Add report.json for 1-epoch run

report.json

@@ -3,7 +3,7 @@
   "model": "teddyoas/Agron (continued pretraining)",
   "data_dir": "/content/data/agron-domain",
   "corpus": {
-    "created": "2026-10-23T09:01:47Z",
+    "created": "2026-10-26T11:57:09Z",
     "tokenizer": "teddyoas/Agron tokenizer.json",
     "seed": 1234,
     "sources": {
@@ -61,410 +61,303 @@
   "hyperparameters": {
     "block_size": 512,
     "batch_size": 16,
-    "grad_accum": 1,
-    "steps": 5000,
+    "grad_accum": 2,
+    "epochs": 1.0,
+    "steps": 7246,
     "lr": 0.0003,
-    "warmup": 200,
+    "warmup": 100,
     "weight_decay": 0.01,
     "grad_clip": 1.0,
     "seed": 42
   },
-  "tokens_per_step": 8192,
-  "total_tokens": 40960000,
-  "train_wall_seconds": 1466.5,
-  "tokens_per_second": 27931.3,
+  "tokens_per_step": 16384,
+  "total_tokens": 118718464,
+  "train_wall_seconds": 3659.7,
+  "tokens_per_second": 32439.5,
   "eval_before": {
     "code": {
-      "loss": 6.7396,
-      "perplexity": 845.25,
+      "loss": 1.4532,
+      "perplexity": 4.28,
       "blocks": 16
     },
     "cyber": {
-      "loss": 7.6322,
-      "perplexity": 2063.68,
+      "loss": 1.358,
+      "perplexity": 3.89,
       "blocks": 16
     },
     "replay": {
-      "loss": 7.2084,
-      "perplexity": 1350.77,
+      "loss": 5.626,
+      "perplexity": 277.55,
       "blocks": 5
     }
   },
   "eval_after": {
     "code": {
-      "loss": 1.4532,
-      "perplexity": 4.28,
+      "loss": 1.2451,
+      "perplexity": 3.47,
       "blocks": 16
     },
     "cyber": {
-      "loss": 1.3581,
-      "perplexity": 3.89,
+      "loss": 1.0636,
+      "perplexity": 2.9,
       "blocks": 16
     },
     "replay": {
-      "loss": 5.626,
-      "perplexity": 277.55,
+      "loss": 5.0238,
+      "perplexity": 151.98,
       "blocks": 5
     }
   },
   "eval_history": [
-    {
-      "step": 250,
-      "code": {
-        "loss": 3.1747,
-        "perplexity": 23.92,
-        "blocks": 16
-      },
-      "cyber": {
-        "loss": 2.8563,
-        "perplexity": 17.4,
-        "blocks": 16
-      },
-      "replay": {
-        "loss": 6.6708,
-        "perplexity": 789.04,
-        "blocks": 5
-      }
-    },
     {
       "step": 500,
       "code": {
-        "loss": 2.6789,
-        "perplexity": 14.57,
+        "loss": 1.5055,
+        "perplexity": 4.51,
         "blocks": 16
       },
       "cyber": {
-        "loss": 2.2694,
-        "perplexity": 9.67,
+        "loss": 1.4013,
+        "perplexity": 4.06,
         "blocks": 16
       },
       "replay": {
-        "loss": 6.5597,
-        "perplexity": 706.09,
-        "blocks": 5
-      }
-    },
-    {
-      "step": 750,
-      "code": {
-        "loss": 2.4011,
-        "perplexity": 11.03,
-        "blocks": 16
-      },
-      "cyber": {
-        "loss": 2.0552,
-        "perplexity": 7.81,
-        "blocks": 16
-      },
-      "replay": {
-        "loss": 6.4247,
-        "perplexity": 616.9,
+        "loss": 5.793,
+        "perplexity": 327.99,
         "blocks": 5
       }
     },
     {
       "step": 1000,
       "code": {
-        "loss": 2.0518,
-        "perplexity": 7.78,
-        "blocks": 16
-      },
-      "cyber": {
-        "loss": 1.9141,
-        "perplexity": 6.78,
-        "blocks": 16
-      },
-      "replay": {
-        "loss": 6.3068,
-        "perplexity": 548.3,
-        "blocks": 5
-      }
-    },
-    {
-      "step": 1250,
-      "code": {
-        "loss": 1.8635,
-        "perplexity": 6.45,
+        "loss": 1.4807,
+        "perplexity": 4.4,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.8036,
-        "perplexity": 6.07,
+        "loss": 1.3501,
+        "perplexity": 3.86,
         "blocks": 16
       },
       "replay": {
-        "loss": 6.1973,
-        "perplexity": 491.41,
+        "loss": 5.6901,
+        "perplexity": 295.92,
         "blocks": 5
       }
     },
     {
       "step": 1500,
       "code": {
-        "loss": 1.7843,
-        "perplexity": 5.96,
-        "blocks": 16
-      },
-      "cyber": {
-        "loss": 1.7277,
-        "perplexity": 5.63,
-        "blocks": 16
-      },
-      "replay": {
-        "loss": 6.1335,
-        "perplexity": 461.03,
-        "blocks": 5
-      }
-    },
-    {
-      "step": 1750,
-      "code": {
-        "loss": 1.7222,
-        "perplexity": 5.6,
+        "loss": 1.4339,
+        "perplexity": 4.19,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.6553,
-        "perplexity": 5.23,
+        "loss": 1.3048,
+        "perplexity": 3.69,
         "blocks": 16
       },
       "replay": {
-        "loss": 6.054,
-        "perplexity": 425.82,
+        "loss": 5.5986,
+        "perplexity": 270.06,
         "blocks": 5
       }
     },
     {
       "step": 2000,
       "code": {
-        "loss": 1.6748,
-        "perplexity": 5.34,
+        "loss": 1.4063,
+        "perplexity": 4.08,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.6081,
-        "perplexity": 4.99,
+        "loss": 1.2648,
+        "perplexity": 3.54,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.9676,
-        "perplexity": 390.59,
-        "blocks": 5
-      }
-    },
-    {
-      "step": 2250,
-      "code": {
-        "loss": 1.6263,
-        "perplexity": 5.09,
-        "blocks": 16
-      },
-      "cyber": {
-        "loss": 1.5535,
-        "perplexity": 4.73,
-        "blocks": 16
-      },
-      "replay": {
-        "loss": 5.9166,
-        "perplexity": 371.16,
+        "loss": 5.5139,
+        "perplexity": 248.12,
         "blocks": 5
       }
     },
     {
       "step": 2500,
       "code": {
-        "loss": 1.5959,
-        "perplexity": 4.93,
+        "loss": 1.3737,
+        "perplexity": 3.95,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.5182,
-        "perplexity": 4.56,
+        "loss": 1.2266,
+        "perplexity": 3.41,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.8646,
-        "perplexity": 352.35,
-        "blocks": 5
-      }
-    },
-    {
-      "step": 2750,
-      "code": {
-        "loss": 1.5603,
-        "perplexity": 4.76,
-        "blocks": 16
-      },
-      "cyber": {
-        "loss": 1.4842,
-        "perplexity": 4.41,
-        "blocks": 16
-      },
-      "replay": {
-        "loss": 5.8129,
-        "perplexity": 334.6,
+        "loss": 5.4098,
+        "perplexity": 223.59,
         "blocks": 5
       }
     },
     {
       "step": 3000,
       "code": {
-        "loss": 1.5311,
-        "perplexity": 4.62,
+        "loss": 1.3446,
+        "perplexity": 3.84,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.458,
-        "perplexity": 4.3,
+        "loss": 1.1879,
+        "perplexity": 3.28,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.7666,
-        "perplexity": 319.44,
+        "loss": 5.3373,
+        "perplexity": 207.95,
         "blocks": 5
       }
     },
     {
-      "step": 3250,
+      "step": 3500,
       "code": {
-        "loss": 1.5151,
-        "perplexity": 4.55,
+        "loss": 1.3275,
+        "perplexity": 3.77,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.4384,
-        "perplexity": 4.21,
+        "loss": 1.1612,
+        "perplexity": 3.19,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.7233,
-        "perplexity": 305.9,
+        "loss": 5.2525,
+        "perplexity": 191.05,
         "blocks": 5
       }
     },
     {
-      "step": 3500,
+      "step": 4000,
       "code": {
-        "loss": 1.4992,
-        "perplexity": 4.48,
+        "loss": 1.3077,
+        "perplexity": 3.7,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.4104,
-        "perplexity": 4.1,
+        "loss": 1.1404,
+        "perplexity": 3.13,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.6973,
-        "perplexity": 298.06,
+        "loss": 5.1941,
+        "perplexity": 180.21,
         "blocks": 5
       }
     },
     {
-      "step": 3750,
+      "step": 4500,
       "code": {
-        "loss": 1.4825,
-        "perplexity": 4.4,
+        "loss": 1.2838,
+        "perplexity": 3.61,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.3947,
-        "perplexity": 4.03,
+        "loss": 1.1155,
+        "perplexity": 3.05,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.6716,
-        "perplexity": 290.51,
+        "loss": 5.1429,
+        "perplexity": 171.22,
         "blocks": 5
       }
     },
     {
-      "step": 4000,
+      "step": 5000,
       "code": {
-        "loss": 1.4692,
-        "perplexity": 4.35,
+        "loss": 1.2708,
+        "perplexity": 3.56,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.3785,
-        "perplexity": 3.97,
+        "loss": 1.0989,
+        "perplexity": 3.0,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.6515,
-        "perplexity": 284.72,
+        "loss": 5.1017,
+        "perplexity": 164.31,
         "blocks": 5
       }
     },
     {
-      "step": 4250,
+      "step": 5500,
       "code": {
-        "loss": 1.461,
-        "perplexity": 4.31,
+        "loss": 1.261,
+        "perplexity": 3.53,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.3678,
-        "perplexity": 3.93,
+        "loss": 1.0821,
+        "perplexity": 2.95,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.638,
-        "perplexity": 280.9,
+        "loss": 5.0702,
+        "perplexity": 159.21,
         "blocks": 5
       }
     },
     {
-      "step": 4500,
+      "step": 6000,
       "code": {
-        "loss": 1.4559,
-        "perplexity": 4.29,
+        "loss": 1.2511,
+        "perplexity": 3.49,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.3612,
-        "perplexity": 3.9,
+        "loss": 1.0722,
+        "perplexity": 2.92,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.6302,
-        "perplexity": 278.72,
+        "loss": 5.0406,
+        "perplexity": 154.56,
         "blocks": 5
       }
     },
     {
-      "step": 4750,
+      "step": 6500,
       "code": {
-        "loss": 1.4533,
-        "perplexity": 4.28,
+        "loss": 1.2471,
+        "perplexity": 3.48,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.3585,
-        "perplexity": 3.89,
+        "loss": 1.0659,
+        "perplexity": 2.9,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.6273,
-        "perplexity": 277.92,
+        "loss": 5.0278,
+        "perplexity": 152.59,
         "blocks": 5
       }
     },
     {
-      "step": 5000,
+      "step": 7000,
       "code": {
-        "loss": 1.4532,
-        "perplexity": 4.28,
+        "loss": 1.2451,
+        "perplexity": 3.47,
         "blocks": 16
       },
       "cyber": {
-        "loss": 1.3581,
-        "perplexity": 3.89,
+        "loss": 1.0637,
+        "perplexity": 2.9,
         "blocks": 16
       },
       "replay": {
-        "loss": 5.626,
-        "perplexity": 277.55,
+        "loss": 5.0241,
+        "perplexity": 152.03,
         "blocks": 5
       }
     }
@@ -472,1204 +365,1738 @@
   "train_loss_history": [
     {
       "step": 25,
-      "loss": 6.523624267578125,
-      "learning_rate": 3.5999999999999994e-05,
-      "grad_norm": 2.373959541320801
+      "loss": 1.465992431640625,
+      "learning_rate": 7.199999999999999e-05,
+      "grad_norm": 0.503962516784668
     },
     {
       "step": 50,
-      "loss": 5.609417724609375,
-      "learning_rate": 7.35e-05,
-      "grad_norm": 1.9777405261993408
+      "loss": 1.4964903259277345,
+      "learning_rate": 0.000147,
+      "grad_norm": 0.5079663991928101
     },
     {
       "step": 75,
-      "loss": 4.9839346313476565,
-      "learning_rate": 0.00011099999999999999,
-      "grad_norm": 1.7172424793243408
+      "loss": 1.4984529113769531,
+      "learning_rate": 0.00022199999999999998,
+      "grad_norm": 0.5481196641921997
     },
     {
       "step": 100,
-      "loss": 4.605402526855468,
-      "learning_rate": 0.00014849999999999998,
-      "grad_norm": 2.054328441619873
+      "loss": 1.53364013671875,
+      "learning_rate": 0.00029699999999999996,
+      "grad_norm": 0.5628758072853088
     },
     {
       "step": 125,
-      "loss": 4.183442687988281,
-      "learning_rate": 0.000186,
-      "grad_norm": 1.8805474042892456
+      "loss": 1.503697509765625,
+      "learning_rate": 0.00029999165063514596,
+      "grad_norm": 0.6324927806854248
     },
     {
       "step": 150,
-      "loss": 3.8408413696289063,
-      "learning_rate": 0.00022349999999999998,
-      "grad_norm": 1.9061907529830933
+      "loss": 1.5965298461914061,
+      "learning_rate": 0.00029996519750734643,
+      "grad_norm": 0.5400546193122864
     },
     {
       "step": 175,
-      "loss": 3.700017395019531,
-      "learning_rate": 0.000261,
-      "grad_norm": 2.0188567638397217
+      "loss": 1.5519804382324218,
+      "learning_rate": 0.0002999206293164506,
+      "grad_norm": 0.5300002098083496
     },
     {
       "step": 200,
-      "loss": 3.5950775146484375,
-      "learning_rate": 0.0002985,
-      "grad_norm": 1.715572714805603
+      "loss": 1.5133120727539062,
+      "learning_rate": 0.0002998579514460781,
+      "grad_norm": 0.5625991821289062
     },
     {
       "step": 225,
-      "loss": 3.416778564453125,
-      "learning_rate": 0.0002999814948722491,
-      "grad_norm": 1.7615081071853638
+      "loss": 1.560133819580078,
+      "learning_rate": 0.00029977717146740954,
+      "grad_norm": 0.5264961123466492
     },
     {
       "step": 250,
-      "loss": 3.2645965576171876,
-      "learning_rate": 0.00029992286819925424,
-      "grad_norm": 1.7984426021575928
+      "loss": 1.5489260864257812,
+      "learning_rate": 0.0002996782991382715,
+      "grad_norm": 0.5348047018051147
     },
     {
       "step": 275,
-      "loss": 3.2022076416015626,
-      "learning_rate": 0.00029982410354917604,
-      "grad_norm": 1.6474525928497314
+      "loss": 1.4816361999511718,
+      "learning_rate": 0.0002995613464019583,
+      "grad_norm": 0.5102686285972595
     },
     {
       "step": 300,
-      "loss": 3.2187335205078127,
-      "learning_rate": 0.00029968522736369965,
-      "grad_norm": 1.5539392232894897
+      "loss": 1.5427243041992187,
+      "learning_rate": 0.00029942632738578904,
+      "grad_norm": 0.9355897307395935
     },
     {
       "step": 325,
-      "loss": 3.002821044921875,
-      "learning_rate": 0.00029950627682333796,
-      "grad_norm": 1.361164927482605
+      "loss": 1.4962777709960937,
+      "learning_rate": 0.00029927325839940107,
+      "grad_norm": 0.515880286693573
     },
     {
       "step": 350,
-      "loss": 3.0067147827148437,
-      "learning_rate": 0.00029928729983747795,
-      "grad_norm": 1.348253607749939
+      "loss": 1.5118930053710937,
+      "learning_rate": 0.0002991021579327801,
+      "grad_norm": 0.5115503072738647
     },
     {
       "step": 375,
-      "loss": 2.90574951171875,
-      "learning_rate": 0.00029902835503155365,
-      "grad_norm": 1.1980832815170288
+      "loss": 1.489414520263672,
+      "learning_rate": 0.0002989130466540265,
+      "grad_norm": 0.5668929815292358
     },
     {
       "step": 400,
-      "loss": 2.8375469970703127,
-      "learning_rate": 0.00029872951173135116,
-      "grad_norm": 1.374573826789856
+      "loss": 1.531691436767578,
+      "learning_rate": 0.00029870594740685855,
+      "grad_norm": 0.5156782269477844
     },
     {
       "step": 425,
-      "loss": 2.8665338134765626,
-      "learning_rate": 0.00029839084994444826,
-      "grad_norm": 1.200546145439148
+      "loss": 1.5007366943359375,
+      "learning_rate": 0.00029848088520785347,
+      "grad_norm": 0.4919520318508148
     },
     {
       "step": 450,
-      "loss": 2.800201721191406,
-      "learning_rate": 0.0002980124603387944,
-      "grad_norm": 1.1094062328338623
+      "loss": 1.5251170349121095,
+      "learning_rate": 0.0002982378872434252,
+      "grad_norm": 0.5143119096755981
     },
     {
       "step": 475,
-      "loss": 2.822390441894531,
-      "learning_rate": 0.0002975944442184368,
-      "grad_norm": 2.7519965171813965
+      "loss": 1.5015919494628907,
+      "learning_rate": 0.0002979769828665405,
+      "grad_norm": 0.5741536021232605
     },
     {
       "step": 500,
-      "loss": 2.68135498046875,
-      "learning_rate": 0.0002971369134963989,
-      "grad_norm": 1.1665791273117065
+      "loss": 1.4777200317382813,
+      "learning_rate": 0.0002976982035931733,
+      "grad_norm": 0.5316178202629089
     },
     {
       "step": 525,
-      "loss": 2.610870056152344,
-      "learning_rate": 0.00029663999066471857,
-      "grad_norm": 1.366045355796814
+      "loss": 1.5096803283691407,
+      "learning_rate": 0.0002974015830984975,
+      "grad_norm": 0.51364666223526
     },
     {
       "step": 550,
-      "loss": 2.605653076171875,
-      "learning_rate": 0.00029610380876165377,
-      "grad_norm": 1.0456777811050415
+      "loss": 1.4988615417480469,
+      "learning_rate": 0.0002970871572128197,
+      "grad_norm": 0.4849585294723511
     },
     {
       "step": 575,
-      "loss": 2.6331027221679686,
-      "learning_rate": 0.0002955285113360653,
-      "grad_norm": 1.0123140811920166
+      "loss": 1.5281857299804686,
+      "learning_rate": 0.00029675496391725037,
+      "grad_norm": 0.4936002492904663
     },
     {
       "step": 600,
-      "loss": 2.589538269042969,
-      "learning_rate": 0.0002949142524089853,
-      "grad_norm": 2.2923152446746826
+      "loss": 1.4581077575683594,
+      "learning_rate": 0.00029640504333911655,
+      "grad_norm": 0.4904490113258362
     },
     {
       "step": 625,
-      "loss": 2.509917907714844,
-      "learning_rate": 0.0002942611964323817,
-      "grad_norm": 1.028593897819519
+      "loss": 1.4236958312988282,
+      "learning_rate": 0.0002960374377471143,
+      "grad_norm": 0.49264708161354065
     },
     {
       "step": 650,
-      "loss": 2.49948974609375,
-      "learning_rate": 0.000293569518245131,
-      "grad_norm": 3.1793365478515625
+      "loss": 1.463341522216797,
+      "learning_rate": 0.0002956521915462028,
+      "grad_norm": 0.5007567405700684
     },
     {
       "step": 675,
-      "loss": 2.4947129821777345,
-      "learning_rate": 0.0002928394030262093,
-      "grad_norm": 1.0351264476776123
+      "loss": 1.4722225952148438,
+      "learning_rate": 0.0002952493512722409,
+      "grad_norm": 0.48851534724235535
     },
     {
       "step": 700,
-      "loss": 2.450889434814453,
-      "learning_rate": 0.0002920710462451155,
-      "grad_norm": 0.9668142199516296
+      "loss": 1.4606297302246094,
+      "learning_rate": 0.00029482896558636517,
+      "grad_norm": 0.5092523694038391
     },
     {
       "step": 725,
-      "loss": 2.437335662841797,
-      "learning_rate": 0.00029126465360953955,
-      "grad_norm": 1.0652600526809692
+      "loss": 1.4079983520507813,
+      "learning_rate": 0.00029439108526911243,
+      "grad_norm": 0.5120623707771301
     },
     {
       "step": 750,
-      "loss": 2.373745574951172,
-      "learning_rate": 0.0002904204410102891,
-      "grad_norm": 0.9787796139717102
+      "loss": 1.472628631591797,
+      "learning_rate": 0.0002939357632142855,
+      "grad_norm": 0.5220867991447449
     },
     {
       "step": 775,
-      "loss": 2.4216960144042967,
-      "learning_rate": 0.0002895386344634909,
-      "grad_norm": 1.6538870334625244
+      "loss": 1.4448049926757813,
+      "learning_rate": 0.0002934630544225637,
+      "grad_norm": 0.510402262210846
     },
     {
       "step": 800,
-      "loss": 2.4106399536132814,
-      "learning_rate": 0.0002886194700500804,
-      "grad_norm": 0.9779872894287109
+      "loss": 1.500727996826172,
+      "learning_rate": 0.0002929730159948594,
+      "grad_norm": 0.46804383397102356
     },
     {
       "step": 825,
-      "loss": 2.28968505859375,
-      "learning_rate": 0.00028766319385259713,
-      "grad_norm": 1.1657065153121948
+      "loss": 1.4285128784179688,
+      "learning_rate": 0.0002924657071254201,
+      "grad_norm": 0.511221170425415
     },
     {
       "step": 850,
-      "loss": 2.351943054199219,
-      "learning_rate": 0.0002866700618893029,
-      "grad_norm": 2.103409767150879
+      "loss": 1.4769186401367187,
+      "learning_rate": 0.0002919411890946786,
+      "grad_norm": 0.5099458694458008
     },
     {
       "step": 875,
-      "loss": 2.2706863403320314,
-      "learning_rate": 0.000285640340045639,
-      "grad_norm": 0.8603836297988892
+      "loss": 1.435450897216797,
+      "learning_rate": 0.0002913995252618499,
+      "grad_norm": 0.513334333896637
     },
     {
       "step": 900,
-      "loss": 2.322249298095703,
-      "learning_rate": 0.0002845743040030426,
-      "grad_norm": 0.8718733191490173
+      "loss": 1.4503627014160156,
+      "learning_rate": 0.0002908407810572784,
+      "grad_norm": 0.48919397592544556
     },
     {
       "step": 925,
-      "loss": 2.206558837890625,
-      "learning_rate": 0.00028347223916513995,
-      "grad_norm": 0.8749960064888
+      "loss": 1.4501026916503905,
+      "learning_rate": 0.00029026502397453395,
+      "grad_norm": 0.46839049458503723
     },
     {
       "step": 950,
-      "loss": 2.2034300231933592,
-      "learning_rate": 0.00028233444058133706,
-      "grad_norm": 0.7962949275970459
+      "loss": 1.4088829040527344,
+      "learning_rate": 0.0002896723235622588,
+      "grad_norm": 0.5264255404472351
     },
     {
       "step": 975,
-      "loss": 2.1959959411621095,
-      "learning_rate": 0.00028116121286782784,
-      "grad_norm": 1.006102204322815
+      "loss": 1.4610905456542969,
+      "learning_rate": 0.0002890627514157666,
+      "grad_norm": 0.4739971458911896
     },
     {
       "step": 1000,
-      "loss": 2.0366529846191406,
-      "learning_rate": 0.00027995287012604075,
-      "grad_norm": 1.1916731595993042
+      "loss": 1.439478759765625,
+      "learning_rate": 0.0002884363811683943,
+      "grad_norm": 0.5118926167488098
     },
     {
       "step": 1025,
-      "loss": 2.1152040100097658,
-      "learning_rate": 0.00027870973585854665,
-      "grad_norm": 0.8370559215545654
+      "loss": 1.3943740844726562,
+      "learning_rate": 0.00028779328848260716,
+      "grad_norm": 0.49222490191459656
     },
     {
       "step": 1050,
-      "loss": 2.12522705078125,
-      "learning_rate": 0.000277432142882449,
-      "grad_norm": 1.0043104887008667
+      "loss": 1.4404747009277343,
+      "learning_rate": 0.0002871335510408592,
+      "grad_norm": 0.48402321338653564
     },
     {
       "step": 1075,
-      "loss": 2.0742945861816406,
-      "learning_rate": 0.00027612043324028087,
-      "grad_norm": 0.8250269293785095
+      "loss": 1.3960848999023439,
+      "learning_rate": 0.00028645724853620995,
+      "grad_norm": 0.48026469349861145
     },
     {
       "step": 1100,
-      "loss": 2.0630601501464843,
-      "learning_rate": 0.00027477495810843166,
-      "grad_norm": 0.8697202205657959
+      "loss": 1.4121771240234375,
+      "learning_rate": 0.0002857644626626973,
+      "grad_norm": 0.5774903297424316
     },
     {
       "step": 1125,
-      "loss": 2.092942199707031,
-      "learning_rate": 0.00027339607770312895,
-      "grad_norm": 0.9798178672790527
+      "loss": 1.3999868774414062,
+      "learning_rate": 0.00028505527710546974,
+      "grad_norm": 0.5563474297523499
     },
     {
       "step": 1150,
-      "loss": 2.0511599731445314,
-      "learning_rate": 0.0002719841611839998,
-      "grad_norm": 0.930115282535553
+      "loss": 1.4009886169433594,
+      "learning_rate": 0.00028432977753067726,
+      "grad_norm": 0.5073521733283997
     },
     {
       "step": 1175,
-      "loss": 1.9910079956054687,
-      "learning_rate": 0.00027053958655523784,
-      "grad_norm": 0.9322160482406616
+      "loss": 1.375696258544922,
+      "learning_rate": 0.0002835880515751236,
+      "grad_norm": 0.5103248357772827
     },
     {
       "step": 1200,
-      "loss": 1.9892750549316407,
-      "learning_rate": 0.00026906274056440215,
-      "grad_norm": 0.8825056552886963
+      "loss": 1.4315164184570313,
+      "learning_rate": 0.00028283018883568,
+      "grad_norm": 0.4903603494167328
     },
     {
       "step": 1225,
-      "loss": 1.9053739929199218,
-      "learning_rate": 0.00026755401859887595,
-      "grad_norm": 0.8298675417900085
+      "loss": 1.4241004943847657,
+      "learning_rate": 0.0002820562808584622,
+      "grad_norm": 0.506152868270874
     },
     {
       "step": 1250,
-      "loss": 1.9155471801757813,
-      "learning_rate": 0.00026601382458001164,
-      "grad_norm": 0.845176637172699
+      "loss": 1.4256643676757812,
+      "learning_rate": 0.0002812664211277724,
+      "grad_norm": 0.4851268529891968
     },
     {
       "step": 1275,
-      "loss": 1.9351853942871093,
-      "learning_rate": 0.0002644425708549913,
-      "grad_norm": 0.8390039801597595
+      "loss": 1.4002693176269532,
+      "learning_rate": 0.00028046070505480665,
+      "grad_norm": 0.5567605495452881
     },
     {
       "step": 1300,
-      "loss": 1.9353437805175782,
-      "learning_rate": 0.00026284067808643143,
-      "grad_norm": 0.8998804092407227
+      "loss": 1.4149452209472657,
+      "learning_rate": 0.0002796392299661297,
+      "grad_norm": 0.49636349081993103
     },
     {
       "step": 1325,
-      "loss": 1.9378793334960938,
-      "learning_rate": 0.0002612085751397613,
-      "grad_norm": 0.8097237944602966
+      "loss": 1.3971769714355469,
+      "learning_rate": 0.0002788020950919183,
+      "grad_norm": 0.4663887917995453
     },
     {
       "step": 1350,
-      "loss": 1.9203353881835938,
-      "learning_rate": 0.00025954669896840544,
-      "grad_norm": 0.7300992012023926
+      "loss": 1.4440313720703124,
+      "learning_rate": 0.0002779494015539749,
+      "grad_norm": 0.5311484932899475
     },
     {
       "step": 1375,
-      "loss": 1.9306573486328125,
-      "learning_rate": 0.0002578554944968006,
-      "grad_norm": 3.613323211669922
+      "loss": 1.4248406982421875,
+      "learning_rate": 0.00027708125235351237,
+      "grad_norm": 0.4884320795536041
     },
     {
       "step": 1400,
-      "loss": 1.841597900390625,
-      "learning_rate": 0.00025613541450127857,
-      "grad_norm": 0.8019481301307678
+      "loss": 1.3670068359375,
+      "learning_rate": 0.00027619775235871217,
+      "grad_norm": 0.5010215640068054
     },
     {
       "step": 1425,
-      "loss": 1.7870706176757813,
-      "learning_rate": 0.0002543869194888471,
-      "grad_norm": 0.8653715252876282
+      "loss": 1.3905780029296875,
+      "learning_rate": 0.0002752990082920567,
+      "grad_norm": 0.5153194069862366
     },
     {
       "step": 1450,
-      "loss": 1.8280177307128906,
-      "learning_rate": 0.0002526104775739014,
-      "grad_norm": 0.7933656573295593
+      "loss": 1.4214459228515626,
+      "learning_rate": 0.0002743851287174378,
+      "grad_norm": 0.4822602868080139
     },
     {
       "step": 1475,
-      "loss": 1.889519500732422,
-      "learning_rate": 0.0002508065643528978,
-      "grad_norm": 0.8022903203964233
+      "loss": 1.4260044860839844,
+      "learning_rate": 0.00027345622402704257,
+      "grad_norm": 0.503484845161438
     },
     {
       "step": 1500,
-      "loss": 1.8756819152832032,
-      "learning_rate": 0.0002489756627770259,
-      "grad_norm": 0.7319203019142151
+      "loss": 1.387130126953125,
+      "learning_rate": 0.00027251240642801893,
+      "grad_norm": 0.5117263793945312
     },
     {
       "step": 1525,
-      "loss": 1.8584066772460937,
-      "learning_rate": 0.0002471182630229103,
-      "grad_norm": 0.8216777443885803
+      "loss": 1.4189169311523437,
+      "learning_rate": 0.000271553789928921,
+      "grad_norm": 0.5162776112556458
     },
     {
       "step": 1550,
-      "loss": 1.8107026672363282,
-      "learning_rate": 0.00024523486236137875,
-      "grad_norm": 0.7631626129150391
+      "loss": 1.3924339294433594,
+      "learning_rate": 0.0002705804903259379,
+      "grad_norm": 0.47961705923080444
     },
     {
       "step": 1575,
-      "loss": 1.9103573608398436,
-      "learning_rate": 0.0002433259650243306,
-      "grad_norm": 0.7555778622627258
+      "loss": 1.43085693359375,
+      "learning_rate": 0.000269592625188906,
+      "grad_norm": 0.4925697147846222
     },
     {
       "step": 1600,
-      "loss": 1.8494841003417968,
-      "learning_rate": 0.00024139208206974187,
-      "grad_norm": 0.7023468613624573
+      "loss": 1.3906794738769532,
+      "learning_rate": 0.00026859031384710686,
+      "grad_norm": 0.5015848278999329
     },
     {
       "step": 1625,
-      "loss": 1.8192034912109376,
-      "learning_rate": 0.00023943373124484234,
-      "grad_norm": 0.745850682258606
+      "loss": 1.4191299438476563,
+      "learning_rate": 0.000267573677374853,
+      "grad_norm": 0.47783127427101135
     },
     {
       "step": 1650,
-      "loss": 1.7671533203125,
-      "learning_rate": 0.00023745143684750263,
-      "grad_norm": 0.8100517392158508
+      "loss": 1.4013314819335938,
+      "learning_rate": 0.0002665428385768629,
+      "grad_norm": 0.44291064143180847
     },
     {
       "step": 1675,
-      "loss": 1.7958676147460937,
-      "learning_rate": 0.00023544572958586645,
-      "grad_norm": 0.8815721273422241
+      "loss": 1.4088816833496094,
+      "learning_rate": 0.00026549792197342627,
+      "grad_norm": 0.4942842721939087
     },
     {
       "step": 1700,
-      "loss": 1.8827743530273438,
-      "learning_rate": 0.00023341714643626747,
-      "grad_norm": 0.7374081015586853
+      "loss": 1.4031414794921875,
+      "learning_rate": 0.0002644390537853633,
+      "grad_norm": 0.4807228744029999
     },
     {
       "step": 1725,
-      "loss": 1.7900999450683595,
-      "learning_rate": 0.00023136623049946779,
-      "grad_norm": 0.7201098203659058
+      "loss": 1.4029275512695312,
+      "learning_rate": 0.0002633663619187772,
+      "grad_norm": 0.46695584058761597
     },
     {
       "step": 1750,
-      "loss": 1.7628277587890624,
-      "learning_rate": 0.0002292935308552567,
-      "grad_norm": 0.8404228091239929
+      "loss": 1.3980467224121094,
+      "learning_rate": 0.000262279975949604,
+      "grad_norm": 0.5140073299407959
     },
     {
       "step": 1775,
-      "loss": 1.7924160766601562,
-      "learning_rate": 0.0002271996024154485,
-      "grad_norm": 1.0276012420654297
+      "loss": 1.3929660034179687,
+      "learning_rate": 0.0002611800271079604,
+      "grad_norm": 0.4695814251899719
     },
     {
       "step": 1800,
-      "loss": 1.7698551940917968,
-      "learning_rate": 0.0002250850057753197,
-      "grad_norm": 0.7282796502113342
+      "loss": 1.3747761535644532,
+      "learning_rate": 0.00026006664826229203,
+      "grad_norm": 0.45960962772369385
     },
     {
       "step": 1825,
-      "loss": 1.7850868225097656,
-      "learning_rate": 0.00022295030706352356,
-      "grad_norm": 0.8056021332740784
+      "loss": 1.373397979736328,
+      "learning_rate": 0.00025893997390332314,
+      "grad_norm": 0.4975016415119171
     },
     {
       "step": 1850,
-      "loss": 1.762191162109375,
-      "learning_rate": 0.00022079607779052418,
-      "grad_norm": 0.7027015089988708
+      "loss": 1.3793846130371095,
+      "learning_rate": 0.00025780014012781113,
+      "grad_norm": 0.48639926314353943
     },
     {
       "step": 1875,
-      "loss": 1.7117097473144531,
-      "learning_rate": 0.00021862289469558904,
-      "grad_norm": 0.6894400715827942
+      "loss": 1.4201519775390625,
+      "learning_rate": 0.0002566472846221067,
+      "grad_norm": 0.48666369915008545
     },
     {
       "step": 1900,
-      "loss": 1.7361253356933595,
-      "learning_rate": 0.0002164313395923822,
-      "grad_norm": 0.770681619644165
+      "loss": 1.4070872497558593,
+      "learning_rate": 0.00025548154664552166,
+      "grad_norm": 0.4822351634502411
     },
     {
       "step": 1925,
-      "loss": 1.7665594482421876,
-      "learning_rate": 0.00021422199921319844,
-      "grad_norm": 0.7259151935577393
+      "loss": 1.4160031127929686,
+      "learning_rate": 0.00025430306701350764,
+      "grad_norm": 0.49669837951660156
     },
     {
       "step": 1950,
-      "loss": 1.7720823669433594,
-      "learning_rate": 0.00021199546505188104,
-      "grad_norm": 0.6277358531951904
+      "loss": 1.4190867614746094,
+      "learning_rate": 0.0002531119880806459,
+      "grad_norm": 0.4919108748435974
     },
     {
       "step": 1975,
-      "loss": 1.7142399597167968,
-      "learning_rate": 0.00020975233320546464,
-      "grad_norm": 2.8367106914520264
+      "loss": 1.4013197326660156,
+      "learning_rate": 0.0002519084537234516,
+      "grad_norm": 0.48018497228622437
     },
     {
       "step": 2000,
-      "loss": 1.7562342834472657,
-      "learning_rate": 0.00020749320421458532,
-      "grad_norm": 0.6662013530731201
+      "loss": 1.3750941467285156,
+      "learning_rate": 0.0002506926093229944,
+      "grad_norm": 0.5125364065170288
     },
     {
       "step": 2025,
-      "loss": 1.7108480834960937,
-      "learning_rate": 0.0002052186829027017,
-      "grad_norm": 0.7505387663841248
+      "loss": 1.3990634155273438,
+      "learning_rate": 0.000249464601747337,
+      "grad_norm": 0.4900771677494049
     },
     {
       "step": 2050,
-      "loss": 1.6644221496582032,
-      "learning_rate": 0.00020292937821416885,
-      "grad_norm": 0.7581669688224792
+      "loss": 1.3854647827148439,
+      "learning_rate": 0.0002482245793337941,
+      "grad_norm": 0.4588489532470703
     },
     {
       "step": 2075,
-      "loss": 1.7933123779296876,
-      "learning_rate": 0.00020062590305120932,
-      "grad_norm": 0.6616817116737366
+      "loss": 1.3835163879394532,
+      "learning_rate": 0.00024697269187101445,
+      "grad_norm": 0.46798691153526306
     },
     {
       "step": 2100,
-      "loss": 1.6594839477539063,
-      "learning_rate": 0.0001983088741098243,
-      "grad_norm": 0.804240882396698
+      "loss": 1.4211228942871095,
+      "learning_rate": 0.0002457090905808864,
+      "grad_norm": 0.4728078842163086
     },
     {
       "step": 2125,
-      "loss": 1.62792236328125,
-      "learning_rate": 0.00019597891171468915,
-      "grad_norm": 0.6666399240493774
+      "loss": 1.4086012268066406,
+      "learning_rate": 0.00024443392810027167,
+      "grad_norm": 0.4489896893501282
     },
     {
       "step": 2150,
-      "loss": 1.7009780883789063,
-      "learning_rate": 0.0001936366396530776,
-      "grad_norm": 0.716991662979126
+      "loss": 1.3630841064453125,
+      "learning_rate": 0.0002431473584625671,
+      "grad_norm": 0.45225420594215393
     },
     {
       "step": 2175,
-      "loss": 1.6474826049804687,
-      "learning_rate": 0.0001912826850078587,
-      "grad_norm": 0.6721972823143005
+      "loss": 1.3702024841308593,
+      "learning_rate": 0.00024184953707909852,
+      "grad_norm": 0.4957192540168762
     },
     {
       "step": 2200,
-      "loss": 1.7127001953125,
-      "learning_rate": 0.00018891767798961175,
-      "grad_norm": 0.9610303044319153
+      "loss": 1.397852783203125,
+      "learning_rate": 0.0002405406207203476,
+      "grad_norm": 0.45574721693992615
     },
     {
       "step": 2225,
-      "loss": 1.6575242614746093,
-      "learning_rate": 0.00018654225176790336,
-      "grad_norm": 0.7149221897125244
+      "loss": 1.4111256408691406,
+      "learning_rate": 0.00023922076749701487,
+      "grad_norm": 0.45370936393737793
     },
     {
       "step": 2250,
-      "loss": 1.660975341796875,
-      "learning_rate": 0.00018415704230177304,
-      "grad_norm": 1.0614609718322754
+      "loss": 1.36464599609375,
+      "learning_rate": 0.00023789013684092053,
+      "grad_norm": 0.45561283826828003
     },
     {
       "step": 2275,
-      "loss": 1.6391448974609375,
-      "learning_rate": 0.00018176268816947136,
-      "grad_norm": 0.8024672269821167
+      "loss": 1.420698699951172,
+      "learning_rate": 0.00023654888948574615,
+      "grad_norm": 0.48367705941200256
     },
     {
       "step": 2300,
-      "loss": 1.6709408569335937,
-      "learning_rate": 0.00017935983039749705,
-      "grad_norm": 0.8126998543739319
+      "loss": 1.3629957580566405,
+      "learning_rate": 0.0002351971874476186,
+      "grad_norm": 0.45925864577293396
     },
     {
       "step": 2325,
-      "loss": 1.6153744506835936,
-      "learning_rate": 0.0001769491122889789,
-      "grad_norm": 0.7124079465866089
+      "loss": 1.3583328247070312,
+      "learning_rate": 0.00023383519400553934,
+      "grad_norm": 0.4538702964782715
     },
     {
       "step": 2350,
-      "loss": 1.6257669067382812,
-      "learning_rate": 0.00017453117925144782,
-      "grad_norm": 0.7417547702789307
+      "loss": 1.3273094177246094,
+      "learning_rate": 0.00023246307368166122,
+      "grad_norm": 0.4757903516292572
     },
     {
       "step": 2375,
-      "loss": 1.6552053833007812,
-      "learning_rate": 0.00017210667862404568,
-      "grad_norm": 0.704951286315918
+      "loss": 1.3557374572753906,
+      "learning_rate": 0.00023108099222141494,
+      "grad_norm": 0.44728630781173706
     },
     {
       "step": 2400,
-      "loss": 1.682860107421875,
-      "learning_rate": 0.0001696762595042171,
-      "grad_norm": 0.8258393406867981
+      "loss": 1.3738319396972656,
+      "learning_rate": 0.0002296891165734876,
+      "grad_norm": 0.5025182366371155
     },
     {
       "step": 2425,
-      "loss": 1.6681719970703126,
-      "learning_rate": 0.00016724057257392998,
-      "grad_norm": 0.6800954341888428
+      "loss": 1.3842662048339844,
+      "learning_rate": 0.00022828761486965658,
+      "grad_norm": 0.46316027641296387
     },
     {
       "step": 2450,
-      "loss": 1.655802001953125,
-      "learning_rate": 0.00016480026992547267,
-      "grad_norm": 0.7170723080635071
+      "loss": 1.4199676513671875,
+      "learning_rate": 0.0002268766564044796,
+      "grad_norm": 0.4852520227432251
     },
     {
       "step": 2475,
-      "loss": 1.6203256225585938,
-      "learning_rate": 0.00016235600488687284,
-      "grad_norm": 0.6897627115249634
+      "loss": 1.3864126586914063,
+      "learning_rate": 0.0002254564116148451,
+      "grad_norm": 0.47896522283554077
     },
     {
       "step": 2500,
-      "loss": 1.692429962158203,
-      "learning_rate": 0.0001599084318469858,
-      "grad_norm": 0.6847426295280457
+      "loss": 1.3791714477539063,
+      "learning_rate": 0.0002240270520593839,
+      "grad_norm": 0.7034990787506104
     },
     {
       "step": 2525,
-      "loss": 1.6337057495117187,
-      "learning_rate": 0.00015745820608029868,
-      "grad_norm": 0.7274415493011475
+      "loss": 1.4165682983398438,
+      "learning_rate": 0.0002225887503977462,
+      "grad_norm": 0.4801398515701294
     },
     {
       "step": 2550,
-      "loss": 1.6076887512207032,
-      "learning_rate": 0.00015500598357149793,
-      "grad_norm": 0.7195432782173157
+      "loss": 1.376103515625,
+      "learning_rate": 0.0002211416803697449,
+      "grad_norm": 0.4766090214252472
     },
     {
       "step": 2575,
-      "loss": 1.619275360107422,
-      "learning_rate": 0.00015255242083984602,
-      "grad_norm": 0.6512688398361206
+      "loss": 1.3946040344238282,
+      "learning_rate": 0.00021968601677436867,
+      "grad_norm": 0.46213752031326294
     },
     {
       "step": 2600,
-      "loss": 1.642754364013672,
-      "learning_rate": 0.0001500981747634155,
-      "grad_norm": 0.6707054376602173
+      "loss": 1.3469049072265624,
+      "learning_rate": 0.00021822193544866712,
+      "grad_norm": 0.48614174127578735
     },
     {
       "step": 2625,
-      "loss": 1.5942057800292968,
-      "learning_rate": 0.00014764390240322691,
-      "grad_norm": 0.6722701191902161
+      "loss": 1.4027534484863282,
+      "learning_rate": 0.00021674961324651061,
+      "grad_norm": 0.46639734506607056
     },
     {
       "step": 2650,
-      "loss": 1.6188661193847655,
-      "learning_rate": 0.0001451902608273374,
-      "grad_norm": 0.6420400738716125
+      "loss": 1.3664921569824218,
+      "learning_rate": 0.000215269228017227,
+      "grad_norm": 0.4851883351802826
     },
     {
       "step": 2675,
-      "loss": 1.6881460571289062,
-      "learning_rate": 0.00014273790693492824,
-      "grad_norm": 0.6697176694869995
+      "loss": 1.3809815979003905,
+      "learning_rate": 0.00021378095858411863,
+      "grad_norm": 0.45697319507598877
     },
     {
       "step": 2700,
-      "loss": 1.6173477172851562,
-      "learning_rate": 0.00014028749728043626,
-      "grad_norm": 0.6940237283706665
+      "loss": 1.3494392395019532,
+      "learning_rate": 0.0002122849847228609,
+      "grad_norm": 0.47020864486694336
     },
     {
       "step": 2725,
-      "loss": 1.6341883850097656,
-      "learning_rate": 0.00013783968789777916,
-      "grad_norm": 0.6811985373497009
+      "loss": 1.338717041015625,
+      "learning_rate": 0.00021078148713978665,
+      "grad_norm": 0.45249685645103455
     },
     {
       "step": 2750,
-      "loss": 1.6167942810058593,
-      "learning_rate": 0.0001353951341247189,
-      "grad_norm": 0.6664125919342041
+      "loss": 1.3287506103515625,
+      "learning_rate": 0.00020927064745005744,
+      "grad_norm": 0.44328123331069946
     },
     {
       "step": 2775,
-      "loss": 1.557093048095703,
-      "learning_rate": 0.00013295449042741183,
-      "grad_norm": 0.690255880355835
+      "loss": 1.3808297729492187,
+      "learning_rate": 0.00020775264815572558,
+      "grad_norm": 0.44321563839912415
     },
     {
       "step": 2800,
-      "loss": 1.5586013793945312,
-      "learning_rate": 0.0001305184102251927,
-      "grad_norm": 0.7270392775535583
+      "loss": 1.3122210693359375,
+      "learning_rate": 0.00020622767262368853,
+      "grad_norm": 0.4848768711090088
     },
     {
       "step": 2825,
-      "loss": 1.5695785522460937,
-      "learning_rate": 0.00012808754571563827,
-      "grad_norm": 0.6706142425537109
+      "loss": 1.3498515319824218,
+      "learning_rate": 0.00020469590506353916,
+      "grad_norm": 0.47914770245552063
     },
     {
       "step": 2850,
-      "loss": 1.5819349670410157,
-      "learning_rate": 0.00012566254769995805,
-      "grad_norm": 0.7445992231369019
+      "loss": 1.3182548522949218,
+      "learning_rate": 0.0002031575305053142,
+      "grad_norm": 0.4577605426311493
     },
     {
       "step": 2875,
-      "loss": 1.594710693359375,
-      "learning_rate": 0.0001232440654087595,
-      "grad_norm": 0.6728308796882629
+      "loss": 1.3279336547851563,
+      "learning_rate": 0.00020161273477714345,
+      "grad_norm": 0.4621855318546295
     },
     {
       "step": 2900,
-      "loss": 1.6170135498046876,
-      "learning_rate": 0.00012083274632823302,
-      "grad_norm": 0.6884806156158447
+      "loss": 1.336206817626953,
+      "learning_rate": 0.0002000617044828027,
+      "grad_norm": 0.4753350019454956
     },
     {
       "step": 2925,
-      "loss": 1.584092559814453,
-      "learning_rate": 0.00011842923602680417,
-      "grad_norm": 0.6766085624694824
+      "loss": 1.3457479858398438,
+      "learning_rate": 0.00019850462697917277,
+      "grad_norm": 0.45626455545425415
     },
     {
       "step": 2950,
-      "loss": 1.6283296203613282,
-      "learning_rate": 0.00011603417798229966,
-      "grad_norm": 0.6982696652412415
+      "loss": 1.3458819580078125,
+      "learning_rate": 0.00019694169035360792,
+      "grad_norm": 0.4753391742706299
     },
     {
       "step": 2975,
-      "loss": 1.58716064453125,
-      "learning_rate": 0.00011364821340967271,
-      "grad_norm": 0.7124394774436951
+      "loss": 1.3453900146484374,
+      "learning_rate": 0.0001953730834012157,
+      "grad_norm": 0.4793599545955658
     },
     {
       "step": 3000,
-      "loss": 1.5275750732421876,
-      "learning_rate": 0.000111271981089334,
-      "grad_norm": 0.7986636757850647
+      "loss": 1.3578317260742188,
+      "learning_rate": 0.00019379899560205134,
+      "grad_norm": 0.4540102779865265
     },
     {
       "step": 3025,
-      "loss": 1.549556884765625,
-      "learning_rate": 0.00010890611719613512,
-      "grad_norm": 0.6931105256080627
+      "loss": 1.3426896667480468,
+      "learning_rate": 0.00019221961709822968,
+      "grad_norm": 0.4781694710254669
     },
     {
       "step": 3050,
-      "loss": 1.617264404296875,
-      "learning_rate": 0.00010655125512904897,
-      "grad_norm": 0.735173761844635
+      "loss": 1.3274795532226562,
+      "learning_rate": 0.00019063513867095678,
+      "grad_norm": 0.4815289080142975
     },
     {
       "step": 3075,
-      "loss": 1.5539808654785157,
-      "learning_rate": 0.00010420802534159346,
-      "grad_norm": 0.6854939460754395
+      "loss": 1.3402708435058595,
+      "learning_rate": 0.00018904575171748458,
+      "grad_norm": 0.44584402441978455
     },
     {
       "step": 3100,
-      "loss": 1.552236328125,
-      "learning_rate": 0.00010187705517304411,
-      "grad_norm": 0.6724647283554077
+      "loss": 1.3204678344726561,
+      "learning_rate": 0.00018745164822799094,
+      "grad_norm": 0.4505169987678528
     },
     {
       "step": 3125,
-      "loss": 1.6483671569824219,
-      "learning_rate": 9.955896868047993e-05,
-      "grad_norm": 0.7743362784385681
+      "loss": 1.3190098571777344,
+      "learning_rate": 0.00018585302076238827,
+      "grad_norm": 0.47513049840927124
     },
     {
       "step": 3150,
-      "loss": 1.5315673828125,
-      "learning_rate": 9.72543864717083e-05,
-      "grad_norm": 0.6886186003684998
+      "loss": 1.318228759765625,
+      "learning_rate": 0.00018425006242706312,
+      "grad_norm": 0.46475449204444885
     },
     {
       "step": 3175,
-      "loss": 1.5273338317871095,
-      "learning_rate": 9.496392553911336e-05,
-      "grad_norm": 0.7097812294960022
+      "loss": 1.3169645690917968,
+      "learning_rate": 0.00018264296685155002,
+      "grad_norm": 0.49327173829078674
     },
     {
       "step": 3200,
-      "loss": 1.5534779357910156,
-      "learning_rate": 9.268819909447217e-05,
-      "grad_norm": 0.7026636004447937
+      "loss": 1.3219915771484374,
+      "learning_rate": 0.00018103192816514195,
+      "grad_norm": 0.4188969135284424
     },
     {
       "step": 3225,
-      "loss": 1.5758627319335938,
-      "learning_rate": 9.042781640478291e-05,
-      "grad_norm": 0.7093859910964966
+      "loss": 1.3042182922363281,
+      "learning_rate": 0.00017941714097344027,
+      "grad_norm": 0.4400496482849121
     },
     {
       "step": 3250,
-      "loss": 1.5612242126464844,
-      "learning_rate": 8.818338262914968e-05,
-      "grad_norm": 0.6529397368431091
+      "loss": 1.388089141845703,
+      "learning_rate": 0.00017779880033484753,
+      "grad_norm": 0.4402545690536499
     },
     {
       "step": 3275,
-      "loss": 1.6053514099121093,
-      "learning_rate": 8.595549865676666e-05,
-      "grad_norm": 0.6446133852005005
+      "loss": 1.2967855834960937,
+      "learning_rate": 0.00017617710173700537,
+      "grad_norm": 0.4542015492916107
     },
     {
       "step": 3300,
-      "loss": 1.4834933471679688,
-      "learning_rate": 8.374476094604538e-05,
-      "grad_norm": 0.6316361427307129
+      "loss": 1.3468402099609376,
+      "learning_rate": 0.00017455224107318037,
+      "grad_norm": 0.44114288687705994
     },
     {
       "step": 3325,
-      "loss": 1.5587440490722657,
-      "learning_rate": 8.155176136492857e-05,
-      "grad_norm": 0.7731664180755615
+      "loss": 1.314374237060547,
+      "learning_rate": 0.0001729244146186013,
+      "grad_norm": 0.448632150888443
     },
     {
       "step": 3350,
-      "loss": 1.5359294128417968,
-      "learning_rate": 7.937708703243285e-05,
-      "grad_norm": 0.7346474528312683
+      "loss": 1.3190969848632812,
+      "learning_rate": 0.00017129381900674997,
+      "grad_norm": 0.49022868275642395
     },
     {
       "step": 3375,
-      "loss": 1.5623774719238281,
-      "learning_rate": 7.722132016146243e-05,
-      "grad_norm": 0.7099519968032837
+      "loss": 1.3036573791503907,
+      "learning_rate": 0.00016966065120560889,
+      "grad_norm": 0.465278685092926
     },
     {
       "step": 3400,
-      "loss": 1.5111256408691407,
-      "learning_rate": 7.508503790293705e-05,
-      "grad_norm": 0.6971424221992493
+      "loss": 1.3122999572753906,
+      "learning_rate": 0.00016802510849386838,
+      "grad_norm": 0.46860554814338684
     },
     {
       "step": 3425,
-      "loss": 1.5142340087890624,
-      "learning_rate": 7.296881219127452e-05,
-      "grad_norm": 0.7202976942062378
+      "loss": 1.3047955322265625,
+      "learning_rate": 0.00016638738843709644,
+      "grad_norm": 0.4665217399597168
     },
     {
       "step": 3450,
-      "loss": 1.5641429138183593,
-      "learning_rate": 7.087320959126997e-05,
-      "grad_norm": 0.6218193173408508
+      "loss": 1.2953912353515624,
+      "learning_rate": 0.0001647476888638736,
+      "grad_norm": 0.4496661424636841
     },
     {
       "step": 3475,
-      "loss": 1.51037109375,
-      "learning_rate": 6.879879114641299e-05,
-      "grad_norm": 0.7012581825256348
+      "loss": 1.2946479797363282,
+      "learning_rate": 0.00016310620784189646,
+      "grad_norm": 0.4517843425273895
     },
     {
       "step": 3500,
-      "loss": 1.5376332092285157,
-      "learning_rate": 6.674611222868252e-05,
-      "grad_norm": 0.7338404059410095
+      "loss": 1.2945361328125,
+      "learning_rate": 0.00016146314365405177,
+      "grad_norm": 0.4854501783847809
     },
     {
       "step": 3525,
-      "loss": 1.5199795532226563,
-      "learning_rate": 6.47157223898604e-05,
-      "grad_norm": 0.6267840266227722
+      "loss": 1.299219970703125,
+      "learning_rate": 0.00015981869477446509,
+      "grad_norm": 0.4721055030822754
     },
     {
       "step": 3550,
-      "loss": 1.5208247375488282,
-      "learning_rate": 6.27081652144031e-05,
-      "grad_norm": 0.6833776235580444
+      "loss": 1.2779254913330078,
+      "learning_rate": 0.00015817305984452593,
+      "grad_norm": 0.4885696768760681
     },
     {
       "step": 3575,
-      "loss": 1.4701828002929687,
-      "learning_rate": 6.0723978173911005e-05,
-      "grad_norm": 0.7513471841812134
+      "loss": 1.3132022094726563,
+      "learning_rate": 0.00015652643764889294,
+      "grad_norm": 0.46799078583717346
     },
     {
       "step": 3600,
-      "loss": 1.5226860046386719,
-      "learning_rate": 5.876369248323416e-05,
-      "grad_norm": 0.6252737641334534
+      "loss": 1.3397958374023438,
+      "learning_rate": 0.00015487902709148153,
+      "grad_norm": 0.4386662244796753
     },
     {
       "step": 3625,
-      "loss": 1.492273712158203,
-      "learning_rate": 5.682783295825345e-05,
-      "grad_norm": 0.7011887431144714
+      "loss": 1.298585205078125,
+      "learning_rate": 0.00015323102717143747,
+      "grad_norm": 0.439731627702713
     },
     {
       "step": 3650,
-      "loss": 1.4881370544433594,
-      "learning_rate": 5.49169178753745e-05,
-      "grad_norm": 0.7054076194763184
+      "loss": 1.2728870391845704,
+      "learning_rate": 0.00015158263695909857,
+      "grad_norm": 0.47465574741363525
     },
     {
       "step": 3675,
-      "loss": 1.47126220703125,
-      "learning_rate": 5.303145883277253e-05,
-      "grad_norm": 0.6423889994621277
+      "loss": 1.3068107604980468,
+      "learning_rate": 0.00014993405557194814,
+      "grad_norm": 0.4406924247741699
     },
     {
       "step": 3700,
-      "loss": 1.5296780395507812,
-      "learning_rate": 5.11719606134253e-05,
-      "grad_norm": 0.7119572162628174
+      "loss": 1.3211735534667968,
+      "learning_rate": 0.00014828548215056232,
+      "grad_norm": 0.4591175317764282
     },
     {
       "step": 3725,
-      "loss": 1.5092683410644532,
-      "learning_rate": 4.933892104997057e-05,
-      "grad_norm": 0.6653545498847961
+      "loss": 1.2942324829101564,
+      "learning_rate": 0.0001466371158345553,
+      "grad_norm": 0.4335545301437378
     },
     {
       "step": 3750,
-      "loss": 1.5583695983886718,
-      "learning_rate": 4.75328308914238e-05,
-      "grad_norm": 0.6844574213027954
+      "loss": 1.3200546264648438,
+      "learning_rate": 0.00014498915573852368,
+      "grad_norm": 0.4724709093570709
     },
     {
       "step": 3775,
-      "loss": 1.475927276611328,
-      "learning_rate": 4.575417367179337e-05,
-      "grad_norm": 0.7387920618057251
+      "loss": 1.2709454345703124,
+      "learning_rate": 0.00014334180092799478,
+      "grad_norm": 0.4954122304916382
     },
     {
       "step": 3800,
-      "loss": 1.5570849609375,
-      "learning_rate": 4.400342558062649e-05,
-      "grad_norm": 0.6857640147209167
+      "loss": 1.307467498779297,
+      "learning_rate": 0.00014169525039538032,
+      "grad_norm": 0.4750748574733734
     },
     {
       "step": 3825,
-      "loss": 1.5318179321289063,
-      "learning_rate": 4.228105533552169e-05,
-      "grad_norm": 0.6937131881713867
+      "loss": 1.2666266632080079,
+      "learning_rate": 0.0001400497030359393,
+      "grad_norm": 0.4433806538581848
     },
     {
       "step": 3850,
-      "loss": 1.515167236328125,
-      "learning_rate": 4.0587524056642065e-05,
-      "grad_norm": 0.7321174740791321
+      "loss": 1.3017340087890625,
+      "learning_rate": 0.00013840535762375225,
+      "grad_norm": 0.44521960616111755
     },
     {
       "step": 3875,
-      "loss": 1.5477534484863282,
-      "learning_rate": 3.892328514326204e-05,
-      "grad_norm": 0.7120669484138489
+      "loss": 1.3061532592773437,
+      "learning_rate": 0.0001367624127877103,
+      "grad_norm": 0.44376805424690247
     },
     {
       "step": 3900,
-      "loss": 1.4948825073242187,
-      "learning_rate": 3.728878415238148e-05,
-      "grad_norm": 0.666736900806427
+      "loss": 1.3032217407226563,
+      "learning_rate": 0.0001351210669875218,
+      "grad_norm": 0.4341404438018799
     },
     {
       "step": 3925,
-      "loss": 1.4762055969238281,
-      "learning_rate": 3.568445867943932e-05,
-      "grad_norm": 0.6510305404663086
+      "loss": 1.272267074584961,
+      "learning_rate": 0.00013348151848973955,
+      "grad_norm": 0.45423921942710876
     },
     {
       "step": 3950,
-      "loss": 1.5333436584472657,
-      "learning_rate": 3.411073824115867e-05,
-      "grad_norm": 0.6959473490715027
+      "loss": 1.339799346923828,
+      "learning_rate": 0.00013184396534381086,
+      "grad_norm": 0.4589066803455353
     },
     {
       "step": 3975,
-      "loss": 1.485062255859375,
-      "learning_rate": 3.2568044160554774e-05,
-      "grad_norm": 0.7163877487182617
+      "loss": 1.2691211700439453,
+      "learning_rate": 0.00013020860535815427,
+      "grad_norm": 0.44028106331825256
     },
     {
       "step": 4000,
-      "loss": 1.4653591918945312,
-      "learning_rate": 3.1056789454136674e-05,
-      "grad_norm": 0.7000799179077148
+      "loss": 1.2714073181152343,
+      "learning_rate": 0.0001285756360762654,
+      "grad_norm": 0.4480900466442108
     },
     {
       "step": 4025,
-      "loss": 1.4768002319335938,
-      "learning_rate": 2.9577378721332843e-05,
-      "grad_norm": 0.6679511666297913
+      "loss": 1.3150112915039063,
+      "learning_rate": 0.00012694525475285442,
+      "grad_norm": 0.471388578414917
     },
     {
       "step": 4050,
-      "loss": 1.5249488830566407,
-      "learning_rate": 2.8130208036169786e-05,
-      "grad_norm": 0.7075392603874207
+      "loss": 1.2618749237060547,
+      "learning_rate": 0.00012531765833001876,
+      "grad_norm": 0.45005708932876587
     },
     {
       "step": 4075,
-      "loss": 1.4851957702636718,
-      "learning_rate": 2.6715664841233835e-05,
-      "grad_norm": 0.5958535671234131
+      "loss": 1.2932235717773437,
+      "learning_rate": 0.00012369304341345355,
+      "grad_norm": 0.45370563864707947
     },
     {
       "step": 4100,
-      "loss": 1.4916766357421876,
-      "learning_rate": 2.5334127843943163e-05,
-      "grad_norm": 0.7006895542144775
+      "loss": 1.214958953857422,
+      "learning_rate": 0.00012207160624870243,
+      "grad_norm": 0.45486775040626526
     },
     {
       "step": 4125,
-      "loss": 1.496282958984375,
-      "learning_rate": 2.398596691515871e-05,
-      "grad_norm": 0.7006640434265137
+      "loss": 1.2909408569335938,
+      "learning_rate": 0.00012045354269745233,
+      "grad_norm": 0.452974796295166
     },
     {
       "step": 4150,
-      "loss": 1.4691181945800782,
-      "learning_rate": 2.267154299016084e-05,
-      "grad_norm": 0.6565421223640442
+      "loss": 1.283765106201172,
+      "learning_rate": 0.00011883904821387391,
+      "grad_norm": 0.46668413281440735
     },
     {
       "step": 4175,
-      "loss": 1.5094065856933594,
-      "learning_rate": 2.139120797201822e-05,
-      "grad_norm": 0.6883091330528259
+      "loss": 1.238913803100586,
+      "learning_rate": 0.0001172283178210121,
+      "grad_norm": 0.4666079580783844
     },
     {
       "step": 4200,
-      "loss": 1.5348017883300782,
-      "learning_rate": 2.0145304637374755e-05,
-      "grad_norm": 0.6652461886405945
+      "loss": 1.303359375,
+      "learning_rate": 0.00011562154608722795,
+      "grad_norm": 0.4673307538032532
     },
     {
       "step": 4225,
-      "loss": 1.494925537109375,
-      "learning_rate": 1.893416654468022e-05,
-      "grad_norm": 0.7023136019706726
+      "loss": 1.2495006561279296,
+      "learning_rate": 0.00011401892710269586,
+      "grad_norm": 0.465811550617218
     },
     {
       "step": 4250,
-      "loss": 1.5202056884765625,
-      "learning_rate": 1.7758117944888416e-05,
-      "grad_norm": 0.6222552061080933
+      "loss": 1.263448257446289,
+      "learning_rate": 0.00011242065445595839,
+      "grad_norm": 0.43971243500709534
     },
     {
       "step": 4275,
-      "loss": 1.4469232177734375,
-      "learning_rate": 1.6617473694647354e-05,
-      "grad_norm": 0.6997081637382507
+      "loss": 1.2942657470703125,
+      "learning_rate": 0.00011082692121054178,
+      "grad_norm": 0.5017976760864258
     },
     {
       "step": 4300,
-      "loss": 1.4701220703125,
-      "learning_rate": 1.5512539172004596e-05,
-      "grad_norm": 1.2036304473876953
+      "loss": 1.2541777038574218,
+      "learning_rate": 0.00010923791988163473,
+      "grad_norm": 0.44632741808891296
     },
     {
       "step": 4325,
-      "loss": 1.4519557189941406,
-      "learning_rate": 1.4443610194650218e-05,
-      "grad_norm": 0.69020676612854
+      "loss": 1.2856007385253907,
+      "learning_rate": 0.0001076538424128336,
+      "grad_norm": 0.47356122732162476
     },
     {
       "step": 4350,
-      "loss": 1.4780242919921875,
-      "learning_rate": 1.3410972940719012e-05,
-      "grad_norm": 0.7094079852104187
+      "loss": 1.24027099609375,
+      "learning_rate": 0.00010607488015295649,
+      "grad_norm": 0.47505736351013184
     },
     {
       "step": 4375,
-      "loss": 1.5312185668945313,
-      "learning_rate": 1.2414903872173964e-05,
-      "grad_norm": 0.7318829298019409
+      "loss": 1.2527085113525391,
+      "learning_rate": 0.00010450122383292921,
+      "grad_norm": 0.4675339460372925
     },
     {
       "step": 4400,
-      "loss": 1.457879638671875,
-      "learning_rate": 1.1455669660790573e-05,
-      "grad_norm": 0.7195532917976379
+      "loss": 1.2761446380615233,
+      "learning_rate": 0.000102933063542746,
+      "grad_norm": 0.4455282390117645
     },
     {
       "step": 4425,
-      "loss": 1.4894285583496094,
-      "learning_rate": 1.0533527116762296e-05,
-      "grad_norm": 0.8031238913536072
+      "loss": 1.2683029174804688,
+      "learning_rate": 0.00010137058870850752,
+      "grad_norm": 0.4367285668849945
     },
     {
       "step": 4450,
-      "loss": 1.5307652282714843,
-      "learning_rate": 9.64872311994641e-06,
-      "grad_norm": 0.71385258436203
+      "loss": 1.2697316741943359,
+      "learning_rate": 9.981398806953917e-05,
+      "grad_norm": 0.4658321738243103
     },
     {
       "step": 4475,
-      "loss": 1.4505300903320313,
-      "learning_rate": 8.80149455376819e-06,
-      "grad_norm": 0.6250471472740173
+      "loss": 1.290264892578125,
+      "learning_rate": 9.826344965559224e-05,
+      "grad_norm": 0.48173364996910095
     },
     {
       "step": 4500,
-      "loss": 1.4776463317871094,
-      "learning_rate": 7.992068241801559e-06,
-      "grad_norm": 0.6997810006141663
+      "loss": 1.2587723541259765,
+      "learning_rate": 9.671916076413083e-05,
+      "grad_norm": 0.5190169215202332
     },
     {
       "step": 4525,
-      "loss": 1.51096923828125,
-      "learning_rate": 7.220660887042829e-06,
-      "grad_norm": 0.6750303506851196
+      "loss": 1.2330872344970703,
+      "learning_rate": 9.51813079377074e-05,
+      "grad_norm": 0.4717499315738678
     },
     {
       "step": 4550,
-      "loss": 1.5360499572753907,
-      "learning_rate": 6.4874790138942145e-06,
-      "grad_norm": 0.729939341545105
+      "loss": 1.2380457305908203,
+      "learning_rate": 9.365007694142903e-05,
+      "grad_norm": 0.42915889620780945
     },
     {
       "step": 4575,
-      "loss": 1.5128431701660157,
-      "learning_rate": 5.792718912872107e-06,
-      "grad_norm": 0.6567699909210205
+      "loss": 1.264954833984375,
+      "learning_rate": 9.212565274051817e-05,
+      "grad_norm": 0.4541546106338501
     },
     {
       "step": 4600,
-      "loss": 1.4098751831054688,
-      "learning_rate": 5.136566588055463e-06,
-      "grad_norm": 0.723537266254425
+      "loss": 1.2391783142089843,
+      "learning_rate": 9.060821947796955e-05,
+      "grad_norm": 0.4281995892524719
     },
     {
       "step": 4625,
-      "loss": 1.4862217712402344,
-      "learning_rate": 4.519197707287986e-06,
-      "grad_norm": 0.676175594329834
+      "loss": 1.2423888397216798,
+      "learning_rate": 8.909796045230688e-05,
+      "grad_norm": 0.445817768573761
     },
     {
       "step": 4650,
-      "loss": 1.4335507202148436,
-      "learning_rate": 3.940777555147567e-06,
-      "grad_norm": 0.6704904437065125
+      "loss": 1.2846461486816407,
+      "learning_rate": 8.759505809544096e-05,
+      "grad_norm": 0.4558369815349579
     },
     {
       "step": 4675,
-      "loss": 1.4516777038574218,
-      "learning_rate": 3.4014609886956477e-06,
-      "grad_norm": 0.6333931088447571
+      "loss": 1.2639895629882814,
+      "learning_rate": 8.609969395063299e-05,
+      "grad_norm": 0.4140309989452362
     },
     {
       "step": 4700,
-      "loss": 1.396420440673828,
-      "learning_rate": 2.9013923960182153e-06,
-      "grad_norm": 0.6822634935379028
+      "loss": 1.1922022247314452,
+      "learning_rate": 8.461204865056494e-05,
+      "grad_norm": 0.42347386479377747
     },
     {
       "step": 4725,
-      "loss": 1.465648193359375,
-      "learning_rate": 2.4407056575696083e-06,
-      "grad_norm": 0.7420686483383179
+      "loss": 1.2776814270019532,
+      "learning_rate": 8.313230189552016e-05,
+      "grad_norm": 0.4576684236526489
     },
     {
       "step": 4750,
-      "loss": 1.45216552734375,
-      "learning_rate": 2.0195241103295567e-06,
-      "grad_norm": 0.6106478571891785
+      "loss": 1.2211253356933593,
+      "learning_rate": 8.166063243167627e-05,
+      "grad_norm": 0.5280000567436218
     },
     {
       "step": 4775,
-      "loss": 1.4659315490722655,
-      "learning_rate": 1.63796051478276e-06,
-      "grad_norm": 0.7448365688323975
+      "loss": 1.2126011657714844,
+      "learning_rate": 8.019721802951368e-05,
+      "grad_norm": 0.45232927799224854
     },
     {
       "step": 4800,
-      "loss": 1.4934619140625,
-      "learning_rate": 1.2961170247303032e-06,
-      "grad_norm": 0.6592292189598083
+      "loss": 1.237452163696289,
+      "learning_rate": 7.874223546234149e-05,
+      "grad_norm": 0.4481607973575592
     },
     {
       "step": 4825,
-      "loss": 1.5013935852050782,
-      "learning_rate": 9.94085159940533e-07,
-      "grad_norm": 0.7112710475921631
+      "loss": 1.198594436645508,
+      "learning_rate": 7.72958604849445e-05,
+      "grad_norm": 0.45539891719818115
     },
     {
       "step": 4850,
-      "loss": 1.4725936889648437,
-      "learning_rate": 7.319457816470753e-07,
-      "grad_norm": 0.6786586046218872
+      "loss": 1.212102279663086,
+      "learning_rate": 7.585826781235274e-05,
+      "grad_norm": 0.45331281423568726
     },
     {
       "step": 4875,
-      "loss": 1.5387237548828125,
-      "learning_rate": 5.097690709002511e-07,
-      "grad_norm": 0.6456375122070312
+      "loss": 1.2540426635742188,
+      "learning_rate": 7.442963109873657e-05,
+      "grad_norm": 0.4342852830886841
     },
     {
       "step": 4900,
-      "loss": 1.512283172607422,
-      "learning_rate": 3.276145097779237e-07,
-      "grad_norm": 0.6460792422294617
+      "loss": 1.2529083251953126,
+      "learning_rate": 7.301012291643042e-05,
+      "grad_norm": 0.4284408688545227
     },
     {
       "step": 4925,
-      "loss": 1.4707493591308594,
-      "learning_rate": 1.8553086546080343e-07,
-      "grad_norm": 0.6637352705001831
+      "loss": 1.2232284545898438,
+      "learning_rate": 7.159991473508666e-05,
+      "grad_norm": 0.4916747808456421
     },
     {
       "step": 4950,
-      "loss": 1.5145166015625,
-      "learning_rate": 8.355617717617502e-08,
-      "grad_norm": 0.6858918070793152
+      "loss": 1.2357479858398437,
+      "learning_rate": 7.019917690096307e-05,
+      "grad_norm": 0.44929102063179016
     },
     {
       "step": 4975,
-      "loss": 1.4747088623046876,
-      "learning_rate": 2.171774601391063e-08,
-      "grad_norm": 0.6949462294578552
+      "loss": 1.2480144500732422,
+      "learning_rate": 6.880807861634563e-05,
+      "grad_norm": 0.4392186403274536
     },
     {
       "step": 5000,
-      "loss": 1.4937205505371094,
-      "learning_rate": 3.212761734983083e-11,
-      "grad_norm": 0.6537016034126282
+      "loss": 1.2522673797607422,
+      "learning_rate": 6.742678791910994e-05,
+      "grad_norm": 0.4787065386772156
+    },
+    {
+      "step": 5025,
+      "loss": 1.253096923828125,
+      "learning_rate": 6.6055471662423e-05,
+      "grad_norm": 0.4620949923992157
+    },
+    {
+      "step": 5050,
+      "loss": 1.2833746337890626,
+      "learning_rate": 6.469429549458793e-05,
+      "grad_norm": 0.4496528208255768
+    },
+    {
+      "step": 5075,
+      "loss": 1.2286983489990235,
+      "learning_rate": 6.334342383903497e-05,
+      "grad_norm": 0.4651983380317688
+    },
+    {
+      "step": 5100,
+      "loss": 1.2594075775146485,
+      "learning_rate": 6.200301987445924e-05,
+      "grad_norm": 0.49068841338157654
+    },
+    {
+      "step": 5125,
+      "loss": 1.2662586975097656,
+      "learning_rate": 6.067324551511042e-05,
+      "grad_norm": 0.4658633768558502
+    },
+    {
+      "step": 5150,
+      "loss": 1.2304222106933593,
+      "learning_rate": 5.935426139123334e-05,
+      "grad_norm": 0.44955214858055115
+    },
+    {
+      "step": 5175,
+      "loss": 1.229621353149414,
+      "learning_rate": 5.80462268296654e-05,
+      "grad_norm": 0.4408133029937744
+    },
+    {
+      "step": 5200,
+      "loss": 1.2133523559570312,
+      "learning_rate": 5.674929983459037e-05,
+      "grad_norm": 0.44262009859085083
+    },
+    {
+      "step": 5225,
+      "loss": 1.2221240234375,
+      "learning_rate": 5.546363706845204e-05,
+      "grad_norm": 0.43076950311660767
+    },
+    {
+      "step": 5250,
+      "loss": 1.2376617431640624,
+      "learning_rate": 5.418939383303067e-05,
+      "grad_norm": 0.45389309525489807
+    },
+    {
+      "step": 5275,
+      "loss": 1.2449069976806642,
+      "learning_rate": 5.2926724050682747e-05,
+      "grad_norm": 0.48846372961997986
+    },
+    {
+      "step": 5300,
+      "loss": 1.2519512939453126,
+      "learning_rate": 5.167578024574836e-05,
+      "grad_norm": 0.4938410520553589
+    },
+    {
+      "step": 5325,
+      "loss": 1.1897734832763671,
+      "learning_rate": 5.0436713526126756e-05,
+      "grad_norm": 0.4553002417087555
+    },
+    {
+      "step": 5350,
+      "loss": 1.232283477783203,
+      "learning_rate": 4.920967356502332e-05,
+      "grad_norm": 0.49487122893333435
+    },
+    {
+      "step": 5375,
+      "loss": 1.2140327453613282,
+      "learning_rate": 4.799480858286981e-05,
+      "grad_norm": 0.4383367896080017
+    },
+    {
+      "step": 5400,
+      "loss": 1.2149247741699218,
+      "learning_rate": 4.679226532941989e-05,
+      "grad_norm": 0.4817405939102173
+    },
+    {
+      "step": 5425,
+      "loss": 1.2574878692626954,
+      "learning_rate": 4.560218906602277e-05,
+      "grad_norm": 0.4749167859554291
+    },
+    {
+      "step": 5450,
+      "loss": 1.2151976013183594,
+      "learning_rate": 4.4424723548076014e-05,
+      "grad_norm": 0.49512583017349243
+    },
+    {
+      "step": 5475,
+      "loss": 1.226508560180664,
+      "learning_rate": 4.32600110076609e-05,
+      "grad_norm": 0.46990203857421875
+    },
+    {
+      "step": 5500,
+      "loss": 1.218799819946289,
+      "learning_rate": 4.210819213636126e-05,
+      "grad_norm": 0.4672454595565796
+    },
+    {
+      "step": 5525,
+      "loss": 1.2233743286132812,
+      "learning_rate": 4.096940606826865e-05,
+      "grad_norm": 0.45172497630119324
+    },
+    {
+      "step": 5550,
+      "loss": 1.194942626953125,
+      "learning_rate": 3.9843790363175865e-05,
+      "grad_norm": 0.4389333426952362
+    },
+    {
+      "step": 5575,
+      "loss": 1.2313622283935546,
+      "learning_rate": 3.8731480989959944e-05,
+      "grad_norm": 0.46159827709198
+    },
+    {
+      "step": 5600,
+      "loss": 1.1869064331054688,
+      "learning_rate": 3.763261231015831e-05,
+      "grad_norm": 0.4371712803840637
+    },
+    {
+      "step": 5625,
+      "loss": 1.2109555053710936,
+      "learning_rate": 3.6547317061738065e-05,
+      "grad_norm": 0.4667040705680847
+    },
+    {
+      "step": 5650,
+      "loss": 1.1982355499267578,
+      "learning_rate": 3.547572634306225e-05,
+      "grad_norm": 0.4180281460285187
+    },
+    {
+      "step": 5675,
+      "loss": 1.2502777099609375,
+      "learning_rate": 3.441796959705373e-05,
+      "grad_norm": 0.4566985070705414
+    },
+    {
+      "step": 5700,
+      "loss": 1.2031982421875,
+      "learning_rate": 3.3374174595558906e-05,
+      "grad_norm": 0.48893609642982483
+    },
+    {
+      "step": 5725,
+      "loss": 1.2035820007324218,
+      "learning_rate": 3.2344467423913764e-05,
+      "grad_norm": 0.47163310647010803
+    },
+    {
+      "step": 5750,
+      "loss": 1.2236875152587892,
+      "learning_rate": 3.1328972465713305e-05,
+      "grad_norm": 0.4129290282726288
+    },
+    {
+      "step": 5775,
+      "loss": 1.209072036743164,
+      "learning_rate": 3.032781238778656e-05,
+      "grad_norm": 0.5045214295387268
+    },
+    {
+      "step": 5800,
+      "loss": 1.2049720001220703,
+      "learning_rate": 2.9341108125378987e-05,
+      "grad_norm": 1.177187442779541
+    },
+    {
+      "step": 5825,
+      "loss": 1.2237082672119142,
+      "learning_rate": 2.836897886754428e-05,
+      "grad_norm": 0.44473758339881897
+    },
+    {
+      "step": 5850,
+      "loss": 1.211065444946289,
+      "learning_rate": 2.741154204274677e-05,
+      "grad_norm": 0.4674845039844513
+    },
+    {
+      "step": 5875,
+      "loss": 1.1834144592285156,
+      "learning_rate": 2.6468913304676588e-05,
+      "grad_norm": 0.4646168649196625
+    },
+    {
+      "step": 5900,
+      "loss": 1.2036126708984376,
+      "learning_rate": 2.5541206518279465e-05,
+      "grad_norm": 0.4711571931838989
+    },
+    {
+      "step": 5925,
+      "loss": 1.2229229736328124,
+      "learning_rate": 2.4628533746002134e-05,
+      "grad_norm": 0.47303342819213867
+    },
+    {
+      "step": 5950,
+      "loss": 1.2234363555908203,
+      "learning_rate": 2.37310052342561e-05,
+      "grad_norm": 0.429867684841156
+    },
+    {
+      "step": 5975,
+      "loss": 1.2309221649169921,
+      "learning_rate": 2.2848729400099936e-05,
+      "grad_norm": 0.5267869830131531
+    },
+    {
+      "step": 6000,
+      "loss": 1.2363741302490234,
+      "learning_rate": 2.198181281814349e-05,
+      "grad_norm": 0.4759523570537567
+    },
+    {
+      "step": 6025,
+      "loss": 1.220240478515625,
+      "learning_rate": 2.1130360207673923e-05,
+      "grad_norm": 0.4383280575275421
+    },
+    {
+      "step": 6050,
+      "loss": 1.1791091918945313,
+      "learning_rate": 2.0294474420006134e-05,
+      "grad_norm": 0.4518462121486664
+    },
+    {
+      "step": 6075,
+      "loss": 1.205836639404297,
+      "learning_rate": 1.9474256426059033e-05,
+      "grad_norm": 0.4878427982330322
+    },
+    {
+      "step": 6100,
+      "loss": 1.2177058410644532,
+      "learning_rate": 1.866980530415844e-05,
+      "grad_norm": 0.45281848311424255
+    },
+    {
+      "step": 6125,
+      "loss": 1.2058243560791015,
+      "learning_rate": 1.7881218228069215e-05,
+      "grad_norm": 0.467458039522171
+    },
+    {
+      "step": 6150,
+      "loss": 1.2168304443359375,
+      "learning_rate": 1.710859045525692e-05,
+      "grad_norm": 0.4621610939502716
+    },
+    {
+      "step": 6175,
+      "loss": 1.239758529663086,
+      "learning_rate": 1.6352015315381306e-05,
+      "grad_norm": 0.44325652718544006
+    },
+    {
+      "step": 6200,
+      "loss": 1.2028849029541016,
+      "learning_rate": 1.5611584199022474e-05,
+      "grad_norm": 0.44522371888160706
+    },
+    {
+      "step": 6225,
+      "loss": 1.2188538360595702,
+      "learning_rate": 1.488738654664135e-05,
+      "grad_norm": 0.4577750861644745
+    },
+    {
+      "step": 6250,
+      "loss": 1.2289091491699218,
+      "learning_rate": 1.4179509837775794e-05,
+      "grad_norm": 0.44208669662475586
+    },
+    {
+      "step": 6275,
+      "loss": 1.233956527709961,
+      "learning_rate": 1.348803958047333e-05,
+      "grad_norm": 0.4531864821910858
+    },
+    {
+      "step": 6300,
+      "loss": 1.2187274169921876,
+      "learning_rate": 1.281305930096243e-05,
+      "grad_norm": 0.4565826654434204
+    },
+    {
+      "step": 6325,
+      "loss": 1.2017286682128907,
+      "learning_rate": 1.215465053356272e-05,
+      "grad_norm": 0.47588875889778137
+    },
+    {
+      "step": 6350,
+      "loss": 1.219265899658203,
+      "learning_rate": 1.1512892810836188e-05,
+      "grad_norm": 0.4373761713504791
+    },
+    {
+      "step": 6375,
+      "loss": 1.2092100524902343,
+      "learning_rate": 1.0887863653980012e-05,
+      "grad_norm": 0.4859849214553833
+    },
+    {
+      "step": 6400,
+      "loss": 1.214217987060547,
+      "learning_rate": 1.0279638563462234e-05,
+      "grad_norm": 0.4610133469104767
+    },
+    {
+      "step": 6425,
+      "loss": 1.1977297973632812,
+      "learning_rate": 9.68829100990191e-06,
+      "grad_norm": 0.46569377183914185
+    },
+    {
+      "step": 6450,
+      "loss": 1.2287699127197265,
+      "learning_rate": 9.113892425194091e-06,
+      "grad_norm": 0.45601311326026917
+    },
+    {
+      "step": 6475,
+      "loss": 1.1981573486328125,
+      "learning_rate": 8.556512193881105e-06,
+      "grad_norm": 0.46404409408569336
+    },
+    {
+      "step": 6500,
+      "loss": 1.2465553283691406,
+      "learning_rate": 8.01621764477146e-06,
+      "grad_norm": 0.47190892696380615
+    },
+    {
+      "step": 6525,
+      "loss": 1.197098846435547,
+      "learning_rate": 7.4930740428066294e-06,
+      "grad_norm": 0.45320889353752136
+    },
+    {
+      "step": 6550,
+      "loss": 1.1829496002197266,
+      "learning_rate": 6.987144581177639e-06,
+      "grad_norm": 0.44584017992019653
+    },
+    {
+      "step": 6575,
+      "loss": 1.2278943634033204,
+      "learning_rate": 6.498490373691289e-06,
+      "grad_norm": 0.4520955979824066
+    },
+    {
+      "step": 6600,
+      "loss": 1.2047061157226562,
+      "learning_rate": 6.027170447388219e-06,
+      "grad_norm": 0.4502451717853546
+    },
+    {
+      "step": 6625,
+      "loss": 1.2211988830566407,
+      "learning_rate": 5.573241735412543e-06,
+      "grad_norm": 0.4682001769542694
+    },
+    {
+      "step": 6650,
+      "loss": 1.2268491363525391,
+      "learning_rate": 5.136759070134638e-06,
+      "grad_norm": 0.44889962673187256
+    },
+    {
+      "step": 6675,
+      "loss": 1.2038793182373047,
+      "learning_rate": 4.7177751765277775e-06,
+      "grad_norm": 0.44545120000839233
+    },
+    {
+      "step": 6700,
+      "loss": 1.23889892578125,
+      "learning_rate": 4.316340665799017e-06,
+      "grad_norm": 0.42067432403564453
+    },
+    {
+      "step": 6725,
+      "loss": 1.1800360107421874,
+      "learning_rate": 3.932504029275779e-06,
+      "grad_norm": 0.44353973865509033
+    },
+    {
+      "step": 6750,
+      "loss": 1.202306671142578,
+      "learning_rate": 3.5663116325482466e-06,
+      "grad_norm": 0.443708211183548
+    },
+    {
+      "step": 6775,
+      "loss": 1.1959423828125,
+      "learning_rate": 3.2178077098686304e-06,
+      "grad_norm": 0.4556666612625122
+    },
+    {
+      "step": 6800,
+      "loss": 1.1825554656982422,
+      "learning_rate": 2.88703435880796e-06,
+      "grad_norm": 0.45425671339035034
+    },
+    {
+      "step": 6825,
+      "loss": 1.1897866058349609,
+      "learning_rate": 2.5740315351707843e-06,
+      "grad_norm": 0.4512813091278076
+    },
+    {
+      "step": 6850,
+      "loss": 1.1878171539306641,
+      "learning_rate": 2.2788370481687965e-06,
+      "grad_norm": 0.43827131390571594
+    },
+    {
+      "step": 6875,
+      "loss": 1.2006049346923828,
+      "learning_rate": 2.0014865558536596e-06,
+      "grad_norm": 0.4461023211479187
+    },
+    {
+      "step": 6900,
+      "loss": 1.179116439819336,
+      "learning_rate": 1.7420135608096297e-06,
+      "grad_norm": 0.43480661511421204
+    },
+    {
+      "step": 6925,
+      "loss": 1.2001779937744141,
+      "learning_rate": 1.5004494061066797e-06,
+      "grad_norm": 0.4782886207103729
+    },
+    {
+      "step": 6950,
+      "loss": 1.2102194976806642,
+      "learning_rate": 1.27682327151436e-06,
+      "grad_norm": 0.4564678370952606
+    },
+    {
+      "step": 6975,
+      "loss": 1.174234848022461,
+      "learning_rate": 1.0711621699770522e-06,
+      "grad_norm": 0.4488446116447449
+    },
+    {
+      "step": 7000,
+      "loss": 1.1999745178222656,
+      "learning_rate": 8.834909443508953e-07,
+      "grad_norm": 0.458385169506073
+    },
+    {
+      "step": 7025,
+      "loss": 1.1988882446289062,
+      "learning_rate": 7.138322644029692e-07,
+      "grad_norm": 0.44504058361053467
+    },
+    {
+      "step": 7050,
+      "loss": 1.2174333953857421,
+      "learning_rate": 5.622066240728029e-07,
+      "grad_norm": 0.430634081363678
+    },
+    {
+      "step": 7075,
+      "loss": 1.2466523742675781,
+      "learning_rate": 4.286323389968871e-07,
+      "grad_norm": 0.4718645513057709
+    },
+    {
+      "step": 7100,
+      "loss": 1.216027069091797,
+      "learning_rate": 3.1312554429617776e-07,
+      "grad_norm": 0.466147243976593
+    },
+    {
+      "step": 7125,
+      "loss": 1.2029841613769532,
+      "learning_rate": 2.1570019262707185e-07,
+      "grad_norm": 0.45927485823631287
+    },
+    {
+      "step": 7150,
+      "loss": 1.1904071044921876,
+      "learning_rate": 1.3636805249603866e-07,
+      "grad_norm": 0.4847685992717743
+    },
+    {
+      "step": 7175,
+      "loss": 1.2437645721435546,
+      "learning_rate": 7.513870683794054e-08,
+      "grad_norm": 0.46799784898757935
+    },
+    {
+      "step": 7200,
+      "loss": 1.1522864532470702,
+      "learning_rate": 3.20195518585753e-08,
+      "grad_norm": 0.4408682882785797
+    },
+    {
+      "step": 7225,
+      "loss": 1.1922702026367187,
+      "learning_rate": 7.015796141157659e-09,
+      "grad_norm": 0.4327032268047333
     }
   ],
-  "final_checkpoint": "/content/agron-domain-run/final"
+  "final_checkpoint": "/content/agron-epoch1/final"
 }