End of training

all_results.json

@@ -1,5 +1,5 @@
 {
-    "epoch": 2.99,
+    "epoch": 3.0,
     "eval_accuracy": 0.7105720509450737,
     "eval_loss": 1.2192070484161377,
     "eval_runtime": 4409.7065,
@@ -7,9 +7,9 @@
     "eval_samples_per_second": 30.601,
     "eval_steps_per_second": 7.65,
     "perplexity": 3.384502922346475,
-    "train_loss": 0.0,
-    "train_runtime": 0.125,
+    "train_loss": 1.602629992106659,
+    "train_runtime": 64476.8032,
     "train_samples": 134942,
-    "train_samples_per_second": 3238688.305,
-    "train_steps_per_second": 404842.038
+    "train_samples_per_second": 6.279,
+    "train_steps_per_second": 3.139
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.99,
-    "train_loss": 0.0,
-    "train_runtime": 0.125,
+    "epoch": 3.0,
+    "train_loss": 1.602629992106659,
+    "train_runtime": 64476.8032,
     "train_samples": 134942,
-    "train_samples_per_second": 3238688.305,
-    "train_steps_per_second": 404842.038
+    "train_samples_per_second": 6.279,
+    "train_steps_per_second": 3.139
 }

trainer_state.json

@@ -1,1237 +1,2449 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.993834479487788,
-  "global_step": 101000,
+  "epoch": 3.0,
+  "global_step": 202413,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.01,
-      "learning_rate": 4.9752983953837645e-05,
-      "loss": 1.339,
+      "learning_rate": 4.9876490146383884e-05,
+      "loss": 1.4737,
       "step": 500
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 4.950596790767528e-05,
-      "loss": 1.3788,
+      "epoch": 0.01,
+      "learning_rate": 4.975298029276776e-05,
+      "loss": 1.4854,
       "step": 1000
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 4.9258951861512924e-05,
-      "loss": 1.378,
+      "epoch": 0.02,
+      "learning_rate": 4.962947043915164e-05,
+      "loss": 1.5299,
       "step": 1500
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 4.901193581535057e-05,
-      "loss": 1.3805,
+      "epoch": 0.03,
+      "learning_rate": 4.950596058553551e-05,
+      "loss": 1.5417,
       "step": 2000
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 4.876491976918821e-05,
-      "loss": 1.3945,
+      "epoch": 0.04,
+      "learning_rate": 4.9382450731919394e-05,
+      "loss": 1.5408,
       "step": 2500
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 4.8517903723025847e-05,
-      "loss": 1.3919,
+      "epoch": 0.04,
+      "learning_rate": 4.9258940878303276e-05,
+      "loss": 1.542,
       "step": 3000
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 4.827088767686349e-05,
-      "loss": 1.4032,
+      "epoch": 0.05,
+      "learning_rate": 4.913543102468715e-05,
+      "loss": 1.5424,
       "step": 3500
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.802387163070113e-05,
-      "loss": 1.4197,
+      "epoch": 0.06,
+      "learning_rate": 4.901192117107103e-05,
+      "loss": 1.5585,
       "step": 4000
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.7776855584538775e-05,
-      "loss": 1.4222,
+      "epoch": 0.07,
+      "learning_rate": 4.8888411317454905e-05,
+      "loss": 1.5505,
       "step": 4500
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.752983953837641e-05,
-      "loss": 1.4057,
+      "epoch": 0.07,
+      "learning_rate": 4.8764901463838787e-05,
+      "loss": 1.5816,
       "step": 5000
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.7282823492214054e-05,
-      "loss": 1.4156,
+      "epoch": 0.08,
+      "learning_rate": 4.864139161022266e-05,
+      "loss": 1.5698,
       "step": 5500
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.70358074460517e-05,
-      "loss": 1.4243,
+      "epoch": 0.09,
+      "learning_rate": 4.851788175660654e-05,
+      "loss": 1.573,
       "step": 6000
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.6788791399889334e-05,
-      "loss": 1.4176,
+      "epoch": 0.1,
+      "learning_rate": 4.839437190299042e-05,
+      "loss": 1.5797,
       "step": 6500
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.6541775353726977e-05,
-      "loss": 1.4254,
+      "epoch": 0.1,
+      "learning_rate": 4.82708620493743e-05,
+      "loss": 1.5854,
       "step": 7000
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.629475930756462e-05,
-      "loss": 1.4088,
+      "epoch": 0.11,
+      "learning_rate": 4.814735219575818e-05,
+      "loss": 1.5935,
       "step": 7500
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.604774326140226e-05,
-      "loss": 1.4307,
+      "epoch": 0.12,
+      "learning_rate": 4.802384234214205e-05,
+      "loss": 1.6072,
       "step": 8000
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.5800727215239905e-05,
-      "loss": 1.4459,
+      "epoch": 0.13,
+      "learning_rate": 4.7900332488525934e-05,
+      "loss": 1.6015,
       "step": 8500
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.555371116907755e-05,
-      "loss": 1.4486,
+      "epoch": 0.13,
+      "learning_rate": 4.7776822634909815e-05,
+      "loss": 1.6086,
       "step": 9000
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.530669512291519e-05,
-      "loss": 1.4432,
+      "epoch": 0.14,
+      "learning_rate": 4.7653312781293696e-05,
+      "loss": 1.5901,
       "step": 9500
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.505967907675283e-05,
-      "loss": 1.4432,
+      "epoch": 0.15,
+      "learning_rate": 4.752980292767758e-05,
+      "loss": 1.5899,
       "step": 10000
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.481266303059047e-05,
-      "loss": 1.4424,
+      "epoch": 0.16,
+      "learning_rate": 4.740629307406145e-05,
+      "loss": 1.5804,
       "step": 10500
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.456564698442811e-05,
-      "loss": 1.4498,
+      "epoch": 0.16,
+      "learning_rate": 4.728278322044533e-05,
+      "loss": 1.6211,
       "step": 11000
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.4318630938265756e-05,
-      "loss": 1.4754,
+      "epoch": 0.17,
+      "learning_rate": 4.715927336682921e-05,
+      "loss": 1.6268,
       "step": 11500
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.407161489210339e-05,
-      "loss": 1.4712,
+      "epoch": 0.18,
+      "learning_rate": 4.703576351321309e-05,
+      "loss": 1.5938,
       "step": 12000
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.3824598845941035e-05,
-      "loss": 1.4552,
+      "epoch": 0.19,
+      "learning_rate": 4.691225365959697e-05,
+      "loss": 1.6143,
       "step": 12500
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.357758279977868e-05,
-      "loss": 1.4759,
+      "epoch": 0.19,
+      "learning_rate": 4.6788743805980844e-05,
+      "loss": 1.5977,
       "step": 13000
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.333056675361632e-05,
-      "loss": 1.4877,
+      "epoch": 0.2,
+      "learning_rate": 4.6665233952364725e-05,
+      "loss": 1.6157,
       "step": 13500
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.308355070745396e-05,
-      "loss": 1.4775,
+      "epoch": 0.21,
+      "learning_rate": 4.65417240987486e-05,
+      "loss": 1.6176,
       "step": 14000
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.28365346612916e-05,
-      "loss": 1.4701,
+      "epoch": 0.21,
+      "learning_rate": 4.641821424513248e-05,
+      "loss": 1.6032,
       "step": 14500
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.258951861512924e-05,
-      "loss": 1.4771,
+      "epoch": 0.22,
+      "learning_rate": 4.6294704391516355e-05,
+      "loss": 1.5959,
       "step": 15000
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.234250256896688e-05,
-      "loss": 1.4856,
+      "epoch": 0.23,
+      "learning_rate": 4.6171194537900236e-05,
+      "loss": 1.6247,
       "step": 15500
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.209548652280452e-05,
-      "loss": 1.4878,
+      "epoch": 0.24,
+      "learning_rate": 4.604768468428412e-05,
+      "loss": 1.615,
       "step": 16000
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.1848470476642165e-05,
-      "loss": 1.5109,
+      "epoch": 0.24,
+      "learning_rate": 4.592417483066799e-05,
+      "loss": 1.6421,
       "step": 16500
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.160145443047981e-05,
-      "loss": 1.5056,
+      "epoch": 0.25,
+      "learning_rate": 4.580066497705187e-05,
+      "loss": 1.6377,
       "step": 17000
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.1354438384317445e-05,
-      "loss": 1.4853,
+      "epoch": 0.26,
+      "learning_rate": 4.567715512343575e-05,
+      "loss": 1.6335,
       "step": 17500
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.110742233815509e-05,
-      "loss": 1.5015,
+      "epoch": 0.27,
+      "learning_rate": 4.555364526981963e-05,
+      "loss": 1.6476,
       "step": 18000
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.086040629199273e-05,
-      "loss": 1.4883,
+      "epoch": 0.27,
+      "learning_rate": 4.543013541620351e-05,
+      "loss": 1.6314,
       "step": 18500
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.061339024583037e-05,
-      "loss": 1.4959,
+      "epoch": 0.28,
+      "learning_rate": 4.5306625562587384e-05,
+      "loss": 1.637,
       "step": 19000
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.036637419966801e-05,
-      "loss": 1.4887,
+      "epoch": 0.29,
+      "learning_rate": 4.5183115708971265e-05,
+      "loss": 1.6289,
       "step": 19500
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.011935815350565e-05,
-      "loss": 1.5087,
+      "epoch": 0.3,
+      "learning_rate": 4.505960585535514e-05,
+      "loss": 1.6434,
       "step": 20000
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 3.9872342107343295e-05,
-      "loss": 1.5022,
+      "epoch": 0.3,
+      "learning_rate": 4.493609600173902e-05,
+      "loss": 1.6458,
       "step": 20500
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 3.962532606118093e-05,
-      "loss": 1.5063,
+      "epoch": 0.31,
+      "learning_rate": 4.48125861481229e-05,
+      "loss": 1.629,
       "step": 21000
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 3.9378310015018575e-05,
-      "loss": 1.5054,
+      "epoch": 0.32,
+      "learning_rate": 4.4689076294506776e-05,
+      "loss": 1.6385,
       "step": 21500
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 3.913129396885622e-05,
-      "loss": 1.5237,
+      "epoch": 0.33,
+      "learning_rate": 4.456556644089066e-05,
+      "loss": 1.6508,
       "step": 22000
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 3.888427792269386e-05,
-      "loss": 1.4989,
+      "epoch": 0.33,
+      "learning_rate": 4.444205658727453e-05,
+      "loss": 1.6715,
       "step": 22500
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 3.86372618765315e-05,
-      "loss": 1.521,
+      "epoch": 0.34,
+      "learning_rate": 4.431854673365841e-05,
+      "loss": 1.6706,
       "step": 23000
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 3.839024583036914e-05,
-      "loss": 1.5334,
+      "epoch": 0.35,
+      "learning_rate": 4.419503688004229e-05,
+      "loss": 1.6633,
       "step": 23500
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 3.814322978420678e-05,
-      "loss": 1.5102,
+      "epoch": 0.36,
+      "learning_rate": 4.407152702642617e-05,
+      "loss": 1.6705,
       "step": 24000
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 3.7896213738044425e-05,
-      "loss": 1.5112,
+      "epoch": 0.36,
+      "learning_rate": 4.394801717281005e-05,
+      "loss": 1.6609,
       "step": 24500
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 3.764919769188207e-05,
-      "loss": 1.534,
+      "epoch": 0.37,
+      "learning_rate": 4.382450731919392e-05,
+      "loss": 1.6349,
       "step": 25000
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 3.740218164571971e-05,
-      "loss": 1.5219,
+      "epoch": 0.38,
+      "learning_rate": 4.3700997465577804e-05,
+      "loss": 1.6829,
       "step": 25500
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 3.7155165599557354e-05,
-      "loss": 1.551,
+      "epoch": 0.39,
+      "learning_rate": 4.3577487611961686e-05,
+      "loss": 1.6589,
       "step": 26000
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 3.690814955339499e-05,
-      "loss": 1.5412,
+      "epoch": 0.39,
+      "learning_rate": 4.345397775834557e-05,
+      "loss": 1.6871,
       "step": 26500
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 3.6661133507232633e-05,
-      "loss": 1.539,
+      "epoch": 0.4,
+      "learning_rate": 4.333046790472944e-05,
+      "loss": 1.6845,
       "step": 27000
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 3.6414117461070276e-05,
-      "loss": 1.5338,
+      "epoch": 0.41,
+      "learning_rate": 4.320695805111332e-05,
+      "loss": 1.6754,
       "step": 27500
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 3.616710141490792e-05,
-      "loss": 1.531,
+      "epoch": 0.41,
+      "learning_rate": 4.30834481974972e-05,
+      "loss": 1.6712,
       "step": 28000
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 3.5920085368745556e-05,
-      "loss": 1.5448,
+      "epoch": 0.42,
+      "learning_rate": 4.295993834388108e-05,
+      "loss": 1.6638,
       "step": 28500
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 3.56730693225832e-05,
-      "loss": 1.5478,
+      "epoch": 0.43,
+      "learning_rate": 4.283642849026496e-05,
+      "loss": 1.6623,
       "step": 29000
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 3.542605327642084e-05,
-      "loss": 1.5441,
+      "epoch": 0.44,
+      "learning_rate": 4.271291863664883e-05,
+      "loss": 1.6832,
       "step": 29500
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 3.517903723025848e-05,
-      "loss": 1.5415,
+      "epoch": 0.44,
+      "learning_rate": 4.2589408783032714e-05,
+      "loss": 1.6583,
       "step": 30000
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 3.493202118409612e-05,
-      "loss": 1.5513,
+      "epoch": 0.45,
+      "learning_rate": 4.2465898929416595e-05,
+      "loss": 1.683,
       "step": 30500
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 3.4685005137933763e-05,
-      "loss": 1.554,
+      "epoch": 0.46,
+      "learning_rate": 4.234238907580047e-05,
+      "loss": 1.6799,
       "step": 31000
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 3.4437989091771406e-05,
-      "loss": 1.5526,
+      "epoch": 0.47,
+      "learning_rate": 4.221887922218435e-05,
+      "loss": 1.6857,
       "step": 31500
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 3.419097304560904e-05,
-      "loss": 1.5608,
+      "epoch": 0.47,
+      "learning_rate": 4.2095369368568225e-05,
+      "loss": 1.6817,
       "step": 32000
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 3.3943956999446686e-05,
-      "loss": 1.5627,
+      "epoch": 0.48,
+      "learning_rate": 4.1971859514952106e-05,
+      "loss": 1.6998,
       "step": 32500
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 3.369694095328433e-05,
-      "loss": 1.5581,
+      "epoch": 0.49,
+      "learning_rate": 4.184834966133598e-05,
+      "loss": 1.7106,
       "step": 33000
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 3.3449924907121965e-05,
-      "loss": 1.5707,
+      "epoch": 0.5,
+      "learning_rate": 4.172483980771986e-05,
+      "loss": 1.7017,
       "step": 33500
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 3.320290886095961e-05,
-      "loss": 1.4586,
+      "epoch": 0.5,
+      "learning_rate": 4.160132995410374e-05,
+      "loss": 1.6984,
       "step": 34000
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 3.295589281479725e-05,
-      "loss": 1.3769,
+      "epoch": 0.51,
+      "learning_rate": 4.147782010048762e-05,
+      "loss": 1.6843,
       "step": 34500
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 3.2708876768634893e-05,
-      "loss": 1.3723,
+      "epoch": 0.52,
+      "learning_rate": 4.13543102468715e-05,
+      "loss": 1.6751,
       "step": 35000
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 3.246186072247253e-05,
-      "loss": 1.3776,
+      "epoch": 0.53,
+      "learning_rate": 4.123080039325537e-05,
+      "loss": 1.7001,
       "step": 35500
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 3.221484467631017e-05,
-      "loss": 1.391,
+      "epoch": 0.53,
+      "learning_rate": 4.1107290539639254e-05,
+      "loss": 1.696,
       "step": 36000
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 3.1967828630147816e-05,
-      "loss": 1.3848,
+      "epoch": 0.54,
+      "learning_rate": 4.0983780686023135e-05,
+      "loss": 1.6911,
       "step": 36500
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 3.172081258398546e-05,
-      "loss": 1.3889,
+      "epoch": 0.55,
+      "learning_rate": 4.086027083240701e-05,
+      "loss": 1.6793,
       "step": 37000
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 3.1473796537823095e-05,
-      "loss": 1.4039,
+      "epoch": 0.56,
+      "learning_rate": 4.073676097879089e-05,
+      "loss": 1.686,
       "step": 37500
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 3.122678049166074e-05,
-      "loss": 1.3999,
+      "epoch": 0.56,
+      "learning_rate": 4.0613251125174765e-05,
+      "loss": 1.6953,
       "step": 38000
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 3.097976444549838e-05,
-      "loss": 1.4151,
+      "epoch": 0.57,
+      "learning_rate": 4.0489741271558646e-05,
+      "loss": 1.6857,
       "step": 38500
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 3.073274839933602e-05,
-      "loss": 1.4157,
+      "epoch": 0.58,
+      "learning_rate": 4.036623141794252e-05,
+      "loss": 1.6844,
       "step": 39000
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 3.0485732353173663e-05,
-      "loss": 1.4274,
+      "epoch": 0.59,
+      "learning_rate": 4.02427215643264e-05,
+      "loss": 1.6924,
       "step": 39500
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 3.0238716307011306e-05,
-      "loss": 1.4204,
+      "epoch": 0.59,
+      "learning_rate": 4.011921171071028e-05,
+      "loss": 1.717,
       "step": 40000
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 2.999170026084895e-05,
-      "loss": 1.4113,
+      "epoch": 0.6,
+      "learning_rate": 3.999570185709416e-05,
+      "loss": 1.6964,
       "step": 40500
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 2.9744684214686585e-05,
-      "loss": 1.4203,
+      "epoch": 0.61,
+      "learning_rate": 3.987219200347804e-05,
+      "loss": 1.699,
       "step": 41000
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 2.9497668168524228e-05,
-      "loss": 1.3833,
+      "epoch": 0.62,
+      "learning_rate": 3.974868214986191e-05,
+      "loss": 1.6946,
       "step": 41500
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 2.925065212236187e-05,
-      "loss": 1.3336,
+      "epoch": 0.62,
+      "learning_rate": 3.9625172296245794e-05,
+      "loss": 1.7071,
       "step": 42000
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 2.9003636076199514e-05,
-      "loss": 1.3271,
+      "epoch": 0.63,
+      "learning_rate": 3.9501662442629675e-05,
+      "loss": 1.7025,
       "step": 42500
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 2.875662003003715e-05,
-      "loss": 1.3437,
+      "epoch": 0.64,
+      "learning_rate": 3.9378152589013556e-05,
+      "loss": 1.6952,
       "step": 43000
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 2.8509603983874793e-05,
-      "loss": 1.3378,
+      "epoch": 0.64,
+      "learning_rate": 3.925464273539744e-05,
+      "loss": 1.7177,
       "step": 43500
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 2.8262587937712436e-05,
-      "loss": 1.3275,
+      "epoch": 0.65,
+      "learning_rate": 3.913113288178131e-05,
+      "loss": 1.7222,
       "step": 44000
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 2.8015571891550076e-05,
-      "loss": 1.352,
+      "epoch": 0.66,
+      "learning_rate": 3.900762302816519e-05,
+      "loss": 1.6949,
       "step": 44500
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 2.776855584538772e-05,
-      "loss": 1.3527,
+      "epoch": 0.67,
+      "learning_rate": 3.888411317454907e-05,
+      "loss": 1.6928,
       "step": 45000
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 2.752153979922536e-05,
-      "loss": 1.4116,
+      "epoch": 0.67,
+      "learning_rate": 3.876060332093295e-05,
+      "loss": 1.707,
       "step": 45500
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 2.7274523753063e-05,
-      "loss": 1.3997,
+      "epoch": 0.68,
+      "learning_rate": 3.863709346731683e-05,
+      "loss": 1.7227,
       "step": 46000
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 2.702750770690064e-05,
-      "loss": 1.4157,
+      "epoch": 0.69,
+      "learning_rate": 3.8513583613700704e-05,
+      "loss": 1.7294,
       "step": 46500
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 2.6780491660738284e-05,
-      "loss": 1.4121,
+      "epoch": 0.7,
+      "learning_rate": 3.8390073760084585e-05,
+      "loss": 1.7305,
       "step": 47000
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 2.6533475614575926e-05,
-      "loss": 1.4252,
+      "epoch": 0.7,
+      "learning_rate": 3.826656390646846e-05,
+      "loss": 1.6871,
       "step": 47500
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 2.6286459568413563e-05,
-      "loss": 1.4227,
+      "epoch": 0.71,
+      "learning_rate": 3.814305405285234e-05,
+      "loss": 1.7227,
       "step": 48000
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 2.6039443522251206e-05,
-      "loss": 1.429,
+      "epoch": 0.72,
+      "learning_rate": 3.8019544199236215e-05,
+      "loss": 1.6963,
       "step": 48500
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 2.579242747608885e-05,
-      "loss": 1.4173,
+      "epoch": 0.73,
+      "learning_rate": 3.7896034345620096e-05,
+      "loss": 1.7107,
       "step": 49000
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 2.554541142992649e-05,
-      "loss": 1.4398,
+      "epoch": 0.73,
+      "learning_rate": 3.777252449200398e-05,
+      "loss": 1.7351,
       "step": 49500
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 2.5298395383764128e-05,
-      "loss": 1.4437,
+      "epoch": 0.74,
+      "learning_rate": 3.764901463838785e-05,
+      "loss": 1.7234,
       "step": 50000
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 2.505137933760177e-05,
-      "loss": 1.4467,
+      "epoch": 0.75,
+      "learning_rate": 3.752550478477173e-05,
+      "loss": 1.715,
       "step": 50500
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 2.4804363291439414e-05,
-      "loss": 1.4396,
+      "epoch": 0.76,
+      "learning_rate": 3.7401994931155607e-05,
+      "loss": 1.7139,
       "step": 51000
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 2.4557347245277053e-05,
-      "loss": 1.444,
+      "epoch": 0.76,
+      "learning_rate": 3.727848507753949e-05,
+      "loss": 1.7475,
       "step": 51500
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 2.4310331199114696e-05,
-      "loss": 1.4635,
+      "epoch": 0.77,
+      "learning_rate": 3.715497522392337e-05,
+      "loss": 1.7425,
       "step": 52000
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 2.4063315152952336e-05,
-      "loss": 1.4614,
+      "epoch": 0.78,
+      "learning_rate": 3.703146537030724e-05,
+      "loss": 1.7259,
       "step": 52500
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 2.381629910678998e-05,
-      "loss": 1.4783,
+      "epoch": 0.79,
+      "learning_rate": 3.6907955516691124e-05,
+      "loss": 1.7436,
       "step": 53000
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 2.356928306062762e-05,
-      "loss": 1.4699,
+      "epoch": 0.79,
+      "learning_rate": 3.6784445663075e-05,
+      "loss": 1.7127,
       "step": 53500
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 2.332226701446526e-05,
-      "loss": 1.4788,
+      "epoch": 0.8,
+      "learning_rate": 3.666093580945888e-05,
+      "loss": 1.7549,
       "step": 54000
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 2.3075250968302904e-05,
-      "loss": 1.4732,
+      "epoch": 0.81,
+      "learning_rate": 3.653742595584276e-05,
+      "loss": 1.7076,
       "step": 54500
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 2.2828234922140544e-05,
-      "loss": 1.4794,
+      "epoch": 0.82,
+      "learning_rate": 3.6413916102226635e-05,
+      "loss": 1.7495,
       "step": 55000
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 2.2581218875978187e-05,
-      "loss": 1.4708,
+      "epoch": 0.82,
+      "learning_rate": 3.6290406248610516e-05,
+      "loss": 1.725,
       "step": 55500
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 2.2334202829815826e-05,
-      "loss": 1.4666,
+      "epoch": 0.83,
+      "learning_rate": 3.616689639499439e-05,
+      "loss": 1.7273,
       "step": 56000
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 2.208718678365347e-05,
-      "loss": 1.4993,
+      "epoch": 0.84,
+      "learning_rate": 3.604338654137827e-05,
+      "loss": 1.7384,
       "step": 56500
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 2.184017073749111e-05,
-      "loss": 1.4921,
+      "epoch": 0.84,
+      "learning_rate": 3.5919876687762146e-05,
+      "loss": 1.7375,
       "step": 57000
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 2.1593154691328748e-05,
-      "loss": 1.482,
+      "epoch": 0.85,
+      "learning_rate": 3.579636683414603e-05,
+      "loss": 1.7467,
       "step": 57500
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 2.134613864516639e-05,
-      "loss": 1.4862,
+      "epoch": 0.86,
+      "learning_rate": 3.567285698052991e-05,
+      "loss": 1.7395,
       "step": 58000
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 2.109912259900403e-05,
-      "loss": 1.4985,
+      "epoch": 0.87,
+      "learning_rate": 3.554934712691378e-05,
+      "loss": 1.7334,
       "step": 58500
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 2.0852106552841674e-05,
-      "loss": 1.5143,
+      "epoch": 0.87,
+      "learning_rate": 3.5425837273297664e-05,
+      "loss": 1.7382,
       "step": 59000
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 2.0605090506679313e-05,
-      "loss": 1.5043,
+      "epoch": 0.88,
+      "learning_rate": 3.530232741968154e-05,
+      "loss": 1.7269,
       "step": 59500
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 2.0358074460516956e-05,
-      "loss": 1.5227,
+      "epoch": 0.89,
+      "learning_rate": 3.5178817566065426e-05,
+      "loss": 1.7411,
       "step": 60000
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 2.0111058414354596e-05,
-      "loss": 1.5202,
+      "epoch": 0.9,
+      "learning_rate": 3.50553077124493e-05,
+      "loss": 1.7492,
       "step": 60500
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 1.986404236819224e-05,
-      "loss": 1.4978,
+      "epoch": 0.9,
+      "learning_rate": 3.493179785883318e-05,
+      "loss": 1.7388,
       "step": 61000
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 1.961702632202988e-05,
-      "loss": 1.5271,
+      "epoch": 0.91,
+      "learning_rate": 3.480828800521706e-05,
+      "loss": 1.7327,
       "step": 61500
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 1.937001027586752e-05,
-      "loss": 1.5253,
+      "epoch": 0.92,
+      "learning_rate": 3.468477815160094e-05,
+      "loss": 1.7584,
       "step": 62000
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 1.9122994229705164e-05,
-      "loss": 1.5317,
+      "epoch": 0.93,
+      "learning_rate": 3.456126829798482e-05,
+      "loss": 1.7346,
       "step": 62500
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 1.8875978183542804e-05,
-      "loss": 1.5253,
+      "epoch": 0.93,
+      "learning_rate": 3.443775844436869e-05,
+      "loss": 1.7543,
       "step": 63000
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 1.8628962137380447e-05,
-      "loss": 1.5434,
+      "epoch": 0.94,
+      "learning_rate": 3.4314248590752574e-05,
+      "loss": 1.7506,
       "step": 63500
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 1.8381946091218086e-05,
-      "loss": 1.5219,
+      "epoch": 0.95,
+      "learning_rate": 3.4190738737136455e-05,
+      "loss": 1.7512,
       "step": 64000
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 1.813493004505573e-05,
-      "loss": 1.537,
+      "epoch": 0.96,
+      "learning_rate": 3.406722888352033e-05,
+      "loss": 1.745,
       "step": 64500
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 1.788791399889337e-05,
-      "loss": 1.5622,
+      "epoch": 0.96,
+      "learning_rate": 3.394371902990421e-05,
+      "loss": 1.7638,
       "step": 65000
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 1.764089795273101e-05,
-      "loss": 1.5391,
+      "epoch": 0.97,
+      "learning_rate": 3.3820209176288085e-05,
+      "loss": 1.7539,
       "step": 65500
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 1.739388190656865e-05,
-      "loss": 1.573,
+      "epoch": 0.98,
+      "learning_rate": 3.3696699322671966e-05,
+      "loss": 1.7421,
       "step": 66000
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 1.714686586040629e-05,
-      "loss": 1.566,
+      "epoch": 0.99,
+      "learning_rate": 3.357318946905584e-05,
+      "loss": 1.7498,
       "step": 66500
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 1.6899849814243934e-05,
-      "loss": 1.5783,
+      "epoch": 0.99,
+      "learning_rate": 3.344967961543972e-05,
+      "loss": 1.7766,
       "step": 67000
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 1.6652833768081573e-05,
-      "loss": 1.5639,
+      "epoch": 1.0,
+      "learning_rate": 3.33261697618236e-05,
+      "loss": 1.7478,
       "step": 67500
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 1.6405817721919216e-05,
-      "loss": 1.3644,
+      "epoch": 1.01,
+      "learning_rate": 3.320265990820748e-05,
+      "loss": 1.5057,
       "step": 68000
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 1.6158801675756856e-05,
-      "loss": 1.3529,
+      "epoch": 1.02,
+      "learning_rate": 3.307915005459136e-05,
+      "loss": 1.5248,
       "step": 68500
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 1.59117856295945e-05,
-      "loss": 1.3552,
+      "epoch": 1.02,
+      "learning_rate": 3.295564020097523e-05,
+      "loss": 1.5344,
       "step": 69000
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 1.566476958343214e-05,
-      "loss": 1.3508,
+      "epoch": 1.03,
+      "learning_rate": 3.2832130347359114e-05,
+      "loss": 1.5177,
       "step": 69500
     },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.2708620493742995e-05,
+      "loss": 1.5255,
+      "step": 70000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.258511064012687e-05,
+      "loss": 1.5388,
+      "step": 70500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.246160078651075e-05,
+      "loss": 1.5187,
+      "step": 71000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.2338090932894625e-05,
+      "loss": 1.5371,
+      "step": 71500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.2214581079278506e-05,
+      "loss": 1.5466,
+      "step": 72000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.209107122566239e-05,
+      "loss": 1.5409,
+      "step": 72500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.196756137204626e-05,
+      "loss": 1.5329,
+      "step": 73000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.184405151843014e-05,
+      "loss": 1.5404,
+      "step": 73500
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.1720541664814017e-05,
+      "loss": 1.5337,
+      "step": 74000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.15970318111979e-05,
+      "loss": 1.5479,
+      "step": 74500
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.147352195758177e-05,
+      "loss": 1.5608,
+      "step": 75000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.135001210396565e-05,
+      "loss": 1.5568,
+      "step": 75500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.1226502250349534e-05,
+      "loss": 1.5503,
+      "step": 76000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.110299239673341e-05,
+      "loss": 1.5678,
+      "step": 76500
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.0979482543117297e-05,
+      "loss": 1.5654,
+      "step": 77000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.085597268950117e-05,
+      "loss": 1.5808,
+      "step": 77500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.073246283588505e-05,
+      "loss": 1.5567,
+      "step": 78000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.0608952982268926e-05,
+      "loss": 1.5816,
+      "step": 78500
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.0485443128652808e-05,
+      "loss": 1.5785,
+      "step": 79000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.0361933275036685e-05,
+      "loss": 1.569,
+      "step": 79500
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.0238423421420563e-05,
+      "loss": 1.5772,
+      "step": 80000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.0114913567804444e-05,
+      "loss": 1.5652,
+      "step": 80500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.9991403714188322e-05,
+      "loss": 1.5571,
+      "step": 81000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.98678938605722e-05,
+      "loss": 1.5757,
+      "step": 81500
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.9744384006956077e-05,
+      "loss": 1.5722,
+      "step": 82000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.9620874153339955e-05,
+      "loss": 1.5799,
+      "step": 82500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.9497364299723833e-05,
+      "loss": 1.5202,
+      "step": 83000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.9373854446107714e-05,
+      "loss": 1.5105,
+      "step": 83500
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.9250344592491592e-05,
+      "loss": 1.5209,
+      "step": 84000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.912683473887547e-05,
+      "loss": 1.5088,
+      "step": 84500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.9003324885259347e-05,
+      "loss": 1.5088,
+      "step": 85000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.8879815031643225e-05,
+      "loss": 1.5258,
+      "step": 85500
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.8756305178027103e-05,
+      "loss": 1.5262,
+      "step": 86000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.8632795324410984e-05,
+      "loss": 1.5241,
+      "step": 86500
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.850928547079486e-05,
+      "loss": 1.5179,
+      "step": 87000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.838577561717874e-05,
+      "loss": 1.5123,
+      "step": 87500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.8262265763562617e-05,
+      "loss": 1.5073,
+      "step": 88000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.8138755909946495e-05,
+      "loss": 1.5279,
+      "step": 88500
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.8015246056330373e-05,
+      "loss": 1.5394,
+      "step": 89000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.7891736202714254e-05,
+      "loss": 1.5415,
+      "step": 89500
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.776822634909813e-05,
+      "loss": 1.5304,
+      "step": 90000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.764471649548201e-05,
+      "loss": 1.587,
+      "step": 90500
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.7521206641865887e-05,
+      "loss": 1.5686,
+      "step": 91000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.7397696788249765e-05,
+      "loss": 1.5662,
+      "step": 91500
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.7274186934633646e-05,
+      "loss": 1.5661,
+      "step": 92000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.7150677081017524e-05,
+      "loss": 1.5823,
+      "step": 92500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.70271672274014e-05,
+      "loss": 1.582,
+      "step": 93000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.690365737378528e-05,
+      "loss": 1.5836,
+      "step": 93500
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.6780147520169157e-05,
+      "loss": 1.5729,
+      "step": 94000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.665663766655304e-05,
+      "loss": 1.5969,
+      "step": 94500
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.653312781293692e-05,
+      "loss": 1.5878,
+      "step": 95000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.6409617959320797e-05,
+      "loss": 1.5956,
+      "step": 95500
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.6286108105704678e-05,
+      "loss": 1.5827,
+      "step": 96000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.6162598252088556e-05,
+      "loss": 1.5935,
+      "step": 96500
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6039088398472433e-05,
+      "loss": 1.5929,
+      "step": 97000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.591557854485631e-05,
+      "loss": 1.5849,
+      "step": 97500
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.579206869124019e-05,
+      "loss": 1.5819,
+      "step": 98000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.5668558837624067e-05,
+      "loss": 1.609,
+      "step": 98500
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5545048984007948e-05,
+      "loss": 1.6012,
+      "step": 99000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5421539130391825e-05,
+      "loss": 1.6138,
+      "step": 99500
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.5298029276775703e-05,
+      "loss": 1.6033,
+      "step": 100000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.517451942315958e-05,
+      "loss": 1.6153,
+      "step": 100500
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.505100956954346e-05,
+      "loss": 1.6065,
+      "step": 101000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.492749971592734e-05,
+      "loss": 1.5978,
+      "step": 101500
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.4803989862311218e-05,
+      "loss": 1.6077,
+      "step": 102000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.4680480008695095e-05,
+      "loss": 1.6067,
+      "step": 102500
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4556970155078973e-05,
+      "loss": 1.6115,
+      "step": 103000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.443346030146285e-05,
+      "loss": 1.6389,
+      "step": 103500
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.430995044784673e-05,
+      "loss": 1.6116,
+      "step": 104000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.418644059423061e-05,
+      "loss": 1.6116,
+      "step": 104500
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4062930740614487e-05,
+      "loss": 1.6344,
+      "step": 105000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.3939420886998365e-05,
+      "loss": 1.643,
+      "step": 105500
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.3815911033382243e-05,
+      "loss": 1.6408,
+      "step": 106000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.369240117976612e-05,
+      "loss": 1.6395,
+      "step": 106500
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3568891326150002e-05,
+      "loss": 1.6242,
+      "step": 107000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.344538147253388e-05,
+      "loss": 1.6318,
+      "step": 107500
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.332187161891776e-05,
+      "loss": 1.6518,
+      "step": 108000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.319836176530164e-05,
+      "loss": 1.6437,
+      "step": 108500
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3074851911685516e-05,
+      "loss": 1.6284,
+      "step": 109000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.2951342058069394e-05,
+      "loss": 1.6549,
+      "step": 109500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.282783220445327e-05,
+      "loss": 1.6254,
+      "step": 110000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2704322350837153e-05,
+      "loss": 1.633,
+      "step": 110500
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.258081249722103e-05,
+      "loss": 1.6299,
+      "step": 111000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2457302643604908e-05,
+      "loss": 1.6273,
+      "step": 111500
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.2333792789988786e-05,
+      "loss": 1.623,
+      "step": 112000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.2210282936372664e-05,
+      "loss": 1.6519,
+      "step": 112500
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.208677308275654e-05,
+      "loss": 1.6663,
+      "step": 113000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.1963263229140423e-05,
+      "loss": 1.6739,
+      "step": 113500
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.18397533755243e-05,
+      "loss": 1.6302,
+      "step": 114000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.1716243521908178e-05,
+      "loss": 1.6341,
+      "step": 114500
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.1592733668292056e-05,
+      "loss": 1.6467,
+      "step": 115000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.1469223814675937e-05,
+      "loss": 1.6351,
+      "step": 115500
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.1345713961059815e-05,
+      "loss": 1.6497,
+      "step": 116000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1222204107443692e-05,
+      "loss": 1.6473,
+      "step": 116500
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1098694253827574e-05,
+      "loss": 1.661,
+      "step": 117000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.097518440021145e-05,
+      "loss": 1.6653,
+      "step": 117500
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.085167454659533e-05,
+      "loss": 1.6754,
+      "step": 118000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.0728164692979207e-05,
+      "loss": 1.6572,
+      "step": 118500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.0604654839363085e-05,
+      "loss": 1.6597,
+      "step": 119000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.0481144985746962e-05,
+      "loss": 1.6732,
+      "step": 119500
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.0357635132130843e-05,
+      "loss": 1.684,
+      "step": 120000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.023412527851472e-05,
+      "loss": 1.6811,
+      "step": 120500
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.01106154248986e-05,
+      "loss": 1.6646,
+      "step": 121000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.9987105571282477e-05,
+      "loss": 1.649,
+      "step": 121500
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.9863595717666354e-05,
+      "loss": 1.6541,
+      "step": 122000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.9740085864050235e-05,
+      "loss": 1.679,
+      "step": 122500
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.9616576010434113e-05,
+      "loss": 1.679,
+      "step": 123000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.949306615681799e-05,
+      "loss": 1.6787,
+      "step": 123500
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.9369556303201872e-05,
+      "loss": 1.6788,
+      "step": 124000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.924604644958575e-05,
+      "loss": 1.6826,
+      "step": 124500
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9122536595969628e-05,
+      "loss": 1.6844,
+      "step": 125000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.8999026742353505e-05,
+      "loss": 1.6819,
+      "step": 125500
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.8875516888737386e-05,
+      "loss": 1.6694,
+      "step": 126000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.8752007035121264e-05,
+      "loss": 1.6982,
+      "step": 126500
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.8628497181505142e-05,
+      "loss": 1.6862,
+      "step": 127000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.850498732788902e-05,
+      "loss": 1.6608,
+      "step": 127500
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.8381477474272897e-05,
+      "loss": 1.6839,
+      "step": 128000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.8257967620656775e-05,
+      "loss": 1.675,
+      "step": 128500
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8134457767040656e-05,
+      "loss": 1.6948,
+      "step": 129000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8010947913424534e-05,
+      "loss": 1.6957,
+      "step": 129500
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7887438059808412e-05,
+      "loss": 1.7279,
+      "step": 130000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.776392820619229e-05,
+      "loss": 1.6884,
+      "step": 130500
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7640418352576167e-05,
+      "loss": 1.685,
+      "step": 131000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.751690849896005e-05,
+      "loss": 1.724,
+      "step": 131500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7393398645343926e-05,
+      "loss": 1.7142,
+      "step": 132000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7269888791727804e-05,
+      "loss": 1.7141,
+      "step": 132500
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7146378938111685e-05,
+      "loss": 1.7107,
+      "step": 133000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7022869084495563e-05,
+      "loss": 1.7284,
+      "step": 133500
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.689935923087944e-05,
+      "loss": 1.7204,
+      "step": 134000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.6775849377263318e-05,
+      "loss": 1.7406,
+      "step": 134500
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.66523395236472e-05,
+      "loss": 1.6716,
+      "step": 135000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.6528829670031077e-05,
+      "loss": 1.484,
+      "step": 135500
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6405319816414955e-05,
+      "loss": 1.4839,
+      "step": 136000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6281809962798833e-05,
+      "loss": 1.4848,
+      "step": 136500
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.615830010918271e-05,
+      "loss": 1.4659,
+      "step": 137000
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6034790255566588e-05,
+      "loss": 1.4711,
+      "step": 137500
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.591128040195047e-05,
+      "loss": 1.4763,
+      "step": 138000
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.5787770548334347e-05,
+      "loss": 1.471,
+      "step": 138500
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5664260694718225e-05,
+      "loss": 1.4713,
+      "step": 139000
+    },
     {
       "epoch": 2.07,
-      "learning_rate": 1.5417753537269785e-05,
-      "loss": 1.3663,
-      "step": 70000
+      "learning_rate": 1.5540750841102102e-05,
+      "loss": 1.4905,
+      "step": 139500
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.541724098748598e-05,
+      "loss": 1.4808,
+      "step": 140000
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.529373113386986e-05,
+      "loss": 1.4741,
+      "step": 140500
     },
     {
       "epoch": 2.09,
-      "learning_rate": 1.5170737491107422e-05,
-      "loss": 1.3622,
-      "step": 70500
+      "learning_rate": 1.5170221280253737e-05,
+      "loss": 1.4827,
+      "step": 141000
     },
     {
       "epoch": 2.1,
-      "learning_rate": 1.4923721444945064e-05,
-      "loss": 1.3698,
-      "step": 71000
+      "learning_rate": 1.5046711426637618e-05,
+      "loss": 1.4905,
+      "step": 141500
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.4923201573021498e-05,
+      "loss": 1.483,
+      "step": 142000
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4799691719405376e-05,
+      "loss": 1.475,
+      "step": 142500
     },
     {
       "epoch": 2.12,
-      "learning_rate": 1.4676705398782705e-05,
-      "loss": 1.3664,
-      "step": 71500
+      "learning_rate": 1.4676181865789253e-05,
+      "loss": 1.4909,
+      "step": 143000
     },
     {
       "epoch": 2.13,
-      "learning_rate": 1.4429689352620346e-05,
-      "loss": 1.3803,
-      "step": 72000
+      "learning_rate": 1.4552672012173133e-05,
+      "loss": 1.4882,
+      "step": 143500
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.442916215855701e-05,
+      "loss": 1.5022,
+      "step": 144000
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4305652304940888e-05,
+      "loss": 1.4856,
+      "step": 144500
     },
     {
       "epoch": 2.15,
-      "learning_rate": 1.4182673306457989e-05,
-      "loss": 1.3713,
-      "step": 72500
+      "learning_rate": 1.4182142451324768e-05,
+      "loss": 1.4868,
+      "step": 145000
     },
     {
       "epoch": 2.16,
-      "learning_rate": 1.3935657260295629e-05,
-      "loss": 1.3897,
-      "step": 73000
+      "learning_rate": 1.4058632597708645e-05,
+      "loss": 1.5068,
+      "step": 145500
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.3935122744092525e-05,
+      "loss": 1.504,
+      "step": 146000
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.3811612890476403e-05,
+      "loss": 1.5079,
+      "step": 146500
     },
     {
       "epoch": 2.18,
-      "learning_rate": 1.3688641214133272e-05,
-      "loss": 1.3869,
-      "step": 73500
+      "learning_rate": 1.368810303686028e-05,
+      "loss": 1.4974,
+      "step": 147000
     },
     {
       "epoch": 2.19,
-      "learning_rate": 1.3441625167970911e-05,
-      "loss": 1.3919,
-      "step": 74000
+      "learning_rate": 1.356459318324416e-05,
+      "loss": 1.5066,
+      "step": 147500
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3441083329628038e-05,
+      "loss": 1.5036,
+      "step": 148000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3317573476011915e-05,
+      "loss": 1.5227,
+      "step": 148500
     },
     {
       "epoch": 2.21,
-      "learning_rate": 1.3194609121808554e-05,
-      "loss": 1.4013,
-      "step": 74500
+      "learning_rate": 1.3194063622395795e-05,
+      "loss": 1.5115,
+      "step": 149000
     },
     {
       "epoch": 2.22,
-      "learning_rate": 1.2947593075646194e-05,
-      "loss": 1.3853,
-      "step": 75000
+      "learning_rate": 1.3070553768779673e-05,
+      "loss": 1.4878,
+      "step": 149500
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.2947043915163554e-05,
+      "loss": 1.5091,
+      "step": 150000
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.2823534061547431e-05,
+      "loss": 1.4812,
+      "step": 150500
     },
     {
       "epoch": 2.24,
-      "learning_rate": 1.2700577029483837e-05,
-      "loss": 1.3776,
-      "step": 75500
+      "learning_rate": 1.270002420793131e-05,
+      "loss": 1.4963,
+      "step": 151000
     },
     {
       "epoch": 2.25,
-      "learning_rate": 1.2453560983321478e-05,
-      "loss": 1.3928,
-      "step": 76000
+      "learning_rate": 1.2576514354315189e-05,
+      "loss": 1.5092,
+      "step": 151500
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2453004500699066e-05,
+      "loss": 1.5039,
+      "step": 152000
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2329494647082946e-05,
+      "loss": 1.5113,
+      "step": 152500
     },
     {
       "epoch": 2.27,
-      "learning_rate": 1.220654493715912e-05,
-      "loss": 1.4051,
-      "step": 76500
+      "learning_rate": 1.2205984793466823e-05,
+      "loss": 1.5233,
+      "step": 153000
     },
     {
       "epoch": 2.28,
-      "learning_rate": 1.195952889099676e-05,
-      "loss": 1.3934,
-      "step": 77000
+      "learning_rate": 1.2082474939850701e-05,
+      "loss": 1.5063,
+      "step": 153500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.195896508623458e-05,
+      "loss": 1.5,
+      "step": 154000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.1835455232618458e-05,
+      "loss": 1.5069,
+      "step": 154500
     },
     {
       "epoch": 2.3,
-      "learning_rate": 1.1712512844834402e-05,
-      "loss": 1.4057,
-      "step": 77500
+      "learning_rate": 1.1711945379002338e-05,
+      "loss": 1.5275,
+      "step": 155000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.1588435525386216e-05,
+      "loss": 1.5147,
+      "step": 155500
     },
     {
       "epoch": 2.31,
-      "learning_rate": 1.1465496798672043e-05,
-      "loss": 1.3998,
-      "step": 78000
+      "learning_rate": 1.1464925671770093e-05,
+      "loss": 1.5021,
+      "step": 156000
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1341415818153973e-05,
+      "loss": 1.5065,
+      "step": 156500
     },
     {
       "epoch": 2.33,
-      "learning_rate": 1.1218480752509682e-05,
-      "loss": 1.4068,
-      "step": 78500
+      "learning_rate": 1.1217905964537852e-05,
+      "loss": 1.5268,
+      "step": 157000
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.109439611092173e-05,
+      "loss": 1.512,
+      "step": 157500
     },
     {
       "epoch": 2.34,
-      "learning_rate": 1.0971464706347324e-05,
-      "loss": 1.409,
-      "step": 79000
+      "learning_rate": 1.0970886257305608e-05,
+      "loss": 1.5226,
+      "step": 158000
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.0847376403689487e-05,
+      "loss": 1.5127,
+      "step": 158500
     },
     {
       "epoch": 2.36,
-      "learning_rate": 1.0724448660184967e-05,
-      "loss": 1.395,
-      "step": 79500
+      "learning_rate": 1.0723866550073365e-05,
+      "loss": 1.4933,
+      "step": 159000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.0600356696457244e-05,
+      "loss": 1.5301,
+      "step": 159500
     },
     {
       "epoch": 2.37,
-      "learning_rate": 1.0477432614022608e-05,
-      "loss": 1.4237,
-      "step": 80000
+      "learning_rate": 1.0476846842841122e-05,
+      "loss": 1.5302,
+      "step": 160000
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0353336989225001e-05,
+      "loss": 1.5325,
+      "step": 160500
     },
     {
       "epoch": 2.39,
-      "learning_rate": 1.023041656786025e-05,
-      "loss": 1.4202,
-      "step": 80500
+      "learning_rate": 1.022982713560888e-05,
+      "loss": 1.5217,
+      "step": 161000
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0106317281992759e-05,
+      "loss": 1.5094,
+      "step": 161500
     },
     {
       "epoch": 2.4,
-      "learning_rate": 9.98340052169789e-06,
-      "loss": 1.3998,
-      "step": 81000
+      "learning_rate": 9.982807428376636e-06,
+      "loss": 1.5011,
+      "step": 162000
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.859297574760514e-06,
+      "loss": 1.523,
+      "step": 162500
     },
     {
       "epoch": 2.42,
-      "learning_rate": 9.736384475535532e-06,
-      "loss": 1.4187,
-      "step": 81500
+      "learning_rate": 9.735787721144394e-06,
+      "loss": 1.5268,
+      "step": 163000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.612277867528271e-06,
+      "loss": 1.526,
+      "step": 163500
     },
     {
       "epoch": 2.43,
-      "learning_rate": 9.489368429373173e-06,
-      "loss": 1.4312,
-      "step": 82000
+      "learning_rate": 9.488768013912149e-06,
+      "loss": 1.5427,
+      "step": 164000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.365258160296028e-06,
+      "loss": 1.5325,
+      "step": 164500
     },
     {
       "epoch": 2.45,
-      "learning_rate": 9.242352383210814e-06,
-      "loss": 1.4199,
-      "step": 82500
+      "learning_rate": 9.241748306679908e-06,
+      "loss": 1.5152,
+      "step": 165000
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.118238453063786e-06,
+      "loss": 1.5045,
+      "step": 165500
     },
     {
       "epoch": 2.46,
-      "learning_rate": 8.995336337048454e-06,
-      "loss": 1.4032,
-      "step": 83000
+      "learning_rate": 8.994728599447665e-06,
+      "loss": 1.5008,
+      "step": 166000
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.871218745831543e-06,
+      "loss": 1.515,
+      "step": 166500
     },
     {
       "epoch": 2.48,
-      "learning_rate": 8.748320290886097e-06,
-      "loss": 1.4152,
-      "step": 83500
+      "learning_rate": 8.74770889221542e-06,
+      "loss": 1.5166,
+      "step": 167000
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.6241990385993e-06,
+      "loss": 1.5336,
+      "step": 167500
     },
     {
       "epoch": 2.49,
-      "learning_rate": 8.501304244723738e-06,
-      "loss": 1.428,
-      "step": 84000
+      "learning_rate": 8.500689184983178e-06,
+      "loss": 1.527,
+      "step": 168000
     },
     {
       "epoch": 2.5,
-      "learning_rate": 8.25428819856138e-06,
-      "loss": 1.4191,
-      "step": 84500
+      "learning_rate": 8.377179331367055e-06,
+      "loss": 1.5306,
+      "step": 168500
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.253669477750937e-06,
+      "loss": 1.5094,
+      "step": 169000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.130159624134814e-06,
+      "loss": 1.5227,
+      "step": 169500
     },
     {
       "epoch": 2.52,
-      "learning_rate": 8.00727215239902e-06,
-      "loss": 1.4356,
-      "step": 85000
+      "learning_rate": 8.006649770518692e-06,
+      "loss": 1.5454,
+      "step": 170000
     },
     {
       "epoch": 2.53,
-      "learning_rate": 7.760256106236662e-06,
-      "loss": 1.443,
-      "step": 85500
+      "learning_rate": 7.883139916902572e-06,
+      "loss": 1.547,
+      "step": 170500
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.75963006328645e-06,
+      "loss": 1.5398,
+      "step": 171000
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.636120209670327e-06,
+      "loss": 1.541,
+      "step": 171500
     },
     {
       "epoch": 2.55,
-      "learning_rate": 7.513240060074303e-06,
-      "loss": 1.4406,
-      "step": 86000
+      "learning_rate": 7.5126103560542065e-06,
+      "loss": 1.5379,
+      "step": 172000
     },
     {
       "epoch": 2.56,
-      "learning_rate": 7.266224013911944e-06,
-      "loss": 1.4256,
-      "step": 86500
+      "learning_rate": 7.389100502438084e-06,
+      "loss": 1.5276,
+      "step": 172500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.265590648821963e-06,
+      "loss": 1.521,
+      "step": 173000
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.142080795205842e-06,
+      "loss": 1.5233,
+      "step": 173500
     },
     {
       "epoch": 2.58,
-      "learning_rate": 7.0192079677495855e-06,
-      "loss": 1.4408,
-      "step": 87000
+      "learning_rate": 7.018570941589721e-06,
+      "loss": 1.5517,
+      "step": 174000
     },
     {
       "epoch": 2.59,
-      "learning_rate": 6.7721919215872276e-06,
-      "loss": 1.4476,
-      "step": 87500
+      "learning_rate": 6.895061087973599e-06,
+      "loss": 1.5515,
+      "step": 174500
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 6.771551234357477e-06,
+      "loss": 1.5417,
+      "step": 175000
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.648041380741356e-06,
+      "loss": 1.5306,
+      "step": 175500
     },
     {
       "epoch": 2.61,
-      "learning_rate": 6.525175875424867e-06,
-      "loss": 1.434,
-      "step": 88000
+      "learning_rate": 6.524531527125234e-06,
+      "loss": 1.5262,
+      "step": 176000
     },
     {
       "epoch": 2.62,
-      "learning_rate": 6.278159829262508e-06,
-      "loss": 1.4378,
-      "step": 88500
+      "learning_rate": 6.401021673509112e-06,
+      "loss": 1.5289,
+      "step": 176500
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.277511819892991e-06,
+      "loss": 1.53,
+      "step": 177000
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.15400196627687e-06,
+      "loss": 1.5116,
+      "step": 177500
     },
     {
       "epoch": 2.64,
-      "learning_rate": 6.0311437831001505e-06,
-      "loss": 1.4276,
-      "step": 89000
+      "learning_rate": 6.030492112660748e-06,
+      "loss": 1.528,
+      "step": 178000
     },
     {
       "epoch": 2.65,
-      "learning_rate": 5.784127736937792e-06,
-      "loss": 1.4491,
-      "step": 89500
+      "learning_rate": 5.906982259044627e-06,
+      "loss": 1.5376,
+      "step": 178500
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.783472405428506e-06,
+      "loss": 1.5413,
+      "step": 179000
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.659962551812384e-06,
+      "loss": 1.5402,
+      "step": 179500
     },
     {
       "epoch": 2.67,
-      "learning_rate": 5.537111690775433e-06,
-      "loss": 1.4583,
-      "step": 90000
+      "learning_rate": 5.536452698196262e-06,
+      "loss": 1.56,
+      "step": 180000
     },
     {
       "epoch": 2.68,
-      "learning_rate": 5.290095644613074e-06,
-      "loss": 1.4546,
-      "step": 90500
+      "learning_rate": 5.412942844580141e-06,
+      "loss": 1.5423,
+      "step": 180500
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.289432990964019e-06,
+      "loss": 1.5453,
+      "step": 181000
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.165923137347898e-06,
+      "loss": 1.5136,
+      "step": 181500
     },
     {
       "epoch": 2.7,
-      "learning_rate": 5.043079598450716e-06,
-      "loss": 1.4338,
-      "step": 91000
+      "learning_rate": 5.0424132837317765e-06,
+      "loss": 1.5321,
+      "step": 182000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 4.918903430115654e-06,
+      "loss": 1.5518,
+      "step": 182500
     },
     {
       "epoch": 2.71,
-      "learning_rate": 4.796063552288357e-06,
-      "loss": 1.4628,
-      "step": 91500
+      "learning_rate": 4.795393576499534e-06,
+      "loss": 1.5524,
+      "step": 183000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.671883722883412e-06,
+      "loss": 1.5433,
+      "step": 183500
     },
     {
       "epoch": 2.73,
-      "learning_rate": 4.549047506125998e-06,
-      "loss": 1.4546,
-      "step": 92000
+      "learning_rate": 4.54837386926729e-06,
+      "loss": 1.5441,
+      "step": 184000
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.424864015651169e-06,
+      "loss": 1.5566,
+      "step": 184500
     },
     {
       "epoch": 2.74,
-      "learning_rate": 4.302031459963639e-06,
-      "loss": 1.449,
-      "step": 92500
+      "learning_rate": 4.301354162035048e-06,
+      "loss": 1.518,
+      "step": 185000
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.177844308418926e-06,
+      "loss": 1.5277,
+      "step": 185500
     },
     {
       "epoch": 2.76,
-      "learning_rate": 4.055015413801281e-06,
-      "loss": 1.4489,
-      "step": 93000
+      "learning_rate": 4.054334454802804e-06,
+      "loss": 1.5427,
+      "step": 186000
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.930824601186683e-06,
+      "loss": 1.5548,
+      "step": 186500
     },
     {
       "epoch": 2.77,
-      "learning_rate": 3.8079993676389218e-06,
-      "loss": 1.4632,
-      "step": 93500
+      "learning_rate": 3.807314747570561e-06,
+      "loss": 1.5419,
+      "step": 187000
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.68380489395444e-06,
+      "loss": 1.5493,
+      "step": 187500
     },
     {
       "epoch": 2.79,
-      "learning_rate": 3.560983321476563e-06,
-      "loss": 1.4739,
-      "step": 94000
+      "learning_rate": 3.5602950403383183e-06,
+      "loss": 1.5695,
+      "step": 188000
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.436785186722197e-06,
+      "loss": 1.5429,
+      "step": 188500
     },
     {
       "epoch": 2.8,
-      "learning_rate": 3.3139672753142047e-06,
-      "loss": 1.4662,
-      "step": 94500
+      "learning_rate": 3.313275333106075e-06,
+      "loss": 1.5547,
+      "step": 189000
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.189765479489954e-06,
+      "loss": 1.5306,
+      "step": 189500
     },
     {
       "epoch": 2.82,
-      "learning_rate": 3.066951229151846e-06,
-      "loss": 1.4496,
-      "step": 95000
+      "learning_rate": 3.0662556258738323e-06,
+      "loss": 1.5336,
+      "step": 190000
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.942745772257711e-06,
+      "loss": 1.5276,
+      "step": 190500
     },
     {
       "epoch": 2.83,
-      "learning_rate": 2.819935182989487e-06,
-      "loss": 1.4544,
-      "step": 95500
+      "learning_rate": 2.8192359186415894e-06,
+      "loss": 1.5453,
+      "step": 191000
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.6957260650254676e-06,
+      "loss": 1.5501,
+      "step": 191500
     },
     {
       "epoch": 2.85,
-      "learning_rate": 2.5729191368271284e-06,
-      "loss": 1.4656,
-      "step": 96000
+      "learning_rate": 2.5722162114093466e-06,
+      "loss": 1.5465,
+      "step": 192000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.448706357793225e-06,
+      "loss": 1.5485,
+      "step": 192500
     },
     {
       "epoch": 2.86,
-      "learning_rate": 2.3259030906647697e-06,
-      "loss": 1.4679,
-      "step": 96500
+      "learning_rate": 2.3251965041771034e-06,
+      "loss": 1.5467,
+      "step": 193000
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.201686650560982e-06,
+      "loss": 1.5694,
+      "step": 193500
     },
     {
       "epoch": 2.88,
-      "learning_rate": 2.078887044502411e-06,
-      "loss": 1.4869,
-      "step": 97000
+      "learning_rate": 2.0781767969448606e-06,
+      "loss": 1.5629,
+      "step": 194000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.9546669433287387e-06,
+      "loss": 1.534,
+      "step": 194500
     },
     {
       "epoch": 2.89,
-      "learning_rate": 1.8318709983400524e-06,
-      "loss": 1.4645,
-      "step": 97500
+      "learning_rate": 1.8311570897126173e-06,
+      "loss": 1.5503,
+      "step": 195000
     },
     {
       "epoch": 2.9,
-      "learning_rate": 1.5848549521776935e-06,
-      "loss": 1.4838,
-      "step": 98000
+      "learning_rate": 1.707647236096496e-06,
+      "loss": 1.5639,
+      "step": 195500
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.5841373824803743e-06,
+      "loss": 1.5585,
+      "step": 196000
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.4606275288642529e-06,
+      "loss": 1.5703,
+      "step": 196500
     },
     {
       "epoch": 2.92,
-      "learning_rate": 1.3378389060153347e-06,
-      "loss": 1.4858,
-      "step": 98500
+      "learning_rate": 1.3371176752481315e-06,
+      "loss": 1.5633,
+      "step": 197000
     },
     {
       "epoch": 2.93,
-      "learning_rate": 1.0908228598529762e-06,
-      "loss": 1.4788,
-      "step": 99000
+      "learning_rate": 1.2136078216320098e-06,
+      "loss": 1.5459,
+      "step": 197500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0900979680158882e-06,
+      "loss": 1.5609,
+      "step": 198000
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.665881143997668e-07,
+      "loss": 1.557,
+      "step": 198500
     },
     {
       "epoch": 2.95,
-      "learning_rate": 8.438068136906174e-07,
-      "loss": 1.4936,
-      "step": 99500
+      "learning_rate": 8.430782607836453e-07,
+      "loss": 1.5799,
+      "step": 199000
     },
     {
       "epoch": 2.96,
-      "learning_rate": 5.967907675282587e-07,
-      "loss": 1.4874,
-      "step": 100000
+      "learning_rate": 7.195684071675239e-07,
+      "loss": 1.5691,
+      "step": 199500
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.960585535514024e-07,
+      "loss": 1.5571,
+      "step": 200000
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.7254869993528084e-07,
+      "loss": 1.5483,
+      "step": 200500
     },
     {
       "epoch": 2.98,
-      "learning_rate": 3.4977472136589993e-07,
-      "loss": 1.4835,
-      "step": 100500
+      "learning_rate": 3.490388463191593e-07,
+      "loss": 1.5659,
+      "step": 201000
     },
     {
       "epoch": 2.99,
-      "learning_rate": 1.0275867520354124e-07,
-      "loss": 1.5014,
-      "step": 101000
+      "learning_rate": 2.2552899270303785e-07,
+      "loss": 1.5672,
+      "step": 201500
     },
     {
       "epoch": 2.99,
-      "step": 101000,
-      "total_flos": 3.751913657865339e+17,
-      "train_loss": 0.0,
-      "train_runtime": 0.125,
-      "train_samples_per_second": 3238688.305,
-      "train_steps_per_second": 404842.038
+      "learning_rate": 1.0201913908691636e-07,
+      "loss": 1.578,
+      "step": 202000
+    },
+    {
+      "epoch": 3.0,
+      "step": 202413,
+      "total_flos": 3.759621873629921e+17,
+      "train_loss": 1.602629992106659,
+      "train_runtime": 64476.8032,
+      "train_samples_per_second": 6.279,
+      "train_steps_per_second": 3.139
     }
   ],
-  "max_steps": 50604,
+  "max_steps": 202413,
   "num_train_epochs": 3,
-  "total_flos": 3.751913657865339e+17,
+  "total_flos": 3.759621873629921e+17,
   "trial_name": null,
   "trial_params": null
 }