robometer-4b-lora-anubis / trainer_state.json
chomeed's picture
LoRA fine-tune on Anubis
c4bfbf8 verified
Raw
History Blame Contribute Delete
173 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.11428571428571428,
"eval_steps": 100,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00011428571428571428,
"grad_norm": 3.4703433513641357,
"learning_rate": 0.0,
"loss": 0.324,
"step": 1
},
{
"epoch": 0.00022857142857142857,
"grad_norm": 2.9177892208099365,
"learning_rate": 2.0000000000000002e-07,
"loss": 0.1644,
"step": 2
},
{
"epoch": 0.00034285714285714285,
"grad_norm": 3.7375617027282715,
"learning_rate": 4.0000000000000003e-07,
"loss": 0.2897,
"step": 3
},
{
"epoch": 0.00045714285714285713,
"grad_norm": 2.2080183029174805,
"learning_rate": 6.000000000000001e-07,
"loss": 0.268,
"step": 4
},
{
"epoch": 0.0005714285714285715,
"grad_norm": 0.6913930177688599,
"learning_rate": 8.000000000000001e-07,
"loss": 0.1106,
"step": 5
},
{
"epoch": 0.0006857142857142857,
"grad_norm": 0.9969800710678101,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.1054,
"step": 6
},
{
"epoch": 0.0008,
"grad_norm": 2.242238998413086,
"learning_rate": 1.2000000000000002e-06,
"loss": 0.3194,
"step": 7
},
{
"epoch": 0.0009142857142857143,
"grad_norm": 1.806040644645691,
"learning_rate": 1.4000000000000001e-06,
"loss": 0.2912,
"step": 8
},
{
"epoch": 0.0010285714285714286,
"grad_norm": 1.1823830604553223,
"learning_rate": 1.6000000000000001e-06,
"loss": 0.2587,
"step": 9
},
{
"epoch": 0.001142857142857143,
"grad_norm": 22.584657669067383,
"learning_rate": 1.8000000000000001e-06,
"loss": 0.5027,
"step": 10
},
{
"epoch": 0.001257142857142857,
"grad_norm": 2.38689923286438,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.1887,
"step": 11
},
{
"epoch": 0.0013714285714285714,
"grad_norm": 2.8879189491271973,
"learning_rate": 2.2e-06,
"loss": 0.4737,
"step": 12
},
{
"epoch": 0.0014857142857142857,
"grad_norm": 1.9093862771987915,
"learning_rate": 2.4000000000000003e-06,
"loss": 0.2569,
"step": 13
},
{
"epoch": 0.0016,
"grad_norm": 10.348184585571289,
"learning_rate": 2.6e-06,
"loss": 1.447,
"step": 14
},
{
"epoch": 0.0017142857142857142,
"grad_norm": 1.6802315711975098,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.1318,
"step": 15
},
{
"epoch": 0.0018285714285714285,
"grad_norm": 1.0553261041641235,
"learning_rate": 3e-06,
"loss": 0.0319,
"step": 16
},
{
"epoch": 0.0019428571428571429,
"grad_norm": 1.3565196990966797,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.2375,
"step": 17
},
{
"epoch": 0.002057142857142857,
"grad_norm": 3.301420211791992,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.3463,
"step": 18
},
{
"epoch": 0.0021714285714285715,
"grad_norm": 3.458338975906372,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.3056,
"step": 19
},
{
"epoch": 0.002285714285714286,
"grad_norm": 4.560081481933594,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.3856,
"step": 20
},
{
"epoch": 0.0024,
"grad_norm": 16.172210693359375,
"learning_rate": 4.000000000000001e-06,
"loss": 0.143,
"step": 21
},
{
"epoch": 0.002514285714285714,
"grad_norm": 2.4496405124664307,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.1181,
"step": 22
},
{
"epoch": 0.0026285714285714285,
"grad_norm": 4.017287731170654,
"learning_rate": 4.4e-06,
"loss": 0.6762,
"step": 23
},
{
"epoch": 0.002742857142857143,
"grad_norm": 1.307288408279419,
"learning_rate": 4.600000000000001e-06,
"loss": 0.2174,
"step": 24
},
{
"epoch": 0.002857142857142857,
"grad_norm": 12.720907211303711,
"learning_rate": 4.800000000000001e-06,
"loss": 1.0025,
"step": 25
},
{
"epoch": 0.0029714285714285715,
"grad_norm": 0.8936770558357239,
"learning_rate": 5e-06,
"loss": 0.1919,
"step": 26
},
{
"epoch": 0.003085714285714286,
"grad_norm": 0.7833129167556763,
"learning_rate": 5.2e-06,
"loss": 0.2149,
"step": 27
},
{
"epoch": 0.0032,
"grad_norm": 1.2059752941131592,
"learning_rate": 5.400000000000001e-06,
"loss": 0.196,
"step": 28
},
{
"epoch": 0.0033142857142857145,
"grad_norm": 3.7683944702148438,
"learning_rate": 5.600000000000001e-06,
"loss": 0.3182,
"step": 29
},
{
"epoch": 0.0034285714285714284,
"grad_norm": 0.9674514532089233,
"learning_rate": 5.8e-06,
"loss": 0.2353,
"step": 30
},
{
"epoch": 0.0035428571428571427,
"grad_norm": 0.8497799038887024,
"learning_rate": 6e-06,
"loss": 0.1224,
"step": 31
},
{
"epoch": 0.003657142857142857,
"grad_norm": 11.138529777526855,
"learning_rate": 6.200000000000001e-06,
"loss": 0.9922,
"step": 32
},
{
"epoch": 0.0037714285714285714,
"grad_norm": 1.0041792392730713,
"learning_rate": 6.4000000000000006e-06,
"loss": 0.2895,
"step": 33
},
{
"epoch": 0.0038857142857142857,
"grad_norm": 1.2810606956481934,
"learning_rate": 6.600000000000001e-06,
"loss": 0.1689,
"step": 34
},
{
"epoch": 0.004,
"grad_norm": 3.3858723640441895,
"learning_rate": 6.800000000000001e-06,
"loss": 0.4356,
"step": 35
},
{
"epoch": 0.004114285714285714,
"grad_norm": 0.5926999449729919,
"learning_rate": 7e-06,
"loss": 0.103,
"step": 36
},
{
"epoch": 0.004228571428571429,
"grad_norm": 12.412001609802246,
"learning_rate": 7.2000000000000005e-06,
"loss": 0.4322,
"step": 37
},
{
"epoch": 0.004342857142857143,
"grad_norm": 0.6445096135139465,
"learning_rate": 7.4e-06,
"loss": 0.1149,
"step": 38
},
{
"epoch": 0.0044571428571428574,
"grad_norm": 0.5666664242744446,
"learning_rate": 7.600000000000001e-06,
"loss": 0.0917,
"step": 39
},
{
"epoch": 0.004571428571428572,
"grad_norm": 20.36823272705078,
"learning_rate": 7.800000000000002e-06,
"loss": 0.7805,
"step": 40
},
{
"epoch": 0.004685714285714286,
"grad_norm": 0.8390617370605469,
"learning_rate": 8.000000000000001e-06,
"loss": 0.208,
"step": 41
},
{
"epoch": 0.0048,
"grad_norm": 6.966379165649414,
"learning_rate": 8.2e-06,
"loss": 0.1552,
"step": 42
},
{
"epoch": 0.004914285714285714,
"grad_norm": 1.1588397026062012,
"learning_rate": 8.400000000000001e-06,
"loss": 0.121,
"step": 43
},
{
"epoch": 0.005028571428571428,
"grad_norm": 6.585827350616455,
"learning_rate": 8.6e-06,
"loss": 0.3496,
"step": 44
},
{
"epoch": 0.005142857142857143,
"grad_norm": 1.3045626878738403,
"learning_rate": 8.8e-06,
"loss": 0.3095,
"step": 45
},
{
"epoch": 0.005257142857142857,
"grad_norm": 0.9495505690574646,
"learning_rate": 9e-06,
"loss": 0.1991,
"step": 46
},
{
"epoch": 0.005371428571428571,
"grad_norm": 0.7030600309371948,
"learning_rate": 9.200000000000002e-06,
"loss": 0.1816,
"step": 47
},
{
"epoch": 0.005485714285714286,
"grad_norm": 0.5981210470199585,
"learning_rate": 9.4e-06,
"loss": 0.1054,
"step": 48
},
{
"epoch": 0.0056,
"grad_norm": 1.8723831176757812,
"learning_rate": 9.600000000000001e-06,
"loss": 0.2229,
"step": 49
},
{
"epoch": 0.005714285714285714,
"grad_norm": 0.6734579801559448,
"learning_rate": 9.800000000000001e-06,
"loss": 0.1186,
"step": 50
},
{
"epoch": 0.005828571428571429,
"grad_norm": 4.051860332489014,
"learning_rate": 1e-05,
"loss": 0.3937,
"step": 51
},
{
"epoch": 0.005942857142857143,
"grad_norm": 0.988488495349884,
"learning_rate": 1.02e-05,
"loss": 0.1942,
"step": 52
},
{
"epoch": 0.006057142857142857,
"grad_norm": 1.15732741355896,
"learning_rate": 1.04e-05,
"loss": 0.2232,
"step": 53
},
{
"epoch": 0.006171428571428572,
"grad_norm": 0.6378714442253113,
"learning_rate": 1.0600000000000002e-05,
"loss": 0.1053,
"step": 54
},
{
"epoch": 0.006285714285714286,
"grad_norm": 3.0973570346832275,
"learning_rate": 1.0800000000000002e-05,
"loss": 0.1852,
"step": 55
},
{
"epoch": 0.0064,
"grad_norm": 0.867125928401947,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.2052,
"step": 56
},
{
"epoch": 0.006514285714285715,
"grad_norm": 1.3798433542251587,
"learning_rate": 1.1200000000000001e-05,
"loss": 0.155,
"step": 57
},
{
"epoch": 0.006628571428571429,
"grad_norm": 0.5028775334358215,
"learning_rate": 1.14e-05,
"loss": 0.0907,
"step": 58
},
{
"epoch": 0.0067428571428571425,
"grad_norm": 11.206276893615723,
"learning_rate": 1.16e-05,
"loss": 0.8554,
"step": 59
},
{
"epoch": 0.006857142857142857,
"grad_norm": 0.36292576789855957,
"learning_rate": 1.18e-05,
"loss": 0.1062,
"step": 60
},
{
"epoch": 0.006971428571428571,
"grad_norm": 1.4553862810134888,
"learning_rate": 1.2e-05,
"loss": 0.2442,
"step": 61
},
{
"epoch": 0.0070857142857142855,
"grad_norm": 32.60602569580078,
"learning_rate": 1.22e-05,
"loss": 2.7087,
"step": 62
},
{
"epoch": 0.0072,
"grad_norm": 22.60771942138672,
"learning_rate": 1.2400000000000002e-05,
"loss": 0.653,
"step": 63
},
{
"epoch": 0.007314285714285714,
"grad_norm": 1.8575780391693115,
"learning_rate": 1.2600000000000001e-05,
"loss": 0.3142,
"step": 64
},
{
"epoch": 0.0074285714285714285,
"grad_norm": 0.3562605679035187,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.0991,
"step": 65
},
{
"epoch": 0.007542857142857143,
"grad_norm": 12.067700386047363,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.3462,
"step": 66
},
{
"epoch": 0.007657142857142857,
"grad_norm": 1.2796293497085571,
"learning_rate": 1.3200000000000002e-05,
"loss": 0.1859,
"step": 67
},
{
"epoch": 0.0077714285714285715,
"grad_norm": 0.7668793201446533,
"learning_rate": 1.3400000000000002e-05,
"loss": 0.1859,
"step": 68
},
{
"epoch": 0.007885714285714286,
"grad_norm": 7.48372745513916,
"learning_rate": 1.3600000000000002e-05,
"loss": 0.281,
"step": 69
},
{
"epoch": 0.008,
"grad_norm": 1.629469633102417,
"learning_rate": 1.38e-05,
"loss": 0.1301,
"step": 70
},
{
"epoch": 0.008114285714285715,
"grad_norm": 1.2479746341705322,
"learning_rate": 1.4e-05,
"loss": 0.144,
"step": 71
},
{
"epoch": 0.008228571428571429,
"grad_norm": 0.5008277297019958,
"learning_rate": 1.4200000000000001e-05,
"loss": 0.1742,
"step": 72
},
{
"epoch": 0.008342857142857143,
"grad_norm": 0.8036436438560486,
"learning_rate": 1.4400000000000001e-05,
"loss": 0.1955,
"step": 73
},
{
"epoch": 0.008457142857142858,
"grad_norm": 0.8107599020004272,
"learning_rate": 1.46e-05,
"loss": 0.0887,
"step": 74
},
{
"epoch": 0.008571428571428572,
"grad_norm": 1.0926967859268188,
"learning_rate": 1.48e-05,
"loss": 0.151,
"step": 75
},
{
"epoch": 0.008685714285714286,
"grad_norm": 7.336767673492432,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.4715,
"step": 76
},
{
"epoch": 0.0088,
"grad_norm": 2.741205930709839,
"learning_rate": 1.5200000000000002e-05,
"loss": 0.3106,
"step": 77
},
{
"epoch": 0.008914285714285715,
"grad_norm": 0.5145547986030579,
"learning_rate": 1.54e-05,
"loss": 0.0079,
"step": 78
},
{
"epoch": 0.00902857142857143,
"grad_norm": 1.00886070728302,
"learning_rate": 1.5600000000000003e-05,
"loss": 0.0989,
"step": 79
},
{
"epoch": 0.009142857142857144,
"grad_norm": 1.0324081182479858,
"learning_rate": 1.58e-05,
"loss": 0.1028,
"step": 80
},
{
"epoch": 0.009257142857142858,
"grad_norm": 0.904427707195282,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.2052,
"step": 81
},
{
"epoch": 0.009371428571428572,
"grad_norm": 1.225813627243042,
"learning_rate": 1.62e-05,
"loss": 0.1882,
"step": 82
},
{
"epoch": 0.009485714285714287,
"grad_norm": 1.0993331670761108,
"learning_rate": 1.64e-05,
"loss": 0.1641,
"step": 83
},
{
"epoch": 0.0096,
"grad_norm": 18.441650390625,
"learning_rate": 1.66e-05,
"loss": 0.7751,
"step": 84
},
{
"epoch": 0.009714285714285713,
"grad_norm": 4.295007705688477,
"learning_rate": 1.6800000000000002e-05,
"loss": 0.0704,
"step": 85
},
{
"epoch": 0.009828571428571428,
"grad_norm": 0.7962968945503235,
"learning_rate": 1.7e-05,
"loss": 0.17,
"step": 86
},
{
"epoch": 0.009942857142857142,
"grad_norm": 9.219826698303223,
"learning_rate": 1.72e-05,
"loss": 0.9301,
"step": 87
},
{
"epoch": 0.010057142857142857,
"grad_norm": 11.585982322692871,
"learning_rate": 1.7400000000000003e-05,
"loss": 0.4984,
"step": 88
},
{
"epoch": 0.01017142857142857,
"grad_norm": 1.313962459564209,
"learning_rate": 1.76e-05,
"loss": 0.2093,
"step": 89
},
{
"epoch": 0.010285714285714285,
"grad_norm": 15.045809745788574,
"learning_rate": 1.7800000000000002e-05,
"loss": 1.0262,
"step": 90
},
{
"epoch": 0.0104,
"grad_norm": 0.8079013228416443,
"learning_rate": 1.8e-05,
"loss": 0.0847,
"step": 91
},
{
"epoch": 0.010514285714285714,
"grad_norm": 1.0459412336349487,
"learning_rate": 1.8200000000000002e-05,
"loss": 0.2191,
"step": 92
},
{
"epoch": 0.010628571428571428,
"grad_norm": 0.9380034804344177,
"learning_rate": 1.8400000000000003e-05,
"loss": 0.176,
"step": 93
},
{
"epoch": 0.010742857142857143,
"grad_norm": 2.09692645072937,
"learning_rate": 1.86e-05,
"loss": 0.2961,
"step": 94
},
{
"epoch": 0.010857142857142857,
"grad_norm": 5.74895715713501,
"learning_rate": 1.88e-05,
"loss": 0.2447,
"step": 95
},
{
"epoch": 0.010971428571428571,
"grad_norm": 3.7462525367736816,
"learning_rate": 1.9e-05,
"loss": 0.2952,
"step": 96
},
{
"epoch": 0.011085714285714286,
"grad_norm": 1.1957429647445679,
"learning_rate": 1.9200000000000003e-05,
"loss": 0.115,
"step": 97
},
{
"epoch": 0.0112,
"grad_norm": 0.5693152546882629,
"learning_rate": 1.94e-05,
"loss": 0.0168,
"step": 98
},
{
"epoch": 0.011314285714285714,
"grad_norm": 1.0899547338485718,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.1928,
"step": 99
},
{
"epoch": 0.011428571428571429,
"grad_norm": 0.6118483543395996,
"learning_rate": 1.98e-05,
"loss": 0.1797,
"step": 100
},
{
"epoch": 0.011542857142857143,
"grad_norm": 0.3671252727508545,
"learning_rate": 2e-05,
"loss": 0.0801,
"step": 101
},
{
"epoch": 0.011657142857142857,
"grad_norm": 0.7626099586486816,
"learning_rate": 1.9999939076577906e-05,
"loss": 0.191,
"step": 102
},
{
"epoch": 0.011771428571428572,
"grad_norm": 1.6209627389907837,
"learning_rate": 1.9999756307053947e-05,
"loss": 0.1627,
"step": 103
},
{
"epoch": 0.011885714285714286,
"grad_norm": 1.5343306064605713,
"learning_rate": 1.9999451693655125e-05,
"loss": 0.2312,
"step": 104
},
{
"epoch": 0.012,
"grad_norm": 0.925738513469696,
"learning_rate": 1.9999025240093045e-05,
"loss": 0.2015,
"step": 105
},
{
"epoch": 0.012114285714285715,
"grad_norm": 2.1403286457061768,
"learning_rate": 1.9998476951563914e-05,
"loss": 0.1291,
"step": 106
},
{
"epoch": 0.012228571428571429,
"grad_norm": 0.8222343325614929,
"learning_rate": 1.9997806834748455e-05,
"loss": 0.0968,
"step": 107
},
{
"epoch": 0.012342857142857143,
"grad_norm": 0.8072381615638733,
"learning_rate": 1.9997014897811834e-05,
"loss": 0.1688,
"step": 108
},
{
"epoch": 0.012457142857142858,
"grad_norm": 1.582384705543518,
"learning_rate": 1.9996101150403543e-05,
"loss": 0.2666,
"step": 109
},
{
"epoch": 0.012571428571428572,
"grad_norm": 6.984894752502441,
"learning_rate": 1.9995065603657317e-05,
"loss": 0.3292,
"step": 110
},
{
"epoch": 0.012685714285714286,
"grad_norm": 1.1681904792785645,
"learning_rate": 1.999390827019096e-05,
"loss": 0.0846,
"step": 111
},
{
"epoch": 0.0128,
"grad_norm": 0.9094777703285217,
"learning_rate": 1.999262916410621e-05,
"loss": 0.0914,
"step": 112
},
{
"epoch": 0.012914285714285715,
"grad_norm": 0.5362823605537415,
"learning_rate": 1.9991228300988586e-05,
"loss": 0.084,
"step": 113
},
{
"epoch": 0.01302857142857143,
"grad_norm": 1.0594851970672607,
"learning_rate": 1.998970569790715e-05,
"loss": 0.132,
"step": 114
},
{
"epoch": 0.013142857142857144,
"grad_norm": 1.316963791847229,
"learning_rate": 1.9988061373414342e-05,
"loss": 0.2268,
"step": 115
},
{
"epoch": 0.013257142857142858,
"grad_norm": 7.658836364746094,
"learning_rate": 1.9986295347545738e-05,
"loss": 0.3578,
"step": 116
},
{
"epoch": 0.01337142857142857,
"grad_norm": 0.8805353045463562,
"learning_rate": 1.9984407641819812e-05,
"loss": 0.1483,
"step": 117
},
{
"epoch": 0.013485714285714285,
"grad_norm": 0.7816386818885803,
"learning_rate": 1.9982398279237657e-05,
"loss": 0.0961,
"step": 118
},
{
"epoch": 0.0136,
"grad_norm": 1.073937177658081,
"learning_rate": 1.9980267284282718e-05,
"loss": 0.1188,
"step": 119
},
{
"epoch": 0.013714285714285714,
"grad_norm": 0.8910870552062988,
"learning_rate": 1.9978014682920503e-05,
"loss": 0.1841,
"step": 120
},
{
"epoch": 0.013828571428571428,
"grad_norm": 0.9908338189125061,
"learning_rate": 1.9975640502598243e-05,
"loss": 0.1633,
"step": 121
},
{
"epoch": 0.013942857142857142,
"grad_norm": 0.5217012166976929,
"learning_rate": 1.997314477224458e-05,
"loss": 0.1017,
"step": 122
},
{
"epoch": 0.014057142857142857,
"grad_norm": 0.9262743592262268,
"learning_rate": 1.9970527522269204e-05,
"loss": 0.1663,
"step": 123
},
{
"epoch": 0.014171428571428571,
"grad_norm": 0.7330294251441956,
"learning_rate": 1.9967788784562474e-05,
"loss": 0.0974,
"step": 124
},
{
"epoch": 0.014285714285714285,
"grad_norm": 2.3975419998168945,
"learning_rate": 1.9964928592495046e-05,
"loss": 0.1281,
"step": 125
},
{
"epoch": 0.0144,
"grad_norm": 4.172351360321045,
"learning_rate": 1.9961946980917457e-05,
"loss": 0.1746,
"step": 126
},
{
"epoch": 0.014514285714285714,
"grad_norm": 0.3429124355316162,
"learning_rate": 1.9958843986159705e-05,
"loss": 0.0731,
"step": 127
},
{
"epoch": 0.014628571428571428,
"grad_norm": 0.6417416930198669,
"learning_rate": 1.99556196460308e-05,
"loss": 0.1037,
"step": 128
},
{
"epoch": 0.014742857142857143,
"grad_norm": 0.7663334608078003,
"learning_rate": 1.9952273999818312e-05,
"loss": 0.0918,
"step": 129
},
{
"epoch": 0.014857142857142857,
"grad_norm": 1.746813416481018,
"learning_rate": 1.9948807088287884e-05,
"loss": 0.2034,
"step": 130
},
{
"epoch": 0.014971428571428571,
"grad_norm": 4.234859943389893,
"learning_rate": 1.9945218953682736e-05,
"loss": 0.1856,
"step": 131
},
{
"epoch": 0.015085714285714286,
"grad_norm": 1.4514203071594238,
"learning_rate": 1.9941509639723155e-05,
"loss": 0.1699,
"step": 132
},
{
"epoch": 0.0152,
"grad_norm": 0.7403672337532043,
"learning_rate": 1.9937679191605964e-05,
"loss": 0.0203,
"step": 133
},
{
"epoch": 0.015314285714285714,
"grad_norm": 0.6282806992530823,
"learning_rate": 1.9933727656003964e-05,
"loss": 0.0893,
"step": 134
},
{
"epoch": 0.015428571428571429,
"grad_norm": 0.8490723371505737,
"learning_rate": 1.992965508106537e-05,
"loss": 0.1563,
"step": 135
},
{
"epoch": 0.015542857142857143,
"grad_norm": 0.5892636775970459,
"learning_rate": 1.9925461516413224e-05,
"loss": 0.0926,
"step": 136
},
{
"epoch": 0.015657142857142856,
"grad_norm": 0.3128161132335663,
"learning_rate": 1.9921147013144782e-05,
"loss": 0.009,
"step": 137
},
{
"epoch": 0.01577142857142857,
"grad_norm": 0.6765373349189758,
"learning_rate": 1.9916711623830904e-05,
"loss": 0.0953,
"step": 138
},
{
"epoch": 0.015885714285714284,
"grad_norm": 0.6584412455558777,
"learning_rate": 1.991215540251542e-05,
"loss": 0.092,
"step": 139
},
{
"epoch": 0.016,
"grad_norm": 1.7359390258789062,
"learning_rate": 1.9907478404714438e-05,
"loss": 0.1502,
"step": 140
},
{
"epoch": 0.016114285714285713,
"grad_norm": 0.8604278564453125,
"learning_rate": 1.9902680687415704e-05,
"loss": 0.1507,
"step": 141
},
{
"epoch": 0.01622857142857143,
"grad_norm": 1.466235876083374,
"learning_rate": 1.989776230907789e-05,
"loss": 0.1256,
"step": 142
},
{
"epoch": 0.01634285714285714,
"grad_norm": 1.634239912033081,
"learning_rate": 1.9892723329629885e-05,
"loss": 0.1888,
"step": 143
},
{
"epoch": 0.016457142857142858,
"grad_norm": 0.8107671737670898,
"learning_rate": 1.988756381047006e-05,
"loss": 0.1522,
"step": 144
},
{
"epoch": 0.01657142857142857,
"grad_norm": 0.8587020039558411,
"learning_rate": 1.988228381446553e-05,
"loss": 0.0855,
"step": 145
},
{
"epoch": 0.016685714285714286,
"grad_norm": 0.6909587979316711,
"learning_rate": 1.9876883405951378e-05,
"loss": 0.0874,
"step": 146
},
{
"epoch": 0.0168,
"grad_norm": 0.7107747197151184,
"learning_rate": 1.987136265072988e-05,
"loss": 0.1611,
"step": 147
},
{
"epoch": 0.016914285714285715,
"grad_norm": 0.7703253030776978,
"learning_rate": 1.9865721616069695e-05,
"loss": 0.1958,
"step": 148
},
{
"epoch": 0.017028571428571428,
"grad_norm": 0.7867444157600403,
"learning_rate": 1.985996037070505e-05,
"loss": 0.1737,
"step": 149
},
{
"epoch": 0.017142857142857144,
"grad_norm": 2.106637954711914,
"learning_rate": 1.9854078984834904e-05,
"loss": 0.2028,
"step": 150
},
{
"epoch": 0.017257142857142856,
"grad_norm": 0.09181072562932968,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.0021,
"step": 151
},
{
"epoch": 0.017371428571428572,
"grad_norm": 1.2006891965866089,
"learning_rate": 1.984195607969242e-05,
"loss": 0.1091,
"step": 152
},
{
"epoch": 0.017485714285714285,
"grad_norm": 0.30722373723983765,
"learning_rate": 1.983571470813386e-05,
"loss": 0.0774,
"step": 153
},
{
"epoch": 0.0176,
"grad_norm": 0.3324722945690155,
"learning_rate": 1.9829353491495545e-05,
"loss": 0.0689,
"step": 154
},
{
"epoch": 0.017714285714285714,
"grad_norm": 7.893986225128174,
"learning_rate": 1.982287250728689e-05,
"loss": 0.5853,
"step": 155
},
{
"epoch": 0.01782857142857143,
"grad_norm": 0.3751853406429291,
"learning_rate": 1.9816271834476642e-05,
"loss": 0.0707,
"step": 156
},
{
"epoch": 0.017942857142857142,
"grad_norm": 1.3349276781082153,
"learning_rate": 1.9809551553491918e-05,
"loss": 0.0783,
"step": 157
},
{
"epoch": 0.01805714285714286,
"grad_norm": 0.4436684846878052,
"learning_rate": 1.9802711746217222e-05,
"loss": 0.0686,
"step": 158
},
{
"epoch": 0.01817142857142857,
"grad_norm": 1.1769542694091797,
"learning_rate": 1.979575249599344e-05,
"loss": 0.0957,
"step": 159
},
{
"epoch": 0.018285714285714287,
"grad_norm": 1.6521239280700684,
"learning_rate": 1.9788673887616852e-05,
"loss": 0.1993,
"step": 160
},
{
"epoch": 0.0184,
"grad_norm": 0.40775999426841736,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.0764,
"step": 161
},
{
"epoch": 0.018514285714285716,
"grad_norm": 1.5056344270706177,
"learning_rate": 1.9774158942860962e-05,
"loss": 0.2197,
"step": 162
},
{
"epoch": 0.01862857142857143,
"grad_norm": 1.272965431213379,
"learning_rate": 1.9766722783341682e-05,
"loss": 0.1882,
"step": 163
},
{
"epoch": 0.018742857142857144,
"grad_norm": 0.3193214535713196,
"learning_rate": 1.9759167619387474e-05,
"loss": 0.0743,
"step": 164
},
{
"epoch": 0.018857142857142857,
"grad_norm": 0.2415762096643448,
"learning_rate": 1.9751493543055634e-05,
"loss": 0.0757,
"step": 165
},
{
"epoch": 0.018971428571428573,
"grad_norm": 0.599396288394928,
"learning_rate": 1.9743700647852356e-05,
"loss": 0.1398,
"step": 166
},
{
"epoch": 0.019085714285714286,
"grad_norm": 0.9281499981880188,
"learning_rate": 1.9735789028731603e-05,
"loss": 0.1892,
"step": 167
},
{
"epoch": 0.0192,
"grad_norm": 1.914065957069397,
"learning_rate": 1.972775878209397e-05,
"loss": 0.1222,
"step": 168
},
{
"epoch": 0.019314285714285714,
"grad_norm": 0.7601314187049866,
"learning_rate": 1.9719610005785466e-05,
"loss": 0.0892,
"step": 169
},
{
"epoch": 0.019428571428571427,
"grad_norm": 1.8253015279769897,
"learning_rate": 1.971134279909636e-05,
"loss": 0.1047,
"step": 170
},
{
"epoch": 0.019542857142857143,
"grad_norm": 0.6221289038658142,
"learning_rate": 1.9702957262759964e-05,
"loss": 0.0845,
"step": 171
},
{
"epoch": 0.019657142857142856,
"grad_norm": 1.6722835302352905,
"learning_rate": 1.9694453498951392e-05,
"loss": 0.174,
"step": 172
},
{
"epoch": 0.019771428571428572,
"grad_norm": 0.8359847664833069,
"learning_rate": 1.9685831611286312e-05,
"loss": 0.1647,
"step": 173
},
{
"epoch": 0.019885714285714284,
"grad_norm": 0.3755902051925659,
"learning_rate": 1.9677091704819714e-05,
"loss": 0.0751,
"step": 174
},
{
"epoch": 0.02,
"grad_norm": 0.6360909938812256,
"learning_rate": 1.9668233886044597e-05,
"loss": 0.1646,
"step": 175
},
{
"epoch": 0.020114285714285713,
"grad_norm": 0.8718181252479553,
"learning_rate": 1.9659258262890683e-05,
"loss": 0.1555,
"step": 176
},
{
"epoch": 0.02022857142857143,
"grad_norm": 1.2214370965957642,
"learning_rate": 1.9650164944723116e-05,
"loss": 0.1722,
"step": 177
},
{
"epoch": 0.02034285714285714,
"grad_norm": 0.60230952501297,
"learning_rate": 1.96409540423411e-05,
"loss": 0.0754,
"step": 178
},
{
"epoch": 0.020457142857142858,
"grad_norm": 0.6614243984222412,
"learning_rate": 1.9631625667976584e-05,
"loss": 0.0969,
"step": 179
},
{
"epoch": 0.02057142857142857,
"grad_norm": 0.5244930982589722,
"learning_rate": 1.9622179935292855e-05,
"loss": 0.1346,
"step": 180
},
{
"epoch": 0.020685714285714286,
"grad_norm": 0.5967604517936707,
"learning_rate": 1.961261695938319e-05,
"loss": 0.0778,
"step": 181
},
{
"epoch": 0.0208,
"grad_norm": 0.868948221206665,
"learning_rate": 1.9602936856769432e-05,
"loss": 0.1562,
"step": 182
},
{
"epoch": 0.020914285714285715,
"grad_norm": 0.6504230499267578,
"learning_rate": 1.9593139745400575e-05,
"loss": 0.1488,
"step": 183
},
{
"epoch": 0.021028571428571428,
"grad_norm": 2.4615097045898438,
"learning_rate": 1.9583225744651334e-05,
"loss": 0.1262,
"step": 184
},
{
"epoch": 0.021142857142857144,
"grad_norm": 1.6685512065887451,
"learning_rate": 1.9573194975320672e-05,
"loss": 0.0417,
"step": 185
},
{
"epoch": 0.021257142857142856,
"grad_norm": 0.4547988772392273,
"learning_rate": 1.9563047559630356e-05,
"loss": 0.0811,
"step": 186
},
{
"epoch": 0.021371428571428572,
"grad_norm": 7.214776039123535,
"learning_rate": 1.9552783621223437e-05,
"loss": 0.1567,
"step": 187
},
{
"epoch": 0.021485714285714285,
"grad_norm": 0.6367964148521423,
"learning_rate": 1.954240328516277e-05,
"loss": 0.0757,
"step": 188
},
{
"epoch": 0.0216,
"grad_norm": 1.4471689462661743,
"learning_rate": 1.9531906677929472e-05,
"loss": 0.1812,
"step": 189
},
{
"epoch": 0.021714285714285714,
"grad_norm": 0.5046927332878113,
"learning_rate": 1.9521293927421388e-05,
"loss": 0.0724,
"step": 190
},
{
"epoch": 0.02182857142857143,
"grad_norm": 0.5476648211479187,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.0727,
"step": 191
},
{
"epoch": 0.021942857142857142,
"grad_norm": 0.8866369724273682,
"learning_rate": 1.9499720515246524e-05,
"loss": 0.177,
"step": 192
},
{
"epoch": 0.02205714285714286,
"grad_norm": 0.34143781661987305,
"learning_rate": 1.9488760116444966e-05,
"loss": 0.0728,
"step": 193
},
{
"epoch": 0.02217142857142857,
"grad_norm": 0.9685168862342834,
"learning_rate": 1.947768410009586e-05,
"loss": 0.093,
"step": 194
},
{
"epoch": 0.022285714285714287,
"grad_norm": 0.7794148921966553,
"learning_rate": 1.9466492601156964e-05,
"loss": 0.1428,
"step": 195
},
{
"epoch": 0.0224,
"grad_norm": 1.6614881753921509,
"learning_rate": 1.945518575599317e-05,
"loss": 0.1724,
"step": 196
},
{
"epoch": 0.022514285714285716,
"grad_norm": 0.5407899618148804,
"learning_rate": 1.944376370237481e-05,
"loss": 0.0718,
"step": 197
},
{
"epoch": 0.02262857142857143,
"grad_norm": 1.0045008659362793,
"learning_rate": 1.943222657947601e-05,
"loss": 0.1789,
"step": 198
},
{
"epoch": 0.022742857142857145,
"grad_norm": 0.4343376159667969,
"learning_rate": 1.942057452787297e-05,
"loss": 0.0631,
"step": 199
},
{
"epoch": 0.022857142857142857,
"grad_norm": 1.169718861579895,
"learning_rate": 1.9408807689542257e-05,
"loss": 0.0934,
"step": 200
},
{
"epoch": 0.02297142857142857,
"grad_norm": 19.440418243408203,
"learning_rate": 1.9396926207859085e-05,
"loss": 1.3785,
"step": 201
},
{
"epoch": 0.023085714285714286,
"grad_norm": 1.1699557304382324,
"learning_rate": 1.938493022759556e-05,
"loss": 0.1625,
"step": 202
},
{
"epoch": 0.0232,
"grad_norm": 0.5967357754707336,
"learning_rate": 1.937281989491892e-05,
"loss": 0.1344,
"step": 203
},
{
"epoch": 0.023314285714285714,
"grad_norm": 0.8613458871841431,
"learning_rate": 1.9360595357389735e-05,
"loss": 0.1544,
"step": 204
},
{
"epoch": 0.023428571428571427,
"grad_norm": 0.4720660448074341,
"learning_rate": 1.9348256763960146e-05,
"loss": 0.0683,
"step": 205
},
{
"epoch": 0.023542857142857143,
"grad_norm": 0.5465097427368164,
"learning_rate": 1.9335804264972018e-05,
"loss": 0.1492,
"step": 206
},
{
"epoch": 0.023657142857142856,
"grad_norm": 0.48513975739479065,
"learning_rate": 1.9323238012155125e-05,
"loss": 0.0682,
"step": 207
},
{
"epoch": 0.023771428571428572,
"grad_norm": 0.6539943814277649,
"learning_rate": 1.9310558158625286e-05,
"loss": 0.1365,
"step": 208
},
{
"epoch": 0.023885714285714284,
"grad_norm": 0.6682160496711731,
"learning_rate": 1.9297764858882516e-05,
"loss": 0.1646,
"step": 209
},
{
"epoch": 0.024,
"grad_norm": 1.122200608253479,
"learning_rate": 1.9284858268809135e-05,
"loss": 0.1789,
"step": 210
},
{
"epoch": 0.024114285714285713,
"grad_norm": 1.3054819107055664,
"learning_rate": 1.9271838545667876e-05,
"loss": 0.0918,
"step": 211
},
{
"epoch": 0.02422857142857143,
"grad_norm": 9.69206714630127,
"learning_rate": 1.925870584809995e-05,
"loss": 0.7861,
"step": 212
},
{
"epoch": 0.024342857142857142,
"grad_norm": 0.4460507929325104,
"learning_rate": 1.9245460336123136e-05,
"loss": 0.0649,
"step": 213
},
{
"epoch": 0.024457142857142858,
"grad_norm": 1.0854889154434204,
"learning_rate": 1.923210217112981e-05,
"loss": 0.1535,
"step": 214
},
{
"epoch": 0.02457142857142857,
"grad_norm": 0.7128372192382812,
"learning_rate": 1.9218631515885007e-05,
"loss": 0.0873,
"step": 215
},
{
"epoch": 0.024685714285714287,
"grad_norm": 3.261958599090576,
"learning_rate": 1.9205048534524405e-05,
"loss": 0.1603,
"step": 216
},
{
"epoch": 0.0248,
"grad_norm": 0.7447339296340942,
"learning_rate": 1.9191353392552346e-05,
"loss": 0.0801,
"step": 217
},
{
"epoch": 0.024914285714285715,
"grad_norm": 2.8254079818725586,
"learning_rate": 1.9177546256839814e-05,
"loss": 0.1528,
"step": 218
},
{
"epoch": 0.025028571428571428,
"grad_norm": 0.8378766775131226,
"learning_rate": 1.9163627295622397e-05,
"loss": 0.1583,
"step": 219
},
{
"epoch": 0.025142857142857144,
"grad_norm": 0.8295012712478638,
"learning_rate": 1.914959667849825e-05,
"loss": 0.142,
"step": 220
},
{
"epoch": 0.025257142857142856,
"grad_norm": 0.6758121848106384,
"learning_rate": 1.913545457642601e-05,
"loss": 0.1376,
"step": 221
},
{
"epoch": 0.025371428571428573,
"grad_norm": 4.832571506500244,
"learning_rate": 1.9121201161722732e-05,
"loss": 0.1677,
"step": 222
},
{
"epoch": 0.025485714285714285,
"grad_norm": 1.1076773405075073,
"learning_rate": 1.910683660806177e-05,
"loss": 0.1401,
"step": 223
},
{
"epoch": 0.0256,
"grad_norm": 2.402377128601074,
"learning_rate": 1.9092361090470688e-05,
"loss": 0.1349,
"step": 224
},
{
"epoch": 0.025714285714285714,
"grad_norm": 1.7865934371948242,
"learning_rate": 1.907777478532909e-05,
"loss": 0.2018,
"step": 225
},
{
"epoch": 0.02582857142857143,
"grad_norm": 0.6209324598312378,
"learning_rate": 1.9063077870366504e-05,
"loss": 0.1414,
"step": 226
},
{
"epoch": 0.025942857142857143,
"grad_norm": 1.7925758361816406,
"learning_rate": 1.9048270524660197e-05,
"loss": 0.1909,
"step": 227
},
{
"epoch": 0.02605714285714286,
"grad_norm": 1.588703989982605,
"learning_rate": 1.903335292863301e-05,
"loss": 0.1025,
"step": 228
},
{
"epoch": 0.02617142857142857,
"grad_norm": 1.9658430814743042,
"learning_rate": 1.901832526405114e-05,
"loss": 0.1645,
"step": 229
},
{
"epoch": 0.026285714285714287,
"grad_norm": 1.1384143829345703,
"learning_rate": 1.9003187714021936e-05,
"loss": 0.0915,
"step": 230
},
{
"epoch": 0.0264,
"grad_norm": 0.04745587334036827,
"learning_rate": 1.8987940462991673e-05,
"loss": 0.0014,
"step": 231
},
{
"epoch": 0.026514285714285716,
"grad_norm": 0.6234807968139648,
"learning_rate": 1.8972583696743284e-05,
"loss": 0.0839,
"step": 232
},
{
"epoch": 0.02662857142857143,
"grad_norm": 0.7348832488059998,
"learning_rate": 1.895711760239413e-05,
"loss": 0.154,
"step": 233
},
{
"epoch": 0.02674285714285714,
"grad_norm": 0.30456089973449707,
"learning_rate": 1.8941542368393683e-05,
"loss": 0.0619,
"step": 234
},
{
"epoch": 0.026857142857142857,
"grad_norm": 0.9593465924263,
"learning_rate": 1.892585818452126e-05,
"loss": 0.1701,
"step": 235
},
{
"epoch": 0.02697142857142857,
"grad_norm": 0.48648354411125183,
"learning_rate": 1.891006524188368e-05,
"loss": 0.0699,
"step": 236
},
{
"epoch": 0.027085714285714286,
"grad_norm": 0.5593926906585693,
"learning_rate": 1.889416373291298e-05,
"loss": 0.0667,
"step": 237
},
{
"epoch": 0.0272,
"grad_norm": 1.7450014352798462,
"learning_rate": 1.8878153851364013e-05,
"loss": 0.1896,
"step": 238
},
{
"epoch": 0.027314285714285715,
"grad_norm": 11.268231391906738,
"learning_rate": 1.8862035792312148e-05,
"loss": 0.1854,
"step": 239
},
{
"epoch": 0.027428571428571427,
"grad_norm": 0.8989130258560181,
"learning_rate": 1.884580975215084e-05,
"loss": 0.1297,
"step": 240
},
{
"epoch": 0.027542857142857143,
"grad_norm": 0.6240429282188416,
"learning_rate": 1.8829475928589272e-05,
"loss": 0.1386,
"step": 241
},
{
"epoch": 0.027657142857142856,
"grad_norm": 0.7558143734931946,
"learning_rate": 1.8813034520649923e-05,
"loss": 0.1303,
"step": 242
},
{
"epoch": 0.027771428571428572,
"grad_norm": 0.6398907899856567,
"learning_rate": 1.879648572866617e-05,
"loss": 0.1534,
"step": 243
},
{
"epoch": 0.027885714285714285,
"grad_norm": 0.7104907035827637,
"learning_rate": 1.8779829754279806e-05,
"loss": 0.0921,
"step": 244
},
{
"epoch": 0.028,
"grad_norm": 0.8072474002838135,
"learning_rate": 1.8763066800438638e-05,
"loss": 0.1373,
"step": 245
},
{
"epoch": 0.028114285714285713,
"grad_norm": 0.5808033347129822,
"learning_rate": 1.874619707139396e-05,
"loss": 0.1341,
"step": 246
},
{
"epoch": 0.02822857142857143,
"grad_norm": 0.6200347542762756,
"learning_rate": 1.8729220772698096e-05,
"loss": 0.0908,
"step": 247
},
{
"epoch": 0.028342857142857142,
"grad_norm": 0.5956852436065674,
"learning_rate": 1.8712138111201898e-05,
"loss": 0.0712,
"step": 248
},
{
"epoch": 0.028457142857142858,
"grad_norm": 1.1164041757583618,
"learning_rate": 1.869494929505219e-05,
"loss": 0.1544,
"step": 249
},
{
"epoch": 0.02857142857142857,
"grad_norm": 0.7912858128547668,
"learning_rate": 1.8677654533689287e-05,
"loss": 0.098,
"step": 250
},
{
"epoch": 0.028685714285714287,
"grad_norm": 1.1630598306655884,
"learning_rate": 1.866025403784439e-05,
"loss": 0.1002,
"step": 251
},
{
"epoch": 0.0288,
"grad_norm": 0.775030255317688,
"learning_rate": 1.864274801953705e-05,
"loss": 0.1416,
"step": 252
},
{
"epoch": 0.028914285714285715,
"grad_norm": 0.823024332523346,
"learning_rate": 1.8625136692072577e-05,
"loss": 0.1194,
"step": 253
},
{
"epoch": 0.029028571428571428,
"grad_norm": 0.6196932792663574,
"learning_rate": 1.860742027003944e-05,
"loss": 0.0802,
"step": 254
},
{
"epoch": 0.029142857142857144,
"grad_norm": 0.6459180116653442,
"learning_rate": 1.8589598969306646e-05,
"loss": 0.0726,
"step": 255
},
{
"epoch": 0.029257142857142857,
"grad_norm": 1.0776869058609009,
"learning_rate": 1.8571673007021124e-05,
"loss": 0.1029,
"step": 256
},
{
"epoch": 0.029371428571428573,
"grad_norm": 0.8157057166099548,
"learning_rate": 1.855364260160507e-05,
"loss": 0.1442,
"step": 257
},
{
"epoch": 0.029485714285714285,
"grad_norm": 0.35648825764656067,
"learning_rate": 1.8535507972753275e-05,
"loss": 0.0027,
"step": 258
},
{
"epoch": 0.0296,
"grad_norm": 0.4941117763519287,
"learning_rate": 1.851726934143048e-05,
"loss": 0.1332,
"step": 259
},
{
"epoch": 0.029714285714285714,
"grad_norm": 0.6328603029251099,
"learning_rate": 1.849892692986864e-05,
"loss": 0.1209,
"step": 260
},
{
"epoch": 0.02982857142857143,
"grad_norm": 1.6049089431762695,
"learning_rate": 1.848048096156426e-05,
"loss": 0.1629,
"step": 261
},
{
"epoch": 0.029942857142857143,
"grad_norm": 0.8721972703933716,
"learning_rate": 1.8461931661275642e-05,
"loss": 0.0839,
"step": 262
},
{
"epoch": 0.03005714285714286,
"grad_norm": 0.45377254486083984,
"learning_rate": 1.8443279255020153e-05,
"loss": 0.0663,
"step": 263
},
{
"epoch": 0.03017142857142857,
"grad_norm": 1.1773452758789062,
"learning_rate": 1.842452397007148e-05,
"loss": 0.1585,
"step": 264
},
{
"epoch": 0.030285714285714287,
"grad_norm": 0.9665262699127197,
"learning_rate": 1.8405666034956842e-05,
"loss": 0.1319,
"step": 265
},
{
"epoch": 0.0304,
"grad_norm": 9.409919738769531,
"learning_rate": 1.8386705679454243e-05,
"loss": 0.3934,
"step": 266
},
{
"epoch": 0.030514285714285716,
"grad_norm": 1.932795524597168,
"learning_rate": 1.836764313458962e-05,
"loss": 0.1116,
"step": 267
},
{
"epoch": 0.03062857142857143,
"grad_norm": 0.6493631601333618,
"learning_rate": 1.8348478632634067e-05,
"loss": 0.1263,
"step": 268
},
{
"epoch": 0.03074285714285714,
"grad_norm": 0.5922803282737732,
"learning_rate": 1.8329212407100996e-05,
"loss": 0.0708,
"step": 269
},
{
"epoch": 0.030857142857142857,
"grad_norm": 0.3190864324569702,
"learning_rate": 1.8309844692743283e-05,
"loss": 0.0596,
"step": 270
},
{
"epoch": 0.03097142857142857,
"grad_norm": 0.5801677107810974,
"learning_rate": 1.8290375725550417e-05,
"loss": 0.1261,
"step": 271
},
{
"epoch": 0.031085714285714286,
"grad_norm": 0.9508682489395142,
"learning_rate": 1.827080574274562e-05,
"loss": 0.129,
"step": 272
},
{
"epoch": 0.0312,
"grad_norm": 0.8393919467926025,
"learning_rate": 1.8251134982782952e-05,
"loss": 0.1328,
"step": 273
},
{
"epoch": 0.03131428571428571,
"grad_norm": 1.055537223815918,
"learning_rate": 1.8231363685344422e-05,
"loss": 0.1375,
"step": 274
},
{
"epoch": 0.03142857142857143,
"grad_norm": 0.8542097806930542,
"learning_rate": 1.821149209133704e-05,
"loss": 0.0916,
"step": 275
},
{
"epoch": 0.03154285714285714,
"grad_norm": 1.0771797895431519,
"learning_rate": 1.819152044288992e-05,
"loss": 0.0886,
"step": 276
},
{
"epoch": 0.031657142857142856,
"grad_norm": 0.5911391973495483,
"learning_rate": 1.8171448983351284e-05,
"loss": 0.1381,
"step": 277
},
{
"epoch": 0.03177142857142857,
"grad_norm": 0.6921252608299255,
"learning_rate": 1.815127795728554e-05,
"loss": 0.1326,
"step": 278
},
{
"epoch": 0.03188571428571429,
"grad_norm": 0.41178473830223083,
"learning_rate": 1.8131007610470278e-05,
"loss": 0.0703,
"step": 279
},
{
"epoch": 0.032,
"grad_norm": 0.3106781840324402,
"learning_rate": 1.8110638189893267e-05,
"loss": 0.0634,
"step": 280
},
{
"epoch": 0.03211428571428571,
"grad_norm": 1.0739474296569824,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.0795,
"step": 281
},
{
"epoch": 0.032228571428571426,
"grad_norm": 0.5694898366928101,
"learning_rate": 1.806960312143802e-05,
"loss": 0.0739,
"step": 282
},
{
"epoch": 0.032342857142857145,
"grad_norm": 0.9528548717498779,
"learning_rate": 1.804893797355914e-05,
"loss": 0.1198,
"step": 283
},
{
"epoch": 0.03245714285714286,
"grad_norm": 1.775436520576477,
"learning_rate": 1.8028174751911147e-05,
"loss": 0.181,
"step": 284
},
{
"epoch": 0.03257142857142857,
"grad_norm": 4.523551940917969,
"learning_rate": 1.8007313709487334e-05,
"loss": 0.1768,
"step": 285
},
{
"epoch": 0.03268571428571428,
"grad_norm": 0.6059966683387756,
"learning_rate": 1.798635510047293e-05,
"loss": 0.1362,
"step": 286
},
{
"epoch": 0.0328,
"grad_norm": 0.9247148633003235,
"learning_rate": 1.7965299180241963e-05,
"loss": 0.1531,
"step": 287
},
{
"epoch": 0.032914285714285715,
"grad_norm": 0.6787970066070557,
"learning_rate": 1.7944146205354182e-05,
"loss": 0.1138,
"step": 288
},
{
"epoch": 0.03302857142857143,
"grad_norm": 0.4064248502254486,
"learning_rate": 1.792289643355191e-05,
"loss": 0.0641,
"step": 289
},
{
"epoch": 0.03314285714285714,
"grad_norm": 0.1638564169406891,
"learning_rate": 1.7901550123756906e-05,
"loss": 0.0056,
"step": 290
},
{
"epoch": 0.03325714285714286,
"grad_norm": 0.46227213740348816,
"learning_rate": 1.788010753606722e-05,
"loss": 0.1205,
"step": 291
},
{
"epoch": 0.03337142857142857,
"grad_norm": 0.06335662305355072,
"learning_rate": 1.785856893175402e-05,
"loss": 0.0005,
"step": 292
},
{
"epoch": 0.033485714285714285,
"grad_norm": 1.006857991218567,
"learning_rate": 1.78369345732584e-05,
"loss": 0.1594,
"step": 293
},
{
"epoch": 0.0336,
"grad_norm": 0.5341035723686218,
"learning_rate": 1.781520472418819e-05,
"loss": 0.1256,
"step": 294
},
{
"epoch": 0.03371428571428572,
"grad_norm": 0.09702645242214203,
"learning_rate": 1.7793379649314743e-05,
"loss": 0.0019,
"step": 295
},
{
"epoch": 0.03382857142857143,
"grad_norm": 1.225579023361206,
"learning_rate": 1.777145961456971e-05,
"loss": 0.0729,
"step": 296
},
{
"epoch": 0.03394285714285714,
"grad_norm": 0.8058451414108276,
"learning_rate": 1.7749444887041797e-05,
"loss": 0.068,
"step": 297
},
{
"epoch": 0.034057142857142855,
"grad_norm": 0.5637487769126892,
"learning_rate": 1.7727335734973512e-05,
"loss": 0.064,
"step": 298
},
{
"epoch": 0.034171428571428575,
"grad_norm": 0.5069957375526428,
"learning_rate": 1.7705132427757895e-05,
"loss": 0.0769,
"step": 299
},
{
"epoch": 0.03428571428571429,
"grad_norm": 6.131114959716797,
"learning_rate": 1.7682835235935236e-05,
"loss": 0.2051,
"step": 300
},
{
"epoch": 0.0344,
"grad_norm": 0.5297291278839111,
"learning_rate": 1.766044443118978e-05,
"loss": 0.1263,
"step": 301
},
{
"epoch": 0.03451428571428571,
"grad_norm": 1.2629313468933105,
"learning_rate": 1.7637960286346423e-05,
"loss": 0.1398,
"step": 302
},
{
"epoch": 0.034628571428571425,
"grad_norm": 0.34277448058128357,
"learning_rate": 1.761538307536737e-05,
"loss": 0.0687,
"step": 303
},
{
"epoch": 0.034742857142857145,
"grad_norm": 0.672459602355957,
"learning_rate": 1.759271307334881e-05,
"loss": 0.0873,
"step": 304
},
{
"epoch": 0.03485714285714286,
"grad_norm": 0.4533510208129883,
"learning_rate": 1.7569950556517566e-05,
"loss": 0.0615,
"step": 305
},
{
"epoch": 0.03497142857142857,
"grad_norm": 0.4920494258403778,
"learning_rate": 1.7547095802227723e-05,
"loss": 0.0669,
"step": 306
},
{
"epoch": 0.03508571428571428,
"grad_norm": 0.5973871350288391,
"learning_rate": 1.7524149088957244e-05,
"loss": 0.1241,
"step": 307
},
{
"epoch": 0.0352,
"grad_norm": 0.740808367729187,
"learning_rate": 1.7501110696304598e-05,
"loss": 0.0726,
"step": 308
},
{
"epoch": 0.035314285714285715,
"grad_norm": 0.9830449223518372,
"learning_rate": 1.747798090498532e-05,
"loss": 0.1402,
"step": 309
},
{
"epoch": 0.03542857142857143,
"grad_norm": 0.22492338716983795,
"learning_rate": 1.7454759996828622e-05,
"loss": 0.0044,
"step": 310
},
{
"epoch": 0.03554285714285714,
"grad_norm": 0.050361260771751404,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.0011,
"step": 311
},
{
"epoch": 0.03565714285714286,
"grad_norm": 0.5040039420127869,
"learning_rate": 1.74080459628675e-05,
"loss": 0.0712,
"step": 312
},
{
"epoch": 0.03577142857142857,
"grad_norm": 0.3765220642089844,
"learning_rate": 1.7384553406258842e-05,
"loss": 0.0734,
"step": 313
},
{
"epoch": 0.035885714285714285,
"grad_norm": 0.7768454551696777,
"learning_rate": 1.7360970871197347e-05,
"loss": 0.1346,
"step": 314
},
{
"epoch": 0.036,
"grad_norm": 0.8529449701309204,
"learning_rate": 1.7337298645028764e-05,
"loss": 0.0697,
"step": 315
},
{
"epoch": 0.03611428571428572,
"grad_norm": 0.5148250460624695,
"learning_rate": 1.7313537016191706e-05,
"loss": 0.0529,
"step": 316
},
{
"epoch": 0.03622857142857143,
"grad_norm": 1.149399995803833,
"learning_rate": 1.7289686274214116e-05,
"loss": 0.0189,
"step": 317
},
{
"epoch": 0.03634285714285714,
"grad_norm": 0.8121666312217712,
"learning_rate": 1.7265746709709762e-05,
"loss": 0.1475,
"step": 318
},
{
"epoch": 0.036457142857142855,
"grad_norm": 0.7753887176513672,
"learning_rate": 1.7241718614374678e-05,
"loss": 0.1372,
"step": 319
},
{
"epoch": 0.036571428571428574,
"grad_norm": 0.4625493288040161,
"learning_rate": 1.7217602280983622e-05,
"loss": 0.0624,
"step": 320
},
{
"epoch": 0.03668571428571429,
"grad_norm": 0.356139600276947,
"learning_rate": 1.7193398003386514e-05,
"loss": 0.0661,
"step": 321
},
{
"epoch": 0.0368,
"grad_norm": 0.4844002425670624,
"learning_rate": 1.716910607650483e-05,
"loss": 0.1136,
"step": 322
},
{
"epoch": 0.03691428571428571,
"grad_norm": 0.5232403874397278,
"learning_rate": 1.7144726796328034e-05,
"loss": 0.129,
"step": 323
},
{
"epoch": 0.03702857142857143,
"grad_norm": 0.3515738248825073,
"learning_rate": 1.712026045990997e-05,
"loss": 0.0517,
"step": 324
},
{
"epoch": 0.037142857142857144,
"grad_norm": 0.640648365020752,
"learning_rate": 1.709570736536521e-05,
"loss": 0.1244,
"step": 325
},
{
"epoch": 0.03725714285714286,
"grad_norm": 1.0593113899230957,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.1342,
"step": 326
},
{
"epoch": 0.03737142857142857,
"grad_norm": 0.7747220993041992,
"learning_rate": 1.7046342099635948e-05,
"loss": 0.0676,
"step": 327
},
{
"epoch": 0.03748571428571429,
"grad_norm": 0.5960583090782166,
"learning_rate": 1.7021530529951627e-05,
"loss": 0.0646,
"step": 328
},
{
"epoch": 0.0376,
"grad_norm": 0.2865906357765198,
"learning_rate": 1.6996633405133656e-05,
"loss": 0.0596,
"step": 329
},
{
"epoch": 0.037714285714285714,
"grad_norm": 0.9084271192550659,
"learning_rate": 1.697165102854565e-05,
"loss": 0.1157,
"step": 330
},
{
"epoch": 0.03782857142857143,
"grad_norm": 0.7081615328788757,
"learning_rate": 1.6946583704589973e-05,
"loss": 0.1042,
"step": 331
},
{
"epoch": 0.037942857142857146,
"grad_norm": 0.647879958152771,
"learning_rate": 1.692143173870407e-05,
"loss": 0.0697,
"step": 332
},
{
"epoch": 0.03805714285714286,
"grad_norm": 0.33049827814102173,
"learning_rate": 1.68961954373567e-05,
"loss": 0.0635,
"step": 333
},
{
"epoch": 0.03817142857142857,
"grad_norm": 1.4319345951080322,
"learning_rate": 1.6870875108044233e-05,
"loss": 0.1559,
"step": 334
},
{
"epoch": 0.038285714285714284,
"grad_norm": 0.3779307007789612,
"learning_rate": 1.684547105928689e-05,
"loss": 0.0695,
"step": 335
},
{
"epoch": 0.0384,
"grad_norm": 0.4206676185131073,
"learning_rate": 1.6819983600624986e-05,
"loss": 0.0591,
"step": 336
},
{
"epoch": 0.038514285714285716,
"grad_norm": 0.5171385407447815,
"learning_rate": 1.6794413042615168e-05,
"loss": 0.0612,
"step": 337
},
{
"epoch": 0.03862857142857143,
"grad_norm": 1.1469932794570923,
"learning_rate": 1.6768759696826608e-05,
"loss": 0.1208,
"step": 338
},
{
"epoch": 0.03874285714285714,
"grad_norm": 1.4118863344192505,
"learning_rate": 1.6743023875837233e-05,
"loss": 0.0847,
"step": 339
},
{
"epoch": 0.038857142857142854,
"grad_norm": 0.6712648272514343,
"learning_rate": 1.6717205893229904e-05,
"loss": 0.0875,
"step": 340
},
{
"epoch": 0.038971428571428574,
"grad_norm": 0.39986902475357056,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.052,
"step": 341
},
{
"epoch": 0.039085714285714286,
"grad_norm": 1.1605353355407715,
"learning_rate": 1.6665324702494524e-05,
"loss": 0.14,
"step": 342
},
{
"epoch": 0.0392,
"grad_norm": 1.3340308666229248,
"learning_rate": 1.6639262126522417e-05,
"loss": 0.1217,
"step": 343
},
{
"epoch": 0.03931428571428571,
"grad_norm": 0.6971845030784607,
"learning_rate": 1.661311865323652e-05,
"loss": 0.1095,
"step": 344
},
{
"epoch": 0.03942857142857143,
"grad_norm": 0.9757205247879028,
"learning_rate": 1.6586894601186804e-05,
"loss": 0.0693,
"step": 345
},
{
"epoch": 0.039542857142857143,
"grad_norm": 0.97724449634552,
"learning_rate": 1.6560590289905074e-05,
"loss": 0.1325,
"step": 346
},
{
"epoch": 0.039657142857142856,
"grad_norm": 0.7021217346191406,
"learning_rate": 1.6534206039901057e-05,
"loss": 0.1265,
"step": 347
},
{
"epoch": 0.03977142857142857,
"grad_norm": 0.7403573989868164,
"learning_rate": 1.650774217265851e-05,
"loss": 0.1144,
"step": 348
},
{
"epoch": 0.03988571428571429,
"grad_norm": 10.570477485656738,
"learning_rate": 1.6481199010631312e-05,
"loss": 0.4382,
"step": 349
},
{
"epoch": 0.04,
"grad_norm": 0.7036963105201721,
"learning_rate": 1.645457687723951e-05,
"loss": 0.1173,
"step": 350
},
{
"epoch": 0.04011428571428571,
"grad_norm": 0.5365110039710999,
"learning_rate": 1.6427876096865394e-05,
"loss": 0.1091,
"step": 351
},
{
"epoch": 0.040228571428571426,
"grad_norm": 1.5085561275482178,
"learning_rate": 1.6401096994849558e-05,
"loss": 0.0837,
"step": 352
},
{
"epoch": 0.040342857142857146,
"grad_norm": 0.6810627579689026,
"learning_rate": 1.63742398974869e-05,
"loss": 0.1134,
"step": 353
},
{
"epoch": 0.04045714285714286,
"grad_norm": 2.38642954826355,
"learning_rate": 1.6347305132022677e-05,
"loss": 0.072,
"step": 354
},
{
"epoch": 0.04057142857142857,
"grad_norm": 0.24412786960601807,
"learning_rate": 1.632029302664851e-05,
"loss": 0.0035,
"step": 355
},
{
"epoch": 0.04068571428571428,
"grad_norm": 0.794260561466217,
"learning_rate": 1.6293203910498375e-05,
"loss": 0.092,
"step": 356
},
{
"epoch": 0.0408,
"grad_norm": 0.48356446623802185,
"learning_rate": 1.6266038113644605e-05,
"loss": 0.0631,
"step": 357
},
{
"epoch": 0.040914285714285716,
"grad_norm": 1.528218150138855,
"learning_rate": 1.6238795967093865e-05,
"loss": 0.12,
"step": 358
},
{
"epoch": 0.04102857142857143,
"grad_norm": 0.5906217098236084,
"learning_rate": 1.6211477802783105e-05,
"loss": 0.0841,
"step": 359
},
{
"epoch": 0.04114285714285714,
"grad_norm": 0.8047304153442383,
"learning_rate": 1.6184083953575543e-05,
"loss": 0.1239,
"step": 360
},
{
"epoch": 0.04125714285714286,
"grad_norm": 0.8435593843460083,
"learning_rate": 1.6156614753256583e-05,
"loss": 0.1249,
"step": 361
},
{
"epoch": 0.04137142857142857,
"grad_norm": 0.9788255095481873,
"learning_rate": 1.6129070536529767e-05,
"loss": 0.0738,
"step": 362
},
{
"epoch": 0.041485714285714285,
"grad_norm": 2.155522108078003,
"learning_rate": 1.610145163901268e-05,
"loss": 0.2152,
"step": 363
},
{
"epoch": 0.0416,
"grad_norm": 0.03900998830795288,
"learning_rate": 1.607375839723287e-05,
"loss": 0.0006,
"step": 364
},
{
"epoch": 0.04171428571428572,
"grad_norm": 1.259210467338562,
"learning_rate": 1.6045991148623752e-05,
"loss": 0.1226,
"step": 365
},
{
"epoch": 0.04182857142857143,
"grad_norm": 1.0150049924850464,
"learning_rate": 1.6018150231520486e-05,
"loss": 0.1433,
"step": 366
},
{
"epoch": 0.04194285714285714,
"grad_norm": 0.4047106206417084,
"learning_rate": 1.599023598515586e-05,
"loss": 0.0554,
"step": 367
},
{
"epoch": 0.042057142857142855,
"grad_norm": 0.9958846569061279,
"learning_rate": 1.5962248749656158e-05,
"loss": 0.1021,
"step": 368
},
{
"epoch": 0.04217142857142857,
"grad_norm": 2.0389113426208496,
"learning_rate": 1.5934188866037017e-05,
"loss": 0.1142,
"step": 369
},
{
"epoch": 0.04228571428571429,
"grad_norm": 1.4610713720321655,
"learning_rate": 1.5906056676199256e-05,
"loss": 0.1299,
"step": 370
},
{
"epoch": 0.0424,
"grad_norm": 0.4549573063850403,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.0704,
"step": 371
},
{
"epoch": 0.04251428571428571,
"grad_norm": 1.9539470672607422,
"learning_rate": 1.584957674987216e-05,
"loss": 0.19,
"step": 372
},
{
"epoch": 0.042628571428571425,
"grad_norm": 6.028285026550293,
"learning_rate": 1.5821229701572897e-05,
"loss": 0.1273,
"step": 373
},
{
"epoch": 0.042742857142857145,
"grad_norm": 0.15216311812400818,
"learning_rate": 1.5792811723426787e-05,
"loss": 0.0036,
"step": 374
},
{
"epoch": 0.04285714285714286,
"grad_norm": 0.821194589138031,
"learning_rate": 1.5764323161697933e-05,
"loss": 0.0638,
"step": 375
},
{
"epoch": 0.04297142857142857,
"grad_norm": 0.6445154547691345,
"learning_rate": 1.573576436351046e-05,
"loss": 0.0678,
"step": 376
},
{
"epoch": 0.04308571428571428,
"grad_norm": 0.7004276514053345,
"learning_rate": 1.570713567684432e-05,
"loss": 0.0744,
"step": 377
},
{
"epoch": 0.0432,
"grad_norm": 0.9961415529251099,
"learning_rate": 1.5678437450531014e-05,
"loss": 0.1367,
"step": 378
},
{
"epoch": 0.043314285714285715,
"grad_norm": 0.32565218210220337,
"learning_rate": 1.564967003424938e-05,
"loss": 0.0074,
"step": 379
},
{
"epoch": 0.04342857142857143,
"grad_norm": 5.462352752685547,
"learning_rate": 1.5620833778521306e-05,
"loss": 0.168,
"step": 380
},
{
"epoch": 0.04354285714285714,
"grad_norm": 16.4879093170166,
"learning_rate": 1.5591929034707468e-05,
"loss": 0.5451,
"step": 381
},
{
"epoch": 0.04365714285714286,
"grad_norm": 0.7500917911529541,
"learning_rate": 1.556295615500305e-05,
"loss": 0.1242,
"step": 382
},
{
"epoch": 0.04377142857142857,
"grad_norm": 0.2675608992576599,
"learning_rate": 1.553391549243344e-05,
"loss": 0.0547,
"step": 383
},
{
"epoch": 0.043885714285714285,
"grad_norm": 0.5758733749389648,
"learning_rate": 1.5504807400849957e-05,
"loss": 0.1173,
"step": 384
},
{
"epoch": 0.044,
"grad_norm": 1.0356460809707642,
"learning_rate": 1.5475632234925505e-05,
"loss": 0.1523,
"step": 385
},
{
"epoch": 0.04411428571428572,
"grad_norm": 1.3973541259765625,
"learning_rate": 1.5446390350150272e-05,
"loss": 0.1352,
"step": 386
},
{
"epoch": 0.04422857142857143,
"grad_norm": 3.244534730911255,
"learning_rate": 1.54170821028274e-05,
"loss": 0.168,
"step": 387
},
{
"epoch": 0.04434285714285714,
"grad_norm": 1.3915773630142212,
"learning_rate": 1.5387707850068633e-05,
"loss": 0.1404,
"step": 388
},
{
"epoch": 0.044457142857142855,
"grad_norm": 0.038858383893966675,
"learning_rate": 1.5358267949789968e-05,
"loss": 0.0009,
"step": 389
},
{
"epoch": 0.044571428571428574,
"grad_norm": 0.768882691860199,
"learning_rate": 1.53287627607073e-05,
"loss": 0.1294,
"step": 390
},
{
"epoch": 0.04468571428571429,
"grad_norm": 0.5438225865364075,
"learning_rate": 1.529919264233205e-05,
"loss": 0.1131,
"step": 391
},
{
"epoch": 0.0448,
"grad_norm": 0.30528900027275085,
"learning_rate": 1.5269557954966777e-05,
"loss": 0.0524,
"step": 392
},
{
"epoch": 0.04491428571428571,
"grad_norm": 0.5019018054008484,
"learning_rate": 1.5239859059700794e-05,
"loss": 0.0633,
"step": 393
},
{
"epoch": 0.04502857142857143,
"grad_norm": 0.5741058588027954,
"learning_rate": 1.5210096318405768e-05,
"loss": 0.1134,
"step": 394
},
{
"epoch": 0.045142857142857144,
"grad_norm": 21.18476676940918,
"learning_rate": 1.5180270093731305e-05,
"loss": 0.6116,
"step": 395
},
{
"epoch": 0.04525714285714286,
"grad_norm": 0.9793661832809448,
"learning_rate": 1.5150380749100545e-05,
"loss": 0.134,
"step": 396
},
{
"epoch": 0.04537142857142857,
"grad_norm": 1.0946067571640015,
"learning_rate": 1.5120428648705716e-05,
"loss": 0.1237,
"step": 397
},
{
"epoch": 0.04548571428571429,
"grad_norm": 0.3047063648700714,
"learning_rate": 1.5090414157503715e-05,
"loss": 0.0037,
"step": 398
},
{
"epoch": 0.0456,
"grad_norm": 0.9988011717796326,
"learning_rate": 1.5060337641211637e-05,
"loss": 0.1216,
"step": 399
},
{
"epoch": 0.045714285714285714,
"grad_norm": 0.6280392408370972,
"learning_rate": 1.5030199466302354e-05,
"loss": 0.0652,
"step": 400
},
{
"epoch": 0.04582857142857143,
"grad_norm": 2.9914512634277344,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.2253,
"step": 401
},
{
"epoch": 0.04594285714285714,
"grad_norm": 0.5957767963409424,
"learning_rate": 1.4969739610275556e-05,
"loss": 0.0603,
"step": 402
},
{
"epoch": 0.04605714285714286,
"grad_norm": 0.5896881818771362,
"learning_rate": 1.493941866584231e-05,
"loss": 0.1138,
"step": 403
},
{
"epoch": 0.04617142857142857,
"grad_norm": 0.329389750957489,
"learning_rate": 1.490903753615141e-05,
"loss": 0.0559,
"step": 404
},
{
"epoch": 0.046285714285714284,
"grad_norm": 1.2978957891464233,
"learning_rate": 1.4878596591387329e-05,
"loss": 0.1582,
"step": 405
},
{
"epoch": 0.0464,
"grad_norm": 2.043853282928467,
"learning_rate": 1.4848096202463373e-05,
"loss": 0.0884,
"step": 406
},
{
"epoch": 0.046514285714285716,
"grad_norm": 0.4522497355937958,
"learning_rate": 1.4817536741017153e-05,
"loss": 0.1062,
"step": 407
},
{
"epoch": 0.04662857142857143,
"grad_norm": 0.07202062010765076,
"learning_rate": 1.478691857940607e-05,
"loss": 0.001,
"step": 408
},
{
"epoch": 0.04674285714285714,
"grad_norm": 1.2533372640609741,
"learning_rate": 1.4756242090702756e-05,
"loss": 0.1259,
"step": 409
},
{
"epoch": 0.046857142857142854,
"grad_norm": 0.35696080327033997,
"learning_rate": 1.4725507648690542e-05,
"loss": 0.0531,
"step": 410
},
{
"epoch": 0.046971428571428574,
"grad_norm": 3.0528464317321777,
"learning_rate": 1.469471562785891e-05,
"loss": 0.1761,
"step": 411
},
{
"epoch": 0.047085714285714286,
"grad_norm": 0.3574976623058319,
"learning_rate": 1.4663866403398915e-05,
"loss": 0.0536,
"step": 412
},
{
"epoch": 0.0472,
"grad_norm": 0.7249162793159485,
"learning_rate": 1.463296035119862e-05,
"loss": 0.1158,
"step": 413
},
{
"epoch": 0.04731428571428571,
"grad_norm": 0.611409604549408,
"learning_rate": 1.4601997847838518e-05,
"loss": 0.1153,
"step": 414
},
{
"epoch": 0.04742857142857143,
"grad_norm": 0.8269656300544739,
"learning_rate": 1.4570979270586944e-05,
"loss": 0.0613,
"step": 415
},
{
"epoch": 0.047542857142857144,
"grad_norm": 0.32490140199661255,
"learning_rate": 1.4539904997395468e-05,
"loss": 0.055,
"step": 416
},
{
"epoch": 0.047657142857142856,
"grad_norm": 0.6147916913032532,
"learning_rate": 1.4508775406894308e-05,
"loss": 0.1244,
"step": 417
},
{
"epoch": 0.04777142857142857,
"grad_norm": 1.3269389867782593,
"learning_rate": 1.4477590878387697e-05,
"loss": 0.1273,
"step": 418
},
{
"epoch": 0.04788571428571429,
"grad_norm": 1.2243558168411255,
"learning_rate": 1.4446351791849276e-05,
"loss": 0.1015,
"step": 419
},
{
"epoch": 0.048,
"grad_norm": 0.3902854323387146,
"learning_rate": 1.4415058527917454e-05,
"loss": 0.0517,
"step": 420
},
{
"epoch": 0.048114285714285714,
"grad_norm": 1.4584861993789673,
"learning_rate": 1.4383711467890776e-05,
"loss": 0.1534,
"step": 421
},
{
"epoch": 0.048228571428571426,
"grad_norm": 0.6895879507064819,
"learning_rate": 1.4352310993723277e-05,
"loss": 0.1108,
"step": 422
},
{
"epoch": 0.048342857142857146,
"grad_norm": 0.24715161323547363,
"learning_rate": 1.4320857488019826e-05,
"loss": 0.0498,
"step": 423
},
{
"epoch": 0.04845714285714286,
"grad_norm": 1.589969277381897,
"learning_rate": 1.4289351334031461e-05,
"loss": 0.0922,
"step": 424
},
{
"epoch": 0.04857142857142857,
"grad_norm": 0.34705668687820435,
"learning_rate": 1.4257792915650728e-05,
"loss": 0.0583,
"step": 425
},
{
"epoch": 0.048685714285714284,
"grad_norm": 0.5411010384559631,
"learning_rate": 1.4226182617406996e-05,
"loss": 0.0643,
"step": 426
},
{
"epoch": 0.0488,
"grad_norm": 8.336661338806152,
"learning_rate": 1.4194520824461773e-05,
"loss": 0.2957,
"step": 427
},
{
"epoch": 0.048914285714285716,
"grad_norm": 1.1309565305709839,
"learning_rate": 1.4162807922604014e-05,
"loss": 0.0792,
"step": 428
},
{
"epoch": 0.04902857142857143,
"grad_norm": 0.954781174659729,
"learning_rate": 1.413104429824542e-05,
"loss": 0.1295,
"step": 429
},
{
"epoch": 0.04914285714285714,
"grad_norm": 1.5289859771728516,
"learning_rate": 1.4099230338415728e-05,
"loss": 0.1119,
"step": 430
},
{
"epoch": 0.04925714285714286,
"grad_norm": 0.39689069986343384,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.054,
"step": 431
},
{
"epoch": 0.04937142857142857,
"grad_norm": 0.6845630407333374,
"learning_rate": 1.4035452963523903e-05,
"loss": 0.061,
"step": 432
},
{
"epoch": 0.049485714285714286,
"grad_norm": 8.508838653564453,
"learning_rate": 1.4003490325568953e-05,
"loss": 0.2188,
"step": 433
},
{
"epoch": 0.0496,
"grad_norm": 0.577778697013855,
"learning_rate": 1.3971478906347806e-05,
"loss": 0.1214,
"step": 434
},
{
"epoch": 0.04971428571428571,
"grad_norm": 0.6180987358093262,
"learning_rate": 1.3939419095909513e-05,
"loss": 0.0591,
"step": 435
},
{
"epoch": 0.04982857142857143,
"grad_norm": 0.6514303088188171,
"learning_rate": 1.3907311284892737e-05,
"loss": 0.059,
"step": 436
},
{
"epoch": 0.04994285714285714,
"grad_norm": 1.722687005996704,
"learning_rate": 1.3875155864521031e-05,
"loss": 0.1435,
"step": 437
},
{
"epoch": 0.050057142857142856,
"grad_norm": 0.6083362102508545,
"learning_rate": 1.3842953226598036e-05,
"loss": 0.1318,
"step": 438
},
{
"epoch": 0.05017142857142857,
"grad_norm": 1.0799511671066284,
"learning_rate": 1.3810703763502744e-05,
"loss": 0.1368,
"step": 439
},
{
"epoch": 0.05028571428571429,
"grad_norm": 1.0777685642242432,
"learning_rate": 1.3778407868184674e-05,
"loss": 0.0805,
"step": 440
},
{
"epoch": 0.0504,
"grad_norm": 1.1632506847381592,
"learning_rate": 1.3746065934159123e-05,
"loss": 0.0796,
"step": 441
},
{
"epoch": 0.05051428571428571,
"grad_norm": 0.8539578318595886,
"learning_rate": 1.371367835550235e-05,
"loss": 0.0694,
"step": 442
},
{
"epoch": 0.050628571428571426,
"grad_norm": 0.7852999567985535,
"learning_rate": 1.3681245526846782e-05,
"loss": 0.0833,
"step": 443
},
{
"epoch": 0.050742857142857145,
"grad_norm": 0.9067091941833496,
"learning_rate": 1.3648767843376196e-05,
"loss": 0.0721,
"step": 444
},
{
"epoch": 0.05085714285714286,
"grad_norm": 0.223627969622612,
"learning_rate": 1.3616245700820922e-05,
"loss": 0.0562,
"step": 445
},
{
"epoch": 0.05097142857142857,
"grad_norm": 0.05723898112773895,
"learning_rate": 1.3583679495453e-05,
"loss": 0.0011,
"step": 446
},
{
"epoch": 0.05108571428571428,
"grad_norm": 0.66180020570755,
"learning_rate": 1.3551069624081372e-05,
"loss": 0.1215,
"step": 447
},
{
"epoch": 0.0512,
"grad_norm": 0.6566020846366882,
"learning_rate": 1.3518416484047018e-05,
"loss": 0.0664,
"step": 448
},
{
"epoch": 0.051314285714285715,
"grad_norm": 1.2952485084533691,
"learning_rate": 1.3485720473218153e-05,
"loss": 0.0776,
"step": 449
},
{
"epoch": 0.05142857142857143,
"grad_norm": 0.811284065246582,
"learning_rate": 1.3452981989985347e-05,
"loss": 0.0683,
"step": 450
},
{
"epoch": 0.05154285714285714,
"grad_norm": 0.520163893699646,
"learning_rate": 1.342020143325669e-05,
"loss": 0.1103,
"step": 451
},
{
"epoch": 0.05165714285714286,
"grad_norm": 0.7408766746520996,
"learning_rate": 1.3387379202452917e-05,
"loss": 0.0706,
"step": 452
},
{
"epoch": 0.05177142857142857,
"grad_norm": 0.9585939049720764,
"learning_rate": 1.3354515697502552e-05,
"loss": 0.1165,
"step": 453
},
{
"epoch": 0.051885714285714285,
"grad_norm": 0.6970547437667847,
"learning_rate": 1.3321611318837033e-05,
"loss": 0.1181,
"step": 454
},
{
"epoch": 0.052,
"grad_norm": 2.3486571311950684,
"learning_rate": 1.3288666467385834e-05,
"loss": 0.1875,
"step": 455
},
{
"epoch": 0.05211428571428572,
"grad_norm": 0.023406101390719414,
"learning_rate": 1.3255681544571568e-05,
"loss": 0.001,
"step": 456
},
{
"epoch": 0.05222857142857143,
"grad_norm": 0.02360491082072258,
"learning_rate": 1.3222656952305113e-05,
"loss": 0.0016,
"step": 457
},
{
"epoch": 0.05234285714285714,
"grad_norm": 0.7878499031066895,
"learning_rate": 1.3189593092980701e-05,
"loss": 0.1372,
"step": 458
},
{
"epoch": 0.052457142857142855,
"grad_norm": 1.05193030834198,
"learning_rate": 1.3156490369471026e-05,
"loss": 0.084,
"step": 459
},
{
"epoch": 0.052571428571428575,
"grad_norm": 0.23702484369277954,
"learning_rate": 1.3123349185122328e-05,
"loss": 0.0041,
"step": 460
},
{
"epoch": 0.05268571428571429,
"grad_norm": 0.5299646258354187,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.0785,
"step": 461
},
{
"epoch": 0.0528,
"grad_norm": 1.4201158285140991,
"learning_rate": 1.3056953049631059e-05,
"loss": 0.1354,
"step": 462
},
{
"epoch": 0.05291428571428571,
"grad_norm": 1.6005371809005737,
"learning_rate": 1.3023698907504447e-05,
"loss": 0.1324,
"step": 463
},
{
"epoch": 0.05302857142857143,
"grad_norm": 1.0204145908355713,
"learning_rate": 1.2990407922560869e-05,
"loss": 0.1147,
"step": 464
},
{
"epoch": 0.053142857142857144,
"grad_norm": 0.8085700273513794,
"learning_rate": 1.2957080500440469e-05,
"loss": 0.1189,
"step": 465
},
{
"epoch": 0.05325714285714286,
"grad_norm": 1.5970574617385864,
"learning_rate": 1.2923717047227368e-05,
"loss": 0.1279,
"step": 466
},
{
"epoch": 0.05337142857142857,
"grad_norm": 0.3705628514289856,
"learning_rate": 1.2890317969444716e-05,
"loss": 0.0629,
"step": 467
},
{
"epoch": 0.05348571428571428,
"grad_norm": 0.12271714955568314,
"learning_rate": 1.2856883674049736e-05,
"loss": 0.0032,
"step": 468
},
{
"epoch": 0.0536,
"grad_norm": 0.5374379754066467,
"learning_rate": 1.2823414568428767e-05,
"loss": 0.0716,
"step": 469
},
{
"epoch": 0.053714285714285714,
"grad_norm": 1.0720727443695068,
"learning_rate": 1.2789911060392295e-05,
"loss": 0.1177,
"step": 470
},
{
"epoch": 0.05382857142857143,
"grad_norm": 0.8524672389030457,
"learning_rate": 1.2756373558169992e-05,
"loss": 0.0714,
"step": 471
},
{
"epoch": 0.05394285714285714,
"grad_norm": 0.5075511336326599,
"learning_rate": 1.2722802470405744e-05,
"loss": 0.0646,
"step": 472
},
{
"epoch": 0.05405714285714286,
"grad_norm": 0.4245651066303253,
"learning_rate": 1.2689198206152657e-05,
"loss": 0.108,
"step": 473
},
{
"epoch": 0.05417142857142857,
"grad_norm": 0.4011399447917938,
"learning_rate": 1.265556117486809e-05,
"loss": 0.0543,
"step": 474
},
{
"epoch": 0.054285714285714284,
"grad_norm": 0.8801376819610596,
"learning_rate": 1.2621891786408648e-05,
"loss": 0.1,
"step": 475
},
{
"epoch": 0.0544,
"grad_norm": 0.6451431512832642,
"learning_rate": 1.2588190451025209e-05,
"loss": 0.1171,
"step": 476
},
{
"epoch": 0.054514285714285717,
"grad_norm": 0.4531909227371216,
"learning_rate": 1.2554457579357906e-05,
"loss": 0.1094,
"step": 477
},
{
"epoch": 0.05462857142857143,
"grad_norm": 0.40501904487609863,
"learning_rate": 1.252069358243114e-05,
"loss": 0.0621,
"step": 478
},
{
"epoch": 0.05474285714285714,
"grad_norm": 0.5643985271453857,
"learning_rate": 1.2486898871648552e-05,
"loss": 0.1248,
"step": 479
},
{
"epoch": 0.054857142857142854,
"grad_norm": 0.6199919581413269,
"learning_rate": 1.2453073858788027e-05,
"loss": 0.1257,
"step": 480
},
{
"epoch": 0.054971428571428574,
"grad_norm": 1.3144861459732056,
"learning_rate": 1.2419218955996677e-05,
"loss": 0.0801,
"step": 481
},
{
"epoch": 0.055085714285714286,
"grad_norm": 0.5956525802612305,
"learning_rate": 1.238533457578581e-05,
"loss": 0.1146,
"step": 482
},
{
"epoch": 0.0552,
"grad_norm": 0.7082200646400452,
"learning_rate": 1.23514211310259e-05,
"loss": 0.0145,
"step": 483
},
{
"epoch": 0.05531428571428571,
"grad_norm": 0.9622437953948975,
"learning_rate": 1.2317479034941572e-05,
"loss": 0.1163,
"step": 484
},
{
"epoch": 0.05542857142857143,
"grad_norm": 1.4974924325942993,
"learning_rate": 1.2283508701106559e-05,
"loss": 0.1291,
"step": 485
},
{
"epoch": 0.055542857142857144,
"grad_norm": 0.929922342300415,
"learning_rate": 1.2249510543438652e-05,
"loss": 0.1288,
"step": 486
},
{
"epoch": 0.055657142857142856,
"grad_norm": 0.030353449285030365,
"learning_rate": 1.2215484976194675e-05,
"loss": 0.0007,
"step": 487
},
{
"epoch": 0.05577142857142857,
"grad_norm": 0.7135401368141174,
"learning_rate": 1.2181432413965428e-05,
"loss": 0.073,
"step": 488
},
{
"epoch": 0.05588571428571429,
"grad_norm": 0.8080915808677673,
"learning_rate": 1.2147353271670634e-05,
"loss": 0.12,
"step": 489
},
{
"epoch": 0.056,
"grad_norm": 0.4979619085788727,
"learning_rate": 1.211324796455389e-05,
"loss": 0.0691,
"step": 490
},
{
"epoch": 0.056114285714285714,
"grad_norm": 0.3718797564506531,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.0558,
"step": 491
},
{
"epoch": 0.056228571428571426,
"grad_norm": 1.747310757637024,
"learning_rate": 1.2044960518417902e-05,
"loss": 0.1442,
"step": 492
},
{
"epoch": 0.056342857142857146,
"grad_norm": 0.9576373100280762,
"learning_rate": 1.2010779211459649e-05,
"loss": 0.0612,
"step": 493
},
{
"epoch": 0.05645714285714286,
"grad_norm": 0.5326647162437439,
"learning_rate": 1.1976573403791263e-05,
"loss": 0.0488,
"step": 494
},
{
"epoch": 0.05657142857142857,
"grad_norm": 0.7388985753059387,
"learning_rate": 1.194234351219972e-05,
"loss": 0.1112,
"step": 495
},
{
"epoch": 0.056685714285714284,
"grad_norm": 1.0324307680130005,
"learning_rate": 1.190808995376545e-05,
"loss": 0.1083,
"step": 496
},
{
"epoch": 0.0568,
"grad_norm": 0.9696743488311768,
"learning_rate": 1.187381314585725e-05,
"loss": 0.1152,
"step": 497
},
{
"epoch": 0.056914285714285716,
"grad_norm": 0.07941851019859314,
"learning_rate": 1.1839513506127202e-05,
"loss": 0.0031,
"step": 498
},
{
"epoch": 0.05702857142857143,
"grad_norm": 0.9655370116233826,
"learning_rate": 1.1805191452505602e-05,
"loss": 0.1085,
"step": 499
},
{
"epoch": 0.05714285714285714,
"grad_norm": 0.1506630927324295,
"learning_rate": 1.1770847403195836e-05,
"loss": 0.0025,
"step": 500
},
{
"epoch": 0.057257142857142854,
"grad_norm": 0.36272644996643066,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.0621,
"step": 501
},
{
"epoch": 0.05737142857142857,
"grad_norm": 0.967313826084137,
"learning_rate": 1.1702094991660326e-05,
"loss": 0.0665,
"step": 502
},
{
"epoch": 0.057485714285714286,
"grad_norm": 0.3336320221424103,
"learning_rate": 1.1667687467161025e-05,
"loss": 0.058,
"step": 503
},
{
"epoch": 0.0576,
"grad_norm": 0.5870653986930847,
"learning_rate": 1.1633259622416224e-05,
"loss": 0.1304,
"step": 504
},
{
"epoch": 0.05771428571428571,
"grad_norm": 0.09200169146060944,
"learning_rate": 1.159881187691835e-05,
"loss": 0.0016,
"step": 505
},
{
"epoch": 0.05782857142857143,
"grad_norm": 0.5145789980888367,
"learning_rate": 1.156434465040231e-05,
"loss": 0.0543,
"step": 506
},
{
"epoch": 0.05794285714285714,
"grad_norm": 1.5270731449127197,
"learning_rate": 1.1529858362840383e-05,
"loss": 0.0939,
"step": 507
},
{
"epoch": 0.058057142857142856,
"grad_norm": 0.8157461285591125,
"learning_rate": 1.1495353434437098e-05,
"loss": 0.0671,
"step": 508
},
{
"epoch": 0.05817142857142857,
"grad_norm": 0.707593560218811,
"learning_rate": 1.1460830285624119e-05,
"loss": 0.1134,
"step": 509
},
{
"epoch": 0.05828571428571429,
"grad_norm": 0.27421197295188904,
"learning_rate": 1.1426289337055119e-05,
"loss": 0.0621,
"step": 510
},
{
"epoch": 0.0584,
"grad_norm": 0.4556952714920044,
"learning_rate": 1.1391731009600655e-05,
"loss": 0.1033,
"step": 511
},
{
"epoch": 0.05851428571428571,
"grad_norm": 0.6568729281425476,
"learning_rate": 1.1357155724343046e-05,
"loss": 0.101,
"step": 512
},
{
"epoch": 0.058628571428571426,
"grad_norm": 0.6892651915550232,
"learning_rate": 1.1322563902571227e-05,
"loss": 0.0651,
"step": 513
},
{
"epoch": 0.058742857142857145,
"grad_norm": 0.45106813311576843,
"learning_rate": 1.128795596577563e-05,
"loss": 0.0619,
"step": 514
},
{
"epoch": 0.05885714285714286,
"grad_norm": 35.790855407714844,
"learning_rate": 1.1253332335643043e-05,
"loss": 0.5985,
"step": 515
},
{
"epoch": 0.05897142857142857,
"grad_norm": 0.3904721140861511,
"learning_rate": 1.1218693434051475e-05,
"loss": 0.0594,
"step": 516
},
{
"epoch": 0.05908571428571428,
"grad_norm": 0.9614586234092712,
"learning_rate": 1.1184039683065014e-05,
"loss": 0.1179,
"step": 517
},
{
"epoch": 0.0592,
"grad_norm": 0.8595171570777893,
"learning_rate": 1.1149371504928667e-05,
"loss": 0.0676,
"step": 518
},
{
"epoch": 0.059314285714285715,
"grad_norm": 0.02296481840312481,
"learning_rate": 1.1114689322063255e-05,
"loss": 0.0012,
"step": 519
},
{
"epoch": 0.05942857142857143,
"grad_norm": 0.8652318716049194,
"learning_rate": 1.1079993557060228e-05,
"loss": 0.1487,
"step": 520
},
{
"epoch": 0.05954285714285714,
"grad_norm": 1.0914485454559326,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.1246,
"step": 521
},
{
"epoch": 0.05965714285714286,
"grad_norm": 0.8204888105392456,
"learning_rate": 1.1010562971829464e-05,
"loss": 0.0573,
"step": 522
},
{
"epoch": 0.05977142857142857,
"grad_norm": 1.5582354068756104,
"learning_rate": 1.0975828997591496e-05,
"loss": 0.1432,
"step": 523
},
{
"epoch": 0.059885714285714285,
"grad_norm": 0.3577067255973816,
"learning_rate": 1.0941083133185146e-05,
"loss": 0.0534,
"step": 524
},
{
"epoch": 0.06,
"grad_norm": 0.6040599346160889,
"learning_rate": 1.0906325801977804e-05,
"loss": 0.1299,
"step": 525
},
{
"epoch": 0.06011428571428572,
"grad_norm": 0.3767750859260559,
"learning_rate": 1.0871557427476585e-05,
"loss": 0.0475,
"step": 526
},
{
"epoch": 0.06022857142857143,
"grad_norm": 0.4475708603858948,
"learning_rate": 1.083677843332316e-05,
"loss": 0.0573,
"step": 527
},
{
"epoch": 0.06034285714285714,
"grad_norm": 0.824868381023407,
"learning_rate": 1.0801989243288588e-05,
"loss": 0.1286,
"step": 528
},
{
"epoch": 0.060457142857142855,
"grad_norm": 0.28535786271095276,
"learning_rate": 1.0767190281268187e-05,
"loss": 0.0523,
"step": 529
},
{
"epoch": 0.060571428571428575,
"grad_norm": 0.3284713327884674,
"learning_rate": 1.0732381971276318e-05,
"loss": 0.0582,
"step": 530
},
{
"epoch": 0.06068571428571429,
"grad_norm": 1.0954184532165527,
"learning_rate": 1.0697564737441254e-05,
"loss": 0.1161,
"step": 531
},
{
"epoch": 0.0608,
"grad_norm": 0.7729164361953735,
"learning_rate": 1.0662739004000005e-05,
"loss": 0.0986,
"step": 532
},
{
"epoch": 0.06091428571428571,
"grad_norm": 1.583968162536621,
"learning_rate": 1.0627905195293135e-05,
"loss": 0.1513,
"step": 533
},
{
"epoch": 0.06102857142857143,
"grad_norm": 1.144153356552124,
"learning_rate": 1.0593063735759619e-05,
"loss": 0.0515,
"step": 534
},
{
"epoch": 0.061142857142857145,
"grad_norm": 1.6097029447555542,
"learning_rate": 1.055821504993164e-05,
"loss": 0.1316,
"step": 535
},
{
"epoch": 0.06125714285714286,
"grad_norm": 0.45332613587379456,
"learning_rate": 1.0523359562429441e-05,
"loss": 0.066,
"step": 536
},
{
"epoch": 0.06137142857142857,
"grad_norm": 0.9633097648620605,
"learning_rate": 1.0488497697956134e-05,
"loss": 0.0629,
"step": 537
},
{
"epoch": 0.06148571428571428,
"grad_norm": 0.6255372166633606,
"learning_rate": 1.0453629881292537e-05,
"loss": 0.1007,
"step": 538
},
{
"epoch": 0.0616,
"grad_norm": 0.7773392796516418,
"learning_rate": 1.0418756537291996e-05,
"loss": 0.1251,
"step": 539
},
{
"epoch": 0.061714285714285715,
"grad_norm": 2.2649662494659424,
"learning_rate": 1.03838780908752e-05,
"loss": 0.1262,
"step": 540
},
{
"epoch": 0.06182857142857143,
"grad_norm": 1.0723952054977417,
"learning_rate": 1.0348994967025012e-05,
"loss": 0.1254,
"step": 541
},
{
"epoch": 0.06194285714285714,
"grad_norm": 0.6209930777549744,
"learning_rate": 1.0314107590781284e-05,
"loss": 0.074,
"step": 542
},
{
"epoch": 0.06205714285714286,
"grad_norm": 0.5363054275512695,
"learning_rate": 1.0279216387235691e-05,
"loss": 0.0677,
"step": 543
},
{
"epoch": 0.06217142857142857,
"grad_norm": 0.6603552103042603,
"learning_rate": 1.0244321781526533e-05,
"loss": 0.0622,
"step": 544
},
{
"epoch": 0.062285714285714285,
"grad_norm": 1.0280457735061646,
"learning_rate": 1.0209424198833571e-05,
"loss": 0.1444,
"step": 545
},
{
"epoch": 0.0624,
"grad_norm": 1.1765459775924683,
"learning_rate": 1.0174524064372837e-05,
"loss": 0.1384,
"step": 546
},
{
"epoch": 0.06251428571428572,
"grad_norm": 0.27345114946365356,
"learning_rate": 1.0139621803391454e-05,
"loss": 0.052,
"step": 547
},
{
"epoch": 0.06262857142857142,
"grad_norm": 1.3890050649642944,
"learning_rate": 1.010471784116246e-05,
"loss": 0.0862,
"step": 548
},
{
"epoch": 0.06274285714285714,
"grad_norm": 0.8524456024169922,
"learning_rate": 1.0069812602979617e-05,
"loss": 0.1128,
"step": 549
},
{
"epoch": 0.06285714285714286,
"grad_norm": 1.3577845096588135,
"learning_rate": 1.0034906514152239e-05,
"loss": 0.1301,
"step": 550
},
{
"epoch": 0.06297142857142857,
"grad_norm": 0.34635528922080994,
"learning_rate": 1e-05,
"loss": 0.0446,
"step": 551
},
{
"epoch": 0.06308571428571429,
"grad_norm": 0.3902726173400879,
"learning_rate": 9.965093485847766e-06,
"loss": 0.0589,
"step": 552
},
{
"epoch": 0.0632,
"grad_norm": 0.7191069722175598,
"learning_rate": 9.930187397020385e-06,
"loss": 0.1069,
"step": 553
},
{
"epoch": 0.06331428571428571,
"grad_norm": 4.232479095458984,
"learning_rate": 9.895282158837545e-06,
"loss": 0.1692,
"step": 554
},
{
"epoch": 0.06342857142857143,
"grad_norm": 0.4231140613555908,
"learning_rate": 9.860378196608549e-06,
"loss": 0.0629,
"step": 555
},
{
"epoch": 0.06354285714285714,
"grad_norm": 1.0060336589813232,
"learning_rate": 9.825475935627165e-06,
"loss": 0.1414,
"step": 556
},
{
"epoch": 0.06365714285714286,
"grad_norm": 0.1287923902273178,
"learning_rate": 9.790575801166432e-06,
"loss": 0.0025,
"step": 557
},
{
"epoch": 0.06377142857142858,
"grad_norm": 1.0859025716781616,
"learning_rate": 9.75567821847347e-06,
"loss": 0.0881,
"step": 558
},
{
"epoch": 0.06388571428571428,
"grad_norm": 0.3275485336780548,
"learning_rate": 9.720783612764314e-06,
"loss": 0.0621,
"step": 559
},
{
"epoch": 0.064,
"grad_norm": 1.073527455329895,
"learning_rate": 9.685892409218718e-06,
"loss": 0.109,
"step": 560
},
{
"epoch": 0.06411428571428572,
"grad_norm": 0.7933323979377747,
"learning_rate": 9.651005032974994e-06,
"loss": 0.0962,
"step": 561
},
{
"epoch": 0.06422857142857143,
"grad_norm": 1.3427356481552124,
"learning_rate": 9.616121909124801e-06,
"loss": 0.1362,
"step": 562
},
{
"epoch": 0.06434285714285715,
"grad_norm": 0.7281093597412109,
"learning_rate": 9.581243462708007e-06,
"loss": 0.0637,
"step": 563
},
{
"epoch": 0.06445714285714285,
"grad_norm": 0.8299837112426758,
"learning_rate": 9.546370118707463e-06,
"loss": 0.1089,
"step": 564
},
{
"epoch": 0.06457142857142857,
"grad_norm": 0.6638592481613159,
"learning_rate": 9.511502302043867e-06,
"loss": 0.102,
"step": 565
},
{
"epoch": 0.06468571428571429,
"grad_norm": 0.37742698192596436,
"learning_rate": 9.476640437570562e-06,
"loss": 0.0552,
"step": 566
},
{
"epoch": 0.0648,
"grad_norm": 0.7892422080039978,
"learning_rate": 9.441784950068362e-06,
"loss": 0.0683,
"step": 567
},
{
"epoch": 0.06491428571428572,
"grad_norm": 0.8673684000968933,
"learning_rate": 9.406936264240386e-06,
"loss": 0.123,
"step": 568
},
{
"epoch": 0.06502857142857142,
"grad_norm": 0.9010658264160156,
"learning_rate": 9.372094804706867e-06,
"loss": 0.0537,
"step": 569
},
{
"epoch": 0.06514285714285714,
"grad_norm": 0.5293774008750916,
"learning_rate": 9.337260996000002e-06,
"loss": 0.1091,
"step": 570
},
{
"epoch": 0.06525714285714286,
"grad_norm": 0.5375731587409973,
"learning_rate": 9.302435262558748e-06,
"loss": 0.0502,
"step": 571
},
{
"epoch": 0.06537142857142857,
"grad_norm": 1.1420235633850098,
"learning_rate": 9.267618028723687e-06,
"loss": 0.1276,
"step": 572
},
{
"epoch": 0.06548571428571429,
"grad_norm": 0.5085283517837524,
"learning_rate": 9.232809718731815e-06,
"loss": 0.0507,
"step": 573
},
{
"epoch": 0.0656,
"grad_norm": 32.80242919921875,
"learning_rate": 9.198010756711413e-06,
"loss": 1.1526,
"step": 574
},
{
"epoch": 0.06571428571428571,
"grad_norm": 2.226266622543335,
"learning_rate": 9.163221566676847e-06,
"loss": 0.0719,
"step": 575
},
{
"epoch": 0.06582857142857143,
"grad_norm": 0.7481557130813599,
"learning_rate": 9.128442572523418e-06,
"loss": 0.1166,
"step": 576
},
{
"epoch": 0.06594285714285714,
"grad_norm": 0.07323756068944931,
"learning_rate": 9.093674198022201e-06,
"loss": 0.0005,
"step": 577
},
{
"epoch": 0.06605714285714286,
"grad_norm": 0.6296484470367432,
"learning_rate": 9.058916866814857e-06,
"loss": 0.0583,
"step": 578
},
{
"epoch": 0.06617142857142858,
"grad_norm": 0.8247835040092468,
"learning_rate": 9.024171002408507e-06,
"loss": 0.1181,
"step": 579
},
{
"epoch": 0.06628571428571428,
"grad_norm": 1.388418436050415,
"learning_rate": 8.989437028170537e-06,
"loss": 0.1206,
"step": 580
},
{
"epoch": 0.0664,
"grad_norm": 0.5962538123130798,
"learning_rate": 8.954715367323468e-06,
"loss": 0.0596,
"step": 581
},
{
"epoch": 0.06651428571428572,
"grad_norm": 0.4225008487701416,
"learning_rate": 8.920006442939772e-06,
"loss": 0.0577,
"step": 582
},
{
"epoch": 0.06662857142857143,
"grad_norm": 0.32645943760871887,
"learning_rate": 8.885310677936746e-06,
"loss": 0.0465,
"step": 583
},
{
"epoch": 0.06674285714285715,
"grad_norm": 0.04318680614233017,
"learning_rate": 8.850628495071336e-06,
"loss": 0.0009,
"step": 584
},
{
"epoch": 0.06685714285714285,
"grad_norm": 0.7562779784202576,
"learning_rate": 8.815960316934991e-06,
"loss": 0.1097,
"step": 585
},
{
"epoch": 0.06697142857142857,
"grad_norm": 0.6235162615776062,
"learning_rate": 8.781306565948528e-06,
"loss": 0.1048,
"step": 586
},
{
"epoch": 0.06708571428571429,
"grad_norm": 2.5611071586608887,
"learning_rate": 8.746667664356957e-06,
"loss": 0.0947,
"step": 587
},
{
"epoch": 0.0672,
"grad_norm": 4.237906455993652,
"learning_rate": 8.712044034224374e-06,
"loss": 0.0215,
"step": 588
},
{
"epoch": 0.06731428571428572,
"grad_norm": 0.07394464313983917,
"learning_rate": 8.677436097428775e-06,
"loss": 0.0006,
"step": 589
},
{
"epoch": 0.06742857142857143,
"grad_norm": 0.350160151720047,
"learning_rate": 8.642844275656957e-06,
"loss": 0.0534,
"step": 590
},
{
"epoch": 0.06754285714285714,
"grad_norm": 0.3278239071369171,
"learning_rate": 8.60826899039935e-06,
"loss": 0.0489,
"step": 591
},
{
"epoch": 0.06765714285714286,
"grad_norm": 0.39994513988494873,
"learning_rate": 8.573710662944884e-06,
"loss": 0.0575,
"step": 592
},
{
"epoch": 0.06777142857142857,
"grad_norm": 0.7466303110122681,
"learning_rate": 8.539169714375885e-06,
"loss": 0.0689,
"step": 593
},
{
"epoch": 0.06788571428571429,
"grad_norm": 1.4406474828720093,
"learning_rate": 8.504646565562907e-06,
"loss": 0.1385,
"step": 594
},
{
"epoch": 0.068,
"grad_norm": 0.4625524878501892,
"learning_rate": 8.47014163715962e-06,
"loss": 0.056,
"step": 595
},
{
"epoch": 0.06811428571428571,
"grad_norm": 0.37473708391189575,
"learning_rate": 8.43565534959769e-06,
"loss": 0.088,
"step": 596
},
{
"epoch": 0.06822857142857143,
"grad_norm": 0.5432204604148865,
"learning_rate": 8.401188123081653e-06,
"loss": 0.0552,
"step": 597
},
{
"epoch": 0.06834285714285715,
"grad_norm": 0.7249435186386108,
"learning_rate": 8.366740377583781e-06,
"loss": 0.1033,
"step": 598
},
{
"epoch": 0.06845714285714286,
"grad_norm": 0.45707064867019653,
"learning_rate": 8.332312532838978e-06,
"loss": 0.0593,
"step": 599
},
{
"epoch": 0.06857142857142857,
"grad_norm": 0.3139670789241791,
"learning_rate": 8.297905008339677e-06,
"loss": 0.0505,
"step": 600
},
{
"epoch": 0.06868571428571428,
"grad_norm": 0.9832450151443481,
"learning_rate": 8.263518223330698e-06,
"loss": 0.122,
"step": 601
},
{
"epoch": 0.0688,
"grad_norm": 0.2890370190143585,
"learning_rate": 8.22915259680417e-06,
"loss": 0.0491,
"step": 602
},
{
"epoch": 0.06891428571428572,
"grad_norm": 0.7719813585281372,
"learning_rate": 8.194808547494401e-06,
"loss": 0.0969,
"step": 603
},
{
"epoch": 0.06902857142857143,
"grad_norm": 1.631966471672058,
"learning_rate": 8.1604864938728e-06,
"loss": 0.1387,
"step": 604
},
{
"epoch": 0.06914285714285714,
"grad_norm": 0.7742996215820312,
"learning_rate": 8.126186854142752e-06,
"loss": 0.0555,
"step": 605
},
{
"epoch": 0.06925714285714285,
"grad_norm": 0.32021498680114746,
"learning_rate": 8.091910046234552e-06,
"loss": 0.0436,
"step": 606
},
{
"epoch": 0.06937142857142857,
"grad_norm": 1.0288447141647339,
"learning_rate": 8.057656487800283e-06,
"loss": 0.1272,
"step": 607
},
{
"epoch": 0.06948571428571429,
"grad_norm": 0.8329721093177795,
"learning_rate": 8.023426596208739e-06,
"loss": 0.1058,
"step": 608
},
{
"epoch": 0.0696,
"grad_norm": 0.7122859954833984,
"learning_rate": 7.989220788540356e-06,
"loss": 0.1078,
"step": 609
},
{
"epoch": 0.06971428571428571,
"grad_norm": 0.4297250211238861,
"learning_rate": 7.955039481582098e-06,
"loss": 0.0416,
"step": 610
},
{
"epoch": 0.06982857142857143,
"grad_norm": 0.5651209950447083,
"learning_rate": 7.92088309182241e-06,
"loss": 0.0987,
"step": 611
},
{
"epoch": 0.06994285714285714,
"grad_norm": 0.6013455986976624,
"learning_rate": 7.886752035446116e-06,
"loss": 0.0598,
"step": 612
},
{
"epoch": 0.07005714285714286,
"grad_norm": 0.5885671377182007,
"learning_rate": 7.852646728329368e-06,
"loss": 0.1115,
"step": 613
},
{
"epoch": 0.07017142857142857,
"grad_norm": 0.9340572953224182,
"learning_rate": 7.818567586034578e-06,
"loss": 0.1253,
"step": 614
},
{
"epoch": 0.07028571428571428,
"grad_norm": 1.5684138536453247,
"learning_rate": 7.784515023805328e-06,
"loss": 0.1398,
"step": 615
},
{
"epoch": 0.0704,
"grad_norm": 0.500484824180603,
"learning_rate": 7.750489456561351e-06,
"loss": 0.0514,
"step": 616
},
{
"epoch": 0.07051428571428571,
"grad_norm": 0.9151150584220886,
"learning_rate": 7.716491298893443e-06,
"loss": 0.1118,
"step": 617
},
{
"epoch": 0.07062857142857143,
"grad_norm": 0.03167599067091942,
"learning_rate": 7.68252096505843e-06,
"loss": 0.0008,
"step": 618
},
{
"epoch": 0.07074285714285715,
"grad_norm": 0.5975403189659119,
"learning_rate": 7.6485788689741e-06,
"loss": 0.0529,
"step": 619
},
{
"epoch": 0.07085714285714285,
"grad_norm": 0.7143279910087585,
"learning_rate": 7.6146654242141935e-06,
"loss": 0.0658,
"step": 620
},
{
"epoch": 0.07097142857142857,
"grad_norm": 2.354229688644409,
"learning_rate": 7.580781044003324e-06,
"loss": 0.178,
"step": 621
},
{
"epoch": 0.07108571428571428,
"grad_norm": 0.650444507598877,
"learning_rate": 7.546926141211975e-06,
"loss": 0.0517,
"step": 622
},
{
"epoch": 0.0712,
"grad_norm": 0.2801244258880615,
"learning_rate": 7.513101128351454e-06,
"loss": 0.0536,
"step": 623
},
{
"epoch": 0.07131428571428572,
"grad_norm": 0.2829851508140564,
"learning_rate": 7.4793064175688635e-06,
"loss": 0.0048,
"step": 624
},
{
"epoch": 0.07142857142857142,
"grad_norm": 0.9836239814758301,
"learning_rate": 7.445542420642097e-06,
"loss": 0.1254,
"step": 625
},
{
"epoch": 0.07154285714285714,
"grad_norm": 0.7138457298278809,
"learning_rate": 7.411809548974792e-06,
"loss": 0.0671,
"step": 626
},
{
"epoch": 0.07165714285714286,
"grad_norm": 0.6020057201385498,
"learning_rate": 7.378108213591355e-06,
"loss": 0.1034,
"step": 627
},
{
"epoch": 0.07177142857142857,
"grad_norm": 0.3739442527294159,
"learning_rate": 7.344438825131912e-06,
"loss": 0.0503,
"step": 628
},
{
"epoch": 0.07188571428571429,
"grad_norm": 0.28676095604896545,
"learning_rate": 7.310801793847344e-06,
"loss": 0.0431,
"step": 629
},
{
"epoch": 0.072,
"grad_norm": 0.839794933795929,
"learning_rate": 7.277197529594257e-06,
"loss": 0.0591,
"step": 630
},
{
"epoch": 0.07211428571428571,
"grad_norm": 0.7637845873832703,
"learning_rate": 7.243626441830009e-06,
"loss": 0.1091,
"step": 631
},
{
"epoch": 0.07222857142857143,
"grad_norm": 0.7257825136184692,
"learning_rate": 7.210088939607709e-06,
"loss": 0.1357,
"step": 632
},
{
"epoch": 0.07234285714285714,
"grad_norm": 1.113267421722412,
"learning_rate": 7.176585431571235e-06,
"loss": 0.0937,
"step": 633
},
{
"epoch": 0.07245714285714286,
"grad_norm": 0.5612810850143433,
"learning_rate": 7.143116325950266e-06,
"loss": 0.0587,
"step": 634
},
{
"epoch": 0.07257142857142856,
"grad_norm": 0.9396265149116516,
"learning_rate": 7.109682030555283e-06,
"loss": 0.1152,
"step": 635
},
{
"epoch": 0.07268571428571428,
"grad_norm": 0.4426088333129883,
"learning_rate": 7.076282952772634e-06,
"loss": 0.0532,
"step": 636
},
{
"epoch": 0.0728,
"grad_norm": 0.4740736782550812,
"learning_rate": 7.042919499559538e-06,
"loss": 0.1005,
"step": 637
},
{
"epoch": 0.07291428571428571,
"grad_norm": 0.7607542872428894,
"learning_rate": 7.009592077439135e-06,
"loss": 0.1098,
"step": 638
},
{
"epoch": 0.07302857142857143,
"grad_norm": 1.8538681268692017,
"learning_rate": 6.976301092495556e-06,
"loss": 0.127,
"step": 639
},
{
"epoch": 0.07314285714285715,
"grad_norm": 1.1161400079727173,
"learning_rate": 6.943046950368944e-06,
"loss": 0.0623,
"step": 640
},
{
"epoch": 0.07325714285714285,
"grad_norm": 2.595522165298462,
"learning_rate": 6.909830056250527e-06,
"loss": 0.1493,
"step": 641
},
{
"epoch": 0.07337142857142857,
"grad_norm": 1.184796690940857,
"learning_rate": 6.876650814877675e-06,
"loss": 0.0769,
"step": 642
},
{
"epoch": 0.07348571428571428,
"grad_norm": 1.257194995880127,
"learning_rate": 6.843509630528977e-06,
"loss": 0.1224,
"step": 643
},
{
"epoch": 0.0736,
"grad_norm": 0.7015554308891296,
"learning_rate": 6.8104069070193e-06,
"loss": 0.0615,
"step": 644
},
{
"epoch": 0.07371428571428572,
"grad_norm": 0.4807669520378113,
"learning_rate": 6.777343047694891e-06,
"loss": 0.0526,
"step": 645
},
{
"epoch": 0.07382857142857142,
"grad_norm": 0.7870861291885376,
"learning_rate": 6.744318455428436e-06,
"loss": 0.1175,
"step": 646
},
{
"epoch": 0.07394285714285714,
"grad_norm": 0.7987018823623657,
"learning_rate": 6.711333532614168e-06,
"loss": 0.0676,
"step": 647
},
{
"epoch": 0.07405714285714286,
"grad_norm": 4.092370986938477,
"learning_rate": 6.67838868116297e-06,
"loss": 0.142,
"step": 648
},
{
"epoch": 0.07417142857142857,
"grad_norm": 4.180191516876221,
"learning_rate": 6.645484302497452e-06,
"loss": 0.2342,
"step": 649
},
{
"epoch": 0.07428571428571429,
"grad_norm": 0.9793099761009216,
"learning_rate": 6.612620797547087e-06,
"loss": 0.0654,
"step": 650
},
{
"epoch": 0.0744,
"grad_norm": 0.45217183232307434,
"learning_rate": 6.579798566743314e-06,
"loss": 0.1056,
"step": 651
},
{
"epoch": 0.07451428571428571,
"grad_norm": 0.6437646150588989,
"learning_rate": 6.547018010014654e-06,
"loss": 0.1176,
"step": 652
},
{
"epoch": 0.07462857142857143,
"grad_norm": 0.39132067561149597,
"learning_rate": 6.5142795267818505e-06,
"loss": 0.0503,
"step": 653
},
{
"epoch": 0.07474285714285714,
"grad_norm": 0.7829943299293518,
"learning_rate": 6.481583515952983e-06,
"loss": 0.0648,
"step": 654
},
{
"epoch": 0.07485714285714286,
"grad_norm": 0.06885892897844315,
"learning_rate": 6.448930375918632e-06,
"loss": 0.0023,
"step": 655
},
{
"epoch": 0.07497142857142858,
"grad_norm": 0.3638521730899811,
"learning_rate": 6.4163205045469975e-06,
"loss": 0.0576,
"step": 656
},
{
"epoch": 0.07508571428571428,
"grad_norm": 1.1378263235092163,
"learning_rate": 6.383754299179079e-06,
"loss": 0.1119,
"step": 657
},
{
"epoch": 0.0752,
"grad_norm": 1.2728840112686157,
"learning_rate": 6.351232156623803e-06,
"loss": 0.1146,
"step": 658
},
{
"epoch": 0.07531428571428571,
"grad_norm": 0.5690306425094604,
"learning_rate": 6.318754473153221e-06,
"loss": 0.1018,
"step": 659
},
{
"epoch": 0.07542857142857143,
"grad_norm": 0.049346674233675,
"learning_rate": 6.286321644497655e-06,
"loss": 0.0012,
"step": 660
},
{
"epoch": 0.07554285714285715,
"grad_norm": 1.1735026836395264,
"learning_rate": 6.25393406584088e-06,
"loss": 0.1096,
"step": 661
},
{
"epoch": 0.07565714285714285,
"grad_norm": 0.5500955581665039,
"learning_rate": 6.22159213181533e-06,
"loss": 0.0649,
"step": 662
},
{
"epoch": 0.07577142857142857,
"grad_norm": 2.1257147789001465,
"learning_rate": 6.18929623649726e-06,
"loss": 0.0978,
"step": 663
},
{
"epoch": 0.07588571428571429,
"grad_norm": 1.1304757595062256,
"learning_rate": 6.157046773401964e-06,
"loss": 0.089,
"step": 664
},
{
"epoch": 0.076,
"grad_norm": 0.8758174180984497,
"learning_rate": 6.124844135478971e-06,
"loss": 0.1075,
"step": 665
},
{
"epoch": 0.07611428571428572,
"grad_norm": 0.20409873127937317,
"learning_rate": 6.092688715107265e-06,
"loss": 0.0477,
"step": 666
},
{
"epoch": 0.07622857142857142,
"grad_norm": 0.7486653327941895,
"learning_rate": 6.06058090409049e-06,
"loss": 0.1084,
"step": 667
},
{
"epoch": 0.07634285714285714,
"grad_norm": 1.3490878343582153,
"learning_rate": 6.028521093652195e-06,
"loss": 0.109,
"step": 668
},
{
"epoch": 0.07645714285714286,
"grad_norm": 0.3590696454048157,
"learning_rate": 5.996509674431053e-06,
"loss": 0.0415,
"step": 669
},
{
"epoch": 0.07657142857142857,
"grad_norm": 0.8520549535751343,
"learning_rate": 5.9645470364761e-06,
"loss": 0.1019,
"step": 670
},
{
"epoch": 0.07668571428571429,
"grad_norm": 0.695841908454895,
"learning_rate": 5.932633569242e-06,
"loss": 0.0672,
"step": 671
},
{
"epoch": 0.0768,
"grad_norm": 0.579662024974823,
"learning_rate": 5.900769661584273e-06,
"loss": 0.1026,
"step": 672
},
{
"epoch": 0.07691428571428571,
"grad_norm": 0.9331521391868591,
"learning_rate": 5.868955701754584e-06,
"loss": 0.1172,
"step": 673
},
{
"epoch": 0.07702857142857143,
"grad_norm": 0.4442518353462219,
"learning_rate": 5.83719207739599e-06,
"loss": 0.0416,
"step": 674
},
{
"epoch": 0.07714285714285714,
"grad_norm": 0.6794987320899963,
"learning_rate": 5.8054791755382286e-06,
"loss": 0.1208,
"step": 675
},
{
"epoch": 0.07725714285714286,
"grad_norm": 3.283614158630371,
"learning_rate": 5.773817382593008e-06,
"loss": 0.1716,
"step": 676
},
{
"epoch": 0.07737142857142858,
"grad_norm": 0.7049353718757629,
"learning_rate": 5.742207084349274e-06,
"loss": 0.0694,
"step": 677
},
{
"epoch": 0.07748571428571428,
"grad_norm": 0.933340311050415,
"learning_rate": 5.710648665968543e-06,
"loss": 0.0611,
"step": 678
},
{
"epoch": 0.0776,
"grad_norm": 0.7235684394836426,
"learning_rate": 5.679142511980176e-06,
"loss": 0.1234,
"step": 679
},
{
"epoch": 0.07771428571428571,
"grad_norm": 1.197269320487976,
"learning_rate": 5.647689006276727e-06,
"loss": 0.1181,
"step": 680
},
{
"epoch": 0.07782857142857143,
"grad_norm": 0.7340664267539978,
"learning_rate": 5.616288532109225e-06,
"loss": 0.0564,
"step": 681
},
{
"epoch": 0.07794285714285715,
"grad_norm": 4.762490749359131,
"learning_rate": 5.584941472082549e-06,
"loss": 0.1237,
"step": 682
},
{
"epoch": 0.07805714285714285,
"grad_norm": 0.4227953553199768,
"learning_rate": 5.553648208150728e-06,
"loss": 0.0594,
"step": 683
},
{
"epoch": 0.07817142857142857,
"grad_norm": 0.3790622353553772,
"learning_rate": 5.522409121612304e-06,
"loss": 0.0679,
"step": 684
},
{
"epoch": 0.07828571428571429,
"grad_norm": 0.6780324578285217,
"learning_rate": 5.491224593105695e-06,
"loss": 0.0691,
"step": 685
},
{
"epoch": 0.0784,
"grad_norm": 0.99502032995224,
"learning_rate": 5.460095002604533e-06,
"loss": 0.1165,
"step": 686
},
{
"epoch": 0.07851428571428572,
"grad_norm": 0.5032870769500732,
"learning_rate": 5.429020729413062e-06,
"loss": 0.0523,
"step": 687
},
{
"epoch": 0.07862857142857142,
"grad_norm": 1.5868319272994995,
"learning_rate": 5.398002152161484e-06,
"loss": 0.08,
"step": 688
},
{
"epoch": 0.07874285714285714,
"grad_norm": 0.6237499713897705,
"learning_rate": 5.367039648801386e-06,
"loss": 0.058,
"step": 689
},
{
"epoch": 0.07885714285714286,
"grad_norm": 0.4059104919433594,
"learning_rate": 5.336133596601089e-06,
"loss": 0.0526,
"step": 690
},
{
"epoch": 0.07897142857142857,
"grad_norm": 1.475788950920105,
"learning_rate": 5.305284372141095e-06,
"loss": 0.129,
"step": 691
},
{
"epoch": 0.07908571428571429,
"grad_norm": 0.3848338723182678,
"learning_rate": 5.274492351309462e-06,
"loss": 0.0534,
"step": 692
},
{
"epoch": 0.0792,
"grad_norm": 1.2723824977874756,
"learning_rate": 5.243757909297247e-06,
"loss": 0.1115,
"step": 693
},
{
"epoch": 0.07931428571428571,
"grad_norm": 0.08853624016046524,
"learning_rate": 5.213081420593933e-06,
"loss": 0.003,
"step": 694
},
{
"epoch": 0.07942857142857143,
"grad_norm": 0.20749793946743011,
"learning_rate": 5.1824632589828465e-06,
"loss": 0.0485,
"step": 695
},
{
"epoch": 0.07954285714285714,
"grad_norm": 0.8824698328971863,
"learning_rate": 5.151903797536631e-06,
"loss": 0.064,
"step": 696
},
{
"epoch": 0.07965714285714286,
"grad_norm": 0.6835923194885254,
"learning_rate": 5.121403408612672e-06,
"loss": 0.1031,
"step": 697
},
{
"epoch": 0.07977142857142858,
"grad_norm": 0.43906140327453613,
"learning_rate": 5.090962463848592e-06,
"loss": 0.0929,
"step": 698
},
{
"epoch": 0.07988571428571428,
"grad_norm": 1.1689563989639282,
"learning_rate": 5.060581334157693e-06,
"loss": 0.135,
"step": 699
},
{
"epoch": 0.08,
"grad_norm": 0.617749810218811,
"learning_rate": 5.030260389724447e-06,
"loss": 0.11,
"step": 700
},
{
"epoch": 0.08011428571428572,
"grad_norm": 0.845007061958313,
"learning_rate": 5.000000000000003e-06,
"loss": 0.1085,
"step": 701
},
{
"epoch": 0.08022857142857143,
"grad_norm": 0.8941121101379395,
"learning_rate": 4.96980053369765e-06,
"loss": 0.0733,
"step": 702
},
{
"epoch": 0.08034285714285715,
"grad_norm": 0.022771498188376427,
"learning_rate": 4.939662358788364e-06,
"loss": 0.0004,
"step": 703
},
{
"epoch": 0.08045714285714285,
"grad_norm": 0.8632171750068665,
"learning_rate": 4.909585842496287e-06,
"loss": 0.095,
"step": 704
},
{
"epoch": 0.08057142857142857,
"grad_norm": 0.9821418523788452,
"learning_rate": 4.879571351294287e-06,
"loss": 0.1441,
"step": 705
},
{
"epoch": 0.08068571428571429,
"grad_norm": 0.9110166430473328,
"learning_rate": 4.849619250899458e-06,
"loss": 0.12,
"step": 706
},
{
"epoch": 0.0808,
"grad_norm": 0.523320198059082,
"learning_rate": 4.8197299062687e-06,
"loss": 0.0658,
"step": 707
},
{
"epoch": 0.08091428571428572,
"grad_norm": 1.6946700811386108,
"learning_rate": 4.78990368159424e-06,
"loss": 0.1316,
"step": 708
},
{
"epoch": 0.08102857142857142,
"grad_norm": 0.9995729923248291,
"learning_rate": 4.76014094029921e-06,
"loss": 0.1176,
"step": 709
},
{
"epoch": 0.08114285714285714,
"grad_norm": 2.1208927631378174,
"learning_rate": 4.7304420450332244e-06,
"loss": 0.1465,
"step": 710
},
{
"epoch": 0.08125714285714286,
"grad_norm": 1.7362505197525024,
"learning_rate": 4.700807357667953e-06,
"loss": 0.144,
"step": 711
},
{
"epoch": 0.08137142857142857,
"grad_norm": 0.7457737922668457,
"learning_rate": 4.671237239292699e-06,
"loss": 0.0722,
"step": 712
},
{
"epoch": 0.08148571428571429,
"grad_norm": 4.4693498611450195,
"learning_rate": 4.641732050210032e-06,
"loss": 0.1765,
"step": 713
},
{
"epoch": 0.0816,
"grad_norm": 0.6143847107887268,
"learning_rate": 4.612292149931369e-06,
"loss": 0.0652,
"step": 714
},
{
"epoch": 0.08171428571428571,
"grad_norm": 1.0831738710403442,
"learning_rate": 4.582917897172603e-06,
"loss": 0.1369,
"step": 715
},
{
"epoch": 0.08182857142857143,
"grad_norm": 0.564538300037384,
"learning_rate": 4.5536096498497295e-06,
"loss": 0.0526,
"step": 716
},
{
"epoch": 0.08194285714285714,
"grad_norm": 0.49217793345451355,
"learning_rate": 4.524367765074499e-06,
"loss": 0.1079,
"step": 717
},
{
"epoch": 0.08205714285714286,
"grad_norm": 0.9084864854812622,
"learning_rate": 4.495192599150045e-06,
"loss": 0.1066,
"step": 718
},
{
"epoch": 0.08217142857142858,
"grad_norm": 1.5947651863098145,
"learning_rate": 4.46608450756656e-06,
"loss": 0.1476,
"step": 719
},
{
"epoch": 0.08228571428571428,
"grad_norm": 1.25254487991333,
"learning_rate": 4.437043844996952e-06,
"loss": 0.1106,
"step": 720
},
{
"epoch": 0.0824,
"grad_norm": 0.904678463935852,
"learning_rate": 4.408070965292534e-06,
"loss": 0.1035,
"step": 721
},
{
"epoch": 0.08251428571428572,
"grad_norm": 0.6488781571388245,
"learning_rate": 4.379166221478697e-06,
"loss": 0.1033,
"step": 722
},
{
"epoch": 0.08262857142857143,
"grad_norm": 0.030457662418484688,
"learning_rate": 4.350329965750622e-06,
"loss": 0.0007,
"step": 723
},
{
"epoch": 0.08274285714285715,
"grad_norm": 0.27059054374694824,
"learning_rate": 4.321562549468991e-06,
"loss": 0.049,
"step": 724
},
{
"epoch": 0.08285714285714285,
"grad_norm": 0.5485967993736267,
"learning_rate": 4.292864323155684e-06,
"loss": 0.0617,
"step": 725
},
{
"epoch": 0.08297142857142857,
"grad_norm": 0.6584226489067078,
"learning_rate": 4.264235636489542e-06,
"loss": 0.099,
"step": 726
},
{
"epoch": 0.08308571428571429,
"grad_norm": 1.0592844486236572,
"learning_rate": 4.235676838302069e-06,
"loss": 0.0962,
"step": 727
},
{
"epoch": 0.0832,
"grad_norm": 0.7608891725540161,
"learning_rate": 4.207188276573214e-06,
"loss": 0.1021,
"step": 728
},
{
"epoch": 0.08331428571428572,
"grad_norm": 0.9272671341896057,
"learning_rate": 4.178770298427107e-06,
"loss": 0.0687,
"step": 729
},
{
"epoch": 0.08342857142857144,
"grad_norm": 3.1046485900878906,
"learning_rate": 4.150423250127846e-06,
"loss": 0.1335,
"step": 730
},
{
"epoch": 0.08354285714285714,
"grad_norm": 0.30813950300216675,
"learning_rate": 4.12214747707527e-06,
"loss": 0.0568,
"step": 731
},
{
"epoch": 0.08365714285714286,
"grad_norm": 0.8978940844535828,
"learning_rate": 4.093943323800746e-06,
"loss": 0.12,
"step": 732
},
{
"epoch": 0.08377142857142857,
"grad_norm": 0.5154088735580444,
"learning_rate": 4.065811133962987e-06,
"loss": 0.0735,
"step": 733
},
{
"epoch": 0.08388571428571429,
"grad_norm": 0.6369738578796387,
"learning_rate": 4.037751250343841e-06,
"loss": 0.0719,
"step": 734
},
{
"epoch": 0.084,
"grad_norm": 0.9210620522499084,
"learning_rate": 4.009764014844143e-06,
"loss": 0.1074,
"step": 735
},
{
"epoch": 0.08411428571428571,
"grad_norm": 0.8698270320892334,
"learning_rate": 3.981849768479516e-06,
"loss": 0.1145,
"step": 736
},
{
"epoch": 0.08422857142857143,
"grad_norm": 1.0493848323822021,
"learning_rate": 3.954008851376252e-06,
"loss": 0.1089,
"step": 737
},
{
"epoch": 0.08434285714285714,
"grad_norm": 1.1828758716583252,
"learning_rate": 3.9262416027671354e-06,
"loss": 0.1344,
"step": 738
},
{
"epoch": 0.08445714285714286,
"grad_norm": 0.7396812438964844,
"learning_rate": 3.898548360987325e-06,
"loss": 0.1137,
"step": 739
},
{
"epoch": 0.08457142857142858,
"grad_norm": 0.5415331721305847,
"learning_rate": 3.8709294634702374e-06,
"loss": 0.0492,
"step": 740
},
{
"epoch": 0.08468571428571428,
"grad_norm": 0.9625293016433716,
"learning_rate": 3.8433852467434175e-06,
"loss": 0.0781,
"step": 741
},
{
"epoch": 0.0848,
"grad_norm": 0.0349762849509716,
"learning_rate": 3.81591604642446e-06,
"loss": 0.0011,
"step": 742
},
{
"epoch": 0.08491428571428572,
"grad_norm": 0.3069891929626465,
"learning_rate": 3.7885221972168974e-06,
"loss": 0.0555,
"step": 743
},
{
"epoch": 0.08502857142857143,
"grad_norm": 0.9470656514167786,
"learning_rate": 3.7612040329061405e-06,
"loss": 0.1108,
"step": 744
},
{
"epoch": 0.08514285714285715,
"grad_norm": 0.8957858681678772,
"learning_rate": 3.7339618863553983e-06,
"loss": 0.1158,
"step": 745
},
{
"epoch": 0.08525714285714285,
"grad_norm": 0.7257146239280701,
"learning_rate": 3.7067960895016277e-06,
"loss": 0.1192,
"step": 746
},
{
"epoch": 0.08537142857142857,
"grad_norm": 1.328791618347168,
"learning_rate": 3.679706973351491e-06,
"loss": 0.1394,
"step": 747
},
{
"epoch": 0.08548571428571429,
"grad_norm": 1.0056930780410767,
"learning_rate": 3.6526948679773256e-06,
"loss": 0.1247,
"step": 748
},
{
"epoch": 0.0856,
"grad_norm": 0.6528775095939636,
"learning_rate": 3.625760102513103e-06,
"loss": 0.1027,
"step": 749
},
{
"epoch": 0.08571428571428572,
"grad_norm": 0.6015933752059937,
"learning_rate": 3.598903005150444e-06,
"loss": 0.0477,
"step": 750
},
{
"epoch": 0.08582857142857143,
"grad_norm": 1.6941736936569214,
"learning_rate": 3.5721239031346067e-06,
"loss": 0.1663,
"step": 751
},
{
"epoch": 0.08594285714285714,
"grad_norm": 0.8489632606506348,
"learning_rate": 3.545423122760493e-06,
"loss": 0.1232,
"step": 752
},
{
"epoch": 0.08605714285714286,
"grad_norm": 0.3967733085155487,
"learning_rate": 3.5188009893686916e-06,
"loss": 0.0617,
"step": 753
},
{
"epoch": 0.08617142857142857,
"grad_norm": 0.5610271096229553,
"learning_rate": 3.492257827341492e-06,
"loss": 0.1116,
"step": 754
},
{
"epoch": 0.08628571428571429,
"grad_norm": 0.4398455023765564,
"learning_rate": 3.4657939600989453e-06,
"loss": 0.053,
"step": 755
},
{
"epoch": 0.0864,
"grad_norm": 1.1644352674484253,
"learning_rate": 3.4394097100949286e-06,
"loss": 0.1103,
"step": 756
},
{
"epoch": 0.08651428571428571,
"grad_norm": 0.022055093199014664,
"learning_rate": 3.4131053988131947e-06,
"loss": 0.0011,
"step": 757
},
{
"epoch": 0.08662857142857143,
"grad_norm": 0.5757044553756714,
"learning_rate": 3.3868813467634833e-06,
"loss": 0.101,
"step": 758
},
{
"epoch": 0.08674285714285715,
"grad_norm": 2.0454261302948,
"learning_rate": 3.360737873477584e-06,
"loss": 0.1212,
"step": 759
},
{
"epoch": 0.08685714285714285,
"grad_norm": 0.5118523836135864,
"learning_rate": 3.3346752975054763e-06,
"loss": 0.0541,
"step": 760
},
{
"epoch": 0.08697142857142857,
"grad_norm": 0.5061915516853333,
"learning_rate": 3.308693936411421e-06,
"loss": 0.0481,
"step": 761
},
{
"epoch": 0.08708571428571428,
"grad_norm": 1.3214257955551147,
"learning_rate": 3.2827941067700996e-06,
"loss": 0.1239,
"step": 762
},
{
"epoch": 0.0872,
"grad_norm": 1.7101081609725952,
"learning_rate": 3.2569761241627694e-06,
"loss": 0.1205,
"step": 763
},
{
"epoch": 0.08731428571428572,
"grad_norm": 0.48467421531677246,
"learning_rate": 3.2312403031733943e-06,
"loss": 0.0526,
"step": 764
},
{
"epoch": 0.08742857142857142,
"grad_norm": 1.0977176427841187,
"learning_rate": 3.2055869573848374e-06,
"loss": 0.1097,
"step": 765
},
{
"epoch": 0.08754285714285714,
"grad_norm": 0.7671828866004944,
"learning_rate": 3.1800163993750166e-06,
"loss": 0.1048,
"step": 766
},
{
"epoch": 0.08765714285714286,
"grad_norm": 1.0653796195983887,
"learning_rate": 3.1545289407131128e-06,
"loss": 0.0641,
"step": 767
},
{
"epoch": 0.08777142857142857,
"grad_norm": 0.9216064214706421,
"learning_rate": 3.1291248919557717e-06,
"loss": 0.1094,
"step": 768
},
{
"epoch": 0.08788571428571429,
"grad_norm": 0.5879707932472229,
"learning_rate": 3.103804562643302e-06,
"loss": 0.0575,
"step": 769
},
{
"epoch": 0.088,
"grad_norm": 0.467585027217865,
"learning_rate": 3.0785682612959334e-06,
"loss": 0.0462,
"step": 770
},
{
"epoch": 0.08811428571428571,
"grad_norm": 0.7656713128089905,
"learning_rate": 3.0534162954100264e-06,
"loss": 0.1113,
"step": 771
},
{
"epoch": 0.08822857142857143,
"grad_norm": 0.021971160545945168,
"learning_rate": 3.028348971454356e-06,
"loss": 0.001,
"step": 772
},
{
"epoch": 0.08834285714285714,
"grad_norm": 0.6035601496696472,
"learning_rate": 3.003366594866345e-06,
"loss": 0.0666,
"step": 773
},
{
"epoch": 0.08845714285714286,
"grad_norm": 0.39699748158454895,
"learning_rate": 2.978469470048376e-06,
"loss": 0.0515,
"step": 774
},
{
"epoch": 0.08857142857142856,
"grad_norm": 0.3424980342388153,
"learning_rate": 2.953657900364053e-06,
"loss": 0.0411,
"step": 775
},
{
"epoch": 0.08868571428571428,
"grad_norm": 0.5518139004707336,
"learning_rate": 2.9289321881345257e-06,
"loss": 0.1024,
"step": 776
},
{
"epoch": 0.0888,
"grad_norm": 0.9892997741699219,
"learning_rate": 2.9042926346347932e-06,
"loss": 0.0682,
"step": 777
},
{
"epoch": 0.08891428571428571,
"grad_norm": 0.5007666945457458,
"learning_rate": 2.8797395400900362e-06,
"loss": 0.0582,
"step": 778
},
{
"epoch": 0.08902857142857143,
"grad_norm": 4.517937183380127,
"learning_rate": 2.855273203671969e-06,
"loss": 0.0759,
"step": 779
},
{
"epoch": 0.08914285714285715,
"grad_norm": 0.925056517124176,
"learning_rate": 2.830893923495173e-06,
"loss": 0.1034,
"step": 780
},
{
"epoch": 0.08925714285714285,
"grad_norm": 0.4738568067550659,
"learning_rate": 2.8066019966134907e-06,
"loss": 0.049,
"step": 781
},
{
"epoch": 0.08937142857142857,
"grad_norm": 0.4791397452354431,
"learning_rate": 2.7823977190163788e-06,
"loss": 0.1022,
"step": 782
},
{
"epoch": 0.08948571428571428,
"grad_norm": 0.4478612542152405,
"learning_rate": 2.7582813856253276e-06,
"loss": 0.0508,
"step": 783
},
{
"epoch": 0.0896,
"grad_norm": 0.6503352522850037,
"learning_rate": 2.7342532902902418e-06,
"loss": 0.1204,
"step": 784
},
{
"epoch": 0.08971428571428572,
"grad_norm": 0.37914592027664185,
"learning_rate": 2.7103137257858867e-06,
"loss": 0.0533,
"step": 785
},
{
"epoch": 0.08982857142857142,
"grad_norm": 0.9657279849052429,
"learning_rate": 2.6864629838082957e-06,
"loss": 0.1114,
"step": 786
},
{
"epoch": 0.08994285714285714,
"grad_norm": 0.9328279495239258,
"learning_rate": 2.6627013549712355e-06,
"loss": 0.1247,
"step": 787
},
{
"epoch": 0.09005714285714286,
"grad_norm": 0.5992411375045776,
"learning_rate": 2.639029128802657e-06,
"loss": 0.0619,
"step": 788
},
{
"epoch": 0.09017142857142857,
"grad_norm": 0.6358429193496704,
"learning_rate": 2.615446593741161e-06,
"loss": 0.1091,
"step": 789
},
{
"epoch": 0.09028571428571429,
"grad_norm": 0.6846811175346375,
"learning_rate": 2.5919540371325005e-06,
"loss": 0.1129,
"step": 790
},
{
"epoch": 0.0904,
"grad_norm": 0.43121761083602905,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.0471,
"step": 791
},
{
"epoch": 0.09051428571428571,
"grad_norm": 0.8021223545074463,
"learning_rate": 2.5452400031713786e-06,
"loss": 0.1161,
"step": 792
},
{
"epoch": 0.09062857142857143,
"grad_norm": 0.8482909798622131,
"learning_rate": 2.522019095014683e-06,
"loss": 0.0526,
"step": 793
},
{
"epoch": 0.09074285714285714,
"grad_norm": 0.5687688589096069,
"learning_rate": 2.4988893036954045e-06,
"loss": 0.0586,
"step": 794
},
{
"epoch": 0.09085714285714286,
"grad_norm": 0.02480258233845234,
"learning_rate": 2.4758509110427576e-06,
"loss": 0.0011,
"step": 795
},
{
"epoch": 0.09097142857142858,
"grad_norm": 6.957435131072998,
"learning_rate": 2.45290419777228e-06,
"loss": 0.2072,
"step": 796
},
{
"epoch": 0.09108571428571428,
"grad_norm": 0.3257977068424225,
"learning_rate": 2.4300494434824373e-06,
"loss": 0.0431,
"step": 797
},
{
"epoch": 0.0912,
"grad_norm": 2.345851182937622,
"learning_rate": 2.407286926651192e-06,
"loss": 0.1417,
"step": 798
},
{
"epoch": 0.09131428571428571,
"grad_norm": 1.0585758686065674,
"learning_rate": 2.3846169246326345e-06,
"loss": 0.0686,
"step": 799
},
{
"epoch": 0.09142857142857143,
"grad_norm": 0.5051372647285461,
"learning_rate": 2.362039713653581e-06,
"loss": 0.056,
"step": 800
},
{
"epoch": 0.09154285714285715,
"grad_norm": 0.37791648507118225,
"learning_rate": 2.339555568810221e-06,
"loss": 0.0511,
"step": 801
},
{
"epoch": 0.09165714285714285,
"grad_norm": 0.3963716924190521,
"learning_rate": 2.317164764064769e-06,
"loss": 0.055,
"step": 802
},
{
"epoch": 0.09177142857142857,
"grad_norm": 0.6705289483070374,
"learning_rate": 2.2948675722421086e-06,
"loss": 0.1039,
"step": 803
},
{
"epoch": 0.09188571428571428,
"grad_norm": 0.26530247926712036,
"learning_rate": 2.27266426502649e-06,
"loss": 0.043,
"step": 804
},
{
"epoch": 0.092,
"grad_norm": 0.6296550631523132,
"learning_rate": 2.2505551129582047e-06,
"loss": 0.1103,
"step": 805
},
{
"epoch": 0.09211428571428572,
"grad_norm": 0.5118422508239746,
"learning_rate": 2.2285403854302912e-06,
"loss": 0.0985,
"step": 806
},
{
"epoch": 0.09222857142857142,
"grad_norm": 12.643540382385254,
"learning_rate": 2.206620350685257e-06,
"loss": 0.3876,
"step": 807
},
{
"epoch": 0.09234285714285714,
"grad_norm": 0.44169461727142334,
"learning_rate": 2.1847952758118118e-06,
"loss": 0.0482,
"step": 808
},
{
"epoch": 0.09245714285714286,
"grad_norm": 0.058004774153232574,
"learning_rate": 2.163065426741603e-06,
"loss": 0.0014,
"step": 809
},
{
"epoch": 0.09257142857142857,
"grad_norm": 0.5432921648025513,
"learning_rate": 2.1414310682459805e-06,
"loss": 0.0495,
"step": 810
},
{
"epoch": 0.09268571428571429,
"grad_norm": 0.8585667610168457,
"learning_rate": 2.119892463932781e-06,
"loss": 0.0546,
"step": 811
},
{
"epoch": 0.0928,
"grad_norm": 1.0429165363311768,
"learning_rate": 2.098449876243096e-06,
"loss": 0.0584,
"step": 812
},
{
"epoch": 0.09291428571428571,
"grad_norm": 0.4918368458747864,
"learning_rate": 2.0771035664480944e-06,
"loss": 0.1071,
"step": 813
},
{
"epoch": 0.09302857142857143,
"grad_norm": 0.8457767963409424,
"learning_rate": 2.0558537946458177e-06,
"loss": 0.1135,
"step": 814
},
{
"epoch": 0.09314285714285714,
"grad_norm": 0.795174777507782,
"learning_rate": 2.0347008197580376e-06,
"loss": 0.1109,
"step": 815
},
{
"epoch": 0.09325714285714286,
"grad_norm": 0.5936285853385925,
"learning_rate": 2.013644899527074e-06,
"loss": 0.0618,
"step": 816
},
{
"epoch": 0.09337142857142858,
"grad_norm": 1.5751475095748901,
"learning_rate": 1.9926862905126663e-06,
"loss": 0.1333,
"step": 817
},
{
"epoch": 0.09348571428571428,
"grad_norm": 0.9550502300262451,
"learning_rate": 1.9718252480888567e-06,
"loss": 0.1003,
"step": 818
},
{
"epoch": 0.0936,
"grad_norm": 0.24387723207473755,
"learning_rate": 1.95106202644086e-06,
"loss": 0.0526,
"step": 819
},
{
"epoch": 0.09371428571428571,
"grad_norm": 0.48181673884391785,
"learning_rate": 1.930396878561983e-06,
"loss": 0.0551,
"step": 820
},
{
"epoch": 0.09382857142857143,
"grad_norm": 1.059958815574646,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.1078,
"step": 821
},
{
"epoch": 0.09394285714285715,
"grad_norm": 0.4085182249546051,
"learning_rate": 1.8893618101067357e-06,
"loss": 0.0925,
"step": 822
},
{
"epoch": 0.09405714285714285,
"grad_norm": 15.399497985839844,
"learning_rate": 1.8689923895297247e-06,
"loss": 0.2978,
"step": 823
},
{
"epoch": 0.09417142857142857,
"grad_norm": 0.07142768055200577,
"learning_rate": 1.848722042714457e-06,
"loss": 0.0025,
"step": 824
},
{
"epoch": 0.09428571428571429,
"grad_norm": 0.022009899839758873,
"learning_rate": 1.8285510166487154e-06,
"loss": 0.0012,
"step": 825
},
{
"epoch": 0.0944,
"grad_norm": 0.6226180195808411,
"learning_rate": 1.808479557110081e-06,
"loss": 0.1225,
"step": 826
},
{
"epoch": 0.09451428571428572,
"grad_norm": 0.29665324091911316,
"learning_rate": 1.7885079086629598e-06,
"loss": 0.0471,
"step": 827
},
{
"epoch": 0.09462857142857142,
"grad_norm": 0.6959401369094849,
"learning_rate": 1.7686363146555807e-06,
"loss": 0.0959,
"step": 828
},
{
"epoch": 0.09474285714285714,
"grad_norm": 1.2162383794784546,
"learning_rate": 1.7488650172170496e-06,
"loss": 0.1143,
"step": 829
},
{
"epoch": 0.09485714285714286,
"grad_norm": 0.7188528180122375,
"learning_rate": 1.7291942572543806e-06,
"loss": 0.1,
"step": 830
},
{
"epoch": 0.09497142857142857,
"grad_norm": 0.38854190707206726,
"learning_rate": 1.709624274449584e-06,
"loss": 0.0601,
"step": 831
},
{
"epoch": 0.09508571428571429,
"grad_norm": 0.6414255499839783,
"learning_rate": 1.6901553072567189e-06,
"loss": 0.1349,
"step": 832
},
{
"epoch": 0.0952,
"grad_norm": 0.3700990378856659,
"learning_rate": 1.6707875928990059e-06,
"loss": 0.055,
"step": 833
},
{
"epoch": 0.09531428571428571,
"grad_norm": 1.3736345767974854,
"learning_rate": 1.651521367365936e-06,
"loss": 0.0779,
"step": 834
},
{
"epoch": 0.09542857142857143,
"grad_norm": 0.4473450183868408,
"learning_rate": 1.6323568654103838e-06,
"loss": 0.0577,
"step": 835
},
{
"epoch": 0.09554285714285714,
"grad_norm": 1.072985053062439,
"learning_rate": 1.6132943205457607e-06,
"loss": 0.1208,
"step": 836
},
{
"epoch": 0.09565714285714286,
"grad_norm": 0.3164627254009247,
"learning_rate": 1.5943339650431578e-06,
"loss": 0.0512,
"step": 837
},
{
"epoch": 0.09577142857142858,
"grad_norm": 14.42309284210205,
"learning_rate": 1.5754760299285255e-06,
"loss": 0.3633,
"step": 838
},
{
"epoch": 0.09588571428571428,
"grad_norm": 1.2983229160308838,
"learning_rate": 1.5567207449798517e-06,
"loss": 0.0966,
"step": 839
},
{
"epoch": 0.096,
"grad_norm": 4.02403450012207,
"learning_rate": 1.538068338724361e-06,
"loss": 0.173,
"step": 840
},
{
"epoch": 0.09611428571428571,
"grad_norm": 0.8422157764434814,
"learning_rate": 1.5195190384357405e-06,
"loss": 0.11,
"step": 841
},
{
"epoch": 0.09622857142857143,
"grad_norm": 0.6441839337348938,
"learning_rate": 1.5010730701313626e-06,
"loss": 0.1106,
"step": 842
},
{
"epoch": 0.09634285714285715,
"grad_norm": 0.4187048077583313,
"learning_rate": 1.4827306585695234e-06,
"loss": 0.0481,
"step": 843
},
{
"epoch": 0.09645714285714285,
"grad_norm": 0.8879269957542419,
"learning_rate": 1.4644920272467245e-06,
"loss": 0.1284,
"step": 844
},
{
"epoch": 0.09657142857142857,
"grad_norm": 0.6119964718818665,
"learning_rate": 1.446357398394934e-06,
"loss": 0.113,
"step": 845
},
{
"epoch": 0.09668571428571429,
"grad_norm": 0.37139254808425903,
"learning_rate": 1.4283269929788779e-06,
"loss": 0.0979,
"step": 846
},
{
"epoch": 0.0968,
"grad_norm": 0.429556667804718,
"learning_rate": 1.4104010306933558e-06,
"loss": 0.102,
"step": 847
},
{
"epoch": 0.09691428571428572,
"grad_norm": 0.37817418575286865,
"learning_rate": 1.3925797299605649e-06,
"loss": 0.0978,
"step": 848
},
{
"epoch": 0.09702857142857142,
"grad_norm": 15.693328857421875,
"learning_rate": 1.3748633079274254e-06,
"loss": 0.474,
"step": 849
},
{
"epoch": 0.09714285714285714,
"grad_norm": 0.37196817994117737,
"learning_rate": 1.3572519804629537e-06,
"loss": 0.0489,
"step": 850
},
{
"epoch": 0.09725714285714286,
"grad_norm": 0.8451738953590393,
"learning_rate": 1.339745962155613e-06,
"loss": 0.0634,
"step": 851
},
{
"epoch": 0.09737142857142857,
"grad_norm": 1.088495135307312,
"learning_rate": 1.322345466310717e-06,
"loss": 0.1227,
"step": 852
},
{
"epoch": 0.09748571428571429,
"grad_norm": 0.9944125413894653,
"learning_rate": 1.30505070494781e-06,
"loss": 0.1146,
"step": 853
},
{
"epoch": 0.0976,
"grad_norm": 0.5955567359924316,
"learning_rate": 1.2878618887981064e-06,
"loss": 0.0661,
"step": 854
},
{
"epoch": 0.09771428571428571,
"grad_norm": 0.3421649932861328,
"learning_rate": 1.2707792273019049e-06,
"loss": 0.0531,
"step": 855
},
{
"epoch": 0.09782857142857143,
"grad_norm": 0.8138859868049622,
"learning_rate": 1.2538029286060428e-06,
"loss": 0.0568,
"step": 856
},
{
"epoch": 0.09794285714285714,
"grad_norm": 0.3583981990814209,
"learning_rate": 1.2369331995613664e-06,
"loss": 0.0487,
"step": 857
},
{
"epoch": 0.09805714285714286,
"grad_norm": 0.5639165639877319,
"learning_rate": 1.2201702457201948e-06,
"loss": 0.1036,
"step": 858
},
{
"epoch": 0.09817142857142858,
"grad_norm": 0.2765410244464874,
"learning_rate": 1.2035142713338366e-06,
"loss": 0.0037,
"step": 859
},
{
"epoch": 0.09828571428571428,
"grad_norm": 6.294540882110596,
"learning_rate": 1.1869654793500784e-06,
"loss": 0.125,
"step": 860
},
{
"epoch": 0.0984,
"grad_norm": 26.169923782348633,
"learning_rate": 1.1705240714107301e-06,
"loss": 0.4837,
"step": 861
},
{
"epoch": 0.09851428571428572,
"grad_norm": 0.8702516555786133,
"learning_rate": 1.1541902478491607e-06,
"loss": 0.1067,
"step": 862
},
{
"epoch": 0.09862857142857143,
"grad_norm": 0.5474616289138794,
"learning_rate": 1.1379642076878528e-06,
"loss": 0.0637,
"step": 863
},
{
"epoch": 0.09874285714285715,
"grad_norm": 0.7787386178970337,
"learning_rate": 1.1218461486359878e-06,
"loss": 0.1067,
"step": 864
},
{
"epoch": 0.09885714285714285,
"grad_norm": 0.9886103868484497,
"learning_rate": 1.1058362670870248e-06,
"loss": 0.1071,
"step": 865
},
{
"epoch": 0.09897142857142857,
"grad_norm": 0.6011041402816772,
"learning_rate": 1.0899347581163222e-06,
"loss": 0.0999,
"step": 866
},
{
"epoch": 0.09908571428571429,
"grad_norm": 0.7078867554664612,
"learning_rate": 1.0741418154787443e-06,
"loss": 0.0533,
"step": 867
},
{
"epoch": 0.0992,
"grad_norm": 0.43726450204849243,
"learning_rate": 1.058457631606319e-06,
"loss": 0.0617,
"step": 868
},
{
"epoch": 0.09931428571428572,
"grad_norm": 0.9017318487167358,
"learning_rate": 1.042882397605871e-06,
"loss": 0.099,
"step": 869
},
{
"epoch": 0.09942857142857142,
"grad_norm": 0.5289992690086365,
"learning_rate": 1.0274163032567165e-06,
"loss": 0.0942,
"step": 870
},
{
"epoch": 0.09954285714285714,
"grad_norm": 15.03108024597168,
"learning_rate": 1.012059537008332e-06,
"loss": 0.3946,
"step": 871
},
{
"epoch": 0.09965714285714286,
"grad_norm": 0.5590927004814148,
"learning_rate": 9.968122859780648e-07,
"loss": 0.0522,
"step": 872
},
{
"epoch": 0.09977142857142857,
"grad_norm": 0.7883051633834839,
"learning_rate": 9.816747359488632e-07,
"loss": 0.0577,
"step": 873
},
{
"epoch": 0.09988571428571429,
"grad_norm": 1.0855271816253662,
"learning_rate": 9.666470713669918e-07,
"loss": 0.0671,
"step": 874
},
{
"epoch": 0.1,
"grad_norm": 0.8798184990882874,
"learning_rate": 9.517294753398066e-07,
"loss": 0.1164,
"step": 875
},
{
"epoch": 0.10011428571428571,
"grad_norm": 0.4173676669597626,
"learning_rate": 9.369221296335007e-07,
"loss": 0.0522,
"step": 876
},
{
"epoch": 0.10022857142857143,
"grad_norm": 0.3149450123310089,
"learning_rate": 9.222252146709143e-07,
"loss": 0.0461,
"step": 877
},
{
"epoch": 0.10034285714285714,
"grad_norm": 0.5355912446975708,
"learning_rate": 9.076389095293148e-07,
"loss": 0.1009,
"step": 878
},
{
"epoch": 0.10045714285714286,
"grad_norm": 0.3526480197906494,
"learning_rate": 8.931633919382299e-07,
"loss": 0.0519,
"step": 879
},
{
"epoch": 0.10057142857142858,
"grad_norm": 0.30637577176094055,
"learning_rate": 8.787988382772705e-07,
"loss": 0.0528,
"step": 880
},
{
"epoch": 0.10068571428571428,
"grad_norm": 0.5830509662628174,
"learning_rate": 8.645454235739903e-07,
"loss": 0.0657,
"step": 881
},
{
"epoch": 0.1008,
"grad_norm": 1.7157456874847412,
"learning_rate": 8.504033215017527e-07,
"loss": 0.139,
"step": 882
},
{
"epoch": 0.10091428571428572,
"grad_norm": 0.7653803825378418,
"learning_rate": 8.363727043776037e-07,
"loss": 0.1097,
"step": 883
},
{
"epoch": 0.10102857142857143,
"grad_norm": 1.1827641725540161,
"learning_rate": 8.224537431601886e-07,
"loss": 0.1272,
"step": 884
},
{
"epoch": 0.10114285714285715,
"grad_norm": 0.44600731134414673,
"learning_rate": 8.086466074476562e-07,
"loss": 0.0523,
"step": 885
},
{
"epoch": 0.10125714285714285,
"grad_norm": 1.7456201314926147,
"learning_rate": 7.949514654755963e-07,
"loss": 0.1368,
"step": 886
},
{
"epoch": 0.10137142857142857,
"grad_norm": 0.0900225043296814,
"learning_rate": 7.81368484114996e-07,
"loss": 0.0031,
"step": 887
},
{
"epoch": 0.10148571428571429,
"grad_norm": 1.5314533710479736,
"learning_rate": 7.678978288701911e-07,
"loss": 0.1307,
"step": 888
},
{
"epoch": 0.1016,
"grad_norm": 0.4106394946575165,
"learning_rate": 7.545396638768698e-07,
"loss": 0.0439,
"step": 889
},
{
"epoch": 0.10171428571428572,
"grad_norm": 0.33473098278045654,
"learning_rate": 7.412941519000527e-07,
"loss": 0.0511,
"step": 890
},
{
"epoch": 0.10182857142857143,
"grad_norm": 1.11957848072052,
"learning_rate": 7.281614543321269e-07,
"loss": 0.1068,
"step": 891
},
{
"epoch": 0.10194285714285714,
"grad_norm": 0.5046252608299255,
"learning_rate": 7.151417311908648e-07,
"loss": 0.0526,
"step": 892
},
{
"epoch": 0.10205714285714286,
"grad_norm": 0.8343473076820374,
"learning_rate": 7.022351411174866e-07,
"loss": 0.1051,
"step": 893
},
{
"epoch": 0.10217142857142857,
"grad_norm": 0.58259117603302,
"learning_rate": 6.894418413747183e-07,
"loss": 0.0562,
"step": 894
},
{
"epoch": 0.10228571428571429,
"grad_norm": 0.8305698037147522,
"learning_rate": 6.767619878448783e-07,
"loss": 0.1011,
"step": 895
},
{
"epoch": 0.1024,
"grad_norm": 1.1420916318893433,
"learning_rate": 6.641957350279838e-07,
"loss": 0.1125,
"step": 896
},
{
"epoch": 0.10251428571428571,
"grad_norm": 0.761816143989563,
"learning_rate": 6.517432360398556e-07,
"loss": 0.061,
"step": 897
},
{
"epoch": 0.10262857142857143,
"grad_norm": 0.5329399108886719,
"learning_rate": 6.394046426102673e-07,
"loss": 0.0547,
"step": 898
},
{
"epoch": 0.10274285714285715,
"grad_norm": 0.8240087032318115,
"learning_rate": 6.271801050810856e-07,
"loss": 0.127,
"step": 899
},
{
"epoch": 0.10285714285714286,
"grad_norm": 0.8073368072509766,
"learning_rate": 6.150697724044407e-07,
"loss": 0.1032,
"step": 900
},
{
"epoch": 0.10297142857142857,
"grad_norm": 0.48127835988998413,
"learning_rate": 6.030737921409169e-07,
"loss": 0.047,
"step": 901
},
{
"epoch": 0.10308571428571428,
"grad_norm": 3.648961305618286,
"learning_rate": 5.911923104577455e-07,
"loss": 0.1384,
"step": 902
},
{
"epoch": 0.1032,
"grad_norm": 0.6189379692077637,
"learning_rate": 5.794254721270331e-07,
"loss": 0.058,
"step": 903
},
{
"epoch": 0.10331428571428572,
"grad_norm": 0.4216036796569824,
"learning_rate": 5.677734205239904e-07,
"loss": 0.0457,
"step": 904
},
{
"epoch": 0.10342857142857143,
"grad_norm": 0.3244432508945465,
"learning_rate": 5.562362976251901e-07,
"loss": 0.051,
"step": 905
},
{
"epoch": 0.10354285714285714,
"grad_norm": 1.1427147388458252,
"learning_rate": 5.448142440068316e-07,
"loss": 0.0607,
"step": 906
},
{
"epoch": 0.10365714285714285,
"grad_norm": 1.9026908874511719,
"learning_rate": 5.335073988430373e-07,
"loss": 0.1094,
"step": 907
},
{
"epoch": 0.10377142857142857,
"grad_norm": 0.41159185767173767,
"learning_rate": 5.223158999041444e-07,
"loss": 0.0545,
"step": 908
},
{
"epoch": 0.10388571428571429,
"grad_norm": 0.32340720295906067,
"learning_rate": 5.112398835550348e-07,
"loss": 0.0461,
"step": 909
},
{
"epoch": 0.104,
"grad_norm": 1.4293159246444702,
"learning_rate": 5.002794847534765e-07,
"loss": 0.1422,
"step": 910
},
{
"epoch": 0.10411428571428571,
"grad_norm": 0.548450767993927,
"learning_rate": 4.894348370484648e-07,
"loss": 0.0555,
"step": 911
},
{
"epoch": 0.10422857142857143,
"grad_norm": 0.6607321500778198,
"learning_rate": 4.787060725786141e-07,
"loss": 0.1172,
"step": 912
},
{
"epoch": 0.10434285714285714,
"grad_norm": 0.4475196599960327,
"learning_rate": 4.6809332207053083e-07,
"loss": 0.0524,
"step": 913
},
{
"epoch": 0.10445714285714286,
"grad_norm": 0.29807189106941223,
"learning_rate": 4.575967148372318e-07,
"loss": 0.048,
"step": 914
},
{
"epoch": 0.10457142857142857,
"grad_norm": 0.5986025929450989,
"learning_rate": 4.4721637877656377e-07,
"loss": 0.0551,
"step": 915
},
{
"epoch": 0.10468571428571428,
"grad_norm": 0.892488420009613,
"learning_rate": 4.3695244036964567e-07,
"loss": 0.1236,
"step": 916
},
{
"epoch": 0.1048,
"grad_norm": 0.9700779318809509,
"learning_rate": 4.268050246793276e-07,
"loss": 0.1262,
"step": 917
},
{
"epoch": 0.10491428571428571,
"grad_norm": 1.149242639541626,
"learning_rate": 4.167742553486676e-07,
"loss": 0.0646,
"step": 918
},
{
"epoch": 0.10502857142857143,
"grad_norm": 32.788822174072266,
"learning_rate": 4.068602545994249e-07,
"loss": 0.3542,
"step": 919
},
{
"epoch": 0.10514285714285715,
"grad_norm": 0.6324719190597534,
"learning_rate": 3.9706314323056936e-07,
"loss": 0.0605,
"step": 920
},
{
"epoch": 0.10525714285714285,
"grad_norm": 0.7762057185173035,
"learning_rate": 3.8738304061681107e-07,
"loss": 0.1008,
"step": 921
},
{
"epoch": 0.10537142857142857,
"grad_norm": 1.0244741439819336,
"learning_rate": 3.7782006470714614e-07,
"loss": 0.1174,
"step": 922
},
{
"epoch": 0.10548571428571428,
"grad_norm": 0.5968824625015259,
"learning_rate": 3.68374332023419e-07,
"loss": 0.0994,
"step": 923
},
{
"epoch": 0.1056,
"grad_norm": 0.3292442858219147,
"learning_rate": 3.590459576589e-07,
"loss": 0.0462,
"step": 924
},
{
"epoch": 0.10571428571428572,
"grad_norm": 1.3567745685577393,
"learning_rate": 3.498350552768859e-07,
"loss": 0.0953,
"step": 925
},
{
"epoch": 0.10582857142857142,
"grad_norm": 2.1283113956451416,
"learning_rate": 3.4074173710931804e-07,
"loss": 0.1185,
"step": 926
},
{
"epoch": 0.10594285714285714,
"grad_norm": 0.9257760643959045,
"learning_rate": 3.3176611395540625e-07,
"loss": 0.0573,
"step": 927
},
{
"epoch": 0.10605714285714286,
"grad_norm": 0.5500800013542175,
"learning_rate": 3.2290829518028867e-07,
"loss": 0.0915,
"step": 928
},
{
"epoch": 0.10617142857142857,
"grad_norm": 0.6672600507736206,
"learning_rate": 3.1416838871368925e-07,
"loss": 0.1121,
"step": 929
},
{
"epoch": 0.10628571428571429,
"grad_norm": 0.43164995312690735,
"learning_rate": 3.0554650104861137e-07,
"loss": 0.0937,
"step": 930
},
{
"epoch": 0.1064,
"grad_norm": 0.2542230188846588,
"learning_rate": 2.970427372400353e-07,
"loss": 0.0521,
"step": 931
},
{
"epoch": 0.10651428571428571,
"grad_norm": 1.5962998867034912,
"learning_rate": 2.8865720090364037e-07,
"loss": 0.1305,
"step": 932
},
{
"epoch": 0.10662857142857143,
"grad_norm": 0.9495799541473389,
"learning_rate": 2.8038999421453827e-07,
"loss": 0.1103,
"step": 933
},
{
"epoch": 0.10674285714285714,
"grad_norm": 1.1988343000411987,
"learning_rate": 2.7224121790603517e-07,
"loss": 0.1105,
"step": 934
},
{
"epoch": 0.10685714285714286,
"grad_norm": 14.932929992675781,
"learning_rate": 2.6421097126839714e-07,
"loss": 0.24,
"step": 935
},
{
"epoch": 0.10697142857142856,
"grad_norm": 0.9639659523963928,
"learning_rate": 2.5629935214764866e-07,
"loss": 0.0762,
"step": 936
},
{
"epoch": 0.10708571428571428,
"grad_norm": 0.617256760597229,
"learning_rate": 2.4850645694436736e-07,
"loss": 0.1135,
"step": 937
},
{
"epoch": 0.1072,
"grad_norm": 1.5766311883926392,
"learning_rate": 2.4083238061252565e-07,
"loss": 0.1265,
"step": 938
},
{
"epoch": 0.10731428571428571,
"grad_norm": 0.03188295662403107,
"learning_rate": 2.332772166583208e-07,
"loss": 0.0014,
"step": 939
},
{
"epoch": 0.10742857142857143,
"grad_norm": 0.7548657059669495,
"learning_rate": 2.2584105713904126e-07,
"loss": 0.0995,
"step": 940
},
{
"epoch": 0.10754285714285715,
"grad_norm": 0.6812636852264404,
"learning_rate": 2.1852399266194312e-07,
"loss": 0.0978,
"step": 941
},
{
"epoch": 0.10765714285714285,
"grad_norm": 0.8354098796844482,
"learning_rate": 2.1132611238315004e-07,
"loss": 0.058,
"step": 942
},
{
"epoch": 0.10777142857142857,
"grad_norm": 0.39750880002975464,
"learning_rate": 2.0424750400655947e-07,
"loss": 0.0534,
"step": 943
},
{
"epoch": 0.10788571428571428,
"grad_norm": 14.927691459655762,
"learning_rate": 1.9728825378278248e-07,
"loss": 0.3998,
"step": 944
},
{
"epoch": 0.108,
"grad_norm": 1.135928750038147,
"learning_rate": 1.9044844650808468e-07,
"loss": 0.1122,
"step": 945
},
{
"epoch": 0.10811428571428572,
"grad_norm": 0.7230623960494995,
"learning_rate": 1.8372816552336025e-07,
"loss": 0.0557,
"step": 946
},
{
"epoch": 0.10822857142857142,
"grad_norm": 0.02265091985464096,
"learning_rate": 1.7712749271311392e-07,
"loss": 0.0009,
"step": 947
},
{
"epoch": 0.10834285714285714,
"grad_norm": 0.9828023314476013,
"learning_rate": 1.706465085044584e-07,
"loss": 0.1148,
"step": 948
},
{
"epoch": 0.10845714285714286,
"grad_norm": 0.5630030035972595,
"learning_rate": 1.6428529186614195e-07,
"loss": 0.1027,
"step": 949
},
{
"epoch": 0.10857142857142857,
"grad_norm": 0.903812050819397,
"learning_rate": 1.580439203075812e-07,
"loss": 0.0649,
"step": 950
},
{
"epoch": 0.10868571428571429,
"grad_norm": 0.4727358818054199,
"learning_rate": 1.519224698779198e-07,
"loss": 0.095,
"step": 951
},
{
"epoch": 0.1088,
"grad_norm": 1.1214702129364014,
"learning_rate": 1.4592101516509916e-07,
"loss": 0.105,
"step": 952
},
{
"epoch": 0.10891428571428571,
"grad_norm": 0.3436620235443115,
"learning_rate": 1.400396292949513e-07,
"loss": 0.0472,
"step": 953
},
{
"epoch": 0.10902857142857143,
"grad_norm": 0.896814227104187,
"learning_rate": 1.3427838393030634e-07,
"loss": 0.1067,
"step": 954
},
{
"epoch": 0.10914285714285714,
"grad_norm": 1.1149619817733765,
"learning_rate": 1.2863734927012094e-07,
"loss": 0.0591,
"step": 955
},
{
"epoch": 0.10925714285714286,
"grad_norm": 0.4961557388305664,
"learning_rate": 1.231165940486234e-07,
"loss": 0.059,
"step": 956
},
{
"epoch": 0.10937142857142858,
"grad_norm": 0.3222205936908722,
"learning_rate": 1.1771618553447217e-07,
"loss": 0.0493,
"step": 957
},
{
"epoch": 0.10948571428571428,
"grad_norm": 0.2927284240722656,
"learning_rate": 1.1243618952994195e-07,
"loss": 0.0461,
"step": 958
},
{
"epoch": 0.1096,
"grad_norm": 0.8068863153457642,
"learning_rate": 1.0727667037011668e-07,
"loss": 0.0725,
"step": 959
},
{
"epoch": 0.10971428571428571,
"grad_norm": 0.7345075011253357,
"learning_rate": 1.0223769092211012e-07,
"loss": 0.117,
"step": 960
},
{
"epoch": 0.10982857142857143,
"grad_norm": 0.6134506464004517,
"learning_rate": 9.731931258429638e-08,
"loss": 0.0541,
"step": 961
},
{
"epoch": 0.10994285714285715,
"grad_norm": 0.7149043083190918,
"learning_rate": 9.252159528556404e-08,
"loss": 0.1058,
"step": 962
},
{
"epoch": 0.11005714285714285,
"grad_norm": 0.6467493176460266,
"learning_rate": 8.784459748458318e-08,
"loss": 0.1027,
"step": 963
},
{
"epoch": 0.11017142857142857,
"grad_norm": 1.1258658170700073,
"learning_rate": 8.328837616909612e-08,
"loss": 0.0533,
"step": 964
},
{
"epoch": 0.11028571428571429,
"grad_norm": 0.4852413833141327,
"learning_rate": 7.885298685522235e-08,
"loss": 0.0623,
"step": 965
},
{
"epoch": 0.1104,
"grad_norm": 0.2948656380176544,
"learning_rate": 7.453848358678018e-08,
"loss": 0.0442,
"step": 966
},
{
"epoch": 0.11051428571428572,
"grad_norm": 0.7246602773666382,
"learning_rate": 7.034491893463059e-08,
"loss": 0.1123,
"step": 967
},
{
"epoch": 0.11062857142857142,
"grad_norm": 1.0135889053344727,
"learning_rate": 6.627234399603554e-08,
"loss": 0.1191,
"step": 968
},
{
"epoch": 0.11074285714285714,
"grad_norm": 1.0500082969665527,
"learning_rate": 6.232080839403631e-08,
"loss": 0.06,
"step": 969
},
{
"epoch": 0.11085714285714286,
"grad_norm": 1.5601297616958618,
"learning_rate": 5.849036027684607e-08,
"loss": 0.0662,
"step": 970
},
{
"epoch": 0.11097142857142857,
"grad_norm": 0.5829685926437378,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.1072,
"step": 971
},
{
"epoch": 0.11108571428571429,
"grad_norm": 0.4342767894268036,
"learning_rate": 5.119291171211793e-08,
"loss": 0.1074,
"step": 972
},
{
"epoch": 0.1112,
"grad_norm": 1.2641159296035767,
"learning_rate": 4.772600018168816e-08,
"loss": 0.1128,
"step": 973
},
{
"epoch": 0.11131428571428571,
"grad_norm": 0.675849974155426,
"learning_rate": 4.438035396920004e-08,
"loss": 0.1174,
"step": 974
},
{
"epoch": 0.11142857142857143,
"grad_norm": 0.9253007173538208,
"learning_rate": 4.115601384029666e-08,
"loss": 0.0635,
"step": 975
},
{
"epoch": 0.11154285714285714,
"grad_norm": 0.6775985956192017,
"learning_rate": 3.805301908254455e-08,
"loss": 0.0999,
"step": 976
},
{
"epoch": 0.11165714285714286,
"grad_norm": 11.044492721557617,
"learning_rate": 3.50714075049563e-08,
"loss": 0.2332,
"step": 977
},
{
"epoch": 0.11177142857142858,
"grad_norm": 0.4669469892978668,
"learning_rate": 3.22112154375287e-08,
"loss": 0.0512,
"step": 978
},
{
"epoch": 0.11188571428571428,
"grad_norm": 0.4130972623825073,
"learning_rate": 2.947247773079753e-08,
"loss": 0.047,
"step": 979
},
{
"epoch": 0.112,
"grad_norm": 0.8283016085624695,
"learning_rate": 2.6855227755419046e-08,
"loss": 0.0552,
"step": 980
},
{
"epoch": 0.11211428571428571,
"grad_norm": 0.8060480952262878,
"learning_rate": 2.4359497401758026e-08,
"loss": 0.0953,
"step": 981
},
{
"epoch": 0.11222857142857143,
"grad_norm": 0.5043060779571533,
"learning_rate": 2.1985317079500358e-08,
"loss": 0.0398,
"step": 982
},
{
"epoch": 0.11234285714285715,
"grad_norm": 0.673374593257904,
"learning_rate": 1.973271571728441e-08,
"loss": 0.0509,
"step": 983
},
{
"epoch": 0.11245714285714285,
"grad_norm": 0.5060809254646301,
"learning_rate": 1.7601720762346895e-08,
"loss": 0.0533,
"step": 984
},
{
"epoch": 0.11257142857142857,
"grad_norm": 3.078909397125244,
"learning_rate": 1.5592358180189782e-08,
"loss": 0.2289,
"step": 985
},
{
"epoch": 0.11268571428571429,
"grad_norm": 1.1784045696258545,
"learning_rate": 1.370465245426167e-08,
"loss": 0.1519,
"step": 986
},
{
"epoch": 0.1128,
"grad_norm": 1.364969253540039,
"learning_rate": 1.1938626585660252e-08,
"loss": 0.1148,
"step": 987
},
{
"epoch": 0.11291428571428572,
"grad_norm": 0.9849162101745605,
"learning_rate": 1.0294302092853647e-08,
"loss": 0.1052,
"step": 988
},
{
"epoch": 0.11302857142857142,
"grad_norm": 0.2593817412853241,
"learning_rate": 8.771699011416169e-09,
"loss": 0.0491,
"step": 989
},
{
"epoch": 0.11314285714285714,
"grad_norm": 1.879010558128357,
"learning_rate": 7.370835893788508e-09,
"loss": 0.1678,
"step": 990
},
{
"epoch": 0.11325714285714286,
"grad_norm": 0.9571077823638916,
"learning_rate": 6.091729809042379e-09,
"loss": 0.1217,
"step": 991
},
{
"epoch": 0.11337142857142857,
"grad_norm": 1.00277578830719,
"learning_rate": 4.9343963426840006e-09,
"loss": 0.1071,
"step": 992
},
{
"epoch": 0.11348571428571429,
"grad_norm": 0.23549872636795044,
"learning_rate": 3.898849596456477e-09,
"loss": 0.0439,
"step": 993
},
{
"epoch": 0.1136,
"grad_norm": 0.448467493057251,
"learning_rate": 2.9851021881688314e-09,
"loss": 0.0499,
"step": 994
},
{
"epoch": 0.11371428571428571,
"grad_norm": 0.2983313798904419,
"learning_rate": 2.193165251545004e-09,
"loss": 0.0408,
"step": 995
},
{
"epoch": 0.11382857142857143,
"grad_norm": 0.5201529264450073,
"learning_rate": 1.5230484360873043e-09,
"loss": 0.1005,
"step": 996
},
{
"epoch": 0.11394285714285714,
"grad_norm": 21.895347595214844,
"learning_rate": 9.74759906957612e-10,
"loss": 1.033,
"step": 997
},
{
"epoch": 0.11405714285714286,
"grad_norm": 2.064412832260132,
"learning_rate": 5.483063448785686e-10,
"loss": 0.138,
"step": 998
},
{
"epoch": 0.11417142857142858,
"grad_norm": 1.3795288801193237,
"learning_rate": 2.436929460525317e-10,
"loss": 0.1468,
"step": 999
},
{
"epoch": 0.11428571428571428,
"grad_norm": 0.28793764114379883,
"learning_rate": 6.092342209607083e-11,
"loss": 0.0478,
"step": 1000
}
],
"logging_steps": 1,
"max_steps": 1000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}