End of training

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 4.98,
-    "eval_accuracy": 0.6914040854429856,
-    "eval_loss": 1.3205146789550781,
-    "eval_runtime": 4403.2676,
+    "epoch": 3.0,
+    "eval_accuracy": 0.7105720509450737,
+    "eval_loss": 1.2192070484161377,
+    "eval_runtime": 4409.7065,
     "eval_samples": 134942,
-    "eval_samples_per_second": 30.646,
-    "eval_steps_per_second": 7.662,
-    "perplexity": 3.7453485333545293,
-    "train_loss": 0.0,
-    "train_runtime": 0.0953,
+    "eval_samples_per_second": 30.601,
+    "eval_steps_per_second": 7.65,
+    "perplexity": 3.384502922346475,
+    "train_loss": 1.4532578269084653,
+    "train_runtime": 40248.475,
     "train_samples": 134942,
-    "train_samples_per_second": 4248275.036,
-    "train_steps_per_second": 1062084.5
+    "train_samples_per_second": 10.058,
+    "train_steps_per_second": 2.515
 }

eval_results.json

@@ -1,10 +1,10 @@
 {
-    "epoch": 4.98,
-    "eval_accuracy": 0.6914040854429856,
-    "eval_loss": 1.3205146789550781,
-    "eval_runtime": 4403.2676,
+    "epoch": 3.0,
+    "eval_accuracy": 0.7105720509450737,
+    "eval_loss": 1.2192070484161377,
+    "eval_runtime": 4409.7065,
     "eval_samples": 134942,
-    "eval_samples_per_second": 30.646,
-    "eval_steps_per_second": 7.662,
-    "perplexity": 3.7453485333545293
+    "eval_samples_per_second": 30.601,
+    "eval_steps_per_second": 7.65,
+    "perplexity": 3.384502922346475
 }

runs/Aug16_00-54-03_586b8982deeb/events.out.tfevents.1660658109.586b8982deeb.565.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0a05f4883f9811fca411ba0edfcb7d3e18a2b1ad1a59290177d3b999c23c07c
+size 369

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 4.98,
-    "train_loss": 0.0,
-    "train_runtime": 0.0953,
+    "epoch": 3.0,
+    "train_loss": 1.4532578269084653,
+    "train_runtime": 40248.475,
     "train_samples": 134942,
-    "train_samples_per_second": 4248275.036,
-    "train_steps_per_second": 1062084.5
+    "train_samples_per_second": 10.058,
+    "train_steps_per_second": 2.515
 }

trainer_state.json

@@ -1,2041 +1,1237 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.979843490633152,
-  "global_step": 168000,
+  "epoch": 3.0,
+  "global_step": 101208,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.01,
-      "learning_rate": 4.9925895186151295e-05,
-      "loss": 1.6378,
+      "learning_rate": 4.9752983953837645e-05,
+      "loss": 1.339,
       "step": 500
     },
     {
       "epoch": 0.03,
-      "learning_rate": 4.985179037230259e-05,
-      "loss": 1.6723,
+      "learning_rate": 4.950596790767528e-05,
+      "loss": 1.3788,
       "step": 1000
     },
     {
       "epoch": 0.04,
-      "learning_rate": 4.977768555845388e-05,
-      "loss": 1.6694,
+      "learning_rate": 4.9258951861512924e-05,
+      "loss": 1.378,
       "step": 1500
     },
     {
       "epoch": 0.06,
-      "learning_rate": 4.970358074460517e-05,
-      "loss": 1.6685,
+      "learning_rate": 4.901193581535057e-05,
+      "loss": 1.3805,
       "step": 2000
     },
     {
       "epoch": 0.07,
-      "learning_rate": 4.9629475930756463e-05,
-      "loss": 1.6812,
+      "learning_rate": 4.876491976918821e-05,
+      "loss": 1.3945,
       "step": 2500
     },
     {
       "epoch": 0.09,
-      "learning_rate": 4.9555371116907756e-05,
-      "loss": 1.6792,
+      "learning_rate": 4.8517903723025847e-05,
+      "loss": 1.3919,
       "step": 3000
     },
     {
       "epoch": 0.1,
-      "learning_rate": 4.948126630305905e-05,
-      "loss": 1.6873,
+      "learning_rate": 4.827088767686349e-05,
+      "loss": 1.4032,
       "step": 3500
     },
     {
       "epoch": 0.12,
-      "learning_rate": 4.940716148921034e-05,
-      "loss": 1.7036,
+      "learning_rate": 4.802387163070113e-05,
+      "loss": 1.4197,
       "step": 4000
     },
     {
       "epoch": 0.13,
-      "learning_rate": 4.866611335072326e-05,
-      "loss": 1.7003,
+      "learning_rate": 4.7776855584538775e-05,
+      "loss": 1.4222,
       "step": 4500
     },
     {
       "epoch": 0.15,
-      "learning_rate": 4.8517903723025847e-05,
-      "loss": 1.6828,
+      "learning_rate": 4.752983953837641e-05,
+      "loss": 1.4057,
       "step": 5000
     },
     {
       "epoch": 0.16,
-      "learning_rate": 4.836969409532843e-05,
-      "loss": 1.6904,
+      "learning_rate": 4.7282823492214054e-05,
+      "loss": 1.4156,
       "step": 5500
     },
     {
       "epoch": 0.18,
-      "learning_rate": 4.8221484467631015e-05,
-      "loss": 1.6975,
+      "learning_rate": 4.70358074460517e-05,
+      "loss": 1.4243,
       "step": 6000
     },
     {
       "epoch": 0.19,
-      "learning_rate": 4.80732748399336e-05,
-      "loss": 1.6904,
+      "learning_rate": 4.6788791399889334e-05,
+      "loss": 1.4176,
       "step": 6500
     },
     {
       "epoch": 0.21,
-      "learning_rate": 4.7925065212236184e-05,
-      "loss": 1.6979,
+      "learning_rate": 4.6541775353726977e-05,
+      "loss": 1.4254,
       "step": 7000
     },
     {
       "epoch": 0.22,
-      "learning_rate": 4.7776855584538775e-05,
-      "loss": 1.6774,
+      "learning_rate": 4.629475930756462e-05,
+      "loss": 1.4088,
       "step": 7500
     },
     {
       "epoch": 0.24,
-      "learning_rate": 4.762864595684136e-05,
-      "loss": 1.6976,
+      "learning_rate": 4.604774326140226e-05,
+      "loss": 1.4307,
       "step": 8000
     },
     {
       "epoch": 0.25,
-      "learning_rate": 4.7480436329143944e-05,
-      "loss": 1.715,
+      "learning_rate": 4.5800727215239905e-05,
+      "loss": 1.4459,
       "step": 8500
     },
     {
       "epoch": 0.27,
-      "learning_rate": 4.733222670144653e-05,
-      "loss": 1.7125,
+      "learning_rate": 4.555371116907755e-05,
+      "loss": 1.4486,
       "step": 9000
     },
     {
       "epoch": 0.28,
-      "learning_rate": 4.718401707374911e-05,
-      "loss": 1.7065,
+      "learning_rate": 4.530669512291519e-05,
+      "loss": 1.4432,
       "step": 9500
     },
     {
       "epoch": 0.3,
-      "learning_rate": 4.70358074460517e-05,
-      "loss": 1.7062,
+      "learning_rate": 4.505967907675283e-05,
+      "loss": 1.4432,
       "step": 10000
     },
     {
       "epoch": 0.31,
-      "learning_rate": 4.688759781835428e-05,
-      "loss": 1.7038,
+      "learning_rate": 4.481266303059047e-05,
+      "loss": 1.4424,
       "step": 10500
     },
     {
       "epoch": 0.33,
-      "learning_rate": 4.6739388190656866e-05,
-      "loss": 1.7092,
+      "learning_rate": 4.456564698442811e-05,
+      "loss": 1.4498,
       "step": 11000
     },
     {
       "epoch": 0.34,
-      "learning_rate": 4.659117856295945e-05,
-      "loss": 1.7377,
+      "learning_rate": 4.4318630938265756e-05,
+      "loss": 1.4754,
       "step": 11500
     },
     {
       "epoch": 0.36,
-      "learning_rate": 4.6442968935262035e-05,
-      "loss": 1.7306,
+      "learning_rate": 4.407161489210339e-05,
+      "loss": 1.4712,
       "step": 12000
     },
     {
       "epoch": 0.37,
-      "learning_rate": 4.629475930756462e-05,
-      "loss": 1.7128,
+      "learning_rate": 4.3824598845941035e-05,
+      "loss": 1.4552,
       "step": 12500
     },
     {
       "epoch": 0.39,
-      "learning_rate": 4.6146549679867204e-05,
-      "loss": 1.7324,
+      "learning_rate": 4.357758279977868e-05,
+      "loss": 1.4759,
       "step": 13000
     },
     {
       "epoch": 0.4,
-      "learning_rate": 4.599834005216979e-05,
-      "loss": 1.7483,
+      "learning_rate": 4.333056675361632e-05,
+      "loss": 1.4877,
       "step": 13500
     },
     {
       "epoch": 0.41,
-      "learning_rate": 4.585013042447238e-05,
-      "loss": 1.7351,
+      "learning_rate": 4.308355070745396e-05,
+      "loss": 1.4775,
       "step": 14000
     },
     {
       "epoch": 0.43,
-      "learning_rate": 4.5701920796774964e-05,
-      "loss": 1.7238,
+      "learning_rate": 4.28365346612916e-05,
+      "loss": 1.4701,
       "step": 14500
     },
     {
       "epoch": 0.44,
-      "learning_rate": 4.555371116907755e-05,
-      "loss": 1.7315,
+      "learning_rate": 4.258951861512924e-05,
+      "loss": 1.4771,
       "step": 15000
     },
     {
       "epoch": 0.46,
-      "learning_rate": 4.540550154138013e-05,
-      "loss": 1.7391,
+      "learning_rate": 4.234250256896688e-05,
+      "loss": 1.4856,
       "step": 15500
     },
     {
       "epoch": 0.47,
-      "learning_rate": 4.525729191368272e-05,
-      "loss": 1.7384,
+      "learning_rate": 4.209548652280452e-05,
+      "loss": 1.4878,
       "step": 16000
     },
     {
       "epoch": 0.49,
-      "learning_rate": 4.51090822859853e-05,
-      "loss": 1.7613,
+      "learning_rate": 4.1848470476642165e-05,
+      "loss": 1.5109,
       "step": 16500
     },
     {
       "epoch": 0.5,
-      "learning_rate": 4.4960872658287886e-05,
-      "loss": 1.7558,
+      "learning_rate": 4.160145443047981e-05,
+      "loss": 1.5056,
       "step": 17000
     },
     {
       "epoch": 0.52,
-      "learning_rate": 4.481266303059047e-05,
-      "loss": 1.7353,
+      "learning_rate": 4.1354438384317445e-05,
+      "loss": 1.4853,
       "step": 17500
     },
     {
       "epoch": 0.53,
-      "learning_rate": 4.4664453402893055e-05,
-      "loss": 1.7513,
+      "learning_rate": 4.110742233815509e-05,
+      "loss": 1.5015,
       "step": 18000
     },
     {
       "epoch": 0.55,
-      "learning_rate": 4.451624377519564e-05,
-      "loss": 1.7365,
+      "learning_rate": 4.086040629199273e-05,
+      "loss": 1.4883,
       "step": 18500
     },
     {
       "epoch": 0.56,
-      "learning_rate": 4.4368034147498224e-05,
-      "loss": 1.7428,
+      "learning_rate": 4.061339024583037e-05,
+      "loss": 1.4959,
       "step": 19000
     },
     {
       "epoch": 0.58,
-      "learning_rate": 4.421982451980081e-05,
-      "loss": 1.7352,
+      "learning_rate": 4.036637419966801e-05,
+      "loss": 1.4887,
       "step": 19500
     },
     {
       "epoch": 0.59,
-      "learning_rate": 4.407161489210339e-05,
-      "loss": 1.7561,
+      "learning_rate": 4.011935815350565e-05,
+      "loss": 1.5087,
       "step": 20000
     },
     {
       "epoch": 0.61,
-      "learning_rate": 4.392340526440598e-05,
-      "loss": 1.7478,
+      "learning_rate": 3.9872342107343295e-05,
+      "loss": 1.5022,
       "step": 20500
     },
     {
       "epoch": 0.62,
-      "learning_rate": 4.377519563670856e-05,
-      "loss": 1.7504,
+      "learning_rate": 3.962532606118093e-05,
+      "loss": 1.5063,
       "step": 21000
     },
     {
       "epoch": 0.64,
-      "learning_rate": 4.3626986009011146e-05,
-      "loss": 1.7484,
+      "learning_rate": 3.9378310015018575e-05,
+      "loss": 1.5054,
       "step": 21500
     },
     {
       "epoch": 0.65,
-      "learning_rate": 4.347877638131374e-05,
-      "loss": 1.7685,
+      "learning_rate": 3.913129396885622e-05,
+      "loss": 1.5237,
       "step": 22000
     },
     {
       "epoch": 0.67,
-      "learning_rate": 4.333056675361632e-05,
-      "loss": 1.743,
+      "learning_rate": 3.888427792269386e-05,
+      "loss": 1.4989,
       "step": 22500
     },
     {
       "epoch": 0.68,
-      "learning_rate": 4.3182357125918906e-05,
-      "loss": 1.7624,
+      "learning_rate": 3.86372618765315e-05,
+      "loss": 1.521,
       "step": 23000
     },
     {
       "epoch": 0.7,
-      "learning_rate": 4.303414749822149e-05,
-      "loss": 1.7769,
+      "learning_rate": 3.839024583036914e-05,
+      "loss": 1.5334,
       "step": 23500
     },
     {
       "epoch": 0.71,
-      "learning_rate": 4.2885937870524075e-05,
-      "loss": 1.7504,
+      "learning_rate": 3.814322978420678e-05,
+      "loss": 1.5102,
       "step": 24000
     },
     {
       "epoch": 0.73,
-      "learning_rate": 4.273772824282666e-05,
-      "loss": 1.7516,
+      "learning_rate": 3.7896213738044425e-05,
+      "loss": 1.5112,
       "step": 24500
     },
     {
       "epoch": 0.74,
-      "learning_rate": 4.258951861512924e-05,
-      "loss": 1.7763,
+      "learning_rate": 3.764919769188207e-05,
+      "loss": 1.534,
       "step": 25000
     },
     {
       "epoch": 0.76,
-      "learning_rate": 4.244130898743183e-05,
-      "loss": 1.7626,
+      "learning_rate": 3.740218164571971e-05,
+      "loss": 1.5219,
       "step": 25500
     },
     {
       "epoch": 0.77,
-      "learning_rate": 4.229309935973441e-05,
-      "loss": 1.7935,
+      "learning_rate": 3.7155165599557354e-05,
+      "loss": 1.551,
       "step": 26000
     },
     {
       "epoch": 0.79,
-      "learning_rate": 4.2144889732036997e-05,
-      "loss": 1.7821,
+      "learning_rate": 3.690814955339499e-05,
+      "loss": 1.5412,
       "step": 26500
     },
     {
       "epoch": 0.8,
-      "learning_rate": 4.199668010433958e-05,
-      "loss": 1.7785,
+      "learning_rate": 3.6661133507232633e-05,
+      "loss": 1.539,
       "step": 27000
     },
     {
       "epoch": 0.82,
-      "learning_rate": 4.1848470476642165e-05,
-      "loss": 1.7732,
+      "learning_rate": 3.6414117461070276e-05,
+      "loss": 1.5338,
       "step": 27500
     },
     {
       "epoch": 0.83,
-      "learning_rate": 4.170026084894475e-05,
-      "loss": 1.7696,
+      "learning_rate": 3.616710141490792e-05,
+      "loss": 1.531,
       "step": 28000
     },
     {
       "epoch": 0.84,
-      "learning_rate": 4.1552051221247334e-05,
-      "loss": 1.7821,
+      "learning_rate": 3.5920085368745556e-05,
+      "loss": 1.5448,
       "step": 28500
     },
     {
       "epoch": 0.86,
-      "learning_rate": 4.140384159354992e-05,
-      "loss": 1.7855,
+      "learning_rate": 3.56730693225832e-05,
+      "loss": 1.5478,
       "step": 29000
     },
     {
       "epoch": 0.87,
-      "learning_rate": 4.12556319658525e-05,
-      "loss": 1.7815,
+      "learning_rate": 3.542605327642084e-05,
+      "loss": 1.5441,
       "step": 29500
     },
     {
       "epoch": 0.89,
-      "learning_rate": 4.110742233815509e-05,
-      "loss": 1.7776,
+      "learning_rate": 3.517903723025848e-05,
+      "loss": 1.5415,
       "step": 30000
     },
     {
       "epoch": 0.9,
-      "learning_rate": 4.095921271045767e-05,
-      "loss": 1.7864,
+      "learning_rate": 3.493202118409612e-05,
+      "loss": 1.5513,
       "step": 30500
     },
     {
       "epoch": 0.92,
-      "learning_rate": 4.0811003082760256e-05,
-      "loss": 1.7894,
+      "learning_rate": 3.4685005137933763e-05,
+      "loss": 1.554,
       "step": 31000
     },
     {
       "epoch": 0.93,
-      "learning_rate": 4.066279345506284e-05,
-      "loss": 1.7879,
+      "learning_rate": 3.4437989091771406e-05,
+      "loss": 1.5526,
       "step": 31500
     },
     {
       "epoch": 0.95,
-      "learning_rate": 4.0514583827365425e-05,
-      "loss": 1.7928,
+      "learning_rate": 3.419097304560904e-05,
+      "loss": 1.5608,
       "step": 32000
     },
     {
       "epoch": 0.96,
-      "learning_rate": 4.036637419966801e-05,
-      "loss": 1.7965,
+      "learning_rate": 3.3943956999446686e-05,
+      "loss": 1.5627,
       "step": 32500
     },
     {
       "epoch": 0.98,
-      "learning_rate": 4.0218164571970594e-05,
-      "loss": 1.7906,
+      "learning_rate": 3.369694095328433e-05,
+      "loss": 1.5581,
       "step": 33000
     },
     {
       "epoch": 0.99,
-      "learning_rate": 4.006995494427318e-05,
-      "loss": 1.8053,
+      "learning_rate": 3.3449924907121965e-05,
+      "loss": 1.5707,
       "step": 33500
     },
     {
       "epoch": 1.01,
-      "learning_rate": 3.992174531657576e-05,
-      "loss": 1.6997,
+      "learning_rate": 3.320290886095961e-05,
+      "loss": 1.4586,
       "step": 34000
     },
     {
       "epoch": 1.02,
-      "learning_rate": 3.977353568887835e-05,
-      "loss": 1.627,
+      "learning_rate": 3.295589281479725e-05,
+      "loss": 1.3769,
       "step": 34500
     },
     {
       "epoch": 1.04,
-      "learning_rate": 3.962532606118093e-05,
-      "loss": 1.6205,
+      "learning_rate": 3.2708876768634893e-05,
+      "loss": 1.3723,
       "step": 35000
     },
     {
       "epoch": 1.05,
-      "learning_rate": 3.9477116433483516e-05,
-      "loss": 1.624,
+      "learning_rate": 3.246186072247253e-05,
+      "loss": 1.3776,
       "step": 35500
     },
     {
       "epoch": 1.07,
-      "learning_rate": 3.93289068057861e-05,
-      "loss": 1.6387,
+      "learning_rate": 3.221484467631017e-05,
+      "loss": 1.391,
       "step": 36000
     },
     {
       "epoch": 1.08,
-      "learning_rate": 3.9180697178088685e-05,
-      "loss": 1.632,
+      "learning_rate": 3.1967828630147816e-05,
+      "loss": 1.3848,
       "step": 36500
     },
     {
       "epoch": 1.1,
-      "learning_rate": 3.9032487550391276e-05,
-      "loss": 1.6339,
+      "learning_rate": 3.172081258398546e-05,
+      "loss": 1.3889,
       "step": 37000
     },
     {
       "epoch": 1.11,
-      "learning_rate": 3.888427792269386e-05,
-      "loss": 1.6487,
+      "learning_rate": 3.1473796537823095e-05,
+      "loss": 1.4039,
       "step": 37500
     },
     {
       "epoch": 1.13,
-      "learning_rate": 3.8736068294996445e-05,
-      "loss": 1.6449,
+      "learning_rate": 3.122678049166074e-05,
+      "loss": 1.3999,
       "step": 38000
     },
     {
       "epoch": 1.14,
-      "learning_rate": 3.858785866729903e-05,
-      "loss": 1.6598,
+      "learning_rate": 3.097976444549838e-05,
+      "loss": 1.4151,
       "step": 38500
     },
     {
       "epoch": 1.16,
-      "learning_rate": 3.8439649039601614e-05,
-      "loss": 1.6621,
+      "learning_rate": 3.073274839933602e-05,
+      "loss": 1.4157,
       "step": 39000
     },
     {
       "epoch": 1.17,
-      "learning_rate": 3.82914394119042e-05,
-      "loss": 1.6735,
+      "learning_rate": 3.0485732353173663e-05,
+      "loss": 1.4274,
       "step": 39500
     },
     {
       "epoch": 1.19,
-      "learning_rate": 3.814322978420678e-05,
-      "loss": 1.665,
+      "learning_rate": 3.0238716307011306e-05,
+      "loss": 1.4204,
       "step": 40000
     },
     {
       "epoch": 1.2,
-      "learning_rate": 3.799502015650937e-05,
-      "loss": 1.6542,
+      "learning_rate": 2.999170026084895e-05,
+      "loss": 1.4113,
       "step": 40500
     },
     {
       "epoch": 1.22,
-      "learning_rate": 3.784681052881195e-05,
-      "loss": 1.6632,
+      "learning_rate": 2.9744684214686585e-05,
+      "loss": 1.4203,
       "step": 41000
     },
     {
       "epoch": 1.23,
-      "learning_rate": 3.7698600901114536e-05,
-      "loss": 1.6096,
+      "learning_rate": 2.9497668168524228e-05,
+      "loss": 1.3833,
       "step": 41500
     },
     {
       "epoch": 1.24,
-      "learning_rate": 3.755039127341713e-05,
-      "loss": 1.5457,
+      "learning_rate": 2.925065212236187e-05,
+      "loss": 1.3336,
       "step": 42000
     },
     {
       "epoch": 1.26,
-      "learning_rate": 3.740218164571971e-05,
-      "loss": 1.538,
+      "learning_rate": 2.9003636076199514e-05,
+      "loss": 1.3271,
       "step": 42500
     },
     {
       "epoch": 1.27,
-      "learning_rate": 3.7253972018022296e-05,
-      "loss": 1.5538,
+      "learning_rate": 2.875662003003715e-05,
+      "loss": 1.3437,
       "step": 43000
     },
     {
       "epoch": 1.29,
-      "learning_rate": 3.710576239032488e-05,
-      "loss": 1.5505,
+      "learning_rate": 2.8509603983874793e-05,
+      "loss": 1.3378,
       "step": 43500
     },
     {
       "epoch": 1.3,
-      "learning_rate": 3.6957552762627465e-05,
-      "loss": 1.5386,
+      "learning_rate": 2.8262587937712436e-05,
+      "loss": 1.3275,
       "step": 44000
     },
     {
       "epoch": 1.32,
-      "learning_rate": 3.680934313493005e-05,
-      "loss": 1.5629,
+      "learning_rate": 2.8015571891550076e-05,
+      "loss": 1.352,
       "step": 44500
     },
     {
       "epoch": 1.33,
-      "learning_rate": 3.6661133507232633e-05,
-      "loss": 1.5639,
+      "learning_rate": 2.776855584538772e-05,
+      "loss": 1.3527,
       "step": 45000
     },
     {
       "epoch": 1.35,
-      "learning_rate": 3.651292387953522e-05,
-      "loss": 1.6376,
+      "learning_rate": 2.752153979922536e-05,
+      "loss": 1.4116,
       "step": 45500
     },
     {
       "epoch": 1.36,
-      "learning_rate": 3.63647142518378e-05,
-      "loss": 1.6273,
+      "learning_rate": 2.7274523753063e-05,
+      "loss": 1.3997,
       "step": 46000
     },
     {
       "epoch": 1.38,
-      "learning_rate": 3.621650462414039e-05,
-      "loss": 1.6433,
+      "learning_rate": 2.702750770690064e-05,
+      "loss": 1.4157,
       "step": 46500
     },
     {
       "epoch": 1.39,
-      "learning_rate": 3.606829499644297e-05,
-      "loss": 1.6421,
+      "learning_rate": 2.6780491660738284e-05,
+      "loss": 1.4121,
       "step": 47000
     },
     {
       "epoch": 1.41,
-      "learning_rate": 3.5920085368745556e-05,
-      "loss": 1.6529,
+      "learning_rate": 2.6533475614575926e-05,
+      "loss": 1.4252,
       "step": 47500
     },
     {
       "epoch": 1.42,
-      "learning_rate": 3.577187574104814e-05,
-      "loss": 1.6501,
+      "learning_rate": 2.6286459568413563e-05,
+      "loss": 1.4227,
       "step": 48000
     },
     {
       "epoch": 1.44,
-      "learning_rate": 3.5623666113350724e-05,
-      "loss": 1.6568,
+      "learning_rate": 2.6039443522251206e-05,
+      "loss": 1.429,
       "step": 48500
     },
     {
       "epoch": 1.45,
-      "learning_rate": 3.547545648565331e-05,
-      "loss": 1.6454,
+      "learning_rate": 2.579242747608885e-05,
+      "loss": 1.4173,
       "step": 49000
     },
     {
       "epoch": 1.47,
-      "learning_rate": 3.532724685795589e-05,
-      "loss": 1.6698,
+      "learning_rate": 2.554541142992649e-05,
+      "loss": 1.4398,
       "step": 49500
     },
     {
       "epoch": 1.48,
-      "learning_rate": 3.517903723025848e-05,
-      "loss": 1.672,
+      "learning_rate": 2.5298395383764128e-05,
+      "loss": 1.4437,
       "step": 50000
     },
     {
       "epoch": 1.5,
-      "learning_rate": 3.503082760256106e-05,
-      "loss": 1.6761,
+      "learning_rate": 2.505137933760177e-05,
+      "loss": 1.4467,
       "step": 50500
     },
     {
       "epoch": 1.51,
-      "learning_rate": 3.4882617974863646e-05,
-      "loss": 1.6686,
+      "learning_rate": 2.4804363291439414e-05,
+      "loss": 1.4396,
       "step": 51000
     },
     {
       "epoch": 1.53,
-      "learning_rate": 3.473440834716624e-05,
-      "loss": 1.6736,
+      "learning_rate": 2.4557347245277053e-05,
+      "loss": 1.444,
       "step": 51500
     },
     {
       "epoch": 1.54,
-      "learning_rate": 3.458619871946882e-05,
-      "loss": 1.6939,
+      "learning_rate": 2.4310331199114696e-05,
+      "loss": 1.4635,
       "step": 52000
     },
     {
       "epoch": 1.56,
-      "learning_rate": 3.4437989091771406e-05,
-      "loss": 1.6905,
+      "learning_rate": 2.4063315152952336e-05,
+      "loss": 1.4614,
       "step": 52500
     },
     {
       "epoch": 1.57,
-      "learning_rate": 3.428977946407399e-05,
-      "loss": 1.7088,
+      "learning_rate": 2.381629910678998e-05,
+      "loss": 1.4783,
       "step": 53000
     },
     {
       "epoch": 1.59,
-      "learning_rate": 3.4141569836376575e-05,
-      "loss": 1.7015,
+      "learning_rate": 2.356928306062762e-05,
+      "loss": 1.4699,
       "step": 53500
     },
     {
       "epoch": 1.6,
-      "learning_rate": 3.399336020867916e-05,
-      "loss": 1.7107,
+      "learning_rate": 2.332226701446526e-05,
+      "loss": 1.4788,
       "step": 54000
     },
     {
       "epoch": 1.62,
-      "learning_rate": 3.3845150580981744e-05,
-      "loss": 1.705,
+      "learning_rate": 2.3075250968302904e-05,
+      "loss": 1.4732,
       "step": 54500
     },
     {
       "epoch": 1.63,
-      "learning_rate": 3.369694095328433e-05,
-      "loss": 1.71,
+      "learning_rate": 2.2828234922140544e-05,
+      "loss": 1.4794,
       "step": 55000
     },
     {
       "epoch": 1.65,
-      "learning_rate": 3.354873132558691e-05,
-      "loss": 1.7015,
+      "learning_rate": 2.2581218875978187e-05,
+      "loss": 1.4708,
       "step": 55500
     },
     {
       "epoch": 1.66,
-      "learning_rate": 3.34005216978895e-05,
-      "loss": 1.6962,
+      "learning_rate": 2.2334202829815826e-05,
+      "loss": 1.4666,
       "step": 56000
     },
     {
       "epoch": 1.67,
-      "learning_rate": 3.325231207019208e-05,
-      "loss": 1.7304,
+      "learning_rate": 2.208718678365347e-05,
+      "loss": 1.4993,
       "step": 56500
     },
     {
       "epoch": 1.69,
-      "learning_rate": 3.3104102442494666e-05,
-      "loss": 1.7234,
+      "learning_rate": 2.184017073749111e-05,
+      "loss": 1.4921,
       "step": 57000
     },
     {
       "epoch": 1.7,
-      "learning_rate": 3.295589281479725e-05,
-      "loss": 1.7124,
+      "learning_rate": 2.1593154691328748e-05,
+      "loss": 1.482,
       "step": 57500
     },
     {
       "epoch": 1.72,
-      "learning_rate": 3.2807683187099835e-05,
-      "loss": 1.7161,
+      "learning_rate": 2.134613864516639e-05,
+      "loss": 1.4862,
       "step": 58000
     },
     {
       "epoch": 1.73,
-      "learning_rate": 3.265947355940242e-05,
-      "loss": 1.7298,
+      "learning_rate": 2.109912259900403e-05,
+      "loss": 1.4985,
       "step": 58500
     },
     {
       "epoch": 1.75,
-      "learning_rate": 3.2511263931705004e-05,
-      "loss": 1.7463,
+      "learning_rate": 2.0852106552841674e-05,
+      "loss": 1.5143,
       "step": 59000
     },
     {
       "epoch": 1.76,
-      "learning_rate": 3.236305430400759e-05,
-      "loss": 1.7358,
+      "learning_rate": 2.0605090506679313e-05,
+      "loss": 1.5043,
       "step": 59500
     },
     {
       "epoch": 1.78,
-      "learning_rate": 3.221484467631017e-05,
-      "loss": 1.7545,
+      "learning_rate": 2.0358074460516956e-05,
+      "loss": 1.5227,
       "step": 60000
     },
     {
       "epoch": 1.79,
-      "learning_rate": 3.206663504861276e-05,
-      "loss": 1.7504,
+      "learning_rate": 2.0111058414354596e-05,
+      "loss": 1.5202,
       "step": 60500
     },
     {
       "epoch": 1.81,
-      "learning_rate": 3.191842542091534e-05,
-      "loss": 1.7291,
+      "learning_rate": 1.986404236819224e-05,
+      "loss": 1.4978,
       "step": 61000
     },
     {
       "epoch": 1.82,
-      "learning_rate": 3.1770215793217926e-05,
-      "loss": 1.7584,
+      "learning_rate": 1.961702632202988e-05,
+      "loss": 1.5271,
       "step": 61500
     },
     {
       "epoch": 1.84,
-      "learning_rate": 3.162200616552051e-05,
-      "loss": 1.7563,
+      "learning_rate": 1.937001027586752e-05,
+      "loss": 1.5253,
       "step": 62000
     },
     {
       "epoch": 1.85,
-      "learning_rate": 3.1473796537823095e-05,
-      "loss": 1.7621,
+      "learning_rate": 1.9122994229705164e-05,
+      "loss": 1.5317,
       "step": 62500
     },
     {
       "epoch": 1.87,
-      "learning_rate": 3.132558691012568e-05,
-      "loss": 1.7567,
+      "learning_rate": 1.8875978183542804e-05,
+      "loss": 1.5253,
       "step": 63000
     },
     {
       "epoch": 1.88,
-      "learning_rate": 3.1177377282428263e-05,
-      "loss": 1.7759,
+      "learning_rate": 1.8628962137380447e-05,
+      "loss": 1.5434,
       "step": 63500
     },
     {
       "epoch": 1.9,
-      "learning_rate": 3.102916765473085e-05,
-      "loss": 1.7544,
+      "learning_rate": 1.8381946091218086e-05,
+      "loss": 1.5219,
       "step": 64000
     },
     {
       "epoch": 1.91,
-      "learning_rate": 3.088095802703343e-05,
-      "loss": 1.7663,
+      "learning_rate": 1.813493004505573e-05,
+      "loss": 1.537,
       "step": 64500
     },
     {
       "epoch": 1.93,
-      "learning_rate": 3.073274839933602e-05,
-      "loss": 1.7943,
+      "learning_rate": 1.788791399889337e-05,
+      "loss": 1.5622,
       "step": 65000
     },
     {
       "epoch": 1.94,
-      "learning_rate": 3.05845387716386e-05,
-      "loss": 1.7706,
+      "learning_rate": 1.764089795273101e-05,
+      "loss": 1.5391,
       "step": 65500
     },
     {
       "epoch": 1.96,
-      "learning_rate": 3.0436329143941196e-05,
-      "loss": 1.8033,
+      "learning_rate": 1.739388190656865e-05,
+      "loss": 1.573,
       "step": 66000
     },
     {
       "epoch": 1.97,
-      "learning_rate": 3.028811951624378e-05,
-      "loss": 1.7962,
+      "learning_rate": 1.714686586040629e-05,
+      "loss": 1.566,
       "step": 66500
     },
     {
       "epoch": 1.99,
-      "learning_rate": 3.0139909888546365e-05,
-      "loss": 1.8109,
+      "learning_rate": 1.6899849814243934e-05,
+      "loss": 1.5783,
       "step": 67000
     },
     {
       "epoch": 2.0,
-      "learning_rate": 2.999170026084895e-05,
-      "loss": 1.7925,
+      "learning_rate": 1.6652833768081573e-05,
+      "loss": 1.5639,
       "step": 67500
     },
     {
       "epoch": 2.02,
-      "learning_rate": 2.9843490633151533e-05,
-      "loss": 1.5786,
+      "learning_rate": 1.6405817721919216e-05,
+      "loss": 1.3644,
       "step": 68000
     },
     {
       "epoch": 2.03,
-      "learning_rate": 2.9695281005454118e-05,
-      "loss": 1.5676,
+      "learning_rate": 1.6158801675756856e-05,
+      "loss": 1.3529,
       "step": 68500
     },
     {
       "epoch": 2.05,
-      "learning_rate": 2.9547071377756702e-05,
-      "loss": 1.5703,
+      "learning_rate": 1.59117856295945e-05,
+      "loss": 1.3552,
       "step": 69000
     },
     {
       "epoch": 2.06,
-      "learning_rate": 2.9398861750059287e-05,
-      "loss": 1.566,
+      "learning_rate": 1.566476958343214e-05,
+      "loss": 1.3508,
       "step": 69500
     },
     {
       "epoch": 2.07,
-      "learning_rate": 2.925065212236187e-05,
-      "loss": 1.5811,
+      "learning_rate": 1.5417753537269785e-05,
+      "loss": 1.3663,
       "step": 70000
     },
     {
       "epoch": 2.09,
-      "learning_rate": 2.9102442494664455e-05,
-      "loss": 1.5771,
+      "learning_rate": 1.5170737491107422e-05,
+      "loss": 1.3622,
       "step": 70500
     },
     {
       "epoch": 2.1,
-      "learning_rate": 2.895423286696704e-05,
-      "loss": 1.5829,
+      "learning_rate": 1.4923721444945064e-05,
+      "loss": 1.3698,
       "step": 71000
     },
     {
       "epoch": 2.12,
-      "learning_rate": 2.8806023239269624e-05,
-      "loss": 1.5809,
+      "learning_rate": 1.4676705398782705e-05,
+      "loss": 1.3664,
       "step": 71500
     },
     {
       "epoch": 2.13,
-      "learning_rate": 2.865781361157221e-05,
-      "loss": 1.5957,
+      "learning_rate": 1.4429689352620346e-05,
+      "loss": 1.3803,
       "step": 72000
     },
     {
       "epoch": 2.15,
-      "learning_rate": 2.8509603983874793e-05,
-      "loss": 1.5873,
+      "learning_rate": 1.4182673306457989e-05,
+      "loss": 1.3713,
       "step": 72500
     },
     {
       "epoch": 2.16,
-      "learning_rate": 2.8361394356177377e-05,
-      "loss": 1.6051,
+      "learning_rate": 1.3935657260295629e-05,
+      "loss": 1.3897,
       "step": 73000
     },
     {
       "epoch": 2.18,
-      "learning_rate": 2.8213184728479962e-05,
-      "loss": 1.6021,
+      "learning_rate": 1.3688641214133272e-05,
+      "loss": 1.3869,
       "step": 73500
     },
     {
       "epoch": 2.19,
-      "learning_rate": 2.8064975100782546e-05,
-      "loss": 1.6071,
+      "learning_rate": 1.3441625167970911e-05,
+      "loss": 1.3919,
       "step": 74000
     },
     {
       "epoch": 2.21,
-      "learning_rate": 2.7916765473085134e-05,
-      "loss": 1.6173,
+      "learning_rate": 1.3194609121808554e-05,
+      "loss": 1.4013,
       "step": 74500
     },
     {
       "epoch": 2.22,
-      "learning_rate": 2.776855584538772e-05,
-      "loss": 1.6023,
+      "learning_rate": 1.2947593075646194e-05,
+      "loss": 1.3853,
       "step": 75000
     },
     {
       "epoch": 2.24,
-      "learning_rate": 2.7620346217690303e-05,
-      "loss": 1.5913,
+      "learning_rate": 1.2700577029483837e-05,
+      "loss": 1.3776,
       "step": 75500
     },
     {
       "epoch": 2.25,
-      "learning_rate": 2.7472136589992887e-05,
-      "loss": 1.6075,
+      "learning_rate": 1.2453560983321478e-05,
+      "loss": 1.3928,
       "step": 76000
     },
     {
       "epoch": 2.27,
-      "learning_rate": 2.7323926962295472e-05,
-      "loss": 1.6212,
+      "learning_rate": 1.220654493715912e-05,
+      "loss": 1.4051,
       "step": 76500
     },
     {
       "epoch": 2.28,
-      "learning_rate": 2.7175717334598056e-05,
-      "loss": 1.6085,
+      "learning_rate": 1.195952889099676e-05,
+      "loss": 1.3934,
       "step": 77000
     },
     {
       "epoch": 2.3,
-      "learning_rate": 2.702750770690064e-05,
-      "loss": 1.6209,
+      "learning_rate": 1.1712512844834402e-05,
+      "loss": 1.4057,
       "step": 77500
     },
     {
       "epoch": 2.31,
-      "learning_rate": 2.6879298079203225e-05,
-      "loss": 1.6171,
+      "learning_rate": 1.1465496798672043e-05,
+      "loss": 1.3998,
       "step": 78000
     },
     {
       "epoch": 2.33,
-      "learning_rate": 2.673108845150581e-05,
-      "loss": 1.6225,
+      "learning_rate": 1.1218480752509682e-05,
+      "loss": 1.4068,
       "step": 78500
     },
     {
       "epoch": 2.34,
-      "learning_rate": 2.6582878823808394e-05,
-      "loss": 1.6236,
+      "learning_rate": 1.0971464706347324e-05,
+      "loss": 1.409,
       "step": 79000
     },
     {
       "epoch": 2.36,
-      "learning_rate": 2.6434669196110978e-05,
-      "loss": 1.6109,
+      "learning_rate": 1.0724448660184967e-05,
+      "loss": 1.395,
       "step": 79500
     },
     {
       "epoch": 2.37,
-      "learning_rate": 2.6286459568413563e-05,
-      "loss": 1.6405,
+      "learning_rate": 1.0477432614022608e-05,
+      "loss": 1.4237,
       "step": 80000
     },
     {
       "epoch": 2.39,
-      "learning_rate": 2.6138249940716147e-05,
-      "loss": 1.6367,
+      "learning_rate": 1.023041656786025e-05,
+      "loss": 1.4202,
       "step": 80500
     },
     {
       "epoch": 2.4,
-      "learning_rate": 2.5990040313018738e-05,
-      "loss": 1.6158,
+      "learning_rate": 9.98340052169789e-06,
+      "loss": 1.3998,
       "step": 81000
     },
     {
       "epoch": 2.42,
-      "learning_rate": 2.5841830685321323e-05,
-      "loss": 1.6349,
+      "learning_rate": 9.736384475535532e-06,
+      "loss": 1.4187,
       "step": 81500
     },
     {
       "epoch": 2.43,
-      "learning_rate": 2.5693621057623907e-05,
-      "loss": 1.6462,
+      "learning_rate": 9.489368429373173e-06,
+      "loss": 1.4312,
       "step": 82000
     },
     {
       "epoch": 2.45,
-      "learning_rate": 2.554541142992649e-05,
-      "loss": 1.6367,
+      "learning_rate": 9.242352383210814e-06,
+      "loss": 1.4199,
       "step": 82500
     },
     {
       "epoch": 2.46,
-      "learning_rate": 2.5397201802229076e-05,
-      "loss": 1.6175,
+      "learning_rate": 8.995336337048454e-06,
+      "loss": 1.4032,
       "step": 83000
     },
     {
       "epoch": 2.48,
-      "learning_rate": 2.524899217453166e-05,
-      "loss": 1.6294,
+      "learning_rate": 8.748320290886097e-06,
+      "loss": 1.4152,
       "step": 83500
     },
     {
       "epoch": 2.49,
-      "learning_rate": 2.5100782546834245e-05,
-      "loss": 1.6447,
+      "learning_rate": 8.501304244723738e-06,
+      "loss": 1.428,
       "step": 84000
     },
     {
       "epoch": 2.5,
-      "learning_rate": 2.495257291913683e-05,
-      "loss": 1.6362,
+      "learning_rate": 8.25428819856138e-06,
+      "loss": 1.4191,
       "step": 84500
     },
     {
       "epoch": 2.52,
-      "learning_rate": 2.4804363291439414e-05,
-      "loss": 1.6515,
+      "learning_rate": 8.00727215239902e-06,
+      "loss": 1.4356,
       "step": 85000
     },
     {
       "epoch": 2.53,
-      "learning_rate": 2.4656153663741998e-05,
-      "loss": 1.6596,
+      "learning_rate": 7.760256106236662e-06,
+      "loss": 1.443,
       "step": 85500
     },
     {
       "epoch": 2.55,
-      "learning_rate": 2.4507944036044582e-05,
-      "loss": 1.6572,
+      "learning_rate": 7.513240060074303e-06,
+      "loss": 1.4406,
       "step": 86000
     },
     {
       "epoch": 2.56,
-      "learning_rate": 2.4359734408347167e-05,
-      "loss": 1.6417,
+      "learning_rate": 7.266224013911944e-06,
+      "loss": 1.4256,
       "step": 86500
     },
     {
       "epoch": 2.58,
-      "learning_rate": 2.421152478064975e-05,
-      "loss": 1.656,
+      "learning_rate": 7.0192079677495855e-06,
+      "loss": 1.4408,
       "step": 87000
     },
     {
       "epoch": 2.59,
-      "learning_rate": 2.4063315152952336e-05,
-      "loss": 1.6642,
+      "learning_rate": 6.7721919215872276e-06,
+      "loss": 1.4476,
       "step": 87500
     },
     {
       "epoch": 2.61,
-      "learning_rate": 2.391510552525492e-05,
-      "loss": 1.6481,
+      "learning_rate": 6.525175875424867e-06,
+      "loss": 1.434,
       "step": 88000
     },
     {
       "epoch": 2.62,
-      "learning_rate": 2.3766895897557508e-05,
-      "loss": 1.6528,
+      "learning_rate": 6.278159829262508e-06,
+      "loss": 1.4378,
       "step": 88500
     },
     {
       "epoch": 2.64,
-      "learning_rate": 2.3618686269860092e-05,
-      "loss": 1.64,
+      "learning_rate": 6.0311437831001505e-06,
+      "loss": 1.4276,
       "step": 89000
     },
     {
       "epoch": 2.65,
-      "learning_rate": 2.3470476642162677e-05,
-      "loss": 1.6621,
+      "learning_rate": 5.784127736937792e-06,
+      "loss": 1.4491,
       "step": 89500
     },
     {
       "epoch": 2.67,
-      "learning_rate": 2.332226701446526e-05,
-      "loss": 1.6731,
+      "learning_rate": 5.537111690775433e-06,
+      "loss": 1.4583,
       "step": 90000
     },
     {
       "epoch": 2.68,
-      "learning_rate": 2.3174057386767846e-05,
-      "loss": 1.6672,
+      "learning_rate": 5.290095644613074e-06,
+      "loss": 1.4546,
       "step": 90500
     },
     {
       "epoch": 2.7,
-      "learning_rate": 2.302584775907043e-05,
-      "loss": 1.6476,
+      "learning_rate": 5.043079598450716e-06,
+      "loss": 1.4338,
       "step": 91000
     },
     {
       "epoch": 2.71,
-      "learning_rate": 2.2877638131373014e-05,
-      "loss": 1.6773,
+      "learning_rate": 4.796063552288357e-06,
+      "loss": 1.4628,
       "step": 91500
     },
     {
       "epoch": 2.73,
-      "learning_rate": 2.2729428503675602e-05,
-      "loss": 1.6671,
+      "learning_rate": 4.549047506125998e-06,
+      "loss": 1.4546,
       "step": 92000
     },
     {
       "epoch": 2.74,
-      "learning_rate": 2.2581218875978187e-05,
-      "loss": 1.6623,
+      "learning_rate": 4.302031459963639e-06,
+      "loss": 1.449,
       "step": 92500
     },
     {
       "epoch": 2.76,
-      "learning_rate": 2.243300924828077e-05,
-      "loss": 1.6602,
+      "learning_rate": 4.055015413801281e-06,
+      "loss": 1.4489,
       "step": 93000
     },
     {
       "epoch": 2.77,
-      "learning_rate": 2.2284799620583355e-05,
-      "loss": 1.6765,
+      "learning_rate": 3.8079993676389218e-06,
+      "loss": 1.4632,
       "step": 93500
     },
     {
       "epoch": 2.79,
-      "learning_rate": 2.213658999288594e-05,
-      "loss": 1.6874,
+      "learning_rate": 3.560983321476563e-06,
+      "loss": 1.4739,
       "step": 94000
     },
     {
       "epoch": 2.8,
-      "learning_rate": 2.1988380365188524e-05,
-      "loss": 1.6777,
+      "learning_rate": 3.3139672753142047e-06,
+      "loss": 1.4662,
       "step": 94500
     },
     {
       "epoch": 2.82,
-      "learning_rate": 2.184017073749111e-05,
-      "loss": 1.6582,
+      "learning_rate": 3.066951229151846e-06,
+      "loss": 1.4496,
       "step": 95000
     },
     {
       "epoch": 2.83,
-      "learning_rate": 2.1691961109793693e-05,
-      "loss": 1.6663,
+      "learning_rate": 2.819935182989487e-06,
+      "loss": 1.4544,
       "step": 95500
     },
     {
       "epoch": 2.85,
-      "learning_rate": 2.1543751482096277e-05,
-      "loss": 1.6773,
+      "learning_rate": 2.5729191368271284e-06,
+      "loss": 1.4656,
       "step": 96000
     },
     {
       "epoch": 2.86,
-      "learning_rate": 2.1395541854398862e-05,
-      "loss": 1.6781,
+      "learning_rate": 2.3259030906647697e-06,
+      "loss": 1.4679,
       "step": 96500
     },
     {
       "epoch": 2.88,
-      "learning_rate": 2.1247332226701446e-05,
-      "loss": 1.6983,
+      "learning_rate": 2.078887044502411e-06,
+      "loss": 1.4869,
       "step": 97000
     },
     {
       "epoch": 2.89,
-      "learning_rate": 2.109912259900403e-05,
-      "loss": 1.6733,
+      "learning_rate": 1.8318709983400524e-06,
+      "loss": 1.4645,
       "step": 97500
     },
     {
       "epoch": 2.9,
-      "learning_rate": 2.0950912971306615e-05,
-      "loss": 1.6931,
+      "learning_rate": 1.5848549521776935e-06,
+      "loss": 1.4838,
       "step": 98000
     },
     {
       "epoch": 2.92,
-      "learning_rate": 2.08027033436092e-05,
-      "loss": 1.6948,
+      "learning_rate": 1.3378389060153347e-06,
+      "loss": 1.4858,
       "step": 98500
     },
     {
       "epoch": 2.93,
-      "learning_rate": 2.0654493715911787e-05,
-      "loss": 1.6866,
+      "learning_rate": 1.0908228598529762e-06,
+      "loss": 1.4788,
       "step": 99000
     },
     {
       "epoch": 2.95,
-      "learning_rate": 2.0506284088214372e-05,
-      "loss": 1.7013,
+      "learning_rate": 8.438068136906174e-07,
+      "loss": 1.4936,
       "step": 99500
     },
     {
       "epoch": 2.96,
-      "learning_rate": 2.0358074460516956e-05,
-      "loss": 1.6963,
+      "learning_rate": 5.967907675282587e-07,
+      "loss": 1.4874,
       "step": 100000
     },
     {
       "epoch": 2.98,
-      "learning_rate": 2.020986483281954e-05,
-      "loss": 1.6904,
+      "learning_rate": 3.4977472136589993e-07,
+      "loss": 1.4835,
       "step": 100500
     },
     {
       "epoch": 2.99,
-      "learning_rate": 2.0061655205122125e-05,
-      "loss": 1.7072,
+      "learning_rate": 1.0275867520354124e-07,
+      "loss": 1.5014,
       "step": 101000
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.991344557742471e-05,
-      "loss": 1.6192,
-      "step": 101500
-    },
-    {
-      "epoch": 3.02,
-      "learning_rate": 1.9765235949727294e-05,
-      "loss": 1.5539,
-      "step": 102000
-    },
-    {
-      "epoch": 3.04,
-      "learning_rate": 1.961702632202988e-05,
-      "loss": 1.5617,
-      "step": 102500
-    },
-    {
-      "epoch": 3.05,
-      "learning_rate": 1.9468816694332466e-05,
-      "loss": 1.5699,
-      "step": 103000
-    },
-    {
-      "epoch": 3.07,
-      "learning_rate": 1.932060706663505e-05,
-      "loss": 1.5772,
-      "step": 103500
-    },
-    {
-      "epoch": 3.08,
-      "learning_rate": 1.9172397438937635e-05,
-      "loss": 1.5697,
-      "step": 104000
-    },
-    {
-      "epoch": 3.1,
-      "learning_rate": 1.902418781124022e-05,
-      "loss": 1.5681,
-      "step": 104500
-    },
-    {
-      "epoch": 3.11,
-      "learning_rate": 1.8875978183542804e-05,
-      "loss": 1.5799,
-      "step": 105000
-    },
-    {
-      "epoch": 3.13,
-      "learning_rate": 1.8727768555845388e-05,
-      "loss": 1.5845,
-      "step": 105500
-    },
-    {
-      "epoch": 3.14,
-      "learning_rate": 1.8579558928147972e-05,
-      "loss": 1.5768,
-      "step": 106000
-    },
-    {
-      "epoch": 3.16,
-      "learning_rate": 1.843134930045056e-05,
-      "loss": 1.5776,
-      "step": 106500
-    },
-    {
-      "epoch": 3.17,
-      "learning_rate": 1.8283139672753145e-05,
-      "loss": 1.5728,
-      "step": 107000
-    },
-    {
-      "epoch": 3.19,
-      "learning_rate": 1.813493004505573e-05,
-      "loss": 1.5831,
-      "step": 107500
-    },
-    {
-      "epoch": 3.2,
-      "learning_rate": 1.7986720417358314e-05,
-      "loss": 1.5911,
-      "step": 108000
-    },
-    {
-      "epoch": 3.22,
-      "learning_rate": 1.7838510789660898e-05,
-      "loss": 1.5932,
-      "step": 108500
-    },
-    {
-      "epoch": 3.23,
-      "learning_rate": 1.7690301161963482e-05,
-      "loss": 1.5702,
-      "step": 109000
-    },
-    {
-      "epoch": 3.25,
-      "learning_rate": 1.7542091534266067e-05,
-      "loss": 1.5901,
-      "step": 109500
-    },
-    {
-      "epoch": 3.26,
-      "learning_rate": 1.739388190656865e-05,
-      "loss": 1.594,
-      "step": 110000
-    },
-    {
-      "epoch": 3.28,
-      "learning_rate": 1.7245672278871236e-05,
-      "loss": 1.5922,
-      "step": 110500
-    },
-    {
-      "epoch": 3.29,
-      "learning_rate": 1.709746265117382e-05,
-      "loss": 1.5669,
-      "step": 111000
-    },
-    {
-      "epoch": 3.31,
-      "learning_rate": 1.6949253023476404e-05,
-      "loss": 1.5945,
-      "step": 111500
-    },
-    {
-      "epoch": 3.32,
-      "learning_rate": 1.680104339577899e-05,
-      "loss": 1.5999,
-      "step": 112000
-    },
-    {
-      "epoch": 3.33,
-      "learning_rate": 1.6652833768081573e-05,
-      "loss": 1.5991,
-      "step": 112500
-    },
-    {
-      "epoch": 3.35,
-      "learning_rate": 1.6504624140384158e-05,
-      "loss": 1.5865,
-      "step": 113000
-    },
-    {
-      "epoch": 3.36,
-      "learning_rate": 1.6356414512686745e-05,
-      "loss": 1.598,
-      "step": 113500
-    },
-    {
-      "epoch": 3.38,
-      "learning_rate": 1.620820488498933e-05,
-      "loss": 1.5958,
-      "step": 114000
-    },
-    {
-      "epoch": 3.39,
-      "learning_rate": 1.6059995257291914e-05,
-      "loss": 1.5898,
-      "step": 114500
-    },
-    {
-      "epoch": 3.41,
-      "learning_rate": 1.59117856295945e-05,
-      "loss": 1.602,
-      "step": 115000
-    },
-    {
-      "epoch": 3.42,
-      "learning_rate": 1.5763576001897083e-05,
-      "loss": 1.5987,
-      "step": 115500
-    },
-    {
-      "epoch": 3.44,
-      "learning_rate": 1.5615366374199667e-05,
-      "loss": 1.5998,
-      "step": 116000
-    },
-    {
-      "epoch": 3.45,
-      "learning_rate": 1.5467156746502255e-05,
-      "loss": 1.6131,
-      "step": 116500
-    },
-    {
-      "epoch": 3.47,
-      "learning_rate": 1.531894711880484e-05,
-      "loss": 1.6055,
-      "step": 117000
-    },
-    {
-      "epoch": 3.48,
-      "learning_rate": 1.5170737491107422e-05,
-      "loss": 1.6107,
-      "step": 117500
-    },
-    {
-      "epoch": 3.5,
-      "learning_rate": 1.5022527863410007e-05,
-      "loss": 1.5989,
-      "step": 118000
-    },
-    {
-      "epoch": 3.51,
-      "learning_rate": 1.4874318235712591e-05,
-      "loss": 1.6009,
-      "step": 118500
-    },
-    {
-      "epoch": 3.53,
-      "learning_rate": 1.4726108608015177e-05,
-      "loss": 1.601,
-      "step": 119000
-    },
-    {
-      "epoch": 3.54,
-      "learning_rate": 1.4577898980317762e-05,
-      "loss": 1.6077,
-      "step": 119500
-    },
-    {
-      "epoch": 3.56,
-      "learning_rate": 1.4429689352620346e-05,
-      "loss": 1.5987,
-      "step": 120000
-    },
-    {
-      "epoch": 3.57,
-      "learning_rate": 1.428147972492293e-05,
-      "loss": 1.6148,
-      "step": 120500
-    },
-    {
-      "epoch": 3.59,
-      "learning_rate": 1.4133270097225518e-05,
-      "loss": 1.6126,
-      "step": 121000
-    },
-    {
-      "epoch": 3.6,
-      "learning_rate": 1.3985060469528103e-05,
-      "loss": 1.6129,
-      "step": 121500
-    },
-    {
-      "epoch": 3.62,
-      "learning_rate": 1.3836850841830687e-05,
-      "loss": 1.6093,
-      "step": 122000
-    },
-    {
-      "epoch": 3.63,
-      "learning_rate": 1.3688641214133272e-05,
-      "loss": 1.6098,
-      "step": 122500
-    },
-    {
-      "epoch": 3.65,
-      "learning_rate": 1.3540431586435856e-05,
-      "loss": 1.6178,
-      "step": 123000
-    },
-    {
-      "epoch": 3.66,
-      "learning_rate": 1.339222195873844e-05,
-      "loss": 1.6175,
-      "step": 123500
-    },
-    {
-      "epoch": 3.68,
-      "learning_rate": 1.3244012331041025e-05,
-      "loss": 1.6113,
-      "step": 124000
-    },
-    {
-      "epoch": 3.69,
-      "learning_rate": 1.309580270334361e-05,
-      "loss": 1.6271,
-      "step": 124500
-    },
-    {
-      "epoch": 3.71,
-      "learning_rate": 1.2947593075646194e-05,
-      "loss": 1.6193,
-      "step": 125000
-    },
-    {
-      "epoch": 3.72,
-      "learning_rate": 1.279938344794878e-05,
-      "loss": 1.621,
-      "step": 125500
-    },
-    {
-      "epoch": 3.73,
-      "learning_rate": 1.2651173820251364e-05,
-      "loss": 1.6174,
-      "step": 126000
-    },
-    {
-      "epoch": 3.75,
-      "learning_rate": 1.2502964192553949e-05,
-      "loss": 1.6252,
-      "step": 126500
-    },
-    {
-      "epoch": 3.76,
-      "learning_rate": 1.2354754564856535e-05,
-      "loss": 1.6591,
-      "step": 127000
-    },
-    {
-      "epoch": 3.78,
-      "learning_rate": 1.220654493715912e-05,
-      "loss": 1.6134,
-      "step": 127500
-    },
-    {
-      "epoch": 3.79,
-      "learning_rate": 1.2058335309461704e-05,
-      "loss": 1.6309,
-      "step": 128000
-    },
-    {
-      "epoch": 3.81,
-      "learning_rate": 1.1910125681764288e-05,
-      "loss": 1.6346,
-      "step": 128500
-    },
-    {
-      "epoch": 3.82,
-      "learning_rate": 1.1761916054066872e-05,
-      "loss": 1.6497,
-      "step": 129000
-    },
-    {
-      "epoch": 3.84,
-      "learning_rate": 1.1613706426369457e-05,
-      "loss": 1.6228,
-      "step": 129500
-    },
-    {
-      "epoch": 3.85,
-      "learning_rate": 1.1465496798672043e-05,
-      "loss": 1.6137,
-      "step": 130000
-    },
-    {
-      "epoch": 3.87,
-      "learning_rate": 1.1317287170974627e-05,
-      "loss": 1.6424,
-      "step": 130500
-    },
-    {
-      "epoch": 3.88,
-      "learning_rate": 1.1169077543277212e-05,
-      "loss": 1.6277,
-      "step": 131000
-    },
-    {
-      "epoch": 3.9,
-      "learning_rate": 1.1020867915579796e-05,
-      "loss": 1.6309,
-      "step": 131500
-    },
-    {
-      "epoch": 3.91,
-      "learning_rate": 1.087265828788238e-05,
-      "loss": 1.6329,
-      "step": 132000
-    },
-    {
-      "epoch": 3.93,
-      "learning_rate": 1.0724448660184967e-05,
-      "loss": 1.6469,
-      "step": 132500
-    },
-    {
-      "epoch": 3.94,
-      "learning_rate": 1.0576239032487551e-05,
-      "loss": 1.628,
-      "step": 133000
-    },
-    {
-      "epoch": 3.96,
-      "learning_rate": 1.0428029404790136e-05,
-      "loss": 1.6392,
-      "step": 133500
-    },
-    {
-      "epoch": 3.97,
-      "learning_rate": 1.0279819777092722e-05,
-      "loss": 1.6337,
-      "step": 134000
-    },
-    {
-      "epoch": 3.99,
-      "learning_rate": 1.0131610149395306e-05,
-      "loss": 1.6455,
-      "step": 134500
-    },
-    {
-      "epoch": 4.0,
-      "learning_rate": 9.98340052169789e-06,
-      "loss": 1.6251,
-      "step": 135000
-    },
-    {
-      "epoch": 4.02,
-      "learning_rate": 9.835190894000475e-06,
-      "loss": 1.5574,
-      "step": 135500
-    },
-    {
-      "epoch": 4.03,
-      "learning_rate": 9.68698126630306e-06,
-      "loss": 1.5637,
-      "step": 136000
-    },
-    {
-      "epoch": 4.05,
-      "learning_rate": 9.538771638605644e-06,
-      "loss": 1.5522,
-      "step": 136500
-    },
-    {
-      "epoch": 4.06,
-      "learning_rate": 9.390562010908228e-06,
-      "loss": 1.5643,
-      "step": 137000
-    },
-    {
-      "epoch": 4.08,
-      "learning_rate": 9.242352383210814e-06,
-      "loss": 1.5713,
-      "step": 137500
-    },
-    {
-      "epoch": 4.09,
-      "learning_rate": 9.094142755513399e-06,
-      "loss": 1.5562,
-      "step": 138000
-    },
-    {
-      "epoch": 4.11,
-      "learning_rate": 8.945933127815983e-06,
-      "loss": 1.561,
-      "step": 138500
-    },
-    {
-      "epoch": 4.12,
-      "learning_rate": 8.797723500118567e-06,
-      "loss": 1.57,
-      "step": 139000
-    },
-    {
-      "epoch": 4.14,
-      "learning_rate": 8.649513872421154e-06,
-      "loss": 1.5747,
-      "step": 139500
-    },
-    {
-      "epoch": 4.15,
-      "learning_rate": 8.501304244723738e-06,
-      "loss": 1.5745,
-      "step": 140000
-    },
-    {
-      "epoch": 4.16,
-      "learning_rate": 8.353094617026322e-06,
-      "loss": 1.5642,
-      "step": 140500
-    },
-    {
-      "epoch": 4.18,
-      "learning_rate": 8.204884989328908e-06,
-      "loss": 1.5595,
-      "step": 141000
-    },
-    {
-      "epoch": 4.19,
-      "learning_rate": 8.056675361631493e-06,
-      "loss": 1.563,
-      "step": 141500
-    },
-    {
-      "epoch": 4.21,
-      "learning_rate": 7.908465733934077e-06,
-      "loss": 1.573,
-      "step": 142000
-    },
-    {
-      "epoch": 4.22,
-      "learning_rate": 7.760256106236662e-06,
-      "loss": 1.5599,
-      "step": 142500
-    },
-    {
-      "epoch": 4.24,
-      "learning_rate": 7.612046478539246e-06,
-      "loss": 1.5728,
-      "step": 143000
-    },
-    {
-      "epoch": 4.25,
-      "learning_rate": 7.4638368508418305e-06,
-      "loss": 1.5663,
-      "step": 143500
-    },
-    {
-      "epoch": 4.27,
-      "learning_rate": 7.315627223144415e-06,
-      "loss": 1.5639,
-      "step": 144000
-    },
-    {
-      "epoch": 4.28,
-      "learning_rate": 7.167417595447e-06,
-      "loss": 1.5638,
-      "step": 144500
-    },
-    {
-      "epoch": 4.3,
-      "learning_rate": 7.0192079677495855e-06,
-      "loss": 1.5848,
-      "step": 145000
-    },
-    {
-      "epoch": 4.31,
-      "learning_rate": 6.870998340052171e-06,
-      "loss": 1.5749,
-      "step": 145500
-    },
-    {
-      "epoch": 4.33,
-      "learning_rate": 6.722788712354755e-06,
-      "loss": 1.5706,
-      "step": 146000
-    },
-    {
-      "epoch": 4.34,
-      "learning_rate": 6.5745790846573396e-06,
-      "loss": 1.5832,
-      "step": 146500
-    },
-    {
-      "epoch": 4.36,
-      "learning_rate": 6.426369456959924e-06,
-      "loss": 1.5897,
-      "step": 147000
-    },
-    {
-      "epoch": 4.37,
-      "learning_rate": 6.278159829262508e-06,
-      "loss": 1.5883,
-      "step": 147500
-    },
-    {
-      "epoch": 4.39,
-      "learning_rate": 6.129950201565094e-06,
-      "loss": 1.6014,
-      "step": 148000
-    },
-    {
-      "epoch": 4.4,
-      "learning_rate": 5.981740573867679e-06,
-      "loss": 1.5756,
-      "step": 148500
-    },
-    {
-      "epoch": 4.42,
-      "learning_rate": 5.833530946170263e-06,
-      "loss": 1.5797,
-      "step": 149000
-    },
-    {
-      "epoch": 4.43,
-      "learning_rate": 5.6853213184728486e-06,
-      "loss": 1.577,
-      "step": 149500
-    },
-    {
-      "epoch": 4.45,
-      "learning_rate": 5.537111690775433e-06,
-      "loss": 1.583,
-      "step": 150000
-    },
-    {
-      "epoch": 4.46,
-      "learning_rate": 5.388902063078017e-06,
-      "loss": 1.5979,
-      "step": 150500
-    },
-    {
-      "epoch": 4.48,
-      "learning_rate": 5.240692435380603e-06,
-      "loss": 1.5847,
-      "step": 151000
-    },
-    {
-      "epoch": 4.49,
-      "learning_rate": 5.092482807683188e-06,
-      "loss": 1.5839,
-      "step": 151500
-    },
-    {
-      "epoch": 4.51,
-      "learning_rate": 4.944273179985772e-06,
-      "loss": 1.5707,
-      "step": 152000
-    },
-    {
-      "epoch": 4.52,
-      "learning_rate": 4.796063552288357e-06,
-      "loss": 1.5825,
-      "step": 152500
-    },
-    {
-      "epoch": 4.54,
-      "learning_rate": 4.647853924590942e-06,
-      "loss": 1.5799,
-      "step": 153000
-    },
-    {
-      "epoch": 4.55,
-      "learning_rate": 4.499644296893526e-06,
-      "loss": 1.5804,
-      "step": 153500
-    },
-    {
-      "epoch": 4.56,
-      "learning_rate": 4.351434669196111e-06,
-      "loss": 1.5841,
-      "step": 154000
-    },
-    {
-      "epoch": 4.58,
-      "learning_rate": 4.203225041498696e-06,
-      "loss": 1.582,
-      "step": 154500
-    },
-    {
-      "epoch": 4.59,
-      "learning_rate": 4.055015413801281e-06,
-      "loss": 1.5959,
-      "step": 155000
-    },
-    {
-      "epoch": 4.61,
-      "learning_rate": 3.906805786103866e-06,
-      "loss": 1.5959,
-      "step": 155500
-    },
-    {
-      "epoch": 4.62,
-      "learning_rate": 3.75859615840645e-06,
-      "loss": 1.6,
-      "step": 156000
-    },
-    {
-      "epoch": 4.64,
-      "learning_rate": 3.6103865307090354e-06,
-      "loss": 1.5994,
-      "step": 156500
-    },
-    {
-      "epoch": 4.65,
-      "learning_rate": 3.46217690301162e-06,
-      "loss": 1.5941,
-      "step": 157000
-    },
-    {
-      "epoch": 4.67,
-      "learning_rate": 3.3139672753142047e-06,
-      "loss": 1.582,
-      "step": 157500
-    },
-    {
-      "epoch": 4.68,
-      "learning_rate": 3.165757647616789e-06,
-      "loss": 1.5842,
-      "step": 158000
-    },
-    {
-      "epoch": 4.7,
-      "learning_rate": 3.017548019919374e-06,
-      "loss": 1.5866,
-      "step": 158500
-    },
-    {
-      "epoch": 4.71,
-      "learning_rate": 2.8693383922219588e-06,
-      "loss": 1.5969,
-      "step": 159000
-    },
-    {
-      "epoch": 4.73,
-      "learning_rate": 2.7211287645245436e-06,
-      "loss": 1.6188,
-      "step": 159500
-    },
-    {
-      "epoch": 4.74,
-      "learning_rate": 2.5729191368271284e-06,
-      "loss": 1.5876,
-      "step": 160000
-    },
-    {
-      "epoch": 4.76,
-      "learning_rate": 2.424709509129713e-06,
-      "loss": 1.6092,
-      "step": 160500
-    },
-    {
-      "epoch": 4.77,
-      "learning_rate": 2.276499881432298e-06,
-      "loss": 1.5818,
-      "step": 161000
-    },
-    {
-      "epoch": 4.79,
-      "learning_rate": 2.1282902537348825e-06,
-      "loss": 1.586,
-      "step": 161500
-    },
-    {
-      "epoch": 4.8,
-      "learning_rate": 1.9800806260374674e-06,
-      "loss": 1.5991,
-      "step": 162000
-    },
-    {
-      "epoch": 4.82,
-      "learning_rate": 1.8318709983400524e-06,
-      "loss": 1.5824,
-      "step": 162500
-    },
-    {
-      "epoch": 4.83,
-      "learning_rate": 1.683661370642637e-06,
-      "loss": 1.5907,
-      "step": 163000
-    },
-    {
-      "epoch": 4.85,
-      "learning_rate": 1.5354517429452217e-06,
-      "loss": 1.607,
-      "step": 163500
-    },
-    {
-      "epoch": 4.86,
-      "learning_rate": 1.3872421152478065e-06,
-      "loss": 1.591,
-      "step": 164000
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 1.2390324875503913e-06,
-      "loss": 1.6106,
-      "step": 164500
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 1.0908228598529762e-06,
-      "loss": 1.6029,
-      "step": 165000
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 9.426132321555609e-07,
-      "loss": 1.6054,
-      "step": 165500
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 7.944036044581456e-07,
-      "loss": 1.6038,
-      "step": 166000
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 6.461939767607305e-07,
-      "loss": 1.6129,
-      "step": 166500
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 4.979843490633152e-07,
-      "loss": 1.603,
-      "step": 167000
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 3.4977472136589993e-07,
-      "loss": 1.5922,
-      "step": 167500
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.0156509366848474e-07,
-      "loss": 1.6177,
-      "step": 168000
-    },
-    {
-      "epoch": 4.98,
-      "step": 168000,
-      "total_flos": 6.240794371172598e+17,
-      "train_loss": 0.0,
-      "train_runtime": 0.0953,
-      "train_samples_per_second": 4248275.036,
-      "train_steps_per_second": 1062084.5
+      "epoch": 3.0,
+      "step": 101208,
+      "total_flos": 3.759621873629921e+17,
+      "train_loss": 1.4532578269084653,
+      "train_runtime": 40248.475,
+      "train_samples_per_second": 10.058,
+      "train_steps_per_second": 2.515
     }
   ],
   "max_steps": 101208,
   "num_train_epochs": 3,
-  "total_flos": 6.240794371172598e+17,
+  "total_flos": 3.759621873629921e+17,
   "trial_name": null,
   "trial_params": null
 }