diff --git "a/trainer_state.json" "b/trainer_state.json"
deleted file mode 100644--- "a/trainer_state.json"
+++ /dev/null
@@ -1,14419 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 1.99768734531586,
-  "eval_steps": 500,
-  "global_step": 2054,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.000973393900064893,
-      "grad_norm": 2.84375,
-      "learning_rate": 5.000000000000001e-07,
-      "loss": 1.585,
-      "step": 1
-    },
-    {
-      "epoch": 0.001946787800129786,
-      "grad_norm": 2.84375,
-      "learning_rate": 1.0000000000000002e-06,
-      "loss": 1.6013,
-      "step": 2
-    },
-    {
-      "epoch": 0.0029201817001946787,
-      "grad_norm": 2.6875,
-      "learning_rate": 1.5e-06,
-      "loss": 1.5481,
-      "step": 3
-    },
-    {
-      "epoch": 0.003893575600259572,
-      "grad_norm": 2.625,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 1.5626,
-      "step": 4
-    },
-    {
-      "epoch": 0.004866969500324465,
-      "grad_norm": 2.5,
-      "learning_rate": 2.5e-06,
-      "loss": 1.5683,
-      "step": 5
-    },
-    {
-      "epoch": 0.005840363400389357,
-      "grad_norm": 2.21875,
-      "learning_rate": 3e-06,
-      "loss": 1.5582,
-      "step": 6
-    },
-    {
-      "epoch": 0.00681375730045425,
-      "grad_norm": 2.0625,
-      "learning_rate": 3.5e-06,
-      "loss": 1.5699,
-      "step": 7
-    },
-    {
-      "epoch": 0.007787151200519144,
-      "grad_norm": 1.703125,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 1.543,
-      "step": 8
-    },
-    {
-      "epoch": 0.008760545100584036,
-      "grad_norm": 1.5546875,
-      "learning_rate": 4.5e-06,
-      "loss": 1.5312,
-      "step": 9
-    },
-    {
-      "epoch": 0.00973393900064893,
-      "grad_norm": 1.3359375,
-      "learning_rate": 5e-06,
-      "loss": 1.563,
-      "step": 10
-    },
-    {
-      "epoch": 0.010707332900713823,
-      "grad_norm": 1.1796875,
-      "learning_rate": 4.999997047105779e-06,
-      "loss": 1.5218,
-      "step": 11
-    },
-    {
-      "epoch": 0.011680726800778715,
-      "grad_norm": 1.140625,
-      "learning_rate": 4.99998818843009e-06,
-      "loss": 1.5607,
-      "step": 12
-    },
-    {
-      "epoch": 0.012654120700843608,
-      "grad_norm": 1.0546875,
-      "learning_rate": 4.9999734239938604e-06,
-      "loss": 1.508,
-      "step": 13
-    },
-    {
-      "epoch": 0.0136275146009085,
-      "grad_norm": 0.98046875,
-      "learning_rate": 4.999952753831968e-06,
-      "loss": 1.4982,
-      "step": 14
-    },
-    {
-      "epoch": 0.014600908500973394,
-      "grad_norm": 0.91015625,
-      "learning_rate": 4.999926177993243e-06,
-      "loss": 1.5067,
-      "step": 15
-    },
-    {
-      "epoch": 0.015574302401038288,
-      "grad_norm": 0.83203125,
-      "learning_rate": 4.999893696540466e-06,
-      "loss": 1.5303,
-      "step": 16
-    },
-    {
-      "epoch": 0.01654769630110318,
-      "grad_norm": 0.77734375,
-      "learning_rate": 4.999855309550366e-06,
-      "loss": 1.5187,
-      "step": 17
-    },
-    {
-      "epoch": 0.01752109020116807,
-      "grad_norm": 0.75390625,
-      "learning_rate": 4.99981101711363e-06,
-      "loss": 1.5402,
-      "step": 18
-    },
-    {
-      "epoch": 0.018494484101232965,
-      "grad_norm": 0.703125,
-      "learning_rate": 4.999760819334887e-06,
-      "loss": 1.4852,
-      "step": 19
-    },
-    {
-      "epoch": 0.01946787800129786,
-      "grad_norm": 0.70703125,
-      "learning_rate": 4.999704716332721e-06,
-      "loss": 1.4951,
-      "step": 20
-    },
-    {
-      "epoch": 0.020441271901362752,
-      "grad_norm": 0.67578125,
-      "learning_rate": 4.999642708239664e-06,
-      "loss": 1.4866,
-      "step": 21
-    },
-    {
-      "epoch": 0.021414665801427646,
-      "grad_norm": 0.66015625,
-      "learning_rate": 4.9995747952022e-06,
-      "loss": 1.4905,
-      "step": 22
-    },
-    {
-      "epoch": 0.022388059701492536,
-      "grad_norm": 0.6328125,
-      "learning_rate": 4.9995009773807614e-06,
-      "loss": 1.4776,
-      "step": 23
-    },
-    {
-      "epoch": 0.02336145360155743,
-      "grad_norm": 0.59765625,
-      "learning_rate": 4.999421254949728e-06,
-      "loss": 1.487,
-      "step": 24
-    },
-    {
-      "epoch": 0.024334847501622323,
-      "grad_norm": 0.5859375,
-      "learning_rate": 4.999335628097429e-06,
-      "loss": 1.4688,
-      "step": 25
-    },
-    {
-      "epoch": 0.025308241401687217,
-      "grad_norm": 0.59375,
-      "learning_rate": 4.999244097026143e-06,
-      "loss": 1.4589,
-      "step": 26
-    },
-    {
-      "epoch": 0.02628163530175211,
-      "grad_norm": 0.57421875,
-      "learning_rate": 4.999146661952095e-06,
-      "loss": 1.4611,
-      "step": 27
-    },
-    {
-      "epoch": 0.027255029201817,
-      "grad_norm": 0.55859375,
-      "learning_rate": 4.999043323105458e-06,
-      "loss": 1.4766,
-      "step": 28
-    },
-    {
-      "epoch": 0.028228423101881894,
-      "grad_norm": 0.546875,
-      "learning_rate": 4.9989340807303496e-06,
-      "loss": 1.4617,
-      "step": 29
-    },
-    {
-      "epoch": 0.029201817001946788,
-      "grad_norm": 0.53125,
-      "learning_rate": 4.998818935084836e-06,
-      "loss": 1.427,
-      "step": 30
-    },
-    {
-      "epoch": 0.03017521090201168,
-      "grad_norm": 0.53515625,
-      "learning_rate": 4.998697886440927e-06,
-      "loss": 1.477,
-      "step": 31
-    },
-    {
-      "epoch": 0.031148604802076575,
-      "grad_norm": 0.53125,
-      "learning_rate": 4.998570935084577e-06,
-      "loss": 1.4606,
-      "step": 32
-    },
-    {
-      "epoch": 0.03212199870214147,
-      "grad_norm": 0.515625,
-      "learning_rate": 4.998438081315686e-06,
-      "loss": 1.4544,
-      "step": 33
-    },
-    {
-      "epoch": 0.03309539260220636,
-      "grad_norm": 0.51171875,
-      "learning_rate": 4.998299325448096e-06,
-      "loss": 1.4324,
-      "step": 34
-    },
-    {
-      "epoch": 0.03406878650227125,
-      "grad_norm": 0.5078125,
-      "learning_rate": 4.998154667809593e-06,
-      "loss": 1.4709,
-      "step": 35
-    },
-    {
-      "epoch": 0.03504218040233614,
-      "grad_norm": 0.494140625,
-      "learning_rate": 4.998004108741903e-06,
-      "loss": 1.4562,
-      "step": 36
-    },
-    {
-      "epoch": 0.036015574302401036,
-      "grad_norm": 0.494140625,
-      "learning_rate": 4.9978476486006945e-06,
-      "loss": 1.457,
-      "step": 37
-    },
-    {
-      "epoch": 0.03698896820246593,
-      "grad_norm": 0.478515625,
-      "learning_rate": 4.9976852877555755e-06,
-      "loss": 1.4127,
-      "step": 38
-    },
-    {
-      "epoch": 0.037962362102530824,
-      "grad_norm": 0.4765625,
-      "learning_rate": 4.997517026590094e-06,
-      "loss": 1.4591,
-      "step": 39
-    },
-    {
-      "epoch": 0.03893575600259572,
-      "grad_norm": 0.470703125,
-      "learning_rate": 4.997342865501734e-06,
-      "loss": 1.4168,
-      "step": 40
-    },
-    {
-      "epoch": 0.03990914990266061,
-      "grad_norm": 0.47265625,
-      "learning_rate": 4.997162804901922e-06,
-      "loss": 1.4558,
-      "step": 41
-    },
-    {
-      "epoch": 0.040882543802725504,
-      "grad_norm": 0.462890625,
-      "learning_rate": 4.996976845216016e-06,
-      "loss": 1.4498,
-      "step": 42
-    },
-    {
-      "epoch": 0.0418559377027904,
-      "grad_norm": 0.453125,
-      "learning_rate": 4.996784986883312e-06,
-      "loss": 1.4169,
-      "step": 43
-    },
-    {
-      "epoch": 0.04282933160285529,
-      "grad_norm": 0.45703125,
-      "learning_rate": 4.99658723035704e-06,
-      "loss": 1.436,
-      "step": 44
-    },
-    {
-      "epoch": 0.04380272550292018,
-      "grad_norm": 0.455078125,
-      "learning_rate": 4.996383576104362e-06,
-      "loss": 1.4431,
-      "step": 45
-    },
-    {
-      "epoch": 0.04477611940298507,
-      "grad_norm": 0.453125,
-      "learning_rate": 4.996174024606375e-06,
-      "loss": 1.4005,
-      "step": 46
-    },
-    {
-      "epoch": 0.045749513303049966,
-      "grad_norm": 0.447265625,
-      "learning_rate": 4.995958576358105e-06,
-      "loss": 1.4259,
-      "step": 47
-    },
-    {
-      "epoch": 0.04672290720311486,
-      "grad_norm": 0.447265625,
-      "learning_rate": 4.995737231868509e-06,
-      "loss": 1.4245,
-      "step": 48
-    },
-    {
-      "epoch": 0.04769630110317975,
-      "grad_norm": 0.439453125,
-      "learning_rate": 4.9955099916604725e-06,
-      "loss": 1.4237,
-      "step": 49
-    },
-    {
-      "epoch": 0.04866969500324465,
-      "grad_norm": 0.4453125,
-      "learning_rate": 4.9952768562708085e-06,
-      "loss": 1.4092,
-      "step": 50
-    },
-    {
-      "epoch": 0.04964308890330954,
-      "grad_norm": 0.443359375,
-      "learning_rate": 4.995037826250257e-06,
-      "loss": 1.4246,
-      "step": 51
-    },
-    {
-      "epoch": 0.050616482803374434,
-      "grad_norm": 0.43359375,
-      "learning_rate": 4.9947929021634815e-06,
-      "loss": 1.4332,
-      "step": 52
-    },
-    {
-      "epoch": 0.05158987670343933,
-      "grad_norm": 0.4296875,
-      "learning_rate": 4.994542084589069e-06,
-      "loss": 1.4002,
-      "step": 53
-    },
-    {
-      "epoch": 0.05256327060350422,
-      "grad_norm": 0.427734375,
-      "learning_rate": 4.994285374119531e-06,
-      "loss": 1.4039,
-      "step": 54
-    },
-    {
-      "epoch": 0.05353666450356911,
-      "grad_norm": 0.4296875,
-      "learning_rate": 4.994022771361299e-06,
-      "loss": 1.3983,
-      "step": 55
-    },
-    {
-      "epoch": 0.054510058403634,
-      "grad_norm": 0.421875,
-      "learning_rate": 4.993754276934722e-06,
-      "loss": 1.4061,
-      "step": 56
-    },
-    {
-      "epoch": 0.055483452303698895,
-      "grad_norm": 0.42578125,
-      "learning_rate": 4.9934798914740705e-06,
-      "loss": 1.3991,
-      "step": 57
-    },
-    {
-      "epoch": 0.05645684620376379,
-      "grad_norm": 0.419921875,
-      "learning_rate": 4.9931996156275285e-06,
-      "loss": 1.3873,
-      "step": 58
-    },
-    {
-      "epoch": 0.05743024010382868,
-      "grad_norm": 0.41796875,
-      "learning_rate": 4.9929134500571954e-06,
-      "loss": 1.4224,
-      "step": 59
-    },
-    {
-      "epoch": 0.058403634003893576,
-      "grad_norm": 0.4140625,
-      "learning_rate": 4.992621395439084e-06,
-      "loss": 1.3786,
-      "step": 60
-    },
-    {
-      "epoch": 0.05937702790395847,
-      "grad_norm": 0.4140625,
-      "learning_rate": 4.992323452463122e-06,
-      "loss": 1.4181,
-      "step": 61
-    },
-    {
-      "epoch": 0.06035042180402336,
-      "grad_norm": 0.41015625,
-      "learning_rate": 4.992019621833142e-06,
-      "loss": 1.398,
-      "step": 62
-    },
-    {
-      "epoch": 0.06132381570408826,
-      "grad_norm": 0.41015625,
-      "learning_rate": 4.9917099042668886e-06,
-      "loss": 1.4016,
-      "step": 63
-    },
-    {
-      "epoch": 0.06229720960415315,
-      "grad_norm": 0.4140625,
-      "learning_rate": 4.9913943004960134e-06,
-      "loss": 1.392,
-      "step": 64
-    },
-    {
-      "epoch": 0.06327060350421804,
-      "grad_norm": 0.404296875,
-      "learning_rate": 4.99107281126607e-06,
-      "loss": 1.3946,
-      "step": 65
-    },
-    {
-      "epoch": 0.06424399740428294,
-      "grad_norm": 0.40625,
-      "learning_rate": 4.99074543733652e-06,
-      "loss": 1.3838,
-      "step": 66
-    },
-    {
-      "epoch": 0.06521739130434782,
-      "grad_norm": 0.416015625,
-      "learning_rate": 4.990412179480722e-06,
-      "loss": 1.4,
-      "step": 67
-    },
-    {
-      "epoch": 0.06619078520441272,
-      "grad_norm": 0.412109375,
-      "learning_rate": 4.9900730384859366e-06,
-      "loss": 1.3964,
-      "step": 68
-    },
-    {
-      "epoch": 0.06716417910447761,
-      "grad_norm": 0.404296875,
-      "learning_rate": 4.989728015153321e-06,
-      "loss": 1.4166,
-      "step": 69
-    },
-    {
-      "epoch": 0.0681375730045425,
-      "grad_norm": 0.400390625,
-      "learning_rate": 4.9893771102979305e-06,
-      "loss": 1.3839,
-      "step": 70
-    },
-    {
-      "epoch": 0.0691109669046074,
-      "grad_norm": 0.3984375,
-      "learning_rate": 4.989020324748713e-06,
-      "loss": 1.3821,
-      "step": 71
-    },
-    {
-      "epoch": 0.07008436080467229,
-      "grad_norm": 0.3984375,
-      "learning_rate": 4.988657659348507e-06,
-      "loss": 1.3899,
-      "step": 72
-    },
-    {
-      "epoch": 0.07105775470473719,
-      "grad_norm": 0.3984375,
-      "learning_rate": 4.988289114954045e-06,
-      "loss": 1.3865,
-      "step": 73
-    },
-    {
-      "epoch": 0.07203114860480207,
-      "grad_norm": 0.3984375,
-      "learning_rate": 4.987914692435942e-06,
-      "loss": 1.3732,
-      "step": 74
-    },
-    {
-      "epoch": 0.07300454250486697,
-      "grad_norm": 0.404296875,
-      "learning_rate": 4.987534392678704e-06,
-      "loss": 1.3853,
-      "step": 75
-    },
-    {
-      "epoch": 0.07397793640493186,
-      "grad_norm": 0.3984375,
-      "learning_rate": 4.987148216580719e-06,
-      "loss": 1.3749,
-      "step": 76
-    },
-    {
-      "epoch": 0.07495133030499676,
-      "grad_norm": 0.41796875,
-      "learning_rate": 4.986756165054256e-06,
-      "loss": 1.4128,
-      "step": 77
-    },
-    {
-      "epoch": 0.07592472420506165,
-      "grad_norm": 0.39453125,
-      "learning_rate": 4.986358239025465e-06,
-      "loss": 1.3262,
-      "step": 78
-    },
-    {
-      "epoch": 0.07689811810512655,
-      "grad_norm": 0.396484375,
-      "learning_rate": 4.985954439434373e-06,
-      "loss": 1.3743,
-      "step": 79
-    },
-    {
-      "epoch": 0.07787151200519143,
-      "grad_norm": 0.396484375,
-      "learning_rate": 4.98554476723488e-06,
-      "loss": 1.3854,
-      "step": 80
-    },
-    {
-      "epoch": 0.07884490590525632,
-      "grad_norm": 0.404296875,
-      "learning_rate": 4.985129223394764e-06,
-      "loss": 1.374,
-      "step": 81
-    },
-    {
-      "epoch": 0.07981829980532122,
-      "grad_norm": 0.400390625,
-      "learning_rate": 4.984707808895668e-06,
-      "loss": 1.3581,
-      "step": 82
-    },
-    {
-      "epoch": 0.08079169370538611,
-      "grad_norm": 0.3984375,
-      "learning_rate": 4.984280524733107e-06,
-      "loss": 1.3607,
-      "step": 83
-    },
-    {
-      "epoch": 0.08176508760545101,
-      "grad_norm": 0.388671875,
-      "learning_rate": 4.983847371916461e-06,
-      "loss": 1.3647,
-      "step": 84
-    },
-    {
-      "epoch": 0.0827384815055159,
-      "grad_norm": 0.388671875,
-      "learning_rate": 4.983408351468973e-06,
-      "loss": 1.3678,
-      "step": 85
-    },
-    {
-      "epoch": 0.0837118754055808,
-      "grad_norm": 0.37890625,
-      "learning_rate": 4.982963464427749e-06,
-      "loss": 1.3478,
-      "step": 86
-    },
-    {
-      "epoch": 0.08468526930564568,
-      "grad_norm": 0.39453125,
-      "learning_rate": 4.982512711843753e-06,
-      "loss": 1.3811,
-      "step": 87
-    },
-    {
-      "epoch": 0.08565866320571058,
-      "grad_norm": 0.388671875,
-      "learning_rate": 4.982056094781801e-06,
-      "loss": 1.3797,
-      "step": 88
-    },
-    {
-      "epoch": 0.08663205710577547,
-      "grad_norm": 0.384765625,
-      "learning_rate": 4.98159361432057e-06,
-      "loss": 1.3658,
-      "step": 89
-    },
-    {
-      "epoch": 0.08760545100584036,
-      "grad_norm": 0.3828125,
-      "learning_rate": 4.9811252715525835e-06,
-      "loss": 1.3723,
-      "step": 90
-    },
-    {
-      "epoch": 0.08857884490590526,
-      "grad_norm": 0.3828125,
-      "learning_rate": 4.980651067584215e-06,
-      "loss": 1.3584,
-      "step": 91
-    },
-    {
-      "epoch": 0.08955223880597014,
-      "grad_norm": 0.37890625,
-      "learning_rate": 4.980171003535684e-06,
-      "loss": 1.3579,
-      "step": 92
-    },
-    {
-      "epoch": 0.09052563270603504,
-      "grad_norm": 0.388671875,
-      "learning_rate": 4.979685080541053e-06,
-      "loss": 1.3621,
-      "step": 93
-    },
-    {
-      "epoch": 0.09149902660609993,
-      "grad_norm": 0.3828125,
-      "learning_rate": 4.979193299748225e-06,
-      "loss": 1.3598,
-      "step": 94
-    },
-    {
-      "epoch": 0.09247242050616483,
-      "grad_norm": 0.37890625,
-      "learning_rate": 4.978695662318943e-06,
-      "loss": 1.3666,
-      "step": 95
-    },
-    {
-      "epoch": 0.09344581440622972,
-      "grad_norm": 0.373046875,
-      "learning_rate": 4.978192169428781e-06,
-      "loss": 1.3569,
-      "step": 96
-    },
-    {
-      "epoch": 0.09441920830629462,
-      "grad_norm": 0.38671875,
-      "learning_rate": 4.9776828222671495e-06,
-      "loss": 1.3424,
-      "step": 97
-    },
-    {
-      "epoch": 0.0953926022063595,
-      "grad_norm": 0.376953125,
-      "learning_rate": 4.977167622037287e-06,
-      "loss": 1.3429,
-      "step": 98
-    },
-    {
-      "epoch": 0.0963659961064244,
-      "grad_norm": 0.375,
-      "learning_rate": 4.9766465699562585e-06,
-      "loss": 1.3491,
-      "step": 99
-    },
-    {
-      "epoch": 0.0973393900064893,
-      "grad_norm": 0.369140625,
-      "learning_rate": 4.976119667254954e-06,
-      "loss": 1.3504,
-      "step": 100
-    },
-    {
-      "epoch": 0.09831278390655418,
-      "grad_norm": 0.384765625,
-      "learning_rate": 4.975586915178084e-06,
-      "loss": 1.345,
-      "step": 101
-    },
-    {
-      "epoch": 0.09928617780661908,
-      "grad_norm": 0.369140625,
-      "learning_rate": 4.975048314984176e-06,
-      "loss": 1.3331,
-      "step": 102
-    },
-    {
-      "epoch": 0.10025957170668397,
-      "grad_norm": 0.390625,
-      "learning_rate": 4.9745038679455735e-06,
-      "loss": 1.3661,
-      "step": 103
-    },
-    {
-      "epoch": 0.10123296560674887,
-      "grad_norm": 0.373046875,
-      "learning_rate": 4.973953575348434e-06,
-      "loss": 1.3396,
-      "step": 104
-    },
-    {
-      "epoch": 0.10220635950681375,
-      "grad_norm": 0.375,
-      "learning_rate": 4.973397438492719e-06,
-      "loss": 1.3396,
-      "step": 105
-    },
-    {
-      "epoch": 0.10317975340687865,
-      "grad_norm": 0.376953125,
-      "learning_rate": 4.972835458692202e-06,
-      "loss": 1.3498,
-      "step": 106
-    },
-    {
-      "epoch": 0.10415314730694354,
-      "grad_norm": 0.373046875,
-      "learning_rate": 4.972267637274455e-06,
-      "loss": 1.3449,
-      "step": 107
-    },
-    {
-      "epoch": 0.10512654120700844,
-      "grad_norm": 0.369140625,
-      "learning_rate": 4.971693975580851e-06,
-      "loss": 1.327,
-      "step": 108
-    },
-    {
-      "epoch": 0.10609993510707333,
-      "grad_norm": 0.3671875,
-      "learning_rate": 4.9711144749665615e-06,
-      "loss": 1.3289,
-      "step": 109
-    },
-    {
-      "epoch": 0.10707332900713822,
-      "grad_norm": 0.37109375,
-      "learning_rate": 4.9705291368005485e-06,
-      "loss": 1.3386,
-      "step": 110
-    },
-    {
-      "epoch": 0.10804672290720312,
-      "grad_norm": 0.3671875,
-      "learning_rate": 4.969937962465566e-06,
-      "loss": 1.3364,
-      "step": 111
-    },
-    {
-      "epoch": 0.109020116807268,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.9693409533581524e-06,
-      "loss": 1.3133,
-      "step": 112
-    },
-    {
-      "epoch": 0.1099935107073329,
-      "grad_norm": 0.37109375,
-      "learning_rate": 4.968738110888635e-06,
-      "loss": 1.3272,
-      "step": 113
-    },
-    {
-      "epoch": 0.11096690460739779,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.968129436481114e-06,
-      "loss": 1.3333,
-      "step": 114
-    },
-    {
-      "epoch": 0.11194029850746269,
-      "grad_norm": 0.369140625,
-      "learning_rate": 4.967514931573473e-06,
-      "loss": 1.3263,
-      "step": 115
-    },
-    {
-      "epoch": 0.11291369240752758,
-      "grad_norm": 0.359375,
-      "learning_rate": 4.966894597617365e-06,
-      "loss": 1.3243,
-      "step": 116
-    },
-    {
-      "epoch": 0.11388708630759248,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.966268436078215e-06,
-      "loss": 1.3502,
-      "step": 117
-    },
-    {
-      "epoch": 0.11486048020765736,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.965636448435214e-06,
-      "loss": 1.3209,
-      "step": 118
-    },
-    {
-      "epoch": 0.11583387410772227,
-      "grad_norm": 0.357421875,
-      "learning_rate": 4.9649986361813155e-06,
-      "loss": 1.3069,
-      "step": 119
-    },
-    {
-      "epoch": 0.11680726800778715,
-      "grad_norm": 0.369140625,
-      "learning_rate": 4.9643550008232335e-06,
-      "loss": 1.3339,
-      "step": 120
-    },
-    {
-      "epoch": 0.11778066190785204,
-      "grad_norm": 0.361328125,
-      "learning_rate": 4.9637055438814385e-06,
-      "loss": 1.3142,
-      "step": 121
-    },
-    {
-      "epoch": 0.11875405580791694,
-      "grad_norm": 0.369140625,
-      "learning_rate": 4.963050266890152e-06,
-      "loss": 1.3347,
-      "step": 122
-    },
-    {
-      "epoch": 0.11972744970798183,
-      "grad_norm": 0.3671875,
-      "learning_rate": 4.962389171397345e-06,
-      "loss": 1.2988,
-      "step": 123
-    },
-    {
-      "epoch": 0.12070084360804673,
-      "grad_norm": 0.361328125,
-      "learning_rate": 4.961722258964734e-06,
-      "loss": 1.3353,
-      "step": 124
-    },
-    {
-      "epoch": 0.12167423750811161,
-      "grad_norm": 0.357421875,
-      "learning_rate": 4.961049531167775e-06,
-      "loss": 1.3181,
-      "step": 125
-    },
-    {
-      "epoch": 0.12264763140817651,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.960370989595665e-06,
-      "loss": 1.2864,
-      "step": 126
-    },
-    {
-      "epoch": 0.1236210253082414,
-      "grad_norm": 0.37109375,
-      "learning_rate": 4.959686635851332e-06,
-      "loss": 1.32,
-      "step": 127
-    },
-    {
-      "epoch": 0.1245944192083063,
-      "grad_norm": 0.361328125,
-      "learning_rate": 4.958996471551436e-06,
-      "loss": 1.3361,
-      "step": 128
-    },
-    {
-      "epoch": 0.1255678131083712,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.958300498326363e-06,
-      "loss": 1.3165,
-      "step": 129
-    },
-    {
-      "epoch": 0.12654120700843607,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.95759871782022e-06,
-      "loss": 1.3161,
-      "step": 130
-    },
-    {
-      "epoch": 0.12751460090850097,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.956891131690835e-06,
-      "loss": 1.3125,
-      "step": 131
-    },
-    {
-      "epoch": 0.12848799480856588,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.956177741609749e-06,
-      "loss": 1.3335,
-      "step": 132
-    },
-    {
-      "epoch": 0.12946138870863075,
-      "grad_norm": 0.361328125,
-      "learning_rate": 4.955458549262216e-06,
-      "loss": 1.3505,
-      "step": 133
-    },
-    {
-      "epoch": 0.13043478260869565,
-      "grad_norm": 0.359375,
-      "learning_rate": 4.954733556347192e-06,
-      "loss": 1.3077,
-      "step": 134
-    },
-    {
-      "epoch": 0.13140817650876055,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.954002764577342e-06,
-      "loss": 1.3213,
-      "step": 135
-    },
-    {
-      "epoch": 0.13238157040882545,
-      "grad_norm": 0.357421875,
-      "learning_rate": 4.953266175679023e-06,
-      "loss": 1.3513,
-      "step": 136
-    },
-    {
-      "epoch": 0.13335496430889032,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.952523791392295e-06,
-      "loss": 1.3394,
-      "step": 137
-    },
-    {
-      "epoch": 0.13432835820895522,
-      "grad_norm": 0.359375,
-      "learning_rate": 4.9517756134709005e-06,
-      "loss": 1.317,
-      "step": 138
-    },
-    {
-      "epoch": 0.13530175210902012,
-      "grad_norm": 0.359375,
-      "learning_rate": 4.951021643682273e-06,
-      "loss": 1.3253,
-      "step": 139
-    },
-    {
-      "epoch": 0.136275146009085,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.950261883807526e-06,
-      "loss": 1.3097,
-      "step": 140
-    },
-    {
-      "epoch": 0.1372485399091499,
-      "grad_norm": 0.357421875,
-      "learning_rate": 4.949496335641453e-06,
-      "loss": 1.3284,
-      "step": 141
-    },
-    {
-      "epoch": 0.1382219338092148,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.948725000992519e-06,
-      "loss": 1.3016,
-      "step": 142
-    },
-    {
-      "epoch": 0.1391953277092797,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.947947881682861e-06,
-      "loss": 1.3289,
-      "step": 143
-    },
-    {
-      "epoch": 0.14016872160934457,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.947164979548279e-06,
-      "loss": 1.3198,
-      "step": 144
-    },
-    {
-      "epoch": 0.14114211550940947,
-      "grad_norm": 0.4609375,
-      "learning_rate": 4.9463762964382355e-06,
-      "loss": 1.328,
-      "step": 145
-    },
-    {
-      "epoch": 0.14211550940947437,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.945581834215848e-06,
-      "loss": 1.31,
-      "step": 146
-    },
-    {
-      "epoch": 0.14308890330953927,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.944781594757887e-06,
-      "loss": 1.299,
-      "step": 147
-    },
-    {
-      "epoch": 0.14406229720960415,
-      "grad_norm": 0.376953125,
-      "learning_rate": 4.9439755799547705e-06,
-      "loss": 1.3003,
-      "step": 148
-    },
-    {
-      "epoch": 0.14503569110966905,
-      "grad_norm": 0.380859375,
-      "learning_rate": 4.94316379171056e-06,
-      "loss": 1.3353,
-      "step": 149
-    },
-    {
-      "epoch": 0.14600908500973395,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.942346231942955e-06,
-      "loss": 1.3057,
-      "step": 150
-    },
-    {
-      "epoch": 0.14698247890979882,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.941522902583289e-06,
-      "loss": 1.2995,
-      "step": 151
-    },
-    {
-      "epoch": 0.14795587280986372,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.9406938055765265e-06,
-      "loss": 1.3182,
-      "step": 152
-    },
-    {
-      "epoch": 0.14892926670992862,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.939858942881256e-06,
-      "loss": 1.3123,
-      "step": 153
-    },
-    {
-      "epoch": 0.14990266060999352,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.939018316469687e-06,
-      "loss": 1.2856,
-      "step": 154
-    },
-    {
-      "epoch": 0.1508760545100584,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.938171928327643e-06,
-      "loss": 1.3181,
-      "step": 155
-    },
-    {
-      "epoch": 0.1518494484101233,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.937319780454559e-06,
-      "loss": 1.2955,
-      "step": 156
-    },
-    {
-      "epoch": 0.1528228423101882,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.9364618748634794e-06,
-      "loss": 1.2884,
-      "step": 157
-    },
-    {
-      "epoch": 0.1537962362102531,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.935598213581047e-06,
-      "loss": 1.3084,
-      "step": 158
-    },
-    {
-      "epoch": 0.15476963011031797,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.934728798647501e-06,
-      "loss": 1.2904,
-      "step": 159
-    },
-    {
-      "epoch": 0.15574302401038287,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.933853632116674e-06,
-      "loss": 1.2794,
-      "step": 160
-    },
-    {
-      "epoch": 0.15671641791044777,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.932972716055987e-06,
-      "loss": 1.2885,
-      "step": 161
-    },
-    {
-      "epoch": 0.15768981181051264,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.93208605254644e-06,
-      "loss": 1.3044,
-      "step": 162
-    },
-    {
-      "epoch": 0.15866320571057754,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.931193643682611e-06,
-      "loss": 1.2846,
-      "step": 163
-    },
-    {
-      "epoch": 0.15963659961064244,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.9302954915726535e-06,
-      "loss": 1.3036,
-      "step": 164
-    },
-    {
-      "epoch": 0.16060999351070734,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.929391598338285e-06,
-      "loss": 1.3031,
-      "step": 165
-    },
-    {
-      "epoch": 0.16158338741077222,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.928481966114785e-06,
-      "loss": 1.2814,
-      "step": 166
-    },
-    {
-      "epoch": 0.16255678131083712,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.927566597050995e-06,
-      "loss": 1.2601,
-      "step": 167
-    },
-    {
-      "epoch": 0.16353017521090202,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.926645493309302e-06,
-      "loss": 1.3052,
-      "step": 168
-    },
-    {
-      "epoch": 0.16450356911096692,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.9257186570656446e-06,
-      "loss": 1.305,
-      "step": 169
-    },
-    {
-      "epoch": 0.1654769630110318,
-      "grad_norm": 0.3671875,
-      "learning_rate": 4.924786090509504e-06,
-      "loss": 1.2964,
-      "step": 170
-    },
-    {
-      "epoch": 0.1664503569110967,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.923847795843894e-06,
-      "loss": 1.2921,
-      "step": 171
-    },
-    {
-      "epoch": 0.1674237508111616,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.9229037752853634e-06,
-      "loss": 1.2761,
-      "step": 172
-    },
-    {
-      "epoch": 0.16839714471122647,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.921954031063987e-06,
-      "loss": 1.2822,
-      "step": 173
-    },
-    {
-      "epoch": 0.16937053861129137,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.92099856542336e-06,
-      "loss": 1.2768,
-      "step": 174
-    },
-    {
-      "epoch": 0.17034393251135627,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.920037380620594e-06,
-      "loss": 1.2932,
-      "step": 175
-    },
-    {
-      "epoch": 0.17131732641142117,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.91907047892631e-06,
-      "loss": 1.3019,
-      "step": 176
-    },
-    {
-      "epoch": 0.17229072031148604,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.918097862624635e-06,
-      "loss": 1.2809,
-      "step": 177
-    },
-    {
-      "epoch": 0.17326411421155094,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.917119534013194e-06,
-      "loss": 1.2763,
-      "step": 178
-    },
-    {
-      "epoch": 0.17423750811161584,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.9161354954031105e-06,
-      "loss": 1.2856,
-      "step": 179
-    },
-    {
-      "epoch": 0.1752109020116807,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.915145749118992e-06,
-      "loss": 1.2842,
-      "step": 180
-    },
-    {
-      "epoch": 0.17618429591174561,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.914150297498932e-06,
-      "loss": 1.2908,
-      "step": 181
-    },
-    {
-      "epoch": 0.17715768981181051,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.913149142894501e-06,
-      "loss": 1.2622,
-      "step": 182
-    },
-    {
-      "epoch": 0.17813108371187542,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.912142287670742e-06,
-      "loss": 1.2957,
-      "step": 183
-    },
-    {
-      "epoch": 0.1791044776119403,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.911129734206164e-06,
-      "loss": 1.2829,
-      "step": 184
-    },
-    {
-      "epoch": 0.1800778715120052,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.91011148489274e-06,
-      "loss": 1.2799,
-      "step": 185
-    },
-    {
-      "epoch": 0.1810512654120701,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.909087542135893e-06,
-      "loss": 1.2664,
-      "step": 186
-    },
-    {
-      "epoch": 0.182024659312135,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.9080579083545e-06,
-      "loss": 1.2715,
-      "step": 187
-    },
-    {
-      "epoch": 0.18299805321219986,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.907022585980881e-06,
-      "loss": 1.2758,
-      "step": 188
-    },
-    {
-      "epoch": 0.18397144711226476,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.905981577460793e-06,
-      "loss": 1.2775,
-      "step": 189
-    },
-    {
-      "epoch": 0.18494484101232966,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.904934885253428e-06,
-      "loss": 1.2712,
-      "step": 190
-    },
-    {
-      "epoch": 0.18591823491239454,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.903882511831401e-06,
-      "loss": 1.2968,
-      "step": 191
-    },
-    {
-      "epoch": 0.18689162881245944,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.9028244596807525e-06,
-      "loss": 1.2776,
-      "step": 192
-    },
-    {
-      "epoch": 0.18786502271252434,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.901760731300933e-06,
-      "loss": 1.256,
-      "step": 193
-    },
-    {
-      "epoch": 0.18883841661258924,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.9006913292048055e-06,
-      "loss": 1.2852,
-      "step": 194
-    },
-    {
-      "epoch": 0.1898118105126541,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.899616255918636e-06,
-      "loss": 1.2798,
-      "step": 195
-    },
-    {
-      "epoch": 0.190785204412719,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.898535513982083e-06,
-      "loss": 1.2732,
-      "step": 196
-    },
-    {
-      "epoch": 0.1917585983127839,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.897449105948205e-06,
-      "loss": 1.2633,
-      "step": 197
-    },
-    {
-      "epoch": 0.1927319922128488,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.896357034383436e-06,
-      "loss": 1.2865,
-      "step": 198
-    },
-    {
-      "epoch": 0.19370538611291369,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.8952593018675955e-06,
-      "loss": 1.2864,
-      "step": 199
-    },
-    {
-      "epoch": 0.1946787800129786,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.894155910993874e-06,
-      "loss": 1.2467,
-      "step": 200
-    },
-    {
-      "epoch": 0.1956521739130435,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.893046864368828e-06,
-      "loss": 1.274,
-      "step": 201
-    },
-    {
-      "epoch": 0.19662556781310836,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.891932164612376e-06,
-      "loss": 1.2746,
-      "step": 202
-    },
-    {
-      "epoch": 0.19759896171317326,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.89081181435779e-06,
-      "loss": 1.2733,
-      "step": 203
-    },
-    {
-      "epoch": 0.19857235561323816,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.889685816251691e-06,
-      "loss": 1.2621,
-      "step": 204
-    },
-    {
-      "epoch": 0.19954574951330306,
-      "grad_norm": 0.3671875,
-      "learning_rate": 4.888554172954041e-06,
-      "loss": 1.2812,
-      "step": 205
-    },
-    {
-      "epoch": 0.20051914341336793,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.887416887138139e-06,
-      "loss": 1.286,
-      "step": 206
-    },
-    {
-      "epoch": 0.20149253731343283,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.886273961490612e-06,
-      "loss": 1.2646,
-      "step": 207
-    },
-    {
-      "epoch": 0.20246593121349774,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.885125398711411e-06,
-      "loss": 1.2767,
-      "step": 208
-    },
-    {
-      "epoch": 0.2034393251135626,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.8839712015138045e-06,
-      "loss": 1.2751,
-      "step": 209
-    },
-    {
-      "epoch": 0.2044127190136275,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.882811372624369e-06,
-      "loss": 1.2858,
-      "step": 210
-    },
-    {
-      "epoch": 0.2053861129136924,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.881645914782988e-06,
-      "loss": 1.2771,
-      "step": 211
-    },
-    {
-      "epoch": 0.2063595068137573,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.880474830742838e-06,
-      "loss": 1.2565,
-      "step": 212
-    },
-    {
-      "epoch": 0.20733290071382218,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.879298123270391e-06,
-      "loss": 1.2459,
-      "step": 213
-    },
-    {
-      "epoch": 0.20830629461388708,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.8781157951454e-06,
-      "loss": 1.2714,
-      "step": 214
-    },
-    {
-      "epoch": 0.20927968851395198,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.876927849160897e-06,
-      "loss": 1.2648,
-      "step": 215
-    },
-    {
-      "epoch": 0.21025308241401688,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.875734288123185e-06,
-      "loss": 1.2913,
-      "step": 216
-    },
-    {
-      "epoch": 0.21122647631408176,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.874535114851832e-06,
-      "loss": 1.255,
-      "step": 217
-    },
-    {
-      "epoch": 0.21219987021414666,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.873330332179663e-06,
-      "loss": 1.2688,
-      "step": 218
-    },
-    {
-      "epoch": 0.21317326411421156,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.872119942952755e-06,
-      "loss": 1.2461,
-      "step": 219
-    },
-    {
-      "epoch": 0.21414665801427643,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.870903950030429e-06,
-      "loss": 1.2685,
-      "step": 220
-    },
-    {
-      "epoch": 0.21512005191434133,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.869682356285244e-06,
-      "loss": 1.2683,
-      "step": 221
-    },
-    {
-      "epoch": 0.21609344581440623,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.86845516460299e-06,
-      "loss": 1.2667,
-      "step": 222
-    },
-    {
-      "epoch": 0.21706683971447113,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.8672223778826786e-06,
-      "loss": 1.2316,
-      "step": 223
-    },
-    {
-      "epoch": 0.218040233614536,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.865983999036543e-06,
-      "loss": 1.233,
-      "step": 224
-    },
-    {
-      "epoch": 0.2190136275146009,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.864740030990025e-06,
-      "loss": 1.2671,
-      "step": 225
-    },
-    {
-      "epoch": 0.2199870214146658,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.863490476681768e-06,
-      "loss": 1.2431,
-      "step": 226
-    },
-    {
-      "epoch": 0.2209604153147307,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.862235339063613e-06,
-      "loss": 1.258,
-      "step": 227
-    },
-    {
-      "epoch": 0.22193380921479558,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.860974621100593e-06,
-      "loss": 1.2264,
-      "step": 228
-    },
-    {
-      "epoch": 0.22290720311486048,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.859708325770919e-06,
-      "loss": 1.2394,
-      "step": 229
-    },
-    {
-      "epoch": 0.22388059701492538,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.858436456065982e-06,
-      "loss": 1.2713,
-      "step": 230
-    },
-    {
-      "epoch": 0.22485399091499025,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.857159014990338e-06,
-      "loss": 1.2514,
-      "step": 231
-    },
-    {
-      "epoch": 0.22582738481505515,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.855876005561706e-06,
-      "loss": 1.2595,
-      "step": 232
-    },
-    {
-      "epoch": 0.22680077871512005,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.85458743081096e-06,
-      "loss": 1.2807,
-      "step": 233
-    },
-    {
-      "epoch": 0.22777417261518496,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.853293293782118e-06,
-      "loss": 1.2597,
-      "step": 234
-    },
-    {
-      "epoch": 0.22874756651524983,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.851993597532341e-06,
-      "loss": 1.2372,
-      "step": 235
-    },
-    {
-      "epoch": 0.22972096041531473,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.850688345131922e-06,
-      "loss": 1.2455,
-      "step": 236
-    },
-    {
-      "epoch": 0.23069435431537963,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.849377539664278e-06,
-      "loss": 1.2385,
-      "step": 237
-    },
-    {
-      "epoch": 0.23166774821544453,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.8480611842259435e-06,
-      "loss": 1.2438,
-      "step": 238
-    },
-    {
-      "epoch": 0.2326411421155094,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.846739281926568e-06,
-      "loss": 1.256,
-      "step": 239
-    },
-    {
-      "epoch": 0.2336145360155743,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.845411835888899e-06,
-      "loss": 1.2447,
-      "step": 240
-    },
-    {
-      "epoch": 0.2345879299156392,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.844078849248785e-06,
-      "loss": 1.2295,
-      "step": 241
-    },
-    {
-      "epoch": 0.23556132381570408,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.842740325155159e-06,
-      "loss": 1.2412,
-      "step": 242
-    },
-    {
-      "epoch": 0.23653471771576898,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.84139626677004e-06,
-      "loss": 1.2739,
-      "step": 243
-    },
-    {
-      "epoch": 0.23750811161583388,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.840046677268514e-06,
-      "loss": 1.2488,
-      "step": 244
-    },
-    {
-      "epoch": 0.23848150551589878,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.838691559838738e-06,
-      "loss": 1.2599,
-      "step": 245
-    },
-    {
-      "epoch": 0.23945489941596365,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.83733091768193e-06,
-      "loss": 1.241,
-      "step": 246
-    },
-    {
-      "epoch": 0.24042829331602855,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.835964754012351e-06,
-      "loss": 1.2532,
-      "step": 247
-    },
-    {
-      "epoch": 0.24140168721609345,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.834593072057313e-06,
-      "loss": 1.2336,
-      "step": 248
-    },
-    {
-      "epoch": 0.24237508111615833,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.833215875057163e-06,
-      "loss": 1.2617,
-      "step": 249
-    },
-    {
-      "epoch": 0.24334847501622323,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.831833166265271e-06,
-      "loss": 1.2501,
-      "step": 250
-    },
-    {
-      "epoch": 0.24432186891628813,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.8304449489480335e-06,
-      "loss": 1.2712,
-      "step": 251
-    },
-    {
-      "epoch": 0.24529526281635303,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.829051226384856e-06,
-      "loss": 1.236,
-      "step": 252
-    },
-    {
-      "epoch": 0.2462686567164179,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.827652001868154e-06,
-      "loss": 1.2453,
-      "step": 253
-    },
-    {
-      "epoch": 0.2472420506164828,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.826247278703333e-06,
-      "loss": 1.2432,
-      "step": 254
-    },
-    {
-      "epoch": 0.2482154445165477,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.8248370602087954e-06,
-      "loss": 1.2688,
-      "step": 255
-    },
-    {
-      "epoch": 0.2491888384166126,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.823421349715921e-06,
-      "loss": 1.2311,
-      "step": 256
-    },
-    {
-      "epoch": 0.2501622323166775,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.8220001505690635e-06,
-      "loss": 1.2459,
-      "step": 257
-    },
-    {
-      "epoch": 0.2511356262167424,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.820573466125544e-06,
-      "loss": 1.2312,
-      "step": 258
-    },
-    {
-      "epoch": 0.25210902011680725,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.819141299755642e-06,
-      "loss": 1.2275,
-      "step": 259
-    },
-    {
-      "epoch": 0.25308241401687215,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.8177036548425856e-06,
-      "loss": 1.2323,
-      "step": 260
-    },
-    {
-      "epoch": 0.25405580791693705,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.816260534782545e-06,
-      "loss": 1.2354,
-      "step": 261
-    },
-    {
-      "epoch": 0.25502920181700195,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.814811942984625e-06,
-      "loss": 1.246,
-      "step": 262
-    },
-    {
-      "epoch": 0.25600259571706685,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.813357882870857e-06,
-      "loss": 1.2261,
-      "step": 263
-    },
-    {
-      "epoch": 0.25697598961713175,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.811898357876189e-06,
-      "loss": 1.2499,
-      "step": 264
-    },
-    {
-      "epoch": 0.25794938351719665,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.810433371448479e-06,
-      "loss": 1.221,
-      "step": 265
-    },
-    {
-      "epoch": 0.2589227774172615,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.8089629270484875e-06,
-      "loss": 1.2249,
-      "step": 266
-    },
-    {
-      "epoch": 0.2598961713173264,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.807487028149868e-06,
-      "loss": 1.2275,
-      "step": 267
-    },
-    {
-      "epoch": 0.2608695652173913,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.80600567823916e-06,
-      "loss": 1.2489,
-      "step": 268
-    },
-    {
-      "epoch": 0.2618429591174562,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.804518880815776e-06,
-      "loss": 1.2254,
-      "step": 269
-    },
-    {
-      "epoch": 0.2628163530175211,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.8030266393920045e-06,
-      "loss": 1.2492,
-      "step": 270
-    },
-    {
-      "epoch": 0.263789746917586,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.801528957492987e-06,
-      "loss": 1.239,
-      "step": 271
-    },
-    {
-      "epoch": 0.2647631408176509,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.800025838656723e-06,
-      "loss": 1.2422,
-      "step": 272
-    },
-    {
-      "epoch": 0.26573653471771574,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.798517286434052e-06,
-      "loss": 1.2337,
-      "step": 273
-    },
-    {
-      "epoch": 0.26670992861778064,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.79700330438865e-06,
-      "loss": 1.2443,
-      "step": 274
-    },
-    {
-      "epoch": 0.26768332251784555,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.795483896097021e-06,
-      "loss": 1.2407,
-      "step": 275
-    },
-    {
-      "epoch": 0.26865671641791045,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.793959065148484e-06,
-      "loss": 1.2472,
-      "step": 276
-    },
-    {
-      "epoch": 0.26963011031797535,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.792428815145175e-06,
-      "loss": 1.2207,
-      "step": 277
-    },
-    {
-      "epoch": 0.27060350421804025,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.790893149702023e-06,
-      "loss": 1.2603,
-      "step": 278
-    },
-    {
-      "epoch": 0.27157689811810515,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.7893520724467566e-06,
-      "loss": 1.2475,
-      "step": 279
-    },
-    {
-      "epoch": 0.27255029201817,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.787805587019885e-06,
-      "loss": 1.2171,
-      "step": 280
-    },
-    {
-      "epoch": 0.2735236859182349,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.786253697074694e-06,
-      "loss": 1.2496,
-      "step": 281
-    },
-    {
-      "epoch": 0.2744970798182998,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.78469640627724e-06,
-      "loss": 1.2153,
-      "step": 282
-    },
-    {
-      "epoch": 0.2754704737183647,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.783133718306331e-06,
-      "loss": 1.2707,
-      "step": 283
-    },
-    {
-      "epoch": 0.2764438676184296,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.781565636853533e-06,
-      "loss": 1.2226,
-      "step": 284
-    },
-    {
-      "epoch": 0.2774172615184945,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.779992165623144e-06,
-      "loss": 1.2133,
-      "step": 285
-    },
-    {
-      "epoch": 0.2783906554185594,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.778413308332204e-06,
-      "loss": 1.2088,
-      "step": 286
-    },
-    {
-      "epoch": 0.2793640493186243,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.776829068710469e-06,
-      "loss": 1.2205,
-      "step": 287
-    },
-    {
-      "epoch": 0.28033744321868914,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.775239450500414e-06,
-      "loss": 1.2188,
-      "step": 288
-    },
-    {
-      "epoch": 0.28131083711875404,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.773644457457218e-06,
-      "loss": 1.2338,
-      "step": 289
-    },
-    {
-      "epoch": 0.28228423101881894,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.772044093348757e-06,
-      "loss": 1.209,
-      "step": 290
-    },
-    {
-      "epoch": 0.28325762491888384,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.770438361955598e-06,
-      "loss": 1.2513,
-      "step": 291
-    },
-    {
-      "epoch": 0.28423101881894874,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.768827267070982e-06,
-      "loss": 1.221,
-      "step": 292
-    },
-    {
-      "epoch": 0.28520441271901364,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.767210812500825e-06,
-      "loss": 1.2144,
-      "step": 293
-    },
-    {
-      "epoch": 0.28617780661907855,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.765589002063702e-06,
-      "loss": 1.2116,
-      "step": 294
-    },
-    {
-      "epoch": 0.2871512005191434,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.763961839590841e-06,
-      "loss": 1.204,
-      "step": 295
-    },
-    {
-      "epoch": 0.2881245944192083,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.762329328926112e-06,
-      "loss": 1.222,
-      "step": 296
-    },
-    {
-      "epoch": 0.2890979883192732,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.760691473926021e-06,
-      "loss": 1.2335,
-      "step": 297
-    },
-    {
-      "epoch": 0.2900713822193381,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.759048278459698e-06,
-      "loss": 1.23,
-      "step": 298
-    },
-    {
-      "epoch": 0.291044776119403,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.757399746408889e-06,
-      "loss": 1.2134,
-      "step": 299
-    },
-    {
-      "epoch": 0.2920181700194679,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.755745881667946e-06,
-      "loss": 1.2191,
-      "step": 300
-    },
-    {
-      "epoch": 0.2929915639195328,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.754086688143819e-06,
-      "loss": 1.2006,
-      "step": 301
-    },
-    {
-      "epoch": 0.29396495781959764,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.752422169756048e-06,
-      "loss": 1.227,
-      "step": 302
-    },
-    {
-      "epoch": 0.29493835171966254,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.750752330436749e-06,
-      "loss": 1.2041,
-      "step": 303
-    },
-    {
-      "epoch": 0.29591174561972744,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.749077174130609e-06,
-      "loss": 1.2204,
-      "step": 304
-    },
-    {
-      "epoch": 0.29688513951979234,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.747396704794876e-06,
-      "loss": 1.2288,
-      "step": 305
-    },
-    {
-      "epoch": 0.29785853341985724,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.745710926399347e-06,
-      "loss": 1.2037,
-      "step": 306
-    },
-    {
-      "epoch": 0.29883192731992214,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.744019842926366e-06,
-      "loss": 1.2168,
-      "step": 307
-    },
-    {
-      "epoch": 0.29980532121998704,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.742323458370801e-06,
-      "loss": 1.2106,
-      "step": 308
-    },
-    {
-      "epoch": 0.3007787151200519,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.740621776740051e-06,
-      "loss": 1.225,
-      "step": 309
-    },
-    {
-      "epoch": 0.3017521090201168,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.738914802054022e-06,
-      "loss": 1.2183,
-      "step": 310
-    },
-    {
-      "epoch": 0.3027255029201817,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.7372025383451285e-06,
-      "loss": 1.23,
-      "step": 311
-    },
-    {
-      "epoch": 0.3036988968202466,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.735484989658275e-06,
-      "loss": 1.2182,
-      "step": 312
-    },
-    {
-      "epoch": 0.3046722907203115,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.733762160050856e-06,
-      "loss": 1.2184,
-      "step": 313
-    },
-    {
-      "epoch": 0.3056456846203764,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.732034053592738e-06,
-      "loss": 1.2021,
-      "step": 314
-    },
-    {
-      "epoch": 0.3066190785204413,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.73030067436625e-06,
-      "loss": 1.2068,
-      "step": 315
-    },
-    {
-      "epoch": 0.3075924724205062,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.728562026466185e-06,
-      "loss": 1.2274,
-      "step": 316
-    },
-    {
-      "epoch": 0.30856586632057104,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.726818113999775e-06,
-      "loss": 1.2237,
-      "step": 317
-    },
-    {
-      "epoch": 0.30953926022063594,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.725068941086693e-06,
-      "loss": 1.2095,
-      "step": 318
-    },
-    {
-      "epoch": 0.31051265412070084,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.723314511859034e-06,
-      "loss": 1.2121,
-      "step": 319
-    },
-    {
-      "epoch": 0.31148604802076574,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.721554830461316e-06,
-      "loss": 1.2385,
-      "step": 320
-    },
-    {
-      "epoch": 0.31245944192083064,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.71978990105046e-06,
-      "loss": 1.2254,
-      "step": 321
-    },
-    {
-      "epoch": 0.31343283582089554,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.718019727795787e-06,
-      "loss": 1.2159,
-      "step": 322
-    },
-    {
-      "epoch": 0.31440622972096044,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.716244314879004e-06,
-      "loss": 1.2199,
-      "step": 323
-    },
-    {
-      "epoch": 0.3153796236210253,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.714463666494195e-06,
-      "loss": 1.2263,
-      "step": 324
-    },
-    {
-      "epoch": 0.3163530175210902,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.712677786847814e-06,
-      "loss": 1.2254,
-      "step": 325
-    },
-    {
-      "epoch": 0.3173264114211551,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.710886680158674e-06,
-      "loss": 1.2196,
-      "step": 326
-    },
-    {
-      "epoch": 0.31829980532122,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.7090903506579314e-06,
-      "loss": 1.2118,
-      "step": 327
-    },
-    {
-      "epoch": 0.3192731992212849,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.707288802589083e-06,
-      "loss": 1.2327,
-      "step": 328
-    },
-    {
-      "epoch": 0.3202465931213498,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.7054820402079535e-06,
-      "loss": 1.2259,
-      "step": 329
-    },
-    {
-      "epoch": 0.3212199870214147,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.7036700677826876e-06,
-      "loss": 1.2155,
-      "step": 330
-    },
-    {
-      "epoch": 0.32219338092147953,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.701852889593733e-06,
-      "loss": 1.2246,
-      "step": 331
-    },
-    {
-      "epoch": 0.32316677482154443,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.70003050993384e-06,
-      "loss": 1.2251,
-      "step": 332
-    },
-    {
-      "epoch": 0.32414016872160933,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.698202933108041e-06,
-      "loss": 1.2039,
-      "step": 333
-    },
-    {
-      "epoch": 0.32511356262167423,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.696370163433652e-06,
-      "loss": 1.2272,
-      "step": 334
-    },
-    {
-      "epoch": 0.32608695652173914,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.694532205240251e-06,
-      "loss": 1.1974,
-      "step": 335
-    },
-    {
-      "epoch": 0.32706035042180404,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.692689062869676e-06,
-      "loss": 1.2181,
-      "step": 336
-    },
-    {
-      "epoch": 0.32803374432186894,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.69084074067601e-06,
-      "loss": 1.22,
-      "step": 337
-    },
-    {
-      "epoch": 0.32900713822193384,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.688987243025574e-06,
-      "loss": 1.2002,
-      "step": 338
-    },
-    {
-      "epoch": 0.3299805321219987,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.687128574296912e-06,
-      "loss": 1.2266,
-      "step": 339
-    },
-    {
-      "epoch": 0.3309539260220636,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.685264738880787e-06,
-      "loss": 1.2075,
-      "step": 340
-    },
-    {
-      "epoch": 0.3319273199221285,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.683395741180166e-06,
-      "loss": 1.2322,
-      "step": 341
-    },
-    {
-      "epoch": 0.3329007138221934,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.681521585610212e-06,
-      "loss": 1.1893,
-      "step": 342
-    },
-    {
-      "epoch": 0.3338741077222583,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.679642276598268e-06,
-      "loss": 1.1961,
-      "step": 343
-    },
-    {
-      "epoch": 0.3348475016223232,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.677757818583859e-06,
-      "loss": 1.2141,
-      "step": 344
-    },
-    {
-      "epoch": 0.3358208955223881,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.675868216018666e-06,
-      "loss": 1.2274,
-      "step": 345
-    },
-    {
-      "epoch": 0.33679428942245293,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.6739734733665275e-06,
-      "loss": 1.1964,
-      "step": 346
-    },
-    {
-      "epoch": 0.33776768332251783,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.672073595103423e-06,
-      "loss": 1.1861,
-      "step": 347
-    },
-    {
-      "epoch": 0.33874107722258273,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.670168585717465e-06,
-      "loss": 1.2064,
-      "step": 348
-    },
-    {
-      "epoch": 0.33971447112264763,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.6682584497088855e-06,
-      "loss": 1.2087,
-      "step": 349
-    },
-    {
-      "epoch": 0.34068786502271253,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.666343191590027e-06,
-      "loss": 1.1953,
-      "step": 350
-    },
-    {
-      "epoch": 0.34166125892277743,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.664422815885336e-06,
-      "loss": 1.2176,
-      "step": 351
-    },
-    {
-      "epoch": 0.34263465282284233,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.6624973271313434e-06,
-      "loss": 1.2124,
-      "step": 352
-    },
-    {
-      "epoch": 0.3436080467229072,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.660566729876661e-06,
-      "loss": 1.2018,
-      "step": 353
-    },
-    {
-      "epoch": 0.3445814406229721,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.65863102868197e-06,
-      "loss": 1.2028,
-      "step": 354
-    },
-    {
-      "epoch": 0.345554834523037,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.6566902281200055e-06,
-      "loss": 1.2266,
-      "step": 355
-    },
-    {
-      "epoch": 0.3465282284231019,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.654744332775552e-06,
-      "loss": 1.2053,
-      "step": 356
-    },
-    {
-      "epoch": 0.3475016223231668,
-      "grad_norm": 0.546875,
-      "learning_rate": 4.6527933472454255e-06,
-      "loss": 1.2024,
-      "step": 357
-    },
-    {
-      "epoch": 0.3484750162232317,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.650837276138471e-06,
-      "loss": 1.1917,
-      "step": 358
-    },
-    {
-      "epoch": 0.3494484101232966,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.648876124075545e-06,
-      "loss": 1.1951,
-      "step": 359
-    },
-    {
-      "epoch": 0.3504218040233614,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.646909895689508e-06,
-      "loss": 1.1942,
-      "step": 360
-    },
-    {
-      "epoch": 0.35139519792342633,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.64493859562521e-06,
-      "loss": 1.1921,
-      "step": 361
-    },
-    {
-      "epoch": 0.35236859182349123,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.642962228539485e-06,
-      "loss": 1.2025,
-      "step": 362
-    },
-    {
-      "epoch": 0.35334198572355613,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.640980799101134e-06,
-      "loss": 1.2043,
-      "step": 363
-    },
-    {
-      "epoch": 0.35431537962362103,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.638994311990919e-06,
-      "loss": 1.2227,
-      "step": 364
-    },
-    {
-      "epoch": 0.35528877352368593,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.637002771901548e-06,
-      "loss": 1.1924,
-      "step": 365
-    },
-    {
-      "epoch": 0.35626216742375083,
-      "grad_norm": 0.486328125,
-      "learning_rate": 4.635006183537668e-06,
-      "loss": 1.1945,
-      "step": 366
-    },
-    {
-      "epoch": 0.35723556132381573,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.633004551615851e-06,
-      "loss": 1.1887,
-      "step": 367
-    },
-    {
-      "epoch": 0.3582089552238806,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.63099788086458e-06,
-      "loss": 1.205,
-      "step": 368
-    },
-    {
-      "epoch": 0.3591823491239455,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.628986176024248e-06,
-      "loss": 1.2369,
-      "step": 369
-    },
-    {
-      "epoch": 0.3601557430240104,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.626969441847133e-06,
-      "loss": 1.2008,
-      "step": 370
-    },
-    {
-      "epoch": 0.3611291369240753,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.624947683097399e-06,
-      "loss": 1.2042,
-      "step": 371
-    },
-    {
-      "epoch": 0.3621025308241402,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.622920904551078e-06,
-      "loss": 1.179,
-      "step": 372
-    },
-    {
-      "epoch": 0.3630759247242051,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.620889110996058e-06,
-      "loss": 1.1847,
-      "step": 373
-    },
-    {
-      "epoch": 0.36404931862427,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.618852307232078e-06,
-      "loss": 1.2136,
-      "step": 374
-    },
-    {
-      "epoch": 0.3650227125243348,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.6168104980707105e-06,
-      "loss": 1.183,
-      "step": 375
-    },
-    {
-      "epoch": 0.3659961064243997,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.614763688335353e-06,
-      "loss": 1.185,
-      "step": 376
-    },
-    {
-      "epoch": 0.3669695003244646,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.612711882861215e-06,
-      "loss": 1.1868,
-      "step": 377
-    },
-    {
-      "epoch": 0.3679428942245295,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.610655086495308e-06,
-      "loss": 1.1888,
-      "step": 378
-    },
-    {
-      "epoch": 0.3689162881245944,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.608593304096434e-06,
-      "loss": 1.1971,
-      "step": 379
-    },
-    {
-      "epoch": 0.3698896820246593,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.606526540535173e-06,
-      "loss": 1.1996,
-      "step": 380
-    },
-    {
-      "epoch": 0.37086307592472423,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.604454800693874e-06,
-      "loss": 1.2032,
-      "step": 381
-    },
-    {
-      "epoch": 0.3718364698247891,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.602378089466637e-06,
-      "loss": 1.213,
-      "step": 382
-    },
-    {
-      "epoch": 0.372809863724854,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.600296411759311e-06,
-      "loss": 1.1809,
-      "step": 383
-    },
-    {
-      "epoch": 0.3737832576249189,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.598209772489475e-06,
-      "loss": 1.178,
-      "step": 384
-    },
-    {
-      "epoch": 0.3747566515249838,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.596118176586429e-06,
-      "loss": 1.181,
-      "step": 385
-    },
-    {
-      "epoch": 0.3757300454250487,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.59402162899118e-06,
-      "loss": 1.2054,
-      "step": 386
-    },
-    {
-      "epoch": 0.3767034393251136,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.591920134656437e-06,
-      "loss": 1.2011,
-      "step": 387
-    },
-    {
-      "epoch": 0.3776768332251785,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.589813698546592e-06,
-      "loss": 1.1845,
-      "step": 388
-    },
-    {
-      "epoch": 0.3786502271252433,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.587702325637712e-06,
-      "loss": 1.1964,
-      "step": 389
-    },
-    {
-      "epoch": 0.3796236210253082,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.585586020917524e-06,
-      "loss": 1.1885,
-      "step": 390
-    },
-    {
-      "epoch": 0.3805970149253731,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.583464789385409e-06,
-      "loss": 1.1947,
-      "step": 391
-    },
-    {
-      "epoch": 0.381570408825438,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.5813386360523835e-06,
-      "loss": 1.1827,
-      "step": 392
-    },
-    {
-      "epoch": 0.3825438027255029,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.579207565941093e-06,
-      "loss": 1.2077,
-      "step": 393
-    },
-    {
-      "epoch": 0.3835171966255678,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.577071584085797e-06,
-      "loss": 1.1792,
-      "step": 394
-    },
-    {
-      "epoch": 0.3844905905256327,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.574930695532357e-06,
-      "loss": 1.1724,
-      "step": 395
-    },
-    {
-      "epoch": 0.3854639844256976,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.572784905338229e-06,
-      "loss": 1.2056,
-      "step": 396
-    },
-    {
-      "epoch": 0.38643737832576247,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.570634218572445e-06,
-      "loss": 1.166,
-      "step": 397
-    },
-    {
-      "epoch": 0.38741077222582737,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.568478640315606e-06,
-      "loss": 1.1755,
-      "step": 398
-    },
-    {
-      "epoch": 0.38838416612589227,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.566318175659868e-06,
-      "loss": 1.2053,
-      "step": 399
-    },
-    {
-      "epoch": 0.3893575600259572,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.564152829708928e-06,
-      "loss": 1.1804,
-      "step": 400
-    },
-    {
-      "epoch": 0.3903309539260221,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.561982607578018e-06,
-      "loss": 1.2036,
-      "step": 401
-    },
-    {
-      "epoch": 0.391304347826087,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.5598075143938855e-06,
-      "loss": 1.1936,
-      "step": 402
-    },
-    {
-      "epoch": 0.3922777417261519,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.557627555294789e-06,
-      "loss": 1.184,
-      "step": 403
-    },
-    {
-      "epoch": 0.3932511356262167,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.555442735430477e-06,
-      "loss": 1.1864,
-      "step": 404
-    },
-    {
-      "epoch": 0.3942245295262816,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.553253059962184e-06,
-      "loss": 1.1955,
-      "step": 405
-    },
-    {
-      "epoch": 0.3951979234263465,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.551058534062614e-06,
-      "loss": 1.1962,
-      "step": 406
-    },
-    {
-      "epoch": 0.3961713173264114,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.548859162915929e-06,
-      "loss": 1.1796,
-      "step": 407
-    },
-    {
-      "epoch": 0.3971447112264763,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.546654951717737e-06,
-      "loss": 1.1852,
-      "step": 408
-    },
-    {
-      "epoch": 0.3981181051265412,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.544445905675082e-06,
-      "loss": 1.1965,
-      "step": 409
-    },
-    {
-      "epoch": 0.3990914990266061,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.5422320300064246e-06,
-      "loss": 1.194,
-      "step": 410
-    },
-    {
-      "epoch": 0.40006489292667097,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.54001332994164e-06,
-      "loss": 1.1922,
-      "step": 411
-    },
-    {
-      "epoch": 0.40103828682673587,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.537789810721996e-06,
-      "loss": 1.2031,
-      "step": 412
-    },
-    {
-      "epoch": 0.40201168072680077,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.535561477600146e-06,
-      "loss": 1.1734,
-      "step": 413
-    },
-    {
-      "epoch": 0.40298507462686567,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.5333283358401155e-06,
-      "loss": 1.184,
-      "step": 414
-    },
-    {
-      "epoch": 0.40395846852693057,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.53109039071729e-06,
-      "loss": 1.192,
-      "step": 415
-    },
-    {
-      "epoch": 0.40493186242699547,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.528847647518403e-06,
-      "loss": 1.1853,
-      "step": 416
-    },
-    {
-      "epoch": 0.40590525632706037,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.526600111541519e-06,
-      "loss": 1.1668,
-      "step": 417
-    },
-    {
-      "epoch": 0.4068786502271252,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.524347788096029e-06,
-      "loss": 1.181,
-      "step": 418
-    },
-    {
-      "epoch": 0.4078520441271901,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.5220906825026295e-06,
-      "loss": 1.1743,
-      "step": 419
-    },
-    {
-      "epoch": 0.408825438027255,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.5198288000933165e-06,
-      "loss": 1.1953,
-      "step": 420
-    },
-    {
-      "epoch": 0.4097988319273199,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.51756214621137e-06,
-      "loss": 1.1808,
-      "step": 421
-    },
-    {
-      "epoch": 0.4107722258273848,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.51529072621134e-06,
-      "loss": 1.1769,
-      "step": 422
-    },
-    {
-      "epoch": 0.4117456197274497,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.513014545459038e-06,
-      "loss": 1.1834,
-      "step": 423
-    },
-    {
-      "epoch": 0.4127190136275146,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.510733609331521e-06,
-      "loss": 1.1732,
-      "step": 424
-    },
-    {
-      "epoch": 0.4136924075275795,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.508447923217079e-06,
-      "loss": 1.1499,
-      "step": 425
-    },
-    {
-      "epoch": 0.41466580142764436,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.506157492515223e-06,
-      "loss": 1.1866,
-      "step": 426
-    },
-    {
-      "epoch": 0.41563919532770927,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.503862322636674e-06,
-      "loss": 1.182,
-      "step": 427
-    },
-    {
-      "epoch": 0.41661258922777417,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.5015624190033445e-06,
-      "loss": 1.1971,
-      "step": 428
-    },
-    {
-      "epoch": 0.41758598312783907,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.499257787048336e-06,
-      "loss": 1.1736,
-      "step": 429
-    },
-    {
-      "epoch": 0.41855937702790397,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.4969484322159125e-06,
-      "loss": 1.1895,
-      "step": 430
-    },
-    {
-      "epoch": 0.41953277092796887,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.494634359961501e-06,
-      "loss": 1.1804,
-      "step": 431
-    },
-    {
-      "epoch": 0.42050616482803377,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.492315575751668e-06,
-      "loss": 1.1797,
-      "step": 432
-    },
-    {
-      "epoch": 0.4214795587280986,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.489992085064115e-06,
-      "loss": 1.1809,
-      "step": 433
-    },
-    {
-      "epoch": 0.4224529526281635,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.487663893387658e-06,
-      "loss": 1.1542,
-      "step": 434
-    },
-    {
-      "epoch": 0.4234263465282284,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.485331006222221e-06,
-      "loss": 1.1896,
-      "step": 435
-    },
-    {
-      "epoch": 0.4243997404282933,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.482993429078818e-06,
-      "loss": 1.1687,
-      "step": 436
-    },
-    {
-      "epoch": 0.4253731343283582,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.480651167479545e-06,
-      "loss": 1.1928,
-      "step": 437
-    },
-    {
-      "epoch": 0.4263465282284231,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.478304226957563e-06,
-      "loss": 1.1903,
-      "step": 438
-    },
-    {
-      "epoch": 0.427319922128488,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.475952613057083e-06,
-      "loss": 1.164,
-      "step": 439
-    },
-    {
-      "epoch": 0.42829331602855286,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.473596331333362e-06,
-      "loss": 1.1756,
-      "step": 440
-    },
-    {
-      "epoch": 0.42926670992861776,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.4712353873526785e-06,
-      "loss": 1.1622,
-      "step": 441
-    },
-    {
-      "epoch": 0.43024010382868266,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.468869786692327e-06,
-      "loss": 1.1759,
-      "step": 442
-    },
-    {
-      "epoch": 0.43121349772874756,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.466499534940602e-06,
-      "loss": 1.184,
-      "step": 443
-    },
-    {
-      "epoch": 0.43218689162881246,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.464124637696786e-06,
-      "loss": 1.1926,
-      "step": 444
-    },
-    {
-      "epoch": 0.43316028552887736,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.461745100571136e-06,
-      "loss": 1.1718,
-      "step": 445
-    },
-    {
-      "epoch": 0.43413367942894227,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.459360929184869e-06,
-      "loss": 1.1872,
-      "step": 446
-    },
-    {
-      "epoch": 0.43510707332900717,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.4569721291701485e-06,
-      "loss": 1.1593,
-      "step": 447
-    },
-    {
-      "epoch": 0.436080467229072,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.454578706170075e-06,
-      "loss": 1.1794,
-      "step": 448
-    },
-    {
-      "epoch": 0.4370538611291369,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.452180665838668e-06,
-      "loss": 1.1764,
-      "step": 449
-    },
-    {
-      "epoch": 0.4380272550292018,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.449778013840854e-06,
-      "loss": 1.1875,
-      "step": 450
-    },
-    {
-      "epoch": 0.4390006489292667,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.4473707558524555e-06,
-      "loss": 1.1611,
-      "step": 451
-    },
-    {
-      "epoch": 0.4399740428293316,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.444958897560175e-06,
-      "loss": 1.1873,
-      "step": 452
-    },
-    {
-      "epoch": 0.4409474367293965,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.442542444661584e-06,
-      "loss": 1.1678,
-      "step": 453
-    },
-    {
-      "epoch": 0.4419208306294614,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.440121402865104e-06,
-      "loss": 1.1832,
-      "step": 454
-    },
-    {
-      "epoch": 0.44289422452952626,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.43769577789e-06,
-      "loss": 1.1594,
-      "step": 455
-    },
-    {
-      "epoch": 0.44386761842959116,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.4352655754663635e-06,
-      "loss": 1.1823,
-      "step": 456
-    },
-    {
-      "epoch": 0.44484101232965606,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.432830801335099e-06,
-      "loss": 1.1574,
-      "step": 457
-    },
-    {
-      "epoch": 0.44581440622972096,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.430391461247911e-06,
-      "loss": 1.1531,
-      "step": 458
-    },
-    {
-      "epoch": 0.44678780012978586,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.42794756096729e-06,
-      "loss": 1.1383,
-      "step": 459
-    },
-    {
-      "epoch": 0.44776119402985076,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.425499106266499e-06,
-      "loss": 1.2007,
-      "step": 460
-    },
-    {
-      "epoch": 0.44873458792991566,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.42304610292956e-06,
-      "loss": 1.1821,
-      "step": 461
-    },
-    {
-      "epoch": 0.4497079818299805,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.42058855675124e-06,
-      "loss": 1.1747,
-      "step": 462
-    },
-    {
-      "epoch": 0.4506813757300454,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.418126473537041e-06,
-      "loss": 1.163,
-      "step": 463
-    },
-    {
-      "epoch": 0.4516547696301103,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.415659859103176e-06,
-      "loss": 1.165,
-      "step": 464
-    },
-    {
-      "epoch": 0.4526281635301752,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.413188719276569e-06,
-      "loss": 1.1847,
-      "step": 465
-    },
-    {
-      "epoch": 0.4536015574302401,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.410713059894831e-06,
-      "loss": 1.1662,
-      "step": 466
-    },
-    {
-      "epoch": 0.454574951330305,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.40823288680625e-06,
-      "loss": 1.1723,
-      "step": 467
-    },
-    {
-      "epoch": 0.4555483452303699,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.405748205869776e-06,
-      "loss": 1.1772,
-      "step": 468
-    },
-    {
-      "epoch": 0.45652173913043476,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.403259022955011e-06,
-      "loss": 1.1551,
-      "step": 469
-    },
-    {
-      "epoch": 0.45749513303049966,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.400765343942189e-06,
-      "loss": 1.1567,
-      "step": 470
-    },
-    {
-      "epoch": 0.45846852693056456,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.398267174722166e-06,
-      "loss": 1.1828,
-      "step": 471
-    },
-    {
-      "epoch": 0.45944192083062946,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.3957645211964065e-06,
-      "loss": 1.1626,
-      "step": 472
-    },
-    {
-      "epoch": 0.46041531473069436,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.393257389276967e-06,
-      "loss": 1.1773,
-      "step": 473
-    },
-    {
-      "epoch": 0.46138870863075926,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.390745784886483e-06,
-      "loss": 1.1845,
-      "step": 474
-    },
-    {
-      "epoch": 0.46236210253082416,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.388229713958157e-06,
-      "loss": 1.1762,
-      "step": 475
-    },
-    {
-      "epoch": 0.46333549643088906,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.3857091824357415e-06,
-      "loss": 1.1789,
-      "step": 476
-    },
-    {
-      "epoch": 0.4643088903309539,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.383184196273528e-06,
-      "loss": 1.1689,
-      "step": 477
-    },
-    {
-      "epoch": 0.4652822842310188,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.380654761436329e-06,
-      "loss": 1.1708,
-      "step": 478
-    },
-    {
-      "epoch": 0.4662556781310837,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.378120883899467e-06,
-      "loss": 1.1806,
-      "step": 479
-    },
-    {
-      "epoch": 0.4672290720311486,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.375582569648761e-06,
-      "loss": 1.1535,
-      "step": 480
-    },
-    {
-      "epoch": 0.4682024659312135,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.37303982468051e-06,
-      "loss": 1.1853,
-      "step": 481
-    },
-    {
-      "epoch": 0.4691758598312784,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.370492655001478e-06,
-      "loss": 1.1766,
-      "step": 482
-    },
-    {
-      "epoch": 0.4701492537313433,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.3679410666288845e-06,
-      "loss": 1.178,
-      "step": 483
-    },
-    {
-      "epoch": 0.47112264763140815,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.3653850655903854e-06,
-      "loss": 1.1828,
-      "step": 484
-    },
-    {
-      "epoch": 0.47209604153147305,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.36282465792406e-06,
-      "loss": 1.166,
-      "step": 485
-    },
-    {
-      "epoch": 0.47306943543153795,
-      "grad_norm": 0.302734375,
-      "learning_rate": 4.360259849678402e-06,
-      "loss": 1.1522,
-      "step": 486
-    },
-    {
-      "epoch": 0.47404282933160286,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.357690646912295e-06,
-      "loss": 1.1681,
-      "step": 487
-    },
-    {
-      "epoch": 0.47501622323166776,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.355117055695005e-06,
-      "loss": 1.1652,
-      "step": 488
-    },
-    {
-      "epoch": 0.47598961713173266,
-      "grad_norm": 0.302734375,
-      "learning_rate": 4.352539082106168e-06,
-      "loss": 1.1571,
-      "step": 489
-    },
-    {
-      "epoch": 0.47696301103179756,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.349956732235772e-06,
-      "loss": 1.1718,
-      "step": 490
-    },
-    {
-      "epoch": 0.4779364049318624,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.347370012184139e-06,
-      "loss": 1.1667,
-      "step": 491
-    },
-    {
-      "epoch": 0.4789097988319273,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.344778928061919e-06,
-      "loss": 1.159,
-      "step": 492
-    },
-    {
-      "epoch": 0.4798831927319922,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.3421834859900695e-06,
-      "loss": 1.17,
-      "step": 493
-    },
-    {
-      "epoch": 0.4808565866320571,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.339583692099843e-06,
-      "loss": 1.1538,
-      "step": 494
-    },
-    {
-      "epoch": 0.481829980532122,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.3369795525327715e-06,
-      "loss": 1.1504,
-      "step": 495
-    },
-    {
-      "epoch": 0.4828033744321869,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.334371073440658e-06,
-      "loss": 1.1759,
-      "step": 496
-    },
-    {
-      "epoch": 0.4837767683322518,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.331758260985548e-06,
-      "loss": 1.1725,
-      "step": 497
-    },
-    {
-      "epoch": 0.48475016223231665,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.329141121339731e-06,
-      "loss": 1.1786,
-      "step": 498
-    },
-    {
-      "epoch": 0.48572355613238155,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.326519660685715e-06,
-      "loss": 1.1781,
-      "step": 499
-    },
-    {
-      "epoch": 0.48669695003244645,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.3238938852162195e-06,
-      "loss": 1.1607,
-      "step": 500
-    },
-    {
-      "epoch": 0.48767034393251135,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.321263801134151e-06,
-      "loss": 1.1665,
-      "step": 501
-    },
-    {
-      "epoch": 0.48864373783257625,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.3186294146526e-06,
-      "loss": 1.1689,
-      "step": 502
-    },
-    {
-      "epoch": 0.48961713173264115,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.315990731994817e-06,
-      "loss": 1.1716,
-      "step": 503
-    },
-    {
-      "epoch": 0.49059052563270605,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.3133477593942015e-06,
-      "loss": 1.1738,
-      "step": 504
-    },
-    {
-      "epoch": 0.49156391953277095,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.310700503094291e-06,
-      "loss": 1.1716,
-      "step": 505
-    },
-    {
-      "epoch": 0.4925373134328358,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.308048969348738e-06,
-      "loss": 1.1547,
-      "step": 506
-    },
-    {
-      "epoch": 0.4935107073329007,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.305393164421301e-06,
-      "loss": 1.1684,
-      "step": 507
-    },
-    {
-      "epoch": 0.4944841012329656,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.30273309458583e-06,
-      "loss": 1.1607,
-      "step": 508
-    },
-    {
-      "epoch": 0.4954574951330305,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.3000687661262495e-06,
-      "loss": 1.1582,
-      "step": 509
-    },
-    {
-      "epoch": 0.4964308890330954,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.297400185336541e-06,
-      "loss": 1.1769,
-      "step": 510
-    },
-    {
-      "epoch": 0.4974042829331603,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.294727358520737e-06,
-      "loss": 1.1715,
-      "step": 511
-    },
-    {
-      "epoch": 0.4983776768332252,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.2920502919928956e-06,
-      "loss": 1.1516,
-      "step": 512
-    },
-    {
-      "epoch": 0.49935107073329005,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.289368992077092e-06,
-      "loss": 1.1522,
-      "step": 513
-    },
-    {
-      "epoch": 0.500324464633355,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.286683465107403e-06,
-      "loss": 1.1532,
-      "step": 514
-    },
-    {
-      "epoch": 0.5012978585334199,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.28399371742789e-06,
-      "loss": 1.1768,
-      "step": 515
-    },
-    {
-      "epoch": 0.5022712524334848,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.2812997553925854e-06,
-      "loss": 1.1675,
-      "step": 516
-    },
-    {
-      "epoch": 0.5032446463335496,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.278601585365477e-06,
-      "loss": 1.1615,
-      "step": 517
-    },
-    {
-      "epoch": 0.5042180402336145,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.275899213720493e-06,
-      "loss": 1.161,
-      "step": 518
-    },
-    {
-      "epoch": 0.5051914341336794,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.273192646841489e-06,
-      "loss": 1.1674,
-      "step": 519
-    },
-    {
-      "epoch": 0.5061648280337443,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.270481891122227e-06,
-      "loss": 1.158,
-      "step": 520
-    },
-    {
-      "epoch": 0.5071382219338092,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.267766952966369e-06,
-      "loss": 1.1423,
-      "step": 521
-    },
-    {
-      "epoch": 0.5081116158338741,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.265047838787455e-06,
-      "loss": 1.1618,
-      "step": 522
-    },
-    {
-      "epoch": 0.509085009733939,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.26232455500889e-06,
-      "loss": 1.1699,
-      "step": 523
-    },
-    {
-      "epoch": 0.5100584036340039,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.259597108063927e-06,
-      "loss": 1.173,
-      "step": 524
-    },
-    {
-      "epoch": 0.5110317975340688,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.256865504395659e-06,
-      "loss": 1.1795,
-      "step": 525
-    },
-    {
-      "epoch": 0.5120051914341337,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.254129750456995e-06,
-      "loss": 1.1652,
-      "step": 526
-    },
-    {
-      "epoch": 0.5129785853341986,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.251389852710647e-06,
-      "loss": 1.1535,
-      "step": 527
-    },
-    {
-      "epoch": 0.5139519792342635,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.2486458176291176e-06,
-      "loss": 1.1548,
-      "step": 528
-    },
-    {
-      "epoch": 0.5149253731343284,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.245897651694685e-06,
-      "loss": 1.1659,
-      "step": 529
-    },
-    {
-      "epoch": 0.5158987670343933,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.243145361399383e-06,
-      "loss": 1.1587,
-      "step": 530
-    },
-    {
-      "epoch": 0.5168721609344581,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.240388953244987e-06,
-      "loss": 1.1506,
-      "step": 531
-    },
-    {
-      "epoch": 0.517845554834523,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.2376284337430054e-06,
-      "loss": 1.1476,
-      "step": 532
-    },
-    {
-      "epoch": 0.5188189487345879,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.234863809414654e-06,
-      "loss": 1.1611,
-      "step": 533
-    },
-    {
-      "epoch": 0.5197923426346528,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.232095086790849e-06,
-      "loss": 1.1484,
-      "step": 534
-    },
-    {
-      "epoch": 0.5207657365347177,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.2293222724121855e-06,
-      "loss": 1.151,
-      "step": 535
-    },
-    {
-      "epoch": 0.5217391304347826,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.226545372828925e-06,
-      "loss": 1.1747,
-      "step": 536
-    },
-    {
-      "epoch": 0.5227125243348475,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.223764394600981e-06,
-      "loss": 1.1428,
-      "step": 537
-    },
-    {
-      "epoch": 0.5236859182349124,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.220979344297901e-06,
-      "loss": 1.1716,
-      "step": 538
-    },
-    {
-      "epoch": 0.5246593121349773,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.218190228498851e-06,
-      "loss": 1.1703,
-      "step": 539
-    },
-    {
-      "epoch": 0.5256327060350422,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.215397053792604e-06,
-      "loss": 1.154,
-      "step": 540
-    },
-    {
-      "epoch": 0.5266060999351071,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.212599826777518e-06,
-      "loss": 1.1275,
-      "step": 541
-    },
-    {
-      "epoch": 0.527579493835172,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.209798554061527e-06,
-      "loss": 1.1573,
-      "step": 542
-    },
-    {
-      "epoch": 0.5285528877352369,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.206993242262118e-06,
-      "loss": 1.1553,
-      "step": 543
-    },
-    {
-      "epoch": 0.5295262816353018,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.204183898006326e-06,
-      "loss": 1.1505,
-      "step": 544
-    },
-    {
-      "epoch": 0.5304996755353667,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.201370527930705e-06,
-      "loss": 1.1725,
-      "step": 545
-    },
-    {
-      "epoch": 0.5314730694354315,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.198553138681324e-06,
-      "loss": 1.173,
-      "step": 546
-    },
-    {
-      "epoch": 0.5324464633354964,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.195731736913743e-06,
-      "loss": 1.1346,
-      "step": 547
-    },
-    {
-      "epoch": 0.5334198572355613,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.192906329293005e-06,
-      "loss": 1.1605,
-      "step": 548
-    },
-    {
-      "epoch": 0.5343932511356262,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.1900769224936125e-06,
-      "loss": 1.1575,
-      "step": 549
-    },
-    {
-      "epoch": 0.5353666450356911,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.187243523199518e-06,
-      "loss": 1.1635,
-      "step": 550
-    },
-    {
-      "epoch": 0.536340038935756,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.184406138104102e-06,
-      "loss": 1.1736,
-      "step": 551
-    },
-    {
-      "epoch": 0.5373134328358209,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.1815647739101665e-06,
-      "loss": 1.1759,
-      "step": 552
-    },
-    {
-      "epoch": 0.5382868267358858,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.178719437329905e-06,
-      "loss": 1.1491,
-      "step": 553
-    },
-    {
-      "epoch": 0.5392602206359507,
-      "grad_norm": 0.302734375,
-      "learning_rate": 4.175870135084905e-06,
-      "loss": 1.1562,
-      "step": 554
-    },
-    {
-      "epoch": 0.5402336145360156,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.173016873906113e-06,
-      "loss": 1.145,
-      "step": 555
-    },
-    {
-      "epoch": 0.5412070084360805,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.170159660533834e-06,
-      "loss": 1.1607,
-      "step": 556
-    },
-    {
-      "epoch": 0.5421804023361454,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.167298501717706e-06,
-      "loss": 1.1648,
-      "step": 557
-    },
-    {
-      "epoch": 0.5431537962362103,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.164433404216689e-06,
-      "loss": 1.1585,
-      "step": 558
-    },
-    {
-      "epoch": 0.5441271901362752,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.161564374799047e-06,
-      "loss": 1.1596,
-      "step": 559
-    },
-    {
-      "epoch": 0.54510058403634,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.158691420242331e-06,
-      "loss": 1.1463,
-      "step": 560
-    },
-    {
-      "epoch": 0.5460739779364049,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.155814547333368e-06,
-      "loss": 1.1629,
-      "step": 561
-    },
-    {
-      "epoch": 0.5470473718364698,
-      "grad_norm": 0.302734375,
-      "learning_rate": 4.1529337628682375e-06,
-      "loss": 1.1543,
-      "step": 562
-    },
-    {
-      "epoch": 0.5480207657365347,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.150049073652262e-06,
-      "loss": 1.176,
-      "step": 563
-    },
-    {
-      "epoch": 0.5489941596365996,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.147160486499985e-06,
-      "loss": 1.1534,
-      "step": 564
-    },
-    {
-      "epoch": 0.5499675535366645,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.144268008235163e-06,
-      "loss": 1.161,
-      "step": 565
-    },
-    {
-      "epoch": 0.5509409474367294,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.14137164569074e-06,
-      "loss": 1.1548,
-      "step": 566
-    },
-    {
-      "epoch": 0.5519143413367943,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.138471405708839e-06,
-      "loss": 1.1705,
-      "step": 567
-    },
-    {
-      "epoch": 0.5528877352368592,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.135567295140741e-06,
-      "loss": 1.1429,
-      "step": 568
-    },
-    {
-      "epoch": 0.5538611291369241,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.132659320846871e-06,
-      "loss": 1.1628,
-      "step": 569
-    },
-    {
-      "epoch": 0.554834523036989,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.129747489696781e-06,
-      "loss": 1.1549,
-      "step": 570
-    },
-    {
-      "epoch": 0.5558079169370539,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.126831808569136e-06,
-      "loss": 1.1526,
-      "step": 571
-    },
-    {
-      "epoch": 0.5567813108371188,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.123912284351693e-06,
-      "loss": 1.1524,
-      "step": 572
-    },
-    {
-      "epoch": 0.5577547047371837,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.120988923941289e-06,
-      "loss": 1.1487,
-      "step": 573
-    },
-    {
-      "epoch": 0.5587280986372486,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.118061734243824e-06,
-      "loss": 1.1507,
-      "step": 574
-    },
-    {
-      "epoch": 0.5597014925373134,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.115130722174242e-06,
-      "loss": 1.1355,
-      "step": 575
-    },
-    {
-      "epoch": 0.5606748864373783,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.112195894656519e-06,
-      "loss": 1.1634,
-      "step": 576
-    },
-    {
-      "epoch": 0.5616482803374432,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.109257258623644e-06,
-      "loss": 1.1654,
-      "step": 577
-    },
-    {
-      "epoch": 0.5626216742375081,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.1063148210176e-06,
-      "loss": 1.1519,
-      "step": 578
-    },
-    {
-      "epoch": 0.563595068137573,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.103368588789353e-06,
-      "loss": 1.1698,
-      "step": 579
-    },
-    {
-      "epoch": 0.5645684620376379,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.100418568898833e-06,
-      "loss": 1.1455,
-      "step": 580
-    },
-    {
-      "epoch": 0.5655418559377028,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.097464768314919e-06,
-      "loss": 1.1384,
-      "step": 581
-    },
-    {
-      "epoch": 0.5665152498377677,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.094507194015417e-06,
-      "loss": 1.1458,
-      "step": 582
-    },
-    {
-      "epoch": 0.5674886437378326,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.091545852987051e-06,
-      "loss": 1.134,
-      "step": 583
-    },
-    {
-      "epoch": 0.5684620376378975,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.0885807522254435e-06,
-      "loss": 1.1547,
-      "step": 584
-    },
-    {
-      "epoch": 0.5694354315379624,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.085611898735098e-06,
-      "loss": 1.1378,
-      "step": 585
-    },
-    {
-      "epoch": 0.5704088254380273,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.08263929952938e-06,
-      "loss": 1.1387,
-      "step": 586
-    },
-    {
-      "epoch": 0.5713822193380922,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.079662961630508e-06,
-      "loss": 1.1365,
-      "step": 587
-    },
-    {
-      "epoch": 0.5723556132381571,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.0766828920695326e-06,
-      "loss": 1.1572,
-      "step": 588
-    },
-    {
-      "epoch": 0.5733290071382219,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.073699097886314e-06,
-      "loss": 1.1731,
-      "step": 589
-    },
-    {
-      "epoch": 0.5743024010382868,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.070711586129519e-06,
-      "loss": 1.1721,
-      "step": 590
-    },
-    {
-      "epoch": 0.5752757949383517,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.0677203638565895e-06,
-      "loss": 1.1325,
-      "step": 591
-    },
-    {
-      "epoch": 0.5762491888384166,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.064725438133738e-06,
-      "loss": 1.1496,
-      "step": 592
-    },
-    {
-      "epoch": 0.5772225827384815,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.061726816035922e-06,
-      "loss": 1.1556,
-      "step": 593
-    },
-    {
-      "epoch": 0.5781959766385464,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.058724504646834e-06,
-      "loss": 1.1618,
-      "step": 594
-    },
-    {
-      "epoch": 0.5791693705386113,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.055718511058879e-06,
-      "loss": 1.161,
-      "step": 595
-    },
-    {
-      "epoch": 0.5801427644386762,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.052708842373164e-06,
-      "loss": 1.1204,
-      "step": 596
-    },
-    {
-      "epoch": 0.5811161583387411,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.049695505699473e-06,
-      "loss": 1.149,
-      "step": 597
-    },
-    {
-      "epoch": 0.582089552238806,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.046678508156259e-06,
-      "loss": 1.1592,
-      "step": 598
-    },
-    {
-      "epoch": 0.5830629461388709,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.0436578568706215e-06,
-      "loss": 1.1502,
-      "step": 599
-    },
-    {
-      "epoch": 0.5840363400389358,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.040633558978291e-06,
-      "loss": 1.1522,
-      "step": 600
-    },
-    {
-      "epoch": 0.5850097339390007,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.037605621623614e-06,
-      "loss": 1.1684,
-      "step": 601
-    },
-    {
-      "epoch": 0.5859831278390656,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.034574051959532e-06,
-      "loss": 1.1303,
-      "step": 602
-    },
-    {
-      "epoch": 0.5869565217391305,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.03153885714757e-06,
-      "loss": 1.159,
-      "step": 603
-    },
-    {
-      "epoch": 0.5879299156391953,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.028500044357816e-06,
-      "loss": 1.1514,
-      "step": 604
-    },
-    {
-      "epoch": 0.5889033095392602,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.025457620768901e-06,
-      "loss": 1.1263,
-      "step": 605
-    },
-    {
-      "epoch": 0.5898767034393251,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.022411593567992e-06,
-      "loss": 1.1438,
-      "step": 606
-    },
-    {
-      "epoch": 0.59085009733939,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.019361969950764e-06,
-      "loss": 1.1459,
-      "step": 607
-    },
-    {
-      "epoch": 0.5918234912394549,
-      "grad_norm": 0.302734375,
-      "learning_rate": 4.016308757121392e-06,
-      "loss": 1.1323,
-      "step": 608
-    },
-    {
-      "epoch": 0.5927968851395198,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.013251962292526e-06,
-      "loss": 1.1366,
-      "step": 609
-    },
-    {
-      "epoch": 0.5937702790395847,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.0101915926852795e-06,
-      "loss": 1.1507,
-      "step": 610
-    },
-    {
-      "epoch": 0.5947436729396496,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.007127655529211e-06,
-      "loss": 1.1468,
-      "step": 611
-    },
-    {
-      "epoch": 0.5957170668397145,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.004060158062306e-06,
-      "loss": 1.1407,
-      "step": 612
-    },
-    {
-      "epoch": 0.5966904607397794,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.0009891075309625e-06,
-      "loss": 1.1478,
-      "step": 613
-    },
-    {
-      "epoch": 0.5976638546398443,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.997914511189968e-06,
-      "loss": 1.161,
-      "step": 614
-    },
-    {
-      "epoch": 0.5986372485399092,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.99483637630249e-06,
-      "loss": 1.147,
-      "step": 615
-    },
-    {
-      "epoch": 0.5996106424399741,
-      "grad_norm": 0.353515625,
-      "learning_rate": 3.991754710140053e-06,
-      "loss": 1.1457,
-      "step": 616
-    },
-    {
-      "epoch": 0.600584036340039,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.9886695199825266e-06,
-      "loss": 1.1505,
-      "step": 617
-    },
-    {
-      "epoch": 0.6015574302401038,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.9855808131181e-06,
-      "loss": 1.1422,
-      "step": 618
-    },
-    {
-      "epoch": 0.6025308241401687,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.982488596843276e-06,
-      "loss": 1.1562,
-      "step": 619
-    },
-    {
-      "epoch": 0.6035042180402336,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.979392878462842e-06,
-      "loss": 1.1552,
-      "step": 620
-    },
-    {
-      "epoch": 0.6044776119402985,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.976293665289862e-06,
-      "loss": 1.136,
-      "step": 621
-    },
-    {
-      "epoch": 0.6054510058403634,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.973190964645655e-06,
-      "loss": 1.1411,
-      "step": 622
-    },
-    {
-      "epoch": 0.6064243997404283,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.970084783859779e-06,
-      "loss": 1.148,
-      "step": 623
-    },
-    {
-      "epoch": 0.6073977936404932,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.966975130270012e-06,
-      "loss": 1.1427,
-      "step": 624
-    },
-    {
-      "epoch": 0.6083711875405581,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.963862011222337e-06,
-      "loss": 1.1386,
-      "step": 625
-    },
-    {
-      "epoch": 0.609344581440623,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.9607454340709215e-06,
-      "loss": 1.1396,
-      "step": 626
-    },
-    {
-      "epoch": 0.6103179753406879,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.957625406178105e-06,
-      "loss": 1.1296,
-      "step": 627
-    },
-    {
-      "epoch": 0.6112913692407528,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.9545019349143784e-06,
-      "loss": 1.1452,
-      "step": 628
-    },
-    {
-      "epoch": 0.6122647631408177,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.951375027658364e-06,
-      "loss": 1.1642,
-      "step": 629
-    },
-    {
-      "epoch": 0.6132381570408826,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.948244691796803e-06,
-      "loss": 1.1405,
-      "step": 630
-    },
-    {
-      "epoch": 0.6142115509409475,
-      "grad_norm": 0.3203125,
-      "learning_rate": 3.945110934724537e-06,
-      "loss": 1.1355,
-      "step": 631
-    },
-    {
-      "epoch": 0.6151849448410124,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.941973763844488e-06,
-      "loss": 1.1419,
-      "step": 632
-    },
-    {
-      "epoch": 0.6161583387410772,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.9388331865676436e-06,
-      "loss": 1.1489,
-      "step": 633
-    },
-    {
-      "epoch": 0.6171317326411421,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.935689210313036e-06,
-      "loss": 1.1359,
-      "step": 634
-    },
-    {
-      "epoch": 0.618105126541207,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.93254184250773e-06,
-      "loss": 1.1236,
-      "step": 635
-    },
-    {
-      "epoch": 0.6190785204412719,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.929391090586801e-06,
-      "loss": 1.1396,
-      "step": 636
-    },
-    {
-      "epoch": 0.6200519143413368,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.926236961993319e-06,
-      "loss": 1.1466,
-      "step": 637
-    },
-    {
-      "epoch": 0.6210253082414017,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.92307946417833e-06,
-      "loss": 1.1496,
-      "step": 638
-    },
-    {
-      "epoch": 0.6219987021414666,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.919918604600838e-06,
-      "loss": 1.139,
-      "step": 639
-    },
-    {
-      "epoch": 0.6229720960415315,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.916754390727795e-06,
-      "loss": 1.1254,
-      "step": 640
-    },
-    {
-      "epoch": 0.6239454899415964,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.913586830034067e-06,
-      "loss": 1.1303,
-      "step": 641
-    },
-    {
-      "epoch": 0.6249188838416613,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.910415930002433e-06,
-      "loss": 1.1203,
-      "step": 642
-    },
-    {
-      "epoch": 0.6258922777417262,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.907241698123559e-06,
-      "loss": 1.1401,
-      "step": 643
-    },
-    {
-      "epoch": 0.6268656716417911,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.904064141895984e-06,
-      "loss": 1.1253,
-      "step": 644
-    },
-    {
-      "epoch": 0.627839065541856,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.9008832688260935e-06,
-      "loss": 1.1549,
-      "step": 645
-    },
-    {
-      "epoch": 0.6288124594419209,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.897699086428117e-06,
-      "loss": 1.1617,
-      "step": 646
-    },
-    {
-      "epoch": 0.6297858533419857,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.8945116022240945e-06,
-      "loss": 1.1339,
-      "step": 647
-    },
-    {
-      "epoch": 0.6307592472420506,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.8913208237438704e-06,
-      "loss": 1.1407,
-      "step": 648
-    },
-    {
-      "epoch": 0.6317326411421155,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.888126758525071e-06,
-      "loss": 1.1467,
-      "step": 649
-    },
-    {
-      "epoch": 0.6327060350421804,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.884929414113082e-06,
-      "loss": 1.11,
-      "step": 650
-    },
-    {
-      "epoch": 0.6336794289422453,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.881728798061043e-06,
-      "loss": 1.1691,
-      "step": 651
-    },
-    {
-      "epoch": 0.6346528228423102,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.878524917929818e-06,
-      "loss": 1.1577,
-      "step": 652
-    },
-    {
-      "epoch": 0.6356262167423751,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.875317781287979e-06,
-      "loss": 1.1608,
-      "step": 653
-    },
-    {
-      "epoch": 0.63659961064244,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.872107395711799e-06,
-      "loss": 1.1551,
-      "step": 654
-    },
-    {
-      "epoch": 0.6375730045425049,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.868893768785217e-06,
-      "loss": 1.1493,
-      "step": 655
-    },
-    {
-      "epoch": 0.6385463984425698,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.865676908099836e-06,
-      "loss": 1.1478,
-      "step": 656
-    },
-    {
-      "epoch": 0.6395197923426347,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.862456821254894e-06,
-      "loss": 1.1385,
-      "step": 657
-    },
-    {
-      "epoch": 0.6404931862426996,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.859233515857253e-06,
-      "loss": 1.1472,
-      "step": 658
-    },
-    {
-      "epoch": 0.6414665801427645,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.8560069995213756e-06,
-      "loss": 1.1435,
-      "step": 659
-    },
-    {
-      "epoch": 0.6424399740428294,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.852777279869312e-06,
-      "loss": 1.1415,
-      "step": 660
-    },
-    {
-      "epoch": 0.6434133679428943,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.849544364530678e-06,
-      "loss": 1.1116,
-      "step": 661
-    },
-    {
-      "epoch": 0.6443867618429591,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.846308261142639e-06,
-      "loss": 1.1247,
-      "step": 662
-    },
-    {
-      "epoch": 0.645360155743024,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.843068977349892e-06,
-      "loss": 1.1312,
-      "step": 663
-    },
-    {
-      "epoch": 0.6463335496430889,
-      "grad_norm": 0.3203125,
-      "learning_rate": 3.839826520804648e-06,
-      "loss": 1.1567,
-      "step": 664
-    },
-    {
-      "epoch": 0.6473069435431538,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.83658089916661e-06,
-      "loss": 1.1262,
-      "step": 665
-    },
-    {
-      "epoch": 0.6482803374432187,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.833332120102961e-06,
-      "loss": 1.1511,
-      "step": 666
-    },
-    {
-      "epoch": 0.6492537313432836,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.830080191288342e-06,
-      "loss": 1.1398,
-      "step": 667
-    },
-    {
-      "epoch": 0.6502271252433485,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.8268251204048335e-06,
-      "loss": 1.1574,
-      "step": 668
-    },
-    {
-      "epoch": 0.6512005191434134,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.823566915141941e-06,
-      "loss": 1.1366,
-      "step": 669
-    },
-    {
-      "epoch": 0.6521739130434783,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.820305583196571e-06,
-      "loss": 1.1221,
-      "step": 670
-    },
-    {
-      "epoch": 0.6531473069435432,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.81704113227302e-06,
-      "loss": 1.1409,
-      "step": 671
-    },
-    {
-      "epoch": 0.6541207008436081,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.8137735700829494e-06,
-      "loss": 1.1469,
-      "step": 672
-    },
-    {
-      "epoch": 0.655094094743673,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.810502904345371e-06,
-      "loss": 1.126,
-      "step": 673
-    },
-    {
-      "epoch": 0.6560674886437379,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.807229142786631e-06,
-      "loss": 1.1344,
-      "step": 674
-    },
-    {
-      "epoch": 0.6570408825438028,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.803952293140385e-06,
-      "loss": 1.1434,
-      "step": 675
-    },
-    {
-      "epoch": 0.6580142764438677,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.800672363147585e-06,
-      "loss": 1.1356,
-      "step": 676
-    },
-    {
-      "epoch": 0.6589876703439325,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.7973893605564615e-06,
-      "loss": 1.1235,
-      "step": 677
-    },
-    {
-      "epoch": 0.6599610642439974,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.7941032931225007e-06,
-      "loss": 1.1105,
-      "step": 678
-    },
-    {
-      "epoch": 0.6609344581440623,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.7908141686084305e-06,
-      "loss": 1.1375,
-      "step": 679
-    },
-    {
-      "epoch": 0.6619078520441272,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.7875219947842006e-06,
-      "loss": 1.1295,
-      "step": 680
-    },
-    {
-      "epoch": 0.6628812459441921,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.784226779426965e-06,
-      "loss": 1.1294,
-      "step": 681
-    },
-    {
-      "epoch": 0.663854639844257,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.78092853032106e-06,
-      "loss": 1.1331,
-      "step": 682
-    },
-    {
-      "epoch": 0.6648280337443219,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.7776272552579896e-06,
-      "loss": 1.1202,
-      "step": 683
-    },
-    {
-      "epoch": 0.6658014276443868,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.774322962036408e-06,
-      "loss": 1.1376,
-      "step": 684
-    },
-    {
-      "epoch": 0.6667748215444517,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.771015658462098e-06,
-      "loss": 1.1353,
-      "step": 685
-    },
-    {
-      "epoch": 0.6677482154445166,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.7677053523479534e-06,
-      "loss": 1.1484,
-      "step": 686
-    },
-    {
-      "epoch": 0.6687216093445815,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.7643920515139597e-06,
-      "loss": 1.129,
-      "step": 687
-    },
-    {
-      "epoch": 0.6696950032446464,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.761075763787181e-06,
-      "loss": 1.1342,
-      "step": 688
-    },
-    {
-      "epoch": 0.6706683971447113,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.7577564970017338e-06,
-      "loss": 1.132,
-      "step": 689
-    },
-    {
-      "epoch": 0.6716417910447762,
-      "grad_norm": 0.3359375,
-      "learning_rate": 3.754434258998772e-06,
-      "loss": 1.1385,
-      "step": 690
-    },
-    {
-      "epoch": 0.672615184944841,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.7511090576264708e-06,
-      "loss": 1.1309,
-      "step": 691
-    },
-    {
-      "epoch": 0.6735885788449059,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.7477809007400045e-06,
-      "loss": 1.1215,
-      "step": 692
-    },
-    {
-      "epoch": 0.6745619727449708,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.744449796201528e-06,
-      "loss": 1.1319,
-      "step": 693
-    },
-    {
-      "epoch": 0.6755353666450357,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.7411157518801622e-06,
-      "loss": 1.1149,
-      "step": 694
-    },
-    {
-      "epoch": 0.6765087605451006,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.737778775651971e-06,
-      "loss": 1.1374,
-      "step": 695
-    },
-    {
-      "epoch": 0.6774821544451655,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.7344388753999434e-06,
-      "loss": 1.1396,
-      "step": 696
-    },
-    {
-      "epoch": 0.6784555483452304,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.731096059013979e-06,
-      "loss": 1.1482,
-      "step": 697
-    },
-    {
-      "epoch": 0.6794289422452953,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.7277503343908627e-06,
-      "loss": 1.1383,
-      "step": 698
-    },
-    {
-      "epoch": 0.6804023361453602,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.7244017094342526e-06,
-      "loss": 1.1184,
-      "step": 699
-    },
-    {
-      "epoch": 0.6813757300454251,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.7210501920546565e-06,
-      "loss": 1.1349,
-      "step": 700
-    },
-    {
-      "epoch": 0.68234912394549,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.7176957901694144e-06,
-      "loss": 1.1601,
-      "step": 701
-    },
-    {
-      "epoch": 0.6833225178455549,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.714338511702683e-06,
-      "loss": 1.1478,
-      "step": 702
-    },
-    {
-      "epoch": 0.6842959117456198,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.7109783645854116e-06,
-      "loss": 1.1334,
-      "step": 703
-    },
-    {
-      "epoch": 0.6852693056456847,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.7076153567553286e-06,
-      "loss": 1.1552,
-      "step": 704
-    },
-    {
-      "epoch": 0.6862426995457496,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.704249496156919e-06,
-      "loss": 1.1386,
-      "step": 705
-    },
-    {
-      "epoch": 0.6872160934458144,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.700880790741405e-06,
-      "loss": 1.1463,
-      "step": 706
-    },
-    {
-      "epoch": 0.6881894873458793,
-      "grad_norm": 0.34765625,
-      "learning_rate": 3.6975092484667336e-06,
-      "loss": 1.146,
-      "step": 707
-    },
-    {
-      "epoch": 0.6891628812459442,
-      "grad_norm": 0.39453125,
-      "learning_rate": 3.69413487729755e-06,
-      "loss": 1.1323,
-      "step": 708
-    },
-    {
-      "epoch": 0.6901362751460091,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.6907576852051836e-06,
-      "loss": 1.1284,
-      "step": 709
-    },
-    {
-      "epoch": 0.691109669046074,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.6873776801676265e-06,
-      "loss": 1.1416,
-      "step": 710
-    },
-    {
-      "epoch": 0.6920830629461389,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.683994870169516e-06,
-      "loss": 1.1301,
-      "step": 711
-    },
-    {
-      "epoch": 0.6930564568462038,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.6806092632021184e-06,
-      "loss": 1.133,
-      "step": 712
-    },
-    {
-      "epoch": 0.6940298507462687,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.6772208672633024e-06,
-      "loss": 1.134,
-      "step": 713
-    },
-    {
-      "epoch": 0.6950032446463336,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.6738296903575303e-06,
-      "loss": 1.1315,
-      "step": 714
-    },
-    {
-      "epoch": 0.6959766385463985,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.6704357404958303e-06,
-      "loss": 1.1401,
-      "step": 715
-    },
-    {
-      "epoch": 0.6969500324464634,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.6670390256957815e-06,
-      "loss": 1.1279,
-      "step": 716
-    },
-    {
-      "epoch": 0.6979234263465283,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.6636395539814975e-06,
-      "loss": 1.1354,
-      "step": 717
-    },
-    {
-      "epoch": 0.6988968202465932,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.6602373333836004e-06,
-      "loss": 1.1141,
-      "step": 718
-    },
-    {
-      "epoch": 0.6998702141466581,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.65683237193921e-06,
-      "loss": 1.1426,
-      "step": 719
-    },
-    {
-      "epoch": 0.7008436080467229,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.6534246776919193e-06,
-      "loss": 1.1411,
-      "step": 720
-    },
-    {
-      "epoch": 0.7018170019467878,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.650014258691774e-06,
-      "loss": 1.1404,
-      "step": 721
-    },
-    {
-      "epoch": 0.7027903958468527,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.646601122995263e-06,
-      "loss": 1.1167,
-      "step": 722
-    },
-    {
-      "epoch": 0.7037637897469176,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.6431852786652866e-06,
-      "loss": 1.1295,
-      "step": 723
-    },
-    {
-      "epoch": 0.7047371836469825,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.6397667337711475e-06,
-      "loss": 1.1363,
-      "step": 724
-    },
-    {
-      "epoch": 0.7057105775470474,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.636345496388527e-06,
-      "loss": 1.1246,
-      "step": 725
-    },
-    {
-      "epoch": 0.7066839714471123,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.6329215745994656e-06,
-      "loss": 1.1514,
-      "step": 726
-    },
-    {
-      "epoch": 0.7076573653471772,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.6294949764923478e-06,
-      "loss": 1.1427,
-      "step": 727
-    },
-    {
-      "epoch": 0.7086307592472421,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.6260657101618784e-06,
-      "loss": 1.1422,
-      "step": 728
-    },
-    {
-      "epoch": 0.709604153147307,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.622633783709066e-06,
-      "loss": 1.1331,
-      "step": 729
-    },
-    {
-      "epoch": 0.7105775470473719,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.619199205241204e-06,
-      "loss": 1.1175,
-      "step": 730
-    },
-    {
-      "epoch": 0.7115509409474368,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.6157619828718477e-06,
-      "loss": 1.1574,
-      "step": 731
-    },
-    {
-      "epoch": 0.7125243348475017,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.612322124720803e-06,
-      "loss": 1.1329,
-      "step": 732
-    },
-    {
-      "epoch": 0.7134977287475666,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.6088796389140984e-06,
-      "loss": 1.1303,
-      "step": 733
-    },
-    {
-      "epoch": 0.7144711226476315,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.605434533583971e-06,
-      "loss": 1.1111,
-      "step": 734
-    },
-    {
-      "epoch": 0.7154445165476963,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.601986816868846e-06,
-      "loss": 1.1348,
-      "step": 735
-    },
-    {
-      "epoch": 0.7164179104477612,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.5985364969133185e-06,
-      "loss": 1.1287,
-      "step": 736
-    },
-    {
-      "epoch": 0.717391304347826,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.595083581868132e-06,
-      "loss": 1.1319,
-      "step": 737
-    },
-    {
-      "epoch": 0.718364698247891,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.5916280798901604e-06,
-      "loss": 1.1394,
-      "step": 738
-    },
-    {
-      "epoch": 0.7193380921479559,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.5881699991423887e-06,
-      "loss": 1.1082,
-      "step": 739
-    },
-    {
-      "epoch": 0.7203114860480208,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.5847093477938955e-06,
-      "loss": 1.1251,
-      "step": 740
-    },
-    {
-      "epoch": 0.7212848799480857,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.5812461340198303e-06,
-      "loss": 1.1486,
-      "step": 741
-    },
-    {
-      "epoch": 0.7222582738481506,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.577780366001396e-06,
-      "loss": 1.1336,
-      "step": 742
-    },
-    {
-      "epoch": 0.7232316677482155,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.5743120519258295e-06,
-      "loss": 1.1346,
-      "step": 743
-    },
-    {
-      "epoch": 0.7242050616482804,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.5708411999863824e-06,
-      "loss": 1.1264,
-      "step": 744
-    },
-    {
-      "epoch": 0.7251784555483453,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.567367818382303e-06,
-      "loss": 1.1237,
-      "step": 745
-    },
-    {
-      "epoch": 0.7261518494484102,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.5638919153188125e-06,
-      "loss": 1.1146,
-      "step": 746
-    },
-    {
-      "epoch": 0.7271252433484751,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.56041349900709e-06,
-      "loss": 1.1083,
-      "step": 747
-    },
-    {
-      "epoch": 0.72809863724854,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.556932577664253e-06,
-      "loss": 1.1202,
-      "step": 748
-    },
-    {
-      "epoch": 0.7290720311486047,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.553449159513335e-06,
-      "loss": 1.1406,
-      "step": 749
-    },
-    {
-      "epoch": 0.7300454250486696,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.5499632527832683e-06,
-      "loss": 1.1324,
-      "step": 750
-    },
-    {
-      "epoch": 0.7310188189487346,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.5464748657088645e-06,
-      "loss": 1.1277,
-      "step": 751
-    },
-    {
-      "epoch": 0.7319922128487995,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.5429840065307924e-06,
-      "loss": 1.128,
-      "step": 752
-    },
-    {
-      "epoch": 0.7329656067488644,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.5394906834955646e-06,
-      "loss": 1.1373,
-      "step": 753
-    },
-    {
-      "epoch": 0.7339390006489293,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.535994904855509e-06,
-      "loss": 1.1305,
-      "step": 754
-    },
-    {
-      "epoch": 0.7349123945489942,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.53249667886876e-06,
-      "loss": 1.13,
-      "step": 755
-    },
-    {
-      "epoch": 0.735885788449059,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.5289960137992296e-06,
-      "loss": 1.1123,
-      "step": 756
-    },
-    {
-      "epoch": 0.736859182349124,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.525492917916592e-06,
-      "loss": 1.1485,
-      "step": 757
-    },
-    {
-      "epoch": 0.7378325762491889,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.521987399496266e-06,
-      "loss": 1.1465,
-      "step": 758
-    },
-    {
-      "epoch": 0.7388059701492538,
-      "grad_norm": 0.322265625,
-      "learning_rate": 3.5184794668193893e-06,
-      "loss": 1.1376,
-      "step": 759
-    },
-    {
-      "epoch": 0.7397793640493187,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.5149691281728083e-06,
-      "loss": 1.116,
-      "step": 760
-    },
-    {
-      "epoch": 0.7407527579493836,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.5114563918490473e-06,
-      "loss": 1.1217,
-      "step": 761
-    },
-    {
-      "epoch": 0.7417261518494485,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.507941266146299e-06,
-      "loss": 1.1159,
-      "step": 762
-    },
-    {
-      "epoch": 0.7426995457495134,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.5044237593683977e-06,
-      "loss": 1.1278,
-      "step": 763
-    },
-    {
-      "epoch": 0.7436729396495781,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.5009038798248047e-06,
-      "loss": 1.1419,
-      "step": 764
-    },
-    {
-      "epoch": 0.744646333549643,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.497381635830585e-06,
-      "loss": 1.1304,
-      "step": 765
-    },
-    {
-      "epoch": 0.745619727449708,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.4938570357063906e-06,
-      "loss": 1.1165,
-      "step": 766
-    },
-    {
-      "epoch": 0.7465931213497728,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.4903300877784362e-06,
-      "loss": 1.1412,
-      "step": 767
-    },
-    {
-      "epoch": 0.7475665152498377,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.486800800378489e-06,
-      "loss": 1.104,
-      "step": 768
-    },
-    {
-      "epoch": 0.7485399091499026,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.483269181843835e-06,
-      "loss": 1.127,
-      "step": 769
-    },
-    {
-      "epoch": 0.7495133030499675,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.4797352405172735e-06,
-      "loss": 1.1248,
-      "step": 770
-    },
-    {
-      "epoch": 0.7504866969500325,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.4761989847470877e-06,
-      "loss": 1.1267,
-      "step": 771
-    },
-    {
-      "epoch": 0.7514600908500974,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.4726604228870286e-06,
-      "loss": 1.1258,
-      "step": 772
-    },
-    {
-      "epoch": 0.7524334847501623,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.469119563296296e-06,
-      "loss": 1.1354,
-      "step": 773
-    },
-    {
-      "epoch": 0.7534068786502272,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.4655764143395163e-06,
-      "loss": 1.1328,
-      "step": 774
-    },
-    {
-      "epoch": 0.754380272550292,
-      "grad_norm": 0.29296875,
-      "learning_rate": 3.462030984386725e-06,
-      "loss": 1.1262,
-      "step": 775
-    },
-    {
-      "epoch": 0.755353666450357,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.4584832818133463e-06,
-      "loss": 1.1316,
-      "step": 776
-    },
-    {
-      "epoch": 0.7563270603504219,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.454933315000172e-06,
-      "loss": 1.1305,
-      "step": 777
-    },
-    {
-      "epoch": 0.7573004542504866,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.4513810923333436e-06,
-      "loss": 1.1495,
-      "step": 778
-    },
-    {
-      "epoch": 0.7582738481505515,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.44782662220433e-06,
-      "loss": 1.1516,
-      "step": 779
-    },
-    {
-      "epoch": 0.7592472420506164,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.444269913009912e-06,
-      "loss": 1.1053,
-      "step": 780
-    },
-    {
-      "epoch": 0.7602206359506813,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.440710973152159e-06,
-      "loss": 1.1105,
-      "step": 781
-    },
-    {
-      "epoch": 0.7611940298507462,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.4371498110384083e-06,
-      "loss": 1.144,
-      "step": 782
-    },
-    {
-      "epoch": 0.7621674237508111,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.4335864350812475e-06,
-      "loss": 1.1367,
-      "step": 783
-    },
-    {
-      "epoch": 0.763140817650876,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.4300208536984963e-06,
-      "loss": 1.126,
-      "step": 784
-    },
-    {
-      "epoch": 0.764114211550941,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.42645307531318e-06,
-      "loss": 1.1217,
-      "step": 785
-    },
-    {
-      "epoch": 0.7650876054510058,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.4228831083535185e-06,
-      "loss": 1.1297,
-      "step": 786
-    },
-    {
-      "epoch": 0.7660609993510707,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.4193109612528972e-06,
-      "loss": 1.1295,
-      "step": 787
-    },
-    {
-      "epoch": 0.7670343932511356,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.4157366424498574e-06,
-      "loss": 1.1299,
-      "step": 788
-    },
-    {
-      "epoch": 0.7680077871512005,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.4121601603880648e-06,
-      "loss": 1.1165,
-      "step": 789
-    },
-    {
-      "epoch": 0.7689811810512654,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.408581523516299e-06,
-      "loss": 1.1152,
-      "step": 790
-    },
-    {
-      "epoch": 0.7699545749513304,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.4050007402884276e-06,
-      "loss": 1.1487,
-      "step": 791
-    },
-    {
-      "epoch": 0.7709279688513953,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.401417819163392e-06,
-      "loss": 1.1263,
-      "step": 792
-    },
-    {
-      "epoch": 0.77190136275146,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.397832768605181e-06,
-      "loss": 1.1111,
-      "step": 793
-    },
-    {
-      "epoch": 0.7728747566515249,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.3942455970828146e-06,
-      "loss": 1.1163,
-      "step": 794
-    },
-    {
-      "epoch": 0.7738481505515898,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.3906563130703218e-06,
-      "loss": 1.124,
-      "step": 795
-    },
-    {
-      "epoch": 0.7748215444516547,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.3870649250467256e-06,
-      "loss": 1.123,
-      "step": 796
-    },
-    {
-      "epoch": 0.7757949383517196,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.383471441496016e-06,
-      "loss": 1.1221,
-      "step": 797
-    },
-    {
-      "epoch": 0.7767683322517845,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.3798758709071354e-06,
-      "loss": 1.1252,
-      "step": 798
-    },
-    {
-      "epoch": 0.7777417261518494,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.376278221773954e-06,
-      "loss": 1.1328,
-      "step": 799
-    },
-    {
-      "epoch": 0.7787151200519143,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.372678502595255e-06,
-      "loss": 1.1363,
-      "step": 800
-    },
-    {
-      "epoch": 0.7796885139519792,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.3690767218747104e-06,
-      "loss": 1.1205,
-      "step": 801
-    },
-    {
-      "epoch": 0.7806619078520441,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.3654728881208607e-06,
-      "loss": 1.1225,
-      "step": 802
-    },
-    {
-      "epoch": 0.781635301752109,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.3618670098470995e-06,
-      "loss": 1.1284,
-      "step": 803
-    },
-    {
-      "epoch": 0.782608695652174,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.3582590955716478e-06,
-      "loss": 1.1041,
-      "step": 804
-    },
-    {
-      "epoch": 0.7835820895522388,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.3546491538175362e-06,
-      "loss": 1.1102,
-      "step": 805
-    },
-    {
-      "epoch": 0.7845554834523037,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.3510371931125875e-06,
-      "loss": 1.1102,
-      "step": 806
-    },
-    {
-      "epoch": 0.7855288773523685,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.3474232219893893e-06,
-      "loss": 1.1212,
-      "step": 807
-    },
-    {
-      "epoch": 0.7865022712524334,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.3438072489852837e-06,
-      "loss": 1.1562,
-      "step": 808
-    },
-    {
-      "epoch": 0.7874756651524983,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.3401892826423382e-06,
-      "loss": 1.1242,
-      "step": 809
-    },
-    {
-      "epoch": 0.7884490590525632,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.33656933150733e-06,
-      "loss": 1.1147,
-      "step": 810
-    },
-    {
-      "epoch": 0.7894224529526281,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.3329474041317262e-06,
-      "loss": 1.1366,
-      "step": 811
-    },
-    {
-      "epoch": 0.790395846852693,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.329323509071661e-06,
-      "loss": 1.1191,
-      "step": 812
-    },
-    {
-      "epoch": 0.7913692407527579,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.3256976548879183e-06,
-      "loss": 1.1141,
-      "step": 813
-    },
-    {
-      "epoch": 0.7923426346528228,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.3220698501459082e-06,
-      "loss": 1.1154,
-      "step": 814
-    },
-    {
-      "epoch": 0.7933160285528877,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.318440103415649e-06,
-      "loss": 1.1292,
-      "step": 815
-    },
-    {
-      "epoch": 0.7942894224529526,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.3148084232717487e-06,
-      "loss": 1.1255,
-      "step": 816
-    },
-    {
-      "epoch": 0.7952628163530175,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.3111748182933805e-06,
-      "loss": 1.116,
-      "step": 817
-    },
-    {
-      "epoch": 0.7962362102530824,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.3075392970642655e-06,
-      "loss": 1.1381,
-      "step": 818
-    },
-    {
-      "epoch": 0.7972096041531473,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.303901868172651e-06,
-      "loss": 1.1258,
-      "step": 819
-    },
-    {
-      "epoch": 0.7981829980532122,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.3002625402112916e-06,
-      "loss": 1.1386,
-      "step": 820
-    },
-    {
-      "epoch": 0.7991563919532771,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.296621321777428e-06,
-      "loss": 1.1484,
-      "step": 821
-    },
-    {
-      "epoch": 0.8001297858533419,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.2929782214727657e-06,
-      "loss": 1.1307,
-      "step": 822
-    },
-    {
-      "epoch": 0.8011031797534068,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.2893332479034564e-06,
-      "loss": 1.121,
-      "step": 823
-    },
-    {
-      "epoch": 0.8020765736534717,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.285686409680079e-06,
-      "loss": 1.136,
-      "step": 824
-    },
-    {
-      "epoch": 0.8030499675535366,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.282037715417613e-06,
-      "loss": 1.1244,
-      "step": 825
-    },
-    {
-      "epoch": 0.8040233614536015,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.2783871737354272e-06,
-      "loss": 1.1264,
-      "step": 826
-    },
-    {
-      "epoch": 0.8049967553536664,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.274734793257251e-06,
-      "loss": 1.1109,
-      "step": 827
-    },
-    {
-      "epoch": 0.8059701492537313,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.2710805826111592e-06,
-      "loss": 1.1233,
-      "step": 828
-    },
-    {
-      "epoch": 0.8069435431537962,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.2674245504295505e-06,
-      "loss": 1.1178,
-      "step": 829
-    },
-    {
-      "epoch": 0.8079169370538611,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.263766705349125e-06,
-      "loss": 1.1146,
-      "step": 830
-    },
-    {
-      "epoch": 0.808890330953926,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.2601070560108667e-06,
-      "loss": 1.1171,
-      "step": 831
-    },
-    {
-      "epoch": 0.8098637248539909,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.256445611060022e-06,
-      "loss": 1.1245,
-      "step": 832
-    },
-    {
-      "epoch": 0.8108371187540558,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.2527823791460778e-06,
-      "loss": 1.1076,
-      "step": 833
-    },
-    {
-      "epoch": 0.8118105126541207,
-      "grad_norm": 0.38671875,
-      "learning_rate": 3.249117368922744e-06,
-      "loss": 1.1124,
-      "step": 834
-    },
-    {
-      "epoch": 0.8127839065541856,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.2454505890479297e-06,
-      "loss": 1.1258,
-      "step": 835
-    },
-    {
-      "epoch": 0.8137573004542504,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.241782048183726e-06,
-      "loss": 1.1111,
-      "step": 836
-    },
-    {
-      "epoch": 0.8147306943543153,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.2381117549963835e-06,
-      "loss": 1.1168,
-      "step": 837
-    },
-    {
-      "epoch": 0.8157040882543802,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.234439718156292e-06,
-      "loss": 1.1316,
-      "step": 838
-    },
-    {
-      "epoch": 0.8166774821544451,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.23076594633796e-06,
-      "loss": 1.1405,
-      "step": 839
-    },
-    {
-      "epoch": 0.81765087605451,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.2270904482199965e-06,
-      "loss": 1.131,
-      "step": 840
-    },
-    {
-      "epoch": 0.8186242699545749,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.2234132324850865e-06,
-      "loss": 1.1048,
-      "step": 841
-    },
-    {
-      "epoch": 0.8195976638546398,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.2197343078199735e-06,
-      "loss": 1.1325,
-      "step": 842
-    },
-    {
-      "epoch": 0.8205710577547047,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.216053682915436e-06,
-      "loss": 1.1423,
-      "step": 843
-    },
-    {
-      "epoch": 0.8215444516547696,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.2123713664662735e-06,
-      "loss": 1.1322,
-      "step": 844
-    },
-    {
-      "epoch": 0.8225178455548345,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.208687367171278e-06,
-      "loss": 1.1316,
-      "step": 845
-    },
-    {
-      "epoch": 0.8234912394548994,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.2050016937332166e-06,
-      "loss": 1.1327,
-      "step": 846
-    },
-    {
-      "epoch": 0.8244646333549643,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.2013143548588134e-06,
-      "loss": 1.1243,
-      "step": 847
-    },
-    {
-      "epoch": 0.8254380272550292,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.1976253592587252e-06,
-      "loss": 1.107,
-      "step": 848
-    },
-    {
-      "epoch": 0.8264114211550941,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.1939347156475244e-06,
-      "loss": 1.1293,
-      "step": 849
-    },
-    {
-      "epoch": 0.827384815055159,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.190242432743673e-06,
-      "loss": 1.1217,
-      "step": 850
-    },
-    {
-      "epoch": 0.8283582089552238,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.186548519269508e-06,
-      "loss": 1.1277,
-      "step": 851
-    },
-    {
-      "epoch": 0.8293316028552887,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.1828529839512196e-06,
-      "loss": 1.1332,
-      "step": 852
-    },
-    {
-      "epoch": 0.8303049967553536,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.1791558355188263e-06,
-      "loss": 1.1258,
-      "step": 853
-    },
-    {
-      "epoch": 0.8312783906554185,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.17545708270616e-06,
-      "loss": 1.1098,
-      "step": 854
-    },
-    {
-      "epoch": 0.8322517845554834,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.1717567342508393e-06,
-      "loss": 1.1232,
-      "step": 855
-    },
-    {
-      "epoch": 0.8332251784555483,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.1680547988942562e-06,
-      "loss": 1.1292,
-      "step": 856
-    },
-    {
-      "epoch": 0.8341985723556132,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.164351285381549e-06,
-      "loss": 1.126,
-      "step": 857
-    },
-    {
-      "epoch": 0.8351719662556781,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.1606462024615838e-06,
-      "loss": 1.1292,
-      "step": 858
-    },
-    {
-      "epoch": 0.836145360155743,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.1569395588869358e-06,
-      "loss": 1.0986,
-      "step": 859
-    },
-    {
-      "epoch": 0.8371187540558079,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.1532313634138663e-06,
-      "loss": 1.13,
-      "step": 860
-    },
-    {
-      "epoch": 0.8380921479558728,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.1495216248023013e-06,
-      "loss": 1.1238,
-      "step": 861
-    },
-    {
-      "epoch": 0.8390655418559377,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.145810351815815e-06,
-      "loss": 1.1416,
-      "step": 862
-    },
-    {
-      "epoch": 0.8400389357560026,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.1420975532216036e-06,
-      "loss": 1.1169,
-      "step": 863
-    },
-    {
-      "epoch": 0.8410123296560675,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.1383832377904676e-06,
-      "loss": 1.1209,
-      "step": 864
-    },
-    {
-      "epoch": 0.8419857235561323,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.1346674142967936e-06,
-      "loss": 1.1124,
-      "step": 865
-    },
-    {
-      "epoch": 0.8429591174561972,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.1309500915185265e-06,
-      "loss": 1.1155,
-      "step": 866
-    },
-    {
-      "epoch": 0.8439325113562621,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.127231278237156e-06,
-      "loss": 1.1183,
-      "step": 867
-    },
-    {
-      "epoch": 0.844905905256327,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.1235109832376927e-06,
-      "loss": 1.1083,
-      "step": 868
-    },
-    {
-      "epoch": 0.8458792991563919,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.1197892153086445e-06,
-      "loss": 1.1132,
-      "step": 869
-    },
-    {
-      "epoch": 0.8468526930564568,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.1160659832420033e-06,
-      "loss": 1.1273,
-      "step": 870
-    },
-    {
-      "epoch": 0.8478260869565217,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.1123412958332155e-06,
-      "loss": 1.1245,
-      "step": 871
-    },
-    {
-      "epoch": 0.8487994808565866,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.108615161881169e-06,
-      "loss": 1.1176,
-      "step": 872
-    },
-    {
-      "epoch": 0.8497728747566515,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.104887590188167e-06,
-      "loss": 1.1051,
-      "step": 873
-    },
-    {
-      "epoch": 0.8507462686567164,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.1011585895599083e-06,
-      "loss": 1.134,
-      "step": 874
-    },
-    {
-      "epoch": 0.8517196625567813,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.0974281688054698e-06,
-      "loss": 1.1388,
-      "step": 875
-    },
-    {
-      "epoch": 0.8526930564568462,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.093696336737281e-06,
-      "loss": 1.1202,
-      "step": 876
-    },
-    {
-      "epoch": 0.8536664503569111,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.0899631021711073e-06,
-      "loss": 1.1155,
-      "step": 877
-    },
-    {
-      "epoch": 0.854639844256976,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.0862284739260247e-06,
-      "loss": 1.1065,
-      "step": 878
-    },
-    {
-      "epoch": 0.8556132381570409,
-      "grad_norm": 0.3203125,
-      "learning_rate": 3.0824924608244035e-06,
-      "loss": 1.1248,
-      "step": 879
-    },
-    {
-      "epoch": 0.8565866320571057,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.078755071691885e-06,
-      "loss": 1.1208,
-      "step": 880
-    },
-    {
-      "epoch": 0.8575600259571706,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.075016315357361e-06,
-      "loss": 1.1144,
-      "step": 881
-    },
-    {
-      "epoch": 0.8585334198572355,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.0712762006529533e-06,
-      "loss": 1.1423,
-      "step": 882
-    },
-    {
-      "epoch": 0.8595068137573004,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.0675347364139917e-06,
-      "loss": 1.1107,
-      "step": 883
-    },
-    {
-      "epoch": 0.8604802076573653,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.0637919314789953e-06,
-      "loss": 1.1298,
-      "step": 884
-    },
-    {
-      "epoch": 0.8614536015574302,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.0600477946896494e-06,
-      "loss": 1.1119,
-      "step": 885
-    },
-    {
-      "epoch": 0.8624269954574951,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.056302334890786e-06,
-      "loss": 1.1101,
-      "step": 886
-    },
-    {
-      "epoch": 0.86340038935756,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.052555560930363e-06,
-      "loss": 1.1208,
-      "step": 887
-    },
-    {
-      "epoch": 0.8643737832576249,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.0488074816594417e-06,
-      "loss": 1.1325,
-      "step": 888
-    },
-    {
-      "epoch": 0.8653471771576898,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.0450581059321664e-06,
-      "loss": 1.1362,
-      "step": 889
-    },
-    {
-      "epoch": 0.8663205710577547,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.0413074426057472e-06,
-      "loss": 1.1273,
-      "step": 890
-    },
-    {
-      "epoch": 0.8672939649578196,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.037555500540431e-06,
-      "loss": 1.1199,
-      "step": 891
-    },
-    {
-      "epoch": 0.8682673588578845,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.0338022885994904e-06,
-      "loss": 1.1174,
-      "step": 892
-    },
-    {
-      "epoch": 0.8692407527579494,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.0300478156491954e-06,
-      "loss": 1.1006,
-      "step": 893
-    },
-    {
-      "epoch": 0.8702141466580143,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.0262920905587945e-06,
-      "loss": 1.126,
-      "step": 894
-    },
-    {
-      "epoch": 0.8711875405580791,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.0225351222004944e-06,
-      "loss": 1.1314,
-      "step": 895
-    },
-    {
-      "epoch": 0.872160934458144,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.018776919449441e-06,
-      "loss": 1.1382,
-      "step": 896
-    },
-    {
-      "epoch": 0.8731343283582089,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.0150174911836926e-06,
-      "loss": 1.1146,
-      "step": 897
-    },
-    {
-      "epoch": 0.8741077222582738,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.011256846284206e-06,
-      "loss": 1.1414,
-      "step": 898
-    },
-    {
-      "epoch": 0.8750811161583387,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.0074949936348084e-06,
-      "loss": 1.1139,
-      "step": 899
-    },
-    {
-      "epoch": 0.8760545100584036,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.003731942122185e-06,
-      "loss": 1.1114,
-      "step": 900
-    },
-    {
-      "epoch": 0.8770279039584685,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.999967700635848e-06,
-      "loss": 1.1207,
-      "step": 901
-    },
-    {
-      "epoch": 0.8780012978585334,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.9962022780681237e-06,
-      "loss": 1.1315,
-      "step": 902
-    },
-    {
-      "epoch": 0.8789746917585983,
-      "grad_norm": 0.310546875,
-      "learning_rate": 2.9924356833141276e-06,
-      "loss": 1.1293,
-      "step": 903
-    },
-    {
-      "epoch": 0.8799480856586632,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.988667925271744e-06,
-      "loss": 1.0981,
-      "step": 904
-    },
-    {
-      "epoch": 0.8809214795587281,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.984899012841607e-06,
-      "loss": 1.1355,
-      "step": 905
-    },
-    {
-      "epoch": 0.881894873458793,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.981128954927075e-06,
-      "loss": 1.1445,
-      "step": 906
-    },
-    {
-      "epoch": 0.8828682673588579,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.9773577604342147e-06,
-      "loss": 1.1198,
-      "step": 907
-    },
-    {
-      "epoch": 0.8838416612589228,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.9735854382717766e-06,
-      "loss": 1.139,
-      "step": 908
-    },
-    {
-      "epoch": 0.8848150551589876,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.969811997351174e-06,
-      "loss": 1.1192,
-      "step": 909
-    },
-    {
-      "epoch": 0.8857884490590525,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.966037446586467e-06,
-      "loss": 1.133,
-      "step": 910
-    },
-    {
-      "epoch": 0.8867618429591174,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.9622617948943326e-06,
-      "loss": 1.1156,
-      "step": 911
-    },
-    {
-      "epoch": 0.8877352368591823,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.958485051194052e-06,
-      "loss": 1.1103,
-      "step": 912
-    },
-    {
-      "epoch": 0.8887086307592472,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.9547072244074853e-06,
-      "loss": 1.1136,
-      "step": 913
-    },
-    {
-      "epoch": 0.8896820246593121,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.95092832345905e-06,
-      "loss": 1.1204,
-      "step": 914
-    },
-    {
-      "epoch": 0.890655418559377,
-      "grad_norm": 0.32421875,
-      "learning_rate": 2.9471483572757013e-06,
-      "loss": 1.1176,
-      "step": 915
-    },
-    {
-      "epoch": 0.8916288124594419,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.9433673347869134e-06,
-      "loss": 1.1384,
-      "step": 916
-    },
-    {
-      "epoch": 0.8926022063595068,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.939585264924652e-06,
-      "loss": 1.1201,
-      "step": 917
-    },
-    {
-      "epoch": 0.8935756002595717,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.93580215662336e-06,
-      "loss": 1.1105,
-      "step": 918
-    },
-    {
-      "epoch": 0.8945489941596366,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.932018018819932e-06,
-      "loss": 1.1264,
-      "step": 919
-    },
-    {
-      "epoch": 0.8955223880597015,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.928232860453694e-06,
-      "loss": 1.1095,
-      "step": 920
-    },
-    {
-      "epoch": 0.8964957819597664,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.924446690466386e-06,
-      "loss": 1.1245,
-      "step": 921
-    },
-    {
-      "epoch": 0.8974691758598313,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.9206595178021337e-06,
-      "loss": 1.1027,
-      "step": 922
-    },
-    {
-      "epoch": 0.8984425697598962,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.9168713514074334e-06,
-      "loss": 1.1309,
-      "step": 923
-    },
-    {
-      "epoch": 0.899415963659961,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.9130822002311294e-06,
-      "loss": 1.0955,
-      "step": 924
-    },
-    {
-      "epoch": 0.9003893575600259,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.9092920732243923e-06,
-      "loss": 1.1209,
-      "step": 925
-    },
-    {
-      "epoch": 0.9013627514600908,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.9055009793406972e-06,
-      "loss": 1.1147,
-      "step": 926
-    },
-    {
-      "epoch": 0.9023361453601557,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.9017089275358017e-06,
-      "loss": 1.1211,
-      "step": 927
-    },
-    {
-      "epoch": 0.9033095392602206,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.8979159267677305e-06,
-      "loss": 1.121,
-      "step": 928
-    },
-    {
-      "epoch": 0.9042829331602855,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.894121985996746e-06,
-      "loss": 1.1152,
-      "step": 929
-    },
-    {
-      "epoch": 0.9052563270603504,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.8903271141853346e-06,
-      "loss": 1.1092,
-      "step": 930
-    },
-    {
-      "epoch": 0.9062297209604153,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.8865313202981784e-06,
-      "loss": 1.126,
-      "step": 931
-    },
-    {
-      "epoch": 0.9072031148604802,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.88273461330214e-06,
-      "loss": 1.1055,
-      "step": 932
-    },
-    {
-      "epoch": 0.9081765087605451,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.87893700216624e-06,
-      "loss": 1.1337,
-      "step": 933
-    },
-    {
-      "epoch": 0.90914990266061,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.8751384958616318e-06,
-      "loss": 1.123,
-      "step": 934
-    },
-    {
-      "epoch": 0.9101232965606749,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.8713391033615865e-06,
-      "loss": 1.1202,
-      "step": 935
-    },
-    {
-      "epoch": 0.9110966904607398,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.8675388336414676e-06,
-      "loss": 1.1279,
-      "step": 936
-    },
-    {
-      "epoch": 0.9120700843608047,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.8637376956787104e-06,
-      "loss": 1.1358,
-      "step": 937
-    },
-    {
-      "epoch": 0.9130434782608695,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.8599356984528014e-06,
-      "loss": 1.1205,
-      "step": 938
-    },
-    {
-      "epoch": 0.9140168721609344,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.856132850945257e-06,
-      "loss": 1.1075,
-      "step": 939
-    },
-    {
-      "epoch": 0.9149902660609993,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.8523291621396015e-06,
-      "loss": 1.1236,
-      "step": 940
-    },
-    {
-      "epoch": 0.9159636599610642,
-      "grad_norm": 0.310546875,
-      "learning_rate": 2.8485246410213497e-06,
-      "loss": 1.1339,
-      "step": 941
-    },
-    {
-      "epoch": 0.9169370538611291,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.8447192965779778e-06,
-      "loss": 1.1243,
-      "step": 942
-    },
-    {
-      "epoch": 0.917910447761194,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.840913137798911e-06,
-      "loss": 1.1191,
-      "step": 943
-    },
-    {
-      "epoch": 0.9188838416612589,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.837106173675498e-06,
-      "loss": 1.1065,
-      "step": 944
-    },
-    {
-      "epoch": 0.9198572355613238,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.8332984132009854e-06,
-      "loss": 1.1333,
-      "step": 945
-    },
-    {
-      "epoch": 0.9208306294613887,
-      "grad_norm": 0.3125,
-      "learning_rate": 2.829489865370507e-06,
-      "loss": 1.1357,
-      "step": 946
-    },
-    {
-      "epoch": 0.9218040233614536,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.825680539181052e-06,
-      "loss": 1.1343,
-      "step": 947
-    },
-    {
-      "epoch": 0.9227774172615185,
-      "grad_norm": 0.310546875,
-      "learning_rate": 2.8218704436314525e-06,
-      "loss": 1.1118,
-      "step": 948
-    },
-    {
-      "epoch": 0.9237508111615834,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.8180595877223537e-06,
-      "loss": 1.1107,
-      "step": 949
-    },
-    {
-      "epoch": 0.9247242050616483,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.8142479804562e-06,
-      "loss": 1.1174,
-      "step": 950
-    },
-    {
-      "epoch": 0.9256975989617132,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.81043563083721e-06,
-      "loss": 1.1143,
-      "step": 951
-    },
-    {
-      "epoch": 0.9266709928617781,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.8066225478713556e-06,
-      "loss": 1.1309,
-      "step": 952
-    },
-    {
-      "epoch": 0.9276443867618429,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.802808740566341e-06,
-      "loss": 1.1266,
-      "step": 953
-    },
-    {
-      "epoch": 0.9286177806619078,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.7989942179315832e-06,
-      "loss": 1.1376,
-      "step": 954
-    },
-    {
-      "epoch": 0.9295911745619727,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.795178988978185e-06,
-      "loss": 1.1132,
-      "step": 955
-    },
-    {
-      "epoch": 0.9305645684620376,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.791363062718923e-06,
-      "loss": 1.1357,
-      "step": 956
-    },
-    {
-      "epoch": 0.9315379623621025,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.7875464481682173e-06,
-      "loss": 1.1072,
-      "step": 957
-    },
-    {
-      "epoch": 0.9325113562621674,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.7837291543421164e-06,
-      "loss": 1.1247,
-      "step": 958
-    },
-    {
-      "epoch": 0.9334847501622323,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.7799111902582697e-06,
-      "loss": 1.1064,
-      "step": 959
-    },
-    {
-      "epoch": 0.9344581440622972,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.7760925649359148e-06,
-      "loss": 1.1073,
-      "step": 960
-    },
-    {
-      "epoch": 0.9354315379623621,
-      "grad_norm": 0.3125,
-      "learning_rate": 2.7722732873958485e-06,
-      "loss": 1.1195,
-      "step": 961
-    },
-    {
-      "epoch": 0.936404931862427,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.768453366660408e-06,
-      "loss": 1.1207,
-      "step": 962
-    },
-    {
-      "epoch": 0.9373783257624919,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.764632811753451e-06,
-      "loss": 1.0924,
-      "step": 963
-    },
-    {
-      "epoch": 0.9383517196625568,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.7608116317003347e-06,
-      "loss": 1.1248,
-      "step": 964
-    },
-    {
-      "epoch": 0.9393251135626217,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.756989835527889e-06,
-      "loss": 1.1071,
-      "step": 965
-    },
-    {
-      "epoch": 0.9402985074626866,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.753167432264404e-06,
-      "loss": 1.1171,
-      "step": 966
-    },
-    {
-      "epoch": 0.9412719013627514,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.7493444309396e-06,
-      "loss": 1.1064,
-      "step": 967
-    },
-    {
-      "epoch": 0.9422452952628163,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.745520840584613e-06,
-      "loss": 1.1149,
-      "step": 968
-    },
-    {
-      "epoch": 0.9432186891628812,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.741696670231969e-06,
-      "loss": 1.1323,
-      "step": 969
-    },
-    {
-      "epoch": 0.9441920830629461,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.7378719289155637e-06,
-      "loss": 1.1234,
-      "step": 970
-    },
-    {
-      "epoch": 0.945165476963011,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.7340466256706426e-06,
-      "loss": 1.1273,
-      "step": 971
-    },
-    {
-      "epoch": 0.9461388708630759,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.7302207695337794e-06,
-      "loss": 1.0925,
-      "step": 972
-    },
-    {
-      "epoch": 0.9471122647631408,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.726394369542852e-06,
-      "loss": 1.1152,
-      "step": 973
-    },
-    {
-      "epoch": 0.9480856586632057,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.7225674347370248e-06,
-      "loss": 1.1286,
-      "step": 974
-    },
-    {
-      "epoch": 0.9490590525632706,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.718739974156722e-06,
-      "loss": 1.1087,
-      "step": 975
-    },
-    {
-      "epoch": 0.9500324464633355,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.714911996843617e-06,
-      "loss": 1.1025,
-      "step": 976
-    },
-    {
-      "epoch": 0.9510058403634004,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.711083511840596e-06,
-      "loss": 1.1089,
-      "step": 977
-    },
-    {
-      "epoch": 0.9519792342634653,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.707254528191749e-06,
-      "loss": 1.1162,
-      "step": 978
-    },
-    {
-      "epoch": 0.9529526281635302,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.7034250549423422e-06,
-      "loss": 1.1187,
-      "step": 979
-    },
-    {
-      "epoch": 0.9539260220635951,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.699595101138801e-06,
-      "loss": 1.1094,
-      "step": 980
-    },
-    {
-      "epoch": 0.95489941596366,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.6957646758286836e-06,
-      "loss": 1.1091,
-      "step": 981
-    },
-    {
-      "epoch": 0.9558728098637248,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.691933788060662e-06,
-      "loss": 1.1277,
-      "step": 982
-    },
-    {
-      "epoch": 0.9568462037637897,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.6881024468845e-06,
-      "loss": 1.1209,
-      "step": 983
-    },
-    {
-      "epoch": 0.9578195976638546,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.684270661351037e-06,
-      "loss": 1.1478,
-      "step": 984
-    },
-    {
-      "epoch": 0.9587929915639195,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.680438440512156e-06,
-      "loss": 1.1162,
-      "step": 985
-    },
-    {
-      "epoch": 0.9597663854639844,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.6766057934207734e-06,
-      "loss": 1.1019,
-      "step": 986
-    },
-    {
-      "epoch": 0.9607397793640493,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.6727727291308083e-06,
-      "loss": 1.1186,
-      "step": 987
-    },
-    {
-      "epoch": 0.9617131732641142,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.6689392566971683e-06,
-      "loss": 1.1224,
-      "step": 988
-    },
-    {
-      "epoch": 0.9626865671641791,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.665105385175725e-06,
-      "loss": 1.1273,
-      "step": 989
-    },
-    {
-      "epoch": 0.963659961064244,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.6612711236232915e-06,
-      "loss": 1.1151,
-      "step": 990
-    },
-    {
-      "epoch": 0.9646333549643089,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.6574364810976018e-06,
-      "loss": 1.104,
-      "step": 991
-    },
-    {
-      "epoch": 0.9656067488643738,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.653601466657293e-06,
-      "loss": 1.1121,
-      "step": 992
-    },
-    {
-      "epoch": 0.9665801427644387,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.6497660893618763e-06,
-      "loss": 1.1205,
-      "step": 993
-    },
-    {
-      "epoch": 0.9675535366645036,
-      "grad_norm": 0.341796875,
-      "learning_rate": 2.645930358271725e-06,
-      "loss": 1.1086,
-      "step": 994
-    },
-    {
-      "epoch": 0.9685269305645685,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.642094282448044e-06,
-      "loss": 1.1169,
-      "step": 995
-    },
-    {
-      "epoch": 0.9695003244646333,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.638257870952854e-06,
-      "loss": 1.1119,
-      "step": 996
-    },
-    {
-      "epoch": 0.9704737183646982,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.63442113284897e-06,
-      "loss": 1.1153,
-      "step": 997
-    },
-    {
-      "epoch": 0.9714471122647631,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.6305840771999764e-06,
-      "loss": 1.1103,
-      "step": 998
-    },
-    {
-      "epoch": 0.972420506164828,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.626746713070209e-06,
-      "loss": 1.1152,
-      "step": 999
-    },
-    {
-      "epoch": 0.9733939000648929,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.6229090495247335e-06,
-      "loss": 1.1115,
-      "step": 1000
-    },
-    {
-      "epoch": 0.9743672939649578,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.6190710956293187e-06,
-      "loss": 1.1288,
-      "step": 1001
-    },
-    {
-      "epoch": 0.9753406878650227,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.615232860450425e-06,
-      "loss": 1.1066,
-      "step": 1002
-    },
-    {
-      "epoch": 0.9763140817650876,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.6113943530551718e-06,
-      "loss": 1.1126,
-      "step": 1003
-    },
-    {
-      "epoch": 0.9772874756651525,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.6075555825113265e-06,
-      "loss": 1.1091,
-      "step": 1004
-    },
-    {
-      "epoch": 0.9782608695652174,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.6037165578872746e-06,
-      "loss": 1.1152,
-      "step": 1005
-    },
-    {
-      "epoch": 0.9792342634652823,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.599877288252003e-06,
-      "loss": 1.1319,
-      "step": 1006
-    },
-    {
-      "epoch": 0.9802076573653472,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.596037782675077e-06,
-      "loss": 1.1063,
-      "step": 1007
-    },
-    {
-      "epoch": 0.9811810512654121,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.592198050226621e-06,
-      "loss": 1.081,
-      "step": 1008
-    },
-    {
-      "epoch": 0.982154445165477,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.588358099977293e-06,
-      "loss": 1.1398,
-      "step": 1009
-    },
-    {
-      "epoch": 0.9831278390655419,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.5845179409982667e-06,
-      "loss": 1.1272,
-      "step": 1010
-    },
-    {
-      "epoch": 0.9841012329656067,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.580677582361208e-06,
-      "loss": 1.104,
-      "step": 1011
-    },
-    {
-      "epoch": 0.9850746268656716,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.5768370331382566e-06,
-      "loss": 1.1011,
-      "step": 1012
-    },
-    {
-      "epoch": 0.9860480207657365,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.572996302402e-06,
-      "loss": 1.1139,
-      "step": 1013
-    },
-    {
-      "epoch": 0.9870214146658014,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.569155399225456e-06,
-      "loss": 1.1077,
-      "step": 1014
-    },
-    {
-      "epoch": 0.9879948085658663,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.565314332682049e-06,
-      "loss": 1.1053,
-      "step": 1015
-    },
-    {
-      "epoch": 0.9889682024659312,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.5614731118455887e-06,
-      "loss": 1.0897,
-      "step": 1016
-    },
-    {
-      "epoch": 0.9899415963659961,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.5576317457902522e-06,
-      "loss": 1.1306,
-      "step": 1017
-    },
-    {
-      "epoch": 0.990914990266061,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.553790243590556e-06,
-      "loss": 1.1,
-      "step": 1018
-    },
-    {
-      "epoch": 0.9918883841661259,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.5499486143213396e-06,
-      "loss": 1.1075,
-      "step": 1019
-    },
-    {
-      "epoch": 0.9928617780661908,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.5461068670577445e-06,
-      "loss": 1.0963,
-      "step": 1020
-    },
-    {
-      "epoch": 0.9938351719662557,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.542265010875188e-06,
-      "loss": 1.128,
-      "step": 1021
-    },
-    {
-      "epoch": 0.9948085658663206,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.5384230548493467e-06,
-      "loss": 1.1246,
-      "step": 1022
-    },
-    {
-      "epoch": 0.9957819597663855,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.534581008056131e-06,
-      "loss": 1.1078,
-      "step": 1023
-    },
-    {
-      "epoch": 0.9967553536664504,
-      "grad_norm": 0.310546875,
-      "learning_rate": 2.5307388795716688e-06,
-      "loss": 1.1128,
-      "step": 1024
-    },
-    {
-      "epoch": 0.9977287475665152,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.5268966784722792e-06,
-      "loss": 1.1342,
-      "step": 1025
-    },
-    {
-      "epoch": 0.9987021414665801,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.5230544138344513e-06,
-      "loss": 1.1108,
-      "step": 1026
-    },
-    {
-      "epoch": 0.999675535366645,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.519212094734828e-06,
-      "loss": 1.108,
-      "step": 1027
-    },
-    {
-      "epoch": 1.00064892926671,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.5153697302501777e-06,
-      "loss": 1.0967,
-      "step": 1028
-    },
-    {
-      "epoch": 1.001622323166775,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.5115273294573774e-06,
-      "loss": 1.1076,
-      "step": 1029
-    },
-    {
-      "epoch": 1.001622323166775,
-      "eval_loss": 1.1358181238174438,
-      "eval_runtime": 3324.4601,
-      "eval_samples_per_second": 12.56,
-      "eval_steps_per_second": 1.57,
-      "step": 1029
-    },
-    {
-      "epoch": 1.0005680204487362,
-      "grad_norm": 0.310546875,
-      "learning_rate": 2.50768490143339e-06,
-      "loss": 1.1097,
-      "step": 1030
-    },
-    {
-      "epoch": 1.0015417697894267,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.50384245525524e-06,
-      "loss": 1.1214,
-      "step": 1031
-    },
-    {
-      "epoch": 1.0025155191301172,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.5e-06,
-      "loss": 1.1221,
-      "step": 1032
-    },
-    {
-      "epoch": 1.0034892684708079,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.4961575447447606e-06,
-      "loss": 1.1036,
-      "step": 1033
-    },
-    {
-      "epoch": 1.0044630178114984,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.492315098566612e-06,
-      "loss": 1.1166,
-      "step": 1034
-    },
-    {
-      "epoch": 1.0054367671521889,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.488472670542624e-06,
-      "loss": 1.1092,
-      "step": 1035
-    },
-    {
-      "epoch": 1.0064105164928794,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.4846302697498227e-06,
-      "loss": 1.1292,
-      "step": 1036
-    },
-    {
-      "epoch": 1.00738426583357,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.4807879052651722e-06,
-      "loss": 1.1315,
-      "step": 1037
-    },
-    {
-      "epoch": 1.0083580151742606,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.476945586165549e-06,
-      "loss": 1.1263,
-      "step": 1038
-    },
-    {
-      "epoch": 1.009331764514951,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.4731033215277216e-06,
-      "loss": 1.1283,
-      "step": 1039
-    },
-    {
-      "epoch": 1.0103055138556416,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.469261120428332e-06,
-      "loss": 1.1199,
-      "step": 1040
-    },
-    {
-      "epoch": 1.011279263196332,
-      "grad_norm": 0.2890625,
-      "learning_rate": 2.4654189919438697e-06,
-      "loss": 1.1186,
-      "step": 1041
-    },
-    {
-      "epoch": 1.0122530125370228,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.461576945150655e-06,
-      "loss": 1.1045,
-      "step": 1042
-    },
-    {
-      "epoch": 1.0132267618777133,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.4577349891248137e-06,
-      "loss": 1.1065,
-      "step": 1043
-    },
-    {
-      "epoch": 1.0142005112184038,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.4538931329422555e-06,
-      "loss": 1.1261,
-      "step": 1044
-    },
-    {
-      "epoch": 1.0151742605590943,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.4500513856786604e-06,
-      "loss": 1.1281,
-      "step": 1045
-    },
-    {
-      "epoch": 1.016148009899785,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.446209756409445e-06,
-      "loss": 1.1396,
-      "step": 1046
-    },
-    {
-      "epoch": 1.0171217592404755,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.4423682542097486e-06,
-      "loss": 1.0971,
-      "step": 1047
-    },
-    {
-      "epoch": 1.018095508581166,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.4385268881544117e-06,
-      "loss": 1.0897,
-      "step": 1048
-    },
-    {
-      "epoch": 1.0190692579218565,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.434685667317952e-06,
-      "loss": 1.1367,
-      "step": 1049
-    },
-    {
-      "epoch": 1.0200430072625473,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.4308446007745452e-06,
-      "loss": 1.1181,
-      "step": 1050
-    },
-    {
-      "epoch": 1.0210167566032378,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.4270036975980006e-06,
-      "loss": 1.1096,
-      "step": 1051
-    },
-    {
-      "epoch": 1.0219905059439283,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.4231629668617434e-06,
-      "loss": 1.1105,
-      "step": 1052
-    },
-    {
-      "epoch": 1.0229642552846188,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.4193224176387926e-06,
-      "loss": 1.111,
-      "step": 1053
-    },
-    {
-      "epoch": 1.0239380046253093,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.415482059001734e-06,
-      "loss": 1.114,
-      "step": 1054
-    },
-    {
-      "epoch": 1.024911753966,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.411641900022708e-06,
-      "loss": 1.1152,
-      "step": 1055
-    },
-    {
-      "epoch": 1.0258855033066905,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.4078019497733803e-06,
-      "loss": 1.098,
-      "step": 1056
-    },
-    {
-      "epoch": 1.026859252647381,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.4039622173249238e-06,
-      "loss": 1.1104,
-      "step": 1057
-    },
-    {
-      "epoch": 1.0278330019880715,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.4001227117479983e-06,
-      "loss": 1.1058,
-      "step": 1058
-    },
-    {
-      "epoch": 1.0288067513287622,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.3962834421127266e-06,
-      "loss": 1.1294,
-      "step": 1059
-    },
-    {
-      "epoch": 1.0297805006694527,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.3924444174886735e-06,
-      "loss": 1.117,
-      "step": 1060
-    },
-    {
-      "epoch": 1.0307542500101432,
-      "grad_norm": 0.322265625,
-      "learning_rate": 2.3886056469448287e-06,
-      "loss": 1.1008,
-      "step": 1061
-    },
-    {
-      "epoch": 1.0317279993508337,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.3847671395495757e-06,
-      "loss": 1.1255,
-      "step": 1062
-    },
-    {
-      "epoch": 1.0327017486915244,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.3809289043706817e-06,
-      "loss": 1.1273,
-      "step": 1063
-    },
-    {
-      "epoch": 1.033675498032215,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.3770909504752677e-06,
-      "loss": 1.1074,
-      "step": 1064
-    },
-    {
-      "epoch": 1.0346492473729054,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.373253286929792e-06,
-      "loss": 1.1135,
-      "step": 1065
-    },
-    {
-      "epoch": 1.035622996713596,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.369415922800025e-06,
-      "loss": 1.0959,
-      "step": 1066
-    },
-    {
-      "epoch": 1.0365967460542864,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.3655788671510314e-06,
-      "loss": 1.0967,
-      "step": 1067
-    },
-    {
-      "epoch": 1.0375704953949771,
-      "grad_norm": 0.3125,
-      "learning_rate": 2.3617421290471463e-06,
-      "loss": 1.12,
-      "step": 1068
-    },
-    {
-      "epoch": 1.0385442447356676,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.357905717551957e-06,
-      "loss": 1.1282,
-      "step": 1069
-    },
-    {
-      "epoch": 1.0395179940763581,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.3540696417282755e-06,
-      "loss": 1.0986,
-      "step": 1070
-    },
-    {
-      "epoch": 1.0404917434170486,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.350233910638124e-06,
-      "loss": 1.1188,
-      "step": 1071
-    },
-    {
-      "epoch": 1.0414654927577394,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.3463985333427076e-06,
-      "loss": 1.1008,
-      "step": 1072
-    },
-    {
-      "epoch": 1.0424392420984299,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.3425635189023987e-06,
-      "loss": 1.1319,
-      "step": 1073
-    },
-    {
-      "epoch": 1.0434129914391204,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.3387288763767097e-06,
-      "loss": 1.1305,
-      "step": 1074
-    },
-    {
-      "epoch": 1.0443867407798109,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.334894614824276e-06,
-      "loss": 1.0987,
-      "step": 1075
-    },
-    {
-      "epoch": 1.0453604901205016,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.3310607433028317e-06,
-      "loss": 1.117,
-      "step": 1076
-    },
-    {
-      "epoch": 1.046334239461192,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.327227270869192e-06,
-      "loss": 1.1093,
-      "step": 1077
-    },
-    {
-      "epoch": 1.0473079888018826,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.3233942065792274e-06,
-      "loss": 1.1209,
-      "step": 1078
-    },
-    {
-      "epoch": 1.048281738142573,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.3195615594878446e-06,
-      "loss": 1.1055,
-      "step": 1079
-    },
-    {
-      "epoch": 1.0492554874832636,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.3157293386489636e-06,
-      "loss": 1.1049,
-      "step": 1080
-    },
-    {
-      "epoch": 1.0502292368239543,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.3118975531155003e-06,
-      "loss": 1.1202,
-      "step": 1081
-    },
-    {
-      "epoch": 1.0512029861646448,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.3080662119393395e-06,
-      "loss": 1.1115,
-      "step": 1082
-    },
-    {
-      "epoch": 1.0521767355053353,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.3042353241713176e-06,
-      "loss": 1.1082,
-      "step": 1083
-    },
-    {
-      "epoch": 1.0531504848460258,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.300404898861199e-06,
-      "loss": 1.1151,
-      "step": 1084
-    },
-    {
-      "epoch": 1.0541242341867165,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.2965749450576573e-06,
-      "loss": 1.1177,
-      "step": 1085
-    },
-    {
-      "epoch": 1.055097983527407,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.292745471808252e-06,
-      "loss": 1.1014,
-      "step": 1086
-    },
-    {
-      "epoch": 1.0560717328680975,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.2889164881594047e-06,
-      "loss": 1.1095,
-      "step": 1087
-    },
-    {
-      "epoch": 1.057045482208788,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.2850880031563845e-06,
-      "loss": 1.1264,
-      "step": 1088
-    },
-    {
-      "epoch": 1.0580192315494787,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.2812600258432783e-06,
-      "loss": 1.1343,
-      "step": 1089
-    },
-    {
-      "epoch": 1.0589929808901692,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.2774325652629765e-06,
-      "loss": 1.118,
-      "step": 1090
-    },
-    {
-      "epoch": 1.0599667302308597,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.273605630457149e-06,
-      "loss": 1.1143,
-      "step": 1091
-    },
-    {
-      "epoch": 1.0609404795715502,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.269779230466221e-06,
-      "loss": 1.1059,
-      "step": 1092
-    },
-    {
-      "epoch": 1.0619142289122407,
-      "grad_norm": 0.2890625,
-      "learning_rate": 2.2659533743293574e-06,
-      "loss": 1.0922,
-      "step": 1093
-    },
-    {
-      "epoch": 1.0628879782529315,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.262128071084437e-06,
-      "loss": 1.1063,
-      "step": 1094
-    },
-    {
-      "epoch": 1.063861727593622,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.2583033297680316e-06,
-      "loss": 1.1176,
-      "step": 1095
-    },
-    {
-      "epoch": 1.0648354769343125,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.254479159415388e-06,
-      "loss": 1.1075,
-      "step": 1096
-    },
-    {
-      "epoch": 1.065809226275003,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.2506555690604e-06,
-      "loss": 1.1095,
-      "step": 1097
-    },
-    {
-      "epoch": 1.0667829756156937,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.246832567735597e-06,
-      "loss": 1.1053,
-      "step": 1098
-    },
-    {
-      "epoch": 1.0677567249563842,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.2430101644721116e-06,
-      "loss": 1.1186,
-      "step": 1099
-    },
-    {
-      "epoch": 1.0687304742970747,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.2391883682996657e-06,
-      "loss": 1.1008,
-      "step": 1100
-    },
-    {
-      "epoch": 1.0697042236377652,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.235367188246549e-06,
-      "loss": 1.1145,
-      "step": 1101
-    },
-    {
-      "epoch": 1.070677972978456,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.2315466333395927e-06,
-      "loss": 1.1062,
-      "step": 1102
-    },
-    {
-      "epoch": 1.0716517223191464,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.227726712604152e-06,
-      "loss": 1.1166,
-      "step": 1103
-    },
-    {
-      "epoch": 1.072625471659837,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.223907435064086e-06,
-      "loss": 1.131,
-      "step": 1104
-    },
-    {
-      "epoch": 1.0735992210005274,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.2200888097417308e-06,
-      "loss": 1.111,
-      "step": 1105
-    },
-    {
-      "epoch": 1.074572970341218,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.2162708456578853e-06,
-      "loss": 1.1044,
-      "step": 1106
-    },
-    {
-      "epoch": 1.0755467196819086,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.212453551831783e-06,
-      "loss": 1.0777,
-      "step": 1107
-    },
-    {
-      "epoch": 1.0765204690225991,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.208636937281077e-06,
-      "loss": 1.1256,
-      "step": 1108
-    },
-    {
-      "epoch": 1.0774942183632896,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.204821011021815e-06,
-      "loss": 1.1179,
-      "step": 1109
-    },
-    {
-      "epoch": 1.0784679677039801,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.2010057820684176e-06,
-      "loss": 1.1123,
-      "step": 1110
-    },
-    {
-      "epoch": 1.0794417170446708,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.1971912594336593e-06,
-      "loss": 1.1165,
-      "step": 1111
-    },
-    {
-      "epoch": 1.0804154663853613,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.1933774521286457e-06,
-      "loss": 1.1145,
-      "step": 1112
-    },
-    {
-      "epoch": 1.0813892157260518,
-      "grad_norm": 0.33203125,
-      "learning_rate": 2.189564369162791e-06,
-      "loss": 1.1141,
-      "step": 1113
-    },
-    {
-      "epoch": 1.0823629650667423,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.185752019543801e-06,
-      "loss": 1.0997,
-      "step": 1114
-    },
-    {
-      "epoch": 1.0833367144074328,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.181940412277647e-06,
-      "loss": 1.1082,
-      "step": 1115
-    },
-    {
-      "epoch": 1.0843104637481236,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.178129556368548e-06,
-      "loss": 1.1159,
-      "step": 1116
-    },
-    {
-      "epoch": 1.085284213088814,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.174319460818948e-06,
-      "loss": 1.1134,
-      "step": 1117
-    },
-    {
-      "epoch": 1.0862579624295046,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.1705101346294936e-06,
-      "loss": 1.1019,
-      "step": 1118
-    },
-    {
-      "epoch": 1.087231711770195,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.166701586799015e-06,
-      "loss": 1.1055,
-      "step": 1119
-    },
-    {
-      "epoch": 1.0882054611108858,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.162893826324503e-06,
-      "loss": 1.1126,
-      "step": 1120
-    },
-    {
-      "epoch": 1.0891792104515763,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.1590868622010892e-06,
-      "loss": 1.1191,
-      "step": 1121
-    },
-    {
-      "epoch": 1.0901529597922668,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.1552807034220226e-06,
-      "loss": 1.1114,
-      "step": 1122
-    },
-    {
-      "epoch": 1.0911267091329573,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.151475358978652e-06,
-      "loss": 1.1206,
-      "step": 1123
-    },
-    {
-      "epoch": 1.092100458473648,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.1476708378603985e-06,
-      "loss": 1.1238,
-      "step": 1124
-    },
-    {
-      "epoch": 1.0930742078143385,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.1438671490547442e-06,
-      "loss": 1.1146,
-      "step": 1125
-    },
-    {
-      "epoch": 1.094047957155029,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.1400643015471994e-06,
-      "loss": 1.1249,
-      "step": 1126
-    },
-    {
-      "epoch": 1.0950217064957195,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.1362623043212904e-06,
-      "loss": 1.1261,
-      "step": 1127
-    },
-    {
-      "epoch": 1.0959954558364102,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.132461166358533e-06,
-      "loss": 1.1253,
-      "step": 1128
-    },
-    {
-      "epoch": 1.0969692051771007,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.128660896638414e-06,
-      "loss": 1.1115,
-      "step": 1129
-    },
-    {
-      "epoch": 1.0979429545177912,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.1248615041383686e-06,
-      "loss": 1.1012,
-      "step": 1130
-    },
-    {
-      "epoch": 1.0989167038584817,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.1210629978337615e-06,
-      "loss": 1.1071,
-      "step": 1131
-    },
-    {
-      "epoch": 1.0998904531991722,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.1172653866978603e-06,
-      "loss": 1.1145,
-      "step": 1132
-    },
-    {
-      "epoch": 1.100864202539863,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.113468679701822e-06,
-      "loss": 1.1067,
-      "step": 1133
-    },
-    {
-      "epoch": 1.1018379518805534,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.1096728858146667e-06,
-      "loss": 1.1113,
-      "step": 1134
-    },
-    {
-      "epoch": 1.102811701221244,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.1058780140032544e-06,
-      "loss": 1.1214,
-      "step": 1135
-    },
-    {
-      "epoch": 1.1037854505619344,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.10208407323227e-06,
-      "loss": 1.107,
-      "step": 1136
-    },
-    {
-      "epoch": 1.1047591999026252,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.098291072464199e-06,
-      "loss": 1.1142,
-      "step": 1137
-    },
-    {
-      "epoch": 1.1057329492433157,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.094499020659304e-06,
-      "loss": 1.129,
-      "step": 1138
-    },
-    {
-      "epoch": 1.1067066985840062,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.0907079267756085e-06,
-      "loss": 1.108,
-      "step": 1139
-    },
-    {
-      "epoch": 1.1076804479246967,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.0869177997688706e-06,
-      "loss": 1.1033,
-      "step": 1140
-    },
-    {
-      "epoch": 1.1086541972653872,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.0831286485925666e-06,
-      "loss": 1.0946,
-      "step": 1141
-    },
-    {
-      "epoch": 1.1096279466060779,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.079340482197867e-06,
-      "loss": 1.1255,
-      "step": 1142
-    },
-    {
-      "epoch": 1.1106016959467684,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.0755533095336147e-06,
-      "loss": 1.1146,
-      "step": 1143
-    },
-    {
-      "epoch": 1.1115754452874589,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.0717671395463063e-06,
-      "loss": 1.1341,
-      "step": 1144
-    },
-    {
-      "epoch": 1.1125491946281494,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.0679819811800687e-06,
-      "loss": 1.0996,
-      "step": 1145
-    },
-    {
-      "epoch": 1.11352294396884,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.0641978433766412e-06,
-      "loss": 1.1194,
-      "step": 1146
-    },
-    {
-      "epoch": 1.1144966933095306,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.060414735075349e-06,
-      "loss": 1.1227,
-      "step": 1147
-    },
-    {
-      "epoch": 1.115470442650221,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.0566326652130874e-06,
-      "loss": 1.1378,
-      "step": 1148
-    },
-    {
-      "epoch": 1.1164441919909116,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.0528516427242987e-06,
-      "loss": 1.111,
-      "step": 1149
-    },
-    {
-      "epoch": 1.1174179413316023,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.049071676540951e-06,
-      "loss": 1.1221,
-      "step": 1150
-    },
-    {
-      "epoch": 1.1183916906722928,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.0452927755925155e-06,
-      "loss": 1.1032,
-      "step": 1151
-    },
-    {
-      "epoch": 1.1193654400129833,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.0415149488059483e-06,
-      "loss": 1.1165,
-      "step": 1152
-    },
-    {
-      "epoch": 1.1203391893536738,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.0377382051056678e-06,
-      "loss": 1.1032,
-      "step": 1153
-    },
-    {
-      "epoch": 1.1213129386943643,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.033962553413534e-06,
-      "loss": 1.0967,
-      "step": 1154
-    },
-    {
-      "epoch": 1.122286688035055,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.0301880026488263e-06,
-      "loss": 1.1204,
-      "step": 1155
-    },
-    {
-      "epoch": 1.1232604373757455,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.026414561728224e-06,
-      "loss": 1.1261,
-      "step": 1156
-    },
-    {
-      "epoch": 1.124234186716436,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.0226422395657857e-06,
-      "loss": 1.1062,
-      "step": 1157
-    },
-    {
-      "epoch": 1.1252079360571265,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.0188710450729255e-06,
-      "loss": 1.1048,
-      "step": 1158
-    },
-    {
-      "epoch": 1.1261816853978173,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.0151009871583933e-06,
-      "loss": 1.1018,
-      "step": 1159
-    },
-    {
-      "epoch": 1.1271554347385078,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.0113320747282563e-06,
-      "loss": 1.1175,
-      "step": 1160
-    },
-    {
-      "epoch": 1.1281291840791983,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.0075643166858733e-06,
-      "loss": 1.1107,
-      "step": 1161
-    },
-    {
-      "epoch": 1.1291029334198888,
-      "grad_norm": 0.310546875,
-      "learning_rate": 2.0037977219318776e-06,
-      "loss": 1.1127,
-      "step": 1162
-    },
-    {
-      "epoch": 1.1300766827605795,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.000032299364153e-06,
-      "loss": 1.1096,
-      "step": 1163
-    },
-    {
-      "epoch": 1.13105043210127,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.9962680578778153e-06,
-      "loss": 1.1111,
-      "step": 1164
-    },
-    {
-      "epoch": 1.1320241814419605,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.992505006365191e-06,
-      "loss": 1.1122,
-      "step": 1165
-    },
-    {
-      "epoch": 1.132997930782651,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.9887431537157944e-06,
-      "loss": 1.1191,
-      "step": 1166
-    },
-    {
-      "epoch": 1.1339716801233415,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.984982508816308e-06,
-      "loss": 1.1103,
-      "step": 1167
-    },
-    {
-      "epoch": 1.1349454294640322,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.9812230805505596e-06,
-      "loss": 1.1242,
-      "step": 1168
-    },
-    {
-      "epoch": 1.1359191788047227,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.977464877799506e-06,
-      "loss": 1.1138,
-      "step": 1169
-    },
-    {
-      "epoch": 1.1368929281454132,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.9737079094412072e-06,
-      "loss": 1.1269,
-      "step": 1170
-    },
-    {
-      "epoch": 1.1378666774861037,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.969952184350806e-06,
-      "loss": 1.105,
-      "step": 1171
-    },
-    {
-      "epoch": 1.1388404268267944,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.96619771140051e-06,
-      "loss": 1.1197,
-      "step": 1172
-    },
-    {
-      "epoch": 1.139814176167485,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.9624444994595694e-06,
-      "loss": 1.108,
-      "step": 1173
-    },
-    {
-      "epoch": 1.1407879255081754,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.958692557394254e-06,
-      "loss": 1.1319,
-      "step": 1174
-    },
-    {
-      "epoch": 1.141761674848866,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.954941894067834e-06,
-      "loss": 1.1308,
-      "step": 1175
-    },
-    {
-      "epoch": 1.1427354241895564,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.951192518340559e-06,
-      "loss": 1.1501,
-      "step": 1176
-    },
-    {
-      "epoch": 1.1437091735302471,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.9474444390696377e-06,
-      "loss": 1.1069,
-      "step": 1177
-    },
-    {
-      "epoch": 1.1446829228709376,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.9436976651092143e-06,
-      "loss": 1.1194,
-      "step": 1178
-    },
-    {
-      "epoch": 1.1456566722116281,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.9399522053103514e-06,
-      "loss": 1.1233,
-      "step": 1179
-    },
-    {
-      "epoch": 1.1466304215523186,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.936208068521005e-06,
-      "loss": 1.1126,
-      "step": 1180
-    },
-    {
-      "epoch": 1.1476041708930094,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.9324652635860087e-06,
-      "loss": 1.1051,
-      "step": 1181
-    },
-    {
-      "epoch": 1.1485779202336999,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.9287237993470475e-06,
-      "loss": 1.124,
-      "step": 1182
-    },
-    {
-      "epoch": 1.1495516695743904,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.9249836846426395e-06,
-      "loss": 1.1133,
-      "step": 1183
-    },
-    {
-      "epoch": 1.1505254189150809,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.9212449283081157e-06,
-      "loss": 1.1117,
-      "step": 1184
-    },
-    {
-      "epoch": 1.1514991682557716,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.9175075391755973e-06,
-      "loss": 1.135,
-      "step": 1185
-    },
-    {
-      "epoch": 1.152472917596462,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.913771526073976e-06,
-      "loss": 1.1045,
-      "step": 1186
-    },
-    {
-      "epoch": 1.1534466669371526,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.910036897828894e-06,
-      "loss": 1.129,
-      "step": 1187
-    },
-    {
-      "epoch": 1.154420416277843,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.9063036632627188e-06,
-      "loss": 1.1113,
-      "step": 1188
-    },
-    {
-      "epoch": 1.1553941656185338,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.9025718311945302e-06,
-      "loss": 1.1178,
-      "step": 1189
-    },
-    {
-      "epoch": 1.1563679149592243,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.898841410440092e-06,
-      "loss": 1.1071,
-      "step": 1190
-    },
-    {
-      "epoch": 1.1573416642999148,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.8951124098118337e-06,
-      "loss": 1.129,
-      "step": 1191
-    },
-    {
-      "epoch": 1.1583154136406053,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.8913848381188316e-06,
-      "loss": 1.1079,
-      "step": 1192
-    },
-    {
-      "epoch": 1.1592891629812958,
-      "grad_norm": 0.3203125,
-      "learning_rate": 1.8876587041667855e-06,
-      "loss": 1.112,
-      "step": 1193
-    },
-    {
-      "epoch": 1.1602629123219865,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.8839340167579977e-06,
-      "loss": 1.1124,
-      "step": 1194
-    },
-    {
-      "epoch": 1.161236661662677,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.8802107846913567e-06,
-      "loss": 1.0919,
-      "step": 1195
-    },
-    {
-      "epoch": 1.1622104110033675,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.8764890167623082e-06,
-      "loss": 1.112,
-      "step": 1196
-    },
-    {
-      "epoch": 1.163184160344058,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.872768721762844e-06,
-      "loss": 1.1075,
-      "step": 1197
-    },
-    {
-      "epoch": 1.1641579096847487,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.8690499084814739e-06,
-      "loss": 1.1096,
-      "step": 1198
-    },
-    {
-      "epoch": 1.1651316590254392,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.865332585703207e-06,
-      "loss": 1.1192,
-      "step": 1199
-    },
-    {
-      "epoch": 1.1661054083661297,
-      "grad_norm": 0.3203125,
-      "learning_rate": 1.8616167622095328e-06,
-      "loss": 1.1225,
-      "step": 1200
-    },
-    {
-      "epoch": 1.1670791577068202,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.8579024467783975e-06,
-      "loss": 1.1096,
-      "step": 1201
-    },
-    {
-      "epoch": 1.1680529070475107,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.854189648184186e-06,
-      "loss": 1.1178,
-      "step": 1202
-    },
-    {
-      "epoch": 1.1690266563882015,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.8504783751976995e-06,
-      "loss": 1.1021,
-      "step": 1203
-    },
-    {
-      "epoch": 1.170000405728892,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.8467686365861343e-06,
-      "loss": 1.1175,
-      "step": 1204
-    },
-    {
-      "epoch": 1.1709741550695825,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.8430604411130642e-06,
-      "loss": 1.1163,
-      "step": 1205
-    },
-    {
-      "epoch": 1.171947904410273,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.8393537975384168e-06,
-      "loss": 1.1116,
-      "step": 1206
-    },
-    {
-      "epoch": 1.1729216537509637,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.8356487146184517e-06,
-      "loss": 1.1107,
-      "step": 1207
-    },
-    {
-      "epoch": 1.1738954030916542,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.8319452011057446e-06,
-      "loss": 1.113,
-      "step": 1208
-    },
-    {
-      "epoch": 1.1748691524323447,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.828243265749161e-06,
-      "loss": 1.1106,
-      "step": 1209
-    },
-    {
-      "epoch": 1.1758429017730352,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.8245429172938414e-06,
-      "loss": 1.1074,
-      "step": 1210
-    },
-    {
-      "epoch": 1.1768166511137257,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.8208441644811741e-06,
-      "loss": 1.1152,
-      "step": 1211
-    },
-    {
-      "epoch": 1.1777904004544164,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.8171470160487806e-06,
-      "loss": 1.1179,
-      "step": 1212
-    },
-    {
-      "epoch": 1.178764149795107,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.8134514807304923e-06,
-      "loss": 1.1096,
-      "step": 1213
-    },
-    {
-      "epoch": 1.1797378991357974,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.8097575672563278e-06,
-      "loss": 1.1059,
-      "step": 1214
-    },
-    {
-      "epoch": 1.1807116484764881,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.8060652843524767e-06,
-      "loss": 1.0962,
-      "step": 1215
-    },
-    {
-      "epoch": 1.1816853978171786,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.8023746407412754e-06,
-      "loss": 1.1279,
-      "step": 1216
-    },
-    {
-      "epoch": 1.1826591471578691,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.7986856451411872e-06,
-      "loss": 1.1116,
-      "step": 1217
-    },
-    {
-      "epoch": 1.1836328964985596,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.7949983062667842e-06,
-      "loss": 1.1243,
-      "step": 1218
-    },
-    {
-      "epoch": 1.1846066458392501,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.791312632828723e-06,
-      "loss": 1.1019,
-      "step": 1219
-    },
-    {
-      "epoch": 1.1855803951799408,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.7876286335337265e-06,
-      "loss": 1.12,
-      "step": 1220
-    },
-    {
-      "epoch": 1.1865541445206313,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.7839463170845641e-06,
-      "loss": 1.0793,
-      "step": 1221
-    },
-    {
-      "epoch": 1.1875278938613218,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.7802656921800277e-06,
-      "loss": 1.0903,
-      "step": 1222
-    },
-    {
-      "epoch": 1.1885016432020123,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.7765867675149145e-06,
-      "loss": 1.1243,
-      "step": 1223
-    },
-    {
-      "epoch": 1.189475392542703,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.7729095517800039e-06,
-      "loss": 1.119,
-      "step": 1224
-    },
-    {
-      "epoch": 1.1904491418833936,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.7692340536620406e-06,
-      "loss": 1.119,
-      "step": 1225
-    },
-    {
-      "epoch": 1.191422891224084,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.7655602818437095e-06,
-      "loss": 1.1193,
-      "step": 1226
-    },
-    {
-      "epoch": 1.1923966405647746,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.7618882450036175e-06,
-      "loss": 1.1055,
-      "step": 1227
-    },
-    {
-      "epoch": 1.193370389905465,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.7582179518162742e-06,
-      "loss": 1.1207,
-      "step": 1228
-    },
-    {
-      "epoch": 1.1943441392461558,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.7545494109520711e-06,
-      "loss": 1.1203,
-      "step": 1229
-    },
-    {
-      "epoch": 1.1953178885868463,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.7508826310772566e-06,
-      "loss": 1.1045,
-      "step": 1230
-    },
-    {
-      "epoch": 1.1962916379275368,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.7472176208539233e-06,
-      "loss": 1.1119,
-      "step": 1231
-    },
-    {
-      "epoch": 1.1972653872682273,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.7435543889399786e-06,
-      "loss": 1.0956,
-      "step": 1232
-    },
-    {
-      "epoch": 1.198239136608918,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.7398929439891344e-06,
-      "loss": 1.1145,
-      "step": 1233
-    },
-    {
-      "epoch": 1.1992128859496085,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.7362332946508759e-06,
-      "loss": 1.1022,
-      "step": 1234
-    },
-    {
-      "epoch": 1.200186635290299,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.7325754495704508e-06,
-      "loss": 1.1177,
-      "step": 1235
-    },
-    {
-      "epoch": 1.2011603846309895,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.728919417388841e-06,
-      "loss": 1.1089,
-      "step": 1236
-    },
-    {
-      "epoch": 1.20213413397168,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.7252652067427497e-06,
-      "loss": 1.1023,
-      "step": 1237
-    },
-    {
-      "epoch": 1.2031078833123707,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.7216128262645734e-06,
-      "loss": 1.1144,
-      "step": 1238
-    },
-    {
-      "epoch": 1.2040816326530612,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.7179622845823875e-06,
-      "loss": 1.1075,
-      "step": 1239
-    },
-    {
-      "epoch": 1.2050553819937517,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.714313590319922e-06,
-      "loss": 1.1142,
-      "step": 1240
-    },
-    {
-      "epoch": 1.2060291313344425,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.7106667520965443e-06,
-      "loss": 1.1008,
-      "step": 1241
-    },
-    {
-      "epoch": 1.207002880675133,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.7070217785272354e-06,
-      "loss": 1.1158,
-      "step": 1242
-    },
-    {
-      "epoch": 1.2079766300158234,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.7033786782225734e-06,
-      "loss": 1.1119,
-      "step": 1243
-    },
-    {
-      "epoch": 1.208950379356514,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.6997374597887086e-06,
-      "loss": 1.1195,
-      "step": 1244
-    },
-    {
-      "epoch": 1.2099241286972044,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.6960981318273492e-06,
-      "loss": 1.1188,
-      "step": 1245
-    },
-    {
-      "epoch": 1.2108978780378952,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.6924607029357354e-06,
-      "loss": 1.113,
-      "step": 1246
-    },
-    {
-      "epoch": 1.2118716273785857,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.6888251817066199e-06,
-      "loss": 1.1185,
-      "step": 1247
-    },
-    {
-      "epoch": 1.2128453767192762,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.685191576728252e-06,
-      "loss": 1.108,
-      "step": 1248
-    },
-    {
-      "epoch": 1.2138191260599667,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.681559896584352e-06,
-      "loss": 1.1152,
-      "step": 1249
-    },
-    {
-      "epoch": 1.2147928754006574,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.6779301498540934e-06,
-      "loss": 1.0984,
-      "step": 1250
-    },
-    {
-      "epoch": 1.215766624741348,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.6743023451120831e-06,
-      "loss": 1.1316,
-      "step": 1251
-    },
-    {
-      "epoch": 1.2167403740820384,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.670676490928339e-06,
-      "loss": 1.0933,
-      "step": 1252
-    },
-    {
-      "epoch": 1.217714123422729,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.6670525958682738e-06,
-      "loss": 1.0946,
-      "step": 1253
-    },
-    {
-      "epoch": 1.2186878727634194,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.6634306684926705e-06,
-      "loss": 1.0899,
-      "step": 1254
-    },
-    {
-      "epoch": 1.21966162210411,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.6598107173576624e-06,
-      "loss": 1.1171,
-      "step": 1255
-    },
-    {
-      "epoch": 1.2206353714448006,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.6561927510147172e-06,
-      "loss": 1.0925,
-      "step": 1256
-    },
-    {
-      "epoch": 1.221609120785491,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.652576778010611e-06,
-      "loss": 1.112,
-      "step": 1257
-    },
-    {
-      "epoch": 1.2225828701261816,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.6489628068874142e-06,
-      "loss": 1.0896,
-      "step": 1258
-    },
-    {
-      "epoch": 1.2235566194668723,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.645350846182465e-06,
-      "loss": 1.121,
-      "step": 1259
-    },
-    {
-      "epoch": 1.2245303688075628,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.641740904428353e-06,
-      "loss": 1.1226,
-      "step": 1260
-    },
-    {
-      "epoch": 1.2255041181482533,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.6381329901529007e-06,
-      "loss": 1.1406,
-      "step": 1261
-    },
-    {
-      "epoch": 1.2264778674889438,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.6345271118791395e-06,
-      "loss": 1.0938,
-      "step": 1262
-    },
-    {
-      "epoch": 1.2274516168296343,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.63092327812529e-06,
-      "loss": 1.1203,
-      "step": 1263
-    },
-    {
-      "epoch": 1.228425366170325,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.6273214974047452e-06,
-      "loss": 1.1144,
-      "step": 1264
-    },
-    {
-      "epoch": 1.2293991155110155,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.6237217782260462e-06,
-      "loss": 1.1032,
-      "step": 1265
-    },
-    {
-      "epoch": 1.230372864851706,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.6201241290928654e-06,
-      "loss": 1.0979,
-      "step": 1266
-    },
-    {
-      "epoch": 1.2313466141923965,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.6165285585039841e-06,
-      "loss": 1.1048,
-      "step": 1267
-    },
-    {
-      "epoch": 1.2323203635330873,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.6129350749532744e-06,
-      "loss": 1.1076,
-      "step": 1268
-    },
-    {
-      "epoch": 1.2332941128737778,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.6093436869296787e-06,
-      "loss": 1.1214,
-      "step": 1269
-    },
-    {
-      "epoch": 1.2342678622144683,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.6057544029171863e-06,
-      "loss": 1.1205,
-      "step": 1270
-    },
-    {
-      "epoch": 1.2352416115551588,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.6021672313948198e-06,
-      "loss": 1.1231,
-      "step": 1271
-    },
-    {
-      "epoch": 1.2362153608958493,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.5985821808366087e-06,
-      "loss": 1.1176,
-      "step": 1272
-    },
-    {
-      "epoch": 1.23718911023654,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.5949992597115726e-06,
-      "loss": 1.1265,
-      "step": 1273
-    },
-    {
-      "epoch": 1.2381628595772305,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.5914184764837024e-06,
-      "loss": 1.1282,
-      "step": 1274
-    },
-    {
-      "epoch": 1.239136608917921,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.587839839611936e-06,
-      "loss": 1.1155,
-      "step": 1275
-    },
-    {
-      "epoch": 1.2401103582586117,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.5842633575501428e-06,
-      "loss": 1.1057,
-      "step": 1276
-    },
-    {
-      "epoch": 1.2410841075993022,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.5806890387471025e-06,
-      "loss": 1.1049,
-      "step": 1277
-    },
-    {
-      "epoch": 1.2420578569399927,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.577116891646482e-06,
-      "loss": 1.1125,
-      "step": 1278
-    },
-    {
-      "epoch": 1.2430316062806832,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.5735469246868207e-06,
-      "loss": 1.1019,
-      "step": 1279
-    },
-    {
-      "epoch": 1.2440053556213737,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.5699791463015045e-06,
-      "loss": 1.114,
-      "step": 1280
-    },
-    {
-      "epoch": 1.2449791049620644,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.566413564918753e-06,
-      "loss": 1.1194,
-      "step": 1281
-    },
-    {
-      "epoch": 1.245952854302755,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.562850188961593e-06,
-      "loss": 1.1103,
-      "step": 1282
-    },
-    {
-      "epoch": 1.2469266036434454,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.5592890268478417e-06,
-      "loss": 1.1121,
-      "step": 1283
-    },
-    {
-      "epoch": 1.247900352984136,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.5557300869900876e-06,
-      "loss": 1.1036,
-      "step": 1284
-    },
-    {
-      "epoch": 1.2488741023248267,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.5521733777956704e-06,
-      "loss": 1.1153,
-      "step": 1285
-    },
-    {
-      "epoch": 1.2498478516655172,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.5486189076666575e-06,
-      "loss": 1.1262,
-      "step": 1286
-    },
-    {
-      "epoch": 1.2508216010062077,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.5450666849998291e-06,
-      "loss": 1.1079,
-      "step": 1287
-    },
-    {
-      "epoch": 1.2517953503468982,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.5415167181866541e-06,
-      "loss": 1.1224,
-      "step": 1288
-    },
-    {
-      "epoch": 1.2527690996875886,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.5379690156132757e-06,
-      "loss": 1.1507,
-      "step": 1289
-    },
-    {
-      "epoch": 1.2537428490282794,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.5344235856604845e-06,
-      "loss": 1.112,
-      "step": 1290
-    },
-    {
-      "epoch": 1.2547165983689699,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.530880436703705e-06,
-      "loss": 1.142,
-      "step": 1291
-    },
-    {
-      "epoch": 1.2556903477096604,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.5273395771129714e-06,
-      "loss": 1.1056,
-      "step": 1292
-    },
-    {
-      "epoch": 1.256664097050351,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.5238010152529125e-06,
-      "loss": 1.1141,
-      "step": 1293
-    },
-    {
-      "epoch": 1.2576378463910416,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.5202647594827269e-06,
-      "loss": 1.1176,
-      "step": 1294
-    },
-    {
-      "epoch": 1.258611595731732,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.5167308181561658e-06,
-      "loss": 1.1059,
-      "step": 1295
-    },
-    {
-      "epoch": 1.2595853450724226,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.5131991996215123e-06,
-      "loss": 1.1009,
-      "step": 1296
-    },
-    {
-      "epoch": 1.260559094413113,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.5096699122215642e-06,
-      "loss": 1.1101,
-      "step": 1297
-    },
-    {
-      "epoch": 1.2615328437538036,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.5061429642936107e-06,
-      "loss": 1.1115,
-      "step": 1298
-    },
-    {
-      "epoch": 1.2625065930944943,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.502618364169416e-06,
-      "loss": 1.1085,
-      "step": 1299
-    },
-    {
-      "epoch": 1.2634803424351848,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.4990961201751957e-06,
-      "loss": 1.1095,
-      "step": 1300
-    },
-    {
-      "epoch": 1.2644540917758753,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.4955762406316023e-06,
-      "loss": 1.1101,
-      "step": 1301
-    },
-    {
-      "epoch": 1.265427841116566,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.4920587338537018e-06,
-      "loss": 1.0996,
-      "step": 1302
-    },
-    {
-      "epoch": 1.2664015904572565,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.488543608150953e-06,
-      "loss": 1.1236,
-      "step": 1303
-    },
-    {
-      "epoch": 1.267375339797947,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.4850308718271927e-06,
-      "loss": 1.1067,
-      "step": 1304
-    },
-    {
-      "epoch": 1.2683490891386375,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.4815205331806113e-06,
-      "loss": 1.1076,
-      "step": 1305
-    },
-    {
-      "epoch": 1.269322838479328,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.4780126005037354e-06,
-      "loss": 1.1076,
-      "step": 1306
-    },
-    {
-      "epoch": 1.2702965878200185,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.474507082083409e-06,
-      "loss": 1.0992,
-      "step": 1307
-    },
-    {
-      "epoch": 1.2712703371607093,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.471003986200771e-06,
-      "loss": 1.1016,
-      "step": 1308
-    },
-    {
-      "epoch": 1.2722440865013998,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.4675033211312401e-06,
-      "loss": 1.119,
-      "step": 1309
-    },
-    {
-      "epoch": 1.2732178358420903,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.4640050951444912e-06,
-      "loss": 1.1419,
-      "step": 1310
-    },
-    {
-      "epoch": 1.274191585182781,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.4605093165044364e-06,
-      "loss": 1.1288,
-      "step": 1311
-    },
-    {
-      "epoch": 1.2751653345234715,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.4570159934692085e-06,
-      "loss": 1.1076,
-      "step": 1312
-    },
-    {
-      "epoch": 1.276139083864162,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.4535251342911367e-06,
-      "loss": 1.0944,
-      "step": 1313
-    },
-    {
-      "epoch": 1.2771128332048525,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.450036747216732e-06,
-      "loss": 1.0937,
-      "step": 1314
-    },
-    {
-      "epoch": 1.278086582545543,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.4465508404866662e-06,
-      "loss": 1.1264,
-      "step": 1315
-    },
-    {
-      "epoch": 1.2790603318862337,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.4430674223357471e-06,
-      "loss": 1.1206,
-      "step": 1316
-    },
-    {
-      "epoch": 1.2800340812269242,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.4395865009929112e-06,
-      "loss": 1.1171,
-      "step": 1317
-    },
-    {
-      "epoch": 1.2810078305676147,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.4361080846811888e-06,
-      "loss": 1.1164,
-      "step": 1318
-    },
-    {
-      "epoch": 1.2819815799083052,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.432632181617698e-06,
-      "loss": 1.0895,
-      "step": 1319
-    },
-    {
-      "epoch": 1.282955329248996,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.4291588000136176e-06,
-      "loss": 1.1169,
-      "step": 1320
-    },
-    {
-      "epoch": 1.2839290785896864,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.425687948074172e-06,
-      "loss": 1.1035,
-      "step": 1321
-    },
-    {
-      "epoch": 1.284902827930377,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.4222196339986053e-06,
-      "loss": 1.1043,
-      "step": 1322
-    },
-    {
-      "epoch": 1.2858765772710674,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.4187538659801706e-06,
-      "loss": 1.0925,
-      "step": 1323
-    },
-    {
-      "epoch": 1.286850326611758,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.415290652206105e-06,
-      "loss": 1.1007,
-      "step": 1324
-    },
-    {
-      "epoch": 1.2878240759524486,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.4118300008576117e-06,
-      "loss": 1.1139,
-      "step": 1325
-    },
-    {
-      "epoch": 1.2887978252931391,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.4083719201098404e-06,
-      "loss": 1.1087,
-      "step": 1326
-    },
-    {
-      "epoch": 1.2897715746338296,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.4049164181318684e-06,
-      "loss": 1.1057,
-      "step": 1327
-    },
-    {
-      "epoch": 1.2907453239745204,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.4014635030866825e-06,
-      "loss": 1.1311,
-      "step": 1328
-    },
-    {
-      "epoch": 1.2917190733152109,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.3980131831311544e-06,
-      "loss": 1.1018,
-      "step": 1329
-    },
-    {
-      "epoch": 1.2926928226559014,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.3945654664160296e-06,
-      "loss": 1.1137,
-      "step": 1330
-    },
-    {
-      "epoch": 1.2936665719965919,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.391120361085903e-06,
-      "loss": 1.1349,
-      "step": 1331
-    },
-    {
-      "epoch": 1.2946403213372824,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.3876778752791973e-06,
-      "loss": 1.1165,
-      "step": 1332
-    },
-    {
-      "epoch": 1.2956140706779729,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.3842380171281522e-06,
-      "loss": 1.0838,
-      "step": 1333
-    },
-    {
-      "epoch": 1.2965878200186636,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.3808007947587976e-06,
-      "loss": 1.1112,
-      "step": 1334
-    },
-    {
-      "epoch": 1.297561569359354,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.3773662162909346e-06,
-      "loss": 1.107,
-      "step": 1335
-    },
-    {
-      "epoch": 1.2985353187000446,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.373934289838122e-06,
-      "loss": 1.1078,
-      "step": 1336
-    },
-    {
-      "epoch": 1.2995090680407353,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.3705050235076524e-06,
-      "loss": 1.1118,
-      "step": 1337
-    },
-    {
-      "epoch": 1.3004828173814258,
-      "grad_norm": 0.2890625,
-      "learning_rate": 1.3670784254005354e-06,
-      "loss": 1.0981,
-      "step": 1338
-    },
-    {
-      "epoch": 1.3014565667221163,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.3636545036114746e-06,
-      "loss": 1.121,
-      "step": 1339
-    },
-    {
-      "epoch": 1.3024303160628068,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.3602332662288536e-06,
-      "loss": 1.1048,
-      "step": 1340
-    },
-    {
-      "epoch": 1.3034040654034973,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.3568147213347144e-06,
-      "loss": 1.0991,
-      "step": 1341
-    },
-    {
-      "epoch": 1.304377814744188,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.353398877004738e-06,
-      "loss": 1.1129,
-      "step": 1342
-    },
-    {
-      "epoch": 1.3053515640848785,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.3499857413082256e-06,
-      "loss": 1.1177,
-      "step": 1343
-    },
-    {
-      "epoch": 1.306325313425569,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.3465753223080824e-06,
-      "loss": 1.1208,
-      "step": 1344
-    },
-    {
-      "epoch": 1.3072990627662595,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.3431676280607903e-06,
-      "loss": 1.1169,
-      "step": 1345
-    },
-    {
-      "epoch": 1.3082728121069502,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.3397626666163998e-06,
-      "loss": 1.1027,
-      "step": 1346
-    },
-    {
-      "epoch": 1.3092465614476407,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.3363604460185031e-06,
-      "loss": 1.0955,
-      "step": 1347
-    },
-    {
-      "epoch": 1.3102203107883312,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.3329609743042183e-06,
-      "loss": 1.1151,
-      "step": 1348
-    },
-    {
-      "epoch": 1.3111940601290217,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.3295642595041703e-06,
-      "loss": 1.0816,
-      "step": 1349
-    },
-    {
-      "epoch": 1.3121678094697122,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.3261703096424695e-06,
-      "loss": 1.1209,
-      "step": 1350
-    },
-    {
-      "epoch": 1.313141558810403,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.322779132736698e-06,
-      "loss": 1.1151,
-      "step": 1351
-    },
-    {
-      "epoch": 1.3141153081510935,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.3193907367978825e-06,
-      "loss": 1.1085,
-      "step": 1352
-    },
-    {
-      "epoch": 1.315089057491784,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.3160051298304843e-06,
-      "loss": 1.1025,
-      "step": 1353
-    },
-    {
-      "epoch": 1.3160628068324747,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.3126223198323752e-06,
-      "loss": 1.0917,
-      "step": 1354
-    },
-    {
-      "epoch": 1.3170365561731652,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.309242314794818e-06,
-      "loss": 1.1167,
-      "step": 1355
-    },
-    {
-      "epoch": 1.3180103055138557,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.30586512270245e-06,
-      "loss": 1.1179,
-      "step": 1356
-    },
-    {
-      "epoch": 1.3189840548545462,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.3024907515332668e-06,
-      "loss": 1.1029,
-      "step": 1357
-    },
-    {
-      "epoch": 1.3199578041952367,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.2991192092585952e-06,
-      "loss": 1.1001,
-      "step": 1358
-    },
-    {
-      "epoch": 1.3209315535359272,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.2957505038430818e-06,
-      "loss": 1.1238,
-      "step": 1359
-    },
-    {
-      "epoch": 1.321905302876618,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.2923846432446712e-06,
-      "loss": 1.0861,
-      "step": 1360
-    },
-    {
-      "epoch": 1.3228790522173084,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.289021635414589e-06,
-      "loss": 1.114,
-      "step": 1361
-    },
-    {
-      "epoch": 1.323852801557999,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.2856614882973178e-06,
-      "loss": 1.1295,
-      "step": 1362
-    },
-    {
-      "epoch": 1.3248265508986896,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.282304209830586e-06,
-      "loss": 1.1232,
-      "step": 1363
-    },
-    {
-      "epoch": 1.3258003002393801,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.2789498079453443e-06,
-      "loss": 1.1087,
-      "step": 1364
-    },
-    {
-      "epoch": 1.3267740495800706,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.2755982905657478e-06,
-      "loss": 1.094,
-      "step": 1365
-    },
-    {
-      "epoch": 1.3277477989207611,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.272249665609137e-06,
-      "loss": 1.0989,
-      "step": 1366
-    },
-    {
-      "epoch": 1.3287215482614516,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.268903940986022e-06,
-      "loss": 1.1004,
-      "step": 1367
-    },
-    {
-      "epoch": 1.3296952976021421,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.265561124600057e-06,
-      "loss": 1.1216,
-      "step": 1368
-    },
-    {
-      "epoch": 1.3306690469428328,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.2622212243480302e-06,
-      "loss": 1.1027,
-      "step": 1369
-    },
-    {
-      "epoch": 1.3316427962835233,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.2588842481198382e-06,
-      "loss": 1.1039,
-      "step": 1370
-    },
-    {
-      "epoch": 1.3326165456242138,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.2555502037984732e-06,
-      "loss": 1.1137,
-      "step": 1371
-    },
-    {
-      "epoch": 1.3335902949649046,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.252219099259996e-06,
-      "loss": 1.1167,
-      "step": 1372
-    },
-    {
-      "epoch": 1.334564044305595,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.248890942373529e-06,
-      "loss": 1.1104,
-      "step": 1373
-    },
-    {
-      "epoch": 1.3355377936462856,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.2455657410012283e-06,
-      "loss": 1.1031,
-      "step": 1374
-    },
-    {
-      "epoch": 1.336511542986976,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.2422435029982669e-06,
-      "loss": 1.1091,
-      "step": 1375
-    },
-    {
-      "epoch": 1.3374852923276666,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.238924236212819e-06,
-      "loss": 1.1163,
-      "step": 1376
-    },
-    {
-      "epoch": 1.3384590416683573,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.235607948486041e-06,
-      "loss": 1.1193,
-      "step": 1377
-    },
-    {
-      "epoch": 1.3394327910090478,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.232294647652048e-06,
-      "loss": 1.1106,
-      "step": 1378
-    },
-    {
-      "epoch": 1.3404065403497383,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.228984341537903e-06,
-      "loss": 1.1168,
-      "step": 1379
-    },
-    {
-      "epoch": 1.3413802896904288,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.2256770379635927e-06,
-      "loss": 1.1177,
-      "step": 1380
-    },
-    {
-      "epoch": 1.3423540390311195,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.2223727447420112e-06,
-      "loss": 1.1128,
-      "step": 1381
-    },
-    {
-      "epoch": 1.34332778837181,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.219071469678941e-06,
-      "loss": 1.1258,
-      "step": 1382
-    },
-    {
-      "epoch": 1.3443015377125005,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.2157732205730355e-06,
-      "loss": 1.1162,
-      "step": 1383
-    },
-    {
-      "epoch": 1.345275287053191,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.2124780052157996e-06,
-      "loss": 1.104,
-      "step": 1384
-    },
-    {
-      "epoch": 1.3462490363938815,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.2091858313915702e-06,
-      "loss": 1.0959,
-      "step": 1385
-    },
-    {
-      "epoch": 1.3472227857345722,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.2058967068775002e-06,
-      "loss": 1.1069,
-      "step": 1386
-    },
-    {
-      "epoch": 1.3481965350752627,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.2026106394435402e-06,
-      "loss": 1.1325,
-      "step": 1387
-    },
-    {
-      "epoch": 1.3491702844159532,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.1993276368524154e-06,
-      "loss": 1.1096,
-      "step": 1388
-    },
-    {
-      "epoch": 1.350144033756644,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.1960477068596155e-06,
-      "loss": 1.1081,
-      "step": 1389
-    },
-    {
-      "epoch": 1.3511177830973344,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.19277085721337e-06,
-      "loss": 1.0919,
-      "step": 1390
-    },
-    {
-      "epoch": 1.352091532438025,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1894970956546296e-06,
-      "loss": 1.1151,
-      "step": 1391
-    },
-    {
-      "epoch": 1.3530652817787154,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.1862264299170517e-06,
-      "loss": 1.1098,
-      "step": 1392
-    },
-    {
-      "epoch": 1.354039031119406,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1829588677269808e-06,
-      "loss": 1.1138,
-      "step": 1393
-    },
-    {
-      "epoch": 1.3550127804600964,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.17969441680343e-06,
-      "loss": 1.1007,
-      "step": 1394
-    },
-    {
-      "epoch": 1.3559865298007872,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.1764330848580603e-06,
-      "loss": 1.1099,
-      "step": 1395
-    },
-    {
-      "epoch": 1.3569602791414777,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.173174879595166e-06,
-      "loss": 1.1043,
-      "step": 1396
-    },
-    {
-      "epoch": 1.3579340284821682,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.169919808711659e-06,
-      "loss": 1.087,
-      "step": 1397
-    },
-    {
-      "epoch": 1.3589077778228589,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.1666678798970397e-06,
-      "loss": 1.1186,
-      "step": 1398
-    },
-    {
-      "epoch": 1.3598815271635494,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1634191008333906e-06,
-      "loss": 1.1124,
-      "step": 1399
-    },
-    {
-      "epoch": 1.3608552765042399,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1601734791953537e-06,
-      "loss": 1.1132,
-      "step": 1400
-    },
-    {
-      "epoch": 1.3618290258449304,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.1569310226501088e-06,
-      "loss": 1.1327,
-      "step": 1401
-    },
-    {
-      "epoch": 1.3628027751856209,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.1536917388573618e-06,
-      "loss": 1.128,
-      "step": 1402
-    },
-    {
-      "epoch": 1.3637765245263116,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.1504556354693227e-06,
-      "loss": 1.1133,
-      "step": 1403
-    },
-    {
-      "epoch": 1.364750273867002,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.1472227201306885e-06,
-      "loss": 1.123,
-      "step": 1404
-    },
-    {
-      "epoch": 1.3657240232076926,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.1439930004786242e-06,
-      "loss": 1.1264,
-      "step": 1405
-    },
-    {
-      "epoch": 1.366697772548383,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.1407664841427472e-06,
-      "loss": 1.1272,
-      "step": 1406
-    },
-    {
-      "epoch": 1.3676715218890738,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.1375431787451063e-06,
-      "loss": 1.1211,
-      "step": 1407
-    },
-    {
-      "epoch": 1.3686452712297643,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.1343230919001647e-06,
-      "loss": 1.1031,
-      "step": 1408
-    },
-    {
-      "epoch": 1.3696190205704548,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1311062312147833e-06,
-      "loss": 1.1072,
-      "step": 1409
-    },
-    {
-      "epoch": 1.3705927699111453,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1278926042882026e-06,
-      "loss": 1.1109,
-      "step": 1410
-    },
-    {
-      "epoch": 1.3715665192518358,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.1246822187120215e-06,
-      "loss": 1.1325,
-      "step": 1411
-    },
-    {
-      "epoch": 1.3725402685925265,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1214750820701828e-06,
-      "loss": 1.1108,
-      "step": 1412
-    },
-    {
-      "epoch": 1.373514017933217,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.1182712019389574e-06,
-      "loss": 1.0976,
-      "step": 1413
-    },
-    {
-      "epoch": 1.3744877672739075,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.115070585886918e-06,
-      "loss": 1.1177,
-      "step": 1414
-    },
-    {
-      "epoch": 1.3754615166145983,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.1118732414749302e-06,
-      "loss": 1.1044,
-      "step": 1415
-    },
-    {
-      "epoch": 1.3764352659552888,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1086791762561298e-06,
-      "loss": 1.1232,
-      "step": 1416
-    },
-    {
-      "epoch": 1.3774090152959793,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.1054883977759067e-06,
-      "loss": 1.1178,
-      "step": 1417
-    },
-    {
-      "epoch": 1.3783827646366698,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1023009135718845e-06,
-      "loss": 1.0969,
-      "step": 1418
-    },
-    {
-      "epoch": 1.3793565139773603,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.099116731173907e-06,
-      "loss": 1.1124,
-      "step": 1419
-    },
-    {
-      "epoch": 1.3803302633180508,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.0959358581040171e-06,
-      "loss": 1.117,
-      "step": 1420
-    },
-    {
-      "epoch": 1.3813040126587415,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.0927583018764406e-06,
-      "loss": 1.1338,
-      "step": 1421
-    },
-    {
-      "epoch": 1.382277761999432,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.089584069997567e-06,
-      "loss": 1.1064,
-      "step": 1422
-    },
-    {
-      "epoch": 1.3832515113401225,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.0864131699659345e-06,
-      "loss": 1.096,
-      "step": 1423
-    },
-    {
-      "epoch": 1.3842252606808132,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.0832456092722063e-06,
-      "loss": 1.1245,
-      "step": 1424
-    },
-    {
-      "epoch": 1.3851990100215037,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.0800813953991615e-06,
-      "loss": 1.1274,
-      "step": 1425
-    },
-    {
-      "epoch": 1.3861727593621942,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.0769205358216707e-06,
-      "loss": 1.1171,
-      "step": 1426
-    },
-    {
-      "epoch": 1.3871465087028847,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.0737630380066822e-06,
-      "loss": 1.0838,
-      "step": 1427
-    },
-    {
-      "epoch": 1.3881202580435752,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.0706089094131988e-06,
-      "loss": 1.1079,
-      "step": 1428
-    },
-    {
-      "epoch": 1.389094007384266,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.0674581574922698e-06,
-      "loss": 1.1162,
-      "step": 1429
-    },
-    {
-      "epoch": 1.3900677567249564,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.064310789686965e-06,
-      "loss": 1.1008,
-      "step": 1430
-    },
-    {
-      "epoch": 1.391041506065647,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.0611668134323577e-06,
-      "loss": 1.1008,
-      "step": 1431
-    },
-    {
-      "epoch": 1.3920152554063374,
-      "grad_norm": 0.2890625,
-      "learning_rate": 1.0580262361555127e-06,
-      "loss": 1.0925,
-      "step": 1432
-    },
-    {
-      "epoch": 1.3929890047470281,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.0548890652754643e-06,
-      "loss": 1.1139,
-      "step": 1433
-    },
-    {
-      "epoch": 1.3939627540877186,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.051755308203198e-06,
-      "loss": 1.1031,
-      "step": 1434
-    },
-    {
-      "epoch": 1.3949365034284091,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.0486249723416371e-06,
-      "loss": 1.0898,
-      "step": 1435
-    },
-    {
-      "epoch": 1.3959102527690996,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.0454980650856226e-06,
-      "loss": 1.1144,
-      "step": 1436
-    },
-    {
-      "epoch": 1.3968840021097901,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.0423745938218951e-06,
-      "loss": 1.1158,
-      "step": 1437
-    },
-    {
-      "epoch": 1.3978577514504809,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.0392545659290789e-06,
-      "loss": 1.0878,
-      "step": 1438
-    },
-    {
-      "epoch": 1.3988315007911714,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.036137988777664e-06,
-      "loss": 1.1157,
-      "step": 1439
-    },
-    {
-      "epoch": 1.3998052501318619,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.0330248697299893e-06,
-      "loss": 1.1215,
-      "step": 1440
-    },
-    {
-      "epoch": 1.4007789994725526,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.029915216140222e-06,
-      "loss": 1.1074,
-      "step": 1441
-    },
-    {
-      "epoch": 1.401752748813243,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.0268090353543456e-06,
-      "loss": 1.1071,
-      "step": 1442
-    },
-    {
-      "epoch": 1.4027264981539336,
-      "grad_norm": 0.34375,
-      "learning_rate": 1.0237063347101394e-06,
-      "loss": 1.1089,
-      "step": 1443
-    },
-    {
-      "epoch": 1.403700247494624,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.0206071215371588e-06,
-      "loss": 1.1046,
-      "step": 1444
-    },
-    {
-      "epoch": 1.4046739968353146,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.0175114031567246e-06,
-      "loss": 1.1176,
-      "step": 1445
-    },
-    {
-      "epoch": 1.405647746176005,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.0144191868819003e-06,
-      "loss": 1.1158,
-      "step": 1446
-    },
-    {
-      "epoch": 1.4066214955166958,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.0113304800174743e-06,
-      "loss": 1.1228,
-      "step": 1447
-    },
-    {
-      "epoch": 1.4075952448573863,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.0082452898599468e-06,
-      "loss": 1.0977,
-      "step": 1448
-    },
-    {
-      "epoch": 1.4085689941980768,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.0051636236975105e-06,
-      "loss": 1.1136,
-      "step": 1449
-    },
-    {
-      "epoch": 1.4095427435387675,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.002085488810033e-06,
-      "loss": 1.0944,
-      "step": 1450
-    },
-    {
-      "epoch": 1.410516492879458,
-      "grad_norm": 0.30078125,
-      "learning_rate": 9.990108924690386e-07,
-      "loss": 1.1053,
-      "step": 1451
-    },
-    {
-      "epoch": 1.4114902422201485,
-      "grad_norm": 0.296875,
-      "learning_rate": 9.95939841937693e-07,
-      "loss": 1.1287,
-      "step": 1452
-    },
-    {
-      "epoch": 1.412463991560839,
-      "grad_norm": 0.29296875,
-      "learning_rate": 9.928723444707895e-07,
-      "loss": 1.1046,
-      "step": 1453
-    },
-    {
-      "epoch": 1.4134377409015295,
-      "grad_norm": 0.302734375,
-      "learning_rate": 9.898084073147207e-07,
-      "loss": 1.1119,
-      "step": 1454
-    },
-    {
-      "epoch": 1.41441149024222,
-      "grad_norm": 0.302734375,
-      "learning_rate": 9.867480377074742e-07,
-      "loss": 1.1054,
-      "step": 1455
-    },
-    {
-      "epoch": 1.4153852395829107,
-      "grad_norm": 0.30078125,
-      "learning_rate": 9.836912428786086e-07,
-      "loss": 1.1022,
-      "step": 1456
-    },
-    {
-      "epoch": 1.4163589889236012,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.80638030049236e-07,
-      "loss": 1.1176,
-      "step": 1457
-    },
-    {
-      "epoch": 1.4173327382642917,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.775884064320087e-07,
-      "loss": 1.0989,
-      "step": 1458
-    },
-    {
-      "epoch": 1.4183064876049825,
-      "grad_norm": 0.298828125,
-      "learning_rate": 9.745423792310996e-07,
-      "loss": 1.1276,
-      "step": 1459
-    },
-    {
-      "epoch": 1.419280236945673,
-      "grad_norm": 0.306640625,
-      "learning_rate": 9.71499955642185e-07,
-      "loss": 1.1279,
-      "step": 1460
-    },
-    {
-      "epoch": 1.4202539862863635,
-      "grad_norm": 0.29296875,
-      "learning_rate": 9.684611428524298e-07,
-      "loss": 1.1057,
-      "step": 1461
-    },
-    {
-      "epoch": 1.421227735627054,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.654259480404676e-07,
-      "loss": 1.1116,
-      "step": 1462
-    },
-    {
-      "epoch": 1.4222014849677445,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.62394378376387e-07,
-      "loss": 1.108,
-      "step": 1463
-    },
-    {
-      "epoch": 1.4231752343084352,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.593664410217092e-07,
-      "loss": 1.1224,
-      "step": 1464
-    },
-    {
-      "epoch": 1.4241489836491257,
-      "grad_norm": 0.30078125,
-      "learning_rate": 9.563421431293794e-07,
-      "loss": 1.1257,
-      "step": 1465
-    },
-    {
-      "epoch": 1.4251227329898162,
-      "grad_norm": 0.296875,
-      "learning_rate": 9.533214918437422e-07,
-      "loss": 1.1039,
-      "step": 1466
-    },
-    {
-      "epoch": 1.4260964823305067,
-      "grad_norm": 0.306640625,
-      "learning_rate": 9.503044943005283e-07,
-      "loss": 1.1298,
-      "step": 1467
-    },
-    {
-      "epoch": 1.4270702316711974,
-      "grad_norm": 0.302734375,
-      "learning_rate": 9.472911576268365e-07,
-      "loss": 1.1206,
-      "step": 1468
-    },
-    {
-      "epoch": 1.428043981011888,
-      "grad_norm": 0.30078125,
-      "learning_rate": 9.442814889411211e-07,
-      "loss": 1.0996,
-      "step": 1469
-    },
-    {
-      "epoch": 1.4290177303525784,
-      "grad_norm": 0.3046875,
-      "learning_rate": 9.412754953531664e-07,
-      "loss": 1.1175,
-      "step": 1470
-    },
-    {
-      "epoch": 1.429991479693269,
-      "grad_norm": 0.306640625,
-      "learning_rate": 9.38273183964078e-07,
-      "loss": 1.1262,
-      "step": 1471
-    },
-    {
-      "epoch": 1.4309652290339594,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.352745618662623e-07,
-      "loss": 1.1026,
-      "step": 1472
-    },
-    {
-      "epoch": 1.4319389783746501,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.322796361434111e-07,
-      "loss": 1.1134,
-      "step": 1473
-    },
-    {
-      "epoch": 1.4329127277153406,
-      "grad_norm": 0.296875,
-      "learning_rate": 9.292884138704819e-07,
-      "loss": 1.1112,
-      "step": 1474
-    },
-    {
-      "epoch": 1.4338864770560311,
-      "grad_norm": 0.296875,
-      "learning_rate": 9.263009021136859e-07,
-      "loss": 1.0991,
-      "step": 1475
-    },
-    {
-      "epoch": 1.4348602263967218,
-      "grad_norm": 0.296875,
-      "learning_rate": 9.233171079304684e-07,
-      "loss": 1.1029,
-      "step": 1476
-    },
-    {
-      "epoch": 1.4358339757374123,
-      "grad_norm": 0.30078125,
-      "learning_rate": 9.203370383694915e-07,
-      "loss": 1.1236,
-      "step": 1477
-    },
-    {
-      "epoch": 1.4368077250781028,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.1736070047062e-07,
-      "loss": 1.0972,
-      "step": 1478
-    },
-    {
-      "epoch": 1.4377814744187933,
-      "grad_norm": 0.302734375,
-      "learning_rate": 9.143881012649033e-07,
-      "loss": 1.1104,
-      "step": 1479
-    },
-    {
-      "epoch": 1.4387552237594838,
-      "grad_norm": 0.29296875,
-      "learning_rate": 9.114192477745568e-07,
-      "loss": 1.0761,
-      "step": 1480
-    },
-    {
-      "epoch": 1.4397289731001743,
-      "grad_norm": 0.310546875,
-      "learning_rate": 9.084541470129491e-07,
-      "loss": 1.1074,
-      "step": 1481
-    },
-    {
-      "epoch": 1.440702722440865,
-      "grad_norm": 0.298828125,
-      "learning_rate": 9.054928059845836e-07,
-      "loss": 1.1008,
-      "step": 1482
-    },
-    {
-      "epoch": 1.4416764717815556,
-      "grad_norm": 0.298828125,
-      "learning_rate": 9.025352316850827e-07,
-      "loss": 1.1095,
-      "step": 1483
-    },
-    {
-      "epoch": 1.442650221122246,
-      "grad_norm": 0.30078125,
-      "learning_rate": 8.995814311011669e-07,
-      "loss": 1.1202,
-      "step": 1484
-    },
-    {
-      "epoch": 1.4436239704629368,
-      "grad_norm": 0.30078125,
-      "learning_rate": 8.966314112106475e-07,
-      "loss": 1.1013,
-      "step": 1485
-    },
-    {
-      "epoch": 1.4445977198036273,
-      "grad_norm": 0.29296875,
-      "learning_rate": 8.936851789824017e-07,
-      "loss": 1.1047,
-      "step": 1486
-    },
-    {
-      "epoch": 1.4455714691443178,
-      "grad_norm": 0.291015625,
-      "learning_rate": 8.907427413763572e-07,
-      "loss": 1.0974,
-      "step": 1487
-    },
-    {
-      "epoch": 1.4465452184850083,
-      "grad_norm": 0.296875,
-      "learning_rate": 8.87804105343481e-07,
-      "loss": 1.1165,
-      "step": 1488
-    },
-    {
-      "epoch": 1.4475189678256988,
-      "grad_norm": 0.30859375,
-      "learning_rate": 8.848692778257589e-07,
-      "loss": 1.1048,
-      "step": 1489
-    },
-    {
-      "epoch": 1.4484927171663895,
-      "grad_norm": 0.3046875,
-      "learning_rate": 8.81938265756177e-07,
-      "loss": 1.0946,
-      "step": 1490
-    },
-    {
-      "epoch": 1.44946646650708,
-      "grad_norm": 0.3046875,
-      "learning_rate": 8.790110760587117e-07,
-      "loss": 1.1094,
-      "step": 1491
-    },
-    {
-      "epoch": 1.4504402158477705,
-      "grad_norm": 0.30078125,
-      "learning_rate": 8.760877156483078e-07,
-      "loss": 1.1113,
-      "step": 1492
-    },
-    {
-      "epoch": 1.451413965188461,
-      "grad_norm": 0.296875,
-      "learning_rate": 8.731681914308643e-07,
-      "loss": 1.1163,
-      "step": 1493
-    },
-    {
-      "epoch": 1.4523877145291517,
-      "grad_norm": 0.29296875,
-      "learning_rate": 8.702525103032186e-07,
-      "loss": 1.1079,
-      "step": 1494
-    },
-    {
-      "epoch": 1.4533614638698422,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.673406791531292e-07,
-      "loss": 1.1144,
-      "step": 1495
-    },
-    {
-      "epoch": 1.4543352132105327,
-      "grad_norm": 0.30078125,
-      "learning_rate": 8.644327048592596e-07,
-      "loss": 1.1114,
-      "step": 1496
-    },
-    {
-      "epoch": 1.4553089625512232,
-      "grad_norm": 0.2890625,
-      "learning_rate": 8.615285942911616e-07,
-      "loss": 1.114,
-      "step": 1497
-    },
-    {
-      "epoch": 1.4562827118919137,
-      "grad_norm": 0.296875,
-      "learning_rate": 8.586283543092603e-07,
-      "loss": 1.1074,
-      "step": 1498
-    },
-    {
-      "epoch": 1.4572564612326044,
-      "grad_norm": 0.294921875,
-      "learning_rate": 8.557319917648385e-07,
-      "loss": 1.1096,
-      "step": 1499
-    },
-    {
-      "epoch": 1.458230210573295,
-      "grad_norm": 0.294921875,
-      "learning_rate": 8.528395135000151e-07,
-      "loss": 1.1129,
-      "step": 1500
-    },
-    {
-      "epoch": 1.4592039599139854,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.499509263477388e-07,
-      "loss": 1.1201,
-      "step": 1501
-    },
-    {
-      "epoch": 1.4601777092546762,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.470662371317631e-07,
-      "loss": 1.1133,
-      "step": 1502
-    },
-    {
-      "epoch": 1.4611514585953667,
-      "grad_norm": 0.302734375,
-      "learning_rate": 8.441854526666324e-07,
-      "loss": 1.1143,
-      "step": 1503
-    },
-    {
-      "epoch": 1.4621252079360572,
-      "grad_norm": 0.296875,
-      "learning_rate": 8.41308579757669e-07,
-      "loss": 1.1227,
-      "step": 1504
-    },
-    {
-      "epoch": 1.4630989572767477,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.384356252009534e-07,
-      "loss": 1.115,
-      "step": 1505
-    },
-    {
-      "epoch": 1.4640727066174382,
-      "grad_norm": 0.296875,
-      "learning_rate": 8.35566595783312e-07,
-      "loss": 1.105,
-      "step": 1506
-    },
-    {
-      "epoch": 1.4650464559581287,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.327014982822948e-07,
-      "loss": 1.1137,
-      "step": 1507
-    },
-    {
-      "epoch": 1.4660202052988194,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.298403394661658e-07,
-      "loss": 1.1059,
-      "step": 1508
-    },
-    {
-      "epoch": 1.4669939546395099,
-      "grad_norm": 0.29296875,
-      "learning_rate": 8.269831260938873e-07,
-      "loss": 1.1023,
-      "step": 1509
-    },
-    {
-      "epoch": 1.4679677039802004,
-      "grad_norm": 0.296875,
-      "learning_rate": 8.241298649150958e-07,
-      "loss": 1.1034,
-      "step": 1510
-    },
-    {
-      "epoch": 1.468941453320891,
-      "grad_norm": 0.294921875,
-      "learning_rate": 8.212805626700945e-07,
-      "loss": 1.1282,
-      "step": 1511
-    },
-    {
-      "epoch": 1.4699152026615816,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.184352260898351e-07,
-      "loss": 1.108,
-      "step": 1512
-    },
-    {
-      "epoch": 1.470888952002272,
-      "grad_norm": 0.296875,
-      "learning_rate": 8.155938618958983e-07,
-      "loss": 1.0938,
-      "step": 1513
-    },
-    {
-      "epoch": 1.4718627013429626,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.12756476800483e-07,
-      "loss": 1.1223,
-      "step": 1514
-    },
-    {
-      "epoch": 1.472836450683653,
-      "grad_norm": 0.29296875,
-      "learning_rate": 8.099230775063879e-07,
-      "loss": 1.093,
-      "step": 1515
-    },
-    {
-      "epoch": 1.4738102000243436,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.070936707069957e-07,
-      "loss": 1.1251,
-      "step": 1516
-    },
-    {
-      "epoch": 1.4747839493650343,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.042682630862572e-07,
-      "loss": 1.1226,
-      "step": 1517
-    },
-    {
-      "epoch": 1.4757576987057248,
-      "grad_norm": 0.296875,
-      "learning_rate": 8.014468613186768e-07,
-      "loss": 1.1099,
-      "step": 1518
-    },
-    {
-      "epoch": 1.4767314480464153,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.986294720692958e-07,
-      "loss": 1.1073,
-      "step": 1519
-    },
-    {
-      "epoch": 1.477705197387106,
-      "grad_norm": 0.302734375,
-      "learning_rate": 7.958161019936747e-07,
-      "loss": 1.1234,
-      "step": 1520
-    },
-    {
-      "epoch": 1.4786789467277965,
-      "grad_norm": 0.30078125,
-      "learning_rate": 7.930067577378817e-07,
-      "loss": 1.1034,
-      "step": 1521
-    },
-    {
-      "epoch": 1.479652696068487,
-      "grad_norm": 0.291015625,
-      "learning_rate": 7.902014459384744e-07,
-      "loss": 1.119,
-      "step": 1522
-    },
-    {
-      "epoch": 1.4806264454091775,
-      "grad_norm": 0.287109375,
-      "learning_rate": 7.874001732224829e-07,
-      "loss": 1.1132,
-      "step": 1523
-    },
-    {
-      "epoch": 1.481600194749868,
-      "grad_norm": 0.296875,
-      "learning_rate": 7.846029462073962e-07,
-      "loss": 1.1071,
-      "step": 1524
-    },
-    {
-      "epoch": 1.4825739440905588,
-      "grad_norm": 0.29296875,
-      "learning_rate": 7.818097715011496e-07,
-      "loss": 1.11,
-      "step": 1525
-    },
-    {
-      "epoch": 1.4835476934312493,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.790206557021001e-07,
-      "loss": 1.1016,
-      "step": 1526
-    },
-    {
-      "epoch": 1.4845214427719398,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.762356053990197e-07,
-      "loss": 1.1194,
-      "step": 1527
-    },
-    {
-      "epoch": 1.4854951921126303,
-      "grad_norm": 0.29296875,
-      "learning_rate": 7.734546271710755e-07,
-      "loss": 1.1116,
-      "step": 1528
-    },
-    {
-      "epoch": 1.486468941453321,
-      "grad_norm": 0.3046875,
-      "learning_rate": 7.706777275878161e-07,
-      "loss": 1.1066,
-      "step": 1529
-    },
-    {
-      "epoch": 1.4874426907940115,
-      "grad_norm": 0.30078125,
-      "learning_rate": 7.67904913209152e-07,
-      "loss": 1.122,
-      "step": 1530
-    },
-    {
-      "epoch": 1.488416440134702,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.651361905853466e-07,
-      "loss": 1.12,
-      "step": 1531
-    },
-    {
-      "epoch": 1.4893901894753925,
-      "grad_norm": 0.3046875,
-      "learning_rate": 7.623715662569955e-07,
-      "loss": 1.1295,
-      "step": 1532
-    },
-    {
-      "epoch": 1.490363938816083,
-      "grad_norm": 0.30078125,
-      "learning_rate": 7.596110467550136e-07,
-      "loss": 1.1232,
-      "step": 1533
-    },
-    {
-      "epoch": 1.4913376881567737,
-      "grad_norm": 0.302734375,
-      "learning_rate": 7.568546386006185e-07,
-      "loss": 1.1285,
-      "step": 1534
-    },
-    {
-      "epoch": 1.4923114374974642,
-      "grad_norm": 0.302734375,
-      "learning_rate": 7.541023483053162e-07,
-      "loss": 1.1139,
-      "step": 1535
-    },
-    {
-      "epoch": 1.4932851868381547,
-      "grad_norm": 0.302734375,
-      "learning_rate": 7.513541823708828e-07,
-      "loss": 1.1002,
-      "step": 1536
-    },
-    {
-      "epoch": 1.4942589361788454,
-      "grad_norm": 0.30078125,
-      "learning_rate": 7.48610147289354e-07,
-      "loss": 1.1188,
-      "step": 1537
-    },
-    {
-      "epoch": 1.495232685519536,
-      "grad_norm": 0.296875,
-      "learning_rate": 7.458702495430056e-07,
-      "loss": 1.1045,
-      "step": 1538
-    },
-    {
-      "epoch": 1.4962064348602264,
-      "grad_norm": 0.298828125,
-      "learning_rate": 7.431344956043415e-07,
-      "loss": 1.1021,
-      "step": 1539
-    },
-    {
-      "epoch": 1.497180184200917,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.404028919360726e-07,
-      "loss": 1.1085,
-      "step": 1540
-    },
-    {
-      "epoch": 1.4981539335416074,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.376754449911108e-07,
-      "loss": 1.1166,
-      "step": 1541
-    },
-    {
-      "epoch": 1.499127682882298,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.349521612125459e-07,
-      "loss": 1.1027,
-      "step": 1542
-    },
-    {
-      "epoch": 1.5001014322229886,
-      "grad_norm": 0.30078125,
-      "learning_rate": 7.322330470336314e-07,
-      "loss": 1.1102,
-      "step": 1543
-    },
-    {
-      "epoch": 1.5010751815636791,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.295181088777734e-07,
-      "loss": 1.1238,
-      "step": 1544
-    },
-    {
-      "epoch": 1.5020489309043699,
-      "grad_norm": 0.30078125,
-      "learning_rate": 7.268073531585126e-07,
-      "loss": 1.1267,
-      "step": 1545
-    },
-    {
-      "epoch": 1.5030226802450604,
-      "grad_norm": 0.296875,
-      "learning_rate": 7.241007862795079e-07,
-      "loss": 1.0929,
-      "step": 1546
-    },
-    {
-      "epoch": 1.5039964295857509,
-      "grad_norm": 0.296875,
-      "learning_rate": 7.213984146345235e-07,
-      "loss": 1.1014,
-      "step": 1547
-    },
-    {
-      "epoch": 1.5049701789264414,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.187002446074156e-07,
-      "loss": 1.1184,
-      "step": 1548
-    },
-    {
-      "epoch": 1.5059439282671319,
-      "grad_norm": 0.302734375,
-      "learning_rate": 7.16006282572111e-07,
-      "loss": 1.1094,
-      "step": 1549
-    },
-    {
-      "epoch": 1.5069176776078224,
-      "grad_norm": 0.296875,
-      "learning_rate": 7.133165348925978e-07,
-      "loss": 1.1025,
-      "step": 1550
-    },
-    {
-      "epoch": 1.5078914269485129,
-      "grad_norm": 0.29296875,
-      "learning_rate": 7.106310079229084e-07,
-      "loss": 1.0988,
-      "step": 1551
-    },
-    {
-      "epoch": 1.5088651762892036,
-      "grad_norm": 0.322265625,
-      "learning_rate": 7.079497080071057e-07,
-      "loss": 1.1063,
-      "step": 1552
-    },
-    {
-      "epoch": 1.509838925629894,
-      "grad_norm": 0.291015625,
-      "learning_rate": 7.052726414792637e-07,
-      "loss": 1.106,
-      "step": 1553
-    },
-    {
-      "epoch": 1.5108126749705848,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.025998146634591e-07,
-      "loss": 1.1027,
-      "step": 1554
-    },
-    {
-      "epoch": 1.5117864243112753,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.999312338737513e-07,
-      "loss": 1.1304,
-      "step": 1555
-    },
-    {
-      "epoch": 1.5127601736519658,
-      "grad_norm": 0.29296875,
-      "learning_rate": 6.972669054141701e-07,
-      "loss": 1.1029,
-      "step": 1556
-    },
-    {
-      "epoch": 1.5137339229926563,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.946068355786992e-07,
-      "loss": 1.0985,
-      "step": 1557
-    },
-    {
-      "epoch": 1.5147076723333468,
-      "grad_norm": 0.302734375,
-      "learning_rate": 6.919510306512633e-07,
-      "loss": 1.1217,
-      "step": 1558
-    },
-    {
-      "epoch": 1.5156814216740373,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.892994969057101e-07,
-      "loss": 1.1093,
-      "step": 1559
-    },
-    {
-      "epoch": 1.5166551710147278,
-      "grad_norm": 0.294921875,
-      "learning_rate": 6.86652240605799e-07,
-      "loss": 1.1157,
-      "step": 1560
-    },
-    {
-      "epoch": 1.5176289203554185,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.840092680051844e-07,
-      "loss": 1.1153,
-      "step": 1561
-    },
-    {
-      "epoch": 1.518602669696109,
-      "grad_norm": 0.3046875,
-      "learning_rate": 6.813705853474014e-07,
-      "loss": 1.1085,
-      "step": 1562
-    },
-    {
-      "epoch": 1.5195764190367997,
-      "grad_norm": 0.294921875,
-      "learning_rate": 6.787361988658489e-07,
-      "loss": 1.0963,
-      "step": 1563
-    },
-    {
-      "epoch": 1.5205501683774902,
-      "grad_norm": 0.29296875,
-      "learning_rate": 6.761061147837808e-07,
-      "loss": 1.1052,
-      "step": 1564
-    },
-    {
-      "epoch": 1.5215239177181807,
-      "grad_norm": 0.294921875,
-      "learning_rate": 6.734803393142849e-07,
-      "loss": 1.1201,
-      "step": 1565
-    },
-    {
-      "epoch": 1.5224976670588712,
-      "grad_norm": 0.29296875,
-      "learning_rate": 6.708588786602699e-07,
-      "loss": 1.1069,
-      "step": 1566
-    },
-    {
-      "epoch": 1.5234714163995617,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.68241739014453e-07,
-      "loss": 1.111,
-      "step": 1567
-    },
-    {
-      "epoch": 1.5244451657402522,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.656289265593441e-07,
-      "loss": 1.1024,
-      "step": 1568
-    },
-    {
-      "epoch": 1.525418915080943,
-      "grad_norm": 0.30078125,
-      "learning_rate": 6.630204474672289e-07,
-      "loss": 1.1128,
-      "step": 1569
-    },
-    {
-      "epoch": 1.5263926644216335,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.604163079001586e-07,
-      "loss": 1.1082,
-      "step": 1570
-    },
-    {
-      "epoch": 1.527366413762324,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.578165140099318e-07,
-      "loss": 1.1091,
-      "step": 1571
-    },
-    {
-      "epoch": 1.5283401631030147,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.55221071938082e-07,
-      "loss": 1.1204,
-      "step": 1572
-    },
-    {
-      "epoch": 1.5293139124437052,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.526299878158615e-07,
-      "loss": 1.1016,
-      "step": 1573
-    },
-    {
-      "epoch": 1.5302876617843957,
-      "grad_norm": 0.294921875,
-      "learning_rate": 6.500432677642282e-07,
-      "loss": 1.1115,
-      "step": 1574
-    },
-    {
-      "epoch": 1.5312614111250862,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.474609178938318e-07,
-      "loss": 1.1099,
-      "step": 1575
-    },
-    {
-      "epoch": 1.5322351604657767,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.448829443049953e-07,
-      "loss": 1.1179,
-      "step": 1576
-    },
-    {
-      "epoch": 1.5332089098064672,
-      "grad_norm": 0.3046875,
-      "learning_rate": 6.423093530877062e-07,
-      "loss": 1.1121,
-      "step": 1577
-    },
-    {
-      "epoch": 1.534182659147158,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.397401503215992e-07,
-      "loss": 1.1062,
-      "step": 1578
-    },
-    {
-      "epoch": 1.5351564084878484,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.371753420759394e-07,
-      "loss": 1.1087,
-      "step": 1579
-    },
-    {
-      "epoch": 1.5361301578285391,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.346149344096151e-07,
-      "loss": 1.0994,
-      "step": 1580
-    },
-    {
-      "epoch": 1.5371039071692296,
-      "grad_norm": 0.3046875,
-      "learning_rate": 6.320589333711163e-07,
-      "loss": 1.1403,
-      "step": 1581
-    },
-    {
-      "epoch": 1.5380776565099201,
-      "grad_norm": 0.291015625,
-      "learning_rate": 6.295073449985223e-07,
-      "loss": 1.0955,
-      "step": 1582
-    },
-    {
-      "epoch": 1.5390514058506106,
-      "grad_norm": 0.294921875,
-      "learning_rate": 6.269601753194907e-07,
-      "loss": 1.1107,
-      "step": 1583
-    },
-    {
-      "epoch": 1.5400251551913011,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.244174303512391e-07,
-      "loss": 1.1051,
-      "step": 1584
-    },
-    {
-      "epoch": 1.5409989045319916,
-      "grad_norm": 0.294921875,
-      "learning_rate": 6.218791161005336e-07,
-      "loss": 1.1048,
-      "step": 1585
-    },
-    {
-      "epoch": 1.5419726538726821,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.193452385636722e-07,
-      "loss": 1.0947,
-      "step": 1586
-    },
-    {
-      "epoch": 1.5429464032133728,
-      "grad_norm": 0.29296875,
-      "learning_rate": 6.168158037264721e-07,
-      "loss": 1.1161,
-      "step": 1587
-    },
-    {
-      "epoch": 1.5439201525540633,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.14290817564259e-07,
-      "loss": 1.114,
-      "step": 1588
-    },
-    {
-      "epoch": 1.544893901894754,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.117702860418434e-07,
-      "loss": 1.1295,
-      "step": 1589
-    },
-    {
-      "epoch": 1.5458676512354446,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.092542151135175e-07,
-      "loss": 1.1165,
-      "step": 1590
-    },
-    {
-      "epoch": 1.546841400576135,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.067426107230342e-07,
-      "loss": 1.1086,
-      "step": 1591
-    },
-    {
-      "epoch": 1.5478151499168256,
-      "grad_norm": 0.294921875,
-      "learning_rate": 6.042354788035943e-07,
-      "loss": 1.1151,
-      "step": 1592
-    },
-    {
-      "epoch": 1.548788899257516,
-      "grad_norm": 0.294921875,
-      "learning_rate": 6.017328252778343e-07,
-      "loss": 1.1167,
-      "step": 1593
-    },
-    {
-      "epoch": 1.5497626485982066,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.992346560578113e-07,
-      "loss": 1.1019,
-      "step": 1594
-    },
-    {
-      "epoch": 1.5507363979388973,
-      "grad_norm": 0.30078125,
-      "learning_rate": 5.967409770449891e-07,
-      "loss": 1.1353,
-      "step": 1595
-    },
-    {
-      "epoch": 1.5517101472795878,
-      "grad_norm": 0.298828125,
-      "learning_rate": 5.942517941302239e-07,
-      "loss": 1.1085,
-      "step": 1596
-    },
-    {
-      "epoch": 1.5526838966202783,
-      "grad_norm": 0.298828125,
-      "learning_rate": 5.917671131937505e-07,
-      "loss": 1.1096,
-      "step": 1597
-    },
-    {
-      "epoch": 1.553657645960969,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.892869401051698e-07,
-      "loss": 1.1153,
-      "step": 1598
-    },
-    {
-      "epoch": 1.5546313953016595,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.868112807234313e-07,
-      "loss": 1.1267,
-      "step": 1599
-    },
-    {
-      "epoch": 1.55560514464235,
-      "grad_norm": 0.287109375,
-      "learning_rate": 5.843401408968244e-07,
-      "loss": 1.1078,
-      "step": 1600
-    },
-    {
-      "epoch": 1.5565788939830405,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.818735264629608e-07,
-      "loss": 1.0801,
-      "step": 1601
-    },
-    {
-      "epoch": 1.557552643323731,
-      "grad_norm": 0.3046875,
-      "learning_rate": 5.794114432487605e-07,
-      "loss": 1.1221,
-      "step": 1602
-    },
-    {
-      "epoch": 1.5585263926644215,
-      "grad_norm": 0.29296875,
-      "learning_rate": 5.769538970704405e-07,
-      "loss": 1.091,
-      "step": 1603
-    },
-    {
-      "epoch": 1.5595001420051122,
-      "grad_norm": 0.298828125,
-      "learning_rate": 5.745008937335023e-07,
-      "loss": 1.1214,
-      "step": 1604
-    },
-    {
-      "epoch": 1.5604738913458027,
-      "grad_norm": 0.29296875,
-      "learning_rate": 5.720524390327109e-07,
-      "loss": 1.1095,
-      "step": 1605
-    },
-    {
-      "epoch": 1.5614476406864934,
-      "grad_norm": 0.298828125,
-      "learning_rate": 5.696085387520894e-07,
-      "loss": 1.1215,
-      "step": 1606
-    },
-    {
-      "epoch": 1.562421390027184,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.671691986649011e-07,
-      "loss": 1.1102,
-      "step": 1607
-    },
-    {
-      "epoch": 1.5633951393678744,
-      "grad_norm": 0.298828125,
-      "learning_rate": 5.647344245336372e-07,
-      "loss": 1.137,
-      "step": 1608
-    },
-    {
-      "epoch": 1.564368888708565,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.623042221100008e-07,
-      "loss": 1.1046,
-      "step": 1609
-    },
-    {
-      "epoch": 1.5653426380492554,
-      "grad_norm": 0.302734375,
-      "learning_rate": 5.598785971348969e-07,
-      "loss": 1.1089,
-      "step": 1610
-    },
-    {
-      "epoch": 1.566316387389946,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.574575553384168e-07,
-      "loss": 1.0944,
-      "step": 1611
-    },
-    {
-      "epoch": 1.5672901367306364,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.550411024398247e-07,
-      "loss": 1.108,
-      "step": 1612
-    },
-    {
-      "epoch": 1.5682638860713272,
-      "grad_norm": 0.302734375,
-      "learning_rate": 5.526292441475448e-07,
-      "loss": 1.1193,
-      "step": 1613
-    },
-    {
-      "epoch": 1.5692376354120177,
-      "grad_norm": 0.29296875,
-      "learning_rate": 5.502219861591471e-07,
-      "loss": 1.0987,
-      "step": 1614
-    },
-    {
-      "epoch": 1.5702113847527084,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.478193341613333e-07,
-      "loss": 1.1086,
-      "step": 1615
-    },
-    {
-      "epoch": 1.5711851340933989,
-      "grad_norm": 0.30078125,
-      "learning_rate": 5.454212938299256e-07,
-      "loss": 1.1278,
-      "step": 1616
-    },
-    {
-      "epoch": 1.5721588834340894,
-      "grad_norm": 0.302734375,
-      "learning_rate": 5.430278708298517e-07,
-      "loss": 1.1181,
-      "step": 1617
-    },
-    {
-      "epoch": 1.5731326327747799,
-      "grad_norm": 0.29296875,
-      "learning_rate": 5.406390708151321e-07,
-      "loss": 1.1023,
-      "step": 1618
-    },
-    {
-      "epoch": 1.5741063821154704,
-      "grad_norm": 0.30078125,
-      "learning_rate": 5.382548994288639e-07,
-      "loss": 1.1081,
-      "step": 1619
-    },
-    {
-      "epoch": 1.5750801314561609,
-      "grad_norm": 0.298828125,
-      "learning_rate": 5.358753623032137e-07,
-      "loss": 1.0891,
-      "step": 1620
-    },
-    {
-      "epoch": 1.5760538807968514,
-      "grad_norm": 0.29296875,
-      "learning_rate": 5.335004650593989e-07,
-      "loss": 1.1097,
-      "step": 1621
-    },
-    {
-      "epoch": 1.577027630137542,
-      "grad_norm": 0.291015625,
-      "learning_rate": 5.311302133076742e-07,
-      "loss": 1.0929,
-      "step": 1622
-    },
-    {
-      "epoch": 1.5780013794782326,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.287646126473226e-07,
-      "loss": 1.1267,
-      "step": 1623
-    },
-    {
-      "epoch": 1.5789751288189233,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.264036686666391e-07,
-      "loss": 1.1193,
-      "step": 1624
-    },
-    {
-      "epoch": 1.5799488781596138,
-      "grad_norm": 0.29296875,
-      "learning_rate": 5.240473869429174e-07,
-      "loss": 1.1163,
-      "step": 1625
-    },
-    {
-      "epoch": 1.5809226275003043,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.216957730424382e-07,
-      "loss": 1.1101,
-      "step": 1626
-    },
-    {
-      "epoch": 1.5818963768409948,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.193488325204551e-07,
-      "loss": 1.1118,
-      "step": 1627
-    },
-    {
-      "epoch": 1.5828701261816853,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.170065709211824e-07,
-      "loss": 1.0995,
-      "step": 1628
-    },
-    {
-      "epoch": 1.5838438755223758,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.146689937777802e-07,
-      "loss": 1.1057,
-      "step": 1629
-    },
-    {
-      "epoch": 1.5848176248630665,
-      "grad_norm": 0.3046875,
-      "learning_rate": 5.123361066123428e-07,
-      "loss": 1.1193,
-      "step": 1630
-    },
-    {
-      "epoch": 1.585791374203757,
-      "grad_norm": 0.298828125,
-      "learning_rate": 5.100079149358864e-07,
-      "loss": 1.0986,
-      "step": 1631
-    },
-    {
-      "epoch": 1.5867651235444478,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.076844242483325e-07,
-      "loss": 1.1291,
-      "step": 1632
-    },
-    {
-      "epoch": 1.5877388728851383,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.053656400384998e-07,
-      "loss": 1.114,
-      "step": 1633
-    },
-    {
-      "epoch": 1.5887126222258288,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.030515677840883e-07,
-      "loss": 1.103,
-      "step": 1634
-    },
-    {
-      "epoch": 1.5896863715665193,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.007422129516648e-07,
-      "loss": 1.1057,
-      "step": 1635
-    },
-    {
-      "epoch": 1.5906601209072098,
-      "grad_norm": 0.287109375,
-      "learning_rate": 4.984375809966552e-07,
-      "loss": 1.0953,
-      "step": 1636
-    },
-    {
-      "epoch": 1.5916338702479003,
-      "grad_norm": 0.28515625,
-      "learning_rate": 4.961376773633273e-07,
-      "loss": 1.1185,
-      "step": 1637
-    },
-    {
-      "epoch": 1.5926076195885908,
-      "grad_norm": 0.291015625,
-      "learning_rate": 4.938425074847778e-07,
-      "loss": 1.1052,
-      "step": 1638
-    },
-    {
-      "epoch": 1.5935813689292815,
-      "grad_norm": 0.29296875,
-      "learning_rate": 4.915520767829219e-07,
-      "loss": 1.101,
-      "step": 1639
-    },
-    {
-      "epoch": 1.594555118269972,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.892663906684794e-07,
-      "loss": 1.1169,
-      "step": 1640
-    },
-    {
-      "epoch": 1.5955288676106627,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.869854545409627e-07,
-      "loss": 1.122,
-      "step": 1641
-    },
-    {
-      "epoch": 1.5965026169513532,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.847092737886608e-07,
-      "loss": 1.1114,
-      "step": 1642
-    },
-    {
-      "epoch": 1.5974763662920437,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.824378537886304e-07,
-      "loss": 1.1021,
-      "step": 1643
-    },
-    {
-      "epoch": 1.5984501156327342,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.801711999066841e-07,
-      "loss": 1.1257,
-      "step": 1644
-    },
-    {
-      "epoch": 1.5994238649734247,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.779093174973709e-07,
-      "loss": 1.1123,
-      "step": 1645
-    },
-    {
-      "epoch": 1.6003976143141152,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.756522119039714e-07,
-      "loss": 1.1184,
-      "step": 1646
-    },
-    {
-      "epoch": 1.6013713636548057,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.7339988845848147e-07,
-      "loss": 1.1143,
-      "step": 1647
-    },
-    {
-      "epoch": 1.6023451129954964,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.711523524815978e-07,
-      "loss": 1.1116,
-      "step": 1648
-    },
-    {
-      "epoch": 1.603318862336187,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.6890960928271026e-07,
-      "loss": 1.1084,
-      "step": 1649
-    },
-    {
-      "epoch": 1.6042926116768776,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.666716641598851e-07,
-      "loss": 1.1397,
-      "step": 1650
-    },
-    {
-      "epoch": 1.6052663610175681,
-      "grad_norm": 0.302734375,
-      "learning_rate": 4.6443852239985517e-07,
-      "loss": 1.1211,
-      "step": 1651
-    },
-    {
-      "epoch": 1.6062401103582586,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.622101892780048e-07,
-      "loss": 1.1128,
-      "step": 1652
-    },
-    {
-      "epoch": 1.6072138596989491,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.599866700583602e-07,
-      "loss": 1.1054,
-      "step": 1653
-    },
-    {
-      "epoch": 1.6081876090396396,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.577679699935755e-07,
-      "loss": 1.1151,
-      "step": 1654
-    },
-    {
-      "epoch": 1.6091613583803301,
-      "grad_norm": 0.291015625,
-      "learning_rate": 4.555540943249187e-07,
-      "loss": 1.0949,
-      "step": 1655
-    },
-    {
-      "epoch": 1.6101351077210209,
-      "grad_norm": 0.302734375,
-      "learning_rate": 4.5334504828226293e-07,
-      "loss": 1.1087,
-      "step": 1656
-    },
-    {
-      "epoch": 1.6111088570617114,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.511408370840722e-07,
-      "loss": 1.1149,
-      "step": 1657
-    },
-    {
-      "epoch": 1.6120826064024019,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.4894146593738724e-07,
-      "loss": 1.1164,
-      "step": 1658
-    },
-    {
-      "epoch": 1.6130563557430926,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.467469400378166e-07,
-      "loss": 1.1162,
-      "step": 1659
-    },
-    {
-      "epoch": 1.614030105083783,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.4455726456952397e-07,
-      "loss": 1.107,
-      "step": 1660
-    },
-    {
-      "epoch": 1.6150038544244736,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.423724447052119e-07,
-      "loss": 1.12,
-      "step": 1661
-    },
-    {
-      "epoch": 1.615977603765164,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.401924856061146e-07,
-      "loss": 1.1048,
-      "step": 1662
-    },
-    {
-      "epoch": 1.6169513531058546,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.380173924219827e-07,
-      "loss": 1.1123,
-      "step": 1663
-    },
-    {
-      "epoch": 1.617925102446545,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.35847170291073e-07,
-      "loss": 1.1198,
-      "step": 1664
-    },
-    {
-      "epoch": 1.6188988517872358,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.3368182434013333e-07,
-      "loss": 1.1071,
-      "step": 1665
-    },
-    {
-      "epoch": 1.6198726011279263,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.315213596843945e-07,
-      "loss": 1.1088,
-      "step": 1666
-    },
-    {
-      "epoch": 1.620846350468617,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.293657814275551e-07,
-      "loss": 1.0989,
-      "step": 1667
-    },
-    {
-      "epoch": 1.6218200998093075,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.272150946617712e-07,
-      "loss": 1.0944,
-      "step": 1668
-    },
-    {
-      "epoch": 1.622793849149998,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.250693044676429e-07,
-      "loss": 1.1041,
-      "step": 1669
-    },
-    {
-      "epoch": 1.6237675984906885,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.229284159142044e-07,
-      "loss": 1.1003,
-      "step": 1670
-    },
-    {
-      "epoch": 1.624741347831379,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.2079243405890785e-07,
-      "loss": 1.1139,
-      "step": 1671
-    },
-    {
-      "epoch": 1.6257150971720695,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.18661363947617e-07,
-      "loss": 1.1323,
-      "step": 1672
-    },
-    {
-      "epoch": 1.62668884651276,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.165352106145915e-07,
-      "loss": 1.1083,
-      "step": 1673
-    },
-    {
-      "epoch": 1.6276625958534507,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.144139790824764e-07,
-      "loss": 1.1139,
-      "step": 1674
-    },
-    {
-      "epoch": 1.6286363451941412,
-      "grad_norm": 0.302734375,
-      "learning_rate": 4.1229767436228805e-07,
-      "loss": 1.1266,
-      "step": 1675
-    },
-    {
-      "epoch": 1.629610094534832,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.1018630145340744e-07,
-      "loss": 1.1049,
-      "step": 1676
-    },
-    {
-      "epoch": 1.6305838438755225,
-      "grad_norm": 0.29296875,
-      "learning_rate": 4.080798653435633e-07,
-      "loss": 1.1058,
-      "step": 1677
-    },
-    {
-      "epoch": 1.631557593216213,
-      "grad_norm": 0.29296875,
-      "learning_rate": 4.0597837100882053e-07,
-      "loss": 1.1059,
-      "step": 1678
-    },
-    {
-      "epoch": 1.6325313425569035,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.0388182341357245e-07,
-      "loss": 1.1189,
-      "step": 1679
-    },
-    {
-      "epoch": 1.633505091897594,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.017902275105262e-07,
-      "loss": 1.1288,
-      "step": 1680
-    },
-    {
-      "epoch": 1.6344788412382845,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.9970358824068966e-07,
-      "loss": 1.1108,
-      "step": 1681
-    },
-    {
-      "epoch": 1.635452590578975,
-      "grad_norm": 0.291015625,
-      "learning_rate": 3.976219105333634e-07,
-      "loss": 1.0847,
-      "step": 1682
-    },
-    {
-      "epoch": 1.6364263399196657,
-      "grad_norm": 0.291015625,
-      "learning_rate": 3.9554519930612683e-07,
-      "loss": 1.1008,
-      "step": 1683
-    },
-    {
-      "epoch": 1.6374000892603562,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.934734594648268e-07,
-      "loss": 1.1105,
-      "step": 1684
-    },
-    {
-      "epoch": 1.638373838601047,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.9140669590356616e-07,
-      "loss": 1.1108,
-      "step": 1685
-    },
-    {
-      "epoch": 1.6393475879417374,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.893449135046923e-07,
-      "loss": 1.1066,
-      "step": 1686
-    },
-    {
-      "epoch": 1.640321337282428,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.87288117138786e-07,
-      "loss": 1.1154,
-      "step": 1687
-    },
-    {
-      "epoch": 1.6412950866231184,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.8523631166464763e-07,
-      "loss": 1.0986,
-      "step": 1688
-    },
-    {
-      "epoch": 1.642268835963809,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.831895019292897e-07,
-      "loss": 1.1149,
-      "step": 1689
-    },
-    {
-      "epoch": 1.6432425853044994,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.811476927679228e-07,
-      "loss": 1.119,
-      "step": 1690
-    },
-    {
-      "epoch": 1.6442163346451901,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.7911088900394215e-07,
-      "loss": 1.1154,
-      "step": 1691
-    },
-    {
-      "epoch": 1.6451900839858806,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.7707909544892284e-07,
-      "loss": 1.1312,
-      "step": 1692
-    },
-    {
-      "epoch": 1.6461638333265713,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.750523169026016e-07,
-      "loss": 1.1131,
-      "step": 1693
-    },
-    {
-      "epoch": 1.6471375826672618,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.7303055815286733e-07,
-      "loss": 1.1099,
-      "step": 1694
-    },
-    {
-      "epoch": 1.6481113320079523,
-      "grad_norm": 0.29296875,
-      "learning_rate": 3.710138239757527e-07,
-      "loss": 1.0925,
-      "step": 1695
-    },
-    {
-      "epoch": 1.6490850813486428,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.690021191354198e-07,
-      "loss": 1.0978,
-      "step": 1696
-    },
-    {
-      "epoch": 1.6500588306893333,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.6699544838415035e-07,
-      "loss": 1.101,
-      "step": 1697
-    },
-    {
-      "epoch": 1.6510325800300238,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.649938164623326e-07,
-      "loss": 1.1098,
-      "step": 1698
-    },
-    {
-      "epoch": 1.6520063293707143,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.6299722809845203e-07,
-      "loss": 1.1178,
-      "step": 1699
-    },
-    {
-      "epoch": 1.652980078711405,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.6100568800908205e-07,
-      "loss": 1.1249,
-      "step": 1700
-    },
-    {
-      "epoch": 1.6539538280520956,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.5901920089886687e-07,
-      "loss": 1.1188,
-      "step": 1701
-    },
-    {
-      "epoch": 1.6549275773927863,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.570377714605158e-07,
-      "loss": 1.1184,
-      "step": 1702
-    },
-    {
-      "epoch": 1.6559013267334768,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.5506140437479076e-07,
-      "loss": 1.1227,
-      "step": 1703
-    },
-    {
-      "epoch": 1.6568750760741673,
-      "grad_norm": 0.29296875,
-      "learning_rate": 3.5309010431049284e-07,
-      "loss": 1.0805,
-      "step": 1704
-    },
-    {
-      "epoch": 1.6578488254148578,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.5112387592445537e-07,
-      "loss": 1.1199,
-      "step": 1705
-    },
-    {
-      "epoch": 1.6588225747555483,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.4916272386152946e-07,
-      "loss": 1.1024,
-      "step": 1706
-    },
-    {
-      "epoch": 1.6597963240962388,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.4720665275457504e-07,
-      "loss": 1.1034,
-      "step": 1707
-    },
-    {
-      "epoch": 1.6607700734369293,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.452556672244492e-07,
-      "loss": 1.1157,
-      "step": 1708
-    },
-    {
-      "epoch": 1.66174382277762,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.4330977187999464e-07,
-      "loss": 1.0978,
-      "step": 1709
-    },
-    {
-      "epoch": 1.6627175721183105,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.413689713180307e-07,
-      "loss": 1.1002,
-      "step": 1710
-    },
-    {
-      "epoch": 1.6636913214590012,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.394332701233391e-07,
-      "loss": 1.0925,
-      "step": 1711
-    },
-    {
-      "epoch": 1.6646650707996917,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.375026728686573e-07,
-      "loss": 1.1029,
-      "step": 1712
-    },
-    {
-      "epoch": 1.6656388201403822,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.3557718411466517e-07,
-      "loss": 1.1128,
-      "step": 1713
-    },
-    {
-      "epoch": 1.6666125694810727,
-      "grad_norm": 0.29296875,
-      "learning_rate": 3.3365680840997337e-07,
-      "loss": 1.1321,
-      "step": 1714
-    },
-    {
-      "epoch": 1.6675863188217632,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.3174155029111505e-07,
-      "loss": 1.1146,
-      "step": 1715
-    },
-    {
-      "epoch": 1.6685600681624537,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.2983141428253565e-07,
-      "loss": 1.1146,
-      "step": 1716
-    },
-    {
-      "epoch": 1.6695338175031444,
-      "grad_norm": 0.29296875,
-      "learning_rate": 3.279264048965769e-07,
-      "loss": 1.1262,
-      "step": 1717
-    },
-    {
-      "epoch": 1.670507566843835,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.260265266334725e-07,
-      "loss": 1.0968,
-      "step": 1718
-    },
-    {
-      "epoch": 1.6714813161845254,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.2413178398133435e-07,
-      "loss": 1.1354,
-      "step": 1719
-    },
-    {
-      "epoch": 1.6724550655252162,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.2224218141614223e-07,
-      "loss": 1.1221,
-      "step": 1720
-    },
-    {
-      "epoch": 1.6734288148659067,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.203577234017324e-07,
-      "loss": 1.1098,
-      "step": 1721
-    },
-    {
-      "epoch": 1.6744025642065972,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.184784143897895e-07,
-      "loss": 1.1072,
-      "step": 1722
-    },
-    {
-      "epoch": 1.6753763135472877,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.1660425881983425e-07,
-      "loss": 1.1273,
-      "step": 1723
-    },
-    {
-      "epoch": 1.6763500628879782,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.147352611192131e-07,
-      "loss": 1.1001,
-      "step": 1724
-    },
-    {
-      "epoch": 1.6773238122286687,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.128714257030882e-07,
-      "loss": 1.1127,
-      "step": 1725
-    },
-    {
-      "epoch": 1.6782975615693594,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.11012756974427e-07,
-      "loss": 1.1242,
-      "step": 1726
-    },
-    {
-      "epoch": 1.6792713109100499,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.0915925932399006e-07,
-      "loss": 1.119,
-      "step": 1727
-    },
-    {
-      "epoch": 1.6802450602507406,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.07310937130324e-07,
-      "loss": 1.1159,
-      "step": 1728
-    },
-    {
-      "epoch": 1.681218809591431,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.05467794759749e-07,
-      "loss": 1.124,
-      "step": 1729
-    },
-    {
-      "epoch": 1.6821925589321216,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.036298365663487e-07,
-      "loss": 1.11,
-      "step": 1730
-    },
-    {
-      "epoch": 1.683166308272812,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.0179706689195895e-07,
-      "loss": 1.1157,
-      "step": 1731
-    },
-    {
-      "epoch": 1.6841400576135026,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.9996949006616096e-07,
-      "loss": 1.1061,
-      "step": 1732
-    },
-    {
-      "epoch": 1.685113806954193,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.981471104062675e-07,
-      "loss": 1.1366,
-      "step": 1733
-    },
-    {
-      "epoch": 1.6860875562948836,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.96329932217313e-07,
-      "loss": 1.1042,
-      "step": 1734
-    },
-    {
-      "epoch": 1.6870613056355743,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.945179597920464e-07,
-      "loss": 1.1147,
-      "step": 1735
-    },
-    {
-      "epoch": 1.6880350549762648,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.927111974109181e-07,
-      "loss": 1.1081,
-      "step": 1736
-    },
-    {
-      "epoch": 1.6890088043169555,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.909096493420699e-07,
-      "loss": 1.1143,
-      "step": 1737
-    },
-    {
-      "epoch": 1.689982553657646,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.891133198413268e-07,
-      "loss": 1.107,
-      "step": 1738
-    },
-    {
-      "epoch": 1.6909563029983365,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.8732221315218576e-07,
-      "loss": 1.0975,
-      "step": 1739
-    },
-    {
-      "epoch": 1.691930052339027,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.8553633350580534e-07,
-      "loss": 1.1052,
-      "step": 1740
-    },
-    {
-      "epoch": 1.6929038016797175,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.8375568512099713e-07,
-      "loss": 1.0811,
-      "step": 1741
-    },
-    {
-      "epoch": 1.693877551020408,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.8198027220421363e-07,
-      "loss": 1.1013,
-      "step": 1742
-    },
-    {
-      "epoch": 1.6948513003610988,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.802100989495407e-07,
-      "loss": 1.1195,
-      "step": 1743
-    },
-    {
-      "epoch": 1.6958250497017893,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.78445169538685e-07,
-      "loss": 1.1275,
-      "step": 1744
-    },
-    {
-      "epoch": 1.6967987990424798,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.766854881409667e-07,
-      "loss": 1.1194,
-      "step": 1745
-    },
-    {
-      "epoch": 1.6977725483831705,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.7493105891330837e-07,
-      "loss": 1.1198,
-      "step": 1746
-    },
-    {
-      "epoch": 1.698746297723861,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.731818860002253e-07,
-      "loss": 1.1095,
-      "step": 1747
-    },
-    {
-      "epoch": 1.6997200470645515,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.714379735338152e-07,
-      "loss": 1.144,
-      "step": 1748
-    },
-    {
-      "epoch": 1.700693796405242,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.696993256337499e-07,
-      "loss": 1.126,
-      "step": 1749
-    },
-    {
-      "epoch": 1.7016675457459325,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.679659464072631e-07,
-      "loss": 1.1129,
-      "step": 1750
-    },
-    {
-      "epoch": 1.702641295086623,
-      "grad_norm": 0.2890625,
-      "learning_rate": 2.6623783994914386e-07,
-      "loss": 1.1011,
-      "step": 1751
-    },
-    {
-      "epoch": 1.7036150444273137,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.645150103417246e-07,
-      "loss": 1.1046,
-      "step": 1752
-    },
-    {
-      "epoch": 1.7045887937680042,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.6279746165487256e-07,
-      "loss": 1.1109,
-      "step": 1753
-    },
-    {
-      "epoch": 1.705562543108695,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.610851979459786e-07,
-      "loss": 1.0963,
-      "step": 1754
-    },
-    {
-      "epoch": 1.7065362924493854,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.5937822325994954e-07,
-      "loss": 1.1119,
-      "step": 1755
-    },
-    {
-      "epoch": 1.707510041790076,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.57676541629199e-07,
-      "loss": 1.113,
-      "step": 1756
-    },
-    {
-      "epoch": 1.7084837911307664,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.559801570736348e-07,
-      "loss": 1.1127,
-      "step": 1757
-    },
-    {
-      "epoch": 1.709457540471457,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.5428907360065256e-07,
-      "loss": 1.1279,
-      "step": 1758
-    },
-    {
-      "epoch": 1.7104312898121474,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.5260329520512524e-07,
-      "loss": 1.1167,
-      "step": 1759
-    },
-    {
-      "epoch": 1.711405039152838,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.5092282586939187e-07,
-      "loss": 1.1002,
-      "step": 1760
-    },
-    {
-      "epoch": 1.7123787884935286,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.4924766956325176e-07,
-      "loss": 1.1066,
-      "step": 1761
-    },
-    {
-      "epoch": 1.7133525378342191,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.4757783024395244e-07,
-      "loss": 1.105,
-      "step": 1762
-    },
-    {
-      "epoch": 1.7143262871749099,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.4591331185618075e-07,
-      "loss": 1.1019,
-      "step": 1763
-    },
-    {
-      "epoch": 1.7153000365156004,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.4425411833205406e-07,
-      "loss": 1.1098,
-      "step": 1764
-    },
-    {
-      "epoch": 1.7162737858562909,
-      "grad_norm": 0.291015625,
-      "learning_rate": 2.4260025359111124e-07,
-      "loss": 1.1156,
-      "step": 1765
-    },
-    {
-      "epoch": 1.7172475351969814,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.4095172154030235e-07,
-      "loss": 1.0934,
-      "step": 1766
-    },
-    {
-      "epoch": 1.7182212845376719,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.393085260739794e-07,
-      "loss": 1.1108,
-      "step": 1767
-    },
-    {
-      "epoch": 1.7191950338783624,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.3767067107388852e-07,
-      "loss": 1.1067,
-      "step": 1768
-    },
-    {
-      "epoch": 1.7201687832190529,
-      "grad_norm": 0.291015625,
-      "learning_rate": 2.3603816040915994e-07,
-      "loss": 1.104,
-      "step": 1769
-    },
-    {
-      "epoch": 1.7211425325597436,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.3441099793629906e-07,
-      "loss": 1.1139,
-      "step": 1770
-    },
-    {
-      "epoch": 1.722116281900434,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.3278918749917528e-07,
-      "loss": 1.1266,
-      "step": 1771
-    },
-    {
-      "epoch": 1.7230900312411248,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.3117273292901865e-07,
-      "loss": 1.1032,
-      "step": 1772
-    },
-    {
-      "epoch": 1.7240637805818153,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.2956163804440296e-07,
-      "loss": 1.1001,
-      "step": 1773
-    },
-    {
-      "epoch": 1.7250375299225058,
-      "grad_norm": 0.291015625,
-      "learning_rate": 2.2795590665124267e-07,
-      "loss": 1.0866,
-      "step": 1774
-    },
-    {
-      "epoch": 1.7260112792631963,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.2635554254278213e-07,
-      "loss": 1.1151,
-      "step": 1775
-    },
-    {
-      "epoch": 1.7269850286038868,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.2476054949958664e-07,
-      "loss": 1.1167,
-      "step": 1776
-    },
-    {
-      "epoch": 1.7279587779445773,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.2317093128953137e-07,
-      "loss": 1.1167,
-      "step": 1777
-    },
-    {
-      "epoch": 1.728932527285268,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.2158669166779667e-07,
-      "loss": 1.1099,
-      "step": 1778
-    },
-    {
-      "epoch": 1.7299062766259585,
-      "grad_norm": 0.291015625,
-      "learning_rate": 2.2000783437685608e-07,
-      "loss": 1.106,
-      "step": 1779
-    },
-    {
-      "epoch": 1.7308800259666492,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.184343631464686e-07,
-      "loss": 1.108,
-      "step": 1780
-    },
-    {
-      "epoch": 1.7318537753073397,
-      "grad_norm": 0.291015625,
-      "learning_rate": 2.1686628169366923e-07,
-      "loss": 1.105,
-      "step": 1781
-    },
-    {
-      "epoch": 1.7328275246480302,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.1530359372276126e-07,
-      "loss": 1.0873,
-      "step": 1782
-    },
-    {
-      "epoch": 1.7338012739887207,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.1374630292530617e-07,
-      "loss": 1.1124,
-      "step": 1783
-    },
-    {
-      "epoch": 1.7347750233294112,
-      "grad_norm": 0.291015625,
-      "learning_rate": 2.1219441298011594e-07,
-      "loss": 1.1201,
-      "step": 1784
-    },
-    {
-      "epoch": 1.7357487726701017,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.1064792755324443e-07,
-      "loss": 1.104,
-      "step": 1785
-    },
-    {
-      "epoch": 1.7367225220107922,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.0910685029797789e-07,
-      "loss": 1.1113,
-      "step": 1786
-    },
-    {
-      "epoch": 1.737696271351483,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.0757118485482592e-07,
-      "loss": 1.1137,
-      "step": 1787
-    },
-    {
-      "epoch": 1.7386700206921735,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.0604093485151548e-07,
-      "loss": 1.1215,
-      "step": 1788
-    },
-    {
-      "epoch": 1.7396437700328642,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.0451610390298027e-07,
-      "loss": 1.1047,
-      "step": 1789
-    },
-    {
-      "epoch": 1.7406175193735547,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.0299669561135048e-07,
-      "loss": 1.1254,
-      "step": 1790
-    },
-    {
-      "epoch": 1.7415912687142452,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.014827135659486e-07,
-      "loss": 1.124,
-      "step": 1791
-    },
-    {
-      "epoch": 1.7425650180549357,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.9997416134327735e-07,
-      "loss": 1.0975,
-      "step": 1792
-    },
-    {
-      "epoch": 1.7435387673956262,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.9847104250701327e-07,
-      "loss": 1.1047,
-      "step": 1793
-    },
-    {
-      "epoch": 1.7445125167363167,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.9697336060799626e-07,
-      "loss": 1.1057,
-      "step": 1794
-    },
-    {
-      "epoch": 1.7454862660770072,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.95481119184224e-07,
-      "loss": 1.1361,
-      "step": 1795
-    },
-    {
-      "epoch": 1.746460015417698,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.9399432176084116e-07,
-      "loss": 1.1217,
-      "step": 1796
-    },
-    {
-      "epoch": 1.7474337647583884,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.925129718501323e-07,
-      "loss": 1.0974,
-      "step": 1797
-    },
-    {
-      "epoch": 1.7484075140990791,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.9103707295151262e-07,
-      "loss": 1.1087,
-      "step": 1798
-    },
-    {
-      "epoch": 1.7493812634397696,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.8956662855152157e-07,
-      "loss": 1.1083,
-      "step": 1799
-    },
-    {
-      "epoch": 1.7503550127804601,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.881016421238119e-07,
-      "loss": 1.1083,
-      "step": 1800
-    },
-    {
-      "epoch": 1.7513287621211506,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.8664211712914382e-07,
-      "loss": 1.1072,
-      "step": 1801
-    },
-    {
-      "epoch": 1.7523025114618411,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.851880570153755e-07,
-      "loss": 1.1317,
-      "step": 1802
-    },
-    {
-      "epoch": 1.7532762608025316,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.8373946521745567e-07,
-      "loss": 1.1298,
-      "step": 1803
-    },
-    {
-      "epoch": 1.7542500101432223,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.82296345157415e-07,
-      "loss": 1.096,
-      "step": 1804
-    },
-    {
-      "epoch": 1.7552237594839128,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.8085870024435849e-07,
-      "loss": 1.1231,
-      "step": 1805
-    },
-    {
-      "epoch": 1.7561975088246033,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.7942653387445607e-07,
-      "loss": 1.0859,
-      "step": 1806
-    },
-    {
-      "epoch": 1.757171258165294,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.7799984943093708e-07,
-      "loss": 1.1322,
-      "step": 1807
-    },
-    {
-      "epoch": 1.7581450075059846,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.7657865028407967e-07,
-      "loss": 1.1104,
-      "step": 1808
-    },
-    {
-      "epoch": 1.759118756846675,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.7516293979120525e-07,
-      "loss": 1.1082,
-      "step": 1809
-    },
-    {
-      "epoch": 1.7600925061873656,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.737527212966672e-07,
-      "loss": 1.1133,
-      "step": 1810
-    },
-    {
-      "epoch": 1.761066255528056,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.7234799813184681e-07,
-      "loss": 1.0896,
-      "step": 1811
-    },
-    {
-      "epoch": 1.7620400048687466,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.7094877361514413e-07,
-      "loss": 1.0921,
-      "step": 1812
-    },
-    {
-      "epoch": 1.7630137542094373,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.6955505105196752e-07,
-      "loss": 1.1113,
-      "step": 1813
-    },
-    {
-      "epoch": 1.7639875035501278,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.6816683373473004e-07,
-      "loss": 1.0998,
-      "step": 1814
-    },
-    {
-      "epoch": 1.7649612528908185,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.667841249428384e-07,
-      "loss": 1.1224,
-      "step": 1815
-    },
-    {
-      "epoch": 1.765935002231509,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.654069279426873e-07,
-      "loss": 1.1041,
-      "step": 1816
-    },
-    {
-      "epoch": 1.7669087515721995,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.640352459876496e-07,
-      "loss": 1.1232,
-      "step": 1817
-    },
-    {
-      "epoch": 1.76788250091289,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.6266908231807127e-07,
-      "loss": 1.1077,
-      "step": 1818
-    },
-    {
-      "epoch": 1.7688562502535805,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.6130844016126164e-07,
-      "loss": 1.09,
-      "step": 1819
-    },
-    {
-      "epoch": 1.769829999594271,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.5995332273148644e-07,
-      "loss": 1.1331,
-      "step": 1820
-    },
-    {
-      "epoch": 1.7708037489349615,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.586037332299606e-07,
-      "loss": 1.1187,
-      "step": 1821
-    },
-    {
-      "epoch": 1.7717774982756522,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.5725967484484046e-07,
-      "loss": 1.1119,
-      "step": 1822
-    },
-    {
-      "epoch": 1.7727512476163427,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.5592115075121512e-07,
-      "loss": 1.1215,
-      "step": 1823
-    },
-    {
-      "epoch": 1.7737249969570335,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.5458816411110107e-07,
-      "loss": 1.1161,
-      "step": 1824
-    },
-    {
-      "epoch": 1.774698746297724,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.5326071807343272e-07,
-      "loss": 1.124,
-      "step": 1825
-    },
-    {
-      "epoch": 1.7756724956384144,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.5193881577405729e-07,
-      "loss": 1.0938,
-      "step": 1826
-    },
-    {
-      "epoch": 1.776646244979105,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.5062246033572313e-07,
-      "loss": 1.1319,
-      "step": 1827
-    },
-    {
-      "epoch": 1.7776199943197954,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.4931165486807896e-07,
-      "loss": 1.1099,
-      "step": 1828
-    },
-    {
-      "epoch": 1.778593743660486,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.4800640246765945e-07,
-      "loss": 1.123,
-      "step": 1829
-    },
-    {
-      "epoch": 1.7795674930011764,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.467067062178823e-07,
-      "loss": 1.1025,
-      "step": 1830
-    },
-    {
-      "epoch": 1.7805412423418672,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.4541256918904069e-07,
-      "loss": 1.1016,
-      "step": 1831
-    },
-    {
-      "epoch": 1.7815149916825577,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.441239944382944e-07,
-      "loss": 1.1238,
-      "step": 1832
-    },
-    {
-      "epoch": 1.7824887410232484,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.4284098500966263e-07,
-      "loss": 1.1093,
-      "step": 1833
-    },
-    {
-      "epoch": 1.783462490363939,
-      "grad_norm": 0.2890625,
-      "learning_rate": 1.415635439340185e-07,
-      "loss": 1.0803,
-      "step": 1834
-    },
-    {
-      "epoch": 1.7844362397046294,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.4029167422908107e-07,
-      "loss": 1.1166,
-      "step": 1835
-    },
-    {
-      "epoch": 1.78540998904532,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.390253788994078e-07,
-      "loss": 1.1173,
-      "step": 1836
-    },
-    {
-      "epoch": 1.7863837383860104,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.3776466093638696e-07,
-      "loss": 1.1279,
-      "step": 1837
-    },
-    {
-      "epoch": 1.7873574877267009,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.365095233182326e-07,
-      "loss": 1.1077,
-      "step": 1838
-    },
-    {
-      "epoch": 1.7883312370673916,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.3525996900997567e-07,
-      "loss": 1.0882,
-      "step": 1839
-    },
-    {
-      "epoch": 1.789304986408082,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.3401600096345696e-07,
-      "loss": 1.1031,
-      "step": 1840
-    },
-    {
-      "epoch": 1.7902787357487728,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.3277762211732159e-07,
-      "loss": 1.111,
-      "step": 1841
-    },
-    {
-      "epoch": 1.7912524850894633,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.315448353970114e-07,
-      "loss": 1.1118,
-      "step": 1842
-    },
-    {
-      "epoch": 1.7922262344301538,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.303176437147563e-07,
-      "loss": 1.1313,
-      "step": 1843
-    },
-    {
-      "epoch": 1.7931999837708443,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.2909604996957093e-07,
-      "loss": 1.1007,
-      "step": 1844
-    },
-    {
-      "epoch": 1.7941737331115348,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.2788005704724549e-07,
-      "loss": 1.1067,
-      "step": 1845
-    },
-    {
-      "epoch": 1.7951474824522253,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.2666966782033745e-07,
-      "loss": 1.1017,
-      "step": 1846
-    },
-    {
-      "epoch": 1.7961212317929158,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.254648851481685e-07,
-      "loss": 1.1316,
-      "step": 1847
-    },
-    {
-      "epoch": 1.7970949811336066,
-      "grad_norm": 0.2890625,
-      "learning_rate": 1.2426571187681536e-07,
-      "loss": 1.1177,
-      "step": 1848
-    },
-    {
-      "epoch": 1.798068730474297,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.230721508391039e-07,
-      "loss": 1.1068,
-      "step": 1849
-    },
-    {
-      "epoch": 1.7990424798149878,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.2188420485460066e-07,
-      "loss": 1.1319,
-      "step": 1850
-    },
-    {
-      "epoch": 1.8000162291556783,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.2070187672960948e-07,
-      "loss": 1.0941,
-      "step": 1851
-    },
-    {
-      "epoch": 1.8009899784963688,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1952516925716197e-07,
-      "loss": 1.1221,
-      "step": 1852
-    },
-    {
-      "epoch": 1.8019637278370593,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.1835408521701259e-07,
-      "loss": 1.0959,
-      "step": 1853
-    },
-    {
-      "epoch": 1.8029374771777498,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.1718862737563063e-07,
-      "loss": 1.083,
-      "step": 1854
-    },
-    {
-      "epoch": 1.8039112265184403,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1602879848619602e-07,
-      "loss": 1.1073,
-      "step": 1855
-    },
-    {
-      "epoch": 1.8048849758591308,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.148746012885893e-07,
-      "loss": 1.1014,
-      "step": 1856
-    },
-    {
-      "epoch": 1.8058587251998215,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.1372603850938885e-07,
-      "loss": 1.091,
-      "step": 1857
-    },
-    {
-      "epoch": 1.806832474540512,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.1258311286186208e-07,
-      "loss": 1.1169,
-      "step": 1858
-    },
-    {
-      "epoch": 1.8078062238812027,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.114458270459598e-07,
-      "loss": 1.1239,
-      "step": 1859
-    },
-    {
-      "epoch": 1.8087799732218932,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.103141837483096e-07,
-      "loss": 1.1037,
-      "step": 1860
-    },
-    {
-      "epoch": 1.8097537225625837,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.0918818564221029e-07,
-      "loss": 1.1123,
-      "step": 1861
-    },
-    {
-      "epoch": 1.8107274719032742,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.0806783538762439e-07,
-      "loss": 1.1231,
-      "step": 1862
-    },
-    {
-      "epoch": 1.8117012212439647,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.0695313563117232e-07,
-      "loss": 1.0973,
-      "step": 1863
-    },
-    {
-      "epoch": 1.8126749705846552,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.0584408900612631e-07,
-      "loss": 1.094,
-      "step": 1864
-    },
-    {
-      "epoch": 1.813648719925346,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.0474069813240505e-07,
-      "loss": 1.1057,
-      "step": 1865
-    },
-    {
-      "epoch": 1.8146224692660364,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.0364296561656467e-07,
-      "loss": 1.1015,
-      "step": 1866
-    },
-    {
-      "epoch": 1.815596218606727,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.025508940517958e-07,
-      "loss": 1.1205,
-      "step": 1867
-    },
-    {
-      "epoch": 1.8165699679474177,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.0146448601791703e-07,
-      "loss": 1.0996,
-      "step": 1868
-    },
-    {
-      "epoch": 1.8175437172881082,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.0038374408136542e-07,
-      "loss": 1.1183,
-      "step": 1869
-    },
-    {
-      "epoch": 1.8185174666287987,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.930867079519485e-08,
-      "loss": 1.097,
-      "step": 1870
-    },
-    {
-      "epoch": 1.8194912159694892,
-      "grad_norm": 0.29296875,
-      "learning_rate": 9.823926869906741e-08,
-      "loss": 1.1089,
-      "step": 1871
-    },
-    {
-      "epoch": 1.8204649653101797,
-      "grad_norm": 0.2890625,
-      "learning_rate": 9.717554031924842e-08,
-      "loss": 1.1049,
-      "step": 1872
-    },
-    {
-      "epoch": 1.8214387146508701,
-      "grad_norm": 0.296875,
-      "learning_rate": 9.611748816859916e-08,
-      "loss": 1.1074,
-      "step": 1873
-    },
-    {
-      "epoch": 1.8224124639915609,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.506511474657282e-08,
-      "loss": 1.109,
-      "step": 1874
-    },
-    {
-      "epoch": 1.8233862133322514,
-      "grad_norm": 0.294921875,
-      "learning_rate": 9.401842253920712e-08,
-      "loss": 1.1026,
-      "step": 1875
-    },
-    {
-      "epoch": 1.824359962672942,
-      "grad_norm": 0.298828125,
-      "learning_rate": 9.297741401911947e-08,
-      "loss": 1.0956,
-      "step": 1876
-    },
-    {
-      "epoch": 1.8253337120136326,
-      "grad_norm": 0.29296875,
-      "learning_rate": 9.194209164550021e-08,
-      "loss": 1.1028,
-      "step": 1877
-    },
-    {
-      "epoch": 1.826307461354323,
-      "grad_norm": 0.296875,
-      "learning_rate": 9.091245786410768e-08,
-      "loss": 1.1045,
-      "step": 1878
-    },
-    {
-      "epoch": 1.8272812106950136,
-      "grad_norm": 0.294921875,
-      "learning_rate": 8.988851510726093e-08,
-      "loss": 1.1145,
-      "step": 1879
-    },
-    {
-      "epoch": 1.828254960035704,
-      "grad_norm": 0.30078125,
-      "learning_rate": 8.88702657938359e-08,
-      "loss": 1.1177,
-      "step": 1880
-    },
-    {
-      "epoch": 1.8292287093763946,
-      "grad_norm": 0.294921875,
-      "learning_rate": 8.785771232925844e-08,
-      "loss": 1.1063,
-      "step": 1881
-    },
-    {
-      "epoch": 1.830202458717085,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.68508571054999e-08,
-      "loss": 1.1085,
-      "step": 1882
-    },
-    {
-      "epoch": 1.8311762080577758,
-      "grad_norm": 0.291015625,
-      "learning_rate": 8.58497025010685e-08,
-      "loss": 1.1116,
-      "step": 1883
-    },
-    {
-      "epoch": 1.8321499573984663,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.485425088100852e-08,
-      "loss": 1.112,
-      "step": 1884
-    },
-    {
-      "epoch": 1.833123706739157,
-      "grad_norm": 0.29296875,
-      "learning_rate": 8.386450459689028e-08,
-      "loss": 1.1031,
-      "step": 1885
-    },
-    {
-      "epoch": 1.8340974560798475,
-      "grad_norm": 0.29296875,
-      "learning_rate": 8.288046598680627e-08,
-      "loss": 1.1208,
-      "step": 1886
-    },
-    {
-      "epoch": 1.835071205420538,
-      "grad_norm": 0.294921875,
-      "learning_rate": 8.190213737536618e-08,
-      "loss": 1.1218,
-      "step": 1887
-    },
-    {
-      "epoch": 1.8360449547612285,
-      "grad_norm": 0.2890625,
-      "learning_rate": 8.092952107369101e-08,
-      "loss": 1.1087,
-      "step": 1888
-    },
-    {
-      "epoch": 1.837018704101919,
-      "grad_norm": 0.29296875,
-      "learning_rate": 7.996261937940675e-08,
-      "loss": 1.1028,
-      "step": 1889
-    },
-    {
-      "epoch": 1.8379924534426095,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.900143457664017e-08,
-      "loss": 1.1153,
-      "step": 1890
-    },
-    {
-      "epoch": 1.8389662027833003,
-      "grad_norm": 0.30078125,
-      "learning_rate": 7.80459689360133e-08,
-      "loss": 1.1015,
-      "step": 1891
-    },
-    {
-      "epoch": 1.8399399521239908,
-      "grad_norm": 0.296875,
-      "learning_rate": 7.709622471463701e-08,
-      "loss": 1.1278,
-      "step": 1892
-    },
-    {
-      "epoch": 1.8409137014646813,
-      "grad_norm": 0.298828125,
-      "learning_rate": 7.61522041561069e-08,
-      "loss": 1.1247,
-      "step": 1893
-    },
-    {
-      "epoch": 1.841887450805372,
-      "grad_norm": 0.298828125,
-      "learning_rate": 7.521390949049717e-08,
-      "loss": 1.1018,
-      "step": 1894
-    },
-    {
-      "epoch": 1.8428612001460625,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.428134293435585e-08,
-      "loss": 1.1085,
-      "step": 1895
-    },
-    {
-      "epoch": 1.843834949486753,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.335450669069877e-08,
-      "loss": 1.1119,
-      "step": 1896
-    },
-    {
-      "epoch": 1.8448086988274435,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.243340294900592e-08,
-      "loss": 1.1115,
-      "step": 1897
-    },
-    {
-      "epoch": 1.845782448168134,
-      "grad_norm": 0.298828125,
-      "learning_rate": 7.151803388521505e-08,
-      "loss": 1.0971,
-      "step": 1898
-    },
-    {
-      "epoch": 1.8467561975088245,
-      "grad_norm": 0.291015625,
-      "learning_rate": 7.060840166171562e-08,
-      "loss": 1.1048,
-      "step": 1899
-    },
-    {
-      "epoch": 1.8477299468495152,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.97045084273465e-08,
-      "loss": 1.1129,
-      "step": 1900
-    },
-    {
-      "epoch": 1.8487036961902057,
-      "grad_norm": 0.30078125,
-      "learning_rate": 6.880635631738908e-08,
-      "loss": 1.0951,
-      "step": 1901
-    },
-    {
-      "epoch": 1.8496774455308964,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.791394745356089e-08,
-      "loss": 1.0993,
-      "step": 1902
-    },
-    {
-      "epoch": 1.850651194871587,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.702728394401337e-08,
-      "loss": 1.0879,
-      "step": 1903
-    },
-    {
-      "epoch": 1.8516249442122774,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.614636788332574e-08,
-      "loss": 1.1205,
-      "step": 1904
-    },
-    {
-      "epoch": 1.852598693552968,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.527120135249975e-08,
-      "loss": 1.1151,
-      "step": 1905
-    },
-    {
-      "epoch": 1.8535724428936584,
-      "grad_norm": 0.30078125,
-      "learning_rate": 6.440178641895389e-08,
-      "loss": 1.1254,
-      "step": 1906
-    },
-    {
-      "epoch": 1.854546192234349,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.353812513652052e-08,
-      "loss": 1.1047,
-      "step": 1907
-    },
-    {
-      "epoch": 1.8555199415750394,
-      "grad_norm": 0.3515625,
-      "learning_rate": 6.268021954544095e-08,
-      "loss": 1.1273,
-      "step": 1908
-    },
-    {
-      "epoch": 1.8564936909157301,
-      "grad_norm": 0.294921875,
-      "learning_rate": 6.182807167235794e-08,
-      "loss": 1.1099,
-      "step": 1909
-    },
-    {
-      "epoch": 1.8574674402564206,
-      "grad_norm": 0.3046875,
-      "learning_rate": 6.09816835303137e-08,
-      "loss": 1.1165,
-      "step": 1910
-    },
-    {
-      "epoch": 1.8584411895971114,
-      "grad_norm": 0.29296875,
-      "learning_rate": 6.01410571187444e-08,
-      "loss": 1.1161,
-      "step": 1911
-    },
-    {
-      "epoch": 1.8594149389378019,
-      "grad_norm": 0.306640625,
-      "learning_rate": 5.9306194423473795e-08,
-      "loss": 1.1211,
-      "step": 1912
-    },
-    {
-      "epoch": 1.8603886882784924,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.847709741671148e-08,
-      "loss": 1.0959,
-      "step": 1913
-    },
-    {
-      "epoch": 1.8613624376191829,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.7653768057045757e-08,
-      "loss": 1.1065,
-      "step": 1914
-    },
-    {
-      "epoch": 1.8623361869598734,
-      "grad_norm": 0.30078125,
-      "learning_rate": 5.6836208289440807e-08,
-      "loss": 1.1176,
-      "step": 1915
-    },
-    {
-      "epoch": 1.8633099363005639,
-      "grad_norm": 0.298828125,
-      "learning_rate": 5.6024420045230065e-08,
-      "loss": 1.117,
-      "step": 1916
-    },
-    {
-      "epoch": 1.8642836856412544,
-      "grad_norm": 0.2890625,
-      "learning_rate": 5.5218405242113415e-08,
-      "loss": 1.1043,
-      "step": 1917
-    },
-    {
-      "epoch": 1.865257434981945,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.441816578415276e-08,
-      "loss": 1.1322,
-      "step": 1918
-    },
-    {
-      "epoch": 1.8662311843226356,
-      "grad_norm": 0.29296875,
-      "learning_rate": 5.362370356176511e-08,
-      "loss": 1.1053,
-      "step": 1919
-    },
-    {
-      "epoch": 1.8672049336633263,
-      "grad_norm": 0.29296875,
-      "learning_rate": 5.283502045172112e-08,
-      "loss": 1.1015,
-      "step": 1920
-    },
-    {
-      "epoch": 1.8681786830040168,
-      "grad_norm": 0.294921875,
-      "learning_rate": 5.205211831713935e-08,
-      "loss": 1.1161,
-      "step": 1921
-    },
-    {
-      "epoch": 1.8691524323447073,
-      "grad_norm": 0.29296875,
-      "learning_rate": 5.127499900748123e-08,
-      "loss": 1.0981,
-      "step": 1922
-    },
-    {
-      "epoch": 1.8701261816853978,
-      "grad_norm": 0.298828125,
-      "learning_rate": 5.0503664358547424e-08,
-      "loss": 1.1168,
-      "step": 1923
-    },
-    {
-      "epoch": 1.8710999310260883,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.9738116192474264e-08,
-      "loss": 1.1257,
-      "step": 1924
-    },
-    {
-      "epoch": 1.8720736803667788,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.897835631772763e-08,
-      "loss": 1.1163,
-      "step": 1925
-    },
-    {
-      "epoch": 1.8730474297074695,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.8224386529099884e-08,
-      "loss": 1.104,
-      "step": 1926
-    },
-    {
-      "epoch": 1.87402117904816,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.7476208607705463e-08,
-      "loss": 1.1185,
-      "step": 1927
-    },
-    {
-      "epoch": 1.8749949283888507,
-      "grad_norm": 0.2890625,
-      "learning_rate": 4.6733824320976674e-08,
-      "loss": 1.1121,
-      "step": 1928
-    },
-    {
-      "epoch": 1.8759686777295412,
-      "grad_norm": 0.30078125,
-      "learning_rate": 4.599723542265927e-08,
-      "loss": 1.117,
-      "step": 1929
-    },
-    {
-      "epoch": 1.8769424270702317,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.526644365280858e-08,
-      "loss": 1.1284,
-      "step": 1930
-    },
-    {
-      "epoch": 1.8779161764109222,
-      "grad_norm": 0.291015625,
-      "learning_rate": 4.454145073778532e-08,
-      "loss": 1.1131,
-      "step": 1931
-    },
-    {
-      "epoch": 1.8788899257516127,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.382225839025117e-08,
-      "loss": 1.1129,
-      "step": 1932
-    },
-    {
-      "epoch": 1.8798636750923032,
-      "grad_norm": 0.29296875,
-      "learning_rate": 4.310886830916544e-08,
-      "loss": 1.1175,
-      "step": 1933
-    },
-    {
-      "epoch": 1.8808374244329937,
-      "grad_norm": 0.302734375,
-      "learning_rate": 4.240128217978062e-08,
-      "loss": 1.1223,
-      "step": 1934
-    },
-    {
-      "epoch": 1.8818111737736845,
-      "grad_norm": 0.2890625,
-      "learning_rate": 4.169950167363768e-08,
-      "loss": 1.091,
-      "step": 1935
-    },
-    {
-      "epoch": 1.882784923114375,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.1003528448564366e-08,
-      "loss": 1.0898,
-      "step": 1936
-    },
-    {
-      "epoch": 1.8837586724550657,
-      "grad_norm": 0.29296875,
-      "learning_rate": 4.031336414866832e-08,
-      "loss": 1.0946,
-      "step": 1937
-    },
-    {
-      "epoch": 1.8847324217957562,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.9629010404335644e-08,
-      "loss": 1.0958,
-      "step": 1938
-    },
-    {
-      "epoch": 1.8857061711364467,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.895046883222508e-08,
-      "loss": 1.1194,
-      "step": 1939
-    },
-    {
-      "epoch": 1.8866799204771372,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.827774103526666e-08,
-      "loss": 1.0904,
-      "step": 1940
-    },
-    {
-      "epoch": 1.8876536698178277,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.761082860265553e-08,
-      "loss": 1.0919,
-      "step": 1941
-    },
-    {
-      "epoch": 1.8886274191585182,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.6949733109848395e-08,
-      "loss": 1.1241,
-      "step": 1942
-    },
-    {
-      "epoch": 1.8896011684992087,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.629445611856186e-08,
-      "loss": 1.0963,
-      "step": 1943
-    },
-    {
-      "epoch": 1.8905749178398994,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.5644999176767116e-08,
-      "loss": 1.1069,
-      "step": 1944
-    },
-    {
-      "epoch": 1.89154866718059,
-      "grad_norm": 0.291015625,
-      "learning_rate": 3.500136381868552e-08,
-      "loss": 1.1014,
-      "step": 1945
-    },
-    {
-      "epoch": 1.8925224165212806,
-      "grad_norm": 0.291015625,
-      "learning_rate": 3.436355156478749e-08,
-      "loss": 1.1004,
-      "step": 1946
-    },
-    {
-      "epoch": 1.8934961658619711,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.3731563921786116e-08,
-      "loss": 1.1019,
-      "step": 1947
-    },
-    {
-      "epoch": 1.8944699152026616,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.310540238263604e-08,
-      "loss": 1.1268,
-      "step": 1948
-    },
-    {
-      "epoch": 1.8954436645433521,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.248506842652793e-08,
-      "loss": 1.1261,
-      "step": 1949
-    },
-    {
-      "epoch": 1.8964174138840426,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.187056351888651e-08,
-      "loss": 1.0993,
-      "step": 1950
-    },
-    {
-      "epoch": 1.8973911632247331,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.1261889111366125e-08,
-      "loss": 1.1099,
-      "step": 1951
-    },
-    {
-      "epoch": 1.8983649125654238,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.065904664184771e-08,
-      "loss": 1.1155,
-      "step": 1952
-    },
-    {
-      "epoch": 1.8993386619061143,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.006203753443487e-08,
-      "loss": 1.129,
-      "step": 1953
-    },
-    {
-      "epoch": 1.9003124112468048,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.9470863199451972e-08,
-      "loss": 1.1008,
-      "step": 1954
-    },
-    {
-      "epoch": 1.9012861605874956,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.888552503343883e-08,
-      "loss": 1.1089,
-      "step": 1955
-    },
-    {
-      "epoch": 1.902259909928186,
-      "grad_norm": 0.2890625,
-      "learning_rate": 2.8306024419148814e-08,
-      "loss": 1.0956,
-      "step": 1956
-    },
-    {
-      "epoch": 1.9032336592688766,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.7732362725545747e-08,
-      "loss": 1.1214,
-      "step": 1957
-    },
-    {
-      "epoch": 1.904207408609567,
-      "grad_norm": 0.291015625,
-      "learning_rate": 2.7164541307798665e-08,
-      "loss": 1.1183,
-      "step": 1958
-    },
-    {
-      "epoch": 1.9051811579502576,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.6602561507281245e-08,
-      "loss": 1.0996,
-      "step": 1959
-    },
-    {
-      "epoch": 1.906154907290948,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.604642465156709e-08,
-      "loss": 1.0992,
-      "step": 1960
-    },
-    {
-      "epoch": 1.9071286566316388,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.5496132054426692e-08,
-      "loss": 1.1059,
-      "step": 1961
-    },
-    {
-      "epoch": 1.9081024059723293,
-      "grad_norm": 0.291015625,
-      "learning_rate": 2.4951685015824623e-08,
-      "loss": 1.0966,
-      "step": 1962
-    },
-    {
-      "epoch": 1.90907615531302,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.4413084821916232e-08,
-      "loss": 1.1148,
-      "step": 1963
-    },
-    {
-      "epoch": 1.9100499046537105,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.3880332745045974e-08,
-      "loss": 1.0985,
-      "step": 1964
-    },
-    {
-      "epoch": 1.911023653994401,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.3353430043741578e-08,
-      "loss": 1.0899,
-      "step": 1965
-    },
-    {
-      "epoch": 1.9119974033350915,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.2832377962713203e-08,
-      "loss": 1.1007,
-      "step": 1966
-    },
-    {
-      "epoch": 1.912971152675782,
-      "grad_norm": 0.30078125,
-      "learning_rate": 2.231717773285097e-08,
-      "loss": 1.118,
-      "step": 1967
-    },
-    {
-      "epoch": 1.9139449020164725,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.1807830571219658e-08,
-      "loss": 1.1238,
-      "step": 1968
-    },
-    {
-      "epoch": 1.914918651357163,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.1304337681057885e-08,
-      "loss": 1.11,
-      "step": 1969
-    },
-    {
-      "epoch": 1.9158924006978537,
-      "grad_norm": 0.2890625,
-      "learning_rate": 2.0806700251775057e-08,
-      "loss": 1.0986,
-      "step": 1970
-    },
-    {
-      "epoch": 1.9168661500385442,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.031491945894748e-08,
-      "loss": 1.1062,
-      "step": 1971
-    },
-    {
-      "epoch": 1.917839899379235,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.9828996464316407e-08,
-      "loss": 1.1132,
-      "step": 1972
-    },
-    {
-      "epoch": 1.9188136487199254,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.934893241578528e-08,
-      "loss": 1.121,
-      "step": 1973
-    },
-    {
-      "epoch": 1.919787398060616,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.8874728447417223e-08,
-      "loss": 1.1201,
-      "step": 1974
-    },
-    {
-      "epoch": 1.9207611474013064,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.8406385679430594e-08,
-      "loss": 1.1185,
-      "step": 1975
-    },
-    {
-      "epoch": 1.921734896741997,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.7943905218199553e-08,
-      "loss": 1.1001,
-      "step": 1976
-    },
-    {
-      "epoch": 1.9227086460826874,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.7487288156248782e-08,
-      "loss": 1.1125,
-      "step": 1977
-    },
-    {
-      "epoch": 1.923682395423378,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.7036535572251267e-08,
-      "loss": 1.1066,
-      "step": 1978
-    },
-    {
-      "epoch": 1.9246561447640687,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.6591648531026906e-08,
-      "loss": 1.1086,
-      "step": 1979
-    },
-    {
-      "epoch": 1.9256298941047592,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.6152628083539458e-08,
-      "loss": 1.1202,
-      "step": 1980
-    },
-    {
-      "epoch": 1.9266036434454499,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.571947526689349e-08,
-      "loss": 1.1126,
-      "step": 1981
-    },
-    {
-      "epoch": 1.9275773927861404,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.529219110433272e-08,
-      "loss": 1.1092,
-      "step": 1982
-    },
-    {
-      "epoch": 1.9285511421268309,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.4870776605236948e-08,
-      "loss": 1.1296,
-      "step": 1983
-    },
-    {
-      "epoch": 1.9295248914675214,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.4455232765120397e-08,
-      "loss": 1.0883,
-      "step": 1984
-    },
-    {
-      "epoch": 1.9304986408082119,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.4045560565628114e-08,
-      "loss": 1.1214,
-      "step": 1985
-    },
-    {
-      "epoch": 1.9314723901489024,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.3641760974535401e-08,
-      "loss": 1.1111,
-      "step": 1986
-    },
-    {
-      "epoch": 1.932446139489593,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.3243834945744216e-08,
-      "loss": 1.1257,
-      "step": 1987
-    },
-    {
-      "epoch": 1.9334198888302836,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.2851783419281504e-08,
-      "loss": 1.1122,
-      "step": 1988
-    },
-    {
-      "epoch": 1.9343936381709743,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.2465607321296424e-08,
-      "loss": 1.113,
-      "step": 1989
-    },
-    {
-      "epoch": 1.9353673875116648,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.2085307564058679e-08,
-      "loss": 1.1144,
-      "step": 1990
-    },
-    {
-      "epoch": 1.9363411368523553,
-      "grad_norm": 0.294921875,
-      "learning_rate": 1.1710885045956022e-08,
-      "loss": 1.1093,
-      "step": 1991
-    },
-    {
-      "epoch": 1.9373148861930458,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.1342340651493144e-08,
-      "loss": 1.1219,
-      "step": 1992
-    },
-    {
-      "epoch": 1.9382886355337363,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.0979675251287514e-08,
-      "loss": 1.0943,
-      "step": 1993
-    },
-    {
-      "epoch": 1.9392623848744268,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.062288970206965e-08,
-      "loss": 1.1022,
-      "step": 1994
-    },
-    {
-      "epoch": 1.9402361342151173,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.0271984846679239e-08,
-      "loss": 1.1162,
-      "step": 1995
-    },
-    {
-      "epoch": 1.941209883555808,
-      "grad_norm": 0.29296875,
-      "learning_rate": 9.926961514064303e-09,
-      "loss": 1.1184,
-      "step": 1996
-    },
-    {
-      "epoch": 1.9421836328964985,
-      "grad_norm": 0.29296875,
-      "learning_rate": 9.587820519278702e-09,
-      "loss": 1.0872,
-      "step": 1997
-    },
-    {
-      "epoch": 1.9431573822371893,
-      "grad_norm": 0.296875,
-      "learning_rate": 9.25456266348046e-09,
-      "loss": 1.1274,
-      "step": 1998
-    },
-    {
-      "epoch": 1.9441311315778798,
-      "grad_norm": 0.294921875,
-      "learning_rate": 8.927188733930115e-09,
-      "loss": 1.1183,
-      "step": 1999
-    },
-    {
-      "epoch": 1.9451048809185703,
-      "grad_norm": 0.298828125,
-      "learning_rate": 8.605699503987374e-09,
-      "loss": 1.127,
-      "step": 2000
-    },
-    {
-      "epoch": 1.9460786302592608,
-      "grad_norm": 0.294921875,
-      "learning_rate": 8.290095733111958e-09,
-      "loss": 1.1123,
-      "step": 2001
-    },
-    {
-      "epoch": 1.9470523795999513,
-      "grad_norm": 0.294921875,
-      "learning_rate": 7.98037816685887e-09,
-      "loss": 1.1038,
-      "step": 2002
-    },
-    {
-      "epoch": 1.9480261289406418,
-      "grad_norm": 0.298828125,
-      "learning_rate": 7.676547536879242e-09,
-      "loss": 1.1245,
-      "step": 2003
-    },
-    {
-      "epoch": 1.9489998782813323,
-      "grad_norm": 0.29296875,
-      "learning_rate": 7.3786045609161586e-09,
-      "loss": 1.1183,
-      "step": 2004
-    },
-    {
-      "epoch": 1.949973627622023,
-      "grad_norm": 0.296875,
-      "learning_rate": 7.086549942805499e-09,
-      "loss": 1.0966,
-      "step": 2005
-    },
-    {
-      "epoch": 1.9509473769627135,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.800384372472324e-09,
-      "loss": 1.1359,
-      "step": 2006
-    },
-    {
-      "epoch": 1.9519211263034042,
-      "grad_norm": 0.298828125,
-      "learning_rate": 6.520108525929769e-09,
-      "loss": 1.1141,
-      "step": 2007
-    },
-    {
-      "epoch": 1.9528948756440947,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.245723065277653e-09,
-      "loss": 1.1267,
-      "step": 2008
-    },
-    {
-      "epoch": 1.9538686249847852,
-      "grad_norm": 0.291015625,
-      "learning_rate": 5.977228638701371e-09,
-      "loss": 1.0831,
-      "step": 2009
-    },
-    {
-      "epoch": 1.9548423743254757,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.714625880469116e-09,
-      "loss": 1.1133,
-      "step": 2010
-    },
-    {
-      "epoch": 1.9558161236661662,
-      "grad_norm": 0.296875,
-      "learning_rate": 5.4579154109313244e-09,
-      "loss": 1.1165,
-      "step": 2011
-    },
-    {
-      "epoch": 1.9567898730068567,
-      "grad_norm": 0.30078125,
-      "learning_rate": 5.20709783651957e-09,
-      "loss": 1.1013,
-      "step": 2012
-    },
-    {
-      "epoch": 1.9577636223475474,
-      "grad_norm": 0.29296875,
-      "learning_rate": 4.962173749743782e-09,
-      "loss": 1.1018,
-      "step": 2013
-    },
-    {
-      "epoch": 1.958737371688238,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.7231437291916925e-09,
-      "loss": 1.1102,
-      "step": 2014
-    },
-    {
-      "epoch": 1.9597111210289284,
-      "grad_norm": 0.294921875,
-      "learning_rate": 4.490008339527729e-09,
-      "loss": 1.1078,
-      "step": 2015
-    },
-    {
-      "epoch": 1.9606848703696191,
-      "grad_norm": 0.302734375,
-      "learning_rate": 4.262768131491346e-09,
-      "loss": 1.1042,
-      "step": 2016
-    },
-    {
-      "epoch": 1.9616586197103096,
-      "grad_norm": 0.298828125,
-      "learning_rate": 4.0414236418956365e-09,
-      "loss": 1.0998,
-      "step": 2017
-    },
-    {
-      "epoch": 1.9626323690510001,
-      "grad_norm": 0.296875,
-      "learning_rate": 3.8259753936256695e-09,
-      "loss": 1.1075,
-      "step": 2018
-    },
-    {
-      "epoch": 1.9636061183916906,
-      "grad_norm": 0.298828125,
-      "learning_rate": 3.6164238956384878e-09,
-      "loss": 1.1129,
-      "step": 2019
-    },
-    {
-      "epoch": 1.9645798677323811,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.412769642961167e-09,
-      "loss": 1.1044,
-      "step": 2020
-    },
-    {
-      "epoch": 1.9655536170730716,
-      "grad_norm": 0.30078125,
-      "learning_rate": 3.2150131166885924e-09,
-      "loss": 1.1089,
-      "step": 2021
-    },
-    {
-      "epoch": 1.9665273664137624,
-      "grad_norm": 0.294921875,
-      "learning_rate": 3.0231547839842945e-09,
-      "loss": 1.1106,
-      "step": 2022
-    },
-    {
-      "epoch": 1.9675011157544529,
-      "grad_norm": 0.29296875,
-      "learning_rate": 2.837195098078227e-09,
-      "loss": 1.1004,
-      "step": 2023
-    },
-    {
-      "epoch": 1.9684748650951436,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.6571344982659342e-09,
-      "loss": 1.1096,
-      "step": 2024
-    },
-    {
-      "epoch": 1.969448614435834,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.4829734099068858e-09,
-      "loss": 1.103,
-      "step": 2025
-    },
-    {
-      "epoch": 1.9704223637765246,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.3147122444250327e-09,
-      "loss": 1.1077,
-      "step": 2026
-    },
-    {
-      "epoch": 1.971396113117215,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.1523513993060297e-09,
-      "loss": 1.1294,
-      "step": 2027
-    },
-    {
-      "epoch": 1.9723698624579056,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.9958912580975152e-09,
-      "loss": 1.0958,
-      "step": 2028
-    },
-    {
-      "epoch": 1.973343611798596,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.8453321904077226e-09,
-      "loss": 1.1004,
-      "step": 2029
-    },
-    {
-      "epoch": 1.9743173611392866,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.7006745519043688e-09,
-      "loss": 1.1096,
-      "step": 2030
-    },
-    {
-      "epoch": 1.9752911104799773,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.561918684314656e-09,
-      "loss": 1.1115,
-      "step": 2031
-    },
-    {
-      "epoch": 1.9762648598206678,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.4290649154238834e-09,
-      "loss": 1.1071,
-      "step": 2032
-    },
-    {
-      "epoch": 1.9772386091613585,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.3021135590740585e-09,
-      "loss": 1.1167,
-      "step": 2033
-    },
-    {
-      "epoch": 1.978212358502049,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.1810649151647313e-09,
-      "loss": 1.1127,
-      "step": 2034
-    },
-    {
-      "epoch": 1.9791861078427395,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.0659192696507724e-09,
-      "loss": 1.1157,
-      "step": 2035
-    },
-    {
-      "epoch": 1.98015985718343,
-      "grad_norm": 0.29296875,
-      "learning_rate": 9.566768945426518e-10,
-      "loss": 1.102,
-      "step": 2036
-    },
-    {
-      "epoch": 1.9811336065241205,
-      "grad_norm": 0.294921875,
-      "learning_rate": 8.533380479050502e-10,
-      "loss": 1.1054,
-      "step": 2037
-    },
-    {
-      "epoch": 1.982107355864811,
-      "grad_norm": 0.298828125,
-      "learning_rate": 7.559029738571366e-10,
-      "loss": 1.1196,
-      "step": 2038
-    },
-    {
-      "epoch": 1.9830811052055017,
-      "grad_norm": 0.29296875,
-      "learning_rate": 6.64371902571459e-10,
-      "loss": 1.0882,
-      "step": 2039
-    },
-    {
-      "epoch": 1.9840548545461922,
-      "grad_norm": 0.291015625,
-      "learning_rate": 5.787450502728331e-10,
-      "loss": 1.0996,
-      "step": 2040
-    },
-    {
-      "epoch": 1.9850286038868827,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.990226192391756e-10,
-      "loss": 1.1174,
-      "step": 2041
-    },
-    {
-      "epoch": 1.9860023532275735,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.2520479780011613e-10,
-      "loss": 1.1182,
-      "step": 2042
-    },
-    {
-      "epoch": 1.986976102568264,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.572917603364423e-10,
-      "loss": 1.125,
-      "step": 2043
-    },
-    {
-      "epoch": 1.9879498519089545,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.952836672800996e-10,
-      "loss": 1.1115,
-      "step": 2044
-    },
-    {
-      "epoch": 1.988923601249645,
-      "grad_norm": 0.296875,
-      "learning_rate": 2.391806651136364e-10,
-      "loss": 1.1109,
-      "step": 2045
-    },
-    {
-      "epoch": 1.9898973505903355,
-      "grad_norm": 0.29296875,
-      "learning_rate": 1.8898288637048122e-10,
-      "loss": 1.1038,
-      "step": 2046
-    },
-    {
-      "epoch": 1.990871099931026,
-      "grad_norm": 0.298828125,
-      "learning_rate": 1.446904496335555e-10,
-      "loss": 1.1211,
-      "step": 2047
-    },
-    {
-      "epoch": 1.9918448492717167,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.0630345953499543e-10,
-      "loss": 1.1194,
-      "step": 2048
-    },
-    {
-      "epoch": 1.9928185986124072,
-      "grad_norm": 0.29296875,
-      "learning_rate": 7.38220067578177e-11,
-      "loss": 1.103,
-      "step": 2049
-    },
-    {
-      "epoch": 1.993792347953098,
-      "grad_norm": 0.296875,
-      "learning_rate": 4.724616803286619e-11,
-      "loss": 1.1095,
-      "step": 2050
-    },
-    {
-      "epoch": 1.9947660972937884,
-      "grad_norm": 0.291015625,
-      "learning_rate": 2.6576006140477394e-11,
-      "loss": 1.1102,
-      "step": 2051
-    },
-    {
-      "epoch": 1.995739846634479,
-      "grad_norm": 0.28515625,
-      "learning_rate": 1.1811569910757937e-11,
-      "loss": 1.092,
-      "step": 2052
-    },
-    {
-      "epoch": 1.9967135959751694,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.952894221641689e-12,
-      "loss": 1.1276,
-      "step": 2053
-    },
-    {
-      "epoch": 1.99768734531586,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.0,
-      "loss": 1.1033,
-      "step": 2054
-    }
-  ],
-  "logging_steps": 1,
-  "max_steps": 2054,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
-  "save_steps": 1027,
-  "stateful_callbacks": {
-    "TrainerControl": {
-      "args": {
-        "should_epoch_stop": false,
-        "should_evaluate": false,
-        "should_log": false,
-        "should_save": true,
-        "should_training_stop": true
-      },
-      "attributes": {}
-    }
-  },
-  "total_flos": 5.4638533821656465e+19,
-  "train_batch_size": 2,
-  "trial_name": null,
-  "trial_params": null
-}