aleegis10's picture
Training in progress, step 200, checkpoint
5a9b2d1 verified
{
"best_metric": 0.4424768090248108,
"best_model_checkpoint": "miner_id_24/checkpoint-150",
"epoch": 1.146131805157593,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0057306590257879654,
"grad_norm": 1.0264272689819336,
"learning_rate": 1e-05,
"loss": 1.1815,
"step": 1
},
{
"epoch": 0.0057306590257879654,
"eval_loss": 4.764197826385498,
"eval_runtime": 31.0879,
"eval_samples_per_second": 9.457,
"eval_steps_per_second": 2.38,
"step": 1
},
{
"epoch": 0.011461318051575931,
"grad_norm": 1.316609501838684,
"learning_rate": 2e-05,
"loss": 1.4512,
"step": 2
},
{
"epoch": 0.017191977077363897,
"grad_norm": 1.4684877395629883,
"learning_rate": 3e-05,
"loss": 1.56,
"step": 3
},
{
"epoch": 0.022922636103151862,
"grad_norm": 1.3467730283737183,
"learning_rate": 4e-05,
"loss": 1.38,
"step": 4
},
{
"epoch": 0.02865329512893983,
"grad_norm": 1.4723625183105469,
"learning_rate": 5e-05,
"loss": 1.4113,
"step": 5
},
{
"epoch": 0.034383954154727794,
"grad_norm": 1.2219895124435425,
"learning_rate": 6e-05,
"loss": 0.9573,
"step": 6
},
{
"epoch": 0.04011461318051576,
"grad_norm": 1.8342928886413574,
"learning_rate": 7e-05,
"loss": 1.0261,
"step": 7
},
{
"epoch": 0.045845272206303724,
"grad_norm": 1.9770305156707764,
"learning_rate": 8e-05,
"loss": 1.1345,
"step": 8
},
{
"epoch": 0.05157593123209169,
"grad_norm": 2.2028651237487793,
"learning_rate": 9e-05,
"loss": 0.9319,
"step": 9
},
{
"epoch": 0.05730659025787966,
"grad_norm": 20.278539657592773,
"learning_rate": 0.0001,
"loss": 2.2523,
"step": 10
},
{
"epoch": 0.06303724928366762,
"grad_norm": 33.522491455078125,
"learning_rate": 9.999316524962345e-05,
"loss": 4.5261,
"step": 11
},
{
"epoch": 0.06876790830945559,
"grad_norm": 15.207009315490723,
"learning_rate": 9.997266286704631e-05,
"loss": 1.4516,
"step": 12
},
{
"epoch": 0.07449856733524356,
"grad_norm": 10.996587753295898,
"learning_rate": 9.993849845741524e-05,
"loss": 0.6496,
"step": 13
},
{
"epoch": 0.08022922636103152,
"grad_norm": 6.977783679962158,
"learning_rate": 9.989068136093873e-05,
"loss": 0.4604,
"step": 14
},
{
"epoch": 0.08595988538681948,
"grad_norm": 39.51494598388672,
"learning_rate": 9.98292246503335e-05,
"loss": 1.4999,
"step": 15
},
{
"epoch": 0.09169054441260745,
"grad_norm": 16.736629486083984,
"learning_rate": 9.975414512725057e-05,
"loss": 0.5388,
"step": 16
},
{
"epoch": 0.09742120343839542,
"grad_norm": 63.65343475341797,
"learning_rate": 9.966546331768191e-05,
"loss": 13.7754,
"step": 17
},
{
"epoch": 0.10315186246418338,
"grad_norm": 64.84332275390625,
"learning_rate": 9.956320346634876e-05,
"loss": 11.4004,
"step": 18
},
{
"epoch": 0.10888252148997135,
"grad_norm": 66.50566101074219,
"learning_rate": 9.944739353007344e-05,
"loss": 5.7751,
"step": 19
},
{
"epoch": 0.11461318051575932,
"grad_norm": 13.467355728149414,
"learning_rate": 9.931806517013612e-05,
"loss": 0.9894,
"step": 20
},
{
"epoch": 0.12034383954154727,
"grad_norm": 3.473876953125,
"learning_rate": 9.917525374361912e-05,
"loss": 0.9739,
"step": 21
},
{
"epoch": 0.12607449856733524,
"grad_norm": 1.947067379951477,
"learning_rate": 9.901899829374047e-05,
"loss": 0.7591,
"step": 22
},
{
"epoch": 0.1318051575931232,
"grad_norm": 1.4497781991958618,
"learning_rate": 9.884934153917997e-05,
"loss": 0.688,
"step": 23
},
{
"epoch": 0.13753581661891118,
"grad_norm": 1.3182792663574219,
"learning_rate": 9.86663298624003e-05,
"loss": 0.6148,
"step": 24
},
{
"epoch": 0.14326647564469913,
"grad_norm": 0.9997177720069885,
"learning_rate": 9.847001329696653e-05,
"loss": 0.6207,
"step": 25
},
{
"epoch": 0.1489971346704871,
"grad_norm": 1.1361724138259888,
"learning_rate": 9.826044551386744e-05,
"loss": 0.5833,
"step": 26
},
{
"epoch": 0.15472779369627507,
"grad_norm": 1.1941437721252441,
"learning_rate": 9.803768380684242e-05,
"loss": 0.8207,
"step": 27
},
{
"epoch": 0.16045845272206305,
"grad_norm": 1.0357601642608643,
"learning_rate": 9.780178907671789e-05,
"loss": 0.6325,
"step": 28
},
{
"epoch": 0.166189111747851,
"grad_norm": 0.8765930533409119,
"learning_rate": 9.755282581475769e-05,
"loss": 0.4919,
"step": 29
},
{
"epoch": 0.17191977077363896,
"grad_norm": 0.9793793559074402,
"learning_rate": 9.729086208503174e-05,
"loss": 0.4398,
"step": 30
},
{
"epoch": 0.17765042979942694,
"grad_norm": 0.9896600842475891,
"learning_rate": 9.701596950580806e-05,
"loss": 0.4801,
"step": 31
},
{
"epoch": 0.1833810888252149,
"grad_norm": 0.8649046421051025,
"learning_rate": 9.672822322997305e-05,
"loss": 0.538,
"step": 32
},
{
"epoch": 0.18911174785100288,
"grad_norm": 1.3178229331970215,
"learning_rate": 9.642770192448536e-05,
"loss": 0.606,
"step": 33
},
{
"epoch": 0.19484240687679083,
"grad_norm": 0.9757937788963318,
"learning_rate": 9.611448774886924e-05,
"loss": 0.5187,
"step": 34
},
{
"epoch": 0.20057306590257878,
"grad_norm": 1.2028745412826538,
"learning_rate": 9.578866633275288e-05,
"loss": 0.6169,
"step": 35
},
{
"epoch": 0.20630372492836677,
"grad_norm": 1.1853832006454468,
"learning_rate": 9.545032675245813e-05,
"loss": 0.6222,
"step": 36
},
{
"epoch": 0.21203438395415472,
"grad_norm": 1.2438991069793701,
"learning_rate": 9.509956150664796e-05,
"loss": 0.4947,
"step": 37
},
{
"epoch": 0.2177650429799427,
"grad_norm": 1.0978509187698364,
"learning_rate": 9.473646649103818e-05,
"loss": 0.5419,
"step": 38
},
{
"epoch": 0.22349570200573066,
"grad_norm": 1.2547284364700317,
"learning_rate": 9.43611409721806e-05,
"loss": 0.6178,
"step": 39
},
{
"epoch": 0.22922636103151864,
"grad_norm": 0.9904249906539917,
"learning_rate": 9.397368756032445e-05,
"loss": 0.5263,
"step": 40
},
{
"epoch": 0.2349570200573066,
"grad_norm": 1.0623767375946045,
"learning_rate": 9.357421218136386e-05,
"loss": 0.5814,
"step": 41
},
{
"epoch": 0.24068767908309455,
"grad_norm": 1.1961649656295776,
"learning_rate": 9.316282404787871e-05,
"loss": 0.5644,
"step": 42
},
{
"epoch": 0.24641833810888253,
"grad_norm": 1.3157678842544556,
"learning_rate": 9.273963562927695e-05,
"loss": 0.4061,
"step": 43
},
{
"epoch": 0.2521489971346705,
"grad_norm": 1.0945324897766113,
"learning_rate": 9.230476262104677e-05,
"loss": 0.9284,
"step": 44
},
{
"epoch": 0.25787965616045844,
"grad_norm": 0.7690916657447815,
"learning_rate": 9.185832391312644e-05,
"loss": 0.9635,
"step": 45
},
{
"epoch": 0.2636103151862464,
"grad_norm": 0.7278021574020386,
"learning_rate": 9.140044155740101e-05,
"loss": 0.9913,
"step": 46
},
{
"epoch": 0.2693409742120344,
"grad_norm": 0.7822308540344238,
"learning_rate": 9.093124073433463e-05,
"loss": 1.002,
"step": 47
},
{
"epoch": 0.27507163323782235,
"grad_norm": 0.725284993648529,
"learning_rate": 9.045084971874738e-05,
"loss": 0.9742,
"step": 48
},
{
"epoch": 0.2808022922636103,
"grad_norm": 0.7749102711677551,
"learning_rate": 8.995939984474624e-05,
"loss": 0.8727,
"step": 49
},
{
"epoch": 0.28653295128939826,
"grad_norm": 0.8219019174575806,
"learning_rate": 8.945702546981969e-05,
"loss": 0.7038,
"step": 50
},
{
"epoch": 0.28653295128939826,
"eval_loss": 0.9488083124160767,
"eval_runtime": 31.9866,
"eval_samples_per_second": 9.191,
"eval_steps_per_second": 2.313,
"step": 50
},
{
"epoch": 0.2922636103151863,
"grad_norm": 1.1375199556350708,
"learning_rate": 8.894386393810563e-05,
"loss": 0.7551,
"step": 51
},
{
"epoch": 0.2979942693409742,
"grad_norm": 1.0667378902435303,
"learning_rate": 8.842005554284296e-05,
"loss": 0.7102,
"step": 52
},
{
"epoch": 0.3037249283667622,
"grad_norm": 4.162139892578125,
"learning_rate": 8.788574348801675e-05,
"loss": 0.8943,
"step": 53
},
{
"epoch": 0.30945558739255014,
"grad_norm": 12.511982917785645,
"learning_rate": 8.73410738492077e-05,
"loss": 1.3943,
"step": 54
},
{
"epoch": 0.3151862464183381,
"grad_norm": 3.844420909881592,
"learning_rate": 8.678619553365659e-05,
"loss": 0.6625,
"step": 55
},
{
"epoch": 0.3209169054441261,
"grad_norm": 18.522705078125,
"learning_rate": 8.622126023955446e-05,
"loss": 0.6189,
"step": 56
},
{
"epoch": 0.32664756446991405,
"grad_norm": 5.122213363647461,
"learning_rate": 8.564642241456986e-05,
"loss": 0.4329,
"step": 57
},
{
"epoch": 0.332378223495702,
"grad_norm": 2.2696518898010254,
"learning_rate": 8.506183921362443e-05,
"loss": 0.3758,
"step": 58
},
{
"epoch": 0.33810888252148996,
"grad_norm": 3.0940942764282227,
"learning_rate": 8.44676704559283e-05,
"loss": 0.3867,
"step": 59
},
{
"epoch": 0.3438395415472779,
"grad_norm": 8.835616111755371,
"learning_rate": 8.386407858128706e-05,
"loss": 0.4398,
"step": 60
},
{
"epoch": 0.3495702005730659,
"grad_norm": 12.437342643737793,
"learning_rate": 8.32512286056924e-05,
"loss": 1.0798,
"step": 61
},
{
"epoch": 0.3553008595988539,
"grad_norm": 3.8784339427948,
"learning_rate": 8.262928807620843e-05,
"loss": 0.7854,
"step": 62
},
{
"epoch": 0.36103151862464183,
"grad_norm": 1.4383920431137085,
"learning_rate": 8.199842702516583e-05,
"loss": 0.5265,
"step": 63
},
{
"epoch": 0.3667621776504298,
"grad_norm": 1.467946171760559,
"learning_rate": 8.135881792367686e-05,
"loss": 0.6669,
"step": 64
},
{
"epoch": 0.37249283667621774,
"grad_norm": 1.0643025636672974,
"learning_rate": 8.07106356344834e-05,
"loss": 0.5215,
"step": 65
},
{
"epoch": 0.37822349570200575,
"grad_norm": 1.2381428480148315,
"learning_rate": 8.005405736415126e-05,
"loss": 0.5034,
"step": 66
},
{
"epoch": 0.3839541547277937,
"grad_norm": 1.444108486175537,
"learning_rate": 7.938926261462366e-05,
"loss": 0.6301,
"step": 67
},
{
"epoch": 0.38968481375358166,
"grad_norm": 1.552503228187561,
"learning_rate": 7.871643313414718e-05,
"loss": 0.6934,
"step": 68
},
{
"epoch": 0.3954154727793696,
"grad_norm": 0.9831112027168274,
"learning_rate": 7.803575286758364e-05,
"loss": 0.6346,
"step": 69
},
{
"epoch": 0.40114613180515757,
"grad_norm": 0.9043542146682739,
"learning_rate": 7.734740790612136e-05,
"loss": 0.536,
"step": 70
},
{
"epoch": 0.4068767908309456,
"grad_norm": 1.1685701608657837,
"learning_rate": 7.66515864363997e-05,
"loss": 0.7253,
"step": 71
},
{
"epoch": 0.41260744985673353,
"grad_norm": 0.9149118661880493,
"learning_rate": 7.594847868906076e-05,
"loss": 0.5738,
"step": 72
},
{
"epoch": 0.4183381088825215,
"grad_norm": 0.9435709714889526,
"learning_rate": 7.52382768867422e-05,
"loss": 0.5919,
"step": 73
},
{
"epoch": 0.42406876790830944,
"grad_norm": 0.794225811958313,
"learning_rate": 7.452117519152542e-05,
"loss": 0.4303,
"step": 74
},
{
"epoch": 0.4297994269340974,
"grad_norm": 0.8737073540687561,
"learning_rate": 7.379736965185368e-05,
"loss": 0.5233,
"step": 75
},
{
"epoch": 0.4355300859598854,
"grad_norm": 0.7666350603103638,
"learning_rate": 7.30670581489344e-05,
"loss": 0.4854,
"step": 76
},
{
"epoch": 0.44126074498567336,
"grad_norm": 1.0823107957839966,
"learning_rate": 7.233044034264034e-05,
"loss": 0.5387,
"step": 77
},
{
"epoch": 0.4469914040114613,
"grad_norm": 0.7942987084388733,
"learning_rate": 7.158771761692464e-05,
"loss": 0.5044,
"step": 78
},
{
"epoch": 0.45272206303724927,
"grad_norm": 1.0654313564300537,
"learning_rate": 7.083909302476453e-05,
"loss": 0.5598,
"step": 79
},
{
"epoch": 0.4584527220630373,
"grad_norm": 0.9917106628417969,
"learning_rate": 7.008477123264848e-05,
"loss": 0.5757,
"step": 80
},
{
"epoch": 0.46418338108882523,
"grad_norm": 0.9490253925323486,
"learning_rate": 6.932495846462261e-05,
"loss": 0.4478,
"step": 81
},
{
"epoch": 0.4699140401146132,
"grad_norm": 0.7999745607376099,
"learning_rate": 6.855986244591104e-05,
"loss": 0.4072,
"step": 82
},
{
"epoch": 0.47564469914040114,
"grad_norm": 1.2711093425750732,
"learning_rate": 6.778969234612584e-05,
"loss": 0.536,
"step": 83
},
{
"epoch": 0.4813753581661891,
"grad_norm": 1.2422999143600464,
"learning_rate": 6.701465872208216e-05,
"loss": 0.674,
"step": 84
},
{
"epoch": 0.4871060171919771,
"grad_norm": 1.09995436668396,
"learning_rate": 6.623497346023418e-05,
"loss": 0.5848,
"step": 85
},
{
"epoch": 0.49283667621776506,
"grad_norm": 1.2638225555419922,
"learning_rate": 6.545084971874738e-05,
"loss": 0.5901,
"step": 86
},
{
"epoch": 0.498567335243553,
"grad_norm": 0.6778761148452759,
"learning_rate": 6.466250186922325e-05,
"loss": 0.7618,
"step": 87
},
{
"epoch": 0.504297994269341,
"grad_norm": 0.6989274621009827,
"learning_rate": 6.387014543809223e-05,
"loss": 0.8459,
"step": 88
},
{
"epoch": 0.5100286532951289,
"grad_norm": 0.6884039044380188,
"learning_rate": 6.307399704769099e-05,
"loss": 0.7707,
"step": 89
},
{
"epoch": 0.5157593123209169,
"grad_norm": 0.708406925201416,
"learning_rate": 6.227427435703997e-05,
"loss": 0.8629,
"step": 90
},
{
"epoch": 0.5214899713467048,
"grad_norm": 0.6902703642845154,
"learning_rate": 6.147119600233758e-05,
"loss": 0.694,
"step": 91
},
{
"epoch": 0.5272206303724928,
"grad_norm": 0.7008216977119446,
"learning_rate": 6.066498153718735e-05,
"loss": 0.7711,
"step": 92
},
{
"epoch": 0.5329512893982808,
"grad_norm": 0.838283121585846,
"learning_rate": 5.985585137257401e-05,
"loss": 0.6012,
"step": 93
},
{
"epoch": 0.5386819484240688,
"grad_norm": 1.0871139764785767,
"learning_rate": 5.90440267166055e-05,
"loss": 0.5829,
"step": 94
},
{
"epoch": 0.5444126074498568,
"grad_norm": 0.9221871495246887,
"learning_rate": 5.8229729514036705e-05,
"loss": 0.5958,
"step": 95
},
{
"epoch": 0.5501432664756447,
"grad_norm": 1.4138253927230835,
"learning_rate": 5.74131823855921e-05,
"loss": 0.49,
"step": 96
},
{
"epoch": 0.5558739255014327,
"grad_norm": 9.99791431427002,
"learning_rate": 5.6594608567103456e-05,
"loss": 0.6695,
"step": 97
},
{
"epoch": 0.5616045845272206,
"grad_norm": 3.215693712234497,
"learning_rate": 5.577423184847932e-05,
"loss": 0.3236,
"step": 98
},
{
"epoch": 0.5673352435530086,
"grad_norm": 9.064443588256836,
"learning_rate": 5.495227651252315e-05,
"loss": 0.491,
"step": 99
},
{
"epoch": 0.5730659025787965,
"grad_norm": 8.719297409057617,
"learning_rate": 5.4128967273616625e-05,
"loss": 0.5477,
"step": 100
},
{
"epoch": 0.5730659025787965,
"eval_loss": 0.5027397871017456,
"eval_runtime": 31.4413,
"eval_samples_per_second": 9.351,
"eval_steps_per_second": 2.354,
"step": 100
},
{
"epoch": 0.5787965616045845,
"grad_norm": 9.894404411315918,
"learning_rate": 5.330452921628497e-05,
"loss": 0.5022,
"step": 101
},
{
"epoch": 0.5845272206303725,
"grad_norm": 1.1394400596618652,
"learning_rate": 5.247918773366112e-05,
"loss": 0.3321,
"step": 102
},
{
"epoch": 0.5902578796561605,
"grad_norm": 0.9384787678718567,
"learning_rate": 5.165316846586541e-05,
"loss": 0.3203,
"step": 103
},
{
"epoch": 0.5959885386819485,
"grad_norm": 1.2583256959915161,
"learning_rate": 5.0826697238317935e-05,
"loss": 0.5917,
"step": 104
},
{
"epoch": 0.6017191977077364,
"grad_norm": 0.9908493161201477,
"learning_rate": 5e-05,
"loss": 0.4444,
"step": 105
},
{
"epoch": 0.6074498567335244,
"grad_norm": 0.7263604402542114,
"learning_rate": 4.917330276168208e-05,
"loss": 0.3582,
"step": 106
},
{
"epoch": 0.6131805157593123,
"grad_norm": 0.9219439029693604,
"learning_rate": 4.834683153413459e-05,
"loss": 0.5481,
"step": 107
},
{
"epoch": 0.6189111747851003,
"grad_norm": 0.877366840839386,
"learning_rate": 4.7520812266338885e-05,
"loss": 0.6433,
"step": 108
},
{
"epoch": 0.6246418338108882,
"grad_norm": 0.8580757975578308,
"learning_rate": 4.669547078371504e-05,
"loss": 0.4153,
"step": 109
},
{
"epoch": 0.6303724928366762,
"grad_norm": 1.123096227645874,
"learning_rate": 4.5871032726383386e-05,
"loss": 0.471,
"step": 110
},
{
"epoch": 0.6361031518624641,
"grad_norm": 0.6214250326156616,
"learning_rate": 4.504772348747687e-05,
"loss": 0.4516,
"step": 111
},
{
"epoch": 0.6418338108882522,
"grad_norm": 0.6955455541610718,
"learning_rate": 4.4225768151520694e-05,
"loss": 0.4404,
"step": 112
},
{
"epoch": 0.6475644699140402,
"grad_norm": 0.9501744508743286,
"learning_rate": 4.3405391432896555e-05,
"loss": 0.525,
"step": 113
},
{
"epoch": 0.6532951289398281,
"grad_norm": 1.006244421005249,
"learning_rate": 4.2586817614407895e-05,
"loss": 0.3878,
"step": 114
},
{
"epoch": 0.6590257879656161,
"grad_norm": 0.8180000185966492,
"learning_rate": 4.17702704859633e-05,
"loss": 0.4943,
"step": 115
},
{
"epoch": 0.664756446991404,
"grad_norm": 0.880533754825592,
"learning_rate": 4.095597328339452e-05,
"loss": 0.6028,
"step": 116
},
{
"epoch": 0.670487106017192,
"grad_norm": 0.806432843208313,
"learning_rate": 4.0144148627425993e-05,
"loss": 0.4096,
"step": 117
},
{
"epoch": 0.6762177650429799,
"grad_norm": 0.8692377805709839,
"learning_rate": 3.933501846281267e-05,
"loss": 0.4796,
"step": 118
},
{
"epoch": 0.6819484240687679,
"grad_norm": 0.923999547958374,
"learning_rate": 3.852880399766243e-05,
"loss": 0.5621,
"step": 119
},
{
"epoch": 0.6876790830945558,
"grad_norm": 0.8302284479141235,
"learning_rate": 3.772572564296005e-05,
"loss": 0.4635,
"step": 120
},
{
"epoch": 0.6934097421203438,
"grad_norm": 0.9233800768852234,
"learning_rate": 3.6926002952309016e-05,
"loss": 0.6338,
"step": 121
},
{
"epoch": 0.6991404011461319,
"grad_norm": 0.9604414701461792,
"learning_rate": 3.612985456190778e-05,
"loss": 0.4962,
"step": 122
},
{
"epoch": 0.7048710601719198,
"grad_norm": 0.9841806888580322,
"learning_rate": 3.533749813077677e-05,
"loss": 0.5442,
"step": 123
},
{
"epoch": 0.7106017191977078,
"grad_norm": 0.9199331998825073,
"learning_rate": 3.4549150281252636e-05,
"loss": 0.4992,
"step": 124
},
{
"epoch": 0.7163323782234957,
"grad_norm": 2.0792620182037354,
"learning_rate": 3.3765026539765834e-05,
"loss": 0.5844,
"step": 125
},
{
"epoch": 0.7220630372492837,
"grad_norm": 1.4456840753555298,
"learning_rate": 3.298534127791785e-05,
"loss": 0.5728,
"step": 126
},
{
"epoch": 0.7277936962750716,
"grad_norm": 1.0397483110427856,
"learning_rate": 3.221030765387417e-05,
"loss": 0.4935,
"step": 127
},
{
"epoch": 0.7335243553008596,
"grad_norm": 1.067703127861023,
"learning_rate": 3.144013755408895e-05,
"loss": 0.4693,
"step": 128
},
{
"epoch": 0.7392550143266475,
"grad_norm": 1.415053129196167,
"learning_rate": 3.0675041535377405e-05,
"loss": 0.5269,
"step": 129
},
{
"epoch": 0.7449856733524355,
"grad_norm": 0.5865670442581177,
"learning_rate": 2.991522876735154e-05,
"loss": 0.6609,
"step": 130
},
{
"epoch": 0.7507163323782235,
"grad_norm": 0.6172277927398682,
"learning_rate": 2.916090697523549e-05,
"loss": 0.7333,
"step": 131
},
{
"epoch": 0.7564469914040115,
"grad_norm": 0.5793328285217285,
"learning_rate": 2.8412282383075363e-05,
"loss": 0.693,
"step": 132
},
{
"epoch": 0.7621776504297995,
"grad_norm": 0.6867206692695618,
"learning_rate": 2.766955965735968e-05,
"loss": 0.7878,
"step": 133
},
{
"epoch": 0.7679083094555874,
"grad_norm": 0.6148689985275269,
"learning_rate": 2.693294185106562e-05,
"loss": 0.7474,
"step": 134
},
{
"epoch": 0.7736389684813754,
"grad_norm": 0.6532262563705444,
"learning_rate": 2.6202630348146324e-05,
"loss": 0.6484,
"step": 135
},
{
"epoch": 0.7793696275071633,
"grad_norm": 0.7750144600868225,
"learning_rate": 2.547882480847461e-05,
"loss": 0.4612,
"step": 136
},
{
"epoch": 0.7851002865329513,
"grad_norm": 1.312247633934021,
"learning_rate": 2.476172311325783e-05,
"loss": 0.6573,
"step": 137
},
{
"epoch": 0.7908309455587392,
"grad_norm": 1.3347526788711548,
"learning_rate": 2.405152131093926e-05,
"loss": 0.4767,
"step": 138
},
{
"epoch": 0.7965616045845272,
"grad_norm": 2.2870190143585205,
"learning_rate": 2.3348413563600325e-05,
"loss": 0.3952,
"step": 139
},
{
"epoch": 0.8022922636103151,
"grad_norm": 1.9249480962753296,
"learning_rate": 2.2652592093878666e-05,
"loss": 0.3543,
"step": 140
},
{
"epoch": 0.8080229226361032,
"grad_norm": 2.5358173847198486,
"learning_rate": 2.196424713241637e-05,
"loss": 0.3907,
"step": 141
},
{
"epoch": 0.8137535816618912,
"grad_norm": 5.4295268058776855,
"learning_rate": 2.128356686585282e-05,
"loss": 0.4112,
"step": 142
},
{
"epoch": 0.8194842406876791,
"grad_norm": 7.659093379974365,
"learning_rate": 2.061073738537635e-05,
"loss": 0.4122,
"step": 143
},
{
"epoch": 0.8252148997134671,
"grad_norm": 0.9804878234863281,
"learning_rate": 1.9945942635848748e-05,
"loss": 0.3241,
"step": 144
},
{
"epoch": 0.830945558739255,
"grad_norm": 1.0490623712539673,
"learning_rate": 1.928936436551661e-05,
"loss": 0.2925,
"step": 145
},
{
"epoch": 0.836676217765043,
"grad_norm": 3.1657533645629883,
"learning_rate": 1.8641182076323148e-05,
"loss": 0.3618,
"step": 146
},
{
"epoch": 0.8424068767908309,
"grad_norm": 0.7988712787628174,
"learning_rate": 1.800157297483417e-05,
"loss": 0.5177,
"step": 147
},
{
"epoch": 0.8481375358166189,
"grad_norm": 0.7606900930404663,
"learning_rate": 1.7370711923791567e-05,
"loss": 0.4489,
"step": 148
},
{
"epoch": 0.8538681948424068,
"grad_norm": 0.8476601243019104,
"learning_rate": 1.6748771394307585e-05,
"loss": 0.4233,
"step": 149
},
{
"epoch": 0.8595988538681948,
"grad_norm": 0.8060834407806396,
"learning_rate": 1.6135921418712956e-05,
"loss": 0.4926,
"step": 150
},
{
"epoch": 0.8595988538681948,
"eval_loss": 0.4424768090248108,
"eval_runtime": 31.7245,
"eval_samples_per_second": 9.267,
"eval_steps_per_second": 2.333,
"step": 150
},
{
"epoch": 0.8653295128939829,
"grad_norm": 0.8667862415313721,
"learning_rate": 1.553232954407171e-05,
"loss": 0.4631,
"step": 151
},
{
"epoch": 0.8710601719197708,
"grad_norm": 0.8326748013496399,
"learning_rate": 1.4938160786375572e-05,
"loss": 0.3921,
"step": 152
},
{
"epoch": 0.8767908309455588,
"grad_norm": 0.7306289672851562,
"learning_rate": 1.435357758543015e-05,
"loss": 0.416,
"step": 153
},
{
"epoch": 0.8825214899713467,
"grad_norm": 0.9903921484947205,
"learning_rate": 1.3778739760445552e-05,
"loss": 0.5748,
"step": 154
},
{
"epoch": 0.8882521489971347,
"grad_norm": 0.9286304116249084,
"learning_rate": 1.3213804466343421e-05,
"loss": 0.4889,
"step": 155
},
{
"epoch": 0.8939828080229226,
"grad_norm": 1.4209407567977905,
"learning_rate": 1.2658926150792322e-05,
"loss": 0.6411,
"step": 156
},
{
"epoch": 0.8997134670487106,
"grad_norm": 0.6917632818222046,
"learning_rate": 1.2114256511983274e-05,
"loss": 0.4044,
"step": 157
},
{
"epoch": 0.9054441260744985,
"grad_norm": 0.680417001247406,
"learning_rate": 1.157994445715706e-05,
"loss": 0.4781,
"step": 158
},
{
"epoch": 0.9111747851002865,
"grad_norm": 0.8488168716430664,
"learning_rate": 1.1056136061894384e-05,
"loss": 0.52,
"step": 159
},
{
"epoch": 0.9169054441260746,
"grad_norm": 0.7880952954292297,
"learning_rate": 1.0542974530180327e-05,
"loss": 0.4502,
"step": 160
},
{
"epoch": 0.9226361031518625,
"grad_norm": 0.7359418272972107,
"learning_rate": 1.0040600155253765e-05,
"loss": 0.3859,
"step": 161
},
{
"epoch": 0.9283667621776505,
"grad_norm": 0.8013064861297607,
"learning_rate": 9.549150281252633e-06,
"loss": 0.4462,
"step": 162
},
{
"epoch": 0.9340974212034384,
"grad_norm": 1.0986295938491821,
"learning_rate": 9.068759265665384e-06,
"loss": 0.6279,
"step": 163
},
{
"epoch": 0.9398280802292264,
"grad_norm": 1.0124629735946655,
"learning_rate": 8.599558442598998e-06,
"loss": 0.3747,
"step": 164
},
{
"epoch": 0.9455587392550143,
"grad_norm": 1.118847370147705,
"learning_rate": 8.141676086873572e-06,
"loss": 0.4289,
"step": 165
},
{
"epoch": 0.9512893982808023,
"grad_norm": 1.3702805042266846,
"learning_rate": 7.695237378953223e-06,
"loss": 0.5166,
"step": 166
},
{
"epoch": 0.9570200573065902,
"grad_norm": 0.9965174794197083,
"learning_rate": 7.260364370723044e-06,
"loss": 0.4632,
"step": 167
},
{
"epoch": 0.9627507163323782,
"grad_norm": 0.9668011665344238,
"learning_rate": 6.837175952121306e-06,
"loss": 0.4677,
"step": 168
},
{
"epoch": 0.9684813753581661,
"grad_norm": 0.984090268611908,
"learning_rate": 6.425787818636131e-06,
"loss": 0.5068,
"step": 169
},
{
"epoch": 0.9742120343839542,
"grad_norm": 1.558980107307434,
"learning_rate": 6.026312439675552e-06,
"loss": 0.3354,
"step": 170
},
{
"epoch": 0.9799426934097422,
"grad_norm": 1.1338611841201782,
"learning_rate": 5.6388590278194096e-06,
"loss": 0.5402,
"step": 171
},
{
"epoch": 0.9856733524355301,
"grad_norm": 1.4251434803009033,
"learning_rate": 5.263533508961827e-06,
"loss": 0.3865,
"step": 172
},
{
"epoch": 0.9914040114613181,
"grad_norm": 0.7832275032997131,
"learning_rate": 4.900438493352055e-06,
"loss": 0.6026,
"step": 173
},
{
"epoch": 0.997134670487106,
"grad_norm": 0.8190169334411621,
"learning_rate": 4.549673247541875e-06,
"loss": 0.347,
"step": 174
},
{
"epoch": 1.002865329512894,
"grad_norm": 2.4558818340301514,
"learning_rate": 4.2113336672471245e-06,
"loss": 0.8194,
"step": 175
},
{
"epoch": 1.008595988538682,
"grad_norm": 0.4555129110813141,
"learning_rate": 3.885512251130763e-06,
"loss": 0.562,
"step": 176
},
{
"epoch": 1.0143266475644699,
"grad_norm": 0.555292010307312,
"learning_rate": 3.5722980755146517e-06,
"loss": 0.6938,
"step": 177
},
{
"epoch": 1.0200573065902578,
"grad_norm": 0.532479465007782,
"learning_rate": 3.271776770026963e-06,
"loss": 0.5869,
"step": 178
},
{
"epoch": 1.0257879656160458,
"grad_norm": 0.586027204990387,
"learning_rate": 2.9840304941919415e-06,
"loss": 0.7192,
"step": 179
},
{
"epoch": 1.0315186246418337,
"grad_norm": 0.6169580817222595,
"learning_rate": 2.7091379149682685e-06,
"loss": 0.6631,
"step": 180
},
{
"epoch": 1.0372492836676217,
"grad_norm": 0.6698386669158936,
"learning_rate": 2.4471741852423237e-06,
"loss": 0.6115,
"step": 181
},
{
"epoch": 1.0429799426934097,
"grad_norm": 0.807712972164154,
"learning_rate": 2.1982109232821178e-06,
"loss": 0.365,
"step": 182
},
{
"epoch": 1.0487106017191976,
"grad_norm": 0.7543205618858337,
"learning_rate": 1.962316193157593e-06,
"loss": 0.3497,
"step": 183
},
{
"epoch": 1.0544412607449858,
"grad_norm": 0.963528037071228,
"learning_rate": 1.7395544861325718e-06,
"loss": 0.4672,
"step": 184
},
{
"epoch": 1.0601719197707737,
"grad_norm": 1.023469090461731,
"learning_rate": 1.5299867030334814e-06,
"loss": 0.169,
"step": 185
},
{
"epoch": 1.0659025787965617,
"grad_norm": 4.012850284576416,
"learning_rate": 1.333670137599713e-06,
"loss": 0.2741,
"step": 186
},
{
"epoch": 1.0716332378223496,
"grad_norm": 2.5467660427093506,
"learning_rate": 1.1506584608200367e-06,
"loss": 0.3445,
"step": 187
},
{
"epoch": 1.0773638968481376,
"grad_norm": 2.2343711853027344,
"learning_rate": 9.810017062595322e-07,
"loss": 0.3052,
"step": 188
},
{
"epoch": 1.0830945558739256,
"grad_norm": 1.2899246215820312,
"learning_rate": 8.247462563808817e-07,
"loss": 0.3261,
"step": 189
},
{
"epoch": 1.0888252148997135,
"grad_norm": 1.2853631973266602,
"learning_rate": 6.819348298638839e-07,
"loss": 0.3044,
"step": 190
},
{
"epoch": 1.0945558739255015,
"grad_norm": 3.3697710037231445,
"learning_rate": 5.526064699265753e-07,
"loss": 0.2744,
"step": 191
},
{
"epoch": 1.1002865329512894,
"grad_norm": 1.8737415075302124,
"learning_rate": 4.367965336512403e-07,
"loss": 0.146,
"step": 192
},
{
"epoch": 1.1060171919770774,
"grad_norm": 0.9376641511917114,
"learning_rate": 3.3453668231809286e-07,
"loss": 0.4814,
"step": 193
},
{
"epoch": 1.1117478510028653,
"grad_norm": 0.7903323173522949,
"learning_rate": 2.458548727494292e-07,
"loss": 0.4596,
"step": 194
},
{
"epoch": 1.1174785100286533,
"grad_norm": 0.6514807939529419,
"learning_rate": 1.7077534966650766e-07,
"loss": 0.302,
"step": 195
},
{
"epoch": 1.1232091690544412,
"grad_norm": 0.6428126692771912,
"learning_rate": 1.0931863906127327e-07,
"loss": 0.2919,
"step": 196
},
{
"epoch": 1.1289398280802292,
"grad_norm": 0.6601886749267578,
"learning_rate": 6.150154258476315e-08,
"loss": 0.2931,
"step": 197
},
{
"epoch": 1.1346704871060171,
"grad_norm": 0.7356408834457397,
"learning_rate": 2.7337132953697554e-08,
"loss": 0.3583,
"step": 198
},
{
"epoch": 1.140401146131805,
"grad_norm": 0.599341630935669,
"learning_rate": 6.834750376549792e-09,
"loss": 0.3106,
"step": 199
},
{
"epoch": 1.146131805157593,
"grad_norm": 0.7014303803443909,
"learning_rate": 0.0,
"loss": 0.4286,
"step": 200
},
{
"epoch": 1.146131805157593,
"eval_loss": 0.4465039372444153,
"eval_runtime": 31.5316,
"eval_samples_per_second": 9.324,
"eval_steps_per_second": 2.347,
"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": 1
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.341690643827917e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}