prxy5607's picture
Training in progress, step 200, checkpoint
ba9694f verified
{
"best_metric": 0.5345189571380615,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 2.7777777777777777,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.013888888888888888,
"grad_norm": 0.3112389147281647,
"learning_rate": 1e-05,
"loss": 1.2095,
"step": 1
},
{
"epoch": 0.013888888888888888,
"eval_loss": 1.1204164028167725,
"eval_runtime": 3.4886,
"eval_samples_per_second": 34.684,
"eval_steps_per_second": 8.886,
"step": 1
},
{
"epoch": 0.027777777777777776,
"grad_norm": 0.3137545585632324,
"learning_rate": 2e-05,
"loss": 1.1913,
"step": 2
},
{
"epoch": 0.041666666666666664,
"grad_norm": 0.3253415822982788,
"learning_rate": 3e-05,
"loss": 1.234,
"step": 3
},
{
"epoch": 0.05555555555555555,
"grad_norm": 0.33316218852996826,
"learning_rate": 4e-05,
"loss": 1.2397,
"step": 4
},
{
"epoch": 0.06944444444444445,
"grad_norm": 0.3142785131931305,
"learning_rate": 5e-05,
"loss": 1.1617,
"step": 5
},
{
"epoch": 0.08333333333333333,
"grad_norm": 0.32327765226364136,
"learning_rate": 6e-05,
"loss": 1.1889,
"step": 6
},
{
"epoch": 0.09722222222222222,
"grad_norm": 0.3067941963672638,
"learning_rate": 7e-05,
"loss": 1.1352,
"step": 7
},
{
"epoch": 0.1111111111111111,
"grad_norm": 0.2889195680618286,
"learning_rate": 8e-05,
"loss": 1.1277,
"step": 8
},
{
"epoch": 0.125,
"grad_norm": 0.25924792885780334,
"learning_rate": 9e-05,
"loss": 1.053,
"step": 9
},
{
"epoch": 0.1388888888888889,
"grad_norm": 0.24769535660743713,
"learning_rate": 0.0001,
"loss": 1.0631,
"step": 10
},
{
"epoch": 0.1527777777777778,
"grad_norm": 0.2467634081840515,
"learning_rate": 9.999316524962345e-05,
"loss": 0.9718,
"step": 11
},
{
"epoch": 0.16666666666666666,
"grad_norm": 0.2721269130706787,
"learning_rate": 9.997266286704631e-05,
"loss": 0.945,
"step": 12
},
{
"epoch": 0.18055555555555555,
"grad_norm": 0.2742646038532257,
"learning_rate": 9.993849845741524e-05,
"loss": 0.8526,
"step": 13
},
{
"epoch": 0.19444444444444445,
"grad_norm": 0.26832282543182373,
"learning_rate": 9.989068136093873e-05,
"loss": 0.7533,
"step": 14
},
{
"epoch": 0.20833333333333334,
"grad_norm": 0.23127856850624084,
"learning_rate": 9.98292246503335e-05,
"loss": 0.6847,
"step": 15
},
{
"epoch": 0.2222222222222222,
"grad_norm": 0.24074342846870422,
"learning_rate": 9.975414512725057e-05,
"loss": 0.7165,
"step": 16
},
{
"epoch": 0.2361111111111111,
"grad_norm": 0.26960524916648865,
"learning_rate": 9.966546331768191e-05,
"loss": 0.6778,
"step": 17
},
{
"epoch": 0.25,
"grad_norm": 0.22152476012706757,
"learning_rate": 9.956320346634876e-05,
"loss": 1.0433,
"step": 18
},
{
"epoch": 0.2638888888888889,
"grad_norm": 0.2142818570137024,
"learning_rate": 9.944739353007344e-05,
"loss": 1.0155,
"step": 19
},
{
"epoch": 0.2777777777777778,
"grad_norm": 0.1960361748933792,
"learning_rate": 9.931806517013612e-05,
"loss": 0.9768,
"step": 20
},
{
"epoch": 0.2916666666666667,
"grad_norm": 0.18991218507289886,
"learning_rate": 9.917525374361912e-05,
"loss": 0.9745,
"step": 21
},
{
"epoch": 0.3055555555555556,
"grad_norm": 0.17349815368652344,
"learning_rate": 9.901899829374047e-05,
"loss": 0.9289,
"step": 22
},
{
"epoch": 0.3194444444444444,
"grad_norm": 0.15921980142593384,
"learning_rate": 9.884934153917997e-05,
"loss": 0.9247,
"step": 23
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.14373350143432617,
"learning_rate": 9.86663298624003e-05,
"loss": 0.8992,
"step": 24
},
{
"epoch": 0.3472222222222222,
"grad_norm": 0.13948096334934235,
"learning_rate": 9.847001329696653e-05,
"loss": 0.8626,
"step": 25
},
{
"epoch": 0.3611111111111111,
"grad_norm": 0.14365331828594208,
"learning_rate": 9.826044551386744e-05,
"loss": 0.8894,
"step": 26
},
{
"epoch": 0.375,
"grad_norm": 0.13908372819423676,
"learning_rate": 9.803768380684242e-05,
"loss": 0.7846,
"step": 27
},
{
"epoch": 0.3888888888888889,
"grad_norm": 0.13948780298233032,
"learning_rate": 9.780178907671789e-05,
"loss": 0.7975,
"step": 28
},
{
"epoch": 0.4027777777777778,
"grad_norm": 0.14315032958984375,
"learning_rate": 9.755282581475769e-05,
"loss": 0.794,
"step": 29
},
{
"epoch": 0.4166666666666667,
"grad_norm": 0.1532944291830063,
"learning_rate": 9.729086208503174e-05,
"loss": 0.7428,
"step": 30
},
{
"epoch": 0.4305555555555556,
"grad_norm": 0.16150806844234467,
"learning_rate": 9.701596950580806e-05,
"loss": 0.6904,
"step": 31
},
{
"epoch": 0.4444444444444444,
"grad_norm": 0.17786817252635956,
"learning_rate": 9.672822322997305e-05,
"loss": 0.6409,
"step": 32
},
{
"epoch": 0.4583333333333333,
"grad_norm": 0.19505026936531067,
"learning_rate": 9.642770192448536e-05,
"loss": 0.5824,
"step": 33
},
{
"epoch": 0.4722222222222222,
"grad_norm": 0.26508426666259766,
"learning_rate": 9.611448774886924e-05,
"loss": 0.4995,
"step": 34
},
{
"epoch": 0.4861111111111111,
"grad_norm": 0.1277182400226593,
"learning_rate": 9.578866633275288e-05,
"loss": 0.9415,
"step": 35
},
{
"epoch": 0.5,
"grad_norm": 0.14160484075546265,
"learning_rate": 9.545032675245813e-05,
"loss": 0.8889,
"step": 36
},
{
"epoch": 0.5138888888888888,
"grad_norm": 0.14977054297924042,
"learning_rate": 9.509956150664796e-05,
"loss": 0.8795,
"step": 37
},
{
"epoch": 0.5277777777777778,
"grad_norm": 0.1421051174402237,
"learning_rate": 9.473646649103818e-05,
"loss": 0.847,
"step": 38
},
{
"epoch": 0.5416666666666666,
"grad_norm": 0.1366896778345108,
"learning_rate": 9.43611409721806e-05,
"loss": 0.8417,
"step": 39
},
{
"epoch": 0.5555555555555556,
"grad_norm": 0.13114619255065918,
"learning_rate": 9.397368756032445e-05,
"loss": 0.8416,
"step": 40
},
{
"epoch": 0.5694444444444444,
"grad_norm": 0.11874759942293167,
"learning_rate": 9.357421218136386e-05,
"loss": 0.7827,
"step": 41
},
{
"epoch": 0.5833333333333334,
"grad_norm": 0.12351527065038681,
"learning_rate": 9.316282404787871e-05,
"loss": 0.7871,
"step": 42
},
{
"epoch": 0.5972222222222222,
"grad_norm": 0.11650143563747406,
"learning_rate": 9.273963562927695e-05,
"loss": 0.7365,
"step": 43
},
{
"epoch": 0.6111111111111112,
"grad_norm": 0.12172406166791916,
"learning_rate": 9.230476262104677e-05,
"loss": 0.773,
"step": 44
},
{
"epoch": 0.625,
"grad_norm": 0.12252252548933029,
"learning_rate": 9.185832391312644e-05,
"loss": 0.7338,
"step": 45
},
{
"epoch": 0.6388888888888888,
"grad_norm": 0.13812412321567535,
"learning_rate": 9.140044155740101e-05,
"loss": 0.6955,
"step": 46
},
{
"epoch": 0.6527777777777778,
"grad_norm": 0.14357773959636688,
"learning_rate": 9.093124073433463e-05,
"loss": 0.6665,
"step": 47
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.1663510501384735,
"learning_rate": 9.045084971874738e-05,
"loss": 0.5623,
"step": 48
},
{
"epoch": 0.6805555555555556,
"grad_norm": 0.16708587110042572,
"learning_rate": 8.995939984474624e-05,
"loss": 0.5097,
"step": 49
},
{
"epoch": 0.6944444444444444,
"grad_norm": 0.18520402908325195,
"learning_rate": 8.945702546981969e-05,
"loss": 0.4955,
"step": 50
},
{
"epoch": 0.6944444444444444,
"eval_loss": 0.6502977013587952,
"eval_runtime": 3.5154,
"eval_samples_per_second": 34.42,
"eval_steps_per_second": 8.818,
"step": 50
},
{
"epoch": 0.7083333333333334,
"grad_norm": 0.20574061572551727,
"learning_rate": 8.894386393810563e-05,
"loss": 0.3752,
"step": 51
},
{
"epoch": 0.7222222222222222,
"grad_norm": 0.145262211561203,
"learning_rate": 8.842005554284296e-05,
"loss": 0.8286,
"step": 52
},
{
"epoch": 0.7361111111111112,
"grad_norm": 0.1392854005098343,
"learning_rate": 8.788574348801675e-05,
"loss": 0.8136,
"step": 53
},
{
"epoch": 0.75,
"grad_norm": 0.14218120276927948,
"learning_rate": 8.73410738492077e-05,
"loss": 0.799,
"step": 54
},
{
"epoch": 0.7638888888888888,
"grad_norm": 0.148985356092453,
"learning_rate": 8.678619553365659e-05,
"loss": 0.7889,
"step": 55
},
{
"epoch": 0.7777777777777778,
"grad_norm": 0.14011260867118835,
"learning_rate": 8.622126023955446e-05,
"loss": 0.7625,
"step": 56
},
{
"epoch": 0.7916666666666666,
"grad_norm": 0.13147369027137756,
"learning_rate": 8.564642241456986e-05,
"loss": 0.7281,
"step": 57
},
{
"epoch": 0.8055555555555556,
"grad_norm": 0.12505590915679932,
"learning_rate": 8.506183921362443e-05,
"loss": 0.6872,
"step": 58
},
{
"epoch": 0.8194444444444444,
"grad_norm": 0.1397821605205536,
"learning_rate": 8.44676704559283e-05,
"loss": 0.7261,
"step": 59
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.1289505511522293,
"learning_rate": 8.386407858128706e-05,
"loss": 0.6839,
"step": 60
},
{
"epoch": 0.8472222222222222,
"grad_norm": 0.12853360176086426,
"learning_rate": 8.32512286056924e-05,
"loss": 0.6608,
"step": 61
},
{
"epoch": 0.8611111111111112,
"grad_norm": 0.14755533635616302,
"learning_rate": 8.262928807620843e-05,
"loss": 0.5972,
"step": 62
},
{
"epoch": 0.875,
"grad_norm": 0.15450289845466614,
"learning_rate": 8.199842702516583e-05,
"loss": 0.6179,
"step": 63
},
{
"epoch": 0.8888888888888888,
"grad_norm": 0.1698022037744522,
"learning_rate": 8.135881792367686e-05,
"loss": 0.551,
"step": 64
},
{
"epoch": 0.9027777777777778,
"grad_norm": 0.16076339781284332,
"learning_rate": 8.07106356344834e-05,
"loss": 0.5141,
"step": 65
},
{
"epoch": 0.9166666666666666,
"grad_norm": 0.17172326147556305,
"learning_rate": 8.005405736415126e-05,
"loss": 0.4842,
"step": 66
},
{
"epoch": 0.9305555555555556,
"grad_norm": 0.17616863548755646,
"learning_rate": 7.938926261462366e-05,
"loss": 0.4328,
"step": 67
},
{
"epoch": 0.9444444444444444,
"grad_norm": 0.20513340830802917,
"learning_rate": 7.871643313414718e-05,
"loss": 0.3587,
"step": 68
},
{
"epoch": 0.9583333333333334,
"grad_norm": 0.15791983902454376,
"learning_rate": 7.803575286758364e-05,
"loss": 0.8081,
"step": 69
},
{
"epoch": 0.9722222222222222,
"grad_norm": 0.15353897213935852,
"learning_rate": 7.734740790612136e-05,
"loss": 0.6442,
"step": 70
},
{
"epoch": 0.9861111111111112,
"grad_norm": 0.17296886444091797,
"learning_rate": 7.66515864363997e-05,
"loss": 0.631,
"step": 71
},
{
"epoch": 1.0,
"grad_norm": 0.20765294134616852,
"learning_rate": 7.594847868906076e-05,
"loss": 0.4446,
"step": 72
},
{
"epoch": 1.0138888888888888,
"grad_norm": 0.17101220786571503,
"learning_rate": 7.52382768867422e-05,
"loss": 0.7825,
"step": 73
},
{
"epoch": 1.0277777777777777,
"grad_norm": 0.14948996901512146,
"learning_rate": 7.452117519152542e-05,
"loss": 0.7529,
"step": 74
},
{
"epoch": 1.0416666666666667,
"grad_norm": 0.13551495969295502,
"learning_rate": 7.379736965185368e-05,
"loss": 0.7079,
"step": 75
},
{
"epoch": 1.0555555555555556,
"grad_norm": 0.13507342338562012,
"learning_rate": 7.30670581489344e-05,
"loss": 0.7062,
"step": 76
},
{
"epoch": 1.0694444444444444,
"grad_norm": 0.13421040773391724,
"learning_rate": 7.233044034264034e-05,
"loss": 0.7302,
"step": 77
},
{
"epoch": 1.0833333333333333,
"grad_norm": 0.13059109449386597,
"learning_rate": 7.158771761692464e-05,
"loss": 0.698,
"step": 78
},
{
"epoch": 1.0972222222222223,
"grad_norm": 0.1350412219762802,
"learning_rate": 7.083909302476453e-05,
"loss": 0.6791,
"step": 79
},
{
"epoch": 1.1111111111111112,
"grad_norm": 0.13963958621025085,
"learning_rate": 7.008477123264848e-05,
"loss": 0.682,
"step": 80
},
{
"epoch": 1.125,
"grad_norm": 0.14932432770729065,
"learning_rate": 6.932495846462261e-05,
"loss": 0.6713,
"step": 81
},
{
"epoch": 1.1388888888888888,
"grad_norm": 0.15794923901557922,
"learning_rate": 6.855986244591104e-05,
"loss": 0.6426,
"step": 82
},
{
"epoch": 1.1527777777777777,
"grad_norm": 0.16218756139278412,
"learning_rate": 6.778969234612584e-05,
"loss": 0.6052,
"step": 83
},
{
"epoch": 1.1666666666666667,
"grad_norm": 0.17144888639450073,
"learning_rate": 6.701465872208216e-05,
"loss": 0.5974,
"step": 84
},
{
"epoch": 1.1805555555555556,
"grad_norm": 0.1637173742055893,
"learning_rate": 6.623497346023418e-05,
"loss": 0.5076,
"step": 85
},
{
"epoch": 1.1944444444444444,
"grad_norm": 0.17121444642543793,
"learning_rate": 6.545084971874738e-05,
"loss": 0.4393,
"step": 86
},
{
"epoch": 1.2083333333333333,
"grad_norm": 0.17634309828281403,
"learning_rate": 6.466250186922325e-05,
"loss": 0.4627,
"step": 87
},
{
"epoch": 1.2222222222222223,
"grad_norm": 0.18657170236110687,
"learning_rate": 6.387014543809223e-05,
"loss": 0.4126,
"step": 88
},
{
"epoch": 1.2361111111111112,
"grad_norm": 0.21576394140720367,
"learning_rate": 6.307399704769099e-05,
"loss": 0.3399,
"step": 89
},
{
"epoch": 1.25,
"grad_norm": 0.1835145205259323,
"learning_rate": 6.227427435703997e-05,
"loss": 0.766,
"step": 90
},
{
"epoch": 1.2638888888888888,
"grad_norm": 0.18964627385139465,
"learning_rate": 6.147119600233758e-05,
"loss": 0.7648,
"step": 91
},
{
"epoch": 1.2777777777777777,
"grad_norm": 0.1734985113143921,
"learning_rate": 6.066498153718735e-05,
"loss": 0.7286,
"step": 92
},
{
"epoch": 1.2916666666666667,
"grad_norm": 0.16397635638713837,
"learning_rate": 5.985585137257401e-05,
"loss": 0.6864,
"step": 93
},
{
"epoch": 1.3055555555555556,
"grad_norm": 0.1572818011045456,
"learning_rate": 5.90440267166055e-05,
"loss": 0.6823,
"step": 94
},
{
"epoch": 1.3194444444444444,
"grad_norm": 0.1458768993616104,
"learning_rate": 5.8229729514036705e-05,
"loss": 0.6733,
"step": 95
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.14720548689365387,
"learning_rate": 5.74131823855921e-05,
"loss": 0.6394,
"step": 96
},
{
"epoch": 1.3472222222222223,
"grad_norm": 0.14361244440078735,
"learning_rate": 5.6594608567103456e-05,
"loss": 0.6245,
"step": 97
},
{
"epoch": 1.3611111111111112,
"grad_norm": 0.1422269195318222,
"learning_rate": 5.577423184847932e-05,
"loss": 0.6202,
"step": 98
},
{
"epoch": 1.375,
"grad_norm": 0.16280454397201538,
"learning_rate": 5.495227651252315e-05,
"loss": 0.6058,
"step": 99
},
{
"epoch": 1.3888888888888888,
"grad_norm": 0.17183168232440948,
"learning_rate": 5.4128967273616625e-05,
"loss": 0.5816,
"step": 100
},
{
"epoch": 1.3888888888888888,
"eval_loss": 0.5693058371543884,
"eval_runtime": 3.6015,
"eval_samples_per_second": 33.597,
"eval_steps_per_second": 8.608,
"step": 100
},
{
"epoch": 1.4027777777777777,
"grad_norm": 0.18475580215454102,
"learning_rate": 5.330452921628497e-05,
"loss": 0.534,
"step": 101
},
{
"epoch": 1.4166666666666667,
"grad_norm": 0.19207347929477692,
"learning_rate": 5.247918773366112e-05,
"loss": 0.5436,
"step": 102
},
{
"epoch": 1.4305555555555556,
"grad_norm": 0.1878172904253006,
"learning_rate": 5.165316846586541e-05,
"loss": 0.497,
"step": 103
},
{
"epoch": 1.4444444444444444,
"grad_norm": 0.20428557693958282,
"learning_rate": 5.0826697238317935e-05,
"loss": 0.4329,
"step": 104
},
{
"epoch": 1.4583333333333333,
"grad_norm": 0.19672515988349915,
"learning_rate": 5e-05,
"loss": 0.3714,
"step": 105
},
{
"epoch": 1.4722222222222223,
"grad_norm": 0.21482190489768982,
"learning_rate": 4.917330276168208e-05,
"loss": 0.3562,
"step": 106
},
{
"epoch": 1.4861111111111112,
"grad_norm": 0.16411441564559937,
"learning_rate": 4.834683153413459e-05,
"loss": 0.7145,
"step": 107
},
{
"epoch": 1.5,
"grad_norm": 0.1770886480808258,
"learning_rate": 4.7520812266338885e-05,
"loss": 0.7146,
"step": 108
},
{
"epoch": 1.5138888888888888,
"grad_norm": 0.17772319912910461,
"learning_rate": 4.669547078371504e-05,
"loss": 0.7416,
"step": 109
},
{
"epoch": 1.5277777777777777,
"grad_norm": 0.16155695915222168,
"learning_rate": 4.5871032726383386e-05,
"loss": 0.6592,
"step": 110
},
{
"epoch": 1.5416666666666665,
"grad_norm": 0.17135368287563324,
"learning_rate": 4.504772348747687e-05,
"loss": 0.6799,
"step": 111
},
{
"epoch": 1.5555555555555556,
"grad_norm": 0.1723327934741974,
"learning_rate": 4.4225768151520694e-05,
"loss": 0.6582,
"step": 112
},
{
"epoch": 1.5694444444444444,
"grad_norm": 0.1537693589925766,
"learning_rate": 4.3405391432896555e-05,
"loss": 0.6093,
"step": 113
},
{
"epoch": 1.5833333333333335,
"grad_norm": 0.14987392723560333,
"learning_rate": 4.2586817614407895e-05,
"loss": 0.619,
"step": 114
},
{
"epoch": 1.5972222222222223,
"grad_norm": 0.16366180777549744,
"learning_rate": 4.17702704859633e-05,
"loss": 0.6228,
"step": 115
},
{
"epoch": 1.6111111111111112,
"grad_norm": 0.15627038478851318,
"learning_rate": 4.095597328339452e-05,
"loss": 0.6095,
"step": 116
},
{
"epoch": 1.625,
"grad_norm": 0.15942201018333435,
"learning_rate": 4.0144148627425993e-05,
"loss": 0.5645,
"step": 117
},
{
"epoch": 1.6388888888888888,
"grad_norm": 0.1791529804468155,
"learning_rate": 3.933501846281267e-05,
"loss": 0.5296,
"step": 118
},
{
"epoch": 1.6527777777777777,
"grad_norm": 0.1955953687429428,
"learning_rate": 3.852880399766243e-05,
"loss": 0.4956,
"step": 119
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.19113066792488098,
"learning_rate": 3.772572564296005e-05,
"loss": 0.4531,
"step": 120
},
{
"epoch": 1.6805555555555556,
"grad_norm": 0.2053869515657425,
"learning_rate": 3.6926002952309016e-05,
"loss": 0.4096,
"step": 121
},
{
"epoch": 1.6944444444444444,
"grad_norm": 0.2251274585723877,
"learning_rate": 3.612985456190778e-05,
"loss": 0.3972,
"step": 122
},
{
"epoch": 1.7083333333333335,
"grad_norm": 0.2621534764766693,
"learning_rate": 3.533749813077677e-05,
"loss": 0.3028,
"step": 123
},
{
"epoch": 1.7222222222222223,
"grad_norm": 0.17428083717823029,
"learning_rate": 3.4549150281252636e-05,
"loss": 0.7743,
"step": 124
},
{
"epoch": 1.7361111111111112,
"grad_norm": 0.16121581196784973,
"learning_rate": 3.3765026539765834e-05,
"loss": 0.7127,
"step": 125
},
{
"epoch": 1.75,
"grad_norm": 0.15852221846580505,
"learning_rate": 3.298534127791785e-05,
"loss": 0.6773,
"step": 126
},
{
"epoch": 1.7638888888888888,
"grad_norm": 0.15743443369865417,
"learning_rate": 3.221030765387417e-05,
"loss": 0.6808,
"step": 127
},
{
"epoch": 1.7777777777777777,
"grad_norm": 0.15846262872219086,
"learning_rate": 3.144013755408895e-05,
"loss": 0.6537,
"step": 128
},
{
"epoch": 1.7916666666666665,
"grad_norm": 0.1547112762928009,
"learning_rate": 3.0675041535377405e-05,
"loss": 0.6391,
"step": 129
},
{
"epoch": 1.8055555555555556,
"grad_norm": 0.15034057199954987,
"learning_rate": 2.991522876735154e-05,
"loss": 0.6197,
"step": 130
},
{
"epoch": 1.8194444444444444,
"grad_norm": 0.14704330265522003,
"learning_rate": 2.916090697523549e-05,
"loss": 0.5884,
"step": 131
},
{
"epoch": 1.8333333333333335,
"grad_norm": 0.15140971541404724,
"learning_rate": 2.8412282383075363e-05,
"loss": 0.5713,
"step": 132
},
{
"epoch": 1.8472222222222223,
"grad_norm": 0.16444961726665497,
"learning_rate": 2.766955965735968e-05,
"loss": 0.5857,
"step": 133
},
{
"epoch": 1.8611111111111112,
"grad_norm": 0.17003598809242249,
"learning_rate": 2.693294185106562e-05,
"loss": 0.5449,
"step": 134
},
{
"epoch": 1.875,
"grad_norm": 0.1749505251646042,
"learning_rate": 2.6202630348146324e-05,
"loss": 0.5208,
"step": 135
},
{
"epoch": 1.8888888888888888,
"grad_norm": 0.1826929748058319,
"learning_rate": 2.547882480847461e-05,
"loss": 0.4426,
"step": 136
},
{
"epoch": 1.9027777777777777,
"grad_norm": 0.19359657168388367,
"learning_rate": 2.476172311325783e-05,
"loss": 0.4069,
"step": 137
},
{
"epoch": 1.9166666666666665,
"grad_norm": 0.18222056329250336,
"learning_rate": 2.405152131093926e-05,
"loss": 0.3872,
"step": 138
},
{
"epoch": 1.9305555555555556,
"grad_norm": 0.20088309049606323,
"learning_rate": 2.3348413563600325e-05,
"loss": 0.3768,
"step": 139
},
{
"epoch": 1.9444444444444444,
"grad_norm": 0.23081806302070618,
"learning_rate": 2.2652592093878666e-05,
"loss": 0.2929,
"step": 140
},
{
"epoch": 1.9583333333333335,
"grad_norm": 0.15289883315563202,
"learning_rate": 2.196424713241637e-05,
"loss": 0.7163,
"step": 141
},
{
"epoch": 1.9722222222222223,
"grad_norm": 0.1538965255022049,
"learning_rate": 2.128356686585282e-05,
"loss": 0.6164,
"step": 142
},
{
"epoch": 1.9861111111111112,
"grad_norm": 0.16435350477695465,
"learning_rate": 2.061073738537635e-05,
"loss": 0.5721,
"step": 143
},
{
"epoch": 2.0,
"grad_norm": 0.22278344631195068,
"learning_rate": 1.9945942635848748e-05,
"loss": 0.3827,
"step": 144
},
{
"epoch": 2.013888888888889,
"grad_norm": 0.14708329737186432,
"learning_rate": 1.928936436551661e-05,
"loss": 0.6964,
"step": 145
},
{
"epoch": 2.0277777777777777,
"grad_norm": 0.1450520157814026,
"learning_rate": 1.8641182076323148e-05,
"loss": 0.6921,
"step": 146
},
{
"epoch": 2.0416666666666665,
"grad_norm": 0.14873582124710083,
"learning_rate": 1.800157297483417e-05,
"loss": 0.6442,
"step": 147
},
{
"epoch": 2.0555555555555554,
"grad_norm": 0.14763230085372925,
"learning_rate": 1.7370711923791567e-05,
"loss": 0.6551,
"step": 148
},
{
"epoch": 2.0694444444444446,
"grad_norm": 0.1448996514081955,
"learning_rate": 1.6748771394307585e-05,
"loss": 0.6288,
"step": 149
},
{
"epoch": 2.0833333333333335,
"grad_norm": 0.1448766440153122,
"learning_rate": 1.6135921418712956e-05,
"loss": 0.6028,
"step": 150
},
{
"epoch": 2.0833333333333335,
"eval_loss": 0.5381028652191162,
"eval_runtime": 3.5449,
"eval_samples_per_second": 34.134,
"eval_steps_per_second": 8.745,
"step": 150
},
{
"epoch": 2.0972222222222223,
"grad_norm": 0.14842510223388672,
"learning_rate": 1.553232954407171e-05,
"loss": 0.5901,
"step": 151
},
{
"epoch": 2.111111111111111,
"grad_norm": 0.15183375775814056,
"learning_rate": 1.4938160786375572e-05,
"loss": 0.5814,
"step": 152
},
{
"epoch": 2.125,
"grad_norm": 0.15480774641036987,
"learning_rate": 1.435357758543015e-05,
"loss": 0.5869,
"step": 153
},
{
"epoch": 2.138888888888889,
"grad_norm": 0.1642688363790512,
"learning_rate": 1.3778739760445552e-05,
"loss": 0.5537,
"step": 154
},
{
"epoch": 2.1527777777777777,
"grad_norm": 0.1688297539949417,
"learning_rate": 1.3213804466343421e-05,
"loss": 0.5308,
"step": 155
},
{
"epoch": 2.1666666666666665,
"grad_norm": 0.17873252928256989,
"learning_rate": 1.2658926150792322e-05,
"loss": 0.5337,
"step": 156
},
{
"epoch": 2.1805555555555554,
"grad_norm": 0.17512454092502594,
"learning_rate": 1.2114256511983274e-05,
"loss": 0.4581,
"step": 157
},
{
"epoch": 2.1944444444444446,
"grad_norm": 0.1957445740699768,
"learning_rate": 1.157994445715706e-05,
"loss": 0.4185,
"step": 158
},
{
"epoch": 2.2083333333333335,
"grad_norm": 0.19680950045585632,
"learning_rate": 1.1056136061894384e-05,
"loss": 0.3692,
"step": 159
},
{
"epoch": 2.2222222222222223,
"grad_norm": 0.2072426974773407,
"learning_rate": 1.0542974530180327e-05,
"loss": 0.3534,
"step": 160
},
{
"epoch": 2.236111111111111,
"grad_norm": 0.24383068084716797,
"learning_rate": 1.0040600155253765e-05,
"loss": 0.2696,
"step": 161
},
{
"epoch": 2.25,
"grad_norm": 0.1489602029323578,
"learning_rate": 9.549150281252633e-06,
"loss": 0.6955,
"step": 162
},
{
"epoch": 2.263888888888889,
"grad_norm": 0.1519065797328949,
"learning_rate": 9.068759265665384e-06,
"loss": 0.7147,
"step": 163
},
{
"epoch": 2.2777777777777777,
"grad_norm": 0.15344607830047607,
"learning_rate": 8.599558442598998e-06,
"loss": 0.6698,
"step": 164
},
{
"epoch": 2.2916666666666665,
"grad_norm": 0.15113775432109833,
"learning_rate": 8.141676086873572e-06,
"loss": 0.6609,
"step": 165
},
{
"epoch": 2.3055555555555554,
"grad_norm": 0.14493389427661896,
"learning_rate": 7.695237378953223e-06,
"loss": 0.6014,
"step": 166
},
{
"epoch": 2.3194444444444446,
"grad_norm": 0.14825144410133362,
"learning_rate": 7.260364370723044e-06,
"loss": 0.5999,
"step": 167
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.15180328488349915,
"learning_rate": 6.837175952121306e-06,
"loss": 0.5974,
"step": 168
},
{
"epoch": 2.3472222222222223,
"grad_norm": 0.1549406200647354,
"learning_rate": 6.425787818636131e-06,
"loss": 0.5761,
"step": 169
},
{
"epoch": 2.361111111111111,
"grad_norm": 0.1564502865076065,
"learning_rate": 6.026312439675552e-06,
"loss": 0.5643,
"step": 170
},
{
"epoch": 2.375,
"grad_norm": 0.15905457735061646,
"learning_rate": 5.6388590278194096e-06,
"loss": 0.5651,
"step": 171
},
{
"epoch": 2.388888888888889,
"grad_norm": 0.17072376608848572,
"learning_rate": 5.263533508961827e-06,
"loss": 0.5345,
"step": 172
},
{
"epoch": 2.4027777777777777,
"grad_norm": 0.1651260256767273,
"learning_rate": 4.900438493352055e-06,
"loss": 0.4983,
"step": 173
},
{
"epoch": 2.4166666666666665,
"grad_norm": 0.18529348075389862,
"learning_rate": 4.549673247541875e-06,
"loss": 0.4954,
"step": 174
},
{
"epoch": 2.4305555555555554,
"grad_norm": 0.17907188832759857,
"learning_rate": 4.2113336672471245e-06,
"loss": 0.4031,
"step": 175
},
{
"epoch": 2.4444444444444446,
"grad_norm": 0.20123593509197235,
"learning_rate": 3.885512251130763e-06,
"loss": 0.3927,
"step": 176
},
{
"epoch": 2.4583333333333335,
"grad_norm": 0.1944931298494339,
"learning_rate": 3.5722980755146517e-06,
"loss": 0.3825,
"step": 177
},
{
"epoch": 2.4722222222222223,
"grad_norm": 0.21767187118530273,
"learning_rate": 3.271776770026963e-06,
"loss": 0.2717,
"step": 178
},
{
"epoch": 2.486111111111111,
"grad_norm": 0.15970081090927124,
"learning_rate": 2.9840304941919415e-06,
"loss": 0.7508,
"step": 179
},
{
"epoch": 2.5,
"grad_norm": 0.15349596738815308,
"learning_rate": 2.7091379149682685e-06,
"loss": 0.6883,
"step": 180
},
{
"epoch": 2.513888888888889,
"grad_norm": 0.15296021103858948,
"learning_rate": 2.4471741852423237e-06,
"loss": 0.6467,
"step": 181
},
{
"epoch": 2.5277777777777777,
"grad_norm": 0.145484060049057,
"learning_rate": 2.1982109232821178e-06,
"loss": 0.6391,
"step": 182
},
{
"epoch": 2.5416666666666665,
"grad_norm": 0.1434708833694458,
"learning_rate": 1.962316193157593e-06,
"loss": 0.6276,
"step": 183
},
{
"epoch": 2.5555555555555554,
"grad_norm": 0.15669021010398865,
"learning_rate": 1.7395544861325718e-06,
"loss": 0.6448,
"step": 184
},
{
"epoch": 2.5694444444444446,
"grad_norm": 0.16533377766609192,
"learning_rate": 1.5299867030334814e-06,
"loss": 0.6437,
"step": 185
},
{
"epoch": 2.5833333333333335,
"grad_norm": 0.15576769411563873,
"learning_rate": 1.333670137599713e-06,
"loss": 0.6014,
"step": 186
},
{
"epoch": 2.5972222222222223,
"grad_norm": 0.1502740979194641,
"learning_rate": 1.1506584608200367e-06,
"loss": 0.5766,
"step": 187
},
{
"epoch": 2.611111111111111,
"grad_norm": 0.16706611216068268,
"learning_rate": 9.810017062595322e-07,
"loss": 0.5761,
"step": 188
},
{
"epoch": 2.625,
"grad_norm": 0.15800993144512177,
"learning_rate": 8.247462563808817e-07,
"loss": 0.5352,
"step": 189
},
{
"epoch": 2.638888888888889,
"grad_norm": 0.18885111808776855,
"learning_rate": 6.819348298638839e-07,
"loss": 0.5341,
"step": 190
},
{
"epoch": 2.6527777777777777,
"grad_norm": 0.1742873340845108,
"learning_rate": 5.526064699265753e-07,
"loss": 0.4581,
"step": 191
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.18888191878795624,
"learning_rate": 4.367965336512403e-07,
"loss": 0.4303,
"step": 192
},
{
"epoch": 2.6805555555555554,
"grad_norm": 0.1930943876504898,
"learning_rate": 3.3453668231809286e-07,
"loss": 0.3706,
"step": 193
},
{
"epoch": 2.6944444444444446,
"grad_norm": 0.2105153501033783,
"learning_rate": 2.458548727494292e-07,
"loss": 0.3677,
"step": 194
},
{
"epoch": 2.7083333333333335,
"grad_norm": 0.219269797205925,
"learning_rate": 1.7077534966650766e-07,
"loss": 0.3111,
"step": 195
},
{
"epoch": 2.7222222222222223,
"grad_norm": 0.16448748111724854,
"learning_rate": 1.0931863906127327e-07,
"loss": 0.6885,
"step": 196
},
{
"epoch": 2.736111111111111,
"grad_norm": 0.14586412906646729,
"learning_rate": 6.150154258476315e-08,
"loss": 0.6617,
"step": 197
},
{
"epoch": 2.75,
"grad_norm": 0.14762835204601288,
"learning_rate": 2.7337132953697554e-08,
"loss": 0.6563,
"step": 198
},
{
"epoch": 2.763888888888889,
"grad_norm": 0.1469719260931015,
"learning_rate": 6.834750376549792e-09,
"loss": 0.6341,
"step": 199
},
{
"epoch": 2.7777777777777777,
"grad_norm": 0.15720607340335846,
"learning_rate": 0.0,
"loss": 0.6374,
"step": 200
},
{
"epoch": 2.7777777777777777,
"eval_loss": 0.5345189571380615,
"eval_runtime": 3.5462,
"eval_samples_per_second": 34.121,
"eval_steps_per_second": 8.742,
"step": 200
}
],
"logging_steps": 1,
"max_steps": 200,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"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": 2.55221805416448e+16,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}