DeepSeek-Math-7B-instruct / trainer_state.json
Zzh-tju's picture
Upload DeepSeek-Math-7B-instruct_epoch2_5e-6
f2a72e9 verified
Raw
History Blame Contribute Delete
356 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 2048,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.000976950083956648,
"grad_norm": 17.968935012817383,
"learning_rate": 0.0,
"loss": 1.2799,
"step": 1
},
{
"epoch": 0.001953900167913296,
"grad_norm": 22.7800235748291,
"learning_rate": 8.064516129032259e-08,
"loss": 1.2072,
"step": 2
},
{
"epoch": 0.0029308502518699433,
"grad_norm": 16.298179626464844,
"learning_rate": 1.6129032258064518e-07,
"loss": 1.411,
"step": 3
},
{
"epoch": 0.003907800335826592,
"grad_norm": 13.078301429748535,
"learning_rate": 2.4193548387096775e-07,
"loss": 1.1789,
"step": 4
},
{
"epoch": 0.00488475041978324,
"grad_norm": 12.133336067199707,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.0316,
"step": 5
},
{
"epoch": 0.005861700503739887,
"grad_norm": 15.982775688171387,
"learning_rate": 4.032258064516129e-07,
"loss": 1.1839,
"step": 6
},
{
"epoch": 0.006838650587696535,
"grad_norm": 13.022838592529297,
"learning_rate": 4.838709677419355e-07,
"loss": 1.1056,
"step": 7
},
{
"epoch": 0.007815600671653183,
"grad_norm": 16.860830307006836,
"learning_rate": 5.645161290322581e-07,
"loss": 1.2288,
"step": 8
},
{
"epoch": 0.008792550755609831,
"grad_norm": 11.351203918457031,
"learning_rate": 6.451612903225807e-07,
"loss": 1.2214,
"step": 9
},
{
"epoch": 0.00976950083956648,
"grad_norm": 15.175298690795898,
"learning_rate": 7.258064516129033e-07,
"loss": 1.421,
"step": 10
},
{
"epoch": 0.010746450923523125,
"grad_norm": 19.758380889892578,
"learning_rate": 8.064516129032258e-07,
"loss": 1.401,
"step": 11
},
{
"epoch": 0.011723401007479773,
"grad_norm": 16.22666358947754,
"learning_rate": 8.870967741935485e-07,
"loss": 1.0867,
"step": 12
},
{
"epoch": 0.012700351091436421,
"grad_norm": 15.286944389343262,
"learning_rate": 9.67741935483871e-07,
"loss": 1.4673,
"step": 13
},
{
"epoch": 0.01367730117539307,
"grad_norm": 16.071199417114258,
"learning_rate": 1.0483870967741936e-06,
"loss": 1.2807,
"step": 14
},
{
"epoch": 0.014654251259349717,
"grad_norm": 12.26354694366455,
"learning_rate": 1.1290322580645162e-06,
"loss": 1.1647,
"step": 15
},
{
"epoch": 0.015631201343306367,
"grad_norm": 15.69483470916748,
"learning_rate": 1.2096774193548388e-06,
"loss": 1.1656,
"step": 16
},
{
"epoch": 0.016608151427263015,
"grad_norm": 19.166576385498047,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.3481,
"step": 17
},
{
"epoch": 0.017585101511219663,
"grad_norm": 15.776891708374023,
"learning_rate": 1.3709677419354838e-06,
"loss": 1.3953,
"step": 18
},
{
"epoch": 0.01856205159517631,
"grad_norm": 14.586052894592285,
"learning_rate": 1.4516129032258066e-06,
"loss": 1.3277,
"step": 19
},
{
"epoch": 0.01953900167913296,
"grad_norm": 17.343685150146484,
"learning_rate": 1.5322580645161292e-06,
"loss": 1.0952,
"step": 20
},
{
"epoch": 0.020515951763089603,
"grad_norm": 14.847457885742188,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.2949,
"step": 21
},
{
"epoch": 0.02149290184704625,
"grad_norm": 14.039813041687012,
"learning_rate": 1.6935483870967742e-06,
"loss": 1.1587,
"step": 22
},
{
"epoch": 0.0224698519310029,
"grad_norm": 14.10567569732666,
"learning_rate": 1.774193548387097e-06,
"loss": 1.1669,
"step": 23
},
{
"epoch": 0.023446802014959547,
"grad_norm": 12.816205978393555,
"learning_rate": 1.8548387096774196e-06,
"loss": 1.114,
"step": 24
},
{
"epoch": 0.024423752098916195,
"grad_norm": 18.03955078125,
"learning_rate": 1.935483870967742e-06,
"loss": 1.7148,
"step": 25
},
{
"epoch": 0.025400702182872843,
"grad_norm": 11.711819648742676,
"learning_rate": 2.0161290322580646e-06,
"loss": 1.1004,
"step": 26
},
{
"epoch": 0.02637765226682949,
"grad_norm": 11.962332725524902,
"learning_rate": 2.096774193548387e-06,
"loss": 1.0374,
"step": 27
},
{
"epoch": 0.02735460235078614,
"grad_norm": 20.176681518554688,
"learning_rate": 2.17741935483871e-06,
"loss": 1.2643,
"step": 28
},
{
"epoch": 0.028331552434742786,
"grad_norm": 10.272699356079102,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.86,
"step": 29
},
{
"epoch": 0.029308502518699434,
"grad_norm": 12.97459888458252,
"learning_rate": 2.338709677419355e-06,
"loss": 1.1146,
"step": 30
},
{
"epoch": 0.030285452602656082,
"grad_norm": 17.359928131103516,
"learning_rate": 2.4193548387096776e-06,
"loss": 1.1275,
"step": 31
},
{
"epoch": 0.031262402686612734,
"grad_norm": 10.485804557800293,
"learning_rate": 2.5e-06,
"loss": 1.0539,
"step": 32
},
{
"epoch": 0.03223935277056938,
"grad_norm": 12.242683410644531,
"learning_rate": 2.580645161290323e-06,
"loss": 1.0711,
"step": 33
},
{
"epoch": 0.03321630285452603,
"grad_norm": 10.652597427368164,
"learning_rate": 2.6612903225806454e-06,
"loss": 0.827,
"step": 34
},
{
"epoch": 0.034193252938482674,
"grad_norm": 11.75671100616455,
"learning_rate": 2.7419354838709676e-06,
"loss": 0.9183,
"step": 35
},
{
"epoch": 0.035170203022439325,
"grad_norm": 11.894708633422852,
"learning_rate": 2.822580645161291e-06,
"loss": 0.8568,
"step": 36
},
{
"epoch": 0.03614715310639597,
"grad_norm": 12.387930870056152,
"learning_rate": 2.903225806451613e-06,
"loss": 0.948,
"step": 37
},
{
"epoch": 0.03712410319035262,
"grad_norm": 23.549697875976562,
"learning_rate": 2.983870967741936e-06,
"loss": 1.0557,
"step": 38
},
{
"epoch": 0.038101053274309266,
"grad_norm": 10.79415225982666,
"learning_rate": 3.0645161290322584e-06,
"loss": 0.8586,
"step": 39
},
{
"epoch": 0.03907800335826592,
"grad_norm": 11.130952835083008,
"learning_rate": 3.145161290322581e-06,
"loss": 0.7007,
"step": 40
},
{
"epoch": 0.04005495344222256,
"grad_norm": 14.665740966796875,
"learning_rate": 3.225806451612903e-06,
"loss": 0.8159,
"step": 41
},
{
"epoch": 0.041031903526179206,
"grad_norm": 9.616032600402832,
"learning_rate": 3.306451612903226e-06,
"loss": 0.6988,
"step": 42
},
{
"epoch": 0.04200885361013586,
"grad_norm": 12.437187194824219,
"learning_rate": 3.3870967741935484e-06,
"loss": 0.6888,
"step": 43
},
{
"epoch": 0.0429858036940925,
"grad_norm": 7.805335998535156,
"learning_rate": 3.4677419354838714e-06,
"loss": 0.564,
"step": 44
},
{
"epoch": 0.04396275377804915,
"grad_norm": 7.527621269226074,
"learning_rate": 3.548387096774194e-06,
"loss": 0.5619,
"step": 45
},
{
"epoch": 0.0449397038620058,
"grad_norm": 10.622644424438477,
"learning_rate": 3.6290322580645166e-06,
"loss": 0.6401,
"step": 46
},
{
"epoch": 0.04591665394596245,
"grad_norm": 9.222088813781738,
"learning_rate": 3.7096774193548392e-06,
"loss": 0.5901,
"step": 47
},
{
"epoch": 0.04689360402991909,
"grad_norm": 7.2194366455078125,
"learning_rate": 3.7903225806451614e-06,
"loss": 0.5456,
"step": 48
},
{
"epoch": 0.047870554113875745,
"grad_norm": 10.335058212280273,
"learning_rate": 3.870967741935484e-06,
"loss": 0.5238,
"step": 49
},
{
"epoch": 0.04884750419783239,
"grad_norm": 7.783710956573486,
"learning_rate": 3.951612903225807e-06,
"loss": 0.4556,
"step": 50
},
{
"epoch": 0.04982445428178904,
"grad_norm": 7.917117118835449,
"learning_rate": 4.032258064516129e-06,
"loss": 0.4581,
"step": 51
},
{
"epoch": 0.050801404365745685,
"grad_norm": 6.636364459991455,
"learning_rate": 4.112903225806452e-06,
"loss": 0.5275,
"step": 52
},
{
"epoch": 0.051778354449702337,
"grad_norm": 6.040541172027588,
"learning_rate": 4.193548387096774e-06,
"loss": 0.4448,
"step": 53
},
{
"epoch": 0.05275530453365898,
"grad_norm": 6.1212263107299805,
"learning_rate": 4.274193548387097e-06,
"loss": 0.3961,
"step": 54
},
{
"epoch": 0.05373225461761563,
"grad_norm": 6.664571285247803,
"learning_rate": 4.35483870967742e-06,
"loss": 0.3631,
"step": 55
},
{
"epoch": 0.05470920470157228,
"grad_norm": 4.888693809509277,
"learning_rate": 4.435483870967742e-06,
"loss": 0.3922,
"step": 56
},
{
"epoch": 0.05568615478552893,
"grad_norm": 5.145596027374268,
"learning_rate": 4.516129032258065e-06,
"loss": 0.3449,
"step": 57
},
{
"epoch": 0.05666310486948557,
"grad_norm": 6.826282978057861,
"learning_rate": 4.596774193548387e-06,
"loss": 0.3537,
"step": 58
},
{
"epoch": 0.057640054953442224,
"grad_norm": 4.439844131469727,
"learning_rate": 4.67741935483871e-06,
"loss": 0.3876,
"step": 59
},
{
"epoch": 0.05861700503739887,
"grad_norm": 4.745813369750977,
"learning_rate": 4.758064516129033e-06,
"loss": 0.3219,
"step": 60
},
{
"epoch": 0.05959395512135552,
"grad_norm": 6.4229736328125,
"learning_rate": 4.838709677419355e-06,
"loss": 0.3467,
"step": 61
},
{
"epoch": 0.060570905205312164,
"grad_norm": 5.087151527404785,
"learning_rate": 4.919354838709678e-06,
"loss": 0.3007,
"step": 62
},
{
"epoch": 0.061547855289268816,
"grad_norm": 4.666023254394531,
"learning_rate": 5e-06,
"loss": 0.3669,
"step": 63
},
{
"epoch": 0.06252480537322547,
"grad_norm": 3.0126302242279053,
"learning_rate": 4.999996872112104e-06,
"loss": 0.2961,
"step": 64
},
{
"epoch": 0.0635017554571821,
"grad_norm": 3.9373786449432373,
"learning_rate": 4.999987488456242e-06,
"loss": 0.3356,
"step": 65
},
{
"epoch": 0.06447870554113876,
"grad_norm": 3.22776460647583,
"learning_rate": 4.999971849055896e-06,
"loss": 0.2988,
"step": 66
},
{
"epoch": 0.06545565562509541,
"grad_norm": 5.045942783355713,
"learning_rate": 4.9999499539501995e-06,
"loss": 0.29,
"step": 67
},
{
"epoch": 0.06643260570905206,
"grad_norm": 4.188644886016846,
"learning_rate": 4.9999218031939415e-06,
"loss": 0.29,
"step": 68
},
{
"epoch": 0.0674095557930087,
"grad_norm": 3.267216682434082,
"learning_rate": 4.999887396857563e-06,
"loss": 0.2885,
"step": 69
},
{
"epoch": 0.06838650587696535,
"grad_norm": 3.768075466156006,
"learning_rate": 4.999846735027161e-06,
"loss": 0.3026,
"step": 70
},
{
"epoch": 0.069363455960922,
"grad_norm": 3.0770761966705322,
"learning_rate": 4.999799817804482e-06,
"loss": 0.2781,
"step": 71
},
{
"epoch": 0.07034040604487865,
"grad_norm": 3.6093029975891113,
"learning_rate": 4.9997466453069295e-06,
"loss": 0.2753,
"step": 72
},
{
"epoch": 0.07131735612883529,
"grad_norm": 4.61515474319458,
"learning_rate": 4.999687217667556e-06,
"loss": 0.2969,
"step": 73
},
{
"epoch": 0.07229430621279194,
"grad_norm": 2.091836452484131,
"learning_rate": 4.999621535035068e-06,
"loss": 0.247,
"step": 74
},
{
"epoch": 0.07327125629674859,
"grad_norm": 2.735579013824463,
"learning_rate": 4.999549597573825e-06,
"loss": 0.2728,
"step": 75
},
{
"epoch": 0.07424820638070524,
"grad_norm": 3.6851677894592285,
"learning_rate": 4.999471405463837e-06,
"loss": 0.2785,
"step": 76
},
{
"epoch": 0.07522515646466188,
"grad_norm": 2.23905086517334,
"learning_rate": 4.999386958900763e-06,
"loss": 0.233,
"step": 77
},
{
"epoch": 0.07620210654861853,
"grad_norm": 3.8556435108184814,
"learning_rate": 4.999296258095916e-06,
"loss": 0.2777,
"step": 78
},
{
"epoch": 0.07717905663257518,
"grad_norm": 3.1029233932495117,
"learning_rate": 4.9991993032762565e-06,
"loss": 0.2318,
"step": 79
},
{
"epoch": 0.07815600671653183,
"grad_norm": 3.8344168663024902,
"learning_rate": 4.999096094684397e-06,
"loss": 0.2938,
"step": 80
},
{
"epoch": 0.07913295680048847,
"grad_norm": 1.9346352815628052,
"learning_rate": 4.998986632578596e-06,
"loss": 0.2201,
"step": 81
},
{
"epoch": 0.08010990688444512,
"grad_norm": 2.4821383953094482,
"learning_rate": 4.998870917232763e-06,
"loss": 0.2349,
"step": 82
},
{
"epoch": 0.08108685696840177,
"grad_norm": 2.1198668479919434,
"learning_rate": 4.998748948936452e-06,
"loss": 0.252,
"step": 83
},
{
"epoch": 0.08206380705235841,
"grad_norm": 2.3148319721221924,
"learning_rate": 4.9986207279948675e-06,
"loss": 0.246,
"step": 84
},
{
"epoch": 0.08304075713631506,
"grad_norm": 2.0130088329315186,
"learning_rate": 4.998486254728856e-06,
"loss": 0.2426,
"step": 85
},
{
"epoch": 0.08401770722027171,
"grad_norm": 2.2017133235931396,
"learning_rate": 4.998345529474913e-06,
"loss": 0.2142,
"step": 86
},
{
"epoch": 0.08499465730422837,
"grad_norm": 3.0237224102020264,
"learning_rate": 4.998198552585177e-06,
"loss": 0.2565,
"step": 87
},
{
"epoch": 0.085971607388185,
"grad_norm": 2.1265227794647217,
"learning_rate": 4.998045324427428e-06,
"loss": 0.2327,
"step": 88
},
{
"epoch": 0.08694855747214165,
"grad_norm": 2.4763548374176025,
"learning_rate": 4.997885845385092e-06,
"loss": 0.2833,
"step": 89
},
{
"epoch": 0.0879255075560983,
"grad_norm": 3.6360650062561035,
"learning_rate": 4.9977201158572344e-06,
"loss": 0.2709,
"step": 90
},
{
"epoch": 0.08890245764005496,
"grad_norm": 1.8734065294265747,
"learning_rate": 4.997548136258561e-06,
"loss": 0.2115,
"step": 91
},
{
"epoch": 0.0898794077240116,
"grad_norm": 2.348198175430298,
"learning_rate": 4.99736990701942e-06,
"loss": 0.2326,
"step": 92
},
{
"epoch": 0.09085635780796825,
"grad_norm": 2.2890987396240234,
"learning_rate": 4.997185428585795e-06,
"loss": 0.208,
"step": 93
},
{
"epoch": 0.0918333078919249,
"grad_norm": 2.5858378410339355,
"learning_rate": 4.996994701419309e-06,
"loss": 0.2217,
"step": 94
},
{
"epoch": 0.09281025797588155,
"grad_norm": 2.160419225692749,
"learning_rate": 4.99679772599722e-06,
"loss": 0.222,
"step": 95
},
{
"epoch": 0.09378720805983819,
"grad_norm": 2.987501859664917,
"learning_rate": 4.996594502812421e-06,
"loss": 0.2521,
"step": 96
},
{
"epoch": 0.09476415814379484,
"grad_norm": 2.2823338508605957,
"learning_rate": 4.996385032373441e-06,
"loss": 0.2104,
"step": 97
},
{
"epoch": 0.09574110822775149,
"grad_norm": 2.4988088607788086,
"learning_rate": 4.9961693152044385e-06,
"loss": 0.222,
"step": 98
},
{
"epoch": 0.09671805831170814,
"grad_norm": 2.2819881439208984,
"learning_rate": 4.995947351845207e-06,
"loss": 0.2042,
"step": 99
},
{
"epoch": 0.09769500839566478,
"grad_norm": 1.820811152458191,
"learning_rate": 4.995719142851166e-06,
"loss": 0.1957,
"step": 100
},
{
"epoch": 0.09867195847962143,
"grad_norm": 3.9323723316192627,
"learning_rate": 4.995484688793365e-06,
"loss": 0.2339,
"step": 101
},
{
"epoch": 0.09964890856357808,
"grad_norm": 2.8077266216278076,
"learning_rate": 4.995243990258481e-06,
"loss": 0.2151,
"step": 102
},
{
"epoch": 0.10062585864753473,
"grad_norm": 1.6394991874694824,
"learning_rate": 4.994997047848818e-06,
"loss": 0.2219,
"step": 103
},
{
"epoch": 0.10160280873149137,
"grad_norm": 3.04624342918396,
"learning_rate": 4.994743862182301e-06,
"loss": 0.1981,
"step": 104
},
{
"epoch": 0.10257975881544802,
"grad_norm": 2.501258611679077,
"learning_rate": 4.994484433892479e-06,
"loss": 0.2088,
"step": 105
},
{
"epoch": 0.10355670889940467,
"grad_norm": 1.729941725730896,
"learning_rate": 4.994218763628522e-06,
"loss": 0.2029,
"step": 106
},
{
"epoch": 0.10453365898336132,
"grad_norm": 3.224109649658203,
"learning_rate": 4.9939468520552205e-06,
"loss": 0.2307,
"step": 107
},
{
"epoch": 0.10551060906731796,
"grad_norm": 1.9119126796722412,
"learning_rate": 4.993668699852982e-06,
"loss": 0.2221,
"step": 108
},
{
"epoch": 0.10648755915127461,
"grad_norm": 1.848638653755188,
"learning_rate": 4.993384307717828e-06,
"loss": 0.2143,
"step": 109
},
{
"epoch": 0.10746450923523126,
"grad_norm": 1.9752260446548462,
"learning_rate": 4.993093676361396e-06,
"loss": 0.2258,
"step": 110
},
{
"epoch": 0.10844145931918792,
"grad_norm": 2.0553202629089355,
"learning_rate": 4.9927968065109366e-06,
"loss": 0.2356,
"step": 111
},
{
"epoch": 0.10941840940314455,
"grad_norm": 1.6603600978851318,
"learning_rate": 4.99249369890931e-06,
"loss": 0.1998,
"step": 112
},
{
"epoch": 0.1103953594871012,
"grad_norm": 4.3979339599609375,
"learning_rate": 4.992184354314986e-06,
"loss": 0.2492,
"step": 113
},
{
"epoch": 0.11137230957105786,
"grad_norm": 1.3203685283660889,
"learning_rate": 4.99186877350204e-06,
"loss": 0.1949,
"step": 114
},
{
"epoch": 0.11234925965501451,
"grad_norm": 1.582054853439331,
"learning_rate": 4.991546957260155e-06,
"loss": 0.1899,
"step": 115
},
{
"epoch": 0.11332620973897115,
"grad_norm": 3.809147357940674,
"learning_rate": 4.991218906394612e-06,
"loss": 0.2529,
"step": 116
},
{
"epoch": 0.1143031598229278,
"grad_norm": 2.083314895629883,
"learning_rate": 4.990884621726298e-06,
"loss": 0.2269,
"step": 117
},
{
"epoch": 0.11528010990688445,
"grad_norm": 2.141798734664917,
"learning_rate": 4.990544104091696e-06,
"loss": 0.1825,
"step": 118
},
{
"epoch": 0.1162570599908411,
"grad_norm": 1.1795802116394043,
"learning_rate": 4.990197354342889e-06,
"loss": 0.1867,
"step": 119
},
{
"epoch": 0.11723401007479774,
"grad_norm": 2.0592150688171387,
"learning_rate": 4.9898443733475495e-06,
"loss": 0.2038,
"step": 120
},
{
"epoch": 0.11821096015875439,
"grad_norm": 2.3183350563049316,
"learning_rate": 4.9894851619889475e-06,
"loss": 0.2277,
"step": 121
},
{
"epoch": 0.11918791024271104,
"grad_norm": 1.8908512592315674,
"learning_rate": 4.989119721165941e-06,
"loss": 0.1979,
"step": 122
},
{
"epoch": 0.12016486032666769,
"grad_norm": 2.6937687397003174,
"learning_rate": 4.9887480517929746e-06,
"loss": 0.2323,
"step": 123
},
{
"epoch": 0.12114181041062433,
"grad_norm": 2.000697135925293,
"learning_rate": 4.988370154800083e-06,
"loss": 0.1972,
"step": 124
},
{
"epoch": 0.12211876049458098,
"grad_norm": 5.163992881774902,
"learning_rate": 4.98798603113288e-06,
"loss": 0.1952,
"step": 125
},
{
"epoch": 0.12309571057853763,
"grad_norm": 1.2718846797943115,
"learning_rate": 4.987595681752564e-06,
"loss": 0.1936,
"step": 126
},
{
"epoch": 0.12407266066249427,
"grad_norm": 1.3900277614593506,
"learning_rate": 4.987199107635907e-06,
"loss": 0.1927,
"step": 127
},
{
"epoch": 0.12504961074645093,
"grad_norm": 1.9162976741790771,
"learning_rate": 4.986796309775264e-06,
"loss": 0.1978,
"step": 128
},
{
"epoch": 0.12602656083040756,
"grad_norm": 2.0890822410583496,
"learning_rate": 4.986387289178559e-06,
"loss": 0.234,
"step": 129
},
{
"epoch": 0.1270035109143642,
"grad_norm": 2.2969601154327393,
"learning_rate": 4.985972046869288e-06,
"loss": 0.2219,
"step": 130
},
{
"epoch": 0.12798046099832086,
"grad_norm": 1.5308568477630615,
"learning_rate": 4.985550583886516e-06,
"loss": 0.2064,
"step": 131
},
{
"epoch": 0.1289574110822775,
"grad_norm": 1.4585882425308228,
"learning_rate": 4.985122901284875e-06,
"loss": 0.1977,
"step": 132
},
{
"epoch": 0.12993436116623416,
"grad_norm": 1.4630974531173706,
"learning_rate": 4.984689000134559e-06,
"loss": 0.1967,
"step": 133
},
{
"epoch": 0.13091131125019081,
"grad_norm": 1.3947391510009766,
"learning_rate": 4.984248881521322e-06,
"loss": 0.2117,
"step": 134
},
{
"epoch": 0.13188826133414747,
"grad_norm": 1.6325701475143433,
"learning_rate": 4.98380254654648e-06,
"loss": 0.186,
"step": 135
},
{
"epoch": 0.13286521141810412,
"grad_norm": 4.865332126617432,
"learning_rate": 4.9833499963269e-06,
"loss": 0.2107,
"step": 136
},
{
"epoch": 0.13384216150206074,
"grad_norm": 1.3951985836029053,
"learning_rate": 4.982891231995005e-06,
"loss": 0.1695,
"step": 137
},
{
"epoch": 0.1348191115860174,
"grad_norm": 2.447169780731201,
"learning_rate": 4.982426254698764e-06,
"loss": 0.2176,
"step": 138
},
{
"epoch": 0.13579606166997404,
"grad_norm": 1.225381851196289,
"learning_rate": 4.981955065601693e-06,
"loss": 0.2107,
"step": 139
},
{
"epoch": 0.1367730117539307,
"grad_norm": 1.4975941181182861,
"learning_rate": 4.981477665882856e-06,
"loss": 0.195,
"step": 140
},
{
"epoch": 0.13774996183788735,
"grad_norm": 2.819326877593994,
"learning_rate": 4.980994056736854e-06,
"loss": 0.2106,
"step": 141
},
{
"epoch": 0.138726911921844,
"grad_norm": 1.3914092779159546,
"learning_rate": 4.980504239373828e-06,
"loss": 0.1818,
"step": 142
},
{
"epoch": 0.13970386200580065,
"grad_norm": 1.275080680847168,
"learning_rate": 4.980008215019452e-06,
"loss": 0.1882,
"step": 143
},
{
"epoch": 0.1406808120897573,
"grad_norm": 1.7779786586761475,
"learning_rate": 4.979505984914934e-06,
"loss": 0.1741,
"step": 144
},
{
"epoch": 0.14165776217371392,
"grad_norm": 1.4576950073242188,
"learning_rate": 4.978997550317008e-06,
"loss": 0.1941,
"step": 145
},
{
"epoch": 0.14263471225767058,
"grad_norm": 1.334631085395813,
"learning_rate": 4.9784829124979365e-06,
"loss": 0.1741,
"step": 146
},
{
"epoch": 0.14361166234162723,
"grad_norm": 2.1714446544647217,
"learning_rate": 4.977962072745503e-06,
"loss": 0.1858,
"step": 147
},
{
"epoch": 0.14458861242558388,
"grad_norm": 2.0816490650177,
"learning_rate": 4.977435032363009e-06,
"loss": 0.2045,
"step": 148
},
{
"epoch": 0.14556556250954053,
"grad_norm": 1.6877857446670532,
"learning_rate": 4.976901792669274e-06,
"loss": 0.2134,
"step": 149
},
{
"epoch": 0.14654251259349718,
"grad_norm": 1.721567153930664,
"learning_rate": 4.97636235499863e-06,
"loss": 0.1691,
"step": 150
},
{
"epoch": 0.14751946267745383,
"grad_norm": 1.3968329429626465,
"learning_rate": 4.975816720700915e-06,
"loss": 0.1814,
"step": 151
},
{
"epoch": 0.14849641276141048,
"grad_norm": 1.2914350032806396,
"learning_rate": 4.975264891141478e-06,
"loss": 0.2027,
"step": 152
},
{
"epoch": 0.1494733628453671,
"grad_norm": 1.7255231142044067,
"learning_rate": 4.974706867701166e-06,
"loss": 0.1999,
"step": 153
},
{
"epoch": 0.15045031292932376,
"grad_norm": 2.8396005630493164,
"learning_rate": 4.974142651776327e-06,
"loss": 0.2043,
"step": 154
},
{
"epoch": 0.1514272630132804,
"grad_norm": 1.509140968322754,
"learning_rate": 4.973572244778806e-06,
"loss": 0.2053,
"step": 155
},
{
"epoch": 0.15240421309723706,
"grad_norm": 1.8194832801818848,
"learning_rate": 4.972995648135937e-06,
"loss": 0.2071,
"step": 156
},
{
"epoch": 0.1533811631811937,
"grad_norm": 1.8108936548233032,
"learning_rate": 4.972412863290542e-06,
"loss": 0.1914,
"step": 157
},
{
"epoch": 0.15435811326515037,
"grad_norm": 1.7810618877410889,
"learning_rate": 4.971823891700933e-06,
"loss": 0.1896,
"step": 158
},
{
"epoch": 0.15533506334910702,
"grad_norm": 2.121969699859619,
"learning_rate": 4.971228734840897e-06,
"loss": 0.1745,
"step": 159
},
{
"epoch": 0.15631201343306367,
"grad_norm": 1.448722004890442,
"learning_rate": 4.970627394199703e-06,
"loss": 0.209,
"step": 160
},
{
"epoch": 0.1572889635170203,
"grad_norm": 1.5792391300201416,
"learning_rate": 4.970019871282091e-06,
"loss": 0.1886,
"step": 161
},
{
"epoch": 0.15826591360097694,
"grad_norm": 1.3888905048370361,
"learning_rate": 4.9694061676082716e-06,
"loss": 0.2009,
"step": 162
},
{
"epoch": 0.1592428636849336,
"grad_norm": 1.5487862825393677,
"learning_rate": 4.968786284713922e-06,
"loss": 0.1789,
"step": 163
},
{
"epoch": 0.16021981376889025,
"grad_norm": 1.1394370794296265,
"learning_rate": 4.968160224150182e-06,
"loss": 0.1805,
"step": 164
},
{
"epoch": 0.1611967638528469,
"grad_norm": 1.8469414710998535,
"learning_rate": 4.967527987483649e-06,
"loss": 0.1984,
"step": 165
},
{
"epoch": 0.16217371393680355,
"grad_norm": 1.388489007949829,
"learning_rate": 4.966889576296376e-06,
"loss": 0.1993,
"step": 166
},
{
"epoch": 0.1631506640207602,
"grad_norm": 1.0046708583831787,
"learning_rate": 4.966244992185865e-06,
"loss": 0.1744,
"step": 167
},
{
"epoch": 0.16412761410471682,
"grad_norm": 1.326556921005249,
"learning_rate": 4.965594236765066e-06,
"loss": 0.1746,
"step": 168
},
{
"epoch": 0.16510456418867347,
"grad_norm": 2.322471857070923,
"learning_rate": 4.9649373116623705e-06,
"loss": 0.1985,
"step": 169
},
{
"epoch": 0.16608151427263013,
"grad_norm": 1.5991991758346558,
"learning_rate": 4.96427421852161e-06,
"loss": 0.1906,
"step": 170
},
{
"epoch": 0.16705846435658678,
"grad_norm": 1.6929835081100464,
"learning_rate": 4.963604959002047e-06,
"loss": 0.1834,
"step": 171
},
{
"epoch": 0.16803541444054343,
"grad_norm": 1.5653678178787231,
"learning_rate": 4.96292953477838e-06,
"loss": 0.1815,
"step": 172
},
{
"epoch": 0.16901236452450008,
"grad_norm": 1.2223589420318604,
"learning_rate": 4.962247947540727e-06,
"loss": 0.1883,
"step": 173
},
{
"epoch": 0.16998931460845673,
"grad_norm": 0.8988173604011536,
"learning_rate": 4.9615601989946326e-06,
"loss": 0.1702,
"step": 174
},
{
"epoch": 0.17096626469241338,
"grad_norm": 1.0329456329345703,
"learning_rate": 4.960866290861055e-06,
"loss": 0.1789,
"step": 175
},
{
"epoch": 0.17194321477637,
"grad_norm": 2.3423991203308105,
"learning_rate": 4.960166224876371e-06,
"loss": 0.2061,
"step": 176
},
{
"epoch": 0.17292016486032666,
"grad_norm": 1.2909654378890991,
"learning_rate": 4.95946000279236e-06,
"loss": 0.1929,
"step": 177
},
{
"epoch": 0.1738971149442833,
"grad_norm": 1.4275271892547607,
"learning_rate": 4.958747626376209e-06,
"loss": 0.1783,
"step": 178
},
{
"epoch": 0.17487406502823996,
"grad_norm": 1.309098482131958,
"learning_rate": 4.958029097410507e-06,
"loss": 0.2004,
"step": 179
},
{
"epoch": 0.1758510151121966,
"grad_norm": 1.9952583312988281,
"learning_rate": 4.957304417693234e-06,
"loss": 0.1978,
"step": 180
},
{
"epoch": 0.17682796519615326,
"grad_norm": 1.8879168033599854,
"learning_rate": 4.956573589037766e-06,
"loss": 0.1988,
"step": 181
},
{
"epoch": 0.17780491528010992,
"grad_norm": 1.4148002862930298,
"learning_rate": 4.9558366132728604e-06,
"loss": 0.1714,
"step": 182
},
{
"epoch": 0.17878186536406657,
"grad_norm": 1.941772699356079,
"learning_rate": 4.9550934922426604e-06,
"loss": 0.2368,
"step": 183
},
{
"epoch": 0.1797588154480232,
"grad_norm": 1.2670023441314697,
"learning_rate": 4.954344227806687e-06,
"loss": 0.1979,
"step": 184
},
{
"epoch": 0.18073576553197984,
"grad_norm": 1.1730530261993408,
"learning_rate": 4.95358882183983e-06,
"loss": 0.1913,
"step": 185
},
{
"epoch": 0.1817127156159365,
"grad_norm": 1.131927251815796,
"learning_rate": 4.95282727623235e-06,
"loss": 0.1858,
"step": 186
},
{
"epoch": 0.18268966569989314,
"grad_norm": 2.0665884017944336,
"learning_rate": 4.952059592889872e-06,
"loss": 0.1903,
"step": 187
},
{
"epoch": 0.1836666157838498,
"grad_norm": 1.1630992889404297,
"learning_rate": 4.951285773733377e-06,
"loss": 0.1698,
"step": 188
},
{
"epoch": 0.18464356586780645,
"grad_norm": 1.262695550918579,
"learning_rate": 4.950505820699199e-06,
"loss": 0.1782,
"step": 189
},
{
"epoch": 0.1856205159517631,
"grad_norm": 1.7037773132324219,
"learning_rate": 4.949719735739025e-06,
"loss": 0.1825,
"step": 190
},
{
"epoch": 0.18659746603571975,
"grad_norm": 1.0291463136672974,
"learning_rate": 4.948927520819882e-06,
"loss": 0.1769,
"step": 191
},
{
"epoch": 0.18757441611967637,
"grad_norm": 2.2458765506744385,
"learning_rate": 4.948129177924139e-06,
"loss": 0.1954,
"step": 192
},
{
"epoch": 0.18855136620363303,
"grad_norm": 1.084694504737854,
"learning_rate": 4.947324709049496e-06,
"loss": 0.1901,
"step": 193
},
{
"epoch": 0.18952831628758968,
"grad_norm": 1.9943842887878418,
"learning_rate": 4.946514116208985e-06,
"loss": 0.1805,
"step": 194
},
{
"epoch": 0.19050526637154633,
"grad_norm": 2.2961764335632324,
"learning_rate": 4.94569740143096e-06,
"loss": 0.1935,
"step": 195
},
{
"epoch": 0.19148221645550298,
"grad_norm": 1.5056332349777222,
"learning_rate": 4.944874566759095e-06,
"loss": 0.1801,
"step": 196
},
{
"epoch": 0.19245916653945963,
"grad_norm": 2.1005048751831055,
"learning_rate": 4.944045614252379e-06,
"loss": 0.1897,
"step": 197
},
{
"epoch": 0.19343611662341628,
"grad_norm": 1.011614203453064,
"learning_rate": 4.943210545985105e-06,
"loss": 0.1797,
"step": 198
},
{
"epoch": 0.19441306670737293,
"grad_norm": 1.0442618131637573,
"learning_rate": 4.942369364046877e-06,
"loss": 0.1827,
"step": 199
},
{
"epoch": 0.19539001679132956,
"grad_norm": 1.0433064699172974,
"learning_rate": 4.941522070542591e-06,
"loss": 0.1694,
"step": 200
},
{
"epoch": 0.1963669668752862,
"grad_norm": 1.0130871534347534,
"learning_rate": 4.940668667592439e-06,
"loss": 0.1786,
"step": 201
},
{
"epoch": 0.19734391695924286,
"grad_norm": 1.100374698638916,
"learning_rate": 4.9398091573318995e-06,
"loss": 0.1759,
"step": 202
},
{
"epoch": 0.1983208670431995,
"grad_norm": 1.3154064416885376,
"learning_rate": 4.938943541911735e-06,
"loss": 0.184,
"step": 203
},
{
"epoch": 0.19929781712715616,
"grad_norm": 1.5183656215667725,
"learning_rate": 4.938071823497982e-06,
"loss": 0.185,
"step": 204
},
{
"epoch": 0.20027476721111281,
"grad_norm": 1.569200038909912,
"learning_rate": 4.9371940042719516e-06,
"loss": 0.1894,
"step": 205
},
{
"epoch": 0.20125171729506947,
"grad_norm": 2.047342538833618,
"learning_rate": 4.93631008643022e-06,
"loss": 0.1969,
"step": 206
},
{
"epoch": 0.2022286673790261,
"grad_norm": 0.9860909581184387,
"learning_rate": 4.935420072184624e-06,
"loss": 0.1627,
"step": 207
},
{
"epoch": 0.20320561746298274,
"grad_norm": 0.8913268446922302,
"learning_rate": 4.934523963762254e-06,
"loss": 0.1839,
"step": 208
},
{
"epoch": 0.2041825675469394,
"grad_norm": 1.0603684186935425,
"learning_rate": 4.933621763405454e-06,
"loss": 0.1836,
"step": 209
},
{
"epoch": 0.20515951763089604,
"grad_norm": 0.8889226913452148,
"learning_rate": 4.932713473371807e-06,
"loss": 0.1746,
"step": 210
},
{
"epoch": 0.2061364677148527,
"grad_norm": 1.3194069862365723,
"learning_rate": 4.931799095934137e-06,
"loss": 0.1752,
"step": 211
},
{
"epoch": 0.20711341779880935,
"grad_norm": 1.4344172477722168,
"learning_rate": 4.930878633380501e-06,
"loss": 0.1776,
"step": 212
},
{
"epoch": 0.208090367882766,
"grad_norm": 1.1890789270401,
"learning_rate": 4.929952088014182e-06,
"loss": 0.201,
"step": 213
},
{
"epoch": 0.20906731796672265,
"grad_norm": 1.1506404876708984,
"learning_rate": 4.929019462153682e-06,
"loss": 0.19,
"step": 214
},
{
"epoch": 0.21004426805067927,
"grad_norm": 1.034428358078003,
"learning_rate": 4.928080758132723e-06,
"loss": 0.1764,
"step": 215
},
{
"epoch": 0.21102121813463592,
"grad_norm": 1.5425089597702026,
"learning_rate": 4.927135978300231e-06,
"loss": 0.1708,
"step": 216
},
{
"epoch": 0.21199816821859258,
"grad_norm": 1.2676106691360474,
"learning_rate": 4.926185125020341e-06,
"loss": 0.1773,
"step": 217
},
{
"epoch": 0.21297511830254923,
"grad_norm": 0.959566056728363,
"learning_rate": 4.92522820067238e-06,
"loss": 0.1806,
"step": 218
},
{
"epoch": 0.21395206838650588,
"grad_norm": 1.3489631414413452,
"learning_rate": 4.924265207650874e-06,
"loss": 0.1856,
"step": 219
},
{
"epoch": 0.21492901847046253,
"grad_norm": 1.06121826171875,
"learning_rate": 4.923296148365526e-06,
"loss": 0.1889,
"step": 220
},
{
"epoch": 0.21590596855441918,
"grad_norm": 1.19551420211792,
"learning_rate": 4.922321025241225e-06,
"loss": 0.1659,
"step": 221
},
{
"epoch": 0.21688291863837583,
"grad_norm": 1.5810611248016357,
"learning_rate": 4.921339840718031e-06,
"loss": 0.1846,
"step": 222
},
{
"epoch": 0.21785986872233246,
"grad_norm": 1.8626632690429688,
"learning_rate": 4.920352597251174e-06,
"loss": 0.2031,
"step": 223
},
{
"epoch": 0.2188368188062891,
"grad_norm": 1.2777925729751587,
"learning_rate": 4.919359297311041e-06,
"loss": 0.1818,
"step": 224
},
{
"epoch": 0.21981376889024576,
"grad_norm": 0.956462025642395,
"learning_rate": 4.9183599433831785e-06,
"loss": 0.1748,
"step": 225
},
{
"epoch": 0.2207907189742024,
"grad_norm": 0.807868242263794,
"learning_rate": 4.91735453796828e-06,
"loss": 0.1672,
"step": 226
},
{
"epoch": 0.22176766905815906,
"grad_norm": 1.7048554420471191,
"learning_rate": 4.9163430835821805e-06,
"loss": 0.1986,
"step": 227
},
{
"epoch": 0.2227446191421157,
"grad_norm": 1.080628514289856,
"learning_rate": 4.915325582755853e-06,
"loss": 0.1742,
"step": 228
},
{
"epoch": 0.22372156922607236,
"grad_norm": 1.4222372770309448,
"learning_rate": 4.914302038035402e-06,
"loss": 0.1582,
"step": 229
},
{
"epoch": 0.22469851931002902,
"grad_norm": 1.040003776550293,
"learning_rate": 4.913272451982053e-06,
"loss": 0.177,
"step": 230
},
{
"epoch": 0.22567546939398564,
"grad_norm": 1.0453258752822876,
"learning_rate": 4.9122368271721485e-06,
"loss": 0.1843,
"step": 231
},
{
"epoch": 0.2266524194779423,
"grad_norm": 1.233941912651062,
"learning_rate": 4.911195166197145e-06,
"loss": 0.1608,
"step": 232
},
{
"epoch": 0.22762936956189894,
"grad_norm": 1.4035524129867554,
"learning_rate": 4.910147471663601e-06,
"loss": 0.2034,
"step": 233
},
{
"epoch": 0.2286063196458556,
"grad_norm": 0.9531283974647522,
"learning_rate": 4.909093746193173e-06,
"loss": 0.1978,
"step": 234
},
{
"epoch": 0.22958326972981224,
"grad_norm": 0.927309513092041,
"learning_rate": 4.908033992422609e-06,
"loss": 0.1871,
"step": 235
},
{
"epoch": 0.2305602198137689,
"grad_norm": 1.0833598375320435,
"learning_rate": 4.9069682130037434e-06,
"loss": 0.1691,
"step": 236
},
{
"epoch": 0.23153716989772555,
"grad_norm": 1.1534310579299927,
"learning_rate": 4.905896410603485e-06,
"loss": 0.2012,
"step": 237
},
{
"epoch": 0.2325141199816822,
"grad_norm": 1.2670401334762573,
"learning_rate": 4.904818587903817e-06,
"loss": 0.1969,
"step": 238
},
{
"epoch": 0.23349107006563882,
"grad_norm": 1.222806692123413,
"learning_rate": 4.903734747601785e-06,
"loss": 0.2041,
"step": 239
},
{
"epoch": 0.23446802014959547,
"grad_norm": 0.9613257646560669,
"learning_rate": 4.902644892409496e-06,
"loss": 0.1745,
"step": 240
},
{
"epoch": 0.23544497023355213,
"grad_norm": 1.2783435583114624,
"learning_rate": 4.901549025054103e-06,
"loss": 0.1828,
"step": 241
},
{
"epoch": 0.23642192031750878,
"grad_norm": 1.9349430799484253,
"learning_rate": 4.900447148277809e-06,
"loss": 0.1778,
"step": 242
},
{
"epoch": 0.23739887040146543,
"grad_norm": 0.9153691530227661,
"learning_rate": 4.899339264837851e-06,
"loss": 0.1673,
"step": 243
},
{
"epoch": 0.23837582048542208,
"grad_norm": 1.086670160293579,
"learning_rate": 4.898225377506496e-06,
"loss": 0.1621,
"step": 244
},
{
"epoch": 0.23935277056937873,
"grad_norm": 0.9247845411300659,
"learning_rate": 4.897105489071036e-06,
"loss": 0.154,
"step": 245
},
{
"epoch": 0.24032972065333538,
"grad_norm": 1.4103556871414185,
"learning_rate": 4.89597960233378e-06,
"loss": 0.1596,
"step": 246
},
{
"epoch": 0.241306670737292,
"grad_norm": 0.8966643214225769,
"learning_rate": 4.894847720112047e-06,
"loss": 0.1672,
"step": 247
},
{
"epoch": 0.24228362082124866,
"grad_norm": 1.4627892971038818,
"learning_rate": 4.893709845238155e-06,
"loss": 0.1829,
"step": 248
},
{
"epoch": 0.2432605709052053,
"grad_norm": 1.492667317390442,
"learning_rate": 4.8925659805594214e-06,
"loss": 0.1889,
"step": 249
},
{
"epoch": 0.24423752098916196,
"grad_norm": 2.349557638168335,
"learning_rate": 4.891416128938151e-06,
"loss": 0.1859,
"step": 250
},
{
"epoch": 0.2452144710731186,
"grad_norm": 1.0932039022445679,
"learning_rate": 4.890260293251629e-06,
"loss": 0.1672,
"step": 251
},
{
"epoch": 0.24619142115707526,
"grad_norm": 0.9675506949424744,
"learning_rate": 4.889098476392115e-06,
"loss": 0.1818,
"step": 252
},
{
"epoch": 0.24716837124103191,
"grad_norm": 0.7860025763511658,
"learning_rate": 4.887930681266835e-06,
"loss": 0.1619,
"step": 253
},
{
"epoch": 0.24814532132498854,
"grad_norm": 0.9406007528305054,
"learning_rate": 4.8867569107979755e-06,
"loss": 0.162,
"step": 254
},
{
"epoch": 0.2491222714089452,
"grad_norm": 2.2786617279052734,
"learning_rate": 4.885577167922672e-06,
"loss": 0.1872,
"step": 255
},
{
"epoch": 0.25009922149290187,
"grad_norm": 1.7823460102081299,
"learning_rate": 4.884391455593011e-06,
"loss": 0.1831,
"step": 256
},
{
"epoch": 0.2510761715768585,
"grad_norm": 0.9489114284515381,
"learning_rate": 4.883199776776011e-06,
"loss": 0.1706,
"step": 257
},
{
"epoch": 0.2520531216608151,
"grad_norm": 1.132427453994751,
"learning_rate": 4.882002134453622e-06,
"loss": 0.1796,
"step": 258
},
{
"epoch": 0.25303007174477177,
"grad_norm": 1.2291914224624634,
"learning_rate": 4.880798531622717e-06,
"loss": 0.1791,
"step": 259
},
{
"epoch": 0.2540070218287284,
"grad_norm": 1.2372835874557495,
"learning_rate": 4.879588971295083e-06,
"loss": 0.1656,
"step": 260
},
{
"epoch": 0.25498397191268507,
"grad_norm": 3.1791579723358154,
"learning_rate": 4.878373456497416e-06,
"loss": 0.1843,
"step": 261
},
{
"epoch": 0.2559609219966417,
"grad_norm": 1.6447696685791016,
"learning_rate": 4.877151990271312e-06,
"loss": 0.1761,
"step": 262
},
{
"epoch": 0.2569378720805984,
"grad_norm": 1.0099893808364868,
"learning_rate": 4.875924575673257e-06,
"loss": 0.1886,
"step": 263
},
{
"epoch": 0.257914822164555,
"grad_norm": 1.0321879386901855,
"learning_rate": 4.874691215774624e-06,
"loss": 0.1621,
"step": 264
},
{
"epoch": 0.2588917722485117,
"grad_norm": 1.142579436302185,
"learning_rate": 4.873451913661662e-06,
"loss": 0.1775,
"step": 265
},
{
"epoch": 0.2598687223324683,
"grad_norm": 0.9944607019424438,
"learning_rate": 4.87220667243549e-06,
"loss": 0.1577,
"step": 266
},
{
"epoch": 0.260845672416425,
"grad_norm": 1.294043779373169,
"learning_rate": 4.8709554952120865e-06,
"loss": 0.1717,
"step": 267
},
{
"epoch": 0.26182262250038163,
"grad_norm": 0.9566840529441833,
"learning_rate": 4.8696983851222875e-06,
"loss": 0.1489,
"step": 268
},
{
"epoch": 0.2627995725843383,
"grad_norm": 0.7195562720298767,
"learning_rate": 4.8684353453117704e-06,
"loss": 0.1621,
"step": 269
},
{
"epoch": 0.26377652266829493,
"grad_norm": 0.9187518358230591,
"learning_rate": 4.867166378941054e-06,
"loss": 0.1609,
"step": 270
},
{
"epoch": 0.2647534727522516,
"grad_norm": 1.0037418603897095,
"learning_rate": 4.865891489185485e-06,
"loss": 0.1686,
"step": 271
},
{
"epoch": 0.26573042283620824,
"grad_norm": 0.9572906494140625,
"learning_rate": 4.864610679235234e-06,
"loss": 0.1548,
"step": 272
},
{
"epoch": 0.2667073729201649,
"grad_norm": 1.7213281393051147,
"learning_rate": 4.863323952295286e-06,
"loss": 0.1816,
"step": 273
},
{
"epoch": 0.2676843230041215,
"grad_norm": 0.7924426794052124,
"learning_rate": 4.862031311585428e-06,
"loss": 0.1573,
"step": 274
},
{
"epoch": 0.26866127308807813,
"grad_norm": 1.6240406036376953,
"learning_rate": 4.860732760340251e-06,
"loss": 0.1842,
"step": 275
},
{
"epoch": 0.2696382231720348,
"grad_norm": 1.8995360136032104,
"learning_rate": 4.859428301809131e-06,
"loss": 0.1678,
"step": 276
},
{
"epoch": 0.27061517325599144,
"grad_norm": 1.3227934837341309,
"learning_rate": 4.858117939256231e-06,
"loss": 0.1684,
"step": 277
},
{
"epoch": 0.2715921233399481,
"grad_norm": 0.8575958609580994,
"learning_rate": 4.856801675960482e-06,
"loss": 0.152,
"step": 278
},
{
"epoch": 0.27256907342390474,
"grad_norm": 0.9428021907806396,
"learning_rate": 4.855479515215583e-06,
"loss": 0.1691,
"step": 279
},
{
"epoch": 0.2735460235078614,
"grad_norm": 1.7851122617721558,
"learning_rate": 4.854151460329994e-06,
"loss": 0.1874,
"step": 280
},
{
"epoch": 0.27452297359181804,
"grad_norm": 1.286556601524353,
"learning_rate": 4.852817514626916e-06,
"loss": 0.1659,
"step": 281
},
{
"epoch": 0.2754999236757747,
"grad_norm": 0.6606729626655579,
"learning_rate": 4.8514776814442975e-06,
"loss": 0.1514,
"step": 282
},
{
"epoch": 0.27647687375973135,
"grad_norm": 1.2322301864624023,
"learning_rate": 4.850131964134818e-06,
"loss": 0.161,
"step": 283
},
{
"epoch": 0.277453823843688,
"grad_norm": 1.3725941181182861,
"learning_rate": 4.8487803660658765e-06,
"loss": 0.1713,
"step": 284
},
{
"epoch": 0.27843077392764465,
"grad_norm": 1.284625768661499,
"learning_rate": 4.847422890619594e-06,
"loss": 0.1868,
"step": 285
},
{
"epoch": 0.2794077240116013,
"grad_norm": 1.9067991971969604,
"learning_rate": 4.846059541192794e-06,
"loss": 0.1787,
"step": 286
},
{
"epoch": 0.28038467409555795,
"grad_norm": 2.312279462814331,
"learning_rate": 4.844690321196999e-06,
"loss": 0.1636,
"step": 287
},
{
"epoch": 0.2813616241795146,
"grad_norm": 1.0614606142044067,
"learning_rate": 4.843315234058423e-06,
"loss": 0.1684,
"step": 288
},
{
"epoch": 0.2823385742634712,
"grad_norm": 0.8870749473571777,
"learning_rate": 4.841934283217961e-06,
"loss": 0.1785,
"step": 289
},
{
"epoch": 0.28331552434742785,
"grad_norm": 0.9860772490501404,
"learning_rate": 4.840547472131181e-06,
"loss": 0.1727,
"step": 290
},
{
"epoch": 0.2842924744313845,
"grad_norm": 1.1348960399627686,
"learning_rate": 4.839154804268313e-06,
"loss": 0.1703,
"step": 291
},
{
"epoch": 0.28526942451534115,
"grad_norm": 0.9704286456108093,
"learning_rate": 4.8377562831142465e-06,
"loss": 0.182,
"step": 292
},
{
"epoch": 0.2862463745992978,
"grad_norm": 0.795126736164093,
"learning_rate": 4.836351912168513e-06,
"loss": 0.1511,
"step": 293
},
{
"epoch": 0.28722332468325446,
"grad_norm": 0.9404454231262207,
"learning_rate": 4.834941694945286e-06,
"loss": 0.1667,
"step": 294
},
{
"epoch": 0.2882002747672111,
"grad_norm": 1.038159728050232,
"learning_rate": 4.833525634973367e-06,
"loss": 0.1898,
"step": 295
},
{
"epoch": 0.28917722485116776,
"grad_norm": 0.9030426144599915,
"learning_rate": 4.832103735796177e-06,
"loss": 0.1618,
"step": 296
},
{
"epoch": 0.2901541749351244,
"grad_norm": 1.3933353424072266,
"learning_rate": 4.830676000971747e-06,
"loss": 0.1683,
"step": 297
},
{
"epoch": 0.29113112501908106,
"grad_norm": 0.9132196307182312,
"learning_rate": 4.829242434072714e-06,
"loss": 0.1564,
"step": 298
},
{
"epoch": 0.2921080751030377,
"grad_norm": 0.9444112777709961,
"learning_rate": 4.827803038686309e-06,
"loss": 0.1665,
"step": 299
},
{
"epoch": 0.29308502518699436,
"grad_norm": 1.2773785591125488,
"learning_rate": 4.826357818414343e-06,
"loss": 0.1772,
"step": 300
},
{
"epoch": 0.294061975270951,
"grad_norm": 1.0572564601898193,
"learning_rate": 4.824906776873208e-06,
"loss": 0.1672,
"step": 301
},
{
"epoch": 0.29503892535490767,
"grad_norm": 1.107437014579773,
"learning_rate": 4.823449917693859e-06,
"loss": 0.1616,
"step": 302
},
{
"epoch": 0.2960158754388643,
"grad_norm": 1.1432325839996338,
"learning_rate": 4.82198724452181e-06,
"loss": 0.1698,
"step": 303
},
{
"epoch": 0.29699282552282097,
"grad_norm": 1.064653754234314,
"learning_rate": 4.820518761017123e-06,
"loss": 0.1741,
"step": 304
},
{
"epoch": 0.29796977560677756,
"grad_norm": 0.9631031155586243,
"learning_rate": 4.819044470854399e-06,
"loss": 0.1786,
"step": 305
},
{
"epoch": 0.2989467256907342,
"grad_norm": 0.8367262482643127,
"learning_rate": 4.817564377722771e-06,
"loss": 0.1766,
"step": 306
},
{
"epoch": 0.29992367577469087,
"grad_norm": 1.6821892261505127,
"learning_rate": 4.81607848532589e-06,
"loss": 0.1621,
"step": 307
},
{
"epoch": 0.3009006258586475,
"grad_norm": 0.7397382259368896,
"learning_rate": 4.81458679738192e-06,
"loss": 0.1569,
"step": 308
},
{
"epoch": 0.30187757594260417,
"grad_norm": 1.6769345998764038,
"learning_rate": 4.813089317623527e-06,
"loss": 0.1732,
"step": 309
},
{
"epoch": 0.3028545260265608,
"grad_norm": 0.8182811737060547,
"learning_rate": 4.8115860497978715e-06,
"loss": 0.151,
"step": 310
},
{
"epoch": 0.3038314761105175,
"grad_norm": 1.0059869289398193,
"learning_rate": 4.810076997666594e-06,
"loss": 0.1497,
"step": 311
},
{
"epoch": 0.3048084261944741,
"grad_norm": 0.8826881051063538,
"learning_rate": 4.808562165005812e-06,
"loss": 0.1696,
"step": 312
},
{
"epoch": 0.3057853762784308,
"grad_norm": 1.1489773988723755,
"learning_rate": 4.807041555606108e-06,
"loss": 0.1603,
"step": 313
},
{
"epoch": 0.3067623263623874,
"grad_norm": 0.824995756149292,
"learning_rate": 4.8055151732725175e-06,
"loss": 0.1638,
"step": 314
},
{
"epoch": 0.3077392764463441,
"grad_norm": 1.4060777425765991,
"learning_rate": 4.803983021824523e-06,
"loss": 0.1567,
"step": 315
},
{
"epoch": 0.30871622653030073,
"grad_norm": 0.9083486199378967,
"learning_rate": 4.802445105096042e-06,
"loss": 0.163,
"step": 316
},
{
"epoch": 0.3096931766142574,
"grad_norm": 1.0700857639312744,
"learning_rate": 4.800901426935422e-06,
"loss": 0.1709,
"step": 317
},
{
"epoch": 0.31067012669821403,
"grad_norm": 0.6764273643493652,
"learning_rate": 4.799351991205421e-06,
"loss": 0.1565,
"step": 318
},
{
"epoch": 0.3116470767821707,
"grad_norm": 1.3011271953582764,
"learning_rate": 4.797796801783211e-06,
"loss": 0.1699,
"step": 319
},
{
"epoch": 0.31262402686612734,
"grad_norm": 0.7218932509422302,
"learning_rate": 4.796235862560357e-06,
"loss": 0.1317,
"step": 320
},
{
"epoch": 0.31360097695008393,
"grad_norm": 0.9350078701972961,
"learning_rate": 4.794669177442815e-06,
"loss": 0.1744,
"step": 321
},
{
"epoch": 0.3145779270340406,
"grad_norm": 1.2761632204055786,
"learning_rate": 4.793096750350915e-06,
"loss": 0.1864,
"step": 322
},
{
"epoch": 0.31555487711799723,
"grad_norm": 3.156892776489258,
"learning_rate": 4.791518585219359e-06,
"loss": 0.1695,
"step": 323
},
{
"epoch": 0.3165318272019539,
"grad_norm": 0.8076464533805847,
"learning_rate": 4.789934685997206e-06,
"loss": 0.1636,
"step": 324
},
{
"epoch": 0.31750877728591054,
"grad_norm": 0.8827851414680481,
"learning_rate": 4.788345056647863e-06,
"loss": 0.1536,
"step": 325
},
{
"epoch": 0.3184857273698672,
"grad_norm": 1.3951282501220703,
"learning_rate": 4.786749701149076e-06,
"loss": 0.1752,
"step": 326
},
{
"epoch": 0.31946267745382384,
"grad_norm": 0.9473510384559631,
"learning_rate": 4.785148623492919e-06,
"loss": 0.1763,
"step": 327
},
{
"epoch": 0.3204396275377805,
"grad_norm": 1.2782249450683594,
"learning_rate": 4.783541827685786e-06,
"loss": 0.1619,
"step": 328
},
{
"epoch": 0.32141657762173714,
"grad_norm": 1.2298985719680786,
"learning_rate": 4.781929317748378e-06,
"loss": 0.1946,
"step": 329
},
{
"epoch": 0.3223935277056938,
"grad_norm": 0.8484151363372803,
"learning_rate": 4.780311097715695e-06,
"loss": 0.169,
"step": 330
},
{
"epoch": 0.32337047778965045,
"grad_norm": 0.8106393814086914,
"learning_rate": 4.778687171637028e-06,
"loss": 0.1808,
"step": 331
},
{
"epoch": 0.3243474278736071,
"grad_norm": 1.1082829236984253,
"learning_rate": 4.77705754357594e-06,
"loss": 0.1519,
"step": 332
},
{
"epoch": 0.32532437795756375,
"grad_norm": 1.0785582065582275,
"learning_rate": 4.77542221761027e-06,
"loss": 0.171,
"step": 333
},
{
"epoch": 0.3263013280415204,
"grad_norm": 0.8459275960922241,
"learning_rate": 4.773781197832109e-06,
"loss": 0.174,
"step": 334
},
{
"epoch": 0.32727827812547705,
"grad_norm": 0.7670328617095947,
"learning_rate": 4.772134488347798e-06,
"loss": 0.1699,
"step": 335
},
{
"epoch": 0.32825522820943365,
"grad_norm": 1.1015170812606812,
"learning_rate": 4.770482093277915e-06,
"loss": 0.1886,
"step": 336
},
{
"epoch": 0.3292321782933903,
"grad_norm": 0.8439568877220154,
"learning_rate": 4.768824016757265e-06,
"loss": 0.1567,
"step": 337
},
{
"epoch": 0.33020912837734695,
"grad_norm": 1.3197076320648193,
"learning_rate": 4.767160262934871e-06,
"loss": 0.1913,
"step": 338
},
{
"epoch": 0.3311860784613036,
"grad_norm": 0.8700889945030212,
"learning_rate": 4.765490835973962e-06,
"loss": 0.1386,
"step": 339
},
{
"epoch": 0.33216302854526025,
"grad_norm": 1.1503537893295288,
"learning_rate": 4.7638157400519595e-06,
"loss": 0.1563,
"step": 340
},
{
"epoch": 0.3331399786292169,
"grad_norm": 1.0167195796966553,
"learning_rate": 4.762134979360476e-06,
"loss": 0.1547,
"step": 341
},
{
"epoch": 0.33411692871317356,
"grad_norm": 1.6731079816818237,
"learning_rate": 4.760448558105294e-06,
"loss": 0.1905,
"step": 342
},
{
"epoch": 0.3350938787971302,
"grad_norm": 0.9303874373435974,
"learning_rate": 4.7587564805063645e-06,
"loss": 0.1686,
"step": 343
},
{
"epoch": 0.33607082888108686,
"grad_norm": 1.4790585041046143,
"learning_rate": 4.757058750797791e-06,
"loss": 0.1805,
"step": 344
},
{
"epoch": 0.3370477789650435,
"grad_norm": 2.3826332092285156,
"learning_rate": 4.755355373227818e-06,
"loss": 0.1775,
"step": 345
},
{
"epoch": 0.33802472904900016,
"grad_norm": 1.202460765838623,
"learning_rate": 4.753646352058827e-06,
"loss": 0.1755,
"step": 346
},
{
"epoch": 0.3390016791329568,
"grad_norm": 1.3687115907669067,
"learning_rate": 4.751931691567318e-06,
"loss": 0.1809,
"step": 347
},
{
"epoch": 0.33997862921691346,
"grad_norm": 1.2868486642837524,
"learning_rate": 4.750211396043904e-06,
"loss": 0.1708,
"step": 348
},
{
"epoch": 0.3409555793008701,
"grad_norm": 1.0164076089859009,
"learning_rate": 4.748485469793299e-06,
"loss": 0.1617,
"step": 349
},
{
"epoch": 0.34193252938482677,
"grad_norm": 1.2062278985977173,
"learning_rate": 4.746753917134304e-06,
"loss": 0.1652,
"step": 350
},
{
"epoch": 0.3429094794687834,
"grad_norm": 1.3958442211151123,
"learning_rate": 4.745016742399804e-06,
"loss": 0.1963,
"step": 351
},
{
"epoch": 0.34388642955274,
"grad_norm": 0.9678618311882019,
"learning_rate": 4.7432739499367475e-06,
"loss": 0.1445,
"step": 352
},
{
"epoch": 0.34486337963669667,
"grad_norm": 0.6724891662597656,
"learning_rate": 4.7415255441061414e-06,
"loss": 0.1626,
"step": 353
},
{
"epoch": 0.3458403297206533,
"grad_norm": 0.7690248489379883,
"learning_rate": 4.739771529283041e-06,
"loss": 0.1727,
"step": 354
},
{
"epoch": 0.34681727980460997,
"grad_norm": 1.0899006128311157,
"learning_rate": 4.738011909856535e-06,
"loss": 0.2008,
"step": 355
},
{
"epoch": 0.3477942298885666,
"grad_norm": 0.7474625110626221,
"learning_rate": 4.736246690229737e-06,
"loss": 0.1588,
"step": 356
},
{
"epoch": 0.34877117997252327,
"grad_norm": 0.8380438089370728,
"learning_rate": 4.734475874819776e-06,
"loss": 0.1615,
"step": 357
},
{
"epoch": 0.3497481300564799,
"grad_norm": 1.187130331993103,
"learning_rate": 4.7326994680577784e-06,
"loss": 0.1614,
"step": 358
},
{
"epoch": 0.3507250801404366,
"grad_norm": 0.8472177982330322,
"learning_rate": 4.730917474388869e-06,
"loss": 0.1601,
"step": 359
},
{
"epoch": 0.3517020302243932,
"grad_norm": 1.0092201232910156,
"learning_rate": 4.729129898272147e-06,
"loss": 0.171,
"step": 360
},
{
"epoch": 0.3526789803083499,
"grad_norm": 1.1632071733474731,
"learning_rate": 4.727336744180681e-06,
"loss": 0.1774,
"step": 361
},
{
"epoch": 0.35365593039230653,
"grad_norm": 1.825290322303772,
"learning_rate": 4.7255380166015016e-06,
"loss": 0.1766,
"step": 362
},
{
"epoch": 0.3546328804762632,
"grad_norm": 1.0044591426849365,
"learning_rate": 4.723733720035583e-06,
"loss": 0.1821,
"step": 363
},
{
"epoch": 0.35560983056021983,
"grad_norm": 1.4767296314239502,
"learning_rate": 4.721923858997833e-06,
"loss": 0.1753,
"step": 364
},
{
"epoch": 0.3565867806441765,
"grad_norm": 0.908411979675293,
"learning_rate": 4.720108438017088e-06,
"loss": 0.1799,
"step": 365
},
{
"epoch": 0.35756373072813313,
"grad_norm": 0.9019033908843994,
"learning_rate": 4.718287461636094e-06,
"loss": 0.1614,
"step": 366
},
{
"epoch": 0.3585406808120898,
"grad_norm": 0.6928865313529968,
"learning_rate": 4.716460934411498e-06,
"loss": 0.1524,
"step": 367
},
{
"epoch": 0.3595176308960464,
"grad_norm": 1.8299281597137451,
"learning_rate": 4.71462886091384e-06,
"loss": 0.1807,
"step": 368
},
{
"epoch": 0.36049458098000303,
"grad_norm": 0.8391110301017761,
"learning_rate": 4.712791245727534e-06,
"loss": 0.1588,
"step": 369
},
{
"epoch": 0.3614715310639597,
"grad_norm": 0.8084884881973267,
"learning_rate": 4.7109480934508646e-06,
"loss": 0.1521,
"step": 370
},
{
"epoch": 0.36244848114791633,
"grad_norm": 1.0511680841445923,
"learning_rate": 4.7090994086959705e-06,
"loss": 0.1656,
"step": 371
},
{
"epoch": 0.363425431231873,
"grad_norm": 1.0038580894470215,
"learning_rate": 4.707245196088835e-06,
"loss": 0.1669,
"step": 372
},
{
"epoch": 0.36440238131582964,
"grad_norm": 1.2239633798599243,
"learning_rate": 4.705385460269273e-06,
"loss": 0.1731,
"step": 373
},
{
"epoch": 0.3653793313997863,
"grad_norm": 0.8391385078430176,
"learning_rate": 4.703520205890921e-06,
"loss": 0.1597,
"step": 374
},
{
"epoch": 0.36635628148374294,
"grad_norm": 0.9824623465538025,
"learning_rate": 4.701649437621224e-06,
"loss": 0.1648,
"step": 375
},
{
"epoch": 0.3673332315676996,
"grad_norm": 1.1654932498931885,
"learning_rate": 4.699773160141425e-06,
"loss": 0.1729,
"step": 376
},
{
"epoch": 0.36831018165165624,
"grad_norm": 0.9378538131713867,
"learning_rate": 4.697891378146552e-06,
"loss": 0.1594,
"step": 377
},
{
"epoch": 0.3692871317356129,
"grad_norm": 1.00264573097229,
"learning_rate": 4.696004096345408e-06,
"loss": 0.1613,
"step": 378
},
{
"epoch": 0.37026408181956955,
"grad_norm": 0.7746912837028503,
"learning_rate": 4.694111319460558e-06,
"loss": 0.1619,
"step": 379
},
{
"epoch": 0.3712410319035262,
"grad_norm": 0.7986708879470825,
"learning_rate": 4.692213052228316e-06,
"loss": 0.1668,
"step": 380
},
{
"epoch": 0.37221798198748285,
"grad_norm": 0.9470774531364441,
"learning_rate": 4.690309299398736e-06,
"loss": 0.1669,
"step": 381
},
{
"epoch": 0.3731949320714395,
"grad_norm": 0.762748122215271,
"learning_rate": 4.6884000657356e-06,
"loss": 0.1528,
"step": 382
},
{
"epoch": 0.3741718821553961,
"grad_norm": 0.7623162269592285,
"learning_rate": 4.686485356016402e-06,
"loss": 0.1695,
"step": 383
},
{
"epoch": 0.37514883223935275,
"grad_norm": 1.0561014413833618,
"learning_rate": 4.6845651750323384e-06,
"loss": 0.1848,
"step": 384
},
{
"epoch": 0.3761257823233094,
"grad_norm": 1.7893115282058716,
"learning_rate": 4.6826395275883e-06,
"loss": 0.1745,
"step": 385
},
{
"epoch": 0.37710273240726605,
"grad_norm": 0.9349923729896545,
"learning_rate": 4.680708418502853e-06,
"loss": 0.1781,
"step": 386
},
{
"epoch": 0.3780796824912227,
"grad_norm": 0.8574498295783997,
"learning_rate": 4.678771852608232e-06,
"loss": 0.1556,
"step": 387
},
{
"epoch": 0.37905663257517935,
"grad_norm": 0.7922853827476501,
"learning_rate": 4.676829834750326e-06,
"loss": 0.1725,
"step": 388
},
{
"epoch": 0.380033582659136,
"grad_norm": 0.9478686451911926,
"learning_rate": 4.674882369788667e-06,
"loss": 0.1604,
"step": 389
},
{
"epoch": 0.38101053274309266,
"grad_norm": 1.0247634649276733,
"learning_rate": 4.672929462596415e-06,
"loss": 0.1714,
"step": 390
},
{
"epoch": 0.3819874828270493,
"grad_norm": 1.026174783706665,
"learning_rate": 4.67097111806035e-06,
"loss": 0.1617,
"step": 391
},
{
"epoch": 0.38296443291100596,
"grad_norm": 0.8140632510185242,
"learning_rate": 4.669007341080859e-06,
"loss": 0.163,
"step": 392
},
{
"epoch": 0.3839413829949626,
"grad_norm": 0.6845609545707703,
"learning_rate": 4.6670381365719215e-06,
"loss": 0.1436,
"step": 393
},
{
"epoch": 0.38491833307891926,
"grad_norm": 0.905922532081604,
"learning_rate": 4.665063509461098e-06,
"loss": 0.1368,
"step": 394
},
{
"epoch": 0.3858952831628759,
"grad_norm": 1.1261165142059326,
"learning_rate": 4.663083464689516e-06,
"loss": 0.1679,
"step": 395
},
{
"epoch": 0.38687223324683256,
"grad_norm": 0.7991992235183716,
"learning_rate": 4.661098007211864e-06,
"loss": 0.1494,
"step": 396
},
{
"epoch": 0.3878491833307892,
"grad_norm": 1.2790718078613281,
"learning_rate": 4.659107141996374e-06,
"loss": 0.1524,
"step": 397
},
{
"epoch": 0.38882613341474587,
"grad_norm": 0.7661685347557068,
"learning_rate": 4.657110874024806e-06,
"loss": 0.1644,
"step": 398
},
{
"epoch": 0.38980308349870246,
"grad_norm": 0.8004336357116699,
"learning_rate": 4.655109208292443e-06,
"loss": 0.1641,
"step": 399
},
{
"epoch": 0.3907800335826591,
"grad_norm": 0.6391324996948242,
"learning_rate": 4.653102149808074e-06,
"loss": 0.1724,
"step": 400
},
{
"epoch": 0.39175698366661577,
"grad_norm": 0.9730057716369629,
"learning_rate": 4.6510897035939825e-06,
"loss": 0.183,
"step": 401
},
{
"epoch": 0.3927339337505724,
"grad_norm": 0.934459388256073,
"learning_rate": 4.649071874685933e-06,
"loss": 0.1592,
"step": 402
},
{
"epoch": 0.39371088383452907,
"grad_norm": 0.9212885499000549,
"learning_rate": 4.647048668133158e-06,
"loss": 0.1609,
"step": 403
},
{
"epoch": 0.3946878339184857,
"grad_norm": 1.1316171884536743,
"learning_rate": 4.645020088998352e-06,
"loss": 0.1986,
"step": 404
},
{
"epoch": 0.39566478400244237,
"grad_norm": 0.8710122108459473,
"learning_rate": 4.6429861423576454e-06,
"loss": 0.1696,
"step": 405
},
{
"epoch": 0.396641734086399,
"grad_norm": 1.0295352935791016,
"learning_rate": 4.640946833300607e-06,
"loss": 0.1656,
"step": 406
},
{
"epoch": 0.3976186841703557,
"grad_norm": 0.813359797000885,
"learning_rate": 4.638902166930218e-06,
"loss": 0.1681,
"step": 407
},
{
"epoch": 0.3985956342543123,
"grad_norm": 0.9952261447906494,
"learning_rate": 4.636852148362871e-06,
"loss": 0.1591,
"step": 408
},
{
"epoch": 0.399572584338269,
"grad_norm": 1.1479076147079468,
"learning_rate": 4.6347967827283455e-06,
"loss": 0.1725,
"step": 409
},
{
"epoch": 0.40054953442222563,
"grad_norm": 0.7456801533699036,
"learning_rate": 4.6327360751698075e-06,
"loss": 0.167,
"step": 410
},
{
"epoch": 0.4015264845061823,
"grad_norm": 0.9170630574226379,
"learning_rate": 4.630670030843785e-06,
"loss": 0.1638,
"step": 411
},
{
"epoch": 0.40250343459013893,
"grad_norm": 0.9080767035484314,
"learning_rate": 4.628598654920161e-06,
"loss": 0.1538,
"step": 412
},
{
"epoch": 0.4034803846740956,
"grad_norm": 1.0291121006011963,
"learning_rate": 4.626521952582164e-06,
"loss": 0.1828,
"step": 413
},
{
"epoch": 0.4044573347580522,
"grad_norm": 0.8592116832733154,
"learning_rate": 4.624439929026344e-06,
"loss": 0.1558,
"step": 414
},
{
"epoch": 0.40543428484200883,
"grad_norm": 0.8687212467193604,
"learning_rate": 4.622352589462573e-06,
"loss": 0.1548,
"step": 415
},
{
"epoch": 0.4064112349259655,
"grad_norm": 0.8444783091545105,
"learning_rate": 4.62025993911402e-06,
"loss": 0.1523,
"step": 416
},
{
"epoch": 0.40738818500992213,
"grad_norm": 0.7917346358299255,
"learning_rate": 4.618161983217148e-06,
"loss": 0.1668,
"step": 417
},
{
"epoch": 0.4083651350938788,
"grad_norm": 0.9798198342323303,
"learning_rate": 4.616058727021691e-06,
"loss": 0.159,
"step": 418
},
{
"epoch": 0.40934208517783544,
"grad_norm": 0.7648982405662537,
"learning_rate": 4.6139501757906515e-06,
"loss": 0.1563,
"step": 419
},
{
"epoch": 0.4103190352617921,
"grad_norm": 0.9149302840232849,
"learning_rate": 4.611836334800276e-06,
"loss": 0.171,
"step": 420
},
{
"epoch": 0.41129598534574874,
"grad_norm": 1.0029511451721191,
"learning_rate": 4.609717209340053e-06,
"loss": 0.1803,
"step": 421
},
{
"epoch": 0.4122729354297054,
"grad_norm": 1.6170082092285156,
"learning_rate": 4.607592804712691e-06,
"loss": 0.1795,
"step": 422
},
{
"epoch": 0.41324988551366204,
"grad_norm": 0.8813459873199463,
"learning_rate": 4.60546312623411e-06,
"loss": 0.1725,
"step": 423
},
{
"epoch": 0.4142268355976187,
"grad_norm": 1.0119832754135132,
"learning_rate": 4.603328179233426e-06,
"loss": 0.1503,
"step": 424
},
{
"epoch": 0.41520378568157534,
"grad_norm": 0.9550930857658386,
"learning_rate": 4.601187969052938e-06,
"loss": 0.1725,
"step": 425
},
{
"epoch": 0.416180735765532,
"grad_norm": 0.9823699593544006,
"learning_rate": 4.5990425010481185e-06,
"loss": 0.1796,
"step": 426
},
{
"epoch": 0.41715768584948865,
"grad_norm": 0.8281930685043335,
"learning_rate": 4.596891780587592e-06,
"loss": 0.1622,
"step": 427
},
{
"epoch": 0.4181346359334453,
"grad_norm": 1.0369951725006104,
"learning_rate": 4.594735813053131e-06,
"loss": 0.1711,
"step": 428
},
{
"epoch": 0.41911158601740195,
"grad_norm": 1.0193263292312622,
"learning_rate": 4.592574603839632e-06,
"loss": 0.1653,
"step": 429
},
{
"epoch": 0.42008853610135855,
"grad_norm": 0.7897890210151672,
"learning_rate": 4.590408158355113e-06,
"loss": 0.1718,
"step": 430
},
{
"epoch": 0.4210654861853152,
"grad_norm": 0.7788593769073486,
"learning_rate": 4.5882364820206925e-06,
"loss": 0.171,
"step": 431
},
{
"epoch": 0.42204243626927185,
"grad_norm": 0.9562849402427673,
"learning_rate": 4.586059580270579e-06,
"loss": 0.1583,
"step": 432
},
{
"epoch": 0.4230193863532285,
"grad_norm": 1.1443227529525757,
"learning_rate": 4.583877458552056e-06,
"loss": 0.1538,
"step": 433
},
{
"epoch": 0.42399633643718515,
"grad_norm": 0.907416045665741,
"learning_rate": 4.581690122325469e-06,
"loss": 0.1498,
"step": 434
},
{
"epoch": 0.4249732865211418,
"grad_norm": 0.9503728747367859,
"learning_rate": 4.579497577064213e-06,
"loss": 0.1776,
"step": 435
},
{
"epoch": 0.42595023660509845,
"grad_norm": 1.2248893976211548,
"learning_rate": 4.577299828254715e-06,
"loss": 0.1605,
"step": 436
},
{
"epoch": 0.4269271866890551,
"grad_norm": 0.9676153659820557,
"learning_rate": 4.575096881396425e-06,
"loss": 0.1669,
"step": 437
},
{
"epoch": 0.42790413677301176,
"grad_norm": 0.9138438701629639,
"learning_rate": 4.5728887420018e-06,
"loss": 0.1711,
"step": 438
},
{
"epoch": 0.4288810868569684,
"grad_norm": 0.8535164594650269,
"learning_rate": 4.57067541559629e-06,
"loss": 0.1637,
"step": 439
},
{
"epoch": 0.42985803694092506,
"grad_norm": 0.7810698747634888,
"learning_rate": 4.568456907718325e-06,
"loss": 0.1728,
"step": 440
},
{
"epoch": 0.4308349870248817,
"grad_norm": 0.7574284672737122,
"learning_rate": 4.566233223919299e-06,
"loss": 0.1526,
"step": 441
},
{
"epoch": 0.43181193710883836,
"grad_norm": 0.9540330171585083,
"learning_rate": 4.5640043697635595e-06,
"loss": 0.1481,
"step": 442
},
{
"epoch": 0.432788887192795,
"grad_norm": 0.6913356781005859,
"learning_rate": 4.561770350828391e-06,
"loss": 0.1576,
"step": 443
},
{
"epoch": 0.43376583727675166,
"grad_norm": 0.8490065336227417,
"learning_rate": 4.559531172704002e-06,
"loss": 0.1544,
"step": 444
},
{
"epoch": 0.4347427873607083,
"grad_norm": 1.0822558403015137,
"learning_rate": 4.557286840993512e-06,
"loss": 0.1828,
"step": 445
},
{
"epoch": 0.4357197374446649,
"grad_norm": 1.0955301523208618,
"learning_rate": 4.555037361312935e-06,
"loss": 0.1627,
"step": 446
},
{
"epoch": 0.43669668752862156,
"grad_norm": 1.0650818347930908,
"learning_rate": 4.552782739291168e-06,
"loss": 0.1625,
"step": 447
},
{
"epoch": 0.4376736376125782,
"grad_norm": 0.879410445690155,
"learning_rate": 4.5505229805699725e-06,
"loss": 0.1694,
"step": 448
},
{
"epoch": 0.43865058769653487,
"grad_norm": 0.6247472763061523,
"learning_rate": 4.548258090803968e-06,
"loss": 0.1343,
"step": 449
},
{
"epoch": 0.4396275377804915,
"grad_norm": 0.8280798196792603,
"learning_rate": 4.54598807566061e-06,
"loss": 0.138,
"step": 450
},
{
"epoch": 0.44060448786444817,
"grad_norm": 0.598173201084137,
"learning_rate": 4.543712940820183e-06,
"loss": 0.1535,
"step": 451
},
{
"epoch": 0.4415814379484048,
"grad_norm": 0.8917229175567627,
"learning_rate": 4.541432691975779e-06,
"loss": 0.1434,
"step": 452
},
{
"epoch": 0.44255838803236147,
"grad_norm": 0.8526452779769897,
"learning_rate": 4.539147334833288e-06,
"loss": 0.1638,
"step": 453
},
{
"epoch": 0.4435353381163181,
"grad_norm": 1.030854344367981,
"learning_rate": 4.536856875111383e-06,
"loss": 0.1672,
"step": 454
},
{
"epoch": 0.4445122882002748,
"grad_norm": 0.6388799548149109,
"learning_rate": 4.534561318541506e-06,
"loss": 0.146,
"step": 455
},
{
"epoch": 0.4454892382842314,
"grad_norm": 0.76018226146698,
"learning_rate": 4.532260670867851e-06,
"loss": 0.1477,
"step": 456
},
{
"epoch": 0.4464661883681881,
"grad_norm": 3.6600043773651123,
"learning_rate": 4.529954937847352e-06,
"loss": 0.176,
"step": 457
},
{
"epoch": 0.44744313845214473,
"grad_norm": 0.9332470893859863,
"learning_rate": 4.527644125249669e-06,
"loss": 0.165,
"step": 458
},
{
"epoch": 0.4484200885361014,
"grad_norm": 1.7258678674697876,
"learning_rate": 4.525328238857173e-06,
"loss": 0.1453,
"step": 459
},
{
"epoch": 0.44939703862005803,
"grad_norm": 0.7266802191734314,
"learning_rate": 4.523007284464928e-06,
"loss": 0.1461,
"step": 460
},
{
"epoch": 0.4503739887040146,
"grad_norm": 1.3688712120056152,
"learning_rate": 4.520681267880685e-06,
"loss": 0.1706,
"step": 461
},
{
"epoch": 0.4513509387879713,
"grad_norm": 1.2537286281585693,
"learning_rate": 4.5183501949248585e-06,
"loss": 0.1711,
"step": 462
},
{
"epoch": 0.45232788887192793,
"grad_norm": 1.1937775611877441,
"learning_rate": 4.516014071430515e-06,
"loss": 0.1847,
"step": 463
},
{
"epoch": 0.4533048389558846,
"grad_norm": 0.8803216814994812,
"learning_rate": 4.513672903243361e-06,
"loss": 0.1734,
"step": 464
},
{
"epoch": 0.45428178903984123,
"grad_norm": 0.6755242347717285,
"learning_rate": 4.5113266962217276e-06,
"loss": 0.1647,
"step": 465
},
{
"epoch": 0.4552587391237979,
"grad_norm": 1.0921308994293213,
"learning_rate": 4.50897545623655e-06,
"loss": 0.1558,
"step": 466
},
{
"epoch": 0.45623568920775454,
"grad_norm": 0.835395872592926,
"learning_rate": 4.506619189171363e-06,
"loss": 0.1638,
"step": 467
},
{
"epoch": 0.4572126392917112,
"grad_norm": 1.1466346979141235,
"learning_rate": 4.504257900922276e-06,
"loss": 0.158,
"step": 468
},
{
"epoch": 0.45818958937566784,
"grad_norm": 1.408265471458435,
"learning_rate": 4.501891597397966e-06,
"loss": 0.1601,
"step": 469
},
{
"epoch": 0.4591665394596245,
"grad_norm": 0.8094568252563477,
"learning_rate": 4.4995202845196574e-06,
"loss": 0.1593,
"step": 470
},
{
"epoch": 0.46014348954358114,
"grad_norm": 0.8101633191108704,
"learning_rate": 4.497143968221113e-06,
"loss": 0.1676,
"step": 471
},
{
"epoch": 0.4611204396275378,
"grad_norm": 1.7777295112609863,
"learning_rate": 4.4947626544486115e-06,
"loss": 0.1471,
"step": 472
},
{
"epoch": 0.46209738971149444,
"grad_norm": 0.6761342287063599,
"learning_rate": 4.49237634916094e-06,
"loss": 0.1473,
"step": 473
},
{
"epoch": 0.4630743397954511,
"grad_norm": 0.821212112903595,
"learning_rate": 4.489985058329375e-06,
"loss": 0.1747,
"step": 474
},
{
"epoch": 0.46405128987940775,
"grad_norm": 1.1308681964874268,
"learning_rate": 4.487588787937667e-06,
"loss": 0.1624,
"step": 475
},
{
"epoch": 0.4650282399633644,
"grad_norm": 0.7441355586051941,
"learning_rate": 4.48518754398203e-06,
"loss": 0.1558,
"step": 476
},
{
"epoch": 0.466005190047321,
"grad_norm": 0.5962588787078857,
"learning_rate": 4.48278133247112e-06,
"loss": 0.1508,
"step": 477
},
{
"epoch": 0.46698214013127765,
"grad_norm": 0.8312955498695374,
"learning_rate": 4.4803701594260256e-06,
"loss": 0.1879,
"step": 478
},
{
"epoch": 0.4679590902152343,
"grad_norm": 0.8293958902359009,
"learning_rate": 4.47795403088025e-06,
"loss": 0.153,
"step": 479
},
{
"epoch": 0.46893604029919095,
"grad_norm": 1.1809464693069458,
"learning_rate": 4.475532952879696e-06,
"loss": 0.1592,
"step": 480
},
{
"epoch": 0.4699129903831476,
"grad_norm": 0.8504133820533752,
"learning_rate": 4.473106931482653e-06,
"loss": 0.1602,
"step": 481
},
{
"epoch": 0.47088994046710425,
"grad_norm": 0.8525364398956299,
"learning_rate": 4.470675972759778e-06,
"loss": 0.1555,
"step": 482
},
{
"epoch": 0.4718668905510609,
"grad_norm": 0.9061461687088013,
"learning_rate": 4.468240082794087e-06,
"loss": 0.1638,
"step": 483
},
{
"epoch": 0.47284384063501755,
"grad_norm": 0.7759461402893066,
"learning_rate": 4.465799267680928e-06,
"loss": 0.158,
"step": 484
},
{
"epoch": 0.4738207907189742,
"grad_norm": 1.092197299003601,
"learning_rate": 4.463353533527981e-06,
"loss": 0.1733,
"step": 485
},
{
"epoch": 0.47479774080293086,
"grad_norm": 0.9862155318260193,
"learning_rate": 4.4609028864552315e-06,
"loss": 0.1565,
"step": 486
},
{
"epoch": 0.4757746908868875,
"grad_norm": 1.069003939628601,
"learning_rate": 4.458447332594959e-06,
"loss": 0.1751,
"step": 487
},
{
"epoch": 0.47675164097084416,
"grad_norm": 1.0645431280136108,
"learning_rate": 4.45598687809172e-06,
"loss": 0.1825,
"step": 488
},
{
"epoch": 0.4777285910548008,
"grad_norm": 0.8188602328300476,
"learning_rate": 4.453521529102337e-06,
"loss": 0.1424,
"step": 489
},
{
"epoch": 0.47870554113875746,
"grad_norm": 0.8234372138977051,
"learning_rate": 4.451051291795877e-06,
"loss": 0.1543,
"step": 490
},
{
"epoch": 0.4796824912227141,
"grad_norm": 0.7755301594734192,
"learning_rate": 4.44857617235364e-06,
"loss": 0.1758,
"step": 491
},
{
"epoch": 0.48065944130667076,
"grad_norm": 0.7147533297538757,
"learning_rate": 4.446096176969144e-06,
"loss": 0.1697,
"step": 492
},
{
"epoch": 0.48163639139062736,
"grad_norm": 0.7931758165359497,
"learning_rate": 4.443611311848106e-06,
"loss": 0.1524,
"step": 493
},
{
"epoch": 0.482613341474584,
"grad_norm": 0.98976731300354,
"learning_rate": 4.441121583208431e-06,
"loss": 0.1605,
"step": 494
},
{
"epoch": 0.48359029155854066,
"grad_norm": 1.1909774541854858,
"learning_rate": 4.438626997280191e-06,
"loss": 0.2115,
"step": 495
},
{
"epoch": 0.4845672416424973,
"grad_norm": 0.7883973121643066,
"learning_rate": 4.436127560305615e-06,
"loss": 0.1642,
"step": 496
},
{
"epoch": 0.48554419172645397,
"grad_norm": 0.7579115033149719,
"learning_rate": 4.4336232785390706e-06,
"loss": 0.1622,
"step": 497
},
{
"epoch": 0.4865211418104106,
"grad_norm": 0.8877824544906616,
"learning_rate": 4.431114158247047e-06,
"loss": 0.1528,
"step": 498
},
{
"epoch": 0.48749809189436727,
"grad_norm": 1.4241032600402832,
"learning_rate": 4.428600205708142e-06,
"loss": 0.1598,
"step": 499
},
{
"epoch": 0.4884750419783239,
"grad_norm": 0.9293645024299622,
"learning_rate": 4.4260814272130435e-06,
"loss": 0.1739,
"step": 500
},
{
"epoch": 0.48945199206228057,
"grad_norm": 0.7388932704925537,
"learning_rate": 4.42355782906452e-06,
"loss": 0.1662,
"step": 501
},
{
"epoch": 0.4904289421462372,
"grad_norm": 1.6770546436309814,
"learning_rate": 4.421029417577394e-06,
"loss": 0.1732,
"step": 502
},
{
"epoch": 0.4914058922301939,
"grad_norm": 0.9671643376350403,
"learning_rate": 4.418496199078538e-06,
"loss": 0.1668,
"step": 503
},
{
"epoch": 0.4923828423141505,
"grad_norm": 0.8135396838188171,
"learning_rate": 4.415958179906851e-06,
"loss": 0.1532,
"step": 504
},
{
"epoch": 0.4933597923981072,
"grad_norm": 0.753450870513916,
"learning_rate": 4.413415366413242e-06,
"loss": 0.1581,
"step": 505
},
{
"epoch": 0.49433674248206383,
"grad_norm": 1.3733445405960083,
"learning_rate": 4.410867764960622e-06,
"loss": 0.16,
"step": 506
},
{
"epoch": 0.4953136925660205,
"grad_norm": 1.0689316987991333,
"learning_rate": 4.408315381923878e-06,
"loss": 0.1525,
"step": 507
},
{
"epoch": 0.4962906426499771,
"grad_norm": 0.7836982607841492,
"learning_rate": 4.405758223689867e-06,
"loss": 0.1525,
"step": 508
},
{
"epoch": 0.4972675927339337,
"grad_norm": 1.3173394203186035,
"learning_rate": 4.4031962966573905e-06,
"loss": 0.171,
"step": 509
},
{
"epoch": 0.4982445428178904,
"grad_norm": 0.8582804799079895,
"learning_rate": 4.4006296072371855e-06,
"loss": 0.1607,
"step": 510
},
{
"epoch": 0.49922149290184703,
"grad_norm": 0.7594783902168274,
"learning_rate": 4.3980581618519055e-06,
"loss": 0.1554,
"step": 511
},
{
"epoch": 0.5001984429858037,
"grad_norm": 0.9228777885437012,
"learning_rate": 4.395481966936106e-06,
"loss": 0.1751,
"step": 512
},
{
"epoch": 0.5011753930697603,
"grad_norm": 0.7919161915779114,
"learning_rate": 4.392901028936224e-06,
"loss": 0.1741,
"step": 513
},
{
"epoch": 0.502152343153717,
"grad_norm": 1.94724702835083,
"learning_rate": 4.390315354310568e-06,
"loss": 0.1628,
"step": 514
},
{
"epoch": 0.5031292932376736,
"grad_norm": 1.2179944515228271,
"learning_rate": 4.387724949529298e-06,
"loss": 0.1521,
"step": 515
},
{
"epoch": 0.5041062433216302,
"grad_norm": 1.060072422027588,
"learning_rate": 4.385129821074413e-06,
"loss": 0.1506,
"step": 516
},
{
"epoch": 0.5050831934055869,
"grad_norm": 0.9509943723678589,
"learning_rate": 4.382529975439727e-06,
"loss": 0.1641,
"step": 517
},
{
"epoch": 0.5060601434895435,
"grad_norm": 0.6980380415916443,
"learning_rate": 4.379925419130862e-06,
"loss": 0.1706,
"step": 518
},
{
"epoch": 0.5070370935735002,
"grad_norm": 1.4049944877624512,
"learning_rate": 4.377316158665224e-06,
"loss": 0.1708,
"step": 519
},
{
"epoch": 0.5080140436574568,
"grad_norm": 0.6877716779708862,
"learning_rate": 4.374702200571996e-06,
"loss": 0.1534,
"step": 520
},
{
"epoch": 0.5089909937414135,
"grad_norm": 1.0414890050888062,
"learning_rate": 4.372083551392108e-06,
"loss": 0.1568,
"step": 521
},
{
"epoch": 0.5099679438253701,
"grad_norm": 0.987546443939209,
"learning_rate": 4.3694602176782365e-06,
"loss": 0.1507,
"step": 522
},
{
"epoch": 0.5109448939093268,
"grad_norm": 0.7913284301757812,
"learning_rate": 4.366832205994774e-06,
"loss": 0.1684,
"step": 523
},
{
"epoch": 0.5119218439932834,
"grad_norm": 0.7315497398376465,
"learning_rate": 4.364199522917824e-06,
"loss": 0.1628,
"step": 524
},
{
"epoch": 0.5128987940772401,
"grad_norm": 0.751228392124176,
"learning_rate": 4.3615621750351735e-06,
"loss": 0.1555,
"step": 525
},
{
"epoch": 0.5138757441611967,
"grad_norm": 1.2990520000457764,
"learning_rate": 4.358920168946288e-06,
"loss": 0.1735,
"step": 526
},
{
"epoch": 0.5148526942451535,
"grad_norm": 0.9432131052017212,
"learning_rate": 4.356273511262283e-06,
"loss": 0.1551,
"step": 527
},
{
"epoch": 0.51582964432911,
"grad_norm": 1.6193960905075073,
"learning_rate": 4.353622208605922e-06,
"loss": 0.1661,
"step": 528
},
{
"epoch": 0.5168065944130668,
"grad_norm": 0.7774462699890137,
"learning_rate": 4.350966267611584e-06,
"loss": 0.1552,
"step": 529
},
{
"epoch": 0.5177835444970234,
"grad_norm": 1.0754817724227905,
"learning_rate": 4.348305694925258e-06,
"loss": 0.1563,
"step": 530
},
{
"epoch": 0.5187604945809801,
"grad_norm": 1.0061575174331665,
"learning_rate": 4.345640497204522e-06,
"loss": 0.1597,
"step": 531
},
{
"epoch": 0.5197374446649367,
"grad_norm": 0.8242843151092529,
"learning_rate": 4.342970681118529e-06,
"loss": 0.1731,
"step": 532
},
{
"epoch": 0.5207143947488933,
"grad_norm": 0.8342220783233643,
"learning_rate": 4.340296253347986e-06,
"loss": 0.1797,
"step": 533
},
{
"epoch": 0.52169134483285,
"grad_norm": 0.8062166571617126,
"learning_rate": 4.337617220585143e-06,
"loss": 0.1614,
"step": 534
},
{
"epoch": 0.5226682949168066,
"grad_norm": 0.8868542909622192,
"learning_rate": 4.334933589533769e-06,
"loss": 0.1653,
"step": 535
},
{
"epoch": 0.5236452450007633,
"grad_norm": 1.0182602405548096,
"learning_rate": 4.332245366909144e-06,
"loss": 0.1436,
"step": 536
},
{
"epoch": 0.5246221950847199,
"grad_norm": 0.8037219643592834,
"learning_rate": 4.329552559438034e-06,
"loss": 0.1511,
"step": 537
},
{
"epoch": 0.5255991451686766,
"grad_norm": 1.2997537851333618,
"learning_rate": 4.32685517385868e-06,
"loss": 0.1687,
"step": 538
},
{
"epoch": 0.5265760952526332,
"grad_norm": 0.9416012167930603,
"learning_rate": 4.324153216920775e-06,
"loss": 0.1525,
"step": 539
},
{
"epoch": 0.5275530453365899,
"grad_norm": 1.6582372188568115,
"learning_rate": 4.321446695385456e-06,
"loss": 0.161,
"step": 540
},
{
"epoch": 0.5285299954205465,
"grad_norm": 1.245970368385315,
"learning_rate": 4.31873561602528e-06,
"loss": 0.1466,
"step": 541
},
{
"epoch": 0.5295069455045032,
"grad_norm": 0.8859710097312927,
"learning_rate": 4.316019985624208e-06,
"loss": 0.1428,
"step": 542
},
{
"epoch": 0.5304838955884598,
"grad_norm": 0.83238285779953,
"learning_rate": 4.313299810977589e-06,
"loss": 0.1474,
"step": 543
},
{
"epoch": 0.5314608456724165,
"grad_norm": 0.6943266987800598,
"learning_rate": 4.310575098892147e-06,
"loss": 0.1622,
"step": 544
},
{
"epoch": 0.5324377957563731,
"grad_norm": 0.5472984910011292,
"learning_rate": 4.307845856185953e-06,
"loss": 0.1488,
"step": 545
},
{
"epoch": 0.5334147458403298,
"grad_norm": 0.9675540328025818,
"learning_rate": 4.305112089688423e-06,
"loss": 0.1516,
"step": 546
},
{
"epoch": 0.5343916959242864,
"grad_norm": 0.903506338596344,
"learning_rate": 4.302373806240289e-06,
"loss": 0.1738,
"step": 547
},
{
"epoch": 0.535368646008243,
"grad_norm": 0.8674273490905762,
"learning_rate": 4.2996310126935825e-06,
"loss": 0.1589,
"step": 548
},
{
"epoch": 0.5363455960921997,
"grad_norm": 0.9862509369850159,
"learning_rate": 4.2968837159116275e-06,
"loss": 0.1598,
"step": 549
},
{
"epoch": 0.5373225461761563,
"grad_norm": 0.8058833479881287,
"learning_rate": 4.294131922769011e-06,
"loss": 0.1561,
"step": 550
},
{
"epoch": 0.538299496260113,
"grad_norm": 0.9268854260444641,
"learning_rate": 4.291375640151574e-06,
"loss": 0.1623,
"step": 551
},
{
"epoch": 0.5392764463440696,
"grad_norm": 0.9943864345550537,
"learning_rate": 4.288614874956391e-06,
"loss": 0.1717,
"step": 552
},
{
"epoch": 0.5402533964280263,
"grad_norm": 0.7909972667694092,
"learning_rate": 4.285849634091753e-06,
"loss": 0.1565,
"step": 553
},
{
"epoch": 0.5412303465119829,
"grad_norm": 1.182213544845581,
"learning_rate": 4.2830799244771524e-06,
"loss": 0.1781,
"step": 554
},
{
"epoch": 0.5422072965959396,
"grad_norm": 0.7538149952888489,
"learning_rate": 4.280305753043259e-06,
"loss": 0.1572,
"step": 555
},
{
"epoch": 0.5431842466798962,
"grad_norm": 0.652836263179779,
"learning_rate": 4.277527126731914e-06,
"loss": 0.1434,
"step": 556
},
{
"epoch": 0.5441611967638529,
"grad_norm": 0.7350523471832275,
"learning_rate": 4.2747440524961005e-06,
"loss": 0.1508,
"step": 557
},
{
"epoch": 0.5451381468478095,
"grad_norm": 1.0987657308578491,
"learning_rate": 4.271956537299934e-06,
"loss": 0.1655,
"step": 558
},
{
"epoch": 0.5461150969317662,
"grad_norm": 0.9032847881317139,
"learning_rate": 4.2691645881186435e-06,
"loss": 0.1615,
"step": 559
},
{
"epoch": 0.5470920470157228,
"grad_norm": 0.7912704944610596,
"learning_rate": 4.266368211938551e-06,
"loss": 0.1556,
"step": 560
},
{
"epoch": 0.5480689970996795,
"grad_norm": 0.7764425277709961,
"learning_rate": 4.263567415757058e-06,
"loss": 0.1567,
"step": 561
},
{
"epoch": 0.5490459471836361,
"grad_norm": 0.8339623808860779,
"learning_rate": 4.260762206582627e-06,
"loss": 0.1785,
"step": 562
},
{
"epoch": 0.5500228972675927,
"grad_norm": 0.8981364369392395,
"learning_rate": 4.25795259143476e-06,
"loss": 0.1604,
"step": 563
},
{
"epoch": 0.5509998473515494,
"grad_norm": 1.2198928594589233,
"learning_rate": 4.255138577343987e-06,
"loss": 0.1577,
"step": 564
},
{
"epoch": 0.551976797435506,
"grad_norm": 1.3460452556610107,
"learning_rate": 4.2523201713518434e-06,
"loss": 0.1599,
"step": 565
},
{
"epoch": 0.5529537475194627,
"grad_norm": 0.6532039642333984,
"learning_rate": 4.249497380510857e-06,
"loss": 0.1498,
"step": 566
},
{
"epoch": 0.5539306976034193,
"grad_norm": 1.4958688020706177,
"learning_rate": 4.246670211884527e-06,
"loss": 0.1695,
"step": 567
},
{
"epoch": 0.554907647687376,
"grad_norm": 0.9645088315010071,
"learning_rate": 4.243838672547303e-06,
"loss": 0.1676,
"step": 568
},
{
"epoch": 0.5558845977713326,
"grad_norm": 0.902220606803894,
"learning_rate": 4.241002769584579e-06,
"loss": 0.1442,
"step": 569
},
{
"epoch": 0.5568615478552893,
"grad_norm": 0.6822658181190491,
"learning_rate": 4.238162510092663e-06,
"loss": 0.1437,
"step": 570
},
{
"epoch": 0.5578384979392459,
"grad_norm": 0.6158772706985474,
"learning_rate": 4.235317901178765e-06,
"loss": 0.1357,
"step": 571
},
{
"epoch": 0.5588154480232026,
"grad_norm": 0.6255518794059753,
"learning_rate": 4.23246894996098e-06,
"loss": 0.1544,
"step": 572
},
{
"epoch": 0.5597923981071592,
"grad_norm": 0.9218140244483948,
"learning_rate": 4.229615663568266e-06,
"loss": 0.1733,
"step": 573
},
{
"epoch": 0.5607693481911159,
"grad_norm": 0.9581194519996643,
"learning_rate": 4.226758049140434e-06,
"loss": 0.1654,
"step": 574
},
{
"epoch": 0.5617462982750725,
"grad_norm": 0.6818786263465881,
"learning_rate": 4.2238961138281205e-06,
"loss": 0.1485,
"step": 575
},
{
"epoch": 0.5627232483590292,
"grad_norm": 0.8091064095497131,
"learning_rate": 4.221029864792776e-06,
"loss": 0.1501,
"step": 576
},
{
"epoch": 0.5637001984429858,
"grad_norm": 0.6479501128196716,
"learning_rate": 4.218159309206645e-06,
"loss": 0.1453,
"step": 577
},
{
"epoch": 0.5646771485269424,
"grad_norm": 0.8743746280670166,
"learning_rate": 4.2152844542527485e-06,
"loss": 0.1585,
"step": 578
},
{
"epoch": 0.5656540986108991,
"grad_norm": 0.8726969361305237,
"learning_rate": 4.212405307124865e-06,
"loss": 0.1521,
"step": 579
},
{
"epoch": 0.5666310486948557,
"grad_norm": 1.0019125938415527,
"learning_rate": 4.209521875027515e-06,
"loss": 0.1758,
"step": 580
},
{
"epoch": 0.5676079987788124,
"grad_norm": 0.8543357253074646,
"learning_rate": 4.2066341651759404e-06,
"loss": 0.1533,
"step": 581
},
{
"epoch": 0.568584948862769,
"grad_norm": 0.9275641441345215,
"learning_rate": 4.203742184796085e-06,
"loss": 0.1469,
"step": 582
},
{
"epoch": 0.5695618989467257,
"grad_norm": 1.0211560726165771,
"learning_rate": 4.200845941124585e-06,
"loss": 0.1591,
"step": 583
},
{
"epoch": 0.5705388490306823,
"grad_norm": 0.9742165803909302,
"learning_rate": 4.197945441408739e-06,
"loss": 0.1524,
"step": 584
},
{
"epoch": 0.571515799114639,
"grad_norm": 0.7164245247840881,
"learning_rate": 4.195040692906497e-06,
"loss": 0.1719,
"step": 585
},
{
"epoch": 0.5724927491985956,
"grad_norm": 0.8360973000526428,
"learning_rate": 4.19213170288644e-06,
"loss": 0.1676,
"step": 586
},
{
"epoch": 0.5734696992825523,
"grad_norm": 0.8007204532623291,
"learning_rate": 4.189218478627767e-06,
"loss": 0.1453,
"step": 587
},
{
"epoch": 0.5744466493665089,
"grad_norm": 0.723944902420044,
"learning_rate": 4.186301027420266e-06,
"loss": 0.1354,
"step": 588
},
{
"epoch": 0.5754235994504656,
"grad_norm": 1.2359501123428345,
"learning_rate": 4.183379356564308e-06,
"loss": 0.1632,
"step": 589
},
{
"epoch": 0.5764005495344222,
"grad_norm": 0.7834674119949341,
"learning_rate": 4.180453473370818e-06,
"loss": 0.1511,
"step": 590
},
{
"epoch": 0.5773774996183789,
"grad_norm": 0.7616752982139587,
"learning_rate": 4.177523385161264e-06,
"loss": 0.1542,
"step": 591
},
{
"epoch": 0.5783544497023355,
"grad_norm": 0.9482843279838562,
"learning_rate": 4.174589099267636e-06,
"loss": 0.1481,
"step": 592
},
{
"epoch": 0.5793313997862922,
"grad_norm": 2.3494958877563477,
"learning_rate": 4.171650623032429e-06,
"loss": 0.1751,
"step": 593
},
{
"epoch": 0.5803083498702488,
"grad_norm": 0.9331839680671692,
"learning_rate": 4.1687079638086215e-06,
"loss": 0.1446,
"step": 594
},
{
"epoch": 0.5812852999542054,
"grad_norm": 0.7390409708023071,
"learning_rate": 4.16576112895966e-06,
"loss": 0.1544,
"step": 595
},
{
"epoch": 0.5822622500381621,
"grad_norm": 1.0247509479522705,
"learning_rate": 4.16281012585944e-06,
"loss": 0.1472,
"step": 596
},
{
"epoch": 0.5832392001221187,
"grad_norm": 0.7131674289703369,
"learning_rate": 4.159854961892287e-06,
"loss": 0.1555,
"step": 597
},
{
"epoch": 0.5842161502060754,
"grad_norm": 0.6914873719215393,
"learning_rate": 4.156895644452937e-06,
"loss": 0.1598,
"step": 598
},
{
"epoch": 0.585193100290032,
"grad_norm": 1.2369025945663452,
"learning_rate": 4.1539321809465225e-06,
"loss": 0.168,
"step": 599
},
{
"epoch": 0.5861700503739887,
"grad_norm": 0.783982515335083,
"learning_rate": 4.1509645787885485e-06,
"loss": 0.1611,
"step": 600
},
{
"epoch": 0.5871470004579453,
"grad_norm": 0.8376875519752502,
"learning_rate": 4.147992845404876e-06,
"loss": 0.1693,
"step": 601
},
{
"epoch": 0.588123950541902,
"grad_norm": 1.4471428394317627,
"learning_rate": 4.145016988231704e-06,
"loss": 0.1831,
"step": 602
},
{
"epoch": 0.5891009006258586,
"grad_norm": 0.9096348881721497,
"learning_rate": 4.14203701471555e-06,
"loss": 0.1519,
"step": 603
},
{
"epoch": 0.5900778507098153,
"grad_norm": 0.7935978174209595,
"learning_rate": 4.1390529323132335e-06,
"loss": 0.1652,
"step": 604
},
{
"epoch": 0.5910548007937719,
"grad_norm": 0.6195600628852844,
"learning_rate": 4.136064748491855e-06,
"loss": 0.1366,
"step": 605
},
{
"epoch": 0.5920317508777286,
"grad_norm": 0.8138124942779541,
"learning_rate": 4.133072470728776e-06,
"loss": 0.1492,
"step": 606
},
{
"epoch": 0.5930087009616852,
"grad_norm": 0.5873351693153381,
"learning_rate": 4.130076106511605e-06,
"loss": 0.1381,
"step": 607
},
{
"epoch": 0.5939856510456419,
"grad_norm": 0.9016026854515076,
"learning_rate": 4.127075663338177e-06,
"loss": 0.1532,
"step": 608
},
{
"epoch": 0.5949626011295985,
"grad_norm": 0.9960969686508179,
"learning_rate": 4.124071148716529e-06,
"loss": 0.1605,
"step": 609
},
{
"epoch": 0.5959395512135551,
"grad_norm": 0.9086290001869202,
"learning_rate": 4.12106257016489e-06,
"loss": 0.1607,
"step": 610
},
{
"epoch": 0.5969165012975118,
"grad_norm": 0.7733098864555359,
"learning_rate": 4.118049935211657e-06,
"loss": 0.1646,
"step": 611
},
{
"epoch": 0.5978934513814684,
"grad_norm": 0.8592734336853027,
"learning_rate": 4.115033251395378e-06,
"loss": 0.1593,
"step": 612
},
{
"epoch": 0.5988704014654251,
"grad_norm": 1.0230865478515625,
"learning_rate": 4.112012526264731e-06,
"loss": 0.1562,
"step": 613
},
{
"epoch": 0.5998473515493817,
"grad_norm": 0.8900034427642822,
"learning_rate": 4.10898776737851e-06,
"loss": 0.1643,
"step": 614
},
{
"epoch": 0.6008243016333384,
"grad_norm": 0.8863382935523987,
"learning_rate": 4.105958982305598e-06,
"loss": 0.1537,
"step": 615
},
{
"epoch": 0.601801251717295,
"grad_norm": 0.8836170434951782,
"learning_rate": 4.102926178624956e-06,
"loss": 0.1587,
"step": 616
},
{
"epoch": 0.6027782018012517,
"grad_norm": 0.7207074165344238,
"learning_rate": 4.0998893639256e-06,
"loss": 0.1549,
"step": 617
},
{
"epoch": 0.6037551518852083,
"grad_norm": 0.8067906498908997,
"learning_rate": 4.096848545806582e-06,
"loss": 0.1505,
"step": 618
},
{
"epoch": 0.604732101969165,
"grad_norm": 0.7560562491416931,
"learning_rate": 4.093803731876973e-06,
"loss": 0.1473,
"step": 619
},
{
"epoch": 0.6057090520531216,
"grad_norm": 0.8147042989730835,
"learning_rate": 4.090754929755843e-06,
"loss": 0.1495,
"step": 620
},
{
"epoch": 0.6066860021370784,
"grad_norm": 0.7013716697692871,
"learning_rate": 4.087702147072241e-06,
"loss": 0.1386,
"step": 621
},
{
"epoch": 0.607662952221035,
"grad_norm": 0.8274866342544556,
"learning_rate": 4.084645391465175e-06,
"loss": 0.1503,
"step": 622
},
{
"epoch": 0.6086399023049917,
"grad_norm": 0.7079525589942932,
"learning_rate": 4.081584670583598e-06,
"loss": 0.1722,
"step": 623
},
{
"epoch": 0.6096168523889482,
"grad_norm": 0.7784177660942078,
"learning_rate": 4.078519992086382e-06,
"loss": 0.1628,
"step": 624
},
{
"epoch": 0.6105938024729048,
"grad_norm": 0.6801629066467285,
"learning_rate": 4.075451363642304e-06,
"loss": 0.1476,
"step": 625
},
{
"epoch": 0.6115707525568616,
"grad_norm": 0.9884294271469116,
"learning_rate": 4.072378792930026e-06,
"loss": 0.1705,
"step": 626
},
{
"epoch": 0.6125477026408181,
"grad_norm": 1.2903105020523071,
"learning_rate": 4.069302287638071e-06,
"loss": 0.1957,
"step": 627
},
{
"epoch": 0.6135246527247749,
"grad_norm": 1.4761219024658203,
"learning_rate": 4.066221855464812e-06,
"loss": 0.1699,
"step": 628
},
{
"epoch": 0.6145016028087315,
"grad_norm": 0.7803857326507568,
"learning_rate": 4.0631375041184455e-06,
"loss": 0.1966,
"step": 629
},
{
"epoch": 0.6154785528926882,
"grad_norm": 0.7521586418151855,
"learning_rate": 4.060049241316975e-06,
"loss": 0.147,
"step": 630
},
{
"epoch": 0.6164555029766448,
"grad_norm": 1.1016544103622437,
"learning_rate": 4.056957074788194e-06,
"loss": 0.1589,
"step": 631
},
{
"epoch": 0.6174324530606015,
"grad_norm": 0.714637041091919,
"learning_rate": 4.0538610122696604e-06,
"loss": 0.1561,
"step": 632
},
{
"epoch": 0.6184094031445581,
"grad_norm": 0.6871845126152039,
"learning_rate": 4.050761061508685e-06,
"loss": 0.1511,
"step": 633
},
{
"epoch": 0.6193863532285148,
"grad_norm": 0.8859542608261108,
"learning_rate": 4.047657230262306e-06,
"loss": 0.1622,
"step": 634
},
{
"epoch": 0.6203633033124714,
"grad_norm": 0.6352054476737976,
"learning_rate": 4.0445495262972724e-06,
"loss": 0.1474,
"step": 635
},
{
"epoch": 0.6213402533964281,
"grad_norm": 0.818693995475769,
"learning_rate": 4.041437957390025e-06,
"loss": 0.1432,
"step": 636
},
{
"epoch": 0.6223172034803847,
"grad_norm": 0.6814315319061279,
"learning_rate": 4.038322531326673e-06,
"loss": 0.1655,
"step": 637
},
{
"epoch": 0.6232941535643414,
"grad_norm": 0.9554708003997803,
"learning_rate": 4.035203255902981e-06,
"loss": 0.1619,
"step": 638
},
{
"epoch": 0.624271103648298,
"grad_norm": 1.0255087614059448,
"learning_rate": 4.032080138924341e-06,
"loss": 0.1656,
"step": 639
},
{
"epoch": 0.6252480537322547,
"grad_norm": 0.8935983777046204,
"learning_rate": 4.028953188205765e-06,
"loss": 0.1649,
"step": 640
},
{
"epoch": 0.6262250038162113,
"grad_norm": 1.1148182153701782,
"learning_rate": 4.025822411571851e-06,
"loss": 0.1446,
"step": 641
},
{
"epoch": 0.6272019539001679,
"grad_norm": 1.0702588558197021,
"learning_rate": 4.022687816856775e-06,
"loss": 0.1503,
"step": 642
},
{
"epoch": 0.6281789039841246,
"grad_norm": 0.822248101234436,
"learning_rate": 4.019549411904265e-06,
"loss": 0.1495,
"step": 643
},
{
"epoch": 0.6291558540680812,
"grad_norm": 0.7461888194084167,
"learning_rate": 4.016407204567585e-06,
"loss": 0.1445,
"step": 644
},
{
"epoch": 0.6301328041520379,
"grad_norm": 0.7315708994865417,
"learning_rate": 4.013261202709512e-06,
"loss": 0.1415,
"step": 645
},
{
"epoch": 0.6311097542359945,
"grad_norm": 0.7259897589683533,
"learning_rate": 4.010111414202319e-06,
"loss": 0.1583,
"step": 646
},
{
"epoch": 0.6320867043199512,
"grad_norm": 0.6060989499092102,
"learning_rate": 4.006957846927755e-06,
"loss": 0.1492,
"step": 647
},
{
"epoch": 0.6330636544039078,
"grad_norm": 1.101295828819275,
"learning_rate": 4.0038005087770225e-06,
"loss": 0.1698,
"step": 648
},
{
"epoch": 0.6340406044878645,
"grad_norm": 1.2705976963043213,
"learning_rate": 4.0006394076507625e-06,
"loss": 0.1642,
"step": 649
},
{
"epoch": 0.6350175545718211,
"grad_norm": 0.9330812692642212,
"learning_rate": 3.997474551459031e-06,
"loss": 0.1517,
"step": 650
},
{
"epoch": 0.6359945046557778,
"grad_norm": 0.6551914811134338,
"learning_rate": 3.99430594812128e-06,
"loss": 0.1662,
"step": 651
},
{
"epoch": 0.6369714547397344,
"grad_norm": 1.0914472341537476,
"learning_rate": 3.991133605566338e-06,
"loss": 0.1601,
"step": 652
},
{
"epoch": 0.6379484048236911,
"grad_norm": 0.736682653427124,
"learning_rate": 3.98795753173239e-06,
"loss": 0.1542,
"step": 653
},
{
"epoch": 0.6389253549076477,
"grad_norm": 0.7676042914390564,
"learning_rate": 3.98477773456696e-06,
"loss": 0.1473,
"step": 654
},
{
"epoch": 0.6399023049916044,
"grad_norm": 0.6700080037117004,
"learning_rate": 3.981594222026886e-06,
"loss": 0.1342,
"step": 655
},
{
"epoch": 0.640879255075561,
"grad_norm": 0.9363613724708557,
"learning_rate": 3.978407002078305e-06,
"loss": 0.1299,
"step": 656
},
{
"epoch": 0.6418562051595176,
"grad_norm": 0.9178924560546875,
"learning_rate": 3.97521608269663e-06,
"loss": 0.1581,
"step": 657
},
{
"epoch": 0.6428331552434743,
"grad_norm": 0.7020032405853271,
"learning_rate": 3.972021471866531e-06,
"loss": 0.1535,
"step": 658
},
{
"epoch": 0.6438101053274309,
"grad_norm": 0.8765652775764465,
"learning_rate": 3.968823177581916e-06,
"loss": 0.1725,
"step": 659
},
{
"epoch": 0.6447870554113876,
"grad_norm": 1.2579203844070435,
"learning_rate": 3.965621207845911e-06,
"loss": 0.1537,
"step": 660
},
{
"epoch": 0.6457640054953442,
"grad_norm": 0.8908601999282837,
"learning_rate": 3.962415570670836e-06,
"loss": 0.1625,
"step": 661
},
{
"epoch": 0.6467409555793009,
"grad_norm": 1.2500725984573364,
"learning_rate": 3.959206274078191e-06,
"loss": 0.149,
"step": 662
},
{
"epoch": 0.6477179056632575,
"grad_norm": 0.8343613147735596,
"learning_rate": 3.9559933260986315e-06,
"loss": 0.1482,
"step": 663
},
{
"epoch": 0.6486948557472142,
"grad_norm": 0.8883591890335083,
"learning_rate": 3.952776734771951e-06,
"loss": 0.1516,
"step": 664
},
{
"epoch": 0.6496718058311708,
"grad_norm": 1.1744972467422485,
"learning_rate": 3.949556508147059e-06,
"loss": 0.1443,
"step": 665
},
{
"epoch": 0.6506487559151275,
"grad_norm": 1.3209812641143799,
"learning_rate": 3.946332654281961e-06,
"loss": 0.1487,
"step": 666
},
{
"epoch": 0.6516257059990841,
"grad_norm": 1.0948824882507324,
"learning_rate": 3.94310518124374e-06,
"loss": 0.1613,
"step": 667
},
{
"epoch": 0.6526026560830408,
"grad_norm": 0.9042884111404419,
"learning_rate": 3.939874097108537e-06,
"loss": 0.1443,
"step": 668
},
{
"epoch": 0.6535796061669974,
"grad_norm": 0.7743211388587952,
"learning_rate": 3.936639409961524e-06,
"loss": 0.1403,
"step": 669
},
{
"epoch": 0.6545565562509541,
"grad_norm": 0.8176175951957703,
"learning_rate": 3.933401127896894e-06,
"loss": 0.1635,
"step": 670
},
{
"epoch": 0.6555335063349107,
"grad_norm": 0.8969285488128662,
"learning_rate": 3.930159259017834e-06,
"loss": 0.1692,
"step": 671
},
{
"epoch": 0.6565104564188673,
"grad_norm": 0.744179368019104,
"learning_rate": 3.9269138114365056e-06,
"loss": 0.1779,
"step": 672
},
{
"epoch": 0.657487406502824,
"grad_norm": 0.7474929094314575,
"learning_rate": 3.923664793274024e-06,
"loss": 0.1367,
"step": 673
},
{
"epoch": 0.6584643565867806,
"grad_norm": 0.8669520616531372,
"learning_rate": 3.920412212660444e-06,
"loss": 0.155,
"step": 674
},
{
"epoch": 0.6594413066707373,
"grad_norm": 0.7737996578216553,
"learning_rate": 3.917156077734728e-06,
"loss": 0.1544,
"step": 675
},
{
"epoch": 0.6604182567546939,
"grad_norm": 0.6811118721961975,
"learning_rate": 3.913896396644739e-06,
"loss": 0.1582,
"step": 676
},
{
"epoch": 0.6613952068386506,
"grad_norm": 0.8076813220977783,
"learning_rate": 3.910633177547209e-06,
"loss": 0.1657,
"step": 677
},
{
"epoch": 0.6623721569226072,
"grad_norm": 0.7231690287590027,
"learning_rate": 3.9073664286077265e-06,
"loss": 0.1359,
"step": 678
},
{
"epoch": 0.6633491070065639,
"grad_norm": 0.8039504885673523,
"learning_rate": 3.904096158000708e-06,
"loss": 0.1464,
"step": 679
},
{
"epoch": 0.6643260570905205,
"grad_norm": 1.062908411026001,
"learning_rate": 3.900822373909389e-06,
"loss": 0.1556,
"step": 680
},
{
"epoch": 0.6653030071744772,
"grad_norm": 0.8012369275093079,
"learning_rate": 3.89754508452579e-06,
"loss": 0.1505,
"step": 681
},
{
"epoch": 0.6662799572584338,
"grad_norm": 1.5198546648025513,
"learning_rate": 3.894264298050709e-06,
"loss": 0.1724,
"step": 682
},
{
"epoch": 0.6672569073423905,
"grad_norm": 1.3087764978408813,
"learning_rate": 3.89098002269369e-06,
"loss": 0.1655,
"step": 683
},
{
"epoch": 0.6682338574263471,
"grad_norm": 1.0281848907470703,
"learning_rate": 3.887692266673009e-06,
"loss": 0.1586,
"step": 684
},
{
"epoch": 0.6692108075103038,
"grad_norm": 0.8611457347869873,
"learning_rate": 3.884401038215652e-06,
"loss": 0.1476,
"step": 685
},
{
"epoch": 0.6701877575942604,
"grad_norm": 0.7394402623176575,
"learning_rate": 3.881106345557295e-06,
"loss": 0.1596,
"step": 686
},
{
"epoch": 0.6711647076782171,
"grad_norm": 1.836513876914978,
"learning_rate": 3.877808196942281e-06,
"loss": 0.1513,
"step": 687
},
{
"epoch": 0.6721416577621737,
"grad_norm": 0.6691869497299194,
"learning_rate": 3.8745066006236e-06,
"loss": 0.1443,
"step": 688
},
{
"epoch": 0.6731186078461303,
"grad_norm": 0.7106708884239197,
"learning_rate": 3.871201564862871e-06,
"loss": 0.139,
"step": 689
},
{
"epoch": 0.674095557930087,
"grad_norm": 0.6661007404327393,
"learning_rate": 3.8678930979303205e-06,
"loss": 0.1503,
"step": 690
},
{
"epoch": 0.6750725080140436,
"grad_norm": 0.7850021123886108,
"learning_rate": 3.864581208104759e-06,
"loss": 0.1408,
"step": 691
},
{
"epoch": 0.6760494580980003,
"grad_norm": 0.7836375832557678,
"learning_rate": 3.861265903673561e-06,
"loss": 0.1574,
"step": 692
},
{
"epoch": 0.6770264081819569,
"grad_norm": 0.670484721660614,
"learning_rate": 3.857947192932649e-06,
"loss": 0.1467,
"step": 693
},
{
"epoch": 0.6780033582659136,
"grad_norm": 0.8295063972473145,
"learning_rate": 3.854625084186466e-06,
"loss": 0.1784,
"step": 694
},
{
"epoch": 0.6789803083498702,
"grad_norm": 1.3509005308151245,
"learning_rate": 3.851299585747959e-06,
"loss": 0.1741,
"step": 695
},
{
"epoch": 0.6799572584338269,
"grad_norm": 1.1903209686279297,
"learning_rate": 3.847970705938558e-06,
"loss": 0.1452,
"step": 696
},
{
"epoch": 0.6809342085177835,
"grad_norm": 0.7440477609634399,
"learning_rate": 3.844638453088151e-06,
"loss": 0.1541,
"step": 697
},
{
"epoch": 0.6819111586017402,
"grad_norm": 0.9636890888214111,
"learning_rate": 3.841302835535071e-06,
"loss": 0.1541,
"step": 698
},
{
"epoch": 0.6828881086856968,
"grad_norm": 1.3575022220611572,
"learning_rate": 3.837963861626067e-06,
"loss": 0.1431,
"step": 699
},
{
"epoch": 0.6838650587696535,
"grad_norm": 0.6622233390808105,
"learning_rate": 3.834621539716289e-06,
"loss": 0.1672,
"step": 700
},
{
"epoch": 0.6848420088536101,
"grad_norm": 1.5222954750061035,
"learning_rate": 3.831275878169263e-06,
"loss": 0.1785,
"step": 701
},
{
"epoch": 0.6858189589375668,
"grad_norm": 0.8201009035110474,
"learning_rate": 3.827926885356872e-06,
"loss": 0.1504,
"step": 702
},
{
"epoch": 0.6867959090215234,
"grad_norm": 0.7373434901237488,
"learning_rate": 3.824574569659335e-06,
"loss": 0.1503,
"step": 703
},
{
"epoch": 0.68777285910548,
"grad_norm": 1.0645610094070435,
"learning_rate": 3.821218939465188e-06,
"loss": 0.1502,
"step": 704
},
{
"epoch": 0.6887498091894367,
"grad_norm": 0.6556426882743835,
"learning_rate": 3.817860003171257e-06,
"loss": 0.1314,
"step": 705
},
{
"epoch": 0.6897267592733933,
"grad_norm": 1.0406161546707153,
"learning_rate": 3.8144977691826435e-06,
"loss": 0.1615,
"step": 706
},
{
"epoch": 0.69070370935735,
"grad_norm": 0.8727053999900818,
"learning_rate": 3.8111322459127002e-06,
"loss": 0.1493,
"step": 707
},
{
"epoch": 0.6916806594413066,
"grad_norm": 0.796471118927002,
"learning_rate": 3.8077634417830117e-06,
"loss": 0.1419,
"step": 708
},
{
"epoch": 0.6926576095252633,
"grad_norm": 0.6850622892379761,
"learning_rate": 3.8043913652233705e-06,
"loss": 0.1545,
"step": 709
},
{
"epoch": 0.6936345596092199,
"grad_norm": 0.8406426906585693,
"learning_rate": 3.8010160246717588e-06,
"loss": 0.1384,
"step": 710
},
{
"epoch": 0.6946115096931766,
"grad_norm": 0.9589312076568604,
"learning_rate": 3.7976374285743257e-06,
"loss": 0.1614,
"step": 711
},
{
"epoch": 0.6955884597771332,
"grad_norm": 0.6751416921615601,
"learning_rate": 3.794255585385368e-06,
"loss": 0.1347,
"step": 712
},
{
"epoch": 0.69656540986109,
"grad_norm": 0.7299074530601501,
"learning_rate": 3.7908705035673056e-06,
"loss": 0.1653,
"step": 713
},
{
"epoch": 0.6975423599450465,
"grad_norm": 0.8144887089729309,
"learning_rate": 3.7874821915906645e-06,
"loss": 0.1652,
"step": 714
},
{
"epoch": 0.6985193100290032,
"grad_norm": 0.7209720015525818,
"learning_rate": 3.784090657934052e-06,
"loss": 0.1572,
"step": 715
},
{
"epoch": 0.6994962601129598,
"grad_norm": 0.7366744875907898,
"learning_rate": 3.7806959110841385e-06,
"loss": 0.1652,
"step": 716
},
{
"epoch": 0.7004732101969166,
"grad_norm": 0.9909793138504028,
"learning_rate": 3.777297959535634e-06,
"loss": 0.1625,
"step": 717
},
{
"epoch": 0.7014501602808731,
"grad_norm": 0.911719560623169,
"learning_rate": 3.7738968117912672e-06,
"loss": 0.1511,
"step": 718
},
{
"epoch": 0.7024271103648297,
"grad_norm": 1.118007779121399,
"learning_rate": 3.7704924763617656e-06,
"loss": 0.15,
"step": 719
},
{
"epoch": 0.7034040604487865,
"grad_norm": 0.8047483563423157,
"learning_rate": 3.7670849617658324e-06,
"loss": 0.1568,
"step": 720
},
{
"epoch": 0.704381010532743,
"grad_norm": 0.7117910385131836,
"learning_rate": 3.763674276530127e-06,
"loss": 0.1542,
"step": 721
},
{
"epoch": 0.7053579606166998,
"grad_norm": 0.8130705952644348,
"learning_rate": 3.760260429189243e-06,
"loss": 0.149,
"step": 722
},
{
"epoch": 0.7063349107006563,
"grad_norm": 1.1423535346984863,
"learning_rate": 3.7568434282856836e-06,
"loss": 0.1456,
"step": 723
},
{
"epoch": 0.7073118607846131,
"grad_norm": 0.8230849504470825,
"learning_rate": 3.753423282369847e-06,
"loss": 0.1501,
"step": 724
},
{
"epoch": 0.7082888108685697,
"grad_norm": 0.7504922747612,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.1678,
"step": 725
},
{
"epoch": 0.7092657609525264,
"grad_norm": 0.6872618198394775,
"learning_rate": 3.746573589742257e-06,
"loss": 0.1444,
"step": 726
},
{
"epoch": 0.710242711036483,
"grad_norm": 0.7755722999572754,
"learning_rate": 3.743144060170559e-06,
"loss": 0.158,
"step": 727
},
{
"epoch": 0.7112196611204397,
"grad_norm": 0.722652018070221,
"learning_rate": 3.7397114198666536e-06,
"loss": 0.131,
"step": 728
},
{
"epoch": 0.7121966112043963,
"grad_norm": 0.7475513219833374,
"learning_rate": 3.7362756774200727e-06,
"loss": 0.1577,
"step": 729
},
{
"epoch": 0.713173561288353,
"grad_norm": 0.8016400933265686,
"learning_rate": 3.7328368414281093e-06,
"loss": 0.1701,
"step": 730
},
{
"epoch": 0.7141505113723096,
"grad_norm": 0.9551611542701721,
"learning_rate": 3.729394920495798e-06,
"loss": 0.1595,
"step": 731
},
{
"epoch": 0.7151274614562663,
"grad_norm": 0.8681173324584961,
"learning_rate": 3.7259499232358937e-06,
"loss": 0.1566,
"step": 732
},
{
"epoch": 0.7161044115402229,
"grad_norm": 1.5914216041564941,
"learning_rate": 3.722501858268849e-06,
"loss": 0.1813,
"step": 733
},
{
"epoch": 0.7170813616241796,
"grad_norm": 0.655326783657074,
"learning_rate": 3.719050734222791e-06,
"loss": 0.1437,
"step": 734
},
{
"epoch": 0.7180583117081362,
"grad_norm": 0.785659909248352,
"learning_rate": 3.7155965597335042e-06,
"loss": 0.1398,
"step": 735
},
{
"epoch": 0.7190352617920928,
"grad_norm": 0.8709280490875244,
"learning_rate": 3.712139343444405e-06,
"loss": 0.1524,
"step": 736
},
{
"epoch": 0.7200122118760495,
"grad_norm": 0.592876672744751,
"learning_rate": 3.708679094006521e-06,
"loss": 0.1598,
"step": 737
},
{
"epoch": 0.7209891619600061,
"grad_norm": 0.8991689682006836,
"learning_rate": 3.7052158200784686e-06,
"loss": 0.1494,
"step": 738
},
{
"epoch": 0.7219661120439628,
"grad_norm": 0.8952021598815918,
"learning_rate": 3.701749530326437e-06,
"loss": 0.1559,
"step": 739
},
{
"epoch": 0.7229430621279194,
"grad_norm": 1.0004162788391113,
"learning_rate": 3.6982802334241573e-06,
"loss": 0.1581,
"step": 740
},
{
"epoch": 0.7239200122118761,
"grad_norm": 1.0609227418899536,
"learning_rate": 3.6948079380528866e-06,
"loss": 0.1599,
"step": 741
},
{
"epoch": 0.7248969622958327,
"grad_norm": 0.8507347702980042,
"learning_rate": 3.691332652901385e-06,
"loss": 0.1635,
"step": 742
},
{
"epoch": 0.7258739123797894,
"grad_norm": 1.0106269121170044,
"learning_rate": 3.6878543866658956e-06,
"loss": 0.1539,
"step": 743
},
{
"epoch": 0.726850862463746,
"grad_norm": 0.7608538269996643,
"learning_rate": 3.68437314805012e-06,
"loss": 0.1674,
"step": 744
},
{
"epoch": 0.7278278125477027,
"grad_norm": 0.6567974090576172,
"learning_rate": 3.680888945765196e-06,
"loss": 0.1404,
"step": 745
},
{
"epoch": 0.7288047626316593,
"grad_norm": 1.413355827331543,
"learning_rate": 3.6774017885296796e-06,
"loss": 0.1536,
"step": 746
},
{
"epoch": 0.729781712715616,
"grad_norm": 0.8005159497261047,
"learning_rate": 3.673911685069521e-06,
"loss": 0.1592,
"step": 747
},
{
"epoch": 0.7307586627995726,
"grad_norm": 0.971828043460846,
"learning_rate": 3.6704186441180424e-06,
"loss": 0.1643,
"step": 748
},
{
"epoch": 0.7317356128835293,
"grad_norm": 0.9104104042053223,
"learning_rate": 3.6669226744159142e-06,
"loss": 0.1496,
"step": 749
},
{
"epoch": 0.7327125629674859,
"grad_norm": 1.1971485614776611,
"learning_rate": 3.663423784711139e-06,
"loss": 0.1522,
"step": 750
},
{
"epoch": 0.7336895130514425,
"grad_norm": 0.7140184640884399,
"learning_rate": 3.6599219837590242e-06,
"loss": 0.1459,
"step": 751
},
{
"epoch": 0.7346664631353992,
"grad_norm": 0.7513206005096436,
"learning_rate": 3.6564172803221633e-06,
"loss": 0.1629,
"step": 752
},
{
"epoch": 0.7356434132193558,
"grad_norm": 0.7880206108093262,
"learning_rate": 3.65290968317041e-06,
"loss": 0.1574,
"step": 753
},
{
"epoch": 0.7366203633033125,
"grad_norm": 0.8470908403396606,
"learning_rate": 3.6493992010808625e-06,
"loss": 0.1708,
"step": 754
},
{
"epoch": 0.7375973133872691,
"grad_norm": 1.215649962425232,
"learning_rate": 3.6458858428378365e-06,
"loss": 0.1553,
"step": 755
},
{
"epoch": 0.7385742634712258,
"grad_norm": 1.10202956199646,
"learning_rate": 3.642369617232843e-06,
"loss": 0.1915,
"step": 756
},
{
"epoch": 0.7395512135551824,
"grad_norm": 0.7796477675437927,
"learning_rate": 3.6388505330645706e-06,
"loss": 0.1504,
"step": 757
},
{
"epoch": 0.7405281636391391,
"grad_norm": 0.9163200259208679,
"learning_rate": 3.63532859913886e-06,
"loss": 0.1616,
"step": 758
},
{
"epoch": 0.7415051137230957,
"grad_norm": 0.669607400894165,
"learning_rate": 3.6318038242686822e-06,
"loss": 0.1398,
"step": 759
},
{
"epoch": 0.7424820638070524,
"grad_norm": 1.0828386545181274,
"learning_rate": 3.628276217274118e-06,
"loss": 0.1526,
"step": 760
},
{
"epoch": 0.743459013891009,
"grad_norm": 0.6801566481590271,
"learning_rate": 3.6247457869823355e-06,
"loss": 0.1459,
"step": 761
},
{
"epoch": 0.7444359639749657,
"grad_norm": 0.8077031373977661,
"learning_rate": 3.621212542227566e-06,
"loss": 0.1477,
"step": 762
},
{
"epoch": 0.7454129140589223,
"grad_norm": 0.7882485389709473,
"learning_rate": 3.617676491851085e-06,
"loss": 0.1626,
"step": 763
},
{
"epoch": 0.746389864142879,
"grad_norm": 1.1144007444381714,
"learning_rate": 3.6141376447011866e-06,
"loss": 0.1662,
"step": 764
},
{
"epoch": 0.7473668142268356,
"grad_norm": 0.8559699654579163,
"learning_rate": 3.610596009633166e-06,
"loss": 0.1577,
"step": 765
},
{
"epoch": 0.7483437643107922,
"grad_norm": 1.1554315090179443,
"learning_rate": 3.607051595509292e-06,
"loss": 0.1941,
"step": 766
},
{
"epoch": 0.7493207143947489,
"grad_norm": 0.7447336316108704,
"learning_rate": 3.6035044111987892e-06,
"loss": 0.1695,
"step": 767
},
{
"epoch": 0.7502976644787055,
"grad_norm": 0.9970964789390564,
"learning_rate": 3.599954465577814e-06,
"loss": 0.169,
"step": 768
},
{
"epoch": 0.7512746145626622,
"grad_norm": 0.8850570917129517,
"learning_rate": 3.5964017675294305e-06,
"loss": 0.1717,
"step": 769
},
{
"epoch": 0.7522515646466188,
"grad_norm": 0.8938663601875305,
"learning_rate": 3.592846325943593e-06,
"loss": 0.155,
"step": 770
},
{
"epoch": 0.7532285147305755,
"grad_norm": 0.7211086750030518,
"learning_rate": 3.5892881497171195e-06,
"loss": 0.1452,
"step": 771
},
{
"epoch": 0.7542054648145321,
"grad_norm": 0.9276571869850159,
"learning_rate": 3.5857272477536705e-06,
"loss": 0.1449,
"step": 772
},
{
"epoch": 0.7551824148984888,
"grad_norm": 1.0998425483703613,
"learning_rate": 3.582163628963728e-06,
"loss": 0.1509,
"step": 773
},
{
"epoch": 0.7561593649824454,
"grad_norm": 0.6526289582252502,
"learning_rate": 3.5785973022645717e-06,
"loss": 0.1517,
"step": 774
},
{
"epoch": 0.7571363150664021,
"grad_norm": 1.0672348737716675,
"learning_rate": 3.5750282765802584e-06,
"loss": 0.1676,
"step": 775
},
{
"epoch": 0.7581132651503587,
"grad_norm": 0.7462525367736816,
"learning_rate": 3.5714565608415975e-06,
"loss": 0.147,
"step": 776
},
{
"epoch": 0.7590902152343154,
"grad_norm": 0.8577287197113037,
"learning_rate": 3.5678821639861305e-06,
"loss": 0.1463,
"step": 777
},
{
"epoch": 0.760067165318272,
"grad_norm": 1.007952332496643,
"learning_rate": 3.5643050949581065e-06,
"loss": 0.1506,
"step": 778
},
{
"epoch": 0.7610441154022287,
"grad_norm": 0.7961283326148987,
"learning_rate": 3.560725362708464e-06,
"loss": 0.1594,
"step": 779
},
{
"epoch": 0.7620210654861853,
"grad_norm": 0.987490713596344,
"learning_rate": 3.557142976194802e-06,
"loss": 0.1587,
"step": 780
},
{
"epoch": 0.7629980155701419,
"grad_norm": 1.0939033031463623,
"learning_rate": 3.5535579443813652e-06,
"loss": 0.1393,
"step": 781
},
{
"epoch": 0.7639749656540986,
"grad_norm": 0.8193220496177673,
"learning_rate": 3.5499702762390133e-06,
"loss": 0.1701,
"step": 782
},
{
"epoch": 0.7649519157380552,
"grad_norm": 0.7307434678077698,
"learning_rate": 3.5463799807452076e-06,
"loss": 0.1519,
"step": 783
},
{
"epoch": 0.7659288658220119,
"grad_norm": 0.6775676012039185,
"learning_rate": 3.54278706688398e-06,
"loss": 0.1561,
"step": 784
},
{
"epoch": 0.7669058159059685,
"grad_norm": 0.7602413892745972,
"learning_rate": 3.5391915436459167e-06,
"loss": 0.1728,
"step": 785
},
{
"epoch": 0.7678827659899252,
"grad_norm": 1.4706379175186157,
"learning_rate": 3.535593420028132e-06,
"loss": 0.1707,
"step": 786
},
{
"epoch": 0.7688597160738818,
"grad_norm": 0.7432350516319275,
"learning_rate": 3.5319927050342497e-06,
"loss": 0.1507,
"step": 787
},
{
"epoch": 0.7698366661578385,
"grad_norm": 1.189603328704834,
"learning_rate": 3.528389407674373e-06,
"loss": 0.1798,
"step": 788
},
{
"epoch": 0.7708136162417951,
"grad_norm": 0.7542485594749451,
"learning_rate": 3.5247835369650718e-06,
"loss": 0.1598,
"step": 789
},
{
"epoch": 0.7717905663257518,
"grad_norm": 0.8822116851806641,
"learning_rate": 3.521175101929353e-06,
"loss": 0.1566,
"step": 790
},
{
"epoch": 0.7727675164097084,
"grad_norm": 0.7665351033210754,
"learning_rate": 3.517564111596643e-06,
"loss": 0.1494,
"step": 791
},
{
"epoch": 0.7737444664936651,
"grad_norm": 0.7099792957305908,
"learning_rate": 3.5139505750027568e-06,
"loss": 0.1528,
"step": 792
},
{
"epoch": 0.7747214165776217,
"grad_norm": 0.6690868735313416,
"learning_rate": 3.510334501189886e-06,
"loss": 0.1343,
"step": 793
},
{
"epoch": 0.7756983666615784,
"grad_norm": 0.8017610311508179,
"learning_rate": 3.5067158992065697e-06,
"loss": 0.1501,
"step": 794
},
{
"epoch": 0.776675316745535,
"grad_norm": 1.0609956979751587,
"learning_rate": 3.503094778107673e-06,
"loss": 0.1651,
"step": 795
},
{
"epoch": 0.7776522668294917,
"grad_norm": 0.6591711640357971,
"learning_rate": 3.4994711469543625e-06,
"loss": 0.1515,
"step": 796
},
{
"epoch": 0.7786292169134483,
"grad_norm": 0.8857126832008362,
"learning_rate": 3.4958450148140906e-06,
"loss": 0.1657,
"step": 797
},
{
"epoch": 0.7796061669974049,
"grad_norm": 0.951080858707428,
"learning_rate": 3.4922163907605637e-06,
"loss": 0.1569,
"step": 798
},
{
"epoch": 0.7805831170813616,
"grad_norm": 0.6247563362121582,
"learning_rate": 3.488585283873726e-06,
"loss": 0.1447,
"step": 799
},
{
"epoch": 0.7815600671653182,
"grad_norm": 0.82416832447052,
"learning_rate": 3.4849517032397327e-06,
"loss": 0.1563,
"step": 800
},
{
"epoch": 0.7825370172492749,
"grad_norm": 0.6015522480010986,
"learning_rate": 3.4813156579509304e-06,
"loss": 0.1411,
"step": 801
},
{
"epoch": 0.7835139673332315,
"grad_norm": 0.9072467088699341,
"learning_rate": 3.4776771571058344e-06,
"loss": 0.1789,
"step": 802
},
{
"epoch": 0.7844909174171882,
"grad_norm": 2.0301270484924316,
"learning_rate": 3.4740362098091003e-06,
"loss": 0.1566,
"step": 803
},
{
"epoch": 0.7854678675011448,
"grad_norm": 1.1363762617111206,
"learning_rate": 3.4703928251715093e-06,
"loss": 0.1574,
"step": 804
},
{
"epoch": 0.7864448175851015,
"grad_norm": 0.7775416374206543,
"learning_rate": 3.46674701230994e-06,
"loss": 0.1531,
"step": 805
},
{
"epoch": 0.7874217676690581,
"grad_norm": 1.1870876550674438,
"learning_rate": 3.4630987803473493e-06,
"loss": 0.1404,
"step": 806
},
{
"epoch": 0.7883987177530148,
"grad_norm": 0.8673657178878784,
"learning_rate": 3.4594481384127424e-06,
"loss": 0.1582,
"step": 807
},
{
"epoch": 0.7893756678369714,
"grad_norm": 0.7943292856216431,
"learning_rate": 3.4557950956411613e-06,
"loss": 0.1517,
"step": 808
},
{
"epoch": 0.7903526179209281,
"grad_norm": 0.8395007252693176,
"learning_rate": 3.4521396611736512e-06,
"loss": 0.1546,
"step": 809
},
{
"epoch": 0.7913295680048847,
"grad_norm": 0.8592638373374939,
"learning_rate": 3.4484818441572444e-06,
"loss": 0.1619,
"step": 810
},
{
"epoch": 0.7923065180888414,
"grad_norm": 1.238155722618103,
"learning_rate": 3.4448216537449326e-06,
"loss": 0.1389,
"step": 811
},
{
"epoch": 0.793283468172798,
"grad_norm": 1.1079182624816895,
"learning_rate": 3.441159099095649e-06,
"loss": 0.1481,
"step": 812
},
{
"epoch": 0.7942604182567546,
"grad_norm": 0.7676824927330017,
"learning_rate": 3.437494189374242e-06,
"loss": 0.1446,
"step": 813
},
{
"epoch": 0.7952373683407113,
"grad_norm": 0.7961470484733582,
"learning_rate": 3.4338269337514527e-06,
"loss": 0.1572,
"step": 814
},
{
"epoch": 0.7962143184246679,
"grad_norm": 0.6730189323425293,
"learning_rate": 3.430157341403893e-06,
"loss": 0.1354,
"step": 815
},
{
"epoch": 0.7971912685086247,
"grad_norm": 0.7726109027862549,
"learning_rate": 3.4264854215140216e-06,
"loss": 0.1619,
"step": 816
},
{
"epoch": 0.7981682185925812,
"grad_norm": 0.7890665531158447,
"learning_rate": 3.422811183270121e-06,
"loss": 0.1664,
"step": 817
},
{
"epoch": 0.799145168676538,
"grad_norm": 0.8949539065361023,
"learning_rate": 3.4191346358662762e-06,
"loss": 0.1512,
"step": 818
},
{
"epoch": 0.8001221187604945,
"grad_norm": 1.0934100151062012,
"learning_rate": 3.4154557885023493e-06,
"loss": 0.1495,
"step": 819
},
{
"epoch": 0.8010990688444513,
"grad_norm": 0.682202160358429,
"learning_rate": 3.4117746503839577e-06,
"loss": 0.1376,
"step": 820
},
{
"epoch": 0.8020760189284079,
"grad_norm": 0.9899504780769348,
"learning_rate": 3.4080912307224524e-06,
"loss": 0.1507,
"step": 821
},
{
"epoch": 0.8030529690123646,
"grad_norm": 0.8517482876777649,
"learning_rate": 3.4044055387348912e-06,
"loss": 0.1442,
"step": 822
},
{
"epoch": 0.8040299190963212,
"grad_norm": 1.0311285257339478,
"learning_rate": 3.4007175836440197e-06,
"loss": 0.1635,
"step": 823
},
{
"epoch": 0.8050068691802779,
"grad_norm": 0.8413298726081848,
"learning_rate": 3.397027374678246e-06,
"loss": 0.1623,
"step": 824
},
{
"epoch": 0.8059838192642345,
"grad_norm": 0.7206878066062927,
"learning_rate": 3.393334921071618e-06,
"loss": 0.1524,
"step": 825
},
{
"epoch": 0.8069607693481912,
"grad_norm": 0.741763174533844,
"learning_rate": 3.3896402320638006e-06,
"loss": 0.1517,
"step": 826
},
{
"epoch": 0.8079377194321478,
"grad_norm": 1.0026910305023193,
"learning_rate": 3.385943316900052e-06,
"loss": 0.1932,
"step": 827
},
{
"epoch": 0.8089146695161044,
"grad_norm": 1.0051147937774658,
"learning_rate": 3.3822441848312017e-06,
"loss": 0.1497,
"step": 828
},
{
"epoch": 0.8098916196000611,
"grad_norm": 0.7903057932853699,
"learning_rate": 3.378542845113625e-06,
"loss": 0.1831,
"step": 829
},
{
"epoch": 0.8108685696840177,
"grad_norm": 0.632512629032135,
"learning_rate": 3.374839307009223e-06,
"loss": 0.1459,
"step": 830
},
{
"epoch": 0.8118455197679744,
"grad_norm": 0.7736039161682129,
"learning_rate": 3.371133579785398e-06,
"loss": 0.152,
"step": 831
},
{
"epoch": 0.812822469851931,
"grad_norm": 1.2243468761444092,
"learning_rate": 3.3674256727150285e-06,
"loss": 0.1673,
"step": 832
},
{
"epoch": 0.8137994199358877,
"grad_norm": 1.4464961290359497,
"learning_rate": 3.363715595076449e-06,
"loss": 0.1715,
"step": 833
},
{
"epoch": 0.8147763700198443,
"grad_norm": 0.6386120319366455,
"learning_rate": 3.3600033561534247e-06,
"loss": 0.1657,
"step": 834
},
{
"epoch": 0.815753320103801,
"grad_norm": 0.6698469519615173,
"learning_rate": 3.356288965235129e-06,
"loss": 0.1456,
"step": 835
},
{
"epoch": 0.8167302701877576,
"grad_norm": 1.1008213758468628,
"learning_rate": 3.352572431616123e-06,
"loss": 0.1656,
"step": 836
},
{
"epoch": 0.8177072202717143,
"grad_norm": 1.0147091150283813,
"learning_rate": 3.348853764596324e-06,
"loss": 0.1502,
"step": 837
},
{
"epoch": 0.8186841703556709,
"grad_norm": 0.6490986943244934,
"learning_rate": 3.3451329734809928e-06,
"loss": 0.1329,
"step": 838
},
{
"epoch": 0.8196611204396276,
"grad_norm": 0.8020153045654297,
"learning_rate": 3.3414100675807026e-06,
"loss": 0.1669,
"step": 839
},
{
"epoch": 0.8206380705235842,
"grad_norm": 1.6853318214416504,
"learning_rate": 3.3376850562113193e-06,
"loss": 0.1685,
"step": 840
},
{
"epoch": 0.8216150206075409,
"grad_norm": 1.2695226669311523,
"learning_rate": 3.3339579486939784e-06,
"loss": 0.1487,
"step": 841
},
{
"epoch": 0.8225919706914975,
"grad_norm": 0.6849395632743835,
"learning_rate": 3.3302287543550582e-06,
"loss": 0.1538,
"step": 842
},
{
"epoch": 0.8235689207754542,
"grad_norm": 0.7239601016044617,
"learning_rate": 3.3264974825261597e-06,
"loss": 0.1531,
"step": 843
},
{
"epoch": 0.8245458708594108,
"grad_norm": 0.693019449710846,
"learning_rate": 3.3227641425440843e-06,
"loss": 0.1548,
"step": 844
},
{
"epoch": 0.8255228209433674,
"grad_norm": 1.082384705543518,
"learning_rate": 3.3190287437508067e-06,
"loss": 0.1509,
"step": 845
},
{
"epoch": 0.8264997710273241,
"grad_norm": 0.705163836479187,
"learning_rate": 3.3152912954934536e-06,
"loss": 0.1513,
"step": 846
},
{
"epoch": 0.8274767211112807,
"grad_norm": 0.8317340016365051,
"learning_rate": 3.3115518071242805e-06,
"loss": 0.1406,
"step": 847
},
{
"epoch": 0.8284536711952374,
"grad_norm": 0.738751232624054,
"learning_rate": 3.3078102880006467e-06,
"loss": 0.15,
"step": 848
},
{
"epoch": 0.829430621279194,
"grad_norm": 0.8009028434753418,
"learning_rate": 3.3040667474849967e-06,
"loss": 0.1446,
"step": 849
},
{
"epoch": 0.8304075713631507,
"grad_norm": 0.737471342086792,
"learning_rate": 3.3003211949448284e-06,
"loss": 0.1434,
"step": 850
},
{
"epoch": 0.8313845214471073,
"grad_norm": 0.8901124596595764,
"learning_rate": 3.296573639752676e-06,
"loss": 0.1376,
"step": 851
},
{
"epoch": 0.832361471531064,
"grad_norm": 0.837114155292511,
"learning_rate": 3.292824091286088e-06,
"loss": 0.1564,
"step": 852
},
{
"epoch": 0.8333384216150206,
"grad_norm": 0.8355427980422974,
"learning_rate": 3.289072558927596e-06,
"loss": 0.1539,
"step": 853
},
{
"epoch": 0.8343153716989773,
"grad_norm": 0.7224215269088745,
"learning_rate": 3.2853190520646995e-06,
"loss": 0.1681,
"step": 854
},
{
"epoch": 0.8352923217829339,
"grad_norm": 0.7744103670120239,
"learning_rate": 3.281563580089836e-06,
"loss": 0.1453,
"step": 855
},
{
"epoch": 0.8362692718668906,
"grad_norm": 0.8624812364578247,
"learning_rate": 3.2778061524003633e-06,
"loss": 0.1594,
"step": 856
},
{
"epoch": 0.8372462219508472,
"grad_norm": 1.0955893993377686,
"learning_rate": 3.2740467783985315e-06,
"loss": 0.1624,
"step": 857
},
{
"epoch": 0.8382231720348039,
"grad_norm": 0.633622407913208,
"learning_rate": 3.2702854674914584e-06,
"loss": 0.1449,
"step": 858
},
{
"epoch": 0.8392001221187605,
"grad_norm": 1.0901026725769043,
"learning_rate": 3.2665222290911145e-06,
"loss": 0.1664,
"step": 859
},
{
"epoch": 0.8401770722027171,
"grad_norm": 1.002231240272522,
"learning_rate": 3.2627570726142882e-06,
"loss": 0.1514,
"step": 860
},
{
"epoch": 0.8411540222866738,
"grad_norm": 1.0149930715560913,
"learning_rate": 3.25899000748257e-06,
"loss": 0.1609,
"step": 861
},
{
"epoch": 0.8421309723706304,
"grad_norm": 0.8724030256271362,
"learning_rate": 3.2552210431223247e-06,
"loss": 0.1499,
"step": 862
},
{
"epoch": 0.8431079224545871,
"grad_norm": 0.7488546967506409,
"learning_rate": 3.251450188964672e-06,
"loss": 0.1474,
"step": 863
},
{
"epoch": 0.8440848725385437,
"grad_norm": 0.6939215660095215,
"learning_rate": 3.2476774544454597e-06,
"loss": 0.1488,
"step": 864
},
{
"epoch": 0.8450618226225004,
"grad_norm": 0.8398666381835938,
"learning_rate": 3.243902849005238e-06,
"loss": 0.148,
"step": 865
},
{
"epoch": 0.846038772706457,
"grad_norm": 0.9901300668716431,
"learning_rate": 3.2401263820892433e-06,
"loss": 0.1561,
"step": 866
},
{
"epoch": 0.8470157227904137,
"grad_norm": 0.6501366496086121,
"learning_rate": 3.2363480631473666e-06,
"loss": 0.1467,
"step": 867
},
{
"epoch": 0.8479926728743703,
"grad_norm": 0.7710810303688049,
"learning_rate": 3.2325679016341347e-06,
"loss": 0.15,
"step": 868
},
{
"epoch": 0.848969622958327,
"grad_norm": 0.7182073593139648,
"learning_rate": 3.228785907008684e-06,
"loss": 0.1408,
"step": 869
},
{
"epoch": 0.8499465730422836,
"grad_norm": 0.7221583724021912,
"learning_rate": 3.22500208873474e-06,
"loss": 0.1461,
"step": 870
},
{
"epoch": 0.8509235231262403,
"grad_norm": 0.7775384187698364,
"learning_rate": 3.221216456280589e-06,
"loss": 0.1494,
"step": 871
},
{
"epoch": 0.8519004732101969,
"grad_norm": 0.7050665020942688,
"learning_rate": 3.217429019119059e-06,
"loss": 0.1451,
"step": 872
},
{
"epoch": 0.8528774232941536,
"grad_norm": 1.0802805423736572,
"learning_rate": 3.2136397867274926e-06,
"loss": 0.1497,
"step": 873
},
{
"epoch": 0.8538543733781102,
"grad_norm": 0.9995020031929016,
"learning_rate": 3.2098487685877257e-06,
"loss": 0.1782,
"step": 874
},
{
"epoch": 0.8548313234620668,
"grad_norm": 0.7245685458183289,
"learning_rate": 3.206055974186061e-06,
"loss": 0.1618,
"step": 875
},
{
"epoch": 0.8558082735460235,
"grad_norm": 0.8504968881607056,
"learning_rate": 3.202261413013248e-06,
"loss": 0.153,
"step": 876
},
{
"epoch": 0.8567852236299801,
"grad_norm": 0.8796176910400391,
"learning_rate": 3.198465094564456e-06,
"loss": 0.1571,
"step": 877
},
{
"epoch": 0.8577621737139368,
"grad_norm": 0.9412023425102234,
"learning_rate": 3.1946670283392523e-06,
"loss": 0.1535,
"step": 878
},
{
"epoch": 0.8587391237978934,
"grad_norm": 1.6855803728103638,
"learning_rate": 3.190867223841576e-06,
"loss": 0.1716,
"step": 879
},
{
"epoch": 0.8597160738818501,
"grad_norm": 0.8716698884963989,
"learning_rate": 3.187065690579718e-06,
"loss": 0.1553,
"step": 880
},
{
"epoch": 0.8606930239658067,
"grad_norm": 1.042761206626892,
"learning_rate": 3.1832624380662943e-06,
"loss": 0.1611,
"step": 881
},
{
"epoch": 0.8616699740497634,
"grad_norm": 1.0470339059829712,
"learning_rate": 3.179457475818222e-06,
"loss": 0.1481,
"step": 882
},
{
"epoch": 0.86264692413372,
"grad_norm": 0.7020443081855774,
"learning_rate": 3.175650813356699e-06,
"loss": 0.1407,
"step": 883
},
{
"epoch": 0.8636238742176767,
"grad_norm": 1.1522868871688843,
"learning_rate": 3.171842460207174e-06,
"loss": 0.1453,
"step": 884
},
{
"epoch": 0.8646008243016333,
"grad_norm": 0.8696357011795044,
"learning_rate": 3.1680324258993302e-06,
"loss": 0.1432,
"step": 885
},
{
"epoch": 0.86557777438559,
"grad_norm": 0.8831077218055725,
"learning_rate": 3.164220719967055e-06,
"loss": 0.1524,
"step": 886
},
{
"epoch": 0.8665547244695466,
"grad_norm": 0.919243335723877,
"learning_rate": 3.160407351948419e-06,
"loss": 0.16,
"step": 887
},
{
"epoch": 0.8675316745535033,
"grad_norm": 0.7314760684967041,
"learning_rate": 3.1565923313856526e-06,
"loss": 0.1692,
"step": 888
},
{
"epoch": 0.8685086246374599,
"grad_norm": 0.7344856858253479,
"learning_rate": 3.152775667825122e-06,
"loss": 0.1358,
"step": 889
},
{
"epoch": 0.8694855747214166,
"grad_norm": 0.6402750611305237,
"learning_rate": 3.1489573708173025e-06,
"loss": 0.1338,
"step": 890
},
{
"epoch": 0.8704625248053732,
"grad_norm": 1.0203721523284912,
"learning_rate": 3.1451374499167586e-06,
"loss": 0.1552,
"step": 891
},
{
"epoch": 0.8714394748893298,
"grad_norm": 0.930683970451355,
"learning_rate": 3.1413159146821193e-06,
"loss": 0.1506,
"step": 892
},
{
"epoch": 0.8724164249732865,
"grad_norm": 0.6897743344306946,
"learning_rate": 3.137492774676049e-06,
"loss": 0.1444,
"step": 893
},
{
"epoch": 0.8733933750572431,
"grad_norm": 0.8135661482810974,
"learning_rate": 3.133668039465232e-06,
"loss": 0.156,
"step": 894
},
{
"epoch": 0.8743703251411998,
"grad_norm": 0.7550363540649414,
"learning_rate": 3.1298417186203415e-06,
"loss": 0.1516,
"step": 895
},
{
"epoch": 0.8753472752251564,
"grad_norm": 0.6833528280258179,
"learning_rate": 3.126013821716022e-06,
"loss": 0.1538,
"step": 896
},
{
"epoch": 0.8763242253091131,
"grad_norm": 0.878643274307251,
"learning_rate": 3.122184358330857e-06,
"loss": 0.1561,
"step": 897
},
{
"epoch": 0.8773011753930697,
"grad_norm": 0.6735469698905945,
"learning_rate": 3.1183533380473525e-06,
"loss": 0.1704,
"step": 898
},
{
"epoch": 0.8782781254770264,
"grad_norm": 0.7926891446113586,
"learning_rate": 3.1145207704519113e-06,
"loss": 0.1366,
"step": 899
},
{
"epoch": 0.879255075560983,
"grad_norm": 0.6722373962402344,
"learning_rate": 3.1106866651348065e-06,
"loss": 0.1613,
"step": 900
},
{
"epoch": 0.8802320256449397,
"grad_norm": 0.8567788600921631,
"learning_rate": 3.1068510316901584e-06,
"loss": 0.1525,
"step": 901
},
{
"epoch": 0.8812089757288963,
"grad_norm": 0.781726598739624,
"learning_rate": 3.1030138797159125e-06,
"loss": 0.1657,
"step": 902
},
{
"epoch": 0.882185925812853,
"grad_norm": 1.2267659902572632,
"learning_rate": 3.0991752188138157e-06,
"loss": 0.1549,
"step": 903
},
{
"epoch": 0.8831628758968096,
"grad_norm": 0.7455759048461914,
"learning_rate": 3.095335058589386e-06,
"loss": 0.1719,
"step": 904
},
{
"epoch": 0.8841398259807663,
"grad_norm": 0.8069635629653931,
"learning_rate": 3.091493408651897e-06,
"loss": 0.1422,
"step": 905
},
{
"epoch": 0.8851167760647229,
"grad_norm": 0.6300619840621948,
"learning_rate": 3.087650278614349e-06,
"loss": 0.1478,
"step": 906
},
{
"epoch": 0.8860937261486795,
"grad_norm": 0.9580602645874023,
"learning_rate": 3.0838056780934473e-06,
"loss": 0.1541,
"step": 907
},
{
"epoch": 0.8870706762326362,
"grad_norm": 0.8519987463951111,
"learning_rate": 3.0799596167095737e-06,
"loss": 0.1484,
"step": 908
},
{
"epoch": 0.8880476263165928,
"grad_norm": 0.6413607597351074,
"learning_rate": 3.0761121040867666e-06,
"loss": 0.1465,
"step": 909
},
{
"epoch": 0.8890245764005495,
"grad_norm": 0.8902720212936401,
"learning_rate": 3.072263149852698e-06,
"loss": 0.1541,
"step": 910
},
{
"epoch": 0.8900015264845061,
"grad_norm": 0.9374000430107117,
"learning_rate": 3.0684127636386475e-06,
"loss": 0.1673,
"step": 911
},
{
"epoch": 0.8909784765684629,
"grad_norm": 0.8860354423522949,
"learning_rate": 3.0645609550794726e-06,
"loss": 0.1574,
"step": 912
},
{
"epoch": 0.8919554266524194,
"grad_norm": 1.2239422798156738,
"learning_rate": 3.060707733813595e-06,
"loss": 0.1438,
"step": 913
},
{
"epoch": 0.8929323767363762,
"grad_norm": 0.6602402329444885,
"learning_rate": 3.05685310948297e-06,
"loss": 0.1613,
"step": 914
},
{
"epoch": 0.8939093268203328,
"grad_norm": 0.7185958623886108,
"learning_rate": 3.0529970917330655e-06,
"loss": 0.1579,
"step": 915
},
{
"epoch": 0.8948862769042895,
"grad_norm": 0.6650615930557251,
"learning_rate": 3.0491396902128318e-06,
"loss": 0.1491,
"step": 916
},
{
"epoch": 0.895863226988246,
"grad_norm": 0.6480940580368042,
"learning_rate": 3.0452809145746863e-06,
"loss": 0.1427,
"step": 917
},
{
"epoch": 0.8968401770722028,
"grad_norm": 0.7698121070861816,
"learning_rate": 3.041420774474483e-06,
"loss": 0.1396,
"step": 918
},
{
"epoch": 0.8978171271561594,
"grad_norm": 0.7171248197555542,
"learning_rate": 3.03755927957149e-06,
"loss": 0.1475,
"step": 919
},
{
"epoch": 0.8987940772401161,
"grad_norm": 1.046448826789856,
"learning_rate": 3.0336964395283645e-06,
"loss": 0.1578,
"step": 920
},
{
"epoch": 0.8997710273240727,
"grad_norm": 0.858893632888794,
"learning_rate": 3.0298322640111333e-06,
"loss": 0.1438,
"step": 921
},
{
"epoch": 0.9007479774080293,
"grad_norm": 0.6682378053665161,
"learning_rate": 3.0259667626891615e-06,
"loss": 0.1457,
"step": 922
},
{
"epoch": 0.901724927491986,
"grad_norm": 0.7226861715316772,
"learning_rate": 3.0220999452351333e-06,
"loss": 0.1616,
"step": 923
},
{
"epoch": 0.9027018775759426,
"grad_norm": 0.7330286502838135,
"learning_rate": 3.0182318213250246e-06,
"loss": 0.1505,
"step": 924
},
{
"epoch": 0.9036788276598993,
"grad_norm": 0.9333467483520508,
"learning_rate": 3.0143624006380833e-06,
"loss": 0.1623,
"step": 925
},
{
"epoch": 0.9046557777438559,
"grad_norm": 0.9768693447113037,
"learning_rate": 3.0104916928568007e-06,
"loss": 0.17,
"step": 926
},
{
"epoch": 0.9056327278278126,
"grad_norm": 1.4236985445022583,
"learning_rate": 3.006619707666888e-06,
"loss": 0.1432,
"step": 927
},
{
"epoch": 0.9066096779117692,
"grad_norm": 0.6060042977333069,
"learning_rate": 3.0027464547572536e-06,
"loss": 0.1418,
"step": 928
},
{
"epoch": 0.9075866279957259,
"grad_norm": 0.8775410652160645,
"learning_rate": 2.9988719438199793e-06,
"loss": 0.161,
"step": 929
},
{
"epoch": 0.9085635780796825,
"grad_norm": 0.7384399771690369,
"learning_rate": 2.994996184550293e-06,
"loss": 0.1612,
"step": 930
},
{
"epoch": 0.9095405281636392,
"grad_norm": 0.7607274055480957,
"learning_rate": 2.9911191866465466e-06,
"loss": 0.1729,
"step": 931
},
{
"epoch": 0.9105174782475958,
"grad_norm": 0.8182145953178406,
"learning_rate": 2.987240959810193e-06,
"loss": 0.1194,
"step": 932
},
{
"epoch": 0.9114944283315525,
"grad_norm": 0.7334820628166199,
"learning_rate": 2.983361513745759e-06,
"loss": 0.131,
"step": 933
},
{
"epoch": 0.9124713784155091,
"grad_norm": 0.7076065540313721,
"learning_rate": 2.979480858160822e-06,
"loss": 0.1426,
"step": 934
},
{
"epoch": 0.9134483284994658,
"grad_norm": 0.698074460029602,
"learning_rate": 2.975599002765987e-06,
"loss": 0.1402,
"step": 935
},
{
"epoch": 0.9144252785834224,
"grad_norm": 0.6762651205062866,
"learning_rate": 2.9717159572748605e-06,
"loss": 0.1347,
"step": 936
},
{
"epoch": 0.9154022286673791,
"grad_norm": 0.7691007256507874,
"learning_rate": 2.9678317314040277e-06,
"loss": 0.1601,
"step": 937
},
{
"epoch": 0.9163791787513357,
"grad_norm": 0.8327917456626892,
"learning_rate": 2.963946334873027e-06,
"loss": 0.1611,
"step": 938
},
{
"epoch": 0.9173561288352923,
"grad_norm": 3.0022335052490234,
"learning_rate": 2.9600597774043256e-06,
"loss": 0.2316,
"step": 939
},
{
"epoch": 0.918333078919249,
"grad_norm": 0.7689866423606873,
"learning_rate": 2.9561720687232963e-06,
"loss": 0.1545,
"step": 940
},
{
"epoch": 0.9193100290032056,
"grad_norm": 0.6273118257522583,
"learning_rate": 2.952283218558194e-06,
"loss": 0.1352,
"step": 941
},
{
"epoch": 0.9202869790871623,
"grad_norm": 0.7392250299453735,
"learning_rate": 2.9483932366401274e-06,
"loss": 0.1567,
"step": 942
},
{
"epoch": 0.9212639291711189,
"grad_norm": 0.6563267111778259,
"learning_rate": 2.9445021327030387e-06,
"loss": 0.1734,
"step": 943
},
{
"epoch": 0.9222408792550756,
"grad_norm": 1.4566086530685425,
"learning_rate": 2.9406099164836772e-06,
"loss": 0.1276,
"step": 944
},
{
"epoch": 0.9232178293390322,
"grad_norm": 0.7569136023521423,
"learning_rate": 2.936716597721576e-06,
"loss": 0.1816,
"step": 945
},
{
"epoch": 0.9241947794229889,
"grad_norm": 0.8260944485664368,
"learning_rate": 2.9328221861590265e-06,
"loss": 0.1673,
"step": 946
},
{
"epoch": 0.9251717295069455,
"grad_norm": 0.709345817565918,
"learning_rate": 2.928926691541056e-06,
"loss": 0.1529,
"step": 947
},
{
"epoch": 0.9261486795909022,
"grad_norm": 0.9544458389282227,
"learning_rate": 2.9250301236153987e-06,
"loss": 0.1643,
"step": 948
},
{
"epoch": 0.9271256296748588,
"grad_norm": 0.8452097773551941,
"learning_rate": 2.921132492132479e-06,
"loss": 0.1636,
"step": 949
},
{
"epoch": 0.9281025797588155,
"grad_norm": 0.7840229272842407,
"learning_rate": 2.9172338068453793e-06,
"loss": 0.1595,
"step": 950
},
{
"epoch": 0.9290795298427721,
"grad_norm": 0.6846920251846313,
"learning_rate": 2.91333407750982e-06,
"loss": 0.1467,
"step": 951
},
{
"epoch": 0.9300564799267288,
"grad_norm": 0.6001102924346924,
"learning_rate": 2.9094333138841348e-06,
"loss": 0.15,
"step": 952
},
{
"epoch": 0.9310334300106854,
"grad_norm": 0.7123298048973083,
"learning_rate": 2.9055315257292427e-06,
"loss": 0.1444,
"step": 953
},
{
"epoch": 0.932010380094642,
"grad_norm": 1.014249324798584,
"learning_rate": 2.9016287228086315e-06,
"loss": 0.1671,
"step": 954
},
{
"epoch": 0.9329873301785987,
"grad_norm": 0.6392649412155151,
"learning_rate": 2.8977249148883226e-06,
"loss": 0.1484,
"step": 955
},
{
"epoch": 0.9339642802625553,
"grad_norm": 1.1468456983566284,
"learning_rate": 2.8938201117368563e-06,
"loss": 0.1588,
"step": 956
},
{
"epoch": 0.934941230346512,
"grad_norm": 0.7079674601554871,
"learning_rate": 2.889914323125261e-06,
"loss": 0.1404,
"step": 957
},
{
"epoch": 0.9359181804304686,
"grad_norm": 0.7039918899536133,
"learning_rate": 2.8860075588270325e-06,
"loss": 0.1509,
"step": 958
},
{
"epoch": 0.9368951305144253,
"grad_norm": 0.8289921879768372,
"learning_rate": 2.882099828618107e-06,
"loss": 0.1613,
"step": 959
},
{
"epoch": 0.9378720805983819,
"grad_norm": 0.84840327501297,
"learning_rate": 2.8781911422768376e-06,
"loss": 0.1489,
"step": 960
},
{
"epoch": 0.9388490306823386,
"grad_norm": 0.6307504773139954,
"learning_rate": 2.8742815095839728e-06,
"loss": 0.141,
"step": 961
},
{
"epoch": 0.9398259807662952,
"grad_norm": 0.9006770253181458,
"learning_rate": 2.870370940322624e-06,
"loss": 0.1718,
"step": 962
},
{
"epoch": 0.9408029308502519,
"grad_norm": 1.9910690784454346,
"learning_rate": 2.86645944427825e-06,
"loss": 0.161,
"step": 963
},
{
"epoch": 0.9417798809342085,
"grad_norm": 1.1320542097091675,
"learning_rate": 2.862547031238628e-06,
"loss": 0.1554,
"step": 964
},
{
"epoch": 0.9427568310181652,
"grad_norm": 0.7087171077728271,
"learning_rate": 2.8586337109938294e-06,
"loss": 0.1491,
"step": 965
},
{
"epoch": 0.9437337811021218,
"grad_norm": 0.6598576307296753,
"learning_rate": 2.8547194933361953e-06,
"loss": 0.1389,
"step": 966
},
{
"epoch": 0.9447107311860785,
"grad_norm": 0.7265792489051819,
"learning_rate": 2.8508043880603135e-06,
"loss": 0.1512,
"step": 967
},
{
"epoch": 0.9456876812700351,
"grad_norm": 0.8819522261619568,
"learning_rate": 2.8468884049629923e-06,
"loss": 0.1445,
"step": 968
},
{
"epoch": 0.9466646313539917,
"grad_norm": 0.5127539038658142,
"learning_rate": 2.842971553843238e-06,
"loss": 0.1479,
"step": 969
},
{
"epoch": 0.9476415814379484,
"grad_norm": 0.9985774755477905,
"learning_rate": 2.8390538445022244e-06,
"loss": 0.1588,
"step": 970
},
{
"epoch": 0.948618531521905,
"grad_norm": 0.7013076543807983,
"learning_rate": 2.8351352867432774e-06,
"loss": 0.1395,
"step": 971
},
{
"epoch": 0.9495954816058617,
"grad_norm": 0.9600759744644165,
"learning_rate": 2.8312158903718457e-06,
"loss": 0.1591,
"step": 972
},
{
"epoch": 0.9505724316898183,
"grad_norm": 0.6649108529090881,
"learning_rate": 2.8272956651954747e-06,
"loss": 0.1337,
"step": 973
},
{
"epoch": 0.951549381773775,
"grad_norm": 1.0358630418777466,
"learning_rate": 2.8233746210237835e-06,
"loss": 0.1554,
"step": 974
},
{
"epoch": 0.9525263318577316,
"grad_norm": 0.6809327602386475,
"learning_rate": 2.8194527676684413e-06,
"loss": 0.1388,
"step": 975
},
{
"epoch": 0.9535032819416883,
"grad_norm": 0.6123958230018616,
"learning_rate": 2.8155301149431434e-06,
"loss": 0.1452,
"step": 976
},
{
"epoch": 0.9544802320256449,
"grad_norm": 1.0365608930587769,
"learning_rate": 2.8116066726635838e-06,
"loss": 0.1578,
"step": 977
},
{
"epoch": 0.9554571821096016,
"grad_norm": 0.8110082745552063,
"learning_rate": 2.8076824506474316e-06,
"loss": 0.1427,
"step": 978
},
{
"epoch": 0.9564341321935582,
"grad_norm": 0.8344182372093201,
"learning_rate": 2.803757458714309e-06,
"loss": 0.1547,
"step": 979
},
{
"epoch": 0.9574110822775149,
"grad_norm": 1.3430052995681763,
"learning_rate": 2.7998317066857637e-06,
"loss": 0.1413,
"step": 980
},
{
"epoch": 0.9583880323614715,
"grad_norm": 0.8442664742469788,
"learning_rate": 2.7959052043852463e-06,
"loss": 0.1547,
"step": 981
},
{
"epoch": 0.9593649824454282,
"grad_norm": 0.8243014812469482,
"learning_rate": 2.7919779616380815e-06,
"loss": 0.1585,
"step": 982
},
{
"epoch": 0.9603419325293848,
"grad_norm": 0.7175701856613159,
"learning_rate": 2.788049988271452e-06,
"loss": 0.1358,
"step": 983
},
{
"epoch": 0.9613188826133415,
"grad_norm": 0.6697768568992615,
"learning_rate": 2.784121294114365e-06,
"loss": 0.1557,
"step": 984
},
{
"epoch": 0.9622958326972981,
"grad_norm": 0.8436161875724792,
"learning_rate": 2.780191888997632e-06,
"loss": 0.1509,
"step": 985
},
{
"epoch": 0.9632727827812547,
"grad_norm": 0.6790250539779663,
"learning_rate": 2.7762617827538446e-06,
"loss": 0.168,
"step": 986
},
{
"epoch": 0.9642497328652114,
"grad_norm": 0.6713331341743469,
"learning_rate": 2.772330985217348e-06,
"loss": 0.137,
"step": 987
},
{
"epoch": 0.965226682949168,
"grad_norm": 0.9197622537612915,
"learning_rate": 2.7683995062242174e-06,
"loss": 0.1399,
"step": 988
},
{
"epoch": 0.9662036330331247,
"grad_norm": 0.9536051154136658,
"learning_rate": 2.7644673556122335e-06,
"loss": 0.1448,
"step": 989
},
{
"epoch": 0.9671805831170813,
"grad_norm": 0.6290590763092041,
"learning_rate": 2.7605345432208576e-06,
"loss": 0.1454,
"step": 990
},
{
"epoch": 0.968157533201038,
"grad_norm": 1.1950160264968872,
"learning_rate": 2.7566010788912056e-06,
"loss": 0.1497,
"step": 991
},
{
"epoch": 0.9691344832849946,
"grad_norm": 0.9107944965362549,
"learning_rate": 2.7526669724660265e-06,
"loss": 0.1411,
"step": 992
},
{
"epoch": 0.9701114333689513,
"grad_norm": 0.9635177254676819,
"learning_rate": 2.7487322337896757e-06,
"loss": 0.1444,
"step": 993
},
{
"epoch": 0.9710883834529079,
"grad_norm": 1.8326715230941772,
"learning_rate": 2.74479687270809e-06,
"loss": 0.1551,
"step": 994
},
{
"epoch": 0.9720653335368646,
"grad_norm": 0.6276825070381165,
"learning_rate": 2.7408608990687643e-06,
"loss": 0.1388,
"step": 995
},
{
"epoch": 0.9730422836208212,
"grad_norm": 1.9748849868774414,
"learning_rate": 2.736924322720726e-06,
"loss": 0.1547,
"step": 996
},
{
"epoch": 0.9740192337047779,
"grad_norm": 0.749406635761261,
"learning_rate": 2.7329871535145105e-06,
"loss": 0.1548,
"step": 997
},
{
"epoch": 0.9749961837887345,
"grad_norm": 0.6552861332893372,
"learning_rate": 2.7290494013021366e-06,
"loss": 0.1418,
"step": 998
},
{
"epoch": 0.9759731338726912,
"grad_norm": 0.5786459445953369,
"learning_rate": 2.725111075937083e-06,
"loss": 0.152,
"step": 999
},
{
"epoch": 0.9769500839566478,
"grad_norm": 1.0821510553359985,
"learning_rate": 2.721172187274262e-06,
"loss": 0.1577,
"step": 1000
},
{
"epoch": 0.9779270340406044,
"grad_norm": 0.7567124962806702,
"learning_rate": 2.717232745169995e-06,
"loss": 0.1535,
"step": 1001
},
{
"epoch": 0.9789039841245611,
"grad_norm": 0.7972353100776672,
"learning_rate": 2.713292759481988e-06,
"loss": 0.1633,
"step": 1002
},
{
"epoch": 0.9798809342085177,
"grad_norm": 0.7528476119041443,
"learning_rate": 2.7093522400693083e-06,
"loss": 0.1418,
"step": 1003
},
{
"epoch": 0.9808578842924744,
"grad_norm": 0.9000647664070129,
"learning_rate": 2.705411196792359e-06,
"loss": 0.1528,
"step": 1004
},
{
"epoch": 0.981834834376431,
"grad_norm": 0.618922770023346,
"learning_rate": 2.7014696395128526e-06,
"loss": 0.1411,
"step": 1005
},
{
"epoch": 0.9828117844603877,
"grad_norm": 0.791508138179779,
"learning_rate": 2.697527578093788e-06,
"loss": 0.1655,
"step": 1006
},
{
"epoch": 0.9837887345443443,
"grad_norm": 0.6102598905563354,
"learning_rate": 2.6935850223994277e-06,
"loss": 0.1581,
"step": 1007
},
{
"epoch": 0.984765684628301,
"grad_norm": 0.7569069862365723,
"learning_rate": 2.6896419822952686e-06,
"loss": 0.1491,
"step": 1008
},
{
"epoch": 0.9857426347122576,
"grad_norm": 0.7554214596748352,
"learning_rate": 2.68569846764802e-06,
"loss": 0.1649,
"step": 1009
},
{
"epoch": 0.9867195847962144,
"grad_norm": 0.7558314800262451,
"learning_rate": 2.681754488325581e-06,
"loss": 0.1324,
"step": 1010
},
{
"epoch": 0.987696534880171,
"grad_norm": 0.9125105142593384,
"learning_rate": 2.677810054197009e-06,
"loss": 0.1622,
"step": 1011
},
{
"epoch": 0.9886734849641277,
"grad_norm": 0.8719031810760498,
"learning_rate": 2.6738651751325055e-06,
"loss": 0.1443,
"step": 1012
},
{
"epoch": 0.9896504350480843,
"grad_norm": 0.9133313298225403,
"learning_rate": 2.66991986100338e-06,
"loss": 0.1695,
"step": 1013
},
{
"epoch": 0.990627385132041,
"grad_norm": 0.6901757717132568,
"learning_rate": 2.6659741216820335e-06,
"loss": 0.1428,
"step": 1014
},
{
"epoch": 0.9916043352159976,
"grad_norm": 0.8047243356704712,
"learning_rate": 2.662027967041929e-06,
"loss": 0.1474,
"step": 1015
},
{
"epoch": 0.9925812852999542,
"grad_norm": 0.774756133556366,
"learning_rate": 2.658081406957572e-06,
"loss": 0.1532,
"step": 1016
},
{
"epoch": 0.9935582353839109,
"grad_norm": 1.1753309965133667,
"learning_rate": 2.6541344513044794e-06,
"loss": 0.136,
"step": 1017
},
{
"epoch": 0.9945351854678675,
"grad_norm": 0.7303087115287781,
"learning_rate": 2.650187109959158e-06,
"loss": 0.1419,
"step": 1018
},
{
"epoch": 0.9955121355518242,
"grad_norm": 0.7605620622634888,
"learning_rate": 2.6462393927990828e-06,
"loss": 0.1379,
"step": 1019
},
{
"epoch": 0.9964890856357808,
"grad_norm": 0.7614297270774841,
"learning_rate": 2.6422913097026666e-06,
"loss": 0.1577,
"step": 1020
},
{
"epoch": 0.9974660357197375,
"grad_norm": 0.6788558959960938,
"learning_rate": 2.6383428705492374e-06,
"loss": 0.1451,
"step": 1021
},
{
"epoch": 0.9984429858036941,
"grad_norm": 0.9079276323318481,
"learning_rate": 2.634394085219016e-06,
"loss": 0.1627,
"step": 1022
},
{
"epoch": 0.9994199358876508,
"grad_norm": 0.6492640972137451,
"learning_rate": 2.6304449635930882e-06,
"loss": 0.135,
"step": 1023
},
{
"epoch": 1.0,
"grad_norm": 0.7287482619285583,
"learning_rate": 2.626495515553384e-06,
"loss": 0.1315,
"step": 1024
},
{
"epoch": 1.0009769500839567,
"grad_norm": 0.48633936047554016,
"learning_rate": 2.6225457509826446e-06,
"loss": 0.1217,
"step": 1025
},
{
"epoch": 1.0019539001679132,
"grad_norm": 0.4515506327152252,
"learning_rate": 2.618595679764408e-06,
"loss": 0.1232,
"step": 1026
},
{
"epoch": 1.00293085025187,
"grad_norm": 0.5885474681854248,
"learning_rate": 2.6146453117829795e-06,
"loss": 0.1196,
"step": 1027
},
{
"epoch": 1.0039078003358266,
"grad_norm": 0.4799521863460541,
"learning_rate": 2.6106946569234043e-06,
"loss": 0.1174,
"step": 1028
},
{
"epoch": 1.0048847504197833,
"grad_norm": 0.5415709614753723,
"learning_rate": 2.606743725071446e-06,
"loss": 0.114,
"step": 1029
},
{
"epoch": 1.0058617005037398,
"grad_norm": 0.6112764477729797,
"learning_rate": 2.602792526113564e-06,
"loss": 0.138,
"step": 1030
},
{
"epoch": 1.0068386505876965,
"grad_norm": 0.5982470512390137,
"learning_rate": 2.598841069936884e-06,
"loss": 0.1159,
"step": 1031
},
{
"epoch": 1.0078156006716532,
"grad_norm": 0.4844081401824951,
"learning_rate": 2.594889366429174e-06,
"loss": 0.1259,
"step": 1032
},
{
"epoch": 1.00879255075561,
"grad_norm": 0.6176791787147522,
"learning_rate": 2.590937425478823e-06,
"loss": 0.1175,
"step": 1033
},
{
"epoch": 1.0097695008395664,
"grad_norm": 0.6051523089408875,
"learning_rate": 2.5869852569748145e-06,
"loss": 0.129,
"step": 1034
},
{
"epoch": 1.010746450923523,
"grad_norm": 0.44821661710739136,
"learning_rate": 2.583032870806701e-06,
"loss": 0.1241,
"step": 1035
},
{
"epoch": 1.0117234010074798,
"grad_norm": 0.8697763085365295,
"learning_rate": 2.5790802768645767e-06,
"loss": 0.1279,
"step": 1036
},
{
"epoch": 1.0127003510914365,
"grad_norm": 0.5205837488174438,
"learning_rate": 2.5751274850390607e-06,
"loss": 0.126,
"step": 1037
},
{
"epoch": 1.013677301175393,
"grad_norm": 0.5221166610717773,
"learning_rate": 2.5711745052212634e-06,
"loss": 0.1331,
"step": 1038
},
{
"epoch": 1.0146542512593497,
"grad_norm": 0.5351077318191528,
"learning_rate": 2.5672213473027683e-06,
"loss": 0.1291,
"step": 1039
},
{
"epoch": 1.0156312013433064,
"grad_norm": 0.5691295862197876,
"learning_rate": 2.563268021175601e-06,
"loss": 0.1164,
"step": 1040
},
{
"epoch": 1.016608151427263,
"grad_norm": 1.0605528354644775,
"learning_rate": 2.559314536732212e-06,
"loss": 0.1163,
"step": 1041
},
{
"epoch": 1.0175851015112196,
"grad_norm": 0.47866785526275635,
"learning_rate": 2.555360903865446e-06,
"loss": 0.1224,
"step": 1042
},
{
"epoch": 1.0185620515951763,
"grad_norm": 0.7376262545585632,
"learning_rate": 2.551407132468519e-06,
"loss": 0.1273,
"step": 1043
},
{
"epoch": 1.019539001679133,
"grad_norm": 0.556888997554779,
"learning_rate": 2.547453232434994e-06,
"loss": 0.1254,
"step": 1044
},
{
"epoch": 1.0205159517630895,
"grad_norm": 0.519315779209137,
"learning_rate": 2.543499213658756e-06,
"loss": 0.131,
"step": 1045
},
{
"epoch": 1.0214929018470462,
"grad_norm": 0.5682300925254822,
"learning_rate": 2.539545086033987e-06,
"loss": 0.1277,
"step": 1046
},
{
"epoch": 1.022469851931003,
"grad_norm": 0.6019615530967712,
"learning_rate": 2.5355908594551416e-06,
"loss": 0.1261,
"step": 1047
},
{
"epoch": 1.0234468020149596,
"grad_norm": 0.5091032981872559,
"learning_rate": 2.5316365438169206e-06,
"loss": 0.1182,
"step": 1048
},
{
"epoch": 1.0244237520989161,
"grad_norm": 0.6306937336921692,
"learning_rate": 2.52768214901425e-06,
"loss": 0.1325,
"step": 1049
},
{
"epoch": 1.0254007021828728,
"grad_norm": 0.4754994511604309,
"learning_rate": 2.523727684942252e-06,
"loss": 0.1443,
"step": 1050
},
{
"epoch": 1.0263776522668295,
"grad_norm": 0.6034619808197021,
"learning_rate": 2.5197731614962236e-06,
"loss": 0.1398,
"step": 1051
},
{
"epoch": 1.0273546023507862,
"grad_norm": 1.2189825773239136,
"learning_rate": 2.515818588571609e-06,
"loss": 0.1387,
"step": 1052
},
{
"epoch": 1.0283315524347427,
"grad_norm": 0.5273181200027466,
"learning_rate": 2.5118639760639775e-06,
"loss": 0.1329,
"step": 1053
},
{
"epoch": 1.0293085025186994,
"grad_norm": 0.6635386943817139,
"learning_rate": 2.5079093338689953e-06,
"loss": 0.1312,
"step": 1054
},
{
"epoch": 1.0302854526026561,
"grad_norm": 1.444678783416748,
"learning_rate": 2.503954671882406e-06,
"loss": 0.1229,
"step": 1055
},
{
"epoch": 1.0312624026866128,
"grad_norm": 0.43653005361557007,
"learning_rate": 2.5e-06,
"loss": 0.1168,
"step": 1056
},
{
"epoch": 1.0322393527705693,
"grad_norm": 0.4668242335319519,
"learning_rate": 2.496045328117595e-06,
"loss": 0.1263,
"step": 1057
},
{
"epoch": 1.033216302854526,
"grad_norm": 0.4796827435493469,
"learning_rate": 2.4920906661310056e-06,
"loss": 0.1155,
"step": 1058
},
{
"epoch": 1.0341932529384827,
"grad_norm": 0.4911864995956421,
"learning_rate": 2.488136023936023e-06,
"loss": 0.1161,
"step": 1059
},
{
"epoch": 1.0351702030224392,
"grad_norm": 0.5216488242149353,
"learning_rate": 2.4841814114283923e-06,
"loss": 0.1265,
"step": 1060
},
{
"epoch": 1.036147153106396,
"grad_norm": 0.6136938333511353,
"learning_rate": 2.480226838503777e-06,
"loss": 0.1467,
"step": 1061
},
{
"epoch": 1.0371241031903526,
"grad_norm": 0.4947086274623871,
"learning_rate": 2.4762723150577485e-06,
"loss": 0.1265,
"step": 1062
},
{
"epoch": 1.0381010532743093,
"grad_norm": 0.552817702293396,
"learning_rate": 2.4723178509857505e-06,
"loss": 0.1291,
"step": 1063
},
{
"epoch": 1.0390780033582658,
"grad_norm": 0.4872208833694458,
"learning_rate": 2.46836345618308e-06,
"loss": 0.1166,
"step": 1064
},
{
"epoch": 1.0400549534422225,
"grad_norm": 0.6919270157814026,
"learning_rate": 2.4644091405448596e-06,
"loss": 0.1214,
"step": 1065
},
{
"epoch": 1.0410319035261792,
"grad_norm": 0.6018043160438538,
"learning_rate": 2.460454913966013e-06,
"loss": 0.1299,
"step": 1066
},
{
"epoch": 1.042008853610136,
"grad_norm": 0.4866807460784912,
"learning_rate": 2.456500786341244e-06,
"loss": 0.1234,
"step": 1067
},
{
"epoch": 1.0429858036940924,
"grad_norm": 0.4290779232978821,
"learning_rate": 2.452546767565007e-06,
"loss": 0.1129,
"step": 1068
},
{
"epoch": 1.0439627537780491,
"grad_norm": 0.5012803077697754,
"learning_rate": 2.4485928675314816e-06,
"loss": 0.1328,
"step": 1069
},
{
"epoch": 1.0449397038620059,
"grad_norm": 0.6870376467704773,
"learning_rate": 2.4446390961345544e-06,
"loss": 0.1175,
"step": 1070
},
{
"epoch": 1.0459166539459623,
"grad_norm": 0.5841708779335022,
"learning_rate": 2.4406854632677883e-06,
"loss": 0.1126,
"step": 1071
},
{
"epoch": 1.046893604029919,
"grad_norm": 0.4253412187099457,
"learning_rate": 2.4367319788244e-06,
"loss": 0.1227,
"step": 1072
},
{
"epoch": 1.0478705541138758,
"grad_norm": 0.6573653221130371,
"learning_rate": 2.4327786526972325e-06,
"loss": 0.1174,
"step": 1073
},
{
"epoch": 1.0488475041978325,
"grad_norm": 0.7503080368041992,
"learning_rate": 2.4288254947787366e-06,
"loss": 0.1331,
"step": 1074
},
{
"epoch": 1.049824454281789,
"grad_norm": 0.5565494298934937,
"learning_rate": 2.4248725149609397e-06,
"loss": 0.1283,
"step": 1075
},
{
"epoch": 1.0508014043657457,
"grad_norm": 0.5634884834289551,
"learning_rate": 2.420919723135424e-06,
"loss": 0.1367,
"step": 1076
},
{
"epoch": 1.0517783544497024,
"grad_norm": 0.4781632721424103,
"learning_rate": 2.4169671291933e-06,
"loss": 0.1247,
"step": 1077
},
{
"epoch": 1.052755304533659,
"grad_norm": 1.1834274530410767,
"learning_rate": 2.413014743025186e-06,
"loss": 0.1283,
"step": 1078
},
{
"epoch": 1.0537322546176155,
"grad_norm": 0.5243507623672485,
"learning_rate": 2.4090625745211776e-06,
"loss": 0.1294,
"step": 1079
},
{
"epoch": 1.0547092047015723,
"grad_norm": 0.560347318649292,
"learning_rate": 2.405110633570827e-06,
"loss": 0.1269,
"step": 1080
},
{
"epoch": 1.055686154785529,
"grad_norm": 0.7651094794273376,
"learning_rate": 2.4011589300631174e-06,
"loss": 0.1276,
"step": 1081
},
{
"epoch": 1.0566631048694857,
"grad_norm": 1.046125888824463,
"learning_rate": 2.3972074738864365e-06,
"loss": 0.119,
"step": 1082
},
{
"epoch": 1.0576400549534422,
"grad_norm": 0.660923182964325,
"learning_rate": 2.3932562749285544e-06,
"loss": 0.1256,
"step": 1083
},
{
"epoch": 1.0586170050373989,
"grad_norm": 1.0061943531036377,
"learning_rate": 2.3893053430765973e-06,
"loss": 0.1214,
"step": 1084
},
{
"epoch": 1.0595939551213556,
"grad_norm": 0.4880746901035309,
"learning_rate": 2.3853546882170217e-06,
"loss": 0.135,
"step": 1085
},
{
"epoch": 1.0605709052053123,
"grad_norm": 0.6813744902610779,
"learning_rate": 2.3814043202355927e-06,
"loss": 0.1086,
"step": 1086
},
{
"epoch": 1.0615478552892688,
"grad_norm": 0.3869533836841583,
"learning_rate": 2.377454249017356e-06,
"loss": 0.1071,
"step": 1087
},
{
"epoch": 1.0625248053732255,
"grad_norm": 0.6334142684936523,
"learning_rate": 2.373504484446617e-06,
"loss": 0.1441,
"step": 1088
},
{
"epoch": 1.0635017554571822,
"grad_norm": 0.689838171005249,
"learning_rate": 2.3695550364069113e-06,
"loss": 0.1262,
"step": 1089
},
{
"epoch": 1.0644787055411387,
"grad_norm": 0.4681420922279358,
"learning_rate": 2.365605914780984e-06,
"loss": 0.1223,
"step": 1090
},
{
"epoch": 1.0654556556250954,
"grad_norm": 0.6027653217315674,
"learning_rate": 2.361657129450764e-06,
"loss": 0.1177,
"step": 1091
},
{
"epoch": 1.066432605709052,
"grad_norm": 0.886193037033081,
"learning_rate": 2.3577086902973347e-06,
"loss": 0.1308,
"step": 1092
},
{
"epoch": 1.0674095557930088,
"grad_norm": 1.470538854598999,
"learning_rate": 2.3537606072009177e-06,
"loss": 0.1266,
"step": 1093
},
{
"epoch": 1.0683865058769653,
"grad_norm": 0.6389652490615845,
"learning_rate": 2.3498128900408425e-06,
"loss": 0.1115,
"step": 1094
},
{
"epoch": 1.069363455960922,
"grad_norm": 0.45969241857528687,
"learning_rate": 2.3458655486955214e-06,
"loss": 0.1198,
"step": 1095
},
{
"epoch": 1.0703404060448787,
"grad_norm": 0.5180190801620483,
"learning_rate": 2.3419185930424283e-06,
"loss": 0.1214,
"step": 1096
},
{
"epoch": 1.0713173561288354,
"grad_norm": 0.5502235889434814,
"learning_rate": 2.3379720329580714e-06,
"loss": 0.1246,
"step": 1097
},
{
"epoch": 1.0722943062127919,
"grad_norm": 0.7179540991783142,
"learning_rate": 2.3340258783179674e-06,
"loss": 0.1358,
"step": 1098
},
{
"epoch": 1.0732712562967486,
"grad_norm": 0.7727946043014526,
"learning_rate": 2.330080138996621e-06,
"loss": 0.1177,
"step": 1099
},
{
"epoch": 1.0742482063807053,
"grad_norm": 0.4668129086494446,
"learning_rate": 2.3261348248674957e-06,
"loss": 0.1193,
"step": 1100
},
{
"epoch": 1.0752251564646618,
"grad_norm": 0.5279784798622131,
"learning_rate": 2.3221899458029917e-06,
"loss": 0.1142,
"step": 1101
},
{
"epoch": 1.0762021065486185,
"grad_norm": 0.6633546352386475,
"learning_rate": 2.31824551167442e-06,
"loss": 0.1236,
"step": 1102
},
{
"epoch": 1.0771790566325752,
"grad_norm": 1.4349162578582764,
"learning_rate": 2.3143015323519804e-06,
"loss": 0.1314,
"step": 1103
},
{
"epoch": 1.078156006716532,
"grad_norm": 0.4960933327674866,
"learning_rate": 2.3103580177047323e-06,
"loss": 0.1189,
"step": 1104
},
{
"epoch": 1.0791329568004884,
"grad_norm": 0.4428045451641083,
"learning_rate": 2.3064149776005727e-06,
"loss": 0.1167,
"step": 1105
},
{
"epoch": 1.080109906884445,
"grad_norm": 0.49342143535614014,
"learning_rate": 2.3024724219062115e-06,
"loss": 0.1217,
"step": 1106
},
{
"epoch": 1.0810868569684018,
"grad_norm": 0.798454761505127,
"learning_rate": 2.298530360487149e-06,
"loss": 0.1306,
"step": 1107
},
{
"epoch": 1.0820638070523585,
"grad_norm": 0.6384590864181519,
"learning_rate": 2.2945888032076422e-06,
"loss": 0.1148,
"step": 1108
},
{
"epoch": 1.083040757136315,
"grad_norm": 0.5483962893486023,
"learning_rate": 2.290647759930692e-06,
"loss": 0.1204,
"step": 1109
},
{
"epoch": 1.0840177072202717,
"grad_norm": 0.773679792881012,
"learning_rate": 2.286707240518013e-06,
"loss": 0.1134,
"step": 1110
},
{
"epoch": 1.0849946573042284,
"grad_norm": 0.4625346064567566,
"learning_rate": 2.2827672548300063e-06,
"loss": 0.115,
"step": 1111
},
{
"epoch": 1.085971607388185,
"grad_norm": 0.9028828144073486,
"learning_rate": 2.278827812725739e-06,
"loss": 0.1241,
"step": 1112
},
{
"epoch": 1.0869485574721416,
"grad_norm": 0.5289633274078369,
"learning_rate": 2.274888924062917e-06,
"loss": 0.1146,
"step": 1113
},
{
"epoch": 1.0879255075560983,
"grad_norm": 0.501119077205658,
"learning_rate": 2.2709505986978633e-06,
"loss": 0.1231,
"step": 1114
},
{
"epoch": 1.088902457640055,
"grad_norm": 1.0217068195343018,
"learning_rate": 2.2670128464854908e-06,
"loss": 0.1457,
"step": 1115
},
{
"epoch": 1.0898794077240117,
"grad_norm": 0.597627580165863,
"learning_rate": 2.263075677279275e-06,
"loss": 0.1243,
"step": 1116
},
{
"epoch": 1.0908563578079682,
"grad_norm": 0.7789493799209595,
"learning_rate": 2.259139100931236e-06,
"loss": 0.1232,
"step": 1117
},
{
"epoch": 1.091833307891925,
"grad_norm": 0.485882431268692,
"learning_rate": 2.2552031272919105e-06,
"loss": 0.1321,
"step": 1118
},
{
"epoch": 1.0928102579758816,
"grad_norm": 0.5639163851737976,
"learning_rate": 2.2512677662103248e-06,
"loss": 0.1244,
"step": 1119
},
{
"epoch": 1.093787208059838,
"grad_norm": 0.5579909682273865,
"learning_rate": 2.2473330275339744e-06,
"loss": 0.1331,
"step": 1120
},
{
"epoch": 1.0947641581437948,
"grad_norm": 0.6492118835449219,
"learning_rate": 2.243398921108795e-06,
"loss": 0.1225,
"step": 1121
},
{
"epoch": 1.0957411082277515,
"grad_norm": 0.4354166090488434,
"learning_rate": 2.239465456779144e-06,
"loss": 0.1185,
"step": 1122
},
{
"epoch": 1.0967180583117082,
"grad_norm": 0.6575405597686768,
"learning_rate": 2.235532644387767e-06,
"loss": 0.1183,
"step": 1123
},
{
"epoch": 1.0976950083956647,
"grad_norm": 0.45513030886650085,
"learning_rate": 2.231600493775783e-06,
"loss": 0.1114,
"step": 1124
},
{
"epoch": 1.0986719584796214,
"grad_norm": 0.6796538829803467,
"learning_rate": 2.227669014782653e-06,
"loss": 0.1293,
"step": 1125
},
{
"epoch": 1.099648908563578,
"grad_norm": 0.5638039112091064,
"learning_rate": 2.2237382172461563e-06,
"loss": 0.1378,
"step": 1126
},
{
"epoch": 1.1006258586475348,
"grad_norm": 0.8584489822387695,
"learning_rate": 2.219808111002369e-06,
"loss": 0.1428,
"step": 1127
},
{
"epoch": 1.1016028087314913,
"grad_norm": 0.5264009237289429,
"learning_rate": 2.2158787058856356e-06,
"loss": 0.1202,
"step": 1128
},
{
"epoch": 1.102579758815448,
"grad_norm": 0.5003917217254639,
"learning_rate": 2.2119500117285485e-06,
"loss": 0.1128,
"step": 1129
},
{
"epoch": 1.1035567088994047,
"grad_norm": 0.6145273447036743,
"learning_rate": 2.208022038361919e-06,
"loss": 0.1196,
"step": 1130
},
{
"epoch": 1.1045336589833614,
"grad_norm": 0.497677206993103,
"learning_rate": 2.204094795614755e-06,
"loss": 0.1145,
"step": 1131
},
{
"epoch": 1.105510609067318,
"grad_norm": 0.6417908072471619,
"learning_rate": 2.2001682933142367e-06,
"loss": 0.1275,
"step": 1132
},
{
"epoch": 1.1064875591512746,
"grad_norm": 0.5320245027542114,
"learning_rate": 2.196242541285692e-06,
"loss": 0.118,
"step": 1133
},
{
"epoch": 1.1074645092352313,
"grad_norm": 1.0965776443481445,
"learning_rate": 2.192317549352569e-06,
"loss": 0.1203,
"step": 1134
},
{
"epoch": 1.108441459319188,
"grad_norm": 1.9805694818496704,
"learning_rate": 2.188393327336417e-06,
"loss": 0.123,
"step": 1135
},
{
"epoch": 1.1094184094031445,
"grad_norm": 0.5877267122268677,
"learning_rate": 2.184469885056857e-06,
"loss": 0.1269,
"step": 1136
},
{
"epoch": 1.1103953594871012,
"grad_norm": 0.4872968792915344,
"learning_rate": 2.180547232331559e-06,
"loss": 0.1194,
"step": 1137
},
{
"epoch": 1.111372309571058,
"grad_norm": 0.5158323645591736,
"learning_rate": 2.1766253789762177e-06,
"loss": 0.1248,
"step": 1138
},
{
"epoch": 1.1123492596550144,
"grad_norm": 0.7523319721221924,
"learning_rate": 2.1727043348045265e-06,
"loss": 0.139,
"step": 1139
},
{
"epoch": 1.1133262097389711,
"grad_norm": 0.8644654154777527,
"learning_rate": 2.168784109628155e-06,
"loss": 0.1436,
"step": 1140
},
{
"epoch": 1.1143031598229278,
"grad_norm": 0.5241405367851257,
"learning_rate": 2.164864713256723e-06,
"loss": 0.116,
"step": 1141
},
{
"epoch": 1.1152801099068845,
"grad_norm": 0.4895915389060974,
"learning_rate": 2.160946155497776e-06,
"loss": 0.1482,
"step": 1142
},
{
"epoch": 1.116257059990841,
"grad_norm": 0.45186084508895874,
"learning_rate": 2.1570284461567633e-06,
"loss": 0.1119,
"step": 1143
},
{
"epoch": 1.1172340100747977,
"grad_norm": 0.5475239753723145,
"learning_rate": 2.1531115950370073e-06,
"loss": 0.1186,
"step": 1144
},
{
"epoch": 1.1182109601587544,
"grad_norm": 0.49343541264533997,
"learning_rate": 2.1491956119396865e-06,
"loss": 0.1181,
"step": 1145
},
{
"epoch": 1.1191879102427111,
"grad_norm": 0.44927504658699036,
"learning_rate": 2.1452805066638056e-06,
"loss": 0.119,
"step": 1146
},
{
"epoch": 1.1201648603266676,
"grad_norm": 0.6453165411949158,
"learning_rate": 2.141366289006172e-06,
"loss": 0.1396,
"step": 1147
},
{
"epoch": 1.1211418104106243,
"grad_norm": 0.9072570204734802,
"learning_rate": 2.1374529687613733e-06,
"loss": 0.1224,
"step": 1148
},
{
"epoch": 1.122118760494581,
"grad_norm": 0.7782952189445496,
"learning_rate": 2.1335405557217507e-06,
"loss": 0.1231,
"step": 1149
},
{
"epoch": 1.1230957105785375,
"grad_norm": 0.5067862868309021,
"learning_rate": 2.129629059677377e-06,
"loss": 0.1195,
"step": 1150
},
{
"epoch": 1.1240726606624942,
"grad_norm": 0.6701231598854065,
"learning_rate": 2.125718490416028e-06,
"loss": 0.1259,
"step": 1151
},
{
"epoch": 1.125049610746451,
"grad_norm": 1.0517492294311523,
"learning_rate": 2.121808857723162e-06,
"loss": 0.1276,
"step": 1152
},
{
"epoch": 1.1260265608304076,
"grad_norm": 0.6626511216163635,
"learning_rate": 2.117900171381894e-06,
"loss": 0.1241,
"step": 1153
},
{
"epoch": 1.1270035109143641,
"grad_norm": 0.47597432136535645,
"learning_rate": 2.113992441172969e-06,
"loss": 0.1144,
"step": 1154
},
{
"epoch": 1.1279804609983208,
"grad_norm": 0.6432631015777588,
"learning_rate": 2.1100856768747403e-06,
"loss": 0.1277,
"step": 1155
},
{
"epoch": 1.1289574110822775,
"grad_norm": 0.5803653001785278,
"learning_rate": 2.1061798882631445e-06,
"loss": 0.1122,
"step": 1156
},
{
"epoch": 1.1299343611662342,
"grad_norm": 0.6182239651679993,
"learning_rate": 2.102275085111678e-06,
"loss": 0.1423,
"step": 1157
},
{
"epoch": 1.1309113112501907,
"grad_norm": 0.5685151815414429,
"learning_rate": 2.0983712771913693e-06,
"loss": 0.1255,
"step": 1158
},
{
"epoch": 1.1318882613341474,
"grad_norm": 0.6879568696022034,
"learning_rate": 2.0944684742707573e-06,
"loss": 0.1122,
"step": 1159
},
{
"epoch": 1.1328652114181041,
"grad_norm": 0.6084652543067932,
"learning_rate": 2.0905666861158665e-06,
"loss": 0.1406,
"step": 1160
},
{
"epoch": 1.1338421615020606,
"grad_norm": 1.02890944480896,
"learning_rate": 2.086665922490181e-06,
"loss": 0.1251,
"step": 1161
},
{
"epoch": 1.1348191115860173,
"grad_norm": 0.5203421115875244,
"learning_rate": 2.0827661931546215e-06,
"loss": 0.1261,
"step": 1162
},
{
"epoch": 1.135796061669974,
"grad_norm": 0.7673782110214233,
"learning_rate": 2.0788675078675215e-06,
"loss": 0.1095,
"step": 1163
},
{
"epoch": 1.1367730117539308,
"grad_norm": 0.8322330117225647,
"learning_rate": 2.0749698763846017e-06,
"loss": 0.1357,
"step": 1164
},
{
"epoch": 1.1377499618378875,
"grad_norm": 0.7669651508331299,
"learning_rate": 2.071073308458945e-06,
"loss": 0.1384,
"step": 1165
},
{
"epoch": 1.138726911921844,
"grad_norm": 0.7110041379928589,
"learning_rate": 2.067177813840974e-06,
"loss": 0.1147,
"step": 1166
},
{
"epoch": 1.1397038620058006,
"grad_norm": 0.6263896822929382,
"learning_rate": 2.0632834022784247e-06,
"loss": 0.1219,
"step": 1167
},
{
"epoch": 1.1406808120897574,
"grad_norm": 0.4878046214580536,
"learning_rate": 2.059390083516323e-06,
"loss": 0.1258,
"step": 1168
},
{
"epoch": 1.1416577621737138,
"grad_norm": 0.6990064382553101,
"learning_rate": 2.0554978672969626e-06,
"loss": 0.1362,
"step": 1169
},
{
"epoch": 1.1426347122576705,
"grad_norm": 0.45585277676582336,
"learning_rate": 2.0516067633598743e-06,
"loss": 0.1139,
"step": 1170
},
{
"epoch": 1.1436116623416273,
"grad_norm": 0.5277510285377502,
"learning_rate": 2.047716781441807e-06,
"loss": 0.1219,
"step": 1171
},
{
"epoch": 1.144588612425584,
"grad_norm": 0.5456313490867615,
"learning_rate": 2.043827931276704e-06,
"loss": 0.113,
"step": 1172
},
{
"epoch": 1.1455655625095404,
"grad_norm": 0.6312161684036255,
"learning_rate": 2.0399402225956757e-06,
"loss": 0.1187,
"step": 1173
},
{
"epoch": 1.1465425125934972,
"grad_norm": 1.6021937131881714,
"learning_rate": 2.0360536651269743e-06,
"loss": 0.147,
"step": 1174
},
{
"epoch": 1.1475194626774539,
"grad_norm": 0.4911890923976898,
"learning_rate": 2.0321682685959727e-06,
"loss": 0.1252,
"step": 1175
},
{
"epoch": 1.1484964127614106,
"grad_norm": 0.661885142326355,
"learning_rate": 2.0282840427251395e-06,
"loss": 0.1418,
"step": 1176
},
{
"epoch": 1.149473362845367,
"grad_norm": 0.5701663494110107,
"learning_rate": 2.024400997234014e-06,
"loss": 0.1215,
"step": 1177
},
{
"epoch": 1.1504503129293238,
"grad_norm": 0.6612499356269836,
"learning_rate": 2.0205191418391787e-06,
"loss": 0.1113,
"step": 1178
},
{
"epoch": 1.1514272630132805,
"grad_norm": 0.7011617422103882,
"learning_rate": 2.016638486254242e-06,
"loss": 0.143,
"step": 1179
},
{
"epoch": 1.152404213097237,
"grad_norm": 1.0519636869430542,
"learning_rate": 2.0127590401898075e-06,
"loss": 0.1223,
"step": 1180
},
{
"epoch": 1.1533811631811937,
"grad_norm": 0.5051503777503967,
"learning_rate": 2.008880813353454e-06,
"loss": 0.1199,
"step": 1181
},
{
"epoch": 1.1543581132651504,
"grad_norm": 0.5821029543876648,
"learning_rate": 2.0050038154497074e-06,
"loss": 0.124,
"step": 1182
},
{
"epoch": 1.155335063349107,
"grad_norm": 0.6066112518310547,
"learning_rate": 2.001128056180021e-06,
"loss": 0.1409,
"step": 1183
},
{
"epoch": 1.1563120134330638,
"grad_norm": 0.6067060232162476,
"learning_rate": 1.997253545242747e-06,
"loss": 0.1334,
"step": 1184
},
{
"epoch": 1.1572889635170203,
"grad_norm": 1.2235379219055176,
"learning_rate": 1.993380292333113e-06,
"loss": 0.1323,
"step": 1185
},
{
"epoch": 1.158265913600977,
"grad_norm": 0.4805576205253601,
"learning_rate": 1.9895083071432e-06,
"loss": 0.1374,
"step": 1186
},
{
"epoch": 1.1592428636849337,
"grad_norm": 0.6857699155807495,
"learning_rate": 1.9856375993619175e-06,
"loss": 0.1218,
"step": 1187
},
{
"epoch": 1.1602198137688902,
"grad_norm": 2.302901268005371,
"learning_rate": 1.9817681786749758e-06,
"loss": 0.1229,
"step": 1188
},
{
"epoch": 1.1611967638528469,
"grad_norm": 1.1203244924545288,
"learning_rate": 1.9779000547648675e-06,
"loss": 0.1265,
"step": 1189
},
{
"epoch": 1.1621737139368036,
"grad_norm": 1.30556321144104,
"learning_rate": 1.974033237310839e-06,
"loss": 0.1595,
"step": 1190
},
{
"epoch": 1.1631506640207603,
"grad_norm": 0.4282764196395874,
"learning_rate": 1.9701677359888667e-06,
"loss": 0.1193,
"step": 1191
},
{
"epoch": 1.1641276141047168,
"grad_norm": 0.5991023182868958,
"learning_rate": 1.966303560471636e-06,
"loss": 0.1293,
"step": 1192
},
{
"epoch": 1.1651045641886735,
"grad_norm": 0.4783615469932556,
"learning_rate": 1.9624407204285114e-06,
"loss": 0.1218,
"step": 1193
},
{
"epoch": 1.1660815142726302,
"grad_norm": 0.6126300692558289,
"learning_rate": 1.958579225525518e-06,
"loss": 0.1193,
"step": 1194
},
{
"epoch": 1.1670584643565869,
"grad_norm": 0.5440214276313782,
"learning_rate": 1.954719085425314e-06,
"loss": 0.1287,
"step": 1195
},
{
"epoch": 1.1680354144405434,
"grad_norm": 0.508152425289154,
"learning_rate": 1.9508603097871686e-06,
"loss": 0.124,
"step": 1196
},
{
"epoch": 1.1690123645245,
"grad_norm": 0.43889936804771423,
"learning_rate": 1.947002908266935e-06,
"loss": 0.1163,
"step": 1197
},
{
"epoch": 1.1699893146084568,
"grad_norm": 0.8772429823875427,
"learning_rate": 1.94314689051703e-06,
"loss": 0.1449,
"step": 1198
},
{
"epoch": 1.1709662646924133,
"grad_norm": 0.9620941877365112,
"learning_rate": 1.9392922661864056e-06,
"loss": 0.1275,
"step": 1199
},
{
"epoch": 1.17194321477637,
"grad_norm": 0.6348773837089539,
"learning_rate": 1.935439044920529e-06,
"loss": 0.1273,
"step": 1200
},
{
"epoch": 1.1729201648603267,
"grad_norm": 0.48426857590675354,
"learning_rate": 1.931587236361354e-06,
"loss": 0.1161,
"step": 1201
},
{
"epoch": 1.1738971149442834,
"grad_norm": 0.5368437170982361,
"learning_rate": 1.9277368501473022e-06,
"loss": 0.1315,
"step": 1202
},
{
"epoch": 1.1748740650282399,
"grad_norm": 0.560650646686554,
"learning_rate": 1.923887895913234e-06,
"loss": 0.1232,
"step": 1203
},
{
"epoch": 1.1758510151121966,
"grad_norm": 0.6062894463539124,
"learning_rate": 1.9200403832904276e-06,
"loss": 0.1061,
"step": 1204
},
{
"epoch": 1.1768279651961533,
"grad_norm": 0.5022917985916138,
"learning_rate": 1.9161943219065535e-06,
"loss": 0.1262,
"step": 1205
},
{
"epoch": 1.17780491528011,
"grad_norm": 0.5425379872322083,
"learning_rate": 1.912349721385651e-06,
"loss": 0.1277,
"step": 1206
},
{
"epoch": 1.1787818653640665,
"grad_norm": 0.43464258313179016,
"learning_rate": 1.9085065913481033e-06,
"loss": 0.1199,
"step": 1207
},
{
"epoch": 1.1797588154480232,
"grad_norm": 0.615777850151062,
"learning_rate": 1.9046649414106152e-06,
"loss": 0.1264,
"step": 1208
},
{
"epoch": 1.18073576553198,
"grad_norm": 0.7756106853485107,
"learning_rate": 1.9008247811861858e-06,
"loss": 0.1263,
"step": 1209
},
{
"epoch": 1.1817127156159364,
"grad_norm": 0.5153846144676208,
"learning_rate": 1.8969861202840877e-06,
"loss": 0.1221,
"step": 1210
},
{
"epoch": 1.182689665699893,
"grad_norm": 0.5160000324249268,
"learning_rate": 1.893148968309842e-06,
"loss": 0.1215,
"step": 1211
},
{
"epoch": 1.1836666157838498,
"grad_norm": 0.8004584312438965,
"learning_rate": 1.8893133348651943e-06,
"loss": 0.1282,
"step": 1212
},
{
"epoch": 1.1846435658678065,
"grad_norm": 0.6641963124275208,
"learning_rate": 1.885479229548089e-06,
"loss": 0.1231,
"step": 1213
},
{
"epoch": 1.1856205159517632,
"grad_norm": 0.5261145234107971,
"learning_rate": 1.8816466619526477e-06,
"loss": 0.1135,
"step": 1214
},
{
"epoch": 1.1865974660357197,
"grad_norm": 0.8460418581962585,
"learning_rate": 1.8778156416691435e-06,
"loss": 0.1255,
"step": 1215
},
{
"epoch": 1.1875744161196764,
"grad_norm": 0.5295639038085938,
"learning_rate": 1.8739861782839795e-06,
"loss": 0.1226,
"step": 1216
},
{
"epoch": 1.188551366203633,
"grad_norm": 0.953988254070282,
"learning_rate": 1.870158281379659e-06,
"loss": 0.1149,
"step": 1217
},
{
"epoch": 1.1895283162875896,
"grad_norm": 1.4285132884979248,
"learning_rate": 1.8663319605347685e-06,
"loss": 0.1329,
"step": 1218
},
{
"epoch": 1.1905052663715463,
"grad_norm": 0.9437524080276489,
"learning_rate": 1.8625072253239518e-06,
"loss": 0.1405,
"step": 1219
},
{
"epoch": 1.191482216455503,
"grad_norm": 0.640650749206543,
"learning_rate": 1.8586840853178818e-06,
"loss": 0.1106,
"step": 1220
},
{
"epoch": 1.1924591665394597,
"grad_norm": 0.5248428583145142,
"learning_rate": 1.8548625500832412e-06,
"loss": 0.1173,
"step": 1221
},
{
"epoch": 1.1934361166234162,
"grad_norm": 0.7590430974960327,
"learning_rate": 1.8510426291826977e-06,
"loss": 0.1255,
"step": 1222
},
{
"epoch": 1.194413066707373,
"grad_norm": 1.049569010734558,
"learning_rate": 1.8472243321748793e-06,
"loss": 0.1286,
"step": 1223
},
{
"epoch": 1.1953900167913296,
"grad_norm": 0.5904210209846497,
"learning_rate": 1.8434076686143482e-06,
"loss": 0.1202,
"step": 1224
},
{
"epoch": 1.1963669668752863,
"grad_norm": 0.8871178030967712,
"learning_rate": 1.839592648051582e-06,
"loss": 0.1194,
"step": 1225
},
{
"epoch": 1.1973439169592428,
"grad_norm": 0.5890358686447144,
"learning_rate": 1.835779280032946e-06,
"loss": 0.1135,
"step": 1226
},
{
"epoch": 1.1983208670431995,
"grad_norm": 0.4250822067260742,
"learning_rate": 1.8319675741006704e-06,
"loss": 0.1219,
"step": 1227
},
{
"epoch": 1.1992978171271562,
"grad_norm": 0.5853110551834106,
"learning_rate": 1.8281575397928264e-06,
"loss": 0.1108,
"step": 1228
},
{
"epoch": 1.2002747672111127,
"grad_norm": 0.4559549391269684,
"learning_rate": 1.824349186643302e-06,
"loss": 0.1104,
"step": 1229
},
{
"epoch": 1.2012517172950694,
"grad_norm": 0.5680705904960632,
"learning_rate": 1.8205425241817776e-06,
"loss": 0.1326,
"step": 1230
},
{
"epoch": 1.2022286673790261,
"grad_norm": 0.863606333732605,
"learning_rate": 1.816737561933707e-06,
"loss": 0.1338,
"step": 1231
},
{
"epoch": 1.2032056174629828,
"grad_norm": 1.035900354385376,
"learning_rate": 1.8129343094202826e-06,
"loss": 0.1335,
"step": 1232
},
{
"epoch": 1.2041825675469393,
"grad_norm": 0.46370747685432434,
"learning_rate": 1.8091327761584247e-06,
"loss": 0.1374,
"step": 1233
},
{
"epoch": 1.205159517630896,
"grad_norm": 0.7283802628517151,
"learning_rate": 1.8053329716607488e-06,
"loss": 0.1272,
"step": 1234
},
{
"epoch": 1.2061364677148527,
"grad_norm": 0.5614640712738037,
"learning_rate": 1.8015349054355447e-06,
"loss": 0.1147,
"step": 1235
},
{
"epoch": 1.2071134177988094,
"grad_norm": 0.6142383217811584,
"learning_rate": 1.7977385869867525e-06,
"loss": 0.1179,
"step": 1236
},
{
"epoch": 1.208090367882766,
"grad_norm": 0.5201336145401001,
"learning_rate": 1.793944025813939e-06,
"loss": 0.1166,
"step": 1237
},
{
"epoch": 1.2090673179667226,
"grad_norm": 1.0300626754760742,
"learning_rate": 1.790151231412275e-06,
"loss": 0.1284,
"step": 1238
},
{
"epoch": 1.2100442680506793,
"grad_norm": 0.5696738362312317,
"learning_rate": 1.7863602132725082e-06,
"loss": 0.1288,
"step": 1239
},
{
"epoch": 1.2110212181346358,
"grad_norm": 0.6256274580955505,
"learning_rate": 1.7825709808809416e-06,
"loss": 0.1155,
"step": 1240
},
{
"epoch": 1.2119981682185925,
"grad_norm": 1.0086814165115356,
"learning_rate": 1.7787835437194117e-06,
"loss": 0.1224,
"step": 1241
},
{
"epoch": 1.2129751183025492,
"grad_norm": 0.7290226221084595,
"learning_rate": 1.774997911265261e-06,
"loss": 0.1192,
"step": 1242
},
{
"epoch": 1.213952068386506,
"grad_norm": 0.5742580890655518,
"learning_rate": 1.771214092991317e-06,
"loss": 0.1184,
"step": 1243
},
{
"epoch": 1.2149290184704626,
"grad_norm": 0.919278621673584,
"learning_rate": 1.7674320983658661e-06,
"loss": 0.1234,
"step": 1244
},
{
"epoch": 1.2159059685544191,
"grad_norm": 0.46248340606689453,
"learning_rate": 1.7636519368526338e-06,
"loss": 0.1059,
"step": 1245
},
{
"epoch": 1.2168829186383758,
"grad_norm": 0.9081810116767883,
"learning_rate": 1.759873617910757e-06,
"loss": 0.1329,
"step": 1246
},
{
"epoch": 1.2178598687223325,
"grad_norm": 0.7663976550102234,
"learning_rate": 1.7560971509947627e-06,
"loss": 0.1272,
"step": 1247
},
{
"epoch": 1.218836818806289,
"grad_norm": 0.6514719724655151,
"learning_rate": 1.7523225455545415e-06,
"loss": 0.1077,
"step": 1248
},
{
"epoch": 1.2198137688902457,
"grad_norm": 0.5231809020042419,
"learning_rate": 1.7485498110353283e-06,
"loss": 0.1162,
"step": 1249
},
{
"epoch": 1.2207907189742024,
"grad_norm": 0.7818597555160522,
"learning_rate": 1.7447789568776757e-06,
"loss": 0.1392,
"step": 1250
},
{
"epoch": 1.2217676690581591,
"grad_norm": 0.5452788472175598,
"learning_rate": 1.7410099925174306e-06,
"loss": 0.1232,
"step": 1251
},
{
"epoch": 1.2227446191421156,
"grad_norm": 0.65793776512146,
"learning_rate": 1.7372429273857122e-06,
"loss": 0.1241,
"step": 1252
},
{
"epoch": 1.2237215692260723,
"grad_norm": 0.4611707925796509,
"learning_rate": 1.7334777709088857e-06,
"loss": 0.1267,
"step": 1253
},
{
"epoch": 1.224698519310029,
"grad_norm": 0.43640777468681335,
"learning_rate": 1.7297145325085418e-06,
"loss": 0.1076,
"step": 1254
},
{
"epoch": 1.2256754693939858,
"grad_norm": 0.5017767548561096,
"learning_rate": 1.7259532216014702e-06,
"loss": 0.1177,
"step": 1255
},
{
"epoch": 1.2266524194779422,
"grad_norm": 0.5469508171081543,
"learning_rate": 1.7221938475996375e-06,
"loss": 0.1215,
"step": 1256
},
{
"epoch": 1.227629369561899,
"grad_norm": 0.543257474899292,
"learning_rate": 1.7184364199101649e-06,
"loss": 0.1262,
"step": 1257
},
{
"epoch": 1.2286063196458556,
"grad_norm": 0.6066008806228638,
"learning_rate": 1.7146809479353013e-06,
"loss": 0.1189,
"step": 1258
},
{
"epoch": 1.2295832697298121,
"grad_norm": 0.8405829668045044,
"learning_rate": 1.7109274410724045e-06,
"loss": 0.122,
"step": 1259
},
{
"epoch": 1.2305602198137688,
"grad_norm": 0.609324038028717,
"learning_rate": 1.7071759087139128e-06,
"loss": 0.1233,
"step": 1260
},
{
"epoch": 1.2315371698977255,
"grad_norm": 0.520811915397644,
"learning_rate": 1.703426360247324e-06,
"loss": 0.1283,
"step": 1261
},
{
"epoch": 1.2325141199816823,
"grad_norm": 0.5709425210952759,
"learning_rate": 1.6996788050551733e-06,
"loss": 0.1294,
"step": 1262
},
{
"epoch": 1.2334910700656387,
"grad_norm": 1.2291886806488037,
"learning_rate": 1.6959332525150046e-06,
"loss": 0.1294,
"step": 1263
},
{
"epoch": 1.2344680201495954,
"grad_norm": 0.5380057096481323,
"learning_rate": 1.6921897119993535e-06,
"loss": 0.1162,
"step": 1264
},
{
"epoch": 1.2354449702335522,
"grad_norm": 0.4606705904006958,
"learning_rate": 1.6884481928757208e-06,
"loss": 0.1147,
"step": 1265
},
{
"epoch": 1.2364219203175089,
"grad_norm": 0.7243968844413757,
"learning_rate": 1.6847087045065468e-06,
"loss": 0.1238,
"step": 1266
},
{
"epoch": 1.2373988704014653,
"grad_norm": 0.49404072761535645,
"learning_rate": 1.6809712562491937e-06,
"loss": 0.1355,
"step": 1267
},
{
"epoch": 1.238375820485422,
"grad_norm": 0.5434608459472656,
"learning_rate": 1.677235857455916e-06,
"loss": 0.1209,
"step": 1268
},
{
"epoch": 1.2393527705693788,
"grad_norm": 0.6871845722198486,
"learning_rate": 1.6735025174738407e-06,
"loss": 0.1204,
"step": 1269
},
{
"epoch": 1.2403297206533355,
"grad_norm": 0.5333722233772278,
"learning_rate": 1.6697712456449432e-06,
"loss": 0.1259,
"step": 1270
},
{
"epoch": 1.241306670737292,
"grad_norm": 0.6351161599159241,
"learning_rate": 1.6660420513060226e-06,
"loss": 0.1293,
"step": 1271
},
{
"epoch": 1.2422836208212487,
"grad_norm": 0.6560023427009583,
"learning_rate": 1.6623149437886813e-06,
"loss": 0.1507,
"step": 1272
},
{
"epoch": 1.2432605709052054,
"grad_norm": 0.5429563522338867,
"learning_rate": 1.6585899324192978e-06,
"loss": 0.1296,
"step": 1273
},
{
"epoch": 1.244237520989162,
"grad_norm": 0.5854029059410095,
"learning_rate": 1.6548670265190076e-06,
"loss": 0.1294,
"step": 1274
},
{
"epoch": 1.2452144710731186,
"grad_norm": 0.5841599702835083,
"learning_rate": 1.6511462354036765e-06,
"loss": 0.1122,
"step": 1275
},
{
"epoch": 1.2461914211570753,
"grad_norm": 0.6738702058792114,
"learning_rate": 1.647427568383878e-06,
"loss": 0.1293,
"step": 1276
},
{
"epoch": 1.247168371241032,
"grad_norm": 0.7624144554138184,
"learning_rate": 1.6437110347648706e-06,
"loss": 0.1154,
"step": 1277
},
{
"epoch": 1.2481453213249885,
"grad_norm": 0.48225340247154236,
"learning_rate": 1.6399966438465768e-06,
"loss": 0.1509,
"step": 1278
},
{
"epoch": 1.2491222714089452,
"grad_norm": 1.0383435487747192,
"learning_rate": 1.6362844049235527e-06,
"loss": 0.1283,
"step": 1279
},
{
"epoch": 1.2500992214929019,
"grad_norm": 0.557651698589325,
"learning_rate": 1.6325743272849726e-06,
"loss": 0.1121,
"step": 1280
},
{
"epoch": 1.2510761715768586,
"grad_norm": 0.8385334014892578,
"learning_rate": 1.628866420214603e-06,
"loss": 0.1393,
"step": 1281
},
{
"epoch": 1.252053121660815,
"grad_norm": 1.031480312347412,
"learning_rate": 1.6251606929907777e-06,
"loss": 0.1219,
"step": 1282
},
{
"epoch": 1.2530300717447718,
"grad_norm": 0.9081658720970154,
"learning_rate": 1.6214571548863756e-06,
"loss": 0.1462,
"step": 1283
},
{
"epoch": 1.2540070218287285,
"grad_norm": 0.5723317861557007,
"learning_rate": 1.617755815168799e-06,
"loss": 0.1246,
"step": 1284
},
{
"epoch": 1.2549839719126852,
"grad_norm": 0.5951042175292969,
"learning_rate": 1.614056683099949e-06,
"loss": 0.122,
"step": 1285
},
{
"epoch": 1.2559609219966417,
"grad_norm": 0.8381775617599487,
"learning_rate": 1.6103597679362005e-06,
"loss": 0.125,
"step": 1286
},
{
"epoch": 1.2569378720805984,
"grad_norm": 0.5216977596282959,
"learning_rate": 1.6066650789283824e-06,
"loss": 0.1315,
"step": 1287
},
{
"epoch": 1.257914822164555,
"grad_norm": 0.5493628978729248,
"learning_rate": 1.6029726253217548e-06,
"loss": 0.1129,
"step": 1288
},
{
"epoch": 1.2588917722485116,
"grad_norm": 0.636802077293396,
"learning_rate": 1.5992824163559809e-06,
"loss": 0.1264,
"step": 1289
},
{
"epoch": 1.2598687223324683,
"grad_norm": 0.6565551161766052,
"learning_rate": 1.5955944612651094e-06,
"loss": 0.1131,
"step": 1290
},
{
"epoch": 1.260845672416425,
"grad_norm": 0.7052702307701111,
"learning_rate": 1.5919087692775485e-06,
"loss": 0.1281,
"step": 1291
},
{
"epoch": 1.2618226225003817,
"grad_norm": 1.0554252862930298,
"learning_rate": 1.5882253496160422e-06,
"loss": 0.1322,
"step": 1292
},
{
"epoch": 1.2627995725843384,
"grad_norm": 0.8078587651252747,
"learning_rate": 1.584544211497652e-06,
"loss": 0.1118,
"step": 1293
},
{
"epoch": 1.2637765226682949,
"grad_norm": 0.6466701030731201,
"learning_rate": 1.5808653641337246e-06,
"loss": 0.1348,
"step": 1294
},
{
"epoch": 1.2647534727522516,
"grad_norm": 0.6934959888458252,
"learning_rate": 1.5771888167298796e-06,
"loss": 0.1199,
"step": 1295
},
{
"epoch": 1.2657304228362083,
"grad_norm": 0.6258717775344849,
"learning_rate": 1.573514578485979e-06,
"loss": 0.1486,
"step": 1296
},
{
"epoch": 1.2667073729201648,
"grad_norm": 1.4640861749649048,
"learning_rate": 1.5698426585961074e-06,
"loss": 0.1434,
"step": 1297
},
{
"epoch": 1.2676843230041215,
"grad_norm": 0.5409911274909973,
"learning_rate": 1.5661730662485477e-06,
"loss": 0.1147,
"step": 1298
},
{
"epoch": 1.2686612730880782,
"grad_norm": 1.5885776281356812,
"learning_rate": 1.5625058106257578e-06,
"loss": 0.1212,
"step": 1299
},
{
"epoch": 1.2696382231720347,
"grad_norm": 1.986743688583374,
"learning_rate": 1.558840900904351e-06,
"loss": 0.1383,
"step": 1300
},
{
"epoch": 1.2706151732559914,
"grad_norm": 0.696490466594696,
"learning_rate": 1.5551783462550683e-06,
"loss": 0.1287,
"step": 1301
},
{
"epoch": 1.271592123339948,
"grad_norm": 0.5482082366943359,
"learning_rate": 1.5515181558427567e-06,
"loss": 0.1293,
"step": 1302
},
{
"epoch": 1.2725690734239048,
"grad_norm": 0.745911717414856,
"learning_rate": 1.5478603388263492e-06,
"loss": 0.1185,
"step": 1303
},
{
"epoch": 1.2735460235078615,
"grad_norm": 0.6497787833213806,
"learning_rate": 1.5442049043588391e-06,
"loss": 0.1165,
"step": 1304
},
{
"epoch": 1.274522973591818,
"grad_norm": 0.9276279211044312,
"learning_rate": 1.5405518615872578e-06,
"loss": 0.1344,
"step": 1305
},
{
"epoch": 1.2754999236757747,
"grad_norm": 0.47579535841941833,
"learning_rate": 1.5369012196526515e-06,
"loss": 0.1138,
"step": 1306
},
{
"epoch": 1.2764768737597314,
"grad_norm": 0.5858950018882751,
"learning_rate": 1.5332529876900598e-06,
"loss": 0.1175,
"step": 1307
},
{
"epoch": 1.2774538238436879,
"grad_norm": 0.4554455280303955,
"learning_rate": 1.5296071748284913e-06,
"loss": 0.1177,
"step": 1308
},
{
"epoch": 1.2784307739276446,
"grad_norm": 0.6461067795753479,
"learning_rate": 1.5259637901909007e-06,
"loss": 0.1277,
"step": 1309
},
{
"epoch": 1.2794077240116013,
"grad_norm": 0.624670684337616,
"learning_rate": 1.522322842894167e-06,
"loss": 0.1224,
"step": 1310
},
{
"epoch": 1.280384674095558,
"grad_norm": 0.48761656880378723,
"learning_rate": 1.5186843420490698e-06,
"loss": 0.1269,
"step": 1311
},
{
"epoch": 1.2813616241795147,
"grad_norm": 1.011336326599121,
"learning_rate": 1.5150482967602681e-06,
"loss": 0.1125,
"step": 1312
},
{
"epoch": 1.2823385742634712,
"grad_norm": 0.5898116230964661,
"learning_rate": 1.5114147161262743e-06,
"loss": 0.1449,
"step": 1313
},
{
"epoch": 1.283315524347428,
"grad_norm": 0.74689781665802,
"learning_rate": 1.5077836092394367e-06,
"loss": 0.1379,
"step": 1314
},
{
"epoch": 1.2842924744313846,
"grad_norm": 0.4546085298061371,
"learning_rate": 1.5041549851859098e-06,
"loss": 0.1099,
"step": 1315
},
{
"epoch": 1.285269424515341,
"grad_norm": 0.6719655394554138,
"learning_rate": 1.500528853045638e-06,
"loss": 0.1226,
"step": 1316
},
{
"epoch": 1.2862463745992978,
"grad_norm": 0.5607120394706726,
"learning_rate": 1.4969052218923285e-06,
"loss": 0.1189,
"step": 1317
},
{
"epoch": 1.2872233246832545,
"grad_norm": 0.5269936919212341,
"learning_rate": 1.4932841007934313e-06,
"loss": 0.1283,
"step": 1318
},
{
"epoch": 1.288200274767211,
"grad_norm": 0.5397964715957642,
"learning_rate": 1.4896654988101147e-06,
"loss": 0.1158,
"step": 1319
},
{
"epoch": 1.2891772248511677,
"grad_norm": 0.4894252419471741,
"learning_rate": 1.4860494249972438e-06,
"loss": 0.1171,
"step": 1320
},
{
"epoch": 1.2901541749351244,
"grad_norm": 0.5752339363098145,
"learning_rate": 1.4824358884033579e-06,
"loss": 0.1206,
"step": 1321
},
{
"epoch": 1.2911311250190811,
"grad_norm": 0.704617977142334,
"learning_rate": 1.4788248980706466e-06,
"loss": 0.1076,
"step": 1322
},
{
"epoch": 1.2921080751030378,
"grad_norm": 0.43897145986557007,
"learning_rate": 1.4752164630349287e-06,
"loss": 0.1235,
"step": 1323
},
{
"epoch": 1.2930850251869943,
"grad_norm": 0.5359245538711548,
"learning_rate": 1.471610592325628e-06,
"loss": 0.1227,
"step": 1324
},
{
"epoch": 1.294061975270951,
"grad_norm": 0.6712575554847717,
"learning_rate": 1.468007294965752e-06,
"loss": 0.1305,
"step": 1325
},
{
"epoch": 1.2950389253549077,
"grad_norm": 0.47084033489227295,
"learning_rate": 1.4644065799718688e-06,
"loss": 0.1104,
"step": 1326
},
{
"epoch": 1.2960158754388642,
"grad_norm": 0.5468409061431885,
"learning_rate": 1.4608084563540837e-06,
"loss": 0.1279,
"step": 1327
},
{
"epoch": 1.296992825522821,
"grad_norm": 0.5525038838386536,
"learning_rate": 1.4572129331160213e-06,
"loss": 0.1308,
"step": 1328
},
{
"epoch": 1.2979697756067776,
"grad_norm": 0.5236685872077942,
"learning_rate": 1.4536200192547933e-06,
"loss": 0.1158,
"step": 1329
},
{
"epoch": 1.298946725690734,
"grad_norm": 0.5852157473564148,
"learning_rate": 1.4500297237609867e-06,
"loss": 0.1181,
"step": 1330
},
{
"epoch": 1.2999236757746908,
"grad_norm": 0.5368656516075134,
"learning_rate": 1.446442055618636e-06,
"loss": 0.1236,
"step": 1331
},
{
"epoch": 1.3009006258586475,
"grad_norm": 0.7352957129478455,
"learning_rate": 1.4428570238051989e-06,
"loss": 0.1211,
"step": 1332
},
{
"epoch": 1.3018775759426042,
"grad_norm": 0.45952096581459045,
"learning_rate": 1.4392746372915368e-06,
"loss": 0.1217,
"step": 1333
},
{
"epoch": 1.302854526026561,
"grad_norm": 0.5984220504760742,
"learning_rate": 1.4356949050418944e-06,
"loss": 0.1122,
"step": 1334
},
{
"epoch": 1.3038314761105174,
"grad_norm": 0.6518990993499756,
"learning_rate": 1.4321178360138705e-06,
"loss": 0.1108,
"step": 1335
},
{
"epoch": 1.3048084261944741,
"grad_norm": 0.47587719559669495,
"learning_rate": 1.4285434391584035e-06,
"loss": 0.1115,
"step": 1336
},
{
"epoch": 1.3057853762784308,
"grad_norm": 0.9126313328742981,
"learning_rate": 1.4249717234197422e-06,
"loss": 0.1318,
"step": 1337
},
{
"epoch": 1.3067623263623873,
"grad_norm": 0.49654099345207214,
"learning_rate": 1.4214026977354283e-06,
"loss": 0.1194,
"step": 1338
},
{
"epoch": 1.307739276446344,
"grad_norm": 0.5105423927307129,
"learning_rate": 1.417836371036273e-06,
"loss": 0.118,
"step": 1339
},
{
"epoch": 1.3087162265303007,
"grad_norm": 1.9939988851547241,
"learning_rate": 1.414272752246331e-06,
"loss": 0.1822,
"step": 1340
},
{
"epoch": 1.3096931766142574,
"grad_norm": 0.9904623627662659,
"learning_rate": 1.4107118502828816e-06,
"loss": 0.1344,
"step": 1341
},
{
"epoch": 1.3106701266982141,
"grad_norm": 0.5835309624671936,
"learning_rate": 1.4071536740564079e-06,
"loss": 0.1161,
"step": 1342
},
{
"epoch": 1.3116470767821706,
"grad_norm": 0.602512001991272,
"learning_rate": 1.40359823247057e-06,
"loss": 0.122,
"step": 1343
},
{
"epoch": 1.3126240268661273,
"grad_norm": 0.7191610932350159,
"learning_rate": 1.4000455344221864e-06,
"loss": 0.1283,
"step": 1344
},
{
"epoch": 1.313600976950084,
"grad_norm": 0.6409730911254883,
"learning_rate": 1.396495588801211e-06,
"loss": 0.1197,
"step": 1345
},
{
"epoch": 1.3145779270340405,
"grad_norm": 0.4960377812385559,
"learning_rate": 1.3929484044907079e-06,
"loss": 0.1198,
"step": 1346
},
{
"epoch": 1.3155548771179972,
"grad_norm": 0.9895467758178711,
"learning_rate": 1.3894039903668346e-06,
"loss": 0.1375,
"step": 1347
},
{
"epoch": 1.316531827201954,
"grad_norm": 0.4482794404029846,
"learning_rate": 1.3858623552988142e-06,
"loss": 0.1106,
"step": 1348
},
{
"epoch": 1.3175087772859104,
"grad_norm": 0.9347596764564514,
"learning_rate": 1.3823235081489156e-06,
"loss": 0.1357,
"step": 1349
},
{
"epoch": 1.3184857273698671,
"grad_norm": 0.6971988677978516,
"learning_rate": 1.3787874577724347e-06,
"loss": 0.1243,
"step": 1350
},
{
"epoch": 1.3194626774538238,
"grad_norm": 1.050885558128357,
"learning_rate": 1.3752542130176649e-06,
"loss": 0.1276,
"step": 1351
},
{
"epoch": 1.3204396275377805,
"grad_norm": 0.47921139001846313,
"learning_rate": 1.3717237827258816e-06,
"loss": 0.1057,
"step": 1352
},
{
"epoch": 1.3214165776217373,
"grad_norm": 0.5367651581764221,
"learning_rate": 1.3681961757313184e-06,
"loss": 0.1293,
"step": 1353
},
{
"epoch": 1.3223935277056937,
"grad_norm": 0.682149350643158,
"learning_rate": 1.3646714008611405e-06,
"loss": 0.1186,
"step": 1354
},
{
"epoch": 1.3233704777896504,
"grad_norm": 0.4652373492717743,
"learning_rate": 1.3611494669354302e-06,
"loss": 0.1099,
"step": 1355
},
{
"epoch": 1.3243474278736072,
"grad_norm": 0.6074487566947937,
"learning_rate": 1.3576303827671584e-06,
"loss": 0.114,
"step": 1356
},
{
"epoch": 1.3253243779575636,
"grad_norm": 0.4402923583984375,
"learning_rate": 1.3541141571621648e-06,
"loss": 0.1155,
"step": 1357
},
{
"epoch": 1.3263013280415203,
"grad_norm": 0.526982307434082,
"learning_rate": 1.3506007989191372e-06,
"loss": 0.1148,
"step": 1358
},
{
"epoch": 1.327278278125477,
"grad_norm": 0.5465818643569946,
"learning_rate": 1.3470903168295904e-06,
"loss": 0.1095,
"step": 1359
},
{
"epoch": 1.3282552282094335,
"grad_norm": 0.5220271348953247,
"learning_rate": 1.3435827196778373e-06,
"loss": 0.129,
"step": 1360
},
{
"epoch": 1.3292321782933902,
"grad_norm": 0.5601693391799927,
"learning_rate": 1.340078016240976e-06,
"loss": 0.1252,
"step": 1361
},
{
"epoch": 1.330209128377347,
"grad_norm": 0.5428116321563721,
"learning_rate": 1.3365762152888612e-06,
"loss": 0.1205,
"step": 1362
},
{
"epoch": 1.3311860784613037,
"grad_norm": 0.4764959514141083,
"learning_rate": 1.3330773255840864e-06,
"loss": 0.1252,
"step": 1363
},
{
"epoch": 1.3321630285452604,
"grad_norm": 0.6474655866622925,
"learning_rate": 1.3295813558819593e-06,
"loss": 0.1231,
"step": 1364
},
{
"epoch": 1.3331399786292168,
"grad_norm": 0.5812271237373352,
"learning_rate": 1.3260883149304793e-06,
"loss": 0.1101,
"step": 1365
},
{
"epoch": 1.3341169287131736,
"grad_norm": 0.5133628845214844,
"learning_rate": 1.3225982114703201e-06,
"loss": 0.1134,
"step": 1366
},
{
"epoch": 1.3350938787971303,
"grad_norm": 0.873123824596405,
"learning_rate": 1.3191110542348048e-06,
"loss": 0.1274,
"step": 1367
},
{
"epoch": 1.3360708288810867,
"grad_norm": 1.164931297302246,
"learning_rate": 1.3156268519498806e-06,
"loss": 0.1327,
"step": 1368
},
{
"epoch": 1.3370477789650435,
"grad_norm": 0.5429942607879639,
"learning_rate": 1.3121456133341048e-06,
"loss": 0.1251,
"step": 1369
},
{
"epoch": 1.3380247290490002,
"grad_norm": 0.5467685461044312,
"learning_rate": 1.308667347098615e-06,
"loss": 0.1105,
"step": 1370
},
{
"epoch": 1.3390016791329569,
"grad_norm": 0.5272745490074158,
"learning_rate": 1.3051920619471145e-06,
"loss": 0.1145,
"step": 1371
},
{
"epoch": 1.3399786292169136,
"grad_norm": 0.613308846950531,
"learning_rate": 1.301719766575844e-06,
"loss": 0.1227,
"step": 1372
},
{
"epoch": 1.34095557930087,
"grad_norm": 1.4428318738937378,
"learning_rate": 1.2982504696735637e-06,
"loss": 0.1453,
"step": 1373
},
{
"epoch": 1.3419325293848268,
"grad_norm": 0.6538224220275879,
"learning_rate": 1.294784179921531e-06,
"loss": 0.1231,
"step": 1374
},
{
"epoch": 1.3429094794687835,
"grad_norm": 0.8577346205711365,
"learning_rate": 1.2913209059934807e-06,
"loss": 0.1442,
"step": 1375
},
{
"epoch": 1.34388642955274,
"grad_norm": 0.5622920393943787,
"learning_rate": 1.2878606565555956e-06,
"loss": 0.1224,
"step": 1376
},
{
"epoch": 1.3448633796366967,
"grad_norm": 0.5000760555267334,
"learning_rate": 1.2844034402664958e-06,
"loss": 0.1103,
"step": 1377
},
{
"epoch": 1.3458403297206534,
"grad_norm": 0.8320068717002869,
"learning_rate": 1.2809492657772105e-06,
"loss": 0.1265,
"step": 1378
},
{
"epoch": 1.3468172798046099,
"grad_norm": 0.6084581613540649,
"learning_rate": 1.2774981417311523e-06,
"loss": 0.109,
"step": 1379
},
{
"epoch": 1.3477942298885666,
"grad_norm": 0.6465567350387573,
"learning_rate": 1.2740500767641065e-06,
"loss": 0.1148,
"step": 1380
},
{
"epoch": 1.3487711799725233,
"grad_norm": 0.5642968416213989,
"learning_rate": 1.270605079504203e-06,
"loss": 0.1262,
"step": 1381
},
{
"epoch": 1.34974813005648,
"grad_norm": 0.4981468915939331,
"learning_rate": 1.2671631585718913e-06,
"loss": 0.1208,
"step": 1382
},
{
"epoch": 1.3507250801404367,
"grad_norm": 0.9176070690155029,
"learning_rate": 1.2637243225799283e-06,
"loss": 0.1257,
"step": 1383
},
{
"epoch": 1.3517020302243932,
"grad_norm": 0.6084444522857666,
"learning_rate": 1.2602885801333468e-06,
"loss": 0.1122,
"step": 1384
},
{
"epoch": 1.3526789803083499,
"grad_norm": 0.7930445671081543,
"learning_rate": 1.256855939829441e-06,
"loss": 0.1232,
"step": 1385
},
{
"epoch": 1.3536559303923066,
"grad_norm": 0.7945284843444824,
"learning_rate": 1.2534264102577445e-06,
"loss": 0.1129,
"step": 1386
},
{
"epoch": 1.354632880476263,
"grad_norm": 0.5566101670265198,
"learning_rate": 1.2500000000000007e-06,
"loss": 0.1221,
"step": 1387
},
{
"epoch": 1.3556098305602198,
"grad_norm": 0.6021673083305359,
"learning_rate": 1.246576717630153e-06,
"loss": 0.1363,
"step": 1388
},
{
"epoch": 1.3565867806441765,
"grad_norm": 0.6514391899108887,
"learning_rate": 1.243156571714317e-06,
"loss": 0.1238,
"step": 1389
},
{
"epoch": 1.3575637307281332,
"grad_norm": 0.7367511987686157,
"learning_rate": 1.2397395708107578e-06,
"loss": 0.1186,
"step": 1390
},
{
"epoch": 1.35854068081209,
"grad_norm": 0.45107749104499817,
"learning_rate": 1.2363257234698728e-06,
"loss": 0.1207,
"step": 1391
},
{
"epoch": 1.3595176308960464,
"grad_norm": 0.6295725703239441,
"learning_rate": 1.2329150382341682e-06,
"loss": 0.125,
"step": 1392
},
{
"epoch": 1.360494580980003,
"grad_norm": 0.6059122085571289,
"learning_rate": 1.2295075236382348e-06,
"loss": 0.1345,
"step": 1393
},
{
"epoch": 1.3614715310639598,
"grad_norm": 0.9611726403236389,
"learning_rate": 1.2261031882087334e-06,
"loss": 0.1326,
"step": 1394
},
{
"epoch": 1.3624484811479163,
"grad_norm": 0.6361368298530579,
"learning_rate": 1.2227020404643671e-06,
"loss": 0.1306,
"step": 1395
},
{
"epoch": 1.363425431231873,
"grad_norm": 0.915302038192749,
"learning_rate": 1.2193040889158621e-06,
"loss": 0.1214,
"step": 1396
},
{
"epoch": 1.3644023813158297,
"grad_norm": 0.5358700156211853,
"learning_rate": 1.215909342065949e-06,
"loss": 0.1168,
"step": 1397
},
{
"epoch": 1.3653793313997862,
"grad_norm": 1.0897305011749268,
"learning_rate": 1.2125178084093363e-06,
"loss": 0.1351,
"step": 1398
},
{
"epoch": 1.3663562814837429,
"grad_norm": 0.7376011610031128,
"learning_rate": 1.2091294964326948e-06,
"loss": 0.12,
"step": 1399
},
{
"epoch": 1.3673332315676996,
"grad_norm": 0.7292815446853638,
"learning_rate": 1.205744414614633e-06,
"loss": 0.123,
"step": 1400
},
{
"epoch": 1.3683101816516563,
"grad_norm": 0.64980149269104,
"learning_rate": 1.2023625714256743e-06,
"loss": 0.1264,
"step": 1401
},
{
"epoch": 1.369287131735613,
"grad_norm": 0.686808168888092,
"learning_rate": 1.198983975328242e-06,
"loss": 0.1144,
"step": 1402
},
{
"epoch": 1.3702640818195695,
"grad_norm": 1.0751445293426514,
"learning_rate": 1.1956086347766306e-06,
"loss": 0.1471,
"step": 1403
},
{
"epoch": 1.3712410319035262,
"grad_norm": 1.1832984685897827,
"learning_rate": 1.192236558216989e-06,
"loss": 0.1328,
"step": 1404
},
{
"epoch": 1.372217981987483,
"grad_norm": 0.5105385780334473,
"learning_rate": 1.1888677540873008e-06,
"loss": 0.1248,
"step": 1405
},
{
"epoch": 1.3731949320714394,
"grad_norm": 0.5958878397941589,
"learning_rate": 1.1855022308173575e-06,
"loss": 0.1215,
"step": 1406
},
{
"epoch": 1.374171882155396,
"grad_norm": 0.4924409091472626,
"learning_rate": 1.1821399968287437e-06,
"loss": 0.1101,
"step": 1407
},
{
"epoch": 1.3751488322393528,
"grad_norm": 0.8076400756835938,
"learning_rate": 1.1787810605348132e-06,
"loss": 0.136,
"step": 1408
},
{
"epoch": 1.3761257823233093,
"grad_norm": 0.5968782305717468,
"learning_rate": 1.175425430340666e-06,
"loss": 0.1424,
"step": 1409
},
{
"epoch": 1.377102732407266,
"grad_norm": 0.5203692317008972,
"learning_rate": 1.1720731146431289e-06,
"loss": 0.1186,
"step": 1410
},
{
"epoch": 1.3780796824912227,
"grad_norm": 0.7054222822189331,
"learning_rate": 1.1687241218307382e-06,
"loss": 0.1332,
"step": 1411
},
{
"epoch": 1.3790566325751794,
"grad_norm": 0.5586547255516052,
"learning_rate": 1.1653784602837115e-06,
"loss": 0.1269,
"step": 1412
},
{
"epoch": 1.3800335826591361,
"grad_norm": 0.9022789001464844,
"learning_rate": 1.162036138373933e-06,
"loss": 0.1225,
"step": 1413
},
{
"epoch": 1.3810105327430926,
"grad_norm": 0.5225563645362854,
"learning_rate": 1.1586971644649297e-06,
"loss": 0.1265,
"step": 1414
},
{
"epoch": 1.3819874828270493,
"grad_norm": 1.30247163772583,
"learning_rate": 1.1553615469118492e-06,
"loss": 0.1232,
"step": 1415
},
{
"epoch": 1.382964432911006,
"grad_norm": 0.4558253288269043,
"learning_rate": 1.1520292940614432e-06,
"loss": 0.1138,
"step": 1416
},
{
"epoch": 1.3839413829949625,
"grad_norm": 0.46534863114356995,
"learning_rate": 1.1487004142520422e-06,
"loss": 0.1269,
"step": 1417
},
{
"epoch": 1.3849183330789192,
"grad_norm": 0.5262455940246582,
"learning_rate": 1.1453749158135347e-06,
"loss": 0.1483,
"step": 1418
},
{
"epoch": 1.385895283162876,
"grad_norm": 0.5273460149765015,
"learning_rate": 1.142052807067352e-06,
"loss": 0.1345,
"step": 1419
},
{
"epoch": 1.3868722332468326,
"grad_norm": 0.530284583568573,
"learning_rate": 1.1387340963264393e-06,
"loss": 0.1269,
"step": 1420
},
{
"epoch": 1.3878491833307893,
"grad_norm": 2.742753267288208,
"learning_rate": 1.135418791895241e-06,
"loss": 0.1764,
"step": 1421
},
{
"epoch": 1.3888261334147458,
"grad_norm": 0.7036141157150269,
"learning_rate": 1.1321069020696794e-06,
"loss": 0.1119,
"step": 1422
},
{
"epoch": 1.3898030834987025,
"grad_norm": 1.1463329792022705,
"learning_rate": 1.1287984351371287e-06,
"loss": 0.1327,
"step": 1423
},
{
"epoch": 1.3907800335826592,
"grad_norm": 0.5199117660522461,
"learning_rate": 1.1254933993764008e-06,
"loss": 0.121,
"step": 1424
},
{
"epoch": 1.3917569836666157,
"grad_norm": 1.0432943105697632,
"learning_rate": 1.122191803057721e-06,
"loss": 0.116,
"step": 1425
},
{
"epoch": 1.3927339337505724,
"grad_norm": 0.47711002826690674,
"learning_rate": 1.1188936544427056e-06,
"loss": 0.1206,
"step": 1426
},
{
"epoch": 1.3937108838345291,
"grad_norm": 0.9329760670661926,
"learning_rate": 1.115598961784348e-06,
"loss": 0.1157,
"step": 1427
},
{
"epoch": 1.3946878339184856,
"grad_norm": 0.7033591866493225,
"learning_rate": 1.1123077333269921e-06,
"loss": 0.1122,
"step": 1428
},
{
"epoch": 1.3956647840024423,
"grad_norm": 0.47941333055496216,
"learning_rate": 1.1090199773063107e-06,
"loss": 0.11,
"step": 1429
},
{
"epoch": 1.396641734086399,
"grad_norm": 0.5806613564491272,
"learning_rate": 1.1057357019492921e-06,
"loss": 0.1342,
"step": 1430
},
{
"epoch": 1.3976186841703557,
"grad_norm": 0.5927656888961792,
"learning_rate": 1.1024549154742098e-06,
"loss": 0.1175,
"step": 1431
},
{
"epoch": 1.3985956342543124,
"grad_norm": 0.5375458598136902,
"learning_rate": 1.0991776260906113e-06,
"loss": 0.1128,
"step": 1432
},
{
"epoch": 1.399572584338269,
"grad_norm": 0.5560141801834106,
"learning_rate": 1.0959038419992929e-06,
"loss": 0.14,
"step": 1433
},
{
"epoch": 1.4005495344222256,
"grad_norm": 0.517528772354126,
"learning_rate": 1.092633571392275e-06,
"loss": 0.1373,
"step": 1434
},
{
"epoch": 1.4015264845061823,
"grad_norm": 1.1873289346694946,
"learning_rate": 1.0893668224527912e-06,
"loss": 0.1419,
"step": 1435
},
{
"epoch": 1.4025034345901388,
"grad_norm": 0.6347547769546509,
"learning_rate": 1.0861036033552618e-06,
"loss": 0.1268,
"step": 1436
},
{
"epoch": 1.4034803846740955,
"grad_norm": 0.5334620475769043,
"learning_rate": 1.0828439222652722e-06,
"loss": 0.1156,
"step": 1437
},
{
"epoch": 1.4044573347580522,
"grad_norm": 0.48120439052581787,
"learning_rate": 1.0795877873395578e-06,
"loss": 0.1231,
"step": 1438
},
{
"epoch": 1.4054342848420087,
"grad_norm": 1.4684795141220093,
"learning_rate": 1.076335206725976e-06,
"loss": 0.1408,
"step": 1439
},
{
"epoch": 1.4064112349259654,
"grad_norm": 0.4773467481136322,
"learning_rate": 1.0730861885634946e-06,
"loss": 0.1186,
"step": 1440
},
{
"epoch": 1.4073881850099221,
"grad_norm": 0.609987199306488,
"learning_rate": 1.0698407409821661e-06,
"loss": 0.1305,
"step": 1441
},
{
"epoch": 1.4083651350938788,
"grad_norm": 0.7426857948303223,
"learning_rate": 1.0665988721031062e-06,
"loss": 0.1358,
"step": 1442
},
{
"epoch": 1.4093420851778355,
"grad_norm": 1.0063998699188232,
"learning_rate": 1.0633605900384763e-06,
"loss": 0.1261,
"step": 1443
},
{
"epoch": 1.410319035261792,
"grad_norm": 0.5790124535560608,
"learning_rate": 1.0601259028914647e-06,
"loss": 0.1219,
"step": 1444
},
{
"epoch": 1.4112959853457487,
"grad_norm": 0.9051598310470581,
"learning_rate": 1.05689481875626e-06,
"loss": 0.1333,
"step": 1445
},
{
"epoch": 1.4122729354297054,
"grad_norm": 0.626784086227417,
"learning_rate": 1.0536673457180393e-06,
"loss": 0.1137,
"step": 1446
},
{
"epoch": 1.413249885513662,
"grad_norm": 1.024578332901001,
"learning_rate": 1.050443491852942e-06,
"loss": 0.1253,
"step": 1447
},
{
"epoch": 1.4142268355976186,
"grad_norm": 0.677089512348175,
"learning_rate": 1.04722326522805e-06,
"loss": 0.1286,
"step": 1448
},
{
"epoch": 1.4152037856815753,
"grad_norm": 0.947273313999176,
"learning_rate": 1.044006673901369e-06,
"loss": 0.1312,
"step": 1449
},
{
"epoch": 1.416180735765532,
"grad_norm": 0.5041372179985046,
"learning_rate": 1.0407937259218101e-06,
"loss": 0.1192,
"step": 1450
},
{
"epoch": 1.4171576858494888,
"grad_norm": 0.7648371458053589,
"learning_rate": 1.0375844293291646e-06,
"loss": 0.1394,
"step": 1451
},
{
"epoch": 1.4181346359334452,
"grad_norm": 0.5033823847770691,
"learning_rate": 1.03437879215409e-06,
"loss": 0.1196,
"step": 1452
},
{
"epoch": 1.419111586017402,
"grad_norm": 0.5057837963104248,
"learning_rate": 1.031176822418084e-06,
"loss": 0.1175,
"step": 1453
},
{
"epoch": 1.4200885361013587,
"grad_norm": 0.5633666515350342,
"learning_rate": 1.0279785281334693e-06,
"loss": 0.1348,
"step": 1454
},
{
"epoch": 1.4210654861853151,
"grad_norm": 0.5748161673545837,
"learning_rate": 1.0247839173033707e-06,
"loss": 0.1176,
"step": 1455
},
{
"epoch": 1.4220424362692718,
"grad_norm": 0.566787838935852,
"learning_rate": 1.0215929979216961e-06,
"loss": 0.1175,
"step": 1456
},
{
"epoch": 1.4230193863532286,
"grad_norm": 0.49881404638290405,
"learning_rate": 1.0184057779731146e-06,
"loss": 0.1262,
"step": 1457
},
{
"epoch": 1.423996336437185,
"grad_norm": 0.4397923946380615,
"learning_rate": 1.0152222654330412e-06,
"loss": 0.1171,
"step": 1458
},
{
"epoch": 1.4249732865211417,
"grad_norm": 0.6777058243751526,
"learning_rate": 1.0120424682676106e-06,
"loss": 0.1348,
"step": 1459
},
{
"epoch": 1.4259502366050985,
"grad_norm": 0.8062659502029419,
"learning_rate": 1.008866394433663e-06,
"loss": 0.1348,
"step": 1460
},
{
"epoch": 1.4269271866890552,
"grad_norm": 0.7809616923332214,
"learning_rate": 1.0056940518787212e-06,
"loss": 0.1176,
"step": 1461
},
{
"epoch": 1.4279041367730119,
"grad_norm": 0.5211665630340576,
"learning_rate": 1.0025254485409694e-06,
"loss": 0.1038,
"step": 1462
},
{
"epoch": 1.4288810868569684,
"grad_norm": 0.5614594221115112,
"learning_rate": 9.993605923492381e-07,
"loss": 0.1336,
"step": 1463
},
{
"epoch": 1.429858036940925,
"grad_norm": 0.7447706460952759,
"learning_rate": 9.961994912229785e-07,
"loss": 0.1414,
"step": 1464
},
{
"epoch": 1.4308349870248818,
"grad_norm": 0.6395076513290405,
"learning_rate": 9.930421530722462e-07,
"loss": 0.1252,
"step": 1465
},
{
"epoch": 1.4318119371088383,
"grad_norm": 0.49954870343208313,
"learning_rate": 9.89888585797682e-07,
"loss": 0.1158,
"step": 1466
},
{
"epoch": 1.432788887192795,
"grad_norm": 0.9796427488327026,
"learning_rate": 9.867387972904885e-07,
"loss": 0.126,
"step": 1467
},
{
"epoch": 1.4337658372767517,
"grad_norm": 0.533560037612915,
"learning_rate": 9.83592795432415e-07,
"loss": 0.1243,
"step": 1468
},
{
"epoch": 1.4347427873607084,
"grad_norm": 0.6036112904548645,
"learning_rate": 9.804505880957351e-07,
"loss": 0.1175,
"step": 1469
},
{
"epoch": 1.4357197374446649,
"grad_norm": 0.663607656955719,
"learning_rate": 9.77312183143225e-07,
"loss": 0.1145,
"step": 1470
},
{
"epoch": 1.4366966875286216,
"grad_norm": 0.7179862260818481,
"learning_rate": 9.741775884281493e-07,
"loss": 0.1203,
"step": 1471
},
{
"epoch": 1.4376736376125783,
"grad_norm": 0.5878673195838928,
"learning_rate": 9.710468117942359e-07,
"loss": 0.1311,
"step": 1472
},
{
"epoch": 1.438650587696535,
"grad_norm": 0.5466660857200623,
"learning_rate": 9.679198610756587e-07,
"loss": 0.1244,
"step": 1473
},
{
"epoch": 1.4396275377804915,
"grad_norm": 0.6389520764350891,
"learning_rate": 9.647967440970207e-07,
"loss": 0.116,
"step": 1474
},
{
"epoch": 1.4406044878644482,
"grad_norm": 0.6984757781028748,
"learning_rate": 9.616774686733275e-07,
"loss": 0.1151,
"step": 1475
},
{
"epoch": 1.4415814379484049,
"grad_norm": 0.9478892683982849,
"learning_rate": 9.585620426099753e-07,
"loss": 0.1283,
"step": 1476
},
{
"epoch": 1.4425583880323614,
"grad_norm": 0.7938877940177917,
"learning_rate": 9.554504737027278e-07,
"loss": 0.1094,
"step": 1477
},
{
"epoch": 1.443535338116318,
"grad_norm": 0.7785425782203674,
"learning_rate": 9.523427697376941e-07,
"loss": 0.128,
"step": 1478
},
{
"epoch": 1.4445122882002748,
"grad_norm": 0.5309463739395142,
"learning_rate": 9.492389384913159e-07,
"loss": 0.1303,
"step": 1479
},
{
"epoch": 1.4454892382842315,
"grad_norm": 0.8360875248908997,
"learning_rate": 9.461389877303409e-07,
"loss": 0.1194,
"step": 1480
},
{
"epoch": 1.4464661883681882,
"grad_norm": 0.5747968554496765,
"learning_rate": 9.430429252118073e-07,
"loss": 0.1141,
"step": 1481
},
{
"epoch": 1.4474431384521447,
"grad_norm": 0.5618856549263,
"learning_rate": 9.39950758683025e-07,
"loss": 0.1292,
"step": 1482
},
{
"epoch": 1.4484200885361014,
"grad_norm": 0.7081732153892517,
"learning_rate": 9.368624958815553e-07,
"loss": 0.1121,
"step": 1483
},
{
"epoch": 1.449397038620058,
"grad_norm": 0.7016134858131409,
"learning_rate": 9.337781445351879e-07,
"loss": 0.1356,
"step": 1484
},
{
"epoch": 1.4503739887040146,
"grad_norm": 0.7247880101203918,
"learning_rate": 9.306977123619293e-07,
"loss": 0.1243,
"step": 1485
},
{
"epoch": 1.4513509387879713,
"grad_norm": 0.46754956245422363,
"learning_rate": 9.276212070699744e-07,
"loss": 0.1235,
"step": 1486
},
{
"epoch": 1.452327888871928,
"grad_norm": 0.4951886236667633,
"learning_rate": 9.245486363576961e-07,
"loss": 0.113,
"step": 1487
},
{
"epoch": 1.4533048389558845,
"grad_norm": 0.4841793179512024,
"learning_rate": 9.21480007913619e-07,
"loss": 0.1168,
"step": 1488
},
{
"epoch": 1.4542817890398412,
"grad_norm": 0.6489046216011047,
"learning_rate": 9.184153294164031e-07,
"loss": 0.1336,
"step": 1489
},
{
"epoch": 1.4552587391237979,
"grad_norm": 0.5661309361457825,
"learning_rate": 9.153546085348253e-07,
"loss": 0.1092,
"step": 1490
},
{
"epoch": 1.4562356892077546,
"grad_norm": 0.5326016545295715,
"learning_rate": 9.122978529277601e-07,
"loss": 0.109,
"step": 1491
},
{
"epoch": 1.4572126392917113,
"grad_norm": 0.46728262305259705,
"learning_rate": 9.092450702441572e-07,
"loss": 0.1119,
"step": 1492
},
{
"epoch": 1.4581895893756678,
"grad_norm": 0.4788103997707367,
"learning_rate": 9.061962681230266e-07,
"loss": 0.1125,
"step": 1493
},
{
"epoch": 1.4591665394596245,
"grad_norm": 0.5560827255249023,
"learning_rate": 9.031514541934186e-07,
"loss": 0.1203,
"step": 1494
},
{
"epoch": 1.4601434895435812,
"grad_norm": 0.6202582716941833,
"learning_rate": 9.001106360744014e-07,
"loss": 0.125,
"step": 1495
},
{
"epoch": 1.4611204396275377,
"grad_norm": 0.7125849723815918,
"learning_rate": 8.970738213750447e-07,
"loss": 0.1262,
"step": 1496
},
{
"epoch": 1.4620973897114944,
"grad_norm": 0.7084099650382996,
"learning_rate": 8.940410176944029e-07,
"loss": 0.1429,
"step": 1497
},
{
"epoch": 1.463074339795451,
"grad_norm": 0.4294019341468811,
"learning_rate": 8.910122326214903e-07,
"loss": 0.1181,
"step": 1498
},
{
"epoch": 1.4640512898794078,
"grad_norm": 0.8556720614433289,
"learning_rate": 8.87987473735269e-07,
"loss": 0.1254,
"step": 1499
},
{
"epoch": 1.4650282399633645,
"grad_norm": 0.5792166590690613,
"learning_rate": 8.849667486046224e-07,
"loss": 0.1145,
"step": 1500
},
{
"epoch": 1.466005190047321,
"grad_norm": 0.5899726748466492,
"learning_rate": 8.819500647883428e-07,
"loss": 0.114,
"step": 1501
},
{
"epoch": 1.4669821401312777,
"grad_norm": 1.1968765258789062,
"learning_rate": 8.789374298351105e-07,
"loss": 0.1257,
"step": 1502
},
{
"epoch": 1.4679590902152344,
"grad_norm": 0.7331739664077759,
"learning_rate": 8.759288512834721e-07,
"loss": 0.1106,
"step": 1503
},
{
"epoch": 1.468936040299191,
"grad_norm": 0.5271918177604675,
"learning_rate": 8.729243366618237e-07,
"loss": 0.1351,
"step": 1504
},
{
"epoch": 1.4699129903831476,
"grad_norm": 0.9830980896949768,
"learning_rate": 8.699238934883949e-07,
"loss": 0.125,
"step": 1505
},
{
"epoch": 1.4708899404671043,
"grad_norm": 0.495226114988327,
"learning_rate": 8.669275292712242e-07,
"loss": 0.1172,
"step": 1506
},
{
"epoch": 1.4718668905510608,
"grad_norm": 0.5556236505508423,
"learning_rate": 8.639352515081465e-07,
"loss": 0.1268,
"step": 1507
},
{
"epoch": 1.4728438406350175,
"grad_norm": 1.2183847427368164,
"learning_rate": 8.60947067686767e-07,
"loss": 0.1322,
"step": 1508
},
{
"epoch": 1.4738207907189742,
"grad_norm": 0.6959921717643738,
"learning_rate": 8.579629852844501e-07,
"loss": 0.1238,
"step": 1509
},
{
"epoch": 1.474797740802931,
"grad_norm": 0.5215861201286316,
"learning_rate": 8.54983011768297e-07,
"loss": 0.1141,
"step": 1510
},
{
"epoch": 1.4757746908868876,
"grad_norm": 0.4962870180606842,
"learning_rate": 8.520071545951253e-07,
"loss": 0.111,
"step": 1511
},
{
"epoch": 1.476751640970844,
"grad_norm": 0.5615082383155823,
"learning_rate": 8.490354212114518e-07,
"loss": 0.1285,
"step": 1512
},
{
"epoch": 1.4777285910548008,
"grad_norm": 1.5830167531967163,
"learning_rate": 8.460678190534779e-07,
"loss": 0.1113,
"step": 1513
},
{
"epoch": 1.4787055411387575,
"grad_norm": 0.6755221486091614,
"learning_rate": 8.43104355547063e-07,
"loss": 0.1222,
"step": 1514
},
{
"epoch": 1.479682491222714,
"grad_norm": 0.6732090711593628,
"learning_rate": 8.401450381077134e-07,
"loss": 0.1208,
"step": 1515
},
{
"epoch": 1.4806594413066707,
"grad_norm": 0.8399896621704102,
"learning_rate": 8.371898741405604e-07,
"loss": 0.1206,
"step": 1516
},
{
"epoch": 1.4816363913906274,
"grad_norm": 0.44740211963653564,
"learning_rate": 8.342388710403398e-07,
"loss": 0.1087,
"step": 1517
},
{
"epoch": 1.482613341474584,
"grad_norm": 0.636900782585144,
"learning_rate": 8.312920361913787e-07,
"loss": 0.1285,
"step": 1518
},
{
"epoch": 1.4835902915585406,
"grad_norm": 0.808541476726532,
"learning_rate": 8.283493769675716e-07,
"loss": 0.1233,
"step": 1519
},
{
"epoch": 1.4845672416424973,
"grad_norm": 0.4924761950969696,
"learning_rate": 8.254109007323638e-07,
"loss": 0.1242,
"step": 1520
},
{
"epoch": 1.485544191726454,
"grad_norm": 0.5257745981216431,
"learning_rate": 8.224766148387369e-07,
"loss": 0.1325,
"step": 1521
},
{
"epoch": 1.4865211418104107,
"grad_norm": 0.8435088396072388,
"learning_rate": 8.195465266291827e-07,
"loss": 0.1261,
"step": 1522
},
{
"epoch": 1.4874980918943672,
"grad_norm": 0.7535191774368286,
"learning_rate": 8.166206434356921e-07,
"loss": 0.1235,
"step": 1523
},
{
"epoch": 1.488475041978324,
"grad_norm": 0.8404972553253174,
"learning_rate": 8.136989725797337e-07,
"loss": 0.1192,
"step": 1524
},
{
"epoch": 1.4894519920622806,
"grad_norm": 0.6082322001457214,
"learning_rate": 8.107815213722331e-07,
"loss": 0.1319,
"step": 1525
},
{
"epoch": 1.4904289421462371,
"grad_norm": 0.5912935137748718,
"learning_rate": 8.078682971135598e-07,
"loss": 0.1413,
"step": 1526
},
{
"epoch": 1.4914058922301938,
"grad_norm": 0.6042731404304504,
"learning_rate": 8.049593070935044e-07,
"loss": 0.1243,
"step": 1527
},
{
"epoch": 1.4923828423141505,
"grad_norm": 0.7142298221588135,
"learning_rate": 8.020545585912617e-07,
"loss": 0.1277,
"step": 1528
},
{
"epoch": 1.4933597923981072,
"grad_norm": 0.8495448231697083,
"learning_rate": 7.991540588754149e-07,
"loss": 0.1253,
"step": 1529
},
{
"epoch": 1.494336742482064,
"grad_norm": 0.7489160299301147,
"learning_rate": 7.962578152039149e-07,
"loss": 0.1108,
"step": 1530
},
{
"epoch": 1.4953136925660204,
"grad_norm": 0.6093230247497559,
"learning_rate": 7.933658348240605e-07,
"loss": 0.1171,
"step": 1531
},
{
"epoch": 1.4962906426499771,
"grad_norm": 0.5270717740058899,
"learning_rate": 7.904781249724858e-07,
"loss": 0.1217,
"step": 1532
},
{
"epoch": 1.4972675927339338,
"grad_norm": 1.3727728128433228,
"learning_rate": 7.875946928751355e-07,
"loss": 0.1316,
"step": 1533
},
{
"epoch": 1.4982445428178903,
"grad_norm": 0.9571304321289062,
"learning_rate": 7.847155457472527e-07,
"loss": 0.1267,
"step": 1534
},
{
"epoch": 1.499221492901847,
"grad_norm": 0.6771973967552185,
"learning_rate": 7.818406907933562e-07,
"loss": 0.1252,
"step": 1535
},
{
"epoch": 1.5001984429858037,
"grad_norm": 0.6739697456359863,
"learning_rate": 7.789701352072249e-07,
"loss": 0.1301,
"step": 1536
},
{
"epoch": 1.5011753930697602,
"grad_norm": 0.6448550224304199,
"learning_rate": 7.761038861718798e-07,
"loss": 0.1231,
"step": 1537
},
{
"epoch": 1.5021523431537172,
"grad_norm": 0.600214958190918,
"learning_rate": 7.732419508595668e-07,
"loss": 0.1329,
"step": 1538
},
{
"epoch": 1.5031292932376736,
"grad_norm": 0.5520655512809753,
"learning_rate": 7.70384336431734e-07,
"loss": 0.123,
"step": 1539
},
{
"epoch": 1.5041062433216301,
"grad_norm": 0.584899365901947,
"learning_rate": 7.675310500390215e-07,
"loss": 0.1339,
"step": 1540
},
{
"epoch": 1.505083193405587,
"grad_norm": 0.6146745085716248,
"learning_rate": 7.646820988212364e-07,
"loss": 0.1166,
"step": 1541
},
{
"epoch": 1.5060601434895435,
"grad_norm": 1.3191120624542236,
"learning_rate": 7.61837489907338e-07,
"loss": 0.1248,
"step": 1542
},
{
"epoch": 1.5070370935735002,
"grad_norm": 0.6798263788223267,
"learning_rate": 7.58997230415422e-07,
"loss": 0.1247,
"step": 1543
},
{
"epoch": 1.508014043657457,
"grad_norm": 0.6484711766242981,
"learning_rate": 7.561613274526971e-07,
"loss": 0.1233,
"step": 1544
},
{
"epoch": 1.5089909937414134,
"grad_norm": 0.5282912254333496,
"learning_rate": 7.533297881154739e-07,
"loss": 0.1133,
"step": 1545
},
{
"epoch": 1.5099679438253701,
"grad_norm": 0.5817572474479675,
"learning_rate": 7.505026194891432e-07,
"loss": 0.1279,
"step": 1546
},
{
"epoch": 1.5109448939093268,
"grad_norm": 0.443130224943161,
"learning_rate": 7.476798286481562e-07,
"loss": 0.1243,
"step": 1547
},
{
"epoch": 1.5119218439932833,
"grad_norm": 0.49105507135391235,
"learning_rate": 7.448614226560127e-07,
"loss": 0.1152,
"step": 1548
},
{
"epoch": 1.5128987940772403,
"grad_norm": 0.42153459787368774,
"learning_rate": 7.420474085652409e-07,
"loss": 0.1163,
"step": 1549
},
{
"epoch": 1.5138757441611967,
"grad_norm": 0.7881332039833069,
"learning_rate": 7.392377934173736e-07,
"loss": 0.1248,
"step": 1550
},
{
"epoch": 1.5148526942451535,
"grad_norm": 0.6543744802474976,
"learning_rate": 7.364325842429417e-07,
"loss": 0.1261,
"step": 1551
},
{
"epoch": 1.5158296443291102,
"grad_norm": 0.4835239052772522,
"learning_rate": 7.336317880614497e-07,
"loss": 0.1248,
"step": 1552
},
{
"epoch": 1.5168065944130666,
"grad_norm": 0.6503763794898987,
"learning_rate": 7.308354118813571e-07,
"loss": 0.1211,
"step": 1553
},
{
"epoch": 1.5177835444970234,
"grad_norm": 0.7182596921920776,
"learning_rate": 7.280434627000669e-07,
"loss": 0.1368,
"step": 1554
},
{
"epoch": 1.51876049458098,
"grad_norm": 0.8125506043434143,
"learning_rate": 7.252559475039004e-07,
"loss": 0.1294,
"step": 1555
},
{
"epoch": 1.5197374446649365,
"grad_norm": 0.6878829002380371,
"learning_rate": 7.224728732680861e-07,
"loss": 0.1257,
"step": 1556
},
{
"epoch": 1.5207143947488933,
"grad_norm": 0.6152113080024719,
"learning_rate": 7.196942469567417e-07,
"loss": 0.1246,
"step": 1557
},
{
"epoch": 1.52169134483285,
"grad_norm": 0.5264148712158203,
"learning_rate": 7.169200755228486e-07,
"loss": 0.1154,
"step": 1558
},
{
"epoch": 1.5226682949168064,
"grad_norm": 1.520292043685913,
"learning_rate": 7.141503659082469e-07,
"loss": 0.1254,
"step": 1559
},
{
"epoch": 1.5236452450007634,
"grad_norm": 0.6638200879096985,
"learning_rate": 7.113851250436099e-07,
"loss": 0.1155,
"step": 1560
},
{
"epoch": 1.5246221950847199,
"grad_norm": 0.4376411437988281,
"learning_rate": 7.086243598484266e-07,
"loss": 0.1217,
"step": 1561
},
{
"epoch": 1.5255991451686766,
"grad_norm": 0.4782516360282898,
"learning_rate": 7.058680772309895e-07,
"loss": 0.1211,
"step": 1562
},
{
"epoch": 1.5265760952526333,
"grad_norm": 0.5473380088806152,
"learning_rate": 7.031162840883735e-07,
"loss": 0.1194,
"step": 1563
},
{
"epoch": 1.5275530453365898,
"grad_norm": 0.6324317455291748,
"learning_rate": 7.003689873064173e-07,
"loss": 0.1169,
"step": 1564
},
{
"epoch": 1.5285299954205465,
"grad_norm": 0.5594613552093506,
"learning_rate": 6.97626193759712e-07,
"loss": 0.1171,
"step": 1565
},
{
"epoch": 1.5295069455045032,
"grad_norm": 0.5535578727722168,
"learning_rate": 6.948879103115769e-07,
"loss": 0.1213,
"step": 1566
},
{
"epoch": 1.5304838955884597,
"grad_norm": 0.6731473207473755,
"learning_rate": 6.921541438140464e-07,
"loss": 0.1102,
"step": 1567
},
{
"epoch": 1.5314608456724166,
"grad_norm": 1.485221266746521,
"learning_rate": 6.894249011078543e-07,
"loss": 0.116,
"step": 1568
},
{
"epoch": 1.532437795756373,
"grad_norm": 0.7399523854255676,
"learning_rate": 6.867001890224109e-07,
"loss": 0.1374,
"step": 1569
},
{
"epoch": 1.5334147458403298,
"grad_norm": 0.5747402310371399,
"learning_rate": 6.839800143757924e-07,
"loss": 0.1336,
"step": 1570
},
{
"epoch": 1.5343916959242865,
"grad_norm": 0.6139041185379028,
"learning_rate": 6.812643839747204e-07,
"loss": 0.1454,
"step": 1571
},
{
"epoch": 1.535368646008243,
"grad_norm": 0.6948770880699158,
"learning_rate": 6.785533046145445e-07,
"loss": 0.1217,
"step": 1572
},
{
"epoch": 1.5363455960921997,
"grad_norm": 0.6427881121635437,
"learning_rate": 6.758467830792254e-07,
"loss": 0.1482,
"step": 1573
},
{
"epoch": 1.5373225461761564,
"grad_norm": 0.5660995841026306,
"learning_rate": 6.731448261413215e-07,
"loss": 0.1121,
"step": 1574
},
{
"epoch": 1.5382994962601129,
"grad_norm": 0.8488437533378601,
"learning_rate": 6.70447440561966e-07,
"loss": 0.1192,
"step": 1575
},
{
"epoch": 1.5392764463440696,
"grad_norm": 0.47182774543762207,
"learning_rate": 6.677546330908563e-07,
"loss": 0.1115,
"step": 1576
},
{
"epoch": 1.5402533964280263,
"grad_norm": 1.0015591382980347,
"learning_rate": 6.650664104662308e-07,
"loss": 0.1208,
"step": 1577
},
{
"epoch": 1.5412303465119828,
"grad_norm": 0.9412240386009216,
"learning_rate": 6.623827794148572e-07,
"loss": 0.1314,
"step": 1578
},
{
"epoch": 1.5422072965959397,
"grad_norm": 0.46412599086761475,
"learning_rate": 6.597037466520139e-07,
"loss": 0.1124,
"step": 1579
},
{
"epoch": 1.5431842466798962,
"grad_norm": 0.4531099200248718,
"learning_rate": 6.570293188814722e-07,
"loss": 0.1218,
"step": 1580
},
{
"epoch": 1.5441611967638529,
"grad_norm": 0.6451519727706909,
"learning_rate": 6.543595027954786e-07,
"loss": 0.1179,
"step": 1581
},
{
"epoch": 1.5451381468478096,
"grad_norm": 0.7565576434135437,
"learning_rate": 6.516943050747435e-07,
"loss": 0.1161,
"step": 1582
},
{
"epoch": 1.546115096931766,
"grad_norm": 0.48450151085853577,
"learning_rate": 6.490337323884171e-07,
"loss": 0.1181,
"step": 1583
},
{
"epoch": 1.5470920470157228,
"grad_norm": 0.6125978827476501,
"learning_rate": 6.463777913940783e-07,
"loss": 0.1109,
"step": 1584
},
{
"epoch": 1.5480689970996795,
"grad_norm": 0.611227810382843,
"learning_rate": 6.437264887377168e-07,
"loss": 0.1201,
"step": 1585
},
{
"epoch": 1.549045947183636,
"grad_norm": 0.5589958429336548,
"learning_rate": 6.410798310537131e-07,
"loss": 0.118,
"step": 1586
},
{
"epoch": 1.5500228972675927,
"grad_norm": 1.73842453956604,
"learning_rate": 6.38437824964827e-07,
"loss": 0.1484,
"step": 1587
},
{
"epoch": 1.5509998473515494,
"grad_norm": 0.795369029045105,
"learning_rate": 6.358004770821774e-07,
"loss": 0.1156,
"step": 1588
},
{
"epoch": 1.5519767974355059,
"grad_norm": 0.6028547286987305,
"learning_rate": 6.331677940052261e-07,
"loss": 0.1278,
"step": 1589
},
{
"epoch": 1.5529537475194628,
"grad_norm": 1.1297847032546997,
"learning_rate": 6.305397823217649e-07,
"loss": 0.1218,
"step": 1590
},
{
"epoch": 1.5539306976034193,
"grad_norm": 0.6576840877532959,
"learning_rate": 6.279164486078926e-07,
"loss": 0.123,
"step": 1591
},
{
"epoch": 1.554907647687376,
"grad_norm": 0.497183233499527,
"learning_rate": 6.252977994280055e-07,
"loss": 0.1265,
"step": 1592
},
{
"epoch": 1.5558845977713327,
"grad_norm": 0.5563151836395264,
"learning_rate": 6.226838413347766e-07,
"loss": 0.1325,
"step": 1593
},
{
"epoch": 1.5568615478552892,
"grad_norm": 0.742324948310852,
"learning_rate": 6.200745808691391e-07,
"loss": 0.1259,
"step": 1594
},
{
"epoch": 1.557838497939246,
"grad_norm": 0.5966019034385681,
"learning_rate": 6.174700245602739e-07,
"loss": 0.1313,
"step": 1595
},
{
"epoch": 1.5588154480232026,
"grad_norm": 0.48993411660194397,
"learning_rate": 6.148701789255881e-07,
"loss": 0.1215,
"step": 1596
},
{
"epoch": 1.559792398107159,
"grad_norm": 0.8104481101036072,
"learning_rate": 6.12275050470702e-07,
"loss": 0.1253,
"step": 1597
},
{
"epoch": 1.560769348191116,
"grad_norm": 0.6388115882873535,
"learning_rate": 6.096846456894326e-07,
"loss": 0.1262,
"step": 1598
},
{
"epoch": 1.5617462982750725,
"grad_norm": 0.5106819272041321,
"learning_rate": 6.070989710637773e-07,
"loss": 0.1198,
"step": 1599
},
{
"epoch": 1.5627232483590292,
"grad_norm": 0.4931243360042572,
"learning_rate": 6.04518033063895e-07,
"loss": 0.1149,
"step": 1600
},
{
"epoch": 1.563700198442986,
"grad_norm": 0.6106110215187073,
"learning_rate": 6.019418381480952e-07,
"loss": 0.1271,
"step": 1601
},
{
"epoch": 1.5646771485269424,
"grad_norm": 0.5603227615356445,
"learning_rate": 5.99370392762815e-07,
"loss": 0.1149,
"step": 1602
},
{
"epoch": 1.565654098610899,
"grad_norm": 0.5077688097953796,
"learning_rate": 5.968037033426108e-07,
"loss": 0.1288,
"step": 1603
},
{
"epoch": 1.5666310486948558,
"grad_norm": 0.621652364730835,
"learning_rate": 5.942417763101346e-07,
"loss": 0.1164,
"step": 1604
},
{
"epoch": 1.5676079987788123,
"grad_norm": 0.548293948173523,
"learning_rate": 5.916846180761229e-07,
"loss": 0.1138,
"step": 1605
},
{
"epoch": 1.568584948862769,
"grad_norm": 0.7004862427711487,
"learning_rate": 5.891322350393791e-07,
"loss": 0.1184,
"step": 1606
},
{
"epoch": 1.5695618989467257,
"grad_norm": 0.43852436542510986,
"learning_rate": 5.86584633586759e-07,
"loss": 0.1215,
"step": 1607
},
{
"epoch": 1.5705388490306822,
"grad_norm": 1.1353037357330322,
"learning_rate": 5.840418200931499e-07,
"loss": 0.1207,
"step": 1608
},
{
"epoch": 1.5715157991146391,
"grad_norm": 0.5289071798324585,
"learning_rate": 5.815038009214625e-07,
"loss": 0.1171,
"step": 1609
},
{
"epoch": 1.5724927491985956,
"grad_norm": 0.5824196934700012,
"learning_rate": 5.78970582422606e-07,
"loss": 0.1241,
"step": 1610
},
{
"epoch": 1.5734696992825523,
"grad_norm": 0.7752619981765747,
"learning_rate": 5.764421709354812e-07,
"loss": 0.123,
"step": 1611
},
{
"epoch": 1.574446649366509,
"grad_norm": 0.5433787107467651,
"learning_rate": 5.739185727869573e-07,
"loss": 0.1344,
"step": 1612
},
{
"epoch": 1.5754235994504655,
"grad_norm": 0.8078683018684387,
"learning_rate": 5.713997942918595e-07,
"loss": 0.1181,
"step": 1613
},
{
"epoch": 1.5764005495344222,
"grad_norm": 0.4675946831703186,
"learning_rate": 5.688858417529539e-07,
"loss": 0.1149,
"step": 1614
},
{
"epoch": 1.577377499618379,
"grad_norm": 0.5040484070777893,
"learning_rate": 5.663767214609301e-07,
"loss": 0.1126,
"step": 1615
},
{
"epoch": 1.5783544497023354,
"grad_norm": 0.5947394967079163,
"learning_rate": 5.638724396943848e-07,
"loss": 0.1289,
"step": 1616
},
{
"epoch": 1.5793313997862923,
"grad_norm": 0.7541037201881409,
"learning_rate": 5.613730027198089e-07,
"loss": 0.1375,
"step": 1617
},
{
"epoch": 1.5803083498702488,
"grad_norm": 0.6051619648933411,
"learning_rate": 5.588784167915696e-07,
"loss": 0.1298,
"step": 1618
},
{
"epoch": 1.5812852999542053,
"grad_norm": 0.5421449542045593,
"learning_rate": 5.563886881518943e-07,
"loss": 0.1211,
"step": 1619
},
{
"epoch": 1.5822622500381622,
"grad_norm": 0.5993972420692444,
"learning_rate": 5.539038230308564e-07,
"loss": 0.111,
"step": 1620
},
{
"epoch": 1.5832392001221187,
"grad_norm": 0.48703792691230774,
"learning_rate": 5.514238276463607e-07,
"loss": 0.1157,
"step": 1621
},
{
"epoch": 1.5842161502060754,
"grad_norm": 0.4974450170993805,
"learning_rate": 5.489487082041236e-07,
"loss": 0.124,
"step": 1622
},
{
"epoch": 1.5851931002900321,
"grad_norm": 0.48971793055534363,
"learning_rate": 5.464784708976639e-07,
"loss": 0.1245,
"step": 1623
},
{
"epoch": 1.5861700503739886,
"grad_norm": 0.4794745147228241,
"learning_rate": 5.4401312190828e-07,
"loss": 0.1259,
"step": 1624
},
{
"epoch": 1.5871470004579453,
"grad_norm": 0.5340744853019714,
"learning_rate": 5.415526674050412e-07,
"loss": 0.1195,
"step": 1625
},
{
"epoch": 1.588123950541902,
"grad_norm": 0.8452017903327942,
"learning_rate": 5.390971135447687e-07,
"loss": 0.1139,
"step": 1626
},
{
"epoch": 1.5891009006258585,
"grad_norm": 0.5496191382408142,
"learning_rate": 5.366464664720195e-07,
"loss": 0.1269,
"step": 1627
},
{
"epoch": 1.5900778507098154,
"grad_norm": 0.4589594602584839,
"learning_rate": 5.342007323190726e-07,
"loss": 0.129,
"step": 1628
},
{
"epoch": 1.591054800793772,
"grad_norm": 0.9928551912307739,
"learning_rate": 5.317599172059149e-07,
"loss": 0.1124,
"step": 1629
},
{
"epoch": 1.5920317508777286,
"grad_norm": 0.8636537194252014,
"learning_rate": 5.29324027240222e-07,
"loss": 0.1209,
"step": 1630
},
{
"epoch": 1.5930087009616853,
"grad_norm": 0.45822155475616455,
"learning_rate": 5.268930685173473e-07,
"loss": 0.108,
"step": 1631
},
{
"epoch": 1.5939856510456418,
"grad_norm": 0.4430014193058014,
"learning_rate": 5.244670471203042e-07,
"loss": 0.1197,
"step": 1632
},
{
"epoch": 1.5949626011295985,
"grad_norm": 0.7680898308753967,
"learning_rate": 5.220459691197502e-07,
"loss": 0.1253,
"step": 1633
},
{
"epoch": 1.5959395512135552,
"grad_norm": 0.7509202361106873,
"learning_rate": 5.196298405739749e-07,
"loss": 0.1298,
"step": 1634
},
{
"epoch": 1.5969165012975117,
"grad_norm": 0.6401910781860352,
"learning_rate": 5.17218667528881e-07,
"loss": 0.1275,
"step": 1635
},
{
"epoch": 1.5978934513814684,
"grad_norm": 0.4642656445503235,
"learning_rate": 5.148124560179707e-07,
"loss": 0.1146,
"step": 1636
},
{
"epoch": 1.5988704014654251,
"grad_norm": 0.5455799102783203,
"learning_rate": 5.124112120623337e-07,
"loss": 0.0989,
"step": 1637
},
{
"epoch": 1.5998473515493816,
"grad_norm": 0.628270149230957,
"learning_rate": 5.10014941670626e-07,
"loss": 0.127,
"step": 1638
},
{
"epoch": 1.6008243016333386,
"grad_norm": 0.6266299486160278,
"learning_rate": 5.076236508390606e-07,
"loss": 0.129,
"step": 1639
},
{
"epoch": 1.601801251717295,
"grad_norm": 0.5914819836616516,
"learning_rate": 5.052373455513895e-07,
"loss": 0.1211,
"step": 1640
},
{
"epoch": 1.6027782018012517,
"grad_norm": 0.5658186078071594,
"learning_rate": 5.02856031778888e-07,
"loss": 0.1214,
"step": 1641
},
{
"epoch": 1.6037551518852085,
"grad_norm": 0.5784573554992676,
"learning_rate": 5.004797154803431e-07,
"loss": 0.1313,
"step": 1642
},
{
"epoch": 1.604732101969165,
"grad_norm": 0.4859617352485657,
"learning_rate": 4.981084026020355e-07,
"loss": 0.106,
"step": 1643
},
{
"epoch": 1.6057090520531216,
"grad_norm": 0.5939765572547913,
"learning_rate": 4.957420990777248e-07,
"loss": 0.1159,
"step": 1644
},
{
"epoch": 1.6066860021370784,
"grad_norm": 0.6516666412353516,
"learning_rate": 4.933808108286381e-07,
"loss": 0.1326,
"step": 1645
},
{
"epoch": 1.6076629522210348,
"grad_norm": 0.995316207408905,
"learning_rate": 4.910245437634501e-07,
"loss": 0.1313,
"step": 1646
},
{
"epoch": 1.6086399023049918,
"grad_norm": 0.49212634563446045,
"learning_rate": 4.886733037782728e-07,
"loss": 0.1212,
"step": 1647
},
{
"epoch": 1.6096168523889482,
"grad_norm": 0.5230310559272766,
"learning_rate": 4.863270967566389e-07,
"loss": 0.1203,
"step": 1648
},
{
"epoch": 1.6105938024729047,
"grad_norm": 0.4029630720615387,
"learning_rate": 4.839859285694854e-07,
"loss": 0.1063,
"step": 1649
},
{
"epoch": 1.6115707525568617,
"grad_norm": 0.499109148979187,
"learning_rate": 4.816498050751422e-07,
"loss": 0.1224,
"step": 1650
},
{
"epoch": 1.6125477026408181,
"grad_norm": 0.9248983860015869,
"learning_rate": 4.793187321193155e-07,
"loss": 0.1458,
"step": 1651
},
{
"epoch": 1.6135246527247749,
"grad_norm": 0.5667829513549805,
"learning_rate": 4.769927155350721e-07,
"loss": 0.1165,
"step": 1652
},
{
"epoch": 1.6145016028087316,
"grad_norm": 0.6089081764221191,
"learning_rate": 4.746717611428278e-07,
"loss": 0.1164,
"step": 1653
},
{
"epoch": 1.615478552892688,
"grad_norm": 0.6611413359642029,
"learning_rate": 4.7235587475033153e-07,
"loss": 0.1333,
"step": 1654
},
{
"epoch": 1.6164555029766448,
"grad_norm": 0.5669116973876953,
"learning_rate": 4.700450621526484e-07,
"loss": 0.1256,
"step": 1655
},
{
"epoch": 1.6174324530606015,
"grad_norm": 0.4340880513191223,
"learning_rate": 4.6773932913215e-07,
"loss": 0.1236,
"step": 1656
},
{
"epoch": 1.618409403144558,
"grad_norm": 0.5093257427215576,
"learning_rate": 4.654386814584944e-07,
"loss": 0.1302,
"step": 1657
},
{
"epoch": 1.6193863532285149,
"grad_norm": 1.0469484329223633,
"learning_rate": 4.631431248886173e-07,
"loss": 0.123,
"step": 1658
},
{
"epoch": 1.6203633033124714,
"grad_norm": 0.46528348326683044,
"learning_rate": 4.6085266516671306e-07,
"loss": 0.1053,
"step": 1659
},
{
"epoch": 1.621340253396428,
"grad_norm": 0.5752331614494324,
"learning_rate": 4.58567308024222e-07,
"loss": 0.1224,
"step": 1660
},
{
"epoch": 1.6223172034803848,
"grad_norm": 0.7798343300819397,
"learning_rate": 4.562870591798177e-07,
"loss": 0.1198,
"step": 1661
},
{
"epoch": 1.6232941535643413,
"grad_norm": 2.5183918476104736,
"learning_rate": 4.540119243393906e-07,
"loss": 0.1166,
"step": 1662
},
{
"epoch": 1.624271103648298,
"grad_norm": 1.0901920795440674,
"learning_rate": 4.5174190919603316e-07,
"loss": 0.1158,
"step": 1663
},
{
"epoch": 1.6252480537322547,
"grad_norm": 0.6507642865180969,
"learning_rate": 4.494770194300288e-07,
"loss": 0.1213,
"step": 1664
},
{
"epoch": 1.6262250038162112,
"grad_norm": 0.48321932554244995,
"learning_rate": 4.4721726070883377e-07,
"loss": 0.1179,
"step": 1665
},
{
"epoch": 1.6272019539001679,
"grad_norm": 0.5404170155525208,
"learning_rate": 4.4496263868706534e-07,
"loss": 0.1205,
"step": 1666
},
{
"epoch": 1.6281789039841246,
"grad_norm": 0.5240875482559204,
"learning_rate": 4.4271315900648785e-07,
"loss": 0.1032,
"step": 1667
},
{
"epoch": 1.629155854068081,
"grad_norm": 1.1204215288162231,
"learning_rate": 4.404688272959981e-07,
"loss": 0.1127,
"step": 1668
},
{
"epoch": 1.630132804152038,
"grad_norm": 0.7335038185119629,
"learning_rate": 4.3822964917160953e-07,
"loss": 0.1253,
"step": 1669
},
{
"epoch": 1.6311097542359945,
"grad_norm": 0.8138047456741333,
"learning_rate": 4.359956302364418e-07,
"loss": 0.1276,
"step": 1670
},
{
"epoch": 1.6320867043199512,
"grad_norm": 0.5346017479896545,
"learning_rate": 4.337667760807018e-07,
"loss": 0.1301,
"step": 1671
},
{
"epoch": 1.6330636544039079,
"grad_norm": 0.5730168223381042,
"learning_rate": 4.3154309228167573e-07,
"loss": 0.1137,
"step": 1672
},
{
"epoch": 1.6340406044878644,
"grad_norm": 0.5700895190238953,
"learning_rate": 4.293245844037111e-07,
"loss": 0.1172,
"step": 1673
},
{
"epoch": 1.635017554571821,
"grad_norm": 1.7504950761795044,
"learning_rate": 4.2711125799820093e-07,
"loss": 0.1306,
"step": 1674
},
{
"epoch": 1.6359945046557778,
"grad_norm": 0.6507771611213684,
"learning_rate": 4.2490311860357597e-07,
"loss": 0.108,
"step": 1675
},
{
"epoch": 1.6369714547397343,
"grad_norm": 0.7692915201187134,
"learning_rate": 4.227001717452864e-07,
"loss": 0.136,
"step": 1676
},
{
"epoch": 1.6379484048236912,
"grad_norm": 0.5343664884567261,
"learning_rate": 4.2050242293578817e-07,
"loss": 0.1314,
"step": 1677
},
{
"epoch": 1.6389253549076477,
"grad_norm": 0.5488064289093018,
"learning_rate": 4.1830987767453174e-07,
"loss": 0.118,
"step": 1678
},
{
"epoch": 1.6399023049916044,
"grad_norm": 0.7980213761329651,
"learning_rate": 4.1612254144794443e-07,
"loss": 0.1191,
"step": 1679
},
{
"epoch": 1.640879255075561,
"grad_norm": 0.5957295894622803,
"learning_rate": 4.139404197294211e-07,
"loss": 0.1238,
"step": 1680
},
{
"epoch": 1.6418562051595176,
"grad_norm": 0.4211692214012146,
"learning_rate": 4.117635179793078e-07,
"loss": 0.1271,
"step": 1681
},
{
"epoch": 1.6428331552434743,
"grad_norm": 0.7664165496826172,
"learning_rate": 4.0959184164488796e-07,
"loss": 0.129,
"step": 1682
},
{
"epoch": 1.643810105327431,
"grad_norm": 0.5844789743423462,
"learning_rate": 4.0742539616036887e-07,
"loss": 0.1318,
"step": 1683
},
{
"epoch": 1.6447870554113875,
"grad_norm": 1.0815391540527344,
"learning_rate": 4.052641869468704e-07,
"loss": 0.1215,
"step": 1684
},
{
"epoch": 1.6457640054953442,
"grad_norm": 0.613914430141449,
"learning_rate": 4.0310821941240807e-07,
"loss": 0.1243,
"step": 1685
},
{
"epoch": 1.646740955579301,
"grad_norm": 0.5809972882270813,
"learning_rate": 4.0095749895188164e-07,
"loss": 0.129,
"step": 1686
},
{
"epoch": 1.6477179056632574,
"grad_norm": 0.6392321586608887,
"learning_rate": 3.9881203094706194e-07,
"loss": 0.1211,
"step": 1687
},
{
"epoch": 1.6486948557472143,
"grad_norm": 0.6616940498352051,
"learning_rate": 3.9667182076657456e-07,
"loss": 0.1272,
"step": 1688
},
{
"epoch": 1.6496718058311708,
"grad_norm": 0.905881404876709,
"learning_rate": 3.945368737658908e-07,
"loss": 0.1364,
"step": 1689
},
{
"epoch": 1.6506487559151275,
"grad_norm": 0.5154632329940796,
"learning_rate": 3.9240719528730986e-07,
"loss": 0.1161,
"step": 1690
},
{
"epoch": 1.6516257059990842,
"grad_norm": 0.6851862668991089,
"learning_rate": 3.9028279065994765e-07,
"loss": 0.1195,
"step": 1691
},
{
"epoch": 1.6526026560830407,
"grad_norm": 1.2716608047485352,
"learning_rate": 3.8816366519972477e-07,
"loss": 0.1311,
"step": 1692
},
{
"epoch": 1.6535796061669974,
"grad_norm": 1.04118812084198,
"learning_rate": 3.860498242093497e-07,
"loss": 0.1211,
"step": 1693
},
{
"epoch": 1.654556556250954,
"grad_norm": 0.5050714612007141,
"learning_rate": 3.83941272978309e-07,
"loss": 0.1162,
"step": 1694
},
{
"epoch": 1.6555335063349106,
"grad_norm": 0.4503592848777771,
"learning_rate": 3.818380167828528e-07,
"loss": 0.1164,
"step": 1695
},
{
"epoch": 1.6565104564188673,
"grad_norm": 0.8338929414749146,
"learning_rate": 3.7974006088598e-07,
"loss": 0.115,
"step": 1696
},
{
"epoch": 1.657487406502824,
"grad_norm": 0.4717448949813843,
"learning_rate": 3.776474105374278e-07,
"loss": 0.1078,
"step": 1697
},
{
"epoch": 1.6584643565867805,
"grad_norm": 0.5584269165992737,
"learning_rate": 3.755600709736565e-07,
"loss": 0.1181,
"step": 1698
},
{
"epoch": 1.6594413066707374,
"grad_norm": 0.7003796100616455,
"learning_rate": 3.7347804741783705e-07,
"loss": 0.1419,
"step": 1699
},
{
"epoch": 1.660418256754694,
"grad_norm": 0.5545423626899719,
"learning_rate": 3.714013450798387e-07,
"loss": 0.1274,
"step": 1700
},
{
"epoch": 1.6613952068386506,
"grad_norm": 0.7139258980751038,
"learning_rate": 3.6932996915621604e-07,
"loss": 0.1272,
"step": 1701
},
{
"epoch": 1.6623721569226073,
"grad_norm": 0.47771155834198,
"learning_rate": 3.672639248301929e-07,
"loss": 0.1323,
"step": 1702
},
{
"epoch": 1.6633491070065638,
"grad_norm": 0.7455830574035645,
"learning_rate": 3.652032172716546e-07,
"loss": 0.1262,
"step": 1703
},
{
"epoch": 1.6643260570905205,
"grad_norm": 0.5801220536231995,
"learning_rate": 3.631478516371306e-07,
"loss": 0.1146,
"step": 1704
},
{
"epoch": 1.6653030071744772,
"grad_norm": 0.5234197378158569,
"learning_rate": 3.610978330697826e-07,
"loss": 0.1214,
"step": 1705
},
{
"epoch": 1.6662799572584337,
"grad_norm": 0.9796130061149597,
"learning_rate": 3.590531666993946e-07,
"loss": 0.1156,
"step": 1706
},
{
"epoch": 1.6672569073423906,
"grad_norm": 0.6356166005134583,
"learning_rate": 3.570138576423551e-07,
"loss": 0.1163,
"step": 1707
},
{
"epoch": 1.668233857426347,
"grad_norm": 1.2422935962677002,
"learning_rate": 3.5497991100164887e-07,
"loss": 0.1311,
"step": 1708
},
{
"epoch": 1.6692108075103038,
"grad_norm": 0.5308371186256409,
"learning_rate": 3.5295133186684177e-07,
"loss": 0.1258,
"step": 1709
},
{
"epoch": 1.6701877575942605,
"grad_norm": 0.42890292406082153,
"learning_rate": 3.5092812531406754e-07,
"loss": 0.1143,
"step": 1710
},
{
"epoch": 1.671164707678217,
"grad_norm": 0.5552493929862976,
"learning_rate": 3.4891029640601813e-07,
"loss": 0.1111,
"step": 1711
},
{
"epoch": 1.6721416577621737,
"grad_norm": 0.995431125164032,
"learning_rate": 3.468978501919265e-07,
"loss": 0.1324,
"step": 1712
},
{
"epoch": 1.6731186078461304,
"grad_norm": 1.1340943574905396,
"learning_rate": 3.448907917075575e-07,
"loss": 0.1296,
"step": 1713
},
{
"epoch": 1.674095557930087,
"grad_norm": 0.5379215478897095,
"learning_rate": 3.428891259751949e-07,
"loss": 0.1161,
"step": 1714
},
{
"epoch": 1.6750725080140436,
"grad_norm": 0.6913337111473083,
"learning_rate": 3.408928580036269e-07,
"loss": 0.1175,
"step": 1715
},
{
"epoch": 1.6760494580980003,
"grad_norm": 0.5137271285057068,
"learning_rate": 3.3890199278813596e-07,
"loss": 0.1338,
"step": 1716
},
{
"epoch": 1.6770264081819568,
"grad_norm": 0.8610045909881592,
"learning_rate": 3.3691653531048475e-07,
"loss": 0.113,
"step": 1717
},
{
"epoch": 1.6780033582659137,
"grad_norm": 0.627528727054596,
"learning_rate": 3.3493649053890325e-07,
"loss": 0.1142,
"step": 1718
},
{
"epoch": 1.6789803083498702,
"grad_norm": 0.9074206948280334,
"learning_rate": 3.3296186342807845e-07,
"loss": 0.1216,
"step": 1719
},
{
"epoch": 1.679957258433827,
"grad_norm": 0.5306910872459412,
"learning_rate": 3.309926589191414e-07,
"loss": 0.1194,
"step": 1720
},
{
"epoch": 1.6809342085177836,
"grad_norm": 0.677941083908081,
"learning_rate": 3.2902888193965033e-07,
"loss": 0.1168,
"step": 1721
},
{
"epoch": 1.6819111586017401,
"grad_norm": 0.6383283734321594,
"learning_rate": 3.270705374035857e-07,
"loss": 0.1311,
"step": 1722
},
{
"epoch": 1.6828881086856968,
"grad_norm": 0.5561057329177856,
"learning_rate": 3.25117630211334e-07,
"loss": 0.1171,
"step": 1723
},
{
"epoch": 1.6838650587696535,
"grad_norm": 1.2501775026321411,
"learning_rate": 3.231701652496741e-07,
"loss": 0.1211,
"step": 1724
},
{
"epoch": 1.68484200885361,
"grad_norm": 0.6360272169113159,
"learning_rate": 3.2122814739176833e-07,
"loss": 0.1088,
"step": 1725
},
{
"epoch": 1.685818958937567,
"grad_norm": 0.7439811825752258,
"learning_rate": 3.1929158149714726e-07,
"loss": 0.1223,
"step": 1726
},
{
"epoch": 1.6867959090215234,
"grad_norm": 0.46923208236694336,
"learning_rate": 3.173604724117002e-07,
"loss": 0.1174,
"step": 1727
},
{
"epoch": 1.68777285910548,
"grad_norm": 0.7019605040550232,
"learning_rate": 3.1543482496766255e-07,
"loss": 0.1059,
"step": 1728
},
{
"epoch": 1.6887498091894368,
"grad_norm": 0.5157560706138611,
"learning_rate": 3.1351464398359897e-07,
"loss": 0.1272,
"step": 1729
},
{
"epoch": 1.6897267592733933,
"grad_norm": 0.681706964969635,
"learning_rate": 3.115999342644002e-07,
"loss": 0.1275,
"step": 1730
},
{
"epoch": 1.69070370935735,
"grad_norm": 0.5303037166595459,
"learning_rate": 3.096907006012639e-07,
"loss": 0.1343,
"step": 1731
},
{
"epoch": 1.6916806594413067,
"grad_norm": 0.46930181980133057,
"learning_rate": 3.077869477716844e-07,
"loss": 0.1146,
"step": 1732
},
{
"epoch": 1.6926576095252632,
"grad_norm": 1.014082431793213,
"learning_rate": 3.0588868053944233e-07,
"loss": 0.1268,
"step": 1733
},
{
"epoch": 1.69363455960922,
"grad_norm": 0.501285970211029,
"learning_rate": 3.0399590365459223e-07,
"loss": 0.1162,
"step": 1734
},
{
"epoch": 1.6946115096931766,
"grad_norm": 0.801666259765625,
"learning_rate": 3.0210862185344856e-07,
"loss": 0.1342,
"step": 1735
},
{
"epoch": 1.6955884597771331,
"grad_norm": 0.5253545045852661,
"learning_rate": 3.002268398585756e-07,
"loss": 0.1173,
"step": 1736
},
{
"epoch": 1.69656540986109,
"grad_norm": 0.7428088188171387,
"learning_rate": 2.983505623787766e-07,
"loss": 0.1281,
"step": 1737
},
{
"epoch": 1.6975423599450465,
"grad_norm": 0.6004049181938171,
"learning_rate": 2.964797941090794e-07,
"loss": 0.1241,
"step": 1738
},
{
"epoch": 1.6985193100290032,
"grad_norm": 0.9722501635551453,
"learning_rate": 2.946145397307276e-07,
"loss": 0.1325,
"step": 1739
},
{
"epoch": 1.69949626011296,
"grad_norm": 0.483163446187973,
"learning_rate": 2.927548039111655e-07,
"loss": 0.1222,
"step": 1740
},
{
"epoch": 1.7004732101969164,
"grad_norm": 0.49213358759880066,
"learning_rate": 2.909005913040297e-07,
"loss": 0.1102,
"step": 1741
},
{
"epoch": 1.7014501602808731,
"grad_norm": 0.5967307090759277,
"learning_rate": 2.8905190654913595e-07,
"loss": 0.1309,
"step": 1742
},
{
"epoch": 1.7024271103648299,
"grad_norm": 0.6980396509170532,
"learning_rate": 2.872087542724669e-07,
"loss": 0.1245,
"step": 1743
},
{
"epoch": 1.7034040604487863,
"grad_norm": 0.7286382913589478,
"learning_rate": 2.8537113908616074e-07,
"loss": 0.1332,
"step": 1744
},
{
"epoch": 1.704381010532743,
"grad_norm": 0.49004125595092773,
"learning_rate": 2.8353906558850203e-07,
"loss": 0.1187,
"step": 1745
},
{
"epoch": 1.7053579606166998,
"grad_norm": 0.8462884426116943,
"learning_rate": 2.8171253836390605e-07,
"loss": 0.1203,
"step": 1746
},
{
"epoch": 1.7063349107006562,
"grad_norm": 0.5028384327888489,
"learning_rate": 2.7989156198291207e-07,
"loss": 0.1114,
"step": 1747
},
{
"epoch": 1.7073118607846132,
"grad_norm": 0.6612257957458496,
"learning_rate": 2.7807614100216707e-07,
"loss": 0.1406,
"step": 1748
},
{
"epoch": 1.7082888108685697,
"grad_norm": 0.628887414932251,
"learning_rate": 2.762662799644178e-07,
"loss": 0.1195,
"step": 1749
},
{
"epoch": 1.7092657609525264,
"grad_norm": 0.6640041470527649,
"learning_rate": 2.7446198339849875e-07,
"loss": 0.1311,
"step": 1750
},
{
"epoch": 1.710242711036483,
"grad_norm": 0.5691184401512146,
"learning_rate": 2.726632558193196e-07,
"loss": 0.1176,
"step": 1751
},
{
"epoch": 1.7112196611204396,
"grad_norm": 0.5496095418930054,
"learning_rate": 2.708701017278545e-07,
"loss": 0.1068,
"step": 1752
},
{
"epoch": 1.7121966112043963,
"grad_norm": 0.5080822110176086,
"learning_rate": 2.6908252561113207e-07,
"loss": 0.1139,
"step": 1753
},
{
"epoch": 1.713173561288353,
"grad_norm": 0.672421395778656,
"learning_rate": 2.673005319422217e-07,
"loss": 0.1202,
"step": 1754
},
{
"epoch": 1.7141505113723094,
"grad_norm": 0.797073483467102,
"learning_rate": 2.6552412518022483e-07,
"loss": 0.1244,
"step": 1755
},
{
"epoch": 1.7151274614562664,
"grad_norm": 0.6942318677902222,
"learning_rate": 2.6375330977026323e-07,
"loss": 0.1302,
"step": 1756
},
{
"epoch": 1.7161044115402229,
"grad_norm": 1.2046414613723755,
"learning_rate": 2.6198809014346545e-07,
"loss": 0.1389,
"step": 1757
},
{
"epoch": 1.7170813616241796,
"grad_norm": 1.2071318626403809,
"learning_rate": 2.602284707169597e-07,
"loss": 0.1341,
"step": 1758
},
{
"epoch": 1.7180583117081363,
"grad_norm": 0.5155503749847412,
"learning_rate": 2.584744558938593e-07,
"loss": 0.1318,
"step": 1759
},
{
"epoch": 1.7190352617920928,
"grad_norm": 0.5806264281272888,
"learning_rate": 2.5672605006325347e-07,
"loss": 0.1142,
"step": 1760
},
{
"epoch": 1.7200122118760495,
"grad_norm": 0.7489654421806335,
"learning_rate": 2.5498325760019646e-07,
"loss": 0.1181,
"step": 1761
},
{
"epoch": 1.7209891619600062,
"grad_norm": 0.6799153685569763,
"learning_rate": 2.532460828656957e-07,
"loss": 0.1318,
"step": 1762
},
{
"epoch": 1.7219661120439627,
"grad_norm": 0.4637848436832428,
"learning_rate": 2.5151453020670127e-07,
"loss": 0.1128,
"step": 1763
},
{
"epoch": 1.7229430621279194,
"grad_norm": 0.6327468156814575,
"learning_rate": 2.497886039560962e-07,
"loss": 0.1191,
"step": 1764
},
{
"epoch": 1.723920012211876,
"grad_norm": 1.051252841949463,
"learning_rate": 2.480683084326821e-07,
"loss": 0.1488,
"step": 1765
},
{
"epoch": 1.7248969622958326,
"grad_norm": 0.7160297632217407,
"learning_rate": 2.463536479411735e-07,
"loss": 0.13,
"step": 1766
},
{
"epoch": 1.7258739123797895,
"grad_norm": 0.5140122771263123,
"learning_rate": 2.4464462677218276e-07,
"loss": 0.1105,
"step": 1767
},
{
"epoch": 1.726850862463746,
"grad_norm": 0.6712512969970703,
"learning_rate": 2.429412492022101e-07,
"loss": 0.1112,
"step": 1768
},
{
"epoch": 1.7278278125477027,
"grad_norm": 0.6085613369941711,
"learning_rate": 2.412435194936358e-07,
"loss": 0.1175,
"step": 1769
},
{
"epoch": 1.7288047626316594,
"grad_norm": 0.8117672801017761,
"learning_rate": 2.3955144189470675e-07,
"loss": 0.1317,
"step": 1770
},
{
"epoch": 1.7297817127156159,
"grad_norm": 0.8529082536697388,
"learning_rate": 2.37865020639525e-07,
"loss": 0.1305,
"step": 1771
},
{
"epoch": 1.7307586627995726,
"grad_norm": 0.5067161917686462,
"learning_rate": 2.361842599480413e-07,
"loss": 0.1253,
"step": 1772
},
{
"epoch": 1.7317356128835293,
"grad_norm": 0.5894564986228943,
"learning_rate": 2.3450916402603884e-07,
"loss": 0.1311,
"step": 1773
},
{
"epoch": 1.7327125629674858,
"grad_norm": 0.4523022770881653,
"learning_rate": 2.328397370651289e-07,
"loss": 0.117,
"step": 1774
},
{
"epoch": 1.7336895130514425,
"grad_norm": 0.5705298781394958,
"learning_rate": 2.3117598324273515e-07,
"loss": 0.1234,
"step": 1775
},
{
"epoch": 1.7346664631353992,
"grad_norm": 0.4949217438697815,
"learning_rate": 2.2951790672208567e-07,
"loss": 0.1263,
"step": 1776
},
{
"epoch": 1.7356434132193557,
"grad_norm": 0.5339197516441345,
"learning_rate": 2.2786551165220273e-07,
"loss": 0.1207,
"step": 1777
},
{
"epoch": 1.7366203633033126,
"grad_norm": 0.5311686992645264,
"learning_rate": 2.2621880216789182e-07,
"loss": 0.1188,
"step": 1778
},
{
"epoch": 1.737597313387269,
"grad_norm": 0.6782169342041016,
"learning_rate": 2.245777823897305e-07,
"loss": 0.1242,
"step": 1779
},
{
"epoch": 1.7385742634712258,
"grad_norm": 0.5356742143630981,
"learning_rate": 2.2294245642406008e-07,
"loss": 0.1287,
"step": 1780
},
{
"epoch": 1.7395512135551825,
"grad_norm": 0.514859676361084,
"learning_rate": 2.2131282836297313e-07,
"loss": 0.1255,
"step": 1781
},
{
"epoch": 1.740528163639139,
"grad_norm": 0.8519387245178223,
"learning_rate": 2.1968890228430517e-07,
"loss": 0.115,
"step": 1782
},
{
"epoch": 1.7415051137230957,
"grad_norm": 1.0648313760757446,
"learning_rate": 2.180706822516232e-07,
"loss": 0.1095,
"step": 1783
},
{
"epoch": 1.7424820638070524,
"grad_norm": 0.7434473037719727,
"learning_rate": 2.1645817231421506e-07,
"loss": 0.1343,
"step": 1784
},
{
"epoch": 1.7434590138910089,
"grad_norm": 0.5473856925964355,
"learning_rate": 2.148513765070817e-07,
"loss": 0.1131,
"step": 1785
},
{
"epoch": 1.7444359639749658,
"grad_norm": 0.5825949907302856,
"learning_rate": 2.1325029885092496e-07,
"loss": 0.116,
"step": 1786
},
{
"epoch": 1.7454129140589223,
"grad_norm": 0.5620967745780945,
"learning_rate": 2.1165494335213754e-07,
"loss": 0.1311,
"step": 1787
},
{
"epoch": 1.746389864142879,
"grad_norm": 0.8868221044540405,
"learning_rate": 2.1006531400279417e-07,
"loss": 0.112,
"step": 1788
},
{
"epoch": 1.7473668142268357,
"grad_norm": 0.5546829104423523,
"learning_rate": 2.0848141478064133e-07,
"loss": 0.1211,
"step": 1789
},
{
"epoch": 1.7483437643107922,
"grad_norm": 0.5511954426765442,
"learning_rate": 2.069032496490858e-07,
"loss": 0.1207,
"step": 1790
},
{
"epoch": 1.749320714394749,
"grad_norm": 0.5309725403785706,
"learning_rate": 2.0533082255718594e-07,
"loss": 0.113,
"step": 1791
},
{
"epoch": 1.7502976644787056,
"grad_norm": 0.5222966074943542,
"learning_rate": 2.0376413743964347e-07,
"loss": 0.1176,
"step": 1792
},
{
"epoch": 1.751274614562662,
"grad_norm": 0.6175430417060852,
"learning_rate": 2.0220319821678964e-07,
"loss": 0.1226,
"step": 1793
},
{
"epoch": 1.7522515646466188,
"grad_norm": 0.44996610283851624,
"learning_rate": 2.0064800879457973e-07,
"loss": 0.1111,
"step": 1794
},
{
"epoch": 1.7532285147305755,
"grad_norm": 0.45626726746559143,
"learning_rate": 1.9909857306457918e-07,
"loss": 0.1134,
"step": 1795
},
{
"epoch": 1.754205464814532,
"grad_norm": 0.6614998579025269,
"learning_rate": 1.9755489490395797e-07,
"loss": 0.1183,
"step": 1796
},
{
"epoch": 1.755182414898489,
"grad_norm": 0.4641120135784149,
"learning_rate": 1.960169781754781e-07,
"loss": 0.1063,
"step": 1797
},
{
"epoch": 1.7561593649824454,
"grad_norm": 0.5206264853477478,
"learning_rate": 1.944848267274832e-07,
"loss": 0.1293,
"step": 1798
},
{
"epoch": 1.757136315066402,
"grad_norm": 0.5581468939781189,
"learning_rate": 1.9295844439389229e-07,
"loss": 0.1127,
"step": 1799
},
{
"epoch": 1.7581132651503588,
"grad_norm": 0.5560063123703003,
"learning_rate": 1.9143783499418812e-07,
"loss": 0.1152,
"step": 1800
},
{
"epoch": 1.7590902152343153,
"grad_norm": 1.344071865081787,
"learning_rate": 1.8992300233340644e-07,
"loss": 0.1229,
"step": 1801
},
{
"epoch": 1.760067165318272,
"grad_norm": 0.629028856754303,
"learning_rate": 1.88413950202129e-07,
"loss": 0.1112,
"step": 1802
},
{
"epoch": 1.7610441154022287,
"grad_norm": 0.7920987010002136,
"learning_rate": 1.8691068237647296e-07,
"loss": 0.1326,
"step": 1803
},
{
"epoch": 1.7620210654861852,
"grad_norm": 0.7801615595817566,
"learning_rate": 1.8541320261808022e-07,
"loss": 0.1236,
"step": 1804
},
{
"epoch": 1.762998015570142,
"grad_norm": 0.4146665632724762,
"learning_rate": 1.8392151467411052e-07,
"loss": 0.1095,
"step": 1805
},
{
"epoch": 1.7639749656540986,
"grad_norm": 0.5060549974441528,
"learning_rate": 1.824356222772297e-07,
"loss": 0.1067,
"step": 1806
},
{
"epoch": 1.764951915738055,
"grad_norm": 0.6613800525665283,
"learning_rate": 1.8095552914560082e-07,
"loss": 0.1353,
"step": 1807
},
{
"epoch": 1.765928865822012,
"grad_norm": 0.701943039894104,
"learning_rate": 1.794812389828776e-07,
"loss": 0.1221,
"step": 1808
},
{
"epoch": 1.7669058159059685,
"grad_norm": 0.583827793598175,
"learning_rate": 1.7801275547819047e-07,
"loss": 0.1246,
"step": 1809
},
{
"epoch": 1.7678827659899252,
"grad_norm": 0.6502670049667358,
"learning_rate": 1.7655008230614112e-07,
"loss": 0.1222,
"step": 1810
},
{
"epoch": 1.768859716073882,
"grad_norm": 0.6035724878311157,
"learning_rate": 1.750932231267921e-07,
"loss": 0.113,
"step": 1811
},
{
"epoch": 1.7698366661578384,
"grad_norm": 0.7206379771232605,
"learning_rate": 1.7364218158565648e-07,
"loss": 0.122,
"step": 1812
},
{
"epoch": 1.7708136162417951,
"grad_norm": 0.8710956573486328,
"learning_rate": 1.7219696131369146e-07,
"loss": 0.1373,
"step": 1813
},
{
"epoch": 1.7717905663257518,
"grad_norm": 1.339104175567627,
"learning_rate": 1.7075756592728615e-07,
"loss": 0.1167,
"step": 1814
},
{
"epoch": 1.7727675164097083,
"grad_norm": 0.7403225898742676,
"learning_rate": 1.6932399902825385e-07,
"loss": 0.1251,
"step": 1815
},
{
"epoch": 1.7737444664936652,
"grad_norm": 0.5638961791992188,
"learning_rate": 1.6789626420382454e-07,
"loss": 0.14,
"step": 1816
},
{
"epoch": 1.7747214165776217,
"grad_norm": 1.1879632472991943,
"learning_rate": 1.6647436502663346e-07,
"loss": 0.1328,
"step": 1817
},
{
"epoch": 1.7756983666615784,
"grad_norm": 1.1180294752120972,
"learning_rate": 1.650583050547136e-07,
"loss": 0.1091,
"step": 1818
},
{
"epoch": 1.7766753167455351,
"grad_norm": 0.7066265344619751,
"learning_rate": 1.6364808783148694e-07,
"loss": 0.1193,
"step": 1819
},
{
"epoch": 1.7776522668294916,
"grad_norm": 0.7320586442947388,
"learning_rate": 1.622437168857541e-07,
"loss": 0.1197,
"step": 1820
},
{
"epoch": 1.7786292169134483,
"grad_norm": 0.8313867449760437,
"learning_rate": 1.6084519573168728e-07,
"loss": 0.1254,
"step": 1821
},
{
"epoch": 1.779606166997405,
"grad_norm": 0.4735081195831299,
"learning_rate": 1.5945252786881998e-07,
"loss": 0.1095,
"step": 1822
},
{
"epoch": 1.7805831170813615,
"grad_norm": 0.8831436038017273,
"learning_rate": 1.580657167820396e-07,
"loss": 0.1164,
"step": 1823
},
{
"epoch": 1.7815600671653182,
"grad_norm": 0.7069382667541504,
"learning_rate": 1.5668476594157768e-07,
"loss": 0.1237,
"step": 1824
},
{
"epoch": 1.782537017249275,
"grad_norm": 0.5078839063644409,
"learning_rate": 1.553096788030023e-07,
"loss": 0.1162,
"step": 1825
},
{
"epoch": 1.7835139673332314,
"grad_norm": 0.6370827555656433,
"learning_rate": 1.5394045880720716e-07,
"loss": 0.1087,
"step": 1826
},
{
"epoch": 1.7844909174171883,
"grad_norm": 0.5774054527282715,
"learning_rate": 1.525771093804068e-07,
"loss": 0.1154,
"step": 1827
},
{
"epoch": 1.7854678675011448,
"grad_norm": 0.8257759213447571,
"learning_rate": 1.5121963393412402e-07,
"loss": 0.1213,
"step": 1828
},
{
"epoch": 1.7864448175851015,
"grad_norm": 0.6230661273002625,
"learning_rate": 1.4986803586518327e-07,
"loss": 0.1218,
"step": 1829
},
{
"epoch": 1.7874217676690582,
"grad_norm": 1.1795037984848022,
"learning_rate": 1.4852231855570277e-07,
"loss": 0.1244,
"step": 1830
},
{
"epoch": 1.7883987177530147,
"grad_norm": 0.7884340882301331,
"learning_rate": 1.4718248537308465e-07,
"loss": 0.1201,
"step": 1831
},
{
"epoch": 1.7893756678369714,
"grad_norm": 0.4623565971851349,
"learning_rate": 1.458485396700071e-07,
"loss": 0.123,
"step": 1832
},
{
"epoch": 1.7903526179209281,
"grad_norm": 0.43081963062286377,
"learning_rate": 1.44520484784417e-07,
"loss": 0.1159,
"step": 1833
},
{
"epoch": 1.7913295680048846,
"grad_norm": 0.5624845027923584,
"learning_rate": 1.4319832403951877e-07,
"loss": 0.1115,
"step": 1834
},
{
"epoch": 1.7923065180888416,
"grad_norm": 0.5004430413246155,
"learning_rate": 1.4188206074376987e-07,
"loss": 0.1269,
"step": 1835
},
{
"epoch": 1.793283468172798,
"grad_norm": 0.9545750617980957,
"learning_rate": 1.405716981908692e-07,
"loss": 0.1273,
"step": 1836
},
{
"epoch": 1.7942604182567545,
"grad_norm": 1.3716719150543213,
"learning_rate": 1.392672396597497e-07,
"loss": 0.1363,
"step": 1837
},
{
"epoch": 1.7952373683407115,
"grad_norm": 0.6497982144355774,
"learning_rate": 1.3796868841457227e-07,
"loss": 0.1249,
"step": 1838
},
{
"epoch": 1.796214318424668,
"grad_norm": 0.7861154675483704,
"learning_rate": 1.366760477047152e-07,
"loss": 0.1347,
"step": 1839
},
{
"epoch": 1.7971912685086247,
"grad_norm": 0.48248550295829773,
"learning_rate": 1.353893207647658e-07,
"loss": 0.1112,
"step": 1840
},
{
"epoch": 1.7981682185925814,
"grad_norm": 1.1630258560180664,
"learning_rate": 1.3410851081451514e-07,
"loss": 0.118,
"step": 1841
},
{
"epoch": 1.7991451686765378,
"grad_norm": 0.48358047008514404,
"learning_rate": 1.328336210589465e-07,
"loss": 0.1103,
"step": 1842
},
{
"epoch": 1.8001221187604945,
"grad_norm": 0.887587308883667,
"learning_rate": 1.3156465468822988e-07,
"loss": 0.1276,
"step": 1843
},
{
"epoch": 1.8010990688444513,
"grad_norm": 0.5231468677520752,
"learning_rate": 1.3030161487771365e-07,
"loss": 0.1317,
"step": 1844
},
{
"epoch": 1.8020760189284077,
"grad_norm": 0.5103966593742371,
"learning_rate": 1.290445047879138e-07,
"loss": 0.1275,
"step": 1845
},
{
"epoch": 1.8030529690123647,
"grad_norm": 0.6896922588348389,
"learning_rate": 1.2779332756451073e-07,
"loss": 0.1273,
"step": 1846
},
{
"epoch": 1.8040299190963212,
"grad_norm": 0.6292285323143005,
"learning_rate": 1.2654808633833865e-07,
"loss": 0.1058,
"step": 1847
},
{
"epoch": 1.8050068691802779,
"grad_norm": 0.8873867988586426,
"learning_rate": 1.2530878422537668e-07,
"loss": 0.1319,
"step": 1848
},
{
"epoch": 1.8059838192642346,
"grad_norm": 0.5508884787559509,
"learning_rate": 1.2407542432674392e-07,
"loss": 0.1283,
"step": 1849
},
{
"epoch": 1.806960769348191,
"grad_norm": 0.5044010877609253,
"learning_rate": 1.2284800972868883e-07,
"loss": 0.1267,
"step": 1850
},
{
"epoch": 1.8079377194321478,
"grad_norm": 0.422820121049881,
"learning_rate": 1.2162654350258403e-07,
"loss": 0.1108,
"step": 1851
},
{
"epoch": 1.8089146695161045,
"grad_norm": 0.8181347846984863,
"learning_rate": 1.2041102870491734e-07,
"loss": 0.1094,
"step": 1852
},
{
"epoch": 1.809891619600061,
"grad_norm": 0.7040722966194153,
"learning_rate": 1.192014683772838e-07,
"loss": 0.1095,
"step": 1853
},
{
"epoch": 1.8108685696840177,
"grad_norm": 0.6295640468597412,
"learning_rate": 1.1799786554637843e-07,
"loss": 0.1155,
"step": 1854
},
{
"epoch": 1.8118455197679744,
"grad_norm": 0.6320077180862427,
"learning_rate": 1.1680022322398926e-07,
"loss": 0.1148,
"step": 1855
},
{
"epoch": 1.8128224698519309,
"grad_norm": 0.6232160329818726,
"learning_rate": 1.1560854440698877e-07,
"loss": 0.1168,
"step": 1856
},
{
"epoch": 1.8137994199358878,
"grad_norm": 0.9959467053413391,
"learning_rate": 1.1442283207732746e-07,
"loss": 0.1135,
"step": 1857
},
{
"epoch": 1.8147763700198443,
"grad_norm": 0.5906278491020203,
"learning_rate": 1.1324308920202559e-07,
"loss": 0.1279,
"step": 1858
},
{
"epoch": 1.815753320103801,
"grad_norm": 0.691768229007721,
"learning_rate": 1.120693187331659e-07,
"loss": 0.121,
"step": 1859
},
{
"epoch": 1.8167302701877577,
"grad_norm": 0.49635595083236694,
"learning_rate": 1.1090152360788558e-07,
"loss": 0.1255,
"step": 1860
},
{
"epoch": 1.8177072202717142,
"grad_norm": 0.7929428815841675,
"learning_rate": 1.0973970674837159e-07,
"loss": 0.1307,
"step": 1861
},
{
"epoch": 1.8186841703556709,
"grad_norm": 0.6797349452972412,
"learning_rate": 1.085838710618492e-07,
"loss": 0.1158,
"step": 1862
},
{
"epoch": 1.8196611204396276,
"grad_norm": 0.4945840537548065,
"learning_rate": 1.0743401944057902e-07,
"loss": 0.1155,
"step": 1863
},
{
"epoch": 1.820638070523584,
"grad_norm": 0.5590537190437317,
"learning_rate": 1.0629015476184557e-07,
"loss": 0.109,
"step": 1864
},
{
"epoch": 1.821615020607541,
"grad_norm": 0.8970704078674316,
"learning_rate": 1.0515227988795369e-07,
"loss": 0.1255,
"step": 1865
},
{
"epoch": 1.8225919706914975,
"grad_norm": 0.6763855218887329,
"learning_rate": 1.040203976662199e-07,
"loss": 0.1142,
"step": 1866
},
{
"epoch": 1.8235689207754542,
"grad_norm": 0.6744996905326843,
"learning_rate": 1.0289451092896446e-07,
"loss": 0.1156,
"step": 1867
},
{
"epoch": 1.824545870859411,
"grad_norm": 0.5310499668121338,
"learning_rate": 1.0177462249350456e-07,
"loss": 0.1218,
"step": 1868
},
{
"epoch": 1.8255228209433674,
"grad_norm": 0.5525161623954773,
"learning_rate": 1.006607351621497e-07,
"loss": 0.1124,
"step": 1869
},
{
"epoch": 1.826499771027324,
"grad_norm": 0.6265363693237305,
"learning_rate": 9.955285172219108e-08,
"loss": 0.124,
"step": 1870
},
{
"epoch": 1.8274767211112808,
"grad_norm": 0.9501376152038574,
"learning_rate": 9.845097494589672e-08,
"loss": 0.1201,
"step": 1871
},
{
"epoch": 1.8284536711952373,
"grad_norm": 0.5876163840293884,
"learning_rate": 9.735510759050492e-08,
"loss": 0.1126,
"step": 1872
},
{
"epoch": 1.829430621279194,
"grad_norm": 0.6965773701667786,
"learning_rate": 9.626525239821521e-08,
"loss": 0.1216,
"step": 1873
},
{
"epoch": 1.8304075713631507,
"grad_norm": 1.1193808317184448,
"learning_rate": 9.518141209618415e-08,
"loss": 0.1271,
"step": 1874
},
{
"epoch": 1.8313845214471072,
"grad_norm": 0.5882210731506348,
"learning_rate": 9.410358939651587e-08,
"loss": 0.1161,
"step": 1875
},
{
"epoch": 1.832361471531064,
"grad_norm": 0.6504080891609192,
"learning_rate": 9.303178699625715e-08,
"loss": 0.1223,
"step": 1876
},
{
"epoch": 1.8333384216150206,
"grad_norm": 0.6243442893028259,
"learning_rate": 9.196600757739065e-08,
"loss": 0.1314,
"step": 1877
},
{
"epoch": 1.8343153716989773,
"grad_norm": 0.5928772687911987,
"learning_rate": 9.090625380682699e-08,
"loss": 0.1132,
"step": 1878
},
{
"epoch": 1.835292321782934,
"grad_norm": 0.6002843976020813,
"learning_rate": 8.985252833639907e-08,
"loss": 0.1238,
"step": 1879
},
{
"epoch": 1.8362692718668905,
"grad_norm": 0.47673922777175903,
"learning_rate": 8.880483380285521e-08,
"loss": 0.1194,
"step": 1880
},
{
"epoch": 1.8372462219508472,
"grad_norm": 0.856458306312561,
"learning_rate": 8.776317282785163e-08,
"loss": 0.1127,
"step": 1881
},
{
"epoch": 1.838223172034804,
"grad_norm": 0.48830950260162354,
"learning_rate": 8.672754801794775e-08,
"loss": 0.1117,
"step": 1882
},
{
"epoch": 1.8392001221187604,
"grad_norm": 0.8956071138381958,
"learning_rate": 8.56979619645984e-08,
"loss": 0.1409,
"step": 1883
},
{
"epoch": 1.840177072202717,
"grad_norm": 1.5788103342056274,
"learning_rate": 8.467441724414688e-08,
"loss": 0.1179,
"step": 1884
},
{
"epoch": 1.8411540222866738,
"grad_norm": 0.5313601493835449,
"learning_rate": 8.365691641782025e-08,
"loss": 0.1282,
"step": 1885
},
{
"epoch": 1.8421309723706303,
"grad_norm": 0.4421827793121338,
"learning_rate": 8.264546203172074e-08,
"loss": 0.1157,
"step": 1886
},
{
"epoch": 1.8431079224545872,
"grad_norm": 0.6887156963348389,
"learning_rate": 8.164005661682155e-08,
"loss": 0.1211,
"step": 1887
},
{
"epoch": 1.8440848725385437,
"grad_norm": 1.5370506048202515,
"learning_rate": 8.064070268895912e-08,
"loss": 0.1143,
"step": 1888
},
{
"epoch": 1.8450618226225004,
"grad_norm": 0.5899321436882019,
"learning_rate": 7.964740274882643e-08,
"loss": 0.1245,
"step": 1889
},
{
"epoch": 1.846038772706457,
"grad_norm": 1.5002104043960571,
"learning_rate": 7.866015928196885e-08,
"loss": 0.1236,
"step": 1890
},
{
"epoch": 1.8470157227904136,
"grad_norm": 0.5591606497764587,
"learning_rate": 7.767897475877584e-08,
"loss": 0.1302,
"step": 1891
},
{
"epoch": 1.8479926728743703,
"grad_norm": 0.7197921276092529,
"learning_rate": 7.67038516344748e-08,
"loss": 0.1269,
"step": 1892
},
{
"epoch": 1.848969622958327,
"grad_norm": 0.6165885925292969,
"learning_rate": 7.573479234912723e-08,
"loss": 0.1189,
"step": 1893
},
{
"epoch": 1.8499465730422835,
"grad_norm": 0.5456331372261047,
"learning_rate": 7.47717993276198e-08,
"loss": 0.1225,
"step": 1894
},
{
"epoch": 1.8509235231262404,
"grad_norm": 0.5044867992401123,
"learning_rate": 7.381487497965995e-08,
"loss": 0.1229,
"step": 1895
},
{
"epoch": 1.851900473210197,
"grad_norm": 0.5167362689971924,
"learning_rate": 7.286402169976975e-08,
"loss": 0.1068,
"step": 1896
},
{
"epoch": 1.8528774232941536,
"grad_norm": 0.501384973526001,
"learning_rate": 7.191924186727817e-08,
"loss": 0.1209,
"step": 1897
},
{
"epoch": 1.8538543733781103,
"grad_norm": 0.5291509032249451,
"learning_rate": 7.098053784631853e-08,
"loss": 0.1038,
"step": 1898
},
{
"epoch": 1.8548313234620668,
"grad_norm": 0.9261486530303955,
"learning_rate": 7.004791198581939e-08,
"loss": 0.1573,
"step": 1899
},
{
"epoch": 1.8558082735460235,
"grad_norm": 0.49648037552833557,
"learning_rate": 6.912136661949953e-08,
"loss": 0.1097,
"step": 1900
},
{
"epoch": 1.8567852236299802,
"grad_norm": 0.5355586409568787,
"learning_rate": 6.820090406586322e-08,
"loss": 0.1068,
"step": 1901
},
{
"epoch": 1.8577621737139367,
"grad_norm": 0.5010924339294434,
"learning_rate": 6.728652662819357e-08,
"loss": 0.1199,
"step": 1902
},
{
"epoch": 1.8587391237978934,
"grad_norm": 0.42572537064552307,
"learning_rate": 6.637823659454673e-08,
"loss": 0.1078,
"step": 1903
},
{
"epoch": 1.8597160738818501,
"grad_norm": 0.6323830485343933,
"learning_rate": 6.547603623774601e-08,
"loss": 0.1174,
"step": 1904
},
{
"epoch": 1.8606930239658066,
"grad_norm": 0.4609481394290924,
"learning_rate": 6.457992781537692e-08,
"loss": 0.1064,
"step": 1905
},
{
"epoch": 1.8616699740497635,
"grad_norm": 0.5903739333152771,
"learning_rate": 6.368991356978077e-08,
"loss": 0.1323,
"step": 1906
},
{
"epoch": 1.86264692413372,
"grad_norm": 1.1212151050567627,
"learning_rate": 6.280599572804885e-08,
"loss": 0.1241,
"step": 1907
},
{
"epoch": 1.8636238742176767,
"grad_norm": 0.5500158667564392,
"learning_rate": 6.192817650201883e-08,
"loss": 0.1243,
"step": 1908
},
{
"epoch": 1.8646008243016334,
"grad_norm": 0.5038639903068542,
"learning_rate": 6.105645808826583e-08,
"loss": 0.1158,
"step": 1909
},
{
"epoch": 1.86557777438559,
"grad_norm": 0.47662466764450073,
"learning_rate": 6.019084266810054e-08,
"loss": 0.1091,
"step": 1910
},
{
"epoch": 1.8665547244695466,
"grad_norm": 0.5451299548149109,
"learning_rate": 5.933133240756117e-08,
"loss": 0.1129,
"step": 1911
},
{
"epoch": 1.8675316745535033,
"grad_norm": 0.5382338166236877,
"learning_rate": 5.8477929457408654e-08,
"loss": 0.1179,
"step": 1912
},
{
"epoch": 1.8685086246374598,
"grad_norm": 0.5455739498138428,
"learning_rate": 5.7630635953122846e-08,
"loss": 0.1249,
"step": 1913
},
{
"epoch": 1.8694855747214167,
"grad_norm": 0.45847466588020325,
"learning_rate": 5.6789454014894716e-08,
"loss": 0.107,
"step": 1914
},
{
"epoch": 1.8704625248053732,
"grad_norm": 0.6337384581565857,
"learning_rate": 5.595438574762191e-08,
"loss": 0.1348,
"step": 1915
},
{
"epoch": 1.8714394748893297,
"grad_norm": 0.7348527908325195,
"learning_rate": 5.512543324090541e-08,
"loss": 0.1324,
"step": 1916
},
{
"epoch": 1.8724164249732866,
"grad_norm": 0.5962925553321838,
"learning_rate": 5.430259856904041e-08,
"loss": 0.1282,
"step": 1917
},
{
"epoch": 1.8733933750572431,
"grad_norm": 0.636215329170227,
"learning_rate": 5.348588379101571e-08,
"loss": 0.1233,
"step": 1918
},
{
"epoch": 1.8743703251411998,
"grad_norm": 0.6858799457550049,
"learning_rate": 5.2675290950504335e-08,
"loss": 0.1357,
"step": 1919
},
{
"epoch": 1.8753472752251565,
"grad_norm": 1.9896303415298462,
"learning_rate": 5.1870822075861526e-08,
"loss": 0.1226,
"step": 1920
},
{
"epoch": 1.876324225309113,
"grad_norm": 0.5214887261390686,
"learning_rate": 5.107247918011815e-08,
"loss": 0.1209,
"step": 1921
},
{
"epoch": 1.8773011753930697,
"grad_norm": 0.7137670516967773,
"learning_rate": 5.028026426097565e-08,
"loss": 0.1029,
"step": 1922
},
{
"epoch": 1.8782781254770264,
"grad_norm": 0.4643041491508484,
"learning_rate": 4.949417930080136e-08,
"loss": 0.1113,
"step": 1923
},
{
"epoch": 1.879255075560983,
"grad_norm": 0.5489664077758789,
"learning_rate": 4.871422626662431e-08,
"loss": 0.1134,
"step": 1924
},
{
"epoch": 1.8802320256449399,
"grad_norm": 0.7097074389457703,
"learning_rate": 4.7940407110128595e-08,
"loss": 0.1212,
"step": 1925
},
{
"epoch": 1.8812089757288963,
"grad_norm": 0.6178134679794312,
"learning_rate": 4.717272376765003e-08,
"loss": 0.1379,
"step": 1926
},
{
"epoch": 1.882185925812853,
"grad_norm": 0.724938154220581,
"learning_rate": 4.6411178160170603e-08,
"loss": 0.1036,
"step": 1927
},
{
"epoch": 1.8831628758968098,
"grad_norm": 0.8065431118011475,
"learning_rate": 4.565577219331374e-08,
"loss": 0.1199,
"step": 1928
},
{
"epoch": 1.8841398259807662,
"grad_norm": 0.5854756236076355,
"learning_rate": 4.490650775733962e-08,
"loss": 0.1141,
"step": 1929
},
{
"epoch": 1.885116776064723,
"grad_norm": 0.9782392978668213,
"learning_rate": 4.416338672714044e-08,
"loss": 0.1225,
"step": 1930
},
{
"epoch": 1.8860937261486797,
"grad_norm": 1.1764185428619385,
"learning_rate": 4.34264109622351e-08,
"loss": 0.1234,
"step": 1931
},
{
"epoch": 1.8870706762326361,
"grad_norm": 0.6142870783805847,
"learning_rate": 4.2695582306766236e-08,
"loss": 0.1199,
"step": 1932
},
{
"epoch": 1.8880476263165928,
"grad_norm": 0.9564851522445679,
"learning_rate": 4.1970902589493765e-08,
"loss": 0.1372,
"step": 1933
},
{
"epoch": 1.8890245764005495,
"grad_norm": 0.6659229397773743,
"learning_rate": 4.125237362379103e-08,
"loss": 0.113,
"step": 1934
},
{
"epoch": 1.890001526484506,
"grad_norm": 1.0215439796447754,
"learning_rate": 4.05399972076409e-08,
"loss": 0.1185,
"step": 1935
},
{
"epoch": 1.890978476568463,
"grad_norm": 0.5739313364028931,
"learning_rate": 3.983377512362968e-08,
"loss": 0.1235,
"step": 1936
},
{
"epoch": 1.8919554266524194,
"grad_norm": 0.570742666721344,
"learning_rate": 3.913370913894488e-08,
"loss": 0.1244,
"step": 1937
},
{
"epoch": 1.8929323767363762,
"grad_norm": 0.6435270309448242,
"learning_rate": 3.843980100536826e-08,
"loss": 0.1151,
"step": 1938
},
{
"epoch": 1.8939093268203329,
"grad_norm": 0.9209471940994263,
"learning_rate": 3.7752052459273383e-08,
"loss": 0.1341,
"step": 1939
},
{
"epoch": 1.8948862769042893,
"grad_norm": 0.5585471987724304,
"learning_rate": 3.707046522162055e-08,
"loss": 0.1272,
"step": 1940
},
{
"epoch": 1.895863226988246,
"grad_norm": 0.5234296917915344,
"learning_rate": 3.6395040997952714e-08,
"loss": 0.1099,
"step": 1941
},
{
"epoch": 1.8968401770722028,
"grad_norm": 0.5270965099334717,
"learning_rate": 3.5725781478390674e-08,
"loss": 0.1152,
"step": 1942
},
{
"epoch": 1.8978171271561592,
"grad_norm": 0.6238093972206116,
"learning_rate": 3.506268833762982e-08,
"loss": 0.1248,
"step": 1943
},
{
"epoch": 1.8987940772401162,
"grad_norm": 0.6873887777328491,
"learning_rate": 3.4405763234934274e-08,
"loss": 0.1266,
"step": 1944
},
{
"epoch": 1.8997710273240727,
"grad_norm": 0.5120057463645935,
"learning_rate": 3.375500781413521e-08,
"loss": 0.1205,
"step": 1945
},
{
"epoch": 1.9007479774080291,
"grad_norm": 0.8953158855438232,
"learning_rate": 3.311042370362422e-08,
"loss": 0.1266,
"step": 1946
},
{
"epoch": 1.901724927491986,
"grad_norm": 0.678451418876648,
"learning_rate": 3.2472012516350793e-08,
"loss": 0.1283,
"step": 1947
},
{
"epoch": 1.9027018775759426,
"grad_norm": 0.8391262888908386,
"learning_rate": 3.18397758498179e-08,
"loss": 0.123,
"step": 1948
},
{
"epoch": 1.9036788276598993,
"grad_norm": 0.881302535533905,
"learning_rate": 3.121371528607808e-08,
"loss": 0.1341,
"step": 1949
},
{
"epoch": 1.904655777743856,
"grad_norm": 0.6028062105178833,
"learning_rate": 3.059383239172875e-08,
"loss": 0.1216,
"step": 1950
},
{
"epoch": 1.9056327278278125,
"grad_norm": 0.5131371021270752,
"learning_rate": 2.9980128717909394e-08,
"loss": 0.1082,
"step": 1951
},
{
"epoch": 1.9066096779117692,
"grad_norm": 0.543569028377533,
"learning_rate": 2.937260580029716e-08,
"loss": 0.1237,
"step": 1952
},
{
"epoch": 1.9075866279957259,
"grad_norm": 0.6273193359375,
"learning_rate": 2.8771265159102945e-08,
"loss": 0.1137,
"step": 1953
},
{
"epoch": 1.9085635780796824,
"grad_norm": 0.5713611245155334,
"learning_rate": 2.8176108299067807e-08,
"loss": 0.1144,
"step": 1954
},
{
"epoch": 1.9095405281636393,
"grad_norm": 0.9177828431129456,
"learning_rate": 2.7587136709458228e-08,
"loss": 0.1282,
"step": 1955
},
{
"epoch": 1.9105174782475958,
"grad_norm": 0.526182234287262,
"learning_rate": 2.7004351864064193e-08,
"loss": 0.1172,
"step": 1956
},
{
"epoch": 1.9114944283315525,
"grad_norm": 0.6265659332275391,
"learning_rate": 2.6427755221194728e-08,
"loss": 0.1179,
"step": 1957
},
{
"epoch": 1.9124713784155092,
"grad_norm": 0.4952566921710968,
"learning_rate": 2.5857348223673194e-08,
"loss": 0.1209,
"step": 1958
},
{
"epoch": 1.9134483284994657,
"grad_norm": 0.7257328629493713,
"learning_rate": 2.529313229883451e-08,
"loss": 0.1114,
"step": 1959
},
{
"epoch": 1.9144252785834224,
"grad_norm": 0.9025228023529053,
"learning_rate": 2.4735108858522927e-08,
"loss": 0.119,
"step": 1960
},
{
"epoch": 1.915402228667379,
"grad_norm": 0.5474231839179993,
"learning_rate": 2.418327929908537e-08,
"loss": 0.1298,
"step": 1961
},
{
"epoch": 1.9163791787513356,
"grad_norm": 0.6201128363609314,
"learning_rate": 2.3637645001371167e-08,
"loss": 0.1353,
"step": 1962
},
{
"epoch": 1.9173561288352923,
"grad_norm": 0.562218964099884,
"learning_rate": 2.3098207330726485e-08,
"loss": 0.119,
"step": 1963
},
{
"epoch": 1.918333078919249,
"grad_norm": 1.3331130743026733,
"learning_rate": 2.2564967636991563e-08,
"loss": 0.1112,
"step": 1964
},
{
"epoch": 1.9193100290032055,
"grad_norm": 0.49711117148399353,
"learning_rate": 2.2037927254497936e-08,
"loss": 0.1142,
"step": 1965
},
{
"epoch": 1.9202869790871624,
"grad_norm": 0.4408424496650696,
"learning_rate": 2.1517087502063994e-08,
"loss": 0.1115,
"step": 1966
},
{
"epoch": 1.9212639291711189,
"grad_norm": 0.6409486532211304,
"learning_rate": 2.1002449682992198e-08,
"loss": 0.1297,
"step": 1967
},
{
"epoch": 1.9222408792550756,
"grad_norm": 0.6932352185249329,
"learning_rate": 2.0494015085066877e-08,
"loss": 0.11,
"step": 1968
},
{
"epoch": 1.9232178293390323,
"grad_norm": 0.5591157078742981,
"learning_rate": 1.999178498054838e-08,
"loss": 0.1131,
"step": 1969
},
{
"epoch": 1.9241947794229888,
"grad_norm": 0.5266435146331787,
"learning_rate": 1.9495760626172532e-08,
"loss": 0.1179,
"step": 1970
},
{
"epoch": 1.9251717295069455,
"grad_norm": 0.8860810399055481,
"learning_rate": 1.9005943263146197e-08,
"loss": 0.1363,
"step": 1971
},
{
"epoch": 1.9261486795909022,
"grad_norm": 0.6715410351753235,
"learning_rate": 1.852233411714449e-08,
"loss": 0.1205,
"step": 1972
},
{
"epoch": 1.9271256296748587,
"grad_norm": 1.7903392314910889,
"learning_rate": 1.8044934398307456e-08,
"loss": 0.1436,
"step": 1973
},
{
"epoch": 1.9281025797588156,
"grad_norm": 0.6085479855537415,
"learning_rate": 1.757374530123729e-08,
"loss": 0.1168,
"step": 1974
},
{
"epoch": 1.929079529842772,
"grad_norm": 0.6084596514701843,
"learning_rate": 1.710876800499556e-08,
"loss": 0.1257,
"step": 1975
},
{
"epoch": 1.9300564799267288,
"grad_norm": 0.5929890871047974,
"learning_rate": 1.6650003673099603e-08,
"loss": 0.106,
"step": 1976
},
{
"epoch": 1.9310334300106855,
"grad_norm": 0.7909138798713684,
"learning_rate": 1.619745345352003e-08,
"loss": 0.1253,
"step": 1977
},
{
"epoch": 1.932010380094642,
"grad_norm": 0.5139659643173218,
"learning_rate": 1.5751118478678218e-08,
"loss": 0.1228,
"step": 1978
},
{
"epoch": 1.9329873301785987,
"grad_norm": 0.41263025999069214,
"learning_rate": 1.5310999865442434e-08,
"loss": 0.1169,
"step": 1979
},
{
"epoch": 1.9339642802625554,
"grad_norm": 0.6405431032180786,
"learning_rate": 1.4877098715126159e-08,
"loss": 0.1284,
"step": 1980
},
{
"epoch": 1.9349412303465119,
"grad_norm": 0.6059233546257019,
"learning_rate": 1.444941611348477e-08,
"loss": 0.1122,
"step": 1981
},
{
"epoch": 1.9359181804304686,
"grad_norm": 0.5469845533370972,
"learning_rate": 1.4027953130713034e-08,
"loss": 0.1135,
"step": 1982
},
{
"epoch": 1.9368951305144253,
"grad_norm": 1.5313292741775513,
"learning_rate": 1.3612710821441777e-08,
"loss": 0.124,
"step": 1983
},
{
"epoch": 1.9378720805983818,
"grad_norm": 0.7973529696464539,
"learning_rate": 1.3203690224736222e-08,
"loss": 0.1246,
"step": 1984
},
{
"epoch": 1.9388490306823387,
"grad_norm": 0.4730254113674164,
"learning_rate": 1.2800892364092932e-08,
"loss": 0.1252,
"step": 1985
},
{
"epoch": 1.9398259807662952,
"grad_norm": 0.4545423984527588,
"learning_rate": 1.2404318247437042e-08,
"loss": 0.1272,
"step": 1986
},
{
"epoch": 1.940802930850252,
"grad_norm": 0.4167299270629883,
"learning_rate": 1.2013968867120308e-08,
"loss": 0.1028,
"step": 1987
},
{
"epoch": 1.9417798809342086,
"grad_norm": 0.4157821834087372,
"learning_rate": 1.16298451999175e-08,
"loss": 0.1257,
"step": 1988
},
{
"epoch": 1.942756831018165,
"grad_norm": 0.6711307764053345,
"learning_rate": 1.1251948207025576e-08,
"loss": 0.1199,
"step": 1989
},
{
"epoch": 1.9437337811021218,
"grad_norm": 0.9208105802536011,
"learning_rate": 1.0880278834060065e-08,
"loss": 0.1275,
"step": 1990
},
{
"epoch": 1.9447107311860785,
"grad_norm": 0.6180605292320251,
"learning_rate": 1.051483801105313e-08,
"loss": 0.1161,
"step": 1991
},
{
"epoch": 1.945687681270035,
"grad_norm": 1.1046503782272339,
"learning_rate": 1.0155626652451068e-08,
"loss": 0.1168,
"step": 1992
},
{
"epoch": 1.9466646313539917,
"grad_norm": 0.6219564080238342,
"learning_rate": 9.802645657112086e-09,
"loss": 0.1113,
"step": 1993
},
{
"epoch": 1.9476415814379484,
"grad_norm": 0.6671769618988037,
"learning_rate": 9.455895908304091e-09,
"loss": 0.1131,
"step": 1994
},
{
"epoch": 1.948618531521905,
"grad_norm": 0.7161427140235901,
"learning_rate": 9.115378273702735e-09,
"loss": 0.1341,
"step": 1995
},
{
"epoch": 1.9495954816058618,
"grad_norm": 0.4451122581958771,
"learning_rate": 8.781093605388924e-09,
"loss": 0.0981,
"step": 1996
},
{
"epoch": 1.9505724316898183,
"grad_norm": 0.557096004486084,
"learning_rate": 8.45304273984604e-09,
"loss": 0.1276,
"step": 1997
},
{
"epoch": 1.951549381773775,
"grad_norm": 0.4758124351501465,
"learning_rate": 8.131226497959944e-09,
"loss": 0.1243,
"step": 1998
},
{
"epoch": 1.9525263318577317,
"grad_norm": 0.6417310237884521,
"learning_rate": 7.81564568501425e-09,
"loss": 0.1242,
"step": 1999
},
{
"epoch": 1.9535032819416882,
"grad_norm": 0.8300948739051819,
"learning_rate": 7.506301090690338e-09,
"loss": 0.1238,
"step": 2000
},
{
"epoch": 1.954480232025645,
"grad_norm": 0.6566423773765564,
"learning_rate": 7.203193489064009e-09,
"loss": 0.1133,
"step": 2001
},
{
"epoch": 1.9554571821096016,
"grad_norm": 0.5134234428405762,
"learning_rate": 6.906323638604662e-09,
"loss": 0.1148,
"step": 2002
},
{
"epoch": 1.956434132193558,
"grad_norm": 0.5473626852035522,
"learning_rate": 6.61569228217307e-09,
"loss": 0.1297,
"step": 2003
},
{
"epoch": 1.957411082277515,
"grad_norm": 0.6833441853523254,
"learning_rate": 6.331300147019159e-09,
"loss": 0.0996,
"step": 2004
},
{
"epoch": 1.9583880323614715,
"grad_norm": 0.5028896927833557,
"learning_rate": 6.053147944779514e-09,
"loss": 0.1063,
"step": 2005
},
{
"epoch": 1.9593649824454282,
"grad_norm": 0.4890092611312866,
"learning_rate": 5.781236371478205e-09,
"loss": 0.1162,
"step": 2006
},
{
"epoch": 1.960341932529385,
"grad_norm": 0.6293298006057739,
"learning_rate": 5.515566107521797e-09,
"loss": 0.1142,
"step": 2007
},
{
"epoch": 1.9613188826133414,
"grad_norm": 0.5001157522201538,
"learning_rate": 5.2561378176999025e-09,
"loss": 0.1159,
"step": 2008
},
{
"epoch": 1.9622958326972981,
"grad_norm": 0.6762978434562683,
"learning_rate": 5.002952151182405e-09,
"loss": 0.1234,
"step": 2009
},
{
"epoch": 1.9632727827812548,
"grad_norm": 0.5457140803337097,
"learning_rate": 4.7560097415189076e-09,
"loss": 0.1323,
"step": 2010
},
{
"epoch": 1.9642497328652113,
"grad_norm": 0.5183831453323364,
"learning_rate": 4.515311206635397e-09,
"loss": 0.1201,
"step": 2011
},
{
"epoch": 1.965226682949168,
"grad_norm": 1.0143280029296875,
"learning_rate": 4.280857148834528e-09,
"loss": 0.125,
"step": 2012
},
{
"epoch": 1.9662036330331247,
"grad_norm": 0.6653719544410706,
"learning_rate": 4.052648154793393e-09,
"loss": 0.1196,
"step": 2013
},
{
"epoch": 1.9671805831170812,
"grad_norm": 0.6077477335929871,
"learning_rate": 3.830684795561035e-09,
"loss": 0.1146,
"step": 2014
},
{
"epoch": 1.9681575332010381,
"grad_norm": 0.6227024793624878,
"learning_rate": 3.614967626559551e-09,
"loss": 0.1431,
"step": 2015
},
{
"epoch": 1.9691344832849946,
"grad_norm": 0.5604516863822937,
"learning_rate": 3.4054971875793763e-09,
"loss": 0.1207,
"step": 2016
},
{
"epoch": 1.9701114333689513,
"grad_norm": 0.6133773922920227,
"learning_rate": 3.2022740027809473e-09,
"loss": 0.121,
"step": 2017
},
{
"epoch": 1.971088383452908,
"grad_norm": 0.5954017043113708,
"learning_rate": 3.005298580691929e-09,
"loss": 0.1143,
"step": 2018
},
{
"epoch": 1.9720653335368645,
"grad_norm": 0.549052894115448,
"learning_rate": 2.8145714142055467e-09,
"loss": 0.1292,
"step": 2019
},
{
"epoch": 1.9730422836208212,
"grad_norm": 0.550966203212738,
"learning_rate": 2.63009298058059e-09,
"loss": 0.1117,
"step": 2020
},
{
"epoch": 1.974019233704778,
"grad_norm": 0.736648678779602,
"learning_rate": 2.4518637414391867e-09,
"loss": 0.1072,
"step": 2021
},
{
"epoch": 1.9749961837887344,
"grad_norm": 0.49929293990135193,
"learning_rate": 2.279884142766531e-09,
"loss": 0.1207,
"step": 2022
},
{
"epoch": 1.9759731338726914,
"grad_norm": 0.4560934007167816,
"learning_rate": 2.1141546149086588e-09,
"loss": 0.1094,
"step": 2023
},
{
"epoch": 1.9769500839566478,
"grad_norm": 0.5404055118560791,
"learning_rate": 1.954675572572451e-09,
"loss": 0.1052,
"step": 2024
},
{
"epoch": 1.9779270340406043,
"grad_norm": 0.587166965007782,
"learning_rate": 1.801447414823687e-09,
"loss": 0.1282,
"step": 2025
},
{
"epoch": 1.9789039841245613,
"grad_norm": 0.7316482067108154,
"learning_rate": 1.6544705250870485e-09,
"loss": 0.1269,
"step": 2026
},
{
"epoch": 1.9798809342085177,
"grad_norm": 0.47679123282432556,
"learning_rate": 1.5137452711441735e-09,
"loss": 0.1181,
"step": 2027
},
{
"epoch": 1.9808578842924744,
"grad_norm": 0.4444611966609955,
"learning_rate": 1.3792720051333807e-09,
"loss": 0.1197,
"step": 2028
},
{
"epoch": 1.9818348343764312,
"grad_norm": 0.6291835904121399,
"learning_rate": 1.2510510635482809e-09,
"loss": 0.1162,
"step": 2029
},
{
"epoch": 1.9828117844603876,
"grad_norm": 0.42497414350509644,
"learning_rate": 1.1290827672380544e-09,
"loss": 0.1072,
"step": 2030
},
{
"epoch": 1.9837887345443443,
"grad_norm": 0.8148401975631714,
"learning_rate": 1.0133674214043987e-09,
"loss": 0.1191,
"step": 2031
},
{
"epoch": 1.984765684628301,
"grad_norm": 3.020845413208008,
"learning_rate": 9.03905315603748e-10,
"loss": 0.1207,
"step": 2032
},
{
"epoch": 1.9857426347122575,
"grad_norm": 0.47406142950057983,
"learning_rate": 8.006967237439433e-10,
"loss": 0.1132,
"step": 2033
},
{
"epoch": 1.9867195847962145,
"grad_norm": 0.5028852224349976,
"learning_rate": 7.03741904084787e-10,
"loss": 0.1085,
"step": 2034
},
{
"epoch": 1.987696534880171,
"grad_norm": 0.49047085642814636,
"learning_rate": 6.130410992377656e-10,
"loss": 0.1286,
"step": 2035
},
{
"epoch": 1.9886734849641277,
"grad_norm": 0.5054574608802795,
"learning_rate": 5.285945361638289e-10,
"loss": 0.1291,
"step": 2036
},
{
"epoch": 1.9896504350480844,
"grad_norm": 0.5029098987579346,
"learning_rate": 4.5040242617505614e-10,
"loss": 0.1217,
"step": 2037
},
{
"epoch": 1.9906273851320408,
"grad_norm": 0.5311554670333862,
"learning_rate": 3.7846496493187945e-10,
"loss": 0.1121,
"step": 2038
},
{
"epoch": 1.9916043352159976,
"grad_norm": 0.5074602365493774,
"learning_rate": 3.127823324444723e-10,
"loss": 0.1209,
"step": 2039
},
{
"epoch": 1.9925812852999543,
"grad_norm": 0.5317457318305969,
"learning_rate": 2.533546930713615e-10,
"loss": 0.1083,
"step": 2040
},
{
"epoch": 1.9935582353839107,
"grad_norm": 1.2235394716262817,
"learning_rate": 2.0018219551831696e-10,
"loss": 0.1229,
"step": 2041
},
{
"epoch": 1.9945351854678675,
"grad_norm": 0.7713088989257812,
"learning_rate": 1.532649728397395e-10,
"loss": 0.1131,
"step": 2042
},
{
"epoch": 1.9955121355518242,
"grad_norm": 0.5145773887634277,
"learning_rate": 1.126031424372731e-10,
"loss": 0.1222,
"step": 2043
},
{
"epoch": 1.9964890856357806,
"grad_norm": 0.9036497473716736,
"learning_rate": 7.819680605924974e-11,
"loss": 0.1263,
"step": 2044
},
{
"epoch": 1.9974660357197376,
"grad_norm": 0.5623131990432739,
"learning_rate": 5.004604980096695e-11,
"loss": 0.1236,
"step": 2045
},
{
"epoch": 1.998442985803694,
"grad_norm": 0.5780706405639648,
"learning_rate": 2.8150944104410327e-11,
"loss": 0.1133,
"step": 2046
},
{
"epoch": 1.9994199358876508,
"grad_norm": 0.5285246968269348,
"learning_rate": 1.251154375797592e-11,
"loss": 0.1165,
"step": 2047
},
{
"epoch": 2.0,
"grad_norm": 1.0938704013824463,
"learning_rate": 3.1278878961926717e-12,
"loss": 0.114,
"step": 2048
},
{
"epoch": 2.0,
"step": 2048,
"total_flos": 6.863462328004772e+17,
"train_loss": 0.16797507342562312,
"train_runtime": 21551.3712,
"train_samples_per_second": 24.318,
"train_steps_per_second": 0.095
}
],
"logging_steps": 1.0,
"max_steps": 2048,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 2000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6.863462328004772e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}