Model save

all_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 1.0,
-    "total_flos": 12929193738240.0,
-    "train_loss": 1.46957281581786,
-    "train_runtime": 1664.2063,
-    "train_samples": 50301,
-    "train_samples_per_second": 4.746,
-    "train_steps_per_second": 0.148
+    "total_flos": 19395247472640.0,
+    "train_loss": 1.5797494724631234,
+    "train_runtime": 5954.1324,
+    "train_samples": 50000,
+    "train_samples_per_second": 8.398,
+    "train_steps_per_second": 0.263
 }

runs/Dec19_02-24-16_gilbreth-j001.rcac.purdue.edu/events.out.tfevents.1734593392.gilbreth-j001.rcac.purdue.edu.118015.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:71e80db590a638eacaa6ab83ba2c948aa8777d86e5c952fcf75da11770bc8fbd
-size 72104
+oid sha256:0fe8795c5e501ac981fd5ee7c6f67f34f3e8c5f082140e2a2a5f695bee831edf
+size 72458

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 1.0,
-    "total_flos": 12929193738240.0,
-    "train_loss": 1.46957281581786,
-    "train_runtime": 1664.2063,
-    "train_samples": 50301,
-    "train_samples_per_second": 4.746,
-    "train_steps_per_second": 0.148
+    "total_flos": 19395247472640.0,
+    "train_loss": 1.5797494724631234,
+    "train_runtime": 5954.1324,
+    "train_samples": 50000,
+    "train_samples_per_second": 8.398,
+    "train_steps_per_second": 0.263
 }

trainer_state.json

@@ -3,381 +3,2222 @@
   "best_model_checkpoint": null,
   "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 247,
+  "global_step": 1563,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.004048582995951417,
-      "grad_norm": 13.330064642581538,
-      "learning_rate": 8.000000000000001e-07,
-      "loss": 1.9385,
+      "epoch": 0.0006397952655150352,
+      "grad_norm": 15.323076201566215,
+      "learning_rate": 1.2738853503184715e-07,
+      "loss": 1.9829,
       "step": 1
     },
     {
-      "epoch": 0.020242914979757085,
-      "grad_norm": 5.939947175837709,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 1.8644,
+      "epoch": 0.003198976327575176,
+      "grad_norm": 14.823043360198175,
+      "learning_rate": 6.369426751592357e-07,
+      "loss": 1.8995,
       "step": 5
     },
     {
-      "epoch": 0.04048582995951417,
-      "grad_norm": 4.3216413283721185,
-      "learning_rate": 8.000000000000001e-06,
-      "loss": 1.6585,
+      "epoch": 0.006397952655150352,
+      "grad_norm": 11.718086782886287,
+      "learning_rate": 1.2738853503184715e-06,
+      "loss": 1.917,
       "step": 10
     },
     {
-      "epoch": 0.06072874493927125,
-      "grad_norm": 3.6726967268003663,
-      "learning_rate": 1.2e-05,
-      "loss": 1.5636,
+      "epoch": 0.009596928982725527,
+      "grad_norm": 10.574744350675175,
+      "learning_rate": 1.9108280254777074e-06,
+      "loss": 1.7,
       "step": 15
     },
     {
-      "epoch": 0.08097165991902834,
-      "grad_norm": 3.5395303268548055,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.5441,
+      "epoch": 0.012795905310300703,
+      "grad_norm": 8.627255062756012,
+      "learning_rate": 2.547770700636943e-06,
+      "loss": 1.6993,
       "step": 20
     },
     {
-      "epoch": 0.10121457489878542,
-      "grad_norm": 3.3745662946717503,
-      "learning_rate": 2e-05,
-      "loss": 1.561,
+      "epoch": 0.01599488163787588,
+      "grad_norm": 9.131793656170954,
+      "learning_rate": 3.1847133757961785e-06,
+      "loss": 1.6257,
       "step": 25
     },
     {
-      "epoch": 0.1214574898785425,
-      "grad_norm": 3.4145999647680654,
-      "learning_rate": 1.9974977965945e-05,
-      "loss": 1.5477,
+      "epoch": 0.019193857965451054,
+      "grad_norm": 7.9893049703700445,
+      "learning_rate": 3.821656050955415e-06,
+      "loss": 1.6272,
       "step": 30
     },
     {
-      "epoch": 0.1417004048582996,
-      "grad_norm": 3.5987577428282123,
-      "learning_rate": 1.9900037084217637e-05,
-      "loss": 1.5828,
+      "epoch": 0.022392834293026232,
+      "grad_norm": 7.733631834659649,
+      "learning_rate": 4.45859872611465e-06,
+      "loss": 1.5541,
       "step": 35
     },
     {
-      "epoch": 0.16194331983805668,
-      "grad_norm": 3.275432270529186,
-      "learning_rate": 1.9775552389476865e-05,
-      "loss": 1.5512,
+      "epoch": 0.025591810620601407,
+      "grad_norm": 7.530183697607551,
+      "learning_rate": 5.095541401273886e-06,
+      "loss": 1.5557,
       "step": 40
     },
     {
-      "epoch": 0.18218623481781376,
-      "grad_norm": 3.42399668876103,
-      "learning_rate": 1.9602146853776894e-05,
-      "loss": 1.5506,
+      "epoch": 0.028790786948176585,
+      "grad_norm": 8.122480826283134,
+      "learning_rate": 5.732484076433121e-06,
+      "loss": 1.5423,
       "step": 45
     },
     {
-      "epoch": 0.20242914979757085,
-      "grad_norm": 3.0319493863000986,
-      "learning_rate": 1.938068826896166e-05,
-      "loss": 1.5446,
+      "epoch": 0.03198976327575176,
+      "grad_norm": 7.846208728274308,
+      "learning_rate": 6.369426751592357e-06,
+      "loss": 1.5706,
       "step": 50
     },
     {
-      "epoch": 0.22267206477732793,
-      "grad_norm": 3.1210434260623106,
-      "learning_rate": 1.911228490388136e-05,
-      "loss": 1.5406,
+      "epoch": 0.035188739603326934,
+      "grad_norm": 7.302277102340616,
+      "learning_rate": 7.006369426751593e-06,
+      "loss": 1.5771,
       "step": 55
     },
     {
-      "epoch": 0.242914979757085,
-      "grad_norm": 2.9914508124552546,
-      "learning_rate": 1.8798279958164295e-05,
-      "loss": 1.5289,
+      "epoch": 0.03838771593090211,
+      "grad_norm": 7.3614411108811115,
+      "learning_rate": 7.64331210191083e-06,
+      "loss": 1.5775,
       "step": 60
     },
     {
-      "epoch": 0.2631578947368421,
-      "grad_norm": 3.204722538572153,
-      "learning_rate": 1.8440244840299507e-05,
-      "loss": 1.5401,
+      "epoch": 0.04158669225847729,
+      "grad_norm": 7.399078296281114,
+      "learning_rate": 8.280254777070064e-06,
+      "loss": 1.5908,
       "step": 65
     },
     {
-      "epoch": 0.2834008097165992,
-      "grad_norm": 3.1147243057747502,
-      "learning_rate": 1.8039971303669407e-05,
-      "loss": 1.5122,
+      "epoch": 0.044785668586052464,
+      "grad_norm": 7.490487624354615,
+      "learning_rate": 8.9171974522293e-06,
+      "loss": 1.5568,
       "step": 70
     },
     {
-      "epoch": 0.30364372469635625,
-      "grad_norm": 3.350754622294628,
-      "learning_rate": 1.7599462479886976e-05,
-      "loss": 1.5253,
+      "epoch": 0.04798464491362764,
+      "grad_norm": 7.564410298064902,
+      "learning_rate": 9.554140127388536e-06,
+      "loss": 1.5511,
       "step": 75
     },
     {
-      "epoch": 0.32388663967611336,
-      "grad_norm": 3.0227437711153238,
-      "learning_rate": 1.712092285431026e-05,
-      "loss": 1.513,
+      "epoch": 0.05118362124120281,
+      "grad_norm": 7.705684760477408,
+      "learning_rate": 1.0191082802547772e-05,
+      "loss": 1.6143,
       "step": 80
     },
     {
-      "epoch": 0.3441295546558704,
-      "grad_norm": 2.898694632209032,
-      "learning_rate": 1.6606747233900816e-05,
-      "loss": 1.5189,
+      "epoch": 0.05438259756877799,
+      "grad_norm": 7.6730816479383535,
+      "learning_rate": 1.0828025477707008e-05,
+      "loss": 1.5796,
       "step": 85
     },
     {
-      "epoch": 0.3643724696356275,
-      "grad_norm": 2.862407919090872,
-      "learning_rate": 1.6059508762635482e-05,
-      "loss": 1.4922,
+      "epoch": 0.05758157389635317,
+      "grad_norm": 7.43682047789693,
+      "learning_rate": 1.1464968152866242e-05,
+      "loss": 1.5761,
       "step": 90
     },
     {
-      "epoch": 0.38461538461538464,
-      "grad_norm": 2.7545502314496306,
-      "learning_rate": 1.54819460444471e-05,
-      "loss": 1.5087,
+      "epoch": 0.060780550223928344,
+      "grad_norm": 11.428826358534083,
+      "learning_rate": 1.2101910828025478e-05,
+      "loss": 1.5828,
       "step": 95
     },
     {
-      "epoch": 0.4048582995951417,
-      "grad_norm": 2.779940917288299,
-      "learning_rate": 1.4876949438136348e-05,
-      "loss": 1.4865,
+      "epoch": 0.06397952655150352,
+      "grad_norm": 7.704374740474158,
+      "learning_rate": 1.2738853503184714e-05,
+      "loss": 1.6069,
       "step": 100
     },
     {
-      "epoch": 0.4251012145748988,
-      "grad_norm": 2.7351259055331902,
-      "learning_rate": 1.424754659284048e-05,
-      "loss": 1.4641,
+      "epoch": 0.0671785028790787,
+      "grad_norm": 7.947848645117073,
+      "learning_rate": 1.337579617834395e-05,
+      "loss": 1.6119,
       "step": 105
     },
     {
-      "epoch": 0.44534412955465585,
-      "grad_norm": 2.7031992560589035,
-      "learning_rate": 1.359688729644536e-05,
-      "loss": 1.4512,
+      "epoch": 0.07037747920665387,
+      "grad_norm": 8.20302637654498,
+      "learning_rate": 1.4012738853503186e-05,
+      "loss": 1.5882,
       "step": 110
     },
     {
-      "epoch": 0.46558704453441296,
-      "grad_norm": 8.916756258638008,
-      "learning_rate": 1.2928227712765504e-05,
-      "loss": 1.4894,
+      "epoch": 0.07357645553422905,
+      "grad_norm": 7.530911397270896,
+      "learning_rate": 1.464968152866242e-05,
+      "loss": 1.6174,
       "step": 115
     },
     {
-      "epoch": 0.48582995951417,
-      "grad_norm": 4.296610600782661,
-      "learning_rate": 1.2244914086375726e-05,
-      "loss": 1.4706,
+      "epoch": 0.07677543186180422,
+      "grad_norm": 7.262293241342851,
+      "learning_rate": 1.528662420382166e-05,
+      "loss": 1.6211,
       "step": 120
     },
     {
-      "epoch": 0.5060728744939271,
-      "grad_norm": 2.735539829128421,
-      "learning_rate": 1.155036599664198e-05,
-      "loss": 1.4668,
+      "epoch": 0.0799744081893794,
+      "grad_norm": 7.723922330201957,
+      "learning_rate": 1.5923566878980894e-05,
+      "loss": 1.6544,
       "step": 125
     },
     {
-      "epoch": 0.5263157894736842,
-      "grad_norm": 2.88320021783325,
-      "learning_rate": 1.0848059244755093e-05,
-      "loss": 1.4391,
+      "epoch": 0.08317338451695458,
+      "grad_norm": 7.13297821284291,
+      "learning_rate": 1.6560509554140128e-05,
+      "loss": 1.6268,
       "step": 130
     },
     {
-      "epoch": 0.5465587044534413,
-      "grad_norm": 2.7000741221149176,
-      "learning_rate": 1.0141508459407622e-05,
-      "loss": 1.4533,
+      "epoch": 0.08637236084452975,
+      "grad_norm": 7.2192751321936655,
+      "learning_rate": 1.7197452229299365e-05,
+      "loss": 1.6534,
       "step": 135
     },
     {
-      "epoch": 0.5668016194331984,
-      "grad_norm": 2.6467423688966245,
-      "learning_rate": 9.434249508162076e-06,
-      "loss": 1.4278,
+      "epoch": 0.08957133717210493,
+      "grad_norm": 6.990889749591103,
+      "learning_rate": 1.78343949044586e-05,
+      "loss": 1.6997,
       "step": 140
     },
     {
-      "epoch": 0.5870445344129555,
-      "grad_norm": 3.1187755082536985,
-      "learning_rate": 8.729821802531213e-06,
-      "loss": 1.4483,
+      "epoch": 0.0927703134996801,
+      "grad_norm": 7.257740391871613,
+      "learning_rate": 1.8471337579617837e-05,
+      "loss": 1.6957,
       "step": 145
     },
     {
-      "epoch": 0.6072874493927125,
-      "grad_norm": 2.6419226226448522,
-      "learning_rate": 8.031750585322948e-06,
-      "loss": 1.4408,
+      "epoch": 0.09596928982725528,
+      "grad_norm": 7.756638020052322,
+      "learning_rate": 1.910828025477707e-05,
+      "loss": 1.6632,
       "step": 150
     },
     {
-      "epoch": 0.6275303643724697,
-      "grad_norm": 2.6407581001383935,
-      "learning_rate": 7.343529288891239e-06,
-      "loss": 1.4166,
+      "epoch": 0.09916826615483046,
+      "grad_norm": 7.370520536957562,
+      "learning_rate": 1.9745222929936306e-05,
+      "loss": 1.7145,
       "step": 155
     },
     {
-      "epoch": 0.6477732793522267,
-      "grad_norm": 2.7267651184883057,
-      "learning_rate": 6.668602052579425e-06,
-      "loss": 1.4262,
+      "epoch": 0.10236724248240563,
+      "grad_norm": 6.877624705149041,
+      "learning_rate": 1.9999775332635076e-05,
+      "loss": 1.6744,
       "step": 160
     },
     {
-      "epoch": 0.6680161943319838,
-      "grad_norm": 2.6623303396026077,
-      "learning_rate": 6.010346486845837e-06,
-      "loss": 1.4019,
+      "epoch": 0.10556621880998081,
+      "grad_norm": 6.884630710566387,
+      "learning_rate": 1.999840240196313e-05,
+      "loss": 1.6345,
       "step": 165
     },
     {
-      "epoch": 0.6882591093117408,
-      "grad_norm": 2.590693399618635,
-      "learning_rate": 5.3720567703270135e-06,
-      "loss": 1.3975,
+      "epoch": 0.10876519513755598,
+      "grad_norm": 7.046088632558083,
+      "learning_rate": 1.9995781526975738e-05,
+      "loss": 1.7073,
       "step": 170
     },
     {
-      "epoch": 0.708502024291498,
-      "grad_norm": 2.6432168771675117,
-      "learning_rate": 4.756927164427685e-06,
-      "loss": 1.4022,
+      "epoch": 0.11196417146513116,
+      "grad_norm": 6.965702919638721,
+      "learning_rate": 1.9991913034795768e-05,
+      "loss": 1.7232,
       "step": 175
     },
     {
-      "epoch": 0.728744939271255,
-      "grad_norm": 2.548470851349637,
-      "learning_rate": 4.168036027937267e-06,
-      "loss": 1.4046,
+      "epoch": 0.11516314779270634,
+      "grad_norm": 7.079402156690239,
+      "learning_rate": 1.9986797408266636e-05,
+      "loss": 1.716,
       "step": 180
     },
     {
-      "epoch": 0.7489878542510121,
-      "grad_norm": 2.5411503805488307,
-      "learning_rate": 3.6083304116701535e-06,
-      "loss": 1.4005,
+      "epoch": 0.1183621241202815,
+      "grad_norm": 7.019445225187399,
+      "learning_rate": 1.9980435285892056e-05,
+      "loss": 1.7822,
       "step": 185
     },
     {
-      "epoch": 0.7692307692307693,
-      "grad_norm": 2.5226203469706676,
-      "learning_rate": 3.0806113102245395e-06,
-      "loss": 1.3827,
+      "epoch": 0.12156110044785669,
+      "grad_norm": 6.811705730204365,
+      "learning_rate": 1.9972827461756335e-05,
+      "loss": 1.7011,
       "step": 190
     },
     {
-      "epoch": 0.7894736842105263,
-      "grad_norm": 2.621236542921664,
-      "learning_rate": 2.587519644666001e-06,
-      "loss": 1.3835,
+      "epoch": 0.12476007677543186,
+      "grad_norm": 7.5561764667749785,
+      "learning_rate": 1.9963974885425267e-05,
+      "loss": 1.6588,
       "step": 195
     },
     {
-      "epoch": 0.8097165991902834,
-      "grad_norm": 2.6155833525581405,
-      "learning_rate": 2.1315230462840985e-06,
-      "loss": 1.3709,
+      "epoch": 0.12795905310300704,
+      "grad_norm": 6.672553387060304,
+      "learning_rate": 1.9953878661827603e-05,
+      "loss": 1.7291,
       "step": 200
     },
     {
-      "epoch": 0.8299595141700404,
-      "grad_norm": 2.605493543515508,
-      "learning_rate": 1.7149035075615795e-06,
-      "loss": 1.3617,
+      "epoch": 0.13115802943058222,
+      "grad_norm": 6.53812108461993,
+      "learning_rate": 1.994254005111715e-05,
+      "loss": 1.7216,
       "step": 205
     },
     {
-      "epoch": 0.8502024291497976,
-      "grad_norm": 2.575872270386264,
-      "learning_rate": 1.339745962155613e-06,
-      "loss": 1.3822,
+      "epoch": 0.1343570057581574,
+      "grad_norm": 6.715574278360845,
+      "learning_rate": 1.992996046851548e-05,
+      "loss": 1.7063,
       "step": 210
     },
     {
-      "epoch": 0.8704453441295547,
-      "grad_norm": 2.5712243306858857,
-      "learning_rate": 1.0079278510416313e-06,
-      "loss": 1.3686,
+      "epoch": 0.13755598208573255,
+      "grad_norm": 7.05804598737254,
+      "learning_rate": 1.9916141484135297e-05,
+      "loss": 1.7471,
       "step": 215
     },
     {
-      "epoch": 0.8906882591093117,
-      "grad_norm": 2.6224297264123373,
-      "learning_rate": 7.211097270349065e-07,
-      "loss": 1.3685,
+      "epoch": 0.14075495841330773,
+      "grad_norm": 6.896788944392643,
+      "learning_rate": 1.990108482278446e-05,
+      "loss": 1.7187,
       "step": 220
     },
     {
-      "epoch": 0.9109311740890689,
-      "grad_norm": 2.496616617072913,
-      "learning_rate": 4.807269447087348e-07,
-      "loss": 1.3757,
+      "epoch": 0.14395393474088292,
+      "grad_norm": 6.495502763769602,
+      "learning_rate": 1.9884792363750684e-05,
+      "loss": 1.7563,
       "step": 225
     },
     {
-      "epoch": 0.9311740890688259,
-      "grad_norm": 2.573961350030146,
-      "learning_rate": 2.879824772962381e-07,
-      "loss": 1.3759,
+      "epoch": 0.1471529110684581,
+      "grad_norm": 7.3712783353199525,
+      "learning_rate": 1.9867266140567024e-05,
+      "loss": 1.7451,
       "step": 230
     },
     {
-      "epoch": 0.951417004048583,
-      "grad_norm": 2.578127235388859,
-      "learning_rate": 1.4384089652291544e-07,
-      "loss": 1.3767,
+      "epoch": 0.15035188739603328,
+      "grad_norm": 6.581973419659573,
+      "learning_rate": 1.9848508340758014e-05,
+      "loss": 1.7456,
       "step": 235
     },
     {
-      "epoch": 0.97165991902834,
-      "grad_norm": 2.452106055618849,
-      "learning_rate": 4.902354549733979e-08,
-      "loss": 1.3693,
+      "epoch": 0.15355086372360843,
+      "grad_norm": 6.696696446630449,
+      "learning_rate": 1.9828521305566647e-05,
+      "loss": 1.7439,
       "step": 240
     },
     {
-      "epoch": 0.9919028340080972,
-      "grad_norm": 2.520963888815248,
-      "learning_rate": 4.00492881678427e-09,
-      "loss": 1.3966,
+      "epoch": 0.15674984005118361,
+      "grad_norm": 6.778517555098042,
+      "learning_rate": 1.9807307529662175e-05,
+      "loss": 1.8145,
       "step": 245
     },
+    {
+      "epoch": 0.1599488163787588,
+      "grad_norm": 7.852925285027587,
+      "learning_rate": 1.9784869660828708e-05,
+      "loss": 1.716,
+      "step": 250
+    },
+    {
+      "epoch": 0.16314779270633398,
+      "grad_norm": 6.847047325407262,
+      "learning_rate": 1.9761210499634754e-05,
+      "loss": 1.7158,
+      "step": 255
+    },
+    {
+      "epoch": 0.16634676903390916,
+      "grad_norm": 6.742686467830304,
+      "learning_rate": 1.973633299908365e-05,
+      "loss": 1.6697,
+      "step": 260
+    },
+    {
+      "epoch": 0.1695457453614843,
+      "grad_norm": 7.615828808218375,
+      "learning_rate": 1.9710240264245005e-05,
+      "loss": 1.7761,
+      "step": 265
+    },
+    {
+      "epoch": 0.1727447216890595,
+      "grad_norm": 6.344736398584996,
+      "learning_rate": 1.9682935551867132e-05,
+      "loss": 1.7123,
+      "step": 270
+    },
+    {
+      "epoch": 0.17594369801663468,
+      "grad_norm": 6.846464024553188,
+      "learning_rate": 1.9654422269970545e-05,
+      "loss": 1.7246,
+      "step": 275
+    },
+    {
+      "epoch": 0.17914267434420986,
+      "grad_norm": 7.172278203486134,
+      "learning_rate": 1.9624703977422624e-05,
+      "loss": 1.7149,
+      "step": 280
+    },
+    {
+      "epoch": 0.18234165067178504,
+      "grad_norm": 6.726988603672391,
+      "learning_rate": 1.959378438349338e-05,
+      "loss": 1.7048,
+      "step": 285
+    },
+    {
+      "epoch": 0.1855406269993602,
+      "grad_norm": 6.943476370244266,
+      "learning_rate": 1.956166734739251e-05,
+      "loss": 1.7305,
+      "step": 290
+    },
+    {
+      "epoch": 0.18873960332693537,
+      "grad_norm": 6.238347919067055,
+      "learning_rate": 1.95283568777877e-05,
+      "loss": 1.7278,
+      "step": 295
+    },
+    {
+      "epoch": 0.19193857965451055,
+      "grad_norm": 6.507904717046459,
+      "learning_rate": 1.9493857132304295e-05,
+      "loss": 1.6888,
+      "step": 300
+    },
+    {
+      "epoch": 0.19513755598208574,
+      "grad_norm": 6.109285663684619,
+      "learning_rate": 1.9458172417006347e-05,
+      "loss": 1.8163,
+      "step": 305
+    },
+    {
+      "epoch": 0.19833653230966092,
+      "grad_norm": 7.408832830125411,
+      "learning_rate": 1.942130718585919e-05,
+      "loss": 1.6901,
+      "step": 310
+    },
+    {
+      "epoch": 0.20153550863723607,
+      "grad_norm": 7.49888517401149,
+      "learning_rate": 1.938326604017349e-05,
+      "loss": 1.7287,
+      "step": 315
+    },
+    {
+      "epoch": 0.20473448496481125,
+      "grad_norm": 6.428397451627141,
+      "learning_rate": 1.9344053728030952e-05,
+      "loss": 1.7148,
+      "step": 320
+    },
+    {
+      "epoch": 0.20793346129238643,
+      "grad_norm": 5.934182360206382,
+      "learning_rate": 1.9303675143691683e-05,
+      "loss": 1.7617,
+      "step": 325
+    },
+    {
+      "epoch": 0.21113243761996162,
+      "grad_norm": 6.861546428758209,
+      "learning_rate": 1.9262135326983326e-05,
+      "loss": 1.7501,
+      "step": 330
+    },
+    {
+      "epoch": 0.2143314139475368,
+      "grad_norm": 6.911577067169947,
+      "learning_rate": 1.921943946267201e-05,
+      "loss": 1.6874,
+      "step": 335
+    },
+    {
+      "epoch": 0.21753039027511195,
+      "grad_norm": 7.155276304372121,
+      "learning_rate": 1.9175592879815217e-05,
+      "loss": 1.7567,
+      "step": 340
+    },
+    {
+      "epoch": 0.22072936660268713,
+      "grad_norm": 6.70480774058875,
+      "learning_rate": 1.9130601051096655e-05,
+      "loss": 1.6877,
+      "step": 345
+    },
+    {
+      "epoch": 0.22392834293026231,
+      "grad_norm": 8.290802152193233,
+      "learning_rate": 1.9084469592143154e-05,
+      "loss": 1.7157,
+      "step": 350
+    },
+    {
+      "epoch": 0.2271273192578375,
+      "grad_norm": 9.885557534520894,
+      "learning_rate": 1.9037204260823788e-05,
+      "loss": 1.6652,
+      "step": 355
+    },
+    {
+      "epoch": 0.23032629558541268,
+      "grad_norm": 6.802707826515414,
+      "learning_rate": 1.89888109565312e-05,
+      "loss": 1.8087,
+      "step": 360
+    },
+    {
+      "epoch": 0.23352527191298783,
+      "grad_norm": 6.50056121244608,
+      "learning_rate": 1.893929571944527e-05,
+      "loss": 1.7308,
+      "step": 365
+    },
+    {
+      "epoch": 0.236724248240563,
+      "grad_norm": 7.16685057869919,
+      "learning_rate": 1.8888664729779205e-05,
+      "loss": 1.7564,
+      "step": 370
+    },
+    {
+      "epoch": 0.2399232245681382,
+      "grad_norm": 6.81270552920015,
+      "learning_rate": 1.883692430700818e-05,
+      "loss": 1.7058,
+      "step": 375
+    },
+    {
+      "epoch": 0.24312220089571338,
+      "grad_norm": 7.226891729121006,
+      "learning_rate": 1.8784080909080568e-05,
+      "loss": 1.6884,
+      "step": 380
+    },
+    {
+      "epoch": 0.24632117722328856,
+      "grad_norm": 6.65472204642506,
+      "learning_rate": 1.8730141131611882e-05,
+      "loss": 1.7218,
+      "step": 385
+    },
+    {
+      "epoch": 0.2495201535508637,
+      "grad_norm": 6.032594521887261,
+      "learning_rate": 1.867511170706157e-05,
+      "loss": 1.6584,
+      "step": 390
+    },
+    {
+      "epoch": 0.2527191298784389,
+      "grad_norm": 6.400600540656193,
+      "learning_rate": 1.861899950389269e-05,
+      "loss": 1.7453,
+      "step": 395
+    },
+    {
+      "epoch": 0.2559181062060141,
+      "grad_norm": 6.170764677850133,
+      "learning_rate": 1.856181152571463e-05,
+      "loss": 1.6645,
+      "step": 400
+    },
+    {
+      "epoch": 0.2591170825335892,
+      "grad_norm": 6.298714448640878,
+      "learning_rate": 1.850355491040897e-05,
+      "loss": 1.6754,
+      "step": 405
+    },
+    {
+      "epoch": 0.26231605886116444,
+      "grad_norm": 5.536349073014043,
+      "learning_rate": 1.8444236929238556e-05,
+      "loss": 1.7155,
+      "step": 410
+    },
+    {
+      "epoch": 0.2655150351887396,
+      "grad_norm": 5.817758192259339,
+      "learning_rate": 1.8383864985939944e-05,
+      "loss": 1.6648,
+      "step": 415
+    },
+    {
+      "epoch": 0.2687140115163148,
+      "grad_norm": 6.302264713196719,
+      "learning_rate": 1.8322446615799317e-05,
+      "loss": 1.6764,
+      "step": 420
+    },
+    {
+      "epoch": 0.27191298784388995,
+      "grad_norm": 6.522045949334604,
+      "learning_rate": 1.825998948471197e-05,
+      "loss": 1.6565,
+      "step": 425
+    },
+    {
+      "epoch": 0.2751119641714651,
+      "grad_norm": 5.637323659265289,
+      "learning_rate": 1.819650138822548e-05,
+      "loss": 1.702,
+      "step": 430
+    },
+    {
+      "epoch": 0.2783109404990403,
+      "grad_norm": 6.095779597956818,
+      "learning_rate": 1.8131990250566733e-05,
+      "loss": 1.5878,
+      "step": 435
+    },
+    {
+      "epoch": 0.28150991682661547,
+      "grad_norm": 6.5246928335576335,
+      "learning_rate": 1.8066464123652857e-05,
+      "loss": 1.7256,
+      "step": 440
+    },
+    {
+      "epoch": 0.2847088931541907,
+      "grad_norm": 8.539143225554797,
+      "learning_rate": 1.7999931186086225e-05,
+      "loss": 1.6771,
+      "step": 445
+    },
+    {
+      "epoch": 0.28790786948176583,
+      "grad_norm": 6.891611040360784,
+      "learning_rate": 1.793239974213364e-05,
+      "loss": 1.6601,
+      "step": 450
+    },
+    {
+      "epoch": 0.291106845809341,
+      "grad_norm": 6.420229773712775,
+      "learning_rate": 1.786387822068987e-05,
+      "loss": 1.7079,
+      "step": 455
+    },
+    {
+      "epoch": 0.2943058221369162,
+      "grad_norm": 6.457834960102284,
+      "learning_rate": 1.7794375174225565e-05,
+      "loss": 1.7232,
+      "step": 460
+    },
+    {
+      "epoch": 0.29750479846449135,
+      "grad_norm": 6.202125604177199,
+      "learning_rate": 1.7723899277719815e-05,
+      "loss": 1.663,
+      "step": 465
+    },
+    {
+      "epoch": 0.30070377479206656,
+      "grad_norm": 6.215588562215511,
+      "learning_rate": 1.7652459327577377e-05,
+      "loss": 1.6851,
+      "step": 470
+    },
+    {
+      "epoch": 0.3039027511196417,
+      "grad_norm": 5.881272618270829,
+      "learning_rate": 1.7580064240530746e-05,
+      "loss": 1.6641,
+      "step": 475
+    },
+    {
+      "epoch": 0.30710172744721687,
+      "grad_norm": 6.32295524386819,
+      "learning_rate": 1.7506723052527243e-05,
+      "loss": 1.6806,
+      "step": 480
+    },
+    {
+      "epoch": 0.3103007037747921,
+      "grad_norm": 6.477043762062937,
+      "learning_rate": 1.7432444917601183e-05,
+      "loss": 1.679,
+      "step": 485
+    },
+    {
+      "epoch": 0.31349968010236723,
+      "grad_norm": 6.183332523319342,
+      "learning_rate": 1.735723910673132e-05,
+      "loss": 1.6618,
+      "step": 490
+    },
+    {
+      "epoch": 0.31669865642994244,
+      "grad_norm": 6.008689157765619,
+      "learning_rate": 1.7281115006683687e-05,
+      "loss": 1.6457,
+      "step": 495
+    },
+    {
+      "epoch": 0.3198976327575176,
+      "grad_norm": 5.46905856782914,
+      "learning_rate": 1.7204082118840035e-05,
+      "loss": 1.69,
+      "step": 500
+    },
+    {
+      "epoch": 0.32309660908509275,
+      "grad_norm": 6.5662567018918665,
+      "learning_rate": 1.712615005801185e-05,
+      "loss": 1.7046,
+      "step": 505
+    },
+    {
+      "epoch": 0.32629558541266795,
+      "grad_norm": 6.004480060557807,
+      "learning_rate": 1.704732855124037e-05,
+      "loss": 1.725,
+      "step": 510
+    },
+    {
+      "epoch": 0.3294945617402431,
+      "grad_norm": 6.079836989604523,
+      "learning_rate": 1.6967627436582445e-05,
+      "loss": 1.7011,
+      "step": 515
+    },
+    {
+      "epoch": 0.3326935380678183,
+      "grad_norm": 6.378804308358523,
+      "learning_rate": 1.6887056661882644e-05,
+      "loss": 1.7008,
+      "step": 520
+    },
+    {
+      "epoch": 0.33589251439539347,
+      "grad_norm": 6.646390728744813,
+      "learning_rate": 1.6805626283531592e-05,
+      "loss": 1.6512,
+      "step": 525
+    },
+    {
+      "epoch": 0.3390914907229686,
+      "grad_norm": 6.213451754689538,
+      "learning_rate": 1.6723346465210815e-05,
+      "loss": 1.6966,
+      "step": 530
+    },
+    {
+      "epoch": 0.34229046705054383,
+      "grad_norm": 6.2533054007830895,
+      "learning_rate": 1.6640227476624163e-05,
+      "loss": 1.7,
+      "step": 535
+    },
+    {
+      "epoch": 0.345489443378119,
+      "grad_norm": 5.900731425972058,
+      "learning_rate": 1.655627969221598e-05,
+      "loss": 1.6349,
+      "step": 540
+    },
+    {
+      "epoch": 0.3486884197056942,
+      "grad_norm": 6.517153961385569,
+      "learning_rate": 1.6471513589876247e-05,
+      "loss": 1.6806,
+      "step": 545
+    },
+    {
+      "epoch": 0.35188739603326935,
+      "grad_norm": 7.471236190819774,
+      "learning_rate": 1.638593974963278e-05,
+      "loss": 1.6744,
+      "step": 550
+    },
+    {
+      "epoch": 0.3550863723608445,
+      "grad_norm": 5.773453148094793,
+      "learning_rate": 1.6299568852330703e-05,
+      "loss": 1.711,
+      "step": 555
+    },
+    {
+      "epoch": 0.3582853486884197,
+      "grad_norm": 6.1708780610342275,
+      "learning_rate": 1.6212411678299306e-05,
+      "loss": 1.6594,
+      "step": 560
+    },
+    {
+      "epoch": 0.36148432501599487,
+      "grad_norm": 5.528698141123457,
+      "learning_rate": 1.612447910600652e-05,
+      "loss": 1.636,
+      "step": 565
+    },
+    {
+      "epoch": 0.3646833013435701,
+      "grad_norm": 5.712561484363837,
+      "learning_rate": 1.6035782110701125e-05,
+      "loss": 1.6579,
+      "step": 570
+    },
+    {
+      "epoch": 0.36788227767114523,
+      "grad_norm": 5.842343602991339,
+      "learning_rate": 1.594633176304287e-05,
+      "loss": 1.674,
+      "step": 575
+    },
+    {
+      "epoch": 0.3710812539987204,
+      "grad_norm": 6.6877770039701785,
+      "learning_rate": 1.5856139227720714e-05,
+      "loss": 1.6873,
+      "step": 580
+    },
+    {
+      "epoch": 0.3742802303262956,
+      "grad_norm": 5.8801742335950395,
+      "learning_rate": 1.5765215762059304e-05,
+      "loss": 1.663,
+      "step": 585
+    },
+    {
+      "epoch": 0.37747920665387075,
+      "grad_norm": 5.932829190253789,
+      "learning_rate": 1.5673572714613886e-05,
+      "loss": 1.6996,
+      "step": 590
+    },
+    {
+      "epoch": 0.38067818298144596,
+      "grad_norm": 6.16552928091471,
+      "learning_rate": 1.558122152375387e-05,
+      "loss": 1.6666,
+      "step": 595
+    },
+    {
+      "epoch": 0.3838771593090211,
+      "grad_norm": 5.683308025237575,
+      "learning_rate": 1.548817371623513e-05,
+      "loss": 1.616,
+      "step": 600
+    },
+    {
+      "epoch": 0.38707613563659626,
+      "grad_norm": 5.221984427669265,
+      "learning_rate": 1.539444090576132e-05,
+      "loss": 1.658,
+      "step": 605
+    },
+    {
+      "epoch": 0.3902751119641715,
+      "grad_norm": 6.046167249724566,
+      "learning_rate": 1.5300034791534297e-05,
+      "loss": 1.6237,
+      "step": 610
+    },
+    {
+      "epoch": 0.3934740882917466,
+      "grad_norm": 5.558556190388178,
+      "learning_rate": 1.520496715679391e-05,
+      "loss": 1.6321,
+      "step": 615
+    },
+    {
+      "epoch": 0.39667306461932184,
+      "grad_norm": 5.687939663666506,
+      "learning_rate": 1.5109249867347276e-05,
+      "loss": 1.6257,
+      "step": 620
+    },
+    {
+      "epoch": 0.399872040946897,
+      "grad_norm": 5.242983409056203,
+      "learning_rate": 1.5012894870087751e-05,
+      "loss": 1.6671,
+      "step": 625
+    },
+    {
+      "epoch": 0.40307101727447214,
+      "grad_norm": 6.789156081853467,
+      "learning_rate": 1.4915914191503792e-05,
+      "loss": 1.6646,
+      "step": 630
+    },
+    {
+      "epoch": 0.40626999360204735,
+      "grad_norm": 6.528979549468428,
+      "learning_rate": 1.4818319936177885e-05,
+      "loss": 1.6607,
+      "step": 635
+    },
+    {
+      "epoch": 0.4094689699296225,
+      "grad_norm": 8.278909196552412,
+      "learning_rate": 1.4720124285275703e-05,
+      "loss": 1.6373,
+      "step": 640
+    },
+    {
+      "epoch": 0.4126679462571977,
+      "grad_norm": 5.5782894278332,
+      "learning_rate": 1.4621339495025731e-05,
+      "loss": 1.6135,
+      "step": 645
+    },
+    {
+      "epoch": 0.41586692258477287,
+      "grad_norm": 5.902082388957196,
+      "learning_rate": 1.4521977895189518e-05,
+      "loss": 1.6117,
+      "step": 650
+    },
+    {
+      "epoch": 0.419065898912348,
+      "grad_norm": 6.256546081010292,
+      "learning_rate": 1.4422051887522735e-05,
+      "loss": 1.6672,
+      "step": 655
+    },
+    {
+      "epoch": 0.42226487523992323,
+      "grad_norm": 5.470666002879475,
+      "learning_rate": 1.4321573944227254e-05,
+      "loss": 1.6357,
+      "step": 660
+    },
+    {
+      "epoch": 0.4254638515674984,
+      "grad_norm": 6.199108039992263,
+      "learning_rate": 1.4220556606394465e-05,
+      "loss": 1.6554,
+      "step": 665
+    },
+    {
+      "epoch": 0.4286628278950736,
+      "grad_norm": 6.36594443161433,
+      "learning_rate": 1.4119012482439929e-05,
+      "loss": 1.6275,
+      "step": 670
+    },
+    {
+      "epoch": 0.43186180422264875,
+      "grad_norm": 5.968167544313682,
+      "learning_rate": 1.4016954246529697e-05,
+      "loss": 1.5722,
+      "step": 675
+    },
+    {
+      "epoch": 0.4350607805502239,
+      "grad_norm": 5.258817540465498,
+      "learning_rate": 1.3914394636998374e-05,
+      "loss": 1.5969,
+      "step": 680
+    },
+    {
+      "epoch": 0.4382597568777991,
+      "grad_norm": 5.695461561961905,
+      "learning_rate": 1.3811346454759211e-05,
+      "loss": 1.6559,
+      "step": 685
+    },
+    {
+      "epoch": 0.44145873320537427,
+      "grad_norm": 5.833470029743587,
+      "learning_rate": 1.3707822561706336e-05,
+      "loss": 1.6515,
+      "step": 690
+    },
+    {
+      "epoch": 0.4446577095329495,
+      "grad_norm": 6.796628890137597,
+      "learning_rate": 1.3603835879109442e-05,
+      "loss": 1.5562,
+      "step": 695
+    },
+    {
+      "epoch": 0.44785668586052463,
+      "grad_norm": 5.42642421978782,
+      "learning_rate": 1.349939938600099e-05,
+      "loss": 1.6267,
+      "step": 700
+    },
+    {
+      "epoch": 0.4510556621880998,
+      "grad_norm": 5.77504172098871,
+      "learning_rate": 1.3394526117556277e-05,
+      "loss": 1.6096,
+      "step": 705
+    },
+    {
+      "epoch": 0.454254638515675,
+      "grad_norm": 5.682023186997677,
+      "learning_rate": 1.3289229163466421e-05,
+      "loss": 1.6159,
+      "step": 710
+    },
+    {
+      "epoch": 0.45745361484325014,
+      "grad_norm": 5.980415972756858,
+      "learning_rate": 1.3183521666304611e-05,
+      "loss": 1.5665,
+      "step": 715
+    },
+    {
+      "epoch": 0.46065259117082535,
+      "grad_norm": 6.277907654795176,
+      "learning_rate": 1.3077416819885707e-05,
+      "loss": 1.6181,
+      "step": 720
+    },
+    {
+      "epoch": 0.4638515674984005,
+      "grad_norm": 5.403104884415684,
+      "learning_rate": 1.297092786761946e-05,
+      "loss": 1.6433,
+      "step": 725
+    },
+    {
+      "epoch": 0.46705054382597566,
+      "grad_norm": 5.7504585319819,
+      "learning_rate": 1.2864068100857565e-05,
+      "loss": 1.5969,
+      "step": 730
+    },
+    {
+      "epoch": 0.47024952015355087,
+      "grad_norm": 5.314291013995357,
+      "learning_rate": 1.2756850857234686e-05,
+      "loss": 1.537,
+      "step": 735
+    },
+    {
+      "epoch": 0.473448496481126,
+      "grad_norm": 11.558618361731758,
+      "learning_rate": 1.2649289519003739e-05,
+      "loss": 1.6282,
+      "step": 740
+    },
+    {
+      "epoch": 0.47664747280870123,
+      "grad_norm": 6.275202197895068,
+      "learning_rate": 1.2541397511365584e-05,
+      "loss": 1.6913,
+      "step": 745
+    },
+    {
+      "epoch": 0.4798464491362764,
+      "grad_norm": 5.848966201444489,
+      "learning_rate": 1.2433188300793399e-05,
+      "loss": 1.6419,
+      "step": 750
+    },
+    {
+      "epoch": 0.48304542546385154,
+      "grad_norm": 5.45346150224839,
+      "learning_rate": 1.2324675393351818e-05,
+      "loss": 1.6389,
+      "step": 755
+    },
+    {
+      "epoch": 0.48624440179142675,
+      "grad_norm": 5.472275124402878,
+      "learning_rate": 1.221587233301123e-05,
+      "loss": 1.6459,
+      "step": 760
+    },
+    {
+      "epoch": 0.4894433781190019,
+      "grad_norm": 5.84560048069194,
+      "learning_rate": 1.2106792699957264e-05,
+      "loss": 1.5909,
+      "step": 765
+    },
+    {
+      "epoch": 0.4926423544465771,
+      "grad_norm": 5.882557613681303,
+      "learning_rate": 1.1997450108895807e-05,
+      "loss": 1.6056,
+      "step": 770
+    },
+    {
+      "epoch": 0.49584133077415227,
+      "grad_norm": 5.597349805942927,
+      "learning_rate": 1.1887858207353678e-05,
+      "loss": 1.5987,
+      "step": 775
+    },
+    {
+      "epoch": 0.4990403071017274,
+      "grad_norm": 5.799922769217321,
+      "learning_rate": 1.1778030673975227e-05,
+      "loss": 1.6334,
+      "step": 780
+    },
+    {
+      "epoch": 0.5022392834293026,
+      "grad_norm": 5.094448679746164,
+      "learning_rate": 1.166798121681505e-05,
+      "loss": 1.5556,
+      "step": 785
+    },
+    {
+      "epoch": 0.5054382597568778,
+      "grad_norm": 6.0295100152848065,
+      "learning_rate": 1.1557723571627016e-05,
+      "loss": 1.5982,
+      "step": 790
+    },
+    {
+      "epoch": 0.508637236084453,
+      "grad_norm": 5.54607761020907,
+      "learning_rate": 1.1447271500149849e-05,
+      "loss": 1.6196,
+      "step": 795
+    },
+    {
+      "epoch": 0.5118362124120281,
+      "grad_norm": 6.170173231609507,
+      "learning_rate": 1.1336638788389473e-05,
+      "loss": 1.5981,
+      "step": 800
+    },
+    {
+      "epoch": 0.5150351887396033,
+      "grad_norm": 5.688859674100479,
+      "learning_rate": 1.122583924489832e-05,
+      "loss": 1.5606,
+      "step": 805
+    },
+    {
+      "epoch": 0.5182341650671785,
+      "grad_norm": 5.419022876394577,
+      "learning_rate": 1.1114886699051803e-05,
+      "loss": 1.6077,
+      "step": 810
+    },
+    {
+      "epoch": 0.5214331413947537,
+      "grad_norm": 5.611555356116697,
+      "learning_rate": 1.1003794999322246e-05,
+      "loss": 1.6013,
+      "step": 815
+    },
+    {
+      "epoch": 0.5246321177223289,
+      "grad_norm": 5.417467087897305,
+      "learning_rate": 1.089257801155037e-05,
+      "loss": 1.5995,
+      "step": 820
+    },
+    {
+      "epoch": 0.527831094049904,
+      "grad_norm": 5.692297421068607,
+      "learning_rate": 1.0781249617214642e-05,
+      "loss": 1.5979,
+      "step": 825
+    },
+    {
+      "epoch": 0.5310300703774792,
+      "grad_norm": 6.805515751516561,
+      "learning_rate": 1.0669823711698668e-05,
+      "loss": 1.6004,
+      "step": 830
+    },
+    {
+      "epoch": 0.5342290467050543,
+      "grad_norm": 5.878915128709559,
+      "learning_rate": 1.0558314202556866e-05,
+      "loss": 1.5874,
+      "step": 835
+    },
+    {
+      "epoch": 0.5374280230326296,
+      "grad_norm": 5.782128606872289,
+      "learning_rate": 1.0446735007778563e-05,
+      "loss": 1.6118,
+      "step": 840
+    },
+    {
+      "epoch": 0.5406269993602048,
+      "grad_norm": 5.475651082333863,
+      "learning_rate": 1.0335100054050877e-05,
+      "loss": 1.541,
+      "step": 845
+    },
+    {
+      "epoch": 0.5438259756877799,
+      "grad_norm": 5.395391495213094,
+      "learning_rate": 1.0223423275020431e-05,
+      "loss": 1.5621,
+      "step": 850
+    },
+    {
+      "epoch": 0.5470249520153551,
+      "grad_norm": 5.221419737221846,
+      "learning_rate": 1.0111718609554254e-05,
+      "loss": 1.588,
+      "step": 855
+    },
+    {
+      "epoch": 0.5502239283429302,
+      "grad_norm": 5.792917502556416,
+      "learning_rate": 1e-05,
+      "loss": 1.5542,
+      "step": 860
+    },
+    {
+      "epoch": 0.5534229046705055,
+      "grad_norm": 5.401563370754301,
+      "learning_rate": 9.888281390445747e-06,
+      "loss": 1.501,
+      "step": 865
+    },
+    {
+      "epoch": 0.5566218809980806,
+      "grad_norm": 5.5625432258357135,
+      "learning_rate": 9.776576724979572e-06,
+      "loss": 1.4808,
+      "step": 870
+    },
+    {
+      "epoch": 0.5598208573256558,
+      "grad_norm": 5.431483361375168,
+      "learning_rate": 9.664899945949128e-06,
+      "loss": 1.6402,
+      "step": 875
+    },
+    {
+      "epoch": 0.5630198336532309,
+      "grad_norm": 5.523596732819312,
+      "learning_rate": 9.553264992221442e-06,
+      "loss": 1.5527,
+      "step": 880
+    },
+    {
+      "epoch": 0.5662188099808061,
+      "grad_norm": 5.760169541114645,
+      "learning_rate": 9.441685797443138e-06,
+      "loss": 1.6082,
+      "step": 885
+    },
+    {
+      "epoch": 0.5694177863083814,
+      "grad_norm": 5.3583708767638045,
+      "learning_rate": 9.330176288301332e-06,
+      "loss": 1.5724,
+      "step": 890
+    },
+    {
+      "epoch": 0.5726167626359565,
+      "grad_norm": 5.246770440344478,
+      "learning_rate": 9.21875038278536e-06,
+      "loss": 1.6719,
+      "step": 895
+    },
+    {
+      "epoch": 0.5758157389635317,
+      "grad_norm": 5.07648174669728,
+      "learning_rate": 9.107421988449632e-06,
+      "loss": 1.5201,
+      "step": 900
+    },
+    {
+      "epoch": 0.5790147152911068,
+      "grad_norm": 5.658963128725044,
+      "learning_rate": 8.996205000677758e-06,
+      "loss": 1.5582,
+      "step": 905
+    },
+    {
+      "epoch": 0.582213691618682,
+      "grad_norm": 6.2255938362038075,
+      "learning_rate": 8.885113300948199e-06,
+      "loss": 1.5257,
+      "step": 910
+    },
+    {
+      "epoch": 0.5854126679462572,
+      "grad_norm": 5.719062994663868,
+      "learning_rate": 8.774160755101685e-06,
+      "loss": 1.5217,
+      "step": 915
+    },
+    {
+      "epoch": 0.5886116442738324,
+      "grad_norm": 5.177110225005978,
+      "learning_rate": 8.663361211610529e-06,
+      "loss": 1.5143,
+      "step": 920
+    },
+    {
+      "epoch": 0.5918106206014075,
+      "grad_norm": 7.670116582941386,
+      "learning_rate": 8.552728499850153e-06,
+      "loss": 1.531,
+      "step": 925
+    },
+    {
+      "epoch": 0.5950095969289827,
+      "grad_norm": 5.070674776722329,
+      "learning_rate": 8.442276428372986e-06,
+      "loss": 1.5987,
+      "step": 930
+    },
+    {
+      "epoch": 0.5982085732565579,
+      "grad_norm": 5.553438306499765,
+      "learning_rate": 8.332018783184952e-06,
+      "loss": 1.5857,
+      "step": 935
+    },
+    {
+      "epoch": 0.6014075495841331,
+      "grad_norm": 5.3841669865680775,
+      "learning_rate": 8.221969326024776e-06,
+      "loss": 1.4961,
+      "step": 940
+    },
+    {
+      "epoch": 0.6046065259117083,
+      "grad_norm": 5.222792058631912,
+      "learning_rate": 8.112141792646326e-06,
+      "loss": 1.5973,
+      "step": 945
+    },
+    {
+      "epoch": 0.6078055022392834,
+      "grad_norm": 5.540974220686552,
+      "learning_rate": 8.002549891104196e-06,
+      "loss": 1.5191,
+      "step": 950
+    },
+    {
+      "epoch": 0.6110044785668586,
+      "grad_norm": 5.699565608658352,
+      "learning_rate": 7.89320730004274e-06,
+      "loss": 1.5266,
+      "step": 955
+    },
+    {
+      "epoch": 0.6142034548944337,
+      "grad_norm": 5.433959384544603,
+      "learning_rate": 7.784127666988774e-06,
+      "loss": 1.579,
+      "step": 960
+    },
+    {
+      "epoch": 0.617402431222009,
+      "grad_norm": 5.876158900491931,
+      "learning_rate": 7.675324606648187e-06,
+      "loss": 1.4826,
+      "step": 965
+    },
+    {
+      "epoch": 0.6206014075495841,
+      "grad_norm": 6.010682309767372,
+      "learning_rate": 7.566811699206604e-06,
+      "loss": 1.5438,
+      "step": 970
+    },
+    {
+      "epoch": 0.6238003838771593,
+      "grad_norm": 5.503850073821662,
+      "learning_rate": 7.458602488634416e-06,
+      "loss": 1.5267,
+      "step": 975
+    },
+    {
+      "epoch": 0.6269993602047345,
+      "grad_norm": 5.8220792177360385,
+      "learning_rate": 7.350710480996266e-06,
+      "loss": 1.5472,
+      "step": 980
+    },
+    {
+      "epoch": 0.6301983365323096,
+      "grad_norm": 5.483991198540493,
+      "learning_rate": 7.243149142765317e-06,
+      "loss": 1.5392,
+      "step": 985
+    },
+    {
+      "epoch": 0.6333973128598849,
+      "grad_norm": 6.004099510058302,
+      "learning_rate": 7.135931899142438e-06,
+      "loss": 1.525,
+      "step": 990
+    },
+    {
+      "epoch": 0.63659628918746,
+      "grad_norm": 5.513848927142678,
+      "learning_rate": 7.029072132380543e-06,
+      "loss": 1.5274,
+      "step": 995
+    },
+    {
+      "epoch": 0.6397952655150352,
+      "grad_norm": 5.16006305552109,
+      "learning_rate": 6.922583180114299e-06,
+      "loss": 1.5518,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6429942418426103,
+      "grad_norm": 5.4181823490399,
+      "learning_rate": 6.816478333695392e-06,
+      "loss": 1.4788,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6461932181701855,
+      "grad_norm": 5.549211186008316,
+      "learning_rate": 6.710770836533584e-06,
+      "loss": 1.5491,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6493921944977608,
+      "grad_norm": 5.197415635909047,
+      "learning_rate": 6.605473882443725e-06,
+      "loss": 1.4588,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6525911708253359,
+      "grad_norm": 5.298268223394051,
+      "learning_rate": 6.5006006139990115e-06,
+      "loss": 1.5561,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6557901471529111,
+      "grad_norm": 5.293116661669245,
+      "learning_rate": 6.396164120890562e-06,
+      "loss": 1.5319,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6589891234804862,
+      "grad_norm": 5.604227519996248,
+      "learning_rate": 6.292177438293665e-06,
+      "loss": 1.5662,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6621880998080614,
+      "grad_norm": 5.576949710298901,
+      "learning_rate": 6.188653545240793e-06,
+      "loss": 1.4851,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6653870761356366,
+      "grad_norm": 5.2704551354406535,
+      "learning_rate": 6.085605363001628e-06,
+      "loss": 1.4728,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6685860524632118,
+      "grad_norm": 5.566308023233107,
+      "learning_rate": 5.983045753470308e-06,
+      "loss": 1.5402,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6717850287907869,
+      "grad_norm": 5.529019248649365,
+      "learning_rate": 5.880987517560075e-06,
+      "loss": 1.4701,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6749840051183621,
+      "grad_norm": 9.623197560440904,
+      "learning_rate": 5.779443393605536e-06,
+      "loss": 1.4883,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6781829814459372,
+      "grad_norm": 5.759928000459023,
+      "learning_rate": 5.678426055772746e-06,
+      "loss": 1.5201,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6813819577735125,
+      "grad_norm": 5.448797711699774,
+      "learning_rate": 5.577948112477271e-06,
+      "loss": 1.5401,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6845809341010877,
+      "grad_norm": 5.302088912757014,
+      "learning_rate": 5.478022104810483e-06,
+      "loss": 1.491,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6877799104286628,
+      "grad_norm": 5.256642109983724,
+      "learning_rate": 5.378660504974271e-06,
+      "loss": 1.4485,
+      "step": 1075
+    },
+    {
+      "epoch": 0.690978886756238,
+      "grad_norm": 5.508918096934492,
+      "learning_rate": 5.2798757147242986e-06,
+      "loss": 1.5478,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6941778630838131,
+      "grad_norm": 6.123712019243396,
+      "learning_rate": 5.1816800638221176e-06,
+      "loss": 1.5032,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6973768394113884,
+      "grad_norm": 5.4952266371272245,
+      "learning_rate": 5.084085808496211e-06,
+      "loss": 1.507,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7005758157389635,
+      "grad_norm": 5.262665430053801,
+      "learning_rate": 4.987105129912252e-06,
+      "loss": 1.4779,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7037747920665387,
+      "grad_norm": 5.635309261763986,
+      "learning_rate": 4.890750132652724e-06,
+      "loss": 1.4992,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7069737683941139,
+      "grad_norm": 5.48037987991141,
+      "learning_rate": 4.795032843206091e-06,
+      "loss": 1.4693,
+      "step": 1105
+    },
+    {
+      "epoch": 0.710172744721689,
+      "grad_norm": 5.877515738094301,
+      "learning_rate": 4.699965208465702e-06,
+      "loss": 1.5334,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7133717210492643,
+      "grad_norm": 5.33814319359391,
+      "learning_rate": 4.605559094238681e-06,
+      "loss": 1.5627,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7165706973768394,
+      "grad_norm": 4.995256226504251,
+      "learning_rate": 4.511826283764872e-06,
+      "loss": 1.5306,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7197696737044146,
+      "grad_norm": 4.622852037416148,
+      "learning_rate": 4.418778476246132e-06,
+      "loss": 1.4096,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7229686500319897,
+      "grad_norm": 5.591264404205622,
+      "learning_rate": 4.326427285386118e-06,
+      "loss": 1.5018,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7261676263595649,
+      "grad_norm": 5.775323577623916,
+      "learning_rate": 4.234784237940705e-06,
+      "loss": 1.4565,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7293666026871402,
+      "grad_norm": 5.167667950217927,
+      "learning_rate": 4.143860772279287e-06,
+      "loss": 1.4658,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7325655790147153,
+      "grad_norm": 5.353427528561617,
+      "learning_rate": 4.053668236957135e-06,
+      "loss": 1.5028,
+      "step": 1145
+    },
+    {
+      "epoch": 0.7357645553422905,
+      "grad_norm": 5.23743913834742,
+      "learning_rate": 3.964217889298882e-06,
+      "loss": 1.4788,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7389635316698656,
+      "grad_norm": 5.18848027092107,
+      "learning_rate": 3.875520893993482e-06,
+      "loss": 1.5132,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7421625079974408,
+      "grad_norm": 5.391670563602561,
+      "learning_rate": 3.787588321700697e-06,
+      "loss": 1.5261,
+      "step": 1160
+    },
+    {
+      "epoch": 0.745361484325016,
+      "grad_norm": 5.029103034550255,
+      "learning_rate": 3.7004311476692977e-06,
+      "loss": 1.5038,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7485604606525912,
+      "grad_norm": 5.3137658148388445,
+      "learning_rate": 3.6140602503672217e-06,
+      "loss": 1.4918,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7517594369801663,
+      "grad_norm": 6.110109081596603,
+      "learning_rate": 3.528486410123758e-06,
+      "loss": 1.4484,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7549584133077415,
+      "grad_norm": 5.2429270345932855,
+      "learning_rate": 3.443720307784022e-06,
+      "loss": 1.4757,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7581573896353166,
+      "grad_norm": 5.585128429142801,
+      "learning_rate": 3.359772523375837e-06,
+      "loss": 1.5187,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7613563659628919,
+      "grad_norm": 5.600172102657646,
+      "learning_rate": 3.276653534789185e-06,
+      "loss": 1.4846,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7645553422904671,
+      "grad_norm": 5.229330933662747,
+      "learning_rate": 3.1943737164684094e-06,
+      "loss": 1.4935,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7677543186180422,
+      "grad_norm": 5.417591664700107,
+      "learning_rate": 3.11294333811736e-06,
+      "loss": 1.4189,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7709532949456174,
+      "grad_norm": 5.064143998955478,
+      "learning_rate": 3.032372563417556e-06,
+      "loss": 1.4799,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7741522712731925,
+      "grad_norm": 5.889945863254974,
+      "learning_rate": 2.952671448759631e-06,
+      "loss": 1.4804,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7773512476007678,
+      "grad_norm": 5.37010722185697,
+      "learning_rate": 2.8738499419881517e-06,
+      "loss": 1.495,
+      "step": 1215
+    },
+    {
+      "epoch": 0.780550223928343,
+      "grad_norm": 5.6156285209396835,
+      "learning_rate": 2.795917881159973e-06,
+      "loss": 1.4205,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7837492002559181,
+      "grad_norm": 6.104878893961979,
+      "learning_rate": 2.718884993316311e-06,
+      "loss": 1.4912,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7869481765834933,
+      "grad_norm": 5.2999746074923575,
+      "learning_rate": 2.642760893268684e-06,
+      "loss": 1.4611,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7901471529110684,
+      "grad_norm": 5.258668581098616,
+      "learning_rate": 2.5675550823988184e-06,
+      "loss": 1.4701,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7933461292386437,
+      "grad_norm": 5.50184667895086,
+      "learning_rate": 2.493276947472756e-06,
+      "loss": 1.4752,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7965451055662188,
+      "grad_norm": 5.352416363928525,
+      "learning_rate": 2.4199357594692564e-06,
+      "loss": 1.4072,
+      "step": 1245
+    },
+    {
+      "epoch": 0.799744081893794,
+      "grad_norm": 5.739241303953708,
+      "learning_rate": 2.347540672422627e-06,
+      "loss": 1.4234,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8029430582213691,
+      "grad_norm": 5.169847996231581,
+      "learning_rate": 2.2761007222801866e-06,
+      "loss": 1.4544,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8061420345489443,
+      "grad_norm": 5.360062306660903,
+      "learning_rate": 2.2056248257744383e-06,
+      "loss": 1.4451,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8093410108765196,
+      "grad_norm": 5.28613855172683,
+      "learning_rate": 2.136121779310132e-06,
+      "loss": 1.4517,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8125399872040947,
+      "grad_norm": 5.54426483916989,
+      "learning_rate": 2.067600257866358e-06,
+      "loss": 1.4603,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8157389635316699,
+      "grad_norm": 5.384745932649186,
+      "learning_rate": 2.000068813913777e-06,
+      "loss": 1.4642,
+      "step": 1275
+    },
+    {
+      "epoch": 0.818937939859245,
+      "grad_norm": 5.296954782854841,
+      "learning_rate": 1.933535876347141e-06,
+      "loss": 1.4828,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8221369161868202,
+      "grad_norm": 5.346333006321434,
+      "learning_rate": 1.8680097494332682e-06,
+      "loss": 1.4567,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8253358925143954,
+      "grad_norm": 4.8329274441724515,
+      "learning_rate": 1.8034986117745245e-06,
+      "loss": 1.4222,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8285348688419706,
+      "grad_norm": 5.928474522129584,
+      "learning_rate": 1.7400105152880331e-06,
+      "loss": 1.4337,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8317338451695457,
+      "grad_norm": 4.869570471894802,
+      "learning_rate": 1.6775533842006853e-06,
+      "loss": 1.5311,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8349328214971209,
+      "grad_norm": 4.8493003716399725,
+      "learning_rate": 1.6161350140600606e-06,
+      "loss": 1.4303,
+      "step": 1305
+    },
+    {
+      "epoch": 0.838131797824696,
+      "grad_norm": 5.384712389114864,
+      "learning_rate": 1.555763070761448e-06,
+      "loss": 1.5058,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8413307741522713,
+      "grad_norm": 4.989379576139708,
+      "learning_rate": 1.496445089591031e-06,
+      "loss": 1.479,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8445297504798465,
+      "grad_norm": 6.107526846697493,
+      "learning_rate": 1.4381884742853723e-06,
+      "loss": 1.4689,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8477287268074216,
+      "grad_norm": 5.898255686483617,
+      "learning_rate": 1.381000496107313e-06,
+      "loss": 1.4327,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8509277031349968,
+      "grad_norm": 5.451329007972122,
+      "learning_rate": 1.3248882929384321e-06,
+      "loss": 1.4044,
+      "step": 1330
+    },
+    {
+      "epoch": 0.8541266794625719,
+      "grad_norm": 5.072053387825864,
+      "learning_rate": 1.2698588683881185e-06,
+      "loss": 1.4606,
+      "step": 1335
+    },
+    {
+      "epoch": 0.8573256557901472,
+      "grad_norm": 5.5982332344576475,
+      "learning_rate": 1.215919090919434e-06,
+      "loss": 1.5089,
+      "step": 1340
+    },
+    {
+      "epoch": 0.8605246321177223,
+      "grad_norm": 5.85576698202184,
+      "learning_rate": 1.1630756929918218e-06,
+      "loss": 1.321,
+      "step": 1345
+    },
+    {
+      "epoch": 0.8637236084452975,
+      "grad_norm": 5.519776724650806,
+      "learning_rate": 1.111335270220798e-06,
+      "loss": 1.3995,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8669225847728727,
+      "grad_norm": 5.155693170774617,
+      "learning_rate": 1.060704280554733e-06,
+      "loss": 1.4438,
+      "step": 1355
+    },
+    {
+      "epoch": 0.8701215611004478,
+      "grad_norm": 4.942994004675328,
+      "learning_rate": 1.0111890434688009e-06,
+      "loss": 1.4739,
+      "step": 1360
+    },
+    {
+      "epoch": 0.8733205374280231,
+      "grad_norm": 5.921993476086954,
+      "learning_rate": 9.62795739176212e-07,
+      "loss": 1.5389,
+      "step": 1365
+    },
+    {
+      "epoch": 0.8765195137555982,
+      "grad_norm": 5.761105437707553,
+      "learning_rate": 9.155304078568495e-07,
+      "loss": 1.3901,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8797184900831734,
+      "grad_norm": 5.479509419610506,
+      "learning_rate": 8.693989489033494e-07,
+      "loss": 1.3581,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8829174664107485,
+      "grad_norm": 5.216115765818895,
+      "learning_rate": 8.244071201847826e-07,
+      "loss": 1.401,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8861164427383237,
+      "grad_norm": 5.2541897927690355,
+      "learning_rate": 7.805605373279934e-07,
+      "loss": 1.434,
+      "step": 1385
+    },
+    {
+      "epoch": 0.889315419065899,
+      "grad_norm": 5.125551484652663,
+      "learning_rate": 7.378646730166783e-07,
+      "loss": 1.4289,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8925143953934741,
+      "grad_norm": 5.717085861671814,
+      "learning_rate": 6.963248563083203e-07,
+      "loss": 1.4579,
+      "step": 1395
+    },
+    {
+      "epoch": 0.8957133717210493,
+      "grad_norm": 4.905074163080772,
+      "learning_rate": 6.559462719690501e-07,
+      "loss": 1.4323,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8989123480486244,
+      "grad_norm": 5.316058579983249,
+      "learning_rate": 6.167339598265109e-07,
+      "loss": 1.4301,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9021113243761996,
+      "grad_norm": 6.134109411578126,
+      "learning_rate": 5.78692814140811e-07,
+      "loss": 1.3989,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9053103007037748,
+      "grad_norm": 5.153844232998973,
+      "learning_rate": 5.418275829936537e-07,
+      "loss": 1.4323,
+      "step": 1415
+    },
+    {
+      "epoch": 0.90850927703135,
+      "grad_norm": 6.057541224818081,
+      "learning_rate": 5.06142867695708e-07,
+      "loss": 1.4211,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9117082533589251,
+      "grad_norm": 5.404909620313045,
+      "learning_rate": 4.716431222122997e-07,
+      "loss": 1.449,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9149072296865003,
+      "grad_norm": 5.461261222086095,
+      "learning_rate": 4.3833265260749157e-07,
+      "loss": 1.4474,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9181062060140754,
+      "grad_norm": 5.57698870599459,
+      "learning_rate": 4.062156165066211e-07,
+      "loss": 1.4625,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9213051823416507,
+      "grad_norm": 5.643478354054604,
+      "learning_rate": 3.752960225773772e-07,
+      "loss": 1.4158,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9245041586692259,
+      "grad_norm": 5.332205735896073,
+      "learning_rate": 3.4557773002945607e-07,
+      "loss": 1.4054,
+      "step": 1445
+    },
+    {
+      "epoch": 0.927703134996801,
+      "grad_norm": 5.810526998669132,
+      "learning_rate": 3.170644481328711e-07,
+      "loss": 1.4248,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9309021113243762,
+      "grad_norm": 5.762100370860069,
+      "learning_rate": 2.8975973575499526e-07,
+      "loss": 1.433,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9341010876519513,
+      "grad_norm": 5.13125458178204,
+      "learning_rate": 2.636670009163522e-07,
+      "loss": 1.4596,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9373000639795266,
+      "grad_norm": 5.032521993257355,
+      "learning_rate": 2.3878950036524963e-07,
+      "loss": 1.352,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9404990403071017,
+      "grad_norm": 5.8054521422829835,
+      "learning_rate": 2.1513033917129334e-07,
+      "loss": 1.43,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9436980166346769,
+      "grad_norm": 6.026904867928668,
+      "learning_rate": 1.9269247033782744e-07,
+      "loss": 1.4344,
+      "step": 1475
+    },
+    {
+      "epoch": 0.946896992962252,
+      "grad_norm": 5.500529775966834,
+      "learning_rate": 1.7147869443335463e-07,
+      "loss": 1.4304,
+      "step": 1480
+    },
+    {
+      "epoch": 0.9500959692898272,
+      "grad_norm": 5.635692434428651,
+      "learning_rate": 1.5149165924199016e-07,
+      "loss": 1.4132,
+      "step": 1485
+    },
+    {
+      "epoch": 0.9532949456174025,
+      "grad_norm": 5.171100680514175,
+      "learning_rate": 1.3273385943297746e-07,
+      "loss": 1.4002,
+      "step": 1490
+    },
+    {
+      "epoch": 0.9564939219449776,
+      "grad_norm": 5.622183510518974,
+      "learning_rate": 1.1520763624931597e-07,
+      "loss": 1.5742,
+      "step": 1495
+    },
+    {
+      "epoch": 0.9596928982725528,
+      "grad_norm": 6.7294220401819524,
+      "learning_rate": 9.891517721554499e-08,
+      "loss": 1.3935,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9628918746001279,
+      "grad_norm": 5.532321436376589,
+      "learning_rate": 8.385851586470318e-08,
+      "loss": 1.3882,
+      "step": 1505
+    },
+    {
+      "epoch": 0.9660908509277031,
+      "grad_norm": 5.525134555665201,
+      "learning_rate": 7.003953148452036e-08,
+      "loss": 1.4797,
+      "step": 1510
+    },
+    {
+      "epoch": 0.9692898272552783,
+      "grad_norm": 5.565575991373979,
+      "learning_rate": 5.745994888285311e-08,
+      "loss": 1.4501,
+      "step": 1515
+    },
+    {
+      "epoch": 0.9724888035828535,
+      "grad_norm": 5.186160389297291,
+      "learning_rate": 4.612133817239905e-08,
+      "loss": 1.4311,
+      "step": 1520
+    },
+    {
+      "epoch": 0.9756877799104287,
+      "grad_norm": 5.359988633375659,
+      "learning_rate": 3.602511457473479e-08,
+      "loss": 1.4544,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9788867562380038,
+      "grad_norm": 5.151877955390032,
+      "learning_rate": 2.7172538243666057e-08,
+      "loss": 1.3686,
+      "step": 1530
+    },
+    {
+      "epoch": 0.982085732565579,
+      "grad_norm": 5.408202048659504,
+      "learning_rate": 1.9564714107945804e-08,
+      "loss": 1.4818,
+      "step": 1535
+    },
+    {
+      "epoch": 0.9852847088931542,
+      "grad_norm": 5.180243679064934,
+      "learning_rate": 1.3202591733365577e-08,
+      "loss": 1.3879,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9884836852207294,
+      "grad_norm": 5.369814685801919,
+      "learning_rate": 8.086965204233688e-09,
+      "loss": 1.463,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9916826615483045,
+      "grad_norm": 5.715937943346943,
+      "learning_rate": 4.218473024261149e-09,
+      "loss": 1.4651,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9948816378758797,
+      "grad_norm": 5.34032956356415,
+      "learning_rate": 1.5975980368709843e-09,
+      "loss": 1.3843,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9980806142034548,
+      "grad_norm": 5.575239367515022,
+      "learning_rate": 2.2466736492643416e-10,
+      "loss": 1.3717,
+      "step": 1560
+    },
     {
       "epoch": 1.0,
-      "eval_loss": 1.3552618026733398,
-      "eval_runtime": 57.2066,
-      "eval_samples_per_second": 23.039,
-      "eval_steps_per_second": 0.734,
-      "step": 247
+      "eval_loss": 1.4683383703231812,
+      "eval_runtime": 18.5499,
+      "eval_samples_per_second": 26.954,
+      "eval_steps_per_second": 0.863,
+      "step": 1563
     },
     {
       "epoch": 1.0,
-      "step": 247,
-      "total_flos": 12929193738240.0,
-      "train_loss": 1.46957281581786,
-      "train_runtime": 1664.2063,
-      "train_samples_per_second": 4.746,
-      "train_steps_per_second": 0.148
+      "step": 1563,
+      "total_flos": 19395247472640.0,
+      "train_loss": 1.5797494724631234,
+      "train_runtime": 5954.1324,
+      "train_samples_per_second": 8.398,
+      "train_steps_per_second": 0.263
     }
   ],
   "logging_steps": 5,
-  "max_steps": 247,
+  "max_steps": 1563,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 500,
@@ -393,7 +2234,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 12929193738240.0,
+  "total_flos": 19395247472640.0,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null