm": 3.752744436264038,
+      "learning_rate": 7.70778543744337e-06,
+      "loss": 1.0687,
+      "step": 4389
+    },
+    {
+      "epoch": 0.33842121492445265,
+      "grad_norm": 4.078830718994141,
+      "learning_rate": 7.706735828418844e-06,
+      "loss": 1.1614,
+      "step": 4390
+    },
+    {
+      "epoch": 0.3384983040394696,
+      "grad_norm": 3.583847999572754,
+      "learning_rate": 7.705686050643354e-06,
+      "loss": 1.0587,
+      "step": 4391
+    },
+    {
+      "epoch": 0.3385753931544866,
+      "grad_norm": 3.779529094696045,
+      "learning_rate": 7.704636104182353e-06,
+      "loss": 1.0217,
+      "step": 4392
+    },
+    {
+      "epoch": 0.33865248226950356,
+      "grad_norm": 3.4878971576690674,
+      "learning_rate": 7.703585989101298e-06,
+      "loss": 0.8588,
+      "step": 4393
+    },
+    {
+      "epoch": 0.3387295713845205,
+      "grad_norm": 3.9853031635284424,
+      "learning_rate": 7.70253570546566e-06,
+      "loss": 1.0825,
+      "step": 4394
+    },
+    {
+      "epoch": 0.33880666049953745,
+      "grad_norm": 3.818143606185913,
+      "learning_rate": 7.701485253340917e-06,
+      "loss": 1.0648,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3388837496145544,
+      "grad_norm": 3.480123281478882,
+      "learning_rate": 7.70043463279256e-06,
+      "loss": 0.9691,
+      "step": 4396
+    },
+    {
+      "epoch": 0.3389608387295714,
+      "grad_norm": 3.3591983318328857,
+      "learning_rate": 7.69938384388609e-06,
+      "loss": 0.8988,
+      "step": 4397
+    },
+    {
+      "epoch": 0.33903792784458836,
+      "grad_norm": 3.428823232650757,
+      "learning_rate": 7.698332886687016e-06,
+      "loss": 0.8853,
+      "step": 4398
+    },
+    {
+      "epoch": 0.33911501695960533,
+      "grad_norm": 3.8618509769439697,
+      "learning_rate": 7.697281761260863e-06,
+      "loss": 1.0317,
+      "step": 4399
+    },
+    {
+      "epoch": 0.33919210607462225,
+      "grad_norm": 3.7116105556488037,
+      "learning_rate": 7.696230467673163e-06,
+      "loss": 1.0838,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3392691951896392,
+      "grad_norm": 3.5771796703338623,
+      "learning_rate": 7.695179005989454e-06,
+      "loss": 0.995,
+      "step": 4401
+    },
+    {
+      "epoch": 0.3393462843046562,
+      "grad_norm": 3.8581273555755615,
+      "learning_rate": 7.694127376275295e-06,
+      "loss": 0.9682,
+      "step": 4402
+    },
+    {
+      "epoch": 0.33942337341967316,
+      "grad_norm": 3.888493776321411,
+      "learning_rate": 7.693075578596247e-06,
+      "loss": 1.0419,
+      "step": 4403
+    },
+    {
+      "epoch": 0.33950046253469013,
+      "grad_norm": 4.1317877769470215,
+      "learning_rate": 7.692023613017884e-06,
+      "loss": 1.0923,
+      "step": 4404
+    },
+    {
+      "epoch": 0.33957755164970704,
+      "grad_norm": 3.4401021003723145,
+      "learning_rate": 7.690971479605793e-06,
+      "loss": 0.9184,
+      "step": 4405
+    },
+    {
+      "epoch": 0.339654640764724,
+      "grad_norm": 3.6244630813598633,
+      "learning_rate": 7.689919178425565e-06,
+      "loss": 0.9051,
+      "step": 4406
+    },
+    {
+      "epoch": 0.339731729879741,
+      "grad_norm": 3.5508549213409424,
+      "learning_rate": 7.688866709542809e-06,
+      "loss": 1.0023,
+      "step": 4407
+    },
+    {
+      "epoch": 0.33980881899475796,
+      "grad_norm": 3.756072521209717,
+      "learning_rate": 7.687814073023138e-06,
+      "loss": 0.9766,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3398859081097749,
+      "grad_norm": 4.206667423248291,
+      "learning_rate": 7.686761268932182e-06,
+      "loss": 1.0179,
+      "step": 4409
+    },
+    {
+      "epoch": 0.33996299722479184,
+      "grad_norm": 3.559318780899048,
+      "learning_rate": 7.685708297335575e-06,
+      "loss": 1.0625,
+      "step": 4410
+    },
+    {
+      "epoch": 0.3400400863398088,
+      "grad_norm": 3.7180213928222656,
+      "learning_rate": 7.684655158298963e-06,
+      "loss": 1.0346,
+      "step": 4411
+    },
+    {
+      "epoch": 0.3401171754548258,
+      "grad_norm": 3.8383474349975586,
+      "learning_rate": 7.68360185188801e-06,
+      "loss": 0.945,
+      "step": 4412
+    },
+    {
+      "epoch": 0.34019426456984275,
+      "grad_norm": 3.727611780166626,
+      "learning_rate": 7.682548378168379e-06,
+      "loss": 1.0908,
+      "step": 4413
+    },
+    {
+      "epoch": 0.3402713536848597,
+      "grad_norm": 3.4572296142578125,
+      "learning_rate": 7.681494737205748e-06,
+      "loss": 0.9101,
+      "step": 4414
+    },
+    {
+      "epoch": 0.34034844279987664,
+      "grad_norm": 4.0167741775512695,
+      "learning_rate": 7.680440929065811e-06,
+      "loss": 1.0263,
+      "step": 4415
+    },
+    {
+      "epoch": 0.3404255319148936,
+      "grad_norm": 4.128690719604492,
+      "learning_rate": 7.679386953814262e-06,
+      "loss": 1.1199,
+      "step": 4416
+    },
+    {
+      "epoch": 0.3405026210299106,
+      "grad_norm": 3.2454535961151123,
+      "learning_rate": 7.678332811516815e-06,
+      "loss": 0.9708,
+      "step": 4417
+    },
+    {
+      "epoch": 0.34057971014492755,
+      "grad_norm": 3.2851145267486572,
+      "learning_rate": 7.677278502239187e-06,
+      "loss": 0.8788,
+      "step": 4418
+    },
+    {
+      "epoch": 0.3406567992599445,
+      "grad_norm": 3.677515983581543,
+      "learning_rate": 7.676224026047112e-06,
+      "loss": 0.9379,
+      "step": 4419
+    },
+    {
+      "epoch": 0.34073388837496144,
+      "grad_norm": 3.384584426879883,
+      "learning_rate": 7.67516938300633e-06,
+      "loss": 0.8972,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3408109774899784,
+      "grad_norm": 3.8261232376098633,
+      "learning_rate": 7.674114573182589e-06,
+      "loss": 1.0569,
+      "step": 4421
+    },
+    {
+      "epoch": 0.3408880666049954,
+      "grad_norm": 3.638108491897583,
+      "learning_rate": 7.673059596641657e-06,
+      "loss": 1.0784,
+      "step": 4422
+    },
+    {
+      "epoch": 0.34096515572001235,
+      "grad_norm": 3.599555730819702,
+      "learning_rate": 7.672004453449301e-06,
+      "loss": 1.0344,
+      "step": 4423
+    },
+    {
+      "epoch": 0.3410422448350293,
+      "grad_norm": 3.3416149616241455,
+      "learning_rate": 7.670949143671306e-06,
+      "loss": 0.8663,
+      "step": 4424
+    },
+    {
+      "epoch": 0.34111933395004623,
+      "grad_norm": 3.894286632537842,
+      "learning_rate": 7.669893667373466e-06,
+      "loss": 0.9725,
+      "step": 4425
+    },
+    {
+      "epoch": 0.3411964230650632,
+      "grad_norm": 3.740086078643799,
+      "learning_rate": 7.668838024621585e-06,
+      "loss": 0.9731,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3412735121800802,
+      "grad_norm": 3.730048179626465,
+      "learning_rate": 7.667782215481474e-06,
+      "loss": 0.912,
+      "step": 4427
+    },
+    {
+      "epoch": 0.34135060129509714,
+      "grad_norm": 3.557861328125,
+      "learning_rate": 7.66672624001896e-06,
+      "loss": 0.9324,
+      "step": 4428
+    },
+    {
+      "epoch": 0.3414276904101141,
+      "grad_norm": 3.767613649368286,
+      "learning_rate": 7.665670098299874e-06,
+      "loss": 0.8642,
+      "step": 4429
+    },
+    {
+      "epoch": 0.34150477952513103,
+      "grad_norm": 4.100453853607178,
+      "learning_rate": 7.664613790390065e-06,
+      "loss": 1.0388,
+      "step": 4430
+    },
+    {
+      "epoch": 0.341581868640148,
+      "grad_norm": 3.8830795288085938,
+      "learning_rate": 7.663557316355386e-06,
+      "loss": 1.005,
+      "step": 4431
+    },
+    {
+      "epoch": 0.34165895775516497,
+      "grad_norm": 3.5792341232299805,
+      "learning_rate": 7.662500676261703e-06,
+      "loss": 0.954,
+      "step": 4432
+    },
+    {
+      "epoch": 0.34173604687018194,
+      "grad_norm": 3.8511476516723633,
+      "learning_rate": 7.661443870174892e-06,
+      "loss": 1.0085,
+      "step": 4433
+    },
+    {
+      "epoch": 0.3418131359851989,
+      "grad_norm": 3.9277069568634033,
+      "learning_rate": 7.66038689816084e-06,
+      "loss": 1.0202,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3418902251002158,
+      "grad_norm": 3.5222573280334473,
+      "learning_rate": 7.659329760285443e-06,
+      "loss": 0.9149,
+      "step": 4435
+    },
+    {
+      "epoch": 0.3419673142152328,
+      "grad_norm": 3.536393404006958,
+      "learning_rate": 7.658272456614609e-06,
+      "loss": 0.9427,
+      "step": 4436
+    },
+    {
+      "epoch": 0.34204440333024977,
+      "grad_norm": 4.047040939331055,
+      "learning_rate": 7.657214987214254e-06,
+      "loss": 1.0167,
+      "step": 4437
+    },
+    {
+      "epoch": 0.34212149244526674,
+      "grad_norm": 3.8045451641082764,
+      "learning_rate": 7.65615735215031e-06,
+      "loss": 1.0209,
+      "step": 4438
+    },
+    {
+      "epoch": 0.3421985815602837,
+      "grad_norm": 3.7481818199157715,
+      "learning_rate": 7.655099551488708e-06,
+      "loss": 1.0584,
+      "step": 4439
+    },
+    {
+      "epoch": 0.3422756706753006,
+      "grad_norm": 3.487293004989624,
+      "learning_rate": 7.6540415852954e-06,
+      "loss": 0.9386,
+      "step": 4440
+    },
+    {
+      "epoch": 0.3423527597903176,
+      "grad_norm": 3.341665267944336,
+      "learning_rate": 7.652983453636346e-06,
+      "loss": 0.935,
+      "step": 4441
+    },
+    {
+      "epoch": 0.34242984890533457,
+      "grad_norm": 3.843733310699463,
+      "learning_rate": 7.651925156577514e-06,
+      "loss": 1.0647,
+      "step": 4442
+    },
+    {
+      "epoch": 0.34250693802035154,
+      "grad_norm": 3.3451244831085205,
+      "learning_rate": 7.650866694184883e-06,
+      "loss": 1.0161,
+      "step": 4443
+    },
+    {
+      "epoch": 0.3425840271353685,
+      "grad_norm": 3.6213300228118896,
+      "learning_rate": 7.649808066524442e-06,
+      "loss": 0.994,
+      "step": 4444
+    },
+    {
+      "epoch": 0.3426611162503854,
+      "grad_norm": 4.26886510848999,
+      "learning_rate": 7.648749273662194e-06,
+      "loss": 0.8898,
+      "step": 4445
+    },
+    {
+      "epoch": 0.3427382053654024,
+      "grad_norm": 3.5714519023895264,
+      "learning_rate": 7.647690315664145e-06,
+      "loss": 1.028,
+      "step": 4446
+    },
+    {
+      "epoch": 0.34281529448041936,
+      "grad_norm": 4.256858825683594,
+      "learning_rate": 7.646631192596317e-06,
+      "loss": 0.9492,
+      "step": 4447
+    },
+    {
+      "epoch": 0.34289238359543633,
+      "grad_norm": 3.999039649963379,
+      "learning_rate": 7.645571904524745e-06,
+      "loss": 1.0845,
+      "step": 4448
+    },
+    {
+      "epoch": 0.3429694727104533,
+      "grad_norm": 3.7288503646850586,
+      "learning_rate": 7.644512451515462e-06,
+      "loss": 0.9635,
+      "step": 4449
+    },
+    {
+      "epoch": 0.3430465618254702,
+      "grad_norm": 3.4463119506835938,
+      "learning_rate": 7.643452833634527e-06,
+      "loss": 0.9414,
+      "step": 4450
+    },
+    {
+      "epoch": 0.3431236509404872,
+      "grad_norm": 3.423847198486328,
+      "learning_rate": 7.642393050947997e-06,
+      "loss": 0.9491,
+      "step": 4451
+    },
+    {
+      "epoch": 0.34320074005550416,
+      "grad_norm": 4.031202793121338,
+      "learning_rate": 7.641333103521945e-06,
+      "loss": 1.1798,
+      "step": 4452
+    },
+    {
+      "epoch": 0.34327782917052113,
+      "grad_norm": 4.064019203186035,
+      "learning_rate": 7.640272991422456e-06,
+      "loss": 1.1341,
+      "step": 4453
+    },
+    {
+      "epoch": 0.3433549182855381,
+      "grad_norm": 3.5513041019439697,
+      "learning_rate": 7.63921271471562e-06,
+      "loss": 1.0552,
+      "step": 4454
+    },
+    {
+      "epoch": 0.343432007400555,
+      "grad_norm": 3.514437437057495,
+      "learning_rate": 7.638152273467538e-06,
+      "loss": 1.086,
+      "step": 4455
+    },
+    {
+      "epoch": 0.343509096515572,
+      "grad_norm": 3.334420680999756,
+      "learning_rate": 7.637091667744326e-06,
+      "loss": 0.8893,
+      "step": 4456
+    },
+    {
+      "epoch": 0.34358618563058896,
+      "grad_norm": 3.7241454124450684,
+      "learning_rate": 7.636030897612108e-06,
+      "loss": 0.9269,
+      "step": 4457
+    },
+    {
+      "epoch": 0.34366327474560593,
+      "grad_norm": 4.0841755867004395,
+      "learning_rate": 7.634969963137015e-06,
+      "loss": 0.9552,
+      "step": 4458
+    },
+    {
+      "epoch": 0.3437403638606229,
+      "grad_norm": 3.496382236480713,
+      "learning_rate": 7.633908864385191e-06,
+      "loss": 0.8975,
+      "step": 4459
+    },
+    {
+      "epoch": 0.3438174529756398,
+      "grad_norm": 3.9809958934783936,
+      "learning_rate": 7.63284760142279e-06,
+      "loss": 1.0904,
+      "step": 4460
+    },
+    {
+      "epoch": 0.3438945420906568,
+      "grad_norm": 3.6975362300872803,
+      "learning_rate": 7.631786174315978e-06,
+      "loss": 0.9786,
+      "step": 4461
+    },
+    {
+      "epoch": 0.34397163120567376,
+      "grad_norm": 4.022948265075684,
+      "learning_rate": 7.630724583130929e-06,
+      "loss": 0.9845,
+      "step": 4462
+    },
+    {
+      "epoch": 0.3440487203206907,
+      "grad_norm": 3.6614363193511963,
+      "learning_rate": 7.629662827933827e-06,
+      "loss": 1.0014,
+      "step": 4463
+    },
+    {
+      "epoch": 0.3441258094357077,
+      "grad_norm": 3.6465506553649902,
+      "learning_rate": 7.628600908790867e-06,
+      "loss": 0.97,
+      "step": 4464
+    },
+    {
+      "epoch": 0.3442028985507246,
+      "grad_norm": 3.7315773963928223,
+      "learning_rate": 7.627538825768253e-06,
+      "loss": 0.9663,
+      "step": 4465
+    },
+    {
+      "epoch": 0.3442799876657416,
+      "grad_norm": 3.598884344100952,
+      "learning_rate": 7.626476578932202e-06,
+      "loss": 0.9673,
+      "step": 4466
+    },
+    {
+      "epoch": 0.34435707678075855,
+      "grad_norm": 3.5586609840393066,
+      "learning_rate": 7.625414168348939e-06,
+      "loss": 1.0037,
+      "step": 4467
+    },
+    {
+      "epoch": 0.3444341658957755,
+      "grad_norm": 3.392164468765259,
+      "learning_rate": 7.6243515940847005e-06,
+      "loss": 1.0022,
+      "step": 4468
+    },
+    {
+      "epoch": 0.3445112550107925,
+      "grad_norm": 3.7672579288482666,
+      "learning_rate": 7.623288856205733e-06,
+      "loss": 0.9914,
+      "step": 4469
+    },
+    {
+      "epoch": 0.3445883441258094,
+      "grad_norm": 3.6310787200927734,
+      "learning_rate": 7.62222595477829e-06,
+      "loss": 0.9816,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3446654332408264,
+      "grad_norm": 3.9517312049865723,
+      "learning_rate": 7.6211628898686386e-06,
+      "loss": 1.0428,
+      "step": 4471
+    },
+    {
+      "epoch": 0.34474252235584335,
+      "grad_norm": 3.518615245819092,
+      "learning_rate": 7.620099661543059e-06,
+      "loss": 0.8673,
+      "step": 4472
+    },
+    {
+      "epoch": 0.3448196114708603,
+      "grad_norm": 3.833643674850464,
+      "learning_rate": 7.619036269867835e-06,
+      "loss": 0.9825,
+      "step": 4473
+    },
+    {
+      "epoch": 0.3448967005858773,
+      "grad_norm": 3.776273012161255,
+      "learning_rate": 7.617972714909263e-06,
+      "loss": 0.9213,
+      "step": 4474
+    },
+    {
+      "epoch": 0.3449737897008942,
+      "grad_norm": 4.035387992858887,
+      "learning_rate": 7.616908996733651e-06,
+      "loss": 0.9992,
+      "step": 4475
+    },
+    {
+      "epoch": 0.3450508788159112,
+      "grad_norm": 3.4117751121520996,
+      "learning_rate": 7.615845115407316e-06,
+      "loss": 0.9576,
+      "step": 4476
+    },
+    {
+      "epoch": 0.34512796793092815,
+      "grad_norm": 3.680464267730713,
+      "learning_rate": 7.6147810709965864e-06,
+      "loss": 0.952,
+      "step": 4477
+    },
+    {
+      "epoch": 0.3452050570459451,
+      "grad_norm": 3.7311110496520996,
+      "learning_rate": 7.6137168635677996e-06,
+      "loss": 1.094,
+      "step": 4478
+    },
+    {
+      "epoch": 0.3452821461609621,
+      "grad_norm": 3.6351912021636963,
+      "learning_rate": 7.612652493187302e-06,
+      "loss": 1.0033,
+      "step": 4479
+    },
+    {
+      "epoch": 0.345359235275979,
+      "grad_norm": 3.4685473442077637,
+      "learning_rate": 7.611587959921453e-06,
+      "loss": 1.0091,
+      "step": 4480
+    },
+    {
+      "epoch": 0.345436324390996,
+      "grad_norm": 3.7951955795288086,
+      "learning_rate": 7.6105232638366225e-06,
+      "loss": 0.9334,
+      "step": 4481
+    },
+    {
+      "epoch": 0.34551341350601295,
+      "grad_norm": 3.6663379669189453,
+      "learning_rate": 7.609458404999186e-06,
+      "loss": 0.998,
+      "step": 4482
+    },
+    {
+      "epoch": 0.3455905026210299,
+      "grad_norm": 3.7089757919311523,
+      "learning_rate": 7.608393383475532e-06,
+      "loss": 0.9697,
+      "step": 4483
+    },
+    {
+      "epoch": 0.3456675917360469,
+      "grad_norm": 3.706681489944458,
+      "learning_rate": 7.607328199332059e-06,
+      "loss": 0.9436,
+      "step": 4484
+    },
+    {
+      "epoch": 0.34574468085106386,
+      "grad_norm": 3.7007081508636475,
+      "learning_rate": 7.606262852635178e-06,
+      "loss": 1.0404,
+      "step": 4485
+    },
+    {
+      "epoch": 0.34582176996608077,
+      "grad_norm": 3.912898063659668,
+      "learning_rate": 7.605197343451305e-06,
+      "loss": 1.0742,
+      "step": 4486
+    },
+    {
+      "epoch": 0.34589885908109774,
+      "grad_norm": 3.450777292251587,
+      "learning_rate": 7.604131671846872e-06,
+      "loss": 0.9221,
+      "step": 4487
+    },
+    {
+      "epoch": 0.3459759481961147,
+      "grad_norm": 3.4001710414886475,
+      "learning_rate": 7.6030658378883145e-06,
+      "loss": 0.9292,
+      "step": 4488
+    },
+    {
+      "epoch": 0.3460530373111317,
+      "grad_norm": 3.4011619091033936,
+      "learning_rate": 7.601999841642085e-06,
+      "loss": 0.9305,
+      "step": 4489
+    },
+    {
+      "epoch": 0.34613012642614865,
+      "grad_norm": 3.7454307079315186,
+      "learning_rate": 7.60093368317464e-06,
+      "loss": 1.0018,
+      "step": 4490
+    },
+    {
+      "epoch": 0.34620721554116557,
+      "grad_norm": 3.7179946899414062,
+      "learning_rate": 7.599867362552451e-06,
+      "loss": 0.9987,
+      "step": 4491
+    },
+    {
+      "epoch": 0.34628430465618254,
+      "grad_norm": 3.8563878536224365,
+      "learning_rate": 7.598800879841998e-06,
+      "loss": 1.0384,
+      "step": 4492
+    },
+    {
+      "epoch": 0.3463613937711995,
+      "grad_norm": 3.63057804107666,
+      "learning_rate": 7.5977342351097694e-06,
+      "loss": 1.0281,
+      "step": 4493
+    },
+    {
+      "epoch": 0.3464384828862165,
+      "grad_norm": 3.7689592838287354,
+      "learning_rate": 7.596667428422264e-06,
+      "loss": 1.0204,
+      "step": 4494
+    },
+    {
+      "epoch": 0.34651557200123345,
+      "grad_norm": 4.023124694824219,
+      "learning_rate": 7.595600459845994e-06,
+      "loss": 1.1758,
+      "step": 4495
+    },
+    {
+      "epoch": 0.34659266111625037,
+      "grad_norm": 3.7010927200317383,
+      "learning_rate": 7.594533329447479e-06,
+      "loss": 1.0144,
+      "step": 4496
+    },
+    {
+      "epoch": 0.34666975023126734,
+      "grad_norm": 3.7244632244110107,
+      "learning_rate": 7.593466037293247e-06,
+      "loss": 0.9995,
+      "step": 4497
+    },
+    {
+      "epoch": 0.3467468393462843,
+      "grad_norm": 3.411987781524658,
+      "learning_rate": 7.5923985834498405e-06,
+      "loss": 0.943,
+      "step": 4498
+    },
+    {
+      "epoch": 0.3468239284613013,
+      "grad_norm": 3.9309608936309814,
+      "learning_rate": 7.5913309679838074e-06,
+      "loss": 0.9554,
+      "step": 4499
+    },
+    {
+      "epoch": 0.34690101757631825,
+      "grad_norm": 3.5526673793792725,
+      "learning_rate": 7.590263190961711e-06,
+      "loss": 0.9885,
+      "step": 4500
+    },
+    {
+      "epoch": 0.34697810669133516,
+      "grad_norm": 3.6589465141296387,
+      "learning_rate": 7.589195252450118e-06,
+      "loss": 1.0778,
+      "step": 4501
+    },
+    {
+      "epoch": 0.34705519580635213,
+      "grad_norm": 3.7149205207824707,
+      "learning_rate": 7.588127152515611e-06,
+      "loss": 1.0306,
+      "step": 4502
+    },
+    {
+      "epoch": 0.3471322849213691,
+      "grad_norm": 4.081293106079102,
+      "learning_rate": 7.587058891224781e-06,
+      "loss": 0.9663,
+      "step": 4503
+    },
+    {
+      "epoch": 0.3472093740363861,
+      "grad_norm": 3.5740342140197754,
+      "learning_rate": 7.585990468644229e-06,
+      "loss": 1.0298,
+      "step": 4504
+    },
+    {
+      "epoch": 0.34728646315140305,
+      "grad_norm": 3.4450860023498535,
+      "learning_rate": 7.584921884840563e-06,
+      "loss": 0.9135,
+      "step": 4505
+    },
+    {
+      "epoch": 0.34736355226641996,
+      "grad_norm": 3.9143495559692383,
+      "learning_rate": 7.583853139880406e-06,
+      "loss": 1.092,
+      "step": 4506
+    },
+    {
+      "epoch": 0.34744064138143693,
+      "grad_norm": 3.25154447555542,
+      "learning_rate": 7.5827842338303866e-06,
+      "loss": 0.9825,
+      "step": 4507
+    },
+    {
+      "epoch": 0.3475177304964539,
+      "grad_norm": 3.853174924850464,
+      "learning_rate": 7.581715166757147e-06,
+      "loss": 0.9713,
+      "step": 4508
+    },
+    {
+      "epoch": 0.3475948196114709,
+      "grad_norm": 3.6071982383728027,
+      "learning_rate": 7.58064593872734e-06,
+      "loss": 0.9837,
+      "step": 4509
+    },
+    {
+      "epoch": 0.34767190872648784,
+      "grad_norm": 3.3574037551879883,
+      "learning_rate": 7.579576549807621e-06,
+      "loss": 0.8727,
+      "step": 4510
+    },
+    {
+      "epoch": 0.34774899784150476,
+      "grad_norm": 3.408640146255493,
+      "learning_rate": 7.578507000064668e-06,
+      "loss": 0.9032,
+      "step": 4511
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 3.5927014350891113,
+      "learning_rate": 7.5774372895651545e-06,
+      "loss": 1.0204,
+      "step": 4512
+    },
+    {
+      "epoch": 0.3479031760715387,
+      "grad_norm": 4.023251533508301,
+      "learning_rate": 7.576367418375776e-06,
+      "loss": 1.0703,
+      "step": 4513
+    },
+    {
+      "epoch": 0.34798026518655567,
+      "grad_norm": 3.498119831085205,
+      "learning_rate": 7.575297386563232e-06,
+      "loss": 1.0155,
+      "step": 4514
+    },
+    {
+      "epoch": 0.34805735430157264,
+      "grad_norm": 4.321279525756836,
+      "learning_rate": 7.574227194194234e-06,
+      "loss": 1.0152,
+      "step": 4515
+    },
+    {
+      "epoch": 0.34813444341658956,
+      "grad_norm": 3.662602186203003,
+      "learning_rate": 7.573156841335503e-06,
+      "loss": 1.0795,
+      "step": 4516
+    },
+    {
+      "epoch": 0.3482115325316065,
+      "grad_norm": 3.4204254150390625,
+      "learning_rate": 7.572086328053769e-06,
+      "loss": 0.9286,
+      "step": 4517
+    },
+    {
+      "epoch": 0.3482886216466235,
+      "grad_norm": 4.003498554229736,
+      "learning_rate": 7.571015654415774e-06,
+      "loss": 0.991,
+      "step": 4518
+    },
+    {
+      "epoch": 0.34836571076164047,
+      "grad_norm": 3.632580518722534,
+      "learning_rate": 7.5699448204882684e-06,
+      "loss": 0.9635,
+      "step": 4519
+    },
+    {
+      "epoch": 0.34844279987665744,
+      "grad_norm": 3.763371706008911,
+      "learning_rate": 7.568873826338015e-06,
+      "loss": 1.0865,
+      "step": 4520
+    },
+    {
+      "epoch": 0.34851988899167435,
+      "grad_norm": 3.367755889892578,
+      "learning_rate": 7.567802672031781e-06,
+      "loss": 0.9446,
+      "step": 4521
+    },
+    {
+      "epoch": 0.3485969781066913,
+      "grad_norm": 3.568023204803467,
+      "learning_rate": 7.56673135763635e-06,
+      "loss": 0.9173,
+      "step": 4522
+    },
+    {
+      "epoch": 0.3486740672217083,
+      "grad_norm": 3.6380043029785156,
+      "learning_rate": 7.565659883218512e-06,
+      "loss": 0.9864,
+      "step": 4523
+    },
+    {
+      "epoch": 0.34875115633672527,
+      "grad_norm": 3.404252290725708,
+      "learning_rate": 7.56458824884507e-06,
+      "loss": 0.9846,
+      "step": 4524
+    },
+    {
+      "epoch": 0.34882824545174224,
+      "grad_norm": 3.7746942043304443,
+      "learning_rate": 7.563516454582831e-06,
+      "loss": 0.9901,
+      "step": 4525
+    },
+    {
+      "epoch": 0.34890533456675915,
+      "grad_norm": 3.6926469802856445,
+      "learning_rate": 7.562444500498618e-06,
+      "loss": 1.0837,
+      "step": 4526
+    },
+    {
+      "epoch": 0.3489824236817761,
+      "grad_norm": 3.7840046882629395,
+      "learning_rate": 7.561372386659262e-06,
+      "loss": 1.1675,
+      "step": 4527
+    },
+    {
+      "epoch": 0.3490595127967931,
+      "grad_norm": 4.055497646331787,
+      "learning_rate": 7.560300113131604e-06,
+      "loss": 1.0338,
+      "step": 4528
+    },
+    {
+      "epoch": 0.34913660191181006,
+      "grad_norm": 4.383268356323242,
+      "learning_rate": 7.559227679982493e-06,
+      "loss": 1.0946,
+      "step": 4529
+    },
+    {
+      "epoch": 0.34921369102682703,
+      "grad_norm": 3.687899351119995,
+      "learning_rate": 7.558155087278791e-06,
+      "loss": 0.8959,
+      "step": 4530
+    },
+    {
+      "epoch": 0.34929078014184395,
+      "grad_norm": 4.111197471618652,
+      "learning_rate": 7.557082335087369e-06,
+      "loss": 0.9816,
+      "step": 4531
+    },
+    {
+      "epoch": 0.3493678692568609,
+      "grad_norm": 3.7356393337249756,
+      "learning_rate": 7.556009423475106e-06,
+      "loss": 0.9332,
+      "step": 4532
+    },
+    {
+      "epoch": 0.3494449583718779,
+      "grad_norm": 3.6677143573760986,
+      "learning_rate": 7.554936352508895e-06,
+      "loss": 1.0524,
+      "step": 4533
+    },
+    {
+      "epoch": 0.34952204748689486,
+      "grad_norm": 3.7901206016540527,
+      "learning_rate": 7.5538631222556325e-06,
+      "loss": 0.9411,
+      "step": 4534
+    },
+    {
+      "epoch": 0.34959913660191183,
+      "grad_norm": 3.7148752212524414,
+      "learning_rate": 7.552789732782233e-06,
+      "loss": 1.0595,
+      "step": 4535
+    },
+    {
+      "epoch": 0.34967622571692875,
+      "grad_norm": 3.53884220123291,
+      "learning_rate": 7.551716184155614e-06,
+      "loss": 0.969,
+      "step": 4536
+    },
+    {
+      "epoch": 0.3497533148319457,
+      "grad_norm": 3.777893543243408,
+      "learning_rate": 7.5506424764427065e-06,
+      "loss": 1.011,
+      "step": 4537
+    },
+    {
+      "epoch": 0.3498304039469627,
+      "grad_norm": 3.4374454021453857,
+      "learning_rate": 7.549568609710451e-06,
+      "loss": 0.9996,
+      "step": 4538
+    },
+    {
+      "epoch": 0.34990749306197966,
+      "grad_norm": 3.2913565635681152,
+      "learning_rate": 7.548494584025797e-06,
+      "loss": 0.9579,
+      "step": 4539
+    },
+    {
+      "epoch": 0.34998458217699663,
+      "grad_norm": 3.424445152282715,
+      "learning_rate": 7.547420399455705e-06,
+      "loss": 0.9822,
+      "step": 4540
+    },
+    {
+      "epoch": 0.35006167129201354,
+      "grad_norm": 3.8123433589935303,
+      "learning_rate": 7.5463460560671446e-06,
+      "loss": 0.9697,
+      "step": 4541
+    },
+    {
+      "epoch": 0.3501387604070305,
+      "grad_norm": 3.7088186740875244,
+      "learning_rate": 7.5452715539270945e-06,
+      "loss": 1.0171,
+      "step": 4542
+    },
+    {
+      "epoch": 0.3502158495220475,
+      "grad_norm": 3.776679754257202,
+      "learning_rate": 7.544196893102547e-06,
+      "loss": 1.0093,
+      "step": 4543
+    },
+    {
+      "epoch": 0.35029293863706445,
+      "grad_norm": 3.353478193283081,
+      "learning_rate": 7.543122073660498e-06,
+      "loss": 0.8819,
+      "step": 4544
+    },
+    {
+      "epoch": 0.3503700277520814,
+      "grad_norm": 3.6976723670959473,
+      "learning_rate": 7.542047095667959e-06,
+      "loss": 1.0509,
+      "step": 4545
+    },
+    {
+      "epoch": 0.35044711686709834,
+      "grad_norm": 4.122467517852783,
+      "learning_rate": 7.540971959191952e-06,
+      "loss": 1.1499,
+      "step": 4546
+    },
+    {
+      "epoch": 0.3505242059821153,
+      "grad_norm": 3.71630859375,
+      "learning_rate": 7.5398966642995e-06,
+      "loss": 1.0094,
+      "step": 4547
+    },
+    {
+      "epoch": 0.3506012950971323,
+      "grad_norm": 3.93642258644104,
+      "learning_rate": 7.538821211057648e-06,
+      "loss": 1.1285,
+      "step": 4548
+    },
+    {
+      "epoch": 0.35067838421214925,
+      "grad_norm": 3.9508731365203857,
+      "learning_rate": 7.537745599533442e-06,
+      "loss": 1.2053,
+      "step": 4549
+    },
+    {
+      "epoch": 0.3507554733271662,
+      "grad_norm": 4.088351726531982,
+      "learning_rate": 7.536669829793939e-06,
+      "loss": 0.9771,
+      "step": 4550
+    },
+    {
+      "epoch": 0.35083256244218314,
+      "grad_norm": 3.7594566345214844,
+      "learning_rate": 7.535593901906212e-06,
+      "loss": 1.0617,
+      "step": 4551
+    },
+    {
+      "epoch": 0.3509096515572001,
+      "grad_norm": 3.9711499214172363,
+      "learning_rate": 7.534517815937336e-06,
+      "loss": 1.0532,
+      "step": 4552
+    },
+    {
+      "epoch": 0.3509867406722171,
+      "grad_norm": 4.007244110107422,
+      "learning_rate": 7.533441571954401e-06,
+      "loss": 1.0847,
+      "step": 4553
+    },
+    {
+      "epoch": 0.35106382978723405,
+      "grad_norm": 3.750323534011841,
+      "learning_rate": 7.532365170024507e-06,
+      "loss": 0.951,
+      "step": 4554
+    },
+    {
+      "epoch": 0.351140918902251,
+      "grad_norm": 3.2767701148986816,
+      "learning_rate": 7.531288610214758e-06,
+      "loss": 1.0074,
+      "step": 4555
+    },
+    {
+      "epoch": 0.35121800801726794,
+      "grad_norm": 4.005361080169678,
+      "learning_rate": 7.530211892592274e-06,
+      "loss": 1.0625,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3512950971322849,
+      "grad_norm": 3.4741199016571045,
+      "learning_rate": 7.529135017224185e-06,
+      "loss": 0.9702,
+      "step": 4557
+    },
+    {
+      "epoch": 0.3513721862473019,
+      "grad_norm": 4.308572292327881,
+      "learning_rate": 7.528057984177624e-06,
+      "loss": 1.1244,
+      "step": 4558
+    },
+    {
+      "epoch": 0.35144927536231885,
+      "grad_norm": 3.521604299545288,
+      "learning_rate": 7.526980793519742e-06,
+      "loss": 0.9298,
+      "step": 4559
+    },
+    {
+      "epoch": 0.3515263644773358,
+      "grad_norm": 3.766963243484497,
+      "learning_rate": 7.525903445317695e-06,
+      "loss": 0.9063,
+      "step": 4560
+    },
+    {
+      "epoch": 0.35160345359235273,
+      "grad_norm": 3.509277582168579,
+      "learning_rate": 7.524825939638649e-06,
+      "loss": 0.9582,
+      "step": 4561
+    },
+    {
+      "epoch": 0.3516805427073697,
+      "grad_norm": 3.2499780654907227,
+      "learning_rate": 7.5237482765497835e-06,
+      "loss": 0.9871,
+      "step": 4562
+    },
+    {
+      "epoch": 0.3517576318223867,
+      "grad_norm": 3.8931937217712402,
+      "learning_rate": 7.522670456118284e-06,
+      "loss": 1.0293,
+      "step": 4563
+    },
+    {
+      "epoch": 0.35183472093740364,
+      "grad_norm": 4.080838203430176,
+      "learning_rate": 7.521592478411346e-06,
+      "loss": 0.9787,
+      "step": 4564
+    },
+    {
+      "epoch": 0.3519118100524206,
+      "grad_norm": 3.7860987186431885,
+      "learning_rate": 7.520514343496177e-06,
+      "loss": 1.0544,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3519888991674376,
+      "grad_norm": 3.832820415496826,
+      "learning_rate": 7.519436051439991e-06,
+      "loss": 1.045,
+      "step": 4566
+    },
+    {
+      "epoch": 0.3520659882824545,
+      "grad_norm": 3.489471912384033,
+      "learning_rate": 7.518357602310018e-06,
+      "loss": 1.0061,
+      "step": 4567
+    },
+    {
+      "epoch": 0.35214307739747147,
+      "grad_norm": 3.8605971336364746,
+      "learning_rate": 7.517278996173489e-06,
+      "loss": 1.1125,
+      "step": 4568
+    },
+    {
+      "epoch": 0.35222016651248844,
+      "grad_norm": 3.234241247177124,
+      "learning_rate": 7.516200233097655e-06,
+      "loss": 0.9031,
+      "step": 4569
+    },
+    {
+      "epoch": 0.3522972556275054,
+      "grad_norm": 3.4962003231048584,
+      "learning_rate": 7.515121313149767e-06,
+      "loss": 1.0268,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3523743447425224,
+      "grad_norm": 4.1234211921691895,
+      "learning_rate": 7.51404223639709e-06,
+      "loss": 1.0732,
+      "step": 4571
+    },
+    {
+      "epoch": 0.3524514338575393,
+      "grad_norm": 3.3490140438079834,
+      "learning_rate": 7.512963002906902e-06,
+      "loss": 0.9798,
+      "step": 4572
+    },
+    {
+      "epoch": 0.35252852297255627,
+      "grad_norm": 3.6099109649658203,
+      "learning_rate": 7.5118836127464845e-06,
+      "loss": 1.0534,
+      "step": 4573
+    },
+    {
+      "epoch": 0.35260561208757324,
+      "grad_norm": 3.4530441761016846,
+      "learning_rate": 7.5108040659831325e-06,
+      "loss": 0.96,
+      "step": 4574
+    },
+    {
+      "epoch": 0.3526827012025902,
+      "grad_norm": 3.9595425128936768,
+      "learning_rate": 7.509724362684153e-06,
+      "loss": 0.99,
+      "step": 4575
+    },
+    {
+      "epoch": 0.3527597903176072,
+      "grad_norm": 3.436499834060669,
+      "learning_rate": 7.508644502916857e-06,
+      "loss": 1.0278,
+      "step": 4576
+    },
+    {
+      "epoch": 0.3528368794326241,
+      "grad_norm": 3.768009901046753,
+      "learning_rate": 7.507564486748567e-06,
+      "loss": 0.9352,
+      "step": 4577
+    },
+    {
+      "epoch": 0.35291396854764107,
+      "grad_norm": 3.8910608291625977,
+      "learning_rate": 7.50648431424662e-06,
+      "loss": 1.1054,
+      "step": 4578
+    },
+    {
+      "epoch": 0.35299105766265804,
+      "grad_norm": 3.816119909286499,
+      "learning_rate": 7.5054039854783565e-06,
+      "loss": 1.0385,
+      "step": 4579
+    },
+    {
+      "epoch": 0.353068146777675,
+      "grad_norm": 3.278871774673462,
+      "learning_rate": 7.5043235005111314e-06,
+      "loss": 0.9679,
+      "step": 4580
+    },
+    {
+      "epoch": 0.353145235892692,
+      "grad_norm": 3.655571222305298,
+      "learning_rate": 7.503242859412306e-06,
+      "loss": 1.0785,
+      "step": 4581
+    },
+    {
+      "epoch": 0.3532223250077089,
+      "grad_norm": 3.78136944770813,
+      "learning_rate": 7.502162062249252e-06,
+      "loss": 0.9802,
+      "step": 4582
+    },
+    {
+      "epoch": 0.35329941412272586,
+      "grad_norm": 3.6029317378997803,
+      "learning_rate": 7.501081109089354e-06,
+      "loss": 0.991,
+      "step": 4583
+    },
+    {
+      "epoch": 0.35337650323774283,
+      "grad_norm": 3.5113279819488525,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.0691,
+      "step": 4584
+    },
+    {
+      "epoch": 0.3534535923527598,
+      "grad_norm": 3.781364679336548,
+      "learning_rate": 7.498918735048596e-06,
+      "loss": 0.9593,
+      "step": 4585
+    },
+    {
+      "epoch": 0.3535306814677768,
+      "grad_norm": 3.832432270050049,
+      "learning_rate": 7.497837314302551e-06,
+      "loss": 0.933,
+      "step": 4586
+    },
+    {
+      "epoch": 0.3536077705827937,
+      "grad_norm": 3.3761022090911865,
+      "learning_rate": 7.496755737829284e-06,
+      "loss": 0.9601,
+      "step": 4587
+    },
+    {
+      "epoch": 0.35368485969781066,
+      "grad_norm": 3.549546241760254,
+      "learning_rate": 7.4956740056962294e-06,
+      "loss": 0.9896,
+      "step": 4588
+    },
+    {
+      "epoch": 0.35376194881282763,
+      "grad_norm": 4.057165622711182,
+      "learning_rate": 7.494592117970826e-06,
+      "loss": 1.0498,
+      "step": 4589
+    },
+    {
+      "epoch": 0.3538390379278446,
+      "grad_norm": 3.55377197265625,
+      "learning_rate": 7.493510074720523e-06,
+      "loss": 0.9978,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3539161270428616,
+      "grad_norm": 3.4123051166534424,
+      "learning_rate": 7.492427876012782e-06,
+      "loss": 0.9557,
+      "step": 4591
+    },
+    {
+      "epoch": 0.3539932161578785,
+      "grad_norm": 3.7144315242767334,
+      "learning_rate": 7.491345521915071e-06,
+      "loss": 0.9637,
+      "step": 4592
+    },
+    {
+      "epoch": 0.35407030527289546,
+      "grad_norm": 3.7125496864318848,
+      "learning_rate": 7.490263012494869e-06,
+      "loss": 0.9757,
+      "step": 4593
+    },
+    {
+      "epoch": 0.35414739438791243,
+      "grad_norm": 3.407020092010498,
+      "learning_rate": 7.489180347819668e-06,
+      "loss": 0.987,
+      "step": 4594
+    },
+    {
+      "epoch": 0.3542244835029294,
+      "grad_norm": 3.3304386138916016,
+      "learning_rate": 7.488097527956962e-06,
+      "loss": 0.9076,
+      "step": 4595
+    },
+    {
+      "epoch": 0.35430157261794637,
+      "grad_norm": 3.7230682373046875,
+      "learning_rate": 7.487014552974263e-06,
+      "loss": 1.0456,
+      "step": 4596
+    },
+    {
+      "epoch": 0.3543786617329633,
+      "grad_norm": 3.5402743816375732,
+      "learning_rate": 7.4859314229390856e-06,
+      "loss": 1.0437,
+      "step": 4597
+    },
+    {
+      "epoch": 0.35445575084798026,
+      "grad_norm": 3.747596025466919,
+      "learning_rate": 7.484848137918958e-06,
+      "loss": 1.0386,
+      "step": 4598
+    },
+    {
+      "epoch": 0.3545328399629972,
+      "grad_norm": 3.624826192855835,
+      "learning_rate": 7.483764697981422e-06,
+      "loss": 0.9824,
+      "step": 4599
+    },
+    {
+      "epoch": 0.3546099290780142,
+      "grad_norm": 3.802626609802246,
+      "learning_rate": 7.482681103194018e-06,
+      "loss": 1.0693,
+      "step": 4600
+    },
+    {
+      "epoch": 0.35468701819303117,
+      "grad_norm": 3.6754977703094482,
+      "learning_rate": 7.481597353624306e-06,
+      "loss": 0.9418,
+      "step": 4601
+    },
+    {
+      "epoch": 0.3547641073080481,
+      "grad_norm": 3.6683568954467773,
+      "learning_rate": 7.480513449339851e-06,
+      "loss": 1.0815,
+      "step": 4602
+    },
+    {
+      "epoch": 0.35484119642306505,
+      "grad_norm": 3.4328901767730713,
+      "learning_rate": 7.47942939040823e-06,
+      "loss": 0.888,
+      "step": 4603
+    },
+    {
+      "epoch": 0.354918285538082,
+      "grad_norm": 3.344658374786377,
+      "learning_rate": 7.478345176897027e-06,
+      "loss": 0.8972,
+      "step": 4604
+    },
+    {
+      "epoch": 0.354995374653099,
+      "grad_norm": 3.514586925506592,
+      "learning_rate": 7.4772608088738395e-06,
+      "loss": 0.8603,
+      "step": 4605
+    },
+    {
+      "epoch": 0.35507246376811596,
+      "grad_norm": 3.7426164150238037,
+      "learning_rate": 7.4761762864062694e-06,
+      "loss": 0.8437,
+      "step": 4606
+    },
+    {
+      "epoch": 0.3551495528831329,
+      "grad_norm": 4.099521636962891,
+      "learning_rate": 7.475091609561933e-06,
+      "loss": 0.9417,
+      "step": 4607
+    },
+    {
+      "epoch": 0.35522664199814985,
+      "grad_norm": 4.204078674316406,
+      "learning_rate": 7.4740067784084534e-06,
+      "loss": 1.0172,
+      "step": 4608
+    },
+    {
+      "epoch": 0.3553037311131668,
+      "grad_norm": 3.2971291542053223,
+      "learning_rate": 7.472921793013463e-06,
+      "loss": 0.9418,
+      "step": 4609
+    },
+    {
+      "epoch": 0.3553808202281838,
+      "grad_norm": 3.2113230228424072,
+      "learning_rate": 7.4718366534446085e-06,
+      "loss": 0.9096,
+      "step": 4610
+    },
+    {
+      "epoch": 0.35545790934320076,
+      "grad_norm": 3.6125385761260986,
+      "learning_rate": 7.4707513597695396e-06,
+      "loss": 0.8899,
+      "step": 4611
+    },
+    {
+      "epoch": 0.3555349984582177,
+      "grad_norm": 3.8694043159484863,
+      "learning_rate": 7.469665912055919e-06,
+      "loss": 1.0594,
+      "step": 4612
+    },
+    {
+      "epoch": 0.35561208757323465,
+      "grad_norm": 3.883565902709961,
+      "learning_rate": 7.4685803103714204e-06,
+      "loss": 0.8622,
+      "step": 4613
+    },
+    {
+      "epoch": 0.3556891766882516,
+      "grad_norm": 3.652117967605591,
+      "learning_rate": 7.4674945547837245e-06,
+      "loss": 1.045,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3557662658032686,
+      "grad_norm": 3.6539828777313232,
+      "learning_rate": 7.466408645360524e-06,
+      "loss": 1.117,
+      "step": 4615
+    },
+    {
+      "epoch": 0.35584335491828556,
+      "grad_norm": 3.4314699172973633,
+      "learning_rate": 7.465322582169516e-06,
+      "loss": 0.8358,
+      "step": 4616
+    },
+    {
+      "epoch": 0.3559204440333025,
+      "grad_norm": 3.411808729171753,
+      "learning_rate": 7.464236365278413e-06,
+      "loss": 0.9704,
+      "step": 4617
+    },
+    {
+      "epoch": 0.35599753314831944,
+      "grad_norm": 3.6066629886627197,
+      "learning_rate": 7.463149994754938e-06,
+      "loss": 0.9963,
+      "step": 4618
+    },
+    {
+      "epoch": 0.3560746222633364,
+      "grad_norm": 3.517822742462158,
+      "learning_rate": 7.462063470666816e-06,
+      "loss": 0.9606,
+      "step": 4619
+    },
+    {
+      "epoch": 0.3561517113783534,
+      "grad_norm": 3.6595687866210938,
+      "learning_rate": 7.460976793081789e-06,
+      "loss": 0.944,
+      "step": 4620
+    },
+    {
+      "epoch": 0.35622880049337036,
+      "grad_norm": 3.9949941635131836,
+      "learning_rate": 7.459889962067603e-06,
+      "loss": 1.1163,
+      "step": 4621
+    },
+    {
+      "epoch": 0.35630588960838727,
+      "grad_norm": 3.570941209793091,
+      "learning_rate": 7.458802977692018e-06,
+      "loss": 1.02,
+      "step": 4622
+    },
+    {
+      "epoch": 0.35638297872340424,
+      "grad_norm": 3.210967540740967,
+      "learning_rate": 7.4577158400228034e-06,
+      "loss": 0.8364,
+      "step": 4623
+    },
+    {
+      "epoch": 0.3564600678384212,
+      "grad_norm": 3.722440719604492,
+      "learning_rate": 7.4566285491277334e-06,
+      "loss": 0.9846,
+      "step": 4624
+    },
+    {
+      "epoch": 0.3565371569534382,
+      "grad_norm": 3.4632368087768555,
+      "learning_rate": 7.455541105074598e-06,
+      "loss": 1.0375,
+      "step": 4625
+    },
+    {
+      "epoch": 0.35661424606845515,
+      "grad_norm": 3.4531891345977783,
+      "learning_rate": 7.454453507931192e-06,
+      "loss": 0.8937,
+      "step": 4626
+    },
+    {
+      "epoch": 0.35669133518347207,
+      "grad_norm": 3.6453826427459717,
+      "learning_rate": 7.45336575776532e-06,
+      "loss": 0.9928,
+      "step": 4627
+    },
+    {
+      "epoch": 0.35676842429848904,
+      "grad_norm": 3.3601088523864746,
+      "learning_rate": 7.452277854644801e-06,
+      "loss": 0.8869,
+      "step": 4628
+    },
+    {
+      "epoch": 0.356845513413506,
+      "grad_norm": 3.36553955078125,
+      "learning_rate": 7.45118979863746e-06,
+      "loss": 0.9932,
+      "step": 4629
+    },
+    {
+      "epoch": 0.356922602528523,
+      "grad_norm": 3.6436736583709717,
+      "learning_rate": 7.450101589811127e-06,
+      "loss": 1.0522,
+      "step": 4630
+    },
+    {
+      "epoch": 0.35699969164353995,
+      "grad_norm": 3.701080083847046,
+      "learning_rate": 7.449013228233651e-06,
+      "loss": 0.7591,
+      "step": 4631
+    },
+    {
+      "epoch": 0.35707678075855687,
+      "grad_norm": 3.6947665214538574,
+      "learning_rate": 7.447924713972883e-06,
+      "loss": 0.9894,
+      "step": 4632
+    },
+    {
+      "epoch": 0.35715386987357384,
+      "grad_norm": 3.8331315517425537,
+      "learning_rate": 7.4468360470966875e-06,
+      "loss": 0.9405,
+      "step": 4633
+    },
+    {
+      "epoch": 0.3572309589885908,
+      "grad_norm": 3.559194803237915,
+      "learning_rate": 7.445747227672937e-06,
+      "loss": 1.0249,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3573080481036078,
+      "grad_norm": 4.069003582000732,
+      "learning_rate": 7.444658255769513e-06,
+      "loss": 1.0037,
+      "step": 4635
+    },
+    {
+      "epoch": 0.35738513721862475,
+      "grad_norm": 3.7197611331939697,
+      "learning_rate": 7.4435691314543105e-06,
+      "loss": 1.0408,
+      "step": 4636
+    },
+    {
+      "epoch": 0.35746222633364166,
+      "grad_norm": 3.3894405364990234,
+      "learning_rate": 7.442479854795226e-06,
+      "loss": 1.0021,
+      "step": 4637
+    },
+    {
+      "epoch": 0.35753931544865863,
+      "grad_norm": 3.269364833831787,
+      "learning_rate": 7.441390425860172e-06,
+      "loss": 0.9195,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3576164045636756,
+      "grad_norm": 3.6949777603149414,
+      "learning_rate": 7.440300844717071e-06,
+      "loss": 1.111,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3576934936786926,
+      "grad_norm": 3.708047389984131,
+      "learning_rate": 7.43921111143385e-06,
+      "loss": 1.0689,
+      "step": 4640
+    },
+    {
+      "epoch": 0.35777058279370955,
+      "grad_norm": 3.682943105697632,
+      "learning_rate": 7.438121226078449e-06,
+      "loss": 1.041,
+      "step": 4641
+    },
+    {
+      "epoch": 0.35784767190872646,
+      "grad_norm": 3.447502613067627,
+      "learning_rate": 7.437031188718818e-06,
+      "loss": 0.9712,
+      "step": 4642
+    },
+    {
+      "epoch": 0.35792476102374343,
+      "grad_norm": 3.6242482662200928,
+      "learning_rate": 7.435940999422913e-06,
+      "loss": 0.9694,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3580018501387604,
+      "grad_norm": 3.254422903060913,
+      "learning_rate": 7.434850658258704e-06,
+      "loss": 0.8993,
+      "step": 4644
+    },
+    {
+      "epoch": 0.3580789392537774,
+      "grad_norm": 3.3965585231781006,
+      "learning_rate": 7.433760165294166e-06,
+      "loss": 0.9396,
+      "step": 4645
+    },
+    {
+      "epoch": 0.35815602836879434,
+      "grad_norm": 3.665678024291992,
+      "learning_rate": 7.4326695205972865e-06,
+      "loss": 1.008,
+      "step": 4646
+    },
+    {
+      "epoch": 0.35823311748381126,
+      "grad_norm": 3.5869436264038086,
+      "learning_rate": 7.431578724236062e-06,
+      "loss": 0.965,
+      "step": 4647
+    },
+    {
+      "epoch": 0.35831020659882823,
+      "grad_norm": 3.809087038040161,
+      "learning_rate": 7.430487776278497e-06,
+      "loss": 0.9678,
+      "step": 4648
+    },
+    {
+      "epoch": 0.3583872957138452,
+      "grad_norm": 3.8319554328918457,
+      "learning_rate": 7.429396676792607e-06,
+      "loss": 1.0688,
+      "step": 4649
+    },
+    {
+      "epoch": 0.35846438482886217,
+      "grad_norm": 3.6857879161834717,
+      "learning_rate": 7.428305425846416e-06,
+      "loss": 1.0452,
+      "step": 4650
+    },
+    {
+      "epoch": 0.35854147394387914,
+      "grad_norm": 3.5876760482788086,
+      "learning_rate": 7.427214023507958e-06,
+      "loss": 1.0092,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3586185630588961,
+      "grad_norm": 3.979463815689087,
+      "learning_rate": 7.426122469845277e-06,
+      "loss": 1.1185,
+      "step": 4652
+    },
+    {
+      "epoch": 0.358695652173913,
+      "grad_norm": 3.456955909729004,
+      "learning_rate": 7.4250307649264265e-06,
+      "loss": 0.9516,
+      "step": 4653
+    },
+    {
+      "epoch": 0.35877274128893,
+      "grad_norm": 3.4941182136535645,
+      "learning_rate": 7.423938908819466e-06,
+      "loss": 0.945,
+      "step": 4654
+    },
+    {
+      "epoch": 0.35884983040394697,
+      "grad_norm": 3.835120677947998,
+      "learning_rate": 7.4228469015924675e-06,
+      "loss": 0.9512,
+      "step": 4655
+    },
+    {
+      "epoch": 0.35892691951896394,
+      "grad_norm": 3.682276964187622,
+      "learning_rate": 7.421754743313514e-06,
+      "loss": 1.0016,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3590040086339809,
+      "grad_norm": 3.423520088195801,
+      "learning_rate": 7.420662434050695e-06,
+      "loss": 0.8797,
+      "step": 4657
+    },
+    {
+      "epoch": 0.3590810977489978,
+      "grad_norm": 3.420644998550415,
+      "learning_rate": 7.41956997387211e-06,
+      "loss": 0.8758,
+      "step": 4658
+    },
+    {
+      "epoch": 0.3591581868640148,
+      "grad_norm": 3.3552937507629395,
+      "learning_rate": 7.418477362845868e-06,
+      "loss": 0.9692,
+      "step": 4659
+    },
+    {
+      "epoch": 0.35923527597903177,
+      "grad_norm": 3.6283655166625977,
+      "learning_rate": 7.41738460104009e-06,
+      "loss": 1.0536,
+      "step": 4660
+    },
+    {
+      "epoch": 0.35931236509404874,
+      "grad_norm": 3.8392016887664795,
+      "learning_rate": 7.4162916885229e-06,
+      "loss": 0.9439,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3593894542090657,
+      "grad_norm": 3.771641492843628,
+      "learning_rate": 7.415198625362437e-06,
+      "loss": 1.0178,
+      "step": 4662
+    },
+    {
+      "epoch": 0.3594665433240826,
+      "grad_norm": 3.600102424621582,
+      "learning_rate": 7.414105411626851e-06,
+      "loss": 1.0072,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3595436324390996,
+      "grad_norm": 3.391453742980957,
+      "learning_rate": 7.413012047384292e-06,
+      "loss": 0.9686,
+      "step": 4664
+    },
+    {
+      "epoch": 0.35962072155411656,
+      "grad_norm": 3.72904634475708,
+      "learning_rate": 7.411918532702932e-06,
+      "loss": 1.0089,
+      "step": 4665
+    },
+    {
+      "epoch": 0.35969781066913353,
+      "grad_norm": 3.5617527961730957,
+      "learning_rate": 7.410824867650943e-06,
+      "loss": 0.9439,
+      "step": 4666
+    },
+    {
+      "epoch": 0.3597748997841505,
+      "grad_norm": 3.4571316242218018,
+      "learning_rate": 7.409731052296508e-06,
+      "loss": 0.9568,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3598519888991674,
+      "grad_norm": 3.765029191970825,
+      "learning_rate": 7.408637086707824e-06,
+      "loss": 1.071,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3599290780141844,
+      "grad_norm": 3.57428240776062,
+      "learning_rate": 7.4075429709530896e-06,
+      "loss": 1.0565,
+      "step": 4669
+    },
+    {
+      "epoch": 0.36000616712920136,
+      "grad_norm": 3.6238112449645996,
+      "learning_rate": 7.4064487051005215e-06,
+      "loss": 1.02,
+      "step": 4670
+    },
+    {
+      "epoch": 0.36008325624421833,
+      "grad_norm": 3.5156960487365723,
+      "learning_rate": 7.405354289218341e-06,
+      "loss": 0.9572,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3601603453592353,
+      "grad_norm": 3.847799062728882,
+      "learning_rate": 7.404259723374778e-06,
+      "loss": 0.8973,
+      "step": 4672
+    },
+    {
+      "epoch": 0.3602374344742522,
+      "grad_norm": 3.390068531036377,
+      "learning_rate": 7.403165007638071e-06,
+      "loss": 0.9345,
+      "step": 4673
+    },
+    {
+      "epoch": 0.3603145235892692,
+      "grad_norm": 4.004146575927734,
+      "learning_rate": 7.402070142076475e-06,
+      "loss": 0.9548,
+      "step": 4674
+    },
+    {
+      "epoch": 0.36039161270428616,
+      "grad_norm": 3.443192720413208,
+      "learning_rate": 7.400975126758244e-06,
+      "loss": 1.0265,
+      "step": 4675
+    },
+    {
+      "epoch": 0.36046870181930313,
+      "grad_norm": 3.7502102851867676,
+      "learning_rate": 7.39987996175165e-06,
+      "loss": 0.9425,
+      "step": 4676
+    },
+    {
+      "epoch": 0.3605457909343201,
+      "grad_norm": 3.6927671432495117,
+      "learning_rate": 7.398784647124969e-06,
+      "loss": 0.9693,
+      "step": 4677
+    },
+    {
+      "epoch": 0.360622880049337,
+      "grad_norm": 3.445166826248169,
+      "learning_rate": 7.397689182946489e-06,
+      "loss": 1.0462,
+      "step": 4678
+    },
+    {
+      "epoch": 0.360699969164354,
+      "grad_norm": 3.651033401489258,
+      "learning_rate": 7.396593569284507e-06,
+      "loss": 1.0167,
+      "step": 4679
+    },
+    {
+      "epoch": 0.36077705827937095,
+      "grad_norm": 3.4636173248291016,
+      "learning_rate": 7.395497806207328e-06,
+      "loss": 0.9994,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3608541473943879,
+      "grad_norm": 3.4041786193847656,
+      "learning_rate": 7.394401893783267e-06,
+      "loss": 0.9804,
+      "step": 4681
+    },
+    {
+      "epoch": 0.3609312365094049,
+      "grad_norm": 3.829814910888672,
+      "learning_rate": 7.393305832080649e-06,
+      "loss": 0.893,
+      "step": 4682
+    },
+    {
+      "epoch": 0.3610083256244218,
+      "grad_norm": 3.543970823287964,
+      "learning_rate": 7.392209621167808e-06,
+      "loss": 0.9415,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3610854147394388,
+      "grad_norm": 3.526637315750122,
+      "learning_rate": 7.3911132611130865e-06,
+      "loss": 0.8818,
+      "step": 4684
+    },
+    {
+      "epoch": 0.36116250385445575,
+      "grad_norm": 3.4459733963012695,
+      "learning_rate": 7.390016751984837e-06,
+      "loss": 0.9302,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3612395929694727,
+      "grad_norm": 3.7464380264282227,
+      "learning_rate": 7.388920093851422e-06,
+      "loss": 0.9382,
+      "step": 4686
+    },
+    {
+      "epoch": 0.3613166820844897,
+      "grad_norm": 3.308462381362915,
+      "learning_rate": 7.387823286781211e-06,
+      "loss": 1.0056,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3613937711995066,
+      "grad_norm": 3.8184940814971924,
+      "learning_rate": 7.386726330842584e-06,
+      "loss": 1.075,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3614708603145236,
+      "grad_norm": 3.9437508583068848,
+      "learning_rate": 7.385629226103932e-06,
+      "loss": 1.0295,
+      "step": 4689
+    },
+    {
+      "epoch": 0.36154794942954055,
+      "grad_norm": 3.679063558578491,
+      "learning_rate": 7.384531972633654e-06,
+      "loss": 1.067,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3616250385445575,
+      "grad_norm": 3.801813840866089,
+      "learning_rate": 7.383434570500156e-06,
+      "loss": 0.9291,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3617021276595745,
+      "grad_norm": 3.7192230224609375,
+      "learning_rate": 7.382337019771859e-06,
+      "loss": 1.0775,
+      "step": 4692
+    },
+    {
+      "epoch": 0.3617792167745914,
+      "grad_norm": 3.7992069721221924,
+      "learning_rate": 7.381239320517185e-06,
+      "loss": 1.0711,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3618563058896084,
+      "grad_norm": 3.7021334171295166,
+      "learning_rate": 7.380141472804572e-06,
+      "loss": 0.9285,
+      "step": 4694
+    },
+    {
+      "epoch": 0.36193339500462535,
+      "grad_norm": 3.7016396522521973,
+      "learning_rate": 7.379043476702468e-06,
+      "loss": 0.9494,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3620104841196423,
+      "grad_norm": 3.413544178009033,
+      "learning_rate": 7.377945332279322e-06,
+      "loss": 0.9474,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3620875732346593,
+      "grad_norm": 3.475022077560425,
+      "learning_rate": 7.376847039603601e-06,
+      "loss": 1.0284,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3621646623496762,
+      "grad_norm": 4.036199569702148,
+      "learning_rate": 7.375748598743777e-06,
+      "loss": 0.9613,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3622417514646932,
+      "grad_norm": 3.2614598274230957,
+      "learning_rate": 7.374650009768332e-06,
+      "loss": 0.9619,
+      "step": 4699
+    },
+    {
+      "epoch": 0.36231884057971014,
+      "grad_norm": 3.5047171115875244,
+      "learning_rate": 7.373551272745757e-06,
+      "loss": 0.995,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3623959296947271,
+      "grad_norm": 3.3241288661956787,
+      "learning_rate": 7.372452387744554e-06,
+      "loss": 0.8698,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3624730188097441,
+      "grad_norm": 3.509653091430664,
+      "learning_rate": 7.371353354833231e-06,
+      "loss": 1.053,
+      "step": 4702
+    },
+    {
+      "epoch": 0.362550107924761,
+      "grad_norm": 4.059185981750488,
+      "learning_rate": 7.370254174080309e-06,
+      "loss": 1.1282,
+      "step": 4703
+    },
+    {
+      "epoch": 0.36262719703977797,
+      "grad_norm": 3.687779426574707,
+      "learning_rate": 7.3691548455543146e-06,
+      "loss": 0.9489,
+      "step": 4704
+    },
+    {
+      "epoch": 0.36270428615479494,
+      "grad_norm": 3.582054615020752,
+      "learning_rate": 7.368055369323787e-06,
+      "loss": 0.9714,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3627813752698119,
+      "grad_norm": 3.6636664867401123,
+      "learning_rate": 7.36695574545727e-06,
+      "loss": 1.0117,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3628584643848289,
+      "grad_norm": 3.7531189918518066,
+      "learning_rate": 7.365855974023321e-06,
+      "loss": 0.9775,
+      "step": 4707
+    },
+    {
+      "epoch": 0.3629355534998458,
+      "grad_norm": 3.9122326374053955,
+      "learning_rate": 7.364756055090506e-06,
+      "loss": 1.0191,
+      "step": 4708
+    },
+    {
+      "epoch": 0.36301264261486277,
+      "grad_norm": 4.153857231140137,
+      "learning_rate": 7.363655988727398e-06,
+      "loss": 1.0468,
+      "step": 4709
+    },
+    {
+      "epoch": 0.36308973172987974,
+      "grad_norm": 3.230003833770752,
+      "learning_rate": 7.36255577500258e-06,
+      "loss": 0.8938,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3631668208448967,
+      "grad_norm": 3.6824822425842285,
+      "learning_rate": 7.3614554139846475e-06,
+      "loss": 0.9479,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3632439099599137,
+      "grad_norm": 3.7477197647094727,
+      "learning_rate": 7.3603549057421975e-06,
+      "loss": 1.0547,
+      "step": 4712
+    },
+    {
+      "epoch": 0.3633209990749306,
+      "grad_norm": 3.6179938316345215,
+      "learning_rate": 7.359254250343846e-06,
+      "loss": 0.9269,
+      "step": 4713
+    },
+    {
+      "epoch": 0.36339808818994757,
+      "grad_norm": 3.606153964996338,
+      "learning_rate": 7.35815344785821e-06,
+      "loss": 1.08,
+      "step": 4714
+    },
+    {
+      "epoch": 0.36347517730496454,
+      "grad_norm": 3.634908676147461,
+      "learning_rate": 7.35705249835392e-06,
+      "loss": 0.8419,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3635522664199815,
+      "grad_norm": 3.726848602294922,
+      "learning_rate": 7.355951401899614e-06,
+      "loss": 1.0174,
+      "step": 4716
+    },
+    {
+      "epoch": 0.3636293555349985,
+      "grad_norm": 3.769972562789917,
+      "learning_rate": 7.35485015856394e-06,
+      "loss": 0.9593,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3637064446500154,
+      "grad_norm": 4.418411731719971,
+      "learning_rate": 7.353748768415554e-06,
+      "loss": 1.0226,
+      "step": 4718
+    },
+    {
+      "epoch": 0.36378353376503236,
+      "grad_norm": 3.3944032192230225,
+      "learning_rate": 7.352647231523124e-06,
+      "loss": 0.9723,
+      "step": 4719
+    },
+    {
+      "epoch": 0.36386062288004933,
+      "grad_norm": 3.6218276023864746,
+      "learning_rate": 7.351545547955321e-06,
+      "loss": 1.0031,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3639377119950663,
+      "grad_norm": 3.672168016433716,
+      "learning_rate": 7.350443717780834e-06,
+      "loss": 0.9893,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3640148011100833,
+      "grad_norm": 3.8659307956695557,
+      "learning_rate": 7.3493417410683545e-06,
+      "loss": 0.9818,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3640918902251002,
+      "grad_norm": 3.814786672592163,
+      "learning_rate": 7.3482396178865846e-06,
+      "loss": 1.0346,
+      "step": 4723
+    },
+    {
+      "epoch": 0.36416897934011716,
+      "grad_norm": 3.595031976699829,
+      "learning_rate": 7.347137348304237e-06,
+      "loss": 0.9958,
+      "step": 4724
+    },
+    {
+      "epoch": 0.36424606845513413,
+      "grad_norm": 3.76802921295166,
+      "learning_rate": 7.346034932390031e-06,
+      "loss": 0.9374,
+      "step": 4725
+    },
+    {
+      "epoch": 0.3643231575701511,
+      "grad_norm": 3.507568120956421,
+      "learning_rate": 7.3449323702126985e-06,
+      "loss": 0.9716,
+      "step": 4726
+    },
+    {
+      "epoch": 0.36440024668516807,
+      "grad_norm": 3.8154513835906982,
+      "learning_rate": 7.343829661840977e-06,
+      "loss": 1.0059,
+      "step": 4727
+    },
+    {
+      "epoch": 0.364477335800185,
+      "grad_norm": 3.319742202758789,
+      "learning_rate": 7.342726807343615e-06,
+      "loss": 0.9758,
+      "step": 4728
+    },
+    {
+      "epoch": 0.36455442491520196,
+      "grad_norm": 3.301845073699951,
+      "learning_rate": 7.341623806789371e-06,
+      "loss": 0.9537,
+      "step": 4729
+    },
+    {
+      "epoch": 0.36463151403021893,
+      "grad_norm": 3.977522611618042,
+      "learning_rate": 7.340520660247008e-06,
+      "loss": 1.0368,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3647086031452359,
+      "grad_norm": 3.8079488277435303,
+      "learning_rate": 7.339417367785304e-06,
+      "loss": 0.9951,
+      "step": 4731
+    },
+    {
+      "epoch": 0.36478569226025287,
+      "grad_norm": 3.3151187896728516,
+      "learning_rate": 7.338313929473046e-06,
+      "loss": 1.0278,
+      "step": 4732
+    },
+    {
+      "epoch": 0.3648627813752698,
+      "grad_norm": 3.971615791320801,
+      "learning_rate": 7.337210345379022e-06,
+      "loss": 0.9549,
+      "step": 4733
+    },
+    {
+      "epoch": 0.36493987049028676,
+      "grad_norm": 3.8018195629119873,
+      "learning_rate": 7.3361066155720385e-06,
+      "loss": 1.0072,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3650169596053037,
+      "grad_norm": 3.6749868392944336,
+      "learning_rate": 7.335002740120907e-06,
+      "loss": 0.9764,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3650940487203207,
+      "grad_norm": 3.655430316925049,
+      "learning_rate": 7.333898719094448e-06,
+      "loss": 0.9881,
+      "step": 4736
+    },
+    {
+      "epoch": 0.36517113783533767,
+      "grad_norm": 3.9375107288360596,
+      "learning_rate": 7.3327945525614906e-06,
+      "loss": 1.0533,
+      "step": 4737
+    },
+    {
+      "epoch": 0.36524822695035464,
+      "grad_norm": 3.5556159019470215,
+      "learning_rate": 7.3316902405908756e-06,
+      "loss": 0.9244,
+      "step": 4738
+    },
+    {
+      "epoch": 0.36532531606537155,
+      "grad_norm": 3.914153575897217,
+      "learning_rate": 7.33058578325145e-06,
+      "loss": 1.0044,
+      "step": 4739
+    },
+    {
+      "epoch": 0.3654024051803885,
+      "grad_norm": 3.557025194168091,
+      "learning_rate": 7.329481180612072e-06,
+      "loss": 0.8395,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3654794942954055,
+      "grad_norm": 3.5067689418792725,
+      "learning_rate": 7.328376432741605e-06,
+      "loss": 1.0821,
+      "step": 4741
+    },
+    {
+      "epoch": 0.36555658341042246,
+      "grad_norm": 3.890801191329956,
+      "learning_rate": 7.327271539708927e-06,
+      "loss": 1.0238,
+      "step": 4742
+    },
+    {
+      "epoch": 0.36563367252543943,
+      "grad_norm": 4.020145893096924,
+      "learning_rate": 7.326166501582922e-06,
+      "loss": 0.9887,
+      "step": 4743
+    },
+    {
+      "epoch": 0.36571076164045635,
+      "grad_norm": 3.579129219055176,
+      "learning_rate": 7.325061318432482e-06,
+      "loss": 1.0122,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3657878507554733,
+      "grad_norm": 3.599555730819702,
+      "learning_rate": 7.323955990326514e-06,
+      "loss": 1.0213,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3658649398704903,
+      "grad_norm": 3.7440710067749023,
+      "learning_rate": 7.322850517333924e-06,
+      "loss": 1.1071,
+      "step": 4746
+    },
+    {
+      "epoch": 0.36594202898550726,
+      "grad_norm": 3.8629238605499268,
+      "learning_rate": 7.321744899523634e-06,
+      "loss": 1.0812,
+      "step": 4747
+    },
+    {
+      "epoch": 0.36601911810052423,
+      "grad_norm": 3.8284130096435547,
+      "learning_rate": 7.320639136964576e-06,
+      "loss": 0.9624,
+      "step": 4748
+    },
+    {
+      "epoch": 0.36609620721554115,
+      "grad_norm": 3.523813247680664,
+      "learning_rate": 7.319533229725685e-06,
+      "loss": 1.0628,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3661732963305581,
+      "grad_norm": 3.4013636112213135,
+      "learning_rate": 7.31842717787591e-06,
+      "loss": 0.9275,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3662503854455751,
+      "grad_norm": 3.6408772468566895,
+      "learning_rate": 7.31732098148421e-06,
+      "loss": 1.0767,
+      "step": 4751
+    },
+    {
+      "epoch": 0.36632747456059206,
+      "grad_norm": 3.796874523162842,
+      "learning_rate": 7.316214640619546e-06,
+      "loss": 1.013,
+      "step": 4752
+    },
+    {
+      "epoch": 0.36640456367560903,
+      "grad_norm": 3.552623748779297,
+      "learning_rate": 7.3151081553508975e-06,
+      "loss": 0.9438,
+      "step": 4753
+    },
+    {
+      "epoch": 0.36648165279062594,
+      "grad_norm": 3.6614010334014893,
+      "learning_rate": 7.314001525747244e-06,
+      "loss": 0.981,
+      "step": 4754
+    },
+    {
+      "epoch": 0.3665587419056429,
+      "grad_norm": 3.726024627685547,
+      "learning_rate": 7.31289475187758e-06,
+      "loss": 0.9823,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3666358310206599,
+      "grad_norm": 3.74556303024292,
+      "learning_rate": 7.311787833810908e-06,
+      "loss": 0.9622,
+      "step": 4756
+    },
+    {
+      "epoch": 0.36671292013567686,
+      "grad_norm": 3.351370096206665,
+      "learning_rate": 7.310680771616238e-06,
+      "loss": 0.9525,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3667900092506938,
+      "grad_norm": 3.492908477783203,
+      "learning_rate": 7.309573565362588e-06,
+      "loss": 0.9845,
+      "step": 4758
+    },
+    {
+      "epoch": 0.36686709836571074,
+      "grad_norm": 3.719702959060669,
+      "learning_rate": 7.308466215118988e-06,
+      "loss": 0.9873,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3669441874807277,
+      "grad_norm": 3.778932809829712,
+      "learning_rate": 7.307358720954476e-06,
+      "loss": 1.0009,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3670212765957447,
+      "grad_norm": 3.6195807456970215,
+      "learning_rate": 7.306251082938096e-06,
+      "loss": 0.8939,
+      "step": 4761
+    },
+    {
+      "epoch": 0.36709836571076165,
+      "grad_norm": 3.6259400844573975,
+      "learning_rate": 7.305143301138908e-06,
+      "loss": 1.0111,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3671754548257786,
+      "grad_norm": 4.1105637550354,
+      "learning_rate": 7.3040353756259726e-06,
+      "loss": 1.0318,
+      "step": 4763
+    },
+    {
+      "epoch": 0.36725254394079554,
+      "grad_norm": 4.061164379119873,
+      "learning_rate": 7.302927306468365e-06,
+      "loss": 0.9631,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3673296330558125,
+      "grad_norm": 3.501284122467041,
+      "learning_rate": 7.301819093735165e-06,
+      "loss": 1.0098,
+      "step": 4765
+    },
+    {
+      "epoch": 0.3674067221708295,
+      "grad_norm": 3.7744534015655518,
+      "learning_rate": 7.3007107374954665e-06,
+      "loss": 1.1278,
+      "step": 4766
+    },
+    {
+      "epoch": 0.36748381128584645,
+      "grad_norm": 3.3170313835144043,
+      "learning_rate": 7.299602237818371e-06,
+      "loss": 1.0022,
+      "step": 4767
+    },
+    {
+      "epoch": 0.3675609004008634,
+      "grad_norm": 3.677799701690674,
+      "learning_rate": 7.298493594772985e-06,
+      "loss": 1.0452,
+      "step": 4768
+    },
+    {
+      "epoch": 0.36763798951588034,
+      "grad_norm": 3.811732292175293,
+      "learning_rate": 7.297384808428428e-06,
+      "loss": 0.8941,
+      "step": 4769
+    },
+    {
+      "epoch": 0.3677150786308973,
+      "grad_norm": 4.003880977630615,
+      "learning_rate": 7.296275878853826e-06,
+      "loss": 1.0617,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3677921677459143,
+      "grad_norm": 3.452606439590454,
+      "learning_rate": 7.295166806118315e-06,
+      "loss": 0.9734,
+      "step": 4771
+    },
+    {
+      "epoch": 0.36786925686093125,
+      "grad_norm": 3.6096274852752686,
+      "learning_rate": 7.294057590291043e-06,
+      "loss": 0.8906,
+      "step": 4772
+    },
+    {
+      "epoch": 0.3679463459759482,
+      "grad_norm": 3.7491250038146973,
+      "learning_rate": 7.2929482314411596e-06,
+      "loss": 0.959,
+      "step": 4773
+    },
+    {
+      "epoch": 0.36802343509096513,
+      "grad_norm": 3.7471847534179688,
+      "learning_rate": 7.291838729637829e-06,
+      "loss": 1.1626,
+      "step": 4774
+    },
+    {
+      "epoch": 0.3681005242059821,
+      "grad_norm": 3.6201562881469727,
+      "learning_rate": 7.290729084950226e-06,
+      "loss": 0.9828,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3681776133209991,
+      "grad_norm": 3.8552613258361816,
+      "learning_rate": 7.289619297447525e-06,
+      "loss": 0.9679,
+      "step": 4776
+    },
+    {
+      "epoch": 0.36825470243601605,
+      "grad_norm": 3.5603456497192383,
+      "learning_rate": 7.288509367198922e-06,
+      "loss": 0.9275,
+      "step": 4777
+    },
+    {
+      "epoch": 0.368331791551033,
+      "grad_norm": 3.6187384128570557,
+      "learning_rate": 7.28739929427361e-06,
+      "loss": 1.0267,
+      "step": 4778
+    },
+    {
+      "epoch": 0.36840888066604993,
+      "grad_norm": 3.5130581855773926,
+      "learning_rate": 7.2862890787408e-06,
+      "loss": 0.9113,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3684859697810669,
+      "grad_norm": 3.5186641216278076,
+      "learning_rate": 7.2851787206697075e-06,
+      "loss": 0.9837,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3685630588960839,
+      "grad_norm": 3.7700576782226562,
+      "learning_rate": 7.284068220129557e-06,
+      "loss": 1.0265,
+      "step": 4781
+    },
+    {
+      "epoch": 0.36864014801110084,
+      "grad_norm": 3.770547389984131,
+      "learning_rate": 7.282957577189581e-06,
+      "loss": 1.0115,
+      "step": 4782
+    },
+    {
+      "epoch": 0.3687172371261178,
+      "grad_norm": 3.8225295543670654,
+      "learning_rate": 7.281846791919025e-06,
+      "loss": 1.1445,
+      "step": 4783
+    },
+    {
+      "epoch": 0.36879432624113473,
+      "grad_norm": 3.4760961532592773,
+      "learning_rate": 7.280735864387138e-06,
+      "loss": 0.9826,
+      "step": 4784
+    },
+    {
+      "epoch": 0.3688714153561517,
+      "grad_norm": 3.5458803176879883,
+      "learning_rate": 7.279624794663184e-06,
+      "loss": 0.9137,
+      "step": 4785
+    },
+    {
+      "epoch": 0.36894850447116867,
+      "grad_norm": 3.6354055404663086,
+      "learning_rate": 7.27851358281643e-06,
+      "loss": 0.9363,
+      "step": 4786
+    },
+    {
+      "epoch": 0.36902559358618564,
+      "grad_norm": 3.4249465465545654,
+      "learning_rate": 7.277402228916155e-06,
+      "loss": 0.9732,
+      "step": 4787
+    },
+    {
+      "epoch": 0.3691026827012026,
+      "grad_norm": 4.003117561340332,
+      "learning_rate": 7.2762907330316446e-06,
+      "loss": 1.0536,
+      "step": 4788
+    },
+    {
+      "epoch": 0.3691797718162195,
+      "grad_norm": 3.547496795654297,
+      "learning_rate": 7.275179095232197e-06,
+      "loss": 0.9676,
+      "step": 4789
+    },
+    {
+      "epoch": 0.3692568609312365,
+      "grad_norm": 3.738219976425171,
+      "learning_rate": 7.274067315587116e-06,
+      "loss": 1.0024,
+      "step": 4790
+    },
+    {
+      "epoch": 0.36933395004625347,
+      "grad_norm": 4.052086353302002,
+      "learning_rate": 7.272955394165717e-06,
+      "loss": 0.9979,
+      "step": 4791
+    },
+    {
+      "epoch": 0.36941103916127044,
+      "grad_norm": 3.6811420917510986,
+      "learning_rate": 7.27184333103732e-06,
+      "loss": 1.0432,
+      "step": 4792
+    },
+    {
+      "epoch": 0.3694881282762874,
+      "grad_norm": 3.4397454261779785,
+      "learning_rate": 7.270731126271257e-06,
+      "loss": 0.8792,
+      "step": 4793
+    },
+    {
+      "epoch": 0.3695652173913043,
+      "grad_norm": 3.70704984664917,
+      "learning_rate": 7.26961877993687e-06,
+      "loss": 0.8806,
+      "step": 4794
+    },
+    {
+      "epoch": 0.3696423065063213,
+      "grad_norm": 4.066430568695068,
+      "learning_rate": 7.268506292103505e-06,
+      "loss": 1.0428,
+      "step": 4795
+    },
+    {
+      "epoch": 0.36971939562133826,
+      "grad_norm": 3.72969388961792,
+      "learning_rate": 7.267393662840525e-06,
+      "loss": 1.0417,
+      "step": 4796
+    },
+    {
+      "epoch": 0.36979648473635524,
+      "grad_norm": 3.8739657402038574,
+      "learning_rate": 7.26628089221729e-06,
+      "loss": 1.0413,
+      "step": 4797
+    },
+    {
+      "epoch": 0.3698735738513722,
+      "grad_norm": 3.4731998443603516,
+      "learning_rate": 7.265167980303179e-06,
+      "loss": 1.0326,
+      "step": 4798
+    },
+    {
+      "epoch": 0.3699506629663891,
+      "grad_norm": 3.697890281677246,
+      "learning_rate": 7.264054927167577e-06,
+      "loss": 1.0505,
+      "step": 4799
+    },
+    {
+      "epoch": 0.3700277520814061,
+      "grad_norm": 3.670499086380005,
+      "learning_rate": 7.2629417328798755e-06,
+      "loss": 0.9704,
+      "step": 4800
+    },
+    {
+      "epoch": 0.37010484119642306,
+      "grad_norm": 3.542783260345459,
+      "learning_rate": 7.2618283975094785e-06,
+      "loss": 1.0082,
+      "step": 4801
+    },
+    {
+      "epoch": 0.37018193031144003,
+      "grad_norm": 3.7526941299438477,
+      "learning_rate": 7.260714921125795e-06,
+      "loss": 1.05,
+      "step": 4802
+    },
+    {
+      "epoch": 0.370259019426457,
+      "grad_norm": 3.547579050064087,
+      "learning_rate": 7.259601303798243e-06,
+      "loss": 0.9653,
+      "step": 4803
+    },
+    {
+      "epoch": 0.3703361085414739,
+      "grad_norm": 3.573000192642212,
+      "learning_rate": 7.258487545596254e-06,
+      "loss": 0.9571,
+      "step": 4804
+    },
+    {
+      "epoch": 0.3704131976564909,
+      "grad_norm": 3.745025873184204,
+      "learning_rate": 7.257373646589262e-06,
+      "loss": 1.0854,
+      "step": 4805
+    },
+    {
+      "epoch": 0.37049028677150786,
+      "grad_norm": 3.414745330810547,
+      "learning_rate": 7.256259606846715e-06,
+      "loss": 0.9017,
+      "step": 4806
+    },
+    {
+      "epoch": 0.37056737588652483,
+      "grad_norm": 3.3791048526763916,
+      "learning_rate": 7.255145426438068e-06,
+      "loss": 0.946,
+      "step": 4807
+    },
+    {
+      "epoch": 0.3706444650015418,
+      "grad_norm": 3.603844165802002,
+      "learning_rate": 7.254031105432781e-06,
+      "loss": 0.956,
+      "step": 4808
+    },
+    {
+      "epoch": 0.3707215541165587,
+      "grad_norm": 4.20005464553833,
+      "learning_rate": 7.252916643900331e-06,
+      "loss": 1.0829,
+      "step": 4809
+    },
+    {
+      "epoch": 0.3707986432315757,
+      "grad_norm": 3.4412059783935547,
+      "learning_rate": 7.251802041910194e-06,
+      "loss": 1.0052,
+      "step": 4810
+    },
+    {
+      "epoch": 0.37087573234659266,
+      "grad_norm": 3.931774616241455,
+      "learning_rate": 7.250687299531864e-06,
+      "loss": 1.0152,
+      "step": 4811
+    },
+    {
+      "epoch": 0.3709528214616096,
+      "grad_norm": 3.7596309185028076,
+      "learning_rate": 7.249572416834838e-06,
+      "loss": 0.9358,
+      "step": 4812
+    },
+    {
+      "epoch": 0.3710299105766266,
+      "grad_norm": 3.739588975906372,
+      "learning_rate": 7.2484573938886215e-06,
+      "loss": 0.9234,
+      "step": 4813
+    },
+    {
+      "epoch": 0.3711069996916435,
+      "grad_norm": 3.4759035110473633,
+      "learning_rate": 7.2473422307627304e-06,
+      "loss": 1.0292,
+      "step": 4814
+    },
+    {
+      "epoch": 0.3711840888066605,
+      "grad_norm": 3.456209659576416,
+      "learning_rate": 7.246226927526693e-06,
+      "loss": 1.0245,
+      "step": 4815
+    },
+    {
+      "epoch": 0.37126117792167745,
+      "grad_norm": 3.7224795818328857,
+      "learning_rate": 7.245111484250038e-06,
+      "loss": 0.9233,
+      "step": 4816
+    },
+    {
+      "epoch": 0.3713382670366944,
+      "grad_norm": 3.3582236766815186,
+      "learning_rate": 7.243995901002312e-06,
+      "loss": 1.0253,
+      "step": 4817
+    },
+    {
+      "epoch": 0.3714153561517114,
+      "grad_norm": 3.8575472831726074,
+      "learning_rate": 7.242880177853062e-06,
+      "loss": 1.0134,
+      "step": 4818
+    },
+    {
+      "epoch": 0.3714924452667283,
+      "grad_norm": 3.423126697540283,
+      "learning_rate": 7.241764314871848e-06,
+      "loss": 0.9355,
+      "step": 4819
+    },
+    {
+      "epoch": 0.3715695343817453,
+      "grad_norm": 3.5420002937316895,
+      "learning_rate": 7.240648312128242e-06,
+      "loss": 0.9646,
+      "step": 4820
+    },
+    {
+      "epoch": 0.37164662349676225,
+      "grad_norm": 3.214689254760742,
+      "learning_rate": 7.239532169691817e-06,
+      "loss": 0.9524,
+      "step": 4821
+    },
+    {
+      "epoch": 0.3717237126117792,
+      "grad_norm": 3.500775098800659,
+      "learning_rate": 7.23841588763216e-06,
+      "loss": 1.0106,
+      "step": 4822
+    },
+    {
+      "epoch": 0.3718008017267962,
+      "grad_norm": 3.302901029586792,
+      "learning_rate": 7.237299466018866e-06,
+      "loss": 0.8822,
+      "step": 4823
+    },
+    {
+      "epoch": 0.37187789084181316,
+      "grad_norm": 3.5195157527923584,
+      "learning_rate": 7.236182904921536e-06,
+      "loss": 1.0378,
+      "step": 4824
+    },
+    {
+      "epoch": 0.3719549799568301,
+      "grad_norm": 3.8534128665924072,
+      "learning_rate": 7.2350662044097854e-06,
+      "loss": 1.0471,
+      "step": 4825
+    },
+    {
+      "epoch": 0.37203206907184705,
+      "grad_norm": 3.6526763439178467,
+      "learning_rate": 7.233949364553232e-06,
+      "loss": 1.0511,
+      "step": 4826
+    },
+    {
+      "epoch": 0.372109158186864,
+      "grad_norm": 3.9592645168304443,
+      "learning_rate": 7.2328323854215044e-06,
+      "loss": 0.9967,
+      "step": 4827
+    },
+    {
+      "epoch": 0.372186247301881,
+      "grad_norm": 3.569300889968872,
+      "learning_rate": 7.231715267084243e-06,
+      "loss": 0.9936,
+      "step": 4828
+    },
+    {
+      "epoch": 0.37226333641689796,
+      "grad_norm": 3.6688168048858643,
+      "learning_rate": 7.2305980096110925e-06,
+      "loss": 1.1023,
+      "step": 4829
+    },
+    {
+      "epoch": 0.3723404255319149,
+      "grad_norm": 3.4594991207122803,
+      "learning_rate": 7.229480613071709e-06,
+      "loss": 0.9457,
+      "step": 4830
+    },
+    {
+      "epoch": 0.37241751464693185,
+      "grad_norm": 3.949270009994507,
+      "learning_rate": 7.228363077535756e-06,
+      "loss": 0.9467,
+      "step": 4831
+    },
+    {
+      "epoch": 0.3724946037619488,
+      "grad_norm": 3.7275776863098145,
+      "learning_rate": 7.227245403072905e-06,
+      "loss": 0.909,
+      "step": 4832
+    },
+    {
+      "epoch": 0.3725716928769658,
+      "grad_norm": 3.561305046081543,
+      "learning_rate": 7.2261275897528405e-06,
+      "loss": 1.1122,
+      "step": 4833
+    },
+    {
+      "epoch": 0.37264878199198276,
+      "grad_norm": 3.612966299057007,
+      "learning_rate": 7.225009637645248e-06,
+      "loss": 1.0967,
+      "step": 4834
+    },
+    {
+      "epoch": 0.3727258711069997,
+      "grad_norm": 3.454404354095459,
+      "learning_rate": 7.223891546819829e-06,
+      "loss": 1.0302,
+      "step": 4835
+    },
+    {
+      "epoch": 0.37280296022201664,
+      "grad_norm": 3.8778316974639893,
+      "learning_rate": 7.222773317346291e-06,
+      "loss": 1.0213,
+      "step": 4836
+    },
+    {
+      "epoch": 0.3728800493370336,
+      "grad_norm": 3.4762234687805176,
+      "learning_rate": 7.221654949294348e-06,
+      "loss": 0.9744,
+      "step": 4837
+    },
+    {
+      "epoch": 0.3729571384520506,
+      "grad_norm": 3.4373879432678223,
+      "learning_rate": 7.220536442733724e-06,
+      "loss": 0.8598,
+      "step": 4838
+    },
+    {
+      "epoch": 0.37303422756706756,
+      "grad_norm": 4.246344089508057,
+      "learning_rate": 7.219417797734155e-06,
+      "loss": 1.0942,
+      "step": 4839
+    },
+    {
+      "epoch": 0.37311131668208447,
+      "grad_norm": 3.5066323280334473,
+      "learning_rate": 7.2182990143653795e-06,
+      "loss": 1.0291,
+      "step": 4840
+    },
+    {
+      "epoch": 0.37318840579710144,
+      "grad_norm": 3.613722562789917,
+      "learning_rate": 7.217180092697152e-06,
+      "loss": 0.9715,
+      "step": 4841
+    },
+    {
+      "epoch": 0.3732654949121184,
+      "grad_norm": 3.9713635444641113,
+      "learning_rate": 7.216061032799225e-06,
+      "loss": 1.0856,
+      "step": 4842
+    },
+    {
+      "epoch": 0.3733425840271354,
+      "grad_norm": 3.5957889556884766,
+      "learning_rate": 7.2149418347413724e-06,
+      "loss": 0.9641,
+      "step": 4843
+    },
+    {
+      "epoch": 0.37341967314215235,
+      "grad_norm": 3.5281994342803955,
+      "learning_rate": 7.213822498593368e-06,
+      "loss": 1.0026,
+      "step": 4844
+    },
+    {
+      "epoch": 0.37349676225716927,
+      "grad_norm": 4.489894390106201,
+      "learning_rate": 7.212703024424995e-06,
+      "loss": 1.061,
+      "step": 4845
+    },
+    {
+      "epoch": 0.37357385137218624,
+      "grad_norm": 4.678754806518555,
+      "learning_rate": 7.211583412306049e-06,
+      "loss": 0.9971,
+      "step": 4846
+    },
+    {
+      "epoch": 0.3736509404872032,
+      "grad_norm": 4.363256454467773,
+      "learning_rate": 7.2104636623063315e-06,
+      "loss": 1.0027,
+      "step": 4847
+    },
+    {
+      "epoch": 0.3737280296022202,
+      "grad_norm": 3.5184109210968018,
+      "learning_rate": 7.209343774495652e-06,
+      "loss": 1.0354,
+      "step": 4848
+    },
+    {
+      "epoch": 0.37380511871723715,
+      "grad_norm": 3.521500587463379,
+      "learning_rate": 7.208223748943832e-06,
+      "loss": 0.9084,
+      "step": 4849
+    },
+    {
+      "epoch": 0.37388220783225407,
+      "grad_norm": 3.56219744682312,
+      "learning_rate": 7.207103585720697e-06,
+      "loss": 1.0089,
+      "step": 4850
+    },
+    {
+      "epoch": 0.37395929694727104,
+      "grad_norm": 3.738111734390259,
+      "learning_rate": 7.2059832848960855e-06,
+      "loss": 1.0386,
+      "step": 4851
+    },
+    {
+      "epoch": 0.374036386062288,
+      "grad_norm": 4.135499954223633,
+      "learning_rate": 7.204862846539841e-06,
+      "loss": 1.0668,
+      "step": 4852
+    },
+    {
+      "epoch": 0.374113475177305,
+      "grad_norm": 3.822765827178955,
+      "learning_rate": 7.2037422707218165e-06,
+      "loss": 1.0167,
+      "step": 4853
+    },
+    {
+      "epoch": 0.37419056429232195,
+      "grad_norm": 3.540203094482422,
+      "learning_rate": 7.202621557511874e-06,
+      "loss": 0.9056,
+      "step": 4854
+    },
+    {
+      "epoch": 0.37426765340733886,
+      "grad_norm": 3.5464091300964355,
+      "learning_rate": 7.201500706979886e-06,
+      "loss": 0.9215,
+      "step": 4855
+    },
+    {
+      "epoch": 0.37434474252235583,
+      "grad_norm": 3.7455790042877197,
+      "learning_rate": 7.20037971919573e-06,
+      "loss": 0.9164,
+      "step": 4856
+    },
+    {
+      "epoch": 0.3744218316373728,
+      "grad_norm": 3.647853374481201,
+      "learning_rate": 7.199258594229297e-06,
+      "loss": 1.0188,
+      "step": 4857
+    },
+    {
+      "epoch": 0.3744989207523898,
+      "grad_norm": 3.654245615005493,
+      "learning_rate": 7.198137332150479e-06,
+      "loss": 1.0136,
+      "step": 4858
+    },
+    {
+      "epoch": 0.37457600986740675,
+      "grad_norm": 3.327587127685547,
+      "learning_rate": 7.197015933029184e-06,
+      "loss": 0.9465,
+      "step": 4859
+    },
+    {
+      "epoch": 0.37465309898242366,
+      "grad_norm": 4.306604385375977,
+      "learning_rate": 7.195894396935324e-06,
+      "loss": 1.0239,
+      "step": 4860
+    },
+    {
+      "epoch": 0.37473018809744063,
+      "grad_norm": 4.178218841552734,
+      "learning_rate": 7.194772723938819e-06,
+      "loss": 1.0016,
+      "step": 4861
+    },
+    {
+      "epoch": 0.3748072772124576,
+      "grad_norm": 4.198592185974121,
+      "learning_rate": 7.193650914109603e-06,
+      "loss": 1.0389,
+      "step": 4862
+    },
+    {
+      "epoch": 0.37488436632747457,
+      "grad_norm": 3.9807186126708984,
+      "learning_rate": 7.192528967517615e-06,
+      "loss": 0.9948,
+      "step": 4863
+    },
+    {
+      "epoch": 0.37496145544249154,
+      "grad_norm": 3.558039426803589,
+      "learning_rate": 7.1914068842328e-06,
+      "loss": 1.0135,
+      "step": 4864
+    },
+    {
+      "epoch": 0.37503854455750846,
+      "grad_norm": 3.5846774578094482,
+      "learning_rate": 7.190284664325116e-06,
+      "loss": 1.0469,
+      "step": 4865
+    },
+    {
+      "epoch": 0.37511563367252543,
+      "grad_norm": 3.2694685459136963,
+      "learning_rate": 7.189162307864525e-06,
+      "loss": 0.9575,
+      "step": 4866
+    },
+    {
+      "epoch": 0.3751927227875424,
+      "grad_norm": 3.3585803508758545,
+      "learning_rate": 7.188039814921004e-06,
+      "loss": 0.9909,
+      "step": 4867
+    },
+    {
+      "epoch": 0.37526981190255937,
+      "grad_norm": 3.606834650039673,
+      "learning_rate": 7.186917185564534e-06,
+      "loss": 0.999,
+      "step": 4868
+    },
+    {
+      "epoch": 0.37534690101757634,
+      "grad_norm": 3.2820441722869873,
+      "learning_rate": 7.185794419865102e-06,
+      "loss": 0.9831,
+      "step": 4869
+    },
+    {
+      "epoch": 0.37542399013259325,
+      "grad_norm": 3.559875011444092,
+      "learning_rate": 7.184671517892707e-06,
+      "loss": 0.9819,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3755010792476102,
+      "grad_norm": 3.5334460735321045,
+      "learning_rate": 7.183548479717361e-06,
+      "loss": 1.031,
+      "step": 4871
+    },
+    {
+      "epoch": 0.3755781683626272,
+      "grad_norm": 3.8719050884246826,
+      "learning_rate": 7.1824253054090735e-06,
+      "loss": 0.9976,
+      "step": 4872
+    },
+    {
+      "epoch": 0.37565525747764417,
+      "grad_norm": 3.7610104084014893,
+      "learning_rate": 7.1813019950378724e-06,
+      "loss": 1.1075,
+      "step": 4873
+    },
+    {
+      "epoch": 0.37573234659266114,
+      "grad_norm": 3.8099474906921387,
+      "learning_rate": 7.1801785486737884e-06,
+      "loss": 1.114,
+      "step": 4874
+    },
+    {
+      "epoch": 0.37580943570767805,
+      "grad_norm": 3.353347063064575,
+      "learning_rate": 7.1790549663868644e-06,
+      "loss": 0.9258,
+      "step": 4875
+    },
+    {
+      "epoch": 0.375886524822695,
+      "grad_norm": 3.220855236053467,
+      "learning_rate": 7.1779312482471475e-06,
+      "loss": 0.9835,
+      "step": 4876
+    },
+    {
+      "epoch": 0.375963613937712,
+      "grad_norm": 3.72371244430542,
+      "learning_rate": 7.176807394324697e-06,
+      "loss": 1.0217,
+      "step": 4877
+    },
+    {
+      "epoch": 0.37604070305272896,
+      "grad_norm": 3.4218595027923584,
+      "learning_rate": 7.1756834046895815e-06,
+      "loss": 0.9421,
+      "step": 4878
+    },
+    {
+      "epoch": 0.37611779216774593,
+      "grad_norm": 3.7744340896606445,
+      "learning_rate": 7.174559279411872e-06,
+      "loss": 0.9703,
+      "step": 4879
+    },
+    {
+      "epoch": 0.37619488128276285,
+      "grad_norm": 3.9162957668304443,
+      "learning_rate": 7.173435018561654e-06,
+      "loss": 1.1117,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3762719703977798,
+      "grad_norm": 3.7875723838806152,
+      "learning_rate": 7.17231062220902e-06,
+      "loss": 1.0843,
+      "step": 4881
+    },
+    {
+      "epoch": 0.3763490595127968,
+      "grad_norm": 3.8338139057159424,
+      "learning_rate": 7.171186090424069e-06,
+      "loss": 1.0342,
+      "step": 4882
+    },
+    {
+      "epoch": 0.37642614862781376,
+      "grad_norm": 3.3888375759124756,
+      "learning_rate": 7.170061423276911e-06,
+      "loss": 1.0047,
+      "step": 4883
+    },
+    {
+      "epoch": 0.37650323774283073,
+      "grad_norm": 3.7450151443481445,
+      "learning_rate": 7.168936620837661e-06,
+      "loss": 0.9721,
+      "step": 4884
+    },
+    {
+      "epoch": 0.37658032685784765,
+      "grad_norm": 3.4551503658294678,
+      "learning_rate": 7.167811683176446e-06,
+      "loss": 0.8549,
+      "step": 4885
+    },
+    {
+      "epoch": 0.3766574159728646,
+      "grad_norm": 3.826219320297241,
+      "learning_rate": 7.166686610363399e-06,
+      "loss": 1.0317,
+      "step": 4886
+    },
+    {
+      "epoch": 0.3767345050878816,
+      "grad_norm": 3.661606788635254,
+      "learning_rate": 7.165561402468666e-06,
+      "loss": 0.9179,
+      "step": 4887
+    },
+    {
+      "epoch": 0.37681159420289856,
+      "grad_norm": 3.9215991497039795,
+      "learning_rate": 7.164436059562393e-06,
+      "loss": 1.1039,
+      "step": 4888
+    },
+    {
+      "epoch": 0.37688868331791553,
+      "grad_norm": 3.983035087585449,
+      "learning_rate": 7.163310581714744e-06,
+      "loss": 0.9932,
+      "step": 4889
+    },
+    {
+      "epoch": 0.37696577243293244,
+      "grad_norm": 3.3933942317962646,
+      "learning_rate": 7.162184968995882e-06,
+      "loss": 0.9746,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3770428615479494,
+      "grad_norm": 3.4548404216766357,
+      "learning_rate": 7.161059221475985e-06,
+      "loss": 0.8202,
+      "step": 4891
+    },
+    {
+      "epoch": 0.3771199506629664,
+      "grad_norm": 3.530745267868042,
+      "learning_rate": 7.15993333922524e-06,
+      "loss": 0.8811,
+      "step": 4892
+    },
+    {
+      "epoch": 0.37719703977798336,
+      "grad_norm": 3.4644525051116943,
+      "learning_rate": 7.158807322313837e-06,
+      "loss": 1.0706,
+      "step": 4893
+    },
+    {
+      "epoch": 0.3772741288930003,
+      "grad_norm": 3.4556727409362793,
+      "learning_rate": 7.157681170811979e-06,
+      "loss": 0.9343,
+      "step": 4894
+    },
+    {
+      "epoch": 0.37735121800801724,
+      "grad_norm": 3.663114547729492,
+      "learning_rate": 7.156554884789874e-06,
+      "loss": 1.0052,
+      "step": 4895
+    },
+    {
+      "epoch": 0.3774283071230342,
+      "grad_norm": 3.9174418449401855,
+      "learning_rate": 7.155428464317741e-06,
+      "loss": 0.9232,
+      "step": 4896
+    },
+    {
+      "epoch": 0.3775053962380512,
+      "grad_norm": 3.8079843521118164,
+      "learning_rate": 7.1543019094658074e-06,
+      "loss": 0.9495,
+      "step": 4897
+    },
+    {
+      "epoch": 0.37758248535306815,
+      "grad_norm": 3.580113649368286,
+      "learning_rate": 7.153175220304305e-06,
+      "loss": 1.0305,
+      "step": 4898
+    },
+    {
+      "epoch": 0.3776595744680851,
+      "grad_norm": 3.540668487548828,
+      "learning_rate": 7.152048396903479e-06,
+      "loss": 0.959,
+      "step": 4899
+    },
+    {
+      "epoch": 0.37773666358310204,
+      "grad_norm": 3.402894973754883,
+      "learning_rate": 7.150921439333584e-06,
+      "loss": 0.9978,
+      "step": 4900
+    },
+    {
+      "epoch": 0.377813752698119,
+      "grad_norm": 3.550427198410034,
+      "learning_rate": 7.149794347664876e-06,
+      "loss": 1.0077,
+      "step": 4901
+    },
+    {
+      "epoch": 0.377890841813136,
+      "grad_norm": 3.6682233810424805,
+      "learning_rate": 7.148667121967625e-06,
+      "loss": 0.9905,
+      "step": 4902
+    },
+    {
+      "epoch": 0.37796793092815295,
+      "grad_norm": 3.8685266971588135,
+      "learning_rate": 7.147539762312107e-06,
+      "loss": 1.0975,
+      "step": 4903
+    },
+    {
+      "epoch": 0.3780450200431699,
+      "grad_norm": 3.448533773422241,
+      "learning_rate": 7.146412268768605e-06,
+      "loss": 0.857,
+      "step": 4904
+    },
+    {
+      "epoch": 0.37812210915818684,
+      "grad_norm": 3.861265182495117,
+      "learning_rate": 7.145284641407418e-06,
+      "loss": 1.0235,
+      "step": 4905
+    },
+    {
+      "epoch": 0.3781991982732038,
+      "grad_norm": 3.419316053390503,
+      "learning_rate": 7.144156880298843e-06,
+      "loss": 1.0753,
+      "step": 4906
+    },
+    {
+      "epoch": 0.3782762873882208,
+      "grad_norm": 3.76204776763916,
+      "learning_rate": 7.143028985513191e-06,
+      "loss": 0.9979,
+      "step": 4907
+    },
+    {
+      "epoch": 0.37835337650323775,
+      "grad_norm": 3.591517686843872,
+      "learning_rate": 7.141900957120781e-06,
+      "loss": 1.0564,
+      "step": 4908
+    },
+    {
+      "epoch": 0.3784304656182547,
+      "grad_norm": 3.76042103767395,
+      "learning_rate": 7.140772795191939e-06,
+      "loss": 1.0371,
+      "step": 4909
+    },
+    {
+      "epoch": 0.3785075547332717,
+      "grad_norm": 4.168672561645508,
+      "learning_rate": 7.139644499797e-06,
+      "loss": 1.0409,
+      "step": 4910
+    },
+    {
+      "epoch": 0.3785846438482886,
+      "grad_norm": 3.331540822982788,
+      "learning_rate": 7.13851607100631e-06,
+      "loss": 0.9319,
+      "step": 4911
+    },
+    {
+      "epoch": 0.3786617329633056,
+      "grad_norm": 3.5602786540985107,
+      "learning_rate": 7.137387508890218e-06,
+      "loss": 0.8784,
+      "step": 4912
+    },
+    {
+      "epoch": 0.37873882207832255,
+      "grad_norm": 3.6382076740264893,
+      "learning_rate": 7.136258813519085e-06,
+      "loss": 1.0411,
+      "step": 4913
+    },
+    {
+      "epoch": 0.3788159111933395,
+      "grad_norm": 3.845329761505127,
+      "learning_rate": 7.135129984963277e-06,
+      "loss": 1.0158,
+      "step": 4914
+    },
+    {
+      "epoch": 0.3788930003083565,
+      "grad_norm": 3.546248197555542,
+      "learning_rate": 7.134001023293173e-06,
+      "loss": 1.044,
+      "step": 4915
+    },
+    {
+      "epoch": 0.3789700894233734,
+      "grad_norm": 3.4055960178375244,
+      "learning_rate": 7.132871928579159e-06,
+      "loss": 0.9337,
+      "step": 4916
+    },
+    {
+      "epoch": 0.3790471785383904,
+      "grad_norm": 3.8326213359832764,
+      "learning_rate": 7.131742700891626e-06,
+      "loss": 1.0974,
+      "step": 4917
+    },
+    {
+      "epoch": 0.37912426765340734,
+      "grad_norm": 3.4271841049194336,
+      "learning_rate": 7.130613340300976e-06,
+      "loss": 0.9532,
+      "step": 4918
+    },
+    {
+      "epoch": 0.3792013567684243,
+      "grad_norm": 3.5200610160827637,
+      "learning_rate": 7.1294838468776195e-06,
+      "loss": 1.0648,
+      "step": 4919
+    },
+    {
+      "epoch": 0.3792784458834413,
+      "grad_norm": 3.484342098236084,
+      "learning_rate": 7.128354220691973e-06,
+      "loss": 0.8582,
+      "step": 4920
+    },
+    {
+      "epoch": 0.3793555349984582,
+      "grad_norm": 3.4747977256774902,
+      "learning_rate": 7.127224461814465e-06,
+      "loss": 0.9661,
+      "step": 4921
+    },
+    {
+      "epoch": 0.37943262411347517,
+      "grad_norm": 3.514315366744995,
+      "learning_rate": 7.126094570315527e-06,
+      "loss": 1.0848,
+      "step": 4922
+    },
+    {
+      "epoch": 0.37950971322849214,
+      "grad_norm": 3.6075520515441895,
+      "learning_rate": 7.124964546265606e-06,
+      "loss": 1.0381,
+      "step": 4923
+    },
+    {
+      "epoch": 0.3795868023435091,
+      "grad_norm": 3.4969682693481445,
+      "learning_rate": 7.1238343897351505e-06,
+      "loss": 0.9393,
+      "step": 4924
+    },
+    {
+      "epoch": 0.3796638914585261,
+      "grad_norm": 3.8373172283172607,
+      "learning_rate": 7.12270410079462e-06,
+      "loss": 1.1534,
+      "step": 4925
+    },
+    {
+      "epoch": 0.379740980573543,
+      "grad_norm": 4.133366584777832,
+      "learning_rate": 7.121573679514484e-06,
+      "loss": 1.1071,
+      "step": 4926
+    },
+    {
+      "epoch": 0.37981806968855997,
+      "grad_norm": 3.4957773685455322,
+      "learning_rate": 7.120443125965215e-06,
+      "loss": 0.9435,
+      "step": 4927
+    },
+    {
+      "epoch": 0.37989515880357694,
+      "grad_norm": 3.298091411590576,
+      "learning_rate": 7.1193124402172995e-06,
+      "loss": 0.8858,
+      "step": 4928
+    },
+    {
+      "epoch": 0.3799722479185939,
+      "grad_norm": 3.720797061920166,
+      "learning_rate": 7.118181622341232e-06,
+      "loss": 1.0036,
+      "step": 4929
+    },
+    {
+      "epoch": 0.3800493370336109,
+      "grad_norm": 3.520009994506836,
+      "learning_rate": 7.117050672407507e-06,
+      "loss": 0.9223,
+      "step": 4930
+    },
+    {
+      "epoch": 0.3801264261486278,
+      "grad_norm": 3.883283853530884,
+      "learning_rate": 7.115919590486638e-06,
+      "loss": 0.8866,
+      "step": 4931
+    },
+    {
+      "epoch": 0.38020351526364476,
+      "grad_norm": 3.637392520904541,
+      "learning_rate": 7.114788376649143e-06,
+      "loss": 0.9867,
+      "step": 4932
+    },
+    {
+      "epoch": 0.38028060437866174,
+      "grad_norm": 3.6200270652770996,
+      "learning_rate": 7.113657030965544e-06,
+      "loss": 0.9339,
+      "step": 4933
+    },
+    {
+      "epoch": 0.3803576934936787,
+      "grad_norm": 3.7726197242736816,
+      "learning_rate": 7.1125255535063766e-06,
+      "loss": 0.9902,
+      "step": 4934
+    },
+    {
+      "epoch": 0.3804347826086957,
+      "grad_norm": 3.6421115398406982,
+      "learning_rate": 7.111393944342182e-06,
+      "loss": 1.0895,
+      "step": 4935
+    },
+    {
+      "epoch": 0.3805118717237126,
+      "grad_norm": 3.6255416870117188,
+      "learning_rate": 7.110262203543509e-06,
+      "loss": 0.9478,
+      "step": 4936
+    },
+    {
+      "epoch": 0.38058896083872956,
+      "grad_norm": 3.6377484798431396,
+      "learning_rate": 7.10913033118092e-06,
+      "loss": 1.0047,
+      "step": 4937
+    },
+    {
+      "epoch": 0.38066604995374653,
+      "grad_norm": 3.65698504447937,
+      "learning_rate": 7.107998327324975e-06,
+      "loss": 1.0363,
+      "step": 4938
+    },
+    {
+      "epoch": 0.3807431390687635,
+      "grad_norm": 3.5486626625061035,
+      "learning_rate": 7.106866192046254e-06,
+      "loss": 1.0382,
+      "step": 4939
+    },
+    {
+      "epoch": 0.3808202281837805,
+      "grad_norm": 3.589175224304199,
+      "learning_rate": 7.105733925415336e-06,
+      "loss": 0.9718,
+      "step": 4940
+    },
+    {
+      "epoch": 0.3808973172987974,
+      "grad_norm": 3.605515956878662,
+      "learning_rate": 7.104601527502815e-06,
+      "loss": 0.9097,
+      "step": 4941
+    },
+    {
+      "epoch": 0.38097440641381436,
+      "grad_norm": 3.868885040283203,
+      "learning_rate": 7.103468998379288e-06,
+      "loss": 1.0441,
+      "step": 4942
+    },
+    {
+      "epoch": 0.38105149552883133,
+      "grad_norm": 3.621760129928589,
+      "learning_rate": 7.102336338115363e-06,
+      "loss": 1.0506,
+      "step": 4943
+    },
+    {
+      "epoch": 0.3811285846438483,
+      "grad_norm": 4.027126789093018,
+      "learning_rate": 7.101203546781655e-06,
+      "loss": 1.0522,
+      "step": 4944
+    },
+    {
+      "epoch": 0.38120567375886527,
+      "grad_norm": 3.5115973949432373,
+      "learning_rate": 7.1000706244487896e-06,
+      "loss": 0.9613,
+      "step": 4945
+    },
+    {
+      "epoch": 0.3812827628738822,
+      "grad_norm": 3.698106050491333,
+      "learning_rate": 7.098937571187397e-06,
+      "loss": 0.9886,
+      "step": 4946
+    },
+    {
+      "epoch": 0.38135985198889916,
+      "grad_norm": 3.994877576828003,
+      "learning_rate": 7.097804387068117e-06,
+      "loss": 1.1597,
+      "step": 4947
+    },
+    {
+      "epoch": 0.3814369411039161,
+      "grad_norm": 3.984495162963867,
+      "learning_rate": 7.0966710721616e-06,
+      "loss": 0.9572,
+      "step": 4948
+    },
+    {
+      "epoch": 0.3815140302189331,
+      "grad_norm": 3.9467110633850098,
+      "learning_rate": 7.095537626538498e-06,
+      "loss": 1.0071,
+      "step": 4949
+    },
+    {
+      "epoch": 0.38159111933395007,
+      "grad_norm": 3.824596643447876,
+      "learning_rate": 7.09440405026948e-06,
+      "loss": 1.0244,
+      "step": 4950
+    },
+    {
+      "epoch": 0.381668208448967,
+      "grad_norm": 3.812662124633789,
+      "learning_rate": 7.093270343425216e-06,
+      "loss": 1.0646,
+      "step": 4951
+    },
+    {
+      "epoch": 0.38174529756398395,
+      "grad_norm": 3.604254961013794,
+      "learning_rate": 7.092136506076387e-06,
+      "loss": 1.0749,
+      "step": 4952
+    },
+    {
+      "epoch": 0.3818223866790009,
+      "grad_norm": 3.493272542953491,
+      "learning_rate": 7.091002538293683e-06,
+      "loss": 0.9379,
+      "step": 4953
+    },
+    {
+      "epoch": 0.3818994757940179,
+      "grad_norm": 3.4740121364593506,
+      "learning_rate": 7.0898684401478e-06,
+      "loss": 0.9795,
+      "step": 4954
+    },
+    {
+      "epoch": 0.38197656490903487,
+      "grad_norm": 4.244378566741943,
+      "learning_rate": 7.088734211709443e-06,
+      "loss": 0.9794,
+      "step": 4955
+    },
+    {
+      "epoch": 0.3820536540240518,
+      "grad_norm": 3.6403074264526367,
+      "learning_rate": 7.087599853049327e-06,
+      "loss": 1.0321,
+      "step": 4956
+    },
+    {
+      "epoch": 0.38213074313906875,
+      "grad_norm": 3.5275707244873047,
+      "learning_rate": 7.086465364238171e-06,
+      "loss": 0.971,
+      "step": 4957
+    },
+    {
+      "epoch": 0.3822078322540857,
+      "grad_norm": 4.26153564453125,
+      "learning_rate": 7.085330745346706e-06,
+      "loss": 1.0188,
+      "step": 4958
+    },
+    {
+      "epoch": 0.3822849213691027,
+      "grad_norm": 3.5671985149383545,
+      "learning_rate": 7.08419599644567e-06,
+      "loss": 1.0241,
+      "step": 4959
+    },
+    {
+      "epoch": 0.38236201048411966,
+      "grad_norm": 3.462334632873535,
+      "learning_rate": 7.083061117605806e-06,
+      "loss": 0.933,
+      "step": 4960
+    },
+    {
+      "epoch": 0.3824390995991366,
+      "grad_norm": 3.856916904449463,
+      "learning_rate": 7.081926108897872e-06,
+      "loss": 0.9347,
+      "step": 4961
+    },
+    {
+      "epoch": 0.38251618871415355,
+      "grad_norm": 3.425292730331421,
+      "learning_rate": 7.080790970392626e-06,
+      "loss": 0.9056,
+      "step": 4962
+    },
+    {
+      "epoch": 0.3825932778291705,
+      "grad_norm": 3.7738280296325684,
+      "learning_rate": 7.07965570216084e-06,
+      "loss": 0.904,
+      "step": 4963
+    },
+    {
+      "epoch": 0.3826703669441875,
+      "grad_norm": 3.893852710723877,
+      "learning_rate": 7.078520304273293e-06,
+      "loss": 0.9253,
+      "step": 4964
+    },
+    {
+      "epoch": 0.38274745605920446,
+      "grad_norm": 3.1743626594543457,
+      "learning_rate": 7.077384776800767e-06,
+      "loss": 0.9373,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3828245451742214,
+      "grad_norm": 3.271000385284424,
+      "learning_rate": 7.076249119814062e-06,
+      "loss": 0.9536,
+      "step": 4966
+    },
+    {
+      "epoch": 0.38290163428923835,
+      "grad_norm": 3.6323864459991455,
+      "learning_rate": 7.075113333383976e-06,
+      "loss": 1.0153,
+      "step": 4967
+    },
+    {
+      "epoch": 0.3829787234042553,
+      "grad_norm": 4.131968975067139,
+      "learning_rate": 7.073977417581321e-06,
+      "loss": 1.0542,
+      "step": 4968
+    },
+    {
+      "epoch": 0.3830558125192723,
+      "grad_norm": 3.279594898223877,
+      "learning_rate": 7.072841372476918e-06,
+      "loss": 0.9579,
+      "step": 4969
+    },
+    {
+      "epoch": 0.38313290163428926,
+      "grad_norm": 3.7253003120422363,
+      "learning_rate": 7.071705198141588e-06,
+      "loss": 0.9364,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3832099907493062,
+      "grad_norm": 3.599295139312744,
+      "learning_rate": 7.070568894646171e-06,
+      "loss": 0.9531,
+      "step": 4971
+    },
+    {
+      "epoch": 0.38328707986432314,
+      "grad_norm": 3.4222543239593506,
+      "learning_rate": 7.0694324620615054e-06,
+      "loss": 1.0856,
+      "step": 4972
+    },
+    {
+      "epoch": 0.3833641689793401,
+      "grad_norm": 3.6698365211486816,
+      "learning_rate": 7.0682959004584436e-06,
+      "loss": 0.9242,
+      "step": 4973
+    },
+    {
+      "epoch": 0.3834412580943571,
+      "grad_norm": 3.4762206077575684,
+      "learning_rate": 7.067159209907845e-06,
+      "loss": 0.933,
+      "step": 4974
+    },
+    {
+      "epoch": 0.38351834720937406,
+      "grad_norm": 3.3484156131744385,
+      "learning_rate": 7.066022390480576e-06,
+      "loss": 0.9227,
+      "step": 4975
+    },
+    {
+      "epoch": 0.38359543632439097,
+      "grad_norm": 4.061807155609131,
+      "learning_rate": 7.06488544224751e-06,
+      "loss": 1.0448,
+      "step": 4976
+    },
+    {
+      "epoch": 0.38367252543940794,
+      "grad_norm": 3.4074602127075195,
+      "learning_rate": 7.0637483652795325e-06,
+      "loss": 0.8229,
+      "step": 4977
+    },
+    {
+      "epoch": 0.3837496145544249,
+      "grad_norm": 3.6092684268951416,
+      "learning_rate": 7.062611159647532e-06,
+      "loss": 1.0388,
+      "step": 4978
+    },
+    {
+      "epoch": 0.3838267036694419,
+      "grad_norm": 3.7201006412506104,
+      "learning_rate": 7.061473825422408e-06,
+      "loss": 0.9342,
+      "step": 4979
+    },
+    {
+      "epoch": 0.38390379278445885,
+      "grad_norm": 3.228909969329834,
+      "learning_rate": 7.060336362675069e-06,
+      "loss": 0.9043,
+      "step": 4980
+    },
+    {
+      "epoch": 0.38398088189947577,
+      "grad_norm": 3.3236916065216064,
+      "learning_rate": 7.059198771476426e-06,
+      "loss": 0.9576,
+      "step": 4981
+    },
+    {
+      "epoch": 0.38405797101449274,
+      "grad_norm": 3.616847038269043,
+      "learning_rate": 7.0580610518974065e-06,
+      "loss": 1.0305,
+      "step": 4982
+    },
+    {
+      "epoch": 0.3841350601295097,
+      "grad_norm": 3.782071590423584,
+      "learning_rate": 7.05692320400894e-06,
+      "loss": 0.9386,
+      "step": 4983
+    },
+    {
+      "epoch": 0.3842121492445267,
+      "grad_norm": 3.3913662433624268,
+      "learning_rate": 7.055785227881963e-06,
+      "loss": 0.9801,
+      "step": 4984
+    },
+    {
+      "epoch": 0.38428923835954365,
+      "grad_norm": 3.7937633991241455,
+      "learning_rate": 7.054647123587426e-06,
+      "loss": 1.0372,
+      "step": 4985
+    },
+    {
+      "epoch": 0.38436632747456057,
+      "grad_norm": 3.505120038986206,
+      "learning_rate": 7.05350889119628e-06,
+      "loss": 0.9769,
+      "step": 4986
+    },
+    {
+      "epoch": 0.38444341658957754,
+      "grad_norm": 3.569885015487671,
+      "learning_rate": 7.052370530779491e-06,
+      "loss": 1.138,
+      "step": 4987
+    },
+    {
+      "epoch": 0.3845205057045945,
+      "grad_norm": 4.187367916107178,
+      "learning_rate": 7.051232042408028e-06,
+      "loss": 0.9448,
+      "step": 4988
+    },
+    {
+      "epoch": 0.3845975948196115,
+      "grad_norm": 3.4536428451538086,
+      "learning_rate": 7.050093426152872e-06,
+      "loss": 0.8804,
+      "step": 4989
+    },
+    {
+      "epoch": 0.38467468393462845,
+      "grad_norm": 3.314223289489746,
+      "learning_rate": 7.0489546820850086e-06,
+      "loss": 0.9758,
+      "step": 4990
+    },
+    {
+      "epoch": 0.38475177304964536,
+      "grad_norm": 3.603670120239258,
+      "learning_rate": 7.047815810275432e-06,
+      "loss": 0.9632,
+      "step": 4991
+    },
+    {
+      "epoch": 0.38482886216466233,
+      "grad_norm": 3.7399637699127197,
+      "learning_rate": 7.046676810795146e-06,
+      "loss": 1.1052,
+      "step": 4992
+    },
+    {
+      "epoch": 0.3849059512796793,
+      "grad_norm": 3.740985870361328,
+      "learning_rate": 7.045537683715161e-06,
+      "loss": 1.0086,
+      "step": 4993
+    },
+    {
+      "epoch": 0.3849830403946963,
+      "grad_norm": 3.6011998653411865,
+      "learning_rate": 7.044398429106495e-06,
+      "loss": 1.0373,
+      "step": 4994
+    },
+    {
+      "epoch": 0.38506012950971324,
+      "grad_norm": 3.508432388305664,
+      "learning_rate": 7.043259047040175e-06,
+      "loss": 1.0279,
+      "step": 4995
+    },
+    {
+      "epoch": 0.3851372186247302,
+      "grad_norm": 3.528916358947754,
+      "learning_rate": 7.042119537587237e-06,
+      "loss": 1.0366,
+      "step": 4996
+    },
+    {
+      "epoch": 0.38521430773974713,
+      "grad_norm": 3.8693268299102783,
+      "learning_rate": 7.04097990081872e-06,
+      "loss": 0.9748,
+      "step": 4997
+    },
+    {
+      "epoch": 0.3852913968547641,
+      "grad_norm": 3.851828098297119,
+      "learning_rate": 7.039840136805679e-06,
+      "loss": 1.0828,
+      "step": 4998
+    },
+    {
+      "epoch": 0.38536848596978107,
+      "grad_norm": 3.7920031547546387,
+      "learning_rate": 7.038700245619169e-06,
+      "loss": 1.1299,
+      "step": 4999
+    },
+    {
+      "epoch": 0.38544557508479804,
+      "grad_norm": 4.139936923980713,
+      "learning_rate": 7.037560227330258e-06,
+      "loss": 0.9281,
+      "step": 5000
+    },
+    {
+      "epoch": 0.385522664199815,
+      "grad_norm": 3.6004490852355957,
+      "learning_rate": 7.0364200820100194e-06,
+      "loss": 1.0097,
+      "step": 5001
+    },
+    {
+      "epoch": 0.38559975331483193,
+      "grad_norm": 3.911167621612549,
+      "learning_rate": 7.035279809729535e-06,
+      "loss": 1.0057,
+      "step": 5002
+    },
+    {
+      "epoch": 0.3856768424298489,
+      "grad_norm": 3.6797120571136475,
+      "learning_rate": 7.0341394105598944e-06,
+      "loss": 0.9339,
+      "step": 5003
+    },
+    {
+      "epoch": 0.38575393154486587,
+      "grad_norm": 3.733839750289917,
+      "learning_rate": 7.0329988845721996e-06,
+      "loss": 0.9404,
+      "step": 5004
+    },
+    {
+      "epoch": 0.38583102065988284,
+      "grad_norm": 3.474809169769287,
+      "learning_rate": 7.0318582318375514e-06,
+      "loss": 1.0068,
+      "step": 5005
+    },
+    {
+      "epoch": 0.3859081097748998,
+      "grad_norm": 4.166741371154785,
+      "learning_rate": 7.030717452427066e-06,
+      "loss": 1.0622,
+      "step": 5006
+    },
+    {
+      "epoch": 0.3859851988899167,
+      "grad_norm": 3.7879574298858643,
+      "learning_rate": 7.0295765464118645e-06,
+      "loss": 0.9869,
+      "step": 5007
+    },
+    {
+      "epoch": 0.3860622880049337,
+      "grad_norm": 3.560980796813965,
+      "learning_rate": 7.028435513863078e-06,
+      "loss": 1.022,
+      "step": 5008
+    },
+    {
+      "epoch": 0.38613937711995067,
+      "grad_norm": 3.5441577434539795,
+      "learning_rate": 7.027294354851842e-06,
+      "loss": 1.014,
+      "step": 5009
+    },
+    {
+      "epoch": 0.38621646623496764,
+      "grad_norm": 3.795830488204956,
+      "learning_rate": 7.0261530694493015e-06,
+      "loss": 0.8879,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3862935553499846,
+      "grad_norm": 3.435690402984619,
+      "learning_rate": 7.025011657726613e-06,
+      "loss": 0.8957,
+      "step": 5011
+    },
+    {
+      "epoch": 0.3863706444650015,
+      "grad_norm": 3.815304756164551,
+      "learning_rate": 7.0238701197549345e-06,
+      "loss": 1.0519,
+      "step": 5012
+    },
+    {
+      "epoch": 0.3864477335800185,
+      "grad_norm": 3.4580211639404297,
+      "learning_rate": 7.0227284556054355e-06,
+      "loss": 0.9664,
+      "step": 5013
+    },
+    {
+      "epoch": 0.38652482269503546,
+      "grad_norm": 3.8483505249023438,
+      "learning_rate": 7.0215866653492936e-06,
+      "loss": 0.9913,
+      "step": 5014
+    },
+    {
+      "epoch": 0.38660191181005243,
+      "grad_norm": 3.6694788932800293,
+      "learning_rate": 7.020444749057693e-06,
+      "loss": 0.9433,
+      "step": 5015
+    },
+    {
+      "epoch": 0.3866790009250694,
+      "grad_norm": 3.6327717304229736,
+      "learning_rate": 7.019302706801826e-06,
+      "loss": 0.9477,
+      "step": 5016
+    },
+    {
+      "epoch": 0.3867560900400863,
+      "grad_norm": 3.4883623123168945,
+      "learning_rate": 7.018160538652895e-06,
+      "loss": 0.9981,
+      "step": 5017
+    },
+    {
+      "epoch": 0.3868331791551033,
+      "grad_norm": 3.332284927368164,
+      "learning_rate": 7.017018244682106e-06,
+      "loss": 0.8992,
+      "step": 5018
+    },
+    {
+      "epoch": 0.38691026827012026,
+      "grad_norm": 3.850799560546875,
+      "learning_rate": 7.015875824960675e-06,
+      "loss": 1.105,
+      "step": 5019
+    },
+    {
+      "epoch": 0.38698735738513723,
+      "grad_norm": 3.9528861045837402,
+      "learning_rate": 7.014733279559829e-06,
+      "loss": 0.9505,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3870644465001542,
+      "grad_norm": 3.9368906021118164,
+      "learning_rate": 7.013590608550796e-06,
+      "loss": 0.9322,
+      "step": 5021
+    },
+    {
+      "epoch": 0.3871415356151711,
+      "grad_norm": 3.622251272201538,
+      "learning_rate": 7.0124478120048164e-06,
+      "loss": 0.9249,
+      "step": 5022
+    },
+    {
+      "epoch": 0.3872186247301881,
+      "grad_norm": 3.4808027744293213,
+      "learning_rate": 7.0113048899931406e-06,
+      "loss": 1.045,
+      "step": 5023
+    },
+    {
+      "epoch": 0.38729571384520506,
+      "grad_norm": 3.806081771850586,
+      "learning_rate": 7.010161842587019e-06,
+      "loss": 0.9854,
+      "step": 5024
+    },
+    {
+      "epoch": 0.38737280296022203,
+      "grad_norm": 3.621647834777832,
+      "learning_rate": 7.009018669857719e-06,
+      "loss": 0.9636,
+      "step": 5025
+    },
+    {
+      "epoch": 0.387449892075239,
+      "grad_norm": 3.846794605255127,
+      "learning_rate": 7.0078753718765105e-06,
+      "loss": 0.957,
+      "step": 5026
+    },
+    {
+      "epoch": 0.3875269811902559,
+      "grad_norm": 3.383054733276367,
+      "learning_rate": 7.00673194871467e-06,
+      "loss": 0.8777,
+      "step": 5027
+    },
+    {
+      "epoch": 0.3876040703052729,
+      "grad_norm": 3.6006200313568115,
+      "learning_rate": 7.005588400443487e-06,
+      "loss": 1.086,
+      "step": 5028
+    },
+    {
+      "epoch": 0.38768115942028986,
+      "grad_norm": 3.8396594524383545,
+      "learning_rate": 7.004444727134254e-06,
+      "loss": 1.1418,
+      "step": 5029
+    },
+    {
+      "epoch": 0.3877582485353068,
+      "grad_norm": 3.7753446102142334,
+      "learning_rate": 7.003300928858273e-06,
+      "loss": 0.9981,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3878353376503238,
+      "grad_norm": 3.5758206844329834,
+      "learning_rate": 7.002157005686855e-06,
+      "loss": 0.9798,
+      "step": 5031
+    },
+    {
+      "epoch": 0.3879124267653407,
+      "grad_norm": 4.047423362731934,
+      "learning_rate": 7.001012957691317e-06,
+      "loss": 1.0324,
+      "step": 5032
+    },
+    {
+      "epoch": 0.3879895158803577,
+      "grad_norm": 3.57684326171875,
+      "learning_rate": 6.999868784942985e-06,
+      "loss": 0.9592,
+      "step": 5033
+    },
+    {
+      "epoch": 0.38806660499537465,
+      "grad_norm": 3.5978035926818848,
+      "learning_rate": 6.998724487513191e-06,
+      "loss": 0.94,
+      "step": 5034
+    },
+    {
+      "epoch": 0.3881436941103916,
+      "grad_norm": 3.3599936962127686,
+      "learning_rate": 6.9975800654732785e-06,
+      "loss": 0.9095,
+      "step": 5035
+    },
+    {
+      "epoch": 0.3882207832254086,
+      "grad_norm": 3.662285566329956,
+      "learning_rate": 6.996435518894593e-06,
+      "loss": 0.9509,
+      "step": 5036
+    },
+    {
+      "epoch": 0.3882978723404255,
+      "grad_norm": 3.5465025901794434,
+      "learning_rate": 6.9952908478484925e-06,
+      "loss": 0.9714,
+      "step": 5037
+    },
+    {
+      "epoch": 0.3883749614554425,
+      "grad_norm": 3.2565081119537354,
+      "learning_rate": 6.994146052406343e-06,
+      "loss": 0.9064,
+      "step": 5038
+    },
+    {
+      "epoch": 0.38845205057045945,
+      "grad_norm": 3.7436461448669434,
+      "learning_rate": 6.993001132639514e-06,
+      "loss": 0.9848,
+      "step": 5039
+    },
+    {
+      "epoch": 0.3885291396854764,
+      "grad_norm": 3.5175907611846924,
+      "learning_rate": 6.991856088619387e-06,
+      "loss": 1.0,
+      "step": 5040
+    },
+    {
+      "epoch": 0.3886062288004934,
+      "grad_norm": 3.567593574523926,
+      "learning_rate": 6.99071092041735e-06,
+      "loss": 1.0659,
+      "step": 5041
+    },
+    {
+      "epoch": 0.3886833179155103,
+      "grad_norm": 3.355001211166382,
+      "learning_rate": 6.989565628104795e-06,
+      "loss": 0.8182,
+      "step": 5042
+    },
+    {
+      "epoch": 0.3887604070305273,
+      "grad_norm": 3.528596878051758,
+      "learning_rate": 6.988420211753129e-06,
+      "loss": 0.9568,
+      "step": 5043
+    },
+    {
+      "epoch": 0.38883749614554425,
+      "grad_norm": 3.691378593444824,
+      "learning_rate": 6.987274671433761e-06,
+      "loss": 1.025,
+      "step": 5044
+    },
+    {
+      "epoch": 0.3889145852605612,
+      "grad_norm": 3.5705032348632812,
+      "learning_rate": 6.986129007218107e-06,
+      "loss": 0.9854,
+      "step": 5045
+    },
+    {
+      "epoch": 0.3889916743755782,
+      "grad_norm": 3.5841269493103027,
+      "learning_rate": 6.984983219177599e-06,
+      "loss": 1.1187,
+      "step": 5046
+    },
+    {
+      "epoch": 0.3890687634905951,
+      "grad_norm": 3.42091965675354,
+      "learning_rate": 6.983837307383666e-06,
+      "loss": 0.9753,
+      "step": 5047
+    },
+    {
+      "epoch": 0.3891458526056121,
+      "grad_norm": 3.6692841053009033,
+      "learning_rate": 6.982691271907752e-06,
+      "loss": 0.8831,
+      "step": 5048
+    },
+    {
+      "epoch": 0.38922294172062905,
+      "grad_norm": 3.5156326293945312,
+      "learning_rate": 6.981545112821306e-06,
+      "loss": 1.0276,
+      "step": 5049
+    },
+    {
+      "epoch": 0.389300030835646,
+      "grad_norm": 3.318147897720337,
+      "learning_rate": 6.980398830195785e-06,
+      "loss": 0.9131,
+      "step": 5050
+    },
+    {
+      "epoch": 0.389377119950663,
+      "grad_norm": 4.0214152336120605,
+      "learning_rate": 6.979252424102654e-06,
+      "loss": 1.0253,
+      "step": 5051
+    },
+    {
+      "epoch": 0.3894542090656799,
+      "grad_norm": 3.803903818130493,
+      "learning_rate": 6.978105894613385e-06,
+      "loss": 1.0602,
+      "step": 5052
+    },
+    {
+      "epoch": 0.38953129818069687,
+      "grad_norm": 3.736529588699341,
+      "learning_rate": 6.976959241799456e-06,
+      "loss": 0.9655,
+      "step": 5053
+    },
+    {
+      "epoch": 0.38960838729571384,
+      "grad_norm": 3.8049023151397705,
+      "learning_rate": 6.97581246573236e-06,
+      "loss": 0.9087,
+      "step": 5054
+    },
+    {
+      "epoch": 0.3896854764107308,
+      "grad_norm": 4.01686954498291,
+      "learning_rate": 6.974665566483588e-06,
+      "loss": 1.1021,
+      "step": 5055
+    },
+    {
+      "epoch": 0.3897625655257478,
+      "grad_norm": 3.589202642440796,
+      "learning_rate": 6.9735185441246466e-06,
+      "loss": 0.9415,
+      "step": 5056
+    },
+    {
+      "epoch": 0.3898396546407647,
+      "grad_norm": 3.462862014770508,
+      "learning_rate": 6.972371398727045e-06,
+      "loss": 0.9772,
+      "step": 5057
+    },
+    {
+      "epoch": 0.38991674375578167,
+      "grad_norm": 3.413395404815674,
+      "learning_rate": 6.971224130362301e-06,
+      "loss": 0.9391,
+      "step": 5058
+    },
+    {
+      "epoch": 0.38999383287079864,
+      "grad_norm": 3.5816166400909424,
+      "learning_rate": 6.970076739101942e-06,
+      "loss": 0.9687,
+      "step": 5059
+    },
+    {
+      "epoch": 0.3900709219858156,
+      "grad_norm": 3.3513641357421875,
+      "learning_rate": 6.968929225017501e-06,
+      "loss": 1.0188,
+      "step": 5060
+    },
+    {
+      "epoch": 0.3901480111008326,
+      "grad_norm": 3.450777292251587,
+      "learning_rate": 6.9677815881805215e-06,
+      "loss": 0.9195,
+      "step": 5061
+    },
+    {
+      "epoch": 0.3902251002158495,
+      "grad_norm": 3.946836471557617,
+      "learning_rate": 6.96663382866255e-06,
+      "loss": 0.9913,
+      "step": 5062
+    },
+    {
+      "epoch": 0.39030218933086647,
+      "grad_norm": 3.5374841690063477,
+      "learning_rate": 6.965485946535145e-06,
+      "loss": 0.945,
+      "step": 5063
+    },
+    {
+      "epoch": 0.39037927844588344,
+      "grad_norm": 3.718808174133301,
+      "learning_rate": 6.964337941869872e-06,
+      "loss": 1.0653,
+      "step": 5064
+    },
+    {
+      "epoch": 0.3904563675609004,
+      "grad_norm": 3.8617405891418457,
+      "learning_rate": 6.963189814738301e-06,
+      "loss": 1.0212,
+      "step": 5065
+    },
+    {
+      "epoch": 0.3905334566759174,
+      "grad_norm": 3.3130338191986084,
+      "learning_rate": 6.962041565212012e-06,
+      "loss": 0.9596,
+      "step": 5066
+    },
+    {
+      "epoch": 0.3906105457909343,
+      "grad_norm": 3.5554182529449463,
+      "learning_rate": 6.960893193362594e-06,
+      "loss": 1.0024,
+      "step": 5067
+    },
+    {
+      "epoch": 0.39068763490595126,
+      "grad_norm": 4.0424604415893555,
+      "learning_rate": 6.959744699261641e-06,
+      "loss": 0.9991,
+      "step": 5068
+    },
+    {
+      "epoch": 0.39076472402096823,
+      "grad_norm": 3.6211678981781006,
+      "learning_rate": 6.9585960829807555e-06,
+      "loss": 1.0449,
+      "step": 5069
+    },
+    {
+      "epoch": 0.3908418131359852,
+      "grad_norm": 3.4850916862487793,
+      "learning_rate": 6.9574473445915495e-06,
+      "loss": 0.9836,
+      "step": 5070
+    },
+    {
+      "epoch": 0.3909189022510022,
+      "grad_norm": 4.005676746368408,
+      "learning_rate": 6.956298484165638e-06,
+      "loss": 0.8692,
+      "step": 5071
+    },
+    {
+      "epoch": 0.3909959913660191,
+      "grad_norm": 3.5592994689941406,
+      "learning_rate": 6.955149501774648e-06,
+      "loss": 1.1276,
+      "step": 5072
+    },
+    {
+      "epoch": 0.39107308048103606,
+      "grad_norm": 3.935800313949585,
+      "learning_rate": 6.954000397490213e-06,
+      "loss": 0.992,
+      "step": 5073
+    },
+    {
+      "epoch": 0.39115016959605303,
+      "grad_norm": 3.423206090927124,
+      "learning_rate": 6.952851171383972e-06,
+      "loss": 0.923,
+      "step": 5074
+    },
+    {
+      "epoch": 0.39122725871107,
+      "grad_norm": 3.449836254119873,
+      "learning_rate": 6.951701823527575e-06,
+      "loss": 0.9982,
+      "step": 5075
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 3.2396328449249268,
+      "learning_rate": 6.950552353992678e-06,
+      "loss": 0.8038,
+      "step": 5076
+    },
+    {
+      "epoch": 0.3913814369411039,
+      "grad_norm": 3.798563003540039,
+      "learning_rate": 6.949402762850943e-06,
+      "loss": 1.0334,
+      "step": 5077
+    },
+    {
+      "epoch": 0.39145852605612086,
+      "grad_norm": 3.7131567001342773,
+      "learning_rate": 6.948253050174043e-06,
+      "loss": 1.0026,
+      "step": 5078
+    },
+    {
+      "epoch": 0.39153561517113783,
+      "grad_norm": 4.187877655029297,
+      "learning_rate": 6.947103216033655e-06,
+      "loss": 1.1898,
+      "step": 5079
+    },
+    {
+      "epoch": 0.3916127042861548,
+      "grad_norm": 3.439176082611084,
+      "learning_rate": 6.945953260501466e-06,
+      "loss": 0.9082,
+      "step": 5080
+    },
+    {
+      "epoch": 0.39168979340117177,
+      "grad_norm": 3.9063847064971924,
+      "learning_rate": 6.9448031836491705e-06,
+      "loss": 1.0431,
+      "step": 5081
+    },
+    {
+      "epoch": 0.39176688251618874,
+      "grad_norm": 3.500028610229492,
+      "learning_rate": 6.943652985548468e-06,
+      "loss": 0.9663,
+      "step": 5082
+    },
+    {
+      "epoch": 0.39184397163120566,
+      "grad_norm": 3.206749439239502,
+      "learning_rate": 6.942502666271069e-06,
+      "loss": 0.8262,
+      "step": 5083
+    },
+    {
+      "epoch": 0.3919210607462226,
+      "grad_norm": 3.7229039669036865,
+      "learning_rate": 6.9413522258886874e-06,
+      "loss": 0.9906,
+      "step": 5084
+    },
+    {
+      "epoch": 0.3919981498612396,
+      "grad_norm": 3.451598882675171,
+      "learning_rate": 6.940201664473051e-06,
+      "loss": 0.9667,
+      "step": 5085
+    },
+    {
+      "epoch": 0.39207523897625657,
+      "grad_norm": 4.100324630737305,
+      "learning_rate": 6.939050982095889e-06,
+      "loss": 0.9991,
+      "step": 5086
+    },
+    {
+      "epoch": 0.39215232809127354,
+      "grad_norm": 3.7283709049224854,
+      "learning_rate": 6.93790017882894e-06,
+      "loss": 0.9816,
+      "step": 5087
+    },
+    {
+      "epoch": 0.39222941720629045,
+      "grad_norm": 3.6680102348327637,
+      "learning_rate": 6.936749254743951e-06,
+      "loss": 0.9915,
+      "step": 5088
+    },
+    {
+      "epoch": 0.3923065063213074,
+      "grad_norm": 3.723318576812744,
+      "learning_rate": 6.935598209912679e-06,
+      "loss": 0.9894,
+      "step": 5089
+    },
+    {
+      "epoch": 0.3923835954363244,
+      "grad_norm": 3.6859655380249023,
+      "learning_rate": 6.934447044406882e-06,
+      "loss": 1.0608,
+      "step": 5090
+    },
+    {
+      "epoch": 0.39246068455134137,
+      "grad_norm": 4.22714900970459,
+      "learning_rate": 6.9332957582983295e-06,
+      "loss": 0.977,
+      "step": 5091
+    },
+    {
+      "epoch": 0.39253777366635834,
+      "grad_norm": 3.4322969913482666,
+      "learning_rate": 6.932144351658801e-06,
+      "loss": 1.0241,
+      "step": 5092
+    },
+    {
+      "epoch": 0.39261486278137525,
+      "grad_norm": 3.7159926891326904,
+      "learning_rate": 6.930992824560078e-06,
+      "loss": 0.901,
+      "step": 5093
+    },
+    {
+      "epoch": 0.3926919518963922,
+      "grad_norm": 3.612903356552124,
+      "learning_rate": 6.9298411770739535e-06,
+      "loss": 0.9303,
+      "step": 5094
+    },
+    {
+      "epoch": 0.3927690410114092,
+      "grad_norm": 3.71382999420166,
+      "learning_rate": 6.9286894092722265e-06,
+      "loss": 0.9735,
+      "step": 5095
+    },
+    {
+      "epoch": 0.39284613012642616,
+      "grad_norm": 3.824847459793091,
+      "learning_rate": 6.9275375212267035e-06,
+      "loss": 1.0065,
+      "step": 5096
+    },
+    {
+      "epoch": 0.39292321924144313,
+      "grad_norm": 3.6681530475616455,
+      "learning_rate": 6.926385513009199e-06,
+      "loss": 1.0493,
+      "step": 5097
+    },
+    {
+      "epoch": 0.39300030835646005,
+      "grad_norm": 3.7533106803894043,
+      "learning_rate": 6.925233384691534e-06,
+      "loss": 1.077,
+      "step": 5098
+    },
+    {
+      "epoch": 0.393077397471477,
+      "grad_norm": 3.778269052505493,
+      "learning_rate": 6.924081136345541e-06,
+      "loss": 0.9683,
+      "step": 5099
+    },
+    {
+      "epoch": 0.393154486586494,
+      "grad_norm": 4.528807163238525,
+      "learning_rate": 6.9229287680430526e-06,
+      "loss": 1.0982,
+      "step": 5100
+    },
+    {
+      "epoch": 0.39323157570151096,
+      "grad_norm": 3.8181581497192383,
+      "learning_rate": 6.921776279855914e-06,
+      "loss": 0.8895,
+      "step": 5101
+    },
+    {
+      "epoch": 0.39330866481652793,
+      "grad_norm": 4.004360198974609,
+      "learning_rate": 6.92062367185598e-06,
+      "loss": 0.9677,
+      "step": 5102
+    },
+    {
+      "epoch": 0.39338575393154485,
+      "grad_norm": 3.5526719093322754,
+      "learning_rate": 6.919470944115104e-06,
+      "loss": 0.9523,
+      "step": 5103
+    },
+    {
+      "epoch": 0.3934628430465618,
+      "grad_norm": 3.5170533657073975,
+      "learning_rate": 6.918318096705157e-06,
+      "loss": 0.9552,
+      "step": 5104
+    },
+    {
+      "epoch": 0.3935399321615788,
+      "grad_norm": 3.4880666732788086,
+      "learning_rate": 6.917165129698013e-06,
+      "loss": 0.9412,
+      "step": 5105
+    },
+    {
+      "epoch": 0.39361702127659576,
+      "grad_norm": 3.667492628097534,
+      "learning_rate": 6.916012043165552e-06,
+      "loss": 0.889,
+      "step": 5106
+    },
+    {
+      "epoch": 0.39369411039161273,
+      "grad_norm": 3.4989163875579834,
+      "learning_rate": 6.9148588371796635e-06,
+      "loss": 1.009,
+      "step": 5107
+    },
+    {
+      "epoch": 0.39377119950662964,
+      "grad_norm": 3.465567111968994,
+      "learning_rate": 6.913705511812243e-06,
+      "loss": 0.9644,
+      "step": 5108
+    },
+    {
+      "epoch": 0.3938482886216466,
+      "grad_norm": 3.9162404537200928,
+      "learning_rate": 6.912552067135195e-06,
+      "loss": 0.9753,
+      "step": 5109
+    },
+    {
+      "epoch": 0.3939253777366636,
+      "grad_norm": 3.494250774383545,
+      "learning_rate": 6.9113985032204325e-06,
+      "loss": 1.0625,
+      "step": 5110
+    },
+    {
+      "epoch": 0.39400246685168056,
+      "grad_norm": 3.6587326526641846,
+      "learning_rate": 6.910244820139871e-06,
+      "loss": 0.9012,
+      "step": 5111
+    },
+    {
+      "epoch": 0.3940795559666975,
+      "grad_norm": 4.449551105499268,
+      "learning_rate": 6.909091017965439e-06,
+      "loss": 1.0422,
+      "step": 5112
+    },
+    {
+      "epoch": 0.39415664508171444,
+      "grad_norm": 3.5880303382873535,
+      "learning_rate": 6.907937096769071e-06,
+      "loss": 1.0439,
+      "step": 5113
+    },
+    {
+      "epoch": 0.3942337341967314,
+      "grad_norm": 4.151397228240967,
+      "learning_rate": 6.906783056622703e-06,
+      "loss": 1.1482,
+      "step": 5114
+    },
+    {
+      "epoch": 0.3943108233117484,
+      "grad_norm": 3.602203607559204,
+      "learning_rate": 6.9056288975982896e-06,
+      "loss": 0.9506,
+      "step": 5115
+    },
+    {
+      "epoch": 0.39438791242676535,
+      "grad_norm": 3.9229793548583984,
+      "learning_rate": 6.904474619767784e-06,
+      "loss": 1.0528,
+      "step": 5116
+    },
+    {
+      "epoch": 0.3944650015417823,
+      "grad_norm": 3.542268991470337,
+      "learning_rate": 6.903320223203148e-06,
+      "loss": 1.0154,
+      "step": 5117
+    },
+    {
+      "epoch": 0.39454209065679924,
+      "grad_norm": 3.427914619445801,
+      "learning_rate": 6.9021657079763545e-06,
+      "loss": 0.9821,
+      "step": 5118
+    },
+    {
+      "epoch": 0.3946191797718162,
+      "grad_norm": 3.257849931716919,
+      "learning_rate": 6.901011074159381e-06,
+      "loss": 1.0341,
+      "step": 5119
+    },
+    {
+      "epoch": 0.3946962688868332,
+      "grad_norm": 3.4966633319854736,
+      "learning_rate": 6.899856321824212e-06,
+      "loss": 0.9723,
+      "step": 5120
+    },
+    {
+      "epoch": 0.39477335800185015,
+      "grad_norm": 3.4750454425811768,
+      "learning_rate": 6.89870145104284e-06,
+      "loss": 0.9318,
+      "step": 5121
+    },
+    {
+      "epoch": 0.3948504471168671,
+      "grad_norm": 4.037153244018555,
+      "learning_rate": 6.897546461887268e-06,
+      "loss": 0.9593,
+      "step": 5122
+    },
+    {
+      "epoch": 0.39492753623188404,
+      "grad_norm": 3.800772190093994,
+      "learning_rate": 6.896391354429501e-06,
+      "loss": 1.0182,
+      "step": 5123
+    },
+    {
+      "epoch": 0.395004625346901,
+      "grad_norm": 3.4899842739105225,
+      "learning_rate": 6.895236128741554e-06,
+      "loss": 1.0558,
+      "step": 5124
+    },
+    {
+      "epoch": 0.395081714461918,
+      "grad_norm": 3.538621187210083,
+      "learning_rate": 6.8940807848954506e-06,
+      "loss": 0.9602,
+      "step": 5125
+    },
+    {
+      "epoch": 0.39515880357693495,
+      "grad_norm": 4.251044273376465,
+      "learning_rate": 6.892925322963221e-06,
+      "loss": 1.1921,
+      "step": 5126
+    },
+    {
+      "epoch": 0.3952358926919519,
+      "grad_norm": 3.6266791820526123,
+      "learning_rate": 6.8917697430169e-06,
+      "loss": 0.9136,
+      "step": 5127
+    },
+    {
+      "epoch": 0.39531298180696883,
+      "grad_norm": 3.5252368450164795,
+      "learning_rate": 6.890614045128533e-06,
+      "loss": 0.911,
+      "step": 5128
+    },
+    {
+      "epoch": 0.3953900709219858,
+      "grad_norm": 4.031729698181152,
+      "learning_rate": 6.889458229370173e-06,
+      "loss": 1.0345,
+      "step": 5129
+    },
+    {
+      "epoch": 0.3954671600370028,
+      "grad_norm": 3.6114397048950195,
+      "learning_rate": 6.888302295813878e-06,
+      "loss": 1.0215,
+      "step": 5130
+    },
+    {
+      "epoch": 0.39554424915201974,
+      "grad_norm": 3.7317934036254883,
+      "learning_rate": 6.887146244531715e-06,
+      "loss": 1.0653,
+      "step": 5131
+    },
+    {
+      "epoch": 0.3956213382670367,
+      "grad_norm": 3.487471580505371,
+      "learning_rate": 6.885990075595757e-06,
+      "loss": 1.0672,
+      "step": 5132
+    },
+    {
+      "epoch": 0.39569842738205363,
+      "grad_norm": 3.7505714893341064,
+      "learning_rate": 6.884833789078084e-06,
+      "loss": 1.0301,
+      "step": 5133
+    },
+    {
+      "epoch": 0.3957755164970706,
+      "grad_norm": 3.2812490463256836,
+      "learning_rate": 6.883677385050788e-06,
+      "loss": 0.9663,
+      "step": 5134
+    },
+    {
+      "epoch": 0.39585260561208757,
+      "grad_norm": 4.058938026428223,
+      "learning_rate": 6.882520863585962e-06,
+      "loss": 1.0798,
+      "step": 5135
+    },
+    {
+      "epoch": 0.39592969472710454,
+      "grad_norm": 3.5618515014648438,
+      "learning_rate": 6.88136422475571e-06,
+      "loss": 0.9328,
+      "step": 5136
+    },
+    {
+      "epoch": 0.3960067838421215,
+      "grad_norm": 3.4688501358032227,
+      "learning_rate": 6.880207468632143e-06,
+      "loss": 0.9176,
+      "step": 5137
+    },
+    {
+      "epoch": 0.3960838729571384,
+      "grad_norm": 3.658219575881958,
+      "learning_rate": 6.8790505952873775e-06,
+      "loss": 0.9566,
+      "step": 5138
+    },
+    {
+      "epoch": 0.3961609620721554,
+      "grad_norm": 3.3856353759765625,
+      "learning_rate": 6.877893604793539e-06,
+      "loss": 1.046,
+      "step": 5139
+    },
+    {
+      "epoch": 0.39623805118717237,
+      "grad_norm": 3.460387706756592,
+      "learning_rate": 6.876736497222762e-06,
+      "loss": 0.9436,
+      "step": 5140
+    },
+    {
+      "epoch": 0.39631514030218934,
+      "grad_norm": 3.813292980194092,
+      "learning_rate": 6.875579272647182e-06,
+      "loss": 1.0252,
+      "step": 5141
+    },
+    {
+      "epoch": 0.3963922294172063,
+      "grad_norm": 3.8571043014526367,
+      "learning_rate": 6.87442193113895e-06,
+      "loss": 1.0681,
+      "step": 5142
+    },
+    {
+      "epoch": 0.3964693185322232,
+      "grad_norm": 3.726741075515747,
+      "learning_rate": 6.873264472770217e-06,
+      "loss": 0.9728,
+      "step": 5143
+    },
+    {
+      "epoch": 0.3965464076472402,
+      "grad_norm": 3.4998526573181152,
+      "learning_rate": 6.8721068976131466e-06,
+      "loss": 0.966,
+      "step": 5144
+    },
+    {
+      "epoch": 0.39662349676225717,
+      "grad_norm": 3.4608278274536133,
+      "learning_rate": 6.870949205739907e-06,
+      "loss": 1.0177,
+      "step": 5145
+    },
+    {
+      "epoch": 0.39670058587727414,
+      "grad_norm": 3.6188559532165527,
+      "learning_rate": 6.869791397222674e-06,
+      "loss": 1.0666,
+      "step": 5146
+    },
+    {
+      "epoch": 0.3967776749922911,
+      "grad_norm": 3.4421889781951904,
+      "learning_rate": 6.868633472133632e-06,
+      "loss": 0.9928,
+      "step": 5147
+    },
+    {
+      "epoch": 0.396854764107308,
+      "grad_norm": 3.740668773651123,
+      "learning_rate": 6.867475430544971e-06,
+      "loss": 1.0269,
+      "step": 5148
+    },
+    {
+      "epoch": 0.396931853222325,
+      "grad_norm": 3.9305145740509033,
+      "learning_rate": 6.866317272528889e-06,
+      "loss": 0.9517,
+      "step": 5149
+    },
+    {
+      "epoch": 0.39700894233734196,
+      "grad_norm": 3.588923931121826,
+      "learning_rate": 6.865158998157591e-06,
+      "loss": 0.9619,
+      "step": 5150
+    },
+    {
+      "epoch": 0.39708603145235893,
+      "grad_norm": 3.9232397079467773,
+      "learning_rate": 6.8640006075032875e-06,
+      "loss": 1.1459,
+      "step": 5151
+    },
+    {
+      "epoch": 0.3971631205673759,
+      "grad_norm": 3.4399147033691406,
+      "learning_rate": 6.862842100638201e-06,
+      "loss": 0.8585,
+      "step": 5152
+    },
+    {
+      "epoch": 0.3972402096823928,
+      "grad_norm": 3.5570197105407715,
+      "learning_rate": 6.861683477634559e-06,
+      "loss": 1.0854,
+      "step": 5153
+    },
+    {
+      "epoch": 0.3973172987974098,
+      "grad_norm": 3.628215789794922,
+      "learning_rate": 6.860524738564591e-06,
+      "loss": 0.9911,
+      "step": 5154
+    },
+    {
+      "epoch": 0.39739438791242676,
+      "grad_norm": 4.154507160186768,
+      "learning_rate": 6.859365883500545e-06,
+      "loss": 1.0277,
+      "step": 5155
+    },
+    {
+      "epoch": 0.39747147702744373,
+      "grad_norm": 3.7878050804138184,
+      "learning_rate": 6.858206912514664e-06,
+      "loss": 1.075,
+      "step": 5156
+    },
+    {
+      "epoch": 0.3975485661424607,
+      "grad_norm": 3.571855068206787,
+      "learning_rate": 6.857047825679206e-06,
+      "loss": 0.9955,
+      "step": 5157
+    },
+    {
+      "epoch": 0.3976256552574776,
+      "grad_norm": 3.5463037490844727,
+      "learning_rate": 6.855888623066434e-06,
+      "loss": 0.969,
+      "step": 5158
+    },
+    {
+      "epoch": 0.3977027443724946,
+      "grad_norm": 3.6123862266540527,
+      "learning_rate": 6.854729304748619e-06,
+      "loss": 0.8627,
+      "step": 5159
+    },
+    {
+      "epoch": 0.39777983348751156,
+      "grad_norm": 4.434622764587402,
+      "learning_rate": 6.8535698707980356e-06,
+      "loss": 0.9712,
+      "step": 5160
+    },
+    {
+      "epoch": 0.39785692260252853,
+      "grad_norm": 3.731267213821411,
+      "learning_rate": 6.852410321286974e-06,
+      "loss": 0.9873,
+      "step": 5161
+    },
+    {
+      "epoch": 0.3979340117175455,
+      "grad_norm": 3.5864498615264893,
+      "learning_rate": 6.85125065628772e-06,
+      "loss": 0.9812,
+      "step": 5162
+    },
+    {
+      "epoch": 0.3980111008325624,
+      "grad_norm": 3.548417329788208,
+      "learning_rate": 6.850090875872577e-06,
+      "loss": 0.9825,
+      "step": 5163
+    },
+    {
+      "epoch": 0.3980881899475794,
+      "grad_norm": 3.9418680667877197,
+      "learning_rate": 6.848930980113848e-06,
+      "loss": 1.1676,
+      "step": 5164
+    },
+    {
+      "epoch": 0.39816527906259636,
+      "grad_norm": 3.502095937728882,
+      "learning_rate": 6.8477709690838486e-06,
+      "loss": 0.9376,
+      "step": 5165
+    },
+    {
+      "epoch": 0.3982423681776133,
+      "grad_norm": 3.5473270416259766,
+      "learning_rate": 6.846610842854902e-06,
+      "loss": 0.9449,
+      "step": 5166
+    },
+    {
+      "epoch": 0.3983194572926303,
+      "grad_norm": 3.6665074825286865,
+      "learning_rate": 6.845450601499331e-06,
+      "loss": 1.0283,
+      "step": 5167
+    },
+    {
+      "epoch": 0.39839654640764727,
+      "grad_norm": 3.162196636199951,
+      "learning_rate": 6.844290245089473e-06,
+      "loss": 0.8484,
+      "step": 5168
+    },
+    {
+      "epoch": 0.3984736355226642,
+      "grad_norm": 3.5580244064331055,
+      "learning_rate": 6.84312977369767e-06,
+      "loss": 0.958,
+      "step": 5169
+    },
+    {
+      "epoch": 0.39855072463768115,
+      "grad_norm": 3.2327682971954346,
+      "learning_rate": 6.841969187396271e-06,
+      "loss": 0.8806,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3986278137526981,
+      "grad_norm": 3.59846568107605,
+      "learning_rate": 6.840808486257634e-06,
+      "loss": 0.8916,
+      "step": 5171
+    },
+    {
+      "epoch": 0.3987049028677151,
+      "grad_norm": 3.460810899734497,
+      "learning_rate": 6.83964767035412e-06,
+      "loss": 1.0101,
+      "step": 5172
+    },
+    {
+      "epoch": 0.39878199198273206,
+      "grad_norm": 3.336289882659912,
+      "learning_rate": 6.838486739758102e-06,
+      "loss": 0.9098,
+      "step": 5173
+    },
+    {
+      "epoch": 0.398859081097749,
+      "grad_norm": 3.8699560165405273,
+      "learning_rate": 6.83732569454196e-06,
+      "loss": 1.0069,
+      "step": 5174
+    },
+    {
+      "epoch": 0.39893617021276595,
+      "grad_norm": 3.5346524715423584,
+      "learning_rate": 6.836164534778074e-06,
+      "loss": 0.9581,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3990132593277829,
+      "grad_norm": 3.468693733215332,
+      "learning_rate": 6.835003260538839e-06,
+      "loss": 0.9279,
+      "step": 5176
+    },
+    {
+      "epoch": 0.3990903484427999,
+      "grad_norm": 3.598850727081299,
+      "learning_rate": 6.833841871896656e-06,
+      "loss": 1.0076,
+      "step": 5177
+    },
+    {
+      "epoch": 0.39916743755781686,
+      "grad_norm": 3.525563955307007,
+      "learning_rate": 6.83268036892393e-06,
+      "loss": 0.9998,
+      "step": 5178
+    },
+    {
+      "epoch": 0.3992445266728338,
+      "grad_norm": 3.8953518867492676,
+      "learning_rate": 6.831518751693073e-06,
+      "loss": 0.9768,
+      "step": 5179
+    },
+    {
+      "epoch": 0.39932161578785075,
+      "grad_norm": 3.505366802215576,
+      "learning_rate": 6.830357020276509e-06,
+      "loss": 1.007,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3993987049028677,
+      "grad_norm": 3.4880192279815674,
+      "learning_rate": 6.829195174746663e-06,
+      "loss": 1.0653,
+      "step": 5181
+    },
+    {
+      "epoch": 0.3994757940178847,
+      "grad_norm": 3.4839048385620117,
+      "learning_rate": 6.828033215175974e-06,
+      "loss": 0.9051,
+      "step": 5182
+    },
+    {
+      "epoch": 0.39955288313290166,
+      "grad_norm": 3.4201345443725586,
+      "learning_rate": 6.826871141636879e-06,
+      "loss": 0.9144,
+      "step": 5183
+    },
+    {
+      "epoch": 0.3996299722479186,
+      "grad_norm": 3.57698917388916,
+      "learning_rate": 6.825708954201832e-06,
+      "loss": 0.9932,
+      "step": 5184
+    },
+    {
+      "epoch": 0.39970706136293555,
+      "grad_norm": 3.7728288173675537,
+      "learning_rate": 6.824546652943287e-06,
+      "loss": 1.0721,
+      "step": 5185
+    },
+    {
+      "epoch": 0.3997841504779525,
+      "grad_norm": 3.7117831707000732,
+      "learning_rate": 6.823384237933706e-06,
+      "loss": 0.9387,
+      "step": 5186
+    },
+    {
+      "epoch": 0.3998612395929695,
+      "grad_norm": 3.299254894256592,
+      "learning_rate": 6.822221709245562e-06,
+      "loss": 0.9593,
+      "step": 5187
+    },
+    {
+      "epoch": 0.39993832870798646,
+      "grad_norm": 3.3196065425872803,
+      "learning_rate": 6.8210590669513325e-06,
+      "loss": 0.9527,
+      "step": 5188
+    },
+    {
+      "epoch": 0.40001541782300337,
+      "grad_norm": 3.578375816345215,
+      "learning_rate": 6.8198963111235e-06,
+      "loss": 0.9586,
+      "step": 5189
+    },
+    {
+      "epoch": 0.40009250693802034,
+      "grad_norm": 3.807652235031128,
+      "learning_rate": 6.8187334418345605e-06,
+      "loss": 0.8759,
+      "step": 5190
+    },
+    {
+      "epoch": 0.4001695960530373,
+      "grad_norm": 3.5907301902770996,
+      "learning_rate": 6.8175704591570105e-06,
+      "loss": 1.0146,
+      "step": 5191
+    },
+    {
+      "epoch": 0.4002466851680543,
+      "grad_norm": 3.224435806274414,
+      "learning_rate": 6.816407363163354e-06,
+      "loss": 0.9449,
+      "step": 5192
+    },
+    {
+      "epoch": 0.40032377428307125,
+      "grad_norm": 3.421980619430542,
+      "learning_rate": 6.815244153926106e-06,
+      "loss": 0.893,
+      "step": 5193
+    },
+    {
+      "epoch": 0.40040086339808817,
+      "grad_norm": 3.3092360496520996,
+      "learning_rate": 6.814080831517787e-06,
+      "loss": 0.8947,
+      "step": 5194
+    },
+    {
+      "epoch": 0.40047795251310514,
+      "grad_norm": 3.3089075088500977,
+      "learning_rate": 6.812917396010924e-06,
+      "loss": 0.8276,
+      "step": 5195
+    },
+    {
+      "epoch": 0.4005550416281221,
+      "grad_norm": 3.9031176567077637,
+      "learning_rate": 6.811753847478051e-06,
+      "loss": 1.0147,
+      "step": 5196
+    },
+    {
+      "epoch": 0.4006321307431391,
+      "grad_norm": 3.475313901901245,
+      "learning_rate": 6.810590185991707e-06,
+      "loss": 0.9532,
+      "step": 5197
+    },
+    {
+      "epoch": 0.40070921985815605,
+      "grad_norm": 3.7122201919555664,
+      "learning_rate": 6.8094264116244434e-06,
+      "loss": 1.0051,
+      "step": 5198
+    },
+    {
+      "epoch": 0.40078630897317297,
+      "grad_norm": 3.6033477783203125,
+      "learning_rate": 6.8082625244488145e-06,
+      "loss": 1.0563,
+      "step": 5199
+    },
+    {
+      "epoch": 0.40086339808818994,
+      "grad_norm": 4.018650054931641,
+      "learning_rate": 6.807098524537381e-06,
+      "loss": 0.922,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4009404872032069,
+      "grad_norm": 3.791778087615967,
+      "learning_rate": 6.8059344119627155e-06,
+      "loss": 1.0258,
+      "step": 5201
+    },
+    {
+      "epoch": 0.4010175763182239,
+      "grad_norm": 3.5043606758117676,
+      "learning_rate": 6.804770186797391e-06,
+      "loss": 1.0035,
+      "step": 5202
+    },
+    {
+      "epoch": 0.40109466543324085,
+      "grad_norm": 4.14236307144165,
+      "learning_rate": 6.8036058491139944e-06,
+      "loss": 0.9195,
+      "step": 5203
+    },
+    {
+      "epoch": 0.40117175454825776,
+      "grad_norm": 3.829864263534546,
+      "learning_rate": 6.802441398985114e-06,
+      "loss": 0.9637,
+      "step": 5204
+    },
+    {
+      "epoch": 0.40124884366327473,
+      "grad_norm": 3.572660446166992,
+      "learning_rate": 6.801276836483346e-06,
+      "loss": 0.8871,
+      "step": 5205
+    },
+    {
+      "epoch": 0.4013259327782917,
+      "grad_norm": 3.405879259109497,
+      "learning_rate": 6.800112161681297e-06,
+      "loss": 0.8893,
+      "step": 5206
+    },
+    {
+      "epoch": 0.4014030218933087,
+      "grad_norm": 3.654069662094116,
+      "learning_rate": 6.798947374651578e-06,
+      "loss": 1.0638,
+      "step": 5207
+    },
+    {
+      "epoch": 0.40148011100832565,
+      "grad_norm": 3.986391067504883,
+      "learning_rate": 6.797782475466806e-06,
+      "loss": 1.0039,
+      "step": 5208
+    },
+    {
+      "epoch": 0.40155720012334256,
+      "grad_norm": 3.699451446533203,
+      "learning_rate": 6.7966174641996085e-06,
+      "loss": 0.9939,
+      "step": 5209
+    },
+    {
+      "epoch": 0.40163428923835953,
+      "grad_norm": 4.058764934539795,
+      "learning_rate": 6.795452340922617e-06,
+      "loss": 0.9998,
+      "step": 5210
+    },
+    {
+      "epoch": 0.4017113783533765,
+      "grad_norm": 4.153298854827881,
+      "learning_rate": 6.7942871057084715e-06,
+      "loss": 1.0098,
+      "step": 5211
+    },
+    {
+      "epoch": 0.4017884674683935,
+      "grad_norm": 3.8170831203460693,
+      "learning_rate": 6.793121758629817e-06,
+      "loss": 1.0502,
+      "step": 5212
+    },
+    {
+      "epoch": 0.40186555658341044,
+      "grad_norm": 4.2434844970703125,
+      "learning_rate": 6.791956299759307e-06,
+      "loss": 0.9763,
+      "step": 5213
+    },
+    {
+      "epoch": 0.40194264569842736,
+      "grad_norm": 4.152831077575684,
+      "learning_rate": 6.790790729169604e-06,
+      "loss": 1.0741,
+      "step": 5214
+    },
+    {
+      "epoch": 0.40201973481344433,
+      "grad_norm": 3.7323899269104004,
+      "learning_rate": 6.7896250469333725e-06,
+      "loss": 0.9627,
+      "step": 5215
+    },
+    {
+      "epoch": 0.4020968239284613,
+      "grad_norm": 4.520125389099121,
+      "learning_rate": 6.78845925312329e-06,
+      "loss": 1.0859,
+      "step": 5216
+    },
+    {
+      "epoch": 0.40217391304347827,
+      "grad_norm": 5.602902889251709,
+      "learning_rate": 6.787293347812034e-06,
+      "loss": 1.0761,
+      "step": 5217
+    },
+    {
+      "epoch": 0.40225100215849524,
+      "grad_norm": 3.4149789810180664,
+      "learning_rate": 6.7861273310722945e-06,
+      "loss": 0.9928,
+      "step": 5218
+    },
+    {
+      "epoch": 0.40232809127351216,
+      "grad_norm": 3.432222366333008,
+      "learning_rate": 6.784961202976768e-06,
+      "loss": 0.9324,
+      "step": 5219
+    },
+    {
+      "epoch": 0.4024051803885291,
+      "grad_norm": 3.2903270721435547,
+      "learning_rate": 6.7837949635981524e-06,
+      "loss": 0.9092,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4024822695035461,
+      "grad_norm": 3.851072072982788,
+      "learning_rate": 6.782628613009161e-06,
+      "loss": 0.9979,
+      "step": 5221
+    },
+    {
+      "epoch": 0.40255935861856307,
+      "grad_norm": 3.850883960723877,
+      "learning_rate": 6.781462151282508e-06,
+      "loss": 0.9482,
+      "step": 5222
+    },
+    {
+      "epoch": 0.40263644773358004,
+      "grad_norm": 3.9080920219421387,
+      "learning_rate": 6.7802955784909165e-06,
+      "loss": 0.9543,
+      "step": 5223
+    },
+    {
+      "epoch": 0.40271353684859695,
+      "grad_norm": 3.799767255783081,
+      "learning_rate": 6.779128894707116e-06,
+      "loss": 0.9217,
+      "step": 5224
+    },
+    {
+      "epoch": 0.4027906259636139,
+      "grad_norm": 3.8918371200561523,
+      "learning_rate": 6.777962100003843e-06,
+      "loss": 1.1152,
+      "step": 5225
+    },
+    {
+      "epoch": 0.4028677150786309,
+      "grad_norm": 3.4487974643707275,
+      "learning_rate": 6.776795194453841e-06,
+      "loss": 1.0196,
+      "step": 5226
+    },
+    {
+      "epoch": 0.40294480419364787,
+      "grad_norm": 3.964271068572998,
+      "learning_rate": 6.7756281781298615e-06,
+      "loss": 1.0352,
+      "step": 5227
+    },
+    {
+      "epoch": 0.40302189330866484,
+      "grad_norm": 3.359325647354126,
+      "learning_rate": 6.7744610511046615e-06,
+      "loss": 0.933,
+      "step": 5228
+    },
+    {
+      "epoch": 0.40309898242368175,
+      "grad_norm": 3.6119518280029297,
+      "learning_rate": 6.773293813451004e-06,
+      "loss": 0.9631,
+      "step": 5229
+    },
+    {
+      "epoch": 0.4031760715386987,
+      "grad_norm": 3.3389744758605957,
+      "learning_rate": 6.772126465241663e-06,
+      "loss": 0.9816,
+      "step": 5230
+    },
+    {
+      "epoch": 0.4032531606537157,
+      "grad_norm": 3.3862054347991943,
+      "learning_rate": 6.7709590065494125e-06,
+      "loss": 0.9248,
+      "step": 5231
+    },
+    {
+      "epoch": 0.40333024976873266,
+      "grad_norm": 3.3513505458831787,
+      "learning_rate": 6.769791437447042e-06,
+      "loss": 1.0087,
+      "step": 5232
+    },
+    {
+      "epoch": 0.40340733888374963,
+      "grad_norm": 3.7678074836730957,
+      "learning_rate": 6.76862375800734e-06,
+      "loss": 0.8853,
+      "step": 5233
+    },
+    {
+      "epoch": 0.40348442799876655,
+      "grad_norm": 3.6383919715881348,
+      "learning_rate": 6.767455968303107e-06,
+      "loss": 0.8316,
+      "step": 5234
+    },
+    {
+      "epoch": 0.4035615171137835,
+      "grad_norm": 3.6129610538482666,
+      "learning_rate": 6.7662880684071495e-06,
+      "loss": 1.0224,
+      "step": 5235
+    },
+    {
+      "epoch": 0.4036386062288005,
+      "grad_norm": 3.6119580268859863,
+      "learning_rate": 6.765120058392278e-06,
+      "loss": 1.0199,
+      "step": 5236
+    },
+    {
+      "epoch": 0.40371569534381746,
+      "grad_norm": 3.5133492946624756,
+      "learning_rate": 6.763951938331313e-06,
+      "loss": 0.8906,
+      "step": 5237
+    },
+    {
+      "epoch": 0.40379278445883443,
+      "grad_norm": 3.35998272895813,
+      "learning_rate": 6.76278370829708e-06,
+      "loss": 0.9159,
+      "step": 5238
+    },
+    {
+      "epoch": 0.40386987357385135,
+      "grad_norm": 3.517707586288452,
+      "learning_rate": 6.761615368362412e-06,
+      "loss": 1.0151,
+      "step": 5239
+    },
+    {
+      "epoch": 0.4039469626888683,
+      "grad_norm": 3.719403028488159,
+      "learning_rate": 6.760446918600151e-06,
+      "loss": 1.0202,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4040240518038853,
+      "grad_norm": 3.614016056060791,
+      "learning_rate": 6.759278359083141e-06,
+      "loss": 1.0067,
+      "step": 5241
+    },
+    {
+      "epoch": 0.40410114091890226,
+      "grad_norm": 3.3453943729400635,
+      "learning_rate": 6.758109689884236e-06,
+      "loss": 1.0046,
+      "step": 5242
+    },
+    {
+      "epoch": 0.40417823003391923,
+      "grad_norm": 3.9464776515960693,
+      "learning_rate": 6.756940911076299e-06,
+      "loss": 1.0484,
+      "step": 5243
+    },
+    {
+      "epoch": 0.40425531914893614,
+      "grad_norm": 3.7068941593170166,
+      "learning_rate": 6.755772022732194e-06,
+      "loss": 0.9906,
+      "step": 5244
+    },
+    {
+      "epoch": 0.4043324082639531,
+      "grad_norm": 3.643101215362549,
+      "learning_rate": 6.754603024924799e-06,
+      "loss": 1.0175,
+      "step": 5245
+    },
+    {
+      "epoch": 0.4044094973789701,
+      "grad_norm": 3.6078343391418457,
+      "learning_rate": 6.753433917726991e-06,
+      "loss": 1.0381,
+      "step": 5246
+    },
+    {
+      "epoch": 0.40448658649398705,
+      "grad_norm": 3.2811944484710693,
+      "learning_rate": 6.75226470121166e-06,
+      "loss": 0.946,
+      "step": 5247
+    },
+    {
+      "epoch": 0.404563675609004,
+      "grad_norm": 3.3540406227111816,
+      "learning_rate": 6.751095375451699e-06,
+      "loss": 1.0243,
+      "step": 5248
+    },
+    {
+      "epoch": 0.404640764724021,
+      "grad_norm": 3.783407211303711,
+      "learning_rate": 6.749925940520012e-06,
+      "loss": 1.0053,
+      "step": 5249
+    },
+    {
+      "epoch": 0.4047178538390379,
+      "grad_norm": 3.748215913772583,
+      "learning_rate": 6.7487563964895066e-06,
+      "loss": 0.9892,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4047949429540549,
+      "grad_norm": 4.270641803741455,
+      "learning_rate": 6.747586743433096e-06,
+      "loss": 1.1076,
+      "step": 5251
+    },
+    {
+      "epoch": 0.40487203206907185,
+      "grad_norm": 3.66277813911438,
+      "learning_rate": 6.746416981423701e-06,
+      "loss": 0.9607,
+      "step": 5252
+    },
+    {
+      "epoch": 0.4049491211840888,
+      "grad_norm": 3.703913927078247,
+      "learning_rate": 6.7452471105342536e-06,
+      "loss": 0.9762,
+      "step": 5253
+    },
+    {
+      "epoch": 0.4050262102991058,
+      "grad_norm": 3.490093469619751,
+      "learning_rate": 6.744077130837687e-06,
+      "loss": 0.943,
+      "step": 5254
+    },
+    {
+      "epoch": 0.4051032994141227,
+      "grad_norm": 3.721001625061035,
+      "learning_rate": 6.742907042406945e-06,
+      "loss": 1.0145,
+      "step": 5255
+    },
+    {
+      "epoch": 0.4051803885291397,
+      "grad_norm": 3.3705310821533203,
+      "learning_rate": 6.741736845314977e-06,
+      "loss": 0.9094,
+      "step": 5256
+    },
+    {
+      "epoch": 0.40525747764415665,
+      "grad_norm": 3.7736752033233643,
+      "learning_rate": 6.7405665396347345e-06,
+      "loss": 0.8783,
+      "step": 5257
+    },
+    {
+      "epoch": 0.4053345667591736,
+      "grad_norm": 3.2952444553375244,
+      "learning_rate": 6.739396125439184e-06,
+      "loss": 0.9976,
+      "step": 5258
+    },
+    {
+      "epoch": 0.4054116558741906,
+      "grad_norm": 3.4504001140594482,
+      "learning_rate": 6.7382256028012945e-06,
+      "loss": 0.9702,
+      "step": 5259
+    },
+    {
+      "epoch": 0.4054887449892075,
+      "grad_norm": 3.792487621307373,
+      "learning_rate": 6.7370549717940405e-06,
+      "loss": 1.0202,
+      "step": 5260
+    },
+    {
+      "epoch": 0.4055658341042245,
+      "grad_norm": 3.9348297119140625,
+      "learning_rate": 6.735884232490405e-06,
+      "loss": 1.0302,
+      "step": 5261
+    },
+    {
+      "epoch": 0.40564292321924145,
+      "grad_norm": 4.174359321594238,
+      "learning_rate": 6.734713384963379e-06,
+      "loss": 1.0545,
+      "step": 5262
+    },
+    {
+      "epoch": 0.4057200123342584,
+      "grad_norm": 3.5805842876434326,
+      "learning_rate": 6.733542429285957e-06,
+      "loss": 1.0303,
+      "step": 5263
+    },
+    {
+      "epoch": 0.4057971014492754,
+      "grad_norm": 3.4815800189971924,
+      "learning_rate": 6.732371365531144e-06,
+      "loss": 0.9535,
+      "step": 5264
+    },
+    {
+      "epoch": 0.4058741905642923,
+      "grad_norm": 3.540489435195923,
+      "learning_rate": 6.731200193771947e-06,
+      "loss": 0.9182,
+      "step": 5265
+    },
+    {
+      "epoch": 0.4059512796793093,
+      "grad_norm": 3.4096593856811523,
+      "learning_rate": 6.730028914081384e-06,
+      "loss": 1.0019,
+      "step": 5266
+    },
+    {
+      "epoch": 0.40602836879432624,
+      "grad_norm": 3.462355136871338,
+      "learning_rate": 6.72885752653248e-06,
+      "loss": 0.9894,
+      "step": 5267
+    },
+    {
+      "epoch": 0.4061054579093432,
+      "grad_norm": 3.640557289123535,
+      "learning_rate": 6.7276860311982614e-06,
+      "loss": 0.9686,
+      "step": 5268
+    },
+    {
+      "epoch": 0.4061825470243602,
+      "grad_norm": 3.729519844055176,
+      "learning_rate": 6.726514428151767e-06,
+      "loss": 1.0958,
+      "step": 5269
+    },
+    {
+      "epoch": 0.4062596361393771,
+      "grad_norm": 3.461050033569336,
+      "learning_rate": 6.725342717466041e-06,
+      "loss": 0.961,
+      "step": 5270
+    },
+    {
+      "epoch": 0.40633672525439407,
+      "grad_norm": 3.6534311771392822,
+      "learning_rate": 6.724170899214131e-06,
+      "loss": 1.0216,
+      "step": 5271
+    },
+    {
+      "epoch": 0.40641381436941104,
+      "grad_norm": 3.3762400150299072,
+      "learning_rate": 6.7229989734690956e-06,
+      "loss": 0.9508,
+      "step": 5272
+    },
+    {
+      "epoch": 0.406490903484428,
+      "grad_norm": 3.6674129962921143,
+      "learning_rate": 6.721826940303999e-06,
+      "loss": 0.8763,
+      "step": 5273
+    },
+    {
+      "epoch": 0.406567992599445,
+      "grad_norm": 3.514495849609375,
+      "learning_rate": 6.720654799791908e-06,
+      "loss": 0.9273,
+      "step": 5274
+    },
+    {
+      "epoch": 0.4066450817144619,
+      "grad_norm": 3.849905490875244,
+      "learning_rate": 6.719482552005903e-06,
+      "loss": 1.0703,
+      "step": 5275
+    },
+    {
+      "epoch": 0.40672217082947887,
+      "grad_norm": 3.744799852371216,
+      "learning_rate": 6.718310197019066e-06,
+      "loss": 1.0608,
+      "step": 5276
+    },
+    {
+      "epoch": 0.40679925994449584,
+      "grad_norm": 3.5321202278137207,
+      "learning_rate": 6.7171377349044875e-06,
+      "loss": 0.9747,
+      "step": 5277
+    },
+    {
+      "epoch": 0.4068763490595128,
+      "grad_norm": 3.6646018028259277,
+      "learning_rate": 6.7159651657352656e-06,
+      "loss": 1.0208,
+      "step": 5278
+    },
+    {
+      "epoch": 0.4069534381745298,
+      "grad_norm": 3.462653636932373,
+      "learning_rate": 6.714792489584501e-06,
+      "loss": 0.9575,
+      "step": 5279
+    },
+    {
+      "epoch": 0.4070305272895467,
+      "grad_norm": 3.5436105728149414,
+      "learning_rate": 6.7136197065253075e-06,
+      "loss": 0.9758,
+      "step": 5280
+    },
+    {
+      "epoch": 0.40710761640456367,
+      "grad_norm": 3.5151219367980957,
+      "learning_rate": 6.7124468166308e-06,
+      "loss": 1.0746,
+      "step": 5281
+    },
+    {
+      "epoch": 0.40718470551958064,
+      "grad_norm": 3.4494199752807617,
+      "learning_rate": 6.711273819974101e-06,
+      "loss": 0.9407,
+      "step": 5282
+    },
+    {
+      "epoch": 0.4072617946345976,
+      "grad_norm": 3.708519220352173,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 1.044,
+      "step": 5283
+    },
+    {
+      "epoch": 0.4073388837496146,
+      "grad_norm": 3.8825523853302,
+      "learning_rate": 6.708927506666664e-06,
+      "loss": 0.9725,
+      "step": 5284
+    },
+    {
+      "epoch": 0.4074159728646315,
+      "grad_norm": 3.7221336364746094,
+      "learning_rate": 6.707754190162203e-06,
+      "loss": 0.9684,
+      "step": 5285
+    },
+    {
+      "epoch": 0.40749306197964846,
+      "grad_norm": 3.6459972858428955,
+      "learning_rate": 6.7065807671881155e-06,
+      "loss": 1.0946,
+      "step": 5286
+    },
+    {
+      "epoch": 0.40757015109466543,
+      "grad_norm": 3.2291955947875977,
+      "learning_rate": 6.705407237817554e-06,
+      "loss": 0.8986,
+      "step": 5287
+    },
+    {
+      "epoch": 0.4076472402096824,
+      "grad_norm": 3.8956899642944336,
+      "learning_rate": 6.704233602123685e-06,
+      "loss": 1.0724,
+      "step": 5288
+    },
+    {
+      "epoch": 0.4077243293246994,
+      "grad_norm": 3.887136459350586,
+      "learning_rate": 6.703059860179677e-06,
+      "loss": 1.0851,
+      "step": 5289
+    },
+    {
+      "epoch": 0.4078014184397163,
+      "grad_norm": 3.6385395526885986,
+      "learning_rate": 6.701886012058706e-06,
+      "loss": 0.9677,
+      "step": 5290
+    },
+    {
+      "epoch": 0.40787850755473326,
+      "grad_norm": 3.614091634750366,
+      "learning_rate": 6.700712057833958e-06,
+      "loss": 0.991,
+      "step": 5291
+    },
+    {
+      "epoch": 0.40795559666975023,
+      "grad_norm": 3.7074344158172607,
+      "learning_rate": 6.69953799757862e-06,
+      "loss": 0.8541,
+      "step": 5292
+    },
+    {
+      "epoch": 0.4080326857847672,
+      "grad_norm": 3.504150152206421,
+      "learning_rate": 6.69836383136589e-06,
+      "loss": 0.942,
+      "step": 5293
+    },
+    {
+      "epoch": 0.4081097748997842,
+      "grad_norm": 3.62353777885437,
+      "learning_rate": 6.697189559268973e-06,
+      "loss": 0.9787,
+      "step": 5294
+    },
+    {
+      "epoch": 0.4081868640148011,
+      "grad_norm": 3.5454580783843994,
+      "learning_rate": 6.696015181361076e-06,
+      "loss": 1.0769,
+      "step": 5295
+    },
+    {
+      "epoch": 0.40826395312981806,
+      "grad_norm": 3.8468873500823975,
+      "learning_rate": 6.694840697715415e-06,
+      "loss": 1.0924,
+      "step": 5296
+    },
+    {
+      "epoch": 0.40834104224483503,
+      "grad_norm": 3.6607978343963623,
+      "learning_rate": 6.693666108405215e-06,
+      "loss": 0.9552,
+      "step": 5297
+    },
+    {
+      "epoch": 0.408418131359852,
+      "grad_norm": 3.949882984161377,
+      "learning_rate": 6.692491413503704e-06,
+      "loss": 0.9816,
+      "step": 5298
+    },
+    {
+      "epoch": 0.40849522047486897,
+      "grad_norm": 3.432330369949341,
+      "learning_rate": 6.691316613084121e-06,
+      "loss": 0.8183,
+      "step": 5299
+    },
+    {
+      "epoch": 0.4085723095898859,
+      "grad_norm": 3.6200854778289795,
+      "learning_rate": 6.690141707219706e-06,
+      "loss": 0.992,
+      "step": 5300
+    },
+    {
+      "epoch": 0.40864939870490286,
+      "grad_norm": 3.9913625717163086,
+      "learning_rate": 6.688966695983708e-06,
+      "loss": 0.9583,
+      "step": 5301
+    },
+    {
+      "epoch": 0.4087264878199198,
+      "grad_norm": 3.4617879390716553,
+      "learning_rate": 6.6877915794493855e-06,
+      "loss": 0.9242,
+      "step": 5302
+    },
+    {
+      "epoch": 0.4088035769349368,
+      "grad_norm": 3.69291615486145,
+      "learning_rate": 6.6866163576899985e-06,
+      "loss": 0.9768,
+      "step": 5303
+    },
+    {
+      "epoch": 0.40888066604995377,
+      "grad_norm": 3.7546987533569336,
+      "learning_rate": 6.6854410307788175e-06,
+      "loss": 0.9785,
+      "step": 5304
+    },
+    {
+      "epoch": 0.4089577551649707,
+      "grad_norm": 3.5481560230255127,
+      "learning_rate": 6.684265598789117e-06,
+      "loss": 1.0147,
+      "step": 5305
+    },
+    {
+      "epoch": 0.40903484427998765,
+      "grad_norm": 4.156661510467529,
+      "learning_rate": 6.683090061794179e-06,
+      "loss": 0.9727,
+      "step": 5306
+    },
+    {
+      "epoch": 0.4091119333950046,
+      "grad_norm": 3.6532485485076904,
+      "learning_rate": 6.6819144198672925e-06,
+      "loss": 0.9808,
+      "step": 5307
+    },
+    {
+      "epoch": 0.4091890225100216,
+      "grad_norm": 3.576277494430542,
+      "learning_rate": 6.6807386730817525e-06,
+      "loss": 0.9245,
+      "step": 5308
+    },
+    {
+      "epoch": 0.40926611162503856,
+      "grad_norm": 3.498910903930664,
+      "learning_rate": 6.679562821510862e-06,
+      "loss": 0.9565,
+      "step": 5309
+    },
+    {
+      "epoch": 0.4093432007400555,
+      "grad_norm": 3.947864294052124,
+      "learning_rate": 6.6783868652279285e-06,
+      "loss": 1.0307,
+      "step": 5310
+    },
+    {
+      "epoch": 0.40942028985507245,
+      "grad_norm": 3.714496612548828,
+      "learning_rate": 6.677210804306266e-06,
+      "loss": 0.9651,
+      "step": 5311
+    },
+    {
+      "epoch": 0.4094973789700894,
+      "grad_norm": 3.6817400455474854,
+      "learning_rate": 6.676034638819197e-06,
+      "loss": 0.932,
+      "step": 5312
+    },
+    {
+      "epoch": 0.4095744680851064,
+      "grad_norm": 3.330031394958496,
+      "learning_rate": 6.674858368840047e-06,
+      "loss": 0.964,
+      "step": 5313
+    },
+    {
+      "epoch": 0.40965155720012336,
+      "grad_norm": 3.4984524250030518,
+      "learning_rate": 6.673681994442153e-06,
+      "loss": 0.9394,
+      "step": 5314
+    },
+    {
+      "epoch": 0.4097286463151403,
+      "grad_norm": 3.6142520904541016,
+      "learning_rate": 6.6725055156988545e-06,
+      "loss": 1.0045,
+      "step": 5315
+    },
+    {
+      "epoch": 0.40980573543015725,
+      "grad_norm": 3.7212975025177,
+      "learning_rate": 6.671328932683499e-06,
+      "loss": 1.0474,
+      "step": 5316
+    },
+    {
+      "epoch": 0.4098828245451742,
+      "grad_norm": 3.747412919998169,
+      "learning_rate": 6.670152245469441e-06,
+      "loss": 1.0149,
+      "step": 5317
+    },
+    {
+      "epoch": 0.4099599136601912,
+      "grad_norm": 3.9034531116485596,
+      "learning_rate": 6.6689754541300426e-06,
+      "loss": 1.1157,
+      "step": 5318
+    },
+    {
+      "epoch": 0.41003700277520816,
+      "grad_norm": 4.028858184814453,
+      "learning_rate": 6.667798558738664e-06,
+      "loss": 1.0,
+      "step": 5319
+    },
+    {
+      "epoch": 0.4101140918902251,
+      "grad_norm": 3.367152690887451,
+      "learning_rate": 6.666621559368687e-06,
+      "loss": 0.9779,
+      "step": 5320
+    },
+    {
+      "epoch": 0.41019118100524204,
+      "grad_norm": 5.969235420227051,
+      "learning_rate": 6.665444456093485e-06,
+      "loss": 1.0506,
+      "step": 5321
+    },
+    {
+      "epoch": 0.410268270120259,
+      "grad_norm": 3.4636454582214355,
+      "learning_rate": 6.664267248986447e-06,
+      "loss": 0.941,
+      "step": 5322
+    },
+    {
+      "epoch": 0.410345359235276,
+      "grad_norm": 3.7874643802642822,
+      "learning_rate": 6.663089938120967e-06,
+      "loss": 1.089,
+      "step": 5323
+    },
+    {
+      "epoch": 0.41042244835029296,
+      "grad_norm": 3.6759419441223145,
+      "learning_rate": 6.6619125235704414e-06,
+      "loss": 0.9959,
+      "step": 5324
+    },
+    {
+      "epoch": 0.41049953746530987,
+      "grad_norm": 3.6804399490356445,
+      "learning_rate": 6.660735005408278e-06,
+      "loss": 1.0022,
+      "step": 5325
+    },
+    {
+      "epoch": 0.41057662658032684,
+      "grad_norm": 3.5297484397888184,
+      "learning_rate": 6.659557383707887e-06,
+      "loss": 1.0264,
+      "step": 5326
+    },
+    {
+      "epoch": 0.4106537156953438,
+      "grad_norm": 3.6392202377319336,
+      "learning_rate": 6.65837965854269e-06,
+      "loss": 0.9898,
+      "step": 5327
+    },
+    {
+      "epoch": 0.4107308048103608,
+      "grad_norm": 3.438964366912842,
+      "learning_rate": 6.65720182998611e-06,
+      "loss": 0.9249,
+      "step": 5328
+    },
+    {
+      "epoch": 0.41080789392537775,
+      "grad_norm": 3.7228152751922607,
+      "learning_rate": 6.656023898111577e-06,
+      "loss": 1.0178,
+      "step": 5329
+    },
+    {
+      "epoch": 0.41088498304039467,
+      "grad_norm": 3.804844856262207,
+      "learning_rate": 6.654845862992532e-06,
+      "loss": 0.8827,
+      "step": 5330
+    },
+    {
+      "epoch": 0.41096207215541164,
+      "grad_norm": 3.4772403240203857,
+      "learning_rate": 6.653667724702419e-06,
+      "loss": 1.0009,
+      "step": 5331
+    },
+    {
+      "epoch": 0.4110391612704286,
+      "grad_norm": 3.5902342796325684,
+      "learning_rate": 6.652489483314686e-06,
+      "loss": 1.0335,
+      "step": 5332
+    },
+    {
+      "epoch": 0.4111162503854456,
+      "grad_norm": 3.510636329650879,
+      "learning_rate": 6.651311138902792e-06,
+      "loss": 1.0807,
+      "step": 5333
+    },
+    {
+      "epoch": 0.41119333950046255,
+      "grad_norm": 3.3497116565704346,
+      "learning_rate": 6.650132691540203e-06,
+      "loss": 0.9117,
+      "step": 5334
+    },
+    {
+      "epoch": 0.4112704286154795,
+      "grad_norm": 3.712567090988159,
+      "learning_rate": 6.648954141300386e-06,
+      "loss": 0.9818,
+      "step": 5335
+    },
+    {
+      "epoch": 0.41134751773049644,
+      "grad_norm": 3.705669641494751,
+      "learning_rate": 6.647775488256818e-06,
+      "loss": 0.957,
+      "step": 5336
+    },
+    {
+      "epoch": 0.4114246068455134,
+      "grad_norm": 3.471576452255249,
+      "learning_rate": 6.646596732482982e-06,
+      "loss": 0.9049,
+      "step": 5337
+    },
+    {
+      "epoch": 0.4115016959605304,
+      "grad_norm": 3.4686636924743652,
+      "learning_rate": 6.645417874052368e-06,
+      "loss": 1.0353,
+      "step": 5338
+    },
+    {
+      "epoch": 0.41157878507554735,
+      "grad_norm": 3.8832995891571045,
+      "learning_rate": 6.644238913038472e-06,
+      "loss": 1.0707,
+      "step": 5339
+    },
+    {
+      "epoch": 0.4116558741905643,
+      "grad_norm": 3.786958694458008,
+      "learning_rate": 6.643059849514795e-06,
+      "loss": 1.0643,
+      "step": 5340
+    },
+    {
+      "epoch": 0.41173296330558123,
+      "grad_norm": 3.698564052581787,
+      "learning_rate": 6.641880683554846e-06,
+      "loss": 1.0087,
+      "step": 5341
+    },
+    {
+      "epoch": 0.4118100524205982,
+      "grad_norm": 3.6572458744049072,
+      "learning_rate": 6.640701415232139e-06,
+      "loss": 0.9285,
+      "step": 5342
+    },
+    {
+      "epoch": 0.4118871415356152,
+      "grad_norm": 3.6328537464141846,
+      "learning_rate": 6.639522044620197e-06,
+      "loss": 1.0668,
+      "step": 5343
+    },
+    {
+      "epoch": 0.41196423065063215,
+      "grad_norm": 3.2839369773864746,
+      "learning_rate": 6.638342571792548e-06,
+      "loss": 0.8407,
+      "step": 5344
+    },
+    {
+      "epoch": 0.4120413197656491,
+      "grad_norm": 3.6459763050079346,
+      "learning_rate": 6.637162996822724e-06,
+      "loss": 0.9962,
+      "step": 5345
+    },
+    {
+      "epoch": 0.41211840888066603,
+      "grad_norm": 3.8563995361328125,
+      "learning_rate": 6.635983319784265e-06,
+      "loss": 0.9631,
+      "step": 5346
+    },
+    {
+      "epoch": 0.412195497995683,
+      "grad_norm": 3.544633388519287,
+      "learning_rate": 6.63480354075072e-06,
+      "loss": 0.8998,
+      "step": 5347
+    },
+    {
+      "epoch": 0.4122725871107,
+      "grad_norm": 3.736931085586548,
+      "learning_rate": 6.633623659795642e-06,
+      "loss": 0.985,
+      "step": 5348
+    },
+    {
+      "epoch": 0.41234967622571694,
+      "grad_norm": 3.825957775115967,
+      "learning_rate": 6.632443676992588e-06,
+      "loss": 0.9953,
+      "step": 5349
+    },
+    {
+      "epoch": 0.4124267653407339,
+      "grad_norm": 3.397263288497925,
+      "learning_rate": 6.631263592415127e-06,
+      "loss": 0.9607,
+      "step": 5350
+    },
+    {
+      "epoch": 0.41250385445575083,
+      "grad_norm": 3.639387845993042,
+      "learning_rate": 6.630083406136829e-06,
+      "loss": 0.9568,
+      "step": 5351
+    },
+    {
+      "epoch": 0.4125809435707678,
+      "grad_norm": 3.8172554969787598,
+      "learning_rate": 6.628903118231274e-06,
+      "loss": 0.9392,
+      "step": 5352
+    },
+    {
+      "epoch": 0.41265803268578477,
+      "grad_norm": 3.8433003425598145,
+      "learning_rate": 6.627722728772044e-06,
+      "loss": 0.987,
+      "step": 5353
+    },
+    {
+      "epoch": 0.41273512180080174,
+      "grad_norm": 3.482104778289795,
+      "learning_rate": 6.6265422378327314e-06,
+      "loss": 0.9403,
+      "step": 5354
+    },
+    {
+      "epoch": 0.4128122109158187,
+      "grad_norm": 3.8958816528320312,
+      "learning_rate": 6.625361645486936e-06,
+      "loss": 0.9892,
+      "step": 5355
+    },
+    {
+      "epoch": 0.4128893000308356,
+      "grad_norm": 3.7703473567962646,
+      "learning_rate": 6.62418095180826e-06,
+      "loss": 0.951,
+      "step": 5356
+    },
+    {
+      "epoch": 0.4129663891458526,
+      "grad_norm": 3.737750291824341,
+      "learning_rate": 6.623000156870313e-06,
+      "loss": 0.9753,
+      "step": 5357
+    },
+    {
+      "epoch": 0.41304347826086957,
+      "grad_norm": 3.6272356510162354,
+      "learning_rate": 6.621819260746713e-06,
+      "loss": 1.0136,
+      "step": 5358
+    },
+    {
+      "epoch": 0.41312056737588654,
+      "grad_norm": 3.797292470932007,
+      "learning_rate": 6.62063826351108e-06,
+      "loss": 1.0478,
+      "step": 5359
+    },
+    {
+      "epoch": 0.4131976564909035,
+      "grad_norm": 4.469625949859619,
+      "learning_rate": 6.619457165237046e-06,
+      "loss": 1.0983,
+      "step": 5360
+    },
+    {
+      "epoch": 0.4132747456059204,
+      "grad_norm": 3.5453269481658936,
+      "learning_rate": 6.618275965998244e-06,
+      "loss": 1.0379,
+      "step": 5361
+    },
+    {
+      "epoch": 0.4133518347209374,
+      "grad_norm": 3.681480646133423,
+      "learning_rate": 6.617094665868317e-06,
+      "loss": 0.9983,
+      "step": 5362
+    },
+    {
+      "epoch": 0.41342892383595437,
+      "grad_norm": 3.6391947269439697,
+      "learning_rate": 6.615913264920912e-06,
+      "loss": 0.9968,
+      "step": 5363
+    },
+    {
+      "epoch": 0.41350601295097134,
+      "grad_norm": 3.4914567470550537,
+      "learning_rate": 6.614731763229686e-06,
+      "loss": 0.9064,
+      "step": 5364
+    },
+    {
+      "epoch": 0.4135831020659883,
+      "grad_norm": 3.5731704235076904,
+      "learning_rate": 6.613550160868297e-06,
+      "loss": 1.0242,
+      "step": 5365
+    },
+    {
+      "epoch": 0.4136601911810052,
+      "grad_norm": 4.226243019104004,
+      "learning_rate": 6.612368457910412e-06,
+      "loss": 1.0554,
+      "step": 5366
+    },
+    {
+      "epoch": 0.4137372802960222,
+      "grad_norm": 3.477130889892578,
+      "learning_rate": 6.611186654429704e-06,
+      "loss": 0.8728,
+      "step": 5367
+    },
+    {
+      "epoch": 0.41381436941103916,
+      "grad_norm": 3.375094175338745,
+      "learning_rate": 6.6100047504998535e-06,
+      "loss": 0.9865,
+      "step": 5368
+    },
+    {
+      "epoch": 0.41389145852605613,
+      "grad_norm": 3.527829647064209,
+      "learning_rate": 6.6088227461945434e-06,
+      "loss": 1.0235,
+      "step": 5369
+    },
+    {
+      "epoch": 0.4139685476410731,
+      "grad_norm": 3.82974910736084,
+      "learning_rate": 6.607640641587469e-06,
+      "loss": 0.9777,
+      "step": 5370
+    },
+    {
+      "epoch": 0.41404563675609,
+      "grad_norm": 3.5548770427703857,
+      "learning_rate": 6.606458436752327e-06,
+      "loss": 0.965,
+      "step": 5371
+    },
+    {
+      "epoch": 0.414122725871107,
+      "grad_norm": 3.665213108062744,
+      "learning_rate": 6.60527613176282e-06,
+      "loss": 0.9712,
+      "step": 5372
+    },
+    {
+      "epoch": 0.41419981498612396,
+      "grad_norm": 3.664933443069458,
+      "learning_rate": 6.604093726692661e-06,
+      "loss": 0.91,
+      "step": 5373
+    },
+    {
+      "epoch": 0.41427690410114093,
+      "grad_norm": 3.6058082580566406,
+      "learning_rate": 6.602911221615567e-06,
+      "loss": 1.0525,
+      "step": 5374
+    },
+    {
+      "epoch": 0.4143539932161579,
+      "grad_norm": 3.68048095703125,
+      "learning_rate": 6.601728616605259e-06,
+      "loss": 0.9375,
+      "step": 5375
+    },
+    {
+      "epoch": 0.4144310823311748,
+      "grad_norm": 3.4503397941589355,
+      "learning_rate": 6.600545911735468e-06,
+      "loss": 0.9779,
+      "step": 5376
+    },
+    {
+      "epoch": 0.4145081714461918,
+      "grad_norm": 3.573887586593628,
+      "learning_rate": 6.599363107079927e-06,
+      "loss": 0.9744,
+      "step": 5377
+    },
+    {
+      "epoch": 0.41458526056120876,
+      "grad_norm": 3.5022692680358887,
+      "learning_rate": 6.59818020271238e-06,
+      "loss": 0.9208,
+      "step": 5378
+    },
+    {
+      "epoch": 0.4146623496762257,
+      "grad_norm": 3.4239299297332764,
+      "learning_rate": 6.596997198706576e-06,
+      "loss": 0.9342,
+      "step": 5379
+    },
+    {
+      "epoch": 0.4147394387912427,
+      "grad_norm": 4.041162014007568,
+      "learning_rate": 6.595814095136267e-06,
+      "loss": 0.8855,
+      "step": 5380
+    },
+    {
+      "epoch": 0.4148165279062596,
+      "grad_norm": 4.2460761070251465,
+      "learning_rate": 6.594630892075213e-06,
+      "loss": 1.0153,
+      "step": 5381
+    },
+    {
+      "epoch": 0.4148936170212766,
+      "grad_norm": 3.367917537689209,
+      "learning_rate": 6.593447589597184e-06,
+      "loss": 0.9544,
+      "step": 5382
+    },
+    {
+      "epoch": 0.41497070613629355,
+      "grad_norm": 3.5523362159729004,
+      "learning_rate": 6.5922641877759484e-06,
+      "loss": 0.9909,
+      "step": 5383
+    },
+    {
+      "epoch": 0.4150477952513105,
+      "grad_norm": 3.4318056106567383,
+      "learning_rate": 6.591080686685289e-06,
+      "loss": 0.9482,
+      "step": 5384
+    },
+    {
+      "epoch": 0.4151248843663275,
+      "grad_norm": 3.5694563388824463,
+      "learning_rate": 6.589897086398989e-06,
+      "loss": 0.9606,
+      "step": 5385
+    },
+    {
+      "epoch": 0.4152019734813444,
+      "grad_norm": 3.1972849369049072,
+      "learning_rate": 6.588713386990837e-06,
+      "loss": 0.9337,
+      "step": 5386
+    },
+    {
+      "epoch": 0.4152790625963614,
+      "grad_norm": 3.958773136138916,
+      "learning_rate": 6.5875295885346356e-06,
+      "loss": 0.8853,
+      "step": 5387
+    },
+    {
+      "epoch": 0.41535615171137835,
+      "grad_norm": 4.037240982055664,
+      "learning_rate": 6.5863456911041865e-06,
+      "loss": 1.0532,
+      "step": 5388
+    },
+    {
+      "epoch": 0.4154332408263953,
+      "grad_norm": 3.808276891708374,
+      "learning_rate": 6.585161694773301e-06,
+      "loss": 1.0308,
+      "step": 5389
+    },
+    {
+      "epoch": 0.4155103299414123,
+      "grad_norm": 3.6094319820404053,
+      "learning_rate": 6.583977599615792e-06,
+      "loss": 1.0333,
+      "step": 5390
+    },
+    {
+      "epoch": 0.4155874190564292,
+      "grad_norm": 3.879518747329712,
+      "learning_rate": 6.582793405705482e-06,
+      "loss": 1.0241,
+      "step": 5391
+    },
+    {
+      "epoch": 0.4156645081714462,
+      "grad_norm": 3.571319818496704,
+      "learning_rate": 6.581609113116203e-06,
+      "loss": 0.9956,
+      "step": 5392
+    },
+    {
+      "epoch": 0.41574159728646315,
+      "grad_norm": 3.7715680599212646,
+      "learning_rate": 6.580424721921785e-06,
+      "loss": 0.9709,
+      "step": 5393
+    },
+    {
+      "epoch": 0.4158186864014801,
+      "grad_norm": 3.707052230834961,
+      "learning_rate": 6.579240232196073e-06,
+      "loss": 0.9661,
+      "step": 5394
+    },
+    {
+      "epoch": 0.4158957755164971,
+      "grad_norm": 3.3030142784118652,
+      "learning_rate": 6.578055644012911e-06,
+      "loss": 0.9728,
+      "step": 5395
+    },
+    {
+      "epoch": 0.415972864631514,
+      "grad_norm": 3.5438342094421387,
+      "learning_rate": 6.576870957446151e-06,
+      "loss": 0.9302,
+      "step": 5396
+    },
+    {
+      "epoch": 0.416049953746531,
+      "grad_norm": 3.5697009563446045,
+      "learning_rate": 6.575686172569655e-06,
+      "loss": 1.1335,
+      "step": 5397
+    },
+    {
+      "epoch": 0.41612704286154795,
+      "grad_norm": 3.6315059661865234,
+      "learning_rate": 6.574501289457287e-06,
+      "loss": 1.1065,
+      "step": 5398
+    },
+    {
+      "epoch": 0.4162041319765649,
+      "grad_norm": 3.3344013690948486,
+      "learning_rate": 6.573316308182917e-06,
+      "loss": 0.9045,
+      "step": 5399
+    },
+    {
+      "epoch": 0.4162812210915819,
+      "grad_norm": 3.6112499237060547,
+      "learning_rate": 6.5721312288204254e-06,
+      "loss": 1.0306,
+      "step": 5400
+    },
+    {
+      "epoch": 0.4163583102065988,
+      "grad_norm": 3.63188099861145,
+      "learning_rate": 6.570946051443693e-06,
+      "loss": 0.9627,
+      "step": 5401
+    },
+    {
+      "epoch": 0.4164353993216158,
+      "grad_norm": 3.795747995376587,
+      "learning_rate": 6.5697607761266105e-06,
+      "loss": 1.1105,
+      "step": 5402
+    },
+    {
+      "epoch": 0.41651248843663274,
+      "grad_norm": 3.7781665325164795,
+      "learning_rate": 6.568575402943073e-06,
+      "loss": 0.9043,
+      "step": 5403
+    },
+    {
+      "epoch": 0.4165895775516497,
+      "grad_norm": 3.3593337535858154,
+      "learning_rate": 6.567389931966983e-06,
+      "loss": 1.0182,
+      "step": 5404
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 4.008146286010742,
+      "learning_rate": 6.566204363272248e-06,
+      "loss": 1.0122,
+      "step": 5405
+    },
+    {
+      "epoch": 0.4167437557816836,
+      "grad_norm": 3.6243855953216553,
+      "learning_rate": 6.565018696932786e-06,
+      "loss": 0.9194,
+      "step": 5406
+    },
+    {
+      "epoch": 0.41682084489670057,
+      "grad_norm": 3.93573260307312,
+      "learning_rate": 6.5638329330225096e-06,
+      "loss": 0.9881,
+      "step": 5407
+    },
+    {
+      "epoch": 0.41689793401171754,
+      "grad_norm": 3.9312803745269775,
+      "learning_rate": 6.5626470716153515e-06,
+      "loss": 0.9584,
+      "step": 5408
+    },
+    {
+      "epoch": 0.4169750231267345,
+      "grad_norm": 3.7972381114959717,
+      "learning_rate": 6.561461112785239e-06,
+      "loss": 0.8669,
+      "step": 5409
+    },
+    {
+      "epoch": 0.4170521122417515,
+      "grad_norm": 3.8700931072235107,
+      "learning_rate": 6.5602750566061154e-06,
+      "loss": 1.0306,
+      "step": 5410
+    },
+    {
+      "epoch": 0.4171292013567684,
+      "grad_norm": 3.5603837966918945,
+      "learning_rate": 6.559088903151923e-06,
+      "loss": 1.0485,
+      "step": 5411
+    },
+    {
+      "epoch": 0.41720629047178537,
+      "grad_norm": 3.537443161010742,
+      "learning_rate": 6.5579026524966106e-06,
+      "loss": 1.0481,
+      "step": 5412
+    },
+    {
+      "epoch": 0.41728337958680234,
+      "grad_norm": 3.662034511566162,
+      "learning_rate": 6.5567163047141395e-06,
+      "loss": 1.0871,
+      "step": 5413
+    },
+    {
+      "epoch": 0.4173604687018193,
+      "grad_norm": 3.207895517349243,
+      "learning_rate": 6.555529859878466e-06,
+      "loss": 0.8474,
+      "step": 5414
+    },
+    {
+      "epoch": 0.4174375578168363,
+      "grad_norm": 3.5553321838378906,
+      "learning_rate": 6.554343318063563e-06,
+      "loss": 1.0164,
+      "step": 5415
+    },
+    {
+      "epoch": 0.4175146469318532,
+      "grad_norm": 4.284764766693115,
+      "learning_rate": 6.553156679343404e-06,
+      "loss": 0.9951,
+      "step": 5416
+    },
+    {
+      "epoch": 0.41759173604687017,
+      "grad_norm": 3.8002851009368896,
+      "learning_rate": 6.551969943791972e-06,
+      "loss": 1.0258,
+      "step": 5417
+    },
+    {
+      "epoch": 0.41766882516188714,
+      "grad_norm": 4.005855560302734,
+      "learning_rate": 6.550783111483249e-06,
+      "loss": 0.9425,
+      "step": 5418
+    },
+    {
+      "epoch": 0.4177459142769041,
+      "grad_norm": 3.5410430431365967,
+      "learning_rate": 6.549596182491233e-06,
+      "loss": 0.9774,
+      "step": 5419
+    },
+    {
+      "epoch": 0.4178230033919211,
+      "grad_norm": 3.704803943634033,
+      "learning_rate": 6.548409156889919e-06,
+      "loss": 0.9802,
+      "step": 5420
+    },
+    {
+      "epoch": 0.41790009250693805,
+      "grad_norm": 3.7099010944366455,
+      "learning_rate": 6.5472220347533145e-06,
+      "loss": 0.9581,
+      "step": 5421
+    },
+    {
+      "epoch": 0.41797718162195496,
+      "grad_norm": 3.5973217487335205,
+      "learning_rate": 6.546034816155429e-06,
+      "loss": 0.9321,
+      "step": 5422
+    },
+    {
+      "epoch": 0.41805427073697193,
+      "grad_norm": 4.097982406616211,
+      "learning_rate": 6.54484750117028e-06,
+      "loss": 1.0699,
+      "step": 5423
+    },
+    {
+      "epoch": 0.4181313598519889,
+      "grad_norm": 3.9153554439544678,
+      "learning_rate": 6.543660089871891e-06,
+      "loss": 0.9045,
+      "step": 5424
+    },
+    {
+      "epoch": 0.4182084489670059,
+      "grad_norm": 3.366682767868042,
+      "learning_rate": 6.5424725823342895e-06,
+      "loss": 1.0124,
+      "step": 5425
+    },
+    {
+      "epoch": 0.41828553808202285,
+      "grad_norm": 3.677605390548706,
+      "learning_rate": 6.54128497863151e-06,
+      "loss": 1.0661,
+      "step": 5426
+    },
+    {
+      "epoch": 0.41836262719703976,
+      "grad_norm": 3.49324369430542,
+      "learning_rate": 6.540097278837596e-06,
+      "loss": 0.9011,
+      "step": 5427
+    },
+    {
+      "epoch": 0.41843971631205673,
+      "grad_norm": 3.3589603900909424,
+      "learning_rate": 6.538909483026593e-06,
+      "loss": 0.9454,
+      "step": 5428
+    },
+    {
+      "epoch": 0.4185168054270737,
+      "grad_norm": 3.4064254760742188,
+      "learning_rate": 6.537721591272553e-06,
+      "loss": 0.9948,
+      "step": 5429
+    },
+    {
+      "epoch": 0.41859389454209067,
+      "grad_norm": 3.785304546356201,
+      "learning_rate": 6.536533603649536e-06,
+      "loss": 0.9664,
+      "step": 5430
+    },
+    {
+      "epoch": 0.41867098365710764,
+      "grad_norm": 3.5690832138061523,
+      "learning_rate": 6.5353455202316075e-06,
+      "loss": 0.9863,
+      "step": 5431
+    },
+    {
+      "epoch": 0.41874807277212456,
+      "grad_norm": 3.667205572128296,
+      "learning_rate": 6.5341573410928376e-06,
+      "loss": 1.0341,
+      "step": 5432
+    },
+    {
+      "epoch": 0.41882516188714153,
+      "grad_norm": 3.1317427158355713,
+      "learning_rate": 6.532969066307302e-06,
+      "loss": 0.9154,
+      "step": 5433
+    },
+    {
+      "epoch": 0.4189022510021585,
+      "grad_norm": 3.749356746673584,
+      "learning_rate": 6.531780695949086e-06,
+      "loss": 0.9678,
+      "step": 5434
+    },
+    {
+      "epoch": 0.41897934011717547,
+      "grad_norm": 3.7041513919830322,
+      "learning_rate": 6.530592230092276e-06,
+      "loss": 1.0266,
+      "step": 5435
+    },
+    {
+      "epoch": 0.41905642923219244,
+      "grad_norm": 3.199434518814087,
+      "learning_rate": 6.529403668810968e-06,
+      "loss": 0.964,
+      "step": 5436
+    },
+    {
+      "epoch": 0.41913351834720936,
+      "grad_norm": 3.5290520191192627,
+      "learning_rate": 6.528215012179262e-06,
+      "loss": 0.9869,
+      "step": 5437
+    },
+    {
+      "epoch": 0.4192106074622263,
+      "grad_norm": 3.9519081115722656,
+      "learning_rate": 6.527026260271265e-06,
+      "loss": 1.0839,
+      "step": 5438
+    },
+    {
+      "epoch": 0.4192876965772433,
+      "grad_norm": 3.520599603652954,
+      "learning_rate": 6.525837413161089e-06,
+      "loss": 1.0857,
+      "step": 5439
+    },
+    {
+      "epoch": 0.41936478569226027,
+      "grad_norm": 3.515936851501465,
+      "learning_rate": 6.524648470922854e-06,
+      "loss": 1.067,
+      "step": 5440
+    },
+    {
+      "epoch": 0.41944187480727724,
+      "grad_norm": 3.486137628555298,
+      "learning_rate": 6.523459433630681e-06,
+      "loss": 0.9586,
+      "step": 5441
+    },
+    {
+      "epoch": 0.41951896392229415,
+      "grad_norm": 3.7735965251922607,
+      "learning_rate": 6.522270301358704e-06,
+      "loss": 1.0085,
+      "step": 5442
+    },
+    {
+      "epoch": 0.4195960530373111,
+      "grad_norm": 3.485811948776245,
+      "learning_rate": 6.521081074181058e-06,
+      "loss": 0.9666,
+      "step": 5443
+    },
+    {
+      "epoch": 0.4196731421523281,
+      "grad_norm": 3.6323955059051514,
+      "learning_rate": 6.519891752171884e-06,
+      "loss": 1.0444,
+      "step": 5444
+    },
+    {
+      "epoch": 0.41975023126734506,
+      "grad_norm": 3.7233123779296875,
+      "learning_rate": 6.518702335405331e-06,
+      "loss": 0.9007,
+      "step": 5445
+    },
+    {
+      "epoch": 0.41982732038236203,
+      "grad_norm": 3.387659788131714,
+      "learning_rate": 6.517512823955554e-06,
+      "loss": 0.9503,
+      "step": 5446
+    },
+    {
+      "epoch": 0.41990440949737895,
+      "grad_norm": 3.6936397552490234,
+      "learning_rate": 6.516323217896712e-06,
+      "loss": 1.012,
+      "step": 5447
+    },
+    {
+      "epoch": 0.4199814986123959,
+      "grad_norm": 3.6477158069610596,
+      "learning_rate": 6.515133517302969e-06,
+      "loss": 1.0494,
+      "step": 5448
+    },
+    {
+      "epoch": 0.4200585877274129,
+      "grad_norm": 3.4203903675079346,
+      "learning_rate": 6.513943722248499e-06,
+      "loss": 0.8738,
+      "step": 5449
+    },
+    {
+      "epoch": 0.42013567684242986,
+      "grad_norm": 3.43131160736084,
+      "learning_rate": 6.512753832807479e-06,
+      "loss": 0.8789,
+      "step": 5450
+    },
+    {
+      "epoch": 0.42021276595744683,
+      "grad_norm": 3.5120580196380615,
+      "learning_rate": 6.5115638490540925e-06,
+      "loss": 0.9803,
+      "step": 5451
+    },
+    {
+      "epoch": 0.42028985507246375,
+      "grad_norm": 3.5579140186309814,
+      "learning_rate": 6.510373771062527e-06,
+      "loss": 0.893,
+      "step": 5452
+    },
+    {
+      "epoch": 0.4203669441874807,
+      "grad_norm": 4.425456523895264,
+      "learning_rate": 6.50918359890698e-06,
+      "loss": 0.9491,
+      "step": 5453
+    },
+    {
+      "epoch": 0.4204440333024977,
+      "grad_norm": 3.6898512840270996,
+      "learning_rate": 6.507993332661653e-06,
+      "loss": 0.9606,
+      "step": 5454
+    },
+    {
+      "epoch": 0.42052112241751466,
+      "grad_norm": 3.4359302520751953,
+      "learning_rate": 6.506802972400751e-06,
+      "loss": 0.9471,
+      "step": 5455
+    },
+    {
+      "epoch": 0.42059821153253163,
+      "grad_norm": 3.5586142539978027,
+      "learning_rate": 6.5056125181984885e-06,
+      "loss": 0.958,
+      "step": 5456
+    },
+    {
+      "epoch": 0.42067530064754854,
+      "grad_norm": 3.591996192932129,
+      "learning_rate": 6.5044219701290825e-06,
+      "loss": 1.0096,
+      "step": 5457
+    },
+    {
+      "epoch": 0.4207523897625655,
+      "grad_norm": 3.7730772495269775,
+      "learning_rate": 6.503231328266758e-06,
+      "loss": 0.9814,
+      "step": 5458
+    },
+    {
+      "epoch": 0.4208294788775825,
+      "grad_norm": 3.378232717514038,
+      "learning_rate": 6.5020405926857465e-06,
+      "loss": 0.915,
+      "step": 5459
+    },
+    {
+      "epoch": 0.42090656799259946,
+      "grad_norm": 3.841001033782959,
+      "learning_rate": 6.500849763460283e-06,
+      "loss": 1.0572,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4209836571076164,
+      "grad_norm": 3.5528838634490967,
+      "learning_rate": 6.4996588406646096e-06,
+      "loss": 1.0324,
+      "step": 5461
+    },
+    {
+      "epoch": 0.42106074622263334,
+      "grad_norm": 3.5531325340270996,
+      "learning_rate": 6.498467824372974e-06,
+      "loss": 0.9234,
+      "step": 5462
+    },
+    {
+      "epoch": 0.4211378353376503,
+      "grad_norm": 3.503661870956421,
+      "learning_rate": 6.497276714659631e-06,
+      "loss": 0.9227,
+      "step": 5463
+    },
+    {
+      "epoch": 0.4212149244526673,
+      "grad_norm": 3.5585978031158447,
+      "learning_rate": 6.49608551159884e-06,
+      "loss": 0.9453,
+      "step": 5464
+    },
+    {
+      "epoch": 0.42129201356768425,
+      "grad_norm": 3.890265941619873,
+      "learning_rate": 6.494894215264864e-06,
+      "loss": 0.9202,
+      "step": 5465
+    },
+    {
+      "epoch": 0.4213691026827012,
+      "grad_norm": 3.9904141426086426,
+      "learning_rate": 6.493702825731977e-06,
+      "loss": 0.9892,
+      "step": 5466
+    },
+    {
+      "epoch": 0.42144619179771814,
+      "grad_norm": 3.911604881286621,
+      "learning_rate": 6.492511343074456e-06,
+      "loss": 1.007,
+      "step": 5467
+    },
+    {
+      "epoch": 0.4215232809127351,
+      "grad_norm": 3.6176915168762207,
+      "learning_rate": 6.491319767366581e-06,
+      "loss": 1.0362,
+      "step": 5468
+    },
+    {
+      "epoch": 0.4216003700277521,
+      "grad_norm": 3.815973997116089,
+      "learning_rate": 6.490128098682642e-06,
+      "loss": 0.9862,
+      "step": 5469
+    },
+    {
+      "epoch": 0.42167745914276905,
+      "grad_norm": 4.3016037940979,
+      "learning_rate": 6.488936337096935e-06,
+      "loss": 0.9584,
+      "step": 5470
+    },
+    {
+      "epoch": 0.421754548257786,
+      "grad_norm": 3.49375319480896,
+      "learning_rate": 6.487744482683758e-06,
+      "loss": 0.9048,
+      "step": 5471
+    },
+    {
+      "epoch": 0.42183163737280294,
+      "grad_norm": 4.007232189178467,
+      "learning_rate": 6.486552535517419e-06,
+      "loss": 1.0449,
+      "step": 5472
+    },
+    {
+      "epoch": 0.4219087264878199,
+      "grad_norm": 3.4516263008117676,
+      "learning_rate": 6.485360495672226e-06,
+      "loss": 0.9911,
+      "step": 5473
+    },
+    {
+      "epoch": 0.4219858156028369,
+      "grad_norm": 3.6937084197998047,
+      "learning_rate": 6.4841683632225005e-06,
+      "loss": 0.9007,
+      "step": 5474
+    },
+    {
+      "epoch": 0.42206290471785385,
+      "grad_norm": 4.0262627601623535,
+      "learning_rate": 6.482976138242564e-06,
+      "loss": 1.1981,
+      "step": 5475
+    },
+    {
+      "epoch": 0.4221399938328708,
+      "grad_norm": 3.6597800254821777,
+      "learning_rate": 6.481783820806745e-06,
+      "loss": 0.898,
+      "step": 5476
+    },
+    {
+      "epoch": 0.42221708294788773,
+      "grad_norm": 3.7206947803497314,
+      "learning_rate": 6.4805914109893805e-06,
+      "loss": 0.9356,
+      "step": 5477
+    },
+    {
+      "epoch": 0.4222941720629047,
+      "grad_norm": 3.3729729652404785,
+      "learning_rate": 6.479398908864809e-06,
+      "loss": 0.9129,
+      "step": 5478
+    },
+    {
+      "epoch": 0.4223712611779217,
+      "grad_norm": 3.6832430362701416,
+      "learning_rate": 6.478206314507378e-06,
+      "loss": 0.9142,
+      "step": 5479
+    },
+    {
+      "epoch": 0.42244835029293865,
+      "grad_norm": 3.764679431915283,
+      "learning_rate": 6.47701362799144e-06,
+      "loss": 1.0789,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4225254394079556,
+      "grad_norm": 3.5273795127868652,
+      "learning_rate": 6.47582084939135e-06,
+      "loss": 0.9582,
+      "step": 5481
+    },
+    {
+      "epoch": 0.42260252852297253,
+      "grad_norm": 3.7401888370513916,
+      "learning_rate": 6.474627978781474e-06,
+      "loss": 1.0972,
+      "step": 5482
+    },
+    {
+      "epoch": 0.4226796176379895,
+      "grad_norm": 3.4443490505218506,
+      "learning_rate": 6.473435016236181e-06,
+      "loss": 0.9941,
+      "step": 5483
+    },
+    {
+      "epoch": 0.4227567067530065,
+      "grad_norm": 3.8706183433532715,
+      "learning_rate": 6.472241961829846e-06,
+      "loss": 0.9752,
+      "step": 5484
+    },
+    {
+      "epoch": 0.42283379586802344,
+      "grad_norm": 3.611464023590088,
+      "learning_rate": 6.47104881563685e-06,
+      "loss": 0.9907,
+      "step": 5485
+    },
+    {
+      "epoch": 0.4229108849830404,
+      "grad_norm": 3.5316667556762695,
+      "learning_rate": 6.469855577731579e-06,
+      "loss": 0.8496,
+      "step": 5486
+    },
+    {
+      "epoch": 0.42298797409805733,
+      "grad_norm": 3.4962635040283203,
+      "learning_rate": 6.468662248188424e-06,
+      "loss": 1.0317,
+      "step": 5487
+    },
+    {
+      "epoch": 0.4230650632130743,
+      "grad_norm": 4.126564025878906,
+      "learning_rate": 6.467468827081786e-06,
+      "loss": 0.9859,
+      "step": 5488
+    },
+    {
+      "epoch": 0.42314215232809127,
+      "grad_norm": 3.768683671951294,
+      "learning_rate": 6.466275314486066e-06,
+      "loss": 0.8975,
+      "step": 5489
+    },
+    {
+      "epoch": 0.42321924144310824,
+      "grad_norm": 3.997591733932495,
+      "learning_rate": 6.4650817104756735e-06,
+      "loss": 0.9813,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4232963305581252,
+      "grad_norm": 3.633054494857788,
+      "learning_rate": 6.463888015125026e-06,
+      "loss": 0.8793,
+      "step": 5491
+    },
+    {
+      "epoch": 0.4233734196731421,
+      "grad_norm": 3.5406949520111084,
+      "learning_rate": 6.4626942285085414e-06,
+      "loss": 0.8675,
+      "step": 5492
+    },
+    {
+      "epoch": 0.4234505087881591,
+      "grad_norm": 3.7187070846557617,
+      "learning_rate": 6.461500350700648e-06,
+      "loss": 1.0153,
+      "step": 5493
+    },
+    {
+      "epoch": 0.42352759790317607,
+      "grad_norm": 3.6261672973632812,
+      "learning_rate": 6.4603063817757746e-06,
+      "loss": 0.9925,
+      "step": 5494
+    },
+    {
+      "epoch": 0.42360468701819304,
+      "grad_norm": 3.54945969581604,
+      "learning_rate": 6.459112321808363e-06,
+      "loss": 1.0348,
+      "step": 5495
+    },
+    {
+      "epoch": 0.42368177613321,
+      "grad_norm": 3.686965227127075,
+      "learning_rate": 6.457918170872855e-06,
+      "loss": 0.9913,
+      "step": 5496
+    },
+    {
+      "epoch": 0.4237588652482269,
+      "grad_norm": 3.7915258407592773,
+      "learning_rate": 6.4567239290437e-06,
+      "loss": 1.025,
+      "step": 5497
+    },
+    {
+      "epoch": 0.4238359543632439,
+      "grad_norm": 3.7425386905670166,
+      "learning_rate": 6.455529596395353e-06,
+      "loss": 0.9447,
+      "step": 5498
+    },
+    {
+      "epoch": 0.42391304347826086,
+      "grad_norm": 3.5631515979766846,
+      "learning_rate": 6.454335173002273e-06,
+      "loss": 0.9225,
+      "step": 5499
+    },
+    {
+      "epoch": 0.42399013259327784,
+      "grad_norm": 3.4709506034851074,
+      "learning_rate": 6.4531406589389275e-06,
+      "loss": 0.9846,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4240672217082948,
+      "grad_norm": 3.696099281311035,
+      "learning_rate": 6.451946054279788e-06,
+      "loss": 0.9241,
+      "step": 5501
+    },
+    {
+      "epoch": 0.4241443108233117,
+      "grad_norm": 3.3619208335876465,
+      "learning_rate": 6.450751359099332e-06,
+      "loss": 0.9098,
+      "step": 5502
+    },
+    {
+      "epoch": 0.4242213999383287,
+      "grad_norm": 3.8258392810821533,
+      "learning_rate": 6.449556573472042e-06,
+      "loss": 1.1477,
+      "step": 5503
+    },
+    {
+      "epoch": 0.42429848905334566,
+      "grad_norm": 3.742774724960327,
+      "learning_rate": 6.448361697472408e-06,
+      "loss": 1.0046,
+      "step": 5504
+    },
+    {
+      "epoch": 0.42437557816836263,
+      "grad_norm": 3.6903765201568604,
+      "learning_rate": 6.447166731174923e-06,
+      "loss": 0.9383,
+      "step": 5505
+    },
+    {
+      "epoch": 0.4244526672833796,
+      "grad_norm": 4.07509183883667,
+      "learning_rate": 6.445971674654087e-06,
+      "loss": 1.0482,
+      "step": 5506
+    },
+    {
+      "epoch": 0.4245297563983966,
+      "grad_norm": 3.969420909881592,
+      "learning_rate": 6.444776527984406e-06,
+      "loss": 0.9808,
+      "step": 5507
+    },
+    {
+      "epoch": 0.4246068455134135,
+      "grad_norm": 3.540527820587158,
+      "learning_rate": 6.443581291240392e-06,
+      "loss": 0.9194,
+      "step": 5508
+    },
+    {
+      "epoch": 0.42468393462843046,
+      "grad_norm": 3.800278663635254,
+      "learning_rate": 6.44238596449656e-06,
+      "loss": 0.9855,
+      "step": 5509
+    },
+    {
+      "epoch": 0.42476102374344743,
+      "grad_norm": 3.841402769088745,
+      "learning_rate": 6.441190547827434e-06,
+      "loss": 1.0468,
+      "step": 5510
+    },
+    {
+      "epoch": 0.4248381128584644,
+      "grad_norm": 3.673065423965454,
+      "learning_rate": 6.439995041307541e-06,
+      "loss": 1.0121,
+      "step": 5511
+    },
+    {
+      "epoch": 0.42491520197348137,
+      "grad_norm": 3.586517810821533,
+      "learning_rate": 6.438799445011415e-06,
+      "loss": 1.0284,
+      "step": 5512
+    },
+    {
+      "epoch": 0.4249922910884983,
+      "grad_norm": 4.116529941558838,
+      "learning_rate": 6.4376037590135955e-06,
+      "loss": 1.06,
+      "step": 5513
+    },
+    {
+      "epoch": 0.42506938020351526,
+      "grad_norm": 3.565775156021118,
+      "learning_rate": 6.436407983388627e-06,
+      "loss": 0.9655,
+      "step": 5514
+    },
+    {
+      "epoch": 0.4251464693185322,
+      "grad_norm": 3.403215169906616,
+      "learning_rate": 6.435212118211062e-06,
+      "loss": 0.9076,
+      "step": 5515
+    },
+    {
+      "epoch": 0.4252235584335492,
+      "grad_norm": 4.086001396179199,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 1.0058,
+      "step": 5516
+    },
+    {
+      "epoch": 0.42530064754856617,
+      "grad_norm": 3.650869131088257,
+      "learning_rate": 6.432820119496363e-06,
+      "loss": 0.941,
+      "step": 5517
+    },
+    {
+      "epoch": 0.4253777366635831,
+      "grad_norm": 3.593628168106079,
+      "learning_rate": 6.431623986108359e-06,
+      "loss": 1.0367,
+      "step": 5518
+    },
+    {
+      "epoch": 0.42545482577860005,
+      "grad_norm": 3.7632596492767334,
+      "learning_rate": 6.430427763466014e-06,
+      "loss": 0.9751,
+      "step": 5519
+    },
+    {
+      "epoch": 0.425531914893617,
+      "grad_norm": 3.490562677383423,
+      "learning_rate": 6.429231451643907e-06,
+      "loss": 0.9234,
+      "step": 5520
+    },
+    {
+      "epoch": 0.425609004008634,
+      "grad_norm": 3.740706443786621,
+      "learning_rate": 6.428035050716621e-06,
+      "loss": 1.023,
+      "step": 5521
+    },
+    {
+      "epoch": 0.42568609312365097,
+      "grad_norm": 3.898951292037964,
+      "learning_rate": 6.426838560758746e-06,
+      "loss": 0.9387,
+      "step": 5522
+    },
+    {
+      "epoch": 0.4257631822386679,
+      "grad_norm": 3.8880972862243652,
+      "learning_rate": 6.425641981844876e-06,
+      "loss": 0.8878,
+      "step": 5523
+    },
+    {
+      "epoch": 0.42584027135368485,
+      "grad_norm": 3.443209648132324,
+      "learning_rate": 6.424445314049611e-06,
+      "loss": 0.8749,
+      "step": 5524
+    },
+    {
+      "epoch": 0.4259173604687018,
+      "grad_norm": 4.383835315704346,
+      "learning_rate": 6.42324855744756e-06,
+      "loss": 1.0518,
+      "step": 5525
+    },
+    {
+      "epoch": 0.4259944495837188,
+      "grad_norm": 3.5138866901397705,
+      "learning_rate": 6.422051712113332e-06,
+      "loss": 1.003,
+      "step": 5526
+    },
+    {
+      "epoch": 0.42607153869873576,
+      "grad_norm": 3.474919080734253,
+      "learning_rate": 6.420854778121543e-06,
+      "loss": 1.0002,
+      "step": 5527
+    },
+    {
+      "epoch": 0.4261486278137527,
+      "grad_norm": 3.776796817779541,
+      "learning_rate": 6.41965775554682e-06,
+      "loss": 1.0651,
+      "step": 5528
+    },
+    {
+      "epoch": 0.42622571692876965,
+      "grad_norm": 3.249838352203369,
+      "learning_rate": 6.418460644463787e-06,
+      "loss": 0.9011,
+      "step": 5529
+    },
+    {
+      "epoch": 0.4263028060437866,
+      "grad_norm": 3.8011159896850586,
+      "learning_rate": 6.41726344494708e-06,
+      "loss": 0.8613,
+      "step": 5530
+    },
+    {
+      "epoch": 0.4263798951588036,
+      "grad_norm": 4.086413383483887,
+      "learning_rate": 6.416066157071338e-06,
+      "loss": 1.0941,
+      "step": 5531
+    },
+    {
+      "epoch": 0.42645698427382056,
+      "grad_norm": 3.662161111831665,
+      "learning_rate": 6.414868780911203e-06,
+      "loss": 1.0212,
+      "step": 5532
+    },
+    {
+      "epoch": 0.4265340733888375,
+      "grad_norm": 3.3614611625671387,
+      "learning_rate": 6.41367131654133e-06,
+      "loss": 0.925,
+      "step": 5533
+    },
+    {
+      "epoch": 0.42661116250385445,
+      "grad_norm": 3.57147216796875,
+      "learning_rate": 6.41247376403637e-06,
+      "loss": 0.9396,
+      "step": 5534
+    },
+    {
+      "epoch": 0.4266882516188714,
+      "grad_norm": 3.516568422317505,
+      "learning_rate": 6.4112761234709866e-06,
+      "loss": 0.8563,
+      "step": 5535
+    },
+    {
+      "epoch": 0.4267653407338884,
+      "grad_norm": 3.589931011199951,
+      "learning_rate": 6.4100783949198465e-06,
+      "loss": 0.9974,
+      "step": 5536
+    },
+    {
+      "epoch": 0.42684242984890536,
+      "grad_norm": 3.5021722316741943,
+      "learning_rate": 6.408880578457622e-06,
+      "loss": 1.0911,
+      "step": 5537
+    },
+    {
+      "epoch": 0.4269195189639223,
+      "grad_norm": 3.5328221321105957,
+      "learning_rate": 6.407682674158988e-06,
+      "loss": 0.8483,
+      "step": 5538
+    },
+    {
+      "epoch": 0.42699660807893924,
+      "grad_norm": 4.008224010467529,
+      "learning_rate": 6.406484682098632e-06,
+      "loss": 1.0525,
+      "step": 5539
+    },
+    {
+      "epoch": 0.4270736971939562,
+      "grad_norm": 3.6036767959594727,
+      "learning_rate": 6.40528660235124e-06,
+      "loss": 0.9958,
+      "step": 5540
+    },
+    {
+      "epoch": 0.4271507863089732,
+      "grad_norm": 4.159304618835449,
+      "learning_rate": 6.404088434991504e-06,
+      "loss": 0.9268,
+      "step": 5541
+    },
+    {
+      "epoch": 0.42722787542399016,
+      "grad_norm": 4.040691375732422,
+      "learning_rate": 6.402890180094129e-06,
+      "loss": 1.1215,
+      "step": 5542
+    },
+    {
+      "epoch": 0.42730496453900707,
+      "grad_norm": 3.66817307472229,
+      "learning_rate": 6.401691837733815e-06,
+      "loss": 0.9277,
+      "step": 5543
+    },
+    {
+      "epoch": 0.42738205365402404,
+      "grad_norm": 4.073860168457031,
+      "learning_rate": 6.4004934079852775e-06,
+      "loss": 0.986,
+      "step": 5544
+    },
+    {
+      "epoch": 0.427459142769041,
+      "grad_norm": 3.9062860012054443,
+      "learning_rate": 6.399294890923227e-06,
+      "loss": 1.0575,
+      "step": 5545
+    },
+    {
+      "epoch": 0.427536231884058,
+      "grad_norm": 3.5801217555999756,
+      "learning_rate": 6.398096286622388e-06,
+      "loss": 0.9702,
+      "step": 5546
+    },
+    {
+      "epoch": 0.42761332099907495,
+      "grad_norm": 3.972623825073242,
+      "learning_rate": 6.396897595157487e-06,
+      "loss": 0.9699,
+      "step": 5547
+    },
+    {
+      "epoch": 0.42769041011409187,
+      "grad_norm": 3.516526222229004,
+      "learning_rate": 6.395698816603253e-06,
+      "loss": 0.9262,
+      "step": 5548
+    },
+    {
+      "epoch": 0.42776749922910884,
+      "grad_norm": 3.5756266117095947,
+      "learning_rate": 6.39449995103443e-06,
+      "loss": 0.9331,
+      "step": 5549
+    },
+    {
+      "epoch": 0.4278445883441258,
+      "grad_norm": 3.659475326538086,
+      "learning_rate": 6.393300998525754e-06,
+      "loss": 0.8752,
+      "step": 5550
+    },
+    {
+      "epoch": 0.4279216774591428,
+      "grad_norm": 4.002374649047852,
+      "learning_rate": 6.392101959151978e-06,
+      "loss": 1.1071,
+      "step": 5551
+    },
+    {
+      "epoch": 0.42799876657415975,
+      "grad_norm": 3.4947450160980225,
+      "learning_rate": 6.390902832987857e-06,
+      "loss": 0.9502,
+      "step": 5552
+    },
+    {
+      "epoch": 0.42807585568917667,
+      "grad_norm": 3.6659510135650635,
+      "learning_rate": 6.389703620108145e-06,
+      "loss": 0.9983,
+      "step": 5553
+    },
+    {
+      "epoch": 0.42815294480419364,
+      "grad_norm": 3.892176628112793,
+      "learning_rate": 6.388504320587611e-06,
+      "loss": 0.9481,
+      "step": 5554
+    },
+    {
+      "epoch": 0.4282300339192106,
+      "grad_norm": 3.580537796020508,
+      "learning_rate": 6.387304934501024e-06,
+      "loss": 0.9526,
+      "step": 5555
+    },
+    {
+      "epoch": 0.4283071230342276,
+      "grad_norm": 3.420477867126465,
+      "learning_rate": 6.386105461923159e-06,
+      "loss": 0.9092,
+      "step": 5556
+    },
+    {
+      "epoch": 0.42838421214924455,
+      "grad_norm": 3.7150497436523438,
+      "learning_rate": 6.384905902928797e-06,
+      "loss": 1.0413,
+      "step": 5557
+    },
+    {
+      "epoch": 0.42846130126426146,
+      "grad_norm": 3.4000418186187744,
+      "learning_rate": 6.383706257592725e-06,
+      "loss": 0.9897,
+      "step": 5558
+    },
+    {
+      "epoch": 0.42853839037927843,
+      "grad_norm": 3.336329698562622,
+      "learning_rate": 6.382506525989734e-06,
+      "loss": 1.0003,
+      "step": 5559
+    },
+    {
+      "epoch": 0.4286154794942954,
+      "grad_norm": 3.510364055633545,
+      "learning_rate": 6.381306708194622e-06,
+      "loss": 0.8863,
+      "step": 5560
+    },
+    {
+      "epoch": 0.4286925686093124,
+      "grad_norm": 3.6670730113983154,
+      "learning_rate": 6.38010680428219e-06,
+      "loss": 1.0318,
+      "step": 5561
+    },
+    {
+      "epoch": 0.42876965772432934,
+      "grad_norm": 3.3945372104644775,
+      "learning_rate": 6.378906814327246e-06,
+      "loss": 1.0504,
+      "step": 5562
+    },
+    {
+      "epoch": 0.42884674683934626,
+      "grad_norm": 4.040469169616699,
+      "learning_rate": 6.377706738404604e-06,
+      "loss": 1.1041,
+      "step": 5563
+    },
+    {
+      "epoch": 0.42892383595436323,
+      "grad_norm": 4.579511642456055,
+      "learning_rate": 6.376506576589082e-06,
+      "loss": 1.0169,
+      "step": 5564
+    },
+    {
+      "epoch": 0.4290009250693802,
+      "grad_norm": 3.5444085597991943,
+      "learning_rate": 6.375306328955506e-06,
+      "loss": 1.0349,
+      "step": 5565
+    },
+    {
+      "epoch": 0.42907801418439717,
+      "grad_norm": 3.6533939838409424,
+      "learning_rate": 6.374105995578701e-06,
+      "loss": 0.9106,
+      "step": 5566
+    },
+    {
+      "epoch": 0.42915510329941414,
+      "grad_norm": 3.527340888977051,
+      "learning_rate": 6.372905576533505e-06,
+      "loss": 0.9116,
+      "step": 5567
+    },
+    {
+      "epoch": 0.42923219241443106,
+      "grad_norm": 3.571380853652954,
+      "learning_rate": 6.371705071894756e-06,
+      "loss": 1.0186,
+      "step": 5568
+    },
+    {
+      "epoch": 0.42930928152944803,
+      "grad_norm": 3.8140509128570557,
+      "learning_rate": 6.3705044817373006e-06,
+      "loss": 0.9015,
+      "step": 5569
+    },
+    {
+      "epoch": 0.429386370644465,
+      "grad_norm": 3.546802520751953,
+      "learning_rate": 6.369303806135989e-06,
+      "loss": 1.0005,
+      "step": 5570
+    },
+    {
+      "epoch": 0.42946345975948197,
+      "grad_norm": 3.889112949371338,
+      "learning_rate": 6.368103045165677e-06,
+      "loss": 1.0829,
+      "step": 5571
+    },
+    {
+      "epoch": 0.42954054887449894,
+      "grad_norm": 3.435272216796875,
+      "learning_rate": 6.366902198901225e-06,
+      "loss": 1.0089,
+      "step": 5572
+    },
+    {
+      "epoch": 0.42961763798951585,
+      "grad_norm": 3.3787803649902344,
+      "learning_rate": 6.365701267417501e-06,
+      "loss": 0.934,
+      "step": 5573
+    },
+    {
+      "epoch": 0.4296947271045328,
+      "grad_norm": 3.8699758052825928,
+      "learning_rate": 6.364500250789375e-06,
+      "loss": 1.1315,
+      "step": 5574
+    },
+    {
+      "epoch": 0.4297718162195498,
+      "grad_norm": 3.6572110652923584,
+      "learning_rate": 6.363299149091726e-06,
+      "loss": 1.0157,
+      "step": 5575
+    },
+    {
+      "epoch": 0.42984890533456677,
+      "grad_norm": 3.4599862098693848,
+      "learning_rate": 6.362097962399436e-06,
+      "loss": 0.9634,
+      "step": 5576
+    },
+    {
+      "epoch": 0.42992599444958374,
+      "grad_norm": 3.603281259536743,
+      "learning_rate": 6.3608966907873925e-06,
+      "loss": 0.9877,
+      "step": 5577
+    },
+    {
+      "epoch": 0.43000308356460065,
+      "grad_norm": 3.7701170444488525,
+      "learning_rate": 6.359695334330488e-06,
+      "loss": 1.0674,
+      "step": 5578
+    },
+    {
+      "epoch": 0.4300801726796176,
+      "grad_norm": 3.7312378883361816,
+      "learning_rate": 6.358493893103623e-06,
+      "loss": 0.9849,
+      "step": 5579
+    },
+    {
+      "epoch": 0.4301572617946346,
+      "grad_norm": 3.3905229568481445,
+      "learning_rate": 6.357292367181698e-06,
+      "loss": 0.8549,
+      "step": 5580
+    },
+    {
+      "epoch": 0.43023435090965156,
+      "grad_norm": 3.478235960006714,
+      "learning_rate": 6.356090756639623e-06,
+      "loss": 0.8909,
+      "step": 5581
+    },
+    {
+      "epoch": 0.43031144002466853,
+      "grad_norm": 3.4896721839904785,
+      "learning_rate": 6.354889061552314e-06,
+      "loss": 0.9935,
+      "step": 5582
+    },
+    {
+      "epoch": 0.43038852913968545,
+      "grad_norm": 3.669250965118408,
+      "learning_rate": 6.353687281994688e-06,
+      "loss": 1.0689,
+      "step": 5583
+    },
+    {
+      "epoch": 0.4304656182547024,
+      "grad_norm": 3.8079261779785156,
+      "learning_rate": 6.352485418041673e-06,
+      "loss": 0.9981,
+      "step": 5584
+    },
+    {
+      "epoch": 0.4305427073697194,
+      "grad_norm": 3.5822927951812744,
+      "learning_rate": 6.351283469768195e-06,
+      "loss": 1.0457,
+      "step": 5585
+    },
+    {
+      "epoch": 0.43061979648473636,
+      "grad_norm": 3.62357759475708,
+      "learning_rate": 6.350081437249191e-06,
+      "loss": 0.9444,
+      "step": 5586
+    },
+    {
+      "epoch": 0.43069688559975333,
+      "grad_norm": 3.627352476119995,
+      "learning_rate": 6.348879320559602e-06,
+      "loss": 0.9715,
+      "step": 5587
+    },
+    {
+      "epoch": 0.43077397471477025,
+      "grad_norm": 3.5861945152282715,
+      "learning_rate": 6.3476771197743735e-06,
+      "loss": 0.9506,
+      "step": 5588
+    },
+    {
+      "epoch": 0.4308510638297872,
+      "grad_norm": 3.5179619789123535,
+      "learning_rate": 6.346474834968458e-06,
+      "loss": 1.0651,
+      "step": 5589
+    },
+    {
+      "epoch": 0.4309281529448042,
+      "grad_norm": 4.048654556274414,
+      "learning_rate": 6.345272466216807e-06,
+      "loss": 1.0671,
+      "step": 5590
+    },
+    {
+      "epoch": 0.43100524205982116,
+      "grad_norm": 3.6038835048675537,
+      "learning_rate": 6.344070013594388e-06,
+      "loss": 1.0723,
+      "step": 5591
+    },
+    {
+      "epoch": 0.43108233117483813,
+      "grad_norm": 3.3801581859588623,
+      "learning_rate": 6.342867477176164e-06,
+      "loss": 0.9918,
+      "step": 5592
+    },
+    {
+      "epoch": 0.4311594202898551,
+      "grad_norm": 4.134720325469971,
+      "learning_rate": 6.3416648570371065e-06,
+      "loss": 1.0093,
+      "step": 5593
+    },
+    {
+      "epoch": 0.431236509404872,
+      "grad_norm": 3.704859495162964,
+      "learning_rate": 6.340462153252195e-06,
+      "loss": 0.953,
+      "step": 5594
+    },
+    {
+      "epoch": 0.431313598519889,
+      "grad_norm": 3.9466912746429443,
+      "learning_rate": 6.339259365896411e-06,
+      "loss": 1.0878,
+      "step": 5595
+    },
+    {
+      "epoch": 0.43139068763490596,
+      "grad_norm": 3.7357873916625977,
+      "learning_rate": 6.33805649504474e-06,
+      "loss": 0.9632,
+      "step": 5596
+    },
+    {
+      "epoch": 0.4314677767499229,
+      "grad_norm": 3.6938416957855225,
+      "learning_rate": 6.336853540772178e-06,
+      "loss": 0.9337,
+      "step": 5597
+    },
+    {
+      "epoch": 0.4315448658649399,
+      "grad_norm": 3.942068099975586,
+      "learning_rate": 6.335650503153721e-06,
+      "loss": 0.974,
+      "step": 5598
+    },
+    {
+      "epoch": 0.4316219549799568,
+      "grad_norm": 4.033871650695801,
+      "learning_rate": 6.334447382264372e-06,
+      "loss": 0.9456,
+      "step": 5599
+    },
+    {
+      "epoch": 0.4316990440949738,
+      "grad_norm": 3.7592177391052246,
+      "learning_rate": 6.333244178179141e-06,
+      "loss": 1.0018,
+      "step": 5600
+    },
+    {
+      "epoch": 0.43177613320999075,
+      "grad_norm": 3.3277747631073,
+      "learning_rate": 6.3320408909730405e-06,
+      "loss": 0.9049,
+      "step": 5601
+    },
+    {
+      "epoch": 0.4318532223250077,
+      "grad_norm": 3.5727672576904297,
+      "learning_rate": 6.330837520721088e-06,
+      "loss": 0.9017,
+      "step": 5602
+    },
+    {
+      "epoch": 0.4319303114400247,
+      "grad_norm": 3.610518455505371,
+      "learning_rate": 6.329634067498311e-06,
+      "loss": 1.0287,
+      "step": 5603
+    },
+    {
+      "epoch": 0.4320074005550416,
+      "grad_norm": 3.8098137378692627,
+      "learning_rate": 6.3284305313797335e-06,
+      "loss": 0.9755,
+      "step": 5604
+    },
+    {
+      "epoch": 0.4320844896700586,
+      "grad_norm": 3.6883254051208496,
+      "learning_rate": 6.327226912440394e-06,
+      "loss": 1.021,
+      "step": 5605
+    },
+    {
+      "epoch": 0.43216157878507555,
+      "grad_norm": 3.6646480560302734,
+      "learning_rate": 6.32602321075533e-06,
+      "loss": 0.9877,
+      "step": 5606
+    },
+    {
+      "epoch": 0.4322386679000925,
+      "grad_norm": 3.433976173400879,
+      "learning_rate": 6.324819426399587e-06,
+      "loss": 0.9482,
+      "step": 5607
+    },
+    {
+      "epoch": 0.4323157570151095,
+      "grad_norm": 3.4387431144714355,
+      "learning_rate": 6.323615559448213e-06,
+      "loss": 0.8817,
+      "step": 5608
+    },
+    {
+      "epoch": 0.4323928461301264,
+      "grad_norm": 3.4525887966156006,
+      "learning_rate": 6.322411609976265e-06,
+      "loss": 0.8509,
+      "step": 5609
+    },
+    {
+      "epoch": 0.4324699352451434,
+      "grad_norm": 3.7124297618865967,
+      "learning_rate": 6.321207578058803e-06,
+      "loss": 0.962,
+      "step": 5610
+    },
+    {
+      "epoch": 0.43254702436016035,
+      "grad_norm": 3.817814350128174,
+      "learning_rate": 6.32000346377089e-06,
+      "loss": 1.051,
+      "step": 5611
+    },
+    {
+      "epoch": 0.4326241134751773,
+      "grad_norm": 3.616466760635376,
+      "learning_rate": 6.318799267187596e-06,
+      "loss": 0.9733,
+      "step": 5612
+    },
+    {
+      "epoch": 0.4327012025901943,
+      "grad_norm": 3.7230231761932373,
+      "learning_rate": 6.317594988384e-06,
+      "loss": 1.0263,
+      "step": 5613
+    },
+    {
+      "epoch": 0.4327782917052112,
+      "grad_norm": 3.632129430770874,
+      "learning_rate": 6.31639062743518e-06,
+      "loss": 0.9545,
+      "step": 5614
+    },
+    {
+      "epoch": 0.4328553808202282,
+      "grad_norm": 4.578291416168213,
+      "learning_rate": 6.315186184416222e-06,
+      "loss": 1.0831,
+      "step": 5615
+    },
+    {
+      "epoch": 0.43293246993524515,
+      "grad_norm": 3.434826374053955,
+      "learning_rate": 6.313981659402218e-06,
+      "loss": 0.9454,
+      "step": 5616
+    },
+    {
+      "epoch": 0.4330095590502621,
+      "grad_norm": 3.2984015941619873,
+      "learning_rate": 6.312777052468262e-06,
+      "loss": 0.9104,
+      "step": 5617
+    },
+    {
+      "epoch": 0.4330866481652791,
+      "grad_norm": 3.349336862564087,
+      "learning_rate": 6.3115723636894565e-06,
+      "loss": 1.0299,
+      "step": 5618
+    },
+    {
+      "epoch": 0.433163737280296,
+      "grad_norm": 3.6265594959259033,
+      "learning_rate": 6.310367593140906e-06,
+      "loss": 0.9828,
+      "step": 5619
+    },
+    {
+      "epoch": 0.433240826395313,
+      "grad_norm": 3.9649040699005127,
+      "learning_rate": 6.3091627408977215e-06,
+      "loss": 0.9659,
+      "step": 5620
+    },
+    {
+      "epoch": 0.43331791551032994,
+      "grad_norm": 3.6127185821533203,
+      "learning_rate": 6.3079578070350235e-06,
+      "loss": 0.8698,
+      "step": 5621
+    },
+    {
+      "epoch": 0.4333950046253469,
+      "grad_norm": 3.442434310913086,
+      "learning_rate": 6.306752791627928e-06,
+      "loss": 0.8926,
+      "step": 5622
+    },
+    {
+      "epoch": 0.4334720937403639,
+      "grad_norm": 3.6911134719848633,
+      "learning_rate": 6.3055476947515635e-06,
+      "loss": 1.0027,
+      "step": 5623
+    },
+    {
+      "epoch": 0.4335491828553808,
+      "grad_norm": 3.4410293102264404,
+      "learning_rate": 6.3043425164810635e-06,
+      "loss": 0.8793,
+      "step": 5624
+    },
+    {
+      "epoch": 0.43362627197039777,
+      "grad_norm": 3.6440846920013428,
+      "learning_rate": 6.303137256891563e-06,
+      "loss": 1.027,
+      "step": 5625
+    },
+    {
+      "epoch": 0.43370336108541474,
+      "grad_norm": 3.805568218231201,
+      "learning_rate": 6.301931916058201e-06,
+      "loss": 0.9033,
+      "step": 5626
+    },
+    {
+      "epoch": 0.4337804502004317,
+      "grad_norm": 3.4418349266052246,
+      "learning_rate": 6.300726494056131e-06,
+      "loss": 0.8662,
+      "step": 5627
+    },
+    {
+      "epoch": 0.4338575393154487,
+      "grad_norm": 3.30560302734375,
+      "learning_rate": 6.299520990960497e-06,
+      "loss": 0.959,
+      "step": 5628
+    },
+    {
+      "epoch": 0.4339346284304656,
+      "grad_norm": 3.5205271244049072,
+      "learning_rate": 6.298315406846462e-06,
+      "loss": 0.9701,
+      "step": 5629
+    },
+    {
+      "epoch": 0.43401171754548257,
+      "grad_norm": 3.5560543537139893,
+      "learning_rate": 6.297109741789184e-06,
+      "loss": 0.8881,
+      "step": 5630
+    },
+    {
+      "epoch": 0.43408880666049954,
+      "grad_norm": 3.4408164024353027,
+      "learning_rate": 6.2959039958638325e-06,
+      "loss": 0.9109,
+      "step": 5631
+    },
+    {
+      "epoch": 0.4341658957755165,
+      "grad_norm": 3.3632290363311768,
+      "learning_rate": 6.2946981691455775e-06,
+      "loss": 0.8839,
+      "step": 5632
+    },
+    {
+      "epoch": 0.4342429848905335,
+      "grad_norm": 3.5588736534118652,
+      "learning_rate": 6.293492261709597e-06,
+      "loss": 0.9453,
+      "step": 5633
+    },
+    {
+      "epoch": 0.4343200740055504,
+      "grad_norm": 3.7234859466552734,
+      "learning_rate": 6.292286273631074e-06,
+      "loss": 1.0011,
+      "step": 5634
+    },
+    {
+      "epoch": 0.43439716312056736,
+      "grad_norm": 3.504631280899048,
+      "learning_rate": 6.291080204985195e-06,
+      "loss": 0.8339,
+      "step": 5635
+    },
+    {
+      "epoch": 0.43447425223558434,
+      "grad_norm": 3.596510171890259,
+      "learning_rate": 6.289874055847152e-06,
+      "loss": 0.8935,
+      "step": 5636
+    },
+    {
+      "epoch": 0.4345513413506013,
+      "grad_norm": 3.4602015018463135,
+      "learning_rate": 6.288667826292142e-06,
+      "loss": 0.9409,
+      "step": 5637
+    },
+    {
+      "epoch": 0.4346284304656183,
+      "grad_norm": 3.5766189098358154,
+      "learning_rate": 6.2874615163953654e-06,
+      "loss": 0.9501,
+      "step": 5638
+    },
+    {
+      "epoch": 0.4347055195806352,
+      "grad_norm": 3.890477180480957,
+      "learning_rate": 6.286255126232032e-06,
+      "loss": 0.8271,
+      "step": 5639
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 3.6183977127075195,
+      "learning_rate": 6.285048655877355e-06,
+      "loss": 0.9268,
+      "step": 5640
+    },
+    {
+      "epoch": 0.43485969781066913,
+      "grad_norm": 3.9595730304718018,
+      "learning_rate": 6.283842105406548e-06,
+      "loss": 1.0355,
+      "step": 5641
+    },
+    {
+      "epoch": 0.4349367869256861,
+      "grad_norm": 3.801222801208496,
+      "learning_rate": 6.282635474894836e-06,
+      "loss": 0.9763,
+      "step": 5642
+    },
+    {
+      "epoch": 0.4350138760407031,
+      "grad_norm": 3.6348776817321777,
+      "learning_rate": 6.281428764417444e-06,
+      "loss": 1.0238,
+      "step": 5643
+    },
+    {
+      "epoch": 0.43509096515572,
+      "grad_norm": 3.195066452026367,
+      "learning_rate": 6.280221974049606e-06,
+      "loss": 0.8491,
+      "step": 5644
+    },
+    {
+      "epoch": 0.43516805427073696,
+      "grad_norm": 3.7148594856262207,
+      "learning_rate": 6.279015103866559e-06,
+      "loss": 0.871,
+      "step": 5645
+    },
+    {
+      "epoch": 0.43524514338575393,
+      "grad_norm": 3.536971092224121,
+      "learning_rate": 6.2778081539435436e-06,
+      "loss": 0.9518,
+      "step": 5646
+    },
+    {
+      "epoch": 0.4353222325007709,
+      "grad_norm": 3.973155975341797,
+      "learning_rate": 6.276601124355807e-06,
+      "loss": 1.012,
+      "step": 5647
+    },
+    {
+      "epoch": 0.43539932161578787,
+      "grad_norm": 3.6045761108398438,
+      "learning_rate": 6.275394015178603e-06,
+      "loss": 0.983,
+      "step": 5648
+    },
+    {
+      "epoch": 0.4354764107308048,
+      "grad_norm": 3.582974433898926,
+      "learning_rate": 6.274186826487187e-06,
+      "loss": 0.9876,
+      "step": 5649
+    },
+    {
+      "epoch": 0.43555349984582176,
+      "grad_norm": 3.7312257289886475,
+      "learning_rate": 6.272979558356821e-06,
+      "loss": 1.029,
+      "step": 5650
+    },
+    {
+      "epoch": 0.4356305889608387,
+      "grad_norm": 3.7958641052246094,
+      "learning_rate": 6.271772210862773e-06,
+      "loss": 0.9989,
+      "step": 5651
+    },
+    {
+      "epoch": 0.4357076780758557,
+      "grad_norm": 3.4761128425598145,
+      "learning_rate": 6.270564784080314e-06,
+      "loss": 1.0541,
+      "step": 5652
+    },
+    {
+      "epoch": 0.43578476719087267,
+      "grad_norm": 3.6425018310546875,
+      "learning_rate": 6.269357278084723e-06,
+      "loss": 1.0119,
+      "step": 5653
+    },
+    {
+      "epoch": 0.4358618563058896,
+      "grad_norm": 3.6321396827697754,
+      "learning_rate": 6.268149692951278e-06,
+      "loss": 1.0429,
+      "step": 5654
+    },
+    {
+      "epoch": 0.43593894542090655,
+      "grad_norm": 3.9710259437561035,
+      "learning_rate": 6.266942028755269e-06,
+      "loss": 0.9756,
+      "step": 5655
+    },
+    {
+      "epoch": 0.4360160345359235,
+      "grad_norm": 3.927583694458008,
+      "learning_rate": 6.265734285571985e-06,
+      "loss": 1.0705,
+      "step": 5656
+    },
+    {
+      "epoch": 0.4360931236509405,
+      "grad_norm": 3.5504817962646484,
+      "learning_rate": 6.2645264634767236e-06,
+      "loss": 1.0056,
+      "step": 5657
+    },
+    {
+      "epoch": 0.43617021276595747,
+      "grad_norm": 3.7636258602142334,
+      "learning_rate": 6.263318562544787e-06,
+      "loss": 1.011,
+      "step": 5658
+    },
+    {
+      "epoch": 0.4362473018809744,
+      "grad_norm": 4.358571529388428,
+      "learning_rate": 6.262110582851481e-06,
+      "loss": 0.9815,
+      "step": 5659
+    },
+    {
+      "epoch": 0.43632439099599135,
+      "grad_norm": 3.5723724365234375,
+      "learning_rate": 6.260902524472116e-06,
+      "loss": 0.8575,
+      "step": 5660
+    },
+    {
+      "epoch": 0.4364014801110083,
+      "grad_norm": 3.437037229537964,
+      "learning_rate": 6.2596943874820105e-06,
+      "loss": 0.9603,
+      "step": 5661
+    },
+    {
+      "epoch": 0.4364785692260253,
+      "grad_norm": 3.732524871826172,
+      "learning_rate": 6.2584861719564825e-06,
+      "loss": 0.8858,
+      "step": 5662
+    },
+    {
+      "epoch": 0.43655565834104226,
+      "grad_norm": 3.429790496826172,
+      "learning_rate": 6.25727787797086e-06,
+      "loss": 1.0058,
+      "step": 5663
+    },
+    {
+      "epoch": 0.4366327474560592,
+      "grad_norm": 3.6642956733703613,
+      "learning_rate": 6.256069505600474e-06,
+      "loss": 0.9124,
+      "step": 5664
+    },
+    {
+      "epoch": 0.43670983657107615,
+      "grad_norm": 3.6058828830718994,
+      "learning_rate": 6.254861054920659e-06,
+      "loss": 0.9339,
+      "step": 5665
+    },
+    {
+      "epoch": 0.4367869256860931,
+      "grad_norm": 3.4865617752075195,
+      "learning_rate": 6.2536525260067575e-06,
+      "loss": 0.9652,
+      "step": 5666
+    },
+    {
+      "epoch": 0.4368640148011101,
+      "grad_norm": 3.6662580966949463,
+      "learning_rate": 6.252443918934112e-06,
+      "loss": 0.9143,
+      "step": 5667
+    },
+    {
+      "epoch": 0.43694110391612706,
+      "grad_norm": 4.207685470581055,
+      "learning_rate": 6.251235233778075e-06,
+      "loss": 0.9565,
+      "step": 5668
+    },
+    {
+      "epoch": 0.437018193031144,
+      "grad_norm": 3.748297929763794,
+      "learning_rate": 6.250026470614003e-06,
+      "loss": 1.1053,
+      "step": 5669
+    },
+    {
+      "epoch": 0.43709528214616095,
+      "grad_norm": 3.7725093364715576,
+      "learning_rate": 6.248817629517253e-06,
+      "loss": 0.9306,
+      "step": 5670
+    },
+    {
+      "epoch": 0.4371723712611779,
+      "grad_norm": 3.868713617324829,
+      "learning_rate": 6.247608710563192e-06,
+      "loss": 0.9549,
+      "step": 5671
+    },
+    {
+      "epoch": 0.4372494603761949,
+      "grad_norm": 3.3361618518829346,
+      "learning_rate": 6.2463997138271905e-06,
+      "loss": 0.9246,
+      "step": 5672
+    },
+    {
+      "epoch": 0.43732654949121186,
+      "grad_norm": 3.693483591079712,
+      "learning_rate": 6.24519063938462e-06,
+      "loss": 0.9676,
+      "step": 5673
+    },
+    {
+      "epoch": 0.4374036386062288,
+      "grad_norm": 3.5362865924835205,
+      "learning_rate": 6.243981487310864e-06,
+      "loss": 1.0187,
+      "step": 5674
+    },
+    {
+      "epoch": 0.43748072772124574,
+      "grad_norm": 3.2017054557800293,
+      "learning_rate": 6.242772257681304e-06,
+      "loss": 0.9676,
+      "step": 5675
+    },
+    {
+      "epoch": 0.4375578168362627,
+      "grad_norm": 3.7719571590423584,
+      "learning_rate": 6.241562950571331e-06,
+      "loss": 0.9562,
+      "step": 5676
+    },
+    {
+      "epoch": 0.4376349059512797,
+      "grad_norm": 3.446551561355591,
+      "learning_rate": 6.240353566056339e-06,
+      "loss": 0.9793,
+      "step": 5677
+    },
+    {
+      "epoch": 0.43771199506629666,
+      "grad_norm": 4.14500617980957,
+      "learning_rate": 6.239144104211724e-06,
+      "loss": 1.0242,
+      "step": 5678
+    },
+    {
+      "epoch": 0.4377890841813136,
+      "grad_norm": 4.008180141448975,
+      "learning_rate": 6.237934565112893e-06,
+      "loss": 1.0776,
+      "step": 5679
+    },
+    {
+      "epoch": 0.43786617329633054,
+      "grad_norm": 3.641493082046509,
+      "learning_rate": 6.2367249488352535e-06,
+      "loss": 1.0238,
+      "step": 5680
+    },
+    {
+      "epoch": 0.4379432624113475,
+      "grad_norm": 3.8391215801239014,
+      "learning_rate": 6.235515255454218e-06,
+      "loss": 1.0155,
+      "step": 5681
+    },
+    {
+      "epoch": 0.4380203515263645,
+      "grad_norm": 3.3518407344818115,
+      "learning_rate": 6.234305485045205e-06,
+      "loss": 0.8861,
+      "step": 5682
+    },
+    {
+      "epoch": 0.43809744064138145,
+      "grad_norm": 3.9264771938323975,
+      "learning_rate": 6.2330956376836384e-06,
+      "loss": 1.0198,
+      "step": 5683
+    },
+    {
+      "epoch": 0.4381745297563984,
+      "grad_norm": 3.6416938304901123,
+      "learning_rate": 6.231885713444944e-06,
+      "loss": 0.9194,
+      "step": 5684
+    },
+    {
+      "epoch": 0.43825161887141534,
+      "grad_norm": 3.5663013458251953,
+      "learning_rate": 6.230675712404557e-06,
+      "loss": 0.8705,
+      "step": 5685
+    },
+    {
+      "epoch": 0.4383287079864323,
+      "grad_norm": 3.5762104988098145,
+      "learning_rate": 6.229465634637912e-06,
+      "loss": 1.0757,
+      "step": 5686
+    },
+    {
+      "epoch": 0.4384057971014493,
+      "grad_norm": 4.03668212890625,
+      "learning_rate": 6.2282554802204535e-06,
+      "loss": 0.9993,
+      "step": 5687
+    },
+    {
+      "epoch": 0.43848288621646625,
+      "grad_norm": 3.1501452922821045,
+      "learning_rate": 6.2270452492276265e-06,
+      "loss": 0.8529,
+      "step": 5688
+    },
+    {
+      "epoch": 0.4385599753314832,
+      "grad_norm": 4.269925117492676,
+      "learning_rate": 6.225834941734883e-06,
+      "loss": 0.9343,
+      "step": 5689
+    },
+    {
+      "epoch": 0.43863706444650014,
+      "grad_norm": 3.7110297679901123,
+      "learning_rate": 6.22462455781768e-06,
+      "loss": 0.9587,
+      "step": 5690
+    },
+    {
+      "epoch": 0.4387141535615171,
+      "grad_norm": 3.938750743865967,
+      "learning_rate": 6.223414097551478e-06,
+      "loss": 0.979,
+      "step": 5691
+    },
+    {
+      "epoch": 0.4387912426765341,
+      "grad_norm": 3.5555200576782227,
+      "learning_rate": 6.2222035610117424e-06,
+      "loss": 0.9671,
+      "step": 5692
+    },
+    {
+      "epoch": 0.43886833179155105,
+      "grad_norm": 3.7331223487854004,
+      "learning_rate": 6.220992948273947e-06,
+      "loss": 0.8927,
+      "step": 5693
+    },
+    {
+      "epoch": 0.438945420906568,
+      "grad_norm": 3.361193895339966,
+      "learning_rate": 6.219782259413562e-06,
+      "loss": 0.8326,
+      "step": 5694
+    },
+    {
+      "epoch": 0.43902251002158493,
+      "grad_norm": 3.461210250854492,
+      "learning_rate": 6.218571494506073e-06,
+      "loss": 0.8966,
+      "step": 5695
+    },
+    {
+      "epoch": 0.4390995991366019,
+      "grad_norm": 3.2159628868103027,
+      "learning_rate": 6.2173606536269605e-06,
+      "loss": 0.8993,
+      "step": 5696
+    },
+    {
+      "epoch": 0.4391766882516189,
+      "grad_norm": 3.616848945617676,
+      "learning_rate": 6.2161497368517175e-06,
+      "loss": 0.9901,
+      "step": 5697
+    },
+    {
+      "epoch": 0.43925377736663584,
+      "grad_norm": 3.6139261722564697,
+      "learning_rate": 6.214938744255837e-06,
+      "loss": 0.8664,
+      "step": 5698
+    },
+    {
+      "epoch": 0.4393308664816528,
+      "grad_norm": 3.526050567626953,
+      "learning_rate": 6.213727675914818e-06,
+      "loss": 1.0925,
+      "step": 5699
+    },
+    {
+      "epoch": 0.43940795559666973,
+      "grad_norm": 3.36438250541687,
+      "learning_rate": 6.212516531904164e-06,
+      "loss": 0.9525,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4394850447116867,
+      "grad_norm": 3.640000343322754,
+      "learning_rate": 6.2113053122993846e-06,
+      "loss": 1.0309,
+      "step": 5701
+    },
+    {
+      "epoch": 0.43956213382670367,
+      "grad_norm": 3.7838895320892334,
+      "learning_rate": 6.210094017175991e-06,
+      "loss": 0.9115,
+      "step": 5702
+    },
+    {
+      "epoch": 0.43963922294172064,
+      "grad_norm": 3.5852866172790527,
+      "learning_rate": 6.208882646609505e-06,
+      "loss": 0.918,
+      "step": 5703
+    },
+    {
+      "epoch": 0.4397163120567376,
+      "grad_norm": 3.5891425609588623,
+      "learning_rate": 6.207671200675446e-06,
+      "loss": 0.961,
+      "step": 5704
+    },
+    {
+      "epoch": 0.4397934011717545,
+      "grad_norm": 3.654510974884033,
+      "learning_rate": 6.206459679449341e-06,
+      "loss": 0.9256,
+      "step": 5705
+    },
+    {
+      "epoch": 0.4398704902867715,
+      "grad_norm": 3.4456582069396973,
+      "learning_rate": 6.205248083006725e-06,
+      "loss": 0.9671,
+      "step": 5706
+    },
+    {
+      "epoch": 0.43994757940178847,
+      "grad_norm": 3.725618362426758,
+      "learning_rate": 6.204036411423133e-06,
+      "loss": 1.0621,
+      "step": 5707
+    },
+    {
+      "epoch": 0.44002466851680544,
+      "grad_norm": 3.431455135345459,
+      "learning_rate": 6.202824664774107e-06,
+      "loss": 0.9588,
+      "step": 5708
+    },
+    {
+      "epoch": 0.4401017576318224,
+      "grad_norm": 3.980421543121338,
+      "learning_rate": 6.201612843135192e-06,
+      "loss": 1.0372,
+      "step": 5709
+    },
+    {
+      "epoch": 0.4401788467468393,
+      "grad_norm": 3.611001491546631,
+      "learning_rate": 6.200400946581939e-06,
+      "loss": 0.9616,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4402559358618563,
+      "grad_norm": 3.704310417175293,
+      "learning_rate": 6.199188975189905e-06,
+      "loss": 1.0496,
+      "step": 5711
+    },
+    {
+      "epoch": 0.44033302497687327,
+      "grad_norm": 3.831941843032837,
+      "learning_rate": 6.197976929034649e-06,
+      "loss": 0.9209,
+      "step": 5712
+    },
+    {
+      "epoch": 0.44041011409189024,
+      "grad_norm": 3.790203094482422,
+      "learning_rate": 6.196764808191735e-06,
+      "loss": 0.9974,
+      "step": 5713
+    },
+    {
+      "epoch": 0.4404872032069072,
+      "grad_norm": 3.4942948818206787,
+      "learning_rate": 6.195552612736735e-06,
+      "loss": 1.0154,
+      "step": 5714
+    },
+    {
+      "epoch": 0.4405642923219241,
+      "grad_norm": 3.5759153366088867,
+      "learning_rate": 6.19434034274522e-06,
+      "loss": 0.9808,
+      "step": 5715
+    },
+    {
+      "epoch": 0.4406413814369411,
+      "grad_norm": 3.54484486579895,
+      "learning_rate": 6.19312799829277e-06,
+      "loss": 0.8969,
+      "step": 5716
+    },
+    {
+      "epoch": 0.44071847055195806,
+      "grad_norm": 3.6887714862823486,
+      "learning_rate": 6.191915579454971e-06,
+      "loss": 0.9693,
+      "step": 5717
+    },
+    {
+      "epoch": 0.44079555966697503,
+      "grad_norm": 3.173921585083008,
+      "learning_rate": 6.1907030863074055e-06,
+      "loss": 0.9008,
+      "step": 5718
+    },
+    {
+      "epoch": 0.440872648781992,
+      "grad_norm": 3.73128080368042,
+      "learning_rate": 6.189490518925673e-06,
+      "loss": 1.0533,
+      "step": 5719
+    },
+    {
+      "epoch": 0.4409497378970089,
+      "grad_norm": 3.7085869312286377,
+      "learning_rate": 6.188277877385365e-06,
+      "loss": 0.9966,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4410268270120259,
+      "grad_norm": 3.4095280170440674,
+      "learning_rate": 6.187065161762085e-06,
+      "loss": 0.9337,
+      "step": 5721
+    },
+    {
+      "epoch": 0.44110391612704286,
+      "grad_norm": 3.37434720993042,
+      "learning_rate": 6.1858523721314425e-06,
+      "loss": 0.9576,
+      "step": 5722
+    },
+    {
+      "epoch": 0.44118100524205983,
+      "grad_norm": 3.524092674255371,
+      "learning_rate": 6.184639508569043e-06,
+      "loss": 0.943,
+      "step": 5723
+    },
+    {
+      "epoch": 0.4412580943570768,
+      "grad_norm": 3.7031326293945312,
+      "learning_rate": 6.183426571150508e-06,
+      "loss": 0.941,
+      "step": 5724
+    },
+    {
+      "epoch": 0.4413351834720937,
+      "grad_norm": 3.827249050140381,
+      "learning_rate": 6.182213559951456e-06,
+      "loss": 0.9173,
+      "step": 5725
+    },
+    {
+      "epoch": 0.4414122725871107,
+      "grad_norm": 3.6542346477508545,
+      "learning_rate": 6.181000475047509e-06,
+      "loss": 0.9171,
+      "step": 5726
+    },
+    {
+      "epoch": 0.44148936170212766,
+      "grad_norm": 3.8441221714019775,
+      "learning_rate": 6.1797873165143005e-06,
+      "loss": 1.0174,
+      "step": 5727
+    },
+    {
+      "epoch": 0.44156645081714463,
+      "grad_norm": 3.4966237545013428,
+      "learning_rate": 6.178574084427464e-06,
+      "loss": 0.9042,
+      "step": 5728
+    },
+    {
+      "epoch": 0.4416435399321616,
+      "grad_norm": 3.398782968521118,
+      "learning_rate": 6.1773607788626355e-06,
+      "loss": 0.9617,
+      "step": 5729
+    },
+    {
+      "epoch": 0.4417206290471785,
+      "grad_norm": 3.6098642349243164,
+      "learning_rate": 6.176147399895461e-06,
+      "loss": 0.9554,
+      "step": 5730
+    },
+    {
+      "epoch": 0.4417977181621955,
+      "grad_norm": 4.328033924102783,
+      "learning_rate": 6.174933947601587e-06,
+      "loss": 1.0728,
+      "step": 5731
+    },
+    {
+      "epoch": 0.44187480727721246,
+      "grad_norm": 3.6380655765533447,
+      "learning_rate": 6.173720422056666e-06,
+      "loss": 0.9786,
+      "step": 5732
+    },
+    {
+      "epoch": 0.4419518963922294,
+      "grad_norm": 3.8986289501190186,
+      "learning_rate": 6.172506823336357e-06,
+      "loss": 0.9966,
+      "step": 5733
+    },
+    {
+      "epoch": 0.4420289855072464,
+      "grad_norm": 3.5367014408111572,
+      "learning_rate": 6.17129315151632e-06,
+      "loss": 0.9705,
+      "step": 5734
+    },
+    {
+      "epoch": 0.4421060746222633,
+      "grad_norm": 3.6071369647979736,
+      "learning_rate": 6.17007940667222e-06,
+      "loss": 0.8371,
+      "step": 5735
+    },
+    {
+      "epoch": 0.4421831637372803,
+      "grad_norm": 3.784031867980957,
+      "learning_rate": 6.16886558887973e-06,
+      "loss": 0.9845,
+      "step": 5736
+    },
+    {
+      "epoch": 0.44226025285229725,
+      "grad_norm": 3.7865121364593506,
+      "learning_rate": 6.167651698214524e-06,
+      "loss": 0.9009,
+      "step": 5737
+    },
+    {
+      "epoch": 0.4423373419673142,
+      "grad_norm": 3.29433274269104,
+      "learning_rate": 6.166437734752282e-06,
+      "loss": 0.8721,
+      "step": 5738
+    },
+    {
+      "epoch": 0.4424144310823312,
+      "grad_norm": 3.7611396312713623,
+      "learning_rate": 6.165223698568689e-06,
+      "loss": 1.0397,
+      "step": 5739
+    },
+    {
+      "epoch": 0.4424915201973481,
+      "grad_norm": 3.6707139015197754,
+      "learning_rate": 6.164009589739431e-06,
+      "loss": 0.9511,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4425686093123651,
+      "grad_norm": 3.516786575317383,
+      "learning_rate": 6.162795408340206e-06,
+      "loss": 0.9583,
+      "step": 5741
+    },
+    {
+      "epoch": 0.44264569842738205,
+      "grad_norm": 3.280879259109497,
+      "learning_rate": 6.161581154446709e-06,
+      "loss": 0.937,
+      "step": 5742
+    },
+    {
+      "epoch": 0.442722787542399,
+      "grad_norm": 3.570941209793091,
+      "learning_rate": 6.1603668281346425e-06,
+      "loss": 1.048,
+      "step": 5743
+    },
+    {
+      "epoch": 0.442799876657416,
+      "grad_norm": 3.6662280559539795,
+      "learning_rate": 6.159152429479714e-06,
+      "loss": 0.9452,
+      "step": 5744
+    },
+    {
+      "epoch": 0.4428769657724329,
+      "grad_norm": 3.7819111347198486,
+      "learning_rate": 6.157937958557635e-06,
+      "loss": 1.0653,
+      "step": 5745
+    },
+    {
+      "epoch": 0.4429540548874499,
+      "grad_norm": 3.768911361694336,
+      "learning_rate": 6.156723415444123e-06,
+      "loss": 0.8976,
+      "step": 5746
+    },
+    {
+      "epoch": 0.44303114400246685,
+      "grad_norm": 3.1364481449127197,
+      "learning_rate": 6.155508800214894e-06,
+      "loss": 0.9241,
+      "step": 5747
+    },
+    {
+      "epoch": 0.4431082331174838,
+      "grad_norm": 4.069226264953613,
+      "learning_rate": 6.154294112945678e-06,
+      "loss": 1.0962,
+      "step": 5748
+    },
+    {
+      "epoch": 0.4431853222325008,
+      "grad_norm": 3.394063711166382,
+      "learning_rate": 6.153079353712201e-06,
+      "loss": 0.8823,
+      "step": 5749
+    },
+    {
+      "epoch": 0.4432624113475177,
+      "grad_norm": 3.523238182067871,
+      "learning_rate": 6.1518645225902e-06,
+      "loss": 1.007,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4433395004625347,
+      "grad_norm": 3.942383050918579,
+      "learning_rate": 6.150649619655411e-06,
+      "loss": 0.9954,
+      "step": 5751
+    },
+    {
+      "epoch": 0.44341658957755165,
+      "grad_norm": 3.7958526611328125,
+      "learning_rate": 6.149434644983576e-06,
+      "loss": 0.9991,
+      "step": 5752
+    },
+    {
+      "epoch": 0.4434936786925686,
+      "grad_norm": 3.4549882411956787,
+      "learning_rate": 6.148219598650444e-06,
+      "loss": 0.9293,
+      "step": 5753
+    },
+    {
+      "epoch": 0.4435707678075856,
+      "grad_norm": 4.054780960083008,
+      "learning_rate": 6.1470044807317695e-06,
+      "loss": 0.8547,
+      "step": 5754
+    },
+    {
+      "epoch": 0.4436478569226025,
+      "grad_norm": 3.937290906906128,
+      "learning_rate": 6.145789291303305e-06,
+      "loss": 0.9548,
+      "step": 5755
+    },
+    {
+      "epoch": 0.44372494603761947,
+      "grad_norm": 3.627803325653076,
+      "learning_rate": 6.144574030440811e-06,
+      "loss": 0.9164,
+      "step": 5756
+    },
+    {
+      "epoch": 0.44380203515263644,
+      "grad_norm": 3.1679458618164062,
+      "learning_rate": 6.143358698220055e-06,
+      "loss": 0.9428,
+      "step": 5757
+    },
+    {
+      "epoch": 0.4438791242676534,
+      "grad_norm": 3.7125744819641113,
+      "learning_rate": 6.142143294716806e-06,
+      "loss": 0.91,
+      "step": 5758
+    },
+    {
+      "epoch": 0.4439562133826704,
+      "grad_norm": 4.144897937774658,
+      "learning_rate": 6.140927820006838e-06,
+      "loss": 1.0763,
+      "step": 5759
+    },
+    {
+      "epoch": 0.4440333024976873,
+      "grad_norm": 3.8823111057281494,
+      "learning_rate": 6.139712274165929e-06,
+      "loss": 0.9906,
+      "step": 5760
+    },
+    {
+      "epoch": 0.44411039161270427,
+      "grad_norm": 3.5604453086853027,
+      "learning_rate": 6.138496657269862e-06,
+      "loss": 0.8591,
+      "step": 5761
+    },
+    {
+      "epoch": 0.44418748072772124,
+      "grad_norm": 3.5651559829711914,
+      "learning_rate": 6.1372809693944255e-06,
+      "loss": 1.0111,
+      "step": 5762
+    },
+    {
+      "epoch": 0.4442645698427382,
+      "grad_norm": 3.4907076358795166,
+      "learning_rate": 6.1360652106154095e-06,
+      "loss": 0.9794,
+      "step": 5763
+    },
+    {
+      "epoch": 0.4443416589577552,
+      "grad_norm": 3.5271456241607666,
+      "learning_rate": 6.134849381008613e-06,
+      "loss": 1.0717,
+      "step": 5764
+    },
+    {
+      "epoch": 0.44441874807277215,
+      "grad_norm": 3.589846611022949,
+      "learning_rate": 6.1336334806498356e-06,
+      "loss": 1.0364,
+      "step": 5765
+    },
+    {
+      "epoch": 0.44449583718778907,
+      "grad_norm": 3.321850299835205,
+      "learning_rate": 6.13241750961488e-06,
+      "loss": 0.9513,
+      "step": 5766
+    },
+    {
+      "epoch": 0.44457292630280604,
+      "grad_norm": 3.9627490043640137,
+      "learning_rate": 6.13120146797956e-06,
+      "loss": 0.9762,
+      "step": 5767
+    },
+    {
+      "epoch": 0.444650015417823,
+      "grad_norm": 3.2399096488952637,
+      "learning_rate": 6.129985355819684e-06,
+      "loss": 0.9448,
+      "step": 5768
+    },
+    {
+      "epoch": 0.44472710453284,
+      "grad_norm": 3.653228521347046,
+      "learning_rate": 6.128769173211075e-06,
+      "loss": 0.9295,
+      "step": 5769
+    },
+    {
+      "epoch": 0.44480419364785695,
+      "grad_norm": 3.379409074783325,
+      "learning_rate": 6.127552920229556e-06,
+      "loss": 0.8892,
+      "step": 5770
+    },
+    {
+      "epoch": 0.44488128276287386,
+      "grad_norm": 3.461467981338501,
+      "learning_rate": 6.126336596950949e-06,
+      "loss": 0.9007,
+      "step": 5771
+    },
+    {
+      "epoch": 0.44495837187789083,
+      "grad_norm": 3.425234794616699,
+      "learning_rate": 6.1251202034510905e-06,
+      "loss": 0.9136,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4450354609929078,
+      "grad_norm": 3.859802007675171,
+      "learning_rate": 6.123903739805815e-06,
+      "loss": 0.9927,
+      "step": 5773
+    },
+    {
+      "epoch": 0.4451125501079248,
+      "grad_norm": 3.9901375770568848,
+      "learning_rate": 6.1226872060909606e-06,
+      "loss": 1.0381,
+      "step": 5774
+    },
+    {
+      "epoch": 0.44518963922294175,
+      "grad_norm": 3.5192387104034424,
+      "learning_rate": 6.121470602382375e-06,
+      "loss": 0.9129,
+      "step": 5775
+    },
+    {
+      "epoch": 0.44526672833795866,
+      "grad_norm": 3.422767400741577,
+      "learning_rate": 6.1202539287559035e-06,
+      "loss": 0.9601,
+      "step": 5776
+    },
+    {
+      "epoch": 0.44534381745297563,
+      "grad_norm": 3.6714818477630615,
+      "learning_rate": 6.119037185287402e-06,
+      "loss": 0.9204,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4454209065679926,
+      "grad_norm": 3.7968335151672363,
+      "learning_rate": 6.1178203720527285e-06,
+      "loss": 0.9916,
+      "step": 5778
+    },
+    {
+      "epoch": 0.4454979956830096,
+      "grad_norm": 3.772531747817993,
+      "learning_rate": 6.116603489127744e-06,
+      "loss": 1.0515,
+      "step": 5779
+    },
+    {
+      "epoch": 0.44557508479802654,
+      "grad_norm": 3.4873077869415283,
+      "learning_rate": 6.1153865365883146e-06,
+      "loss": 0.9606,
+      "step": 5780
+    },
+    {
+      "epoch": 0.44565217391304346,
+      "grad_norm": 3.7624447345733643,
+      "learning_rate": 6.114169514510312e-06,
+      "loss": 0.9,
+      "step": 5781
+    },
+    {
+      "epoch": 0.44572926302806043,
+      "grad_norm": 3.8201286792755127,
+      "learning_rate": 6.112952422969611e-06,
+      "loss": 0.9601,
+      "step": 5782
+    },
+    {
+      "epoch": 0.4458063521430774,
+      "grad_norm": 4.063814640045166,
+      "learning_rate": 6.111735262042088e-06,
+      "loss": 0.9778,
+      "step": 5783
+    },
+    {
+      "epoch": 0.44588344125809437,
+      "grad_norm": 3.7615840435028076,
+      "learning_rate": 6.11051803180363e-06,
+      "loss": 0.9212,
+      "step": 5784
+    },
+    {
+      "epoch": 0.44596053037311134,
+      "grad_norm": 3.5406413078308105,
+      "learning_rate": 6.109300732330126e-06,
+      "loss": 0.9414,
+      "step": 5785
+    },
+    {
+      "epoch": 0.44603761948812826,
+      "grad_norm": 3.9752652645111084,
+      "learning_rate": 6.1080833636974655e-06,
+      "loss": 1.0694,
+      "step": 5786
+    },
+    {
+      "epoch": 0.4461147086031452,
+      "grad_norm": 3.561919689178467,
+      "learning_rate": 6.106865925981547e-06,
+      "loss": 0.9985,
+      "step": 5787
+    },
+    {
+      "epoch": 0.4461917977181622,
+      "grad_norm": 3.6946768760681152,
+      "learning_rate": 6.105648419258271e-06,
+      "loss": 1.0049,
+      "step": 5788
+    },
+    {
+      "epoch": 0.44626888683317917,
+      "grad_norm": 3.748198986053467,
+      "learning_rate": 6.10443084360354e-06,
+      "loss": 0.9642,
+      "step": 5789
+    },
+    {
+      "epoch": 0.44634597594819614,
+      "grad_norm": 3.938978672027588,
+      "learning_rate": 6.103213199093267e-06,
+      "loss": 1.0989,
+      "step": 5790
+    },
+    {
+      "epoch": 0.44642306506321305,
+      "grad_norm": 3.319303035736084,
+      "learning_rate": 6.101995485803367e-06,
+      "loss": 1.0071,
+      "step": 5791
+    },
+    {
+      "epoch": 0.44650015417823,
+      "grad_norm": 3.9416778087615967,
+      "learning_rate": 6.100777703809753e-06,
+      "loss": 1.038,
+      "step": 5792
+    },
+    {
+      "epoch": 0.446577243293247,
+      "grad_norm": 3.5364232063293457,
+      "learning_rate": 6.0995598531883504e-06,
+      "loss": 0.862,
+      "step": 5793
+    },
+    {
+      "epoch": 0.44665433240826397,
+      "grad_norm": 3.440767526626587,
+      "learning_rate": 6.098341934015088e-06,
+      "loss": 0.9708,
+      "step": 5794
+    },
+    {
+      "epoch": 0.44673142152328094,
+      "grad_norm": 3.8628475666046143,
+      "learning_rate": 6.097123946365893e-06,
+      "loss": 0.936,
+      "step": 5795
+    },
+    {
+      "epoch": 0.44680851063829785,
+      "grad_norm": 3.6398117542266846,
+      "learning_rate": 6.095905890316701e-06,
+      "loss": 1.0004,
+      "step": 5796
+    },
+    {
+      "epoch": 0.4468855997533148,
+      "grad_norm": 4.144468307495117,
+      "learning_rate": 6.094687765943455e-06,
+      "loss": 1.0691,
+      "step": 5797
+    },
+    {
+      "epoch": 0.4469626888683318,
+      "grad_norm": 3.321753740310669,
+      "learning_rate": 6.093469573322094e-06,
+      "loss": 1.0334,
+      "step": 5798
+    },
+    {
+      "epoch": 0.44703977798334876,
+      "grad_norm": 3.512993574142456,
+      "learning_rate": 6.09225131252857e-06,
+      "loss": 1.0446,
+      "step": 5799
+    },
+    {
+      "epoch": 0.44711686709836573,
+      "grad_norm": 3.447844982147217,
+      "learning_rate": 6.091032983638833e-06,
+      "loss": 0.9838,
+      "step": 5800
+    },
+    {
+      "epoch": 0.44719395621338265,
+      "grad_norm": 3.6095683574676514,
+      "learning_rate": 6.0898145867288395e-06,
+      "loss": 0.9409,
+      "step": 5801
+    },
+    {
+      "epoch": 0.4472710453283996,
+      "grad_norm": 3.7775659561157227,
+      "learning_rate": 6.088596121874552e-06,
+      "loss": 1.0464,
+      "step": 5802
+    },
+    {
+      "epoch": 0.4473481344434166,
+      "grad_norm": 3.305006980895996,
+      "learning_rate": 6.087377589151933e-06,
+      "loss": 0.9671,
+      "step": 5803
+    },
+    {
+      "epoch": 0.44742522355843356,
+      "grad_norm": 3.450387477874756,
+      "learning_rate": 6.086158988636953e-06,
+      "loss": 0.9715,
+      "step": 5804
+    },
+    {
+      "epoch": 0.44750231267345053,
+      "grad_norm": 3.848440647125244,
+      "learning_rate": 6.084940320405587e-06,
+      "loss": 1.0587,
+      "step": 5805
+    },
+    {
+      "epoch": 0.44757940178846745,
+      "grad_norm": 3.4928512573242188,
+      "learning_rate": 6.08372158453381e-06,
+      "loss": 0.922,
+      "step": 5806
+    },
+    {
+      "epoch": 0.4476564909034844,
+      "grad_norm": 3.3881824016571045,
+      "learning_rate": 6.082502781097603e-06,
+      "loss": 0.9111,
+      "step": 5807
+    },
+    {
+      "epoch": 0.4477335800185014,
+      "grad_norm": 3.2955267429351807,
+      "learning_rate": 6.081283910172956e-06,
+      "loss": 0.9145,
+      "step": 5808
+    },
+    {
+      "epoch": 0.44781066913351836,
+      "grad_norm": 3.5794951915740967,
+      "learning_rate": 6.080064971835857e-06,
+      "loss": 0.9883,
+      "step": 5809
+    },
+    {
+      "epoch": 0.44788775824853533,
+      "grad_norm": 3.3705379962921143,
+      "learning_rate": 6.078845966162302e-06,
+      "loss": 0.9652,
+      "step": 5810
+    },
+    {
+      "epoch": 0.44796484736355224,
+      "grad_norm": 3.782891273498535,
+      "learning_rate": 6.077626893228288e-06,
+      "loss": 1.0398,
+      "step": 5811
+    },
+    {
+      "epoch": 0.4480419364785692,
+      "grad_norm": 3.3794403076171875,
+      "learning_rate": 6.076407753109818e-06,
+      "loss": 0.9309,
+      "step": 5812
+    },
+    {
+      "epoch": 0.4481190255935862,
+      "grad_norm": 3.325587272644043,
+      "learning_rate": 6.0751885458829e-06,
+      "loss": 0.9469,
+      "step": 5813
+    },
+    {
+      "epoch": 0.44819611470860315,
+      "grad_norm": 3.5415618419647217,
+      "learning_rate": 6.073969271623543e-06,
+      "loss": 0.9648,
+      "step": 5814
+    },
+    {
+      "epoch": 0.4482732038236201,
+      "grad_norm": 3.56663179397583,
+      "learning_rate": 6.072749930407767e-06,
+      "loss": 1.0515,
+      "step": 5815
+    },
+    {
+      "epoch": 0.44835029293863704,
+      "grad_norm": 4.031259536743164,
+      "learning_rate": 6.071530522311586e-06,
+      "loss": 1.0355,
+      "step": 5816
+    },
+    {
+      "epoch": 0.448427382053654,
+      "grad_norm": 3.370077610015869,
+      "learning_rate": 6.070311047411027e-06,
+      "loss": 0.8723,
+      "step": 5817
+    },
+    {
+      "epoch": 0.448504471168671,
+      "grad_norm": 3.9571077823638916,
+      "learning_rate": 6.069091505782119e-06,
+      "loss": 0.9908,
+      "step": 5818
+    },
+    {
+      "epoch": 0.44858156028368795,
+      "grad_norm": 3.552165985107422,
+      "learning_rate": 6.067871897500891e-06,
+      "loss": 0.9056,
+      "step": 5819
+    },
+    {
+      "epoch": 0.4486586493987049,
+      "grad_norm": 3.8415639400482178,
+      "learning_rate": 6.066652222643381e-06,
+      "loss": 1.0353,
+      "step": 5820
+    },
+    {
+      "epoch": 0.44873573851372184,
+      "grad_norm": 3.4832634925842285,
+      "learning_rate": 6.0654324812856305e-06,
+      "loss": 0.9881,
+      "step": 5821
+    },
+    {
+      "epoch": 0.4488128276287388,
+      "grad_norm": 3.5420501232147217,
+      "learning_rate": 6.064212673503682e-06,
+      "loss": 0.9398,
+      "step": 5822
+    },
+    {
+      "epoch": 0.4488899167437558,
+      "grad_norm": 3.6258022785186768,
+      "learning_rate": 6.062992799373585e-06,
+      "loss": 1.0994,
+      "step": 5823
+    },
+    {
+      "epoch": 0.44896700585877275,
+      "grad_norm": 3.758962631225586,
+      "learning_rate": 6.061772858971392e-06,
+      "loss": 0.9196,
+      "step": 5824
+    },
+    {
+      "epoch": 0.4490440949737897,
+      "grad_norm": 3.7305667400360107,
+      "learning_rate": 6.060552852373161e-06,
+      "loss": 0.9636,
+      "step": 5825
+    },
+    {
+      "epoch": 0.44912118408880664,
+      "grad_norm": 3.6690361499786377,
+      "learning_rate": 6.059332779654953e-06,
+      "loss": 1.0034,
+      "step": 5826
+    },
+    {
+      "epoch": 0.4491982732038236,
+      "grad_norm": 3.7183585166931152,
+      "learning_rate": 6.05811264089283e-06,
+      "loss": 1.0183,
+      "step": 5827
+    },
+    {
+      "epoch": 0.4492753623188406,
+      "grad_norm": 3.5112946033477783,
+      "learning_rate": 6.056892436162866e-06,
+      "loss": 0.9799,
+      "step": 5828
+    },
+    {
+      "epoch": 0.44935245143385755,
+      "grad_norm": 3.783031940460205,
+      "learning_rate": 6.055672165541132e-06,
+      "loss": 0.9735,
+      "step": 5829
+    },
+    {
+      "epoch": 0.4494295405488745,
+      "grad_norm": 4.117906093597412,
+      "learning_rate": 6.0544518291037055e-06,
+      "loss": 1.0281,
+      "step": 5830
+    },
+    {
+      "epoch": 0.44950662966389143,
+      "grad_norm": 3.5739166736602783,
+      "learning_rate": 6.05323142692667e-06,
+      "loss": 1.0257,
+      "step": 5831
+    },
+    {
+      "epoch": 0.4495837187789084,
+      "grad_norm": 3.51526141166687,
+      "learning_rate": 6.052010959086109e-06,
+      "loss": 0.9378,
+      "step": 5832
+    },
+    {
+      "epoch": 0.4496608078939254,
+      "grad_norm": 4.066465377807617,
+      "learning_rate": 6.050790425658112e-06,
+      "loss": 1.0126,
+      "step": 5833
+    },
+    {
+      "epoch": 0.44973789700894234,
+      "grad_norm": 3.460562229156494,
+      "learning_rate": 6.049569826718776e-06,
+      "loss": 0.9873,
+      "step": 5834
+    },
+    {
+      "epoch": 0.4498149861239593,
+      "grad_norm": 3.6307222843170166,
+      "learning_rate": 6.048349162344196e-06,
+      "loss": 0.9094,
+      "step": 5835
+    },
+    {
+      "epoch": 0.44989207523897623,
+      "grad_norm": 4.065826892852783,
+      "learning_rate": 6.047128432610476e-06,
+      "loss": 0.8306,
+      "step": 5836
+    },
+    {
+      "epoch": 0.4499691643539932,
+      "grad_norm": 3.523407459259033,
+      "learning_rate": 6.045907637593722e-06,
+      "loss": 0.9603,
+      "step": 5837
+    },
+    {
+      "epoch": 0.45004625346901017,
+      "grad_norm": 3.768115997314453,
+      "learning_rate": 6.044686777370042e-06,
+      "loss": 1.0456,
+      "step": 5838
+    },
+    {
+      "epoch": 0.45012334258402714,
+      "grad_norm": 3.527888059616089,
+      "learning_rate": 6.043465852015553e-06,
+      "loss": 1.0582,
+      "step": 5839
+    },
+    {
+      "epoch": 0.4502004316990441,
+      "grad_norm": 3.7846176624298096,
+      "learning_rate": 6.042244861606373e-06,
+      "loss": 1.0712,
+      "step": 5840
+    },
+    {
+      "epoch": 0.450277520814061,
+      "grad_norm": 3.638317584991455,
+      "learning_rate": 6.041023806218622e-06,
+      "loss": 0.9448,
+      "step": 5841
+    },
+    {
+      "epoch": 0.450354609929078,
+      "grad_norm": 3.303644895553589,
+      "learning_rate": 6.03980268592843e-06,
+      "loss": 0.914,
+      "step": 5842
+    },
+    {
+      "epoch": 0.45043169904409497,
+      "grad_norm": 3.3192503452301025,
+      "learning_rate": 6.0385815008119254e-06,
+      "loss": 0.9118,
+      "step": 5843
+    },
+    {
+      "epoch": 0.45050878815911194,
+      "grad_norm": 3.531160831451416,
+      "learning_rate": 6.037360250945243e-06,
+      "loss": 0.966,
+      "step": 5844
+    },
+    {
+      "epoch": 0.4505858772741289,
+      "grad_norm": 3.8585782051086426,
+      "learning_rate": 6.036138936404521e-06,
+      "loss": 0.9231,
+      "step": 5845
+    },
+    {
+      "epoch": 0.4506629663891459,
+      "grad_norm": 3.3526854515075684,
+      "learning_rate": 6.034917557265903e-06,
+      "loss": 0.9663,
+      "step": 5846
+    },
+    {
+      "epoch": 0.4507400555041628,
+      "grad_norm": 3.64371919631958,
+      "learning_rate": 6.033696113605536e-06,
+      "loss": 0.9702,
+      "step": 5847
+    },
+    {
+      "epoch": 0.45081714461917977,
+      "grad_norm": 3.5181994438171387,
+      "learning_rate": 6.0324746054995685e-06,
+      "loss": 0.9026,
+      "step": 5848
+    },
+    {
+      "epoch": 0.45089423373419674,
+      "grad_norm": 3.487680196762085,
+      "learning_rate": 6.031253033024158e-06,
+      "loss": 0.9488,
+      "step": 5849
+    },
+    {
+      "epoch": 0.4509713228492137,
+      "grad_norm": 3.600533962249756,
+      "learning_rate": 6.030031396255462e-06,
+      "loss": 1.0593,
+      "step": 5850
+    },
+    {
+      "epoch": 0.4510484119642307,
+      "grad_norm": 3.4745328426361084,
+      "learning_rate": 6.028809695269641e-06,
+      "loss": 0.937,
+      "step": 5851
+    },
+    {
+      "epoch": 0.4511255010792476,
+      "grad_norm": 3.617902994155884,
+      "learning_rate": 6.027587930142866e-06,
+      "loss": 0.9552,
+      "step": 5852
+    },
+    {
+      "epoch": 0.45120259019426456,
+      "grad_norm": 3.4752037525177,
+      "learning_rate": 6.026366100951304e-06,
+      "loss": 0.9546,
+      "step": 5853
+    },
+    {
+      "epoch": 0.45127967930928153,
+      "grad_norm": 3.3312973976135254,
+      "learning_rate": 6.025144207771132e-06,
+      "loss": 0.9142,
+      "step": 5854
+    },
+    {
+      "epoch": 0.4513567684242985,
+      "grad_norm": 4.001164436340332,
+      "learning_rate": 6.0239222506785285e-06,
+      "loss": 1.0792,
+      "step": 5855
+    },
+    {
+      "epoch": 0.4514338575393155,
+      "grad_norm": 3.3809666633605957,
+      "learning_rate": 6.0227002297496765e-06,
+      "loss": 0.9077,
+      "step": 5856
+    },
+    {
+      "epoch": 0.4515109466543324,
+      "grad_norm": 3.638946771621704,
+      "learning_rate": 6.02147814506076e-06,
+      "loss": 0.9814,
+      "step": 5857
+    },
+    {
+      "epoch": 0.45158803576934936,
+      "grad_norm": 3.711111545562744,
+      "learning_rate": 6.020255996687973e-06,
+      "loss": 0.9074,
+      "step": 5858
+    },
+    {
+      "epoch": 0.45166512488436633,
+      "grad_norm": 3.797400712966919,
+      "learning_rate": 6.019033784707507e-06,
+      "loss": 1.0543,
+      "step": 5859
+    },
+    {
+      "epoch": 0.4517422139993833,
+      "grad_norm": 3.4948596954345703,
+      "learning_rate": 6.017811509195565e-06,
+      "loss": 0.9662,
+      "step": 5860
+    },
+    {
+      "epoch": 0.4518193031144003,
+      "grad_norm": 3.3635761737823486,
+      "learning_rate": 6.0165891702283444e-06,
+      "loss": 0.8524,
+      "step": 5861
+    },
+    {
+      "epoch": 0.4518963922294172,
+      "grad_norm": 3.4985485076904297,
+      "learning_rate": 6.015366767882054e-06,
+      "loss": 0.9585,
+      "step": 5862
+    },
+    {
+      "epoch": 0.45197348134443416,
+      "grad_norm": 3.706798791885376,
+      "learning_rate": 6.014144302232906e-06,
+      "loss": 1.0279,
+      "step": 5863
+    },
+    {
+      "epoch": 0.45205057045945113,
+      "grad_norm": 3.334749460220337,
+      "learning_rate": 6.012921773357112e-06,
+      "loss": 1.0438,
+      "step": 5864
+    },
+    {
+      "epoch": 0.4521276595744681,
+      "grad_norm": 4.0636091232299805,
+      "learning_rate": 6.011699181330891e-06,
+      "loss": 1.0819,
+      "step": 5865
+    },
+    {
+      "epoch": 0.45220474868948507,
+      "grad_norm": 3.427274703979492,
+      "learning_rate": 6.0104765262304676e-06,
+      "loss": 0.9706,
+      "step": 5866
+    },
+    {
+      "epoch": 0.452281837804502,
+      "grad_norm": 3.34578800201416,
+      "learning_rate": 6.009253808132064e-06,
+      "loss": 0.8081,
+      "step": 5867
+    },
+    {
+      "epoch": 0.45235892691951896,
+      "grad_norm": 4.062625408172607,
+      "learning_rate": 6.008031027111913e-06,
+      "loss": 1.0353,
+      "step": 5868
+    },
+    {
+      "epoch": 0.4524360160345359,
+      "grad_norm": 3.454693078994751,
+      "learning_rate": 6.00680818324625e-06,
+      "loss": 0.9642,
+      "step": 5869
+    },
+    {
+      "epoch": 0.4525131051495529,
+      "grad_norm": 3.7603514194488525,
+      "learning_rate": 6.00558527661131e-06,
+      "loss": 0.9058,
+      "step": 5870
+    },
+    {
+      "epoch": 0.45259019426456987,
+      "grad_norm": 3.822753667831421,
+      "learning_rate": 6.004362307283335e-06,
+      "loss": 1.0457,
+      "step": 5871
+    },
+    {
+      "epoch": 0.4526672833795868,
+      "grad_norm": 3.4553167819976807,
+      "learning_rate": 6.003139275338573e-06,
+      "loss": 0.9248,
+      "step": 5872
+    },
+    {
+      "epoch": 0.45274437249460375,
+      "grad_norm": 3.557408332824707,
+      "learning_rate": 6.001916180853271e-06,
+      "loss": 1.0115,
+      "step": 5873
+    },
+    {
+      "epoch": 0.4528214616096207,
+      "grad_norm": 3.668583393096924,
+      "learning_rate": 6.0006930239036865e-06,
+      "loss": 1.0528,
+      "step": 5874
+    },
+    {
+      "epoch": 0.4528985507246377,
+      "grad_norm": 3.564915418624878,
+      "learning_rate": 5.999469804566074e-06,
+      "loss": 0.986,
+      "step": 5875
+    },
+    {
+      "epoch": 0.45297563983965466,
+      "grad_norm": 3.27848744392395,
+      "learning_rate": 5.998246522916695e-06,
+      "loss": 0.8666,
+      "step": 5876
+    },
+    {
+      "epoch": 0.4530527289546716,
+      "grad_norm": 3.95578932762146,
+      "learning_rate": 5.997023179031815e-06,
+      "loss": 1.0293,
+      "step": 5877
+    },
+    {
+      "epoch": 0.45312981806968855,
+      "grad_norm": 3.2330212593078613,
+      "learning_rate": 5.995799772987705e-06,
+      "loss": 0.8926,
+      "step": 5878
+    },
+    {
+      "epoch": 0.4532069071847055,
+      "grad_norm": 3.7614970207214355,
+      "learning_rate": 5.994576304860636e-06,
+      "loss": 0.9709,
+      "step": 5879
+    },
+    {
+      "epoch": 0.4532839962997225,
+      "grad_norm": 3.9075229167938232,
+      "learning_rate": 5.993352774726885e-06,
+      "loss": 0.9729,
+      "step": 5880
+    },
+    {
+      "epoch": 0.45336108541473946,
+      "grad_norm": 3.5691709518432617,
+      "learning_rate": 5.992129182662733e-06,
+      "loss": 0.9145,
+      "step": 5881
+    },
+    {
+      "epoch": 0.4534381745297564,
+      "grad_norm": 3.983722686767578,
+      "learning_rate": 5.990905528744466e-06,
+      "loss": 0.9745,
+      "step": 5882
+    },
+    {
+      "epoch": 0.45351526364477335,
+      "grad_norm": 3.6609230041503906,
+      "learning_rate": 5.98968181304837e-06,
+      "loss": 0.9585,
+      "step": 5883
+    },
+    {
+      "epoch": 0.4535923527597903,
+      "grad_norm": 3.4482264518737793,
+      "learning_rate": 5.98845803565074e-06,
+      "loss": 1.0082,
+      "step": 5884
+    },
+    {
+      "epoch": 0.4536694418748073,
+      "grad_norm": 3.4838855266571045,
+      "learning_rate": 5.987234196627869e-06,
+      "loss": 0.8997,
+      "step": 5885
+    },
+    {
+      "epoch": 0.45374653098982426,
+      "grad_norm": 3.4517669677734375,
+      "learning_rate": 5.986010296056059e-06,
+      "loss": 1.0314,
+      "step": 5886
+    },
+    {
+      "epoch": 0.4538236201048412,
+      "grad_norm": 3.517615556716919,
+      "learning_rate": 5.984786334011617e-06,
+      "loss": 0.9729,
+      "step": 5887
+    },
+    {
+      "epoch": 0.45390070921985815,
+      "grad_norm": 3.7334585189819336,
+      "learning_rate": 5.983562310570844e-06,
+      "loss": 0.9424,
+      "step": 5888
+    },
+    {
+      "epoch": 0.4539777983348751,
+      "grad_norm": 3.1004371643066406,
+      "learning_rate": 5.982338225810056e-06,
+      "loss": 0.8848,
+      "step": 5889
+    },
+    {
+      "epoch": 0.4540548874498921,
+      "grad_norm": 3.9406261444091797,
+      "learning_rate": 5.9811140798055674e-06,
+      "loss": 1.0372,
+      "step": 5890
+    },
+    {
+      "epoch": 0.45413197656490906,
+      "grad_norm": 3.5485575199127197,
+      "learning_rate": 5.9798898726336965e-06,
+      "loss": 0.8899,
+      "step": 5891
+    },
+    {
+      "epoch": 0.45420906567992597,
+      "grad_norm": 3.7856481075286865,
+      "learning_rate": 5.978665604370767e-06,
+      "loss": 0.86,
+      "step": 5892
+    },
+    {
+      "epoch": 0.45428615479494294,
+      "grad_norm": 3.8746819496154785,
+      "learning_rate": 5.977441275093108e-06,
+      "loss": 0.8669,
+      "step": 5893
+    },
+    {
+      "epoch": 0.4543632439099599,
+      "grad_norm": 3.443347215652466,
+      "learning_rate": 5.976216884877045e-06,
+      "loss": 0.8895,
+      "step": 5894
+    },
+    {
+      "epoch": 0.4544403330249769,
+      "grad_norm": 3.6835126876831055,
+      "learning_rate": 5.974992433798916e-06,
+      "loss": 0.9634,
+      "step": 5895
+    },
+    {
+      "epoch": 0.45451742213999385,
+      "grad_norm": 4.053088665008545,
+      "learning_rate": 5.973767921935059e-06,
+      "loss": 0.933,
+      "step": 5896
+    },
+    {
+      "epoch": 0.45459451125501077,
+      "grad_norm": 3.9018402099609375,
+      "learning_rate": 5.972543349361813e-06,
+      "loss": 0.8993,
+      "step": 5897
+    },
+    {
+      "epoch": 0.45467160037002774,
+      "grad_norm": 3.5930070877075195,
+      "learning_rate": 5.97131871615553e-06,
+      "loss": 1.0334,
+      "step": 5898
+    },
+    {
+      "epoch": 0.4547486894850447,
+      "grad_norm": 3.705622434616089,
+      "learning_rate": 5.970094022392553e-06,
+      "loss": 0.8647,
+      "step": 5899
+    },
+    {
+      "epoch": 0.4548257786000617,
+      "grad_norm": 3.460371971130371,
+      "learning_rate": 5.968869268149239e-06,
+      "loss": 0.8912,
+      "step": 5900
+    },
+    {
+      "epoch": 0.45490286771507865,
+      "grad_norm": 3.655982255935669,
+      "learning_rate": 5.967644453501944e-06,
+      "loss": 1.004,
+      "step": 5901
+    },
+    {
+      "epoch": 0.45497995683009557,
+      "grad_norm": 3.730700969696045,
+      "learning_rate": 5.966419578527027e-06,
+      "loss": 0.9948,
+      "step": 5902
+    },
+    {
+      "epoch": 0.45505704594511254,
+      "grad_norm": 3.825910806655884,
+      "learning_rate": 5.965194643300858e-06,
+      "loss": 1.0192,
+      "step": 5903
+    },
+    {
+      "epoch": 0.4551341350601295,
+      "grad_norm": 3.2197327613830566,
+      "learning_rate": 5.9639696478997985e-06,
+      "loss": 0.8807,
+      "step": 5904
+    },
+    {
+      "epoch": 0.4552112241751465,
+      "grad_norm": 3.4035542011260986,
+      "learning_rate": 5.962744592400226e-06,
+      "loss": 0.8079,
+      "step": 5905
+    },
+    {
+      "epoch": 0.45528831329016345,
+      "grad_norm": 3.4863269329071045,
+      "learning_rate": 5.961519476878513e-06,
+      "loss": 0.8259,
+      "step": 5906
+    },
+    {
+      "epoch": 0.45536540240518036,
+      "grad_norm": 3.9149959087371826,
+      "learning_rate": 5.960294301411041e-06,
+      "loss": 0.9398,
+      "step": 5907
+    },
+    {
+      "epoch": 0.45544249152019733,
+      "grad_norm": 3.782867670059204,
+      "learning_rate": 5.959069066074195e-06,
+      "loss": 1.1127,
+      "step": 5908
+    },
+    {
+      "epoch": 0.4555195806352143,
+      "grad_norm": 3.4562859535217285,
+      "learning_rate": 5.957843770944357e-06,
+      "loss": 0.9791,
+      "step": 5909
+    },
+    {
+      "epoch": 0.4555966697502313,
+      "grad_norm": 3.538780689239502,
+      "learning_rate": 5.956618416097921e-06,
+      "loss": 0.965,
+      "step": 5910
+    },
+    {
+      "epoch": 0.45567375886524825,
+      "grad_norm": 3.6963050365448,
+      "learning_rate": 5.955393001611283e-06,
+      "loss": 1.0603,
+      "step": 5911
+    },
+    {
+      "epoch": 0.45575084798026516,
+      "grad_norm": 3.4785306453704834,
+      "learning_rate": 5.954167527560837e-06,
+      "loss": 0.9417,
+      "step": 5912
+    },
+    {
+      "epoch": 0.45582793709528213,
+      "grad_norm": 3.757713556289673,
+      "learning_rate": 5.952941994022988e-06,
+      "loss": 1.0384,
+      "step": 5913
+    },
+    {
+      "epoch": 0.4559050262102991,
+      "grad_norm": 3.715573310852051,
+      "learning_rate": 5.951716401074143e-06,
+      "loss": 0.9363,
+      "step": 5914
+    },
+    {
+      "epoch": 0.4559821153253161,
+      "grad_norm": 3.4896347522735596,
+      "learning_rate": 5.9504907487907086e-06,
+      "loss": 1.0017,
+      "step": 5915
+    },
+    {
+      "epoch": 0.45605920444033304,
+      "grad_norm": 3.7236409187316895,
+      "learning_rate": 5.949265037249096e-06,
+      "loss": 1.0504,
+      "step": 5916
+    },
+    {
+      "epoch": 0.45613629355534996,
+      "grad_norm": 3.842233657836914,
+      "learning_rate": 5.948039266525728e-06,
+      "loss": 1.0006,
+      "step": 5917
+    },
+    {
+      "epoch": 0.45621338267036693,
+      "grad_norm": 3.868191719055176,
+      "learning_rate": 5.946813436697021e-06,
+      "loss": 1.0097,
+      "step": 5918
+    },
+    {
+      "epoch": 0.4562904717853839,
+      "grad_norm": 3.6733784675598145,
+      "learning_rate": 5.9455875478394e-06,
+      "loss": 1.0119,
+      "step": 5919
+    },
+    {
+      "epoch": 0.45636756090040087,
+      "grad_norm": 3.975710153579712,
+      "learning_rate": 5.944361600029291e-06,
+      "loss": 1.0206,
+      "step": 5920
+    },
+    {
+      "epoch": 0.45644465001541784,
+      "grad_norm": 3.2998998165130615,
+      "learning_rate": 5.9431355933431285e-06,
+      "loss": 0.8406,
+      "step": 5921
+    },
+    {
+      "epoch": 0.45652173913043476,
+      "grad_norm": 3.8829758167266846,
+      "learning_rate": 5.941909527857348e-06,
+      "loss": 0.9318,
+      "step": 5922
+    },
+    {
+      "epoch": 0.4565988282454517,
+      "grad_norm": 3.46357798576355,
+      "learning_rate": 5.940683403648384e-06,
+      "loss": 0.927,
+      "step": 5923
+    },
+    {
+      "epoch": 0.4566759173604687,
+      "grad_norm": 3.7031266689300537,
+      "learning_rate": 5.939457220792684e-06,
+      "loss": 0.9815,
+      "step": 5924
+    },
+    {
+      "epoch": 0.45675300647548567,
+      "grad_norm": 3.6570425033569336,
+      "learning_rate": 5.938230979366691e-06,
+      "loss": 1.0229,
+      "step": 5925
+    },
+    {
+      "epoch": 0.45683009559050264,
+      "grad_norm": 3.5698962211608887,
+      "learning_rate": 5.937004679446854e-06,
+      "loss": 0.9961,
+      "step": 5926
+    },
+    {
+      "epoch": 0.45690718470551955,
+      "grad_norm": 3.748589515686035,
+      "learning_rate": 5.935778321109631e-06,
+      "loss": 0.9987,
+      "step": 5927
+    },
+    {
+      "epoch": 0.4569842738205365,
+      "grad_norm": 3.5768892765045166,
+      "learning_rate": 5.934551904431473e-06,
+      "loss": 0.9889,
+      "step": 5928
+    },
+    {
+      "epoch": 0.4570613629355535,
+      "grad_norm": 3.2622320652008057,
+      "learning_rate": 5.933325429488847e-06,
+      "loss": 0.9082,
+      "step": 5929
+    },
+    {
+      "epoch": 0.45713845205057047,
+      "grad_norm": 3.7464730739593506,
+      "learning_rate": 5.9320988963582125e-06,
+      "loss": 1.0409,
+      "step": 5930
+    },
+    {
+      "epoch": 0.45721554116558744,
+      "grad_norm": 3.6139042377471924,
+      "learning_rate": 5.93087230511604e-06,
+      "loss": 1.0702,
+      "step": 5931
+    },
+    {
+      "epoch": 0.4572926302806044,
+      "grad_norm": 3.8561818599700928,
+      "learning_rate": 5.929645655838801e-06,
+      "loss": 1.1142,
+      "step": 5932
+    },
+    {
+      "epoch": 0.4573697193956213,
+      "grad_norm": 3.627636194229126,
+      "learning_rate": 5.9284189486029684e-06,
+      "loss": 0.9967,
+      "step": 5933
+    },
+    {
+      "epoch": 0.4574468085106383,
+      "grad_norm": 4.026650428771973,
+      "learning_rate": 5.927192183485023e-06,
+      "loss": 1.0336,
+      "step": 5934
+    },
+    {
+      "epoch": 0.45752389762565526,
+      "grad_norm": 3.624974012374878,
+      "learning_rate": 5.925965360561448e-06,
+      "loss": 0.9791,
+      "step": 5935
+    },
+    {
+      "epoch": 0.45760098674067223,
+      "grad_norm": 3.952138662338257,
+      "learning_rate": 5.924738479908728e-06,
+      "loss": 0.9589,
+      "step": 5936
+    },
+    {
+      "epoch": 0.4576780758556892,
+      "grad_norm": 4.19709587097168,
+      "learning_rate": 5.923511541603353e-06,
+      "loss": 1.0059,
+      "step": 5937
+    },
+    {
+      "epoch": 0.4577551649707061,
+      "grad_norm": 4.1676530838012695,
+      "learning_rate": 5.922284545721817e-06,
+      "loss": 1.0851,
+      "step": 5938
+    },
+    {
+      "epoch": 0.4578322540857231,
+      "grad_norm": 3.9117021560668945,
+      "learning_rate": 5.921057492340614e-06,
+      "loss": 1.0179,
+      "step": 5939
+    },
+    {
+      "epoch": 0.45790934320074006,
+      "grad_norm": 3.553438186645508,
+      "learning_rate": 5.919830381536249e-06,
+      "loss": 0.9542,
+      "step": 5940
+    },
+    {
+      "epoch": 0.45798643231575703,
+      "grad_norm": 3.3051836490631104,
+      "learning_rate": 5.918603213385223e-06,
+      "loss": 0.9783,
+      "step": 5941
+    },
+    {
+      "epoch": 0.458063521430774,
+      "grad_norm": 4.240174770355225,
+      "learning_rate": 5.917375987964044e-06,
+      "loss": 1.0063,
+      "step": 5942
+    },
+    {
+      "epoch": 0.4581406105457909,
+      "grad_norm": 3.5065135955810547,
+      "learning_rate": 5.916148705349224e-06,
+      "loss": 0.8109,
+      "step": 5943
+    },
+    {
+      "epoch": 0.4582176996608079,
+      "grad_norm": 3.668158531188965,
+      "learning_rate": 5.914921365617276e-06,
+      "loss": 0.879,
+      "step": 5944
+    },
+    {
+      "epoch": 0.45829478877582486,
+      "grad_norm": 3.529534339904785,
+      "learning_rate": 5.9136939688447205e-06,
+      "loss": 0.9102,
+      "step": 5945
+    },
+    {
+      "epoch": 0.45837187789084183,
+      "grad_norm": 3.5600032806396484,
+      "learning_rate": 5.912466515108078e-06,
+      "loss": 0.9238,
+      "step": 5946
+    },
+    {
+      "epoch": 0.4584489670058588,
+      "grad_norm": 3.8036692142486572,
+      "learning_rate": 5.911239004483874e-06,
+      "loss": 1.1042,
+      "step": 5947
+    },
+    {
+      "epoch": 0.4585260561208757,
+      "grad_norm": 3.5569405555725098,
+      "learning_rate": 5.9100114370486375e-06,
+      "loss": 0.9662,
+      "step": 5948
+    },
+    {
+      "epoch": 0.4586031452358927,
+      "grad_norm": 3.4714467525482178,
+      "learning_rate": 5.9087838128789e-06,
+      "loss": 0.9444,
+      "step": 5949
+    },
+    {
+      "epoch": 0.45868023435090965,
+      "grad_norm": 3.8040552139282227,
+      "learning_rate": 5.9075561320511994e-06,
+      "loss": 0.9892,
+      "step": 5950
+    },
+    {
+      "epoch": 0.4587573234659266,
+      "grad_norm": 3.3058183193206787,
+      "learning_rate": 5.906328394642075e-06,
+      "loss": 1.0018,
+      "step": 5951
+    },
+    {
+      "epoch": 0.4588344125809436,
+      "grad_norm": 3.314758539199829,
+      "learning_rate": 5.905100600728067e-06,
+      "loss": 0.831,
+      "step": 5952
+    },
+    {
+      "epoch": 0.4589115016959605,
+      "grad_norm": 3.5863702297210693,
+      "learning_rate": 5.903872750385726e-06,
+      "loss": 0.9611,
+      "step": 5953
+    },
+    {
+      "epoch": 0.4589885908109775,
+      "grad_norm": 3.328200101852417,
+      "learning_rate": 5.902644843691601e-06,
+      "loss": 0.9919,
+      "step": 5954
+    },
+    {
+      "epoch": 0.45906567992599445,
+      "grad_norm": 3.5870213508605957,
+      "learning_rate": 5.901416880722242e-06,
+      "loss": 0.9036,
+      "step": 5955
+    },
+    {
+      "epoch": 0.4591427690410114,
+      "grad_norm": 3.978027582168579,
+      "learning_rate": 5.900188861554213e-06,
+      "loss": 0.9305,
+      "step": 5956
+    },
+    {
+      "epoch": 0.4592198581560284,
+      "grad_norm": 3.5891008377075195,
+      "learning_rate": 5.898960786264067e-06,
+      "loss": 1.0179,
+      "step": 5957
+    },
+    {
+      "epoch": 0.4592969472710453,
+      "grad_norm": 3.536594867706299,
+      "learning_rate": 5.897732654928373e-06,
+      "loss": 0.9776,
+      "step": 5958
+    },
+    {
+      "epoch": 0.4593740363860623,
+      "grad_norm": 3.7737505435943604,
+      "learning_rate": 5.896504467623698e-06,
+      "loss": 1.121,
+      "step": 5959
+    },
+    {
+      "epoch": 0.45945112550107925,
+      "grad_norm": 3.9404971599578857,
+      "learning_rate": 5.895276224426613e-06,
+      "loss": 1.0788,
+      "step": 5960
+    },
+    {
+      "epoch": 0.4595282146160962,
+      "grad_norm": 3.931784152984619,
+      "learning_rate": 5.894047925413691e-06,
+      "loss": 0.988,
+      "step": 5961
+    },
+    {
+      "epoch": 0.4596053037311132,
+      "grad_norm": 3.751889228820801,
+      "learning_rate": 5.892819570661511e-06,
+      "loss": 0.9438,
+      "step": 5962
+    },
+    {
+      "epoch": 0.4596823928461301,
+      "grad_norm": 4.397334098815918,
+      "learning_rate": 5.891591160246655e-06,
+      "loss": 0.9381,
+      "step": 5963
+    },
+    {
+      "epoch": 0.4597594819611471,
+      "grad_norm": 3.8391776084899902,
+      "learning_rate": 5.890362694245709e-06,
+      "loss": 0.8604,
+      "step": 5964
+    },
+    {
+      "epoch": 0.45983657107616405,
+      "grad_norm": 3.5146422386169434,
+      "learning_rate": 5.889134172735259e-06,
+      "loss": 1.0192,
+      "step": 5965
+    },
+    {
+      "epoch": 0.459913660191181,
+      "grad_norm": 3.9917097091674805,
+      "learning_rate": 5.887905595791899e-06,
+      "loss": 1.0062,
+      "step": 5966
+    },
+    {
+      "epoch": 0.459990749306198,
+      "grad_norm": 3.921919345855713,
+      "learning_rate": 5.886676963492224e-06,
+      "loss": 0.9134,
+      "step": 5967
+    },
+    {
+      "epoch": 0.4600678384212149,
+      "grad_norm": 3.3720543384552,
+      "learning_rate": 5.885448275912832e-06,
+      "loss": 0.8687,
+      "step": 5968
+    },
+    {
+      "epoch": 0.4601449275362319,
+      "grad_norm": 3.6408445835113525,
+      "learning_rate": 5.884219533130325e-06,
+      "loss": 1.0733,
+      "step": 5969
+    },
+    {
+      "epoch": 0.46022201665124884,
+      "grad_norm": 3.81526255607605,
+      "learning_rate": 5.882990735221312e-06,
+      "loss": 0.9489,
+      "step": 5970
+    },
+    {
+      "epoch": 0.4602991057662658,
+      "grad_norm": 4.006250858306885,
+      "learning_rate": 5.881761882262398e-06,
+      "loss": 0.9353,
+      "step": 5971
+    },
+    {
+      "epoch": 0.4603761948812828,
+      "grad_norm": 3.7868504524230957,
+      "learning_rate": 5.880532974330197e-06,
+      "loss": 1.0833,
+      "step": 5972
+    },
+    {
+      "epoch": 0.4604532839962997,
+      "grad_norm": 3.490978956222534,
+      "learning_rate": 5.879304011501327e-06,
+      "loss": 1.0075,
+      "step": 5973
+    },
+    {
+      "epoch": 0.46053037311131667,
+      "grad_norm": 3.4989683628082275,
+      "learning_rate": 5.878074993852405e-06,
+      "loss": 1.0013,
+      "step": 5974
+    },
+    {
+      "epoch": 0.46060746222633364,
+      "grad_norm": 3.448817729949951,
+      "learning_rate": 5.876845921460055e-06,
+      "loss": 0.8911,
+      "step": 5975
+    },
+    {
+      "epoch": 0.4606845513413506,
+      "grad_norm": 3.613193988800049,
+      "learning_rate": 5.875616794400902e-06,
+      "loss": 0.9418,
+      "step": 5976
+    },
+    {
+      "epoch": 0.4607616404563676,
+      "grad_norm": 3.5848329067230225,
+      "learning_rate": 5.874387612751579e-06,
+      "loss": 1.0202,
+      "step": 5977
+    },
+    {
+      "epoch": 0.4608387295713845,
+      "grad_norm": 3.828852415084839,
+      "learning_rate": 5.8731583765887156e-06,
+      "loss": 1.0841,
+      "step": 5978
+    },
+    {
+      "epoch": 0.46091581868640147,
+      "grad_norm": 3.3265674114227295,
+      "learning_rate": 5.87192908598895e-06,
+      "loss": 0.8589,
+      "step": 5979
+    },
+    {
+      "epoch": 0.46099290780141844,
+      "grad_norm": 3.5973331928253174,
+      "learning_rate": 5.870699741028922e-06,
+      "loss": 0.9637,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4610699969164354,
+      "grad_norm": 3.2978997230529785,
+      "learning_rate": 5.869470341785274e-06,
+      "loss": 0.9569,
+      "step": 5981
+    },
+    {
+      "epoch": 0.4611470860314524,
+      "grad_norm": 3.7744216918945312,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.9753,
+      "step": 5982
+    },
+    {
+      "epoch": 0.4612241751464693,
+      "grad_norm": 3.4844672679901123,
+      "learning_rate": 5.8670113807537095e-06,
+      "loss": 0.9662,
+      "step": 5983
+    },
+    {
+      "epoch": 0.46130126426148627,
+      "grad_norm": 3.5159308910369873,
+      "learning_rate": 5.865781819119096e-06,
+      "loss": 0.9549,
+      "step": 5984
+    },
+    {
+      "epoch": 0.46137835337650324,
+      "grad_norm": 3.471386194229126,
+      "learning_rate": 5.864552203507472e-06,
+      "loss": 0.9481,
+      "step": 5985
+    },
+    {
+      "epoch": 0.4614554424915202,
+      "grad_norm": 3.8364601135253906,
+      "learning_rate": 5.863322533995495e-06,
+      "loss": 0.9589,
+      "step": 5986
+    },
+    {
+      "epoch": 0.4615325316065372,
+      "grad_norm": 3.745004415512085,
+      "learning_rate": 5.862092810659829e-06,
+      "loss": 0.9863,
+      "step": 5987
+    },
+    {
+      "epoch": 0.4616096207215541,
+      "grad_norm": 3.7336912155151367,
+      "learning_rate": 5.860863033577141e-06,
+      "loss": 1.0487,
+      "step": 5988
+    },
+    {
+      "epoch": 0.46168670983657106,
+      "grad_norm": 3.603816509246826,
+      "learning_rate": 5.859633202824101e-06,
+      "loss": 1.0228,
+      "step": 5989
+    },
+    {
+      "epoch": 0.46176379895158803,
+      "grad_norm": 3.479917049407959,
+      "learning_rate": 5.858403318477384e-06,
+      "loss": 0.9293,
+      "step": 5990
+    },
+    {
+      "epoch": 0.461840888066605,
+      "grad_norm": 3.8286755084991455,
+      "learning_rate": 5.857173380613665e-06,
+      "loss": 1.0632,
+      "step": 5991
+    },
+    {
+      "epoch": 0.461917977181622,
+      "grad_norm": 3.5856165885925293,
+      "learning_rate": 5.855943389309626e-06,
+      "loss": 1.0216,
+      "step": 5992
+    },
+    {
+      "epoch": 0.4619950662966389,
+      "grad_norm": 4.016223907470703,
+      "learning_rate": 5.8547133446419495e-06,
+      "loss": 0.9528,
+      "step": 5993
+    },
+    {
+      "epoch": 0.46207215541165586,
+      "grad_norm": 3.4276955127716064,
+      "learning_rate": 5.853483246687323e-06,
+      "loss": 1.0603,
+      "step": 5994
+    },
+    {
+      "epoch": 0.46214924452667283,
+      "grad_norm": 3.522810459136963,
+      "learning_rate": 5.852253095522435e-06,
+      "loss": 0.9513,
+      "step": 5995
+    },
+    {
+      "epoch": 0.4622263336416898,
+      "grad_norm": 3.607531785964966,
+      "learning_rate": 5.851022891223982e-06,
+      "loss": 0.9798,
+      "step": 5996
+    },
+    {
+      "epoch": 0.46230342275670677,
+      "grad_norm": 3.3895387649536133,
+      "learning_rate": 5.849792633868659e-06,
+      "loss": 0.9795,
+      "step": 5997
+    },
+    {
+      "epoch": 0.4623805118717237,
+      "grad_norm": 4.266413688659668,
+      "learning_rate": 5.848562323533165e-06,
+      "loss": 0.9955,
+      "step": 5998
+    },
+    {
+      "epoch": 0.46245760098674066,
+      "grad_norm": 3.460974931716919,
+      "learning_rate": 5.8473319602942065e-06,
+      "loss": 0.9737,
+      "step": 5999
+    },
+    {
+      "epoch": 0.46253469010175763,
+      "grad_norm": 3.6543521881103516,
+      "learning_rate": 5.846101544228488e-06,
+      "loss": 0.898,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4626117792167746,
+      "grad_norm": 3.48816180229187,
+      "learning_rate": 5.84487107541272e-06,
+      "loss": 0.9795,
+      "step": 6001
+    },
+    {
+      "epoch": 0.46268886833179157,
+      "grad_norm": 3.694929838180542,
+      "learning_rate": 5.843640553923618e-06,
+      "loss": 0.9415,
+      "step": 6002
+    },
+    {
+      "epoch": 0.4627659574468085,
+      "grad_norm": 3.641181230545044,
+      "learning_rate": 5.842409979837894e-06,
+      "loss": 0.9422,
+      "step": 6003
+    },
+    {
+      "epoch": 0.46284304656182546,
+      "grad_norm": 5.558597087860107,
+      "learning_rate": 5.841179353232273e-06,
+      "loss": 0.9911,
+      "step": 6004
+    },
+    {
+      "epoch": 0.4629201356768424,
+      "grad_norm": 3.562645196914673,
+      "learning_rate": 5.839948674183476e-06,
+      "loss": 1.0034,
+      "step": 6005
+    },
+    {
+      "epoch": 0.4629972247918594,
+      "grad_norm": 3.8084611892700195,
+      "learning_rate": 5.8387179427682265e-06,
+      "loss": 1.0519,
+      "step": 6006
+    },
+    {
+      "epoch": 0.46307431390687637,
+      "grad_norm": 3.4862160682678223,
+      "learning_rate": 5.837487159063259e-06,
+      "loss": 0.9177,
+      "step": 6007
+    },
+    {
+      "epoch": 0.4631514030218933,
+      "grad_norm": 3.7302255630493164,
+      "learning_rate": 5.836256323145304e-06,
+      "loss": 0.9929,
+      "step": 6008
+    },
+    {
+      "epoch": 0.46322849213691025,
+      "grad_norm": 4.211978912353516,
+      "learning_rate": 5.835025435091099e-06,
+      "loss": 1.0944,
+      "step": 6009
+    },
+    {
+      "epoch": 0.4633055812519272,
+      "grad_norm": 3.287256956100464,
+      "learning_rate": 5.8337944949773825e-06,
+      "loss": 0.9778,
+      "step": 6010
+    },
+    {
+      "epoch": 0.4633826703669442,
+      "grad_norm": 3.5103518962860107,
+      "learning_rate": 5.832563502880898e-06,
+      "loss": 0.9329,
+      "step": 6011
+    },
+    {
+      "epoch": 0.46345975948196116,
+      "grad_norm": 3.8197262287139893,
+      "learning_rate": 5.831332458878391e-06,
+      "loss": 1.0127,
+      "step": 6012
+    },
+    {
+      "epoch": 0.4635368485969781,
+      "grad_norm": 3.5835108757019043,
+      "learning_rate": 5.830101363046611e-06,
+      "loss": 1.0235,
+      "step": 6013
+    },
+    {
+      "epoch": 0.46361393771199505,
+      "grad_norm": 3.5647315979003906,
+      "learning_rate": 5.82887021546231e-06,
+      "loss": 0.9782,
+      "step": 6014
+    },
+    {
+      "epoch": 0.463691026827012,
+      "grad_norm": 3.430677652359009,
+      "learning_rate": 5.827639016202244e-06,
+      "loss": 1.0517,
+      "step": 6015
+    },
+    {
+      "epoch": 0.463768115942029,
+      "grad_norm": 4.109676837921143,
+      "learning_rate": 5.826407765343172e-06,
+      "loss": 0.8909,
+      "step": 6016
+    },
+    {
+      "epoch": 0.46384520505704596,
+      "grad_norm": 5.432858943939209,
+      "learning_rate": 5.825176462961854e-06,
+      "loss": 0.9727,
+      "step": 6017
+    },
+    {
+      "epoch": 0.46392229417206293,
+      "grad_norm": 3.428158760070801,
+      "learning_rate": 5.82394510913506e-06,
+      "loss": 0.9286,
+      "step": 6018
+    },
+    {
+      "epoch": 0.46399938328707985,
+      "grad_norm": 3.5705456733703613,
+      "learning_rate": 5.822713703939554e-06,
+      "loss": 0.9397,
+      "step": 6019
+    },
+    {
+      "epoch": 0.4640764724020968,
+      "grad_norm": 3.492794990539551,
+      "learning_rate": 5.82148224745211e-06,
+      "loss": 0.943,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4641535615171138,
+      "grad_norm": 3.4019174575805664,
+      "learning_rate": 5.820250739749502e-06,
+      "loss": 0.851,
+      "step": 6021
+    },
+    {
+      "epoch": 0.46423065063213076,
+      "grad_norm": 3.375807046890259,
+      "learning_rate": 5.819019180908509e-06,
+      "loss": 0.9877,
+      "step": 6022
+    },
+    {
+      "epoch": 0.46430773974714773,
+      "grad_norm": 3.3332841396331787,
+      "learning_rate": 5.817787571005913e-06,
+      "loss": 0.9421,
+      "step": 6023
+    },
+    {
+      "epoch": 0.46438482886216464,
+      "grad_norm": 3.4136815071105957,
+      "learning_rate": 5.8165559101184955e-06,
+      "loss": 1.0011,
+      "step": 6024
+    },
+    {
+      "epoch": 0.4644619179771816,
+      "grad_norm": 3.7559056282043457,
+      "learning_rate": 5.8153241983230464e-06,
+      "loss": 1.0067,
+      "step": 6025
+    },
+    {
+      "epoch": 0.4645390070921986,
+      "grad_norm": 3.7920331954956055,
+      "learning_rate": 5.814092435696358e-06,
+      "loss": 0.9631,
+      "step": 6026
+    },
+    {
+      "epoch": 0.46461609620721556,
+      "grad_norm": 3.2758357524871826,
+      "learning_rate": 5.81286062231522e-06,
+      "loss": 0.857,
+      "step": 6027
+    },
+    {
+      "epoch": 0.4646931853222325,
+      "grad_norm": 3.8748836517333984,
+      "learning_rate": 5.811628758256433e-06,
+      "loss": 1.0907,
+      "step": 6028
+    },
+    {
+      "epoch": 0.46477027443724944,
+      "grad_norm": 3.435882568359375,
+      "learning_rate": 5.8103968435967965e-06,
+      "loss": 0.8907,
+      "step": 6029
+    },
+    {
+      "epoch": 0.4648473635522664,
+      "grad_norm": 3.716470956802368,
+      "learning_rate": 5.809164878413114e-06,
+      "loss": 0.9661,
+      "step": 6030
+    },
+    {
+      "epoch": 0.4649244526672834,
+      "grad_norm": 3.935591697692871,
+      "learning_rate": 5.807932862782193e-06,
+      "loss": 1.005,
+      "step": 6031
+    },
+    {
+      "epoch": 0.46500154178230035,
+      "grad_norm": 3.4351768493652344,
+      "learning_rate": 5.8067007967808405e-06,
+      "loss": 0.9601,
+      "step": 6032
+    },
+    {
+      "epoch": 0.4650786308973173,
+      "grad_norm": 3.390435218811035,
+      "learning_rate": 5.805468680485874e-06,
+      "loss": 0.9108,
+      "step": 6033
+    },
+    {
+      "epoch": 0.46515572001233424,
+      "grad_norm": 3.5145466327667236,
+      "learning_rate": 5.804236513974104e-06,
+      "loss": 0.9561,
+      "step": 6034
+    },
+    {
+      "epoch": 0.4652328091273512,
+      "grad_norm": 3.6625545024871826,
+      "learning_rate": 5.8030042973223545e-06,
+      "loss": 0.896,
+      "step": 6035
+    },
+    {
+      "epoch": 0.4653098982423682,
+      "grad_norm": 3.7747929096221924,
+      "learning_rate": 5.801772030607445e-06,
+      "loss": 0.9278,
+      "step": 6036
+    },
+    {
+      "epoch": 0.46538698735738515,
+      "grad_norm": 3.4798998832702637,
+      "learning_rate": 5.800539713906203e-06,
+      "loss": 1.0179,
+      "step": 6037
+    },
+    {
+      "epoch": 0.4654640764724021,
+      "grad_norm": 3.7864067554473877,
+      "learning_rate": 5.799307347295455e-06,
+      "loss": 0.9142,
+      "step": 6038
+    },
+    {
+      "epoch": 0.46554116558741904,
+      "grad_norm": 3.8589086532592773,
+      "learning_rate": 5.798074930852035e-06,
+      "loss": 0.8484,
+      "step": 6039
+    },
+    {
+      "epoch": 0.465618254702436,
+      "grad_norm": 3.9709889888763428,
+      "learning_rate": 5.796842464652774e-06,
+      "loss": 1.1772,
+      "step": 6040
+    },
+    {
+      "epoch": 0.465695343817453,
+      "grad_norm": 3.7802610397338867,
+      "learning_rate": 5.7956099487745135e-06,
+      "loss": 1.0716,
+      "step": 6041
+    },
+    {
+      "epoch": 0.46577243293246995,
+      "grad_norm": 3.4964866638183594,
+      "learning_rate": 5.794377383294094e-06,
+      "loss": 0.9444,
+      "step": 6042
+    },
+    {
+      "epoch": 0.4658495220474869,
+      "grad_norm": 3.519301176071167,
+      "learning_rate": 5.7931447682883565e-06,
+      "loss": 1.0331,
+      "step": 6043
+    },
+    {
+      "epoch": 0.46592661116250383,
+      "grad_norm": 3.5994677543640137,
+      "learning_rate": 5.791912103834154e-06,
+      "loss": 0.9118,
+      "step": 6044
+    },
+    {
+      "epoch": 0.4660037002775208,
+      "grad_norm": 3.669741630554199,
+      "learning_rate": 5.79067939000833e-06,
+      "loss": 0.9314,
+      "step": 6045
+    },
+    {
+      "epoch": 0.4660807893925378,
+      "grad_norm": 3.6019272804260254,
+      "learning_rate": 5.7894466268877426e-06,
+      "loss": 0.9013,
+      "step": 6046
+    },
+    {
+      "epoch": 0.46615787850755475,
+      "grad_norm": 4.028010845184326,
+      "learning_rate": 5.788213814549247e-06,
+      "loss": 0.9965,
+      "step": 6047
+    },
+    {
+      "epoch": 0.4662349676225717,
+      "grad_norm": 3.4716103076934814,
+      "learning_rate": 5.786980953069702e-06,
+      "loss": 0.8853,
+      "step": 6048
+    },
+    {
+      "epoch": 0.46631205673758863,
+      "grad_norm": 3.9504270553588867,
+      "learning_rate": 5.785748042525969e-06,
+      "loss": 1.1174,
+      "step": 6049
+    },
+    {
+      "epoch": 0.4663891458526056,
+      "grad_norm": 3.453526258468628,
+      "learning_rate": 5.784515082994917e-06,
+      "loss": 1.031,
+      "step": 6050
+    },
+    {
+      "epoch": 0.4664662349676226,
+      "grad_norm": 3.568610668182373,
+      "learning_rate": 5.783282074553412e-06,
+      "loss": 0.9396,
+      "step": 6051
+    },
+    {
+      "epoch": 0.46654332408263954,
+      "grad_norm": 3.7350780963897705,
+      "learning_rate": 5.782049017278326e-06,
+      "loss": 0.9443,
+      "step": 6052
+    },
+    {
+      "epoch": 0.4666204131976565,
+      "grad_norm": 4.244739532470703,
+      "learning_rate": 5.7808159112465344e-06,
+      "loss": 1.0614,
+      "step": 6053
+    },
+    {
+      "epoch": 0.46669750231267343,
+      "grad_norm": 3.688356637954712,
+      "learning_rate": 5.779582756534914e-06,
+      "loss": 0.8876,
+      "step": 6054
+    },
+    {
+      "epoch": 0.4667745914276904,
+      "grad_norm": 3.4765431880950928,
+      "learning_rate": 5.778349553220348e-06,
+      "loss": 0.9501,
+      "step": 6055
+    },
+    {
+      "epoch": 0.46685168054270737,
+      "grad_norm": 3.3060991764068604,
+      "learning_rate": 5.777116301379717e-06,
+      "loss": 0.9299,
+      "step": 6056
+    },
+    {
+      "epoch": 0.46692876965772434,
+      "grad_norm": 3.5781679153442383,
+      "learning_rate": 5.775883001089911e-06,
+      "loss": 0.9658,
+      "step": 6057
+    },
+    {
+      "epoch": 0.4670058587727413,
+      "grad_norm": 3.767371892929077,
+      "learning_rate": 5.7746496524278176e-06,
+      "loss": 1.0249,
+      "step": 6058
+    },
+    {
+      "epoch": 0.4670829478877582,
+      "grad_norm": 3.511653184890747,
+      "learning_rate": 5.7734162554703284e-06,
+      "loss": 0.8842,
+      "step": 6059
+    },
+    {
+      "epoch": 0.4671600370027752,
+      "grad_norm": 3.546081781387329,
+      "learning_rate": 5.7721828102943445e-06,
+      "loss": 1.0454,
+      "step": 6060
+    },
+    {
+      "epoch": 0.46723712611779217,
+      "grad_norm": 3.7766621112823486,
+      "learning_rate": 5.770949316976759e-06,
+      "loss": 0.9635,
+      "step": 6061
+    },
+    {
+      "epoch": 0.46731421523280914,
+      "grad_norm": 3.4135143756866455,
+      "learning_rate": 5.7697157755944775e-06,
+      "loss": 0.9405,
+      "step": 6062
+    },
+    {
+      "epoch": 0.4673913043478261,
+      "grad_norm": 3.601234197616577,
+      "learning_rate": 5.768482186224405e-06,
+      "loss": 0.9234,
+      "step": 6063
+    },
+    {
+      "epoch": 0.467468393462843,
+      "grad_norm": 3.9338316917419434,
+      "learning_rate": 5.7672485489434456e-06,
+      "loss": 1.032,
+      "step": 6064
+    },
+    {
+      "epoch": 0.46754548257786,
+      "grad_norm": 4.0997633934021,
+      "learning_rate": 5.766014863828515e-06,
+      "loss": 0.956,
+      "step": 6065
+    },
+    {
+      "epoch": 0.46762257169287696,
+      "grad_norm": 3.778200387954712,
+      "learning_rate": 5.764781130956525e-06,
+      "loss": 0.9533,
+      "step": 6066
+    },
+    {
+      "epoch": 0.46769966080789394,
+      "grad_norm": 3.8161983489990234,
+      "learning_rate": 5.763547350404391e-06,
+      "loss": 0.9875,
+      "step": 6067
+    },
+    {
+      "epoch": 0.4677767499229109,
+      "grad_norm": 3.7838432788848877,
+      "learning_rate": 5.762313522249036e-06,
+      "loss": 0.9397,
+      "step": 6068
+    },
+    {
+      "epoch": 0.4678538390379278,
+      "grad_norm": 3.383547067642212,
+      "learning_rate": 5.761079646567379e-06,
+      "loss": 0.95,
+      "step": 6069
+    },
+    {
+      "epoch": 0.4679309281529448,
+      "grad_norm": 3.7536256313323975,
+      "learning_rate": 5.7598457234363484e-06,
+      "loss": 1.0254,
+      "step": 6070
+    },
+    {
+      "epoch": 0.46800801726796176,
+      "grad_norm": 3.5894887447357178,
+      "learning_rate": 5.7586117529328735e-06,
+      "loss": 0.9251,
+      "step": 6071
+    },
+    {
+      "epoch": 0.46808510638297873,
+      "grad_norm": 3.460188627243042,
+      "learning_rate": 5.757377735133882e-06,
+      "loss": 0.9488,
+      "step": 6072
+    },
+    {
+      "epoch": 0.4681621954979957,
+      "grad_norm": 3.609666109085083,
+      "learning_rate": 5.756143670116315e-06,
+      "loss": 0.8807,
+      "step": 6073
+    },
+    {
+      "epoch": 0.4682392846130126,
+      "grad_norm": 3.767493724822998,
+      "learning_rate": 5.754909557957104e-06,
+      "loss": 1.1737,
+      "step": 6074
+    },
+    {
+      "epoch": 0.4683163737280296,
+      "grad_norm": 3.6794867515563965,
+      "learning_rate": 5.753675398733192e-06,
+      "loss": 1.0523,
+      "step": 6075
+    },
+    {
+      "epoch": 0.46839346284304656,
+      "grad_norm": 3.692859649658203,
+      "learning_rate": 5.752441192521523e-06,
+      "loss": 0.9905,
+      "step": 6076
+    },
+    {
+      "epoch": 0.46847055195806353,
+      "grad_norm": 3.4956281185150146,
+      "learning_rate": 5.751206939399041e-06,
+      "loss": 0.8996,
+      "step": 6077
+    },
+    {
+      "epoch": 0.4685476410730805,
+      "grad_norm": 3.6580235958099365,
+      "learning_rate": 5.749972639442698e-06,
+      "loss": 1.0527,
+      "step": 6078
+    },
+    {
+      "epoch": 0.4686247301880974,
+      "grad_norm": 3.3004517555236816,
+      "learning_rate": 5.748738292729445e-06,
+      "loss": 1.0016,
+      "step": 6079
+    },
+    {
+      "epoch": 0.4687018193031144,
+      "grad_norm": 3.7859244346618652,
+      "learning_rate": 5.747503899336238e-06,
+      "loss": 1.0551,
+      "step": 6080
+    },
+    {
+      "epoch": 0.46877890841813136,
+      "grad_norm": 3.9992268085479736,
+      "learning_rate": 5.746269459340034e-06,
+      "loss": 0.9909,
+      "step": 6081
+    },
+    {
+      "epoch": 0.4688559975331483,
+      "grad_norm": 3.6973602771759033,
+      "learning_rate": 5.7450349728177945e-06,
+      "loss": 0.9538,
+      "step": 6082
+    },
+    {
+      "epoch": 0.4689330866481653,
+      "grad_norm": 3.4329545497894287,
+      "learning_rate": 5.743800439846482e-06,
+      "loss": 0.9311,
+      "step": 6083
+    },
+    {
+      "epoch": 0.4690101757631822,
+      "grad_norm": 3.846623182296753,
+      "learning_rate": 5.742565860503066e-06,
+      "loss": 0.9475,
+      "step": 6084
+    },
+    {
+      "epoch": 0.4690872648781992,
+      "grad_norm": 4.217624664306641,
+      "learning_rate": 5.741331234864513e-06,
+      "loss": 1.204,
+      "step": 6085
+    },
+    {
+      "epoch": 0.46916435399321615,
+      "grad_norm": 4.2322258949279785,
+      "learning_rate": 5.740096563007797e-06,
+      "loss": 0.9926,
+      "step": 6086
+    },
+    {
+      "epoch": 0.4692414431082331,
+      "grad_norm": 4.023532867431641,
+      "learning_rate": 5.738861845009894e-06,
+      "loss": 1.0084,
+      "step": 6087
+    },
+    {
+      "epoch": 0.4693185322232501,
+      "grad_norm": 3.6507327556610107,
+      "learning_rate": 5.737627080947781e-06,
+      "loss": 0.9628,
+      "step": 6088
+    },
+    {
+      "epoch": 0.469395621338267,
+      "grad_norm": 3.5047266483306885,
+      "learning_rate": 5.73639227089844e-06,
+      "loss": 0.9771,
+      "step": 6089
+    },
+    {
+      "epoch": 0.469472710453284,
+      "grad_norm": 3.76613450050354,
+      "learning_rate": 5.735157414938855e-06,
+      "loss": 0.981,
+      "step": 6090
+    },
+    {
+      "epoch": 0.46954979956830095,
+      "grad_norm": 3.5989575386047363,
+      "learning_rate": 5.733922513146013e-06,
+      "loss": 0.8601,
+      "step": 6091
+    },
+    {
+      "epoch": 0.4696268886833179,
+      "grad_norm": 4.787153720855713,
+      "learning_rate": 5.732687565596904e-06,
+      "loss": 1.1043,
+      "step": 6092
+    },
+    {
+      "epoch": 0.4697039777983349,
+      "grad_norm": 3.478583812713623,
+      "learning_rate": 5.731452572368517e-06,
+      "loss": 0.9255,
+      "step": 6093
+    },
+    {
+      "epoch": 0.4697810669133518,
+      "grad_norm": 3.7596375942230225,
+      "learning_rate": 5.730217533537853e-06,
+      "loss": 1.1279,
+      "step": 6094
+    },
+    {
+      "epoch": 0.4698581560283688,
+      "grad_norm": 3.626962661743164,
+      "learning_rate": 5.728982449181907e-06,
+      "loss": 1.0483,
+      "step": 6095
+    },
+    {
+      "epoch": 0.46993524514338575,
+      "grad_norm": 3.686464548110962,
+      "learning_rate": 5.72774731937768e-06,
+      "loss": 0.9567,
+      "step": 6096
+    },
+    {
+      "epoch": 0.4700123342584027,
+      "grad_norm": 4.225037574768066,
+      "learning_rate": 5.7265121442021784e-06,
+      "loss": 1.0122,
+      "step": 6097
+    },
+    {
+      "epoch": 0.4700894233734197,
+      "grad_norm": 3.9033050537109375,
+      "learning_rate": 5.725276923732406e-06,
+      "loss": 0.9657,
+      "step": 6098
+    },
+    {
+      "epoch": 0.4701665124884366,
+      "grad_norm": 3.35575532913208,
+      "learning_rate": 5.724041658045374e-06,
+      "loss": 0.9243,
+      "step": 6099
+    },
+    {
+      "epoch": 0.4702436016034536,
+      "grad_norm": 3.7150776386260986,
+      "learning_rate": 5.722806347218095e-06,
+      "loss": 0.98,
+      "step": 6100
+    },
+    {
+      "epoch": 0.47032069071847055,
+      "grad_norm": 3.7382919788360596,
+      "learning_rate": 5.721570991327583e-06,
+      "loss": 1.1083,
+      "step": 6101
+    },
+    {
+      "epoch": 0.4703977798334875,
+      "grad_norm": 3.8335604667663574,
+      "learning_rate": 5.720335590450858e-06,
+      "loss": 0.9726,
+      "step": 6102
+    },
+    {
+      "epoch": 0.4704748689485045,
+      "grad_norm": 3.5674989223480225,
+      "learning_rate": 5.719100144664938e-06,
+      "loss": 1.0163,
+      "step": 6103
+    },
+    {
+      "epoch": 0.47055195806352146,
+      "grad_norm": 3.2261412143707275,
+      "learning_rate": 5.71786465404685e-06,
+      "loss": 0.9454,
+      "step": 6104
+    },
+    {
+      "epoch": 0.4706290471785384,
+      "grad_norm": 3.453087091445923,
+      "learning_rate": 5.716629118673618e-06,
+      "loss": 0.8817,
+      "step": 6105
+    },
+    {
+      "epoch": 0.47070613629355534,
+      "grad_norm": 3.143789768218994,
+      "learning_rate": 5.7153935386222715e-06,
+      "loss": 0.9472,
+      "step": 6106
+    },
+    {
+      "epoch": 0.4707832254085723,
+      "grad_norm": 3.5191619396209717,
+      "learning_rate": 5.714157913969843e-06,
+      "loss": 0.9291,
+      "step": 6107
+    },
+    {
+      "epoch": 0.4708603145235893,
+      "grad_norm": 3.2179696559906006,
+      "learning_rate": 5.712922244793369e-06,
+      "loss": 0.9505,
+      "step": 6108
+    },
+    {
+      "epoch": 0.47093740363860626,
+      "grad_norm": 3.557729959487915,
+      "learning_rate": 5.711686531169883e-06,
+      "loss": 0.999,
+      "step": 6109
+    },
+    {
+      "epoch": 0.47101449275362317,
+      "grad_norm": 3.841547727584839,
+      "learning_rate": 5.710450773176428e-06,
+      "loss": 0.987,
+      "step": 6110
+    },
+    {
+      "epoch": 0.47109158186864014,
+      "grad_norm": 3.502171039581299,
+      "learning_rate": 5.709214970890049e-06,
+      "loss": 0.9596,
+      "step": 6111
+    },
+    {
+      "epoch": 0.4711686709836571,
+      "grad_norm": 3.86859393119812,
+      "learning_rate": 5.707979124387788e-06,
+      "loss": 0.9704,
+      "step": 6112
+    },
+    {
+      "epoch": 0.4712457600986741,
+      "grad_norm": 3.8467605113983154,
+      "learning_rate": 5.706743233746695e-06,
+      "loss": 0.9401,
+      "step": 6113
+    },
+    {
+      "epoch": 0.47132284921369105,
+      "grad_norm": 3.454573392868042,
+      "learning_rate": 5.705507299043822e-06,
+      "loss": 0.8615,
+      "step": 6114
+    },
+    {
+      "epoch": 0.47139993832870797,
+      "grad_norm": 3.3528759479522705,
+      "learning_rate": 5.704271320356223e-06,
+      "loss": 0.8713,
+      "step": 6115
+    },
+    {
+      "epoch": 0.47147702744372494,
+      "grad_norm": 3.8421576023101807,
+      "learning_rate": 5.703035297760956e-06,
+      "loss": 1.066,
+      "step": 6116
+    },
+    {
+      "epoch": 0.4715541165587419,
+      "grad_norm": 3.7155203819274902,
+      "learning_rate": 5.7017992313350765e-06,
+      "loss": 1.0062,
+      "step": 6117
+    },
+    {
+      "epoch": 0.4716312056737589,
+      "grad_norm": 3.5856966972351074,
+      "learning_rate": 5.700563121155651e-06,
+      "loss": 1.0004,
+      "step": 6118
+    },
+    {
+      "epoch": 0.47170829478877585,
+      "grad_norm": 4.203090190887451,
+      "learning_rate": 5.699326967299743e-06,
+      "loss": 1.0252,
+      "step": 6119
+    },
+    {
+      "epoch": 0.47178538390379277,
+      "grad_norm": 3.741010904312134,
+      "learning_rate": 5.698090769844421e-06,
+      "loss": 1.0402,
+      "step": 6120
+    },
+    {
+      "epoch": 0.47186247301880974,
+      "grad_norm": 3.547619104385376,
+      "learning_rate": 5.696854528866755e-06,
+      "loss": 0.8535,
+      "step": 6121
+    },
+    {
+      "epoch": 0.4719395621338267,
+      "grad_norm": 3.611316442489624,
+      "learning_rate": 5.695618244443818e-06,
+      "loss": 1.0149,
+      "step": 6122
+    },
+    {
+      "epoch": 0.4720166512488437,
+      "grad_norm": 3.562122344970703,
+      "learning_rate": 5.694381916652686e-06,
+      "loss": 0.9871,
+      "step": 6123
+    },
+    {
+      "epoch": 0.47209374036386065,
+      "grad_norm": 3.538245916366577,
+      "learning_rate": 5.693145545570439e-06,
+      "loss": 0.8019,
+      "step": 6124
+    },
+    {
+      "epoch": 0.47217082947887756,
+      "grad_norm": 3.3615951538085938,
+      "learning_rate": 5.691909131274156e-06,
+      "loss": 0.9203,
+      "step": 6125
+    },
+    {
+      "epoch": 0.47224791859389453,
+      "grad_norm": 3.911485433578491,
+      "learning_rate": 5.6906726738409215e-06,
+      "loss": 0.9443,
+      "step": 6126
+    },
+    {
+      "epoch": 0.4723250077089115,
+      "grad_norm": 3.790710926055908,
+      "learning_rate": 5.689436173347825e-06,
+      "loss": 0.9956,
+      "step": 6127
+    },
+    {
+      "epoch": 0.4724020968239285,
+      "grad_norm": 3.447726249694824,
+      "learning_rate": 5.688199629871952e-06,
+      "loss": 0.9532,
+      "step": 6128
+    },
+    {
+      "epoch": 0.47247918593894545,
+      "grad_norm": 3.5070300102233887,
+      "learning_rate": 5.686963043490398e-06,
+      "loss": 0.9041,
+      "step": 6129
+    },
+    {
+      "epoch": 0.47255627505396236,
+      "grad_norm": 3.885089635848999,
+      "learning_rate": 5.6857264142802535e-06,
+      "loss": 1.0288,
+      "step": 6130
+    },
+    {
+      "epoch": 0.47263336416897933,
+      "grad_norm": 3.5010008811950684,
+      "learning_rate": 5.68448974231862e-06,
+      "loss": 0.8738,
+      "step": 6131
+    },
+    {
+      "epoch": 0.4727104532839963,
+      "grad_norm": 3.6111185550689697,
+      "learning_rate": 5.683253027682597e-06,
+      "loss": 1.0589,
+      "step": 6132
+    },
+    {
+      "epoch": 0.47278754239901327,
+      "grad_norm": 3.557774305343628,
+      "learning_rate": 5.682016270449286e-06,
+      "loss": 1.004,
+      "step": 6133
+    },
+    {
+      "epoch": 0.47286463151403024,
+      "grad_norm": 3.7765564918518066,
+      "learning_rate": 5.680779470695791e-06,
+      "loss": 0.9748,
+      "step": 6134
+    },
+    {
+      "epoch": 0.47294172062904716,
+      "grad_norm": 3.426037549972534,
+      "learning_rate": 5.679542628499224e-06,
+      "loss": 0.8693,
+      "step": 6135
+    },
+    {
+      "epoch": 0.47301880974406413,
+      "grad_norm": 3.553133487701416,
+      "learning_rate": 5.678305743936692e-06,
+      "loss": 0.9648,
+      "step": 6136
+    },
+    {
+      "epoch": 0.4730958988590811,
+      "grad_norm": 3.720686197280884,
+      "learning_rate": 5.677068817085311e-06,
+      "loss": 1.0429,
+      "step": 6137
+    },
+    {
+      "epoch": 0.47317298797409807,
+      "grad_norm": 3.6694304943084717,
+      "learning_rate": 5.675831848022195e-06,
+      "loss": 0.9811,
+      "step": 6138
+    },
+    {
+      "epoch": 0.47325007708911504,
+      "grad_norm": 3.528355121612549,
+      "learning_rate": 5.674594836824463e-06,
+      "loss": 1.0162,
+      "step": 6139
+    },
+    {
+      "epoch": 0.47332716620413195,
+      "grad_norm": 3.9201457500457764,
+      "learning_rate": 5.673357783569238e-06,
+      "loss": 1.0571,
+      "step": 6140
+    },
+    {
+      "epoch": 0.4734042553191489,
+      "grad_norm": 3.4825947284698486,
+      "learning_rate": 5.672120688333642e-06,
+      "loss": 0.8942,
+      "step": 6141
+    },
+    {
+      "epoch": 0.4734813444341659,
+      "grad_norm": 3.4814279079437256,
+      "learning_rate": 5.6708835511948035e-06,
+      "loss": 0.8944,
+      "step": 6142
+    },
+    {
+      "epoch": 0.47355843354918287,
+      "grad_norm": 3.8149969577789307,
+      "learning_rate": 5.669646372229849e-06,
+      "loss": 0.9892,
+      "step": 6143
+    },
+    {
+      "epoch": 0.47363552266419984,
+      "grad_norm": 3.37603497505188,
+      "learning_rate": 5.6684091515159105e-06,
+      "loss": 0.8202,
+      "step": 6144
+    },
+    {
+      "epoch": 0.47371261177921675,
+      "grad_norm": 3.226973056793213,
+      "learning_rate": 5.667171889130125e-06,
+      "loss": 0.9294,
+      "step": 6145
+    },
+    {
+      "epoch": 0.4737897008942337,
+      "grad_norm": 3.621938467025757,
+      "learning_rate": 5.6659345851496265e-06,
+      "loss": 1.0182,
+      "step": 6146
+    },
+    {
+      "epoch": 0.4738667900092507,
+      "grad_norm": 3.5621347427368164,
+      "learning_rate": 5.6646972396515555e-06,
+      "loss": 0.9215,
+      "step": 6147
+    },
+    {
+      "epoch": 0.47394387912426766,
+      "grad_norm": 4.244327545166016,
+      "learning_rate": 5.663459852713055e-06,
+      "loss": 1.0142,
+      "step": 6148
+    },
+    {
+      "epoch": 0.47402096823928463,
+      "grad_norm": 3.6657721996307373,
+      "learning_rate": 5.662222424411268e-06,
+      "loss": 1.0159,
+      "step": 6149
+    },
+    {
+      "epoch": 0.47409805735430155,
+      "grad_norm": 3.6783275604248047,
+      "learning_rate": 5.660984954823342e-06,
+      "loss": 1.0533,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4741751464693185,
+      "grad_norm": 3.747204542160034,
+      "learning_rate": 5.659747444026429e-06,
+      "loss": 1.04,
+      "step": 6151
+    },
+    {
+      "epoch": 0.4742522355843355,
+      "grad_norm": 3.452821969985962,
+      "learning_rate": 5.658509892097679e-06,
+      "loss": 0.9084,
+      "step": 6152
+    },
+    {
+      "epoch": 0.47432932469935246,
+      "grad_norm": 3.9301180839538574,
+      "learning_rate": 5.657272299114248e-06,
+      "loss": 0.8685,
+      "step": 6153
+    },
+    {
+      "epoch": 0.47440641381436943,
+      "grad_norm": 3.441697120666504,
+      "learning_rate": 5.656034665153294e-06,
+      "loss": 1.0103,
+      "step": 6154
+    },
+    {
+      "epoch": 0.47448350292938635,
+      "grad_norm": 3.8465073108673096,
+      "learning_rate": 5.654796990291974e-06,
+      "loss": 1.0009,
+      "step": 6155
+    },
+    {
+      "epoch": 0.4745605920444033,
+      "grad_norm": 3.6904850006103516,
+      "learning_rate": 5.6535592746074554e-06,
+      "loss": 0.9484,
+      "step": 6156
+    },
+    {
+      "epoch": 0.4746376811594203,
+      "grad_norm": 3.667325019836426,
+      "learning_rate": 5.6523215181769e-06,
+      "loss": 0.9456,
+      "step": 6157
+    },
+    {
+      "epoch": 0.47471477027443726,
+      "grad_norm": 3.483203172683716,
+      "learning_rate": 5.651083721077475e-06,
+      "loss": 1.0455,
+      "step": 6158
+    },
+    {
+      "epoch": 0.47479185938945423,
+      "grad_norm": 3.435410499572754,
+      "learning_rate": 5.649845883386355e-06,
+      "loss": 0.8986,
+      "step": 6159
+    },
+    {
+      "epoch": 0.47486894850447114,
+      "grad_norm": 3.477846384048462,
+      "learning_rate": 5.6486080051807066e-06,
+      "loss": 0.8902,
+      "step": 6160
+    },
+    {
+      "epoch": 0.4749460376194881,
+      "grad_norm": 3.9039952754974365,
+      "learning_rate": 5.647370086537709e-06,
+      "loss": 1.0054,
+      "step": 6161
+    },
+    {
+      "epoch": 0.4750231267345051,
+      "grad_norm": 3.6027846336364746,
+      "learning_rate": 5.646132127534541e-06,
+      "loss": 1.1071,
+      "step": 6162
+    },
+    {
+      "epoch": 0.47510021584952206,
+      "grad_norm": 4.0150370597839355,
+      "learning_rate": 5.6448941282483795e-06,
+      "loss": 0.9272,
+      "step": 6163
+    },
+    {
+      "epoch": 0.475177304964539,
+      "grad_norm": 3.5615651607513428,
+      "learning_rate": 5.64365608875641e-06,
+      "loss": 0.9647,
+      "step": 6164
+    },
+    {
+      "epoch": 0.47525439407955594,
+      "grad_norm": 3.8740906715393066,
+      "learning_rate": 5.6424180091358175e-06,
+      "loss": 1.0794,
+      "step": 6165
+    },
+    {
+      "epoch": 0.4753314831945729,
+      "grad_norm": 3.5737662315368652,
+      "learning_rate": 5.641179889463788e-06,
+      "loss": 0.9521,
+      "step": 6166
+    },
+    {
+      "epoch": 0.4754085723095899,
+      "grad_norm": 3.58293080329895,
+      "learning_rate": 5.639941729817514e-06,
+      "loss": 1.0373,
+      "step": 6167
+    },
+    {
+      "epoch": 0.47548566142460685,
+      "grad_norm": 3.5029170513153076,
+      "learning_rate": 5.638703530274187e-06,
+      "loss": 1.0038,
+      "step": 6168
+    },
+    {
+      "epoch": 0.4755627505396238,
+      "grad_norm": 3.324521064758301,
+      "learning_rate": 5.637465290911004e-06,
+      "loss": 0.9587,
+      "step": 6169
+    },
+    {
+      "epoch": 0.47563983965464074,
+      "grad_norm": 3.4700326919555664,
+      "learning_rate": 5.63622701180516e-06,
+      "loss": 0.9135,
+      "step": 6170
+    },
+    {
+      "epoch": 0.4757169287696577,
+      "grad_norm": 3.6384148597717285,
+      "learning_rate": 5.634988693033857e-06,
+      "loss": 0.868,
+      "step": 6171
+    },
+    {
+      "epoch": 0.4757940178846747,
+      "grad_norm": 3.50246000289917,
+      "learning_rate": 5.6337503346743e-06,
+      "loss": 0.8695,
+      "step": 6172
+    },
+    {
+      "epoch": 0.47587110699969165,
+      "grad_norm": 3.835340738296509,
+      "learning_rate": 5.632511936803689e-06,
+      "loss": 1.0377,
+      "step": 6173
+    },
+    {
+      "epoch": 0.4759481961147086,
+      "grad_norm": 3.505685567855835,
+      "learning_rate": 5.631273499499236e-06,
+      "loss": 0.902,
+      "step": 6174
+    },
+    {
+      "epoch": 0.47602528522972554,
+      "grad_norm": 3.740123987197876,
+      "learning_rate": 5.630035022838151e-06,
+      "loss": 1.0631,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4761023743447425,
+      "grad_norm": 3.739340305328369,
+      "learning_rate": 5.628796506897642e-06,
+      "loss": 1.0683,
+      "step": 6176
+    },
+    {
+      "epoch": 0.4761794634597595,
+      "grad_norm": 3.675077438354492,
+      "learning_rate": 5.6275579517549306e-06,
+      "loss": 1.016,
+      "step": 6177
+    },
+    {
+      "epoch": 0.47625655257477645,
+      "grad_norm": 3.387910842895508,
+      "learning_rate": 5.62631935748723e-06,
+      "loss": 0.8774,
+      "step": 6178
+    },
+    {
+      "epoch": 0.4763336416897934,
+      "grad_norm": 3.4367949962615967,
+      "learning_rate": 5.625080724171761e-06,
+      "loss": 0.9452,
+      "step": 6179
+    },
+    {
+      "epoch": 0.47641073080481033,
+      "grad_norm": 3.6171581745147705,
+      "learning_rate": 5.623842051885747e-06,
+      "loss": 0.9478,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4764878199198273,
+      "grad_norm": 3.505159854888916,
+      "learning_rate": 5.622603340706411e-06,
+      "loss": 0.9781,
+      "step": 6181
+    },
+    {
+      "epoch": 0.4765649090348443,
+      "grad_norm": 3.581728219985962,
+      "learning_rate": 5.621364590710981e-06,
+      "loss": 0.9409,
+      "step": 6182
+    },
+    {
+      "epoch": 0.47664199814986125,
+      "grad_norm": 3.7894256114959717,
+      "learning_rate": 5.620125801976687e-06,
+      "loss": 0.9496,
+      "step": 6183
+    },
+    {
+      "epoch": 0.4767190872648782,
+      "grad_norm": 3.7810251712799072,
+      "learning_rate": 5.6188869745807614e-06,
+      "loss": 1.0094,
+      "step": 6184
+    },
+    {
+      "epoch": 0.47679617637989513,
+      "grad_norm": 3.6951205730438232,
+      "learning_rate": 5.6176481086004395e-06,
+      "loss": 0.8738,
+      "step": 6185
+    },
+    {
+      "epoch": 0.4768732654949121,
+      "grad_norm": 3.578882932662964,
+      "learning_rate": 5.6164092041129544e-06,
+      "loss": 0.9924,
+      "step": 6186
+    },
+    {
+      "epoch": 0.4769503546099291,
+      "grad_norm": 3.560987949371338,
+      "learning_rate": 5.615170261195549e-06,
+      "loss": 0.9863,
+      "step": 6187
+    },
+    {
+      "epoch": 0.47702744372494604,
+      "grad_norm": 3.433151960372925,
+      "learning_rate": 5.613931279925465e-06,
+      "loss": 0.9885,
+      "step": 6188
+    },
+    {
+      "epoch": 0.477104532839963,
+      "grad_norm": 3.652071237564087,
+      "learning_rate": 5.612692260379945e-06,
+      "loss": 0.9823,
+      "step": 6189
+    },
+    {
+      "epoch": 0.47718162195498,
+      "grad_norm": 3.784919261932373,
+      "learning_rate": 5.611453202636236e-06,
+      "loss": 1.0343,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4772587110699969,
+      "grad_norm": 4.090549945831299,
+      "learning_rate": 5.610214106771585e-06,
+      "loss": 1.0221,
+      "step": 6191
+    },
+    {
+      "epoch": 0.47733580018501387,
+      "grad_norm": 3.437002182006836,
+      "learning_rate": 5.608974972863245e-06,
+      "loss": 0.9871,
+      "step": 6192
+    },
+    {
+      "epoch": 0.47741288930003084,
+      "grad_norm": 3.5076773166656494,
+      "learning_rate": 5.6077358009884705e-06,
+      "loss": 0.8728,
+      "step": 6193
+    },
+    {
+      "epoch": 0.4774899784150478,
+      "grad_norm": 3.560811996459961,
+      "learning_rate": 5.606496591224516e-06,
+      "loss": 1.0416,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4775670675300648,
+      "grad_norm": 3.939362049102783,
+      "learning_rate": 5.60525734364864e-06,
+      "loss": 1.0182,
+      "step": 6195
+    },
+    {
+      "epoch": 0.4776441566450817,
+      "grad_norm": 3.568834066390991,
+      "learning_rate": 5.604018058338104e-06,
+      "loss": 0.8795,
+      "step": 6196
+    },
+    {
+      "epoch": 0.47772124576009867,
+      "grad_norm": 3.514498710632324,
+      "learning_rate": 5.602778735370169e-06,
+      "loss": 0.9279,
+      "step": 6197
+    },
+    {
+      "epoch": 0.47779833487511564,
+      "grad_norm": 3.4076337814331055,
+      "learning_rate": 5.601539374822103e-06,
+      "loss": 0.9863,
+      "step": 6198
+    },
+    {
+      "epoch": 0.4778754239901326,
+      "grad_norm": 3.399474859237671,
+      "learning_rate": 5.600299976771172e-06,
+      "loss": 0.9236,
+      "step": 6199
+    },
+    {
+      "epoch": 0.4779525131051496,
+      "grad_norm": 3.3580496311187744,
+      "learning_rate": 5.5990605412946466e-06,
+      "loss": 0.9451,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4780296022201665,
+      "grad_norm": 3.756740093231201,
+      "learning_rate": 5.597821068469799e-06,
+      "loss": 1.0399,
+      "step": 6201
+    },
+    {
+      "epoch": 0.47810669133518346,
+      "grad_norm": 3.5835258960723877,
+      "learning_rate": 5.596581558373903e-06,
+      "loss": 0.9042,
+      "step": 6202
+    },
+    {
+      "epoch": 0.47818378045020044,
+      "grad_norm": 3.941892385482788,
+      "learning_rate": 5.595342011084237e-06,
+      "loss": 1.0083,
+      "step": 6203
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 3.4614713191986084,
+      "learning_rate": 5.594102426678082e-06,
+      "loss": 0.9398,
+      "step": 6204
+    },
+    {
+      "epoch": 0.4783379586802344,
+      "grad_norm": 3.545919179916382,
+      "learning_rate": 5.592862805232714e-06,
+      "loss": 1.0258,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4784150477952513,
+      "grad_norm": 3.4241621494293213,
+      "learning_rate": 5.591623146825423e-06,
+      "loss": 0.8465,
+      "step": 6206
+    },
+    {
+      "epoch": 0.47849213691026826,
+      "grad_norm": 3.324366807937622,
+      "learning_rate": 5.590383451533493e-06,
+      "loss": 0.9298,
+      "step": 6207
+    },
+    {
+      "epoch": 0.47856922602528523,
+      "grad_norm": 3.5040102005004883,
+      "learning_rate": 5.589143719434211e-06,
+      "loss": 0.9745,
+      "step": 6208
+    },
+    {
+      "epoch": 0.4786463151403022,
+      "grad_norm": 3.782944440841675,
+      "learning_rate": 5.587903950604872e-06,
+      "loss": 1.0106,
+      "step": 6209
+    },
+    {
+      "epoch": 0.4787234042553192,
+      "grad_norm": 3.437720537185669,
+      "learning_rate": 5.586664145122764e-06,
+      "loss": 0.9793,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4788004933703361,
+      "grad_norm": 4.048053741455078,
+      "learning_rate": 5.585424303065186e-06,
+      "loss": 1.0484,
+      "step": 6211
+    },
+    {
+      "epoch": 0.47887758248535306,
+      "grad_norm": 3.8065176010131836,
+      "learning_rate": 5.5841844245094345e-06,
+      "loss": 0.9522,
+      "step": 6212
+    },
+    {
+      "epoch": 0.47895467160037003,
+      "grad_norm": 3.6226398944854736,
+      "learning_rate": 5.582944509532809e-06,
+      "loss": 0.974,
+      "step": 6213
+    },
+    {
+      "epoch": 0.479031760715387,
+      "grad_norm": 4.275754451751709,
+      "learning_rate": 5.581704558212615e-06,
+      "loss": 1.037,
+      "step": 6214
+    },
+    {
+      "epoch": 0.47910884983040397,
+      "grad_norm": 3.503654718399048,
+      "learning_rate": 5.5804645706261515e-06,
+      "loss": 1.0253,
+      "step": 6215
+    },
+    {
+      "epoch": 0.4791859389454209,
+      "grad_norm": 3.7088232040405273,
+      "learning_rate": 5.57922454685073e-06,
+      "loss": 1.0347,
+      "step": 6216
+    },
+    {
+      "epoch": 0.47926302806043786,
+      "grad_norm": 3.6020233631134033,
+      "learning_rate": 5.577984486963658e-06,
+      "loss": 0.9362,
+      "step": 6217
+    },
+    {
+      "epoch": 0.4793401171754548,
+      "grad_norm": 3.534729242324829,
+      "learning_rate": 5.576744391042246e-06,
+      "loss": 1.0147,
+      "step": 6218
+    },
+    {
+      "epoch": 0.4794172062904718,
+      "grad_norm": 3.644331693649292,
+      "learning_rate": 5.575504259163807e-06,
+      "loss": 0.9651,
+      "step": 6219
+    },
+    {
+      "epoch": 0.47949429540548877,
+      "grad_norm": 3.5781354904174805,
+      "learning_rate": 5.5742640914056615e-06,
+      "loss": 0.9953,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4795713845205057,
+      "grad_norm": 3.2821385860443115,
+      "learning_rate": 5.573023887845122e-06,
+      "loss": 0.9041,
+      "step": 6221
+    },
+    {
+      "epoch": 0.47964847363552265,
+      "grad_norm": 3.6788458824157715,
+      "learning_rate": 5.57178364855951e-06,
+      "loss": 1.1181,
+      "step": 6222
+    },
+    {
+      "epoch": 0.4797255627505396,
+      "grad_norm": 3.8085317611694336,
+      "learning_rate": 5.57054337362615e-06,
+      "loss": 0.8443,
+      "step": 6223
+    },
+    {
+      "epoch": 0.4798026518655566,
+      "grad_norm": 3.922316789627075,
+      "learning_rate": 5.569303063122364e-06,
+      "loss": 1.0382,
+      "step": 6224
+    },
+    {
+      "epoch": 0.47987974098057357,
+      "grad_norm": 3.4889650344848633,
+      "learning_rate": 5.568062717125483e-06,
+      "loss": 0.8957,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4799568300955905,
+      "grad_norm": 3.5855491161346436,
+      "learning_rate": 5.5668223357128325e-06,
+      "loss": 0.9909,
+      "step": 6226
+    },
+    {
+      "epoch": 0.48003391921060745,
+      "grad_norm": 3.6739110946655273,
+      "learning_rate": 5.5655819189617445e-06,
+      "loss": 0.8751,
+      "step": 6227
+    },
+    {
+      "epoch": 0.4801110083256244,
+      "grad_norm": 3.6241581439971924,
+      "learning_rate": 5.564341466949553e-06,
+      "loss": 0.9153,
+      "step": 6228
+    },
+    {
+      "epoch": 0.4801880974406414,
+      "grad_norm": 3.6307456493377686,
+      "learning_rate": 5.5631009797535955e-06,
+      "loss": 0.9424,
+      "step": 6229
+    },
+    {
+      "epoch": 0.48026518655565836,
+      "grad_norm": 3.7684690952301025,
+      "learning_rate": 5.561860457451207e-06,
+      "loss": 0.9433,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4803422756706753,
+      "grad_norm": 3.5321202278137207,
+      "learning_rate": 5.560619900119729e-06,
+      "loss": 0.9128,
+      "step": 6231
+    },
+    {
+      "epoch": 0.48041936478569225,
+      "grad_norm": 3.7544522285461426,
+      "learning_rate": 5.5593793078365036e-06,
+      "loss": 1.0287,
+      "step": 6232
+    },
+    {
+      "epoch": 0.4804964539007092,
+      "grad_norm": 3.454862356185913,
+      "learning_rate": 5.5581386806788765e-06,
+      "loss": 0.9557,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4805735430157262,
+      "grad_norm": 3.669862747192383,
+      "learning_rate": 5.5568980187241915e-06,
+      "loss": 1.0802,
+      "step": 6234
+    },
+    {
+      "epoch": 0.48065063213074316,
+      "grad_norm": 3.8794069290161133,
+      "learning_rate": 5.5556573220498e-06,
+      "loss": 0.9532,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4807277212457601,
+      "grad_norm": 3.671264171600342,
+      "learning_rate": 5.554416590733054e-06,
+      "loss": 0.9586,
+      "step": 6236
+    },
+    {
+      "epoch": 0.48080481036077705,
+      "grad_norm": 3.6556334495544434,
+      "learning_rate": 5.553175824851304e-06,
+      "loss": 0.8997,
+      "step": 6237
+    },
+    {
+      "epoch": 0.480881899475794,
+      "grad_norm": 3.8295559883117676,
+      "learning_rate": 5.551935024481906e-06,
+      "loss": 1.142,
+      "step": 6238
+    },
+    {
+      "epoch": 0.480958988590811,
+      "grad_norm": 3.51387357711792,
+      "learning_rate": 5.5506941897022175e-06,
+      "loss": 0.9573,
+      "step": 6239
+    },
+    {
+      "epoch": 0.48103607770582796,
+      "grad_norm": 3.6698529720306396,
+      "learning_rate": 5.549453320589598e-06,
+      "loss": 0.9651,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4811131668208449,
+      "grad_norm": 3.5171773433685303,
+      "learning_rate": 5.54821241722141e-06,
+      "loss": 0.8298,
+      "step": 6241
+    },
+    {
+      "epoch": 0.48119025593586184,
+      "grad_norm": 4.172097682952881,
+      "learning_rate": 5.5469714796750175e-06,
+      "loss": 1.0308,
+      "step": 6242
+    },
+    {
+      "epoch": 0.4812673450508788,
+      "grad_norm": 3.4253857135772705,
+      "learning_rate": 5.5457305080277855e-06,
+      "loss": 0.9464,
+      "step": 6243
+    },
+    {
+      "epoch": 0.4813444341658958,
+      "grad_norm": 3.268526554107666,
+      "learning_rate": 5.544489502357085e-06,
+      "loss": 0.91,
+      "step": 6244
+    },
+    {
+      "epoch": 0.48142152328091276,
+      "grad_norm": 3.801954507827759,
+      "learning_rate": 5.543248462740281e-06,
+      "loss": 0.9596,
+      "step": 6245
+    },
+    {
+      "epoch": 0.48149861239592967,
+      "grad_norm": 3.7973077297210693,
+      "learning_rate": 5.542007389254749e-06,
+      "loss": 1.0331,
+      "step": 6246
+    },
+    {
+      "epoch": 0.48157570151094664,
+      "grad_norm": 3.8994927406311035,
+      "learning_rate": 5.540766281977865e-06,
+      "loss": 0.9226,
+      "step": 6247
+    },
+    {
+      "epoch": 0.4816527906259636,
+      "grad_norm": 3.246563196182251,
+      "learning_rate": 5.539525140987003e-06,
+      "loss": 0.8147,
+      "step": 6248
+    },
+    {
+      "epoch": 0.4817298797409806,
+      "grad_norm": 3.5368645191192627,
+      "learning_rate": 5.538283966359545e-06,
+      "loss": 0.9997,
+      "step": 6249
+    },
+    {
+      "epoch": 0.48180696885599755,
+      "grad_norm": 4.268583297729492,
+      "learning_rate": 5.537042758172866e-06,
+      "loss": 1.0206,
+      "step": 6250
+    },
+    {
+      "epoch": 0.48188405797101447,
+      "grad_norm": 3.504702568054199,
+      "learning_rate": 5.535801516504354e-06,
+      "loss": 0.8864,
+      "step": 6251
+    },
+    {
+      "epoch": 0.48196114708603144,
+      "grad_norm": 3.5273540019989014,
+      "learning_rate": 5.534560241431393e-06,
+      "loss": 0.9426,
+      "step": 6252
+    },
+    {
+      "epoch": 0.4820382362010484,
+      "grad_norm": 3.7020440101623535,
+      "learning_rate": 5.533318933031368e-06,
+      "loss": 1.0103,
+      "step": 6253
+    },
+    {
+      "epoch": 0.4821153253160654,
+      "grad_norm": 3.3781721591949463,
+      "learning_rate": 5.532077591381672e-06,
+      "loss": 0.8451,
+      "step": 6254
+    },
+    {
+      "epoch": 0.48219241443108235,
+      "grad_norm": 3.646610975265503,
+      "learning_rate": 5.530836216559692e-06,
+      "loss": 1.0605,
+      "step": 6255
+    },
+    {
+      "epoch": 0.48226950354609927,
+      "grad_norm": 3.75663423538208,
+      "learning_rate": 5.5295948086428245e-06,
+      "loss": 1.0645,
+      "step": 6256
+    },
+    {
+      "epoch": 0.48234659266111624,
+      "grad_norm": 3.898165702819824,
+      "learning_rate": 5.528353367708462e-06,
+      "loss": 1.0525,
+      "step": 6257
+    },
+    {
+      "epoch": 0.4824236817761332,
+      "grad_norm": 3.5271053314208984,
+      "learning_rate": 5.527111893834004e-06,
+      "loss": 0.8598,
+      "step": 6258
+    },
+    {
+      "epoch": 0.4825007708911502,
+      "grad_norm": 3.4338533878326416,
+      "learning_rate": 5.52587038709685e-06,
+      "loss": 0.9286,
+      "step": 6259
+    },
+    {
+      "epoch": 0.48257786000616715,
+      "grad_norm": 4.084996223449707,
+      "learning_rate": 5.5246288475744016e-06,
+      "loss": 1.1074,
+      "step": 6260
+    },
+    {
+      "epoch": 0.48265494912118406,
+      "grad_norm": 3.887216806411743,
+      "learning_rate": 5.52338727534406e-06,
+      "loss": 0.8895,
+      "step": 6261
+    },
+    {
+      "epoch": 0.48273203823620103,
+      "grad_norm": 3.458646774291992,
+      "learning_rate": 5.522145670483233e-06,
+      "loss": 0.938,
+      "step": 6262
+    },
+    {
+      "epoch": 0.482809127351218,
+      "grad_norm": 3.734957456588745,
+      "learning_rate": 5.520904033069328e-06,
+      "loss": 1.0008,
+      "step": 6263
+    },
+    {
+      "epoch": 0.482886216466235,
+      "grad_norm": 3.3216323852539062,
+      "learning_rate": 5.519662363179754e-06,
+      "loss": 1.0191,
+      "step": 6264
+    },
+    {
+      "epoch": 0.48296330558125194,
+      "grad_norm": 3.6311726570129395,
+      "learning_rate": 5.518420660891924e-06,
+      "loss": 0.9028,
+      "step": 6265
+    },
+    {
+      "epoch": 0.48304039469626886,
+      "grad_norm": 3.6047608852386475,
+      "learning_rate": 5.51717892628325e-06,
+      "loss": 0.9282,
+      "step": 6266
+    },
+    {
+      "epoch": 0.48311748381128583,
+      "grad_norm": 3.5420308113098145,
+      "learning_rate": 5.515937159431147e-06,
+      "loss": 0.8657,
+      "step": 6267
+    },
+    {
+      "epoch": 0.4831945729263028,
+      "grad_norm": 3.5414748191833496,
+      "learning_rate": 5.514695360413037e-06,
+      "loss": 0.9715,
+      "step": 6268
+    },
+    {
+      "epoch": 0.48327166204131977,
+      "grad_norm": 3.513932228088379,
+      "learning_rate": 5.5134535293063355e-06,
+      "loss": 0.8782,
+      "step": 6269
+    },
+    {
+      "epoch": 0.48334875115633674,
+      "grad_norm": 3.6613216400146484,
+      "learning_rate": 5.512211666188465e-06,
+      "loss": 1.0077,
+      "step": 6270
+    },
+    {
+      "epoch": 0.48342584027135366,
+      "grad_norm": 4.074599742889404,
+      "learning_rate": 5.510969771136852e-06,
+      "loss": 0.9404,
+      "step": 6271
+    },
+    {
+      "epoch": 0.48350292938637063,
+      "grad_norm": 3.5803050994873047,
+      "learning_rate": 5.509727844228917e-06,
+      "loss": 0.9411,
+      "step": 6272
+    },
+    {
+      "epoch": 0.4835800185013876,
+      "grad_norm": 3.4778783321380615,
+      "learning_rate": 5.5084858855420945e-06,
+      "loss": 0.917,
+      "step": 6273
+    },
+    {
+      "epoch": 0.48365710761640457,
+      "grad_norm": 3.6360175609588623,
+      "learning_rate": 5.507243895153808e-06,
+      "loss": 1.0272,
+      "step": 6274
+    },
+    {
+      "epoch": 0.48373419673142154,
+      "grad_norm": 3.4350030422210693,
+      "learning_rate": 5.506001873141493e-06,
+      "loss": 0.8781,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4838112858464385,
+      "grad_norm": 3.641852378845215,
+      "learning_rate": 5.504759819582581e-06,
+      "loss": 0.9962,
+      "step": 6276
+    },
+    {
+      "epoch": 0.4838883749614554,
+      "grad_norm": 3.9919748306274414,
+      "learning_rate": 5.5035177345545085e-06,
+      "loss": 0.9799,
+      "step": 6277
+    },
+    {
+      "epoch": 0.4839654640764724,
+      "grad_norm": 3.5844922065734863,
+      "learning_rate": 5.502275618134715e-06,
+      "loss": 0.9189,
+      "step": 6278
+    },
+    {
+      "epoch": 0.48404255319148937,
+      "grad_norm": 3.5310840606689453,
+      "learning_rate": 5.5010334704006364e-06,
+      "loss": 0.9687,
+      "step": 6279
+    },
+    {
+      "epoch": 0.48411964230650634,
+      "grad_norm": 4.053980350494385,
+      "learning_rate": 5.499791291429714e-06,
+      "loss": 0.8998,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4841967314215233,
+      "grad_norm": 3.388463020324707,
+      "learning_rate": 5.498549081299397e-06,
+      "loss": 0.9235,
+      "step": 6281
+    },
+    {
+      "epoch": 0.4842738205365402,
+      "grad_norm": 3.387619972229004,
+      "learning_rate": 5.497306840087124e-06,
+      "loss": 0.9145,
+      "step": 6282
+    },
+    {
+      "epoch": 0.4843509096515572,
+      "grad_norm": 3.7833995819091797,
+      "learning_rate": 5.496064567870346e-06,
+      "loss": 0.9596,
+      "step": 6283
+    },
+    {
+      "epoch": 0.48442799876657416,
+      "grad_norm": 3.767986536026001,
+      "learning_rate": 5.494822264726512e-06,
+      "loss": 1.0506,
+      "step": 6284
+    },
+    {
+      "epoch": 0.48450508788159113,
+      "grad_norm": 3.8580732345581055,
+      "learning_rate": 5.4935799307330715e-06,
+      "loss": 1.1381,
+      "step": 6285
+    },
+    {
+      "epoch": 0.4845821769966081,
+      "grad_norm": 3.7080979347229004,
+      "learning_rate": 5.492337565967479e-06,
+      "loss": 0.9691,
+      "step": 6286
+    },
+    {
+      "epoch": 0.484659266111625,
+      "grad_norm": 3.4065332412719727,
+      "learning_rate": 5.491095170507189e-06,
+      "loss": 0.908,
+      "step": 6287
+    },
+    {
+      "epoch": 0.484736355226642,
+      "grad_norm": 3.787928819656372,
+      "learning_rate": 5.4898527444296586e-06,
+      "loss": 0.8999,
+      "step": 6288
+    },
+    {
+      "epoch": 0.48481344434165896,
+      "grad_norm": 3.6937663555145264,
+      "learning_rate": 5.488610287812348e-06,
+      "loss": 0.976,
+      "step": 6289
+    },
+    {
+      "epoch": 0.48489053345667593,
+      "grad_norm": 3.6718273162841797,
+      "learning_rate": 5.487367800732715e-06,
+      "loss": 0.9509,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4849676225716929,
+      "grad_norm": 3.9421653747558594,
+      "learning_rate": 5.486125283268223e-06,
+      "loss": 0.9137,
+      "step": 6291
+    },
+    {
+      "epoch": 0.4850447116867098,
+      "grad_norm": 3.546727418899536,
+      "learning_rate": 5.48488273549634e-06,
+      "loss": 0.9182,
+      "step": 6292
+    },
+    {
+      "epoch": 0.4851218008017268,
+      "grad_norm": 3.6610233783721924,
+      "learning_rate": 5.483640157494528e-06,
+      "loss": 0.9811,
+      "step": 6293
+    },
+    {
+      "epoch": 0.48519888991674376,
+      "grad_norm": 3.579312801361084,
+      "learning_rate": 5.482397549340256e-06,
+      "loss": 0.879,
+      "step": 6294
+    },
+    {
+      "epoch": 0.48527597903176073,
+      "grad_norm": 3.8670926094055176,
+      "learning_rate": 5.481154911110995e-06,
+      "loss": 1.0397,
+      "step": 6295
+    },
+    {
+      "epoch": 0.4853530681467777,
+      "grad_norm": 3.6566736698150635,
+      "learning_rate": 5.4799122428842185e-06,
+      "loss": 1.0115,
+      "step": 6296
+    },
+    {
+      "epoch": 0.4854301572617946,
+      "grad_norm": 3.503852605819702,
+      "learning_rate": 5.478669544737401e-06,
+      "loss": 1.0862,
+      "step": 6297
+    },
+    {
+      "epoch": 0.4855072463768116,
+      "grad_norm": 3.8774027824401855,
+      "learning_rate": 5.477426816748014e-06,
+      "loss": 1.0201,
+      "step": 6298
+    },
+    {
+      "epoch": 0.48558433549182856,
+      "grad_norm": 3.327522039413452,
+      "learning_rate": 5.476184058993539e-06,
+      "loss": 0.9576,
+      "step": 6299
+    },
+    {
+      "epoch": 0.4856614246068455,
+      "grad_norm": 3.4257946014404297,
+      "learning_rate": 5.4749412715514525e-06,
+      "loss": 0.9476,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4857385137218625,
+      "grad_norm": 4.544504165649414,
+      "learning_rate": 5.473698454499239e-06,
+      "loss": 1.0781,
+      "step": 6301
+    },
+    {
+      "epoch": 0.4858156028368794,
+      "grad_norm": 3.3634119033813477,
+      "learning_rate": 5.47245560791438e-06,
+      "loss": 0.936,
+      "step": 6302
+    },
+    {
+      "epoch": 0.4858926919518964,
+      "grad_norm": 3.5162649154663086,
+      "learning_rate": 5.47121273187436e-06,
+      "loss": 1.0161,
+      "step": 6303
+    },
+    {
+      "epoch": 0.48596978106691335,
+      "grad_norm": 3.8736398220062256,
+      "learning_rate": 5.4699698264566665e-06,
+      "loss": 0.9404,
+      "step": 6304
+    },
+    {
+      "epoch": 0.4860468701819303,
+      "grad_norm": 3.300621509552002,
+      "learning_rate": 5.468726891738789e-06,
+      "loss": 0.9441,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4861239592969473,
+      "grad_norm": 3.4393205642700195,
+      "learning_rate": 5.467483927798217e-06,
+      "loss": 0.9706,
+      "step": 6306
+    },
+    {
+      "epoch": 0.4862010484119642,
+      "grad_norm": 3.7706298828125,
+      "learning_rate": 5.4662409347124436e-06,
+      "loss": 0.9214,
+      "step": 6307
+    },
+    {
+      "epoch": 0.4862781375269812,
+      "grad_norm": 3.6923952102661133,
+      "learning_rate": 5.464997912558963e-06,
+      "loss": 0.8973,
+      "step": 6308
+    },
+    {
+      "epoch": 0.48635522664199815,
+      "grad_norm": 3.481060743331909,
+      "learning_rate": 5.46375486141527e-06,
+      "loss": 0.867,
+      "step": 6309
+    },
+    {
+      "epoch": 0.4864323157570151,
+      "grad_norm": 3.6562881469726562,
+      "learning_rate": 5.462511781358866e-06,
+      "loss": 0.9865,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4865094048720321,
+      "grad_norm": 3.6033737659454346,
+      "learning_rate": 5.461268672467245e-06,
+      "loss": 1.0277,
+      "step": 6311
+    },
+    {
+      "epoch": 0.486586493987049,
+      "grad_norm": 3.8710803985595703,
+      "learning_rate": 5.460025534817911e-06,
+      "loss": 0.985,
+      "step": 6312
+    },
+    {
+      "epoch": 0.486663583102066,
+      "grad_norm": 3.498439073562622,
+      "learning_rate": 5.458782368488369e-06,
+      "loss": 1.0291,
+      "step": 6313
+    },
+    {
+      "epoch": 0.48674067221708295,
+      "grad_norm": 3.638658046722412,
+      "learning_rate": 5.4575391735561216e-06,
+      "loss": 0.9331,
+      "step": 6314
+    },
+    {
+      "epoch": 0.4868177613320999,
+      "grad_norm": 3.5828909873962402,
+      "learning_rate": 5.456295950098676e-06,
+      "loss": 1.0794,
+      "step": 6315
+    },
+    {
+      "epoch": 0.4868948504471169,
+      "grad_norm": 3.282355308532715,
+      "learning_rate": 5.45505269819354e-06,
+      "loss": 0.9155,
+      "step": 6316
+    },
+    {
+      "epoch": 0.4869719395621338,
+      "grad_norm": 3.461933135986328,
+      "learning_rate": 5.453809417918227e-06,
+      "loss": 0.9608,
+      "step": 6317
+    },
+    {
+      "epoch": 0.4870490286771508,
+      "grad_norm": 3.6630191802978516,
+      "learning_rate": 5.452566109350248e-06,
+      "loss": 1.0117,
+      "step": 6318
+    },
+    {
+      "epoch": 0.48712611779216775,
+      "grad_norm": 3.436558485031128,
+      "learning_rate": 5.451322772567114e-06,
+      "loss": 0.8844,
+      "step": 6319
+    },
+    {
+      "epoch": 0.4872032069071847,
+      "grad_norm": 3.7390990257263184,
+      "learning_rate": 5.450079407646343e-06,
+      "loss": 1.0276,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4872802960222017,
+      "grad_norm": 3.4452319145202637,
+      "learning_rate": 5.448836014665453e-06,
+      "loss": 0.8985,
+      "step": 6321
+    },
+    {
+      "epoch": 0.4873573851372186,
+      "grad_norm": 3.383016347885132,
+      "learning_rate": 5.447592593701961e-06,
+      "loss": 0.969,
+      "step": 6322
+    },
+    {
+      "epoch": 0.48743447425223557,
+      "grad_norm": 3.7448880672454834,
+      "learning_rate": 5.446349144833389e-06,
+      "loss": 1.0328,
+      "step": 6323
+    },
+    {
+      "epoch": 0.48751156336725254,
+      "grad_norm": 3.59957218170166,
+      "learning_rate": 5.44510566813726e-06,
+      "loss": 1.0494,
+      "step": 6324
+    },
+    {
+      "epoch": 0.4875886524822695,
+      "grad_norm": 3.316612720489502,
+      "learning_rate": 5.443862163691097e-06,
+      "loss": 1.0422,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4876657415972865,
+      "grad_norm": 3.4998679161071777,
+      "learning_rate": 5.442618631572428e-06,
+      "loss": 0.9016,
+      "step": 6326
+    },
+    {
+      "epoch": 0.4877428307123034,
+      "grad_norm": 3.8718786239624023,
+      "learning_rate": 5.44137507185878e-06,
+      "loss": 0.9931,
+      "step": 6327
+    },
+    {
+      "epoch": 0.48781991982732037,
+      "grad_norm": 3.7852871417999268,
+      "learning_rate": 5.440131484627681e-06,
+      "loss": 0.9991,
+      "step": 6328
+    },
+    {
+      "epoch": 0.48789700894233734,
+      "grad_norm": 3.4875009059906006,
+      "learning_rate": 5.438887869956664e-06,
+      "loss": 0.9499,
+      "step": 6329
+    },
+    {
+      "epoch": 0.4879740980573543,
+      "grad_norm": 3.2732343673706055,
+      "learning_rate": 5.437644227923261e-06,
+      "loss": 1.0069,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4880511871723713,
+      "grad_norm": 3.7164502143859863,
+      "learning_rate": 5.4364005586050075e-06,
+      "loss": 1.0977,
+      "step": 6331
+    },
+    {
+      "epoch": 0.4881282762873882,
+      "grad_norm": 3.7878737449645996,
+      "learning_rate": 5.4351568620794395e-06,
+      "loss": 1.0551,
+      "step": 6332
+    },
+    {
+      "epoch": 0.48820536540240517,
+      "grad_norm": 3.5380685329437256,
+      "learning_rate": 5.433913138424094e-06,
+      "loss": 0.8823,
+      "step": 6333
+    },
+    {
+      "epoch": 0.48828245451742214,
+      "grad_norm": 3.8582963943481445,
+      "learning_rate": 5.4326693877165125e-06,
+      "loss": 0.9363,
+      "step": 6334
+    },
+    {
+      "epoch": 0.4883595436324391,
+      "grad_norm": 3.4837450981140137,
+      "learning_rate": 5.431425610034235e-06,
+      "loss": 0.8931,
+      "step": 6335
+    },
+    {
+      "epoch": 0.4884366327474561,
+      "grad_norm": 4.541477680206299,
+      "learning_rate": 5.430181805454805e-06,
+      "loss": 0.9786,
+      "step": 6336
+    },
+    {
+      "epoch": 0.488513721862473,
+      "grad_norm": 3.627601146697998,
+      "learning_rate": 5.428937974055769e-06,
+      "loss": 0.9365,
+      "step": 6337
+    },
+    {
+      "epoch": 0.48859081097748996,
+      "grad_norm": 3.725983142852783,
+      "learning_rate": 5.427694115914669e-06,
+      "loss": 0.9854,
+      "step": 6338
+    },
+    {
+      "epoch": 0.48866790009250693,
+      "grad_norm": 3.6275992393493652,
+      "learning_rate": 5.426450231109058e-06,
+      "loss": 0.9704,
+      "step": 6339
+    },
+    {
+      "epoch": 0.4887449892075239,
+      "grad_norm": 3.6166484355926514,
+      "learning_rate": 5.425206319716483e-06,
+      "loss": 1.0062,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4888220783225409,
+      "grad_norm": 3.833855152130127,
+      "learning_rate": 5.423962381814496e-06,
+      "loss": 0.983,
+      "step": 6341
+    },
+    {
+      "epoch": 0.4888991674375578,
+      "grad_norm": 3.7868459224700928,
+      "learning_rate": 5.422718417480651e-06,
+      "loss": 0.9064,
+      "step": 6342
+    },
+    {
+      "epoch": 0.48897625655257476,
+      "grad_norm": 3.3974921703338623,
+      "learning_rate": 5.421474426792501e-06,
+      "loss": 0.965,
+      "step": 6343
+    },
+    {
+      "epoch": 0.48905334566759173,
+      "grad_norm": 3.812864065170288,
+      "learning_rate": 5.420230409827604e-06,
+      "loss": 1.0489,
+      "step": 6344
+    },
+    {
+      "epoch": 0.4891304347826087,
+      "grad_norm": 3.837310552597046,
+      "learning_rate": 5.418986366663518e-06,
+      "loss": 0.9851,
+      "step": 6345
+    },
+    {
+      "epoch": 0.4892075238976257,
+      "grad_norm": 3.419736385345459,
+      "learning_rate": 5.4177422973778015e-06,
+      "loss": 0.9357,
+      "step": 6346
+    },
+    {
+      "epoch": 0.4892846130126426,
+      "grad_norm": 3.845294237136841,
+      "learning_rate": 5.416498202048016e-06,
+      "loss": 1.0626,
+      "step": 6347
+    },
+    {
+      "epoch": 0.48936170212765956,
+      "grad_norm": 3.5029916763305664,
+      "learning_rate": 5.415254080751725e-06,
+      "loss": 0.9745,
+      "step": 6348
+    },
+    {
+      "epoch": 0.48943879124267653,
+      "grad_norm": 3.959080696105957,
+      "learning_rate": 5.414009933566492e-06,
+      "loss": 1.0339,
+      "step": 6349
+    },
+    {
+      "epoch": 0.4895158803576935,
+      "grad_norm": 3.6839077472686768,
+      "learning_rate": 5.412765760569886e-06,
+      "loss": 0.9517,
+      "step": 6350
+    },
+    {
+      "epoch": 0.48959296947271047,
+      "grad_norm": 3.2615315914154053,
+      "learning_rate": 5.41152156183947e-06,
+      "loss": 0.8866,
+      "step": 6351
+    },
+    {
+      "epoch": 0.4896700585877274,
+      "grad_norm": 3.5984644889831543,
+      "learning_rate": 5.410277337452817e-06,
+      "loss": 0.9906,
+      "step": 6352
+    },
+    {
+      "epoch": 0.48974714770274436,
+      "grad_norm": 3.3124704360961914,
+      "learning_rate": 5.409033087487498e-06,
+      "loss": 0.8656,
+      "step": 6353
+    },
+    {
+      "epoch": 0.4898242368177613,
+      "grad_norm": 3.2847061157226562,
+      "learning_rate": 5.407788812021082e-06,
+      "loss": 0.8378,
+      "step": 6354
+    },
+    {
+      "epoch": 0.4899013259327783,
+      "grad_norm": 3.559105157852173,
+      "learning_rate": 5.406544511131146e-06,
+      "loss": 1.0279,
+      "step": 6355
+    },
+    {
+      "epoch": 0.48997841504779527,
+      "grad_norm": 3.41558575630188,
+      "learning_rate": 5.405300184895268e-06,
+      "loss": 0.9121,
+      "step": 6356
+    },
+    {
+      "epoch": 0.4900555041628122,
+      "grad_norm": 3.4175400733947754,
+      "learning_rate": 5.40405583339102e-06,
+      "loss": 0.8206,
+      "step": 6357
+    },
+    {
+      "epoch": 0.49013259327782915,
+      "grad_norm": 3.7878329753875732,
+      "learning_rate": 5.402811456695985e-06,
+      "loss": 0.9255,
+      "step": 6358
+    },
+    {
+      "epoch": 0.4902096823928461,
+      "grad_norm": 3.816235303878784,
+      "learning_rate": 5.401567054887741e-06,
+      "loss": 1.1515,
+      "step": 6359
+    },
+    {
+      "epoch": 0.4902867715078631,
+      "grad_norm": 3.481640100479126,
+      "learning_rate": 5.400322628043869e-06,
+      "loss": 0.8833,
+      "step": 6360
+    },
+    {
+      "epoch": 0.49036386062288007,
+      "grad_norm": 3.5390591621398926,
+      "learning_rate": 5.399078176241958e-06,
+      "loss": 0.9331,
+      "step": 6361
+    },
+    {
+      "epoch": 0.49044094973789704,
+      "grad_norm": 3.5117104053497314,
+      "learning_rate": 5.397833699559587e-06,
+      "loss": 0.994,
+      "step": 6362
+    },
+    {
+      "epoch": 0.49051803885291395,
+      "grad_norm": 3.922240972518921,
+      "learning_rate": 5.396589198074347e-06,
+      "loss": 1.063,
+      "step": 6363
+    },
+    {
+      "epoch": 0.4905951279679309,
+      "grad_norm": 3.7044625282287598,
+      "learning_rate": 5.3953446718638235e-06,
+      "loss": 1.0509,
+      "step": 6364
+    },
+    {
+      "epoch": 0.4906722170829479,
+      "grad_norm": 3.929018259048462,
+      "learning_rate": 5.394100121005608e-06,
+      "loss": 1.0066,
+      "step": 6365
+    },
+    {
+      "epoch": 0.49074930619796486,
+      "grad_norm": 3.511359214782715,
+      "learning_rate": 5.39285554557729e-06,
+      "loss": 0.959,
+      "step": 6366
+    },
+    {
+      "epoch": 0.49082639531298183,
+      "grad_norm": 3.469261646270752,
+      "learning_rate": 5.391610945656464e-06,
+      "loss": 1.0333,
+      "step": 6367
+    },
+    {
+      "epoch": 0.49090348442799875,
+      "grad_norm": 3.302649736404419,
+      "learning_rate": 5.390366321320723e-06,
+      "loss": 0.9105,
+      "step": 6368
+    },
+    {
+      "epoch": 0.4909805735430157,
+      "grad_norm": 3.9019734859466553,
+      "learning_rate": 5.389121672647666e-06,
+      "loss": 1.1229,
+      "step": 6369
+    },
+    {
+      "epoch": 0.4910576626580327,
+      "grad_norm": 3.741459846496582,
+      "learning_rate": 5.387876999714885e-06,
+      "loss": 1.0295,
+      "step": 6370
+    },
+    {
+      "epoch": 0.49113475177304966,
+      "grad_norm": 3.3575656414031982,
+      "learning_rate": 5.386632302599985e-06,
+      "loss": 0.9632,
+      "step": 6371
+    },
+    {
+      "epoch": 0.49121184088806663,
+      "grad_norm": 3.4289495944976807,
+      "learning_rate": 5.385387581380561e-06,
+      "loss": 0.9537,
+      "step": 6372
+    },
+    {
+      "epoch": 0.49128893000308355,
+      "grad_norm": 3.718967914581299,
+      "learning_rate": 5.384142836134217e-06,
+      "loss": 0.953,
+      "step": 6373
+    },
+    {
+      "epoch": 0.4913660191181005,
+      "grad_norm": 3.8961894512176514,
+      "learning_rate": 5.382898066938559e-06,
+      "loss": 0.9896,
+      "step": 6374
+    },
+    {
+      "epoch": 0.4914431082331175,
+      "grad_norm": 3.611521005630493,
+      "learning_rate": 5.381653273871188e-06,
+      "loss": 1.0437,
+      "step": 6375
+    },
+    {
+      "epoch": 0.49152019734813446,
+      "grad_norm": 3.494535446166992,
+      "learning_rate": 5.380408457009711e-06,
+      "loss": 0.9011,
+      "step": 6376
+    },
+    {
+      "epoch": 0.49159728646315143,
+      "grad_norm": 3.385871648788452,
+      "learning_rate": 5.37916361643174e-06,
+      "loss": 0.9788,
+      "step": 6377
+    },
+    {
+      "epoch": 0.49167437557816834,
+      "grad_norm": 3.541168212890625,
+      "learning_rate": 5.377918752214878e-06,
+      "loss": 0.9361,
+      "step": 6378
+    },
+    {
+      "epoch": 0.4917514646931853,
+      "grad_norm": 3.743074417114258,
+      "learning_rate": 5.37667386443674e-06,
+      "loss": 0.9046,
+      "step": 6379
+    },
+    {
+      "epoch": 0.4918285538082023,
+      "grad_norm": 3.602600574493408,
+      "learning_rate": 5.375428953174939e-06,
+      "loss": 1.0171,
+      "step": 6380
+    },
+    {
+      "epoch": 0.49190564292321926,
+      "grad_norm": 3.6449949741363525,
+      "learning_rate": 5.374184018507086e-06,
+      "loss": 1.0,
+      "step": 6381
+    },
+    {
+      "epoch": 0.4919827320382362,
+      "grad_norm": 3.660562515258789,
+      "learning_rate": 5.372939060510797e-06,
+      "loss": 1.0855,
+      "step": 6382
+    },
+    {
+      "epoch": 0.49205982115325314,
+      "grad_norm": 3.2304728031158447,
+      "learning_rate": 5.371694079263688e-06,
+      "loss": 0.8088,
+      "step": 6383
+    },
+    {
+      "epoch": 0.4921369102682701,
+      "grad_norm": 3.455740451812744,
+      "learning_rate": 5.37044907484338e-06,
+      "loss": 0.9095,
+      "step": 6384
+    },
+    {
+      "epoch": 0.4922139993832871,
+      "grad_norm": 3.787229061126709,
+      "learning_rate": 5.369204047327491e-06,
+      "loss": 0.9225,
+      "step": 6385
+    },
+    {
+      "epoch": 0.49229108849830405,
+      "grad_norm": 3.7648141384124756,
+      "learning_rate": 5.367958996793641e-06,
+      "loss": 0.9692,
+      "step": 6386
+    },
+    {
+      "epoch": 0.492368177613321,
+      "grad_norm": 3.6073803901672363,
+      "learning_rate": 5.366713923319455e-06,
+      "loss": 0.9865,
+      "step": 6387
+    },
+    {
+      "epoch": 0.49244526672833794,
+      "grad_norm": 3.283367872238159,
+      "learning_rate": 5.365468826982553e-06,
+      "loss": 0.991,
+      "step": 6388
+    },
+    {
+      "epoch": 0.4925223558433549,
+      "grad_norm": 3.9354822635650635,
+      "learning_rate": 5.364223707860563e-06,
+      "loss": 0.9923,
+      "step": 6389
+    },
+    {
+      "epoch": 0.4925994449583719,
+      "grad_norm": 3.2957966327667236,
+      "learning_rate": 5.362978566031112e-06,
+      "loss": 0.9225,
+      "step": 6390
+    },
+    {
+      "epoch": 0.49267653407338885,
+      "grad_norm": 3.9109017848968506,
+      "learning_rate": 5.361733401571826e-06,
+      "loss": 0.8456,
+      "step": 6391
+    },
+    {
+      "epoch": 0.4927536231884058,
+      "grad_norm": 3.773970603942871,
+      "learning_rate": 5.360488214560336e-06,
+      "loss": 0.9238,
+      "step": 6392
+    },
+    {
+      "epoch": 0.49283071230342274,
+      "grad_norm": 3.3572847843170166,
+      "learning_rate": 5.359243005074274e-06,
+      "loss": 0.9182,
+      "step": 6393
+    },
+    {
+      "epoch": 0.4929078014184397,
+      "grad_norm": 3.6737236976623535,
+      "learning_rate": 5.35799777319127e-06,
+      "loss": 0.994,
+      "step": 6394
+    },
+    {
+      "epoch": 0.4929848905334567,
+      "grad_norm": 4.093838214874268,
+      "learning_rate": 5.356752518988961e-06,
+      "loss": 1.013,
+      "step": 6395
+    },
+    {
+      "epoch": 0.49306197964847365,
+      "grad_norm": 3.745969295501709,
+      "learning_rate": 5.355507242544978e-06,
+      "loss": 0.9943,
+      "step": 6396
+    },
+    {
+      "epoch": 0.4931390687634906,
+      "grad_norm": 3.748204469680786,
+      "learning_rate": 5.35426194393696e-06,
+      "loss": 0.8421,
+      "step": 6397
+    },
+    {
+      "epoch": 0.49321615787850753,
+      "grad_norm": 3.6849241256713867,
+      "learning_rate": 5.3530166232425454e-06,
+      "loss": 0.8745,
+      "step": 6398
+    },
+    {
+      "epoch": 0.4932932469935245,
+      "grad_norm": 3.651228189468384,
+      "learning_rate": 5.351771280539372e-06,
+      "loss": 1.0544,
+      "step": 6399
+    },
+    {
+      "epoch": 0.4933703361085415,
+      "grad_norm": 4.058038234710693,
+      "learning_rate": 5.35052591590508e-06,
+      "loss": 1.0527,
+      "step": 6400
+    },
+    {
+      "epoch": 0.49344742522355844,
+      "grad_norm": 4.054307460784912,
+      "learning_rate": 5.349280529417316e-06,
+      "loss": 0.9891,
+      "step": 6401
+    },
+    {
+      "epoch": 0.4935245143385754,
+      "grad_norm": 3.8959100246429443,
+      "learning_rate": 5.348035121153716e-06,
+      "loss": 1.0433,
+      "step": 6402
+    },
+    {
+      "epoch": 0.49360160345359233,
+      "grad_norm": 3.7364206314086914,
+      "learning_rate": 5.346789691191931e-06,
+      "loss": 0.9561,
+      "step": 6403
+    },
+    {
+      "epoch": 0.4936786925686093,
+      "grad_norm": 4.071012496948242,
+      "learning_rate": 5.3455442396096045e-06,
+      "loss": 0.9654,
+      "step": 6404
+    },
+    {
+      "epoch": 0.49375578168362627,
+      "grad_norm": 3.570204019546509,
+      "learning_rate": 5.344298766484382e-06,
+      "loss": 0.9465,
+      "step": 6405
+    },
+    {
+      "epoch": 0.49383287079864324,
+      "grad_norm": 3.7908997535705566,
+      "learning_rate": 5.343053271893919e-06,
+      "loss": 1.0456,
+      "step": 6406
+    },
+    {
+      "epoch": 0.4939099599136602,
+      "grad_norm": 3.9499940872192383,
+      "learning_rate": 5.3418077559158575e-06,
+      "loss": 1.0423,
+      "step": 6407
+    },
+    {
+      "epoch": 0.4939870490286771,
+      "grad_norm": 3.3631200790405273,
+      "learning_rate": 5.340562218627854e-06,
+      "loss": 0.9541,
+      "step": 6408
+    },
+    {
+      "epoch": 0.4940641381436941,
+      "grad_norm": 3.548306941986084,
+      "learning_rate": 5.339316660107561e-06,
+      "loss": 0.9153,
+      "step": 6409
+    },
+    {
+      "epoch": 0.49414122725871107,
+      "grad_norm": 3.9176387786865234,
+      "learning_rate": 5.3380710804326304e-06,
+      "loss": 0.9603,
+      "step": 6410
+    },
+    {
+      "epoch": 0.49421831637372804,
+      "grad_norm": 4.336452484130859,
+      "learning_rate": 5.33682547968072e-06,
+      "loss": 1.0792,
+      "step": 6411
+    },
+    {
+      "epoch": 0.494295405488745,
+      "grad_norm": 3.4913458824157715,
+      "learning_rate": 5.3355798579294834e-06,
+      "loss": 0.9811,
+      "step": 6412
+    },
+    {
+      "epoch": 0.4943724946037619,
+      "grad_norm": 3.5579373836517334,
+      "learning_rate": 5.334334215256582e-06,
+      "loss": 0.9181,
+      "step": 6413
+    },
+    {
+      "epoch": 0.4944495837187789,
+      "grad_norm": 3.506436586380005,
+      "learning_rate": 5.333088551739674e-06,
+      "loss": 0.9107,
+      "step": 6414
+    },
+    {
+      "epoch": 0.49452667283379587,
+      "grad_norm": 3.682651996612549,
+      "learning_rate": 5.3318428674564196e-06,
+      "loss": 0.989,
+      "step": 6415
+    },
+    {
+      "epoch": 0.49460376194881284,
+      "grad_norm": 3.929532527923584,
+      "learning_rate": 5.330597162484481e-06,
+      "loss": 0.8853,
+      "step": 6416
+    },
+    {
+      "epoch": 0.4946808510638298,
+      "grad_norm": 3.4837441444396973,
+      "learning_rate": 5.329351436901522e-06,
+      "loss": 0.904,
+      "step": 6417
+    },
+    {
+      "epoch": 0.4947579401788467,
+      "grad_norm": 3.567117691040039,
+      "learning_rate": 5.3281056907852054e-06,
+      "loss": 0.9982,
+      "step": 6418
+    },
+    {
+      "epoch": 0.4948350292938637,
+      "grad_norm": 3.655592918395996,
+      "learning_rate": 5.3268599242132e-06,
+      "loss": 0.9322,
+      "step": 6419
+    },
+    {
+      "epoch": 0.49491211840888066,
+      "grad_norm": 3.478694200515747,
+      "learning_rate": 5.32561413726317e-06,
+      "loss": 0.9695,
+      "step": 6420
+    },
+    {
+      "epoch": 0.49498920752389763,
+      "grad_norm": 3.512316942214966,
+      "learning_rate": 5.324368330012785e-06,
+      "loss": 0.9506,
+      "step": 6421
+    },
+    {
+      "epoch": 0.4950662966389146,
+      "grad_norm": 3.5826921463012695,
+      "learning_rate": 5.323122502539715e-06,
+      "loss": 0.9291,
+      "step": 6422
+    },
+    {
+      "epoch": 0.4951433857539315,
+      "grad_norm": 3.367875337600708,
+      "learning_rate": 5.32187665492163e-06,
+      "loss": 1.0068,
+      "step": 6423
+    },
+    {
+      "epoch": 0.4952204748689485,
+      "grad_norm": 3.737788200378418,
+      "learning_rate": 5.320630787236203e-06,
+      "loss": 1.0055,
+      "step": 6424
+    },
+    {
+      "epoch": 0.49529756398396546,
+      "grad_norm": 3.3039023876190186,
+      "learning_rate": 5.3193848995611075e-06,
+      "loss": 1.0002,
+      "step": 6425
+    },
+    {
+      "epoch": 0.49537465309898243,
+      "grad_norm": 3.7930006980895996,
+      "learning_rate": 5.318138991974016e-06,
+      "loss": 1.1429,
+      "step": 6426
+    },
+    {
+      "epoch": 0.4954517422139994,
+      "grad_norm": 3.4855892658233643,
+      "learning_rate": 5.316893064552607e-06,
+      "loss": 0.8802,
+      "step": 6427
+    },
+    {
+      "epoch": 0.4955288313290163,
+      "grad_norm": 3.870096206665039,
+      "learning_rate": 5.315647117374556e-06,
+      "loss": 1.0267,
+      "step": 6428
+    },
+    {
+      "epoch": 0.4956059204440333,
+      "grad_norm": 3.2889981269836426,
+      "learning_rate": 5.314401150517543e-06,
+      "loss": 0.9023,
+      "step": 6429
+    },
+    {
+      "epoch": 0.49568300955905026,
+      "grad_norm": 3.54913330078125,
+      "learning_rate": 5.313155164059247e-06,
+      "loss": 1.023,
+      "step": 6430
+    },
+    {
+      "epoch": 0.49576009867406723,
+      "grad_norm": 3.7795615196228027,
+      "learning_rate": 5.311909158077347e-06,
+      "loss": 1.0105,
+      "step": 6431
+    },
+    {
+      "epoch": 0.4958371877890842,
+      "grad_norm": 3.661637306213379,
+      "learning_rate": 5.310663132649526e-06,
+      "loss": 1.0102,
+      "step": 6432
+    },
+    {
+      "epoch": 0.4959142769041011,
+      "grad_norm": 3.3759381771087646,
+      "learning_rate": 5.30941708785347e-06,
+      "loss": 0.88,
+      "step": 6433
+    },
+    {
+      "epoch": 0.4959913660191181,
+      "grad_norm": 3.458970546722412,
+      "learning_rate": 5.3081710237668595e-06,
+      "loss": 0.9249,
+      "step": 6434
+    },
+    {
+      "epoch": 0.49606845513413506,
+      "grad_norm": 3.838961362838745,
+      "learning_rate": 5.306924940467383e-06,
+      "loss": 1.082,
+      "step": 6435
+    },
+    {
+      "epoch": 0.496145544249152,
+      "grad_norm": 3.4712164402008057,
+      "learning_rate": 5.3056788380327255e-06,
+      "loss": 0.8727,
+      "step": 6436
+    },
+    {
+      "epoch": 0.496222633364169,
+      "grad_norm": 3.2734153270721436,
+      "learning_rate": 5.304432716540574e-06,
+      "loss": 0.8984,
+      "step": 6437
+    },
+    {
+      "epoch": 0.4962997224791859,
+      "grad_norm": 3.8021316528320312,
+      "learning_rate": 5.303186576068621e-06,
+      "loss": 1.051,
+      "step": 6438
+    },
+    {
+      "epoch": 0.4963768115942029,
+      "grad_norm": 3.4965784549713135,
+      "learning_rate": 5.301940416694554e-06,
+      "loss": 0.8909,
+      "step": 6439
+    },
+    {
+      "epoch": 0.49645390070921985,
+      "grad_norm": 3.5403904914855957,
+      "learning_rate": 5.300694238496067e-06,
+      "loss": 0.9986,
+      "step": 6440
+    },
+    {
+      "epoch": 0.4965309898242368,
+      "grad_norm": 3.335237503051758,
+      "learning_rate": 5.299448041550852e-06,
+      "loss": 0.9258,
+      "step": 6441
+    },
+    {
+      "epoch": 0.4966080789392538,
+      "grad_norm": 3.8967440128326416,
+      "learning_rate": 5.2982018259366004e-06,
+      "loss": 1.0064,
+      "step": 6442
+    },
+    {
+      "epoch": 0.4966851680542707,
+      "grad_norm": 3.5903615951538086,
+      "learning_rate": 5.296955591731012e-06,
+      "loss": 0.9383,
+      "step": 6443
+    },
+    {
+      "epoch": 0.4967622571692877,
+      "grad_norm": 3.501309871673584,
+      "learning_rate": 5.295709339011779e-06,
+      "loss": 0.895,
+      "step": 6444
+    },
+    {
+      "epoch": 0.49683934628430465,
+      "grad_norm": 3.6556556224823,
+      "learning_rate": 5.2944630678566e-06,
+      "loss": 0.9801,
+      "step": 6445
+    },
+    {
+      "epoch": 0.4969164353993216,
+      "grad_norm": 3.3892955780029297,
+      "learning_rate": 5.293216778343175e-06,
+      "loss": 0.8553,
+      "step": 6446
+    },
+    {
+      "epoch": 0.4969935245143386,
+      "grad_norm": 3.6636784076690674,
+      "learning_rate": 5.2919704705492e-06,
+      "loss": 0.9744,
+      "step": 6447
+    },
+    {
+      "epoch": 0.49707061362935556,
+      "grad_norm": 3.7394680976867676,
+      "learning_rate": 5.290724144552379e-06,
+      "loss": 0.9662,
+      "step": 6448
+    },
+    {
+      "epoch": 0.4971477027443725,
+      "grad_norm": 3.7077627182006836,
+      "learning_rate": 5.2894778004304146e-06,
+      "loss": 0.8284,
+      "step": 6449
+    },
+    {
+      "epoch": 0.49722479185938945,
+      "grad_norm": 3.826967716217041,
+      "learning_rate": 5.288231438261008e-06,
+      "loss": 0.8625,
+      "step": 6450
+    },
+    {
+      "epoch": 0.4973018809744064,
+      "grad_norm": 3.9576635360717773,
+      "learning_rate": 5.286985058121865e-06,
+      "loss": 0.9651,
+      "step": 6451
+    },
+    {
+      "epoch": 0.4973789700894234,
+      "grad_norm": 3.6830098628997803,
+      "learning_rate": 5.285738660090688e-06,
+      "loss": 1.1414,
+      "step": 6452
+    },
+    {
+      "epoch": 0.49745605920444036,
+      "grad_norm": 3.410430908203125,
+      "learning_rate": 5.284492244245187e-06,
+      "loss": 0.8979,
+      "step": 6453
+    },
+    {
+      "epoch": 0.4975331483194573,
+      "grad_norm": 3.495670795440674,
+      "learning_rate": 5.283245810663068e-06,
+      "loss": 0.8779,
+      "step": 6454
+    },
+    {
+      "epoch": 0.49761023743447425,
+      "grad_norm": 3.549842596054077,
+      "learning_rate": 5.281999359422039e-06,
+      "loss": 0.8665,
+      "step": 6455
+    },
+    {
+      "epoch": 0.4976873265494912,
+      "grad_norm": 3.7550711631774902,
+      "learning_rate": 5.28075289059981e-06,
+      "loss": 0.9765,
+      "step": 6456
+    },
+    {
+      "epoch": 0.4977644156645082,
+      "grad_norm": 3.494971990585327,
+      "learning_rate": 5.279506404274094e-06,
+      "loss": 0.9499,
+      "step": 6457
+    },
+    {
+      "epoch": 0.49784150477952516,
+      "grad_norm": 3.89752197265625,
+      "learning_rate": 5.2782599005226e-06,
+      "loss": 1.0034,
+      "step": 6458
+    },
+    {
+      "epoch": 0.49791859389454207,
+      "grad_norm": 3.6488029956817627,
+      "learning_rate": 5.277013379423043e-06,
+      "loss": 1.037,
+      "step": 6459
+    },
+    {
+      "epoch": 0.49799568300955904,
+      "grad_norm": 3.672584295272827,
+      "learning_rate": 5.275766841053136e-06,
+      "loss": 1.0282,
+      "step": 6460
+    },
+    {
+      "epoch": 0.498072772124576,
+      "grad_norm": 3.7023403644561768,
+      "learning_rate": 5.2745202854905946e-06,
+      "loss": 0.9883,
+      "step": 6461
+    },
+    {
+      "epoch": 0.498149861239593,
+      "grad_norm": 3.654310941696167,
+      "learning_rate": 5.273273712813135e-06,
+      "loss": 0.9355,
+      "step": 6462
+    },
+    {
+      "epoch": 0.49822695035460995,
+      "grad_norm": 3.7347729206085205,
+      "learning_rate": 5.272027123098475e-06,
+      "loss": 1.0296,
+      "step": 6463
+    },
+    {
+      "epoch": 0.49830403946962687,
+      "grad_norm": 3.6972110271453857,
+      "learning_rate": 5.2707805164243335e-06,
+      "loss": 0.8939,
+      "step": 6464
+    },
+    {
+      "epoch": 0.49838112858464384,
+      "grad_norm": 3.7917075157165527,
+      "learning_rate": 5.269533892868428e-06,
+      "loss": 0.959,
+      "step": 6465
+    },
+    {
+      "epoch": 0.4984582176996608,
+      "grad_norm": 3.710469961166382,
+      "learning_rate": 5.268287252508481e-06,
+      "loss": 1.0708,
+      "step": 6466
+    },
+    {
+      "epoch": 0.4985353068146778,
+      "grad_norm": 3.4452409744262695,
+      "learning_rate": 5.2670405954222134e-06,
+      "loss": 0.9258,
+      "step": 6467
+    },
+    {
+      "epoch": 0.49861239592969475,
+      "grad_norm": 3.4874088764190674,
+      "learning_rate": 5.265793921687347e-06,
+      "loss": 0.9412,
+      "step": 6468
+    },
+    {
+      "epoch": 0.49868948504471167,
+      "grad_norm": 3.5041942596435547,
+      "learning_rate": 5.264547231381606e-06,
+      "loss": 0.8923,
+      "step": 6469
+    },
+    {
+      "epoch": 0.49876657415972864,
+      "grad_norm": 3.852966785430908,
+      "learning_rate": 5.263300524582717e-06,
+      "loss": 1.0018,
+      "step": 6470
+    },
+    {
+      "epoch": 0.4988436632747456,
+      "grad_norm": 3.813434600830078,
+      "learning_rate": 5.2620538013684005e-06,
+      "loss": 1.0497,
+      "step": 6471
+    },
+    {
+      "epoch": 0.4989207523897626,
+      "grad_norm": 3.9356420040130615,
+      "learning_rate": 5.26080706181639e-06,
+      "loss": 0.9555,
+      "step": 6472
+    },
+    {
+      "epoch": 0.49899784150477955,
+      "grad_norm": 3.891258716583252,
+      "learning_rate": 5.259560306004409e-06,
+      "loss": 0.9037,
+      "step": 6473
+    },
+    {
+      "epoch": 0.49907493061979646,
+      "grad_norm": 3.454547643661499,
+      "learning_rate": 5.258313534010187e-06,
+      "loss": 0.8702,
+      "step": 6474
+    },
+    {
+      "epoch": 0.49915201973481343,
+      "grad_norm": 3.695034980773926,
+      "learning_rate": 5.257066745911454e-06,
+      "loss": 0.8333,
+      "step": 6475
+    },
+    {
+      "epoch": 0.4992291088498304,
+      "grad_norm": 3.522186040878296,
+      "learning_rate": 5.2558199417859415e-06,
+      "loss": 0.948,
+      "step": 6476
+    },
+    {
+      "epoch": 0.4993061979648474,
+      "grad_norm": 3.7446117401123047,
+      "learning_rate": 5.25457312171138e-06,
+      "loss": 0.9535,
+      "step": 6477
+    },
+    {
+      "epoch": 0.49938328707986435,
+      "grad_norm": 3.8174216747283936,
+      "learning_rate": 5.253326285765502e-06,
+      "loss": 0.9085,
+      "step": 6478
+    },
+    {
+      "epoch": 0.49946037619488126,
+      "grad_norm": 3.598435640335083,
+      "learning_rate": 5.252079434026043e-06,
+      "loss": 0.9694,
+      "step": 6479
+    },
+    {
+      "epoch": 0.49953746530989823,
+      "grad_norm": 4.055591583251953,
+      "learning_rate": 5.250832566570736e-06,
+      "loss": 0.998,
+      "step": 6480
+    },
+    {
+      "epoch": 0.4996145544249152,
+      "grad_norm": 3.655466318130493,
+      "learning_rate": 5.2495856834773195e-06,
+      "loss": 0.8998,
+      "step": 6481
+    },
+    {
+      "epoch": 0.4996916435399322,
+      "grad_norm": 3.346543788909912,
+      "learning_rate": 5.248338784823526e-06,
+      "loss": 0.9427,
+      "step": 6482
+    },
+    {
+      "epoch": 0.49976873265494914,
+      "grad_norm": 3.703274965286255,
+      "learning_rate": 5.2470918706870975e-06,
+      "loss": 0.8559,
+      "step": 6483
+    },
+    {
+      "epoch": 0.49984582176996606,
+      "grad_norm": 3.716444492340088,
+      "learning_rate": 5.245844941145769e-06,
+      "loss": 1.017,
+      "step": 6484
+    },
+    {
+      "epoch": 0.49992291088498303,
+      "grad_norm": 3.966560125350952,
+      "learning_rate": 5.244597996277282e-06,
+      "loss": 0.8255,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.69476318359375,
+      "learning_rate": 5.243351036159377e-06,
+      "loss": 0.8626,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5000770891150169,
+      "grad_norm": 3.7048497200012207,
+      "learning_rate": 5.242104060869796e-06,
+      "loss": 0.9411,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5001541782300339,
+      "grad_norm": 3.5426058769226074,
+      "learning_rate": 5.2408570704862795e-06,
+      "loss": 0.9875,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5002312673450509,
+      "grad_norm": 3.579333782196045,
+      "learning_rate": 5.239610065086574e-06,
+      "loss": 1.0032,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5003083564600679,
+      "grad_norm": 3.5092763900756836,
+      "learning_rate": 5.238363044748419e-06,
+      "loss": 0.8675,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5003854455750848,
+      "grad_norm": 4.748165130615234,
+      "learning_rate": 5.237116009549565e-06,
+      "loss": 1.0171,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5004625346901017,
+      "grad_norm": 3.549006938934326,
+      "learning_rate": 5.235868959567755e-06,
+      "loss": 0.968,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5005396238051187,
+      "grad_norm": 4.191278457641602,
+      "learning_rate": 5.2346218948807345e-06,
+      "loss": 0.9313,
+      "step": 6493
+    },
+    {
+      "epoch": 0.5006167129201357,
+      "grad_norm": 3.531649589538574,
+      "learning_rate": 5.233374815566258e-06,
+      "loss": 0.8907,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5006938020351527,
+      "grad_norm": 3.5904295444488525,
+      "learning_rate": 5.232127721702069e-06,
+      "loss": 0.9111,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5007708911501696,
+      "grad_norm": 3.7466869354248047,
+      "learning_rate": 5.230880613365918e-06,
+      "loss": 0.9371,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5008479802651865,
+      "grad_norm": 3.7517306804656982,
+      "learning_rate": 5.229633490635558e-06,
+      "loss": 1.0557,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5009250693802035,
+      "grad_norm": 3.6371867656707764,
+      "learning_rate": 5.228386353588737e-06,
+      "loss": 1.0518,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5010021584952205,
+      "grad_norm": 3.3489437103271484,
+      "learning_rate": 5.2271392023032115e-06,
+      "loss": 0.9757,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5010792476102375,
+      "grad_norm": 3.531639575958252,
+      "learning_rate": 5.225892036856734e-06,
+      "loss": 0.9355,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5011563367252544,
+      "grad_norm": 3.9964072704315186,
+      "learning_rate": 5.224644857327055e-06,
+      "loss": 0.9836,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5012334258402713,
+      "grad_norm": 3.8607258796691895,
+      "learning_rate": 5.223397663791935e-06,
+      "loss": 1.0546,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5013105149552883,
+      "grad_norm": 3.4289846420288086,
+      "learning_rate": 5.222150456329127e-06,
+      "loss": 0.8784,
+      "step": 6503
+    },
+    {
+      "epoch": 0.5013876040703052,
+      "grad_norm": 3.9486873149871826,
+      "learning_rate": 5.220903235016388e-06,
+      "loss": 1.0408,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5014646931853223,
+      "grad_norm": 3.4474892616271973,
+      "learning_rate": 5.2196559999314765e-06,
+      "loss": 0.8746,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5015417823003392,
+      "grad_norm": 3.749389171600342,
+      "learning_rate": 5.218408751152152e-06,
+      "loss": 1.0857,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5016188714153561,
+      "grad_norm": 3.59549880027771,
+      "learning_rate": 5.217161488756172e-06,
+      "loss": 0.9574,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5016959605303731,
+      "grad_norm": 3.458582878112793,
+      "learning_rate": 5.215914212821298e-06,
+      "loss": 0.9595,
+      "step": 6508
+    },
+    {
+      "epoch": 0.50177304964539,
+      "grad_norm": 3.547104835510254,
+      "learning_rate": 5.214666923425291e-06,
+      "loss": 0.9317,
+      "step": 6509
+    },
+    {
+      "epoch": 0.5018501387604071,
+      "grad_norm": 3.983787775039673,
+      "learning_rate": 5.213419620645914e-06,
+      "loss": 1.096,
+      "step": 6510
+    },
+    {
+      "epoch": 0.501927227875424,
+      "grad_norm": 3.5246825218200684,
+      "learning_rate": 5.212172304560928e-06,
+      "loss": 0.9323,
+      "step": 6511
+    },
+    {
+      "epoch": 0.5020043169904409,
+      "grad_norm": 3.895063877105713,
+      "learning_rate": 5.2109249752480985e-06,
+      "loss": 0.9427,
+      "step": 6512
+    },
+    {
+      "epoch": 0.5020814061054579,
+      "grad_norm": 4.591753005981445,
+      "learning_rate": 5.20967763278519e-06,
+      "loss": 1.0114,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5021584952204748,
+      "grad_norm": 4.11314582824707,
+      "learning_rate": 5.208430277249965e-06,
+      "loss": 1.0685,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5022355843354919,
+      "grad_norm": 3.453099012374878,
+      "learning_rate": 5.207182908720192e-06,
+      "loss": 0.9276,
+      "step": 6515
+    },
+    {
+      "epoch": 0.5023126734505088,
+      "grad_norm": 3.538799285888672,
+      "learning_rate": 5.205935527273638e-06,
+      "loss": 0.9903,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5023897625655257,
+      "grad_norm": 3.893820285797119,
+      "learning_rate": 5.204688132988071e-06,
+      "loss": 0.8881,
+      "step": 6517
+    },
+    {
+      "epoch": 0.5024668516805427,
+      "grad_norm": 4.181878089904785,
+      "learning_rate": 5.203440725941259e-06,
+      "loss": 0.9407,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5025439407955596,
+      "grad_norm": 3.6423377990722656,
+      "learning_rate": 5.2021933062109705e-06,
+      "loss": 1.031,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5026210299105767,
+      "grad_norm": 3.723810911178589,
+      "learning_rate": 5.200945873874979e-06,
+      "loss": 1.0118,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5026981190255936,
+      "grad_norm": 3.465095043182373,
+      "learning_rate": 5.19969842901105e-06,
+      "loss": 0.942,
+      "step": 6521
+    },
+    {
+      "epoch": 0.5027752081406105,
+      "grad_norm": 3.6051387786865234,
+      "learning_rate": 5.198450971696959e-06,
+      "loss": 0.8803,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5028522972556275,
+      "grad_norm": 4.09804630279541,
+      "learning_rate": 5.197203502010478e-06,
+      "loss": 0.8432,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5029293863706444,
+      "grad_norm": 4.076051712036133,
+      "learning_rate": 5.19595602002938e-06,
+      "loss": 0.9762,
+      "step": 6524
+    },
+    {
+      "epoch": 0.5030064754856615,
+      "grad_norm": 3.7992618083953857,
+      "learning_rate": 5.194708525831439e-06,
+      "loss": 0.8912,
+      "step": 6525
+    },
+    {
+      "epoch": 0.5030835646006784,
+      "grad_norm": 3.59840989112854,
+      "learning_rate": 5.1934610194944306e-06,
+      "loss": 0.874,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5031606537156953,
+      "grad_norm": 3.5978050231933594,
+      "learning_rate": 5.192213501096129e-06,
+      "loss": 0.9612,
+      "step": 6527
+    },
+    {
+      "epoch": 0.5032377428307123,
+      "grad_norm": 3.7155649662017822,
+      "learning_rate": 5.1909659707143105e-06,
+      "loss": 1.042,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5033148319457292,
+      "grad_norm": 3.7677228450775146,
+      "learning_rate": 5.189718428426753e-06,
+      "loss": 1.0446,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5033919210607463,
+      "grad_norm": 3.759453773498535,
+      "learning_rate": 5.188470874311234e-06,
+      "loss": 1.0825,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5034690101757632,
+      "grad_norm": 3.622924327850342,
+      "learning_rate": 5.187223308445534e-06,
+      "loss": 0.9703,
+      "step": 6531
+    },
+    {
+      "epoch": 0.5035460992907801,
+      "grad_norm": 3.712109088897705,
+      "learning_rate": 5.185975730907428e-06,
+      "loss": 1.0215,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5036231884057971,
+      "grad_norm": 3.5897741317749023,
+      "learning_rate": 5.184728141774699e-06,
+      "loss": 0.919,
+      "step": 6533
+    },
+    {
+      "epoch": 0.503700277520814,
+      "grad_norm": 3.6697773933410645,
+      "learning_rate": 5.183480541125128e-06,
+      "loss": 1.0769,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5037773666358311,
+      "grad_norm": 3.3042874336242676,
+      "learning_rate": 5.182232929036495e-06,
+      "loss": 0.9561,
+      "step": 6535
+    },
+    {
+      "epoch": 0.503854455750848,
+      "grad_norm": 3.4754981994628906,
+      "learning_rate": 5.180985305586581e-06,
+      "loss": 0.9105,
+      "step": 6536
+    },
+    {
+      "epoch": 0.5039315448658649,
+      "grad_norm": 3.574561834335327,
+      "learning_rate": 5.179737670853173e-06,
+      "loss": 1.0152,
+      "step": 6537
+    },
+    {
+      "epoch": 0.5040086339808819,
+      "grad_norm": 3.8162965774536133,
+      "learning_rate": 5.17849002491405e-06,
+      "loss": 0.8418,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5040857230958988,
+      "grad_norm": 3.736490249633789,
+      "learning_rate": 5.177242367846999e-06,
+      "loss": 1.0162,
+      "step": 6539
+    },
+    {
+      "epoch": 0.5041628122109159,
+      "grad_norm": 3.601933240890503,
+      "learning_rate": 5.175994699729806e-06,
+      "loss": 0.9982,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5042399013259328,
+      "grad_norm": 4.0474443435668945,
+      "learning_rate": 5.174747020640253e-06,
+      "loss": 1.0384,
+      "step": 6541
+    },
+    {
+      "epoch": 0.5043169904409497,
+      "grad_norm": 3.8889949321746826,
+      "learning_rate": 5.17349933065613e-06,
+      "loss": 0.8445,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5043940795559667,
+      "grad_norm": 3.7362418174743652,
+      "learning_rate": 5.1722516298552206e-06,
+      "loss": 0.9535,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5044711686709836,
+      "grad_norm": 3.572842597961426,
+      "learning_rate": 5.171003918315316e-06,
+      "loss": 0.9203,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5045482577860007,
+      "grad_norm": 3.86114501953125,
+      "learning_rate": 5.169756196114202e-06,
+      "loss": 0.9648,
+      "step": 6545
+    },
+    {
+      "epoch": 0.5046253469010176,
+      "grad_norm": 3.9800467491149902,
+      "learning_rate": 5.1685084633296665e-06,
+      "loss": 0.9466,
+      "step": 6546
+    },
+    {
+      "epoch": 0.5047024360160345,
+      "grad_norm": 3.3460676670074463,
+      "learning_rate": 5.167260720039504e-06,
+      "loss": 0.8801,
+      "step": 6547
+    },
+    {
+      "epoch": 0.5047795251310515,
+      "grad_norm": 3.8302457332611084,
+      "learning_rate": 5.1660129663215e-06,
+      "loss": 1.0093,
+      "step": 6548
+    },
+    {
+      "epoch": 0.5048566142460684,
+      "grad_norm": 4.078864574432373,
+      "learning_rate": 5.164765202253448e-06,
+      "loss": 1.0053,
+      "step": 6549
+    },
+    {
+      "epoch": 0.5049337033610855,
+      "grad_norm": 3.9565751552581787,
+      "learning_rate": 5.163517427913139e-06,
+      "loss": 1.0029,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5050107924761024,
+      "grad_norm": 3.460613250732422,
+      "learning_rate": 5.162269643378365e-06,
+      "loss": 0.9372,
+      "step": 6551
+    },
+    {
+      "epoch": 0.5050878815911193,
+      "grad_norm": 3.854912757873535,
+      "learning_rate": 5.161021848726919e-06,
+      "loss": 0.9244,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5051649707061363,
+      "grad_norm": 3.741788148880005,
+      "learning_rate": 5.159774044036595e-06,
+      "loss": 0.9823,
+      "step": 6553
+    },
+    {
+      "epoch": 0.5052420598211532,
+      "grad_norm": 3.4191572666168213,
+      "learning_rate": 5.1585262293851865e-06,
+      "loss": 0.9462,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5053191489361702,
+      "grad_norm": 3.4478116035461426,
+      "learning_rate": 5.1572784048504894e-06,
+      "loss": 0.9788,
+      "step": 6555
+    },
+    {
+      "epoch": 0.5053962380511872,
+      "grad_norm": 3.257162094116211,
+      "learning_rate": 5.156030570510298e-06,
+      "loss": 0.8265,
+      "step": 6556
+    },
+    {
+      "epoch": 0.5054733271662041,
+      "grad_norm": 3.8727734088897705,
+      "learning_rate": 5.154782726442409e-06,
+      "loss": 0.9721,
+      "step": 6557
+    },
+    {
+      "epoch": 0.5055504162812211,
+      "grad_norm": 3.7539451122283936,
+      "learning_rate": 5.153534872724618e-06,
+      "loss": 0.9972,
+      "step": 6558
+    },
+    {
+      "epoch": 0.505627505396238,
+      "grad_norm": 3.6078336238861084,
+      "learning_rate": 5.152287009434723e-06,
+      "loss": 1.0208,
+      "step": 6559
+    },
+    {
+      "epoch": 0.505704594511255,
+      "grad_norm": 3.6347334384918213,
+      "learning_rate": 5.1510391366505204e-06,
+      "loss": 0.9033,
+      "step": 6560
+    },
+    {
+      "epoch": 0.505781683626272,
+      "grad_norm": 3.9092540740966797,
+      "learning_rate": 5.149791254449812e-06,
+      "loss": 0.9415,
+      "step": 6561
+    },
+    {
+      "epoch": 0.5058587727412889,
+      "grad_norm": 3.516451597213745,
+      "learning_rate": 5.148543362910393e-06,
+      "loss": 1.0037,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5059358618563059,
+      "grad_norm": 4.396605968475342,
+      "learning_rate": 5.147295462110066e-06,
+      "loss": 1.0104,
+      "step": 6563
+    },
+    {
+      "epoch": 0.5060129509713228,
+      "grad_norm": 4.63103723526001,
+      "learning_rate": 5.14604755212663e-06,
+      "loss": 1.0168,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5060900400863398,
+      "grad_norm": 3.5548949241638184,
+      "learning_rate": 5.144799633037884e-06,
+      "loss": 0.8663,
+      "step": 6565
+    },
+    {
+      "epoch": 0.5061671292013568,
+      "grad_norm": 3.333975315093994,
+      "learning_rate": 5.143551704921632e-06,
+      "loss": 0.8946,
+      "step": 6566
+    },
+    {
+      "epoch": 0.5062442183163737,
+      "grad_norm": 4.707441806793213,
+      "learning_rate": 5.142303767855674e-06,
+      "loss": 0.9074,
+      "step": 6567
+    },
+    {
+      "epoch": 0.5063213074313907,
+      "grad_norm": 3.944218397140503,
+      "learning_rate": 5.141055821917814e-06,
+      "loss": 0.9857,
+      "step": 6568
+    },
+    {
+      "epoch": 0.5063983965464076,
+      "grad_norm": 3.700016975402832,
+      "learning_rate": 5.139807867185853e-06,
+      "loss": 0.9898,
+      "step": 6569
+    },
+    {
+      "epoch": 0.5064754856614246,
+      "grad_norm": 3.7743771076202393,
+      "learning_rate": 5.138559903737596e-06,
+      "loss": 0.9928,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5065525747764416,
+      "grad_norm": 3.386918067932129,
+      "learning_rate": 5.137311931650847e-06,
+      "loss": 0.9561,
+      "step": 6571
+    },
+    {
+      "epoch": 0.5066296638914586,
+      "grad_norm": 3.41721510887146,
+      "learning_rate": 5.136063951003409e-06,
+      "loss": 0.9306,
+      "step": 6572
+    },
+    {
+      "epoch": 0.5067067530064755,
+      "grad_norm": 3.741901397705078,
+      "learning_rate": 5.134815961873089e-06,
+      "loss": 1.0163,
+      "step": 6573
+    },
+    {
+      "epoch": 0.5067838421214924,
+      "grad_norm": 3.600267171859741,
+      "learning_rate": 5.133567964337693e-06,
+      "loss": 1.0534,
+      "step": 6574
+    },
+    {
+      "epoch": 0.5068609312365094,
+      "grad_norm": 3.748769760131836,
+      "learning_rate": 5.132319958475025e-06,
+      "loss": 0.8298,
+      "step": 6575
+    },
+    {
+      "epoch": 0.5069380203515264,
+      "grad_norm": 3.435581684112549,
+      "learning_rate": 5.131071944362893e-06,
+      "loss": 0.932,
+      "step": 6576
+    },
+    {
+      "epoch": 0.5070151094665434,
+      "grad_norm": 3.6183292865753174,
+      "learning_rate": 5.129823922079105e-06,
+      "loss": 0.9721,
+      "step": 6577
+    },
+    {
+      "epoch": 0.5070921985815603,
+      "grad_norm": 3.3050825595855713,
+      "learning_rate": 5.128575891701467e-06,
+      "loss": 0.9811,
+      "step": 6578
+    },
+    {
+      "epoch": 0.5071692876965772,
+      "grad_norm": 3.783656120300293,
+      "learning_rate": 5.127327853307788e-06,
+      "loss": 0.8849,
+      "step": 6579
+    },
+    {
+      "epoch": 0.5072463768115942,
+      "grad_norm": 3.649453639984131,
+      "learning_rate": 5.126079806975877e-06,
+      "loss": 1.0261,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5073234659266112,
+      "grad_norm": 3.476015567779541,
+      "learning_rate": 5.124831752783543e-06,
+      "loss": 0.9128,
+      "step": 6581
+    },
+    {
+      "epoch": 0.5074005550416282,
+      "grad_norm": 3.546862840652466,
+      "learning_rate": 5.123583690808596e-06,
+      "loss": 1.0003,
+      "step": 6582
+    },
+    {
+      "epoch": 0.5074776441566451,
+      "grad_norm": 3.592933177947998,
+      "learning_rate": 5.122335621128844e-06,
+      "loss": 0.9616,
+      "step": 6583
+    },
+    {
+      "epoch": 0.507554733271662,
+      "grad_norm": 3.251722812652588,
+      "learning_rate": 5.121087543822103e-06,
+      "loss": 0.9076,
+      "step": 6584
+    },
+    {
+      "epoch": 0.507631822386679,
+      "grad_norm": 3.8498170375823975,
+      "learning_rate": 5.119839458966179e-06,
+      "loss": 0.9246,
+      "step": 6585
+    },
+    {
+      "epoch": 0.507708911501696,
+      "grad_norm": 3.848869800567627,
+      "learning_rate": 5.118591366638885e-06,
+      "loss": 0.9279,
+      "step": 6586
+    },
+    {
+      "epoch": 0.507786000616713,
+      "grad_norm": 3.7381491661071777,
+      "learning_rate": 5.117343266918035e-06,
+      "loss": 0.9364,
+      "step": 6587
+    },
+    {
+      "epoch": 0.5078630897317299,
+      "grad_norm": 3.8815040588378906,
+      "learning_rate": 5.116095159881438e-06,
+      "loss": 0.9837,
+      "step": 6588
+    },
+    {
+      "epoch": 0.5079401788467468,
+      "grad_norm": 3.340867519378662,
+      "learning_rate": 5.11484704560691e-06,
+      "loss": 0.9044,
+      "step": 6589
+    },
+    {
+      "epoch": 0.5080172679617638,
+      "grad_norm": 3.5740439891815186,
+      "learning_rate": 5.113598924172264e-06,
+      "loss": 0.9594,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5080943570767807,
+      "grad_norm": 3.547769784927368,
+      "learning_rate": 5.112350795655313e-06,
+      "loss": 0.8794,
+      "step": 6591
+    },
+    {
+      "epoch": 0.5081714461917978,
+      "grad_norm": 3.2799148559570312,
+      "learning_rate": 5.1111026601338735e-06,
+      "loss": 0.9809,
+      "step": 6592
+    },
+    {
+      "epoch": 0.5082485353068147,
+      "grad_norm": 3.7307305335998535,
+      "learning_rate": 5.109854517685756e-06,
+      "loss": 0.9173,
+      "step": 6593
+    },
+    {
+      "epoch": 0.5083256244218316,
+      "grad_norm": 3.4130656719207764,
+      "learning_rate": 5.108606368388779e-06,
+      "loss": 0.9393,
+      "step": 6594
+    },
+    {
+      "epoch": 0.5084027135368486,
+      "grad_norm": 3.4056098461151123,
+      "learning_rate": 5.107358212320758e-06,
+      "loss": 0.958,
+      "step": 6595
+    },
+    {
+      "epoch": 0.5084798026518655,
+      "grad_norm": 3.6553897857666016,
+      "learning_rate": 5.106110049559507e-06,
+      "loss": 0.9729,
+      "step": 6596
+    },
+    {
+      "epoch": 0.5085568917668826,
+      "grad_norm": 4.16010046005249,
+      "learning_rate": 5.1048618801828454e-06,
+      "loss": 0.9938,
+      "step": 6597
+    },
+    {
+      "epoch": 0.5086339808818995,
+      "grad_norm": 3.6851933002471924,
+      "learning_rate": 5.1036137042685885e-06,
+      "loss": 0.9174,
+      "step": 6598
+    },
+    {
+      "epoch": 0.5087110699969164,
+      "grad_norm": 3.8315787315368652,
+      "learning_rate": 5.1023655218945534e-06,
+      "loss": 0.9586,
+      "step": 6599
+    },
+    {
+      "epoch": 0.5087881591119334,
+      "grad_norm": 3.3780720233917236,
+      "learning_rate": 5.101117333138558e-06,
+      "loss": 0.891,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5088652482269503,
+      "grad_norm": 3.6677708625793457,
+      "learning_rate": 5.099869138078421e-06,
+      "loss": 0.924,
+      "step": 6601
+    },
+    {
+      "epoch": 0.5089423373419674,
+      "grad_norm": 3.382268190383911,
+      "learning_rate": 5.09862093679196e-06,
+      "loss": 0.9927,
+      "step": 6602
+    },
+    {
+      "epoch": 0.5090194264569843,
+      "grad_norm": 3.3687045574188232,
+      "learning_rate": 5.097372729356997e-06,
+      "loss": 0.9537,
+      "step": 6603
+    },
+    {
+      "epoch": 0.5090965155720012,
+      "grad_norm": 3.567380905151367,
+      "learning_rate": 5.096124515851344e-06,
+      "loss": 0.9901,
+      "step": 6604
+    },
+    {
+      "epoch": 0.5091736046870182,
+      "grad_norm": 3.5415704250335693,
+      "learning_rate": 5.094876296352829e-06,
+      "loss": 1.0561,
+      "step": 6605
+    },
+    {
+      "epoch": 0.5092506938020351,
+      "grad_norm": 4.186346530914307,
+      "learning_rate": 5.093628070939266e-06,
+      "loss": 1.0447,
+      "step": 6606
+    },
+    {
+      "epoch": 0.5093277829170522,
+      "grad_norm": 3.475534677505493,
+      "learning_rate": 5.0923798396884785e-06,
+      "loss": 0.9391,
+      "step": 6607
+    },
+    {
+      "epoch": 0.5094048720320691,
+      "grad_norm": 3.3865292072296143,
+      "learning_rate": 5.0911316026782865e-06,
+      "loss": 1.018,
+      "step": 6608
+    },
+    {
+      "epoch": 0.509481961147086,
+      "grad_norm": 3.4238085746765137,
+      "learning_rate": 5.089883359986512e-06,
+      "loss": 0.9044,
+      "step": 6609
+    },
+    {
+      "epoch": 0.509559050262103,
+      "grad_norm": 3.9409427642822266,
+      "learning_rate": 5.088635111690974e-06,
+      "loss": 0.9538,
+      "step": 6610
+    },
+    {
+      "epoch": 0.5096361393771199,
+      "grad_norm": 3.5488216876983643,
+      "learning_rate": 5.087386857869496e-06,
+      "loss": 0.9652,
+      "step": 6611
+    },
+    {
+      "epoch": 0.509713228492137,
+      "grad_norm": 3.679537534713745,
+      "learning_rate": 5.086138598599901e-06,
+      "loss": 0.9817,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5097903176071539,
+      "grad_norm": 3.6628053188323975,
+      "learning_rate": 5.08489033396001e-06,
+      "loss": 1.0088,
+      "step": 6613
+    },
+    {
+      "epoch": 0.5098674067221708,
+      "grad_norm": 3.9006776809692383,
+      "learning_rate": 5.083642064027646e-06,
+      "loss": 0.9371,
+      "step": 6614
+    },
+    {
+      "epoch": 0.5099444958371878,
+      "grad_norm": 3.7066752910614014,
+      "learning_rate": 5.082393788880633e-06,
+      "loss": 0.8783,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5100215849522047,
+      "grad_norm": 3.3197736740112305,
+      "learning_rate": 5.081145508596794e-06,
+      "loss": 0.9278,
+      "step": 6616
+    },
+    {
+      "epoch": 0.5100986740672218,
+      "grad_norm": 3.691556930541992,
+      "learning_rate": 5.079897223253953e-06,
+      "loss": 1.0941,
+      "step": 6617
+    },
+    {
+      "epoch": 0.5101757631822387,
+      "grad_norm": 3.5036165714263916,
+      "learning_rate": 5.078648932929933e-06,
+      "loss": 0.903,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5102528522972556,
+      "grad_norm": 3.766944408416748,
+      "learning_rate": 5.077400637702561e-06,
+      "loss": 0.8997,
+      "step": 6619
+    },
+    {
+      "epoch": 0.5103299414122726,
+      "grad_norm": 3.8213517665863037,
+      "learning_rate": 5.076152337649658e-06,
+      "loss": 0.955,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5104070305272895,
+      "grad_norm": 3.9807519912719727,
+      "learning_rate": 5.074904032849052e-06,
+      "loss": 0.9852,
+      "step": 6621
+    },
+    {
+      "epoch": 0.5104841196423066,
+      "grad_norm": 3.3743574619293213,
+      "learning_rate": 5.0736557233785685e-06,
+      "loss": 0.8994,
+      "step": 6622
+    },
+    {
+      "epoch": 0.5105612087573235,
+      "grad_norm": 3.691159963607788,
+      "learning_rate": 5.072407409316031e-06,
+      "loss": 1.0397,
+      "step": 6623
+    },
+    {
+      "epoch": 0.5106382978723404,
+      "grad_norm": 3.5302302837371826,
+      "learning_rate": 5.071159090739266e-06,
+      "loss": 0.8242,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5107153869873574,
+      "grad_norm": 4.0092926025390625,
+      "learning_rate": 5.069910767726103e-06,
+      "loss": 0.9717,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5107924761023743,
+      "grad_norm": 4.246231555938721,
+      "learning_rate": 5.068662440354362e-06,
+      "loss": 1.1298,
+      "step": 6626
+    },
+    {
+      "epoch": 0.5108695652173914,
+      "grad_norm": 3.4625141620635986,
+      "learning_rate": 5.067414108701876e-06,
+      "loss": 0.9316,
+      "step": 6627
+    },
+    {
+      "epoch": 0.5109466543324083,
+      "grad_norm": 3.577906847000122,
+      "learning_rate": 5.066165772846468e-06,
+      "loss": 0.9469,
+      "step": 6628
+    },
+    {
+      "epoch": 0.5110237434474252,
+      "grad_norm": 3.831422805786133,
+      "learning_rate": 5.064917432865968e-06,
+      "loss": 0.9549,
+      "step": 6629
+    },
+    {
+      "epoch": 0.5111008325624422,
+      "grad_norm": 3.276904344558716,
+      "learning_rate": 5.063669088838201e-06,
+      "loss": 0.8982,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5111779216774591,
+      "grad_norm": 3.7430529594421387,
+      "learning_rate": 5.0624207408409964e-06,
+      "loss": 0.9587,
+      "step": 6631
+    },
+    {
+      "epoch": 0.5112550107924761,
+      "grad_norm": 3.5662450790405273,
+      "learning_rate": 5.061172388952184e-06,
+      "loss": 0.8223,
+      "step": 6632
+    },
+    {
+      "epoch": 0.5113320999074931,
+      "grad_norm": 4.065229892730713,
+      "learning_rate": 5.059924033249587e-06,
+      "loss": 0.9091,
+      "step": 6633
+    },
+    {
+      "epoch": 0.51140918902251,
+      "grad_norm": 3.51495623588562,
+      "learning_rate": 5.058675673811038e-06,
+      "loss": 0.9732,
+      "step": 6634
+    },
+    {
+      "epoch": 0.511486278137527,
+      "grad_norm": 3.6267058849334717,
+      "learning_rate": 5.057427310714366e-06,
+      "loss": 0.9541,
+      "step": 6635
+    },
+    {
+      "epoch": 0.5115633672525439,
+      "grad_norm": 3.579072952270508,
+      "learning_rate": 5.056178944037396e-06,
+      "loss": 0.9344,
+      "step": 6636
+    },
+    {
+      "epoch": 0.511640456367561,
+      "grad_norm": 3.7198750972747803,
+      "learning_rate": 5.054930573857961e-06,
+      "loss": 0.9487,
+      "step": 6637
+    },
+    {
+      "epoch": 0.5117175454825779,
+      "grad_norm": 3.6153461933135986,
+      "learning_rate": 5.05368220025389e-06,
+      "loss": 0.9901,
+      "step": 6638
+    },
+    {
+      "epoch": 0.5117946345975948,
+      "grad_norm": 3.899430513381958,
+      "learning_rate": 5.052433823303011e-06,
+      "loss": 0.9722,
+      "step": 6639
+    },
+    {
+      "epoch": 0.5118717237126118,
+      "grad_norm": 3.687140703201294,
+      "learning_rate": 5.0511854430831564e-06,
+      "loss": 0.9444,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5119488128276287,
+      "grad_norm": 3.4240894317626953,
+      "learning_rate": 5.0499370596721516e-06,
+      "loss": 0.9509,
+      "step": 6641
+    },
+    {
+      "epoch": 0.5120259019426457,
+      "grad_norm": 3.548834800720215,
+      "learning_rate": 5.048688673147831e-06,
+      "loss": 0.8858,
+      "step": 6642
+    },
+    {
+      "epoch": 0.5121029910576627,
+      "grad_norm": 3.7235825061798096,
+      "learning_rate": 5.047440283588025e-06,
+      "loss": 0.899,
+      "step": 6643
+    },
+    {
+      "epoch": 0.5121800801726796,
+      "grad_norm": 3.849551200866699,
+      "learning_rate": 5.0461918910705625e-06,
+      "loss": 0.9537,
+      "step": 6644
+    },
+    {
+      "epoch": 0.5122571692876966,
+      "grad_norm": 3.7150797843933105,
+      "learning_rate": 5.044943495673273e-06,
+      "loss": 0.9883,
+      "step": 6645
+    },
+    {
+      "epoch": 0.5123342584027135,
+      "grad_norm": 4.125245571136475,
+      "learning_rate": 5.043695097473992e-06,
+      "loss": 0.7933,
+      "step": 6646
+    },
+    {
+      "epoch": 0.5124113475177305,
+      "grad_norm": 3.6767773628234863,
+      "learning_rate": 5.042446696550546e-06,
+      "loss": 0.8519,
+      "step": 6647
+    },
+    {
+      "epoch": 0.5124884366327475,
+      "grad_norm": 3.659404754638672,
+      "learning_rate": 5.041198292980768e-06,
+      "loss": 0.8724,
+      "step": 6648
+    },
+    {
+      "epoch": 0.5125655257477644,
+      "grad_norm": 3.8711514472961426,
+      "learning_rate": 5.03994988684249e-06,
+      "loss": 0.8896,
+      "step": 6649
+    },
+    {
+      "epoch": 0.5126426148627814,
+      "grad_norm": 4.328100681304932,
+      "learning_rate": 5.038701478213545e-06,
+      "loss": 0.8541,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5127197039777983,
+      "grad_norm": 4.055181503295898,
+      "learning_rate": 5.037453067171763e-06,
+      "loss": 0.9379,
+      "step": 6651
+    },
+    {
+      "epoch": 0.5127967930928153,
+      "grad_norm": 3.8188493251800537,
+      "learning_rate": 5.036204653794975e-06,
+      "loss": 1.0316,
+      "step": 6652
+    },
+    {
+      "epoch": 0.5128738822078323,
+      "grad_norm": 3.723499059677124,
+      "learning_rate": 5.034956238161015e-06,
+      "loss": 1.0024,
+      "step": 6653
+    },
+    {
+      "epoch": 0.5129509713228492,
+      "grad_norm": 3.589024782180786,
+      "learning_rate": 5.033707820347715e-06,
+      "loss": 0.9118,
+      "step": 6654
+    },
+    {
+      "epoch": 0.5130280604378662,
+      "grad_norm": 4.008199691772461,
+      "learning_rate": 5.0324594004329065e-06,
+      "loss": 0.956,
+      "step": 6655
+    },
+    {
+      "epoch": 0.5131051495528831,
+      "grad_norm": 3.509976387023926,
+      "learning_rate": 5.031210978494423e-06,
+      "loss": 0.9386,
+      "step": 6656
+    },
+    {
+      "epoch": 0.5131822386679001,
+      "grad_norm": 3.630739212036133,
+      "learning_rate": 5.029962554610096e-06,
+      "loss": 0.9486,
+      "step": 6657
+    },
+    {
+      "epoch": 0.513259327782917,
+      "grad_norm": 3.7410812377929688,
+      "learning_rate": 5.0287141288577575e-06,
+      "loss": 0.933,
+      "step": 6658
+    },
+    {
+      "epoch": 0.513336416897934,
+      "grad_norm": 3.538902997970581,
+      "learning_rate": 5.027465701315245e-06,
+      "loss": 0.9184,
+      "step": 6659
+    },
+    {
+      "epoch": 0.513413506012951,
+      "grad_norm": 3.678654670715332,
+      "learning_rate": 5.026217272060386e-06,
+      "loss": 0.8957,
+      "step": 6660
+    },
+    {
+      "epoch": 0.5134905951279679,
+      "grad_norm": 3.5913567543029785,
+      "learning_rate": 5.024968841171016e-06,
+      "loss": 1.0199,
+      "step": 6661
+    },
+    {
+      "epoch": 0.5135676842429849,
+      "grad_norm": 3.574263334274292,
+      "learning_rate": 5.023720408724971e-06,
+      "loss": 0.9235,
+      "step": 6662
+    },
+    {
+      "epoch": 0.5136447733580018,
+      "grad_norm": 4.819427490234375,
+      "learning_rate": 5.022471974800077e-06,
+      "loss": 0.8857,
+      "step": 6663
+    },
+    {
+      "epoch": 0.5137218624730188,
+      "grad_norm": 3.4264743328094482,
+      "learning_rate": 5.021223539474175e-06,
+      "loss": 0.9493,
+      "step": 6664
+    },
+    {
+      "epoch": 0.5137989515880358,
+      "grad_norm": 4.047597885131836,
+      "learning_rate": 5.019975102825093e-06,
+      "loss": 0.9116,
+      "step": 6665
+    },
+    {
+      "epoch": 0.5138760407030527,
+      "grad_norm": 3.5010175704956055,
+      "learning_rate": 5.018726664930667e-06,
+      "loss": 0.9236,
+      "step": 6666
+    },
+    {
+      "epoch": 0.5139531298180697,
+      "grad_norm": 3.643869638442993,
+      "learning_rate": 5.017478225868733e-06,
+      "loss": 0.945,
+      "step": 6667
+    },
+    {
+      "epoch": 0.5140302189330866,
+      "grad_norm": 3.658156156539917,
+      "learning_rate": 5.0162297857171196e-06,
+      "loss": 0.8691,
+      "step": 6668
+    },
+    {
+      "epoch": 0.5141073080481036,
+      "grad_norm": 3.7067670822143555,
+      "learning_rate": 5.0149813445536634e-06,
+      "loss": 0.9861,
+      "step": 6669
+    },
+    {
+      "epoch": 0.5141843971631206,
+      "grad_norm": 3.7881758213043213,
+      "learning_rate": 5.0137329024561985e-06,
+      "loss": 0.9236,
+      "step": 6670
+    },
+    {
+      "epoch": 0.5142614862781375,
+      "grad_norm": 3.391571521759033,
+      "learning_rate": 5.012484459502558e-06,
+      "loss": 0.9774,
+      "step": 6671
+    },
+    {
+      "epoch": 0.5143385753931545,
+      "grad_norm": 3.8176701068878174,
+      "learning_rate": 5.011236015770577e-06,
+      "loss": 0.8668,
+      "step": 6672
+    },
+    {
+      "epoch": 0.5144156645081714,
+      "grad_norm": 3.639432191848755,
+      "learning_rate": 5.0099875713380875e-06,
+      "loss": 0.879,
+      "step": 6673
+    },
+    {
+      "epoch": 0.5144927536231884,
+      "grad_norm": 3.630164861679077,
+      "learning_rate": 5.008739126282924e-06,
+      "loss": 0.9275,
+      "step": 6674
+    },
+    {
+      "epoch": 0.5145698427382054,
+      "grad_norm": 3.4726369380950928,
+      "learning_rate": 5.007490680682924e-06,
+      "loss": 0.9402,
+      "step": 6675
+    },
+    {
+      "epoch": 0.5146469318532223,
+      "grad_norm": 4.002774715423584,
+      "learning_rate": 5.006242234615916e-06,
+      "loss": 0.9698,
+      "step": 6676
+    },
+    {
+      "epoch": 0.5147240209682393,
+      "grad_norm": 3.349799871444702,
+      "learning_rate": 5.004993788159739e-06,
+      "loss": 0.913,
+      "step": 6677
+    },
+    {
+      "epoch": 0.5148011100832562,
+      "grad_norm": 3.616485595703125,
+      "learning_rate": 5.003745341392225e-06,
+      "loss": 0.8706,
+      "step": 6678
+    },
+    {
+      "epoch": 0.5148781991982732,
+      "grad_norm": 3.556090831756592,
+      "learning_rate": 5.0024968943912055e-06,
+      "loss": 0.9315,
+      "step": 6679
+    },
+    {
+      "epoch": 0.5149552883132902,
+      "grad_norm": 3.7224857807159424,
+      "learning_rate": 5.001248447234521e-06,
+      "loss": 0.8919,
+      "step": 6680
+    },
+    {
+      "epoch": 0.5150323774283071,
+      "grad_norm": 3.654167890548706,
+      "learning_rate": 5e-06,
+      "loss": 0.8902,
+      "step": 6681
+    },
+    {
+      "epoch": 0.5151094665433241,
+      "grad_norm": 4.025610446929932,
+      "learning_rate": 4.99875155276548e-06,
+      "loss": 0.9718,
+      "step": 6682
+    },
+    {
+      "epoch": 0.515186555658341,
+      "grad_norm": 3.4189112186431885,
+      "learning_rate": 4.997503105608795e-06,
+      "loss": 0.9592,
+      "step": 6683
+    },
+    {
+      "epoch": 0.515263644773358,
+      "grad_norm": 3.5572469234466553,
+      "learning_rate": 4.996254658607778e-06,
+      "loss": 0.9258,
+      "step": 6684
+    },
+    {
+      "epoch": 0.515340733888375,
+      "grad_norm": 3.8865630626678467,
+      "learning_rate": 4.995006211840263e-06,
+      "loss": 0.9746,
+      "step": 6685
+    },
+    {
+      "epoch": 0.5154178230033919,
+      "grad_norm": 3.353334426879883,
+      "learning_rate": 4.993757765384085e-06,
+      "loss": 1.0067,
+      "step": 6686
+    },
+    {
+      "epoch": 0.5154949121184089,
+      "grad_norm": 3.3230345249176025,
+      "learning_rate": 4.992509319317078e-06,
+      "loss": 0.8983,
+      "step": 6687
+    },
+    {
+      "epoch": 0.5155720012334258,
+      "grad_norm": 3.8244221210479736,
+      "learning_rate": 4.991260873717077e-06,
+      "loss": 1.0921,
+      "step": 6688
+    },
+    {
+      "epoch": 0.5156490903484428,
+      "grad_norm": 3.557263135910034,
+      "learning_rate": 4.990012428661914e-06,
+      "loss": 1.0434,
+      "step": 6689
+    },
+    {
+      "epoch": 0.5157261794634598,
+      "grad_norm": 3.5607333183288574,
+      "learning_rate": 4.988763984229425e-06,
+      "loss": 1.0169,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5158032685784767,
+      "grad_norm": 3.5967090129852295,
+      "learning_rate": 4.987515540497444e-06,
+      "loss": 1.0117,
+      "step": 6691
+    },
+    {
+      "epoch": 0.5158803576934937,
+      "grad_norm": 3.827690839767456,
+      "learning_rate": 4.986267097543803e-06,
+      "loss": 0.9017,
+      "step": 6692
+    },
+    {
+      "epoch": 0.5159574468085106,
+      "grad_norm": 3.67106294631958,
+      "learning_rate": 4.985018655446337e-06,
+      "loss": 0.9155,
+      "step": 6693
+    },
+    {
+      "epoch": 0.5160345359235275,
+      "grad_norm": 3.8216536045074463,
+      "learning_rate": 4.983770214282883e-06,
+      "loss": 0.9669,
+      "step": 6694
+    },
+    {
+      "epoch": 0.5161116250385446,
+      "grad_norm": 4.562589645385742,
+      "learning_rate": 4.9825217741312695e-06,
+      "loss": 0.9065,
+      "step": 6695
+    },
+    {
+      "epoch": 0.5161887141535615,
+      "grad_norm": 3.8225772380828857,
+      "learning_rate": 4.9812733350693335e-06,
+      "loss": 1.104,
+      "step": 6696
+    },
+    {
+      "epoch": 0.5162658032685785,
+      "grad_norm": 3.3218376636505127,
+      "learning_rate": 4.980024897174909e-06,
+      "loss": 0.8526,
+      "step": 6697
+    },
+    {
+      "epoch": 0.5163428923835954,
+      "grad_norm": 3.461397886276245,
+      "learning_rate": 4.978776460525827e-06,
+      "loss": 1.0473,
+      "step": 6698
+    },
+    {
+      "epoch": 0.5164199814986123,
+      "grad_norm": 4.374227523803711,
+      "learning_rate": 4.977528025199925e-06,
+      "loss": 1.0439,
+      "step": 6699
+    },
+    {
+      "epoch": 0.5164970706136294,
+      "grad_norm": 3.6719889640808105,
+      "learning_rate": 4.976279591275033e-06,
+      "loss": 0.9636,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5165741597286463,
+      "grad_norm": 3.455756425857544,
+      "learning_rate": 4.975031158828985e-06,
+      "loss": 0.9633,
+      "step": 6701
+    },
+    {
+      "epoch": 0.5166512488436633,
+      "grad_norm": 3.604682445526123,
+      "learning_rate": 4.9737827279396165e-06,
+      "loss": 0.8877,
+      "step": 6702
+    },
+    {
+      "epoch": 0.5167283379586802,
+      "grad_norm": 4.052313327789307,
+      "learning_rate": 4.9725342986847575e-06,
+      "loss": 0.9772,
+      "step": 6703
+    },
+    {
+      "epoch": 0.5168054270736971,
+      "grad_norm": 3.812180757522583,
+      "learning_rate": 4.971285871142242e-06,
+      "loss": 0.9781,
+      "step": 6704
+    },
+    {
+      "epoch": 0.5168825161887142,
+      "grad_norm": 3.562392234802246,
+      "learning_rate": 4.9700374453899046e-06,
+      "loss": 0.918,
+      "step": 6705
+    },
+    {
+      "epoch": 0.5169596053037311,
+      "grad_norm": 3.3591978549957275,
+      "learning_rate": 4.968789021505578e-06,
+      "loss": 0.9033,
+      "step": 6706
+    },
+    {
+      "epoch": 0.5170366944187481,
+      "grad_norm": 3.2979397773742676,
+      "learning_rate": 4.967540599567094e-06,
+      "loss": 0.9739,
+      "step": 6707
+    },
+    {
+      "epoch": 0.517113783533765,
+      "grad_norm": 3.202559232711792,
+      "learning_rate": 4.9662921796522856e-06,
+      "loss": 0.7981,
+      "step": 6708
+    },
+    {
+      "epoch": 0.5171908726487819,
+      "grad_norm": 3.387112617492676,
+      "learning_rate": 4.965043761838986e-06,
+      "loss": 0.911,
+      "step": 6709
+    },
+    {
+      "epoch": 0.517267961763799,
+      "grad_norm": 3.931759834289551,
+      "learning_rate": 4.963795346205026e-06,
+      "loss": 1.0281,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5173450508788159,
+      "grad_norm": 3.8739635944366455,
+      "learning_rate": 4.962546932828238e-06,
+      "loss": 1.0021,
+      "step": 6711
+    },
+    {
+      "epoch": 0.5174221399938329,
+      "grad_norm": 3.6043684482574463,
+      "learning_rate": 4.961298521786456e-06,
+      "loss": 0.9596,
+      "step": 6712
+    },
+    {
+      "epoch": 0.5174992291088498,
+      "grad_norm": 3.7522406578063965,
+      "learning_rate": 4.96005011315751e-06,
+      "loss": 0.9987,
+      "step": 6713
+    },
+    {
+      "epoch": 0.5175763182238667,
+      "grad_norm": 3.633314609527588,
+      "learning_rate": 4.958801707019234e-06,
+      "loss": 0.9147,
+      "step": 6714
+    },
+    {
+      "epoch": 0.5176534073388838,
+      "grad_norm": 3.7346205711364746,
+      "learning_rate": 4.957553303449456e-06,
+      "loss": 0.9952,
+      "step": 6715
+    },
+    {
+      "epoch": 0.5177304964539007,
+      "grad_norm": 3.443331003189087,
+      "learning_rate": 4.9563049025260105e-06,
+      "loss": 0.8854,
+      "step": 6716
+    },
+    {
+      "epoch": 0.5178075855689177,
+      "grad_norm": 3.816298723220825,
+      "learning_rate": 4.9550565043267285e-06,
+      "loss": 1.0634,
+      "step": 6717
+    },
+    {
+      "epoch": 0.5178846746839346,
+      "grad_norm": 3.5989532470703125,
+      "learning_rate": 4.95380810892944e-06,
+      "loss": 0.9299,
+      "step": 6718
+    },
+    {
+      "epoch": 0.5179617637989515,
+      "grad_norm": 3.5715746879577637,
+      "learning_rate": 4.952559716411976e-06,
+      "loss": 0.9129,
+      "step": 6719
+    },
+    {
+      "epoch": 0.5180388529139686,
+      "grad_norm": 3.993812322616577,
+      "learning_rate": 4.95131132685217e-06,
+      "loss": 0.9028,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5181159420289855,
+      "grad_norm": 3.4312379360198975,
+      "learning_rate": 4.950062940327849e-06,
+      "loss": 1.039,
+      "step": 6721
+    },
+    {
+      "epoch": 0.5181930311440025,
+      "grad_norm": 3.6945204734802246,
+      "learning_rate": 4.948814556916845e-06,
+      "loss": 0.9212,
+      "step": 6722
+    },
+    {
+      "epoch": 0.5182701202590194,
+      "grad_norm": 3.5042591094970703,
+      "learning_rate": 4.94756617669699e-06,
+      "loss": 0.9631,
+      "step": 6723
+    },
+    {
+      "epoch": 0.5183472093740363,
+      "grad_norm": 3.6338624954223633,
+      "learning_rate": 4.946317799746111e-06,
+      "loss": 1.0292,
+      "step": 6724
+    },
+    {
+      "epoch": 0.5184242984890534,
+      "grad_norm": 3.668816566467285,
+      "learning_rate": 4.94506942614204e-06,
+      "loss": 0.9855,
+      "step": 6725
+    },
+    {
+      "epoch": 0.5185013876040703,
+      "grad_norm": 4.495701313018799,
+      "learning_rate": 4.9438210559626045e-06,
+      "loss": 0.9881,
+      "step": 6726
+    },
+    {
+      "epoch": 0.5185784767190873,
+      "grad_norm": 3.3641388416290283,
+      "learning_rate": 4.942572689285636e-06,
+      "loss": 0.9117,
+      "step": 6727
+    },
+    {
+      "epoch": 0.5186555658341042,
+      "grad_norm": 4.106291770935059,
+      "learning_rate": 4.941324326188963e-06,
+      "loss": 0.9193,
+      "step": 6728
+    },
+    {
+      "epoch": 0.5187326549491211,
+      "grad_norm": 3.9092233180999756,
+      "learning_rate": 4.940075966750414e-06,
+      "loss": 1.0645,
+      "step": 6729
+    },
+    {
+      "epoch": 0.5188097440641382,
+      "grad_norm": 4.035946846008301,
+      "learning_rate": 4.938827611047818e-06,
+      "loss": 1.0177,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5188868331791551,
+      "grad_norm": 3.45704984664917,
+      "learning_rate": 4.937579259159004e-06,
+      "loss": 0.9508,
+      "step": 6731
+    },
+    {
+      "epoch": 0.5189639222941721,
+      "grad_norm": 3.8246238231658936,
+      "learning_rate": 4.9363309111618e-06,
+      "loss": 1.014,
+      "step": 6732
+    },
+    {
+      "epoch": 0.519041011409189,
+      "grad_norm": 3.2717719078063965,
+      "learning_rate": 4.935082567134033e-06,
+      "loss": 0.8324,
+      "step": 6733
+    },
+    {
+      "epoch": 0.5191181005242059,
+      "grad_norm": 3.4391167163848877,
+      "learning_rate": 4.933834227153533e-06,
+      "loss": 0.8627,
+      "step": 6734
+    },
+    {
+      "epoch": 0.519195189639223,
+      "grad_norm": 3.452841281890869,
+      "learning_rate": 4.9325858912981265e-06,
+      "loss": 1.0195,
+      "step": 6735
+    },
+    {
+      "epoch": 0.5192722787542399,
+      "grad_norm": 3.5897481441497803,
+      "learning_rate": 4.93133755964564e-06,
+      "loss": 0.8639,
+      "step": 6736
+    },
+    {
+      "epoch": 0.5193493678692569,
+      "grad_norm": 3.7452785968780518,
+      "learning_rate": 4.9300892322739e-06,
+      "loss": 0.9422,
+      "step": 6737
+    },
+    {
+      "epoch": 0.5194264569842738,
+      "grad_norm": 3.614225149154663,
+      "learning_rate": 4.928840909260735e-06,
+      "loss": 0.9332,
+      "step": 6738
+    },
+    {
+      "epoch": 0.5195035460992907,
+      "grad_norm": 3.500887870788574,
+      "learning_rate": 4.927592590683972e-06,
+      "loss": 0.9325,
+      "step": 6739
+    },
+    {
+      "epoch": 0.5195806352143078,
+      "grad_norm": 3.3096768856048584,
+      "learning_rate": 4.926344276621434e-06,
+      "loss": 0.8339,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5196577243293247,
+      "grad_norm": 3.481545925140381,
+      "learning_rate": 4.925095967150949e-06,
+      "loss": 1.0082,
+      "step": 6741
+    },
+    {
+      "epoch": 0.5197348134443417,
+      "grad_norm": 3.6488254070281982,
+      "learning_rate": 4.923847662350344e-06,
+      "loss": 1.0195,
+      "step": 6742
+    },
+    {
+      "epoch": 0.5198119025593586,
+      "grad_norm": 3.6356041431427,
+      "learning_rate": 4.922599362297442e-06,
+      "loss": 0.9489,
+      "step": 6743
+    },
+    {
+      "epoch": 0.5198889916743756,
+      "grad_norm": 3.7208831310272217,
+      "learning_rate": 4.921351067070068e-06,
+      "loss": 0.9843,
+      "step": 6744
+    },
+    {
+      "epoch": 0.5199660807893925,
+      "grad_norm": 3.6727988719940186,
+      "learning_rate": 4.920102776746049e-06,
+      "loss": 0.9307,
+      "step": 6745
+    },
+    {
+      "epoch": 0.5200431699044095,
+      "grad_norm": 3.49745774269104,
+      "learning_rate": 4.918854491403208e-06,
+      "loss": 0.8772,
+      "step": 6746
+    },
+    {
+      "epoch": 0.5201202590194265,
+      "grad_norm": 3.6098008155822754,
+      "learning_rate": 4.91760621111937e-06,
+      "loss": 0.9431,
+      "step": 6747
+    },
+    {
+      "epoch": 0.5201973481344434,
+      "grad_norm": 3.857675552368164,
+      "learning_rate": 4.916357935972354e-06,
+      "loss": 0.9809,
+      "step": 6748
+    },
+    {
+      "epoch": 0.5202744372494604,
+      "grad_norm": 3.8687050342559814,
+      "learning_rate": 4.915109666039991e-06,
+      "loss": 1.0598,
+      "step": 6749
+    },
+    {
+      "epoch": 0.5203515263644773,
+      "grad_norm": 3.66666841506958,
+      "learning_rate": 4.9138614014001e-06,
+      "loss": 0.844,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5204286154794943,
+      "grad_norm": 3.5708515644073486,
+      "learning_rate": 4.912613142130504e-06,
+      "loss": 1.031,
+      "step": 6751
+    },
+    {
+      "epoch": 0.5205057045945113,
+      "grad_norm": 3.5527634620666504,
+      "learning_rate": 4.911364888309027e-06,
+      "loss": 0.9202,
+      "step": 6752
+    },
+    {
+      "epoch": 0.5205827937095282,
+      "grad_norm": 3.893904447555542,
+      "learning_rate": 4.910116640013489e-06,
+      "loss": 0.8644,
+      "step": 6753
+    },
+    {
+      "epoch": 0.5206598828245452,
+      "grad_norm": 3.3729803562164307,
+      "learning_rate": 4.908868397321714e-06,
+      "loss": 0.848,
+      "step": 6754
+    },
+    {
+      "epoch": 0.5207369719395621,
+      "grad_norm": 3.5364203453063965,
+      "learning_rate": 4.907620160311522e-06,
+      "loss": 0.9323,
+      "step": 6755
+    },
+    {
+      "epoch": 0.5208140610545791,
+      "grad_norm": 3.7312333583831787,
+      "learning_rate": 4.906371929060734e-06,
+      "loss": 0.9715,
+      "step": 6756
+    },
+    {
+      "epoch": 0.5208911501695961,
+      "grad_norm": 3.8178627490997314,
+      "learning_rate": 4.905123703647173e-06,
+      "loss": 0.8637,
+      "step": 6757
+    },
+    {
+      "epoch": 0.520968239284613,
+      "grad_norm": 4.14602518081665,
+      "learning_rate": 4.903875484148656e-06,
+      "loss": 1.0387,
+      "step": 6758
+    },
+    {
+      "epoch": 0.52104532839963,
+      "grad_norm": 4.017736911773682,
+      "learning_rate": 4.902627270643006e-06,
+      "loss": 1.0268,
+      "step": 6759
+    },
+    {
+      "epoch": 0.5211224175146469,
+      "grad_norm": 3.643479347229004,
+      "learning_rate": 4.901379063208041e-06,
+      "loss": 1.0273,
+      "step": 6760
+    },
+    {
+      "epoch": 0.5211995066296639,
+      "grad_norm": 3.823141574859619,
+      "learning_rate": 4.90013086192158e-06,
+      "loss": 0.9976,
+      "step": 6761
+    },
+    {
+      "epoch": 0.5212765957446809,
+      "grad_norm": 3.5352299213409424,
+      "learning_rate": 4.898882666861444e-06,
+      "loss": 1.007,
+      "step": 6762
+    },
+    {
+      "epoch": 0.5213536848596978,
+      "grad_norm": 3.585045576095581,
+      "learning_rate": 4.897634478105448e-06,
+      "loss": 0.9049,
+      "step": 6763
+    },
+    {
+      "epoch": 0.5214307739747148,
+      "grad_norm": 3.5495879650115967,
+      "learning_rate": 4.896386295731413e-06,
+      "loss": 0.9165,
+      "step": 6764
+    },
+    {
+      "epoch": 0.5215078630897317,
+      "grad_norm": 3.4797205924987793,
+      "learning_rate": 4.895138119817156e-06,
+      "loss": 0.991,
+      "step": 6765
+    },
+    {
+      "epoch": 0.5215849522047487,
+      "grad_norm": 3.806225538253784,
+      "learning_rate": 4.8938899504404944e-06,
+      "loss": 0.867,
+      "step": 6766
+    },
+    {
+      "epoch": 0.5216620413197657,
+      "grad_norm": 3.340394973754883,
+      "learning_rate": 4.892641787679244e-06,
+      "loss": 0.8744,
+      "step": 6767
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 3.620096206665039,
+      "learning_rate": 4.891393631611223e-06,
+      "loss": 1.0585,
+      "step": 6768
+    },
+    {
+      "epoch": 0.5218162195497996,
+      "grad_norm": 3.8515403270721436,
+      "learning_rate": 4.8901454823142454e-06,
+      "loss": 1.0662,
+      "step": 6769
+    },
+    {
+      "epoch": 0.5218933086648165,
+      "grad_norm": 4.0987420082092285,
+      "learning_rate": 4.888897339866129e-06,
+      "loss": 1.0095,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5219703977798335,
+      "grad_norm": 3.7818572521209717,
+      "learning_rate": 4.887649204344688e-06,
+      "loss": 0.9227,
+      "step": 6771
+    },
+    {
+      "epoch": 0.5220474868948505,
+      "grad_norm": 3.931004285812378,
+      "learning_rate": 4.886401075827737e-06,
+      "loss": 0.9917,
+      "step": 6772
+    },
+    {
+      "epoch": 0.5221245760098674,
+      "grad_norm": 3.2993662357330322,
+      "learning_rate": 4.8851529543930915e-06,
+      "loss": 0.7365,
+      "step": 6773
+    },
+    {
+      "epoch": 0.5222016651248844,
+      "grad_norm": 3.5455148220062256,
+      "learning_rate": 4.883904840118563e-06,
+      "loss": 0.9588,
+      "step": 6774
+    },
+    {
+      "epoch": 0.5222787542399013,
+      "grad_norm": 3.565535306930542,
+      "learning_rate": 4.882656733081967e-06,
+      "loss": 0.9395,
+      "step": 6775
+    },
+    {
+      "epoch": 0.5223558433549182,
+      "grad_norm": 4.117207050323486,
+      "learning_rate": 4.881408633361116e-06,
+      "loss": 1.1194,
+      "step": 6776
+    },
+    {
+      "epoch": 0.5224329324699353,
+      "grad_norm": 3.6578311920166016,
+      "learning_rate": 4.880160541033822e-06,
+      "loss": 0.9629,
+      "step": 6777
+    },
+    {
+      "epoch": 0.5225100215849522,
+      "grad_norm": 4.134820938110352,
+      "learning_rate": 4.878912456177898e-06,
+      "loss": 0.9406,
+      "step": 6778
+    },
+    {
+      "epoch": 0.5225871106999692,
+      "grad_norm": 3.462031841278076,
+      "learning_rate": 4.8776643788711565e-06,
+      "loss": 0.8621,
+      "step": 6779
+    },
+    {
+      "epoch": 0.5226641998149861,
+      "grad_norm": 3.252340793609619,
+      "learning_rate": 4.876416309191406e-06,
+      "loss": 0.9939,
+      "step": 6780
+    },
+    {
+      "epoch": 0.522741288930003,
+      "grad_norm": 3.3890349864959717,
+      "learning_rate": 4.8751682472164585e-06,
+      "loss": 0.9544,
+      "step": 6781
+    },
+    {
+      "epoch": 0.5228183780450201,
+      "grad_norm": 3.870720624923706,
+      "learning_rate": 4.873920193024125e-06,
+      "loss": 0.9931,
+      "step": 6782
+    },
+    {
+      "epoch": 0.522895467160037,
+      "grad_norm": 3.4139373302459717,
+      "learning_rate": 4.872672146692214e-06,
+      "loss": 0.8944,
+      "step": 6783
+    },
+    {
+      "epoch": 0.522972556275054,
+      "grad_norm": 4.145259380340576,
+      "learning_rate": 4.871424108298536e-06,
+      "loss": 1.0266,
+      "step": 6784
+    },
+    {
+      "epoch": 0.5230496453900709,
+      "grad_norm": 3.556744337081909,
+      "learning_rate": 4.8701760779208975e-06,
+      "loss": 0.8754,
+      "step": 6785
+    },
+    {
+      "epoch": 0.5231267345050878,
+      "grad_norm": 3.8212878704071045,
+      "learning_rate": 4.8689280556371084e-06,
+      "loss": 1.0068,
+      "step": 6786
+    },
+    {
+      "epoch": 0.5232038236201049,
+      "grad_norm": 4.2917399406433105,
+      "learning_rate": 4.867680041524977e-06,
+      "loss": 1.0567,
+      "step": 6787
+    },
+    {
+      "epoch": 0.5232809127351218,
+      "grad_norm": 3.264390230178833,
+      "learning_rate": 4.866432035662309e-06,
+      "loss": 0.9187,
+      "step": 6788
+    },
+    {
+      "epoch": 0.5233580018501388,
+      "grad_norm": 3.5728774070739746,
+      "learning_rate": 4.865184038126912e-06,
+      "loss": 0.9863,
+      "step": 6789
+    },
+    {
+      "epoch": 0.5234350909651557,
+      "grad_norm": 4.23891544342041,
+      "learning_rate": 4.863936048996593e-06,
+      "loss": 0.8834,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5235121800801726,
+      "grad_norm": 3.400658130645752,
+      "learning_rate": 4.862688068349156e-06,
+      "loss": 0.8721,
+      "step": 6791
+    },
+    {
+      "epoch": 0.5235892691951897,
+      "grad_norm": 3.758239507675171,
+      "learning_rate": 4.861440096262404e-06,
+      "loss": 0.8671,
+      "step": 6792
+    },
+    {
+      "epoch": 0.5236663583102066,
+      "grad_norm": 3.4603350162506104,
+      "learning_rate": 4.860192132814148e-06,
+      "loss": 0.9106,
+      "step": 6793
+    },
+    {
+      "epoch": 0.5237434474252236,
+      "grad_norm": 3.423933506011963,
+      "learning_rate": 4.858944178082188e-06,
+      "loss": 0.9117,
+      "step": 6794
+    },
+    {
+      "epoch": 0.5238205365402405,
+      "grad_norm": 3.4197144508361816,
+      "learning_rate": 4.857696232144327e-06,
+      "loss": 0.9417,
+      "step": 6795
+    },
+    {
+      "epoch": 0.5238976256552574,
+      "grad_norm": 3.6277825832366943,
+      "learning_rate": 4.8564482950783685e-06,
+      "loss": 0.9014,
+      "step": 6796
+    },
+    {
+      "epoch": 0.5239747147702745,
+      "grad_norm": 3.313845157623291,
+      "learning_rate": 4.855200366962116e-06,
+      "loss": 1.0002,
+      "step": 6797
+    },
+    {
+      "epoch": 0.5240518038852914,
+      "grad_norm": 3.5861170291900635,
+      "learning_rate": 4.853952447873371e-06,
+      "loss": 0.9123,
+      "step": 6798
+    },
+    {
+      "epoch": 0.5241288930003084,
+      "grad_norm": 3.7218551635742188,
+      "learning_rate": 4.852704537889934e-06,
+      "loss": 0.988,
+      "step": 6799
+    },
+    {
+      "epoch": 0.5242059821153253,
+      "grad_norm": 3.6301844120025635,
+      "learning_rate": 4.851456637089607e-06,
+      "loss": 0.9492,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5242830712303422,
+      "grad_norm": 3.789766788482666,
+      "learning_rate": 4.850208745550189e-06,
+      "loss": 0.8719,
+      "step": 6801
+    },
+    {
+      "epoch": 0.5243601603453593,
+      "grad_norm": 3.7619876861572266,
+      "learning_rate": 4.84896086334948e-06,
+      "loss": 1.0564,
+      "step": 6802
+    },
+    {
+      "epoch": 0.5244372494603762,
+      "grad_norm": 3.331724166870117,
+      "learning_rate": 4.8477129905652785e-06,
+      "loss": 0.9431,
+      "step": 6803
+    },
+    {
+      "epoch": 0.5245143385753932,
+      "grad_norm": 3.551481246948242,
+      "learning_rate": 4.8464651272753835e-06,
+      "loss": 0.9178,
+      "step": 6804
+    },
+    {
+      "epoch": 0.5245914276904101,
+      "grad_norm": 3.397059440612793,
+      "learning_rate": 4.845217273557593e-06,
+      "loss": 1.0281,
+      "step": 6805
+    },
+    {
+      "epoch": 0.524668516805427,
+      "grad_norm": 3.586930990219116,
+      "learning_rate": 4.843969429489703e-06,
+      "loss": 1.0194,
+      "step": 6806
+    },
+    {
+      "epoch": 0.5247456059204441,
+      "grad_norm": 3.493337392807007,
+      "learning_rate": 4.842721595149511e-06,
+      "loss": 0.9052,
+      "step": 6807
+    },
+    {
+      "epoch": 0.524822695035461,
+      "grad_norm": 3.607456684112549,
+      "learning_rate": 4.841473770614814e-06,
+      "loss": 0.8831,
+      "step": 6808
+    },
+    {
+      "epoch": 0.524899784150478,
+      "grad_norm": 4.256878852844238,
+      "learning_rate": 4.840225955963406e-06,
+      "loss": 0.9496,
+      "step": 6809
+    },
+    {
+      "epoch": 0.5249768732654949,
+      "grad_norm": 3.489366292953491,
+      "learning_rate": 4.838978151273082e-06,
+      "loss": 0.8825,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5250539623805118,
+      "grad_norm": 3.237924814224243,
+      "learning_rate": 4.8377303566216365e-06,
+      "loss": 0.8377,
+      "step": 6811
+    },
+    {
+      "epoch": 0.5251310514955289,
+      "grad_norm": 3.6981184482574463,
+      "learning_rate": 4.836482572086862e-06,
+      "loss": 0.975,
+      "step": 6812
+    },
+    {
+      "epoch": 0.5252081406105458,
+      "grad_norm": 3.9798178672790527,
+      "learning_rate": 4.8352347977465535e-06,
+      "loss": 1.0573,
+      "step": 6813
+    },
+    {
+      "epoch": 0.5252852297255628,
+      "grad_norm": 3.7807118892669678,
+      "learning_rate": 4.833987033678501e-06,
+      "loss": 1.055,
+      "step": 6814
+    },
+    {
+      "epoch": 0.5253623188405797,
+      "grad_norm": 3.7995917797088623,
+      "learning_rate": 4.8327392799604974e-06,
+      "loss": 1.0315,
+      "step": 6815
+    },
+    {
+      "epoch": 0.5254394079555966,
+      "grad_norm": 3.570406675338745,
+      "learning_rate": 4.831491536670334e-06,
+      "loss": 0.8707,
+      "step": 6816
+    },
+    {
+      "epoch": 0.5255164970706137,
+      "grad_norm": 3.5016283988952637,
+      "learning_rate": 4.8302438038858005e-06,
+      "loss": 1.0007,
+      "step": 6817
+    },
+    {
+      "epoch": 0.5255935861856306,
+      "grad_norm": 3.539832830429077,
+      "learning_rate": 4.828996081684685e-06,
+      "loss": 0.8653,
+      "step": 6818
+    },
+    {
+      "epoch": 0.5256706753006476,
+      "grad_norm": 3.599923610687256,
+      "learning_rate": 4.82774837014478e-06,
+      "loss": 0.9648,
+      "step": 6819
+    },
+    {
+      "epoch": 0.5257477644156645,
+      "grad_norm": 3.7027664184570312,
+      "learning_rate": 4.826500669343872e-06,
+      "loss": 1.0022,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5258248535306814,
+      "grad_norm": 3.6279308795928955,
+      "learning_rate": 4.8252529793597484e-06,
+      "loss": 0.978,
+      "step": 6821
+    },
+    {
+      "epoch": 0.5259019426456985,
+      "grad_norm": 3.6738011837005615,
+      "learning_rate": 4.824005300270196e-06,
+      "loss": 0.9233,
+      "step": 6822
+    },
+    {
+      "epoch": 0.5259790317607154,
+      "grad_norm": 3.7691490650177,
+      "learning_rate": 4.8227576321530015e-06,
+      "loss": 0.9322,
+      "step": 6823
+    },
+    {
+      "epoch": 0.5260561208757324,
+      "grad_norm": 3.266089916229248,
+      "learning_rate": 4.821509975085952e-06,
+      "loss": 0.8356,
+      "step": 6824
+    },
+    {
+      "epoch": 0.5261332099907493,
+      "grad_norm": 3.7823686599731445,
+      "learning_rate": 4.820262329146829e-06,
+      "loss": 0.9445,
+      "step": 6825
+    },
+    {
+      "epoch": 0.5262102991057662,
+      "grad_norm": 3.267396926879883,
+      "learning_rate": 4.81901469441342e-06,
+      "loss": 0.7796,
+      "step": 6826
+    },
+    {
+      "epoch": 0.5262873882207832,
+      "grad_norm": 3.59197998046875,
+      "learning_rate": 4.8177670709635085e-06,
+      "loss": 0.9793,
+      "step": 6827
+    },
+    {
+      "epoch": 0.5263644773358002,
+      "grad_norm": 3.956080913543701,
+      "learning_rate": 4.8165194588748745e-06,
+      "loss": 0.9847,
+      "step": 6828
+    },
+    {
+      "epoch": 0.5264415664508172,
+      "grad_norm": 3.319718837738037,
+      "learning_rate": 4.815271858225303e-06,
+      "loss": 0.7932,
+      "step": 6829
+    },
+    {
+      "epoch": 0.5265186555658341,
+      "grad_norm": 3.7214298248291016,
+      "learning_rate": 4.814024269092575e-06,
+      "loss": 0.9562,
+      "step": 6830
+    },
+    {
+      "epoch": 0.526595744680851,
+      "grad_norm": 3.7417917251586914,
+      "learning_rate": 4.812776691554469e-06,
+      "loss": 1.0291,
+      "step": 6831
+    },
+    {
+      "epoch": 0.526672833795868,
+      "grad_norm": 3.410249948501587,
+      "learning_rate": 4.811529125688768e-06,
+      "loss": 0.9492,
+      "step": 6832
+    },
+    {
+      "epoch": 0.526749922910885,
+      "grad_norm": 3.934293031692505,
+      "learning_rate": 4.8102815715732495e-06,
+      "loss": 1.0856,
+      "step": 6833
+    },
+    {
+      "epoch": 0.526827012025902,
+      "grad_norm": 3.7458932399749756,
+      "learning_rate": 4.809034029285691e-06,
+      "loss": 0.9264,
+      "step": 6834
+    },
+    {
+      "epoch": 0.5269041011409189,
+      "grad_norm": 3.63541579246521,
+      "learning_rate": 4.807786498903874e-06,
+      "loss": 0.9759,
+      "step": 6835
+    },
+    {
+      "epoch": 0.5269811902559358,
+      "grad_norm": 3.5854616165161133,
+      "learning_rate": 4.806538980505572e-06,
+      "loss": 1.0086,
+      "step": 6836
+    },
+    {
+      "epoch": 0.5270582793709528,
+      "grad_norm": 3.602339267730713,
+      "learning_rate": 4.805291474168561e-06,
+      "loss": 0.8689,
+      "step": 6837
+    },
+    {
+      "epoch": 0.5271353684859698,
+      "grad_norm": 3.346555471420288,
+      "learning_rate": 4.80404397997062e-06,
+      "loss": 0.867,
+      "step": 6838
+    },
+    {
+      "epoch": 0.5272124576009868,
+      "grad_norm": 3.9737861156463623,
+      "learning_rate": 4.802796497989523e-06,
+      "loss": 0.9963,
+      "step": 6839
+    },
+    {
+      "epoch": 0.5272895467160037,
+      "grad_norm": 3.378373146057129,
+      "learning_rate": 4.801549028303042e-06,
+      "loss": 0.8988,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5273666358310206,
+      "grad_norm": 3.7955613136291504,
+      "learning_rate": 4.800301570988951e-06,
+      "loss": 0.9651,
+      "step": 6841
+    },
+    {
+      "epoch": 0.5274437249460376,
+      "grad_norm": 3.713027238845825,
+      "learning_rate": 4.799054126125023e-06,
+      "loss": 0.9965,
+      "step": 6842
+    },
+    {
+      "epoch": 0.5275208140610546,
+      "grad_norm": 3.291476011276245,
+      "learning_rate": 4.79780669378903e-06,
+      "loss": 0.9719,
+      "step": 6843
+    },
+    {
+      "epoch": 0.5275979031760716,
+      "grad_norm": 3.837628126144409,
+      "learning_rate": 4.796559274058741e-06,
+      "loss": 0.8665,
+      "step": 6844
+    },
+    {
+      "epoch": 0.5276749922910885,
+      "grad_norm": 3.562666177749634,
+      "learning_rate": 4.79531186701193e-06,
+      "loss": 0.9667,
+      "step": 6845
+    },
+    {
+      "epoch": 0.5277520814061054,
+      "grad_norm": 3.401890754699707,
+      "learning_rate": 4.7940644727263615e-06,
+      "loss": 0.8789,
+      "step": 6846
+    },
+    {
+      "epoch": 0.5278291705211224,
+      "grad_norm": 3.439560651779175,
+      "learning_rate": 4.7928170912798075e-06,
+      "loss": 0.8647,
+      "step": 6847
+    },
+    {
+      "epoch": 0.5279062596361394,
+      "grad_norm": 3.419785737991333,
+      "learning_rate": 4.791569722750036e-06,
+      "loss": 0.8197,
+      "step": 6848
+    },
+    {
+      "epoch": 0.5279833487511564,
+      "grad_norm": 3.7677369117736816,
+      "learning_rate": 4.790322367214812e-06,
+      "loss": 0.9698,
+      "step": 6849
+    },
+    {
+      "epoch": 0.5280604378661733,
+      "grad_norm": 3.788708448410034,
+      "learning_rate": 4.789075024751903e-06,
+      "loss": 0.9806,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5281375269811902,
+      "grad_norm": 3.4838881492614746,
+      "learning_rate": 4.7878276954390726e-06,
+      "loss": 0.9545,
+      "step": 6851
+    },
+    {
+      "epoch": 0.5282146160962072,
+      "grad_norm": 3.561063528060913,
+      "learning_rate": 4.786580379354087e-06,
+      "loss": 0.9546,
+      "step": 6852
+    },
+    {
+      "epoch": 0.5282917052112241,
+      "grad_norm": 3.6301426887512207,
+      "learning_rate": 4.78533307657471e-06,
+      "loss": 0.9259,
+      "step": 6853
+    },
+    {
+      "epoch": 0.5283687943262412,
+      "grad_norm": 3.6689276695251465,
+      "learning_rate": 4.784085787178703e-06,
+      "loss": 0.9707,
+      "step": 6854
+    },
+    {
+      "epoch": 0.5284458834412581,
+      "grad_norm": 3.4558374881744385,
+      "learning_rate": 4.782838511243828e-06,
+      "loss": 0.8788,
+      "step": 6855
+    },
+    {
+      "epoch": 0.528522972556275,
+      "grad_norm": 3.636366367340088,
+      "learning_rate": 4.78159124884785e-06,
+      "loss": 0.9705,
+      "step": 6856
+    },
+    {
+      "epoch": 0.528600061671292,
+      "grad_norm": 3.362304925918579,
+      "learning_rate": 4.780344000068524e-06,
+      "loss": 0.898,
+      "step": 6857
+    },
+    {
+      "epoch": 0.528677150786309,
+      "grad_norm": 3.9811975955963135,
+      "learning_rate": 4.779096764983614e-06,
+      "loss": 0.9952,
+      "step": 6858
+    },
+    {
+      "epoch": 0.528754239901326,
+      "grad_norm": 3.3469631671905518,
+      "learning_rate": 4.777849543670875e-06,
+      "loss": 0.91,
+      "step": 6859
+    },
+    {
+      "epoch": 0.5288313290163429,
+      "grad_norm": 3.896716594696045,
+      "learning_rate": 4.776602336208066e-06,
+      "loss": 1.0635,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5289084181313598,
+      "grad_norm": 3.27893328666687,
+      "learning_rate": 4.775355142672946e-06,
+      "loss": 0.9398,
+      "step": 6861
+    },
+    {
+      "epoch": 0.5289855072463768,
+      "grad_norm": 3.709341049194336,
+      "learning_rate": 4.774107963143269e-06,
+      "loss": 0.8442,
+      "step": 6862
+    },
+    {
+      "epoch": 0.5290625963613937,
+      "grad_norm": 4.007583141326904,
+      "learning_rate": 4.772860797696789e-06,
+      "loss": 0.9884,
+      "step": 6863
+    },
+    {
+      "epoch": 0.5291396854764108,
+      "grad_norm": 3.773080587387085,
+      "learning_rate": 4.771613646411264e-06,
+      "loss": 1.0555,
+      "step": 6864
+    },
+    {
+      "epoch": 0.5292167745914277,
+      "grad_norm": 3.6124513149261475,
+      "learning_rate": 4.770366509364444e-06,
+      "loss": 1.0263,
+      "step": 6865
+    },
+    {
+      "epoch": 0.5292938637064446,
+      "grad_norm": 3.6658780574798584,
+      "learning_rate": 4.7691193866340825e-06,
+      "loss": 0.9506,
+      "step": 6866
+    },
+    {
+      "epoch": 0.5293709528214616,
+      "grad_norm": 3.8184187412261963,
+      "learning_rate": 4.767872278297933e-06,
+      "loss": 0.9266,
+      "step": 6867
+    },
+    {
+      "epoch": 0.5294480419364785,
+      "grad_norm": 3.761213779449463,
+      "learning_rate": 4.766625184433744e-06,
+      "loss": 0.9132,
+      "step": 6868
+    },
+    {
+      "epoch": 0.5295251310514956,
+      "grad_norm": 3.682281970977783,
+      "learning_rate": 4.765378105119266e-06,
+      "loss": 0.9613,
+      "step": 6869
+    },
+    {
+      "epoch": 0.5296022201665125,
+      "grad_norm": 4.122997760772705,
+      "learning_rate": 4.7641310404322475e-06,
+      "loss": 1.1595,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5296793092815294,
+      "grad_norm": 3.7581260204315186,
+      "learning_rate": 4.762883990450437e-06,
+      "loss": 0.8558,
+      "step": 6871
+    },
+    {
+      "epoch": 0.5297563983965464,
+      "grad_norm": 3.5212559700012207,
+      "learning_rate": 4.761636955251584e-06,
+      "loss": 0.9798,
+      "step": 6872
+    },
+    {
+      "epoch": 0.5298334875115633,
+      "grad_norm": 3.686304807662964,
+      "learning_rate": 4.760389934913429e-06,
+      "loss": 0.9631,
+      "step": 6873
+    },
+    {
+      "epoch": 0.5299105766265804,
+      "grad_norm": 3.6530697345733643,
+      "learning_rate": 4.759142929513722e-06,
+      "loss": 0.9296,
+      "step": 6874
+    },
+    {
+      "epoch": 0.5299876657415973,
+      "grad_norm": 3.5719711780548096,
+      "learning_rate": 4.757895939130206e-06,
+      "loss": 1.0083,
+      "step": 6875
+    },
+    {
+      "epoch": 0.5300647548566142,
+      "grad_norm": 4.652621746063232,
+      "learning_rate": 4.7566489638406245e-06,
+      "loss": 0.8969,
+      "step": 6876
+    },
+    {
+      "epoch": 0.5301418439716312,
+      "grad_norm": 3.799062490463257,
+      "learning_rate": 4.755402003722719e-06,
+      "loss": 1.0924,
+      "step": 6877
+    },
+    {
+      "epoch": 0.5302189330866481,
+      "grad_norm": 3.6467134952545166,
+      "learning_rate": 4.754155058854233e-06,
+      "loss": 1.0393,
+      "step": 6878
+    },
+    {
+      "epoch": 0.5302960222016652,
+      "grad_norm": 3.597630500793457,
+      "learning_rate": 4.752908129312905e-06,
+      "loss": 1.0003,
+      "step": 6879
+    },
+    {
+      "epoch": 0.5303731113166821,
+      "grad_norm": 3.513861894607544,
+      "learning_rate": 4.751661215176476e-06,
+      "loss": 0.9198,
+      "step": 6880
+    },
+    {
+      "epoch": 0.530450200431699,
+      "grad_norm": 3.668997049331665,
+      "learning_rate": 4.750414316522681e-06,
+      "loss": 0.9596,
+      "step": 6881
+    },
+    {
+      "epoch": 0.530527289546716,
+      "grad_norm": 3.9775052070617676,
+      "learning_rate": 4.7491674334292646e-06,
+      "loss": 1.0214,
+      "step": 6882
+    },
+    {
+      "epoch": 0.5306043786617329,
+      "grad_norm": 3.3533380031585693,
+      "learning_rate": 4.747920565973957e-06,
+      "loss": 0.9444,
+      "step": 6883
+    },
+    {
+      "epoch": 0.53068146777675,
+      "grad_norm": 3.5897982120513916,
+      "learning_rate": 4.746673714234498e-06,
+      "loss": 0.9681,
+      "step": 6884
+    },
+    {
+      "epoch": 0.5307585568917669,
+      "grad_norm": 3.429609537124634,
+      "learning_rate": 4.745426878288621e-06,
+      "loss": 0.9459,
+      "step": 6885
+    },
+    {
+      "epoch": 0.5308356460067838,
+      "grad_norm": 3.4824554920196533,
+      "learning_rate": 4.744180058214059e-06,
+      "loss": 1.0649,
+      "step": 6886
+    },
+    {
+      "epoch": 0.5309127351218008,
+      "grad_norm": 3.27522611618042,
+      "learning_rate": 4.742933254088547e-06,
+      "loss": 0.8855,
+      "step": 6887
+    },
+    {
+      "epoch": 0.5309898242368177,
+      "grad_norm": 3.366482973098755,
+      "learning_rate": 4.741686465989814e-06,
+      "loss": 0.8464,
+      "step": 6888
+    },
+    {
+      "epoch": 0.5310669133518348,
+      "grad_norm": 3.3331689834594727,
+      "learning_rate": 4.740439693995592e-06,
+      "loss": 0.8612,
+      "step": 6889
+    },
+    {
+      "epoch": 0.5311440024668517,
+      "grad_norm": 3.410470485687256,
+      "learning_rate": 4.739192938183611e-06,
+      "loss": 1.1055,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5312210915818686,
+      "grad_norm": 3.858147144317627,
+      "learning_rate": 4.7379461986315995e-06,
+      "loss": 0.9747,
+      "step": 6891
+    },
+    {
+      "epoch": 0.5312981806968856,
+      "grad_norm": 3.707534074783325,
+      "learning_rate": 4.736699475417285e-06,
+      "loss": 0.8828,
+      "step": 6892
+    },
+    {
+      "epoch": 0.5313752698119025,
+      "grad_norm": 3.58382248878479,
+      "learning_rate": 4.735452768618395e-06,
+      "loss": 0.9151,
+      "step": 6893
+    },
+    {
+      "epoch": 0.5314523589269196,
+      "grad_norm": 3.5018043518066406,
+      "learning_rate": 4.734206078312654e-06,
+      "loss": 1.0009,
+      "step": 6894
+    },
+    {
+      "epoch": 0.5315294480419365,
+      "grad_norm": 4.095643520355225,
+      "learning_rate": 4.732959404577787e-06,
+      "loss": 0.9141,
+      "step": 6895
+    },
+    {
+      "epoch": 0.5316065371569534,
+      "grad_norm": 3.533647060394287,
+      "learning_rate": 4.731712747491521e-06,
+      "loss": 0.9161,
+      "step": 6896
+    },
+    {
+      "epoch": 0.5316836262719704,
+      "grad_norm": 3.72644305229187,
+      "learning_rate": 4.730466107131574e-06,
+      "loss": 0.9536,
+      "step": 6897
+    },
+    {
+      "epoch": 0.5317607153869873,
+      "grad_norm": 3.7895758152008057,
+      "learning_rate": 4.729219483575669e-06,
+      "loss": 0.9062,
+      "step": 6898
+    },
+    {
+      "epoch": 0.5318378045020044,
+      "grad_norm": 3.6426897048950195,
+      "learning_rate": 4.7279728769015266e-06,
+      "loss": 0.9667,
+      "step": 6899
+    },
+    {
+      "epoch": 0.5319148936170213,
+      "grad_norm": 3.507504463195801,
+      "learning_rate": 4.726726287186866e-06,
+      "loss": 0.8926,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5319919827320382,
+      "grad_norm": 3.811716318130493,
+      "learning_rate": 4.725479714509408e-06,
+      "loss": 0.9849,
+      "step": 6901
+    },
+    {
+      "epoch": 0.5320690718470552,
+      "grad_norm": 3.6967577934265137,
+      "learning_rate": 4.724233158946866e-06,
+      "loss": 1.0016,
+      "step": 6902
+    },
+    {
+      "epoch": 0.5321461609620721,
+      "grad_norm": 3.3442699909210205,
+      "learning_rate": 4.722986620576959e-06,
+      "loss": 0.8726,
+      "step": 6903
+    },
+    {
+      "epoch": 0.5322232500770891,
+      "grad_norm": 3.455738067626953,
+      "learning_rate": 4.721740099477402e-06,
+      "loss": 0.9078,
+      "step": 6904
+    },
+    {
+      "epoch": 0.5323003391921061,
+      "grad_norm": 3.491821765899658,
+      "learning_rate": 4.720493595725908e-06,
+      "loss": 0.9688,
+      "step": 6905
+    },
+    {
+      "epoch": 0.532377428307123,
+      "grad_norm": 3.792142868041992,
+      "learning_rate": 4.719247109400192e-06,
+      "loss": 0.9775,
+      "step": 6906
+    },
+    {
+      "epoch": 0.53245451742214,
+      "grad_norm": 3.832749605178833,
+      "learning_rate": 4.718000640577963e-06,
+      "loss": 1.0054,
+      "step": 6907
+    },
+    {
+      "epoch": 0.5325316065371569,
+      "grad_norm": 3.693619728088379,
+      "learning_rate": 4.716754189336934e-06,
+      "loss": 0.9271,
+      "step": 6908
+    },
+    {
+      "epoch": 0.532608695652174,
+      "grad_norm": 3.5125207901000977,
+      "learning_rate": 4.715507755754815e-06,
+      "loss": 0.8987,
+      "step": 6909
+    },
+    {
+      "epoch": 0.5326857847671909,
+      "grad_norm": 3.387834310531616,
+      "learning_rate": 4.714261339909313e-06,
+      "loss": 0.9932,
+      "step": 6910
+    },
+    {
+      "epoch": 0.5327628738822078,
+      "grad_norm": 3.5012171268463135,
+      "learning_rate": 4.713014941878137e-06,
+      "loss": 0.8553,
+      "step": 6911
+    },
+    {
+      "epoch": 0.5328399629972248,
+      "grad_norm": 3.9803214073181152,
+      "learning_rate": 4.711768561738993e-06,
+      "loss": 1.0452,
+      "step": 6912
+    },
+    {
+      "epoch": 0.5329170521122417,
+      "grad_norm": 3.555570363998413,
+      "learning_rate": 4.710522199569586e-06,
+      "loss": 1.0523,
+      "step": 6913
+    },
+    {
+      "epoch": 0.5329941412272587,
+      "grad_norm": 3.6644721031188965,
+      "learning_rate": 4.7092758554476215e-06,
+      "loss": 0.9008,
+      "step": 6914
+    },
+    {
+      "epoch": 0.5330712303422757,
+      "grad_norm": 3.4861278533935547,
+      "learning_rate": 4.708029529450802e-06,
+      "loss": 0.9557,
+      "step": 6915
+    },
+    {
+      "epoch": 0.5331483194572927,
+      "grad_norm": 3.4730987548828125,
+      "learning_rate": 4.706783221656828e-06,
+      "loss": 0.9147,
+      "step": 6916
+    },
+    {
+      "epoch": 0.5332254085723096,
+      "grad_norm": 3.5856359004974365,
+      "learning_rate": 4.705536932143403e-06,
+      "loss": 0.8769,
+      "step": 6917
+    },
+    {
+      "epoch": 0.5333024976873265,
+      "grad_norm": 3.6524572372436523,
+      "learning_rate": 4.704290660988224e-06,
+      "loss": 0.9084,
+      "step": 6918
+    },
+    {
+      "epoch": 0.5333795868023435,
+      "grad_norm": 3.7509005069732666,
+      "learning_rate": 4.70304440826899e-06,
+      "loss": 0.9222,
+      "step": 6919
+    },
+    {
+      "epoch": 0.5334566759173605,
+      "grad_norm": 3.5913355350494385,
+      "learning_rate": 4.701798174063402e-06,
+      "loss": 0.9637,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5335337650323775,
+      "grad_norm": 3.756228446960449,
+      "learning_rate": 4.700551958449151e-06,
+      "loss": 0.9755,
+      "step": 6921
+    },
+    {
+      "epoch": 0.5336108541473944,
+      "grad_norm": 3.908708333969116,
+      "learning_rate": 4.699305761503935e-06,
+      "loss": 1.0154,
+      "step": 6922
+    },
+    {
+      "epoch": 0.5336879432624113,
+      "grad_norm": 3.226095676422119,
+      "learning_rate": 4.698059583305447e-06,
+      "loss": 0.855,
+      "step": 6923
+    },
+    {
+      "epoch": 0.5337650323774283,
+      "grad_norm": 3.847050428390503,
+      "learning_rate": 4.696813423931381e-06,
+      "loss": 0.9876,
+      "step": 6924
+    },
+    {
+      "epoch": 0.5338421214924453,
+      "grad_norm": 3.6332948207855225,
+      "learning_rate": 4.695567283459426e-06,
+      "loss": 0.9547,
+      "step": 6925
+    },
+    {
+      "epoch": 0.5339192106074623,
+      "grad_norm": 3.968454122543335,
+      "learning_rate": 4.694321161967275e-06,
+      "loss": 0.9856,
+      "step": 6926
+    },
+    {
+      "epoch": 0.5339962997224792,
+      "grad_norm": 3.6489675045013428,
+      "learning_rate": 4.693075059532619e-06,
+      "loss": 0.9324,
+      "step": 6927
+    },
+    {
+      "epoch": 0.5340733888374961,
+      "grad_norm": 3.761967182159424,
+      "learning_rate": 4.6918289762331405e-06,
+      "loss": 0.9429,
+      "step": 6928
+    },
+    {
+      "epoch": 0.5341504779525131,
+      "grad_norm": 3.6253440380096436,
+      "learning_rate": 4.69058291214653e-06,
+      "loss": 0.9184,
+      "step": 6929
+    },
+    {
+      "epoch": 0.53422756706753,
+      "grad_norm": 3.8986105918884277,
+      "learning_rate": 4.689336867350474e-06,
+      "loss": 0.9953,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5343046561825471,
+      "grad_norm": 3.430659055709839,
+      "learning_rate": 4.688090841922654e-06,
+      "loss": 0.9899,
+      "step": 6931
+    },
+    {
+      "epoch": 0.534381745297564,
+      "grad_norm": 3.6624832153320312,
+      "learning_rate": 4.686844835940754e-06,
+      "loss": 1.0142,
+      "step": 6932
+    },
+    {
+      "epoch": 0.5344588344125809,
+      "grad_norm": 3.478834390640259,
+      "learning_rate": 4.685598849482458e-06,
+      "loss": 0.9255,
+      "step": 6933
+    },
+    {
+      "epoch": 0.5345359235275979,
+      "grad_norm": 3.8827271461486816,
+      "learning_rate": 4.6843528826254444e-06,
+      "loss": 1.0197,
+      "step": 6934
+    },
+    {
+      "epoch": 0.5346130126426148,
+      "grad_norm": 3.9673526287078857,
+      "learning_rate": 4.683106935447395e-06,
+      "loss": 0.8611,
+      "step": 6935
+    },
+    {
+      "epoch": 0.5346901017576319,
+      "grad_norm": 3.5819478034973145,
+      "learning_rate": 4.681861008025985e-06,
+      "loss": 0.9893,
+      "step": 6936
+    },
+    {
+      "epoch": 0.5347671908726488,
+      "grad_norm": 3.0837767124176025,
+      "learning_rate": 4.680615100438894e-06,
+      "loss": 0.9102,
+      "step": 6937
+    },
+    {
+      "epoch": 0.5348442799876657,
+      "grad_norm": 3.779120922088623,
+      "learning_rate": 4.679369212763799e-06,
+      "loss": 0.9429,
+      "step": 6938
+    },
+    {
+      "epoch": 0.5349213691026827,
+      "grad_norm": 3.4904768466949463,
+      "learning_rate": 4.678123345078371e-06,
+      "loss": 0.9892,
+      "step": 6939
+    },
+    {
+      "epoch": 0.5349984582176996,
+      "grad_norm": 3.5319314002990723,
+      "learning_rate": 4.676877497460286e-06,
+      "loss": 0.9813,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5350755473327167,
+      "grad_norm": 4.093774795532227,
+      "learning_rate": 4.675631669987217e-06,
+      "loss": 1.0125,
+      "step": 6941
+    },
+    {
+      "epoch": 0.5351526364477336,
+      "grad_norm": 3.4952986240386963,
+      "learning_rate": 4.674385862736832e-06,
+      "loss": 0.9641,
+      "step": 6942
+    },
+    {
+      "epoch": 0.5352297255627505,
+      "grad_norm": 3.800241470336914,
+      "learning_rate": 4.673140075786801e-06,
+      "loss": 0.9452,
+      "step": 6943
+    },
+    {
+      "epoch": 0.5353068146777675,
+      "grad_norm": 3.972257614135742,
+      "learning_rate": 4.671894309214796e-06,
+      "loss": 0.9489,
+      "step": 6944
+    },
+    {
+      "epoch": 0.5353839037927844,
+      "grad_norm": 3.972130060195923,
+      "learning_rate": 4.67064856309848e-06,
+      "loss": 1.0165,
+      "step": 6945
+    },
+    {
+      "epoch": 0.5354609929078015,
+      "grad_norm": 3.7796213626861572,
+      "learning_rate": 4.669402837515521e-06,
+      "loss": 0.8956,
+      "step": 6946
+    },
+    {
+      "epoch": 0.5355380820228184,
+      "grad_norm": 3.68583345413208,
+      "learning_rate": 4.668157132543582e-06,
+      "loss": 0.8976,
+      "step": 6947
+    },
+    {
+      "epoch": 0.5356151711378353,
+      "grad_norm": 3.53948974609375,
+      "learning_rate": 4.666911448260327e-06,
+      "loss": 0.9083,
+      "step": 6948
+    },
+    {
+      "epoch": 0.5356922602528523,
+      "grad_norm": 3.7999818325042725,
+      "learning_rate": 4.665665784743419e-06,
+      "loss": 1.0559,
+      "step": 6949
+    },
+    {
+      "epoch": 0.5357693493678692,
+      "grad_norm": 3.492060661315918,
+      "learning_rate": 4.664420142070517e-06,
+      "loss": 0.9844,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5358464384828863,
+      "grad_norm": 3.376392364501953,
+      "learning_rate": 4.663174520319282e-06,
+      "loss": 1.0456,
+      "step": 6951
+    },
+    {
+      "epoch": 0.5359235275979032,
+      "grad_norm": 3.963566303253174,
+      "learning_rate": 4.661928919567371e-06,
+      "loss": 1.0061,
+      "step": 6952
+    },
+    {
+      "epoch": 0.5360006167129201,
+      "grad_norm": 3.267530679702759,
+      "learning_rate": 4.660683339892441e-06,
+      "loss": 0.8218,
+      "step": 6953
+    },
+    {
+      "epoch": 0.5360777058279371,
+      "grad_norm": 3.88205623626709,
+      "learning_rate": 4.659437781372147e-06,
+      "loss": 0.9725,
+      "step": 6954
+    },
+    {
+      "epoch": 0.536154794942954,
+      "grad_norm": 3.550692081451416,
+      "learning_rate": 4.658192244084143e-06,
+      "loss": 0.9525,
+      "step": 6955
+    },
+    {
+      "epoch": 0.5362318840579711,
+      "grad_norm": 3.2505733966827393,
+      "learning_rate": 4.656946728106083e-06,
+      "loss": 0.8847,
+      "step": 6956
+    },
+    {
+      "epoch": 0.536308973172988,
+      "grad_norm": 3.462864398956299,
+      "learning_rate": 4.6557012335156185e-06,
+      "loss": 0.8702,
+      "step": 6957
+    },
+    {
+      "epoch": 0.5363860622880049,
+      "grad_norm": 3.4686732292175293,
+      "learning_rate": 4.654455760390398e-06,
+      "loss": 0.8376,
+      "step": 6958
+    },
+    {
+      "epoch": 0.5364631514030219,
+      "grad_norm": 3.5081446170806885,
+      "learning_rate": 4.653210308808071e-06,
+      "loss": 0.8766,
+      "step": 6959
+    },
+    {
+      "epoch": 0.5365402405180388,
+      "grad_norm": 3.306342124938965,
+      "learning_rate": 4.651964878846285e-06,
+      "loss": 0.8244,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5366173296330559,
+      "grad_norm": 4.375651836395264,
+      "learning_rate": 4.6507194705826866e-06,
+      "loss": 1.039,
+      "step": 6961
+    },
+    {
+      "epoch": 0.5366944187480728,
+      "grad_norm": 3.5302603244781494,
+      "learning_rate": 4.649474084094921e-06,
+      "loss": 0.9445,
+      "step": 6962
+    },
+    {
+      "epoch": 0.5367715078630897,
+      "grad_norm": 3.6997478008270264,
+      "learning_rate": 4.64822871946063e-06,
+      "loss": 0.9587,
+      "step": 6963
+    },
+    {
+      "epoch": 0.5368485969781067,
+      "grad_norm": 3.9736435413360596,
+      "learning_rate": 4.646983376757457e-06,
+      "loss": 0.955,
+      "step": 6964
+    },
+    {
+      "epoch": 0.5369256860931236,
+      "grad_norm": 3.3650596141815186,
+      "learning_rate": 4.645738056063042e-06,
+      "loss": 0.9348,
+      "step": 6965
+    },
+    {
+      "epoch": 0.5370027752081407,
+      "grad_norm": 3.6180570125579834,
+      "learning_rate": 4.6444927574550245e-06,
+      "loss": 0.7941,
+      "step": 6966
+    },
+    {
+      "epoch": 0.5370798643231576,
+      "grad_norm": 3.7963132858276367,
+      "learning_rate": 4.6432474810110415e-06,
+      "loss": 0.8803,
+      "step": 6967
+    },
+    {
+      "epoch": 0.5371569534381745,
+      "grad_norm": 3.3811287879943848,
+      "learning_rate": 4.642002226808732e-06,
+      "loss": 0.7674,
+      "step": 6968
+    },
+    {
+      "epoch": 0.5372340425531915,
+      "grad_norm": 3.9231021404266357,
+      "learning_rate": 4.640756994925726e-06,
+      "loss": 1.0332,
+      "step": 6969
+    },
+    {
+      "epoch": 0.5373111316682084,
+      "grad_norm": 3.508418321609497,
+      "learning_rate": 4.639511785439664e-06,
+      "loss": 0.9099,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5373882207832255,
+      "grad_norm": 3.725053310394287,
+      "learning_rate": 4.638266598428175e-06,
+      "loss": 0.8434,
+      "step": 6971
+    },
+    {
+      "epoch": 0.5374653098982424,
+      "grad_norm": 3.9625444412231445,
+      "learning_rate": 4.63702143396889e-06,
+      "loss": 0.8459,
+      "step": 6972
+    },
+    {
+      "epoch": 0.5375423990132593,
+      "grad_norm": 3.8281447887420654,
+      "learning_rate": 4.635776292139437e-06,
+      "loss": 0.9814,
+      "step": 6973
+    },
+    {
+      "epoch": 0.5376194881282763,
+      "grad_norm": 4.015565872192383,
+      "learning_rate": 4.634531173017448e-06,
+      "loss": 1.0362,
+      "step": 6974
+    },
+    {
+      "epoch": 0.5376965772432932,
+      "grad_norm": 3.4385459423065186,
+      "learning_rate": 4.633286076680546e-06,
+      "loss": 0.8451,
+      "step": 6975
+    },
+    {
+      "epoch": 0.5377736663583103,
+      "grad_norm": 3.9049465656280518,
+      "learning_rate": 4.632041003206359e-06,
+      "loss": 0.7773,
+      "step": 6976
+    },
+    {
+      "epoch": 0.5378507554733272,
+      "grad_norm": 5.6488142013549805,
+      "learning_rate": 4.630795952672509e-06,
+      "loss": 0.9841,
+      "step": 6977
+    },
+    {
+      "epoch": 0.5379278445883441,
+      "grad_norm": 3.4264445304870605,
+      "learning_rate": 4.62955092515662e-06,
+      "loss": 0.9391,
+      "step": 6978
+    },
+    {
+      "epoch": 0.5380049337033611,
+      "grad_norm": 3.8330531120300293,
+      "learning_rate": 4.628305920736311e-06,
+      "loss": 1.0133,
+      "step": 6979
+    },
+    {
+      "epoch": 0.538082022818378,
+      "grad_norm": 3.940965414047241,
+      "learning_rate": 4.627060939489204e-06,
+      "loss": 0.959,
+      "step": 6980
+    },
+    {
+      "epoch": 0.538159111933395,
+      "grad_norm": 3.4086647033691406,
+      "learning_rate": 4.625815981492916e-06,
+      "loss": 0.95,
+      "step": 6981
+    },
+    {
+      "epoch": 0.538236201048412,
+      "grad_norm": 3.4563660621643066,
+      "learning_rate": 4.6245710468250626e-06,
+      "loss": 0.9469,
+      "step": 6982
+    },
+    {
+      "epoch": 0.5383132901634289,
+      "grad_norm": 3.675934076309204,
+      "learning_rate": 4.623326135563261e-06,
+      "loss": 0.8943,
+      "step": 6983
+    },
+    {
+      "epoch": 0.5383903792784459,
+      "grad_norm": 3.754777431488037,
+      "learning_rate": 4.622081247785123e-06,
+      "loss": 0.9712,
+      "step": 6984
+    },
+    {
+      "epoch": 0.5384674683934628,
+      "grad_norm": 3.8146581649780273,
+      "learning_rate": 4.620836383568262e-06,
+      "loss": 0.9867,
+      "step": 6985
+    },
+    {
+      "epoch": 0.5385445575084798,
+      "grad_norm": 3.5257890224456787,
+      "learning_rate": 4.61959154299029e-06,
+      "loss": 0.9331,
+      "step": 6986
+    },
+    {
+      "epoch": 0.5386216466234968,
+      "grad_norm": 3.52950119972229,
+      "learning_rate": 4.618346726128814e-06,
+      "loss": 0.8518,
+      "step": 6987
+    },
+    {
+      "epoch": 0.5386987357385137,
+      "grad_norm": 3.408184766769409,
+      "learning_rate": 4.6171019330614424e-06,
+      "loss": 0.9116,
+      "step": 6988
+    },
+    {
+      "epoch": 0.5387758248535307,
+      "grad_norm": 4.128994941711426,
+      "learning_rate": 4.615857163865784e-06,
+      "loss": 0.9861,
+      "step": 6989
+    },
+    {
+      "epoch": 0.5388529139685476,
+      "grad_norm": 3.886155366897583,
+      "learning_rate": 4.614612418619441e-06,
+      "loss": 1.0266,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5389300030835646,
+      "grad_norm": 3.525916814804077,
+      "learning_rate": 4.613367697400017e-06,
+      "loss": 0.8755,
+      "step": 6991
+    },
+    {
+      "epoch": 0.5390070921985816,
+      "grad_norm": 3.6894612312316895,
+      "learning_rate": 4.612123000285116e-06,
+      "loss": 0.917,
+      "step": 6992
+    },
+    {
+      "epoch": 0.5390841813135985,
+      "grad_norm": 3.610119342803955,
+      "learning_rate": 4.610878327352336e-06,
+      "loss": 0.9675,
+      "step": 6993
+    },
+    {
+      "epoch": 0.5391612704286155,
+      "grad_norm": 3.5024073123931885,
+      "learning_rate": 4.6096336786792775e-06,
+      "loss": 0.9345,
+      "step": 6994
+    },
+    {
+      "epoch": 0.5392383595436324,
+      "grad_norm": 3.419611930847168,
+      "learning_rate": 4.608389054343537e-06,
+      "loss": 0.8538,
+      "step": 6995
+    },
+    {
+      "epoch": 0.5393154486586494,
+      "grad_norm": 3.4121594429016113,
+      "learning_rate": 4.607144454422711e-06,
+      "loss": 0.9487,
+      "step": 6996
+    },
+    {
+      "epoch": 0.5393925377736664,
+      "grad_norm": 3.849426746368408,
+      "learning_rate": 4.605899878994395e-06,
+      "loss": 0.9854,
+      "step": 6997
+    },
+    {
+      "epoch": 0.5394696268886833,
+      "grad_norm": 3.513972759246826,
+      "learning_rate": 4.604655328136177e-06,
+      "loss": 0.948,
+      "step": 6998
+    },
+    {
+      "epoch": 0.5395467160037003,
+      "grad_norm": 3.6121532917022705,
+      "learning_rate": 4.603410801925654e-06,
+      "loss": 1.0326,
+      "step": 6999
+    },
+    {
+      "epoch": 0.5396238051187172,
+      "grad_norm": 3.5565638542175293,
+      "learning_rate": 4.602166300440415e-06,
+      "loss": 0.8536,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5397008942337342,
+      "grad_norm": 3.5659432411193848,
+      "learning_rate": 4.600921823758044e-06,
+      "loss": 0.9094,
+      "step": 7001
+    },
+    {
+      "epoch": 0.5397779833487512,
+      "grad_norm": 3.8643648624420166,
+      "learning_rate": 4.599677371956132e-06,
+      "loss": 0.9769,
+      "step": 7002
+    },
+    {
+      "epoch": 0.5398550724637681,
+      "grad_norm": 3.5835399627685547,
+      "learning_rate": 4.598432945112262e-06,
+      "loss": 0.8676,
+      "step": 7003
+    },
+    {
+      "epoch": 0.5399321615787851,
+      "grad_norm": 3.427264451980591,
+      "learning_rate": 4.597188543304017e-06,
+      "loss": 0.9429,
+      "step": 7004
+    },
+    {
+      "epoch": 0.540009250693802,
+      "grad_norm": 4.065280437469482,
+      "learning_rate": 4.595944166608982e-06,
+      "loss": 0.8582,
+      "step": 7005
+    },
+    {
+      "epoch": 0.540086339808819,
+      "grad_norm": 3.8755879402160645,
+      "learning_rate": 4.594699815104735e-06,
+      "loss": 0.9051,
+      "step": 7006
+    },
+    {
+      "epoch": 0.540163428923836,
+      "grad_norm": 3.609768867492676,
+      "learning_rate": 4.593455488868855e-06,
+      "loss": 0.9795,
+      "step": 7007
+    },
+    {
+      "epoch": 0.5402405180388529,
+      "grad_norm": 3.6061177253723145,
+      "learning_rate": 4.59221118797892e-06,
+      "loss": 0.8821,
+      "step": 7008
+    },
+    {
+      "epoch": 0.5403176071538699,
+      "grad_norm": 3.737497568130493,
+      "learning_rate": 4.590966912512505e-06,
+      "loss": 0.8979,
+      "step": 7009
+    },
+    {
+      "epoch": 0.5403946962688868,
+      "grad_norm": 3.619135618209839,
+      "learning_rate": 4.589722662547185e-06,
+      "loss": 0.8114,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5404717853839038,
+      "grad_norm": 3.6571645736694336,
+      "learning_rate": 4.588478438160532e-06,
+      "loss": 0.9064,
+      "step": 7011
+    },
+    {
+      "epoch": 0.5405488744989208,
+      "grad_norm": 3.8513779640197754,
+      "learning_rate": 4.5872342394301175e-06,
+      "loss": 0.9858,
+      "step": 7012
+    },
+    {
+      "epoch": 0.5406259636139377,
+      "grad_norm": 3.6383817195892334,
+      "learning_rate": 4.58599006643351e-06,
+      "loss": 0.979,
+      "step": 7013
+    },
+    {
+      "epoch": 0.5407030527289547,
+      "grad_norm": 3.927966356277466,
+      "learning_rate": 4.584745919248275e-06,
+      "loss": 1.0223,
+      "step": 7014
+    },
+    {
+      "epoch": 0.5407801418439716,
+      "grad_norm": 3.733217716217041,
+      "learning_rate": 4.5835017979519844e-06,
+      "loss": 0.9701,
+      "step": 7015
+    },
+    {
+      "epoch": 0.5408572309589886,
+      "grad_norm": 3.937772035598755,
+      "learning_rate": 4.582257702622199e-06,
+      "loss": 0.9882,
+      "step": 7016
+    },
+    {
+      "epoch": 0.5409343200740055,
+      "grad_norm": 3.6561355590820312,
+      "learning_rate": 4.581013633336483e-06,
+      "loss": 0.8608,
+      "step": 7017
+    },
+    {
+      "epoch": 0.5410114091890225,
+      "grad_norm": 3.8480238914489746,
+      "learning_rate": 4.579769590172396e-06,
+      "loss": 0.9592,
+      "step": 7018
+    },
+    {
+      "epoch": 0.5410884983040395,
+      "grad_norm": 3.8090572357177734,
+      "learning_rate": 4.578525573207499e-06,
+      "loss": 0.9319,
+      "step": 7019
+    },
+    {
+      "epoch": 0.5411655874190564,
+      "grad_norm": 3.5534706115722656,
+      "learning_rate": 4.57728158251935e-06,
+      "loss": 0.8965,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5412426765340734,
+      "grad_norm": 3.8051340579986572,
+      "learning_rate": 4.576037618185504e-06,
+      "loss": 1.0375,
+      "step": 7021
+    },
+    {
+      "epoch": 0.5413197656490903,
+      "grad_norm": 3.5272462368011475,
+      "learning_rate": 4.574793680283518e-06,
+      "loss": 0.8829,
+      "step": 7022
+    },
+    {
+      "epoch": 0.5413968547641073,
+      "grad_norm": 3.668534278869629,
+      "learning_rate": 4.573549768890944e-06,
+      "loss": 1.0591,
+      "step": 7023
+    },
+    {
+      "epoch": 0.5414739438791243,
+      "grad_norm": 3.377134084701538,
+      "learning_rate": 4.572305884085332e-06,
+      "loss": 0.9829,
+      "step": 7024
+    },
+    {
+      "epoch": 0.5415510329941412,
+      "grad_norm": 3.962786912918091,
+      "learning_rate": 4.5710620259442325e-06,
+      "loss": 1.0349,
+      "step": 7025
+    },
+    {
+      "epoch": 0.5416281221091582,
+      "grad_norm": 3.8737564086914062,
+      "learning_rate": 4.5698181945451966e-06,
+      "loss": 0.995,
+      "step": 7026
+    },
+    {
+      "epoch": 0.5417052112241751,
+      "grad_norm": 3.767915725708008,
+      "learning_rate": 4.568574389965766e-06,
+      "loss": 0.8897,
+      "step": 7027
+    },
+    {
+      "epoch": 0.5417823003391921,
+      "grad_norm": 3.9697983264923096,
+      "learning_rate": 4.567330612283488e-06,
+      "loss": 0.8637,
+      "step": 7028
+    },
+    {
+      "epoch": 0.5418593894542091,
+      "grad_norm": 3.8324038982391357,
+      "learning_rate": 4.566086861575908e-06,
+      "loss": 1.0064,
+      "step": 7029
+    },
+    {
+      "epoch": 0.541936478569226,
+      "grad_norm": 3.6003153324127197,
+      "learning_rate": 4.564843137920561e-06,
+      "loss": 0.9691,
+      "step": 7030
+    },
+    {
+      "epoch": 0.542013567684243,
+      "grad_norm": 3.5627217292785645,
+      "learning_rate": 4.563599441394994e-06,
+      "loss": 0.86,
+      "step": 7031
+    },
+    {
+      "epoch": 0.5420906567992599,
+      "grad_norm": 3.6470260620117188,
+      "learning_rate": 4.56235577207674e-06,
+      "loss": 0.9588,
+      "step": 7032
+    },
+    {
+      "epoch": 0.5421677459142769,
+      "grad_norm": 4.533912181854248,
+      "learning_rate": 4.561112130043337e-06,
+      "loss": 1.0559,
+      "step": 7033
+    },
+    {
+      "epoch": 0.5422448350292939,
+      "grad_norm": 3.7450127601623535,
+      "learning_rate": 4.5598685153723205e-06,
+      "loss": 0.9593,
+      "step": 7034
+    },
+    {
+      "epoch": 0.5423219241443108,
+      "grad_norm": 3.7098305225372314,
+      "learning_rate": 4.558624928141222e-06,
+      "loss": 1.0346,
+      "step": 7035
+    },
+    {
+      "epoch": 0.5423990132593278,
+      "grad_norm": 3.9631752967834473,
+      "learning_rate": 4.557381368427573e-06,
+      "loss": 0.9668,
+      "step": 7036
+    },
+    {
+      "epoch": 0.5424761023743447,
+      "grad_norm": 3.5435612201690674,
+      "learning_rate": 4.5561378363089035e-06,
+      "loss": 1.0405,
+      "step": 7037
+    },
+    {
+      "epoch": 0.5425531914893617,
+      "grad_norm": 3.9541807174682617,
+      "learning_rate": 4.554894331862741e-06,
+      "loss": 0.9185,
+      "step": 7038
+    },
+    {
+      "epoch": 0.5426302806043787,
+      "grad_norm": 3.6304094791412354,
+      "learning_rate": 4.553650855166612e-06,
+      "loss": 0.9789,
+      "step": 7039
+    },
+    {
+      "epoch": 0.5427073697193956,
+      "grad_norm": 3.823951005935669,
+      "learning_rate": 4.55240740629804e-06,
+      "loss": 0.9947,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5427844588344126,
+      "grad_norm": 3.5936341285705566,
+      "learning_rate": 4.551163985334548e-06,
+      "loss": 0.971,
+      "step": 7041
+    },
+    {
+      "epoch": 0.5428615479494295,
+      "grad_norm": 3.330362319946289,
+      "learning_rate": 4.549920592353659e-06,
+      "loss": 0.8836,
+      "step": 7042
+    },
+    {
+      "epoch": 0.5429386370644464,
+      "grad_norm": 3.501664400100708,
+      "learning_rate": 4.548677227432888e-06,
+      "loss": 0.9594,
+      "step": 7043
+    },
+    {
+      "epoch": 0.5430157261794635,
+      "grad_norm": 3.7609310150146484,
+      "learning_rate": 4.547433890649754e-06,
+      "loss": 0.83,
+      "step": 7044
+    },
+    {
+      "epoch": 0.5430928152944804,
+      "grad_norm": 3.811150312423706,
+      "learning_rate": 4.546190582081774e-06,
+      "loss": 0.9383,
+      "step": 7045
+    },
+    {
+      "epoch": 0.5431699044094974,
+      "grad_norm": 3.409965991973877,
+      "learning_rate": 4.544947301806461e-06,
+      "loss": 0.894,
+      "step": 7046
+    },
+    {
+      "epoch": 0.5432469935245143,
+      "grad_norm": 3.6347696781158447,
+      "learning_rate": 4.543704049901326e-06,
+      "loss": 0.8651,
+      "step": 7047
+    },
+    {
+      "epoch": 0.5433240826395312,
+      "grad_norm": 3.6981849670410156,
+      "learning_rate": 4.542460826443881e-06,
+      "loss": 0.964,
+      "step": 7048
+    },
+    {
+      "epoch": 0.5434011717545483,
+      "grad_norm": 3.7254111766815186,
+      "learning_rate": 4.541217631511633e-06,
+      "loss": 0.9934,
+      "step": 7049
+    },
+    {
+      "epoch": 0.5434782608695652,
+      "grad_norm": 3.680558443069458,
+      "learning_rate": 4.5399744651820915e-06,
+      "loss": 0.7968,
+      "step": 7050
+    },
+    {
+      "epoch": 0.5435553499845822,
+      "grad_norm": 3.6431643962860107,
+      "learning_rate": 4.538731327532758e-06,
+      "loss": 0.9735,
+      "step": 7051
+    },
+    {
+      "epoch": 0.5436324390995991,
+      "grad_norm": 3.7536261081695557,
+      "learning_rate": 4.5374882186411375e-06,
+      "loss": 0.8981,
+      "step": 7052
+    },
+    {
+      "epoch": 0.543709528214616,
+      "grad_norm": 3.773322582244873,
+      "learning_rate": 4.536245138584731e-06,
+      "loss": 1.0596,
+      "step": 7053
+    },
+    {
+      "epoch": 0.5437866173296331,
+      "grad_norm": 3.827707052230835,
+      "learning_rate": 4.5350020874410385e-06,
+      "loss": 0.9203,
+      "step": 7054
+    },
+    {
+      "epoch": 0.54386370644465,
+      "grad_norm": 3.5610642433166504,
+      "learning_rate": 4.533759065287557e-06,
+      "loss": 0.9087,
+      "step": 7055
+    },
+    {
+      "epoch": 0.543940795559667,
+      "grad_norm": 3.6839845180511475,
+      "learning_rate": 4.5325160722017845e-06,
+      "loss": 0.8665,
+      "step": 7056
+    },
+    {
+      "epoch": 0.5440178846746839,
+      "grad_norm": 3.7331130504608154,
+      "learning_rate": 4.531273108261213e-06,
+      "loss": 0.8876,
+      "step": 7057
+    },
+    {
+      "epoch": 0.5440949737897008,
+      "grad_norm": 3.359189510345459,
+      "learning_rate": 4.530030173543334e-06,
+      "loss": 0.9535,
+      "step": 7058
+    },
+    {
+      "epoch": 0.5441720629047179,
+      "grad_norm": 3.495391368865967,
+      "learning_rate": 4.528787268125641e-06,
+      "loss": 0.8966,
+      "step": 7059
+    },
+    {
+      "epoch": 0.5442491520197348,
+      "grad_norm": 3.6998205184936523,
+      "learning_rate": 4.527544392085622e-06,
+      "loss": 0.9623,
+      "step": 7060
+    },
+    {
+      "epoch": 0.5443262411347518,
+      "grad_norm": 3.4941718578338623,
+      "learning_rate": 4.526301545500762e-06,
+      "loss": 1.0587,
+      "step": 7061
+    },
+    {
+      "epoch": 0.5444033302497687,
+      "grad_norm": 3.546320676803589,
+      "learning_rate": 4.5250587284485474e-06,
+      "loss": 0.9509,
+      "step": 7062
+    },
+    {
+      "epoch": 0.5444804193647856,
+      "grad_norm": 3.8419322967529297,
+      "learning_rate": 4.523815941006463e-06,
+      "loss": 0.9845,
+      "step": 7063
+    },
+    {
+      "epoch": 0.5445575084798027,
+      "grad_norm": 3.7636489868164062,
+      "learning_rate": 4.522573183251987e-06,
+      "loss": 1.0808,
+      "step": 7064
+    },
+    {
+      "epoch": 0.5446345975948196,
+      "grad_norm": 3.6135027408599854,
+      "learning_rate": 4.5213304552626e-06,
+      "loss": 0.9379,
+      "step": 7065
+    },
+    {
+      "epoch": 0.5447116867098366,
+      "grad_norm": 4.0867414474487305,
+      "learning_rate": 4.5200877571157815e-06,
+      "loss": 0.969,
+      "step": 7066
+    },
+    {
+      "epoch": 0.5447887758248535,
+      "grad_norm": 3.768380641937256,
+      "learning_rate": 4.518845088889004e-06,
+      "loss": 1.017,
+      "step": 7067
+    },
+    {
+      "epoch": 0.5448658649398704,
+      "grad_norm": 3.526736259460449,
+      "learning_rate": 4.517602450659746e-06,
+      "loss": 0.8962,
+      "step": 7068
+    },
+    {
+      "epoch": 0.5449429540548875,
+      "grad_norm": 3.6190271377563477,
+      "learning_rate": 4.516359842505475e-06,
+      "loss": 0.9257,
+      "step": 7069
+    },
+    {
+      "epoch": 0.5450200431699044,
+      "grad_norm": 3.3780858516693115,
+      "learning_rate": 4.515117264503662e-06,
+      "loss": 0.9142,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5450971322849214,
+      "grad_norm": 3.3571228981018066,
+      "learning_rate": 4.513874716731778e-06,
+      "loss": 0.9285,
+      "step": 7071
+    },
+    {
+      "epoch": 0.5451742213999383,
+      "grad_norm": 3.818427801132202,
+      "learning_rate": 4.5126321992672866e-06,
+      "loss": 0.928,
+      "step": 7072
+    },
+    {
+      "epoch": 0.5452513105149552,
+      "grad_norm": 3.895630359649658,
+      "learning_rate": 4.511389712187654e-06,
+      "loss": 0.9358,
+      "step": 7073
+    },
+    {
+      "epoch": 0.5453283996299723,
+      "grad_norm": 3.877424955368042,
+      "learning_rate": 4.510147255570342e-06,
+      "loss": 1.0102,
+      "step": 7074
+    },
+    {
+      "epoch": 0.5454054887449892,
+      "grad_norm": 3.911123752593994,
+      "learning_rate": 4.508904829492812e-06,
+      "loss": 0.9505,
+      "step": 7075
+    },
+    {
+      "epoch": 0.5454825778600062,
+      "grad_norm": 3.7010762691497803,
+      "learning_rate": 4.507662434032522e-06,
+      "loss": 0.98,
+      "step": 7076
+    },
+    {
+      "epoch": 0.5455596669750231,
+      "grad_norm": 3.6135289669036865,
+      "learning_rate": 4.50642006926693e-06,
+      "loss": 0.9494,
+      "step": 7077
+    },
+    {
+      "epoch": 0.54563675609004,
+      "grad_norm": 3.883880138397217,
+      "learning_rate": 4.505177735273489e-06,
+      "loss": 0.9959,
+      "step": 7078
+    },
+    {
+      "epoch": 0.5457138452050571,
+      "grad_norm": 3.969799518585205,
+      "learning_rate": 4.503935432129656e-06,
+      "loss": 0.9272,
+      "step": 7079
+    },
+    {
+      "epoch": 0.545790934320074,
+      "grad_norm": 4.497955799102783,
+      "learning_rate": 4.5026931599128774e-06,
+      "loss": 0.9667,
+      "step": 7080
+    },
+    {
+      "epoch": 0.545868023435091,
+      "grad_norm": 3.585750102996826,
+      "learning_rate": 4.501450918700605e-06,
+      "loss": 0.9303,
+      "step": 7081
+    },
+    {
+      "epoch": 0.5459451125501079,
+      "grad_norm": 3.6245148181915283,
+      "learning_rate": 4.5002087085702865e-06,
+      "loss": 0.9388,
+      "step": 7082
+    },
+    {
+      "epoch": 0.5460222016651248,
+      "grad_norm": 3.6533985137939453,
+      "learning_rate": 4.498966529599365e-06,
+      "loss": 1.0059,
+      "step": 7083
+    },
+    {
+      "epoch": 0.5460992907801419,
+      "grad_norm": 3.751574754714966,
+      "learning_rate": 4.497724381865287e-06,
+      "loss": 0.9515,
+      "step": 7084
+    },
+    {
+      "epoch": 0.5461763798951588,
+      "grad_norm": 3.4590702056884766,
+      "learning_rate": 4.496482265445492e-06,
+      "loss": 0.9293,
+      "step": 7085
+    },
+    {
+      "epoch": 0.5462534690101758,
+      "grad_norm": 3.8915021419525146,
+      "learning_rate": 4.49524018041742e-06,
+      "loss": 1.051,
+      "step": 7086
+    },
+    {
+      "epoch": 0.5463305581251927,
+      "grad_norm": 4.0933990478515625,
+      "learning_rate": 4.493998126858508e-06,
+      "loss": 0.9782,
+      "step": 7087
+    },
+    {
+      "epoch": 0.5464076472402097,
+      "grad_norm": 3.402745485305786,
+      "learning_rate": 4.492756104846193e-06,
+      "loss": 0.9318,
+      "step": 7088
+    },
+    {
+      "epoch": 0.5464847363552267,
+      "grad_norm": 3.519670009613037,
+      "learning_rate": 4.491514114457907e-06,
+      "loss": 0.9108,
+      "step": 7089
+    },
+    {
+      "epoch": 0.5465618254702436,
+      "grad_norm": 3.736721992492676,
+      "learning_rate": 4.490272155771084e-06,
+      "loss": 0.8759,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5466389145852606,
+      "grad_norm": 3.819697141647339,
+      "learning_rate": 4.489030228863151e-06,
+      "loss": 1.0047,
+      "step": 7091
+    },
+    {
+      "epoch": 0.5467160037002775,
+      "grad_norm": 3.8831934928894043,
+      "learning_rate": 4.487788333811536e-06,
+      "loss": 0.9225,
+      "step": 7092
+    },
+    {
+      "epoch": 0.5467930928152945,
+      "grad_norm": 3.722339153289795,
+      "learning_rate": 4.486546470693667e-06,
+      "loss": 0.8511,
+      "step": 7093
+    },
+    {
+      "epoch": 0.5468701819303114,
+      "grad_norm": 3.325698137283325,
+      "learning_rate": 4.4853046395869655e-06,
+      "loss": 0.8824,
+      "step": 7094
+    },
+    {
+      "epoch": 0.5469472710453284,
+      "grad_norm": 3.8227202892303467,
+      "learning_rate": 4.484062840568854e-06,
+      "loss": 0.918,
+      "step": 7095
+    },
+    {
+      "epoch": 0.5470243601603454,
+      "grad_norm": 3.414175033569336,
+      "learning_rate": 4.4828210737167535e-06,
+      "loss": 0.9069,
+      "step": 7096
+    },
+    {
+      "epoch": 0.5471014492753623,
+      "grad_norm": 3.996797800064087,
+      "learning_rate": 4.481579339108079e-06,
+      "loss": 1.0326,
+      "step": 7097
+    },
+    {
+      "epoch": 0.5471785383903793,
+      "grad_norm": 3.6179771423339844,
+      "learning_rate": 4.480337636820249e-06,
+      "loss": 0.9529,
+      "step": 7098
+    },
+    {
+      "epoch": 0.5472556275053962,
+      "grad_norm": 3.6945784091949463,
+      "learning_rate": 4.479095966930674e-06,
+      "loss": 1.1146,
+      "step": 7099
+    },
+    {
+      "epoch": 0.5473327166204132,
+      "grad_norm": 3.5141420364379883,
+      "learning_rate": 4.477854329516769e-06,
+      "loss": 0.8478,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5474098057354302,
+      "grad_norm": 3.9102392196655273,
+      "learning_rate": 4.476612724655943e-06,
+      "loss": 0.9833,
+      "step": 7101
+    },
+    {
+      "epoch": 0.5474868948504471,
+      "grad_norm": 3.6566107273101807,
+      "learning_rate": 4.475371152425599e-06,
+      "loss": 1.0023,
+      "step": 7102
+    },
+    {
+      "epoch": 0.5475639839654641,
+      "grad_norm": 3.598893880844116,
+      "learning_rate": 4.47412961290315e-06,
+      "loss": 0.9257,
+      "step": 7103
+    },
+    {
+      "epoch": 0.547641073080481,
+      "grad_norm": 3.900968313217163,
+      "learning_rate": 4.472888106165995e-06,
+      "loss": 0.9892,
+      "step": 7104
+    },
+    {
+      "epoch": 0.547718162195498,
+      "grad_norm": 3.7693870067596436,
+      "learning_rate": 4.471646632291539e-06,
+      "loss": 0.9436,
+      "step": 7105
+    },
+    {
+      "epoch": 0.547795251310515,
+      "grad_norm": 3.838836193084717,
+      "learning_rate": 4.470405191357176e-06,
+      "loss": 0.9466,
+      "step": 7106
+    },
+    {
+      "epoch": 0.5478723404255319,
+      "grad_norm": 3.1654272079467773,
+      "learning_rate": 4.469163783440308e-06,
+      "loss": 0.823,
+      "step": 7107
+    },
+    {
+      "epoch": 0.5479494295405489,
+      "grad_norm": 3.3918416500091553,
+      "learning_rate": 4.467922408618329e-06,
+      "loss": 0.9723,
+      "step": 7108
+    },
+    {
+      "epoch": 0.5480265186555658,
+      "grad_norm": 4.015628814697266,
+      "learning_rate": 4.466681066968632e-06,
+      "loss": 0.9698,
+      "step": 7109
+    },
+    {
+      "epoch": 0.5481036077705828,
+      "grad_norm": 3.8677635192871094,
+      "learning_rate": 4.465439758568607e-06,
+      "loss": 0.9895,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5481806968855998,
+      "grad_norm": 3.577852249145508,
+      "learning_rate": 4.464198483495647e-06,
+      "loss": 0.9398,
+      "step": 7111
+    },
+    {
+      "epoch": 0.5482577860006167,
+      "grad_norm": 3.617177724838257,
+      "learning_rate": 4.462957241827135e-06,
+      "loss": 0.9324,
+      "step": 7112
+    },
+    {
+      "epoch": 0.5483348751156337,
+      "grad_norm": 3.606205463409424,
+      "learning_rate": 4.461716033640457e-06,
+      "loss": 0.916,
+      "step": 7113
+    },
+    {
+      "epoch": 0.5484119642306506,
+      "grad_norm": 3.655104637145996,
+      "learning_rate": 4.460474859012998e-06,
+      "loss": 0.9226,
+      "step": 7114
+    },
+    {
+      "epoch": 0.5484890533456676,
+      "grad_norm": 3.676701068878174,
+      "learning_rate": 4.4592337180221365e-06,
+      "loss": 0.9754,
+      "step": 7115
+    },
+    {
+      "epoch": 0.5485661424606846,
+      "grad_norm": 3.2556674480438232,
+      "learning_rate": 4.457992610745252e-06,
+      "loss": 0.8332,
+      "step": 7116
+    },
+    {
+      "epoch": 0.5486432315757015,
+      "grad_norm": 3.65635347366333,
+      "learning_rate": 4.456751537259721e-06,
+      "loss": 0.9277,
+      "step": 7117
+    },
+    {
+      "epoch": 0.5487203206907185,
+      "grad_norm": 3.583652973175049,
+      "learning_rate": 4.455510497642918e-06,
+      "loss": 0.9282,
+      "step": 7118
+    },
+    {
+      "epoch": 0.5487974098057354,
+      "grad_norm": 3.441347599029541,
+      "learning_rate": 4.454269491972215e-06,
+      "loss": 0.8934,
+      "step": 7119
+    },
+    {
+      "epoch": 0.5488744989207524,
+      "grad_norm": 3.9320881366729736,
+      "learning_rate": 4.453028520324984e-06,
+      "loss": 0.9556,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5489515880357694,
+      "grad_norm": 3.3631417751312256,
+      "learning_rate": 4.4517875827785905e-06,
+      "loss": 0.9234,
+      "step": 7121
+    },
+    {
+      "epoch": 0.5490286771507863,
+      "grad_norm": 3.7885899543762207,
+      "learning_rate": 4.450546679410403e-06,
+      "loss": 1.0157,
+      "step": 7122
+    },
+    {
+      "epoch": 0.5491057662658033,
+      "grad_norm": 3.6120402812957764,
+      "learning_rate": 4.449305810297784e-06,
+      "loss": 0.8965,
+      "step": 7123
+    },
+    {
+      "epoch": 0.5491828553808202,
+      "grad_norm": 3.5970468521118164,
+      "learning_rate": 4.448064975518096e-06,
+      "loss": 0.9291,
+      "step": 7124
+    },
+    {
+      "epoch": 0.5492599444958371,
+      "grad_norm": 3.6783077716827393,
+      "learning_rate": 4.446824175148699e-06,
+      "loss": 0.82,
+      "step": 7125
+    },
+    {
+      "epoch": 0.5493370336108542,
+      "grad_norm": 3.4367566108703613,
+      "learning_rate": 4.445583409266947e-06,
+      "loss": 1.0236,
+      "step": 7126
+    },
+    {
+      "epoch": 0.5494141227258711,
+      "grad_norm": 4.298835277557373,
+      "learning_rate": 4.444342677950201e-06,
+      "loss": 1.0732,
+      "step": 7127
+    },
+    {
+      "epoch": 0.5494912118408881,
+      "grad_norm": 3.573850631713867,
+      "learning_rate": 4.443101981275809e-06,
+      "loss": 0.9049,
+      "step": 7128
+    },
+    {
+      "epoch": 0.549568300955905,
+      "grad_norm": 3.9243526458740234,
+      "learning_rate": 4.441861319321125e-06,
+      "loss": 1.0487,
+      "step": 7129
+    },
+    {
+      "epoch": 0.549645390070922,
+      "grad_norm": 3.5670223236083984,
+      "learning_rate": 4.440620692163498e-06,
+      "loss": 0.9543,
+      "step": 7130
+    },
+    {
+      "epoch": 0.549722479185939,
+      "grad_norm": 5.046988010406494,
+      "learning_rate": 4.439380099880272e-06,
+      "loss": 1.1399,
+      "step": 7131
+    },
+    {
+      "epoch": 0.5497995683009559,
+      "grad_norm": 4.248262405395508,
+      "learning_rate": 4.438139542548795e-06,
+      "loss": 0.9838,
+      "step": 7132
+    },
+    {
+      "epoch": 0.5498766574159729,
+      "grad_norm": 3.758612632751465,
+      "learning_rate": 4.436899020246407e-06,
+      "loss": 0.8835,
+      "step": 7133
+    },
+    {
+      "epoch": 0.5499537465309898,
+      "grad_norm": 3.501319408416748,
+      "learning_rate": 4.435658533050448e-06,
+      "loss": 0.8992,
+      "step": 7134
+    },
+    {
+      "epoch": 0.5500308356460067,
+      "grad_norm": 3.7768168449401855,
+      "learning_rate": 4.434418081038256e-06,
+      "loss": 0.9216,
+      "step": 7135
+    },
+    {
+      "epoch": 0.5501079247610238,
+      "grad_norm": 3.8390257358551025,
+      "learning_rate": 4.43317766428717e-06,
+      "loss": 0.997,
+      "step": 7136
+    },
+    {
+      "epoch": 0.5501850138760407,
+      "grad_norm": 3.9735095500946045,
+      "learning_rate": 4.4319372828745185e-06,
+      "loss": 1.0097,
+      "step": 7137
+    },
+    {
+      "epoch": 0.5502621029910577,
+      "grad_norm": 3.6092162132263184,
+      "learning_rate": 4.430696936877638e-06,
+      "loss": 0.9526,
+      "step": 7138
+    },
+    {
+      "epoch": 0.5503391921060746,
+      "grad_norm": 3.396688461303711,
+      "learning_rate": 4.429456626373853e-06,
+      "loss": 0.8063,
+      "step": 7139
+    },
+    {
+      "epoch": 0.5504162812210915,
+      "grad_norm": 3.1804189682006836,
+      "learning_rate": 4.428216351440492e-06,
+      "loss": 0.733,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5504933703361086,
+      "grad_norm": 3.653811454772949,
+      "learning_rate": 4.4269761121548815e-06,
+      "loss": 1.1142,
+      "step": 7141
+    },
+    {
+      "epoch": 0.5505704594511255,
+      "grad_norm": 3.4570720195770264,
+      "learning_rate": 4.425735908594342e-06,
+      "loss": 0.8599,
+      "step": 7142
+    },
+    {
+      "epoch": 0.5506475485661425,
+      "grad_norm": 3.407050609588623,
+      "learning_rate": 4.424495740836193e-06,
+      "loss": 0.8526,
+      "step": 7143
+    },
+    {
+      "epoch": 0.5507246376811594,
+      "grad_norm": 4.278409481048584,
+      "learning_rate": 4.4232556089577575e-06,
+      "loss": 1.0937,
+      "step": 7144
+    },
+    {
+      "epoch": 0.5508017267961763,
+      "grad_norm": 3.6288211345672607,
+      "learning_rate": 4.4220155130363445e-06,
+      "loss": 1.0142,
+      "step": 7145
+    },
+    {
+      "epoch": 0.5508788159111934,
+      "grad_norm": 4.041045188903809,
+      "learning_rate": 4.420775453149273e-06,
+      "loss": 0.955,
+      "step": 7146
+    },
+    {
+      "epoch": 0.5509559050262103,
+      "grad_norm": 3.636233329772949,
+      "learning_rate": 4.4195354293738484e-06,
+      "loss": 0.937,
+      "step": 7147
+    },
+    {
+      "epoch": 0.5510329941412273,
+      "grad_norm": 3.6226770877838135,
+      "learning_rate": 4.418295441787387e-06,
+      "loss": 0.9552,
+      "step": 7148
+    },
+    {
+      "epoch": 0.5511100832562442,
+      "grad_norm": 3.7020938396453857,
+      "learning_rate": 4.417055490467191e-06,
+      "loss": 1.002,
+      "step": 7149
+    },
+    {
+      "epoch": 0.5511871723712611,
+      "grad_norm": 3.8105432987213135,
+      "learning_rate": 4.4158155754905655e-06,
+      "loss": 0.9974,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5512642614862782,
+      "grad_norm": 3.620668888092041,
+      "learning_rate": 4.414575696934814e-06,
+      "loss": 0.9996,
+      "step": 7151
+    },
+    {
+      "epoch": 0.5513413506012951,
+      "grad_norm": 3.2780075073242188,
+      "learning_rate": 4.413335854877237e-06,
+      "loss": 0.9061,
+      "step": 7152
+    },
+    {
+      "epoch": 0.5514184397163121,
+      "grad_norm": 3.7314953804016113,
+      "learning_rate": 4.41209604939513e-06,
+      "loss": 1.008,
+      "step": 7153
+    },
+    {
+      "epoch": 0.551495528831329,
+      "grad_norm": 3.6991615295410156,
+      "learning_rate": 4.4108562805657886e-06,
+      "loss": 0.9696,
+      "step": 7154
+    },
+    {
+      "epoch": 0.5515726179463459,
+      "grad_norm": 3.4623279571533203,
+      "learning_rate": 4.409616548466508e-06,
+      "loss": 0.8454,
+      "step": 7155
+    },
+    {
+      "epoch": 0.551649707061363,
+      "grad_norm": 3.6211092472076416,
+      "learning_rate": 4.408376853174578e-06,
+      "loss": 0.9226,
+      "step": 7156
+    },
+    {
+      "epoch": 0.5517267961763799,
+      "grad_norm": 3.806649684906006,
+      "learning_rate": 4.407137194767286e-06,
+      "loss": 0.9433,
+      "step": 7157
+    },
+    {
+      "epoch": 0.5518038852913969,
+      "grad_norm": 3.5186119079589844,
+      "learning_rate": 4.40589757332192e-06,
+      "loss": 0.9126,
+      "step": 7158
+    },
+    {
+      "epoch": 0.5518809744064138,
+      "grad_norm": 3.941894054412842,
+      "learning_rate": 4.404657988915764e-06,
+      "loss": 1.053,
+      "step": 7159
+    },
+    {
+      "epoch": 0.5519580635214307,
+      "grad_norm": 3.503800868988037,
+      "learning_rate": 4.4034184416260975e-06,
+      "loss": 0.8895,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5520351526364478,
+      "grad_norm": 3.7842636108398438,
+      "learning_rate": 4.402178931530202e-06,
+      "loss": 0.9407,
+      "step": 7161
+    },
+    {
+      "epoch": 0.5521122417514647,
+      "grad_norm": 3.6619977951049805,
+      "learning_rate": 4.400939458705356e-06,
+      "loss": 0.9676,
+      "step": 7162
+    },
+    {
+      "epoch": 0.5521893308664817,
+      "grad_norm": 3.5588581562042236,
+      "learning_rate": 4.3997000232288295e-06,
+      "loss": 0.8884,
+      "step": 7163
+    },
+    {
+      "epoch": 0.5522664199814986,
+      "grad_norm": 3.8872478008270264,
+      "learning_rate": 4.398460625177899e-06,
+      "loss": 1.0666,
+      "step": 7164
+    },
+    {
+      "epoch": 0.5523435090965155,
+      "grad_norm": 3.6066267490386963,
+      "learning_rate": 4.397221264629833e-06,
+      "loss": 0.8131,
+      "step": 7165
+    },
+    {
+      "epoch": 0.5524205982115326,
+      "grad_norm": 4.402515411376953,
+      "learning_rate": 4.395981941661897e-06,
+      "loss": 0.9393,
+      "step": 7166
+    },
+    {
+      "epoch": 0.5524976873265495,
+      "grad_norm": 3.6887872219085693,
+      "learning_rate": 4.394742656351362e-06,
+      "loss": 0.8782,
+      "step": 7167
+    },
+    {
+      "epoch": 0.5525747764415665,
+      "grad_norm": 3.382742404937744,
+      "learning_rate": 4.393503408775485e-06,
+      "loss": 0.9317,
+      "step": 7168
+    },
+    {
+      "epoch": 0.5526518655565834,
+      "grad_norm": 3.7242555618286133,
+      "learning_rate": 4.39226419901153e-06,
+      "loss": 0.9919,
+      "step": 7169
+    },
+    {
+      "epoch": 0.5527289546716003,
+      "grad_norm": 3.5612120628356934,
+      "learning_rate": 4.391025027136756e-06,
+      "loss": 1.0148,
+      "step": 7170
+    },
+    {
+      "epoch": 0.5528060437866174,
+      "grad_norm": 3.597330331802368,
+      "learning_rate": 4.389785893228416e-06,
+      "loss": 0.9333,
+      "step": 7171
+    },
+    {
+      "epoch": 0.5528831329016343,
+      "grad_norm": 3.885812520980835,
+      "learning_rate": 4.388546797363767e-06,
+      "loss": 0.949,
+      "step": 7172
+    },
+    {
+      "epoch": 0.5529602220166513,
+      "grad_norm": 3.399912118911743,
+      "learning_rate": 4.387307739620057e-06,
+      "loss": 0.906,
+      "step": 7173
+    },
+    {
+      "epoch": 0.5530373111316682,
+      "grad_norm": 3.6478137969970703,
+      "learning_rate": 4.386068720074536e-06,
+      "loss": 0.9246,
+      "step": 7174
+    },
+    {
+      "epoch": 0.5531144002466851,
+      "grad_norm": 3.986933946609497,
+      "learning_rate": 4.384829738804452e-06,
+      "loss": 0.9272,
+      "step": 7175
+    },
+    {
+      "epoch": 0.5531914893617021,
+      "grad_norm": 3.6242878437042236,
+      "learning_rate": 4.383590795887046e-06,
+      "loss": 0.9373,
+      "step": 7176
+    },
+    {
+      "epoch": 0.5532685784767191,
+      "grad_norm": 3.7404351234436035,
+      "learning_rate": 4.382351891399562e-06,
+      "loss": 0.9936,
+      "step": 7177
+    },
+    {
+      "epoch": 0.5533456675917361,
+      "grad_norm": 3.527928590774536,
+      "learning_rate": 4.38111302541924e-06,
+      "loss": 0.947,
+      "step": 7178
+    },
+    {
+      "epoch": 0.553422756706753,
+      "grad_norm": 3.5108187198638916,
+      "learning_rate": 4.379874198023314e-06,
+      "loss": 0.8112,
+      "step": 7179
+    },
+    {
+      "epoch": 0.5534998458217699,
+      "grad_norm": 3.2686216831207275,
+      "learning_rate": 4.37863540928902e-06,
+      "loss": 0.8961,
+      "step": 7180
+    },
+    {
+      "epoch": 0.553576934936787,
+      "grad_norm": 3.910783290863037,
+      "learning_rate": 4.3773966592935914e-06,
+      "loss": 1.0448,
+      "step": 7181
+    },
+    {
+      "epoch": 0.5536540240518039,
+      "grad_norm": 3.619373083114624,
+      "learning_rate": 4.3761579481142555e-06,
+      "loss": 0.9193,
+      "step": 7182
+    },
+    {
+      "epoch": 0.5537311131668209,
+      "grad_norm": 3.356902837753296,
+      "learning_rate": 4.374919275828241e-06,
+      "loss": 0.8701,
+      "step": 7183
+    },
+    {
+      "epoch": 0.5538082022818378,
+      "grad_norm": 4.081925392150879,
+      "learning_rate": 4.373680642512773e-06,
+      "loss": 1.0359,
+      "step": 7184
+    },
+    {
+      "epoch": 0.5538852913968547,
+      "grad_norm": 3.6913883686065674,
+      "learning_rate": 4.372442048245071e-06,
+      "loss": 1.0076,
+      "step": 7185
+    },
+    {
+      "epoch": 0.5539623805118717,
+      "grad_norm": 3.53983736038208,
+      "learning_rate": 4.371203493102359e-06,
+      "loss": 0.9063,
+      "step": 7186
+    },
+    {
+      "epoch": 0.5540394696268887,
+      "grad_norm": 3.321823835372925,
+      "learning_rate": 4.369964977161853e-06,
+      "loss": 0.8075,
+      "step": 7187
+    },
+    {
+      "epoch": 0.5541165587419057,
+      "grad_norm": 3.6654202938079834,
+      "learning_rate": 4.368726500500765e-06,
+      "loss": 0.9791,
+      "step": 7188
+    },
+    {
+      "epoch": 0.5541936478569226,
+      "grad_norm": 3.718724012374878,
+      "learning_rate": 4.367488063196313e-06,
+      "loss": 0.9182,
+      "step": 7189
+    },
+    {
+      "epoch": 0.5542707369719395,
+      "grad_norm": 3.7318592071533203,
+      "learning_rate": 4.366249665325703e-06,
+      "loss": 0.9435,
+      "step": 7190
+    },
+    {
+      "epoch": 0.5543478260869565,
+      "grad_norm": 3.3900840282440186,
+      "learning_rate": 4.3650113069661426e-06,
+      "loss": 0.7534,
+      "step": 7191
+    },
+    {
+      "epoch": 0.5544249152019735,
+      "grad_norm": 3.4240376949310303,
+      "learning_rate": 4.36377298819484e-06,
+      "loss": 0.8855,
+      "step": 7192
+    },
+    {
+      "epoch": 0.5545020043169905,
+      "grad_norm": 4.262863636016846,
+      "learning_rate": 4.3625347090889975e-06,
+      "loss": 0.9735,
+      "step": 7193
+    },
+    {
+      "epoch": 0.5545790934320074,
+      "grad_norm": 4.42769193649292,
+      "learning_rate": 4.361296469725813e-06,
+      "loss": 1.1063,
+      "step": 7194
+    },
+    {
+      "epoch": 0.5546561825470243,
+      "grad_norm": 4.841814041137695,
+      "learning_rate": 4.360058270182486e-06,
+      "loss": 0.9777,
+      "step": 7195
+    },
+    {
+      "epoch": 0.5547332716620413,
+      "grad_norm": 3.457386016845703,
+      "learning_rate": 4.3588201105362125e-06,
+      "loss": 0.9072,
+      "step": 7196
+    },
+    {
+      "epoch": 0.5548103607770583,
+      "grad_norm": 3.6573784351348877,
+      "learning_rate": 4.357581990864183e-06,
+      "loss": 0.8965,
+      "step": 7197
+    },
+    {
+      "epoch": 0.5548874498920753,
+      "grad_norm": 3.5548505783081055,
+      "learning_rate": 4.3563439112435904e-06,
+      "loss": 0.7159,
+      "step": 7198
+    },
+    {
+      "epoch": 0.5549645390070922,
+      "grad_norm": 3.5277884006500244,
+      "learning_rate": 4.355105871751621e-06,
+      "loss": 1.0291,
+      "step": 7199
+    },
+    {
+      "epoch": 0.5550416281221091,
+      "grad_norm": 3.4423916339874268,
+      "learning_rate": 4.35386787246546e-06,
+      "loss": 0.8138,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5551187172371261,
+      "grad_norm": 4.191361904144287,
+      "learning_rate": 4.352629913462292e-06,
+      "loss": 0.9909,
+      "step": 7201
+    },
+    {
+      "epoch": 0.555195806352143,
+      "grad_norm": 3.601893663406372,
+      "learning_rate": 4.351391994819294e-06,
+      "loss": 0.9646,
+      "step": 7202
+    },
+    {
+      "epoch": 0.5552728954671601,
+      "grad_norm": 3.6490609645843506,
+      "learning_rate": 4.350154116613648e-06,
+      "loss": 0.9988,
+      "step": 7203
+    },
+    {
+      "epoch": 0.555349984582177,
+      "grad_norm": 3.3986620903015137,
+      "learning_rate": 4.348916278922526e-06,
+      "loss": 0.9039,
+      "step": 7204
+    },
+    {
+      "epoch": 0.5554270736971939,
+      "grad_norm": 4.264974117279053,
+      "learning_rate": 4.347678481823102e-06,
+      "loss": 1.0964,
+      "step": 7205
+    },
+    {
+      "epoch": 0.5555041628122109,
+      "grad_norm": 4.056002140045166,
+      "learning_rate": 4.346440725392546e-06,
+      "loss": 1.044,
+      "step": 7206
+    },
+    {
+      "epoch": 0.5555812519272278,
+      "grad_norm": 3.6536853313446045,
+      "learning_rate": 4.345203009708027e-06,
+      "loss": 1.0003,
+      "step": 7207
+    },
+    {
+      "epoch": 0.5556583410422449,
+      "grad_norm": 3.690647840499878,
+      "learning_rate": 4.343965334846708e-06,
+      "loss": 0.9672,
+      "step": 7208
+    },
+    {
+      "epoch": 0.5557354301572618,
+      "grad_norm": 3.597975969314575,
+      "learning_rate": 4.3427277008857526e-06,
+      "loss": 0.9204,
+      "step": 7209
+    },
+    {
+      "epoch": 0.5558125192722787,
+      "grad_norm": 3.5312154293060303,
+      "learning_rate": 4.341490107902323e-06,
+      "loss": 0.9343,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5558896083872957,
+      "grad_norm": 3.8047034740448,
+      "learning_rate": 4.340252555973572e-06,
+      "loss": 1.0683,
+      "step": 7211
+    },
+    {
+      "epoch": 0.5559666975023126,
+      "grad_norm": 3.7284018993377686,
+      "learning_rate": 4.339015045176659e-06,
+      "loss": 0.8455,
+      "step": 7212
+    },
+    {
+      "epoch": 0.5560437866173297,
+      "grad_norm": 4.019394397735596,
+      "learning_rate": 4.3377775755887335e-06,
+      "loss": 1.0309,
+      "step": 7213
+    },
+    {
+      "epoch": 0.5561208757323466,
+      "grad_norm": 4.077296733856201,
+      "learning_rate": 4.336540147286946e-06,
+      "loss": 1.0705,
+      "step": 7214
+    },
+    {
+      "epoch": 0.5561979648473635,
+      "grad_norm": 3.3389627933502197,
+      "learning_rate": 4.335302760348446e-06,
+      "loss": 0.8988,
+      "step": 7215
+    },
+    {
+      "epoch": 0.5562750539623805,
+      "grad_norm": 3.577312469482422,
+      "learning_rate": 4.334065414850375e-06,
+      "loss": 0.9091,
+      "step": 7216
+    },
+    {
+      "epoch": 0.5563521430773974,
+      "grad_norm": 3.472486734390259,
+      "learning_rate": 4.3328281108698765e-06,
+      "loss": 0.9721,
+      "step": 7217
+    },
+    {
+      "epoch": 0.5564292321924145,
+      "grad_norm": 4.345758438110352,
+      "learning_rate": 4.33159084848409e-06,
+      "loss": 1.0537,
+      "step": 7218
+    },
+    {
+      "epoch": 0.5565063213074314,
+      "grad_norm": 3.5897371768951416,
+      "learning_rate": 4.330353627770153e-06,
+      "loss": 0.8934,
+      "step": 7219
+    },
+    {
+      "epoch": 0.5565834104224483,
+      "grad_norm": 4.042248725891113,
+      "learning_rate": 4.329116448805198e-06,
+      "loss": 1.1139,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5566604995374653,
+      "grad_norm": 4.091678619384766,
+      "learning_rate": 4.327879311666359e-06,
+      "loss": 0.9429,
+      "step": 7221
+    },
+    {
+      "epoch": 0.5567375886524822,
+      "grad_norm": 3.7830512523651123,
+      "learning_rate": 4.326642216430763e-06,
+      "loss": 1.0191,
+      "step": 7222
+    },
+    {
+      "epoch": 0.5568146777674993,
+      "grad_norm": 3.766197919845581,
+      "learning_rate": 4.325405163175539e-06,
+      "loss": 0.983,
+      "step": 7223
+    },
+    {
+      "epoch": 0.5568917668825162,
+      "grad_norm": 3.5098490715026855,
+      "learning_rate": 4.3241681519778074e-06,
+      "loss": 0.9068,
+      "step": 7224
+    },
+    {
+      "epoch": 0.5569688559975331,
+      "grad_norm": 3.6200573444366455,
+      "learning_rate": 4.322931182914692e-06,
+      "loss": 0.8358,
+      "step": 7225
+    },
+    {
+      "epoch": 0.5570459451125501,
+      "grad_norm": 3.716994285583496,
+      "learning_rate": 4.321694256063311e-06,
+      "loss": 1.0499,
+      "step": 7226
+    },
+    {
+      "epoch": 0.557123034227567,
+      "grad_norm": 3.880415439605713,
+      "learning_rate": 4.3204573715007795e-06,
+      "loss": 0.9322,
+      "step": 7227
+    },
+    {
+      "epoch": 0.5572001233425841,
+      "grad_norm": 3.300236940383911,
+      "learning_rate": 4.3192205293042105e-06,
+      "loss": 0.8236,
+      "step": 7228
+    },
+    {
+      "epoch": 0.557277212457601,
+      "grad_norm": 4.212109565734863,
+      "learning_rate": 4.317983729550718e-06,
+      "loss": 0.9395,
+      "step": 7229
+    },
+    {
+      "epoch": 0.5573543015726179,
+      "grad_norm": 3.5828144550323486,
+      "learning_rate": 4.316746972317406e-06,
+      "loss": 1.0149,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5574313906876349,
+      "grad_norm": 3.6307485103607178,
+      "learning_rate": 4.315510257681381e-06,
+      "loss": 1.0208,
+      "step": 7231
+    },
+    {
+      "epoch": 0.5575084798026518,
+      "grad_norm": 3.832481622695923,
+      "learning_rate": 4.314273585719749e-06,
+      "loss": 1.0138,
+      "step": 7232
+    },
+    {
+      "epoch": 0.5575855689176689,
+      "grad_norm": 3.533522367477417,
+      "learning_rate": 4.313036956509605e-06,
+      "loss": 1.0133,
+      "step": 7233
+    },
+    {
+      "epoch": 0.5576626580326858,
+      "grad_norm": 3.565697193145752,
+      "learning_rate": 4.3118003701280504e-06,
+      "loss": 1.0029,
+      "step": 7234
+    },
+    {
+      "epoch": 0.5577397471477027,
+      "grad_norm": 3.788064956665039,
+      "learning_rate": 4.310563826652175e-06,
+      "loss": 0.9481,
+      "step": 7235
+    },
+    {
+      "epoch": 0.5578168362627197,
+      "grad_norm": 3.4550724029541016,
+      "learning_rate": 4.3093273261590785e-06,
+      "loss": 0.8953,
+      "step": 7236
+    },
+    {
+      "epoch": 0.5578939253777366,
+      "grad_norm": 3.3875648975372314,
+      "learning_rate": 4.308090868725844e-06,
+      "loss": 0.928,
+      "step": 7237
+    },
+    {
+      "epoch": 0.5579710144927537,
+      "grad_norm": 4.244081497192383,
+      "learning_rate": 4.306854454429561e-06,
+      "loss": 0.9824,
+      "step": 7238
+    },
+    {
+      "epoch": 0.5580481036077706,
+      "grad_norm": 3.3602986335754395,
+      "learning_rate": 4.3056180833473135e-06,
+      "loss": 0.812,
+      "step": 7239
+    },
+    {
+      "epoch": 0.5581251927227875,
+      "grad_norm": 3.6101772785186768,
+      "learning_rate": 4.304381755556182e-06,
+      "loss": 0.8871,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5582022818378045,
+      "grad_norm": 3.709418535232544,
+      "learning_rate": 4.303145471133246e-06,
+      "loss": 0.9672,
+      "step": 7241
+    },
+    {
+      "epoch": 0.5582793709528214,
+      "grad_norm": 3.9773929119110107,
+      "learning_rate": 4.301909230155579e-06,
+      "loss": 0.9546,
+      "step": 7242
+    },
+    {
+      "epoch": 0.5583564600678385,
+      "grad_norm": 3.4679555892944336,
+      "learning_rate": 4.300673032700256e-06,
+      "loss": 0.8538,
+      "step": 7243
+    },
+    {
+      "epoch": 0.5584335491828554,
+      "grad_norm": 3.525493860244751,
+      "learning_rate": 4.2994368788443496e-06,
+      "loss": 0.9495,
+      "step": 7244
+    },
+    {
+      "epoch": 0.5585106382978723,
+      "grad_norm": 3.462528705596924,
+      "learning_rate": 4.298200768664924e-06,
+      "loss": 0.9793,
+      "step": 7245
+    },
+    {
+      "epoch": 0.5585877274128893,
+      "grad_norm": 3.402458906173706,
+      "learning_rate": 4.296964702239046e-06,
+      "loss": 0.8481,
+      "step": 7246
+    },
+    {
+      "epoch": 0.5586648165279062,
+      "grad_norm": 3.4688830375671387,
+      "learning_rate": 4.295728679643778e-06,
+      "loss": 0.898,
+      "step": 7247
+    },
+    {
+      "epoch": 0.5587419056429233,
+      "grad_norm": 3.5132532119750977,
+      "learning_rate": 4.2944927009561786e-06,
+      "loss": 0.9714,
+      "step": 7248
+    },
+    {
+      "epoch": 0.5588189947579402,
+      "grad_norm": 3.7128775119781494,
+      "learning_rate": 4.293256766253307e-06,
+      "loss": 1.0389,
+      "step": 7249
+    },
+    {
+      "epoch": 0.5588960838729571,
+      "grad_norm": 3.6854405403137207,
+      "learning_rate": 4.2920208756122136e-06,
+      "loss": 0.9308,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5589731729879741,
+      "grad_norm": 3.7184994220733643,
+      "learning_rate": 4.290785029109953e-06,
+      "loss": 0.9581,
+      "step": 7251
+    },
+    {
+      "epoch": 0.559050262102991,
+      "grad_norm": 3.4944541454315186,
+      "learning_rate": 4.2895492268235725e-06,
+      "loss": 0.9439,
+      "step": 7252
+    },
+    {
+      "epoch": 0.559127351218008,
+      "grad_norm": 3.616788387298584,
+      "learning_rate": 4.288313468830119e-06,
+      "loss": 0.8744,
+      "step": 7253
+    },
+    {
+      "epoch": 0.559204440333025,
+      "grad_norm": 3.7554306983947754,
+      "learning_rate": 4.2870777552066336e-06,
+      "loss": 0.8796,
+      "step": 7254
+    },
+    {
+      "epoch": 0.559281529448042,
+      "grad_norm": 3.7401881217956543,
+      "learning_rate": 4.285842086030159e-06,
+      "loss": 0.8975,
+      "step": 7255
+    },
+    {
+      "epoch": 0.5593586185630589,
+      "grad_norm": 3.8018808364868164,
+      "learning_rate": 4.28460646137773e-06,
+      "loss": 0.827,
+      "step": 7256
+    },
+    {
+      "epoch": 0.5594357076780758,
+      "grad_norm": 3.6006662845611572,
+      "learning_rate": 4.283370881326384e-06,
+      "loss": 0.9216,
+      "step": 7257
+    },
+    {
+      "epoch": 0.5595127967930928,
+      "grad_norm": 3.953252077102661,
+      "learning_rate": 4.2821353459531525e-06,
+      "loss": 0.9649,
+      "step": 7258
+    },
+    {
+      "epoch": 0.5595898859081098,
+      "grad_norm": 3.724891185760498,
+      "learning_rate": 4.2808998553350625e-06,
+      "loss": 1.0039,
+      "step": 7259
+    },
+    {
+      "epoch": 0.5596669750231268,
+      "grad_norm": 3.795192241668701,
+      "learning_rate": 4.279664409549145e-06,
+      "loss": 0.9401,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5597440641381437,
+      "grad_norm": 4.127142906188965,
+      "learning_rate": 4.2784290086724186e-06,
+      "loss": 0.9922,
+      "step": 7261
+    },
+    {
+      "epoch": 0.5598211532531606,
+      "grad_norm": 3.5862457752227783,
+      "learning_rate": 4.277193652781906e-06,
+      "loss": 0.9433,
+      "step": 7262
+    },
+    {
+      "epoch": 0.5598982423681776,
+      "grad_norm": 4.025807857513428,
+      "learning_rate": 4.275958341954628e-06,
+      "loss": 0.8679,
+      "step": 7263
+    },
+    {
+      "epoch": 0.5599753314831946,
+      "grad_norm": 3.7022628784179688,
+      "learning_rate": 4.274723076267596e-06,
+      "loss": 0.8451,
+      "step": 7264
+    },
+    {
+      "epoch": 0.5600524205982116,
+      "grad_norm": 3.6183743476867676,
+      "learning_rate": 4.273487855797823e-06,
+      "loss": 0.8063,
+      "step": 7265
+    },
+    {
+      "epoch": 0.5601295097132285,
+      "grad_norm": 3.7264175415039062,
+      "learning_rate": 4.272252680622321e-06,
+      "loss": 0.8843,
+      "step": 7266
+    },
+    {
+      "epoch": 0.5602065988282454,
+      "grad_norm": 3.512242317199707,
+      "learning_rate": 4.271017550818095e-06,
+      "loss": 0.925,
+      "step": 7267
+    },
+    {
+      "epoch": 0.5602836879432624,
+      "grad_norm": 3.9300243854522705,
+      "learning_rate": 4.269782466462149e-06,
+      "loss": 0.949,
+      "step": 7268
+    },
+    {
+      "epoch": 0.5603607770582794,
+      "grad_norm": 3.0627076625823975,
+      "learning_rate": 4.268547427631485e-06,
+      "loss": 0.8389,
+      "step": 7269
+    },
+    {
+      "epoch": 0.5604378661732964,
+      "grad_norm": 3.468635082244873,
+      "learning_rate": 4.267312434403099e-06,
+      "loss": 0.911,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5605149552883133,
+      "grad_norm": 3.891932249069214,
+      "learning_rate": 4.26607748685399e-06,
+      "loss": 0.9965,
+      "step": 7271
+    },
+    {
+      "epoch": 0.5605920444033302,
+      "grad_norm": 3.7001166343688965,
+      "learning_rate": 4.2648425850611465e-06,
+      "loss": 0.8025,
+      "step": 7272
+    },
+    {
+      "epoch": 0.5606691335183472,
+      "grad_norm": 3.481006145477295,
+      "learning_rate": 4.2636077291015615e-06,
+      "loss": 0.8563,
+      "step": 7273
+    },
+    {
+      "epoch": 0.5607462226333642,
+      "grad_norm": 3.4894566535949707,
+      "learning_rate": 4.262372919052221e-06,
+      "loss": 0.9797,
+      "step": 7274
+    },
+    {
+      "epoch": 0.5608233117483812,
+      "grad_norm": 3.728573799133301,
+      "learning_rate": 4.2611381549901085e-06,
+      "loss": 0.8734,
+      "step": 7275
+    },
+    {
+      "epoch": 0.5609004008633981,
+      "grad_norm": 3.5050055980682373,
+      "learning_rate": 4.259903436992204e-06,
+      "loss": 0.8772,
+      "step": 7276
+    },
+    {
+      "epoch": 0.560977489978415,
+      "grad_norm": 3.7864038944244385,
+      "learning_rate": 4.258668765135489e-06,
+      "loss": 0.939,
+      "step": 7277
+    },
+    {
+      "epoch": 0.561054579093432,
+      "grad_norm": 3.818662405014038,
+      "learning_rate": 4.257434139496937e-06,
+      "loss": 1.0586,
+      "step": 7278
+    },
+    {
+      "epoch": 0.561131668208449,
+      "grad_norm": 4.216384410858154,
+      "learning_rate": 4.25619956015352e-06,
+      "loss": 1.0841,
+      "step": 7279
+    },
+    {
+      "epoch": 0.561208757323466,
+      "grad_norm": 3.877438545227051,
+      "learning_rate": 4.254965027182206e-06,
+      "loss": 0.8282,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5612858464384829,
+      "grad_norm": 3.9610254764556885,
+      "learning_rate": 4.253730540659966e-06,
+      "loss": 1.0798,
+      "step": 7281
+    },
+    {
+      "epoch": 0.5613629355534998,
+      "grad_norm": 3.5456466674804688,
+      "learning_rate": 4.252496100663762e-06,
+      "loss": 0.8978,
+      "step": 7282
+    },
+    {
+      "epoch": 0.5614400246685168,
+      "grad_norm": 3.5786006450653076,
+      "learning_rate": 4.2512617072705546e-06,
+      "loss": 0.956,
+      "step": 7283
+    },
+    {
+      "epoch": 0.5615171137835338,
+      "grad_norm": 3.5269174575805664,
+      "learning_rate": 4.250027360557302e-06,
+      "loss": 0.8546,
+      "step": 7284
+    },
+    {
+      "epoch": 0.5615942028985508,
+      "grad_norm": 3.479971408843994,
+      "learning_rate": 4.248793060600959e-06,
+      "loss": 0.832,
+      "step": 7285
+    },
+    {
+      "epoch": 0.5616712920135677,
+      "grad_norm": 3.701209783554077,
+      "learning_rate": 4.247558807478478e-06,
+      "loss": 1.0244,
+      "step": 7286
+    },
+    {
+      "epoch": 0.5617483811285846,
+      "grad_norm": 3.449061393737793,
+      "learning_rate": 4.246324601266809e-06,
+      "loss": 0.8604,
+      "step": 7287
+    },
+    {
+      "epoch": 0.5618254702436016,
+      "grad_norm": 3.8695156574249268,
+      "learning_rate": 4.245090442042897e-06,
+      "loss": 1.0627,
+      "step": 7288
+    },
+    {
+      "epoch": 0.5619025593586185,
+      "grad_norm": 3.7182259559631348,
+      "learning_rate": 4.243856329883687e-06,
+      "loss": 0.9507,
+      "step": 7289
+    },
+    {
+      "epoch": 0.5619796484736356,
+      "grad_norm": 3.4913182258605957,
+      "learning_rate": 4.242622264866118e-06,
+      "loss": 0.9745,
+      "step": 7290
+    },
+    {
+      "epoch": 0.5620567375886525,
+      "grad_norm": 3.5762779712677,
+      "learning_rate": 4.241388247067128e-06,
+      "loss": 0.899,
+      "step": 7291
+    },
+    {
+      "epoch": 0.5621338267036694,
+      "grad_norm": 3.425102472305298,
+      "learning_rate": 4.240154276563652e-06,
+      "loss": 0.9087,
+      "step": 7292
+    },
+    {
+      "epoch": 0.5622109158186864,
+      "grad_norm": 3.7708728313446045,
+      "learning_rate": 4.238920353432623e-06,
+      "loss": 0.9664,
+      "step": 7293
+    },
+    {
+      "epoch": 0.5622880049337033,
+      "grad_norm": 3.508423089981079,
+      "learning_rate": 4.237686477750966e-06,
+      "loss": 0.9499,
+      "step": 7294
+    },
+    {
+      "epoch": 0.5623650940487204,
+      "grad_norm": 3.62243390083313,
+      "learning_rate": 4.23645264959561e-06,
+      "loss": 0.9453,
+      "step": 7295
+    },
+    {
+      "epoch": 0.5624421831637373,
+      "grad_norm": 3.7550387382507324,
+      "learning_rate": 4.235218869043477e-06,
+      "loss": 0.9756,
+      "step": 7296
+    },
+    {
+      "epoch": 0.5625192722787542,
+      "grad_norm": 3.69486927986145,
+      "learning_rate": 4.233985136171487e-06,
+      "loss": 0.9725,
+      "step": 7297
+    },
+    {
+      "epoch": 0.5625963613937712,
+      "grad_norm": 3.8165245056152344,
+      "learning_rate": 4.232751451056555e-06,
+      "loss": 0.917,
+      "step": 7298
+    },
+    {
+      "epoch": 0.5626734505087881,
+      "grad_norm": 3.5180726051330566,
+      "learning_rate": 4.231517813775597e-06,
+      "loss": 0.8487,
+      "step": 7299
+    },
+    {
+      "epoch": 0.5627505396238052,
+      "grad_norm": 3.76409912109375,
+      "learning_rate": 4.230284224405523e-06,
+      "loss": 0.9902,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5628276287388221,
+      "grad_norm": 3.4409983158111572,
+      "learning_rate": 4.2290506830232415e-06,
+      "loss": 0.8749,
+      "step": 7301
+    },
+    {
+      "epoch": 0.562904717853839,
+      "grad_norm": 3.861907958984375,
+      "learning_rate": 4.227817189705657e-06,
+      "loss": 0.8812,
+      "step": 7302
+    },
+    {
+      "epoch": 0.562981806968856,
+      "grad_norm": 3.839994192123413,
+      "learning_rate": 4.226583744529672e-06,
+      "loss": 0.954,
+      "step": 7303
+    },
+    {
+      "epoch": 0.5630588960838729,
+      "grad_norm": 4.001551151275635,
+      "learning_rate": 4.225350347572185e-06,
+      "loss": 0.8083,
+      "step": 7304
+    },
+    {
+      "epoch": 0.56313598519889,
+      "grad_norm": 3.7375106811523438,
+      "learning_rate": 4.22411699891009e-06,
+      "loss": 1.0402,
+      "step": 7305
+    },
+    {
+      "epoch": 0.5632130743139069,
+      "grad_norm": 3.5613088607788086,
+      "learning_rate": 4.2228836986202845e-06,
+      "loss": 0.9585,
+      "step": 7306
+    },
+    {
+      "epoch": 0.5632901634289238,
+      "grad_norm": 3.4407291412353516,
+      "learning_rate": 4.221650446779653e-06,
+      "loss": 0.9678,
+      "step": 7307
+    },
+    {
+      "epoch": 0.5633672525439408,
+      "grad_norm": 3.501506805419922,
+      "learning_rate": 4.2204172434650866e-06,
+      "loss": 0.9946,
+      "step": 7308
+    },
+    {
+      "epoch": 0.5634443416589577,
+      "grad_norm": 3.91636061668396,
+      "learning_rate": 4.219184088753467e-06,
+      "loss": 1.0566,
+      "step": 7309
+    },
+    {
+      "epoch": 0.5635214307739748,
+      "grad_norm": 3.84093976020813,
+      "learning_rate": 4.217950982721675e-06,
+      "loss": 0.9006,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5635985198889917,
+      "grad_norm": 3.702842950820923,
+      "learning_rate": 4.21671792544659e-06,
+      "loss": 0.9995,
+      "step": 7311
+    },
+    {
+      "epoch": 0.5636756090040086,
+      "grad_norm": 3.6409497261047363,
+      "learning_rate": 4.215484917005085e-06,
+      "loss": 0.9295,
+      "step": 7312
+    },
+    {
+      "epoch": 0.5637526981190256,
+      "grad_norm": 3.425840377807617,
+      "learning_rate": 4.2142519574740315e-06,
+      "loss": 0.8669,
+      "step": 7313
+    },
+    {
+      "epoch": 0.5638297872340425,
+      "grad_norm": 3.576141595840454,
+      "learning_rate": 4.2130190469303005e-06,
+      "loss": 0.9028,
+      "step": 7314
+    },
+    {
+      "epoch": 0.5639068763490596,
+      "grad_norm": 3.4973714351654053,
+      "learning_rate": 4.211786185450756e-06,
+      "loss": 0.8018,
+      "step": 7315
+    },
+    {
+      "epoch": 0.5639839654640765,
+      "grad_norm": 3.3027119636535645,
+      "learning_rate": 4.210553373112259e-06,
+      "loss": 0.8958,
+      "step": 7316
+    },
+    {
+      "epoch": 0.5640610545790934,
+      "grad_norm": 4.032995223999023,
+      "learning_rate": 4.209320609991672e-06,
+      "loss": 0.9982,
+      "step": 7317
+    },
+    {
+      "epoch": 0.5641381436941104,
+      "grad_norm": 3.718665838241577,
+      "learning_rate": 4.208087896165849e-06,
+      "loss": 0.9066,
+      "step": 7318
+    },
+    {
+      "epoch": 0.5642152328091273,
+      "grad_norm": 3.670677900314331,
+      "learning_rate": 4.206855231711645e-06,
+      "loss": 1.0318,
+      "step": 7319
+    },
+    {
+      "epoch": 0.5642923219241444,
+      "grad_norm": 3.2902212142944336,
+      "learning_rate": 4.205622616705909e-06,
+      "loss": 0.896,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5643694110391613,
+      "grad_norm": 3.813958168029785,
+      "learning_rate": 4.204390051225488e-06,
+      "loss": 1.0047,
+      "step": 7321
+    },
+    {
+      "epoch": 0.5644465001541782,
+      "grad_norm": 3.3876993656158447,
+      "learning_rate": 4.203157535347229e-06,
+      "loss": 0.9258,
+      "step": 7322
+    },
+    {
+      "epoch": 0.5645235892691952,
+      "grad_norm": 3.70788311958313,
+      "learning_rate": 4.201925069147969e-06,
+      "loss": 0.962,
+      "step": 7323
+    },
+    {
+      "epoch": 0.5646006783842121,
+      "grad_norm": 3.678077459335327,
+      "learning_rate": 4.200692652704545e-06,
+      "loss": 1.0408,
+      "step": 7324
+    },
+    {
+      "epoch": 0.5646777674992292,
+      "grad_norm": 3.8763816356658936,
+      "learning_rate": 4.199460286093797e-06,
+      "loss": 0.9782,
+      "step": 7325
+    },
+    {
+      "epoch": 0.5647548566142461,
+      "grad_norm": 3.756704807281494,
+      "learning_rate": 4.198227969392556e-06,
+      "loss": 0.9667,
+      "step": 7326
+    },
+    {
+      "epoch": 0.564831945729263,
+      "grad_norm": 3.946401596069336,
+      "learning_rate": 4.196995702677646e-06,
+      "loss": 0.9345,
+      "step": 7327
+    },
+    {
+      "epoch": 0.56490903484428,
+      "grad_norm": 3.7415993213653564,
+      "learning_rate": 4.195763486025896e-06,
+      "loss": 0.9629,
+      "step": 7328
+    },
+    {
+      "epoch": 0.5649861239592969,
+      "grad_norm": 3.9311485290527344,
+      "learning_rate": 4.194531319514128e-06,
+      "loss": 0.9507,
+      "step": 7329
+    },
+    {
+      "epoch": 0.565063213074314,
+      "grad_norm": 3.924248456954956,
+      "learning_rate": 4.1932992032191595e-06,
+      "loss": 0.9141,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5651403021893309,
+      "grad_norm": 3.478628396987915,
+      "learning_rate": 4.1920671372178075e-06,
+      "loss": 1.0346,
+      "step": 7331
+    },
+    {
+      "epoch": 0.5652173913043478,
+      "grad_norm": 3.4320709705352783,
+      "learning_rate": 4.190835121586887e-06,
+      "loss": 0.9594,
+      "step": 7332
+    },
+    {
+      "epoch": 0.5652944804193648,
+      "grad_norm": 3.83650541305542,
+      "learning_rate": 4.189603156403204e-06,
+      "loss": 0.9973,
+      "step": 7333
+    },
+    {
+      "epoch": 0.5653715695343817,
+      "grad_norm": 3.5375068187713623,
+      "learning_rate": 4.188371241743568e-06,
+      "loss": 0.9137,
+      "step": 7334
+    },
+    {
+      "epoch": 0.5654486586493987,
+      "grad_norm": 4.036672592163086,
+      "learning_rate": 4.187139377684781e-06,
+      "loss": 1.0529,
+      "step": 7335
+    },
+    {
+      "epoch": 0.5655257477644157,
+      "grad_norm": 3.483830213546753,
+      "learning_rate": 4.185907564303644e-06,
+      "loss": 0.8893,
+      "step": 7336
+    },
+    {
+      "epoch": 0.5656028368794326,
+      "grad_norm": 3.592637300491333,
+      "learning_rate": 4.184675801676955e-06,
+      "loss": 1.0526,
+      "step": 7337
+    },
+    {
+      "epoch": 0.5656799259944496,
+      "grad_norm": 3.425968647003174,
+      "learning_rate": 4.183444089881506e-06,
+      "loss": 0.9941,
+      "step": 7338
+    },
+    {
+      "epoch": 0.5657570151094665,
+      "grad_norm": 3.730457305908203,
+      "learning_rate": 4.182212428994088e-06,
+      "loss": 0.9906,
+      "step": 7339
+    },
+    {
+      "epoch": 0.5658341042244835,
+      "grad_norm": 3.9241623878479004,
+      "learning_rate": 4.1809808190914925e-06,
+      "loss": 0.9277,
+      "step": 7340
+    },
+    {
+      "epoch": 0.5659111933395005,
+      "grad_norm": 3.7960057258605957,
+      "learning_rate": 4.1797492602504985e-06,
+      "loss": 0.8966,
+      "step": 7341
+    },
+    {
+      "epoch": 0.5659882824545174,
+      "grad_norm": 3.747955083847046,
+      "learning_rate": 4.178517752547891e-06,
+      "loss": 1.0132,
+      "step": 7342
+    },
+    {
+      "epoch": 0.5660653715695344,
+      "grad_norm": 3.649674654006958,
+      "learning_rate": 4.177286296060448e-06,
+      "loss": 0.8899,
+      "step": 7343
+    },
+    {
+      "epoch": 0.5661424606845513,
+      "grad_norm": 3.396299123764038,
+      "learning_rate": 4.176054890864942e-06,
+      "loss": 0.9728,
+      "step": 7344
+    },
+    {
+      "epoch": 0.5662195497995683,
+      "grad_norm": 3.8877921104431152,
+      "learning_rate": 4.174823537038147e-06,
+      "loss": 1.0016,
+      "step": 7345
+    },
+    {
+      "epoch": 0.5662966389145853,
+      "grad_norm": 3.6778106689453125,
+      "learning_rate": 4.173592234656831e-06,
+      "loss": 0.9368,
+      "step": 7346
+    },
+    {
+      "epoch": 0.5663737280296022,
+      "grad_norm": 3.6797902584075928,
+      "learning_rate": 4.172360983797757e-06,
+      "loss": 0.9625,
+      "step": 7347
+    },
+    {
+      "epoch": 0.5664508171446192,
+      "grad_norm": 4.3902268409729,
+      "learning_rate": 4.171129784537692e-06,
+      "loss": 0.961,
+      "step": 7348
+    },
+    {
+      "epoch": 0.5665279062596361,
+      "grad_norm": 3.7198588848114014,
+      "learning_rate": 4.16989863695339e-06,
+      "loss": 0.9612,
+      "step": 7349
+    },
+    {
+      "epoch": 0.5666049953746531,
+      "grad_norm": 3.431483507156372,
+      "learning_rate": 4.16866754112161e-06,
+      "loss": 0.8565,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5666820844896701,
+      "grad_norm": 3.6897926330566406,
+      "learning_rate": 4.167436497119103e-06,
+      "loss": 0.9245,
+      "step": 7351
+    },
+    {
+      "epoch": 0.566759173604687,
+      "grad_norm": 3.8588991165161133,
+      "learning_rate": 4.166205505022618e-06,
+      "loss": 1.04,
+      "step": 7352
+    },
+    {
+      "epoch": 0.566836262719704,
+      "grad_norm": 3.7585813999176025,
+      "learning_rate": 4.1649745649089015e-06,
+      "loss": 0.8811,
+      "step": 7353
+    },
+    {
+      "epoch": 0.5669133518347209,
+      "grad_norm": 3.642169713973999,
+      "learning_rate": 4.163743676854697e-06,
+      "loss": 0.8462,
+      "step": 7354
+    },
+    {
+      "epoch": 0.5669904409497379,
+      "grad_norm": 3.2136752605438232,
+      "learning_rate": 4.162512840936742e-06,
+      "loss": 0.9009,
+      "step": 7355
+    },
+    {
+      "epoch": 0.5670675300647549,
+      "grad_norm": 3.7039544582366943,
+      "learning_rate": 4.161282057231776e-06,
+      "loss": 0.9187,
+      "step": 7356
+    },
+    {
+      "epoch": 0.5671446191797718,
+      "grad_norm": 3.5421409606933594,
+      "learning_rate": 4.160051325816528e-06,
+      "loss": 0.9465,
+      "step": 7357
+    },
+    {
+      "epoch": 0.5672217082947888,
+      "grad_norm": 3.6218063831329346,
+      "learning_rate": 4.158820646767729e-06,
+      "loss": 1.0368,
+      "step": 7358
+    },
+    {
+      "epoch": 0.5672987974098057,
+      "grad_norm": 3.5446629524230957,
+      "learning_rate": 4.157590020162108e-06,
+      "loss": 1.0172,
+      "step": 7359
+    },
+    {
+      "epoch": 0.5673758865248227,
+      "grad_norm": 3.736348867416382,
+      "learning_rate": 4.156359446076385e-06,
+      "loss": 0.9261,
+      "step": 7360
+    },
+    {
+      "epoch": 0.5674529756398397,
+      "grad_norm": 3.8830409049987793,
+      "learning_rate": 4.1551289245872815e-06,
+      "loss": 1.0677,
+      "step": 7361
+    },
+    {
+      "epoch": 0.5675300647548566,
+      "grad_norm": 3.8011181354522705,
+      "learning_rate": 4.153898455771514e-06,
+      "loss": 0.9841,
+      "step": 7362
+    },
+    {
+      "epoch": 0.5676071538698736,
+      "grad_norm": 3.9767227172851562,
+      "learning_rate": 4.152668039705796e-06,
+      "loss": 1.0165,
+      "step": 7363
+    },
+    {
+      "epoch": 0.5676842429848905,
+      "grad_norm": 3.4617295265197754,
+      "learning_rate": 4.151437676466836e-06,
+      "loss": 0.9734,
+      "step": 7364
+    },
+    {
+      "epoch": 0.5677613320999075,
+      "grad_norm": 3.2695651054382324,
+      "learning_rate": 4.150207366131344e-06,
+      "loss": 0.9571,
+      "step": 7365
+    },
+    {
+      "epoch": 0.5678384212149244,
+      "grad_norm": 3.5925192832946777,
+      "learning_rate": 4.14897710877602e-06,
+      "loss": 0.9581,
+      "step": 7366
+    },
+    {
+      "epoch": 0.5679155103299414,
+      "grad_norm": 3.435485363006592,
+      "learning_rate": 4.147746904477567e-06,
+      "loss": 0.9586,
+      "step": 7367
+    },
+    {
+      "epoch": 0.5679925994449584,
+      "grad_norm": 3.665407657623291,
+      "learning_rate": 4.146516753312677e-06,
+      "loss": 1.023,
+      "step": 7368
+    },
+    {
+      "epoch": 0.5680696885599753,
+      "grad_norm": 3.7188949584960938,
+      "learning_rate": 4.145286655358051e-06,
+      "loss": 0.9351,
+      "step": 7369
+    },
+    {
+      "epoch": 0.5681467776749923,
+      "grad_norm": 4.338947296142578,
+      "learning_rate": 4.144056610690375e-06,
+      "loss": 0.9218,
+      "step": 7370
+    },
+    {
+      "epoch": 0.5682238667900092,
+      "grad_norm": 4.026747703552246,
+      "learning_rate": 4.142826619386334e-06,
+      "loss": 1.0557,
+      "step": 7371
+    },
+    {
+      "epoch": 0.5683009559050262,
+      "grad_norm": 3.3105926513671875,
+      "learning_rate": 4.141596681522616e-06,
+      "loss": 0.8553,
+      "step": 7372
+    },
+    {
+      "epoch": 0.5683780450200432,
+      "grad_norm": 3.3676323890686035,
+      "learning_rate": 4.140366797175899e-06,
+      "loss": 0.8005,
+      "step": 7373
+    },
+    {
+      "epoch": 0.5684551341350601,
+      "grad_norm": 3.641796588897705,
+      "learning_rate": 4.13913696642286e-06,
+      "loss": 0.8849,
+      "step": 7374
+    },
+    {
+      "epoch": 0.5685322232500771,
+      "grad_norm": 3.88008975982666,
+      "learning_rate": 4.137907189340172e-06,
+      "loss": 0.9257,
+      "step": 7375
+    },
+    {
+      "epoch": 0.568609312365094,
+      "grad_norm": 3.845065116882324,
+      "learning_rate": 4.136677466004506e-06,
+      "loss": 0.9235,
+      "step": 7376
+    },
+    {
+      "epoch": 0.568686401480111,
+      "grad_norm": 3.5287108421325684,
+      "learning_rate": 4.13544779649253e-06,
+      "loss": 0.9357,
+      "step": 7377
+    },
+    {
+      "epoch": 0.568763490595128,
+      "grad_norm": 3.5272388458251953,
+      "learning_rate": 4.1342181808809046e-06,
+      "loss": 0.9282,
+      "step": 7378
+    },
+    {
+      "epoch": 0.5688405797101449,
+      "grad_norm": 4.141164779663086,
+      "learning_rate": 4.132988619246291e-06,
+      "loss": 0.8634,
+      "step": 7379
+    },
+    {
+      "epoch": 0.5689176688251619,
+      "grad_norm": 3.388904571533203,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.9187,
+      "step": 7380
+    },
+    {
+      "epoch": 0.5689947579401788,
+      "grad_norm": 3.592479944229126,
+      "learning_rate": 4.130529658214728e-06,
+      "loss": 0.9531,
+      "step": 7381
+    },
+    {
+      "epoch": 0.5690718470551958,
+      "grad_norm": 3.7211825847625732,
+      "learning_rate": 4.1293002589710795e-06,
+      "loss": 1.0478,
+      "step": 7382
+    },
+    {
+      "epoch": 0.5691489361702128,
+      "grad_norm": 3.6857645511627197,
+      "learning_rate": 4.128070914011052e-06,
+      "loss": 1.0107,
+      "step": 7383
+    },
+    {
+      "epoch": 0.5692260252852297,
+      "grad_norm": 3.7964882850646973,
+      "learning_rate": 4.126841623411286e-06,
+      "loss": 1.0194,
+      "step": 7384
+    },
+    {
+      "epoch": 0.5693031144002467,
+      "grad_norm": 3.608358860015869,
+      "learning_rate": 4.125612387248423e-06,
+      "loss": 0.8752,
+      "step": 7385
+    },
+    {
+      "epoch": 0.5693802035152636,
+      "grad_norm": 3.4516360759735107,
+      "learning_rate": 4.124383205599099e-06,
+      "loss": 0.9234,
+      "step": 7386
+    },
+    {
+      "epoch": 0.5694572926302806,
+      "grad_norm": 3.3848817348480225,
+      "learning_rate": 4.123154078539946e-06,
+      "loss": 0.8775,
+      "step": 7387
+    },
+    {
+      "epoch": 0.5695343817452976,
+      "grad_norm": 3.8858206272125244,
+      "learning_rate": 4.121925006147597e-06,
+      "loss": 1.0166,
+      "step": 7388
+    },
+    {
+      "epoch": 0.5696114708603145,
+      "grad_norm": 3.7127907276153564,
+      "learning_rate": 4.120695988498674e-06,
+      "loss": 0.9005,
+      "step": 7389
+    },
+    {
+      "epoch": 0.5696885599753315,
+      "grad_norm": 3.9486067295074463,
+      "learning_rate": 4.119467025669803e-06,
+      "loss": 0.9782,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5697656490903484,
+      "grad_norm": 3.396491765975952,
+      "learning_rate": 4.118238117737604e-06,
+      "loss": 0.8566,
+      "step": 7391
+    },
+    {
+      "epoch": 0.5698427382053654,
+      "grad_norm": 3.7514426708221436,
+      "learning_rate": 4.1170092647786895e-06,
+      "loss": 0.9061,
+      "step": 7392
+    },
+    {
+      "epoch": 0.5699198273203824,
+      "grad_norm": 3.4897749423980713,
+      "learning_rate": 4.115780466869676e-06,
+      "loss": 0.9237,
+      "step": 7393
+    },
+    {
+      "epoch": 0.5699969164353993,
+      "grad_norm": 3.4681146144866943,
+      "learning_rate": 4.11455172408717e-06,
+      "loss": 0.9261,
+      "step": 7394
+    },
+    {
+      "epoch": 0.5700740055504163,
+      "grad_norm": 3.301769733428955,
+      "learning_rate": 4.113323036507778e-06,
+      "loss": 0.8158,
+      "step": 7395
+    },
+    {
+      "epoch": 0.5701510946654332,
+      "grad_norm": 3.7004668712615967,
+      "learning_rate": 4.112094404208102e-06,
+      "loss": 0.8965,
+      "step": 7396
+    },
+    {
+      "epoch": 0.5702281837804501,
+      "grad_norm": 3.7279725074768066,
+      "learning_rate": 4.110865827264742e-06,
+      "loss": 0.9718,
+      "step": 7397
+    },
+    {
+      "epoch": 0.5703052728954672,
+      "grad_norm": 3.387369394302368,
+      "learning_rate": 4.109637305754293e-06,
+      "loss": 0.9405,
+      "step": 7398
+    },
+    {
+      "epoch": 0.5703823620104841,
+      "grad_norm": 3.686414957046509,
+      "learning_rate": 4.108408839753346e-06,
+      "loss": 0.8854,
+      "step": 7399
+    },
+    {
+      "epoch": 0.5704594511255011,
+      "grad_norm": 3.8577775955200195,
+      "learning_rate": 4.107180429338491e-06,
+      "loss": 0.9723,
+      "step": 7400
+    },
+    {
+      "epoch": 0.570536540240518,
+      "grad_norm": 3.784442901611328,
+      "learning_rate": 4.105952074586311e-06,
+      "loss": 0.8432,
+      "step": 7401
+    },
+    {
+      "epoch": 0.570613629355535,
+      "grad_norm": 4.02522611618042,
+      "learning_rate": 4.10472377557339e-06,
+      "loss": 0.9356,
+      "step": 7402
+    },
+    {
+      "epoch": 0.570690718470552,
+      "grad_norm": 4.0557966232299805,
+      "learning_rate": 4.103495532376304e-06,
+      "loss": 0.9714,
+      "step": 7403
+    },
+    {
+      "epoch": 0.5707678075855689,
+      "grad_norm": 4.144707679748535,
+      "learning_rate": 4.1022673450716295e-06,
+      "loss": 1.0333,
+      "step": 7404
+    },
+    {
+      "epoch": 0.5708448967005859,
+      "grad_norm": 3.7137889862060547,
+      "learning_rate": 4.101039213735935e-06,
+      "loss": 1.0662,
+      "step": 7405
+    },
+    {
+      "epoch": 0.5709219858156028,
+      "grad_norm": 3.3066835403442383,
+      "learning_rate": 4.09981113844579e-06,
+      "loss": 0.7372,
+      "step": 7406
+    },
+    {
+      "epoch": 0.5709990749306197,
+      "grad_norm": 3.355921983718872,
+      "learning_rate": 4.098583119277759e-06,
+      "loss": 0.9921,
+      "step": 7407
+    },
+    {
+      "epoch": 0.5710761640456368,
+      "grad_norm": 3.732572555541992,
+      "learning_rate": 4.097355156308402e-06,
+      "loss": 0.9918,
+      "step": 7408
+    },
+    {
+      "epoch": 0.5711532531606537,
+      "grad_norm": 3.7717831134796143,
+      "learning_rate": 4.096127249614276e-06,
+      "loss": 0.8853,
+      "step": 7409
+    },
+    {
+      "epoch": 0.5712303422756707,
+      "grad_norm": 3.5494415760040283,
+      "learning_rate": 4.094899399271935e-06,
+      "loss": 1.0293,
+      "step": 7410
+    },
+    {
+      "epoch": 0.5713074313906876,
+      "grad_norm": 3.6082026958465576,
+      "learning_rate": 4.093671605357928e-06,
+      "loss": 0.9877,
+      "step": 7411
+    },
+    {
+      "epoch": 0.5713845205057045,
+      "grad_norm": 3.492135524749756,
+      "learning_rate": 4.092443867948801e-06,
+      "loss": 0.8388,
+      "step": 7412
+    },
+    {
+      "epoch": 0.5714616096207216,
+      "grad_norm": 3.4625587463378906,
+      "learning_rate": 4.0912161871211e-06,
+      "loss": 0.8672,
+      "step": 7413
+    },
+    {
+      "epoch": 0.5715386987357385,
+      "grad_norm": 3.4896044731140137,
+      "learning_rate": 4.089988562951363e-06,
+      "loss": 0.933,
+      "step": 7414
+    },
+    {
+      "epoch": 0.5716157878507555,
+      "grad_norm": 3.792921781539917,
+      "learning_rate": 4.088760995516127e-06,
+      "loss": 0.9755,
+      "step": 7415
+    },
+    {
+      "epoch": 0.5716928769657724,
+      "grad_norm": 4.106998443603516,
+      "learning_rate": 4.0875334848919225e-06,
+      "loss": 0.9181,
+      "step": 7416
+    },
+    {
+      "epoch": 0.5717699660807893,
+      "grad_norm": 3.3308520317077637,
+      "learning_rate": 4.08630603115528e-06,
+      "loss": 0.9823,
+      "step": 7417
+    },
+    {
+      "epoch": 0.5718470551958064,
+      "grad_norm": 3.626404285430908,
+      "learning_rate": 4.085078634382724e-06,
+      "loss": 0.924,
+      "step": 7418
+    },
+    {
+      "epoch": 0.5719241443108233,
+      "grad_norm": 3.646789312362671,
+      "learning_rate": 4.083851294650776e-06,
+      "loss": 0.9783,
+      "step": 7419
+    },
+    {
+      "epoch": 0.5720012334258403,
+      "grad_norm": 3.3674442768096924,
+      "learning_rate": 4.0826240120359565e-06,
+      "loss": 0.8714,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5720783225408572,
+      "grad_norm": 3.333601474761963,
+      "learning_rate": 4.081396786614777e-06,
+      "loss": 0.9889,
+      "step": 7421
+    },
+    {
+      "epoch": 0.5721554116558741,
+      "grad_norm": 3.5855748653411865,
+      "learning_rate": 4.080169618463751e-06,
+      "loss": 0.8852,
+      "step": 7422
+    },
+    {
+      "epoch": 0.5722325007708912,
+      "grad_norm": 3.490710496902466,
+      "learning_rate": 4.078942507659386e-06,
+      "loss": 0.9468,
+      "step": 7423
+    },
+    {
+      "epoch": 0.5723095898859081,
+      "grad_norm": 3.4587180614471436,
+      "learning_rate": 4.0777154542781846e-06,
+      "loss": 0.9592,
+      "step": 7424
+    },
+    {
+      "epoch": 0.5723866790009251,
+      "grad_norm": 3.637395143508911,
+      "learning_rate": 4.076488458396649e-06,
+      "loss": 0.9181,
+      "step": 7425
+    },
+    {
+      "epoch": 0.572463768115942,
+      "grad_norm": 3.4306821823120117,
+      "learning_rate": 4.075261520091273e-06,
+      "loss": 0.9677,
+      "step": 7426
+    },
+    {
+      "epoch": 0.572540857230959,
+      "grad_norm": 3.6101748943328857,
+      "learning_rate": 4.074034639438553e-06,
+      "loss": 0.9951,
+      "step": 7427
+    },
+    {
+      "epoch": 0.572617946345976,
+      "grad_norm": 3.5071911811828613,
+      "learning_rate": 4.072807816514978e-06,
+      "loss": 0.878,
+      "step": 7428
+    },
+    {
+      "epoch": 0.5726950354609929,
+      "grad_norm": 3.8580219745635986,
+      "learning_rate": 4.071581051397033e-06,
+      "loss": 0.8271,
+      "step": 7429
+    },
+    {
+      "epoch": 0.5727721245760099,
+      "grad_norm": 3.526630401611328,
+      "learning_rate": 4.070354344161201e-06,
+      "loss": 0.9959,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5728492136910268,
+      "grad_norm": 3.5120689868927,
+      "learning_rate": 4.069127694883962e-06,
+      "loss": 0.8585,
+      "step": 7431
+    },
+    {
+      "epoch": 0.5729263028060438,
+      "grad_norm": 3.920745611190796,
+      "learning_rate": 4.067901103641788e-06,
+      "loss": 0.9903,
+      "step": 7432
+    },
+    {
+      "epoch": 0.5730033919210608,
+      "grad_norm": 3.5974879264831543,
+      "learning_rate": 4.066674570511156e-06,
+      "loss": 0.9851,
+      "step": 7433
+    },
+    {
+      "epoch": 0.5730804810360777,
+      "grad_norm": 3.3706092834472656,
+      "learning_rate": 4.065448095568527e-06,
+      "loss": 0.9094,
+      "step": 7434
+    },
+    {
+      "epoch": 0.5731575701510947,
+      "grad_norm": 3.542107582092285,
+      "learning_rate": 4.064221678890371e-06,
+      "loss": 0.9867,
+      "step": 7435
+    },
+    {
+      "epoch": 0.5732346592661116,
+      "grad_norm": 3.5891056060791016,
+      "learning_rate": 4.062995320553147e-06,
+      "loss": 0.8284,
+      "step": 7436
+    },
+    {
+      "epoch": 0.5733117483811286,
+      "grad_norm": 3.375009059906006,
+      "learning_rate": 4.061769020633311e-06,
+      "loss": 0.86,
+      "step": 7437
+    },
+    {
+      "epoch": 0.5733888374961456,
+      "grad_norm": 3.44054913520813,
+      "learning_rate": 4.060542779207317e-06,
+      "loss": 0.9128,
+      "step": 7438
+    },
+    {
+      "epoch": 0.5734659266111625,
+      "grad_norm": 3.8439524173736572,
+      "learning_rate": 4.059316596351617e-06,
+      "loss": 1.0774,
+      "step": 7439
+    },
+    {
+      "epoch": 0.5735430157261795,
+      "grad_norm": 4.025179862976074,
+      "learning_rate": 4.058090472142654e-06,
+      "loss": 1.0527,
+      "step": 7440
+    },
+    {
+      "epoch": 0.5736201048411964,
+      "grad_norm": 3.771923780441284,
+      "learning_rate": 4.056864406656872e-06,
+      "loss": 0.9199,
+      "step": 7441
+    },
+    {
+      "epoch": 0.5736971939562134,
+      "grad_norm": 3.6226768493652344,
+      "learning_rate": 4.05563839997071e-06,
+      "loss": 0.8523,
+      "step": 7442
+    },
+    {
+      "epoch": 0.5737742830712304,
+      "grad_norm": 3.7793591022491455,
+      "learning_rate": 4.054412452160601e-06,
+      "loss": 0.9109,
+      "step": 7443
+    },
+    {
+      "epoch": 0.5738513721862473,
+      "grad_norm": 3.322064161300659,
+      "learning_rate": 4.053186563302981e-06,
+      "loss": 0.8126,
+      "step": 7444
+    },
+    {
+      "epoch": 0.5739284613012643,
+      "grad_norm": 3.742952585220337,
+      "learning_rate": 4.051960733474273e-06,
+      "loss": 1.0401,
+      "step": 7445
+    },
+    {
+      "epoch": 0.5740055504162812,
+      "grad_norm": 3.36472487449646,
+      "learning_rate": 4.0507349627509045e-06,
+      "loss": 0.7989,
+      "step": 7446
+    },
+    {
+      "epoch": 0.5740826395312982,
+      "grad_norm": 3.582036018371582,
+      "learning_rate": 4.049509251209295e-06,
+      "loss": 0.9565,
+      "step": 7447
+    },
+    {
+      "epoch": 0.5741597286463151,
+      "grad_norm": 3.877749443054199,
+      "learning_rate": 4.048283598925859e-06,
+      "loss": 0.8715,
+      "step": 7448
+    },
+    {
+      "epoch": 0.5742368177613321,
+      "grad_norm": 3.445472002029419,
+      "learning_rate": 4.0470580059770125e-06,
+      "loss": 0.9293,
+      "step": 7449
+    },
+    {
+      "epoch": 0.5743139068763491,
+      "grad_norm": 3.6132113933563232,
+      "learning_rate": 4.045832472439165e-06,
+      "loss": 1.0043,
+      "step": 7450
+    },
+    {
+      "epoch": 0.574390995991366,
+      "grad_norm": 3.4551825523376465,
+      "learning_rate": 4.04460699838872e-06,
+      "loss": 0.8898,
+      "step": 7451
+    },
+    {
+      "epoch": 0.574468085106383,
+      "grad_norm": 3.322658061981201,
+      "learning_rate": 4.0433815839020815e-06,
+      "loss": 0.8621,
+      "step": 7452
+    },
+    {
+      "epoch": 0.5745451742214,
+      "grad_norm": 3.8116767406463623,
+      "learning_rate": 4.042156229055645e-06,
+      "loss": 0.9973,
+      "step": 7453
+    },
+    {
+      "epoch": 0.5746222633364169,
+      "grad_norm": 3.5855953693389893,
+      "learning_rate": 4.040930933925808e-06,
+      "loss": 1.0915,
+      "step": 7454
+    },
+    {
+      "epoch": 0.5746993524514339,
+      "grad_norm": 3.5874369144439697,
+      "learning_rate": 4.039705698588961e-06,
+      "loss": 1.0041,
+      "step": 7455
+    },
+    {
+      "epoch": 0.5747764415664508,
+      "grad_norm": 4.597201824188232,
+      "learning_rate": 4.038480523121488e-06,
+      "loss": 0.851,
+      "step": 7456
+    },
+    {
+      "epoch": 0.5748535306814678,
+      "grad_norm": 3.7259809970855713,
+      "learning_rate": 4.037255407599775e-06,
+      "loss": 0.9158,
+      "step": 7457
+    },
+    {
+      "epoch": 0.5749306197964847,
+      "grad_norm": 3.8748040199279785,
+      "learning_rate": 4.0360303521002014e-06,
+      "loss": 0.9733,
+      "step": 7458
+    },
+    {
+      "epoch": 0.5750077089115017,
+      "grad_norm": 3.888749599456787,
+      "learning_rate": 4.034805356699144e-06,
+      "loss": 1.0054,
+      "step": 7459
+    },
+    {
+      "epoch": 0.5750847980265187,
+      "grad_norm": 3.4167816638946533,
+      "learning_rate": 4.033580421472973e-06,
+      "loss": 0.8625,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5751618871415356,
+      "grad_norm": 4.041205406188965,
+      "learning_rate": 4.032355546498057e-06,
+      "loss": 0.9047,
+      "step": 7461
+    },
+    {
+      "epoch": 0.5752389762565526,
+      "grad_norm": 4.094997882843018,
+      "learning_rate": 4.031130731850762e-06,
+      "loss": 0.9403,
+      "step": 7462
+    },
+    {
+      "epoch": 0.5753160653715695,
+      "grad_norm": 3.3104913234710693,
+      "learning_rate": 4.029905977607448e-06,
+      "loss": 0.8705,
+      "step": 7463
+    },
+    {
+      "epoch": 0.5753931544865865,
+      "grad_norm": 3.7642409801483154,
+      "learning_rate": 4.028681283844471e-06,
+      "loss": 0.961,
+      "step": 7464
+    },
+    {
+      "epoch": 0.5754702436016035,
+      "grad_norm": 3.346750497817993,
+      "learning_rate": 4.027456650638187e-06,
+      "loss": 0.8146,
+      "step": 7465
+    },
+    {
+      "epoch": 0.5755473327166204,
+      "grad_norm": 3.714411735534668,
+      "learning_rate": 4.026232078064942e-06,
+      "loss": 1.0303,
+      "step": 7466
+    },
+    {
+      "epoch": 0.5756244218316374,
+      "grad_norm": 3.7148983478546143,
+      "learning_rate": 4.025007566201085e-06,
+      "loss": 0.8392,
+      "step": 7467
+    },
+    {
+      "epoch": 0.5757015109466543,
+      "grad_norm": 3.490588426589966,
+      "learning_rate": 4.023783115122956e-06,
+      "loss": 0.9568,
+      "step": 7468
+    },
+    {
+      "epoch": 0.5757786000616713,
+      "grad_norm": 3.651613235473633,
+      "learning_rate": 4.0225587249068945e-06,
+      "loss": 0.972,
+      "step": 7469
+    },
+    {
+      "epoch": 0.5758556891766883,
+      "grad_norm": 3.942833185195923,
+      "learning_rate": 4.021334395629234e-06,
+      "loss": 1.0118,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5759327782917052,
+      "grad_norm": 3.5396273136138916,
+      "learning_rate": 4.020110127366304e-06,
+      "loss": 0.8425,
+      "step": 7471
+    },
+    {
+      "epoch": 0.5760098674067222,
+      "grad_norm": 3.949531078338623,
+      "learning_rate": 4.018885920194434e-06,
+      "loss": 0.9779,
+      "step": 7472
+    },
+    {
+      "epoch": 0.5760869565217391,
+      "grad_norm": 4.047253131866455,
+      "learning_rate": 4.0176617741899456e-06,
+      "loss": 1.018,
+      "step": 7473
+    },
+    {
+      "epoch": 0.576164045636756,
+      "grad_norm": 4.037176609039307,
+      "learning_rate": 4.016437689429158e-06,
+      "loss": 0.949,
+      "step": 7474
+    },
+    {
+      "epoch": 0.5762411347517731,
+      "grad_norm": 3.5076513290405273,
+      "learning_rate": 4.015213665988385e-06,
+      "loss": 0.9106,
+      "step": 7475
+    },
+    {
+      "epoch": 0.57631822386679,
+      "grad_norm": 3.6117470264434814,
+      "learning_rate": 4.013989703943942e-06,
+      "loss": 0.9441,
+      "step": 7476
+    },
+    {
+      "epoch": 0.576395312981807,
+      "grad_norm": 3.7086315155029297,
+      "learning_rate": 4.012765803372132e-06,
+      "loss": 0.9981,
+      "step": 7477
+    },
+    {
+      "epoch": 0.5764724020968239,
+      "grad_norm": 3.609908103942871,
+      "learning_rate": 4.011541964349261e-06,
+      "loss": 0.9198,
+      "step": 7478
+    },
+    {
+      "epoch": 0.5765494912118408,
+      "grad_norm": 3.6467490196228027,
+      "learning_rate": 4.010318186951631e-06,
+      "loss": 0.9531,
+      "step": 7479
+    },
+    {
+      "epoch": 0.5766265803268579,
+      "grad_norm": 3.7244629859924316,
+      "learning_rate": 4.009094471255536e-06,
+      "loss": 0.9511,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5767036694418748,
+      "grad_norm": 4.267176628112793,
+      "learning_rate": 4.007870817337268e-06,
+      "loss": 0.9585,
+      "step": 7481
+    },
+    {
+      "epoch": 0.5767807585568918,
+      "grad_norm": 4.043630599975586,
+      "learning_rate": 4.006647225273116e-06,
+      "loss": 0.9615,
+      "step": 7482
+    },
+    {
+      "epoch": 0.5768578476719087,
+      "grad_norm": 3.869731903076172,
+      "learning_rate": 4.005423695139366e-06,
+      "loss": 0.9452,
+      "step": 7483
+    },
+    {
+      "epoch": 0.5769349367869256,
+      "grad_norm": 3.4697086811065674,
+      "learning_rate": 4.004200227012296e-06,
+      "loss": 0.7982,
+      "step": 7484
+    },
+    {
+      "epoch": 0.5770120259019427,
+      "grad_norm": 3.798818826675415,
+      "learning_rate": 4.002976820968186e-06,
+      "loss": 0.9925,
+      "step": 7485
+    },
+    {
+      "epoch": 0.5770891150169596,
+      "grad_norm": 3.5721566677093506,
+      "learning_rate": 4.001753477083306e-06,
+      "loss": 0.8914,
+      "step": 7486
+    },
+    {
+      "epoch": 0.5771662041319766,
+      "grad_norm": 3.3828439712524414,
+      "learning_rate": 4.000530195433928e-06,
+      "loss": 0.8027,
+      "step": 7487
+    },
+    {
+      "epoch": 0.5772432932469935,
+      "grad_norm": 3.4918408393859863,
+      "learning_rate": 3.999306976096315e-06,
+      "loss": 0.8907,
+      "step": 7488
+    },
+    {
+      "epoch": 0.5773203823620104,
+      "grad_norm": 3.524773359298706,
+      "learning_rate": 3.9980838191467296e-06,
+      "loss": 0.9068,
+      "step": 7489
+    },
+    {
+      "epoch": 0.5773974714770275,
+      "grad_norm": 3.529926300048828,
+      "learning_rate": 3.996860724661429e-06,
+      "loss": 0.9297,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5774745605920444,
+      "grad_norm": 3.6954150199890137,
+      "learning_rate": 3.995637692716666e-06,
+      "loss": 0.9739,
+      "step": 7491
+    },
+    {
+      "epoch": 0.5775516497070614,
+      "grad_norm": 3.9414443969726562,
+      "learning_rate": 3.994414723388693e-06,
+      "loss": 1.0595,
+      "step": 7492
+    },
+    {
+      "epoch": 0.5776287388220783,
+      "grad_norm": 3.7672884464263916,
+      "learning_rate": 3.993191816753753e-06,
+      "loss": 0.9773,
+      "step": 7493
+    },
+    {
+      "epoch": 0.5777058279370952,
+      "grad_norm": 3.520125389099121,
+      "learning_rate": 3.991968972888088e-06,
+      "loss": 0.8697,
+      "step": 7494
+    },
+    {
+      "epoch": 0.5777829170521123,
+      "grad_norm": 3.5860490798950195,
+      "learning_rate": 3.990746191867938e-06,
+      "loss": 0.9923,
+      "step": 7495
+    },
+    {
+      "epoch": 0.5778600061671292,
+      "grad_norm": 3.428400993347168,
+      "learning_rate": 3.989523473769536e-06,
+      "loss": 0.7466,
+      "step": 7496
+    },
+    {
+      "epoch": 0.5779370952821462,
+      "grad_norm": 3.6904683113098145,
+      "learning_rate": 3.98830081866911e-06,
+      "loss": 1.0291,
+      "step": 7497
+    },
+    {
+      "epoch": 0.5780141843971631,
+      "grad_norm": 3.687467336654663,
+      "learning_rate": 3.987078226642891e-06,
+      "loss": 0.8533,
+      "step": 7498
+    },
+    {
+      "epoch": 0.57809127351218,
+      "grad_norm": 3.687129020690918,
+      "learning_rate": 3.985855697767097e-06,
+      "loss": 0.9885,
+      "step": 7499
+    },
+    {
+      "epoch": 0.5781683626271971,
+      "grad_norm": 3.5919816493988037,
+      "learning_rate": 3.984633232117948e-06,
+      "loss": 1.0096,
+      "step": 7500
+    },
+    {
+      "epoch": 0.578245451742214,
+      "grad_norm": 3.3104560375213623,
+      "learning_rate": 3.983410829771656e-06,
+      "loss": 0.9715,
+      "step": 7501
+    },
+    {
+      "epoch": 0.578322540857231,
+      "grad_norm": 3.539083957672119,
+      "learning_rate": 3.982188490804437e-06,
+      "loss": 0.8584,
+      "step": 7502
+    },
+    {
+      "epoch": 0.5783996299722479,
+      "grad_norm": 3.4675793647766113,
+      "learning_rate": 3.980966215292493e-06,
+      "loss": 0.903,
+      "step": 7503
+    },
+    {
+      "epoch": 0.5784767190872648,
+      "grad_norm": 3.8223822116851807,
+      "learning_rate": 3.979744003312027e-06,
+      "loss": 0.9112,
+      "step": 7504
+    },
+    {
+      "epoch": 0.5785538082022819,
+      "grad_norm": 3.458359956741333,
+      "learning_rate": 3.97852185493924e-06,
+      "loss": 1.0016,
+      "step": 7505
+    },
+    {
+      "epoch": 0.5786308973172988,
+      "grad_norm": 3.756317138671875,
+      "learning_rate": 3.977299770250324e-06,
+      "loss": 0.9623,
+      "step": 7506
+    },
+    {
+      "epoch": 0.5787079864323158,
+      "grad_norm": 3.434952735900879,
+      "learning_rate": 3.976077749321472e-06,
+      "loss": 0.8595,
+      "step": 7507
+    },
+    {
+      "epoch": 0.5787850755473327,
+      "grad_norm": 3.5256009101867676,
+      "learning_rate": 3.974855792228868e-06,
+      "loss": 0.9876,
+      "step": 7508
+    },
+    {
+      "epoch": 0.5788621646623496,
+      "grad_norm": 3.5869874954223633,
+      "learning_rate": 3.973633899048696e-06,
+      "loss": 0.9544,
+      "step": 7509
+    },
+    {
+      "epoch": 0.5789392537773667,
+      "grad_norm": 3.856430768966675,
+      "learning_rate": 3.9724120698571354e-06,
+      "loss": 1.058,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5790163428923836,
+      "grad_norm": 3.683361053466797,
+      "learning_rate": 3.971190304730359e-06,
+      "loss": 0.9864,
+      "step": 7511
+    },
+    {
+      "epoch": 0.5790934320074006,
+      "grad_norm": 3.582198143005371,
+      "learning_rate": 3.969968603744539e-06,
+      "loss": 0.8897,
+      "step": 7512
+    },
+    {
+      "epoch": 0.5791705211224175,
+      "grad_norm": 3.6801340579986572,
+      "learning_rate": 3.968746966975844e-06,
+      "loss": 0.9485,
+      "step": 7513
+    },
+    {
+      "epoch": 0.5792476102374344,
+      "grad_norm": 3.6165666580200195,
+      "learning_rate": 3.967525394500432e-06,
+      "loss": 0.9006,
+      "step": 7514
+    },
+    {
+      "epoch": 0.5793246993524515,
+      "grad_norm": 3.775045394897461,
+      "learning_rate": 3.966303886394465e-06,
+      "loss": 0.9171,
+      "step": 7515
+    },
+    {
+      "epoch": 0.5794017884674684,
+      "grad_norm": 3.5905048847198486,
+      "learning_rate": 3.965082442734098e-06,
+      "loss": 0.8586,
+      "step": 7516
+    },
+    {
+      "epoch": 0.5794788775824854,
+      "grad_norm": 3.667304277420044,
+      "learning_rate": 3.96386106359548e-06,
+      "loss": 0.9203,
+      "step": 7517
+    },
+    {
+      "epoch": 0.5795559666975023,
+      "grad_norm": 3.824692726135254,
+      "learning_rate": 3.96263974905476e-06,
+      "loss": 0.8806,
+      "step": 7518
+    },
+    {
+      "epoch": 0.5796330558125192,
+      "grad_norm": 3.4772675037384033,
+      "learning_rate": 3.961418499188076e-06,
+      "loss": 0.8067,
+      "step": 7519
+    },
+    {
+      "epoch": 0.5797101449275363,
+      "grad_norm": 3.65657377243042,
+      "learning_rate": 3.960197314071571e-06,
+      "loss": 1.0266,
+      "step": 7520
+    },
+    {
+      "epoch": 0.5797872340425532,
+      "grad_norm": 3.3273978233337402,
+      "learning_rate": 3.9589761937813795e-06,
+      "loss": 0.82,
+      "step": 7521
+    },
+    {
+      "epoch": 0.5798643231575702,
+      "grad_norm": 3.5966200828552246,
+      "learning_rate": 3.957755138393629e-06,
+      "loss": 0.8728,
+      "step": 7522
+    },
+    {
+      "epoch": 0.5799414122725871,
+      "grad_norm": 3.8184077739715576,
+      "learning_rate": 3.9565341479844476e-06,
+      "loss": 1.0214,
+      "step": 7523
+    },
+    {
+      "epoch": 0.580018501387604,
+      "grad_norm": 3.5819761753082275,
+      "learning_rate": 3.9553132226299595e-06,
+      "loss": 0.8721,
+      "step": 7524
+    },
+    {
+      "epoch": 0.580095590502621,
+      "grad_norm": 3.768620491027832,
+      "learning_rate": 3.95409236240628e-06,
+      "loss": 0.905,
+      "step": 7525
+    },
+    {
+      "epoch": 0.580172679617638,
+      "grad_norm": 3.6173248291015625,
+      "learning_rate": 3.952871567389525e-06,
+      "loss": 0.8692,
+      "step": 7526
+    },
+    {
+      "epoch": 0.580249768732655,
+      "grad_norm": 3.3437893390655518,
+      "learning_rate": 3.951650837655805e-06,
+      "loss": 0.982,
+      "step": 7527
+    },
+    {
+      "epoch": 0.5803268578476719,
+      "grad_norm": 3.5120673179626465,
+      "learning_rate": 3.9504301732812255e-06,
+      "loss": 0.9568,
+      "step": 7528
+    },
+    {
+      "epoch": 0.5804039469626888,
+      "grad_norm": 3.586928606033325,
+      "learning_rate": 3.949209574341889e-06,
+      "loss": 0.9897,
+      "step": 7529
+    },
+    {
+      "epoch": 0.5804810360777058,
+      "grad_norm": 3.88667631149292,
+      "learning_rate": 3.947989040913893e-06,
+      "loss": 1.0405,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5805581251927228,
+      "grad_norm": 3.989518165588379,
+      "learning_rate": 3.946768573073332e-06,
+      "loss": 0.9536,
+      "step": 7531
+    },
+    {
+      "epoch": 0.5806352143077398,
+      "grad_norm": 3.595759630203247,
+      "learning_rate": 3.945548170896296e-06,
+      "loss": 0.9909,
+      "step": 7532
+    },
+    {
+      "epoch": 0.5807123034227567,
+      "grad_norm": 3.8122918605804443,
+      "learning_rate": 3.9443278344588696e-06,
+      "loss": 0.9124,
+      "step": 7533
+    },
+    {
+      "epoch": 0.5807893925377736,
+      "grad_norm": 3.4445278644561768,
+      "learning_rate": 3.943107563837135e-06,
+      "loss": 0.862,
+      "step": 7534
+    },
+    {
+      "epoch": 0.5808664816527906,
+      "grad_norm": 3.818312168121338,
+      "learning_rate": 3.941887359107172e-06,
+      "loss": 0.9938,
+      "step": 7535
+    },
+    {
+      "epoch": 0.5809435707678076,
+      "grad_norm": 3.6092066764831543,
+      "learning_rate": 3.94066722034505e-06,
+      "loss": 0.9801,
+      "step": 7536
+    },
+    {
+      "epoch": 0.5810206598828246,
+      "grad_norm": 3.58729887008667,
+      "learning_rate": 3.939447147626842e-06,
+      "loss": 0.8662,
+      "step": 7537
+    },
+    {
+      "epoch": 0.5810977489978415,
+      "grad_norm": 4.0023298263549805,
+      "learning_rate": 3.93822714102861e-06,
+      "loss": 0.929,
+      "step": 7538
+    },
+    {
+      "epoch": 0.5811748381128584,
+      "grad_norm": 3.671865940093994,
+      "learning_rate": 3.937007200626417e-06,
+      "loss": 1.0063,
+      "step": 7539
+    },
+    {
+      "epoch": 0.5812519272278754,
+      "grad_norm": 3.9772789478302,
+      "learning_rate": 3.935787326496321e-06,
+      "loss": 1.1299,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5813290163428924,
+      "grad_norm": 3.3404088020324707,
+      "learning_rate": 3.934567518714372e-06,
+      "loss": 0.9902,
+      "step": 7541
+    },
+    {
+      "epoch": 0.5814061054579094,
+      "grad_norm": 3.876286506652832,
+      "learning_rate": 3.9333477773566204e-06,
+      "loss": 0.9492,
+      "step": 7542
+    },
+    {
+      "epoch": 0.5814831945729263,
+      "grad_norm": 3.982120990753174,
+      "learning_rate": 3.932128102499111e-06,
+      "loss": 1.0037,
+      "step": 7543
+    },
+    {
+      "epoch": 0.5815602836879432,
+      "grad_norm": 4.225841999053955,
+      "learning_rate": 3.930908494217884e-06,
+      "loss": 0.8912,
+      "step": 7544
+    },
+    {
+      "epoch": 0.5816373728029602,
+      "grad_norm": 3.5947415828704834,
+      "learning_rate": 3.929688952588974e-06,
+      "loss": 1.0022,
+      "step": 7545
+    },
+    {
+      "epoch": 0.5817144619179772,
+      "grad_norm": 3.5156421661376953,
+      "learning_rate": 3.928469477688415e-06,
+      "loss": 0.8517,
+      "step": 7546
+    },
+    {
+      "epoch": 0.5817915510329942,
+      "grad_norm": 3.9697585105895996,
+      "learning_rate": 3.927250069592236e-06,
+      "loss": 0.868,
+      "step": 7547
+    },
+    {
+      "epoch": 0.5818686401480111,
+      "grad_norm": 3.4268386363983154,
+      "learning_rate": 3.926030728376458e-06,
+      "loss": 0.8795,
+      "step": 7548
+    },
+    {
+      "epoch": 0.581945729263028,
+      "grad_norm": 3.7398383617401123,
+      "learning_rate": 3.924811454117101e-06,
+      "loss": 0.963,
+      "step": 7549
+    },
+    {
+      "epoch": 0.582022818378045,
+      "grad_norm": 3.5979716777801514,
+      "learning_rate": 3.923592246890183e-06,
+      "loss": 0.9222,
+      "step": 7550
+    },
+    {
+      "epoch": 0.582099907493062,
+      "grad_norm": 3.304446220397949,
+      "learning_rate": 3.922373106771713e-06,
+      "loss": 0.895,
+      "step": 7551
+    },
+    {
+      "epoch": 0.582176996608079,
+      "grad_norm": 3.7752010822296143,
+      "learning_rate": 3.921154033837699e-06,
+      "loss": 1.0166,
+      "step": 7552
+    },
+    {
+      "epoch": 0.5822540857230959,
+      "grad_norm": 3.516181707382202,
+      "learning_rate": 3.919935028164143e-06,
+      "loss": 0.9529,
+      "step": 7553
+    },
+    {
+      "epoch": 0.5823311748381128,
+      "grad_norm": 3.4951820373535156,
+      "learning_rate": 3.9187160898270435e-06,
+      "loss": 0.9276,
+      "step": 7554
+    },
+    {
+      "epoch": 0.5824082639531298,
+      "grad_norm": 3.369300365447998,
+      "learning_rate": 3.917497218902398e-06,
+      "loss": 0.9608,
+      "step": 7555
+    },
+    {
+      "epoch": 0.5824853530681467,
+      "grad_norm": 3.4424049854278564,
+      "learning_rate": 3.916278415466193e-06,
+      "loss": 0.9437,
+      "step": 7556
+    },
+    {
+      "epoch": 0.5825624421831638,
+      "grad_norm": 3.40697979927063,
+      "learning_rate": 3.9150596795944155e-06,
+      "loss": 0.9566,
+      "step": 7557
+    },
+    {
+      "epoch": 0.5826395312981807,
+      "grad_norm": 3.608034133911133,
+      "learning_rate": 3.913841011363049e-06,
+      "loss": 0.9437,
+      "step": 7558
+    },
+    {
+      "epoch": 0.5827166204131976,
+      "grad_norm": 3.5082850456237793,
+      "learning_rate": 3.912622410848069e-06,
+      "loss": 0.8977,
+      "step": 7559
+    },
+    {
+      "epoch": 0.5827937095282146,
+      "grad_norm": 3.802832841873169,
+      "learning_rate": 3.91140387812545e-06,
+      "loss": 0.9256,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5828707986432315,
+      "grad_norm": 3.8759782314300537,
+      "learning_rate": 3.910185413271162e-06,
+      "loss": 0.8519,
+      "step": 7561
+    },
+    {
+      "epoch": 0.5829478877582486,
+      "grad_norm": 3.485323429107666,
+      "learning_rate": 3.908967016361169e-06,
+      "loss": 0.9075,
+      "step": 7562
+    },
+    {
+      "epoch": 0.5830249768732655,
+      "grad_norm": 3.791802167892456,
+      "learning_rate": 3.907748687471431e-06,
+      "loss": 0.9889,
+      "step": 7563
+    },
+    {
+      "epoch": 0.5831020659882824,
+      "grad_norm": 3.474576950073242,
+      "learning_rate": 3.906530426677907e-06,
+      "loss": 0.8918,
+      "step": 7564
+    },
+    {
+      "epoch": 0.5831791551032994,
+      "grad_norm": 3.3780479431152344,
+      "learning_rate": 3.905312234056547e-06,
+      "loss": 0.8903,
+      "step": 7565
+    },
+    {
+      "epoch": 0.5832562442183163,
+      "grad_norm": 3.6616179943084717,
+      "learning_rate": 3.904094109683301e-06,
+      "loss": 0.9756,
+      "step": 7566
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 3.564548969268799,
+      "learning_rate": 3.902876053634109e-06,
+      "loss": 0.8836,
+      "step": 7567
+    },
+    {
+      "epoch": 0.5834104224483503,
+      "grad_norm": 3.944483518600464,
+      "learning_rate": 3.901658065984913e-06,
+      "loss": 0.9277,
+      "step": 7568
+    },
+    {
+      "epoch": 0.5834875115633672,
+      "grad_norm": 3.700523614883423,
+      "learning_rate": 3.90044014681165e-06,
+      "loss": 1.0303,
+      "step": 7569
+    },
+    {
+      "epoch": 0.5835646006783842,
+      "grad_norm": 3.6363394260406494,
+      "learning_rate": 3.899222296190248e-06,
+      "loss": 0.8375,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5836416897934011,
+      "grad_norm": 3.8112072944641113,
+      "learning_rate": 3.898004514196635e-06,
+      "loss": 0.9404,
+      "step": 7571
+    },
+    {
+      "epoch": 0.5837187789084182,
+      "grad_norm": 3.8142549991607666,
+      "learning_rate": 3.896786800906734e-06,
+      "loss": 0.908,
+      "step": 7572
+    },
+    {
+      "epoch": 0.5837958680234351,
+      "grad_norm": 3.6068294048309326,
+      "learning_rate": 3.8955691563964605e-06,
+      "loss": 0.8951,
+      "step": 7573
+    },
+    {
+      "epoch": 0.583872957138452,
+      "grad_norm": 3.8355910778045654,
+      "learning_rate": 3.894351580741731e-06,
+      "loss": 0.8517,
+      "step": 7574
+    },
+    {
+      "epoch": 0.583950046253469,
+      "grad_norm": 3.838613986968994,
+      "learning_rate": 3.893134074018454e-06,
+      "loss": 1.0072,
+      "step": 7575
+    },
+    {
+      "epoch": 0.5840271353684859,
+      "grad_norm": 3.381627321243286,
+      "learning_rate": 3.891916636302535e-06,
+      "loss": 0.9339,
+      "step": 7576
+    },
+    {
+      "epoch": 0.584104224483503,
+      "grad_norm": 3.5409560203552246,
+      "learning_rate": 3.890699267669876e-06,
+      "loss": 0.8285,
+      "step": 7577
+    },
+    {
+      "epoch": 0.5841813135985199,
+      "grad_norm": 3.678004741668701,
+      "learning_rate": 3.889481968196371e-06,
+      "loss": 0.9596,
+      "step": 7578
+    },
+    {
+      "epoch": 0.5842584027135368,
+      "grad_norm": 4.06223726272583,
+      "learning_rate": 3.888264737957913e-06,
+      "loss": 1.0516,
+      "step": 7579
+    },
+    {
+      "epoch": 0.5843354918285538,
+      "grad_norm": 3.5083158016204834,
+      "learning_rate": 3.887047577030393e-06,
+      "loss": 0.8042,
+      "step": 7580
+    },
+    {
+      "epoch": 0.5844125809435707,
+      "grad_norm": 3.717420816421509,
+      "learning_rate": 3.8858304854896906e-06,
+      "loss": 1.0194,
+      "step": 7581
+    },
+    {
+      "epoch": 0.5844896700585878,
+      "grad_norm": 3.8681986331939697,
+      "learning_rate": 3.884613463411687e-06,
+      "loss": 0.9029,
+      "step": 7582
+    },
+    {
+      "epoch": 0.5845667591736047,
+      "grad_norm": 3.5388240814208984,
+      "learning_rate": 3.883396510872259e-06,
+      "loss": 0.9348,
+      "step": 7583
+    },
+    {
+      "epoch": 0.5846438482886216,
+      "grad_norm": 3.749166488647461,
+      "learning_rate": 3.882179627947273e-06,
+      "loss": 0.9825,
+      "step": 7584
+    },
+    {
+      "epoch": 0.5847209374036386,
+      "grad_norm": 3.7191250324249268,
+      "learning_rate": 3.8809628147126e-06,
+      "loss": 0.9845,
+      "step": 7585
+    },
+    {
+      "epoch": 0.5847980265186555,
+      "grad_norm": 3.772263526916504,
+      "learning_rate": 3.879746071244099e-06,
+      "loss": 0.8341,
+      "step": 7586
+    },
+    {
+      "epoch": 0.5848751156336726,
+      "grad_norm": 3.8493549823760986,
+      "learning_rate": 3.8785293976176285e-06,
+      "loss": 1.0723,
+      "step": 7587
+    },
+    {
+      "epoch": 0.5849522047486895,
+      "grad_norm": 3.944638967514038,
+      "learning_rate": 3.877312793909043e-06,
+      "loss": 1.0129,
+      "step": 7588
+    },
+    {
+      "epoch": 0.5850292938637064,
+      "grad_norm": 3.4591333866119385,
+      "learning_rate": 3.876096260194188e-06,
+      "loss": 1.0135,
+      "step": 7589
+    },
+    {
+      "epoch": 0.5851063829787234,
+      "grad_norm": 3.324209451675415,
+      "learning_rate": 3.87487979654891e-06,
+      "loss": 0.8952,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5851834720937403,
+      "grad_norm": 3.500072956085205,
+      "learning_rate": 3.873663403049051e-06,
+      "loss": 0.8838,
+      "step": 7591
+    },
+    {
+      "epoch": 0.5852605612087574,
+      "grad_norm": 3.5885183811187744,
+      "learning_rate": 3.872447079770446e-06,
+      "loss": 0.9321,
+      "step": 7592
+    },
+    {
+      "epoch": 0.5853376503237743,
+      "grad_norm": 3.4161670207977295,
+      "learning_rate": 3.871230826788925e-06,
+      "loss": 0.9088,
+      "step": 7593
+    },
+    {
+      "epoch": 0.5854147394387912,
+      "grad_norm": 3.7749154567718506,
+      "learning_rate": 3.870014644180315e-06,
+      "loss": 0.9423,
+      "step": 7594
+    },
+    {
+      "epoch": 0.5854918285538082,
+      "grad_norm": 3.5398154258728027,
+      "learning_rate": 3.868798532020442e-06,
+      "loss": 0.932,
+      "step": 7595
+    },
+    {
+      "epoch": 0.5855689176688251,
+      "grad_norm": 3.844788074493408,
+      "learning_rate": 3.86758249038512e-06,
+      "loss": 0.9811,
+      "step": 7596
+    },
+    {
+      "epoch": 0.5856460067838422,
+      "grad_norm": 3.943168878555298,
+      "learning_rate": 3.866366519350165e-06,
+      "loss": 0.9965,
+      "step": 7597
+    },
+    {
+      "epoch": 0.5857230958988591,
+      "grad_norm": 3.5133891105651855,
+      "learning_rate": 3.865150618991388e-06,
+      "loss": 0.9619,
+      "step": 7598
+    },
+    {
+      "epoch": 0.5858001850138761,
+      "grad_norm": 3.773076295852661,
+      "learning_rate": 3.8639347893845905e-06,
+      "loss": 0.8272,
+      "step": 7599
+    },
+    {
+      "epoch": 0.585877274128893,
+      "grad_norm": 3.4386439323425293,
+      "learning_rate": 3.862719030605575e-06,
+      "loss": 0.921,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5859543632439099,
+      "grad_norm": 3.826698064804077,
+      "learning_rate": 3.861503342730139e-06,
+      "loss": 0.8279,
+      "step": 7601
+    },
+    {
+      "epoch": 0.586031452358927,
+      "grad_norm": 3.6429896354675293,
+      "learning_rate": 3.860287725834072e-06,
+      "loss": 0.8615,
+      "step": 7602
+    },
+    {
+      "epoch": 0.5861085414739439,
+      "grad_norm": 3.78666353225708,
+      "learning_rate": 3.859072179993164e-06,
+      "loss": 0.9863,
+      "step": 7603
+    },
+    {
+      "epoch": 0.5861856305889609,
+      "grad_norm": 3.275872230529785,
+      "learning_rate": 3.857856705283195e-06,
+      "loss": 0.7179,
+      "step": 7604
+    },
+    {
+      "epoch": 0.5862627197039778,
+      "grad_norm": 3.5281994342803955,
+      "learning_rate": 3.856641301779946e-06,
+      "loss": 0.9711,
+      "step": 7605
+    },
+    {
+      "epoch": 0.5863398088189947,
+      "grad_norm": 3.61147141456604,
+      "learning_rate": 3.855425969559191e-06,
+      "loss": 0.9409,
+      "step": 7606
+    },
+    {
+      "epoch": 0.5864168979340117,
+      "grad_norm": 3.7258195877075195,
+      "learning_rate": 3.854210708696697e-06,
+      "loss": 0.8085,
+      "step": 7607
+    },
+    {
+      "epoch": 0.5864939870490287,
+      "grad_norm": 3.8481357097625732,
+      "learning_rate": 3.852995519268231e-06,
+      "loss": 0.9489,
+      "step": 7608
+    },
+    {
+      "epoch": 0.5865710761640457,
+      "grad_norm": 4.056955337524414,
+      "learning_rate": 3.851780401349557e-06,
+      "loss": 1.0348,
+      "step": 7609
+    },
+    {
+      "epoch": 0.5866481652790626,
+      "grad_norm": 3.392282485961914,
+      "learning_rate": 3.8505653550164255e-06,
+      "loss": 0.8725,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5867252543940795,
+      "grad_norm": 3.6491785049438477,
+      "learning_rate": 3.849350380344591e-06,
+      "loss": 0.8837,
+      "step": 7611
+    },
+    {
+      "epoch": 0.5868023435090965,
+      "grad_norm": 3.8450396060943604,
+      "learning_rate": 3.8481354774098025e-06,
+      "loss": 0.9088,
+      "step": 7612
+    },
+    {
+      "epoch": 0.5868794326241135,
+      "grad_norm": 3.7339627742767334,
+      "learning_rate": 3.8469206462878e-06,
+      "loss": 0.9026,
+      "step": 7613
+    },
+    {
+      "epoch": 0.5869565217391305,
+      "grad_norm": 3.736210346221924,
+      "learning_rate": 3.845705887054324e-06,
+      "loss": 0.8872,
+      "step": 7614
+    },
+    {
+      "epoch": 0.5870336108541474,
+      "grad_norm": 3.851383924484253,
+      "learning_rate": 3.844491199785107e-06,
+      "loss": 0.9875,
+      "step": 7615
+    },
+    {
+      "epoch": 0.5871106999691643,
+      "grad_norm": 3.5837175846099854,
+      "learning_rate": 3.843276584555879e-06,
+      "loss": 0.9964,
+      "step": 7616
+    },
+    {
+      "epoch": 0.5871877890841813,
+      "grad_norm": 3.548447847366333,
+      "learning_rate": 3.842062041442366e-06,
+      "loss": 0.9743,
+      "step": 7617
+    },
+    {
+      "epoch": 0.5872648781991983,
+      "grad_norm": 3.630624532699585,
+      "learning_rate": 3.8408475705202876e-06,
+      "loss": 0.9896,
+      "step": 7618
+    },
+    {
+      "epoch": 0.5873419673142153,
+      "grad_norm": 3.6596896648406982,
+      "learning_rate": 3.839633171865359e-06,
+      "loss": 0.8792,
+      "step": 7619
+    },
+    {
+      "epoch": 0.5874190564292322,
+      "grad_norm": 3.846327066421509,
+      "learning_rate": 3.838418845553293e-06,
+      "loss": 0.9892,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5874961455442491,
+      "grad_norm": 3.916472911834717,
+      "learning_rate": 3.837204591659795e-06,
+      "loss": 1.0246,
+      "step": 7621
+    },
+    {
+      "epoch": 0.5875732346592661,
+      "grad_norm": 3.52048397064209,
+      "learning_rate": 3.83599041026057e-06,
+      "loss": 0.7953,
+      "step": 7622
+    },
+    {
+      "epoch": 0.5876503237742831,
+      "grad_norm": 3.525165557861328,
+      "learning_rate": 3.834776301431314e-06,
+      "loss": 0.8311,
+      "step": 7623
+    },
+    {
+      "epoch": 0.5877274128893001,
+      "grad_norm": 3.754812240600586,
+      "learning_rate": 3.83356226524772e-06,
+      "loss": 0.9588,
+      "step": 7624
+    },
+    {
+      "epoch": 0.587804502004317,
+      "grad_norm": 3.5467734336853027,
+      "learning_rate": 3.832348301785479e-06,
+      "loss": 0.9716,
+      "step": 7625
+    },
+    {
+      "epoch": 0.5878815911193339,
+      "grad_norm": 3.7867307662963867,
+      "learning_rate": 3.831134411120273e-06,
+      "loss": 1.0345,
+      "step": 7626
+    },
+    {
+      "epoch": 0.5879586802343509,
+      "grad_norm": 3.4668374061584473,
+      "learning_rate": 3.8299205933277814e-06,
+      "loss": 0.8672,
+      "step": 7627
+    },
+    {
+      "epoch": 0.5880357693493679,
+      "grad_norm": 3.5994906425476074,
+      "learning_rate": 3.8287068484836835e-06,
+      "loss": 0.8797,
+      "step": 7628
+    },
+    {
+      "epoch": 0.5881128584643849,
+      "grad_norm": 3.47133731842041,
+      "learning_rate": 3.827493176663645e-06,
+      "loss": 0.8707,
+      "step": 7629
+    },
+    {
+      "epoch": 0.5881899475794018,
+      "grad_norm": 3.8516485691070557,
+      "learning_rate": 3.826279577943335e-06,
+      "loss": 0.9896,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5882670366944187,
+      "grad_norm": 3.4910974502563477,
+      "learning_rate": 3.825066052398416e-06,
+      "loss": 0.9694,
+      "step": 7631
+    },
+    {
+      "epoch": 0.5883441258094357,
+      "grad_norm": 3.6999757289886475,
+      "learning_rate": 3.823852600104542e-06,
+      "loss": 0.8824,
+      "step": 7632
+    },
+    {
+      "epoch": 0.5884212149244527,
+      "grad_norm": 3.632878541946411,
+      "learning_rate": 3.822639221137368e-06,
+      "loss": 0.9755,
+      "step": 7633
+    },
+    {
+      "epoch": 0.5884983040394697,
+      "grad_norm": 3.42018461227417,
+      "learning_rate": 3.821425915572537e-06,
+      "loss": 0.902,
+      "step": 7634
+    },
+    {
+      "epoch": 0.5885753931544866,
+      "grad_norm": 3.5439507961273193,
+      "learning_rate": 3.8202126834857e-06,
+      "loss": 0.8284,
+      "step": 7635
+    },
+    {
+      "epoch": 0.5886524822695035,
+      "grad_norm": 3.7671468257904053,
+      "learning_rate": 3.818999524952491e-06,
+      "loss": 0.9682,
+      "step": 7636
+    },
+    {
+      "epoch": 0.5887295713845205,
+      "grad_norm": 3.287444829940796,
+      "learning_rate": 3.817786440048545e-06,
+      "loss": 0.9903,
+      "step": 7637
+    },
+    {
+      "epoch": 0.5888066604995374,
+      "grad_norm": 3.516890525817871,
+      "learning_rate": 3.8165734288494925e-06,
+      "loss": 0.9282,
+      "step": 7638
+    },
+    {
+      "epoch": 0.5888837496145545,
+      "grad_norm": 3.4296483993530273,
+      "learning_rate": 3.815360491430956e-06,
+      "loss": 0.8815,
+      "step": 7639
+    },
+    {
+      "epoch": 0.5889608387295714,
+      "grad_norm": 3.8724493980407715,
+      "learning_rate": 3.8141476278685596e-06,
+      "loss": 0.8815,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5890379278445883,
+      "grad_norm": 3.6292412281036377,
+      "learning_rate": 3.812934838237915e-06,
+      "loss": 0.9961,
+      "step": 7641
+    },
+    {
+      "epoch": 0.5891150169596053,
+      "grad_norm": 3.47171688079834,
+      "learning_rate": 3.811722122614636e-06,
+      "loss": 0.9523,
+      "step": 7642
+    },
+    {
+      "epoch": 0.5891921060746222,
+      "grad_norm": 3.357323408126831,
+      "learning_rate": 3.8105094810743286e-06,
+      "loss": 0.959,
+      "step": 7643
+    },
+    {
+      "epoch": 0.5892691951896393,
+      "grad_norm": 3.610004425048828,
+      "learning_rate": 3.809296913692594e-06,
+      "loss": 0.9088,
+      "step": 7644
+    },
+    {
+      "epoch": 0.5893462843046562,
+      "grad_norm": 3.6334939002990723,
+      "learning_rate": 3.80808442054503e-06,
+      "loss": 0.9915,
+      "step": 7645
+    },
+    {
+      "epoch": 0.5894233734196731,
+      "grad_norm": 3.537346601486206,
+      "learning_rate": 3.80687200170723e-06,
+      "loss": 0.8867,
+      "step": 7646
+    },
+    {
+      "epoch": 0.5895004625346901,
+      "grad_norm": 3.5367846488952637,
+      "learning_rate": 3.805659657254781e-06,
+      "loss": 0.9774,
+      "step": 7647
+    },
+    {
+      "epoch": 0.589577551649707,
+      "grad_norm": 3.7989237308502197,
+      "learning_rate": 3.8044473872632663e-06,
+      "loss": 0.959,
+      "step": 7648
+    },
+    {
+      "epoch": 0.5896546407647241,
+      "grad_norm": 3.7498884201049805,
+      "learning_rate": 3.8032351918082665e-06,
+      "loss": 0.9014,
+      "step": 7649
+    },
+    {
+      "epoch": 0.589731729879741,
+      "grad_norm": 3.9480056762695312,
+      "learning_rate": 3.8020230709653527e-06,
+      "loss": 0.9617,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5898088189947579,
+      "grad_norm": 3.7743465900421143,
+      "learning_rate": 3.800811024810097e-06,
+      "loss": 1.0255,
+      "step": 7651
+    },
+    {
+      "epoch": 0.5898859081097749,
+      "grad_norm": 3.489729166030884,
+      "learning_rate": 3.7995990534180627e-06,
+      "loss": 0.929,
+      "step": 7652
+    },
+    {
+      "epoch": 0.5899629972247918,
+      "grad_norm": 4.14895486831665,
+      "learning_rate": 3.7983871568648095e-06,
+      "loss": 1.0033,
+      "step": 7653
+    },
+    {
+      "epoch": 0.5900400863398089,
+      "grad_norm": 3.2822515964508057,
+      "learning_rate": 3.7971753352258955e-06,
+      "loss": 0.8825,
+      "step": 7654
+    },
+    {
+      "epoch": 0.5901171754548258,
+      "grad_norm": 3.469475507736206,
+      "learning_rate": 3.7959635885768686e-06,
+      "loss": 0.8867,
+      "step": 7655
+    },
+    {
+      "epoch": 0.5901942645698427,
+      "grad_norm": 3.7009308338165283,
+      "learning_rate": 3.7947519169932754e-06,
+      "loss": 0.9416,
+      "step": 7656
+    },
+    {
+      "epoch": 0.5902713536848597,
+      "grad_norm": 3.662630319595337,
+      "learning_rate": 3.79354032055066e-06,
+      "loss": 0.9328,
+      "step": 7657
+    },
+    {
+      "epoch": 0.5903484427998766,
+      "grad_norm": 3.5414748191833496,
+      "learning_rate": 3.7923287993245556e-06,
+      "loss": 0.8743,
+      "step": 7658
+    },
+    {
+      "epoch": 0.5904255319148937,
+      "grad_norm": 3.310471773147583,
+      "learning_rate": 3.791117353390496e-06,
+      "loss": 0.7666,
+      "step": 7659
+    },
+    {
+      "epoch": 0.5905026210299106,
+      "grad_norm": 3.43023943901062,
+      "learning_rate": 3.78990598282401e-06,
+      "loss": 0.8936,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5905797101449275,
+      "grad_norm": 3.8442037105560303,
+      "learning_rate": 3.788694687700617e-06,
+      "loss": 1.0593,
+      "step": 7661
+    },
+    {
+      "epoch": 0.5906567992599445,
+      "grad_norm": 3.430696487426758,
+      "learning_rate": 3.787483468095838e-06,
+      "loss": 0.8401,
+      "step": 7662
+    },
+    {
+      "epoch": 0.5907338883749614,
+      "grad_norm": 3.727947950363159,
+      "learning_rate": 3.786272324085184e-06,
+      "loss": 0.9685,
+      "step": 7663
+    },
+    {
+      "epoch": 0.5908109774899785,
+      "grad_norm": 3.676292657852173,
+      "learning_rate": 3.7850612557441648e-06,
+      "loss": 1.0102,
+      "step": 7664
+    },
+    {
+      "epoch": 0.5908880666049954,
+      "grad_norm": 3.6820626258850098,
+      "learning_rate": 3.783850263148284e-06,
+      "loss": 0.9545,
+      "step": 7665
+    },
+    {
+      "epoch": 0.5909651557200123,
+      "grad_norm": 3.582075357437134,
+      "learning_rate": 3.7826393463730403e-06,
+      "loss": 0.8736,
+      "step": 7666
+    },
+    {
+      "epoch": 0.5910422448350293,
+      "grad_norm": 3.6667416095733643,
+      "learning_rate": 3.7814285054939285e-06,
+      "loss": 0.9364,
+      "step": 7667
+    },
+    {
+      "epoch": 0.5911193339500462,
+      "grad_norm": 3.8910152912139893,
+      "learning_rate": 3.7802177405864392e-06,
+      "loss": 1.0272,
+      "step": 7668
+    },
+    {
+      "epoch": 0.5911964230650633,
+      "grad_norm": 3.327561378479004,
+      "learning_rate": 3.779007051726056e-06,
+      "loss": 0.7072,
+      "step": 7669
+    },
+    {
+      "epoch": 0.5912735121800802,
+      "grad_norm": 3.4671108722686768,
+      "learning_rate": 3.77779643898826e-06,
+      "loss": 0.8911,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5913506012950971,
+      "grad_norm": 3.4955384731292725,
+      "learning_rate": 3.7765859024485246e-06,
+      "loss": 0.7711,
+      "step": 7671
+    },
+    {
+      "epoch": 0.5914276904101141,
+      "grad_norm": 3.6591098308563232,
+      "learning_rate": 3.7753754421823225e-06,
+      "loss": 0.9811,
+      "step": 7672
+    },
+    {
+      "epoch": 0.591504779525131,
+      "grad_norm": 3.4825854301452637,
+      "learning_rate": 3.7741650582651195e-06,
+      "loss": 0.8656,
+      "step": 7673
+    },
+    {
+      "epoch": 0.5915818686401481,
+      "grad_norm": 3.4312667846679688,
+      "learning_rate": 3.7729547507723764e-06,
+      "loss": 0.8741,
+      "step": 7674
+    },
+    {
+      "epoch": 0.591658957755165,
+      "grad_norm": 3.5050833225250244,
+      "learning_rate": 3.771744519779548e-06,
+      "loss": 0.8984,
+      "step": 7675
+    },
+    {
+      "epoch": 0.5917360468701819,
+      "grad_norm": 3.687978982925415,
+      "learning_rate": 3.7705343653620894e-06,
+      "loss": 0.8962,
+      "step": 7676
+    },
+    {
+      "epoch": 0.5918131359851989,
+      "grad_norm": 3.417243480682373,
+      "learning_rate": 3.769324287595445e-06,
+      "loss": 0.922,
+      "step": 7677
+    },
+    {
+      "epoch": 0.5918902251002158,
+      "grad_norm": 3.6329598426818848,
+      "learning_rate": 3.7681142865550555e-06,
+      "loss": 0.9152,
+      "step": 7678
+    },
+    {
+      "epoch": 0.5919673142152329,
+      "grad_norm": 3.985581398010254,
+      "learning_rate": 3.766904362316362e-06,
+      "loss": 0.9494,
+      "step": 7679
+    },
+    {
+      "epoch": 0.5920444033302498,
+      "grad_norm": 3.8809006214141846,
+      "learning_rate": 3.765694514954796e-06,
+      "loss": 0.9822,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5921214924452667,
+      "grad_norm": 3.394606828689575,
+      "learning_rate": 3.7644847445457826e-06,
+      "loss": 0.8432,
+      "step": 7681
+    },
+    {
+      "epoch": 0.5921985815602837,
+      "grad_norm": 3.865122079849243,
+      "learning_rate": 3.7632750511647478e-06,
+      "loss": 0.9003,
+      "step": 7682
+    },
+    {
+      "epoch": 0.5922756706753006,
+      "grad_norm": 3.693389892578125,
+      "learning_rate": 3.7620654348871083e-06,
+      "loss": 0.9377,
+      "step": 7683
+    },
+    {
+      "epoch": 0.5923527597903177,
+      "grad_norm": 3.6535701751708984,
+      "learning_rate": 3.760855895788277e-06,
+      "loss": 0.9298,
+      "step": 7684
+    },
+    {
+      "epoch": 0.5924298489053346,
+      "grad_norm": 3.484705686569214,
+      "learning_rate": 3.759646433943662e-06,
+      "loss": 0.9926,
+      "step": 7685
+    },
+    {
+      "epoch": 0.5925069380203515,
+      "grad_norm": 3.9103033542633057,
+      "learning_rate": 3.7584370494286697e-06,
+      "loss": 1.0298,
+      "step": 7686
+    },
+    {
+      "epoch": 0.5925840271353685,
+      "grad_norm": 3.3555185794830322,
+      "learning_rate": 3.7572277423186964e-06,
+      "loss": 0.8897,
+      "step": 7687
+    },
+    {
+      "epoch": 0.5926611162503854,
+      "grad_norm": 3.7804808616638184,
+      "learning_rate": 3.7560185126891375e-06,
+      "loss": 0.9413,
+      "step": 7688
+    },
+    {
+      "epoch": 0.5927382053654024,
+      "grad_norm": 3.7043962478637695,
+      "learning_rate": 3.75480936061538e-06,
+      "loss": 0.8945,
+      "step": 7689
+    },
+    {
+      "epoch": 0.5928152944804194,
+      "grad_norm": 3.398715019226074,
+      "learning_rate": 3.753600286172811e-06,
+      "loss": 0.8635,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5928923835954363,
+      "grad_norm": 3.560133934020996,
+      "learning_rate": 3.7523912894368093e-06,
+      "loss": 0.9516,
+      "step": 7691
+    },
+    {
+      "epoch": 0.5929694727104533,
+      "grad_norm": 3.709573745727539,
+      "learning_rate": 3.751182370482748e-06,
+      "loss": 0.9026,
+      "step": 7692
+    },
+    {
+      "epoch": 0.5930465618254702,
+      "grad_norm": 3.8817639350891113,
+      "learning_rate": 3.7499735293859985e-06,
+      "loss": 0.9197,
+      "step": 7693
+    },
+    {
+      "epoch": 0.5931236509404872,
+      "grad_norm": 3.9092764854431152,
+      "learning_rate": 3.7487647662219263e-06,
+      "loss": 0.8946,
+      "step": 7694
+    },
+    {
+      "epoch": 0.5932007400555042,
+      "grad_norm": 3.4209654331207275,
+      "learning_rate": 3.7475560810658896e-06,
+      "loss": 0.8795,
+      "step": 7695
+    },
+    {
+      "epoch": 0.5932778291705211,
+      "grad_norm": 3.7244107723236084,
+      "learning_rate": 3.746347473993245e-06,
+      "loss": 0.9669,
+      "step": 7696
+    },
+    {
+      "epoch": 0.5933549182855381,
+      "grad_norm": 3.8219058513641357,
+      "learning_rate": 3.745138945079343e-06,
+      "loss": 0.9342,
+      "step": 7697
+    },
+    {
+      "epoch": 0.593432007400555,
+      "grad_norm": 3.733274459838867,
+      "learning_rate": 3.7439304943995274e-06,
+      "loss": 1.05,
+      "step": 7698
+    },
+    {
+      "epoch": 0.593509096515572,
+      "grad_norm": 3.3344218730926514,
+      "learning_rate": 3.742722122029142e-06,
+      "loss": 0.8988,
+      "step": 7699
+    },
+    {
+      "epoch": 0.593586185630589,
+      "grad_norm": 3.70858097076416,
+      "learning_rate": 3.741513828043519e-06,
+      "loss": 0.9501,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5936632747456059,
+      "grad_norm": 3.8456056118011475,
+      "learning_rate": 3.7403056125179916e-06,
+      "loss": 0.9357,
+      "step": 7701
+    },
+    {
+      "epoch": 0.5937403638606229,
+      "grad_norm": 3.785587787628174,
+      "learning_rate": 3.739097475527885e-06,
+      "loss": 0.9546,
+      "step": 7702
+    },
+    {
+      "epoch": 0.5938174529756398,
+      "grad_norm": 3.5768582820892334,
+      "learning_rate": 3.7378894171485203e-06,
+      "loss": 0.8708,
+      "step": 7703
+    },
+    {
+      "epoch": 0.5938945420906568,
+      "grad_norm": 3.9242618083953857,
+      "learning_rate": 3.736681437455214e-06,
+      "loss": 1.0719,
+      "step": 7704
+    },
+    {
+      "epoch": 0.5939716312056738,
+      "grad_norm": 3.8821098804473877,
+      "learning_rate": 3.7354735365232777e-06,
+      "loss": 0.914,
+      "step": 7705
+    },
+    {
+      "epoch": 0.5940487203206907,
+      "grad_norm": 3.5671024322509766,
+      "learning_rate": 3.7342657144280162e-06,
+      "loss": 1.0382,
+      "step": 7706
+    },
+    {
+      "epoch": 0.5941258094357077,
+      "grad_norm": 4.059629917144775,
+      "learning_rate": 3.7330579712447324e-06,
+      "loss": 0.9346,
+      "step": 7707
+    },
+    {
+      "epoch": 0.5942028985507246,
+      "grad_norm": 3.79659366607666,
+      "learning_rate": 3.7318503070487235e-06,
+      "loss": 0.9125,
+      "step": 7708
+    },
+    {
+      "epoch": 0.5942799876657416,
+      "grad_norm": 3.4974241256713867,
+      "learning_rate": 3.7306427219152786e-06,
+      "loss": 0.9265,
+      "step": 7709
+    },
+    {
+      "epoch": 0.5943570767807586,
+      "grad_norm": 3.5367648601531982,
+      "learning_rate": 3.7294352159196865e-06,
+      "loss": 0.9146,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5944341658957755,
+      "grad_norm": 3.3127148151397705,
+      "learning_rate": 3.7282277891372287e-06,
+      "loss": 0.9294,
+      "step": 7711
+    },
+    {
+      "epoch": 0.5945112550107925,
+      "grad_norm": 3.6922547817230225,
+      "learning_rate": 3.72702044164318e-06,
+      "loss": 0.9764,
+      "step": 7712
+    },
+    {
+      "epoch": 0.5945883441258094,
+      "grad_norm": 3.6928598880767822,
+      "learning_rate": 3.7258131735128156e-06,
+      "loss": 1.0221,
+      "step": 7713
+    },
+    {
+      "epoch": 0.5946654332408264,
+      "grad_norm": 3.2286767959594727,
+      "learning_rate": 3.7246059848213996e-06,
+      "loss": 0.736,
+      "step": 7714
+    },
+    {
+      "epoch": 0.5947425223558434,
+      "grad_norm": 3.4887712001800537,
+      "learning_rate": 3.7233988756441953e-06,
+      "loss": 0.9037,
+      "step": 7715
+    },
+    {
+      "epoch": 0.5948196114708603,
+      "grad_norm": 4.033304691314697,
+      "learning_rate": 3.72219184605646e-06,
+      "loss": 0.956,
+      "step": 7716
+    },
+    {
+      "epoch": 0.5948967005858773,
+      "grad_norm": 3.5219640731811523,
+      "learning_rate": 3.720984896133444e-06,
+      "loss": 0.9298,
+      "step": 7717
+    },
+    {
+      "epoch": 0.5949737897008942,
+      "grad_norm": 3.6129660606384277,
+      "learning_rate": 3.719778025950397e-06,
+      "loss": 0.9154,
+      "step": 7718
+    },
+    {
+      "epoch": 0.5950508788159112,
+      "grad_norm": 3.6359000205993652,
+      "learning_rate": 3.7185712355825577e-06,
+      "loss": 0.9319,
+      "step": 7719
+    },
+    {
+      "epoch": 0.5951279679309281,
+      "grad_norm": 3.6915714740753174,
+      "learning_rate": 3.717364525105166e-06,
+      "loss": 0.899,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5952050570459451,
+      "grad_norm": 3.943981409072876,
+      "learning_rate": 3.716157894593454e-06,
+      "loss": 1.0859,
+      "step": 7721
+    },
+    {
+      "epoch": 0.5952821461609621,
+      "grad_norm": 3.356869697570801,
+      "learning_rate": 3.714951344122647e-06,
+      "loss": 0.8368,
+      "step": 7722
+    },
+    {
+      "epoch": 0.595359235275979,
+      "grad_norm": 3.6155312061309814,
+      "learning_rate": 3.7137448737679677e-06,
+      "loss": 0.8818,
+      "step": 7723
+    },
+    {
+      "epoch": 0.595436324390996,
+      "grad_norm": 3.962568759918213,
+      "learning_rate": 3.712538483604634e-06,
+      "loss": 0.93,
+      "step": 7724
+    },
+    {
+      "epoch": 0.595513413506013,
+      "grad_norm": 3.563605785369873,
+      "learning_rate": 3.7113321737078587e-06,
+      "loss": 0.872,
+      "step": 7725
+    },
+    {
+      "epoch": 0.5955905026210299,
+      "grad_norm": 3.954345226287842,
+      "learning_rate": 3.710125944152849e-06,
+      "loss": 0.9366,
+      "step": 7726
+    },
+    {
+      "epoch": 0.5956675917360469,
+      "grad_norm": 3.7304627895355225,
+      "learning_rate": 3.7089197950148054e-06,
+      "loss": 0.9616,
+      "step": 7727
+    },
+    {
+      "epoch": 0.5957446808510638,
+      "grad_norm": 3.6339216232299805,
+      "learning_rate": 3.7077137263689266e-06,
+      "loss": 0.8768,
+      "step": 7728
+    },
+    {
+      "epoch": 0.5958217699660808,
+      "grad_norm": 3.9956276416778564,
+      "learning_rate": 3.706507738290403e-06,
+      "loss": 1.0666,
+      "step": 7729
+    },
+    {
+      "epoch": 0.5958988590810977,
+      "grad_norm": 3.52579402923584,
+      "learning_rate": 3.705301830854423e-06,
+      "loss": 0.9173,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5959759481961147,
+      "grad_norm": 3.9058704376220703,
+      "learning_rate": 3.7040960041361696e-06,
+      "loss": 1.0185,
+      "step": 7731
+    },
+    {
+      "epoch": 0.5960530373111317,
+      "grad_norm": 4.03030252456665,
+      "learning_rate": 3.7028902582108174e-06,
+      "loss": 1.025,
+      "step": 7732
+    },
+    {
+      "epoch": 0.5961301264261486,
+      "grad_norm": 4.030737400054932,
+      "learning_rate": 3.7016845931535395e-06,
+      "loss": 0.9925,
+      "step": 7733
+    },
+    {
+      "epoch": 0.5962072155411656,
+      "grad_norm": 3.7005774974823,
+      "learning_rate": 3.7004790090395043e-06,
+      "loss": 0.9436,
+      "step": 7734
+    },
+    {
+      "epoch": 0.5962843046561825,
+      "grad_norm": 3.9587509632110596,
+      "learning_rate": 3.699273505943871e-06,
+      "loss": 1.0326,
+      "step": 7735
+    },
+    {
+      "epoch": 0.5963613937711995,
+      "grad_norm": 3.745950937271118,
+      "learning_rate": 3.6980680839418e-06,
+      "loss": 0.8556,
+      "step": 7736
+    },
+    {
+      "epoch": 0.5964384828862165,
+      "grad_norm": 3.4118099212646484,
+      "learning_rate": 3.696862743108439e-06,
+      "loss": 0.9266,
+      "step": 7737
+    },
+    {
+      "epoch": 0.5965155720012334,
+      "grad_norm": 3.6588337421417236,
+      "learning_rate": 3.6956574835189374e-06,
+      "loss": 0.9452,
+      "step": 7738
+    },
+    {
+      "epoch": 0.5965926611162504,
+      "grad_norm": 3.32086443901062,
+      "learning_rate": 3.694452305248437e-06,
+      "loss": 0.9146,
+      "step": 7739
+    },
+    {
+      "epoch": 0.5966697502312673,
+      "grad_norm": 3.491138458251953,
+      "learning_rate": 3.693247208372074e-06,
+      "loss": 0.7812,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5967468393462843,
+      "grad_norm": 3.2790908813476562,
+      "learning_rate": 3.6920421929649786e-06,
+      "loss": 0.7925,
+      "step": 7741
+    },
+    {
+      "epoch": 0.5968239284613013,
+      "grad_norm": 3.902376174926758,
+      "learning_rate": 3.690837259102279e-06,
+      "loss": 0.991,
+      "step": 7742
+    },
+    {
+      "epoch": 0.5969010175763182,
+      "grad_norm": 3.6180334091186523,
+      "learning_rate": 3.689632406859096e-06,
+      "loss": 0.8429,
+      "step": 7743
+    },
+    {
+      "epoch": 0.5969781066913352,
+      "grad_norm": 4.053879737854004,
+      "learning_rate": 3.688427636310545e-06,
+      "loss": 1.1251,
+      "step": 7744
+    },
+    {
+      "epoch": 0.5970551958063521,
+      "grad_norm": 4.163644790649414,
+      "learning_rate": 3.68722294753174e-06,
+      "loss": 0.9068,
+      "step": 7745
+    },
+    {
+      "epoch": 0.597132284921369,
+      "grad_norm": 3.3807566165924072,
+      "learning_rate": 3.6860183405977833e-06,
+      "loss": 0.8411,
+      "step": 7746
+    },
+    {
+      "epoch": 0.5972093740363861,
+      "grad_norm": 3.3274459838867188,
+      "learning_rate": 3.6848138155837786e-06,
+      "loss": 0.7565,
+      "step": 7747
+    },
+    {
+      "epoch": 0.597286463151403,
+      "grad_norm": 3.35614013671875,
+      "learning_rate": 3.683609372564821e-06,
+      "loss": 0.7392,
+      "step": 7748
+    },
+    {
+      "epoch": 0.59736355226642,
+      "grad_norm": 4.015716075897217,
+      "learning_rate": 3.6824050116160002e-06,
+      "loss": 1.0421,
+      "step": 7749
+    },
+    {
+      "epoch": 0.5974406413814369,
+      "grad_norm": 3.730281114578247,
+      "learning_rate": 3.681200732812405e-06,
+      "loss": 1.0248,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5975177304964538,
+      "grad_norm": 3.4438388347625732,
+      "learning_rate": 3.6799965362291123e-06,
+      "loss": 0.8048,
+      "step": 7751
+    },
+    {
+      "epoch": 0.5975948196114709,
+      "grad_norm": 3.696259021759033,
+      "learning_rate": 3.678792421941199e-06,
+      "loss": 0.9177,
+      "step": 7752
+    },
+    {
+      "epoch": 0.5976719087264878,
+      "grad_norm": 3.7113966941833496,
+      "learning_rate": 3.677588390023736e-06,
+      "loss": 0.9895,
+      "step": 7753
+    },
+    {
+      "epoch": 0.5977489978415048,
+      "grad_norm": 3.3123090267181396,
+      "learning_rate": 3.6763844405517877e-06,
+      "loss": 0.8952,
+      "step": 7754
+    },
+    {
+      "epoch": 0.5978260869565217,
+      "grad_norm": 3.5116400718688965,
+      "learning_rate": 3.6751805736004148e-06,
+      "loss": 0.9271,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5979031760715386,
+      "grad_norm": 3.612499237060547,
+      "learning_rate": 3.673976789244672e-06,
+      "loss": 0.8571,
+      "step": 7756
+    },
+    {
+      "epoch": 0.5979802651865557,
+      "grad_norm": 3.808100938796997,
+      "learning_rate": 3.672773087559608e-06,
+      "loss": 0.9372,
+      "step": 7757
+    },
+    {
+      "epoch": 0.5980573543015726,
+      "grad_norm": 3.6125247478485107,
+      "learning_rate": 3.6715694686202686e-06,
+      "loss": 0.93,
+      "step": 7758
+    },
+    {
+      "epoch": 0.5981344434165896,
+      "grad_norm": 3.831362724304199,
+      "learning_rate": 3.6703659325016927e-06,
+      "loss": 0.8422,
+      "step": 7759
+    },
+    {
+      "epoch": 0.5982115325316065,
+      "grad_norm": 3.4651618003845215,
+      "learning_rate": 3.669162479278914e-06,
+      "loss": 0.9065,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5982886216466234,
+      "grad_norm": 3.6976826190948486,
+      "learning_rate": 3.667959109026963e-06,
+      "loss": 0.8673,
+      "step": 7761
+    },
+    {
+      "epoch": 0.5983657107616405,
+      "grad_norm": 3.865349292755127,
+      "learning_rate": 3.6667558218208617e-06,
+      "loss": 0.988,
+      "step": 7762
+    },
+    {
+      "epoch": 0.5984427998766574,
+      "grad_norm": 3.4592785835266113,
+      "learning_rate": 3.665552617735629e-06,
+      "loss": 0.9485,
+      "step": 7763
+    },
+    {
+      "epoch": 0.5985198889916744,
+      "grad_norm": 3.4235899448394775,
+      "learning_rate": 3.6643494968462822e-06,
+      "loss": 0.867,
+      "step": 7764
+    },
+    {
+      "epoch": 0.5985969781066913,
+      "grad_norm": 3.647528648376465,
+      "learning_rate": 3.663146459227824e-06,
+      "loss": 0.851,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5986740672217082,
+      "grad_norm": 3.5159714221954346,
+      "learning_rate": 3.6619435049552633e-06,
+      "loss": 0.908,
+      "step": 7766
+    },
+    {
+      "epoch": 0.5987511563367253,
+      "grad_norm": 3.558000326156616,
+      "learning_rate": 3.66074063410359e-06,
+      "loss": 0.9187,
+      "step": 7767
+    },
+    {
+      "epoch": 0.5988282454517422,
+      "grad_norm": 3.671823263168335,
+      "learning_rate": 3.659537846747806e-06,
+      "loss": 0.9016,
+      "step": 7768
+    },
+    {
+      "epoch": 0.5989053345667592,
+      "grad_norm": 3.4551210403442383,
+      "learning_rate": 3.658335142962894e-06,
+      "loss": 0.8888,
+      "step": 7769
+    },
+    {
+      "epoch": 0.5989824236817761,
+      "grad_norm": 3.604067087173462,
+      "learning_rate": 3.657132522823837e-06,
+      "loss": 0.9862,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5990595127967931,
+      "grad_norm": 3.8700289726257324,
+      "learning_rate": 3.655929986405613e-06,
+      "loss": 0.963,
+      "step": 7771
+    },
+    {
+      "epoch": 0.5991366019118101,
+      "grad_norm": 3.566358804702759,
+      "learning_rate": 3.6547275337831922e-06,
+      "loss": 0.8178,
+      "step": 7772
+    },
+    {
+      "epoch": 0.599213691026827,
+      "grad_norm": 3.576695203781128,
+      "learning_rate": 3.653525165031543e-06,
+      "loss": 0.9241,
+      "step": 7773
+    },
+    {
+      "epoch": 0.599290780141844,
+      "grad_norm": 3.5884556770324707,
+      "learning_rate": 3.6523228802256264e-06,
+      "loss": 0.8936,
+      "step": 7774
+    },
+    {
+      "epoch": 0.5993678692568609,
+      "grad_norm": 3.7305357456207275,
+      "learning_rate": 3.651120679440398e-06,
+      "loss": 0.8829,
+      "step": 7775
+    },
+    {
+      "epoch": 0.599444958371878,
+      "grad_norm": 3.6430912017822266,
+      "learning_rate": 3.6499185627508098e-06,
+      "loss": 0.9241,
+      "step": 7776
+    },
+    {
+      "epoch": 0.5995220474868949,
+      "grad_norm": 3.6668894290924072,
+      "learning_rate": 3.648716530231806e-06,
+      "loss": 0.8658,
+      "step": 7777
+    },
+    {
+      "epoch": 0.5995991366019118,
+      "grad_norm": 3.5166971683502197,
+      "learning_rate": 3.6475145819583285e-06,
+      "loss": 0.9614,
+      "step": 7778
+    },
+    {
+      "epoch": 0.5996762257169288,
+      "grad_norm": 3.9668848514556885,
+      "learning_rate": 3.6463127180053125e-06,
+      "loss": 1.0337,
+      "step": 7779
+    },
+    {
+      "epoch": 0.5997533148319457,
+      "grad_norm": 3.5384409427642822,
+      "learning_rate": 3.6451109384476875e-06,
+      "loss": 0.9416,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5998304039469627,
+      "grad_norm": 3.4631733894348145,
+      "learning_rate": 3.6439092433603775e-06,
+      "loss": 0.819,
+      "step": 7781
+    },
+    {
+      "epoch": 0.5999074930619797,
+      "grad_norm": 3.906912326812744,
+      "learning_rate": 3.642707632818304e-06,
+      "loss": 0.908,
+      "step": 7782
+    },
+    {
+      "epoch": 0.5999845821769966,
+      "grad_norm": 3.743074893951416,
+      "learning_rate": 3.641506106896379e-06,
+      "loss": 0.9959,
+      "step": 7783
+    },
+    {
+      "epoch": 0.6000616712920136,
+      "grad_norm": 3.657150983810425,
+      "learning_rate": 3.640304665669514e-06,
+      "loss": 0.9171,
+      "step": 7784
+    },
+    {
+      "epoch": 0.6001387604070305,
+      "grad_norm": 3.718385934829712,
+      "learning_rate": 3.639103309212609e-06,
+      "loss": 0.8975,
+      "step": 7785
+    },
+    {
+      "epoch": 0.6002158495220475,
+      "grad_norm": 3.5247554779052734,
+      "learning_rate": 3.6379020376005646e-06,
+      "loss": 0.8734,
+      "step": 7786
+    },
+    {
+      "epoch": 0.6002929386370645,
+      "grad_norm": 3.818239688873291,
+      "learning_rate": 3.6367008509082757e-06,
+      "loss": 0.9179,
+      "step": 7787
+    },
+    {
+      "epoch": 0.6003700277520814,
+      "grad_norm": 4.0803728103637695,
+      "learning_rate": 3.6354997492106258e-06,
+      "loss": 0.9927,
+      "step": 7788
+    },
+    {
+      "epoch": 0.6004471168670984,
+      "grad_norm": 3.478654384613037,
+      "learning_rate": 3.634298732582501e-06,
+      "loss": 0.9289,
+      "step": 7789
+    },
+    {
+      "epoch": 0.6005242059821153,
+      "grad_norm": 3.6446428298950195,
+      "learning_rate": 3.6330978010987767e-06,
+      "loss": 0.9966,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6006012950971323,
+      "grad_norm": 3.805497407913208,
+      "learning_rate": 3.6318969548343246e-06,
+      "loss": 1.0771,
+      "step": 7791
+    },
+    {
+      "epoch": 0.6006783842121493,
+      "grad_norm": 3.4235615730285645,
+      "learning_rate": 3.6306961938640116e-06,
+      "loss": 0.9165,
+      "step": 7792
+    },
+    {
+      "epoch": 0.6007554733271662,
+      "grad_norm": 3.7066256999969482,
+      "learning_rate": 3.6294955182627007e-06,
+      "loss": 0.8412,
+      "step": 7793
+    },
+    {
+      "epoch": 0.6008325624421832,
+      "grad_norm": 3.6571426391601562,
+      "learning_rate": 3.628294928105245e-06,
+      "loss": 0.8972,
+      "step": 7794
+    },
+    {
+      "epoch": 0.6009096515572001,
+      "grad_norm": 3.8981006145477295,
+      "learning_rate": 3.6270944234664972e-06,
+      "loss": 0.8919,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6009867406722171,
+      "grad_norm": 3.7625627517700195,
+      "learning_rate": 3.6258940044213004e-06,
+      "loss": 0.8221,
+      "step": 7796
+    },
+    {
+      "epoch": 0.601063829787234,
+      "grad_norm": 4.007872104644775,
+      "learning_rate": 3.6246936710444957e-06,
+      "loss": 0.9362,
+      "step": 7797
+    },
+    {
+      "epoch": 0.601140918902251,
+      "grad_norm": 3.732297420501709,
+      "learning_rate": 3.623493423410919e-06,
+      "loss": 0.8811,
+      "step": 7798
+    },
+    {
+      "epoch": 0.601218008017268,
+      "grad_norm": 3.483485698699951,
+      "learning_rate": 3.6222932615953977e-06,
+      "loss": 0.9207,
+      "step": 7799
+    },
+    {
+      "epoch": 0.6012950971322849,
+      "grad_norm": 3.7129147052764893,
+      "learning_rate": 3.6210931856727547e-06,
+      "loss": 0.8513,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6013721862473019,
+      "grad_norm": 3.39215087890625,
+      "learning_rate": 3.619893195717813e-06,
+      "loss": 0.8216,
+      "step": 7801
+    },
+    {
+      "epoch": 0.6014492753623188,
+      "grad_norm": 3.39719295501709,
+      "learning_rate": 3.6186932918053806e-06,
+      "loss": 1.0014,
+      "step": 7802
+    },
+    {
+      "epoch": 0.6015263644773358,
+      "grad_norm": 3.573181629180908,
+      "learning_rate": 3.6174934740102672e-06,
+      "loss": 0.864,
+      "step": 7803
+    },
+    {
+      "epoch": 0.6016034535923528,
+      "grad_norm": 3.7137699127197266,
+      "learning_rate": 3.6162937424072775e-06,
+      "loss": 1.0308,
+      "step": 7804
+    },
+    {
+      "epoch": 0.6016805427073697,
+      "grad_norm": 3.913328170776367,
+      "learning_rate": 3.6150940970712045e-06,
+      "loss": 0.9982,
+      "step": 7805
+    },
+    {
+      "epoch": 0.6017576318223867,
+      "grad_norm": 3.644329071044922,
+      "learning_rate": 3.6138945380768442e-06,
+      "loss": 0.9262,
+      "step": 7806
+    },
+    {
+      "epoch": 0.6018347209374036,
+      "grad_norm": 3.936444044113159,
+      "learning_rate": 3.612695065498979e-06,
+      "loss": 0.8839,
+      "step": 7807
+    },
+    {
+      "epoch": 0.6019118100524206,
+      "grad_norm": 3.725987434387207,
+      "learning_rate": 3.6114956794123913e-06,
+      "loss": 0.9187,
+      "step": 7808
+    },
+    {
+      "epoch": 0.6019888991674376,
+      "grad_norm": 3.515115261077881,
+      "learning_rate": 3.6102963798918576e-06,
+      "loss": 0.9055,
+      "step": 7809
+    },
+    {
+      "epoch": 0.6020659882824545,
+      "grad_norm": 3.660930871963501,
+      "learning_rate": 3.6090971670121468e-06,
+      "loss": 0.8665,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6021430773974715,
+      "grad_norm": 3.7818009853363037,
+      "learning_rate": 3.6078980408480217e-06,
+      "loss": 1.0293,
+      "step": 7811
+    },
+    {
+      "epoch": 0.6022201665124884,
+      "grad_norm": 4.332160949707031,
+      "learning_rate": 3.606699001474246e-06,
+      "loss": 1.1111,
+      "step": 7812
+    },
+    {
+      "epoch": 0.6022972556275054,
+      "grad_norm": 3.5895073413848877,
+      "learning_rate": 3.6055000489655725e-06,
+      "loss": 0.8604,
+      "step": 7813
+    },
+    {
+      "epoch": 0.6023743447425224,
+      "grad_norm": 3.754880666732788,
+      "learning_rate": 3.6043011833967465e-06,
+      "loss": 0.9298,
+      "step": 7814
+    },
+    {
+      "epoch": 0.6024514338575393,
+      "grad_norm": 4.089043140411377,
+      "learning_rate": 3.6031024048425146e-06,
+      "loss": 1.0692,
+      "step": 7815
+    },
+    {
+      "epoch": 0.6025285229725563,
+      "grad_norm": 3.798560619354248,
+      "learning_rate": 3.6019037133776136e-06,
+      "loss": 0.918,
+      "step": 7816
+    },
+    {
+      "epoch": 0.6026056120875732,
+      "grad_norm": 3.821998357772827,
+      "learning_rate": 3.6007051090767737e-06,
+      "loss": 0.9856,
+      "step": 7817
+    },
+    {
+      "epoch": 0.6026827012025902,
+      "grad_norm": 3.517061948776245,
+      "learning_rate": 3.5995065920147233e-06,
+      "loss": 0.8683,
+      "step": 7818
+    },
+    {
+      "epoch": 0.6027597903176072,
+      "grad_norm": 3.6391801834106445,
+      "learning_rate": 3.598308162266185e-06,
+      "loss": 0.8445,
+      "step": 7819
+    },
+    {
+      "epoch": 0.6028368794326241,
+      "grad_norm": 3.6682894229888916,
+      "learning_rate": 3.5971098199058716e-06,
+      "loss": 0.9821,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6029139685476411,
+      "grad_norm": 4.120725154876709,
+      "learning_rate": 3.5959115650084954e-06,
+      "loss": 0.8732,
+      "step": 7821
+    },
+    {
+      "epoch": 0.602991057662658,
+      "grad_norm": 3.8292622566223145,
+      "learning_rate": 3.5947133976487623e-06,
+      "loss": 0.986,
+      "step": 7822
+    },
+    {
+      "epoch": 0.603068146777675,
+      "grad_norm": 3.57987904548645,
+      "learning_rate": 3.5935153179013692e-06,
+      "loss": 0.8983,
+      "step": 7823
+    },
+    {
+      "epoch": 0.603145235892692,
+      "grad_norm": 3.4487192630767822,
+      "learning_rate": 3.592317325841014e-06,
+      "loss": 0.8779,
+      "step": 7824
+    },
+    {
+      "epoch": 0.6032223250077089,
+      "grad_norm": 3.598322629928589,
+      "learning_rate": 3.5911194215423805e-06,
+      "loss": 0.9429,
+      "step": 7825
+    },
+    {
+      "epoch": 0.6032994141227259,
+      "grad_norm": 3.9131698608398438,
+      "learning_rate": 3.5899216050801548e-06,
+      "loss": 1.0573,
+      "step": 7826
+    },
+    {
+      "epoch": 0.6033765032377428,
+      "grad_norm": 3.5165915489196777,
+      "learning_rate": 3.5887238765290155e-06,
+      "loss": 0.8065,
+      "step": 7827
+    },
+    {
+      "epoch": 0.6034535923527597,
+      "grad_norm": 3.4802615642547607,
+      "learning_rate": 3.5875262359636316e-06,
+      "loss": 0.9431,
+      "step": 7828
+    },
+    {
+      "epoch": 0.6035306814677768,
+      "grad_norm": 4.010217666625977,
+      "learning_rate": 3.586328683458672e-06,
+      "loss": 0.9236,
+      "step": 7829
+    },
+    {
+      "epoch": 0.6036077705827937,
+      "grad_norm": 3.4742398262023926,
+      "learning_rate": 3.5851312190887975e-06,
+      "loss": 0.9148,
+      "step": 7830
+    },
+    {
+      "epoch": 0.6036848596978107,
+      "grad_norm": 3.8777101039886475,
+      "learning_rate": 3.583933842928664e-06,
+      "loss": 0.9534,
+      "step": 7831
+    },
+    {
+      "epoch": 0.6037619488128276,
+      "grad_norm": 3.97666335105896,
+      "learning_rate": 3.5827365550529215e-06,
+      "loss": 1.0859,
+      "step": 7832
+    },
+    {
+      "epoch": 0.6038390379278445,
+      "grad_norm": 3.1657111644744873,
+      "learning_rate": 3.581539355536214e-06,
+      "loss": 0.9061,
+      "step": 7833
+    },
+    {
+      "epoch": 0.6039161270428616,
+      "grad_norm": 3.550740957260132,
+      "learning_rate": 3.580342244453181e-06,
+      "loss": 0.9728,
+      "step": 7834
+    },
+    {
+      "epoch": 0.6039932161578785,
+      "grad_norm": 3.700817108154297,
+      "learning_rate": 3.5791452218784572e-06,
+      "loss": 0.9626,
+      "step": 7835
+    },
+    {
+      "epoch": 0.6040703052728955,
+      "grad_norm": 3.693549633026123,
+      "learning_rate": 3.57794828788667e-06,
+      "loss": 1.0247,
+      "step": 7836
+    },
+    {
+      "epoch": 0.6041473943879124,
+      "grad_norm": 3.4788200855255127,
+      "learning_rate": 3.5767514425524413e-06,
+      "loss": 0.9126,
+      "step": 7837
+    },
+    {
+      "epoch": 0.6042244835029293,
+      "grad_norm": 3.526183843612671,
+      "learning_rate": 3.57555468595039e-06,
+      "loss": 1.0105,
+      "step": 7838
+    },
+    {
+      "epoch": 0.6043015726179464,
+      "grad_norm": 3.612081289291382,
+      "learning_rate": 3.5743580181551265e-06,
+      "loss": 0.9198,
+      "step": 7839
+    },
+    {
+      "epoch": 0.6043786617329633,
+      "grad_norm": 3.6444640159606934,
+      "learning_rate": 3.5731614392412557e-06,
+      "loss": 0.9105,
+      "step": 7840
+    },
+    {
+      "epoch": 0.6044557508479803,
+      "grad_norm": 3.4978742599487305,
+      "learning_rate": 3.571964949283381e-06,
+      "loss": 0.8368,
+      "step": 7841
+    },
+    {
+      "epoch": 0.6045328399629972,
+      "grad_norm": 3.5534303188323975,
+      "learning_rate": 3.5707685483560948e-06,
+      "loss": 0.9524,
+      "step": 7842
+    },
+    {
+      "epoch": 0.6046099290780141,
+      "grad_norm": 3.6037681102752686,
+      "learning_rate": 3.569572236533988e-06,
+      "loss": 0.9408,
+      "step": 7843
+    },
+    {
+      "epoch": 0.6046870181930312,
+      "grad_norm": 3.651366949081421,
+      "learning_rate": 3.5683760138916433e-06,
+      "loss": 0.8935,
+      "step": 7844
+    },
+    {
+      "epoch": 0.6047641073080481,
+      "grad_norm": 3.578256368637085,
+      "learning_rate": 3.5671798805036396e-06,
+      "loss": 0.9498,
+      "step": 7845
+    },
+    {
+      "epoch": 0.6048411964230651,
+      "grad_norm": 4.041118621826172,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 0.9394,
+      "step": 7846
+    },
+    {
+      "epoch": 0.604918285538082,
+      "grad_norm": 3.86403489112854,
+      "learning_rate": 3.564787881788941e-06,
+      "loss": 0.9925,
+      "step": 7847
+    },
+    {
+      "epoch": 0.6049953746530989,
+      "grad_norm": 3.811384916305542,
+      "learning_rate": 3.5635920166113735e-06,
+      "loss": 0.9659,
+      "step": 7848
+    },
+    {
+      "epoch": 0.605072463768116,
+      "grad_norm": 3.6318352222442627,
+      "learning_rate": 3.5623962409864066e-06,
+      "loss": 0.8375,
+      "step": 7849
+    },
+    {
+      "epoch": 0.6051495528831329,
+      "grad_norm": 3.6698193550109863,
+      "learning_rate": 3.5612005549885865e-06,
+      "loss": 1.0244,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6052266419981499,
+      "grad_norm": 3.5718917846679688,
+      "learning_rate": 3.5600049586924607e-06,
+      "loss": 0.9182,
+      "step": 7851
+    },
+    {
+      "epoch": 0.6053037311131668,
+      "grad_norm": 3.9225900173187256,
+      "learning_rate": 3.5588094521725687e-06,
+      "loss": 0.9336,
+      "step": 7852
+    },
+    {
+      "epoch": 0.6053808202281837,
+      "grad_norm": 3.1916608810424805,
+      "learning_rate": 3.5576140355034415e-06,
+      "loss": 0.8681,
+      "step": 7853
+    },
+    {
+      "epoch": 0.6054579093432008,
+      "grad_norm": 3.703646659851074,
+      "learning_rate": 3.5564187087596116e-06,
+      "loss": 0.9635,
+      "step": 7854
+    },
+    {
+      "epoch": 0.6055349984582177,
+      "grad_norm": 3.636467218399048,
+      "learning_rate": 3.5552234720155963e-06,
+      "loss": 0.8925,
+      "step": 7855
+    },
+    {
+      "epoch": 0.6056120875732347,
+      "grad_norm": 3.733719825744629,
+      "learning_rate": 3.554028325345914e-06,
+      "loss": 0.9168,
+      "step": 7856
+    },
+    {
+      "epoch": 0.6056891766882516,
+      "grad_norm": 3.497403860092163,
+      "learning_rate": 3.552833268825078e-06,
+      "loss": 0.9535,
+      "step": 7857
+    },
+    {
+      "epoch": 0.6057662658032685,
+      "grad_norm": 4.8122382164001465,
+      "learning_rate": 3.5516383025275925e-06,
+      "loss": 0.9782,
+      "step": 7858
+    },
+    {
+      "epoch": 0.6058433549182856,
+      "grad_norm": 3.9689764976501465,
+      "learning_rate": 3.5504434265279587e-06,
+      "loss": 1.006,
+      "step": 7859
+    },
+    {
+      "epoch": 0.6059204440333025,
+      "grad_norm": 3.636197566986084,
+      "learning_rate": 3.5492486409006684e-06,
+      "loss": 0.9264,
+      "step": 7860
+    },
+    {
+      "epoch": 0.6059975331483195,
+      "grad_norm": 4.189083099365234,
+      "learning_rate": 3.5480539457202127e-06,
+      "loss": 0.9417,
+      "step": 7861
+    },
+    {
+      "epoch": 0.6060746222633364,
+      "grad_norm": 3.744331121444702,
+      "learning_rate": 3.546859341061073e-06,
+      "loss": 0.8417,
+      "step": 7862
+    },
+    {
+      "epoch": 0.6061517113783533,
+      "grad_norm": 3.3087058067321777,
+      "learning_rate": 3.545664826997727e-06,
+      "loss": 0.911,
+      "step": 7863
+    },
+    {
+      "epoch": 0.6062288004933704,
+      "grad_norm": 3.681582450866699,
+      "learning_rate": 3.5444704036046485e-06,
+      "loss": 0.9636,
+      "step": 7864
+    },
+    {
+      "epoch": 0.6063058896083873,
+      "grad_norm": 3.7215073108673096,
+      "learning_rate": 3.543276070956301e-06,
+      "loss": 0.9945,
+      "step": 7865
+    },
+    {
+      "epoch": 0.6063829787234043,
+      "grad_norm": 3.6413660049438477,
+      "learning_rate": 3.542081829127145e-06,
+      "loss": 0.8434,
+      "step": 7866
+    },
+    {
+      "epoch": 0.6064600678384212,
+      "grad_norm": 3.857893705368042,
+      "learning_rate": 3.540887678191638e-06,
+      "loss": 0.9055,
+      "step": 7867
+    },
+    {
+      "epoch": 0.6065371569534381,
+      "grad_norm": 3.426194906234741,
+      "learning_rate": 3.539693618224226e-06,
+      "loss": 0.9556,
+      "step": 7868
+    },
+    {
+      "epoch": 0.6066142460684552,
+      "grad_norm": 3.5567467212677,
+      "learning_rate": 3.538499649299354e-06,
+      "loss": 0.9462,
+      "step": 7869
+    },
+    {
+      "epoch": 0.6066913351834721,
+      "grad_norm": 3.7717530727386475,
+      "learning_rate": 3.5373057714914607e-06,
+      "loss": 0.9073,
+      "step": 7870
+    },
+    {
+      "epoch": 0.6067684242984891,
+      "grad_norm": 4.004831790924072,
+      "learning_rate": 3.5361119848749755e-06,
+      "loss": 0.9405,
+      "step": 7871
+    },
+    {
+      "epoch": 0.606845513413506,
+      "grad_norm": 3.8351056575775146,
+      "learning_rate": 3.5349182895243274e-06,
+      "loss": 1.0261,
+      "step": 7872
+    },
+    {
+      "epoch": 0.6069226025285229,
+      "grad_norm": 3.8197968006134033,
+      "learning_rate": 3.533724685513935e-06,
+      "loss": 0.9427,
+      "step": 7873
+    },
+    {
+      "epoch": 0.60699969164354,
+      "grad_norm": 3.7887625694274902,
+      "learning_rate": 3.5325311729182155e-06,
+      "loss": 1.1297,
+      "step": 7874
+    },
+    {
+      "epoch": 0.6070767807585569,
+      "grad_norm": 3.8279473781585693,
+      "learning_rate": 3.5313377518115772e-06,
+      "loss": 0.9428,
+      "step": 7875
+    },
+    {
+      "epoch": 0.6071538698735739,
+      "grad_norm": 3.926093101501465,
+      "learning_rate": 3.530144422268423e-06,
+      "loss": 0.8763,
+      "step": 7876
+    },
+    {
+      "epoch": 0.6072309589885908,
+      "grad_norm": 3.941603183746338,
+      "learning_rate": 3.528951184363151e-06,
+      "loss": 1.0133,
+      "step": 7877
+    },
+    {
+      "epoch": 0.6073080481036077,
+      "grad_norm": 3.6375105381011963,
+      "learning_rate": 3.5277580381701553e-06,
+      "loss": 0.9153,
+      "step": 7878
+    },
+    {
+      "epoch": 0.6073851372186247,
+      "grad_norm": 3.378063678741455,
+      "learning_rate": 3.52656498376382e-06,
+      "loss": 0.8914,
+      "step": 7879
+    },
+    {
+      "epoch": 0.6074622263336417,
+      "grad_norm": 3.7920594215393066,
+      "learning_rate": 3.5253720212185284e-06,
+      "loss": 1.0532,
+      "step": 7880
+    },
+    {
+      "epoch": 0.6075393154486587,
+      "grad_norm": 3.6664042472839355,
+      "learning_rate": 3.524179150608652e-06,
+      "loss": 0.7833,
+      "step": 7881
+    },
+    {
+      "epoch": 0.6076164045636756,
+      "grad_norm": 3.3724348545074463,
+      "learning_rate": 3.5229863720085623e-06,
+      "loss": 0.9102,
+      "step": 7882
+    },
+    {
+      "epoch": 0.6076934936786925,
+      "grad_norm": 3.686537504196167,
+      "learning_rate": 3.521793685492624e-06,
+      "loss": 0.9696,
+      "step": 7883
+    },
+    {
+      "epoch": 0.6077705827937095,
+      "grad_norm": 3.747305154800415,
+      "learning_rate": 3.5206010911351924e-06,
+      "loss": 1.0284,
+      "step": 7884
+    },
+    {
+      "epoch": 0.6078476719087265,
+      "grad_norm": 3.655407667160034,
+      "learning_rate": 3.5194085890106204e-06,
+      "loss": 0.9406,
+      "step": 7885
+    },
+    {
+      "epoch": 0.6079247610237435,
+      "grad_norm": 3.6293342113494873,
+      "learning_rate": 3.518216179193257e-06,
+      "loss": 0.9043,
+      "step": 7886
+    },
+    {
+      "epoch": 0.6080018501387604,
+      "grad_norm": 3.54872465133667,
+      "learning_rate": 3.517023861757438e-06,
+      "loss": 1.0235,
+      "step": 7887
+    },
+    {
+      "epoch": 0.6080789392537773,
+      "grad_norm": 4.1367106437683105,
+      "learning_rate": 3.5158316367775016e-06,
+      "loss": 0.968,
+      "step": 7888
+    },
+    {
+      "epoch": 0.6081560283687943,
+      "grad_norm": 3.3810555934906006,
+      "learning_rate": 3.514639504327776e-06,
+      "loss": 0.8467,
+      "step": 7889
+    },
+    {
+      "epoch": 0.6082331174838113,
+      "grad_norm": 3.417717456817627,
+      "learning_rate": 3.5134474644825847e-06,
+      "loss": 0.8455,
+      "step": 7890
+    },
+    {
+      "epoch": 0.6083102065988283,
+      "grad_norm": 3.642557144165039,
+      "learning_rate": 3.512255517316245e-06,
+      "loss": 0.9056,
+      "step": 7891
+    },
+    {
+      "epoch": 0.6083872957138452,
+      "grad_norm": 3.8719959259033203,
+      "learning_rate": 3.5110636629030674e-06,
+      "loss": 0.9462,
+      "step": 7892
+    },
+    {
+      "epoch": 0.6084643848288621,
+      "grad_norm": 3.6329193115234375,
+      "learning_rate": 3.509871901317359e-06,
+      "loss": 0.9395,
+      "step": 7893
+    },
+    {
+      "epoch": 0.6085414739438791,
+      "grad_norm": 4.278444290161133,
+      "learning_rate": 3.5086802326334217e-06,
+      "loss": 0.946,
+      "step": 7894
+    },
+    {
+      "epoch": 0.6086185630588961,
+      "grad_norm": 3.6533453464508057,
+      "learning_rate": 3.507488656925547e-06,
+      "loss": 0.9943,
+      "step": 7895
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 3.3873438835144043,
+      "learning_rate": 3.5062971742680244e-06,
+      "loss": 0.9275,
+      "step": 7896
+    },
+    {
+      "epoch": 0.60877274128893,
+      "grad_norm": 3.647336006164551,
+      "learning_rate": 3.5051057847351377e-06,
+      "loss": 0.8534,
+      "step": 7897
+    },
+    {
+      "epoch": 0.6088498304039469,
+      "grad_norm": 3.4682047367095947,
+      "learning_rate": 3.503914488401163e-06,
+      "loss": 0.9414,
+      "step": 7898
+    },
+    {
+      "epoch": 0.6089269195189639,
+      "grad_norm": 3.890171527862549,
+      "learning_rate": 3.5027232853403705e-06,
+      "loss": 1.0898,
+      "step": 7899
+    },
+    {
+      "epoch": 0.6090040086339809,
+      "grad_norm": 3.491349220275879,
+      "learning_rate": 3.501532175627026e-06,
+      "loss": 0.9413,
+      "step": 7900
+    },
+    {
+      "epoch": 0.6090810977489979,
+      "grad_norm": 3.580474376678467,
+      "learning_rate": 3.5003411593353913e-06,
+      "loss": 0.9499,
+      "step": 7901
+    },
+    {
+      "epoch": 0.6091581868640148,
+      "grad_norm": 3.8808116912841797,
+      "learning_rate": 3.4991502365397177e-06,
+      "loss": 0.9255,
+      "step": 7902
+    },
+    {
+      "epoch": 0.6092352759790317,
+      "grad_norm": 3.8387744426727295,
+      "learning_rate": 3.4979594073142535e-06,
+      "loss": 0.7857,
+      "step": 7903
+    },
+    {
+      "epoch": 0.6093123650940487,
+      "grad_norm": 3.8449535369873047,
+      "learning_rate": 3.4967686717332426e-06,
+      "loss": 0.851,
+      "step": 7904
+    },
+    {
+      "epoch": 0.6093894542090657,
+      "grad_norm": 4.096821308135986,
+      "learning_rate": 3.495578029870918e-06,
+      "loss": 0.9903,
+      "step": 7905
+    },
+    {
+      "epoch": 0.6094665433240827,
+      "grad_norm": 3.537116050720215,
+      "learning_rate": 3.4943874818015115e-06,
+      "loss": 0.9367,
+      "step": 7906
+    },
+    {
+      "epoch": 0.6095436324390996,
+      "grad_norm": 3.9148404598236084,
+      "learning_rate": 3.493197027599249e-06,
+      "loss": 1.0676,
+      "step": 7907
+    },
+    {
+      "epoch": 0.6096207215541165,
+      "grad_norm": 3.719594717025757,
+      "learning_rate": 3.492006667338347e-06,
+      "loss": 0.8925,
+      "step": 7908
+    },
+    {
+      "epoch": 0.6096978106691335,
+      "grad_norm": 3.5892961025238037,
+      "learning_rate": 3.49081640109302e-06,
+      "loss": 0.9494,
+      "step": 7909
+    },
+    {
+      "epoch": 0.6097748997841504,
+      "grad_norm": 3.9413259029388428,
+      "learning_rate": 3.489626228937473e-06,
+      "loss": 0.8696,
+      "step": 7910
+    },
+    {
+      "epoch": 0.6098519888991675,
+      "grad_norm": 3.645066499710083,
+      "learning_rate": 3.4884361509459088e-06,
+      "loss": 0.9745,
+      "step": 7911
+    },
+    {
+      "epoch": 0.6099290780141844,
+      "grad_norm": 3.4567646980285645,
+      "learning_rate": 3.4872461671925227e-06,
+      "loss": 0.721,
+      "step": 7912
+    },
+    {
+      "epoch": 0.6100061671292013,
+      "grad_norm": 3.5239381790161133,
+      "learning_rate": 3.486056277751502e-06,
+      "loss": 0.9016,
+      "step": 7913
+    },
+    {
+      "epoch": 0.6100832562442183,
+      "grad_norm": 3.2868807315826416,
+      "learning_rate": 3.484866482697032e-06,
+      "loss": 0.888,
+      "step": 7914
+    },
+    {
+      "epoch": 0.6101603453592352,
+      "grad_norm": 3.2013051509857178,
+      "learning_rate": 3.4836767821032902e-06,
+      "loss": 0.8686,
+      "step": 7915
+    },
+    {
+      "epoch": 0.6102374344742523,
+      "grad_norm": 3.507582426071167,
+      "learning_rate": 3.4824871760444465e-06,
+      "loss": 1.027,
+      "step": 7916
+    },
+    {
+      "epoch": 0.6103145235892692,
+      "grad_norm": 3.7498128414154053,
+      "learning_rate": 3.48129766459467e-06,
+      "loss": 1.0387,
+      "step": 7917
+    },
+    {
+      "epoch": 0.6103916127042861,
+      "grad_norm": 3.7105958461761475,
+      "learning_rate": 3.480108247828117e-06,
+      "loss": 0.9652,
+      "step": 7918
+    },
+    {
+      "epoch": 0.6104687018193031,
+      "grad_norm": 3.8917605876922607,
+      "learning_rate": 3.478918925818943e-06,
+      "loss": 0.9192,
+      "step": 7919
+    },
+    {
+      "epoch": 0.61054579093432,
+      "grad_norm": 3.6685686111450195,
+      "learning_rate": 3.4777296986412972e-06,
+      "loss": 0.9294,
+      "step": 7920
+    },
+    {
+      "epoch": 0.6106228800493371,
+      "grad_norm": 3.4759345054626465,
+      "learning_rate": 3.4765405663693196e-06,
+      "loss": 0.8923,
+      "step": 7921
+    },
+    {
+      "epoch": 0.610699969164354,
+      "grad_norm": 3.711207151412964,
+      "learning_rate": 3.475351529077148e-06,
+      "loss": 0.9937,
+      "step": 7922
+    },
+    {
+      "epoch": 0.6107770582793709,
+      "grad_norm": 3.594850540161133,
+      "learning_rate": 3.474162586838913e-06,
+      "loss": 0.8816,
+      "step": 7923
+    },
+    {
+      "epoch": 0.6108541473943879,
+      "grad_norm": 3.8033182621002197,
+      "learning_rate": 3.4729737397287365e-06,
+      "loss": 0.9382,
+      "step": 7924
+    },
+    {
+      "epoch": 0.6109312365094048,
+      "grad_norm": 3.380199432373047,
+      "learning_rate": 3.4717849878207387e-06,
+      "loss": 0.8568,
+      "step": 7925
+    },
+    {
+      "epoch": 0.6110083256244219,
+      "grad_norm": 3.973435640335083,
+      "learning_rate": 3.4705963311890334e-06,
+      "loss": 0.8793,
+      "step": 7926
+    },
+    {
+      "epoch": 0.6110854147394388,
+      "grad_norm": 3.6724209785461426,
+      "learning_rate": 3.469407769907725e-06,
+      "loss": 0.8936,
+      "step": 7927
+    },
+    {
+      "epoch": 0.6111625038544557,
+      "grad_norm": 3.7241382598876953,
+      "learning_rate": 3.468219304050916e-06,
+      "loss": 0.9815,
+      "step": 7928
+    },
+    {
+      "epoch": 0.6112395929694727,
+      "grad_norm": 3.8722872734069824,
+      "learning_rate": 3.4670309336926988e-06,
+      "loss": 1.0309,
+      "step": 7929
+    },
+    {
+      "epoch": 0.6113166820844896,
+      "grad_norm": 3.580594301223755,
+      "learning_rate": 3.4658426589071637e-06,
+      "loss": 0.834,
+      "step": 7930
+    },
+    {
+      "epoch": 0.6113937711995067,
+      "grad_norm": 3.637439012527466,
+      "learning_rate": 3.4646544797683946e-06,
+      "loss": 0.9892,
+      "step": 7931
+    },
+    {
+      "epoch": 0.6114708603145236,
+      "grad_norm": 3.7209787368774414,
+      "learning_rate": 3.4634663963504654e-06,
+      "loss": 0.8565,
+      "step": 7932
+    },
+    {
+      "epoch": 0.6115479494295405,
+      "grad_norm": 3.676300525665283,
+      "learning_rate": 3.4622784087274487e-06,
+      "loss": 1.0424,
+      "step": 7933
+    },
+    {
+      "epoch": 0.6116250385445575,
+      "grad_norm": 4.965175151824951,
+      "learning_rate": 3.46109051697341e-06,
+      "loss": 0.9192,
+      "step": 7934
+    },
+    {
+      "epoch": 0.6117021276595744,
+      "grad_norm": 3.5655887126922607,
+      "learning_rate": 3.459902721162406e-06,
+      "loss": 0.958,
+      "step": 7935
+    },
+    {
+      "epoch": 0.6117792167745915,
+      "grad_norm": 3.9590542316436768,
+      "learning_rate": 3.4587150213684917e-06,
+      "loss": 0.9491,
+      "step": 7936
+    },
+    {
+      "epoch": 0.6118563058896084,
+      "grad_norm": 3.773998498916626,
+      "learning_rate": 3.4575274176657134e-06,
+      "loss": 0.9361,
+      "step": 7937
+    },
+    {
+      "epoch": 0.6119333950046254,
+      "grad_norm": 3.76638126373291,
+      "learning_rate": 3.4563399101281116e-06,
+      "loss": 0.9906,
+      "step": 7938
+    },
+    {
+      "epoch": 0.6120104841196423,
+      "grad_norm": 3.668475389480591,
+      "learning_rate": 3.4551524988297224e-06,
+      "loss": 0.782,
+      "step": 7939
+    },
+    {
+      "epoch": 0.6120875732346592,
+      "grad_norm": 3.6993038654327393,
+      "learning_rate": 3.453965183844573e-06,
+      "loss": 0.9687,
+      "step": 7940
+    },
+    {
+      "epoch": 0.6121646623496763,
+      "grad_norm": 4.174620628356934,
+      "learning_rate": 3.4527779652466876e-06,
+      "loss": 0.9987,
+      "step": 7941
+    },
+    {
+      "epoch": 0.6122417514646932,
+      "grad_norm": 3.559088945388794,
+      "learning_rate": 3.451590843110083e-06,
+      "loss": 1.0611,
+      "step": 7942
+    },
+    {
+      "epoch": 0.6123188405797102,
+      "grad_norm": 3.860713243484497,
+      "learning_rate": 3.4504038175087697e-06,
+      "loss": 0.9319,
+      "step": 7943
+    },
+    {
+      "epoch": 0.6123959296947271,
+      "grad_norm": 3.864760160446167,
+      "learning_rate": 3.449216888516751e-06,
+      "loss": 0.8688,
+      "step": 7944
+    },
+    {
+      "epoch": 0.612473018809744,
+      "grad_norm": 3.620938539505005,
+      "learning_rate": 3.448030056208029e-06,
+      "loss": 0.8719,
+      "step": 7945
+    },
+    {
+      "epoch": 0.6125501079247611,
+      "grad_norm": 3.6368861198425293,
+      "learning_rate": 3.4468433206565964e-06,
+      "loss": 0.9333,
+      "step": 7946
+    },
+    {
+      "epoch": 0.612627197039778,
+      "grad_norm": 3.713548421859741,
+      "learning_rate": 3.445656681936438e-06,
+      "loss": 1.0769,
+      "step": 7947
+    },
+    {
+      "epoch": 0.612704286154795,
+      "grad_norm": 3.9157612323760986,
+      "learning_rate": 3.4444701401215343e-06,
+      "loss": 0.8697,
+      "step": 7948
+    },
+    {
+      "epoch": 0.6127813752698119,
+      "grad_norm": 3.3834121227264404,
+      "learning_rate": 3.443283695285863e-06,
+      "loss": 0.7926,
+      "step": 7949
+    },
+    {
+      "epoch": 0.6128584643848288,
+      "grad_norm": 3.601344108581543,
+      "learning_rate": 3.4420973475033894e-06,
+      "loss": 0.922,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6129355534998459,
+      "grad_norm": 3.9263248443603516,
+      "learning_rate": 3.4409110968480773e-06,
+      "loss": 0.8751,
+      "step": 7951
+    },
+    {
+      "epoch": 0.6130126426148628,
+      "grad_norm": 3.377624988555908,
+      "learning_rate": 3.439724943393885e-06,
+      "loss": 0.8985,
+      "step": 7952
+    },
+    {
+      "epoch": 0.6130897317298798,
+      "grad_norm": 3.6270108222961426,
+      "learning_rate": 3.438538887214761e-06,
+      "loss": 0.965,
+      "step": 7953
+    },
+    {
+      "epoch": 0.6131668208448967,
+      "grad_norm": 3.6317126750946045,
+      "learning_rate": 3.4373529283846498e-06,
+      "loss": 0.8884,
+      "step": 7954
+    },
+    {
+      "epoch": 0.6132439099599136,
+      "grad_norm": 3.744412422180176,
+      "learning_rate": 3.4361670669774917e-06,
+      "loss": 0.9421,
+      "step": 7955
+    },
+    {
+      "epoch": 0.6133209990749307,
+      "grad_norm": 3.9327218532562256,
+      "learning_rate": 3.4349813030672165e-06,
+      "loss": 0.894,
+      "step": 7956
+    },
+    {
+      "epoch": 0.6133980881899476,
+      "grad_norm": 3.5057458877563477,
+      "learning_rate": 3.4337956367277524e-06,
+      "loss": 0.9483,
+      "step": 7957
+    },
+    {
+      "epoch": 0.6134751773049646,
+      "grad_norm": 3.859243869781494,
+      "learning_rate": 3.432610068033018e-06,
+      "loss": 1.0009,
+      "step": 7958
+    },
+    {
+      "epoch": 0.6135522664199815,
+      "grad_norm": 3.938803195953369,
+      "learning_rate": 3.4314245970569283e-06,
+      "loss": 0.8712,
+      "step": 7959
+    },
+    {
+      "epoch": 0.6136293555349984,
+      "grad_norm": 3.743238687515259,
+      "learning_rate": 3.4302392238733916e-06,
+      "loss": 0.9356,
+      "step": 7960
+    },
+    {
+      "epoch": 0.6137064446500154,
+      "grad_norm": 3.5383241176605225,
+      "learning_rate": 3.429053948556309e-06,
+      "loss": 1.0357,
+      "step": 7961
+    },
+    {
+      "epoch": 0.6137835337650324,
+      "grad_norm": 3.5702669620513916,
+      "learning_rate": 3.4278687711795754e-06,
+      "loss": 0.9484,
+      "step": 7962
+    },
+    {
+      "epoch": 0.6138606228800494,
+      "grad_norm": 3.585568904876709,
+      "learning_rate": 3.4266836918170844e-06,
+      "loss": 0.8646,
+      "step": 7963
+    },
+    {
+      "epoch": 0.6139377119950663,
+      "grad_norm": 3.544301748275757,
+      "learning_rate": 3.425498710542714e-06,
+      "loss": 0.9274,
+      "step": 7964
+    },
+    {
+      "epoch": 0.6140148011100832,
+      "grad_norm": 3.528738498687744,
+      "learning_rate": 3.424313827430347e-06,
+      "loss": 0.8791,
+      "step": 7965
+    },
+    {
+      "epoch": 0.6140918902251002,
+      "grad_norm": 3.8694934844970703,
+      "learning_rate": 3.42312904255385e-06,
+      "loss": 0.9614,
+      "step": 7966
+    },
+    {
+      "epoch": 0.6141689793401172,
+      "grad_norm": 3.590520143508911,
+      "learning_rate": 3.4219443559870906e-06,
+      "loss": 0.9202,
+      "step": 7967
+    },
+    {
+      "epoch": 0.6142460684551342,
+      "grad_norm": 3.699113607406616,
+      "learning_rate": 3.4207597678039293e-06,
+      "loss": 0.9423,
+      "step": 7968
+    },
+    {
+      "epoch": 0.6143231575701511,
+      "grad_norm": 3.463752508163452,
+      "learning_rate": 3.4195752780782156e-06,
+      "loss": 0.9054,
+      "step": 7969
+    },
+    {
+      "epoch": 0.614400246685168,
+      "grad_norm": 3.7661490440368652,
+      "learning_rate": 3.4183908868837988e-06,
+      "loss": 0.8484,
+      "step": 7970
+    },
+    {
+      "epoch": 0.614477335800185,
+      "grad_norm": 3.7998509407043457,
+      "learning_rate": 3.4172065942945194e-06,
+      "loss": 0.9527,
+      "step": 7971
+    },
+    {
+      "epoch": 0.614554424915202,
+      "grad_norm": 3.6133172512054443,
+      "learning_rate": 3.4160224003842103e-06,
+      "loss": 0.9766,
+      "step": 7972
+    },
+    {
+      "epoch": 0.614631514030219,
+      "grad_norm": 3.8746337890625,
+      "learning_rate": 3.414838305226701e-06,
+      "loss": 0.9468,
+      "step": 7973
+    },
+    {
+      "epoch": 0.6147086031452359,
+      "grad_norm": 3.8318541049957275,
+      "learning_rate": 3.4136543088958147e-06,
+      "loss": 0.9815,
+      "step": 7974
+    },
+    {
+      "epoch": 0.6147856922602528,
+      "grad_norm": 3.35244083404541,
+      "learning_rate": 3.4124704114653653e-06,
+      "loss": 0.83,
+      "step": 7975
+    },
+    {
+      "epoch": 0.6148627813752698,
+      "grad_norm": 3.994019031524658,
+      "learning_rate": 3.4112866130091648e-06,
+      "loss": 0.915,
+      "step": 7976
+    },
+    {
+      "epoch": 0.6149398704902868,
+      "grad_norm": 3.6324615478515625,
+      "learning_rate": 3.4101029136010144e-06,
+      "loss": 0.8744,
+      "step": 7977
+    },
+    {
+      "epoch": 0.6150169596053038,
+      "grad_norm": 3.4820597171783447,
+      "learning_rate": 3.4089193133147136e-06,
+      "loss": 0.8885,
+      "step": 7978
+    },
+    {
+      "epoch": 0.6150940487203207,
+      "grad_norm": 3.644369602203369,
+      "learning_rate": 3.4077358122240532e-06,
+      "loss": 0.996,
+      "step": 7979
+    },
+    {
+      "epoch": 0.6151711378353376,
+      "grad_norm": 3.64266300201416,
+      "learning_rate": 3.4065524104028185e-06,
+      "loss": 0.9692,
+      "step": 7980
+    },
+    {
+      "epoch": 0.6152482269503546,
+      "grad_norm": 3.786377191543579,
+      "learning_rate": 3.405369107924788e-06,
+      "loss": 0.9424,
+      "step": 7981
+    },
+    {
+      "epoch": 0.6153253160653716,
+      "grad_norm": 3.296741008758545,
+      "learning_rate": 3.404185904863736e-06,
+      "loss": 0.8991,
+      "step": 7982
+    },
+    {
+      "epoch": 0.6154024051803886,
+      "grad_norm": 3.8608269691467285,
+      "learning_rate": 3.4030028012934264e-06,
+      "loss": 0.9511,
+      "step": 7983
+    },
+    {
+      "epoch": 0.6154794942954055,
+      "grad_norm": 3.641268730163574,
+      "learning_rate": 3.401819797287621e-06,
+      "loss": 0.9326,
+      "step": 7984
+    },
+    {
+      "epoch": 0.6155565834104224,
+      "grad_norm": 3.7189748287200928,
+      "learning_rate": 3.400636892920076e-06,
+      "loss": 0.901,
+      "step": 7985
+    },
+    {
+      "epoch": 0.6156336725254394,
+      "grad_norm": 3.6073148250579834,
+      "learning_rate": 3.3994540882645353e-06,
+      "loss": 0.9221,
+      "step": 7986
+    },
+    {
+      "epoch": 0.6157107616404563,
+      "grad_norm": 3.328263282775879,
+      "learning_rate": 3.3982713833947447e-06,
+      "loss": 0.8374,
+      "step": 7987
+    },
+    {
+      "epoch": 0.6157878507554734,
+      "grad_norm": 3.7613234519958496,
+      "learning_rate": 3.397088778384434e-06,
+      "loss": 0.8323,
+      "step": 7988
+    },
+    {
+      "epoch": 0.6158649398704903,
+      "grad_norm": 3.8894762992858887,
+      "learning_rate": 3.39590627330734e-06,
+      "loss": 0.9505,
+      "step": 7989
+    },
+    {
+      "epoch": 0.6159420289855072,
+      "grad_norm": 3.8909053802490234,
+      "learning_rate": 3.3947238682371803e-06,
+      "loss": 0.9096,
+      "step": 7990
+    },
+    {
+      "epoch": 0.6160191181005242,
+      "grad_norm": 3.90527081489563,
+      "learning_rate": 3.3935415632476736e-06,
+      "loss": 0.9503,
+      "step": 7991
+    },
+    {
+      "epoch": 0.6160962072155411,
+      "grad_norm": 3.739124298095703,
+      "learning_rate": 3.392359358412532e-06,
+      "loss": 0.9485,
+      "step": 7992
+    },
+    {
+      "epoch": 0.6161732963305582,
+      "grad_norm": 3.6810646057128906,
+      "learning_rate": 3.3911772538054565e-06,
+      "loss": 0.9624,
+      "step": 7993
+    },
+    {
+      "epoch": 0.6162503854455751,
+      "grad_norm": 3.561396598815918,
+      "learning_rate": 3.3899952495001486e-06,
+      "loss": 0.9049,
+      "step": 7994
+    },
+    {
+      "epoch": 0.616327474560592,
+      "grad_norm": 4.142645835876465,
+      "learning_rate": 3.3888133455702964e-06,
+      "loss": 0.9669,
+      "step": 7995
+    },
+    {
+      "epoch": 0.616404563675609,
+      "grad_norm": 3.6202030181884766,
+      "learning_rate": 3.387631542089589e-06,
+      "loss": 0.9981,
+      "step": 7996
+    },
+    {
+      "epoch": 0.616481652790626,
+      "grad_norm": 3.5489907264709473,
+      "learning_rate": 3.3864498391317047e-06,
+      "loss": 0.9013,
+      "step": 7997
+    },
+    {
+      "epoch": 0.616558741905643,
+      "grad_norm": 3.5619547367095947,
+      "learning_rate": 3.385268236770315e-06,
+      "loss": 0.8911,
+      "step": 7998
+    },
+    {
+      "epoch": 0.6166358310206599,
+      "grad_norm": 3.5192930698394775,
+      "learning_rate": 3.3840867350790872e-06,
+      "loss": 0.9838,
+      "step": 7999
+    },
+    {
+      "epoch": 0.6167129201356768,
+      "grad_norm": 3.7306177616119385,
+      "learning_rate": 3.3829053341316843e-06,
+      "loss": 0.9779,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6167900092506938,
+      "grad_norm": 4.072969436645508,
+      "learning_rate": 3.3817240340017575e-06,
+      "loss": 0.9314,
+      "step": 8001
+    },
+    {
+      "epoch": 0.6168670983657107,
+      "grad_norm": 4.617973804473877,
+      "learning_rate": 3.3805428347629554e-06,
+      "loss": 0.9557,
+      "step": 8002
+    },
+    {
+      "epoch": 0.6169441874807278,
+      "grad_norm": 3.65148663520813,
+      "learning_rate": 3.379361736488922e-06,
+      "loss": 0.9209,
+      "step": 8003
+    },
+    {
+      "epoch": 0.6170212765957447,
+      "grad_norm": 3.826519727706909,
+      "learning_rate": 3.3781807392532893e-06,
+      "loss": 1.0016,
+      "step": 8004
+    },
+    {
+      "epoch": 0.6170983657107616,
+      "grad_norm": 3.791816473007202,
+      "learning_rate": 3.376999843129689e-06,
+      "loss": 0.8873,
+      "step": 8005
+    },
+    {
+      "epoch": 0.6171754548257786,
+      "grad_norm": 3.338088274002075,
+      "learning_rate": 3.375819048191742e-06,
+      "loss": 0.9135,
+      "step": 8006
+    },
+    {
+      "epoch": 0.6172525439407955,
+      "grad_norm": 3.48136043548584,
+      "learning_rate": 3.3746383545130646e-06,
+      "loss": 0.8923,
+      "step": 8007
+    },
+    {
+      "epoch": 0.6173296330558126,
+      "grad_norm": 3.3619744777679443,
+      "learning_rate": 3.37345776216727e-06,
+      "loss": 0.9164,
+      "step": 8008
+    },
+    {
+      "epoch": 0.6174067221708295,
+      "grad_norm": 3.594543218612671,
+      "learning_rate": 3.3722772712279583e-06,
+      "loss": 0.9498,
+      "step": 8009
+    },
+    {
+      "epoch": 0.6174838112858464,
+      "grad_norm": 3.30615496635437,
+      "learning_rate": 3.3710968817687284e-06,
+      "loss": 0.8533,
+      "step": 8010
+    },
+    {
+      "epoch": 0.6175609004008634,
+      "grad_norm": 3.645437479019165,
+      "learning_rate": 3.369916593863173e-06,
+      "loss": 0.9808,
+      "step": 8011
+    },
+    {
+      "epoch": 0.6176379895158803,
+      "grad_norm": 3.909817695617676,
+      "learning_rate": 3.3687364075848738e-06,
+      "loss": 1.0065,
+      "step": 8012
+    },
+    {
+      "epoch": 0.6177150786308974,
+      "grad_norm": 4.033278465270996,
+      "learning_rate": 3.3675563230074126e-06,
+      "loss": 0.9278,
+      "step": 8013
+    },
+    {
+      "epoch": 0.6177921677459143,
+      "grad_norm": 3.6392765045166016,
+      "learning_rate": 3.366376340204359e-06,
+      "loss": 0.8868,
+      "step": 8014
+    },
+    {
+      "epoch": 0.6178692568609312,
+      "grad_norm": 3.7995078563690186,
+      "learning_rate": 3.3651964592492805e-06,
+      "loss": 1.0029,
+      "step": 8015
+    },
+    {
+      "epoch": 0.6179463459759482,
+      "grad_norm": 4.048414707183838,
+      "learning_rate": 3.3640166802157357e-06,
+      "loss": 0.9189,
+      "step": 8016
+    },
+    {
+      "epoch": 0.6180234350909651,
+      "grad_norm": 3.517993927001953,
+      "learning_rate": 3.362837003177278e-06,
+      "loss": 0.9468,
+      "step": 8017
+    },
+    {
+      "epoch": 0.6181005242059822,
+      "grad_norm": 3.915440559387207,
+      "learning_rate": 3.3616574282074533e-06,
+      "loss": 0.9799,
+      "step": 8018
+    },
+    {
+      "epoch": 0.6181776133209991,
+      "grad_norm": 3.5532066822052,
+      "learning_rate": 3.360477955379804e-06,
+      "loss": 0.8911,
+      "step": 8019
+    },
+    {
+      "epoch": 0.618254702436016,
+      "grad_norm": 3.4006614685058594,
+      "learning_rate": 3.3592985847678617e-06,
+      "loss": 0.9145,
+      "step": 8020
+    },
+    {
+      "epoch": 0.618331791551033,
+      "grad_norm": 3.6829326152801514,
+      "learning_rate": 3.3581193164451555e-06,
+      "loss": 1.0141,
+      "step": 8021
+    },
+    {
+      "epoch": 0.6184088806660499,
+      "grad_norm": 3.5860280990600586,
+      "learning_rate": 3.3569401504852073e-06,
+      "loss": 0.8548,
+      "step": 8022
+    },
+    {
+      "epoch": 0.618485969781067,
+      "grad_norm": 3.718271255493164,
+      "learning_rate": 3.35576108696153e-06,
+      "loss": 0.9893,
+      "step": 8023
+    },
+    {
+      "epoch": 0.6185630588960839,
+      "grad_norm": 3.679704189300537,
+      "learning_rate": 3.354582125947634e-06,
+      "loss": 0.9285,
+      "step": 8024
+    },
+    {
+      "epoch": 0.6186401480111008,
+      "grad_norm": 3.5176236629486084,
+      "learning_rate": 3.3534032675170205e-06,
+      "loss": 0.9224,
+      "step": 8025
+    },
+    {
+      "epoch": 0.6187172371261178,
+      "grad_norm": 3.7090749740600586,
+      "learning_rate": 3.3522245117431845e-06,
+      "loss": 0.8834,
+      "step": 8026
+    },
+    {
+      "epoch": 0.6187943262411347,
+      "grad_norm": 3.7500243186950684,
+      "learning_rate": 3.3510458586996173e-06,
+      "loss": 1.0048,
+      "step": 8027
+    },
+    {
+      "epoch": 0.6188714153561518,
+      "grad_norm": 3.5407137870788574,
+      "learning_rate": 3.3498673084598e-06,
+      "loss": 1.0076,
+      "step": 8028
+    },
+    {
+      "epoch": 0.6189485044711687,
+      "grad_norm": 3.900679588317871,
+      "learning_rate": 3.348688861097209e-06,
+      "loss": 0.9704,
+      "step": 8029
+    },
+    {
+      "epoch": 0.6190255935861856,
+      "grad_norm": 3.705470323562622,
+      "learning_rate": 3.347510516685317e-06,
+      "loss": 0.9472,
+      "step": 8030
+    },
+    {
+      "epoch": 0.6191026827012026,
+      "grad_norm": 3.6836581230163574,
+      "learning_rate": 3.3463322752975846e-06,
+      "loss": 0.8369,
+      "step": 8031
+    },
+    {
+      "epoch": 0.6191797718162195,
+      "grad_norm": 3.5624961853027344,
+      "learning_rate": 3.34515413700747e-06,
+      "loss": 0.9787,
+      "step": 8032
+    },
+    {
+      "epoch": 0.6192568609312366,
+      "grad_norm": 3.8100168704986572,
+      "learning_rate": 3.3439761018884233e-06,
+      "loss": 1.0565,
+      "step": 8033
+    },
+    {
+      "epoch": 0.6193339500462535,
+      "grad_norm": 3.6078031063079834,
+      "learning_rate": 3.342798170013892e-06,
+      "loss": 0.8859,
+      "step": 8034
+    },
+    {
+      "epoch": 0.6194110391612704,
+      "grad_norm": 3.7805142402648926,
+      "learning_rate": 3.3416203414573113e-06,
+      "loss": 0.9629,
+      "step": 8035
+    },
+    {
+      "epoch": 0.6194881282762874,
+      "grad_norm": 3.473383903503418,
+      "learning_rate": 3.3404426162921123e-06,
+      "loss": 0.8921,
+      "step": 8036
+    },
+    {
+      "epoch": 0.6195652173913043,
+      "grad_norm": 3.560927629470825,
+      "learning_rate": 3.3392649945917235e-06,
+      "loss": 0.8671,
+      "step": 8037
+    },
+    {
+      "epoch": 0.6196423065063213,
+      "grad_norm": 3.80720591545105,
+      "learning_rate": 3.338087476429559e-06,
+      "loss": 0.9247,
+      "step": 8038
+    },
+    {
+      "epoch": 0.6197193956213383,
+      "grad_norm": 3.5851659774780273,
+      "learning_rate": 3.3369100618790336e-06,
+      "loss": 0.9808,
+      "step": 8039
+    },
+    {
+      "epoch": 0.6197964847363552,
+      "grad_norm": 3.476499319076538,
+      "learning_rate": 3.335732751013553e-06,
+      "loss": 0.8289,
+      "step": 8040
+    },
+    {
+      "epoch": 0.6198735738513722,
+      "grad_norm": 3.651381731033325,
+      "learning_rate": 3.334555543906516e-06,
+      "loss": 0.8949,
+      "step": 8041
+    },
+    {
+      "epoch": 0.6199506629663891,
+      "grad_norm": 3.6143252849578857,
+      "learning_rate": 3.333378440631315e-06,
+      "loss": 0.8938,
+      "step": 8042
+    },
+    {
+      "epoch": 0.6200277520814061,
+      "grad_norm": 3.81123685836792,
+      "learning_rate": 3.3322014412613364e-06,
+      "loss": 0.8147,
+      "step": 8043
+    },
+    {
+      "epoch": 0.6201048411964231,
+      "grad_norm": 3.5519094467163086,
+      "learning_rate": 3.3310245458699595e-06,
+      "loss": 0.8491,
+      "step": 8044
+    },
+    {
+      "epoch": 0.62018193031144,
+      "grad_norm": 3.531566619873047,
+      "learning_rate": 3.3298477545305595e-06,
+      "loss": 0.8893,
+      "step": 8045
+    },
+    {
+      "epoch": 0.620259019426457,
+      "grad_norm": 3.8642385005950928,
+      "learning_rate": 3.3286710673165012e-06,
+      "loss": 0.9373,
+      "step": 8046
+    },
+    {
+      "epoch": 0.6203361085414739,
+      "grad_norm": 4.416684627532959,
+      "learning_rate": 3.3274944843011463e-06,
+      "loss": 1.0649,
+      "step": 8047
+    },
+    {
+      "epoch": 0.620413197656491,
+      "grad_norm": 3.7867462635040283,
+      "learning_rate": 3.3263180055578493e-06,
+      "loss": 0.9491,
+      "step": 8048
+    },
+    {
+      "epoch": 0.6204902867715079,
+      "grad_norm": 3.4176316261291504,
+      "learning_rate": 3.325141631159954e-06,
+      "loss": 0.9123,
+      "step": 8049
+    },
+    {
+      "epoch": 0.6205673758865248,
+      "grad_norm": 3.9313580989837646,
+      "learning_rate": 3.3239653611808054e-06,
+      "loss": 0.938,
+      "step": 8050
+    },
+    {
+      "epoch": 0.6206444650015418,
+      "grad_norm": 3.718086004257202,
+      "learning_rate": 3.3227891956937366e-06,
+      "loss": 0.9519,
+      "step": 8051
+    },
+    {
+      "epoch": 0.6207215541165587,
+      "grad_norm": 3.9227771759033203,
+      "learning_rate": 3.321613134772073e-06,
+      "loss": 1.0329,
+      "step": 8052
+    },
+    {
+      "epoch": 0.6207986432315757,
+      "grad_norm": 3.459362506866455,
+      "learning_rate": 3.32043717848914e-06,
+      "loss": 0.92,
+      "step": 8053
+    },
+    {
+      "epoch": 0.6208757323465927,
+      "grad_norm": 3.933849811553955,
+      "learning_rate": 3.3192613269182483e-06,
+      "loss": 0.9814,
+      "step": 8054
+    },
+    {
+      "epoch": 0.6209528214616096,
+      "grad_norm": 4.124648094177246,
+      "learning_rate": 3.3180855801327083e-06,
+      "loss": 0.979,
+      "step": 8055
+    },
+    {
+      "epoch": 0.6210299105766266,
+      "grad_norm": 3.7071919441223145,
+      "learning_rate": 3.3169099382058233e-06,
+      "loss": 0.9179,
+      "step": 8056
+    },
+    {
+      "epoch": 0.6211069996916435,
+      "grad_norm": 3.726698398590088,
+      "learning_rate": 3.315734401210885e-06,
+      "loss": 0.9425,
+      "step": 8057
+    },
+    {
+      "epoch": 0.6211840888066605,
+      "grad_norm": 3.652276039123535,
+      "learning_rate": 3.3145589692211837e-06,
+      "loss": 0.9226,
+      "step": 8058
+    },
+    {
+      "epoch": 0.6212611779216775,
+      "grad_norm": 3.5816662311553955,
+      "learning_rate": 3.313383642310003e-06,
+      "loss": 0.7326,
+      "step": 8059
+    },
+    {
+      "epoch": 0.6213382670366944,
+      "grad_norm": 3.948348045349121,
+      "learning_rate": 3.3122084205506153e-06,
+      "loss": 0.9797,
+      "step": 8060
+    },
+    {
+      "epoch": 0.6214153561517114,
+      "grad_norm": 3.8742499351501465,
+      "learning_rate": 3.3110333040162934e-06,
+      "loss": 1.0117,
+      "step": 8061
+    },
+    {
+      "epoch": 0.6214924452667283,
+      "grad_norm": 3.3892946243286133,
+      "learning_rate": 3.309858292780296e-06,
+      "loss": 0.8751,
+      "step": 8062
+    },
+    {
+      "epoch": 0.6215695343817453,
+      "grad_norm": 3.410844087600708,
+      "learning_rate": 3.308683386915881e-06,
+      "loss": 0.8663,
+      "step": 8063
+    },
+    {
+      "epoch": 0.6216466234967623,
+      "grad_norm": 3.326235055923462,
+      "learning_rate": 3.3075085864962975e-06,
+      "loss": 0.8682,
+      "step": 8064
+    },
+    {
+      "epoch": 0.6217237126117792,
+      "grad_norm": 4.0448079109191895,
+      "learning_rate": 3.306333891594787e-06,
+      "loss": 0.8579,
+      "step": 8065
+    },
+    {
+      "epoch": 0.6218008017267962,
+      "grad_norm": 3.6627490520477295,
+      "learning_rate": 3.3051593022845867e-06,
+      "loss": 0.9246,
+      "step": 8066
+    },
+    {
+      "epoch": 0.6218778908418131,
+      "grad_norm": 3.7059576511383057,
+      "learning_rate": 3.3039848186389278e-06,
+      "loss": 0.7995,
+      "step": 8067
+    },
+    {
+      "epoch": 0.6219549799568301,
+      "grad_norm": 3.518906831741333,
+      "learning_rate": 3.3028104407310302e-06,
+      "loss": 0.9744,
+      "step": 8068
+    },
+    {
+      "epoch": 0.622032069071847,
+      "grad_norm": 3.5456724166870117,
+      "learning_rate": 3.3016361686341115e-06,
+      "loss": 0.9983,
+      "step": 8069
+    },
+    {
+      "epoch": 0.622109158186864,
+      "grad_norm": 4.1023101806640625,
+      "learning_rate": 3.3004620024213824e-06,
+      "loss": 0.9888,
+      "step": 8070
+    },
+    {
+      "epoch": 0.622186247301881,
+      "grad_norm": 3.6297264099121094,
+      "learning_rate": 3.2992879421660447e-06,
+      "loss": 0.9784,
+      "step": 8071
+    },
+    {
+      "epoch": 0.6222633364168979,
+      "grad_norm": 3.9697625637054443,
+      "learning_rate": 3.2981139879412965e-06,
+      "loss": 1.038,
+      "step": 8072
+    },
+    {
+      "epoch": 0.6223404255319149,
+      "grad_norm": 3.5402143001556396,
+      "learning_rate": 3.296940139820326e-06,
+      "loss": 0.8862,
+      "step": 8073
+    },
+    {
+      "epoch": 0.6224175146469318,
+      "grad_norm": 3.9144885540008545,
+      "learning_rate": 3.2957663978763172e-06,
+      "loss": 0.9882,
+      "step": 8074
+    },
+    {
+      "epoch": 0.6224946037619488,
+      "grad_norm": 3.5432751178741455,
+      "learning_rate": 3.294592762182448e-06,
+      "loss": 0.9141,
+      "step": 8075
+    },
+    {
+      "epoch": 0.6225716928769658,
+      "grad_norm": 3.5189805030822754,
+      "learning_rate": 3.2934192328118866e-06,
+      "loss": 0.7741,
+      "step": 8076
+    },
+    {
+      "epoch": 0.6226487819919827,
+      "grad_norm": 3.9740869998931885,
+      "learning_rate": 3.292245809837796e-06,
+      "loss": 0.9351,
+      "step": 8077
+    },
+    {
+      "epoch": 0.6227258711069997,
+      "grad_norm": 3.5222132205963135,
+      "learning_rate": 3.2910724933333365e-06,
+      "loss": 0.9749,
+      "step": 8078
+    },
+    {
+      "epoch": 0.6228029602220166,
+      "grad_norm": 4.026155948638916,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 0.9626,
+      "step": 8079
+    },
+    {
+      "epoch": 0.6228800493370336,
+      "grad_norm": 3.8821022510528564,
+      "learning_rate": 3.2887261800258985e-06,
+      "loss": 1.0371,
+      "step": 8080
+    },
+    {
+      "epoch": 0.6229571384520506,
+      "grad_norm": 3.676262855529785,
+      "learning_rate": 3.287553183369201e-06,
+      "loss": 1.0328,
+      "step": 8081
+    },
+    {
+      "epoch": 0.6230342275670675,
+      "grad_norm": 3.552816152572632,
+      "learning_rate": 3.2863802934746938e-06,
+      "loss": 0.8604,
+      "step": 8082
+    },
+    {
+      "epoch": 0.6231113166820845,
+      "grad_norm": 4.076422214508057,
+      "learning_rate": 3.285207510415499e-06,
+      "loss": 0.8237,
+      "step": 8083
+    },
+    {
+      "epoch": 0.6231884057971014,
+      "grad_norm": 3.790148973464966,
+      "learning_rate": 3.2840348342647357e-06,
+      "loss": 0.8949,
+      "step": 8084
+    },
+    {
+      "epoch": 0.6232654949121184,
+      "grad_norm": 3.885585308074951,
+      "learning_rate": 3.2828622650955133e-06,
+      "loss": 0.9832,
+      "step": 8085
+    },
+    {
+      "epoch": 0.6233425840271354,
+      "grad_norm": 4.014956474304199,
+      "learning_rate": 3.281689802980934e-06,
+      "loss": 0.9888,
+      "step": 8086
+    },
+    {
+      "epoch": 0.6234196731421523,
+      "grad_norm": 3.65321946144104,
+      "learning_rate": 3.280517447994097e-06,
+      "loss": 0.9624,
+      "step": 8087
+    },
+    {
+      "epoch": 0.6234967622571693,
+      "grad_norm": 3.9095733165740967,
+      "learning_rate": 3.279345200208093e-06,
+      "loss": 0.8794,
+      "step": 8088
+    },
+    {
+      "epoch": 0.6235738513721862,
+      "grad_norm": 3.463289737701416,
+      "learning_rate": 3.2781730596960027e-06,
+      "loss": 0.9902,
+      "step": 8089
+    },
+    {
+      "epoch": 0.6236509404872032,
+      "grad_norm": 4.12209939956665,
+      "learning_rate": 3.277001026530906e-06,
+      "loss": 0.8898,
+      "step": 8090
+    },
+    {
+      "epoch": 0.6237280296022202,
+      "grad_norm": 3.618441104888916,
+      "learning_rate": 3.27582910078587e-06,
+      "loss": 0.9217,
+      "step": 8091
+    },
+    {
+      "epoch": 0.6238051187172371,
+      "grad_norm": 3.7258079051971436,
+      "learning_rate": 3.2746572825339607e-06,
+      "loss": 0.9826,
+      "step": 8092
+    },
+    {
+      "epoch": 0.6238822078322541,
+      "grad_norm": 3.7279441356658936,
+      "learning_rate": 3.273485571848234e-06,
+      "loss": 0.9179,
+      "step": 8093
+    },
+    {
+      "epoch": 0.623959296947271,
+      "grad_norm": 3.8348352909088135,
+      "learning_rate": 3.27231396880174e-06,
+      "loss": 1.006,
+      "step": 8094
+    },
+    {
+      "epoch": 0.624036386062288,
+      "grad_norm": 3.53607177734375,
+      "learning_rate": 3.2711424734675212e-06,
+      "loss": 0.8691,
+      "step": 8095
+    },
+    {
+      "epoch": 0.624113475177305,
+      "grad_norm": 3.5375266075134277,
+      "learning_rate": 3.2699710859186175e-06,
+      "loss": 0.9688,
+      "step": 8096
+    },
+    {
+      "epoch": 0.6241905642923219,
+      "grad_norm": 4.027289390563965,
+      "learning_rate": 3.2687998062280546e-06,
+      "loss": 1.0257,
+      "step": 8097
+    },
+    {
+      "epoch": 0.6242676534073389,
+      "grad_norm": 3.652118682861328,
+      "learning_rate": 3.2676286344688575e-06,
+      "loss": 0.923,
+      "step": 8098
+    },
+    {
+      "epoch": 0.6243447425223558,
+      "grad_norm": 4.014166355133057,
+      "learning_rate": 3.2664575707140446e-06,
+      "loss": 0.951,
+      "step": 8099
+    },
+    {
+      "epoch": 0.6244218316373727,
+      "grad_norm": 3.562201738357544,
+      "learning_rate": 3.265286615036622e-06,
+      "loss": 0.9807,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6244989207523898,
+      "grad_norm": 3.267332077026367,
+      "learning_rate": 3.2641157675095965e-06,
+      "loss": 0.9209,
+      "step": 8101
+    },
+    {
+      "epoch": 0.6245760098674067,
+      "grad_norm": 3.832848310470581,
+      "learning_rate": 3.2629450282059603e-06,
+      "loss": 0.8897,
+      "step": 8102
+    },
+    {
+      "epoch": 0.6246530989824237,
+      "grad_norm": 3.6828267574310303,
+      "learning_rate": 3.2617743971987063e-06,
+      "loss": 0.8706,
+      "step": 8103
+    },
+    {
+      "epoch": 0.6247301880974406,
+      "grad_norm": 3.771867275238037,
+      "learning_rate": 3.260603874560817e-06,
+      "loss": 0.9786,
+      "step": 8104
+    },
+    {
+      "epoch": 0.6248072772124575,
+      "grad_norm": 3.621521234512329,
+      "learning_rate": 3.2594334603652668e-06,
+      "loss": 0.9209,
+      "step": 8105
+    },
+    {
+      "epoch": 0.6248843663274746,
+      "grad_norm": 3.851837635040283,
+      "learning_rate": 3.2582631546850254e-06,
+      "loss": 0.893,
+      "step": 8106
+    },
+    {
+      "epoch": 0.6249614554424915,
+      "grad_norm": 4.190458297729492,
+      "learning_rate": 3.2570929575930563e-06,
+      "loss": 0.9945,
+      "step": 8107
+    },
+    {
+      "epoch": 0.6250385445575085,
+      "grad_norm": 3.653280019760132,
+      "learning_rate": 3.2559228691623134e-06,
+      "loss": 0.9732,
+      "step": 8108
+    },
+    {
+      "epoch": 0.6251156336725254,
+      "grad_norm": 3.965890645980835,
+      "learning_rate": 3.254752889465749e-06,
+      "loss": 0.9156,
+      "step": 8109
+    },
+    {
+      "epoch": 0.6251927227875425,
+      "grad_norm": 3.464805841445923,
+      "learning_rate": 3.2535830185763006e-06,
+      "loss": 0.7856,
+      "step": 8110
+    },
+    {
+      "epoch": 0.6252698119025594,
+      "grad_norm": 3.8827240467071533,
+      "learning_rate": 3.252413256566907e-06,
+      "loss": 0.9146,
+      "step": 8111
+    },
+    {
+      "epoch": 0.6253469010175763,
+      "grad_norm": 3.4386796951293945,
+      "learning_rate": 3.2512436035104968e-06,
+      "loss": 0.9379,
+      "step": 8112
+    },
+    {
+      "epoch": 0.6254239901325933,
+      "grad_norm": 3.7330002784729004,
+      "learning_rate": 3.2500740594799895e-06,
+      "loss": 0.942,
+      "step": 8113
+    },
+    {
+      "epoch": 0.6255010792476102,
+      "grad_norm": 3.6872072219848633,
+      "learning_rate": 3.2489046245483016e-06,
+      "loss": 0.8975,
+      "step": 8114
+    },
+    {
+      "epoch": 0.6255781683626273,
+      "grad_norm": 3.910029888153076,
+      "learning_rate": 3.2477352987883426e-06,
+      "loss": 1.0826,
+      "step": 8115
+    },
+    {
+      "epoch": 0.6256552574776442,
+      "grad_norm": 4.114757537841797,
+      "learning_rate": 3.246566082273011e-06,
+      "loss": 0.9402,
+      "step": 8116
+    },
+    {
+      "epoch": 0.6257323465926611,
+      "grad_norm": 3.61517071723938,
+      "learning_rate": 3.245396975075203e-06,
+      "loss": 0.9002,
+      "step": 8117
+    },
+    {
+      "epoch": 0.6258094357076781,
+      "grad_norm": 3.6216084957122803,
+      "learning_rate": 3.2442279772678076e-06,
+      "loss": 0.9916,
+      "step": 8118
+    },
+    {
+      "epoch": 0.625886524822695,
+      "grad_norm": 3.727811336517334,
+      "learning_rate": 3.2430590889237027e-06,
+      "loss": 1.0077,
+      "step": 8119
+    },
+    {
+      "epoch": 0.625963613937712,
+      "grad_norm": 3.697965145111084,
+      "learning_rate": 3.241890310115766e-06,
+      "loss": 1.0235,
+      "step": 8120
+    },
+    {
+      "epoch": 0.626040703052729,
+      "grad_norm": 3.661677122116089,
+      "learning_rate": 3.2407216409168597e-06,
+      "loss": 0.8098,
+      "step": 8121
+    },
+    {
+      "epoch": 0.6261177921677459,
+      "grad_norm": 3.530116558074951,
+      "learning_rate": 3.2395530813998506e-06,
+      "loss": 0.904,
+      "step": 8122
+    },
+    {
+      "epoch": 0.6261948812827629,
+      "grad_norm": 3.739652633666992,
+      "learning_rate": 3.2383846316375876e-06,
+      "loss": 0.9422,
+      "step": 8123
+    },
+    {
+      "epoch": 0.6262719703977798,
+      "grad_norm": 4.194034576416016,
+      "learning_rate": 3.2372162917029205e-06,
+      "loss": 1.0163,
+      "step": 8124
+    },
+    {
+      "epoch": 0.6263490595127968,
+      "grad_norm": 3.3283534049987793,
+      "learning_rate": 3.236048061668688e-06,
+      "loss": 0.7624,
+      "step": 8125
+    },
+    {
+      "epoch": 0.6264261486278138,
+      "grad_norm": 3.7306947708129883,
+      "learning_rate": 3.2348799416077225e-06,
+      "loss": 0.9407,
+      "step": 8126
+    },
+    {
+      "epoch": 0.6265032377428307,
+      "grad_norm": 3.8820812702178955,
+      "learning_rate": 3.2337119315928513e-06,
+      "loss": 0.9529,
+      "step": 8127
+    },
+    {
+      "epoch": 0.6265803268578477,
+      "grad_norm": 3.882148027420044,
+      "learning_rate": 3.2325440316968927e-06,
+      "loss": 0.93,
+      "step": 8128
+    },
+    {
+      "epoch": 0.6266574159728646,
+      "grad_norm": 3.439239025115967,
+      "learning_rate": 3.2313762419926597e-06,
+      "loss": 0.9563,
+      "step": 8129
+    },
+    {
+      "epoch": 0.6267345050878816,
+      "grad_norm": 3.5904295444488525,
+      "learning_rate": 3.2302085625529596e-06,
+      "loss": 1.0631,
+      "step": 8130
+    },
+    {
+      "epoch": 0.6268115942028986,
+      "grad_norm": 3.6750895977020264,
+      "learning_rate": 3.2290409934505884e-06,
+      "loss": 0.9154,
+      "step": 8131
+    },
+    {
+      "epoch": 0.6268886833179155,
+      "grad_norm": 3.750882625579834,
+      "learning_rate": 3.227873534758339e-06,
+      "loss": 0.8514,
+      "step": 8132
+    },
+    {
+      "epoch": 0.6269657724329325,
+      "grad_norm": 3.87109637260437,
+      "learning_rate": 3.226706186548998e-06,
+      "loss": 0.9765,
+      "step": 8133
+    },
+    {
+      "epoch": 0.6270428615479494,
+      "grad_norm": 4.06139612197876,
+      "learning_rate": 3.22553894889534e-06,
+      "loss": 0.9913,
+      "step": 8134
+    },
+    {
+      "epoch": 0.6271199506629664,
+      "grad_norm": 3.7100861072540283,
+      "learning_rate": 3.2243718218701393e-06,
+      "loss": 1.0381,
+      "step": 8135
+    },
+    {
+      "epoch": 0.6271970397779834,
+      "grad_norm": 3.1716091632843018,
+      "learning_rate": 3.2232048055461608e-06,
+      "loss": 0.8383,
+      "step": 8136
+    },
+    {
+      "epoch": 0.6272741288930003,
+      "grad_norm": 3.715208053588867,
+      "learning_rate": 3.2220378999961588e-06,
+      "loss": 0.943,
+      "step": 8137
+    },
+    {
+      "epoch": 0.6273512180080173,
+      "grad_norm": 3.427281618118286,
+      "learning_rate": 3.2208711052928867e-06,
+      "loss": 0.9138,
+      "step": 8138
+    },
+    {
+      "epoch": 0.6274283071230342,
+      "grad_norm": 3.5145509243011475,
+      "learning_rate": 3.219704421509085e-06,
+      "loss": 0.8756,
+      "step": 8139
+    },
+    {
+      "epoch": 0.6275053962380512,
+      "grad_norm": 4.032465934753418,
+      "learning_rate": 3.218537848717493e-06,
+      "loss": 0.9441,
+      "step": 8140
+    },
+    {
+      "epoch": 0.6275824853530682,
+      "grad_norm": 3.699963331222534,
+      "learning_rate": 3.2173713869908406e-06,
+      "loss": 0.9135,
+      "step": 8141
+    },
+    {
+      "epoch": 0.6276595744680851,
+      "grad_norm": 3.3463034629821777,
+      "learning_rate": 3.2162050364018484e-06,
+      "loss": 0.8761,
+      "step": 8142
+    },
+    {
+      "epoch": 0.6277366635831021,
+      "grad_norm": 3.821621894836426,
+      "learning_rate": 3.215038797023234e-06,
+      "loss": 0.9123,
+      "step": 8143
+    },
+    {
+      "epoch": 0.627813752698119,
+      "grad_norm": 3.723275899887085,
+      "learning_rate": 3.2138726689277067e-06,
+      "loss": 0.9333,
+      "step": 8144
+    },
+    {
+      "epoch": 0.627890841813136,
+      "grad_norm": 3.7927725315093994,
+      "learning_rate": 3.2127066521879674e-06,
+      "loss": 0.9102,
+      "step": 8145
+    },
+    {
+      "epoch": 0.627967930928153,
+      "grad_norm": 3.668217182159424,
+      "learning_rate": 3.211540746876711e-06,
+      "loss": 0.9221,
+      "step": 8146
+    },
+    {
+      "epoch": 0.6280450200431699,
+      "grad_norm": 3.8914999961853027,
+      "learning_rate": 3.2103749530666283e-06,
+      "loss": 0.9604,
+      "step": 8147
+    },
+    {
+      "epoch": 0.6281221091581869,
+      "grad_norm": 3.947474718093872,
+      "learning_rate": 3.2092092708303973e-06,
+      "loss": 0.9765,
+      "step": 8148
+    },
+    {
+      "epoch": 0.6281991982732038,
+      "grad_norm": 3.575840711593628,
+      "learning_rate": 3.2080437002406943e-06,
+      "loss": 0.9536,
+      "step": 8149
+    },
+    {
+      "epoch": 0.6282762873882208,
+      "grad_norm": 3.5878658294677734,
+      "learning_rate": 3.206878241370185e-06,
+      "loss": 0.9539,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6283533765032377,
+      "grad_norm": 4.496673107147217,
+      "learning_rate": 3.2057128942915306e-06,
+      "loss": 0.9298,
+      "step": 8151
+    },
+    {
+      "epoch": 0.6284304656182547,
+      "grad_norm": 3.7084543704986572,
+      "learning_rate": 3.204547659077385e-06,
+      "loss": 0.9288,
+      "step": 8152
+    },
+    {
+      "epoch": 0.6285075547332717,
+      "grad_norm": 3.8590118885040283,
+      "learning_rate": 3.2033825358003936e-06,
+      "loss": 0.9057,
+      "step": 8153
+    },
+    {
+      "epoch": 0.6285846438482886,
+      "grad_norm": 3.7039713859558105,
+      "learning_rate": 3.2022175245331954e-06,
+      "loss": 0.9582,
+      "step": 8154
+    },
+    {
+      "epoch": 0.6286617329633056,
+      "grad_norm": 3.4202165603637695,
+      "learning_rate": 3.2010526253484246e-06,
+      "loss": 0.8993,
+      "step": 8155
+    },
+    {
+      "epoch": 0.6287388220783225,
+      "grad_norm": 3.575178861618042,
+      "learning_rate": 3.199887838318705e-06,
+      "loss": 0.8471,
+      "step": 8156
+    },
+    {
+      "epoch": 0.6288159111933395,
+      "grad_norm": 3.737819194793701,
+      "learning_rate": 3.1987231635166565e-06,
+      "loss": 0.8481,
+      "step": 8157
+    },
+    {
+      "epoch": 0.6288930003083565,
+      "grad_norm": 3.638150930404663,
+      "learning_rate": 3.197558601014889e-06,
+      "loss": 0.9176,
+      "step": 8158
+    },
+    {
+      "epoch": 0.6289700894233734,
+      "grad_norm": 3.68680477142334,
+      "learning_rate": 3.1963941508860076e-06,
+      "loss": 0.9781,
+      "step": 8159
+    },
+    {
+      "epoch": 0.6290471785383904,
+      "grad_norm": 3.491912841796875,
+      "learning_rate": 3.1952298132026107e-06,
+      "loss": 0.9208,
+      "step": 8160
+    },
+    {
+      "epoch": 0.6291242676534073,
+      "grad_norm": 3.932940721511841,
+      "learning_rate": 3.194065588037286e-06,
+      "loss": 0.9005,
+      "step": 8161
+    },
+    {
+      "epoch": 0.6292013567684243,
+      "grad_norm": 3.7449350357055664,
+      "learning_rate": 3.1929014754626197e-06,
+      "loss": 1.0379,
+      "step": 8162
+    },
+    {
+      "epoch": 0.6292784458834413,
+      "grad_norm": 3.840050220489502,
+      "learning_rate": 3.191737475551188e-06,
+      "loss": 0.9366,
+      "step": 8163
+    },
+    {
+      "epoch": 0.6293555349984582,
+      "grad_norm": 3.678901195526123,
+      "learning_rate": 3.1905735883755582e-06,
+      "loss": 0.8937,
+      "step": 8164
+    },
+    {
+      "epoch": 0.6294326241134752,
+      "grad_norm": 4.070024013519287,
+      "learning_rate": 3.1894098140082943e-06,
+      "loss": 1.0219,
+      "step": 8165
+    },
+    {
+      "epoch": 0.6295097132284921,
+      "grad_norm": 3.705054759979248,
+      "learning_rate": 3.18824615252195e-06,
+      "loss": 1.0407,
+      "step": 8166
+    },
+    {
+      "epoch": 0.629586802343509,
+      "grad_norm": 3.3765339851379395,
+      "learning_rate": 3.1870826039890766e-06,
+      "loss": 0.7974,
+      "step": 8167
+    },
+    {
+      "epoch": 0.6296638914585261,
+      "grad_norm": 3.9743378162384033,
+      "learning_rate": 3.1859191684822132e-06,
+      "loss": 0.9537,
+      "step": 8168
+    },
+    {
+      "epoch": 0.629740980573543,
+      "grad_norm": 3.8855860233306885,
+      "learning_rate": 3.1847558460738936e-06,
+      "loss": 0.9006,
+      "step": 8169
+    },
+    {
+      "epoch": 0.62981806968856,
+      "grad_norm": 3.418191909790039,
+      "learning_rate": 3.1835926368366465e-06,
+      "loss": 0.8984,
+      "step": 8170
+    },
+    {
+      "epoch": 0.6298951588035769,
+      "grad_norm": 3.7673134803771973,
+      "learning_rate": 3.1824295408429907e-06,
+      "loss": 0.9385,
+      "step": 8171
+    },
+    {
+      "epoch": 0.6299722479185939,
+      "grad_norm": 3.6283762454986572,
+      "learning_rate": 3.181266558165439e-06,
+      "loss": 0.94,
+      "step": 8172
+    },
+    {
+      "epoch": 0.6300493370336109,
+      "grad_norm": 3.8887081146240234,
+      "learning_rate": 3.180103688876499e-06,
+      "loss": 1.0551,
+      "step": 8173
+    },
+    {
+      "epoch": 0.6301264261486278,
+      "grad_norm": 4.249144554138184,
+      "learning_rate": 3.1789409330486683e-06,
+      "loss": 1.1189,
+      "step": 8174
+    },
+    {
+      "epoch": 0.6302035152636448,
+      "grad_norm": 3.9418768882751465,
+      "learning_rate": 3.1777782907544392e-06,
+      "loss": 0.9109,
+      "step": 8175
+    },
+    {
+      "epoch": 0.6302806043786617,
+      "grad_norm": 3.891338348388672,
+      "learning_rate": 3.176615762066295e-06,
+      "loss": 0.9539,
+      "step": 8176
+    },
+    {
+      "epoch": 0.6303576934936787,
+      "grad_norm": 3.6028475761413574,
+      "learning_rate": 3.175453347056715e-06,
+      "loss": 0.8756,
+      "step": 8177
+    },
+    {
+      "epoch": 0.6304347826086957,
+      "grad_norm": 3.6023495197296143,
+      "learning_rate": 3.17429104579817e-06,
+      "loss": 0.9605,
+      "step": 8178
+    },
+    {
+      "epoch": 0.6305118717237126,
+      "grad_norm": 4.147401332855225,
+      "learning_rate": 3.1731288583631214e-06,
+      "loss": 0.9424,
+      "step": 8179
+    },
+    {
+      "epoch": 0.6305889608387296,
+      "grad_norm": 3.7113969326019287,
+      "learning_rate": 3.1719667848240276e-06,
+      "loss": 0.9408,
+      "step": 8180
+    },
+    {
+      "epoch": 0.6306660499537465,
+      "grad_norm": 3.5659947395324707,
+      "learning_rate": 3.1708048252533376e-06,
+      "loss": 0.9135,
+      "step": 8181
+    },
+    {
+      "epoch": 0.6307431390687634,
+      "grad_norm": 3.6994526386260986,
+      "learning_rate": 3.1696429797234918e-06,
+      "loss": 0.8658,
+      "step": 8182
+    },
+    {
+      "epoch": 0.6308202281837805,
+      "grad_norm": 3.7129364013671875,
+      "learning_rate": 3.168481248306927e-06,
+      "loss": 0.9827,
+      "step": 8183
+    },
+    {
+      "epoch": 0.6308973172987974,
+      "grad_norm": 3.838097095489502,
+      "learning_rate": 3.1673196310760723e-06,
+      "loss": 0.9963,
+      "step": 8184
+    },
+    {
+      "epoch": 0.6309744064138144,
+      "grad_norm": 4.0744242668151855,
+      "learning_rate": 3.166158128103345e-06,
+      "loss": 0.9455,
+      "step": 8185
+    },
+    {
+      "epoch": 0.6310514955288313,
+      "grad_norm": 3.745927572250366,
+      "learning_rate": 3.164996739461162e-06,
+      "loss": 0.8731,
+      "step": 8186
+    },
+    {
+      "epoch": 0.6311285846438482,
+      "grad_norm": 3.8844759464263916,
+      "learning_rate": 3.163835465221927e-06,
+      "loss": 0.8394,
+      "step": 8187
+    },
+    {
+      "epoch": 0.6312056737588653,
+      "grad_norm": 3.549527406692505,
+      "learning_rate": 3.162674305458042e-06,
+      "loss": 0.9107,
+      "step": 8188
+    },
+    {
+      "epoch": 0.6312827628738822,
+      "grad_norm": 3.6786959171295166,
+      "learning_rate": 3.1615132602418986e-06,
+      "loss": 0.928,
+      "step": 8189
+    },
+    {
+      "epoch": 0.6313598519888992,
+      "grad_norm": 3.693917751312256,
+      "learning_rate": 3.160352329645881e-06,
+      "loss": 1.0224,
+      "step": 8190
+    },
+    {
+      "epoch": 0.6314369411039161,
+      "grad_norm": 3.7990047931671143,
+      "learning_rate": 3.159191513742368e-06,
+      "loss": 0.8727,
+      "step": 8191
+    },
+    {
+      "epoch": 0.631514030218933,
+      "grad_norm": 3.8932435512542725,
+      "learning_rate": 3.158030812603731e-06,
+      "loss": 0.9425,
+      "step": 8192
+    },
+    {
+      "epoch": 0.6315911193339501,
+      "grad_norm": 3.7320001125335693,
+      "learning_rate": 3.156870226302332e-06,
+      "loss": 0.804,
+      "step": 8193
+    },
+    {
+      "epoch": 0.631668208448967,
+      "grad_norm": 3.685913562774658,
+      "learning_rate": 3.155709754910529e-06,
+      "loss": 0.9927,
+      "step": 8194
+    },
+    {
+      "epoch": 0.631745297563984,
+      "grad_norm": 4.240044593811035,
+      "learning_rate": 3.1545493985006713e-06,
+      "loss": 1.0915,
+      "step": 8195
+    },
+    {
+      "epoch": 0.6318223866790009,
+      "grad_norm": 4.133537292480469,
+      "learning_rate": 3.1533891571451002e-06,
+      "loss": 0.9907,
+      "step": 8196
+    },
+    {
+      "epoch": 0.6318994757940178,
+      "grad_norm": 3.4915730953216553,
+      "learning_rate": 3.152229030916152e-06,
+      "loss": 0.8439,
+      "step": 8197
+    },
+    {
+      "epoch": 0.6319765649090349,
+      "grad_norm": 3.9785425662994385,
+      "learning_rate": 3.1510690198861533e-06,
+      "loss": 0.9674,
+      "step": 8198
+    },
+    {
+      "epoch": 0.6320536540240518,
+      "grad_norm": 3.388223886489868,
+      "learning_rate": 3.149909124127425e-06,
+      "loss": 0.8956,
+      "step": 8199
+    },
+    {
+      "epoch": 0.6321307431390688,
+      "grad_norm": 3.3853883743286133,
+      "learning_rate": 3.148749343712282e-06,
+      "loss": 1.0348,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6322078322540857,
+      "grad_norm": 3.7054595947265625,
+      "learning_rate": 3.1475896787130287e-06,
+      "loss": 0.9161,
+      "step": 8201
+    },
+    {
+      "epoch": 0.6322849213691026,
+      "grad_norm": 4.188024997711182,
+      "learning_rate": 3.146430129201965e-06,
+      "loss": 1.0198,
+      "step": 8202
+    },
+    {
+      "epoch": 0.6323620104841197,
+      "grad_norm": 3.5970795154571533,
+      "learning_rate": 3.1452706952513836e-06,
+      "loss": 1.0027,
+      "step": 8203
+    },
+    {
+      "epoch": 0.6324390995991366,
+      "grad_norm": 4.296309947967529,
+      "learning_rate": 3.144111376933568e-06,
+      "loss": 0.9452,
+      "step": 8204
+    },
+    {
+      "epoch": 0.6325161887141536,
+      "grad_norm": 3.8872005939483643,
+      "learning_rate": 3.142952174320797e-06,
+      "loss": 0.985,
+      "step": 8205
+    },
+    {
+      "epoch": 0.6325932778291705,
+      "grad_norm": 3.649055004119873,
+      "learning_rate": 3.1417930874853386e-06,
+      "loss": 0.8932,
+      "step": 8206
+    },
+    {
+      "epoch": 0.6326703669441874,
+      "grad_norm": 3.9435665607452393,
+      "learning_rate": 3.1406341164994574e-06,
+      "loss": 0.9035,
+      "step": 8207
+    },
+    {
+      "epoch": 0.6327474560592045,
+      "grad_norm": 3.8601176738739014,
+      "learning_rate": 3.1394752614354106e-06,
+      "loss": 0.8653,
+      "step": 8208
+    },
+    {
+      "epoch": 0.6328245451742214,
+      "grad_norm": 3.7992372512817383,
+      "learning_rate": 3.1383165223654444e-06,
+      "loss": 0.9261,
+      "step": 8209
+    },
+    {
+      "epoch": 0.6329016342892384,
+      "grad_norm": 3.477597236633301,
+      "learning_rate": 3.137157899361799e-06,
+      "loss": 0.9783,
+      "step": 8210
+    },
+    {
+      "epoch": 0.6329787234042553,
+      "grad_norm": 3.3875021934509277,
+      "learning_rate": 3.1359993924967124e-06,
+      "loss": 0.8615,
+      "step": 8211
+    },
+    {
+      "epoch": 0.6330558125192722,
+      "grad_norm": 3.7047359943389893,
+      "learning_rate": 3.1348410018424115e-06,
+      "loss": 1.0487,
+      "step": 8212
+    },
+    {
+      "epoch": 0.6331329016342893,
+      "grad_norm": 3.5324010848999023,
+      "learning_rate": 3.1336827274711124e-06,
+      "loss": 0.8742,
+      "step": 8213
+    },
+    {
+      "epoch": 0.6332099907493062,
+      "grad_norm": 3.7448694705963135,
+      "learning_rate": 3.1325245694550292e-06,
+      "loss": 0.9452,
+      "step": 8214
+    },
+    {
+      "epoch": 0.6332870798643232,
+      "grad_norm": 4.035982608795166,
+      "learning_rate": 3.1313665278663686e-06,
+      "loss": 1.0678,
+      "step": 8215
+    },
+    {
+      "epoch": 0.6333641689793401,
+      "grad_norm": 3.4959230422973633,
+      "learning_rate": 3.130208602777326e-06,
+      "loss": 0.8715,
+      "step": 8216
+    },
+    {
+      "epoch": 0.633441258094357,
+      "grad_norm": 4.09905481338501,
+      "learning_rate": 3.1290507942600936e-06,
+      "loss": 0.9712,
+      "step": 8217
+    },
+    {
+      "epoch": 0.633518347209374,
+      "grad_norm": 4.000631332397461,
+      "learning_rate": 3.1278931023868543e-06,
+      "loss": 0.8744,
+      "step": 8218
+    },
+    {
+      "epoch": 0.633595436324391,
+      "grad_norm": 3.4521965980529785,
+      "learning_rate": 3.126735527229784e-06,
+      "loss": 0.8514,
+      "step": 8219
+    },
+    {
+      "epoch": 0.633672525439408,
+      "grad_norm": 3.4077889919281006,
+      "learning_rate": 3.125578068861051e-06,
+      "loss": 0.9584,
+      "step": 8220
+    },
+    {
+      "epoch": 0.6337496145544249,
+      "grad_norm": 3.76359224319458,
+      "learning_rate": 3.124420727352819e-06,
+      "loss": 0.8964,
+      "step": 8221
+    },
+    {
+      "epoch": 0.6338267036694418,
+      "grad_norm": 3.4744224548339844,
+      "learning_rate": 3.1232635027772397e-06,
+      "loss": 0.9265,
+      "step": 8222
+    },
+    {
+      "epoch": 0.6339037927844589,
+      "grad_norm": 3.7400758266448975,
+      "learning_rate": 3.122106395206462e-06,
+      "loss": 0.9967,
+      "step": 8223
+    },
+    {
+      "epoch": 0.6339808818994758,
+      "grad_norm": 3.5406832695007324,
+      "learning_rate": 3.1209494047126233e-06,
+      "loss": 0.9603,
+      "step": 8224
+    },
+    {
+      "epoch": 0.6340579710144928,
+      "grad_norm": 3.479562520980835,
+      "learning_rate": 3.119792531367858e-06,
+      "loss": 0.8891,
+      "step": 8225
+    },
+    {
+      "epoch": 0.6341350601295097,
+      "grad_norm": 3.7099175453186035,
+      "learning_rate": 3.1186357752442915e-06,
+      "loss": 0.9093,
+      "step": 8226
+    },
+    {
+      "epoch": 0.6342121492445266,
+      "grad_norm": 4.117136478424072,
+      "learning_rate": 3.1174791364140394e-06,
+      "loss": 0.8646,
+      "step": 8227
+    },
+    {
+      "epoch": 0.6342892383595437,
+      "grad_norm": 3.6180074214935303,
+      "learning_rate": 3.1163226149492133e-06,
+      "loss": 0.7977,
+      "step": 8228
+    },
+    {
+      "epoch": 0.6343663274745606,
+      "grad_norm": 3.70717191696167,
+      "learning_rate": 3.1151662109219173e-06,
+      "loss": 0.8866,
+      "step": 8229
+    },
+    {
+      "epoch": 0.6344434165895776,
+      "grad_norm": 3.557631492614746,
+      "learning_rate": 3.1140099244042454e-06,
+      "loss": 0.8787,
+      "step": 8230
+    },
+    {
+      "epoch": 0.6345205057045945,
+      "grad_norm": 3.4483120441436768,
+      "learning_rate": 3.1128537554682868e-06,
+      "loss": 0.8503,
+      "step": 8231
+    },
+    {
+      "epoch": 0.6345975948196114,
+      "grad_norm": 3.401564121246338,
+      "learning_rate": 3.111697704186124e-06,
+      "loss": 0.8023,
+      "step": 8232
+    },
+    {
+      "epoch": 0.6346746839346284,
+      "grad_norm": 3.990302562713623,
+      "learning_rate": 3.110541770629828e-06,
+      "loss": 0.9897,
+      "step": 8233
+    },
+    {
+      "epoch": 0.6347517730496454,
+      "grad_norm": 3.496260166168213,
+      "learning_rate": 3.109385954871469e-06,
+      "loss": 1.0002,
+      "step": 8234
+    },
+    {
+      "epoch": 0.6348288621646624,
+      "grad_norm": 3.497117757797241,
+      "learning_rate": 3.108230256983102e-06,
+      "loss": 0.9513,
+      "step": 8235
+    },
+    {
+      "epoch": 0.6349059512796793,
+      "grad_norm": 3.9657294750213623,
+      "learning_rate": 3.107074677036781e-06,
+      "loss": 0.96,
+      "step": 8236
+    },
+    {
+      "epoch": 0.6349830403946962,
+      "grad_norm": 3.5663013458251953,
+      "learning_rate": 3.1059192151045507e-06,
+      "loss": 0.8036,
+      "step": 8237
+    },
+    {
+      "epoch": 0.6350601295097132,
+      "grad_norm": 3.7324275970458984,
+      "learning_rate": 3.104763871258447e-06,
+      "loss": 0.971,
+      "step": 8238
+    },
+    {
+      "epoch": 0.6351372186247302,
+      "grad_norm": 4.242516994476318,
+      "learning_rate": 3.1036086455705006e-06,
+      "loss": 1.0418,
+      "step": 8239
+    },
+    {
+      "epoch": 0.6352143077397472,
+      "grad_norm": 3.5605618953704834,
+      "learning_rate": 3.102453538112734e-06,
+      "loss": 0.9253,
+      "step": 8240
+    },
+    {
+      "epoch": 0.6352913968547641,
+      "grad_norm": 3.748979330062866,
+      "learning_rate": 3.1012985489571613e-06,
+      "loss": 0.972,
+      "step": 8241
+    },
+    {
+      "epoch": 0.635368485969781,
+      "grad_norm": 3.898329734802246,
+      "learning_rate": 3.10014367817579e-06,
+      "loss": 1.0175,
+      "step": 8242
+    },
+    {
+      "epoch": 0.635445575084798,
+      "grad_norm": 3.484175682067871,
+      "learning_rate": 3.098988925840621e-06,
+      "loss": 0.8839,
+      "step": 8243
+    },
+    {
+      "epoch": 0.635522664199815,
+      "grad_norm": 3.634960412979126,
+      "learning_rate": 3.097834292023647e-06,
+      "loss": 0.8899,
+      "step": 8244
+    },
+    {
+      "epoch": 0.635599753314832,
+      "grad_norm": 3.7136380672454834,
+      "learning_rate": 3.096679776796854e-06,
+      "loss": 0.9849,
+      "step": 8245
+    },
+    {
+      "epoch": 0.6356768424298489,
+      "grad_norm": 3.5262043476104736,
+      "learning_rate": 3.0955253802322183e-06,
+      "loss": 0.8174,
+      "step": 8246
+    },
+    {
+      "epoch": 0.6357539315448658,
+      "grad_norm": 3.480539321899414,
+      "learning_rate": 3.094371102401712e-06,
+      "loss": 0.8963,
+      "step": 8247
+    },
+    {
+      "epoch": 0.6358310206598828,
+      "grad_norm": 3.8594253063201904,
+      "learning_rate": 3.0932169433772986e-06,
+      "loss": 0.9647,
+      "step": 8248
+    },
+    {
+      "epoch": 0.6359081097748998,
+      "grad_norm": 3.4248909950256348,
+      "learning_rate": 3.0920629032309323e-06,
+      "loss": 0.8798,
+      "step": 8249
+    },
+    {
+      "epoch": 0.6359851988899168,
+      "grad_norm": 3.841808319091797,
+      "learning_rate": 3.090908982034563e-06,
+      "loss": 0.9655,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6360622880049337,
+      "grad_norm": 3.5504653453826904,
+      "learning_rate": 3.0897551798601312e-06,
+      "loss": 0.9077,
+      "step": 8251
+    },
+    {
+      "epoch": 0.6361393771199506,
+      "grad_norm": 3.5525784492492676,
+      "learning_rate": 3.0886014967795696e-06,
+      "loss": 0.8604,
+      "step": 8252
+    },
+    {
+      "epoch": 0.6362164662349676,
+      "grad_norm": 3.50411057472229,
+      "learning_rate": 3.087447932864807e-06,
+      "loss": 0.9307,
+      "step": 8253
+    },
+    {
+      "epoch": 0.6362935553499846,
+      "grad_norm": 3.3735544681549072,
+      "learning_rate": 3.086294488187758e-06,
+      "loss": 0.8722,
+      "step": 8254
+    },
+    {
+      "epoch": 0.6363706444650016,
+      "grad_norm": 3.8651130199432373,
+      "learning_rate": 3.0851411628203378e-06,
+      "loss": 0.9636,
+      "step": 8255
+    },
+    {
+      "epoch": 0.6364477335800185,
+      "grad_norm": 4.263221263885498,
+      "learning_rate": 3.083987956834449e-06,
+      "loss": 0.946,
+      "step": 8256
+    },
+    {
+      "epoch": 0.6365248226950354,
+      "grad_norm": 3.6470301151275635,
+      "learning_rate": 3.082834870301987e-06,
+      "loss": 0.9204,
+      "step": 8257
+    },
+    {
+      "epoch": 0.6366019118100524,
+      "grad_norm": 3.6749472618103027,
+      "learning_rate": 3.081681903294843e-06,
+      "loss": 1.0162,
+      "step": 8258
+    },
+    {
+      "epoch": 0.6366790009250693,
+      "grad_norm": 3.7581326961517334,
+      "learning_rate": 3.080529055884896e-06,
+      "loss": 0.8681,
+      "step": 8259
+    },
+    {
+      "epoch": 0.6367560900400864,
+      "grad_norm": 3.6538708209991455,
+      "learning_rate": 3.0793763281440225e-06,
+      "loss": 0.9529,
+      "step": 8260
+    },
+    {
+      "epoch": 0.6368331791551033,
+      "grad_norm": 3.6864404678344727,
+      "learning_rate": 3.0782237201440863e-06,
+      "loss": 1.0218,
+      "step": 8261
+    },
+    {
+      "epoch": 0.6369102682701202,
+      "grad_norm": 3.453056573867798,
+      "learning_rate": 3.077071231956948e-06,
+      "loss": 0.8975,
+      "step": 8262
+    },
+    {
+      "epoch": 0.6369873573851372,
+      "grad_norm": 3.943230390548706,
+      "learning_rate": 3.07591886365446e-06,
+      "loss": 0.9366,
+      "step": 8263
+    },
+    {
+      "epoch": 0.6370644465001541,
+      "grad_norm": 3.7501060962677,
+      "learning_rate": 3.0747666153084656e-06,
+      "loss": 0.9883,
+      "step": 8264
+    },
+    {
+      "epoch": 0.6371415356151712,
+      "grad_norm": 3.9778292179107666,
+      "learning_rate": 3.0736144869908015e-06,
+      "loss": 0.9615,
+      "step": 8265
+    },
+    {
+      "epoch": 0.6372186247301881,
+      "grad_norm": 3.7227420806884766,
+      "learning_rate": 3.072462478773298e-06,
+      "loss": 0.9409,
+      "step": 8266
+    },
+    {
+      "epoch": 0.637295713845205,
+      "grad_norm": 3.480952501296997,
+      "learning_rate": 3.071310590727775e-06,
+      "loss": 0.8985,
+      "step": 8267
+    },
+    {
+      "epoch": 0.637372802960222,
+      "grad_norm": 3.7210822105407715,
+      "learning_rate": 3.0701588229260478e-06,
+      "loss": 0.93,
+      "step": 8268
+    },
+    {
+      "epoch": 0.6374498920752389,
+      "grad_norm": 3.6173181533813477,
+      "learning_rate": 3.0690071754399236e-06,
+      "loss": 0.9362,
+      "step": 8269
+    },
+    {
+      "epoch": 0.637526981190256,
+      "grad_norm": 3.7527718544006348,
+      "learning_rate": 3.0678556483412005e-06,
+      "loss": 0.9783,
+      "step": 8270
+    },
+    {
+      "epoch": 0.6376040703052729,
+      "grad_norm": 3.4101197719573975,
+      "learning_rate": 3.066704241701672e-06,
+      "loss": 0.8431,
+      "step": 8271
+    },
+    {
+      "epoch": 0.6376811594202898,
+      "grad_norm": 3.5629935264587402,
+      "learning_rate": 3.06555295559312e-06,
+      "loss": 0.9556,
+      "step": 8272
+    },
+    {
+      "epoch": 0.6377582485353068,
+      "grad_norm": 3.466757297515869,
+      "learning_rate": 3.0644017900873225e-06,
+      "loss": 0.7801,
+      "step": 8273
+    },
+    {
+      "epoch": 0.6378353376503237,
+      "grad_norm": 3.4312379360198975,
+      "learning_rate": 3.06325074525605e-06,
+      "loss": 0.8572,
+      "step": 8274
+    },
+    {
+      "epoch": 0.6379124267653408,
+      "grad_norm": 3.8068385124206543,
+      "learning_rate": 3.0620998211710617e-06,
+      "loss": 0.9986,
+      "step": 8275
+    },
+    {
+      "epoch": 0.6379895158803577,
+      "grad_norm": 3.604086399078369,
+      "learning_rate": 3.0609490179041124e-06,
+      "loss": 0.9039,
+      "step": 8276
+    },
+    {
+      "epoch": 0.6380666049953746,
+      "grad_norm": 4.103906631469727,
+      "learning_rate": 3.059798335526951e-06,
+      "loss": 0.9925,
+      "step": 8277
+    },
+    {
+      "epoch": 0.6381436941103916,
+      "grad_norm": 3.9479622840881348,
+      "learning_rate": 3.0586477741113134e-06,
+      "loss": 0.931,
+      "step": 8278
+    },
+    {
+      "epoch": 0.6382207832254085,
+      "grad_norm": 3.976950168609619,
+      "learning_rate": 3.0574973337289327e-06,
+      "loss": 0.8651,
+      "step": 8279
+    },
+    {
+      "epoch": 0.6382978723404256,
+      "grad_norm": 3.6093039512634277,
+      "learning_rate": 3.0563470144515337e-06,
+      "loss": 0.9052,
+      "step": 8280
+    },
+    {
+      "epoch": 0.6383749614554425,
+      "grad_norm": 3.744586229324341,
+      "learning_rate": 3.0551968163508303e-06,
+      "loss": 0.9146,
+      "step": 8281
+    },
+    {
+      "epoch": 0.6384520505704595,
+      "grad_norm": 3.3914260864257812,
+      "learning_rate": 3.054046739498535e-06,
+      "loss": 0.8381,
+      "step": 8282
+    },
+    {
+      "epoch": 0.6385291396854764,
+      "grad_norm": 3.5384232997894287,
+      "learning_rate": 3.052896783966346e-06,
+      "loss": 0.8195,
+      "step": 8283
+    },
+    {
+      "epoch": 0.6386062288004933,
+      "grad_norm": 3.407332420349121,
+      "learning_rate": 3.051746949825958e-06,
+      "loss": 0.9074,
+      "step": 8284
+    },
+    {
+      "epoch": 0.6386833179155104,
+      "grad_norm": 4.286930084228516,
+      "learning_rate": 3.0505972371490584e-06,
+      "loss": 1.0354,
+      "step": 8285
+    },
+    {
+      "epoch": 0.6387604070305273,
+      "grad_norm": 3.8700149059295654,
+      "learning_rate": 3.0494476460073237e-06,
+      "loss": 0.9878,
+      "step": 8286
+    },
+    {
+      "epoch": 0.6388374961455443,
+      "grad_norm": 4.059021472930908,
+      "learning_rate": 3.048298176472426e-06,
+      "loss": 0.9244,
+      "step": 8287
+    },
+    {
+      "epoch": 0.6389145852605612,
+      "grad_norm": 3.820525646209717,
+      "learning_rate": 3.04714882861603e-06,
+      "loss": 1.104,
+      "step": 8288
+    },
+    {
+      "epoch": 0.6389916743755781,
+      "grad_norm": 3.6468265056610107,
+      "learning_rate": 3.04599960250979e-06,
+      "loss": 0.9207,
+      "step": 8289
+    },
+    {
+      "epoch": 0.6390687634905952,
+      "grad_norm": 3.7343029975891113,
+      "learning_rate": 3.044850498225354e-06,
+      "loss": 0.8921,
+      "step": 8290
+    },
+    {
+      "epoch": 0.6391458526056121,
+      "grad_norm": 3.6490302085876465,
+      "learning_rate": 3.0437015158343654e-06,
+      "loss": 0.9089,
+      "step": 8291
+    },
+    {
+      "epoch": 0.6392229417206291,
+      "grad_norm": 3.772552490234375,
+      "learning_rate": 3.0425526554084526e-06,
+      "loss": 0.9522,
+      "step": 8292
+    },
+    {
+      "epoch": 0.639300030835646,
+      "grad_norm": 3.4605162143707275,
+      "learning_rate": 3.0414039170192466e-06,
+      "loss": 0.8253,
+      "step": 8293
+    },
+    {
+      "epoch": 0.6393771199506629,
+      "grad_norm": 3.6937575340270996,
+      "learning_rate": 3.0402553007383607e-06,
+      "loss": 0.7973,
+      "step": 8294
+    },
+    {
+      "epoch": 0.63945420906568,
+      "grad_norm": 3.3988912105560303,
+      "learning_rate": 3.0391068066374073e-06,
+      "loss": 0.8901,
+      "step": 8295
+    },
+    {
+      "epoch": 0.6395312981806969,
+      "grad_norm": 3.475071668624878,
+      "learning_rate": 3.03795843478799e-06,
+      "loss": 0.8047,
+      "step": 8296
+    },
+    {
+      "epoch": 0.6396083872957139,
+      "grad_norm": 3.693103075027466,
+      "learning_rate": 3.0368101852617017e-06,
+      "loss": 0.926,
+      "step": 8297
+    },
+    {
+      "epoch": 0.6396854764107308,
+      "grad_norm": 3.585578203201294,
+      "learning_rate": 3.03566205813013e-06,
+      "loss": 0.9834,
+      "step": 8298
+    },
+    {
+      "epoch": 0.6397625655257477,
+      "grad_norm": 3.4358348846435547,
+      "learning_rate": 3.034514053464854e-06,
+      "loss": 0.8158,
+      "step": 8299
+    },
+    {
+      "epoch": 0.6398396546407648,
+      "grad_norm": 3.7729804515838623,
+      "learning_rate": 3.0333661713374506e-06,
+      "loss": 0.8983,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6399167437557817,
+      "grad_norm": 3.628103256225586,
+      "learning_rate": 3.0322184118194797e-06,
+      "loss": 0.8904,
+      "step": 8301
+    },
+    {
+      "epoch": 0.6399938328707987,
+      "grad_norm": 3.9679834842681885,
+      "learning_rate": 3.0310707749824987e-06,
+      "loss": 0.894,
+      "step": 8302
+    },
+    {
+      "epoch": 0.6400709219858156,
+      "grad_norm": 3.610292911529541,
+      "learning_rate": 3.0299232608980587e-06,
+      "loss": 0.9125,
+      "step": 8303
+    },
+    {
+      "epoch": 0.6401480111008325,
+      "grad_norm": 3.601590394973755,
+      "learning_rate": 3.0287758696377e-06,
+      "loss": 0.8971,
+      "step": 8304
+    },
+    {
+      "epoch": 0.6402251002158496,
+      "grad_norm": 3.709563732147217,
+      "learning_rate": 3.0276286012729563e-06,
+      "loss": 0.9295,
+      "step": 8305
+    },
+    {
+      "epoch": 0.6403021893308665,
+      "grad_norm": 3.6046383380889893,
+      "learning_rate": 3.0264814558753543e-06,
+      "loss": 0.9875,
+      "step": 8306
+    },
+    {
+      "epoch": 0.6403792784458835,
+      "grad_norm": 3.460169792175293,
+      "learning_rate": 3.0253344335164114e-06,
+      "loss": 0.9901,
+      "step": 8307
+    },
+    {
+      "epoch": 0.6404563675609004,
+      "grad_norm": 3.704393148422241,
+      "learning_rate": 3.0241875342676413e-06,
+      "loss": 0.9067,
+      "step": 8308
+    },
+    {
+      "epoch": 0.6405334566759173,
+      "grad_norm": 3.8875725269317627,
+      "learning_rate": 3.023040758200544e-06,
+      "loss": 0.9543,
+      "step": 8309
+    },
+    {
+      "epoch": 0.6406105457909343,
+      "grad_norm": 3.9328582286834717,
+      "learning_rate": 3.0218941053866167e-06,
+      "loss": 1.0593,
+      "step": 8310
+    },
+    {
+      "epoch": 0.6406876349059513,
+      "grad_norm": 4.249375343322754,
+      "learning_rate": 3.020747575897348e-06,
+      "loss": 0.9223,
+      "step": 8311
+    },
+    {
+      "epoch": 0.6407647240209683,
+      "grad_norm": 3.789259672164917,
+      "learning_rate": 3.019601169804216e-06,
+      "loss": 0.9254,
+      "step": 8312
+    },
+    {
+      "epoch": 0.6408418131359852,
+      "grad_norm": 3.5472848415374756,
+      "learning_rate": 3.0184548871786946e-06,
+      "loss": 0.9175,
+      "step": 8313
+    },
+    {
+      "epoch": 0.6409189022510021,
+      "grad_norm": 3.7821688652038574,
+      "learning_rate": 3.0173087280922493e-06,
+      "loss": 0.8734,
+      "step": 8314
+    },
+    {
+      "epoch": 0.6409959913660191,
+      "grad_norm": 3.4091010093688965,
+      "learning_rate": 3.0161626926163344e-06,
+      "loss": 0.9446,
+      "step": 8315
+    },
+    {
+      "epoch": 0.6410730804810361,
+      "grad_norm": 3.822620391845703,
+      "learning_rate": 3.015016780822402e-06,
+      "loss": 0.9002,
+      "step": 8316
+    },
+    {
+      "epoch": 0.6411501695960531,
+      "grad_norm": 3.6239142417907715,
+      "learning_rate": 3.013870992781894e-06,
+      "loss": 0.8363,
+      "step": 8317
+    },
+    {
+      "epoch": 0.64122725871107,
+      "grad_norm": 3.812986373901367,
+      "learning_rate": 3.0127253285662415e-06,
+      "loss": 0.9568,
+      "step": 8318
+    },
+    {
+      "epoch": 0.6413043478260869,
+      "grad_norm": 3.996857166290283,
+      "learning_rate": 3.0115797882468733e-06,
+      "loss": 0.863,
+      "step": 8319
+    },
+    {
+      "epoch": 0.6413814369411039,
+      "grad_norm": 3.7889609336853027,
+      "learning_rate": 3.0104343718952065e-06,
+      "loss": 1.0847,
+      "step": 8320
+    },
+    {
+      "epoch": 0.6414585260561209,
+      "grad_norm": 3.7491159439086914,
+      "learning_rate": 3.0092890795826524e-06,
+      "loss": 0.8731,
+      "step": 8321
+    },
+    {
+      "epoch": 0.6415356151711379,
+      "grad_norm": 3.6667377948760986,
+      "learning_rate": 3.0081439113806145e-06,
+      "loss": 0.8165,
+      "step": 8322
+    },
+    {
+      "epoch": 0.6416127042861548,
+      "grad_norm": 4.124176502227783,
+      "learning_rate": 3.0069988673604866e-06,
+      "loss": 1.0136,
+      "step": 8323
+    },
+    {
+      "epoch": 0.6416897934011717,
+      "grad_norm": 4.588274955749512,
+      "learning_rate": 3.0058539475936577e-06,
+      "loss": 1.0123,
+      "step": 8324
+    },
+    {
+      "epoch": 0.6417668825161887,
+      "grad_norm": 3.610478401184082,
+      "learning_rate": 3.0047091521515083e-06,
+      "loss": 0.868,
+      "step": 8325
+    },
+    {
+      "epoch": 0.6418439716312057,
+      "grad_norm": 3.607025384902954,
+      "learning_rate": 3.0035644811054083e-06,
+      "loss": 0.8376,
+      "step": 8326
+    },
+    {
+      "epoch": 0.6419210607462227,
+      "grad_norm": 3.4604544639587402,
+      "learning_rate": 3.002419934526723e-06,
+      "loss": 0.898,
+      "step": 8327
+    },
+    {
+      "epoch": 0.6419981498612396,
+      "grad_norm": 3.819643497467041,
+      "learning_rate": 3.00127551248681e-06,
+      "loss": 0.9074,
+      "step": 8328
+    },
+    {
+      "epoch": 0.6420752389762565,
+      "grad_norm": 3.4127678871154785,
+      "learning_rate": 3.000131215057016e-06,
+      "loss": 0.818,
+      "step": 8329
+    },
+    {
+      "epoch": 0.6421523280912735,
+      "grad_norm": 3.94305157661438,
+      "learning_rate": 2.998987042308685e-06,
+      "loss": 1.0063,
+      "step": 8330
+    },
+    {
+      "epoch": 0.6422294172062905,
+      "grad_norm": 3.653914213180542,
+      "learning_rate": 2.9978429943131466e-06,
+      "loss": 0.9659,
+      "step": 8331
+    },
+    {
+      "epoch": 0.6423065063213075,
+      "grad_norm": 3.779825210571289,
+      "learning_rate": 2.9966990711417283e-06,
+      "loss": 0.875,
+      "step": 8332
+    },
+    {
+      "epoch": 0.6423835954363244,
+      "grad_norm": 3.4283056259155273,
+      "learning_rate": 2.9955552728657483e-06,
+      "loss": 0.8167,
+      "step": 8333
+    },
+    {
+      "epoch": 0.6424606845513413,
+      "grad_norm": 3.8111422061920166,
+      "learning_rate": 2.994411599556515e-06,
+      "loss": 0.8542,
+      "step": 8334
+    },
+    {
+      "epoch": 0.6425377736663583,
+      "grad_norm": 3.65975284576416,
+      "learning_rate": 2.993268051285331e-06,
+      "loss": 0.8386,
+      "step": 8335
+    },
+    {
+      "epoch": 0.6426148627813753,
+      "grad_norm": 3.7688958644866943,
+      "learning_rate": 2.9921246281234924e-06,
+      "loss": 0.9334,
+      "step": 8336
+    },
+    {
+      "epoch": 0.6426919518963923,
+      "grad_norm": 3.5940749645233154,
+      "learning_rate": 2.990981330142283e-06,
+      "loss": 0.9051,
+      "step": 8337
+    },
+    {
+      "epoch": 0.6427690410114092,
+      "grad_norm": 3.8755249977111816,
+      "learning_rate": 2.989838157412982e-06,
+      "loss": 0.9065,
+      "step": 8338
+    },
+    {
+      "epoch": 0.6428461301264261,
+      "grad_norm": 4.170990943908691,
+      "learning_rate": 2.9886951100068628e-06,
+      "loss": 1.0524,
+      "step": 8339
+    },
+    {
+      "epoch": 0.6429232192414431,
+      "grad_norm": 3.61470890045166,
+      "learning_rate": 2.9875521879951852e-06,
+      "loss": 0.951,
+      "step": 8340
+    },
+    {
+      "epoch": 0.64300030835646,
+      "grad_norm": 3.7371344566345215,
+      "learning_rate": 2.9864093914492076e-06,
+      "loss": 0.9648,
+      "step": 8341
+    },
+    {
+      "epoch": 0.6430773974714771,
+      "grad_norm": 3.93182635307312,
+      "learning_rate": 2.9852667204401742e-06,
+      "loss": 1.0087,
+      "step": 8342
+    },
+    {
+      "epoch": 0.643154486586494,
+      "grad_norm": 3.8426945209503174,
+      "learning_rate": 2.9841241750393257e-06,
+      "loss": 0.9239,
+      "step": 8343
+    },
+    {
+      "epoch": 0.6432315757015109,
+      "grad_norm": 3.605802297592163,
+      "learning_rate": 2.9829817553178945e-06,
+      "loss": 0.9312,
+      "step": 8344
+    },
+    {
+      "epoch": 0.6433086648165279,
+      "grad_norm": 3.347144603729248,
+      "learning_rate": 2.981839461347105e-06,
+      "loss": 0.8394,
+      "step": 8345
+    },
+    {
+      "epoch": 0.6433857539315448,
+      "grad_norm": 3.462236166000366,
+      "learning_rate": 2.980697293198174e-06,
+      "loss": 0.9387,
+      "step": 8346
+    },
+    {
+      "epoch": 0.6434628430465619,
+      "grad_norm": 3.7107419967651367,
+      "learning_rate": 2.9795552509423075e-06,
+      "loss": 0.7882,
+      "step": 8347
+    },
+    {
+      "epoch": 0.6435399321615788,
+      "grad_norm": 3.59287166595459,
+      "learning_rate": 2.9784133346507077e-06,
+      "loss": 0.9505,
+      "step": 8348
+    },
+    {
+      "epoch": 0.6436170212765957,
+      "grad_norm": 3.7674858570098877,
+      "learning_rate": 2.9772715443945653e-06,
+      "loss": 0.9572,
+      "step": 8349
+    },
+    {
+      "epoch": 0.6436941103916127,
+      "grad_norm": 4.125513076782227,
+      "learning_rate": 2.9761298802450667e-06,
+      "loss": 0.9329,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6437711995066296,
+      "grad_norm": 3.9114468097686768,
+      "learning_rate": 2.974988342273388e-06,
+      "loss": 0.8998,
+      "step": 8351
+    },
+    {
+      "epoch": 0.6438482886216467,
+      "grad_norm": 3.4808576107025146,
+      "learning_rate": 2.9738469305506985e-06,
+      "loss": 0.865,
+      "step": 8352
+    },
+    {
+      "epoch": 0.6439253777366636,
+      "grad_norm": 3.3427348136901855,
+      "learning_rate": 2.972705645148159e-06,
+      "loss": 0.9336,
+      "step": 8353
+    },
+    {
+      "epoch": 0.6440024668516805,
+      "grad_norm": 3.8578758239746094,
+      "learning_rate": 2.9715644861369235e-06,
+      "loss": 0.9053,
+      "step": 8354
+    },
+    {
+      "epoch": 0.6440795559666975,
+      "grad_norm": 3.955080986022949,
+      "learning_rate": 2.970423453588136e-06,
+      "loss": 1.0019,
+      "step": 8355
+    },
+    {
+      "epoch": 0.6441566450817144,
+      "grad_norm": 3.818481922149658,
+      "learning_rate": 2.969282547572936e-06,
+      "loss": 0.9418,
+      "step": 8356
+    },
+    {
+      "epoch": 0.6442337341967315,
+      "grad_norm": 3.748859405517578,
+      "learning_rate": 2.9681417681624502e-06,
+      "loss": 1.0381,
+      "step": 8357
+    },
+    {
+      "epoch": 0.6443108233117484,
+      "grad_norm": 4.283618450164795,
+      "learning_rate": 2.967001115427802e-06,
+      "loss": 1.0513,
+      "step": 8358
+    },
+    {
+      "epoch": 0.6443879124267653,
+      "grad_norm": 3.595768690109253,
+      "learning_rate": 2.965860589440106e-06,
+      "loss": 0.8752,
+      "step": 8359
+    },
+    {
+      "epoch": 0.6444650015417823,
+      "grad_norm": 3.731633424758911,
+      "learning_rate": 2.964720190270467e-06,
+      "loss": 0.9551,
+      "step": 8360
+    },
+    {
+      "epoch": 0.6445420906567992,
+      "grad_norm": 3.895328998565674,
+      "learning_rate": 2.9635799179899822e-06,
+      "loss": 0.8959,
+      "step": 8361
+    },
+    {
+      "epoch": 0.6446191797718163,
+      "grad_norm": 3.839773654937744,
+      "learning_rate": 2.962439772669744e-06,
+      "loss": 0.9535,
+      "step": 8362
+    },
+    {
+      "epoch": 0.6446962688868332,
+      "grad_norm": 3.9109835624694824,
+      "learning_rate": 2.961299754380832e-06,
+      "loss": 0.8728,
+      "step": 8363
+    },
+    {
+      "epoch": 0.6447733580018501,
+      "grad_norm": 3.788618803024292,
+      "learning_rate": 2.960159863194322e-06,
+      "loss": 0.9667,
+      "step": 8364
+    },
+    {
+      "epoch": 0.6448504471168671,
+      "grad_norm": 4.191736698150635,
+      "learning_rate": 2.9590200991812804e-06,
+      "loss": 1.0588,
+      "step": 8365
+    },
+    {
+      "epoch": 0.644927536231884,
+      "grad_norm": 3.4580607414245605,
+      "learning_rate": 2.9578804624127644e-06,
+      "loss": 0.849,
+      "step": 8366
+    },
+    {
+      "epoch": 0.6450046253469011,
+      "grad_norm": 3.6437692642211914,
+      "learning_rate": 2.9567409529598267e-06,
+      "loss": 0.9661,
+      "step": 8367
+    },
+    {
+      "epoch": 0.645081714461918,
+      "grad_norm": 3.486619472503662,
+      "learning_rate": 2.9556015708935064e-06,
+      "loss": 0.855,
+      "step": 8368
+    },
+    {
+      "epoch": 0.6451588035769349,
+      "grad_norm": 3.5905685424804688,
+      "learning_rate": 2.9544623162848406e-06,
+      "loss": 0.9604,
+      "step": 8369
+    },
+    {
+      "epoch": 0.6452358926919519,
+      "grad_norm": 3.5381009578704834,
+      "learning_rate": 2.9533231892048557e-06,
+      "loss": 0.8643,
+      "step": 8370
+    },
+    {
+      "epoch": 0.6453129818069688,
+      "grad_norm": 3.4007678031921387,
+      "learning_rate": 2.9521841897245694e-06,
+      "loss": 0.8136,
+      "step": 8371
+    },
+    {
+      "epoch": 0.6453900709219859,
+      "grad_norm": 3.962644338607788,
+      "learning_rate": 2.9510453179149923e-06,
+      "loss": 0.9401,
+      "step": 8372
+    },
+    {
+      "epoch": 0.6454671600370028,
+      "grad_norm": 3.902945041656494,
+      "learning_rate": 2.94990657384713e-06,
+      "loss": 0.8761,
+      "step": 8373
+    },
+    {
+      "epoch": 0.6455442491520197,
+      "grad_norm": 3.3986709117889404,
+      "learning_rate": 2.9487679575919733e-06,
+      "loss": 0.9076,
+      "step": 8374
+    },
+    {
+      "epoch": 0.6456213382670367,
+      "grad_norm": 3.4570724964141846,
+      "learning_rate": 2.947629469220511e-06,
+      "loss": 0.8969,
+      "step": 8375
+    },
+    {
+      "epoch": 0.6456984273820536,
+      "grad_norm": 3.3817062377929688,
+      "learning_rate": 2.9464911088037225e-06,
+      "loss": 0.7891,
+      "step": 8376
+    },
+    {
+      "epoch": 0.6457755164970707,
+      "grad_norm": 4.3606696128845215,
+      "learning_rate": 2.945352876412577e-06,
+      "loss": 0.8679,
+      "step": 8377
+    },
+    {
+      "epoch": 0.6458526056120876,
+      "grad_norm": 3.8586175441741943,
+      "learning_rate": 2.94421477211804e-06,
+      "loss": 0.9431,
+      "step": 8378
+    },
+    {
+      "epoch": 0.6459296947271045,
+      "grad_norm": 3.5619122982025146,
+      "learning_rate": 2.9430767959910633e-06,
+      "loss": 0.8947,
+      "step": 8379
+    },
+    {
+      "epoch": 0.6460067838421215,
+      "grad_norm": 3.578460693359375,
+      "learning_rate": 2.941938948102595e-06,
+      "loss": 0.9447,
+      "step": 8380
+    },
+    {
+      "epoch": 0.6460838729571384,
+      "grad_norm": 3.719597101211548,
+      "learning_rate": 2.9408012285235753e-06,
+      "loss": 0.9453,
+      "step": 8381
+    },
+    {
+      "epoch": 0.6461609620721555,
+      "grad_norm": 3.9067883491516113,
+      "learning_rate": 2.939663637324934e-06,
+      "loss": 0.9575,
+      "step": 8382
+    },
+    {
+      "epoch": 0.6462380511871724,
+      "grad_norm": 3.6373391151428223,
+      "learning_rate": 2.9385261745775932e-06,
+      "loss": 0.8788,
+      "step": 8383
+    },
+    {
+      "epoch": 0.6463151403021893,
+      "grad_norm": 3.917569160461426,
+      "learning_rate": 2.93738884035247e-06,
+      "loss": 0.9704,
+      "step": 8384
+    },
+    {
+      "epoch": 0.6463922294172063,
+      "grad_norm": 3.8104026317596436,
+      "learning_rate": 2.9362516347204695e-06,
+      "loss": 0.9048,
+      "step": 8385
+    },
+    {
+      "epoch": 0.6464693185322232,
+      "grad_norm": 3.8119072914123535,
+      "learning_rate": 2.9351145577524908e-06,
+      "loss": 0.9322,
+      "step": 8386
+    },
+    {
+      "epoch": 0.6465464076472403,
+      "grad_norm": 3.7479817867279053,
+      "learning_rate": 2.9339776095194245e-06,
+      "loss": 0.9542,
+      "step": 8387
+    },
+    {
+      "epoch": 0.6466234967622572,
+      "grad_norm": 3.774230718612671,
+      "learning_rate": 2.9328407900921555e-06,
+      "loss": 0.9974,
+      "step": 8388
+    },
+    {
+      "epoch": 0.6467005858772741,
+      "grad_norm": 3.979708671569824,
+      "learning_rate": 2.931704099541557e-06,
+      "loss": 1.0159,
+      "step": 8389
+    },
+    {
+      "epoch": 0.6467776749922911,
+      "grad_norm": 3.6764190196990967,
+      "learning_rate": 2.9305675379384945e-06,
+      "loss": 0.9333,
+      "step": 8390
+    },
+    {
+      "epoch": 0.646854764107308,
+      "grad_norm": 3.608494997024536,
+      "learning_rate": 2.9294311053538306e-06,
+      "loss": 0.8975,
+      "step": 8391
+    },
+    {
+      "epoch": 0.646931853222325,
+      "grad_norm": 3.679772138595581,
+      "learning_rate": 2.9282948018584116e-06,
+      "loss": 0.8043,
+      "step": 8392
+    },
+    {
+      "epoch": 0.647008942337342,
+      "grad_norm": 3.3427774906158447,
+      "learning_rate": 2.9271586275230823e-06,
+      "loss": 0.7306,
+      "step": 8393
+    },
+    {
+      "epoch": 0.6470860314523589,
+      "grad_norm": 3.9158496856689453,
+      "learning_rate": 2.926022582418678e-06,
+      "loss": 0.7937,
+      "step": 8394
+    },
+    {
+      "epoch": 0.6471631205673759,
+      "grad_norm": 3.52773380279541,
+      "learning_rate": 2.9248866666160236e-06,
+      "loss": 0.9314,
+      "step": 8395
+    },
+    {
+      "epoch": 0.6472402096823928,
+      "grad_norm": 4.079388618469238,
+      "learning_rate": 2.9237508801859394e-06,
+      "loss": 1.0586,
+      "step": 8396
+    },
+    {
+      "epoch": 0.6473172987974098,
+      "grad_norm": 3.9858343601226807,
+      "learning_rate": 2.922615223199233e-06,
+      "loss": 0.9373,
+      "step": 8397
+    },
+    {
+      "epoch": 0.6473943879124268,
+      "grad_norm": 3.436227798461914,
+      "learning_rate": 2.921479695726709e-06,
+      "loss": 0.9468,
+      "step": 8398
+    },
+    {
+      "epoch": 0.6474714770274437,
+      "grad_norm": 3.483386278152466,
+      "learning_rate": 2.9203442978391618e-06,
+      "loss": 0.9141,
+      "step": 8399
+    },
+    {
+      "epoch": 0.6475485661424607,
+      "grad_norm": 3.5396289825439453,
+      "learning_rate": 2.9192090296073755e-06,
+      "loss": 0.93,
+      "step": 8400
+    },
+    {
+      "epoch": 0.6476256552574776,
+      "grad_norm": 3.383061170578003,
+      "learning_rate": 2.91807389110213e-06,
+      "loss": 0.8128,
+      "step": 8401
+    },
+    {
+      "epoch": 0.6477027443724946,
+      "grad_norm": 3.429288864135742,
+      "learning_rate": 2.916938882394194e-06,
+      "loss": 0.8793,
+      "step": 8402
+    },
+    {
+      "epoch": 0.6477798334875116,
+      "grad_norm": 3.536487340927124,
+      "learning_rate": 2.9158040035543334e-06,
+      "loss": 0.9317,
+      "step": 8403
+    },
+    {
+      "epoch": 0.6478569226025285,
+      "grad_norm": 3.7233903408050537,
+      "learning_rate": 2.914669254653296e-06,
+      "loss": 0.8154,
+      "step": 8404
+    },
+    {
+      "epoch": 0.6479340117175455,
+      "grad_norm": 3.469984292984009,
+      "learning_rate": 2.9135346357618304e-06,
+      "loss": 0.9265,
+      "step": 8405
+    },
+    {
+      "epoch": 0.6480111008325624,
+      "grad_norm": 3.6804935932159424,
+      "learning_rate": 2.912400146950675e-06,
+      "loss": 1.0209,
+      "step": 8406
+    },
+    {
+      "epoch": 0.6480881899475794,
+      "grad_norm": 3.8416061401367188,
+      "learning_rate": 2.9112657882905573e-06,
+      "loss": 0.9045,
+      "step": 8407
+    },
+    {
+      "epoch": 0.6481652790625964,
+      "grad_norm": 3.884570360183716,
+      "learning_rate": 2.910131559852201e-06,
+      "loss": 0.8565,
+      "step": 8408
+    },
+    {
+      "epoch": 0.6482423681776133,
+      "grad_norm": 3.519493341445923,
+      "learning_rate": 2.9089974617063188e-06,
+      "loss": 0.9198,
+      "step": 8409
+    },
+    {
+      "epoch": 0.6483194572926303,
+      "grad_norm": 4.712925434112549,
+      "learning_rate": 2.9078634939236146e-06,
+      "loss": 1.0443,
+      "step": 8410
+    },
+    {
+      "epoch": 0.6483965464076472,
+      "grad_norm": 6.0556559562683105,
+      "learning_rate": 2.9067296565747856e-06,
+      "loss": 0.8174,
+      "step": 8411
+    },
+    {
+      "epoch": 0.6484736355226642,
+      "grad_norm": 3.844569206237793,
+      "learning_rate": 2.905595949730521e-06,
+      "loss": 1.0147,
+      "step": 8412
+    },
+    {
+      "epoch": 0.6485507246376812,
+      "grad_norm": 3.9868366718292236,
+      "learning_rate": 2.9044623734615018e-06,
+      "loss": 0.9485,
+      "step": 8413
+    },
+    {
+      "epoch": 0.6486278137526981,
+      "grad_norm": 3.9000837802886963,
+      "learning_rate": 2.903328927838403e-06,
+      "loss": 1.0351,
+      "step": 8414
+    },
+    {
+      "epoch": 0.6487049028677151,
+      "grad_norm": 3.564354419708252,
+      "learning_rate": 2.902195612931884e-06,
+      "loss": 0.892,
+      "step": 8415
+    },
+    {
+      "epoch": 0.648781991982732,
+      "grad_norm": 3.669095277786255,
+      "learning_rate": 2.901062428812604e-06,
+      "loss": 0.9128,
+      "step": 8416
+    },
+    {
+      "epoch": 0.648859081097749,
+      "grad_norm": 3.598200798034668,
+      "learning_rate": 2.8999293755512113e-06,
+      "loss": 0.8725,
+      "step": 8417
+    },
+    {
+      "epoch": 0.648936170212766,
+      "grad_norm": 3.7269628047943115,
+      "learning_rate": 2.8987964532183454e-06,
+      "loss": 0.8105,
+      "step": 8418
+    },
+    {
+      "epoch": 0.6490132593277829,
+      "grad_norm": 3.622880458831787,
+      "learning_rate": 2.897663661884638e-06,
+      "loss": 0.8715,
+      "step": 8419
+    },
+    {
+      "epoch": 0.6490903484427999,
+      "grad_norm": 3.9800972938537598,
+      "learning_rate": 2.896531001620715e-06,
+      "loss": 0.9407,
+      "step": 8420
+    },
+    {
+      "epoch": 0.6491674375578168,
+      "grad_norm": 4.113578796386719,
+      "learning_rate": 2.8953984724971874e-06,
+      "loss": 0.9153,
+      "step": 8421
+    },
+    {
+      "epoch": 0.6492445266728338,
+      "grad_norm": 4.180689811706543,
+      "learning_rate": 2.8942660745846657e-06,
+      "loss": 0.8857,
+      "step": 8422
+    },
+    {
+      "epoch": 0.6493216157878507,
+      "grad_norm": 4.276498794555664,
+      "learning_rate": 2.8931338079537487e-06,
+      "loss": 0.9398,
+      "step": 8423
+    },
+    {
+      "epoch": 0.6493987049028677,
+      "grad_norm": 3.5343918800354004,
+      "learning_rate": 2.892001672675026e-06,
+      "loss": 0.9164,
+      "step": 8424
+    },
+    {
+      "epoch": 0.6494757940178847,
+      "grad_norm": 3.7870051860809326,
+      "learning_rate": 2.890869668819084e-06,
+      "loss": 1.0146,
+      "step": 8425
+    },
+    {
+      "epoch": 0.6495528831329016,
+      "grad_norm": 3.560620069503784,
+      "learning_rate": 2.8897377964564922e-06,
+      "loss": 0.8416,
+      "step": 8426
+    },
+    {
+      "epoch": 0.6496299722479186,
+      "grad_norm": 3.742684841156006,
+      "learning_rate": 2.88860605565782e-06,
+      "loss": 0.9449,
+      "step": 8427
+    },
+    {
+      "epoch": 0.6497070613629355,
+      "grad_norm": 3.494896411895752,
+      "learning_rate": 2.887474446493625e-06,
+      "loss": 0.8439,
+      "step": 8428
+    },
+    {
+      "epoch": 0.6497841504779525,
+      "grad_norm": 3.8634705543518066,
+      "learning_rate": 2.886342969034457e-06,
+      "loss": 0.9453,
+      "step": 8429
+    },
+    {
+      "epoch": 0.6498612395929695,
+      "grad_norm": 3.891278028488159,
+      "learning_rate": 2.8852116233508604e-06,
+      "loss": 1.0619,
+      "step": 8430
+    },
+    {
+      "epoch": 0.6499383287079864,
+      "grad_norm": 3.782010793685913,
+      "learning_rate": 2.8840804095133617e-06,
+      "loss": 1.0296,
+      "step": 8431
+    },
+    {
+      "epoch": 0.6500154178230034,
+      "grad_norm": 3.684227466583252,
+      "learning_rate": 2.8829493275924935e-06,
+      "loss": 0.9614,
+      "step": 8432
+    },
+    {
+      "epoch": 0.6500925069380203,
+      "grad_norm": 3.9234118461608887,
+      "learning_rate": 2.8818183776587694e-06,
+      "loss": 1.0703,
+      "step": 8433
+    },
+    {
+      "epoch": 0.6501695960530373,
+      "grad_norm": 3.863373279571533,
+      "learning_rate": 2.8806875597827e-06,
+      "loss": 0.9689,
+      "step": 8434
+    },
+    {
+      "epoch": 0.6502466851680543,
+      "grad_norm": 3.6144754886627197,
+      "learning_rate": 2.8795568740347868e-06,
+      "loss": 0.8714,
+      "step": 8435
+    },
+    {
+      "epoch": 0.6503237742830712,
+      "grad_norm": 4.281971454620361,
+      "learning_rate": 2.878426320485518e-06,
+      "loss": 0.9231,
+      "step": 8436
+    },
+    {
+      "epoch": 0.6504008633980882,
+      "grad_norm": 4.0023040771484375,
+      "learning_rate": 2.87729589920538e-06,
+      "loss": 0.9989,
+      "step": 8437
+    },
+    {
+      "epoch": 0.6504779525131051,
+      "grad_norm": 3.5423178672790527,
+      "learning_rate": 2.8761656102648495e-06,
+      "loss": 0.9044,
+      "step": 8438
+    },
+    {
+      "epoch": 0.650555041628122,
+      "grad_norm": 3.6947436332702637,
+      "learning_rate": 2.875035453734394e-06,
+      "loss": 0.9198,
+      "step": 8439
+    },
+    {
+      "epoch": 0.6506321307431391,
+      "grad_norm": 3.7778401374816895,
+      "learning_rate": 2.8739054296844736e-06,
+      "loss": 0.9065,
+      "step": 8440
+    },
+    {
+      "epoch": 0.650709219858156,
+      "grad_norm": 3.7728586196899414,
+      "learning_rate": 2.8727755381855373e-06,
+      "loss": 0.9014,
+      "step": 8441
+    },
+    {
+      "epoch": 0.650786308973173,
+      "grad_norm": 3.6431329250335693,
+      "learning_rate": 2.8716457793080277e-06,
+      "loss": 0.9221,
+      "step": 8442
+    },
+    {
+      "epoch": 0.6508633980881899,
+      "grad_norm": 3.698190927505493,
+      "learning_rate": 2.8705161531223814e-06,
+      "loss": 0.9884,
+      "step": 8443
+    },
+    {
+      "epoch": 0.6509404872032069,
+      "grad_norm": 4.110282897949219,
+      "learning_rate": 2.8693866596990237e-06,
+      "loss": 0.9542,
+      "step": 8444
+    },
+    {
+      "epoch": 0.6510175763182239,
+      "grad_norm": 4.268514156341553,
+      "learning_rate": 2.868257299108374e-06,
+      "loss": 1.0497,
+      "step": 8445
+    },
+    {
+      "epoch": 0.6510946654332408,
+      "grad_norm": 3.500208616256714,
+      "learning_rate": 2.8671280714208427e-06,
+      "loss": 0.8821,
+      "step": 8446
+    },
+    {
+      "epoch": 0.6511717545482578,
+      "grad_norm": 3.6203670501708984,
+      "learning_rate": 2.8659989767068277e-06,
+      "loss": 0.8401,
+      "step": 8447
+    },
+    {
+      "epoch": 0.6512488436632747,
+      "grad_norm": 3.4511590003967285,
+      "learning_rate": 2.8648700150367236e-06,
+      "loss": 0.8388,
+      "step": 8448
+    },
+    {
+      "epoch": 0.6513259327782918,
+      "grad_norm": 3.800976514816284,
+      "learning_rate": 2.863741186480917e-06,
+      "loss": 0.977,
+      "step": 8449
+    },
+    {
+      "epoch": 0.6514030218933087,
+      "grad_norm": 3.96077561378479,
+      "learning_rate": 2.862612491109783e-06,
+      "loss": 1.0604,
+      "step": 8450
+    },
+    {
+      "epoch": 0.6514801110083256,
+      "grad_norm": 3.6005845069885254,
+      "learning_rate": 2.861483928993692e-06,
+      "loss": 1.0056,
+      "step": 8451
+    },
+    {
+      "epoch": 0.6515572001233426,
+      "grad_norm": 3.862231492996216,
+      "learning_rate": 2.8603555002030005e-06,
+      "loss": 0.9447,
+      "step": 8452
+    },
+    {
+      "epoch": 0.6516342892383595,
+      "grad_norm": 3.786400318145752,
+      "learning_rate": 2.8592272048080623e-06,
+      "loss": 0.9996,
+      "step": 8453
+    },
+    {
+      "epoch": 0.6517113783533766,
+      "grad_norm": 4.21661901473999,
+      "learning_rate": 2.8580990428792205e-06,
+      "loss": 0.9962,
+      "step": 8454
+    },
+    {
+      "epoch": 0.6517884674683935,
+      "grad_norm": 3.853390693664551,
+      "learning_rate": 2.85697101448681e-06,
+      "loss": 0.9301,
+      "step": 8455
+    },
+    {
+      "epoch": 0.6518655565834104,
+      "grad_norm": 3.9790003299713135,
+      "learning_rate": 2.855843119701158e-06,
+      "loss": 0.8526,
+      "step": 8456
+    },
+    {
+      "epoch": 0.6519426456984274,
+      "grad_norm": 4.153457164764404,
+      "learning_rate": 2.8547153585925842e-06,
+      "loss": 0.9439,
+      "step": 8457
+    },
+    {
+      "epoch": 0.6520197348134443,
+      "grad_norm": 3.5101563930511475,
+      "learning_rate": 2.853587731231395e-06,
+      "loss": 0.8871,
+      "step": 8458
+    },
+    {
+      "epoch": 0.6520968239284614,
+      "grad_norm": 3.815025806427002,
+      "learning_rate": 2.8524602376878952e-06,
+      "loss": 0.9554,
+      "step": 8459
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 3.934506893157959,
+      "learning_rate": 2.8513328780323763e-06,
+      "loss": 0.8709,
+      "step": 8460
+    },
+    {
+      "epoch": 0.6522510021584952,
+      "grad_norm": 3.9895248413085938,
+      "learning_rate": 2.850205652335124e-06,
+      "loss": 0.976,
+      "step": 8461
+    },
+    {
+      "epoch": 0.6523280912735122,
+      "grad_norm": 3.5615084171295166,
+      "learning_rate": 2.8490785606664177e-06,
+      "loss": 0.8848,
+      "step": 8462
+    },
+    {
+      "epoch": 0.6524051803885291,
+      "grad_norm": 3.748917579650879,
+      "learning_rate": 2.8479516030965215e-06,
+      "loss": 0.8879,
+      "step": 8463
+    },
+    {
+      "epoch": 0.6524822695035462,
+      "grad_norm": 3.557553768157959,
+      "learning_rate": 2.8468247796956967e-06,
+      "loss": 0.8766,
+      "step": 8464
+    },
+    {
+      "epoch": 0.6525593586185631,
+      "grad_norm": 3.534771203994751,
+      "learning_rate": 2.845698090534196e-06,
+      "loss": 0.9342,
+      "step": 8465
+    },
+    {
+      "epoch": 0.65263644773358,
+      "grad_norm": 3.2332229614257812,
+      "learning_rate": 2.8445715356822605e-06,
+      "loss": 0.7536,
+      "step": 8466
+    },
+    {
+      "epoch": 0.652713536848597,
+      "grad_norm": 3.904291868209839,
+      "learning_rate": 2.8434451152101273e-06,
+      "loss": 0.9559,
+      "step": 8467
+    },
+    {
+      "epoch": 0.6527906259636139,
+      "grad_norm": 3.66575288772583,
+      "learning_rate": 2.8423188291880243e-06,
+      "loss": 1.0307,
+      "step": 8468
+    },
+    {
+      "epoch": 0.652867715078631,
+      "grad_norm": 3.674485206604004,
+      "learning_rate": 2.8411926776861653e-06,
+      "loss": 0.9073,
+      "step": 8469
+    },
+    {
+      "epoch": 0.6529448041936479,
+      "grad_norm": 3.7279632091522217,
+      "learning_rate": 2.840066660774762e-06,
+      "loss": 0.9714,
+      "step": 8470
+    },
+    {
+      "epoch": 0.6530218933086648,
+      "grad_norm": 3.9103338718414307,
+      "learning_rate": 2.838940778524015e-06,
+      "loss": 0.8716,
+      "step": 8471
+    },
+    {
+      "epoch": 0.6530989824236818,
+      "grad_norm": 3.741309404373169,
+      "learning_rate": 2.8378150310041197e-06,
+      "loss": 0.8169,
+      "step": 8472
+    },
+    {
+      "epoch": 0.6531760715386987,
+      "grad_norm": 3.7075369358062744,
+      "learning_rate": 2.83668941828526e-06,
+      "loss": 0.9703,
+      "step": 8473
+    },
+    {
+      "epoch": 0.6532531606537157,
+      "grad_norm": 4.015072822570801,
+      "learning_rate": 2.835563940437609e-06,
+      "loss": 0.975,
+      "step": 8474
+    },
+    {
+      "epoch": 0.6533302497687327,
+      "grad_norm": 4.320268630981445,
+      "learning_rate": 2.8344385975313366e-06,
+      "loss": 0.9171,
+      "step": 8475
+    },
+    {
+      "epoch": 0.6534073388837496,
+      "grad_norm": 3.988584518432617,
+      "learning_rate": 2.8333133896365993e-06,
+      "loss": 0.9586,
+      "step": 8476
+    },
+    {
+      "epoch": 0.6534844279987666,
+      "grad_norm": 3.7726614475250244,
+      "learning_rate": 2.8321883168235555e-06,
+      "loss": 0.8503,
+      "step": 8477
+    },
+    {
+      "epoch": 0.6535615171137835,
+      "grad_norm": 3.7435741424560547,
+      "learning_rate": 2.8310633791623403e-06,
+      "loss": 0.9676,
+      "step": 8478
+    },
+    {
+      "epoch": 0.6536386062288005,
+      "grad_norm": 3.893402099609375,
+      "learning_rate": 2.8299385767230904e-06,
+      "loss": 0.8198,
+      "step": 8479
+    },
+    {
+      "epoch": 0.6537156953438175,
+      "grad_norm": 3.577685594558716,
+      "learning_rate": 2.8288139095759315e-06,
+      "loss": 0.9281,
+      "step": 8480
+    },
+    {
+      "epoch": 0.6537927844588344,
+      "grad_norm": 3.464526891708374,
+      "learning_rate": 2.8276893777909796e-06,
+      "loss": 0.8223,
+      "step": 8481
+    },
+    {
+      "epoch": 0.6538698735738514,
+      "grad_norm": 3.8149683475494385,
+      "learning_rate": 2.826564981438345e-06,
+      "loss": 0.9708,
+      "step": 8482
+    },
+    {
+      "epoch": 0.6539469626888683,
+      "grad_norm": 3.6663248538970947,
+      "learning_rate": 2.825440720588129e-06,
+      "loss": 0.8898,
+      "step": 8483
+    },
+    {
+      "epoch": 0.6540240518038853,
+      "grad_norm": 3.7101478576660156,
+      "learning_rate": 2.82431659531042e-06,
+      "loss": 0.9334,
+      "step": 8484
+    },
+    {
+      "epoch": 0.6541011409189023,
+      "grad_norm": 3.7153666019439697,
+      "learning_rate": 2.823192605675303e-06,
+      "loss": 0.8837,
+      "step": 8485
+    },
+    {
+      "epoch": 0.6541782300339192,
+      "grad_norm": 3.713289260864258,
+      "learning_rate": 2.822068751752853e-06,
+      "loss": 0.8889,
+      "step": 8486
+    },
+    {
+      "epoch": 0.6542553191489362,
+      "grad_norm": 3.4118247032165527,
+      "learning_rate": 2.820945033613136e-06,
+      "loss": 0.9104,
+      "step": 8487
+    },
+    {
+      "epoch": 0.6543324082639531,
+      "grad_norm": 3.6996970176696777,
+      "learning_rate": 2.8198214513262132e-06,
+      "loss": 0.8539,
+      "step": 8488
+    },
+    {
+      "epoch": 0.6544094973789701,
+      "grad_norm": 3.8898186683654785,
+      "learning_rate": 2.8186980049621292e-06,
+      "loss": 0.9048,
+      "step": 8489
+    },
+    {
+      "epoch": 0.654486586493987,
+      "grad_norm": 3.663559675216675,
+      "learning_rate": 2.8175746945909277e-06,
+      "loss": 0.798,
+      "step": 8490
+    },
+    {
+      "epoch": 0.654563675609004,
+      "grad_norm": 4.062438488006592,
+      "learning_rate": 2.816451520282641e-06,
+      "loss": 0.8531,
+      "step": 8491
+    },
+    {
+      "epoch": 0.654640764724021,
+      "grad_norm": 3.281128168106079,
+      "learning_rate": 2.8153284821072925e-06,
+      "loss": 0.8421,
+      "step": 8492
+    },
+    {
+      "epoch": 0.6547178538390379,
+      "grad_norm": 3.5460658073425293,
+      "learning_rate": 2.814205580134899e-06,
+      "loss": 0.9501,
+      "step": 8493
+    },
+    {
+      "epoch": 0.6547949429540549,
+      "grad_norm": 3.5473179817199707,
+      "learning_rate": 2.813082814435469e-06,
+      "loss": 0.8635,
+      "step": 8494
+    },
+    {
+      "epoch": 0.6548720320690719,
+      "grad_norm": 3.5406644344329834,
+      "learning_rate": 2.8119601850789966e-06,
+      "loss": 0.8653,
+      "step": 8495
+    },
+    {
+      "epoch": 0.6549491211840888,
+      "grad_norm": 4.069020748138428,
+      "learning_rate": 2.810837692135475e-06,
+      "loss": 0.9973,
+      "step": 8496
+    },
+    {
+      "epoch": 0.6550262102991058,
+      "grad_norm": 3.7306931018829346,
+      "learning_rate": 2.809715335674885e-06,
+      "loss": 1.004,
+      "step": 8497
+    },
+    {
+      "epoch": 0.6551032994141227,
+      "grad_norm": 4.0841288566589355,
+      "learning_rate": 2.8085931157672007e-06,
+      "loss": 0.9853,
+      "step": 8498
+    },
+    {
+      "epoch": 0.6551803885291397,
+      "grad_norm": 3.389622211456299,
+      "learning_rate": 2.807471032482387e-06,
+      "loss": 0.9221,
+      "step": 8499
+    },
+    {
+      "epoch": 0.6552574776441566,
+      "grad_norm": 3.7449147701263428,
+      "learning_rate": 2.8063490858903987e-06,
+      "loss": 0.9093,
+      "step": 8500
+    },
+    {
+      "epoch": 0.6553345667591736,
+      "grad_norm": 3.7935571670532227,
+      "learning_rate": 2.805227276061182e-06,
+      "loss": 0.9671,
+      "step": 8501
+    },
+    {
+      "epoch": 0.6554116558741906,
+      "grad_norm": 3.392277717590332,
+      "learning_rate": 2.8041056030646787e-06,
+      "loss": 0.9458,
+      "step": 8502
+    },
+    {
+      "epoch": 0.6554887449892075,
+      "grad_norm": 4.1073150634765625,
+      "learning_rate": 2.8029840669708176e-06,
+      "loss": 0.9082,
+      "step": 8503
+    },
+    {
+      "epoch": 0.6555658341042245,
+      "grad_norm": 3.53625750541687,
+      "learning_rate": 2.8018626678495215e-06,
+      "loss": 0.9127,
+      "step": 8504
+    },
+    {
+      "epoch": 0.6556429232192414,
+      "grad_norm": 3.6561434268951416,
+      "learning_rate": 2.8007414057707054e-06,
+      "loss": 0.9273,
+      "step": 8505
+    },
+    {
+      "epoch": 0.6557200123342584,
+      "grad_norm": 3.346360683441162,
+      "learning_rate": 2.7996202808042706e-06,
+      "loss": 0.8784,
+      "step": 8506
+    },
+    {
+      "epoch": 0.6557971014492754,
+      "grad_norm": 3.969512462615967,
+      "learning_rate": 2.7984992930201156e-06,
+      "loss": 0.8833,
+      "step": 8507
+    },
+    {
+      "epoch": 0.6558741905642923,
+      "grad_norm": 3.466351270675659,
+      "learning_rate": 2.7973784424881273e-06,
+      "loss": 0.8353,
+      "step": 8508
+    },
+    {
+      "epoch": 0.6559512796793093,
+      "grad_norm": 3.9077634811401367,
+      "learning_rate": 2.7962577292781856e-06,
+      "loss": 0.9578,
+      "step": 8509
+    },
+    {
+      "epoch": 0.6560283687943262,
+      "grad_norm": 4.4450225830078125,
+      "learning_rate": 2.795137153460163e-06,
+      "loss": 1.0021,
+      "step": 8510
+    },
+    {
+      "epoch": 0.6561054579093432,
+      "grad_norm": 3.7983291149139404,
+      "learning_rate": 2.794016715103918e-06,
+      "loss": 0.9893,
+      "step": 8511
+    },
+    {
+      "epoch": 0.6561825470243602,
+      "grad_norm": 3.427762985229492,
+      "learning_rate": 2.7928964142793048e-06,
+      "loss": 0.8019,
+      "step": 8512
+    },
+    {
+      "epoch": 0.6562596361393771,
+      "grad_norm": 4.073552131652832,
+      "learning_rate": 2.7917762510561698e-06,
+      "loss": 0.9104,
+      "step": 8513
+    },
+    {
+      "epoch": 0.6563367252543941,
+      "grad_norm": 3.4051129817962646,
+      "learning_rate": 2.790656225504349e-06,
+      "loss": 0.8763,
+      "step": 8514
+    },
+    {
+      "epoch": 0.656413814369411,
+      "grad_norm": 3.3330910205841064,
+      "learning_rate": 2.789536337693672e-06,
+      "loss": 0.8136,
+      "step": 8515
+    },
+    {
+      "epoch": 0.656490903484428,
+      "grad_norm": 3.8815250396728516,
+      "learning_rate": 2.7884165876939537e-06,
+      "loss": 1.0265,
+      "step": 8516
+    },
+    {
+      "epoch": 0.656567992599445,
+      "grad_norm": 3.9009790420532227,
+      "learning_rate": 2.787296975575008e-06,
+      "loss": 0.9191,
+      "step": 8517
+    },
+    {
+      "epoch": 0.6566450817144619,
+      "grad_norm": 3.823681354522705,
+      "learning_rate": 2.786177501406635e-06,
+      "loss": 1.0026,
+      "step": 8518
+    },
+    {
+      "epoch": 0.6567221708294789,
+      "grad_norm": 3.889641761779785,
+      "learning_rate": 2.7850581652586296e-06,
+      "loss": 0.8824,
+      "step": 8519
+    },
+    {
+      "epoch": 0.6567992599444958,
+      "grad_norm": 3.67506742477417,
+      "learning_rate": 2.7839389672007753e-06,
+      "loss": 0.9501,
+      "step": 8520
+    },
+    {
+      "epoch": 0.6568763490595128,
+      "grad_norm": 4.033994197845459,
+      "learning_rate": 2.7828199073028504e-06,
+      "loss": 0.9688,
+      "step": 8521
+    },
+    {
+      "epoch": 0.6569534381745298,
+      "grad_norm": 3.685767889022827,
+      "learning_rate": 2.78170098563462e-06,
+      "loss": 1.0076,
+      "step": 8522
+    },
+    {
+      "epoch": 0.6570305272895467,
+      "grad_norm": 3.811617374420166,
+      "learning_rate": 2.7805822022658452e-06,
+      "loss": 0.8409,
+      "step": 8523
+    },
+    {
+      "epoch": 0.6571076164045637,
+      "grad_norm": 3.2844462394714355,
+      "learning_rate": 2.7794635572662755e-06,
+      "loss": 0.8843,
+      "step": 8524
+    },
+    {
+      "epoch": 0.6571847055195806,
+      "grad_norm": 3.5882797241210938,
+      "learning_rate": 2.778345050705654e-06,
+      "loss": 0.8763,
+      "step": 8525
+    },
+    {
+      "epoch": 0.6572617946345976,
+      "grad_norm": 4.07551383972168,
+      "learning_rate": 2.7772266826537103e-06,
+      "loss": 0.9604,
+      "step": 8526
+    },
+    {
+      "epoch": 0.6573388837496146,
+      "grad_norm": 3.5869812965393066,
+      "learning_rate": 2.7761084531801707e-06,
+      "loss": 1.0136,
+      "step": 8527
+    },
+    {
+      "epoch": 0.6574159728646315,
+      "grad_norm": 3.7623794078826904,
+      "learning_rate": 2.774990362354752e-06,
+      "loss": 0.9086,
+      "step": 8528
+    },
+    {
+      "epoch": 0.6574930619796485,
+      "grad_norm": 3.60953950881958,
+      "learning_rate": 2.7738724102471603e-06,
+      "loss": 0.9353,
+      "step": 8529
+    },
+    {
+      "epoch": 0.6575701510946654,
+      "grad_norm": 3.6601200103759766,
+      "learning_rate": 2.772754596927094e-06,
+      "loss": 0.9573,
+      "step": 8530
+    },
+    {
+      "epoch": 0.6576472402096823,
+      "grad_norm": 3.794666051864624,
+      "learning_rate": 2.771636922464246e-06,
+      "loss": 0.8832,
+      "step": 8531
+    },
+    {
+      "epoch": 0.6577243293246994,
+      "grad_norm": 3.9697749614715576,
+      "learning_rate": 2.7705193869282924e-06,
+      "loss": 1.0734,
+      "step": 8532
+    },
+    {
+      "epoch": 0.6578014184397163,
+      "grad_norm": 3.8079090118408203,
+      "learning_rate": 2.7694019903889087e-06,
+      "loss": 0.969,
+      "step": 8533
+    },
+    {
+      "epoch": 0.6578785075547333,
+      "grad_norm": 4.480514049530029,
+      "learning_rate": 2.768284732915758e-06,
+      "loss": 1.0543,
+      "step": 8534
+    },
+    {
+      "epoch": 0.6579555966697502,
+      "grad_norm": 3.4868109226226807,
+      "learning_rate": 2.767167614578496e-06,
+      "loss": 0.9419,
+      "step": 8535
+    },
+    {
+      "epoch": 0.6580326857847671,
+      "grad_norm": 3.8815977573394775,
+      "learning_rate": 2.7660506354467708e-06,
+      "loss": 0.9569,
+      "step": 8536
+    },
+    {
+      "epoch": 0.6581097748997842,
+      "grad_norm": 3.8802473545074463,
+      "learning_rate": 2.7649337955902167e-06,
+      "loss": 0.9196,
+      "step": 8537
+    },
+    {
+      "epoch": 0.6581868640148011,
+      "grad_norm": 3.837989330291748,
+      "learning_rate": 2.763817095078465e-06,
+      "loss": 0.9033,
+      "step": 8538
+    },
+    {
+      "epoch": 0.6582639531298181,
+      "grad_norm": 3.4889914989471436,
+      "learning_rate": 2.762700533981136e-06,
+      "loss": 0.8692,
+      "step": 8539
+    },
+    {
+      "epoch": 0.658341042244835,
+      "grad_norm": 3.921881914138794,
+      "learning_rate": 2.761584112367841e-06,
+      "loss": 0.9108,
+      "step": 8540
+    },
+    {
+      "epoch": 0.6584181313598519,
+      "grad_norm": 3.6214163303375244,
+      "learning_rate": 2.7604678303081833e-06,
+      "loss": 0.9357,
+      "step": 8541
+    },
+    {
+      "epoch": 0.658495220474869,
+      "grad_norm": 3.711642265319824,
+      "learning_rate": 2.7593516878717604e-06,
+      "loss": 0.8258,
+      "step": 8542
+    },
+    {
+      "epoch": 0.6585723095898859,
+      "grad_norm": 3.5625925064086914,
+      "learning_rate": 2.758235685128152e-06,
+      "loss": 0.9205,
+      "step": 8543
+    },
+    {
+      "epoch": 0.6586493987049029,
+      "grad_norm": 3.9220776557922363,
+      "learning_rate": 2.75711982214694e-06,
+      "loss": 0.9736,
+      "step": 8544
+    },
+    {
+      "epoch": 0.6587264878199198,
+      "grad_norm": 3.7191176414489746,
+      "learning_rate": 2.7560040989976894e-06,
+      "loss": 0.8053,
+      "step": 8545
+    },
+    {
+      "epoch": 0.6588035769349367,
+      "grad_norm": 3.6026618480682373,
+      "learning_rate": 2.7548885157499617e-06,
+      "loss": 0.9155,
+      "step": 8546
+    },
+    {
+      "epoch": 0.6588806660499538,
+      "grad_norm": 3.873753786087036,
+      "learning_rate": 2.7537730724733092e-06,
+      "loss": 0.9101,
+      "step": 8547
+    },
+    {
+      "epoch": 0.6589577551649707,
+      "grad_norm": 3.89041805267334,
+      "learning_rate": 2.7526577692372704e-06,
+      "loss": 0.783,
+      "step": 8548
+    },
+    {
+      "epoch": 0.6590348442799877,
+      "grad_norm": 3.7555131912231445,
+      "learning_rate": 2.75154260611138e-06,
+      "loss": 1.0063,
+      "step": 8549
+    },
+    {
+      "epoch": 0.6591119333950046,
+      "grad_norm": 3.4990248680114746,
+      "learning_rate": 2.750427583165164e-06,
+      "loss": 0.9204,
+      "step": 8550
+    },
+    {
+      "epoch": 0.6591890225100215,
+      "grad_norm": 3.4794487953186035,
+      "learning_rate": 2.7493127004681365e-06,
+      "loss": 0.9723,
+      "step": 8551
+    },
+    {
+      "epoch": 0.6592661116250386,
+      "grad_norm": 3.5538179874420166,
+      "learning_rate": 2.7481979580898056e-06,
+      "loss": 0.8747,
+      "step": 8552
+    },
+    {
+      "epoch": 0.6593432007400555,
+      "grad_norm": 3.8357620239257812,
+      "learning_rate": 2.7470833560996717e-06,
+      "loss": 1.0114,
+      "step": 8553
+    },
+    {
+      "epoch": 0.6594202898550725,
+      "grad_norm": 3.8175249099731445,
+      "learning_rate": 2.7459688945672204e-06,
+      "loss": 0.9458,
+      "step": 8554
+    },
+    {
+      "epoch": 0.6594973789700894,
+      "grad_norm": 3.8220913410186768,
+      "learning_rate": 2.744854573561935e-06,
+      "loss": 0.9557,
+      "step": 8555
+    },
+    {
+      "epoch": 0.6595744680851063,
+      "grad_norm": 4.2890944480896,
+      "learning_rate": 2.7437403931532867e-06,
+      "loss": 1.043,
+      "step": 8556
+    },
+    {
+      "epoch": 0.6596515572001234,
+      "grad_norm": 4.034401893615723,
+      "learning_rate": 2.7426263534107394e-06,
+      "loss": 0.9287,
+      "step": 8557
+    },
+    {
+      "epoch": 0.6597286463151403,
+      "grad_norm": 3.621591091156006,
+      "learning_rate": 2.7415124544037497e-06,
+      "loss": 0.968,
+      "step": 8558
+    },
+    {
+      "epoch": 0.6598057354301573,
+      "grad_norm": 3.7988715171813965,
+      "learning_rate": 2.74039869620176e-06,
+      "loss": 0.9679,
+      "step": 8559
+    },
+    {
+      "epoch": 0.6598828245451742,
+      "grad_norm": 3.6728246212005615,
+      "learning_rate": 2.7392850788742088e-06,
+      "loss": 0.8546,
+      "step": 8560
+    },
+    {
+      "epoch": 0.6599599136601911,
+      "grad_norm": 3.391083002090454,
+      "learning_rate": 2.7381716024905236e-06,
+      "loss": 0.8998,
+      "step": 8561
+    },
+    {
+      "epoch": 0.6600370027752082,
+      "grad_norm": 3.6413748264312744,
+      "learning_rate": 2.7370582671201253e-06,
+      "loss": 1.021,
+      "step": 8562
+    },
+    {
+      "epoch": 0.6601140918902251,
+      "grad_norm": 3.9035537242889404,
+      "learning_rate": 2.7359450728324254e-06,
+      "loss": 0.9325,
+      "step": 8563
+    },
+    {
+      "epoch": 0.6601911810052421,
+      "grad_norm": 3.603443145751953,
+      "learning_rate": 2.734832019696821e-06,
+      "loss": 0.943,
+      "step": 8564
+    },
+    {
+      "epoch": 0.660268270120259,
+      "grad_norm": 4.327446937561035,
+      "learning_rate": 2.7337191077827107e-06,
+      "loss": 0.94,
+      "step": 8565
+    },
+    {
+      "epoch": 0.6603453592352759,
+      "grad_norm": 3.7741518020629883,
+      "learning_rate": 2.732606337159476e-06,
+      "loss": 0.8859,
+      "step": 8566
+    },
+    {
+      "epoch": 0.660422448350293,
+      "grad_norm": 3.900160074234009,
+      "learning_rate": 2.731493707896493e-06,
+      "loss": 0.8494,
+      "step": 8567
+    },
+    {
+      "epoch": 0.6604995374653099,
+      "grad_norm": 4.106858253479004,
+      "learning_rate": 2.730381220063131e-06,
+      "loss": 0.9732,
+      "step": 8568
+    },
+    {
+      "epoch": 0.6605766265803269,
+      "grad_norm": 3.8731374740600586,
+      "learning_rate": 2.729268873728743e-06,
+      "loss": 0.9365,
+      "step": 8569
+    },
+    {
+      "epoch": 0.6606537156953438,
+      "grad_norm": 4.0173659324646,
+      "learning_rate": 2.7281566689626804e-06,
+      "loss": 0.9892,
+      "step": 8570
+    },
+    {
+      "epoch": 0.6607308048103607,
+      "grad_norm": 3.8889834880828857,
+      "learning_rate": 2.7270446058342836e-06,
+      "loss": 0.9164,
+      "step": 8571
+    },
+    {
+      "epoch": 0.6608078939253778,
+      "grad_norm": 4.212670803070068,
+      "learning_rate": 2.7259326844128832e-06,
+      "loss": 1.0294,
+      "step": 8572
+    },
+    {
+      "epoch": 0.6608849830403947,
+      "grad_norm": 4.06730318069458,
+      "learning_rate": 2.724820904767804e-06,
+      "loss": 0.9905,
+      "step": 8573
+    },
+    {
+      "epoch": 0.6609620721554117,
+      "grad_norm": 3.9518141746520996,
+      "learning_rate": 2.7237092669683562e-06,
+      "loss": 0.9677,
+      "step": 8574
+    },
+    {
+      "epoch": 0.6610391612704286,
+      "grad_norm": 3.7986531257629395,
+      "learning_rate": 2.722597771083847e-06,
+      "loss": 0.8483,
+      "step": 8575
+    },
+    {
+      "epoch": 0.6611162503854455,
+      "grad_norm": 4.189377784729004,
+      "learning_rate": 2.721486417183571e-06,
+      "loss": 1.0454,
+      "step": 8576
+    },
+    {
+      "epoch": 0.6611933395004626,
+      "grad_norm": 3.7273128032684326,
+      "learning_rate": 2.7203752053368167e-06,
+      "loss": 0.928,
+      "step": 8577
+    },
+    {
+      "epoch": 0.6612704286154795,
+      "grad_norm": 3.6438515186309814,
+      "learning_rate": 2.7192641356128613e-06,
+      "loss": 0.9048,
+      "step": 8578
+    },
+    {
+      "epoch": 0.6613475177304965,
+      "grad_norm": 3.6332359313964844,
+      "learning_rate": 2.718153208080977e-06,
+      "loss": 0.9138,
+      "step": 8579
+    },
+    {
+      "epoch": 0.6614246068455134,
+      "grad_norm": 3.474804162979126,
+      "learning_rate": 2.7170424228104207e-06,
+      "loss": 0.8285,
+      "step": 8580
+    },
+    {
+      "epoch": 0.6615016959605303,
+      "grad_norm": 3.8621301651000977,
+      "learning_rate": 2.7159317798704452e-06,
+      "loss": 0.9538,
+      "step": 8581
+    },
+    {
+      "epoch": 0.6615787850755473,
+      "grad_norm": 4.121014595031738,
+      "learning_rate": 2.7148212793302938e-06,
+      "loss": 0.9663,
+      "step": 8582
+    },
+    {
+      "epoch": 0.6616558741905643,
+      "grad_norm": 3.514923572540283,
+      "learning_rate": 2.7137109212592006e-06,
+      "loss": 0.8936,
+      "step": 8583
+    },
+    {
+      "epoch": 0.6617329633055813,
+      "grad_norm": 3.3887276649475098,
+      "learning_rate": 2.712600705726392e-06,
+      "loss": 0.9403,
+      "step": 8584
+    },
+    {
+      "epoch": 0.6618100524205982,
+      "grad_norm": 3.4407477378845215,
+      "learning_rate": 2.7114906328010803e-06,
+      "loss": 0.9276,
+      "step": 8585
+    },
+    {
+      "epoch": 0.6618871415356151,
+      "grad_norm": 3.4507927894592285,
+      "learning_rate": 2.7103807025524764e-06,
+      "loss": 0.7437,
+      "step": 8586
+    },
+    {
+      "epoch": 0.6619642306506321,
+      "grad_norm": 3.491410255432129,
+      "learning_rate": 2.7092709150497764e-06,
+      "loss": 0.8576,
+      "step": 8587
+    },
+    {
+      "epoch": 0.6620413197656491,
+      "grad_norm": 3.461198091506958,
+      "learning_rate": 2.708161270362171e-06,
+      "loss": 0.8797,
+      "step": 8588
+    },
+    {
+      "epoch": 0.6621184088806661,
+      "grad_norm": 3.8753464221954346,
+      "learning_rate": 2.7070517685588417e-06,
+      "loss": 0.8896,
+      "step": 8589
+    },
+    {
+      "epoch": 0.662195497995683,
+      "grad_norm": 3.9158358573913574,
+      "learning_rate": 2.70594240970896e-06,
+      "loss": 0.8949,
+      "step": 8590
+    },
+    {
+      "epoch": 0.6622725871106999,
+      "grad_norm": 12.948868751525879,
+      "learning_rate": 2.7048331938816863e-06,
+      "loss": 0.8589,
+      "step": 8591
+    },
+    {
+      "epoch": 0.6623496762257169,
+      "grad_norm": 3.7209019660949707,
+      "learning_rate": 2.7037241211461752e-06,
+      "loss": 0.9651,
+      "step": 8592
+    },
+    {
+      "epoch": 0.6624267653407339,
+      "grad_norm": 3.9368491172790527,
+      "learning_rate": 2.7026151915715735e-06,
+      "loss": 0.9401,
+      "step": 8593
+    },
+    {
+      "epoch": 0.6625038544557509,
+      "grad_norm": 3.6502487659454346,
+      "learning_rate": 2.7015064052270156e-06,
+      "loss": 0.9531,
+      "step": 8594
+    },
+    {
+      "epoch": 0.6625809435707678,
+      "grad_norm": 3.4706952571868896,
+      "learning_rate": 2.700397762181631e-06,
+      "loss": 0.8791,
+      "step": 8595
+    },
+    {
+      "epoch": 0.6626580326857847,
+      "grad_norm": 3.8239970207214355,
+      "learning_rate": 2.6992892625045343e-06,
+      "loss": 1.0303,
+      "step": 8596
+    },
+    {
+      "epoch": 0.6627351218008017,
+      "grad_norm": 3.561873435974121,
+      "learning_rate": 2.698180906264837e-06,
+      "loss": 0.87,
+      "step": 8597
+    },
+    {
+      "epoch": 0.6628122109158187,
+      "grad_norm": 3.5032331943511963,
+      "learning_rate": 2.697072693531637e-06,
+      "loss": 0.9174,
+      "step": 8598
+    },
+    {
+      "epoch": 0.6628893000308357,
+      "grad_norm": 3.578794479370117,
+      "learning_rate": 2.695964624374029e-06,
+      "loss": 0.926,
+      "step": 8599
+    },
+    {
+      "epoch": 0.6629663891458526,
+      "grad_norm": 3.583988904953003,
+      "learning_rate": 2.6948566988610938e-06,
+      "loss": 0.8603,
+      "step": 8600
+    },
+    {
+      "epoch": 0.6630434782608695,
+      "grad_norm": 3.555171489715576,
+      "learning_rate": 2.693748917061906e-06,
+      "loss": 0.8882,
+      "step": 8601
+    },
+    {
+      "epoch": 0.6631205673758865,
+      "grad_norm": 3.73492431640625,
+      "learning_rate": 2.692641279045527e-06,
+      "loss": 0.961,
+      "step": 8602
+    },
+    {
+      "epoch": 0.6631976564909035,
+      "grad_norm": 3.6243600845336914,
+      "learning_rate": 2.691533784881014e-06,
+      "loss": 0.8199,
+      "step": 8603
+    },
+    {
+      "epoch": 0.6632747456059205,
+      "grad_norm": 3.813415288925171,
+      "learning_rate": 2.6904264346374135e-06,
+      "loss": 0.8965,
+      "step": 8604
+    },
+    {
+      "epoch": 0.6633518347209374,
+      "grad_norm": 3.7679946422576904,
+      "learning_rate": 2.6893192283837636e-06,
+      "loss": 0.9742,
+      "step": 8605
+    },
+    {
+      "epoch": 0.6634289238359543,
+      "grad_norm": 3.7962825298309326,
+      "learning_rate": 2.688212166189095e-06,
+      "loss": 0.9346,
+      "step": 8606
+    },
+    {
+      "epoch": 0.6635060129509713,
+      "grad_norm": 3.616288900375366,
+      "learning_rate": 2.6871052481224217e-06,
+      "loss": 0.9487,
+      "step": 8607
+    },
+    {
+      "epoch": 0.6635831020659883,
+      "grad_norm": 3.7229197025299072,
+      "learning_rate": 2.685998474252758e-06,
+      "loss": 0.8765,
+      "step": 8608
+    },
+    {
+      "epoch": 0.6636601911810053,
+      "grad_norm": 3.9187211990356445,
+      "learning_rate": 2.684891844649103e-06,
+      "loss": 0.9206,
+      "step": 8609
+    },
+    {
+      "epoch": 0.6637372802960222,
+      "grad_norm": 3.438185453414917,
+      "learning_rate": 2.6837853593804554e-06,
+      "loss": 0.871,
+      "step": 8610
+    },
+    {
+      "epoch": 0.6638143694110391,
+      "grad_norm": 3.9521586894989014,
+      "learning_rate": 2.682679018515792e-06,
+      "loss": 0.9515,
+      "step": 8611
+    },
+    {
+      "epoch": 0.6638914585260561,
+      "grad_norm": 3.899041175842285,
+      "learning_rate": 2.68157282212409e-06,
+      "loss": 0.9129,
+      "step": 8612
+    },
+    {
+      "epoch": 0.663968547641073,
+      "grad_norm": 3.531867265701294,
+      "learning_rate": 2.680466770274316e-06,
+      "loss": 0.8427,
+      "step": 8613
+    },
+    {
+      "epoch": 0.6640456367560901,
+      "grad_norm": 3.4896748065948486,
+      "learning_rate": 2.679360863035425e-06,
+      "loss": 0.7873,
+      "step": 8614
+    },
+    {
+      "epoch": 0.664122725871107,
+      "grad_norm": 3.5182361602783203,
+      "learning_rate": 2.678255100476366e-06,
+      "loss": 0.8737,
+      "step": 8615
+    },
+    {
+      "epoch": 0.6641998149861239,
+      "grad_norm": 3.8621270656585693,
+      "learning_rate": 2.6771494826660782e-06,
+      "loss": 0.9002,
+      "step": 8616
+    },
+    {
+      "epoch": 0.6642769041011409,
+      "grad_norm": 3.783743143081665,
+      "learning_rate": 2.6760440096734875e-06,
+      "loss": 0.9574,
+      "step": 8617
+    },
+    {
+      "epoch": 0.6643539932161578,
+      "grad_norm": 3.6670119762420654,
+      "learning_rate": 2.674938681567517e-06,
+      "loss": 0.824,
+      "step": 8618
+    },
+    {
+      "epoch": 0.6644310823311749,
+      "grad_norm": 3.6787564754486084,
+      "learning_rate": 2.6738334984170785e-06,
+      "loss": 0.9541,
+      "step": 8619
+    },
+    {
+      "epoch": 0.6645081714461918,
+      "grad_norm": 3.8661928176879883,
+      "learning_rate": 2.672728460291073e-06,
+      "loss": 0.9713,
+      "step": 8620
+    },
+    {
+      "epoch": 0.6645852605612088,
+      "grad_norm": 5.067914009094238,
+      "learning_rate": 2.671623567258398e-06,
+      "loss": 1.0446,
+      "step": 8621
+    },
+    {
+      "epoch": 0.6646623496762257,
+      "grad_norm": 3.7516796588897705,
+      "learning_rate": 2.6705188193879316e-06,
+      "loss": 0.8746,
+      "step": 8622
+    },
+    {
+      "epoch": 0.6647394387912426,
+      "grad_norm": 3.7899248600006104,
+      "learning_rate": 2.669414216748552e-06,
+      "loss": 0.9415,
+      "step": 8623
+    },
+    {
+      "epoch": 0.6648165279062597,
+      "grad_norm": 3.813930034637451,
+      "learning_rate": 2.6683097594091257e-06,
+      "loss": 1.0627,
+      "step": 8624
+    },
+    {
+      "epoch": 0.6648936170212766,
+      "grad_norm": 4.484930992126465,
+      "learning_rate": 2.6672054474385102e-06,
+      "loss": 1.0192,
+      "step": 8625
+    },
+    {
+      "epoch": 0.6649707061362936,
+      "grad_norm": 3.5336852073669434,
+      "learning_rate": 2.666101280905553e-06,
+      "loss": 0.8844,
+      "step": 8626
+    },
+    {
+      "epoch": 0.6650477952513105,
+      "grad_norm": 3.831902503967285,
+      "learning_rate": 2.664997259879095e-06,
+      "loss": 0.8492,
+      "step": 8627
+    },
+    {
+      "epoch": 0.6651248843663274,
+      "grad_norm": 3.4217164516448975,
+      "learning_rate": 2.663893384427963e-06,
+      "loss": 0.9379,
+      "step": 8628
+    },
+    {
+      "epoch": 0.6652019734813445,
+      "grad_norm": 3.7807199954986572,
+      "learning_rate": 2.66278965462098e-06,
+      "loss": 0.9947,
+      "step": 8629
+    },
+    {
+      "epoch": 0.6652790625963614,
+      "grad_norm": 4.046280384063721,
+      "learning_rate": 2.661686070526956e-06,
+      "loss": 0.9283,
+      "step": 8630
+    },
+    {
+      "epoch": 0.6653561517113784,
+      "grad_norm": 3.804257392883301,
+      "learning_rate": 2.6605826322146954e-06,
+      "loss": 1.0612,
+      "step": 8631
+    },
+    {
+      "epoch": 0.6654332408263953,
+      "grad_norm": 3.5956528186798096,
+      "learning_rate": 2.659479339752994e-06,
+      "loss": 0.8869,
+      "step": 8632
+    },
+    {
+      "epoch": 0.6655103299414122,
+      "grad_norm": 3.916138172149658,
+      "learning_rate": 2.6583761932106323e-06,
+      "loss": 0.8914,
+      "step": 8633
+    },
+    {
+      "epoch": 0.6655874190564293,
+      "grad_norm": 4.105411052703857,
+      "learning_rate": 2.6572731926563867e-06,
+      "loss": 0.961,
+      "step": 8634
+    },
+    {
+      "epoch": 0.6656645081714462,
+      "grad_norm": 3.324625015258789,
+      "learning_rate": 2.6561703381590244e-06,
+      "loss": 0.8609,
+      "step": 8635
+    },
+    {
+      "epoch": 0.6657415972864632,
+      "grad_norm": 3.7240700721740723,
+      "learning_rate": 2.6550676297873023e-06,
+      "loss": 0.8107,
+      "step": 8636
+    },
+    {
+      "epoch": 0.6658186864014801,
+      "grad_norm": 3.6989731788635254,
+      "learning_rate": 2.6539650676099687e-06,
+      "loss": 0.8776,
+      "step": 8637
+    },
+    {
+      "epoch": 0.665895775516497,
+      "grad_norm": 3.833456039428711,
+      "learning_rate": 2.652862651695765e-06,
+      "loss": 0.9177,
+      "step": 8638
+    },
+    {
+      "epoch": 0.6659728646315141,
+      "grad_norm": 3.728043794631958,
+      "learning_rate": 2.651760382113417e-06,
+      "loss": 0.9688,
+      "step": 8639
+    },
+    {
+      "epoch": 0.666049953746531,
+      "grad_norm": 3.813891649246216,
+      "learning_rate": 2.6506582589316463e-06,
+      "loss": 0.9131,
+      "step": 8640
+    },
+    {
+      "epoch": 0.666127042861548,
+      "grad_norm": 3.5366103649139404,
+      "learning_rate": 2.649556282219167e-06,
+      "loss": 0.9076,
+      "step": 8641
+    },
+    {
+      "epoch": 0.6662041319765649,
+      "grad_norm": 3.920154333114624,
+      "learning_rate": 2.64845445204468e-06,
+      "loss": 0.9324,
+      "step": 8642
+    },
+    {
+      "epoch": 0.6662812210915818,
+      "grad_norm": 4.236634731292725,
+      "learning_rate": 2.64735276847688e-06,
+      "loss": 0.8984,
+      "step": 8643
+    },
+    {
+      "epoch": 0.6663583102065989,
+      "grad_norm": 3.766432762145996,
+      "learning_rate": 2.646251231584449e-06,
+      "loss": 0.9886,
+      "step": 8644
+    },
+    {
+      "epoch": 0.6664353993216158,
+      "grad_norm": 3.6179211139678955,
+      "learning_rate": 2.6451498414360633e-06,
+      "loss": 0.7535,
+      "step": 8645
+    },
+    {
+      "epoch": 0.6665124884366328,
+      "grad_norm": 3.5795180797576904,
+      "learning_rate": 2.644048598100388e-06,
+      "loss": 0.8096,
+      "step": 8646
+    },
+    {
+      "epoch": 0.6665895775516497,
+      "grad_norm": 3.80401873588562,
+      "learning_rate": 2.642947501646082e-06,
+      "loss": 0.9594,
+      "step": 8647
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 3.3994295597076416,
+      "learning_rate": 2.6418465521417914e-06,
+      "loss": 0.7756,
+      "step": 8648
+    },
+    {
+      "epoch": 0.6667437557816837,
+      "grad_norm": 3.6575098037719727,
+      "learning_rate": 2.6407457496561573e-06,
+      "loss": 0.9767,
+      "step": 8649
+    },
+    {
+      "epoch": 0.6668208448967006,
+      "grad_norm": 3.2171177864074707,
+      "learning_rate": 2.6396450942578046e-06,
+      "loss": 0.8447,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6668979340117176,
+      "grad_norm": 3.508512496948242,
+      "learning_rate": 2.638544586015356e-06,
+      "loss": 0.9398,
+      "step": 8651
+    },
+    {
+      "epoch": 0.6669750231267345,
+      "grad_norm": 3.8532354831695557,
+      "learning_rate": 2.6374442249974214e-06,
+      "loss": 0.988,
+      "step": 8652
+    },
+    {
+      "epoch": 0.6670521122417514,
+      "grad_norm": 4.1069416999816895,
+      "learning_rate": 2.6363440112726037e-06,
+      "loss": 0.9103,
+      "step": 8653
+    },
+    {
+      "epoch": 0.6671292013567685,
+      "grad_norm": 3.5245847702026367,
+      "learning_rate": 2.6352439449094952e-06,
+      "loss": 0.9636,
+      "step": 8654
+    },
+    {
+      "epoch": 0.6672062904717854,
+      "grad_norm": 3.4307544231414795,
+      "learning_rate": 2.6341440259766792e-06,
+      "loss": 0.8636,
+      "step": 8655
+    },
+    {
+      "epoch": 0.6672833795868024,
+      "grad_norm": 3.6783254146575928,
+      "learning_rate": 2.6330442545427303e-06,
+      "loss": 0.8899,
+      "step": 8656
+    },
+    {
+      "epoch": 0.6673604687018193,
+      "grad_norm": 3.7773871421813965,
+      "learning_rate": 2.6319446306762136e-06,
+      "loss": 0.9547,
+      "step": 8657
+    },
+    {
+      "epoch": 0.6674375578168362,
+      "grad_norm": 3.804189920425415,
+      "learning_rate": 2.6308451544456863e-06,
+      "loss": 0.9195,
+      "step": 8658
+    },
+    {
+      "epoch": 0.6675146469318533,
+      "grad_norm": 3.798102855682373,
+      "learning_rate": 2.6297458259196915e-06,
+      "loss": 0.9327,
+      "step": 8659
+    },
+    {
+      "epoch": 0.6675917360468702,
+      "grad_norm": 4.101744174957275,
+      "learning_rate": 2.628646645166769e-06,
+      "loss": 1.0034,
+      "step": 8660
+    },
+    {
+      "epoch": 0.6676688251618872,
+      "grad_norm": 3.816171884536743,
+      "learning_rate": 2.6275476122554467e-06,
+      "loss": 1.0159,
+      "step": 8661
+    },
+    {
+      "epoch": 0.6677459142769041,
+      "grad_norm": 3.46830677986145,
+      "learning_rate": 2.6264487272542427e-06,
+      "loss": 0.9187,
+      "step": 8662
+    },
+    {
+      "epoch": 0.667823003391921,
+      "grad_norm": 3.302978515625,
+      "learning_rate": 2.625349990231668e-06,
+      "loss": 0.942,
+      "step": 8663
+    },
+    {
+      "epoch": 0.667900092506938,
+      "grad_norm": 3.929262399673462,
+      "learning_rate": 2.624251401256225e-06,
+      "loss": 0.986,
+      "step": 8664
+    },
+    {
+      "epoch": 0.667977181621955,
+      "grad_norm": 4.437450408935547,
+      "learning_rate": 2.623152960396401e-06,
+      "loss": 0.8384,
+      "step": 8665
+    },
+    {
+      "epoch": 0.668054270736972,
+      "grad_norm": 3.7824137210845947,
+      "learning_rate": 2.622054667720679e-06,
+      "loss": 0.9588,
+      "step": 8666
+    },
+    {
+      "epoch": 0.6681313598519889,
+      "grad_norm": 3.94724702835083,
+      "learning_rate": 2.6209565232975342e-06,
+      "loss": 0.875,
+      "step": 8667
+    },
+    {
+      "epoch": 0.6682084489670058,
+      "grad_norm": 3.6828274726867676,
+      "learning_rate": 2.619858527195428e-06,
+      "loss": 0.914,
+      "step": 8668
+    },
+    {
+      "epoch": 0.6682855380820228,
+      "grad_norm": 3.615562677383423,
+      "learning_rate": 2.618760679482817e-06,
+      "loss": 0.9406,
+      "step": 8669
+    },
+    {
+      "epoch": 0.6683626271970398,
+      "grad_norm": 3.8135929107666016,
+      "learning_rate": 2.617662980228144e-06,
+      "loss": 0.8714,
+      "step": 8670
+    },
+    {
+      "epoch": 0.6684397163120568,
+      "grad_norm": 3.3155691623687744,
+      "learning_rate": 2.6165654294998442e-06,
+      "loss": 0.7839,
+      "step": 8671
+    },
+    {
+      "epoch": 0.6685168054270737,
+      "grad_norm": 3.6126980781555176,
+      "learning_rate": 2.6154680273663468e-06,
+      "loss": 0.8381,
+      "step": 8672
+    },
+    {
+      "epoch": 0.6685938945420906,
+      "grad_norm": 4.058097839355469,
+      "learning_rate": 2.614370773896069e-06,
+      "loss": 1.006,
+      "step": 8673
+    },
+    {
+      "epoch": 0.6686709836571076,
+      "grad_norm": 3.903541088104248,
+      "learning_rate": 2.613273669157417e-06,
+      "loss": 0.8644,
+      "step": 8674
+    },
+    {
+      "epoch": 0.6687480727721246,
+      "grad_norm": 3.7401607036590576,
+      "learning_rate": 2.612176713218792e-06,
+      "loss": 0.9988,
+      "step": 8675
+    },
+    {
+      "epoch": 0.6688251618871416,
+      "grad_norm": 3.917459011077881,
+      "learning_rate": 2.61107990614858e-06,
+      "loss": 0.9585,
+      "step": 8676
+    },
+    {
+      "epoch": 0.6689022510021585,
+      "grad_norm": 4.191850185394287,
+      "learning_rate": 2.6099832480151642e-06,
+      "loss": 0.931,
+      "step": 8677
+    },
+    {
+      "epoch": 0.6689793401171754,
+      "grad_norm": 3.9755399227142334,
+      "learning_rate": 2.6088867388869143e-06,
+      "loss": 1.0736,
+      "step": 8678
+    },
+    {
+      "epoch": 0.6690564292321924,
+      "grad_norm": 3.440880298614502,
+      "learning_rate": 2.6077903788321924e-06,
+      "loss": 1.0059,
+      "step": 8679
+    },
+    {
+      "epoch": 0.6691335183472094,
+      "grad_norm": 3.5236377716064453,
+      "learning_rate": 2.606694167919353e-06,
+      "loss": 0.9046,
+      "step": 8680
+    },
+    {
+      "epoch": 0.6692106074622264,
+      "grad_norm": 3.7371933460235596,
+      "learning_rate": 2.6055981062167348e-06,
+      "loss": 0.9506,
+      "step": 8681
+    },
+    {
+      "epoch": 0.6692876965772433,
+      "grad_norm": 3.5741524696350098,
+      "learning_rate": 2.6045021937926738e-06,
+      "loss": 0.8337,
+      "step": 8682
+    },
+    {
+      "epoch": 0.6693647856922602,
+      "grad_norm": 3.534313678741455,
+      "learning_rate": 2.6034064307154944e-06,
+      "loss": 0.8608,
+      "step": 8683
+    },
+    {
+      "epoch": 0.6694418748072772,
+      "grad_norm": 4.026943206787109,
+      "learning_rate": 2.6023108170535115e-06,
+      "loss": 0.909,
+      "step": 8684
+    },
+    {
+      "epoch": 0.6695189639222942,
+      "grad_norm": 3.7044715881347656,
+      "learning_rate": 2.6012153528750316e-06,
+      "loss": 0.9678,
+      "step": 8685
+    },
+    {
+      "epoch": 0.6695960530373112,
+      "grad_norm": 3.3647007942199707,
+      "learning_rate": 2.6001200382483525e-06,
+      "loss": 0.7775,
+      "step": 8686
+    },
+    {
+      "epoch": 0.6696731421523281,
+      "grad_norm": 3.695740222930908,
+      "learning_rate": 2.599024873241758e-06,
+      "loss": 0.9464,
+      "step": 8687
+    },
+    {
+      "epoch": 0.669750231267345,
+      "grad_norm": 3.711057662963867,
+      "learning_rate": 2.5979298579235276e-06,
+      "loss": 0.9555,
+      "step": 8688
+    },
+    {
+      "epoch": 0.669827320382362,
+      "grad_norm": 3.6604957580566406,
+      "learning_rate": 2.5968349923619297e-06,
+      "loss": 0.9571,
+      "step": 8689
+    },
+    {
+      "epoch": 0.669904409497379,
+      "grad_norm": 3.690217971801758,
+      "learning_rate": 2.5957402766252238e-06,
+      "loss": 0.8819,
+      "step": 8690
+    },
+    {
+      "epoch": 0.669981498612396,
+      "grad_norm": 3.983525514602661,
+      "learning_rate": 2.594645710781661e-06,
+      "loss": 0.9634,
+      "step": 8691
+    },
+    {
+      "epoch": 0.6700585877274129,
+      "grad_norm": 3.4786341190338135,
+      "learning_rate": 2.59355129489948e-06,
+      "loss": 0.9043,
+      "step": 8692
+    },
+    {
+      "epoch": 0.6701356768424298,
+      "grad_norm": 3.7870969772338867,
+      "learning_rate": 2.592457029046912e-06,
+      "loss": 0.8991,
+      "step": 8693
+    },
+    {
+      "epoch": 0.6702127659574468,
+      "grad_norm": 3.683763027191162,
+      "learning_rate": 2.5913629132921784e-06,
+      "loss": 0.8989,
+      "step": 8694
+    },
+    {
+      "epoch": 0.6702898550724637,
+      "grad_norm": 3.569939136505127,
+      "learning_rate": 2.5902689477034936e-06,
+      "loss": 0.9178,
+      "step": 8695
+    },
+    {
+      "epoch": 0.6703669441874808,
+      "grad_norm": 3.9141666889190674,
+      "learning_rate": 2.589175132349059e-06,
+      "loss": 0.9937,
+      "step": 8696
+    },
+    {
+      "epoch": 0.6704440333024977,
+      "grad_norm": 3.4919772148132324,
+      "learning_rate": 2.588081467297069e-06,
+      "loss": 0.9049,
+      "step": 8697
+    },
+    {
+      "epoch": 0.6705211224175146,
+      "grad_norm": 3.8324713706970215,
+      "learning_rate": 2.5869879526157073e-06,
+      "loss": 1.0839,
+      "step": 8698
+    },
+    {
+      "epoch": 0.6705982115325316,
+      "grad_norm": 4.001489639282227,
+      "learning_rate": 2.58589458837315e-06,
+      "loss": 0.9274,
+      "step": 8699
+    },
+    {
+      "epoch": 0.6706753006475485,
+      "grad_norm": 4.185301780700684,
+      "learning_rate": 2.584801374637562e-06,
+      "loss": 0.9855,
+      "step": 8700
+    },
+    {
+      "epoch": 0.6707523897625656,
+      "grad_norm": 3.5428168773651123,
+      "learning_rate": 2.583708311477102e-06,
+      "loss": 0.9486,
+      "step": 8701
+    },
+    {
+      "epoch": 0.6708294788775825,
+      "grad_norm": 3.9238369464874268,
+      "learning_rate": 2.582615398959912e-06,
+      "loss": 0.9111,
+      "step": 8702
+    },
+    {
+      "epoch": 0.6709065679925994,
+      "grad_norm": 4.060305595397949,
+      "learning_rate": 2.5815226371541325e-06,
+      "loss": 0.9043,
+      "step": 8703
+    },
+    {
+      "epoch": 0.6709836571076164,
+      "grad_norm": 3.543778896331787,
+      "learning_rate": 2.5804300261278903e-06,
+      "loss": 0.8829,
+      "step": 8704
+    },
+    {
+      "epoch": 0.6710607462226333,
+      "grad_norm": 3.358534336090088,
+      "learning_rate": 2.579337565949305e-06,
+      "loss": 0.9352,
+      "step": 8705
+    },
+    {
+      "epoch": 0.6711378353376504,
+      "grad_norm": 4.072521686553955,
+      "learning_rate": 2.578245256686488e-06,
+      "loss": 0.9195,
+      "step": 8706
+    },
+    {
+      "epoch": 0.6712149244526673,
+      "grad_norm": 3.970937728881836,
+      "learning_rate": 2.577153098407534e-06,
+      "loss": 0.878,
+      "step": 8707
+    },
+    {
+      "epoch": 0.6712920135676842,
+      "grad_norm": 3.6512842178344727,
+      "learning_rate": 2.576061091180536e-06,
+      "loss": 0.8582,
+      "step": 8708
+    },
+    {
+      "epoch": 0.6713691026827012,
+      "grad_norm": 4.31977653503418,
+      "learning_rate": 2.5749692350735756e-06,
+      "loss": 0.8524,
+      "step": 8709
+    },
+    {
+      "epoch": 0.6714461917977181,
+      "grad_norm": 3.4686012268066406,
+      "learning_rate": 2.5738775301547238e-06,
+      "loss": 0.8528,
+      "step": 8710
+    },
+    {
+      "epoch": 0.6715232809127352,
+      "grad_norm": 3.668100595474243,
+      "learning_rate": 2.572785976492043e-06,
+      "loss": 0.8513,
+      "step": 8711
+    },
+    {
+      "epoch": 0.6716003700277521,
+      "grad_norm": 3.792353868484497,
+      "learning_rate": 2.5716945741535864e-06,
+      "loss": 0.7951,
+      "step": 8712
+    },
+    {
+      "epoch": 0.671677459142769,
+      "grad_norm": 3.5469858646392822,
+      "learning_rate": 2.570603323207396e-06,
+      "loss": 0.868,
+      "step": 8713
+    },
+    {
+      "epoch": 0.671754548257786,
+      "grad_norm": 3.5623743534088135,
+      "learning_rate": 2.5695122237215053e-06,
+      "loss": 0.7898,
+      "step": 8714
+    },
+    {
+      "epoch": 0.6718316373728029,
+      "grad_norm": 5.158936023712158,
+      "learning_rate": 2.56842127576394e-06,
+      "loss": 0.8926,
+      "step": 8715
+    },
+    {
+      "epoch": 0.67190872648782,
+      "grad_norm": 3.4081263542175293,
+      "learning_rate": 2.5673304794027143e-06,
+      "loss": 0.8877,
+      "step": 8716
+    },
+    {
+      "epoch": 0.6719858156028369,
+      "grad_norm": 3.6705238819122314,
+      "learning_rate": 2.566239834705837e-06,
+      "loss": 0.8883,
+      "step": 8717
+    },
+    {
+      "epoch": 0.6720629047178538,
+      "grad_norm": 3.546715497970581,
+      "learning_rate": 2.5651493417412986e-06,
+      "loss": 0.7861,
+      "step": 8718
+    },
+    {
+      "epoch": 0.6721399938328708,
+      "grad_norm": 3.700578212738037,
+      "learning_rate": 2.5640590005770883e-06,
+      "loss": 0.867,
+      "step": 8719
+    },
+    {
+      "epoch": 0.6722170829478877,
+      "grad_norm": 4.1538472175598145,
+      "learning_rate": 2.5629688112811835e-06,
+      "loss": 0.8547,
+      "step": 8720
+    },
+    {
+      "epoch": 0.6722941720629048,
+      "grad_norm": 3.4912803173065186,
+      "learning_rate": 2.561878773921551e-06,
+      "loss": 0.7669,
+      "step": 8721
+    },
+    {
+      "epoch": 0.6723712611779217,
+      "grad_norm": 3.5746119022369385,
+      "learning_rate": 2.56078888856615e-06,
+      "loss": 0.87,
+      "step": 8722
+    },
+    {
+      "epoch": 0.6724483502929386,
+      "grad_norm": 4.2662811279296875,
+      "learning_rate": 2.559699155282931e-06,
+      "loss": 0.9485,
+      "step": 8723
+    },
+    {
+      "epoch": 0.6725254394079556,
+      "grad_norm": 4.090740203857422,
+      "learning_rate": 2.558609574139829e-06,
+      "loss": 0.8954,
+      "step": 8724
+    },
+    {
+      "epoch": 0.6726025285229725,
+      "grad_norm": 3.631227970123291,
+      "learning_rate": 2.557520145204776e-06,
+      "loss": 0.9711,
+      "step": 8725
+    },
+    {
+      "epoch": 0.6726796176379896,
+      "grad_norm": 4.057615756988525,
+      "learning_rate": 2.5564308685456916e-06,
+      "loss": 1.0421,
+      "step": 8726
+    },
+    {
+      "epoch": 0.6727567067530065,
+      "grad_norm": 3.319061040878296,
+      "learning_rate": 2.555341744230487e-06,
+      "loss": 0.8149,
+      "step": 8727
+    },
+    {
+      "epoch": 0.6728337958680234,
+      "grad_norm": 3.737069606781006,
+      "learning_rate": 2.5542527723270653e-06,
+      "loss": 0.9622,
+      "step": 8728
+    },
+    {
+      "epoch": 0.6729108849830404,
+      "grad_norm": 4.169541358947754,
+      "learning_rate": 2.553163952903315e-06,
+      "loss": 0.9304,
+      "step": 8729
+    },
+    {
+      "epoch": 0.6729879740980573,
+      "grad_norm": 3.955720901489258,
+      "learning_rate": 2.552075286027119e-06,
+      "loss": 0.9517,
+      "step": 8730
+    },
+    {
+      "epoch": 0.6730650632130744,
+      "grad_norm": 3.720717191696167,
+      "learning_rate": 2.5509867717663516e-06,
+      "loss": 0.8208,
+      "step": 8731
+    },
+    {
+      "epoch": 0.6731421523280913,
+      "grad_norm": 4.123862266540527,
+      "learning_rate": 2.5498984101888747e-06,
+      "loss": 0.8568,
+      "step": 8732
+    },
+    {
+      "epoch": 0.6732192414431082,
+      "grad_norm": 4.052290439605713,
+      "learning_rate": 2.5488102013625425e-06,
+      "loss": 1.0482,
+      "step": 8733
+    },
+    {
+      "epoch": 0.6732963305581252,
+      "grad_norm": 3.5889370441436768,
+      "learning_rate": 2.5477221453552015e-06,
+      "loss": 0.9129,
+      "step": 8734
+    },
+    {
+      "epoch": 0.6733734196731421,
+      "grad_norm": 4.095925331115723,
+      "learning_rate": 2.5466342422346813e-06,
+      "loss": 0.9051,
+      "step": 8735
+    },
+    {
+      "epoch": 0.6734505087881592,
+      "grad_norm": 3.5070693492889404,
+      "learning_rate": 2.5455464920688105e-06,
+      "loss": 0.894,
+      "step": 8736
+    },
+    {
+      "epoch": 0.6735275979031761,
+      "grad_norm": 3.590304136276245,
+      "learning_rate": 2.5444588949254035e-06,
+      "loss": 0.9069,
+      "step": 8737
+    },
+    {
+      "epoch": 0.673604687018193,
+      "grad_norm": 3.7862966060638428,
+      "learning_rate": 2.5433714508722674e-06,
+      "loss": 0.9632,
+      "step": 8738
+    },
+    {
+      "epoch": 0.67368177613321,
+      "grad_norm": 3.6480305194854736,
+      "learning_rate": 2.5422841599771995e-06,
+      "loss": 0.9864,
+      "step": 8739
+    },
+    {
+      "epoch": 0.6737588652482269,
+      "grad_norm": 3.975153923034668,
+      "learning_rate": 2.541197022307984e-06,
+      "loss": 0.8769,
+      "step": 8740
+    },
+    {
+      "epoch": 0.673835954363244,
+      "grad_norm": 3.6856942176818848,
+      "learning_rate": 2.5401100379323994e-06,
+      "loss": 0.9895,
+      "step": 8741
+    },
+    {
+      "epoch": 0.6739130434782609,
+      "grad_norm": 3.8510794639587402,
+      "learning_rate": 2.539023206918212e-06,
+      "loss": 0.842,
+      "step": 8742
+    },
+    {
+      "epoch": 0.6739901325932778,
+      "grad_norm": 3.5220234394073486,
+      "learning_rate": 2.537936529333186e-06,
+      "loss": 0.8066,
+      "step": 8743
+    },
+    {
+      "epoch": 0.6740672217082948,
+      "grad_norm": 3.950100898742676,
+      "learning_rate": 2.536850005245064e-06,
+      "loss": 1.1013,
+      "step": 8744
+    },
+    {
+      "epoch": 0.6741443108233117,
+      "grad_norm": 3.5776264667510986,
+      "learning_rate": 2.535763634721587e-06,
+      "loss": 0.9417,
+      "step": 8745
+    },
+    {
+      "epoch": 0.6742213999383287,
+      "grad_norm": 3.58642578125,
+      "learning_rate": 2.5346774178304847e-06,
+      "loss": 0.8659,
+      "step": 8746
+    },
+    {
+      "epoch": 0.6742984890533457,
+      "grad_norm": 4.12030029296875,
+      "learning_rate": 2.5335913546394776e-06,
+      "loss": 0.9124,
+      "step": 8747
+    },
+    {
+      "epoch": 0.6743755781683626,
+      "grad_norm": 3.6494553089141846,
+      "learning_rate": 2.532505445216275e-06,
+      "loss": 0.9372,
+      "step": 8748
+    },
+    {
+      "epoch": 0.6744526672833796,
+      "grad_norm": 3.9090404510498047,
+      "learning_rate": 2.5314196896285804e-06,
+      "loss": 0.9739,
+      "step": 8749
+    },
+    {
+      "epoch": 0.6745297563983965,
+      "grad_norm": 3.881211280822754,
+      "learning_rate": 2.530334087944082e-06,
+      "loss": 0.935,
+      "step": 8750
+    },
+    {
+      "epoch": 0.6746068455134135,
+      "grad_norm": 3.7291910648345947,
+      "learning_rate": 2.5292486402304617e-06,
+      "loss": 0.83,
+      "step": 8751
+    },
+    {
+      "epoch": 0.6746839346284305,
+      "grad_norm": 3.3748013973236084,
+      "learning_rate": 2.5281633465553923e-06,
+      "loss": 0.8803,
+      "step": 8752
+    },
+    {
+      "epoch": 0.6747610237434474,
+      "grad_norm": 3.430265188217163,
+      "learning_rate": 2.527078206986537e-06,
+      "loss": 0.9526,
+      "step": 8753
+    },
+    {
+      "epoch": 0.6748381128584644,
+      "grad_norm": 3.7419652938842773,
+      "learning_rate": 2.525993221591549e-06,
+      "loss": 0.8539,
+      "step": 8754
+    },
+    {
+      "epoch": 0.6749152019734813,
+      "grad_norm": 3.45969557762146,
+      "learning_rate": 2.524908390438069e-06,
+      "loss": 0.9154,
+      "step": 8755
+    },
+    {
+      "epoch": 0.6749922910884983,
+      "grad_norm": 4.311490535736084,
+      "learning_rate": 2.5238237135937314e-06,
+      "loss": 0.9163,
+      "step": 8756
+    },
+    {
+      "epoch": 0.6750693802035153,
+      "grad_norm": 3.4476709365844727,
+      "learning_rate": 2.5227391911261617e-06,
+      "loss": 0.7989,
+      "step": 8757
+    },
+    {
+      "epoch": 0.6751464693185322,
+      "grad_norm": 3.477334976196289,
+      "learning_rate": 2.5216548231029726e-06,
+      "loss": 0.8753,
+      "step": 8758
+    },
+    {
+      "epoch": 0.6752235584335492,
+      "grad_norm": 3.5356128215789795,
+      "learning_rate": 2.5205706095917703e-06,
+      "loss": 0.9253,
+      "step": 8759
+    },
+    {
+      "epoch": 0.6753006475485661,
+      "grad_norm": 3.273407459259033,
+      "learning_rate": 2.5194865506601507e-06,
+      "loss": 0.9407,
+      "step": 8760
+    },
+    {
+      "epoch": 0.6753777366635831,
+      "grad_norm": 3.790802001953125,
+      "learning_rate": 2.518402646375696e-06,
+      "loss": 0.9698,
+      "step": 8761
+    },
+    {
+      "epoch": 0.6754548257786,
+      "grad_norm": 3.5114316940307617,
+      "learning_rate": 2.5173188968059836e-06,
+      "loss": 0.9552,
+      "step": 8762
+    },
+    {
+      "epoch": 0.675531914893617,
+      "grad_norm": 3.583873987197876,
+      "learning_rate": 2.5162353020185798e-06,
+      "loss": 0.9253,
+      "step": 8763
+    },
+    {
+      "epoch": 0.675609004008634,
+      "grad_norm": 3.8350162506103516,
+      "learning_rate": 2.5151518620810413e-06,
+      "loss": 0.8797,
+      "step": 8764
+    },
+    {
+      "epoch": 0.6756860931236509,
+      "grad_norm": 3.6162362098693848,
+      "learning_rate": 2.514068577060917e-06,
+      "loss": 0.9368,
+      "step": 8765
+    },
+    {
+      "epoch": 0.6757631822386679,
+      "grad_norm": 3.375894546508789,
+      "learning_rate": 2.51298544702574e-06,
+      "loss": 0.8251,
+      "step": 8766
+    },
+    {
+      "epoch": 0.6758402713536849,
+      "grad_norm": 3.4712395668029785,
+      "learning_rate": 2.5119024720430396e-06,
+      "loss": 0.8507,
+      "step": 8767
+    },
+    {
+      "epoch": 0.6759173604687018,
+      "grad_norm": 3.914985179901123,
+      "learning_rate": 2.5108196521803336e-06,
+      "loss": 0.9589,
+      "step": 8768
+    },
+    {
+      "epoch": 0.6759944495837188,
+      "grad_norm": 4.005273342132568,
+      "learning_rate": 2.50973698750513e-06,
+      "loss": 0.8717,
+      "step": 8769
+    },
+    {
+      "epoch": 0.6760715386987357,
+      "grad_norm": 3.7250683307647705,
+      "learning_rate": 2.508654478084929e-06,
+      "loss": 0.9127,
+      "step": 8770
+    },
+    {
+      "epoch": 0.6761486278137527,
+      "grad_norm": 3.753918170928955,
+      "learning_rate": 2.5075721239872196e-06,
+      "loss": 0.8688,
+      "step": 8771
+    },
+    {
+      "epoch": 0.6762257169287696,
+      "grad_norm": 3.8328804969787598,
+      "learning_rate": 2.5064899252794783e-06,
+      "loss": 0.9741,
+      "step": 8772
+    },
+    {
+      "epoch": 0.6763028060437866,
+      "grad_norm": 3.605774402618408,
+      "learning_rate": 2.5054078820291754e-06,
+      "loss": 0.9508,
+      "step": 8773
+    },
+    {
+      "epoch": 0.6763798951588036,
+      "grad_norm": 3.709643840789795,
+      "learning_rate": 2.5043259943037714e-06,
+      "loss": 0.9305,
+      "step": 8774
+    },
+    {
+      "epoch": 0.6764569842738205,
+      "grad_norm": 3.6249752044677734,
+      "learning_rate": 2.5032442621707164e-06,
+      "loss": 0.9045,
+      "step": 8775
+    },
+    {
+      "epoch": 0.6765340733888375,
+      "grad_norm": 3.642636299133301,
+      "learning_rate": 2.5021626856974524e-06,
+      "loss": 0.9249,
+      "step": 8776
+    },
+    {
+      "epoch": 0.6766111625038544,
+      "grad_norm": 3.822664976119995,
+      "learning_rate": 2.501081264951406e-06,
+      "loss": 1.0165,
+      "step": 8777
+    },
+    {
+      "epoch": 0.6766882516188714,
+      "grad_norm": 3.8768982887268066,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.9671,
+      "step": 8778
+    },
+    {
+      "epoch": 0.6767653407338884,
+      "grad_norm": 3.6057324409484863,
+      "learning_rate": 2.498918890910648e-06,
+      "loss": 0.8731,
+      "step": 8779
+    },
+    {
+      "epoch": 0.6768424298489053,
+      "grad_norm": 3.6541295051574707,
+      "learning_rate": 2.497837937750749e-06,
+      "loss": 0.9921,
+      "step": 8780
+    },
+    {
+      "epoch": 0.6769195189639223,
+      "grad_norm": 3.908982038497925,
+      "learning_rate": 2.496757140587695e-06,
+      "loss": 0.9808,
+      "step": 8781
+    },
+    {
+      "epoch": 0.6769966080789392,
+      "grad_norm": 3.8420374393463135,
+      "learning_rate": 2.495676499488871e-06,
+      "loss": 1.0192,
+      "step": 8782
+    },
+    {
+      "epoch": 0.6770736971939562,
+      "grad_norm": 3.714198350906372,
+      "learning_rate": 2.494596014521645e-06,
+      "loss": 0.9404,
+      "step": 8783
+    },
+    {
+      "epoch": 0.6771507863089732,
+      "grad_norm": 3.6355252265930176,
+      "learning_rate": 2.4935156857533816e-06,
+      "loss": 0.9077,
+      "step": 8784
+    },
+    {
+      "epoch": 0.6772278754239901,
+      "grad_norm": 3.312899112701416,
+      "learning_rate": 2.4924355132514337e-06,
+      "loss": 0.8434,
+      "step": 8785
+    },
+    {
+      "epoch": 0.6773049645390071,
+      "grad_norm": 3.788271188735962,
+      "learning_rate": 2.491355497083145e-06,
+      "loss": 0.909,
+      "step": 8786
+    },
+    {
+      "epoch": 0.677382053654024,
+      "grad_norm": 3.835744619369507,
+      "learning_rate": 2.490275637315848e-06,
+      "loss": 0.9578,
+      "step": 8787
+    },
+    {
+      "epoch": 0.677459142769041,
+      "grad_norm": 3.8753020763397217,
+      "learning_rate": 2.489195934016867e-06,
+      "loss": 1.0083,
+      "step": 8788
+    },
+    {
+      "epoch": 0.677536231884058,
+      "grad_norm": 3.9721779823303223,
+      "learning_rate": 2.488116387253516e-06,
+      "loss": 0.9055,
+      "step": 8789
+    },
+    {
+      "epoch": 0.6776133209990749,
+      "grad_norm": 4.252837181091309,
+      "learning_rate": 2.487036997093098e-06,
+      "loss": 0.9204,
+      "step": 8790
+    },
+    {
+      "epoch": 0.6776904101140919,
+      "grad_norm": 3.932285785675049,
+      "learning_rate": 2.485957763602911e-06,
+      "loss": 0.9707,
+      "step": 8791
+    },
+    {
+      "epoch": 0.6777674992291088,
+      "grad_norm": 3.8516898155212402,
+      "learning_rate": 2.4848786868502344e-06,
+      "loss": 0.9746,
+      "step": 8792
+    },
+    {
+      "epoch": 0.6778445883441259,
+      "grad_norm": 3.652773141860962,
+      "learning_rate": 2.483799766902346e-06,
+      "loss": 0.8415,
+      "step": 8793
+    },
+    {
+      "epoch": 0.6779216774591428,
+      "grad_norm": 3.671022891998291,
+      "learning_rate": 2.48272100382651e-06,
+      "loss": 0.9036,
+      "step": 8794
+    },
+    {
+      "epoch": 0.6779987665741597,
+      "grad_norm": 3.8974409103393555,
+      "learning_rate": 2.481642397689982e-06,
+      "loss": 1.0436,
+      "step": 8795
+    },
+    {
+      "epoch": 0.6780758556891767,
+      "grad_norm": 4.345669269561768,
+      "learning_rate": 2.4805639485600087e-06,
+      "loss": 1.0052,
+      "step": 8796
+    },
+    {
+      "epoch": 0.6781529448041936,
+      "grad_norm": 3.7597620487213135,
+      "learning_rate": 2.4794856565038256e-06,
+      "loss": 0.8635,
+      "step": 8797
+    },
+    {
+      "epoch": 0.6782300339192107,
+      "grad_norm": 3.701038360595703,
+      "learning_rate": 2.478407521588656e-06,
+      "loss": 0.8971,
+      "step": 8798
+    },
+    {
+      "epoch": 0.6783071230342276,
+      "grad_norm": 3.521311044692993,
+      "learning_rate": 2.4773295438817178e-06,
+      "loss": 0.9003,
+      "step": 8799
+    },
+    {
+      "epoch": 0.6783842121492445,
+      "grad_norm": 3.864536762237549,
+      "learning_rate": 2.476251723450217e-06,
+      "loss": 1.0305,
+      "step": 8800
+    },
+    {
+      "epoch": 0.6784613012642615,
+      "grad_norm": 3.7340474128723145,
+      "learning_rate": 2.475174060361351e-06,
+      "loss": 0.991,
+      "step": 8801
+    },
+    {
+      "epoch": 0.6785383903792784,
+      "grad_norm": 3.8703043460845947,
+      "learning_rate": 2.4740965546823076e-06,
+      "loss": 0.9927,
+      "step": 8802
+    },
+    {
+      "epoch": 0.6786154794942955,
+      "grad_norm": 3.6807234287261963,
+      "learning_rate": 2.47301920648026e-06,
+      "loss": 0.9202,
+      "step": 8803
+    },
+    {
+      "epoch": 0.6786925686093124,
+      "grad_norm": 3.687746047973633,
+      "learning_rate": 2.471942015822377e-06,
+      "loss": 0.9262,
+      "step": 8804
+    },
+    {
+      "epoch": 0.6787696577243293,
+      "grad_norm": 3.667069435119629,
+      "learning_rate": 2.4708649827758167e-06,
+      "loss": 0.9589,
+      "step": 8805
+    },
+    {
+      "epoch": 0.6788467468393463,
+      "grad_norm": 3.750629425048828,
+      "learning_rate": 2.4697881074077263e-06,
+      "loss": 0.8988,
+      "step": 8806
+    },
+    {
+      "epoch": 0.6789238359543632,
+      "grad_norm": 3.8485107421875,
+      "learning_rate": 2.468711389785242e-06,
+      "loss": 1.0432,
+      "step": 8807
+    },
+    {
+      "epoch": 0.6790009250693803,
+      "grad_norm": 3.7426390647888184,
+      "learning_rate": 2.4676348299754956e-06,
+      "loss": 0.9872,
+      "step": 8808
+    },
+    {
+      "epoch": 0.6790780141843972,
+      "grad_norm": 3.8060035705566406,
+      "learning_rate": 2.4665584280456002e-06,
+      "loss": 0.9285,
+      "step": 8809
+    },
+    {
+      "epoch": 0.6791551032994141,
+      "grad_norm": 3.9451937675476074,
+      "learning_rate": 2.4654821840626653e-06,
+      "loss": 1.0148,
+      "step": 8810
+    },
+    {
+      "epoch": 0.6792321924144311,
+      "grad_norm": 3.2711398601531982,
+      "learning_rate": 2.4644060980937894e-06,
+      "loss": 0.8548,
+      "step": 8811
+    },
+    {
+      "epoch": 0.679309281529448,
+      "grad_norm": 3.3944523334503174,
+      "learning_rate": 2.4633301702060613e-06,
+      "loss": 0.889,
+      "step": 8812
+    },
+    {
+      "epoch": 0.679386370644465,
+      "grad_norm": 3.9511239528656006,
+      "learning_rate": 2.4622544004665617e-06,
+      "loss": 0.8868,
+      "step": 8813
+    },
+    {
+      "epoch": 0.679463459759482,
+      "grad_norm": 3.569582462310791,
+      "learning_rate": 2.4611787889423546e-06,
+      "loss": 0.8662,
+      "step": 8814
+    },
+    {
+      "epoch": 0.6795405488744989,
+      "grad_norm": 3.395447254180908,
+      "learning_rate": 2.460103335700501e-06,
+      "loss": 0.8403,
+      "step": 8815
+    },
+    {
+      "epoch": 0.6796176379895159,
+      "grad_norm": 3.699881076812744,
+      "learning_rate": 2.45902804080805e-06,
+      "loss": 0.8871,
+      "step": 8816
+    },
+    {
+      "epoch": 0.6796947271045328,
+      "grad_norm": 3.846745491027832,
+      "learning_rate": 2.457952904332041e-06,
+      "loss": 0.8642,
+      "step": 8817
+    },
+    {
+      "epoch": 0.6797718162195499,
+      "grad_norm": 3.8669497966766357,
+      "learning_rate": 2.4568779263395026e-06,
+      "loss": 0.9709,
+      "step": 8818
+    },
+    {
+      "epoch": 0.6798489053345668,
+      "grad_norm": 3.6306495666503906,
+      "learning_rate": 2.455803106897456e-06,
+      "loss": 0.9367,
+      "step": 8819
+    },
+    {
+      "epoch": 0.6799259944495837,
+      "grad_norm": 3.4192559719085693,
+      "learning_rate": 2.454728446072907e-06,
+      "loss": 0.7778,
+      "step": 8820
+    },
+    {
+      "epoch": 0.6800030835646007,
+      "grad_norm": 3.654353141784668,
+      "learning_rate": 2.4536539439328575e-06,
+      "loss": 0.9357,
+      "step": 8821
+    },
+    {
+      "epoch": 0.6800801726796176,
+      "grad_norm": 3.7927420139312744,
+      "learning_rate": 2.4525796005442963e-06,
+      "loss": 0.9209,
+      "step": 8822
+    },
+    {
+      "epoch": 0.6801572617946346,
+      "grad_norm": 3.8942689895629883,
+      "learning_rate": 2.451505415974204e-06,
+      "loss": 0.824,
+      "step": 8823
+    },
+    {
+      "epoch": 0.6802343509096516,
+      "grad_norm": 4.133578777313232,
+      "learning_rate": 2.4504313902895523e-06,
+      "loss": 0.9132,
+      "step": 8824
+    },
+    {
+      "epoch": 0.6803114400246685,
+      "grad_norm": 4.291555404663086,
+      "learning_rate": 2.449357523557296e-06,
+      "loss": 0.9479,
+      "step": 8825
+    },
+    {
+      "epoch": 0.6803885291396855,
+      "grad_norm": 3.8000755310058594,
+      "learning_rate": 2.4482838158443883e-06,
+      "loss": 0.8301,
+      "step": 8826
+    },
+    {
+      "epoch": 0.6804656182547024,
+      "grad_norm": 3.7706243991851807,
+      "learning_rate": 2.4472102672177693e-06,
+      "loss": 1.0852,
+      "step": 8827
+    },
+    {
+      "epoch": 0.6805427073697194,
+      "grad_norm": 3.8600974082946777,
+      "learning_rate": 2.4461368777443683e-06,
+      "loss": 0.9142,
+      "step": 8828
+    },
+    {
+      "epoch": 0.6806197964847364,
+      "grad_norm": 3.7134435176849365,
+      "learning_rate": 2.4450636474911072e-06,
+      "loss": 0.9089,
+      "step": 8829
+    },
+    {
+      "epoch": 0.6806968855997533,
+      "grad_norm": 3.886906862258911,
+      "learning_rate": 2.4439905765248944e-06,
+      "loss": 0.9055,
+      "step": 8830
+    },
+    {
+      "epoch": 0.6807739747147703,
+      "grad_norm": 3.8458032608032227,
+      "learning_rate": 2.442917664912632e-06,
+      "loss": 0.9037,
+      "step": 8831
+    },
+    {
+      "epoch": 0.6808510638297872,
+      "grad_norm": 3.480123281478882,
+      "learning_rate": 2.441844912721209e-06,
+      "loss": 0.8902,
+      "step": 8832
+    },
+    {
+      "epoch": 0.6809281529448042,
+      "grad_norm": 3.9139695167541504,
+      "learning_rate": 2.4407723200175064e-06,
+      "loss": 0.9043,
+      "step": 8833
+    },
+    {
+      "epoch": 0.6810052420598212,
+      "grad_norm": 3.7381532192230225,
+      "learning_rate": 2.439699886868398e-06,
+      "loss": 1.0108,
+      "step": 8834
+    },
+    {
+      "epoch": 0.6810823311748381,
+      "grad_norm": 3.5520753860473633,
+      "learning_rate": 2.438627613340739e-06,
+      "loss": 0.9035,
+      "step": 8835
+    },
+    {
+      "epoch": 0.6811594202898551,
+      "grad_norm": 4.716245651245117,
+      "learning_rate": 2.4375554995013827e-06,
+      "loss": 0.9881,
+      "step": 8836
+    },
+    {
+      "epoch": 0.681236509404872,
+      "grad_norm": 3.765180826187134,
+      "learning_rate": 2.43648354541717e-06,
+      "loss": 0.9127,
+      "step": 8837
+    },
+    {
+      "epoch": 0.681313598519889,
+      "grad_norm": 3.5906150341033936,
+      "learning_rate": 2.435411751154931e-06,
+      "loss": 0.9026,
+      "step": 8838
+    },
+    {
+      "epoch": 0.681390687634906,
+      "grad_norm": 3.9749667644500732,
+      "learning_rate": 2.4343401167814895e-06,
+      "loss": 0.8457,
+      "step": 8839
+    },
+    {
+      "epoch": 0.6814677767499229,
+      "grad_norm": 4.670644760131836,
+      "learning_rate": 2.4332686423636516e-06,
+      "loss": 0.9388,
+      "step": 8840
+    },
+    {
+      "epoch": 0.6815448658649399,
+      "grad_norm": 3.8460373878479004,
+      "learning_rate": 2.43219732796822e-06,
+      "loss": 0.8584,
+      "step": 8841
+    },
+    {
+      "epoch": 0.6816219549799568,
+      "grad_norm": 3.915990114212036,
+      "learning_rate": 2.431126173661986e-06,
+      "loss": 0.9734,
+      "step": 8842
+    },
+    {
+      "epoch": 0.6816990440949738,
+      "grad_norm": 3.3147318363189697,
+      "learning_rate": 2.430055179511731e-06,
+      "loss": 0.8453,
+      "step": 8843
+    },
+    {
+      "epoch": 0.6817761332099908,
+      "grad_norm": 3.620931386947632,
+      "learning_rate": 2.4289843455842255e-06,
+      "loss": 0.8603,
+      "step": 8844
+    },
+    {
+      "epoch": 0.6818532223250077,
+      "grad_norm": 5.175318717956543,
+      "learning_rate": 2.4279136719462326e-06,
+      "loss": 0.8268,
+      "step": 8845
+    },
+    {
+      "epoch": 0.6819303114400247,
+      "grad_norm": 3.6000447273254395,
+      "learning_rate": 2.4268431586644985e-06,
+      "loss": 0.8702,
+      "step": 8846
+    },
+    {
+      "epoch": 0.6820074005550416,
+      "grad_norm": 3.783622980117798,
+      "learning_rate": 2.4257728058057673e-06,
+      "loss": 0.97,
+      "step": 8847
+    },
+    {
+      "epoch": 0.6820844896700586,
+      "grad_norm": 3.600386381149292,
+      "learning_rate": 2.424702613436769e-06,
+      "loss": 1.039,
+      "step": 8848
+    },
+    {
+      "epoch": 0.6821615787850756,
+      "grad_norm": 3.5093846321105957,
+      "learning_rate": 2.423632581624225e-06,
+      "loss": 0.9421,
+      "step": 8849
+    },
+    {
+      "epoch": 0.6822386679000925,
+      "grad_norm": 3.684954881668091,
+      "learning_rate": 2.422562710434848e-06,
+      "loss": 0.8655,
+      "step": 8850
+    },
+    {
+      "epoch": 0.6823157570151095,
+      "grad_norm": 3.958432912826538,
+      "learning_rate": 2.4214929999353354e-06,
+      "loss": 0.8486,
+      "step": 8851
+    },
+    {
+      "epoch": 0.6823928461301264,
+      "grad_norm": 4.018614768981934,
+      "learning_rate": 2.4204234501923797e-06,
+      "loss": 0.9712,
+      "step": 8852
+    },
+    {
+      "epoch": 0.6824699352451434,
+      "grad_norm": 3.787519931793213,
+      "learning_rate": 2.419354061272662e-06,
+      "loss": 0.9451,
+      "step": 8853
+    },
+    {
+      "epoch": 0.6825470243601603,
+      "grad_norm": 3.5393474102020264,
+      "learning_rate": 2.4182848332428534e-06,
+      "loss": 0.9476,
+      "step": 8854
+    },
+    {
+      "epoch": 0.6826241134751773,
+      "grad_norm": 3.7970054149627686,
+      "learning_rate": 2.417215766169614e-06,
+      "loss": 0.9967,
+      "step": 8855
+    },
+    {
+      "epoch": 0.6827012025901943,
+      "grad_norm": 3.7069692611694336,
+      "learning_rate": 2.4161468601195966e-06,
+      "loss": 0.9612,
+      "step": 8856
+    },
+    {
+      "epoch": 0.6827782917052112,
+      "grad_norm": 3.631960868835449,
+      "learning_rate": 2.4150781151594386e-06,
+      "loss": 0.9315,
+      "step": 8857
+    },
+    {
+      "epoch": 0.6828553808202282,
+      "grad_norm": 3.616215229034424,
+      "learning_rate": 2.4140095313557734e-06,
+      "loss": 0.9053,
+      "step": 8858
+    },
+    {
+      "epoch": 0.6829324699352451,
+      "grad_norm": 3.7601921558380127,
+      "learning_rate": 2.4129411087752196e-06,
+      "loss": 1.0008,
+      "step": 8859
+    },
+    {
+      "epoch": 0.6830095590502621,
+      "grad_norm": 3.73475980758667,
+      "learning_rate": 2.4118728474843894e-06,
+      "loss": 0.9158,
+      "step": 8860
+    },
+    {
+      "epoch": 0.6830866481652791,
+      "grad_norm": 3.5904901027679443,
+      "learning_rate": 2.4108047475498847e-06,
+      "loss": 0.9244,
+      "step": 8861
+    },
+    {
+      "epoch": 0.683163737280296,
+      "grad_norm": 3.4571268558502197,
+      "learning_rate": 2.4097368090382922e-06,
+      "loss": 0.8297,
+      "step": 8862
+    },
+    {
+      "epoch": 0.683240826395313,
+      "grad_norm": 3.620143413543701,
+      "learning_rate": 2.4086690320161947e-06,
+      "loss": 0.8506,
+      "step": 8863
+    },
+    {
+      "epoch": 0.6833179155103299,
+      "grad_norm": 3.556610107421875,
+      "learning_rate": 2.407601416550162e-06,
+      "loss": 0.9505,
+      "step": 8864
+    },
+    {
+      "epoch": 0.6833950046253469,
+      "grad_norm": 3.6465303897857666,
+      "learning_rate": 2.4065339627067545e-06,
+      "loss": 0.9715,
+      "step": 8865
+    },
+    {
+      "epoch": 0.6834720937403639,
+      "grad_norm": 3.4857358932495117,
+      "learning_rate": 2.4054666705525227e-06,
+      "loss": 0.7897,
+      "step": 8866
+    },
+    {
+      "epoch": 0.6835491828553808,
+      "grad_norm": 3.4617760181427,
+      "learning_rate": 2.404399540154008e-06,
+      "loss": 0.8728,
+      "step": 8867
+    },
+    {
+      "epoch": 0.6836262719703978,
+      "grad_norm": 3.6965367794036865,
+      "learning_rate": 2.403332571577738e-06,
+      "loss": 0.8499,
+      "step": 8868
+    },
+    {
+      "epoch": 0.6837033610854147,
+      "grad_norm": 4.105483055114746,
+      "learning_rate": 2.402265764890233e-06,
+      "loss": 0.9697,
+      "step": 8869
+    },
+    {
+      "epoch": 0.6837804502004317,
+      "grad_norm": 4.040141582489014,
+      "learning_rate": 2.4011991201580036e-06,
+      "loss": 0.8711,
+      "step": 8870
+    },
+    {
+      "epoch": 0.6838575393154487,
+      "grad_norm": 3.5217995643615723,
+      "learning_rate": 2.4001326374475496e-06,
+      "loss": 0.8271,
+      "step": 8871
+    },
+    {
+      "epoch": 0.6839346284304656,
+      "grad_norm": 3.501316785812378,
+      "learning_rate": 2.3990663168253627e-06,
+      "loss": 0.9758,
+      "step": 8872
+    },
+    {
+      "epoch": 0.6840117175454826,
+      "grad_norm": 3.6326067447662354,
+      "learning_rate": 2.398000158357918e-06,
+      "loss": 0.909,
+      "step": 8873
+    },
+    {
+      "epoch": 0.6840888066604995,
+      "grad_norm": 3.510653257369995,
+      "learning_rate": 2.396934162111686e-06,
+      "loss": 0.9052,
+      "step": 8874
+    },
+    {
+      "epoch": 0.6841658957755165,
+      "grad_norm": 3.3554039001464844,
+      "learning_rate": 2.395868328153129e-06,
+      "loss": 0.9057,
+      "step": 8875
+    },
+    {
+      "epoch": 0.6842429848905335,
+      "grad_norm": 3.7715797424316406,
+      "learning_rate": 2.3948026565486963e-06,
+      "loss": 0.9396,
+      "step": 8876
+    },
+    {
+      "epoch": 0.6843200740055504,
+      "grad_norm": 3.4584174156188965,
+      "learning_rate": 2.3937371473648237e-06,
+      "loss": 0.8345,
+      "step": 8877
+    },
+    {
+      "epoch": 0.6843971631205674,
+      "grad_norm": 3.387113571166992,
+      "learning_rate": 2.3926718006679416e-06,
+      "loss": 0.8837,
+      "step": 8878
+    },
+    {
+      "epoch": 0.6844742522355843,
+      "grad_norm": 3.546274423599243,
+      "learning_rate": 2.391606616524469e-06,
+      "loss": 0.8913,
+      "step": 8879
+    },
+    {
+      "epoch": 0.6845513413506013,
+      "grad_norm": 3.3328139781951904,
+      "learning_rate": 2.390541595000815e-06,
+      "loss": 0.8801,
+      "step": 8880
+    },
+    {
+      "epoch": 0.6846284304656183,
+      "grad_norm": 3.5848608016967773,
+      "learning_rate": 2.3894767361633775e-06,
+      "loss": 0.8967,
+      "step": 8881
+    },
+    {
+      "epoch": 0.6847055195806352,
+      "grad_norm": 3.5277299880981445,
+      "learning_rate": 2.388412040078547e-06,
+      "loss": 0.8776,
+      "step": 8882
+    },
+    {
+      "epoch": 0.6847826086956522,
+      "grad_norm": 3.7416014671325684,
+      "learning_rate": 2.387347506812698e-06,
+      "loss": 0.811,
+      "step": 8883
+    },
+    {
+      "epoch": 0.6848596978106691,
+      "grad_norm": 3.776992082595825,
+      "learning_rate": 2.3862831364322013e-06,
+      "loss": 0.966,
+      "step": 8884
+    },
+    {
+      "epoch": 0.684936786925686,
+      "grad_norm": 3.511543035507202,
+      "learning_rate": 2.385218929003414e-06,
+      "loss": 0.9115,
+      "step": 8885
+    },
+    {
+      "epoch": 0.6850138760407031,
+      "grad_norm": 3.6545498371124268,
+      "learning_rate": 2.3841548845926844e-06,
+      "loss": 0.9297,
+      "step": 8886
+    },
+    {
+      "epoch": 0.68509096515572,
+      "grad_norm": 4.169460773468018,
+      "learning_rate": 2.3830910032663514e-06,
+      "loss": 0.9614,
+      "step": 8887
+    },
+    {
+      "epoch": 0.685168054270737,
+      "grad_norm": 3.843756675720215,
+      "learning_rate": 2.3820272850907393e-06,
+      "loss": 0.8832,
+      "step": 8888
+    },
+    {
+      "epoch": 0.6852451433857539,
+      "grad_norm": 3.783128499984741,
+      "learning_rate": 2.380963730132167e-06,
+      "loss": 0.8474,
+      "step": 8889
+    },
+    {
+      "epoch": 0.6853222325007708,
+      "grad_norm": 3.7089338302612305,
+      "learning_rate": 2.3799003384569417e-06,
+      "loss": 0.941,
+      "step": 8890
+    },
+    {
+      "epoch": 0.6853993216157879,
+      "grad_norm": 3.5771591663360596,
+      "learning_rate": 2.3788371101313614e-06,
+      "loss": 0.9073,
+      "step": 8891
+    },
+    {
+      "epoch": 0.6854764107308048,
+      "grad_norm": 3.6812937259674072,
+      "learning_rate": 2.3777740452217106e-06,
+      "loss": 0.9433,
+      "step": 8892
+    },
+    {
+      "epoch": 0.6855534998458218,
+      "grad_norm": 3.610495090484619,
+      "learning_rate": 2.3767111437942703e-06,
+      "loss": 0.9288,
+      "step": 8893
+    },
+    {
+      "epoch": 0.6856305889608387,
+      "grad_norm": 3.8256382942199707,
+      "learning_rate": 2.375648405915301e-06,
+      "loss": 0.8738,
+      "step": 8894
+    },
+    {
+      "epoch": 0.6857076780758556,
+      "grad_norm": 3.9262545108795166,
+      "learning_rate": 2.374585831651062e-06,
+      "loss": 0.8454,
+      "step": 8895
+    },
+    {
+      "epoch": 0.6857847671908727,
+      "grad_norm": 3.498192071914673,
+      "learning_rate": 2.3735234210677994e-06,
+      "loss": 0.8577,
+      "step": 8896
+    },
+    {
+      "epoch": 0.6858618563058896,
+      "grad_norm": 3.9072468280792236,
+      "learning_rate": 2.3724611742317483e-06,
+      "loss": 0.906,
+      "step": 8897
+    },
+    {
+      "epoch": 0.6859389454209066,
+      "grad_norm": 3.8992998600006104,
+      "learning_rate": 2.3713990912091367e-06,
+      "loss": 0.8959,
+      "step": 8898
+    },
+    {
+      "epoch": 0.6860160345359235,
+      "grad_norm": 3.6476802825927734,
+      "learning_rate": 2.3703371720661754e-06,
+      "loss": 0.8936,
+      "step": 8899
+    },
+    {
+      "epoch": 0.6860931236509404,
+      "grad_norm": 3.575178384780884,
+      "learning_rate": 2.369275416869073e-06,
+      "loss": 0.9295,
+      "step": 8900
+    },
+    {
+      "epoch": 0.6861702127659575,
+      "grad_norm": 4.027018070220947,
+      "learning_rate": 2.3682138256840224e-06,
+      "loss": 0.9805,
+      "step": 8901
+    },
+    {
+      "epoch": 0.6862473018809744,
+      "grad_norm": 3.9200379848480225,
+      "learning_rate": 2.3671523985772105e-06,
+      "loss": 0.9032,
+      "step": 8902
+    },
+    {
+      "epoch": 0.6863243909959914,
+      "grad_norm": 3.6151862144470215,
+      "learning_rate": 2.36609113561481e-06,
+      "loss": 0.8806,
+      "step": 8903
+    },
+    {
+      "epoch": 0.6864014801110083,
+      "grad_norm": 4.015349388122559,
+      "learning_rate": 2.365030036862988e-06,
+      "loss": 0.9448,
+      "step": 8904
+    },
+    {
+      "epoch": 0.6864785692260252,
+      "grad_norm": 3.450439691543579,
+      "learning_rate": 2.363969102387894e-06,
+      "loss": 0.9339,
+      "step": 8905
+    },
+    {
+      "epoch": 0.6865556583410423,
+      "grad_norm": 3.489480972290039,
+      "learning_rate": 2.3629083322556744e-06,
+      "loss": 0.8266,
+      "step": 8906
+    },
+    {
+      "epoch": 0.6866327474560592,
+      "grad_norm": 3.841921091079712,
+      "learning_rate": 2.361847726532463e-06,
+      "loss": 0.9075,
+      "step": 8907
+    },
+    {
+      "epoch": 0.6867098365710762,
+      "grad_norm": 3.6356115341186523,
+      "learning_rate": 2.3607872852843814e-06,
+      "loss": 0.9278,
+      "step": 8908
+    },
+    {
+      "epoch": 0.6867869256860931,
+      "grad_norm": 3.600667715072632,
+      "learning_rate": 2.3597270085775466e-06,
+      "loss": 0.9311,
+      "step": 8909
+    },
+    {
+      "epoch": 0.68686401480111,
+      "grad_norm": 3.5631704330444336,
+      "learning_rate": 2.3586668964780562e-06,
+      "loss": 0.8496,
+      "step": 8910
+    },
+    {
+      "epoch": 0.6869411039161271,
+      "grad_norm": 3.3869268894195557,
+      "learning_rate": 2.3576069490520046e-06,
+      "loss": 0.94,
+      "step": 8911
+    },
+    {
+      "epoch": 0.687018193031144,
+      "grad_norm": 4.510827541351318,
+      "learning_rate": 2.3565471663654752e-06,
+      "loss": 0.8451,
+      "step": 8912
+    },
+    {
+      "epoch": 0.687095282146161,
+      "grad_norm": 3.5590405464172363,
+      "learning_rate": 2.355487548484539e-06,
+      "loss": 0.9092,
+      "step": 8913
+    },
+    {
+      "epoch": 0.6871723712611779,
+      "grad_norm": 3.764672040939331,
+      "learning_rate": 2.3544280954752573e-06,
+      "loss": 0.9336,
+      "step": 8914
+    },
+    {
+      "epoch": 0.6872494603761948,
+      "grad_norm": 3.8435845375061035,
+      "learning_rate": 2.3533688074036847e-06,
+      "loss": 0.9875,
+      "step": 8915
+    },
+    {
+      "epoch": 0.6873265494912119,
+      "grad_norm": 3.645751714706421,
+      "learning_rate": 2.3523096843358573e-06,
+      "loss": 0.8304,
+      "step": 8916
+    },
+    {
+      "epoch": 0.6874036386062288,
+      "grad_norm": 3.7767014503479004,
+      "learning_rate": 2.3512507263378086e-06,
+      "loss": 0.8513,
+      "step": 8917
+    },
+    {
+      "epoch": 0.6874807277212458,
+      "grad_norm": 3.934950590133667,
+      "learning_rate": 2.3501919334755584e-06,
+      "loss": 0.8667,
+      "step": 8918
+    },
+    {
+      "epoch": 0.6875578168362627,
+      "grad_norm": 3.6691348552703857,
+      "learning_rate": 2.3491333058151185e-06,
+      "loss": 1.0102,
+      "step": 8919
+    },
+    {
+      "epoch": 0.6876349059512796,
+      "grad_norm": 3.4543542861938477,
+      "learning_rate": 2.3480748434224872e-06,
+      "loss": 0.8902,
+      "step": 8920
+    },
+    {
+      "epoch": 0.6877119950662967,
+      "grad_norm": 4.049129486083984,
+      "learning_rate": 2.347016546363654e-06,
+      "loss": 0.821,
+      "step": 8921
+    },
+    {
+      "epoch": 0.6877890841813136,
+      "grad_norm": 3.604243755340576,
+      "learning_rate": 2.3459584147046e-06,
+      "loss": 0.8626,
+      "step": 8922
+    },
+    {
+      "epoch": 0.6878661732963306,
+      "grad_norm": 3.946105480194092,
+      "learning_rate": 2.344900448511293e-06,
+      "loss": 0.9558,
+      "step": 8923
+    },
+    {
+      "epoch": 0.6879432624113475,
+      "grad_norm": 3.4957942962646484,
+      "learning_rate": 2.3438426478496936e-06,
+      "loss": 0.8554,
+      "step": 8924
+    },
+    {
+      "epoch": 0.6880203515263644,
+      "grad_norm": 3.9918699264526367,
+      "learning_rate": 2.3427850127857464e-06,
+      "loss": 1.0795,
+      "step": 8925
+    },
+    {
+      "epoch": 0.6880974406413815,
+      "grad_norm": 3.549349069595337,
+      "learning_rate": 2.341727543385392e-06,
+      "loss": 0.8196,
+      "step": 8926
+    },
+    {
+      "epoch": 0.6881745297563984,
+      "grad_norm": 3.535104751586914,
+      "learning_rate": 2.3406702397145574e-06,
+      "loss": 0.8999,
+      "step": 8927
+    },
+    {
+      "epoch": 0.6882516188714154,
+      "grad_norm": 3.535996913909912,
+      "learning_rate": 2.3396131018391606e-06,
+      "loss": 0.8936,
+      "step": 8928
+    },
+    {
+      "epoch": 0.6883287079864323,
+      "grad_norm": 4.0836262702941895,
+      "learning_rate": 2.3385561298251087e-06,
+      "loss": 0.9796,
+      "step": 8929
+    },
+    {
+      "epoch": 0.6884057971014492,
+      "grad_norm": 3.8373196125030518,
+      "learning_rate": 2.337499323738299e-06,
+      "loss": 0.9182,
+      "step": 8930
+    },
+    {
+      "epoch": 0.6884828862164662,
+      "grad_norm": 3.3470261096954346,
+      "learning_rate": 2.3364426836446163e-06,
+      "loss": 0.8113,
+      "step": 8931
+    },
+    {
+      "epoch": 0.6885599753314832,
+      "grad_norm": 3.9608960151672363,
+      "learning_rate": 2.3353862096099365e-06,
+      "loss": 0.9545,
+      "step": 8932
+    },
+    {
+      "epoch": 0.6886370644465002,
+      "grad_norm": 3.7849230766296387,
+      "learning_rate": 2.3343299017001263e-06,
+      "loss": 0.8956,
+      "step": 8933
+    },
+    {
+      "epoch": 0.6887141535615171,
+      "grad_norm": 4.301094055175781,
+      "learning_rate": 2.333273759981041e-06,
+      "loss": 0.9049,
+      "step": 8934
+    },
+    {
+      "epoch": 0.688791242676534,
+      "grad_norm": 3.382276773452759,
+      "learning_rate": 2.3322177845185278e-06,
+      "loss": 0.8616,
+      "step": 8935
+    },
+    {
+      "epoch": 0.688868331791551,
+      "grad_norm": 3.7838985919952393,
+      "learning_rate": 2.3311619753784167e-06,
+      "loss": 0.9545,
+      "step": 8936
+    },
+    {
+      "epoch": 0.688945420906568,
+      "grad_norm": 3.419304609298706,
+      "learning_rate": 2.3301063326265338e-06,
+      "loss": 0.9138,
+      "step": 8937
+    },
+    {
+      "epoch": 0.689022510021585,
+      "grad_norm": 3.7308456897735596,
+      "learning_rate": 2.329050856328694e-06,
+      "loss": 0.8699,
+      "step": 8938
+    },
+    {
+      "epoch": 0.6890995991366019,
+      "grad_norm": 3.8383610248565674,
+      "learning_rate": 2.3279955465506995e-06,
+      "loss": 0.9504,
+      "step": 8939
+    },
+    {
+      "epoch": 0.6891766882516188,
+      "grad_norm": 3.7427611351013184,
+      "learning_rate": 2.3269404033583443e-06,
+      "loss": 0.9492,
+      "step": 8940
+    },
+    {
+      "epoch": 0.6892537773666358,
+      "grad_norm": 3.5722649097442627,
+      "learning_rate": 2.3258854268174125e-06,
+      "loss": 0.9321,
+      "step": 8941
+    },
+    {
+      "epoch": 0.6893308664816528,
+      "grad_norm": 4.43306827545166,
+      "learning_rate": 2.324830616993673e-06,
+      "loss": 1.0846,
+      "step": 8942
+    },
+    {
+      "epoch": 0.6894079555966698,
+      "grad_norm": 3.7848153114318848,
+      "learning_rate": 2.3237759739528893e-06,
+      "loss": 0.8663,
+      "step": 8943
+    },
+    {
+      "epoch": 0.6894850447116867,
+      "grad_norm": 4.198482513427734,
+      "learning_rate": 2.3227214977608136e-06,
+      "loss": 0.9388,
+      "step": 8944
+    },
+    {
+      "epoch": 0.6895621338267036,
+      "grad_norm": 3.4994449615478516,
+      "learning_rate": 2.321667188483186e-06,
+      "loss": 0.8521,
+      "step": 8945
+    },
+    {
+      "epoch": 0.6896392229417206,
+      "grad_norm": 3.8813703060150146,
+      "learning_rate": 2.3206130461857403e-06,
+      "loss": 1.072,
+      "step": 8946
+    },
+    {
+      "epoch": 0.6897163120567376,
+      "grad_norm": 3.928469181060791,
+      "learning_rate": 2.3195590709341915e-06,
+      "loss": 0.8955,
+      "step": 8947
+    },
+    {
+      "epoch": 0.6897934011717546,
+      "grad_norm": 3.67775821685791,
+      "learning_rate": 2.3185052627942528e-06,
+      "loss": 0.9373,
+      "step": 8948
+    },
+    {
+      "epoch": 0.6898704902867715,
+      "grad_norm": 3.703199863433838,
+      "learning_rate": 2.317451621831623e-06,
+      "loss": 0.9561,
+      "step": 8949
+    },
+    {
+      "epoch": 0.6899475794017884,
+      "grad_norm": 3.4708447456359863,
+      "learning_rate": 2.3163981481119913e-06,
+      "loss": 0.8132,
+      "step": 8950
+    },
+    {
+      "epoch": 0.6900246685168054,
+      "grad_norm": 3.7089951038360596,
+      "learning_rate": 2.3153448417010367e-06,
+      "loss": 0.9367,
+      "step": 8951
+    },
+    {
+      "epoch": 0.6901017576318224,
+      "grad_norm": 3.567253351211548,
+      "learning_rate": 2.314291702664428e-06,
+      "loss": 0.9247,
+      "step": 8952
+    },
+    {
+      "epoch": 0.6901788467468394,
+      "grad_norm": 4.459877967834473,
+      "learning_rate": 2.3132387310678212e-06,
+      "loss": 0.9534,
+      "step": 8953
+    },
+    {
+      "epoch": 0.6902559358618563,
+      "grad_norm": 3.3123741149902344,
+      "learning_rate": 2.3121859269768637e-06,
+      "loss": 0.8037,
+      "step": 8954
+    },
+    {
+      "epoch": 0.6903330249768732,
+      "grad_norm": 3.7758708000183105,
+      "learning_rate": 2.3111332904571933e-06,
+      "loss": 0.8445,
+      "step": 8955
+    },
+    {
+      "epoch": 0.6904101140918902,
+      "grad_norm": 3.962944507598877,
+      "learning_rate": 2.3100808215744365e-06,
+      "loss": 0.9568,
+      "step": 8956
+    },
+    {
+      "epoch": 0.6904872032069072,
+      "grad_norm": 3.808398723602295,
+      "learning_rate": 2.3090285203942104e-06,
+      "loss": 0.9115,
+      "step": 8957
+    },
+    {
+      "epoch": 0.6905642923219242,
+      "grad_norm": 3.7572476863861084,
+      "learning_rate": 2.3079763869821176e-06,
+      "loss": 0.9322,
+      "step": 8958
+    },
+    {
+      "epoch": 0.6906413814369411,
+      "grad_norm": 3.5704262256622314,
+      "learning_rate": 2.3069244214037546e-06,
+      "loss": 0.8228,
+      "step": 8959
+    },
+    {
+      "epoch": 0.690718470551958,
+      "grad_norm": 3.9301810264587402,
+      "learning_rate": 2.3058726237247063e-06,
+      "loss": 0.903,
+      "step": 8960
+    },
+    {
+      "epoch": 0.690795559666975,
+      "grad_norm": 3.7818596363067627,
+      "learning_rate": 2.3048209940105465e-06,
+      "loss": 0.8574,
+      "step": 8961
+    },
+    {
+      "epoch": 0.690872648781992,
+      "grad_norm": 3.3645167350769043,
+      "learning_rate": 2.3037695323268394e-06,
+      "loss": 0.9248,
+      "step": 8962
+    },
+    {
+      "epoch": 0.690949737897009,
+      "grad_norm": 3.7038087844848633,
+      "learning_rate": 2.3027182387391373e-06,
+      "loss": 0.8641,
+      "step": 8963
+    },
+    {
+      "epoch": 0.6910268270120259,
+      "grad_norm": 3.863682985305786,
+      "learning_rate": 2.3016671133129837e-06,
+      "loss": 0.9759,
+      "step": 8964
+    },
+    {
+      "epoch": 0.6911039161270429,
+      "grad_norm": 4.024723529815674,
+      "learning_rate": 2.3006161561139105e-06,
+      "loss": 0.8978,
+      "step": 8965
+    },
+    {
+      "epoch": 0.6911810052420598,
+      "grad_norm": 3.979031562805176,
+      "learning_rate": 2.29956536720744e-06,
+      "loss": 0.95,
+      "step": 8966
+    },
+    {
+      "epoch": 0.6912580943570767,
+      "grad_norm": 4.345090389251709,
+      "learning_rate": 2.298514746659084e-06,
+      "loss": 1.0485,
+      "step": 8967
+    },
+    {
+      "epoch": 0.6913351834720938,
+      "grad_norm": 3.6718671321868896,
+      "learning_rate": 2.2974642945343405e-06,
+      "loss": 0.9435,
+      "step": 8968
+    },
+    {
+      "epoch": 0.6914122725871107,
+      "grad_norm": 4.286196231842041,
+      "learning_rate": 2.2964140108987015e-06,
+      "loss": 0.9947,
+      "step": 8969
+    },
+    {
+      "epoch": 0.6914893617021277,
+      "grad_norm": 3.392043352127075,
+      "learning_rate": 2.2953638958176465e-06,
+      "loss": 0.7898,
+      "step": 8970
+    },
+    {
+      "epoch": 0.6915664508171446,
+      "grad_norm": 3.7056403160095215,
+      "learning_rate": 2.2943139493566456e-06,
+      "loss": 1.0118,
+      "step": 8971
+    },
+    {
+      "epoch": 0.6916435399321615,
+      "grad_norm": 3.6872429847717285,
+      "learning_rate": 2.293264171581158e-06,
+      "loss": 0.999,
+      "step": 8972
+    },
+    {
+      "epoch": 0.6917206290471786,
+      "grad_norm": 3.7803478240966797,
+      "learning_rate": 2.2922145625566293e-06,
+      "loss": 0.9195,
+      "step": 8973
+    },
+    {
+      "epoch": 0.6917977181621955,
+      "grad_norm": 4.061209201812744,
+      "learning_rate": 2.2911651223484987e-06,
+      "loss": 0.9921,
+      "step": 8974
+    },
+    {
+      "epoch": 0.6918748072772125,
+      "grad_norm": 3.8408148288726807,
+      "learning_rate": 2.2901158510221934e-06,
+      "loss": 0.9547,
+      "step": 8975
+    },
+    {
+      "epoch": 0.6919518963922294,
+      "grad_norm": 4.113273620605469,
+      "learning_rate": 2.2890667486431296e-06,
+      "loss": 0.9756,
+      "step": 8976
+    },
+    {
+      "epoch": 0.6920289855072463,
+      "grad_norm": 3.689815044403076,
+      "learning_rate": 2.288017815276714e-06,
+      "loss": 0.8866,
+      "step": 8977
+    },
+    {
+      "epoch": 0.6921060746222634,
+      "grad_norm": 3.9249446392059326,
+      "learning_rate": 2.2869690509883435e-06,
+      "loss": 0.9308,
+      "step": 8978
+    },
+    {
+      "epoch": 0.6921831637372803,
+      "grad_norm": 3.9732306003570557,
+      "learning_rate": 2.2859204558433994e-06,
+      "loss": 0.8548,
+      "step": 8979
+    },
+    {
+      "epoch": 0.6922602528522973,
+      "grad_norm": 3.4670419692993164,
+      "learning_rate": 2.284872029907259e-06,
+      "loss": 0.9897,
+      "step": 8980
+    },
+    {
+      "epoch": 0.6923373419673142,
+      "grad_norm": 3.4217910766601562,
+      "learning_rate": 2.283823773245285e-06,
+      "loss": 0.8297,
+      "step": 8981
+    },
+    {
+      "epoch": 0.6924144310823311,
+      "grad_norm": 3.647320032119751,
+      "learning_rate": 2.2827756859228316e-06,
+      "loss": 0.8505,
+      "step": 8982
+    },
+    {
+      "epoch": 0.6924915201973482,
+      "grad_norm": 3.502896785736084,
+      "learning_rate": 2.281727768005243e-06,
+      "loss": 0.8633,
+      "step": 8983
+    },
+    {
+      "epoch": 0.6925686093123651,
+      "grad_norm": 3.6835694313049316,
+      "learning_rate": 2.2806800195578484e-06,
+      "loss": 0.9993,
+      "step": 8984
+    },
+    {
+      "epoch": 0.6926456984273821,
+      "grad_norm": 3.4930477142333984,
+      "learning_rate": 2.2796324406459703e-06,
+      "loss": 0.7549,
+      "step": 8985
+    },
+    {
+      "epoch": 0.692722787542399,
+      "grad_norm": 3.864713668823242,
+      "learning_rate": 2.2785850313349207e-06,
+      "loss": 0.9116,
+      "step": 8986
+    },
+    {
+      "epoch": 0.6927998766574159,
+      "grad_norm": 3.770353317260742,
+      "learning_rate": 2.277537791689999e-06,
+      "loss": 0.9246,
+      "step": 8987
+    },
+    {
+      "epoch": 0.692876965772433,
+      "grad_norm": 3.649230718612671,
+      "learning_rate": 2.276490721776497e-06,
+      "loss": 0.9455,
+      "step": 8988
+    },
+    {
+      "epoch": 0.6929540548874499,
+      "grad_norm": 3.8266777992248535,
+      "learning_rate": 2.2754438216596946e-06,
+      "loss": 0.9737,
+      "step": 8989
+    },
+    {
+      "epoch": 0.6930311440024669,
+      "grad_norm": 3.5182337760925293,
+      "learning_rate": 2.2743970914048575e-06,
+      "loss": 0.9442,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6931082331174838,
+      "grad_norm": 4.085440158843994,
+      "learning_rate": 2.273350531077245e-06,
+      "loss": 1.0016,
+      "step": 8991
+    },
+    {
+      "epoch": 0.6931853222325007,
+      "grad_norm": 3.814785957336426,
+      "learning_rate": 2.2723041407421055e-06,
+      "loss": 0.9506,
+      "step": 8992
+    },
+    {
+      "epoch": 0.6932624113475178,
+      "grad_norm": 3.8445160388946533,
+      "learning_rate": 2.2712579204646755e-06,
+      "loss": 0.9151,
+      "step": 8993
+    },
+    {
+      "epoch": 0.6933395004625347,
+      "grad_norm": 3.886320114135742,
+      "learning_rate": 2.270211870310184e-06,
+      "loss": 0.9457,
+      "step": 8994
+    },
+    {
+      "epoch": 0.6934165895775517,
+      "grad_norm": 4.0595011711120605,
+      "learning_rate": 2.269165990343842e-06,
+      "loss": 0.9247,
+      "step": 8995
+    },
+    {
+      "epoch": 0.6934936786925686,
+      "grad_norm": 3.6124463081359863,
+      "learning_rate": 2.2681202806308572e-06,
+      "loss": 0.8495,
+      "step": 8996
+    },
+    {
+      "epoch": 0.6935707678075855,
+      "grad_norm": 3.549788236618042,
+      "learning_rate": 2.2670747412364243e-06,
+      "loss": 0.8262,
+      "step": 8997
+    },
+    {
+      "epoch": 0.6936478569226026,
+      "grad_norm": 3.6086161136627197,
+      "learning_rate": 2.266029372225727e-06,
+      "loss": 0.9215,
+      "step": 8998
+    },
+    {
+      "epoch": 0.6937249460376195,
+      "grad_norm": 3.8933184146881104,
+      "learning_rate": 2.2649841736639387e-06,
+      "loss": 0.8623,
+      "step": 8999
+    },
+    {
+      "epoch": 0.6938020351526365,
+      "grad_norm": 3.9083821773529053,
+      "learning_rate": 2.2639391456162246e-06,
+      "loss": 1.0115,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6938791242676534,
+      "grad_norm": 3.4845428466796875,
+      "learning_rate": 2.262894288147732e-06,
+      "loss": 0.8953,
+      "step": 9001
+    },
+    {
+      "epoch": 0.6939562133826703,
+      "grad_norm": 3.57886004447937,
+      "learning_rate": 2.2618496013236046e-06,
+      "loss": 0.8971,
+      "step": 9002
+    },
+    {
+      "epoch": 0.6940333024976874,
+      "grad_norm": 3.5042362213134766,
+      "learning_rate": 2.2608050852089738e-06,
+      "loss": 0.8637,
+      "step": 9003
+    },
+    {
+      "epoch": 0.6941103916127043,
+      "grad_norm": 3.460904598236084,
+      "learning_rate": 2.2597607398689588e-06,
+      "loss": 0.8448,
+      "step": 9004
+    },
+    {
+      "epoch": 0.6941874807277213,
+      "grad_norm": 3.7049005031585693,
+      "learning_rate": 2.2587165653686714e-06,
+      "loss": 0.9066,
+      "step": 9005
+    },
+    {
+      "epoch": 0.6942645698427382,
+      "grad_norm": 3.4977948665618896,
+      "learning_rate": 2.257672561773207e-06,
+      "loss": 0.8732,
+      "step": 9006
+    },
+    {
+      "epoch": 0.6943416589577551,
+      "grad_norm": 4.0038676261901855,
+      "learning_rate": 2.256628729147654e-06,
+      "loss": 0.9104,
+      "step": 9007
+    },
+    {
+      "epoch": 0.6944187480727722,
+      "grad_norm": 3.465859889984131,
+      "learning_rate": 2.255585067557093e-06,
+      "loss": 0.8565,
+      "step": 9008
+    },
+    {
+      "epoch": 0.6944958371877891,
+      "grad_norm": 3.844653606414795,
+      "learning_rate": 2.254541577066589e-06,
+      "loss": 0.949,
+      "step": 9009
+    },
+    {
+      "epoch": 0.6945729263028061,
+      "grad_norm": 3.570842742919922,
+      "learning_rate": 2.2534982577412013e-06,
+      "loss": 0.9655,
+      "step": 9010
+    },
+    {
+      "epoch": 0.694650015417823,
+      "grad_norm": 3.9422519207000732,
+      "learning_rate": 2.2524551096459703e-06,
+      "loss": 0.9421,
+      "step": 9011
+    },
+    {
+      "epoch": 0.6947271045328399,
+      "grad_norm": 3.6970112323760986,
+      "learning_rate": 2.251412132845933e-06,
+      "loss": 0.9201,
+      "step": 9012
+    },
+    {
+      "epoch": 0.694804193647857,
+      "grad_norm": 3.5936856269836426,
+      "learning_rate": 2.2503693274061145e-06,
+      "loss": 0.8951,
+      "step": 9013
+    },
+    {
+      "epoch": 0.6948812827628739,
+      "grad_norm": 3.511066436767578,
+      "learning_rate": 2.2493266933915274e-06,
+      "loss": 0.845,
+      "step": 9014
+    },
+    {
+      "epoch": 0.6949583718778909,
+      "grad_norm": 3.7108609676361084,
+      "learning_rate": 2.2482842308671766e-06,
+      "loss": 0.97,
+      "step": 9015
+    },
+    {
+      "epoch": 0.6950354609929078,
+      "grad_norm": 3.6902565956115723,
+      "learning_rate": 2.247241939898051e-06,
+      "loss": 0.959,
+      "step": 9016
+    },
+    {
+      "epoch": 0.6951125501079247,
+      "grad_norm": 3.712344169616699,
+      "learning_rate": 2.2461998205491335e-06,
+      "loss": 0.9824,
+      "step": 9017
+    },
+    {
+      "epoch": 0.6951896392229417,
+      "grad_norm": 3.7560031414031982,
+      "learning_rate": 2.2451578728853944e-06,
+      "loss": 0.8713,
+      "step": 9018
+    },
+    {
+      "epoch": 0.6952667283379587,
+      "grad_norm": 3.6209821701049805,
+      "learning_rate": 2.244116096971795e-06,
+      "loss": 0.8807,
+      "step": 9019
+    },
+    {
+      "epoch": 0.6953438174529757,
+      "grad_norm": 3.638458490371704,
+      "learning_rate": 2.2430744928732854e-06,
+      "loss": 0.8903,
+      "step": 9020
+    },
+    {
+      "epoch": 0.6954209065679926,
+      "grad_norm": 3.5852041244506836,
+      "learning_rate": 2.242033060654801e-06,
+      "loss": 0.9105,
+      "step": 9021
+    },
+    {
+      "epoch": 0.6954979956830095,
+      "grad_norm": 3.6906344890594482,
+      "learning_rate": 2.240991800381271e-06,
+      "loss": 0.9511,
+      "step": 9022
+    },
+    {
+      "epoch": 0.6955750847980265,
+      "grad_norm": 3.609318971633911,
+      "learning_rate": 2.2399507121176127e-06,
+      "loss": 0.9687,
+      "step": 9023
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 3.8320932388305664,
+      "learning_rate": 2.238909795928734e-06,
+      "loss": 0.9885,
+      "step": 9024
+    },
+    {
+      "epoch": 0.6957292630280605,
+      "grad_norm": 3.6839687824249268,
+      "learning_rate": 2.2378690518795284e-06,
+      "loss": 0.8995,
+      "step": 9025
+    },
+    {
+      "epoch": 0.6958063521430774,
+      "grad_norm": 4.18240213394165,
+      "learning_rate": 2.2368284800348845e-06,
+      "loss": 0.9525,
+      "step": 9026
+    },
+    {
+      "epoch": 0.6958834412580943,
+      "grad_norm": 3.770507574081421,
+      "learning_rate": 2.2357880804596714e-06,
+      "loss": 0.9616,
+      "step": 9027
+    },
+    {
+      "epoch": 0.6959605303731113,
+      "grad_norm": 3.8755156993865967,
+      "learning_rate": 2.2347478532187562e-06,
+      "loss": 0.8957,
+      "step": 9028
+    },
+    {
+      "epoch": 0.6960376194881283,
+      "grad_norm": 3.3811306953430176,
+      "learning_rate": 2.2337077983769902e-06,
+      "loss": 0.8254,
+      "step": 9029
+    },
+    {
+      "epoch": 0.6961147086031453,
+      "grad_norm": 3.608715295791626,
+      "learning_rate": 2.232667915999216e-06,
+      "loss": 0.8726,
+      "step": 9030
+    },
+    {
+      "epoch": 0.6961917977181622,
+      "grad_norm": 3.5603222846984863,
+      "learning_rate": 2.2316282061502666e-06,
+      "loss": 0.9169,
+      "step": 9031
+    },
+    {
+      "epoch": 0.6962688868331791,
+      "grad_norm": 3.8038291931152344,
+      "learning_rate": 2.230588668894959e-06,
+      "loss": 0.874,
+      "step": 9032
+    },
+    {
+      "epoch": 0.6963459759481961,
+      "grad_norm": 3.717751979827881,
+      "learning_rate": 2.229549304298105e-06,
+      "loss": 0.9489,
+      "step": 9033
+    },
+    {
+      "epoch": 0.696423065063213,
+      "grad_norm": 4.030141830444336,
+      "learning_rate": 2.228510112424503e-06,
+      "loss": 0.9211,
+      "step": 9034
+    },
+    {
+      "epoch": 0.6965001541782301,
+      "grad_norm": 4.214062213897705,
+      "learning_rate": 2.2274710933389423e-06,
+      "loss": 0.9463,
+      "step": 9035
+    },
+    {
+      "epoch": 0.696577243293247,
+      "grad_norm": 3.714287519454956,
+      "learning_rate": 2.226432247106199e-06,
+      "loss": 0.9016,
+      "step": 9036
+    },
+    {
+      "epoch": 0.6966543324082639,
+      "grad_norm": 3.579857110977173,
+      "learning_rate": 2.225393573791042e-06,
+      "loss": 0.8215,
+      "step": 9037
+    },
+    {
+      "epoch": 0.6967314215232809,
+      "grad_norm": 3.9160430431365967,
+      "learning_rate": 2.2243550734582243e-06,
+      "loss": 0.8289,
+      "step": 9038
+    },
+    {
+      "epoch": 0.6968085106382979,
+      "grad_norm": 3.4613399505615234,
+      "learning_rate": 2.223316746172492e-06,
+      "loss": 0.8294,
+      "step": 9039
+    },
+    {
+      "epoch": 0.6968855997533149,
+      "grad_norm": 3.514765977859497,
+      "learning_rate": 2.22227859199858e-06,
+      "loss": 0.9109,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6969626888683318,
+      "grad_norm": 3.923628807067871,
+      "learning_rate": 2.2212406110012113e-06,
+      "loss": 0.9976,
+      "step": 9041
+    },
+    {
+      "epoch": 0.6970397779833487,
+      "grad_norm": 3.940729856491089,
+      "learning_rate": 2.2202028032451013e-06,
+      "loss": 0.9827,
+      "step": 9042
+    },
+    {
+      "epoch": 0.6971168670983657,
+      "grad_norm": 3.6197614669799805,
+      "learning_rate": 2.2191651687949473e-06,
+      "loss": 0.991,
+      "step": 9043
+    },
+    {
+      "epoch": 0.6971939562133826,
+      "grad_norm": 3.794126510620117,
+      "learning_rate": 2.218127707715443e-06,
+      "loss": 0.9555,
+      "step": 9044
+    },
+    {
+      "epoch": 0.6972710453283997,
+      "grad_norm": 3.6639597415924072,
+      "learning_rate": 2.2170904200712684e-06,
+      "loss": 0.968,
+      "step": 9045
+    },
+    {
+      "epoch": 0.6973481344434166,
+      "grad_norm": 3.9340016841888428,
+      "learning_rate": 2.216053305927093e-06,
+      "loss": 0.9202,
+      "step": 9046
+    },
+    {
+      "epoch": 0.6974252235584335,
+      "grad_norm": 3.675588846206665,
+      "learning_rate": 2.2150163653475752e-06,
+      "loss": 0.7814,
+      "step": 9047
+    },
+    {
+      "epoch": 0.6975023126734505,
+      "grad_norm": 3.601801633834839,
+      "learning_rate": 2.2139795983973654e-06,
+      "loss": 0.8856,
+      "step": 9048
+    },
+    {
+      "epoch": 0.6975794017884674,
+      "grad_norm": 3.440999746322632,
+      "learning_rate": 2.2129430051410967e-06,
+      "loss": 0.9017,
+      "step": 9049
+    },
+    {
+      "epoch": 0.6976564909034845,
+      "grad_norm": 3.5719244480133057,
+      "learning_rate": 2.211906585643397e-06,
+      "loss": 0.8295,
+      "step": 9050
+    },
+    {
+      "epoch": 0.6977335800185014,
+      "grad_norm": 4.056939601898193,
+      "learning_rate": 2.210870339968881e-06,
+      "loss": 0.9759,
+      "step": 9051
+    },
+    {
+      "epoch": 0.6978106691335183,
+      "grad_norm": 3.9863460063934326,
+      "learning_rate": 2.2098342681821555e-06,
+      "loss": 0.9698,
+      "step": 9052
+    },
+    {
+      "epoch": 0.6978877582485353,
+      "grad_norm": 3.962463140487671,
+      "learning_rate": 2.2087983703478118e-06,
+      "loss": 0.9996,
+      "step": 9053
+    },
+    {
+      "epoch": 0.6979648473635522,
+      "grad_norm": 3.7303974628448486,
+      "learning_rate": 2.207762646530434e-06,
+      "loss": 0.9972,
+      "step": 9054
+    },
+    {
+      "epoch": 0.6980419364785693,
+      "grad_norm": 3.459636688232422,
+      "learning_rate": 2.2067270967945936e-06,
+      "loss": 0.8566,
+      "step": 9055
+    },
+    {
+      "epoch": 0.6981190255935862,
+      "grad_norm": 3.582228183746338,
+      "learning_rate": 2.2056917212048522e-06,
+      "loss": 0.9435,
+      "step": 9056
+    },
+    {
+      "epoch": 0.6981961147086031,
+      "grad_norm": 3.553546905517578,
+      "learning_rate": 2.2046565198257595e-06,
+      "loss": 0.9016,
+      "step": 9057
+    },
+    {
+      "epoch": 0.6982732038236201,
+      "grad_norm": 3.933868646621704,
+      "learning_rate": 2.203621492721858e-06,
+      "loss": 0.9479,
+      "step": 9058
+    },
+    {
+      "epoch": 0.698350292938637,
+      "grad_norm": 3.995857000350952,
+      "learning_rate": 2.2025866399576713e-06,
+      "loss": 0.98,
+      "step": 9059
+    },
+    {
+      "epoch": 0.6984273820536541,
+      "grad_norm": 3.842726469039917,
+      "learning_rate": 2.2015519615977193e-06,
+      "loss": 0.9618,
+      "step": 9060
+    },
+    {
+      "epoch": 0.698504471168671,
+      "grad_norm": 3.7340540885925293,
+      "learning_rate": 2.2005174577065085e-06,
+      "loss": 0.9015,
+      "step": 9061
+    },
+    {
+      "epoch": 0.6985815602836879,
+      "grad_norm": 3.7171812057495117,
+      "learning_rate": 2.1994831283485363e-06,
+      "loss": 0.9996,
+      "step": 9062
+    },
+    {
+      "epoch": 0.6986586493987049,
+      "grad_norm": 3.576258420944214,
+      "learning_rate": 2.198448973588288e-06,
+      "loss": 0.8816,
+      "step": 9063
+    },
+    {
+      "epoch": 0.6987357385137218,
+      "grad_norm": 3.6319639682769775,
+      "learning_rate": 2.197414993490235e-06,
+      "loss": 0.9423,
+      "step": 9064
+    },
+    {
+      "epoch": 0.6988128276287389,
+      "grad_norm": 3.287196159362793,
+      "learning_rate": 2.1963811881188423e-06,
+      "loss": 0.8847,
+      "step": 9065
+    },
+    {
+      "epoch": 0.6988899167437558,
+      "grad_norm": 3.3368561267852783,
+      "learning_rate": 2.1953475575385618e-06,
+      "loss": 0.8119,
+      "step": 9066
+    },
+    {
+      "epoch": 0.6989670058587727,
+      "grad_norm": 3.6596055030822754,
+      "learning_rate": 2.1943141018138357e-06,
+      "loss": 1.0,
+      "step": 9067
+    },
+    {
+      "epoch": 0.6990440949737897,
+      "grad_norm": 3.594658851623535,
+      "learning_rate": 2.1932808210090963e-06,
+      "loss": 0.867,
+      "step": 9068
+    },
+    {
+      "epoch": 0.6991211840888066,
+      "grad_norm": 4.087553977966309,
+      "learning_rate": 2.1922477151887595e-06,
+      "loss": 0.915,
+      "step": 9069
+    },
+    {
+      "epoch": 0.6991982732038237,
+      "grad_norm": 3.692249059677124,
+      "learning_rate": 2.1912147844172354e-06,
+      "loss": 0.8633,
+      "step": 9070
+    },
+    {
+      "epoch": 0.6992753623188406,
+      "grad_norm": 3.7603821754455566,
+      "learning_rate": 2.1901820287589227e-06,
+      "loss": 0.8733,
+      "step": 9071
+    },
+    {
+      "epoch": 0.6993524514338575,
+      "grad_norm": 3.858159303665161,
+      "learning_rate": 2.189149448278208e-06,
+      "loss": 0.7904,
+      "step": 9072
+    },
+    {
+      "epoch": 0.6994295405488745,
+      "grad_norm": 3.873694896697998,
+      "learning_rate": 2.188117043039468e-06,
+      "loss": 0.9063,
+      "step": 9073
+    },
+    {
+      "epoch": 0.6995066296638914,
+      "grad_norm": 3.4330713748931885,
+      "learning_rate": 2.187084813107069e-06,
+      "loss": 0.9582,
+      "step": 9074
+    },
+    {
+      "epoch": 0.6995837187789085,
+      "grad_norm": 3.4164810180664062,
+      "learning_rate": 2.186052758545361e-06,
+      "loss": 0.8979,
+      "step": 9075
+    },
+    {
+      "epoch": 0.6996608078939254,
+      "grad_norm": 3.9505198001861572,
+      "learning_rate": 2.18502087941869e-06,
+      "loss": 0.9866,
+      "step": 9076
+    },
+    {
+      "epoch": 0.6997378970089423,
+      "grad_norm": 3.860844373703003,
+      "learning_rate": 2.183989175791388e-06,
+      "loss": 0.9133,
+      "step": 9077
+    },
+    {
+      "epoch": 0.6998149861239593,
+      "grad_norm": 3.6044540405273438,
+      "learning_rate": 2.1829576477277765e-06,
+      "loss": 0.9124,
+      "step": 9078
+    },
+    {
+      "epoch": 0.6998920752389762,
+      "grad_norm": 3.7315444946289062,
+      "learning_rate": 2.181926295292167e-06,
+      "loss": 0.9567,
+      "step": 9079
+    },
+    {
+      "epoch": 0.6999691643539933,
+      "grad_norm": 3.6717987060546875,
+      "learning_rate": 2.180895118548857e-06,
+      "loss": 0.9217,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7000462534690102,
+      "grad_norm": 4.249053478240967,
+      "learning_rate": 2.1798641175621354e-06,
+      "loss": 1.0054,
+      "step": 9081
+    },
+    {
+      "epoch": 0.7001233425840271,
+      "grad_norm": 3.520657539367676,
+      "learning_rate": 2.17883329239628e-06,
+      "loss": 0.8778,
+      "step": 9082
+    },
+    {
+      "epoch": 0.7002004316990441,
+      "grad_norm": 3.7700884342193604,
+      "learning_rate": 2.177802643115558e-06,
+      "loss": 0.9893,
+      "step": 9083
+    },
+    {
+      "epoch": 0.700277520814061,
+      "grad_norm": 3.6206729412078857,
+      "learning_rate": 2.1767721697842244e-06,
+      "loss": 0.8638,
+      "step": 9084
+    },
+    {
+      "epoch": 0.700354609929078,
+      "grad_norm": 3.6218421459198,
+      "learning_rate": 2.1757418724665263e-06,
+      "loss": 0.9983,
+      "step": 9085
+    },
+    {
+      "epoch": 0.700431699044095,
+      "grad_norm": 3.8477892875671387,
+      "learning_rate": 2.1747117512266928e-06,
+      "loss": 1.0336,
+      "step": 9086
+    },
+    {
+      "epoch": 0.7005087881591119,
+      "grad_norm": 3.7004141807556152,
+      "learning_rate": 2.1736818061289492e-06,
+      "loss": 0.8383,
+      "step": 9087
+    },
+    {
+      "epoch": 0.7005858772741289,
+      "grad_norm": 3.667736291885376,
+      "learning_rate": 2.1726520372375076e-06,
+      "loss": 0.9925,
+      "step": 9088
+    },
+    {
+      "epoch": 0.7006629663891458,
+      "grad_norm": 3.574789047241211,
+      "learning_rate": 2.1716224446165678e-06,
+      "loss": 0.7904,
+      "step": 9089
+    },
+    {
+      "epoch": 0.7007400555041629,
+      "grad_norm": 3.753038167953491,
+      "learning_rate": 2.1705930283303222e-06,
+      "loss": 0.8674,
+      "step": 9090
+    },
+    {
+      "epoch": 0.7008171446191798,
+      "grad_norm": 3.9094438552856445,
+      "learning_rate": 2.1695637884429456e-06,
+      "loss": 0.9338,
+      "step": 9091
+    },
+    {
+      "epoch": 0.7008942337341967,
+      "grad_norm": 3.6531260013580322,
+      "learning_rate": 2.1685347250186073e-06,
+      "loss": 0.9804,
+      "step": 9092
+    },
+    {
+      "epoch": 0.7009713228492137,
+      "grad_norm": 3.6628949642181396,
+      "learning_rate": 2.1675058381214647e-06,
+      "loss": 0.9991,
+      "step": 9093
+    },
+    {
+      "epoch": 0.7010484119642306,
+      "grad_norm": 3.8137435913085938,
+      "learning_rate": 2.166477127815663e-06,
+      "loss": 0.8553,
+      "step": 9094
+    },
+    {
+      "epoch": 0.7011255010792476,
+      "grad_norm": 3.7563507556915283,
+      "learning_rate": 2.1654485941653374e-06,
+      "loss": 0.8153,
+      "step": 9095
+    },
+    {
+      "epoch": 0.7012025901942646,
+      "grad_norm": 3.9121322631835938,
+      "learning_rate": 2.164420237234611e-06,
+      "loss": 0.9299,
+      "step": 9096
+    },
+    {
+      "epoch": 0.7012796793092815,
+      "grad_norm": 3.449474811553955,
+      "learning_rate": 2.163392057087597e-06,
+      "loss": 0.9556,
+      "step": 9097
+    },
+    {
+      "epoch": 0.7013567684242985,
+      "grad_norm": 3.6520328521728516,
+      "learning_rate": 2.1623640537883977e-06,
+      "loss": 0.8858,
+      "step": 9098
+    },
+    {
+      "epoch": 0.7014338575393154,
+      "grad_norm": 3.4059646129608154,
+      "learning_rate": 2.1613362274011025e-06,
+      "loss": 0.8044,
+      "step": 9099
+    },
+    {
+      "epoch": 0.7015109466543324,
+      "grad_norm": 3.759131669998169,
+      "learning_rate": 2.1603085779897935e-06,
+      "loss": 0.8184,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7015880357693494,
+      "grad_norm": 3.839765787124634,
+      "learning_rate": 2.1592811056185363e-06,
+      "loss": 0.9364,
+      "step": 9101
+    },
+    {
+      "epoch": 0.7016651248843663,
+      "grad_norm": 3.4155290126800537,
+      "learning_rate": 2.1582538103513896e-06,
+      "loss": 0.835,
+      "step": 9102
+    },
+    {
+      "epoch": 0.7017422139993833,
+      "grad_norm": 3.8770790100097656,
+      "learning_rate": 2.1572266922524e-06,
+      "loss": 1.0204,
+      "step": 9103
+    },
+    {
+      "epoch": 0.7018193031144002,
+      "grad_norm": 3.4858505725860596,
+      "learning_rate": 2.1561997513856027e-06,
+      "loss": 0.9334,
+      "step": 9104
+    },
+    {
+      "epoch": 0.7018963922294172,
+      "grad_norm": 3.834233283996582,
+      "learning_rate": 2.155172987815024e-06,
+      "loss": 0.9136,
+      "step": 9105
+    },
+    {
+      "epoch": 0.7019734813444342,
+      "grad_norm": 4.0034050941467285,
+      "learning_rate": 2.1541464016046747e-06,
+      "loss": 1.0584,
+      "step": 9106
+    },
+    {
+      "epoch": 0.7020505704594511,
+      "grad_norm": 3.6104769706726074,
+      "learning_rate": 2.153119992818558e-06,
+      "loss": 0.9131,
+      "step": 9107
+    },
+    {
+      "epoch": 0.7021276595744681,
+      "grad_norm": 3.7609310150146484,
+      "learning_rate": 2.152093761520665e-06,
+      "loss": 0.9231,
+      "step": 9108
+    },
+    {
+      "epoch": 0.702204748689485,
+      "grad_norm": 3.517526388168335,
+      "learning_rate": 2.151067707774977e-06,
+      "loss": 0.9278,
+      "step": 9109
+    },
+    {
+      "epoch": 0.702281837804502,
+      "grad_norm": 3.942485809326172,
+      "learning_rate": 2.150041831645462e-06,
+      "loss": 1.0325,
+      "step": 9110
+    },
+    {
+      "epoch": 0.702358926919519,
+      "grad_norm": 3.776834487915039,
+      "learning_rate": 2.14901613319608e-06,
+      "loss": 0.9948,
+      "step": 9111
+    },
+    {
+      "epoch": 0.7024360160345359,
+      "grad_norm": 3.855884552001953,
+      "learning_rate": 2.147990612490775e-06,
+      "loss": 1.0399,
+      "step": 9112
+    },
+    {
+      "epoch": 0.7025131051495529,
+      "grad_norm": 4.162666320800781,
+      "learning_rate": 2.1469652695934847e-06,
+      "loss": 0.8753,
+      "step": 9113
+    },
+    {
+      "epoch": 0.7025901942645698,
+      "grad_norm": 3.925414800643921,
+      "learning_rate": 2.145940104568133e-06,
+      "loss": 0.9324,
+      "step": 9114
+    },
+    {
+      "epoch": 0.7026672833795868,
+      "grad_norm": 3.718440294265747,
+      "learning_rate": 2.1449151174786343e-06,
+      "loss": 0.8593,
+      "step": 9115
+    },
+    {
+      "epoch": 0.7027443724946038,
+      "grad_norm": 3.670015335083008,
+      "learning_rate": 2.143890308388893e-06,
+      "loss": 0.7962,
+      "step": 9116
+    },
+    {
+      "epoch": 0.7028214616096207,
+      "grad_norm": 3.8344802856445312,
+      "learning_rate": 2.142865677362797e-06,
+      "loss": 0.9722,
+      "step": 9117
+    },
+    {
+      "epoch": 0.7028985507246377,
+      "grad_norm": 3.786147117614746,
+      "learning_rate": 2.141841224464229e-06,
+      "loss": 0.8701,
+      "step": 9118
+    },
+    {
+      "epoch": 0.7029756398396546,
+      "grad_norm": 3.831787109375,
+      "learning_rate": 2.1408169497570576e-06,
+      "loss": 0.8794,
+      "step": 9119
+    },
+    {
+      "epoch": 0.7030527289546716,
+      "grad_norm": 3.9900169372558594,
+      "learning_rate": 2.139792853305141e-06,
+      "loss": 1.055,
+      "step": 9120
+    },
+    {
+      "epoch": 0.7031298180696886,
+      "grad_norm": 3.954350233078003,
+      "learning_rate": 2.1387689351723266e-06,
+      "loss": 0.952,
+      "step": 9121
+    },
+    {
+      "epoch": 0.7032069071847055,
+      "grad_norm": 3.8101086616516113,
+      "learning_rate": 2.1377451954224526e-06,
+      "loss": 0.94,
+      "step": 9122
+    },
+    {
+      "epoch": 0.7032839962997225,
+      "grad_norm": 4.078210353851318,
+      "learning_rate": 2.13672163411934e-06,
+      "loss": 0.9399,
+      "step": 9123
+    },
+    {
+      "epoch": 0.7033610854147394,
+      "grad_norm": 3.5562384128570557,
+      "learning_rate": 2.1356982513268034e-06,
+      "loss": 0.806,
+      "step": 9124
+    },
+    {
+      "epoch": 0.7034381745297564,
+      "grad_norm": 3.3538687229156494,
+      "learning_rate": 2.134675047108647e-06,
+      "loss": 0.8706,
+      "step": 9125
+    },
+    {
+      "epoch": 0.7035152636447733,
+      "grad_norm": 4.081210136413574,
+      "learning_rate": 2.133652021528661e-06,
+      "loss": 0.8758,
+      "step": 9126
+    },
+    {
+      "epoch": 0.7035923527597903,
+      "grad_norm": 4.063713550567627,
+      "learning_rate": 2.1326291746506283e-06,
+      "loss": 1.0671,
+      "step": 9127
+    },
+    {
+      "epoch": 0.7036694418748073,
+      "grad_norm": 3.4613239765167236,
+      "learning_rate": 2.131606506538314e-06,
+      "loss": 0.8831,
+      "step": 9128
+    },
+    {
+      "epoch": 0.7037465309898242,
+      "grad_norm": 3.84372878074646,
+      "learning_rate": 2.1305840172554786e-06,
+      "loss": 1.0199,
+      "step": 9129
+    },
+    {
+      "epoch": 0.7038236201048412,
+      "grad_norm": 3.5245513916015625,
+      "learning_rate": 2.1295617068658685e-06,
+      "loss": 0.8048,
+      "step": 9130
+    },
+    {
+      "epoch": 0.7039007092198581,
+      "grad_norm": 3.294621706008911,
+      "learning_rate": 2.1285395754332198e-06,
+      "loss": 0.8333,
+      "step": 9131
+    },
+    {
+      "epoch": 0.7039777983348752,
+      "grad_norm": 3.9054811000823975,
+      "learning_rate": 2.1275176230212568e-06,
+      "loss": 0.8569,
+      "step": 9132
+    },
+    {
+      "epoch": 0.7040548874498921,
+      "grad_norm": 3.4768083095550537,
+      "learning_rate": 2.1264958496936945e-06,
+      "loss": 0.8138,
+      "step": 9133
+    },
+    {
+      "epoch": 0.704131976564909,
+      "grad_norm": 3.709878921508789,
+      "learning_rate": 2.125474255514232e-06,
+      "loss": 0.925,
+      "step": 9134
+    },
+    {
+      "epoch": 0.704209065679926,
+      "grad_norm": 3.6700382232666016,
+      "learning_rate": 2.1244528405465626e-06,
+      "loss": 0.9451,
+      "step": 9135
+    },
+    {
+      "epoch": 0.7042861547949429,
+      "grad_norm": 3.370394468307495,
+      "learning_rate": 2.123431604854365e-06,
+      "loss": 0.8803,
+      "step": 9136
+    },
+    {
+      "epoch": 0.70436324390996,
+      "grad_norm": 3.7478532791137695,
+      "learning_rate": 2.1224105485013096e-06,
+      "loss": 0.8755,
+      "step": 9137
+    },
+    {
+      "epoch": 0.7044403330249769,
+      "grad_norm": 3.732043981552124,
+      "learning_rate": 2.121389671551054e-06,
+      "loss": 0.9241,
+      "step": 9138
+    },
+    {
+      "epoch": 0.7045174221399938,
+      "grad_norm": 3.701848268508911,
+      "learning_rate": 2.120368974067242e-06,
+      "loss": 0.9204,
+      "step": 9139
+    },
+    {
+      "epoch": 0.7045945112550108,
+      "grad_norm": 3.8429481983184814,
+      "learning_rate": 2.11934845611351e-06,
+      "loss": 1.0072,
+      "step": 9140
+    },
+    {
+      "epoch": 0.7046716003700277,
+      "grad_norm": 3.9448812007904053,
+      "learning_rate": 2.1183281177534828e-06,
+      "loss": 0.945,
+      "step": 9141
+    },
+    {
+      "epoch": 0.7047486894850448,
+      "grad_norm": 3.5992908477783203,
+      "learning_rate": 2.1173079590507735e-06,
+      "loss": 0.9431,
+      "step": 9142
+    },
+    {
+      "epoch": 0.7048257786000617,
+      "grad_norm": 3.7120702266693115,
+      "learning_rate": 2.1162879800689845e-06,
+      "loss": 0.7797,
+      "step": 9143
+    },
+    {
+      "epoch": 0.7049028677150786,
+      "grad_norm": 3.8240060806274414,
+      "learning_rate": 2.1152681808717033e-06,
+      "loss": 0.9258,
+      "step": 9144
+    },
+    {
+      "epoch": 0.7049799568300956,
+      "grad_norm": 3.821179151535034,
+      "learning_rate": 2.1142485615225104e-06,
+      "loss": 0.869,
+      "step": 9145
+    },
+    {
+      "epoch": 0.7050570459451125,
+      "grad_norm": 4.0263142585754395,
+      "learning_rate": 2.1132291220849743e-06,
+      "loss": 0.9052,
+      "step": 9146
+    },
+    {
+      "epoch": 0.7051341350601296,
+      "grad_norm": 3.7679524421691895,
+      "learning_rate": 2.1122098626226512e-06,
+      "loss": 1.0384,
+      "step": 9147
+    },
+    {
+      "epoch": 0.7052112241751465,
+      "grad_norm": 3.6580440998077393,
+      "learning_rate": 2.111190783199089e-06,
+      "loss": 0.9052,
+      "step": 9148
+    },
+    {
+      "epoch": 0.7052883132901634,
+      "grad_norm": 3.7239935398101807,
+      "learning_rate": 2.110171883877818e-06,
+      "loss": 0.9718,
+      "step": 9149
+    },
+    {
+      "epoch": 0.7053654024051804,
+      "grad_norm": 4.641191482543945,
+      "learning_rate": 2.1091531647223632e-06,
+      "loss": 1.015,
+      "step": 9150
+    },
+    {
+      "epoch": 0.7054424915201973,
+      "grad_norm": 3.7391200065612793,
+      "learning_rate": 2.1081346257962367e-06,
+      "loss": 0.9439,
+      "step": 9151
+    },
+    {
+      "epoch": 0.7055195806352144,
+      "grad_norm": 3.7103328704833984,
+      "learning_rate": 2.1071162671629386e-06,
+      "loss": 0.9708,
+      "step": 9152
+    },
+    {
+      "epoch": 0.7055966697502313,
+      "grad_norm": 3.7474148273468018,
+      "learning_rate": 2.1060980888859608e-06,
+      "loss": 0.9439,
+      "step": 9153
+    },
+    {
+      "epoch": 0.7056737588652482,
+      "grad_norm": 3.799514055252075,
+      "learning_rate": 2.1050800910287774e-06,
+      "loss": 0.961,
+      "step": 9154
+    },
+    {
+      "epoch": 0.7057508479802652,
+      "grad_norm": 3.7274961471557617,
+      "learning_rate": 2.1040622736548566e-06,
+      "loss": 0.871,
+      "step": 9155
+    },
+    {
+      "epoch": 0.7058279370952821,
+      "grad_norm": 4.0381669998168945,
+      "learning_rate": 2.1030446368276547e-06,
+      "loss": 0.8929,
+      "step": 9156
+    },
+    {
+      "epoch": 0.7059050262102992,
+      "grad_norm": 3.871047019958496,
+      "learning_rate": 2.102027180610616e-06,
+      "loss": 0.9771,
+      "step": 9157
+    },
+    {
+      "epoch": 0.7059821153253161,
+      "grad_norm": 3.7191429138183594,
+      "learning_rate": 2.1010099050671745e-06,
+      "loss": 0.9002,
+      "step": 9158
+    },
+    {
+      "epoch": 0.706059204440333,
+      "grad_norm": 3.723458766937256,
+      "learning_rate": 2.099992810260752e-06,
+      "loss": 0.9915,
+      "step": 9159
+    },
+    {
+      "epoch": 0.70613629355535,
+      "grad_norm": 4.22058629989624,
+      "learning_rate": 2.098975896254757e-06,
+      "loss": 1.0003,
+      "step": 9160
+    },
+    {
+      "epoch": 0.7062133826703669,
+      "grad_norm": 3.7381343841552734,
+      "learning_rate": 2.0979591631125896e-06,
+      "loss": 0.9054,
+      "step": 9161
+    },
+    {
+      "epoch": 0.706290471785384,
+      "grad_norm": 3.518643379211426,
+      "learning_rate": 2.096942610897639e-06,
+      "loss": 0.9081,
+      "step": 9162
+    },
+    {
+      "epoch": 0.7063675609004009,
+      "grad_norm": 3.9908714294433594,
+      "learning_rate": 2.0959262396732804e-06,
+      "loss": 0.9465,
+      "step": 9163
+    },
+    {
+      "epoch": 0.7064446500154178,
+      "grad_norm": 3.8830153942108154,
+      "learning_rate": 2.094910049502882e-06,
+      "loss": 0.8796,
+      "step": 9164
+    },
+    {
+      "epoch": 0.7065217391304348,
+      "grad_norm": 3.6630706787109375,
+      "learning_rate": 2.093894040449795e-06,
+      "loss": 0.8835,
+      "step": 9165
+    },
+    {
+      "epoch": 0.7065988282454517,
+      "grad_norm": 3.9855260848999023,
+      "learning_rate": 2.0928782125773636e-06,
+      "loss": 1.0637,
+      "step": 9166
+    },
+    {
+      "epoch": 0.7066759173604688,
+      "grad_norm": 3.887150764465332,
+      "learning_rate": 2.091862565948919e-06,
+      "loss": 0.9717,
+      "step": 9167
+    },
+    {
+      "epoch": 0.7067530064754857,
+      "grad_norm": 3.4647603034973145,
+      "learning_rate": 2.0908471006277816e-06,
+      "loss": 0.92,
+      "step": 9168
+    },
+    {
+      "epoch": 0.7068300955905026,
+      "grad_norm": 3.9277379512786865,
+      "learning_rate": 2.089831816677261e-06,
+      "loss": 0.9405,
+      "step": 9169
+    },
+    {
+      "epoch": 0.7069071847055196,
+      "grad_norm": 3.964707136154175,
+      "learning_rate": 2.088816714160656e-06,
+      "loss": 1.0286,
+      "step": 9170
+    },
+    {
+      "epoch": 0.7069842738205365,
+      "grad_norm": 3.6983249187469482,
+      "learning_rate": 2.0878017931412493e-06,
+      "loss": 1.0521,
+      "step": 9171
+    },
+    {
+      "epoch": 0.7070613629355536,
+      "grad_norm": 3.797630548477173,
+      "learning_rate": 2.0867870536823185e-06,
+      "loss": 0.9313,
+      "step": 9172
+    },
+    {
+      "epoch": 0.7071384520505705,
+      "grad_norm": 3.392491340637207,
+      "learning_rate": 2.0857724958471273e-06,
+      "loss": 0.742,
+      "step": 9173
+    },
+    {
+      "epoch": 0.7072155411655874,
+      "grad_norm": 3.4460275173187256,
+      "learning_rate": 2.0847581196989277e-06,
+      "loss": 0.8896,
+      "step": 9174
+    },
+    {
+      "epoch": 0.7072926302806044,
+      "grad_norm": 4.643573760986328,
+      "learning_rate": 2.0837439253009623e-06,
+      "loss": 0.9268,
+      "step": 9175
+    },
+    {
+      "epoch": 0.7073697193956213,
+      "grad_norm": 3.4909539222717285,
+      "learning_rate": 2.0827299127164574e-06,
+      "loss": 0.9444,
+      "step": 9176
+    },
+    {
+      "epoch": 0.7074468085106383,
+      "grad_norm": 3.7072372436523438,
+      "learning_rate": 2.0817160820086342e-06,
+      "loss": 0.9454,
+      "step": 9177
+    },
+    {
+      "epoch": 0.7075238976256553,
+      "grad_norm": 3.618695020675659,
+      "learning_rate": 2.080702433240699e-06,
+      "loss": 0.9146,
+      "step": 9178
+    },
+    {
+      "epoch": 0.7076009867406722,
+      "grad_norm": 3.594953775405884,
+      "learning_rate": 2.079688966475847e-06,
+      "loss": 0.9275,
+      "step": 9179
+    },
+    {
+      "epoch": 0.7076780758556892,
+      "grad_norm": 3.649266004562378,
+      "learning_rate": 2.078675681777264e-06,
+      "loss": 0.8821,
+      "step": 9180
+    },
+    {
+      "epoch": 0.7077551649707061,
+      "grad_norm": 3.623678207397461,
+      "learning_rate": 2.077662579208124e-06,
+      "loss": 0.9072,
+      "step": 9181
+    },
+    {
+      "epoch": 0.7078322540857231,
+      "grad_norm": 3.6807639598846436,
+      "learning_rate": 2.0766496588315853e-06,
+      "loss": 0.9662,
+      "step": 9182
+    },
+    {
+      "epoch": 0.7079093432007401,
+      "grad_norm": 3.5517539978027344,
+      "learning_rate": 2.0756369207107997e-06,
+      "loss": 0.8944,
+      "step": 9183
+    },
+    {
+      "epoch": 0.707986432315757,
+      "grad_norm": 3.9906022548675537,
+      "learning_rate": 2.0746243649089065e-06,
+      "loss": 0.8717,
+      "step": 9184
+    },
+    {
+      "epoch": 0.708063521430774,
+      "grad_norm": 3.7278897762298584,
+      "learning_rate": 2.0736119914890335e-06,
+      "loss": 0.9515,
+      "step": 9185
+    },
+    {
+      "epoch": 0.7081406105457909,
+      "grad_norm": 4.302172660827637,
+      "learning_rate": 2.072599800514296e-06,
+      "loss": 0.9035,
+      "step": 9186
+    },
+    {
+      "epoch": 0.7082176996608079,
+      "grad_norm": 4.022290229797363,
+      "learning_rate": 2.0715877920478e-06,
+      "loss": 0.8397,
+      "step": 9187
+    },
+    {
+      "epoch": 0.7082947887758249,
+      "grad_norm": 3.2679741382598877,
+      "learning_rate": 2.0705759661526387e-06,
+      "loss": 0.7864,
+      "step": 9188
+    },
+    {
+      "epoch": 0.7083718778908418,
+      "grad_norm": 3.836318254470825,
+      "learning_rate": 2.069564322891894e-06,
+      "loss": 0.9187,
+      "step": 9189
+    },
+    {
+      "epoch": 0.7084489670058588,
+      "grad_norm": 4.069839000701904,
+      "learning_rate": 2.068552862328637e-06,
+      "loss": 1.0155,
+      "step": 9190
+    },
+    {
+      "epoch": 0.7085260561208757,
+      "grad_norm": 3.7692060470581055,
+      "learning_rate": 2.067541584525927e-06,
+      "loss": 0.8143,
+      "step": 9191
+    },
+    {
+      "epoch": 0.7086031452358927,
+      "grad_norm": 3.5922646522521973,
+      "learning_rate": 2.0665304895468114e-06,
+      "loss": 0.8584,
+      "step": 9192
+    },
+    {
+      "epoch": 0.7086802343509097,
+      "grad_norm": 3.649376630783081,
+      "learning_rate": 2.065519577454326e-06,
+      "loss": 0.9411,
+      "step": 9193
+    },
+    {
+      "epoch": 0.7087573234659266,
+      "grad_norm": 3.5520145893096924,
+      "learning_rate": 2.0645088483114974e-06,
+      "loss": 0.8421,
+      "step": 9194
+    },
+    {
+      "epoch": 0.7088344125809436,
+      "grad_norm": 4.203880310058594,
+      "learning_rate": 2.0634983021813385e-06,
+      "loss": 0.9849,
+      "step": 9195
+    },
+    {
+      "epoch": 0.7089115016959605,
+      "grad_norm": 3.6634862422943115,
+      "learning_rate": 2.062487939126854e-06,
+      "loss": 0.9275,
+      "step": 9196
+    },
+    {
+      "epoch": 0.7089885908109775,
+      "grad_norm": 3.71562123298645,
+      "learning_rate": 2.0614777592110306e-06,
+      "loss": 0.8916,
+      "step": 9197
+    },
+    {
+      "epoch": 0.7090656799259945,
+      "grad_norm": 3.9342494010925293,
+      "learning_rate": 2.06046776249685e-06,
+      "loss": 0.8933,
+      "step": 9198
+    },
+    {
+      "epoch": 0.7091427690410114,
+      "grad_norm": 3.345992088317871,
+      "learning_rate": 2.0594579490472803e-06,
+      "loss": 0.8077,
+      "step": 9199
+    },
+    {
+      "epoch": 0.7092198581560284,
+      "grad_norm": 3.8320610523223877,
+      "learning_rate": 2.058448318925278e-06,
+      "loss": 0.9362,
+      "step": 9200
+    },
+    {
+      "epoch": 0.7092969472710453,
+      "grad_norm": 3.6400604248046875,
+      "learning_rate": 2.0574388721937905e-06,
+      "loss": 0.8823,
+      "step": 9201
+    },
+    {
+      "epoch": 0.7093740363860623,
+      "grad_norm": 3.9985427856445312,
+      "learning_rate": 2.056429608915747e-06,
+      "loss": 1.0019,
+      "step": 9202
+    },
+    {
+      "epoch": 0.7094511255010792,
+      "grad_norm": 3.3883984088897705,
+      "learning_rate": 2.0554205291540724e-06,
+      "loss": 0.8021,
+      "step": 9203
+    },
+    {
+      "epoch": 0.7095282146160962,
+      "grad_norm": 3.7378809452056885,
+      "learning_rate": 2.0544116329716773e-06,
+      "loss": 0.8293,
+      "step": 9204
+    },
+    {
+      "epoch": 0.7096053037311132,
+      "grad_norm": 3.601414918899536,
+      "learning_rate": 2.0534029204314613e-06,
+      "loss": 0.8926,
+      "step": 9205
+    },
+    {
+      "epoch": 0.7096823928461301,
+      "grad_norm": 3.2823150157928467,
+      "learning_rate": 2.052394391596313e-06,
+      "loss": 0.7918,
+      "step": 9206
+    },
+    {
+      "epoch": 0.7097594819611471,
+      "grad_norm": 4.111335277557373,
+      "learning_rate": 2.0513860465291097e-06,
+      "loss": 0.8922,
+      "step": 9207
+    },
+    {
+      "epoch": 0.709836571076164,
+      "grad_norm": 3.9760074615478516,
+      "learning_rate": 2.0503778852927134e-06,
+      "loss": 0.9042,
+      "step": 9208
+    },
+    {
+      "epoch": 0.709913660191181,
+      "grad_norm": 3.659054756164551,
+      "learning_rate": 2.0493699079499797e-06,
+      "loss": 0.8971,
+      "step": 9209
+    },
+    {
+      "epoch": 0.709990749306198,
+      "grad_norm": 3.8829121589660645,
+      "learning_rate": 2.04836211456375e-06,
+      "loss": 0.8574,
+      "step": 9210
+    },
+    {
+      "epoch": 0.7100678384212149,
+      "grad_norm": 3.769808292388916,
+      "learning_rate": 2.0473545051968557e-06,
+      "loss": 0.777,
+      "step": 9211
+    },
+    {
+      "epoch": 0.7101449275362319,
+      "grad_norm": 3.7595462799072266,
+      "learning_rate": 2.0463470799121177e-06,
+      "loss": 0.8837,
+      "step": 9212
+    },
+    {
+      "epoch": 0.7102220166512488,
+      "grad_norm": 3.825510025024414,
+      "learning_rate": 2.04533983877234e-06,
+      "loss": 0.9327,
+      "step": 9213
+    },
+    {
+      "epoch": 0.7102991057662658,
+      "grad_norm": 3.5481746196746826,
+      "learning_rate": 2.0443327818403213e-06,
+      "loss": 0.9789,
+      "step": 9214
+    },
+    {
+      "epoch": 0.7103761948812828,
+      "grad_norm": 3.591843366622925,
+      "learning_rate": 2.0433259091788453e-06,
+      "loss": 0.9067,
+      "step": 9215
+    },
+    {
+      "epoch": 0.7104532839962997,
+      "grad_norm": 4.088891983032227,
+      "learning_rate": 2.042319220850686e-06,
+      "loss": 0.9353,
+      "step": 9216
+    },
+    {
+      "epoch": 0.7105303731113167,
+      "grad_norm": 3.6502561569213867,
+      "learning_rate": 2.0413127169186053e-06,
+      "loss": 0.9367,
+      "step": 9217
+    },
+    {
+      "epoch": 0.7106074622263336,
+      "grad_norm": 3.281320810317993,
+      "learning_rate": 2.0403063974453547e-06,
+      "loss": 0.8194,
+      "step": 9218
+    },
+    {
+      "epoch": 0.7106845513413506,
+      "grad_norm": 3.4343175888061523,
+      "learning_rate": 2.03930026249367e-06,
+      "loss": 0.7826,
+      "step": 9219
+    },
+    {
+      "epoch": 0.7107616404563676,
+      "grad_norm": 3.537015438079834,
+      "learning_rate": 2.03829431212628e-06,
+      "loss": 0.9891,
+      "step": 9220
+    },
+    {
+      "epoch": 0.7108387295713845,
+      "grad_norm": 3.6004459857940674,
+      "learning_rate": 2.0372885464059004e-06,
+      "loss": 0.9879,
+      "step": 9221
+    },
+    {
+      "epoch": 0.7109158186864015,
+      "grad_norm": 3.736265182495117,
+      "learning_rate": 2.036282965395236e-06,
+      "loss": 0.8828,
+      "step": 9222
+    },
+    {
+      "epoch": 0.7109929078014184,
+      "grad_norm": 3.9425337314605713,
+      "learning_rate": 2.035277569156981e-06,
+      "loss": 1.03,
+      "step": 9223
+    },
+    {
+      "epoch": 0.7110699969164354,
+      "grad_norm": 3.6420586109161377,
+      "learning_rate": 2.0342723577538125e-06,
+      "loss": 1.0069,
+      "step": 9224
+    },
+    {
+      "epoch": 0.7111470860314524,
+      "grad_norm": 3.814761161804199,
+      "learning_rate": 2.0332673312484037e-06,
+      "loss": 1.0226,
+      "step": 9225
+    },
+    {
+      "epoch": 0.7112241751464693,
+      "grad_norm": 3.7190582752227783,
+      "learning_rate": 2.0322624897034102e-06,
+      "loss": 0.8863,
+      "step": 9226
+    },
+    {
+      "epoch": 0.7113012642614863,
+      "grad_norm": 3.9217827320098877,
+      "learning_rate": 2.0312578331814812e-06,
+      "loss": 0.9611,
+      "step": 9227
+    },
+    {
+      "epoch": 0.7113783533765032,
+      "grad_norm": 3.8105664253234863,
+      "learning_rate": 2.030253361745251e-06,
+      "loss": 0.9657,
+      "step": 9228
+    },
+    {
+      "epoch": 0.7114554424915202,
+      "grad_norm": 3.696140766143799,
+      "learning_rate": 2.0292490754573425e-06,
+      "loss": 0.9549,
+      "step": 9229
+    },
+    {
+      "epoch": 0.7115325316065372,
+      "grad_norm": 3.7637147903442383,
+      "learning_rate": 2.0282449743803684e-06,
+      "loss": 0.9717,
+      "step": 9230
+    },
+    {
+      "epoch": 0.7116096207215541,
+      "grad_norm": 3.541837692260742,
+      "learning_rate": 2.0272410585769284e-06,
+      "loss": 0.8398,
+      "step": 9231
+    },
+    {
+      "epoch": 0.7116867098365711,
+      "grad_norm": 3.7558746337890625,
+      "learning_rate": 2.0262373281096133e-06,
+      "loss": 0.9397,
+      "step": 9232
+    },
+    {
+      "epoch": 0.711763798951588,
+      "grad_norm": 3.951739549636841,
+      "learning_rate": 2.025233783041e-06,
+      "loss": 1.0248,
+      "step": 9233
+    },
+    {
+      "epoch": 0.711840888066605,
+      "grad_norm": 3.583740472793579,
+      "learning_rate": 2.0242304234336525e-06,
+      "loss": 0.9207,
+      "step": 9234
+    },
+    {
+      "epoch": 0.711917977181622,
+      "grad_norm": 3.8319308757781982,
+      "learning_rate": 2.0232272493501253e-06,
+      "loss": 0.9674,
+      "step": 9235
+    },
+    {
+      "epoch": 0.7119950662966389,
+      "grad_norm": 3.4977939128875732,
+      "learning_rate": 2.022224260852963e-06,
+      "loss": 0.8959,
+      "step": 9236
+    },
+    {
+      "epoch": 0.7120721554116559,
+      "grad_norm": 3.4424352645874023,
+      "learning_rate": 2.021221458004695e-06,
+      "loss": 0.8286,
+      "step": 9237
+    },
+    {
+      "epoch": 0.7121492445266728,
+      "grad_norm": 3.8004579544067383,
+      "learning_rate": 2.020218840867842e-06,
+      "loss": 1.0063,
+      "step": 9238
+    },
+    {
+      "epoch": 0.7122263336416897,
+      "grad_norm": 3.6114730834960938,
+      "learning_rate": 2.019216409504913e-06,
+      "loss": 0.8109,
+      "step": 9239
+    },
+    {
+      "epoch": 0.7123034227567068,
+      "grad_norm": 3.892796516418457,
+      "learning_rate": 2.018214163978402e-06,
+      "loss": 1.0058,
+      "step": 9240
+    },
+    {
+      "epoch": 0.7123805118717237,
+      "grad_norm": 3.875223159790039,
+      "learning_rate": 2.0172121043507943e-06,
+      "loss": 0.9591,
+      "step": 9241
+    },
+    {
+      "epoch": 0.7124576009867407,
+      "grad_norm": 4.0388665199279785,
+      "learning_rate": 2.016210230684564e-06,
+      "loss": 0.9484,
+      "step": 9242
+    },
+    {
+      "epoch": 0.7125346901017576,
+      "grad_norm": 3.8614680767059326,
+      "learning_rate": 2.015208543042172e-06,
+      "loss": 1.0,
+      "step": 9243
+    },
+    {
+      "epoch": 0.7126117792167745,
+      "grad_norm": 3.793128252029419,
+      "learning_rate": 2.0142070414860704e-06,
+      "loss": 0.9818,
+      "step": 9244
+    },
+    {
+      "epoch": 0.7126888683317916,
+      "grad_norm": 3.489015817642212,
+      "learning_rate": 2.0132057260786943e-06,
+      "loss": 0.9183,
+      "step": 9245
+    },
+    {
+      "epoch": 0.7127659574468085,
+      "grad_norm": 3.5036089420318604,
+      "learning_rate": 2.012204596882472e-06,
+      "loss": 0.9126,
+      "step": 9246
+    },
+    {
+      "epoch": 0.7128430465618255,
+      "grad_norm": 3.7784531116485596,
+      "learning_rate": 2.011203653959819e-06,
+      "loss": 0.902,
+      "step": 9247
+    },
+    {
+      "epoch": 0.7129201356768424,
+      "grad_norm": 3.950711488723755,
+      "learning_rate": 2.0102028973731393e-06,
+      "loss": 0.9668,
+      "step": 9248
+    },
+    {
+      "epoch": 0.7129972247918593,
+      "grad_norm": 3.538501501083374,
+      "learning_rate": 2.0092023271848254e-06,
+      "loss": 0.9317,
+      "step": 9249
+    },
+    {
+      "epoch": 0.7130743139068764,
+      "grad_norm": 3.5862019062042236,
+      "learning_rate": 2.008201943457255e-06,
+      "loss": 0.8727,
+      "step": 9250
+    },
+    {
+      "epoch": 0.7131514030218933,
+      "grad_norm": 3.7320566177368164,
+      "learning_rate": 2.007201746252799e-06,
+      "loss": 1.0035,
+      "step": 9251
+    },
+    {
+      "epoch": 0.7132284921369103,
+      "grad_norm": 3.5628163814544678,
+      "learning_rate": 2.0062017356338136e-06,
+      "loss": 0.8501,
+      "step": 9252
+    },
+    {
+      "epoch": 0.7133055812519272,
+      "grad_norm": 3.4357900619506836,
+      "learning_rate": 2.0052019116626446e-06,
+      "loss": 0.867,
+      "step": 9253
+    },
+    {
+      "epoch": 0.7133826703669441,
+      "grad_norm": 3.7662131786346436,
+      "learning_rate": 2.0042022744016264e-06,
+      "loss": 1.0013,
+      "step": 9254
+    },
+    {
+      "epoch": 0.7134597594819612,
+      "grad_norm": 3.847144842147827,
+      "learning_rate": 2.0032028239130824e-06,
+      "loss": 1.0047,
+      "step": 9255
+    },
+    {
+      "epoch": 0.7135368485969781,
+      "grad_norm": 3.619068145751953,
+      "learning_rate": 2.00220356025932e-06,
+      "loss": 0.8234,
+      "step": 9256
+    },
+    {
+      "epoch": 0.7136139377119951,
+      "grad_norm": 3.6342787742614746,
+      "learning_rate": 2.001204483502639e-06,
+      "loss": 0.8697,
+      "step": 9257
+    },
+    {
+      "epoch": 0.713691026827012,
+      "grad_norm": 3.920424222946167,
+      "learning_rate": 2.000205593705328e-06,
+      "loss": 1.0292,
+      "step": 9258
+    },
+    {
+      "epoch": 0.7137681159420289,
+      "grad_norm": 3.7946994304656982,
+      "learning_rate": 1.9992068909296607e-06,
+      "loss": 0.8509,
+      "step": 9259
+    },
+    {
+      "epoch": 0.713845205057046,
+      "grad_norm": 4.217837810516357,
+      "learning_rate": 1.998208375237905e-06,
+      "loss": 0.9054,
+      "step": 9260
+    },
+    {
+      "epoch": 0.7139222941720629,
+      "grad_norm": 4.02250862121582,
+      "learning_rate": 1.9972100466923083e-06,
+      "loss": 0.8969,
+      "step": 9261
+    },
+    {
+      "epoch": 0.7139993832870799,
+      "grad_norm": 3.8427062034606934,
+      "learning_rate": 1.996211905355114e-06,
+      "loss": 0.8616,
+      "step": 9262
+    },
+    {
+      "epoch": 0.7140764724020968,
+      "grad_norm": 3.860753059387207,
+      "learning_rate": 1.9952139512885497e-06,
+      "loss": 1.0435,
+      "step": 9263
+    },
+    {
+      "epoch": 0.7141535615171137,
+      "grad_norm": 3.5810599327087402,
+      "learning_rate": 1.9942161845548334e-06,
+      "loss": 0.8754,
+      "step": 9264
+    },
+    {
+      "epoch": 0.7142306506321308,
+      "grad_norm": 3.545610189437866,
+      "learning_rate": 1.993218605216171e-06,
+      "loss": 0.8673,
+      "step": 9265
+    },
+    {
+      "epoch": 0.7143077397471477,
+      "grad_norm": 3.7623043060302734,
+      "learning_rate": 1.9922212133347575e-06,
+      "loss": 1.0008,
+      "step": 9266
+    },
+    {
+      "epoch": 0.7143848288621647,
+      "grad_norm": 3.5907156467437744,
+      "learning_rate": 1.991224008972772e-06,
+      "loss": 0.9327,
+      "step": 9267
+    },
+    {
+      "epoch": 0.7144619179771816,
+      "grad_norm": 3.638535261154175,
+      "learning_rate": 1.9902269921923867e-06,
+      "loss": 0.924,
+      "step": 9268
+    },
+    {
+      "epoch": 0.7145390070921985,
+      "grad_norm": 3.881423234939575,
+      "learning_rate": 1.9892301630557604e-06,
+      "loss": 0.9654,
+      "step": 9269
+    },
+    {
+      "epoch": 0.7146160962072156,
+      "grad_norm": 3.544396162033081,
+      "learning_rate": 1.9882335216250402e-06,
+      "loss": 0.8691,
+      "step": 9270
+    },
+    {
+      "epoch": 0.7146931853222325,
+      "grad_norm": 3.863416910171509,
+      "learning_rate": 1.987237067962363e-06,
+      "loss": 0.8315,
+      "step": 9271
+    },
+    {
+      "epoch": 0.7147702744372495,
+      "grad_norm": 3.7197678089141846,
+      "learning_rate": 1.9862408021298503e-06,
+      "loss": 0.9305,
+      "step": 9272
+    },
+    {
+      "epoch": 0.7148473635522664,
+      "grad_norm": 3.4291746616363525,
+      "learning_rate": 1.9852447241896122e-06,
+      "loss": 0.8485,
+      "step": 9273
+    },
+    {
+      "epoch": 0.7149244526672833,
+      "grad_norm": 3.7952208518981934,
+      "learning_rate": 1.984248834203754e-06,
+      "loss": 0.8727,
+      "step": 9274
+    },
+    {
+      "epoch": 0.7150015417823004,
+      "grad_norm": 3.9852194786071777,
+      "learning_rate": 1.9832531322343617e-06,
+      "loss": 0.9672,
+      "step": 9275
+    },
+    {
+      "epoch": 0.7150786308973173,
+      "grad_norm": 3.7083189487457275,
+      "learning_rate": 1.982257618343515e-06,
+      "loss": 0.8642,
+      "step": 9276
+    },
+    {
+      "epoch": 0.7151557200123343,
+      "grad_norm": 3.6518466472625732,
+      "learning_rate": 1.981262292593274e-06,
+      "loss": 0.8827,
+      "step": 9277
+    },
+    {
+      "epoch": 0.7152328091273512,
+      "grad_norm": 3.622223138809204,
+      "learning_rate": 1.9802671550456948e-06,
+      "loss": 0.8975,
+      "step": 9278
+    },
+    {
+      "epoch": 0.7153098982423681,
+      "grad_norm": 3.709808349609375,
+      "learning_rate": 1.979272205762819e-06,
+      "loss": 0.9319,
+      "step": 9279
+    },
+    {
+      "epoch": 0.7153869873573852,
+      "grad_norm": 3.7233026027679443,
+      "learning_rate": 1.978277444806676e-06,
+      "loss": 0.855,
+      "step": 9280
+    },
+    {
+      "epoch": 0.7154640764724021,
+      "grad_norm": 3.8098905086517334,
+      "learning_rate": 1.9772828722392866e-06,
+      "loss": 0.895,
+      "step": 9281
+    },
+    {
+      "epoch": 0.7155411655874191,
+      "grad_norm": 3.943906545639038,
+      "learning_rate": 1.9762884881226535e-06,
+      "loss": 0.8825,
+      "step": 9282
+    },
+    {
+      "epoch": 0.715618254702436,
+      "grad_norm": 3.8374009132385254,
+      "learning_rate": 1.9752942925187725e-06,
+      "loss": 0.7061,
+      "step": 9283
+    },
+    {
+      "epoch": 0.7156953438174529,
+      "grad_norm": 3.700810194015503,
+      "learning_rate": 1.974300285489627e-06,
+      "loss": 0.9142,
+      "step": 9284
+    },
+    {
+      "epoch": 0.71577243293247,
+      "grad_norm": 3.885406494140625,
+      "learning_rate": 1.9733064670971886e-06,
+      "loss": 0.9404,
+      "step": 9285
+    },
+    {
+      "epoch": 0.7158495220474869,
+      "grad_norm": 3.8215103149414062,
+      "learning_rate": 1.972312837403416e-06,
+      "loss": 0.9591,
+      "step": 9286
+    },
+    {
+      "epoch": 0.7159266111625039,
+      "grad_norm": 3.734654426574707,
+      "learning_rate": 1.971319396470259e-06,
+      "loss": 0.9052,
+      "step": 9287
+    },
+    {
+      "epoch": 0.7160037002775208,
+      "grad_norm": 3.6638259887695312,
+      "learning_rate": 1.97032614435965e-06,
+      "loss": 1.0389,
+      "step": 9288
+    },
+    {
+      "epoch": 0.7160807893925377,
+      "grad_norm": 3.528590440750122,
+      "learning_rate": 1.969333081133515e-06,
+      "loss": 0.8955,
+      "step": 9289
+    },
+    {
+      "epoch": 0.7161578785075547,
+      "grad_norm": 3.670738458633423,
+      "learning_rate": 1.9683402068537654e-06,
+      "loss": 1.0466,
+      "step": 9290
+    },
+    {
+      "epoch": 0.7162349676225717,
+      "grad_norm": 3.703134775161743,
+      "learning_rate": 1.9673475215823035e-06,
+      "loss": 0.9594,
+      "step": 9291
+    },
+    {
+      "epoch": 0.7163120567375887,
+      "grad_norm": 4.054429054260254,
+      "learning_rate": 1.966355025381018e-06,
+      "loss": 0.9606,
+      "step": 9292
+    },
+    {
+      "epoch": 0.7163891458526056,
+      "grad_norm": 3.620304584503174,
+      "learning_rate": 1.965362718311784e-06,
+      "loss": 0.8857,
+      "step": 9293
+    },
+    {
+      "epoch": 0.7164662349676225,
+      "grad_norm": 3.4398069381713867,
+      "learning_rate": 1.9643706004364675e-06,
+      "loss": 0.7762,
+      "step": 9294
+    },
+    {
+      "epoch": 0.7165433240826395,
+      "grad_norm": 3.881047487258911,
+      "learning_rate": 1.9633786718169217e-06,
+      "loss": 0.9335,
+      "step": 9295
+    },
+    {
+      "epoch": 0.7166204131976565,
+      "grad_norm": 3.718986749649048,
+      "learning_rate": 1.9623869325149893e-06,
+      "loss": 0.8663,
+      "step": 9296
+    },
+    {
+      "epoch": 0.7166975023126735,
+      "grad_norm": 3.492274522781372,
+      "learning_rate": 1.9613953825925008e-06,
+      "loss": 0.9198,
+      "step": 9297
+    },
+    {
+      "epoch": 0.7167745914276904,
+      "grad_norm": 3.6691811084747314,
+      "learning_rate": 1.960404022111271e-06,
+      "loss": 0.8376,
+      "step": 9298
+    },
+    {
+      "epoch": 0.7168516805427073,
+      "grad_norm": 3.32060170173645,
+      "learning_rate": 1.9594128511331082e-06,
+      "loss": 0.7495,
+      "step": 9299
+    },
+    {
+      "epoch": 0.7169287696577243,
+      "grad_norm": 3.6289753913879395,
+      "learning_rate": 1.958421869719807e-06,
+      "loss": 0.9179,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7170058587727413,
+      "grad_norm": 3.9919261932373047,
+      "learning_rate": 1.957431077933149e-06,
+      "loss": 0.8687,
+      "step": 9301
+    },
+    {
+      "epoch": 0.7170829478877583,
+      "grad_norm": 4.032829761505127,
+      "learning_rate": 1.9564404758349055e-06,
+      "loss": 0.9729,
+      "step": 9302
+    },
+    {
+      "epoch": 0.7171600370027752,
+      "grad_norm": 3.717048406600952,
+      "learning_rate": 1.955450063486837e-06,
+      "loss": 0.9274,
+      "step": 9303
+    },
+    {
+      "epoch": 0.7172371261177922,
+      "grad_norm": 3.426722288131714,
+      "learning_rate": 1.954459840950687e-06,
+      "loss": 0.8553,
+      "step": 9304
+    },
+    {
+      "epoch": 0.7173142152328091,
+      "grad_norm": 3.5215723514556885,
+      "learning_rate": 1.9534698082881926e-06,
+      "loss": 0.9,
+      "step": 9305
+    },
+    {
+      "epoch": 0.717391304347826,
+      "grad_norm": 3.6140124797821045,
+      "learning_rate": 1.9524799655610776e-06,
+      "loss": 0.8674,
+      "step": 9306
+    },
+    {
+      "epoch": 0.7174683934628431,
+      "grad_norm": 3.1654090881347656,
+      "learning_rate": 1.951490312831053e-06,
+      "loss": 0.7923,
+      "step": 9307
+    },
+    {
+      "epoch": 0.71754548257786,
+      "grad_norm": 3.4995226860046387,
+      "learning_rate": 1.9505008501598204e-06,
+      "loss": 0.9468,
+      "step": 9308
+    },
+    {
+      "epoch": 0.717622571692877,
+      "grad_norm": 3.9290666580200195,
+      "learning_rate": 1.949511577609065e-06,
+      "loss": 0.8838,
+      "step": 9309
+    },
+    {
+      "epoch": 0.7176996608078939,
+      "grad_norm": 4.191568851470947,
+      "learning_rate": 1.948522495240463e-06,
+      "loss": 0.8792,
+      "step": 9310
+    },
+    {
+      "epoch": 0.7177767499229109,
+      "grad_norm": 4.151352882385254,
+      "learning_rate": 1.94753360311568e-06,
+      "loss": 0.9023,
+      "step": 9311
+    },
+    {
+      "epoch": 0.7178538390379279,
+      "grad_norm": 3.6860830783843994,
+      "learning_rate": 1.946544901296367e-06,
+      "loss": 0.9006,
+      "step": 9312
+    },
+    {
+      "epoch": 0.7179309281529448,
+      "grad_norm": 3.961071491241455,
+      "learning_rate": 1.945556389844166e-06,
+      "loss": 0.8171,
+      "step": 9313
+    },
+    {
+      "epoch": 0.7180080172679618,
+      "grad_norm": 3.5512242317199707,
+      "learning_rate": 1.9445680688207065e-06,
+      "loss": 1.0231,
+      "step": 9314
+    },
+    {
+      "epoch": 0.7180851063829787,
+      "grad_norm": 4.143185615539551,
+      "learning_rate": 1.943579938287601e-06,
+      "loss": 0.9645,
+      "step": 9315
+    },
+    {
+      "epoch": 0.7181621954979956,
+      "grad_norm": 3.999950885772705,
+      "learning_rate": 1.942591998306457e-06,
+      "loss": 0.9124,
+      "step": 9316
+    },
+    {
+      "epoch": 0.7182392846130127,
+      "grad_norm": 3.6654207706451416,
+      "learning_rate": 1.941604248938867e-06,
+      "loss": 0.8595,
+      "step": 9317
+    },
+    {
+      "epoch": 0.7183163737280296,
+      "grad_norm": 4.063483238220215,
+      "learning_rate": 1.9406166902464128e-06,
+      "loss": 0.9909,
+      "step": 9318
+    },
+    {
+      "epoch": 0.7183934628430466,
+      "grad_norm": 3.7724883556365967,
+      "learning_rate": 1.9396293222906626e-06,
+      "loss": 0.9458,
+      "step": 9319
+    },
+    {
+      "epoch": 0.7184705519580635,
+      "grad_norm": 3.940067768096924,
+      "learning_rate": 1.9386421451331737e-06,
+      "loss": 0.9825,
+      "step": 9320
+    },
+    {
+      "epoch": 0.7185476410730804,
+      "grad_norm": 3.5647549629211426,
+      "learning_rate": 1.9376551588354924e-06,
+      "loss": 0.9541,
+      "step": 9321
+    },
+    {
+      "epoch": 0.7186247301880975,
+      "grad_norm": 3.541923761367798,
+      "learning_rate": 1.936668363459152e-06,
+      "loss": 0.9573,
+      "step": 9322
+    },
+    {
+      "epoch": 0.7187018193031144,
+      "grad_norm": 3.8034353256225586,
+      "learning_rate": 1.9356817590656734e-06,
+      "loss": 0.7969,
+      "step": 9323
+    },
+    {
+      "epoch": 0.7187789084181314,
+      "grad_norm": 3.5387001037597656,
+      "learning_rate": 1.934695345716568e-06,
+      "loss": 0.8029,
+      "step": 9324
+    },
+    {
+      "epoch": 0.7188559975331483,
+      "grad_norm": 3.763958215713501,
+      "learning_rate": 1.933709123473331e-06,
+      "loss": 0.8748,
+      "step": 9325
+    },
+    {
+      "epoch": 0.7189330866481652,
+      "grad_norm": 3.617594003677368,
+      "learning_rate": 1.9327230923974487e-06,
+      "loss": 0.8477,
+      "step": 9326
+    },
+    {
+      "epoch": 0.7190101757631823,
+      "grad_norm": 3.5570931434631348,
+      "learning_rate": 1.931737252550396e-06,
+      "loss": 0.8856,
+      "step": 9327
+    },
+    {
+      "epoch": 0.7190872648781992,
+      "grad_norm": 3.4663596153259277,
+      "learning_rate": 1.9307516039936354e-06,
+      "loss": 0.874,
+      "step": 9328
+    },
+    {
+      "epoch": 0.7191643539932162,
+      "grad_norm": 3.960660696029663,
+      "learning_rate": 1.9297661467886177e-06,
+      "loss": 0.9692,
+      "step": 9329
+    },
+    {
+      "epoch": 0.7192414431082331,
+      "grad_norm": 3.4713823795318604,
+      "learning_rate": 1.928780880996777e-06,
+      "loss": 0.8727,
+      "step": 9330
+    },
+    {
+      "epoch": 0.71931853222325,
+      "grad_norm": 3.68634295463562,
+      "learning_rate": 1.9277958066795426e-06,
+      "loss": 0.8905,
+      "step": 9331
+    },
+    {
+      "epoch": 0.7193956213382671,
+      "grad_norm": 3.8657889366149902,
+      "learning_rate": 1.9268109238983287e-06,
+      "loss": 0.9138,
+      "step": 9332
+    },
+    {
+      "epoch": 0.719472710453284,
+      "grad_norm": 3.750303268432617,
+      "learning_rate": 1.925826232714537e-06,
+      "loss": 0.9105,
+      "step": 9333
+    },
+    {
+      "epoch": 0.719549799568301,
+      "grad_norm": 3.9123423099517822,
+      "learning_rate": 1.924841733189558e-06,
+      "loss": 0.8826,
+      "step": 9334
+    },
+    {
+      "epoch": 0.7196268886833179,
+      "grad_norm": 3.7364614009857178,
+      "learning_rate": 1.923857425384772e-06,
+      "loss": 0.9367,
+      "step": 9335
+    },
+    {
+      "epoch": 0.7197039777983348,
+      "grad_norm": 4.046642303466797,
+      "learning_rate": 1.922873309361542e-06,
+      "loss": 1.0428,
+      "step": 9336
+    },
+    {
+      "epoch": 0.7197810669133519,
+      "grad_norm": 3.7967915534973145,
+      "learning_rate": 1.921889385181225e-06,
+      "loss": 0.9271,
+      "step": 9337
+    },
+    {
+      "epoch": 0.7198581560283688,
+      "grad_norm": 4.126679420471191,
+      "learning_rate": 1.9209056529051617e-06,
+      "loss": 0.9439,
+      "step": 9338
+    },
+    {
+      "epoch": 0.7199352451433858,
+      "grad_norm": 3.514425039291382,
+      "learning_rate": 1.9199221125946847e-06,
+      "loss": 0.8919,
+      "step": 9339
+    },
+    {
+      "epoch": 0.7200123342584027,
+      "grad_norm": 3.4573843479156494,
+      "learning_rate": 1.9189387643111135e-06,
+      "loss": 0.8665,
+      "step": 9340
+    },
+    {
+      "epoch": 0.7200894233734196,
+      "grad_norm": 4.3528242111206055,
+      "learning_rate": 1.9179556081157513e-06,
+      "loss": 0.9908,
+      "step": 9341
+    },
+    {
+      "epoch": 0.7201665124884367,
+      "grad_norm": 3.473971366882324,
+      "learning_rate": 1.9169726440698945e-06,
+      "loss": 0.8575,
+      "step": 9342
+    },
+    {
+      "epoch": 0.7202436016034536,
+      "grad_norm": 3.958357334136963,
+      "learning_rate": 1.9159898722348264e-06,
+      "loss": 0.9238,
+      "step": 9343
+    },
+    {
+      "epoch": 0.7203206907184706,
+      "grad_norm": 3.5286219120025635,
+      "learning_rate": 1.9150072926718166e-06,
+      "loss": 0.9294,
+      "step": 9344
+    },
+    {
+      "epoch": 0.7203977798334875,
+      "grad_norm": 3.8079938888549805,
+      "learning_rate": 1.914024905442127e-06,
+      "loss": 0.9186,
+      "step": 9345
+    },
+    {
+      "epoch": 0.7204748689485044,
+      "grad_norm": 3.7631888389587402,
+      "learning_rate": 1.9130427106069993e-06,
+      "loss": 0.9976,
+      "step": 9346
+    },
+    {
+      "epoch": 0.7205519580635215,
+      "grad_norm": 3.6607329845428467,
+      "learning_rate": 1.912060708227671e-06,
+      "loss": 0.9171,
+      "step": 9347
+    },
+    {
+      "epoch": 0.7206290471785384,
+      "grad_norm": 3.550433874130249,
+      "learning_rate": 1.911078898365365e-06,
+      "loss": 0.8487,
+      "step": 9348
+    },
+    {
+      "epoch": 0.7207061362935554,
+      "grad_norm": 3.631561040878296,
+      "learning_rate": 1.9100972810812918e-06,
+      "loss": 0.9102,
+      "step": 9349
+    },
+    {
+      "epoch": 0.7207832254085723,
+      "grad_norm": 4.037540435791016,
+      "learning_rate": 1.90911585643665e-06,
+      "loss": 0.9158,
+      "step": 9350
+    },
+    {
+      "epoch": 0.7208603145235892,
+      "grad_norm": 4.127115249633789,
+      "learning_rate": 1.908134624492628e-06,
+      "loss": 0.9296,
+      "step": 9351
+    },
+    {
+      "epoch": 0.7209374036386063,
+      "grad_norm": 4.0273613929748535,
+      "learning_rate": 1.9071535853103978e-06,
+      "loss": 0.9744,
+      "step": 9352
+    },
+    {
+      "epoch": 0.7210144927536232,
+      "grad_norm": 4.10391092300415,
+      "learning_rate": 1.9061727389511226e-06,
+      "loss": 0.8842,
+      "step": 9353
+    },
+    {
+      "epoch": 0.7210915818686402,
+      "grad_norm": 3.6492302417755127,
+      "learning_rate": 1.9051920854759543e-06,
+      "loss": 0.9429,
+      "step": 9354
+    },
+    {
+      "epoch": 0.7211686709836571,
+      "grad_norm": 3.974447011947632,
+      "learning_rate": 1.9042116249460307e-06,
+      "loss": 0.9477,
+      "step": 9355
+    },
+    {
+      "epoch": 0.721245760098674,
+      "grad_norm": 3.8912482261657715,
+      "learning_rate": 1.903231357422481e-06,
+      "loss": 0.8826,
+      "step": 9356
+    },
+    {
+      "epoch": 0.721322849213691,
+      "grad_norm": 3.689361810684204,
+      "learning_rate": 1.9022512829664153e-06,
+      "loss": 0.9352,
+      "step": 9357
+    },
+    {
+      "epoch": 0.721399938328708,
+      "grad_norm": 3.682281732559204,
+      "learning_rate": 1.9012714016389388e-06,
+      "loss": 0.8895,
+      "step": 9358
+    },
+    {
+      "epoch": 0.721477027443725,
+      "grad_norm": 3.9238123893737793,
+      "learning_rate": 1.9002917135011413e-06,
+      "loss": 0.9899,
+      "step": 9359
+    },
+    {
+      "epoch": 0.7215541165587419,
+      "grad_norm": 3.783236503601074,
+      "learning_rate": 1.899312218614102e-06,
+      "loss": 0.9622,
+      "step": 9360
+    },
+    {
+      "epoch": 0.7216312056737588,
+      "grad_norm": 3.7680461406707764,
+      "learning_rate": 1.898332917038887e-06,
+      "loss": 0.9219,
+      "step": 9361
+    },
+    {
+      "epoch": 0.7217082947887759,
+      "grad_norm": 3.3629398345947266,
+      "learning_rate": 1.8973538088365507e-06,
+      "loss": 0.8667,
+      "step": 9362
+    },
+    {
+      "epoch": 0.7217853839037928,
+      "grad_norm": 3.6810696125030518,
+      "learning_rate": 1.8963748940681349e-06,
+      "loss": 1.0191,
+      "step": 9363
+    },
+    {
+      "epoch": 0.7218624730188098,
+      "grad_norm": 4.012876987457275,
+      "learning_rate": 1.8953961727946706e-06,
+      "loss": 0.8726,
+      "step": 9364
+    },
+    {
+      "epoch": 0.7219395621338267,
+      "grad_norm": 3.692685842514038,
+      "learning_rate": 1.8944176450771761e-06,
+      "loss": 0.8478,
+      "step": 9365
+    },
+    {
+      "epoch": 0.7220166512488436,
+      "grad_norm": 3.4944231510162354,
+      "learning_rate": 1.893439310976659e-06,
+      "loss": 0.8727,
+      "step": 9366
+    },
+    {
+      "epoch": 0.7220937403638606,
+      "grad_norm": 3.8247053623199463,
+      "learning_rate": 1.8924611705541095e-06,
+      "loss": 0.9374,
+      "step": 9367
+    },
+    {
+      "epoch": 0.7221708294788776,
+      "grad_norm": 4.079497814178467,
+      "learning_rate": 1.8914832238705117e-06,
+      "loss": 0.9631,
+      "step": 9368
+    },
+    {
+      "epoch": 0.7222479185938946,
+      "grad_norm": 4.07694149017334,
+      "learning_rate": 1.8905054709868354e-06,
+      "loss": 1.0642,
+      "step": 9369
+    },
+    {
+      "epoch": 0.7223250077089115,
+      "grad_norm": 3.5892646312713623,
+      "learning_rate": 1.8895279119640387e-06,
+      "loss": 0.9717,
+      "step": 9370
+    },
+    {
+      "epoch": 0.7224020968239284,
+      "grad_norm": 3.906541347503662,
+      "learning_rate": 1.8885505468630673e-06,
+      "loss": 0.9333,
+      "step": 9371
+    },
+    {
+      "epoch": 0.7224791859389454,
+      "grad_norm": 4.1830830574035645,
+      "learning_rate": 1.887573375744856e-06,
+      "loss": 0.993,
+      "step": 9372
+    },
+    {
+      "epoch": 0.7225562750539624,
+      "grad_norm": 3.726923942565918,
+      "learning_rate": 1.8865963986703234e-06,
+      "loss": 0.9537,
+      "step": 9373
+    },
+    {
+      "epoch": 0.7226333641689794,
+      "grad_norm": 3.74369215965271,
+      "learning_rate": 1.885619615700381e-06,
+      "loss": 0.9731,
+      "step": 9374
+    },
+    {
+      "epoch": 0.7227104532839963,
+      "grad_norm": 4.148539066314697,
+      "learning_rate": 1.8846430268959253e-06,
+      "loss": 1.0761,
+      "step": 9375
+    },
+    {
+      "epoch": 0.7227875423990132,
+      "grad_norm": 3.4325528144836426,
+      "learning_rate": 1.883666632317842e-06,
+      "loss": 0.846,
+      "step": 9376
+    },
+    {
+      "epoch": 0.7228646315140302,
+      "grad_norm": 3.434077024459839,
+      "learning_rate": 1.8826904320270068e-06,
+      "loss": 0.7253,
+      "step": 9377
+    },
+    {
+      "epoch": 0.7229417206290472,
+      "grad_norm": 3.7879035472869873,
+      "learning_rate": 1.8817144260842757e-06,
+      "loss": 0.8692,
+      "step": 9378
+    },
+    {
+      "epoch": 0.7230188097440642,
+      "grad_norm": 3.6962223052978516,
+      "learning_rate": 1.8807386145505002e-06,
+      "loss": 0.8621,
+      "step": 9379
+    },
+    {
+      "epoch": 0.7230958988590811,
+      "grad_norm": 3.504754066467285,
+      "learning_rate": 1.8797629974865172e-06,
+      "loss": 0.8153,
+      "step": 9380
+    },
+    {
+      "epoch": 0.723172987974098,
+      "grad_norm": 3.4584434032440186,
+      "learning_rate": 1.8787875749531509e-06,
+      "loss": 0.9573,
+      "step": 9381
+    },
+    {
+      "epoch": 0.723250077089115,
+      "grad_norm": 3.89050555229187,
+      "learning_rate": 1.8778123470112141e-06,
+      "loss": 0.8954,
+      "step": 9382
+    },
+    {
+      "epoch": 0.723327166204132,
+      "grad_norm": 3.5211403369903564,
+      "learning_rate": 1.8768373137215096e-06,
+      "loss": 0.8079,
+      "step": 9383
+    },
+    {
+      "epoch": 0.723404255319149,
+      "grad_norm": 3.81539249420166,
+      "learning_rate": 1.8758624751448213e-06,
+      "loss": 0.9524,
+      "step": 9384
+    },
+    {
+      "epoch": 0.7234813444341659,
+      "grad_norm": 3.4943387508392334,
+      "learning_rate": 1.8748878313419271e-06,
+      "loss": 0.8294,
+      "step": 9385
+    },
+    {
+      "epoch": 0.7235584335491828,
+      "grad_norm": 3.9146976470947266,
+      "learning_rate": 1.873913382373591e-06,
+      "loss": 0.9008,
+      "step": 9386
+    },
+    {
+      "epoch": 0.7236355226641998,
+      "grad_norm": 3.7792716026306152,
+      "learning_rate": 1.872939128300566e-06,
+      "loss": 1.0432,
+      "step": 9387
+    },
+    {
+      "epoch": 0.7237126117792168,
+      "grad_norm": 4.022826671600342,
+      "learning_rate": 1.8719650691835917e-06,
+      "loss": 0.9682,
+      "step": 9388
+    },
+    {
+      "epoch": 0.7237897008942338,
+      "grad_norm": 3.6135926246643066,
+      "learning_rate": 1.8709912050833933e-06,
+      "loss": 0.8809,
+      "step": 9389
+    },
+    {
+      "epoch": 0.7238667900092507,
+      "grad_norm": 3.3770127296447754,
+      "learning_rate": 1.8700175360606882e-06,
+      "loss": 0.8576,
+      "step": 9390
+    },
+    {
+      "epoch": 0.7239438791242676,
+      "grad_norm": 3.8401501178741455,
+      "learning_rate": 1.869044062176179e-06,
+      "loss": 0.975,
+      "step": 9391
+    },
+    {
+      "epoch": 0.7240209682392846,
+      "grad_norm": 3.574070453643799,
+      "learning_rate": 1.8680707834905565e-06,
+      "loss": 0.8732,
+      "step": 9392
+    },
+    {
+      "epoch": 0.7240980573543015,
+      "grad_norm": 3.733696699142456,
+      "learning_rate": 1.8670977000645018e-06,
+      "loss": 1.0079,
+      "step": 9393
+    },
+    {
+      "epoch": 0.7241751464693186,
+      "grad_norm": 3.893634557723999,
+      "learning_rate": 1.8661248119586784e-06,
+      "loss": 0.9188,
+      "step": 9394
+    },
+    {
+      "epoch": 0.7242522355843355,
+      "grad_norm": 3.5450663566589355,
+      "learning_rate": 1.865152119233742e-06,
+      "loss": 0.9227,
+      "step": 9395
+    },
+    {
+      "epoch": 0.7243293246993524,
+      "grad_norm": 4.02310848236084,
+      "learning_rate": 1.864179621950335e-06,
+      "loss": 0.8968,
+      "step": 9396
+    },
+    {
+      "epoch": 0.7244064138143694,
+      "grad_norm": 3.913698434829712,
+      "learning_rate": 1.8632073201690882e-06,
+      "loss": 0.8523,
+      "step": 9397
+    },
+    {
+      "epoch": 0.7244835029293863,
+      "grad_norm": 3.8382716178894043,
+      "learning_rate": 1.8622352139506184e-06,
+      "loss": 0.8597,
+      "step": 9398
+    },
+    {
+      "epoch": 0.7245605920444034,
+      "grad_norm": 4.016849040985107,
+      "learning_rate": 1.8612633033555345e-06,
+      "loss": 0.7716,
+      "step": 9399
+    },
+    {
+      "epoch": 0.7246376811594203,
+      "grad_norm": 3.819669485092163,
+      "learning_rate": 1.8602915884444257e-06,
+      "loss": 0.9032,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7247147702744372,
+      "grad_norm": 4.067957878112793,
+      "learning_rate": 1.859320069277875e-06,
+      "loss": 0.9601,
+      "step": 9401
+    },
+    {
+      "epoch": 0.7247918593894542,
+      "grad_norm": 3.471204996109009,
+      "learning_rate": 1.8583487459164528e-06,
+      "loss": 0.9013,
+      "step": 9402
+    },
+    {
+      "epoch": 0.7248689485044711,
+      "grad_norm": 3.366598606109619,
+      "learning_rate": 1.8573776184207148e-06,
+      "loss": 0.9391,
+      "step": 9403
+    },
+    {
+      "epoch": 0.7249460376194882,
+      "grad_norm": 3.657421827316284,
+      "learning_rate": 1.8564066868512082e-06,
+      "loss": 0.8154,
+      "step": 9404
+    },
+    {
+      "epoch": 0.7250231267345051,
+      "grad_norm": 3.919306755065918,
+      "learning_rate": 1.8554359512684617e-06,
+      "loss": 0.9312,
+      "step": 9405
+    },
+    {
+      "epoch": 0.725100215849522,
+      "grad_norm": 3.4179515838623047,
+      "learning_rate": 1.8544654117329958e-06,
+      "loss": 0.8741,
+      "step": 9406
+    },
+    {
+      "epoch": 0.725177304964539,
+      "grad_norm": 3.9625890254974365,
+      "learning_rate": 1.8534950683053215e-06,
+      "loss": 0.8815,
+      "step": 9407
+    },
+    {
+      "epoch": 0.7252543940795559,
+      "grad_norm": 3.6159121990203857,
+      "learning_rate": 1.8525249210459345e-06,
+      "loss": 0.8737,
+      "step": 9408
+    },
+    {
+      "epoch": 0.725331483194573,
+      "grad_norm": 3.8844573497772217,
+      "learning_rate": 1.8515549700153185e-06,
+      "loss": 0.8763,
+      "step": 9409
+    },
+    {
+      "epoch": 0.7254085723095899,
+      "grad_norm": 3.9960150718688965,
+      "learning_rate": 1.8505852152739423e-06,
+      "loss": 1.0845,
+      "step": 9410
+    },
+    {
+      "epoch": 0.7254856614246068,
+      "grad_norm": 3.7577638626098633,
+      "learning_rate": 1.849615656882267e-06,
+      "loss": 1.0138,
+      "step": 9411
+    },
+    {
+      "epoch": 0.7255627505396238,
+      "grad_norm": 3.612480401992798,
+      "learning_rate": 1.8486462949007388e-06,
+      "loss": 0.9291,
+      "step": 9412
+    },
+    {
+      "epoch": 0.7256398396546407,
+      "grad_norm": 3.4609127044677734,
+      "learning_rate": 1.8476771293897932e-06,
+      "loss": 0.9554,
+      "step": 9413
+    },
+    {
+      "epoch": 0.7257169287696578,
+      "grad_norm": 3.4932973384857178,
+      "learning_rate": 1.846708160409854e-06,
+      "loss": 0.96,
+      "step": 9414
+    },
+    {
+      "epoch": 0.7257940178846747,
+      "grad_norm": 3.711120367050171,
+      "learning_rate": 1.8457393880213282e-06,
+      "loss": 0.9477,
+      "step": 9415
+    },
+    {
+      "epoch": 0.7258711069996916,
+      "grad_norm": 3.978654384613037,
+      "learning_rate": 1.8447708122846148e-06,
+      "loss": 0.9769,
+      "step": 9416
+    },
+    {
+      "epoch": 0.7259481961147086,
+      "grad_norm": 3.9434118270874023,
+      "learning_rate": 1.8438024332601002e-06,
+      "loss": 0.9147,
+      "step": 9417
+    },
+    {
+      "epoch": 0.7260252852297255,
+      "grad_norm": 3.865705966949463,
+      "learning_rate": 1.8428342510081571e-06,
+      "loss": 0.9856,
+      "step": 9418
+    },
+    {
+      "epoch": 0.7261023743447426,
+      "grad_norm": 3.8399107456207275,
+      "learning_rate": 1.8418662655891472e-06,
+      "loss": 0.9694,
+      "step": 9419
+    },
+    {
+      "epoch": 0.7261794634597595,
+      "grad_norm": 3.621166467666626,
+      "learning_rate": 1.8408984770634209e-06,
+      "loss": 0.9042,
+      "step": 9420
+    },
+    {
+      "epoch": 0.7262565525747764,
+      "grad_norm": 3.5802974700927734,
+      "learning_rate": 1.8399308854913118e-06,
+      "loss": 0.8997,
+      "step": 9421
+    },
+    {
+      "epoch": 0.7263336416897934,
+      "grad_norm": 3.3973002433776855,
+      "learning_rate": 1.8389634909331449e-06,
+      "loss": 0.8711,
+      "step": 9422
+    },
+    {
+      "epoch": 0.7264107308048103,
+      "grad_norm": 3.4591875076293945,
+      "learning_rate": 1.8379962934492335e-06,
+      "loss": 0.8588,
+      "step": 9423
+    },
+    {
+      "epoch": 0.7264878199198274,
+      "grad_norm": 3.682715654373169,
+      "learning_rate": 1.8370292930998768e-06,
+      "loss": 0.859,
+      "step": 9424
+    },
+    {
+      "epoch": 0.7265649090348443,
+      "grad_norm": 3.5783348083496094,
+      "learning_rate": 1.8360624899453638e-06,
+      "loss": 0.7877,
+      "step": 9425
+    },
+    {
+      "epoch": 0.7266419981498612,
+      "grad_norm": 3.5001380443573,
+      "learning_rate": 1.8350958840459665e-06,
+      "loss": 0.8595,
+      "step": 9426
+    },
+    {
+      "epoch": 0.7267190872648782,
+      "grad_norm": 3.755657434463501,
+      "learning_rate": 1.8341294754619487e-06,
+      "loss": 0.9371,
+      "step": 9427
+    },
+    {
+      "epoch": 0.7267961763798951,
+      "grad_norm": 4.00222635269165,
+      "learning_rate": 1.8331632642535623e-06,
+      "loss": 0.9562,
+      "step": 9428
+    },
+    {
+      "epoch": 0.7268732654949122,
+      "grad_norm": 3.519622802734375,
+      "learning_rate": 1.8321972504810448e-06,
+      "loss": 0.9217,
+      "step": 9429
+    },
+    {
+      "epoch": 0.7269503546099291,
+      "grad_norm": 3.830078363418579,
+      "learning_rate": 1.8312314342046222e-06,
+      "loss": 0.9358,
+      "step": 9430
+    },
+    {
+      "epoch": 0.727027443724946,
+      "grad_norm": 3.737502336502075,
+      "learning_rate": 1.8302658154845099e-06,
+      "loss": 0.9166,
+      "step": 9431
+    },
+    {
+      "epoch": 0.727104532839963,
+      "grad_norm": 3.694373369216919,
+      "learning_rate": 1.8293003943809062e-06,
+      "loss": 0.8676,
+      "step": 9432
+    },
+    {
+      "epoch": 0.7271816219549799,
+      "grad_norm": 3.707840919494629,
+      "learning_rate": 1.828335170954001e-06,
+      "loss": 0.8953,
+      "step": 9433
+    },
+    {
+      "epoch": 0.727258711069997,
+      "grad_norm": 3.8381030559539795,
+      "learning_rate": 1.8273701452639713e-06,
+      "loss": 0.9618,
+      "step": 9434
+    },
+    {
+      "epoch": 0.7273358001850139,
+      "grad_norm": 3.7728703022003174,
+      "learning_rate": 1.8264053173709817e-06,
+      "loss": 0.8564,
+      "step": 9435
+    },
+    {
+      "epoch": 0.7274128893000308,
+      "grad_norm": 3.9917378425598145,
+      "learning_rate": 1.825440687335186e-06,
+      "loss": 1.0076,
+      "step": 9436
+    },
+    {
+      "epoch": 0.7274899784150478,
+      "grad_norm": 3.8551135063171387,
+      "learning_rate": 1.82447625521672e-06,
+      "loss": 0.9929,
+      "step": 9437
+    },
+    {
+      "epoch": 0.7275670675300647,
+      "grad_norm": 3.5121772289276123,
+      "learning_rate": 1.8235120210757134e-06,
+      "loss": 0.8955,
+      "step": 9438
+    },
+    {
+      "epoch": 0.7276441566450818,
+      "grad_norm": 4.1116509437561035,
+      "learning_rate": 1.8225479849722804e-06,
+      "loss": 0.9207,
+      "step": 9439
+    },
+    {
+      "epoch": 0.7277212457600987,
+      "grad_norm": 3.6008291244506836,
+      "learning_rate": 1.8215841469665247e-06,
+      "loss": 0.9125,
+      "step": 9440
+    },
+    {
+      "epoch": 0.7277983348751156,
+      "grad_norm": 3.791553020477295,
+      "learning_rate": 1.8206205071185373e-06,
+      "loss": 0.8539,
+      "step": 9441
+    },
+    {
+      "epoch": 0.7278754239901326,
+      "grad_norm": 3.53488826751709,
+      "learning_rate": 1.8196570654883932e-06,
+      "loss": 0.8545,
+      "step": 9442
+    },
+    {
+      "epoch": 0.7279525131051495,
+      "grad_norm": 3.721651554107666,
+      "learning_rate": 1.8186938221361594e-06,
+      "loss": 1.0581,
+      "step": 9443
+    },
+    {
+      "epoch": 0.7280296022201665,
+      "grad_norm": 3.7198684215545654,
+      "learning_rate": 1.8177307771218894e-06,
+      "loss": 0.8781,
+      "step": 9444
+    },
+    {
+      "epoch": 0.7281066913351835,
+      "grad_norm": 3.6171069145202637,
+      "learning_rate": 1.8167679305056247e-06,
+      "loss": 0.9443,
+      "step": 9445
+    },
+    {
+      "epoch": 0.7281837804502004,
+      "grad_norm": 3.718351125717163,
+      "learning_rate": 1.8158052823473927e-06,
+      "loss": 0.9222,
+      "step": 9446
+    },
+    {
+      "epoch": 0.7282608695652174,
+      "grad_norm": 3.887049913406372,
+      "learning_rate": 1.8148428327072114e-06,
+      "loss": 0.9062,
+      "step": 9447
+    },
+    {
+      "epoch": 0.7283379586802343,
+      "grad_norm": 3.688441514968872,
+      "learning_rate": 1.8138805816450815e-06,
+      "loss": 0.9262,
+      "step": 9448
+    },
+    {
+      "epoch": 0.7284150477952513,
+      "grad_norm": 3.61232852935791,
+      "learning_rate": 1.812918529220996e-06,
+      "loss": 0.8991,
+      "step": 9449
+    },
+    {
+      "epoch": 0.7284921369102683,
+      "grad_norm": 3.5513792037963867,
+      "learning_rate": 1.8119566754949324e-06,
+      "loss": 0.8965,
+      "step": 9450
+    },
+    {
+      "epoch": 0.7285692260252852,
+      "grad_norm": 3.7241334915161133,
+      "learning_rate": 1.8109950205268624e-06,
+      "loss": 0.8641,
+      "step": 9451
+    },
+    {
+      "epoch": 0.7286463151403022,
+      "grad_norm": 3.3493247032165527,
+      "learning_rate": 1.8100335643767347e-06,
+      "loss": 0.9665,
+      "step": 9452
+    },
+    {
+      "epoch": 0.7287234042553191,
+      "grad_norm": 3.7719924449920654,
+      "learning_rate": 1.809072307104493e-06,
+      "loss": 0.8762,
+      "step": 9453
+    },
+    {
+      "epoch": 0.7288004933703361,
+      "grad_norm": 3.6483349800109863,
+      "learning_rate": 1.8081112487700665e-06,
+      "loss": 0.8928,
+      "step": 9454
+    },
+    {
+      "epoch": 0.7288775824853531,
+      "grad_norm": 3.8290278911590576,
+      "learning_rate": 1.8071503894333725e-06,
+      "loss": 0.9746,
+      "step": 9455
+    },
+    {
+      "epoch": 0.72895467160037,
+      "grad_norm": 3.715212345123291,
+      "learning_rate": 1.8061897291543157e-06,
+      "loss": 0.8306,
+      "step": 9456
+    },
+    {
+      "epoch": 0.729031760715387,
+      "grad_norm": 3.781547784805298,
+      "learning_rate": 1.8052292679927896e-06,
+      "loss": 0.8701,
+      "step": 9457
+    },
+    {
+      "epoch": 0.7291088498304039,
+      "grad_norm": 3.933978796005249,
+      "learning_rate": 1.804269006008671e-06,
+      "loss": 0.9048,
+      "step": 9458
+    },
+    {
+      "epoch": 0.7291859389454209,
+      "grad_norm": 3.5607521533966064,
+      "learning_rate": 1.803308943261829e-06,
+      "loss": 0.861,
+      "step": 9459
+    },
+    {
+      "epoch": 0.7292630280604379,
+      "grad_norm": 3.4771389961242676,
+      "learning_rate": 1.802349079812118e-06,
+      "loss": 0.9354,
+      "step": 9460
+    },
+    {
+      "epoch": 0.7293401171754548,
+      "grad_norm": 3.6316308975219727,
+      "learning_rate": 1.8013894157193807e-06,
+      "loss": 1.0234,
+      "step": 9461
+    },
+    {
+      "epoch": 0.7294172062904718,
+      "grad_norm": 3.486682176589966,
+      "learning_rate": 1.8004299510434493e-06,
+      "loss": 0.8481,
+      "step": 9462
+    },
+    {
+      "epoch": 0.7294942954054887,
+      "grad_norm": 3.479135036468506,
+      "learning_rate": 1.7994706858441375e-06,
+      "loss": 0.9336,
+      "step": 9463
+    },
+    {
+      "epoch": 0.7295713845205057,
+      "grad_norm": 3.551004409790039,
+      "learning_rate": 1.7985116201812524e-06,
+      "loss": 0.937,
+      "step": 9464
+    },
+    {
+      "epoch": 0.7296484736355227,
+      "grad_norm": 3.698634386062622,
+      "learning_rate": 1.7975527541145865e-06,
+      "loss": 0.9793,
+      "step": 9465
+    },
+    {
+      "epoch": 0.7297255627505396,
+      "grad_norm": 3.525210380554199,
+      "learning_rate": 1.7965940877039211e-06,
+      "loss": 1.0082,
+      "step": 9466
+    },
+    {
+      "epoch": 0.7298026518655566,
+      "grad_norm": 3.7876412868499756,
+      "learning_rate": 1.7956356210090236e-06,
+      "loss": 0.8984,
+      "step": 9467
+    },
+    {
+      "epoch": 0.7298797409805735,
+      "grad_norm": 4.059507369995117,
+      "learning_rate": 1.7946773540896506e-06,
+      "loss": 0.9558,
+      "step": 9468
+    },
+    {
+      "epoch": 0.7299568300955905,
+      "grad_norm": 3.954590320587158,
+      "learning_rate": 1.793719287005542e-06,
+      "loss": 1.0237,
+      "step": 9469
+    },
+    {
+      "epoch": 0.7300339192106075,
+      "grad_norm": 4.025545597076416,
+      "learning_rate": 1.7927614198164306e-06,
+      "loss": 1.0038,
+      "step": 9470
+    },
+    {
+      "epoch": 0.7301110083256244,
+      "grad_norm": 3.4767184257507324,
+      "learning_rate": 1.7918037525820336e-06,
+      "loss": 0.832,
+      "step": 9471
+    },
+    {
+      "epoch": 0.7301880974406414,
+      "grad_norm": 3.732330322265625,
+      "learning_rate": 1.7908462853620568e-06,
+      "loss": 0.9538,
+      "step": 9472
+    },
+    {
+      "epoch": 0.7302651865556583,
+      "grad_norm": 3.6224257946014404,
+      "learning_rate": 1.7898890182161954e-06,
+      "loss": 0.9717,
+      "step": 9473
+    },
+    {
+      "epoch": 0.7303422756706753,
+      "grad_norm": 3.483696460723877,
+      "learning_rate": 1.7889319512041264e-06,
+      "loss": 0.8152,
+      "step": 9474
+    },
+    {
+      "epoch": 0.7304193647856922,
+      "grad_norm": 3.9409186840057373,
+      "learning_rate": 1.7879750843855197e-06,
+      "loss": 0.9242,
+      "step": 9475
+    },
+    {
+      "epoch": 0.7304964539007093,
+      "grad_norm": 3.357489585876465,
+      "learning_rate": 1.7870184178200312e-06,
+      "loss": 0.9181,
+      "step": 9476
+    },
+    {
+      "epoch": 0.7305735430157262,
+      "grad_norm": 3.646585702896118,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 0.9368,
+      "step": 9477
+    },
+    {
+      "epoch": 0.7306506321307431,
+      "grad_norm": 3.7453439235687256,
+      "learning_rate": 1.7851056856869681e-06,
+      "loss": 0.9319,
+      "step": 9478
+    },
+    {
+      "epoch": 0.7307277212457601,
+      "grad_norm": 3.725774049758911,
+      "learning_rate": 1.7841496202386437e-06,
+      "loss": 0.8849,
+      "step": 9479
+    },
+    {
+      "epoch": 0.730804810360777,
+      "grad_norm": 3.687467098236084,
+      "learning_rate": 1.7831937552819345e-06,
+      "loss": 0.9166,
+      "step": 9480
+    },
+    {
+      "epoch": 0.7308818994757941,
+      "grad_norm": 3.666307210922241,
+      "learning_rate": 1.7822380908764336e-06,
+      "loss": 0.768,
+      "step": 9481
+    },
+    {
+      "epoch": 0.730958988590811,
+      "grad_norm": 3.638786554336548,
+      "learning_rate": 1.781282627081723e-06,
+      "loss": 0.8445,
+      "step": 9482
+    },
+    {
+      "epoch": 0.7310360777058279,
+      "grad_norm": 3.7186498641967773,
+      "learning_rate": 1.7803273639573704e-06,
+      "loss": 0.9877,
+      "step": 9483
+    },
+    {
+      "epoch": 0.7311131668208449,
+      "grad_norm": 4.096242904663086,
+      "learning_rate": 1.7793723015629333e-06,
+      "loss": 0.9338,
+      "step": 9484
+    },
+    {
+      "epoch": 0.7311902559358618,
+      "grad_norm": 3.3239645957946777,
+      "learning_rate": 1.7784174399579513e-06,
+      "loss": 0.7036,
+      "step": 9485
+    },
+    {
+      "epoch": 0.7312673450508789,
+      "grad_norm": 3.5666675567626953,
+      "learning_rate": 1.7774627792019567e-06,
+      "loss": 0.8775,
+      "step": 9486
+    },
+    {
+      "epoch": 0.7313444341658958,
+      "grad_norm": 3.7199814319610596,
+      "learning_rate": 1.7765083193544679e-06,
+      "loss": 0.974,
+      "step": 9487
+    },
+    {
+      "epoch": 0.7314215232809127,
+      "grad_norm": 3.733372688293457,
+      "learning_rate": 1.775554060474991e-06,
+      "loss": 0.9032,
+      "step": 9488
+    },
+    {
+      "epoch": 0.7314986123959297,
+      "grad_norm": 4.118587493896484,
+      "learning_rate": 1.7746000026230198e-06,
+      "loss": 0.8837,
+      "step": 9489
+    },
+    {
+      "epoch": 0.7315757015109466,
+      "grad_norm": 3.630469560623169,
+      "learning_rate": 1.7736461458580324e-06,
+      "loss": 0.8297,
+      "step": 9490
+    },
+    {
+      "epoch": 0.7316527906259637,
+      "grad_norm": 3.6379334926605225,
+      "learning_rate": 1.7726924902394976e-06,
+      "loss": 0.7553,
+      "step": 9491
+    },
+    {
+      "epoch": 0.7317298797409806,
+      "grad_norm": 3.688769817352295,
+      "learning_rate": 1.7717390358268716e-06,
+      "loss": 0.9635,
+      "step": 9492
+    },
+    {
+      "epoch": 0.7318069688559975,
+      "grad_norm": 3.543370485305786,
+      "learning_rate": 1.7707857826795971e-06,
+      "loss": 0.8589,
+      "step": 9493
+    },
+    {
+      "epoch": 0.7318840579710145,
+      "grad_norm": 3.616456985473633,
+      "learning_rate": 1.7698327308571045e-06,
+      "loss": 0.8856,
+      "step": 9494
+    },
+    {
+      "epoch": 0.7319611470860314,
+      "grad_norm": 3.9555208683013916,
+      "learning_rate": 1.7688798804188118e-06,
+      "loss": 0.9327,
+      "step": 9495
+    },
+    {
+      "epoch": 0.7320382362010485,
+      "grad_norm": 3.7948670387268066,
+      "learning_rate": 1.767927231424124e-06,
+      "loss": 0.9202,
+      "step": 9496
+    },
+    {
+      "epoch": 0.7321153253160654,
+      "grad_norm": 3.7444069385528564,
+      "learning_rate": 1.766974783932434e-06,
+      "loss": 0.8917,
+      "step": 9497
+    },
+    {
+      "epoch": 0.7321924144310823,
+      "grad_norm": 3.439579486846924,
+      "learning_rate": 1.766022538003122e-06,
+      "loss": 0.8429,
+      "step": 9498
+    },
+    {
+      "epoch": 0.7322695035460993,
+      "grad_norm": 3.7228710651397705,
+      "learning_rate": 1.7650704936955577e-06,
+      "loss": 0.8607,
+      "step": 9499
+    },
+    {
+      "epoch": 0.7323465926611162,
+      "grad_norm": 3.5966405868530273,
+      "learning_rate": 1.7641186510690916e-06,
+      "loss": 0.8669,
+      "step": 9500
+    },
+    {
+      "epoch": 0.7324236817761333,
+      "grad_norm": 3.7146801948547363,
+      "learning_rate": 1.763167010183069e-06,
+      "loss": 0.8685,
+      "step": 9501
+    },
+    {
+      "epoch": 0.7325007708911502,
+      "grad_norm": 3.474900484085083,
+      "learning_rate": 1.7622155710968187e-06,
+      "loss": 0.8428,
+      "step": 9502
+    },
+    {
+      "epoch": 0.7325778600061671,
+      "grad_norm": 4.253158092498779,
+      "learning_rate": 1.761264333869659e-06,
+      "loss": 0.925,
+      "step": 9503
+    },
+    {
+      "epoch": 0.7326549491211841,
+      "grad_norm": 4.116722106933594,
+      "learning_rate": 1.7603132985608945e-06,
+      "loss": 0.8248,
+      "step": 9504
+    },
+    {
+      "epoch": 0.732732038236201,
+      "grad_norm": 4.452220916748047,
+      "learning_rate": 1.7593624652298186e-06,
+      "loss": 1.0207,
+      "step": 9505
+    },
+    {
+      "epoch": 0.7328091273512181,
+      "grad_norm": 3.5835022926330566,
+      "learning_rate": 1.7584118339357076e-06,
+      "loss": 0.8986,
+      "step": 9506
+    },
+    {
+      "epoch": 0.732886216466235,
+      "grad_norm": 3.83974289894104,
+      "learning_rate": 1.7574614047378297e-06,
+      "loss": 0.9099,
+      "step": 9507
+    },
+    {
+      "epoch": 0.7329633055812519,
+      "grad_norm": 3.6427395343780518,
+      "learning_rate": 1.7565111776954401e-06,
+      "loss": 0.9213,
+      "step": 9508
+    },
+    {
+      "epoch": 0.7330403946962689,
+      "grad_norm": 3.9767303466796875,
+      "learning_rate": 1.7555611528677803e-06,
+      "loss": 0.9611,
+      "step": 9509
+    },
+    {
+      "epoch": 0.7331174838112858,
+      "grad_norm": 3.73007869720459,
+      "learning_rate": 1.7546113303140806e-06,
+      "loss": 0.9217,
+      "step": 9510
+    },
+    {
+      "epoch": 0.7331945729263029,
+      "grad_norm": 3.5436551570892334,
+      "learning_rate": 1.7536617100935544e-06,
+      "loss": 0.8044,
+      "step": 9511
+    },
+    {
+      "epoch": 0.7332716620413198,
+      "grad_norm": 3.655442237854004,
+      "learning_rate": 1.7527122922654077e-06,
+      "loss": 0.8905,
+      "step": 9512
+    },
+    {
+      "epoch": 0.7333487511563367,
+      "grad_norm": 3.6758716106414795,
+      "learning_rate": 1.751763076888831e-06,
+      "loss": 0.9612,
+      "step": 9513
+    },
+    {
+      "epoch": 0.7334258402713537,
+      "grad_norm": 3.906611204147339,
+      "learning_rate": 1.7508140640230037e-06,
+      "loss": 0.9355,
+      "step": 9514
+    },
+    {
+      "epoch": 0.7335029293863706,
+      "grad_norm": 3.8882009983062744,
+      "learning_rate": 1.7498652537270916e-06,
+      "loss": 0.8655,
+      "step": 9515
+    },
+    {
+      "epoch": 0.7335800185013877,
+      "grad_norm": 3.8022232055664062,
+      "learning_rate": 1.7489166460602496e-06,
+      "loss": 0.9754,
+      "step": 9516
+    },
+    {
+      "epoch": 0.7336571076164046,
+      "grad_norm": 3.5740768909454346,
+      "learning_rate": 1.7479682410816156e-06,
+      "loss": 0.9293,
+      "step": 9517
+    },
+    {
+      "epoch": 0.7337341967314215,
+      "grad_norm": 3.5046918392181396,
+      "learning_rate": 1.7470200388503184e-06,
+      "loss": 0.9062,
+      "step": 9518
+    },
+    {
+      "epoch": 0.7338112858464385,
+      "grad_norm": 3.434080123901367,
+      "learning_rate": 1.7460720394254748e-06,
+      "loss": 0.8451,
+      "step": 9519
+    },
+    {
+      "epoch": 0.7338883749614554,
+      "grad_norm": 3.992218255996704,
+      "learning_rate": 1.7451242428661868e-06,
+      "loss": 0.9591,
+      "step": 9520
+    },
+    {
+      "epoch": 0.7339654640764725,
+      "grad_norm": 3.7113566398620605,
+      "learning_rate": 1.7441766492315465e-06,
+      "loss": 0.9268,
+      "step": 9521
+    },
+    {
+      "epoch": 0.7340425531914894,
+      "grad_norm": 3.7934823036193848,
+      "learning_rate": 1.7432292585806277e-06,
+      "loss": 0.8757,
+      "step": 9522
+    },
+    {
+      "epoch": 0.7341196423065063,
+      "grad_norm": 3.7156882286071777,
+      "learning_rate": 1.742282070972498e-06,
+      "loss": 0.9287,
+      "step": 9523
+    },
+    {
+      "epoch": 0.7341967314215233,
+      "grad_norm": 3.4963877201080322,
+      "learning_rate": 1.7413350864662088e-06,
+      "loss": 0.8622,
+      "step": 9524
+    },
+    {
+      "epoch": 0.7342738205365402,
+      "grad_norm": 3.512899160385132,
+      "learning_rate": 1.7403883051207997e-06,
+      "loss": 0.8646,
+      "step": 9525
+    },
+    {
+      "epoch": 0.7343509096515572,
+      "grad_norm": 3.7643868923187256,
+      "learning_rate": 1.739441726995298e-06,
+      "loss": 0.9348,
+      "step": 9526
+    },
+    {
+      "epoch": 0.7344279987665742,
+      "grad_norm": 3.798492431640625,
+      "learning_rate": 1.7384953521487191e-06,
+      "loss": 0.9234,
+      "step": 9527
+    },
+    {
+      "epoch": 0.7345050878815911,
+      "grad_norm": 3.6906044483184814,
+      "learning_rate": 1.737549180640062e-06,
+      "loss": 0.983,
+      "step": 9528
+    },
+    {
+      "epoch": 0.7345821769966081,
+      "grad_norm": 3.6873204708099365,
+      "learning_rate": 1.7366032125283167e-06,
+      "loss": 0.9359,
+      "step": 9529
+    },
+    {
+      "epoch": 0.734659266111625,
+      "grad_norm": 3.9488134384155273,
+      "learning_rate": 1.7356574478724593e-06,
+      "loss": 0.9906,
+      "step": 9530
+    },
+    {
+      "epoch": 0.734736355226642,
+      "grad_norm": 3.8538029193878174,
+      "learning_rate": 1.7347118867314538e-06,
+      "loss": 0.9494,
+      "step": 9531
+    },
+    {
+      "epoch": 0.734813444341659,
+      "grad_norm": 3.5995075702667236,
+      "learning_rate": 1.7337665291642524e-06,
+      "loss": 0.8652,
+      "step": 9532
+    },
+    {
+      "epoch": 0.7348905334566759,
+      "grad_norm": 3.970221519470215,
+      "learning_rate": 1.7328213752297902e-06,
+      "loss": 0.9299,
+      "step": 9533
+    },
+    {
+      "epoch": 0.7349676225716929,
+      "grad_norm": 3.456852674484253,
+      "learning_rate": 1.7318764249869934e-06,
+      "loss": 0.9381,
+      "step": 9534
+    },
+    {
+      "epoch": 0.7350447116867098,
+      "grad_norm": 3.492973566055298,
+      "learning_rate": 1.730931678494776e-06,
+      "loss": 0.9255,
+      "step": 9535
+    },
+    {
+      "epoch": 0.7351218008017268,
+      "grad_norm": 3.6652021408081055,
+      "learning_rate": 1.7299871358120373e-06,
+      "loss": 0.9341,
+      "step": 9536
+    },
+    {
+      "epoch": 0.7351988899167438,
+      "grad_norm": 3.6004602909088135,
+      "learning_rate": 1.729042796997667e-06,
+      "loss": 0.9393,
+      "step": 9537
+    },
+    {
+      "epoch": 0.7352759790317607,
+      "grad_norm": 3.624760150909424,
+      "learning_rate": 1.7280986621105355e-06,
+      "loss": 0.9113,
+      "step": 9538
+    },
+    {
+      "epoch": 0.7353530681467777,
+      "grad_norm": 3.834989309310913,
+      "learning_rate": 1.7271547312095055e-06,
+      "loss": 0.9726,
+      "step": 9539
+    },
+    {
+      "epoch": 0.7354301572617946,
+      "grad_norm": 3.8131542205810547,
+      "learning_rate": 1.7262110043534285e-06,
+      "loss": 0.9446,
+      "step": 9540
+    },
+    {
+      "epoch": 0.7355072463768116,
+      "grad_norm": 3.5170748233795166,
+      "learning_rate": 1.7252674816011405e-06,
+      "loss": 0.8592,
+      "step": 9541
+    },
+    {
+      "epoch": 0.7355843354918286,
+      "grad_norm": 3.7116787433624268,
+      "learning_rate": 1.7243241630114665e-06,
+      "loss": 0.9016,
+      "step": 9542
+    },
+    {
+      "epoch": 0.7356614246068455,
+      "grad_norm": 3.888503313064575,
+      "learning_rate": 1.723381048643214e-06,
+      "loss": 0.938,
+      "step": 9543
+    },
+    {
+      "epoch": 0.7357385137218625,
+      "grad_norm": 3.5734593868255615,
+      "learning_rate": 1.722438138555183e-06,
+      "loss": 0.8522,
+      "step": 9544
+    },
+    {
+      "epoch": 0.7358156028368794,
+      "grad_norm": 3.421564817428589,
+      "learning_rate": 1.7214954328061588e-06,
+      "loss": 0.9318,
+      "step": 9545
+    },
+    {
+      "epoch": 0.7358926919518964,
+      "grad_norm": 3.7123959064483643,
+      "learning_rate": 1.7205529314549153e-06,
+      "loss": 0.9123,
+      "step": 9546
+    },
+    {
+      "epoch": 0.7359697810669134,
+      "grad_norm": 3.348281145095825,
+      "learning_rate": 1.7196106345602126e-06,
+      "loss": 0.8201,
+      "step": 9547
+    },
+    {
+      "epoch": 0.7360468701819303,
+      "grad_norm": 4.12083625793457,
+      "learning_rate": 1.7186685421807964e-06,
+      "loss": 0.9747,
+      "step": 9548
+    },
+    {
+      "epoch": 0.7361239592969473,
+      "grad_norm": 3.7178947925567627,
+      "learning_rate": 1.7177266543754018e-06,
+      "loss": 0.9056,
+      "step": 9549
+    },
+    {
+      "epoch": 0.7362010484119642,
+      "grad_norm": 4.042630672454834,
+      "learning_rate": 1.7167849712027507e-06,
+      "loss": 0.9108,
+      "step": 9550
+    },
+    {
+      "epoch": 0.7362781375269812,
+      "grad_norm": 3.405878782272339,
+      "learning_rate": 1.7158434927215528e-06,
+      "loss": 0.8586,
+      "step": 9551
+    },
+    {
+      "epoch": 0.7363552266419982,
+      "grad_norm": 3.563534736633301,
+      "learning_rate": 1.7149022189905041e-06,
+      "loss": 0.8012,
+      "step": 9552
+    },
+    {
+      "epoch": 0.7364323157570151,
+      "grad_norm": 3.4462010860443115,
+      "learning_rate": 1.7139611500682896e-06,
+      "loss": 0.874,
+      "step": 9553
+    },
+    {
+      "epoch": 0.7365094048720321,
+      "grad_norm": 3.5920019149780273,
+      "learning_rate": 1.713020286013577e-06,
+      "loss": 0.8469,
+      "step": 9554
+    },
+    {
+      "epoch": 0.736586493987049,
+      "grad_norm": 3.872889995574951,
+      "learning_rate": 1.7120796268850254e-06,
+      "loss": 0.8842,
+      "step": 9555
+    },
+    {
+      "epoch": 0.736663583102066,
+      "grad_norm": 4.169905185699463,
+      "learning_rate": 1.7111391727412807e-06,
+      "loss": 1.0117,
+      "step": 9556
+    },
+    {
+      "epoch": 0.736740672217083,
+      "grad_norm": 4.358822345733643,
+      "learning_rate": 1.7101989236409754e-06,
+      "loss": 0.9126,
+      "step": 9557
+    },
+    {
+      "epoch": 0.7368177613320999,
+      "grad_norm": 3.7961339950561523,
+      "learning_rate": 1.7092588796427306e-06,
+      "loss": 1.0046,
+      "step": 9558
+    },
+    {
+      "epoch": 0.7368948504471169,
+      "grad_norm": 3.6711761951446533,
+      "learning_rate": 1.7083190408051498e-06,
+      "loss": 0.8484,
+      "step": 9559
+    },
+    {
+      "epoch": 0.7369719395621338,
+      "grad_norm": 3.702301502227783,
+      "learning_rate": 1.7073794071868283e-06,
+      "loss": 0.8271,
+      "step": 9560
+    },
+    {
+      "epoch": 0.7370490286771508,
+      "grad_norm": 3.758284330368042,
+      "learning_rate": 1.7064399788463482e-06,
+      "loss": 1.0023,
+      "step": 9561
+    },
+    {
+      "epoch": 0.7371261177921677,
+      "grad_norm": 3.5002236366271973,
+      "learning_rate": 1.7055007558422776e-06,
+      "loss": 0.8328,
+      "step": 9562
+    },
+    {
+      "epoch": 0.7372032069071847,
+      "grad_norm": 3.4732120037078857,
+      "learning_rate": 1.7045617382331726e-06,
+      "loss": 0.794,
+      "step": 9563
+    },
+    {
+      "epoch": 0.7372802960222017,
+      "grad_norm": 4.162208080291748,
+      "learning_rate": 1.7036229260775766e-06,
+      "loss": 1.033,
+      "step": 9564
+    },
+    {
+      "epoch": 0.7373573851372186,
+      "grad_norm": 3.8932626247406006,
+      "learning_rate": 1.7026843194340176e-06,
+      "loss": 0.9436,
+      "step": 9565
+    },
+    {
+      "epoch": 0.7374344742522356,
+      "grad_norm": 3.9444522857666016,
+      "learning_rate": 1.7017459183610142e-06,
+      "loss": 0.9912,
+      "step": 9566
+    },
+    {
+      "epoch": 0.7375115633672525,
+      "grad_norm": 3.844452381134033,
+      "learning_rate": 1.70080772291707e-06,
+      "loss": 0.8048,
+      "step": 9567
+    },
+    {
+      "epoch": 0.7375886524822695,
+      "grad_norm": 3.734955072402954,
+      "learning_rate": 1.6998697331606777e-06,
+      "loss": 0.9294,
+      "step": 9568
+    },
+    {
+      "epoch": 0.7376657415972865,
+      "grad_norm": 3.940300226211548,
+      "learning_rate": 1.6989319491503175e-06,
+      "loss": 0.9772,
+      "step": 9569
+    },
+    {
+      "epoch": 0.7377428307123034,
+      "grad_norm": 3.8375890254974365,
+      "learning_rate": 1.697994370944452e-06,
+      "loss": 0.9545,
+      "step": 9570
+    },
+    {
+      "epoch": 0.7378199198273204,
+      "grad_norm": 3.75097393989563,
+      "learning_rate": 1.6970569986015356e-06,
+      "loss": 0.9672,
+      "step": 9571
+    },
+    {
+      "epoch": 0.7378970089423373,
+      "grad_norm": 3.56791353225708,
+      "learning_rate": 1.6961198321800092e-06,
+      "loss": 0.9334,
+      "step": 9572
+    },
+    {
+      "epoch": 0.7379740980573543,
+      "grad_norm": 3.3650386333465576,
+      "learning_rate": 1.6951828717383e-06,
+      "loss": 0.8871,
+      "step": 9573
+    },
+    {
+      "epoch": 0.7380511871723713,
+      "grad_norm": 3.7570347785949707,
+      "learning_rate": 1.694246117334823e-06,
+      "loss": 0.9818,
+      "step": 9574
+    },
+    {
+      "epoch": 0.7381282762873882,
+      "grad_norm": 3.496140241622925,
+      "learning_rate": 1.6933095690279805e-06,
+      "loss": 0.907,
+      "step": 9575
+    },
+    {
+      "epoch": 0.7382053654024052,
+      "grad_norm": 3.67641019821167,
+      "learning_rate": 1.6923732268761594e-06,
+      "loss": 0.8191,
+      "step": 9576
+    },
+    {
+      "epoch": 0.7382824545174221,
+      "grad_norm": 3.6642329692840576,
+      "learning_rate": 1.691437090937737e-06,
+      "loss": 0.8536,
+      "step": 9577
+    },
+    {
+      "epoch": 0.738359543632439,
+      "grad_norm": 3.9450578689575195,
+      "learning_rate": 1.6905011612710764e-06,
+      "loss": 0.8556,
+      "step": 9578
+    },
+    {
+      "epoch": 0.7384366327474561,
+      "grad_norm": 4.003316402435303,
+      "learning_rate": 1.6895654379345282e-06,
+      "loss": 0.9955,
+      "step": 9579
+    },
+    {
+      "epoch": 0.738513721862473,
+      "grad_norm": 3.8521080017089844,
+      "learning_rate": 1.6886299209864316e-06,
+      "loss": 0.888,
+      "step": 9580
+    },
+    {
+      "epoch": 0.73859081097749,
+      "grad_norm": 3.9132065773010254,
+      "learning_rate": 1.687694610485107e-06,
+      "loss": 0.9501,
+      "step": 9581
+    },
+    {
+      "epoch": 0.7386679000925069,
+      "grad_norm": 3.6591548919677734,
+      "learning_rate": 1.6867595064888693e-06,
+      "loss": 0.8745,
+      "step": 9582
+    },
+    {
+      "epoch": 0.7387449892075239,
+      "grad_norm": 4.419125556945801,
+      "learning_rate": 1.685824609056015e-06,
+      "loss": 0.9704,
+      "step": 9583
+    },
+    {
+      "epoch": 0.7388220783225409,
+      "grad_norm": 3.558927297592163,
+      "learning_rate": 1.6848899182448347e-06,
+      "loss": 0.9493,
+      "step": 9584
+    },
+    {
+      "epoch": 0.7388991674375578,
+      "grad_norm": 3.7360217571258545,
+      "learning_rate": 1.6839554341135973e-06,
+      "loss": 0.9127,
+      "step": 9585
+    },
+    {
+      "epoch": 0.7389762565525748,
+      "grad_norm": 4.130924701690674,
+      "learning_rate": 1.683021156720564e-06,
+      "loss": 0.8786,
+      "step": 9586
+    },
+    {
+      "epoch": 0.7390533456675917,
+      "grad_norm": 3.717590808868408,
+      "learning_rate": 1.6820870861239824e-06,
+      "loss": 0.8723,
+      "step": 9587
+    },
+    {
+      "epoch": 0.7391304347826086,
+      "grad_norm": 3.745699405670166,
+      "learning_rate": 1.6811532223820875e-06,
+      "loss": 0.8268,
+      "step": 9588
+    },
+    {
+      "epoch": 0.7392075238976257,
+      "grad_norm": 3.7815754413604736,
+      "learning_rate": 1.6802195655531e-06,
+      "loss": 0.9505,
+      "step": 9589
+    },
+    {
+      "epoch": 0.7392846130126426,
+      "grad_norm": 4.17587947845459,
+      "learning_rate": 1.6792861156952312e-06,
+      "loss": 0.9202,
+      "step": 9590
+    },
+    {
+      "epoch": 0.7393617021276596,
+      "grad_norm": 3.6216659545898438,
+      "learning_rate": 1.6783528728666725e-06,
+      "loss": 0.8253,
+      "step": 9591
+    },
+    {
+      "epoch": 0.7394387912426765,
+      "grad_norm": 3.885458469390869,
+      "learning_rate": 1.677419837125609e-06,
+      "loss": 1.0538,
+      "step": 9592
+    },
+    {
+      "epoch": 0.7395158803576934,
+      "grad_norm": 3.8355486392974854,
+      "learning_rate": 1.676487008530211e-06,
+      "loss": 0.9182,
+      "step": 9593
+    },
+    {
+      "epoch": 0.7395929694727105,
+      "grad_norm": 4.003592014312744,
+      "learning_rate": 1.6755543871386342e-06,
+      "loss": 0.9035,
+      "step": 9594
+    },
+    {
+      "epoch": 0.7396700585877274,
+      "grad_norm": 3.761218309402466,
+      "learning_rate": 1.6746219730090262e-06,
+      "loss": 0.8466,
+      "step": 9595
+    },
+    {
+      "epoch": 0.7397471477027444,
+      "grad_norm": 3.5276873111724854,
+      "learning_rate": 1.6736897661995132e-06,
+      "loss": 0.8705,
+      "step": 9596
+    },
+    {
+      "epoch": 0.7398242368177613,
+      "grad_norm": 3.678708791732788,
+      "learning_rate": 1.6727577667682165e-06,
+      "loss": 0.8528,
+      "step": 9597
+    },
+    {
+      "epoch": 0.7399013259327782,
+      "grad_norm": 3.7199313640594482,
+      "learning_rate": 1.6718259747732407e-06,
+      "loss": 0.9636,
+      "step": 9598
+    },
+    {
+      "epoch": 0.7399784150477953,
+      "grad_norm": 3.733581304550171,
+      "learning_rate": 1.6708943902726783e-06,
+      "loss": 0.9271,
+      "step": 9599
+    },
+    {
+      "epoch": 0.7400555041628122,
+      "grad_norm": 3.6299502849578857,
+      "learning_rate": 1.6699630133246087e-06,
+      "loss": 0.9428,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7401325932778292,
+      "grad_norm": 3.8488755226135254,
+      "learning_rate": 1.6690318439871e-06,
+      "loss": 0.7749,
+      "step": 9601
+    },
+    {
+      "epoch": 0.7402096823928461,
+      "grad_norm": 3.60003399848938,
+      "learning_rate": 1.6681008823182027e-06,
+      "loss": 0.9667,
+      "step": 9602
+    },
+    {
+      "epoch": 0.740286771507863,
+      "grad_norm": 3.882335662841797,
+      "learning_rate": 1.6671701283759596e-06,
+      "loss": 0.9634,
+      "step": 9603
+    },
+    {
+      "epoch": 0.7403638606228801,
+      "grad_norm": 3.8450510501861572,
+      "learning_rate": 1.6662395822183975e-06,
+      "loss": 0.7925,
+      "step": 9604
+    },
+    {
+      "epoch": 0.740440949737897,
+      "grad_norm": 3.650480031967163,
+      "learning_rate": 1.6653092439035312e-06,
+      "loss": 0.8782,
+      "step": 9605
+    },
+    {
+      "epoch": 0.740518038852914,
+      "grad_norm": 3.5009539127349854,
+      "learning_rate": 1.6643791134893644e-06,
+      "loss": 0.952,
+      "step": 9606
+    },
+    {
+      "epoch": 0.7405951279679309,
+      "grad_norm": 3.5242550373077393,
+      "learning_rate": 1.6634491910338829e-06,
+      "loss": 0.9471,
+      "step": 9607
+    },
+    {
+      "epoch": 0.7406722170829478,
+      "grad_norm": 3.7609641551971436,
+      "learning_rate": 1.6625194765950636e-06,
+      "loss": 0.8637,
+      "step": 9608
+    },
+    {
+      "epoch": 0.7407493061979649,
+      "grad_norm": 3.836843729019165,
+      "learning_rate": 1.6615899702308696e-06,
+      "loss": 0.9317,
+      "step": 9609
+    },
+    {
+      "epoch": 0.7408263953129818,
+      "grad_norm": 4.039276599884033,
+      "learning_rate": 1.6606606719992513e-06,
+      "loss": 0.9741,
+      "step": 9610
+    },
+    {
+      "epoch": 0.7409034844279988,
+      "grad_norm": 3.873875379562378,
+      "learning_rate": 1.6597315819581449e-06,
+      "loss": 0.8736,
+      "step": 9611
+    },
+    {
+      "epoch": 0.7409805735430157,
+      "grad_norm": 3.878174066543579,
+      "learning_rate": 1.6588027001654765e-06,
+      "loss": 0.9261,
+      "step": 9612
+    },
+    {
+      "epoch": 0.7410576626580326,
+      "grad_norm": 3.606624126434326,
+      "learning_rate": 1.6578740266791532e-06,
+      "loss": 0.8731,
+      "step": 9613
+    },
+    {
+      "epoch": 0.7411347517730497,
+      "grad_norm": 3.7116219997406006,
+      "learning_rate": 1.6569455615570757e-06,
+      "loss": 0.8865,
+      "step": 9614
+    },
+    {
+      "epoch": 0.7412118408880666,
+      "grad_norm": 3.811680316925049,
+      "learning_rate": 1.6560173048571277e-06,
+      "loss": 0.938,
+      "step": 9615
+    },
+    {
+      "epoch": 0.7412889300030836,
+      "grad_norm": 3.9601168632507324,
+      "learning_rate": 1.6550892566371823e-06,
+      "loss": 0.9875,
+      "step": 9616
+    },
+    {
+      "epoch": 0.7413660191181005,
+      "grad_norm": 3.370882034301758,
+      "learning_rate": 1.654161416955099e-06,
+      "loss": 0.878,
+      "step": 9617
+    },
+    {
+      "epoch": 0.7414431082331174,
+      "grad_norm": 4.08239221572876,
+      "learning_rate": 1.653233785868722e-06,
+      "loss": 1.0293,
+      "step": 9618
+    },
+    {
+      "epoch": 0.7415201973481345,
+      "grad_norm": 3.562781572341919,
+      "learning_rate": 1.6523063634358844e-06,
+      "loss": 0.8308,
+      "step": 9619
+    },
+    {
+      "epoch": 0.7415972864631514,
+      "grad_norm": 3.801961898803711,
+      "learning_rate": 1.6513791497144071e-06,
+      "loss": 0.9176,
+      "step": 9620
+    },
+    {
+      "epoch": 0.7416743755781684,
+      "grad_norm": 3.9425201416015625,
+      "learning_rate": 1.6504521447620969e-06,
+      "loss": 0.9052,
+      "step": 9621
+    },
+    {
+      "epoch": 0.7417514646931853,
+      "grad_norm": 3.5563175678253174,
+      "learning_rate": 1.649525348636748e-06,
+      "loss": 1.0543,
+      "step": 9622
+    },
+    {
+      "epoch": 0.7418285538082022,
+      "grad_norm": 3.3809213638305664,
+      "learning_rate": 1.6485987613961423e-06,
+      "loss": 0.8341,
+      "step": 9623
+    },
+    {
+      "epoch": 0.7419056429232193,
+      "grad_norm": 3.588660717010498,
+      "learning_rate": 1.6476723830980451e-06,
+      "loss": 0.7934,
+      "step": 9624
+    },
+    {
+      "epoch": 0.7419827320382362,
+      "grad_norm": 3.640718460083008,
+      "learning_rate": 1.6467462138002126e-06,
+      "loss": 0.9354,
+      "step": 9625
+    },
+    {
+      "epoch": 0.7420598211532532,
+      "grad_norm": 3.681979179382324,
+      "learning_rate": 1.6458202535603867e-06,
+      "loss": 0.9655,
+      "step": 9626
+    },
+    {
+      "epoch": 0.7421369102682701,
+      "grad_norm": 4.223269939422607,
+      "learning_rate": 1.6448945024362962e-06,
+      "loss": 0.965,
+      "step": 9627
+    },
+    {
+      "epoch": 0.742213999383287,
+      "grad_norm": 3.7559590339660645,
+      "learning_rate": 1.6439689604856568e-06,
+      "loss": 0.9417,
+      "step": 9628
+    },
+    {
+      "epoch": 0.742291088498304,
+      "grad_norm": 3.631321668624878,
+      "learning_rate": 1.6430436277661715e-06,
+      "loss": 0.8356,
+      "step": 9629
+    },
+    {
+      "epoch": 0.742368177613321,
+      "grad_norm": 3.696200370788574,
+      "learning_rate": 1.6421185043355304e-06,
+      "loss": 0.9243,
+      "step": 9630
+    },
+    {
+      "epoch": 0.742445266728338,
+      "grad_norm": 3.6229546070098877,
+      "learning_rate": 1.6411935902514086e-06,
+      "loss": 0.8577,
+      "step": 9631
+    },
+    {
+      "epoch": 0.7425223558433549,
+      "grad_norm": 3.96211314201355,
+      "learning_rate": 1.6402688855714733e-06,
+      "loss": 0.9646,
+      "step": 9632
+    },
+    {
+      "epoch": 0.7425994449583718,
+      "grad_norm": 3.6700477600097656,
+      "learning_rate": 1.6393443903533707e-06,
+      "loss": 0.9983,
+      "step": 9633
+    },
+    {
+      "epoch": 0.7426765340733888,
+      "grad_norm": 4.209187030792236,
+      "learning_rate": 1.6384201046547399e-06,
+      "loss": 1.0566,
+      "step": 9634
+    },
+    {
+      "epoch": 0.7427536231884058,
+      "grad_norm": 3.9726881980895996,
+      "learning_rate": 1.6374960285332053e-06,
+      "loss": 1.0066,
+      "step": 9635
+    },
+    {
+      "epoch": 0.7428307123034228,
+      "grad_norm": 3.917154550552368,
+      "learning_rate": 1.6365721620463786e-06,
+      "loss": 0.9755,
+      "step": 9636
+    },
+    {
+      "epoch": 0.7429078014184397,
+      "grad_norm": 4.186873435974121,
+      "learning_rate": 1.6356485052518578e-06,
+      "loss": 0.9928,
+      "step": 9637
+    },
+    {
+      "epoch": 0.7429848905334566,
+      "grad_norm": 3.5882840156555176,
+      "learning_rate": 1.6347250582072305e-06,
+      "loss": 0.8609,
+      "step": 9638
+    },
+    {
+      "epoch": 0.7430619796484736,
+      "grad_norm": 3.6521730422973633,
+      "learning_rate": 1.6338018209700647e-06,
+      "loss": 0.9443,
+      "step": 9639
+    },
+    {
+      "epoch": 0.7431390687634906,
+      "grad_norm": 3.2607851028442383,
+      "learning_rate": 1.6328787935979207e-06,
+      "loss": 0.8569,
+      "step": 9640
+    },
+    {
+      "epoch": 0.7432161578785076,
+      "grad_norm": 3.7790210247039795,
+      "learning_rate": 1.6319559761483461e-06,
+      "loss": 0.8764,
+      "step": 9641
+    },
+    {
+      "epoch": 0.7432932469935245,
+      "grad_norm": 3.771569013595581,
+      "learning_rate": 1.631033368678872e-06,
+      "loss": 0.9934,
+      "step": 9642
+    },
+    {
+      "epoch": 0.7433703361085414,
+      "grad_norm": 3.3746719360351562,
+      "learning_rate": 1.6301109712470214e-06,
+      "loss": 0.7936,
+      "step": 9643
+    },
+    {
+      "epoch": 0.7434474252235584,
+      "grad_norm": 3.9737672805786133,
+      "learning_rate": 1.6291887839102966e-06,
+      "loss": 0.9538,
+      "step": 9644
+    },
+    {
+      "epoch": 0.7435245143385754,
+      "grad_norm": 4.225729465484619,
+      "learning_rate": 1.6282668067261935e-06,
+      "loss": 0.8667,
+      "step": 9645
+    },
+    {
+      "epoch": 0.7436016034535924,
+      "grad_norm": 3.5801405906677246,
+      "learning_rate": 1.6273450397521922e-06,
+      "loss": 0.8745,
+      "step": 9646
+    },
+    {
+      "epoch": 0.7436786925686093,
+      "grad_norm": 3.57384991645813,
+      "learning_rate": 1.6264234830457603e-06,
+      "loss": 0.8146,
+      "step": 9647
+    },
+    {
+      "epoch": 0.7437557816836263,
+      "grad_norm": 3.640131711959839,
+      "learning_rate": 1.625502136664352e-06,
+      "loss": 0.9404,
+      "step": 9648
+    },
+    {
+      "epoch": 0.7438328707986432,
+      "grad_norm": 3.9445488452911377,
+      "learning_rate": 1.62458100066541e-06,
+      "loss": 1.0291,
+      "step": 9649
+    },
+    {
+      "epoch": 0.7439099599136602,
+      "grad_norm": 3.508561134338379,
+      "learning_rate": 1.6236600751063597e-06,
+      "loss": 0.8162,
+      "step": 9650
+    },
+    {
+      "epoch": 0.7439870490286772,
+      "grad_norm": 3.63757586479187,
+      "learning_rate": 1.6227393600446168e-06,
+      "loss": 0.8668,
+      "step": 9651
+    },
+    {
+      "epoch": 0.7440641381436941,
+      "grad_norm": 3.6220133304595947,
+      "learning_rate": 1.6218188555375836e-06,
+      "loss": 0.8303,
+      "step": 9652
+    },
+    {
+      "epoch": 0.7441412272587111,
+      "grad_norm": 3.732778310775757,
+      "learning_rate": 1.6208985616426488e-06,
+      "loss": 0.9131,
+      "step": 9653
+    },
+    {
+      "epoch": 0.744218316373728,
+      "grad_norm": 3.7289443016052246,
+      "learning_rate": 1.619978478417189e-06,
+      "loss": 1.0164,
+      "step": 9654
+    },
+    {
+      "epoch": 0.744295405488745,
+      "grad_norm": 3.5976722240448,
+      "learning_rate": 1.6190586059185642e-06,
+      "loss": 0.8991,
+      "step": 9655
+    },
+    {
+      "epoch": 0.744372494603762,
+      "grad_norm": 3.97445011138916,
+      "learning_rate": 1.618138944204125e-06,
+      "loss": 0.9376,
+      "step": 9656
+    },
+    {
+      "epoch": 0.7444495837187789,
+      "grad_norm": 3.541771173477173,
+      "learning_rate": 1.617219493331208e-06,
+      "loss": 0.929,
+      "step": 9657
+    },
+    {
+      "epoch": 0.7445266728337959,
+      "grad_norm": 3.4124934673309326,
+      "learning_rate": 1.6163002533571348e-06,
+      "loss": 0.8519,
+      "step": 9658
+    },
+    {
+      "epoch": 0.7446037619488128,
+      "grad_norm": 3.9390342235565186,
+      "learning_rate": 1.6153812243392169e-06,
+      "loss": 0.8352,
+      "step": 9659
+    },
+    {
+      "epoch": 0.7446808510638298,
+      "grad_norm": 3.7517385482788086,
+      "learning_rate": 1.6144624063347514e-06,
+      "loss": 1.0166,
+      "step": 9660
+    },
+    {
+      "epoch": 0.7447579401788468,
+      "grad_norm": 3.8233649730682373,
+      "learning_rate": 1.6135437994010195e-06,
+      "loss": 0.8967,
+      "step": 9661
+    },
+    {
+      "epoch": 0.7448350292938637,
+      "grad_norm": 4.343902111053467,
+      "learning_rate": 1.6126254035952926e-06,
+      "loss": 0.9621,
+      "step": 9662
+    },
+    {
+      "epoch": 0.7449121184088807,
+      "grad_norm": 3.828294038772583,
+      "learning_rate": 1.6117072189748285e-06,
+      "loss": 0.9517,
+      "step": 9663
+    },
+    {
+      "epoch": 0.7449892075238976,
+      "grad_norm": 4.02465295791626,
+      "learning_rate": 1.6107892455968704e-06,
+      "loss": 0.9339,
+      "step": 9664
+    },
+    {
+      "epoch": 0.7450662966389145,
+      "grad_norm": 3.764190673828125,
+      "learning_rate": 1.6098714835186512e-06,
+      "loss": 0.9529,
+      "step": 9665
+    },
+    {
+      "epoch": 0.7451433857539316,
+      "grad_norm": 3.6371004581451416,
+      "learning_rate": 1.6089539327973857e-06,
+      "loss": 0.8936,
+      "step": 9666
+    },
+    {
+      "epoch": 0.7452204748689485,
+      "grad_norm": 3.513137102127075,
+      "learning_rate": 1.6080365934902798e-06,
+      "loss": 0.9337,
+      "step": 9667
+    },
+    {
+      "epoch": 0.7452975639839655,
+      "grad_norm": 4.115825653076172,
+      "learning_rate": 1.6071194656545246e-06,
+      "loss": 0.9739,
+      "step": 9668
+    },
+    {
+      "epoch": 0.7453746530989824,
+      "grad_norm": 4.290143966674805,
+      "learning_rate": 1.6062025493472988e-06,
+      "loss": 1.0652,
+      "step": 9669
+    },
+    {
+      "epoch": 0.7454517422139993,
+      "grad_norm": 3.6070821285247803,
+      "learning_rate": 1.6052858446257674e-06,
+      "loss": 0.866,
+      "step": 9670
+    },
+    {
+      "epoch": 0.7455288313290164,
+      "grad_norm": 3.7635834217071533,
+      "learning_rate": 1.6043693515470833e-06,
+      "loss": 0.8835,
+      "step": 9671
+    },
+    {
+      "epoch": 0.7456059204440333,
+      "grad_norm": 3.815763473510742,
+      "learning_rate": 1.6034530701683804e-06,
+      "loss": 0.8308,
+      "step": 9672
+    },
+    {
+      "epoch": 0.7456830095590503,
+      "grad_norm": 3.4556825160980225,
+      "learning_rate": 1.6025370005467889e-06,
+      "loss": 0.8331,
+      "step": 9673
+    },
+    {
+      "epoch": 0.7457600986740672,
+      "grad_norm": 3.467095375061035,
+      "learning_rate": 1.6016211427394196e-06,
+      "loss": 0.7999,
+      "step": 9674
+    },
+    {
+      "epoch": 0.7458371877890841,
+      "grad_norm": 4.090667724609375,
+      "learning_rate": 1.600705496803373e-06,
+      "loss": 0.8447,
+      "step": 9675
+    },
+    {
+      "epoch": 0.7459142769041012,
+      "grad_norm": 3.751192092895508,
+      "learning_rate": 1.5997900627957318e-06,
+      "loss": 0.9856,
+      "step": 9676
+    },
+    {
+      "epoch": 0.7459913660191181,
+      "grad_norm": 3.7970805168151855,
+      "learning_rate": 1.5988748407735698e-06,
+      "loss": 0.8911,
+      "step": 9677
+    },
+    {
+      "epoch": 0.7460684551341351,
+      "grad_norm": 3.705199956893921,
+      "learning_rate": 1.597959830793947e-06,
+      "loss": 0.8308,
+      "step": 9678
+    },
+    {
+      "epoch": 0.746145544249152,
+      "grad_norm": 3.4573512077331543,
+      "learning_rate": 1.5970450329139087e-06,
+      "loss": 0.9452,
+      "step": 9679
+    },
+    {
+      "epoch": 0.7462226333641689,
+      "grad_norm": 3.6191141605377197,
+      "learning_rate": 1.5961304471904897e-06,
+      "loss": 0.953,
+      "step": 9680
+    },
+    {
+      "epoch": 0.746299722479186,
+      "grad_norm": 4.017669677734375,
+      "learning_rate": 1.595216073680707e-06,
+      "loss": 0.9276,
+      "step": 9681
+    },
+    {
+      "epoch": 0.7463768115942029,
+      "grad_norm": 3.8973758220672607,
+      "learning_rate": 1.5943019124415687e-06,
+      "loss": 0.8892,
+      "step": 9682
+    },
+    {
+      "epoch": 0.7464539007092199,
+      "grad_norm": 3.952500104904175,
+      "learning_rate": 1.593387963530067e-06,
+      "loss": 0.9615,
+      "step": 9683
+    },
+    {
+      "epoch": 0.7465309898242368,
+      "grad_norm": 3.6438992023468018,
+      "learning_rate": 1.5924742270031823e-06,
+      "loss": 0.8843,
+      "step": 9684
+    },
+    {
+      "epoch": 0.7466080789392537,
+      "grad_norm": 3.6509311199188232,
+      "learning_rate": 1.591560702917882e-06,
+      "loss": 0.9018,
+      "step": 9685
+    },
+    {
+      "epoch": 0.7466851680542708,
+      "grad_norm": 4.112860202789307,
+      "learning_rate": 1.5906473913311204e-06,
+      "loss": 0.9559,
+      "step": 9686
+    },
+    {
+      "epoch": 0.7467622571692877,
+      "grad_norm": 3.7320613861083984,
+      "learning_rate": 1.589734292299835e-06,
+      "loss": 0.9533,
+      "step": 9687
+    },
+    {
+      "epoch": 0.7468393462843047,
+      "grad_norm": 3.6934304237365723,
+      "learning_rate": 1.5888214058809543e-06,
+      "loss": 0.9533,
+      "step": 9688
+    },
+    {
+      "epoch": 0.7469164353993216,
+      "grad_norm": 3.7583377361297607,
+      "learning_rate": 1.587908732131392e-06,
+      "loss": 0.8712,
+      "step": 9689
+    },
+    {
+      "epoch": 0.7469935245143385,
+      "grad_norm": 3.579685926437378,
+      "learning_rate": 1.5869962711080483e-06,
+      "loss": 0.9753,
+      "step": 9690
+    },
+    {
+      "epoch": 0.7470706136293556,
+      "grad_norm": 3.7619788646698,
+      "learning_rate": 1.5860840228678126e-06,
+      "loss": 0.9095,
+      "step": 9691
+    },
+    {
+      "epoch": 0.7471477027443725,
+      "grad_norm": 3.7893290519714355,
+      "learning_rate": 1.5851719874675552e-06,
+      "loss": 0.9442,
+      "step": 9692
+    },
+    {
+      "epoch": 0.7472247918593895,
+      "grad_norm": 3.8225247859954834,
+      "learning_rate": 1.5842601649641388e-06,
+      "loss": 0.8569,
+      "step": 9693
+    },
+    {
+      "epoch": 0.7473018809744064,
+      "grad_norm": 3.59032940864563,
+      "learning_rate": 1.58334855541441e-06,
+      "loss": 0.9241,
+      "step": 9694
+    },
+    {
+      "epoch": 0.7473789700894233,
+      "grad_norm": 3.8772151470184326,
+      "learning_rate": 1.5824371588752042e-06,
+      "loss": 0.8766,
+      "step": 9695
+    },
+    {
+      "epoch": 0.7474560592044404,
+      "grad_norm": 3.8452553749084473,
+      "learning_rate": 1.5815259754033407e-06,
+      "loss": 0.9639,
+      "step": 9696
+    },
+    {
+      "epoch": 0.7475331483194573,
+      "grad_norm": 4.228387355804443,
+      "learning_rate": 1.5806150050556302e-06,
+      "loss": 1.0822,
+      "step": 9697
+    },
+    {
+      "epoch": 0.7476102374344743,
+      "grad_norm": 3.8701577186584473,
+      "learning_rate": 1.579704247888863e-06,
+      "loss": 0.9458,
+      "step": 9698
+    },
+    {
+      "epoch": 0.7476873265494912,
+      "grad_norm": 4.092501640319824,
+      "learning_rate": 1.5787937039598217e-06,
+      "loss": 0.9963,
+      "step": 9699
+    },
+    {
+      "epoch": 0.7477644156645081,
+      "grad_norm": 3.4330146312713623,
+      "learning_rate": 1.5778833733252735e-06,
+      "loss": 0.889,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7478415047795252,
+      "grad_norm": 3.8023056983947754,
+      "learning_rate": 1.5769732560419742e-06,
+      "loss": 0.8818,
+      "step": 9701
+    },
+    {
+      "epoch": 0.7479185938945421,
+      "grad_norm": 3.714808464050293,
+      "learning_rate": 1.5760633521666652e-06,
+      "loss": 1.0705,
+      "step": 9702
+    },
+    {
+      "epoch": 0.7479956830095591,
+      "grad_norm": 4.097212314605713,
+      "learning_rate": 1.5751536617560715e-06,
+      "loss": 0.8879,
+      "step": 9703
+    },
+    {
+      "epoch": 0.748072772124576,
+      "grad_norm": 3.7691214084625244,
+      "learning_rate": 1.57424418486691e-06,
+      "loss": 1.0334,
+      "step": 9704
+    },
+    {
+      "epoch": 0.7481498612395929,
+      "grad_norm": 3.9305317401885986,
+      "learning_rate": 1.5733349215558802e-06,
+      "loss": 1.0065,
+      "step": 9705
+    },
+    {
+      "epoch": 0.74822695035461,
+      "grad_norm": 4.348329067230225,
+      "learning_rate": 1.5724258718796714e-06,
+      "loss": 1.1138,
+      "step": 9706
+    },
+    {
+      "epoch": 0.7483040394696269,
+      "grad_norm": 4.216433048248291,
+      "learning_rate": 1.5715170358949572e-06,
+      "loss": 1.008,
+      "step": 9707
+    },
+    {
+      "epoch": 0.7483811285846439,
+      "grad_norm": 3.6003470420837402,
+      "learning_rate": 1.5706084136584016e-06,
+      "loss": 0.8693,
+      "step": 9708
+    },
+    {
+      "epoch": 0.7484582176996608,
+      "grad_norm": 4.085456848144531,
+      "learning_rate": 1.5697000052266475e-06,
+      "loss": 0.8981,
+      "step": 9709
+    },
+    {
+      "epoch": 0.7485353068146777,
+      "grad_norm": 3.8469676971435547,
+      "learning_rate": 1.5687918106563326e-06,
+      "loss": 0.838,
+      "step": 9710
+    },
+    {
+      "epoch": 0.7486123959296948,
+      "grad_norm": 3.8500454425811768,
+      "learning_rate": 1.5678838300040783e-06,
+      "loss": 0.9659,
+      "step": 9711
+    },
+    {
+      "epoch": 0.7486894850447117,
+      "grad_norm": 4.028532028198242,
+      "learning_rate": 1.5669760633264908e-06,
+      "loss": 0.9956,
+      "step": 9712
+    },
+    {
+      "epoch": 0.7487665741597287,
+      "grad_norm": 3.699516534805298,
+      "learning_rate": 1.5660685106801677e-06,
+      "loss": 0.9479,
+      "step": 9713
+    },
+    {
+      "epoch": 0.7488436632747456,
+      "grad_norm": 3.5519814491271973,
+      "learning_rate": 1.5651611721216865e-06,
+      "loss": 0.8591,
+      "step": 9714
+    },
+    {
+      "epoch": 0.7489207523897625,
+      "grad_norm": 3.9628407955169678,
+      "learning_rate": 1.5642540477076169e-06,
+      "loss": 1.0351,
+      "step": 9715
+    },
+    {
+      "epoch": 0.7489978415047795,
+      "grad_norm": 3.8148396015167236,
+      "learning_rate": 1.5633471374945113e-06,
+      "loss": 1.0393,
+      "step": 9716
+    },
+    {
+      "epoch": 0.7490749306197965,
+      "grad_norm": 4.2555999755859375,
+      "learning_rate": 1.5624404415389166e-06,
+      "loss": 0.8896,
+      "step": 9717
+    },
+    {
+      "epoch": 0.7491520197348135,
+      "grad_norm": 3.7964389324188232,
+      "learning_rate": 1.5615339598973544e-06,
+      "loss": 0.9671,
+      "step": 9718
+    },
+    {
+      "epoch": 0.7492291088498304,
+      "grad_norm": 3.563861608505249,
+      "learning_rate": 1.560627692626342e-06,
+      "loss": 0.9568,
+      "step": 9719
+    },
+    {
+      "epoch": 0.7493061979648473,
+      "grad_norm": 4.131255626678467,
+      "learning_rate": 1.55972163978238e-06,
+      "loss": 0.8972,
+      "step": 9720
+    },
+    {
+      "epoch": 0.7493832870798643,
+      "grad_norm": 3.94917893409729,
+      "learning_rate": 1.5588158014219563e-06,
+      "loss": 0.9817,
+      "step": 9721
+    },
+    {
+      "epoch": 0.7494603761948813,
+      "grad_norm": 3.44014835357666,
+      "learning_rate": 1.5579101776015443e-06,
+      "loss": 0.8567,
+      "step": 9722
+    },
+    {
+      "epoch": 0.7495374653098983,
+      "grad_norm": 3.9482109546661377,
+      "learning_rate": 1.557004768377608e-06,
+      "loss": 0.9041,
+      "step": 9723
+    },
+    {
+      "epoch": 0.7496145544249152,
+      "grad_norm": 3.8712923526763916,
+      "learning_rate": 1.556099573806591e-06,
+      "loss": 0.9819,
+      "step": 9724
+    },
+    {
+      "epoch": 0.7496916435399321,
+      "grad_norm": 3.6934661865234375,
+      "learning_rate": 1.5551945939449287e-06,
+      "loss": 0.9745,
+      "step": 9725
+    },
+    {
+      "epoch": 0.7497687326549491,
+      "grad_norm": 3.556561231613159,
+      "learning_rate": 1.5542898288490426e-06,
+      "loss": 0.9324,
+      "step": 9726
+    },
+    {
+      "epoch": 0.7498458217699661,
+      "grad_norm": 3.59233021736145,
+      "learning_rate": 1.5533852785753401e-06,
+      "loss": 0.8831,
+      "step": 9727
+    },
+    {
+      "epoch": 0.7499229108849831,
+      "grad_norm": 3.663569450378418,
+      "learning_rate": 1.5524809431802162e-06,
+      "loss": 1.0045,
+      "step": 9728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9396040439605713,
+      "learning_rate": 1.551576822720049e-06,
+      "loss": 0.9329,
+      "step": 9729
+    },
+    {
+      "epoch": 0.7500770891150169,
+      "grad_norm": 3.6608591079711914,
+      "learning_rate": 1.550672917251207e-06,
+      "loss": 0.8829,
+      "step": 9730
+    },
+    {
+      "epoch": 0.7501541782300339,
+      "grad_norm": 3.617097854614258,
+      "learning_rate": 1.5497692268300442e-06,
+      "loss": 0.9165,
+      "step": 9731
+    },
+    {
+      "epoch": 0.7502312673450509,
+      "grad_norm": 3.6227474212646484,
+      "learning_rate": 1.5488657515129001e-06,
+      "loss": 0.8429,
+      "step": 9732
+    },
+    {
+      "epoch": 0.7503083564600679,
+      "grad_norm": 3.4037904739379883,
+      "learning_rate": 1.5479624913561037e-06,
+      "loss": 0.8714,
+      "step": 9733
+    },
+    {
+      "epoch": 0.7503854455750848,
+      "grad_norm": 4.304060459136963,
+      "learning_rate": 1.5470594464159682e-06,
+      "loss": 0.9096,
+      "step": 9734
+    },
+    {
+      "epoch": 0.7504625346901017,
+      "grad_norm": 3.9946959018707275,
+      "learning_rate": 1.5461566167487918e-06,
+      "loss": 0.9868,
+      "step": 9735
+    },
+    {
+      "epoch": 0.7505396238051187,
+      "grad_norm": 3.8413825035095215,
+      "learning_rate": 1.5452540024108625e-06,
+      "loss": 0.9049,
+      "step": 9736
+    },
+    {
+      "epoch": 0.7506167129201357,
+      "grad_norm": 3.6537039279937744,
+      "learning_rate": 1.5443516034584533e-06,
+      "loss": 0.9031,
+      "step": 9737
+    },
+    {
+      "epoch": 0.7506938020351527,
+      "grad_norm": 3.7991750240325928,
+      "learning_rate": 1.5434494199478245e-06,
+      "loss": 0.8957,
+      "step": 9738
+    },
+    {
+      "epoch": 0.7507708911501696,
+      "grad_norm": 3.8140668869018555,
+      "learning_rate": 1.5425474519352234e-06,
+      "loss": 0.9202,
+      "step": 9739
+    },
+    {
+      "epoch": 0.7508479802651865,
+      "grad_norm": 3.7266762256622314,
+      "learning_rate": 1.5416456994768813e-06,
+      "loss": 0.919,
+      "step": 9740
+    },
+    {
+      "epoch": 0.7509250693802035,
+      "grad_norm": 3.538825511932373,
+      "learning_rate": 1.5407441626290181e-06,
+      "loss": 0.872,
+      "step": 9741
+    },
+    {
+      "epoch": 0.7510021584952205,
+      "grad_norm": 3.42117977142334,
+      "learning_rate": 1.5398428414478407e-06,
+      "loss": 0.7402,
+      "step": 9742
+    },
+    {
+      "epoch": 0.7510792476102375,
+      "grad_norm": 3.397956132888794,
+      "learning_rate": 1.5389417359895415e-06,
+      "loss": 0.8549,
+      "step": 9743
+    },
+    {
+      "epoch": 0.7511563367252544,
+      "grad_norm": 3.6643259525299072,
+      "learning_rate": 1.5380408463102998e-06,
+      "loss": 0.8128,
+      "step": 9744
+    },
+    {
+      "epoch": 0.7512334258402713,
+      "grad_norm": 3.688371419906616,
+      "learning_rate": 1.5371401724662826e-06,
+      "loss": 0.8901,
+      "step": 9745
+    },
+    {
+      "epoch": 0.7513105149552883,
+      "grad_norm": 3.9233813285827637,
+      "learning_rate": 1.5362397145136398e-06,
+      "loss": 0.9399,
+      "step": 9746
+    },
+    {
+      "epoch": 0.7513876040703052,
+      "grad_norm": 3.741755247116089,
+      "learning_rate": 1.5353394725085113e-06,
+      "loss": 0.8973,
+      "step": 9747
+    },
+    {
+      "epoch": 0.7514646931853223,
+      "grad_norm": 3.6673974990844727,
+      "learning_rate": 1.5344394465070234e-06,
+      "loss": 0.8966,
+      "step": 9748
+    },
+    {
+      "epoch": 0.7515417823003392,
+      "grad_norm": 3.5368854999542236,
+      "learning_rate": 1.5335396365652865e-06,
+      "loss": 0.9082,
+      "step": 9749
+    },
+    {
+      "epoch": 0.7516188714153561,
+      "grad_norm": 3.6051862239837646,
+      "learning_rate": 1.5326400427394023e-06,
+      "loss": 0.9469,
+      "step": 9750
+    },
+    {
+      "epoch": 0.7516959605303731,
+      "grad_norm": 3.7850236892700195,
+      "learning_rate": 1.5317406650854515e-06,
+      "loss": 0.8775,
+      "step": 9751
+    },
+    {
+      "epoch": 0.75177304964539,
+      "grad_norm": 3.5623207092285156,
+      "learning_rate": 1.5308415036595076e-06,
+      "loss": 0.799,
+      "step": 9752
+    },
+    {
+      "epoch": 0.7518501387604071,
+      "grad_norm": 4.15452241897583,
+      "learning_rate": 1.5299425585176292e-06,
+      "loss": 0.9851,
+      "step": 9753
+    },
+    {
+      "epoch": 0.751927227875424,
+      "grad_norm": 3.781522274017334,
+      "learning_rate": 1.52904382971586e-06,
+      "loss": 1.033,
+      "step": 9754
+    },
+    {
+      "epoch": 0.7520043169904409,
+      "grad_norm": 3.5940423011779785,
+      "learning_rate": 1.528145317310231e-06,
+      "loss": 0.9551,
+      "step": 9755
+    },
+    {
+      "epoch": 0.7520814061054579,
+      "grad_norm": 3.518636465072632,
+      "learning_rate": 1.527247021356763e-06,
+      "loss": 0.8302,
+      "step": 9756
+    },
+    {
+      "epoch": 0.7521584952204748,
+      "grad_norm": 3.90459942817688,
+      "learning_rate": 1.5263489419114552e-06,
+      "loss": 0.9807,
+      "step": 9757
+    },
+    {
+      "epoch": 0.7522355843354919,
+      "grad_norm": 3.519850492477417,
+      "learning_rate": 1.5254510790303e-06,
+      "loss": 0.9089,
+      "step": 9758
+    },
+    {
+      "epoch": 0.7523126734505088,
+      "grad_norm": 3.474034547805786,
+      "learning_rate": 1.5245534327692751e-06,
+      "loss": 0.8635,
+      "step": 9759
+    },
+    {
+      "epoch": 0.7523897625655257,
+      "grad_norm": 3.7472171783447266,
+      "learning_rate": 1.5236560031843445e-06,
+      "loss": 0.8809,
+      "step": 9760
+    },
+    {
+      "epoch": 0.7524668516805427,
+      "grad_norm": 3.900770902633667,
+      "learning_rate": 1.5227587903314568e-06,
+      "loss": 0.9003,
+      "step": 9761
+    },
+    {
+      "epoch": 0.7525439407955596,
+      "grad_norm": 3.990259885787964,
+      "learning_rate": 1.5218617942665497e-06,
+      "loss": 0.8674,
+      "step": 9762
+    },
+    {
+      "epoch": 0.7526210299105767,
+      "grad_norm": 3.8502142429351807,
+      "learning_rate": 1.5209650150455462e-06,
+      "loss": 0.8873,
+      "step": 9763
+    },
+    {
+      "epoch": 0.7526981190255936,
+      "grad_norm": 3.540637969970703,
+      "learning_rate": 1.5200684527243552e-06,
+      "loss": 0.8397,
+      "step": 9764
+    },
+    {
+      "epoch": 0.7527752081406105,
+      "grad_norm": 3.576173782348633,
+      "learning_rate": 1.5191721073588755e-06,
+      "loss": 0.9391,
+      "step": 9765
+    },
+    {
+      "epoch": 0.7528522972556275,
+      "grad_norm": 3.8144445419311523,
+      "learning_rate": 1.518275979004985e-06,
+      "loss": 0.9871,
+      "step": 9766
+    },
+    {
+      "epoch": 0.7529293863706444,
+      "grad_norm": 4.009185314178467,
+      "learning_rate": 1.517380067718555e-06,
+      "loss": 0.9879,
+      "step": 9767
+    },
+    {
+      "epoch": 0.7530064754856615,
+      "grad_norm": 3.6188137531280518,
+      "learning_rate": 1.5164843735554408e-06,
+      "loss": 0.8921,
+      "step": 9768
+    },
+    {
+      "epoch": 0.7530835646006784,
+      "grad_norm": 3.5967025756835938,
+      "learning_rate": 1.5155888965714843e-06,
+      "loss": 0.8747,
+      "step": 9769
+    },
+    {
+      "epoch": 0.7531606537156953,
+      "grad_norm": 3.705009937286377,
+      "learning_rate": 1.514693636822514e-06,
+      "loss": 0.8078,
+      "step": 9770
+    },
+    {
+      "epoch": 0.7532377428307123,
+      "grad_norm": 3.469721794128418,
+      "learning_rate": 1.5137985943643463e-06,
+      "loss": 0.8209,
+      "step": 9771
+    },
+    {
+      "epoch": 0.7533148319457292,
+      "grad_norm": 3.523388385772705,
+      "learning_rate": 1.5129037692527794e-06,
+      "loss": 0.8536,
+      "step": 9772
+    },
+    {
+      "epoch": 0.7533919210607463,
+      "grad_norm": 4.049753665924072,
+      "learning_rate": 1.5120091615436016e-06,
+      "loss": 0.8941,
+      "step": 9773
+    },
+    {
+      "epoch": 0.7534690101757632,
+      "grad_norm": 3.7952821254730225,
+      "learning_rate": 1.5111147712925884e-06,
+      "loss": 0.884,
+      "step": 9774
+    },
+    {
+      "epoch": 0.7535460992907801,
+      "grad_norm": 3.82209849357605,
+      "learning_rate": 1.5102205985554992e-06,
+      "loss": 0.7733,
+      "step": 9775
+    },
+    {
+      "epoch": 0.7536231884057971,
+      "grad_norm": 3.5352354049682617,
+      "learning_rate": 1.5093266433880837e-06,
+      "loss": 0.8167,
+      "step": 9776
+    },
+    {
+      "epoch": 0.753700277520814,
+      "grad_norm": 3.5197014808654785,
+      "learning_rate": 1.5084329058460716e-06,
+      "loss": 0.9545,
+      "step": 9777
+    },
+    {
+      "epoch": 0.7537773666358311,
+      "grad_norm": 3.3797638416290283,
+      "learning_rate": 1.5075393859851844e-06,
+      "loss": 0.7952,
+      "step": 9778
+    },
+    {
+      "epoch": 0.753854455750848,
+      "grad_norm": 3.590395927429199,
+      "learning_rate": 1.5066460838611292e-06,
+      "loss": 0.8798,
+      "step": 9779
+    },
+    {
+      "epoch": 0.7539315448658649,
+      "grad_norm": 3.962991952896118,
+      "learning_rate": 1.505752999529597e-06,
+      "loss": 0.8392,
+      "step": 9780
+    },
+    {
+      "epoch": 0.7540086339808819,
+      "grad_norm": 3.7387423515319824,
+      "learning_rate": 1.5048601330462693e-06,
+      "loss": 0.9752,
+      "step": 9781
+    },
+    {
+      "epoch": 0.7540857230958988,
+      "grad_norm": 3.6760013103485107,
+      "learning_rate": 1.5039674844668112e-06,
+      "loss": 0.8867,
+      "step": 9782
+    },
+    {
+      "epoch": 0.7541628122109159,
+      "grad_norm": 3.807915449142456,
+      "learning_rate": 1.503075053846873e-06,
+      "loss": 0.9201,
+      "step": 9783
+    },
+    {
+      "epoch": 0.7542399013259328,
+      "grad_norm": 3.836941957473755,
+      "learning_rate": 1.5021828412420942e-06,
+      "loss": 0.9706,
+      "step": 9784
+    },
+    {
+      "epoch": 0.7543169904409497,
+      "grad_norm": 3.9428610801696777,
+      "learning_rate": 1.501290846708099e-06,
+      "loss": 0.893,
+      "step": 9785
+    },
+    {
+      "epoch": 0.7543940795559667,
+      "grad_norm": 3.6269469261169434,
+      "learning_rate": 1.5003990703004994e-06,
+      "loss": 0.8229,
+      "step": 9786
+    },
+    {
+      "epoch": 0.7544711686709836,
+      "grad_norm": 3.6933364868164062,
+      "learning_rate": 1.4995075120748948e-06,
+      "loss": 0.945,
+      "step": 9787
+    },
+    {
+      "epoch": 0.7545482577860007,
+      "grad_norm": 3.697035312652588,
+      "learning_rate": 1.498616172086866e-06,
+      "loss": 0.9209,
+      "step": 9788
+    },
+    {
+      "epoch": 0.7546253469010176,
+      "grad_norm": 3.5978927612304688,
+      "learning_rate": 1.4977250503919839e-06,
+      "loss": 0.961,
+      "step": 9789
+    },
+    {
+      "epoch": 0.7547024360160345,
+      "grad_norm": 3.69553804397583,
+      "learning_rate": 1.4968341470458064e-06,
+      "loss": 0.7702,
+      "step": 9790
+    },
+    {
+      "epoch": 0.7547795251310515,
+      "grad_norm": 3.7523858547210693,
+      "learning_rate": 1.495943462103877e-06,
+      "loss": 1.0264,
+      "step": 9791
+    },
+    {
+      "epoch": 0.7548566142460684,
+      "grad_norm": 3.5115909576416016,
+      "learning_rate": 1.495052995621724e-06,
+      "loss": 0.8715,
+      "step": 9792
+    },
+    {
+      "epoch": 0.7549337033610855,
+      "grad_norm": 4.038724422454834,
+      "learning_rate": 1.4941627476548665e-06,
+      "loss": 0.9478,
+      "step": 9793
+    },
+    {
+      "epoch": 0.7550107924761024,
+      "grad_norm": 3.764338970184326,
+      "learning_rate": 1.4932727182588025e-06,
+      "loss": 0.9636,
+      "step": 9794
+    },
+    {
+      "epoch": 0.7550878815911193,
+      "grad_norm": 3.6738407611846924,
+      "learning_rate": 1.4923829074890222e-06,
+      "loss": 0.9713,
+      "step": 9795
+    },
+    {
+      "epoch": 0.7551649707061363,
+      "grad_norm": 3.553886890411377,
+      "learning_rate": 1.491493315401002e-06,
+      "loss": 0.8905,
+      "step": 9796
+    },
+    {
+      "epoch": 0.7552420598211532,
+      "grad_norm": 3.702172040939331,
+      "learning_rate": 1.4906039420502022e-06,
+      "loss": 0.906,
+      "step": 9797
+    },
+    {
+      "epoch": 0.7553191489361702,
+      "grad_norm": 3.901486396789551,
+      "learning_rate": 1.4897147874920726e-06,
+      "loss": 0.8935,
+      "step": 9798
+    },
+    {
+      "epoch": 0.7553962380511872,
+      "grad_norm": 3.898955821990967,
+      "learning_rate": 1.4888258517820442e-06,
+      "loss": 0.83,
+      "step": 9799
+    },
+    {
+      "epoch": 0.7554733271662041,
+      "grad_norm": 3.40311598777771,
+      "learning_rate": 1.4879371349755395e-06,
+      "loss": 0.79,
+      "step": 9800
+    },
+    {
+      "epoch": 0.7555504162812211,
+      "grad_norm": 3.6599347591400146,
+      "learning_rate": 1.4870486371279647e-06,
+      "loss": 0.9212,
+      "step": 9801
+    },
+    {
+      "epoch": 0.755627505396238,
+      "grad_norm": 3.6006016731262207,
+      "learning_rate": 1.486160358294713e-06,
+      "loss": 0.914,
+      "step": 9802
+    },
+    {
+      "epoch": 0.755704594511255,
+      "grad_norm": 3.814122438430786,
+      "learning_rate": 1.4852722985311647e-06,
+      "loss": 0.965,
+      "step": 9803
+    },
+    {
+      "epoch": 0.755781683626272,
+      "grad_norm": 3.5150370597839355,
+      "learning_rate": 1.4843844578926863e-06,
+      "loss": 0.8573,
+      "step": 9804
+    },
+    {
+      "epoch": 0.7558587727412889,
+      "grad_norm": 3.7043817043304443,
+      "learning_rate": 1.483496836434627e-06,
+      "loss": 0.9301,
+      "step": 9805
+    },
+    {
+      "epoch": 0.7559358618563059,
+      "grad_norm": 3.6749911308288574,
+      "learning_rate": 1.4826094342123282e-06,
+      "loss": 0.8818,
+      "step": 9806
+    },
+    {
+      "epoch": 0.7560129509713228,
+      "grad_norm": 4.301516532897949,
+      "learning_rate": 1.4817222512811146e-06,
+      "loss": 0.9437,
+      "step": 9807
+    },
+    {
+      "epoch": 0.7560900400863398,
+      "grad_norm": 4.242541313171387,
+      "learning_rate": 1.4808352876962984e-06,
+      "loss": 0.9117,
+      "step": 9808
+    },
+    {
+      "epoch": 0.7561671292013568,
+      "grad_norm": 3.7353222370147705,
+      "learning_rate": 1.4799485435131745e-06,
+      "loss": 0.8453,
+      "step": 9809
+    },
+    {
+      "epoch": 0.7562442183163737,
+      "grad_norm": 3.8031511306762695,
+      "learning_rate": 1.4790620187870275e-06,
+      "loss": 0.9015,
+      "step": 9810
+    },
+    {
+      "epoch": 0.7563213074313907,
+      "grad_norm": 4.096917629241943,
+      "learning_rate": 1.478175713573129e-06,
+      "loss": 0.951,
+      "step": 9811
+    },
+    {
+      "epoch": 0.7563983965464076,
+      "grad_norm": 3.349189043045044,
+      "learning_rate": 1.477289627926734e-06,
+      "loss": 0.8813,
+      "step": 9812
+    },
+    {
+      "epoch": 0.7564754856614246,
+      "grad_norm": 3.9046249389648438,
+      "learning_rate": 1.476403761903088e-06,
+      "loss": 0.8709,
+      "step": 9813
+    },
+    {
+      "epoch": 0.7565525747764416,
+      "grad_norm": 4.114388465881348,
+      "learning_rate": 1.4755181155574166e-06,
+      "loss": 0.9669,
+      "step": 9814
+    },
+    {
+      "epoch": 0.7566296638914586,
+      "grad_norm": 3.496314287185669,
+      "learning_rate": 1.4746326889449375e-06,
+      "loss": 0.9647,
+      "step": 9815
+    },
+    {
+      "epoch": 0.7567067530064755,
+      "grad_norm": 3.8805534839630127,
+      "learning_rate": 1.4737474821208513e-06,
+      "loss": 0.8806,
+      "step": 9816
+    },
+    {
+      "epoch": 0.7567838421214924,
+      "grad_norm": 3.560910224914551,
+      "learning_rate": 1.472862495140347e-06,
+      "loss": 0.7968,
+      "step": 9817
+    },
+    {
+      "epoch": 0.7568609312365094,
+      "grad_norm": 3.6736807823181152,
+      "learning_rate": 1.4719777280585983e-06,
+      "loss": 0.8541,
+      "step": 9818
+    },
+    {
+      "epoch": 0.7569380203515264,
+      "grad_norm": 4.000894069671631,
+      "learning_rate": 1.4710931809307677e-06,
+      "loss": 0.9703,
+      "step": 9819
+    },
+    {
+      "epoch": 0.7570151094665434,
+      "grad_norm": 4.015716075897217,
+      "learning_rate": 1.4702088538119996e-06,
+      "loss": 1.0352,
+      "step": 9820
+    },
+    {
+      "epoch": 0.7570921985815603,
+      "grad_norm": 3.6512205600738525,
+      "learning_rate": 1.4693247467574273e-06,
+      "loss": 0.9386,
+      "step": 9821
+    },
+    {
+      "epoch": 0.7571692876965772,
+      "grad_norm": 3.792320728302002,
+      "learning_rate": 1.4684408598221722e-06,
+      "loss": 0.9188,
+      "step": 9822
+    },
+    {
+      "epoch": 0.7572463768115942,
+      "grad_norm": 3.727431297302246,
+      "learning_rate": 1.4675571930613385e-06,
+      "loss": 0.8135,
+      "step": 9823
+    },
+    {
+      "epoch": 0.7573234659266112,
+      "grad_norm": 3.9126195907592773,
+      "learning_rate": 1.4666737465300202e-06,
+      "loss": 0.958,
+      "step": 9824
+    },
+    {
+      "epoch": 0.7574005550416282,
+      "grad_norm": 3.749931573867798,
+      "learning_rate": 1.4657905202832928e-06,
+      "loss": 0.8214,
+      "step": 9825
+    },
+    {
+      "epoch": 0.7574776441566451,
+      "grad_norm": 3.8023717403411865,
+      "learning_rate": 1.4649075143762225e-06,
+      "loss": 0.9376,
+      "step": 9826
+    },
+    {
+      "epoch": 0.757554733271662,
+      "grad_norm": 3.8662075996398926,
+      "learning_rate": 1.4640247288638603e-06,
+      "loss": 0.9735,
+      "step": 9827
+    },
+    {
+      "epoch": 0.757631822386679,
+      "grad_norm": 3.9173572063446045,
+      "learning_rate": 1.463142163801242e-06,
+      "loss": 0.9421,
+      "step": 9828
+    },
+    {
+      "epoch": 0.757708911501696,
+      "grad_norm": 3.737765312194824,
+      "learning_rate": 1.4622598192433928e-06,
+      "loss": 0.9159,
+      "step": 9829
+    },
+    {
+      "epoch": 0.757786000616713,
+      "grad_norm": 3.5803210735321045,
+      "learning_rate": 1.461377695245323e-06,
+      "loss": 0.8864,
+      "step": 9830
+    },
+    {
+      "epoch": 0.7578630897317299,
+      "grad_norm": 3.9155960083007812,
+      "learning_rate": 1.460495791862025e-06,
+      "loss": 0.9516,
+      "step": 9831
+    },
+    {
+      "epoch": 0.7579401788467468,
+      "grad_norm": 3.639129161834717,
+      "learning_rate": 1.4596141091484828e-06,
+      "loss": 0.8667,
+      "step": 9832
+    },
+    {
+      "epoch": 0.7580172679617638,
+      "grad_norm": 3.9074134826660156,
+      "learning_rate": 1.4587326471596647e-06,
+      "loss": 0.9008,
+      "step": 9833
+    },
+    {
+      "epoch": 0.7580943570767807,
+      "grad_norm": 3.6987130641937256,
+      "learning_rate": 1.4578514059505256e-06,
+      "loss": 0.9457,
+      "step": 9834
+    },
+    {
+      "epoch": 0.7581714461917978,
+      "grad_norm": 3.8588478565216064,
+      "learning_rate": 1.4569703855760076e-06,
+      "loss": 0.9084,
+      "step": 9835
+    },
+    {
+      "epoch": 0.7582485353068147,
+      "grad_norm": 3.9179983139038086,
+      "learning_rate": 1.4560895860910345e-06,
+      "loss": 0.9484,
+      "step": 9836
+    },
+    {
+      "epoch": 0.7583256244218316,
+      "grad_norm": 3.7824649810791016,
+      "learning_rate": 1.4552090075505215e-06,
+      "loss": 0.945,
+      "step": 9837
+    },
+    {
+      "epoch": 0.7584027135368486,
+      "grad_norm": 3.568527936935425,
+      "learning_rate": 1.454328650009368e-06,
+      "loss": 0.9535,
+      "step": 9838
+    },
+    {
+      "epoch": 0.7584798026518655,
+      "grad_norm": 3.9821395874023438,
+      "learning_rate": 1.4534485135224597e-06,
+      "loss": 0.8473,
+      "step": 9839
+    },
+    {
+      "epoch": 0.7585568917668826,
+      "grad_norm": 3.9586429595947266,
+      "learning_rate": 1.452568598144668e-06,
+      "loss": 0.9471,
+      "step": 9840
+    },
+    {
+      "epoch": 0.7586339808818995,
+      "grad_norm": 4.391334533691406,
+      "learning_rate": 1.4516889039308535e-06,
+      "loss": 1.0125,
+      "step": 9841
+    },
+    {
+      "epoch": 0.7587110699969164,
+      "grad_norm": 3.699589252471924,
+      "learning_rate": 1.4508094309358573e-06,
+      "loss": 0.8814,
+      "step": 9842
+    },
+    {
+      "epoch": 0.7587881591119334,
+      "grad_norm": 4.033127784729004,
+      "learning_rate": 1.4499301792145109e-06,
+      "loss": 0.9453,
+      "step": 9843
+    },
+    {
+      "epoch": 0.7588652482269503,
+      "grad_norm": 3.849324941635132,
+      "learning_rate": 1.449051148821632e-06,
+      "loss": 0.8697,
+      "step": 9844
+    },
+    {
+      "epoch": 0.7589423373419674,
+      "grad_norm": 3.5151302814483643,
+      "learning_rate": 1.4481723398120228e-06,
+      "loss": 0.9318,
+      "step": 9845
+    },
+    {
+      "epoch": 0.7590194264569843,
+      "grad_norm": 3.9107677936553955,
+      "learning_rate": 1.4472937522404744e-06,
+      "loss": 0.9946,
+      "step": 9846
+    },
+    {
+      "epoch": 0.7590965155720012,
+      "grad_norm": 3.9752719402313232,
+      "learning_rate": 1.4464153861617597e-06,
+      "loss": 0.8786,
+      "step": 9847
+    },
+    {
+      "epoch": 0.7591736046870182,
+      "grad_norm": 3.470489978790283,
+      "learning_rate": 1.4455372416306407e-06,
+      "loss": 0.8099,
+      "step": 9848
+    },
+    {
+      "epoch": 0.7592506938020351,
+      "grad_norm": 4.009558200836182,
+      "learning_rate": 1.4446593187018637e-06,
+      "loss": 0.9382,
+      "step": 9849
+    },
+    {
+      "epoch": 0.7593277829170522,
+      "grad_norm": 4.106441020965576,
+      "learning_rate": 1.4437816174301684e-06,
+      "loss": 0.9847,
+      "step": 9850
+    },
+    {
+      "epoch": 0.7594048720320691,
+      "grad_norm": 3.834660768508911,
+      "learning_rate": 1.442904137870269e-06,
+      "loss": 0.8793,
+      "step": 9851
+    },
+    {
+      "epoch": 0.759481961147086,
+      "grad_norm": 3.583991050720215,
+      "learning_rate": 1.4420268800768744e-06,
+      "loss": 0.8814,
+      "step": 9852
+    },
+    {
+      "epoch": 0.759559050262103,
+      "grad_norm": 3.8111629486083984,
+      "learning_rate": 1.4411498441046761e-06,
+      "loss": 0.9838,
+      "step": 9853
+    },
+    {
+      "epoch": 0.7596361393771199,
+      "grad_norm": 3.4147655963897705,
+      "learning_rate": 1.4402730300083534e-06,
+      "loss": 0.8173,
+      "step": 9854
+    },
+    {
+      "epoch": 0.759713228492137,
+      "grad_norm": 3.8176896572113037,
+      "learning_rate": 1.4393964378425712e-06,
+      "loss": 0.8702,
+      "step": 9855
+    },
+    {
+      "epoch": 0.7597903176071539,
+      "grad_norm": 3.6424262523651123,
+      "learning_rate": 1.438520067661982e-06,
+      "loss": 0.7725,
+      "step": 9856
+    },
+    {
+      "epoch": 0.7598674067221708,
+      "grad_norm": 3.9817745685577393,
+      "learning_rate": 1.4376439195212194e-06,
+      "loss": 0.9865,
+      "step": 9857
+    },
+    {
+      "epoch": 0.7599444958371878,
+      "grad_norm": 3.7620887756347656,
+      "learning_rate": 1.4367679934749085e-06,
+      "loss": 1.0232,
+      "step": 9858
+    },
+    {
+      "epoch": 0.7600215849522047,
+      "grad_norm": 3.7138636112213135,
+      "learning_rate": 1.4358922895776584e-06,
+      "loss": 0.9584,
+      "step": 9859
+    },
+    {
+      "epoch": 0.7600986740672218,
+      "grad_norm": 3.80833101272583,
+      "learning_rate": 1.4350168078840653e-06,
+      "loss": 0.9049,
+      "step": 9860
+    },
+    {
+      "epoch": 0.7601757631822387,
+      "grad_norm": 3.6317269802093506,
+      "learning_rate": 1.4341415484487126e-06,
+      "loss": 0.8555,
+      "step": 9861
+    },
+    {
+      "epoch": 0.7602528522972556,
+      "grad_norm": 3.839651107788086,
+      "learning_rate": 1.4332665113261645e-06,
+      "loss": 0.8839,
+      "step": 9862
+    },
+    {
+      "epoch": 0.7603299414122726,
+      "grad_norm": 4.206575870513916,
+      "learning_rate": 1.4323916965709766e-06,
+      "loss": 0.9231,
+      "step": 9863
+    },
+    {
+      "epoch": 0.7604070305272895,
+      "grad_norm": 3.6243884563446045,
+      "learning_rate": 1.4315171042376897e-06,
+      "loss": 0.953,
+      "step": 9864
+    },
+    {
+      "epoch": 0.7604841196423066,
+      "grad_norm": 3.7419850826263428,
+      "learning_rate": 1.43064273438083e-06,
+      "loss": 1.0172,
+      "step": 9865
+    },
+    {
+      "epoch": 0.7605612087573235,
+      "grad_norm": 3.832564353942871,
+      "learning_rate": 1.4297685870549088e-06,
+      "loss": 0.8665,
+      "step": 9866
+    },
+    {
+      "epoch": 0.7606382978723404,
+      "grad_norm": 3.582911491394043,
+      "learning_rate": 1.4288946623144272e-06,
+      "loss": 0.8848,
+      "step": 9867
+    },
+    {
+      "epoch": 0.7607153869873574,
+      "grad_norm": 3.7686104774475098,
+      "learning_rate": 1.4280209602138673e-06,
+      "loss": 0.932,
+      "step": 9868
+    },
+    {
+      "epoch": 0.7607924761023743,
+      "grad_norm": 3.6485867500305176,
+      "learning_rate": 1.4271474808077e-06,
+      "loss": 0.9416,
+      "step": 9869
+    },
+    {
+      "epoch": 0.7608695652173914,
+      "grad_norm": 3.6767213344573975,
+      "learning_rate": 1.4262742241503836e-06,
+      "loss": 0.9576,
+      "step": 9870
+    },
+    {
+      "epoch": 0.7609466543324083,
+      "grad_norm": 3.4594292640686035,
+      "learning_rate": 1.42540119029636e-06,
+      "loss": 0.8465,
+      "step": 9871
+    },
+    {
+      "epoch": 0.7610237434474252,
+      "grad_norm": 3.935392141342163,
+      "learning_rate": 1.4245283793000608e-06,
+      "loss": 0.9516,
+      "step": 9872
+    },
+    {
+      "epoch": 0.7611008325624422,
+      "grad_norm": 3.8447792530059814,
+      "learning_rate": 1.4236557912158977e-06,
+      "loss": 0.8472,
+      "step": 9873
+    },
+    {
+      "epoch": 0.7611779216774591,
+      "grad_norm": 3.4628243446350098,
+      "learning_rate": 1.4227834260982732e-06,
+      "loss": 0.861,
+      "step": 9874
+    },
+    {
+      "epoch": 0.7612550107924761,
+      "grad_norm": 3.9336140155792236,
+      "learning_rate": 1.4219112840015759e-06,
+      "loss": 0.9094,
+      "step": 9875
+    },
+    {
+      "epoch": 0.7613320999074931,
+      "grad_norm": 4.2745041847229,
+      "learning_rate": 1.421039364980178e-06,
+      "loss": 0.9124,
+      "step": 9876
+    },
+    {
+      "epoch": 0.76140918902251,
+      "grad_norm": 4.0710625648498535,
+      "learning_rate": 1.4201676690884403e-06,
+      "loss": 0.7956,
+      "step": 9877
+    },
+    {
+      "epoch": 0.761486278137527,
+      "grad_norm": 3.9371089935302734,
+      "learning_rate": 1.4192961963807094e-06,
+      "loss": 0.9185,
+      "step": 9878
+    },
+    {
+      "epoch": 0.7615633672525439,
+      "grad_norm": 3.549879789352417,
+      "learning_rate": 1.4184249469113138e-06,
+      "loss": 0.8869,
+      "step": 9879
+    },
+    {
+      "epoch": 0.761640456367561,
+      "grad_norm": 3.896291494369507,
+      "learning_rate": 1.417553920734574e-06,
+      "loss": 0.8783,
+      "step": 9880
+    },
+    {
+      "epoch": 0.7617175454825779,
+      "grad_norm": 3.623753070831299,
+      "learning_rate": 1.4166831179047923e-06,
+      "loss": 0.8513,
+      "step": 9881
+    },
+    {
+      "epoch": 0.7617946345975948,
+      "grad_norm": 3.9215145111083984,
+      "learning_rate": 1.4158125384762606e-06,
+      "loss": 0.8208,
+      "step": 9882
+    },
+    {
+      "epoch": 0.7618717237126118,
+      "grad_norm": 3.590507745742798,
+      "learning_rate": 1.4149421825032556e-06,
+      "loss": 0.918,
+      "step": 9883
+    },
+    {
+      "epoch": 0.7619488128276287,
+      "grad_norm": 4.319372177124023,
+      "learning_rate": 1.4140720500400363e-06,
+      "loss": 0.9689,
+      "step": 9884
+    },
+    {
+      "epoch": 0.7620259019426457,
+      "grad_norm": 3.8288538455963135,
+      "learning_rate": 1.4132021411408526e-06,
+      "loss": 0.9748,
+      "step": 9885
+    },
+    {
+      "epoch": 0.7621029910576627,
+      "grad_norm": 4.058238506317139,
+      "learning_rate": 1.4123324558599389e-06,
+      "loss": 1.0233,
+      "step": 9886
+    },
+    {
+      "epoch": 0.7621800801726796,
+      "grad_norm": 3.779918670654297,
+      "learning_rate": 1.4114629942515156e-06,
+      "loss": 0.8905,
+      "step": 9887
+    },
+    {
+      "epoch": 0.7622571692876966,
+      "grad_norm": 3.943232774734497,
+      "learning_rate": 1.410593756369789e-06,
+      "loss": 0.8821,
+      "step": 9888
+    },
+    {
+      "epoch": 0.7623342584027135,
+      "grad_norm": 3.8813998699188232,
+      "learning_rate": 1.4097247422689537e-06,
+      "loss": 0.9586,
+      "step": 9889
+    },
+    {
+      "epoch": 0.7624113475177305,
+      "grad_norm": 4.069021224975586,
+      "learning_rate": 1.408855952003184e-06,
+      "loss": 0.9016,
+      "step": 9890
+    },
+    {
+      "epoch": 0.7624884366327475,
+      "grad_norm": 3.904691457748413,
+      "learning_rate": 1.4079873856266468e-06,
+      "loss": 0.8317,
+      "step": 9891
+    },
+    {
+      "epoch": 0.7625655257477644,
+      "grad_norm": 3.729013681411743,
+      "learning_rate": 1.4071190431934934e-06,
+      "loss": 1.0032,
+      "step": 9892
+    },
+    {
+      "epoch": 0.7626426148627814,
+      "grad_norm": 3.8653934001922607,
+      "learning_rate": 1.4062509247578586e-06,
+      "loss": 0.9653,
+      "step": 9893
+    },
+    {
+      "epoch": 0.7627197039777983,
+      "grad_norm": 3.8940982818603516,
+      "learning_rate": 1.405383030373867e-06,
+      "loss": 1.0486,
+      "step": 9894
+    },
+    {
+      "epoch": 0.7627967930928153,
+      "grad_norm": 3.855316162109375,
+      "learning_rate": 1.4045153600956257e-06,
+      "loss": 1.0069,
+      "step": 9895
+    },
+    {
+      "epoch": 0.7628738822078323,
+      "grad_norm": 4.0627827644348145,
+      "learning_rate": 1.4036479139772309e-06,
+      "loss": 0.9452,
+      "step": 9896
+    },
+    {
+      "epoch": 0.7629509713228492,
+      "grad_norm": 3.806206464767456,
+      "learning_rate": 1.402780692072762e-06,
+      "loss": 0.8991,
+      "step": 9897
+    },
+    {
+      "epoch": 0.7630280604378662,
+      "grad_norm": 4.070652484893799,
+      "learning_rate": 1.4019136944362882e-06,
+      "loss": 0.9424,
+      "step": 9898
+    },
+    {
+      "epoch": 0.7631051495528831,
+      "grad_norm": 3.583237409591675,
+      "learning_rate": 1.401046921121859e-06,
+      "loss": 0.9428,
+      "step": 9899
+    },
+    {
+      "epoch": 0.7631822386679001,
+      "grad_norm": 3.9483799934387207,
+      "learning_rate": 1.400180372183515e-06,
+      "loss": 1.017,
+      "step": 9900
+    },
+    {
+      "epoch": 0.763259327782917,
+      "grad_norm": 3.897564172744751,
+      "learning_rate": 1.3993140476752808e-06,
+      "loss": 0.8972,
+      "step": 9901
+    },
+    {
+      "epoch": 0.763336416897934,
+      "grad_norm": 3.622753620147705,
+      "learning_rate": 1.3984479476511676e-06,
+      "loss": 0.925,
+      "step": 9902
+    },
+    {
+      "epoch": 0.763413506012951,
+      "grad_norm": 3.872459888458252,
+      "learning_rate": 1.3975820721651718e-06,
+      "loss": 0.9007,
+      "step": 9903
+    },
+    {
+      "epoch": 0.7634905951279679,
+      "grad_norm": 3.8231959342956543,
+      "learning_rate": 1.3967164212712774e-06,
+      "loss": 0.8777,
+      "step": 9904
+    },
+    {
+      "epoch": 0.7635676842429849,
+      "grad_norm": 3.6052405834198,
+      "learning_rate": 1.3958509950234516e-06,
+      "loss": 0.926,
+      "step": 9905
+    },
+    {
+      "epoch": 0.7636447733580018,
+      "grad_norm": 4.091972351074219,
+      "learning_rate": 1.3949857934756495e-06,
+      "loss": 0.9525,
+      "step": 9906
+    },
+    {
+      "epoch": 0.7637218624730188,
+      "grad_norm": 3.3216168880462646,
+      "learning_rate": 1.394120816681812e-06,
+      "loss": 0.8179,
+      "step": 9907
+    },
+    {
+      "epoch": 0.7637989515880358,
+      "grad_norm": 3.8997809886932373,
+      "learning_rate": 1.3932560646958665e-06,
+      "loss": 0.9877,
+      "step": 9908
+    },
+    {
+      "epoch": 0.7638760407030527,
+      "grad_norm": 3.5005760192871094,
+      "learning_rate": 1.3923915375717272e-06,
+      "loss": 0.9585,
+      "step": 9909
+    },
+    {
+      "epoch": 0.7639531298180697,
+      "grad_norm": 3.4733173847198486,
+      "learning_rate": 1.3915272353632896e-06,
+      "loss": 0.873,
+      "step": 9910
+    },
+    {
+      "epoch": 0.7640302189330866,
+      "grad_norm": 3.862825393676758,
+      "learning_rate": 1.39066315812444e-06,
+      "loss": 0.8944,
+      "step": 9911
+    },
+    {
+      "epoch": 0.7641073080481036,
+      "grad_norm": 3.6997947692871094,
+      "learning_rate": 1.3897993059090492e-06,
+      "loss": 0.9553,
+      "step": 9912
+    },
+    {
+      "epoch": 0.7641843971631206,
+      "grad_norm": 4.415987491607666,
+      "learning_rate": 1.388935678770974e-06,
+      "loss": 0.8898,
+      "step": 9913
+    },
+    {
+      "epoch": 0.7642614862781375,
+      "grad_norm": 3.8351826667785645,
+      "learning_rate": 1.3880722767640575e-06,
+      "loss": 1.0199,
+      "step": 9914
+    },
+    {
+      "epoch": 0.7643385753931545,
+      "grad_norm": 3.8005855083465576,
+      "learning_rate": 1.387209099942129e-06,
+      "loss": 0.8509,
+      "step": 9915
+    },
+    {
+      "epoch": 0.7644156645081714,
+      "grad_norm": 3.49656343460083,
+      "learning_rate": 1.3863461483590008e-06,
+      "loss": 0.8576,
+      "step": 9916
+    },
+    {
+      "epoch": 0.7644927536231884,
+      "grad_norm": 3.7017486095428467,
+      "learning_rate": 1.3854834220684743e-06,
+      "loss": 0.976,
+      "step": 9917
+    },
+    {
+      "epoch": 0.7645698427382054,
+      "grad_norm": 3.82146954536438,
+      "learning_rate": 1.3846209211243366e-06,
+      "loss": 0.8766,
+      "step": 9918
+    },
+    {
+      "epoch": 0.7646469318532223,
+      "grad_norm": 3.6849117279052734,
+      "learning_rate": 1.3837586455803599e-06,
+      "loss": 0.9259,
+      "step": 9919
+    },
+    {
+      "epoch": 0.7647240209682393,
+      "grad_norm": 3.7132599353790283,
+      "learning_rate": 1.382896595490304e-06,
+      "loss": 0.853,
+      "step": 9920
+    },
+    {
+      "epoch": 0.7648011100832562,
+      "grad_norm": 4.089300632476807,
+      "learning_rate": 1.3820347709079103e-06,
+      "loss": 0.9161,
+      "step": 9921
+    },
+    {
+      "epoch": 0.7648781991982732,
+      "grad_norm": 3.6763312816619873,
+      "learning_rate": 1.3811731718869108e-06,
+      "loss": 0.9985,
+      "step": 9922
+    },
+    {
+      "epoch": 0.7649552883132902,
+      "grad_norm": 3.901153326034546,
+      "learning_rate": 1.3803117984810221e-06,
+      "loss": 0.9706,
+      "step": 9923
+    },
+    {
+      "epoch": 0.7650323774283071,
+      "grad_norm": 3.6680846214294434,
+      "learning_rate": 1.3794506507439454e-06,
+      "loss": 0.9672,
+      "step": 9924
+    },
+    {
+      "epoch": 0.7651094665433241,
+      "grad_norm": 3.704328775405884,
+      "learning_rate": 1.3785897287293693e-06,
+      "loss": 0.9395,
+      "step": 9925
+    },
+    {
+      "epoch": 0.765186555658341,
+      "grad_norm": 3.6774888038635254,
+      "learning_rate": 1.3777290324909698e-06,
+      "loss": 1.0224,
+      "step": 9926
+    },
+    {
+      "epoch": 0.765263644773358,
+      "grad_norm": 3.7571513652801514,
+      "learning_rate": 1.376868562082403e-06,
+      "loss": 1.0518,
+      "step": 9927
+    },
+    {
+      "epoch": 0.765340733888375,
+      "grad_norm": 3.6580474376678467,
+      "learning_rate": 1.3760083175573168e-06,
+      "loss": 0.8153,
+      "step": 9928
+    },
+    {
+      "epoch": 0.7654178230033919,
+      "grad_norm": 4.089507579803467,
+      "learning_rate": 1.3751482989693433e-06,
+      "loss": 0.9439,
+      "step": 9929
+    },
+    {
+      "epoch": 0.7654949121184089,
+      "grad_norm": 3.566601037979126,
+      "learning_rate": 1.374288506372099e-06,
+      "loss": 0.8285,
+      "step": 9930
+    },
+    {
+      "epoch": 0.7655720012334258,
+      "grad_norm": 3.4016520977020264,
+      "learning_rate": 1.3734289398191902e-06,
+      "loss": 0.8906,
+      "step": 9931
+    },
+    {
+      "epoch": 0.7656490903484428,
+      "grad_norm": 3.8981709480285645,
+      "learning_rate": 1.372569599364203e-06,
+      "loss": 1.0273,
+      "step": 9932
+    },
+    {
+      "epoch": 0.7657261794634598,
+      "grad_norm": 4.178238868713379,
+      "learning_rate": 1.3717104850607144e-06,
+      "loss": 1.0948,
+      "step": 9933
+    },
+    {
+      "epoch": 0.7658032685784767,
+      "grad_norm": 3.904507875442505,
+      "learning_rate": 1.3708515969622854e-06,
+      "loss": 1.0076,
+      "step": 9934
+    },
+    {
+      "epoch": 0.7658803576934937,
+      "grad_norm": 3.62880802154541,
+      "learning_rate": 1.369992935122464e-06,
+      "loss": 0.9251,
+      "step": 9935
+    },
+    {
+      "epoch": 0.7659574468085106,
+      "grad_norm": 3.961493968963623,
+      "learning_rate": 1.369134499594782e-06,
+      "loss": 0.9443,
+      "step": 9936
+    },
+    {
+      "epoch": 0.7660345359235275,
+      "grad_norm": 3.5568950176239014,
+      "learning_rate": 1.3682762904327613e-06,
+      "loss": 0.9349,
+      "step": 9937
+    },
+    {
+      "epoch": 0.7661116250385446,
+      "grad_norm": 4.020711898803711,
+      "learning_rate": 1.3674183076899016e-06,
+      "loss": 0.8839,
+      "step": 9938
+    },
+    {
+      "epoch": 0.7661887141535615,
+      "grad_norm": 4.319362163543701,
+      "learning_rate": 1.3665605514196984e-06,
+      "loss": 1.0175,
+      "step": 9939
+    },
+    {
+      "epoch": 0.7662658032685785,
+      "grad_norm": 3.667689085006714,
+      "learning_rate": 1.3657030216756263e-06,
+      "loss": 0.8792,
+      "step": 9940
+    },
+    {
+      "epoch": 0.7663428923835954,
+      "grad_norm": 3.632988214492798,
+      "learning_rate": 1.3648457185111502e-06,
+      "loss": 0.8438,
+      "step": 9941
+    },
+    {
+      "epoch": 0.7664199814986123,
+      "grad_norm": 3.730689525604248,
+      "learning_rate": 1.363988641979715e-06,
+      "loss": 0.9797,
+      "step": 9942
+    },
+    {
+      "epoch": 0.7664970706136294,
+      "grad_norm": 3.5583460330963135,
+      "learning_rate": 1.3631317921347564e-06,
+      "loss": 0.7986,
+      "step": 9943
+    },
+    {
+      "epoch": 0.7665741597286463,
+      "grad_norm": 3.7527616024017334,
+      "learning_rate": 1.3622751690296947e-06,
+      "loss": 0.9162,
+      "step": 9944
+    },
+    {
+      "epoch": 0.7666512488436633,
+      "grad_norm": 3.690645694732666,
+      "learning_rate": 1.3614187727179368e-06,
+      "loss": 0.9612,
+      "step": 9945
+    },
+    {
+      "epoch": 0.7667283379586802,
+      "grad_norm": 3.8504841327667236,
+      "learning_rate": 1.3605626032528746e-06,
+      "loss": 0.8196,
+      "step": 9946
+    },
+    {
+      "epoch": 0.7668054270736971,
+      "grad_norm": 3.3687868118286133,
+      "learning_rate": 1.3597066606878834e-06,
+      "loss": 0.8102,
+      "step": 9947
+    },
+    {
+      "epoch": 0.7668825161887142,
+      "grad_norm": 3.635887622833252,
+      "learning_rate": 1.3588509450763281e-06,
+      "loss": 0.8814,
+      "step": 9948
+    },
+    {
+      "epoch": 0.7669596053037311,
+      "grad_norm": 3.819847822189331,
+      "learning_rate": 1.3579954564715587e-06,
+      "loss": 0.8737,
+      "step": 9949
+    },
+    {
+      "epoch": 0.7670366944187481,
+      "grad_norm": 3.8392279148101807,
+      "learning_rate": 1.3571401949269103e-06,
+      "loss": 0.9532,
+      "step": 9950
+    },
+    {
+      "epoch": 0.767113783533765,
+      "grad_norm": 3.4588663578033447,
+      "learning_rate": 1.356285160495704e-06,
+      "loss": 0.7837,
+      "step": 9951
+    },
+    {
+      "epoch": 0.7671908726487819,
+      "grad_norm": 3.9829211235046387,
+      "learning_rate": 1.3554303532312475e-06,
+      "loss": 0.9798,
+      "step": 9952
+    },
+    {
+      "epoch": 0.767267961763799,
+      "grad_norm": 3.6021299362182617,
+      "learning_rate": 1.354575773186832e-06,
+      "loss": 0.8786,
+      "step": 9953
+    },
+    {
+      "epoch": 0.7673450508788159,
+      "grad_norm": 3.810025453567505,
+      "learning_rate": 1.353721420415736e-06,
+      "loss": 0.9356,
+      "step": 9954
+    },
+    {
+      "epoch": 0.7674221399938329,
+      "grad_norm": 3.6363253593444824,
+      "learning_rate": 1.3528672949712257e-06,
+      "loss": 0.9352,
+      "step": 9955
+    },
+    {
+      "epoch": 0.7674992291088498,
+      "grad_norm": 3.648545980453491,
+      "learning_rate": 1.3520133969065502e-06,
+      "loss": 0.8348,
+      "step": 9956
+    },
+    {
+      "epoch": 0.7675763182238667,
+      "grad_norm": 3.741356611251831,
+      "learning_rate": 1.3511597262749476e-06,
+      "loss": 0.8785,
+      "step": 9957
+    },
+    {
+      "epoch": 0.7676534073388838,
+      "grad_norm": 3.7994656562805176,
+      "learning_rate": 1.3503062831296372e-06,
+      "loss": 0.9241,
+      "step": 9958
+    },
+    {
+      "epoch": 0.7677304964539007,
+      "grad_norm": 3.677051544189453,
+      "learning_rate": 1.349453067523827e-06,
+      "loss": 0.9047,
+      "step": 9959
+    },
+    {
+      "epoch": 0.7678075855689177,
+      "grad_norm": 3.7097387313842773,
+      "learning_rate": 1.3486000795107118e-06,
+      "loss": 0.9048,
+      "step": 9960
+    },
+    {
+      "epoch": 0.7678846746839346,
+      "grad_norm": 3.4744412899017334,
+      "learning_rate": 1.3477473191434703e-06,
+      "loss": 0.939,
+      "step": 9961
+    },
+    {
+      "epoch": 0.7679617637989515,
+      "grad_norm": 4.095146179199219,
+      "learning_rate": 1.346894786475268e-06,
+      "loss": 1.0014,
+      "step": 9962
+    },
+    {
+      "epoch": 0.7680388529139686,
+      "grad_norm": 3.7162740230560303,
+      "learning_rate": 1.3460424815592577e-06,
+      "loss": 0.8576,
+      "step": 9963
+    },
+    {
+      "epoch": 0.7681159420289855,
+      "grad_norm": 3.45137095451355,
+      "learning_rate": 1.3451904044485725e-06,
+      "loss": 0.8469,
+      "step": 9964
+    },
+    {
+      "epoch": 0.7681930311440025,
+      "grad_norm": 3.4185662269592285,
+      "learning_rate": 1.3443385551963373e-06,
+      "loss": 0.741,
+      "step": 9965
+    },
+    {
+      "epoch": 0.7682701202590194,
+      "grad_norm": 3.571277618408203,
+      "learning_rate": 1.3434869338556594e-06,
+      "loss": 0.9274,
+      "step": 9966
+    },
+    {
+      "epoch": 0.7683472093740363,
+      "grad_norm": 3.5546271800994873,
+      "learning_rate": 1.3426355404796337e-06,
+      "loss": 0.8564,
+      "step": 9967
+    },
+    {
+      "epoch": 0.7684242984890534,
+      "grad_norm": 3.800027370452881,
+      "learning_rate": 1.341784375121342e-06,
+      "loss": 0.9594,
+      "step": 9968
+    },
+    {
+      "epoch": 0.7685013876040703,
+      "grad_norm": 3.928795337677002,
+      "learning_rate": 1.3409334378338461e-06,
+      "loss": 0.9045,
+      "step": 9969
+    },
+    {
+      "epoch": 0.7685784767190873,
+      "grad_norm": 3.6231064796447754,
+      "learning_rate": 1.3400827286702001e-06,
+      "loss": 0.9071,
+      "step": 9970
+    },
+    {
+      "epoch": 0.7686555658341042,
+      "grad_norm": 3.7267343997955322,
+      "learning_rate": 1.3392322476834402e-06,
+      "loss": 0.9177,
+      "step": 9971
+    },
+    {
+      "epoch": 0.7687326549491211,
+      "grad_norm": 3.9702486991882324,
+      "learning_rate": 1.3383819949265908e-06,
+      "loss": 0.9202,
+      "step": 9972
+    },
+    {
+      "epoch": 0.7688097440641382,
+      "grad_norm": 3.364682912826538,
+      "learning_rate": 1.3375319704526595e-06,
+      "loss": 0.8446,
+      "step": 9973
+    },
+    {
+      "epoch": 0.7688868331791551,
+      "grad_norm": 4.229642868041992,
+      "learning_rate": 1.336682174314643e-06,
+      "loss": 1.0238,
+      "step": 9974
+    },
+    {
+      "epoch": 0.7689639222941721,
+      "grad_norm": 3.9624364376068115,
+      "learning_rate": 1.3358326065655187e-06,
+      "loss": 1.0482,
+      "step": 9975
+    },
+    {
+      "epoch": 0.769041011409189,
+      "grad_norm": 3.632380485534668,
+      "learning_rate": 1.334983267258254e-06,
+      "loss": 0.8454,
+      "step": 9976
+    },
+    {
+      "epoch": 0.7691181005242059,
+      "grad_norm": 3.768838882446289,
+      "learning_rate": 1.334134156445801e-06,
+      "loss": 0.8613,
+      "step": 9977
+    },
+    {
+      "epoch": 0.769195189639223,
+      "grad_norm": 3.4363794326782227,
+      "learning_rate": 1.3332852741810975e-06,
+      "loss": 0.8998,
+      "step": 9978
+    },
+    {
+      "epoch": 0.7692722787542399,
+      "grad_norm": 3.6031389236450195,
+      "learning_rate": 1.332436620517068e-06,
+      "loss": 0.8144,
+      "step": 9979
+    },
+    {
+      "epoch": 0.7693493678692569,
+      "grad_norm": 3.714672327041626,
+      "learning_rate": 1.331588195506619e-06,
+      "loss": 1.0322,
+      "step": 9980
+    },
+    {
+      "epoch": 0.7694264569842738,
+      "grad_norm": 3.482741117477417,
+      "learning_rate": 1.330739999202647e-06,
+      "loss": 0.8334,
+      "step": 9981
+    },
+    {
+      "epoch": 0.7695035460992907,
+      "grad_norm": 4.036420822143555,
+      "learning_rate": 1.3298920316580304e-06,
+      "loss": 0.9349,
+      "step": 9982
+    },
+    {
+      "epoch": 0.7695806352143078,
+      "grad_norm": 4.182904243469238,
+      "learning_rate": 1.3290442929256415e-06,
+      "loss": 0.8901,
+      "step": 9983
+    },
+    {
+      "epoch": 0.7696577243293247,
+      "grad_norm": 3.8596582412719727,
+      "learning_rate": 1.3281967830583264e-06,
+      "loss": 0.9771,
+      "step": 9984
+    },
+    {
+      "epoch": 0.7697348134443417,
+      "grad_norm": 3.6666831970214844,
+      "learning_rate": 1.3273495021089255e-06,
+      "loss": 0.8917,
+      "step": 9985
+    },
+    {
+      "epoch": 0.7698119025593586,
+      "grad_norm": 3.531444787979126,
+      "learning_rate": 1.326502450130262e-06,
+      "loss": 0.8601,
+      "step": 9986
+    },
+    {
+      "epoch": 0.7698889916743756,
+      "grad_norm": 4.0722551345825195,
+      "learning_rate": 1.3256556271751454e-06,
+      "loss": 1.0251,
+      "step": 9987
+    },
+    {
+      "epoch": 0.7699660807893925,
+      "grad_norm": 3.8301305770874023,
+      "learning_rate": 1.3248090332963697e-06,
+      "loss": 0.9047,
+      "step": 9988
+    },
+    {
+      "epoch": 0.7700431699044095,
+      "grad_norm": 3.6400532722473145,
+      "learning_rate": 1.323962668546719e-06,
+      "loss": 0.995,
+      "step": 9989
+    },
+    {
+      "epoch": 0.7701202590194265,
+      "grad_norm": 4.178936958312988,
+      "learning_rate": 1.3231165329789546e-06,
+      "loss": 0.9048,
+      "step": 9990
+    },
+    {
+      "epoch": 0.7701973481344434,
+      "grad_norm": 4.014671325683594,
+      "learning_rate": 1.3222706266458323e-06,
+      "loss": 0.9091,
+      "step": 9991
+    },
+    {
+      "epoch": 0.7702744372494604,
+      "grad_norm": 4.306642055511475,
+      "learning_rate": 1.3214249496000887e-06,
+      "loss": 1.0158,
+      "step": 9992
+    },
+    {
+      "epoch": 0.7703515263644773,
+      "grad_norm": 3.796880006790161,
+      "learning_rate": 1.3205795018944473e-06,
+      "loss": 0.9725,
+      "step": 9993
+    },
+    {
+      "epoch": 0.7704286154794943,
+      "grad_norm": 3.244234800338745,
+      "learning_rate": 1.3197342835816196e-06,
+      "loss": 0.7616,
+      "step": 9994
+    },
+    {
+      "epoch": 0.7705057045945113,
+      "grad_norm": 3.749753713607788,
+      "learning_rate": 1.3188892947142973e-06,
+      "loss": 0.825,
+      "step": 9995
+    },
+    {
+      "epoch": 0.7705827937095282,
+      "grad_norm": 3.7566635608673096,
+      "learning_rate": 1.3180445353451621e-06,
+      "loss": 0.91,
+      "step": 9996
+    },
+    {
+      "epoch": 0.7706598828245452,
+      "grad_norm": 3.844532012939453,
+      "learning_rate": 1.3172000055268814e-06,
+      "loss": 0.96,
+      "step": 9997
+    },
+    {
+      "epoch": 0.7707369719395621,
+      "grad_norm": 3.6308090686798096,
+      "learning_rate": 1.3163557053121062e-06,
+      "loss": 0.9925,
+      "step": 9998
+    },
+    {
+      "epoch": 0.7708140610545791,
+      "grad_norm": 3.9852662086486816,
+      "learning_rate": 1.3155116347534746e-06,
+      "loss": 1.0156,
+      "step": 9999
+    },
+    {
+      "epoch": 0.7708911501695961,
+      "grad_norm": 3.719912052154541,
+      "learning_rate": 1.3146677939036118e-06,
+      "loss": 0.9678,
+      "step": 10000
+    },
+    {
+      "epoch": 0.770968239284613,
+      "grad_norm": 3.3894362449645996,
+      "learning_rate": 1.3138241828151238e-06,
+      "loss": 0.8484,
+      "step": 10001
+    },
+    {
+      "epoch": 0.77104532839963,
+      "grad_norm": 3.582899808883667,
+      "learning_rate": 1.3129808015406064e-06,
+      "loss": 0.7916,
+      "step": 10002
+    },
+    {
+      "epoch": 0.7711224175146469,
+      "grad_norm": 3.8180840015411377,
+      "learning_rate": 1.312137650132641e-06,
+      "loss": 0.9036,
+      "step": 10003
+    },
+    {
+      "epoch": 0.7711995066296639,
+      "grad_norm": 3.5141336917877197,
+      "learning_rate": 1.3112947286437927e-06,
+      "loss": 0.8419,
+      "step": 10004
+    },
+    {
+      "epoch": 0.7712765957446809,
+      "grad_norm": 3.5062694549560547,
+      "learning_rate": 1.3104520371266155e-06,
+      "loss": 0.8971,
+      "step": 10005
+    },
+    {
+      "epoch": 0.7713536848596978,
+      "grad_norm": 3.4877374172210693,
+      "learning_rate": 1.3096095756336442e-06,
+      "loss": 0.9091,
+      "step": 10006
+    },
+    {
+      "epoch": 0.7714307739747148,
+      "grad_norm": 4.043210029602051,
+      "learning_rate": 1.308767344217402e-06,
+      "loss": 1.0359,
+      "step": 10007
+    },
+    {
+      "epoch": 0.7715078630897317,
+      "grad_norm": 4.2239990234375,
+      "learning_rate": 1.307925342930399e-06,
+      "loss": 0.9755,
+      "step": 10008
+    },
+    {
+      "epoch": 0.7715849522047487,
+      "grad_norm": 3.9452037811279297,
+      "learning_rate": 1.3070835718251284e-06,
+      "loss": 0.963,
+      "step": 10009
+    },
+    {
+      "epoch": 0.7716620413197657,
+      "grad_norm": 3.688035249710083,
+      "learning_rate": 1.306242030954072e-06,
+      "loss": 0.9624,
+      "step": 10010
+    },
+    {
+      "epoch": 0.7717391304347826,
+      "grad_norm": 3.4075801372528076,
+      "learning_rate": 1.3054007203696955e-06,
+      "loss": 0.8649,
+      "step": 10011
+    },
+    {
+      "epoch": 0.7718162195497996,
+      "grad_norm": 3.897404909133911,
+      "learning_rate": 1.3045596401244477e-06,
+      "loss": 0.8981,
+      "step": 10012
+    },
+    {
+      "epoch": 0.7718933086648165,
+      "grad_norm": 3.891972303390503,
+      "learning_rate": 1.303718790270767e-06,
+      "loss": 0.9878,
+      "step": 10013
+    },
+    {
+      "epoch": 0.7719703977798335,
+      "grad_norm": 3.599642515182495,
+      "learning_rate": 1.3028781708610766e-06,
+      "loss": 0.8243,
+      "step": 10014
+    },
+    {
+      "epoch": 0.7720474868948505,
+      "grad_norm": 3.8693175315856934,
+      "learning_rate": 1.3020377819477843e-06,
+      "loss": 0.9088,
+      "step": 10015
+    },
+    {
+      "epoch": 0.7721245760098674,
+      "grad_norm": 3.6031887531280518,
+      "learning_rate": 1.3011976235832852e-06,
+      "loss": 0.8812,
+      "step": 10016
+    },
+    {
+      "epoch": 0.7722016651248844,
+      "grad_norm": 3.770428419113159,
+      "learning_rate": 1.3003576958199565e-06,
+      "loss": 1.026,
+      "step": 10017
+    },
+    {
+      "epoch": 0.7722787542399013,
+      "grad_norm": 3.2302563190460205,
+      "learning_rate": 1.2995179987101648e-06,
+      "loss": 0.7229,
+      "step": 10018
+    },
+    {
+      "epoch": 0.7723558433549182,
+      "grad_norm": 3.593198537826538,
+      "learning_rate": 1.29867853230626e-06,
+      "loss": 0.7879,
+      "step": 10019
+    },
+    {
+      "epoch": 0.7724329324699353,
+      "grad_norm": 4.028049945831299,
+      "learning_rate": 1.297839296660579e-06,
+      "loss": 0.8327,
+      "step": 10020
+    },
+    {
+      "epoch": 0.7725100215849522,
+      "grad_norm": 3.5654919147491455,
+      "learning_rate": 1.2970002918254443e-06,
+      "loss": 0.9508,
+      "step": 10021
+    },
+    {
+      "epoch": 0.7725871106999692,
+      "grad_norm": 3.5768983364105225,
+      "learning_rate": 1.2961615178531644e-06,
+      "loss": 0.8049,
+      "step": 10022
+    },
+    {
+      "epoch": 0.7726641998149861,
+      "grad_norm": 3.564749240875244,
+      "learning_rate": 1.29532297479603e-06,
+      "loss": 0.8298,
+      "step": 10023
+    },
+    {
+      "epoch": 0.772741288930003,
+      "grad_norm": 3.7964117527008057,
+      "learning_rate": 1.2944846627063208e-06,
+      "loss": 0.8518,
+      "step": 10024
+    },
+    {
+      "epoch": 0.7728183780450201,
+      "grad_norm": 3.8929343223571777,
+      "learning_rate": 1.2936465816363014e-06,
+      "loss": 0.8663,
+      "step": 10025
+    },
+    {
+      "epoch": 0.772895467160037,
+      "grad_norm": 4.378154277801514,
+      "learning_rate": 1.2928087316382225e-06,
+      "loss": 0.9825,
+      "step": 10026
+    },
+    {
+      "epoch": 0.772972556275054,
+      "grad_norm": 3.699422836303711,
+      "learning_rate": 1.2919711127643186e-06,
+      "loss": 0.939,
+      "step": 10027
+    },
+    {
+      "epoch": 0.7730496453900709,
+      "grad_norm": 3.876960277557373,
+      "learning_rate": 1.2911337250668115e-06,
+      "loss": 0.9534,
+      "step": 10028
+    },
+    {
+      "epoch": 0.7731267345050878,
+      "grad_norm": 4.108030319213867,
+      "learning_rate": 1.290296568597908e-06,
+      "loss": 0.9403,
+      "step": 10029
+    },
+    {
+      "epoch": 0.7732038236201049,
+      "grad_norm": 3.5450022220611572,
+      "learning_rate": 1.2894596434098006e-06,
+      "loss": 0.9481,
+      "step": 10030
+    },
+    {
+      "epoch": 0.7732809127351218,
+      "grad_norm": 3.4901676177978516,
+      "learning_rate": 1.2886229495546687e-06,
+      "loss": 0.8106,
+      "step": 10031
+    },
+    {
+      "epoch": 0.7733580018501388,
+      "grad_norm": 4.202146053314209,
+      "learning_rate": 1.2877864870846724e-06,
+      "loss": 0.8711,
+      "step": 10032
+    },
+    {
+      "epoch": 0.7734350909651557,
+      "grad_norm": 3.778503656387329,
+      "learning_rate": 1.2869502560519626e-06,
+      "loss": 1.053,
+      "step": 10033
+    },
+    {
+      "epoch": 0.7735121800801726,
+      "grad_norm": 3.5146751403808594,
+      "learning_rate": 1.2861142565086737e-06,
+      "loss": 0.7681,
+      "step": 10034
+    },
+    {
+      "epoch": 0.7735892691951897,
+      "grad_norm": 3.5477395057678223,
+      "learning_rate": 1.2852784885069265e-06,
+      "loss": 0.9164,
+      "step": 10035
+    },
+    {
+      "epoch": 0.7736663583102066,
+      "grad_norm": 3.6575543880462646,
+      "learning_rate": 1.284442952098826e-06,
+      "loss": 0.8941,
+      "step": 10036
+    },
+    {
+      "epoch": 0.7737434474252236,
+      "grad_norm": 3.6682982444763184,
+      "learning_rate": 1.2836076473364662e-06,
+      "loss": 0.824,
+      "step": 10037
+    },
+    {
+      "epoch": 0.7738205365402405,
+      "grad_norm": 3.7592432498931885,
+      "learning_rate": 1.2827725742719205e-06,
+      "loss": 0.9658,
+      "step": 10038
+    },
+    {
+      "epoch": 0.7738976256552574,
+      "grad_norm": 3.7127466201782227,
+      "learning_rate": 1.2819377329572525e-06,
+      "loss": 0.8969,
+      "step": 10039
+    },
+    {
+      "epoch": 0.7739747147702745,
+      "grad_norm": 3.9338138103485107,
+      "learning_rate": 1.2811031234445103e-06,
+      "loss": 1.0264,
+      "step": 10040
+    },
+    {
+      "epoch": 0.7740518038852914,
+      "grad_norm": 3.961344003677368,
+      "learning_rate": 1.2802687457857277e-06,
+      "loss": 0.9453,
+      "step": 10041
+    },
+    {
+      "epoch": 0.7741288930003084,
+      "grad_norm": 3.9601597785949707,
+      "learning_rate": 1.2794346000329256e-06,
+      "loss": 0.9532,
+      "step": 10042
+    },
+    {
+      "epoch": 0.7742059821153253,
+      "grad_norm": 3.8072171211242676,
+      "learning_rate": 1.278600686238105e-06,
+      "loss": 1.0072,
+      "step": 10043
+    },
+    {
+      "epoch": 0.7742830712303422,
+      "grad_norm": 3.795827865600586,
+      "learning_rate": 1.2777670044532586e-06,
+      "loss": 0.875,
+      "step": 10044
+    },
+    {
+      "epoch": 0.7743601603453593,
+      "grad_norm": 3.6865553855895996,
+      "learning_rate": 1.2769335547303613e-06,
+      "loss": 0.9439,
+      "step": 10045
+    },
+    {
+      "epoch": 0.7744372494603762,
+      "grad_norm": 3.5264663696289062,
+      "learning_rate": 1.2761003371213743e-06,
+      "loss": 0.942,
+      "step": 10046
+    },
+    {
+      "epoch": 0.7745143385753932,
+      "grad_norm": 3.729469060897827,
+      "learning_rate": 1.2752673516782448e-06,
+      "loss": 0.9807,
+      "step": 10047
+    },
+    {
+      "epoch": 0.7745914276904101,
+      "grad_norm": 3.5642600059509277,
+      "learning_rate": 1.2744345984529066e-06,
+      "loss": 0.9168,
+      "step": 10048
+    },
+    {
+      "epoch": 0.774668516805427,
+      "grad_norm": 3.684962749481201,
+      "learning_rate": 1.2736020774972746e-06,
+      "loss": 0.9065,
+      "step": 10049
+    },
+    {
+      "epoch": 0.7747456059204441,
+      "grad_norm": 3.697866916656494,
+      "learning_rate": 1.2727697888632534e-06,
+      "loss": 0.8338,
+      "step": 10050
+    },
+    {
+      "epoch": 0.774822695035461,
+      "grad_norm": 4.074203968048096,
+      "learning_rate": 1.271937732602732e-06,
+      "loss": 0.868,
+      "step": 10051
+    },
+    {
+      "epoch": 0.774899784150478,
+      "grad_norm": 3.6111466884613037,
+      "learning_rate": 1.2711059087675853e-06,
+      "loss": 0.9527,
+      "step": 10052
+    },
+    {
+      "epoch": 0.7749768732654949,
+      "grad_norm": 3.5815255641937256,
+      "learning_rate": 1.2702743174096737e-06,
+      "loss": 0.9145,
+      "step": 10053
+    },
+    {
+      "epoch": 0.7750539623805118,
+      "grad_norm": 3.476964235305786,
+      "learning_rate": 1.2694429585808404e-06,
+      "loss": 0.8759,
+      "step": 10054
+    },
+    {
+      "epoch": 0.7751310514955289,
+      "grad_norm": 3.8654966354370117,
+      "learning_rate": 1.2686118323329178e-06,
+      "loss": 0.9899,
+      "step": 10055
+    },
+    {
+      "epoch": 0.7752081406105458,
+      "grad_norm": 3.7267918586730957,
+      "learning_rate": 1.267780938717722e-06,
+      "loss": 0.9701,
+      "step": 10056
+    },
+    {
+      "epoch": 0.7752852297255628,
+      "grad_norm": 4.117705821990967,
+      "learning_rate": 1.2669502777870546e-06,
+      "loss": 0.8462,
+      "step": 10057
+    },
+    {
+      "epoch": 0.7753623188405797,
+      "grad_norm": 3.576138496398926,
+      "learning_rate": 1.266119849592704e-06,
+      "loss": 0.9018,
+      "step": 10058
+    },
+    {
+      "epoch": 0.7754394079555966,
+      "grad_norm": 5.330030918121338,
+      "learning_rate": 1.2652896541864435e-06,
+      "loss": 0.8455,
+      "step": 10059
+    },
+    {
+      "epoch": 0.7755164970706137,
+      "grad_norm": 3.8612890243530273,
+      "learning_rate": 1.264459691620029e-06,
+      "loss": 0.9188,
+      "step": 10060
+    },
+    {
+      "epoch": 0.7755935861856306,
+      "grad_norm": 4.6686787605285645,
+      "learning_rate": 1.2636299619452064e-06,
+      "loss": 0.9712,
+      "step": 10061
+    },
+    {
+      "epoch": 0.7756706753006476,
+      "grad_norm": 4.158293724060059,
+      "learning_rate": 1.2628004652137044e-06,
+      "loss": 0.8994,
+      "step": 10062
+    },
+    {
+      "epoch": 0.7757477644156645,
+      "grad_norm": 3.727849245071411,
+      "learning_rate": 1.2619712014772378e-06,
+      "loss": 0.8811,
+      "step": 10063
+    },
+    {
+      "epoch": 0.7758248535306814,
+      "grad_norm": 3.6993374824523926,
+      "learning_rate": 1.2611421707875083e-06,
+      "loss": 0.848,
+      "step": 10064
+    },
+    {
+      "epoch": 0.7759019426456985,
+      "grad_norm": 3.298645496368408,
+      "learning_rate": 1.2603133731961992e-06,
+      "loss": 0.8555,
+      "step": 10065
+    },
+    {
+      "epoch": 0.7759790317607154,
+      "grad_norm": 3.6150310039520264,
+      "learning_rate": 1.2594848087549826e-06,
+      "loss": 0.9416,
+      "step": 10066
+    },
+    {
+      "epoch": 0.7760561208757324,
+      "grad_norm": 3.694746732711792,
+      "learning_rate": 1.2586564775155158e-06,
+      "loss": 0.8568,
+      "step": 10067
+    },
+    {
+      "epoch": 0.7761332099907493,
+      "grad_norm": 4.369163990020752,
+      "learning_rate": 1.257828379529441e-06,
+      "loss": 1.0183,
+      "step": 10068
+    },
+    {
+      "epoch": 0.7762102991057662,
+      "grad_norm": 3.4145572185516357,
+      "learning_rate": 1.257000514848385e-06,
+      "loss": 0.874,
+      "step": 10069
+    },
+    {
+      "epoch": 0.7762873882207832,
+      "grad_norm": 4.185209274291992,
+      "learning_rate": 1.2561728835239633e-06,
+      "loss": 0.8839,
+      "step": 10070
+    },
+    {
+      "epoch": 0.7763644773358002,
+      "grad_norm": 3.814661979675293,
+      "learning_rate": 1.255345485607769e-06,
+      "loss": 0.9982,
+      "step": 10071
+    },
+    {
+      "epoch": 0.7764415664508172,
+      "grad_norm": 3.9218826293945312,
+      "learning_rate": 1.2545183211513918e-06,
+      "loss": 0.933,
+      "step": 10072
+    },
+    {
+      "epoch": 0.7765186555658341,
+      "grad_norm": 4.283771514892578,
+      "learning_rate": 1.2536913902063985e-06,
+      "loss": 0.9881,
+      "step": 10073
+    },
+    {
+      "epoch": 0.776595744680851,
+      "grad_norm": 3.5074286460876465,
+      "learning_rate": 1.252864692824346e-06,
+      "loss": 0.9492,
+      "step": 10074
+    },
+    {
+      "epoch": 0.776672833795868,
+      "grad_norm": 3.372987747192383,
+      "learning_rate": 1.2520382290567717e-06,
+      "loss": 0.8843,
+      "step": 10075
+    },
+    {
+      "epoch": 0.776749922910885,
+      "grad_norm": 3.437117099761963,
+      "learning_rate": 1.2512119989552023e-06,
+      "loss": 0.7762,
+      "step": 10076
+    },
+    {
+      "epoch": 0.776827012025902,
+      "grad_norm": 3.731052875518799,
+      "learning_rate": 1.2503860025711494e-06,
+      "loss": 1.0258,
+      "step": 10077
+    },
+    {
+      "epoch": 0.7769041011409189,
+      "grad_norm": 3.422943592071533,
+      "learning_rate": 1.2495602399561096e-06,
+      "loss": 0.7939,
+      "step": 10078
+    },
+    {
+      "epoch": 0.7769811902559358,
+      "grad_norm": 3.4586517810821533,
+      "learning_rate": 1.248734711161566e-06,
+      "loss": 0.7924,
+      "step": 10079
+    },
+    {
+      "epoch": 0.7770582793709528,
+      "grad_norm": 3.3821561336517334,
+      "learning_rate": 1.2479094162389838e-06,
+      "loss": 0.8364,
+      "step": 10080
+    },
+    {
+      "epoch": 0.7771353684859698,
+      "grad_norm": 4.26823091506958,
+      "learning_rate": 1.2470843552398166e-06,
+      "loss": 0.9628,
+      "step": 10081
+    },
+    {
+      "epoch": 0.7772124576009868,
+      "grad_norm": 3.7963852882385254,
+      "learning_rate": 1.2462595282155032e-06,
+      "loss": 0.9598,
+      "step": 10082
+    },
+    {
+      "epoch": 0.7772895467160037,
+      "grad_norm": 3.440056800842285,
+      "learning_rate": 1.2454349352174666e-06,
+      "loss": 0.887,
+      "step": 10083
+    },
+    {
+      "epoch": 0.7773666358310206,
+      "grad_norm": 3.769310235977173,
+      "learning_rate": 1.2446105762971167e-06,
+      "loss": 0.9397,
+      "step": 10084
+    },
+    {
+      "epoch": 0.7774437249460376,
+      "grad_norm": 3.7799243927001953,
+      "learning_rate": 1.2437864515058495e-06,
+      "loss": 0.9923,
+      "step": 10085
+    },
+    {
+      "epoch": 0.7775208140610546,
+      "grad_norm": 3.681706428527832,
+      "learning_rate": 1.2429625608950412e-06,
+      "loss": 0.9673,
+      "step": 10086
+    },
+    {
+      "epoch": 0.7775979031760716,
+      "grad_norm": 3.8595070838928223,
+      "learning_rate": 1.2421389045160593e-06,
+      "loss": 0.8813,
+      "step": 10087
+    },
+    {
+      "epoch": 0.7776749922910885,
+      "grad_norm": 4.8416361808776855,
+      "learning_rate": 1.2413154824202545e-06,
+      "loss": 0.9058,
+      "step": 10088
+    },
+    {
+      "epoch": 0.7777520814061054,
+      "grad_norm": 3.877955198287964,
+      "learning_rate": 1.240492294658962e-06,
+      "loss": 0.9176,
+      "step": 10089
+    },
+    {
+      "epoch": 0.7778291705211224,
+      "grad_norm": 3.9161739349365234,
+      "learning_rate": 1.239669341283506e-06,
+      "loss": 0.9904,
+      "step": 10090
+    },
+    {
+      "epoch": 0.7779062596361394,
+      "grad_norm": 3.6513936519622803,
+      "learning_rate": 1.2388466223451895e-06,
+      "loss": 0.8185,
+      "step": 10091
+    },
+    {
+      "epoch": 0.7779833487511564,
+      "grad_norm": 3.919912099838257,
+      "learning_rate": 1.2380241378953067e-06,
+      "loss": 0.8561,
+      "step": 10092
+    },
+    {
+      "epoch": 0.7780604378661733,
+      "grad_norm": 3.2892303466796875,
+      "learning_rate": 1.237201887985135e-06,
+      "loss": 0.902,
+      "step": 10093
+    },
+    {
+      "epoch": 0.7781375269811902,
+      "grad_norm": 3.583991527557373,
+      "learning_rate": 1.2363798726659377e-06,
+      "loss": 0.8996,
+      "step": 10094
+    },
+    {
+      "epoch": 0.7782146160962072,
+      "grad_norm": 3.9452333450317383,
+      "learning_rate": 1.235558091988963e-06,
+      "loss": 1.0222,
+      "step": 10095
+    },
+    {
+      "epoch": 0.7782917052112241,
+      "grad_norm": 3.6943013668060303,
+      "learning_rate": 1.234736546005446e-06,
+      "loss": 0.9208,
+      "step": 10096
+    },
+    {
+      "epoch": 0.7783687943262412,
+      "grad_norm": 3.9956469535827637,
+      "learning_rate": 1.2339152347666033e-06,
+      "loss": 1.0566,
+      "step": 10097
+    },
+    {
+      "epoch": 0.7784458834412581,
+      "grad_norm": 3.802483558654785,
+      "learning_rate": 1.2330941583236406e-06,
+      "loss": 0.9223,
+      "step": 10098
+    },
+    {
+      "epoch": 0.778522972556275,
+      "grad_norm": 3.4962236881256104,
+      "learning_rate": 1.2322733167277479e-06,
+      "loss": 0.8389,
+      "step": 10099
+    },
+    {
+      "epoch": 0.778600061671292,
+      "grad_norm": 3.899533271789551,
+      "learning_rate": 1.2314527100301005e-06,
+      "loss": 1.0017,
+      "step": 10100
+    },
+    {
+      "epoch": 0.778677150786309,
+      "grad_norm": 3.7576115131378174,
+      "learning_rate": 1.2306323382818596e-06,
+      "loss": 0.9071,
+      "step": 10101
+    },
+    {
+      "epoch": 0.778754239901326,
+      "grad_norm": 3.894066333770752,
+      "learning_rate": 1.2298122015341696e-06,
+      "loss": 1.0241,
+      "step": 10102
+    },
+    {
+      "epoch": 0.7788313290163429,
+      "grad_norm": 3.69673490524292,
+      "learning_rate": 1.2289922998381625e-06,
+      "loss": 0.9332,
+      "step": 10103
+    },
+    {
+      "epoch": 0.7789084181313598,
+      "grad_norm": 4.03447151184082,
+      "learning_rate": 1.2281726332449544e-06,
+      "loss": 1.017,
+      "step": 10104
+    },
+    {
+      "epoch": 0.7789855072463768,
+      "grad_norm": 3.7014174461364746,
+      "learning_rate": 1.2273532018056482e-06,
+      "loss": 0.9177,
+      "step": 10105
+    },
+    {
+      "epoch": 0.7790625963613937,
+      "grad_norm": 4.122037410736084,
+      "learning_rate": 1.226534005571331e-06,
+      "loss": 0.918,
+      "step": 10106
+    },
+    {
+      "epoch": 0.7791396854764108,
+      "grad_norm": 3.817849636077881,
+      "learning_rate": 1.2257150445930765e-06,
+      "loss": 0.8933,
+      "step": 10107
+    },
+    {
+      "epoch": 0.7792167745914277,
+      "grad_norm": 3.191657304763794,
+      "learning_rate": 1.2248963189219398e-06,
+      "loss": 0.7782,
+      "step": 10108
+    },
+    {
+      "epoch": 0.7792938637064446,
+      "grad_norm": 3.7507643699645996,
+      "learning_rate": 1.224077828608966e-06,
+      "loss": 0.9169,
+      "step": 10109
+    },
+    {
+      "epoch": 0.7793709528214616,
+      "grad_norm": 3.5642547607421875,
+      "learning_rate": 1.223259573705184e-06,
+      "loss": 0.9119,
+      "step": 10110
+    },
+    {
+      "epoch": 0.7794480419364785,
+      "grad_norm": 3.764009952545166,
+      "learning_rate": 1.2224415542616069e-06,
+      "loss": 0.8835,
+      "step": 10111
+    },
+    {
+      "epoch": 0.7795251310514956,
+      "grad_norm": 3.5799670219421387,
+      "learning_rate": 1.2216237703292361e-06,
+      "loss": 0.9183,
+      "step": 10112
+    },
+    {
+      "epoch": 0.7796022201665125,
+      "grad_norm": 3.8919787406921387,
+      "learning_rate": 1.2208062219590527e-06,
+      "loss": 0.9795,
+      "step": 10113
+    },
+    {
+      "epoch": 0.7796793092815294,
+      "grad_norm": 3.9707539081573486,
+      "learning_rate": 1.2199889092020289e-06,
+      "loss": 1.004,
+      "step": 10114
+    },
+    {
+      "epoch": 0.7797563983965464,
+      "grad_norm": 3.72212815284729,
+      "learning_rate": 1.2191718321091178e-06,
+      "loss": 0.9145,
+      "step": 10115
+    },
+    {
+      "epoch": 0.7798334875115633,
+      "grad_norm": 3.555389642715454,
+      "learning_rate": 1.2183549907312624e-06,
+      "loss": 0.8632,
+      "step": 10116
+    },
+    {
+      "epoch": 0.7799105766265804,
+      "grad_norm": 3.579590320587158,
+      "learning_rate": 1.2175383851193901e-06,
+      "loss": 0.8866,
+      "step": 10117
+    },
+    {
+      "epoch": 0.7799876657415973,
+      "grad_norm": 4.378693103790283,
+      "learning_rate": 1.2167220153244075e-06,
+      "loss": 0.8158,
+      "step": 10118
+    },
+    {
+      "epoch": 0.7800647548566142,
+      "grad_norm": 3.398848056793213,
+      "learning_rate": 1.2159058813972135e-06,
+      "loss": 0.9214,
+      "step": 10119
+    },
+    {
+      "epoch": 0.7801418439716312,
+      "grad_norm": 3.9123952388763428,
+      "learning_rate": 1.2150899833886892e-06,
+      "loss": 0.8256,
+      "step": 10120
+    },
+    {
+      "epoch": 0.7802189330866481,
+      "grad_norm": 4.100687026977539,
+      "learning_rate": 1.214274321349702e-06,
+      "loss": 0.9506,
+      "step": 10121
+    },
+    {
+      "epoch": 0.7802960222016652,
+      "grad_norm": 3.491703748703003,
+      "learning_rate": 1.2134588953311056e-06,
+      "loss": 0.8823,
+      "step": 10122
+    },
+    {
+      "epoch": 0.7803731113166821,
+      "grad_norm": 3.760547637939453,
+      "learning_rate": 1.2126437053837348e-06,
+      "loss": 0.9935,
+      "step": 10123
+    },
+    {
+      "epoch": 0.780450200431699,
+      "grad_norm": 3.883286237716675,
+      "learning_rate": 1.2118287515584132e-06,
+      "loss": 0.8716,
+      "step": 10124
+    },
+    {
+      "epoch": 0.780527289546716,
+      "grad_norm": 3.697366714477539,
+      "learning_rate": 1.2110140339059501e-06,
+      "loss": 0.8956,
+      "step": 10125
+    },
+    {
+      "epoch": 0.7806043786617329,
+      "grad_norm": 3.8138182163238525,
+      "learning_rate": 1.2101995524771376e-06,
+      "loss": 0.8893,
+      "step": 10126
+    },
+    {
+      "epoch": 0.78068146777675,
+      "grad_norm": 3.8224334716796875,
+      "learning_rate": 1.2093853073227574e-06,
+      "loss": 0.9656,
+      "step": 10127
+    },
+    {
+      "epoch": 0.7807585568917669,
+      "grad_norm": 3.8924286365509033,
+      "learning_rate": 1.2085712984935693e-06,
+      "loss": 0.8809,
+      "step": 10128
+    },
+    {
+      "epoch": 0.7808356460067838,
+      "grad_norm": 3.684929847717285,
+      "learning_rate": 1.2077575260403247e-06,
+      "loss": 1.0676,
+      "step": 10129
+    },
+    {
+      "epoch": 0.7809127351218008,
+      "grad_norm": 3.5521864891052246,
+      "learning_rate": 1.2069439900137575e-06,
+      "loss": 0.839,
+      "step": 10130
+    },
+    {
+      "epoch": 0.7809898242368177,
+      "grad_norm": 3.7578213214874268,
+      "learning_rate": 1.2061306904645875e-06,
+      "loss": 0.9186,
+      "step": 10131
+    },
+    {
+      "epoch": 0.7810669133518348,
+      "grad_norm": 3.4750614166259766,
+      "learning_rate": 1.2053176274435209e-06,
+      "loss": 0.86,
+      "step": 10132
+    },
+    {
+      "epoch": 0.7811440024668517,
+      "grad_norm": 3.816967487335205,
+      "learning_rate": 1.2045048010012477e-06,
+      "loss": 0.9405,
+      "step": 10133
+    },
+    {
+      "epoch": 0.7812210915818686,
+      "grad_norm": 4.11097526550293,
+      "learning_rate": 1.2036922111884414e-06,
+      "loss": 1.0416,
+      "step": 10134
+    },
+    {
+      "epoch": 0.7812981806968856,
+      "grad_norm": 3.8455522060394287,
+      "learning_rate": 1.2028798580557644e-06,
+      "loss": 0.9472,
+      "step": 10135
+    },
+    {
+      "epoch": 0.7813752698119025,
+      "grad_norm": 3.7609121799468994,
+      "learning_rate": 1.2020677416538623e-06,
+      "loss": 0.9292,
+      "step": 10136
+    },
+    {
+      "epoch": 0.7814523589269196,
+      "grad_norm": 3.555689573287964,
+      "learning_rate": 1.201255862033367e-06,
+      "loss": 0.9046,
+      "step": 10137
+    },
+    {
+      "epoch": 0.7815294480419365,
+      "grad_norm": 3.5667762756347656,
+      "learning_rate": 1.2004442192448956e-06,
+      "loss": 0.8375,
+      "step": 10138
+    },
+    {
+      "epoch": 0.7816065371569534,
+      "grad_norm": 3.609121322631836,
+      "learning_rate": 1.1996328133390472e-06,
+      "loss": 0.9084,
+      "step": 10139
+    },
+    {
+      "epoch": 0.7816836262719704,
+      "grad_norm": 3.747819423675537,
+      "learning_rate": 1.1988216443664102e-06,
+      "loss": 0.8834,
+      "step": 10140
+    },
+    {
+      "epoch": 0.7817607153869873,
+      "grad_norm": 3.649038314819336,
+      "learning_rate": 1.198010712377557e-06,
+      "loss": 0.9357,
+      "step": 10141
+    },
+    {
+      "epoch": 0.7818378045020044,
+      "grad_norm": 3.519754409790039,
+      "learning_rate": 1.1972000174230452e-06,
+      "loss": 0.8994,
+      "step": 10142
+    },
+    {
+      "epoch": 0.7819148936170213,
+      "grad_norm": 3.7660558223724365,
+      "learning_rate": 1.1963895595534164e-06,
+      "loss": 0.8964,
+      "step": 10143
+    },
+    {
+      "epoch": 0.7819919827320382,
+      "grad_norm": 3.712606430053711,
+      "learning_rate": 1.195579338819201e-06,
+      "loss": 0.8096,
+      "step": 10144
+    },
+    {
+      "epoch": 0.7820690718470552,
+      "grad_norm": 3.9107108116149902,
+      "learning_rate": 1.194769355270909e-06,
+      "loss": 0.998,
+      "step": 10145
+    },
+    {
+      "epoch": 0.7821461609620721,
+      "grad_norm": 3.800839424133301,
+      "learning_rate": 1.1939596089590394e-06,
+      "loss": 0.8565,
+      "step": 10146
+    },
+    {
+      "epoch": 0.7822232500770891,
+      "grad_norm": 3.84592342376709,
+      "learning_rate": 1.1931500999340768e-06,
+      "loss": 0.9408,
+      "step": 10147
+    },
+    {
+      "epoch": 0.7823003391921061,
+      "grad_norm": 3.9664783477783203,
+      "learning_rate": 1.1923408282464888e-06,
+      "loss": 1.0395,
+      "step": 10148
+    },
+    {
+      "epoch": 0.782377428307123,
+      "grad_norm": 3.7878577709198,
+      "learning_rate": 1.1915317939467314e-06,
+      "loss": 0.964,
+      "step": 10149
+    },
+    {
+      "epoch": 0.78245451742214,
+      "grad_norm": 3.499857187271118,
+      "learning_rate": 1.1907229970852407e-06,
+      "loss": 0.8254,
+      "step": 10150
+    },
+    {
+      "epoch": 0.7825316065371569,
+      "grad_norm": 3.702730894088745,
+      "learning_rate": 1.1899144377124422e-06,
+      "loss": 0.9191,
+      "step": 10151
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 3.7440195083618164,
+      "learning_rate": 1.1891061158787459e-06,
+      "loss": 0.8737,
+      "step": 10152
+    },
+    {
+      "epoch": 0.7826857847671909,
+      "grad_norm": 3.973545551300049,
+      "learning_rate": 1.1882980316345461e-06,
+      "loss": 1.0105,
+      "step": 10153
+    },
+    {
+      "epoch": 0.7827628738822078,
+      "grad_norm": 3.627835273742676,
+      "learning_rate": 1.1874901850302223e-06,
+      "loss": 0.8726,
+      "step": 10154
+    },
+    {
+      "epoch": 0.7828399629972248,
+      "grad_norm": 4.042776107788086,
+      "learning_rate": 1.1866825761161417e-06,
+      "loss": 0.9184,
+      "step": 10155
+    },
+    {
+      "epoch": 0.7829170521122417,
+      "grad_norm": 3.9499711990356445,
+      "learning_rate": 1.1858752049426513e-06,
+      "loss": 1.001,
+      "step": 10156
+    },
+    {
+      "epoch": 0.7829941412272587,
+      "grad_norm": 3.8737709522247314,
+      "learning_rate": 1.1850680715600883e-06,
+      "loss": 0.9999,
+      "step": 10157
+    },
+    {
+      "epoch": 0.7830712303422757,
+      "grad_norm": 3.580709934234619,
+      "learning_rate": 1.1842611760187722e-06,
+      "loss": 0.8882,
+      "step": 10158
+    },
+    {
+      "epoch": 0.7831483194572927,
+      "grad_norm": 3.639634847640991,
+      "learning_rate": 1.1834545183690105e-06,
+      "loss": 0.8136,
+      "step": 10159
+    },
+    {
+      "epoch": 0.7832254085723096,
+      "grad_norm": 3.59987735748291,
+      "learning_rate": 1.1826480986610928e-06,
+      "loss": 0.8607,
+      "step": 10160
+    },
+    {
+      "epoch": 0.7833024976873265,
+      "grad_norm": 3.7638700008392334,
+      "learning_rate": 1.1818419169452954e-06,
+      "loss": 0.836,
+      "step": 10161
+    },
+    {
+      "epoch": 0.7833795868023435,
+      "grad_norm": 3.5380563735961914,
+      "learning_rate": 1.1810359732718795e-06,
+      "loss": 0.847,
+      "step": 10162
+    },
+    {
+      "epoch": 0.7834566759173605,
+      "grad_norm": 4.049480438232422,
+      "learning_rate": 1.1802302676910926e-06,
+      "loss": 1.0036,
+      "step": 10163
+    },
+    {
+      "epoch": 0.7835337650323775,
+      "grad_norm": 3.6710615158081055,
+      "learning_rate": 1.1794248002531644e-06,
+      "loss": 0.8508,
+      "step": 10164
+    },
+    {
+      "epoch": 0.7836108541473944,
+      "grad_norm": 3.6866486072540283,
+      "learning_rate": 1.1786195710083149e-06,
+      "loss": 0.8394,
+      "step": 10165
+    },
+    {
+      "epoch": 0.7836879432624113,
+      "grad_norm": 4.017769813537598,
+      "learning_rate": 1.1778145800067419e-06,
+      "loss": 0.9146,
+      "step": 10166
+    },
+    {
+      "epoch": 0.7837650323774283,
+      "grad_norm": 3.859618902206421,
+      "learning_rate": 1.177009827298634e-06,
+      "loss": 0.9376,
+      "step": 10167
+    },
+    {
+      "epoch": 0.7838421214924453,
+      "grad_norm": 3.7508604526519775,
+      "learning_rate": 1.1762053129341643e-06,
+      "loss": 0.8773,
+      "step": 10168
+    },
+    {
+      "epoch": 0.7839192106074623,
+      "grad_norm": 4.094524383544922,
+      "learning_rate": 1.175401036963489e-06,
+      "loss": 0.9331,
+      "step": 10169
+    },
+    {
+      "epoch": 0.7839962997224792,
+      "grad_norm": 3.7034811973571777,
+      "learning_rate": 1.1745969994367524e-06,
+      "loss": 0.7125,
+      "step": 10170
+    },
+    {
+      "epoch": 0.7840733888374961,
+      "grad_norm": 3.425140857696533,
+      "learning_rate": 1.1737932004040792e-06,
+      "loss": 0.8507,
+      "step": 10171
+    },
+    {
+      "epoch": 0.7841504779525131,
+      "grad_norm": 3.6695098876953125,
+      "learning_rate": 1.1729896399155831e-06,
+      "loss": 0.96,
+      "step": 10172
+    },
+    {
+      "epoch": 0.78422756706753,
+      "grad_norm": 3.393888235092163,
+      "learning_rate": 1.1721863180213627e-06,
+      "loss": 0.7526,
+      "step": 10173
+    },
+    {
+      "epoch": 0.7843046561825471,
+      "grad_norm": 3.838538408279419,
+      "learning_rate": 1.171383234771501e-06,
+      "loss": 0.9423,
+      "step": 10174
+    },
+    {
+      "epoch": 0.784381745297564,
+      "grad_norm": 3.5390851497650146,
+      "learning_rate": 1.1705803902160668e-06,
+      "loss": 0.9341,
+      "step": 10175
+    },
+    {
+      "epoch": 0.7844588344125809,
+      "grad_norm": 3.473173141479492,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.903,
+      "step": 10176
+    },
+    {
+      "epoch": 0.7845359235275979,
+      "grad_norm": 4.052275657653809,
+      "learning_rate": 1.1689754173886725e-06,
+      "loss": 0.9877,
+      "step": 10177
+    },
+    {
+      "epoch": 0.7846130126426148,
+      "grad_norm": 3.684760808944702,
+      "learning_rate": 1.1681732892167757e-06,
+      "loss": 0.9197,
+      "step": 10178
+    },
+    {
+      "epoch": 0.7846901017576319,
+      "grad_norm": 3.7572269439697266,
+      "learning_rate": 1.1673713999394287e-06,
+      "loss": 0.9295,
+      "step": 10179
+    },
+    {
+      "epoch": 0.7847671908726488,
+      "grad_norm": 3.714064359664917,
+      "learning_rate": 1.1665697496066253e-06,
+      "loss": 0.8446,
+      "step": 10180
+    },
+    {
+      "epoch": 0.7848442799876657,
+      "grad_norm": 3.450218439102173,
+      "learning_rate": 1.1657683382683454e-06,
+      "loss": 0.9163,
+      "step": 10181
+    },
+    {
+      "epoch": 0.7849213691026827,
+      "grad_norm": 3.6925554275512695,
+      "learning_rate": 1.1649671659745504e-06,
+      "loss": 0.9408,
+      "step": 10182
+    },
+    {
+      "epoch": 0.7849984582176996,
+      "grad_norm": 4.052911758422852,
+      "learning_rate": 1.1641662327751907e-06,
+      "loss": 0.9494,
+      "step": 10183
+    },
+    {
+      "epoch": 0.7850755473327167,
+      "grad_norm": 4.215100288391113,
+      "learning_rate": 1.1633655387201998e-06,
+      "loss": 0.8627,
+      "step": 10184
+    },
+    {
+      "epoch": 0.7851526364477336,
+      "grad_norm": 3.8449103832244873,
+      "learning_rate": 1.162565083859497e-06,
+      "loss": 0.9384,
+      "step": 10185
+    },
+    {
+      "epoch": 0.7852297255627505,
+      "grad_norm": 3.831585645675659,
+      "learning_rate": 1.1617648682429882e-06,
+      "loss": 0.8777,
+      "step": 10186
+    },
+    {
+      "epoch": 0.7853068146777675,
+      "grad_norm": 3.922975540161133,
+      "learning_rate": 1.1609648919205596e-06,
+      "loss": 0.9134,
+      "step": 10187
+    },
+    {
+      "epoch": 0.7853839037927844,
+      "grad_norm": 3.6187238693237305,
+      "learning_rate": 1.1601651549420873e-06,
+      "loss": 0.9355,
+      "step": 10188
+    },
+    {
+      "epoch": 0.7854609929078015,
+      "grad_norm": 3.704779624938965,
+      "learning_rate": 1.1593656573574302e-06,
+      "loss": 0.9219,
+      "step": 10189
+    },
+    {
+      "epoch": 0.7855380820228184,
+      "grad_norm": 4.0944719314575195,
+      "learning_rate": 1.1585663992164336e-06,
+      "loss": 0.9159,
+      "step": 10190
+    },
+    {
+      "epoch": 0.7856151711378353,
+      "grad_norm": 3.6618051528930664,
+      "learning_rate": 1.1577673805689266e-06,
+      "loss": 0.8657,
+      "step": 10191
+    },
+    {
+      "epoch": 0.7856922602528523,
+      "grad_norm": 3.9331610202789307,
+      "learning_rate": 1.1569686014647253e-06,
+      "loss": 0.9718,
+      "step": 10192
+    },
+    {
+      "epoch": 0.7857693493678692,
+      "grad_norm": 3.8769805431365967,
+      "learning_rate": 1.156170061953627e-06,
+      "loss": 0.846,
+      "step": 10193
+    },
+    {
+      "epoch": 0.7858464384828863,
+      "grad_norm": 3.4525961875915527,
+      "learning_rate": 1.1553717620854176e-06,
+      "loss": 1.0036,
+      "step": 10194
+    },
+    {
+      "epoch": 0.7859235275979032,
+      "grad_norm": 3.376194715499878,
+      "learning_rate": 1.1545737019098668e-06,
+      "loss": 0.8905,
+      "step": 10195
+    },
+    {
+      "epoch": 0.7860006167129201,
+      "grad_norm": 3.572995185852051,
+      "learning_rate": 1.1537758814767298e-06,
+      "loss": 0.8374,
+      "step": 10196
+    },
+    {
+      "epoch": 0.7860777058279371,
+      "grad_norm": 3.6022350788116455,
+      "learning_rate": 1.1529783008357476e-06,
+      "loss": 0.9317,
+      "step": 10197
+    },
+    {
+      "epoch": 0.786154794942954,
+      "grad_norm": 3.5814099311828613,
+      "learning_rate": 1.152180960036643e-06,
+      "loss": 0.8554,
+      "step": 10198
+    },
+    {
+      "epoch": 0.7862318840579711,
+      "grad_norm": 3.6974618434906006,
+      "learning_rate": 1.151383859129127e-06,
+      "loss": 0.7929,
+      "step": 10199
+    },
+    {
+      "epoch": 0.786308973172988,
+      "grad_norm": 3.654067039489746,
+      "learning_rate": 1.1505869981628953e-06,
+      "loss": 0.8319,
+      "step": 10200
+    },
+    {
+      "epoch": 0.7863860622880049,
+      "grad_norm": 3.9007177352905273,
+      "learning_rate": 1.1497903771876272e-06,
+      "loss": 1.0742,
+      "step": 10201
+    },
+    {
+      "epoch": 0.7864631514030219,
+      "grad_norm": 4.104714870452881,
+      "learning_rate": 1.1489939962529884e-06,
+      "loss": 0.9199,
+      "step": 10202
+    },
+    {
+      "epoch": 0.7865402405180388,
+      "grad_norm": 3.8968212604522705,
+      "learning_rate": 1.148197855408631e-06,
+      "loss": 0.9078,
+      "step": 10203
+    },
+    {
+      "epoch": 0.7866173296330559,
+      "grad_norm": 4.120811939239502,
+      "learning_rate": 1.1474019547041848e-06,
+      "loss": 0.9516,
+      "step": 10204
+    },
+    {
+      "epoch": 0.7866944187480728,
+      "grad_norm": 3.669375419616699,
+      "learning_rate": 1.1466062941892754e-06,
+      "loss": 0.9964,
+      "step": 10205
+    },
+    {
+      "epoch": 0.7867715078630897,
+      "grad_norm": 3.683593988418579,
+      "learning_rate": 1.145810873913506e-06,
+      "loss": 0.8877,
+      "step": 10206
+    },
+    {
+      "epoch": 0.7868485969781067,
+      "grad_norm": 3.3355050086975098,
+      "learning_rate": 1.1450156939264689e-06,
+      "loss": 0.8196,
+      "step": 10207
+    },
+    {
+      "epoch": 0.7869256860931236,
+      "grad_norm": 3.5134360790252686,
+      "learning_rate": 1.144220754277736e-06,
+      "loss": 0.9914,
+      "step": 10208
+    },
+    {
+      "epoch": 0.7870027752081407,
+      "grad_norm": 4.333958625793457,
+      "learning_rate": 1.1434260550168702e-06,
+      "loss": 1.0483,
+      "step": 10209
+    },
+    {
+      "epoch": 0.7870798643231576,
+      "grad_norm": 3.63576340675354,
+      "learning_rate": 1.1426315961934159e-06,
+      "loss": 0.848,
+      "step": 10210
+    },
+    {
+      "epoch": 0.7871569534381745,
+      "grad_norm": 3.60412859916687,
+      "learning_rate": 1.1418373778569036e-06,
+      "loss": 0.8695,
+      "step": 10211
+    },
+    {
+      "epoch": 0.7872340425531915,
+      "grad_norm": 3.5103275775909424,
+      "learning_rate": 1.1410434000568487e-06,
+      "loss": 0.932,
+      "step": 10212
+    },
+    {
+      "epoch": 0.7873111316682084,
+      "grad_norm": 4.256019115447998,
+      "learning_rate": 1.1402496628427534e-06,
+      "loss": 0.9959,
+      "step": 10213
+    },
+    {
+      "epoch": 0.7873882207832255,
+      "grad_norm": 3.469325304031372,
+      "learning_rate": 1.1394561662641e-06,
+      "loss": 0.8449,
+      "step": 10214
+    },
+    {
+      "epoch": 0.7874653098982424,
+      "grad_norm": 3.975343942642212,
+      "learning_rate": 1.1386629103703606e-06,
+      "loss": 0.9372,
+      "step": 10215
+    },
+    {
+      "epoch": 0.7875423990132593,
+      "grad_norm": 3.5165517330169678,
+      "learning_rate": 1.13786989521099e-06,
+      "loss": 0.9793,
+      "step": 10216
+    },
+    {
+      "epoch": 0.7876194881282763,
+      "grad_norm": 3.727750062942505,
+      "learning_rate": 1.1370771208354291e-06,
+      "loss": 0.9962,
+      "step": 10217
+    },
+    {
+      "epoch": 0.7876965772432932,
+      "grad_norm": 3.6379904747009277,
+      "learning_rate": 1.1362845872931044e-06,
+      "loss": 0.8594,
+      "step": 10218
+    },
+    {
+      "epoch": 0.7877736663583103,
+      "grad_norm": 3.5502896308898926,
+      "learning_rate": 1.1354922946334241e-06,
+      "loss": 0.8649,
+      "step": 10219
+    },
+    {
+      "epoch": 0.7878507554733272,
+      "grad_norm": 3.6985366344451904,
+      "learning_rate": 1.1347002429057835e-06,
+      "loss": 0.9357,
+      "step": 10220
+    },
+    {
+      "epoch": 0.7879278445883441,
+      "grad_norm": 3.5729541778564453,
+      "learning_rate": 1.1339084321595644e-06,
+      "loss": 0.892,
+      "step": 10221
+    },
+    {
+      "epoch": 0.7880049337033611,
+      "grad_norm": 3.6966254711151123,
+      "learning_rate": 1.133116862444132e-06,
+      "loss": 0.9705,
+      "step": 10222
+    },
+    {
+      "epoch": 0.788082022818378,
+      "grad_norm": 3.8931562900543213,
+      "learning_rate": 1.1323255338088368e-06,
+      "loss": 0.8877,
+      "step": 10223
+    },
+    {
+      "epoch": 0.788159111933395,
+      "grad_norm": 3.7230522632598877,
+      "learning_rate": 1.131534446303012e-06,
+      "loss": 0.9154,
+      "step": 10224
+    },
+    {
+      "epoch": 0.788236201048412,
+      "grad_norm": 3.777541399002075,
+      "learning_rate": 1.1307435999759796e-06,
+      "loss": 0.8524,
+      "step": 10225
+    },
+    {
+      "epoch": 0.7883132901634289,
+      "grad_norm": 3.533332586288452,
+      "learning_rate": 1.1299529948770443e-06,
+      "loss": 0.9151,
+      "step": 10226
+    },
+    {
+      "epoch": 0.7883903792784459,
+      "grad_norm": 3.7804453372955322,
+      "learning_rate": 1.129162631055496e-06,
+      "loss": 0.8325,
+      "step": 10227
+    },
+    {
+      "epoch": 0.7884674683934628,
+      "grad_norm": 3.4261248111724854,
+      "learning_rate": 1.1283725085606101e-06,
+      "loss": 0.8707,
+      "step": 10228
+    },
+    {
+      "epoch": 0.7885445575084798,
+      "grad_norm": 3.9265434741973877,
+      "learning_rate": 1.1275826274416485e-06,
+      "loss": 0.9651,
+      "step": 10229
+    },
+    {
+      "epoch": 0.7886216466234968,
+      "grad_norm": 3.6662166118621826,
+      "learning_rate": 1.1267929877478522e-06,
+      "loss": 0.9031,
+      "step": 10230
+    },
+    {
+      "epoch": 0.7886987357385137,
+      "grad_norm": 3.6205952167510986,
+      "learning_rate": 1.1260035895284538e-06,
+      "loss": 0.935,
+      "step": 10231
+    },
+    {
+      "epoch": 0.7887758248535307,
+      "grad_norm": 3.919410467147827,
+      "learning_rate": 1.1252144328326676e-06,
+      "loss": 0.8459,
+      "step": 10232
+    },
+    {
+      "epoch": 0.7888529139685476,
+      "grad_norm": 3.570185422897339,
+      "learning_rate": 1.124425517709693e-06,
+      "loss": 0.846,
+      "step": 10233
+    },
+    {
+      "epoch": 0.7889300030835646,
+      "grad_norm": 3.781740427017212,
+      "learning_rate": 1.123636844208717e-06,
+      "loss": 0.8971,
+      "step": 10234
+    },
+    {
+      "epoch": 0.7890070921985816,
+      "grad_norm": 3.7370405197143555,
+      "learning_rate": 1.1228484123789064e-06,
+      "loss": 0.958,
+      "step": 10235
+    },
+    {
+      "epoch": 0.7890841813135985,
+      "grad_norm": 3.855154037475586,
+      "learning_rate": 1.1220602222694166e-06,
+      "loss": 0.9874,
+      "step": 10236
+    },
+    {
+      "epoch": 0.7891612704286155,
+      "grad_norm": 3.9142956733703613,
+      "learning_rate": 1.1212722739293875e-06,
+      "loss": 0.903,
+      "step": 10237
+    },
+    {
+      "epoch": 0.7892383595436324,
+      "grad_norm": 4.107082366943359,
+      "learning_rate": 1.120484567407944e-06,
+      "loss": 1.0201,
+      "step": 10238
+    },
+    {
+      "epoch": 0.7893154486586494,
+      "grad_norm": 3.537972927093506,
+      "learning_rate": 1.1196971027541953e-06,
+      "loss": 0.8097,
+      "step": 10239
+    },
+    {
+      "epoch": 0.7893925377736664,
+      "grad_norm": 3.690303087234497,
+      "learning_rate": 1.1189098800172365e-06,
+      "loss": 0.9472,
+      "step": 10240
+    },
+    {
+      "epoch": 0.7894696268886833,
+      "grad_norm": 4.062992572784424,
+      "learning_rate": 1.1181228992461451e-06,
+      "loss": 0.8588,
+      "step": 10241
+    },
+    {
+      "epoch": 0.7895467160037003,
+      "grad_norm": 3.6081042289733887,
+      "learning_rate": 1.1173361604899857e-06,
+      "loss": 0.8476,
+      "step": 10242
+    },
+    {
+      "epoch": 0.7896238051187172,
+      "grad_norm": 3.469123601913452,
+      "learning_rate": 1.1165496637978086e-06,
+      "loss": 0.8059,
+      "step": 10243
+    },
+    {
+      "epoch": 0.7897008942337342,
+      "grad_norm": 3.6479201316833496,
+      "learning_rate": 1.1157634092186464e-06,
+      "loss": 0.8624,
+      "step": 10244
+    },
+    {
+      "epoch": 0.7897779833487512,
+      "grad_norm": 3.684077024459839,
+      "learning_rate": 1.1149773968015205e-06,
+      "loss": 0.9347,
+      "step": 10245
+    },
+    {
+      "epoch": 0.7898550724637681,
+      "grad_norm": 4.0022382736206055,
+      "learning_rate": 1.114191626595431e-06,
+      "loss": 0.9736,
+      "step": 10246
+    },
+    {
+      "epoch": 0.7899321615787851,
+      "grad_norm": 3.8564565181732178,
+      "learning_rate": 1.1134060986493688e-06,
+      "loss": 0.834,
+      "step": 10247
+    },
+    {
+      "epoch": 0.790009250693802,
+      "grad_norm": 3.95080828666687,
+      "learning_rate": 1.1126208130123056e-06,
+      "loss": 0.8655,
+      "step": 10248
+    },
+    {
+      "epoch": 0.790086339808819,
+      "grad_norm": 3.4541101455688477,
+      "learning_rate": 1.1118357697332027e-06,
+      "loss": 0.8184,
+      "step": 10249
+    },
+    {
+      "epoch": 0.790163428923836,
+      "grad_norm": 3.7209436893463135,
+      "learning_rate": 1.1110509688610038e-06,
+      "loss": 0.8719,
+      "step": 10250
+    },
+    {
+      "epoch": 0.7902405180388529,
+      "grad_norm": 3.6410865783691406,
+      "learning_rate": 1.1102664104446342e-06,
+      "loss": 0.9483,
+      "step": 10251
+    },
+    {
+      "epoch": 0.7903176071538699,
+      "grad_norm": 3.3770508766174316,
+      "learning_rate": 1.1094820945330088e-06,
+      "loss": 0.7337,
+      "step": 10252
+    },
+    {
+      "epoch": 0.7903946962688868,
+      "grad_norm": 3.8358688354492188,
+      "learning_rate": 1.1086980211750247e-06,
+      "loss": 0.8971,
+      "step": 10253
+    },
+    {
+      "epoch": 0.7904717853839038,
+      "grad_norm": 3.432882308959961,
+      "learning_rate": 1.1079141904195662e-06,
+      "loss": 0.8461,
+      "step": 10254
+    },
+    {
+      "epoch": 0.7905488744989208,
+      "grad_norm": 4.271652698516846,
+      "learning_rate": 1.107130602315501e-06,
+      "loss": 1.0068,
+      "step": 10255
+    },
+    {
+      "epoch": 0.7906259636139377,
+      "grad_norm": 3.7254769802093506,
+      "learning_rate": 1.1063472569116802e-06,
+      "loss": 0.8175,
+      "step": 10256
+    },
+    {
+      "epoch": 0.7907030527289547,
+      "grad_norm": 3.4626779556274414,
+      "learning_rate": 1.1055641542569418e-06,
+      "loss": 0.7748,
+      "step": 10257
+    },
+    {
+      "epoch": 0.7907801418439716,
+      "grad_norm": 4.178032398223877,
+      "learning_rate": 1.1047812944001084e-06,
+      "loss": 0.9467,
+      "step": 10258
+    },
+    {
+      "epoch": 0.7908572309589886,
+      "grad_norm": 3.711949348449707,
+      "learning_rate": 1.103998677389988e-06,
+      "loss": 0.9139,
+      "step": 10259
+    },
+    {
+      "epoch": 0.7909343200740055,
+      "grad_norm": 4.120387554168701,
+      "learning_rate": 1.1032163032753717e-06,
+      "loss": 0.9809,
+      "step": 10260
+    },
+    {
+      "epoch": 0.7910114091890225,
+      "grad_norm": 3.4602792263031006,
+      "learning_rate": 1.1024341721050385e-06,
+      "loss": 0.9068,
+      "step": 10261
+    },
+    {
+      "epoch": 0.7910884983040395,
+      "grad_norm": 3.6511247158050537,
+      "learning_rate": 1.1016522839277471e-06,
+      "loss": 0.8453,
+      "step": 10262
+    },
+    {
+      "epoch": 0.7911655874190564,
+      "grad_norm": 4.048055648803711,
+      "learning_rate": 1.1008706387922457e-06,
+      "loss": 1.0161,
+      "step": 10263
+    },
+    {
+      "epoch": 0.7912426765340734,
+      "grad_norm": 3.4524850845336914,
+      "learning_rate": 1.100089236747266e-06,
+      "loss": 0.8793,
+      "step": 10264
+    },
+    {
+      "epoch": 0.7913197656490903,
+      "grad_norm": 3.5564897060394287,
+      "learning_rate": 1.0993080778415245e-06,
+      "loss": 0.8971,
+      "step": 10265
+    },
+    {
+      "epoch": 0.7913968547641073,
+      "grad_norm": 4.036877155303955,
+      "learning_rate": 1.0985271621237231e-06,
+      "loss": 0.91,
+      "step": 10266
+    },
+    {
+      "epoch": 0.7914739438791243,
+      "grad_norm": 3.743246078491211,
+      "learning_rate": 1.0977464896425461e-06,
+      "loss": 0.9205,
+      "step": 10267
+    },
+    {
+      "epoch": 0.7915510329941412,
+      "grad_norm": 3.4432597160339355,
+      "learning_rate": 1.0969660604466648e-06,
+      "loss": 0.8664,
+      "step": 10268
+    },
+    {
+      "epoch": 0.7916281221091582,
+      "grad_norm": 3.663029670715332,
+      "learning_rate": 1.0961858745847348e-06,
+      "loss": 0.9574,
+      "step": 10269
+    },
+    {
+      "epoch": 0.7917052112241751,
+      "grad_norm": 3.6875574588775635,
+      "learning_rate": 1.0954059321053978e-06,
+      "loss": 0.859,
+      "step": 10270
+    },
+    {
+      "epoch": 0.7917823003391921,
+      "grad_norm": 3.6623551845550537,
+      "learning_rate": 1.0946262330572798e-06,
+      "loss": 0.8892,
+      "step": 10271
+    },
+    {
+      "epoch": 0.7918593894542091,
+      "grad_norm": 3.574294328689575,
+      "learning_rate": 1.0938467774889883e-06,
+      "loss": 0.9423,
+      "step": 10272
+    },
+    {
+      "epoch": 0.791936478569226,
+      "grad_norm": 3.773029088973999,
+      "learning_rate": 1.0930675654491197e-06,
+      "loss": 0.8767,
+      "step": 10273
+    },
+    {
+      "epoch": 0.792013567684243,
+      "grad_norm": 3.5403318405151367,
+      "learning_rate": 1.092288596986254e-06,
+      "loss": 0.962,
+      "step": 10274
+    },
+    {
+      "epoch": 0.7920906567992599,
+      "grad_norm": 3.787517786026001,
+      "learning_rate": 1.0915098721489553e-06,
+      "loss": 0.9419,
+      "step": 10275
+    },
+    {
+      "epoch": 0.7921677459142769,
+      "grad_norm": 3.657186508178711,
+      "learning_rate": 1.0907313909857737e-06,
+      "loss": 0.9917,
+      "step": 10276
+    },
+    {
+      "epoch": 0.7922448350292939,
+      "grad_norm": 3.8778676986694336,
+      "learning_rate": 1.0899531535452452e-06,
+      "loss": 0.8329,
+      "step": 10277
+    },
+    {
+      "epoch": 0.7923219241443108,
+      "grad_norm": 4.028141975402832,
+      "learning_rate": 1.0891751598758849e-06,
+      "loss": 0.9051,
+      "step": 10278
+    },
+    {
+      "epoch": 0.7923990132593278,
+      "grad_norm": 3.770726442337036,
+      "learning_rate": 1.0883974100261985e-06,
+      "loss": 0.8382,
+      "step": 10279
+    },
+    {
+      "epoch": 0.7924761023743447,
+      "grad_norm": 3.7787578105926514,
+      "learning_rate": 1.0876199040446754e-06,
+      "loss": 0.8499,
+      "step": 10280
+    },
+    {
+      "epoch": 0.7925531914893617,
+      "grad_norm": 3.5443973541259766,
+      "learning_rate": 1.0868426419797883e-06,
+      "loss": 0.9412,
+      "step": 10281
+    },
+    {
+      "epoch": 0.7926302806043787,
+      "grad_norm": 3.392566680908203,
+      "learning_rate": 1.0860656238799971e-06,
+      "loss": 0.8903,
+      "step": 10282
+    },
+    {
+      "epoch": 0.7927073697193956,
+      "grad_norm": 3.8018105030059814,
+      "learning_rate": 1.0852888497937424e-06,
+      "loss": 0.8888,
+      "step": 10283
+    },
+    {
+      "epoch": 0.7927844588344126,
+      "grad_norm": 3.710118293762207,
+      "learning_rate": 1.0845123197694528e-06,
+      "loss": 1.0176,
+      "step": 10284
+    },
+    {
+      "epoch": 0.7928615479494295,
+      "grad_norm": 3.820878028869629,
+      "learning_rate": 1.0837360338555414e-06,
+      "loss": 0.9367,
+      "step": 10285
+    },
+    {
+      "epoch": 0.7929386370644464,
+      "grad_norm": 4.015862941741943,
+      "learning_rate": 1.0829599921004054e-06,
+      "loss": 0.8729,
+      "step": 10286
+    },
+    {
+      "epoch": 0.7930157261794635,
+      "grad_norm": 3.3819711208343506,
+      "learning_rate": 1.0821841945524265e-06,
+      "loss": 0.8165,
+      "step": 10287
+    },
+    {
+      "epoch": 0.7930928152944804,
+      "grad_norm": 4.338066101074219,
+      "learning_rate": 1.081408641259974e-06,
+      "loss": 1.0214,
+      "step": 10288
+    },
+    {
+      "epoch": 0.7931699044094974,
+      "grad_norm": 3.5678458213806152,
+      "learning_rate": 1.0806333322713964e-06,
+      "loss": 0.8869,
+      "step": 10289
+    },
+    {
+      "epoch": 0.7932469935245143,
+      "grad_norm": 4.043673515319824,
+      "learning_rate": 1.0798582676350316e-06,
+      "loss": 0.9494,
+      "step": 10290
+    },
+    {
+      "epoch": 0.7933240826395312,
+      "grad_norm": 3.7308008670806885,
+      "learning_rate": 1.0790834473992013e-06,
+      "loss": 0.7882,
+      "step": 10291
+    },
+    {
+      "epoch": 0.7934011717545483,
+      "grad_norm": 4.306490421295166,
+      "learning_rate": 1.0783088716122102e-06,
+      "loss": 0.995,
+      "step": 10292
+    },
+    {
+      "epoch": 0.7934782608695652,
+      "grad_norm": 3.6858410835266113,
+      "learning_rate": 1.0775345403223509e-06,
+      "loss": 0.8796,
+      "step": 10293
+    },
+    {
+      "epoch": 0.7935553499845822,
+      "grad_norm": 3.7026453018188477,
+      "learning_rate": 1.0767604535778976e-06,
+      "loss": 0.9955,
+      "step": 10294
+    },
+    {
+      "epoch": 0.7936324390995991,
+      "grad_norm": 3.596752882003784,
+      "learning_rate": 1.075986611427111e-06,
+      "loss": 0.8235,
+      "step": 10295
+    },
+    {
+      "epoch": 0.793709528214616,
+      "grad_norm": 3.608722448348999,
+      "learning_rate": 1.0752130139182364e-06,
+      "loss": 0.8598,
+      "step": 10296
+    },
+    {
+      "epoch": 0.7937866173296331,
+      "grad_norm": 3.686950922012329,
+      "learning_rate": 1.0744396610995033e-06,
+      "loss": 0.8529,
+      "step": 10297
+    },
+    {
+      "epoch": 0.79386370644465,
+      "grad_norm": 4.59524393081665,
+      "learning_rate": 1.0736665530191276e-06,
+      "loss": 1.1376,
+      "step": 10298
+    },
+    {
+      "epoch": 0.793940795559667,
+      "grad_norm": 3.7717888355255127,
+      "learning_rate": 1.072893689725306e-06,
+      "loss": 0.8147,
+      "step": 10299
+    },
+    {
+      "epoch": 0.7940178846746839,
+      "grad_norm": 3.5233025550842285,
+      "learning_rate": 1.072121071266224e-06,
+      "loss": 0.8799,
+      "step": 10300
+    },
+    {
+      "epoch": 0.7940949737897008,
+      "grad_norm": 3.7410104274749756,
+      "learning_rate": 1.07134869769005e-06,
+      "loss": 0.9286,
+      "step": 10301
+    },
+    {
+      "epoch": 0.7941720629047179,
+      "grad_norm": 3.774850606918335,
+      "learning_rate": 1.0705765690449376e-06,
+      "loss": 0.9598,
+      "step": 10302
+    },
+    {
+      "epoch": 0.7942491520197348,
+      "grad_norm": 4.010867595672607,
+      "learning_rate": 1.0698046853790268e-06,
+      "loss": 0.9681,
+      "step": 10303
+    },
+    {
+      "epoch": 0.7943262411347518,
+      "grad_norm": 3.8371970653533936,
+      "learning_rate": 1.0690330467404375e-06,
+      "loss": 0.9313,
+      "step": 10304
+    },
+    {
+      "epoch": 0.7944033302497687,
+      "grad_norm": 3.689708709716797,
+      "learning_rate": 1.0682616531772782e-06,
+      "loss": 0.9279,
+      "step": 10305
+    },
+    {
+      "epoch": 0.7944804193647856,
+      "grad_norm": 3.87109112739563,
+      "learning_rate": 1.0674905047376423e-06,
+      "loss": 0.8764,
+      "step": 10306
+    },
+    {
+      "epoch": 0.7945575084798027,
+      "grad_norm": 3.5622525215148926,
+      "learning_rate": 1.0667196014696062e-06,
+      "loss": 0.9274,
+      "step": 10307
+    },
+    {
+      "epoch": 0.7946345975948196,
+      "grad_norm": 3.48157000541687,
+      "learning_rate": 1.0659489434212323e-06,
+      "loss": 0.8604,
+      "step": 10308
+    },
+    {
+      "epoch": 0.7947116867098366,
+      "grad_norm": 3.320413112640381,
+      "learning_rate": 1.0651785306405683e-06,
+      "loss": 0.803,
+      "step": 10309
+    },
+    {
+      "epoch": 0.7947887758248535,
+      "grad_norm": 3.5544357299804688,
+      "learning_rate": 1.064408363175642e-06,
+      "loss": 0.9412,
+      "step": 10310
+    },
+    {
+      "epoch": 0.7948658649398704,
+      "grad_norm": 3.2090461254119873,
+      "learning_rate": 1.0636384410744716e-06,
+      "loss": 0.8046,
+      "step": 10311
+    },
+    {
+      "epoch": 0.7949429540548875,
+      "grad_norm": 3.8545830249786377,
+      "learning_rate": 1.0628687643850572e-06,
+      "loss": 0.9692,
+      "step": 10312
+    },
+    {
+      "epoch": 0.7950200431699044,
+      "grad_norm": 3.9732425212860107,
+      "learning_rate": 1.0620993331553848e-06,
+      "loss": 0.8421,
+      "step": 10313
+    },
+    {
+      "epoch": 0.7950971322849214,
+      "grad_norm": 3.6110129356384277,
+      "learning_rate": 1.0613301474334254e-06,
+      "loss": 0.943,
+      "step": 10314
+    },
+    {
+      "epoch": 0.7951742213999383,
+      "grad_norm": 3.558922529220581,
+      "learning_rate": 1.0605612072671305e-06,
+      "loss": 0.8885,
+      "step": 10315
+    },
+    {
+      "epoch": 0.7952513105149552,
+      "grad_norm": 3.5651447772979736,
+      "learning_rate": 1.0597925127044423e-06,
+      "loss": 0.8734,
+      "step": 10316
+    },
+    {
+      "epoch": 0.7953283996299723,
+      "grad_norm": 3.470515251159668,
+      "learning_rate": 1.0590240637932835e-06,
+      "loss": 0.7586,
+      "step": 10317
+    },
+    {
+      "epoch": 0.7954054887449892,
+      "grad_norm": 3.696683883666992,
+      "learning_rate": 1.0582558605815636e-06,
+      "loss": 0.8775,
+      "step": 10318
+    },
+    {
+      "epoch": 0.7954825778600062,
+      "grad_norm": 3.392326593399048,
+      "learning_rate": 1.0574879031171776e-06,
+      "loss": 0.7721,
+      "step": 10319
+    },
+    {
+      "epoch": 0.7955596669750231,
+      "grad_norm": 4.037016868591309,
+      "learning_rate": 1.0567201914480002e-06,
+      "loss": 0.8883,
+      "step": 10320
+    },
+    {
+      "epoch": 0.79563675609004,
+      "grad_norm": 3.5130562782287598,
+      "learning_rate": 1.0559527256218959e-06,
+      "loss": 0.861,
+      "step": 10321
+    },
+    {
+      "epoch": 0.7957138452050571,
+      "grad_norm": 3.640817165374756,
+      "learning_rate": 1.0551855056867132e-06,
+      "loss": 0.9281,
+      "step": 10322
+    },
+    {
+      "epoch": 0.795790934320074,
+      "grad_norm": 4.14249324798584,
+      "learning_rate": 1.0544185316902834e-06,
+      "loss": 0.9328,
+      "step": 10323
+    },
+    {
+      "epoch": 0.795868023435091,
+      "grad_norm": 3.433025598526001,
+      "learning_rate": 1.0536518036804228e-06,
+      "loss": 0.8419,
+      "step": 10324
+    },
+    {
+      "epoch": 0.7959451125501079,
+      "grad_norm": 3.647221565246582,
+      "learning_rate": 1.0528853217049357e-06,
+      "loss": 0.9425,
+      "step": 10325
+    },
+    {
+      "epoch": 0.7960222016651248,
+      "grad_norm": 3.8280515670776367,
+      "learning_rate": 1.0521190858116042e-06,
+      "loss": 0.9141,
+      "step": 10326
+    },
+    {
+      "epoch": 0.7960992907801419,
+      "grad_norm": 3.922826051712036,
+      "learning_rate": 1.0513530960482016e-06,
+      "loss": 1.0064,
+      "step": 10327
+    },
+    {
+      "epoch": 0.7961763798951588,
+      "grad_norm": 3.8651580810546875,
+      "learning_rate": 1.0505873524624821e-06,
+      "loss": 1.0625,
+      "step": 10328
+    },
+    {
+      "epoch": 0.7962534690101758,
+      "grad_norm": 3.5953264236450195,
+      "learning_rate": 1.0498218551021876e-06,
+      "loss": 0.8368,
+      "step": 10329
+    },
+    {
+      "epoch": 0.7963305581251927,
+      "grad_norm": 3.8444368839263916,
+      "learning_rate": 1.0490566040150428e-06,
+      "loss": 0.9272,
+      "step": 10330
+    },
+    {
+      "epoch": 0.7964076472402097,
+      "grad_norm": 3.7892086505889893,
+      "learning_rate": 1.0482915992487546e-06,
+      "loss": 0.8233,
+      "step": 10331
+    },
+    {
+      "epoch": 0.7964847363552267,
+      "grad_norm": 3.583220958709717,
+      "learning_rate": 1.0475268408510192e-06,
+      "loss": 1.0282,
+      "step": 10332
+    },
+    {
+      "epoch": 0.7965618254702436,
+      "grad_norm": 3.683847427368164,
+      "learning_rate": 1.046762328869515e-06,
+      "loss": 0.8625,
+      "step": 10333
+    },
+    {
+      "epoch": 0.7966389145852606,
+      "grad_norm": 3.9110031127929688,
+      "learning_rate": 1.0459980633519052e-06,
+      "loss": 0.9429,
+      "step": 10334
+    },
+    {
+      "epoch": 0.7967160037002775,
+      "grad_norm": 3.5115268230438232,
+      "learning_rate": 1.0452340443458376e-06,
+      "loss": 0.8812,
+      "step": 10335
+    },
+    {
+      "epoch": 0.7967930928152945,
+      "grad_norm": 3.59769868850708,
+      "learning_rate": 1.0444702718989452e-06,
+      "loss": 0.8564,
+      "step": 10336
+    },
+    {
+      "epoch": 0.7968701819303114,
+      "grad_norm": 3.9897191524505615,
+      "learning_rate": 1.043706746058845e-06,
+      "loss": 0.8961,
+      "step": 10337
+    },
+    {
+      "epoch": 0.7969472710453284,
+      "grad_norm": 3.6812877655029297,
+      "learning_rate": 1.0429434668731393e-06,
+      "loss": 1.0195,
+      "step": 10338
+    },
+    {
+      "epoch": 0.7970243601603454,
+      "grad_norm": 4.031727313995361,
+      "learning_rate": 1.0421804343894142e-06,
+      "loss": 0.9039,
+      "step": 10339
+    },
+    {
+      "epoch": 0.7971014492753623,
+      "grad_norm": 4.048935413360596,
+      "learning_rate": 1.0414176486552424e-06,
+      "loss": 0.8887,
+      "step": 10340
+    },
+    {
+      "epoch": 0.7971785383903793,
+      "grad_norm": 3.7354636192321777,
+      "learning_rate": 1.0406551097181765e-06,
+      "loss": 0.9092,
+      "step": 10341
+    },
+    {
+      "epoch": 0.7972556275053962,
+      "grad_norm": 3.63582706451416,
+      "learning_rate": 1.0398928176257588e-06,
+      "loss": 0.9253,
+      "step": 10342
+    },
+    {
+      "epoch": 0.7973327166204132,
+      "grad_norm": 3.9569201469421387,
+      "learning_rate": 1.039130772425514e-06,
+      "loss": 1.0315,
+      "step": 10343
+    },
+    {
+      "epoch": 0.7974098057354302,
+      "grad_norm": 3.3830461502075195,
+      "learning_rate": 1.0383689741649516e-06,
+      "loss": 0.8325,
+      "step": 10344
+    },
+    {
+      "epoch": 0.7974868948504471,
+      "grad_norm": 4.122878074645996,
+      "learning_rate": 1.0376074228915662e-06,
+      "loss": 1.0077,
+      "step": 10345
+    },
+    {
+      "epoch": 0.7975639839654641,
+      "grad_norm": 3.5504045486450195,
+      "learning_rate": 1.0368461186528366e-06,
+      "loss": 0.9757,
+      "step": 10346
+    },
+    {
+      "epoch": 0.797641073080481,
+      "grad_norm": 3.6346943378448486,
+      "learning_rate": 1.0360850614962254e-06,
+      "loss": 0.7473,
+      "step": 10347
+    },
+    {
+      "epoch": 0.797718162195498,
+      "grad_norm": 3.489760637283325,
+      "learning_rate": 1.0353242514691807e-06,
+      "loss": 0.8558,
+      "step": 10348
+    },
+    {
+      "epoch": 0.797795251310515,
+      "grad_norm": 4.047338485717773,
+      "learning_rate": 1.0345636886191351e-06,
+      "loss": 0.9747,
+      "step": 10349
+    },
+    {
+      "epoch": 0.7978723404255319,
+      "grad_norm": 3.327815055847168,
+      "learning_rate": 1.0338033729935064e-06,
+      "loss": 0.9214,
+      "step": 10350
+    },
+    {
+      "epoch": 0.7979494295405489,
+      "grad_norm": 3.7090442180633545,
+      "learning_rate": 1.0330433046396971e-06,
+      "loss": 0.9179,
+      "step": 10351
+    },
+    {
+      "epoch": 0.7980265186555658,
+      "grad_norm": 3.697265863418579,
+      "learning_rate": 1.032283483605091e-06,
+      "loss": 0.8827,
+      "step": 10352
+    },
+    {
+      "epoch": 0.7981036077705828,
+      "grad_norm": 3.3949902057647705,
+      "learning_rate": 1.03152390993706e-06,
+      "loss": 0.9044,
+      "step": 10353
+    },
+    {
+      "epoch": 0.7981806968855998,
+      "grad_norm": 3.660855531692505,
+      "learning_rate": 1.0307645836829604e-06,
+      "loss": 0.9089,
+      "step": 10354
+    },
+    {
+      "epoch": 0.7982577860006167,
+      "grad_norm": 3.7583930492401123,
+      "learning_rate": 1.030005504890132e-06,
+      "loss": 0.9184,
+      "step": 10355
+    },
+    {
+      "epoch": 0.7983348751156337,
+      "grad_norm": 4.092621803283691,
+      "learning_rate": 1.0292466736058988e-06,
+      "loss": 0.919,
+      "step": 10356
+    },
+    {
+      "epoch": 0.7984119642306506,
+      "grad_norm": 3.7729806900024414,
+      "learning_rate": 1.0284880898775717e-06,
+      "loss": 0.8264,
+      "step": 10357
+    },
+    {
+      "epoch": 0.7984890533456676,
+      "grad_norm": 3.7292611598968506,
+      "learning_rate": 1.0277297537524422e-06,
+      "loss": 0.7999,
+      "step": 10358
+    },
+    {
+      "epoch": 0.7985661424606846,
+      "grad_norm": 3.5362532138824463,
+      "learning_rate": 1.0269716652777894e-06,
+      "loss": 0.9376,
+      "step": 10359
+    },
+    {
+      "epoch": 0.7986432315757015,
+      "grad_norm": 3.8130404949188232,
+      "learning_rate": 1.0262138245008768e-06,
+      "loss": 0.9412,
+      "step": 10360
+    },
+    {
+      "epoch": 0.7987203206907185,
+      "grad_norm": 3.703113555908203,
+      "learning_rate": 1.0254562314689508e-06,
+      "loss": 0.9883,
+      "step": 10361
+    },
+    {
+      "epoch": 0.7987974098057354,
+      "grad_norm": 3.831934928894043,
+      "learning_rate": 1.0246988862292462e-06,
+      "loss": 0.8621,
+      "step": 10362
+    },
+    {
+      "epoch": 0.7988744989207524,
+      "grad_norm": 3.592346668243408,
+      "learning_rate": 1.023941788828976e-06,
+      "loss": 0.9581,
+      "step": 10363
+    },
+    {
+      "epoch": 0.7989515880357694,
+      "grad_norm": 3.945913553237915,
+      "learning_rate": 1.023184939315342e-06,
+      "loss": 0.9714,
+      "step": 10364
+    },
+    {
+      "epoch": 0.7990286771507863,
+      "grad_norm": 3.960003614425659,
+      "learning_rate": 1.0224283377355316e-06,
+      "loss": 0.9178,
+      "step": 10365
+    },
+    {
+      "epoch": 0.7991057662658033,
+      "grad_norm": 3.8358631134033203,
+      "learning_rate": 1.021671984136713e-06,
+      "loss": 0.9071,
+      "step": 10366
+    },
+    {
+      "epoch": 0.7991828553808202,
+      "grad_norm": 3.6046762466430664,
+      "learning_rate": 1.020915878566044e-06,
+      "loss": 0.8723,
+      "step": 10367
+    },
+    {
+      "epoch": 0.7992599444958371,
+      "grad_norm": 3.6357581615448,
+      "learning_rate": 1.0201600210706596e-06,
+      "loss": 0.8464,
+      "step": 10368
+    },
+    {
+      "epoch": 0.7993370336108542,
+      "grad_norm": 3.683241605758667,
+      "learning_rate": 1.0194044116976864e-06,
+      "loss": 0.9561,
+      "step": 10369
+    },
+    {
+      "epoch": 0.7994141227258711,
+      "grad_norm": 4.788916110992432,
+      "learning_rate": 1.018649050494232e-06,
+      "loss": 1.021,
+      "step": 10370
+    },
+    {
+      "epoch": 0.7994912118408881,
+      "grad_norm": 4.036243438720703,
+      "learning_rate": 1.0178939375073892e-06,
+      "loss": 1.0098,
+      "step": 10371
+    },
+    {
+      "epoch": 0.799568300955905,
+      "grad_norm": 3.611304998397827,
+      "learning_rate": 1.0171390727842357e-06,
+      "loss": 0.8585,
+      "step": 10372
+    },
+    {
+      "epoch": 0.799645390070922,
+      "grad_norm": 3.9461710453033447,
+      "learning_rate": 1.0163844563718344e-06,
+      "loss": 0.9371,
+      "step": 10373
+    },
+    {
+      "epoch": 0.799722479185939,
+      "grad_norm": 3.3774936199188232,
+      "learning_rate": 1.0156300883172292e-06,
+      "loss": 0.8,
+      "step": 10374
+    },
+    {
+      "epoch": 0.7997995683009559,
+      "grad_norm": 3.748626947402954,
+      "learning_rate": 1.0148759686674532e-06,
+      "loss": 0.8262,
+      "step": 10375
+    },
+    {
+      "epoch": 0.7998766574159729,
+      "grad_norm": 3.4377031326293945,
+      "learning_rate": 1.0141220974695199e-06,
+      "loss": 0.794,
+      "step": 10376
+    },
+    {
+      "epoch": 0.7999537465309898,
+      "grad_norm": 3.6376664638519287,
+      "learning_rate": 1.0133684747704314e-06,
+      "loss": 0.8805,
+      "step": 10377
+    },
+    {
+      "epoch": 0.8000308356460067,
+      "grad_norm": 3.6229360103607178,
+      "learning_rate": 1.012615100617172e-06,
+      "loss": 0.8017,
+      "step": 10378
+    },
+    {
+      "epoch": 0.8001079247610238,
+      "grad_norm": 3.5345287322998047,
+      "learning_rate": 1.0118619750567082e-06,
+      "loss": 0.8068,
+      "step": 10379
+    },
+    {
+      "epoch": 0.8001850138760407,
+      "grad_norm": 3.6617918014526367,
+      "learning_rate": 1.0111090981359961e-06,
+      "loss": 1.0157,
+      "step": 10380
+    },
+    {
+      "epoch": 0.8002621029910577,
+      "grad_norm": 3.9429585933685303,
+      "learning_rate": 1.0103564699019707e-06,
+      "loss": 0.9599,
+      "step": 10381
+    },
+    {
+      "epoch": 0.8003391921060746,
+      "grad_norm": 3.5763978958129883,
+      "learning_rate": 1.009604090401558e-06,
+      "loss": 0.9006,
+      "step": 10382
+    },
+    {
+      "epoch": 0.8004162812210915,
+      "grad_norm": 3.58232045173645,
+      "learning_rate": 1.008851959681665e-06,
+      "loss": 0.8655,
+      "step": 10383
+    },
+    {
+      "epoch": 0.8004933703361086,
+      "grad_norm": 3.727440118789673,
+      "learning_rate": 1.0081000777891803e-06,
+      "loss": 0.9123,
+      "step": 10384
+    },
+    {
+      "epoch": 0.8005704594511255,
+      "grad_norm": 3.6381850242614746,
+      "learning_rate": 1.0073484447709809e-06,
+      "loss": 0.8735,
+      "step": 10385
+    },
+    {
+      "epoch": 0.8006475485661425,
+      "grad_norm": 3.738128662109375,
+      "learning_rate": 1.0065970606739273e-06,
+      "loss": 0.8284,
+      "step": 10386
+    },
+    {
+      "epoch": 0.8007246376811594,
+      "grad_norm": 3.4860498905181885,
+      "learning_rate": 1.005845925544865e-06,
+      "loss": 0.8732,
+      "step": 10387
+    },
+    {
+      "epoch": 0.8008017267961763,
+      "grad_norm": 3.461859703063965,
+      "learning_rate": 1.0050950394306242e-06,
+      "loss": 0.9177,
+      "step": 10388
+    },
+    {
+      "epoch": 0.8008788159111934,
+      "grad_norm": 3.968186855316162,
+      "learning_rate": 1.0043444023780164e-06,
+      "loss": 0.9015,
+      "step": 10389
+    },
+    {
+      "epoch": 0.8009559050262103,
+      "grad_norm": 3.562736749649048,
+      "learning_rate": 1.0035940144338406e-06,
+      "loss": 0.7696,
+      "step": 10390
+    },
+    {
+      "epoch": 0.8010329941412273,
+      "grad_norm": 4.058101654052734,
+      "learning_rate": 1.0028438756448805e-06,
+      "loss": 0.9197,
+      "step": 10391
+    },
+    {
+      "epoch": 0.8011100832562442,
+      "grad_norm": 4.087806224822998,
+      "learning_rate": 1.0020939860579033e-06,
+      "loss": 0.9222,
+      "step": 10392
+    },
+    {
+      "epoch": 0.8011871723712611,
+      "grad_norm": 3.800830364227295,
+      "learning_rate": 1.0013443457196598e-06,
+      "loss": 1.0234,
+      "step": 10393
+    },
+    {
+      "epoch": 0.8012642614862782,
+      "grad_norm": 3.52839994430542,
+      "learning_rate": 1.0005949546768879e-06,
+      "loss": 0.8291,
+      "step": 10394
+    },
+    {
+      "epoch": 0.8013413506012951,
+      "grad_norm": 3.8631598949432373,
+      "learning_rate": 9.998458129763062e-07,
+      "loss": 0.8932,
+      "step": 10395
+    },
+    {
+      "epoch": 0.8014184397163121,
+      "grad_norm": 3.8592798709869385,
+      "learning_rate": 9.990969206646205e-07,
+      "loss": 1.0186,
+      "step": 10396
+    },
+    {
+      "epoch": 0.801495528831329,
+      "grad_norm": 3.7347230911254883,
+      "learning_rate": 9.983482777885211e-07,
+      "loss": 0.9141,
+      "step": 10397
+    },
+    {
+      "epoch": 0.8015726179463459,
+      "grad_norm": 3.569770574569702,
+      "learning_rate": 9.975998843946811e-07,
+      "loss": 0.8909,
+      "step": 10398
+    },
+    {
+      "epoch": 0.801649707061363,
+      "grad_norm": 3.699913740158081,
+      "learning_rate": 9.968517405297607e-07,
+      "loss": 0.916,
+      "step": 10399
+    },
+    {
+      "epoch": 0.8017267961763799,
+      "grad_norm": 3.733384609222412,
+      "learning_rate": 9.961038462403999e-07,
+      "loss": 0.8737,
+      "step": 10400
+    },
+    {
+      "epoch": 0.8018038852913969,
+      "grad_norm": 3.7092156410217285,
+      "learning_rate": 9.953562015732281e-07,
+      "loss": 0.9746,
+      "step": 10401
+    },
+    {
+      "epoch": 0.8018809744064138,
+      "grad_norm": 3.567920446395874,
+      "learning_rate": 9.94608806574856e-07,
+      "loss": 0.8904,
+      "step": 10402
+    },
+    {
+      "epoch": 0.8019580635214307,
+      "grad_norm": 3.6356654167175293,
+      "learning_rate": 9.93861661291881e-07,
+      "loss": 0.9383,
+      "step": 10403
+    },
+    {
+      "epoch": 0.8020351526364478,
+      "grad_norm": 3.344292640686035,
+      "learning_rate": 9.931147657708823e-07,
+      "loss": 0.7388,
+      "step": 10404
+    },
+    {
+      "epoch": 0.8021122417514647,
+      "grad_norm": 3.598586082458496,
+      "learning_rate": 9.92368120058428e-07,
+      "loss": 0.7381,
+      "step": 10405
+    },
+    {
+      "epoch": 0.8021893308664817,
+      "grad_norm": 3.6138226985931396,
+      "learning_rate": 9.916217242010634e-07,
+      "loss": 0.8098,
+      "step": 10406
+    },
+    {
+      "epoch": 0.8022664199814986,
+      "grad_norm": 3.5014567375183105,
+      "learning_rate": 9.908755782453245e-07,
+      "loss": 0.8243,
+      "step": 10407
+    },
+    {
+      "epoch": 0.8023435090965155,
+      "grad_norm": 3.7613766193389893,
+      "learning_rate": 9.901296822377293e-07,
+      "loss": 0.8441,
+      "step": 10408
+    },
+    {
+      "epoch": 0.8024205982115326,
+      "grad_norm": 3.4418587684631348,
+      "learning_rate": 9.893840362247809e-07,
+      "loss": 0.8912,
+      "step": 10409
+    },
+    {
+      "epoch": 0.8024976873265495,
+      "grad_norm": 3.7355403900146484,
+      "learning_rate": 9.88638640252968e-07,
+      "loss": 0.9684,
+      "step": 10410
+    },
+    {
+      "epoch": 0.8025747764415665,
+      "grad_norm": 3.9282965660095215,
+      "learning_rate": 9.87893494368759e-07,
+      "loss": 0.9324,
+      "step": 10411
+    },
+    {
+      "epoch": 0.8026518655565834,
+      "grad_norm": 3.668252944946289,
+      "learning_rate": 9.871485986186114e-07,
+      "loss": 0.9229,
+      "step": 10412
+    },
+    {
+      "epoch": 0.8027289546716003,
+      "grad_norm": 3.8640308380126953,
+      "learning_rate": 9.86403953048965e-07,
+      "loss": 0.9495,
+      "step": 10413
+    },
+    {
+      "epoch": 0.8028060437866174,
+      "grad_norm": 3.8201818466186523,
+      "learning_rate": 9.856595577062456e-07,
+      "loss": 0.8985,
+      "step": 10414
+    },
+    {
+      "epoch": 0.8028831329016343,
+      "grad_norm": 3.5459721088409424,
+      "learning_rate": 9.849154126368638e-07,
+      "loss": 0.9009,
+      "step": 10415
+    },
+    {
+      "epoch": 0.8029602220166513,
+      "grad_norm": 3.3526997566223145,
+      "learning_rate": 9.841715178872092e-07,
+      "loss": 0.8461,
+      "step": 10416
+    },
+    {
+      "epoch": 0.8030373111316682,
+      "grad_norm": 3.9801721572875977,
+      "learning_rate": 9.834278735036623e-07,
+      "loss": 0.9436,
+      "step": 10417
+    },
+    {
+      "epoch": 0.8031144002466851,
+      "grad_norm": 3.7711634635925293,
+      "learning_rate": 9.826844795325852e-07,
+      "loss": 1.0106,
+      "step": 10418
+    },
+    {
+      "epoch": 0.8031914893617021,
+      "grad_norm": 3.627129316329956,
+      "learning_rate": 9.819413360203244e-07,
+      "loss": 1.0238,
+      "step": 10419
+    },
+    {
+      "epoch": 0.8032685784767191,
+      "grad_norm": 3.4935200214385986,
+      "learning_rate": 9.811984430132116e-07,
+      "loss": 0.7833,
+      "step": 10420
+    },
+    {
+      "epoch": 0.8033456675917361,
+      "grad_norm": 3.4968650341033936,
+      "learning_rate": 9.804558005575632e-07,
+      "loss": 0.9387,
+      "step": 10421
+    },
+    {
+      "epoch": 0.803422756706753,
+      "grad_norm": 3.8687143325805664,
+      "learning_rate": 9.797134086996762e-07,
+      "loss": 0.8498,
+      "step": 10422
+    },
+    {
+      "epoch": 0.8034998458217699,
+      "grad_norm": 3.631547451019287,
+      "learning_rate": 9.78971267485837e-07,
+      "loss": 0.8742,
+      "step": 10423
+    },
+    {
+      "epoch": 0.803576934936787,
+      "grad_norm": 3.5340750217437744,
+      "learning_rate": 9.782293769623136e-07,
+      "loss": 0.9018,
+      "step": 10424
+    },
+    {
+      "epoch": 0.8036540240518039,
+      "grad_norm": 3.6497690677642822,
+      "learning_rate": 9.774877371753594e-07,
+      "loss": 0.8692,
+      "step": 10425
+    },
+    {
+      "epoch": 0.8037311131668209,
+      "grad_norm": 4.077952861785889,
+      "learning_rate": 9.767463481712113e-07,
+      "loss": 0.867,
+      "step": 10426
+    },
+    {
+      "epoch": 0.8038082022818378,
+      "grad_norm": 3.7168338298797607,
+      "learning_rate": 9.760052099960921e-07,
+      "loss": 0.8681,
+      "step": 10427
+    },
+    {
+      "epoch": 0.8038852913968547,
+      "grad_norm": 3.6925764083862305,
+      "learning_rate": 9.752643226962066e-07,
+      "loss": 0.9753,
+      "step": 10428
+    },
+    {
+      "epoch": 0.8039623805118717,
+      "grad_norm": 3.3511204719543457,
+      "learning_rate": 9.745236863177465e-07,
+      "loss": 0.8573,
+      "step": 10429
+    },
+    {
+      "epoch": 0.8040394696268887,
+      "grad_norm": 3.528261661529541,
+      "learning_rate": 9.737833009068859e-07,
+      "loss": 0.958,
+      "step": 10430
+    },
+    {
+      "epoch": 0.8041165587419057,
+      "grad_norm": 3.954437255859375,
+      "learning_rate": 9.73043166509785e-07,
+      "loss": 1.0196,
+      "step": 10431
+    },
+    {
+      "epoch": 0.8041936478569226,
+      "grad_norm": 3.7125864028930664,
+      "learning_rate": 9.723032831725859e-07,
+      "loss": 0.9288,
+      "step": 10432
+    },
+    {
+      "epoch": 0.8042707369719395,
+      "grad_norm": 3.7825093269348145,
+      "learning_rate": 9.715636509414168e-07,
+      "loss": 0.967,
+      "step": 10433
+    },
+    {
+      "epoch": 0.8043478260869565,
+      "grad_norm": 3.930483818054199,
+      "learning_rate": 9.708242698623898e-07,
+      "loss": 0.8991,
+      "step": 10434
+    },
+    {
+      "epoch": 0.8044249152019735,
+      "grad_norm": 3.8229269981384277,
+      "learning_rate": 9.700851399816026e-07,
+      "loss": 0.8454,
+      "step": 10435
+    },
+    {
+      "epoch": 0.8045020043169905,
+      "grad_norm": 3.621720790863037,
+      "learning_rate": 9.693462613451365e-07,
+      "loss": 0.9305,
+      "step": 10436
+    },
+    {
+      "epoch": 0.8045790934320074,
+      "grad_norm": 3.6314635276794434,
+      "learning_rate": 9.686076339990546e-07,
+      "loss": 0.8921,
+      "step": 10437
+    },
+    {
+      "epoch": 0.8046561825470243,
+      "grad_norm": 3.537757635116577,
+      "learning_rate": 9.678692579894072e-07,
+      "loss": 0.9057,
+      "step": 10438
+    },
+    {
+      "epoch": 0.8047332716620413,
+      "grad_norm": 3.6594655513763428,
+      "learning_rate": 9.671311333622292e-07,
+      "loss": 0.9033,
+      "step": 10439
+    },
+    {
+      "epoch": 0.8048103607770583,
+      "grad_norm": 4.193519592285156,
+      "learning_rate": 9.663932601635378e-07,
+      "loss": 0.9157,
+      "step": 10440
+    },
+    {
+      "epoch": 0.8048874498920753,
+      "grad_norm": 3.8593950271606445,
+      "learning_rate": 9.656556384393362e-07,
+      "loss": 0.9297,
+      "step": 10441
+    },
+    {
+      "epoch": 0.8049645390070922,
+      "grad_norm": 3.904437780380249,
+      "learning_rate": 9.649182682356122e-07,
+      "loss": 0.8452,
+      "step": 10442
+    },
+    {
+      "epoch": 0.8050416281221091,
+      "grad_norm": 4.147661209106445,
+      "learning_rate": 9.64181149598335e-07,
+      "loss": 0.9162,
+      "step": 10443
+    },
+    {
+      "epoch": 0.8051187172371261,
+      "grad_norm": 4.427166938781738,
+      "learning_rate": 9.634442825734609e-07,
+      "loss": 0.919,
+      "step": 10444
+    },
+    {
+      "epoch": 0.805195806352143,
+      "grad_norm": 3.7740793228149414,
+      "learning_rate": 9.627076672069302e-07,
+      "loss": 0.8899,
+      "step": 10445
+    },
+    {
+      "epoch": 0.8052728954671601,
+      "grad_norm": 3.5739588737487793,
+      "learning_rate": 9.619713035446666e-07,
+      "loss": 0.9004,
+      "step": 10446
+    },
+    {
+      "epoch": 0.805349984582177,
+      "grad_norm": 3.8233814239501953,
+      "learning_rate": 9.612351916325795e-07,
+      "loss": 0.9768,
+      "step": 10447
+    },
+    {
+      "epoch": 0.8054270736971939,
+      "grad_norm": 4.079464912414551,
+      "learning_rate": 9.604993315165607e-07,
+      "loss": 1.0417,
+      "step": 10448
+    },
+    {
+      "epoch": 0.8055041628122109,
+      "grad_norm": 3.4598426818847656,
+      "learning_rate": 9.597637232424866e-07,
+      "loss": 0.8658,
+      "step": 10449
+    },
+    {
+      "epoch": 0.8055812519272278,
+      "grad_norm": 3.5878536701202393,
+      "learning_rate": 9.590283668562195e-07,
+      "loss": 0.8726,
+      "step": 10450
+    },
+    {
+      "epoch": 0.8056583410422449,
+      "grad_norm": 3.5726075172424316,
+      "learning_rate": 9.582932624036052e-07,
+      "loss": 0.8884,
+      "step": 10451
+    },
+    {
+      "epoch": 0.8057354301572618,
+      "grad_norm": 3.6233930587768555,
+      "learning_rate": 9.575584099304735e-07,
+      "loss": 0.8807,
+      "step": 10452
+    },
+    {
+      "epoch": 0.8058125192722787,
+      "grad_norm": 3.603677272796631,
+      "learning_rate": 9.5682380948264e-07,
+      "loss": 0.9608,
+      "step": 10453
+    },
+    {
+      "epoch": 0.8058896083872957,
+      "grad_norm": 3.6978039741516113,
+      "learning_rate": 9.560894611059001e-07,
+      "loss": 0.906,
+      "step": 10454
+    },
+    {
+      "epoch": 0.8059666975023126,
+      "grad_norm": 3.5961194038391113,
+      "learning_rate": 9.55355364846039e-07,
+      "loss": 0.9432,
+      "step": 10455
+    },
+    {
+      "epoch": 0.8060437866173297,
+      "grad_norm": 3.6993865966796875,
+      "learning_rate": 9.546215207488225e-07,
+      "loss": 0.9392,
+      "step": 10456
+    },
+    {
+      "epoch": 0.8061208757323466,
+      "grad_norm": 4.084522724151611,
+      "learning_rate": 9.538879288600028e-07,
+      "loss": 0.8663,
+      "step": 10457
+    },
+    {
+      "epoch": 0.8061979648473635,
+      "grad_norm": 3.5728228092193604,
+      "learning_rate": 9.531545892253169e-07,
+      "loss": 0.9397,
+      "step": 10458
+    },
+    {
+      "epoch": 0.8062750539623805,
+      "grad_norm": 3.443225383758545,
+      "learning_rate": 9.52421501890482e-07,
+      "loss": 0.8319,
+      "step": 10459
+    },
+    {
+      "epoch": 0.8063521430773974,
+      "grad_norm": 3.599980115890503,
+      "learning_rate": 9.516886669012032e-07,
+      "loss": 0.9463,
+      "step": 10460
+    },
+    {
+      "epoch": 0.8064292321924145,
+      "grad_norm": 3.455798387527466,
+      "learning_rate": 9.509560843031695e-07,
+      "loss": 0.8954,
+      "step": 10461
+    },
+    {
+      "epoch": 0.8065063213074314,
+      "grad_norm": 3.5912976264953613,
+      "learning_rate": 9.502237541420534e-07,
+      "loss": 0.9573,
+      "step": 10462
+    },
+    {
+      "epoch": 0.8065834104224483,
+      "grad_norm": 3.59824275970459,
+      "learning_rate": 9.494916764635126e-07,
+      "loss": 0.8395,
+      "step": 10463
+    },
+    {
+      "epoch": 0.8066604995374653,
+      "grad_norm": 3.764808416366577,
+      "learning_rate": 9.487598513131868e-07,
+      "loss": 0.8776,
+      "step": 10464
+    },
+    {
+      "epoch": 0.8067375886524822,
+      "grad_norm": 3.6447689533233643,
+      "learning_rate": 9.480282787367029e-07,
+      "loss": 0.8773,
+      "step": 10465
+    },
+    {
+      "epoch": 0.8068146777674993,
+      "grad_norm": 3.842656373977661,
+      "learning_rate": 9.472969587796694e-07,
+      "loss": 0.9713,
+      "step": 10466
+    },
+    {
+      "epoch": 0.8068917668825162,
+      "grad_norm": 3.9619832038879395,
+      "learning_rate": 9.465658914876813e-07,
+      "loss": 0.871,
+      "step": 10467
+    },
+    {
+      "epoch": 0.8069688559975331,
+      "grad_norm": 3.368314743041992,
+      "learning_rate": 9.458350769063162e-07,
+      "loss": 0.9105,
+      "step": 10468
+    },
+    {
+      "epoch": 0.8070459451125501,
+      "grad_norm": 3.7825496196746826,
+      "learning_rate": 9.451045150811377e-07,
+      "loss": 0.9072,
+      "step": 10469
+    },
+    {
+      "epoch": 0.807123034227567,
+      "grad_norm": 4.548655033111572,
+      "learning_rate": 9.443742060576916e-07,
+      "loss": 0.8509,
+      "step": 10470
+    },
+    {
+      "epoch": 0.8072001233425841,
+      "grad_norm": 3.4712092876434326,
+      "learning_rate": 9.436441498815086e-07,
+      "loss": 0.8334,
+      "step": 10471
+    },
+    {
+      "epoch": 0.807277212457601,
+      "grad_norm": 3.7380151748657227,
+      "learning_rate": 9.42914346598105e-07,
+      "loss": 0.8441,
+      "step": 10472
+    },
+    {
+      "epoch": 0.8073543015726179,
+      "grad_norm": 4.4421467781066895,
+      "learning_rate": 9.421847962529806e-07,
+      "loss": 0.8698,
+      "step": 10473
+    },
+    {
+      "epoch": 0.8074313906876349,
+      "grad_norm": 3.492043972015381,
+      "learning_rate": 9.414554988916175e-07,
+      "loss": 0.86,
+      "step": 10474
+    },
+    {
+      "epoch": 0.8075084798026518,
+      "grad_norm": 3.6429760456085205,
+      "learning_rate": 9.407264545594841e-07,
+      "loss": 0.9613,
+      "step": 10475
+    },
+    {
+      "epoch": 0.8075855689176689,
+      "grad_norm": 4.198061466217041,
+      "learning_rate": 9.399976633020325e-07,
+      "loss": 1.0213,
+      "step": 10476
+    },
+    {
+      "epoch": 0.8076626580326858,
+      "grad_norm": 4.003988265991211,
+      "learning_rate": 9.392691251646991e-07,
+      "loss": 0.9604,
+      "step": 10477
+    },
+    {
+      "epoch": 0.8077397471477027,
+      "grad_norm": 3.8301596641540527,
+      "learning_rate": 9.385408401929052e-07,
+      "loss": 0.9308,
+      "step": 10478
+    },
+    {
+      "epoch": 0.8078168362627197,
+      "grad_norm": 3.498955726623535,
+      "learning_rate": 9.378128084320559e-07,
+      "loss": 0.9101,
+      "step": 10479
+    },
+    {
+      "epoch": 0.8078939253777366,
+      "grad_norm": 3.979250907897949,
+      "learning_rate": 9.370850299275386e-07,
+      "loss": 0.976,
+      "step": 10480
+    },
+    {
+      "epoch": 0.8079710144927537,
+      "grad_norm": 3.989082098007202,
+      "learning_rate": 9.36357504724727e-07,
+      "loss": 0.9461,
+      "step": 10481
+    },
+    {
+      "epoch": 0.8080481036077706,
+      "grad_norm": 3.72397518157959,
+      "learning_rate": 9.356302328689787e-07,
+      "loss": 0.9482,
+      "step": 10482
+    },
+    {
+      "epoch": 0.8081251927227875,
+      "grad_norm": 3.821753740310669,
+      "learning_rate": 9.349032144056358e-07,
+      "loss": 0.9482,
+      "step": 10483
+    },
+    {
+      "epoch": 0.8082022818378045,
+      "grad_norm": 3.7636606693267822,
+      "learning_rate": 9.341764493800248e-07,
+      "loss": 0.7877,
+      "step": 10484
+    },
+    {
+      "epoch": 0.8082793709528214,
+      "grad_norm": 3.7938175201416016,
+      "learning_rate": 9.334499378374534e-07,
+      "loss": 0.8539,
+      "step": 10485
+    },
+    {
+      "epoch": 0.8083564600678385,
+      "grad_norm": 3.736424446105957,
+      "learning_rate": 9.327236798232176e-07,
+      "loss": 0.9258,
+      "step": 10486
+    },
+    {
+      "epoch": 0.8084335491828554,
+      "grad_norm": 3.6965887546539307,
+      "learning_rate": 9.319976753825949e-07,
+      "loss": 0.871,
+      "step": 10487
+    },
+    {
+      "epoch": 0.8085106382978723,
+      "grad_norm": 3.8846166133880615,
+      "learning_rate": 9.312719245608487e-07,
+      "loss": 0.9277,
+      "step": 10488
+    },
+    {
+      "epoch": 0.8085877274128893,
+      "grad_norm": 4.221765518188477,
+      "learning_rate": 9.305464274032256e-07,
+      "loss": 0.9001,
+      "step": 10489
+    },
+    {
+      "epoch": 0.8086648165279062,
+      "grad_norm": 3.9422693252563477,
+      "learning_rate": 9.298211839549576e-07,
+      "loss": 0.9941,
+      "step": 10490
+    },
+    {
+      "epoch": 0.8087419056429233,
+      "grad_norm": 3.6819467544555664,
+      "learning_rate": 9.290961942612576e-07,
+      "loss": 0.855,
+      "step": 10491
+    },
+    {
+      "epoch": 0.8088189947579402,
+      "grad_norm": 3.909057140350342,
+      "learning_rate": 9.283714583673264e-07,
+      "loss": 0.8636,
+      "step": 10492
+    },
+    {
+      "epoch": 0.8088960838729571,
+      "grad_norm": 3.8165695667266846,
+      "learning_rate": 9.276469763183471e-07,
+      "loss": 0.9215,
+      "step": 10493
+    },
+    {
+      "epoch": 0.8089731729879741,
+      "grad_norm": 3.7066540718078613,
+      "learning_rate": 9.269227481594872e-07,
+      "loss": 0.8403,
+      "step": 10494
+    },
+    {
+      "epoch": 0.809050262102991,
+      "grad_norm": 3.6968142986297607,
+      "learning_rate": 9.261987739359007e-07,
+      "loss": 0.9137,
+      "step": 10495
+    },
+    {
+      "epoch": 0.809127351218008,
+      "grad_norm": 3.936868906021118,
+      "learning_rate": 9.254750536927204e-07,
+      "loss": 1.0582,
+      "step": 10496
+    },
+    {
+      "epoch": 0.809204440333025,
+      "grad_norm": 3.842881441116333,
+      "learning_rate": 9.247515874750679e-07,
+      "loss": 0.9131,
+      "step": 10497
+    },
+    {
+      "epoch": 0.809281529448042,
+      "grad_norm": 3.542695999145508,
+      "learning_rate": 9.240283753280477e-07,
+      "loss": 0.7972,
+      "step": 10498
+    },
+    {
+      "epoch": 0.8093586185630589,
+      "grad_norm": 3.8389735221862793,
+      "learning_rate": 9.233054172967481e-07,
+      "loss": 0.9359,
+      "step": 10499
+    },
+    {
+      "epoch": 0.8094357076780758,
+      "grad_norm": 3.6747260093688965,
+      "learning_rate": 9.225827134262422e-07,
+      "loss": 0.901,
+      "step": 10500
+    },
+    {
+      "epoch": 0.8095127967930928,
+      "grad_norm": 4.066951751708984,
+      "learning_rate": 9.218602637615882e-07,
+      "loss": 0.8817,
+      "step": 10501
+    },
+    {
+      "epoch": 0.8095898859081098,
+      "grad_norm": 3.7741777896881104,
+      "learning_rate": 9.211380683478238e-07,
+      "loss": 1.0278,
+      "step": 10502
+    },
+    {
+      "epoch": 0.8096669750231268,
+      "grad_norm": 4.318873882293701,
+      "learning_rate": 9.204161272299761e-07,
+      "loss": 1.0011,
+      "step": 10503
+    },
+    {
+      "epoch": 0.8097440641381437,
+      "grad_norm": 3.677009105682373,
+      "learning_rate": 9.196944404530544e-07,
+      "loss": 0.8841,
+      "step": 10504
+    },
+    {
+      "epoch": 0.8098211532531606,
+      "grad_norm": 3.3423690795898438,
+      "learning_rate": 9.189730080620512e-07,
+      "loss": 0.7182,
+      "step": 10505
+    },
+    {
+      "epoch": 0.8098982423681776,
+      "grad_norm": 3.752957344055176,
+      "learning_rate": 9.182518301019466e-07,
+      "loss": 0.9169,
+      "step": 10506
+    },
+    {
+      "epoch": 0.8099753314831946,
+      "grad_norm": 4.208918571472168,
+      "learning_rate": 9.175309066176991e-07,
+      "loss": 0.9648,
+      "step": 10507
+    },
+    {
+      "epoch": 0.8100524205982116,
+      "grad_norm": 3.9163053035736084,
+      "learning_rate": 9.168102376542559e-07,
+      "loss": 0.8975,
+      "step": 10508
+    },
+    {
+      "epoch": 0.8101295097132285,
+      "grad_norm": 3.690941333770752,
+      "learning_rate": 9.160898232565468e-07,
+      "loss": 0.8554,
+      "step": 10509
+    },
+    {
+      "epoch": 0.8102065988282454,
+      "grad_norm": 3.9642856121063232,
+      "learning_rate": 9.153696634694865e-07,
+      "loss": 1.0219,
+      "step": 10510
+    },
+    {
+      "epoch": 0.8102836879432624,
+      "grad_norm": 3.2848901748657227,
+      "learning_rate": 9.146497583379737e-07,
+      "loss": 0.8234,
+      "step": 10511
+    },
+    {
+      "epoch": 0.8103607770582794,
+      "grad_norm": 3.8563730716705322,
+      "learning_rate": 9.139301079068891e-07,
+      "loss": 0.9495,
+      "step": 10512
+    },
+    {
+      "epoch": 0.8104378661732964,
+      "grad_norm": 3.7783915996551514,
+      "learning_rate": 9.132107122210998e-07,
+      "loss": 0.9717,
+      "step": 10513
+    },
+    {
+      "epoch": 0.8105149552883133,
+      "grad_norm": 3.5279428958892822,
+      "learning_rate": 9.124915713254551e-07,
+      "loss": 0.9075,
+      "step": 10514
+    },
+    {
+      "epoch": 0.8105920444033302,
+      "grad_norm": 3.8966267108917236,
+      "learning_rate": 9.117726852647924e-07,
+      "loss": 0.9536,
+      "step": 10515
+    },
+    {
+      "epoch": 0.8106691335183472,
+      "grad_norm": 3.5992579460144043,
+      "learning_rate": 9.110540540839307e-07,
+      "loss": 0.866,
+      "step": 10516
+    },
+    {
+      "epoch": 0.8107462226333642,
+      "grad_norm": 3.7475366592407227,
+      "learning_rate": 9.103356778276701e-07,
+      "loss": 0.9517,
+      "step": 10517
+    },
+    {
+      "epoch": 0.8108233117483812,
+      "grad_norm": 3.9139578342437744,
+      "learning_rate": 9.096175565407994e-07,
+      "loss": 0.9181,
+      "step": 10518
+    },
+    {
+      "epoch": 0.8109004008633981,
+      "grad_norm": 3.3899145126342773,
+      "learning_rate": 9.088996902680891e-07,
+      "loss": 0.8865,
+      "step": 10519
+    },
+    {
+      "epoch": 0.810977489978415,
+      "grad_norm": 3.5167202949523926,
+      "learning_rate": 9.08182079054295e-07,
+      "loss": 0.9024,
+      "step": 10520
+    },
+    {
+      "epoch": 0.811054579093432,
+      "grad_norm": 4.085640907287598,
+      "learning_rate": 9.074647229441575e-07,
+      "loss": 0.9606,
+      "step": 10521
+    },
+    {
+      "epoch": 0.811131668208449,
+      "grad_norm": 3.6280179023742676,
+      "learning_rate": 9.067476219823973e-07,
+      "loss": 0.963,
+      "step": 10522
+    },
+    {
+      "epoch": 0.811208757323466,
+      "grad_norm": 4.181948184967041,
+      "learning_rate": 9.060307762137233e-07,
+      "loss": 0.9805,
+      "step": 10523
+    },
+    {
+      "epoch": 0.8112858464384829,
+      "grad_norm": 3.7412025928497314,
+      "learning_rate": 9.053141856828274e-07,
+      "loss": 0.888,
+      "step": 10524
+    },
+    {
+      "epoch": 0.8113629355534998,
+      "grad_norm": 3.9258925914764404,
+      "learning_rate": 9.04597850434385e-07,
+      "loss": 1.0001,
+      "step": 10525
+    },
+    {
+      "epoch": 0.8114400246685168,
+      "grad_norm": 3.6146883964538574,
+      "learning_rate": 9.038817705130559e-07,
+      "loss": 0.9586,
+      "step": 10526
+    },
+    {
+      "epoch": 0.8115171137835338,
+      "grad_norm": 3.679476499557495,
+      "learning_rate": 9.031659459634856e-07,
+      "loss": 0.9259,
+      "step": 10527
+    },
+    {
+      "epoch": 0.8115942028985508,
+      "grad_norm": 4.017477512359619,
+      "learning_rate": 9.024503768302989e-07,
+      "loss": 0.9798,
+      "step": 10528
+    },
+    {
+      "epoch": 0.8116712920135677,
+      "grad_norm": 3.543410301208496,
+      "learning_rate": 9.017350631581095e-07,
+      "loss": 0.9361,
+      "step": 10529
+    },
+    {
+      "epoch": 0.8117483811285846,
+      "grad_norm": 3.8175089359283447,
+      "learning_rate": 9.010200049915141e-07,
+      "loss": 0.851,
+      "step": 10530
+    },
+    {
+      "epoch": 0.8118254702436016,
+      "grad_norm": 4.348031997680664,
+      "learning_rate": 9.003052023750913e-07,
+      "loss": 0.9564,
+      "step": 10531
+    },
+    {
+      "epoch": 0.8119025593586185,
+      "grad_norm": 3.7295446395874023,
+      "learning_rate": 8.995906553534084e-07,
+      "loss": 0.8679,
+      "step": 10532
+    },
+    {
+      "epoch": 0.8119796484736356,
+      "grad_norm": 3.961146116256714,
+      "learning_rate": 8.988763639710097e-07,
+      "loss": 0.9047,
+      "step": 10533
+    },
+    {
+      "epoch": 0.8120567375886525,
+      "grad_norm": 3.5502161979675293,
+      "learning_rate": 8.981623282724295e-07,
+      "loss": 0.9201,
+      "step": 10534
+    },
+    {
+      "epoch": 0.8121338267036694,
+      "grad_norm": 4.238675594329834,
+      "learning_rate": 8.974485483021844e-07,
+      "loss": 0.8612,
+      "step": 10535
+    },
+    {
+      "epoch": 0.8122109158186864,
+      "grad_norm": 3.8428282737731934,
+      "learning_rate": 8.967350241047745e-07,
+      "loss": 0.887,
+      "step": 10536
+    },
+    {
+      "epoch": 0.8122880049337033,
+      "grad_norm": 3.5076427459716797,
+      "learning_rate": 8.960217557246842e-07,
+      "loss": 0.9022,
+      "step": 10537
+    },
+    {
+      "epoch": 0.8123650940487204,
+      "grad_norm": 3.7073769569396973,
+      "learning_rate": 8.953087432063839e-07,
+      "loss": 0.9396,
+      "step": 10538
+    },
+    {
+      "epoch": 0.8124421831637373,
+      "grad_norm": 4.131918430328369,
+      "learning_rate": 8.945959865943238e-07,
+      "loss": 0.9772,
+      "step": 10539
+    },
+    {
+      "epoch": 0.8125192722787542,
+      "grad_norm": 3.5746655464172363,
+      "learning_rate": 8.938834859329415e-07,
+      "loss": 0.8473,
+      "step": 10540
+    },
+    {
+      "epoch": 0.8125963613937712,
+      "grad_norm": 3.607383966445923,
+      "learning_rate": 8.931712412666571e-07,
+      "loss": 0.9258,
+      "step": 10541
+    },
+    {
+      "epoch": 0.8126734505087881,
+      "grad_norm": 3.5877065658569336,
+      "learning_rate": 8.924592526398762e-07,
+      "loss": 0.8664,
+      "step": 10542
+    },
+    {
+      "epoch": 0.8127505396238052,
+      "grad_norm": 3.302738904953003,
+      "learning_rate": 8.917475200969889e-07,
+      "loss": 0.7732,
+      "step": 10543
+    },
+    {
+      "epoch": 0.8128276287388221,
+      "grad_norm": 3.6982762813568115,
+      "learning_rate": 8.910360436823651e-07,
+      "loss": 0.9946,
+      "step": 10544
+    },
+    {
+      "epoch": 0.812904717853839,
+      "grad_norm": 3.758657932281494,
+      "learning_rate": 8.903248234403633e-07,
+      "loss": 0.8554,
+      "step": 10545
+    },
+    {
+      "epoch": 0.812981806968856,
+      "grad_norm": 3.797997236251831,
+      "learning_rate": 8.896138594153242e-07,
+      "loss": 0.9052,
+      "step": 10546
+    },
+    {
+      "epoch": 0.8130588960838729,
+      "grad_norm": 4.33210563659668,
+      "learning_rate": 8.889031516515729e-07,
+      "loss": 1.022,
+      "step": 10547
+    },
+    {
+      "epoch": 0.81313598519889,
+      "grad_norm": 3.558067560195923,
+      "learning_rate": 8.881927001934177e-07,
+      "loss": 0.829,
+      "step": 10548
+    },
+    {
+      "epoch": 0.8132130743139069,
+      "grad_norm": 3.977452516555786,
+      "learning_rate": 8.874825050851532e-07,
+      "loss": 0.9614,
+      "step": 10549
+    },
+    {
+      "epoch": 0.8132901634289238,
+      "grad_norm": 4.086385250091553,
+      "learning_rate": 8.867725663710547e-07,
+      "loss": 1.0238,
+      "step": 10550
+    },
+    {
+      "epoch": 0.8133672525439408,
+      "grad_norm": 4.283471584320068,
+      "learning_rate": 8.860628840953833e-07,
+      "loss": 0.8701,
+      "step": 10551
+    },
+    {
+      "epoch": 0.8134443416589577,
+      "grad_norm": 4.118513584136963,
+      "learning_rate": 8.853534583023843e-07,
+      "loss": 0.9455,
+      "step": 10552
+    },
+    {
+      "epoch": 0.8135214307739748,
+      "grad_norm": 3.710076332092285,
+      "learning_rate": 8.846442890362872e-07,
+      "loss": 0.8362,
+      "step": 10553
+    },
+    {
+      "epoch": 0.8135985198889917,
+      "grad_norm": 3.5066654682159424,
+      "learning_rate": 8.839353763413056e-07,
+      "loss": 0.8795,
+      "step": 10554
+    },
+    {
+      "epoch": 0.8136756090040086,
+      "grad_norm": 3.774261236190796,
+      "learning_rate": 8.832267202616346e-07,
+      "loss": 0.9621,
+      "step": 10555
+    },
+    {
+      "epoch": 0.8137526981190256,
+      "grad_norm": 3.7309882640838623,
+      "learning_rate": 8.825183208414562e-07,
+      "loss": 0.9751,
+      "step": 10556
+    },
+    {
+      "epoch": 0.8138297872340425,
+      "grad_norm": 3.934528112411499,
+      "learning_rate": 8.818101781249355e-07,
+      "loss": 0.9044,
+      "step": 10557
+    },
+    {
+      "epoch": 0.8139068763490596,
+      "grad_norm": 3.5573508739471436,
+      "learning_rate": 8.811022921562218e-07,
+      "loss": 0.8235,
+      "step": 10558
+    },
+    {
+      "epoch": 0.8139839654640765,
+      "grad_norm": 3.88106632232666,
+      "learning_rate": 8.803946629794475e-07,
+      "loss": 0.9063,
+      "step": 10559
+    },
+    {
+      "epoch": 0.8140610545790934,
+      "grad_norm": 3.713273048400879,
+      "learning_rate": 8.796872906387299e-07,
+      "loss": 0.8801,
+      "step": 10560
+    },
+    {
+      "epoch": 0.8141381436941104,
+      "grad_norm": 3.7265689373016357,
+      "learning_rate": 8.789801751781707e-07,
+      "loss": 0.9106,
+      "step": 10561
+    },
+    {
+      "epoch": 0.8142152328091273,
+      "grad_norm": 3.56809401512146,
+      "learning_rate": 8.782733166418539e-07,
+      "loss": 0.9087,
+      "step": 10562
+    },
+    {
+      "epoch": 0.8142923219241444,
+      "grad_norm": 4.446261405944824,
+      "learning_rate": 8.775667150738487e-07,
+      "loss": 0.9448,
+      "step": 10563
+    },
+    {
+      "epoch": 0.8143694110391613,
+      "grad_norm": 3.8217625617980957,
+      "learning_rate": 8.768603705182094e-07,
+      "loss": 0.8059,
+      "step": 10564
+    },
+    {
+      "epoch": 0.8144465001541782,
+      "grad_norm": 3.5717344284057617,
+      "learning_rate": 8.761542830189706e-07,
+      "loss": 0.859,
+      "step": 10565
+    },
+    {
+      "epoch": 0.8145235892691952,
+      "grad_norm": 3.8214683532714844,
+      "learning_rate": 8.754484526201546e-07,
+      "loss": 0.8514,
+      "step": 10566
+    },
+    {
+      "epoch": 0.8146006783842121,
+      "grad_norm": 4.17495059967041,
+      "learning_rate": 8.747428793657658e-07,
+      "loss": 0.9882,
+      "step": 10567
+    },
+    {
+      "epoch": 0.8146777674992292,
+      "grad_norm": 4.129594326019287,
+      "learning_rate": 8.740375632997927e-07,
+      "loss": 0.9762,
+      "step": 10568
+    },
+    {
+      "epoch": 0.8147548566142461,
+      "grad_norm": 3.8099358081817627,
+      "learning_rate": 8.733325044662106e-07,
+      "loss": 0.9268,
+      "step": 10569
+    },
+    {
+      "epoch": 0.814831945729263,
+      "grad_norm": 4.046210765838623,
+      "learning_rate": 8.726277029089725e-07,
+      "loss": 0.8863,
+      "step": 10570
+    },
+    {
+      "epoch": 0.81490903484428,
+      "grad_norm": 4.233992099761963,
+      "learning_rate": 8.719231586720211e-07,
+      "loss": 0.9588,
+      "step": 10571
+    },
+    {
+      "epoch": 0.8149861239592969,
+      "grad_norm": 3.290083646774292,
+      "learning_rate": 8.712188717992814e-07,
+      "loss": 0.8304,
+      "step": 10572
+    },
+    {
+      "epoch": 0.815063213074314,
+      "grad_norm": 3.646782875061035,
+      "learning_rate": 8.70514842334661e-07,
+      "loss": 0.9994,
+      "step": 10573
+    },
+    {
+      "epoch": 0.8151403021893309,
+      "grad_norm": 3.9229447841644287,
+      "learning_rate": 8.69811070322053e-07,
+      "loss": 0.8786,
+      "step": 10574
+    },
+    {
+      "epoch": 0.8152173913043478,
+      "grad_norm": 3.7923243045806885,
+      "learning_rate": 8.691075558053353e-07,
+      "loss": 0.936,
+      "step": 10575
+    },
+    {
+      "epoch": 0.8152944804193648,
+      "grad_norm": 3.350618600845337,
+      "learning_rate": 8.684042988283659e-07,
+      "loss": 0.858,
+      "step": 10576
+    },
+    {
+      "epoch": 0.8153715695343817,
+      "grad_norm": 4.026613235473633,
+      "learning_rate": 8.677012994349904e-07,
+      "loss": 0.9224,
+      "step": 10577
+    },
+    {
+      "epoch": 0.8154486586493987,
+      "grad_norm": 3.9146225452423096,
+      "learning_rate": 8.669985576690371e-07,
+      "loss": 0.8832,
+      "step": 10578
+    },
+    {
+      "epoch": 0.8155257477644157,
+      "grad_norm": 3.648346185684204,
+      "learning_rate": 8.662960735743181e-07,
+      "loss": 0.7847,
+      "step": 10579
+    },
+    {
+      "epoch": 0.8156028368794326,
+      "grad_norm": 3.5210020542144775,
+      "learning_rate": 8.655938471946313e-07,
+      "loss": 0.8185,
+      "step": 10580
+    },
+    {
+      "epoch": 0.8156799259944496,
+      "grad_norm": 4.143526554107666,
+      "learning_rate": 8.648918785737542e-07,
+      "loss": 0.8891,
+      "step": 10581
+    },
+    {
+      "epoch": 0.8157570151094665,
+      "grad_norm": 3.710387706756592,
+      "learning_rate": 8.641901677554526e-07,
+      "loss": 0.8862,
+      "step": 10582
+    },
+    {
+      "epoch": 0.8158341042244835,
+      "grad_norm": 3.5736660957336426,
+      "learning_rate": 8.634887147834736e-07,
+      "loss": 0.8868,
+      "step": 10583
+    },
+    {
+      "epoch": 0.8159111933395005,
+      "grad_norm": 3.8351974487304688,
+      "learning_rate": 8.6278751970155e-07,
+      "loss": 0.9472,
+      "step": 10584
+    },
+    {
+      "epoch": 0.8159882824545174,
+      "grad_norm": 3.8771169185638428,
+      "learning_rate": 8.620865825533975e-07,
+      "loss": 0.9296,
+      "step": 10585
+    },
+    {
+      "epoch": 0.8160653715695344,
+      "grad_norm": 3.649178981781006,
+      "learning_rate": 8.613859033827165e-07,
+      "loss": 0.9505,
+      "step": 10586
+    },
+    {
+      "epoch": 0.8161424606845513,
+      "grad_norm": 3.742682456970215,
+      "learning_rate": 8.606854822331895e-07,
+      "loss": 0.8974,
+      "step": 10587
+    },
+    {
+      "epoch": 0.8162195497995683,
+      "grad_norm": 4.2910237312316895,
+      "learning_rate": 8.599853191484842e-07,
+      "loss": 0.9299,
+      "step": 10588
+    },
+    {
+      "epoch": 0.8162966389145853,
+      "grad_norm": 3.860732078552246,
+      "learning_rate": 8.592854141722523e-07,
+      "loss": 1.0105,
+      "step": 10589
+    },
+    {
+      "epoch": 0.8163737280296022,
+      "grad_norm": 3.7548787593841553,
+      "learning_rate": 8.585857673481302e-07,
+      "loss": 0.9132,
+      "step": 10590
+    },
+    {
+      "epoch": 0.8164508171446192,
+      "grad_norm": 3.929503917694092,
+      "learning_rate": 8.578863787197372e-07,
+      "loss": 0.9521,
+      "step": 10591
+    },
+    {
+      "epoch": 0.8165279062596361,
+      "grad_norm": 4.102705955505371,
+      "learning_rate": 8.571872483306748e-07,
+      "loss": 0.9715,
+      "step": 10592
+    },
+    {
+      "epoch": 0.8166049953746531,
+      "grad_norm": 3.499938488006592,
+      "learning_rate": 8.564883762245313e-07,
+      "loss": 0.9841,
+      "step": 10593
+    },
+    {
+      "epoch": 0.8166820844896701,
+      "grad_norm": 4.210318565368652,
+      "learning_rate": 8.55789762444878e-07,
+      "loss": 0.9848,
+      "step": 10594
+    },
+    {
+      "epoch": 0.816759173604687,
+      "grad_norm": 3.527986526489258,
+      "learning_rate": 8.550914070352695e-07,
+      "loss": 0.9264,
+      "step": 10595
+    },
+    {
+      "epoch": 0.816836262719704,
+      "grad_norm": 3.987180471420288,
+      "learning_rate": 8.54393310039246e-07,
+      "loss": 0.9992,
+      "step": 10596
+    },
+    {
+      "epoch": 0.8169133518347209,
+      "grad_norm": 3.61505389213562,
+      "learning_rate": 8.536954715003276e-07,
+      "loss": 0.8168,
+      "step": 10597
+    },
+    {
+      "epoch": 0.8169904409497379,
+      "grad_norm": 3.735048294067383,
+      "learning_rate": 8.529978914620219e-07,
+      "loss": 0.8828,
+      "step": 10598
+    },
+    {
+      "epoch": 0.8170675300647549,
+      "grad_norm": 4.112254619598389,
+      "learning_rate": 8.523005699678205e-07,
+      "loss": 0.9597,
+      "step": 10599
+    },
+    {
+      "epoch": 0.8171446191797718,
+      "grad_norm": 4.099547386169434,
+      "learning_rate": 8.516035070611967e-07,
+      "loss": 0.9305,
+      "step": 10600
+    },
+    {
+      "epoch": 0.8172217082947888,
+      "grad_norm": 3.74941086769104,
+      "learning_rate": 8.509067027856093e-07,
+      "loss": 0.8205,
+      "step": 10601
+    },
+    {
+      "epoch": 0.8172987974098057,
+      "grad_norm": 3.4929075241088867,
+      "learning_rate": 8.502101571845001e-07,
+      "loss": 0.9547,
+      "step": 10602
+    },
+    {
+      "epoch": 0.8173758865248227,
+      "grad_norm": 4.058771133422852,
+      "learning_rate": 8.495138703012957e-07,
+      "loss": 0.983,
+      "step": 10603
+    },
+    {
+      "epoch": 0.8174529756398397,
+      "grad_norm": 3.9835104942321777,
+      "learning_rate": 8.488178421794047e-07,
+      "loss": 0.9632,
+      "step": 10604
+    },
+    {
+      "epoch": 0.8175300647548566,
+      "grad_norm": 4.225916862487793,
+      "learning_rate": 8.481220728622225e-07,
+      "loss": 0.9406,
+      "step": 10605
+    },
+    {
+      "epoch": 0.8176071538698736,
+      "grad_norm": 3.349649429321289,
+      "learning_rate": 8.474265623931272e-07,
+      "loss": 0.7666,
+      "step": 10606
+    },
+    {
+      "epoch": 0.8176842429848905,
+      "grad_norm": 3.722669839859009,
+      "learning_rate": 8.467313108154773e-07,
+      "loss": 0.939,
+      "step": 10607
+    },
+    {
+      "epoch": 0.8177613320999075,
+      "grad_norm": 3.4461019039154053,
+      "learning_rate": 8.4603631817262e-07,
+      "loss": 0.8237,
+      "step": 10608
+    },
+    {
+      "epoch": 0.8178384212149244,
+      "grad_norm": 3.774502992630005,
+      "learning_rate": 8.453415845078844e-07,
+      "loss": 0.8583,
+      "step": 10609
+    },
+    {
+      "epoch": 0.8179155103299414,
+      "grad_norm": 3.614598512649536,
+      "learning_rate": 8.446471098645831e-07,
+      "loss": 0.9551,
+      "step": 10610
+    },
+    {
+      "epoch": 0.8179925994449584,
+      "grad_norm": 3.7169055938720703,
+      "learning_rate": 8.439528942860137e-07,
+      "loss": 0.8629,
+      "step": 10611
+    },
+    {
+      "epoch": 0.8180696885599753,
+      "grad_norm": 3.7345728874206543,
+      "learning_rate": 8.432589378154582e-07,
+      "loss": 0.9106,
+      "step": 10612
+    },
+    {
+      "epoch": 0.8181467776749923,
+      "grad_norm": 4.463329315185547,
+      "learning_rate": 8.425652404961781e-07,
+      "loss": 0.9058,
+      "step": 10613
+    },
+    {
+      "epoch": 0.8182238667900092,
+      "grad_norm": 3.3270933628082275,
+      "learning_rate": 8.418718023714235e-07,
+      "loss": 0.8002,
+      "step": 10614
+    },
+    {
+      "epoch": 0.8183009559050262,
+      "grad_norm": 3.8927876949310303,
+      "learning_rate": 8.41178623484426e-07,
+      "loss": 0.8976,
+      "step": 10615
+    },
+    {
+      "epoch": 0.8183780450200432,
+      "grad_norm": 3.9593920707702637,
+      "learning_rate": 8.404857038784026e-07,
+      "loss": 0.9744,
+      "step": 10616
+    },
+    {
+      "epoch": 0.8184551341350601,
+      "grad_norm": 3.8429269790649414,
+      "learning_rate": 8.39793043596554e-07,
+      "loss": 0.8321,
+      "step": 10617
+    },
+    {
+      "epoch": 0.8185322232500771,
+      "grad_norm": 3.442072629928589,
+      "learning_rate": 8.391006426820619e-07,
+      "loss": 0.8208,
+      "step": 10618
+    },
+    {
+      "epoch": 0.818609312365094,
+      "grad_norm": 3.8619630336761475,
+      "learning_rate": 8.384085011780946e-07,
+      "loss": 0.8541,
+      "step": 10619
+    },
+    {
+      "epoch": 0.818686401480111,
+      "grad_norm": 3.815845489501953,
+      "learning_rate": 8.377166191278036e-07,
+      "loss": 0.9521,
+      "step": 10620
+    },
+    {
+      "epoch": 0.818763490595128,
+      "grad_norm": 3.9525768756866455,
+      "learning_rate": 8.370249965743249e-07,
+      "loss": 0.9698,
+      "step": 10621
+    },
+    {
+      "epoch": 0.8188405797101449,
+      "grad_norm": 3.71791934967041,
+      "learning_rate": 8.363336335607769e-07,
+      "loss": 1.037,
+      "step": 10622
+    },
+    {
+      "epoch": 0.8189176688251619,
+      "grad_norm": 3.4764459133148193,
+      "learning_rate": 8.356425301302639e-07,
+      "loss": 0.8571,
+      "step": 10623
+    },
+    {
+      "epoch": 0.8189947579401788,
+      "grad_norm": 4.280252933502197,
+      "learning_rate": 8.3495168632587e-07,
+      "loss": 1.067,
+      "step": 10624
+    },
+    {
+      "epoch": 0.8190718470551958,
+      "grad_norm": 3.945054769515991,
+      "learning_rate": 8.342611021906672e-07,
+      "loss": 0.9154,
+      "step": 10625
+    },
+    {
+      "epoch": 0.8191489361702128,
+      "grad_norm": 3.6436848640441895,
+      "learning_rate": 8.335707777677099e-07,
+      "loss": 0.8429,
+      "step": 10626
+    },
+    {
+      "epoch": 0.8192260252852297,
+      "grad_norm": 3.7490007877349854,
+      "learning_rate": 8.32880713100036e-07,
+      "loss": 0.9217,
+      "step": 10627
+    },
+    {
+      "epoch": 0.8193031144002467,
+      "grad_norm": 3.5114471912384033,
+      "learning_rate": 8.321909082306684e-07,
+      "loss": 0.9029,
+      "step": 10628
+    },
+    {
+      "epoch": 0.8193802035152636,
+      "grad_norm": 3.8217241764068604,
+      "learning_rate": 8.315013632026114e-07,
+      "loss": 0.861,
+      "step": 10629
+    },
+    {
+      "epoch": 0.8194572926302806,
+      "grad_norm": 3.3438620567321777,
+      "learning_rate": 8.308120780588553e-07,
+      "loss": 0.7675,
+      "step": 10630
+    },
+    {
+      "epoch": 0.8195343817452976,
+      "grad_norm": 4.367412090301514,
+      "learning_rate": 8.301230528423726e-07,
+      "loss": 0.934,
+      "step": 10631
+    },
+    {
+      "epoch": 0.8196114708603145,
+      "grad_norm": 4.256845951080322,
+      "learning_rate": 8.29434287596122e-07,
+      "loss": 0.9082,
+      "step": 10632
+    },
+    {
+      "epoch": 0.8196885599753315,
+      "grad_norm": 3.7340166568756104,
+      "learning_rate": 8.287457823630429e-07,
+      "loss": 0.9485,
+      "step": 10633
+    },
+    {
+      "epoch": 0.8197656490903484,
+      "grad_norm": 3.8312957286834717,
+      "learning_rate": 8.280575371860627e-07,
+      "loss": 0.8524,
+      "step": 10634
+    },
+    {
+      "epoch": 0.8198427382053654,
+      "grad_norm": 3.471848726272583,
+      "learning_rate": 8.273695521080866e-07,
+      "loss": 0.8596,
+      "step": 10635
+    },
+    {
+      "epoch": 0.8199198273203824,
+      "grad_norm": 3.743417978286743,
+      "learning_rate": 8.266818271720078e-07,
+      "loss": 0.984,
+      "step": 10636
+    },
+    {
+      "epoch": 0.8199969164353993,
+      "grad_norm": 3.6192376613616943,
+      "learning_rate": 8.259943624207034e-07,
+      "loss": 0.9746,
+      "step": 10637
+    },
+    {
+      "epoch": 0.8200740055504163,
+      "grad_norm": 3.581130027770996,
+      "learning_rate": 8.253071578970329e-07,
+      "loss": 0.8544,
+      "step": 10638
+    },
+    {
+      "epoch": 0.8201510946654332,
+      "grad_norm": 3.57409405708313,
+      "learning_rate": 8.246202136438408e-07,
+      "loss": 0.7728,
+      "step": 10639
+    },
+    {
+      "epoch": 0.8202281837804501,
+      "grad_norm": 3.536517858505249,
+      "learning_rate": 8.239335297039525e-07,
+      "loss": 0.8483,
+      "step": 10640
+    },
+    {
+      "epoch": 0.8203052728954672,
+      "grad_norm": 3.8591723442077637,
+      "learning_rate": 8.232471061201797e-07,
+      "loss": 0.9938,
+      "step": 10641
+    },
+    {
+      "epoch": 0.8203823620104841,
+      "grad_norm": 3.6102797985076904,
+      "learning_rate": 8.225609429353187e-07,
+      "loss": 0.9334,
+      "step": 10642
+    },
+    {
+      "epoch": 0.8204594511255011,
+      "grad_norm": 3.6389827728271484,
+      "learning_rate": 8.218750401921466e-07,
+      "loss": 0.8946,
+      "step": 10643
+    },
+    {
+      "epoch": 0.820536540240518,
+      "grad_norm": 3.4284937381744385,
+      "learning_rate": 8.211893979334284e-07,
+      "loss": 0.7989,
+      "step": 10644
+    },
+    {
+      "epoch": 0.820613629355535,
+      "grad_norm": 3.6757824420928955,
+      "learning_rate": 8.205040162019073e-07,
+      "loss": 0.8724,
+      "step": 10645
+    },
+    {
+      "epoch": 0.820690718470552,
+      "grad_norm": 4.049697399139404,
+      "learning_rate": 8.198188950403146e-07,
+      "loss": 0.9266,
+      "step": 10646
+    },
+    {
+      "epoch": 0.8207678075855689,
+      "grad_norm": 3.7433371543884277,
+      "learning_rate": 8.191340344913629e-07,
+      "loss": 0.9927,
+      "step": 10647
+    },
+    {
+      "epoch": 0.8208448967005859,
+      "grad_norm": 3.705571174621582,
+      "learning_rate": 8.184494345977517e-07,
+      "loss": 0.8741,
+      "step": 10648
+    },
+    {
+      "epoch": 0.8209219858156028,
+      "grad_norm": 3.779428482055664,
+      "learning_rate": 8.177650954021632e-07,
+      "loss": 0.9699,
+      "step": 10649
+    },
+    {
+      "epoch": 0.8209990749306197,
+      "grad_norm": 3.6336560249328613,
+      "learning_rate": 8.170810169472593e-07,
+      "loss": 0.8423,
+      "step": 10650
+    },
+    {
+      "epoch": 0.8210761640456368,
+      "grad_norm": 3.8812954425811768,
+      "learning_rate": 8.163971992756897e-07,
+      "loss": 0.9858,
+      "step": 10651
+    },
+    {
+      "epoch": 0.8211532531606537,
+      "grad_norm": 3.7961621284484863,
+      "learning_rate": 8.157136424300877e-07,
+      "loss": 0.9314,
+      "step": 10652
+    },
+    {
+      "epoch": 0.8212303422756707,
+      "grad_norm": 3.7296488285064697,
+      "learning_rate": 8.15030346453069e-07,
+      "loss": 1.0059,
+      "step": 10653
+    },
+    {
+      "epoch": 0.8213074313906876,
+      "grad_norm": 3.9131686687469482,
+      "learning_rate": 8.143473113872353e-07,
+      "loss": 0.8374,
+      "step": 10654
+    },
+    {
+      "epoch": 0.8213845205057045,
+      "grad_norm": 4.030622482299805,
+      "learning_rate": 8.136645372751667e-07,
+      "loss": 0.8713,
+      "step": 10655
+    },
+    {
+      "epoch": 0.8214616096207216,
+      "grad_norm": 3.774033308029175,
+      "learning_rate": 8.129820241594333e-07,
+      "loss": 0.9483,
+      "step": 10656
+    },
+    {
+      "epoch": 0.8215386987357385,
+      "grad_norm": 3.8504068851470947,
+      "learning_rate": 8.122997720825853e-07,
+      "loss": 1.045,
+      "step": 10657
+    },
+    {
+      "epoch": 0.8216157878507555,
+      "grad_norm": 3.6778900623321533,
+      "learning_rate": 8.116177810871578e-07,
+      "loss": 0.8602,
+      "step": 10658
+    },
+    {
+      "epoch": 0.8216928769657724,
+      "grad_norm": 3.6927058696746826,
+      "learning_rate": 8.109360512156695e-07,
+      "loss": 0.9418,
+      "step": 10659
+    },
+    {
+      "epoch": 0.8217699660807893,
+      "grad_norm": 3.3587496280670166,
+      "learning_rate": 8.10254582510624e-07,
+      "loss": 0.8898,
+      "step": 10660
+    },
+    {
+      "epoch": 0.8218470551958064,
+      "grad_norm": 3.8647491931915283,
+      "learning_rate": 8.095733750145046e-07,
+      "loss": 0.9488,
+      "step": 10661
+    },
+    {
+      "epoch": 0.8219241443108233,
+      "grad_norm": 3.846730947494507,
+      "learning_rate": 8.088924287697824e-07,
+      "loss": 0.9386,
+      "step": 10662
+    },
+    {
+      "epoch": 0.8220012334258403,
+      "grad_norm": 3.8304007053375244,
+      "learning_rate": 8.082117438189113e-07,
+      "loss": 0.9066,
+      "step": 10663
+    },
+    {
+      "epoch": 0.8220783225408572,
+      "grad_norm": 3.793449878692627,
+      "learning_rate": 8.075313202043277e-07,
+      "loss": 0.8526,
+      "step": 10664
+    },
+    {
+      "epoch": 0.8221554116558741,
+      "grad_norm": 4.21845006942749,
+      "learning_rate": 8.068511579684546e-07,
+      "loss": 1.0152,
+      "step": 10665
+    },
+    {
+      "epoch": 0.8222325007708912,
+      "grad_norm": 3.6791608333587646,
+      "learning_rate": 8.06171257153694e-07,
+      "loss": 0.9214,
+      "step": 10666
+    },
+    {
+      "epoch": 0.8223095898859081,
+      "grad_norm": 3.7897164821624756,
+      "learning_rate": 8.054916178024347e-07,
+      "loss": 0.9361,
+      "step": 10667
+    },
+    {
+      "epoch": 0.8223866790009251,
+      "grad_norm": 3.942296266555786,
+      "learning_rate": 8.04812239957049e-07,
+      "loss": 1.037,
+      "step": 10668
+    },
+    {
+      "epoch": 0.822463768115942,
+      "grad_norm": 3.920706272125244,
+      "learning_rate": 8.041331236598926e-07,
+      "loss": 0.8321,
+      "step": 10669
+    },
+    {
+      "epoch": 0.822540857230959,
+      "grad_norm": 3.6586873531341553,
+      "learning_rate": 8.034542689533053e-07,
+      "loss": 0.9324,
+      "step": 10670
+    },
+    {
+      "epoch": 0.822617946345976,
+      "grad_norm": 3.7215545177459717,
+      "learning_rate": 8.027756758796107e-07,
+      "loss": 0.884,
+      "step": 10671
+    },
+    {
+      "epoch": 0.8226950354609929,
+      "grad_norm": 3.4907851219177246,
+      "learning_rate": 8.020973444811142e-07,
+      "loss": 0.801,
+      "step": 10672
+    },
+    {
+      "epoch": 0.8227721245760099,
+      "grad_norm": 3.6691055297851562,
+      "learning_rate": 8.014192748001059e-07,
+      "loss": 0.9102,
+      "step": 10673
+    },
+    {
+      "epoch": 0.8228492136910268,
+      "grad_norm": 3.8015921115875244,
+      "learning_rate": 8.007414668788616e-07,
+      "loss": 0.924,
+      "step": 10674
+    },
+    {
+      "epoch": 0.8229263028060438,
+      "grad_norm": 3.784499168395996,
+      "learning_rate": 8.000639207596377e-07,
+      "loss": 0.928,
+      "step": 10675
+    },
+    {
+      "epoch": 0.8230033919210608,
+      "grad_norm": 3.817927122116089,
+      "learning_rate": 7.993866364846781e-07,
+      "loss": 0.892,
+      "step": 10676
+    },
+    {
+      "epoch": 0.8230804810360777,
+      "grad_norm": 3.8663864135742188,
+      "learning_rate": 7.987096140962047e-07,
+      "loss": 0.8445,
+      "step": 10677
+    },
+    {
+      "epoch": 0.8231575701510947,
+      "grad_norm": 4.127407073974609,
+      "learning_rate": 7.980328536364279e-07,
+      "loss": 0.914,
+      "step": 10678
+    },
+    {
+      "epoch": 0.8232346592661116,
+      "grad_norm": 3.5715925693511963,
+      "learning_rate": 7.9735635514754e-07,
+      "loss": 0.924,
+      "step": 10679
+    },
+    {
+      "epoch": 0.8233117483811286,
+      "grad_norm": 3.3841607570648193,
+      "learning_rate": 7.966801186717172e-07,
+      "loss": 0.805,
+      "step": 10680
+    },
+    {
+      "epoch": 0.8233888374961456,
+      "grad_norm": 3.75689435005188,
+      "learning_rate": 7.960041442511196e-07,
+      "loss": 0.8059,
+      "step": 10681
+    },
+    {
+      "epoch": 0.8234659266111625,
+      "grad_norm": 3.8247830867767334,
+      "learning_rate": 7.95328431927892e-07,
+      "loss": 0.9118,
+      "step": 10682
+    },
+    {
+      "epoch": 0.8235430157261795,
+      "grad_norm": 4.0702385902404785,
+      "learning_rate": 7.946529817441584e-07,
+      "loss": 0.9658,
+      "step": 10683
+    },
+    {
+      "epoch": 0.8236201048411964,
+      "grad_norm": 3.605431318283081,
+      "learning_rate": 7.939777937420317e-07,
+      "loss": 0.9196,
+      "step": 10684
+    },
+    {
+      "epoch": 0.8236971939562134,
+      "grad_norm": 3.7044453620910645,
+      "learning_rate": 7.933028679636057e-07,
+      "loss": 0.908,
+      "step": 10685
+    },
+    {
+      "epoch": 0.8237742830712304,
+      "grad_norm": 3.886737585067749,
+      "learning_rate": 7.926282044509593e-07,
+      "loss": 0.9069,
+      "step": 10686
+    },
+    {
+      "epoch": 0.8238513721862473,
+      "grad_norm": 3.6313858032226562,
+      "learning_rate": 7.919538032461544e-07,
+      "loss": 0.9743,
+      "step": 10687
+    },
+    {
+      "epoch": 0.8239284613012643,
+      "grad_norm": 3.607703924179077,
+      "learning_rate": 7.912796643912352e-07,
+      "loss": 0.8587,
+      "step": 10688
+    },
+    {
+      "epoch": 0.8240055504162812,
+      "grad_norm": 3.6826751232147217,
+      "learning_rate": 7.906057879282314e-07,
+      "loss": 0.9056,
+      "step": 10689
+    },
+    {
+      "epoch": 0.8240826395312982,
+      "grad_norm": 4.050000190734863,
+      "learning_rate": 7.89932173899155e-07,
+      "loss": 0.8867,
+      "step": 10690
+    },
+    {
+      "epoch": 0.8241597286463151,
+      "grad_norm": 3.8793132305145264,
+      "learning_rate": 7.89258822346004e-07,
+      "loss": 0.8438,
+      "step": 10691
+    },
+    {
+      "epoch": 0.8242368177613321,
+      "grad_norm": 3.8853375911712646,
+      "learning_rate": 7.885857333107566e-07,
+      "loss": 0.8774,
+      "step": 10692
+    },
+    {
+      "epoch": 0.8243139068763491,
+      "grad_norm": 3.8986237049102783,
+      "learning_rate": 7.87912906835378e-07,
+      "loss": 0.9329,
+      "step": 10693
+    },
+    {
+      "epoch": 0.824390995991366,
+      "grad_norm": 3.8146276473999023,
+      "learning_rate": 7.872403429618141e-07,
+      "loss": 0.9111,
+      "step": 10694
+    },
+    {
+      "epoch": 0.824468085106383,
+      "grad_norm": 3.530385732650757,
+      "learning_rate": 7.865680417319965e-07,
+      "loss": 0.9361,
+      "step": 10695
+    },
+    {
+      "epoch": 0.8245451742214,
+      "grad_norm": 3.7008519172668457,
+      "learning_rate": 7.858960031878399e-07,
+      "loss": 0.905,
+      "step": 10696
+    },
+    {
+      "epoch": 0.8246222633364169,
+      "grad_norm": 3.490631103515625,
+      "learning_rate": 7.852242273712429e-07,
+      "loss": 0.8935,
+      "step": 10697
+    },
+    {
+      "epoch": 0.8246993524514339,
+      "grad_norm": 4.064510345458984,
+      "learning_rate": 7.845527143240856e-07,
+      "loss": 0.93,
+      "step": 10698
+    },
+    {
+      "epoch": 0.8247764415664508,
+      "grad_norm": 3.9456470012664795,
+      "learning_rate": 7.838814640882342e-07,
+      "loss": 0.9109,
+      "step": 10699
+    },
+    {
+      "epoch": 0.8248535306814678,
+      "grad_norm": 4.111172676086426,
+      "learning_rate": 7.832104767055376e-07,
+      "loss": 0.9342,
+      "step": 10700
+    },
+    {
+      "epoch": 0.8249306197964847,
+      "grad_norm": 3.4403626918792725,
+      "learning_rate": 7.82539752217828e-07,
+      "loss": 0.972,
+      "step": 10701
+    },
+    {
+      "epoch": 0.8250077089115017,
+      "grad_norm": 3.8715031147003174,
+      "learning_rate": 7.81869290666924e-07,
+      "loss": 0.7921,
+      "step": 10702
+    },
+    {
+      "epoch": 0.8250847980265187,
+      "grad_norm": 3.712979793548584,
+      "learning_rate": 7.811990920946216e-07,
+      "loss": 1.0237,
+      "step": 10703
+    },
+    {
+      "epoch": 0.8251618871415356,
+      "grad_norm": 3.841836452484131,
+      "learning_rate": 7.805291565427065e-07,
+      "loss": 1.0109,
+      "step": 10704
+    },
+    {
+      "epoch": 0.8252389762565526,
+      "grad_norm": 3.964392900466919,
+      "learning_rate": 7.79859484052945e-07,
+      "loss": 0.8749,
+      "step": 10705
+    },
+    {
+      "epoch": 0.8253160653715695,
+      "grad_norm": 3.964123010635376,
+      "learning_rate": 7.791900746670883e-07,
+      "loss": 0.9281,
+      "step": 10706
+    },
+    {
+      "epoch": 0.8253931544865865,
+      "grad_norm": 3.7393078804016113,
+      "learning_rate": 7.785209284268702e-07,
+      "loss": 0.8826,
+      "step": 10707
+    },
+    {
+      "epoch": 0.8254702436016035,
+      "grad_norm": 3.9628257751464844,
+      "learning_rate": 7.778520453740096e-07,
+      "loss": 0.9923,
+      "step": 10708
+    },
+    {
+      "epoch": 0.8255473327166204,
+      "grad_norm": 3.574497699737549,
+      "learning_rate": 7.771834255502059e-07,
+      "loss": 0.8425,
+      "step": 10709
+    },
+    {
+      "epoch": 0.8256244218316374,
+      "grad_norm": 3.801563262939453,
+      "learning_rate": 7.765150689971451e-07,
+      "loss": 0.8815,
+      "step": 10710
+    },
+    {
+      "epoch": 0.8257015109466543,
+      "grad_norm": 3.759230613708496,
+      "learning_rate": 7.758469757564957e-07,
+      "loss": 0.8957,
+      "step": 10711
+    },
+    {
+      "epoch": 0.8257786000616713,
+      "grad_norm": 3.4054291248321533,
+      "learning_rate": 7.751791458699099e-07,
+      "loss": 0.8231,
+      "step": 10712
+    },
+    {
+      "epoch": 0.8258556891766883,
+      "grad_norm": 3.554560422897339,
+      "learning_rate": 7.745115793790247e-07,
+      "loss": 0.9277,
+      "step": 10713
+    },
+    {
+      "epoch": 0.8259327782917052,
+      "grad_norm": 3.6641149520874023,
+      "learning_rate": 7.738442763254573e-07,
+      "loss": 0.9879,
+      "step": 10714
+    },
+    {
+      "epoch": 0.8260098674067222,
+      "grad_norm": 3.7249083518981934,
+      "learning_rate": 7.731772367508111e-07,
+      "loss": 0.8303,
+      "step": 10715
+    },
+    {
+      "epoch": 0.8260869565217391,
+      "grad_norm": 3.779303789138794,
+      "learning_rate": 7.725104606966727e-07,
+      "loss": 0.9721,
+      "step": 10716
+    },
+    {
+      "epoch": 0.826164045636756,
+      "grad_norm": 3.8608760833740234,
+      "learning_rate": 7.718439482046125e-07,
+      "loss": 0.871,
+      "step": 10717
+    },
+    {
+      "epoch": 0.8262411347517731,
+      "grad_norm": 3.9529666900634766,
+      "learning_rate": 7.711776993161841e-07,
+      "loss": 1.0389,
+      "step": 10718
+    },
+    {
+      "epoch": 0.82631822386679,
+      "grad_norm": 3.6020145416259766,
+      "learning_rate": 7.705117140729257e-07,
+      "loss": 0.7875,
+      "step": 10719
+    },
+    {
+      "epoch": 0.826395312981807,
+      "grad_norm": 4.178860664367676,
+      "learning_rate": 7.698459925163554e-07,
+      "loss": 1.0956,
+      "step": 10720
+    },
+    {
+      "epoch": 0.8264724020968239,
+      "grad_norm": 3.685406446456909,
+      "learning_rate": 7.691805346879794e-07,
+      "loss": 0.9837,
+      "step": 10721
+    },
+    {
+      "epoch": 0.8265494912118408,
+      "grad_norm": 3.7086429595947266,
+      "learning_rate": 7.685153406292845e-07,
+      "loss": 1.0095,
+      "step": 10722
+    },
+    {
+      "epoch": 0.8266265803268579,
+      "grad_norm": 3.816329002380371,
+      "learning_rate": 7.67850410381743e-07,
+      "loss": 0.9181,
+      "step": 10723
+    },
+    {
+      "epoch": 0.8267036694418748,
+      "grad_norm": 3.774265766143799,
+      "learning_rate": 7.671857439868107e-07,
+      "loss": 0.8757,
+      "step": 10724
+    },
+    {
+      "epoch": 0.8267807585568918,
+      "grad_norm": 4.03645133972168,
+      "learning_rate": 7.665213414859235e-07,
+      "loss": 0.9412,
+      "step": 10725
+    },
+    {
+      "epoch": 0.8268578476719087,
+      "grad_norm": 3.7689404487609863,
+      "learning_rate": 7.658572029205052e-07,
+      "loss": 0.8735,
+      "step": 10726
+    },
+    {
+      "epoch": 0.8269349367869256,
+      "grad_norm": 3.5467817783355713,
+      "learning_rate": 7.651933283319613e-07,
+      "loss": 0.8445,
+      "step": 10727
+    },
+    {
+      "epoch": 0.8270120259019427,
+      "grad_norm": 4.0023088455200195,
+      "learning_rate": 7.645297177616806e-07,
+      "loss": 0.9186,
+      "step": 10728
+    },
+    {
+      "epoch": 0.8270891150169596,
+      "grad_norm": 3.534449577331543,
+      "learning_rate": 7.638663712510358e-07,
+      "loss": 0.8886,
+      "step": 10729
+    },
+    {
+      "epoch": 0.8271662041319766,
+      "grad_norm": 3.6895134449005127,
+      "learning_rate": 7.632032888413848e-07,
+      "loss": 0.798,
+      "step": 10730
+    },
+    {
+      "epoch": 0.8272432932469935,
+      "grad_norm": 3.8588826656341553,
+      "learning_rate": 7.625404705740641e-07,
+      "loss": 0.8937,
+      "step": 10731
+    },
+    {
+      "epoch": 0.8273203823620104,
+      "grad_norm": 3.6557419300079346,
+      "learning_rate": 7.618779164903988e-07,
+      "loss": 0.8956,
+      "step": 10732
+    },
+    {
+      "epoch": 0.8273974714770275,
+      "grad_norm": 3.7859385013580322,
+      "learning_rate": 7.612156266316962e-07,
+      "loss": 0.927,
+      "step": 10733
+    },
+    {
+      "epoch": 0.8274745605920444,
+      "grad_norm": 3.476879358291626,
+      "learning_rate": 7.60553601039245e-07,
+      "loss": 0.869,
+      "step": 10734
+    },
+    {
+      "epoch": 0.8275516497070614,
+      "grad_norm": 3.6959054470062256,
+      "learning_rate": 7.598918397543209e-07,
+      "loss": 0.8605,
+      "step": 10735
+    },
+    {
+      "epoch": 0.8276287388220783,
+      "grad_norm": 3.761538505554199,
+      "learning_rate": 7.592303428181802e-07,
+      "loss": 0.9357,
+      "step": 10736
+    },
+    {
+      "epoch": 0.8277058279370952,
+      "grad_norm": 3.5777299404144287,
+      "learning_rate": 7.585691102720643e-07,
+      "loss": 0.8868,
+      "step": 10737
+    },
+    {
+      "epoch": 0.8277829170521123,
+      "grad_norm": 3.6994411945343018,
+      "learning_rate": 7.579081421571976e-07,
+      "loss": 0.8553,
+      "step": 10738
+    },
+    {
+      "epoch": 0.8278600061671292,
+      "grad_norm": 3.478304862976074,
+      "learning_rate": 7.572474385147888e-07,
+      "loss": 0.8501,
+      "step": 10739
+    },
+    {
+      "epoch": 0.8279370952821462,
+      "grad_norm": 3.7842917442321777,
+      "learning_rate": 7.56586999386027e-07,
+      "loss": 0.7966,
+      "step": 10740
+    },
+    {
+      "epoch": 0.8280141843971631,
+      "grad_norm": 3.9170117378234863,
+      "learning_rate": 7.559268248120887e-07,
+      "loss": 1.0096,
+      "step": 10741
+    },
+    {
+      "epoch": 0.82809127351218,
+      "grad_norm": 3.4382944107055664,
+      "learning_rate": 7.55266914834133e-07,
+      "loss": 0.8735,
+      "step": 10742
+    },
+    {
+      "epoch": 0.8281683626271971,
+      "grad_norm": 3.6370790004730225,
+      "learning_rate": 7.546072694933004e-07,
+      "loss": 0.9112,
+      "step": 10743
+    },
+    {
+      "epoch": 0.828245451742214,
+      "grad_norm": 4.061718463897705,
+      "learning_rate": 7.539478888307173e-07,
+      "loss": 1.021,
+      "step": 10744
+    },
+    {
+      "epoch": 0.828322540857231,
+      "grad_norm": 3.542180061340332,
+      "learning_rate": 7.532887728874933e-07,
+      "loss": 0.8531,
+      "step": 10745
+    },
+    {
+      "epoch": 0.8283996299722479,
+      "grad_norm": 3.725696325302124,
+      "learning_rate": 7.526299217047195e-07,
+      "loss": 1.0043,
+      "step": 10746
+    },
+    {
+      "epoch": 0.8284767190872648,
+      "grad_norm": 3.6887924671173096,
+      "learning_rate": 7.51971335323472e-07,
+      "loss": 0.9136,
+      "step": 10747
+    },
+    {
+      "epoch": 0.8285538082022819,
+      "grad_norm": 3.769517660140991,
+      "learning_rate": 7.513130137848101e-07,
+      "loss": 0.8516,
+      "step": 10748
+    },
+    {
+      "epoch": 0.8286308973172988,
+      "grad_norm": 3.488903522491455,
+      "learning_rate": 7.506549571297783e-07,
+      "loss": 0.7989,
+      "step": 10749
+    },
+    {
+      "epoch": 0.8287079864323158,
+      "grad_norm": 3.627612352371216,
+      "learning_rate": 7.499971653994026e-07,
+      "loss": 0.8622,
+      "step": 10750
+    },
+    {
+      "epoch": 0.8287850755473327,
+      "grad_norm": 3.5904018878936768,
+      "learning_rate": 7.493396386346913e-07,
+      "loss": 0.8568,
+      "step": 10751
+    },
+    {
+      "epoch": 0.8288621646623496,
+      "grad_norm": 3.8833725452423096,
+      "learning_rate": 7.486823768766388e-07,
+      "loss": 0.8596,
+      "step": 10752
+    },
+    {
+      "epoch": 0.8289392537773667,
+      "grad_norm": 3.913623809814453,
+      "learning_rate": 7.480253801662219e-07,
+      "loss": 1.0065,
+      "step": 10753
+    },
+    {
+      "epoch": 0.8290163428923836,
+      "grad_norm": 3.8650975227355957,
+      "learning_rate": 7.473686485444009e-07,
+      "loss": 0.8731,
+      "step": 10754
+    },
+    {
+      "epoch": 0.8290934320074006,
+      "grad_norm": 3.8849239349365234,
+      "learning_rate": 7.467121820521194e-07,
+      "loss": 0.9002,
+      "step": 10755
+    },
+    {
+      "epoch": 0.8291705211224175,
+      "grad_norm": 3.877471685409546,
+      "learning_rate": 7.46055980730307e-07,
+      "loss": 0.8938,
+      "step": 10756
+    },
+    {
+      "epoch": 0.8292476102374344,
+      "grad_norm": 3.7837424278259277,
+      "learning_rate": 7.454000446198706e-07,
+      "loss": 0.9011,
+      "step": 10757
+    },
+    {
+      "epoch": 0.8293246993524515,
+      "grad_norm": 3.6548290252685547,
+      "learning_rate": 7.447443737617066e-07,
+      "loss": 0.9058,
+      "step": 10758
+    },
+    {
+      "epoch": 0.8294017884674684,
+      "grad_norm": 3.6717207431793213,
+      "learning_rate": 7.440889681966923e-07,
+      "loss": 0.9047,
+      "step": 10759
+    },
+    {
+      "epoch": 0.8294788775824854,
+      "grad_norm": 3.3943278789520264,
+      "learning_rate": 7.434338279656889e-07,
+      "loss": 0.8617,
+      "step": 10760
+    },
+    {
+      "epoch": 0.8295559666975023,
+      "grad_norm": 3.89233660697937,
+      "learning_rate": 7.427789531095425e-07,
+      "loss": 0.9176,
+      "step": 10761
+    },
+    {
+      "epoch": 0.8296330558125192,
+      "grad_norm": 3.8601460456848145,
+      "learning_rate": 7.421243436690778e-07,
+      "loss": 0.8272,
+      "step": 10762
+    },
+    {
+      "epoch": 0.8297101449275363,
+      "grad_norm": 3.9159624576568604,
+      "learning_rate": 7.414699996851088e-07,
+      "loss": 0.977,
+      "step": 10763
+    },
+    {
+      "epoch": 0.8297872340425532,
+      "grad_norm": 3.4939470291137695,
+      "learning_rate": 7.408159211984294e-07,
+      "loss": 0.8306,
+      "step": 10764
+    },
+    {
+      "epoch": 0.8298643231575702,
+      "grad_norm": 3.7393009662628174,
+      "learning_rate": 7.401621082498189e-07,
+      "loss": 0.9836,
+      "step": 10765
+    },
+    {
+      "epoch": 0.8299414122725871,
+      "grad_norm": 3.700958013534546,
+      "learning_rate": 7.395085608800384e-07,
+      "loss": 0.854,
+      "step": 10766
+    },
+    {
+      "epoch": 0.830018501387604,
+      "grad_norm": 4.00097131729126,
+      "learning_rate": 7.388552791298343e-07,
+      "loss": 0.9573,
+      "step": 10767
+    },
+    {
+      "epoch": 0.830095590502621,
+      "grad_norm": 3.7124149799346924,
+      "learning_rate": 7.382022630399338e-07,
+      "loss": 0.925,
+      "step": 10768
+    },
+    {
+      "epoch": 0.830172679617638,
+      "grad_norm": 3.643216371536255,
+      "learning_rate": 7.375495126510496e-07,
+      "loss": 0.8348,
+      "step": 10769
+    },
+    {
+      "epoch": 0.830249768732655,
+      "grad_norm": 11.0379638671875,
+      "learning_rate": 7.368970280038778e-07,
+      "loss": 0.8255,
+      "step": 10770
+    },
+    {
+      "epoch": 0.8303268578476719,
+      "grad_norm": 3.5740654468536377,
+      "learning_rate": 7.362448091390966e-07,
+      "loss": 0.8926,
+      "step": 10771
+    },
+    {
+      "epoch": 0.8304039469626888,
+      "grad_norm": 3.759594440460205,
+      "learning_rate": 7.355928560973707e-07,
+      "loss": 0.8353,
+      "step": 10772
+    },
+    {
+      "epoch": 0.8304810360777058,
+      "grad_norm": 3.6525447368621826,
+      "learning_rate": 7.349411689193426e-07,
+      "loss": 0.8027,
+      "step": 10773
+    },
+    {
+      "epoch": 0.8305581251927228,
+      "grad_norm": 3.7419726848602295,
+      "learning_rate": 7.34289747645644e-07,
+      "loss": 0.8801,
+      "step": 10774
+    },
+    {
+      "epoch": 0.8306352143077398,
+      "grad_norm": 3.5162227153778076,
+      "learning_rate": 7.336385923168865e-07,
+      "loss": 0.8258,
+      "step": 10775
+    },
+    {
+      "epoch": 0.8307123034227567,
+      "grad_norm": 3.819016933441162,
+      "learning_rate": 7.329877029736665e-07,
+      "loss": 0.9143,
+      "step": 10776
+    },
+    {
+      "epoch": 0.8307893925377736,
+      "grad_norm": 3.756284475326538,
+      "learning_rate": 7.323370796565637e-07,
+      "loss": 0.9504,
+      "step": 10777
+    },
+    {
+      "epoch": 0.8308664816527906,
+      "grad_norm": 3.6436846256256104,
+      "learning_rate": 7.31686722406143e-07,
+      "loss": 0.9968,
+      "step": 10778
+    },
+    {
+      "epoch": 0.8309435707678076,
+      "grad_norm": 3.641303300857544,
+      "learning_rate": 7.310366312629475e-07,
+      "loss": 0.949,
+      "step": 10779
+    },
+    {
+      "epoch": 0.8310206598828246,
+      "grad_norm": 3.9875218868255615,
+      "learning_rate": 7.303868062675073e-07,
+      "loss": 0.9923,
+      "step": 10780
+    },
+    {
+      "epoch": 0.8310977489978415,
+      "grad_norm": 3.8598506450653076,
+      "learning_rate": 7.297372474603381e-07,
+      "loss": 1.0225,
+      "step": 10781
+    },
+    {
+      "epoch": 0.8311748381128584,
+      "grad_norm": 3.6692776679992676,
+      "learning_rate": 7.290879548819363e-07,
+      "loss": 0.8962,
+      "step": 10782
+    },
+    {
+      "epoch": 0.8312519272278754,
+      "grad_norm": 3.7915146350860596,
+      "learning_rate": 7.284389285727799e-07,
+      "loss": 0.953,
+      "step": 10783
+    },
+    {
+      "epoch": 0.8313290163428924,
+      "grad_norm": 3.9026849269866943,
+      "learning_rate": 7.277901685733335e-07,
+      "loss": 0.9768,
+      "step": 10784
+    },
+    {
+      "epoch": 0.8314061054579094,
+      "grad_norm": 3.8057961463928223,
+      "learning_rate": 7.271416749240435e-07,
+      "loss": 0.9183,
+      "step": 10785
+    },
+    {
+      "epoch": 0.8314831945729263,
+      "grad_norm": 3.617309808731079,
+      "learning_rate": 7.264934476653401e-07,
+      "loss": 0.9085,
+      "step": 10786
+    },
+    {
+      "epoch": 0.8315602836879432,
+      "grad_norm": 3.680267572402954,
+      "learning_rate": 7.258454868376385e-07,
+      "loss": 0.8003,
+      "step": 10787
+    },
+    {
+      "epoch": 0.8316373728029602,
+      "grad_norm": 3.6309967041015625,
+      "learning_rate": 7.251977924813336e-07,
+      "loss": 0.9443,
+      "step": 10788
+    },
+    {
+      "epoch": 0.8317144619179772,
+      "grad_norm": 3.5176992416381836,
+      "learning_rate": 7.245503646368063e-07,
+      "loss": 0.8685,
+      "step": 10789
+    },
+    {
+      "epoch": 0.8317915510329942,
+      "grad_norm": 3.7374002933502197,
+      "learning_rate": 7.239032033444205e-07,
+      "loss": 0.8479,
+      "step": 10790
+    },
+    {
+      "epoch": 0.8318686401480111,
+      "grad_norm": 4.19711446762085,
+      "learning_rate": 7.232563086445238e-07,
+      "loss": 0.9312,
+      "step": 10791
+    },
+    {
+      "epoch": 0.831945729263028,
+      "grad_norm": 3.7935373783111572,
+      "learning_rate": 7.226096805774463e-07,
+      "loss": 1.0013,
+      "step": 10792
+    },
+    {
+      "epoch": 0.832022818378045,
+      "grad_norm": 3.825516700744629,
+      "learning_rate": 7.21963319183503e-07,
+      "loss": 0.8755,
+      "step": 10793
+    },
+    {
+      "epoch": 0.832099907493062,
+      "grad_norm": 3.986665725708008,
+      "learning_rate": 7.213172245029892e-07,
+      "loss": 0.889,
+      "step": 10794
+    },
+    {
+      "epoch": 0.832176996608079,
+      "grad_norm": 3.7849297523498535,
+      "learning_rate": 7.206713965761864e-07,
+      "loss": 0.9522,
+      "step": 10795
+    },
+    {
+      "epoch": 0.8322540857230959,
+      "grad_norm": 3.8678035736083984,
+      "learning_rate": 7.200258354433593e-07,
+      "loss": 0.9428,
+      "step": 10796
+    },
+    {
+      "epoch": 0.8323311748381128,
+      "grad_norm": 3.5105371475219727,
+      "learning_rate": 7.193805411447546e-07,
+      "loss": 0.7586,
+      "step": 10797
+    },
+    {
+      "epoch": 0.8324082639531298,
+      "grad_norm": 4.333858966827393,
+      "learning_rate": 7.187355137206042e-07,
+      "loss": 0.8917,
+      "step": 10798
+    },
+    {
+      "epoch": 0.8324853530681467,
+      "grad_norm": 3.7516279220581055,
+      "learning_rate": 7.180907532111203e-07,
+      "loss": 0.9158,
+      "step": 10799
+    },
+    {
+      "epoch": 0.8325624421831638,
+      "grad_norm": 3.8405489921569824,
+      "learning_rate": 7.174462596565012e-07,
+      "loss": 0.9124,
+      "step": 10800
+    },
+    {
+      "epoch": 0.8326395312981807,
+      "grad_norm": 4.030445098876953,
+      "learning_rate": 7.168020330969283e-07,
+      "loss": 0.9262,
+      "step": 10801
+    },
+    {
+      "epoch": 0.8327166204131976,
+      "grad_norm": 4.261043548583984,
+      "learning_rate": 7.161580735725648e-07,
+      "loss": 1.0228,
+      "step": 10802
+    },
+    {
+      "epoch": 0.8327937095282146,
+      "grad_norm": 4.2174973487854,
+      "learning_rate": 7.155143811235593e-07,
+      "loss": 1.0124,
+      "step": 10803
+    },
+    {
+      "epoch": 0.8328707986432315,
+      "grad_norm": 4.092016696929932,
+      "learning_rate": 7.14870955790043e-07,
+      "loss": 0.8879,
+      "step": 10804
+    },
+    {
+      "epoch": 0.8329478877582486,
+      "grad_norm": 3.6013119220733643,
+      "learning_rate": 7.142277976121287e-07,
+      "loss": 0.8734,
+      "step": 10805
+    },
+    {
+      "epoch": 0.8330249768732655,
+      "grad_norm": 4.047834396362305,
+      "learning_rate": 7.135849066299144e-07,
+      "loss": 0.9181,
+      "step": 10806
+    },
+    {
+      "epoch": 0.8331020659882824,
+      "grad_norm": 3.696688413619995,
+      "learning_rate": 7.129422828834809e-07,
+      "loss": 0.8078,
+      "step": 10807
+    },
+    {
+      "epoch": 0.8331791551032994,
+      "grad_norm": 3.918166399002075,
+      "learning_rate": 7.122999264128933e-07,
+      "loss": 0.8543,
+      "step": 10808
+    },
+    {
+      "epoch": 0.8332562442183163,
+      "grad_norm": 3.7116942405700684,
+      "learning_rate": 7.116578372581995e-07,
+      "loss": 0.8861,
+      "step": 10809
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 3.7015819549560547,
+      "learning_rate": 7.110160154594287e-07,
+      "loss": 0.8536,
+      "step": 10810
+    },
+    {
+      "epoch": 0.8334104224483503,
+      "grad_norm": 3.7466604709625244,
+      "learning_rate": 7.10374461056596e-07,
+      "loss": 0.9696,
+      "step": 10811
+    },
+    {
+      "epoch": 0.8334875115633672,
+      "grad_norm": 3.6178178787231445,
+      "learning_rate": 7.097331740896995e-07,
+      "loss": 0.9173,
+      "step": 10812
+    },
+    {
+      "epoch": 0.8335646006783842,
+      "grad_norm": 3.8105711936950684,
+      "learning_rate": 7.090921545987195e-07,
+      "loss": 0.9118,
+      "step": 10813
+    },
+    {
+      "epoch": 0.8336416897934011,
+      "grad_norm": 3.8142049312591553,
+      "learning_rate": 7.0845140262362e-07,
+      "loss": 0.7892,
+      "step": 10814
+    },
+    {
+      "epoch": 0.8337187789084182,
+      "grad_norm": 3.651261329650879,
+      "learning_rate": 7.078109182043508e-07,
+      "loss": 0.9293,
+      "step": 10815
+    },
+    {
+      "epoch": 0.8337958680234351,
+      "grad_norm": 3.8389205932617188,
+      "learning_rate": 7.071707013808399e-07,
+      "loss": 0.8746,
+      "step": 10816
+    },
+    {
+      "epoch": 0.833872957138452,
+      "grad_norm": 4.682302951812744,
+      "learning_rate": 7.065307521930026e-07,
+      "loss": 0.8675,
+      "step": 10817
+    },
+    {
+      "epoch": 0.833950046253469,
+      "grad_norm": 3.9559519290924072,
+      "learning_rate": 7.058910706807359e-07,
+      "loss": 0.9481,
+      "step": 10818
+    },
+    {
+      "epoch": 0.8340271353684859,
+      "grad_norm": 3.810323715209961,
+      "learning_rate": 7.052516568839218e-07,
+      "loss": 0.9161,
+      "step": 10819
+    },
+    {
+      "epoch": 0.834104224483503,
+      "grad_norm": 4.13281774520874,
+      "learning_rate": 7.046125108424245e-07,
+      "loss": 0.9867,
+      "step": 10820
+    },
+    {
+      "epoch": 0.8341813135985199,
+      "grad_norm": 3.446863889694214,
+      "learning_rate": 7.039736325960899e-07,
+      "loss": 0.8842,
+      "step": 10821
+    },
+    {
+      "epoch": 0.8342584027135368,
+      "grad_norm": 3.840688943862915,
+      "learning_rate": 7.033350221847496e-07,
+      "loss": 0.9211,
+      "step": 10822
+    },
+    {
+      "epoch": 0.8343354918285538,
+      "grad_norm": 3.8792099952697754,
+      "learning_rate": 7.026966796482177e-07,
+      "loss": 0.8376,
+      "step": 10823
+    },
+    {
+      "epoch": 0.8344125809435707,
+      "grad_norm": 3.9304957389831543,
+      "learning_rate": 7.020586050262912e-07,
+      "loss": 0.919,
+      "step": 10824
+    },
+    {
+      "epoch": 0.8344896700585878,
+      "grad_norm": 3.9042816162109375,
+      "learning_rate": 7.014207983587517e-07,
+      "loss": 0.8255,
+      "step": 10825
+    },
+    {
+      "epoch": 0.8345667591736047,
+      "grad_norm": 3.54953670501709,
+      "learning_rate": 7.00783259685362e-07,
+      "loss": 0.8494,
+      "step": 10826
+    },
+    {
+      "epoch": 0.8346438482886216,
+      "grad_norm": 3.699927568435669,
+      "learning_rate": 7.0014598904587e-07,
+      "loss": 0.7763,
+      "step": 10827
+    },
+    {
+      "epoch": 0.8347209374036386,
+      "grad_norm": 3.908418655395508,
+      "learning_rate": 6.995089864800059e-07,
+      "loss": 0.9824,
+      "step": 10828
+    },
+    {
+      "epoch": 0.8347980265186555,
+      "grad_norm": 3.541356325149536,
+      "learning_rate": 6.988722520274838e-07,
+      "loss": 0.7651,
+      "step": 10829
+    },
+    {
+      "epoch": 0.8348751156336726,
+      "grad_norm": 3.6384029388427734,
+      "learning_rate": 6.98235785728002e-07,
+      "loss": 0.9051,
+      "step": 10830
+    },
+    {
+      "epoch": 0.8349522047486895,
+      "grad_norm": 3.667262315750122,
+      "learning_rate": 6.975995876212383e-07,
+      "loss": 0.9101,
+      "step": 10831
+    },
+    {
+      "epoch": 0.8350292938637064,
+      "grad_norm": 3.559422492980957,
+      "learning_rate": 6.969636577468575e-07,
+      "loss": 0.8595,
+      "step": 10832
+    },
+    {
+      "epoch": 0.8351063829787234,
+      "grad_norm": 4.325561046600342,
+      "learning_rate": 6.963279961445068e-07,
+      "loss": 0.9312,
+      "step": 10833
+    },
+    {
+      "epoch": 0.8351834720937403,
+      "grad_norm": 4.272604465484619,
+      "learning_rate": 6.956926028538163e-07,
+      "loss": 0.8992,
+      "step": 10834
+    },
+    {
+      "epoch": 0.8352605612087574,
+      "grad_norm": 3.4892845153808594,
+      "learning_rate": 6.950574779144004e-07,
+      "loss": 0.8594,
+      "step": 10835
+    },
+    {
+      "epoch": 0.8353376503237743,
+      "grad_norm": 3.5458579063415527,
+      "learning_rate": 6.944226213658534e-07,
+      "loss": 0.9747,
+      "step": 10836
+    },
+    {
+      "epoch": 0.8354147394387912,
+      "grad_norm": 3.427591562271118,
+      "learning_rate": 6.937880332477576e-07,
+      "loss": 0.9044,
+      "step": 10837
+    },
+    {
+      "epoch": 0.8354918285538082,
+      "grad_norm": 4.049854755401611,
+      "learning_rate": 6.931537135996747e-07,
+      "loss": 0.9564,
+      "step": 10838
+    },
+    {
+      "epoch": 0.8355689176688251,
+      "grad_norm": 3.7413408756256104,
+      "learning_rate": 6.925196624611525e-07,
+      "loss": 0.9341,
+      "step": 10839
+    },
+    {
+      "epoch": 0.8356460067838422,
+      "grad_norm": 3.709747076034546,
+      "learning_rate": 6.918858798717204e-07,
+      "loss": 0.9402,
+      "step": 10840
+    },
+    {
+      "epoch": 0.8357230958988591,
+      "grad_norm": 4.0322465896606445,
+      "learning_rate": 6.912523658708919e-07,
+      "loss": 0.91,
+      "step": 10841
+    },
+    {
+      "epoch": 0.8358001850138761,
+      "grad_norm": 3.899702310562134,
+      "learning_rate": 6.906191204981621e-07,
+      "loss": 0.9262,
+      "step": 10842
+    },
+    {
+      "epoch": 0.835877274128893,
+      "grad_norm": 3.4912753105163574,
+      "learning_rate": 6.899861437930116e-07,
+      "loss": 0.8982,
+      "step": 10843
+    },
+    {
+      "epoch": 0.8359543632439099,
+      "grad_norm": 3.437044620513916,
+      "learning_rate": 6.893534357949022e-07,
+      "loss": 0.8238,
+      "step": 10844
+    },
+    {
+      "epoch": 0.836031452358927,
+      "grad_norm": 3.966545820236206,
+      "learning_rate": 6.887209965432812e-07,
+      "loss": 0.977,
+      "step": 10845
+    },
+    {
+      "epoch": 0.8361085414739439,
+      "grad_norm": 3.3362553119659424,
+      "learning_rate": 6.880888260775786e-07,
+      "loss": 0.877,
+      "step": 10846
+    },
+    {
+      "epoch": 0.8361856305889609,
+      "grad_norm": 3.762795925140381,
+      "learning_rate": 6.874569244372042e-07,
+      "loss": 0.9176,
+      "step": 10847
+    },
+    {
+      "epoch": 0.8362627197039778,
+      "grad_norm": 3.6273088455200195,
+      "learning_rate": 6.868252916615553e-07,
+      "loss": 0.8417,
+      "step": 10848
+    },
+    {
+      "epoch": 0.8363398088189947,
+      "grad_norm": 3.5637712478637695,
+      "learning_rate": 6.861939277900115e-07,
+      "loss": 0.8944,
+      "step": 10849
+    },
+    {
+      "epoch": 0.8364168979340117,
+      "grad_norm": 3.720207929611206,
+      "learning_rate": 6.855628328619341e-07,
+      "loss": 1.0064,
+      "step": 10850
+    },
+    {
+      "epoch": 0.8364939870490287,
+      "grad_norm": 3.5952982902526855,
+      "learning_rate": 6.849320069166693e-07,
+      "loss": 0.9475,
+      "step": 10851
+    },
+    {
+      "epoch": 0.8365710761640457,
+      "grad_norm": 3.7892141342163086,
+      "learning_rate": 6.843014499935463e-07,
+      "loss": 0.9439,
+      "step": 10852
+    },
+    {
+      "epoch": 0.8366481652790626,
+      "grad_norm": 3.6238391399383545,
+      "learning_rate": 6.836711621318753e-07,
+      "loss": 0.8605,
+      "step": 10853
+    },
+    {
+      "epoch": 0.8367252543940795,
+      "grad_norm": 3.9006173610687256,
+      "learning_rate": 6.830411433709527e-07,
+      "loss": 0.8448,
+      "step": 10854
+    },
+    {
+      "epoch": 0.8368023435090965,
+      "grad_norm": 3.381910800933838,
+      "learning_rate": 6.824113937500565e-07,
+      "loss": 0.9104,
+      "step": 10855
+    },
+    {
+      "epoch": 0.8368794326241135,
+      "grad_norm": 3.998650312423706,
+      "learning_rate": 6.817819133084485e-07,
+      "loss": 1.0149,
+      "step": 10856
+    },
+    {
+      "epoch": 0.8369565217391305,
+      "grad_norm": 3.4496870040893555,
+      "learning_rate": 6.811527020853748e-07,
+      "loss": 0.8017,
+      "step": 10857
+    },
+    {
+      "epoch": 0.8370336108541474,
+      "grad_norm": 3.7553465366363525,
+      "learning_rate": 6.805237601200615e-07,
+      "loss": 0.7632,
+      "step": 10858
+    },
+    {
+      "epoch": 0.8371106999691643,
+      "grad_norm": 4.353603839874268,
+      "learning_rate": 6.798950874517201e-07,
+      "loss": 0.9318,
+      "step": 10859
+    },
+    {
+      "epoch": 0.8371877890841813,
+      "grad_norm": 3.8863165378570557,
+      "learning_rate": 6.792666841195455e-07,
+      "loss": 0.904,
+      "step": 10860
+    },
+    {
+      "epoch": 0.8372648781991983,
+      "grad_norm": 3.8300702571868896,
+      "learning_rate": 6.786385501627157e-07,
+      "loss": 0.8406,
+      "step": 10861
+    },
+    {
+      "epoch": 0.8373419673142153,
+      "grad_norm": 3.770839214324951,
+      "learning_rate": 6.780106856203916e-07,
+      "loss": 0.9262,
+      "step": 10862
+    },
+    {
+      "epoch": 0.8374190564292322,
+      "grad_norm": 3.5766117572784424,
+      "learning_rate": 6.77383090531718e-07,
+      "loss": 0.868,
+      "step": 10863
+    },
+    {
+      "epoch": 0.8374961455442491,
+      "grad_norm": 3.751206159591675,
+      "learning_rate": 6.767557649358203e-07,
+      "loss": 0.8759,
+      "step": 10864
+    },
+    {
+      "epoch": 0.8375732346592661,
+      "grad_norm": 3.9734339714050293,
+      "learning_rate": 6.7612870887181e-07,
+      "loss": 0.8378,
+      "step": 10865
+    },
+    {
+      "epoch": 0.8376503237742831,
+      "grad_norm": 3.7045814990997314,
+      "learning_rate": 6.755019223787807e-07,
+      "loss": 0.9453,
+      "step": 10866
+    },
+    {
+      "epoch": 0.8377274128893001,
+      "grad_norm": 3.88385272026062,
+      "learning_rate": 6.748754054958095e-07,
+      "loss": 0.9384,
+      "step": 10867
+    },
+    {
+      "epoch": 0.837804502004317,
+      "grad_norm": 4.15552282333374,
+      "learning_rate": 6.742491582619559e-07,
+      "loss": 0.9795,
+      "step": 10868
+    },
+    {
+      "epoch": 0.8378815911193339,
+      "grad_norm": 3.2853026390075684,
+      "learning_rate": 6.736231807162641e-07,
+      "loss": 0.843,
+      "step": 10869
+    },
+    {
+      "epoch": 0.8379586802343509,
+      "grad_norm": 3.701760768890381,
+      "learning_rate": 6.7299747289776e-07,
+      "loss": 0.9865,
+      "step": 10870
+    },
+    {
+      "epoch": 0.8380357693493679,
+      "grad_norm": 3.7172200679779053,
+      "learning_rate": 6.723720348454538e-07,
+      "loss": 0.8868,
+      "step": 10871
+    },
+    {
+      "epoch": 0.8381128584643849,
+      "grad_norm": 3.8276724815368652,
+      "learning_rate": 6.717468665983384e-07,
+      "loss": 0.9274,
+      "step": 10872
+    },
+    {
+      "epoch": 0.8381899475794018,
+      "grad_norm": 3.7223312854766846,
+      "learning_rate": 6.711219681953885e-07,
+      "loss": 0.8895,
+      "step": 10873
+    },
+    {
+      "epoch": 0.8382670366944187,
+      "grad_norm": 3.972644329071045,
+      "learning_rate": 6.704973396755638e-07,
+      "loss": 0.9633,
+      "step": 10874
+    },
+    {
+      "epoch": 0.8383441258094357,
+      "grad_norm": 3.7583634853363037,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.8806,
+      "step": 10875
+    },
+    {
+      "epoch": 0.8384212149244527,
+      "grad_norm": 4.294461727142334,
+      "learning_rate": 6.692488924410434e-07,
+      "loss": 0.957,
+      "step": 10876
+    },
+    {
+      "epoch": 0.8384983040394697,
+      "grad_norm": 3.7632694244384766,
+      "learning_rate": 6.686250738041816e-07,
+      "loss": 0.8539,
+      "step": 10877
+    },
+    {
+      "epoch": 0.8385753931544866,
+      "grad_norm": 3.696632146835327,
+      "learning_rate": 6.68001525206115e-07,
+      "loss": 0.9736,
+      "step": 10878
+    },
+    {
+      "epoch": 0.8386524822695035,
+      "grad_norm": 3.7102134227752686,
+      "learning_rate": 6.673782466857165e-07,
+      "loss": 0.9112,
+      "step": 10879
+    },
+    {
+      "epoch": 0.8387295713845205,
+      "grad_norm": 4.029872417449951,
+      "learning_rate": 6.667552382818449e-07,
+      "loss": 0.9195,
+      "step": 10880
+    },
+    {
+      "epoch": 0.8388066604995374,
+      "grad_norm": 4.231304168701172,
+      "learning_rate": 6.661325000333419e-07,
+      "loss": 0.9635,
+      "step": 10881
+    },
+    {
+      "epoch": 0.8388837496145545,
+      "grad_norm": 3.9188613891601562,
+      "learning_rate": 6.655100319790314e-07,
+      "loss": 0.7859,
+      "step": 10882
+    },
+    {
+      "epoch": 0.8389608387295714,
+      "grad_norm": 3.596547842025757,
+      "learning_rate": 6.64887834157723e-07,
+      "loss": 0.8746,
+      "step": 10883
+    },
+    {
+      "epoch": 0.8390379278445883,
+      "grad_norm": 3.9885101318359375,
+      "learning_rate": 6.642659066082046e-07,
+      "loss": 0.922,
+      "step": 10884
+    },
+    {
+      "epoch": 0.8391150169596053,
+      "grad_norm": 3.787560224533081,
+      "learning_rate": 6.636442493692518e-07,
+      "loss": 0.894,
+      "step": 10885
+    },
+    {
+      "epoch": 0.8391921060746222,
+      "grad_norm": 3.884455680847168,
+      "learning_rate": 6.630228624796215e-07,
+      "loss": 0.9345,
+      "step": 10886
+    },
+    {
+      "epoch": 0.8392691951896393,
+      "grad_norm": 3.6436421871185303,
+      "learning_rate": 6.624017459780541e-07,
+      "loss": 0.837,
+      "step": 10887
+    },
+    {
+      "epoch": 0.8393462843046562,
+      "grad_norm": 3.4954371452331543,
+      "learning_rate": 6.617808999032727e-07,
+      "loss": 0.8966,
+      "step": 10888
+    },
+    {
+      "epoch": 0.8394233734196731,
+      "grad_norm": 4.002804279327393,
+      "learning_rate": 6.611603242939846e-07,
+      "loss": 1.0232,
+      "step": 10889
+    },
+    {
+      "epoch": 0.8395004625346901,
+      "grad_norm": 3.798150062561035,
+      "learning_rate": 6.605400191888784e-07,
+      "loss": 0.9284,
+      "step": 10890
+    },
+    {
+      "epoch": 0.839577551649707,
+      "grad_norm": 3.5775742530822754,
+      "learning_rate": 6.599199846266274e-07,
+      "loss": 0.8892,
+      "step": 10891
+    },
+    {
+      "epoch": 0.8396546407647241,
+      "grad_norm": 3.491213321685791,
+      "learning_rate": 6.59300220645887e-07,
+      "loss": 0.9179,
+      "step": 10892
+    },
+    {
+      "epoch": 0.839731729879741,
+      "grad_norm": 3.7937722206115723,
+      "learning_rate": 6.586807272852969e-07,
+      "loss": 0.9093,
+      "step": 10893
+    },
+    {
+      "epoch": 0.8398088189947579,
+      "grad_norm": 4.150206089019775,
+      "learning_rate": 6.580615045834803e-07,
+      "loss": 0.9983,
+      "step": 10894
+    },
+    {
+      "epoch": 0.8398859081097749,
+      "grad_norm": 3.7913031578063965,
+      "learning_rate": 6.574425525790407e-07,
+      "loss": 0.8851,
+      "step": 10895
+    },
+    {
+      "epoch": 0.8399629972247918,
+      "grad_norm": 3.3941149711608887,
+      "learning_rate": 6.568238713105668e-07,
+      "loss": 0.8295,
+      "step": 10896
+    },
+    {
+      "epoch": 0.8400400863398089,
+      "grad_norm": 3.6495449542999268,
+      "learning_rate": 6.562054608166307e-07,
+      "loss": 0.7818,
+      "step": 10897
+    },
+    {
+      "epoch": 0.8401171754548258,
+      "grad_norm": 3.510993719100952,
+      "learning_rate": 6.555873211357872e-07,
+      "loss": 0.8457,
+      "step": 10898
+    },
+    {
+      "epoch": 0.8401942645698427,
+      "grad_norm": 3.769568920135498,
+      "learning_rate": 6.549694523065742e-07,
+      "loss": 0.9033,
+      "step": 10899
+    },
+    {
+      "epoch": 0.8402713536848597,
+      "grad_norm": 3.4650299549102783,
+      "learning_rate": 6.543518543675132e-07,
+      "loss": 0.7952,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8403484427998766,
+      "grad_norm": 3.828594923019409,
+      "learning_rate": 6.537345273571061e-07,
+      "loss": 1.0226,
+      "step": 10901
+    },
+    {
+      "epoch": 0.8404255319148937,
+      "grad_norm": 4.020386695861816,
+      "learning_rate": 6.531174713138416e-07,
+      "loss": 0.9334,
+      "step": 10902
+    },
+    {
+      "epoch": 0.8405026210299106,
+      "grad_norm": 4.063125133514404,
+      "learning_rate": 6.525006862761895e-07,
+      "loss": 0.8969,
+      "step": 10903
+    },
+    {
+      "epoch": 0.8405797101449275,
+      "grad_norm": 4.093360900878906,
+      "learning_rate": 6.51884172282603e-07,
+      "loss": 1.0452,
+      "step": 10904
+    },
+    {
+      "epoch": 0.8406567992599445,
+      "grad_norm": 3.666809320449829,
+      "learning_rate": 6.512679293715208e-07,
+      "loss": 0.9207,
+      "step": 10905
+    },
+    {
+      "epoch": 0.8407338883749614,
+      "grad_norm": 3.7799835205078125,
+      "learning_rate": 6.506519575813591e-07,
+      "loss": 0.8436,
+      "step": 10906
+    },
+    {
+      "epoch": 0.8408109774899785,
+      "grad_norm": 3.6833157539367676,
+      "learning_rate": 6.500362569505215e-07,
+      "loss": 0.9263,
+      "step": 10907
+    },
+    {
+      "epoch": 0.8408880666049954,
+      "grad_norm": 3.6458330154418945,
+      "learning_rate": 6.494208275173947e-07,
+      "loss": 0.9459,
+      "step": 10908
+    },
+    {
+      "epoch": 0.8409651557200123,
+      "grad_norm": 3.5508716106414795,
+      "learning_rate": 6.488056693203471e-07,
+      "loss": 0.7733,
+      "step": 10909
+    },
+    {
+      "epoch": 0.8410422448350293,
+      "grad_norm": 3.958599805831909,
+      "learning_rate": 6.481907823977307e-07,
+      "loss": 0.9815,
+      "step": 10910
+    },
+    {
+      "epoch": 0.8411193339500462,
+      "grad_norm": 4.111875057220459,
+      "learning_rate": 6.47576166787881e-07,
+      "loss": 0.9399,
+      "step": 10911
+    },
+    {
+      "epoch": 0.8411964230650633,
+      "grad_norm": 3.8634867668151855,
+      "learning_rate": 6.469618225291141e-07,
+      "loss": 0.8141,
+      "step": 10912
+    },
+    {
+      "epoch": 0.8412735121800802,
+      "grad_norm": 4.0127716064453125,
+      "learning_rate": 6.463477496597332e-07,
+      "loss": 0.9476,
+      "step": 10913
+    },
+    {
+      "epoch": 0.8413506012950971,
+      "grad_norm": 3.996286153793335,
+      "learning_rate": 6.45733948218022e-07,
+      "loss": 0.9159,
+      "step": 10914
+    },
+    {
+      "epoch": 0.8414276904101141,
+      "grad_norm": 3.8326454162597656,
+      "learning_rate": 6.451204182422488e-07,
+      "loss": 0.8948,
+      "step": 10915
+    },
+    {
+      "epoch": 0.841504779525131,
+      "grad_norm": 3.672430992126465,
+      "learning_rate": 6.44507159770662e-07,
+      "loss": 0.9338,
+      "step": 10916
+    },
+    {
+      "epoch": 0.8415818686401481,
+      "grad_norm": 3.8770806789398193,
+      "learning_rate": 6.43894172841496e-07,
+      "loss": 0.9311,
+      "step": 10917
+    },
+    {
+      "epoch": 0.841658957755165,
+      "grad_norm": 3.768202304840088,
+      "learning_rate": 6.43281457492968e-07,
+      "loss": 0.9567,
+      "step": 10918
+    },
+    {
+      "epoch": 0.8417360468701819,
+      "grad_norm": 4.294236660003662,
+      "learning_rate": 6.426690137632763e-07,
+      "loss": 0.9294,
+      "step": 10919
+    },
+    {
+      "epoch": 0.8418131359851989,
+      "grad_norm": 3.6829004287719727,
+      "learning_rate": 6.420568416906059e-07,
+      "loss": 0.9141,
+      "step": 10920
+    },
+    {
+      "epoch": 0.8418902251002158,
+      "grad_norm": 4.125067710876465,
+      "learning_rate": 6.414449413131202e-07,
+      "loss": 0.9829,
+      "step": 10921
+    },
+    {
+      "epoch": 0.8419673142152329,
+      "grad_norm": 3.662050247192383,
+      "learning_rate": 6.408333126689686e-07,
+      "loss": 0.9487,
+      "step": 10922
+    },
+    {
+      "epoch": 0.8420444033302498,
+      "grad_norm": 4.144391059875488,
+      "learning_rate": 6.402219557962835e-07,
+      "loss": 0.9397,
+      "step": 10923
+    },
+    {
+      "epoch": 0.8421214924452667,
+      "grad_norm": 4.148819446563721,
+      "learning_rate": 6.396108707331794e-07,
+      "loss": 1.0001,
+      "step": 10924
+    },
+    {
+      "epoch": 0.8421985815602837,
+      "grad_norm": 3.679710865020752,
+      "learning_rate": 6.390000575177546e-07,
+      "loss": 0.9325,
+      "step": 10925
+    },
+    {
+      "epoch": 0.8422756706753006,
+      "grad_norm": 3.7735555171966553,
+      "learning_rate": 6.38389516188091e-07,
+      "loss": 0.9016,
+      "step": 10926
+    },
+    {
+      "epoch": 0.8423527597903177,
+      "grad_norm": 4.009374618530273,
+      "learning_rate": 6.377792467822502e-07,
+      "loss": 0.9601,
+      "step": 10927
+    },
+    {
+      "epoch": 0.8424298489053346,
+      "grad_norm": 3.806119441986084,
+      "learning_rate": 6.371692493382814e-07,
+      "loss": 0.9,
+      "step": 10928
+    },
+    {
+      "epoch": 0.8425069380203515,
+      "grad_norm": 3.881964921951294,
+      "learning_rate": 6.36559523894214e-07,
+      "loss": 0.9717,
+      "step": 10929
+    },
+    {
+      "epoch": 0.8425840271353685,
+      "grad_norm": 3.9612395763397217,
+      "learning_rate": 6.359500704880617e-07,
+      "loss": 0.955,
+      "step": 10930
+    },
+    {
+      "epoch": 0.8426611162503854,
+      "grad_norm": 3.793069839477539,
+      "learning_rate": 6.353408891578212e-07,
+      "loss": 0.8759,
+      "step": 10931
+    },
+    {
+      "epoch": 0.8427382053654024,
+      "grad_norm": 3.9942023754119873,
+      "learning_rate": 6.347319799414702e-07,
+      "loss": 0.9591,
+      "step": 10932
+    },
+    {
+      "epoch": 0.8428152944804194,
+      "grad_norm": 3.749751567840576,
+      "learning_rate": 6.341233428769722e-07,
+      "loss": 0.8971,
+      "step": 10933
+    },
+    {
+      "epoch": 0.8428923835954363,
+      "grad_norm": 3.6659231185913086,
+      "learning_rate": 6.33514978002272e-07,
+      "loss": 0.7227,
+      "step": 10934
+    },
+    {
+      "epoch": 0.8429694727104533,
+      "grad_norm": 3.700681686401367,
+      "learning_rate": 6.329068853552983e-07,
+      "loss": 0.8878,
+      "step": 10935
+    },
+    {
+      "epoch": 0.8430465618254702,
+      "grad_norm": 3.738607883453369,
+      "learning_rate": 6.322990649739624e-07,
+      "loss": 0.924,
+      "step": 10936
+    },
+    {
+      "epoch": 0.8431236509404872,
+      "grad_norm": 3.6381099224090576,
+      "learning_rate": 6.316915168961602e-07,
+      "loss": 0.9039,
+      "step": 10937
+    },
+    {
+      "epoch": 0.8432007400555042,
+      "grad_norm": 3.76474928855896,
+      "learning_rate": 6.310842411597667e-07,
+      "loss": 0.9657,
+      "step": 10938
+    },
+    {
+      "epoch": 0.8432778291705211,
+      "grad_norm": 3.532024383544922,
+      "learning_rate": 6.304772378026441e-07,
+      "loss": 0.8998,
+      "step": 10939
+    },
+    {
+      "epoch": 0.8433549182855381,
+      "grad_norm": 3.453786611557007,
+      "learning_rate": 6.298705068626348e-07,
+      "loss": 0.7852,
+      "step": 10940
+    },
+    {
+      "epoch": 0.843432007400555,
+      "grad_norm": 3.8487155437469482,
+      "learning_rate": 6.292640483775664e-07,
+      "loss": 0.8156,
+      "step": 10941
+    },
+    {
+      "epoch": 0.843509096515572,
+      "grad_norm": 3.691965341567993,
+      "learning_rate": 6.286578623852485e-07,
+      "loss": 0.9317,
+      "step": 10942
+    },
+    {
+      "epoch": 0.843586185630589,
+      "grad_norm": 3.6872313022613525,
+      "learning_rate": 6.280519489234721e-07,
+      "loss": 0.8926,
+      "step": 10943
+    },
+    {
+      "epoch": 0.8436632747456059,
+      "grad_norm": 3.968344211578369,
+      "learning_rate": 6.274463080300142e-07,
+      "loss": 0.9674,
+      "step": 10944
+    },
+    {
+      "epoch": 0.8437403638606229,
+      "grad_norm": 3.671273708343506,
+      "learning_rate": 6.268409397426323e-07,
+      "loss": 0.8654,
+      "step": 10945
+    },
+    {
+      "epoch": 0.8438174529756398,
+      "grad_norm": 3.5088984966278076,
+      "learning_rate": 6.26235844099069e-07,
+      "loss": 0.8773,
+      "step": 10946
+    },
+    {
+      "epoch": 0.8438945420906568,
+      "grad_norm": 3.9786536693573,
+      "learning_rate": 6.256310211370486e-07,
+      "loss": 0.8877,
+      "step": 10947
+    },
+    {
+      "epoch": 0.8439716312056738,
+      "grad_norm": 3.580746650695801,
+      "learning_rate": 6.2502647089428e-07,
+      "loss": 0.888,
+      "step": 10948
+    },
+    {
+      "epoch": 0.8440487203206907,
+      "grad_norm": 3.8491809368133545,
+      "learning_rate": 6.244221934084504e-07,
+      "loss": 0.9276,
+      "step": 10949
+    },
+    {
+      "epoch": 0.8441258094357077,
+      "grad_norm": 3.9452064037323,
+      "learning_rate": 6.238181887172362e-07,
+      "loss": 0.9588,
+      "step": 10950
+    },
+    {
+      "epoch": 0.8442028985507246,
+      "grad_norm": 3.862786054611206,
+      "learning_rate": 6.232144568582926e-07,
+      "loss": 0.8399,
+      "step": 10951
+    },
+    {
+      "epoch": 0.8442799876657416,
+      "grad_norm": 3.535684823989868,
+      "learning_rate": 6.226109978692596e-07,
+      "loss": 0.8093,
+      "step": 10952
+    },
+    {
+      "epoch": 0.8443570767807586,
+      "grad_norm": 3.716644048690796,
+      "learning_rate": 6.220078117877615e-07,
+      "loss": 0.9213,
+      "step": 10953
+    },
+    {
+      "epoch": 0.8444341658957755,
+      "grad_norm": 3.8293979167938232,
+      "learning_rate": 6.214048986514004e-07,
+      "loss": 0.9022,
+      "step": 10954
+    },
+    {
+      "epoch": 0.8445112550107925,
+      "grad_norm": 3.8130550384521484,
+      "learning_rate": 6.208022584977668e-07,
+      "loss": 0.9534,
+      "step": 10955
+    },
+    {
+      "epoch": 0.8445883441258094,
+      "grad_norm": 3.5550150871276855,
+      "learning_rate": 6.201998913644319e-07,
+      "loss": 0.8563,
+      "step": 10956
+    },
+    {
+      "epoch": 0.8446654332408264,
+      "grad_norm": 3.5891706943511963,
+      "learning_rate": 6.195977972889505e-07,
+      "loss": 0.8518,
+      "step": 10957
+    },
+    {
+      "epoch": 0.8447425223558434,
+      "grad_norm": 3.988762855529785,
+      "learning_rate": 6.189959763088593e-07,
+      "loss": 0.9264,
+      "step": 10958
+    },
+    {
+      "epoch": 0.8448196114708603,
+      "grad_norm": 3.7621021270751953,
+      "learning_rate": 6.183944284616794e-07,
+      "loss": 0.8383,
+      "step": 10959
+    },
+    {
+      "epoch": 0.8448967005858773,
+      "grad_norm": 3.256624460220337,
+      "learning_rate": 6.177931537849141e-07,
+      "loss": 0.7408,
+      "step": 10960
+    },
+    {
+      "epoch": 0.8449737897008942,
+      "grad_norm": 3.6678712368011475,
+      "learning_rate": 6.17192152316049e-07,
+      "loss": 0.8796,
+      "step": 10961
+    },
+    {
+      "epoch": 0.8450508788159112,
+      "grad_norm": 3.6658334732055664,
+      "learning_rate": 6.165914240925547e-07,
+      "loss": 0.9107,
+      "step": 10962
+    },
+    {
+      "epoch": 0.8451279679309281,
+      "grad_norm": 3.5411484241485596,
+      "learning_rate": 6.15990969151884e-07,
+      "loss": 0.8073,
+      "step": 10963
+    },
+    {
+      "epoch": 0.8452050570459451,
+      "grad_norm": 3.56990385055542,
+      "learning_rate": 6.153907875314697e-07,
+      "loss": 0.8502,
+      "step": 10964
+    },
+    {
+      "epoch": 0.8452821461609621,
+      "grad_norm": 4.044010639190674,
+      "learning_rate": 6.147908792687307e-07,
+      "loss": 0.9423,
+      "step": 10965
+    },
+    {
+      "epoch": 0.845359235275979,
+      "grad_norm": 3.776015043258667,
+      "learning_rate": 6.141912444010694e-07,
+      "loss": 0.9912,
+      "step": 10966
+    },
+    {
+      "epoch": 0.845436324390996,
+      "grad_norm": 3.4817419052124023,
+      "learning_rate": 6.135918829658694e-07,
+      "loss": 0.8841,
+      "step": 10967
+    },
+    {
+      "epoch": 0.845513413506013,
+      "grad_norm": 3.5754287242889404,
+      "learning_rate": 6.129927950004988e-07,
+      "loss": 0.8635,
+      "step": 10968
+    },
+    {
+      "epoch": 0.8455905026210299,
+      "grad_norm": 3.851569414138794,
+      "learning_rate": 6.123939805423051e-07,
+      "loss": 1.0293,
+      "step": 10969
+    },
+    {
+      "epoch": 0.8456675917360469,
+      "grad_norm": 3.6617162227630615,
+      "learning_rate": 6.117954396286236e-07,
+      "loss": 0.864,
+      "step": 10970
+    },
+    {
+      "epoch": 0.8457446808510638,
+      "grad_norm": 3.597431182861328,
+      "learning_rate": 6.111971722967686e-07,
+      "loss": 0.9426,
+      "step": 10971
+    },
+    {
+      "epoch": 0.8458217699660808,
+      "grad_norm": 3.592806816101074,
+      "learning_rate": 6.105991785840398e-07,
+      "loss": 0.9014,
+      "step": 10972
+    },
+    {
+      "epoch": 0.8458988590810977,
+      "grad_norm": 3.4167540073394775,
+      "learning_rate": 6.100014585277187e-07,
+      "loss": 0.8378,
+      "step": 10973
+    },
+    {
+      "epoch": 0.8459759481961147,
+      "grad_norm": 4.124277114868164,
+      "learning_rate": 6.094040121650719e-07,
+      "loss": 0.9724,
+      "step": 10974
+    },
+    {
+      "epoch": 0.8460530373111317,
+      "grad_norm": 4.155192852020264,
+      "learning_rate": 6.088068395333441e-07,
+      "loss": 0.9636,
+      "step": 10975
+    },
+    {
+      "epoch": 0.8461301264261486,
+      "grad_norm": 3.687875270843506,
+      "learning_rate": 6.082099406697673e-07,
+      "loss": 0.935,
+      "step": 10976
+    },
+    {
+      "epoch": 0.8462072155411656,
+      "grad_norm": 3.8519952297210693,
+      "learning_rate": 6.076133156115549e-07,
+      "loss": 0.8827,
+      "step": 10977
+    },
+    {
+      "epoch": 0.8462843046561825,
+      "grad_norm": 3.6095309257507324,
+      "learning_rate": 6.070169643959034e-07,
+      "loss": 0.9423,
+      "step": 10978
+    },
+    {
+      "epoch": 0.8463613937711995,
+      "grad_norm": 3.734135866165161,
+      "learning_rate": 6.064208870599935e-07,
+      "loss": 0.9421,
+      "step": 10979
+    },
+    {
+      "epoch": 0.8464384828862165,
+      "grad_norm": 3.8327412605285645,
+      "learning_rate": 6.058250836409856e-07,
+      "loss": 0.9157,
+      "step": 10980
+    },
+    {
+      "epoch": 0.8465155720012334,
+      "grad_norm": 4.3431782722473145,
+      "learning_rate": 6.052295541760256e-07,
+      "loss": 1.0126,
+      "step": 10981
+    },
+    {
+      "epoch": 0.8465926611162504,
+      "grad_norm": 3.996701717376709,
+      "learning_rate": 6.046342987022419e-07,
+      "loss": 0.9892,
+      "step": 10982
+    },
+    {
+      "epoch": 0.8466697502312673,
+      "grad_norm": 4.109698295593262,
+      "learning_rate": 6.04039317256745e-07,
+      "loss": 0.9538,
+      "step": 10983
+    },
+    {
+      "epoch": 0.8467468393462843,
+      "grad_norm": 3.8181276321411133,
+      "learning_rate": 6.0344460987663e-07,
+      "loss": 0.8675,
+      "step": 10984
+    },
+    {
+      "epoch": 0.8468239284613013,
+      "grad_norm": 3.624297618865967,
+      "learning_rate": 6.028501765989736e-07,
+      "loss": 0.9302,
+      "step": 10985
+    },
+    {
+      "epoch": 0.8469010175763182,
+      "grad_norm": 3.416637659072876,
+      "learning_rate": 6.02256017460835e-07,
+      "loss": 0.8451,
+      "step": 10986
+    },
+    {
+      "epoch": 0.8469781066913352,
+      "grad_norm": 3.6329896450042725,
+      "learning_rate": 6.016621324992566e-07,
+      "loss": 0.9613,
+      "step": 10987
+    },
+    {
+      "epoch": 0.8470551958063521,
+      "grad_norm": 4.0961408615112305,
+      "learning_rate": 6.010685217512647e-07,
+      "loss": 0.9912,
+      "step": 10988
+    },
+    {
+      "epoch": 0.847132284921369,
+      "grad_norm": 3.716765880584717,
+      "learning_rate": 6.004751852538682e-07,
+      "loss": 0.9723,
+      "step": 10989
+    },
+    {
+      "epoch": 0.8472093740363861,
+      "grad_norm": 3.594797372817993,
+      "learning_rate": 5.998821230440588e-07,
+      "loss": 0.8827,
+      "step": 10990
+    },
+    {
+      "epoch": 0.847286463151403,
+      "grad_norm": 4.180978298187256,
+      "learning_rate": 5.992893351588097e-07,
+      "loss": 1.004,
+      "step": 10991
+    },
+    {
+      "epoch": 0.84736355226642,
+      "grad_norm": 3.7273788452148438,
+      "learning_rate": 5.986968216350786e-07,
+      "loss": 0.859,
+      "step": 10992
+    },
+    {
+      "epoch": 0.8474406413814369,
+      "grad_norm": 3.659938097000122,
+      "learning_rate": 5.981045825098053e-07,
+      "loss": 0.9083,
+      "step": 10993
+    },
+    {
+      "epoch": 0.8475177304964538,
+      "grad_norm": 4.153972625732422,
+      "learning_rate": 5.975126178199136e-07,
+      "loss": 0.8852,
+      "step": 10994
+    },
+    {
+      "epoch": 0.8475948196114709,
+      "grad_norm": 3.5449161529541016,
+      "learning_rate": 5.969209276023091e-07,
+      "loss": 0.7538,
+      "step": 10995
+    },
+    {
+      "epoch": 0.8476719087264878,
+      "grad_norm": 3.8162269592285156,
+      "learning_rate": 5.963295118938816e-07,
+      "loss": 0.8447,
+      "step": 10996
+    },
+    {
+      "epoch": 0.8477489978415048,
+      "grad_norm": 3.6209535598754883,
+      "learning_rate": 5.957383707315006e-07,
+      "loss": 0.9266,
+      "step": 10997
+    },
+    {
+      "epoch": 0.8478260869565217,
+      "grad_norm": 3.3889384269714355,
+      "learning_rate": 5.951475041520222e-07,
+      "loss": 0.7844,
+      "step": 10998
+    },
+    {
+      "epoch": 0.8479031760715386,
+      "grad_norm": 3.7911598682403564,
+      "learning_rate": 5.945569121922834e-07,
+      "loss": 0.876,
+      "step": 10999
+    },
+    {
+      "epoch": 0.8479802651865557,
+      "grad_norm": 3.6760973930358887,
+      "learning_rate": 5.939665948891049e-07,
+      "loss": 0.9148,
+      "step": 11000
+    },
+    {
+      "epoch": 0.8480573543015726,
+      "grad_norm": 3.681337833404541,
+      "learning_rate": 5.9337655227929e-07,
+      "loss": 0.9059,
+      "step": 11001
+    },
+    {
+      "epoch": 0.8481344434165896,
+      "grad_norm": 3.6630451679229736,
+      "learning_rate": 5.927867843996243e-07,
+      "loss": 0.8611,
+      "step": 11002
+    },
+    {
+      "epoch": 0.8482115325316065,
+      "grad_norm": 3.8612163066864014,
+      "learning_rate": 5.921972912868768e-07,
+      "loss": 0.9276,
+      "step": 11003
+    },
+    {
+      "epoch": 0.8482886216466234,
+      "grad_norm": 3.941199541091919,
+      "learning_rate": 5.916080729778e-07,
+      "loss": 0.9488,
+      "step": 11004
+    },
+    {
+      "epoch": 0.8483657107616405,
+      "grad_norm": 3.7736704349517822,
+      "learning_rate": 5.910191295091289e-07,
+      "loss": 0.8789,
+      "step": 11005
+    },
+    {
+      "epoch": 0.8484427998766574,
+      "grad_norm": 3.812868595123291,
+      "learning_rate": 5.904304609175798e-07,
+      "loss": 0.9216,
+      "step": 11006
+    },
+    {
+      "epoch": 0.8485198889916744,
+      "grad_norm": 3.7537074089050293,
+      "learning_rate": 5.89842067239853e-07,
+      "loss": 0.8657,
+      "step": 11007
+    },
+    {
+      "epoch": 0.8485969781066913,
+      "grad_norm": 3.538891553878784,
+      "learning_rate": 5.892539485126331e-07,
+      "loss": 0.9183,
+      "step": 11008
+    },
+    {
+      "epoch": 0.8486740672217082,
+      "grad_norm": 3.7746341228485107,
+      "learning_rate": 5.886661047725856e-07,
+      "loss": 0.9663,
+      "step": 11009
+    },
+    {
+      "epoch": 0.8487511563367253,
+      "grad_norm": 4.052391529083252,
+      "learning_rate": 5.880785360563596e-07,
+      "loss": 0.9895,
+      "step": 11010
+    },
+    {
+      "epoch": 0.8488282454517422,
+      "grad_norm": 3.68890643119812,
+      "learning_rate": 5.87491242400588e-07,
+      "loss": 0.836,
+      "step": 11011
+    },
+    {
+      "epoch": 0.8489053345667592,
+      "grad_norm": 3.5953805446624756,
+      "learning_rate": 5.869042238418832e-07,
+      "loss": 0.9303,
+      "step": 11012
+    },
+    {
+      "epoch": 0.8489824236817761,
+      "grad_norm": 3.813616991043091,
+      "learning_rate": 5.863174804168442e-07,
+      "loss": 0.9442,
+      "step": 11013
+    },
+    {
+      "epoch": 0.8490595127967931,
+      "grad_norm": 3.9077954292297363,
+      "learning_rate": 5.857310121620513e-07,
+      "loss": 0.9025,
+      "step": 11014
+    },
+    {
+      "epoch": 0.8491366019118101,
+      "grad_norm": 4.491962909698486,
+      "learning_rate": 5.851448191140674e-07,
+      "loss": 0.9682,
+      "step": 11015
+    },
+    {
+      "epoch": 0.849213691026827,
+      "grad_norm": 3.854057550430298,
+      "learning_rate": 5.845589013094405e-07,
+      "loss": 0.9459,
+      "step": 11016
+    },
+    {
+      "epoch": 0.849290780141844,
+      "grad_norm": 3.7456696033477783,
+      "learning_rate": 5.839732587846963e-07,
+      "loss": 0.9548,
+      "step": 11017
+    },
+    {
+      "epoch": 0.8493678692568609,
+      "grad_norm": 3.6205060482025146,
+      "learning_rate": 5.833878915763485e-07,
+      "loss": 0.9676,
+      "step": 11018
+    },
+    {
+      "epoch": 0.849444958371878,
+      "grad_norm": 3.759831428527832,
+      "learning_rate": 5.828027997208918e-07,
+      "loss": 0.8192,
+      "step": 11019
+    },
+    {
+      "epoch": 0.8495220474868949,
+      "grad_norm": 3.9140279293060303,
+      "learning_rate": 5.822179832548025e-07,
+      "loss": 0.9411,
+      "step": 11020
+    },
+    {
+      "epoch": 0.8495991366019118,
+      "grad_norm": 3.6115550994873047,
+      "learning_rate": 5.81633442214542e-07,
+      "loss": 0.8644,
+      "step": 11021
+    },
+    {
+      "epoch": 0.8496762257169288,
+      "grad_norm": 3.4248645305633545,
+      "learning_rate": 5.810491766365545e-07,
+      "loss": 0.897,
+      "step": 11022
+    },
+    {
+      "epoch": 0.8497533148319457,
+      "grad_norm": 4.218714237213135,
+      "learning_rate": 5.80465186557263e-07,
+      "loss": 0.9441,
+      "step": 11023
+    },
+    {
+      "epoch": 0.8498304039469627,
+      "grad_norm": 3.3824057579040527,
+      "learning_rate": 5.798814720130779e-07,
+      "loss": 0.8109,
+      "step": 11024
+    },
+    {
+      "epoch": 0.8499074930619797,
+      "grad_norm": 3.555572032928467,
+      "learning_rate": 5.792980330403908e-07,
+      "loss": 0.9638,
+      "step": 11025
+    },
+    {
+      "epoch": 0.8499845821769966,
+      "grad_norm": 3.7353179454803467,
+      "learning_rate": 5.787148696755757e-07,
+      "loss": 0.997,
+      "step": 11026
+    },
+    {
+      "epoch": 0.8500616712920136,
+      "grad_norm": 3.5979185104370117,
+      "learning_rate": 5.781319819549913e-07,
+      "loss": 0.8548,
+      "step": 11027
+    },
+    {
+      "epoch": 0.8501387604070305,
+      "grad_norm": 3.9905002117156982,
+      "learning_rate": 5.775493699149754e-07,
+      "loss": 0.9858,
+      "step": 11028
+    },
+    {
+      "epoch": 0.8502158495220475,
+      "grad_norm": 3.6577179431915283,
+      "learning_rate": 5.769670335918515e-07,
+      "loss": 0.9216,
+      "step": 11029
+    },
+    {
+      "epoch": 0.8502929386370645,
+      "grad_norm": 4.757742404937744,
+      "learning_rate": 5.763849730219257e-07,
+      "loss": 1.0233,
+      "step": 11030
+    },
+    {
+      "epoch": 0.8503700277520814,
+      "grad_norm": 3.6999893188476562,
+      "learning_rate": 5.758031882414861e-07,
+      "loss": 0.9356,
+      "step": 11031
+    },
+    {
+      "epoch": 0.8504471168670984,
+      "grad_norm": 4.301717281341553,
+      "learning_rate": 5.752216792868048e-07,
+      "loss": 1.0229,
+      "step": 11032
+    },
+    {
+      "epoch": 0.8505242059821153,
+      "grad_norm": 3.8095266819000244,
+      "learning_rate": 5.746404461941358e-07,
+      "loss": 0.965,
+      "step": 11033
+    },
+    {
+      "epoch": 0.8506012950971323,
+      "grad_norm": 3.8716020584106445,
+      "learning_rate": 5.740594889997147e-07,
+      "loss": 0.8976,
+      "step": 11034
+    },
+    {
+      "epoch": 0.8506783842121493,
+      "grad_norm": 3.362471103668213,
+      "learning_rate": 5.734788077397618e-07,
+      "loss": 0.8177,
+      "step": 11035
+    },
+    {
+      "epoch": 0.8507554733271662,
+      "grad_norm": 3.7064013481140137,
+      "learning_rate": 5.728984024504796e-07,
+      "loss": 0.9328,
+      "step": 11036
+    },
+    {
+      "epoch": 0.8508325624421832,
+      "grad_norm": 3.902492046356201,
+      "learning_rate": 5.723182731680538e-07,
+      "loss": 1.0963,
+      "step": 11037
+    },
+    {
+      "epoch": 0.8509096515572001,
+      "grad_norm": 3.7971115112304688,
+      "learning_rate": 5.717384199286529e-07,
+      "loss": 0.9039,
+      "step": 11038
+    },
+    {
+      "epoch": 0.8509867406722171,
+      "grad_norm": 3.7627763748168945,
+      "learning_rate": 5.711588427684262e-07,
+      "loss": 0.9493,
+      "step": 11039
+    },
+    {
+      "epoch": 0.851063829787234,
+      "grad_norm": 3.705435037612915,
+      "learning_rate": 5.705795417235077e-07,
+      "loss": 0.9674,
+      "step": 11040
+    },
+    {
+      "epoch": 0.851140918902251,
+      "grad_norm": 4.405561447143555,
+      "learning_rate": 5.700005168300144e-07,
+      "loss": 0.9634,
+      "step": 11041
+    },
+    {
+      "epoch": 0.851218008017268,
+      "grad_norm": 3.6750857830047607,
+      "learning_rate": 5.694217681240455e-07,
+      "loss": 0.9182,
+      "step": 11042
+    },
+    {
+      "epoch": 0.8512950971322849,
+      "grad_norm": 3.666693925857544,
+      "learning_rate": 5.688432956416823e-07,
+      "loss": 0.8737,
+      "step": 11043
+    },
+    {
+      "epoch": 0.8513721862473019,
+      "grad_norm": 3.796339988708496,
+      "learning_rate": 5.682650994189915e-07,
+      "loss": 0.9079,
+      "step": 11044
+    },
+    {
+      "epoch": 0.8514492753623188,
+      "grad_norm": 3.8772926330566406,
+      "learning_rate": 5.676871794920169e-07,
+      "loss": 0.8889,
+      "step": 11045
+    },
+    {
+      "epoch": 0.8515263644773358,
+      "grad_norm": 3.8630027770996094,
+      "learning_rate": 5.671095358967926e-07,
+      "loss": 0.9146,
+      "step": 11046
+    },
+    {
+      "epoch": 0.8516034535923528,
+      "grad_norm": 3.968269109725952,
+      "learning_rate": 5.665321686693298e-07,
+      "loss": 0.9438,
+      "step": 11047
+    },
+    {
+      "epoch": 0.8516805427073697,
+      "grad_norm": 3.7893552780151367,
+      "learning_rate": 5.659550778456258e-07,
+      "loss": 0.9303,
+      "step": 11048
+    },
+    {
+      "epoch": 0.8517576318223867,
+      "grad_norm": 3.5288009643554688,
+      "learning_rate": 5.653782634616573e-07,
+      "loss": 0.8899,
+      "step": 11049
+    },
+    {
+      "epoch": 0.8518347209374036,
+      "grad_norm": 4.320144176483154,
+      "learning_rate": 5.648017255533866e-07,
+      "loss": 0.9022,
+      "step": 11050
+    },
+    {
+      "epoch": 0.8519118100524206,
+      "grad_norm": 3.9753165245056152,
+      "learning_rate": 5.642254641567579e-07,
+      "loss": 0.8956,
+      "step": 11051
+    },
+    {
+      "epoch": 0.8519888991674376,
+      "grad_norm": 3.734354257583618,
+      "learning_rate": 5.636494793076974e-07,
+      "loss": 0.9239,
+      "step": 11052
+    },
+    {
+      "epoch": 0.8520659882824545,
+      "grad_norm": 3.6947646141052246,
+      "learning_rate": 5.630737710421174e-07,
+      "loss": 0.9127,
+      "step": 11053
+    },
+    {
+      "epoch": 0.8521430773974715,
+      "grad_norm": 3.678510904312134,
+      "learning_rate": 5.624983393959066e-07,
+      "loss": 0.8578,
+      "step": 11054
+    },
+    {
+      "epoch": 0.8522201665124884,
+      "grad_norm": 3.4292240142822266,
+      "learning_rate": 5.619231844049422e-07,
+      "loss": 0.9145,
+      "step": 11055
+    },
+    {
+      "epoch": 0.8522972556275054,
+      "grad_norm": 4.060187816619873,
+      "learning_rate": 5.613483061050818e-07,
+      "loss": 0.9417,
+      "step": 11056
+    },
+    {
+      "epoch": 0.8523743447425224,
+      "grad_norm": 3.9965662956237793,
+      "learning_rate": 5.607737045321666e-07,
+      "loss": 0.7447,
+      "step": 11057
+    },
+    {
+      "epoch": 0.8524514338575393,
+      "grad_norm": 3.6497106552124023,
+      "learning_rate": 5.60199379722019e-07,
+      "loss": 0.8786,
+      "step": 11058
+    },
+    {
+      "epoch": 0.8525285229725563,
+      "grad_norm": 3.614408493041992,
+      "learning_rate": 5.596253317104472e-07,
+      "loss": 0.8301,
+      "step": 11059
+    },
+    {
+      "epoch": 0.8526056120875732,
+      "grad_norm": 4.088712692260742,
+      "learning_rate": 5.590515605332369e-07,
+      "loss": 1.0084,
+      "step": 11060
+    },
+    {
+      "epoch": 0.8526827012025902,
+      "grad_norm": 3.6503963470458984,
+      "learning_rate": 5.584780662261624e-07,
+      "loss": 0.8587,
+      "step": 11061
+    },
+    {
+      "epoch": 0.8527597903176072,
+      "grad_norm": 3.714747667312622,
+      "learning_rate": 5.579048488249766e-07,
+      "loss": 0.9238,
+      "step": 11062
+    },
+    {
+      "epoch": 0.8528368794326241,
+      "grad_norm": 4.048036098480225,
+      "learning_rate": 5.573319083654177e-07,
+      "loss": 0.953,
+      "step": 11063
+    },
+    {
+      "epoch": 0.8529139685476411,
+      "grad_norm": 3.8897385597229004,
+      "learning_rate": 5.56759244883206e-07,
+      "loss": 0.8999,
+      "step": 11064
+    },
+    {
+      "epoch": 0.852991057662658,
+      "grad_norm": 4.280935287475586,
+      "learning_rate": 5.561868584140423e-07,
+      "loss": 0.8316,
+      "step": 11065
+    },
+    {
+      "epoch": 0.853068146777675,
+      "grad_norm": 3.4700443744659424,
+      "learning_rate": 5.556147489936131e-07,
+      "loss": 0.8984,
+      "step": 11066
+    },
+    {
+      "epoch": 0.853145235892692,
+      "grad_norm": 3.7145278453826904,
+      "learning_rate": 5.55042916657586e-07,
+      "loss": 0.911,
+      "step": 11067
+    },
+    {
+      "epoch": 0.8532223250077089,
+      "grad_norm": 3.821202278137207,
+      "learning_rate": 5.54471361441612e-07,
+      "loss": 1.0008,
+      "step": 11068
+    },
+    {
+      "epoch": 0.8532994141227259,
+      "grad_norm": 3.5977516174316406,
+      "learning_rate": 5.539000833813246e-07,
+      "loss": 0.9296,
+      "step": 11069
+    },
+    {
+      "epoch": 0.8533765032377428,
+      "grad_norm": 3.5324435234069824,
+      "learning_rate": 5.533290825123411e-07,
+      "loss": 0.7846,
+      "step": 11070
+    },
+    {
+      "epoch": 0.8534535923527597,
+      "grad_norm": 3.96653151512146,
+      "learning_rate": 5.527583588702584e-07,
+      "loss": 0.9285,
+      "step": 11071
+    },
+    {
+      "epoch": 0.8535306814677768,
+      "grad_norm": 4.007519721984863,
+      "learning_rate": 5.521879124906587e-07,
+      "loss": 0.961,
+      "step": 11072
+    },
+    {
+      "epoch": 0.8536077705827937,
+      "grad_norm": 3.8947818279266357,
+      "learning_rate": 5.516177434091074e-07,
+      "loss": 0.8486,
+      "step": 11073
+    },
+    {
+      "epoch": 0.8536848596978107,
+      "grad_norm": 3.4934535026550293,
+      "learning_rate": 5.510478516611512e-07,
+      "loss": 0.8873,
+      "step": 11074
+    },
+    {
+      "epoch": 0.8537619488128276,
+      "grad_norm": 3.4623193740844727,
+      "learning_rate": 5.504782372823203e-07,
+      "loss": 0.9679,
+      "step": 11075
+    },
+    {
+      "epoch": 0.8538390379278445,
+      "grad_norm": 3.704373836517334,
+      "learning_rate": 5.499089003081259e-07,
+      "loss": 0.8837,
+      "step": 11076
+    },
+    {
+      "epoch": 0.8539161270428616,
+      "grad_norm": 3.9960663318634033,
+      "learning_rate": 5.493398407740641e-07,
+      "loss": 0.9834,
+      "step": 11077
+    },
+    {
+      "epoch": 0.8539932161578785,
+      "grad_norm": 4.131714820861816,
+      "learning_rate": 5.48771058715612e-07,
+      "loss": 1.0009,
+      "step": 11078
+    },
+    {
+      "epoch": 0.8540703052728955,
+      "grad_norm": 3.8496928215026855,
+      "learning_rate": 5.482025541682312e-07,
+      "loss": 0.902,
+      "step": 11079
+    },
+    {
+      "epoch": 0.8541473943879124,
+      "grad_norm": 3.6283254623413086,
+      "learning_rate": 5.476343271673651e-07,
+      "loss": 0.9259,
+      "step": 11080
+    },
+    {
+      "epoch": 0.8542244835029293,
+      "grad_norm": 3.9260733127593994,
+      "learning_rate": 5.470663777484402e-07,
+      "loss": 0.9205,
+      "step": 11081
+    },
+    {
+      "epoch": 0.8543015726179464,
+      "grad_norm": 3.7307844161987305,
+      "learning_rate": 5.464987059468629e-07,
+      "loss": 0.9092,
+      "step": 11082
+    },
+    {
+      "epoch": 0.8543786617329633,
+      "grad_norm": 3.9613077640533447,
+      "learning_rate": 5.459313117980264e-07,
+      "loss": 0.9234,
+      "step": 11083
+    },
+    {
+      "epoch": 0.8544557508479803,
+      "grad_norm": 3.670541524887085,
+      "learning_rate": 5.45364195337304e-07,
+      "loss": 0.9093,
+      "step": 11084
+    },
+    {
+      "epoch": 0.8545328399629972,
+      "grad_norm": 3.992414951324463,
+      "learning_rate": 5.447973566000531e-07,
+      "loss": 0.9303,
+      "step": 11085
+    },
+    {
+      "epoch": 0.8546099290780141,
+      "grad_norm": 3.7107622623443604,
+      "learning_rate": 5.442307956216142e-07,
+      "loss": 0.8985,
+      "step": 11086
+    },
+    {
+      "epoch": 0.8546870181930312,
+      "grad_norm": 4.1595354080200195,
+      "learning_rate": 5.436645124373069e-07,
+      "loss": 0.906,
+      "step": 11087
+    },
+    {
+      "epoch": 0.8547641073080481,
+      "grad_norm": 3.343167543411255,
+      "learning_rate": 5.430985070824374e-07,
+      "loss": 0.7814,
+      "step": 11088
+    },
+    {
+      "epoch": 0.8548411964230651,
+      "grad_norm": 3.637519598007202,
+      "learning_rate": 5.42532779592293e-07,
+      "loss": 0.8609,
+      "step": 11089
+    },
+    {
+      "epoch": 0.854918285538082,
+      "grad_norm": 3.568153142929077,
+      "learning_rate": 5.419673300021427e-07,
+      "loss": 0.9346,
+      "step": 11090
+    },
+    {
+      "epoch": 0.8549953746530989,
+      "grad_norm": 3.780703067779541,
+      "learning_rate": 5.414021583472429e-07,
+      "loss": 0.9486,
+      "step": 11091
+    },
+    {
+      "epoch": 0.855072463768116,
+      "grad_norm": 3.555079460144043,
+      "learning_rate": 5.408372646628257e-07,
+      "loss": 0.8887,
+      "step": 11092
+    },
+    {
+      "epoch": 0.8551495528831329,
+      "grad_norm": 3.8127081394195557,
+      "learning_rate": 5.40272648984111e-07,
+      "loss": 0.8914,
+      "step": 11093
+    },
+    {
+      "epoch": 0.8552266419981499,
+      "grad_norm": 3.8799734115600586,
+      "learning_rate": 5.397083113462986e-07,
+      "loss": 0.8731,
+      "step": 11094
+    },
+    {
+      "epoch": 0.8553037311131668,
+      "grad_norm": 3.8874175548553467,
+      "learning_rate": 5.391442517845724e-07,
+      "loss": 0.8681,
+      "step": 11095
+    },
+    {
+      "epoch": 0.8553808202281837,
+      "grad_norm": 3.862079381942749,
+      "learning_rate": 5.385804703341002e-07,
+      "loss": 0.895,
+      "step": 11096
+    },
+    {
+      "epoch": 0.8554579093432008,
+      "grad_norm": 3.89817476272583,
+      "learning_rate": 5.380169670300284e-07,
+      "loss": 0.8392,
+      "step": 11097
+    },
+    {
+      "epoch": 0.8555349984582177,
+      "grad_norm": 3.441835403442383,
+      "learning_rate": 5.374537419074893e-07,
+      "loss": 0.8703,
+      "step": 11098
+    },
+    {
+      "epoch": 0.8556120875732347,
+      "grad_norm": 3.845632314682007,
+      "learning_rate": 5.368907950015972e-07,
+      "loss": 0.8489,
+      "step": 11099
+    },
+    {
+      "epoch": 0.8556891766882516,
+      "grad_norm": 3.8767735958099365,
+      "learning_rate": 5.36328126347449e-07,
+      "loss": 0.8637,
+      "step": 11100
+    },
+    {
+      "epoch": 0.8557662658032685,
+      "grad_norm": 3.9100561141967773,
+      "learning_rate": 5.357657359801249e-07,
+      "loss": 0.951,
+      "step": 11101
+    },
+    {
+      "epoch": 0.8558433549182856,
+      "grad_norm": 3.784698486328125,
+      "learning_rate": 5.352036239346858e-07,
+      "loss": 0.9673,
+      "step": 11102
+    },
+    {
+      "epoch": 0.8559204440333025,
+      "grad_norm": 3.8566789627075195,
+      "learning_rate": 5.34641790246177e-07,
+      "loss": 0.8962,
+      "step": 11103
+    },
+    {
+      "epoch": 0.8559975331483195,
+      "grad_norm": 3.3040380477905273,
+      "learning_rate": 5.340802349496254e-07,
+      "loss": 0.8545,
+      "step": 11104
+    },
+    {
+      "epoch": 0.8560746222633364,
+      "grad_norm": 3.585635185241699,
+      "learning_rate": 5.335189580800421e-07,
+      "loss": 0.8885,
+      "step": 11105
+    },
+    {
+      "epoch": 0.8561517113783533,
+      "grad_norm": 3.8033549785614014,
+      "learning_rate": 5.329579596724188e-07,
+      "loss": 0.8385,
+      "step": 11106
+    },
+    {
+      "epoch": 0.8562288004933704,
+      "grad_norm": 3.958906888961792,
+      "learning_rate": 5.323972397617327e-07,
+      "loss": 0.9254,
+      "step": 11107
+    },
+    {
+      "epoch": 0.8563058896083873,
+      "grad_norm": 3.711397647857666,
+      "learning_rate": 5.318367983829393e-07,
+      "loss": 0.9813,
+      "step": 11108
+    },
+    {
+      "epoch": 0.8563829787234043,
+      "grad_norm": 3.6522676944732666,
+      "learning_rate": 5.312766355709803e-07,
+      "loss": 0.8936,
+      "step": 11109
+    },
+    {
+      "epoch": 0.8564600678384212,
+      "grad_norm": 4.029238700866699,
+      "learning_rate": 5.307167513607786e-07,
+      "loss": 1.009,
+      "step": 11110
+    },
+    {
+      "epoch": 0.8565371569534381,
+      "grad_norm": 3.7067220211029053,
+      "learning_rate": 5.301571457872407e-07,
+      "loss": 0.9072,
+      "step": 11111
+    },
+    {
+      "epoch": 0.8566142460684552,
+      "grad_norm": 4.054270267486572,
+      "learning_rate": 5.295978188852557e-07,
+      "loss": 1.0588,
+      "step": 11112
+    },
+    {
+      "epoch": 0.8566913351834721,
+      "grad_norm": 3.697340250015259,
+      "learning_rate": 5.290387706896933e-07,
+      "loss": 0.9871,
+      "step": 11113
+    },
+    {
+      "epoch": 0.8567684242984891,
+      "grad_norm": 4.2902092933654785,
+      "learning_rate": 5.284800012354075e-07,
+      "loss": 0.8552,
+      "step": 11114
+    },
+    {
+      "epoch": 0.856845513413506,
+      "grad_norm": 3.741992950439453,
+      "learning_rate": 5.279215105572355e-07,
+      "loss": 0.8789,
+      "step": 11115
+    },
+    {
+      "epoch": 0.8569226025285229,
+      "grad_norm": 3.8144991397857666,
+      "learning_rate": 5.273632986899951e-07,
+      "loss": 0.8931,
+      "step": 11116
+    },
+    {
+      "epoch": 0.85699969164354,
+      "grad_norm": 3.5984315872192383,
+      "learning_rate": 5.268053656684891e-07,
+      "loss": 0.9031,
+      "step": 11117
+    },
+    {
+      "epoch": 0.8570767807585569,
+      "grad_norm": 4.40741491317749,
+      "learning_rate": 5.262477115275022e-07,
+      "loss": 0.9021,
+      "step": 11118
+    },
+    {
+      "epoch": 0.8571538698735739,
+      "grad_norm": 3.9442696571350098,
+      "learning_rate": 5.256903363017995e-07,
+      "loss": 1.0166,
+      "step": 11119
+    },
+    {
+      "epoch": 0.8572309589885908,
+      "grad_norm": 3.6909403800964355,
+      "learning_rate": 5.251332400261311e-07,
+      "loss": 0.8354,
+      "step": 11120
+    },
+    {
+      "epoch": 0.8573080481036077,
+      "grad_norm": 3.769570827484131,
+      "learning_rate": 5.24576422735229e-07,
+      "loss": 0.9404,
+      "step": 11121
+    },
+    {
+      "epoch": 0.8573851372186247,
+      "grad_norm": 3.7346904277801514,
+      "learning_rate": 5.240198844638084e-07,
+      "loss": 0.9651,
+      "step": 11122
+    },
+    {
+      "epoch": 0.8574622263336417,
+      "grad_norm": 3.3625786304473877,
+      "learning_rate": 5.234636252465675e-07,
+      "loss": 0.9208,
+      "step": 11123
+    },
+    {
+      "epoch": 0.8575393154486587,
+      "grad_norm": 3.8647756576538086,
+      "learning_rate": 5.229076451181836e-07,
+      "loss": 0.9616,
+      "step": 11124
+    },
+    {
+      "epoch": 0.8576164045636756,
+      "grad_norm": 3.784053325653076,
+      "learning_rate": 5.223519441133206e-07,
+      "loss": 0.9683,
+      "step": 11125
+    },
+    {
+      "epoch": 0.8576934936786925,
+      "grad_norm": 3.712662696838379,
+      "learning_rate": 5.217965222666239e-07,
+      "loss": 0.8376,
+      "step": 11126
+    },
+    {
+      "epoch": 0.8577705827937095,
+      "grad_norm": 4.219189167022705,
+      "learning_rate": 5.212413796127208e-07,
+      "loss": 1.0494,
+      "step": 11127
+    },
+    {
+      "epoch": 0.8578476719087265,
+      "grad_norm": 3.7683613300323486,
+      "learning_rate": 5.206865161862212e-07,
+      "loss": 0.9379,
+      "step": 11128
+    },
+    {
+      "epoch": 0.8579247610237435,
+      "grad_norm": 3.439199447631836,
+      "learning_rate": 5.201319320217196e-07,
+      "loss": 0.8908,
+      "step": 11129
+    },
+    {
+      "epoch": 0.8580018501387604,
+      "grad_norm": 3.727071762084961,
+      "learning_rate": 5.195776271537894e-07,
+      "loss": 0.8725,
+      "step": 11130
+    },
+    {
+      "epoch": 0.8580789392537773,
+      "grad_norm": 3.9292681217193604,
+      "learning_rate": 5.190236016169892e-07,
+      "loss": 0.8883,
+      "step": 11131
+    },
+    {
+      "epoch": 0.8581560283687943,
+      "grad_norm": 4.457842826843262,
+      "learning_rate": 5.184698554458595e-07,
+      "loss": 0.8266,
+      "step": 11132
+    },
+    {
+      "epoch": 0.8582331174838113,
+      "grad_norm": 4.050070762634277,
+      "learning_rate": 5.179163886749244e-07,
+      "loss": 1.0174,
+      "step": 11133
+    },
+    {
+      "epoch": 0.8583102065988283,
+      "grad_norm": 3.961498498916626,
+      "learning_rate": 5.173632013386892e-07,
+      "loss": 0.9666,
+      "step": 11134
+    },
+    {
+      "epoch": 0.8583872957138452,
+      "grad_norm": 3.9840786457061768,
+      "learning_rate": 5.168102934716419e-07,
+      "loss": 0.857,
+      "step": 11135
+    },
+    {
+      "epoch": 0.8584643848288621,
+      "grad_norm": 3.528966188430786,
+      "learning_rate": 5.162576651082541e-07,
+      "loss": 0.9403,
+      "step": 11136
+    },
+    {
+      "epoch": 0.8585414739438791,
+      "grad_norm": 3.7930431365966797,
+      "learning_rate": 5.157053162829789e-07,
+      "loss": 0.919,
+      "step": 11137
+    },
+    {
+      "epoch": 0.8586185630588961,
+      "grad_norm": 3.9015920162200928,
+      "learning_rate": 5.151532470302523e-07,
+      "loss": 0.9518,
+      "step": 11138
+    },
+    {
+      "epoch": 0.8586956521739131,
+      "grad_norm": 3.5874063968658447,
+      "learning_rate": 5.146014573844943e-07,
+      "loss": 0.8774,
+      "step": 11139
+    },
+    {
+      "epoch": 0.85877274128893,
+      "grad_norm": 3.664182662963867,
+      "learning_rate": 5.140499473801036e-07,
+      "loss": 0.9138,
+      "step": 11140
+    },
+    {
+      "epoch": 0.8588498304039469,
+      "grad_norm": 3.809180736541748,
+      "learning_rate": 5.134987170514654e-07,
+      "loss": 0.8837,
+      "step": 11141
+    },
+    {
+      "epoch": 0.8589269195189639,
+      "grad_norm": 4.012628078460693,
+      "learning_rate": 5.129477664329463e-07,
+      "loss": 0.9214,
+      "step": 11142
+    },
+    {
+      "epoch": 0.8590040086339809,
+      "grad_norm": 3.4249749183654785,
+      "learning_rate": 5.123970955588947e-07,
+      "loss": 0.6997,
+      "step": 11143
+    },
+    {
+      "epoch": 0.8590810977489979,
+      "grad_norm": 3.523764133453369,
+      "learning_rate": 5.118467044636438e-07,
+      "loss": 0.8808,
+      "step": 11144
+    },
+    {
+      "epoch": 0.8591581868640148,
+      "grad_norm": 3.6512789726257324,
+      "learning_rate": 5.112965931815045e-07,
+      "loss": 0.9734,
+      "step": 11145
+    },
+    {
+      "epoch": 0.8592352759790317,
+      "grad_norm": 3.7311131954193115,
+      "learning_rate": 5.107467617467754e-07,
+      "loss": 0.961,
+      "step": 11146
+    },
+    {
+      "epoch": 0.8593123650940487,
+      "grad_norm": 3.953731060028076,
+      "learning_rate": 5.101972101937352e-07,
+      "loss": 0.9679,
+      "step": 11147
+    },
+    {
+      "epoch": 0.8593894542090657,
+      "grad_norm": 3.7409465312957764,
+      "learning_rate": 5.096479385566455e-07,
+      "loss": 0.9236,
+      "step": 11148
+    },
+    {
+      "epoch": 0.8594665433240827,
+      "grad_norm": 3.9198498725891113,
+      "learning_rate": 5.090989468697515e-07,
+      "loss": 0.9617,
+      "step": 11149
+    },
+    {
+      "epoch": 0.8595436324390996,
+      "grad_norm": 3.991605520248413,
+      "learning_rate": 5.085502351672788e-07,
+      "loss": 0.8857,
+      "step": 11150
+    },
+    {
+      "epoch": 0.8596207215541165,
+      "grad_norm": 3.804913282394409,
+      "learning_rate": 5.080018034834367e-07,
+      "loss": 0.9456,
+      "step": 11151
+    },
+    {
+      "epoch": 0.8596978106691335,
+      "grad_norm": 3.8616576194763184,
+      "learning_rate": 5.074536518524176e-07,
+      "loss": 0.9924,
+      "step": 11152
+    },
+    {
+      "epoch": 0.8597748997841504,
+      "grad_norm": 3.636204957962036,
+      "learning_rate": 5.06905780308396e-07,
+      "loss": 1.0058,
+      "step": 11153
+    },
+    {
+      "epoch": 0.8598519888991675,
+      "grad_norm": 4.016709327697754,
+      "learning_rate": 5.063581888855285e-07,
+      "loss": 0.9465,
+      "step": 11154
+    },
+    {
+      "epoch": 0.8599290780141844,
+      "grad_norm": 4.032883644104004,
+      "learning_rate": 5.058108776179555e-07,
+      "loss": 0.9796,
+      "step": 11155
+    },
+    {
+      "epoch": 0.8600061671292013,
+      "grad_norm": 3.659719705581665,
+      "learning_rate": 5.05263846539798e-07,
+      "loss": 0.9076,
+      "step": 11156
+    },
+    {
+      "epoch": 0.8600832562442183,
+      "grad_norm": 4.1367950439453125,
+      "learning_rate": 5.0471709568516e-07,
+      "loss": 0.9214,
+      "step": 11157
+    },
+    {
+      "epoch": 0.8601603453592352,
+      "grad_norm": 4.058650493621826,
+      "learning_rate": 5.041706250881301e-07,
+      "loss": 1.0593,
+      "step": 11158
+    },
+    {
+      "epoch": 0.8602374344742523,
+      "grad_norm": 3.848299980163574,
+      "learning_rate": 5.036244347827768e-07,
+      "loss": 0.8446,
+      "step": 11159
+    },
+    {
+      "epoch": 0.8603145235892692,
+      "grad_norm": 3.458164930343628,
+      "learning_rate": 5.030785248031534e-07,
+      "loss": 0.877,
+      "step": 11160
+    },
+    {
+      "epoch": 0.8603916127042861,
+      "grad_norm": 3.6949539184570312,
+      "learning_rate": 5.025328951832936e-07,
+      "loss": 0.8685,
+      "step": 11161
+    },
+    {
+      "epoch": 0.8604687018193031,
+      "grad_norm": 4.433669090270996,
+      "learning_rate": 5.019875459572143e-07,
+      "loss": 0.9717,
+      "step": 11162
+    },
+    {
+      "epoch": 0.86054579093432,
+      "grad_norm": 3.662029266357422,
+      "learning_rate": 5.014424771589155e-07,
+      "loss": 0.8557,
+      "step": 11163
+    },
+    {
+      "epoch": 0.8606228800493371,
+      "grad_norm": 3.8234245777130127,
+      "learning_rate": 5.008976888223799e-07,
+      "loss": 0.9186,
+      "step": 11164
+    },
+    {
+      "epoch": 0.860699969164354,
+      "grad_norm": 3.7250325679779053,
+      "learning_rate": 5.003531809815721e-07,
+      "loss": 0.9464,
+      "step": 11165
+    },
+    {
+      "epoch": 0.8607770582793709,
+      "grad_norm": 3.8604021072387695,
+      "learning_rate": 4.998089536704399e-07,
+      "loss": 1.0107,
+      "step": 11166
+    },
+    {
+      "epoch": 0.8608541473943879,
+      "grad_norm": 3.525672674179077,
+      "learning_rate": 4.992650069229116e-07,
+      "loss": 0.9176,
+      "step": 11167
+    },
+    {
+      "epoch": 0.8609312365094048,
+      "grad_norm": 3.9230077266693115,
+      "learning_rate": 4.987213407729002e-07,
+      "loss": 0.8406,
+      "step": 11168
+    },
+    {
+      "epoch": 0.8610083256244219,
+      "grad_norm": 3.6223180294036865,
+      "learning_rate": 4.981779552543004e-07,
+      "loss": 0.8505,
+      "step": 11169
+    },
+    {
+      "epoch": 0.8610854147394388,
+      "grad_norm": 4.130772590637207,
+      "learning_rate": 4.976348504009899e-07,
+      "loss": 0.9977,
+      "step": 11170
+    },
+    {
+      "epoch": 0.8611625038544557,
+      "grad_norm": 3.5042903423309326,
+      "learning_rate": 4.970920262468282e-07,
+      "loss": 0.8921,
+      "step": 11171
+    },
+    {
+      "epoch": 0.8612395929694727,
+      "grad_norm": 4.355722427368164,
+      "learning_rate": 4.965494828256573e-07,
+      "loss": 0.8054,
+      "step": 11172
+    },
+    {
+      "epoch": 0.8613166820844896,
+      "grad_norm": 3.7091667652130127,
+      "learning_rate": 4.960072201713018e-07,
+      "loss": 0.8937,
+      "step": 11173
+    },
+    {
+      "epoch": 0.8613937711995067,
+      "grad_norm": 4.190767288208008,
+      "learning_rate": 4.954652383175696e-07,
+      "loss": 0.9878,
+      "step": 11174
+    },
+    {
+      "epoch": 0.8614708603145236,
+      "grad_norm": 3.6260433197021484,
+      "learning_rate": 4.949235372982503e-07,
+      "loss": 0.9711,
+      "step": 11175
+    },
+    {
+      "epoch": 0.8615479494295405,
+      "grad_norm": 3.458486318588257,
+      "learning_rate": 4.943821171471158e-07,
+      "loss": 0.8357,
+      "step": 11176
+    },
+    {
+      "epoch": 0.8616250385445575,
+      "grad_norm": 3.68377423286438,
+      "learning_rate": 4.93840977897922e-07,
+      "loss": 0.8426,
+      "step": 11177
+    },
+    {
+      "epoch": 0.8617021276595744,
+      "grad_norm": 3.642500877380371,
+      "learning_rate": 4.933001195844034e-07,
+      "loss": 0.8853,
+      "step": 11178
+    },
+    {
+      "epoch": 0.8617792167745915,
+      "grad_norm": 3.628203868865967,
+      "learning_rate": 4.927595422402826e-07,
+      "loss": 0.9523,
+      "step": 11179
+    },
+    {
+      "epoch": 0.8618563058896084,
+      "grad_norm": 3.9804065227508545,
+      "learning_rate": 4.922192458992609e-07,
+      "loss": 1.0091,
+      "step": 11180
+    },
+    {
+      "epoch": 0.8619333950046254,
+      "grad_norm": 4.004647731781006,
+      "learning_rate": 4.916792305950235e-07,
+      "loss": 1.0064,
+      "step": 11181
+    },
+    {
+      "epoch": 0.8620104841196423,
+      "grad_norm": 3.7954509258270264,
+      "learning_rate": 4.911394963612359e-07,
+      "loss": 0.919,
+      "step": 11182
+    },
+    {
+      "epoch": 0.8620875732346592,
+      "grad_norm": 3.5825188159942627,
+      "learning_rate": 4.906000432315489e-07,
+      "loss": 0.9094,
+      "step": 11183
+    },
+    {
+      "epoch": 0.8621646623496763,
+      "grad_norm": 3.406367778778076,
+      "learning_rate": 4.900608712395943e-07,
+      "loss": 0.917,
+      "step": 11184
+    },
+    {
+      "epoch": 0.8622417514646932,
+      "grad_norm": 3.9264237880706787,
+      "learning_rate": 4.895219804189871e-07,
+      "loss": 0.8757,
+      "step": 11185
+    },
+    {
+      "epoch": 0.8623188405797102,
+      "grad_norm": 3.8309881687164307,
+      "learning_rate": 4.889833708033248e-07,
+      "loss": 0.9881,
+      "step": 11186
+    },
+    {
+      "epoch": 0.8623959296947271,
+      "grad_norm": 3.63254714012146,
+      "learning_rate": 4.884450424261849e-07,
+      "loss": 0.8489,
+      "step": 11187
+    },
+    {
+      "epoch": 0.862473018809744,
+      "grad_norm": 3.7826976776123047,
+      "learning_rate": 4.879069953211312e-07,
+      "loss": 0.9321,
+      "step": 11188
+    },
+    {
+      "epoch": 0.8625501079247611,
+      "grad_norm": 3.793102979660034,
+      "learning_rate": 4.873692295217076e-07,
+      "loss": 0.9313,
+      "step": 11189
+    },
+    {
+      "epoch": 0.862627197039778,
+      "grad_norm": 3.558684825897217,
+      "learning_rate": 4.868317450614407e-07,
+      "loss": 0.9335,
+      "step": 11190
+    },
+    {
+      "epoch": 0.862704286154795,
+      "grad_norm": 4.0856194496154785,
+      "learning_rate": 4.862945419738401e-07,
+      "loss": 0.9869,
+      "step": 11191
+    },
+    {
+      "epoch": 0.8627813752698119,
+      "grad_norm": 3.5604751110076904,
+      "learning_rate": 4.857576202923986e-07,
+      "loss": 0.8317,
+      "step": 11192
+    },
+    {
+      "epoch": 0.8628584643848288,
+      "grad_norm": 3.5392532348632812,
+      "learning_rate": 4.852209800505892e-07,
+      "loss": 0.8233,
+      "step": 11193
+    },
+    {
+      "epoch": 0.8629355534998459,
+      "grad_norm": 3.6630494594573975,
+      "learning_rate": 4.846846212818684e-07,
+      "loss": 0.8782,
+      "step": 11194
+    },
+    {
+      "epoch": 0.8630126426148628,
+      "grad_norm": 3.6344716548919678,
+      "learning_rate": 4.841485440196763e-07,
+      "loss": 0.8732,
+      "step": 11195
+    },
+    {
+      "epoch": 0.8630897317298798,
+      "grad_norm": 3.5769948959350586,
+      "learning_rate": 4.836127482974346e-07,
+      "loss": 0.8491,
+      "step": 11196
+    },
+    {
+      "epoch": 0.8631668208448967,
+      "grad_norm": 3.778287649154663,
+      "learning_rate": 4.830772341485479e-07,
+      "loss": 0.9085,
+      "step": 11197
+    },
+    {
+      "epoch": 0.8632439099599136,
+      "grad_norm": 3.6463840007781982,
+      "learning_rate": 4.825420016064009e-07,
+      "loss": 0.9255,
+      "step": 11198
+    },
+    {
+      "epoch": 0.8633209990749307,
+      "grad_norm": 3.8667242527008057,
+      "learning_rate": 4.820070507043633e-07,
+      "loss": 0.931,
+      "step": 11199
+    },
+    {
+      "epoch": 0.8633980881899476,
+      "grad_norm": 3.825806140899658,
+      "learning_rate": 4.814723814757871e-07,
+      "loss": 0.9954,
+      "step": 11200
+    },
+    {
+      "epoch": 0.8634751773049646,
+      "grad_norm": 3.5652754306793213,
+      "learning_rate": 4.80937993954006e-07,
+      "loss": 0.8503,
+      "step": 11201
+    },
+    {
+      "epoch": 0.8635522664199815,
+      "grad_norm": 3.5560030937194824,
+      "learning_rate": 4.804038881723361e-07,
+      "loss": 0.8271,
+      "step": 11202
+    },
+    {
+      "epoch": 0.8636293555349984,
+      "grad_norm": 3.7864773273468018,
+      "learning_rate": 4.798700641640768e-07,
+      "loss": 0.8704,
+      "step": 11203
+    },
+    {
+      "epoch": 0.8637064446500154,
+      "grad_norm": 3.6049001216888428,
+      "learning_rate": 4.793365219625079e-07,
+      "loss": 0.9203,
+      "step": 11204
+    },
+    {
+      "epoch": 0.8637835337650324,
+      "grad_norm": 3.6420271396636963,
+      "learning_rate": 4.788032616008936e-07,
+      "loss": 0.966,
+      "step": 11205
+    },
+    {
+      "epoch": 0.8638606228800494,
+      "grad_norm": 3.835549831390381,
+      "learning_rate": 4.782702831124803e-07,
+      "loss": 0.9633,
+      "step": 11206
+    },
+    {
+      "epoch": 0.8639377119950663,
+      "grad_norm": 3.744598388671875,
+      "learning_rate": 4.777375865304962e-07,
+      "loss": 0.8078,
+      "step": 11207
+    },
+    {
+      "epoch": 0.8640148011100832,
+      "grad_norm": 3.756577491760254,
+      "learning_rate": 4.772051718881532e-07,
+      "loss": 0.9566,
+      "step": 11208
+    },
+    {
+      "epoch": 0.8640918902251002,
+      "grad_norm": 3.558971643447876,
+      "learning_rate": 4.766730392186425e-07,
+      "loss": 0.9583,
+      "step": 11209
+    },
+    {
+      "epoch": 0.8641689793401172,
+      "grad_norm": 3.711453676223755,
+      "learning_rate": 4.761411885551409e-07,
+      "loss": 0.9485,
+      "step": 11210
+    },
+    {
+      "epoch": 0.8642460684551342,
+      "grad_norm": 3.7305915355682373,
+      "learning_rate": 4.7560961993080636e-07,
+      "loss": 0.9207,
+      "step": 11211
+    },
+    {
+      "epoch": 0.8643231575701511,
+      "grad_norm": 3.7739152908325195,
+      "learning_rate": 4.750783333787795e-07,
+      "loss": 0.8622,
+      "step": 11212
+    },
+    {
+      "epoch": 0.864400246685168,
+      "grad_norm": 3.8151278495788574,
+      "learning_rate": 4.745473289321839e-07,
+      "loss": 0.7792,
+      "step": 11213
+    },
+    {
+      "epoch": 0.864477335800185,
+      "grad_norm": 3.5902225971221924,
+      "learning_rate": 4.7401660662412477e-07,
+      "loss": 0.9519,
+      "step": 11214
+    },
+    {
+      "epoch": 0.864554424915202,
+      "grad_norm": 3.8135595321655273,
+      "learning_rate": 4.7348616648768886e-07,
+      "loss": 0.8904,
+      "step": 11215
+    },
+    {
+      "epoch": 0.864631514030219,
+      "grad_norm": 3.9646761417388916,
+      "learning_rate": 4.729560085559476e-07,
+      "loss": 0.8801,
+      "step": 11216
+    },
+    {
+      "epoch": 0.8647086031452359,
+      "grad_norm": 4.04872465133667,
+      "learning_rate": 4.7242613286195227e-07,
+      "loss": 1.0091,
+      "step": 11217
+    },
+    {
+      "epoch": 0.8647856922602528,
+      "grad_norm": 3.746007204055786,
+      "learning_rate": 4.718965394387387e-07,
+      "loss": 0.9377,
+      "step": 11218
+    },
+    {
+      "epoch": 0.8648627813752698,
+      "grad_norm": 3.77582049369812,
+      "learning_rate": 4.7136722831932546e-07,
+      "loss": 0.9175,
+      "step": 11219
+    },
+    {
+      "epoch": 0.8649398704902868,
+      "grad_norm": 3.808149576187134,
+      "learning_rate": 4.7083819953671007e-07,
+      "loss": 0.8849,
+      "step": 11220
+    },
+    {
+      "epoch": 0.8650169596053038,
+      "grad_norm": 3.639953851699829,
+      "learning_rate": 4.703094531238761e-07,
+      "loss": 0.8691,
+      "step": 11221
+    },
+    {
+      "epoch": 0.8650940487203207,
+      "grad_norm": 4.364095211029053,
+      "learning_rate": 4.6978098911378776e-07,
+      "loss": 0.9547,
+      "step": 11222
+    },
+    {
+      "epoch": 0.8651711378353376,
+      "grad_norm": 3.8730828762054443,
+      "learning_rate": 4.6925280753939097e-07,
+      "loss": 0.9314,
+      "step": 11223
+    },
+    {
+      "epoch": 0.8652482269503546,
+      "grad_norm": 3.9587454795837402,
+      "learning_rate": 4.687249084336182e-07,
+      "loss": 0.9972,
+      "step": 11224
+    },
+    {
+      "epoch": 0.8653253160653716,
+      "grad_norm": 3.558377504348755,
+      "learning_rate": 4.6819729182937865e-07,
+      "loss": 0.8813,
+      "step": 11225
+    },
+    {
+      "epoch": 0.8654024051803886,
+      "grad_norm": 3.674800395965576,
+      "learning_rate": 4.676699577595667e-07,
+      "loss": 0.8476,
+      "step": 11226
+    },
+    {
+      "epoch": 0.8654794942954055,
+      "grad_norm": 3.645923376083374,
+      "learning_rate": 4.6714290625705983e-07,
+      "loss": 0.9952,
+      "step": 11227
+    },
+    {
+      "epoch": 0.8655565834104224,
+      "grad_norm": 4.369297981262207,
+      "learning_rate": 4.666161373547162e-07,
+      "loss": 0.9502,
+      "step": 11228
+    },
+    {
+      "epoch": 0.8656336725254394,
+      "grad_norm": 3.8362910747528076,
+      "learning_rate": 4.660896510853785e-07,
+      "loss": 0.8966,
+      "step": 11229
+    },
+    {
+      "epoch": 0.8657107616404563,
+      "grad_norm": 4.300778388977051,
+      "learning_rate": 4.655634474818682e-07,
+      "loss": 1.024,
+      "step": 11230
+    },
+    {
+      "epoch": 0.8657878507554734,
+      "grad_norm": 3.7804079055786133,
+      "learning_rate": 4.650375265769924e-07,
+      "loss": 0.8685,
+      "step": 11231
+    },
+    {
+      "epoch": 0.8658649398704903,
+      "grad_norm": 3.908895492553711,
+      "learning_rate": 4.645118884035399e-07,
+      "loss": 0.9242,
+      "step": 11232
+    },
+    {
+      "epoch": 0.8659420289855072,
+      "grad_norm": 3.7380692958831787,
+      "learning_rate": 4.639865329942811e-07,
+      "loss": 0.8432,
+      "step": 11233
+    },
+    {
+      "epoch": 0.8660191181005242,
+      "grad_norm": 3.6968612670898438,
+      "learning_rate": 4.6346146038197035e-07,
+      "loss": 0.9012,
+      "step": 11234
+    },
+    {
+      "epoch": 0.8660962072155411,
+      "grad_norm": 3.667461395263672,
+      "learning_rate": 4.6293667059934154e-07,
+      "loss": 0.8959,
+      "step": 11235
+    },
+    {
+      "epoch": 0.8661732963305582,
+      "grad_norm": 3.7316794395446777,
+      "learning_rate": 4.624121636791129e-07,
+      "loss": 0.9219,
+      "step": 11236
+    },
+    {
+      "epoch": 0.8662503854455751,
+      "grad_norm": 3.7822744846343994,
+      "learning_rate": 4.6188793965398493e-07,
+      "loss": 0.9128,
+      "step": 11237
+    },
+    {
+      "epoch": 0.866327474560592,
+      "grad_norm": 3.7413852214813232,
+      "learning_rate": 4.6136399855664093e-07,
+      "loss": 0.9516,
+      "step": 11238
+    },
+    {
+      "epoch": 0.866404563675609,
+      "grad_norm": 3.792145013809204,
+      "learning_rate": 4.6084034041974533e-07,
+      "loss": 0.9049,
+      "step": 11239
+    },
+    {
+      "epoch": 0.866481652790626,
+      "grad_norm": 4.004141330718994,
+      "learning_rate": 4.603169652759465e-07,
+      "loss": 0.8256,
+      "step": 11240
+    },
+    {
+      "epoch": 0.866558741905643,
+      "grad_norm": 4.334702968597412,
+      "learning_rate": 4.5979387315787215e-07,
+      "loss": 0.9637,
+      "step": 11241
+    },
+    {
+      "epoch": 0.8666358310206599,
+      "grad_norm": 3.4571897983551025,
+      "learning_rate": 4.592710640981352e-07,
+      "loss": 0.8333,
+      "step": 11242
+    },
+    {
+      "epoch": 0.8667129201356768,
+      "grad_norm": 3.4184558391571045,
+      "learning_rate": 4.5874853812933107e-07,
+      "loss": 0.8278,
+      "step": 11243
+    },
+    {
+      "epoch": 0.8667900092506938,
+      "grad_norm": 3.8723700046539307,
+      "learning_rate": 4.582262952840355e-07,
+      "loss": 0.9484,
+      "step": 11244
+    },
+    {
+      "epoch": 0.8668670983657107,
+      "grad_norm": 3.8549139499664307,
+      "learning_rate": 4.5770433559480854e-07,
+      "loss": 1.0091,
+      "step": 11245
+    },
+    {
+      "epoch": 0.8669441874807278,
+      "grad_norm": 3.5513107776641846,
+      "learning_rate": 4.571826590941908e-07,
+      "loss": 0.8842,
+      "step": 11246
+    },
+    {
+      "epoch": 0.8670212765957447,
+      "grad_norm": 3.3549818992614746,
+      "learning_rate": 4.5666126581470625e-07,
+      "loss": 0.8307,
+      "step": 11247
+    },
+    {
+      "epoch": 0.8670983657107616,
+      "grad_norm": 3.714850664138794,
+      "learning_rate": 4.561401557888606e-07,
+      "loss": 0.8353,
+      "step": 11248
+    },
+    {
+      "epoch": 0.8671754548257786,
+      "grad_norm": 3.587151527404785,
+      "learning_rate": 4.5561932904914395e-07,
+      "loss": 0.9133,
+      "step": 11249
+    },
+    {
+      "epoch": 0.8672525439407955,
+      "grad_norm": 3.682015895843506,
+      "learning_rate": 4.5509878562802536e-07,
+      "loss": 0.8385,
+      "step": 11250
+    },
+    {
+      "epoch": 0.8673296330558126,
+      "grad_norm": 3.6053788661956787,
+      "learning_rate": 4.5457852555796044e-07,
+      "loss": 0.8828,
+      "step": 11251
+    },
+    {
+      "epoch": 0.8674067221708295,
+      "grad_norm": 3.85351824760437,
+      "learning_rate": 4.5405854887138165e-07,
+      "loss": 0.8888,
+      "step": 11252
+    },
+    {
+      "epoch": 0.8674838112858464,
+      "grad_norm": 4.928849220275879,
+      "learning_rate": 4.5353885560070867e-07,
+      "loss": 0.8688,
+      "step": 11253
+    },
+    {
+      "epoch": 0.8675609004008634,
+      "grad_norm": 3.783874034881592,
+      "learning_rate": 4.53019445778341e-07,
+      "loss": 0.8879,
+      "step": 11254
+    },
+    {
+      "epoch": 0.8676379895158803,
+      "grad_norm": 3.7811548709869385,
+      "learning_rate": 4.5250031943666174e-07,
+      "loss": 0.9084,
+      "step": 11255
+    },
+    {
+      "epoch": 0.8677150786308974,
+      "grad_norm": 3.439472198486328,
+      "learning_rate": 4.5198147660803605e-07,
+      "loss": 0.9078,
+      "step": 11256
+    },
+    {
+      "epoch": 0.8677921677459143,
+      "grad_norm": 3.847801923751831,
+      "learning_rate": 4.5146291732480975e-07,
+      "loss": 0.9516,
+      "step": 11257
+    },
+    {
+      "epoch": 0.8678692568609312,
+      "grad_norm": 3.690554141998291,
+      "learning_rate": 4.5094464161931305e-07,
+      "loss": 0.9484,
+      "step": 11258
+    },
+    {
+      "epoch": 0.8679463459759482,
+      "grad_norm": 3.9348044395446777,
+      "learning_rate": 4.5042664952385797e-07,
+      "loss": 0.9123,
+      "step": 11259
+    },
+    {
+      "epoch": 0.8680234350909651,
+      "grad_norm": 4.005280494689941,
+      "learning_rate": 4.499089410707386e-07,
+      "loss": 0.9793,
+      "step": 11260
+    },
+    {
+      "epoch": 0.8681005242059822,
+      "grad_norm": 4.255978584289551,
+      "learning_rate": 4.493915162922308e-07,
+      "loss": 0.8457,
+      "step": 11261
+    },
+    {
+      "epoch": 0.8681776133209991,
+      "grad_norm": 4.279727935791016,
+      "learning_rate": 4.4887437522059487e-07,
+      "loss": 0.9742,
+      "step": 11262
+    },
+    {
+      "epoch": 0.868254702436016,
+      "grad_norm": 3.771867036819458,
+      "learning_rate": 4.4835751788807e-07,
+      "loss": 0.9787,
+      "step": 11263
+    },
+    {
+      "epoch": 0.868331791551033,
+      "grad_norm": 3.949169635772705,
+      "learning_rate": 4.4784094432687997e-07,
+      "loss": 0.9254,
+      "step": 11264
+    },
+    {
+      "epoch": 0.8684088806660499,
+      "grad_norm": 4.027673244476318,
+      "learning_rate": 4.4732465456923113e-07,
+      "loss": 0.9364,
+      "step": 11265
+    },
+    {
+      "epoch": 0.868485969781067,
+      "grad_norm": 3.5182044506073,
+      "learning_rate": 4.468086486473111e-07,
+      "loss": 1.0193,
+      "step": 11266
+    },
+    {
+      "epoch": 0.8685630588960839,
+      "grad_norm": 3.757246494293213,
+      "learning_rate": 4.462929265932897e-07,
+      "loss": 0.9157,
+      "step": 11267
+    },
+    {
+      "epoch": 0.8686401480111008,
+      "grad_norm": 3.785848379135132,
+      "learning_rate": 4.457774884393207e-07,
+      "loss": 0.8675,
+      "step": 11268
+    },
+    {
+      "epoch": 0.8687172371261178,
+      "grad_norm": 4.203325271606445,
+      "learning_rate": 4.452623342175383e-07,
+      "loss": 0.9538,
+      "step": 11269
+    },
+    {
+      "epoch": 0.8687943262411347,
+      "grad_norm": 3.8824243545532227,
+      "learning_rate": 4.447474639600596e-07,
+      "loss": 0.9154,
+      "step": 11270
+    },
+    {
+      "epoch": 0.8688714153561518,
+      "grad_norm": 3.8376996517181396,
+      "learning_rate": 4.442328776989846e-07,
+      "loss": 0.9482,
+      "step": 11271
+    },
+    {
+      "epoch": 0.8689485044711687,
+      "grad_norm": 3.8418562412261963,
+      "learning_rate": 4.437185754663953e-07,
+      "loss": 0.9741,
+      "step": 11272
+    },
+    {
+      "epoch": 0.8690255935861856,
+      "grad_norm": 3.5321462154388428,
+      "learning_rate": 4.432045572943544e-07,
+      "loss": 0.885,
+      "step": 11273
+    },
+    {
+      "epoch": 0.8691026827012026,
+      "grad_norm": 4.0153374671936035,
+      "learning_rate": 4.4269082321490906e-07,
+      "loss": 0.9149,
+      "step": 11274
+    },
+    {
+      "epoch": 0.8691797718162195,
+      "grad_norm": 4.0503129959106445,
+      "learning_rate": 4.4217737326008814e-07,
+      "loss": 0.9256,
+      "step": 11275
+    },
+    {
+      "epoch": 0.8692568609312366,
+      "grad_norm": 3.9747254848480225,
+      "learning_rate": 4.4166420746190206e-07,
+      "loss": 0.9465,
+      "step": 11276
+    },
+    {
+      "epoch": 0.8693339500462535,
+      "grad_norm": 3.6550421714782715,
+      "learning_rate": 4.411513258523459e-07,
+      "loss": 0.8072,
+      "step": 11277
+    },
+    {
+      "epoch": 0.8694110391612704,
+      "grad_norm": 3.5995473861694336,
+      "learning_rate": 4.406387284633923e-07,
+      "loss": 0.7394,
+      "step": 11278
+    },
+    {
+      "epoch": 0.8694881282762874,
+      "grad_norm": 3.6607563495635986,
+      "learning_rate": 4.4012641532700075e-07,
+      "loss": 0.9314,
+      "step": 11279
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 3.560035228729248,
+      "learning_rate": 4.3961438647511066e-07,
+      "loss": 0.8134,
+      "step": 11280
+    },
+    {
+      "epoch": 0.8696423065063213,
+      "grad_norm": 3.919574022293091,
+      "learning_rate": 4.391026419396449e-07,
+      "loss": 0.9771,
+      "step": 11281
+    },
+    {
+      "epoch": 0.8697193956213383,
+      "grad_norm": 3.7704415321350098,
+      "learning_rate": 4.3859118175250905e-07,
+      "loss": 0.8802,
+      "step": 11282
+    },
+    {
+      "epoch": 0.8697964847363552,
+      "grad_norm": 3.845430850982666,
+      "learning_rate": 4.3808000594558754e-07,
+      "loss": 0.8919,
+      "step": 11283
+    },
+    {
+      "epoch": 0.8698735738513722,
+      "grad_norm": 3.642606735229492,
+      "learning_rate": 4.3756911455075155e-07,
+      "loss": 0.8085,
+      "step": 11284
+    },
+    {
+      "epoch": 0.8699506629663891,
+      "grad_norm": 3.7264983654022217,
+      "learning_rate": 4.3705850759985127e-07,
+      "loss": 0.9121,
+      "step": 11285
+    },
+    {
+      "epoch": 0.8700277520814061,
+      "grad_norm": 3.6813814640045166,
+      "learning_rate": 4.3654818512472106e-07,
+      "loss": 0.8442,
+      "step": 11286
+    },
+    {
+      "epoch": 0.8701048411964231,
+      "grad_norm": 4.019516468048096,
+      "learning_rate": 4.3603814715717674e-07,
+      "loss": 0.8202,
+      "step": 11287
+    },
+    {
+      "epoch": 0.87018193031144,
+      "grad_norm": 3.6544294357299805,
+      "learning_rate": 4.355283937290178e-07,
+      "loss": 0.8273,
+      "step": 11288
+    },
+    {
+      "epoch": 0.870259019426457,
+      "grad_norm": 3.649919271469116,
+      "learning_rate": 4.350189248720221e-07,
+      "loss": 0.876,
+      "step": 11289
+    },
+    {
+      "epoch": 0.8703361085414739,
+      "grad_norm": 3.7568275928497314,
+      "learning_rate": 4.3450974061795437e-07,
+      "loss": 0.9599,
+      "step": 11290
+    },
+    {
+      "epoch": 0.870413197656491,
+      "grad_norm": 3.7730679512023926,
+      "learning_rate": 4.3400084099855854e-07,
+      "loss": 0.9461,
+      "step": 11291
+    },
+    {
+      "epoch": 0.8704902867715079,
+      "grad_norm": 4.046023845672607,
+      "learning_rate": 4.334922260455626e-07,
+      "loss": 0.9321,
+      "step": 11292
+    },
+    {
+      "epoch": 0.8705673758865248,
+      "grad_norm": 3.993216037750244,
+      "learning_rate": 4.3298389579067677e-07,
+      "loss": 0.8991,
+      "step": 11293
+    },
+    {
+      "epoch": 0.8706444650015418,
+      "grad_norm": 3.6614949703216553,
+      "learning_rate": 4.324758502655907e-07,
+      "loss": 0.9043,
+      "step": 11294
+    },
+    {
+      "epoch": 0.8707215541165587,
+      "grad_norm": 3.688453197479248,
+      "learning_rate": 4.3196808950197954e-07,
+      "loss": 0.9041,
+      "step": 11295
+    },
+    {
+      "epoch": 0.8707986432315757,
+      "grad_norm": 3.937119245529175,
+      "learning_rate": 4.314606135314997e-07,
+      "loss": 0.9584,
+      "step": 11296
+    },
+    {
+      "epoch": 0.8708757323465927,
+      "grad_norm": 3.8757388591766357,
+      "learning_rate": 4.3095342238578974e-07,
+      "loss": 0.8873,
+      "step": 11297
+    },
+    {
+      "epoch": 0.8709528214616096,
+      "grad_norm": 3.7967705726623535,
+      "learning_rate": 4.304465160964699e-07,
+      "loss": 0.8722,
+      "step": 11298
+    },
+    {
+      "epoch": 0.8710299105766266,
+      "grad_norm": 3.7907609939575195,
+      "learning_rate": 4.299398946951444e-07,
+      "loss": 0.9177,
+      "step": 11299
+    },
+    {
+      "epoch": 0.8711069996916435,
+      "grad_norm": 3.48050594329834,
+      "learning_rate": 4.294335582133968e-07,
+      "loss": 0.9796,
+      "step": 11300
+    },
+    {
+      "epoch": 0.8711840888066605,
+      "grad_norm": 3.578596830368042,
+      "learning_rate": 4.289275066827947e-07,
+      "loss": 0.9109,
+      "step": 11301
+    },
+    {
+      "epoch": 0.8712611779216775,
+      "grad_norm": 3.6778552532196045,
+      "learning_rate": 4.284217401348889e-07,
+      "loss": 0.9468,
+      "step": 11302
+    },
+    {
+      "epoch": 0.8713382670366944,
+      "grad_norm": 4.028721809387207,
+      "learning_rate": 4.2791625860121087e-07,
+      "loss": 0.8429,
+      "step": 11303
+    },
+    {
+      "epoch": 0.8714153561517114,
+      "grad_norm": 3.7571585178375244,
+      "learning_rate": 4.2741106211327544e-07,
+      "loss": 0.9363,
+      "step": 11304
+    },
+    {
+      "epoch": 0.8714924452667283,
+      "grad_norm": 3.5501813888549805,
+      "learning_rate": 4.2690615070257737e-07,
+      "loss": 0.8177,
+      "step": 11305
+    },
+    {
+      "epoch": 0.8715695343817453,
+      "grad_norm": 3.985402822494507,
+      "learning_rate": 4.264015244005959e-07,
+      "loss": 0.8521,
+      "step": 11306
+    },
+    {
+      "epoch": 0.8716466234967623,
+      "grad_norm": 3.679849624633789,
+      "learning_rate": 4.258971832387926e-07,
+      "loss": 0.9388,
+      "step": 11307
+    },
+    {
+      "epoch": 0.8717237126117792,
+      "grad_norm": 3.909078359603882,
+      "learning_rate": 4.253931272486095e-07,
+      "loss": 0.9455,
+      "step": 11308
+    },
+    {
+      "epoch": 0.8718008017267962,
+      "grad_norm": 3.8097403049468994,
+      "learning_rate": 4.2488935646147253e-07,
+      "loss": 0.9776,
+      "step": 11309
+    },
+    {
+      "epoch": 0.8718778908418131,
+      "grad_norm": 3.7239317893981934,
+      "learning_rate": 4.2438587090879e-07,
+      "loss": 0.8996,
+      "step": 11310
+    },
+    {
+      "epoch": 0.8719549799568301,
+      "grad_norm": 3.8013432025909424,
+      "learning_rate": 4.23882670621949e-07,
+      "loss": 0.9272,
+      "step": 11311
+    },
+    {
+      "epoch": 0.872032069071847,
+      "grad_norm": 4.090017795562744,
+      "learning_rate": 4.2337975563232437e-07,
+      "loss": 0.8814,
+      "step": 11312
+    },
+    {
+      "epoch": 0.872109158186864,
+      "grad_norm": 3.7773220539093018,
+      "learning_rate": 4.2287712597126884e-07,
+      "loss": 0.924,
+      "step": 11313
+    },
+    {
+      "epoch": 0.872186247301881,
+      "grad_norm": 3.6810662746429443,
+      "learning_rate": 4.223747816701196e-07,
+      "loss": 0.8808,
+      "step": 11314
+    },
+    {
+      "epoch": 0.8722633364168979,
+      "grad_norm": 3.542295455932617,
+      "learning_rate": 4.2187272276019373e-07,
+      "loss": 0.7952,
+      "step": 11315
+    },
+    {
+      "epoch": 0.8723404255319149,
+      "grad_norm": 3.5852906703948975,
+      "learning_rate": 4.2137094927279296e-07,
+      "loss": 0.8106,
+      "step": 11316
+    },
+    {
+      "epoch": 0.8724175146469318,
+      "grad_norm": 3.841078281402588,
+      "learning_rate": 4.208694612392006e-07,
+      "loss": 0.9806,
+      "step": 11317
+    },
+    {
+      "epoch": 0.8724946037619488,
+      "grad_norm": 3.797222852706909,
+      "learning_rate": 4.2036825869068097e-07,
+      "loss": 0.9715,
+      "step": 11318
+    },
+    {
+      "epoch": 0.8725716928769658,
+      "grad_norm": 3.6867878437042236,
+      "learning_rate": 4.19867341658482e-07,
+      "loss": 0.7848,
+      "step": 11319
+    },
+    {
+      "epoch": 0.8726487819919827,
+      "grad_norm": 3.439988613128662,
+      "learning_rate": 4.1936671017383356e-07,
+      "loss": 0.8382,
+      "step": 11320
+    },
+    {
+      "epoch": 0.8727258711069997,
+      "grad_norm": 3.4686200618743896,
+      "learning_rate": 4.188663642679469e-07,
+      "loss": 0.7944,
+      "step": 11321
+    },
+    {
+      "epoch": 0.8728029602220166,
+      "grad_norm": 3.7739365100860596,
+      "learning_rate": 4.1836630397201593e-07,
+      "loss": 0.8806,
+      "step": 11322
+    },
+    {
+      "epoch": 0.8728800493370336,
+      "grad_norm": 4.114104747772217,
+      "learning_rate": 4.1786652931721683e-07,
+      "loss": 0.8682,
+      "step": 11323
+    },
+    {
+      "epoch": 0.8729571384520506,
+      "grad_norm": 3.7291269302368164,
+      "learning_rate": 4.17367040334708e-07,
+      "loss": 0.8022,
+      "step": 11324
+    },
+    {
+      "epoch": 0.8730342275670675,
+      "grad_norm": 4.308863162994385,
+      "learning_rate": 4.1686783705563115e-07,
+      "loss": 1.0564,
+      "step": 11325
+    },
+    {
+      "epoch": 0.8731113166820845,
+      "grad_norm": 3.66945481300354,
+      "learning_rate": 4.163689195111076e-07,
+      "loss": 0.8329,
+      "step": 11326
+    },
+    {
+      "epoch": 0.8731884057971014,
+      "grad_norm": 3.6699118614196777,
+      "learning_rate": 4.158702877322418e-07,
+      "loss": 0.8636,
+      "step": 11327
+    },
+    {
+      "epoch": 0.8732654949121184,
+      "grad_norm": 3.587113618850708,
+      "learning_rate": 4.153719417501223e-07,
+      "loss": 0.8866,
+      "step": 11328
+    },
+    {
+      "epoch": 0.8733425840271354,
+      "grad_norm": 3.8809823989868164,
+      "learning_rate": 4.148738815958181e-07,
+      "loss": 0.8977,
+      "step": 11329
+    },
+    {
+      "epoch": 0.8734196731421523,
+      "grad_norm": 3.6669769287109375,
+      "learning_rate": 4.14376107300381e-07,
+      "loss": 0.8677,
+      "step": 11330
+    },
+    {
+      "epoch": 0.8734967622571693,
+      "grad_norm": 3.631700038909912,
+      "learning_rate": 4.1387861889484294e-07,
+      "loss": 0.9199,
+      "step": 11331
+    },
+    {
+      "epoch": 0.8735738513721862,
+      "grad_norm": 4.055285930633545,
+      "learning_rate": 4.1338141641022125e-07,
+      "loss": 0.9046,
+      "step": 11332
+    },
+    {
+      "epoch": 0.8736509404872032,
+      "grad_norm": 3.6422579288482666,
+      "learning_rate": 4.128844998775133e-07,
+      "loss": 1.0013,
+      "step": 11333
+    },
+    {
+      "epoch": 0.8737280296022202,
+      "grad_norm": 3.686168909072876,
+      "learning_rate": 4.1238786932769947e-07,
+      "loss": 0.9467,
+      "step": 11334
+    },
+    {
+      "epoch": 0.8738051187172371,
+      "grad_norm": 3.7894532680511475,
+      "learning_rate": 4.118915247917421e-07,
+      "loss": 0.884,
+      "step": 11335
+    },
+    {
+      "epoch": 0.8738822078322541,
+      "grad_norm": 3.6736958026885986,
+      "learning_rate": 4.113954663005865e-07,
+      "loss": 0.9269,
+      "step": 11336
+    },
+    {
+      "epoch": 0.873959296947271,
+      "grad_norm": 3.9054605960845947,
+      "learning_rate": 4.108996938851578e-07,
+      "loss": 0.9955,
+      "step": 11337
+    },
+    {
+      "epoch": 0.874036386062288,
+      "grad_norm": 3.832731008529663,
+      "learning_rate": 4.104042075763659e-07,
+      "loss": 0.96,
+      "step": 11338
+    },
+    {
+      "epoch": 0.874113475177305,
+      "grad_norm": 3.614189386367798,
+      "learning_rate": 4.0990900740510155e-07,
+      "loss": 0.8904,
+      "step": 11339
+    },
+    {
+      "epoch": 0.8741905642923219,
+      "grad_norm": 3.8343393802642822,
+      "learning_rate": 4.094140934022378e-07,
+      "loss": 0.8866,
+      "step": 11340
+    },
+    {
+      "epoch": 0.8742676534073389,
+      "grad_norm": 4.005188465118408,
+      "learning_rate": 4.089194655986306e-07,
+      "loss": 1.0047,
+      "step": 11341
+    },
+    {
+      "epoch": 0.8743447425223558,
+      "grad_norm": 3.9065210819244385,
+      "learning_rate": 4.084251240251164e-07,
+      "loss": 0.992,
+      "step": 11342
+    },
+    {
+      "epoch": 0.8744218316373727,
+      "grad_norm": 3.6721572875976562,
+      "learning_rate": 4.079310687125154e-07,
+      "loss": 0.8878,
+      "step": 11343
+    },
+    {
+      "epoch": 0.8744989207523898,
+      "grad_norm": 3.8441555500030518,
+      "learning_rate": 4.0743729969162924e-07,
+      "loss": 0.9359,
+      "step": 11344
+    },
+    {
+      "epoch": 0.8745760098674067,
+      "grad_norm": 3.992184638977051,
+      "learning_rate": 4.0694381699324157e-07,
+      "loss": 0.9161,
+      "step": 11345
+    },
+    {
+      "epoch": 0.8746530989824237,
+      "grad_norm": 3.973717451095581,
+      "learning_rate": 4.0645062064811945e-07,
+      "loss": 0.9411,
+      "step": 11346
+    },
+    {
+      "epoch": 0.8747301880974406,
+      "grad_norm": 3.883294105529785,
+      "learning_rate": 4.059577106870111e-07,
+      "loss": 0.9405,
+      "step": 11347
+    },
+    {
+      "epoch": 0.8748072772124575,
+      "grad_norm": 3.9654433727264404,
+      "learning_rate": 4.054650871406451e-07,
+      "loss": 0.9231,
+      "step": 11348
+    },
+    {
+      "epoch": 0.8748843663274746,
+      "grad_norm": 3.9687893390655518,
+      "learning_rate": 4.0497275003973604e-07,
+      "loss": 0.9068,
+      "step": 11349
+    },
+    {
+      "epoch": 0.8749614554424915,
+      "grad_norm": 3.720082998275757,
+      "learning_rate": 4.044806994149769e-07,
+      "loss": 0.9742,
+      "step": 11350
+    },
+    {
+      "epoch": 0.8750385445575085,
+      "grad_norm": 3.890352964401245,
+      "learning_rate": 4.039889352970461e-07,
+      "loss": 0.9705,
+      "step": 11351
+    },
+    {
+      "epoch": 0.8751156336725254,
+      "grad_norm": 4.000323295593262,
+      "learning_rate": 4.0349745771660233e-07,
+      "loss": 0.9054,
+      "step": 11352
+    },
+    {
+      "epoch": 0.8751927227875425,
+      "grad_norm": 4.110193729400635,
+      "learning_rate": 4.03006266704285e-07,
+      "loss": 0.8724,
+      "step": 11353
+    },
+    {
+      "epoch": 0.8752698119025594,
+      "grad_norm": 3.643488645553589,
+      "learning_rate": 4.0251536229071906e-07,
+      "loss": 0.8721,
+      "step": 11354
+    },
+    {
+      "epoch": 0.8753469010175763,
+      "grad_norm": 3.35498309135437,
+      "learning_rate": 4.0202474450650786e-07,
+      "loss": 0.8364,
+      "step": 11355
+    },
+    {
+      "epoch": 0.8754239901325933,
+      "grad_norm": 3.5333361625671387,
+      "learning_rate": 4.0153441338224187e-07,
+      "loss": 0.8993,
+      "step": 11356
+    },
+    {
+      "epoch": 0.8755010792476102,
+      "grad_norm": 3.7463607788085938,
+      "learning_rate": 4.010443689484894e-07,
+      "loss": 0.8257,
+      "step": 11357
+    },
+    {
+      "epoch": 0.8755781683626273,
+      "grad_norm": 3.9253594875335693,
+      "learning_rate": 4.0055461123580166e-07,
+      "loss": 0.921,
+      "step": 11358
+    },
+    {
+      "epoch": 0.8756552574776442,
+      "grad_norm": 4.455300331115723,
+      "learning_rate": 4.0006514027471243e-07,
+      "loss": 0.9877,
+      "step": 11359
+    },
+    {
+      "epoch": 0.8757323465926611,
+      "grad_norm": 3.6334636211395264,
+      "learning_rate": 3.9957595609573794e-07,
+      "loss": 0.824,
+      "step": 11360
+    },
+    {
+      "epoch": 0.8758094357076781,
+      "grad_norm": 4.01237154006958,
+      "learning_rate": 3.99087058729376e-07,
+      "loss": 0.8933,
+      "step": 11361
+    },
+    {
+      "epoch": 0.875886524822695,
+      "grad_norm": 3.8893368244171143,
+      "learning_rate": 3.985984482061089e-07,
+      "loss": 0.9028,
+      "step": 11362
+    },
+    {
+      "epoch": 0.875963613937712,
+      "grad_norm": 3.884305238723755,
+      "learning_rate": 3.981101245563962e-07,
+      "loss": 0.9565,
+      "step": 11363
+    },
+    {
+      "epoch": 0.876040703052729,
+      "grad_norm": 4.136764049530029,
+      "learning_rate": 3.976220878106829e-07,
+      "loss": 0.958,
+      "step": 11364
+    },
+    {
+      "epoch": 0.8761177921677459,
+      "grad_norm": 4.275717258453369,
+      "learning_rate": 3.971343379993964e-07,
+      "loss": 0.9941,
+      "step": 11365
+    },
+    {
+      "epoch": 0.8761948812827629,
+      "grad_norm": 3.4475440979003906,
+      "learning_rate": 3.9664687515294566e-07,
+      "loss": 0.8929,
+      "step": 11366
+    },
+    {
+      "epoch": 0.8762719703977798,
+      "grad_norm": 3.3091351985931396,
+      "learning_rate": 3.9615969930172027e-07,
+      "loss": 0.7643,
+      "step": 11367
+    },
+    {
+      "epoch": 0.8763490595127968,
+      "grad_norm": 3.8550055027008057,
+      "learning_rate": 3.9567281047609427e-07,
+      "loss": 0.8662,
+      "step": 11368
+    },
+    {
+      "epoch": 0.8764261486278138,
+      "grad_norm": 3.535909414291382,
+      "learning_rate": 3.9518620870642176e-07,
+      "loss": 0.9508,
+      "step": 11369
+    },
+    {
+      "epoch": 0.8765032377428307,
+      "grad_norm": 4.034812927246094,
+      "learning_rate": 3.946998940230401e-07,
+      "loss": 0.8804,
+      "step": 11370
+    },
+    {
+      "epoch": 0.8765803268578477,
+      "grad_norm": 3.9649436473846436,
+      "learning_rate": 3.9421386645626894e-07,
+      "loss": 1.0498,
+      "step": 11371
+    },
+    {
+      "epoch": 0.8766574159728646,
+      "grad_norm": 3.772454261779785,
+      "learning_rate": 3.9372812603640844e-07,
+      "loss": 0.9142,
+      "step": 11372
+    },
+    {
+      "epoch": 0.8767345050878816,
+      "grad_norm": 3.668459415435791,
+      "learning_rate": 3.932426727937444e-07,
+      "loss": 0.8981,
+      "step": 11373
+    },
+    {
+      "epoch": 0.8768115942028986,
+      "grad_norm": 4.301527500152588,
+      "learning_rate": 3.927575067585393e-07,
+      "loss": 0.7799,
+      "step": 11374
+    },
+    {
+      "epoch": 0.8768886833179155,
+      "grad_norm": 3.583970546722412,
+      "learning_rate": 3.9227262796104226e-07,
+      "loss": 0.9783,
+      "step": 11375
+    },
+    {
+      "epoch": 0.8769657724329325,
+      "grad_norm": 4.123873710632324,
+      "learning_rate": 3.91788036431483e-07,
+      "loss": 0.9984,
+      "step": 11376
+    },
+    {
+      "epoch": 0.8770428615479494,
+      "grad_norm": 3.809326648712158,
+      "learning_rate": 3.9130373220007347e-07,
+      "loss": 0.9792,
+      "step": 11377
+    },
+    {
+      "epoch": 0.8771199506629664,
+      "grad_norm": 3.716824769973755,
+      "learning_rate": 3.9081971529700726e-07,
+      "loss": 0.8683,
+      "step": 11378
+    },
+    {
+      "epoch": 0.8771970397779834,
+      "grad_norm": 3.9373066425323486,
+      "learning_rate": 3.903359857524597e-07,
+      "loss": 0.8693,
+      "step": 11379
+    },
+    {
+      "epoch": 0.8772741288930003,
+      "grad_norm": 3.775418996810913,
+      "learning_rate": 3.898525435965894e-07,
+      "loss": 0.8815,
+      "step": 11380
+    },
+    {
+      "epoch": 0.8773512180080173,
+      "grad_norm": 3.630040168762207,
+      "learning_rate": 3.893693888595368e-07,
+      "loss": 0.9739,
+      "step": 11381
+    },
+    {
+      "epoch": 0.8774283071230342,
+      "grad_norm": 3.7354750633239746,
+      "learning_rate": 3.8888652157142327e-07,
+      "loss": 0.8161,
+      "step": 11382
+    },
+    {
+      "epoch": 0.8775053962380512,
+      "grad_norm": 3.6053590774536133,
+      "learning_rate": 3.8840394176235365e-07,
+      "loss": 0.9304,
+      "step": 11383
+    },
+    {
+      "epoch": 0.8775824853530682,
+      "grad_norm": 3.883542537689209,
+      "learning_rate": 3.87921649462415e-07,
+      "loss": 0.8643,
+      "step": 11384
+    },
+    {
+      "epoch": 0.8776595744680851,
+      "grad_norm": 3.824704647064209,
+      "learning_rate": 3.8743964470167427e-07,
+      "loss": 0.8747,
+      "step": 11385
+    },
+    {
+      "epoch": 0.8777366635831021,
+      "grad_norm": 3.9502909183502197,
+      "learning_rate": 3.8695792751018257e-07,
+      "loss": 0.9046,
+      "step": 11386
+    },
+    {
+      "epoch": 0.877813752698119,
+      "grad_norm": 3.645355224609375,
+      "learning_rate": 3.864764979179725e-07,
+      "loss": 0.89,
+      "step": 11387
+    },
+    {
+      "epoch": 0.877890841813136,
+      "grad_norm": 3.920626163482666,
+      "learning_rate": 3.859953559550589e-07,
+      "loss": 1.0779,
+      "step": 11388
+    },
+    {
+      "epoch": 0.877967930928153,
+      "grad_norm": 3.494645595550537,
+      "learning_rate": 3.855145016514389e-07,
+      "loss": 0.9319,
+      "step": 11389
+    },
+    {
+      "epoch": 0.8780450200431699,
+      "grad_norm": 3.413755178451538,
+      "learning_rate": 3.8503393503708965e-07,
+      "loss": 0.9129,
+      "step": 11390
+    },
+    {
+      "epoch": 0.8781221091581869,
+      "grad_norm": 3.9198455810546875,
+      "learning_rate": 3.8455365614197327e-07,
+      "loss": 1.0013,
+      "step": 11391
+    },
+    {
+      "epoch": 0.8781991982732038,
+      "grad_norm": 4.102309703826904,
+      "learning_rate": 3.840736649960325e-07,
+      "loss": 0.9704,
+      "step": 11392
+    },
+    {
+      "epoch": 0.8782762873882208,
+      "grad_norm": 3.7239162921905518,
+      "learning_rate": 3.8359396162919225e-07,
+      "loss": 0.8796,
+      "step": 11393
+    },
+    {
+      "epoch": 0.8783533765032377,
+      "grad_norm": 3.941939115524292,
+      "learning_rate": 3.831145460713592e-07,
+      "loss": 0.9241,
+      "step": 11394
+    },
+    {
+      "epoch": 0.8784304656182547,
+      "grad_norm": 3.810060977935791,
+      "learning_rate": 3.826354183524239e-07,
+      "loss": 0.8778,
+      "step": 11395
+    },
+    {
+      "epoch": 0.8785075547332717,
+      "grad_norm": 3.513016700744629,
+      "learning_rate": 3.821565785022552e-07,
+      "loss": 0.9103,
+      "step": 11396
+    },
+    {
+      "epoch": 0.8785846438482886,
+      "grad_norm": 3.6620430946350098,
+      "learning_rate": 3.816780265507075e-07,
+      "loss": 0.9333,
+      "step": 11397
+    },
+    {
+      "epoch": 0.8786617329633056,
+      "grad_norm": 3.8001415729522705,
+      "learning_rate": 3.811997625276165e-07,
+      "loss": 0.8179,
+      "step": 11398
+    },
+    {
+      "epoch": 0.8787388220783225,
+      "grad_norm": 3.8821539878845215,
+      "learning_rate": 3.807217864627982e-07,
+      "loss": 0.9516,
+      "step": 11399
+    },
+    {
+      "epoch": 0.8788159111933395,
+      "grad_norm": 3.6851859092712402,
+      "learning_rate": 3.802440983860528e-07,
+      "loss": 0.9707,
+      "step": 11400
+    },
+    {
+      "epoch": 0.8788930003083565,
+      "grad_norm": 4.240994453430176,
+      "learning_rate": 3.7976669832716193e-07,
+      "loss": 0.9472,
+      "step": 11401
+    },
+    {
+      "epoch": 0.8789700894233734,
+      "grad_norm": 3.83020281791687,
+      "learning_rate": 3.7928958631588795e-07,
+      "loss": 0.9476,
+      "step": 11402
+    },
+    {
+      "epoch": 0.8790471785383904,
+      "grad_norm": 3.228642702102661,
+      "learning_rate": 3.788127623819776e-07,
+      "loss": 0.7694,
+      "step": 11403
+    },
+    {
+      "epoch": 0.8791242676534073,
+      "grad_norm": 3.589622735977173,
+      "learning_rate": 3.783362265551577e-07,
+      "loss": 0.9188,
+      "step": 11404
+    },
+    {
+      "epoch": 0.8792013567684243,
+      "grad_norm": 3.4474847316741943,
+      "learning_rate": 3.7785997886513827e-07,
+      "loss": 0.8424,
+      "step": 11405
+    },
+    {
+      "epoch": 0.8792784458834413,
+      "grad_norm": 3.7509865760803223,
+      "learning_rate": 3.7738401934161006e-07,
+      "loss": 0.883,
+      "step": 11406
+    },
+    {
+      "epoch": 0.8793555349984582,
+      "grad_norm": 3.862302541732788,
+      "learning_rate": 3.7690834801424714e-07,
+      "loss": 0.8483,
+      "step": 11407
+    },
+    {
+      "epoch": 0.8794326241134752,
+      "grad_norm": 3.5118508338928223,
+      "learning_rate": 3.764329649127046e-07,
+      "loss": 0.8729,
+      "step": 11408
+    },
+    {
+      "epoch": 0.8795097132284921,
+      "grad_norm": 3.7637479305267334,
+      "learning_rate": 3.759578700666211e-07,
+      "loss": 0.9572,
+      "step": 11409
+    },
+    {
+      "epoch": 0.879586802343509,
+      "grad_norm": 3.7532765865325928,
+      "learning_rate": 3.754830635056167e-07,
+      "loss": 0.9194,
+      "step": 11410
+    },
+    {
+      "epoch": 0.8796638914585261,
+      "grad_norm": 3.501901865005493,
+      "learning_rate": 3.750085452592911e-07,
+      "loss": 0.7931,
+      "step": 11411
+    },
+    {
+      "epoch": 0.879740980573543,
+      "grad_norm": 3.859379529953003,
+      "learning_rate": 3.745343153572295e-07,
+      "loss": 0.8634,
+      "step": 11412
+    },
+    {
+      "epoch": 0.87981806968856,
+      "grad_norm": 3.430656671524048,
+      "learning_rate": 3.7406037382899716e-07,
+      "loss": 0.9078,
+      "step": 11413
+    },
+    {
+      "epoch": 0.8798951588035769,
+      "grad_norm": 3.642101526260376,
+      "learning_rate": 3.735867207041427e-07,
+      "loss": 0.9528,
+      "step": 11414
+    },
+    {
+      "epoch": 0.8799722479185939,
+      "grad_norm": 3.803781270980835,
+      "learning_rate": 3.7311335601219524e-07,
+      "loss": 0.8334,
+      "step": 11415
+    },
+    {
+      "epoch": 0.8800493370336109,
+      "grad_norm": 3.886075496673584,
+      "learning_rate": 3.7264027978266727e-07,
+      "loss": 0.945,
+      "step": 11416
+    },
+    {
+      "epoch": 0.8801264261486278,
+      "grad_norm": 3.6445107460021973,
+      "learning_rate": 3.7216749204505187e-07,
+      "loss": 0.8265,
+      "step": 11417
+    },
+    {
+      "epoch": 0.8802035152636448,
+      "grad_norm": 4.004867076873779,
+      "learning_rate": 3.7169499282882435e-07,
+      "loss": 0.8832,
+      "step": 11418
+    },
+    {
+      "epoch": 0.8802806043786617,
+      "grad_norm": 3.6422877311706543,
+      "learning_rate": 3.71222782163444e-07,
+      "loss": 0.982,
+      "step": 11419
+    },
+    {
+      "epoch": 0.8803576934936787,
+      "grad_norm": 3.556857109069824,
+      "learning_rate": 3.707508600783499e-07,
+      "loss": 0.8123,
+      "step": 11420
+    },
+    {
+      "epoch": 0.8804347826086957,
+      "grad_norm": 3.529580593109131,
+      "learning_rate": 3.7027922660296533e-07,
+      "loss": 0.8967,
+      "step": 11421
+    },
+    {
+      "epoch": 0.8805118717237126,
+      "grad_norm": 3.779604911804199,
+      "learning_rate": 3.6980788176669223e-07,
+      "loss": 0.8938,
+      "step": 11422
+    },
+    {
+      "epoch": 0.8805889608387296,
+      "grad_norm": 3.4977400302886963,
+      "learning_rate": 3.6933682559891717e-07,
+      "loss": 0.9388,
+      "step": 11423
+    },
+    {
+      "epoch": 0.8806660499537465,
+      "grad_norm": 3.975605010986328,
+      "learning_rate": 3.6886605812900766e-07,
+      "loss": 0.9501,
+      "step": 11424
+    },
+    {
+      "epoch": 0.8807431390687634,
+      "grad_norm": 4.2428202629089355,
+      "learning_rate": 3.683955793863148e-07,
+      "loss": 0.9384,
+      "step": 11425
+    },
+    {
+      "epoch": 0.8808202281837805,
+      "grad_norm": 3.509303092956543,
+      "learning_rate": 3.6792538940017054e-07,
+      "loss": 0.9074,
+      "step": 11426
+    },
+    {
+      "epoch": 0.8808973172987974,
+      "grad_norm": 3.5925185680389404,
+      "learning_rate": 3.6745548819988763e-07,
+      "loss": 0.7849,
+      "step": 11427
+    },
+    {
+      "epoch": 0.8809744064138144,
+      "grad_norm": 3.5892629623413086,
+      "learning_rate": 3.669858758147621e-07,
+      "loss": 0.8397,
+      "step": 11428
+    },
+    {
+      "epoch": 0.8810514955288313,
+      "grad_norm": 3.527111291885376,
+      "learning_rate": 3.6651655227407214e-07,
+      "loss": 0.7728,
+      "step": 11429
+    },
+    {
+      "epoch": 0.8811285846438482,
+      "grad_norm": 3.720457077026367,
+      "learning_rate": 3.660475176070777e-07,
+      "loss": 0.8121,
+      "step": 11430
+    },
+    {
+      "epoch": 0.8812056737588653,
+      "grad_norm": 3.873420000076294,
+      "learning_rate": 3.6557877184302093e-07,
+      "loss": 0.9066,
+      "step": 11431
+    },
+    {
+      "epoch": 0.8812827628738822,
+      "grad_norm": 3.740910530090332,
+      "learning_rate": 3.6511031501112625e-07,
+      "loss": 0.8271,
+      "step": 11432
+    },
+    {
+      "epoch": 0.8813598519888992,
+      "grad_norm": 3.7194406986236572,
+      "learning_rate": 3.6464214714059754e-07,
+      "loss": 0.9262,
+      "step": 11433
+    },
+    {
+      "epoch": 0.8814369411039161,
+      "grad_norm": 3.7724802494049072,
+      "learning_rate": 3.641742682606242e-07,
+      "loss": 0.7748,
+      "step": 11434
+    },
+    {
+      "epoch": 0.881514030218933,
+      "grad_norm": 4.035959243774414,
+      "learning_rate": 3.637066784003757e-07,
+      "loss": 0.9279,
+      "step": 11435
+    },
+    {
+      "epoch": 0.8815911193339501,
+      "grad_norm": 3.918487310409546,
+      "learning_rate": 3.632393775890036e-07,
+      "loss": 0.9514,
+      "step": 11436
+    },
+    {
+      "epoch": 0.881668208448967,
+      "grad_norm": 3.338373899459839,
+      "learning_rate": 3.627723658556426e-07,
+      "loss": 0.7877,
+      "step": 11437
+    },
+    {
+      "epoch": 0.881745297563984,
+      "grad_norm": 3.8025522232055664,
+      "learning_rate": 3.6230564322940754e-07,
+      "loss": 0.9449,
+      "step": 11438
+    },
+    {
+      "epoch": 0.8818223866790009,
+      "grad_norm": 3.8101162910461426,
+      "learning_rate": 3.618392097393958e-07,
+      "loss": 0.9916,
+      "step": 11439
+    },
+    {
+      "epoch": 0.8818994757940178,
+      "grad_norm": 3.876401662826538,
+      "learning_rate": 3.6137306541468797e-07,
+      "loss": 0.8952,
+      "step": 11440
+    },
+    {
+      "epoch": 0.8819765649090349,
+      "grad_norm": 3.733764171600342,
+      "learning_rate": 3.6090721028434527e-07,
+      "loss": 0.8225,
+      "step": 11441
+    },
+    {
+      "epoch": 0.8820536540240518,
+      "grad_norm": 3.6853973865509033,
+      "learning_rate": 3.604416443774117e-07,
+      "loss": 0.8797,
+      "step": 11442
+    },
+    {
+      "epoch": 0.8821307431390688,
+      "grad_norm": 3.8799946308135986,
+      "learning_rate": 3.599763677229134e-07,
+      "loss": 0.9028,
+      "step": 11443
+    },
+    {
+      "epoch": 0.8822078322540857,
+      "grad_norm": 3.8343660831451416,
+      "learning_rate": 3.595113803498557e-07,
+      "loss": 0.9354,
+      "step": 11444
+    },
+    {
+      "epoch": 0.8822849213691026,
+      "grad_norm": 3.665975332260132,
+      "learning_rate": 3.5904668228723074e-07,
+      "loss": 0.9204,
+      "step": 11445
+    },
+    {
+      "epoch": 0.8823620104841197,
+      "grad_norm": 3.6539506912231445,
+      "learning_rate": 3.5858227356400876e-07,
+      "loss": 0.8875,
+      "step": 11446
+    },
+    {
+      "epoch": 0.8824390995991366,
+      "grad_norm": 3.846554756164551,
+      "learning_rate": 3.581181542091444e-07,
+      "loss": 0.875,
+      "step": 11447
+    },
+    {
+      "epoch": 0.8825161887141536,
+      "grad_norm": 3.7280449867248535,
+      "learning_rate": 3.576543242515712e-07,
+      "loss": 0.9288,
+      "step": 11448
+    },
+    {
+      "epoch": 0.8825932778291705,
+      "grad_norm": 3.6032848358154297,
+      "learning_rate": 3.571907837202077e-07,
+      "loss": 0.8565,
+      "step": 11449
+    },
+    {
+      "epoch": 0.8826703669441874,
+      "grad_norm": 3.6691172122955322,
+      "learning_rate": 3.5672753264395345e-07,
+      "loss": 0.8965,
+      "step": 11450
+    },
+    {
+      "epoch": 0.8827474560592045,
+      "grad_norm": 3.7892770767211914,
+      "learning_rate": 3.5626457105168876e-07,
+      "loss": 0.8547,
+      "step": 11451
+    },
+    {
+      "epoch": 0.8828245451742214,
+      "grad_norm": 4.1033477783203125,
+      "learning_rate": 3.558018989722778e-07,
+      "loss": 0.9716,
+      "step": 11452
+    },
+    {
+      "epoch": 0.8829016342892384,
+      "grad_norm": 3.6659538745880127,
+      "learning_rate": 3.5533951643456686e-07,
+      "loss": 0.8052,
+      "step": 11453
+    },
+    {
+      "epoch": 0.8829787234042553,
+      "grad_norm": 4.2055864334106445,
+      "learning_rate": 3.548774234673802e-07,
+      "loss": 0.8538,
+      "step": 11454
+    },
+    {
+      "epoch": 0.8830558125192722,
+      "grad_norm": 3.6786320209503174,
+      "learning_rate": 3.5441562009952856e-07,
+      "loss": 0.9322,
+      "step": 11455
+    },
+    {
+      "epoch": 0.8831329016342893,
+      "grad_norm": 3.6481385231018066,
+      "learning_rate": 3.5395410635980343e-07,
+      "loss": 0.8989,
+      "step": 11456
+    },
+    {
+      "epoch": 0.8832099907493062,
+      "grad_norm": 3.7394192218780518,
+      "learning_rate": 3.5349288227697675e-07,
+      "loss": 0.9274,
+      "step": 11457
+    },
+    {
+      "epoch": 0.8832870798643232,
+      "grad_norm": 3.672476053237915,
+      "learning_rate": 3.5303194787980497e-07,
+      "loss": 0.9608,
+      "step": 11458
+    },
+    {
+      "epoch": 0.8833641689793401,
+      "grad_norm": 3.525527238845825,
+      "learning_rate": 3.5257130319702347e-07,
+      "loss": 0.8911,
+      "step": 11459
+    },
+    {
+      "epoch": 0.883441258094357,
+      "grad_norm": 3.9847187995910645,
+      "learning_rate": 3.5211094825735147e-07,
+      "loss": 0.895,
+      "step": 11460
+    },
+    {
+      "epoch": 0.883518347209374,
+      "grad_norm": 4.699793338775635,
+      "learning_rate": 3.516508830894894e-07,
+      "loss": 0.956,
+      "step": 11461
+    },
+    {
+      "epoch": 0.883595436324391,
+      "grad_norm": 3.901806354522705,
+      "learning_rate": 3.5119110772212084e-07,
+      "loss": 0.9902,
+      "step": 11462
+    },
+    {
+      "epoch": 0.883672525439408,
+      "grad_norm": 4.073935031890869,
+      "learning_rate": 3.5073162218390967e-07,
+      "loss": 0.8234,
+      "step": 11463
+    },
+    {
+      "epoch": 0.8837496145544249,
+      "grad_norm": 3.7054238319396973,
+      "learning_rate": 3.502724265035035e-07,
+      "loss": 0.8809,
+      "step": 11464
+    },
+    {
+      "epoch": 0.8838267036694418,
+      "grad_norm": 3.653473377227783,
+      "learning_rate": 3.498135207095288e-07,
+      "loss": 0.8325,
+      "step": 11465
+    },
+    {
+      "epoch": 0.8839037927844589,
+      "grad_norm": 3.7188761234283447,
+      "learning_rate": 3.4935490483059775e-07,
+      "loss": 0.8455,
+      "step": 11466
+    },
+    {
+      "epoch": 0.8839808818994758,
+      "grad_norm": 4.025871276855469,
+      "learning_rate": 3.4889657889530195e-07,
+      "loss": 0.9839,
+      "step": 11467
+    },
+    {
+      "epoch": 0.8840579710144928,
+      "grad_norm": 3.9531667232513428,
+      "learning_rate": 3.4843854293221515e-07,
+      "loss": 0.9269,
+      "step": 11468
+    },
+    {
+      "epoch": 0.8841350601295097,
+      "grad_norm": 3.4958362579345703,
+      "learning_rate": 3.479807969698956e-07,
+      "loss": 0.8985,
+      "step": 11469
+    },
+    {
+      "epoch": 0.8842121492445266,
+      "grad_norm": 3.8494677543640137,
+      "learning_rate": 3.475233410368789e-07,
+      "loss": 0.9038,
+      "step": 11470
+    },
+    {
+      "epoch": 0.8842892383595437,
+      "grad_norm": 3.6966187953948975,
+      "learning_rate": 3.470661751616866e-07,
+      "loss": 0.816,
+      "step": 11471
+    },
+    {
+      "epoch": 0.8843663274745606,
+      "grad_norm": 3.576230049133301,
+      "learning_rate": 3.466092993728193e-07,
+      "loss": 0.8531,
+      "step": 11472
+    },
+    {
+      "epoch": 0.8844434165895776,
+      "grad_norm": 3.6708030700683594,
+      "learning_rate": 3.461527136987619e-07,
+      "loss": 0.8799,
+      "step": 11473
+    },
+    {
+      "epoch": 0.8845205057045945,
+      "grad_norm": 4.025277614593506,
+      "learning_rate": 3.4569641816798115e-07,
+      "loss": 0.9407,
+      "step": 11474
+    },
+    {
+      "epoch": 0.8845975948196114,
+      "grad_norm": 3.651562452316284,
+      "learning_rate": 3.452404128089221e-07,
+      "loss": 0.8917,
+      "step": 11475
+    },
+    {
+      "epoch": 0.8846746839346284,
+      "grad_norm": 3.617121934890747,
+      "learning_rate": 3.447846976500163e-07,
+      "loss": 0.9336,
+      "step": 11476
+    },
+    {
+      "epoch": 0.8847517730496454,
+      "grad_norm": 3.4860620498657227,
+      "learning_rate": 3.44329272719674e-07,
+      "loss": 0.8401,
+      "step": 11477
+    },
+    {
+      "epoch": 0.8848288621646624,
+      "grad_norm": 4.102723121643066,
+      "learning_rate": 3.4387413804628955e-07,
+      "loss": 1.0335,
+      "step": 11478
+    },
+    {
+      "epoch": 0.8849059512796793,
+      "grad_norm": 3.53051495552063,
+      "learning_rate": 3.434192936582381e-07,
+      "loss": 0.8077,
+      "step": 11479
+    },
+    {
+      "epoch": 0.8849830403946962,
+      "grad_norm": 3.604583978652954,
+      "learning_rate": 3.42964739583877e-07,
+      "loss": 0.977,
+      "step": 11480
+    },
+    {
+      "epoch": 0.8850601295097132,
+      "grad_norm": 3.8658695220947266,
+      "learning_rate": 3.425104758515441e-07,
+      "loss": 0.887,
+      "step": 11481
+    },
+    {
+      "epoch": 0.8851372186247302,
+      "grad_norm": 3.5702638626098633,
+      "learning_rate": 3.420565024895617e-07,
+      "loss": 0.9748,
+      "step": 11482
+    },
+    {
+      "epoch": 0.8852143077397472,
+      "grad_norm": 3.790424108505249,
+      "learning_rate": 3.416028195262322e-07,
+      "loss": 0.9366,
+      "step": 11483
+    },
+    {
+      "epoch": 0.8852913968547641,
+      "grad_norm": 4.021109580993652,
+      "learning_rate": 3.411494269898402e-07,
+      "loss": 1.0176,
+      "step": 11484
+    },
+    {
+      "epoch": 0.885368485969781,
+      "grad_norm": 4.051353454589844,
+      "learning_rate": 3.406963249086537e-07,
+      "loss": 0.8541,
+      "step": 11485
+    },
+    {
+      "epoch": 0.885445575084798,
+      "grad_norm": 3.7868709564208984,
+      "learning_rate": 3.4024351331091945e-07,
+      "loss": 0.8638,
+      "step": 11486
+    },
+    {
+      "epoch": 0.885522664199815,
+      "grad_norm": 3.931445598602295,
+      "learning_rate": 3.3979099222486824e-07,
+      "loss": 0.9284,
+      "step": 11487
+    },
+    {
+      "epoch": 0.885599753314832,
+      "grad_norm": 3.800612449645996,
+      "learning_rate": 3.3933876167871196e-07,
+      "loss": 0.8934,
+      "step": 11488
+    },
+    {
+      "epoch": 0.8856768424298489,
+      "grad_norm": 3.67228627204895,
+      "learning_rate": 3.388868217006469e-07,
+      "loss": 0.8676,
+      "step": 11489
+    },
+    {
+      "epoch": 0.8857539315448658,
+      "grad_norm": 4.013760089874268,
+      "learning_rate": 3.3843517231884894e-07,
+      "loss": 0.8973,
+      "step": 11490
+    },
+    {
+      "epoch": 0.8858310206598828,
+      "grad_norm": 3.6695284843444824,
+      "learning_rate": 3.379838135614738e-07,
+      "loss": 0.781,
+      "step": 11491
+    },
+    {
+      "epoch": 0.8859081097748998,
+      "grad_norm": 3.716569185256958,
+      "learning_rate": 3.375327454566629e-07,
+      "loss": 0.9104,
+      "step": 11492
+    },
+    {
+      "epoch": 0.8859851988899168,
+      "grad_norm": 3.993039608001709,
+      "learning_rate": 3.370819680325377e-07,
+      "loss": 0.8933,
+      "step": 11493
+    },
+    {
+      "epoch": 0.8860622880049337,
+      "grad_norm": 3.30863356590271,
+      "learning_rate": 3.3663148131720223e-07,
+      "loss": 0.799,
+      "step": 11494
+    },
+    {
+      "epoch": 0.8861393771199506,
+      "grad_norm": 3.9631307125091553,
+      "learning_rate": 3.3618128533874196e-07,
+      "loss": 0.8422,
+      "step": 11495
+    },
+    {
+      "epoch": 0.8862164662349676,
+      "grad_norm": 3.697521209716797,
+      "learning_rate": 3.357313801252238e-07,
+      "loss": 0.8291,
+      "step": 11496
+    },
+    {
+      "epoch": 0.8862935553499846,
+      "grad_norm": 4.11827278137207,
+      "learning_rate": 3.3528176570469697e-07,
+      "loss": 0.9668,
+      "step": 11497
+    },
+    {
+      "epoch": 0.8863706444650016,
+      "grad_norm": 4.193023204803467,
+      "learning_rate": 3.348324421051929e-07,
+      "loss": 0.896,
+      "step": 11498
+    },
+    {
+      "epoch": 0.8864477335800185,
+      "grad_norm": 3.8270843029022217,
+      "learning_rate": 3.3438340935472436e-07,
+      "loss": 0.9042,
+      "step": 11499
+    },
+    {
+      "epoch": 0.8865248226950354,
+      "grad_norm": 3.8805384635925293,
+      "learning_rate": 3.3393466748128657e-07,
+      "loss": 0.9534,
+      "step": 11500
+    },
+    {
+      "epoch": 0.8866019118100524,
+      "grad_norm": 3.6519830226898193,
+      "learning_rate": 3.3348621651285663e-07,
+      "loss": 0.8594,
+      "step": 11501
+    },
+    {
+      "epoch": 0.8866790009250693,
+      "grad_norm": 3.8050684928894043,
+      "learning_rate": 3.330380564773922e-07,
+      "loss": 0.8864,
+      "step": 11502
+    },
+    {
+      "epoch": 0.8867560900400864,
+      "grad_norm": 4.044971942901611,
+      "learning_rate": 3.3259018740283366e-07,
+      "loss": 0.9039,
+      "step": 11503
+    },
+    {
+      "epoch": 0.8868331791551033,
+      "grad_norm": 3.61731219291687,
+      "learning_rate": 3.321426093171043e-07,
+      "loss": 0.8472,
+      "step": 11504
+    },
+    {
+      "epoch": 0.8869102682701202,
+      "grad_norm": 3.757559061050415,
+      "learning_rate": 3.316953222481073e-07,
+      "loss": 0.9608,
+      "step": 11505
+    },
+    {
+      "epoch": 0.8869873573851372,
+      "grad_norm": 4.073364734649658,
+      "learning_rate": 3.312483262237304e-07,
+      "loss": 0.97,
+      "step": 11506
+    },
+    {
+      "epoch": 0.8870644465001541,
+      "grad_norm": 3.8168318271636963,
+      "learning_rate": 3.30801621271839e-07,
+      "loss": 1.0477,
+      "step": 11507
+    },
+    {
+      "epoch": 0.8871415356151712,
+      "grad_norm": 3.800417184829712,
+      "learning_rate": 3.303552074202848e-07,
+      "loss": 0.9133,
+      "step": 11508
+    },
+    {
+      "epoch": 0.8872186247301881,
+      "grad_norm": 3.7072558403015137,
+      "learning_rate": 3.299090846968983e-07,
+      "loss": 0.8998,
+      "step": 11509
+    },
+    {
+      "epoch": 0.887295713845205,
+      "grad_norm": 3.8859128952026367,
+      "learning_rate": 3.294632531294933e-07,
+      "loss": 0.8885,
+      "step": 11510
+    },
+    {
+      "epoch": 0.887372802960222,
+      "grad_norm": 3.608797311782837,
+      "learning_rate": 3.290177127458655e-07,
+      "loss": 0.9306,
+      "step": 11511
+    },
+    {
+      "epoch": 0.8874498920752389,
+      "grad_norm": 3.846325635910034,
+      "learning_rate": 3.285724635737919e-07,
+      "loss": 0.8752,
+      "step": 11512
+    },
+    {
+      "epoch": 0.887526981190256,
+      "grad_norm": 3.467984199523926,
+      "learning_rate": 3.281275056410305e-07,
+      "loss": 0.7637,
+      "step": 11513
+    },
+    {
+      "epoch": 0.8876040703052729,
+      "grad_norm": 3.4977262020111084,
+      "learning_rate": 3.276828389753234e-07,
+      "loss": 0.8379,
+      "step": 11514
+    },
+    {
+      "epoch": 0.8876811594202898,
+      "grad_norm": 3.824713945388794,
+      "learning_rate": 3.2723846360439236e-07,
+      "loss": 0.8603,
+      "step": 11515
+    },
+    {
+      "epoch": 0.8877582485353068,
+      "grad_norm": 3.6566162109375,
+      "learning_rate": 3.267943795559425e-07,
+      "loss": 0.9688,
+      "step": 11516
+    },
+    {
+      "epoch": 0.8878353376503237,
+      "grad_norm": 3.705967664718628,
+      "learning_rate": 3.263505868576605e-07,
+      "loss": 0.9366,
+      "step": 11517
+    },
+    {
+      "epoch": 0.8879124267653408,
+      "grad_norm": 3.7640185356140137,
+      "learning_rate": 3.259070855372132e-07,
+      "loss": 0.8421,
+      "step": 11518
+    },
+    {
+      "epoch": 0.8879895158803577,
+      "grad_norm": 3.6393392086029053,
+      "learning_rate": 3.2546387562225166e-07,
+      "loss": 0.9986,
+      "step": 11519
+    },
+    {
+      "epoch": 0.8880666049953746,
+      "grad_norm": 3.8339908123016357,
+      "learning_rate": 3.2502095714040673e-07,
+      "loss": 0.8996,
+      "step": 11520
+    },
+    {
+      "epoch": 0.8881436941103916,
+      "grad_norm": 3.783637285232544,
+      "learning_rate": 3.245783301192934e-07,
+      "loss": 0.9308,
+      "step": 11521
+    },
+    {
+      "epoch": 0.8882207832254085,
+      "grad_norm": 3.7414445877075195,
+      "learning_rate": 3.2413599458650744e-07,
+      "loss": 0.8207,
+      "step": 11522
+    },
+    {
+      "epoch": 0.8882978723404256,
+      "grad_norm": 3.6002695560455322,
+      "learning_rate": 3.2369395056962404e-07,
+      "loss": 0.8007,
+      "step": 11523
+    },
+    {
+      "epoch": 0.8883749614554425,
+      "grad_norm": 3.9991655349731445,
+      "learning_rate": 3.232521980962039e-07,
+      "loss": 0.8721,
+      "step": 11524
+    },
+    {
+      "epoch": 0.8884520505704595,
+      "grad_norm": 3.785975933074951,
+      "learning_rate": 3.2281073719378773e-07,
+      "loss": 0.9526,
+      "step": 11525
+    },
+    {
+      "epoch": 0.8885291396854764,
+      "grad_norm": 3.625488758087158,
+      "learning_rate": 3.22369567889898e-07,
+      "loss": 0.8719,
+      "step": 11526
+    },
+    {
+      "epoch": 0.8886062288004933,
+      "grad_norm": 3.615448236465454,
+      "learning_rate": 3.2192869021203997e-07,
+      "loss": 0.8264,
+      "step": 11527
+    },
+    {
+      "epoch": 0.8886833179155104,
+      "grad_norm": 3.5204501152038574,
+      "learning_rate": 3.214881041877005e-07,
+      "loss": 0.8942,
+      "step": 11528
+    },
+    {
+      "epoch": 0.8887604070305273,
+      "grad_norm": 3.519294261932373,
+      "learning_rate": 3.210478098443459e-07,
+      "loss": 0.7734,
+      "step": 11529
+    },
+    {
+      "epoch": 0.8888374961455443,
+      "grad_norm": 3.4945321083068848,
+      "learning_rate": 3.206078072094276e-07,
+      "loss": 0.8511,
+      "step": 11530
+    },
+    {
+      "epoch": 0.8889145852605612,
+      "grad_norm": 3.7900755405426025,
+      "learning_rate": 3.201680963103776e-07,
+      "loss": 0.8575,
+      "step": 11531
+    },
+    {
+      "epoch": 0.8889916743755781,
+      "grad_norm": 3.660367012023926,
+      "learning_rate": 3.197286771746094e-07,
+      "loss": 0.91,
+      "step": 11532
+    },
+    {
+      "epoch": 0.8890687634905952,
+      "grad_norm": 3.8760204315185547,
+      "learning_rate": 3.192895498295179e-07,
+      "loss": 0.8206,
+      "step": 11533
+    },
+    {
+      "epoch": 0.8891458526056121,
+      "grad_norm": 3.871697425842285,
+      "learning_rate": 3.188507143024816e-07,
+      "loss": 0.9093,
+      "step": 11534
+    },
+    {
+      "epoch": 0.8892229417206291,
+      "grad_norm": 3.5889370441436768,
+      "learning_rate": 3.184121706208587e-07,
+      "loss": 0.8424,
+      "step": 11535
+    },
+    {
+      "epoch": 0.889300030835646,
+      "grad_norm": 3.674396514892578,
+      "learning_rate": 3.1797391881199014e-07,
+      "loss": 0.9136,
+      "step": 11536
+    },
+    {
+      "epoch": 0.8893771199506629,
+      "grad_norm": 3.540079116821289,
+      "learning_rate": 3.17535958903199e-07,
+      "loss": 0.8915,
+      "step": 11537
+    },
+    {
+      "epoch": 0.88945420906568,
+      "grad_norm": 3.643944501876831,
+      "learning_rate": 3.170982909217907e-07,
+      "loss": 0.9742,
+      "step": 11538
+    },
+    {
+      "epoch": 0.8895312981806969,
+      "grad_norm": 3.8346221446990967,
+      "learning_rate": 3.1666091489505004e-07,
+      "loss": 0.9745,
+      "step": 11539
+    },
+    {
+      "epoch": 0.8896083872957139,
+      "grad_norm": 3.8320624828338623,
+      "learning_rate": 3.162238308502452e-07,
+      "loss": 0.8823,
+      "step": 11540
+    },
+    {
+      "epoch": 0.8896854764107308,
+      "grad_norm": 3.508303642272949,
+      "learning_rate": 3.157870388146267e-07,
+      "loss": 0.8418,
+      "step": 11541
+    },
+    {
+      "epoch": 0.8897625655257477,
+      "grad_norm": 3.801394462585449,
+      "learning_rate": 3.1535053881542657e-07,
+      "loss": 0.9079,
+      "step": 11542
+    },
+    {
+      "epoch": 0.8898396546407648,
+      "grad_norm": 3.286583662033081,
+      "learning_rate": 3.14914330879858e-07,
+      "loss": 0.7761,
+      "step": 11543
+    },
+    {
+      "epoch": 0.8899167437557817,
+      "grad_norm": 4.076812267303467,
+      "learning_rate": 3.14478415035116e-07,
+      "loss": 1.0037,
+      "step": 11544
+    },
+    {
+      "epoch": 0.8899938328707987,
+      "grad_norm": 3.5942256450653076,
+      "learning_rate": 3.140427913083777e-07,
+      "loss": 0.8007,
+      "step": 11545
+    },
+    {
+      "epoch": 0.8900709219858156,
+      "grad_norm": 3.80216121673584,
+      "learning_rate": 3.136074597268024e-07,
+      "loss": 0.9157,
+      "step": 11546
+    },
+    {
+      "epoch": 0.8901480111008325,
+      "grad_norm": 4.44173002243042,
+      "learning_rate": 3.1317242031753013e-07,
+      "loss": 1.0875,
+      "step": 11547
+    },
+    {
+      "epoch": 0.8902251002158496,
+      "grad_norm": 3.757503032684326,
+      "learning_rate": 3.127376731076837e-07,
+      "loss": 0.9564,
+      "step": 11548
+    },
+    {
+      "epoch": 0.8903021893308665,
+      "grad_norm": 3.6914825439453125,
+      "learning_rate": 3.1230321812436846e-07,
+      "loss": 0.8826,
+      "step": 11549
+    },
+    {
+      "epoch": 0.8903792784458835,
+      "grad_norm": 3.583136796951294,
+      "learning_rate": 3.118690553946685e-07,
+      "loss": 0.8588,
+      "step": 11550
+    },
+    {
+      "epoch": 0.8904563675609004,
+      "grad_norm": 3.7965340614318848,
+      "learning_rate": 3.114351849456526e-07,
+      "loss": 0.9515,
+      "step": 11551
+    },
+    {
+      "epoch": 0.8905334566759173,
+      "grad_norm": 3.5539650917053223,
+      "learning_rate": 3.110016068043703e-07,
+      "loss": 0.9266,
+      "step": 11552
+    },
+    {
+      "epoch": 0.8906105457909343,
+      "grad_norm": 3.8235104084014893,
+      "learning_rate": 3.105683209978527e-07,
+      "loss": 0.9676,
+      "step": 11553
+    },
+    {
+      "epoch": 0.8906876349059513,
+      "grad_norm": 3.7492964267730713,
+      "learning_rate": 3.101353275531138e-07,
+      "loss": 0.8673,
+      "step": 11554
+    },
+    {
+      "epoch": 0.8907647240209683,
+      "grad_norm": 3.6325714588165283,
+      "learning_rate": 3.09702626497147e-07,
+      "loss": 0.8821,
+      "step": 11555
+    },
+    {
+      "epoch": 0.8908418131359852,
+      "grad_norm": 3.75046443939209,
+      "learning_rate": 3.092702178569301e-07,
+      "loss": 0.9256,
+      "step": 11556
+    },
+    {
+      "epoch": 0.8909189022510021,
+      "grad_norm": 3.687927007675171,
+      "learning_rate": 3.088381016594211e-07,
+      "loss": 0.8995,
+      "step": 11557
+    },
+    {
+      "epoch": 0.8909959913660191,
+      "grad_norm": 3.565046787261963,
+      "learning_rate": 3.0840627793156e-07,
+      "loss": 0.8913,
+      "step": 11558
+    },
+    {
+      "epoch": 0.8910730804810361,
+      "grad_norm": 3.7868244647979736,
+      "learning_rate": 3.0797474670026917e-07,
+      "loss": 0.9228,
+      "step": 11559
+    },
+    {
+      "epoch": 0.8911501695960531,
+      "grad_norm": 3.6656594276428223,
+      "learning_rate": 3.0754350799245323e-07,
+      "loss": 0.9629,
+      "step": 11560
+    },
+    {
+      "epoch": 0.89122725871107,
+      "grad_norm": 3.7980310916900635,
+      "learning_rate": 3.0711256183499574e-07,
+      "loss": 0.9249,
+      "step": 11561
+    },
+    {
+      "epoch": 0.8913043478260869,
+      "grad_norm": 3.5572173595428467,
+      "learning_rate": 3.066819082547651e-07,
+      "loss": 0.7583,
+      "step": 11562
+    },
+    {
+      "epoch": 0.8913814369411039,
+      "grad_norm": 3.639159679412842,
+      "learning_rate": 3.062515472786098e-07,
+      "loss": 1.0125,
+      "step": 11563
+    },
+    {
+      "epoch": 0.8914585260561209,
+      "grad_norm": 3.9139461517333984,
+      "learning_rate": 3.0582147893336136e-07,
+      "loss": 0.994,
+      "step": 11564
+    },
+    {
+      "epoch": 0.8915356151711379,
+      "grad_norm": 3.53010630607605,
+      "learning_rate": 3.0539170324583257e-07,
+      "loss": 0.8378,
+      "step": 11565
+    },
+    {
+      "epoch": 0.8916127042861548,
+      "grad_norm": 3.5645971298217773,
+      "learning_rate": 3.049622202428165e-07,
+      "loss": 0.8547,
+      "step": 11566
+    },
+    {
+      "epoch": 0.8916897934011717,
+      "grad_norm": 3.8837499618530273,
+      "learning_rate": 3.045330299510896e-07,
+      "loss": 0.9795,
+      "step": 11567
+    },
+    {
+      "epoch": 0.8917668825161887,
+      "grad_norm": 4.236751079559326,
+      "learning_rate": 3.041041323974098e-07,
+      "loss": 0.9627,
+      "step": 11568
+    },
+    {
+      "epoch": 0.8918439716312057,
+      "grad_norm": 4.3294477462768555,
+      "learning_rate": 3.0367552760851684e-07,
+      "loss": 0.8706,
+      "step": 11569
+    },
+    {
+      "epoch": 0.8919210607462227,
+      "grad_norm": 3.358781337738037,
+      "learning_rate": 3.0324721561113213e-07,
+      "loss": 0.8451,
+      "step": 11570
+    },
+    {
+      "epoch": 0.8919981498612396,
+      "grad_norm": 3.78355073928833,
+      "learning_rate": 3.028191964319582e-07,
+      "loss": 0.771,
+      "step": 11571
+    },
+    {
+      "epoch": 0.8920752389762565,
+      "grad_norm": 3.7563531398773193,
+      "learning_rate": 3.023914700976799e-07,
+      "loss": 0.8986,
+      "step": 11572
+    },
+    {
+      "epoch": 0.8921523280912735,
+      "grad_norm": 3.557497024536133,
+      "learning_rate": 3.019640366349641e-07,
+      "loss": 0.9661,
+      "step": 11573
+    },
+    {
+      "epoch": 0.8922294172062905,
+      "grad_norm": 3.693223476409912,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.8886,
+      "step": 11574
+    },
+    {
+      "epoch": 0.8923065063213075,
+      "grad_norm": 3.728806972503662,
+      "learning_rate": 3.0111004843079327e-07,
+      "loss": 0.8569,
+      "step": 11575
+    },
+    {
+      "epoch": 0.8923835954363244,
+      "grad_norm": 3.8139472007751465,
+      "learning_rate": 3.0068349374258175e-07,
+      "loss": 0.8939,
+      "step": 11576
+    },
+    {
+      "epoch": 0.8924606845513413,
+      "grad_norm": 3.815910577774048,
+      "learning_rate": 3.002572320324143e-07,
+      "loss": 0.8666,
+      "step": 11577
+    },
+    {
+      "epoch": 0.8925377736663583,
+      "grad_norm": 3.8948774337768555,
+      "learning_rate": 2.99831263326868e-07,
+      "loss": 0.9402,
+      "step": 11578
+    },
+    {
+      "epoch": 0.8926148627813753,
+      "grad_norm": 4.020652770996094,
+      "learning_rate": 2.994055876525004e-07,
+      "loss": 1.011,
+      "step": 11579
+    },
+    {
+      "epoch": 0.8926919518963923,
+      "grad_norm": 3.6577086448669434,
+      "learning_rate": 2.9898020503584977e-07,
+      "loss": 0.9003,
+      "step": 11580
+    },
+    {
+      "epoch": 0.8927690410114092,
+      "grad_norm": 3.762552499771118,
+      "learning_rate": 2.985551155034355e-07,
+      "loss": 0.9369,
+      "step": 11581
+    },
+    {
+      "epoch": 0.8928461301264261,
+      "grad_norm": 3.9927804470062256,
+      "learning_rate": 2.9813031908176025e-07,
+      "loss": 0.8735,
+      "step": 11582
+    },
+    {
+      "epoch": 0.8929232192414431,
+      "grad_norm": 3.5767393112182617,
+      "learning_rate": 2.977058157973084e-07,
+      "loss": 0.9675,
+      "step": 11583
+    },
+    {
+      "epoch": 0.89300030835646,
+      "grad_norm": 3.9472174644470215,
+      "learning_rate": 2.9728160567654483e-07,
+      "loss": 0.9394,
+      "step": 11584
+    },
+    {
+      "epoch": 0.8930773974714771,
+      "grad_norm": 3.4129271507263184,
+      "learning_rate": 2.968576887459168e-07,
+      "loss": 0.7976,
+      "step": 11585
+    },
+    {
+      "epoch": 0.893154486586494,
+      "grad_norm": 3.768298864364624,
+      "learning_rate": 2.964340650318548e-07,
+      "loss": 0.9286,
+      "step": 11586
+    },
+    {
+      "epoch": 0.8932315757015109,
+      "grad_norm": 3.6881837844848633,
+      "learning_rate": 2.9601073456076766e-07,
+      "loss": 0.8934,
+      "step": 11587
+    },
+    {
+      "epoch": 0.8933086648165279,
+      "grad_norm": 3.6824235916137695,
+      "learning_rate": 2.955876973590488e-07,
+      "loss": 0.9171,
+      "step": 11588
+    },
+    {
+      "epoch": 0.8933857539315448,
+      "grad_norm": 3.3715460300445557,
+      "learning_rate": 2.9516495345307207e-07,
+      "loss": 0.8354,
+      "step": 11589
+    },
+    {
+      "epoch": 0.8934628430465619,
+      "grad_norm": 3.638155698776245,
+      "learning_rate": 2.947425028691936e-07,
+      "loss": 0.8242,
+      "step": 11590
+    },
+    {
+      "epoch": 0.8935399321615788,
+      "grad_norm": 3.8177831172943115,
+      "learning_rate": 2.943203456337512e-07,
+      "loss": 1.0496,
+      "step": 11591
+    },
+    {
+      "epoch": 0.8936170212765957,
+      "grad_norm": 3.5661537647247314,
+      "learning_rate": 2.938984817730639e-07,
+      "loss": 0.8795,
+      "step": 11592
+    },
+    {
+      "epoch": 0.8936941103916127,
+      "grad_norm": 3.6265878677368164,
+      "learning_rate": 2.9347691131343223e-07,
+      "loss": 0.95,
+      "step": 11593
+    },
+    {
+      "epoch": 0.8937711995066296,
+      "grad_norm": 3.493656635284424,
+      "learning_rate": 2.930556342811397e-07,
+      "loss": 0.8027,
+      "step": 11594
+    },
+    {
+      "epoch": 0.8938482886216467,
+      "grad_norm": 3.3560476303100586,
+      "learning_rate": 2.926346507024502e-07,
+      "loss": 0.8137,
+      "step": 11595
+    },
+    {
+      "epoch": 0.8939253777366636,
+      "grad_norm": 3.855767011642456,
+      "learning_rate": 2.922139606036106e-07,
+      "loss": 0.9452,
+      "step": 11596
+    },
+    {
+      "epoch": 0.8940024668516805,
+      "grad_norm": 4.110727787017822,
+      "learning_rate": 2.917935640108488e-07,
+      "loss": 1.0172,
+      "step": 11597
+    },
+    {
+      "epoch": 0.8940795559666975,
+      "grad_norm": 3.6166462898254395,
+      "learning_rate": 2.9137346095037324e-07,
+      "loss": 0.9122,
+      "step": 11598
+    },
+    {
+      "epoch": 0.8941566450817144,
+      "grad_norm": 3.6462063789367676,
+      "learning_rate": 2.909536514483752e-07,
+      "loss": 0.831,
+      "step": 11599
+    },
+    {
+      "epoch": 0.8942337341967315,
+      "grad_norm": 3.6715078353881836,
+      "learning_rate": 2.9053413553102874e-07,
+      "loss": 0.8714,
+      "step": 11600
+    },
+    {
+      "epoch": 0.8943108233117484,
+      "grad_norm": 3.616772174835205,
+      "learning_rate": 2.901149132244879e-07,
+      "loss": 1.0388,
+      "step": 11601
+    },
+    {
+      "epoch": 0.8943879124267653,
+      "grad_norm": 4.153509616851807,
+      "learning_rate": 2.896959845548902e-07,
+      "loss": 0.8361,
+      "step": 11602
+    },
+    {
+      "epoch": 0.8944650015417823,
+      "grad_norm": 3.7695071697235107,
+      "learning_rate": 2.892773495483514e-07,
+      "loss": 0.9392,
+      "step": 11603
+    },
+    {
+      "epoch": 0.8945420906567992,
+      "grad_norm": 3.27496600151062,
+      "learning_rate": 2.8885900823097223e-07,
+      "loss": 0.8339,
+      "step": 11604
+    },
+    {
+      "epoch": 0.8946191797718163,
+      "grad_norm": 4.047286510467529,
+      "learning_rate": 2.8844096062883466e-07,
+      "loss": 0.9914,
+      "step": 11605
+    },
+    {
+      "epoch": 0.8946962688868332,
+      "grad_norm": 3.8037588596343994,
+      "learning_rate": 2.880232067680011e-07,
+      "loss": 0.8849,
+      "step": 11606
+    },
+    {
+      "epoch": 0.8947733580018501,
+      "grad_norm": 3.4862048625946045,
+      "learning_rate": 2.876057466745169e-07,
+      "loss": 0.8648,
+      "step": 11607
+    },
+    {
+      "epoch": 0.8948504471168671,
+      "grad_norm": 3.446492910385132,
+      "learning_rate": 2.87188580374409e-07,
+      "loss": 0.8775,
+      "step": 11608
+    },
+    {
+      "epoch": 0.894927536231884,
+      "grad_norm": 3.7354066371917725,
+      "learning_rate": 2.867717078936838e-07,
+      "loss": 0.9698,
+      "step": 11609
+    },
+    {
+      "epoch": 0.8950046253469011,
+      "grad_norm": 4.134057998657227,
+      "learning_rate": 2.863551292583322e-07,
+      "loss": 0.9169,
+      "step": 11610
+    },
+    {
+      "epoch": 0.895081714461918,
+      "grad_norm": 3.7595643997192383,
+      "learning_rate": 2.8593884449432617e-07,
+      "loss": 0.9883,
+      "step": 11611
+    },
+    {
+      "epoch": 0.8951588035769349,
+      "grad_norm": 3.875776529312134,
+      "learning_rate": 2.8552285362761833e-07,
+      "loss": 0.985,
+      "step": 11612
+    },
+    {
+      "epoch": 0.8952358926919519,
+      "grad_norm": 3.6377480030059814,
+      "learning_rate": 2.85107156684144e-07,
+      "loss": 0.8584,
+      "step": 11613
+    },
+    {
+      "epoch": 0.8953129818069688,
+      "grad_norm": 3.419172763824463,
+      "learning_rate": 2.846917536898192e-07,
+      "loss": 0.8829,
+      "step": 11614
+    },
+    {
+      "epoch": 0.8953900709219859,
+      "grad_norm": 3.9442739486694336,
+      "learning_rate": 2.8427664467054194e-07,
+      "loss": 0.9979,
+      "step": 11615
+    },
+    {
+      "epoch": 0.8954671600370028,
+      "grad_norm": 3.6538686752319336,
+      "learning_rate": 2.838618296521922e-07,
+      "loss": 0.9496,
+      "step": 11616
+    },
+    {
+      "epoch": 0.8955442491520197,
+      "grad_norm": 3.4656448364257812,
+      "learning_rate": 2.8344730866063264e-07,
+      "loss": 0.7275,
+      "step": 11617
+    },
+    {
+      "epoch": 0.8956213382670367,
+      "grad_norm": 3.758117437362671,
+      "learning_rate": 2.8303308172170587e-07,
+      "loss": 0.8674,
+      "step": 11618
+    },
+    {
+      "epoch": 0.8956984273820536,
+      "grad_norm": 3.560523271560669,
+      "learning_rate": 2.826191488612362e-07,
+      "loss": 0.8544,
+      "step": 11619
+    },
+    {
+      "epoch": 0.8957755164970707,
+      "grad_norm": 3.549065351486206,
+      "learning_rate": 2.822055101050308e-07,
+      "loss": 0.8741,
+      "step": 11620
+    },
+    {
+      "epoch": 0.8958526056120876,
+      "grad_norm": 3.7114226818084717,
+      "learning_rate": 2.817921654788769e-07,
+      "loss": 0.9257,
+      "step": 11621
+    },
+    {
+      "epoch": 0.8959296947271045,
+      "grad_norm": 4.023944854736328,
+      "learning_rate": 2.813791150085454e-07,
+      "loss": 1.1294,
+      "step": 11622
+    },
+    {
+      "epoch": 0.8960067838421215,
+      "grad_norm": 3.9397988319396973,
+      "learning_rate": 2.8096635871978915e-07,
+      "loss": 0.906,
+      "step": 11623
+    },
+    {
+      "epoch": 0.8960838729571384,
+      "grad_norm": 3.5151617527008057,
+      "learning_rate": 2.8055389663833923e-07,
+      "loss": 0.9842,
+      "step": 11624
+    },
+    {
+      "epoch": 0.8961609620721555,
+      "grad_norm": 3.4286890029907227,
+      "learning_rate": 2.801417287899111e-07,
+      "loss": 0.9019,
+      "step": 11625
+    },
+    {
+      "epoch": 0.8962380511871724,
+      "grad_norm": 3.755769968032837,
+      "learning_rate": 2.7972985520020093e-07,
+      "loss": 0.9232,
+      "step": 11626
+    },
+    {
+      "epoch": 0.8963151403021893,
+      "grad_norm": 3.760594606399536,
+      "learning_rate": 2.793182758948881e-07,
+      "loss": 0.8456,
+      "step": 11627
+    },
+    {
+      "epoch": 0.8963922294172063,
+      "grad_norm": 3.722588300704956,
+      "learning_rate": 2.7890699089963225e-07,
+      "loss": 0.923,
+      "step": 11628
+    },
+    {
+      "epoch": 0.8964693185322232,
+      "grad_norm": 3.614448070526123,
+      "learning_rate": 2.784960002400733e-07,
+      "loss": 1.0738,
+      "step": 11629
+    },
+    {
+      "epoch": 0.8965464076472403,
+      "grad_norm": 3.7189695835113525,
+      "learning_rate": 2.7808530394183577e-07,
+      "loss": 0.9993,
+      "step": 11630
+    },
+    {
+      "epoch": 0.8966234967622572,
+      "grad_norm": 3.6947999000549316,
+      "learning_rate": 2.776749020305236e-07,
+      "loss": 0.8291,
+      "step": 11631
+    },
+    {
+      "epoch": 0.8967005858772741,
+      "grad_norm": 3.6183085441589355,
+      "learning_rate": 2.7726479453172415e-07,
+      "loss": 0.7955,
+      "step": 11632
+    },
+    {
+      "epoch": 0.8967776749922911,
+      "grad_norm": 3.457646608352661,
+      "learning_rate": 2.7685498147100533e-07,
+      "loss": 0.8498,
+      "step": 11633
+    },
+    {
+      "epoch": 0.896854764107308,
+      "grad_norm": 3.7649595737457275,
+      "learning_rate": 2.7644546287391716e-07,
+      "loss": 0.8975,
+      "step": 11634
+    },
+    {
+      "epoch": 0.896931853222325,
+      "grad_norm": 3.469334840774536,
+      "learning_rate": 2.760362387659893e-07,
+      "loss": 0.8225,
+      "step": 11635
+    },
+    {
+      "epoch": 0.897008942337342,
+      "grad_norm": 3.689354181289673,
+      "learning_rate": 2.756273091727363e-07,
+      "loss": 0.9019,
+      "step": 11636
+    },
+    {
+      "epoch": 0.8970860314523589,
+      "grad_norm": 3.99678635597229,
+      "learning_rate": 2.7521867411965273e-07,
+      "loss": 0.9491,
+      "step": 11637
+    },
+    {
+      "epoch": 0.8971631205673759,
+      "grad_norm": 3.96669602394104,
+      "learning_rate": 2.7481033363221385e-07,
+      "loss": 0.8561,
+      "step": 11638
+    },
+    {
+      "epoch": 0.8972402096823928,
+      "grad_norm": 3.8000166416168213,
+      "learning_rate": 2.744022877358793e-07,
+      "loss": 0.9234,
+      "step": 11639
+    },
+    {
+      "epoch": 0.8973172987974098,
+      "grad_norm": 3.587712287902832,
+      "learning_rate": 2.73994536456087e-07,
+      "loss": 0.905,
+      "step": 11640
+    },
+    {
+      "epoch": 0.8973943879124268,
+      "grad_norm": 3.3833603858947754,
+      "learning_rate": 2.735870798182588e-07,
+      "loss": 0.8504,
+      "step": 11641
+    },
+    {
+      "epoch": 0.8974714770274437,
+      "grad_norm": 3.685027837753296,
+      "learning_rate": 2.7317991784779727e-07,
+      "loss": 0.9113,
+      "step": 11642
+    },
+    {
+      "epoch": 0.8975485661424607,
+      "grad_norm": 3.7003281116485596,
+      "learning_rate": 2.727730505700871e-07,
+      "loss": 0.8788,
+      "step": 11643
+    },
+    {
+      "epoch": 0.8976256552574776,
+      "grad_norm": 3.5367331504821777,
+      "learning_rate": 2.7236647801049456e-07,
+      "loss": 0.879,
+      "step": 11644
+    },
+    {
+      "epoch": 0.8977027443724946,
+      "grad_norm": 4.164783000946045,
+      "learning_rate": 2.719602001943672e-07,
+      "loss": 1.0591,
+      "step": 11645
+    },
+    {
+      "epoch": 0.8977798334875116,
+      "grad_norm": 3.9324138164520264,
+      "learning_rate": 2.7155421714703424e-07,
+      "loss": 0.9406,
+      "step": 11646
+    },
+    {
+      "epoch": 0.8978569226025285,
+      "grad_norm": 4.007849216461182,
+      "learning_rate": 2.7114852889380594e-07,
+      "loss": 0.9425,
+      "step": 11647
+    },
+    {
+      "epoch": 0.8979340117175455,
+      "grad_norm": 3.8484272956848145,
+      "learning_rate": 2.7074313545997546e-07,
+      "loss": 0.9126,
+      "step": 11648
+    },
+    {
+      "epoch": 0.8980111008325624,
+      "grad_norm": 4.0231804847717285,
+      "learning_rate": 2.703380368708175e-07,
+      "loss": 0.8856,
+      "step": 11649
+    },
+    {
+      "epoch": 0.8980881899475794,
+      "grad_norm": 3.677838087081909,
+      "learning_rate": 2.6993323315158804e-07,
+      "loss": 0.9547,
+      "step": 11650
+    },
+    {
+      "epoch": 0.8981652790625964,
+      "grad_norm": 3.767169237136841,
+      "learning_rate": 2.6952872432752295e-07,
+      "loss": 0.9004,
+      "step": 11651
+    },
+    {
+      "epoch": 0.8982423681776133,
+      "grad_norm": 3.888305187225342,
+      "learning_rate": 2.691245104238421e-07,
+      "loss": 0.9836,
+      "step": 11652
+    },
+    {
+      "epoch": 0.8983194572926303,
+      "grad_norm": 4.260769367218018,
+      "learning_rate": 2.687205914657465e-07,
+      "loss": 0.9681,
+      "step": 11653
+    },
+    {
+      "epoch": 0.8983965464076472,
+      "grad_norm": 3.9282066822052,
+      "learning_rate": 2.6831696747841804e-07,
+      "loss": 0.8736,
+      "step": 11654
+    },
+    {
+      "epoch": 0.8984736355226642,
+      "grad_norm": 4.230199337005615,
+      "learning_rate": 2.679136384870201e-07,
+      "loss": 1.0049,
+      "step": 11655
+    },
+    {
+      "epoch": 0.8985507246376812,
+      "grad_norm": 3.359044075012207,
+      "learning_rate": 2.6751060451670033e-07,
+      "loss": 0.7274,
+      "step": 11656
+    },
+    {
+      "epoch": 0.8986278137526981,
+      "grad_norm": 3.813596487045288,
+      "learning_rate": 2.671078655925829e-07,
+      "loss": 0.8514,
+      "step": 11657
+    },
+    {
+      "epoch": 0.8987049028677151,
+      "grad_norm": 3.9940125942230225,
+      "learning_rate": 2.6670542173977745e-07,
+      "loss": 0.9324,
+      "step": 11658
+    },
+    {
+      "epoch": 0.898781991982732,
+      "grad_norm": 3.4712913036346436,
+      "learning_rate": 2.6630327298337535e-07,
+      "loss": 0.8621,
+      "step": 11659
+    },
+    {
+      "epoch": 0.898859081097749,
+      "grad_norm": 3.4401824474334717,
+      "learning_rate": 2.6590141934844715e-07,
+      "loss": 0.7482,
+      "step": 11660
+    },
+    {
+      "epoch": 0.898936170212766,
+      "grad_norm": 4.060207366943359,
+      "learning_rate": 2.654998608600479e-07,
+      "loss": 1.0239,
+      "step": 11661
+    },
+    {
+      "epoch": 0.8990132593277829,
+      "grad_norm": 3.8618593215942383,
+      "learning_rate": 2.6509859754321076e-07,
+      "loss": 0.9923,
+      "step": 11662
+    },
+    {
+      "epoch": 0.8990903484427999,
+      "grad_norm": 3.749800682067871,
+      "learning_rate": 2.6469762942295363e-07,
+      "loss": 0.9303,
+      "step": 11663
+    },
+    {
+      "epoch": 0.8991674375578168,
+      "grad_norm": 3.6844890117645264,
+      "learning_rate": 2.642969565242748e-07,
+      "loss": 0.9534,
+      "step": 11664
+    },
+    {
+      "epoch": 0.8992445266728338,
+      "grad_norm": 3.837385892868042,
+      "learning_rate": 2.6389657887215314e-07,
+      "loss": 0.9022,
+      "step": 11665
+    },
+    {
+      "epoch": 0.8993216157878507,
+      "grad_norm": 3.7096149921417236,
+      "learning_rate": 2.6349649649155153e-07,
+      "loss": 0.9856,
+      "step": 11666
+    },
+    {
+      "epoch": 0.8993987049028677,
+      "grad_norm": 4.130624771118164,
+      "learning_rate": 2.6309670940741215e-07,
+      "loss": 0.9756,
+      "step": 11667
+    },
+    {
+      "epoch": 0.8994757940178847,
+      "grad_norm": 4.128146171569824,
+      "learning_rate": 2.6269721764466016e-07,
+      "loss": 0.9506,
+      "step": 11668
+    },
+    {
+      "epoch": 0.8995528831329016,
+      "grad_norm": 3.8928797245025635,
+      "learning_rate": 2.6229802122820114e-07,
+      "loss": 0.9731,
+      "step": 11669
+    },
+    {
+      "epoch": 0.8996299722479186,
+      "grad_norm": 4.215887069702148,
+      "learning_rate": 2.618991201829235e-07,
+      "loss": 0.937,
+      "step": 11670
+    },
+    {
+      "epoch": 0.8997070613629355,
+      "grad_norm": 3.51582407951355,
+      "learning_rate": 2.6150051453369684e-07,
+      "loss": 0.8767,
+      "step": 11671
+    },
+    {
+      "epoch": 0.8997841504779525,
+      "grad_norm": 3.87080979347229,
+      "learning_rate": 2.6110220430537124e-07,
+      "loss": 0.866,
+      "step": 11672
+    },
+    {
+      "epoch": 0.8998612395929695,
+      "grad_norm": 3.93489670753479,
+      "learning_rate": 2.607041895227802e-07,
+      "loss": 0.9668,
+      "step": 11673
+    },
+    {
+      "epoch": 0.8999383287079864,
+      "grad_norm": 3.6503822803497314,
+      "learning_rate": 2.603064702107372e-07,
+      "loss": 0.8732,
+      "step": 11674
+    },
+    {
+      "epoch": 0.9000154178230034,
+      "grad_norm": 3.783892869949341,
+      "learning_rate": 2.599090463940385e-07,
+      "loss": 0.9098,
+      "step": 11675
+    },
+    {
+      "epoch": 0.9000925069380203,
+      "grad_norm": 3.5961296558380127,
+      "learning_rate": 2.5951191809746146e-07,
+      "loss": 0.8894,
+      "step": 11676
+    },
+    {
+      "epoch": 0.9001695960530373,
+      "grad_norm": 3.5369625091552734,
+      "learning_rate": 2.591150853457641e-07,
+      "loss": 0.8137,
+      "step": 11677
+    },
+    {
+      "epoch": 0.9002466851680543,
+      "grad_norm": 3.896723508834839,
+      "learning_rate": 2.587185481636878e-07,
+      "loss": 0.8884,
+      "step": 11678
+    },
+    {
+      "epoch": 0.9003237742830712,
+      "grad_norm": 3.630908250808716,
+      "learning_rate": 2.583223065759538e-07,
+      "loss": 0.9381,
+      "step": 11679
+    },
+    {
+      "epoch": 0.9004008633980882,
+      "grad_norm": 3.8402626514434814,
+      "learning_rate": 2.5792636060726684e-07,
+      "loss": 0.7879,
+      "step": 11680
+    },
+    {
+      "epoch": 0.9004779525131051,
+      "grad_norm": 3.4314825534820557,
+      "learning_rate": 2.5753071028231104e-07,
+      "loss": 0.7755,
+      "step": 11681
+    },
+    {
+      "epoch": 0.900555041628122,
+      "grad_norm": 3.852926015853882,
+      "learning_rate": 2.571353556257544e-07,
+      "loss": 0.8611,
+      "step": 11682
+    },
+    {
+      "epoch": 0.9006321307431391,
+      "grad_norm": 3.5122294425964355,
+      "learning_rate": 2.56740296662244e-07,
+      "loss": 0.8702,
+      "step": 11683
+    },
+    {
+      "epoch": 0.900709219858156,
+      "grad_norm": 3.657466411590576,
+      "learning_rate": 2.5634553341640943e-07,
+      "loss": 0.8854,
+      "step": 11684
+    },
+    {
+      "epoch": 0.900786308973173,
+      "grad_norm": 3.6080384254455566,
+      "learning_rate": 2.5595106591286335e-07,
+      "loss": 0.8445,
+      "step": 11685
+    },
+    {
+      "epoch": 0.9008633980881899,
+      "grad_norm": 3.5531303882598877,
+      "learning_rate": 2.555568941761982e-07,
+      "loss": 0.8449,
+      "step": 11686
+    },
+    {
+      "epoch": 0.9009404872032069,
+      "grad_norm": 3.68255352973938,
+      "learning_rate": 2.5516301823098944e-07,
+      "loss": 0.9405,
+      "step": 11687
+    },
+    {
+      "epoch": 0.9010175763182239,
+      "grad_norm": 3.9351344108581543,
+      "learning_rate": 2.547694381017912e-07,
+      "loss": 0.8447,
+      "step": 11688
+    },
+    {
+      "epoch": 0.9010946654332408,
+      "grad_norm": 4.277076244354248,
+      "learning_rate": 2.5437615381314284e-07,
+      "loss": 0.8522,
+      "step": 11689
+    },
+    {
+      "epoch": 0.9011717545482578,
+      "grad_norm": 3.9808146953582764,
+      "learning_rate": 2.5398316538956246e-07,
+      "loss": 0.8612,
+      "step": 11690
+    },
+    {
+      "epoch": 0.9012488436632747,
+      "grad_norm": 3.763561725616455,
+      "learning_rate": 2.5359047285555214e-07,
+      "loss": 0.9626,
+      "step": 11691
+    },
+    {
+      "epoch": 0.9013259327782918,
+      "grad_norm": 3.6474595069885254,
+      "learning_rate": 2.5319807623559287e-07,
+      "loss": 0.9554,
+      "step": 11692
+    },
+    {
+      "epoch": 0.9014030218933087,
+      "grad_norm": 4.234410762786865,
+      "learning_rate": 2.5280597555415067e-07,
+      "loss": 0.8349,
+      "step": 11693
+    },
+    {
+      "epoch": 0.9014801110083256,
+      "grad_norm": 3.9878194332122803,
+      "learning_rate": 2.524141708356681e-07,
+      "loss": 0.7692,
+      "step": 11694
+    },
+    {
+      "epoch": 0.9015572001233426,
+      "grad_norm": 3.9393675327301025,
+      "learning_rate": 2.52022662104574e-07,
+      "loss": 0.9336,
+      "step": 11695
+    },
+    {
+      "epoch": 0.9016342892383595,
+      "grad_norm": 3.7222704887390137,
+      "learning_rate": 2.5163144938527675e-07,
+      "loss": 0.8467,
+      "step": 11696
+    },
+    {
+      "epoch": 0.9017113783533766,
+      "grad_norm": 3.774393081665039,
+      "learning_rate": 2.5124053270216553e-07,
+      "loss": 0.9367,
+      "step": 11697
+    },
+    {
+      "epoch": 0.9017884674683935,
+      "grad_norm": 4.058575630187988,
+      "learning_rate": 2.5084991207961374e-07,
+      "loss": 0.896,
+      "step": 11698
+    },
+    {
+      "epoch": 0.9018655565834104,
+      "grad_norm": 3.4426233768463135,
+      "learning_rate": 2.50459587541973e-07,
+      "loss": 0.8279,
+      "step": 11699
+    },
+    {
+      "epoch": 0.9019426456984274,
+      "grad_norm": 3.468212842941284,
+      "learning_rate": 2.5006955911357823e-07,
+      "loss": 0.8706,
+      "step": 11700
+    },
+    {
+      "epoch": 0.9020197348134443,
+      "grad_norm": 3.775519609451294,
+      "learning_rate": 2.496798268187456e-07,
+      "loss": 0.8586,
+      "step": 11701
+    },
+    {
+      "epoch": 0.9020968239284614,
+      "grad_norm": 3.4390523433685303,
+      "learning_rate": 2.492903906817734e-07,
+      "loss": 0.781,
+      "step": 11702
+    },
+    {
+      "epoch": 0.9021739130434783,
+      "grad_norm": 3.695444107055664,
+      "learning_rate": 2.4890125072694114e-07,
+      "loss": 0.9408,
+      "step": 11703
+    },
+    {
+      "epoch": 0.9022510021584952,
+      "grad_norm": 3.9369289875030518,
+      "learning_rate": 2.4851240697850997e-07,
+      "loss": 0.8868,
+      "step": 11704
+    },
+    {
+      "epoch": 0.9023280912735122,
+      "grad_norm": 3.8416519165039062,
+      "learning_rate": 2.481238594607205e-07,
+      "loss": 0.8003,
+      "step": 11705
+    },
+    {
+      "epoch": 0.9024051803885291,
+      "grad_norm": 4.129755020141602,
+      "learning_rate": 2.477356081977983e-07,
+      "loss": 0.9659,
+      "step": 11706
+    },
+    {
+      "epoch": 0.9024822695035462,
+      "grad_norm": 4.082637786865234,
+      "learning_rate": 2.4734765321394793e-07,
+      "loss": 0.9412,
+      "step": 11707
+    },
+    {
+      "epoch": 0.9025593586185631,
+      "grad_norm": 3.7622687816619873,
+      "learning_rate": 2.469599945333567e-07,
+      "loss": 0.9492,
+      "step": 11708
+    },
+    {
+      "epoch": 0.90263644773358,
+      "grad_norm": 4.142516136169434,
+      "learning_rate": 2.465726321801942e-07,
+      "loss": 1.0039,
+      "step": 11709
+    },
+    {
+      "epoch": 0.902713536848597,
+      "grad_norm": 3.67720103263855,
+      "learning_rate": 2.4618556617860777e-07,
+      "loss": 0.9048,
+      "step": 11710
+    },
+    {
+      "epoch": 0.9027906259636139,
+      "grad_norm": 3.787903308868408,
+      "learning_rate": 2.45798796552732e-07,
+      "loss": 0.8038,
+      "step": 11711
+    },
+    {
+      "epoch": 0.902867715078631,
+      "grad_norm": 4.044528007507324,
+      "learning_rate": 2.454123233266781e-07,
+      "loss": 0.9192,
+      "step": 11712
+    },
+    {
+      "epoch": 0.9029448041936479,
+      "grad_norm": 3.6001529693603516,
+      "learning_rate": 2.4502614652454184e-07,
+      "loss": 0.8692,
+      "step": 11713
+    },
+    {
+      "epoch": 0.9030218933086648,
+      "grad_norm": 3.8225014209747314,
+      "learning_rate": 2.4464026617039784e-07,
+      "loss": 0.871,
+      "step": 11714
+    },
+    {
+      "epoch": 0.9030989824236818,
+      "grad_norm": 3.8792974948883057,
+      "learning_rate": 2.4425468228830516e-07,
+      "loss": 0.808,
+      "step": 11715
+    },
+    {
+      "epoch": 0.9031760715386987,
+      "grad_norm": 3.640425682067871,
+      "learning_rate": 2.438693949023019e-07,
+      "loss": 0.7728,
+      "step": 11716
+    },
+    {
+      "epoch": 0.9032531606537157,
+      "grad_norm": 4.195462703704834,
+      "learning_rate": 2.434844040364098e-07,
+      "loss": 0.9055,
+      "step": 11717
+    },
+    {
+      "epoch": 0.9033302497687327,
+      "grad_norm": 3.6664235591888428,
+      "learning_rate": 2.4309970971462984e-07,
+      "loss": 0.8798,
+      "step": 11718
+    },
+    {
+      "epoch": 0.9034073388837496,
+      "grad_norm": 3.623297691345215,
+      "learning_rate": 2.427153119609477e-07,
+      "loss": 0.8227,
+      "step": 11719
+    },
+    {
+      "epoch": 0.9034844279987666,
+      "grad_norm": 3.899339199066162,
+      "learning_rate": 2.4233121079932585e-07,
+      "loss": 0.884,
+      "step": 11720
+    },
+    {
+      "epoch": 0.9035615171137835,
+      "grad_norm": 3.89997935295105,
+      "learning_rate": 2.4194740625371303e-07,
+      "loss": 0.8831,
+      "step": 11721
+    },
+    {
+      "epoch": 0.9036386062288005,
+      "grad_norm": 3.7729170322418213,
+      "learning_rate": 2.4156389834803616e-07,
+      "loss": 0.8071,
+      "step": 11722
+    },
+    {
+      "epoch": 0.9037156953438175,
+      "grad_norm": 3.915086507797241,
+      "learning_rate": 2.411806871062061e-07,
+      "loss": 1.0403,
+      "step": 11723
+    },
+    {
+      "epoch": 0.9037927844588344,
+      "grad_norm": 3.809762716293335,
+      "learning_rate": 2.4079777255211434e-07,
+      "loss": 0.864,
+      "step": 11724
+    },
+    {
+      "epoch": 0.9038698735738514,
+      "grad_norm": 4.3212714195251465,
+      "learning_rate": 2.4041515470963227e-07,
+      "loss": 1.0022,
+      "step": 11725
+    },
+    {
+      "epoch": 0.9039469626888683,
+      "grad_norm": 3.7529261112213135,
+      "learning_rate": 2.400328336026148e-07,
+      "loss": 0.9564,
+      "step": 11726
+    },
+    {
+      "epoch": 0.9040240518038853,
+      "grad_norm": 3.535996675491333,
+      "learning_rate": 2.396508092548977e-07,
+      "loss": 0.8769,
+      "step": 11727
+    },
+    {
+      "epoch": 0.9041011409189023,
+      "grad_norm": 3.646723747253418,
+      "learning_rate": 2.392690816902976e-07,
+      "loss": 0.8884,
+      "step": 11728
+    },
+    {
+      "epoch": 0.9041782300339192,
+      "grad_norm": 4.039508819580078,
+      "learning_rate": 2.3888765093261434e-07,
+      "loss": 0.8506,
+      "step": 11729
+    },
+    {
+      "epoch": 0.9042553191489362,
+      "grad_norm": 4.302793502807617,
+      "learning_rate": 2.385065170056283e-07,
+      "loss": 0.9042,
+      "step": 11730
+    },
+    {
+      "epoch": 0.9043324082639531,
+      "grad_norm": 3.591334104537964,
+      "learning_rate": 2.3812567993309943e-07,
+      "loss": 0.7687,
+      "step": 11731
+    },
+    {
+      "epoch": 0.9044094973789701,
+      "grad_norm": 3.6448166370391846,
+      "learning_rate": 2.3774513973877254e-07,
+      "loss": 0.8568,
+      "step": 11732
+    },
+    {
+      "epoch": 0.904486586493987,
+      "grad_norm": 3.65342116355896,
+      "learning_rate": 2.3736489644637152e-07,
+      "loss": 0.8879,
+      "step": 11733
+    },
+    {
+      "epoch": 0.904563675609004,
+      "grad_norm": 3.8169586658477783,
+      "learning_rate": 2.3698495007960286e-07,
+      "loss": 0.9232,
+      "step": 11734
+    },
+    {
+      "epoch": 0.904640764724021,
+      "grad_norm": 3.728663921356201,
+      "learning_rate": 2.3660530066215493e-07,
+      "loss": 0.8023,
+      "step": 11735
+    },
+    {
+      "epoch": 0.9047178538390379,
+      "grad_norm": 3.7461793422698975,
+      "learning_rate": 2.3622594821769595e-07,
+      "loss": 0.8047,
+      "step": 11736
+    },
+    {
+      "epoch": 0.9047949429540549,
+      "grad_norm": 3.480088472366333,
+      "learning_rate": 2.358468927698765e-07,
+      "loss": 0.9529,
+      "step": 11737
+    },
+    {
+      "epoch": 0.9048720320690719,
+      "grad_norm": 3.421665668487549,
+      "learning_rate": 2.354681343423293e-07,
+      "loss": 0.8599,
+      "step": 11738
+    },
+    {
+      "epoch": 0.9049491211840888,
+      "grad_norm": 3.9536914825439453,
+      "learning_rate": 2.350896729586677e-07,
+      "loss": 0.7775,
+      "step": 11739
+    },
+    {
+      "epoch": 0.9050262102991058,
+      "grad_norm": 3.6503090858459473,
+      "learning_rate": 2.347115086424867e-07,
+      "loss": 0.9699,
+      "step": 11740
+    },
+    {
+      "epoch": 0.9051032994141227,
+      "grad_norm": 3.8761227130889893,
+      "learning_rate": 2.3433364141736414e-07,
+      "loss": 0.9433,
+      "step": 11741
+    },
+    {
+      "epoch": 0.9051803885291397,
+      "grad_norm": 3.4743881225585938,
+      "learning_rate": 2.3395607130685616e-07,
+      "loss": 0.9262,
+      "step": 11742
+    },
+    {
+      "epoch": 0.9052574776441566,
+      "grad_norm": 3.9404661655426025,
+      "learning_rate": 2.3357879833450335e-07,
+      "loss": 0.8681,
+      "step": 11743
+    },
+    {
+      "epoch": 0.9053345667591736,
+      "grad_norm": 3.5697805881500244,
+      "learning_rate": 2.332018225238264e-07,
+      "loss": 0.7547,
+      "step": 11744
+    },
+    {
+      "epoch": 0.9054116558741906,
+      "grad_norm": 3.835613965988159,
+      "learning_rate": 2.3282514389832755e-07,
+      "loss": 0.9139,
+      "step": 11745
+    },
+    {
+      "epoch": 0.9054887449892075,
+      "grad_norm": 3.496504306793213,
+      "learning_rate": 2.3244876248149196e-07,
+      "loss": 0.9144,
+      "step": 11746
+    },
+    {
+      "epoch": 0.9055658341042245,
+      "grad_norm": 3.6928272247314453,
+      "learning_rate": 2.3207267829678416e-07,
+      "loss": 0.8003,
+      "step": 11747
+    },
+    {
+      "epoch": 0.9056429232192414,
+      "grad_norm": 3.711456537246704,
+      "learning_rate": 2.3169689136765038e-07,
+      "loss": 0.9611,
+      "step": 11748
+    },
+    {
+      "epoch": 0.9057200123342584,
+      "grad_norm": 3.7150790691375732,
+      "learning_rate": 2.3132140171752027e-07,
+      "loss": 0.9735,
+      "step": 11749
+    },
+    {
+      "epoch": 0.9057971014492754,
+      "grad_norm": 3.9875683784484863,
+      "learning_rate": 2.3094620936980283e-07,
+      "loss": 0.9415,
+      "step": 11750
+    },
+    {
+      "epoch": 0.9058741905642923,
+      "grad_norm": 3.7568795680999756,
+      "learning_rate": 2.3057131434788994e-07,
+      "loss": 0.9143,
+      "step": 11751
+    },
+    {
+      "epoch": 0.9059512796793093,
+      "grad_norm": 3.826282262802124,
+      "learning_rate": 2.3019671667515454e-07,
+      "loss": 0.9529,
+      "step": 11752
+    },
+    {
+      "epoch": 0.9060283687943262,
+      "grad_norm": 3.865046262741089,
+      "learning_rate": 2.2982241637494962e-07,
+      "loss": 1.0432,
+      "step": 11753
+    },
+    {
+      "epoch": 0.9061054579093432,
+      "grad_norm": 3.8025035858154297,
+      "learning_rate": 2.2944841347061153e-07,
+      "loss": 0.8933,
+      "step": 11754
+    },
+    {
+      "epoch": 0.9061825470243602,
+      "grad_norm": 3.6848461627960205,
+      "learning_rate": 2.2907470798545772e-07,
+      "loss": 0.9039,
+      "step": 11755
+    },
+    {
+      "epoch": 0.9062596361393771,
+      "grad_norm": 3.549743413925171,
+      "learning_rate": 2.2870129994278733e-07,
+      "loss": 0.9509,
+      "step": 11756
+    },
+    {
+      "epoch": 0.9063367252543941,
+      "grad_norm": 3.480574607849121,
+      "learning_rate": 2.2832818936587954e-07,
+      "loss": 0.7861,
+      "step": 11757
+    },
+    {
+      "epoch": 0.906413814369411,
+      "grad_norm": 3.917457342147827,
+      "learning_rate": 2.2795537627799512e-07,
+      "loss": 0.773,
+      "step": 11758
+    },
+    {
+      "epoch": 0.906490903484428,
+      "grad_norm": 3.638928174972534,
+      "learning_rate": 2.2758286070237889e-07,
+      "loss": 0.9302,
+      "step": 11759
+    },
+    {
+      "epoch": 0.906567992599445,
+      "grad_norm": 3.471503257751465,
+      "learning_rate": 2.2721064266225335e-07,
+      "loss": 0.8307,
+      "step": 11760
+    },
+    {
+      "epoch": 0.9066450817144619,
+      "grad_norm": 3.8338592052459717,
+      "learning_rate": 2.2683872218082659e-07,
+      "loss": 0.8046,
+      "step": 11761
+    },
+    {
+      "epoch": 0.9067221708294789,
+      "grad_norm": 3.566448450088501,
+      "learning_rate": 2.2646709928128397e-07,
+      "loss": 0.8662,
+      "step": 11762
+    },
+    {
+      "epoch": 0.9067992599444958,
+      "grad_norm": 3.907142162322998,
+      "learning_rate": 2.2609577398679472e-07,
+      "loss": 0.9882,
+      "step": 11763
+    },
+    {
+      "epoch": 0.9068763490595128,
+      "grad_norm": 3.876858949661255,
+      "learning_rate": 2.2572474632050977e-07,
+      "loss": 0.8642,
+      "step": 11764
+    },
+    {
+      "epoch": 0.9069534381745298,
+      "grad_norm": 3.5765106678009033,
+      "learning_rate": 2.253540163055601e-07,
+      "loss": 0.829,
+      "step": 11765
+    },
+    {
+      "epoch": 0.9070305272895467,
+      "grad_norm": 4.412302494049072,
+      "learning_rate": 2.249835839650588e-07,
+      "loss": 1.0133,
+      "step": 11766
+    },
+    {
+      "epoch": 0.9071076164045637,
+      "grad_norm": 3.776357412338257,
+      "learning_rate": 2.2461344932210084e-07,
+      "loss": 0.8659,
+      "step": 11767
+    },
+    {
+      "epoch": 0.9071847055195806,
+      "grad_norm": 4.0795392990112305,
+      "learning_rate": 2.2424361239976212e-07,
+      "loss": 0.9052,
+      "step": 11768
+    },
+    {
+      "epoch": 0.9072617946345976,
+      "grad_norm": 3.6457784175872803,
+      "learning_rate": 2.2387407322109922e-07,
+      "loss": 0.8973,
+      "step": 11769
+    },
+    {
+      "epoch": 0.9073388837496146,
+      "grad_norm": 3.8819169998168945,
+      "learning_rate": 2.2350483180915206e-07,
+      "loss": 0.9255,
+      "step": 11770
+    },
+    {
+      "epoch": 0.9074159728646315,
+      "grad_norm": 4.195042610168457,
+      "learning_rate": 2.2313588818694055e-07,
+      "loss": 0.9305,
+      "step": 11771
+    },
+    {
+      "epoch": 0.9074930619796485,
+      "grad_norm": 3.5375936031341553,
+      "learning_rate": 2.2276724237746684e-07,
+      "loss": 0.9001,
+      "step": 11772
+    },
+    {
+      "epoch": 0.9075701510946654,
+      "grad_norm": 4.148855209350586,
+      "learning_rate": 2.223988944037131e-07,
+      "loss": 0.9946,
+      "step": 11773
+    },
+    {
+      "epoch": 0.9076472402096823,
+      "grad_norm": 3.680772304534912,
+      "learning_rate": 2.220308442886443e-07,
+      "loss": 0.8558,
+      "step": 11774
+    },
+    {
+      "epoch": 0.9077243293246994,
+      "grad_norm": 3.458876848220825,
+      "learning_rate": 2.2166309205520707e-07,
+      "loss": 0.9057,
+      "step": 11775
+    },
+    {
+      "epoch": 0.9078014184397163,
+      "grad_norm": 3.533604145050049,
+      "learning_rate": 2.2129563772632755e-07,
+      "loss": 0.8633,
+      "step": 11776
+    },
+    {
+      "epoch": 0.9078785075547333,
+      "grad_norm": 4.154286861419678,
+      "learning_rate": 2.2092848132491628e-07,
+      "loss": 1.0149,
+      "step": 11777
+    },
+    {
+      "epoch": 0.9079555966697502,
+      "grad_norm": 3.6534740924835205,
+      "learning_rate": 2.205616228738633e-07,
+      "loss": 0.7558,
+      "step": 11778
+    },
+    {
+      "epoch": 0.9080326857847671,
+      "grad_norm": 3.837028741836548,
+      "learning_rate": 2.201950623960386e-07,
+      "loss": 0.9613,
+      "step": 11779
+    },
+    {
+      "epoch": 0.9081097748997842,
+      "grad_norm": 3.6078686714172363,
+      "learning_rate": 2.1982879991429728e-07,
+      "loss": 0.8294,
+      "step": 11780
+    },
+    {
+      "epoch": 0.9081868640148011,
+      "grad_norm": 3.634082794189453,
+      "learning_rate": 2.1946283545147274e-07,
+      "loss": 0.9769,
+      "step": 11781
+    },
+    {
+      "epoch": 0.9082639531298181,
+      "grad_norm": 3.799694776535034,
+      "learning_rate": 2.1909716903038114e-07,
+      "loss": 0.945,
+      "step": 11782
+    },
+    {
+      "epoch": 0.908341042244835,
+      "grad_norm": 3.5379395484924316,
+      "learning_rate": 2.1873180067382095e-07,
+      "loss": 0.8928,
+      "step": 11783
+    },
+    {
+      "epoch": 0.9084181313598519,
+      "grad_norm": 3.9974184036254883,
+      "learning_rate": 2.1836673040456947e-07,
+      "loss": 0.946,
+      "step": 11784
+    },
+    {
+      "epoch": 0.908495220474869,
+      "grad_norm": 3.915295124053955,
+      "learning_rate": 2.1800195824538794e-07,
+      "loss": 1.0305,
+      "step": 11785
+    },
+    {
+      "epoch": 0.9085723095898859,
+      "grad_norm": 3.7968573570251465,
+      "learning_rate": 2.1763748421901764e-07,
+      "loss": 0.8681,
+      "step": 11786
+    },
+    {
+      "epoch": 0.9086493987049029,
+      "grad_norm": 3.79858136177063,
+      "learning_rate": 2.172733083481815e-07,
+      "loss": 0.8694,
+      "step": 11787
+    },
+    {
+      "epoch": 0.9087264878199198,
+      "grad_norm": 3.833800792694092,
+      "learning_rate": 2.1690943065558412e-07,
+      "loss": 0.9167,
+      "step": 11788
+    },
+    {
+      "epoch": 0.9088035769349367,
+      "grad_norm": 3.841184616088867,
+      "learning_rate": 2.1654585116391236e-07,
+      "loss": 0.8676,
+      "step": 11789
+    },
+    {
+      "epoch": 0.9088806660499538,
+      "grad_norm": 3.632209062576294,
+      "learning_rate": 2.1618256989583197e-07,
+      "loss": 0.9878,
+      "step": 11790
+    },
+    {
+      "epoch": 0.9089577551649707,
+      "grad_norm": 4.1442742347717285,
+      "learning_rate": 2.1581958687399206e-07,
+      "loss": 0.9334,
+      "step": 11791
+    },
+    {
+      "epoch": 0.9090348442799877,
+      "grad_norm": 3.917574167251587,
+      "learning_rate": 2.1545690212102344e-07,
+      "loss": 0.822,
+      "step": 11792
+    },
+    {
+      "epoch": 0.9091119333950046,
+      "grad_norm": 4.023653507232666,
+      "learning_rate": 2.1509451565953688e-07,
+      "loss": 0.9154,
+      "step": 11793
+    },
+    {
+      "epoch": 0.9091890225100215,
+      "grad_norm": 3.650171995162964,
+      "learning_rate": 2.1473242751212653e-07,
+      "loss": 0.7958,
+      "step": 11794
+    },
+    {
+      "epoch": 0.9092661116250386,
+      "grad_norm": 3.726482391357422,
+      "learning_rate": 2.143706377013649e-07,
+      "loss": 0.9593,
+      "step": 11795
+    },
+    {
+      "epoch": 0.9093432007400555,
+      "grad_norm": 3.3501317501068115,
+      "learning_rate": 2.140091462498084e-07,
+      "loss": 0.9107,
+      "step": 11796
+    },
+    {
+      "epoch": 0.9094202898550725,
+      "grad_norm": 3.5928847789764404,
+      "learning_rate": 2.1364795317999455e-07,
+      "loss": 0.845,
+      "step": 11797
+    },
+    {
+      "epoch": 0.9094973789700894,
+      "grad_norm": 3.925431489944458,
+      "learning_rate": 2.13287058514442e-07,
+      "loss": 0.9261,
+      "step": 11798
+    },
+    {
+      "epoch": 0.9095744680851063,
+      "grad_norm": 4.184495449066162,
+      "learning_rate": 2.1292646227564995e-07,
+      "loss": 1.0248,
+      "step": 11799
+    },
+    {
+      "epoch": 0.9096515572001234,
+      "grad_norm": 3.948833703994751,
+      "learning_rate": 2.125661644860999e-07,
+      "loss": 0.9499,
+      "step": 11800
+    },
+    {
+      "epoch": 0.9097286463151403,
+      "grad_norm": 3.7556886672973633,
+      "learning_rate": 2.1220616516825497e-07,
+      "loss": 0.909,
+      "step": 11801
+    },
+    {
+      "epoch": 0.9098057354301573,
+      "grad_norm": 3.4662973880767822,
+      "learning_rate": 2.1184646434455947e-07,
+      "loss": 0.9219,
+      "step": 11802
+    },
+    {
+      "epoch": 0.9098828245451742,
+      "grad_norm": 3.4703729152679443,
+      "learning_rate": 2.1148706203743762e-07,
+      "loss": 0.8062,
+      "step": 11803
+    },
+    {
+      "epoch": 0.9099599136601911,
+      "grad_norm": 3.3203537464141846,
+      "learning_rate": 2.111279582692982e-07,
+      "loss": 0.8091,
+      "step": 11804
+    },
+    {
+      "epoch": 0.9100370027752082,
+      "grad_norm": 3.9638702869415283,
+      "learning_rate": 2.1076915306252776e-07,
+      "loss": 0.9664,
+      "step": 11805
+    },
+    {
+      "epoch": 0.9101140918902251,
+      "grad_norm": 3.615192174911499,
+      "learning_rate": 2.104106464394967e-07,
+      "loss": 0.8336,
+      "step": 11806
+    },
+    {
+      "epoch": 0.9101911810052421,
+      "grad_norm": 4.019468784332275,
+      "learning_rate": 2.1005243842255552e-07,
+      "loss": 0.9677,
+      "step": 11807
+    },
+    {
+      "epoch": 0.910268270120259,
+      "grad_norm": 3.62318754196167,
+      "learning_rate": 2.0969452903403742e-07,
+      "loss": 0.902,
+      "step": 11808
+    },
+    {
+      "epoch": 0.9103453592352759,
+      "grad_norm": 3.653689384460449,
+      "learning_rate": 2.0933691829625624e-07,
+      "loss": 0.903,
+      "step": 11809
+    },
+    {
+      "epoch": 0.910422448350293,
+      "grad_norm": 3.717650890350342,
+      "learning_rate": 2.0897960623150581e-07,
+      "loss": 0.9309,
+      "step": 11810
+    },
+    {
+      "epoch": 0.9104995374653099,
+      "grad_norm": 3.7195537090301514,
+      "learning_rate": 2.0862259286206387e-07,
+      "loss": 0.9048,
+      "step": 11811
+    },
+    {
+      "epoch": 0.9105766265803269,
+      "grad_norm": 3.824012279510498,
+      "learning_rate": 2.0826587821018818e-07,
+      "loss": 0.8867,
+      "step": 11812
+    },
+    {
+      "epoch": 0.9106537156953438,
+      "grad_norm": 3.8283004760742188,
+      "learning_rate": 2.0790946229811758e-07,
+      "loss": 0.789,
+      "step": 11813
+    },
+    {
+      "epoch": 0.9107308048103607,
+      "grad_norm": 3.8848516941070557,
+      "learning_rate": 2.0755334514807379e-07,
+      "loss": 0.8425,
+      "step": 11814
+    },
+    {
+      "epoch": 0.9108078939253778,
+      "grad_norm": 3.4980616569519043,
+      "learning_rate": 2.0719752678225846e-07,
+      "loss": 0.849,
+      "step": 11815
+    },
+    {
+      "epoch": 0.9108849830403947,
+      "grad_norm": 3.96785044670105,
+      "learning_rate": 2.0684200722285386e-07,
+      "loss": 0.8735,
+      "step": 11816
+    },
+    {
+      "epoch": 0.9109620721554117,
+      "grad_norm": 3.910421133041382,
+      "learning_rate": 2.0648678649202615e-07,
+      "loss": 1.0155,
+      "step": 11817
+    },
+    {
+      "epoch": 0.9110391612704286,
+      "grad_norm": 3.8571994304656982,
+      "learning_rate": 2.0613186461192092e-07,
+      "loss": 0.7947,
+      "step": 11818
+    },
+    {
+      "epoch": 0.9111162503854455,
+      "grad_norm": 3.5077316761016846,
+      "learning_rate": 2.057772416046655e-07,
+      "loss": 0.8644,
+      "step": 11819
+    },
+    {
+      "epoch": 0.9111933395004626,
+      "grad_norm": 4.075991153717041,
+      "learning_rate": 2.0542291749237053e-07,
+      "loss": 0.8711,
+      "step": 11820
+    },
+    {
+      "epoch": 0.9112704286154795,
+      "grad_norm": 3.9802606105804443,
+      "learning_rate": 2.050688922971239e-07,
+      "loss": 0.7961,
+      "step": 11821
+    },
+    {
+      "epoch": 0.9113475177304965,
+      "grad_norm": 4.114253520965576,
+      "learning_rate": 2.0471516604099794e-07,
+      "loss": 0.8708,
+      "step": 11822
+    },
+    {
+      "epoch": 0.9114246068455134,
+      "grad_norm": 3.611233711242676,
+      "learning_rate": 2.0436173874604615e-07,
+      "loss": 0.7748,
+      "step": 11823
+    },
+    {
+      "epoch": 0.9115016959605303,
+      "grad_norm": 3.610714912414551,
+      "learning_rate": 2.0400861043430254e-07,
+      "loss": 0.6573,
+      "step": 11824
+    },
+    {
+      "epoch": 0.9115787850755473,
+      "grad_norm": 3.904794216156006,
+      "learning_rate": 2.036557811277834e-07,
+      "loss": 0.9248,
+      "step": 11825
+    },
+    {
+      "epoch": 0.9116558741905643,
+      "grad_norm": 3.6194839477539062,
+      "learning_rate": 2.0330325084848612e-07,
+      "loss": 0.837,
+      "step": 11826
+    },
+    {
+      "epoch": 0.9117329633055813,
+      "grad_norm": 3.733680248260498,
+      "learning_rate": 2.029510196183876e-07,
+      "loss": 0.8843,
+      "step": 11827
+    },
+    {
+      "epoch": 0.9118100524205982,
+      "grad_norm": 3.8289010524749756,
+      "learning_rate": 2.02599087459448e-07,
+      "loss": 0.9078,
+      "step": 11828
+    },
+    {
+      "epoch": 0.9118871415356151,
+      "grad_norm": 3.486313581466675,
+      "learning_rate": 2.0224745439360928e-07,
+      "loss": 0.8279,
+      "step": 11829
+    },
+    {
+      "epoch": 0.9119642306506321,
+      "grad_norm": 3.5716099739074707,
+      "learning_rate": 2.0189612044279384e-07,
+      "loss": 0.9217,
+      "step": 11830
+    },
+    {
+      "epoch": 0.9120413197656491,
+      "grad_norm": 3.4846935272216797,
+      "learning_rate": 2.015450856289053e-07,
+      "loss": 0.7435,
+      "step": 11831
+    },
+    {
+      "epoch": 0.9121184088806661,
+      "grad_norm": 3.8818373680114746,
+      "learning_rate": 2.0119434997382893e-07,
+      "loss": 0.9561,
+      "step": 11832
+    },
+    {
+      "epoch": 0.912195497995683,
+      "grad_norm": 3.465576171875,
+      "learning_rate": 2.0084391349943055e-07,
+      "loss": 0.8902,
+      "step": 11833
+    },
+    {
+      "epoch": 0.9122725871106999,
+      "grad_norm": 4.149135589599609,
+      "learning_rate": 2.0049377622755885e-07,
+      "loss": 0.9379,
+      "step": 11834
+    },
+    {
+      "epoch": 0.9123496762257169,
+      "grad_norm": 4.032505512237549,
+      "learning_rate": 2.0014393818004295e-07,
+      "loss": 0.8065,
+      "step": 11835
+    },
+    {
+      "epoch": 0.9124267653407339,
+      "grad_norm": 4.063005447387695,
+      "learning_rate": 1.9979439937869383e-07,
+      "loss": 1.0096,
+      "step": 11836
+    },
+    {
+      "epoch": 0.9125038544557509,
+      "grad_norm": 3.7113261222839355,
+      "learning_rate": 1.9944515984530343e-07,
+      "loss": 0.9335,
+      "step": 11837
+    },
+    {
+      "epoch": 0.9125809435707678,
+      "grad_norm": 4.403497219085693,
+      "learning_rate": 1.9909621960164382e-07,
+      "loss": 0.9352,
+      "step": 11838
+    },
+    {
+      "epoch": 0.9126580326857847,
+      "grad_norm": 3.810023069381714,
+      "learning_rate": 1.9874757866947036e-07,
+      "loss": 0.9586,
+      "step": 11839
+    },
+    {
+      "epoch": 0.9127351218008017,
+      "grad_norm": 3.5744166374206543,
+      "learning_rate": 1.98399237070519e-07,
+      "loss": 0.8838,
+      "step": 11840
+    },
+    {
+      "epoch": 0.9128122109158187,
+      "grad_norm": 3.613818407058716,
+      "learning_rate": 1.9805119482650737e-07,
+      "loss": 0.8913,
+      "step": 11841
+    },
+    {
+      "epoch": 0.9128893000308357,
+      "grad_norm": 3.6662750244140625,
+      "learning_rate": 1.977034519591342e-07,
+      "loss": 0.9187,
+      "step": 11842
+    },
+    {
+      "epoch": 0.9129663891458526,
+      "grad_norm": 4.007206439971924,
+      "learning_rate": 1.9735600849007774e-07,
+      "loss": 0.8897,
+      "step": 11843
+    },
+    {
+      "epoch": 0.9130434782608695,
+      "grad_norm": 4.291807651519775,
+      "learning_rate": 1.970088644410012e-07,
+      "loss": 0.8999,
+      "step": 11844
+    },
+    {
+      "epoch": 0.9131205673758865,
+      "grad_norm": 3.8497557640075684,
+      "learning_rate": 1.966620198335467e-07,
+      "loss": 0.8368,
+      "step": 11845
+    },
+    {
+      "epoch": 0.9131976564909035,
+      "grad_norm": 3.969374179840088,
+      "learning_rate": 1.963154746893392e-07,
+      "loss": 0.8243,
+      "step": 11846
+    },
+    {
+      "epoch": 0.9132747456059205,
+      "grad_norm": 3.7255005836486816,
+      "learning_rate": 1.9596922902998193e-07,
+      "loss": 0.8739,
+      "step": 11847
+    },
+    {
+      "epoch": 0.9133518347209374,
+      "grad_norm": 3.940896511077881,
+      "learning_rate": 1.956232828770621e-07,
+      "loss": 0.8913,
+      "step": 11848
+    },
+    {
+      "epoch": 0.9134289238359543,
+      "grad_norm": 3.7923643589019775,
+      "learning_rate": 1.9527763625214857e-07,
+      "loss": 0.906,
+      "step": 11849
+    },
+    {
+      "epoch": 0.9135060129509713,
+      "grad_norm": 3.9090356826782227,
+      "learning_rate": 1.9493228917679018e-07,
+      "loss": 0.9555,
+      "step": 11850
+    },
+    {
+      "epoch": 0.9135831020659883,
+      "grad_norm": 3.677847146987915,
+      "learning_rate": 1.9458724167251753e-07,
+      "loss": 0.8394,
+      "step": 11851
+    },
+    {
+      "epoch": 0.9136601911810053,
+      "grad_norm": 3.4867942333221436,
+      "learning_rate": 1.942424937608428e-07,
+      "loss": 0.8895,
+      "step": 11852
+    },
+    {
+      "epoch": 0.9137372802960222,
+      "grad_norm": 3.678907632827759,
+      "learning_rate": 1.9389804546325885e-07,
+      "loss": 0.832,
+      "step": 11853
+    },
+    {
+      "epoch": 0.9138143694110391,
+      "grad_norm": 3.7241756916046143,
+      "learning_rate": 1.935538968012396e-07,
+      "loss": 0.8667,
+      "step": 11854
+    },
+    {
+      "epoch": 0.9138914585260561,
+      "grad_norm": 3.8677916526794434,
+      "learning_rate": 1.9321004779624232e-07,
+      "loss": 0.8854,
+      "step": 11855
+    },
+    {
+      "epoch": 0.913968547641073,
+      "grad_norm": 3.4867103099823,
+      "learning_rate": 1.9286649846970318e-07,
+      "loss": 0.8591,
+      "step": 11856
+    },
+    {
+      "epoch": 0.9140456367560901,
+      "grad_norm": 3.9573867321014404,
+      "learning_rate": 1.9252324884304174e-07,
+      "loss": 0.8497,
+      "step": 11857
+    },
+    {
+      "epoch": 0.914122725871107,
+      "grad_norm": 3.7366578578948975,
+      "learning_rate": 1.9218029893765643e-07,
+      "loss": 0.9457,
+      "step": 11858
+    },
+    {
+      "epoch": 0.9141998149861239,
+      "grad_norm": 3.685131549835205,
+      "learning_rate": 1.9183764877492905e-07,
+      "loss": 0.8712,
+      "step": 11859
+    },
+    {
+      "epoch": 0.9142769041011409,
+      "grad_norm": 3.735658645629883,
+      "learning_rate": 1.914952983762225e-07,
+      "loss": 0.8748,
+      "step": 11860
+    },
+    {
+      "epoch": 0.9143539932161578,
+      "grad_norm": 4.044561386108398,
+      "learning_rate": 1.9115324776288024e-07,
+      "loss": 0.938,
+      "step": 11861
+    },
+    {
+      "epoch": 0.9144310823311749,
+      "grad_norm": 3.9689042568206787,
+      "learning_rate": 1.9081149695622693e-07,
+      "loss": 0.8393,
+      "step": 11862
+    },
+    {
+      "epoch": 0.9145081714461918,
+      "grad_norm": 3.726811170578003,
+      "learning_rate": 1.9047004597757045e-07,
+      "loss": 0.7973,
+      "step": 11863
+    },
+    {
+      "epoch": 0.9145852605612088,
+      "grad_norm": 3.691999673843384,
+      "learning_rate": 1.9012889484819665e-07,
+      "loss": 0.8179,
+      "step": 11864
+    },
+    {
+      "epoch": 0.9146623496762257,
+      "grad_norm": 3.436737298965454,
+      "learning_rate": 1.8978804358937508e-07,
+      "loss": 0.8289,
+      "step": 11865
+    },
+    {
+      "epoch": 0.9147394387912426,
+      "grad_norm": 4.035014629364014,
+      "learning_rate": 1.8944749222235658e-07,
+      "loss": 0.9811,
+      "step": 11866
+    },
+    {
+      "epoch": 0.9148165279062597,
+      "grad_norm": 3.6229660511016846,
+      "learning_rate": 1.8910724076837196e-07,
+      "loss": 0.9873,
+      "step": 11867
+    },
+    {
+      "epoch": 0.9148936170212766,
+      "grad_norm": 3.7031772136688232,
+      "learning_rate": 1.887672892486353e-07,
+      "loss": 1.0282,
+      "step": 11868
+    },
+    {
+      "epoch": 0.9149707061362936,
+      "grad_norm": 3.981867551803589,
+      "learning_rate": 1.8842763768434024e-07,
+      "loss": 0.9978,
+      "step": 11869
+    },
+    {
+      "epoch": 0.9150477952513105,
+      "grad_norm": 3.8303933143615723,
+      "learning_rate": 1.880882860966615e-07,
+      "loss": 0.9158,
+      "step": 11870
+    },
+    {
+      "epoch": 0.9151248843663274,
+      "grad_norm": 3.6401233673095703,
+      "learning_rate": 1.8774923450675665e-07,
+      "loss": 0.8627,
+      "step": 11871
+    },
+    {
+      "epoch": 0.9152019734813445,
+      "grad_norm": 3.968207359313965,
+      "learning_rate": 1.8741048293576424e-07,
+      "loss": 0.9463,
+      "step": 11872
+    },
+    {
+      "epoch": 0.9152790625963614,
+      "grad_norm": 3.688127040863037,
+      "learning_rate": 1.8707203140480245e-07,
+      "loss": 0.9945,
+      "step": 11873
+    },
+    {
+      "epoch": 0.9153561517113784,
+      "grad_norm": 3.7214889526367188,
+      "learning_rate": 1.8673387993497383e-07,
+      "loss": 0.8459,
+      "step": 11874
+    },
+    {
+      "epoch": 0.9154332408263953,
+      "grad_norm": 3.942366600036621,
+      "learning_rate": 1.8639602854735872e-07,
+      "loss": 0.7865,
+      "step": 11875
+    },
+    {
+      "epoch": 0.9155103299414122,
+      "grad_norm": 3.5286405086517334,
+      "learning_rate": 1.8605847726302085e-07,
+      "loss": 0.9035,
+      "step": 11876
+    },
+    {
+      "epoch": 0.9155874190564293,
+      "grad_norm": 3.608853816986084,
+      "learning_rate": 1.8572122610300447e-07,
+      "loss": 0.8938,
+      "step": 11877
+    },
+    {
+      "epoch": 0.9156645081714462,
+      "grad_norm": 3.8086750507354736,
+      "learning_rate": 1.8538427508833612e-07,
+      "loss": 0.8832,
+      "step": 11878
+    },
+    {
+      "epoch": 0.9157415972864632,
+      "grad_norm": 3.3695833683013916,
+      "learning_rate": 1.8504762424002342e-07,
+      "loss": 0.9145,
+      "step": 11879
+    },
+    {
+      "epoch": 0.9158186864014801,
+      "grad_norm": 3.7012627124786377,
+      "learning_rate": 1.8471127357905348e-07,
+      "loss": 0.9761,
+      "step": 11880
+    },
+    {
+      "epoch": 0.915895775516497,
+      "grad_norm": 3.448864459991455,
+      "learning_rate": 1.843752231263962e-07,
+      "loss": 0.7754,
+      "step": 11881
+    },
+    {
+      "epoch": 0.9159728646315141,
+      "grad_norm": 3.5888924598693848,
+      "learning_rate": 1.8403947290300318e-07,
+      "loss": 0.936,
+      "step": 11882
+    },
+    {
+      "epoch": 0.916049953746531,
+      "grad_norm": 3.8859617710113525,
+      "learning_rate": 1.8370402292980703e-07,
+      "loss": 0.9449,
+      "step": 11883
+    },
+    {
+      "epoch": 0.916127042861548,
+      "grad_norm": 3.799428939819336,
+      "learning_rate": 1.8336887322772e-07,
+      "loss": 0.8585,
+      "step": 11884
+    },
+    {
+      "epoch": 0.9162041319765649,
+      "grad_norm": 3.7906007766723633,
+      "learning_rate": 1.8303402381763924e-07,
+      "loss": 0.9822,
+      "step": 11885
+    },
+    {
+      "epoch": 0.9162812210915818,
+      "grad_norm": 3.4046058654785156,
+      "learning_rate": 1.8269947472043804e-07,
+      "loss": 0.8123,
+      "step": 11886
+    },
+    {
+      "epoch": 0.9163583102065989,
+      "grad_norm": 3.974215030670166,
+      "learning_rate": 1.823652259569747e-07,
+      "loss": 0.9087,
+      "step": 11887
+    },
+    {
+      "epoch": 0.9164353993216158,
+      "grad_norm": 3.9902284145355225,
+      "learning_rate": 1.8203127754808923e-07,
+      "loss": 0.937,
+      "step": 11888
+    },
+    {
+      "epoch": 0.9165124884366328,
+      "grad_norm": 3.7469322681427,
+      "learning_rate": 1.8169762951460112e-07,
+      "loss": 0.8929,
+      "step": 11889
+    },
+    {
+      "epoch": 0.9165895775516497,
+      "grad_norm": 3.5326128005981445,
+      "learning_rate": 1.8136428187731037e-07,
+      "loss": 0.8825,
+      "step": 11890
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 3.375762462615967,
+      "learning_rate": 1.810312346570009e-07,
+      "loss": 0.8043,
+      "step": 11891
+    },
+    {
+      "epoch": 0.9167437557816837,
+      "grad_norm": 3.9036519527435303,
+      "learning_rate": 1.8069848787443556e-07,
+      "loss": 1.036,
+      "step": 11892
+    },
+    {
+      "epoch": 0.9168208448967006,
+      "grad_norm": 3.671597719192505,
+      "learning_rate": 1.8036604155035942e-07,
+      "loss": 0.9171,
+      "step": 11893
+    },
+    {
+      "epoch": 0.9168979340117176,
+      "grad_norm": 3.8690693378448486,
+      "learning_rate": 1.8003389570549978e-07,
+      "loss": 0.8524,
+      "step": 11894
+    },
+    {
+      "epoch": 0.9169750231267345,
+      "grad_norm": 3.334073543548584,
+      "learning_rate": 1.7970205036056287e-07,
+      "loss": 0.9197,
+      "step": 11895
+    },
+    {
+      "epoch": 0.9170521122417514,
+      "grad_norm": 3.826357126235962,
+      "learning_rate": 1.793705055362388e-07,
+      "loss": 0.8555,
+      "step": 11896
+    },
+    {
+      "epoch": 0.9171292013567685,
+      "grad_norm": 3.7970800399780273,
+      "learning_rate": 1.7903926125319603e-07,
+      "loss": 0.845,
+      "step": 11897
+    },
+    {
+      "epoch": 0.9172062904717854,
+      "grad_norm": 3.9418017864227295,
+      "learning_rate": 1.7870831753208752e-07,
+      "loss": 0.9457,
+      "step": 11898
+    },
+    {
+      "epoch": 0.9172833795868024,
+      "grad_norm": 4.235414981842041,
+      "learning_rate": 1.7837767439354503e-07,
+      "loss": 0.9059,
+      "step": 11899
+    },
+    {
+      "epoch": 0.9173604687018193,
+      "grad_norm": 3.8422882556915283,
+      "learning_rate": 1.7804733185818378e-07,
+      "loss": 0.852,
+      "step": 11900
+    },
+    {
+      "epoch": 0.9174375578168362,
+      "grad_norm": 4.103250026702881,
+      "learning_rate": 1.7771728994659677e-07,
+      "loss": 0.8487,
+      "step": 11901
+    },
+    {
+      "epoch": 0.9175146469318533,
+      "grad_norm": 3.8441503047943115,
+      "learning_rate": 1.7738754867936193e-07,
+      "loss": 0.8913,
+      "step": 11902
+    },
+    {
+      "epoch": 0.9175917360468702,
+      "grad_norm": 3.94001841545105,
+      "learning_rate": 1.7705810807703616e-07,
+      "loss": 0.9332,
+      "step": 11903
+    },
+    {
+      "epoch": 0.9176688251618872,
+      "grad_norm": 3.453479766845703,
+      "learning_rate": 1.7672896816015861e-07,
+      "loss": 0.917,
+      "step": 11904
+    },
+    {
+      "epoch": 0.9177459142769041,
+      "grad_norm": 3.9532957077026367,
+      "learning_rate": 1.7640012894925008e-07,
+      "loss": 0.8864,
+      "step": 11905
+    },
+    {
+      "epoch": 0.917823003391921,
+      "grad_norm": 3.767510414123535,
+      "learning_rate": 1.7607159046481138e-07,
+      "loss": 0.9912,
+      "step": 11906
+    },
+    {
+      "epoch": 0.917900092506938,
+      "grad_norm": 3.960230588912964,
+      "learning_rate": 1.7574335272732445e-07,
+      "loss": 0.8642,
+      "step": 11907
+    },
+    {
+      "epoch": 0.917977181621955,
+      "grad_norm": 4.039930820465088,
+      "learning_rate": 1.7541541575725464e-07,
+      "loss": 1.0177,
+      "step": 11908
+    },
+    {
+      "epoch": 0.918054270736972,
+      "grad_norm": 4.087274074554443,
+      "learning_rate": 1.7508777957504662e-07,
+      "loss": 0.9013,
+      "step": 11909
+    },
+    {
+      "epoch": 0.9181313598519889,
+      "grad_norm": 3.746781826019287,
+      "learning_rate": 1.7476044420112637e-07,
+      "loss": 0.9299,
+      "step": 11910
+    },
+    {
+      "epoch": 0.9182084489670058,
+      "grad_norm": 3.875746250152588,
+      "learning_rate": 1.744334096559025e-07,
+      "loss": 0.8717,
+      "step": 11911
+    },
+    {
+      "epoch": 0.9182855380820228,
+      "grad_norm": 4.113644123077393,
+      "learning_rate": 1.741066759597626e-07,
+      "loss": 1.0246,
+      "step": 11912
+    },
+    {
+      "epoch": 0.9183626271970398,
+      "grad_norm": 3.610800266265869,
+      "learning_rate": 1.7378024313307763e-07,
+      "loss": 0.9421,
+      "step": 11913
+    },
+    {
+      "epoch": 0.9184397163120568,
+      "grad_norm": 4.1041107177734375,
+      "learning_rate": 1.7345411119619847e-07,
+      "loss": 0.8065,
+      "step": 11914
+    },
+    {
+      "epoch": 0.9185168054270737,
+      "grad_norm": 3.635951519012451,
+      "learning_rate": 1.7312828016945836e-07,
+      "loss": 0.9826,
+      "step": 11915
+    },
+    {
+      "epoch": 0.9185938945420906,
+      "grad_norm": 3.9342377185821533,
+      "learning_rate": 1.728027500731716e-07,
+      "loss": 0.981,
+      "step": 11916
+    },
+    {
+      "epoch": 0.9186709836571076,
+      "grad_norm": 3.4716575145721436,
+      "learning_rate": 1.7247752092763247e-07,
+      "loss": 0.9107,
+      "step": 11917
+    },
+    {
+      "epoch": 0.9187480727721246,
+      "grad_norm": 3.786710262298584,
+      "learning_rate": 1.7215259275311703e-07,
+      "loss": 0.9527,
+      "step": 11918
+    },
+    {
+      "epoch": 0.9188251618871416,
+      "grad_norm": 3.892406940460205,
+      "learning_rate": 1.71827965569884e-07,
+      "loss": 0.7812,
+      "step": 11919
+    },
+    {
+      "epoch": 0.9189022510021585,
+      "grad_norm": 3.8469479084014893,
+      "learning_rate": 1.7150363939817117e-07,
+      "loss": 0.919,
+      "step": 11920
+    },
+    {
+      "epoch": 0.9189793401171754,
+      "grad_norm": 3.4932711124420166,
+      "learning_rate": 1.7117961425819897e-07,
+      "loss": 0.9501,
+      "step": 11921
+    },
+    {
+      "epoch": 0.9190564292321924,
+      "grad_norm": 3.6135263442993164,
+      "learning_rate": 1.70855890170169e-07,
+      "loss": 0.8402,
+      "step": 11922
+    },
+    {
+      "epoch": 0.9191335183472094,
+      "grad_norm": 4.184671878814697,
+      "learning_rate": 1.7053246715426297e-07,
+      "loss": 0.8431,
+      "step": 11923
+    },
+    {
+      "epoch": 0.9192106074622264,
+      "grad_norm": 3.7733452320098877,
+      "learning_rate": 1.702093452306458e-07,
+      "loss": 0.9064,
+      "step": 11924
+    },
+    {
+      "epoch": 0.9192876965772433,
+      "grad_norm": 3.719705820083618,
+      "learning_rate": 1.698865244194614e-07,
+      "loss": 0.8943,
+      "step": 11925
+    },
+    {
+      "epoch": 0.9193647856922602,
+      "grad_norm": 3.8758575916290283,
+      "learning_rate": 1.6956400474083644e-07,
+      "loss": 0.9384,
+      "step": 11926
+    },
+    {
+      "epoch": 0.9194418748072772,
+      "grad_norm": 3.6898250579833984,
+      "learning_rate": 1.6924178621487875e-07,
+      "loss": 0.8039,
+      "step": 11927
+    },
+    {
+      "epoch": 0.9195189639222942,
+      "grad_norm": 3.731034755706787,
+      "learning_rate": 1.689198688616761e-07,
+      "loss": 0.9268,
+      "step": 11928
+    },
+    {
+      "epoch": 0.9195960530373112,
+      "grad_norm": 3.5098345279693604,
+      "learning_rate": 1.685982527012986e-07,
+      "loss": 0.8259,
+      "step": 11929
+    },
+    {
+      "epoch": 0.9196731421523281,
+      "grad_norm": 3.9864370822906494,
+      "learning_rate": 1.6827693775379794e-07,
+      "loss": 0.8976,
+      "step": 11930
+    },
+    {
+      "epoch": 0.919750231267345,
+      "grad_norm": 3.444135904312134,
+      "learning_rate": 1.6795592403920591e-07,
+      "loss": 0.8563,
+      "step": 11931
+    },
+    {
+      "epoch": 0.919827320382362,
+      "grad_norm": 3.5715203285217285,
+      "learning_rate": 1.6763521157753647e-07,
+      "loss": 0.8746,
+      "step": 11932
+    },
+    {
+      "epoch": 0.919904409497379,
+      "grad_norm": 4.088859558105469,
+      "learning_rate": 1.6731480038878368e-07,
+      "loss": 0.9353,
+      "step": 11933
+    },
+    {
+      "epoch": 0.919981498612396,
+      "grad_norm": 3.915494918823242,
+      "learning_rate": 1.6699469049292427e-07,
+      "loss": 0.946,
+      "step": 11934
+    },
+    {
+      "epoch": 0.9200585877274129,
+      "grad_norm": 3.923362970352173,
+      "learning_rate": 1.666748819099151e-07,
+      "loss": 0.9018,
+      "step": 11935
+    },
+    {
+      "epoch": 0.9201356768424298,
+      "grad_norm": 3.7624075412750244,
+      "learning_rate": 1.6635537465969463e-07,
+      "loss": 0.8156,
+      "step": 11936
+    },
+    {
+      "epoch": 0.9202127659574468,
+      "grad_norm": 3.9974863529205322,
+      "learning_rate": 1.6603616876218308e-07,
+      "loss": 0.8917,
+      "step": 11937
+    },
+    {
+      "epoch": 0.9202898550724637,
+      "grad_norm": 4.101132392883301,
+      "learning_rate": 1.6571726423727953e-07,
+      "loss": 0.9253,
+      "step": 11938
+    },
+    {
+      "epoch": 0.9203669441874808,
+      "grad_norm": 4.09736442565918,
+      "learning_rate": 1.6539866110486802e-07,
+      "loss": 0.9177,
+      "step": 11939
+    },
+    {
+      "epoch": 0.9204440333024977,
+      "grad_norm": 3.8575503826141357,
+      "learning_rate": 1.6508035938481048e-07,
+      "loss": 0.9692,
+      "step": 11940
+    },
+    {
+      "epoch": 0.9205211224175146,
+      "grad_norm": 3.686838150024414,
+      "learning_rate": 1.6476235909695158e-07,
+      "loss": 0.9461,
+      "step": 11941
+    },
+    {
+      "epoch": 0.9205982115325316,
+      "grad_norm": 3.914886236190796,
+      "learning_rate": 1.6444466026111826e-07,
+      "loss": 0.9885,
+      "step": 11942
+    },
+    {
+      "epoch": 0.9206753006475485,
+      "grad_norm": 4.105428695678711,
+      "learning_rate": 1.6412726289711578e-07,
+      "loss": 0.9431,
+      "step": 11943
+    },
+    {
+      "epoch": 0.9207523897625656,
+      "grad_norm": 6.005154132843018,
+      "learning_rate": 1.6381016702473273e-07,
+      "loss": 0.7875,
+      "step": 11944
+    },
+    {
+      "epoch": 0.9208294788775825,
+      "grad_norm": 4.297434329986572,
+      "learning_rate": 1.6349337266373832e-07,
+      "loss": 1.0278,
+      "step": 11945
+    },
+    {
+      "epoch": 0.9209065679925994,
+      "grad_norm": 3.489044189453125,
+      "learning_rate": 1.631768798338834e-07,
+      "loss": 0.828,
+      "step": 11946
+    },
+    {
+      "epoch": 0.9209836571076164,
+      "grad_norm": 3.5091493129730225,
+      "learning_rate": 1.6286068855489946e-07,
+      "loss": 0.804,
+      "step": 11947
+    },
+    {
+      "epoch": 0.9210607462226333,
+      "grad_norm": 3.7728817462921143,
+      "learning_rate": 1.625447988465001e-07,
+      "loss": 0.9729,
+      "step": 11948
+    },
+    {
+      "epoch": 0.9211378353376504,
+      "grad_norm": 3.835939884185791,
+      "learning_rate": 1.6222921072837794e-07,
+      "loss": 0.9318,
+      "step": 11949
+    },
+    {
+      "epoch": 0.9212149244526673,
+      "grad_norm": 3.666374921798706,
+      "learning_rate": 1.6191392422020892e-07,
+      "loss": 0.922,
+      "step": 11950
+    },
+    {
+      "epoch": 0.9212920135676842,
+      "grad_norm": 3.756239652633667,
+      "learning_rate": 1.6159893934165006e-07,
+      "loss": 0.9781,
+      "step": 11951
+    },
+    {
+      "epoch": 0.9213691026827012,
+      "grad_norm": 3.4061012268066406,
+      "learning_rate": 1.6128425611233844e-07,
+      "loss": 0.7702,
+      "step": 11952
+    },
+    {
+      "epoch": 0.9214461917977181,
+      "grad_norm": 3.5357954502105713,
+      "learning_rate": 1.6096987455189338e-07,
+      "loss": 0.8715,
+      "step": 11953
+    },
+    {
+      "epoch": 0.9215232809127352,
+      "grad_norm": 4.1103596687316895,
+      "learning_rate": 1.6065579467991422e-07,
+      "loss": 0.8386,
+      "step": 11954
+    },
+    {
+      "epoch": 0.9216003700277521,
+      "grad_norm": 3.631218433380127,
+      "learning_rate": 1.6034201651598304e-07,
+      "loss": 0.7605,
+      "step": 11955
+    },
+    {
+      "epoch": 0.921677459142769,
+      "grad_norm": 3.968055248260498,
+      "learning_rate": 1.600285400796614e-07,
+      "loss": 1.0534,
+      "step": 11956
+    },
+    {
+      "epoch": 0.921754548257786,
+      "grad_norm": 4.089054107666016,
+      "learning_rate": 1.597153653904937e-07,
+      "loss": 0.9012,
+      "step": 11957
+    },
+    {
+      "epoch": 0.9218316373728029,
+      "grad_norm": 3.858142375946045,
+      "learning_rate": 1.594024924680043e-07,
+      "loss": 0.8826,
+      "step": 11958
+    },
+    {
+      "epoch": 0.92190872648782,
+      "grad_norm": 3.8919715881347656,
+      "learning_rate": 1.5908992133170041e-07,
+      "loss": 0.7672,
+      "step": 11959
+    },
+    {
+      "epoch": 0.9219858156028369,
+      "grad_norm": 4.094879150390625,
+      "learning_rate": 1.5877765200106697e-07,
+      "loss": 1.0555,
+      "step": 11960
+    },
+    {
+      "epoch": 0.9220629047178538,
+      "grad_norm": 3.7169480323791504,
+      "learning_rate": 1.5846568449557397e-07,
+      "loss": 0.7915,
+      "step": 11961
+    },
+    {
+      "epoch": 0.9221399938328708,
+      "grad_norm": 3.7812538146972656,
+      "learning_rate": 1.5815401883467086e-07,
+      "loss": 0.8998,
+      "step": 11962
+    },
+    {
+      "epoch": 0.9222170829478877,
+      "grad_norm": 3.6201839447021484,
+      "learning_rate": 1.578426550377876e-07,
+      "loss": 0.8315,
+      "step": 11963
+    },
+    {
+      "epoch": 0.9222941720629048,
+      "grad_norm": 3.6739892959594727,
+      "learning_rate": 1.5753159312433762e-07,
+      "loss": 0.9893,
+      "step": 11964
+    },
+    {
+      "epoch": 0.9223712611779217,
+      "grad_norm": 3.594050884246826,
+      "learning_rate": 1.5722083311371206e-07,
+      "loss": 0.8761,
+      "step": 11965
+    },
+    {
+      "epoch": 0.9224483502929386,
+      "grad_norm": 3.51835298538208,
+      "learning_rate": 1.569103750252865e-07,
+      "loss": 0.9165,
+      "step": 11966
+    },
+    {
+      "epoch": 0.9225254394079556,
+      "grad_norm": 3.550865650177002,
+      "learning_rate": 1.5660021887841603e-07,
+      "loss": 0.8594,
+      "step": 11967
+    },
+    {
+      "epoch": 0.9226025285229725,
+      "grad_norm": 3.8896725177764893,
+      "learning_rate": 1.5629036469243686e-07,
+      "loss": 0.8592,
+      "step": 11968
+    },
+    {
+      "epoch": 0.9226796176379896,
+      "grad_norm": 3.8969407081604004,
+      "learning_rate": 1.5598081248666742e-07,
+      "loss": 0.8356,
+      "step": 11969
+    },
+    {
+      "epoch": 0.9227567067530065,
+      "grad_norm": 3.598137378692627,
+      "learning_rate": 1.5567156228040726e-07,
+      "loss": 0.8937,
+      "step": 11970
+    },
+    {
+      "epoch": 0.9228337958680234,
+      "grad_norm": 3.7268011569976807,
+      "learning_rate": 1.553626140929354e-07,
+      "loss": 0.9288,
+      "step": 11971
+    },
+    {
+      "epoch": 0.9229108849830404,
+      "grad_norm": 3.700021266937256,
+      "learning_rate": 1.5505396794351312e-07,
+      "loss": 0.8649,
+      "step": 11972
+    },
+    {
+      "epoch": 0.9229879740980573,
+      "grad_norm": 3.5971262454986572,
+      "learning_rate": 1.5474562385138337e-07,
+      "loss": 0.8113,
+      "step": 11973
+    },
+    {
+      "epoch": 0.9230650632130744,
+      "grad_norm": 5.08428955078125,
+      "learning_rate": 1.5443758183576962e-07,
+      "loss": 0.9799,
+      "step": 11974
+    },
+    {
+      "epoch": 0.9231421523280913,
+      "grad_norm": 4.136900424957275,
+      "learning_rate": 1.5412984191587766e-07,
+      "loss": 0.9596,
+      "step": 11975
+    },
+    {
+      "epoch": 0.9232192414431082,
+      "grad_norm": 3.8938400745391846,
+      "learning_rate": 1.5382240411089156e-07,
+      "loss": 0.9409,
+      "step": 11976
+    },
+    {
+      "epoch": 0.9232963305581252,
+      "grad_norm": 4.196589469909668,
+      "learning_rate": 1.5351526843997988e-07,
+      "loss": 0.96,
+      "step": 11977
+    },
+    {
+      "epoch": 0.9233734196731421,
+      "grad_norm": 4.031776428222656,
+      "learning_rate": 1.532084349222912e-07,
+      "loss": 0.9206,
+      "step": 11978
+    },
+    {
+      "epoch": 0.9234505087881592,
+      "grad_norm": 3.581867218017578,
+      "learning_rate": 1.5290190357695466e-07,
+      "loss": 0.8353,
+      "step": 11979
+    },
+    {
+      "epoch": 0.9235275979031761,
+      "grad_norm": 3.6931517124176025,
+      "learning_rate": 1.5259567442308e-07,
+      "loss": 0.7995,
+      "step": 11980
+    },
+    {
+      "epoch": 0.923604687018193,
+      "grad_norm": 3.691195011138916,
+      "learning_rate": 1.5228974747975965e-07,
+      "loss": 0.8052,
+      "step": 11981
+    },
+    {
+      "epoch": 0.92368177613321,
+      "grad_norm": 3.95929217338562,
+      "learning_rate": 1.5198412276606622e-07,
+      "loss": 0.9099,
+      "step": 11982
+    },
+    {
+      "epoch": 0.9237588652482269,
+      "grad_norm": 3.7074384689331055,
+      "learning_rate": 1.5167880030105497e-07,
+      "loss": 0.8664,
+      "step": 11983
+    },
+    {
+      "epoch": 0.923835954363244,
+      "grad_norm": 4.384084224700928,
+      "learning_rate": 1.5137378010376015e-07,
+      "loss": 1.0385,
+      "step": 11984
+    },
+    {
+      "epoch": 0.9239130434782609,
+      "grad_norm": 3.809328556060791,
+      "learning_rate": 1.5106906219319871e-07,
+      "loss": 0.8989,
+      "step": 11985
+    },
+    {
+      "epoch": 0.9239901325932778,
+      "grad_norm": 3.73268461227417,
+      "learning_rate": 1.5076464658836775e-07,
+      "loss": 0.9453,
+      "step": 11986
+    },
+    {
+      "epoch": 0.9240672217082948,
+      "grad_norm": 4.032083988189697,
+      "learning_rate": 1.504605333082465e-07,
+      "loss": 0.9235,
+      "step": 11987
+    },
+    {
+      "epoch": 0.9241443108233117,
+      "grad_norm": 3.3622939586639404,
+      "learning_rate": 1.5015672237179424e-07,
+      "loss": 0.8173,
+      "step": 11988
+    },
+    {
+      "epoch": 0.9242213999383287,
+      "grad_norm": 3.6310486793518066,
+      "learning_rate": 1.498532137979525e-07,
+      "loss": 0.9421,
+      "step": 11989
+    },
+    {
+      "epoch": 0.9242984890533457,
+      "grad_norm": 3.632511615753174,
+      "learning_rate": 1.495500076056433e-07,
+      "loss": 0.8727,
+      "step": 11990
+    },
+    {
+      "epoch": 0.9243755781683626,
+      "grad_norm": 4.108543872833252,
+      "learning_rate": 1.4924710381376995e-07,
+      "loss": 0.984,
+      "step": 11991
+    },
+    {
+      "epoch": 0.9244526672833796,
+      "grad_norm": 3.7871105670928955,
+      "learning_rate": 1.4894450244121727e-07,
+      "loss": 0.8612,
+      "step": 11992
+    },
+    {
+      "epoch": 0.9245297563983965,
+      "grad_norm": 3.8213963508605957,
+      "learning_rate": 1.486422035068502e-07,
+      "loss": 1.1303,
+      "step": 11993
+    },
+    {
+      "epoch": 0.9246068455134135,
+      "grad_norm": 3.63419508934021,
+      "learning_rate": 1.483402070295159e-07,
+      "loss": 0.804,
+      "step": 11994
+    },
+    {
+      "epoch": 0.9246839346284305,
+      "grad_norm": 3.4561123847961426,
+      "learning_rate": 1.4803851302804261e-07,
+      "loss": 0.9273,
+      "step": 11995
+    },
+    {
+      "epoch": 0.9247610237434474,
+      "grad_norm": 3.9428532123565674,
+      "learning_rate": 1.477371215212392e-07,
+      "loss": 0.9112,
+      "step": 11996
+    },
+    {
+      "epoch": 0.9248381128584644,
+      "grad_norm": 3.147937774658203,
+      "learning_rate": 1.474360325278956e-07,
+      "loss": 0.8637,
+      "step": 11997
+    },
+    {
+      "epoch": 0.9249152019734813,
+      "grad_norm": 3.6595137119293213,
+      "learning_rate": 1.4713524606678298e-07,
+      "loss": 0.9017,
+      "step": 11998
+    },
+    {
+      "epoch": 0.9249922910884983,
+      "grad_norm": 4.111579895019531,
+      "learning_rate": 1.4683476215665405e-07,
+      "loss": 1.0554,
+      "step": 11999
+    },
+    {
+      "epoch": 0.9250693802035153,
+      "grad_norm": 3.70733642578125,
+      "learning_rate": 1.465345808162427e-07,
+      "loss": 0.9537,
+      "step": 12000
+    },
+    {
+      "epoch": 0.9251464693185322,
+      "grad_norm": 3.986865758895874,
+      "learning_rate": 1.4623470206426404e-07,
+      "loss": 1.0235,
+      "step": 12001
+    },
+    {
+      "epoch": 0.9252235584335492,
+      "grad_norm": 3.7944588661193848,
+      "learning_rate": 1.4593512591941304e-07,
+      "loss": 0.8312,
+      "step": 12002
+    },
+    {
+      "epoch": 0.9253006475485661,
+      "grad_norm": 3.5245468616485596,
+      "learning_rate": 1.4563585240036705e-07,
+      "loss": 0.8643,
+      "step": 12003
+    },
+    {
+      "epoch": 0.9253777366635831,
+      "grad_norm": 3.854820966720581,
+      "learning_rate": 1.4533688152578384e-07,
+      "loss": 0.8276,
+      "step": 12004
+    },
+    {
+      "epoch": 0.9254548257786,
+      "grad_norm": 4.025414943695068,
+      "learning_rate": 1.4503821331430358e-07,
+      "loss": 0.859,
+      "step": 12005
+    },
+    {
+      "epoch": 0.925531914893617,
+      "grad_norm": 3.724771499633789,
+      "learning_rate": 1.447398477845463e-07,
+      "loss": 0.8583,
+      "step": 12006
+    },
+    {
+      "epoch": 0.925609004008634,
+      "grad_norm": 3.5540995597839355,
+      "learning_rate": 1.444417849551133e-07,
+      "loss": 0.871,
+      "step": 12007
+    },
+    {
+      "epoch": 0.9256860931236509,
+      "grad_norm": 3.8475286960601807,
+      "learning_rate": 1.4414402484458746e-07,
+      "loss": 0.9248,
+      "step": 12008
+    },
+    {
+      "epoch": 0.9257631822386679,
+      "grad_norm": 3.745168924331665,
+      "learning_rate": 1.438465674715328e-07,
+      "loss": 0.9334,
+      "step": 12009
+    },
+    {
+      "epoch": 0.9258402713536849,
+      "grad_norm": 4.050620079040527,
+      "learning_rate": 1.4354941285449342e-07,
+      "loss": 0.9622,
+      "step": 12010
+    },
+    {
+      "epoch": 0.9259173604687018,
+      "grad_norm": 3.5585291385650635,
+      "learning_rate": 1.4325256101199615e-07,
+      "loss": 0.9217,
+      "step": 12011
+    },
+    {
+      "epoch": 0.9259944495837188,
+      "grad_norm": 3.585474729537964,
+      "learning_rate": 1.4295601196254838e-07,
+      "loss": 0.8582,
+      "step": 12012
+    },
+    {
+      "epoch": 0.9260715386987357,
+      "grad_norm": 3.88104510307312,
+      "learning_rate": 1.4265976572463815e-07,
+      "loss": 0.9336,
+      "step": 12013
+    },
+    {
+      "epoch": 0.9261486278137527,
+      "grad_norm": 3.3760218620300293,
+      "learning_rate": 1.4236382231673395e-07,
+      "loss": 0.8343,
+      "step": 12014
+    },
+    {
+      "epoch": 0.9262257169287696,
+      "grad_norm": 3.9959182739257812,
+      "learning_rate": 1.420681817572872e-07,
+      "loss": 0.9329,
+      "step": 12015
+    },
+    {
+      "epoch": 0.9263028060437866,
+      "grad_norm": 3.396411895751953,
+      "learning_rate": 1.417728440647298e-07,
+      "loss": 0.7385,
+      "step": 12016
+    },
+    {
+      "epoch": 0.9263798951588036,
+      "grad_norm": 3.7224302291870117,
+      "learning_rate": 1.4147780925747368e-07,
+      "loss": 0.8898,
+      "step": 12017
+    },
+    {
+      "epoch": 0.9264569842738205,
+      "grad_norm": 3.9120421409606934,
+      "learning_rate": 1.4118307735391412e-07,
+      "loss": 0.8889,
+      "step": 12018
+    },
+    {
+      "epoch": 0.9265340733888375,
+      "grad_norm": 3.755709171295166,
+      "learning_rate": 1.4088864837242422e-07,
+      "loss": 0.9396,
+      "step": 12019
+    },
+    {
+      "epoch": 0.9266111625038544,
+      "grad_norm": 3.8816263675689697,
+      "learning_rate": 1.4059452233136094e-07,
+      "loss": 0.8704,
+      "step": 12020
+    },
+    {
+      "epoch": 0.9266882516188714,
+      "grad_norm": 3.5586330890655518,
+      "learning_rate": 1.4030069924906241e-07,
+      "loss": 0.8279,
+      "step": 12021
+    },
+    {
+      "epoch": 0.9267653407338884,
+      "grad_norm": 3.6975462436676025,
+      "learning_rate": 1.4000717914384677e-07,
+      "loss": 0.9152,
+      "step": 12022
+    },
+    {
+      "epoch": 0.9268424298489053,
+      "grad_norm": 3.672626256942749,
+      "learning_rate": 1.397139620340121e-07,
+      "loss": 0.8989,
+      "step": 12023
+    },
+    {
+      "epoch": 0.9269195189639223,
+      "grad_norm": 3.6923696994781494,
+      "learning_rate": 1.3942104793783996e-07,
+      "loss": 0.8973,
+      "step": 12024
+    },
+    {
+      "epoch": 0.9269966080789392,
+      "grad_norm": 3.7403361797332764,
+      "learning_rate": 1.391284368735918e-07,
+      "loss": 0.8737,
+      "step": 12025
+    },
+    {
+      "epoch": 0.9270736971939562,
+      "grad_norm": 4.060670852661133,
+      "learning_rate": 1.388361288595108e-07,
+      "loss": 0.9774,
+      "step": 12026
+    },
+    {
+      "epoch": 0.9271507863089732,
+      "grad_norm": 3.923124313354492,
+      "learning_rate": 1.3854412391382078e-07,
+      "loss": 0.9751,
+      "step": 12027
+    },
+    {
+      "epoch": 0.9272278754239901,
+      "grad_norm": 3.894160032272339,
+      "learning_rate": 1.3825242205472599e-07,
+      "loss": 0.919,
+      "step": 12028
+    },
+    {
+      "epoch": 0.9273049645390071,
+      "grad_norm": 3.898775815963745,
+      "learning_rate": 1.3796102330041305e-07,
+      "loss": 0.9102,
+      "step": 12029
+    },
+    {
+      "epoch": 0.927382053654024,
+      "grad_norm": 3.7741925716400146,
+      "learning_rate": 1.3766992766904908e-07,
+      "loss": 0.7993,
+      "step": 12030
+    },
+    {
+      "epoch": 0.927459142769041,
+      "grad_norm": 3.740205764770508,
+      "learning_rate": 1.3737913517878286e-07,
+      "loss": 0.9111,
+      "step": 12031
+    },
+    {
+      "epoch": 0.927536231884058,
+      "grad_norm": 3.5820446014404297,
+      "learning_rate": 1.370886458477433e-07,
+      "loss": 0.8983,
+      "step": 12032
+    },
+    {
+      "epoch": 0.9276133209990749,
+      "grad_norm": 3.634716510772705,
+      "learning_rate": 1.3679845969404138e-07,
+      "loss": 0.7987,
+      "step": 12033
+    },
+    {
+      "epoch": 0.9276904101140919,
+      "grad_norm": 3.8393595218658447,
+      "learning_rate": 1.3650857673576767e-07,
+      "loss": 0.8793,
+      "step": 12034
+    },
+    {
+      "epoch": 0.9277674992291088,
+      "grad_norm": 3.9487619400024414,
+      "learning_rate": 1.362189969909955e-07,
+      "loss": 0.8779,
+      "step": 12035
+    },
+    {
+      "epoch": 0.9278445883441259,
+      "grad_norm": 3.9395010471343994,
+      "learning_rate": 1.3592972047777874e-07,
+      "loss": 0.9886,
+      "step": 12036
+    },
+    {
+      "epoch": 0.9279216774591428,
+      "grad_norm": 3.642089366912842,
+      "learning_rate": 1.3564074721415243e-07,
+      "loss": 0.8256,
+      "step": 12037
+    },
+    {
+      "epoch": 0.9279987665741597,
+      "grad_norm": 3.643852472305298,
+      "learning_rate": 1.3535207721813327e-07,
+      "loss": 0.8572,
+      "step": 12038
+    },
+    {
+      "epoch": 0.9280758556891767,
+      "grad_norm": 3.6606554985046387,
+      "learning_rate": 1.3506371050771626e-07,
+      "loss": 0.9307,
+      "step": 12039
+    },
+    {
+      "epoch": 0.9281529448041936,
+      "grad_norm": 3.6610138416290283,
+      "learning_rate": 1.3477564710088097e-07,
+      "loss": 0.8941,
+      "step": 12040
+    },
+    {
+      "epoch": 0.9282300339192107,
+      "grad_norm": 3.769508123397827,
+      "learning_rate": 1.3448788701558635e-07,
+      "loss": 0.8769,
+      "step": 12041
+    },
+    {
+      "epoch": 0.9283071230342276,
+      "grad_norm": 3.8001794815063477,
+      "learning_rate": 1.3420043026977303e-07,
+      "loss": 0.9133,
+      "step": 12042
+    },
+    {
+      "epoch": 0.9283842121492445,
+      "grad_norm": 3.7807233333587646,
+      "learning_rate": 1.3391327688136225e-07,
+      "loss": 0.9274,
+      "step": 12043
+    },
+    {
+      "epoch": 0.9284613012642615,
+      "grad_norm": 3.714263677597046,
+      "learning_rate": 1.3362642686825688e-07,
+      "loss": 0.8069,
+      "step": 12044
+    },
+    {
+      "epoch": 0.9285383903792784,
+      "grad_norm": 3.7317070960998535,
+      "learning_rate": 1.3333988024833989e-07,
+      "loss": 0.9593,
+      "step": 12045
+    },
+    {
+      "epoch": 0.9286154794942955,
+      "grad_norm": 3.5440969467163086,
+      "learning_rate": 1.3305363703947581e-07,
+      "loss": 0.8853,
+      "step": 12046
+    },
+    {
+      "epoch": 0.9286925686093124,
+      "grad_norm": 3.542323112487793,
+      "learning_rate": 1.3276769725951155e-07,
+      "loss": 0.8715,
+      "step": 12047
+    },
+    {
+      "epoch": 0.9287696577243293,
+      "grad_norm": 3.7308967113494873,
+      "learning_rate": 1.3248206092627281e-07,
+      "loss": 0.8932,
+      "step": 12048
+    },
+    {
+      "epoch": 0.9288467468393463,
+      "grad_norm": 3.898761034011841,
+      "learning_rate": 1.3219672805756922e-07,
+      "loss": 0.9775,
+      "step": 12049
+    },
+    {
+      "epoch": 0.9289238359543632,
+      "grad_norm": 3.488557815551758,
+      "learning_rate": 1.319116986711877e-07,
+      "loss": 0.9247,
+      "step": 12050
+    },
+    {
+      "epoch": 0.9290009250693803,
+      "grad_norm": 4.142662048339844,
+      "learning_rate": 1.31626972784899e-07,
+      "loss": 0.9129,
+      "step": 12051
+    },
+    {
+      "epoch": 0.9290780141843972,
+      "grad_norm": 3.6546530723571777,
+      "learning_rate": 1.3134255041645505e-07,
+      "loss": 0.8448,
+      "step": 12052
+    },
+    {
+      "epoch": 0.9291551032994141,
+      "grad_norm": 3.8783202171325684,
+      "learning_rate": 1.310584315835872e-07,
+      "loss": 0.938,
+      "step": 12053
+    },
+    {
+      "epoch": 0.9292321924144311,
+      "grad_norm": 3.5673296451568604,
+      "learning_rate": 1.3077461630400967e-07,
+      "loss": 0.9903,
+      "step": 12054
+    },
+    {
+      "epoch": 0.929309281529448,
+      "grad_norm": 3.872450590133667,
+      "learning_rate": 1.3049110459541658e-07,
+      "loss": 0.8605,
+      "step": 12055
+    },
+    {
+      "epoch": 0.929386370644465,
+      "grad_norm": 3.5778377056121826,
+      "learning_rate": 1.3020789647548326e-07,
+      "loss": 0.8794,
+      "step": 12056
+    },
+    {
+      "epoch": 0.929463459759482,
+      "grad_norm": 3.7152254581451416,
+      "learning_rate": 1.2992499196186614e-07,
+      "loss": 0.8878,
+      "step": 12057
+    },
+    {
+      "epoch": 0.9295405488744989,
+      "grad_norm": 3.6128242015838623,
+      "learning_rate": 1.296423910722028e-07,
+      "loss": 0.9107,
+      "step": 12058
+    },
+    {
+      "epoch": 0.9296176379895159,
+      "grad_norm": 3.4194366931915283,
+      "learning_rate": 1.293600938241124e-07,
+      "loss": 0.8509,
+      "step": 12059
+    },
+    {
+      "epoch": 0.9296947271045328,
+      "grad_norm": 3.806325912475586,
+      "learning_rate": 1.2907810023519485e-07,
+      "loss": 0.8803,
+      "step": 12060
+    },
+    {
+      "epoch": 0.9297718162195499,
+      "grad_norm": 3.6678504943847656,
+      "learning_rate": 1.2879641032302993e-07,
+      "loss": 0.9138,
+      "step": 12061
+    },
+    {
+      "epoch": 0.9298489053345668,
+      "grad_norm": 3.877365827560425,
+      "learning_rate": 1.2851502410518025e-07,
+      "loss": 0.9103,
+      "step": 12062
+    },
+    {
+      "epoch": 0.9299259944495837,
+      "grad_norm": 3.8869688510894775,
+      "learning_rate": 1.28233941599189e-07,
+      "loss": 0.9129,
+      "step": 12063
+    },
+    {
+      "epoch": 0.9300030835646007,
+      "grad_norm": 4.313940525054932,
+      "learning_rate": 1.279531628225794e-07,
+      "loss": 0.8989,
+      "step": 12064
+    },
+    {
+      "epoch": 0.9300801726796176,
+      "grad_norm": 3.7056541442871094,
+      "learning_rate": 1.2767268779285803e-07,
+      "loss": 0.8635,
+      "step": 12065
+    },
+    {
+      "epoch": 0.9301572617946346,
+      "grad_norm": 3.957458019256592,
+      "learning_rate": 1.2739251652750917e-07,
+      "loss": 0.8531,
+      "step": 12066
+    },
+    {
+      "epoch": 0.9302343509096516,
+      "grad_norm": 4.063552379608154,
+      "learning_rate": 1.2711264904400167e-07,
+      "loss": 1.0151,
+      "step": 12067
+    },
+    {
+      "epoch": 0.9303114400246685,
+      "grad_norm": 4.064858436584473,
+      "learning_rate": 1.2683308535978323e-07,
+      "loss": 0.9269,
+      "step": 12068
+    },
+    {
+      "epoch": 0.9303885291396855,
+      "grad_norm": 3.7394871711730957,
+      "learning_rate": 1.2655382549228267e-07,
+      "loss": 0.8913,
+      "step": 12069
+    },
+    {
+      "epoch": 0.9304656182547024,
+      "grad_norm": 3.8202192783355713,
+      "learning_rate": 1.2627486945891166e-07,
+      "loss": 0.9727,
+      "step": 12070
+    },
+    {
+      "epoch": 0.9305427073697194,
+      "grad_norm": 3.5857367515563965,
+      "learning_rate": 1.2599621727706014e-07,
+      "loss": 0.9116,
+      "step": 12071
+    },
+    {
+      "epoch": 0.9306197964847364,
+      "grad_norm": 3.656193494796753,
+      "learning_rate": 1.2571786896410144e-07,
+      "loss": 0.8853,
+      "step": 12072
+    },
+    {
+      "epoch": 0.9306968855997533,
+      "grad_norm": 3.4861977100372314,
+      "learning_rate": 1.2543982453738945e-07,
+      "loss": 0.8542,
+      "step": 12073
+    },
+    {
+      "epoch": 0.9307739747147703,
+      "grad_norm": 3.6881868839263916,
+      "learning_rate": 1.251620840142581e-07,
+      "loss": 0.9306,
+      "step": 12074
+    },
+    {
+      "epoch": 0.9308510638297872,
+      "grad_norm": 4.058561325073242,
+      "learning_rate": 1.2488464741202355e-07,
+      "loss": 0.9352,
+      "step": 12075
+    },
+    {
+      "epoch": 0.9309281529448042,
+      "grad_norm": 4.049665927886963,
+      "learning_rate": 1.2460751474798249e-07,
+      "loss": 1.0101,
+      "step": 12076
+    },
+    {
+      "epoch": 0.9310052420598212,
+      "grad_norm": 3.6814520359039307,
+      "learning_rate": 1.2433068603941223e-07,
+      "loss": 0.881,
+      "step": 12077
+    },
+    {
+      "epoch": 0.9310823311748381,
+      "grad_norm": 4.561155796051025,
+      "learning_rate": 1.2405416130357172e-07,
+      "loss": 0.9776,
+      "step": 12078
+    },
+    {
+      "epoch": 0.9311594202898551,
+      "grad_norm": 4.181887149810791,
+      "learning_rate": 1.2377794055770164e-07,
+      "loss": 0.8847,
+      "step": 12079
+    },
+    {
+      "epoch": 0.931236509404872,
+      "grad_norm": 3.861226797103882,
+      "learning_rate": 1.235020238190221e-07,
+      "loss": 1.0193,
+      "step": 12080
+    },
+    {
+      "epoch": 0.931313598519889,
+      "grad_norm": 3.324911594390869,
+      "learning_rate": 1.23226411104736e-07,
+      "loss": 0.8281,
+      "step": 12081
+    },
+    {
+      "epoch": 0.931390687634906,
+      "grad_norm": 3.70337176322937,
+      "learning_rate": 1.2295110243202458e-07,
+      "loss": 0.9571,
+      "step": 12082
+    },
+    {
+      "epoch": 0.9314677767499229,
+      "grad_norm": 3.8327345848083496,
+      "learning_rate": 1.2267609781805356e-07,
+      "loss": 0.8885,
+      "step": 12083
+    },
+    {
+      "epoch": 0.9315448658649399,
+      "grad_norm": 3.6880276203155518,
+      "learning_rate": 1.2240139727996757e-07,
+      "loss": 0.8384,
+      "step": 12084
+    },
+    {
+      "epoch": 0.9316219549799568,
+      "grad_norm": 3.634875774383545,
+      "learning_rate": 1.2212700083489236e-07,
+      "loss": 0.9202,
+      "step": 12085
+    },
+    {
+      "epoch": 0.9316990440949738,
+      "grad_norm": 3.5088517665863037,
+      "learning_rate": 1.2185290849993648e-07,
+      "loss": 0.8589,
+      "step": 12086
+    },
+    {
+      "epoch": 0.9317761332099908,
+      "grad_norm": 3.7651207447052,
+      "learning_rate": 1.2157912029218676e-07,
+      "loss": 0.899,
+      "step": 12087
+    },
+    {
+      "epoch": 0.9318532223250077,
+      "grad_norm": 3.844053030014038,
+      "learning_rate": 1.2130563622871239e-07,
+      "loss": 0.913,
+      "step": 12088
+    },
+    {
+      "epoch": 0.9319303114400247,
+      "grad_norm": 3.8823323249816895,
+      "learning_rate": 1.2103245632656414e-07,
+      "loss": 0.9003,
+      "step": 12089
+    },
+    {
+      "epoch": 0.9320074005550416,
+      "grad_norm": 3.5348687171936035,
+      "learning_rate": 1.2075958060277394e-07,
+      "loss": 0.8039,
+      "step": 12090
+    },
+    {
+      "epoch": 0.9320844896700586,
+      "grad_norm": 3.1809444427490234,
+      "learning_rate": 1.2048700907435318e-07,
+      "loss": 0.7699,
+      "step": 12091
+    },
+    {
+      "epoch": 0.9321615787850756,
+      "grad_norm": 3.506540298461914,
+      "learning_rate": 1.2021474175829662e-07,
+      "loss": 0.9134,
+      "step": 12092
+    },
+    {
+      "epoch": 0.9322386679000925,
+      "grad_norm": 3.255476236343384,
+      "learning_rate": 1.1994277867157734e-07,
+      "loss": 0.8231,
+      "step": 12093
+    },
+    {
+      "epoch": 0.9323157570151095,
+      "grad_norm": 3.484185218811035,
+      "learning_rate": 1.196711198311512e-07,
+      "loss": 0.8325,
+      "step": 12094
+    },
+    {
+      "epoch": 0.9323928461301264,
+      "grad_norm": 4.07226037979126,
+      "learning_rate": 1.1939976525395468e-07,
+      "loss": 0.9041,
+      "step": 12095
+    },
+    {
+      "epoch": 0.9324699352451434,
+      "grad_norm": 3.9831414222717285,
+      "learning_rate": 1.1912871495690592e-07,
+      "loss": 0.9138,
+      "step": 12096
+    },
+    {
+      "epoch": 0.9325470243601603,
+      "grad_norm": 3.949620246887207,
+      "learning_rate": 1.1885796895690304e-07,
+      "loss": 1.0008,
+      "step": 12097
+    },
+    {
+      "epoch": 0.9326241134751773,
+      "grad_norm": 3.8057408332824707,
+      "learning_rate": 1.185875272708259e-07,
+      "loss": 0.9066,
+      "step": 12098
+    },
+    {
+      "epoch": 0.9327012025901943,
+      "grad_norm": 4.090064525604248,
+      "learning_rate": 1.1831738991553432e-07,
+      "loss": 0.9784,
+      "step": 12099
+    },
+    {
+      "epoch": 0.9327782917052112,
+      "grad_norm": 4.04171895980835,
+      "learning_rate": 1.1804755690787095e-07,
+      "loss": 0.98,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9328553808202282,
+      "grad_norm": 3.672153949737549,
+      "learning_rate": 1.177780282646579e-07,
+      "loss": 0.8458,
+      "step": 12101
+    },
+    {
+      "epoch": 0.9329324699352451,
+      "grad_norm": 3.4648172855377197,
+      "learning_rate": 1.1750880400269948e-07,
+      "loss": 0.8031,
+      "step": 12102
+    },
+    {
+      "epoch": 0.9330095590502621,
+      "grad_norm": 3.73508882522583,
+      "learning_rate": 1.1723988413878063e-07,
+      "loss": 0.9876,
+      "step": 12103
+    },
+    {
+      "epoch": 0.9330866481652791,
+      "grad_norm": 3.7413759231567383,
+      "learning_rate": 1.1697126868966569e-07,
+      "loss": 0.8011,
+      "step": 12104
+    },
+    {
+      "epoch": 0.933163737280296,
+      "grad_norm": 3.7317357063293457,
+      "learning_rate": 1.1670295767210238e-07,
+      "loss": 0.8842,
+      "step": 12105
+    },
+    {
+      "epoch": 0.933240826395313,
+      "grad_norm": 3.6849234104156494,
+      "learning_rate": 1.1643495110281844e-07,
+      "loss": 0.9667,
+      "step": 12106
+    },
+    {
+      "epoch": 0.9333179155103299,
+      "grad_norm": 3.7106947898864746,
+      "learning_rate": 1.1616724899852217e-07,
+      "loss": 0.8664,
+      "step": 12107
+    },
+    {
+      "epoch": 0.9333950046253469,
+      "grad_norm": 3.798051595687866,
+      "learning_rate": 1.158998513759052e-07,
+      "loss": 0.8214,
+      "step": 12108
+    },
+    {
+      "epoch": 0.9334720937403639,
+      "grad_norm": 4.954608917236328,
+      "learning_rate": 1.156327582516359e-07,
+      "loss": 0.9682,
+      "step": 12109
+    },
+    {
+      "epoch": 0.9335491828553808,
+      "grad_norm": 3.375283718109131,
+      "learning_rate": 1.1536596964236757e-07,
+      "loss": 0.7497,
+      "step": 12110
+    },
+    {
+      "epoch": 0.9336262719703978,
+      "grad_norm": 3.573298931121826,
+      "learning_rate": 1.1509948556473306e-07,
+      "loss": 0.8748,
+      "step": 12111
+    },
+    {
+      "epoch": 0.9337033610854147,
+      "grad_norm": 3.768425464630127,
+      "learning_rate": 1.1483330603534625e-07,
+      "loss": 0.8943,
+      "step": 12112
+    },
+    {
+      "epoch": 0.9337804502004317,
+      "grad_norm": 3.941859245300293,
+      "learning_rate": 1.1456743107080171e-07,
+      "loss": 0.9499,
+      "step": 12113
+    },
+    {
+      "epoch": 0.9338575393154487,
+      "grad_norm": 3.9781956672668457,
+      "learning_rate": 1.1430186068767557e-07,
+      "loss": 0.8871,
+      "step": 12114
+    },
+    {
+      "epoch": 0.9339346284304656,
+      "grad_norm": 3.6682260036468506,
+      "learning_rate": 1.1403659490252462e-07,
+      "loss": 0.8444,
+      "step": 12115
+    },
+    {
+      "epoch": 0.9340117175454826,
+      "grad_norm": 3.898146629333496,
+      "learning_rate": 1.137716337318867e-07,
+      "loss": 0.9247,
+      "step": 12116
+    },
+    {
+      "epoch": 0.9340888066604995,
+      "grad_norm": 4.3249335289001465,
+      "learning_rate": 1.1350697719228144e-07,
+      "loss": 0.9518,
+      "step": 12117
+    },
+    {
+      "epoch": 0.9341658957755165,
+      "grad_norm": 3.901083469390869,
+      "learning_rate": 1.1324262530020835e-07,
+      "loss": 0.9795,
+      "step": 12118
+    },
+    {
+      "epoch": 0.9342429848905335,
+      "grad_norm": 3.6279914379119873,
+      "learning_rate": 1.1297857807214818e-07,
+      "loss": 0.8613,
+      "step": 12119
+    },
+    {
+      "epoch": 0.9343200740055504,
+      "grad_norm": 3.867549180984497,
+      "learning_rate": 1.127148355245633e-07,
+      "loss": 0.8051,
+      "step": 12120
+    },
+    {
+      "epoch": 0.9343971631205674,
+      "grad_norm": 3.700650453567505,
+      "learning_rate": 1.1245139767389612e-07,
+      "loss": 0.8942,
+      "step": 12121
+    },
+    {
+      "epoch": 0.9344742522355843,
+      "grad_norm": 3.632929563522339,
+      "learning_rate": 1.1218826453657127e-07,
+      "loss": 0.8625,
+      "step": 12122
+    },
+    {
+      "epoch": 0.9345513413506013,
+      "grad_norm": 3.861298084259033,
+      "learning_rate": 1.1192543612899398e-07,
+      "loss": 0.9062,
+      "step": 12123
+    },
+    {
+      "epoch": 0.9346284304656183,
+      "grad_norm": 3.5877022743225098,
+      "learning_rate": 1.1166291246754945e-07,
+      "loss": 0.8753,
+      "step": 12124
+    },
+    {
+      "epoch": 0.9347055195806352,
+      "grad_norm": 3.9185619354248047,
+      "learning_rate": 1.114006935686046e-07,
+      "loss": 0.8794,
+      "step": 12125
+    },
+    {
+      "epoch": 0.9347826086956522,
+      "grad_norm": 3.9409501552581787,
+      "learning_rate": 1.1113877944850804e-07,
+      "loss": 1.0012,
+      "step": 12126
+    },
+    {
+      "epoch": 0.9348596978106691,
+      "grad_norm": 3.9709725379943848,
+      "learning_rate": 1.1087717012358834e-07,
+      "loss": 1.0408,
+      "step": 12127
+    },
+    {
+      "epoch": 0.934936786925686,
+      "grad_norm": 3.867388963699341,
+      "learning_rate": 1.1061586561015636e-07,
+      "loss": 0.9123,
+      "step": 12128
+    },
+    {
+      "epoch": 0.9350138760407031,
+      "grad_norm": 3.8618171215057373,
+      "learning_rate": 1.1035486592450184e-07,
+      "loss": 0.8511,
+      "step": 12129
+    },
+    {
+      "epoch": 0.93509096515572,
+      "grad_norm": 3.811495542526245,
+      "learning_rate": 1.1009417108289733e-07,
+      "loss": 0.8261,
+      "step": 12130
+    },
+    {
+      "epoch": 0.935168054270737,
+      "grad_norm": 4.648553848266602,
+      "learning_rate": 1.0983378110159593e-07,
+      "loss": 0.9152,
+      "step": 12131
+    },
+    {
+      "epoch": 0.9352451433857539,
+      "grad_norm": 3.401317834854126,
+      "learning_rate": 1.0957369599683132e-07,
+      "loss": 0.9336,
+      "step": 12132
+    },
+    {
+      "epoch": 0.9353222325007708,
+      "grad_norm": 4.083096504211426,
+      "learning_rate": 1.0931391578481832e-07,
+      "loss": 0.9892,
+      "step": 12133
+    },
+    {
+      "epoch": 0.9353993216157879,
+      "grad_norm": 3.8430960178375244,
+      "learning_rate": 1.0905444048175396e-07,
+      "loss": 0.8581,
+      "step": 12134
+    },
+    {
+      "epoch": 0.9354764107308048,
+      "grad_norm": 3.859454393386841,
+      "learning_rate": 1.0879527010381419e-07,
+      "loss": 0.9705,
+      "step": 12135
+    },
+    {
+      "epoch": 0.9355534998458218,
+      "grad_norm": 4.2190375328063965,
+      "learning_rate": 1.0853640466715664e-07,
+      "loss": 0.8858,
+      "step": 12136
+    },
+    {
+      "epoch": 0.9356305889608387,
+      "grad_norm": 3.7572779655456543,
+      "learning_rate": 1.0827784418792064e-07,
+      "loss": 0.9528,
+      "step": 12137
+    },
+    {
+      "epoch": 0.9357076780758556,
+      "grad_norm": 3.6124427318573,
+      "learning_rate": 1.0801958868222662e-07,
+      "loss": 0.8229,
+      "step": 12138
+    },
+    {
+      "epoch": 0.9357847671908727,
+      "grad_norm": 4.138119220733643,
+      "learning_rate": 1.0776163816617446e-07,
+      "loss": 0.9593,
+      "step": 12139
+    },
+    {
+      "epoch": 0.9358618563058896,
+      "grad_norm": 3.941408634185791,
+      "learning_rate": 1.0750399265584744e-07,
+      "loss": 0.8894,
+      "step": 12140
+    },
+    {
+      "epoch": 0.9359389454209066,
+      "grad_norm": 3.9744157791137695,
+      "learning_rate": 1.0724665216730711e-07,
+      "loss": 0.9538,
+      "step": 12141
+    },
+    {
+      "epoch": 0.9360160345359235,
+      "grad_norm": 3.580399513244629,
+      "learning_rate": 1.0698961671659791e-07,
+      "loss": 0.8409,
+      "step": 12142
+    },
+    {
+      "epoch": 0.9360931236509404,
+      "grad_norm": 3.7937207221984863,
+      "learning_rate": 1.0673288631974421e-07,
+      "loss": 1.0333,
+      "step": 12143
+    },
+    {
+      "epoch": 0.9361702127659575,
+      "grad_norm": 3.6252706050872803,
+      "learning_rate": 1.0647646099275267e-07,
+      "loss": 0.8437,
+      "step": 12144
+    },
+    {
+      "epoch": 0.9362473018809744,
+      "grad_norm": 4.086170673370361,
+      "learning_rate": 1.0622034075160936e-07,
+      "loss": 0.8064,
+      "step": 12145
+    },
+    {
+      "epoch": 0.9363243909959914,
+      "grad_norm": 3.6738433837890625,
+      "learning_rate": 1.059645256122821e-07,
+      "loss": 0.9238,
+      "step": 12146
+    },
+    {
+      "epoch": 0.9364014801110083,
+      "grad_norm": 4.102461338043213,
+      "learning_rate": 1.0570901559072033e-07,
+      "loss": 0.9676,
+      "step": 12147
+    },
+    {
+      "epoch": 0.9364785692260252,
+      "grad_norm": 4.378781318664551,
+      "learning_rate": 1.0545381070285243e-07,
+      "loss": 0.8863,
+      "step": 12148
+    },
+    {
+      "epoch": 0.9365556583410423,
+      "grad_norm": 3.481828212738037,
+      "learning_rate": 1.0519891096459067e-07,
+      "loss": 0.827,
+      "step": 12149
+    },
+    {
+      "epoch": 0.9366327474560592,
+      "grad_norm": 3.6842973232269287,
+      "learning_rate": 1.0494431639182567e-07,
+      "loss": 0.8607,
+      "step": 12150
+    },
+    {
+      "epoch": 0.9367098365710762,
+      "grad_norm": 3.917614221572876,
+      "learning_rate": 1.0469002700043029e-07,
+      "loss": 0.8737,
+      "step": 12151
+    },
+    {
+      "epoch": 0.9367869256860931,
+      "grad_norm": 3.5234785079956055,
+      "learning_rate": 1.0443604280625852e-07,
+      "loss": 0.8082,
+      "step": 12152
+    },
+    {
+      "epoch": 0.93686401480111,
+      "grad_norm": 3.9610044956207275,
+      "learning_rate": 1.0418236382514435e-07,
+      "loss": 0.9952,
+      "step": 12153
+    },
+    {
+      "epoch": 0.9369411039161271,
+      "grad_norm": 3.7560231685638428,
+      "learning_rate": 1.0392899007290347e-07,
+      "loss": 0.8605,
+      "step": 12154
+    },
+    {
+      "epoch": 0.937018193031144,
+      "grad_norm": 3.415227174758911,
+      "learning_rate": 1.036759215653338e-07,
+      "loss": 0.8353,
+      "step": 12155
+    },
+    {
+      "epoch": 0.937095282146161,
+      "grad_norm": 3.7259321212768555,
+      "learning_rate": 1.0342315831821104e-07,
+      "loss": 0.9268,
+      "step": 12156
+    },
+    {
+      "epoch": 0.9371723712611779,
+      "grad_norm": 3.7746005058288574,
+      "learning_rate": 1.0317070034729426e-07,
+      "loss": 0.8681,
+      "step": 12157
+    },
+    {
+      "epoch": 0.9372494603761948,
+      "grad_norm": 3.6794521808624268,
+      "learning_rate": 1.0291854766832254e-07,
+      "loss": 0.9124,
+      "step": 12158
+    },
+    {
+      "epoch": 0.9373265494912119,
+      "grad_norm": 3.675534725189209,
+      "learning_rate": 1.026667002970172e-07,
+      "loss": 0.9868,
+      "step": 12159
+    },
+    {
+      "epoch": 0.9374036386062288,
+      "grad_norm": 3.834210157394409,
+      "learning_rate": 1.0241515824907955e-07,
+      "loss": 0.8566,
+      "step": 12160
+    },
+    {
+      "epoch": 0.9374807277212458,
+      "grad_norm": 3.795119285583496,
+      "learning_rate": 1.021639215401915e-07,
+      "loss": 0.9076,
+      "step": 12161
+    },
+    {
+      "epoch": 0.9375578168362627,
+      "grad_norm": 3.7029600143432617,
+      "learning_rate": 1.0191299018601608e-07,
+      "loss": 0.8642,
+      "step": 12162
+    },
+    {
+      "epoch": 0.9376349059512796,
+      "grad_norm": 3.834038734436035,
+      "learning_rate": 1.0166236420219744e-07,
+      "loss": 0.9159,
+      "step": 12163
+    },
+    {
+      "epoch": 0.9377119950662967,
+      "grad_norm": 3.733107328414917,
+      "learning_rate": 1.0141204360436197e-07,
+      "loss": 0.8474,
+      "step": 12164
+    },
+    {
+      "epoch": 0.9377890841813136,
+      "grad_norm": 3.8652164936065674,
+      "learning_rate": 1.011620284081144e-07,
+      "loss": 0.9073,
+      "step": 12165
+    },
+    {
+      "epoch": 0.9378661732963306,
+      "grad_norm": 4.108710289001465,
+      "learning_rate": 1.0091231862904394e-07,
+      "loss": 0.8865,
+      "step": 12166
+    },
+    {
+      "epoch": 0.9379432624113475,
+      "grad_norm": 3.595679998397827,
+      "learning_rate": 1.0066291428271646e-07,
+      "loss": 0.9258,
+      "step": 12167
+    },
+    {
+      "epoch": 0.9380203515263644,
+      "grad_norm": 3.6943235397338867,
+      "learning_rate": 1.0041381538468175e-07,
+      "loss": 0.863,
+      "step": 12168
+    },
+    {
+      "epoch": 0.9380974406413815,
+      "grad_norm": 3.979586124420166,
+      "learning_rate": 1.0016502195047017e-07,
+      "loss": 0.8683,
+      "step": 12169
+    },
+    {
+      "epoch": 0.9381745297563984,
+      "grad_norm": 3.542391777038574,
+      "learning_rate": 9.991653399559265e-08,
+      "loss": 0.8578,
+      "step": 12170
+    },
+    {
+      "epoch": 0.9382516188714154,
+      "grad_norm": 3.59480357170105,
+      "learning_rate": 9.966835153554177e-08,
+      "loss": 0.9473,
+      "step": 12171
+    },
+    {
+      "epoch": 0.9383287079864323,
+      "grad_norm": 3.7652363777160645,
+      "learning_rate": 9.942047458578852e-08,
+      "loss": 0.9378,
+      "step": 12172
+    },
+    {
+      "epoch": 0.9384057971014492,
+      "grad_norm": 3.717331886291504,
+      "learning_rate": 9.917290316178884e-08,
+      "loss": 0.8719,
+      "step": 12173
+    },
+    {
+      "epoch": 0.9384828862164662,
+      "grad_norm": 4.402118682861328,
+      "learning_rate": 9.892563727897597e-08,
+      "loss": 0.9718,
+      "step": 12174
+    },
+    {
+      "epoch": 0.9385599753314832,
+      "grad_norm": 3.7598085403442383,
+      "learning_rate": 9.867867695276645e-08,
+      "loss": 0.8839,
+      "step": 12175
+    },
+    {
+      "epoch": 0.9386370644465002,
+      "grad_norm": 3.4538068771362305,
+      "learning_rate": 9.843202219855685e-08,
+      "loss": 0.8984,
+      "step": 12176
+    },
+    {
+      "epoch": 0.9387141535615171,
+      "grad_norm": 3.9770476818084717,
+      "learning_rate": 9.818567303172544e-08,
+      "loss": 0.8237,
+      "step": 12177
+    },
+    {
+      "epoch": 0.938791242676534,
+      "grad_norm": 3.9193336963653564,
+      "learning_rate": 9.793962946762936e-08,
+      "loss": 0.9098,
+      "step": 12178
+    },
+    {
+      "epoch": 0.938868331791551,
+      "grad_norm": 3.7343082427978516,
+      "learning_rate": 9.769389152160913e-08,
+      "loss": 0.9654,
+      "step": 12179
+    },
+    {
+      "epoch": 0.938945420906568,
+      "grad_norm": 3.766138792037964,
+      "learning_rate": 9.744845920898527e-08,
+      "loss": 0.9089,
+      "step": 12180
+    },
+    {
+      "epoch": 0.939022510021585,
+      "grad_norm": 4.211320877075195,
+      "learning_rate": 9.720333254505887e-08,
+      "loss": 0.9806,
+      "step": 12181
+    },
+    {
+      "epoch": 0.9390995991366019,
+      "grad_norm": 4.284248352050781,
+      "learning_rate": 9.695851154511271e-08,
+      "loss": 1.0172,
+      "step": 12182
+    },
+    {
+      "epoch": 0.9391766882516188,
+      "grad_norm": 3.9053781032562256,
+      "learning_rate": 9.671399622440957e-08,
+      "loss": 0.8795,
+      "step": 12183
+    },
+    {
+      "epoch": 0.9392537773666358,
+      "grad_norm": 3.633023262023926,
+      "learning_rate": 9.646978659819394e-08,
+      "loss": 0.8426,
+      "step": 12184
+    },
+    {
+      "epoch": 0.9393308664816528,
+      "grad_norm": 4.009373188018799,
+      "learning_rate": 9.622588268169141e-08,
+      "loss": 0.8787,
+      "step": 12185
+    },
+    {
+      "epoch": 0.9394079555966698,
+      "grad_norm": 3.9296786785125732,
+      "learning_rate": 9.598228449010704e-08,
+      "loss": 1.0206,
+      "step": 12186
+    },
+    {
+      "epoch": 0.9394850447116867,
+      "grad_norm": 3.5266220569610596,
+      "learning_rate": 9.573899203862925e-08,
+      "loss": 0.925,
+      "step": 12187
+    },
+    {
+      "epoch": 0.9395621338267036,
+      "grad_norm": 3.396301746368408,
+      "learning_rate": 9.549600534242587e-08,
+      "loss": 0.831,
+      "step": 12188
+    },
+    {
+      "epoch": 0.9396392229417206,
+      "grad_norm": 3.696995735168457,
+      "learning_rate": 9.525332441664481e-08,
+      "loss": 0.8887,
+      "step": 12189
+    },
+    {
+      "epoch": 0.9397163120567376,
+      "grad_norm": 3.7658541202545166,
+      "learning_rate": 9.50109492764173e-08,
+      "loss": 0.9345,
+      "step": 12190
+    },
+    {
+      "epoch": 0.9397934011717546,
+      "grad_norm": 3.884662389755249,
+      "learning_rate": 9.476887993685291e-08,
+      "loss": 0.9221,
+      "step": 12191
+    },
+    {
+      "epoch": 0.9398704902867715,
+      "grad_norm": 3.74465274810791,
+      "learning_rate": 9.452711641304402e-08,
+      "loss": 0.889,
+      "step": 12192
+    },
+    {
+      "epoch": 0.9399475794017884,
+      "grad_norm": 3.8537025451660156,
+      "learning_rate": 9.428565872006356e-08,
+      "loss": 1.0205,
+      "step": 12193
+    },
+    {
+      "epoch": 0.9400246685168054,
+      "grad_norm": 3.9346094131469727,
+      "learning_rate": 9.404450687296507e-08,
+      "loss": 0.835,
+      "step": 12194
+    },
+    {
+      "epoch": 0.9401017576318224,
+      "grad_norm": 3.567856550216675,
+      "learning_rate": 9.380366088678261e-08,
+      "loss": 0.8092,
+      "step": 12195
+    },
+    {
+      "epoch": 0.9401788467468394,
+      "grad_norm": 3.484372854232788,
+      "learning_rate": 9.356312077653196e-08,
+      "loss": 0.9011,
+      "step": 12196
+    },
+    {
+      "epoch": 0.9402559358618563,
+      "grad_norm": 4.030934810638428,
+      "learning_rate": 9.332288655720945e-08,
+      "loss": 0.9122,
+      "step": 12197
+    },
+    {
+      "epoch": 0.9403330249768732,
+      "grad_norm": 5.114222526550293,
+      "learning_rate": 9.308295824379365e-08,
+      "loss": 0.9911,
+      "step": 12198
+    },
+    {
+      "epoch": 0.9404101140918902,
+      "grad_norm": 3.6418612003326416,
+      "learning_rate": 9.284333585124094e-08,
+      "loss": 0.9481,
+      "step": 12199
+    },
+    {
+      "epoch": 0.9404872032069072,
+      "grad_norm": 3.4810874462127686,
+      "learning_rate": 9.260401939449215e-08,
+      "loss": 0.9044,
+      "step": 12200
+    },
+    {
+      "epoch": 0.9405642923219242,
+      "grad_norm": 3.6137795448303223,
+      "learning_rate": 9.236500888846589e-08,
+      "loss": 0.7977,
+      "step": 12201
+    },
+    {
+      "epoch": 0.9406413814369411,
+      "grad_norm": 3.7133419513702393,
+      "learning_rate": 9.212630434806413e-08,
+      "loss": 0.9542,
+      "step": 12202
+    },
+    {
+      "epoch": 0.940718470551958,
+      "grad_norm": 3.7179129123687744,
+      "learning_rate": 9.188790578816942e-08,
+      "loss": 0.8459,
+      "step": 12203
+    },
+    {
+      "epoch": 0.940795559666975,
+      "grad_norm": 3.615569591522217,
+      "learning_rate": 9.164981322364375e-08,
+      "loss": 0.8417,
+      "step": 12204
+    },
+    {
+      "epoch": 0.940872648781992,
+      "grad_norm": 3.824866533279419,
+      "learning_rate": 9.141202666933135e-08,
+      "loss": 1.0083,
+      "step": 12205
+    },
+    {
+      "epoch": 0.940949737897009,
+      "grad_norm": 3.547849655151367,
+      "learning_rate": 9.11745461400565e-08,
+      "loss": 0.8361,
+      "step": 12206
+    },
+    {
+      "epoch": 0.9410268270120259,
+      "grad_norm": 3.908607244491577,
+      "learning_rate": 9.09373716506251e-08,
+      "loss": 0.8185,
+      "step": 12207
+    },
+    {
+      "epoch": 0.9411039161270429,
+      "grad_norm": 3.6814157962799072,
+      "learning_rate": 9.07005032158248e-08,
+      "loss": 0.8893,
+      "step": 12208
+    },
+    {
+      "epoch": 0.9411810052420598,
+      "grad_norm": 3.7791712284088135,
+      "learning_rate": 9.046394085042154e-08,
+      "loss": 0.9866,
+      "step": 12209
+    },
+    {
+      "epoch": 0.9412580943570767,
+      "grad_norm": 3.7115676403045654,
+      "learning_rate": 9.022768456916409e-08,
+      "loss": 1.0297,
+      "step": 12210
+    },
+    {
+      "epoch": 0.9413351834720938,
+      "grad_norm": 3.7487754821777344,
+      "learning_rate": 8.999173438678233e-08,
+      "loss": 0.9212,
+      "step": 12211
+    },
+    {
+      "epoch": 0.9414122725871107,
+      "grad_norm": 3.4741318225860596,
+      "learning_rate": 8.975609031798671e-08,
+      "loss": 0.9159,
+      "step": 12212
+    },
+    {
+      "epoch": 0.9414893617021277,
+      "grad_norm": 3.7162187099456787,
+      "learning_rate": 8.952075237746771e-08,
+      "loss": 0.939,
+      "step": 12213
+    },
+    {
+      "epoch": 0.9415664508171446,
+      "grad_norm": 3.8283746242523193,
+      "learning_rate": 8.928572057989804e-08,
+      "loss": 0.8525,
+      "step": 12214
+    },
+    {
+      "epoch": 0.9416435399321615,
+      "grad_norm": 3.759662628173828,
+      "learning_rate": 8.905099493993041e-08,
+      "loss": 0.879,
+      "step": 12215
+    },
+    {
+      "epoch": 0.9417206290471786,
+      "grad_norm": 3.5151100158691406,
+      "learning_rate": 8.881657547219869e-08,
+      "loss": 0.8647,
+      "step": 12216
+    },
+    {
+      "epoch": 0.9417977181621955,
+      "grad_norm": 4.150439739227295,
+      "learning_rate": 8.858246219131784e-08,
+      "loss": 0.9453,
+      "step": 12217
+    },
+    {
+      "epoch": 0.9418748072772125,
+      "grad_norm": 3.5774085521698,
+      "learning_rate": 8.83486551118834e-08,
+      "loss": 0.8436,
+      "step": 12218
+    },
+    {
+      "epoch": 0.9419518963922294,
+      "grad_norm": 3.561293601989746,
+      "learning_rate": 8.811515424847261e-08,
+      "loss": 0.7775,
+      "step": 12219
+    },
+    {
+      "epoch": 0.9420289855072463,
+      "grad_norm": 4.2305827140808105,
+      "learning_rate": 8.788195961564273e-08,
+      "loss": 0.9484,
+      "step": 12220
+    },
+    {
+      "epoch": 0.9421060746222634,
+      "grad_norm": 3.9733986854553223,
+      "learning_rate": 8.764907122793154e-08,
+      "loss": 0.8226,
+      "step": 12221
+    },
+    {
+      "epoch": 0.9421831637372803,
+      "grad_norm": 3.7328057289123535,
+      "learning_rate": 8.741648909985967e-08,
+      "loss": 1.0214,
+      "step": 12222
+    },
+    {
+      "epoch": 0.9422602528522973,
+      "grad_norm": 3.4936575889587402,
+      "learning_rate": 8.718421324592608e-08,
+      "loss": 0.9083,
+      "step": 12223
+    },
+    {
+      "epoch": 0.9423373419673142,
+      "grad_norm": 3.8838651180267334,
+      "learning_rate": 8.695224368061305e-08,
+      "loss": 0.922,
+      "step": 12224
+    },
+    {
+      "epoch": 0.9424144310823311,
+      "grad_norm": 3.990166664123535,
+      "learning_rate": 8.672058041838294e-08,
+      "loss": 0.8594,
+      "step": 12225
+    },
+    {
+      "epoch": 0.9424915201973482,
+      "grad_norm": 3.8123323917388916,
+      "learning_rate": 8.64892234736775e-08,
+      "loss": 0.8615,
+      "step": 12226
+    },
+    {
+      "epoch": 0.9425686093123651,
+      "grad_norm": 3.807555675506592,
+      "learning_rate": 8.625817286092075e-08,
+      "loss": 0.8749,
+      "step": 12227
+    },
+    {
+      "epoch": 0.9426456984273821,
+      "grad_norm": 3.557279586791992,
+      "learning_rate": 8.602742859451841e-08,
+      "loss": 0.8594,
+      "step": 12228
+    },
+    {
+      "epoch": 0.942722787542399,
+      "grad_norm": 4.088743209838867,
+      "learning_rate": 8.579699068885616e-08,
+      "loss": 0.9473,
+      "step": 12229
+    },
+    {
+      "epoch": 0.9427998766574159,
+      "grad_norm": 3.422257661819458,
+      "learning_rate": 8.556685915830033e-08,
+      "loss": 0.8581,
+      "step": 12230
+    },
+    {
+      "epoch": 0.942876965772433,
+      "grad_norm": 3.8169891834259033,
+      "learning_rate": 8.533703401719773e-08,
+      "loss": 0.9015,
+      "step": 12231
+    },
+    {
+      "epoch": 0.9429540548874499,
+      "grad_norm": 3.840945243835449,
+      "learning_rate": 8.510751527987748e-08,
+      "loss": 0.9821,
+      "step": 12232
+    },
+    {
+      "epoch": 0.9430311440024669,
+      "grad_norm": 3.5995681285858154,
+      "learning_rate": 8.487830296064869e-08,
+      "loss": 0.9348,
+      "step": 12233
+    },
+    {
+      "epoch": 0.9431082331174838,
+      "grad_norm": 3.9970502853393555,
+      "learning_rate": 8.46493970738016e-08,
+      "loss": 0.9943,
+      "step": 12234
+    },
+    {
+      "epoch": 0.9431853222325007,
+      "grad_norm": 3.852111339569092,
+      "learning_rate": 8.442079763360755e-08,
+      "loss": 0.9228,
+      "step": 12235
+    },
+    {
+      "epoch": 0.9432624113475178,
+      "grad_norm": 3.759093761444092,
+      "learning_rate": 8.419250465431905e-08,
+      "loss": 0.8915,
+      "step": 12236
+    },
+    {
+      "epoch": 0.9433395004625347,
+      "grad_norm": 3.7389092445373535,
+      "learning_rate": 8.396451815016749e-08,
+      "loss": 0.8838,
+      "step": 12237
+    },
+    {
+      "epoch": 0.9434165895775517,
+      "grad_norm": 3.6164088249206543,
+      "learning_rate": 8.373683813536703e-08,
+      "loss": 0.8513,
+      "step": 12238
+    },
+    {
+      "epoch": 0.9434936786925686,
+      "grad_norm": 3.7332277297973633,
+      "learning_rate": 8.350946462411303e-08,
+      "loss": 0.9333,
+      "step": 12239
+    },
+    {
+      "epoch": 0.9435707678075855,
+      "grad_norm": 3.6654908657073975,
+      "learning_rate": 8.328239763058077e-08,
+      "loss": 0.8103,
+      "step": 12240
+    },
+    {
+      "epoch": 0.9436478569226026,
+      "grad_norm": 3.48644757270813,
+      "learning_rate": 8.305563716892728e-08,
+      "loss": 0.9154,
+      "step": 12241
+    },
+    {
+      "epoch": 0.9437249460376195,
+      "grad_norm": 3.515536308288574,
+      "learning_rate": 8.282918325328848e-08,
+      "loss": 0.9073,
+      "step": 12242
+    },
+    {
+      "epoch": 0.9438020351526365,
+      "grad_norm": 3.6415224075317383,
+      "learning_rate": 8.260303589778362e-08,
+      "loss": 0.849,
+      "step": 12243
+    },
+    {
+      "epoch": 0.9438791242676534,
+      "grad_norm": 3.827643394470215,
+      "learning_rate": 8.237719511651199e-08,
+      "loss": 0.8901,
+      "step": 12244
+    },
+    {
+      "epoch": 0.9439562133826703,
+      "grad_norm": 4.021296501159668,
+      "learning_rate": 8.215166092355286e-08,
+      "loss": 1.0162,
+      "step": 12245
+    },
+    {
+      "epoch": 0.9440333024976874,
+      "grad_norm": 3.674678325653076,
+      "learning_rate": 8.192643333296779e-08,
+      "loss": 0.8896,
+      "step": 12246
+    },
+    {
+      "epoch": 0.9441103916127043,
+      "grad_norm": 3.961991310119629,
+      "learning_rate": 8.17015123587983e-08,
+      "loss": 0.9414,
+      "step": 12247
+    },
+    {
+      "epoch": 0.9441874807277213,
+      "grad_norm": 3.9013025760650635,
+      "learning_rate": 8.147689801506653e-08,
+      "loss": 0.9172,
+      "step": 12248
+    },
+    {
+      "epoch": 0.9442645698427382,
+      "grad_norm": 3.474181890487671,
+      "learning_rate": 8.125259031577681e-08,
+      "loss": 0.8584,
+      "step": 12249
+    },
+    {
+      "epoch": 0.9443416589577551,
+      "grad_norm": 4.0712103843688965,
+      "learning_rate": 8.102858927491297e-08,
+      "loss": 0.9646,
+      "step": 12250
+    },
+    {
+      "epoch": 0.9444187480727722,
+      "grad_norm": 3.6317129135131836,
+      "learning_rate": 8.080489490644106e-08,
+      "loss": 0.825,
+      "step": 12251
+    },
+    {
+      "epoch": 0.9444958371877891,
+      "grad_norm": 3.92905855178833,
+      "learning_rate": 8.058150722430658e-08,
+      "loss": 0.9223,
+      "step": 12252
+    },
+    {
+      "epoch": 0.9445729263028061,
+      "grad_norm": 4.44641637802124,
+      "learning_rate": 8.035842624243673e-08,
+      "loss": 0.9286,
+      "step": 12253
+    },
+    {
+      "epoch": 0.944650015417823,
+      "grad_norm": 3.717348337173462,
+      "learning_rate": 8.013565197473927e-08,
+      "loss": 0.8958,
+      "step": 12254
+    },
+    {
+      "epoch": 0.9447271045328399,
+      "grad_norm": 3.6619021892547607,
+      "learning_rate": 7.991318443510365e-08,
+      "loss": 0.873,
+      "step": 12255
+    },
+    {
+      "epoch": 0.944804193647857,
+      "grad_norm": 3.7299957275390625,
+      "learning_rate": 7.969102363739933e-08,
+      "loss": 0.9392,
+      "step": 12256
+    },
+    {
+      "epoch": 0.9448812827628739,
+      "grad_norm": 3.6785383224487305,
+      "learning_rate": 7.946916959547635e-08,
+      "loss": 0.8582,
+      "step": 12257
+    },
+    {
+      "epoch": 0.9449583718778909,
+      "grad_norm": 3.580245018005371,
+      "learning_rate": 7.92476223231664e-08,
+      "loss": 0.8853,
+      "step": 12258
+    },
+    {
+      "epoch": 0.9450354609929078,
+      "grad_norm": 3.7513821125030518,
+      "learning_rate": 7.902638183428235e-08,
+      "loss": 0.9091,
+      "step": 12259
+    },
+    {
+      "epoch": 0.9451125501079247,
+      "grad_norm": 3.6423444747924805,
+      "learning_rate": 7.880544814261704e-08,
+      "loss": 0.9026,
+      "step": 12260
+    },
+    {
+      "epoch": 0.9451896392229417,
+      "grad_norm": 3.6739706993103027,
+      "learning_rate": 7.858482126194445e-08,
+      "loss": 0.8349,
+      "step": 12261
+    },
+    {
+      "epoch": 0.9452667283379587,
+      "grad_norm": 4.1562628746032715,
+      "learning_rate": 7.836450120601968e-08,
+      "loss": 0.8982,
+      "step": 12262
+    },
+    {
+      "epoch": 0.9453438174529757,
+      "grad_norm": 3.613546133041382,
+      "learning_rate": 7.814448798857843e-08,
+      "loss": 0.8372,
+      "step": 12263
+    },
+    {
+      "epoch": 0.9454209065679926,
+      "grad_norm": 3.739797353744507,
+      "learning_rate": 7.792478162333694e-08,
+      "loss": 0.9476,
+      "step": 12264
+    },
+    {
+      "epoch": 0.9454979956830095,
+      "grad_norm": 4.127760410308838,
+      "learning_rate": 7.77053821239937e-08,
+      "loss": 0.9036,
+      "step": 12265
+    },
+    {
+      "epoch": 0.9455750847980265,
+      "grad_norm": 3.717101812362671,
+      "learning_rate": 7.748628950422666e-08,
+      "loss": 0.8591,
+      "step": 12266
+    },
+    {
+      "epoch": 0.9456521739130435,
+      "grad_norm": 3.6435585021972656,
+      "learning_rate": 7.726750377769488e-08,
+      "loss": 0.9294,
+      "step": 12267
+    },
+    {
+      "epoch": 0.9457292630280605,
+      "grad_norm": 3.5347166061401367,
+      "learning_rate": 7.704902495803911e-08,
+      "loss": 0.8237,
+      "step": 12268
+    },
+    {
+      "epoch": 0.9458063521430774,
+      "grad_norm": 3.792682647705078,
+      "learning_rate": 7.683085305887961e-08,
+      "loss": 0.9064,
+      "step": 12269
+    },
+    {
+      "epoch": 0.9458834412580943,
+      "grad_norm": 3.9385581016540527,
+      "learning_rate": 7.661298809381878e-08,
+      "loss": 0.9554,
+      "step": 12270
+    },
+    {
+      "epoch": 0.9459605303731113,
+      "grad_norm": 3.6666650772094727,
+      "learning_rate": 7.639543007643913e-08,
+      "loss": 0.898,
+      "step": 12271
+    },
+    {
+      "epoch": 0.9460376194881283,
+      "grad_norm": 3.755439043045044,
+      "learning_rate": 7.617817902030478e-08,
+      "loss": 0.892,
+      "step": 12272
+    },
+    {
+      "epoch": 0.9461147086031453,
+      "grad_norm": 4.0008225440979,
+      "learning_rate": 7.59612349389599e-08,
+      "loss": 0.8153,
+      "step": 12273
+    },
+    {
+      "epoch": 0.9461917977181622,
+      "grad_norm": 3.6859991550445557,
+      "learning_rate": 7.574459784592981e-08,
+      "loss": 0.8639,
+      "step": 12274
+    },
+    {
+      "epoch": 0.9462688868331791,
+      "grad_norm": 3.624303102493286,
+      "learning_rate": 7.552826775472033e-08,
+      "loss": 0.8132,
+      "step": 12275
+    },
+    {
+      "epoch": 0.9463459759481961,
+      "grad_norm": 4.032271385192871,
+      "learning_rate": 7.531224467881848e-08,
+      "loss": 1.0711,
+      "step": 12276
+    },
+    {
+      "epoch": 0.946423065063213,
+      "grad_norm": 3.7532832622528076,
+      "learning_rate": 7.509652863169348e-08,
+      "loss": 0.8494,
+      "step": 12277
+    },
+    {
+      "epoch": 0.9465001541782301,
+      "grad_norm": 4.128763675689697,
+      "learning_rate": 7.48811196267929e-08,
+      "loss": 0.8642,
+      "step": 12278
+    },
+    {
+      "epoch": 0.946577243293247,
+      "grad_norm": 3.744234561920166,
+      "learning_rate": 7.466601767754655e-08,
+      "loss": 0.9608,
+      "step": 12279
+    },
+    {
+      "epoch": 0.9466543324082639,
+      "grad_norm": 3.6240429878234863,
+      "learning_rate": 7.445122279736484e-08,
+      "loss": 0.8315,
+      "step": 12280
+    },
+    {
+      "epoch": 0.9467314215232809,
+      "grad_norm": 3.724736213684082,
+      "learning_rate": 7.423673499963924e-08,
+      "loss": 0.8643,
+      "step": 12281
+    },
+    {
+      "epoch": 0.9468085106382979,
+      "grad_norm": 3.973217248916626,
+      "learning_rate": 7.402255429774241e-08,
+      "loss": 0.9679,
+      "step": 12282
+    },
+    {
+      "epoch": 0.9468855997533149,
+      "grad_norm": 3.8771355152130127,
+      "learning_rate": 7.380868070502644e-08,
+      "loss": 0.8696,
+      "step": 12283
+    },
+    {
+      "epoch": 0.9469626888683318,
+      "grad_norm": 3.8972392082214355,
+      "learning_rate": 7.359511423482679e-08,
+      "loss": 1.005,
+      "step": 12284
+    },
+    {
+      "epoch": 0.9470397779833487,
+      "grad_norm": 3.7604198455810547,
+      "learning_rate": 7.338185490045668e-08,
+      "loss": 0.867,
+      "step": 12285
+    },
+    {
+      "epoch": 0.9471168670983657,
+      "grad_norm": 3.7321414947509766,
+      "learning_rate": 7.316890271521215e-08,
+      "loss": 0.8899,
+      "step": 12286
+    },
+    {
+      "epoch": 0.9471939562133826,
+      "grad_norm": 4.230839252471924,
+      "learning_rate": 7.29562576923698e-08,
+      "loss": 1.034,
+      "step": 12287
+    },
+    {
+      "epoch": 0.9472710453283997,
+      "grad_norm": 3.6010680198669434,
+      "learning_rate": 7.274391984518736e-08,
+      "loss": 0.8038,
+      "step": 12288
+    },
+    {
+      "epoch": 0.9473481344434166,
+      "grad_norm": 3.6797475814819336,
+      "learning_rate": 7.253188918690257e-08,
+      "loss": 0.9017,
+      "step": 12289
+    },
+    {
+      "epoch": 0.9474252235584335,
+      "grad_norm": 3.575794219970703,
+      "learning_rate": 7.232016573073431e-08,
+      "loss": 0.8756,
+      "step": 12290
+    },
+    {
+      "epoch": 0.9475023126734505,
+      "grad_norm": 3.789335012435913,
+      "learning_rate": 7.210874948988255e-08,
+      "loss": 0.9401,
+      "step": 12291
+    },
+    {
+      "epoch": 0.9475794017884674,
+      "grad_norm": 3.835794448852539,
+      "learning_rate": 7.189764047752789e-08,
+      "loss": 0.9796,
+      "step": 12292
+    },
+    {
+      "epoch": 0.9476564909034845,
+      "grad_norm": 3.511427879333496,
+      "learning_rate": 7.168683870683258e-08,
+      "loss": 0.9066,
+      "step": 12293
+    },
+    {
+      "epoch": 0.9477335800185014,
+      "grad_norm": 3.5862207412719727,
+      "learning_rate": 7.14763441909383e-08,
+      "loss": 0.9302,
+      "step": 12294
+    },
+    {
+      "epoch": 0.9478106691335183,
+      "grad_norm": 3.8762261867523193,
+      "learning_rate": 7.126615694296846e-08,
+      "loss": 0.9133,
+      "step": 12295
+    },
+    {
+      "epoch": 0.9478877582485353,
+      "grad_norm": 3.6678431034088135,
+      "learning_rate": 7.105627697602702e-08,
+      "loss": 0.8997,
+      "step": 12296
+    },
+    {
+      "epoch": 0.9479648473635522,
+      "grad_norm": 3.454246997833252,
+      "learning_rate": 7.084670430319907e-08,
+      "loss": 0.7763,
+      "step": 12297
+    },
+    {
+      "epoch": 0.9480419364785693,
+      "grad_norm": 3.5856475830078125,
+      "learning_rate": 7.063743893755026e-08,
+      "loss": 0.8533,
+      "step": 12298
+    },
+    {
+      "epoch": 0.9481190255935862,
+      "grad_norm": 3.7448766231536865,
+      "learning_rate": 7.042848089212794e-08,
+      "loss": 0.9361,
+      "step": 12299
+    },
+    {
+      "epoch": 0.9481961147086031,
+      "grad_norm": 3.805290460586548,
+      "learning_rate": 7.021983017995836e-08,
+      "loss": 0.9013,
+      "step": 12300
+    },
+    {
+      "epoch": 0.9482732038236201,
+      "grad_norm": 3.3226075172424316,
+      "learning_rate": 7.001148681405056e-08,
+      "loss": 0.7962,
+      "step": 12301
+    },
+    {
+      "epoch": 0.948350292938637,
+      "grad_norm": 3.614866256713867,
+      "learning_rate": 6.980345080739303e-08,
+      "loss": 0.7797,
+      "step": 12302
+    },
+    {
+      "epoch": 0.9484273820536541,
+      "grad_norm": 3.74783992767334,
+      "learning_rate": 6.959572217295651e-08,
+      "loss": 0.9274,
+      "step": 12303
+    },
+    {
+      "epoch": 0.948504471168671,
+      "grad_norm": 4.042082786560059,
+      "learning_rate": 6.938830092369176e-08,
+      "loss": 1.0048,
+      "step": 12304
+    },
+    {
+      "epoch": 0.9485815602836879,
+      "grad_norm": 3.7380220890045166,
+      "learning_rate": 6.918118707253007e-08,
+      "loss": 0.9577,
+      "step": 12305
+    },
+    {
+      "epoch": 0.9486586493987049,
+      "grad_norm": 4.00551700592041,
+      "learning_rate": 6.897438063238393e-08,
+      "loss": 0.8565,
+      "step": 12306
+    },
+    {
+      "epoch": 0.9487357385137218,
+      "grad_norm": 4.03279972076416,
+      "learning_rate": 6.87678816161469e-08,
+      "loss": 1.0123,
+      "step": 12307
+    },
+    {
+      "epoch": 0.9488128276287389,
+      "grad_norm": 3.53090500831604,
+      "learning_rate": 6.856169003669256e-08,
+      "loss": 0.7968,
+      "step": 12308
+    },
+    {
+      "epoch": 0.9488899167437558,
+      "grad_norm": 3.888162612915039,
+      "learning_rate": 6.835580590687618e-08,
+      "loss": 0.9289,
+      "step": 12309
+    },
+    {
+      "epoch": 0.9489670058587727,
+      "grad_norm": 3.883291721343994,
+      "learning_rate": 6.815022923953418e-08,
+      "loss": 0.8446,
+      "step": 12310
+    },
+    {
+      "epoch": 0.9490440949737897,
+      "grad_norm": 4.122531414031982,
+      "learning_rate": 6.794496004748241e-08,
+      "loss": 0.8171,
+      "step": 12311
+    },
+    {
+      "epoch": 0.9491211840888066,
+      "grad_norm": 3.75138521194458,
+      "learning_rate": 6.773999834351841e-08,
+      "loss": 0.8468,
+      "step": 12312
+    },
+    {
+      "epoch": 0.9491982732038237,
+      "grad_norm": 3.7109479904174805,
+      "learning_rate": 6.753534414042085e-08,
+      "loss": 0.9143,
+      "step": 12313
+    },
+    {
+      "epoch": 0.9492753623188406,
+      "grad_norm": 3.7108957767486572,
+      "learning_rate": 6.733099745094896e-08,
+      "loss": 0.9051,
+      "step": 12314
+    },
+    {
+      "epoch": 0.9493524514338575,
+      "grad_norm": 3.8768532276153564,
+      "learning_rate": 6.712695828784254e-08,
+      "loss": 0.8712,
+      "step": 12315
+    },
+    {
+      "epoch": 0.9494295405488745,
+      "grad_norm": 3.5676753520965576,
+      "learning_rate": 6.6923226663822e-08,
+      "loss": 0.8435,
+      "step": 12316
+    },
+    {
+      "epoch": 0.9495066296638914,
+      "grad_norm": 3.522658348083496,
+      "learning_rate": 6.671980259158883e-08,
+      "loss": 0.7721,
+      "step": 12317
+    },
+    {
+      "epoch": 0.9495837187789085,
+      "grad_norm": 3.815915584564209,
+      "learning_rate": 6.651668608382622e-08,
+      "loss": 0.9029,
+      "step": 12318
+    },
+    {
+      "epoch": 0.9496608078939254,
+      "grad_norm": 3.938446044921875,
+      "learning_rate": 6.631387715319681e-08,
+      "loss": 0.9521,
+      "step": 12319
+    },
+    {
+      "epoch": 0.9497378970089423,
+      "grad_norm": 3.745887517929077,
+      "learning_rate": 6.611137581234495e-08,
+      "loss": 0.9028,
+      "step": 12320
+    },
+    {
+      "epoch": 0.9498149861239593,
+      "grad_norm": 3.692595958709717,
+      "learning_rate": 6.590918207389608e-08,
+      "loss": 0.8865,
+      "step": 12321
+    },
+    {
+      "epoch": 0.9498920752389762,
+      "grad_norm": 4.061500549316406,
+      "learning_rate": 6.57072959504551e-08,
+      "loss": 0.9498,
+      "step": 12322
+    },
+    {
+      "epoch": 0.9499691643539933,
+      "grad_norm": 3.645247459411621,
+      "learning_rate": 6.550571745460865e-08,
+      "loss": 0.8695,
+      "step": 12323
+    },
+    {
+      "epoch": 0.9500462534690102,
+      "grad_norm": 3.4942803382873535,
+      "learning_rate": 6.530444659892443e-08,
+      "loss": 0.8362,
+      "step": 12324
+    },
+    {
+      "epoch": 0.9501233425840271,
+      "grad_norm": 3.572190284729004,
+      "learning_rate": 6.510348339595074e-08,
+      "loss": 0.9694,
+      "step": 12325
+    },
+    {
+      "epoch": 0.9502004316990441,
+      "grad_norm": 3.5825698375701904,
+      "learning_rate": 6.490282785821645e-08,
+      "loss": 0.8123,
+      "step": 12326
+    },
+    {
+      "epoch": 0.950277520814061,
+      "grad_norm": 3.692664861679077,
+      "learning_rate": 6.470247999823099e-08,
+      "loss": 0.897,
+      "step": 12327
+    },
+    {
+      "epoch": 0.950354609929078,
+      "grad_norm": 3.427412271499634,
+      "learning_rate": 6.450243982848548e-08,
+      "loss": 0.8236,
+      "step": 12328
+    },
+    {
+      "epoch": 0.950431699044095,
+      "grad_norm": 3.9332494735717773,
+      "learning_rate": 6.430270736145106e-08,
+      "loss": 0.9017,
+      "step": 12329
+    },
+    {
+      "epoch": 0.9505087881591119,
+      "grad_norm": 3.665562629699707,
+      "learning_rate": 6.410328260957998e-08,
+      "loss": 0.8807,
+      "step": 12330
+    },
+    {
+      "epoch": 0.9505858772741289,
+      "grad_norm": 3.466925859451294,
+      "learning_rate": 6.390416558530622e-08,
+      "loss": 0.8508,
+      "step": 12331
+    },
+    {
+      "epoch": 0.9506629663891458,
+      "grad_norm": 3.790743112564087,
+      "learning_rate": 6.370535630104257e-08,
+      "loss": 0.8721,
+      "step": 12332
+    },
+    {
+      "epoch": 0.9507400555041629,
+      "grad_norm": 3.7184832096099854,
+      "learning_rate": 6.350685476918416e-08,
+      "loss": 0.964,
+      "step": 12333
+    },
+    {
+      "epoch": 0.9508171446191798,
+      "grad_norm": 3.781548500061035,
+      "learning_rate": 6.33086610021072e-08,
+      "loss": 0.9842,
+      "step": 12334
+    },
+    {
+      "epoch": 0.9508942337341967,
+      "grad_norm": 4.001351356506348,
+      "learning_rate": 6.311077501216677e-08,
+      "loss": 0.9441,
+      "step": 12335
+    },
+    {
+      "epoch": 0.9509713228492137,
+      "grad_norm": 3.9916863441467285,
+      "learning_rate": 6.291319681170139e-08,
+      "loss": 0.9009,
+      "step": 12336
+    },
+    {
+      "epoch": 0.9510484119642306,
+      "grad_norm": 4.026021480560303,
+      "learning_rate": 6.271592641302781e-08,
+      "loss": 0.9658,
+      "step": 12337
+    },
+    {
+      "epoch": 0.9511255010792476,
+      "grad_norm": 3.522988796234131,
+      "learning_rate": 6.251896382844569e-08,
+      "loss": 0.796,
+      "step": 12338
+    },
+    {
+      "epoch": 0.9512025901942646,
+      "grad_norm": 3.4949891567230225,
+      "learning_rate": 6.232230907023407e-08,
+      "loss": 0.8099,
+      "step": 12339
+    },
+    {
+      "epoch": 0.9512796793092815,
+      "grad_norm": 3.4826712608337402,
+      "learning_rate": 6.21259621506537e-08,
+      "loss": 0.9186,
+      "step": 12340
+    },
+    {
+      "epoch": 0.9513567684242985,
+      "grad_norm": 3.609074115753174,
+      "learning_rate": 6.192992308194534e-08,
+      "loss": 0.8708,
+      "step": 12341
+    },
+    {
+      "epoch": 0.9514338575393154,
+      "grad_norm": 3.599756956100464,
+      "learning_rate": 6.173419187633201e-08,
+      "loss": 0.7601,
+      "step": 12342
+    },
+    {
+      "epoch": 0.9515109466543324,
+      "grad_norm": 3.8272032737731934,
+      "learning_rate": 6.15387685460156e-08,
+      "loss": 0.9467,
+      "step": 12343
+    },
+    {
+      "epoch": 0.9515880357693494,
+      "grad_norm": 3.584209680557251,
+      "learning_rate": 6.134365310317969e-08,
+      "loss": 0.898,
+      "step": 12344
+    },
+    {
+      "epoch": 0.9516651248843663,
+      "grad_norm": 3.9957098960876465,
+      "learning_rate": 6.114884555998957e-08,
+      "loss": 0.9181,
+      "step": 12345
+    },
+    {
+      "epoch": 0.9517422139993833,
+      "grad_norm": 3.9217257499694824,
+      "learning_rate": 6.09543459285894e-08,
+      "loss": 0.8856,
+      "step": 12346
+    },
+    {
+      "epoch": 0.9518193031144002,
+      "grad_norm": 3.2795722484588623,
+      "learning_rate": 6.07601542211067e-08,
+      "loss": 0.8311,
+      "step": 12347
+    },
+    {
+      "epoch": 0.9518963922294172,
+      "grad_norm": 3.9861719608306885,
+      "learning_rate": 6.05662704496468e-08,
+      "loss": 0.9533,
+      "step": 12348
+    },
+    {
+      "epoch": 0.9519734813444342,
+      "grad_norm": 3.6338775157928467,
+      "learning_rate": 6.03726946262978e-08,
+      "loss": 0.8222,
+      "step": 12349
+    },
+    {
+      "epoch": 0.9520505704594511,
+      "grad_norm": 4.2912726402282715,
+      "learning_rate": 6.017942676312837e-08,
+      "loss": 0.9267,
+      "step": 12350
+    },
+    {
+      "epoch": 0.9521276595744681,
+      "grad_norm": 4.117364883422852,
+      "learning_rate": 5.998646687218779e-08,
+      "loss": 0.9223,
+      "step": 12351
+    },
+    {
+      "epoch": 0.952204748689485,
+      "grad_norm": 4.384626865386963,
+      "learning_rate": 5.97938149655064e-08,
+      "loss": 1.0203,
+      "step": 12352
+    },
+    {
+      "epoch": 0.952281837804502,
+      "grad_norm": 3.674335241317749,
+      "learning_rate": 5.960147105509406e-08,
+      "loss": 0.8207,
+      "step": 12353
+    },
+    {
+      "epoch": 0.952358926919519,
+      "grad_norm": 3.7455525398254395,
+      "learning_rate": 5.9409435152943395e-08,
+      "loss": 0.8943,
+      "step": 12354
+    },
+    {
+      "epoch": 0.9524360160345359,
+      "grad_norm": 3.3691940307617188,
+      "learning_rate": 5.9217707271025937e-08,
+      "loss": 0.899,
+      "step": 12355
+    },
+    {
+      "epoch": 0.9525131051495529,
+      "grad_norm": 3.5628163814544678,
+      "learning_rate": 5.9026287421296014e-08,
+      "loss": 0.8902,
+      "step": 12356
+    },
+    {
+      "epoch": 0.9525901942645698,
+      "grad_norm": 3.971393585205078,
+      "learning_rate": 5.883517561568686e-08,
+      "loss": 0.8993,
+      "step": 12357
+    },
+    {
+      "epoch": 0.9526672833795868,
+      "grad_norm": 3.777592420578003,
+      "learning_rate": 5.8644371866113934e-08,
+      "loss": 0.8346,
+      "step": 12358
+    },
+    {
+      "epoch": 0.9527443724946038,
+      "grad_norm": 3.8741729259490967,
+      "learning_rate": 5.845387618447218e-08,
+      "loss": 0.8584,
+      "step": 12359
+    },
+    {
+      "epoch": 0.9528214616096207,
+      "grad_norm": 3.5481209754943848,
+      "learning_rate": 5.82636885826382e-08,
+      "loss": 0.802,
+      "step": 12360
+    },
+    {
+      "epoch": 0.9528985507246377,
+      "grad_norm": 3.961042642593384,
+      "learning_rate": 5.807380907246918e-08,
+      "loss": 0.9015,
+      "step": 12361
+    },
+    {
+      "epoch": 0.9529756398396546,
+      "grad_norm": 3.8133602142333984,
+      "learning_rate": 5.788423766580342e-08,
+      "loss": 0.7618,
+      "step": 12362
+    },
+    {
+      "epoch": 0.9530527289546716,
+      "grad_norm": 3.639708995819092,
+      "learning_rate": 5.769497437446037e-08,
+      "loss": 0.794,
+      "step": 12363
+    },
+    {
+      "epoch": 0.9531298180696886,
+      "grad_norm": 3.4450173377990723,
+      "learning_rate": 5.7506019210237794e-08,
+      "loss": 0.8323,
+      "step": 12364
+    },
+    {
+      "epoch": 0.9532069071847055,
+      "grad_norm": 3.6423511505126953,
+      "learning_rate": 5.73173721849174e-08,
+      "loss": 0.8228,
+      "step": 12365
+    },
+    {
+      "epoch": 0.9532839962997225,
+      "grad_norm": 3.7283174991607666,
+      "learning_rate": 5.7129033310260316e-08,
+      "loss": 0.9807,
+      "step": 12366
+    },
+    {
+      "epoch": 0.9533610854147394,
+      "grad_norm": 3.4573850631713867,
+      "learning_rate": 5.694100259800772e-08,
+      "loss": 0.8502,
+      "step": 12367
+    },
+    {
+      "epoch": 0.9534381745297564,
+      "grad_norm": 3.561988592147827,
+      "learning_rate": 5.675328005988301e-08,
+      "loss": 0.9494,
+      "step": 12368
+    },
+    {
+      "epoch": 0.9535152636447733,
+      "grad_norm": 3.7618987560272217,
+      "learning_rate": 5.6565865707590153e-08,
+      "loss": 0.9674,
+      "step": 12369
+    },
+    {
+      "epoch": 0.9535923527597903,
+      "grad_norm": 3.7699737548828125,
+      "learning_rate": 5.637875955281202e-08,
+      "loss": 0.7715,
+      "step": 12370
+    },
+    {
+      "epoch": 0.9536694418748073,
+      "grad_norm": 3.5663082599639893,
+      "learning_rate": 5.6191961607214853e-08,
+      "loss": 0.8386,
+      "step": 12371
+    },
+    {
+      "epoch": 0.9537465309898242,
+      "grad_norm": 3.4624545574188232,
+      "learning_rate": 5.6005471882444316e-08,
+      "loss": 0.8488,
+      "step": 12372
+    },
+    {
+      "epoch": 0.9538236201048412,
+      "grad_norm": 4.092374801635742,
+      "learning_rate": 5.581929039012668e-08,
+      "loss": 0.9231,
+      "step": 12373
+    },
+    {
+      "epoch": 0.9539007092198581,
+      "grad_norm": 4.778739929199219,
+      "learning_rate": 5.563341714186987e-08,
+      "loss": 1.0022,
+      "step": 12374
+    },
+    {
+      "epoch": 0.9539777983348752,
+      "grad_norm": 3.724924325942993,
+      "learning_rate": 5.5447852149262406e-08,
+      "loss": 0.9458,
+      "step": 12375
+    },
+    {
+      "epoch": 0.9540548874498921,
+      "grad_norm": 3.464205026626587,
+      "learning_rate": 5.5262595423872244e-08,
+      "loss": 0.8716,
+      "step": 12376
+    },
+    {
+      "epoch": 0.954131976564909,
+      "grad_norm": 3.98112416267395,
+      "learning_rate": 5.507764697725015e-08,
+      "loss": 1.0144,
+      "step": 12377
+    },
+    {
+      "epoch": 0.954209065679926,
+      "grad_norm": 3.9500741958618164,
+      "learning_rate": 5.4893006820926355e-08,
+      "loss": 0.9449,
+      "step": 12378
+    },
+    {
+      "epoch": 0.9542861547949429,
+      "grad_norm": 3.776728391647339,
+      "learning_rate": 5.4708674966412744e-08,
+      "loss": 0.8444,
+      "step": 12379
+    },
+    {
+      "epoch": 0.95436324390996,
+      "grad_norm": 4.004519939422607,
+      "learning_rate": 5.4524651425200135e-08,
+      "loss": 0.8892,
+      "step": 12380
+    },
+    {
+      "epoch": 0.9544403330249769,
+      "grad_norm": 3.8165571689605713,
+      "learning_rate": 5.434093620876213e-08,
+      "loss": 0.9224,
+      "step": 12381
+    },
+    {
+      "epoch": 0.9545174221399938,
+      "grad_norm": 3.499563217163086,
+      "learning_rate": 5.4157529328552896e-08,
+      "loss": 0.7655,
+      "step": 12382
+    },
+    {
+      "epoch": 0.9545945112550108,
+      "grad_norm": 3.9584450721740723,
+      "learning_rate": 5.397443079600662e-08,
+      "loss": 0.8664,
+      "step": 12383
+    },
+    {
+      "epoch": 0.9546716003700277,
+      "grad_norm": 4.017448425292969,
+      "learning_rate": 5.379164062253861e-08,
+      "loss": 0.9921,
+      "step": 12384
+    },
+    {
+      "epoch": 0.9547486894850448,
+      "grad_norm": 3.8151862621307373,
+      "learning_rate": 5.3609158819544205e-08,
+      "loss": 0.8458,
+      "step": 12385
+    },
+    {
+      "epoch": 0.9548257786000617,
+      "grad_norm": 3.502845287322998,
+      "learning_rate": 5.3426985398400965e-08,
+      "loss": 0.861,
+      "step": 12386
+    },
+    {
+      "epoch": 0.9549028677150786,
+      "grad_norm": 3.5837574005126953,
+      "learning_rate": 5.324512037046647e-08,
+      "loss": 0.9356,
+      "step": 12387
+    },
+    {
+      "epoch": 0.9549799568300956,
+      "grad_norm": 3.933488368988037,
+      "learning_rate": 5.306356374707833e-08,
+      "loss": 0.9829,
+      "step": 12388
+    },
+    {
+      "epoch": 0.9550570459451125,
+      "grad_norm": 3.528136968612671,
+      "learning_rate": 5.288231553955636e-08,
+      "loss": 0.9494,
+      "step": 12389
+    },
+    {
+      "epoch": 0.9551341350601296,
+      "grad_norm": 3.7610669136047363,
+      "learning_rate": 5.270137575920098e-08,
+      "loss": 0.9087,
+      "step": 12390
+    },
+    {
+      "epoch": 0.9552112241751465,
+      "grad_norm": 3.8250885009765625,
+      "learning_rate": 5.2520744417290936e-08,
+      "loss": 0.9419,
+      "step": 12391
+    },
+    {
+      "epoch": 0.9552883132901634,
+      "grad_norm": 3.721879720687866,
+      "learning_rate": 5.2340421525089445e-08,
+      "loss": 0.9422,
+      "step": 12392
+    },
+    {
+      "epoch": 0.9553654024051804,
+      "grad_norm": 3.6027677059173584,
+      "learning_rate": 5.2160407093838074e-08,
+      "loss": 0.9309,
+      "step": 12393
+    },
+    {
+      "epoch": 0.9554424915201973,
+      "grad_norm": 3.6538522243499756,
+      "learning_rate": 5.19807011347595e-08,
+      "loss": 0.9611,
+      "step": 12394
+    },
+    {
+      "epoch": 0.9555195806352144,
+      "grad_norm": 4.078005790710449,
+      "learning_rate": 5.180130365905811e-08,
+      "loss": 0.9666,
+      "step": 12395
+    },
+    {
+      "epoch": 0.9555966697502313,
+      "grad_norm": 3.643521785736084,
+      "learning_rate": 5.162221467791772e-08,
+      "loss": 0.8401,
+      "step": 12396
+    },
+    {
+      "epoch": 0.9556737588652482,
+      "grad_norm": 3.5786333084106445,
+      "learning_rate": 5.1443434202504414e-08,
+      "loss": 0.8743,
+      "step": 12397
+    },
+    {
+      "epoch": 0.9557508479802652,
+      "grad_norm": 3.833845853805542,
+      "learning_rate": 5.1264962243963155e-08,
+      "loss": 0.9121,
+      "step": 12398
+    },
+    {
+      "epoch": 0.9558279370952821,
+      "grad_norm": 3.96075439453125,
+      "learning_rate": 5.1086798813421715e-08,
+      "loss": 0.9589,
+      "step": 12399
+    },
+    {
+      "epoch": 0.9559050262102992,
+      "grad_norm": 3.644322156906128,
+      "learning_rate": 5.090894392198731e-08,
+      "loss": 0.9057,
+      "step": 12400
+    },
+    {
+      "epoch": 0.9559821153253161,
+      "grad_norm": 3.6770665645599365,
+      "learning_rate": 5.07313975807483e-08,
+      "loss": 0.8914,
+      "step": 12401
+    },
+    {
+      "epoch": 0.956059204440333,
+      "grad_norm": 3.862185001373291,
+      "learning_rate": 5.0554159800773604e-08,
+      "loss": 0.7902,
+      "step": 12402
+    },
+    {
+      "epoch": 0.95613629355535,
+      "grad_norm": 3.8168914318084717,
+      "learning_rate": 5.037723059311273e-08,
+      "loss": 0.8907,
+      "step": 12403
+    },
+    {
+      "epoch": 0.9562133826703669,
+      "grad_norm": 3.6184260845184326,
+      "learning_rate": 5.0200609968797386e-08,
+      "loss": 0.886,
+      "step": 12404
+    },
+    {
+      "epoch": 0.956290471785384,
+      "grad_norm": 3.8705575466156006,
+      "learning_rate": 5.0024297938838227e-08,
+      "loss": 0.9254,
+      "step": 12405
+    },
+    {
+      "epoch": 0.9563675609004009,
+      "grad_norm": 3.5495188236236572,
+      "learning_rate": 4.9848294514227566e-08,
+      "loss": 0.88,
+      "step": 12406
+    },
+    {
+      "epoch": 0.9564446500154178,
+      "grad_norm": 3.8909826278686523,
+      "learning_rate": 4.9672599705938294e-08,
+      "loss": 1.0299,
+      "step": 12407
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 3.9004476070404053,
+      "learning_rate": 4.949721352492387e-08,
+      "loss": 0.9281,
+      "step": 12408
+    },
+    {
+      "epoch": 0.9565988282454517,
+      "grad_norm": 3.852919816970825,
+      "learning_rate": 4.9322135982118877e-08,
+      "loss": 0.9431,
+      "step": 12409
+    },
+    {
+      "epoch": 0.9566759173604688,
+      "grad_norm": 3.943549633026123,
+      "learning_rate": 4.914736708843848e-08,
+      "loss": 0.905,
+      "step": 12410
+    },
+    {
+      "epoch": 0.9567530064754857,
+      "grad_norm": 3.6017589569091797,
+      "learning_rate": 4.89729068547784e-08,
+      "loss": 0.9505,
+      "step": 12411
+    },
+    {
+      "epoch": 0.9568300955905026,
+      "grad_norm": 3.4761962890625,
+      "learning_rate": 4.8798755292016055e-08,
+      "loss": 0.8673,
+      "step": 12412
+    },
+    {
+      "epoch": 0.9569071847055196,
+      "grad_norm": 3.7245309352874756,
+      "learning_rate": 4.8624912411007754e-08,
+      "loss": 0.8425,
+      "step": 12413
+    },
+    {
+      "epoch": 0.9569842738205365,
+      "grad_norm": 3.6705353260040283,
+      "learning_rate": 4.8451378222592605e-08,
+      "loss": 0.8727,
+      "step": 12414
+    },
+    {
+      "epoch": 0.9570613629355536,
+      "grad_norm": 3.829969644546509,
+      "learning_rate": 4.827815273758973e-08,
+      "loss": 1.01,
+      "step": 12415
+    },
+    {
+      "epoch": 0.9571384520505705,
+      "grad_norm": 3.7990779876708984,
+      "learning_rate": 4.8105235966797727e-08,
+      "loss": 0.8892,
+      "step": 12416
+    },
+    {
+      "epoch": 0.9572155411655874,
+      "grad_norm": 3.6120383739471436,
+      "learning_rate": 4.793262792099851e-08,
+      "loss": 0.8526,
+      "step": 12417
+    },
+    {
+      "epoch": 0.9572926302806044,
+      "grad_norm": 3.7359561920166016,
+      "learning_rate": 4.776032861095181e-08,
+      "loss": 0.9674,
+      "step": 12418
+    },
+    {
+      "epoch": 0.9573697193956213,
+      "grad_norm": 3.7942633628845215,
+      "learning_rate": 4.758833804740015e-08,
+      "loss": 0.8739,
+      "step": 12419
+    },
+    {
+      "epoch": 0.9574468085106383,
+      "grad_norm": 4.12003755569458,
+      "learning_rate": 4.7416656241067175e-08,
+      "loss": 1.0081,
+      "step": 12420
+    },
+    {
+      "epoch": 0.9575238976256553,
+      "grad_norm": 3.9636361598968506,
+      "learning_rate": 4.724528320265542e-08,
+      "loss": 0.9851,
+      "step": 12421
+    },
+    {
+      "epoch": 0.9576009867406722,
+      "grad_norm": 3.665433406829834,
+      "learning_rate": 4.7074218942849134e-08,
+      "loss": 0.9087,
+      "step": 12422
+    },
+    {
+      "epoch": 0.9576780758556892,
+      "grad_norm": 3.686065673828125,
+      "learning_rate": 4.6903463472313114e-08,
+      "loss": 0.8049,
+      "step": 12423
+    },
+    {
+      "epoch": 0.9577551649707061,
+      "grad_norm": 3.6736299991607666,
+      "learning_rate": 4.6733016801693845e-08,
+      "loss": 0.9044,
+      "step": 12424
+    },
+    {
+      "epoch": 0.9578322540857231,
+      "grad_norm": 3.865446090698242,
+      "learning_rate": 4.6562878941617264e-08,
+      "loss": 1.0022,
+      "step": 12425
+    },
+    {
+      "epoch": 0.9579093432007401,
+      "grad_norm": 3.7858636379241943,
+      "learning_rate": 4.6393049902690446e-08,
+      "loss": 0.8994,
+      "step": 12426
+    },
+    {
+      "epoch": 0.957986432315757,
+      "grad_norm": 3.769582986831665,
+      "learning_rate": 4.6223529695502144e-08,
+      "loss": 0.8945,
+      "step": 12427
+    },
+    {
+      "epoch": 0.958063521430774,
+      "grad_norm": 3.338371753692627,
+      "learning_rate": 4.605431833062002e-08,
+      "loss": 0.8453,
+      "step": 12428
+    },
+    {
+      "epoch": 0.9581406105457909,
+      "grad_norm": 4.088877201080322,
+      "learning_rate": 4.588541581859396e-08,
+      "loss": 0.9191,
+      "step": 12429
+    },
+    {
+      "epoch": 0.9582176996608079,
+      "grad_norm": 3.5991716384887695,
+      "learning_rate": 4.5716822169954436e-08,
+      "loss": 0.9485,
+      "step": 12430
+    },
+    {
+      "epoch": 0.9582947887758249,
+      "grad_norm": 3.5620510578155518,
+      "learning_rate": 4.554853739521192e-08,
+      "loss": 0.9463,
+      "step": 12431
+    },
+    {
+      "epoch": 0.9583718778908418,
+      "grad_norm": 3.7672712802886963,
+      "learning_rate": 4.5380561504858586e-08,
+      "loss": 0.87,
+      "step": 12432
+    },
+    {
+      "epoch": 0.9584489670058588,
+      "grad_norm": 4.017693996429443,
+      "learning_rate": 4.521289450936661e-08,
+      "loss": 0.9304,
+      "step": 12433
+    },
+    {
+      "epoch": 0.9585260561208757,
+      "grad_norm": 3.617096185684204,
+      "learning_rate": 4.504553641918874e-08,
+      "loss": 0.8937,
+      "step": 12434
+    },
+    {
+      "epoch": 0.9586031452358927,
+      "grad_norm": 3.6982741355895996,
+      "learning_rate": 4.4878487244759964e-08,
+      "loss": 0.8577,
+      "step": 12435
+    },
+    {
+      "epoch": 0.9586802343509097,
+      "grad_norm": 3.6797823905944824,
+      "learning_rate": 4.471174699649361e-08,
+      "loss": 0.8549,
+      "step": 12436
+    },
+    {
+      "epoch": 0.9587573234659266,
+      "grad_norm": 4.134285926818848,
+      "learning_rate": 4.4545315684785815e-08,
+      "loss": 0.9191,
+      "step": 12437
+    },
+    {
+      "epoch": 0.9588344125809436,
+      "grad_norm": 3.8390088081359863,
+      "learning_rate": 4.437919332001328e-08,
+      "loss": 0.9041,
+      "step": 12438
+    },
+    {
+      "epoch": 0.9589115016959605,
+      "grad_norm": 3.7085273265838623,
+      "learning_rate": 4.421337991253161e-08,
+      "loss": 0.9955,
+      "step": 12439
+    },
+    {
+      "epoch": 0.9589885908109775,
+      "grad_norm": 4.025529384613037,
+      "learning_rate": 4.4047875472679194e-08,
+      "loss": 1.0332,
+      "step": 12440
+    },
+    {
+      "epoch": 0.9590656799259945,
+      "grad_norm": 3.5782909393310547,
+      "learning_rate": 4.3882680010774445e-08,
+      "loss": 0.988,
+      "step": 12441
+    },
+    {
+      "epoch": 0.9591427690410114,
+      "grad_norm": 3.8593509197235107,
+      "learning_rate": 4.3717793537115806e-08,
+      "loss": 0.9466,
+      "step": 12442
+    },
+    {
+      "epoch": 0.9592198581560284,
+      "grad_norm": 3.907071590423584,
+      "learning_rate": 4.355321606198393e-08,
+      "loss": 0.8672,
+      "step": 12443
+    },
+    {
+      "epoch": 0.9592969472710453,
+      "grad_norm": 4.352113246917725,
+      "learning_rate": 4.338894759563894e-08,
+      "loss": 0.9618,
+      "step": 12444
+    },
+    {
+      "epoch": 0.9593740363860623,
+      "grad_norm": 3.6431379318237305,
+      "learning_rate": 4.3224988148321545e-08,
+      "loss": 0.8185,
+      "step": 12445
+    },
+    {
+      "epoch": 0.9594511255010792,
+      "grad_norm": 3.652925729751587,
+      "learning_rate": 4.306133773025467e-08,
+      "loss": 0.8715,
+      "step": 12446
+    },
+    {
+      "epoch": 0.9595282146160962,
+      "grad_norm": 3.6305699348449707,
+      "learning_rate": 4.2897996351640715e-08,
+      "loss": 0.7884,
+      "step": 12447
+    },
+    {
+      "epoch": 0.9596053037311132,
+      "grad_norm": 3.712374448776245,
+      "learning_rate": 4.2734964022663194e-08,
+      "loss": 0.8094,
+      "step": 12448
+    },
+    {
+      "epoch": 0.9596823928461301,
+      "grad_norm": 3.571690797805786,
+      "learning_rate": 4.257224075348676e-08,
+      "loss": 0.7927,
+      "step": 12449
+    },
+    {
+      "epoch": 0.9597594819611471,
+      "grad_norm": 3.573230743408203,
+      "learning_rate": 4.240982655425552e-08,
+      "loss": 0.8144,
+      "step": 12450
+    },
+    {
+      "epoch": 0.959836571076164,
+      "grad_norm": 4.072946548461914,
+      "learning_rate": 4.224772143509526e-08,
+      "loss": 0.9237,
+      "step": 12451
+    },
+    {
+      "epoch": 0.959913660191181,
+      "grad_norm": 3.8101742267608643,
+      "learning_rate": 4.2085925406112894e-08,
+      "loss": 0.9525,
+      "step": 12452
+    },
+    {
+      "epoch": 0.959990749306198,
+      "grad_norm": 3.8451669216156006,
+      "learning_rate": 4.192443847739536e-08,
+      "loss": 0.8578,
+      "step": 12453
+    },
+    {
+      "epoch": 0.9600678384212149,
+      "grad_norm": 3.733152389526367,
+      "learning_rate": 4.1763260659011265e-08,
+      "loss": 0.9414,
+      "step": 12454
+    },
+    {
+      "epoch": 0.9601449275362319,
+      "grad_norm": 3.55338716506958,
+      "learning_rate": 4.160239196100757e-08,
+      "loss": 0.8958,
+      "step": 12455
+    },
+    {
+      "epoch": 0.9602220166512488,
+      "grad_norm": 3.659872531890869,
+      "learning_rate": 4.144183239341515e-08,
+      "loss": 0.9185,
+      "step": 12456
+    },
+    {
+      "epoch": 0.9602991057662658,
+      "grad_norm": 3.818180561065674,
+      "learning_rate": 4.1281581966243214e-08,
+      "loss": 0.9281,
+      "step": 12457
+    },
+    {
+      "epoch": 0.9603761948812828,
+      "grad_norm": 4.13081169128418,
+      "learning_rate": 4.1121640689482655e-08,
+      "loss": 0.992,
+      "step": 12458
+    },
+    {
+      "epoch": 0.9604532839962997,
+      "grad_norm": 3.2538528442382812,
+      "learning_rate": 4.0962008573105505e-08,
+      "loss": 0.7597,
+      "step": 12459
+    },
+    {
+      "epoch": 0.9605303731113167,
+      "grad_norm": 3.9792425632476807,
+      "learning_rate": 4.080268562706324e-08,
+      "loss": 0.8961,
+      "step": 12460
+    },
+    {
+      "epoch": 0.9606074622263336,
+      "grad_norm": 4.055764198303223,
+      "learning_rate": 4.0643671861289035e-08,
+      "loss": 0.8922,
+      "step": 12461
+    },
+    {
+      "epoch": 0.9606845513413506,
+      "grad_norm": 3.6515095233917236,
+      "learning_rate": 4.048496728569717e-08,
+      "loss": 0.8648,
+      "step": 12462
+    },
+    {
+      "epoch": 0.9607616404563676,
+      "grad_norm": 3.2547214031219482,
+      "learning_rate": 4.0326571910180855e-08,
+      "loss": 0.7415,
+      "step": 12463
+    },
+    {
+      "epoch": 0.9608387295713845,
+      "grad_norm": 3.92317271232605,
+      "learning_rate": 4.016848574461718e-08,
+      "loss": 0.9378,
+      "step": 12464
+    },
+    {
+      "epoch": 0.9609158186864015,
+      "grad_norm": 3.4518179893493652,
+      "learning_rate": 4.001070879885993e-08,
+      "loss": 0.7162,
+      "step": 12465
+    },
+    {
+      "epoch": 0.9609929078014184,
+      "grad_norm": 3.502856969833374,
+      "learning_rate": 3.9853241082746795e-08,
+      "loss": 0.8978,
+      "step": 12466
+    },
+    {
+      "epoch": 0.9610699969164354,
+      "grad_norm": 3.2858731746673584,
+      "learning_rate": 3.9696082606094366e-08,
+      "loss": 0.8471,
+      "step": 12467
+    },
+    {
+      "epoch": 0.9611470860314524,
+      "grad_norm": 3.7062535285949707,
+      "learning_rate": 3.953923337870147e-08,
+      "loss": 0.893,
+      "step": 12468
+    },
+    {
+      "epoch": 0.9612241751464693,
+      "grad_norm": 4.334238529205322,
+      "learning_rate": 3.938269341034695e-08,
+      "loss": 0.9144,
+      "step": 12469
+    },
+    {
+      "epoch": 0.9613012642614863,
+      "grad_norm": 3.6347129344940186,
+      "learning_rate": 3.922646271078911e-08,
+      "loss": 0.8314,
+      "step": 12470
+    },
+    {
+      "epoch": 0.9613783533765032,
+      "grad_norm": 4.064388751983643,
+      "learning_rate": 3.907054128976906e-08,
+      "loss": 0.9495,
+      "step": 12471
+    },
+    {
+      "epoch": 0.9614554424915202,
+      "grad_norm": 3.732004165649414,
+      "learning_rate": 3.8914929157007895e-08,
+      "loss": 0.9922,
+      "step": 12472
+    },
+    {
+      "epoch": 0.9615325316065372,
+      "grad_norm": 3.932603120803833,
+      "learning_rate": 3.87596263222062e-08,
+      "loss": 0.8979,
+      "step": 12473
+    },
+    {
+      "epoch": 0.9616096207215541,
+      "grad_norm": 3.8675358295440674,
+      "learning_rate": 3.860463279504678e-08,
+      "loss": 0.9025,
+      "step": 12474
+    },
+    {
+      "epoch": 0.9616867098365711,
+      "grad_norm": 3.780653715133667,
+      "learning_rate": 3.8449948585193577e-08,
+      "loss": 0.9724,
+      "step": 12475
+    },
+    {
+      "epoch": 0.961763798951588,
+      "grad_norm": 3.6268699169158936,
+      "learning_rate": 3.829557370228887e-08,
+      "loss": 0.8795,
+      "step": 12476
+    },
+    {
+      "epoch": 0.961840888066605,
+      "grad_norm": 3.926316976547241,
+      "learning_rate": 3.814150815595774e-08,
+      "loss": 0.8567,
+      "step": 12477
+    },
+    {
+      "epoch": 0.961917977181622,
+      "grad_norm": 4.029868125915527,
+      "learning_rate": 3.798775195580584e-08,
+      "loss": 0.837,
+      "step": 12478
+    },
+    {
+      "epoch": 0.9619950662966389,
+      "grad_norm": 3.769047260284424,
+      "learning_rate": 3.783430511141828e-08,
+      "loss": 0.7122,
+      "step": 12479
+    },
+    {
+      "epoch": 0.9620721554116559,
+      "grad_norm": 3.4524729251861572,
+      "learning_rate": 3.76811676323624e-08,
+      "loss": 0.7635,
+      "step": 12480
+    },
+    {
+      "epoch": 0.9621492445266728,
+      "grad_norm": 4.048440933227539,
+      "learning_rate": 3.752833952818502e-08,
+      "loss": 0.8472,
+      "step": 12481
+    },
+    {
+      "epoch": 0.9622263336416897,
+      "grad_norm": 3.5961873531341553,
+      "learning_rate": 3.737582080841462e-08,
+      "loss": 0.9834,
+      "step": 12482
+    },
+    {
+      "epoch": 0.9623034227567068,
+      "grad_norm": 4.128652095794678,
+      "learning_rate": 3.722361148255971e-08,
+      "loss": 0.8462,
+      "step": 12483
+    },
+    {
+      "epoch": 0.9623805118717237,
+      "grad_norm": 3.8181169033050537,
+      "learning_rate": 3.707171156010936e-08,
+      "loss": 0.9219,
+      "step": 12484
+    },
+    {
+      "epoch": 0.9624576009867407,
+      "grad_norm": 3.621793031692505,
+      "learning_rate": 3.692012105053433e-08,
+      "loss": 0.8931,
+      "step": 12485
+    },
+    {
+      "epoch": 0.9625346901017576,
+      "grad_norm": 3.9379565715789795,
+      "learning_rate": 3.6768839963285395e-08,
+      "loss": 0.8306,
+      "step": 12486
+    },
+    {
+      "epoch": 0.9626117792167745,
+      "grad_norm": 3.8023934364318848,
+      "learning_rate": 3.66178683077939e-08,
+      "loss": 0.97,
+      "step": 12487
+    },
+    {
+      "epoch": 0.9626888683317916,
+      "grad_norm": 3.626626968383789,
+      "learning_rate": 3.646720609347232e-08,
+      "loss": 0.9197,
+      "step": 12488
+    },
+    {
+      "epoch": 0.9627659574468085,
+      "grad_norm": 3.600571393966675,
+      "learning_rate": 3.6316853329713686e-08,
+      "loss": 0.9092,
+      "step": 12489
+    },
+    {
+      "epoch": 0.9628430465618255,
+      "grad_norm": 3.4085299968719482,
+      "learning_rate": 3.616681002589162e-08,
+      "loss": 0.7884,
+      "step": 12490
+    },
+    {
+      "epoch": 0.9629201356768424,
+      "grad_norm": 3.688203811645508,
+      "learning_rate": 3.601707619136086e-08,
+      "loss": 0.7688,
+      "step": 12491
+    },
+    {
+      "epoch": 0.9629972247918593,
+      "grad_norm": 3.6305623054504395,
+      "learning_rate": 3.586765183545615e-08,
+      "loss": 0.9648,
+      "step": 12492
+    },
+    {
+      "epoch": 0.9630743139068764,
+      "grad_norm": 3.8015410900115967,
+      "learning_rate": 3.571853696749339e-08,
+      "loss": 0.9645,
+      "step": 12493
+    },
+    {
+      "epoch": 0.9631514030218933,
+      "grad_norm": 3.945363998413086,
+      "learning_rate": 3.556973159676902e-08,
+      "loss": 0.92,
+      "step": 12494
+    },
+    {
+      "epoch": 0.9632284921369103,
+      "grad_norm": 3.6869564056396484,
+      "learning_rate": 3.542123573256062e-08,
+      "loss": 0.8454,
+      "step": 12495
+    },
+    {
+      "epoch": 0.9633055812519272,
+      "grad_norm": 4.102055549621582,
+      "learning_rate": 3.5273049384126345e-08,
+      "loss": 0.9861,
+      "step": 12496
+    },
+    {
+      "epoch": 0.9633826703669441,
+      "grad_norm": 4.952552318572998,
+      "learning_rate": 3.51251725607038e-08,
+      "loss": 0.8793,
+      "step": 12497
+    },
+    {
+      "epoch": 0.9634597594819612,
+      "grad_norm": 3.5862975120544434,
+      "learning_rate": 3.497760527151284e-08,
+      "loss": 0.8908,
+      "step": 12498
+    },
+    {
+      "epoch": 0.9635368485969781,
+      "grad_norm": 3.9423680305480957,
+      "learning_rate": 3.4830347525754425e-08,
+      "loss": 0.8734,
+      "step": 12499
+    },
+    {
+      "epoch": 0.9636139377119951,
+      "grad_norm": 4.143372058868408,
+      "learning_rate": 3.4683399332607893e-08,
+      "loss": 0.9683,
+      "step": 12500
+    },
+    {
+      "epoch": 0.963691026827012,
+      "grad_norm": 3.9046006202697754,
+      "learning_rate": 3.45367607012359e-08,
+      "loss": 0.9009,
+      "step": 12501
+    },
+    {
+      "epoch": 0.9637681159420289,
+      "grad_norm": 3.8234989643096924,
+      "learning_rate": 3.439043164078004e-08,
+      "loss": 1.0309,
+      "step": 12502
+    },
+    {
+      "epoch": 0.963845205057046,
+      "grad_norm": 3.920905828475952,
+      "learning_rate": 3.424441216036301e-08,
+      "loss": 0.9936,
+      "step": 12503
+    },
+    {
+      "epoch": 0.9639222941720629,
+      "grad_norm": 3.868464231491089,
+      "learning_rate": 3.409870226908863e-08,
+      "loss": 0.8631,
+      "step": 12504
+    },
+    {
+      "epoch": 0.9639993832870799,
+      "grad_norm": 3.8888497352600098,
+      "learning_rate": 3.3953301976040743e-08,
+      "loss": 0.9221,
+      "step": 12505
+    },
+    {
+      "epoch": 0.9640764724020968,
+      "grad_norm": 3.7030375003814697,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 0.8446,
+      "step": 12506
+    },
+    {
+      "epoch": 0.9641535615171137,
+      "grad_norm": 4.315587520599365,
+      "learning_rate": 3.36634302208666e-08,
+      "loss": 0.9185,
+      "step": 12507
+    },
+    {
+      "epoch": 0.9642306506321308,
+      "grad_norm": 3.6543986797332764,
+      "learning_rate": 3.351895877681255e-08,
+      "loss": 0.9264,
+      "step": 12508
+    },
+    {
+      "epoch": 0.9643077397471477,
+      "grad_norm": 3.710772752761841,
+      "learning_rate": 3.3374796967128866e-08,
+      "loss": 0.9397,
+      "step": 12509
+    },
+    {
+      "epoch": 0.9643848288621647,
+      "grad_norm": 3.6390933990478516,
+      "learning_rate": 3.3230944800803355e-08,
+      "loss": 0.8503,
+      "step": 12510
+    },
+    {
+      "epoch": 0.9644619179771816,
+      "grad_norm": 4.070563793182373,
+      "learning_rate": 3.3087402286805514e-08,
+      "loss": 0.8884,
+      "step": 12511
+    },
+    {
+      "epoch": 0.9645390070921985,
+      "grad_norm": 3.739609718322754,
+      "learning_rate": 3.294416943408374e-08,
+      "loss": 0.9062,
+      "step": 12512
+    },
+    {
+      "epoch": 0.9646160962072156,
+      "grad_norm": 3.311293840408325,
+      "learning_rate": 3.2801246251568106e-08,
+      "loss": 0.8365,
+      "step": 12513
+    },
+    {
+      "epoch": 0.9646931853222325,
+      "grad_norm": 3.451474666595459,
+      "learning_rate": 3.2658632748168714e-08,
+      "loss": 0.829,
+      "step": 12514
+    },
+    {
+      "epoch": 0.9647702744372495,
+      "grad_norm": 3.717940092086792,
+      "learning_rate": 3.2516328932776786e-08,
+      "loss": 0.8624,
+      "step": 12515
+    },
+    {
+      "epoch": 0.9648473635522664,
+      "grad_norm": 3.5586538314819336,
+      "learning_rate": 3.2374334814265216e-08,
+      "loss": 0.8706,
+      "step": 12516
+    },
+    {
+      "epoch": 0.9649244526672833,
+      "grad_norm": 3.5937881469726562,
+      "learning_rate": 3.223265040148527e-08,
+      "loss": 0.9791,
+      "step": 12517
+    },
+    {
+      "epoch": 0.9650015417823004,
+      "grad_norm": 3.6897778511047363,
+      "learning_rate": 3.209127570327153e-08,
+      "loss": 0.8629,
+      "step": 12518
+    },
+    {
+      "epoch": 0.9650786308973173,
+      "grad_norm": 3.6691691875457764,
+      "learning_rate": 3.19502107284364e-08,
+      "loss": 0.9409,
+      "step": 12519
+    },
+    {
+      "epoch": 0.9651557200123343,
+      "grad_norm": 3.7458150386810303,
+      "learning_rate": 3.180945548577619e-08,
+      "loss": 0.9058,
+      "step": 12520
+    },
+    {
+      "epoch": 0.9652328091273512,
+      "grad_norm": 3.7683491706848145,
+      "learning_rate": 3.166900998406497e-08,
+      "loss": 0.9012,
+      "step": 12521
+    },
+    {
+      "epoch": 0.9653098982423681,
+      "grad_norm": 3.705265522003174,
+      "learning_rate": 3.1528874232059635e-08,
+      "loss": 0.813,
+      "step": 12522
+    },
+    {
+      "epoch": 0.9653869873573852,
+      "grad_norm": 3.670219898223877,
+      "learning_rate": 3.138904823849653e-08,
+      "loss": 0.8558,
+      "step": 12523
+    },
+    {
+      "epoch": 0.9654640764724021,
+      "grad_norm": 3.741985321044922,
+      "learning_rate": 3.124953201209313e-08,
+      "loss": 0.9858,
+      "step": 12524
+    },
+    {
+      "epoch": 0.9655411655874191,
+      "grad_norm": 3.5063395500183105,
+      "learning_rate": 3.1110325561547473e-08,
+      "loss": 0.7742,
+      "step": 12525
+    },
+    {
+      "epoch": 0.965618254702436,
+      "grad_norm": 3.4746077060699463,
+      "learning_rate": 3.097142889553872e-08,
+      "loss": 0.8353,
+      "step": 12526
+    },
+    {
+      "epoch": 0.9656953438174529,
+      "grad_norm": 3.630375385284424,
+      "learning_rate": 3.083284202272607e-08,
+      "loss": 0.84,
+      "step": 12527
+    },
+    {
+      "epoch": 0.96577243293247,
+      "grad_norm": 4.338189125061035,
+      "learning_rate": 3.069456495174983e-08,
+      "loss": 0.8878,
+      "step": 12528
+    },
+    {
+      "epoch": 0.9658495220474869,
+      "grad_norm": 3.834465503692627,
+      "learning_rate": 3.0556597691230317e-08,
+      "loss": 0.8431,
+      "step": 12529
+    },
+    {
+      "epoch": 0.9659266111625039,
+      "grad_norm": 3.5252437591552734,
+      "learning_rate": 3.04189402497701e-08,
+      "loss": 0.7253,
+      "step": 12530
+    },
+    {
+      "epoch": 0.9660037002775208,
+      "grad_norm": 3.7228147983551025,
+      "learning_rate": 3.028159263595065e-08,
+      "loss": 0.9094,
+      "step": 12531
+    },
+    {
+      "epoch": 0.9660807893925377,
+      "grad_norm": 4.035754680633545,
+      "learning_rate": 3.0144554858334563e-08,
+      "loss": 0.994,
+      "step": 12532
+    },
+    {
+      "epoch": 0.9661578785075547,
+      "grad_norm": 4.05665922164917,
+      "learning_rate": 3.000782692546667e-08,
+      "loss": 0.869,
+      "step": 12533
+    },
+    {
+      "epoch": 0.9662349676225717,
+      "grad_norm": 3.910731554031372,
+      "learning_rate": 2.9871408845870144e-08,
+      "loss": 0.9663,
+      "step": 12534
+    },
+    {
+      "epoch": 0.9663120567375887,
+      "grad_norm": 4.1619696617126465,
+      "learning_rate": 2.973530062805041e-08,
+      "loss": 0.8635,
+      "step": 12535
+    },
+    {
+      "epoch": 0.9663891458526056,
+      "grad_norm": 3.829831838607788,
+      "learning_rate": 2.959950228049291e-08,
+      "loss": 0.9295,
+      "step": 12536
+    },
+    {
+      "epoch": 0.9664662349676225,
+      "grad_norm": 3.9368739128112793,
+      "learning_rate": 2.9464013811664192e-08,
+      "loss": 0.8815,
+      "step": 12537
+    },
+    {
+      "epoch": 0.9665433240826395,
+      "grad_norm": 3.4988670349121094,
+      "learning_rate": 2.9328835230011398e-08,
+      "loss": 0.8505,
+      "step": 12538
+    },
+    {
+      "epoch": 0.9666204131976565,
+      "grad_norm": 3.812965154647827,
+      "learning_rate": 2.9193966543961673e-08,
+      "loss": 0.8958,
+      "step": 12539
+    },
+    {
+      "epoch": 0.9666975023126735,
+      "grad_norm": 3.911647081375122,
+      "learning_rate": 2.905940776192384e-08,
+      "loss": 0.9759,
+      "step": 12540
+    },
+    {
+      "epoch": 0.9667745914276904,
+      "grad_norm": 3.842432975769043,
+      "learning_rate": 2.89251588922862e-08,
+      "loss": 0.9148,
+      "step": 12541
+    },
+    {
+      "epoch": 0.9668516805427073,
+      "grad_norm": 3.5029232501983643,
+      "learning_rate": 2.8791219943419824e-08,
+      "loss": 0.7411,
+      "step": 12542
+    },
+    {
+      "epoch": 0.9669287696577243,
+      "grad_norm": 3.9447033405303955,
+      "learning_rate": 2.8657590923673596e-08,
+      "loss": 0.9414,
+      "step": 12543
+    },
+    {
+      "epoch": 0.9670058587727413,
+      "grad_norm": 3.7002828121185303,
+      "learning_rate": 2.852427184138029e-08,
+      "loss": 0.9014,
+      "step": 12544
+    },
+    {
+      "epoch": 0.9670829478877583,
+      "grad_norm": 3.4802746772766113,
+      "learning_rate": 2.839126270484993e-08,
+      "loss": 0.8871,
+      "step": 12545
+    },
+    {
+      "epoch": 0.9671600370027752,
+      "grad_norm": 3.994755744934082,
+      "learning_rate": 2.825856352237588e-08,
+      "loss": 0.8622,
+      "step": 12546
+    },
+    {
+      "epoch": 0.9672371261177922,
+      "grad_norm": 3.6245877742767334,
+      "learning_rate": 2.812617430223097e-08,
+      "loss": 0.802,
+      "step": 12547
+    },
+    {
+      "epoch": 0.9673142152328091,
+      "grad_norm": 4.391573429107666,
+      "learning_rate": 2.7994095052669147e-08,
+      "loss": 0.9172,
+      "step": 12548
+    },
+    {
+      "epoch": 0.967391304347826,
+      "grad_norm": 3.789491653442383,
+      "learning_rate": 2.786232578192494e-08,
+      "loss": 0.9283,
+      "step": 12549
+    },
+    {
+      "epoch": 0.9674683934628431,
+      "grad_norm": 3.5741639137268066,
+      "learning_rate": 2.773086649821344e-08,
+      "loss": 0.9144,
+      "step": 12550
+    },
+    {
+      "epoch": 0.96754548257786,
+      "grad_norm": 3.636545181274414,
+      "learning_rate": 2.7599717209730316e-08,
+      "loss": 0.823,
+      "step": 12551
+    },
+    {
+      "epoch": 0.967622571692877,
+      "grad_norm": 3.7562508583068848,
+      "learning_rate": 2.7468877924651804e-08,
+      "loss": 0.9054,
+      "step": 12552
+    },
+    {
+      "epoch": 0.9676996608078939,
+      "grad_norm": 3.873007297515869,
+      "learning_rate": 2.733834865113527e-08,
+      "loss": 0.9583,
+      "step": 12553
+    },
+    {
+      "epoch": 0.9677767499229109,
+      "grad_norm": 3.7551496028900146,
+      "learning_rate": 2.7208129397319206e-08,
+      "loss": 0.8355,
+      "step": 12554
+    },
+    {
+      "epoch": 0.9678538390379279,
+      "grad_norm": 3.8303260803222656,
+      "learning_rate": 2.7078220171321556e-08,
+      "loss": 0.8735,
+      "step": 12555
+    },
+    {
+      "epoch": 0.9679309281529448,
+      "grad_norm": 3.5815868377685547,
+      "learning_rate": 2.6948620981240848e-08,
+      "loss": 0.7389,
+      "step": 12556
+    },
+    {
+      "epoch": 0.9680080172679618,
+      "grad_norm": 3.6111199855804443,
+      "learning_rate": 2.6819331835157836e-08,
+      "loss": 0.7886,
+      "step": 12557
+    },
+    {
+      "epoch": 0.9680851063829787,
+      "grad_norm": 3.4187819957733154,
+      "learning_rate": 2.669035274113274e-08,
+      "loss": 0.8699,
+      "step": 12558
+    },
+    {
+      "epoch": 0.9681621954979956,
+      "grad_norm": 3.8764898777008057,
+      "learning_rate": 2.6561683707206354e-08,
+      "loss": 0.921,
+      "step": 12559
+    },
+    {
+      "epoch": 0.9682392846130127,
+      "grad_norm": 3.894237756729126,
+      "learning_rate": 2.6433324741401702e-08,
+      "loss": 0.9434,
+      "step": 12560
+    },
+    {
+      "epoch": 0.9683163737280296,
+      "grad_norm": 3.7127158641815186,
+      "learning_rate": 2.6305275851720158e-08,
+      "loss": 0.8949,
+      "step": 12561
+    },
+    {
+      "epoch": 0.9683934628430466,
+      "grad_norm": 3.729252576828003,
+      "learning_rate": 2.6177537046144786e-08,
+      "loss": 0.8965,
+      "step": 12562
+    },
+    {
+      "epoch": 0.9684705519580635,
+      "grad_norm": 4.151895523071289,
+      "learning_rate": 2.6050108332640323e-08,
+      "loss": 1.0355,
+      "step": 12563
+    },
+    {
+      "epoch": 0.9685476410730804,
+      "grad_norm": 3.9273674488067627,
+      "learning_rate": 2.5922989719150414e-08,
+      "loss": 0.8994,
+      "step": 12564
+    },
+    {
+      "epoch": 0.9686247301880975,
+      "grad_norm": 3.550187349319458,
+      "learning_rate": 2.5796181213601503e-08,
+      "loss": 0.9353,
+      "step": 12565
+    },
+    {
+      "epoch": 0.9687018193031144,
+      "grad_norm": 3.781679153442383,
+      "learning_rate": 2.5669682823897813e-08,
+      "loss": 0.9055,
+      "step": 12566
+    },
+    {
+      "epoch": 0.9687789084181314,
+      "grad_norm": 3.9291346073150635,
+      "learning_rate": 2.554349455792693e-08,
+      "loss": 0.8633,
+      "step": 12567
+    },
+    {
+      "epoch": 0.9688559975331483,
+      "grad_norm": 4.16334867477417,
+      "learning_rate": 2.541761642355589e-08,
+      "loss": 0.9427,
+      "step": 12568
+    },
+    {
+      "epoch": 0.9689330866481652,
+      "grad_norm": 3.531303882598877,
+      "learning_rate": 2.529204842863231e-08,
+      "loss": 0.773,
+      "step": 12569
+    },
+    {
+      "epoch": 0.9690101757631823,
+      "grad_norm": 3.517418146133423,
+      "learning_rate": 2.5166790580984923e-08,
+      "loss": 0.8888,
+      "step": 12570
+    },
+    {
+      "epoch": 0.9690872648781992,
+      "grad_norm": 3.5748257637023926,
+      "learning_rate": 2.504184288842304e-08,
+      "loss": 0.9294,
+      "step": 12571
+    },
+    {
+      "epoch": 0.9691643539932162,
+      "grad_norm": 3.5963714122772217,
+      "learning_rate": 2.4917205358735986e-08,
+      "loss": 0.9087,
+      "step": 12572
+    },
+    {
+      "epoch": 0.9692414431082331,
+      "grad_norm": 3.668978691101074,
+      "learning_rate": 2.4792877999694764e-08,
+      "loss": 0.8818,
+      "step": 12573
+    },
+    {
+      "epoch": 0.96931853222325,
+      "grad_norm": 3.8249216079711914,
+      "learning_rate": 2.46688608190504e-08,
+      "loss": 0.9889,
+      "step": 12574
+    },
+    {
+      "epoch": 0.9693956213382671,
+      "grad_norm": 3.7434964179992676,
+      "learning_rate": 2.4545153824534483e-08,
+      "loss": 0.906,
+      "step": 12575
+    },
+    {
+      "epoch": 0.969472710453284,
+      "grad_norm": 3.613534450531006,
+      "learning_rate": 2.4421757023859737e-08,
+      "loss": 0.8281,
+      "step": 12576
+    },
+    {
+      "epoch": 0.969549799568301,
+      "grad_norm": 3.6317696571350098,
+      "learning_rate": 2.4298670424718895e-08,
+      "loss": 0.8908,
+      "step": 12577
+    },
+    {
+      "epoch": 0.9696268886833179,
+      "grad_norm": 3.982708215713501,
+      "learning_rate": 2.417589403478693e-08,
+      "loss": 0.857,
+      "step": 12578
+    },
+    {
+      "epoch": 0.9697039777983348,
+      "grad_norm": 3.8874056339263916,
+      "learning_rate": 2.4053427861716605e-08,
+      "loss": 0.9243,
+      "step": 12579
+    },
+    {
+      "epoch": 0.9697810669133519,
+      "grad_norm": 3.79288649559021,
+      "learning_rate": 2.3931271913144595e-08,
+      "loss": 0.9047,
+      "step": 12580
+    },
+    {
+      "epoch": 0.9698581560283688,
+      "grad_norm": 3.9113500118255615,
+      "learning_rate": 2.3809426196685913e-08,
+      "loss": 1.0422,
+      "step": 12581
+    },
+    {
+      "epoch": 0.9699352451433858,
+      "grad_norm": 3.4592325687408447,
+      "learning_rate": 2.3687890719937266e-08,
+      "loss": 0.7883,
+      "step": 12582
+    },
+    {
+      "epoch": 0.9700123342584027,
+      "grad_norm": 4.037449836730957,
+      "learning_rate": 2.3566665490475372e-08,
+      "loss": 0.9285,
+      "step": 12583
+    },
+    {
+      "epoch": 0.9700894233734196,
+      "grad_norm": 3.595924139022827,
+      "learning_rate": 2.3445750515858067e-08,
+      "loss": 0.8422,
+      "step": 12584
+    },
+    {
+      "epoch": 0.9701665124884367,
+      "grad_norm": 3.691532611846924,
+      "learning_rate": 2.3325145803624327e-08,
+      "loss": 0.9454,
+      "step": 12585
+    },
+    {
+      "epoch": 0.9702436016034536,
+      "grad_norm": 3.8150765895843506,
+      "learning_rate": 2.3204851361293136e-08,
+      "loss": 0.944,
+      "step": 12586
+    },
+    {
+      "epoch": 0.9703206907184706,
+      "grad_norm": 4.036136627197266,
+      "learning_rate": 2.308486719636349e-08,
+      "loss": 0.9124,
+      "step": 12587
+    },
+    {
+      "epoch": 0.9703977798334875,
+      "grad_norm": 3.491523504257202,
+      "learning_rate": 2.2965193316316635e-08,
+      "loss": 0.9327,
+      "step": 12588
+    },
+    {
+      "epoch": 0.9704748689485044,
+      "grad_norm": 4.169717788696289,
+      "learning_rate": 2.2845829728613268e-08,
+      "loss": 0.8412,
+      "step": 12589
+    },
+    {
+      "epoch": 0.9705519580635215,
+      "grad_norm": 3.619222640991211,
+      "learning_rate": 2.2726776440694653e-08,
+      "loss": 0.8712,
+      "step": 12590
+    },
+    {
+      "epoch": 0.9706290471785384,
+      "grad_norm": 3.93977427482605,
+      "learning_rate": 2.2608033459983747e-08,
+      "loss": 0.9405,
+      "step": 12591
+    },
+    {
+      "epoch": 0.9707061362935554,
+      "grad_norm": 3.866957187652588,
+      "learning_rate": 2.2489600793883516e-08,
+      "loss": 0.9402,
+      "step": 12592
+    },
+    {
+      "epoch": 0.9707832254085723,
+      "grad_norm": 3.519181966781616,
+      "learning_rate": 2.2371478449777495e-08,
+      "loss": 0.8973,
+      "step": 12593
+    },
+    {
+      "epoch": 0.9708603145235892,
+      "grad_norm": 3.5281713008880615,
+      "learning_rate": 2.2253666435029797e-08,
+      "loss": 0.839,
+      "step": 12594
+    },
+    {
+      "epoch": 0.9709374036386063,
+      "grad_norm": 3.604396343231201,
+      "learning_rate": 2.2136164756985655e-08,
+      "loss": 0.8367,
+      "step": 12595
+    },
+    {
+      "epoch": 0.9710144927536232,
+      "grad_norm": 3.734344720840454,
+      "learning_rate": 2.201897342297088e-08,
+      "loss": 0.9964,
+      "step": 12596
+    },
+    {
+      "epoch": 0.9710915818686402,
+      "grad_norm": 4.021172523498535,
+      "learning_rate": 2.190209244029129e-08,
+      "loss": 0.7933,
+      "step": 12597
+    },
+    {
+      "epoch": 0.9711686709836571,
+      "grad_norm": 3.9066355228424072,
+      "learning_rate": 2.1785521816233834e-08,
+      "loss": 1.0668,
+      "step": 12598
+    },
+    {
+      "epoch": 0.971245760098674,
+      "grad_norm": 4.022904396057129,
+      "learning_rate": 2.166926155806659e-08,
+      "loss": 0.9402,
+      "step": 12599
+    },
+    {
+      "epoch": 0.971322849213691,
+      "grad_norm": 3.6551458835601807,
+      "learning_rate": 2.155331167303709e-08,
+      "loss": 0.8959,
+      "step": 12600
+    },
+    {
+      "epoch": 0.971399938328708,
+      "grad_norm": 3.6164748668670654,
+      "learning_rate": 2.1437672168374557e-08,
+      "loss": 0.8252,
+      "step": 12601
+    },
+    {
+      "epoch": 0.971477027443725,
+      "grad_norm": 3.7705652713775635,
+      "learning_rate": 2.1322343051289328e-08,
+      "loss": 0.976,
+      "step": 12602
+    },
+    {
+      "epoch": 0.9715541165587419,
+      "grad_norm": 3.6110386848449707,
+      "learning_rate": 2.1207324328970103e-08,
+      "loss": 0.925,
+      "step": 12603
+    },
+    {
+      "epoch": 0.9716312056737588,
+      "grad_norm": 3.571214199066162,
+      "learning_rate": 2.1092616008588364e-08,
+      "loss": 0.9206,
+      "step": 12604
+    },
+    {
+      "epoch": 0.9717082947887759,
+      "grad_norm": 3.6576857566833496,
+      "learning_rate": 2.0978218097295612e-08,
+      "loss": 0.8903,
+      "step": 12605
+    },
+    {
+      "epoch": 0.9717853839037928,
+      "grad_norm": 3.6607632637023926,
+      "learning_rate": 2.086413060222392e-08,
+      "loss": 0.8791,
+      "step": 12606
+    },
+    {
+      "epoch": 0.9718624730188098,
+      "grad_norm": 3.758730888366699,
+      "learning_rate": 2.0750353530485932e-08,
+      "loss": 0.8838,
+      "step": 12607
+    },
+    {
+      "epoch": 0.9719395621338267,
+      "grad_norm": 4.265690326690674,
+      "learning_rate": 2.0636886889175978e-08,
+      "loss": 0.9786,
+      "step": 12608
+    },
+    {
+      "epoch": 0.9720166512488436,
+      "grad_norm": 3.9939539432525635,
+      "learning_rate": 2.0523730685366726e-08,
+      "loss": 0.9212,
+      "step": 12609
+    },
+    {
+      "epoch": 0.9720937403638606,
+      "grad_norm": 3.4965243339538574,
+      "learning_rate": 2.0410884926113094e-08,
+      "loss": 0.9082,
+      "step": 12610
+    },
+    {
+      "epoch": 0.9721708294788776,
+      "grad_norm": 4.189631938934326,
+      "learning_rate": 2.0298349618451673e-08,
+      "loss": 0.9899,
+      "step": 12611
+    },
+    {
+      "epoch": 0.9722479185938946,
+      "grad_norm": 3.7480947971343994,
+      "learning_rate": 2.0186124769396855e-08,
+      "loss": 0.9383,
+      "step": 12612
+    },
+    {
+      "epoch": 0.9723250077089115,
+      "grad_norm": 3.6697635650634766,
+      "learning_rate": 2.0074210385946925e-08,
+      "loss": 0.9815,
+      "step": 12613
+    },
+    {
+      "epoch": 0.9724020968239284,
+      "grad_norm": 3.7893688678741455,
+      "learning_rate": 1.996260647507797e-08,
+      "loss": 1.0289,
+      "step": 12614
+    },
+    {
+      "epoch": 0.9724791859389454,
+      "grad_norm": 3.7555932998657227,
+      "learning_rate": 1.9851313043747767e-08,
+      "loss": 0.8935,
+      "step": 12615
+    },
+    {
+      "epoch": 0.9725562750539624,
+      "grad_norm": 3.678717851638794,
+      "learning_rate": 1.9740330098895755e-08,
+      "loss": 0.8674,
+      "step": 12616
+    },
+    {
+      "epoch": 0.9726333641689794,
+      "grad_norm": 3.5781774520874023,
+      "learning_rate": 1.962965764744029e-08,
+      "loss": 0.9078,
+      "step": 12617
+    },
+    {
+      "epoch": 0.9727104532839963,
+      "grad_norm": 3.7126963138580322,
+      "learning_rate": 1.951929569628197e-08,
+      "loss": 0.8724,
+      "step": 12618
+    },
+    {
+      "epoch": 0.9727875423990132,
+      "grad_norm": 3.61206316947937,
+      "learning_rate": 1.9409244252301396e-08,
+      "loss": 0.788,
+      "step": 12619
+    },
+    {
+      "epoch": 0.9728646315140302,
+      "grad_norm": 3.783534049987793,
+      "learning_rate": 1.929950332235864e-08,
+      "loss": 0.9197,
+      "step": 12620
+    },
+    {
+      "epoch": 0.9729417206290472,
+      "grad_norm": 3.796304225921631,
+      "learning_rate": 1.9190072913296555e-08,
+      "loss": 0.8634,
+      "step": 12621
+    },
+    {
+      "epoch": 0.9730188097440642,
+      "grad_norm": 4.072221755981445,
+      "learning_rate": 1.908095303193691e-08,
+      "loss": 0.9239,
+      "step": 12622
+    },
+    {
+      "epoch": 0.9730958988590811,
+      "grad_norm": 4.112748146057129,
+      "learning_rate": 1.8972143685083155e-08,
+      "loss": 0.9495,
+      "step": 12623
+    },
+    {
+      "epoch": 0.973172987974098,
+      "grad_norm": 3.8566768169403076,
+      "learning_rate": 1.886364487951875e-08,
+      "loss": 0.9331,
+      "step": 12624
+    },
+    {
+      "epoch": 0.973250077089115,
+      "grad_norm": 3.7345011234283447,
+      "learning_rate": 1.8755456622008283e-08,
+      "loss": 0.8506,
+      "step": 12625
+    },
+    {
+      "epoch": 0.973327166204132,
+      "grad_norm": 4.711991310119629,
+      "learning_rate": 1.864757891929636e-08,
+      "loss": 0.934,
+      "step": 12626
+    },
+    {
+      "epoch": 0.973404255319149,
+      "grad_norm": 4.006261825561523,
+      "learning_rate": 1.8540011778108714e-08,
+      "loss": 1.0348,
+      "step": 12627
+    },
+    {
+      "epoch": 0.9734813444341659,
+      "grad_norm": 3.980159282684326,
+      "learning_rate": 1.84327552051522e-08,
+      "loss": 0.8444,
+      "step": 12628
+    },
+    {
+      "epoch": 0.9735584335491828,
+      "grad_norm": 3.6981513500213623,
+      "learning_rate": 1.8325809207112576e-08,
+      "loss": 0.9964,
+      "step": 12629
+    },
+    {
+      "epoch": 0.9736355226641998,
+      "grad_norm": 3.5920050144195557,
+      "learning_rate": 1.8219173790658406e-08,
+      "loss": 0.8259,
+      "step": 12630
+    },
+    {
+      "epoch": 0.9737126117792168,
+      "grad_norm": 3.436185836791992,
+      "learning_rate": 1.8112848962437146e-08,
+      "loss": 0.8754,
+      "step": 12631
+    },
+    {
+      "epoch": 0.9737897008942338,
+      "grad_norm": 3.517681360244751,
+      "learning_rate": 1.800683472907794e-08,
+      "loss": 0.8284,
+      "step": 12632
+    },
+    {
+      "epoch": 0.9738667900092507,
+      "grad_norm": 3.979491710662842,
+      "learning_rate": 1.7901131097190494e-08,
+      "loss": 0.9857,
+      "step": 12633
+    },
+    {
+      "epoch": 0.9739438791242676,
+      "grad_norm": 3.7884466648101807,
+      "learning_rate": 1.7795738073364543e-08,
+      "loss": 0.8687,
+      "step": 12634
+    },
+    {
+      "epoch": 0.9740209682392846,
+      "grad_norm": 3.576528787612915,
+      "learning_rate": 1.7690655664170388e-08,
+      "loss": 0.7963,
+      "step": 12635
+    },
+    {
+      "epoch": 0.9740980573543015,
+      "grad_norm": 4.090680122375488,
+      "learning_rate": 1.7585883876160003e-08,
+      "loss": 0.8982,
+      "step": 12636
+    },
+    {
+      "epoch": 0.9741751464693186,
+      "grad_norm": 3.728907346725464,
+      "learning_rate": 1.7481422715865394e-08,
+      "loss": 0.8825,
+      "step": 12637
+    },
+    {
+      "epoch": 0.9742522355843355,
+      "grad_norm": 3.741525411605835,
+      "learning_rate": 1.737727218979912e-08,
+      "loss": 0.9252,
+      "step": 12638
+    },
+    {
+      "epoch": 0.9743293246993524,
+      "grad_norm": 3.8002474308013916,
+      "learning_rate": 1.7273432304453774e-08,
+      "loss": 0.7882,
+      "step": 12639
+    },
+    {
+      "epoch": 0.9744064138143694,
+      "grad_norm": 3.745436191558838,
+      "learning_rate": 1.7169903066303618e-08,
+      "loss": 0.9055,
+      "step": 12640
+    },
+    {
+      "epoch": 0.9744835029293863,
+      "grad_norm": 3.9796302318573,
+      "learning_rate": 1.7066684481803486e-08,
+      "loss": 0.9246,
+      "step": 12641
+    },
+    {
+      "epoch": 0.9745605920444034,
+      "grad_norm": 3.8389484882354736,
+      "learning_rate": 1.6963776557388235e-08,
+      "loss": 0.82,
+      "step": 12642
+    },
+    {
+      "epoch": 0.9746376811594203,
+      "grad_norm": 3.872706651687622,
+      "learning_rate": 1.686117929947384e-08,
+      "loss": 0.9159,
+      "step": 12643
+    },
+    {
+      "epoch": 0.9747147702744372,
+      "grad_norm": 4.000880241394043,
+      "learning_rate": 1.6758892714456853e-08,
+      "loss": 0.8,
+      "step": 12644
+    },
+    {
+      "epoch": 0.9747918593894542,
+      "grad_norm": 3.820415735244751,
+      "learning_rate": 1.665691680871384e-08,
+      "loss": 0.9869,
+      "step": 12645
+    },
+    {
+      "epoch": 0.9748689485044711,
+      "grad_norm": 3.617678165435791,
+      "learning_rate": 1.6555251588602496e-08,
+      "loss": 0.9262,
+      "step": 12646
+    },
+    {
+      "epoch": 0.9749460376194882,
+      "grad_norm": 3.6912779808044434,
+      "learning_rate": 1.6453897060461076e-08,
+      "loss": 0.9492,
+      "step": 12647
+    },
+    {
+      "epoch": 0.9750231267345051,
+      "grad_norm": 3.9584383964538574,
+      "learning_rate": 1.6352853230609534e-08,
+      "loss": 0.972,
+      "step": 12648
+    },
+    {
+      "epoch": 0.975100215849522,
+      "grad_norm": 3.8923022747039795,
+      "learning_rate": 1.6252120105345604e-08,
+      "loss": 0.993,
+      "step": 12649
+    },
+    {
+      "epoch": 0.975177304964539,
+      "grad_norm": 3.5396828651428223,
+      "learning_rate": 1.6151697690951484e-08,
+      "loss": 0.83,
+      "step": 12650
+    },
+    {
+      "epoch": 0.9752543940795559,
+      "grad_norm": 3.774773359298706,
+      "learning_rate": 1.6051585993686614e-08,
+      "loss": 0.8335,
+      "step": 12651
+    },
+    {
+      "epoch": 0.975331483194573,
+      "grad_norm": 3.4751946926116943,
+      "learning_rate": 1.5951785019792666e-08,
+      "loss": 0.8478,
+      "step": 12652
+    },
+    {
+      "epoch": 0.9754085723095899,
+      "grad_norm": 3.7243905067443848,
+      "learning_rate": 1.5852294775491882e-08,
+      "loss": 0.8084,
+      "step": 12653
+    },
+    {
+      "epoch": 0.9754856614246068,
+      "grad_norm": 3.5833497047424316,
+      "learning_rate": 1.575311526698653e-08,
+      "loss": 0.7971,
+      "step": 12654
+    },
+    {
+      "epoch": 0.9755627505396238,
+      "grad_norm": 3.848736524581909,
+      "learning_rate": 1.56542465004611e-08,
+      "loss": 0.9441,
+      "step": 12655
+    },
+    {
+      "epoch": 0.9756398396546407,
+      "grad_norm": 3.8537116050720215,
+      "learning_rate": 1.5555688482078446e-08,
+      "loss": 0.8782,
+      "step": 12656
+    },
+    {
+      "epoch": 0.9757169287696578,
+      "grad_norm": 3.7755062580108643,
+      "learning_rate": 1.545744121798365e-08,
+      "loss": 0.823,
+      "step": 12657
+    },
+    {
+      "epoch": 0.9757940178846747,
+      "grad_norm": 3.62941837310791,
+      "learning_rate": 1.535950471430181e-08,
+      "loss": 0.8467,
+      "step": 12658
+    },
+    {
+      "epoch": 0.9758711069996916,
+      "grad_norm": 3.677624464035034,
+      "learning_rate": 1.52618789771386e-08,
+      "loss": 0.9067,
+      "step": 12659
+    },
+    {
+      "epoch": 0.9759481961147086,
+      "grad_norm": 3.693296432495117,
+      "learning_rate": 1.516456401258082e-08,
+      "loss": 0.8758,
+      "step": 12660
+    },
+    {
+      "epoch": 0.9760252852297255,
+      "grad_norm": 4.068735122680664,
+      "learning_rate": 1.5067559826695277e-08,
+      "loss": 1.0129,
+      "step": 12661
+    },
+    {
+      "epoch": 0.9761023743447426,
+      "grad_norm": 3.9047787189483643,
+      "learning_rate": 1.4970866425529916e-08,
+      "loss": 0.8848,
+      "step": 12662
+    },
+    {
+      "epoch": 0.9761794634597595,
+      "grad_norm": 3.9053268432617188,
+      "learning_rate": 1.4874483815112694e-08,
+      "loss": 0.9414,
+      "step": 12663
+    },
+    {
+      "epoch": 0.9762565525747764,
+      "grad_norm": 3.717155933380127,
+      "learning_rate": 1.477841200145269e-08,
+      "loss": 0.8827,
+      "step": 12664
+    },
+    {
+      "epoch": 0.9763336416897934,
+      "grad_norm": 3.714543581008911,
+      "learning_rate": 1.468265099053956e-08,
+      "loss": 0.9911,
+      "step": 12665
+    },
+    {
+      "epoch": 0.9764107308048103,
+      "grad_norm": 3.9254050254821777,
+      "learning_rate": 1.4587200788343524e-08,
+      "loss": 0.9414,
+      "step": 12666
+    },
+    {
+      "epoch": 0.9764878199198274,
+      "grad_norm": 3.731003999710083,
+      "learning_rate": 1.4492061400815383e-08,
+      "loss": 0.8444,
+      "step": 12667
+    },
+    {
+      "epoch": 0.9765649090348443,
+      "grad_norm": 3.458054542541504,
+      "learning_rate": 1.4397232833887053e-08,
+      "loss": 0.717,
+      "step": 12668
+    },
+    {
+      "epoch": 0.9766419981498612,
+      "grad_norm": 4.42410945892334,
+      "learning_rate": 1.4302715093469365e-08,
+      "loss": 0.8526,
+      "step": 12669
+    },
+    {
+      "epoch": 0.9767190872648782,
+      "grad_norm": 3.618612289428711,
+      "learning_rate": 1.4208508185456493e-08,
+      "loss": 0.9025,
+      "step": 12670
+    },
+    {
+      "epoch": 0.9767961763798951,
+      "grad_norm": 3.267320156097412,
+      "learning_rate": 1.4114612115720961e-08,
+      "loss": 0.8451,
+      "step": 12671
+    },
+    {
+      "epoch": 0.9768732654949122,
+      "grad_norm": 3.8861756324768066,
+      "learning_rate": 1.4021026890116418e-08,
+      "loss": 0.9779,
+      "step": 12672
+    },
+    {
+      "epoch": 0.9769503546099291,
+      "grad_norm": 3.7133381366729736,
+      "learning_rate": 1.3927752514478198e-08,
+      "loss": 0.9188,
+      "step": 12673
+    },
+    {
+      "epoch": 0.977027443724946,
+      "grad_norm": 4.010810375213623,
+      "learning_rate": 1.383478899462054e-08,
+      "loss": 0.9097,
+      "step": 12674
+    },
+    {
+      "epoch": 0.977104532839963,
+      "grad_norm": 4.118958473205566,
+      "learning_rate": 1.3742136336340473e-08,
+      "loss": 0.9476,
+      "step": 12675
+    },
+    {
+      "epoch": 0.9771816219549799,
+      "grad_norm": 4.091235160827637,
+      "learning_rate": 1.3649794545413376e-08,
+      "loss": 0.9998,
+      "step": 12676
+    },
+    {
+      "epoch": 0.977258711069997,
+      "grad_norm": 3.626465082168579,
+      "learning_rate": 1.3557763627596309e-08,
+      "loss": 0.9915,
+      "step": 12677
+    },
+    {
+      "epoch": 0.9773358001850139,
+      "grad_norm": 3.511397361755371,
+      "learning_rate": 1.3466043588628019e-08,
+      "loss": 0.9487,
+      "step": 12678
+    },
+    {
+      "epoch": 0.9774128893000308,
+      "grad_norm": 3.758230447769165,
+      "learning_rate": 1.3374634434225043e-08,
+      "loss": 0.9622,
+      "step": 12679
+    },
+    {
+      "epoch": 0.9774899784150478,
+      "grad_norm": 4.067485332489014,
+      "learning_rate": 1.3283536170087818e-08,
+      "loss": 1.0281,
+      "step": 12680
+    },
+    {
+      "epoch": 0.9775670675300647,
+      "grad_norm": 3.6686248779296875,
+      "learning_rate": 1.3192748801895139e-08,
+      "loss": 0.86,
+      "step": 12681
+    },
+    {
+      "epoch": 0.9776441566450818,
+      "grad_norm": 3.7029106616973877,
+      "learning_rate": 1.3102272335307476e-08,
+      "loss": 0.7519,
+      "step": 12682
+    },
+    {
+      "epoch": 0.9777212457600987,
+      "grad_norm": 3.997795343399048,
+      "learning_rate": 1.301210677596476e-08,
+      "loss": 0.8996,
+      "step": 12683
+    },
+    {
+      "epoch": 0.9777983348751156,
+      "grad_norm": 3.61710262298584,
+      "learning_rate": 1.2922252129489165e-08,
+      "loss": 0.9548,
+      "step": 12684
+    },
+    {
+      "epoch": 0.9778754239901326,
+      "grad_norm": 3.6802361011505127,
+      "learning_rate": 1.2832708401482319e-08,
+      "loss": 0.8405,
+      "step": 12685
+    },
+    {
+      "epoch": 0.9779525131051495,
+      "grad_norm": 3.580125331878662,
+      "learning_rate": 1.2743475597526978e-08,
+      "loss": 0.8522,
+      "step": 12686
+    },
+    {
+      "epoch": 0.9780296022201665,
+      "grad_norm": 3.6615421772003174,
+      "learning_rate": 1.265455372318647e-08,
+      "loss": 0.9192,
+      "step": 12687
+    },
+    {
+      "epoch": 0.9781066913351835,
+      "grad_norm": 3.9896399974823,
+      "learning_rate": 1.2565942784004692e-08,
+      "loss": 0.9484,
+      "step": 12688
+    },
+    {
+      "epoch": 0.9781837804502004,
+      "grad_norm": 4.462338447570801,
+      "learning_rate": 1.2477642785505006e-08,
+      "loss": 0.9435,
+      "step": 12689
+    },
+    {
+      "epoch": 0.9782608695652174,
+      "grad_norm": 3.643880605697632,
+      "learning_rate": 1.2389653733193563e-08,
+      "loss": 0.8626,
+      "step": 12690
+    },
+    {
+      "epoch": 0.9783379586802343,
+      "grad_norm": 3.689103841781616,
+      "learning_rate": 1.2301975632555973e-08,
+      "loss": 0.826,
+      "step": 12691
+    },
+    {
+      "epoch": 0.9784150477952513,
+      "grad_norm": 3.740290880203247,
+      "learning_rate": 1.2214608489057866e-08,
+      "loss": 0.8738,
+      "step": 12692
+    },
+    {
+      "epoch": 0.9784921369102683,
+      "grad_norm": 3.896019697189331,
+      "learning_rate": 1.2127552308147106e-08,
+      "loss": 0.9251,
+      "step": 12693
+    },
+    {
+      "epoch": 0.9785692260252852,
+      "grad_norm": 3.6664927005767822,
+      "learning_rate": 1.2040807095249907e-08,
+      "loss": 0.9496,
+      "step": 12694
+    },
+    {
+      "epoch": 0.9786463151403022,
+      "grad_norm": 3.8327767848968506,
+      "learning_rate": 1.1954372855775275e-08,
+      "loss": 0.991,
+      "step": 12695
+    },
+    {
+      "epoch": 0.9787234042553191,
+      "grad_norm": 3.853940486907959,
+      "learning_rate": 1.1868249595111681e-08,
+      "loss": 0.8489,
+      "step": 12696
+    },
+    {
+      "epoch": 0.9788004933703361,
+      "grad_norm": 3.322674512863159,
+      "learning_rate": 1.1782437318628714e-08,
+      "loss": 0.8567,
+      "step": 12697
+    },
+    {
+      "epoch": 0.9788775824853531,
+      "grad_norm": 3.6797235012054443,
+      "learning_rate": 1.1696936031676542e-08,
+      "loss": 0.8655,
+      "step": 12698
+    },
+    {
+      "epoch": 0.97895467160037,
+      "grad_norm": 4.247450351715088,
+      "learning_rate": 1.161174573958479e-08,
+      "loss": 1.0226,
+      "step": 12699
+    },
+    {
+      "epoch": 0.979031760715387,
+      "grad_norm": 3.664731979370117,
+      "learning_rate": 1.152686644766532e-08,
+      "loss": 0.8835,
+      "step": 12700
+    },
+    {
+      "epoch": 0.9791088498304039,
+      "grad_norm": 3.965914249420166,
+      "learning_rate": 1.1442298161209453e-08,
+      "loss": 0.8815,
+      "step": 12701
+    },
+    {
+      "epoch": 0.9791859389454209,
+      "grad_norm": 4.060098648071289,
+      "learning_rate": 1.1358040885490196e-08,
+      "loss": 0.8988,
+      "step": 12702
+    },
+    {
+      "epoch": 0.9792630280604379,
+      "grad_norm": 3.575415849685669,
+      "learning_rate": 1.1274094625760013e-08,
+      "loss": 0.8422,
+      "step": 12703
+    },
+    {
+      "epoch": 0.9793401171754548,
+      "grad_norm": 3.5373342037200928,
+      "learning_rate": 1.119045938725305e-08,
+      "loss": 1.0148,
+      "step": 12704
+    },
+    {
+      "epoch": 0.9794172062904718,
+      "grad_norm": 3.689316511154175,
+      "learning_rate": 1.110713517518347e-08,
+      "loss": 0.9632,
+      "step": 12705
+    },
+    {
+      "epoch": 0.9794942954054887,
+      "grad_norm": 3.855285882949829,
+      "learning_rate": 1.1024121994745452e-08,
+      "loss": 0.929,
+      "step": 12706
+    },
+    {
+      "epoch": 0.9795713845205057,
+      "grad_norm": 3.655165910720825,
+      "learning_rate": 1.0941419851114853e-08,
+      "loss": 0.9428,
+      "step": 12707
+    },
+    {
+      "epoch": 0.9796484736355227,
+      "grad_norm": 3.5580670833587646,
+      "learning_rate": 1.0859028749447552e-08,
+      "loss": 0.8268,
+      "step": 12708
+    },
+    {
+      "epoch": 0.9797255627505396,
+      "grad_norm": 3.758411407470703,
+      "learning_rate": 1.0776948694881107e-08,
+      "loss": 0.9141,
+      "step": 12709
+    },
+    {
+      "epoch": 0.9798026518655566,
+      "grad_norm": 3.422999858856201,
+      "learning_rate": 1.0695179692531421e-08,
+      "loss": 0.8323,
+      "step": 12710
+    },
+    {
+      "epoch": 0.9798797409805735,
+      "grad_norm": 3.754002809524536,
+      "learning_rate": 1.0613721747497196e-08,
+      "loss": 0.942,
+      "step": 12711
+    },
+    {
+      "epoch": 0.9799568300955905,
+      "grad_norm": 3.7836527824401855,
+      "learning_rate": 1.0532574864856592e-08,
+      "loss": 0.9146,
+      "step": 12712
+    },
+    {
+      "epoch": 0.9800339192106075,
+      "grad_norm": 3.721501350402832,
+      "learning_rate": 1.0451739049668896e-08,
+      "loss": 0.9326,
+      "step": 12713
+    },
+    {
+      "epoch": 0.9801110083256244,
+      "grad_norm": 3.7368814945220947,
+      "learning_rate": 1.0371214306973964e-08,
+      "loss": 0.9373,
+      "step": 12714
+    },
+    {
+      "epoch": 0.9801880974406414,
+      "grad_norm": 3.6165754795074463,
+      "learning_rate": 1.029100064179167e-08,
+      "loss": 0.9012,
+      "step": 12715
+    },
+    {
+      "epoch": 0.9802651865556583,
+      "grad_norm": 3.814692497253418,
+      "learning_rate": 1.0211098059123015e-08,
+      "loss": 0.8587,
+      "step": 12716
+    },
+    {
+      "epoch": 0.9803422756706753,
+      "grad_norm": 3.516226053237915,
+      "learning_rate": 1.0131506563950121e-08,
+      "loss": 0.9172,
+      "step": 12717
+    },
+    {
+      "epoch": 0.9804193647856922,
+      "grad_norm": 3.7366223335266113,
+      "learning_rate": 1.0052226161234024e-08,
+      "loss": 0.8711,
+      "step": 12718
+    },
+    {
+      "epoch": 0.9804964539007093,
+      "grad_norm": 3.5088720321655273,
+      "learning_rate": 9.97325685591799e-09,
+      "loss": 0.8817,
+      "step": 12719
+    },
+    {
+      "epoch": 0.9805735430157262,
+      "grad_norm": 3.6304702758789062,
+      "learning_rate": 9.894598652925857e-09,
+      "loss": 0.9045,
+      "step": 12720
+    },
+    {
+      "epoch": 0.9806506321307431,
+      "grad_norm": 3.697366952896118,
+      "learning_rate": 9.816251557160927e-09,
+      "loss": 0.9645,
+      "step": 12721
+    },
+    {
+      "epoch": 0.9807277212457601,
+      "grad_norm": 3.8746864795684814,
+      "learning_rate": 9.738215573507625e-09,
+      "loss": 0.9756,
+      "step": 12722
+    },
+    {
+      "epoch": 0.980804810360777,
+      "grad_norm": 3.758707046508789,
+      "learning_rate": 9.660490706831505e-09,
+      "loss": 0.8504,
+      "step": 12723
+    },
+    {
+      "epoch": 0.9808818994757941,
+      "grad_norm": 3.641058921813965,
+      "learning_rate": 9.583076961978132e-09,
+      "loss": 0.7918,
+      "step": 12724
+    },
+    {
+      "epoch": 0.980958988590811,
+      "grad_norm": 3.7192978858947754,
+      "learning_rate": 9.505974343774205e-09,
+      "loss": 0.9248,
+      "step": 12725
+    },
+    {
+      "epoch": 0.9810360777058279,
+      "grad_norm": 3.7372353076934814,
+      "learning_rate": 9.429182857025876e-09,
+      "loss": 0.8648,
+      "step": 12726
+    },
+    {
+      "epoch": 0.9811131668208449,
+      "grad_norm": 3.754185199737549,
+      "learning_rate": 9.352702506521539e-09,
+      "loss": 0.8707,
+      "step": 12727
+    },
+    {
+      "epoch": 0.9811902559358618,
+      "grad_norm": 3.9802560806274414,
+      "learning_rate": 9.276533297028489e-09,
+      "loss": 0.9176,
+      "step": 12728
+    },
+    {
+      "epoch": 0.9812673450508789,
+      "grad_norm": 3.81831431388855,
+      "learning_rate": 9.200675233296263e-09,
+      "loss": 0.8515,
+      "step": 12729
+    },
+    {
+      "epoch": 0.9813444341658958,
+      "grad_norm": 3.8928399085998535,
+      "learning_rate": 9.125128320053856e-09,
+      "loss": 0.9243,
+      "step": 12730
+    },
+    {
+      "epoch": 0.9814215232809127,
+      "grad_norm": 3.8177058696746826,
+      "learning_rate": 9.049892562011387e-09,
+      "loss": 0.908,
+      "step": 12731
+    },
+    {
+      "epoch": 0.9814986123959297,
+      "grad_norm": 3.94755220413208,
+      "learning_rate": 8.974967963858994e-09,
+      "loss": 0.906,
+      "step": 12732
+    },
+    {
+      "epoch": 0.9815757015109466,
+      "grad_norm": 3.647578477859497,
+      "learning_rate": 8.900354530268497e-09,
+      "loss": 0.8664,
+      "step": 12733
+    },
+    {
+      "epoch": 0.9816527906259637,
+      "grad_norm": 3.8050804138183594,
+      "learning_rate": 8.826052265891172e-09,
+      "loss": 0.9537,
+      "step": 12734
+    },
+    {
+      "epoch": 0.9817298797409806,
+      "grad_norm": 3.629077672958374,
+      "learning_rate": 8.752061175359983e-09,
+      "loss": 0.8645,
+      "step": 12735
+    },
+    {
+      "epoch": 0.9818069688559975,
+      "grad_norm": 3.9921088218688965,
+      "learning_rate": 8.67838126328735e-09,
+      "loss": 0.98,
+      "step": 12736
+    },
+    {
+      "epoch": 0.9818840579710145,
+      "grad_norm": 3.7594244480133057,
+      "learning_rate": 8.605012534266265e-09,
+      "loss": 0.8998,
+      "step": 12737
+    },
+    {
+      "epoch": 0.9819611470860314,
+      "grad_norm": 3.63808536529541,
+      "learning_rate": 8.53195499287196e-09,
+      "loss": 0.9154,
+      "step": 12738
+    },
+    {
+      "epoch": 0.9820382362010485,
+      "grad_norm": 3.701803207397461,
+      "learning_rate": 8.459208643659122e-09,
+      "loss": 0.8308,
+      "step": 12739
+    },
+    {
+      "epoch": 0.9821153253160654,
+      "grad_norm": 3.5713162422180176,
+      "learning_rate": 8.386773491162458e-09,
+      "loss": 0.9392,
+      "step": 12740
+    },
+    {
+      "epoch": 0.9821924144310823,
+      "grad_norm": 3.8430564403533936,
+      "learning_rate": 8.314649539898357e-09,
+      "loss": 0.8847,
+      "step": 12741
+    },
+    {
+      "epoch": 0.9822695035460993,
+      "grad_norm": 3.8548264503479004,
+      "learning_rate": 8.242836794362662e-09,
+      "loss": 0.9908,
+      "step": 12742
+    },
+    {
+      "epoch": 0.9823465926611162,
+      "grad_norm": 4.117403984069824,
+      "learning_rate": 8.17133525903402e-09,
+      "loss": 0.8986,
+      "step": 12743
+    },
+    {
+      "epoch": 0.9824236817761333,
+      "grad_norm": 3.5124740600585938,
+      "learning_rate": 8.100144938368304e-09,
+      "loss": 0.9653,
+      "step": 12744
+    },
+    {
+      "epoch": 0.9825007708911502,
+      "grad_norm": 3.7108638286590576,
+      "learning_rate": 8.029265836805855e-09,
+      "loss": 0.8482,
+      "step": 12745
+    },
+    {
+      "epoch": 0.9825778600061671,
+      "grad_norm": 3.5516583919525146,
+      "learning_rate": 7.958697958763695e-09,
+      "loss": 0.9324,
+      "step": 12746
+    },
+    {
+      "epoch": 0.9826549491211841,
+      "grad_norm": 4.330380439758301,
+      "learning_rate": 7.888441308642746e-09,
+      "loss": 0.9597,
+      "step": 12747
+    },
+    {
+      "epoch": 0.982732038236201,
+      "grad_norm": 3.6358656883239746,
+      "learning_rate": 7.818495890822842e-09,
+      "loss": 0.8769,
+      "step": 12748
+    },
+    {
+      "epoch": 0.9828091273512181,
+      "grad_norm": 3.5097153186798096,
+      "learning_rate": 7.748861709664379e-09,
+      "loss": 0.8299,
+      "step": 12749
+    },
+    {
+      "epoch": 0.982886216466235,
+      "grad_norm": 4.351414203643799,
+      "learning_rate": 7.679538769508887e-09,
+      "loss": 0.9594,
+      "step": 12750
+    },
+    {
+      "epoch": 0.9829633055812519,
+      "grad_norm": 4.271960735321045,
+      "learning_rate": 7.61052707467791e-09,
+      "loss": 0.8921,
+      "step": 12751
+    },
+    {
+      "epoch": 0.9830403946962689,
+      "grad_norm": 3.7134621143341064,
+      "learning_rate": 7.541826629474669e-09,
+      "loss": 0.8811,
+      "step": 12752
+    },
+    {
+      "epoch": 0.9831174838112858,
+      "grad_norm": 3.69516921043396,
+      "learning_rate": 7.473437438181852e-09,
+      "loss": 0.9587,
+      "step": 12753
+    },
+    {
+      "epoch": 0.9831945729263029,
+      "grad_norm": 3.7787835597991943,
+      "learning_rate": 7.40535950506327e-09,
+      "loss": 0.9491,
+      "step": 12754
+    },
+    {
+      "epoch": 0.9832716620413198,
+      "grad_norm": 4.08289909362793,
+      "learning_rate": 7.3375928343633054e-09,
+      "loss": 0.9131,
+      "step": 12755
+    },
+    {
+      "epoch": 0.9833487511563367,
+      "grad_norm": 3.624502658843994,
+      "learning_rate": 7.2701374303063565e-09,
+      "loss": 0.9144,
+      "step": 12756
+    },
+    {
+      "epoch": 0.9834258402713537,
+      "grad_norm": 3.4605467319488525,
+      "learning_rate": 7.202993297099059e-09,
+      "loss": 0.8354,
+      "step": 12757
+    },
+    {
+      "epoch": 0.9835029293863706,
+      "grad_norm": 3.8539695739746094,
+      "learning_rate": 7.136160438925843e-09,
+      "loss": 0.8254,
+      "step": 12758
+    },
+    {
+      "epoch": 0.9835800185013877,
+      "grad_norm": 3.464372396469116,
+      "learning_rate": 7.069638859955041e-09,
+      "loss": 0.8553,
+      "step": 12759
+    },
+    {
+      "epoch": 0.9836571076164046,
+      "grad_norm": 3.6082799434661865,
+      "learning_rate": 7.003428564332782e-09,
+      "loss": 0.9381,
+      "step": 12760
+    },
+    {
+      "epoch": 0.9837341967314215,
+      "grad_norm": 3.6667985916137695,
+      "learning_rate": 6.937529556187428e-09,
+      "loss": 0.962,
+      "step": 12761
+    },
+    {
+      "epoch": 0.9838112858464385,
+      "grad_norm": 3.6042168140411377,
+      "learning_rate": 6.871941839627916e-09,
+      "loss": 0.8476,
+      "step": 12762
+    },
+    {
+      "epoch": 0.9838883749614554,
+      "grad_norm": 3.740255832672119,
+      "learning_rate": 6.8066654187420865e-09,
+      "loss": 0.9196,
+      "step": 12763
+    },
+    {
+      "epoch": 0.9839654640764725,
+      "grad_norm": 3.6301140785217285,
+      "learning_rate": 6.741700297600573e-09,
+      "loss": 0.7799,
+      "step": 12764
+    },
+    {
+      "epoch": 0.9840425531914894,
+      "grad_norm": 3.6463210582733154,
+      "learning_rate": 6.677046480252913e-09,
+      "loss": 0.8718,
+      "step": 12765
+    },
+    {
+      "epoch": 0.9841196423065063,
+      "grad_norm": 4.018143653869629,
+      "learning_rate": 6.6127039707308826e-09,
+      "loss": 1.0302,
+      "step": 12766
+    },
+    {
+      "epoch": 0.9841967314215233,
+      "grad_norm": 3.6346871852874756,
+      "learning_rate": 6.548672773045161e-09,
+      "loss": 0.9883,
+      "step": 12767
+    },
+    {
+      "epoch": 0.9842738205365402,
+      "grad_norm": 3.8458900451660156,
+      "learning_rate": 6.4849528911881125e-09,
+      "loss": 0.8449,
+      "step": 12768
+    },
+    {
+      "epoch": 0.9843509096515572,
+      "grad_norm": 3.548745632171631,
+      "learning_rate": 6.421544329131557e-09,
+      "loss": 0.738,
+      "step": 12769
+    },
+    {
+      "epoch": 0.9844279987665742,
+      "grad_norm": 3.7064075469970703,
+      "learning_rate": 6.358447090829556e-09,
+      "loss": 0.8474,
+      "step": 12770
+    },
+    {
+      "epoch": 0.9845050878815911,
+      "grad_norm": 3.6911072731018066,
+      "learning_rate": 6.295661180216184e-09,
+      "loss": 0.911,
+      "step": 12771
+    },
+    {
+      "epoch": 0.9845821769966081,
+      "grad_norm": 3.657993793487549,
+      "learning_rate": 6.2331866012044215e-09,
+      "loss": 0.9808,
+      "step": 12772
+    },
+    {
+      "epoch": 0.984659266111625,
+      "grad_norm": 3.437884569168091,
+      "learning_rate": 6.171023357690598e-09,
+      "loss": 0.8563,
+      "step": 12773
+    },
+    {
+      "epoch": 0.984736355226642,
+      "grad_norm": 3.694701910018921,
+      "learning_rate": 6.109171453549944e-09,
+      "loss": 0.798,
+      "step": 12774
+    },
+    {
+      "epoch": 0.984813444341659,
+      "grad_norm": 3.68932843208313,
+      "learning_rate": 6.0476308926377125e-09,
+      "loss": 0.8745,
+      "step": 12775
+    },
+    {
+      "epoch": 0.9848905334566759,
+      "grad_norm": 3.8350744247436523,
+      "learning_rate": 5.986401678791942e-09,
+      "loss": 0.9399,
+      "step": 12776
+    },
+    {
+      "epoch": 0.9849676225716929,
+      "grad_norm": 4.023548126220703,
+      "learning_rate": 5.9254838158295805e-09,
+      "loss": 0.835,
+      "step": 12777
+    },
+    {
+      "epoch": 0.9850447116867098,
+      "grad_norm": 3.844338893890381,
+      "learning_rate": 5.864877307547589e-09,
+      "loss": 1.0251,
+      "step": 12778
+    },
+    {
+      "epoch": 0.9851218008017268,
+      "grad_norm": 3.5062670707702637,
+      "learning_rate": 5.804582157725724e-09,
+      "loss": 0.8741,
+      "step": 12779
+    },
+    {
+      "epoch": 0.9851988899167438,
+      "grad_norm": 3.6520488262176514,
+      "learning_rate": 5.744598370122645e-09,
+      "loss": 0.9983,
+      "step": 12780
+    },
+    {
+      "epoch": 0.9852759790317607,
+      "grad_norm": 3.605093479156494,
+      "learning_rate": 5.684925948477582e-09,
+      "loss": 0.8435,
+      "step": 12781
+    },
+    {
+      "epoch": 0.9853530681467777,
+      "grad_norm": 3.9544193744659424,
+      "learning_rate": 5.625564896511448e-09,
+      "loss": 0.899,
+      "step": 12782
+    },
+    {
+      "epoch": 0.9854301572617946,
+      "grad_norm": 3.8571465015411377,
+      "learning_rate": 5.566515217924617e-09,
+      "loss": 0.8574,
+      "step": 12783
+    },
+    {
+      "epoch": 0.9855072463768116,
+      "grad_norm": 3.823538303375244,
+      "learning_rate": 5.507776916398588e-09,
+      "loss": 0.9406,
+      "step": 12784
+    },
+    {
+      "epoch": 0.9855843354918286,
+      "grad_norm": 4.045183181762695,
+      "learning_rate": 5.4493499955959874e-09,
+      "loss": 0.804,
+      "step": 12785
+    },
+    {
+      "epoch": 0.9856614246068455,
+      "grad_norm": 3.724013090133667,
+      "learning_rate": 5.391234459158901e-09,
+      "loss": 0.9158,
+      "step": 12786
+    },
+    {
+      "epoch": 0.9857385137218625,
+      "grad_norm": 3.712146759033203,
+      "learning_rate": 5.333430310709986e-09,
+      "loss": 1.0271,
+      "step": 12787
+    },
+    {
+      "epoch": 0.9858156028368794,
+      "grad_norm": 4.106431484222412,
+      "learning_rate": 5.2759375538541376e-09,
+      "loss": 0.9228,
+      "step": 12788
+    },
+    {
+      "epoch": 0.9858926919518964,
+      "grad_norm": 3.821424961090088,
+      "learning_rate": 5.218756192175156e-09,
+      "loss": 0.974,
+      "step": 12789
+    },
+    {
+      "epoch": 0.9859697810669134,
+      "grad_norm": 3.7011265754699707,
+      "learning_rate": 5.161886229238522e-09,
+      "loss": 0.81,
+      "step": 12790
+    },
+    {
+      "epoch": 0.9860468701819303,
+      "grad_norm": 3.595111608505249,
+      "learning_rate": 5.105327668588622e-09,
+      "loss": 0.9212,
+      "step": 12791
+    },
+    {
+      "epoch": 0.9861239592969473,
+      "grad_norm": 3.862673759460449,
+      "learning_rate": 5.049080513752636e-09,
+      "loss": 0.9619,
+      "step": 12792
+    },
+    {
+      "epoch": 0.9862010484119642,
+      "grad_norm": 3.584799289703369,
+      "learning_rate": 4.993144768237201e-09,
+      "loss": 0.9546,
+      "step": 12793
+    },
+    {
+      "epoch": 0.9862781375269812,
+      "grad_norm": 3.9195289611816406,
+      "learning_rate": 4.937520435528975e-09,
+      "loss": 0.9763,
+      "step": 12794
+    },
+    {
+      "epoch": 0.9863552266419982,
+      "grad_norm": 3.7915120124816895,
+      "learning_rate": 4.8822075190962936e-09,
+      "loss": 0.7143,
+      "step": 12795
+    },
+    {
+      "epoch": 0.9864323157570151,
+      "grad_norm": 3.788917303085327,
+      "learning_rate": 4.827206022388065e-09,
+      "loss": 0.9824,
+      "step": 12796
+    },
+    {
+      "epoch": 0.9865094048720321,
+      "grad_norm": 3.605607509613037,
+      "learning_rate": 4.772515948832657e-09,
+      "loss": 0.8145,
+      "step": 12797
+    },
+    {
+      "epoch": 0.986586493987049,
+      "grad_norm": 3.643545150756836,
+      "learning_rate": 4.718137301839565e-09,
+      "loss": 0.8363,
+      "step": 12798
+    },
+    {
+      "epoch": 0.986663583102066,
+      "grad_norm": 4.066047668457031,
+      "learning_rate": 4.66407008479941e-09,
+      "loss": 0.9509,
+      "step": 12799
+    },
+    {
+      "epoch": 0.986740672217083,
+      "grad_norm": 3.583989381790161,
+      "learning_rate": 4.6103143010833854e-09,
+      "loss": 0.8493,
+      "step": 12800
+    },
+    {
+      "epoch": 0.9868177613320999,
+      "grad_norm": 3.6887717247009277,
+      "learning_rate": 4.556869954042143e-09,
+      "loss": 0.8542,
+      "step": 12801
+    },
+    {
+      "epoch": 0.9868948504471169,
+      "grad_norm": 3.8332138061523438,
+      "learning_rate": 4.5037370470085715e-09,
+      "loss": 0.9282,
+      "step": 12802
+    },
+    {
+      "epoch": 0.9869719395621338,
+      "grad_norm": 3.6815474033355713,
+      "learning_rate": 4.450915583293913e-09,
+      "loss": 0.8847,
+      "step": 12803
+    },
+    {
+      "epoch": 0.9870490286771508,
+      "grad_norm": 3.505082607269287,
+      "learning_rate": 4.398405566192754e-09,
+      "loss": 0.7977,
+      "step": 12804
+    },
+    {
+      "epoch": 0.9871261177921677,
+      "grad_norm": 3.825004816055298,
+      "learning_rate": 4.346206998977476e-09,
+      "loss": 1.0071,
+      "step": 12805
+    },
+    {
+      "epoch": 0.9872032069071847,
+      "grad_norm": 4.440659999847412,
+      "learning_rate": 4.294319884903253e-09,
+      "loss": 1.0099,
+      "step": 12806
+    },
+    {
+      "epoch": 0.9872802960222017,
+      "grad_norm": 3.277602195739746,
+      "learning_rate": 4.24274422720472e-09,
+      "loss": 0.8623,
+      "step": 12807
+    },
+    {
+      "epoch": 0.9873573851372186,
+      "grad_norm": 3.6194324493408203,
+      "learning_rate": 4.191480029097639e-09,
+      "loss": 0.9843,
+      "step": 12808
+    },
+    {
+      "epoch": 0.9874344742522356,
+      "grad_norm": 3.5237951278686523,
+      "learning_rate": 4.140527293777785e-09,
+      "loss": 0.7906,
+      "step": 12809
+    },
+    {
+      "epoch": 0.9875115633672525,
+      "grad_norm": 4.282720565795898,
+      "learning_rate": 4.089886024421508e-09,
+      "loss": 0.9852,
+      "step": 12810
+    },
+    {
+      "epoch": 0.9875886524822695,
+      "grad_norm": 3.8325626850128174,
+      "learning_rate": 4.039556224186836e-09,
+      "loss": 0.9432,
+      "step": 12811
+    },
+    {
+      "epoch": 0.9876657415972865,
+      "grad_norm": 3.7518670558929443,
+      "learning_rate": 3.989537896210704e-09,
+      "loss": 0.8936,
+      "step": 12812
+    },
+    {
+      "epoch": 0.9877428307123034,
+      "grad_norm": 3.3281097412109375,
+      "learning_rate": 3.93983104361173e-09,
+      "loss": 0.8703,
+      "step": 12813
+    },
+    {
+      "epoch": 0.9878199198273204,
+      "grad_norm": 3.881019353866577,
+      "learning_rate": 3.890435669489656e-09,
+      "loss": 0.8354,
+      "step": 12814
+    },
+    {
+      "epoch": 0.9878970089423373,
+      "grad_norm": 3.5037472248077393,
+      "learning_rate": 3.84135177692313e-09,
+      "loss": 0.9291,
+      "step": 12815
+    },
+    {
+      "epoch": 0.9879740980573543,
+      "grad_norm": 3.938447952270508,
+      "learning_rate": 3.792579368972482e-09,
+      "loss": 0.9214,
+      "step": 12816
+    },
+    {
+      "epoch": 0.9880511871723713,
+      "grad_norm": 4.022164344787598,
+      "learning_rate": 3.744118448678058e-09,
+      "loss": 0.8379,
+      "step": 12817
+    },
+    {
+      "epoch": 0.9881282762873882,
+      "grad_norm": 3.696821451187134,
+      "learning_rate": 3.6959690190618847e-09,
+      "loss": 0.8365,
+      "step": 12818
+    },
+    {
+      "epoch": 0.9882053654024052,
+      "grad_norm": 3.4798009395599365,
+      "learning_rate": 3.6481310831260054e-09,
+      "loss": 0.8654,
+      "step": 12819
+    },
+    {
+      "epoch": 0.9882824545174221,
+      "grad_norm": 3.6241512298583984,
+      "learning_rate": 3.600604643851924e-09,
+      "loss": 0.8805,
+      "step": 12820
+    },
+    {
+      "epoch": 0.988359543632439,
+      "grad_norm": 3.9195075035095215,
+      "learning_rate": 3.5533897042033805e-09,
+      "loss": 0.9811,
+      "step": 12821
+    },
+    {
+      "epoch": 0.9884366327474561,
+      "grad_norm": 4.072207927703857,
+      "learning_rate": 3.5064862671230217e-09,
+      "loss": 0.9203,
+      "step": 12822
+    },
+    {
+      "epoch": 0.988513721862473,
+      "grad_norm": 4.044565200805664,
+      "learning_rate": 3.4598943355362845e-09,
+      "loss": 0.9196,
+      "step": 12823
+    },
+    {
+      "epoch": 0.98859081097749,
+      "grad_norm": 3.762471914291382,
+      "learning_rate": 3.4136139123475134e-09,
+      "loss": 0.8305,
+      "step": 12824
+    },
+    {
+      "epoch": 0.9886679000925069,
+      "grad_norm": 3.648581027984619,
+      "learning_rate": 3.3676450004416215e-09,
+      "loss": 0.8875,
+      "step": 12825
+    },
+    {
+      "epoch": 0.9887449892075239,
+      "grad_norm": 3.582662343978882,
+      "learning_rate": 3.321987602685206e-09,
+      "loss": 0.8568,
+      "step": 12826
+    },
+    {
+      "epoch": 0.9888220783225409,
+      "grad_norm": 3.838958978652954,
+      "learning_rate": 3.276641721924323e-09,
+      "loss": 0.9401,
+      "step": 12827
+    },
+    {
+      "epoch": 0.9888991674375578,
+      "grad_norm": 3.704432964324951,
+      "learning_rate": 3.2316073609856e-09,
+      "loss": 0.8952,
+      "step": 12828
+    },
+    {
+      "epoch": 0.9889762565525748,
+      "grad_norm": 3.5584044456481934,
+      "learning_rate": 3.186884522677902e-09,
+      "loss": 1.0251,
+      "step": 12829
+    },
+    {
+      "epoch": 0.9890533456675917,
+      "grad_norm": 3.896476984024048,
+      "learning_rate": 3.1424732097884434e-09,
+      "loss": 0.888,
+      "step": 12830
+    },
+    {
+      "epoch": 0.9891304347826086,
+      "grad_norm": 3.5131354331970215,
+      "learning_rate": 3.098373425086676e-09,
+      "loss": 0.8817,
+      "step": 12831
+    },
+    {
+      "epoch": 0.9892075238976257,
+      "grad_norm": 3.6488070487976074,
+      "learning_rate": 3.0545851713215113e-09,
+      "loss": 0.8898,
+      "step": 12832
+    },
+    {
+      "epoch": 0.9892846130126426,
+      "grad_norm": 3.468665361404419,
+      "learning_rate": 3.011108451222988e-09,
+      "loss": 0.8596,
+      "step": 12833
+    },
+    {
+      "epoch": 0.9893617021276596,
+      "grad_norm": 3.568416118621826,
+      "learning_rate": 2.9679432675017163e-09,
+      "loss": 0.9725,
+      "step": 12834
+    },
+    {
+      "epoch": 0.9894387912426765,
+      "grad_norm": 4.020820617675781,
+      "learning_rate": 2.9250896228494307e-09,
+      "loss": 0.8734,
+      "step": 12835
+    },
+    {
+      "epoch": 0.9895158803576934,
+      "grad_norm": 3.5080959796905518,
+      "learning_rate": 2.8825475199367735e-09,
+      "loss": 0.8361,
+      "step": 12836
+    },
+    {
+      "epoch": 0.9895929694727105,
+      "grad_norm": 3.717933416366577,
+      "learning_rate": 2.8403169614166226e-09,
+      "loss": 0.7573,
+      "step": 12837
+    },
+    {
+      "epoch": 0.9896700585877274,
+      "grad_norm": 3.621506690979004,
+      "learning_rate": 2.7983979499218715e-09,
+      "loss": 0.903,
+      "step": 12838
+    },
+    {
+      "epoch": 0.9897471477027444,
+      "grad_norm": 3.9174814224243164,
+      "learning_rate": 2.756790488065986e-09,
+      "loss": 0.8788,
+      "step": 12839
+    },
+    {
+      "epoch": 0.9898242368177613,
+      "grad_norm": 4.379630088806152,
+      "learning_rate": 2.7154945784424456e-09,
+      "loss": 0.912,
+      "step": 12840
+    },
+    {
+      "epoch": 0.9899013259327782,
+      "grad_norm": 3.8400206565856934,
+      "learning_rate": 2.6745102236264142e-09,
+      "loss": 0.9096,
+      "step": 12841
+    },
+    {
+      "epoch": 0.9899784150477953,
+      "grad_norm": 3.97227144241333,
+      "learning_rate": 2.633837426173069e-09,
+      "loss": 0.8001,
+      "step": 12842
+    },
+    {
+      "epoch": 0.9900555041628122,
+      "grad_norm": 3.7930757999420166,
+      "learning_rate": 2.5934761886176052e-09,
+      "loss": 0.886,
+      "step": 12843
+    },
+    {
+      "epoch": 0.9901325932778292,
+      "grad_norm": 3.9105029106140137,
+      "learning_rate": 2.5534265134768977e-09,
+      "loss": 0.9109,
+      "step": 12844
+    },
+    {
+      "epoch": 0.9902096823928461,
+      "grad_norm": 3.479025363922119,
+      "learning_rate": 2.513688403247283e-09,
+      "loss": 0.8513,
+      "step": 12845
+    },
+    {
+      "epoch": 0.990286771507863,
+      "grad_norm": 4.080109119415283,
+      "learning_rate": 2.474261860406779e-09,
+      "loss": 0.8726,
+      "step": 12846
+    },
+    {
+      "epoch": 0.9903638606228801,
+      "grad_norm": 3.6820366382598877,
+      "learning_rate": 2.4351468874134197e-09,
+      "loss": 0.8961,
+      "step": 12847
+    },
+    {
+      "epoch": 0.990440949737897,
+      "grad_norm": 3.6713099479675293,
+      "learning_rate": 2.396343486705255e-09,
+      "loss": 0.9043,
+      "step": 12848
+    },
+    {
+      "epoch": 0.990518038852914,
+      "grad_norm": 3.635523796081543,
+      "learning_rate": 2.3578516607020154e-09,
+      "loss": 0.8773,
+      "step": 12849
+    },
+    {
+      "epoch": 0.9905951279679309,
+      "grad_norm": 3.570716142654419,
+      "learning_rate": 2.319671411802893e-09,
+      "loss": 0.9127,
+      "step": 12850
+    },
+    {
+      "epoch": 0.9906722170829478,
+      "grad_norm": 3.6702277660369873,
+      "learning_rate": 2.281802742388761e-09,
+      "loss": 0.852,
+      "step": 12851
+    },
+    {
+      "epoch": 0.9907493061979649,
+      "grad_norm": 3.9197945594787598,
+      "learning_rate": 2.2442456548205094e-09,
+      "loss": 0.9476,
+      "step": 12852
+    },
+    {
+      "epoch": 0.9908263953129818,
+      "grad_norm": 3.8219149112701416,
+      "learning_rate": 2.207000151439598e-09,
+      "loss": 0.9928,
+      "step": 12853
+    },
+    {
+      "epoch": 0.9909034844279988,
+      "grad_norm": 3.7937533855438232,
+      "learning_rate": 2.170066234568058e-09,
+      "loss": 1.061,
+      "step": 12854
+    },
+    {
+      "epoch": 0.9909805735430157,
+      "grad_norm": 3.5674586296081543,
+      "learning_rate": 2.133443906508492e-09,
+      "loss": 0.8648,
+      "step": 12855
+    },
+    {
+      "epoch": 0.9910576626580326,
+      "grad_norm": 3.8357954025268555,
+      "learning_rate": 2.097133169543519e-09,
+      "loss": 0.9061,
+      "step": 12856
+    },
+    {
+      "epoch": 0.9911347517730497,
+      "grad_norm": 3.6642403602600098,
+      "learning_rate": 2.0611340259379942e-09,
+      "loss": 0.9031,
+      "step": 12857
+    },
+    {
+      "epoch": 0.9912118408880666,
+      "grad_norm": 3.718217134475708,
+      "learning_rate": 2.0254464779356776e-09,
+      "loss": 0.8826,
+      "step": 12858
+    },
+    {
+      "epoch": 0.9912889300030836,
+      "grad_norm": 3.5286970138549805,
+      "learning_rate": 1.990070527761456e-09,
+      "loss": 0.8263,
+      "step": 12859
+    },
+    {
+      "epoch": 0.9913660191181005,
+      "grad_norm": 3.934288263320923,
+      "learning_rate": 1.9550061776213435e-09,
+      "loss": 0.8804,
+      "step": 12860
+    },
+    {
+      "epoch": 0.9914431082331174,
+      "grad_norm": 3.582671880722046,
+      "learning_rate": 1.9202534297008134e-09,
+      "loss": 0.8939,
+      "step": 12861
+    },
+    {
+      "epoch": 0.9915201973481345,
+      "grad_norm": 3.763944387435913,
+      "learning_rate": 1.8858122861664664e-09,
+      "loss": 0.9044,
+      "step": 12862
+    },
+    {
+      "epoch": 0.9915972864631514,
+      "grad_norm": 4.1658196449279785,
+      "learning_rate": 1.8516827491660282e-09,
+      "loss": 0.8877,
+      "step": 12863
+    },
+    {
+      "epoch": 0.9916743755781684,
+      "grad_norm": 3.695138692855835,
+      "learning_rate": 1.817864820827242e-09,
+      "loss": 0.8791,
+      "step": 12864
+    },
+    {
+      "epoch": 0.9917514646931853,
+      "grad_norm": 4.022881984710693,
+      "learning_rate": 1.784358503258421e-09,
+      "loss": 1.0657,
+      "step": 12865
+    },
+    {
+      "epoch": 0.9918285538082022,
+      "grad_norm": 3.7440459728240967,
+      "learning_rate": 1.7511637985478947e-09,
+      "loss": 0.8766,
+      "step": 12866
+    },
+    {
+      "epoch": 0.9919056429232193,
+      "grad_norm": 3.9772725105285645,
+      "learning_rate": 1.718280708766229e-09,
+      "loss": 0.9796,
+      "step": 12867
+    },
+    {
+      "epoch": 0.9919827320382362,
+      "grad_norm": 3.865461826324463,
+      "learning_rate": 1.6857092359628957e-09,
+      "loss": 0.8336,
+      "step": 12868
+    },
+    {
+      "epoch": 0.9920598211532532,
+      "grad_norm": 4.357512474060059,
+      "learning_rate": 1.6534493821684927e-09,
+      "loss": 1.0112,
+      "step": 12869
+    },
+    {
+      "epoch": 0.9921369102682701,
+      "grad_norm": 3.5063059329986572,
+      "learning_rate": 1.621501149394744e-09,
+      "loss": 0.8071,
+      "step": 12870
+    },
+    {
+      "epoch": 0.992213999383287,
+      "grad_norm": 3.5108940601348877,
+      "learning_rate": 1.5898645396328349e-09,
+      "loss": 0.9605,
+      "step": 12871
+    },
+    {
+      "epoch": 0.992291088498304,
+      "grad_norm": 3.727933883666992,
+      "learning_rate": 1.5585395548556314e-09,
+      "loss": 0.9568,
+      "step": 12872
+    },
+    {
+      "epoch": 0.992368177613321,
+      "grad_norm": 3.5517189502716064,
+      "learning_rate": 1.5275261970154608e-09,
+      "loss": 0.9227,
+      "step": 12873
+    },
+    {
+      "epoch": 0.992445266728338,
+      "grad_norm": 3.8274848461151123,
+      "learning_rate": 1.4968244680468868e-09,
+      "loss": 1.0268,
+      "step": 12874
+    },
+    {
+      "epoch": 0.9925223558433549,
+      "grad_norm": 3.797487258911133,
+      "learning_rate": 1.4664343698628236e-09,
+      "loss": 0.7404,
+      "step": 12875
+    },
+    {
+      "epoch": 0.9925994449583718,
+      "grad_norm": 3.6216273307800293,
+      "learning_rate": 1.436355904358977e-09,
+      "loss": 0.919,
+      "step": 12876
+    },
+    {
+      "epoch": 0.9926765340733888,
+      "grad_norm": 3.420397996902466,
+      "learning_rate": 1.4065890734099586e-09,
+      "loss": 0.8691,
+      "step": 12877
+    },
+    {
+      "epoch": 0.9927536231884058,
+      "grad_norm": 3.8331243991851807,
+      "learning_rate": 1.3771338788715061e-09,
+      "loss": 0.945,
+      "step": 12878
+    },
+    {
+      "epoch": 0.9928307123034228,
+      "grad_norm": 3.9298641681671143,
+      "learning_rate": 1.3479903225804836e-09,
+      "loss": 0.937,
+      "step": 12879
+    },
+    {
+      "epoch": 0.9929078014184397,
+      "grad_norm": 3.827240467071533,
+      "learning_rate": 1.319158406353771e-09,
+      "loss": 0.9731,
+      "step": 12880
+    },
+    {
+      "epoch": 0.9929848905334566,
+      "grad_norm": 3.64302921295166,
+      "learning_rate": 1.2906381319882644e-09,
+      "loss": 0.8572,
+      "step": 12881
+    },
+    {
+      "epoch": 0.9930619796484736,
+      "grad_norm": 3.7599618434906006,
+      "learning_rate": 1.2624295012625409e-09,
+      "loss": 0.954,
+      "step": 12882
+    },
+    {
+      "epoch": 0.9931390687634906,
+      "grad_norm": 3.944673776626587,
+      "learning_rate": 1.2345325159357491e-09,
+      "loss": 0.9175,
+      "step": 12883
+    },
+    {
+      "epoch": 0.9932161578785076,
+      "grad_norm": 3.6056995391845703,
+      "learning_rate": 1.2069471777459429e-09,
+      "loss": 0.8907,
+      "step": 12884
+    },
+    {
+      "epoch": 0.9932932469935245,
+      "grad_norm": 3.5463924407958984,
+      "learning_rate": 1.1796734884139682e-09,
+      "loss": 0.8499,
+      "step": 12885
+    },
+    {
+      "epoch": 0.9933703361085414,
+      "grad_norm": 3.643226385116577,
+      "learning_rate": 1.1527114496395764e-09,
+      "loss": 0.9093,
+      "step": 12886
+    },
+    {
+      "epoch": 0.9934474252235584,
+      "grad_norm": 3.62485671043396,
+      "learning_rate": 1.126061063103645e-09,
+      "loss": 0.9344,
+      "step": 12887
+    },
+    {
+      "epoch": 0.9935245143385754,
+      "grad_norm": 3.879370927810669,
+      "learning_rate": 1.0997223304687333e-09,
+      "loss": 0.9439,
+      "step": 12888
+    },
+    {
+      "epoch": 0.9936016034535924,
+      "grad_norm": 3.560511350631714,
+      "learning_rate": 1.0736952533757506e-09,
+      "loss": 0.8394,
+      "step": 12889
+    },
+    {
+      "epoch": 0.9936786925686093,
+      "grad_norm": 3.8150246143341064,
+      "learning_rate": 1.047979833447843e-09,
+      "loss": 0.7913,
+      "step": 12890
+    },
+    {
+      "epoch": 0.9937557816836263,
+      "grad_norm": 4.272899150848389,
+      "learning_rate": 1.0225760722876177e-09,
+      "loss": 0.9468,
+      "step": 12891
+    },
+    {
+      "epoch": 0.9938328707986432,
+      "grad_norm": 3.9249536991119385,
+      "learning_rate": 9.974839714799178e-10,
+      "loss": 0.8953,
+      "step": 12892
+    },
+    {
+      "epoch": 0.9939099599136602,
+      "grad_norm": 3.497441291809082,
+      "learning_rate": 9.727035325884925e-10,
+      "loss": 0.822,
+      "step": 12893
+    },
+    {
+      "epoch": 0.9939870490286772,
+      "grad_norm": 4.1843647956848145,
+      "learning_rate": 9.482347571587724e-10,
+      "loss": 0.9528,
+      "step": 12894
+    },
+    {
+      "epoch": 0.9940641381436941,
+      "grad_norm": 3.7656443119049072,
+      "learning_rate": 9.240776467150936e-10,
+      "loss": 0.9781,
+      "step": 12895
+    },
+    {
+      "epoch": 0.9941412272587111,
+      "grad_norm": 3.6662378311157227,
+      "learning_rate": 9.002322027651389e-10,
+      "loss": 0.889,
+      "step": 12896
+    },
+    {
+      "epoch": 0.994218316373728,
+      "grad_norm": 4.327576637268066,
+      "learning_rate": 8.766984267938316e-10,
+      "loss": 0.8852,
+      "step": 12897
+    },
+    {
+      "epoch": 0.994295405488745,
+      "grad_norm": 3.467092275619507,
+      "learning_rate": 8.534763202699969e-10,
+      "loss": 0.8667,
+      "step": 12898
+    },
+    {
+      "epoch": 0.994372494603762,
+      "grad_norm": 3.563298463821411,
+      "learning_rate": 8.305658846402554e-10,
+      "loss": 0.9111,
+      "step": 12899
+    },
+    {
+      "epoch": 0.9944495837187789,
+      "grad_norm": 3.706197738647461,
+      "learning_rate": 8.079671213334639e-10,
+      "loss": 0.854,
+      "step": 12900
+    },
+    {
+      "epoch": 0.9945266728337959,
+      "grad_norm": 4.062145233154297,
+      "learning_rate": 7.856800317584956e-10,
+      "loss": 0.8797,
+      "step": 12901
+    },
+    {
+      "epoch": 0.9946037619488128,
+      "grad_norm": 3.602550506591797,
+      "learning_rate": 7.637046173047946e-10,
+      "loss": 0.8933,
+      "step": 12902
+    },
+    {
+      "epoch": 0.9946808510638298,
+      "grad_norm": 4.166884422302246,
+      "learning_rate": 7.420408793423762e-10,
+      "loss": 0.9677,
+      "step": 12903
+    },
+    {
+      "epoch": 0.9947579401788468,
+      "grad_norm": 4.013248443603516,
+      "learning_rate": 7.206888192218265e-10,
+      "loss": 0.9302,
+      "step": 12904
+    },
+    {
+      "epoch": 0.9948350292938637,
+      "grad_norm": 3.9088120460510254,
+      "learning_rate": 6.99648438274303e-10,
+      "loss": 0.8768,
+      "step": 12905
+    },
+    {
+      "epoch": 0.9949121184088807,
+      "grad_norm": 3.6118009090423584,
+      "learning_rate": 6.789197378115342e-10,
+      "loss": 0.9307,
+      "step": 12906
+    },
+    {
+      "epoch": 0.9949892075238976,
+      "grad_norm": 3.630589485168457,
+      "learning_rate": 6.585027191263748e-10,
+      "loss": 0.9369,
+      "step": 12907
+    },
+    {
+      "epoch": 0.9950662966389145,
+      "grad_norm": 3.9359915256500244,
+      "learning_rate": 6.383973834911406e-10,
+      "loss": 0.8476,
+      "step": 12908
+    },
+    {
+      "epoch": 0.9951433857539316,
+      "grad_norm": 3.3206164836883545,
+      "learning_rate": 6.186037321592731e-10,
+      "loss": 0.7947,
+      "step": 12909
+    },
+    {
+      "epoch": 0.9952204748689485,
+      "grad_norm": 3.8254833221435547,
+      "learning_rate": 5.991217663653404e-10,
+      "loss": 0.859,
+      "step": 12910
+    },
+    {
+      "epoch": 0.9952975639839655,
+      "grad_norm": 3.701080322265625,
+      "learning_rate": 5.799514873233714e-10,
+      "loss": 0.8847,
+      "step": 12911
+    },
+    {
+      "epoch": 0.9953746530989824,
+      "grad_norm": 3.581287145614624,
+      "learning_rate": 5.610928962290763e-10,
+      "loss": 0.9241,
+      "step": 12912
+    },
+    {
+      "epoch": 0.9954517422139993,
+      "grad_norm": 3.830244302749634,
+      "learning_rate": 5.425459942576261e-10,
+      "loss": 0.9187,
+      "step": 12913
+    },
+    {
+      "epoch": 0.9955288313290164,
+      "grad_norm": 3.582486867904663,
+      "learning_rate": 5.243107825653182e-10,
+      "loss": 0.8323,
+      "step": 12914
+    },
+    {
+      "epoch": 0.9956059204440333,
+      "grad_norm": 3.7536768913269043,
+      "learning_rate": 5.06387262289576e-10,
+      "loss": 0.9389,
+      "step": 12915
+    },
+    {
+      "epoch": 0.9956830095590503,
+      "grad_norm": 3.8179402351379395,
+      "learning_rate": 4.887754345478391e-10,
+      "loss": 0.8959,
+      "step": 12916
+    },
+    {
+      "epoch": 0.9957600986740672,
+      "grad_norm": 3.915501356124878,
+      "learning_rate": 4.714753004375627e-10,
+      "loss": 0.9829,
+      "step": 12917
+    },
+    {
+      "epoch": 0.9958371877890841,
+      "grad_norm": 3.4252476692199707,
+      "learning_rate": 4.5448686103732876e-10,
+      "loss": 0.816,
+      "step": 12918
+    },
+    {
+      "epoch": 0.9959142769041012,
+      "grad_norm": 4.076374053955078,
+      "learning_rate": 4.37810117406845e-10,
+      "loss": 0.9458,
+      "step": 12919
+    },
+    {
+      "epoch": 0.9959913660191181,
+      "grad_norm": 3.622620105743408,
+      "learning_rate": 4.2144507058528016e-10,
+      "loss": 0.8332,
+      "step": 12920
+    },
+    {
+      "epoch": 0.9960684551341351,
+      "grad_norm": 3.3440213203430176,
+      "learning_rate": 4.053917215934844e-10,
+      "loss": 0.8837,
+      "step": 12921
+    },
+    {
+      "epoch": 0.996145544249152,
+      "grad_norm": 3.92189884185791,
+      "learning_rate": 3.8965007143176857e-10,
+      "loss": 0.9619,
+      "step": 12922
+    },
+    {
+      "epoch": 0.9962226333641689,
+      "grad_norm": 4.633378505706787,
+      "learning_rate": 3.742201210815699e-10,
+      "loss": 0.9774,
+      "step": 12923
+    },
+    {
+      "epoch": 0.996299722479186,
+      "grad_norm": 3.8096113204956055,
+      "learning_rate": 3.5910187150545174e-10,
+      "loss": 0.9167,
+      "step": 12924
+    },
+    {
+      "epoch": 0.9963768115942029,
+      "grad_norm": 3.8679616451263428,
+      "learning_rate": 3.4429532364488315e-10,
+      "loss": 0.844,
+      "step": 12925
+    },
+    {
+      "epoch": 0.9964539007092199,
+      "grad_norm": 4.342710494995117,
+      "learning_rate": 3.298004784241249e-10,
+      "loss": 0.9341,
+      "step": 12926
+    },
+    {
+      "epoch": 0.9965309898242368,
+      "grad_norm": 4.704819679260254,
+      "learning_rate": 3.156173367457882e-10,
+      "loss": 0.9207,
+      "step": 12927
+    },
+    {
+      "epoch": 0.9966080789392537,
+      "grad_norm": 3.6841025352478027,
+      "learning_rate": 3.01745899495276e-10,
+      "loss": 0.8167,
+      "step": 12928
+    },
+    {
+      "epoch": 0.9966851680542708,
+      "grad_norm": 3.9212276935577393,
+      "learning_rate": 2.8818616753634177e-10,
+      "loss": 0.9434,
+      "step": 12929
+    },
+    {
+      "epoch": 0.9967622571692877,
+      "grad_norm": 3.5783891677856445,
+      "learning_rate": 2.749381417155306e-10,
+      "loss": 0.8192,
+      "step": 12930
+    },
+    {
+      "epoch": 0.9968393462843047,
+      "grad_norm": 3.933195114135742,
+      "learning_rate": 2.6200182285718303e-10,
+      "loss": 0.8573,
+      "step": 12931
+    },
+    {
+      "epoch": 0.9969164353993216,
+      "grad_norm": 4.147609710693359,
+      "learning_rate": 2.493772117695414e-10,
+      "loss": 0.9402,
+      "step": 12932
+    },
+    {
+      "epoch": 0.9969935245143385,
+      "grad_norm": 3.782233476638794,
+      "learning_rate": 2.3706430923808867e-10,
+      "loss": 0.969,
+      "step": 12933
+    },
+    {
+      "epoch": 0.9970706136293556,
+      "grad_norm": 3.95326566696167,
+      "learning_rate": 2.250631160316541e-10,
+      "loss": 1.0036,
+      "step": 12934
+    },
+    {
+      "epoch": 0.9971477027443725,
+      "grad_norm": 3.8084826469421387,
+      "learning_rate": 2.1337363289797296e-10,
+      "loss": 0.9255,
+      "step": 12935
+    },
+    {
+      "epoch": 0.9972247918593895,
+      "grad_norm": 4.2080535888671875,
+      "learning_rate": 2.0199586056590669e-10,
+      "loss": 0.9461,
+      "step": 12936
+    },
+    {
+      "epoch": 0.9973018809744064,
+      "grad_norm": 4.278135299682617,
+      "learning_rate": 1.909297997448878e-10,
+      "loss": 0.9677,
+      "step": 12937
+    },
+    {
+      "epoch": 0.9973789700894233,
+      "grad_norm": 3.921271800994873,
+      "learning_rate": 1.8017545112491984e-10,
+      "loss": 0.9193,
+      "step": 12938
+    },
+    {
+      "epoch": 0.9974560592044404,
+      "grad_norm": 3.6912286281585693,
+      "learning_rate": 1.697328153760225e-10,
+      "loss": 0.9062,
+      "step": 12939
+    },
+    {
+      "epoch": 0.9975331483194573,
+      "grad_norm": 3.613358736038208,
+      "learning_rate": 1.5960189314934149e-10,
+      "loss": 0.8264,
+      "step": 12940
+    },
+    {
+      "epoch": 0.9976102374344743,
+      "grad_norm": 4.033813953399658,
+      "learning_rate": 1.4978268507659376e-10,
+      "loss": 0.9703,
+      "step": 12941
+    },
+    {
+      "epoch": 0.9976873265494912,
+      "grad_norm": 3.622581958770752,
+      "learning_rate": 1.402751917700673e-10,
+      "loss": 0.8894,
+      "step": 12942
+    },
+    {
+      "epoch": 0.9977644156645081,
+      "grad_norm": 3.6066484451293945,
+      "learning_rate": 1.3107941382262124e-10,
+      "loss": 0.8503,
+      "step": 12943
+    },
+    {
+      "epoch": 0.9978415047795252,
+      "grad_norm": 3.685361385345459,
+      "learning_rate": 1.221953518071306e-10,
+      "loss": 0.8426,
+      "step": 12944
+    },
+    {
+      "epoch": 0.9979185938945421,
+      "grad_norm": 4.148501873016357,
+      "learning_rate": 1.1362300627815182e-10,
+      "loss": 0.9235,
+      "step": 12945
+    },
+    {
+      "epoch": 0.9979956830095591,
+      "grad_norm": 3.7229340076446533,
+      "learning_rate": 1.0536237776970215e-10,
+      "loss": 0.8819,
+      "step": 12946
+    },
+    {
+      "epoch": 0.998072772124576,
+      "grad_norm": 3.745091199874878,
+      "learning_rate": 9.741346679636998e-11,
+      "loss": 0.8432,
+      "step": 12947
+    },
+    {
+      "epoch": 0.9981498612395929,
+      "grad_norm": 3.642552137374878,
+      "learning_rate": 8.9776273854425e-11,
+      "loss": 0.7828,
+      "step": 12948
+    },
+    {
+      "epoch": 0.99822695035461,
+      "grad_norm": 4.068527698516846,
+      "learning_rate": 8.245079941959777e-11,
+      "loss": 0.8816,
+      "step": 12949
+    },
+    {
+      "epoch": 0.9983040394696269,
+      "grad_norm": 3.732131242752075,
+      "learning_rate": 7.543704394874507e-11,
+      "loss": 0.8963,
+      "step": 12950
+    },
+    {
+      "epoch": 0.9983811285846439,
+      "grad_norm": 4.001811504364014,
+      "learning_rate": 6.873500787873965e-11,
+      "loss": 0.9997,
+      "step": 12951
+    },
+    {
+      "epoch": 0.9984582176996608,
+      "grad_norm": 3.770373821258545,
+      "learning_rate": 6.234469162813561e-11,
+      "loss": 0.8653,
+      "step": 12952
+    },
+    {
+      "epoch": 0.9985353068146777,
+      "grad_norm": 3.548311710357666,
+      "learning_rate": 5.6266095595503e-11,
+      "loss": 0.8459,
+      "step": 12953
+    },
+    {
+      "epoch": 0.9986123959296948,
+      "grad_norm": 3.578949451446533,
+      "learning_rate": 5.049922015887276e-11,
+      "loss": 0.8083,
+      "step": 12954
+    },
+    {
+      "epoch": 0.9986894850447117,
+      "grad_norm": 3.936638593673706,
+      "learning_rate": 4.504406567795716e-11,
+      "loss": 0.9448,
+      "step": 12955
+    },
+    {
+      "epoch": 0.9987665741597287,
+      "grad_norm": 3.4746735095977783,
+      "learning_rate": 3.990063249359466e-11,
+      "loss": 0.8742,
+      "step": 12956
+    },
+    {
+      "epoch": 0.9988436632747456,
+      "grad_norm": 3.4810571670532227,
+      "learning_rate": 3.506892092552949e-11,
+      "loss": 0.8603,
+      "step": 12957
+    },
+    {
+      "epoch": 0.9989207523897625,
+      "grad_norm": 3.837407112121582,
+      "learning_rate": 3.054893127574232e-11,
+      "loss": 1.0381,
+      "step": 12958
+    },
+    {
+      "epoch": 0.9989978415047795,
+      "grad_norm": 3.7372279167175293,
+      "learning_rate": 2.6340663825674683e-11,
+      "loss": 0.8691,
+      "step": 12959
+    },
+    {
+      "epoch": 0.9990749306197965,
+      "grad_norm": 3.8073391914367676,
+      "learning_rate": 2.2444118837339212e-11,
+      "loss": 0.945,
+      "step": 12960
+    },
+    {
+      "epoch": 0.9991520197348135,
+      "grad_norm": 3.483799457550049,
+      "learning_rate": 1.8859296554429862e-11,
+      "loss": 0.8425,
+      "step": 12961
+    },
+    {
+      "epoch": 0.9992291088498304,
+      "grad_norm": 4.023273468017578,
+      "learning_rate": 1.5586197200101462e-11,
+      "loss": 0.905,
+      "step": 12962
+    },
+    {
+      "epoch": 0.9993061979648473,
+      "grad_norm": 3.542368173599243,
+      "learning_rate": 1.2624820978079932e-11,
+      "loss": 0.922,
+      "step": 12963
+    },
+    {
+      "epoch": 0.9993832870798643,
+      "grad_norm": 3.746952772140503,
+      "learning_rate": 9.975168073772523e-12,
+      "loss": 0.7879,
+      "step": 12964
+    },
+    {
+      "epoch": 0.9994603761948813,
+      "grad_norm": 3.6323859691619873,
+      "learning_rate": 7.63723865149224e-12,
+      "loss": 0.9078,
+      "step": 12965
+    },
+    {
+      "epoch": 0.9995374653098983,
+      "grad_norm": 3.747823476791382,
+      "learning_rate": 5.611032857788523e-12,
+      "loss": 0.8839,
+      "step": 12966
+    },
+    {
+      "epoch": 0.9996145544249152,
+      "grad_norm": 3.6762750148773193,
+      "learning_rate": 3.896550818116574e-12,
+      "loss": 0.86,
+      "step": 12967
+    },
+    {
+      "epoch": 0.9996916435399321,
+      "grad_norm": 3.4373939037323,
+      "learning_rate": 2.4937926401680245e-12,
+      "loss": 0.904,
+      "step": 12968
+    },
+    {
+      "epoch": 0.9997687326549491,
+      "grad_norm": 3.7813000679016113,
+      "learning_rate": 1.402758411095384e-12,
+      "loss": 0.9159,
+      "step": 12969
+    },
+    {
+      "epoch": 0.9998458217699661,
+      "grad_norm": 3.9511356353759766,
+      "learning_rate": 6.234481991773678e-13,
+      "loss": 0.864,
+      "step": 12970
+    },
+    {
+      "epoch": 0.9999229108849831,
+      "grad_norm": 4.2751851081848145,
+      "learning_rate": 1.5586205215356586e-13,
+      "loss": 0.8958,
+      "step": 12971
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.090433120727539,
+      "learning_rate": 0.0,
+      "loss": 0.8966,
+      "step": 12972
+    },
+    {
+      "epoch": 1.0,
+      "step": 12972,
+      "total_flos": 2.0878280805119427e+18,
+      "train_loss": 0.3483946987583135,
+      "train_runtime": 106063.3145,
+      "train_samples_per_second": 3.914,
+      "train_steps_per_second": 0.122
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12972,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.0878280805119427e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}