| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 2860, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 5.587369652925466, | |
| "learning_rate": 1.7482517482517483e-08, | |
| "loss": 0.8755, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 5.500934350042366, | |
| "learning_rate": 8.741258741258742e-08, | |
| "loss": 0.8733, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.778154128284402, | |
| "learning_rate": 1.7482517482517484e-07, | |
| "loss": 0.8812, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.157465345945996, | |
| "learning_rate": 2.622377622377623e-07, | |
| "loss": 0.8595, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.01463883737159, | |
| "learning_rate": 3.496503496503497e-07, | |
| "loss": 0.8627, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 2.545403355372151, | |
| "learning_rate": 4.3706293706293707e-07, | |
| "loss": 0.8508, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.2986226033618844, | |
| "learning_rate": 5.244755244755246e-07, | |
| "loss": 0.8633, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.0882287948426668, | |
| "learning_rate": 6.118881118881119e-07, | |
| "loss": 0.822, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.9905683806362928, | |
| "learning_rate": 6.993006993006994e-07, | |
| "loss": 0.8195, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.8546442781858452, | |
| "learning_rate": 7.867132867132868e-07, | |
| "loss": 0.8423, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.9539088930354693, | |
| "learning_rate": 8.741258741258741e-07, | |
| "loss": 0.7999, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.8944469761122955, | |
| "learning_rate": 9.615384615384617e-07, | |
| "loss": 0.8234, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.976723504907432, | |
| "learning_rate": 1.0489510489510491e-06, | |
| "loss": 0.8047, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.8383713435930058, | |
| "learning_rate": 1.1363636363636364e-06, | |
| "loss": 0.8046, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.9093043777289103, | |
| "learning_rate": 1.2237762237762238e-06, | |
| "loss": 0.7935, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.8651426939128903, | |
| "learning_rate": 1.3111888111888113e-06, | |
| "loss": 0.8036, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.8471862984018954, | |
| "learning_rate": 1.3986013986013987e-06, | |
| "loss": 0.8034, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.9262668779136634, | |
| "learning_rate": 1.486013986013986e-06, | |
| "loss": 0.8008, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.9222277764812408, | |
| "learning_rate": 1.5734265734265736e-06, | |
| "loss": 0.7835, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.8688469770089136, | |
| "learning_rate": 1.660839160839161e-06, | |
| "loss": 0.7967, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.8814770181802394, | |
| "learning_rate": 1.7482517482517483e-06, | |
| "loss": 0.7998, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 2.037977585299295, | |
| "learning_rate": 1.8356643356643357e-06, | |
| "loss": 0.8056, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.001270020691376, | |
| "learning_rate": 1.9230769230769234e-06, | |
| "loss": 0.7756, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.9295825867916314, | |
| "learning_rate": 2.0104895104895104e-06, | |
| "loss": 0.788, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 2.0467456581571577, | |
| "learning_rate": 2.0979020979020983e-06, | |
| "loss": 0.7852, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 2.140303916653523, | |
| "learning_rate": 2.1853146853146857e-06, | |
| "loss": 0.7749, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 2.093503142861556, | |
| "learning_rate": 2.2727272727272728e-06, | |
| "loss": 0.7905, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 2.064301983027922, | |
| "learning_rate": 2.36013986013986e-06, | |
| "loss": 0.7659, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.997358747946483, | |
| "learning_rate": 2.4475524475524477e-06, | |
| "loss": 0.7831, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.1027121897036274, | |
| "learning_rate": 2.534965034965035e-06, | |
| "loss": 0.7702, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.1291996409910015, | |
| "learning_rate": 2.6223776223776225e-06, | |
| "loss": 0.7658, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 41.9980343074638, | |
| "learning_rate": 2.70979020979021e-06, | |
| "loss": 0.7668, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 2.198061904985496, | |
| "learning_rate": 2.7972027972027974e-06, | |
| "loss": 0.7742, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 2.129322746365122, | |
| "learning_rate": 2.8846153846153845e-06, | |
| "loss": 0.7458, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 2.0235052008880867, | |
| "learning_rate": 2.972027972027972e-06, | |
| "loss": 0.7434, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 2.0254135513369156, | |
| "learning_rate": 3.0594405594405598e-06, | |
| "loss": 0.7338, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.985119509495075, | |
| "learning_rate": 3.1468531468531472e-06, | |
| "loss": 0.7341, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.250959168282524, | |
| "learning_rate": 3.2342657342657347e-06, | |
| "loss": 0.7307, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 1.9968379378052319, | |
| "learning_rate": 3.321678321678322e-06, | |
| "loss": 0.7385, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 2.630990506587568, | |
| "learning_rate": 3.409090909090909e-06, | |
| "loss": 0.7328, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 3.57779109043985, | |
| "learning_rate": 3.4965034965034966e-06, | |
| "loss": 0.7429, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 26.759360492511966, | |
| "learning_rate": 3.583916083916084e-06, | |
| "loss": 0.7541, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 2.779819046550619, | |
| "learning_rate": 3.6713286713286715e-06, | |
| "loss": 0.7457, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.9676288431898625, | |
| "learning_rate": 3.7587412587412593e-06, | |
| "loss": 0.7347, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.7250524315438622, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.7383, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.8839750025590591, | |
| "learning_rate": 3.933566433566433e-06, | |
| "loss": 0.7318, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.6586996717057443, | |
| "learning_rate": 4.020979020979021e-06, | |
| "loss": 0.7227, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 3.585620352276218, | |
| "learning_rate": 4.108391608391608e-06, | |
| "loss": 0.7051, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.676940387665252, | |
| "learning_rate": 4.195804195804197e-06, | |
| "loss": 0.7296, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.4297941750638832, | |
| "learning_rate": 4.283216783216784e-06, | |
| "loss": 0.7034, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.5752492466565178, | |
| "learning_rate": 4.3706293706293715e-06, | |
| "loss": 0.7114, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 23.603921846562006, | |
| "learning_rate": 4.458041958041958e-06, | |
| "loss": 0.7189, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.4838961986379053, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 0.7134, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.553567707360373, | |
| "learning_rate": 4.632867132867133e-06, | |
| "loss": 0.7244, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.6872621755065427, | |
| "learning_rate": 4.72027972027972e-06, | |
| "loss": 0.7221, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 26.978554548474214, | |
| "learning_rate": 4.807692307692308e-06, | |
| "loss": 0.6954, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 1.5442233324126737, | |
| "learning_rate": 4.895104895104895e-06, | |
| "loss": 0.7098, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.5826467174296213, | |
| "learning_rate": 4.982517482517483e-06, | |
| "loss": 0.6956, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.529332769657097, | |
| "learning_rate": 4.999970207167611e-06, | |
| "loss": 0.7054, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.4134102929361827, | |
| "learning_rate": 4.999849175003032e-06, | |
| "loss": 0.6925, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.5561380803552274, | |
| "learning_rate": 4.999635045958129e-06, | |
| "loss": 0.6777, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 1.7122419800735684, | |
| "learning_rate": 4.999327828007281e-06, | |
| "loss": 0.6782, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.884234946585116, | |
| "learning_rate": 4.998927532591592e-06, | |
| "loss": 0.714, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.5140404948967683, | |
| "learning_rate": 4.998434174618464e-06, | |
| "loss": 0.6969, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.5157313273979758, | |
| "learning_rate": 4.997847772461038e-06, | |
| "loss": 0.6937, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.4006324095309106, | |
| "learning_rate": 4.997168347957521e-06, | |
| "loss": 0.6707, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.4555169305723101, | |
| "learning_rate": 4.996395926410354e-06, | |
| "loss": 0.6901, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.4699402702938433, | |
| "learning_rate": 4.995530536585293e-06, | |
| "loss": 0.6954, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.423966140797764, | |
| "learning_rate": 4.994572210710315e-06, | |
| "loss": 0.6834, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.4114672539255717, | |
| "learning_rate": 4.993520984474435e-06, | |
| "loss": 0.7091, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 1.4495020382615775, | |
| "learning_rate": 4.9923768970263675e-06, | |
| "loss": 0.6909, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.5759646703195747, | |
| "learning_rate": 4.991139990973071e-06, | |
| "loss": 0.6943, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 1.5472655442648218, | |
| "learning_rate": 4.989810312378165e-06, | |
| "loss": 0.699, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.4257086908354295, | |
| "learning_rate": 4.988387910760206e-06, | |
| "loss": 0.688, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.4575544094857862, | |
| "learning_rate": 4.986872839090853e-06, | |
| "loss": 0.6715, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.5103716324462753, | |
| "learning_rate": 4.985265153792887e-06, | |
| "loss": 0.6818, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.8027777365944309, | |
| "learning_rate": 4.983564914738113e-06, | |
| "loss": 0.6927, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 1.4256278816465284, | |
| "learning_rate": 4.981772185245135e-06, | |
| "loss": 0.6886, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 1.795771667278661, | |
| "learning_rate": 4.9798870320769884e-06, | |
| "loss": 0.6716, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.3370646701877693, | |
| "learning_rate": 4.9779095254386605e-06, | |
| "loss": 0.6939, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.4886996027080686, | |
| "learning_rate": 4.975839738974473e-06, | |
| "loss": 0.698, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 1.4334461526468911, | |
| "learning_rate": 4.9736777497653425e-06, | |
| "loss": 0.6706, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.500198901246391, | |
| "learning_rate": 4.971423638325906e-06, | |
| "loss": 0.6687, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.5396586819167384, | |
| "learning_rate": 4.969077488601525e-06, | |
| "loss": 0.6734, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.4725153996617695, | |
| "learning_rate": 4.966639387965158e-06, | |
| "loss": 0.6999, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.5885070209974053, | |
| "learning_rate": 4.964109427214111e-06, | |
| "loss": 0.6853, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.7768817102056875, | |
| "learning_rate": 4.961487700566646e-06, | |
| "loss": 0.6758, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 1.4156688514395073, | |
| "learning_rate": 4.958774305658484e-06, | |
| "loss": 0.682, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.615821708366966, | |
| "learning_rate": 4.955969343539162e-06, | |
| "loss": 0.6623, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 1.5778334799669915, | |
| "learning_rate": 4.95307291866827e-06, | |
| "loss": 0.7118, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.5459039639149776, | |
| "learning_rate": 4.9500851389115645e-06, | |
| "loss": 0.6822, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.950166551192635, | |
| "learning_rate": 4.947006115536947e-06, | |
| "loss": 0.6896, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.4583416417638757, | |
| "learning_rate": 4.943835963210324e-06, | |
| "loss": 0.6737, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 1.3818410898501687, | |
| "learning_rate": 4.9405747999913355e-06, | |
| "loss": 0.6663, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.642028714364656, | |
| "learning_rate": 4.937222747328956e-06, | |
| "loss": 0.6702, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.4162244952189953, | |
| "learning_rate": 4.933779930056975e-06, | |
| "loss": 0.6709, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.5391809516965524, | |
| "learning_rate": 4.9302464763893474e-06, | |
| "loss": 0.6581, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.4577041041003882, | |
| "learning_rate": 4.926622517915417e-06, | |
| "loss": 0.6523, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 1.4232896131606396, | |
| "learning_rate": 4.9229081895950185e-06, | |
| "loss": 0.6828, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.5308862157877259, | |
| "learning_rate": 4.9191036297534455e-06, | |
| "loss": 0.6671, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.4998597450310402, | |
| "learning_rate": 4.91520898007631e-06, | |
| "loss": 0.6559, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.5868254943762528, | |
| "learning_rate": 4.911224385604255e-06, | |
| "loss": 0.6731, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.5018477168697573, | |
| "learning_rate": 4.907149994727559e-06, | |
| "loss": 0.6643, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.4716960928109946, | |
| "learning_rate": 4.902985959180608e-06, | |
| "loss": 0.6556, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.4130064465598204, | |
| "learning_rate": 4.8987324340362445e-06, | |
| "loss": 0.6675, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.4999119765126108, | |
| "learning_rate": 4.894389577699994e-06, | |
| "loss": 0.6484, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.4835440823652937, | |
| "learning_rate": 4.889957551904164e-06, | |
| "loss": 0.6566, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.5704867480691889, | |
| "learning_rate": 4.885436521701824e-06, | |
| "loss": 0.6589, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.5154576884271134, | |
| "learning_rate": 4.8808266554606535e-06, | |
| "loss": 0.6476, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.4656996812743353, | |
| "learning_rate": 4.876128124856676e-06, | |
| "loss": 0.6652, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.6670494857895617, | |
| "learning_rate": 4.8713411048678635e-06, | |
| "loss": 0.6711, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 12.070438425219901, | |
| "learning_rate": 4.866465773767625e-06, | |
| "loss": 0.6941, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.9537677638893325, | |
| "learning_rate": 4.861502313118157e-06, | |
| "loss": 0.6606, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 2.0272499350973807, | |
| "learning_rate": 4.856450907763693e-06, | |
| "loss": 0.6506, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.7554436009464944, | |
| "learning_rate": 4.851311745823616e-06, | |
| "loss": 0.6695, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.6609551310539246, | |
| "eval_runtime": 330.32, | |
| "eval_samples_per_second": 22.769, | |
| "eval_steps_per_second": 0.357, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 2.145254776756945, | |
| "learning_rate": 4.846085018685449e-06, | |
| "loss": 0.6166, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.7225148009689535, | |
| "learning_rate": 4.84077092099773e-06, | |
| "loss": 0.5855, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.5799570982565634, | |
| "learning_rate": 4.835369650662767e-06, | |
| "loss": 0.589, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 1.7146719749288577, | |
| "learning_rate": 4.829881408829262e-06, | |
| "loss": 0.5809, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 1.6210603375147141, | |
| "learning_rate": 4.824306399884822e-06, | |
| "loss": 0.5619, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 1.8183201949889993, | |
| "learning_rate": 4.81864483144835e-06, | |
| "loss": 0.5821, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 1.5245577439422557, | |
| "learning_rate": 4.81289691436231e-06, | |
| "loss": 0.5622, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.6453822729343333, | |
| "learning_rate": 4.807062862684874e-06, | |
| "loss": 0.6047, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.7446190682981846, | |
| "learning_rate": 4.801142893681955e-06, | |
| "loss": 0.5696, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.6147794969895501, | |
| "learning_rate": 4.795137227819113e-06, | |
| "loss": 0.5523, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 1.6086737084266636, | |
| "learning_rate": 4.7890460887533415e-06, | |
| "loss": 0.5871, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.5378514408370791, | |
| "learning_rate": 4.782869703324746e-06, | |
| "loss": 0.5687, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 1.6893413050433177, | |
| "learning_rate": 4.7766083015480876e-06, | |
| "loss": 0.5748, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.6460419873518315, | |
| "learning_rate": 4.770262116604224e-06, | |
| "loss": 0.566, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 1.5227788895596694, | |
| "learning_rate": 4.763831384831421e-06, | |
| "loss": 0.542, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 2.0335648439912455, | |
| "learning_rate": 4.757316345716554e-06, | |
| "loss": 0.5659, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 1.5041239731809826, | |
| "learning_rate": 4.750717241886186e-06, | |
| "loss": 0.5529, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 1.900645473248817, | |
| "learning_rate": 4.744034319097536e-06, | |
| "loss": 0.5726, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.9943298775514051, | |
| "learning_rate": 4.7372678262293235e-06, | |
| "loss": 0.5709, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 1.6801680941350656, | |
| "learning_rate": 4.7304180152725035e-06, | |
| "loss": 0.5592, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 1.6643180642477764, | |
| "learning_rate": 4.723485141320877e-06, | |
| "loss": 0.5755, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 1.7468633841451968, | |
| "learning_rate": 4.716469462561595e-06, | |
| "loss": 0.5737, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.565224204964231, | |
| "learning_rate": 4.709371240265543e-06, | |
| "loss": 0.5713, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.704523210767029, | |
| "learning_rate": 4.702190738777608e-06, | |
| "loss": 0.5765, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 1.643895772580424, | |
| "learning_rate": 4.69492822550684e-06, | |
| "loss": 0.5646, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 1.8094188227885684, | |
| "learning_rate": 4.687583970916487e-06, | |
| "loss": 0.5652, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 1.7680146082507486, | |
| "learning_rate": 4.680158248513924e-06, | |
| "loss": 0.5789, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.900227044096463, | |
| "learning_rate": 4.6726513348404736e-06, | |
| "loss": 0.5806, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.6673076686826593, | |
| "learning_rate": 4.665063509461098e-06, | |
| "loss": 0.5685, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 1.8123226938487078, | |
| "learning_rate": 4.657395054953992e-06, | |
| "loss": 0.5886, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 1.7238311806282647, | |
| "learning_rate": 4.649646256900064e-06, | |
| "loss": 0.5635, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.6188026163610223, | |
| "learning_rate": 4.641817403872293e-06, | |
| "loss": 0.5731, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.4913698739002144, | |
| "learning_rate": 4.633908787424986e-06, | |
| "loss": 0.5732, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.9372955869658495, | |
| "learning_rate": 4.625920702082918e-06, | |
| "loss": 0.5571, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 1.7995058667015358, | |
| "learning_rate": 4.617853445330367e-06, | |
| "loss": 0.5607, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 1.6274220841761087, | |
| "learning_rate": 4.6097073176000325e-06, | |
| "loss": 0.559, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.6385978867845423, | |
| "learning_rate": 4.601482622261848e-06, | |
| "loss": 0.5737, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 1.9049655839724005, | |
| "learning_rate": 4.593179665611685e-06, | |
| "loss": 0.5578, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 1.8343333848444217, | |
| "learning_rate": 4.584798756859941e-06, | |
| "loss": 0.5666, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.6845673541242205, | |
| "learning_rate": 4.5763402081200295e-06, | |
| "loss": 0.5718, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.772559396539105, | |
| "learning_rate": 4.567804334396756e-06, | |
| "loss": 0.5716, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 1.748801886170466, | |
| "learning_rate": 4.559191453574582e-06, | |
| "loss": 0.5426, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 1.5929925572389767, | |
| "learning_rate": 4.550501886405795e-06, | |
| "loss": 0.5748, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 1.9926799279560756, | |
| "learning_rate": 4.541735956498555e-06, | |
| "loss": 0.5704, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 1.6789209872224784, | |
| "learning_rate": 4.532893990304848e-06, | |
| "loss": 0.5581, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.7322217739168355, | |
| "learning_rate": 4.523976317108326e-06, | |
| "loss": 0.5608, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 29.58409303047747, | |
| "learning_rate": 4.514983269012048e-06, | |
| "loss": 0.5806, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 2.307561877023173, | |
| "learning_rate": 4.5059151809261085e-06, | |
| "loss": 0.5521, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 1.8846862840887817, | |
| "learning_rate": 4.496772390555164e-06, | |
| "loss": 0.5516, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 2.8880758982566186, | |
| "learning_rate": 4.487555238385862e-06, | |
| "loss": 0.5643, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.7726833097207872, | |
| "learning_rate": 4.478264067674155e-06, | |
| "loss": 0.5603, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 1.9926460899793028, | |
| "learning_rate": 4.4688992244325215e-06, | |
| "loss": 0.5648, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.959915507617199, | |
| "learning_rate": 4.459461057417078e-06, | |
| "loss": 0.5408, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 1.8617496123839707, | |
| "learning_rate": 4.449949918114593e-06, | |
| "loss": 0.5528, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 1.9538793006795439, | |
| "learning_rate": 4.440366160729393e-06, | |
| "loss": 0.5674, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.8693300905967214, | |
| "learning_rate": 4.430710142170176e-06, | |
| "loss": 0.5578, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.9612803924260176, | |
| "learning_rate": 4.420982222036719e-06, | |
| "loss": 0.5435, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.995791143771985, | |
| "learning_rate": 4.411182762606484e-06, | |
| "loss": 0.5613, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 5.458759085906611, | |
| "learning_rate": 4.401312128821131e-06, | |
| "loss": 0.5535, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 3.017681632112604, | |
| "learning_rate": 4.391370688272919e-06, | |
| "loss": 0.5761, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 1.6117486364142628, | |
| "learning_rate": 4.381358811191025e-06, | |
| "loss": 0.5424, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 1.6889717368205845, | |
| "learning_rate": 4.3712768704277535e-06, | |
| "loss": 0.5578, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 1.6263535777974591, | |
| "learning_rate": 4.361125241444647e-06, | |
| "loss": 0.5517, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 1.5391004007544808, | |
| "learning_rate": 4.350904302298511e-06, | |
| "loss": 0.5475, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 3.919504264594871, | |
| "learning_rate": 4.3406144336273284e-06, | |
| "loss": 0.5488, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.7006307109389633, | |
| "learning_rate": 4.330256018636086e-06, | |
| "loss": 0.5385, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 1.6821530119030959, | |
| "learning_rate": 4.319829443082506e-06, | |
| "loss": 0.5392, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.81327230754709, | |
| "learning_rate": 4.309335095262675e-06, | |
| "loss": 0.5463, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.7582986070443545, | |
| "learning_rate": 4.298773365996591e-06, | |
| "loss": 0.5698, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 1.6167984983083563, | |
| "learning_rate": 4.288144648613601e-06, | |
| "loss": 0.5451, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 3.2311390071409125, | |
| "learning_rate": 4.277449338937754e-06, | |
| "loss": 0.5516, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.6218961380455157, | |
| "learning_rate": 4.266687835273071e-06, | |
| "loss": 0.538, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.6454054434252245, | |
| "learning_rate": 4.255860538388694e-06, | |
| "loss": 0.5678, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 1.9639702122144873, | |
| "learning_rate": 4.244967851503975e-06, | |
| "loss": 0.5446, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 1.7028448251386163, | |
| "learning_rate": 4.234010180273455e-06, | |
| "loss": 0.5297, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 1.689519412982808, | |
| "learning_rate": 4.2229879327717545e-06, | |
| "loss": 0.5432, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.64362516189522, | |
| "learning_rate": 4.211901519478382e-06, | |
| "loss": 0.5414, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 1.9767208825933924, | |
| "learning_rate": 4.200751353262442e-06, | |
| "loss": 0.5578, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 1.7757696899561655, | |
| "learning_rate": 4.1895378493672615e-06, | |
| "loss": 0.5447, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.7488808503021864, | |
| "learning_rate": 4.178261425394926e-06, | |
| "loss": 0.5362, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.720469955142088, | |
| "learning_rate": 4.16692250129073e-06, | |
| "loss": 0.5339, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 1.656259449924422, | |
| "learning_rate": 4.15552149932753e-06, | |
| "loss": 0.5324, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 1.605065327847084, | |
| "learning_rate": 4.144058844090032e-06, | |
| "loss": 0.5421, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 2.400014331232814, | |
| "learning_rate": 4.1325349624589625e-06, | |
| "loss": 0.5345, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 1.8217771753224867, | |
| "learning_rate": 4.120950283595188e-06, | |
| "loss": 0.5645, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 1.7606143199953959, | |
| "learning_rate": 4.109305238923718e-06, | |
| "loss": 0.5385, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.854649950712452, | |
| "learning_rate": 4.09760026211765e-06, | |
| "loss": 0.537, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 1.834481039824983, | |
| "learning_rate": 4.0858357890820115e-06, | |
| "loss": 0.5256, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 1.7881039817362976, | |
| "learning_rate": 4.0740122579375284e-06, | |
| "loss": 0.5492, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 1.99136744573053, | |
| "learning_rate": 4.062130109004313e-06, | |
| "loss": 0.5423, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 2.067114672482736, | |
| "learning_rate": 4.0501897847854596e-06, | |
| "loss": 0.5248, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.3619737058273698, | |
| "learning_rate": 4.038191729950569e-06, | |
| "loss": 0.5548, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 1.6422923327155017, | |
| "learning_rate": 4.026136391319187e-06, | |
| "loss": 0.5661, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 1.7084270040887584, | |
| "learning_rate": 4.014024217844167e-06, | |
| "loss": 0.5413, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 2.4216667822493387, | |
| "learning_rate": 4.001855660594948e-06, | |
| "loss": 0.5251, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 1.6894496135090933, | |
| "learning_rate": 3.989631172740756e-06, | |
| "loss": 0.5282, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 2.3374959162830224, | |
| "learning_rate": 3.97735120953373e-06, | |
| "loss": 0.5375, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 1.6036838276324374, | |
| "learning_rate": 3.965016228291966e-06, | |
| "loss": 0.5482, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 1.5838188110377205, | |
| "learning_rate": 3.9526266883824865e-06, | |
| "loss": 0.5406, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 2.925850402777526, | |
| "learning_rate": 3.940183051204133e-06, | |
| "loss": 0.5514, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.8345221327535834, | |
| "learning_rate": 3.927685780170385e-06, | |
| "loss": 0.5351, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 1.7672217360317701, | |
| "learning_rate": 3.915135340692098e-06, | |
| "loss": 0.5431, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.7690557869414958, | |
| "learning_rate": 3.902532200160174e-06, | |
| "loss": 0.5332, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 1.8016714829119445, | |
| "learning_rate": 3.889876827928156e-06, | |
| "loss": 0.5547, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 1.8101317872578178, | |
| "learning_rate": 3.877169695294749e-06, | |
| "loss": 0.5152, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.863558573257381, | |
| "learning_rate": 3.8644112754862614e-06, | |
| "loss": 0.5496, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 1.6834072929471855, | |
| "learning_rate": 3.8516020436389945e-06, | |
| "loss": 0.5403, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 1.8343601221708872, | |
| "learning_rate": 3.838742476781535e-06, | |
| "loss": 0.5449, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 1.8111180639710172, | |
| "learning_rate": 3.825833053816998e-06, | |
| "loss": 0.5177, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 1.760616458847631, | |
| "learning_rate": 3.812874255505191e-06, | |
| "loss": 0.5282, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 1.662863038395529, | |
| "learning_rate": 3.7998665644447064e-06, | |
| "loss": 0.5462, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.5940854942690412, | |
| "learning_rate": 3.786810465054953e-06, | |
| "loss": 0.5155, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.7490423857630772, | |
| "learning_rate": 3.773706443558112e-06, | |
| "loss": 0.5333, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 1.7805842103529974, | |
| "learning_rate": 3.7605549879610346e-06, | |
| "loss": 0.523, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.48952990770339966, | |
| "eval_runtime": 329.9798, | |
| "eval_samples_per_second": 22.792, | |
| "eval_steps_per_second": 0.358, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.530130888128286, | |
| "learning_rate": 3.747356588037064e-06, | |
| "loss": 0.4829, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 2.3455796414814656, | |
| "learning_rate": 3.7341117353077964e-06, | |
| "loss": 0.4185, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 11.689021838621713, | |
| "learning_rate": 3.7208209230247785e-06, | |
| "loss": 0.4431, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 2.136634988191819, | |
| "learning_rate": 3.7074846461511336e-06, | |
| "loss": 0.4166, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 1.8007601624193452, | |
| "learning_rate": 3.694103401343136e-06, | |
| "loss": 0.4247, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 1.8280089181701968, | |
| "learning_rate": 3.6806776869317074e-06, | |
| "loss": 0.4184, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 1.8498758253178664, | |
| "learning_rate": 3.667208002903863e-06, | |
| "loss": 0.4015, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 1.7616805224163088, | |
| "learning_rate": 3.6536948508840915e-06, | |
| "loss": 0.4139, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 1.7368842598902192, | |
| "learning_rate": 3.6401387341156715e-06, | |
| "loss": 0.3943, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.735642439390496, | |
| "learning_rate": 3.6265401574419316e-06, | |
| "loss": 0.4362, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 1.9606530829602056, | |
| "learning_rate": 3.6128996272874523e-06, | |
| "loss": 0.4177, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 2.045121741468278, | |
| "learning_rate": 3.5992176516392007e-06, | |
| "loss": 0.4191, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 1.8838455742670526, | |
| "learning_rate": 3.5854947400276164e-06, | |
| "loss": 0.414, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 2.0713934176853366, | |
| "learning_rate": 3.5717314035076355e-06, | |
| "loss": 0.423, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 1.9586979622942737, | |
| "learning_rate": 3.5579281546396582e-06, | |
| "loss": 0.4089, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 1.6793543323852747, | |
| "learning_rate": 3.54408550747046e-06, | |
| "loss": 0.4045, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 1.8609322408923432, | |
| "learning_rate": 3.530203977514049e-06, | |
| "loss": 0.4216, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 1.9301155764363938, | |
| "learning_rate": 3.516284081732466e-06, | |
| "loss": 0.4141, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 1.8203713381481375, | |
| "learning_rate": 3.5023263385165346e-06, | |
| "loss": 0.4193, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 1.986035063455061, | |
| "learning_rate": 3.4883312676665537e-06, | |
| "loss": 0.4102, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 1.932701768003649, | |
| "learning_rate": 3.4742993903729423e-06, | |
| "loss": 0.4096, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 2.072592848243265, | |
| "learning_rate": 3.460231229196826e-06, | |
| "loss": 0.4356, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 1.8608782853161558, | |
| "learning_rate": 3.446127308050579e-06, | |
| "loss": 0.422, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.8587160248700436, | |
| "learning_rate": 3.431988152178315e-06, | |
| "loss": 0.4055, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 1.7275839544864644, | |
| "learning_rate": 3.4178142881363192e-06, | |
| "loss": 0.3994, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 1.7408220116241242, | |
| "learning_rate": 3.4036062437734484e-06, | |
| "loss": 0.4247, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 1.8684812995394824, | |
| "learning_rate": 3.3893645482114663e-06, | |
| "loss": 0.4013, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 1.9169356401008089, | |
| "learning_rate": 3.3750897318253407e-06, | |
| "loss": 0.4436, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 1.9965429262903887, | |
| "learning_rate": 3.3607823262234936e-06, | |
| "loss": 0.4052, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 1.676108249215827, | |
| "learning_rate": 3.3464428642280004e-06, | |
| "loss": 0.4043, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 1.759052520530038, | |
| "learning_rate": 3.3320718798547503e-06, | |
| "loss": 0.3891, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 1.8158002528007304, | |
| "learning_rate": 3.3176699082935546e-06, | |
| "loss": 0.4048, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 1.8540661043983968, | |
| "learning_rate": 3.303237485888221e-06, | |
| "loss": 0.413, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 1.8368083328897664, | |
| "learning_rate": 3.2887751501165755e-06, | |
| "loss": 0.4134, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 1.943248250811944, | |
| "learning_rate": 3.2742834395704486e-06, | |
| "loss": 0.4225, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 2.023022399396665, | |
| "learning_rate": 3.2597628939356174e-06, | |
| "loss": 0.4045, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 2.0127015131643575, | |
| "learning_rate": 3.2452140539717047e-06, | |
| "loss": 0.4157, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 1.9502787177069507, | |
| "learning_rate": 3.2306374614920434e-06, | |
| "loss": 0.4037, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 1.938220761297596, | |
| "learning_rate": 3.2160336593434977e-06, | |
| "loss": 0.4065, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 1.8287659925527244, | |
| "learning_rate": 3.201403191386247e-06, | |
| "loss": 0.4249, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 2.115173595697826, | |
| "learning_rate": 3.1867466024735327e-06, | |
| "loss": 0.4083, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 1.8653664947018882, | |
| "learning_rate": 3.1720644384313647e-06, | |
| "loss": 0.4311, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 2.0817854592111025, | |
| "learning_rate": 3.1573572460381992e-06, | |
| "loss": 0.4184, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 1.9348176314156256, | |
| "learning_rate": 3.1426255730045703e-06, | |
| "loss": 0.4231, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 1.9693517430562117, | |
| "learning_rate": 3.127869967952698e-06, | |
| "loss": 0.4206, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.8456135208525017, | |
| "learning_rate": 3.1130909803960533e-06, | |
| "loss": 0.3978, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.9626926891274619, | |
| "learning_rate": 3.0982891607188948e-06, | |
| "loss": 0.4064, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 2.026442414272127, | |
| "learning_rate": 3.0834650601557724e-06, | |
| "loss": 0.4149, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 2.063023886136181, | |
| "learning_rate": 3.068619230770999e-06, | |
| "loss": 0.4118, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 1.9784482842258986, | |
| "learning_rate": 3.0537522254380902e-06, | |
| "loss": 0.4026, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 1.921786550245392, | |
| "learning_rate": 3.0388645978191745e-06, | |
| "loss": 0.4044, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 1.9377855933194186, | |
| "learning_rate": 3.0239569023443756e-06, | |
| "loss": 0.4294, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 1.8926358118848756, | |
| "learning_rate": 3.0090296941911633e-06, | |
| "loss": 0.4131, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 1.8195567532258379, | |
| "learning_rate": 2.9940835292636806e-06, | |
| "loss": 0.416, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 1.9219319571815525, | |
| "learning_rate": 2.9791189641720385e-06, | |
| "loss": 0.4023, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 1.9615684580403125, | |
| "learning_rate": 2.9641365562115886e-06, | |
| "loss": 0.4256, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 2.2610537590860145, | |
| "learning_rate": 2.949136863342169e-06, | |
| "loss": 0.4101, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.9357969284665328, | |
| "learning_rate": 2.9341204441673267e-06, | |
| "loss": 0.4202, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 1.8520492192305298, | |
| "learning_rate": 2.9190878579135077e-06, | |
| "loss": 0.4155, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 1.9246616441113338, | |
| "learning_rate": 2.904039664409244e-06, | |
| "loss": 0.4056, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 2.030321498137203, | |
| "learning_rate": 2.888976424064289e-06, | |
| "loss": 0.4176, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 2.238812621763622, | |
| "learning_rate": 2.8738986978487625e-06, | |
| "loss": 0.4049, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 2.259109743694307, | |
| "learning_rate": 2.8588070472722486e-06, | |
| "loss": 0.3983, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 2.437226455840826, | |
| "learning_rate": 2.8437020343628896e-06, | |
| "loss": 0.3935, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 2.346717301291877, | |
| "learning_rate": 2.8285842216464544e-06, | |
| "loss": 0.4093, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 2.4385081373620197, | |
| "learning_rate": 2.813454172125389e-06, | |
| "loss": 0.4007, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 2.0580498358246557, | |
| "learning_rate": 2.79831244925785e-06, | |
| "loss": 0.4012, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 1.8904939031404155, | |
| "learning_rate": 2.783159616936723e-06, | |
| "loss": 0.4247, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.939667430958579, | |
| "learning_rate": 2.76799623946862e-06, | |
| "loss": 0.3983, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.9363488042991572, | |
| "learning_rate": 2.7528228815528622e-06, | |
| "loss": 0.4054, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 1.9510664856850064, | |
| "learning_rate": 2.7376401082604563e-06, | |
| "loss": 0.4098, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 1.838136670411088, | |
| "learning_rate": 2.722448485013046e-06, | |
| "loss": 0.3997, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 1.792723761937039, | |
| "learning_rate": 2.707248577561854e-06, | |
| "loss": 0.4029, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 1.8102828207451078, | |
| "learning_rate": 2.6920409519666173e-06, | |
| "loss": 0.4249, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 2.1220720458624465, | |
| "learning_rate": 2.6768261745745037e-06, | |
| "loss": 0.4106, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 1.812641713649887, | |
| "learning_rate": 2.6616048119990214e-06, | |
| "loss": 0.3944, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 2.4648885003841046, | |
| "learning_rate": 2.6463774310989154e-06, | |
| "loss": 0.3971, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 1.85108667649988, | |
| "learning_rate": 2.6311445989570633e-06, | |
| "loss": 0.3975, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 1.7524914725271679, | |
| "learning_rate": 2.615906882859349e-06, | |
| "loss": 0.392, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 2.0229317958148867, | |
| "learning_rate": 2.6006648502735384e-06, | |
| "loss": 0.4123, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 1.8570564954534334, | |
| "learning_rate": 2.585419068828152e-06, | |
| "loss": 0.4174, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 1.8937349012574363, | |
| "learning_rate": 2.5701701062913194e-06, | |
| "loss": 0.4265, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 1.7590069556724997, | |
| "learning_rate": 2.554918530549637e-06, | |
| "loss": 0.3863, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 2.096408205769661, | |
| "learning_rate": 2.53966490958702e-06, | |
| "loss": 0.396, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 1.962464797958802, | |
| "learning_rate": 2.5244098114635503e-06, | |
| "loss": 0.4142, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 1.8458575978769718, | |
| "learning_rate": 2.5091538042943183e-06, | |
| "loss": 0.3946, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 1.8269976161284471, | |
| "learning_rate": 2.4938974562282708e-06, | |
| "loss": 0.4127, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 2.01278268309966, | |
| "learning_rate": 2.4786413354270494e-06, | |
| "loss": 0.4032, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 1.7849542357371841, | |
| "learning_rate": 2.4633860100438317e-06, | |
| "loss": 0.3905, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 69.63657032212085, | |
| "learning_rate": 2.4481320482021716e-06, | |
| "loss": 0.4038, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 2.1742738648421023, | |
| "learning_rate": 2.4328800179748475e-06, | |
| "loss": 0.3888, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 2.8552125373844985, | |
| "learning_rate": 2.4176304873626983e-06, | |
| "loss": 0.4066, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 2.1841551939924244, | |
| "learning_rate": 2.4023840242734774e-06, | |
| "loss": 0.3997, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 2.194040510042996, | |
| "learning_rate": 2.3871411965006985e-06, | |
| "loss": 0.4102, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 2.3916077476597475, | |
| "learning_rate": 2.3719025717024946e-06, | |
| "loss": 0.4014, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 1.9141926801542115, | |
| "learning_rate": 2.3566687173804747e-06, | |
| "loss": 0.3977, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 1.8144717702331585, | |
| "learning_rate": 2.341440200858589e-06, | |
| "loss": 0.3964, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 1.851593659208308, | |
| "learning_rate": 2.3262175892620064e-06, | |
| "loss": 0.4085, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 1.8686287238442707, | |
| "learning_rate": 2.311001449495986e-06, | |
| "loss": 0.4002, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 1.898471354316753, | |
| "learning_rate": 2.2957923482247745e-06, | |
| "loss": 0.399, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.8645673290000866, | |
| "learning_rate": 2.280590851850493e-06, | |
| "loss": 0.4025, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.8195968879508377, | |
| "learning_rate": 2.265397526492052e-06, | |
| "loss": 0.3942, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 2.133179133462322, | |
| "learning_rate": 2.2502129379640644e-06, | |
| "loss": 0.4135, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 1.790567048936156, | |
| "learning_rate": 2.235037651755773e-06, | |
| "loss": 0.3889, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 1.7690615361620154, | |
| "learning_rate": 2.2198722330099964e-06, | |
| "loss": 0.3971, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 2.0091067058646543, | |
| "learning_rate": 2.2047172465020757e-06, | |
| "loss": 0.3953, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 1.9832659283001421, | |
| "learning_rate": 2.1895732566188475e-06, | |
| "loss": 0.396, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 1.7695360334830164, | |
| "learning_rate": 2.1744408273376204e-06, | |
| "loss": 0.3864, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 1.9266047240521331, | |
| "learning_rate": 2.159320522205179e-06, | |
| "loss": 0.3902, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 1.9067745089946206, | |
| "learning_rate": 2.1442129043167877e-06, | |
| "loss": 0.4048, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 1.8121353203087656, | |
| "learning_rate": 2.1291185362952274e-06, | |
| "loss": 0.3949, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 1.8495794192789976, | |
| "learning_rate": 2.114037980269842e-06, | |
| "loss": 0.4038, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 1.912104342675586, | |
| "learning_rate": 2.0989717978555992e-06, | |
| "loss": 0.3919, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 1.8275208022684235, | |
| "learning_rate": 2.0839205501321844e-06, | |
| "loss": 0.4066, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.057130603594985, | |
| "learning_rate": 2.0688847976230952e-06, | |
| "loss": 0.3858, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.377420037984848, | |
| "eval_runtime": 330.5022, | |
| "eval_samples_per_second": 22.756, | |
| "eval_steps_per_second": 0.357, | |
| "step": 1716 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 3.3388932256530697, | |
| "learning_rate": 2.0538651002747745e-06, | |
| "loss": 0.2972, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 2.8981240612419423, | |
| "learning_rate": 2.0388620174357542e-06, | |
| "loss": 0.3002, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 2.205591361217027, | |
| "learning_rate": 2.023876107835825e-06, | |
| "loss": 0.3135, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 2.1495938527408835, | |
| "learning_rate": 2.008907929565231e-06, | |
| "loss": 0.3126, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 2.3520040686767167, | |
| "learning_rate": 1.993958040053881e-06, | |
| "loss": 0.3209, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 2.0650913293200848, | |
| "learning_rate": 1.9790269960505947e-06, | |
| "loss": 0.2944, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 2.1220138588069704, | |
| "learning_rate": 1.9641153536023646e-06, | |
| "loss": 0.308, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 2.150988951055739, | |
| "learning_rate": 1.9492236680336486e-06, | |
| "loss": 0.2931, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 2.124000470558878, | |
| "learning_rate": 1.934352493925695e-06, | |
| "loss": 0.2995, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 1.9934393020695829, | |
| "learning_rate": 1.9195023850958812e-06, | |
| "loss": 0.2926, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 2.154505232085453, | |
| "learning_rate": 1.9046738945770932e-06, | |
| "loss": 0.3057, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 2.1900958746909565, | |
| "learning_rate": 1.889867574597129e-06, | |
| "loss": 0.3012, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 1.8955650762693546, | |
| "learning_rate": 1.875083976558136e-06, | |
| "loss": 0.2838, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 2.0442738100465783, | |
| "learning_rate": 1.860323651016072e-06, | |
| "loss": 0.2977, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 2.110043278615698, | |
| "learning_rate": 1.8455871476602023e-06, | |
| "loss": 0.2984, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 2.3136510034991282, | |
| "learning_rate": 1.8308750152926338e-06, | |
| "loss": 0.2927, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 2.0131063879803826, | |
| "learning_rate": 1.8161878018078693e-06, | |
| "loss": 0.2908, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 2.1355586799767674, | |
| "learning_rate": 1.8015260541724128e-06, | |
| "loss": 0.2905, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 2.0664019629242087, | |
| "learning_rate": 1.7868903184043888e-06, | |
| "loss": 0.2955, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 2.188481583652719, | |
| "learning_rate": 1.772281139553218e-06, | |
| "loss": 0.3071, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 2.0847120701500774, | |
| "learning_rate": 1.7576990616793139e-06, | |
| "loss": 0.2908, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 2.018064853932535, | |
| "learning_rate": 1.74314462783382e-06, | |
| "loss": 0.2967, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 2.0242538886830843, | |
| "learning_rate": 1.7286183800383937e-06, | |
| "loss": 0.3017, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 2.145977098033245, | |
| "learning_rate": 1.714120859265011e-06, | |
| "loss": 0.2872, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 1.9029268616901622, | |
| "learning_rate": 1.6996526054158283e-06, | |
| "loss": 0.2888, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 4.096519984140158, | |
| "learning_rate": 1.685214157303069e-06, | |
| "loss": 0.2966, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 2.1188967978481013, | |
| "learning_rate": 1.6708060526289648e-06, | |
| "loss": 0.299, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 2.7278823372957666, | |
| "learning_rate": 1.6564288279657253e-06, | |
| "loss": 0.3034, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 2.0002017052238257, | |
| "learning_rate": 1.6420830187355572e-06, | |
| "loss": 0.2958, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "grad_norm": 3.6893974280355097, | |
| "learning_rate": 1.6277691591907272e-06, | |
| "loss": 0.2954, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 1.8667517556916042, | |
| "learning_rate": 1.613487782393661e-06, | |
| "loss": 0.2933, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "grad_norm": 2.0838269309261364, | |
| "learning_rate": 1.599239420197098e-06, | |
| "loss": 0.297, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 1.9755359084479687, | |
| "learning_rate": 1.5850246032242766e-06, | |
| "loss": 0.2909, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 2096.7504686136917, | |
| "learning_rate": 1.5708438608491816e-06, | |
| "loss": 0.2951, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 2.918573785412608, | |
| "learning_rate": 1.556697721176823e-06, | |
| "loss": 0.3012, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 4.274328804415084, | |
| "learning_rate": 1.5425867110235717e-06, | |
| "loss": 0.2876, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 5.92446369343208, | |
| "learning_rate": 1.5285113558975429e-06, | |
| "loss": 0.3466, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 5.122609680763134, | |
| "learning_rate": 1.5144721799790194e-06, | |
| "loss": 0.3023, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 5.036779121717785, | |
| "learning_rate": 1.5004697061009372e-06, | |
| "loss": 0.3013, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 4.110935956575215, | |
| "learning_rate": 1.486504455729408e-06, | |
| "loss": 0.2915, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 2.867011840176696, | |
| "learning_rate": 1.4725769489443082e-06, | |
| "loss": 0.2822, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 2.9818315723679327, | |
| "learning_rate": 1.4586877044199015e-06, | |
| "loss": 0.286, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 3.284507097488489, | |
| "learning_rate": 1.4448372394055249e-06, | |
| "loss": 0.3031, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "grad_norm": 2.6446231205609907, | |
| "learning_rate": 1.431026069706335e-06, | |
| "loss": 0.2951, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 2.429736926817698, | |
| "learning_rate": 1.4172547096640837e-06, | |
| "loss": 0.2948, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 2.4087174238529565, | |
| "learning_rate": 1.4035236721379758e-06, | |
| "loss": 0.3079, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 2.346665896073256, | |
| "learning_rate": 1.3898334684855647e-06, | |
| "loss": 0.2884, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "grad_norm": 2.2112665369477806, | |
| "learning_rate": 1.376184608543709e-06, | |
| "loss": 0.2784, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 2.3555732075174705, | |
| "learning_rate": 1.3625776006095882e-06, | |
| "loss": 0.2951, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 2.453057863316602, | |
| "learning_rate": 1.3490129514217665e-06, | |
| "loss": 0.2912, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 2.349215074609662, | |
| "learning_rate": 1.3354911661413305e-06, | |
| "loss": 0.3175, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 2.369663823495378, | |
| "learning_rate": 1.3220127483330714e-06, | |
| "loss": 0.3115, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 2.353086780550841, | |
| "learning_rate": 1.3085781999467303e-06, | |
| "loss": 0.2974, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 2.283791495168621, | |
| "learning_rate": 1.2951880212983106e-06, | |
| "loss": 0.2963, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 2.285838470923522, | |
| "learning_rate": 1.2818427110514382e-06, | |
| "loss": 0.2925, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 2.0947843282235863, | |
| "learning_rate": 1.2685427661987975e-06, | |
| "loss": 0.2944, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 2.181297764037496, | |
| "learning_rate": 1.2552886820436208e-06, | |
| "loss": 0.2821, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 2.2572030477763847, | |
| "learning_rate": 1.2420809521812406e-06, | |
| "loss": 0.299, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 2.2492785376097677, | |
| "learning_rate": 1.2289200684807098e-06, | |
| "loss": 0.3049, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 2.334560206327675, | |
| "learning_rate": 1.2158065210664848e-06, | |
| "loss": 0.2877, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "grad_norm": 2.0972126060506375, | |
| "learning_rate": 1.2027407983001683e-06, | |
| "loss": 0.2818, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 2.545800792955665, | |
| "learning_rate": 1.189723386762328e-06, | |
| "loss": 0.2927, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 2.2889188636916624, | |
| "learning_rate": 1.1767547712343722e-06, | |
| "loss": 0.3015, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 2.073915350133538, | |
| "learning_rate": 1.1638354346804974e-06, | |
| "loss": 0.2784, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 2.2964925609294826, | |
| "learning_rate": 1.1509658582297025e-06, | |
| "loss": 0.2906, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 2.12295197546072, | |
| "learning_rate": 1.1381465211578673e-06, | |
| "loss": 0.2894, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 21.13479225301652, | |
| "learning_rate": 1.1253779008699131e-06, | |
| "loss": 0.2995, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "grad_norm": 1.9921395090376566, | |
| "learning_rate": 1.1126604728820102e-06, | |
| "loss": 0.2766, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 2.1541306330776258, | |
| "learning_rate": 1.0999947108038816e-06, | |
| "loss": 0.3002, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 2.3718321918672824, | |
| "learning_rate": 1.0873810863211595e-06, | |
| "loss": 0.2894, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 2.106333020947831, | |
| "learning_rate": 1.074820069177816e-06, | |
| "loss": 0.3026, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 2.232973434321066, | |
| "learning_rate": 1.0623121271586806e-06, | |
| "loss": 0.287, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 2.150526937883329, | |
| "learning_rate": 1.049857726072005e-06, | |
| "loss": 0.3079, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 2.4299034348212034, | |
| "learning_rate": 1.037457329732127e-06, | |
| "loss": 0.2912, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 2.2110962609469644, | |
| "learning_rate": 1.0251113999421936e-06, | |
| "loss": 0.2983, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 2.2895025988377857, | |
| "learning_rate": 1.0128203964769602e-06, | |
| "loss": 0.2804, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 1.9663163475739167, | |
| "learning_rate": 1.0005847770656757e-06, | |
| "loss": 0.3099, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 1.955773416240966, | |
| "learning_rate": 9.884049973750268e-07, | |
| "loss": 0.2843, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 2.151841568064239, | |
| "learning_rate": 9.762815109921762e-07, | |
| "loss": 0.2898, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 2.2055014532745925, | |
| "learning_rate": 9.642147694078664e-07, | |
| "loss": 0.2922, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 2.0548163347288955, | |
| "learning_rate": 9.522052219996072e-07, | |
| "loss": 0.2875, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 2.0264167858867554, | |
| "learning_rate": 9.402533160149415e-07, | |
| "loss": 0.2916, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 1.9310228298997407, | |
| "learning_rate": 9.283594965547846e-07, | |
| "loss": 0.288, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 2.0775492416338994, | |
| "learning_rate": 9.165242065568547e-07, | |
| "loss": 0.3013, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 4.273341280017886, | |
| "learning_rate": 9.047478867791732e-07, | |
| "loss": 0.2963, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 2.2046805609188, | |
| "learning_rate": 8.930309757836517e-07, | |
| "loss": 0.2884, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 2.0943612862709022, | |
| "learning_rate": 8.813739099197597e-07, | |
| "loss": 0.288, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 2.470082871683493, | |
| "learning_rate": 8.697771233082744e-07, | |
| "loss": 0.2958, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 2.514580950169932, | |
| "learning_rate": 8.582410478251119e-07, | |
| "loss": 0.2877, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 2.4225946060971326, | |
| "learning_rate": 8.467661130852464e-07, | |
| "loss": 0.2964, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 2.5213364012816752, | |
| "learning_rate": 8.353527464267105e-07, | |
| "loss": 0.3139, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 2.1383450216449766, | |
| "learning_rate": 8.240013728946786e-07, | |
| "loss": 0.3011, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 2.275623309460668, | |
| "learning_rate": 8.127124152256408e-07, | |
| "loss": 0.2877, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 2.3231659654583945, | |
| "learning_rate": 8.014862938316542e-07, | |
| "loss": 0.2877, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 2.034432326023906, | |
| "learning_rate": 7.903234267846965e-07, | |
| "loss": 0.2891, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 2.1427143907161024, | |
| "learning_rate": 7.792242298010847e-07, | |
| "loss": 0.2905, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 1.992125321368494, | |
| "learning_rate": 7.681891162260016e-07, | |
| "loss": 0.2904, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 2.2912621172371073, | |
| "learning_rate": 7.572184970181005e-07, | |
| "loss": 0.2931, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 2.1841846462580303, | |
| "learning_rate": 7.463127807341966e-07, | |
| "loss": 0.2888, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 2.4186238871530947, | |
| "learning_rate": 7.354723735140609e-07, | |
| "loss": 0.2972, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 2.165303668523032, | |
| "learning_rate": 7.246976790652843e-07, | |
| "loss": 0.2894, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 1.971010064873144, | |
| "learning_rate": 7.139890986482515e-07, | |
| "loss": 0.2801, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 2.226759681716193, | |
| "learning_rate": 7.033470310611945e-07, | |
| "loss": 0.2936, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 2.128956719636366, | |
| "learning_rate": 6.927718726253379e-07, | |
| "loss": 0.2873, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 1.9680023419439248, | |
| "learning_rate": 6.822640171701486e-07, | |
| "loss": 0.2946, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 2.2622423856131344, | |
| "learning_rate": 6.718238560186572e-07, | |
| "loss": 0.2825, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 1.947066365404126, | |
| "learning_rate": 6.614517779728943e-07, | |
| "loss": 0.2893, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 2.2897303224060064, | |
| "learning_rate": 6.511481692994077e-07, | |
| "loss": 0.2816, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 2.1661490360203937, | |
| "learning_rate": 6.409134137148737e-07, | |
| "loss": 0.2961, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 2.163171587598097, | |
| "learning_rate": 6.307478923718171e-07, | |
| "loss": 0.2858, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 2.03844883078776, | |
| "learning_rate": 6.206519838444044e-07, | |
| "loss": 0.2802, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 1.950829082933861, | |
| "learning_rate": 6.106260641143547e-07, | |
| "loss": 0.3058, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 1.971999267799782, | |
| "learning_rate": 6.006705065569329e-07, | |
| "loss": 0.2809, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 1.9523923513832082, | |
| "learning_rate": 5.907856819270471e-07, | |
| "loss": 0.281, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.3295736014842987, | |
| "eval_runtime": 330.2445, | |
| "eval_samples_per_second": 22.774, | |
| "eval_steps_per_second": 0.357, | |
| "step": 2288 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.044140540928026, | |
| "learning_rate": 5.809719583454415e-07, | |
| "loss": 0.2618, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 2.5480857730573305, | |
| "learning_rate": 5.712297012849826e-07, | |
| "loss": 0.2349, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 2.2011961310601276, | |
| "learning_rate": 5.615592735570563e-07, | |
| "loss": 0.2555, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 2.1457420228511497, | |
| "learning_rate": 5.519610352980501e-07, | |
| "loss": 0.2208, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 1.989514783963616, | |
| "learning_rate": 5.424353439559446e-07, | |
| "loss": 0.24, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 5.657828438864317, | |
| "learning_rate": 5.32982554277e-07, | |
| "loss": 0.2344, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 2.1124182295185783, | |
| "learning_rate": 5.236030182925475e-07, | |
| "loss": 0.2332, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 2.1054589415417797, | |
| "learning_rate": 5.142970853058743e-07, | |
| "loss": 0.2264, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "grad_norm": 2.1186956472831557, | |
| "learning_rate": 5.050651018792213e-07, | |
| "loss": 0.2331, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 1.9513447351491608, | |
| "learning_rate": 4.959074118208726e-07, | |
| "loss": 0.2243, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "grad_norm": 2.020062184794575, | |
| "learning_rate": 4.868243561723535e-07, | |
| "loss": 0.2467, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 2.180369971274554, | |
| "learning_rate": 4.7781627319573e-07, | |
| "loss": 0.2396, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 1.9826937910162328, | |
| "learning_rate": 4.688834983610083e-07, | |
| "loss": 0.23, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 1.9374964597970792, | |
| "learning_rate": 4.6002636433364836e-07, | |
| "loss": 0.233, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 2.2161679673685364, | |
| "learning_rate": 4.512452009621665e-07, | |
| "loss": 0.2355, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 2.0656182008910258, | |
| "learning_rate": 4.4254033526585917e-07, | |
| "loss": 0.2298, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 1.9301077360385739, | |
| "learning_rate": 4.3391209142261996e-07, | |
| "loss": 0.2244, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "grad_norm": 2.019772633601012, | |
| "learning_rate": 4.2536079075686687e-07, | |
| "loss": 0.239, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 1.9236814219014258, | |
| "learning_rate": 4.168867517275807e-07, | |
| "loss": 0.2294, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 1.9496838165175443, | |
| "learning_rate": 4.0849028991643726e-07, | |
| "loss": 0.2351, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "grad_norm": 2.0751662501448687, | |
| "learning_rate": 4.0017171801606245e-07, | |
| "loss": 0.2328, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "grad_norm": 2.030778712243648, | |
| "learning_rate": 3.9193134581838375e-07, | |
| "loss": 0.2193, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 2.1108537998820696, | |
| "learning_rate": 3.8376948020309083e-07, | |
| "loss": 0.2259, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 2.0388571280858003, | |
| "learning_rate": 3.756864251262143e-07, | |
| "loss": 0.2335, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 2.4800994906904616, | |
| "learning_rate": 3.6768248160879786e-07, | |
| "loss": 0.2359, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "grad_norm": 2.0463977880911193, | |
| "learning_rate": 3.597579477256932e-07, | |
| "loss": 0.2296, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "grad_norm": 1.980879433572184, | |
| "learning_rate": 3.51913118594458e-07, | |
| "loss": 0.2244, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 2.201271813612607, | |
| "learning_rate": 3.4414828636436525e-07, | |
| "loss": 0.2355, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 1.9358796946633714, | |
| "learning_rate": 3.364637402055235e-07, | |
| "loss": 0.2235, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "grad_norm": 1.9943772531112582, | |
| "learning_rate": 3.288597662981069e-07, | |
| "loss": 0.2363, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 1.9701699768295131, | |
| "learning_rate": 3.2133664782169947e-07, | |
| "loss": 0.2211, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 1.9846134376402154, | |
| "learning_rate": 3.138946649447483e-07, | |
| "loss": 0.2421, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "grad_norm": 2.1267366719554084, | |
| "learning_rate": 3.0653409481412906e-07, | |
| "loss": 0.232, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 2.025895740497588, | |
| "learning_rate": 2.9925521154482577e-07, | |
| "loss": 0.2327, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 2.0801594057057304, | |
| "learning_rate": 2.9205828620972267e-07, | |
| "loss": 0.229, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 2.1905576705856413, | |
| "learning_rate": 2.8494358682950725e-07, | |
| "loss": 0.2343, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 2.104131043625037, | |
| "learning_rate": 2.779113783626916e-07, | |
| "loss": 0.2324, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "grad_norm": 1.9860664411869802, | |
| "learning_rate": 2.70961922695743e-07, | |
| "loss": 0.2238, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 2.0126712420413324, | |
| "learning_rate": 2.6409547863333246e-07, | |
| "loss": 0.2203, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 1.9772873060488734, | |
| "learning_rate": 2.573123018886961e-07, | |
| "loss": 0.231, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "grad_norm": 1.8887328536766113, | |
| "learning_rate": 2.5061264507411057e-07, | |
| "loss": 0.2135, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 2.1093653326387933, | |
| "learning_rate": 2.4399675769148784e-07, | |
| "loss": 0.2262, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 2.209645499404587, | |
| "learning_rate": 2.37464886123083e-07, | |
| "loss": 0.2307, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 2.0867824783223208, | |
| "learning_rate": 2.3101727362231762e-07, | |
| "loss": 0.2363, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 2.1576991729700814, | |
| "learning_rate": 2.2465416030472227e-07, | |
| "loss": 0.2331, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 2.0751768584129495, | |
| "learning_rate": 2.18375783138991e-07, | |
| "loss": 0.2243, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 1.8719170813482977, | |
| "learning_rate": 2.1218237593816305e-07, | |
| "loss": 0.2366, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 1.9719981089033465, | |
| "learning_rate": 2.0607416935090746e-07, | |
| "loss": 0.2239, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "grad_norm": 2.1100472183339276, | |
| "learning_rate": 2.0005139085293945e-07, | |
| "loss": 0.2384, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "grad_norm": 2.194632892084896, | |
| "learning_rate": 1.941142647385469e-07, | |
| "loss": 0.2311, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 2.0340858963446475, | |
| "learning_rate": 1.882630121122353e-07, | |
| "loss": 0.2358, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 1.9791028474343204, | |
| "learning_rate": 1.8249785088049894e-07, | |
| "loss": 0.2389, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 2.2469036964697082, | |
| "learning_rate": 1.7681899574369916e-07, | |
| "loss": 0.2373, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "grad_norm": 1.9763859240449058, | |
| "learning_rate": 1.7122665818807478e-07, | |
| "loss": 0.2242, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 2.2911402959565224, | |
| "learning_rate": 1.6572104647786247e-07, | |
| "loss": 0.2411, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 2.083914931198256, | |
| "learning_rate": 1.6030236564754054e-07, | |
| "loss": 0.2228, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "grad_norm": 1.9520019226109886, | |
| "learning_rate": 1.5497081749419745e-07, | |
| "loss": 0.223, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 2.000958322889037, | |
| "learning_rate": 1.497266005700107e-07, | |
| "loss": 0.2366, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "grad_norm": 2.255516690447902, | |
| "learning_rate": 1.4456991017485737e-07, | |
| "loss": 0.2228, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "grad_norm": 1.9859817862123228, | |
| "learning_rate": 1.3950093834903865e-07, | |
| "loss": 0.2277, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "grad_norm": 1.9478379602583347, | |
| "learning_rate": 1.3451987386612852e-07, | |
| "loss": 0.2303, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "grad_norm": 2.0551925040719135, | |
| "learning_rate": 1.2962690222594444e-07, | |
| "loss": 0.2213, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 1.9946612862734567, | |
| "learning_rate": 1.2482220564763669e-07, | |
| "loss": 0.2332, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 2.013120854912158, | |
| "learning_rate": 1.2010596306290588e-07, | |
| "loss": 0.2171, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "grad_norm": 3.943001009546542, | |
| "learning_rate": 1.154783501093365e-07, | |
| "loss": 0.2351, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "grad_norm": 2.077111667283586, | |
| "learning_rate": 1.1093953912385769e-07, | |
| "loss": 0.2227, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 1.9319320014149188, | |
| "learning_rate": 1.0648969913632401e-07, | |
| "loss": 0.2349, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 2.0904879562505125, | |
| "learning_rate": 1.0212899586322183e-07, | |
| "loss": 0.2353, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 1.9353613819189108, | |
| "learning_rate": 9.785759170149622e-08, | |
| "loss": 0.2207, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "grad_norm": 2.070941108809218, | |
| "learning_rate": 9.36756457225052e-08, | |
| "loss": 0.2333, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 1.962848104620626, | |
| "learning_rate": 8.958331366609424e-08, | |
| "loss": 0.2268, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 2.0575315291559857, | |
| "learning_rate": 8.55807479347967e-08, | |
| "loss": 0.2253, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "grad_norm": 2.107360332698912, | |
| "learning_rate": 8.166809758815897e-08, | |
| "loss": 0.224, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 2.144090062153445, | |
| "learning_rate": 7.784550833718707e-08, | |
| "loss": 0.2365, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 1.9482188832166278, | |
| "learning_rate": 7.411312253892466e-08, | |
| "loss": 0.2294, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "grad_norm": 2.1524181670324616, | |
| "learning_rate": 7.047107919114588e-08, | |
| "loss": 0.2336, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 2.152296301431861, | |
| "learning_rate": 6.691951392718332e-08, | |
| "loss": 0.2258, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "grad_norm": 1.9081910207653139, | |
| "learning_rate": 6.345855901087522e-08, | |
| "loss": 0.2338, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 1.938526307346203, | |
| "learning_rate": 6.008834333163876e-08, | |
| "loss": 0.2195, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 1.9617786271701327, | |
| "learning_rate": 5.680899239967369e-08, | |
| "loss": 0.2337, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 2.070106598519499, | |
| "learning_rate": 5.3620628341283234e-08, | |
| "loss": 0.2339, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 2.063857678666163, | |
| "learning_rate": 5.052336989433082e-08, | |
| "loss": 0.2424, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "grad_norm": 1.9748870315279254, | |
| "learning_rate": 4.75173324038139e-08, | |
| "loss": 0.2284, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 2.001809998998157, | |
| "learning_rate": 4.4602627817571245e-08, | |
| "loss": 0.2358, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "grad_norm": 1.9729922140795935, | |
| "learning_rate": 4.1779364682113796e-08, | |
| "loss": 0.233, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 2.045685009725174, | |
| "learning_rate": 3.904764813858014e-08, | |
| "loss": 0.2323, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "grad_norm": 1.9591449865359234, | |
| "learning_rate": 3.640757991882349e-08, | |
| "loss": 0.2349, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "grad_norm": 1.9911019888259411, | |
| "learning_rate": 3.385925834162113e-08, | |
| "loss": 0.2257, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "grad_norm": 2.050714072028577, | |
| "learning_rate": 3.1402778309014284e-08, | |
| "loss": 0.2299, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "grad_norm": 2.0356703719044518, | |
| "learning_rate": 2.903823130277289e-08, | |
| "loss": 0.2277, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 1.9823356155432006, | |
| "learning_rate": 2.676570538098944e-08, | |
| "loss": 0.2431, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 2.0568572715035063, | |
| "learning_rate": 2.4585285174799357e-08, | |
| "loss": 0.2331, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "grad_norm": 2.195948388536599, | |
| "learning_rate": 2.2497051885228825e-08, | |
| "loss": 0.2308, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "grad_norm": 2.1647004516523523, | |
| "learning_rate": 2.050108328017164e-08, | |
| "loss": 0.2418, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "grad_norm": 2.3975608426150656, | |
| "learning_rate": 1.8597453691492628e-08, | |
| "loss": 0.2318, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "grad_norm": 2.0876506445205716, | |
| "learning_rate": 1.678623401225876e-08, | |
| "loss": 0.2227, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 1.9235370393907136, | |
| "learning_rate": 1.5067491694100156e-08, | |
| "loss": 0.2317, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "grad_norm": 2.166013576809247, | |
| "learning_rate": 1.3441290744697632e-08, | |
| "loss": 0.2281, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "grad_norm": 1.9685518937742286, | |
| "learning_rate": 1.1907691725398795e-08, | |
| "loss": 0.2238, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "grad_norm": 1.9483454828149802, | |
| "learning_rate": 1.0466751748963443e-08, | |
| "loss": 0.2304, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "grad_norm": 1.9753685457102212, | |
| "learning_rate": 9.118524477434999e-09, | |
| "loss": 0.2365, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 1.8854947658053076, | |
| "learning_rate": 7.863060120144316e-09, | |
| "loss": 0.2142, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 2.154842589588192, | |
| "learning_rate": 6.700405431837587e-09, | |
| "loss": 0.2315, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 2.0037090722639506, | |
| "learning_rate": 5.6306037109371544e-09, | |
| "loss": 0.2225, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "grad_norm": 1.9552444706301575, | |
| "learning_rate": 4.653694797927544e-09, | |
| "loss": 0.2331, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "grad_norm": 2.563219810008139, | |
| "learning_rate": 3.769715073872749e-09, | |
| "loss": 0.2407, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "grad_norm": 1.9200483422222523, | |
| "learning_rate": 2.978697459060098e-09, | |
| "loss": 0.23, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 1.9640394162849764, | |
| "learning_rate": 2.280671411776514e-09, | |
| "loss": 0.2264, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "grad_norm": 2.1505121388781467, | |
| "learning_rate": 1.6756629272085545e-09, | |
| "loss": 0.2278, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 1.9911030010645636, | |
| "learning_rate": 1.1636945364768005e-09, | |
| "loss": 0.2341, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 2.0018264068324187, | |
| "learning_rate": 7.447853057954146e-10, | |
| "loss": 0.2185, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 1.9862221260079735, | |
| "learning_rate": 4.1895083576271035e-10, | |
| "loss": 0.221, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "grad_norm": 1.9962774038489774, | |
| "learning_rate": 1.8620326077967155e-10, | |
| "loss": 0.2316, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "grad_norm": 1.9534249748428105, | |
| "learning_rate": 4.6551248598647457e-11, | |
| "loss": 0.2333, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.0900193031037806, | |
| "learning_rate": 0.0, | |
| "loss": 0.229, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.3284313678741455, | |
| "eval_runtime": 329.743, | |
| "eval_samples_per_second": 22.809, | |
| "eval_steps_per_second": 0.358, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 2860, | |
| "total_flos": 2395303260979200.0, | |
| "train_loss": 0.44255039914921446, | |
| "train_runtime": 63778.4258, | |
| "train_samples_per_second": 5.733, | |
| "train_steps_per_second": 0.045 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 2860, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 2395303260979200.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |