mbart-translation-zh-yue / trainer_state.json
indiejoseph's picture
End of training
8cb98ab
raw
history blame
31 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 124030,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 4.9800854631943886e-05,
"loss": 1.6607,
"step": 500
},
{
"epoch": 0.04,
"learning_rate": 4.9599290494235265e-05,
"loss": 0.8757,
"step": 1000
},
{
"epoch": 0.06,
"learning_rate": 4.9397726356526644e-05,
"loss": 0.6738,
"step": 1500
},
{
"epoch": 0.08,
"learning_rate": 4.919616221881803e-05,
"loss": 0.595,
"step": 2000
},
{
"epoch": 0.1,
"learning_rate": 4.899500120938483e-05,
"loss": 0.5675,
"step": 2500
},
{
"epoch": 0.12,
"learning_rate": 4.879343707167621e-05,
"loss": 0.53,
"step": 3000
},
{
"epoch": 0.14,
"learning_rate": 4.859187293396759e-05,
"loss": 0.4989,
"step": 3500
},
{
"epoch": 0.16,
"learning_rate": 4.839030879625897e-05,
"loss": 0.4899,
"step": 4000
},
{
"epoch": 0.18,
"learning_rate": 4.818874465855036e-05,
"loss": 0.4695,
"step": 4500
},
{
"epoch": 0.2,
"learning_rate": 4.798718052084174e-05,
"loss": 0.4498,
"step": 5000
},
{
"epoch": 0.22,
"learning_rate": 4.7785616383133116e-05,
"loss": 0.4292,
"step": 5500
},
{
"epoch": 0.24,
"learning_rate": 4.7584052245424496e-05,
"loss": 0.4233,
"step": 6000
},
{
"epoch": 0.26,
"learning_rate": 4.7382488107715875e-05,
"loss": 0.401,
"step": 6500
},
{
"epoch": 0.28,
"learning_rate": 4.7180923970007254e-05,
"loss": 0.4,
"step": 7000
},
{
"epoch": 0.3,
"learning_rate": 4.697935983229864e-05,
"loss": 0.3965,
"step": 7500
},
{
"epoch": 0.32,
"learning_rate": 4.677779569459002e-05,
"loss": 0.3708,
"step": 8000
},
{
"epoch": 0.34,
"learning_rate": 4.65762315568814e-05,
"loss": 0.3906,
"step": 8500
},
{
"epoch": 0.36,
"learning_rate": 4.6374667419172785e-05,
"loss": 0.3615,
"step": 9000
},
{
"epoch": 0.38,
"learning_rate": 4.617350640973958e-05,
"loss": 0.3599,
"step": 9500
},
{
"epoch": 0.4,
"learning_rate": 4.597194227203096e-05,
"loss": 0.3704,
"step": 10000
},
{
"epoch": 0.42,
"learning_rate": 4.577037813432235e-05,
"loss": 0.3596,
"step": 10500
},
{
"epoch": 0.44,
"learning_rate": 4.5568813996613726e-05,
"loss": 0.3331,
"step": 11000
},
{
"epoch": 0.46,
"learning_rate": 4.5367249858905106e-05,
"loss": 0.3304,
"step": 11500
},
{
"epoch": 0.48,
"learning_rate": 4.5165685721196485e-05,
"loss": 0.3274,
"step": 12000
},
{
"epoch": 0.5,
"learning_rate": 4.4964121583487864e-05,
"loss": 0.3036,
"step": 12500
},
{
"epoch": 0.52,
"learning_rate": 4.476255744577925e-05,
"loss": 0.3131,
"step": 13000
},
{
"epoch": 0.54,
"learning_rate": 4.456139643634605e-05,
"loss": 0.293,
"step": 13500
},
{
"epoch": 0.56,
"learning_rate": 4.4359832298637426e-05,
"loss": 0.3054,
"step": 14000
},
{
"epoch": 0.58,
"learning_rate": 4.415826816092881e-05,
"loss": 0.3031,
"step": 14500
},
{
"epoch": 0.6,
"learning_rate": 4.395670402322019e-05,
"loss": 0.2922,
"step": 15000
},
{
"epoch": 0.62,
"learning_rate": 4.375513988551157e-05,
"loss": 0.2755,
"step": 15500
},
{
"epoch": 0.65,
"learning_rate": 4.355357574780296e-05,
"loss": 0.289,
"step": 16000
},
{
"epoch": 0.67,
"learning_rate": 4.3352011610094336e-05,
"loss": 0.2926,
"step": 16500
},
{
"epoch": 0.69,
"learning_rate": 4.3150447472385716e-05,
"loss": 0.2773,
"step": 17000
},
{
"epoch": 0.71,
"learning_rate": 4.2948883334677095e-05,
"loss": 0.2623,
"step": 17500
},
{
"epoch": 0.73,
"learning_rate": 4.2747319196968474e-05,
"loss": 0.2658,
"step": 18000
},
{
"epoch": 0.75,
"learning_rate": 4.254615818753527e-05,
"loss": 0.2614,
"step": 18500
},
{
"epoch": 0.77,
"learning_rate": 4.234459404982666e-05,
"loss": 0.2652,
"step": 19000
},
{
"epoch": 0.79,
"learning_rate": 4.2143029912118036e-05,
"loss": 0.255,
"step": 19500
},
{
"epoch": 0.81,
"learning_rate": 4.1941465774409415e-05,
"loss": 0.2497,
"step": 20000
},
{
"epoch": 0.83,
"learning_rate": 4.17399016367008e-05,
"loss": 0.2468,
"step": 20500
},
{
"epoch": 0.85,
"learning_rate": 4.15387406272676e-05,
"loss": 0.2533,
"step": 21000
},
{
"epoch": 0.87,
"learning_rate": 4.1337176489558984e-05,
"loss": 0.2427,
"step": 21500
},
{
"epoch": 0.89,
"learning_rate": 4.1135612351850364e-05,
"loss": 0.2355,
"step": 22000
},
{
"epoch": 0.91,
"learning_rate": 4.093404821414174e-05,
"loss": 0.2402,
"step": 22500
},
{
"epoch": 0.93,
"learning_rate": 4.073248407643312e-05,
"loss": 0.2456,
"step": 23000
},
{
"epoch": 0.95,
"learning_rate": 4.05309199387245e-05,
"loss": 0.2199,
"step": 23500
},
{
"epoch": 0.97,
"learning_rate": 4.032935580101588e-05,
"loss": 0.2198,
"step": 24000
},
{
"epoch": 0.99,
"learning_rate": 4.012779166330727e-05,
"loss": 0.2297,
"step": 24500
},
{
"epoch": 1.01,
"learning_rate": 3.9926227525598646e-05,
"loss": 0.1927,
"step": 25000
},
{
"epoch": 1.03,
"learning_rate": 3.9724663387890025e-05,
"loss": 0.1625,
"step": 25500
},
{
"epoch": 1.05,
"learning_rate": 3.952350237845683e-05,
"loss": 0.1594,
"step": 26000
},
{
"epoch": 1.07,
"learning_rate": 3.932193824074821e-05,
"loss": 0.1577,
"step": 26500
},
{
"epoch": 1.09,
"learning_rate": 3.912037410303959e-05,
"loss": 0.1661,
"step": 27000
},
{
"epoch": 1.11,
"learning_rate": 3.891921309360639e-05,
"loss": 0.1562,
"step": 27500
},
{
"epoch": 1.13,
"learning_rate": 3.871764895589777e-05,
"loss": 0.162,
"step": 28000
},
{
"epoch": 1.15,
"learning_rate": 3.851608481818915e-05,
"loss": 0.1603,
"step": 28500
},
{
"epoch": 1.17,
"learning_rate": 3.831452068048053e-05,
"loss": 0.1629,
"step": 29000
},
{
"epoch": 1.19,
"learning_rate": 3.811295654277191e-05,
"loss": 0.1507,
"step": 29500
},
{
"epoch": 1.21,
"learning_rate": 3.7911392405063294e-05,
"loss": 0.1525,
"step": 30000
},
{
"epoch": 1.23,
"learning_rate": 3.770982826735467e-05,
"loss": 0.1552,
"step": 30500
},
{
"epoch": 1.25,
"learning_rate": 3.750866725792147e-05,
"loss": 0.1518,
"step": 31000
},
{
"epoch": 1.27,
"learning_rate": 3.7307103120212856e-05,
"loss": 0.145,
"step": 31500
},
{
"epoch": 1.29,
"learning_rate": 3.7105538982504235e-05,
"loss": 0.1429,
"step": 32000
},
{
"epoch": 1.31,
"learning_rate": 3.6903974844795615e-05,
"loss": 0.1547,
"step": 32500
},
{
"epoch": 1.33,
"learning_rate": 3.6702410707087e-05,
"loss": 0.1565,
"step": 33000
},
{
"epoch": 1.35,
"learning_rate": 3.650084656937838e-05,
"loss": 0.1533,
"step": 33500
},
{
"epoch": 1.37,
"learning_rate": 3.629928243166976e-05,
"loss": 0.1424,
"step": 34000
},
{
"epoch": 1.39,
"learning_rate": 3.609771829396114e-05,
"loss": 0.141,
"step": 34500
},
{
"epoch": 1.41,
"learning_rate": 3.589615415625252e-05,
"loss": 0.1423,
"step": 35000
},
{
"epoch": 1.43,
"learning_rate": 3.56945900185439e-05,
"loss": 0.141,
"step": 35500
},
{
"epoch": 1.45,
"learning_rate": 3.54934290091107e-05,
"loss": 0.1443,
"step": 36000
},
{
"epoch": 1.47,
"learning_rate": 3.529186487140208e-05,
"loss": 0.1327,
"step": 36500
},
{
"epoch": 1.49,
"learning_rate": 3.5090300733693466e-05,
"loss": 0.1412,
"step": 37000
},
{
"epoch": 1.51,
"learning_rate": 3.488913972426026e-05,
"loss": 0.1397,
"step": 37500
},
{
"epoch": 1.53,
"learning_rate": 3.468757558655164e-05,
"loss": 0.129,
"step": 38000
},
{
"epoch": 1.55,
"learning_rate": 3.448601144884303e-05,
"loss": 0.1338,
"step": 38500
},
{
"epoch": 1.57,
"learning_rate": 3.428444731113441e-05,
"loss": 0.1301,
"step": 39000
},
{
"epoch": 1.59,
"learning_rate": 3.40832863017012e-05,
"loss": 0.1269,
"step": 39500
},
{
"epoch": 1.61,
"learning_rate": 3.388172216399258e-05,
"loss": 0.1299,
"step": 40000
},
{
"epoch": 1.63,
"learning_rate": 3.368015802628396e-05,
"loss": 0.1251,
"step": 40500
},
{
"epoch": 1.65,
"learning_rate": 3.3478997016850766e-05,
"loss": 0.1249,
"step": 41000
},
{
"epoch": 1.67,
"learning_rate": 3.3277432879142145e-05,
"loss": 0.1317,
"step": 41500
},
{
"epoch": 1.69,
"learning_rate": 3.3075868741433525e-05,
"loss": 0.1306,
"step": 42000
},
{
"epoch": 1.71,
"learning_rate": 3.287430460372491e-05,
"loss": 0.1201,
"step": 42500
},
{
"epoch": 1.73,
"learning_rate": 3.267274046601629e-05,
"loss": 0.1245,
"step": 43000
},
{
"epoch": 1.75,
"learning_rate": 3.247117632830767e-05,
"loss": 0.1223,
"step": 43500
},
{
"epoch": 1.77,
"learning_rate": 3.2269612190599055e-05,
"loss": 0.1131,
"step": 44000
},
{
"epoch": 1.79,
"learning_rate": 3.2068048052890435e-05,
"loss": 0.1195,
"step": 44500
},
{
"epoch": 1.81,
"learning_rate": 3.1866483915181814e-05,
"loss": 0.1212,
"step": 45000
},
{
"epoch": 1.83,
"learning_rate": 3.166491977747319e-05,
"loss": 0.1212,
"step": 45500
},
{
"epoch": 1.85,
"learning_rate": 3.146335563976457e-05,
"loss": 0.1035,
"step": 46000
},
{
"epoch": 1.87,
"learning_rate": 3.126179150205595e-05,
"loss": 0.1135,
"step": 46500
},
{
"epoch": 1.89,
"learning_rate": 3.106022736434734e-05,
"loss": 0.1098,
"step": 47000
},
{
"epoch": 1.91,
"learning_rate": 3.085866322663872e-05,
"loss": 0.1103,
"step": 47500
},
{
"epoch": 1.94,
"learning_rate": 3.0657099088930096e-05,
"loss": 0.1008,
"step": 48000
},
{
"epoch": 1.96,
"learning_rate": 3.0455534951221482e-05,
"loss": 0.1041,
"step": 48500
},
{
"epoch": 1.98,
"learning_rate": 3.025437394178828e-05,
"loss": 0.103,
"step": 49000
},
{
"epoch": 2.0,
"learning_rate": 3.005321293235508e-05,
"loss": 0.1027,
"step": 49500
},
{
"epoch": 2.02,
"learning_rate": 2.985164879464646e-05,
"loss": 0.0724,
"step": 50000
},
{
"epoch": 2.04,
"learning_rate": 2.9650084656937838e-05,
"loss": 0.0548,
"step": 50500
},
{
"epoch": 2.06,
"learning_rate": 2.9448520519229224e-05,
"loss": 0.0611,
"step": 51000
},
{
"epoch": 2.08,
"learning_rate": 2.9246956381520603e-05,
"loss": 0.0599,
"step": 51500
},
{
"epoch": 2.1,
"learning_rate": 2.9045392243811982e-05,
"loss": 0.0568,
"step": 52000
},
{
"epoch": 2.12,
"learning_rate": 2.8843828106103365e-05,
"loss": 0.0575,
"step": 52500
},
{
"epoch": 2.14,
"learning_rate": 2.8642667096670162e-05,
"loss": 0.0548,
"step": 53000
},
{
"epoch": 2.16,
"learning_rate": 2.844110295896154e-05,
"loss": 0.0554,
"step": 53500
},
{
"epoch": 2.18,
"learning_rate": 2.8239538821252927e-05,
"loss": 0.0569,
"step": 54000
},
{
"epoch": 2.2,
"learning_rate": 2.8037974683544303e-05,
"loss": 0.0565,
"step": 54500
},
{
"epoch": 2.22,
"learning_rate": 2.78368136741111e-05,
"loss": 0.0599,
"step": 55000
},
{
"epoch": 2.24,
"learning_rate": 2.7635249536402486e-05,
"loss": 0.0539,
"step": 55500
},
{
"epoch": 2.26,
"learning_rate": 2.7433685398693865e-05,
"loss": 0.0565,
"step": 56000
},
{
"epoch": 2.28,
"learning_rate": 2.7232524389260665e-05,
"loss": 0.0528,
"step": 56500
},
{
"epoch": 2.3,
"learning_rate": 2.7030960251552044e-05,
"loss": 0.0534,
"step": 57000
},
{
"epoch": 2.32,
"learning_rate": 2.6829396113843424e-05,
"loss": 0.0557,
"step": 57500
},
{
"epoch": 2.34,
"learning_rate": 2.662783197613481e-05,
"loss": 0.0545,
"step": 58000
},
{
"epoch": 2.36,
"learning_rate": 2.642626783842619e-05,
"loss": 0.0535,
"step": 58500
},
{
"epoch": 2.38,
"learning_rate": 2.622470370071757e-05,
"loss": 0.0596,
"step": 59000
},
{
"epoch": 2.4,
"learning_rate": 2.602313956300895e-05,
"loss": 0.0557,
"step": 59500
},
{
"epoch": 2.42,
"learning_rate": 2.582157542530033e-05,
"loss": 0.0533,
"step": 60000
},
{
"epoch": 2.44,
"learning_rate": 2.562001128759171e-05,
"loss": 0.0543,
"step": 60500
},
{
"epoch": 2.46,
"learning_rate": 2.5418447149883096e-05,
"loss": 0.051,
"step": 61000
},
{
"epoch": 2.48,
"learning_rate": 2.5216883012174475e-05,
"loss": 0.0508,
"step": 61500
},
{
"epoch": 2.5,
"learning_rate": 2.5015318874465854e-05,
"loss": 0.0505,
"step": 62000
},
{
"epoch": 2.52,
"learning_rate": 2.4814157865032654e-05,
"loss": 0.0507,
"step": 62500
},
{
"epoch": 2.54,
"learning_rate": 2.4612593727324034e-05,
"loss": 0.0519,
"step": 63000
},
{
"epoch": 2.56,
"learning_rate": 2.4411029589615416e-05,
"loss": 0.0489,
"step": 63500
},
{
"epoch": 2.58,
"learning_rate": 2.42094654519068e-05,
"loss": 0.0512,
"step": 64000
},
{
"epoch": 2.6,
"learning_rate": 2.4007901314198178e-05,
"loss": 0.0519,
"step": 64500
},
{
"epoch": 2.62,
"learning_rate": 2.3806740304764978e-05,
"loss": 0.0478,
"step": 65000
},
{
"epoch": 2.64,
"learning_rate": 2.3605176167056358e-05,
"loss": 0.0466,
"step": 65500
},
{
"epoch": 2.66,
"learning_rate": 2.340361202934774e-05,
"loss": 0.049,
"step": 66000
},
{
"epoch": 2.68,
"learning_rate": 2.320204789163912e-05,
"loss": 0.0459,
"step": 66500
},
{
"epoch": 2.7,
"learning_rate": 2.3000483753930502e-05,
"loss": 0.0488,
"step": 67000
},
{
"epoch": 2.72,
"learning_rate": 2.2799322744497302e-05,
"loss": 0.0453,
"step": 67500
},
{
"epoch": 2.74,
"learning_rate": 2.259775860678868e-05,
"loss": 0.0444,
"step": 68000
},
{
"epoch": 2.76,
"learning_rate": 2.239619446908006e-05,
"loss": 0.0452,
"step": 68500
},
{
"epoch": 2.78,
"learning_rate": 2.2194630331371444e-05,
"loss": 0.0424,
"step": 69000
},
{
"epoch": 2.8,
"learning_rate": 2.1993066193662826e-05,
"loss": 0.0442,
"step": 69500
},
{
"epoch": 2.82,
"learning_rate": 2.1791905184229623e-05,
"loss": 0.0383,
"step": 70000
},
{
"epoch": 2.84,
"learning_rate": 2.1590341046521002e-05,
"loss": 0.0446,
"step": 70500
},
{
"epoch": 2.86,
"learning_rate": 2.1388776908812385e-05,
"loss": 0.0419,
"step": 71000
},
{
"epoch": 2.88,
"learning_rate": 2.1187212771103764e-05,
"loss": 0.0418,
"step": 71500
},
{
"epoch": 2.9,
"learning_rate": 2.0986051761670564e-05,
"loss": 0.0409,
"step": 72000
},
{
"epoch": 2.92,
"learning_rate": 2.0784487623961947e-05,
"loss": 0.0378,
"step": 72500
},
{
"epoch": 2.94,
"learning_rate": 2.0582923486253326e-05,
"loss": 0.0363,
"step": 73000
},
{
"epoch": 2.96,
"learning_rate": 2.0381359348544705e-05,
"loss": 0.0363,
"step": 73500
},
{
"epoch": 2.98,
"learning_rate": 2.0179795210836088e-05,
"loss": 0.0366,
"step": 74000
},
{
"epoch": 3.0,
"learning_rate": 1.9978634201402888e-05,
"loss": 0.0342,
"step": 74500
},
{
"epoch": 3.02,
"learning_rate": 1.977707006369427e-05,
"loss": 0.0198,
"step": 75000
},
{
"epoch": 3.04,
"learning_rate": 1.957550592598565e-05,
"loss": 0.0209,
"step": 75500
},
{
"epoch": 3.06,
"learning_rate": 1.937394178827703e-05,
"loss": 0.0186,
"step": 76000
},
{
"epoch": 3.08,
"learning_rate": 1.9172377650568412e-05,
"loss": 0.0183,
"step": 76500
},
{
"epoch": 3.1,
"learning_rate": 1.897081351285979e-05,
"loss": 0.0187,
"step": 77000
},
{
"epoch": 3.12,
"learning_rate": 1.8769249375151174e-05,
"loss": 0.0192,
"step": 77500
},
{
"epoch": 3.14,
"learning_rate": 1.8567685237442557e-05,
"loss": 0.0163,
"step": 78000
},
{
"epoch": 3.16,
"learning_rate": 1.8366121099733936e-05,
"loss": 0.0218,
"step": 78500
},
{
"epoch": 3.18,
"learning_rate": 1.8164556962025315e-05,
"loss": 0.0179,
"step": 79000
},
{
"epoch": 3.2,
"learning_rate": 1.7962992824316698e-05,
"loss": 0.0201,
"step": 79500
},
{
"epoch": 3.23,
"learning_rate": 1.776142868660808e-05,
"loss": 0.0181,
"step": 80000
},
{
"epoch": 3.25,
"learning_rate": 1.755986454889946e-05,
"loss": 0.0211,
"step": 80500
},
{
"epoch": 3.27,
"learning_rate": 1.735870353946626e-05,
"loss": 0.0182,
"step": 81000
},
{
"epoch": 3.29,
"learning_rate": 1.7157542530033057e-05,
"loss": 0.0199,
"step": 81500
},
{
"epoch": 3.31,
"learning_rate": 1.695597839232444e-05,
"loss": 0.0173,
"step": 82000
},
{
"epoch": 3.33,
"learning_rate": 1.675441425461582e-05,
"loss": 0.0195,
"step": 82500
},
{
"epoch": 3.35,
"learning_rate": 1.65528501169072e-05,
"loss": 0.0212,
"step": 83000
},
{
"epoch": 3.37,
"learning_rate": 1.6351285979198584e-05,
"loss": 0.0186,
"step": 83500
},
{
"epoch": 3.39,
"learning_rate": 1.6150124969765377e-05,
"loss": 0.0188,
"step": 84000
},
{
"epoch": 3.41,
"learning_rate": 1.5948963960332177e-05,
"loss": 0.018,
"step": 84500
},
{
"epoch": 3.43,
"learning_rate": 1.5747802950898977e-05,
"loss": 0.0202,
"step": 85000
},
{
"epoch": 3.45,
"learning_rate": 1.554623881319036e-05,
"loss": 0.0186,
"step": 85500
},
{
"epoch": 3.47,
"learning_rate": 1.534467467548174e-05,
"loss": 0.0182,
"step": 86000
},
{
"epoch": 3.49,
"learning_rate": 1.514311053777312e-05,
"loss": 0.0172,
"step": 86500
},
{
"epoch": 3.51,
"learning_rate": 1.4941546400064501e-05,
"loss": 0.0164,
"step": 87000
},
{
"epoch": 3.53,
"learning_rate": 1.4739982262355884e-05,
"loss": 0.015,
"step": 87500
},
{
"epoch": 3.55,
"learning_rate": 1.4538418124647263e-05,
"loss": 0.0171,
"step": 88000
},
{
"epoch": 3.57,
"learning_rate": 1.4336853986938644e-05,
"loss": 0.0148,
"step": 88500
},
{
"epoch": 3.59,
"learning_rate": 1.4135289849230027e-05,
"loss": 0.0168,
"step": 89000
},
{
"epoch": 3.61,
"learning_rate": 1.3933725711521406e-05,
"loss": 0.0161,
"step": 89500
},
{
"epoch": 3.63,
"learning_rate": 1.3732564702088205e-05,
"loss": 0.0152,
"step": 90000
},
{
"epoch": 3.65,
"learning_rate": 1.3531000564379587e-05,
"loss": 0.0155,
"step": 90500
},
{
"epoch": 3.67,
"learning_rate": 1.3329436426670968e-05,
"loss": 0.0155,
"step": 91000
},
{
"epoch": 3.69,
"learning_rate": 1.3127872288962348e-05,
"loss": 0.0137,
"step": 91500
},
{
"epoch": 3.71,
"learning_rate": 1.2926711279529146e-05,
"loss": 0.0154,
"step": 92000
},
{
"epoch": 3.73,
"learning_rate": 1.2725147141820529e-05,
"loss": 0.0152,
"step": 92500
},
{
"epoch": 3.75,
"learning_rate": 1.2523583004111908e-05,
"loss": 0.0156,
"step": 93000
},
{
"epoch": 3.77,
"learning_rate": 1.232201886640329e-05,
"loss": 0.0147,
"step": 93500
},
{
"epoch": 3.79,
"learning_rate": 1.2120454728694672e-05,
"loss": 0.0158,
"step": 94000
},
{
"epoch": 3.81,
"learning_rate": 1.1918890590986053e-05,
"loss": 0.015,
"step": 94500
},
{
"epoch": 3.83,
"learning_rate": 1.1717326453277434e-05,
"loss": 0.0131,
"step": 95000
},
{
"epoch": 3.85,
"learning_rate": 1.1515762315568815e-05,
"loss": 0.0145,
"step": 95500
},
{
"epoch": 3.87,
"learning_rate": 1.1314198177860195e-05,
"loss": 0.0136,
"step": 96000
},
{
"epoch": 3.89,
"learning_rate": 1.1112634040151578e-05,
"loss": 0.0121,
"step": 96500
},
{
"epoch": 3.91,
"learning_rate": 1.0911069902442957e-05,
"loss": 0.0125,
"step": 97000
},
{
"epoch": 3.93,
"learning_rate": 1.0709505764734338e-05,
"loss": 0.0116,
"step": 97500
},
{
"epoch": 3.95,
"learning_rate": 1.0508344755301137e-05,
"loss": 0.0105,
"step": 98000
},
{
"epoch": 3.97,
"learning_rate": 1.0306780617592518e-05,
"loss": 0.0128,
"step": 98500
},
{
"epoch": 3.99,
"learning_rate": 1.01052164798839e-05,
"loss": 0.0115,
"step": 99000
},
{
"epoch": 4.01,
"learning_rate": 9.90365234217528e-06,
"loss": 0.009,
"step": 99500
},
{
"epoch": 4.03,
"learning_rate": 9.70208820446666e-06,
"loss": 0.0071,
"step": 100000
},
{
"epoch": 4.05,
"learning_rate": 9.500524066758043e-06,
"loss": 0.006,
"step": 100500
},
{
"epoch": 4.07,
"learning_rate": 9.298959929049424e-06,
"loss": 0.0063,
"step": 101000
},
{
"epoch": 4.09,
"learning_rate": 9.097395791340805e-06,
"loss": 0.006,
"step": 101500
},
{
"epoch": 4.11,
"learning_rate": 8.895831653632186e-06,
"loss": 0.0065,
"step": 102000
},
{
"epoch": 4.13,
"learning_rate": 8.694267515923567e-06,
"loss": 0.0059,
"step": 102500
},
{
"epoch": 4.15,
"learning_rate": 8.492703378214948e-06,
"loss": 0.0063,
"step": 103000
},
{
"epoch": 4.17,
"learning_rate": 8.291542368781747e-06,
"loss": 0.007,
"step": 103500
},
{
"epoch": 4.19,
"learning_rate": 8.089978231073128e-06,
"loss": 0.0059,
"step": 104000
},
{
"epoch": 4.21,
"learning_rate": 7.888414093364509e-06,
"loss": 0.0062,
"step": 104500
},
{
"epoch": 4.23,
"learning_rate": 7.686849955655891e-06,
"loss": 0.0055,
"step": 105000
},
{
"epoch": 4.25,
"learning_rate": 7.4852858179472715e-06,
"loss": 0.0055,
"step": 105500
},
{
"epoch": 4.27,
"learning_rate": 7.283721680238652e-06,
"loss": 0.0048,
"step": 106000
},
{
"epoch": 4.29,
"learning_rate": 7.0821575425300335e-06,
"loss": 0.0063,
"step": 106500
},
{
"epoch": 4.31,
"learning_rate": 6.880996533096831e-06,
"loss": 0.0059,
"step": 107000
},
{
"epoch": 4.33,
"learning_rate": 6.679432395388213e-06,
"loss": 0.0052,
"step": 107500
},
{
"epoch": 4.35,
"learning_rate": 6.477868257679594e-06,
"loss": 0.005,
"step": 108000
},
{
"epoch": 4.37,
"learning_rate": 6.276304119970976e-06,
"loss": 0.0043,
"step": 108500
},
{
"epoch": 4.39,
"learning_rate": 6.074739982262357e-06,
"loss": 0.0056,
"step": 109000
},
{
"epoch": 4.41,
"learning_rate": 5.873175844553738e-06,
"loss": 0.0046,
"step": 109500
},
{
"epoch": 4.43,
"learning_rate": 5.672014835120536e-06,
"loss": 0.0056,
"step": 110000
},
{
"epoch": 4.45,
"learning_rate": 5.470450697411917e-06,
"loss": 0.0049,
"step": 110500
},
{
"epoch": 4.47,
"learning_rate": 5.269289687978715e-06,
"loss": 0.006,
"step": 111000
},
{
"epoch": 4.49,
"learning_rate": 5.067725550270096e-06,
"loss": 0.0042,
"step": 111500
},
{
"epoch": 4.52,
"learning_rate": 4.866161412561477e-06,
"loss": 0.0048,
"step": 112000
},
{
"epoch": 4.54,
"learning_rate": 4.664597274852858e-06,
"loss": 0.0065,
"step": 112500
},
{
"epoch": 4.56,
"learning_rate": 4.463033137144239e-06,
"loss": 0.0052,
"step": 113000
},
{
"epoch": 4.58,
"learning_rate": 4.261468999435621e-06,
"loss": 0.0051,
"step": 113500
},
{
"epoch": 4.6,
"learning_rate": 4.059904861727002e-06,
"loss": 0.0037,
"step": 114000
},
{
"epoch": 4.62,
"learning_rate": 3.858340724018382e-06,
"loss": 0.0038,
"step": 114500
},
{
"epoch": 4.64,
"learning_rate": 3.656776586309764e-06,
"loss": 0.0044,
"step": 115000
},
{
"epoch": 4.66,
"learning_rate": 3.455615576876562e-06,
"loss": 0.0038,
"step": 115500
},
{
"epoch": 4.68,
"learning_rate": 3.2544545674433607e-06,
"loss": 0.0046,
"step": 116000
},
{
"epoch": 4.7,
"learning_rate": 3.0528904297347417e-06,
"loss": 0.0044,
"step": 116500
},
{
"epoch": 4.72,
"learning_rate": 2.8513262920261227e-06,
"loss": 0.0038,
"step": 117000
},
{
"epoch": 4.74,
"learning_rate": 2.649762154317504e-06,
"loss": 0.0041,
"step": 117500
},
{
"epoch": 4.76,
"learning_rate": 2.448198016608885e-06,
"loss": 0.0035,
"step": 118000
},
{
"epoch": 4.78,
"learning_rate": 2.246633878900266e-06,
"loss": 0.0038,
"step": 118500
},
{
"epoch": 4.8,
"learning_rate": 2.0450697411916475e-06,
"loss": 0.0042,
"step": 119000
},
{
"epoch": 4.82,
"learning_rate": 1.8435056034830283e-06,
"loss": 0.0036,
"step": 119500
},
{
"epoch": 4.84,
"learning_rate": 1.6419414657744097e-06,
"loss": 0.0028,
"step": 120000
},
{
"epoch": 4.86,
"learning_rate": 1.4407804563412078e-06,
"loss": 0.0031,
"step": 120500
},
{
"epoch": 4.88,
"learning_rate": 1.2392163186325888e-06,
"loss": 0.0024,
"step": 121000
},
{
"epoch": 4.9,
"learning_rate": 1.03765218092397e-06,
"loss": 0.0037,
"step": 121500
},
{
"epoch": 4.92,
"learning_rate": 8.364911714907685e-07,
"loss": 0.0031,
"step": 122000
},
{
"epoch": 4.94,
"learning_rate": 6.349270337821494e-07,
"loss": 0.0022,
"step": 122500
},
{
"epoch": 4.96,
"learning_rate": 4.3336289607353065e-07,
"loss": 0.004,
"step": 123000
},
{
"epoch": 4.98,
"learning_rate": 2.317987583649117e-07,
"loss": 0.0029,
"step": 123500
},
{
"epoch": 5.0,
"learning_rate": 3.0637748931710075e-08,
"loss": 0.0033,
"step": 124000
},
{
"epoch": 5.0,
"step": 124030,
"total_flos": 1.3242508872843264e+17,
"train_loss": 0.11724786504293357,
"train_runtime": 24264.2225,
"train_samples_per_second": 40.892,
"train_steps_per_second": 5.112
}
],
"logging_steps": 500,
"max_steps": 124030,
"num_train_epochs": 5,
"save_steps": 500,
"total_flos": 1.3242508872843264e+17,
"trial_name": null,
"trial_params": null
}