Model save

all_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 1.0,
-    "total_flos": 11777605632000.0,
-    "train_loss": 1.5383918571472168,
-    "train_runtime": 1636.4794,
-    "train_samples": 50301,
-    "train_samples_per_second": 4.388,
-    "train_steps_per_second": 0.137
+    "total_flos": 17631303106560.0,
+    "train_loss": 1.5734304829583439,
+    "train_runtime": 5865.6403,
+    "train_samples": 50000,
+    "train_samples_per_second": 8.524,
+    "train_steps_per_second": 0.266
 }

runs/Dec19_00-39-24_gilbreth-j001.rcac.purdue.edu/events.out.tfevents.1734586892.gilbreth-j001.rcac.purdue.edu.92305.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:223681660f8b5aaf68d0bb309401b6594b099f299ee4fedb436f1df5f3dc90f2
-size 72314
+oid sha256:96f04f30ad5a9d1dfaa14f51300aa16424425a96fb06d9ca40d336322e4d6213
+size 72668

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 1.0,
-    "total_flos": 11777605632000.0,
-    "train_loss": 1.5383918571472168,
-    "train_runtime": 1636.4794,
-    "train_samples": 50301,
-    "train_samples_per_second": 4.388,
-    "train_steps_per_second": 0.137
+    "total_flos": 17631303106560.0,
+    "train_loss": 1.5734304829583439,
+    "train_runtime": 5865.6403,
+    "train_samples": 50000,
+    "train_samples_per_second": 8.524,
+    "train_steps_per_second": 0.266
 }

trainer_state.json

@@ -3,353 +3,2222 @@
   "best_model_checkpoint": null,
   "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 225,
+  "global_step": 1563,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0044444444444444444,
-      "grad_norm": 16.902174244065552,
-      "learning_rate": 8.695652173913044e-07,
-      "loss": 2.0963,
+      "epoch": 0.0006397952655150352,
+      "grad_norm": 7.394080615918296,
+      "learning_rate": 1.2738853503184715e-07,
+      "loss": 2.0305,
       "step": 1
     },
     {
-      "epoch": 0.022222222222222223,
-      "grad_norm": 7.009398687249402,
-      "learning_rate": 4.347826086956522e-06,
-      "loss": 2.0386,
+      "epoch": 0.003198976327575176,
+      "grad_norm": 6.988123754534515,
+      "learning_rate": 6.369426751592357e-07,
+      "loss": 1.9432,
       "step": 5
     },
     {
-      "epoch": 0.044444444444444446,
-      "grad_norm": 3.847165431793462,
-      "learning_rate": 8.695652173913044e-06,
-      "loss": 1.8407,
+      "epoch": 0.006397952655150352,
+      "grad_norm": 6.389634280271642,
+      "learning_rate": 1.2738853503184715e-06,
+      "loss": 2.003,
       "step": 10
     },
     {
-      "epoch": 0.06666666666666667,
-      "grad_norm": 1.8435387905194667,
-      "learning_rate": 1.3043478260869566e-05,
-      "loss": 1.7046,
+      "epoch": 0.009596928982725527,
+      "grad_norm": 4.815514441114148,
+      "learning_rate": 1.9108280254777074e-06,
+      "loss": 1.8623,
       "step": 15
     },
     {
-      "epoch": 0.08888888888888889,
-      "grad_norm": 1.5416577043323643,
-      "learning_rate": 1.739130434782609e-05,
-      "loss": 1.6393,
+      "epoch": 0.012795905310300703,
+      "grad_norm": 4.254380514471883,
+      "learning_rate": 2.547770700636943e-06,
+      "loss": 1.8562,
       "step": 20
     },
     {
-      "epoch": 0.1111111111111111,
-      "grad_norm": 1.4298947508822206,
-      "learning_rate": 1.999516282291988e-05,
-      "loss": 1.6362,
+      "epoch": 0.01599488163787588,
+      "grad_norm": 3.992810395688516,
+      "learning_rate": 3.1847133757961785e-06,
+      "loss": 1.7801,
       "step": 25
     },
     {
-      "epoch": 0.13333333333333333,
-      "grad_norm": 1.3933473839401893,
-      "learning_rate": 1.9940798309400527e-05,
-      "loss": 1.6027,
+      "epoch": 0.019193857965451054,
+      "grad_norm": 3.7657804821863046,
+      "learning_rate": 3.821656050955415e-06,
+      "loss": 1.7563,
       "step": 30
     },
     {
-      "epoch": 0.15555555555555556,
-      "grad_norm": 1.518431185821292,
-      "learning_rate": 1.982635248222264e-05,
-      "loss": 1.6116,
+      "epoch": 0.022392834293026232,
+      "grad_norm": 3.5537232539203414,
+      "learning_rate": 4.45859872611465e-06,
+      "loss": 1.6801,
       "step": 35
     },
     {
-      "epoch": 0.17777777777777778,
-      "grad_norm": 1.4964220141465232,
-      "learning_rate": 1.9652517041934357e-05,
-      "loss": 1.6207,
+      "epoch": 0.025591810620601407,
+      "grad_norm": 3.22896206640905,
+      "learning_rate": 5.095541401273886e-06,
+      "loss": 1.6648,
       "step": 40
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 1.3273747890888088,
-      "learning_rate": 1.9420342634699893e-05,
-      "loss": 1.5851,
+      "epoch": 0.028790786948176585,
+      "grad_norm": 3.403433611826566,
+      "learning_rate": 5.732484076433121e-06,
+      "loss": 1.6292,
       "step": 45
     },
     {
-      "epoch": 0.2222222222222222,
-      "grad_norm": 1.3559947683219293,
-      "learning_rate": 1.913123250228619e-05,
-      "loss": 1.5698,
+      "epoch": 0.03198976327575176,
+      "grad_norm": 3.4679951074737265,
+      "learning_rate": 6.369426751592357e-06,
+      "loss": 1.6593,
       "step": 50
     },
     {
-      "epoch": 0.24444444444444444,
-      "grad_norm": 1.4346793931071362,
-      "learning_rate": 1.878693400099269e-05,
-      "loss": 1.5886,
+      "epoch": 0.035188739603326934,
+      "grad_norm": 3.156456048051111,
+      "learning_rate": 7.006369426751593e-06,
+      "loss": 1.6613,
       "step": 55
     },
     {
-      "epoch": 0.26666666666666666,
-      "grad_norm": 1.4956510210420204,
-      "learning_rate": 1.8389528040783014e-05,
-      "loss": 1.5665,
+      "epoch": 0.03838771593090211,
+      "grad_norm": 3.086307127274183,
+      "learning_rate": 7.64331210191083e-06,
+      "loss": 1.6561,
       "step": 60
     },
     {
-      "epoch": 0.28888888888888886,
-      "grad_norm": 1.2995167467095625,
-      "learning_rate": 1.7941416508447537e-05,
-      "loss": 1.5525,
+      "epoch": 0.04158669225847729,
+      "grad_norm": 3.172716776400222,
+      "learning_rate": 8.280254777070064e-06,
+      "loss": 1.6567,
       "step": 65
     },
     {
-      "epoch": 0.3111111111111111,
-      "grad_norm": 1.3135664549473283,
-      "learning_rate": 1.7445307750810153e-05,
-      "loss": 1.5339,
+      "epoch": 0.044785668586052464,
+      "grad_norm": 3.302196355125012,
+      "learning_rate": 8.9171974522293e-06,
+      "loss": 1.636,
       "step": 70
     },
     {
-      "epoch": 0.3333333333333333,
-      "grad_norm": 1.2301844065790093,
-      "learning_rate": 1.690420020571747e-05,
-      "loss": 1.5359,
+      "epoch": 0.04798464491362764,
+      "grad_norm": 3.4368418751057446,
+      "learning_rate": 9.554140127388536e-06,
+      "loss": 1.6074,
       "step": 75
     },
     {
-      "epoch": 0.35555555555555557,
-      "grad_norm": 1.241596243663996,
-      "learning_rate": 1.6321364279743267e-05,
-      "loss": 1.5251,
+      "epoch": 0.05118362124120281,
+      "grad_norm": 3.1917269735549625,
+      "learning_rate": 1.0191082802547772e-05,
+      "loss": 1.6616,
       "step": 80
     },
     {
-      "epoch": 0.37777777777777777,
-      "grad_norm": 1.3171185796153217,
-      "learning_rate": 1.570032258213783e-05,
-      "loss": 1.5166,
+      "epoch": 0.05438259756877799,
+      "grad_norm": 3.3917094379269206,
+      "learning_rate": 1.0828025477707008e-05,
+      "loss": 1.6223,
       "step": 85
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 1.2799777894642266,
-      "learning_rate": 1.50448286344864e-05,
-      "loss": 1.5316,
+      "epoch": 0.05758157389635317,
+      "grad_norm": 3.2284167705131375,
+      "learning_rate": 1.1464968152866242e-05,
+      "loss": 1.6181,
       "step": 90
     },
     {
-      "epoch": 0.4222222222222222,
-      "grad_norm": 1.2503649653962412,
-      "learning_rate": 1.4358844184753713e-05,
-      "loss": 1.5384,
+      "epoch": 0.060780550223928344,
+      "grad_norm": 3.1149786469782645,
+      "learning_rate": 1.2101910828025478e-05,
+      "loss": 1.6075,
       "step": 95
     },
     {
-      "epoch": 0.4444444444444444,
-      "grad_norm": 1.2234087623622982,
-      "learning_rate": 1.3646515262826551e-05,
-      "loss": 1.5068,
+      "epoch": 0.06397952655150352,
+      "grad_norm": 3.2532995287482813,
+      "learning_rate": 1.2738853503184714e-05,
+      "loss": 1.6283,
       "step": 100
     },
     {
-      "epoch": 0.4666666666666667,
-      "grad_norm": 1.2352190020575813,
-      "learning_rate": 1.2912147122272523e-05,
-      "loss": 1.5157,
+      "epoch": 0.0671785028790787,
+      "grad_norm": 3.4021186832363957,
+      "learning_rate": 1.337579617834395e-05,
+      "loss": 1.6266,
       "step": 105
     },
     {
-      "epoch": 0.4888888888888889,
-      "grad_norm": 1.2137578254278512,
-      "learning_rate": 1.2160178219764838e-05,
-      "loss": 1.4907,
+      "epoch": 0.07037747920665387,
+      "grad_norm": 3.49083558012156,
+      "learning_rate": 1.4012738853503186e-05,
+      "loss": 1.613,
       "step": 110
     },
     {
-      "epoch": 0.5111111111111111,
-      "grad_norm": 1.2333755755507012,
-      "learning_rate": 1.1395153389439232e-05,
-      "loss": 1.4864,
+      "epoch": 0.07357645553422905,
+      "grad_norm": 3.197274845693601,
+      "learning_rate": 1.464968152866242e-05,
+      "loss": 1.6213,
       "step": 115
     },
     {
-      "epoch": 0.5333333333333333,
-      "grad_norm": 1.239317098102939,
-      "learning_rate": 1.0621696374314807e-05,
-      "loss": 1.4872,
+      "epoch": 0.07677543186180422,
+      "grad_norm": 3.15979564252741,
+      "learning_rate": 1.528662420382166e-05,
+      "loss": 1.6228,
       "step": 120
     },
     {
-      "epoch": 0.5555555555555556,
-      "grad_norm": 1.1935616120101977,
-      "learning_rate": 9.844481880796492e-06,
-      "loss": 1.5288,
+      "epoch": 0.0799744081893794,
+      "grad_norm": 3.308420733595098,
+      "learning_rate": 1.5923566878980894e-05,
+      "loss": 1.6463,
       "step": 125
     },
     {
-      "epoch": 0.5777777777777777,
-      "grad_norm": 1.167038478793757,
-      "learning_rate": 9.068207325159285e-06,
-      "loss": 1.4918,
+      "epoch": 0.08317338451695458,
+      "grad_norm": 3.1632407335742525,
+      "learning_rate": 1.6560509554140128e-05,
+      "loss": 1.6282,
       "step": 130
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 1.2606116528575086,
-      "learning_rate": 8.297564442776014e-06,
-      "loss": 1.4989,
+      "epoch": 0.08637236084452975,
+      "grad_norm": 3.1631210507580163,
+      "learning_rate": 1.7197452229299365e-05,
+      "loss": 1.6365,
       "step": 135
     },
     {
-      "epoch": 0.6222222222222222,
-      "grad_norm": 1.2053439169353648,
-      "learning_rate": 7.537210931679988e-06,
-      "loss": 1.499,
+      "epoch": 0.08957133717210493,
+      "grad_norm": 3.189101122277436,
+      "learning_rate": 1.78343949044586e-05,
+      "loss": 1.6921,
       "step": 140
     },
     {
-      "epoch": 0.6444444444444445,
-      "grad_norm": 1.2078865568862998,
-      "learning_rate": 6.791742301846325e-06,
-      "loss": 1.5059,
+      "epoch": 0.0927703134996801,
+      "grad_norm": 3.3936802906197086,
+      "learning_rate": 1.8471337579617837e-05,
+      "loss": 1.6784,
       "step": 145
     },
     {
-      "epoch": 0.6666666666666666,
-      "grad_norm": 1.181773149819734,
-      "learning_rate": 6.065664100332478e-06,
-      "loss": 1.482,
+      "epoch": 0.09596928982725528,
+      "grad_norm": 3.1547164505418053,
+      "learning_rate": 1.910828025477707e-05,
+      "loss": 1.6326,
       "step": 150
     },
     {
-      "epoch": 0.6888888888888889,
-      "grad_norm": 1.1586295971120542,
-      "learning_rate": 5.3633646801467255e-06,
-      "loss": 1.4776,
+      "epoch": 0.09916826615483046,
+      "grad_norm": 3.1142443602043923,
+      "learning_rate": 1.9745222929936306e-05,
+      "loss": 1.6999,
       "step": 155
     },
     {
-      "epoch": 0.7111111111111111,
-      "grad_norm": 1.1426794972054166,
-      "learning_rate": 4.6890886774272485e-06,
-      "loss": 1.4794,
+      "epoch": 0.10236724248240563,
+      "grad_norm": 3.0457646937300837,
+      "learning_rate": 1.9999775332635076e-05,
+      "loss": 1.6366,
       "step": 160
     },
     {
-      "epoch": 0.7333333333333333,
-      "grad_norm": 1.2053169655060443,
-      "learning_rate": 4.046911357233343e-06,
-      "loss": 1.4616,
+      "epoch": 0.10556621880998081,
+      "grad_norm": 3.0372482724825525,
+      "learning_rate": 1.999840240196313e-05,
+      "loss": 1.6012,
       "step": 165
     },
     {
-      "epoch": 0.7555555555555555,
-      "grad_norm": 1.1647726703293662,
-      "learning_rate": 3.440713983000601e-06,
-      "loss": 1.4921,
+      "epoch": 0.10876519513755598,
+      "grad_norm": 3.211113164573299,
+      "learning_rate": 1.9995781526975738e-05,
+      "loss": 1.6572,
       "step": 170
     },
     {
-      "epoch": 0.7777777777777778,
-      "grad_norm": 1.1515379297439883,
-      "learning_rate": 2.8741603585249312e-06,
-      "loss": 1.4614,
+      "epoch": 0.11196417146513116,
+      "grad_norm": 2.8705010535094875,
+      "learning_rate": 1.9991913034795768e-05,
+      "loss": 1.6988,
       "step": 175
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 1.1677038359307508,
-      "learning_rate": 2.3506746842535244e-06,
-      "loss": 1.4821,
+      "epoch": 0.11516314779270634,
+      "grad_norm": 3.075350854034086,
+      "learning_rate": 1.9986797408266636e-05,
+      "loss": 1.6673,
       "step": 180
     },
     {
-      "epoch": 0.8222222222222222,
-      "grad_norm": 1.157505348471252,
-      "learning_rate": 1.8734208617174986e-06,
-      "loss": 1.4643,
+      "epoch": 0.1183621241202815,
+      "grad_norm": 2.8930449752507803,
+      "learning_rate": 1.9980435285892056e-05,
+      "loss": 1.7404,
       "step": 185
     },
     {
-      "epoch": 0.8444444444444444,
-      "grad_norm": 1.1694689724300111,
-      "learning_rate": 1.4452833711883629e-06,
-      "loss": 1.4734,
+      "epoch": 0.12156110044785669,
+      "grad_norm": 2.997023488149924,
+      "learning_rate": 1.9972827461756335e-05,
+      "loss": 1.6516,
       "step": 190
     },
     {
-      "epoch": 0.8666666666666667,
-      "grad_norm": 1.1512148511863012,
-      "learning_rate": 1.0688498381320855e-06,
-      "loss": 1.4435,
+      "epoch": 0.12476007677543186,
+      "grad_norm": 3.0525943500136594,
+      "learning_rate": 1.9963974885425267e-05,
+      "loss": 1.6142,
       "step": 195
     },
     {
-      "epoch": 0.8888888888888888,
-      "grad_norm": 1.1323651049667787,
-      "learning_rate": 7.463953938275859e-07,
-      "loss": 1.447,
+      "epoch": 0.12795905310300704,
+      "grad_norm": 3.032903483565325,
+      "learning_rate": 1.9953878661827603e-05,
+      "loss": 1.6857,
       "step": 200
     },
     {
-      "epoch": 0.9111111111111111,
-      "grad_norm": 1.1423892630102332,
-      "learning_rate": 4.798689246727006e-07,
-      "loss": 1.4368,
+      "epoch": 0.13115802943058222,
+      "grad_norm": 2.8448154468271416,
+      "learning_rate": 1.994254005111715e-05,
+      "loss": 1.6578,
       "step": 205
     },
     {
-      "epoch": 0.9333333333333333,
-      "grad_norm": 1.1409480033032104,
-      "learning_rate": 2.708812932856253e-07,
-      "loss": 1.4754,
+      "epoch": 0.1343570057581574,
+      "grad_norm": 2.9094110750911395,
+      "learning_rate": 1.992996046851548e-05,
+      "loss": 1.6407,
       "step": 210
     },
     {
-      "epoch": 0.9555555555555556,
-      "grad_norm": 1.1449815403515953,
-      "learning_rate": 1.206956025924333e-07,
-      "loss": 1.473,
+      "epoch": 0.13755598208573255,
+      "grad_norm": 2.9037930260697182,
+      "learning_rate": 1.9916141484135297e-05,
+      "loss": 1.6858,
       "step": 215
     },
     {
-      "epoch": 0.9777777777777777,
-      "grad_norm": 1.1871235983216775,
-      "learning_rate": 3.0219561743707326e-08,
-      "loss": 1.4529,
+      "epoch": 0.14075495841330773,
+      "grad_norm": 2.8984434910658834,
+      "learning_rate": 1.990108482278446e-05,
+      "loss": 1.6633,
       "step": 220
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 1.1249933194182675,
-      "learning_rate": 0.0,
-      "loss": 1.4638,
+      "epoch": 0.14395393474088292,
+      "grad_norm": 2.7339743533257255,
+      "learning_rate": 1.9884792363750684e-05,
+      "loss": 1.7033,
       "step": 225
     },
     {
-      "epoch": 1.0,
-      "eval_loss": 1.4236433506011963,
-      "eval_runtime": 54.2532,
-      "eval_samples_per_second": 22.063,
-      "eval_steps_per_second": 0.7,
-      "step": 225
+      "epoch": 0.1471529110684581,
+      "grad_norm": 2.945202398619138,
+      "learning_rate": 1.9867266140567024e-05,
+      "loss": 1.6713,
+      "step": 230
+    },
+    {
+      "epoch": 0.15035188739603328,
+      "grad_norm": 2.7793460261472984,
+      "learning_rate": 1.9848508340758014e-05,
+      "loss": 1.694,
+      "step": 235
+    },
+    {
+      "epoch": 0.15355086372360843,
+      "grad_norm": 2.803292273662775,
+      "learning_rate": 1.9828521305566647e-05,
+      "loss": 1.6802,
+      "step": 240
+    },
+    {
+      "epoch": 0.15674984005118361,
+      "grad_norm": 2.7907576962985816,
+      "learning_rate": 1.9807307529662175e-05,
+      "loss": 1.7578,
+      "step": 245
+    },
+    {
+      "epoch": 0.1599488163787588,
+      "grad_norm": 2.905959935226371,
+      "learning_rate": 1.9784869660828708e-05,
+      "loss": 1.6559,
+      "step": 250
+    },
+    {
+      "epoch": 0.16314779270633398,
+      "grad_norm": 3.000333928564516,
+      "learning_rate": 1.9761210499634754e-05,
+      "loss": 1.6492,
+      "step": 255
+    },
+    {
+      "epoch": 0.16634676903390916,
+      "grad_norm": 2.8638985431140584,
+      "learning_rate": 1.973633299908365e-05,
+      "loss": 1.6195,
+      "step": 260
+    },
+    {
+      "epoch": 0.1695457453614843,
+      "grad_norm": 3.1338815620020557,
+      "learning_rate": 1.9710240264245005e-05,
+      "loss": 1.7182,
+      "step": 265
+    },
+    {
+      "epoch": 0.1727447216890595,
+      "grad_norm": 2.735922550084266,
+      "learning_rate": 1.9682935551867132e-05,
+      "loss": 1.6404,
+      "step": 270
+    },
+    {
+      "epoch": 0.17594369801663468,
+      "grad_norm": 2.7772426797312852,
+      "learning_rate": 1.9654422269970545e-05,
+      "loss": 1.6768,
+      "step": 275
+    },
+    {
+      "epoch": 0.17914267434420986,
+      "grad_norm": 2.9636810758877568,
+      "learning_rate": 1.9624703977422624e-05,
+      "loss": 1.6454,
+      "step": 280
+    },
+    {
+      "epoch": 0.18234165067178504,
+      "grad_norm": 2.7840158936469974,
+      "learning_rate": 1.959378438349338e-05,
+      "loss": 1.6454,
+      "step": 285
+    },
+    {
+      "epoch": 0.1855406269993602,
+      "grad_norm": 2.8310598910760163,
+      "learning_rate": 1.956166734739251e-05,
+      "loss": 1.6648,
+      "step": 290
+    },
+    {
+      "epoch": 0.18873960332693537,
+      "grad_norm": 2.714786142604982,
+      "learning_rate": 1.95283568777877e-05,
+      "loss": 1.6666,
+      "step": 295
+    },
+    {
+      "epoch": 0.19193857965451055,
+      "grad_norm": 2.9756653362793504,
+      "learning_rate": 1.9493857132304295e-05,
+      "loss": 1.6332,
+      "step": 300
+    },
+    {
+      "epoch": 0.19513755598208574,
+      "grad_norm": 2.663479791243424,
+      "learning_rate": 1.9458172417006347e-05,
+      "loss": 1.7517,
+      "step": 305
+    },
+    {
+      "epoch": 0.19833653230966092,
+      "grad_norm": 2.787541324182297,
+      "learning_rate": 1.942130718585919e-05,
+      "loss": 1.6247,
+      "step": 310
+    },
+    {
+      "epoch": 0.20153550863723607,
+      "grad_norm": 2.9978313664286005,
+      "learning_rate": 1.938326604017349e-05,
+      "loss": 1.6562,
+      "step": 315
+    },
+    {
+      "epoch": 0.20473448496481125,
+      "grad_norm": 2.8618671523467274,
+      "learning_rate": 1.9344053728030952e-05,
+      "loss": 1.6709,
+      "step": 320
+    },
+    {
+      "epoch": 0.20793346129238643,
+      "grad_norm": 2.691578696625765,
+      "learning_rate": 1.9303675143691683e-05,
+      "loss": 1.6802,
+      "step": 325
+    },
+    {
+      "epoch": 0.21113243761996162,
+      "grad_norm": 2.8907063332178002,
+      "learning_rate": 1.9262135326983326e-05,
+      "loss": 1.7082,
+      "step": 330
+    },
+    {
+      "epoch": 0.2143314139475368,
+      "grad_norm": 2.9725776193939706,
+      "learning_rate": 1.921943946267201e-05,
+      "loss": 1.626,
+      "step": 335
+    },
+    {
+      "epoch": 0.21753039027511195,
+      "grad_norm": 2.8615891525639974,
+      "learning_rate": 1.9175592879815217e-05,
+      "loss": 1.6959,
+      "step": 340
+    },
+    {
+      "epoch": 0.22072936660268713,
+      "grad_norm": 2.8919731781020857,
+      "learning_rate": 1.9130601051096655e-05,
+      "loss": 1.6268,
+      "step": 345
+    },
+    {
+      "epoch": 0.22392834293026231,
+      "grad_norm": 2.9924454366979343,
+      "learning_rate": 1.9084469592143154e-05,
+      "loss": 1.6537,
+      "step": 350
+    },
+    {
+      "epoch": 0.2271273192578375,
+      "grad_norm": 2.781125301226174,
+      "learning_rate": 1.9037204260823788e-05,
+      "loss": 1.6043,
+      "step": 355
+    },
+    {
+      "epoch": 0.23032629558541268,
+      "grad_norm": 2.8421749381590007,
+      "learning_rate": 1.89888109565312e-05,
+      "loss": 1.7348,
+      "step": 360
+    },
+    {
+      "epoch": 0.23352527191298783,
+      "grad_norm": 2.7140292893955418,
+      "learning_rate": 1.893929571944527e-05,
+      "loss": 1.6701,
+      "step": 365
+    },
+    {
+      "epoch": 0.236724248240563,
+      "grad_norm": 2.946356544423079,
+      "learning_rate": 1.8888664729779205e-05,
+      "loss": 1.6862,
+      "step": 370
+    },
+    {
+      "epoch": 0.2399232245681382,
+      "grad_norm": 2.839893452790028,
+      "learning_rate": 1.883692430700818e-05,
+      "loss": 1.6464,
+      "step": 375
+    },
+    {
+      "epoch": 0.24312220089571338,
+      "grad_norm": 3.0052515914390736,
+      "learning_rate": 1.8784080909080568e-05,
+      "loss": 1.6302,
+      "step": 380
+    },
+    {
+      "epoch": 0.24632117722328856,
+      "grad_norm": 2.6820394128865805,
+      "learning_rate": 1.8730141131611882e-05,
+      "loss": 1.6555,
+      "step": 385
+    },
+    {
+      "epoch": 0.2495201535508637,
+      "grad_norm": 2.634143252481978,
+      "learning_rate": 1.867511170706157e-05,
+      "loss": 1.6004,
+      "step": 390
+    },
+    {
+      "epoch": 0.2527191298784389,
+      "grad_norm": 2.7745186332846634,
+      "learning_rate": 1.861899950389269e-05,
+      "loss": 1.6785,
+      "step": 395
+    },
+    {
+      "epoch": 0.2559181062060141,
+      "grad_norm": 2.7324330094342284,
+      "learning_rate": 1.856181152571463e-05,
+      "loss": 1.6163,
+      "step": 400
+    },
+    {
+      "epoch": 0.2591170825335892,
+      "grad_norm": 2.9535285722755775,
+      "learning_rate": 1.850355491040897e-05,
+      "loss": 1.6173,
+      "step": 405
+    },
+    {
+      "epoch": 0.26231605886116444,
+      "grad_norm": 2.9149621130030536,
+      "learning_rate": 1.8444236929238556e-05,
+      "loss": 1.6581,
+      "step": 410
+    },
+    {
+      "epoch": 0.2655150351887396,
+      "grad_norm": 2.6067627751049094,
+      "learning_rate": 1.8383864985939944e-05,
+      "loss": 1.6164,
+      "step": 415
+    },
+    {
+      "epoch": 0.2687140115163148,
+      "grad_norm": 2.7496620680313786,
+      "learning_rate": 1.8322446615799317e-05,
+      "loss": 1.617,
+      "step": 420
+    },
+    {
+      "epoch": 0.27191298784388995,
+      "grad_norm": 2.7116588879744805,
+      "learning_rate": 1.825998948471197e-05,
+      "loss": 1.5945,
+      "step": 425
+    },
+    {
+      "epoch": 0.2751119641714651,
+      "grad_norm": 2.6386390309428003,
+      "learning_rate": 1.819650138822548e-05,
+      "loss": 1.6485,
+      "step": 430
+    },
+    {
+      "epoch": 0.2783109404990403,
+      "grad_norm": 2.4403141850907986,
+      "learning_rate": 1.8131990250566733e-05,
+      "loss": 1.5296,
+      "step": 435
+    },
+    {
+      "epoch": 0.28150991682661547,
+      "grad_norm": 2.690663842410636,
+      "learning_rate": 1.8066464123652857e-05,
+      "loss": 1.6769,
+      "step": 440
+    },
+    {
+      "epoch": 0.2847088931541907,
+      "grad_norm": 2.4053093033366375,
+      "learning_rate": 1.7999931186086225e-05,
+      "loss": 1.616,
+      "step": 445
+    },
+    {
+      "epoch": 0.28790786948176583,
+      "grad_norm": 2.655658834658945,
+      "learning_rate": 1.793239974213364e-05,
+      "loss": 1.585,
+      "step": 450
+    },
+    {
+      "epoch": 0.291106845809341,
+      "grad_norm": 2.7334995955366566,
+      "learning_rate": 1.786387822068987e-05,
+      "loss": 1.6489,
+      "step": 455
+    },
+    {
+      "epoch": 0.2943058221369162,
+      "grad_norm": 2.7528589911283365,
+      "learning_rate": 1.7794375174225565e-05,
+      "loss": 1.6601,
+      "step": 460
+    },
+    {
+      "epoch": 0.29750479846449135,
+      "grad_norm": 2.5929884221027937,
+      "learning_rate": 1.7723899277719815e-05,
+      "loss": 1.6194,
+      "step": 465
+    },
+    {
+      "epoch": 0.30070377479206656,
+      "grad_norm": 2.67165006812173,
+      "learning_rate": 1.7652459327577377e-05,
+      "loss": 1.6281,
+      "step": 470
+    },
+    {
+      "epoch": 0.3039027511196417,
+      "grad_norm": 2.5803192411036164,
+      "learning_rate": 1.7580064240530746e-05,
+      "loss": 1.6054,
+      "step": 475
+    },
+    {
+      "epoch": 0.30710172744721687,
+      "grad_norm": 2.7629487887046844,
+      "learning_rate": 1.7506723052527243e-05,
+      "loss": 1.6157,
+      "step": 480
+    },
+    {
+      "epoch": 0.3103007037747921,
+      "grad_norm": 2.818119974179455,
+      "learning_rate": 1.7432444917601183e-05,
+      "loss": 1.629,
+      "step": 485
+    },
+    {
+      "epoch": 0.31349968010236723,
+      "grad_norm": 2.6002526488275213,
+      "learning_rate": 1.735723910673132e-05,
+      "loss": 1.6047,
+      "step": 490
+    },
+    {
+      "epoch": 0.31669865642994244,
+      "grad_norm": 2.637737666270034,
+      "learning_rate": 1.7281115006683687e-05,
+      "loss": 1.594,
+      "step": 495
+    },
+    {
+      "epoch": 0.3198976327575176,
+      "grad_norm": 2.450430698280842,
+      "learning_rate": 1.7204082118840035e-05,
+      "loss": 1.6339,
+      "step": 500
+    },
+    {
+      "epoch": 0.32309660908509275,
+      "grad_norm": 2.6454010985724237,
+      "learning_rate": 1.712615005801185e-05,
+      "loss": 1.6465,
+      "step": 505
+    },
+    {
+      "epoch": 0.32629558541266795,
+      "grad_norm": 2.632364572436156,
+      "learning_rate": 1.704732855124037e-05,
+      "loss": 1.668,
+      "step": 510
+    },
+    {
+      "epoch": 0.3294945617402431,
+      "grad_norm": 2.6939062066847734,
+      "learning_rate": 1.6967627436582445e-05,
+      "loss": 1.6473,
+      "step": 515
+    },
+    {
+      "epoch": 0.3326935380678183,
+      "grad_norm": 2.704537381781226,
+      "learning_rate": 1.6887056661882644e-05,
+      "loss": 1.6555,
+      "step": 520
+    },
+    {
+      "epoch": 0.33589251439539347,
+      "grad_norm": 2.673030043305989,
+      "learning_rate": 1.6805626283531592e-05,
+      "loss": 1.5817,
+      "step": 525
+    },
+    {
+      "epoch": 0.3390914907229686,
+      "grad_norm": 2.5627610542356263,
+      "learning_rate": 1.6723346465210815e-05,
+      "loss": 1.6425,
+      "step": 530
+    },
+    {
+      "epoch": 0.34229046705054383,
+      "grad_norm": 2.490955616710756,
+      "learning_rate": 1.6640227476624163e-05,
+      "loss": 1.652,
+      "step": 535
+    },
+    {
+      "epoch": 0.345489443378119,
+      "grad_norm": 2.7836678293691914,
+      "learning_rate": 1.655627969221598e-05,
+      "loss": 1.5775,
+      "step": 540
+    },
+    {
+      "epoch": 0.3486884197056942,
+      "grad_norm": 2.879150296148446,
+      "learning_rate": 1.6471513589876247e-05,
+      "loss": 1.6193,
+      "step": 545
+    },
+    {
+      "epoch": 0.35188739603326935,
+      "grad_norm": 2.544183815219073,
+      "learning_rate": 1.638593974963278e-05,
+      "loss": 1.6253,
+      "step": 550
+    },
+    {
+      "epoch": 0.3550863723608445,
+      "grad_norm": 2.5247797493256354,
+      "learning_rate": 1.6299568852330703e-05,
+      "loss": 1.6498,
+      "step": 555
+    },
+    {
+      "epoch": 0.3582853486884197,
+      "grad_norm": 2.641960224176829,
+      "learning_rate": 1.6212411678299306e-05,
+      "loss": 1.5938,
+      "step": 560
+    },
+    {
+      "epoch": 0.36148432501599487,
+      "grad_norm": 2.449768437546663,
+      "learning_rate": 1.612447910600652e-05,
+      "loss": 1.5764,
+      "step": 565
+    },
+    {
+      "epoch": 0.3646833013435701,
+      "grad_norm": 2.578583632586866,
+      "learning_rate": 1.6035782110701125e-05,
+      "loss": 1.6131,
+      "step": 570
+    },
+    {
+      "epoch": 0.36788227767114523,
+      "grad_norm": 2.480736343141984,
+      "learning_rate": 1.594633176304287e-05,
+      "loss": 1.6181,
+      "step": 575
+    },
+    {
+      "epoch": 0.3710812539987204,
+      "grad_norm": 2.652942682958262,
+      "learning_rate": 1.5856139227720714e-05,
+      "loss": 1.6279,
+      "step": 580
+    },
+    {
+      "epoch": 0.3742802303262956,
+      "grad_norm": 2.5484066228370748,
+      "learning_rate": 1.5765215762059304e-05,
+      "loss": 1.6044,
+      "step": 585
+    },
+    {
+      "epoch": 0.37747920665387075,
+      "grad_norm": 2.6484189766310617,
+      "learning_rate": 1.5673572714613886e-05,
+      "loss": 1.6565,
+      "step": 590
+    },
+    {
+      "epoch": 0.38067818298144596,
+      "grad_norm": 2.707940008356439,
+      "learning_rate": 1.558122152375387e-05,
+      "loss": 1.6113,
+      "step": 595
+    },
+    {
+      "epoch": 0.3838771593090211,
+      "grad_norm": 2.5070863505818606,
+      "learning_rate": 1.548817371623513e-05,
+      "loss": 1.5619,
+      "step": 600
+    },
+    {
+      "epoch": 0.38707613563659626,
+      "grad_norm": 2.3650716848486333,
+      "learning_rate": 1.539444090576132e-05,
+      "loss": 1.6134,
+      "step": 605
+    },
+    {
+      "epoch": 0.3902751119641715,
+      "grad_norm": 2.6861687501614466,
+      "learning_rate": 1.5300034791534297e-05,
+      "loss": 1.5806,
+      "step": 610
+    },
+    {
+      "epoch": 0.3934740882917466,
+      "grad_norm": 2.4191496743332093,
+      "learning_rate": 1.520496715679391e-05,
+      "loss": 1.5797,
+      "step": 615
+    },
+    {
+      "epoch": 0.39667306461932184,
+      "grad_norm": 2.550326271295577,
+      "learning_rate": 1.5109249867347276e-05,
+      "loss": 1.5974,
+      "step": 620
+    },
+    {
+      "epoch": 0.399872040946897,
+      "grad_norm": 2.3102647791766864,
+      "learning_rate": 1.5012894870087751e-05,
+      "loss": 1.6202,
+      "step": 625
+    },
+    {
+      "epoch": 0.40307101727447214,
+      "grad_norm": 2.59762028303461,
+      "learning_rate": 1.4915914191503792e-05,
+      "loss": 1.6274,
+      "step": 630
+    },
+    {
+      "epoch": 0.40626999360204735,
+      "grad_norm": 2.881971956760912,
+      "learning_rate": 1.4818319936177885e-05,
+      "loss": 1.5946,
+      "step": 635
+    },
+    {
+      "epoch": 0.4094689699296225,
+      "grad_norm": 2.569538647562056,
+      "learning_rate": 1.4720124285275703e-05,
+      "loss": 1.5929,
+      "step": 640
+    },
+    {
+      "epoch": 0.4126679462571977,
+      "grad_norm": 2.5291866469053033,
+      "learning_rate": 1.4621339495025731e-05,
+      "loss": 1.5741,
+      "step": 645
+    },
+    {
+      "epoch": 0.41586692258477287,
+      "grad_norm": 2.688772146899729,
+      "learning_rate": 1.4521977895189518e-05,
+      "loss": 1.5687,
+      "step": 650
+    },
+    {
+      "epoch": 0.419065898912348,
+      "grad_norm": 2.8684978363355933,
+      "learning_rate": 1.4422051887522735e-05,
+      "loss": 1.6226,
+      "step": 655
+    },
+    {
+      "epoch": 0.42226487523992323,
+      "grad_norm": 2.3545531570138434,
+      "learning_rate": 1.4321573944227254e-05,
+      "loss": 1.5946,
+      "step": 660
+    },
+    {
+      "epoch": 0.4254638515674984,
+      "grad_norm": 2.509229712523995,
+      "learning_rate": 1.4220556606394465e-05,
+      "loss": 1.6031,
+      "step": 665
+    },
+    {
+      "epoch": 0.4286628278950736,
+      "grad_norm": 2.6522491425171713,
+      "learning_rate": 1.4119012482439929e-05,
+      "loss": 1.5893,
+      "step": 670
+    },
+    {
+      "epoch": 0.43186180422264875,
+      "grad_norm": 2.6103836152679714,
+      "learning_rate": 1.4016954246529697e-05,
+      "loss": 1.5309,
+      "step": 675
+    },
+    {
+      "epoch": 0.4350607805502239,
+      "grad_norm": 2.4597418020462323,
+      "learning_rate": 1.3914394636998374e-05,
+      "loss": 1.5641,
+      "step": 680
+    },
+    {
+      "epoch": 0.4382597568777991,
+      "grad_norm": 2.593897343213478,
+      "learning_rate": 1.3811346454759211e-05,
+      "loss": 1.6196,
+      "step": 685
+    },
+    {
+      "epoch": 0.44145873320537427,
+      "grad_norm": 2.6002108820265457,
+      "learning_rate": 1.3707822561706336e-05,
+      "loss": 1.6235,
+      "step": 690
+    },
+    {
+      "epoch": 0.4446577095329495,
+      "grad_norm": 3.0420801400215853,
+      "learning_rate": 1.3603835879109442e-05,
+      "loss": 1.5103,
+      "step": 695
+    },
+    {
+      "epoch": 0.44785668586052463,
+      "grad_norm": 2.466641868555013,
+      "learning_rate": 1.349939938600099e-05,
+      "loss": 1.5902,
+      "step": 700
+    },
+    {
+      "epoch": 0.4510556621880998,
+      "grad_norm": 2.571792638474329,
+      "learning_rate": 1.3394526117556277e-05,
+      "loss": 1.5681,
+      "step": 705
+    },
+    {
+      "epoch": 0.454254638515675,
+      "grad_norm": 2.550162054129212,
+      "learning_rate": 1.3289229163466421e-05,
+      "loss": 1.5775,
+      "step": 710
+    },
+    {
+      "epoch": 0.45745361484325014,
+      "grad_norm": 2.618009664290396,
+      "learning_rate": 1.3183521666304611e-05,
+      "loss": 1.5413,
+      "step": 715
+    },
+    {
+      "epoch": 0.46065259117082535,
+      "grad_norm": 2.733491223792463,
+      "learning_rate": 1.3077416819885707e-05,
+      "loss": 1.5837,
+      "step": 720
+    },
+    {
+      "epoch": 0.4638515674984005,
+      "grad_norm": 2.5097687704688925,
+      "learning_rate": 1.297092786761946e-05,
+      "loss": 1.611,
+      "step": 725
+    },
+    {
+      "epoch": 0.46705054382597566,
+      "grad_norm": 2.5184864174917982,
+      "learning_rate": 1.2864068100857565e-05,
+      "loss": 1.5697,
+      "step": 730
+    },
+    {
+      "epoch": 0.47024952015355087,
+      "grad_norm": 2.3816287608148494,
+      "learning_rate": 1.2756850857234686e-05,
+      "loss": 1.5068,
+      "step": 735
+    },
+    {
+      "epoch": 0.473448496481126,
+      "grad_norm": 2.5411269367566853,
+      "learning_rate": 1.2649289519003739e-05,
+      "loss": 1.5925,
+      "step": 740
+    },
+    {
+      "epoch": 0.47664747280870123,
+      "grad_norm": 2.6112309877103104,
+      "learning_rate": 1.2541397511365584e-05,
+      "loss": 1.5981,
+      "step": 745
+    },
+    {
+      "epoch": 0.4798464491362764,
+      "grad_norm": 2.5171345270617427,
+      "learning_rate": 1.2433188300793399e-05,
+      "loss": 1.5989,
+      "step": 750
+    },
+    {
+      "epoch": 0.48304542546385154,
+      "grad_norm": 2.5466621441825215,
+      "learning_rate": 1.2324675393351818e-05,
+      "loss": 1.6118,
+      "step": 755
+    },
+    {
+      "epoch": 0.48624440179142675,
+      "grad_norm": 2.4943933560857503,
+      "learning_rate": 1.221587233301123e-05,
+      "loss": 1.6128,
+      "step": 760
+    },
+    {
+      "epoch": 0.4894433781190019,
+      "grad_norm": 2.672340704509595,
+      "learning_rate": 1.2106792699957264e-05,
+      "loss": 1.5632,
+      "step": 765
+    },
+    {
+      "epoch": 0.4926423544465771,
+      "grad_norm": 2.4667369084948354,
+      "learning_rate": 1.1997450108895807e-05,
+      "loss": 1.5842,
+      "step": 770
+    },
+    {
+      "epoch": 0.49584133077415227,
+      "grad_norm": 2.511794674207891,
+      "learning_rate": 1.1887858207353678e-05,
+      "loss": 1.5659,
+      "step": 775
+    },
+    {
+      "epoch": 0.4990403071017274,
+      "grad_norm": 2.6412164194277743,
+      "learning_rate": 1.1778030673975227e-05,
+      "loss": 1.61,
+      "step": 780
+    },
+    {
+      "epoch": 0.5022392834293026,
+      "grad_norm": 2.478596228285701,
+      "learning_rate": 1.166798121681505e-05,
+      "loss": 1.522,
+      "step": 785
+    },
+    {
+      "epoch": 0.5054382597568778,
+      "grad_norm": 2.595939842592459,
+      "learning_rate": 1.1557723571627016e-05,
+      "loss": 1.5636,
+      "step": 790
+    },
+    {
+      "epoch": 0.508637236084453,
+      "grad_norm": 2.432326751699925,
+      "learning_rate": 1.1447271500149849e-05,
+      "loss": 1.5878,
+      "step": 795
+    },
+    {
+      "epoch": 0.5118362124120281,
+      "grad_norm": 2.6385262661777453,
+      "learning_rate": 1.1336638788389473e-05,
+      "loss": 1.5783,
+      "step": 800
+    },
+    {
+      "epoch": 0.5150351887396033,
+      "grad_norm": 2.555868661788137,
+      "learning_rate": 1.122583924489832e-05,
+      "loss": 1.5321,
+      "step": 805
+    },
+    {
+      "epoch": 0.5182341650671785,
+      "grad_norm": 2.4875755865322295,
+      "learning_rate": 1.1114886699051803e-05,
+      "loss": 1.5803,
+      "step": 810
+    },
+    {
+      "epoch": 0.5214331413947537,
+      "grad_norm": 2.5436639429408694,
+      "learning_rate": 1.1003794999322246e-05,
+      "loss": 1.5801,
+      "step": 815
+    },
+    {
+      "epoch": 0.5246321177223289,
+      "grad_norm": 2.5086288677269093,
+      "learning_rate": 1.089257801155037e-05,
+      "loss": 1.5719,
+      "step": 820
+    },
+    {
+      "epoch": 0.527831094049904,
+      "grad_norm": 2.53995010922951,
+      "learning_rate": 1.0781249617214642e-05,
+      "loss": 1.5719,
+      "step": 825
+    },
+    {
+      "epoch": 0.5310300703774792,
+      "grad_norm": 2.507596140428045,
+      "learning_rate": 1.0669823711698668e-05,
+      "loss": 1.5836,
+      "step": 830
+    },
+    {
+      "epoch": 0.5342290467050543,
+      "grad_norm": 2.5252340375115314,
+      "learning_rate": 1.0558314202556866e-05,
+      "loss": 1.5615,
+      "step": 835
+    },
+    {
+      "epoch": 0.5374280230326296,
+      "grad_norm": 2.595739313200728,
+      "learning_rate": 1.0446735007778563e-05,
+      "loss": 1.588,
+      "step": 840
+    },
+    {
+      "epoch": 0.5406269993602048,
+      "grad_norm": 2.5624942589229898,
+      "learning_rate": 1.0335100054050877e-05,
+      "loss": 1.5346,
+      "step": 845
+    },
+    {
+      "epoch": 0.5438259756877799,
+      "grad_norm": 2.46500986708702,
+      "learning_rate": 1.0223423275020431e-05,
+      "loss": 1.5449,
+      "step": 850
+    },
+    {
+      "epoch": 0.5470249520153551,
+      "grad_norm": 2.394167818299305,
+      "learning_rate": 1.0111718609554254e-05,
+      "loss": 1.5705,
+      "step": 855
+    },
+    {
+      "epoch": 0.5502239283429302,
+      "grad_norm": 2.6197177214921283,
+      "learning_rate": 1e-05,
+      "loss": 1.5355,
+      "step": 860
+    },
+    {
+      "epoch": 0.5534229046705055,
+      "grad_norm": 2.509456810430575,
+      "learning_rate": 9.888281390445747e-06,
+      "loss": 1.4882,
+      "step": 865
+    },
+    {
+      "epoch": 0.5566218809980806,
+      "grad_norm": 2.3500525614804,
+      "learning_rate": 9.776576724979572e-06,
+      "loss": 1.4693,
+      "step": 870
+    },
+    {
+      "epoch": 0.5598208573256558,
+      "grad_norm": 2.554313130115952,
+      "learning_rate": 9.664899945949128e-06,
+      "loss": 1.6321,
+      "step": 875
+    },
+    {
+      "epoch": 0.5630198336532309,
+      "grad_norm": 2.4292831573691287,
+      "learning_rate": 9.553264992221442e-06,
+      "loss": 1.5272,
+      "step": 880
+    },
+    {
+      "epoch": 0.5662188099808061,
+      "grad_norm": 2.5933356852122,
+      "learning_rate": 9.441685797443138e-06,
+      "loss": 1.5924,
+      "step": 885
+    },
+    {
+      "epoch": 0.5694177863083814,
+      "grad_norm": 2.3651540877215904,
+      "learning_rate": 9.330176288301332e-06,
+      "loss": 1.565,
+      "step": 890
+    },
+    {
+      "epoch": 0.5726167626359565,
+      "grad_norm": 2.4590451802020734,
+      "learning_rate": 9.21875038278536e-06,
+      "loss": 1.6479,
+      "step": 895
+    },
+    {
+      "epoch": 0.5758157389635317,
+      "grad_norm": 2.3702199508365185,
+      "learning_rate": 9.107421988449632e-06,
+      "loss": 1.5073,
+      "step": 900
+    },
+    {
+      "epoch": 0.5790147152911068,
+      "grad_norm": 2.5796981737540645,
+      "learning_rate": 8.996205000677758e-06,
+      "loss": 1.5437,
+      "step": 905
+    },
+    {
+      "epoch": 0.582213691618682,
+      "grad_norm": 2.6018058977122855,
+      "learning_rate": 8.885113300948199e-06,
+      "loss": 1.5271,
+      "step": 910
+    },
+    {
+      "epoch": 0.5854126679462572,
+      "grad_norm": 2.6014600033150956,
+      "learning_rate": 8.774160755101685e-06,
+      "loss": 1.5063,
+      "step": 915
+    },
+    {
+      "epoch": 0.5886116442738324,
+      "grad_norm": 2.422587375645625,
+      "learning_rate": 8.663361211610529e-06,
+      "loss": 1.4959,
+      "step": 920
+    },
+    {
+      "epoch": 0.5918106206014075,
+      "grad_norm": 2.5294546007457135,
+      "learning_rate": 8.552728499850153e-06,
+      "loss": 1.5216,
+      "step": 925
+    },
+    {
+      "epoch": 0.5950095969289827,
+      "grad_norm": 2.3542539093499344,
+      "learning_rate": 8.442276428372986e-06,
+      "loss": 1.5783,
+      "step": 930
+    },
+    {
+      "epoch": 0.5982085732565579,
+      "grad_norm": 2.5576779252359296,
+      "learning_rate": 8.332018783184952e-06,
+      "loss": 1.5815,
+      "step": 935
+    },
+    {
+      "epoch": 0.6014075495841331,
+      "grad_norm": 2.4091583012332243,
+      "learning_rate": 8.221969326024776e-06,
+      "loss": 1.4926,
+      "step": 940
+    },
+    {
+      "epoch": 0.6046065259117083,
+      "grad_norm": 2.4157626052804115,
+      "learning_rate": 8.112141792646326e-06,
+      "loss": 1.5961,
+      "step": 945
+    },
+    {
+      "epoch": 0.6078055022392834,
+      "grad_norm": 2.728678425231519,
+      "learning_rate": 8.002549891104196e-06,
+      "loss": 1.5204,
+      "step": 950
+    },
+    {
+      "epoch": 0.6110044785668586,
+      "grad_norm": 2.693508248242465,
+      "learning_rate": 7.89320730004274e-06,
+      "loss": 1.5202,
+      "step": 955
+    },
+    {
+      "epoch": 0.6142034548944337,
+      "grad_norm": 2.4606255061660742,
+      "learning_rate": 7.784127666988774e-06,
+      "loss": 1.5788,
+      "step": 960
+    },
+    {
+      "epoch": 0.617402431222009,
+      "grad_norm": 2.4415648713028584,
+      "learning_rate": 7.675324606648187e-06,
+      "loss": 1.4761,
+      "step": 965
+    },
+    {
+      "epoch": 0.6206014075495841,
+      "grad_norm": 2.6515670463268144,
+      "learning_rate": 7.566811699206604e-06,
+      "loss": 1.5426,
+      "step": 970
+    },
+    {
+      "epoch": 0.6238003838771593,
+      "grad_norm": 2.464998512720093,
+      "learning_rate": 7.458602488634416e-06,
+      "loss": 1.5232,
+      "step": 975
+    },
+    {
+      "epoch": 0.6269993602047345,
+      "grad_norm": 2.576825954460212,
+      "learning_rate": 7.350710480996266e-06,
+      "loss": 1.5492,
+      "step": 980
+    },
+    {
+      "epoch": 0.6301983365323096,
+      "grad_norm": 2.6343391183245477,
+      "learning_rate": 7.243149142765317e-06,
+      "loss": 1.5366,
+      "step": 985
+    },
+    {
+      "epoch": 0.6333973128598849,
+      "grad_norm": 2.7039136946619378,
+      "learning_rate": 7.135931899142438e-06,
+      "loss": 1.5308,
+      "step": 990
+    },
+    {
+      "epoch": 0.63659628918746,
+      "grad_norm": 2.487153313721114,
+      "learning_rate": 7.029072132380543e-06,
+      "loss": 1.5458,
+      "step": 995
+    },
+    {
+      "epoch": 0.6397952655150352,
+      "grad_norm": 2.4436815591141055,
+      "learning_rate": 6.922583180114299e-06,
+      "loss": 1.5467,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6429942418426103,
+      "grad_norm": 2.5798180176201737,
+      "learning_rate": 6.816478333695392e-06,
+      "loss": 1.4839,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6461932181701855,
+      "grad_norm": 2.4447683187654317,
+      "learning_rate": 6.710770836533584e-06,
+      "loss": 1.555,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6493921944977608,
+      "grad_norm": 2.466415598823523,
+      "learning_rate": 6.605473882443725e-06,
+      "loss": 1.4658,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6525911708253359,
+      "grad_norm": 2.447032435340423,
+      "learning_rate": 6.5006006139990115e-06,
+      "loss": 1.56,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6557901471529111,
+      "grad_norm": 2.487772254910239,
+      "learning_rate": 6.396164120890562e-06,
+      "loss": 1.5423,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6589891234804862,
+      "grad_norm": 2.375322480797237,
+      "learning_rate": 6.292177438293665e-06,
+      "loss": 1.5784,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6621880998080614,
+      "grad_norm": 2.63673786424632,
+      "learning_rate": 6.188653545240793e-06,
+      "loss": 1.4851,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6653870761356366,
+      "grad_norm": 2.372452515646806,
+      "learning_rate": 6.085605363001628e-06,
+      "loss": 1.4868,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6685860524632118,
+      "grad_norm": 2.5610106308123,
+      "learning_rate": 5.983045753470308e-06,
+      "loss": 1.5517,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6717850287907869,
+      "grad_norm": 2.4328153274434534,
+      "learning_rate": 5.880987517560075e-06,
+      "loss": 1.4631,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6749840051183621,
+      "grad_norm": 2.441028073653641,
+      "learning_rate": 5.779443393605536e-06,
+      "loss": 1.4843,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6781829814459372,
+      "grad_norm": 2.7162623374699346,
+      "learning_rate": 5.678426055772746e-06,
+      "loss": 1.5357,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6813819577735125,
+      "grad_norm": 2.4530732028130067,
+      "learning_rate": 5.577948112477271e-06,
+      "loss": 1.5454,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6845809341010877,
+      "grad_norm": 2.4890182788114696,
+      "learning_rate": 5.478022104810483e-06,
+      "loss": 1.5102,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6877799104286628,
+      "grad_norm": 2.42286314600682,
+      "learning_rate": 5.378660504974271e-06,
+      "loss": 1.4676,
+      "step": 1075
+    },
+    {
+      "epoch": 0.690978886756238,
+      "grad_norm": 2.4804784717437576,
+      "learning_rate": 5.2798757147242986e-06,
+      "loss": 1.567,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6941778630838131,
+      "grad_norm": 2.7295949677609577,
+      "learning_rate": 5.1816800638221176e-06,
+      "loss": 1.5184,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6973768394113884,
+      "grad_norm": 2.537601563528886,
+      "learning_rate": 5.084085808496211e-06,
+      "loss": 1.5231,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7005758157389635,
+      "grad_norm": 2.4956593302493406,
+      "learning_rate": 4.987105129912252e-06,
+      "loss": 1.486,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7037747920665387,
+      "grad_norm": 2.798138201731909,
+      "learning_rate": 4.890750132652724e-06,
+      "loss": 1.5195,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7069737683941139,
+      "grad_norm": 2.571458670426184,
+      "learning_rate": 4.795032843206091e-06,
+      "loss": 1.4892,
+      "step": 1105
+    },
+    {
+      "epoch": 0.710172744721689,
+      "grad_norm": 2.8172099210546744,
+      "learning_rate": 4.699965208465702e-06,
+      "loss": 1.5646,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7133717210492643,
+      "grad_norm": 2.5674334312495093,
+      "learning_rate": 4.605559094238681e-06,
+      "loss": 1.5747,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7165706973768394,
+      "grad_norm": 2.3341361723396075,
+      "learning_rate": 4.511826283764872e-06,
+      "loss": 1.548,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7197696737044146,
+      "grad_norm": 2.366148041831113,
+      "learning_rate": 4.418778476246132e-06,
+      "loss": 1.4312,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7229686500319897,
+      "grad_norm": 2.5343317843332955,
+      "learning_rate": 4.326427285386118e-06,
+      "loss": 1.5053,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7261676263595649,
+      "grad_norm": 2.359774519977056,
+      "learning_rate": 4.234784237940705e-06,
+      "loss": 1.4653,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7293666026871402,
+      "grad_norm": 2.4400916788306772,
+      "learning_rate": 4.143860772279287e-06,
+      "loss": 1.4899,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7325655790147153,
+      "grad_norm": 2.623904935527164,
+      "learning_rate": 4.053668236957135e-06,
+      "loss": 1.5307,
+      "step": 1145
+    },
+    {
+      "epoch": 0.7357645553422905,
+      "grad_norm": 2.4701682729848256,
+      "learning_rate": 3.964217889298882e-06,
+      "loss": 1.4983,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7389635316698656,
+      "grad_norm": 2.4296405777379104,
+      "learning_rate": 3.875520893993482e-06,
+      "loss": 1.53,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7421625079974408,
+      "grad_norm": 2.623309835362006,
+      "learning_rate": 3.787588321700697e-06,
+      "loss": 1.5473,
+      "step": 1160
+    },
+    {
+      "epoch": 0.745361484325016,
+      "grad_norm": 2.3592387568189963,
+      "learning_rate": 3.7004311476692977e-06,
+      "loss": 1.534,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7485604606525912,
+      "grad_norm": 2.484561103362345,
+      "learning_rate": 3.6140602503672217e-06,
+      "loss": 1.516,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7517594369801663,
+      "grad_norm": 2.5714704618567272,
+      "learning_rate": 3.528486410123758e-06,
+      "loss": 1.4717,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7549584133077415,
+      "grad_norm": 2.473733289812066,
+      "learning_rate": 3.443720307784022e-06,
+      "loss": 1.4859,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7581573896353166,
+      "grad_norm": 2.5807040815191384,
+      "learning_rate": 3.359772523375837e-06,
+      "loss": 1.5437,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7613563659628919,
+      "grad_norm": 2.6351719725329157,
+      "learning_rate": 3.276653534789185e-06,
+      "loss": 1.5152,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7645553422904671,
+      "grad_norm": 2.481998923911893,
+      "learning_rate": 3.1943737164684094e-06,
+      "loss": 1.509,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7677543186180422,
+      "grad_norm": 2.4407227083761223,
+      "learning_rate": 3.11294333811736e-06,
+      "loss": 1.4421,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7709532949456174,
+      "grad_norm": 2.427463345964753,
+      "learning_rate": 3.032372563417556e-06,
+      "loss": 1.5026,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7741522712731925,
+      "grad_norm": 2.7179825626721947,
+      "learning_rate": 2.952671448759631e-06,
+      "loss": 1.5082,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7773512476007678,
+      "grad_norm": 2.49463783683528,
+      "learning_rate": 2.8738499419881517e-06,
+      "loss": 1.5174,
+      "step": 1215
+    },
+    {
+      "epoch": 0.780550223928343,
+      "grad_norm": 2.7159929171274317,
+      "learning_rate": 2.795917881159973e-06,
+      "loss": 1.4513,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7837492002559181,
+      "grad_norm": 2.499199393921607,
+      "learning_rate": 2.718884993316311e-06,
+      "loss": 1.5236,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7869481765834933,
+      "grad_norm": 2.494526532816132,
+      "learning_rate": 2.642760893268684e-06,
+      "loss": 1.492,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7901471529110684,
+      "grad_norm": 2.4433657766906984,
+      "learning_rate": 2.5675550823988184e-06,
+      "loss": 1.5034,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7933461292386437,
+      "grad_norm": 2.591611879746854,
+      "learning_rate": 2.493276947472756e-06,
+      "loss": 1.5017,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7965451055662188,
+      "grad_norm": 2.608689307398732,
+      "learning_rate": 2.4199357594692564e-06,
+      "loss": 1.4498,
+      "step": 1245
+    },
+    {
+      "epoch": 0.799744081893794,
+      "grad_norm": 2.5807529550129478,
+      "learning_rate": 2.347540672422627e-06,
+      "loss": 1.4488,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8029430582213691,
+      "grad_norm": 2.444046888012205,
+      "learning_rate": 2.2761007222801866e-06,
+      "loss": 1.4879,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8061420345489443,
+      "grad_norm": 2.556218040341285,
+      "learning_rate": 2.2056248257744383e-06,
+      "loss": 1.4814,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8093410108765196,
+      "grad_norm": 2.3761884198072525,
+      "learning_rate": 2.136121779310132e-06,
+      "loss": 1.4938,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8125399872040947,
+      "grad_norm": 2.555747029316681,
+      "learning_rate": 2.067600257866358e-06,
+      "loss": 1.5016,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8157389635316699,
+      "grad_norm": 2.607092244917964,
+      "learning_rate": 2.000068813913777e-06,
+      "loss": 1.5008,
+      "step": 1275
+    },
+    {
+      "epoch": 0.818937939859245,
+      "grad_norm": 2.3856644731191503,
+      "learning_rate": 1.933535876347141e-06,
+      "loss": 1.5232,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8221369161868202,
+      "grad_norm": 2.557854119192549,
+      "learning_rate": 1.8680097494332682e-06,
+      "loss": 1.492,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8253358925143954,
+      "grad_norm": 2.3969485171992697,
+      "learning_rate": 1.8034986117745245e-06,
+      "loss": 1.4561,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8285348688419706,
+      "grad_norm": 2.6466425327738823,
+      "learning_rate": 1.7400105152880331e-06,
+      "loss": 1.4648,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8317338451695457,
+      "grad_norm": 2.586825079027672,
+      "learning_rate": 1.6775533842006853e-06,
+      "loss": 1.5776,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8349328214971209,
+      "grad_norm": 2.3870646040382453,
+      "learning_rate": 1.6161350140600606e-06,
+      "loss": 1.4682,
+      "step": 1305
+    },
+    {
+      "epoch": 0.838131797824696,
+      "grad_norm": 2.5215206212469217,
+      "learning_rate": 1.555763070761448e-06,
+      "loss": 1.5412,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8413307741522713,
+      "grad_norm": 2.3464462532435917,
+      "learning_rate": 1.496445089591031e-06,
+      "loss": 1.5152,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8445297504798465,
+      "grad_norm": 2.8068099408312164,
+      "learning_rate": 1.4381884742853723e-06,
+      "loss": 1.5109,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8477287268074216,
+      "grad_norm": 2.5858712843708047,
+      "learning_rate": 1.381000496107313e-06,
+      "loss": 1.463,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8509277031349968,
+      "grad_norm": 2.6346898060885446,
+      "learning_rate": 1.3248882929384321e-06,
+      "loss": 1.4372,
+      "step": 1330
+    },
+    {
+      "epoch": 0.8541266794625719,
+      "grad_norm": 2.4502197712622005,
+      "learning_rate": 1.2698588683881185e-06,
+      "loss": 1.5104,
+      "step": 1335
+    },
+    {
+      "epoch": 0.8573256557901472,
+      "grad_norm": 2.6070455124090723,
+      "learning_rate": 1.215919090919434e-06,
+      "loss": 1.5453,
+      "step": 1340
+    },
+    {
+      "epoch": 0.8605246321177223,
+      "grad_norm": 2.613859058553336,
+      "learning_rate": 1.1630756929918218e-06,
+      "loss": 1.3629,
+      "step": 1345
+    },
+    {
+      "epoch": 0.8637236084452975,
+      "grad_norm": 2.555619592197141,
+      "learning_rate": 1.111335270220798e-06,
+      "loss": 1.4444,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8669225847728727,
+      "grad_norm": 2.391401860119296,
+      "learning_rate": 1.060704280554733e-06,
+      "loss": 1.4823,
+      "step": 1355
+    },
+    {
+      "epoch": 0.8701215611004478,
+      "grad_norm": 2.3676422364993845,
+      "learning_rate": 1.0111890434688009e-06,
+      "loss": 1.516,
+      "step": 1360
+    },
+    {
+      "epoch": 0.8733205374280231,
+      "grad_norm": 2.7016986166296455,
+      "learning_rate": 9.62795739176212e-07,
+      "loss": 1.5778,
+      "step": 1365
+    },
+    {
+      "epoch": 0.8765195137555982,
+      "grad_norm": 2.552693201900853,
+      "learning_rate": 9.155304078568495e-07,
+      "loss": 1.4319,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8797184900831734,
+      "grad_norm": 2.584684528802189,
+      "learning_rate": 8.693989489033494e-07,
+      "loss": 1.407,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8829174664107485,
+      "grad_norm": 2.4070124325794056,
+      "learning_rate": 8.244071201847826e-07,
+      "loss": 1.4465,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8861164427383237,
+      "grad_norm": 2.40972912643448,
+      "learning_rate": 7.805605373279934e-07,
+      "loss": 1.467,
+      "step": 1385
+    },
+    {
+      "epoch": 0.889315419065899,
+      "grad_norm": 2.3832038340879427,
+      "learning_rate": 7.378646730166783e-07,
+      "loss": 1.4711,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8925143953934741,
+      "grad_norm": 2.71612283778794,
+      "learning_rate": 6.963248563083203e-07,
+      "loss": 1.4954,
+      "step": 1395
+    },
+    {
+      "epoch": 0.8957133717210493,
+      "grad_norm": 2.363104397198656,
+      "learning_rate": 6.559462719690501e-07,
+      "loss": 1.471,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8989123480486244,
+      "grad_norm": 2.4973910430866404,
+      "learning_rate": 6.167339598265109e-07,
+      "loss": 1.4736,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9021113243761996,
+      "grad_norm": 2.8272793794206494,
+      "learning_rate": 5.78692814140811e-07,
+      "loss": 1.4466,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9053103007037748,
+      "grad_norm": 2.378150271799849,
+      "learning_rate": 5.418275829936537e-07,
+      "loss": 1.4711,
+      "step": 1415
+    },
+    {
+      "epoch": 0.90850927703135,
+      "grad_norm": 2.6449329469693383,
+      "learning_rate": 5.06142867695708e-07,
+      "loss": 1.4544,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9117082533589251,
+      "grad_norm": 2.4686342917623842,
+      "learning_rate": 4.716431222122997e-07,
+      "loss": 1.4822,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9149072296865003,
+      "grad_norm": 2.5348886312509595,
+      "learning_rate": 4.3833265260749157e-07,
+      "loss": 1.4774,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9181062060140754,
+      "grad_norm": 2.573276808796142,
+      "learning_rate": 4.062156165066211e-07,
+      "loss": 1.5098,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9213051823416507,
+      "grad_norm": 2.4977650539207854,
+      "learning_rate": 3.752960225773772e-07,
+      "loss": 1.4559,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9245041586692259,
+      "grad_norm": 2.3752215307243594,
+      "learning_rate": 3.4557773002945607e-07,
+      "loss": 1.4437,
+      "step": 1445
+    },
+    {
+      "epoch": 0.927703134996801,
+      "grad_norm": 2.5462973017160584,
+      "learning_rate": 3.170644481328711e-07,
+      "loss": 1.4723,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9309021113243762,
+      "grad_norm": 2.595249620944453,
+      "learning_rate": 2.8975973575499526e-07,
+      "loss": 1.4827,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9341010876519513,
+      "grad_norm": 2.455367708761631,
+      "learning_rate": 2.636670009163522e-07,
+      "loss": 1.4999,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9373000639795266,
+      "grad_norm": 2.3405669217821887,
+      "learning_rate": 2.3878950036524963e-07,
+      "loss": 1.3879,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9404990403071017,
+      "grad_norm": 2.6237619212466092,
+      "learning_rate": 2.1513033917129334e-07,
+      "loss": 1.4763,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9436980166346769,
+      "grad_norm": 2.6691853952475024,
+      "learning_rate": 1.9269247033782744e-07,
+      "loss": 1.4632,
+      "step": 1475
+    },
+    {
+      "epoch": 0.946896992962252,
+      "grad_norm": 2.5498332777296624,
+      "learning_rate": 1.7147869443335463e-07,
+      "loss": 1.474,
+      "step": 1480
+    },
+    {
+      "epoch": 0.9500959692898272,
+      "grad_norm": 2.459870224044313,
+      "learning_rate": 1.5149165924199016e-07,
+      "loss": 1.4654,
+      "step": 1485
+    },
+    {
+      "epoch": 0.9532949456174025,
+      "grad_norm": 2.511293723941215,
+      "learning_rate": 1.3273385943297746e-07,
+      "loss": 1.4418,
+      "step": 1490
+    },
+    {
+      "epoch": 0.9564939219449776,
+      "grad_norm": 2.56678375061047,
+      "learning_rate": 1.1520763624931597e-07,
+      "loss": 1.6092,
+      "step": 1495
+    },
+    {
+      "epoch": 0.9596928982725528,
+      "grad_norm": 2.6727043053301394,
+      "learning_rate": 9.891517721554499e-08,
+      "loss": 1.4307,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9628918746001279,
+      "grad_norm": 2.5296611789075136,
+      "learning_rate": 8.385851586470318e-08,
+      "loss": 1.4335,
+      "step": 1505
+    },
+    {
+      "epoch": 0.9660908509277031,
+      "grad_norm": 2.583371271354282,
+      "learning_rate": 7.003953148452036e-08,
+      "loss": 1.525,
+      "step": 1510
+    },
+    {
+      "epoch": 0.9692898272552783,
+      "grad_norm": 2.5459755085081186,
+      "learning_rate": 5.745994888285311e-08,
+      "loss": 1.4828,
+      "step": 1515
+    },
+    {
+      "epoch": 0.9724888035828535,
+      "grad_norm": 2.4989846165964,
+      "learning_rate": 4.612133817239905e-08,
+      "loss": 1.4797,
+      "step": 1520
+    },
+    {
+      "epoch": 0.9756877799104287,
+      "grad_norm": 2.5699920030762646,
+      "learning_rate": 3.602511457473479e-08,
+      "loss": 1.4892,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9788867562380038,
+      "grad_norm": 2.3932347022659544,
+      "learning_rate": 2.7172538243666057e-08,
+      "loss": 1.4259,
+      "step": 1530
+    },
+    {
+      "epoch": 0.982085732565579,
+      "grad_norm": 2.502306737907492,
+      "learning_rate": 1.9564714107945804e-08,
+      "loss": 1.533,
+      "step": 1535
+    },
+    {
+      "epoch": 0.9852847088931542,
+      "grad_norm": 2.399652752946487,
+      "learning_rate": 1.3202591733365577e-08,
+      "loss": 1.4265,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9884836852207294,
+      "grad_norm": 2.4710638611047795,
+      "learning_rate": 8.086965204233688e-09,
+      "loss": 1.5123,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9916826615483045,
+      "grad_norm": 2.6100528553962663,
+      "learning_rate": 4.218473024261149e-09,
+      "loss": 1.5122,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9948816378758797,
+      "grad_norm": 2.5237627996475465,
+      "learning_rate": 1.5975980368709843e-09,
+      "loss": 1.426,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9980806142034548,
+      "grad_norm": 2.508010573111159,
+      "learning_rate": 2.2466736492643416e-10,
+      "loss": 1.4184,
+      "step": 1560
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.526857614517212,
+      "eval_runtime": 18.9319,
+      "eval_samples_per_second": 26.41,
+      "eval_steps_per_second": 0.845,
+      "step": 1563
     },
     {
       "epoch": 1.0,
-      "step": 225,
-      "total_flos": 11777605632000.0,
-      "train_loss": 1.5383918571472168,
-      "train_runtime": 1636.4794,
-      "train_samples_per_second": 4.388,
-      "train_steps_per_second": 0.137
+      "step": 1563,
+      "total_flos": 17631303106560.0,
+      "train_loss": 1.5734304829583439,
+      "train_runtime": 5865.6403,
+      "train_samples_per_second": 8.524,
+      "train_steps_per_second": 0.266
     }
   ],
   "logging_steps": 5,
-  "max_steps": 225,
+  "max_steps": 1563,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 500,
@@ -365,7 +2234,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 11777605632000.0,
+  "total_flos": 17631303106560.0,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null