prxy5608's picture
Training in progress, step 200, checkpoint
40e428e verified
{
"best_metric": 1.2549843788146973,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 1.21580547112462,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0060790273556231,
"grad_norm": 1.7974525690078735,
"learning_rate": 1e-05,
"loss": 1.9276,
"step": 1
},
{
"epoch": 0.0060790273556231,
"eval_loss": 2.0459976196289062,
"eval_runtime": 30.7741,
"eval_samples_per_second": 9.001,
"eval_steps_per_second": 2.275,
"step": 1
},
{
"epoch": 0.0121580547112462,
"grad_norm": 1.986690640449524,
"learning_rate": 2e-05,
"loss": 1.9162,
"step": 2
},
{
"epoch": 0.0182370820668693,
"grad_norm": 2.001444101333618,
"learning_rate": 3e-05,
"loss": 2.0916,
"step": 3
},
{
"epoch": 0.0243161094224924,
"grad_norm": 1.7187578678131104,
"learning_rate": 4e-05,
"loss": 2.1186,
"step": 4
},
{
"epoch": 0.030395136778115502,
"grad_norm": 1.712262511253357,
"learning_rate": 5e-05,
"loss": 2.0101,
"step": 5
},
{
"epoch": 0.0364741641337386,
"grad_norm": 1.5926772356033325,
"learning_rate": 6e-05,
"loss": 1.8638,
"step": 6
},
{
"epoch": 0.0425531914893617,
"grad_norm": 1.5831444263458252,
"learning_rate": 7e-05,
"loss": 1.7983,
"step": 7
},
{
"epoch": 0.0486322188449848,
"grad_norm": 1.9558838605880737,
"learning_rate": 8e-05,
"loss": 1.648,
"step": 8
},
{
"epoch": 0.0547112462006079,
"grad_norm": 2.270124912261963,
"learning_rate": 9e-05,
"loss": 1.7939,
"step": 9
},
{
"epoch": 0.060790273556231005,
"grad_norm": 1.7760719060897827,
"learning_rate": 0.0001,
"loss": 1.6603,
"step": 10
},
{
"epoch": 0.0668693009118541,
"grad_norm": 1.6471532583236694,
"learning_rate": 9.999316524962345e-05,
"loss": 1.5516,
"step": 11
},
{
"epoch": 0.0729483282674772,
"grad_norm": 1.803841233253479,
"learning_rate": 9.997266286704631e-05,
"loss": 1.6288,
"step": 12
},
{
"epoch": 0.0790273556231003,
"grad_norm": 1.6245392560958862,
"learning_rate": 9.993849845741524e-05,
"loss": 1.4918,
"step": 13
},
{
"epoch": 0.0851063829787234,
"grad_norm": 1.519576072692871,
"learning_rate": 9.989068136093873e-05,
"loss": 1.5378,
"step": 14
},
{
"epoch": 0.0911854103343465,
"grad_norm": 1.5418260097503662,
"learning_rate": 9.98292246503335e-05,
"loss": 1.4792,
"step": 15
},
{
"epoch": 0.0972644376899696,
"grad_norm": 1.6473251581192017,
"learning_rate": 9.975414512725057e-05,
"loss": 1.6151,
"step": 16
},
{
"epoch": 0.1033434650455927,
"grad_norm": 1.3600949048995972,
"learning_rate": 9.966546331768191e-05,
"loss": 1.6516,
"step": 17
},
{
"epoch": 0.1094224924012158,
"grad_norm": 1.528100609779358,
"learning_rate": 9.956320346634876e-05,
"loss": 1.6456,
"step": 18
},
{
"epoch": 0.11550151975683891,
"grad_norm": 1.5808520317077637,
"learning_rate": 9.944739353007344e-05,
"loss": 1.8078,
"step": 19
},
{
"epoch": 0.12158054711246201,
"grad_norm": 1.488606572151184,
"learning_rate": 9.931806517013612e-05,
"loss": 1.516,
"step": 20
},
{
"epoch": 0.1276595744680851,
"grad_norm": 1.6215261220932007,
"learning_rate": 9.917525374361912e-05,
"loss": 1.6613,
"step": 21
},
{
"epoch": 0.1337386018237082,
"grad_norm": 1.5725288391113281,
"learning_rate": 9.901899829374047e-05,
"loss": 1.7446,
"step": 22
},
{
"epoch": 0.1398176291793313,
"grad_norm": 1.3139268159866333,
"learning_rate": 9.884934153917997e-05,
"loss": 1.5082,
"step": 23
},
{
"epoch": 0.1458966565349544,
"grad_norm": 1.2928907871246338,
"learning_rate": 9.86663298624003e-05,
"loss": 1.2943,
"step": 24
},
{
"epoch": 0.1519756838905775,
"grad_norm": 1.195099949836731,
"learning_rate": 9.847001329696653e-05,
"loss": 1.3026,
"step": 25
},
{
"epoch": 0.1580547112462006,
"grad_norm": 1.235960841178894,
"learning_rate": 9.826044551386744e-05,
"loss": 1.3095,
"step": 26
},
{
"epoch": 0.1641337386018237,
"grad_norm": 1.1664656400680542,
"learning_rate": 9.803768380684242e-05,
"loss": 1.259,
"step": 27
},
{
"epoch": 0.1702127659574468,
"grad_norm": 1.1557590961456299,
"learning_rate": 9.780178907671789e-05,
"loss": 1.2139,
"step": 28
},
{
"epoch": 0.1762917933130699,
"grad_norm": 1.283929467201233,
"learning_rate": 9.755282581475769e-05,
"loss": 1.2244,
"step": 29
},
{
"epoch": 0.182370820668693,
"grad_norm": 1.0923078060150146,
"learning_rate": 9.729086208503174e-05,
"loss": 1.2023,
"step": 30
},
{
"epoch": 0.1884498480243161,
"grad_norm": 1.1785950660705566,
"learning_rate": 9.701596950580806e-05,
"loss": 1.2532,
"step": 31
},
{
"epoch": 0.1945288753799392,
"grad_norm": 1.342201590538025,
"learning_rate": 9.672822322997305e-05,
"loss": 1.2825,
"step": 32
},
{
"epoch": 0.2006079027355623,
"grad_norm": 1.2498522996902466,
"learning_rate": 9.642770192448536e-05,
"loss": 1.28,
"step": 33
},
{
"epoch": 0.2066869300911854,
"grad_norm": 1.133314609527588,
"learning_rate": 9.611448774886924e-05,
"loss": 1.183,
"step": 34
},
{
"epoch": 0.2127659574468085,
"grad_norm": 1.1778188943862915,
"learning_rate": 9.578866633275288e-05,
"loss": 1.2689,
"step": 35
},
{
"epoch": 0.2188449848024316,
"grad_norm": 1.0845565795898438,
"learning_rate": 9.545032675245813e-05,
"loss": 1.1862,
"step": 36
},
{
"epoch": 0.22492401215805471,
"grad_norm": 1.1704156398773193,
"learning_rate": 9.509956150664796e-05,
"loss": 1.1542,
"step": 37
},
{
"epoch": 0.23100303951367782,
"grad_norm": 1.1439368724822998,
"learning_rate": 9.473646649103818e-05,
"loss": 1.3128,
"step": 38
},
{
"epoch": 0.23708206686930092,
"grad_norm": 1.2115826606750488,
"learning_rate": 9.43611409721806e-05,
"loss": 1.326,
"step": 39
},
{
"epoch": 0.24316109422492402,
"grad_norm": 1.1417663097381592,
"learning_rate": 9.397368756032445e-05,
"loss": 1.2613,
"step": 40
},
{
"epoch": 0.24924012158054712,
"grad_norm": 1.2824070453643799,
"learning_rate": 9.357421218136386e-05,
"loss": 1.3291,
"step": 41
},
{
"epoch": 0.2553191489361702,
"grad_norm": 0.9674400687217712,
"learning_rate": 9.316282404787871e-05,
"loss": 1.5845,
"step": 42
},
{
"epoch": 0.2613981762917933,
"grad_norm": 1.0582726001739502,
"learning_rate": 9.273963562927695e-05,
"loss": 1.4711,
"step": 43
},
{
"epoch": 0.2674772036474164,
"grad_norm": 1.121254563331604,
"learning_rate": 9.230476262104677e-05,
"loss": 1.5071,
"step": 44
},
{
"epoch": 0.2735562310030395,
"grad_norm": 1.112056016921997,
"learning_rate": 9.185832391312644e-05,
"loss": 1.4977,
"step": 45
},
{
"epoch": 0.2796352583586626,
"grad_norm": 1.1874089241027832,
"learning_rate": 9.140044155740101e-05,
"loss": 1.6197,
"step": 46
},
{
"epoch": 0.2857142857142857,
"grad_norm": 1.1099293231964111,
"learning_rate": 9.093124073433463e-05,
"loss": 1.4242,
"step": 47
},
{
"epoch": 0.2917933130699088,
"grad_norm": 1.0958565473556519,
"learning_rate": 9.045084971874738e-05,
"loss": 1.3368,
"step": 48
},
{
"epoch": 0.2978723404255319,
"grad_norm": 1.2167139053344727,
"learning_rate": 8.995939984474624e-05,
"loss": 1.391,
"step": 49
},
{
"epoch": 0.303951367781155,
"grad_norm": 2.04726505279541,
"learning_rate": 8.945702546981969e-05,
"loss": 1.4067,
"step": 50
},
{
"epoch": 0.303951367781155,
"eval_loss": 1.3686360120773315,
"eval_runtime": 31.3208,
"eval_samples_per_second": 8.844,
"eval_steps_per_second": 2.235,
"step": 50
},
{
"epoch": 0.3100303951367781,
"grad_norm": 2.002873420715332,
"learning_rate": 8.894386393810563e-05,
"loss": 1.449,
"step": 51
},
{
"epoch": 0.3161094224924012,
"grad_norm": 1.227084755897522,
"learning_rate": 8.842005554284296e-05,
"loss": 1.264,
"step": 52
},
{
"epoch": 0.3221884498480243,
"grad_norm": 1.2173935174942017,
"learning_rate": 8.788574348801675e-05,
"loss": 1.4194,
"step": 53
},
{
"epoch": 0.3282674772036474,
"grad_norm": 1.3812220096588135,
"learning_rate": 8.73410738492077e-05,
"loss": 1.4863,
"step": 54
},
{
"epoch": 0.3343465045592705,
"grad_norm": 1.3314387798309326,
"learning_rate": 8.678619553365659e-05,
"loss": 1.4746,
"step": 55
},
{
"epoch": 0.3404255319148936,
"grad_norm": 1.4191101789474487,
"learning_rate": 8.622126023955446e-05,
"loss": 1.5348,
"step": 56
},
{
"epoch": 0.3465045592705167,
"grad_norm": 1.325097918510437,
"learning_rate": 8.564642241456986e-05,
"loss": 1.5035,
"step": 57
},
{
"epoch": 0.3525835866261398,
"grad_norm": 1.4912933111190796,
"learning_rate": 8.506183921362443e-05,
"loss": 1.716,
"step": 58
},
{
"epoch": 0.3586626139817629,
"grad_norm": 1.5581443309783936,
"learning_rate": 8.44676704559283e-05,
"loss": 1.5603,
"step": 59
},
{
"epoch": 0.364741641337386,
"grad_norm": 1.4817399978637695,
"learning_rate": 8.386407858128706e-05,
"loss": 1.6258,
"step": 60
},
{
"epoch": 0.3708206686930091,
"grad_norm": 1.4714860916137695,
"learning_rate": 8.32512286056924e-05,
"loss": 1.4344,
"step": 61
},
{
"epoch": 0.3768996960486322,
"grad_norm": 1.2636381387710571,
"learning_rate": 8.262928807620843e-05,
"loss": 1.1586,
"step": 62
},
{
"epoch": 0.3829787234042553,
"grad_norm": 1.4186855554580688,
"learning_rate": 8.199842702516583e-05,
"loss": 1.1606,
"step": 63
},
{
"epoch": 0.3890577507598784,
"grad_norm": 1.0884548425674438,
"learning_rate": 8.135881792367686e-05,
"loss": 1.1811,
"step": 64
},
{
"epoch": 0.3951367781155015,
"grad_norm": 1.1721493005752563,
"learning_rate": 8.07106356344834e-05,
"loss": 1.2581,
"step": 65
},
{
"epoch": 0.4012158054711246,
"grad_norm": 1.2545545101165771,
"learning_rate": 8.005405736415126e-05,
"loss": 1.4144,
"step": 66
},
{
"epoch": 0.4072948328267477,
"grad_norm": 1.2253715991973877,
"learning_rate": 7.938926261462366e-05,
"loss": 1.1576,
"step": 67
},
{
"epoch": 0.4133738601823708,
"grad_norm": 1.1156668663024902,
"learning_rate": 7.871643313414718e-05,
"loss": 1.1589,
"step": 68
},
{
"epoch": 0.4194528875379939,
"grad_norm": 1.2024389505386353,
"learning_rate": 7.803575286758364e-05,
"loss": 1.3169,
"step": 69
},
{
"epoch": 0.425531914893617,
"grad_norm": 1.0439965724945068,
"learning_rate": 7.734740790612136e-05,
"loss": 1.2513,
"step": 70
},
{
"epoch": 0.4316109422492401,
"grad_norm": 1.163286805152893,
"learning_rate": 7.66515864363997e-05,
"loss": 1.2245,
"step": 71
},
{
"epoch": 0.4376899696048632,
"grad_norm": 1.049737572669983,
"learning_rate": 7.594847868906076e-05,
"loss": 1.1199,
"step": 72
},
{
"epoch": 0.44376899696048633,
"grad_norm": 1.2103822231292725,
"learning_rate": 7.52382768867422e-05,
"loss": 1.2207,
"step": 73
},
{
"epoch": 0.44984802431610943,
"grad_norm": 1.0623281002044678,
"learning_rate": 7.452117519152542e-05,
"loss": 1.2003,
"step": 74
},
{
"epoch": 0.45592705167173253,
"grad_norm": 1.0393519401550293,
"learning_rate": 7.379736965185368e-05,
"loss": 1.117,
"step": 75
},
{
"epoch": 0.46200607902735563,
"grad_norm": 1.1172492504119873,
"learning_rate": 7.30670581489344e-05,
"loss": 1.2259,
"step": 76
},
{
"epoch": 0.46808510638297873,
"grad_norm": 1.2103140354156494,
"learning_rate": 7.233044034264034e-05,
"loss": 1.222,
"step": 77
},
{
"epoch": 0.47416413373860183,
"grad_norm": 1.1494786739349365,
"learning_rate": 7.158771761692464e-05,
"loss": 1.2799,
"step": 78
},
{
"epoch": 0.48024316109422494,
"grad_norm": 1.22427237033844,
"learning_rate": 7.083909302476453e-05,
"loss": 1.1671,
"step": 79
},
{
"epoch": 0.48632218844984804,
"grad_norm": 1.0408166646957397,
"learning_rate": 7.008477123264848e-05,
"loss": 1.2081,
"step": 80
},
{
"epoch": 0.49240121580547114,
"grad_norm": 1.0972356796264648,
"learning_rate": 6.932495846462261e-05,
"loss": 1.2123,
"step": 81
},
{
"epoch": 0.49848024316109424,
"grad_norm": 1.2304025888442993,
"learning_rate": 6.855986244591104e-05,
"loss": 1.3152,
"step": 82
},
{
"epoch": 0.5045592705167173,
"grad_norm": 0.9390163421630859,
"learning_rate": 6.778969234612584e-05,
"loss": 1.4002,
"step": 83
},
{
"epoch": 0.5106382978723404,
"grad_norm": 1.102756381034851,
"learning_rate": 6.701465872208216e-05,
"loss": 1.5442,
"step": 84
},
{
"epoch": 0.5167173252279635,
"grad_norm": 0.9713530540466309,
"learning_rate": 6.623497346023418e-05,
"loss": 1.4247,
"step": 85
},
{
"epoch": 0.5227963525835866,
"grad_norm": 1.088732123374939,
"learning_rate": 6.545084971874738e-05,
"loss": 1.4056,
"step": 86
},
{
"epoch": 0.5288753799392097,
"grad_norm": 1.030799150466919,
"learning_rate": 6.466250186922325e-05,
"loss": 1.4886,
"step": 87
},
{
"epoch": 0.5349544072948328,
"grad_norm": 1.0953729152679443,
"learning_rate": 6.387014543809223e-05,
"loss": 1.3793,
"step": 88
},
{
"epoch": 0.541033434650456,
"grad_norm": 1.212327241897583,
"learning_rate": 6.307399704769099e-05,
"loss": 1.2817,
"step": 89
},
{
"epoch": 0.547112462006079,
"grad_norm": 1.0665987730026245,
"learning_rate": 6.227427435703997e-05,
"loss": 1.2388,
"step": 90
},
{
"epoch": 0.5531914893617021,
"grad_norm": 1.158036470413208,
"learning_rate": 6.147119600233758e-05,
"loss": 1.2811,
"step": 91
},
{
"epoch": 0.5592705167173252,
"grad_norm": 1.1704466342926025,
"learning_rate": 6.066498153718735e-05,
"loss": 1.21,
"step": 92
},
{
"epoch": 0.5653495440729484,
"grad_norm": 1.1668264865875244,
"learning_rate": 5.985585137257401e-05,
"loss": 1.3044,
"step": 93
},
{
"epoch": 0.5714285714285714,
"grad_norm": 1.4689115285873413,
"learning_rate": 5.90440267166055e-05,
"loss": 1.2798,
"step": 94
},
{
"epoch": 0.5775075987841946,
"grad_norm": 1.411950945854187,
"learning_rate": 5.8229729514036705e-05,
"loss": 1.3352,
"step": 95
},
{
"epoch": 0.5835866261398176,
"grad_norm": 1.4480981826782227,
"learning_rate": 5.74131823855921e-05,
"loss": 1.4022,
"step": 96
},
{
"epoch": 0.5896656534954408,
"grad_norm": 1.184134840965271,
"learning_rate": 5.6594608567103456e-05,
"loss": 1.2582,
"step": 97
},
{
"epoch": 0.5957446808510638,
"grad_norm": 1.4779952764511108,
"learning_rate": 5.577423184847932e-05,
"loss": 1.5624,
"step": 98
},
{
"epoch": 0.601823708206687,
"grad_norm": 1.3613331317901611,
"learning_rate": 5.495227651252315e-05,
"loss": 1.2237,
"step": 99
},
{
"epoch": 0.60790273556231,
"grad_norm": 1.5509363412857056,
"learning_rate": 5.4128967273616625e-05,
"loss": 1.4074,
"step": 100
},
{
"epoch": 0.60790273556231,
"eval_loss": 1.2909982204437256,
"eval_runtime": 31.3491,
"eval_samples_per_second": 8.836,
"eval_steps_per_second": 2.233,
"step": 100
},
{
"epoch": 0.6139817629179332,
"grad_norm": 1.3716071844100952,
"learning_rate": 5.330452921628497e-05,
"loss": 1.4074,
"step": 101
},
{
"epoch": 0.6200607902735562,
"grad_norm": 1.7256039381027222,
"learning_rate": 5.247918773366112e-05,
"loss": 1.4873,
"step": 102
},
{
"epoch": 0.6261398176291794,
"grad_norm": 1.1996338367462158,
"learning_rate": 5.165316846586541e-05,
"loss": 1.2525,
"step": 103
},
{
"epoch": 0.6322188449848024,
"grad_norm": 1.0866903066635132,
"learning_rate": 5.0826697238317935e-05,
"loss": 1.1553,
"step": 104
},
{
"epoch": 0.6382978723404256,
"grad_norm": 1.0325300693511963,
"learning_rate": 5e-05,
"loss": 1.2234,
"step": 105
},
{
"epoch": 0.6443768996960486,
"grad_norm": 1.040966272354126,
"learning_rate": 4.917330276168208e-05,
"loss": 1.2523,
"step": 106
},
{
"epoch": 0.6504559270516718,
"grad_norm": 1.0111114978790283,
"learning_rate": 4.834683153413459e-05,
"loss": 1.128,
"step": 107
},
{
"epoch": 0.6565349544072948,
"grad_norm": 1.0495020151138306,
"learning_rate": 4.7520812266338885e-05,
"loss": 1.1945,
"step": 108
},
{
"epoch": 0.662613981762918,
"grad_norm": 1.0427861213684082,
"learning_rate": 4.669547078371504e-05,
"loss": 1.1904,
"step": 109
},
{
"epoch": 0.668693009118541,
"grad_norm": 1.045060396194458,
"learning_rate": 4.5871032726383386e-05,
"loss": 1.1415,
"step": 110
},
{
"epoch": 0.6747720364741642,
"grad_norm": 1.1028318405151367,
"learning_rate": 4.504772348747687e-05,
"loss": 1.2223,
"step": 111
},
{
"epoch": 0.6808510638297872,
"grad_norm": 1.0551676750183105,
"learning_rate": 4.4225768151520694e-05,
"loss": 1.1616,
"step": 112
},
{
"epoch": 0.6869300911854104,
"grad_norm": 1.101637840270996,
"learning_rate": 4.3405391432896555e-05,
"loss": 1.2207,
"step": 113
},
{
"epoch": 0.6930091185410334,
"grad_norm": 1.0059363842010498,
"learning_rate": 4.2586817614407895e-05,
"loss": 1.1746,
"step": 114
},
{
"epoch": 0.6990881458966566,
"grad_norm": 1.111784815788269,
"learning_rate": 4.17702704859633e-05,
"loss": 1.0745,
"step": 115
},
{
"epoch": 0.7051671732522796,
"grad_norm": 0.9831715226173401,
"learning_rate": 4.095597328339452e-05,
"loss": 1.1745,
"step": 116
},
{
"epoch": 0.7112462006079028,
"grad_norm": 1.20254647731781,
"learning_rate": 4.0144148627425993e-05,
"loss": 1.1417,
"step": 117
},
{
"epoch": 0.7173252279635258,
"grad_norm": 1.0116372108459473,
"learning_rate": 3.933501846281267e-05,
"loss": 1.1745,
"step": 118
},
{
"epoch": 0.723404255319149,
"grad_norm": 1.083243727684021,
"learning_rate": 3.852880399766243e-05,
"loss": 1.098,
"step": 119
},
{
"epoch": 0.729483282674772,
"grad_norm": 1.0191985368728638,
"learning_rate": 3.772572564296005e-05,
"loss": 1.2014,
"step": 120
},
{
"epoch": 0.7355623100303952,
"grad_norm": 1.1605379581451416,
"learning_rate": 3.6926002952309016e-05,
"loss": 1.2191,
"step": 121
},
{
"epoch": 0.7416413373860182,
"grad_norm": 1.0989474058151245,
"learning_rate": 3.612985456190778e-05,
"loss": 1.2149,
"step": 122
},
{
"epoch": 0.7477203647416414,
"grad_norm": 1.201255202293396,
"learning_rate": 3.533749813077677e-05,
"loss": 1.2719,
"step": 123
},
{
"epoch": 0.7537993920972644,
"grad_norm": 3.3416824340820312,
"learning_rate": 3.4549150281252636e-05,
"loss": 1.3992,
"step": 124
},
{
"epoch": 0.7598784194528876,
"grad_norm": 1.069623589515686,
"learning_rate": 3.3765026539765834e-05,
"loss": 1.3996,
"step": 125
},
{
"epoch": 0.7659574468085106,
"grad_norm": 1.0454785823822021,
"learning_rate": 3.298534127791785e-05,
"loss": 1.4316,
"step": 126
},
{
"epoch": 0.7720364741641338,
"grad_norm": 1.1272077560424805,
"learning_rate": 3.221030765387417e-05,
"loss": 1.4332,
"step": 127
},
{
"epoch": 0.7781155015197568,
"grad_norm": 1.0027071237564087,
"learning_rate": 3.144013755408895e-05,
"loss": 1.2585,
"step": 128
},
{
"epoch": 0.78419452887538,
"grad_norm": 1.2847915887832642,
"learning_rate": 3.0675041535377405e-05,
"loss": 1.436,
"step": 129
},
{
"epoch": 0.790273556231003,
"grad_norm": 1.1256495714187622,
"learning_rate": 2.991522876735154e-05,
"loss": 1.2759,
"step": 130
},
{
"epoch": 0.7963525835866262,
"grad_norm": 2.605994462966919,
"learning_rate": 2.916090697523549e-05,
"loss": 1.5428,
"step": 131
},
{
"epoch": 0.8024316109422492,
"grad_norm": 1.5910120010375977,
"learning_rate": 2.8412282383075363e-05,
"loss": 1.3673,
"step": 132
},
{
"epoch": 0.8085106382978723,
"grad_norm": 1.200351357460022,
"learning_rate": 2.766955965735968e-05,
"loss": 1.4734,
"step": 133
},
{
"epoch": 0.8145896656534954,
"grad_norm": 1.1978181600570679,
"learning_rate": 2.693294185106562e-05,
"loss": 1.319,
"step": 134
},
{
"epoch": 0.8206686930091185,
"grad_norm": 1.2815990447998047,
"learning_rate": 2.6202630348146324e-05,
"loss": 1.1354,
"step": 135
},
{
"epoch": 0.8267477203647416,
"grad_norm": 1.2171591520309448,
"learning_rate": 2.547882480847461e-05,
"loss": 1.4609,
"step": 136
},
{
"epoch": 0.8328267477203647,
"grad_norm": 1.3670789003372192,
"learning_rate": 2.476172311325783e-05,
"loss": 1.2709,
"step": 137
},
{
"epoch": 0.8389057750759878,
"grad_norm": 1.2800545692443848,
"learning_rate": 2.405152131093926e-05,
"loss": 1.245,
"step": 138
},
{
"epoch": 0.8449848024316109,
"grad_norm": 1.3339649438858032,
"learning_rate": 2.3348413563600325e-05,
"loss": 1.2444,
"step": 139
},
{
"epoch": 0.851063829787234,
"grad_norm": 1.6725680828094482,
"learning_rate": 2.2652592093878666e-05,
"loss": 1.2912,
"step": 140
},
{
"epoch": 0.8571428571428571,
"grad_norm": 1.3804372549057007,
"learning_rate": 2.196424713241637e-05,
"loss": 1.3963,
"step": 141
},
{
"epoch": 0.8632218844984803,
"grad_norm": 1.4840611219406128,
"learning_rate": 2.128356686585282e-05,
"loss": 1.3154,
"step": 142
},
{
"epoch": 0.8693009118541033,
"grad_norm": 1.4211379289627075,
"learning_rate": 2.061073738537635e-05,
"loss": 1.3298,
"step": 143
},
{
"epoch": 0.8753799392097265,
"grad_norm": 1.1204715967178345,
"learning_rate": 1.9945942635848748e-05,
"loss": 1.207,
"step": 144
},
{
"epoch": 0.8814589665653495,
"grad_norm": 1.0703834295272827,
"learning_rate": 1.928936436551661e-05,
"loss": 1.0602,
"step": 145
},
{
"epoch": 0.8875379939209727,
"grad_norm": 0.9957065582275391,
"learning_rate": 1.8641182076323148e-05,
"loss": 1.1588,
"step": 146
},
{
"epoch": 0.8936170212765957,
"grad_norm": 1.0328831672668457,
"learning_rate": 1.800157297483417e-05,
"loss": 1.1604,
"step": 147
},
{
"epoch": 0.8996960486322189,
"grad_norm": 0.949932873249054,
"learning_rate": 1.7370711923791567e-05,
"loss": 1.0872,
"step": 148
},
{
"epoch": 0.9057750759878419,
"grad_norm": 1.0512586832046509,
"learning_rate": 1.6748771394307585e-05,
"loss": 1.2157,
"step": 149
},
{
"epoch": 0.9118541033434651,
"grad_norm": 0.9940697550773621,
"learning_rate": 1.6135921418712956e-05,
"loss": 1.2266,
"step": 150
},
{
"epoch": 0.9118541033434651,
"eval_loss": 1.2603729963302612,
"eval_runtime": 31.0535,
"eval_samples_per_second": 8.92,
"eval_steps_per_second": 2.254,
"step": 150
},
{
"epoch": 0.9179331306990881,
"grad_norm": 1.0907458066940308,
"learning_rate": 1.553232954407171e-05,
"loss": 1.2081,
"step": 151
},
{
"epoch": 0.9240121580547113,
"grad_norm": 1.0060489177703857,
"learning_rate": 1.4938160786375572e-05,
"loss": 1.072,
"step": 152
},
{
"epoch": 0.9300911854103343,
"grad_norm": 0.9514420032501221,
"learning_rate": 1.435357758543015e-05,
"loss": 1.0736,
"step": 153
},
{
"epoch": 0.9361702127659575,
"grad_norm": 0.9664386510848999,
"learning_rate": 1.3778739760445552e-05,
"loss": 1.14,
"step": 154
},
{
"epoch": 0.9422492401215805,
"grad_norm": 1.0399470329284668,
"learning_rate": 1.3213804466343421e-05,
"loss": 1.2073,
"step": 155
},
{
"epoch": 0.9483282674772037,
"grad_norm": 0.9771557450294495,
"learning_rate": 1.2658926150792322e-05,
"loss": 1.0803,
"step": 156
},
{
"epoch": 0.9544072948328267,
"grad_norm": 1.0296918153762817,
"learning_rate": 1.2114256511983274e-05,
"loss": 1.2187,
"step": 157
},
{
"epoch": 0.9604863221884499,
"grad_norm": 1.1126008033752441,
"learning_rate": 1.157994445715706e-05,
"loss": 1.1505,
"step": 158
},
{
"epoch": 0.9665653495440729,
"grad_norm": 0.949400007724762,
"learning_rate": 1.1056136061894384e-05,
"loss": 1.0784,
"step": 159
},
{
"epoch": 0.9726443768996961,
"grad_norm": 1.0279536247253418,
"learning_rate": 1.0542974530180327e-05,
"loss": 1.1437,
"step": 160
},
{
"epoch": 0.9787234042553191,
"grad_norm": 1.022508144378662,
"learning_rate": 1.0040600155253765e-05,
"loss": 1.1306,
"step": 161
},
{
"epoch": 0.9848024316109423,
"grad_norm": 1.011339545249939,
"learning_rate": 9.549150281252633e-06,
"loss": 1.2241,
"step": 162
},
{
"epoch": 0.9908814589665653,
"grad_norm": 1.1113312244415283,
"learning_rate": 9.068759265665384e-06,
"loss": 1.0358,
"step": 163
},
{
"epoch": 0.9969604863221885,
"grad_norm": 1.3528187274932861,
"learning_rate": 8.599558442598998e-06,
"loss": 1.2746,
"step": 164
},
{
"epoch": 1.0030395136778116,
"grad_norm": 1.632208228111267,
"learning_rate": 8.141676086873572e-06,
"loss": 1.8117,
"step": 165
},
{
"epoch": 1.0091185410334347,
"grad_norm": 0.8551629185676575,
"learning_rate": 7.695237378953223e-06,
"loss": 1.1323,
"step": 166
},
{
"epoch": 1.0151975683890577,
"grad_norm": 0.8589297533035278,
"learning_rate": 7.260364370723044e-06,
"loss": 1.1267,
"step": 167
},
{
"epoch": 1.0212765957446808,
"grad_norm": 0.9016435742378235,
"learning_rate": 6.837175952121306e-06,
"loss": 1.1859,
"step": 168
},
{
"epoch": 1.027355623100304,
"grad_norm": 0.9005184173583984,
"learning_rate": 6.425787818636131e-06,
"loss": 1.2905,
"step": 169
},
{
"epoch": 1.033434650455927,
"grad_norm": 1.3948642015457153,
"learning_rate": 6.026312439675552e-06,
"loss": 1.0782,
"step": 170
},
{
"epoch": 1.0395136778115501,
"grad_norm": 0.9876579642295837,
"learning_rate": 5.6388590278194096e-06,
"loss": 1.1003,
"step": 171
},
{
"epoch": 1.0455927051671732,
"grad_norm": 0.9763656258583069,
"learning_rate": 5.263533508961827e-06,
"loss": 1.0394,
"step": 172
},
{
"epoch": 1.0516717325227964,
"grad_norm": 1.0157183408737183,
"learning_rate": 4.900438493352055e-06,
"loss": 1.057,
"step": 173
},
{
"epoch": 1.0577507598784195,
"grad_norm": 0.9586958289146423,
"learning_rate": 4.549673247541875e-06,
"loss": 1.0878,
"step": 174
},
{
"epoch": 1.0638297872340425,
"grad_norm": 1.0164352655410767,
"learning_rate": 4.2113336672471245e-06,
"loss": 0.9211,
"step": 175
},
{
"epoch": 1.0699088145896656,
"grad_norm": 0.9395712614059448,
"learning_rate": 3.885512251130763e-06,
"loss": 0.9316,
"step": 176
},
{
"epoch": 1.0759878419452888,
"grad_norm": 1.0541352033615112,
"learning_rate": 3.5722980755146517e-06,
"loss": 0.9426,
"step": 177
},
{
"epoch": 1.082066869300912,
"grad_norm": 1.0821480751037598,
"learning_rate": 3.271776770026963e-06,
"loss": 1.0249,
"step": 178
},
{
"epoch": 1.088145896656535,
"grad_norm": 1.0913549661636353,
"learning_rate": 2.9840304941919415e-06,
"loss": 0.9365,
"step": 179
},
{
"epoch": 1.094224924012158,
"grad_norm": 1.069218397140503,
"learning_rate": 2.7091379149682685e-06,
"loss": 1.049,
"step": 180
},
{
"epoch": 1.1003039513677813,
"grad_norm": 1.2155698537826538,
"learning_rate": 2.4471741852423237e-06,
"loss": 1.056,
"step": 181
},
{
"epoch": 1.1063829787234043,
"grad_norm": 1.283200740814209,
"learning_rate": 2.1982109232821178e-06,
"loss": 1.1162,
"step": 182
},
{
"epoch": 1.1124620060790273,
"grad_norm": 1.1658945083618164,
"learning_rate": 1.962316193157593e-06,
"loss": 0.9398,
"step": 183
},
{
"epoch": 1.1185410334346504,
"grad_norm": 1.1515864133834839,
"learning_rate": 1.7395544861325718e-06,
"loss": 1.0499,
"step": 184
},
{
"epoch": 1.1246200607902737,
"grad_norm": 1.2257872819900513,
"learning_rate": 1.5299867030334814e-06,
"loss": 1.1993,
"step": 185
},
{
"epoch": 1.1306990881458967,
"grad_norm": 1.0769529342651367,
"learning_rate": 1.333670137599713e-06,
"loss": 1.0041,
"step": 186
},
{
"epoch": 1.1367781155015197,
"grad_norm": 0.891404390335083,
"learning_rate": 1.1506584608200367e-06,
"loss": 0.8925,
"step": 187
},
{
"epoch": 1.1428571428571428,
"grad_norm": 0.8577249646186829,
"learning_rate": 9.810017062595322e-07,
"loss": 0.9949,
"step": 188
},
{
"epoch": 1.148936170212766,
"grad_norm": 0.8312423825263977,
"learning_rate": 8.247462563808817e-07,
"loss": 0.8597,
"step": 189
},
{
"epoch": 1.155015197568389,
"grad_norm": 0.917024552822113,
"learning_rate": 6.819348298638839e-07,
"loss": 0.8803,
"step": 190
},
{
"epoch": 1.1610942249240122,
"grad_norm": 0.9672223329544067,
"learning_rate": 5.526064699265753e-07,
"loss": 0.9759,
"step": 191
},
{
"epoch": 1.1671732522796352,
"grad_norm": 0.8610408306121826,
"learning_rate": 4.367965336512403e-07,
"loss": 0.876,
"step": 192
},
{
"epoch": 1.1732522796352582,
"grad_norm": 0.8313782811164856,
"learning_rate": 3.3453668231809286e-07,
"loss": 0.9894,
"step": 193
},
{
"epoch": 1.1793313069908815,
"grad_norm": 0.7922574877738953,
"learning_rate": 2.458548727494292e-07,
"loss": 0.8427,
"step": 194
},
{
"epoch": 1.1854103343465046,
"grad_norm": 0.9984447360038757,
"learning_rate": 1.7077534966650766e-07,
"loss": 0.8597,
"step": 195
},
{
"epoch": 1.1914893617021276,
"grad_norm": 0.9787684679031372,
"learning_rate": 1.0931863906127327e-07,
"loss": 1.0736,
"step": 196
},
{
"epoch": 1.1975683890577509,
"grad_norm": 0.8788344860076904,
"learning_rate": 6.150154258476315e-08,
"loss": 0.9099,
"step": 197
},
{
"epoch": 1.203647416413374,
"grad_norm": 0.8538684248924255,
"learning_rate": 2.7337132953697554e-08,
"loss": 0.8865,
"step": 198
},
{
"epoch": 1.209726443768997,
"grad_norm": 0.8741906881332397,
"learning_rate": 6.834750376549792e-09,
"loss": 0.9334,
"step": 199
},
{
"epoch": 1.21580547112462,
"grad_norm": 0.8052499294281006,
"learning_rate": 0.0,
"loss": 0.9124,
"step": 200
},
{
"epoch": 1.21580547112462,
"eval_loss": 1.2549843788146973,
"eval_runtime": 31.3414,
"eval_samples_per_second": 8.838,
"eval_steps_per_second": 2.233,
"step": 200
}
],
"logging_steps": 1,
"max_steps": 200,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 50,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 5,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.603794878135009e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}