diff --git "a/checkpoint-1100/trainer_state.json" "b/checkpoint-1100/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1100/trainer_state.json"
@@ -0,0 +1,7769 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9836512261580381,
+  "eval_steps": 184,
+  "global_step": 1100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0018165304268846503,
+      "grad_norm": 38.5,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.8686,
+      "step": 1
+    },
+    {
+      "epoch": 0.0018165304268846503,
+      "eval_loss": 0.866981029510498,
+      "eval_runtime": 34.9333,
+      "eval_samples_per_second": 5.725,
+      "eval_steps_per_second": 1.431,
+      "step": 1
+    },
+    {
+      "epoch": 0.0036330608537693005,
+      "grad_norm": 45.5,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.8814,
+      "step": 2
+    },
+    {
+      "epoch": 0.005449591280653951,
+      "grad_norm": 34.75,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.8713,
+      "step": 3
+    },
+    {
+      "epoch": 0.007266121707538601,
+      "grad_norm": 26.625,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.8797,
+      "step": 4
+    },
+    {
+      "epoch": 0.009082652134423252,
+      "grad_norm": 48.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.8693,
+      "step": 5
+    },
+    {
+      "epoch": 0.010899182561307902,
+      "grad_norm": 102.5,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.877,
+      "step": 6
+    },
+    {
+      "epoch": 0.012715712988192553,
+      "grad_norm": 73.5,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.8732,
+      "step": 7
+    },
+    {
+      "epoch": 0.014532243415077202,
+      "grad_norm": 143.0,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.8666,
+      "step": 8
+    },
+    {
+      "epoch": 0.01634877384196185,
+      "grad_norm": 58.75,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.8559,
+      "step": 9
+    },
+    {
+      "epoch": 0.018165304268846504,
+      "grad_norm": 56.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8493,
+      "step": 10
+    },
+    {
+      "epoch": 0.019981834695731154,
+      "grad_norm": 68.5,
+      "learning_rate": 2.2e-06,
+      "loss": 0.8313,
+      "step": 11
+    },
+    {
+      "epoch": 0.021798365122615803,
+      "grad_norm": 76.5,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.8413,
+      "step": 12
+    },
+    {
+      "epoch": 0.023614895549500452,
+      "grad_norm": 73.0,
+      "learning_rate": 2.6e-06,
+      "loss": 0.8287,
+      "step": 13
+    },
+    {
+      "epoch": 0.025431425976385105,
+      "grad_norm": 21.875,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.8349,
+      "step": 14
+    },
+    {
+      "epoch": 0.027247956403269755,
+      "grad_norm": 62.75,
+      "learning_rate": 3e-06,
+      "loss": 0.8052,
+      "step": 15
+    },
+    {
+      "epoch": 0.029064486830154404,
+      "grad_norm": 95.0,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.8093,
+      "step": 16
+    },
+    {
+      "epoch": 0.030881017257039057,
+      "grad_norm": 26.625,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.7839,
+      "step": 17
+    },
+    {
+      "epoch": 0.0326975476839237,
+      "grad_norm": 10.9375,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.796,
+      "step": 18
+    },
+    {
+      "epoch": 0.03451407811080836,
+      "grad_norm": 10.5,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.7889,
+      "step": 19
+    },
+    {
+      "epoch": 0.03633060853769301,
+      "grad_norm": 11.8125,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.776,
+      "step": 20
+    },
+    {
+      "epoch": 0.03814713896457766,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.7824,
+      "step": 21
+    },
+    {
+      "epoch": 0.03996366939146231,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.4e-06,
+      "loss": 0.7597,
+      "step": 22
+    },
+    {
+      "epoch": 0.04178019981834696,
+      "grad_norm": 16.25,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.7666,
+      "step": 23
+    },
+    {
+      "epoch": 0.043596730245231606,
+      "grad_norm": 11.75,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.7579,
+      "step": 24
+    },
+    {
+      "epoch": 0.045413260672116255,
+      "grad_norm": 67.5,
+      "learning_rate": 5e-06,
+      "loss": 0.7548,
+      "step": 25
+    },
+    {
+      "epoch": 0.047229791099000905,
+      "grad_norm": 78.5,
+      "learning_rate": 5.2e-06,
+      "loss": 0.754,
+      "step": 26
+    },
+    {
+      "epoch": 0.04904632152588556,
+      "grad_norm": 71.5,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 0.7631,
+      "step": 27
+    },
+    {
+      "epoch": 0.05086285195277021,
+      "grad_norm": 47.25,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.7508,
+      "step": 28
+    },
+    {
+      "epoch": 0.05267938237965486,
+      "grad_norm": 14.6875,
+      "learning_rate": 5.8e-06,
+      "loss": 0.7522,
+      "step": 29
+    },
+    {
+      "epoch": 0.05449591280653951,
+      "grad_norm": 7.5625,
+      "learning_rate": 6e-06,
+      "loss": 0.7282,
+      "step": 30
+    },
+    {
+      "epoch": 0.05631244323342416,
+      "grad_norm": 8.375,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 0.7432,
+      "step": 31
+    },
+    {
+      "epoch": 0.05812897366030881,
+      "grad_norm": 11.9375,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.7436,
+      "step": 32
+    },
+    {
+      "epoch": 0.05994550408719346,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 0.7403,
+      "step": 33
+    },
+    {
+      "epoch": 0.061762034514078114,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.7307,
+      "step": 34
+    },
+    {
+      "epoch": 0.06357856494096276,
+      "grad_norm": 2.171875,
+      "learning_rate": 7e-06,
+      "loss": 0.7297,
+      "step": 35
+    },
+    {
+      "epoch": 0.0653950953678474,
+      "grad_norm": 35.0,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.7254,
+      "step": 36
+    },
+    {
+      "epoch": 0.06721162579473206,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.4e-06,
+      "loss": 0.726,
+      "step": 37
+    },
+    {
+      "epoch": 0.06902815622161672,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.7124,
+      "step": 38
+    },
+    {
+      "epoch": 0.07084468664850137,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 0.7159,
+      "step": 39
+    },
+    {
+      "epoch": 0.07266121707538602,
+      "grad_norm": 5.78125,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.7216,
+      "step": 40
+    },
+    {
+      "epoch": 0.07447774750227067,
+      "grad_norm": 6.4375,
+      "learning_rate": 8.2e-06,
+      "loss": 0.7005,
+      "step": 41
+    },
+    {
+      "epoch": 0.07629427792915532,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.7049,
+      "step": 42
+    },
+    {
+      "epoch": 0.07811080835603997,
+      "grad_norm": 5.28125,
+      "learning_rate": 8.6e-06,
+      "loss": 0.6968,
+      "step": 43
+    },
+    {
+      "epoch": 0.07992733878292461,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.8e-06,
+      "loss": 0.6837,
+      "step": 44
+    },
+    {
+      "epoch": 0.08174386920980926,
+      "grad_norm": 2.25,
+      "learning_rate": 9e-06,
+      "loss": 0.6973,
+      "step": 45
+    },
+    {
+      "epoch": 0.08356039963669391,
+      "grad_norm": 24.25,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.6861,
+      "step": 46
+    },
+    {
+      "epoch": 0.08537693006357856,
+      "grad_norm": 14.375,
+      "learning_rate": 9.4e-06,
+      "loss": 0.6894,
+      "step": 47
+    },
+    {
+      "epoch": 0.08719346049046321,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.6908,
+      "step": 48
+    },
+    {
+      "epoch": 0.08900999091734786,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.696,
+      "step": 49
+    },
+    {
+      "epoch": 0.09082652134423251,
+      "grad_norm": 5.875,
+      "learning_rate": 1e-05,
+      "loss": 0.6732,
+      "step": 50
+    },
+    {
+      "epoch": 0.09264305177111716,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.02e-05,
+      "loss": 0.6765,
+      "step": 51
+    },
+    {
+      "epoch": 0.09445958219800181,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.04e-05,
+      "loss": 0.663,
+      "step": 52
+    },
+    {
+      "epoch": 0.09627611262488647,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.0600000000000002e-05,
+      "loss": 0.678,
+      "step": 53
+    },
+    {
+      "epoch": 0.09809264305177112,
+      "grad_norm": 4.8125,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 0.6655,
+      "step": 54
+    },
+    {
+      "epoch": 0.09990917347865577,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.6566,
+      "step": 55
+    },
+    {
+      "epoch": 0.10172570390554042,
+      "grad_norm": 2.375,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.6616,
+      "step": 56
+    },
+    {
+      "epoch": 0.10354223433242507,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.14e-05,
+      "loss": 0.6524,
+      "step": 57
+    },
+    {
+      "epoch": 0.10535876475930972,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.16e-05,
+      "loss": 0.6599,
+      "step": 58
+    },
+    {
+      "epoch": 0.10717529518619437,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.18e-05,
+      "loss": 0.6574,
+      "step": 59
+    },
+    {
+      "epoch": 0.10899182561307902,
+      "grad_norm": 3.0,
+      "learning_rate": 1.2e-05,
+      "loss": 0.6606,
+      "step": 60
+    },
+    {
+      "epoch": 0.11080835603996367,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.22e-05,
+      "loss": 0.6541,
+      "step": 61
+    },
+    {
+      "epoch": 0.11262488646684832,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 0.6331,
+      "step": 62
+    },
+    {
+      "epoch": 0.11444141689373297,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.6632,
+      "step": 63
+    },
+    {
+      "epoch": 0.11625794732061762,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.6609,
+      "step": 64
+    },
+    {
+      "epoch": 0.11807447774750227,
+      "grad_norm": 6.875,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.6529,
+      "step": 65
+    },
+    {
+      "epoch": 0.11989100817438691,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 0.6448,
+      "step": 66
+    },
+    {
+      "epoch": 0.12170753860127158,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 0.6516,
+      "step": 67
+    },
+    {
+      "epoch": 0.12352406902815623,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.6437,
+      "step": 68
+    },
+    {
+      "epoch": 0.12534059945504086,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.38e-05,
+      "loss": 0.635,
+      "step": 69
+    },
+    {
+      "epoch": 0.1271571298819255,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.4e-05,
+      "loss": 0.6374,
+      "step": 70
+    },
+    {
+      "epoch": 0.12897366030881016,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4200000000000001e-05,
+      "loss": 0.648,
+      "step": 71
+    },
+    {
+      "epoch": 0.1307901907356948,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 0.6387,
+      "step": 72
+    },
+    {
+      "epoch": 0.13260672116257946,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.46e-05,
+      "loss": 0.6249,
+      "step": 73
+    },
+    {
+      "epoch": 0.1344232515894641,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.48e-05,
+      "loss": 0.6266,
+      "step": 74
+    },
+    {
+      "epoch": 0.1362397820163488,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.6143,
+      "step": 75
+    },
+    {
+      "epoch": 0.13805631244323344,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 0.6152,
+      "step": 76
+    },
+    {
+      "epoch": 0.13987284287011809,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.54e-05,
+      "loss": 0.6254,
+      "step": 77
+    },
+    {
+      "epoch": 0.14168937329700274,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 0.6048,
+      "step": 78
+    },
+    {
+      "epoch": 0.14350590372388738,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.58e-05,
+      "loss": 0.6151,
+      "step": 79
+    },
+    {
+      "epoch": 0.14532243415077203,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.611,
+      "step": 80
+    },
+    {
+      "epoch": 0.14713896457765668,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.62e-05,
+      "loss": 0.6195,
+      "step": 81
+    },
+    {
+      "epoch": 0.14895549500454133,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.64e-05,
+      "loss": 0.6244,
+      "step": 82
+    },
+    {
+      "epoch": 0.15077202543142598,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.66e-05,
+      "loss": 0.6003,
+      "step": 83
+    },
+    {
+      "epoch": 0.15258855585831063,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.6004,
+      "step": 84
+    },
+    {
+      "epoch": 0.15440508628519528,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7e-05,
+      "loss": 0.6015,
+      "step": 85
+    },
+    {
+      "epoch": 0.15622161671207993,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.72e-05,
+      "loss": 0.6015,
+      "step": 86
+    },
+    {
+      "epoch": 0.15803814713896458,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7400000000000003e-05,
+      "loss": 0.596,
+      "step": 87
+    },
+    {
+      "epoch": 0.15985467756584923,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.76e-05,
+      "loss": 0.6109,
+      "step": 88
+    },
+    {
+      "epoch": 0.16167120799273388,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7800000000000002e-05,
+      "loss": 0.5912,
+      "step": 89
+    },
+    {
+      "epoch": 0.16348773841961853,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6066,
+      "step": 90
+    },
+    {
+      "epoch": 0.16530426884650318,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 0.6014,
+      "step": 91
+    },
+    {
+      "epoch": 0.16712079927338783,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 0.5858,
+      "step": 92
+    },
+    {
+      "epoch": 0.16893732970027248,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.86e-05,
+      "loss": 0.6087,
+      "step": 93
+    },
+    {
+      "epoch": 0.17075386012715713,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.88e-05,
+      "loss": 0.609,
+      "step": 94
+    },
+    {
+      "epoch": 0.17257039055404177,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9e-05,
+      "loss": 0.5747,
+      "step": 95
+    },
+    {
+      "epoch": 0.17438692098092642,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.5767,
+      "step": 96
+    },
+    {
+      "epoch": 0.17620345140781107,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.94e-05,
+      "loss": 0.5859,
+      "step": 97
+    },
+    {
+      "epoch": 0.17801998183469572,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.5856,
+      "step": 98
+    },
+    {
+      "epoch": 0.17983651226158037,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.98e-05,
+      "loss": 0.5988,
+      "step": 99
+    },
+    {
+      "epoch": 0.18165304268846502,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2e-05,
+      "loss": 0.5928,
+      "step": 100
+    },
+    {
+      "epoch": 0.18346957311534967,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9999950652018585e-05,
+      "loss": 0.5682,
+      "step": 101
+    },
+    {
+      "epoch": 0.18528610354223432,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.999980260856137e-05,
+      "loss": 0.5799,
+      "step": 102
+    },
+    {
+      "epoch": 0.18710263396911897,
+      "grad_norm": 1.75,
+      "learning_rate": 1.99995558710895e-05,
+      "loss": 0.5892,
+      "step": 103
+    },
+    {
+      "epoch": 0.18891916439600362,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9999210442038164e-05,
+      "loss": 0.5722,
+      "step": 104
+    },
+    {
+      "epoch": 0.1907356948228883,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9998766324816606e-05,
+      "loss": 0.573,
+      "step": 105
+    },
+    {
+      "epoch": 0.19255222524977295,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9998223523808092e-05,
+      "loss": 0.5635,
+      "step": 106
+    },
+    {
+      "epoch": 0.1943687556766576,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9997582044369843e-05,
+      "loss": 0.5743,
+      "step": 107
+    },
+    {
+      "epoch": 0.19618528610354224,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9996841892833e-05,
+      "loss": 0.5653,
+      "step": 108
+    },
+    {
+      "epoch": 0.1980018165304269,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9996003076502567e-05,
+      "loss": 0.5792,
+      "step": 109
+    },
+    {
+      "epoch": 0.19981834695731154,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9995065603657317e-05,
+      "loss": 0.5746,
+      "step": 110
+    },
+    {
+      "epoch": 0.2016348773841962,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9994029483549732e-05,
+      "loss": 0.5647,
+      "step": 111
+    },
+    {
+      "epoch": 0.20345140781108084,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9992894726405894e-05,
+      "loss": 0.5647,
+      "step": 112
+    },
+    {
+      "epoch": 0.2052679382379655,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9991661343425402e-05,
+      "loss": 0.5704,
+      "step": 113
+    },
+    {
+      "epoch": 0.20708446866485014,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.999032934678125e-05,
+      "loss": 0.5676,
+      "step": 114
+    },
+    {
+      "epoch": 0.2089009990917348,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9988898749619702e-05,
+      "loss": 0.568,
+      "step": 115
+    },
+    {
+      "epoch": 0.21071752951861944,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.998736956606018e-05,
+      "loss": 0.5689,
+      "step": 116
+    },
+    {
+      "epoch": 0.2125340599455041,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9985741811195098e-05,
+      "loss": 0.5741,
+      "step": 117
+    },
+    {
+      "epoch": 0.21435059037238874,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.998401550108975e-05,
+      "loss": 0.569,
+      "step": 118
+    },
+    {
+      "epoch": 0.2161671207992734,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9982190652782122e-05,
+      "loss": 0.5636,
+      "step": 119
+    },
+    {
+      "epoch": 0.21798365122615804,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9980267284282718e-05,
+      "loss": 0.5531,
+      "step": 120
+    },
+    {
+      "epoch": 0.2198001816530427,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.997824541457442e-05,
+      "loss": 0.5594,
+      "step": 121
+    },
+    {
+      "epoch": 0.22161671207992734,
+      "grad_norm": 1.125,
+      "learning_rate": 1.9976125063612254e-05,
+      "loss": 0.5459,
+      "step": 122
+    },
+    {
+      "epoch": 0.22343324250681199,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.997390625232324e-05,
+      "loss": 0.5517,
+      "step": 123
+    },
+    {
+      "epoch": 0.22524977293369663,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.997158900260614e-05,
+      "loss": 0.5602,
+      "step": 124
+    },
+    {
+      "epoch": 0.22706630336058128,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.9969173337331283e-05,
+      "loss": 0.5568,
+      "step": 125
+    },
+    {
+      "epoch": 0.22888283378746593,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.99666592803403e-05,
+      "loss": 0.5667,
+      "step": 126
+    },
+    {
+      "epoch": 0.23069936421435058,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9964046856445926e-05,
+      "loss": 0.5614,
+      "step": 127
+    },
+    {
+      "epoch": 0.23251589464123523,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.9961336091431728e-05,
+      "loss": 0.5445,
+      "step": 128
+    },
+    {
+      "epoch": 0.23433242506811988,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.995852701205186e-05,
+      "loss": 0.5668,
+      "step": 129
+    },
+    {
+      "epoch": 0.23614895549500453,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.99556196460308e-05,
+      "loss": 0.5474,
+      "step": 130
+    },
+    {
+      "epoch": 0.23796548592188918,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9952614022063085e-05,
+      "loss": 0.5516,
+      "step": 131
+    },
+    {
+      "epoch": 0.23978201634877383,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.9949510169813006e-05,
+      "loss": 0.5472,
+      "step": 132
+    },
+    {
+      "epoch": 0.24159854677565848,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9946308119914323e-05,
+      "loss": 0.5502,
+      "step": 133
+    },
+    {
+      "epoch": 0.24341507720254316,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.994300790396999e-05,
+      "loss": 0.5493,
+      "step": 134
+    },
+    {
+      "epoch": 0.2452316076294278,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.99396095545518e-05,
+      "loss": 0.5548,
+      "step": 135
+    },
+    {
+      "epoch": 0.24704813805631246,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9936113105200085e-05,
+      "loss": 0.5576,
+      "step": 136
+    },
+    {
+      "epoch": 0.2488646684831971,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9932518590423396e-05,
+      "loss": 0.5506,
+      "step": 137
+    },
+    {
+      "epoch": 0.2506811989100817,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9928826045698138e-05,
+      "loss": 0.5462,
+      "step": 138
+    },
+    {
+      "epoch": 0.2524977293369664,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.992503550746824e-05,
+      "loss": 0.5623,
+      "step": 139
+    },
+    {
+      "epoch": 0.254314259763851,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.5529,
+      "step": 140
+    },
+    {
+      "epoch": 0.2561307901907357,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9917160601105632e-05,
+      "loss": 0.55,
+      "step": 141
+    },
+    {
+      "epoch": 0.2579473206176203,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9913076310695068e-05,
+      "loss": 0.5379,
+      "step": 142
+    },
+    {
+      "epoch": 0.259763851044505,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.990889418222339e-05,
+      "loss": 0.5436,
+      "step": 143
+    },
+    {
+      "epoch": 0.2615803814713896,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9904614256966514e-05,
+      "loss": 0.5423,
+      "step": 144
+    },
+    {
+      "epoch": 0.2633969118982743,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.990023657716558e-05,
+      "loss": 0.5336,
+      "step": 145
+    },
+    {
+      "epoch": 0.2652134423251589,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.989576118602651e-05,
+      "loss": 0.5353,
+      "step": 146
+    },
+    {
+      "epoch": 0.2670299727520436,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.989118812771962e-05,
+      "loss": 0.5512,
+      "step": 147
+    },
+    {
+      "epoch": 0.2688465031789282,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.988651744737914e-05,
+      "loss": 0.5402,
+      "step": 148
+    },
+    {
+      "epoch": 0.2706630336058129,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9881749191102807e-05,
+      "loss": 0.5418,
+      "step": 149
+    },
+    {
+      "epoch": 0.2724795640326976,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 0.5424,
+      "step": 150
+    },
+    {
+      "epoch": 0.2742960944595822,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9871920139948193e-05,
+      "loss": 0.522,
+      "step": 151
+    },
+    {
+      "epoch": 0.2761126248864669,
+      "grad_norm": 1.125,
+      "learning_rate": 1.986685944207868e-05,
+      "loss": 0.5352,
+      "step": 152
+    },
+    {
+      "epoch": 0.2779291553133515,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9861701362289892e-05,
+      "loss": 0.5331,
+      "step": 153
+    },
+    {
+      "epoch": 0.27974568574023617,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9856445951489984e-05,
+      "loss": 0.5345,
+      "step": 154
+    },
+    {
+      "epoch": 0.2815622161671208,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.985109326154774e-05,
+      "loss": 0.5413,
+      "step": 155
+    },
+    {
+      "epoch": 0.28337874659400547,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9845643345292055e-05,
+      "loss": 0.5436,
+      "step": 156
+    },
+    {
+      "epoch": 0.2851952770208901,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9840096256511398e-05,
+      "loss": 0.5292,
+      "step": 157
+    },
+    {
+      "epoch": 0.28701180744777477,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.98344520499533e-05,
+      "loss": 0.5216,
+      "step": 158
+    },
+    {
+      "epoch": 0.2888283378746594,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9828710781323793e-05,
+      "loss": 0.5282,
+      "step": 159
+    },
+    {
+      "epoch": 0.29064486830154407,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.982287250728689e-05,
+      "loss": 0.5299,
+      "step": 160
+    },
+    {
+      "epoch": 0.2924613987284287,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.9816937285463992e-05,
+      "loss": 0.5237,
+      "step": 161
+    },
+    {
+      "epoch": 0.29427792915531337,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.981090517443334e-05,
+      "loss": 0.5231,
+      "step": 162
+    },
+    {
+      "epoch": 0.296094459582198,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9804776233729446e-05,
+      "loss": 0.544,
+      "step": 163
+    },
+    {
+      "epoch": 0.29791099000908267,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.979855052384247e-05,
+      "loss": 0.5439,
+      "step": 164
+    },
+    {
+      "epoch": 0.2997275204359673,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.979222810621766e-05,
+      "loss": 0.5332,
+      "step": 165
+    },
+    {
+      "epoch": 0.30154405086285196,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.978580904325472e-05,
+      "loss": 0.5308,
+      "step": 166
+    },
+    {
+      "epoch": 0.3033605812897366,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.977929339830722e-05,
+      "loss": 0.5259,
+      "step": 167
+    },
+    {
+      "epoch": 0.30517711171662126,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9772681235681936e-05,
+      "loss": 0.5262,
+      "step": 168
+    },
+    {
+      "epoch": 0.3069936421435059,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.976597262063825e-05,
+      "loss": 0.5348,
+      "step": 169
+    },
+    {
+      "epoch": 0.30881017257039056,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9759167619387474e-05,
+      "loss": 0.5412,
+      "step": 170
+    },
+    {
+      "epoch": 0.3106267029972752,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9752266299092234e-05,
+      "loss": 0.5194,
+      "step": 171
+    },
+    {
+      "epoch": 0.31244323342415986,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9745268727865774e-05,
+      "loss": 0.5355,
+      "step": 172
+    },
+    {
+      "epoch": 0.3142597638510445,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9738174974771288e-05,
+      "loss": 0.5241,
+      "step": 173
+    },
+    {
+      "epoch": 0.31607629427792916,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9730985109821268e-05,
+      "loss": 0.5317,
+      "step": 174
+    },
+    {
+      "epoch": 0.3178928247048138,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9723699203976768e-05,
+      "loss": 0.5216,
+      "step": 175
+    },
+    {
+      "epoch": 0.31970935513169846,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.971631732914674e-05,
+      "loss": 0.5123,
+      "step": 176
+    },
+    {
+      "epoch": 0.3215258855585831,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9708839558187313e-05,
+      "loss": 0.5209,
+      "step": 177
+    },
+    {
+      "epoch": 0.32334241598546776,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.970126596490106e-05,
+      "loss": 0.5271,
+      "step": 178
+    },
+    {
+      "epoch": 0.32515894641235243,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.9693596624036294e-05,
+      "loss": 0.5136,
+      "step": 179
+    },
+    {
+      "epoch": 0.32697547683923706,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.5257,
+      "step": 180
+    },
+    {
+      "epoch": 0.32879200726612173,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9677971003288657e-05,
+      "loss": 0.5186,
+      "step": 181
+    },
+    {
+      "epoch": 0.33060853769300635,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9670014877624353e-05,
+      "loss": 0.5198,
+      "step": 182
+    },
+    {
+      "epoch": 0.33242506811989103,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.966196331281715e-05,
+      "loss": 0.5207,
+      "step": 183
+    },
+    {
+      "epoch": 0.33424159854677565,
+      "grad_norm": 1.25,
+      "learning_rate": 1.965381638833274e-05,
+      "loss": 0.514,
+      "step": 184
+    },
+    {
+      "epoch": 0.33424159854677565,
+      "eval_loss": 0.5189763903617859,
+      "eval_runtime": 34.9275,
+      "eval_samples_per_second": 5.726,
+      "eval_steps_per_second": 1.432,
+      "step": 184
+    },
+    {
+      "epoch": 0.33605812897366033,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9645574184577982e-05,
+      "loss": 0.5245,
+      "step": 185
+    },
+    {
+      "epoch": 0.33787465940054495,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.96372367829001e-05,
+      "loss": 0.5097,
+      "step": 186
+    },
+    {
+      "epoch": 0.33969118982742963,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9628804265585878e-05,
+      "loss": 0.5277,
+      "step": 187
+    },
+    {
+      "epoch": 0.34150772025431425,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.962027671586086e-05,
+      "loss": 0.5154,
+      "step": 188
+    },
+    {
+      "epoch": 0.34332425068119893,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.961165421788852e-05,
+      "loss": 0.5241,
+      "step": 189
+    },
+    {
+      "epoch": 0.34514078110808355,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9602936856769432e-05,
+      "loss": 0.5145,
+      "step": 190
+    },
+    {
+      "epoch": 0.3469573115349682,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.959412471854043e-05,
+      "loss": 0.5147,
+      "step": 191
+    },
+    {
+      "epoch": 0.34877384196185285,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.958521789017376e-05,
+      "loss": 0.5267,
+      "step": 192
+    },
+    {
+      "epoch": 0.3505903723887375,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9576216459576222e-05,
+      "loss": 0.5065,
+      "step": 193
+    },
+    {
+      "epoch": 0.35240690281562215,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.9567120515588307e-05,
+      "loss": 0.5087,
+      "step": 194
+    },
+    {
+      "epoch": 0.3542234332425068,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9557930147983303e-05,
+      "loss": 0.524,
+      "step": 195
+    },
+    {
+      "epoch": 0.35603996366939145,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9548645447466433e-05,
+      "loss": 0.5176,
+      "step": 196
+    },
+    {
+      "epoch": 0.3578564940962761,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9539266505673938e-05,
+      "loss": 0.5081,
+      "step": 197
+    },
+    {
+      "epoch": 0.35967302452316074,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.952979341517219e-05,
+      "loss": 0.5296,
+      "step": 198
+    },
+    {
+      "epoch": 0.3614895549500454,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9520226269456767e-05,
+      "loss": 0.4961,
+      "step": 199
+    },
+    {
+      "epoch": 0.36330608537693004,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.5091,
+      "step": 200
+    },
+    {
+      "epoch": 0.3651226158038147,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9500810191007717e-05,
+      "loss": 0.5169,
+      "step": 201
+    },
+    {
+      "epoch": 0.36693914623069934,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9490961449902946e-05,
+      "loss": 0.5044,
+      "step": 202
+    },
+    {
+      "epoch": 0.368755676657584,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.948101903684032e-05,
+      "loss": 0.5146,
+      "step": 203
+    },
+    {
+      "epoch": 0.37057220708446864,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9470983049947446e-05,
+      "loss": 0.5171,
+      "step": 204
+    },
+    {
+      "epoch": 0.3723887375113533,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9460853588275454e-05,
+      "loss": 0.5081,
+      "step": 205
+    },
+    {
+      "epoch": 0.37420526793823794,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.945063075179805e-05,
+      "loss": 0.5098,
+      "step": 206
+    },
+    {
+      "epoch": 0.3760217983651226,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.94403146414105e-05,
+      "loss": 0.5025,
+      "step": 207
+    },
+    {
+      "epoch": 0.37783832879200724,
+      "grad_norm": 1.125,
+      "learning_rate": 1.9429905358928648e-05,
+      "loss": 0.5231,
+      "step": 208
+    },
+    {
+      "epoch": 0.3796548592188919,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9419403007087908e-05,
+      "loss": 0.505,
+      "step": 209
+    },
+    {
+      "epoch": 0.3814713896457766,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9408807689542257e-05,
+      "loss": 0.4997,
+      "step": 210
+    },
+    {
+      "epoch": 0.3832879200726612,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.9398119510863197e-05,
+      "loss": 0.5099,
+      "step": 211
+    },
+    {
+      "epoch": 0.3851044504995459,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.9387338576538743e-05,
+      "loss": 0.5036,
+      "step": 212
+    },
+    {
+      "epoch": 0.3869209809264305,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9376464992972358e-05,
+      "loss": 0.5158,
+      "step": 213
+    },
+    {
+      "epoch": 0.3887375113533152,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9365498867481926e-05,
+      "loss": 0.5072,
+      "step": 214
+    },
+    {
+      "epoch": 0.3905540417801998,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.9354440308298676e-05,
+      "loss": 0.5058,
+      "step": 215
+    },
+    {
+      "epoch": 0.3923705722070845,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9343289424566122e-05,
+      "loss": 0.4971,
+      "step": 216
+    },
+    {
+      "epoch": 0.3941871026339691,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.9332046326338985e-05,
+      "loss": 0.504,
+      "step": 217
+    },
+    {
+      "epoch": 0.3960036330608538,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.932071112458211e-05,
+      "loss": 0.5174,
+      "step": 218
+    },
+    {
+      "epoch": 0.3978201634877384,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.930928393116936e-05,
+      "loss": 0.5216,
+      "step": 219
+    },
+    {
+      "epoch": 0.3996366939146231,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 0.4991,
+      "step": 220
+    },
+    {
+      "epoch": 0.4014532243415077,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9286154021410177e-05,
+      "loss": 0.4929,
+      "step": 221
+    },
+    {
+      "epoch": 0.4032697547683924,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9274451533346617e-05,
+      "loss": 0.507,
+      "step": 222
+    },
+    {
+      "epoch": 0.405086285195277,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.926265751019067e-05,
+      "loss": 0.4942,
+      "step": 223
+    },
+    {
+      "epoch": 0.4069028156221617,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.925077206834458e-05,
+      "loss": 0.4991,
+      "step": 224
+    },
+    {
+      "epoch": 0.4087193460490463,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 0.5008,
+      "step": 225
+    },
+    {
+      "epoch": 0.410535876475931,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.922672739870115e-05,
+      "loss": 0.4975,
+      "step": 226
+    },
+    {
+      "epoch": 0.4123524069028156,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9214568408214986e-05,
+      "loss": 0.4997,
+      "step": 227
+    },
+    {
+      "epoch": 0.4141689373297003,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9202318473658707e-05,
+      "loss": 0.5072,
+      "step": 228
+    },
+    {
+      "epoch": 0.4159854677565849,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9189977715934214e-05,
+      "loss": 0.5099,
+      "step": 229
+    },
+    {
+      "epoch": 0.4178019981834696,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9177546256839814e-05,
+      "loss": 0.508,
+      "step": 230
+    },
+    {
+      "epoch": 0.4196185286103542,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.916502421906898e-05,
+      "loss": 0.5061,
+      "step": 231
+    },
+    {
+      "epoch": 0.4214350590372389,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.9152411726209176e-05,
+      "loss": 0.511,
+      "step": 232
+    },
+    {
+      "epoch": 0.4232515894641235,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.913970890274061e-05,
+      "loss": 0.4973,
+      "step": 233
+    },
+    {
+      "epoch": 0.4250681198910082,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.912691587403503e-05,
+      "loss": 0.5123,
+      "step": 234
+    },
+    {
+      "epoch": 0.4268846503178928,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9114032766354453e-05,
+      "loss": 0.5033,
+      "step": 235
+    },
+    {
+      "epoch": 0.4287011807447775,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9101059706849957e-05,
+      "loss": 0.5,
+      "step": 236
+    },
+    {
+      "epoch": 0.4305177111716621,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9087996823560404e-05,
+      "loss": 0.4999,
+      "step": 237
+    },
+    {
+      "epoch": 0.4323342415985468,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.907484424541117e-05,
+      "loss": 0.5152,
+      "step": 238
+    },
+    {
+      "epoch": 0.43415077202543145,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.9061602102212898e-05,
+      "loss": 0.4996,
+      "step": 239
+    },
+    {
+      "epoch": 0.4359673024523161,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9048270524660197e-05,
+      "loss": 0.4941,
+      "step": 240
+    },
+    {
+      "epoch": 0.43778383287920075,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.903484964433035e-05,
+      "loss": 0.4895,
+      "step": 241
+    },
+    {
+      "epoch": 0.4396003633060854,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.902133959368203e-05,
+      "loss": 0.5014,
+      "step": 242
+    },
+    {
+      "epoch": 0.44141689373297005,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.9007740506053983e-05,
+      "loss": 0.4971,
+      "step": 243
+    },
+    {
+      "epoch": 0.44323342415985467,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.899405251566371e-05,
+      "loss": 0.4969,
+      "step": 244
+    },
+    {
+      "epoch": 0.44504995458673935,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.8980275757606157e-05,
+      "loss": 0.4883,
+      "step": 245
+    },
+    {
+      "epoch": 0.44686648501362397,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.896641036785236e-05,
+      "loss": 0.5083,
+      "step": 246
+    },
+    {
+      "epoch": 0.44868301544050865,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.8952456483248117e-05,
+      "loss": 0.4947,
+      "step": 247
+    },
+    {
+      "epoch": 0.45049954586739327,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.893841424151264e-05,
+      "loss": 0.5,
+      "step": 248
+    },
+    {
+      "epoch": 0.45231607629427795,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.892428378123718e-05,
+      "loss": 0.4889,
+      "step": 249
+    },
+    {
+      "epoch": 0.45413260672116257,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.4956,
+      "step": 250
+    },
+    {
+      "epoch": 0.45594913714804725,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8895758763783383e-05,
+      "loss": 0.5052,
+      "step": 251
+    },
+    {
+      "epoch": 0.45776566757493187,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8881364488135448e-05,
+      "loss": 0.4849,
+      "step": 252
+    },
+    {
+      "epoch": 0.45958219800181654,
+      "grad_norm": 1.25,
+      "learning_rate": 1.8866882557005567e-05,
+      "loss": 0.4914,
+      "step": 253
+    },
+    {
+      "epoch": 0.46139872842870117,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.8852313113324553e-05,
+      "loss": 0.5086,
+      "step": 254
+    },
+    {
+      "epoch": 0.46321525885558584,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.8837656300886937e-05,
+      "loss": 0.5043,
+      "step": 255
+    },
+    {
+      "epoch": 0.46503178928247046,
+      "grad_norm": 1.125,
+      "learning_rate": 1.8822912264349535e-05,
+      "loss": 0.4941,
+      "step": 256
+    },
+    {
+      "epoch": 0.46684831970935514,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8808081149230036e-05,
+      "loss": 0.497,
+      "step": 257
+    },
+    {
+      "epoch": 0.46866485013623976,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.8793163101905562e-05,
+      "loss": 0.4948,
+      "step": 258
+    },
+    {
+      "epoch": 0.47048138056312444,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.877815826961122e-05,
+      "loss": 0.481,
+      "step": 259
+    },
+    {
+      "epoch": 0.47229791099000906,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 0.4863,
+      "step": 260
+    },
+    {
+      "epoch": 0.47411444141689374,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.8747888843334528e-05,
+      "loss": 0.4854,
+      "step": 261
+    },
+    {
+      "epoch": 0.47593097184377836,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.8732624548099204e-05,
+      "loss": 0.487,
+      "step": 262
+    },
+    {
+      "epoch": 0.47774750227066304,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.8717274065385092e-05,
+      "loss": 0.4972,
+      "step": 263
+    },
+    {
+      "epoch": 0.47956403269754766,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.870183754669526e-05,
+      "loss": 0.4982,
+      "step": 264
+    },
+    {
+      "epoch": 0.48138056312443234,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8686315144381914e-05,
+      "loss": 0.5032,
+      "step": 265
+    },
+    {
+      "epoch": 0.48319709355131696,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.86707070116449e-05,
+      "loss": 0.4951,
+      "step": 266
+    },
+    {
+      "epoch": 0.48501362397820164,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8655013302530193e-05,
+      "loss": 0.487,
+      "step": 267
+    },
+    {
+      "epoch": 0.4868301544050863,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.8639234171928355e-05,
+      "loss": 0.4897,
+      "step": 268
+    },
+    {
+      "epoch": 0.48864668483197093,
+      "grad_norm": 1.375,
+      "learning_rate": 1.862336977557304e-05,
+      "loss": 0.4806,
+      "step": 269
+    },
+    {
+      "epoch": 0.4904632152588556,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.860742027003944e-05,
+      "loss": 0.4892,
+      "step": 270
+    },
+    {
+      "epoch": 0.49227974568574023,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8591385812742724e-05,
+      "loss": 0.4881,
+      "step": 271
+    },
+    {
+      "epoch": 0.4940962761126249,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8575266561936526e-05,
+      "loss": 0.4903,
+      "step": 272
+    },
+    {
+      "epoch": 0.49591280653950953,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.855906267671133e-05,
+      "loss": 0.4995,
+      "step": 273
+    },
+    {
+      "epoch": 0.4977293369663942,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.8542774316992953e-05,
+      "loss": 0.4713,
+      "step": 274
+    },
+    {
+      "epoch": 0.49954586739327883,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.8526401643540924e-05,
+      "loss": 0.4769,
+      "step": 275
+    },
+    {
+      "epoch": 0.5013623978201635,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.850994481794692e-05,
+      "loss": 0.4778,
+      "step": 276
+    },
+    {
+      "epoch": 0.5031789282470481,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.8493404002633167e-05,
+      "loss": 0.4859,
+      "step": 277
+    },
+    {
+      "epoch": 0.5049954586739328,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.8476779360850833e-05,
+      "loss": 0.4899,
+      "step": 278
+    },
+    {
+      "epoch": 0.5068119891008175,
+      "grad_norm": 1.0,
+      "learning_rate": 1.8460071056678424e-05,
+      "loss": 0.4879,
+      "step": 279
+    },
+    {
+      "epoch": 0.508628519527702,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8443279255020153e-05,
+      "loss": 0.485,
+      "step": 280
+    },
+    {
+      "epoch": 0.5104450499545867,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8426404121604324e-05,
+      "loss": 0.4809,
+      "step": 281
+    },
+    {
+      "epoch": 0.5122615803814714,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.8409445822981694e-05,
+      "loss": 0.4929,
+      "step": 282
+    },
+    {
+      "epoch": 0.5140781108083561,
+      "grad_norm": 1.0,
+      "learning_rate": 1.8392404526523816e-05,
+      "loss": 0.4864,
+      "step": 283
+    },
+    {
+      "epoch": 0.5158946412352406,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.837528040042142e-05,
+      "loss": 0.4846,
+      "step": 284
+    },
+    {
+      "epoch": 0.5177111716621253,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.8358073613682705e-05,
+      "loss": 0.47,
+      "step": 285
+    },
+    {
+      "epoch": 0.51952770208901,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.8340784336131715e-05,
+      "loss": 0.4803,
+      "step": 286
+    },
+    {
+      "epoch": 0.5213442325158947,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.8323412738406638e-05,
+      "loss": 0.4951,
+      "step": 287
+    },
+    {
+      "epoch": 0.5231607629427792,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.830595899195813e-05,
+      "loss": 0.4787,
+      "step": 288
+    },
+    {
+      "epoch": 0.5249772933696639,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.828842326904762e-05,
+      "loss": 0.4841,
+      "step": 289
+    },
+    {
+      "epoch": 0.5267938237965486,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.827080574274562e-05,
+      "loss": 0.4903,
+      "step": 290
+    },
+    {
+      "epoch": 0.5286103542234333,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.825310658693e-05,
+      "loss": 0.4875,
+      "step": 291
+    },
+    {
+      "epoch": 0.5304268846503178,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8235325976284276e-05,
+      "loss": 0.4772,
+      "step": 292
+    },
+    {
+      "epoch": 0.5322434150772025,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.8217464086295904e-05,
+      "loss": 0.4805,
+      "step": 293
+    },
+    {
+      "epoch": 0.5340599455040872,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.8199521093254524e-05,
+      "loss": 0.4964,
+      "step": 294
+    },
+    {
+      "epoch": 0.5358764759309719,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.8181497174250236e-05,
+      "loss": 0.486,
+      "step": 295
+    },
+    {
+      "epoch": 0.5376930063578564,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.816339250717184e-05,
+      "loss": 0.4828,
+      "step": 296
+    },
+    {
+      "epoch": 0.5395095367847411,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8145207270705095e-05,
+      "loss": 0.4772,
+      "step": 297
+    },
+    {
+      "epoch": 0.5413260672116258,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.812694164433094e-05,
+      "loss": 0.4739,
+      "step": 298
+    },
+    {
+      "epoch": 0.5431425976385105,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8108595808323736e-05,
+      "loss": 0.4746,
+      "step": 299
+    },
+    {
+      "epoch": 0.5449591280653951,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.4777,
+      "step": 300
+    },
+    {
+      "epoch": 0.5467756584922797,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8071664232464005e-05,
+      "loss": 0.4825,
+      "step": 301
+    },
+    {
+      "epoch": 0.5485921889191644,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.8053078857111218e-05,
+      "loss": 0.4796,
+      "step": 302
+    },
+    {
+      "epoch": 0.5504087193460491,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8034414001121278e-05,
+      "loss": 0.4947,
+      "step": 303
+    },
+    {
+      "epoch": 0.5522252497729337,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8015669848708768e-05,
+      "loss": 0.4741,
+      "step": 304
+    },
+    {
+      "epoch": 0.5540417801998183,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.799684658487091e-05,
+      "loss": 0.4742,
+      "step": 305
+    },
+    {
+      "epoch": 0.555858310626703,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.7977944395385713e-05,
+      "loss": 0.4848,
+      "step": 306
+    },
+    {
+      "epoch": 0.5576748410535877,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.795896346681016e-05,
+      "loss": 0.4763,
+      "step": 307
+    },
+    {
+      "epoch": 0.5594913714804723,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.7939903986478354e-05,
+      "loss": 0.4814,
+      "step": 308
+    },
+    {
+      "epoch": 0.5613079019073569,
+      "grad_norm": 1.125,
+      "learning_rate": 1.7920766142499673e-05,
+      "loss": 0.4778,
+      "step": 309
+    },
+    {
+      "epoch": 0.5631244323342416,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.7901550123756906e-05,
+      "loss": 0.4781,
+      "step": 310
+    },
+    {
+      "epoch": 0.5649409627611263,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.78822561199044e-05,
+      "loss": 0.4747,
+      "step": 311
+    },
+    {
+      "epoch": 0.5667574931880109,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.786288432136619e-05,
+      "loss": 0.4736,
+      "step": 312
+    },
+    {
+      "epoch": 0.5685740236148955,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7843434919334103e-05,
+      "loss": 0.4803,
+      "step": 313
+    },
+    {
+      "epoch": 0.5703905540417802,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.7823908105765883e-05,
+      "loss": 0.4774,
+      "step": 314
+    },
+    {
+      "epoch": 0.5722070844686649,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.7804304073383298e-05,
+      "loss": 0.4864,
+      "step": 315
+    },
+    {
+      "epoch": 0.5740236148955495,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.7784623015670237e-05,
+      "loss": 0.4705,
+      "step": 316
+    },
+    {
+      "epoch": 0.5758401453224341,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.7764865126870788e-05,
+      "loss": 0.4672,
+      "step": 317
+    },
+    {
+      "epoch": 0.5776566757493188,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7745030601987338e-05,
+      "loss": 0.4758,
+      "step": 318
+    },
+    {
+      "epoch": 0.5794732061762035,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.7725119636778644e-05,
+      "loss": 0.473,
+      "step": 319
+    },
+    {
+      "epoch": 0.5812897366030881,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.7705132427757895e-05,
+      "loss": 0.4677,
+      "step": 320
+    },
+    {
+      "epoch": 0.5831062670299727,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.7685069172190766e-05,
+      "loss": 0.4733,
+      "step": 321
+    },
+    {
+      "epoch": 0.5849227974568574,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.76649300680935e-05,
+      "loss": 0.4816,
+      "step": 322
+    },
+    {
+      "epoch": 0.5867393278837421,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.764471531423092e-05,
+      "loss": 0.4918,
+      "step": 323
+    },
+    {
+      "epoch": 0.5885558583106267,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.762442511011448e-05,
+      "loss": 0.4732,
+      "step": 324
+    },
+    {
+      "epoch": 0.5903723887375113,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.7604059656000313e-05,
+      "loss": 0.4685,
+      "step": 325
+    },
+    {
+      "epoch": 0.592188919164396,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.7583619152887222e-05,
+      "loss": 0.4748,
+      "step": 326
+    },
+    {
+      "epoch": 0.5940054495912807,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.756310380251472e-05,
+      "loss": 0.4669,
+      "step": 327
+    },
+    {
+      "epoch": 0.5958219800181653,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.754251380736104e-05,
+      "loss": 0.466,
+      "step": 328
+    },
+    {
+      "epoch": 0.59763851044505,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7521849370641116e-05,
+      "loss": 0.4591,
+      "step": 329
+    },
+    {
+      "epoch": 0.5994550408719346,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.7501110696304598e-05,
+      "loss": 0.4607,
+      "step": 330
+    },
+    {
+      "epoch": 0.6012715712988193,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.7480297989033824e-05,
+      "loss": 0.4698,
+      "step": 331
+    },
+    {
+      "epoch": 0.6030881017257039,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.7459411454241822e-05,
+      "loss": 0.4713,
+      "step": 332
+    },
+    {
+      "epoch": 0.6049046321525886,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.7438451298070252e-05,
+      "loss": 0.4632,
+      "step": 333
+    },
+    {
+      "epoch": 0.6067211625794732,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.7417417727387392e-05,
+      "loss": 0.4675,
+      "step": 334
+    },
+    {
+      "epoch": 0.6085376930063578,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.73963109497861e-05,
+      "loss": 0.4812,
+      "step": 335
+    },
+    {
+      "epoch": 0.6103542234332425,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.737513117358174e-05,
+      "loss": 0.4746,
+      "step": 336
+    },
+    {
+      "epoch": 0.6121707538601272,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.735387860781016e-05,
+      "loss": 0.468,
+      "step": 337
+    },
+    {
+      "epoch": 0.6139872842870118,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.7332553462225604e-05,
+      "loss": 0.4721,
+      "step": 338
+    },
+    {
+      "epoch": 0.6158038147138964,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.7311155947298644e-05,
+      "loss": 0.4696,
+      "step": 339
+    },
+    {
+      "epoch": 0.6176203451407811,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.7289686274214116e-05,
+      "loss": 0.478,
+      "step": 340
+    },
+    {
+      "epoch": 0.6194368755676658,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.726814465486903e-05,
+      "loss": 0.47,
+      "step": 341
+    },
+    {
+      "epoch": 0.6212534059945504,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.7246531301870467e-05,
+      "loss": 0.4697,
+      "step": 342
+    },
+    {
+      "epoch": 0.623069936421435,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.7224846428533498e-05,
+      "loss": 0.4752,
+      "step": 343
+    },
+    {
+      "epoch": 0.6248864668483197,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.720309024887907e-05,
+      "loss": 0.4713,
+      "step": 344
+    },
+    {
+      "epoch": 0.6267029972752044,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.718126297763189e-05,
+      "loss": 0.4718,
+      "step": 345
+    },
+    {
+      "epoch": 0.628519527702089,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.7159364830218312e-05,
+      "loss": 0.4798,
+      "step": 346
+    },
+    {
+      "epoch": 0.6303360581289736,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7137396022764216e-05,
+      "loss": 0.4691,
+      "step": 347
+    },
+    {
+      "epoch": 0.6321525885558583,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.7115356772092858e-05,
+      "loss": 0.4685,
+      "step": 348
+    },
+    {
+      "epoch": 0.633969118982743,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.709324729572274e-05,
+      "loss": 0.4716,
+      "step": 349
+    },
+    {
+      "epoch": 0.6357856494096276,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.4775,
+      "step": 350
+    },
+    {
+      "epoch": 0.6376021798365122,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.7048818539423616e-05,
+      "loss": 0.473,
+      "step": 351
+    },
+    {
+      "epoch": 0.6394187102633969,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.7026499697988496e-05,
+      "loss": 0.4677,
+      "step": 352
+    },
+    {
+      "epoch": 0.6412352406902816,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.7004111507838067e-05,
+      "loss": 0.4711,
+      "step": 353
+    },
+    {
+      "epoch": 0.6430517711171662,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.698165418993473e-05,
+      "loss": 0.4677,
+      "step": 354
+    },
+    {
+      "epoch": 0.6448683015440508,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.6959127965923144e-05,
+      "loss": 0.4761,
+      "step": 355
+    },
+    {
+      "epoch": 0.6466848319709355,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.693653305812805e-05,
+      "loss": 0.4593,
+      "step": 356
+    },
+    {
+      "epoch": 0.6485013623978202,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.6913869689552066e-05,
+      "loss": 0.4805,
+      "step": 357
+    },
+    {
+      "epoch": 0.6503178928247049,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.6891138083873486e-05,
+      "loss": 0.4702,
+      "step": 358
+    },
+    {
+      "epoch": 0.6521344232515894,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.6868338465444086e-05,
+      "loss": 0.4714,
+      "step": 359
+    },
+    {
+      "epoch": 0.6539509536784741,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.684547105928689e-05,
+      "loss": 0.4627,
+      "step": 360
+    },
+    {
+      "epoch": 0.6557674841053588,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.6822536091093967e-05,
+      "loss": 0.4633,
+      "step": 361
+    },
+    {
+      "epoch": 0.6575840145322435,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.6799533787224192e-05,
+      "loss": 0.4751,
+      "step": 362
+    },
+    {
+      "epoch": 0.659400544959128,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.6776464374701026e-05,
+      "loss": 0.4731,
+      "step": 363
+    },
+    {
+      "epoch": 0.6612170753860127,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.6753328081210244e-05,
+      "loss": 0.4631,
+      "step": 364
+    },
+    {
+      "epoch": 0.6630336058128974,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.6730125135097736e-05,
+      "loss": 0.4576,
+      "step": 365
+    },
+    {
+      "epoch": 0.6648501362397821,
+      "grad_norm": 1.0,
+      "learning_rate": 1.6706855765367202e-05,
+      "loss": 0.4645,
+      "step": 366
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.6683520201677933e-05,
+      "loss": 0.4628,
+      "step": 367
+    },
+    {
+      "epoch": 0.6684831970935513,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.666011867434252e-05,
+      "loss": 0.4769,
+      "step": 368
+    },
+    {
+      "epoch": 0.6684831970935513,
+      "eval_loss": 0.4684128165245056,
+      "eval_runtime": 34.9278,
+      "eval_samples_per_second": 5.726,
+      "eval_steps_per_second": 1.432,
+      "step": 368
+    },
+    {
+      "epoch": 0.670299727520436,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.6636651414324586e-05,
+      "loss": 0.466,
+      "step": 369
+    },
+    {
+      "epoch": 0.6721162579473207,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.661311865323652e-05,
+      "loss": 0.4649,
+      "step": 370
+    },
+    {
+      "epoch": 0.6739327883742052,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.6589520623337173e-05,
+      "loss": 0.4734,
+      "step": 371
+    },
+    {
+      "epoch": 0.6757493188010899,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.6565857557529567e-05,
+      "loss": 0.4594,
+      "step": 372
+    },
+    {
+      "epoch": 0.6775658492279746,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.6542129689358613e-05,
+      "loss": 0.4647,
+      "step": 373
+    },
+    {
+      "epoch": 0.6793823796548593,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.651833725300879e-05,
+      "loss": 0.4553,
+      "step": 374
+    },
+    {
+      "epoch": 0.6811989100817438,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.6494480483301836e-05,
+      "loss": 0.4703,
+      "step": 375
+    },
+    {
+      "epoch": 0.6830154405086285,
+      "grad_norm": 1.0,
+      "learning_rate": 1.6470559615694445e-05,
+      "loss": 0.4627,
+      "step": 376
+    },
+    {
+      "epoch": 0.6848319709355132,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.6446574886275914e-05,
+      "loss": 0.4641,
+      "step": 377
+    },
+    {
+      "epoch": 0.6866485013623979,
+      "grad_norm": 1.125,
+      "learning_rate": 1.6422526531765846e-05,
+      "loss": 0.4709,
+      "step": 378
+    },
+    {
+      "epoch": 0.6884650317892824,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.6398414789511784e-05,
+      "loss": 0.4615,
+      "step": 379
+    },
+    {
+      "epoch": 0.6902815622161671,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 0.4565,
+      "step": 380
+    },
+    {
+      "epoch": 0.6920980926430518,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.6350002094287608e-05,
+      "loss": 0.4635,
+      "step": 381
+    },
+    {
+      "epoch": 0.6939146230699365,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.6325701619131246e-05,
+      "loss": 0.4538,
+      "step": 382
+    },
+    {
+      "epoch": 0.695731153496821,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.6301338711853695e-05,
+      "loss": 0.4598,
+      "step": 383
+    },
+    {
+      "epoch": 0.6975476839237057,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.6276913612907005e-05,
+      "loss": 0.4586,
+      "step": 384
+    },
+    {
+      "epoch": 0.6993642143505904,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.6252426563357054e-05,
+      "loss": 0.4695,
+      "step": 385
+    },
+    {
+      "epoch": 0.701180744777475,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.6227877804881126e-05,
+      "loss": 0.4743,
+      "step": 386
+    },
+    {
+      "epoch": 0.7029972752043597,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.6203267579765563e-05,
+      "loss": 0.4701,
+      "step": 387
+    },
+    {
+      "epoch": 0.7048138056312443,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.6178596130903345e-05,
+      "loss": 0.4578,
+      "step": 388
+    },
+    {
+      "epoch": 0.706630336058129,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.6153863701791717e-05,
+      "loss": 0.4635,
+      "step": 389
+    },
+    {
+      "epoch": 0.7084468664850136,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.6129070536529767e-05,
+      "loss": 0.4641,
+      "step": 390
+    },
+    {
+      "epoch": 0.7102633969118983,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.6104216879816027e-05,
+      "loss": 0.4715,
+      "step": 391
+    },
+    {
+      "epoch": 0.7120799273387829,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.6079302976946055e-05,
+      "loss": 0.4641,
+      "step": 392
+    },
+    {
+      "epoch": 0.7138964577656676,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.6054329073810016e-05,
+      "loss": 0.4569,
+      "step": 393
+    },
+    {
+      "epoch": 0.7157129881925522,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.602929541689025e-05,
+      "loss": 0.461,
+      "step": 394
+    },
+    {
+      "epoch": 0.7175295186194369,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.6004202253258844e-05,
+      "loss": 0.4575,
+      "step": 395
+    },
+    {
+      "epoch": 0.7193460490463215,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.597904983057519e-05,
+      "loss": 0.4693,
+      "step": 396
+    },
+    {
+      "epoch": 0.7211625794732062,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.595383839708355e-05,
+      "loss": 0.46,
+      "step": 397
+    },
+    {
+      "epoch": 0.7229791099000908,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.5928568201610593e-05,
+      "loss": 0.4654,
+      "step": 398
+    },
+    {
+      "epoch": 0.7247956403269755,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.5903239493562948e-05,
+      "loss": 0.4575,
+      "step": 399
+    },
+    {
+      "epoch": 0.7266121707538601,
+      "grad_norm": 1.0,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.4535,
+      "step": 400
+    },
+    {
+      "epoch": 0.7284287011807448,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.5852407540255103e-05,
+      "loss": 0.4591,
+      "step": 401
+    },
+    {
+      "epoch": 0.7302452316076294,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.5826904796685763e-05,
+      "loss": 0.4627,
+      "step": 402
+    },
+    {
+      "epoch": 0.7320617620345141,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.5801344543918495e-05,
+      "loss": 0.4614,
+      "step": 403
+    },
+    {
+      "epoch": 0.7338782924613987,
+      "grad_norm": 1.0,
+      "learning_rate": 1.5775727034222675e-05,
+      "loss": 0.4685,
+      "step": 404
+    },
+    {
+      "epoch": 0.7356948228882834,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.575005252043279e-05,
+      "loss": 0.4719,
+      "step": 405
+    },
+    {
+      "epoch": 0.737511353315168,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.572432125594591e-05,
+      "loss": 0.4637,
+      "step": 406
+    },
+    {
+      "epoch": 0.7393278837420527,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.5698533494719238e-05,
+      "loss": 0.459,
+      "step": 407
+    },
+    {
+      "epoch": 0.7411444141689373,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.567268949126757e-05,
+      "loss": 0.4762,
+      "step": 408
+    },
+    {
+      "epoch": 0.742960944595822,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.5646789500660772e-05,
+      "loss": 0.457,
+      "step": 409
+    },
+    {
+      "epoch": 0.7447774750227066,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.5620833778521306e-05,
+      "loss": 0.4666,
+      "step": 410
+    },
+    {
+      "epoch": 0.7465940054495913,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.5594822581021673e-05,
+      "loss": 0.4609,
+      "step": 411
+    },
+    {
+      "epoch": 0.7484105358764759,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.556875616488188e-05,
+      "loss": 0.4581,
+      "step": 412
+    },
+    {
+      "epoch": 0.7502270663033606,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.5542634787366942e-05,
+      "loss": 0.4659,
+      "step": 413
+    },
+    {
+      "epoch": 0.7520435967302452,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.5516458706284306e-05,
+      "loss": 0.4686,
+      "step": 414
+    },
+    {
+      "epoch": 0.7538601271571299,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.549022817998132e-05,
+      "loss": 0.454,
+      "step": 415
+    },
+    {
+      "epoch": 0.7556766575840145,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.5463943467342694e-05,
+      "loss": 0.4529,
+      "step": 416
+    },
+    {
+      "epoch": 0.7574931880108992,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.5437604827787925e-05,
+      "loss": 0.4668,
+      "step": 417
+    },
+    {
+      "epoch": 0.7593097184377838,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.541121252126876e-05,
+      "loss": 0.4632,
+      "step": 418
+    },
+    {
+      "epoch": 0.7611262488646685,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.5384766808266603e-05,
+      "loss": 0.4579,
+      "step": 419
+    },
+    {
+      "epoch": 0.7629427792915532,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.5358267949789968e-05,
+      "loss": 0.4507,
+      "step": 420
+    },
+    {
+      "epoch": 0.7647593097184378,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.5331716207371888e-05,
+      "loss": 0.4527,
+      "step": 421
+    },
+    {
+      "epoch": 0.7665758401453224,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.5305111843067343e-05,
+      "loss": 0.4684,
+      "step": 422
+    },
+    {
+      "epoch": 0.7683923705722071,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.5278455119450666e-05,
+      "loss": 0.4613,
+      "step": 423
+    },
+    {
+      "epoch": 0.7702089009990918,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.5251746299612959e-05,
+      "loss": 0.4552,
+      "step": 424
+    },
+    {
+      "epoch": 0.7720254314259763,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.5224985647159489e-05,
+      "loss": 0.4582,
+      "step": 425
+    },
+    {
+      "epoch": 0.773841961852861,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5198173426207095e-05,
+      "loss": 0.4601,
+      "step": 426
+    },
+    {
+      "epoch": 0.7756584922797457,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.5171309901381572e-05,
+      "loss": 0.4487,
+      "step": 427
+    },
+    {
+      "epoch": 0.7774750227066304,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.5144395337815066e-05,
+      "loss": 0.4581,
+      "step": 428
+    },
+    {
+      "epoch": 0.779291553133515,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5117430001143451e-05,
+      "loss": 0.4645,
+      "step": 429
+    },
+    {
+      "epoch": 0.7811080835603996,
+      "grad_norm": 1.0,
+      "learning_rate": 1.5090414157503715e-05,
+      "loss": 0.4586,
+      "step": 430
+    },
+    {
+      "epoch": 0.7829246139872843,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.5063348073531325e-05,
+      "loss": 0.4542,
+      "step": 431
+    },
+    {
+      "epoch": 0.784741144414169,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.503623201635761e-05,
+      "loss": 0.4485,
+      "step": 432
+    },
+    {
+      "epoch": 0.7865576748410535,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.5009066253607101e-05,
+      "loss": 0.4696,
+      "step": 433
+    },
+    {
+      "epoch": 0.7883742052679382,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.498185105339491e-05,
+      "loss": 0.4636,
+      "step": 434
+    },
+    {
+      "epoch": 0.7901907356948229,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.4954586684324077e-05,
+      "loss": 0.4597,
+      "step": 435
+    },
+    {
+      "epoch": 0.7920072661217076,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.4927273415482916e-05,
+      "loss": 0.453,
+      "step": 436
+    },
+    {
+      "epoch": 0.7938237965485921,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.4899911516442367e-05,
+      "loss": 0.457,
+      "step": 437
+    },
+    {
+      "epoch": 0.7956403269754768,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.4872501257253325e-05,
+      "loss": 0.4625,
+      "step": 438
+    },
+    {
+      "epoch": 0.7974568574023615,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.484504290844398e-05,
+      "loss": 0.4496,
+      "step": 439
+    },
+    {
+      "epoch": 0.7992733878292462,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.4817536741017153e-05,
+      "loss": 0.4508,
+      "step": 440
+    },
+    {
+      "epoch": 0.8010899182561307,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.4789983026447612e-05,
+      "loss": 0.4589,
+      "step": 441
+    },
+    {
+      "epoch": 0.8029064486830154,
+      "grad_norm": 1.0,
+      "learning_rate": 1.4762382036679393e-05,
+      "loss": 0.4499,
+      "step": 442
+    },
+    {
+      "epoch": 0.8047229791099001,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.4734734044123123e-05,
+      "loss": 0.4476,
+      "step": 443
+    },
+    {
+      "epoch": 0.8065395095367848,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.470703932165333e-05,
+      "loss": 0.452,
+      "step": 444
+    },
+    {
+      "epoch": 0.8083560399636693,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.4679298142605735e-05,
+      "loss": 0.4438,
+      "step": 445
+    },
+    {
+      "epoch": 0.810172570390554,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.4651510780774585e-05,
+      "loss": 0.4426,
+      "step": 446
+    },
+    {
+      "epoch": 0.8119891008174387,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.462367751040992e-05,
+      "loss": 0.451,
+      "step": 447
+    },
+    {
+      "epoch": 0.8138056312443234,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.4595798606214882e-05,
+      "loss": 0.4637,
+      "step": 448
+    },
+    {
+      "epoch": 0.815622161671208,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.4567874343342996e-05,
+      "loss": 0.4576,
+      "step": 449
+    },
+    {
+      "epoch": 0.8174386920980926,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 0.4755,
+      "step": 450
+    },
+    {
+      "epoch": 0.8192552225249773,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.4511890844418453e-05,
+      "loss": 0.4501,
+      "step": 451
+    },
+    {
+      "epoch": 0.821071752951862,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.4483832160900326e-05,
+      "loss": 0.4471,
+      "step": 452
+    },
+    {
+      "epoch": 0.8228882833787466,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.4455729223768966e-05,
+      "loss": 0.4491,
+      "step": 453
+    },
+    {
+      "epoch": 0.8247048138056312,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.442758231038902e-05,
+      "loss": 0.4429,
+      "step": 454
+    },
+    {
+      "epoch": 0.8265213442325159,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.4399391698559153e-05,
+      "loss": 0.4597,
+      "step": 455
+    },
+    {
+      "epoch": 0.8283378746594006,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.437115766650933e-05,
+      "loss": 0.4525,
+      "step": 456
+    },
+    {
+      "epoch": 0.8301544050862852,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.4342880492898048e-05,
+      "loss": 0.4568,
+      "step": 457
+    },
+    {
+      "epoch": 0.8319709355131698,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.4314560456809592e-05,
+      "loss": 0.4491,
+      "step": 458
+    },
+    {
+      "epoch": 0.8337874659400545,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.4286197837751286e-05,
+      "loss": 0.4501,
+      "step": 459
+    },
+    {
+      "epoch": 0.8356039963669392,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.4257792915650728e-05,
+      "loss": 0.4601,
+      "step": 460
+    },
+    {
+      "epoch": 0.8374205267938238,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.4229345970853032e-05,
+      "loss": 0.4474,
+      "step": 461
+    },
+    {
+      "epoch": 0.8392370572207084,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.4200857284118067e-05,
+      "loss": 0.4454,
+      "step": 462
+    },
+    {
+      "epoch": 0.8410535876475931,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.4172327136617656e-05,
+      "loss": 0.4454,
+      "step": 463
+    },
+    {
+      "epoch": 0.8428701180744778,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.4143755809932843e-05,
+      "loss": 0.4471,
+      "step": 464
+    },
+    {
+      "epoch": 0.8446866485013624,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.411514358605109e-05,
+      "loss": 0.4513,
+      "step": 465
+    },
+    {
+      "epoch": 0.846503178928247,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.4086490747363492e-05,
+      "loss": 0.4702,
+      "step": 466
+    },
+    {
+      "epoch": 0.8483197093551317,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.4057797576662e-05,
+      "loss": 0.4456,
+      "step": 467
+    },
+    {
+      "epoch": 0.8501362397820164,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.4029064357136628e-05,
+      "loss": 0.4347,
+      "step": 468
+    },
+    {
+      "epoch": 0.851952770208901,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.4000291372372647e-05,
+      "loss": 0.4546,
+      "step": 469
+    },
+    {
+      "epoch": 0.8537693006357856,
+      "grad_norm": 1.0,
+      "learning_rate": 1.3971478906347806e-05,
+      "loss": 0.4448,
+      "step": 470
+    },
+    {
+      "epoch": 0.8555858310626703,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.3942627243429512e-05,
+      "loss": 0.4443,
+      "step": 471
+    },
+    {
+      "epoch": 0.857402361489555,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.3913736668372027e-05,
+      "loss": 0.4598,
+      "step": 472
+    },
+    {
+      "epoch": 0.8592188919164396,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.3884807466313664e-05,
+      "loss": 0.444,
+      "step": 473
+    },
+    {
+      "epoch": 0.8610354223433242,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.3855839922773968e-05,
+      "loss": 0.4509,
+      "step": 474
+    },
+    {
+      "epoch": 0.8628519527702089,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 0.4655,
+      "step": 475
+    },
+    {
+      "epoch": 0.8646684831970936,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.3797790955218014e-05,
+      "loss": 0.4516,
+      "step": 476
+    },
+    {
+      "epoch": 0.8664850136239782,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.3768710104121628e-05,
+      "loss": 0.458,
+      "step": 477
+    },
+    {
+      "epoch": 0.8683015440508629,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.3739592057378005e-05,
+      "loss": 0.4561,
+      "step": 478
+    },
+    {
+      "epoch": 0.8701180744777475,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.3710437102370511e-05,
+      "loss": 0.4539,
+      "step": 479
+    },
+    {
+      "epoch": 0.8719346049046321,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.3681245526846782e-05,
+      "loss": 0.4428,
+      "step": 480
+    },
+    {
+      "epoch": 0.8737511353315168,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.365201761891588e-05,
+      "loss": 0.4426,
+      "step": 481
+    },
+    {
+      "epoch": 0.8755676657584015,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.3622753667045459e-05,
+      "loss": 0.4518,
+      "step": 482
+    },
+    {
+      "epoch": 0.8773841961852861,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.3593453960058909e-05,
+      "loss": 0.4453,
+      "step": 483
+    },
+    {
+      "epoch": 0.8792007266121707,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.3564118787132507e-05,
+      "loss": 0.4581,
+      "step": 484
+    },
+    {
+      "epoch": 0.8810172570390554,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.3534748437792573e-05,
+      "loss": 0.4543,
+      "step": 485
+    },
+    {
+      "epoch": 0.8828337874659401,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.350534320191259e-05,
+      "loss": 0.4446,
+      "step": 486
+    },
+    {
+      "epoch": 0.8846503178928247,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.347590336971037e-05,
+      "loss": 0.4394,
+      "step": 487
+    },
+    {
+      "epoch": 0.8864668483197093,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.344642923174517e-05,
+      "loss": 0.4523,
+      "step": 488
+    },
+    {
+      "epoch": 0.888283378746594,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.3416921078914835e-05,
+      "loss": 0.4602,
+      "step": 489
+    },
+    {
+      "epoch": 0.8900999091734787,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.3387379202452917e-05,
+      "loss": 0.4418,
+      "step": 490
+    },
+    {
+      "epoch": 0.8919164396003633,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.3357803893925807e-05,
+      "loss": 0.4399,
+      "step": 491
+    },
+    {
+      "epoch": 0.8937329700272479,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.3328195445229869e-05,
+      "loss": 0.4492,
+      "step": 492
+    },
+    {
+      "epoch": 0.8955495004541326,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.329855414858853e-05,
+      "loss": 0.4605,
+      "step": 493
+    },
+    {
+      "epoch": 0.8973660308810173,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.3268880296549424e-05,
+      "loss": 0.4497,
+      "step": 494
+    },
+    {
+      "epoch": 0.8991825613079019,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.3239174181981496e-05,
+      "loss": 0.4604,
+      "step": 495
+    },
+    {
+      "epoch": 0.9009990917347865,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.3209436098072095e-05,
+      "loss": 0.4542,
+      "step": 496
+    },
+    {
+      "epoch": 0.9028156221616712,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.317966633832411e-05,
+      "loss": 0.4468,
+      "step": 497
+    },
+    {
+      "epoch": 0.9046321525885559,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.3149865196553049e-05,
+      "loss": 0.4623,
+      "step": 498
+    },
+    {
+      "epoch": 0.9064486830154405,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.3120032966884151e-05,
+      "loss": 0.4479,
+      "step": 499
+    },
+    {
+      "epoch": 0.9082652134423251,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.456,
+      "step": 500
+    },
+    {
+      "epoch": 0.9100817438692098,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.306027642188501e-05,
+      "loss": 0.4532,
+      "step": 501
+    },
+    {
+      "epoch": 0.9118982742960945,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.3030352696327741e-05,
+      "loss": 0.4517,
+      "step": 502
+    },
+    {
+      "epoch": 0.9137148047229791,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.3000399062412763e-05,
+      "loss": 0.4304,
+      "step": 503
+    },
+    {
+      "epoch": 0.9155313351498637,
+      "grad_norm": 1.0,
+      "learning_rate": 1.297041581577035e-05,
+      "loss": 0.4352,
+      "step": 504
+    },
+    {
+      "epoch": 0.9173478655767484,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.294040325232304e-05,
+      "loss": 0.4319,
+      "step": 505
+    },
+    {
+      "epoch": 0.9191643960036331,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.2910361668282718e-05,
+      "loss": 0.451,
+      "step": 506
+    },
+    {
+      "epoch": 0.9209809264305178,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.2880291360147694e-05,
+      "loss": 0.4431,
+      "step": 507
+    },
+    {
+      "epoch": 0.9227974568574023,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.2850192624699762e-05,
+      "loss": 0.4596,
+      "step": 508
+    },
+    {
+      "epoch": 0.924613987284287,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.2820065759001295e-05,
+      "loss": 0.4516,
+      "step": 509
+    },
+    {
+      "epoch": 0.9264305177111717,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.2789911060392295e-05,
+      "loss": 0.4492,
+      "step": 510
+    },
+    {
+      "epoch": 0.9282470481380564,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.2759728826487461e-05,
+      "loss": 0.4476,
+      "step": 511
+    },
+    {
+      "epoch": 0.9300635785649409,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.2729519355173254e-05,
+      "loss": 0.4468,
+      "step": 512
+    },
+    {
+      "epoch": 0.9318801089918256,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.2699282944604968e-05,
+      "loss": 0.4413,
+      "step": 513
+    },
+    {
+      "epoch": 0.9336966394187103,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.2669019893203758e-05,
+      "loss": 0.4409,
+      "step": 514
+    },
+    {
+      "epoch": 0.935513169845595,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.2638730499653731e-05,
+      "loss": 0.4381,
+      "step": 515
+    },
+    {
+      "epoch": 0.9373297002724795,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.2608415062898971e-05,
+      "loss": 0.4573,
+      "step": 516
+    },
+    {
+      "epoch": 0.9391462306993642,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.25780738821406e-05,
+      "loss": 0.4509,
+      "step": 517
+    },
+    {
+      "epoch": 0.9409627611262489,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.2547707256833823e-05,
+      "loss": 0.4566,
+      "step": 518
+    },
+    {
+      "epoch": 0.9427792915531336,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.2517315486684973e-05,
+      "loss": 0.4379,
+      "step": 519
+    },
+    {
+      "epoch": 0.9445958219800181,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.2486898871648552e-05,
+      "loss": 0.439,
+      "step": 520
+    },
+    {
+      "epoch": 0.9464123524069028,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.2456457711924266e-05,
+      "loss": 0.4341,
+      "step": 521
+    },
+    {
+      "epoch": 0.9482288828337875,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.2425992307954075e-05,
+      "loss": 0.4375,
+      "step": 522
+    },
+    {
+      "epoch": 0.9500454132606722,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.2395502960419221e-05,
+      "loss": 0.4627,
+      "step": 523
+    },
+    {
+      "epoch": 0.9518619436875567,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.236498997023725e-05,
+      "loss": 0.445,
+      "step": 524
+    },
+    {
+      "epoch": 0.9536784741144414,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2334453638559057e-05,
+      "loss": 0.4422,
+      "step": 525
+    },
+    {
+      "epoch": 0.9554950045413261,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2303894266765908e-05,
+      "loss": 0.4506,
+      "step": 526
+    },
+    {
+      "epoch": 0.9573115349682108,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.2273312156466466e-05,
+      "loss": 0.4475,
+      "step": 527
+    },
+    {
+      "epoch": 0.9591280653950953,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.2242707609493814e-05,
+      "loss": 0.436,
+      "step": 528
+    },
+    {
+      "epoch": 0.96094459582198,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.2212080927902474e-05,
+      "loss": 0.4527,
+      "step": 529
+    },
+    {
+      "epoch": 0.9627611262488647,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2181432413965428e-05,
+      "loss": 0.445,
+      "step": 530
+    },
+    {
+      "epoch": 0.9645776566757494,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.2150762370171137e-05,
+      "loss": 0.4447,
+      "step": 531
+    },
+    {
+      "epoch": 0.9663941871026339,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.212007109922055e-05,
+      "loss": 0.4446,
+      "step": 532
+    },
+    {
+      "epoch": 0.9682107175295186,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.2089358904024117e-05,
+      "loss": 0.4394,
+      "step": 533
+    },
+    {
+      "epoch": 0.9700272479564033,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.2058626087698814e-05,
+      "loss": 0.4407,
+      "step": 534
+    },
+    {
+      "epoch": 0.971843778383288,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.2027872953565125e-05,
+      "loss": 0.4396,
+      "step": 535
+    },
+    {
+      "epoch": 0.9736603088101726,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.1997099805144071e-05,
+      "loss": 0.4557,
+      "step": 536
+    },
+    {
+      "epoch": 0.9754768392370572,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.19663069461542e-05,
+      "loss": 0.4415,
+      "step": 537
+    },
+    {
+      "epoch": 0.9772933696639419,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.1935494680508606e-05,
+      "loss": 0.4474,
+      "step": 538
+    },
+    {
+      "epoch": 0.9791099000908265,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1904663312311902e-05,
+      "loss": 0.4546,
+      "step": 539
+    },
+    {
+      "epoch": 0.9809264305177112,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.187381314585725e-05,
+      "loss": 0.4402,
+      "step": 540
+    },
+    {
+      "epoch": 0.9827429609445958,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1842944485623335e-05,
+      "loss": 0.4257,
+      "step": 541
+    },
+    {
+      "epoch": 0.9845594913714805,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1812057636271374e-05,
+      "loss": 0.4409,
+      "step": 542
+    },
+    {
+      "epoch": 0.9863760217983651,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.17811529026421e-05,
+      "loss": 0.4522,
+      "step": 543
+    },
+    {
+      "epoch": 0.9881925522252498,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.1750230589752763e-05,
+      "loss": 0.4594,
+      "step": 544
+    },
+    {
+      "epoch": 0.9900090826521344,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.1719291002794096e-05,
+      "loss": 0.437,
+      "step": 545
+    },
+    {
+      "epoch": 0.9918256130790191,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.1688334447127338e-05,
+      "loss": 0.4266,
+      "step": 546
+    },
+    {
+      "epoch": 0.9936421435059037,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1657361228281198e-05,
+      "loss": 0.4377,
+      "step": 547
+    },
+    {
+      "epoch": 0.9954586739327884,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.1626371651948839e-05,
+      "loss": 0.4373,
+      "step": 548
+    },
+    {
+      "epoch": 0.997275204359673,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.1595366023984864e-05,
+      "loss": 0.4395,
+      "step": 549
+    },
+    {
+      "epoch": 0.9990917347865577,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.4454,
+      "step": 550
+    },
+    {
+      "epoch": 1.0009082652134422,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.1533307837369607e-05,
+      "loss": 0.4246,
+      "step": 551
+    },
+    {
+      "epoch": 1.002724795640327,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.1502255891207572e-05,
+      "loss": 0.4394,
+      "step": 552
+    },
+    {
+      "epoch": 1.002724795640327,
+      "eval_loss": 0.4440397024154663,
+      "eval_runtime": 34.9277,
+      "eval_samples_per_second": 5.726,
+      "eval_steps_per_second": 1.432,
+      "step": 552
+    },
+    {
+      "epoch": 1.0045413260672116,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.1471189118386374e-05,
+      "loss": 0.4373,
+      "step": 553
+    },
+    {
+      "epoch": 1.0063578564940963,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.1440107825522522e-05,
+      "loss": 0.4495,
+      "step": 554
+    },
+    {
+      "epoch": 1.008174386920981,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.1409012319375828e-05,
+      "loss": 0.4417,
+      "step": 555
+    },
+    {
+      "epoch": 1.0099909173478656,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.137790290684638e-05,
+      "loss": 0.4462,
+      "step": 556
+    },
+    {
+      "epoch": 1.0118074477747503,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.1346779894971526e-05,
+      "loss": 0.4426,
+      "step": 557
+    },
+    {
+      "epoch": 1.013623978201635,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.1315643590922827e-05,
+      "loss": 0.4388,
+      "step": 558
+    },
+    {
+      "epoch": 1.0009082652134422,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.128449430200303e-05,
+      "loss": 0.4019,
+      "step": 559
+    },
+    {
+      "epoch": 1.002724795640327,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1253332335643043e-05,
+      "loss": 0.3465,
+      "step": 560
+    },
+    {
+      "epoch": 1.0045413260672116,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.1222157999398895e-05,
+      "loss": 0.3613,
+      "step": 561
+    },
+    {
+      "epoch": 1.0063578564940963,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.11909716009487e-05,
+      "loss": 0.354,
+      "step": 562
+    },
+    {
+      "epoch": 1.008174386920981,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1159773448089615e-05,
+      "loss": 0.3524,
+      "step": 563
+    },
+    {
+      "epoch": 1.0099909173478656,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1128563848734817e-05,
+      "loss": 0.3622,
+      "step": 564
+    },
+    {
+      "epoch": 1.0118074477747503,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.1097343110910452e-05,
+      "loss": 0.3519,
+      "step": 565
+    },
+    {
+      "epoch": 1.013623978201635,
+      "grad_norm": 1.125,
+      "learning_rate": 1.10661115427526e-05,
+      "loss": 0.3445,
+      "step": 566
+    },
+    {
+      "epoch": 1.0154405086285194,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1034869452504227e-05,
+      "loss": 0.3572,
+      "step": 567
+    },
+    {
+      "epoch": 1.017257039055404,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.1003617148512149e-05,
+      "loss": 0.361,
+      "step": 568
+    },
+    {
+      "epoch": 1.0190735694822888,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0972354939223997e-05,
+      "loss": 0.3498,
+      "step": 569
+    },
+    {
+      "epoch": 1.0208900999091735,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.0941083133185146e-05,
+      "loss": 0.3508,
+      "step": 570
+    },
+    {
+      "epoch": 1.0227066303360581,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.0909802039035702e-05,
+      "loss": 0.3432,
+      "step": 571
+    },
+    {
+      "epoch": 1.0245231607629428,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.0878511965507435e-05,
+      "loss": 0.3498,
+      "step": 572
+    },
+    {
+      "epoch": 1.0263396911898275,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.0847213221420735e-05,
+      "loss": 0.3472,
+      "step": 573
+    },
+    {
+      "epoch": 1.0281562216167122,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.0815906115681579e-05,
+      "loss": 0.3442,
+      "step": 574
+    },
+    {
+      "epoch": 1.0299727520435966,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.0784590957278452e-05,
+      "loss": 0.3472,
+      "step": 575
+    },
+    {
+      "epoch": 1.0317892824704813,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.0753268055279328e-05,
+      "loss": 0.3509,
+      "step": 576
+    },
+    {
+      "epoch": 1.033605812897366,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.072193771882861e-05,
+      "loss": 0.3451,
+      "step": 577
+    },
+    {
+      "epoch": 1.0354223433242506,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.0690600257144062e-05,
+      "loss": 0.3414,
+      "step": 578
+    },
+    {
+      "epoch": 1.0372388737511353,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.065925597951378e-05,
+      "loss": 0.3427,
+      "step": 579
+    },
+    {
+      "epoch": 1.03905540417802,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.0627905195293135e-05,
+      "loss": 0.3464,
+      "step": 580
+    },
+    {
+      "epoch": 1.0408719346049047,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.059654821390171e-05,
+      "loss": 0.3456,
+      "step": 581
+    },
+    {
+      "epoch": 1.0426884650317894,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.0565185344820248e-05,
+      "loss": 0.3387,
+      "step": 582
+    },
+    {
+      "epoch": 1.044504995458674,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.0533816897587605e-05,
+      "loss": 0.3437,
+      "step": 583
+    },
+    {
+      "epoch": 1.0463215258855585,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.0502443181797696e-05,
+      "loss": 0.3472,
+      "step": 584
+    },
+    {
+      "epoch": 1.0481380563124432,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.0471064507096427e-05,
+      "loss": 0.3563,
+      "step": 585
+    },
+    {
+      "epoch": 1.0499545867393278,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.043968118317865e-05,
+      "loss": 0.3459,
+      "step": 586
+    },
+    {
+      "epoch": 1.0517711171662125,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.0408293519785103e-05,
+      "loss": 0.3446,
+      "step": 587
+    },
+    {
+      "epoch": 1.0535876475930972,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.0376901826699349e-05,
+      "loss": 0.3554,
+      "step": 588
+    },
+    {
+      "epoch": 1.0554041780199819,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.0345506413744726e-05,
+      "loss": 0.3508,
+      "step": 589
+    },
+    {
+      "epoch": 1.0572207084468666,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.0314107590781284e-05,
+      "loss": 0.3443,
+      "step": 590
+    },
+    {
+      "epoch": 1.0590372388737512,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.0282705667702734e-05,
+      "loss": 0.353,
+      "step": 591
+    },
+    {
+      "epoch": 1.0608537693006357,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.0251300954433377e-05,
+      "loss": 0.3386,
+      "step": 592
+    },
+    {
+      "epoch": 1.0626702997275204,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.0219893760925053e-05,
+      "loss": 0.3438,
+      "step": 593
+    },
+    {
+      "epoch": 1.064486830154405,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.0188484397154083e-05,
+      "loss": 0.3446,
+      "step": 594
+    },
+    {
+      "epoch": 1.0663033605812897,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.0157073173118207e-05,
+      "loss": 0.3429,
+      "step": 595
+    },
+    {
+      "epoch": 1.0681198910081744,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0125660398833528e-05,
+      "loss": 0.3435,
+      "step": 596
+    },
+    {
+      "epoch": 1.069936421435059,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.0094246384331444e-05,
+      "loss": 0.3465,
+      "step": 597
+    },
+    {
+      "epoch": 1.0717529518619437,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.0062831439655591e-05,
+      "loss": 0.3429,
+      "step": 598
+    },
+    {
+      "epoch": 1.0735694822888284,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0031415874858796e-05,
+      "loss": 0.3505,
+      "step": 599
+    },
+    {
+      "epoch": 1.0753860127157129,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1e-05,
+      "loss": 0.338,
+      "step": 600
+    },
+    {
+      "epoch": 1.0772025431425976,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.968584125141206e-06,
+      "loss": 0.3485,
+      "step": 601
+    },
+    {
+      "epoch": 1.0790190735694822,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.937168560344412e-06,
+      "loss": 0.3498,
+      "step": 602
+    },
+    {
+      "epoch": 1.080835603996367,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.905753615668561e-06,
+      "loss": 0.3521,
+      "step": 603
+    },
+    {
+      "epoch": 1.0826521344232516,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.874339601166474e-06,
+      "loss": 0.361,
+      "step": 604
+    },
+    {
+      "epoch": 1.0844686648501363,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.842926826881796e-06,
+      "loss": 0.3517,
+      "step": 605
+    },
+    {
+      "epoch": 1.086285195277021,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.81151560284592e-06,
+      "loss": 0.3379,
+      "step": 606
+    },
+    {
+      "epoch": 1.0881017257039056,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.78010623907495e-06,
+      "loss": 0.346,
+      "step": 607
+    },
+    {
+      "epoch": 1.0899182561307903,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.748699045566626e-06,
+      "loss": 0.3352,
+      "step": 608
+    },
+    {
+      "epoch": 1.0917347865576748,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.717294332297269e-06,
+      "loss": 0.3544,
+      "step": 609
+    },
+    {
+      "epoch": 1.0935513169845594,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.685892409218718e-06,
+      "loss": 0.3468,
+      "step": 610
+    },
+    {
+      "epoch": 1.095367847411444,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.654493586255279e-06,
+      "loss": 0.3537,
+      "step": 611
+    },
+    {
+      "epoch": 1.0971843778383288,
+      "grad_norm": 0.875,
+      "learning_rate": 9.623098173300655e-06,
+      "loss": 0.3467,
+      "step": 612
+    },
+    {
+      "epoch": 1.0990009082652135,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.5917064802149e-06,
+      "loss": 0.3459,
+      "step": 613
+    },
+    {
+      "epoch": 1.1008174386920981,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.560318816821354e-06,
+      "loss": 0.3423,
+      "step": 614
+    },
+    {
+      "epoch": 1.1026339691189828,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.528935492903575e-06,
+      "loss": 0.3393,
+      "step": 615
+    },
+    {
+      "epoch": 1.1044504995458675,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.497556818202306e-06,
+      "loss": 0.3552,
+      "step": 616
+    },
+    {
+      "epoch": 1.106267029972752,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.466183102412397e-06,
+      "loss": 0.3435,
+      "step": 617
+    },
+    {
+      "epoch": 1.1080835603996366,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.434814655179756e-06,
+      "loss": 0.3462,
+      "step": 618
+    },
+    {
+      "epoch": 1.1099000908265213,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.403451786098295e-06,
+      "loss": 0.3423,
+      "step": 619
+    },
+    {
+      "epoch": 1.111716621253406,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.372094804706867e-06,
+      "loss": 0.3488,
+      "step": 620
+    },
+    {
+      "epoch": 1.1135331516802907,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.340744020486223e-06,
+      "loss": 0.333,
+      "step": 621
+    },
+    {
+      "epoch": 1.1153496821071753,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.309399742855943e-06,
+      "loss": 0.3453,
+      "step": 622
+    },
+    {
+      "epoch": 1.11716621253406,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.278062281171394e-06,
+      "loss": 0.3413,
+      "step": 623
+    },
+    {
+      "epoch": 1.1189827429609447,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.246731944720675e-06,
+      "loss": 0.3358,
+      "step": 624
+    },
+    {
+      "epoch": 1.1207992733878291,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.215409042721553e-06,
+      "loss": 0.3415,
+      "step": 625
+    },
+    {
+      "epoch": 1.1226158038147138,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.184093884318426e-06,
+      "loss": 0.3415,
+      "step": 626
+    },
+    {
+      "epoch": 1.1244323342415985,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.152786778579266e-06,
+      "loss": 0.3512,
+      "step": 627
+    },
+    {
+      "epoch": 1.1262488646684832,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.121488034492569e-06,
+      "loss": 0.3364,
+      "step": 628
+    },
+    {
+      "epoch": 1.1280653950953679,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.090197960964301e-06,
+      "loss": 0.3414,
+      "step": 629
+    },
+    {
+      "epoch": 1.1298819255222525,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.058916866814857e-06,
+      "loss": 0.3365,
+      "step": 630
+    },
+    {
+      "epoch": 1.1316984559491372,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.027645060776008e-06,
+      "loss": 0.3416,
+      "step": 631
+    },
+    {
+      "epoch": 1.1335149863760219,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.996382851487851e-06,
+      "loss": 0.3467,
+      "step": 632
+    },
+    {
+      "epoch": 1.1353315168029066,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.965130547495777e-06,
+      "loss": 0.3402,
+      "step": 633
+    },
+    {
+      "epoch": 1.137148047229791,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.933888457247402e-06,
+      "loss": 0.3476,
+      "step": 634
+    },
+    {
+      "epoch": 1.1389645776566757,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.902656889089548e-06,
+      "loss": 0.3437,
+      "step": 635
+    },
+    {
+      "epoch": 1.1407811080835604,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.871436151265183e-06,
+      "loss": 0.3459,
+      "step": 636
+    },
+    {
+      "epoch": 1.142597638510445,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.840226551910387e-06,
+      "loss": 0.3525,
+      "step": 637
+    },
+    {
+      "epoch": 1.1444141689373297,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.809028399051302e-06,
+      "loss": 0.3364,
+      "step": 638
+    },
+    {
+      "epoch": 1.1462306993642144,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.777842000601106e-06,
+      "loss": 0.3422,
+      "step": 639
+    },
+    {
+      "epoch": 1.148047229791099,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.746667664356957e-06,
+      "loss": 0.3532,
+      "step": 640
+    },
+    {
+      "epoch": 1.1498637602179835,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.715505697996972e-06,
+      "loss": 0.3468,
+      "step": 641
+    },
+    {
+      "epoch": 1.1516802906448682,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.684356409077177e-06,
+      "loss": 0.3462,
+      "step": 642
+    },
+    {
+      "epoch": 1.1534968210717529,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.653220105028476e-06,
+      "loss": 0.3409,
+      "step": 643
+    },
+    {
+      "epoch": 1.1553133514986376,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.62209709315362e-06,
+      "loss": 0.3475,
+      "step": 644
+    },
+    {
+      "epoch": 1.1571298819255222,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.590987680624174e-06,
+      "loss": 0.3365,
+      "step": 645
+    },
+    {
+      "epoch": 1.158946412352407,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.559892174477478e-06,
+      "loss": 0.3369,
+      "step": 646
+    },
+    {
+      "epoch": 1.1607629427792916,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.528810881613626e-06,
+      "loss": 0.3328,
+      "step": 647
+    },
+    {
+      "epoch": 1.1625794732061763,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.49774410879243e-06,
+      "loss": 0.3401,
+      "step": 648
+    },
+    {
+      "epoch": 1.164396003633061,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.466692162630393e-06,
+      "loss": 0.3511,
+      "step": 649
+    },
+    {
+      "epoch": 1.1662125340599454,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 0.3406,
+      "step": 650
+    },
+    {
+      "epoch": 1.16802906448683,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.404633976015136e-06,
+      "loss": 0.3422,
+      "step": 651
+    },
+    {
+      "epoch": 1.1698455949137148,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.373628348051165e-06,
+      "loss": 0.3444,
+      "step": 652
+    },
+    {
+      "epoch": 1.1716621253405994,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.342638771718804e-06,
+      "loss": 0.3422,
+      "step": 653
+    },
+    {
+      "epoch": 1.1734786557674841,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.311665552872662e-06,
+      "loss": 0.3518,
+      "step": 654
+    },
+    {
+      "epoch": 1.1752951861943688,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.280708997205904e-06,
+      "loss": 0.3381,
+      "step": 655
+    },
+    {
+      "epoch": 1.1771117166212535,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.249769410247239e-06,
+      "loss": 0.3423,
+      "step": 656
+    },
+    {
+      "epoch": 1.1789282470481381,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.218847097357898e-06,
+      "loss": 0.3313,
+      "step": 657
+    },
+    {
+      "epoch": 1.1807447774750228,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.187942363728626e-06,
+      "loss": 0.3463,
+      "step": 658
+    },
+    {
+      "epoch": 1.1825613079019073,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.157055514376667e-06,
+      "loss": 0.3397,
+      "step": 659
+    },
+    {
+      "epoch": 1.184377838328792,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.126186854142752e-06,
+      "loss": 0.3409,
+      "step": 660
+    },
+    {
+      "epoch": 1.1861943687556766,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.095336687688102e-06,
+      "loss": 0.3326,
+      "step": 661
+    },
+    {
+      "epoch": 1.1880108991825613,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.064505319491398e-06,
+      "loss": 0.3404,
+      "step": 662
+    },
+    {
+      "epoch": 1.189827429609446,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.033693053845801e-06,
+      "loss": 0.34,
+      "step": 663
+    },
+    {
+      "epoch": 1.1916439600363307,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.00290019485593e-06,
+      "loss": 0.3429,
+      "step": 664
+    },
+    {
+      "epoch": 1.1934604904632153,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.972127046434878e-06,
+      "loss": 0.3455,
+      "step": 665
+    },
+    {
+      "epoch": 1.1952770208900998,
+      "grad_norm": 1.0,
+      "learning_rate": 7.94137391230119e-06,
+      "loss": 0.3388,
+      "step": 666
+    },
+    {
+      "epoch": 1.1970935513169845,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.910641095975886e-06,
+      "loss": 0.3381,
+      "step": 667
+    },
+    {
+      "epoch": 1.1989100817438691,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.879928900779457e-06,
+      "loss": 0.3498,
+      "step": 668
+    },
+    {
+      "epoch": 1.2007266121707538,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.84923762982887e-06,
+      "loss": 0.3364,
+      "step": 669
+    },
+    {
+      "epoch": 1.2025431425976385,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.818567586034578e-06,
+      "loss": 0.3342,
+      "step": 670
+    },
+    {
+      "epoch": 1.2043596730245232,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.787919072097531e-06,
+      "loss": 0.3367,
+      "step": 671
+    },
+    {
+      "epoch": 1.2061762034514079,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.757292390506191e-06,
+      "loss": 0.3368,
+      "step": 672
+    },
+    {
+      "epoch": 1.2079927338782925,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.726687843533539e-06,
+      "loss": 0.3439,
+      "step": 673
+    },
+    {
+      "epoch": 1.2098092643051772,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.696105733234099e-06,
+      "loss": 0.3384,
+      "step": 674
+    },
+    {
+      "epoch": 1.2116257947320617,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.66554636144095e-06,
+      "loss": 0.3425,
+      "step": 675
+    },
+    {
+      "epoch": 1.2134423251589463,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.635010029762755e-06,
+      "loss": 0.3453,
+      "step": 676
+    },
+    {
+      "epoch": 1.215258855585831,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.604497039580785e-06,
+      "loss": 0.3526,
+      "step": 677
+    },
+    {
+      "epoch": 1.2170753860127157,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.574007692045928e-06,
+      "loss": 0.3433,
+      "step": 678
+    },
+    {
+      "epoch": 1.2188919164396004,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.543542288075739e-06,
+      "loss": 0.3406,
+      "step": 679
+    },
+    {
+      "epoch": 1.220708446866485,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.513101128351454e-06,
+      "loss": 0.3413,
+      "step": 680
+    },
+    {
+      "epoch": 1.2225249772933697,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.482684513315031e-06,
+      "loss": 0.3306,
+      "step": 681
+    },
+    {
+      "epoch": 1.2243415077202544,
+      "grad_norm": 0.875,
+      "learning_rate": 7.4522927431661805e-06,
+      "loss": 0.3413,
+      "step": 682
+    },
+    {
+      "epoch": 1.226158038147139,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.421926117859403e-06,
+      "loss": 0.3347,
+      "step": 683
+    },
+    {
+      "epoch": 1.2279745685740235,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.391584937101034e-06,
+      "loss": 0.3496,
+      "step": 684
+    },
+    {
+      "epoch": 1.2297910990009082,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.361269500346274e-06,
+      "loss": 0.3389,
+      "step": 685
+    },
+    {
+      "epoch": 1.231607629427793,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.330980106796247e-06,
+      "loss": 0.3411,
+      "step": 686
+    },
+    {
+      "epoch": 1.2334241598546776,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.300717055395039e-06,
+      "loss": 0.3486,
+      "step": 687
+    },
+    {
+      "epoch": 1.2352406902815622,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.27048064482675e-06,
+      "loss": 0.3451,
+      "step": 688
+    },
+    {
+      "epoch": 1.237057220708447,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.240271173512545e-06,
+      "loss": 0.342,
+      "step": 689
+    },
+    {
+      "epoch": 1.2388737511353316,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.210088939607709e-06,
+      "loss": 0.3494,
+      "step": 690
+    },
+    {
+      "epoch": 1.240690281562216,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.179934240998707e-06,
+      "loss": 0.3402,
+      "step": 691
+    },
+    {
+      "epoch": 1.2425068119891007,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.149807375300239e-06,
+      "loss": 0.3352,
+      "step": 692
+    },
+    {
+      "epoch": 1.2443233424159854,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.119708639852312e-06,
+      "loss": 0.342,
+      "step": 693
+    },
+    {
+      "epoch": 1.24613987284287,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.0896383317172845e-06,
+      "loss": 0.3467,
+      "step": 694
+    },
+    {
+      "epoch": 1.2479564032697548,
+      "grad_norm": 0.875,
+      "learning_rate": 7.059596747676963e-06,
+      "loss": 0.3498,
+      "step": 695
+    },
+    {
+      "epoch": 1.2497729336966394,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.029584184229653e-06,
+      "loss": 0.3516,
+      "step": 696
+    },
+    {
+      "epoch": 1.2515894641235241,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.99960093758724e-06,
+      "loss": 0.336,
+      "step": 697
+    },
+    {
+      "epoch": 1.2534059945504088,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.969647303672262e-06,
+      "loss": 0.331,
+      "step": 698
+    },
+    {
+      "epoch": 1.2552225249772935,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.9397235781149945e-06,
+      "loss": 0.3454,
+      "step": 699
+    },
+    {
+      "epoch": 1.257039055404178,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.3413,
+      "step": 700
+    },
+    {
+      "epoch": 1.2588555858310626,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.879967033115853e-06,
+      "loss": 0.3498,
+      "step": 701
+    },
+    {
+      "epoch": 1.2606721162579473,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.850134803446955e-06,
+      "loss": 0.3397,
+      "step": 702
+    },
+    {
+      "epoch": 1.262488646684832,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.820333661675893e-06,
+      "loss": 0.3487,
+      "step": 703
+    },
+    {
+      "epoch": 1.2643051771117166,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.790563901927907e-06,
+      "loss": 0.345,
+      "step": 704
+    },
+    {
+      "epoch": 1.2661217075386013,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.7608258180185085e-06,
+      "loss": 0.3491,
+      "step": 705
+    },
+    {
+      "epoch": 1.267938237965486,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.731119703450577e-06,
+      "loss": 0.3435,
+      "step": 706
+    },
+    {
+      "epoch": 1.2697547683923704,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.701445851411472e-06,
+      "loss": 0.3416,
+      "step": 707
+    },
+    {
+      "epoch": 1.2715712988192553,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.671804554770135e-06,
+      "loss": 0.3492,
+      "step": 708
+    },
+    {
+      "epoch": 1.2733878292461398,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.642196106074195e-06,
+      "loss": 0.3336,
+      "step": 709
+    },
+    {
+      "epoch": 1.2752043596730245,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.612620797547087e-06,
+      "loss": 0.3355,
+      "step": 710
+    },
+    {
+      "epoch": 1.2770208900999092,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.583078921085167e-06,
+      "loss": 0.3373,
+      "step": 711
+    },
+    {
+      "epoch": 1.2788374205267938,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.553570768254831e-06,
+      "loss": 0.3432,
+      "step": 712
+    },
+    {
+      "epoch": 1.2806539509536785,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.524096630289632e-06,
+      "loss": 0.3476,
+      "step": 713
+    },
+    {
+      "epoch": 1.2824704813805632,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.494656798087412e-06,
+      "loss": 0.3346,
+      "step": 714
+    },
+    {
+      "epoch": 1.2842870118074479,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.465251562207431e-06,
+      "loss": 0.3354,
+      "step": 715
+    },
+    {
+      "epoch": 1.2861035422343323,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.435881212867494e-06,
+      "loss": 0.3356,
+      "step": 716
+    },
+    {
+      "epoch": 1.287920072661217,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.406546039941095e-06,
+      "loss": 0.3426,
+      "step": 717
+    },
+    {
+      "epoch": 1.2897366030881017,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.377246332954544e-06,
+      "loss": 0.3406,
+      "step": 718
+    },
+    {
+      "epoch": 1.2915531335149864,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.3479823810841235e-06,
+      "loss": 0.3449,
+      "step": 719
+    },
+    {
+      "epoch": 1.293369663941871,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.318754473153221e-06,
+      "loss": 0.3411,
+      "step": 720
+    },
+    {
+      "epoch": 1.2951861943687557,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.289562897629492e-06,
+      "loss": 0.3363,
+      "step": 721
+    },
+    {
+      "epoch": 1.2970027247956404,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.260407942621998e-06,
+      "loss": 0.3453,
+      "step": 722
+    },
+    {
+      "epoch": 1.298819255222525,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.231289895878375e-06,
+      "loss": 0.3467,
+      "step": 723
+    },
+    {
+      "epoch": 1.3006357856494097,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.202209044781991e-06,
+      "loss": 0.3431,
+      "step": 724
+    },
+    {
+      "epoch": 1.3024523160762942,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 0.3407,
+      "step": 725
+    },
+    {
+      "epoch": 1.3042688465031789,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.144160077226035e-06,
+      "loss": 0.3436,
+      "step": 726
+    },
+    {
+      "epoch": 1.3060853769300635,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.115192533686341e-06,
+      "loss": 0.3355,
+      "step": 727
+    },
+    {
+      "epoch": 1.3079019073569482,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.086263331627976e-06,
+      "loss": 0.3365,
+      "step": 728
+    },
+    {
+      "epoch": 1.309718437783833,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.05737275657049e-06,
+      "loss": 0.3462,
+      "step": 729
+    },
+    {
+      "epoch": 1.3115349682107176,
+      "grad_norm": 0.875,
+      "learning_rate": 6.028521093652195e-06,
+      "loss": 0.3438,
+      "step": 730
+    },
+    {
+      "epoch": 1.3133514986376023,
+      "grad_norm": 0.86328125,
+      "learning_rate": 5.9997086276273545e-06,
+      "loss": 0.3435,
+      "step": 731
+    },
+    {
+      "epoch": 1.3151680290644867,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.970935642863375e-06,
+      "loss": 0.3446,
+      "step": 732
+    },
+    {
+      "epoch": 1.3169845594913716,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.942202423338001e-06,
+      "loss": 0.3313,
+      "step": 733
+    },
+    {
+      "epoch": 1.318801089918256,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.913509252636511e-06,
+      "loss": 0.3389,
+      "step": 734
+    },
+    {
+      "epoch": 1.3206176203451407,
+      "grad_norm": 0.86328125,
+      "learning_rate": 5.884856413948913e-06,
+      "loss": 0.3492,
+      "step": 735
+    },
+    {
+      "epoch": 1.3224341507720254,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.85624419006716e-06,
+      "loss": 0.3399,
+      "step": 736
+    },
+    {
+      "epoch": 1.3224341507720254,
+      "eval_loss": 0.44357970356941223,
+      "eval_runtime": 34.9269,
+      "eval_samples_per_second": 5.726,
+      "eval_steps_per_second": 1.432,
+      "step": 736
+    },
+    {
+      "epoch": 1.32425068119891,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.8276728633823494e-06,
+      "loss": 0.3642,
+      "step": 737
+    },
+    {
+      "epoch": 1.3260672116257948,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.799142715881938e-06,
+      "loss": 0.3379,
+      "step": 738
+    },
+    {
+      "epoch": 1.3278837420526794,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.770654029146969e-06,
+      "loss": 0.3337,
+      "step": 739
+    },
+    {
+      "epoch": 1.3297002724795641,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.742207084349274e-06,
+      "loss": 0.3453,
+      "step": 740
+    },
+    {
+      "epoch": 1.3315168029064486,
+      "grad_norm": 0.89453125,
+      "learning_rate": 5.713802162248718e-06,
+      "loss": 0.3498,
+      "step": 741
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.875,
+      "learning_rate": 5.685439543190409e-06,
+      "loss": 0.3521,
+      "step": 742
+    },
+    {
+      "epoch": 1.335149863760218,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.657119507101955e-06,
+      "loss": 0.3488,
+      "step": 743
+    },
+    {
+      "epoch": 1.3369663941871026,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.628842333490674e-06,
+      "loss": 0.342,
+      "step": 744
+    },
+    {
+      "epoch": 1.3387829246139873,
+      "grad_norm": 0.875,
+      "learning_rate": 5.600608301440848e-06,
+      "loss": 0.3399,
+      "step": 745
+    },
+    {
+      "epoch": 1.340599455040872,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.572417689610987e-06,
+      "loss": 0.3481,
+      "step": 746
+    },
+    {
+      "epoch": 1.3424159854677566,
+      "grad_norm": 0.84375,
+      "learning_rate": 5.544270776231038e-06,
+      "loss": 0.3387,
+      "step": 747
+    },
+    {
+      "epoch": 1.344232515894641,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.516167839099679e-06,
+      "loss": 0.3283,
+      "step": 748
+    },
+    {
+      "epoch": 1.346049046321526,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.488109155581549e-06,
+      "loss": 0.3349,
+      "step": 749
+    },
+    {
+      "epoch": 1.3478655767484105,
+      "grad_norm": 0.89453125,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 0.3481,
+      "step": 750
+    },
+    {
+      "epoch": 1.3496821071752951,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.432125656657004e-06,
+      "loss": 0.34,
+      "step": 751
+    },
+    {
+      "epoch": 1.3514986376021798,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.404201393785123e-06,
+      "loss": 0.3452,
+      "step": 752
+    },
+    {
+      "epoch": 1.3533151680290645,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.376322489590085e-06,
+      "loss": 0.3392,
+      "step": 753
+    },
+    {
+      "epoch": 1.3551316984559492,
+      "grad_norm": 0.8828125,
+      "learning_rate": 5.348489219225417e-06,
+      "loss": 0.3452,
+      "step": 754
+    },
+    {
+      "epoch": 1.3569482288828338,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.3207018573942684e-06,
+      "loss": 0.3487,
+      "step": 755
+    },
+    {
+      "epoch": 1.3587647593097185,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.292960678346674e-06,
+      "loss": 0.3471,
+      "step": 756
+    },
+    {
+      "epoch": 1.360581289736603,
+      "grad_norm": 0.8828125,
+      "learning_rate": 5.2652659558768795e-06,
+      "loss": 0.3419,
+      "step": 757
+    },
+    {
+      "epoch": 1.3623978201634879,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.237617963320608e-06,
+      "loss": 0.3497,
+      "step": 758
+    },
+    {
+      "epoch": 1.3642143505903723,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.2100169735523906e-06,
+      "loss": 0.3418,
+      "step": 759
+    },
+    {
+      "epoch": 1.366030881017257,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.1824632589828465e-06,
+      "loss": 0.343,
+      "step": 760
+    },
+    {
+      "epoch": 1.3678474114441417,
+      "grad_norm": 0.87890625,
+      "learning_rate": 5.154957091556021e-06,
+      "loss": 0.3426,
+      "step": 761
+    },
+    {
+      "epoch": 1.3696639418710264,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.127498742746675e-06,
+      "loss": 0.3331,
+      "step": 762
+    },
+    {
+      "epoch": 1.371480472297911,
+      "grad_norm": 0.83984375,
+      "learning_rate": 5.100088483557635e-06,
+      "loss": 0.341,
+      "step": 763
+    },
+    {
+      "epoch": 1.3732970027247957,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.072726584517086e-06,
+      "loss": 0.3512,
+      "step": 764
+    },
+    {
+      "epoch": 1.3751135331516804,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.045413315675925e-06,
+      "loss": 0.3398,
+      "step": 765
+    },
+    {
+      "epoch": 1.3769300635785648,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.018148946605092e-06,
+      "loss": 0.3407,
+      "step": 766
+    },
+    {
+      "epoch": 1.3787465940054495,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.9909337463929e-06,
+      "loss": 0.3533,
+      "step": 767
+    },
+    {
+      "epoch": 1.3805631244323342,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.9637679836423926e-06,
+      "loss": 0.3406,
+      "step": 768
+    },
+    {
+      "epoch": 1.3823796548592189,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.936651926468673e-06,
+      "loss": 0.344,
+      "step": 769
+    },
+    {
+      "epoch": 1.3841961852861036,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.909585842496287e-06,
+      "loss": 0.3563,
+      "step": 770
+    },
+    {
+      "epoch": 1.3860127157129882,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.882569998856549e-06,
+      "loss": 0.3295,
+      "step": 771
+    },
+    {
+      "epoch": 1.387829246139873,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.855604662184935e-06,
+      "loss": 0.3418,
+      "step": 772
+    },
+    {
+      "epoch": 1.3896457765667574,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.828690098618429e-06,
+      "loss": 0.3368,
+      "step": 773
+    },
+    {
+      "epoch": 1.3914623069936423,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.801826573792905e-06,
+      "loss": 0.3392,
+      "step": 774
+    },
+    {
+      "epoch": 1.3932788374205267,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.775014352840512e-06,
+      "loss": 0.3567,
+      "step": 775
+    },
+    {
+      "epoch": 1.3950953678474114,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.7482537003870425e-06,
+      "loss": 0.3446,
+      "step": 776
+    },
+    {
+      "epoch": 1.396911898274296,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.721544880549337e-06,
+      "loss": 0.3317,
+      "step": 777
+    },
+    {
+      "epoch": 1.3987284287011807,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.694888156932657e-06,
+      "loss": 0.3478,
+      "step": 778
+    },
+    {
+      "epoch": 1.4005449591280654,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.668283792628114e-06,
+      "loss": 0.3356,
+      "step": 779
+    },
+    {
+      "epoch": 1.40236148955495,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.641732050210032e-06,
+      "loss": 0.345,
+      "step": 780
+    },
+    {
+      "epoch": 1.4041780199818348,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.6152331917333985e-06,
+      "loss": 0.3434,
+      "step": 781
+    },
+    {
+      "epoch": 1.4059945504087192,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.588787478731242e-06,
+      "loss": 0.3433,
+      "step": 782
+    },
+    {
+      "epoch": 1.4078110808356041,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.562395172212074e-06,
+      "loss": 0.3407,
+      "step": 783
+    },
+    {
+      "epoch": 1.4096276112624886,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.53605653265731e-06,
+      "loss": 0.3374,
+      "step": 784
+    },
+    {
+      "epoch": 1.4114441416893733,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.509771820018682e-06,
+      "loss": 0.3409,
+      "step": 785
+    },
+    {
+      "epoch": 1.413260672116258,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.483541293715699e-06,
+      "loss": 0.3414,
+      "step": 786
+    },
+    {
+      "epoch": 1.4150772025431426,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.457365212633058e-06,
+      "loss": 0.3407,
+      "step": 787
+    },
+    {
+      "epoch": 1.4168937329700273,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.4312438351181246e-06,
+      "loss": 0.3453,
+      "step": 788
+    },
+    {
+      "epoch": 1.418710263396912,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.405177418978331e-06,
+      "loss": 0.3411,
+      "step": 789
+    },
+    {
+      "epoch": 1.4205267938237967,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.379166221478697e-06,
+      "loss": 0.3407,
+      "step": 790
+    },
+    {
+      "epoch": 1.422343324250681,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.353210499339231e-06,
+      "loss": 0.3279,
+      "step": 791
+    },
+    {
+      "epoch": 1.4241598546775658,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.3273105087324375e-06,
+      "loss": 0.3343,
+      "step": 792
+    },
+    {
+      "epoch": 1.4259763851044505,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.301466505280763e-06,
+      "loss": 0.337,
+      "step": 793
+    },
+    {
+      "epoch": 1.4277929155313351,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.275678744054094e-06,
+      "loss": 0.345,
+      "step": 794
+    },
+    {
+      "epoch": 1.4296094459582198,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.249947479567218e-06,
+      "loss": 0.338,
+      "step": 795
+    },
+    {
+      "epoch": 1.4314259763851045,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.224272965777326e-06,
+      "loss": 0.3353,
+      "step": 796
+    },
+    {
+      "epoch": 1.4332425068119892,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.1986554560815095e-06,
+      "loss": 0.3419,
+      "step": 797
+    },
+    {
+      "epoch": 1.4350590372388736,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.173095203314241e-06,
+      "loss": 0.3414,
+      "step": 798
+    },
+    {
+      "epoch": 1.4368755676657585,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.1475924597449025e-06,
+      "loss": 0.3426,
+      "step": 799
+    },
+    {
+      "epoch": 1.438692098092643,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.3359,
+      "step": 800
+    },
+    {
+      "epoch": 1.4405086285195277,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.096760506437057e-06,
+      "loss": 0.3328,
+      "step": 801
+    },
+    {
+      "epoch": 1.4423251589464123,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.071431798389408e-06,
+      "loss": 0.3424,
+      "step": 802
+    },
+    {
+      "epoch": 1.444141689373297,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.046161602916453e-06,
+      "loss": 0.3317,
+      "step": 803
+    },
+    {
+      "epoch": 1.4459582198001817,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.020950169424815e-06,
+      "loss": 0.3338,
+      "step": 804
+    },
+    {
+      "epoch": 1.4477747502270664,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.9957977467411615e-06,
+      "loss": 0.3407,
+      "step": 805
+    },
+    {
+      "epoch": 1.449591280653951,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.970704583109755e-06,
+      "loss": 0.3311,
+      "step": 806
+    },
+    {
+      "epoch": 1.4514078110808355,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.945670926189987e-06,
+      "loss": 0.3431,
+      "step": 807
+    },
+    {
+      "epoch": 1.4532243415077202,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.920697023053949e-06,
+      "loss": 0.3485,
+      "step": 808
+    },
+    {
+      "epoch": 1.4550408719346049,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.895783120183975e-06,
+      "loss": 0.3456,
+      "step": 809
+    },
+    {
+      "epoch": 1.4568574023614895,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.8709294634702374e-06,
+      "loss": 0.3464,
+      "step": 810
+    },
+    {
+      "epoch": 1.4586739327883742,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.846136298208285e-06,
+      "loss": 0.3385,
+      "step": 811
+    },
+    {
+      "epoch": 1.4604904632152589,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.821403869096658e-06,
+      "loss": 0.3457,
+      "step": 812
+    },
+    {
+      "epoch": 1.4623069936421436,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.7967324202344433e-06,
+      "loss": 0.3526,
+      "step": 813
+    },
+    {
+      "epoch": 1.4641235240690282,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.772122195118877e-06,
+      "loss": 0.3345,
+      "step": 814
+    },
+    {
+      "epoch": 1.465940054495913,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.747573436642952e-06,
+      "loss": 0.345,
+      "step": 815
+    },
+    {
+      "epoch": 1.4677565849227974,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.723086387092997e-06,
+      "loss": 0.341,
+      "step": 816
+    },
+    {
+      "epoch": 1.469573115349682,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.6986612881463114e-06,
+      "loss": 0.3393,
+      "step": 817
+    },
+    {
+      "epoch": 1.4713896457765667,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.674298380868756e-06,
+      "loss": 0.3461,
+      "step": 818
+    },
+    {
+      "epoch": 1.4732061762034514,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.649997905712396e-06,
+      "loss": 0.3431,
+      "step": 819
+    },
+    {
+      "epoch": 1.475022706630336,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.625760102513103e-06,
+      "loss": 0.3357,
+      "step": 820
+    },
+    {
+      "epoch": 1.4768392370572208,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.601585210488218e-06,
+      "loss": 0.3422,
+      "step": 821
+    },
+    {
+      "epoch": 1.4786557674841054,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.5774734682341563e-06,
+      "loss": 0.3447,
+      "step": 822
+    },
+    {
+      "epoch": 1.4804722979109899,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.5534251137240883e-06,
+      "loss": 0.3364,
+      "step": 823
+    },
+    {
+      "epoch": 1.4822888283378748,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.5294403843055604e-06,
+      "loss": 0.3326,
+      "step": 824
+    },
+    {
+      "epoch": 1.4841053587647592,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.505519516698165e-06,
+      "loss": 0.3369,
+      "step": 825
+    },
+    {
+      "epoch": 1.485921889191644,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.4816627469912147e-06,
+      "loss": 0.3453,
+      "step": 826
+    },
+    {
+      "epoch": 1.4877384196185286,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.4578703106413903e-06,
+      "loss": 0.3452,
+      "step": 827
+    },
+    {
+      "epoch": 1.4895549500454133,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.4341424424704373e-06,
+      "loss": 0.3401,
+      "step": 828
+    },
+    {
+      "epoch": 1.491371480472298,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.4104793766628307e-06,
+      "loss": 0.3314,
+      "step": 829
+    },
+    {
+      "epoch": 1.4931880108991826,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.3868813467634833e-06,
+      "loss": 0.3407,
+      "step": 830
+    },
+    {
+      "epoch": 1.4950045413260673,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.3633485856754143e-06,
+      "loss": 0.3467,
+      "step": 831
+    },
+    {
+      "epoch": 1.4968210717529518,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.3398813256574847e-06,
+      "loss": 0.3419,
+      "step": 832
+    },
+    {
+      "epoch": 1.4986376021798364,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.316479798322072e-06,
+      "loss": 0.3404,
+      "step": 833
+    },
+    {
+      "epoch": 1.5004541326067211,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.2931442346328e-06,
+      "loss": 0.3349,
+      "step": 834
+    },
+    {
+      "epoch": 1.5022706630336058,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.2698748649022693e-06,
+      "loss": 0.3362,
+      "step": 835
+    },
+    {
+      "epoch": 1.5040871934604905,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.2466719187897555e-06,
+      "loss": 0.3388,
+      "step": 836
+    },
+    {
+      "epoch": 1.5059037238873751,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.223535625298979e-06,
+      "loss": 0.3294,
+      "step": 837
+    },
+    {
+      "epoch": 1.5077202543142598,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.200466212775808e-06,
+      "loss": 0.336,
+      "step": 838
+    },
+    {
+      "epoch": 1.5095367847411443,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.1774639089060364e-06,
+      "loss": 0.3325,
+      "step": 839
+    },
+    {
+      "epoch": 1.5113533151680292,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.1545289407131128e-06,
+      "loss": 0.3389,
+      "step": 840
+    },
+    {
+      "epoch": 1.5131698455949136,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.1316615345559188e-06,
+      "loss": 0.3342,
+      "step": 841
+    },
+    {
+      "epoch": 1.5149863760217985,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.108861916126518e-06,
+      "loss": 0.3461,
+      "step": 842
+    },
+    {
+      "epoch": 1.516802906448683,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.086130310447937e-06,
+      "loss": 0.3425,
+      "step": 843
+    },
+    {
+      "epoch": 1.5186194368755677,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.063466941871952e-06,
+      "loss": 0.3495,
+      "step": 844
+    },
+    {
+      "epoch": 1.5204359673024523,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.040872034076857e-06,
+      "loss": 0.3436,
+      "step": 845
+    },
+    {
+      "epoch": 1.522252497729337,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.0183458100652752e-06,
+      "loss": 0.337,
+      "step": 846
+    },
+    {
+      "epoch": 1.5240690281562217,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.9958884921619368e-06,
+      "loss": 0.3339,
+      "step": 847
+    },
+    {
+      "epoch": 1.5258855585831061,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.9735003020115095e-06,
+      "loss": 0.3345,
+      "step": 848
+    },
+    {
+      "epoch": 1.527702089009991,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.9511814605763855e-06,
+      "loss": 0.3447,
+      "step": 849
+    },
+    {
+      "epoch": 1.5295186194368755,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.343,
+      "step": 850
+    },
+    {
+      "epoch": 1.5313351498637602,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.9067527042772638e-06,
+      "loss": 0.3298,
+      "step": 851
+    },
+    {
+      "epoch": 1.5331516802906449,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.884643227907147e-06,
+      "loss": 0.3433,
+      "step": 852
+    },
+    {
+      "epoch": 1.5349682107175295,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.8626039772357884e-06,
+      "loss": 0.3347,
+      "step": 853
+    },
+    {
+      "epoch": 1.5367847411444142,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.840635169781688e-06,
+      "loss": 0.3462,
+      "step": 854
+    },
+    {
+      "epoch": 1.5386012715712987,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.8187370223681134e-06,
+      "loss": 0.3375,
+      "step": 855
+    },
+    {
+      "epoch": 1.5404178019981836,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.796909751120931e-06,
+      "loss": 0.3402,
+      "step": 856
+    },
+    {
+      "epoch": 1.542234332425068,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.7751535714665025e-06,
+      "loss": 0.3466,
+      "step": 857
+    },
+    {
+      "epoch": 1.544050862851953,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.7534686981295335e-06,
+      "loss": 0.3502,
+      "step": 858
+    },
+    {
+      "epoch": 1.5458673932788374,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.7318553451309726e-06,
+      "loss": 0.3383,
+      "step": 859
+    },
+    {
+      "epoch": 1.547683923705722,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.7103137257858867e-06,
+      "loss": 0.3427,
+      "step": 860
+    },
+    {
+      "epoch": 1.5495004541326067,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.6888440527013595e-06,
+      "loss": 0.3398,
+      "step": 861
+    },
+    {
+      "epoch": 1.5513169845594914,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.667446537774402e-06,
+      "loss": 0.3353,
+      "step": 862
+    },
+    {
+      "epoch": 1.553133514986376,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.646121392189841e-06,
+      "loss": 0.3464,
+      "step": 863
+    },
+    {
+      "epoch": 1.5549500454132605,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.624868826418262e-06,
+      "loss": 0.3435,
+      "step": 864
+    },
+    {
+      "epoch": 1.5567665758401454,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.603689050213902e-06,
+      "loss": 0.3358,
+      "step": 865
+    },
+    {
+      "epoch": 1.55858310626703,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.5825822726126095e-06,
+      "loss": 0.333,
+      "step": 866
+    },
+    {
+      "epoch": 1.5603996366939146,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.561548701929749e-06,
+      "loss": 0.3379,
+      "step": 867
+    },
+    {
+      "epoch": 1.5622161671207992,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.5405885457581793e-06,
+      "loss": 0.3401,
+      "step": 868
+    },
+    {
+      "epoch": 1.564032697547684,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.5197020109661775e-06,
+      "loss": 0.3421,
+      "step": 869
+    },
+    {
+      "epoch": 1.5658492279745686,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.4988893036954045e-06,
+      "loss": 0.3337,
+      "step": 870
+    },
+    {
+      "epoch": 1.5676657584014533,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.4781506293588876e-06,
+      "loss": 0.3446,
+      "step": 871
+    },
+    {
+      "epoch": 1.569482288828338,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.4574861926389615e-06,
+      "loss": 0.3334,
+      "step": 872
+    },
+    {
+      "epoch": 1.5712988192552224,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.436896197485282e-06,
+      "loss": 0.3385,
+      "step": 873
+    },
+    {
+      "epoch": 1.5731153496821073,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.4163808471127815e-06,
+      "loss": 0.3416,
+      "step": 874
+    },
+    {
+      "epoch": 1.5749318801089918,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.395940343999691e-06,
+      "loss": 0.3371,
+      "step": 875
+    },
+    {
+      "epoch": 1.5767484105358764,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.37557488988552e-06,
+      "loss": 0.3413,
+      "step": 876
+    },
+    {
+      "epoch": 1.5785649409627611,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.3552846857690847e-06,
+      "loss": 0.3454,
+      "step": 877
+    },
+    {
+      "epoch": 1.5803814713896458,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.335069931906503e-06,
+      "loss": 0.3287,
+      "step": 878
+    },
+    {
+      "epoch": 1.5821980018165305,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.3149308278092343e-06,
+      "loss": 0.3271,
+      "step": 879
+    },
+    {
+      "epoch": 1.584014532243415,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.2948675722421086e-06,
+      "loss": 0.3418,
+      "step": 880
+    },
+    {
+      "epoch": 1.5858310626702998,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.2748803632213556e-06,
+      "loss": 0.352,
+      "step": 881
+    },
+    {
+      "epoch": 1.5876475930971843,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.254969398012663e-06,
+      "loss": 0.3318,
+      "step": 882
+    },
+    {
+      "epoch": 1.5894641235240692,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.2351348731292134e-06,
+      "loss": 0.3436,
+      "step": 883
+    },
+    {
+      "epoch": 1.5912806539509536,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.215376984329767e-06,
+      "loss": 0.3498,
+      "step": 884
+    },
+    {
+      "epoch": 1.5930971843778383,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.195695926616702e-06,
+      "loss": 0.343,
+      "step": 885
+    },
+    {
+      "epoch": 1.594913714804723,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.1760918942341193e-06,
+      "loss": 0.3329,
+      "step": 886
+    },
+    {
+      "epoch": 1.5967302452316077,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.1565650806658977e-06,
+      "loss": 0.3402,
+      "step": 887
+    },
+    {
+      "epoch": 1.5985467756584923,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.1371156786338108e-06,
+      "loss": 0.3334,
+      "step": 888
+    },
+    {
+      "epoch": 1.6003633060853768,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.117743880095601e-06,
+      "loss": 0.3472,
+      "step": 889
+    },
+    {
+      "epoch": 1.6021798365122617,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.098449876243096e-06,
+      "loss": 0.3428,
+      "step": 890
+    },
+    {
+      "epoch": 1.6039963669391462,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.0792338575003303e-06,
+      "loss": 0.3436,
+      "step": 891
+    },
+    {
+      "epoch": 1.6058128973660308,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.0600960135216463e-06,
+      "loss": 0.3301,
+      "step": 892
+    },
+    {
+      "epoch": 1.6076294277929155,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.041036533189842e-06,
+      "loss": 0.3524,
+      "step": 893
+    },
+    {
+      "epoch": 1.6094459582198002,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.022055604614289e-06,
+      "loss": 0.3451,
+      "step": 894
+    },
+    {
+      "epoch": 1.6112624886466849,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.0031534151290944e-06,
+      "loss": 0.3428,
+      "step": 895
+    },
+    {
+      "epoch": 1.6130790190735693,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.984330151291233e-06,
+      "loss": 0.3421,
+      "step": 896
+    },
+    {
+      "epoch": 1.6148955495004542,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.965585998878724e-06,
+      "loss": 0.3392,
+      "step": 897
+    },
+    {
+      "epoch": 1.6167120799273387,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9469211428887813e-06,
+      "loss": 0.3494,
+      "step": 898
+    },
+    {
+      "epoch": 1.6185286103542236,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.928335767535997e-06,
+      "loss": 0.3348,
+      "step": 899
+    },
+    {
+      "epoch": 1.620345140781108,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.3453,
+      "step": 900
+    },
+    {
+      "epoch": 1.6221616712079927,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.8914041916762648e-06,
+      "loss": 0.3375,
+      "step": 901
+    },
+    {
+      "epoch": 1.6239782016348774,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.8730583556690607e-06,
+      "loss": 0.3459,
+      "step": 902
+    },
+    {
+      "epoch": 1.625794732061762,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.8547927292949053e-06,
+      "loss": 0.3314,
+      "step": 903
+    },
+    {
+      "epoch": 1.6276112624886467,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.8366074928281608e-06,
+      "loss": 0.3385,
+      "step": 904
+    },
+    {
+      "epoch": 1.6294277929155312,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.818502825749764e-06,
+      "loss": 0.3377,
+      "step": 905
+    },
+    {
+      "epoch": 1.631244323342416,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.8004789067454763e-06,
+      "loss": 0.3427,
+      "step": 906
+    },
+    {
+      "epoch": 1.6330608537693005,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.7825359137040987e-06,
+      "loss": 0.3375,
+      "step": 907
+    },
+    {
+      "epoch": 1.6348773841961854,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.7646740237157256e-06,
+      "loss": 0.3496,
+      "step": 908
+    },
+    {
+      "epoch": 1.63669391462307,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.7468934130700044e-06,
+      "loss": 0.3354,
+      "step": 909
+    },
+    {
+      "epoch": 1.6385104450499546,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.7291942572543806e-06,
+      "loss": 0.3357,
+      "step": 910
+    },
+    {
+      "epoch": 1.6403269754768393,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.7115767309523811e-06,
+      "loss": 0.3357,
+      "step": 911
+    },
+    {
+      "epoch": 1.642143505903724,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.6940410080418723e-06,
+      "loss": 0.347,
+      "step": 912
+    },
+    {
+      "epoch": 1.6439600363306086,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.6765872615933676e-06,
+      "loss": 0.3419,
+      "step": 913
+    },
+    {
+      "epoch": 1.645776566757493,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.6592156638682887e-06,
+      "loss": 0.3332,
+      "step": 914
+    },
+    {
+      "epoch": 1.647593097184378,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.6419263863172997e-06,
+      "loss": 0.3347,
+      "step": 915
+    },
+    {
+      "epoch": 1.6494096276112624,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6247195995785836e-06,
+      "loss": 0.3377,
+      "step": 916
+    },
+    {
+      "epoch": 1.651226158038147,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.6075954734761844e-06,
+      "loss": 0.3375,
+      "step": 917
+    },
+    {
+      "epoch": 1.6530426884650318,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.5905541770183096e-06,
+      "loss": 0.3358,
+      "step": 918
+    },
+    {
+      "epoch": 1.6548592188919165,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5735958783956795e-06,
+      "loss": 0.3416,
+      "step": 919
+    },
+    {
+      "epoch": 1.6566757493188011,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5567207449798517e-06,
+      "loss": 0.3394,
+      "step": 920
+    },
+    {
+      "epoch": 1.6566757493188011,
+      "eval_loss": 0.4413210153579712,
+      "eval_runtime": 34.9166,
+      "eval_samples_per_second": 5.728,
+      "eval_steps_per_second": 1.432,
+      "step": 920
+    },
+    {
+      "epoch": 1.6584922797456856,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5399289433215792e-06,
+      "loss": 0.3422,
+      "step": 921
+    },
+    {
+      "epoch": 1.6603088101725705,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.52322063914917e-06,
+      "loss": 0.3395,
+      "step": 922
+    },
+    {
+      "epoch": 1.662125340599455,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5065959973668355e-06,
+      "loss": 0.3285,
+      "step": 923
+    },
+    {
+      "epoch": 1.6639418710263398,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.490055182053083e-06,
+      "loss": 0.3453,
+      "step": 924
+    },
+    {
+      "epoch": 1.6657584014532243,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4735983564590784e-06,
+      "loss": 0.3375,
+      "step": 925
+    },
+    {
+      "epoch": 1.667574931880109,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4572256830070497e-06,
+      "loss": 0.3266,
+      "step": 926
+    },
+    {
+      "epoch": 1.6693914623069936,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4409373232886703e-06,
+      "loss": 0.3434,
+      "step": 927
+    },
+    {
+      "epoch": 1.6712079927338783,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4247334380634792e-06,
+      "loss": 0.3397,
+      "step": 928
+    },
+    {
+      "epoch": 1.673024523160763,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.408614187257279e-06,
+      "loss": 0.3388,
+      "step": 929
+    },
+    {
+      "epoch": 1.6748410535876475,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3925797299605649e-06,
+      "loss": 0.3447,
+      "step": 930
+    },
+    {
+      "epoch": 1.6766575840145324,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.3766302244269624e-06,
+      "loss": 0.3504,
+      "step": 931
+    },
+    {
+      "epoch": 1.6784741144414168,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3607658280716474e-06,
+      "loss": 0.3421,
+      "step": 932
+    },
+    {
+      "epoch": 1.6802906448683017,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.3449866974698123e-06,
+      "loss": 0.3419,
+      "step": 933
+    },
+    {
+      "epoch": 1.6821071752951862,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3292929883550998e-06,
+      "loss": 0.349,
+      "step": 934
+    },
+    {
+      "epoch": 1.6839237057220708,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3136848556180893e-06,
+      "loss": 0.3437,
+      "step": 935
+    },
+    {
+      "epoch": 1.6857402361489555,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2981624533047432e-06,
+      "loss": 0.3374,
+      "step": 936
+    },
+    {
+      "epoch": 1.6875567665758402,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.2827259346149123e-06,
+      "loss": 0.3435,
+      "step": 937
+    },
+    {
+      "epoch": 1.6893732970027249,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2673754519008008e-06,
+      "loss": 0.3402,
+      "step": 938
+    },
+    {
+      "epoch": 1.6911898274296093,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.2521111566654732e-06,
+      "loss": 0.3458,
+      "step": 939
+    },
+    {
+      "epoch": 1.6930063578564942,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.2369331995613664e-06,
+      "loss": 0.3275,
+      "step": 940
+    },
+    {
+      "epoch": 1.6948228882833787,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.2218417303887842e-06,
+      "loss": 0.3333,
+      "step": 941
+    },
+    {
+      "epoch": 1.6966394187102634,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.206836898094439e-06,
+      "loss": 0.3348,
+      "step": 942
+    },
+    {
+      "epoch": 1.698455949137148,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.1919188507699641e-06,
+      "loss": 0.3346,
+      "step": 943
+    },
+    {
+      "epoch": 1.7002724795640327,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1770877356504684e-06,
+      "loss": 0.3426,
+      "step": 944
+    },
+    {
+      "epoch": 1.7020890099909174,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1623436991130654e-06,
+      "loss": 0.3413,
+      "step": 945
+    },
+    {
+      "epoch": 1.7039055404178018,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1476868866754488e-06,
+      "loss": 0.3441,
+      "step": 946
+    },
+    {
+      "epoch": 1.7057220708446867,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1331174429944346e-06,
+      "loss": 0.3334,
+      "step": 947
+    },
+    {
+      "epoch": 1.7075386012715712,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1186355118645552e-06,
+      "loss": 0.3338,
+      "step": 948
+    },
+    {
+      "epoch": 1.709355131698456,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1042412362166221e-06,
+      "loss": 0.3419,
+      "step": 949
+    },
+    {
+      "epoch": 1.7111716621253406,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.0899347581163222e-06,
+      "loss": 0.34,
+      "step": 950
+    },
+    {
+      "epoch": 1.7129881925522252,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0757162187628223e-06,
+      "loss": 0.3387,
+      "step": 951
+    },
+    {
+      "epoch": 1.71480472297911,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0615857584873624e-06,
+      "loss": 0.3465,
+      "step": 952
+    },
+    {
+      "epoch": 1.7166212534059946,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0475435167518843e-06,
+      "loss": 0.3368,
+      "step": 953
+    },
+    {
+      "epoch": 1.7184377838328793,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0335896321476413e-06,
+      "loss": 0.3313,
+      "step": 954
+    },
+    {
+      "epoch": 1.7202543142597637,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0197242423938447e-06,
+      "loss": 0.346,
+      "step": 955
+    },
+    {
+      "epoch": 1.7220708446866486,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.0059474843362893e-06,
+      "loss": 0.3416,
+      "step": 956
+    },
+    {
+      "epoch": 1.723887375113533,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.922594939460195e-07,
+      "loss": 0.3455,
+      "step": 957
+    },
+    {
+      "epoch": 1.7257039055404177,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.786604063179728e-07,
+      "loss": 0.3378,
+      "step": 958
+    },
+    {
+      "epoch": 1.7275204359673024,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.651503556696519e-07,
+      "loss": 0.3271,
+      "step": 959
+    },
+    {
+      "epoch": 1.729336966394187,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.517294753398066e-07,
+      "loss": 0.3473,
+      "step": 960
+    },
+    {
+      "epoch": 1.7311534968210718,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.383978977871022e-07,
+      "loss": 0.3402,
+      "step": 961
+    },
+    {
+      "epoch": 1.7329700272479565,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.251557545888312e-07,
+      "loss": 0.3387,
+      "step": 962
+    },
+    {
+      "epoch": 1.7347865576748411,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.120031764395987e-07,
+      "loss": 0.3401,
+      "step": 963
+    },
+    {
+      "epoch": 1.7366030881017256,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.989402931500434e-07,
+      "loss": 0.3368,
+      "step": 964
+    },
+    {
+      "epoch": 1.7384196185286105,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.859672336455471e-07,
+      "loss": 0.34,
+      "step": 965
+    },
+    {
+      "epoch": 1.740236148955495,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.730841259649725e-07,
+      "loss": 0.331,
+      "step": 966
+    },
+    {
+      "epoch": 1.7420526793823796,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.602910972593892e-07,
+      "loss": 0.3439,
+      "step": 967
+    },
+    {
+      "epoch": 1.7438692098092643,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.475882737908248e-07,
+      "loss": 0.3418,
+      "step": 968
+    },
+    {
+      "epoch": 1.745685740236149,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.349757809310211e-07,
+      "loss": 0.3327,
+      "step": 969
+    },
+    {
+      "epoch": 1.7475022706630337,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.224537431601886e-07,
+      "loss": 0.3323,
+      "step": 970
+    },
+    {
+      "epoch": 1.749318801089918,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.100222840657879e-07,
+      "loss": 0.3308,
+      "step": 971
+    },
+    {
+      "epoch": 1.751135331516803,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.976815263412963e-07,
+      "loss": 0.3357,
+      "step": 972
+    },
+    {
+      "epoch": 1.7529518619436875,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.854315917850163e-07,
+      "loss": 0.3355,
+      "step": 973
+    },
+    {
+      "epoch": 1.7547683923705724,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.732726012988512e-07,
+      "loss": 0.3402,
+      "step": 974
+    },
+    {
+      "epoch": 1.7565849227974568,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 0.3361,
+      "step": 975
+    },
+    {
+      "epoch": 1.7584014532243415,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.492279316554207e-07,
+      "loss": 0.3162,
+      "step": 976
+    },
+    {
+      "epoch": 1.7602179836512262,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.373424898093339e-07,
+      "loss": 0.3439,
+      "step": 977
+    },
+    {
+      "epoch": 1.7620345140781108,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.255484666533874e-07,
+      "loss": 0.3486,
+      "step": 978
+    },
+    {
+      "epoch": 1.7638510445049955,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.138459785898266e-07,
+      "loss": 0.3391,
+      "step": 979
+    },
+    {
+      "epoch": 1.76566757493188,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.022351411174866e-07,
+      "loss": 0.3232,
+      "step": 980
+    },
+    {
+      "epoch": 1.7674841053587649,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.907160688306425e-07,
+      "loss": 0.3404,
+      "step": 981
+    },
+    {
+      "epoch": 1.7693006357856493,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.792888754178906e-07,
+      "loss": 0.3346,
+      "step": 982
+    },
+    {
+      "epoch": 1.771117166212534,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.679536736610137e-07,
+      "loss": 0.3356,
+      "step": 983
+    },
+    {
+      "epoch": 1.7729336966394187,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.567105754338798e-07,
+      "loss": 0.3357,
+      "step": 984
+    },
+    {
+      "epoch": 1.7747502270663034,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.455596917013274e-07,
+      "loss": 0.3487,
+      "step": 985
+    },
+    {
+      "epoch": 1.776566757493188,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.345011325180772e-07,
+      "loss": 0.3414,
+      "step": 986
+    },
+    {
+      "epoch": 1.7783832879200725,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.235350070276447e-07,
+      "loss": 0.3531,
+      "step": 987
+    },
+    {
+      "epoch": 1.7801998183469574,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.126614234612593e-07,
+      "loss": 0.3411,
+      "step": 988
+    },
+    {
+      "epoch": 1.7820163487738419,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.018804891368035e-07,
+      "loss": 0.348,
+      "step": 989
+    },
+    {
+      "epoch": 1.7838328792007268,
+      "grad_norm": 0.84375,
+      "learning_rate": 5.911923104577455e-07,
+      "loss": 0.3399,
+      "step": 990
+    },
+    {
+      "epoch": 1.7856494096276112,
+      "grad_norm": 0.83984375,
+      "learning_rate": 5.805969929120947e-07,
+      "loss": 0.3364,
+      "step": 991
+    },
+    {
+      "epoch": 1.7874659400544959,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.700946410713548e-07,
+      "loss": 0.3374,
+      "step": 992
+    },
+    {
+      "epoch": 1.7892824704813806,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.596853585895034e-07,
+      "loss": 0.3366,
+      "step": 993
+    },
+    {
+      "epoch": 1.7910990009082652,
+      "grad_norm": 0.8203125,
+      "learning_rate": 5.49369248201953e-07,
+      "loss": 0.3326,
+      "step": 994
+    },
+    {
+      "epoch": 1.79291553133515,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.391464117245471e-07,
+      "loss": 0.3353,
+      "step": 995
+    },
+    {
+      "epoch": 1.7947320617620344,
+      "grad_norm": 0.83984375,
+      "learning_rate": 5.290169500525577e-07,
+      "loss": 0.3411,
+      "step": 996
+    },
+    {
+      "epoch": 1.7965485921889193,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.189809631596798e-07,
+      "loss": 0.3382,
+      "step": 997
+    },
+    {
+      "epoch": 1.7983651226158037,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.090385500970551e-07,
+      "loss": 0.3354,
+      "step": 998
+    },
+    {
+      "epoch": 1.8001816530426886,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.99189808992282e-07,
+      "loss": 0.339,
+      "step": 999
+    },
+    {
+      "epoch": 1.801998183469573,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.335,
+      "step": 1000
+    },
+    {
+      "epoch": 1.8038147138964578,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.797737305432337e-07,
+      "loss": 0.3448,
+      "step": 1001
+    },
+    {
+      "epoch": 1.8056312443233424,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.702065848278126e-07,
+      "loss": 0.3465,
+      "step": 1002
+    },
+    {
+      "epoch": 1.807447774750227,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.6073349432606554e-07,
+      "loss": 0.3416,
+      "step": 1003
+    },
+    {
+      "epoch": 1.8092643051771118,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.5135455253357053e-07,
+      "loss": 0.3414,
+      "step": 1004
+    },
+    {
+      "epoch": 1.8110808356039962,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.420698520166988e-07,
+      "loss": 0.3592,
+      "step": 1005
+    },
+    {
+      "epoch": 1.8128973660308811,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.3287948441169457e-07,
+      "loss": 0.3369,
+      "step": 1006
+    },
+    {
+      "epoch": 1.8147138964577656,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.2378354042377776e-07,
+      "loss": 0.3476,
+      "step": 1007
+    },
+    {
+      "epoch": 1.8165304268846503,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.1478210982624055e-07,
+      "loss": 0.3308,
+      "step": 1008
+    },
+    {
+      "epoch": 1.818346957311535,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.0587528145957235e-07,
+      "loss": 0.3428,
+      "step": 1009
+    },
+    {
+      "epoch": 1.8201634877384196,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.9706314323056936e-07,
+      "loss": 0.3389,
+      "step": 1010
+    },
+    {
+      "epoch": 1.8219800181653043,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.883457821114811e-07,
+      "loss": 0.3419,
+      "step": 1011
+    },
+    {
+      "epoch": 1.8237965485921888,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.7972328413914074e-07,
+      "loss": 0.3424,
+      "step": 1012
+    },
+    {
+      "epoch": 1.8256130790190737,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.711957344141237e-07,
+      "loss": 0.339,
+      "step": 1013
+    },
+    {
+      "epoch": 1.8274296094459581,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.627632170999029e-07,
+      "loss": 0.3358,
+      "step": 1014
+    },
+    {
+      "epoch": 1.829246139872843,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.544258154220193e-07,
+      "loss": 0.3537,
+      "step": 1015
+    },
+    {
+      "epoch": 1.8310626702997275,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.4618361166726123e-07,
+      "loss": 0.3359,
+      "step": 1016
+    },
+    {
+      "epoch": 1.8328792007266121,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.380366871828522e-07,
+      "loss": 0.3417,
+      "step": 1017
+    },
+    {
+      "epoch": 1.8346957311534968,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.2998512237565005e-07,
+      "loss": 0.3258,
+      "step": 1018
+    },
+    {
+      "epoch": 1.8365122615803815,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.2202899671134546e-07,
+      "loss": 0.3375,
+      "step": 1019
+    },
+    {
+      "epoch": 1.8383287920072662,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.1416838871368925e-07,
+      "loss": 0.3391,
+      "step": 1020
+    },
+    {
+      "epoch": 1.8401453224341506,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.064033759637064e-07,
+      "loss": 0.3351,
+      "step": 1021
+    },
+    {
+      "epoch": 1.8419618528610355,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.987340350989421e-07,
+      "loss": 0.3357,
+      "step": 1022
+    },
+    {
+      "epoch": 1.84377838328792,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.911604418126901e-07,
+      "loss": 0.3276,
+      "step": 1023
+    },
+    {
+      "epoch": 1.8455949137148049,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.836826708532603e-07,
+      "loss": 0.3315,
+      "step": 1024
+    },
+    {
+      "epoch": 1.8474114441416893,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.7630079602323447e-07,
+      "loss": 0.3437,
+      "step": 1025
+    },
+    {
+      "epoch": 1.849227974568574,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.6901489017873375e-07,
+      "loss": 0.3405,
+      "step": 1026
+    },
+    {
+      "epoch": 1.8510445049954587,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.6182502522871135e-07,
+      "loss": 0.3362,
+      "step": 1027
+    },
+    {
+      "epoch": 1.8528610354223434,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.547312721342277e-07,
+      "loss": 0.3458,
+      "step": 1028
+    },
+    {
+      "epoch": 1.854677565849228,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.4773370090776625e-07,
+      "loss": 0.3307,
+      "step": 1029
+    },
+    {
+      "epoch": 1.8564940962761125,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.4083238061252565e-07,
+      "loss": 0.3316,
+      "step": 1030
+    },
+    {
+      "epoch": 1.8583106267029974,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.3402737936175423e-07,
+      "loss": 0.337,
+      "step": 1031
+    },
+    {
+      "epoch": 1.8601271571298819,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.273187643180652e-07,
+      "loss": 0.3435,
+      "step": 1032
+    },
+    {
+      "epoch": 1.8619436875567665,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.2070660169278168e-07,
+      "loss": 0.342,
+      "step": 1033
+    },
+    {
+      "epoch": 1.8637602179836512,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.1419095674527934e-07,
+      "loss": 0.3315,
+      "step": 1034
+    },
+    {
+      "epoch": 1.865576748410536,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.077718937823414e-07,
+      "loss": 0.3403,
+      "step": 1035
+    },
+    {
+      "epoch": 1.8673932788374206,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.014494761575314e-07,
+      "loss": 0.3413,
+      "step": 1036
+    },
+    {
+      "epoch": 1.869209809264305,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9522376627055585e-07,
+      "loss": 0.3406,
+      "step": 1037
+    },
+    {
+      "epoch": 1.87102633969119,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.8909482556666026e-07,
+      "loss": 0.3497,
+      "step": 1038
+    },
+    {
+      "epoch": 1.8728428701180744,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.8306271453601198e-07,
+      "loss": 0.335,
+      "step": 1039
+    },
+    {
+      "epoch": 1.8746594005449593,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.7712749271311392e-07,
+      "loss": 0.3393,
+      "step": 1040
+    },
+    {
+      "epoch": 1.8764759309718437,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.7128921867620828e-07,
+      "loss": 0.3358,
+      "step": 1041
+    },
+    {
+      "epoch": 1.8782924613987284,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6554795004670389e-07,
+      "loss": 0.3496,
+      "step": 1042
+    },
+    {
+      "epoch": 1.880108991825613,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5990374348860304e-07,
+      "loss": 0.3355,
+      "step": 1043
+    },
+    {
+      "epoch": 1.8819255222524978,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.543566547079467e-07,
+      "loss": 0.3359,
+      "step": 1044
+    },
+    {
+      "epoch": 1.8837420526793824,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4890673845226133e-07,
+      "loss": 0.3332,
+      "step": 1045
+    },
+    {
+      "epoch": 1.885558583106267,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4355404851001953e-07,
+      "loss": 0.3322,
+      "step": 1046
+    },
+    {
+      "epoch": 1.8873751135331518,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.3829863771011253e-07,
+      "loss": 0.3518,
+      "step": 1047
+    },
+    {
+      "epoch": 1.8891916439600362,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3314055792131964e-07,
+      "loss": 0.3448,
+      "step": 1048
+    },
+    {
+      "epoch": 1.891008174386921,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.280798600518085e-07,
+      "loss": 0.3416,
+      "step": 1049
+    },
+    {
+      "epoch": 1.8928247048138056,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.231165940486234e-07,
+      "loss": 0.3347,
+      "step": 1050
+    },
+    {
+      "epoch": 1.8946412352406903,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1825080889719565e-07,
+      "loss": 0.3415,
+      "step": 1051
+    },
+    {
+      "epoch": 1.896457765667575,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.134825526208605e-07,
+      "loss": 0.3393,
+      "step": 1052
+    },
+    {
+      "epoch": 1.8982742960944596,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.0881187228038214e-07,
+      "loss": 0.3379,
+      "step": 1053
+    },
+    {
+      "epoch": 1.9000908265213443,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0423881397349067e-07,
+      "loss": 0.3417,
+      "step": 1054
+    },
+    {
+      "epoch": 1.9019073569482288,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.976342283442464e-08,
+      "loss": 0.3306,
+      "step": 1055
+    },
+    {
+      "epoch": 1.9037238873751137,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.538574303348813e-08,
+      "loss": 0.3326,
+      "step": 1056
+    },
+    {
+      "epoch": 1.9055404178019981,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.110581777661331e-08,
+      "loss": 0.3413,
+      "step": 1057
+    },
+    {
+      "epoch": 1.9073569482288828,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.692368930493522e-08,
+      "loss": 0.3436,
+      "step": 1058
+    },
+    {
+      "epoch": 1.9091734786557675,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.283939889437209e-08,
+      "loss": 0.3417,
+      "step": 1059
+    },
+    {
+      "epoch": 1.9109900090826522,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.885298685522235e-08,
+      "loss": 0.3512,
+      "step": 1060
+    },
+    {
+      "epoch": 1.9128065395095368,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.496449253176274e-08,
+      "loss": 0.3308,
+      "step": 1061
+    },
+    {
+      "epoch": 1.9146230699364213,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.117395430186414e-08,
+      "loss": 0.3314,
+      "step": 1062
+    },
+    {
+      "epoch": 1.9164396003633062,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.748140957660632e-08,
+      "loss": 0.3381,
+      "step": 1063
+    },
+    {
+      "epoch": 1.9182561307901906,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.388689479991606e-08,
+      "loss": 0.3856,
+      "step": 1064
+    },
+    {
+      "epoch": 1.9200726612170755,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.039044544820404e-08,
+      "loss": 0.3441,
+      "step": 1065
+    },
+    {
+      "epoch": 1.92188919164396,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.699209603001077e-08,
+      "loss": 0.3349,
+      "step": 1066
+    },
+    {
+      "epoch": 1.9237057220708447,
+      "grad_norm": 0.86328125,
+      "learning_rate": 5.369188008567672e-08,
+      "loss": 0.3335,
+      "step": 1067
+    },
+    {
+      "epoch": 1.9255222524977293,
+      "grad_norm": 0.8203125,
+      "learning_rate": 5.048983018699827e-08,
+      "loss": 0.3462,
+      "step": 1068
+    },
+    {
+      "epoch": 1.927338782924614,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.7385977936916796e-08,
+      "loss": 0.3315,
+      "step": 1069
+    },
+    {
+      "epoch": 1.9291553133514987,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.438035396920004e-08,
+      "loss": 0.337,
+      "step": 1070
+    },
+    {
+      "epoch": 1.9309718437783832,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.147298794814347e-08,
+      "loss": 0.3505,
+      "step": 1071
+    },
+    {
+      "epoch": 1.932788374205268,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.866390856827495e-08,
+      "loss": 0.3388,
+      "step": 1072
+    },
+    {
+      "epoch": 1.9346049046321525,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.595314355407609e-08,
+      "loss": 0.3372,
+      "step": 1073
+    },
+    {
+      "epoch": 1.9364214350590372,
+      "grad_norm": 1.6796875,
+      "learning_rate": 3.3340719659701315e-08,
+      "loss": 0.3567,
+      "step": 1074
+    },
+    {
+      "epoch": 1.9382379654859219,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.082666266872036e-08,
+      "loss": 0.347,
+      "step": 1075
+    },
+    {
+      "epoch": 1.9400544959128065,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.8410997393860663e-08,
+      "loss": 0.3401,
+      "step": 1076
+    },
+    {
+      "epoch": 1.9418710263396912,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.6093747676763093e-08,
+      "loss": 0.3449,
+      "step": 1077
+    },
+    {
+      "epoch": 1.9436875567665757,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.3874936387747738e-08,
+      "loss": 0.34,
+      "step": 1078
+    },
+    {
+      "epoch": 1.9455040871934606,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.175458542558517e-08,
+      "loss": 0.3355,
+      "step": 1079
+    },
+    {
+      "epoch": 1.947320617620345,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.973271571728441e-08,
+      "loss": 0.3348,
+      "step": 1080
+    },
+    {
+      "epoch": 1.94913714804723,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.7809347217881966e-08,
+      "loss": 0.3439,
+      "step": 1081
+    },
+    {
+      "epoch": 1.9509536784741144,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5984498910249778e-08,
+      "loss": 0.3394,
+      "step": 1082
+    },
+    {
+      "epoch": 1.952770208900999,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.425818880490315e-08,
+      "loss": 0.3359,
+      "step": 1083
+    },
+    {
+      "epoch": 1.9545867393278837,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2630433939825326e-08,
+      "loss": 0.338,
+      "step": 1084
+    },
+    {
+      "epoch": 1.9564032697547684,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1101250380300965e-08,
+      "loss": 0.3383,
+      "step": 1085
+    },
+    {
+      "epoch": 1.958219800181653,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.670653218752935e-09,
+      "loss": 0.3502,
+      "step": 1086
+    },
+    {
+      "epoch": 1.9600363306085375,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.33865657459909e-09,
+      "loss": 0.3376,
+      "step": 1087
+    },
+    {
+      "epoch": 1.9618528610354224,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.105273594107953e-09,
+      "loss": 0.3473,
+      "step": 1088
+    },
+    {
+      "epoch": 1.963669391462307,
+      "grad_norm": 0.8203125,
+      "learning_rate": 5.970516450271025e-09,
+      "loss": 0.3386,
+      "step": 1089
+    },
+    {
+      "epoch": 1.9654859218891918,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.9343963426840006e-09,
+      "loss": 0.3425,
+      "step": 1090
+    },
+    {
+      "epoch": 1.9673024523160763,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.996923497434635e-09,
+      "loss": 0.3356,
+      "step": 1091
+    },
+    {
+      "epoch": 1.969118982742961,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.1581071670006013e-09,
+      "loss": 0.3389,
+      "step": 1092
+    },
+    {
+      "epoch": 1.9709355131698456,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.417955630159563e-09,
+      "loss": 0.348,
+      "step": 1093
+    },
+    {
+      "epoch": 1.9727520435967303,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.776476191910348e-09,
+      "loss": 0.331,
+      "step": 1094
+    },
+    {
+      "epoch": 1.974568574023615,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.233675183394123e-09,
+      "loss": 0.3397,
+      "step": 1095
+    },
+    {
+      "epoch": 1.9763851044504994,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.895579618388827e-10,
+      "loss": 0.3351,
+      "step": 1096
+    },
+    {
+      "epoch": 1.9782016348773843,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.441289105017177e-10,
+      "loss": 0.333,
+      "step": 1097
+    },
+    {
+      "epoch": 1.9800181653042688,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9739143862884668e-10,
+      "loss": 0.3517,
+      "step": 1098
+    },
+    {
+      "epoch": 1.9818346957311535,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.934798141786878e-11,
+      "loss": 0.3322,
+      "step": 1099
+    },
+    {
+      "epoch": 1.9836512261580381,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0,
+      "loss": 0.343,
+      "step": 1100
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1100,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 550,
+  "total_flos": 1.2984648423230669e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}