| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 931, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010741138560687433, | |
| "grad_norm": 22.91085751330113, | |
| "learning_rate": 1.0638297872340426e-07, | |
| "loss": 1.3283, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0053705692803437165, | |
| "grad_norm": 20.288173310717095, | |
| "learning_rate": 5.319148936170213e-07, | |
| "loss": 1.3223, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010741138560687433, | |
| "grad_norm": 8.769257685223952, | |
| "learning_rate": 1.0638297872340427e-06, | |
| "loss": 1.1992, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01611170784103115, | |
| "grad_norm": 8.213788629020675, | |
| "learning_rate": 1.595744680851064e-06, | |
| "loss": 1.0448, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.021482277121374866, | |
| "grad_norm": 3.1269756077572763, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "loss": 0.9214, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02685284640171858, | |
| "grad_norm": 2.4050803000452614, | |
| "learning_rate": 2.6595744680851065e-06, | |
| "loss": 0.866, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0322234156820623, | |
| "grad_norm": 2.265373196640215, | |
| "learning_rate": 3.191489361702128e-06, | |
| "loss": 0.8439, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03759398496240601, | |
| "grad_norm": 2.3896948871947226, | |
| "learning_rate": 3.723404255319149e-06, | |
| "loss": 0.8168, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04296455424274973, | |
| "grad_norm": 2.2142847727025745, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 0.8029, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04833512352309345, | |
| "grad_norm": 2.615386014469481, | |
| "learning_rate": 4.787234042553192e-06, | |
| "loss": 0.7905, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05370569280343716, | |
| "grad_norm": 2.4265488613802453, | |
| "learning_rate": 5.319148936170213e-06, | |
| "loss": 0.7719, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05907626208378088, | |
| "grad_norm": 2.3528826351801997, | |
| "learning_rate": 5.851063829787235e-06, | |
| "loss": 0.7623, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0644468313641246, | |
| "grad_norm": 2.399639169785202, | |
| "learning_rate": 6.382978723404256e-06, | |
| "loss": 0.7299, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06981740064446831, | |
| "grad_norm": 2.3134344809673117, | |
| "learning_rate": 6.914893617021278e-06, | |
| "loss": 0.7273, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07518796992481203, | |
| "grad_norm": 2.4544444295200263, | |
| "learning_rate": 7.446808510638298e-06, | |
| "loss": 0.7256, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08055853920515575, | |
| "grad_norm": 2.396543440801372, | |
| "learning_rate": 7.97872340425532e-06, | |
| "loss": 0.7205, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08592910848549946, | |
| "grad_norm": 2.334064555082653, | |
| "learning_rate": 8.510638297872341e-06, | |
| "loss": 0.7084, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09129967776584318, | |
| "grad_norm": 2.3827808498057625, | |
| "learning_rate": 9.042553191489362e-06, | |
| "loss": 0.6977, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0966702470461869, | |
| "grad_norm": 2.218302672309819, | |
| "learning_rate": 9.574468085106385e-06, | |
| "loss": 0.7009, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10204081632653061, | |
| "grad_norm": 2.190020287235194, | |
| "learning_rate": 9.999964780082996e-06, | |
| "loss": 0.6972, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10741138560687433, | |
| "grad_norm": 2.3225529019292446, | |
| "learning_rate": 9.998732135085665e-06, | |
| "loss": 0.6886, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11278195488721804, | |
| "grad_norm": 2.2648438359026692, | |
| "learning_rate": 9.995738990383743e-06, | |
| "loss": 0.6858, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11815252416756176, | |
| "grad_norm": 2.1052133128056227, | |
| "learning_rate": 9.990986400130607e-06, | |
| "loss": 0.6776, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12352309344790548, | |
| "grad_norm": 2.2304556352160976, | |
| "learning_rate": 9.984476038137437e-06, | |
| "loss": 0.6804, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1288936627282492, | |
| "grad_norm": 2.1997916038296537, | |
| "learning_rate": 9.97621019728372e-06, | |
| "loss": 0.6823, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13426423200859292, | |
| "grad_norm": 2.097027456664302, | |
| "learning_rate": 9.966191788709716e-06, | |
| "loss": 0.6724, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.13963480128893663, | |
| "grad_norm": 2.0742523673376296, | |
| "learning_rate": 9.954424340791195e-06, | |
| "loss": 0.6671, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14500537056928034, | |
| "grad_norm": 2.130928347720073, | |
| "learning_rate": 9.940911997896774e-06, | |
| "loss": 0.659, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.15037593984962405, | |
| "grad_norm": 2.524340696530577, | |
| "learning_rate": 9.925659518928316e-06, | |
| "loss": 0.6591, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.15574650912996776, | |
| "grad_norm": 2.2589554106263505, | |
| "learning_rate": 9.908672275644898e-06, | |
| "loss": 0.6626, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1611170784103115, | |
| "grad_norm": 2.0855685765048513, | |
| "learning_rate": 9.889956250770933e-06, | |
| "loss": 0.6507, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16648764769065522, | |
| "grad_norm": 1.9782423537163931, | |
| "learning_rate": 9.86951803588912e-06, | |
| "loss": 0.6607, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.17185821697099893, | |
| "grad_norm": 2.096081337731493, | |
| "learning_rate": 9.847364829118963e-06, | |
| "loss": 0.6443, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17722878625134264, | |
| "grad_norm": 2.00136387648786, | |
| "learning_rate": 9.82350443258166e-06, | |
| "loss": 0.6501, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.18259935553168635, | |
| "grad_norm": 1.9722375580956812, | |
| "learning_rate": 9.797945249652295e-06, | |
| "loss": 0.636, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.18796992481203006, | |
| "grad_norm": 1.9754080013745825, | |
| "learning_rate": 9.770696282000245e-06, | |
| "loss": 0.6337, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1933404940923738, | |
| "grad_norm": 2.100922511277814, | |
| "learning_rate": 9.741767126418898e-06, | |
| "loss": 0.6253, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19871106337271752, | |
| "grad_norm": 2.062680632629392, | |
| "learning_rate": 9.711167971445766e-06, | |
| "loss": 0.627, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.20408163265306123, | |
| "grad_norm": 2.021855902031637, | |
| "learning_rate": 9.67890959377418e-06, | |
| "loss": 0.6265, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.20945220193340494, | |
| "grad_norm": 2.09605687700288, | |
| "learning_rate": 9.645003354457872e-06, | |
| "loss": 0.6162, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.21482277121374865, | |
| "grad_norm": 1.9814268678267672, | |
| "learning_rate": 9.60946119490972e-06, | |
| "loss": 0.6261, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22019334049409237, | |
| "grad_norm": 2.035391481068172, | |
| "learning_rate": 9.57229563269612e-06, | |
| "loss": 0.6191, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.22556390977443608, | |
| "grad_norm": 2.0644565465007236, | |
| "learning_rate": 9.533519757128426e-06, | |
| "loss": 0.6077, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.23093447905477982, | |
| "grad_norm": 1.9165388707830449, | |
| "learning_rate": 9.49314722465304e-06, | |
| "loss": 0.6202, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.23630504833512353, | |
| "grad_norm": 1.8612773035887493, | |
| "learning_rate": 9.451192254041759e-06, | |
| "loss": 0.6071, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.24167561761546724, | |
| "grad_norm": 1.9195944320987655, | |
| "learning_rate": 9.407669621384073e-06, | |
| "loss": 0.6019, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.24704618689581095, | |
| "grad_norm": 1.9206849100012615, | |
| "learning_rate": 9.362594654883185e-06, | |
| "loss": 0.6049, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.25241675617615467, | |
| "grad_norm": 2.002233465125814, | |
| "learning_rate": 9.31598322945759e-06, | |
| "loss": 0.5965, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2577873254564984, | |
| "grad_norm": 1.8894973687886405, | |
| "learning_rate": 9.267851761150092e-06, | |
| "loss": 0.5989, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 2.072122453837846, | |
| "learning_rate": 9.218217201346251e-06, | |
| "loss": 0.5981, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.26852846401718583, | |
| "grad_norm": 1.9464798434012267, | |
| "learning_rate": 9.167097030804289e-06, | |
| "loss": 0.6001, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2738990332975295, | |
| "grad_norm": 2.001735827137876, | |
| "learning_rate": 9.114509253498554e-06, | |
| "loss": 0.5876, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.27926960257787325, | |
| "grad_norm": 1.9075332129313654, | |
| "learning_rate": 9.060472390278717e-06, | |
| "loss": 0.5832, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.284640171858217, | |
| "grad_norm": 2.0633775192756176, | |
| "learning_rate": 9.005005472346923e-06, | |
| "loss": 0.5824, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2900107411385607, | |
| "grad_norm": 1.8960322375768555, | |
| "learning_rate": 8.948128034555212e-06, | |
| "loss": 0.5921, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2953813104189044, | |
| "grad_norm": 1.9405235062135664, | |
| "learning_rate": 8.889860108525544e-06, | |
| "loss": 0.5735, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3007518796992481, | |
| "grad_norm": 2.080310458221302, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5625, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.30612244897959184, | |
| "grad_norm": 2.0619022130305784, | |
| "learning_rate": 8.76923535958783e-06, | |
| "loss": 0.5717, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.31149301825993553, | |
| "grad_norm": 1.9068651106559482, | |
| "learning_rate": 8.706921019419237e-06, | |
| "loss": 0.5591, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.31686358754027927, | |
| "grad_norm": 1.927395166865176, | |
| "learning_rate": 8.643301141529619e-06, | |
| "loss": 0.5878, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.322234156820623, | |
| "grad_norm": 1.9978060531735273, | |
| "learning_rate": 8.578398132155846e-06, | |
| "loss": 0.5745, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3276047261009667, | |
| "grad_norm": 1.865419008290949, | |
| "learning_rate": 8.512234849439887e-06, | |
| "loss": 0.5691, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.33297529538131043, | |
| "grad_norm": 1.8439698944678748, | |
| "learning_rate": 8.444834595378434e-06, | |
| "loss": 0.5476, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3383458646616541, | |
| "grad_norm": 1.9373123471183016, | |
| "learning_rate": 8.376221107616187e-06, | |
| "loss": 0.5463, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.34371643394199786, | |
| "grad_norm": 1.9054864517788856, | |
| "learning_rate": 8.306418551085707e-06, | |
| "loss": 0.5533, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.34908700322234154, | |
| "grad_norm": 1.8674970659214487, | |
| "learning_rate": 8.23545150949679e-06, | |
| "loss": 0.5459, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3544575725026853, | |
| "grad_norm": 1.9490992668548852, | |
| "learning_rate": 8.163344976678342e-06, | |
| "loss": 0.5557, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.359828141783029, | |
| "grad_norm": 2.0923835009467218, | |
| "learning_rate": 8.090124347775837e-06, | |
| "loss": 0.5469, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3651987110633727, | |
| "grad_norm": 1.9541886722127315, | |
| "learning_rate": 8.0158154103074e-06, | |
| "loss": 0.5523, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.37056928034371645, | |
| "grad_norm": 2.013088562636803, | |
| "learning_rate": 7.940444335081733e-06, | |
| "loss": 0.5418, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.37593984962406013, | |
| "grad_norm": 1.934795964689943, | |
| "learning_rate": 7.864037666981037e-06, | |
| "loss": 0.5166, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.38131041890440387, | |
| "grad_norm": 1.9010673558732618, | |
| "learning_rate": 7.786622315612182e-06, | |
| "loss": 0.5359, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3866809881847476, | |
| "grad_norm": 1.9210934256304713, | |
| "learning_rate": 7.708225545829446e-06, | |
| "loss": 0.5327, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3920515574650913, | |
| "grad_norm": 1.9088636324595145, | |
| "learning_rate": 7.6288749681321115e-06, | |
| "loss": 0.5303, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.39742212674543503, | |
| "grad_norm": 1.9478986396760145, | |
| "learning_rate": 7.548598528940354e-06, | |
| "loss": 0.5207, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4027926960257787, | |
| "grad_norm": 2.0389508957804727, | |
| "learning_rate": 7.4674245007528135e-06, | |
| "loss": 0.5355, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.40816326530612246, | |
| "grad_norm": 1.900432029568095, | |
| "learning_rate": 7.385381472189321e-06, | |
| "loss": 0.5264, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.41353383458646614, | |
| "grad_norm": 2.103053330699666, | |
| "learning_rate": 7.302498337922293e-06, | |
| "loss": 0.5138, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.4189044038668099, | |
| "grad_norm": 1.9990494257399714, | |
| "learning_rate": 7.218804288500343e-06, | |
| "loss": 0.5104, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.4242749731471536, | |
| "grad_norm": 2.008836602068826, | |
| "learning_rate": 7.134328800067684e-06, | |
| "loss": 0.5072, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4296455424274973, | |
| "grad_norm": 1.898533967889375, | |
| "learning_rate": 7.049101623982938e-06, | |
| "loss": 0.5037, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.43501611170784105, | |
| "grad_norm": 1.9911662902437104, | |
| "learning_rate": 6.963152776341044e-06, | |
| "loss": 0.504, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.44038668098818473, | |
| "grad_norm": 2.0295509745248506, | |
| "learning_rate": 6.876512527401897e-06, | |
| "loss": 0.5049, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.44575725026852847, | |
| "grad_norm": 2.0181826919469956, | |
| "learning_rate": 6.789211390929497e-06, | |
| "loss": 0.4963, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.45112781954887216, | |
| "grad_norm": 1.892141276395263, | |
| "learning_rate": 6.701280113445324e-06, | |
| "loss": 0.5002, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4564983888292159, | |
| "grad_norm": 1.9070052783863771, | |
| "learning_rate": 6.6127496633997475e-06, | |
| "loss": 0.485, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.46186895810955964, | |
| "grad_norm": 2.1387072725193454, | |
| "learning_rate": 6.523651220265269e-06, | |
| "loss": 0.4967, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4672395273899033, | |
| "grad_norm": 2.085588228406015, | |
| "learning_rate": 6.434016163555452e-06, | |
| "loss": 0.501, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.47261009667024706, | |
| "grad_norm": 1.9089216655576455, | |
| "learning_rate": 6.343876061773385e-06, | |
| "loss": 0.4937, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.47798066595059074, | |
| "grad_norm": 1.9312633313300902, | |
| "learning_rate": 6.2532626612936035e-06, | |
| "loss": 0.4839, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.4833512352309345, | |
| "grad_norm": 1.9571336743336616, | |
| "learning_rate": 6.162207875181354e-06, | |
| "loss": 0.476, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.48872180451127817, | |
| "grad_norm": 1.9801969405557864, | |
| "learning_rate": 6.070743771953157e-06, | |
| "loss": 0.4764, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.4940923737916219, | |
| "grad_norm": 1.9655943801740696, | |
| "learning_rate": 5.978902564282616e-06, | |
| "loss": 0.4765, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.49946294307196565, | |
| "grad_norm": 1.9737596326633116, | |
| "learning_rate": 5.886716597655472e-06, | |
| "loss": 0.4735, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.5048335123523093, | |
| "grad_norm": 1.95318643736562, | |
| "learning_rate": 5.7942183389778536e-06, | |
| "loss": 0.4679, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5102040816326531, | |
| "grad_norm": 1.9041916424192031, | |
| "learning_rate": 5.701440365141799e-06, | |
| "loss": 0.4766, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5155746509129968, | |
| "grad_norm": 1.9562750863298486, | |
| "learning_rate": 5.608415351552014e-06, | |
| "loss": 0.4592, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5209452201933404, | |
| "grad_norm": 1.9361504397416536, | |
| "learning_rate": 5.515176060617945e-06, | |
| "loss": 0.4781, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 1.899883908382976, | |
| "learning_rate": 5.421755330215223e-06, | |
| "loss": 0.4656, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5316863587540279, | |
| "grad_norm": 1.9431192925662886, | |
| "learning_rate": 5.328186062120509e-06, | |
| "loss": 0.4633, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5370569280343717, | |
| "grad_norm": 2.0449974415384164, | |
| "learning_rate": 5.23450121042383e-06, | |
| "loss": 0.4683, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5424274973147154, | |
| "grad_norm": 2.0390170181760237, | |
| "learning_rate": 5.140733769922525e-06, | |
| "loss": 0.4476, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.547798066595059, | |
| "grad_norm": 1.9021329921771841, | |
| "learning_rate": 5.0469167645008245e-06, | |
| "loss": 0.4453, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5531686358754028, | |
| "grad_norm": 1.9763527996853345, | |
| "learning_rate": 4.953083235499177e-06, | |
| "loss": 0.4514, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5585392051557465, | |
| "grad_norm": 2.0561752573847323, | |
| "learning_rate": 4.859266230077474e-06, | |
| "loss": 0.4297, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5639097744360902, | |
| "grad_norm": 1.8434878503038103, | |
| "learning_rate": 4.7654987895761705e-06, | |
| "loss": 0.4275, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.569280343716434, | |
| "grad_norm": 1.83689219445279, | |
| "learning_rate": 4.671813937879494e-06, | |
| "loss": 0.4503, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5746509129967776, | |
| "grad_norm": 1.9537201020249528, | |
| "learning_rate": 4.5782446697847775e-06, | |
| "loss": 0.4392, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5800214822771214, | |
| "grad_norm": 1.8533716853221458, | |
| "learning_rate": 4.484823939382056e-06, | |
| "loss": 0.4334, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5853920515574651, | |
| "grad_norm": 1.961161386672394, | |
| "learning_rate": 4.391584648447989e-06, | |
| "loss": 0.4421, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5907626208378088, | |
| "grad_norm": 2.08054114617343, | |
| "learning_rate": 4.298559634858202e-06, | |
| "loss": 0.4283, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5961331901181526, | |
| "grad_norm": 1.8677160254958194, | |
| "learning_rate": 4.205781661022146e-06, | |
| "loss": 0.4381, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.6015037593984962, | |
| "grad_norm": 1.920752668532558, | |
| "learning_rate": 4.1132834023445304e-06, | |
| "loss": 0.4341, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.60687432867884, | |
| "grad_norm": 1.9452189289759259, | |
| "learning_rate": 4.021097435717386e-06, | |
| "loss": 0.415, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.6122448979591837, | |
| "grad_norm": 1.9132748664941985, | |
| "learning_rate": 3.929256228046845e-06, | |
| "loss": 0.4336, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6176154672395274, | |
| "grad_norm": 1.993290738709928, | |
| "learning_rate": 3.837792124818647e-06, | |
| "loss": 0.422, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6229860365198711, | |
| "grad_norm": 1.9913961970653589, | |
| "learning_rate": 3.7467373387063973e-06, | |
| "loss": 0.4178, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6283566058002148, | |
| "grad_norm": 1.8427119621080716, | |
| "learning_rate": 3.656123938226618e-06, | |
| "loss": 0.428, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6337271750805585, | |
| "grad_norm": 1.968721400472942, | |
| "learning_rate": 3.5659838364445505e-06, | |
| "loss": 0.4172, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6390977443609023, | |
| "grad_norm": 1.96678689056989, | |
| "learning_rate": 3.476348779734732e-06, | |
| "loss": 0.4167, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.644468313641246, | |
| "grad_norm": 2.0428744345313317, | |
| "learning_rate": 3.387250336600254e-06, | |
| "loss": 0.4217, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6498388829215896, | |
| "grad_norm": 1.911511183929403, | |
| "learning_rate": 3.298719886554677e-06, | |
| "loss": 0.4148, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6552094522019334, | |
| "grad_norm": 1.8297131756573501, | |
| "learning_rate": 3.2107886090705035e-06, | |
| "loss": 0.4113, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6605800214822771, | |
| "grad_norm": 1.8374139904627782, | |
| "learning_rate": 3.1234874725981045e-06, | |
| "loss": 0.4117, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6659505907626209, | |
| "grad_norm": 1.9133897000669955, | |
| "learning_rate": 3.036847223658958e-06, | |
| "loss": 0.4045, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6713211600429646, | |
| "grad_norm": 1.8730189334453256, | |
| "learning_rate": 2.950898376017064e-06, | |
| "loss": 0.4095, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6766917293233082, | |
| "grad_norm": 1.8500933581783825, | |
| "learning_rate": 2.865671199932318e-06, | |
| "loss": 0.4012, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.682062298603652, | |
| "grad_norm": 1.7967563876892727, | |
| "learning_rate": 2.781195711499658e-06, | |
| "loss": 0.3941, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6874328678839957, | |
| "grad_norm": 1.8747314764617706, | |
| "learning_rate": 2.697501662077707e-06, | |
| "loss": 0.393, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6928034371643395, | |
| "grad_norm": 1.7802207847491662, | |
| "learning_rate": 2.6146185278106807e-06, | |
| "loss": 0.4017, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6981740064446831, | |
| "grad_norm": 2.02161990229544, | |
| "learning_rate": 2.5325754992471886e-06, | |
| "loss": 0.3995, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7035445757250268, | |
| "grad_norm": 1.8593472428410407, | |
| "learning_rate": 2.4514014710596467e-06, | |
| "loss": 0.3901, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.7089151450053706, | |
| "grad_norm": 1.8511721146012157, | |
| "learning_rate": 2.3711250318678906e-06, | |
| "loss": 0.3968, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.8665899913322235, | |
| "learning_rate": 2.2917744541705544e-06, | |
| "loss": 0.3908, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.719656283566058, | |
| "grad_norm": 1.8754563393501016, | |
| "learning_rate": 2.2133776843878185e-06, | |
| "loss": 0.3889, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7250268528464017, | |
| "grad_norm": 1.973295379140291, | |
| "learning_rate": 2.1359623330189655e-06, | |
| "loss": 0.3842, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7303974221267454, | |
| "grad_norm": 1.8678018351775936, | |
| "learning_rate": 2.059555664918268e-06, | |
| "loss": 0.3865, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7357679914070892, | |
| "grad_norm": 1.83999361738213, | |
| "learning_rate": 1.9841845896926022e-06, | |
| "loss": 0.3868, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7411385606874329, | |
| "grad_norm": 1.9587826124368826, | |
| "learning_rate": 1.9098756522241634e-06, | |
| "loss": 0.3868, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7465091299677766, | |
| "grad_norm": 1.7998085254745122, | |
| "learning_rate": 1.8366550233216584e-06, | |
| "loss": 0.3817, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7518796992481203, | |
| "grad_norm": 1.8019229893567459, | |
| "learning_rate": 1.7645484905032129e-06, | |
| "loss": 0.3756, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.757250268528464, | |
| "grad_norm": 1.7736589226596267, | |
| "learning_rate": 1.6935814489142937e-06, | |
| "loss": 0.382, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7626208378088077, | |
| "grad_norm": 1.8736562250816207, | |
| "learning_rate": 1.6237788923838149e-06, | |
| "loss": 0.3742, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7679914070891515, | |
| "grad_norm": 1.8367933341446385, | |
| "learning_rate": 1.555165404621567e-06, | |
| "loss": 0.3764, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7733619763694952, | |
| "grad_norm": 1.742864313155285, | |
| "learning_rate": 1.487765150560116e-06, | |
| "loss": 0.3742, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7787325456498388, | |
| "grad_norm": 1.7722081095014115, | |
| "learning_rate": 1.4216018678441558e-06, | |
| "loss": 0.3678, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7841031149301826, | |
| "grad_norm": 1.9069716881049699, | |
| "learning_rate": 1.3566988584703817e-06, | |
| "loss": 0.362, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 1.9399530643340799, | |
| "learning_rate": 1.293078980580766e-06, | |
| "loss": 0.3665, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7948442534908701, | |
| "grad_norm": 1.8012127743281918, | |
| "learning_rate": 1.2307646404121692e-06, | |
| "loss": 0.3714, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.8002148227712137, | |
| "grad_norm": 1.7867785582011508, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.3708, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.8055853920515574, | |
| "grad_norm": 1.7589328804606057, | |
| "learning_rate": 1.1101398914744565e-06, | |
| "loss": 0.3643, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8109559613319012, | |
| "grad_norm": 1.8095464000864174, | |
| "learning_rate": 1.0518719654447896e-06, | |
| "loss": 0.3717, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 1.7251086607823238, | |
| "learning_rate": 9.949945276530782e-07, | |
| "loss": 0.3629, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.8216970998925887, | |
| "grad_norm": 1.8028411997450835, | |
| "learning_rate": 9.395276097212841e-07, | |
| "loss": 0.374, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.8270676691729323, | |
| "grad_norm": 1.7477569693149821, | |
| "learning_rate": 8.854907465014479e-07, | |
| "loss": 0.3624, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.832438238453276, | |
| "grad_norm": 1.7785432606954263, | |
| "learning_rate": 8.329029691957124e-07, | |
| "loss": 0.3606, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8378088077336198, | |
| "grad_norm": 1.7600532012822097, | |
| "learning_rate": 7.817827986537508e-07, | |
| "loss": 0.3518, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8431793770139635, | |
| "grad_norm": 1.8290298911632301, | |
| "learning_rate": 7.321482388499096e-07, | |
| "loss": 0.3558, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.8485499462943072, | |
| "grad_norm": 1.736051089190557, | |
| "learning_rate": 6.840167705424106e-07, | |
| "loss": 0.3443, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8539205155746509, | |
| "grad_norm": 1.7339705264072511, | |
| "learning_rate": 6.374053451168166e-07, | |
| "loss": 0.3552, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8592910848549946, | |
| "grad_norm": 1.7168831046635995, | |
| "learning_rate": 5.92330378615929e-07, | |
| "loss": 0.3597, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8646616541353384, | |
| "grad_norm": 1.7597885378945268, | |
| "learning_rate": 5.488077459582425e-07, | |
| "loss": 0.3477, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8700322234156821, | |
| "grad_norm": 1.7454773401665622, | |
| "learning_rate": 5.068527753469604e-07, | |
| "loss": 0.347, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8754027926960258, | |
| "grad_norm": 2.1001488618217494, | |
| "learning_rate": 4.664802428715753e-07, | |
| "loss": 0.3624, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.8807733619763695, | |
| "grad_norm": 1.8348931889008009, | |
| "learning_rate": 4.2770436730388166e-07, | |
| "loss": 0.3493, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8861439312567132, | |
| "grad_norm": 1.8482665559640643, | |
| "learning_rate": 3.9053880509028086e-07, | |
| "loss": 0.3575, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8915145005370569, | |
| "grad_norm": 1.7654284228336778, | |
| "learning_rate": 3.549966455421305e-07, | |
| "loss": 0.3521, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8968850698174007, | |
| "grad_norm": 1.8224060382887075, | |
| "learning_rate": 3.2109040622582186e-07, | |
| "loss": 0.3488, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.9022556390977443, | |
| "grad_norm": 1.8156671777429907, | |
| "learning_rate": 2.8883202855423676e-07, | |
| "loss": 0.3475, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.907626208378088, | |
| "grad_norm": 1.7401116306912476, | |
| "learning_rate": 2.582328735811029e-07, | |
| "loss": 0.3534, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.9129967776584318, | |
| "grad_norm": 1.812051193355519, | |
| "learning_rate": 2.2930371799975593e-07, | |
| "loss": 0.3468, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9183673469387755, | |
| "grad_norm": 1.7410480531490395, | |
| "learning_rate": 2.0205475034770606e-07, | |
| "loss": 0.3543, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.9237379162191193, | |
| "grad_norm": 1.9139735870352734, | |
| "learning_rate": 1.7649556741833995e-07, | |
| "loss": 0.3529, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9291084854994629, | |
| "grad_norm": 1.771391905028915, | |
| "learning_rate": 1.5263517088103862e-07, | |
| "loss": 0.3437, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.9344790547798066, | |
| "grad_norm": 1.7345022358606093, | |
| "learning_rate": 1.304819641108801e-07, | |
| "loss": 0.3503, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9398496240601504, | |
| "grad_norm": 1.7801094287349393, | |
| "learning_rate": 1.1004374922906846e-07, | |
| "loss": 0.3451, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9452201933404941, | |
| "grad_norm": 1.8071597427760844, | |
| "learning_rate": 9.132772435510362e-08, | |
| "loss": 0.3508, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9505907626208379, | |
| "grad_norm": 1.7309028056205316, | |
| "learning_rate": 7.434048107168523e-08, | |
| "loss": 0.3493, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.9559613319011815, | |
| "grad_norm": 1.8154045704256447, | |
| "learning_rate": 5.908800210322696e-08, | |
| "loss": 0.3574, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9613319011815252, | |
| "grad_norm": 1.8355027069478766, | |
| "learning_rate": 4.55756592088058e-08, | |
| "loss": 0.3525, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.966702470461869, | |
| "grad_norm": 1.8104910905814395, | |
| "learning_rate": 3.3808211290284886e-08, | |
| "loss": 0.3397, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9720730397422127, | |
| "grad_norm": 1.8129460361338274, | |
| "learning_rate": 2.378980271628195e-08, | |
| "loss": 0.3451, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9774436090225563, | |
| "grad_norm": 1.712268760399258, | |
| "learning_rate": 1.552396186256411e-08, | |
| "loss": 0.3483, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9828141783029001, | |
| "grad_norm": 1.6906868953312937, | |
| "learning_rate": 9.013599869394096e-09, | |
| "loss": 0.3499, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9881847475832438, | |
| "grad_norm": 1.8154004316415735, | |
| "learning_rate": 4.261009616257638e-09, | |
| "loss": 0.3496, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9935553168635876, | |
| "grad_norm": 1.7609273835040906, | |
| "learning_rate": 1.2678649143349485e-09, | |
| "loss": 0.3491, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.9989258861439313, | |
| "grad_norm": 1.7702401819369071, | |
| "learning_rate": 3.5219917003948003e-11, | |
| "loss": 0.3483, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.3050035536289215, | |
| "eval_runtime": 1.1835, | |
| "eval_samples_per_second": 1.69, | |
| "eval_steps_per_second": 0.845, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 931, | |
| "total_flos": 194827769610240.0, | |
| "train_loss": 0.5089763147490365, | |
| "train_runtime": 18320.1118, | |
| "train_samples_per_second": 1.626, | |
| "train_steps_per_second": 0.051 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 931, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 194827769610240.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |