| { | |
| "best_metric": 0.5130746364593506, | |
| "best_model_checkpoint": "models/khmer-llama32-qa/checkpoint-9000", | |
| "epoch": 3.0, | |
| "eval_steps": 1000, | |
| "global_step": 9591, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006255864873318737, | |
| "grad_norm": 3.055704355239868, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "loss": 3.599, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.012511729746637473, | |
| "grad_norm": 3.5084385871887207, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 3.6639, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01876759461995621, | |
| "grad_norm": 3.2163877487182617, | |
| "learning_rate": 1.25e-06, | |
| "loss": 3.6221, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.025023459493274947, | |
| "grad_norm": 2.483070135116577, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 3.5293, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03127932436659368, | |
| "grad_norm": 2.087958335876465, | |
| "learning_rate": 2.0833333333333334e-06, | |
| "loss": 3.486, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.03753518923991242, | |
| "grad_norm": 1.9473289251327515, | |
| "learning_rate": 2.5e-06, | |
| "loss": 3.3948, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.043791054113231156, | |
| "grad_norm": 1.4656407833099365, | |
| "learning_rate": 2.916666666666667e-06, | |
| "loss": 3.1533, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.05004691898654989, | |
| "grad_norm": 1.4900797605514526, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 2.9367, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.05630278385986863, | |
| "grad_norm": 1.2561100721359253, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 2.8441, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.06255864873318737, | |
| "grad_norm": 1.502004623413086, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 2.6692, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.0688145136065061, | |
| "grad_norm": 1.1987361907958984, | |
| "learning_rate": 4.583333333333333e-06, | |
| "loss": 2.618, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.07507037847982484, | |
| "grad_norm": 1.2664542198181152, | |
| "learning_rate": 5e-06, | |
| "loss": 2.5955, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.08132624335314358, | |
| "grad_norm": 1.4493029117584229, | |
| "learning_rate": 5.416666666666667e-06, | |
| "loss": 2.5506, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.08758210822646231, | |
| "grad_norm": 1.3738720417022705, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 2.5202, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.09383797309978105, | |
| "grad_norm": 1.5998780727386475, | |
| "learning_rate": 6.25e-06, | |
| "loss": 2.4909, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.10009383797309979, | |
| "grad_norm": 1.4477788209915161, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 2.4929, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.10634970284641852, | |
| "grad_norm": 1.4700181484222412, | |
| "learning_rate": 7.083333333333335e-06, | |
| "loss": 2.4574, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.11260556771973726, | |
| "grad_norm": 1.9266761541366577, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 2.4146, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.118861432593056, | |
| "grad_norm": 2.079688787460327, | |
| "learning_rate": 7.916666666666667e-06, | |
| "loss": 2.3837, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.12511729746637473, | |
| "grad_norm": 1.6938954591751099, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 2.3659, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.13137316233969346, | |
| "grad_norm": 1.6853476762771606, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 2.3636, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.1376290272130122, | |
| "grad_norm": 1.5365005731582642, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 2.3139, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.14388489208633093, | |
| "grad_norm": 1.6722062826156616, | |
| "learning_rate": 9.583333333333335e-06, | |
| "loss": 2.3738, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.15014075695964968, | |
| "grad_norm": 1.6911437511444092, | |
| "learning_rate": 1e-05, | |
| "loss": 2.246, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.1563966218329684, | |
| "grad_norm": 1.9284753799438477, | |
| "learning_rate": 1.0416666666666668e-05, | |
| "loss": 2.2457, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.16265248670628715, | |
| "grad_norm": 1.870841145515442, | |
| "learning_rate": 1.0833333333333334e-05, | |
| "loss": 2.1937, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.16890835157960588, | |
| "grad_norm": 2.1168696880340576, | |
| "learning_rate": 1.125e-05, | |
| "loss": 2.2104, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.17516421645292463, | |
| "grad_norm": 2.3863697052001953, | |
| "learning_rate": 1.1666666666666668e-05, | |
| "loss": 2.1708, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.18142008132624335, | |
| "grad_norm": 2.1332266330718994, | |
| "learning_rate": 1.2083333333333333e-05, | |
| "loss": 2.1147, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.1876759461995621, | |
| "grad_norm": 2.657927989959717, | |
| "learning_rate": 1.25e-05, | |
| "loss": 2.1468, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.19393181107288082, | |
| "grad_norm": 2.723944664001465, | |
| "learning_rate": 1.2916666666666668e-05, | |
| "loss": 2.1126, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.20018767594619957, | |
| "grad_norm": 2.2403364181518555, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 2.132, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.2064435408195183, | |
| "grad_norm": 2.1621572971343994, | |
| "learning_rate": 1.375e-05, | |
| "loss": 2.0652, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.21269940569283705, | |
| "grad_norm": 2.283856153488159, | |
| "learning_rate": 1.416666666666667e-05, | |
| "loss": 2.0652, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.21895527056615577, | |
| "grad_norm": 2.3808491230010986, | |
| "learning_rate": 1.4583333333333333e-05, | |
| "loss": 2.0647, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.22521113543947452, | |
| "grad_norm": 2.655351400375366, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 2.0804, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.23146700031279324, | |
| "grad_norm": 2.5698490142822266, | |
| "learning_rate": 1.5416666666666668e-05, | |
| "loss": 2.0677, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.237722865186112, | |
| "grad_norm": 3.3763697147369385, | |
| "learning_rate": 1.5833333333333333e-05, | |
| "loss": 2.0432, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.24397873005943072, | |
| "grad_norm": 2.749903917312622, | |
| "learning_rate": 1.6250000000000002e-05, | |
| "loss": 2.0451, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.25023459493274947, | |
| "grad_norm": 3.3386733531951904, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 2.0461, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.2564904598060682, | |
| "grad_norm": 2.211704730987549, | |
| "learning_rate": 1.7083333333333333e-05, | |
| "loss": 2.0056, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.2627463246793869, | |
| "grad_norm": 2.886216640472412, | |
| "learning_rate": 1.7500000000000002e-05, | |
| "loss": 1.9902, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.26900218955270566, | |
| "grad_norm": 2.8164501190185547, | |
| "learning_rate": 1.7916666666666667e-05, | |
| "loss": 2.0382, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.2752580544260244, | |
| "grad_norm": 3.3984410762786865, | |
| "learning_rate": 1.8333333333333333e-05, | |
| "loss": 2.0245, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.2815139192993431, | |
| "grad_norm": 2.9318149089813232, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 1.981, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.28776978417266186, | |
| "grad_norm": 2.9568192958831787, | |
| "learning_rate": 1.916666666666667e-05, | |
| "loss": 1.981, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.2940256490459806, | |
| "grad_norm": 3.3105645179748535, | |
| "learning_rate": 1.9583333333333333e-05, | |
| "loss": 1.9575, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.30028151391929936, | |
| "grad_norm": 3.094958782196045, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0318, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.30653737879261805, | |
| "grad_norm": 2.9924981594085693, | |
| "learning_rate": 1.9953655428107984e-05, | |
| "loss": 1.9318, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.3127932436659368, | |
| "grad_norm": 3.0640549659729004, | |
| "learning_rate": 1.9907310856215966e-05, | |
| "loss": 1.9675, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3127932436659368, | |
| "eval_loss": 0.6163395643234253, | |
| "eval_runtime": 30.1355, | |
| "eval_samples_per_second": 16.061, | |
| "eval_steps_per_second": 4.015, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.31904910853925555, | |
| "grad_norm": 3.016735076904297, | |
| "learning_rate": 1.986096628432395e-05, | |
| "loss": 1.9317, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.3253049734125743, | |
| "grad_norm": 3.159961462020874, | |
| "learning_rate": 1.9814621712431933e-05, | |
| "loss": 1.9003, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.331560838285893, | |
| "grad_norm": 3.248335123062134, | |
| "learning_rate": 1.9768277140539916e-05, | |
| "loss": 1.9733, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.33781670315921175, | |
| "grad_norm": 3.2077865600585938, | |
| "learning_rate": 1.9721932568647898e-05, | |
| "loss": 1.8866, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.3440725680325305, | |
| "grad_norm": 3.3022334575653076, | |
| "learning_rate": 1.967558799675588e-05, | |
| "loss": 1.9616, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.35032843290584925, | |
| "grad_norm": 3.258803606033325, | |
| "learning_rate": 1.9629243424863865e-05, | |
| "loss": 1.9183, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.35658429777916795, | |
| "grad_norm": 3.348259687423706, | |
| "learning_rate": 1.9582898852971847e-05, | |
| "loss": 1.8972, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.3628401626524867, | |
| "grad_norm": 2.9948360919952393, | |
| "learning_rate": 1.953655428107983e-05, | |
| "loss": 1.8953, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.36909602752580545, | |
| "grad_norm": 3.072226047515869, | |
| "learning_rate": 1.949020970918781e-05, | |
| "loss": 1.909, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.3753518923991242, | |
| "grad_norm": 2.7837533950805664, | |
| "learning_rate": 1.9443865137295797e-05, | |
| "loss": 1.9396, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.3816077572724429, | |
| "grad_norm": 3.2200827598571777, | |
| "learning_rate": 1.939752056540378e-05, | |
| "loss": 1.8605, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.38786362214576164, | |
| "grad_norm": 2.9632668495178223, | |
| "learning_rate": 1.935117599351176e-05, | |
| "loss": 1.8823, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.3941194870190804, | |
| "grad_norm": 3.762718915939331, | |
| "learning_rate": 1.9304831421619747e-05, | |
| "loss": 1.861, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.40037535189239915, | |
| "grad_norm": 3.123317003250122, | |
| "learning_rate": 1.9258486849727725e-05, | |
| "loss": 1.8385, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.40663121676571784, | |
| "grad_norm": 3.2393641471862793, | |
| "learning_rate": 1.921214227783571e-05, | |
| "loss": 1.8507, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.4128870816390366, | |
| "grad_norm": 3.06451153755188, | |
| "learning_rate": 1.9165797705943693e-05, | |
| "loss": 1.8872, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.41914294651235534, | |
| "grad_norm": 3.2108166217803955, | |
| "learning_rate": 1.9119453134051675e-05, | |
| "loss": 1.9376, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.4253988113856741, | |
| "grad_norm": 3.1738028526306152, | |
| "learning_rate": 1.907310856215966e-05, | |
| "loss": 1.8453, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.4316546762589928, | |
| "grad_norm": 3.297335147857666, | |
| "learning_rate": 1.902676399026764e-05, | |
| "loss": 1.8411, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.43791054113231154, | |
| "grad_norm": 3.6973254680633545, | |
| "learning_rate": 1.8980419418375625e-05, | |
| "loss": 1.8655, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.4441664060056303, | |
| "grad_norm": 3.067793846130371, | |
| "learning_rate": 1.8934074846483607e-05, | |
| "loss": 1.8269, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.45042227087894904, | |
| "grad_norm": 3.414215564727783, | |
| "learning_rate": 1.888773027459159e-05, | |
| "loss": 1.8114, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.45667813575226773, | |
| "grad_norm": 3.223034620285034, | |
| "learning_rate": 1.8841385702699574e-05, | |
| "loss": 1.8345, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.4629340006255865, | |
| "grad_norm": 3.184588670730591, | |
| "learning_rate": 1.8795041130807557e-05, | |
| "loss": 1.8334, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.46918986549890523, | |
| "grad_norm": 3.343078374862671, | |
| "learning_rate": 1.874869655891554e-05, | |
| "loss": 1.7791, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.475445730372224, | |
| "grad_norm": 3.3436670303344727, | |
| "learning_rate": 1.870235198702352e-05, | |
| "loss": 1.8317, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.4817015952455427, | |
| "grad_norm": 3.7679710388183594, | |
| "learning_rate": 1.8656007415131506e-05, | |
| "loss": 1.8179, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.48795746011886143, | |
| "grad_norm": 3.780228614807129, | |
| "learning_rate": 1.8609662843239485e-05, | |
| "loss": 1.8051, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.4942133249921802, | |
| "grad_norm": 3.061527729034424, | |
| "learning_rate": 1.856331827134747e-05, | |
| "loss": 1.8413, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.5004691898654989, | |
| "grad_norm": 3.53220796585083, | |
| "learning_rate": 1.8516973699455452e-05, | |
| "loss": 1.8018, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.5067250547388177, | |
| "grad_norm": 3.7156918048858643, | |
| "learning_rate": 1.8470629127563435e-05, | |
| "loss": 1.771, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.5129809196121364, | |
| "grad_norm": 3.174913167953491, | |
| "learning_rate": 1.842428455567142e-05, | |
| "loss": 1.792, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.5192367844854551, | |
| "grad_norm": 3.3128390312194824, | |
| "learning_rate": 1.83779399837794e-05, | |
| "loss": 1.7571, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.5254926493587738, | |
| "grad_norm": 3.163397789001465, | |
| "learning_rate": 1.8331595411887384e-05, | |
| "loss": 1.8294, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.5317485142320926, | |
| "grad_norm": 3.1041367053985596, | |
| "learning_rate": 1.8285250839995366e-05, | |
| "loss": 1.7722, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.5380043791054113, | |
| "grad_norm": 3.4675991535186768, | |
| "learning_rate": 1.823890626810335e-05, | |
| "loss": 1.7617, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.5442602439787301, | |
| "grad_norm": 3.572848081588745, | |
| "learning_rate": 1.8192561696211334e-05, | |
| "loss": 1.7121, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.5505161088520488, | |
| "grad_norm": 3.1788299083709717, | |
| "learning_rate": 1.8146217124319316e-05, | |
| "loss": 1.7865, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.5567719737253676, | |
| "grad_norm": 3.3534226417541504, | |
| "learning_rate": 1.8099872552427298e-05, | |
| "loss": 1.8233, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.5630278385986862, | |
| "grad_norm": 3.2126591205596924, | |
| "learning_rate": 1.805352798053528e-05, | |
| "loss": 1.7321, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.569283703472005, | |
| "grad_norm": 3.1480772495269775, | |
| "learning_rate": 1.8007183408643266e-05, | |
| "loss": 1.8045, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.5755395683453237, | |
| "grad_norm": 3.242344617843628, | |
| "learning_rate": 1.7960838836751248e-05, | |
| "loss": 1.7182, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.5817954332186425, | |
| "grad_norm": 3.085733413696289, | |
| "learning_rate": 1.791449426485923e-05, | |
| "loss": 1.8152, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.5880512980919612, | |
| "grad_norm": 3.510521173477173, | |
| "learning_rate": 1.7868149692967212e-05, | |
| "loss": 1.7413, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.59430716296528, | |
| "grad_norm": 3.4337544441223145, | |
| "learning_rate": 1.7821805121075194e-05, | |
| "loss": 1.8148, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.6005630278385987, | |
| "grad_norm": 2.888315200805664, | |
| "learning_rate": 1.777546054918318e-05, | |
| "loss": 1.7372, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.6068188927119175, | |
| "grad_norm": 3.148432970046997, | |
| "learning_rate": 1.7729115977291162e-05, | |
| "loss": 1.8036, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.6130747575852361, | |
| "grad_norm": 3.4758594036102295, | |
| "learning_rate": 1.7682771405399144e-05, | |
| "loss": 1.7447, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.6193306224585549, | |
| "grad_norm": 3.720221519470215, | |
| "learning_rate": 1.763642683350713e-05, | |
| "loss": 1.6979, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.6255864873318736, | |
| "grad_norm": 3.6428353786468506, | |
| "learning_rate": 1.7590082261615108e-05, | |
| "loss": 1.7339, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.6255864873318736, | |
| "eval_loss": 0.5785377621650696, | |
| "eval_runtime": 30.1197, | |
| "eval_samples_per_second": 16.069, | |
| "eval_steps_per_second": 4.017, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.6318423522051924, | |
| "grad_norm": 3.3335535526275635, | |
| "learning_rate": 1.7543737689723093e-05, | |
| "loss": 1.7016, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.6380982170785111, | |
| "grad_norm": 2.9847285747528076, | |
| "learning_rate": 1.7497393117831076e-05, | |
| "loss": 1.6967, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.6443540819518299, | |
| "grad_norm": 3.235311269760132, | |
| "learning_rate": 1.7451048545939058e-05, | |
| "loss": 1.733, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.6506099468251486, | |
| "grad_norm": 3.2147135734558105, | |
| "learning_rate": 1.740470397404704e-05, | |
| "loss": 1.7548, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.6568658116984674, | |
| "grad_norm": 3.2119052410125732, | |
| "learning_rate": 1.7358359402155025e-05, | |
| "loss": 1.6748, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.663121676571786, | |
| "grad_norm": 3.278869390487671, | |
| "learning_rate": 1.7312014830263007e-05, | |
| "loss": 1.7599, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.6693775414451048, | |
| "grad_norm": 3.781015396118164, | |
| "learning_rate": 1.726567025837099e-05, | |
| "loss": 1.7298, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.6756334063184235, | |
| "grad_norm": 3.6497671604156494, | |
| "learning_rate": 1.721932568647897e-05, | |
| "loss": 1.7324, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.6818892711917423, | |
| "grad_norm": 3.0687382221221924, | |
| "learning_rate": 1.7172981114586954e-05, | |
| "loss": 1.6643, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.688145136065061, | |
| "grad_norm": 3.9177842140197754, | |
| "learning_rate": 1.712663654269494e-05, | |
| "loss": 1.6868, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.6944010009383798, | |
| "grad_norm": 3.525784969329834, | |
| "learning_rate": 1.708029197080292e-05, | |
| "loss": 1.7393, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.7006568658116985, | |
| "grad_norm": 3.3422930240631104, | |
| "learning_rate": 1.7033947398910903e-05, | |
| "loss": 1.6745, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.7069127306850173, | |
| "grad_norm": 3.4408421516418457, | |
| "learning_rate": 1.698760282701889e-05, | |
| "loss": 1.7166, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.7131685955583359, | |
| "grad_norm": 3.9245855808258057, | |
| "learning_rate": 1.6941258255126868e-05, | |
| "loss": 1.7218, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.7194244604316546, | |
| "grad_norm": 3.502345561981201, | |
| "learning_rate": 1.6894913683234853e-05, | |
| "loss": 1.7191, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.7256803253049734, | |
| "grad_norm": 3.451737642288208, | |
| "learning_rate": 1.6848569111342835e-05, | |
| "loss": 1.726, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.7319361901782921, | |
| "grad_norm": 3.112499952316284, | |
| "learning_rate": 1.6802224539450817e-05, | |
| "loss": 1.6989, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.7381920550516109, | |
| "grad_norm": 4.081657409667969, | |
| "learning_rate": 1.6755879967558803e-05, | |
| "loss": 1.7184, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.7444479199249296, | |
| "grad_norm": 3.0992517471313477, | |
| "learning_rate": 1.6709535395666785e-05, | |
| "loss": 1.6656, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.7507037847982484, | |
| "grad_norm": 3.080566644668579, | |
| "learning_rate": 1.6663190823774767e-05, | |
| "loss": 1.6727, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.7569596496715671, | |
| "grad_norm": 2.938995361328125, | |
| "learning_rate": 1.661684625188275e-05, | |
| "loss": 1.6483, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.7632155145448858, | |
| "grad_norm": 3.2587146759033203, | |
| "learning_rate": 1.657050167999073e-05, | |
| "loss": 1.6965, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.7694713794182045, | |
| "grad_norm": 3.5692856311798096, | |
| "learning_rate": 1.6524157108098717e-05, | |
| "loss": 1.7098, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.7757272442915233, | |
| "grad_norm": 3.1614418029785156, | |
| "learning_rate": 1.64778125362067e-05, | |
| "loss": 1.6576, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.781983109164842, | |
| "grad_norm": 3.1559157371520996, | |
| "learning_rate": 1.643146796431468e-05, | |
| "loss": 1.6498, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.7882389740381608, | |
| "grad_norm": 3.6744728088378906, | |
| "learning_rate": 1.6385123392422663e-05, | |
| "loss": 1.6789, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.7944948389114795, | |
| "grad_norm": 3.6650450229644775, | |
| "learning_rate": 1.633877882053065e-05, | |
| "loss": 1.6372, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.8007507037847983, | |
| "grad_norm": 3.7447121143341064, | |
| "learning_rate": 1.629243424863863e-05, | |
| "loss": 1.7067, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.8070065686581169, | |
| "grad_norm": 3.40779972076416, | |
| "learning_rate": 1.6246089676746613e-05, | |
| "loss": 1.6942, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.8132624335314357, | |
| "grad_norm": 4.096377849578857, | |
| "learning_rate": 1.6199745104854595e-05, | |
| "loss": 1.6124, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.8195182984047544, | |
| "grad_norm": 3.5456223487854004, | |
| "learning_rate": 1.6153400532962577e-05, | |
| "loss": 1.6957, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.8257741632780732, | |
| "grad_norm": 4.157273292541504, | |
| "learning_rate": 1.6107055961070562e-05, | |
| "loss": 1.6718, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.8320300281513919, | |
| "grad_norm": 3.7936980724334717, | |
| "learning_rate": 1.6060711389178544e-05, | |
| "loss": 1.6235, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.8382858930247107, | |
| "grad_norm": 3.590808629989624, | |
| "learning_rate": 1.6014366817286526e-05, | |
| "loss": 1.6704, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.8445417578980294, | |
| "grad_norm": 3.088557243347168, | |
| "learning_rate": 1.596802224539451e-05, | |
| "loss": 1.6886, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.8507976227713482, | |
| "grad_norm": 3.6176650524139404, | |
| "learning_rate": 1.592167767350249e-05, | |
| "loss": 1.6178, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.8570534876446668, | |
| "grad_norm": 3.519657611846924, | |
| "learning_rate": 1.5875333101610476e-05, | |
| "loss": 1.6596, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.8633093525179856, | |
| "grad_norm": 3.0401337146759033, | |
| "learning_rate": 1.5828988529718458e-05, | |
| "loss": 1.6518, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 3.584986925125122, | |
| "learning_rate": 1.578264395782644e-05, | |
| "loss": 1.6248, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.8758210822646231, | |
| "grad_norm": 3.4703197479248047, | |
| "learning_rate": 1.5736299385934422e-05, | |
| "loss": 1.6753, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.8820769471379418, | |
| "grad_norm": 3.361783266067505, | |
| "learning_rate": 1.5689954814042408e-05, | |
| "loss": 1.6769, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.8883328120112606, | |
| "grad_norm": 3.613750696182251, | |
| "learning_rate": 1.564361024215039e-05, | |
| "loss": 1.7109, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.8945886768845793, | |
| "grad_norm": 3.510986566543579, | |
| "learning_rate": 1.5597265670258372e-05, | |
| "loss": 1.6455, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.9008445417578981, | |
| "grad_norm": 3.334446668624878, | |
| "learning_rate": 1.5550921098366358e-05, | |
| "loss": 1.6255, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.9071004066312167, | |
| "grad_norm": 3.8680272102355957, | |
| "learning_rate": 1.5504576526474336e-05, | |
| "loss": 1.6302, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.9133562715045355, | |
| "grad_norm": 3.6026196479797363, | |
| "learning_rate": 1.5458231954582322e-05, | |
| "loss": 1.6466, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.9196121363778542, | |
| "grad_norm": 3.609692096710205, | |
| "learning_rate": 1.5411887382690304e-05, | |
| "loss": 1.633, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.925868001251173, | |
| "grad_norm": 3.4523870944976807, | |
| "learning_rate": 1.5365542810798286e-05, | |
| "loss": 1.6967, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.9321238661244917, | |
| "grad_norm": 3.257688283920288, | |
| "learning_rate": 1.531919823890627e-05, | |
| "loss": 1.6019, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.9383797309978105, | |
| "grad_norm": 3.10917067527771, | |
| "learning_rate": 1.527285366701425e-05, | |
| "loss": 1.6436, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.9383797309978105, | |
| "eval_loss": 0.5556857585906982, | |
| "eval_runtime": 30.1131, | |
| "eval_samples_per_second": 16.073, | |
| "eval_steps_per_second": 4.018, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.9446355958711292, | |
| "grad_norm": 3.7206368446350098, | |
| "learning_rate": 1.5226509095122236e-05, | |
| "loss": 1.6323, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.950891460744448, | |
| "grad_norm": 3.273854970932007, | |
| "learning_rate": 1.518016452323022e-05, | |
| "loss": 1.6645, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.9571473256177666, | |
| "grad_norm": 3.4478001594543457, | |
| "learning_rate": 1.51338199513382e-05, | |
| "loss": 1.6323, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.9634031904910854, | |
| "grad_norm": 3.5107407569885254, | |
| "learning_rate": 1.5087475379446184e-05, | |
| "loss": 1.6106, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.9696590553644041, | |
| "grad_norm": 3.4260244369506836, | |
| "learning_rate": 1.5041130807554167e-05, | |
| "loss": 1.6129, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.9759149202377229, | |
| "grad_norm": 3.3112871646881104, | |
| "learning_rate": 1.499478623566215e-05, | |
| "loss": 1.6468, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.9821707851110416, | |
| "grad_norm": 3.370237112045288, | |
| "learning_rate": 1.4948441663770132e-05, | |
| "loss": 1.6002, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.9884266499843604, | |
| "grad_norm": 2.750504732131958, | |
| "learning_rate": 1.4902097091878115e-05, | |
| "loss": 1.6065, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.9946825148576791, | |
| "grad_norm": 4.291706085205078, | |
| "learning_rate": 1.4855752519986097e-05, | |
| "loss": 1.5884, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 1.0009383797309979, | |
| "grad_norm": 3.447493553161621, | |
| "learning_rate": 1.4809407948094081e-05, | |
| "loss": 1.6045, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.0071942446043165, | |
| "grad_norm": 3.561293125152588, | |
| "learning_rate": 1.4763063376202062e-05, | |
| "loss": 1.601, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 1.0134501094776354, | |
| "grad_norm": 3.4609010219573975, | |
| "learning_rate": 1.4716718804310045e-05, | |
| "loss": 1.5634, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 1.019705974350954, | |
| "grad_norm": 3.336390733718872, | |
| "learning_rate": 1.467037423241803e-05, | |
| "loss": 1.6336, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 1.0259618392242729, | |
| "grad_norm": 3.7268803119659424, | |
| "learning_rate": 1.4624029660526011e-05, | |
| "loss": 1.5949, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 1.0322177040975915, | |
| "grad_norm": 3.0219619274139404, | |
| "learning_rate": 1.4577685088633995e-05, | |
| "loss": 1.5786, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.0384735689709101, | |
| "grad_norm": 3.6862988471984863, | |
| "learning_rate": 1.4531340516741979e-05, | |
| "loss": 1.5208, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 1.044729433844229, | |
| "grad_norm": 3.8871073722839355, | |
| "learning_rate": 1.448499594484996e-05, | |
| "loss": 1.594, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 1.0509852987175476, | |
| "grad_norm": 3.6962697505950928, | |
| "learning_rate": 1.4438651372957943e-05, | |
| "loss": 1.5132, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 1.0572411635908665, | |
| "grad_norm": 3.7267775535583496, | |
| "learning_rate": 1.4392306801065927e-05, | |
| "loss": 1.5618, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 1.0634970284641851, | |
| "grad_norm": 3.993413209915161, | |
| "learning_rate": 1.4345962229173909e-05, | |
| "loss": 1.5639, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.069752893337504, | |
| "grad_norm": 3.4021973609924316, | |
| "learning_rate": 1.4299617657281893e-05, | |
| "loss": 1.591, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 1.0760087582108226, | |
| "grad_norm": 3.457179307937622, | |
| "learning_rate": 1.4253273085389877e-05, | |
| "loss": 1.5748, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 1.0822646230841413, | |
| "grad_norm": 3.792189836502075, | |
| "learning_rate": 1.4206928513497857e-05, | |
| "loss": 1.5604, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 1.0885204879574601, | |
| "grad_norm": 3.5839691162109375, | |
| "learning_rate": 1.416058394160584e-05, | |
| "loss": 1.5709, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 1.0947763528307788, | |
| "grad_norm": 3.1503920555114746, | |
| "learning_rate": 1.4114239369713823e-05, | |
| "loss": 1.6201, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.1010322177040976, | |
| "grad_norm": 3.971402645111084, | |
| "learning_rate": 1.4067894797821807e-05, | |
| "loss": 1.6339, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 1.1072880825774163, | |
| "grad_norm": 2.93398118019104, | |
| "learning_rate": 1.402155022592979e-05, | |
| "loss": 1.6061, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 1.1135439474507351, | |
| "grad_norm": 3.298301935195923, | |
| "learning_rate": 1.3975205654037771e-05, | |
| "loss": 1.5666, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 1.1197998123240538, | |
| "grad_norm": 3.154184341430664, | |
| "learning_rate": 1.3928861082145755e-05, | |
| "loss": 1.5712, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 1.1260556771973724, | |
| "grad_norm": 4.005284309387207, | |
| "learning_rate": 1.3882516510253738e-05, | |
| "loss": 1.5215, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.1323115420706913, | |
| "grad_norm": 4.161669731140137, | |
| "learning_rate": 1.383617193836172e-05, | |
| "loss": 1.5316, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 1.13856740694401, | |
| "grad_norm": 3.5575239658355713, | |
| "learning_rate": 1.3789827366469704e-05, | |
| "loss": 1.5181, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 1.1448232718173288, | |
| "grad_norm": 3.849310874938965, | |
| "learning_rate": 1.3743482794577686e-05, | |
| "loss": 1.553, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 1.1510791366906474, | |
| "grad_norm": 4.078603744506836, | |
| "learning_rate": 1.3697138222685669e-05, | |
| "loss": 1.6065, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 1.1573350015639663, | |
| "grad_norm": 3.6123812198638916, | |
| "learning_rate": 1.3650793650793652e-05, | |
| "loss": 1.6441, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.163590866437285, | |
| "grad_norm": 4.202884674072266, | |
| "learning_rate": 1.3604449078901634e-05, | |
| "loss": 1.5529, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 1.1698467313106038, | |
| "grad_norm": 3.2778208255767822, | |
| "learning_rate": 1.3558104507009617e-05, | |
| "loss": 1.5742, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 1.1761025961839224, | |
| "grad_norm": 4.048972129821777, | |
| "learning_rate": 1.35117599351176e-05, | |
| "loss": 1.5131, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 1.182358461057241, | |
| "grad_norm": 3.4141221046447754, | |
| "learning_rate": 1.3465415363225582e-05, | |
| "loss": 1.5249, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 1.18861432593056, | |
| "grad_norm": 4.267421722412109, | |
| "learning_rate": 1.3419070791333566e-05, | |
| "loss": 1.5395, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.1948701908038786, | |
| "grad_norm": 3.1491217613220215, | |
| "learning_rate": 1.337272621944155e-05, | |
| "loss": 1.5512, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 1.2011260556771974, | |
| "grad_norm": 3.626814365386963, | |
| "learning_rate": 1.332638164754953e-05, | |
| "loss": 1.5097, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 1.207381920550516, | |
| "grad_norm": 3.499701738357544, | |
| "learning_rate": 1.3280037075657514e-05, | |
| "loss": 1.5498, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 1.213637785423835, | |
| "grad_norm": 3.646066188812256, | |
| "learning_rate": 1.3233692503765498e-05, | |
| "loss": 1.5804, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 1.2198936502971536, | |
| "grad_norm": 3.6452980041503906, | |
| "learning_rate": 1.318734793187348e-05, | |
| "loss": 1.6128, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.2261495151704724, | |
| "grad_norm": 3.4716873168945312, | |
| "learning_rate": 1.3141003359981464e-05, | |
| "loss": 1.5725, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 1.232405380043791, | |
| "grad_norm": 3.3494322299957275, | |
| "learning_rate": 1.3094658788089448e-05, | |
| "loss": 1.539, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 1.2386612449171097, | |
| "grad_norm": 3.9730496406555176, | |
| "learning_rate": 1.3048314216197428e-05, | |
| "loss": 1.5518, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 1.2449171097904286, | |
| "grad_norm": 3.71327805519104, | |
| "learning_rate": 1.3001969644305412e-05, | |
| "loss": 1.5592, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 1.2511729746637472, | |
| "grad_norm": 3.4022958278656006, | |
| "learning_rate": 1.2955625072413394e-05, | |
| "loss": 1.5271, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.2511729746637472, | |
| "eval_loss": 0.5431435704231262, | |
| "eval_runtime": 30.1654, | |
| "eval_samples_per_second": 16.045, | |
| "eval_steps_per_second": 4.011, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.257428839537066, | |
| "grad_norm": 4.227381229400635, | |
| "learning_rate": 1.2909280500521378e-05, | |
| "loss": 1.4989, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 1.2636847044103847, | |
| "grad_norm": 3.459625720977783, | |
| "learning_rate": 1.2862935928629362e-05, | |
| "loss": 1.5381, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 1.2699405692837034, | |
| "grad_norm": 3.8662827014923096, | |
| "learning_rate": 1.2816591356737342e-05, | |
| "loss": 1.5592, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 1.2761964341570222, | |
| "grad_norm": 3.6900293827056885, | |
| "learning_rate": 1.2770246784845326e-05, | |
| "loss": 1.542, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 1.282452299030341, | |
| "grad_norm": 3.6139461994171143, | |
| "learning_rate": 1.272390221295331e-05, | |
| "loss": 1.5312, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.2887081639036597, | |
| "grad_norm": 4.2868332862854, | |
| "learning_rate": 1.2677557641061292e-05, | |
| "loss": 1.5769, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 1.2949640287769784, | |
| "grad_norm": 3.5073251724243164, | |
| "learning_rate": 1.2631213069169275e-05, | |
| "loss": 1.5122, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 1.3012198936502972, | |
| "grad_norm": 3.2796294689178467, | |
| "learning_rate": 1.2584868497277258e-05, | |
| "loss": 1.5677, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 1.3074757585236159, | |
| "grad_norm": 4.9055376052856445, | |
| "learning_rate": 1.253852392538524e-05, | |
| "loss": 1.5355, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 1.3137316233969347, | |
| "grad_norm": 3.3482460975646973, | |
| "learning_rate": 1.2492179353493223e-05, | |
| "loss": 1.5339, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.3199874882702534, | |
| "grad_norm": 3.3556487560272217, | |
| "learning_rate": 1.2445834781601207e-05, | |
| "loss": 1.5132, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 1.326243353143572, | |
| "grad_norm": 3.3929593563079834, | |
| "learning_rate": 1.2399490209709188e-05, | |
| "loss": 1.5163, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 1.3324992180168909, | |
| "grad_norm": 3.4471447467803955, | |
| "learning_rate": 1.2353145637817171e-05, | |
| "loss": 1.5829, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 1.3387550828902095, | |
| "grad_norm": 4.140485763549805, | |
| "learning_rate": 1.2306801065925154e-05, | |
| "loss": 1.4735, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 1.3450109477635284, | |
| "grad_norm": 3.6784679889678955, | |
| "learning_rate": 1.2260456494033137e-05, | |
| "loss": 1.5415, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.351266812636847, | |
| "grad_norm": 3.488356113433838, | |
| "learning_rate": 1.2214111922141121e-05, | |
| "loss": 1.5178, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 1.3575226775101656, | |
| "grad_norm": 3.4582056999206543, | |
| "learning_rate": 1.2167767350249102e-05, | |
| "loss": 1.5464, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 1.3637785423834845, | |
| "grad_norm": 3.233997344970703, | |
| "learning_rate": 1.2121422778357085e-05, | |
| "loss": 1.5184, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 1.3700344072568034, | |
| "grad_norm": 4.5231547355651855, | |
| "learning_rate": 1.2075078206465069e-05, | |
| "loss": 1.5492, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 1.376290272130122, | |
| "grad_norm": 3.368537425994873, | |
| "learning_rate": 1.2028733634573051e-05, | |
| "loss": 1.5519, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.3825461370034406, | |
| "grad_norm": 3.618436336517334, | |
| "learning_rate": 1.1982389062681035e-05, | |
| "loss": 1.5166, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 1.3888020018767595, | |
| "grad_norm": 3.74650502204895, | |
| "learning_rate": 1.1936044490789019e-05, | |
| "loss": 1.5313, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 1.3950578667500781, | |
| "grad_norm": 3.186457395553589, | |
| "learning_rate": 1.1889699918897e-05, | |
| "loss": 1.502, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 1.401313731623397, | |
| "grad_norm": 3.0362093448638916, | |
| "learning_rate": 1.1843355347004983e-05, | |
| "loss": 1.507, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 1.4075695964967156, | |
| "grad_norm": 3.121917486190796, | |
| "learning_rate": 1.1797010775112965e-05, | |
| "loss": 1.5148, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.4138254613700343, | |
| "grad_norm": 3.965427875518799, | |
| "learning_rate": 1.1750666203220949e-05, | |
| "loss": 1.5764, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 1.4200813262433531, | |
| "grad_norm": 4.197850227355957, | |
| "learning_rate": 1.1704321631328933e-05, | |
| "loss": 1.5838, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 1.426337191116672, | |
| "grad_norm": 3.8980863094329834, | |
| "learning_rate": 1.1657977059436913e-05, | |
| "loss": 1.5039, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 1.4325930559899906, | |
| "grad_norm": 3.8204169273376465, | |
| "learning_rate": 1.1611632487544897e-05, | |
| "loss": 1.5211, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 1.4388489208633093, | |
| "grad_norm": 4.281759262084961, | |
| "learning_rate": 1.156528791565288e-05, | |
| "loss": 1.5597, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.4451047857366282, | |
| "grad_norm": 3.467996835708618, | |
| "learning_rate": 1.1518943343760863e-05, | |
| "loss": 1.4983, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 1.4513606506099468, | |
| "grad_norm": 3.524231195449829, | |
| "learning_rate": 1.1472598771868847e-05, | |
| "loss": 1.488, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 1.4576165154832657, | |
| "grad_norm": 3.9286301136016846, | |
| "learning_rate": 1.142625419997683e-05, | |
| "loss": 1.5314, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 1.4638723803565843, | |
| "grad_norm": 3.6402275562286377, | |
| "learning_rate": 1.137990962808481e-05, | |
| "loss": 1.4996, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 1.470128245229903, | |
| "grad_norm": 3.587326765060425, | |
| "learning_rate": 1.1333565056192795e-05, | |
| "loss": 1.5019, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.4763841101032218, | |
| "grad_norm": 3.7835605144500732, | |
| "learning_rate": 1.1287220484300778e-05, | |
| "loss": 1.5088, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 1.4826399749765404, | |
| "grad_norm": 3.7514798641204834, | |
| "learning_rate": 1.124087591240876e-05, | |
| "loss": 1.5111, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 1.4888958398498593, | |
| "grad_norm": 3.885326623916626, | |
| "learning_rate": 1.1194531340516743e-05, | |
| "loss": 1.5366, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 1.495151704723178, | |
| "grad_norm": 3.762674331665039, | |
| "learning_rate": 1.1148186768624725e-05, | |
| "loss": 1.5278, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 1.5014075695964966, | |
| "grad_norm": 4.042783260345459, | |
| "learning_rate": 1.1101842196732708e-05, | |
| "loss": 1.5166, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.5076634344698154, | |
| "grad_norm": 4.318854331970215, | |
| "learning_rate": 1.1055497624840692e-05, | |
| "loss": 1.4609, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 1.5139192993431343, | |
| "grad_norm": 3.423506498336792, | |
| "learning_rate": 1.1009153052948673e-05, | |
| "loss": 1.5034, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 1.520175164216453, | |
| "grad_norm": 3.9718987941741943, | |
| "learning_rate": 1.0962808481056656e-05, | |
| "loss": 1.4938, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 1.5264310290897716, | |
| "grad_norm": 3.8156068325042725, | |
| "learning_rate": 1.091646390916464e-05, | |
| "loss": 1.5344, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 1.5326868939630904, | |
| "grad_norm": 3.734156370162964, | |
| "learning_rate": 1.0870119337272622e-05, | |
| "loss": 1.4922, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 1.5389427588364093, | |
| "grad_norm": 3.59214186668396, | |
| "learning_rate": 1.0823774765380606e-05, | |
| "loss": 1.5005, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 1.545198623709728, | |
| "grad_norm": 3.830472946166992, | |
| "learning_rate": 1.077743019348859e-05, | |
| "loss": 1.435, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 1.5514544885830466, | |
| "grad_norm": 3.512565851211548, | |
| "learning_rate": 1.073108562159657e-05, | |
| "loss": 1.5366, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 1.5577103534563652, | |
| "grad_norm": 4.332693576812744, | |
| "learning_rate": 1.0684741049704554e-05, | |
| "loss": 1.5388, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 1.563966218329684, | |
| "grad_norm": 4.484791278839111, | |
| "learning_rate": 1.0638396477812538e-05, | |
| "loss": 1.4778, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.563966218329684, | |
| "eval_loss": 0.5325935482978821, | |
| "eval_runtime": 30.2979, | |
| "eval_samples_per_second": 15.975, | |
| "eval_steps_per_second": 3.994, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.570222083203003, | |
| "grad_norm": 3.519361972808838, | |
| "learning_rate": 1.059205190592052e-05, | |
| "loss": 1.5333, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 1.5764779480763216, | |
| "grad_norm": 4.101566791534424, | |
| "learning_rate": 1.0545707334028504e-05, | |
| "loss": 1.4594, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 1.5827338129496402, | |
| "grad_norm": 3.4529476165771484, | |
| "learning_rate": 1.0499362762136484e-05, | |
| "loss": 1.5263, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 1.588989677822959, | |
| "grad_norm": 3.918497085571289, | |
| "learning_rate": 1.0453018190244468e-05, | |
| "loss": 1.5055, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 1.5952455426962777, | |
| "grad_norm": 4.017299175262451, | |
| "learning_rate": 1.0406673618352452e-05, | |
| "loss": 1.5578, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 1.6015014075695966, | |
| "grad_norm": 3.6538748741149902, | |
| "learning_rate": 1.0360329046460434e-05, | |
| "loss": 1.4732, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 1.6077572724429152, | |
| "grad_norm": 3.3641505241394043, | |
| "learning_rate": 1.0313984474568418e-05, | |
| "loss": 1.5595, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 1.6140131373162339, | |
| "grad_norm": 4.010868549346924, | |
| "learning_rate": 1.0267639902676401e-05, | |
| "loss": 1.4979, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 1.6202690021895527, | |
| "grad_norm": 3.882277488708496, | |
| "learning_rate": 1.0221295330784382e-05, | |
| "loss": 1.4327, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 1.6265248670628716, | |
| "grad_norm": 3.3188819885253906, | |
| "learning_rate": 1.0174950758892366e-05, | |
| "loss": 1.499, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.6327807319361902, | |
| "grad_norm": 3.892775058746338, | |
| "learning_rate": 1.012860618700035e-05, | |
| "loss": 1.4713, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 1.6390365968095089, | |
| "grad_norm": 4.000039100646973, | |
| "learning_rate": 1.0082261615108331e-05, | |
| "loss": 1.4392, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 1.6452924616828275, | |
| "grad_norm": 3.8597168922424316, | |
| "learning_rate": 1.0035917043216314e-05, | |
| "loss": 1.5019, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 1.6515483265561464, | |
| "grad_norm": 3.5325751304626465, | |
| "learning_rate": 9.989572471324297e-06, | |
| "loss": 1.5157, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 1.6578041914294652, | |
| "grad_norm": 3.6834187507629395, | |
| "learning_rate": 9.94322789943228e-06, | |
| "loss": 1.4627, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 1.6640600563027839, | |
| "grad_norm": 3.7806499004364014, | |
| "learning_rate": 9.896883327540263e-06, | |
| "loss": 1.4858, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 1.6703159211761025, | |
| "grad_norm": 4.8793044090271, | |
| "learning_rate": 9.850538755648245e-06, | |
| "loss": 1.5186, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 1.6765717860494214, | |
| "grad_norm": 3.3358120918273926, | |
| "learning_rate": 9.804194183756227e-06, | |
| "loss": 1.476, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 1.6828276509227402, | |
| "grad_norm": 3.562429904937744, | |
| "learning_rate": 9.757849611864211e-06, | |
| "loss": 1.4551, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 1.6890835157960589, | |
| "grad_norm": 3.358164072036743, | |
| "learning_rate": 9.711505039972195e-06, | |
| "loss": 1.4587, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.6953393806693775, | |
| "grad_norm": 3.7120015621185303, | |
| "learning_rate": 9.665160468080177e-06, | |
| "loss": 1.509, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 1.7015952455426961, | |
| "grad_norm": 3.393617868423462, | |
| "learning_rate": 9.61881589618816e-06, | |
| "loss": 1.4466, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 1.707851110416015, | |
| "grad_norm": 3.6086132526397705, | |
| "learning_rate": 9.572471324296143e-06, | |
| "loss": 1.5047, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 1.7141069752893339, | |
| "grad_norm": 3.6902570724487305, | |
| "learning_rate": 9.526126752404125e-06, | |
| "loss": 1.4928, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 1.7203628401626525, | |
| "grad_norm": 3.3542566299438477, | |
| "learning_rate": 9.479782180512109e-06, | |
| "loss": 1.516, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.7266187050359711, | |
| "grad_norm": 3.9466335773468018, | |
| "learning_rate": 9.433437608620091e-06, | |
| "loss": 1.5194, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 1.73287456990929, | |
| "grad_norm": 3.526545286178589, | |
| "learning_rate": 9.387093036728075e-06, | |
| "loss": 1.469, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 3.920001745223999, | |
| "learning_rate": 9.340748464836057e-06, | |
| "loss": 1.4877, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 1.7453862996559275, | |
| "grad_norm": 3.415998935699463, | |
| "learning_rate": 9.294403892944039e-06, | |
| "loss": 1.477, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 1.7516421645292461, | |
| "grad_norm": 4.152552127838135, | |
| "learning_rate": 9.248059321052023e-06, | |
| "loss": 1.4598, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.7578980294025648, | |
| "grad_norm": 4.323616027832031, | |
| "learning_rate": 9.201714749160005e-06, | |
| "loss": 1.4412, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 1.7641538942758836, | |
| "grad_norm": 4.302744388580322, | |
| "learning_rate": 9.155370177267989e-06, | |
| "loss": 1.4885, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 1.7704097591492025, | |
| "grad_norm": 3.2638041973114014, | |
| "learning_rate": 9.10902560537597e-06, | |
| "loss": 1.4948, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 1.7766656240225212, | |
| "grad_norm": 3.4382920265197754, | |
| "learning_rate": 9.062681033483955e-06, | |
| "loss": 1.4491, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 1.7829214888958398, | |
| "grad_norm": 3.907999038696289, | |
| "learning_rate": 9.016336461591937e-06, | |
| "loss": 1.4573, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 1.7891773537691584, | |
| "grad_norm": 4.269818305969238, | |
| "learning_rate": 8.969991889699919e-06, | |
| "loss": 1.4932, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 1.7954332186424773, | |
| "grad_norm": 4.269619941711426, | |
| "learning_rate": 8.923647317807903e-06, | |
| "loss": 1.4468, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 1.8016890835157962, | |
| "grad_norm": 3.4848062992095947, | |
| "learning_rate": 8.877302745915886e-06, | |
| "loss": 1.4679, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 1.8079449483891148, | |
| "grad_norm": 3.567715644836426, | |
| "learning_rate": 8.830958174023868e-06, | |
| "loss": 1.4974, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 1.8142008132624334, | |
| "grad_norm": 4.031379222869873, | |
| "learning_rate": 8.78461360213185e-06, | |
| "loss": 1.4192, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.8204566781357523, | |
| "grad_norm": 3.2717881202697754, | |
| "learning_rate": 8.738269030239834e-06, | |
| "loss": 1.4905, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 1.8267125430090712, | |
| "grad_norm": 3.7909555435180664, | |
| "learning_rate": 8.691924458347816e-06, | |
| "loss": 1.4595, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 1.8329684078823898, | |
| "grad_norm": 3.242023468017578, | |
| "learning_rate": 8.645579886455799e-06, | |
| "loss": 1.4484, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 1.8392242727557084, | |
| "grad_norm": 3.605093240737915, | |
| "learning_rate": 8.599235314563782e-06, | |
| "loss": 1.4915, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 1.845480137629027, | |
| "grad_norm": 3.832170248031616, | |
| "learning_rate": 8.552890742671766e-06, | |
| "loss": 1.476, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 1.851736002502346, | |
| "grad_norm": 3.6190032958984375, | |
| "learning_rate": 8.506546170779748e-06, | |
| "loss": 1.4179, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 1.8579918673756648, | |
| "grad_norm": 4.795033931732178, | |
| "learning_rate": 8.46020159888773e-06, | |
| "loss": 1.4823, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 1.8642477322489834, | |
| "grad_norm": 4.25372314453125, | |
| "learning_rate": 8.413857026995714e-06, | |
| "loss": 1.4547, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 1.870503597122302, | |
| "grad_norm": 3.29866099357605, | |
| "learning_rate": 8.367512455103696e-06, | |
| "loss": 1.4366, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 1.876759461995621, | |
| "grad_norm": 3.771183967590332, | |
| "learning_rate": 8.32116788321168e-06, | |
| "loss": 1.5107, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.876759461995621, | |
| "eval_loss": 0.5249506831169128, | |
| "eval_runtime": 30.3312, | |
| "eval_samples_per_second": 15.957, | |
| "eval_steps_per_second": 3.989, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.8830153268689396, | |
| "grad_norm": 3.4355130195617676, | |
| "learning_rate": 8.274823311319662e-06, | |
| "loss": 1.4839, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 1.8892711917422584, | |
| "grad_norm": 3.5628068447113037, | |
| "learning_rate": 8.228478739427646e-06, | |
| "loss": 1.5049, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 1.895527056615577, | |
| "grad_norm": 3.52653169631958, | |
| "learning_rate": 8.182134167535628e-06, | |
| "loss": 1.4729, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 1.9017829214888957, | |
| "grad_norm": 4.024752140045166, | |
| "learning_rate": 8.13578959564361e-06, | |
| "loss": 1.4413, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 1.9080387863622146, | |
| "grad_norm": 4.114953994750977, | |
| "learning_rate": 8.089445023751594e-06, | |
| "loss": 1.4517, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 1.9142946512355334, | |
| "grad_norm": 3.832791328430176, | |
| "learning_rate": 8.043100451859576e-06, | |
| "loss": 1.4576, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 1.920550516108852, | |
| "grad_norm": 3.5607106685638428, | |
| "learning_rate": 7.99675587996756e-06, | |
| "loss": 1.4813, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 1.9268063809821707, | |
| "grad_norm": 3.79740309715271, | |
| "learning_rate": 7.950411308075542e-06, | |
| "loss": 1.4886, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 1.9330622458554894, | |
| "grad_norm": 3.8134922981262207, | |
| "learning_rate": 7.904066736183526e-06, | |
| "loss": 1.425, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 1.9393181107288082, | |
| "grad_norm": 4.097330570220947, | |
| "learning_rate": 7.857722164291508e-06, | |
| "loss": 1.5026, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.945573975602127, | |
| "grad_norm": 3.597710609436035, | |
| "learning_rate": 7.81137759239949e-06, | |
| "loss": 1.4355, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 1.9518298404754457, | |
| "grad_norm": 4.481216907501221, | |
| "learning_rate": 7.765033020507474e-06, | |
| "loss": 1.4761, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 1.9580857053487644, | |
| "grad_norm": 3.9572675228118896, | |
| "learning_rate": 7.718688448615457e-06, | |
| "loss": 1.4684, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 1.9643415702220832, | |
| "grad_norm": 3.0749871730804443, | |
| "learning_rate": 7.67234387672344e-06, | |
| "loss": 1.4369, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 1.970597435095402, | |
| "grad_norm": 4.008276462554932, | |
| "learning_rate": 7.625999304831422e-06, | |
| "loss": 1.4414, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 1.9768532999687207, | |
| "grad_norm": 3.3780014514923096, | |
| "learning_rate": 7.5796547329394054e-06, | |
| "loss": 1.4795, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 1.9831091648420394, | |
| "grad_norm": 3.802067756652832, | |
| "learning_rate": 7.533310161047388e-06, | |
| "loss": 1.4506, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 1.989365029715358, | |
| "grad_norm": 3.7218985557556152, | |
| "learning_rate": 7.4869655891553705e-06, | |
| "loss": 1.4379, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 1.9956208945886769, | |
| "grad_norm": 3.222632646560669, | |
| "learning_rate": 7.440621017263354e-06, | |
| "loss": 1.4354, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 2.0018767594619957, | |
| "grad_norm": 3.8669259548187256, | |
| "learning_rate": 7.394276445371336e-06, | |
| "loss": 1.4626, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 2.0081326243353144, | |
| "grad_norm": 3.748908281326294, | |
| "learning_rate": 7.347931873479319e-06, | |
| "loss": 1.4119, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 2.014388489208633, | |
| "grad_norm": 3.8527488708496094, | |
| "learning_rate": 7.301587301587301e-06, | |
| "loss": 1.468, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 2.0206443540819516, | |
| "grad_norm": 4.265289783477783, | |
| "learning_rate": 7.255242729695285e-06, | |
| "loss": 1.411, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 2.0269002189552707, | |
| "grad_norm": 3.580428123474121, | |
| "learning_rate": 7.208898157803268e-06, | |
| "loss": 1.4447, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 2.0331560838285894, | |
| "grad_norm": 3.5705726146698, | |
| "learning_rate": 7.16255358591125e-06, | |
| "loss": 1.4475, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.039411948701908, | |
| "grad_norm": 4.200178146362305, | |
| "learning_rate": 7.116209014019233e-06, | |
| "loss": 1.4493, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 2.0456678135752266, | |
| "grad_norm": 3.9091575145721436, | |
| "learning_rate": 7.069864442127217e-06, | |
| "loss": 1.4842, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 2.0519236784485457, | |
| "grad_norm": 4.206997394561768, | |
| "learning_rate": 7.023519870235199e-06, | |
| "loss": 1.4528, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 2.0581795433218644, | |
| "grad_norm": 3.746002435684204, | |
| "learning_rate": 6.977175298343182e-06, | |
| "loss": 1.4081, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 2.064435408195183, | |
| "grad_norm": 4.167978286743164, | |
| "learning_rate": 6.930830726451166e-06, | |
| "loss": 1.4367, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 2.0706912730685016, | |
| "grad_norm": 3.922278881072998, | |
| "learning_rate": 6.884486154559148e-06, | |
| "loss": 1.4745, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 2.0769471379418203, | |
| "grad_norm": 4.390613555908203, | |
| "learning_rate": 6.838141582667131e-06, | |
| "loss": 1.4456, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 2.0832030028151394, | |
| "grad_norm": 3.3324151039123535, | |
| "learning_rate": 6.791797010775113e-06, | |
| "loss": 1.4202, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 2.089458867688458, | |
| "grad_norm": 3.823288917541504, | |
| "learning_rate": 6.745452438883097e-06, | |
| "loss": 1.4624, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 2.0957147325617766, | |
| "grad_norm": 3.162973403930664, | |
| "learning_rate": 6.699107866991079e-06, | |
| "loss": 1.4481, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 2.1019705974350953, | |
| "grad_norm": 3.888578176498413, | |
| "learning_rate": 6.652763295099062e-06, | |
| "loss": 1.3965, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 2.108226462308414, | |
| "grad_norm": 3.485076665878296, | |
| "learning_rate": 6.606418723207046e-06, | |
| "loss": 1.4791, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 2.114482327181733, | |
| "grad_norm": 4.454657554626465, | |
| "learning_rate": 6.560074151315028e-06, | |
| "loss": 1.4502, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 2.1207381920550517, | |
| "grad_norm": 4.14171838760376, | |
| "learning_rate": 6.513729579423011e-06, | |
| "loss": 1.4569, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 2.1269940569283703, | |
| "grad_norm": 3.889117956161499, | |
| "learning_rate": 6.467385007530993e-06, | |
| "loss": 1.4173, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 2.133249921801689, | |
| "grad_norm": 4.004756927490234, | |
| "learning_rate": 6.4210404356389765e-06, | |
| "loss": 1.4278, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 2.139505786675008, | |
| "grad_norm": 3.7619662284851074, | |
| "learning_rate": 6.3746958637469595e-06, | |
| "loss": 1.4279, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 2.1457616515483267, | |
| "grad_norm": 4.004810333251953, | |
| "learning_rate": 6.3283512918549416e-06, | |
| "loss": 1.4657, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 2.1520175164216453, | |
| "grad_norm": 5.289163112640381, | |
| "learning_rate": 6.282006719962925e-06, | |
| "loss": 1.4271, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 2.158273381294964, | |
| "grad_norm": 3.6333744525909424, | |
| "learning_rate": 6.2356621480709074e-06, | |
| "loss": 1.4337, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 2.1645292461682826, | |
| "grad_norm": 4.031957626342773, | |
| "learning_rate": 6.18931757617889e-06, | |
| "loss": 1.4133, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 2.1707851110416017, | |
| "grad_norm": 3.917013168334961, | |
| "learning_rate": 6.142973004286873e-06, | |
| "loss": 1.4533, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 2.1770409759149203, | |
| "grad_norm": 3.9864284992218018, | |
| "learning_rate": 6.096628432394856e-06, | |
| "loss": 1.4549, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 2.183296840788239, | |
| "grad_norm": 3.6606876850128174, | |
| "learning_rate": 6.050283860502839e-06, | |
| "loss": 1.4091, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 2.1895527056615576, | |
| "grad_norm": 4.302699565887451, | |
| "learning_rate": 6.003939288610821e-06, | |
| "loss": 1.4914, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.1895527056615576, | |
| "eval_loss": 0.5183842778205872, | |
| "eval_runtime": 30.127, | |
| "eval_samples_per_second": 16.065, | |
| "eval_steps_per_second": 4.016, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.1958085705348767, | |
| "grad_norm": 4.072482109069824, | |
| "learning_rate": 5.957594716718805e-06, | |
| "loss": 1.4146, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 2.2020644354081953, | |
| "grad_norm": 3.948002815246582, | |
| "learning_rate": 5.911250144826788e-06, | |
| "loss": 1.4575, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 2.208320300281514, | |
| "grad_norm": 3.2207224369049072, | |
| "learning_rate": 5.86490557293477e-06, | |
| "loss": 1.3987, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 2.2145761651548326, | |
| "grad_norm": 3.5877556800842285, | |
| "learning_rate": 5.818561001042753e-06, | |
| "loss": 1.4113, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 2.220832030028151, | |
| "grad_norm": 4.221316337585449, | |
| "learning_rate": 5.772216429150737e-06, | |
| "loss": 1.4649, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 2.2270878949014703, | |
| "grad_norm": 3.4622902870178223, | |
| "learning_rate": 5.725871857258719e-06, | |
| "loss": 1.4291, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 2.233343759774789, | |
| "grad_norm": 3.8495419025421143, | |
| "learning_rate": 5.679527285366702e-06, | |
| "loss": 1.4492, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 2.2395996246481076, | |
| "grad_norm": 4.676374912261963, | |
| "learning_rate": 5.633182713474685e-06, | |
| "loss": 1.4219, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 2.245855489521426, | |
| "grad_norm": 5.073587894439697, | |
| "learning_rate": 5.586838141582668e-06, | |
| "loss": 1.4784, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 2.252111354394745, | |
| "grad_norm": 3.6623458862304688, | |
| "learning_rate": 5.54049356969065e-06, | |
| "loss": 1.4654, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 2.258367219268064, | |
| "grad_norm": 3.8176751136779785, | |
| "learning_rate": 5.494148997798633e-06, | |
| "loss": 1.384, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 2.2646230841413826, | |
| "grad_norm": 4.066217422485352, | |
| "learning_rate": 5.447804425906617e-06, | |
| "loss": 1.4125, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 2.270878949014701, | |
| "grad_norm": 3.977444887161255, | |
| "learning_rate": 5.401459854014599e-06, | |
| "loss": 1.4492, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 2.27713481388802, | |
| "grad_norm": 3.8999898433685303, | |
| "learning_rate": 5.355115282122582e-06, | |
| "loss": 1.3982, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 2.283390678761339, | |
| "grad_norm": 3.845292329788208, | |
| "learning_rate": 5.3087707102305655e-06, | |
| "loss": 1.4199, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 2.2896465436346576, | |
| "grad_norm": 4.052536964416504, | |
| "learning_rate": 5.262426138338548e-06, | |
| "loss": 1.4316, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 2.2959024085079762, | |
| "grad_norm": 5.119332790374756, | |
| "learning_rate": 5.2160815664465305e-06, | |
| "loss": 1.4084, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 2.302158273381295, | |
| "grad_norm": 4.279220104217529, | |
| "learning_rate": 5.169736994554513e-06, | |
| "loss": 1.4383, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 2.308414138254614, | |
| "grad_norm": 4.149873733520508, | |
| "learning_rate": 5.123392422662496e-06, | |
| "loss": 1.4221, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 2.3146700031279326, | |
| "grad_norm": 4.548859119415283, | |
| "learning_rate": 5.077047850770479e-06, | |
| "loss": 1.3936, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 2.3209258680012512, | |
| "grad_norm": 4.159961700439453, | |
| "learning_rate": 5.0307032788784615e-06, | |
| "loss": 1.4052, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 2.32718173287457, | |
| "grad_norm": 3.7540910243988037, | |
| "learning_rate": 4.984358706986444e-06, | |
| "loss": 1.416, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 2.3334375977478885, | |
| "grad_norm": 4.482418537139893, | |
| "learning_rate": 4.938014135094427e-06, | |
| "loss": 1.4299, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 2.3396934626212076, | |
| "grad_norm": 4.642965316772461, | |
| "learning_rate": 4.89166956320241e-06, | |
| "loss": 1.3794, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 2.3459493274945262, | |
| "grad_norm": 3.7981739044189453, | |
| "learning_rate": 4.845324991310393e-06, | |
| "loss": 1.4147, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.352205192367845, | |
| "grad_norm": 3.6449689865112305, | |
| "learning_rate": 4.798980419418376e-06, | |
| "loss": 1.3822, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 2.3584610572411635, | |
| "grad_norm": 4.532881736755371, | |
| "learning_rate": 4.752635847526359e-06, | |
| "loss": 1.4283, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 2.364716922114482, | |
| "grad_norm": 4.486178874969482, | |
| "learning_rate": 4.706291275634341e-06, | |
| "loss": 1.4043, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 2.3709727869878012, | |
| "grad_norm": 3.9950759410858154, | |
| "learning_rate": 4.659946703742325e-06, | |
| "loss": 1.4632, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 2.37722865186112, | |
| "grad_norm": 3.486673355102539, | |
| "learning_rate": 4.613602131850307e-06, | |
| "loss": 1.4193, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 2.3834845167344385, | |
| "grad_norm": 3.896305799484253, | |
| "learning_rate": 4.56725755995829e-06, | |
| "loss": 1.3879, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 2.389740381607757, | |
| "grad_norm": 3.072448492050171, | |
| "learning_rate": 4.520912988066273e-06, | |
| "loss": 1.4173, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 2.395996246481076, | |
| "grad_norm": 3.844083547592163, | |
| "learning_rate": 4.474568416174256e-06, | |
| "loss": 1.3865, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 2.402252111354395, | |
| "grad_norm": 4.288559436798096, | |
| "learning_rate": 4.428223844282239e-06, | |
| "loss": 1.4044, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 2.4085079762277135, | |
| "grad_norm": 3.5817108154296875, | |
| "learning_rate": 4.381879272390222e-06, | |
| "loss": 1.4128, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 2.414763841101032, | |
| "grad_norm": 4.0619282722473145, | |
| "learning_rate": 4.335534700498205e-06, | |
| "loss": 1.4721, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 2.421019705974351, | |
| "grad_norm": 3.2147204875946045, | |
| "learning_rate": 4.289190128606187e-06, | |
| "loss": 1.4621, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 2.42727557084767, | |
| "grad_norm": 3.896071195602417, | |
| "learning_rate": 4.24284555671417e-06, | |
| "loss": 1.4041, | |
| "step": 7760 | |
| }, | |
| { | |
| "epoch": 2.4335314357209885, | |
| "grad_norm": 3.7601754665374756, | |
| "learning_rate": 4.196500984822154e-06, | |
| "loss": 1.3929, | |
| "step": 7780 | |
| }, | |
| { | |
| "epoch": 2.439787300594307, | |
| "grad_norm": 3.753279685974121, | |
| "learning_rate": 4.150156412930136e-06, | |
| "loss": 1.4418, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 2.446043165467626, | |
| "grad_norm": 3.9221978187561035, | |
| "learning_rate": 4.103811841038119e-06, | |
| "loss": 1.3875, | |
| "step": 7820 | |
| }, | |
| { | |
| "epoch": 2.452299030340945, | |
| "grad_norm": 3.614114761352539, | |
| "learning_rate": 4.057467269146102e-06, | |
| "loss": 1.3828, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 2.4585548952142635, | |
| "grad_norm": 4.317903518676758, | |
| "learning_rate": 4.0111226972540846e-06, | |
| "loss": 1.4378, | |
| "step": 7860 | |
| }, | |
| { | |
| "epoch": 2.464810760087582, | |
| "grad_norm": 4.5976738929748535, | |
| "learning_rate": 3.964778125362067e-06, | |
| "loss": 1.3863, | |
| "step": 7880 | |
| }, | |
| { | |
| "epoch": 2.471066624960901, | |
| "grad_norm": 4.07506799697876, | |
| "learning_rate": 3.9184335534700504e-06, | |
| "loss": 1.3808, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 2.4773224898342194, | |
| "grad_norm": 4.548733711242676, | |
| "learning_rate": 3.872088981578033e-06, | |
| "loss": 1.4063, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 2.4835783547075385, | |
| "grad_norm": 4.166104316711426, | |
| "learning_rate": 3.8257444096860155e-06, | |
| "loss": 1.3962, | |
| "step": 7940 | |
| }, | |
| { | |
| "epoch": 2.489834219580857, | |
| "grad_norm": 3.7219245433807373, | |
| "learning_rate": 3.779399837793999e-06, | |
| "loss": 1.4458, | |
| "step": 7960 | |
| }, | |
| { | |
| "epoch": 2.496090084454176, | |
| "grad_norm": 4.242513179779053, | |
| "learning_rate": 3.7330552659019814e-06, | |
| "loss": 1.4342, | |
| "step": 7980 | |
| }, | |
| { | |
| "epoch": 2.5023459493274944, | |
| "grad_norm": 3.4067800045013428, | |
| "learning_rate": 3.6867106940099643e-06, | |
| "loss": 1.3884, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.5023459493274944, | |
| "eval_loss": 0.5156850814819336, | |
| "eval_runtime": 30.2, | |
| "eval_samples_per_second": 16.027, | |
| "eval_steps_per_second": 4.007, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.508601814200813, | |
| "grad_norm": 3.91298508644104, | |
| "learning_rate": 3.640366122117947e-06, | |
| "loss": 1.3653, | |
| "step": 8020 | |
| }, | |
| { | |
| "epoch": 2.514857679074132, | |
| "grad_norm": 3.830792188644409, | |
| "learning_rate": 3.5940215502259302e-06, | |
| "loss": 1.3945, | |
| "step": 8040 | |
| }, | |
| { | |
| "epoch": 2.521113543947451, | |
| "grad_norm": 3.802927017211914, | |
| "learning_rate": 3.5476769783339127e-06, | |
| "loss": 1.3996, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 2.5273694088207694, | |
| "grad_norm": 4.000018119812012, | |
| "learning_rate": 3.5013324064418957e-06, | |
| "loss": 1.4302, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 2.533625273694088, | |
| "grad_norm": 4.325403690338135, | |
| "learning_rate": 3.454987834549879e-06, | |
| "loss": 1.4083, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 2.5398811385674067, | |
| "grad_norm": 3.761460781097412, | |
| "learning_rate": 3.4086432626578616e-06, | |
| "loss": 1.3648, | |
| "step": 8120 | |
| }, | |
| { | |
| "epoch": 2.546137003440726, | |
| "grad_norm": 3.756761312484741, | |
| "learning_rate": 3.3622986907658445e-06, | |
| "loss": 1.3559, | |
| "step": 8140 | |
| }, | |
| { | |
| "epoch": 2.5523928683140444, | |
| "grad_norm": 3.938476085662842, | |
| "learning_rate": 3.315954118873827e-06, | |
| "loss": 1.4314, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 2.558648733187363, | |
| "grad_norm": 3.8850269317626953, | |
| "learning_rate": 3.26960954698181e-06, | |
| "loss": 1.4357, | |
| "step": 8180 | |
| }, | |
| { | |
| "epoch": 2.564904598060682, | |
| "grad_norm": 4.117141246795654, | |
| "learning_rate": 3.2232649750897925e-06, | |
| "loss": 1.3784, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 2.5711604629340004, | |
| "grad_norm": 3.2003445625305176, | |
| "learning_rate": 3.176920403197776e-06, | |
| "loss": 1.4187, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 2.5774163278073194, | |
| "grad_norm": 3.5441741943359375, | |
| "learning_rate": 3.130575831305759e-06, | |
| "loss": 1.3897, | |
| "step": 8240 | |
| }, | |
| { | |
| "epoch": 2.583672192680638, | |
| "grad_norm": 4.082095623016357, | |
| "learning_rate": 3.0842312594137413e-06, | |
| "loss": 1.3901, | |
| "step": 8260 | |
| }, | |
| { | |
| "epoch": 2.5899280575539567, | |
| "grad_norm": 3.878877878189087, | |
| "learning_rate": 3.0378866875217243e-06, | |
| "loss": 1.4273, | |
| "step": 8280 | |
| }, | |
| { | |
| "epoch": 2.596183922427276, | |
| "grad_norm": 4.063547134399414, | |
| "learning_rate": 2.991542115629707e-06, | |
| "loss": 1.4536, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 2.6024397873005944, | |
| "grad_norm": 5.6043596267700195, | |
| "learning_rate": 2.94519754373769e-06, | |
| "loss": 1.3799, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 2.608695652173913, | |
| "grad_norm": 3.9058849811553955, | |
| "learning_rate": 2.8988529718456727e-06, | |
| "loss": 1.4384, | |
| "step": 8340 | |
| }, | |
| { | |
| "epoch": 2.6149515170472317, | |
| "grad_norm": 4.544114112854004, | |
| "learning_rate": 2.8525083999536556e-06, | |
| "loss": 1.4154, | |
| "step": 8360 | |
| }, | |
| { | |
| "epoch": 2.6212073819205504, | |
| "grad_norm": 4.125789165496826, | |
| "learning_rate": 2.8061638280616386e-06, | |
| "loss": 1.348, | |
| "step": 8380 | |
| }, | |
| { | |
| "epoch": 2.6274632467938694, | |
| "grad_norm": 3.849442958831787, | |
| "learning_rate": 2.7598192561696215e-06, | |
| "loss": 1.3859, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 2.633719111667188, | |
| "grad_norm": 3.8428611755371094, | |
| "learning_rate": 2.7134746842776045e-06, | |
| "loss": 1.3743, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 2.6399749765405067, | |
| "grad_norm": 3.9532744884490967, | |
| "learning_rate": 2.667130112385587e-06, | |
| "loss": 1.3664, | |
| "step": 8440 | |
| }, | |
| { | |
| "epoch": 2.6462308414138254, | |
| "grad_norm": 4.1256232261657715, | |
| "learning_rate": 2.62078554049357e-06, | |
| "loss": 1.4149, | |
| "step": 8460 | |
| }, | |
| { | |
| "epoch": 2.652486706287144, | |
| "grad_norm": 4.687183856964111, | |
| "learning_rate": 2.5744409686015525e-06, | |
| "loss": 1.3974, | |
| "step": 8480 | |
| }, | |
| { | |
| "epoch": 2.658742571160463, | |
| "grad_norm": 4.005310535430908, | |
| "learning_rate": 2.528096396709536e-06, | |
| "loss": 1.3903, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.6649984360337817, | |
| "grad_norm": 3.9469504356384277, | |
| "learning_rate": 2.4817518248175183e-06, | |
| "loss": 1.3635, | |
| "step": 8520 | |
| }, | |
| { | |
| "epoch": 2.6712543009071004, | |
| "grad_norm": 3.4386444091796875, | |
| "learning_rate": 2.4354072529255013e-06, | |
| "loss": 1.4357, | |
| "step": 8540 | |
| }, | |
| { | |
| "epoch": 2.677510165780419, | |
| "grad_norm": 3.956700563430786, | |
| "learning_rate": 2.3890626810334842e-06, | |
| "loss": 1.3847, | |
| "step": 8560 | |
| }, | |
| { | |
| "epoch": 2.6837660306537376, | |
| "grad_norm": 3.6167428493499756, | |
| "learning_rate": 2.3427181091414668e-06, | |
| "loss": 1.3627, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 2.6900218955270567, | |
| "grad_norm": 4.528804779052734, | |
| "learning_rate": 2.2963735372494497e-06, | |
| "loss": 1.3817, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 2.6962777604003754, | |
| "grad_norm": 4.087806701660156, | |
| "learning_rate": 2.2500289653574326e-06, | |
| "loss": 1.4025, | |
| "step": 8620 | |
| }, | |
| { | |
| "epoch": 2.702533625273694, | |
| "grad_norm": 3.7826669216156006, | |
| "learning_rate": 2.2036843934654156e-06, | |
| "loss": 1.3574, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 2.708789490147013, | |
| "grad_norm": 3.944565534591675, | |
| "learning_rate": 2.1573398215733985e-06, | |
| "loss": 1.4142, | |
| "step": 8660 | |
| }, | |
| { | |
| "epoch": 2.7150453550203313, | |
| "grad_norm": 3.3899853229522705, | |
| "learning_rate": 2.110995249681381e-06, | |
| "loss": 1.41, | |
| "step": 8680 | |
| }, | |
| { | |
| "epoch": 2.7213012198936504, | |
| "grad_norm": 3.7669284343719482, | |
| "learning_rate": 2.064650677789364e-06, | |
| "loss": 1.434, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 2.727557084766969, | |
| "grad_norm": 4.533022880554199, | |
| "learning_rate": 2.018306105897347e-06, | |
| "loss": 1.4266, | |
| "step": 8720 | |
| }, | |
| { | |
| "epoch": 2.7338129496402876, | |
| "grad_norm": 4.264920711517334, | |
| "learning_rate": 1.97196153400533e-06, | |
| "loss": 1.3866, | |
| "step": 8740 | |
| }, | |
| { | |
| "epoch": 2.7400688145136067, | |
| "grad_norm": 3.9496846199035645, | |
| "learning_rate": 1.9256169621133124e-06, | |
| "loss": 1.3872, | |
| "step": 8760 | |
| }, | |
| { | |
| "epoch": 2.7463246793869254, | |
| "grad_norm": 3.966773271560669, | |
| "learning_rate": 1.8792723902212956e-06, | |
| "loss": 1.4064, | |
| "step": 8780 | |
| }, | |
| { | |
| "epoch": 2.752580544260244, | |
| "grad_norm": 3.9621760845184326, | |
| "learning_rate": 1.8329278183292783e-06, | |
| "loss": 1.372, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 2.7588364091335627, | |
| "grad_norm": 3.506624698638916, | |
| "learning_rate": 1.7865832464372612e-06, | |
| "loss": 1.355, | |
| "step": 8820 | |
| }, | |
| { | |
| "epoch": 2.7650922740068813, | |
| "grad_norm": 4.213022708892822, | |
| "learning_rate": 1.740238674545244e-06, | |
| "loss": 1.4216, | |
| "step": 8840 | |
| }, | |
| { | |
| "epoch": 2.7713481388802004, | |
| "grad_norm": 4.381209373474121, | |
| "learning_rate": 1.693894102653227e-06, | |
| "loss": 1.3929, | |
| "step": 8860 | |
| }, | |
| { | |
| "epoch": 2.777604003753519, | |
| "grad_norm": 3.455552101135254, | |
| "learning_rate": 1.6475495307612097e-06, | |
| "loss": 1.3575, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 2.7838598686268377, | |
| "grad_norm": 4.170933723449707, | |
| "learning_rate": 1.6012049588691924e-06, | |
| "loss": 1.3772, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 2.7901157335001563, | |
| "grad_norm": 3.778902292251587, | |
| "learning_rate": 1.5548603869771755e-06, | |
| "loss": 1.3941, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 2.796371598373475, | |
| "grad_norm": 3.5581791400909424, | |
| "learning_rate": 1.5085158150851583e-06, | |
| "loss": 1.3799, | |
| "step": 8940 | |
| }, | |
| { | |
| "epoch": 2.802627463246794, | |
| "grad_norm": 3.6567673683166504, | |
| "learning_rate": 1.4621712431931412e-06, | |
| "loss": 1.3902, | |
| "step": 8960 | |
| }, | |
| { | |
| "epoch": 2.8088833281201127, | |
| "grad_norm": 4.234646797180176, | |
| "learning_rate": 1.415826671301124e-06, | |
| "loss": 1.3859, | |
| "step": 8980 | |
| }, | |
| { | |
| "epoch": 2.8151391929934313, | |
| "grad_norm": 3.7421681880950928, | |
| "learning_rate": 1.369482099409107e-06, | |
| "loss": 1.357, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.8151391929934313, | |
| "eval_loss": 0.5130746364593506, | |
| "eval_runtime": 30.1582, | |
| "eval_samples_per_second": 16.049, | |
| "eval_steps_per_second": 4.012, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.82139505786675, | |
| "grad_norm": 4.607457160949707, | |
| "learning_rate": 1.3231375275170896e-06, | |
| "loss": 1.4083, | |
| "step": 9020 | |
| }, | |
| { | |
| "epoch": 2.8276509227400686, | |
| "grad_norm": 3.6216139793395996, | |
| "learning_rate": 1.2767929556250724e-06, | |
| "loss": 1.3613, | |
| "step": 9040 | |
| }, | |
| { | |
| "epoch": 2.8339067876133877, | |
| "grad_norm": 4.228906154632568, | |
| "learning_rate": 1.2304483837330553e-06, | |
| "loss": 1.4016, | |
| "step": 9060 | |
| }, | |
| { | |
| "epoch": 2.8401626524867063, | |
| "grad_norm": 4.767721652984619, | |
| "learning_rate": 1.1841038118410383e-06, | |
| "loss": 1.3985, | |
| "step": 9080 | |
| }, | |
| { | |
| "epoch": 2.846418517360025, | |
| "grad_norm": 3.784308671951294, | |
| "learning_rate": 1.137759239949021e-06, | |
| "loss": 1.3853, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 2.852674382233344, | |
| "grad_norm": 4.14461088180542, | |
| "learning_rate": 1.091414668057004e-06, | |
| "loss": 1.3303, | |
| "step": 9120 | |
| }, | |
| { | |
| "epoch": 2.8589302471066627, | |
| "grad_norm": 3.3787248134613037, | |
| "learning_rate": 1.0450700961649869e-06, | |
| "loss": 1.4045, | |
| "step": 9140 | |
| }, | |
| { | |
| "epoch": 2.8651861119799813, | |
| "grad_norm": 3.9108362197875977, | |
| "learning_rate": 9.987255242729696e-07, | |
| "loss": 1.3806, | |
| "step": 9160 | |
| }, | |
| { | |
| "epoch": 2.8714419768533, | |
| "grad_norm": 3.8654673099517822, | |
| "learning_rate": 9.523809523809525e-07, | |
| "loss": 1.367, | |
| "step": 9180 | |
| }, | |
| { | |
| "epoch": 2.8776978417266186, | |
| "grad_norm": 3.9487152099609375, | |
| "learning_rate": 9.060363804889354e-07, | |
| "loss": 1.4199, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 2.8839537065999377, | |
| "grad_norm": 3.804671049118042, | |
| "learning_rate": 8.596918085969181e-07, | |
| "loss": 1.4252, | |
| "step": 9220 | |
| }, | |
| { | |
| "epoch": 2.8902095714732563, | |
| "grad_norm": 3.471071243286133, | |
| "learning_rate": 8.13347236704901e-07, | |
| "loss": 1.4102, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 2.896465436346575, | |
| "grad_norm": 3.498152256011963, | |
| "learning_rate": 7.670026648128838e-07, | |
| "loss": 1.3805, | |
| "step": 9260 | |
| }, | |
| { | |
| "epoch": 2.9027213012198936, | |
| "grad_norm": 4.2036871910095215, | |
| "learning_rate": 7.206580929208668e-07, | |
| "loss": 1.3746, | |
| "step": 9280 | |
| }, | |
| { | |
| "epoch": 2.908977166093212, | |
| "grad_norm": 3.783219337463379, | |
| "learning_rate": 6.743135210288496e-07, | |
| "loss": 1.3919, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 2.9152330309665313, | |
| "grad_norm": 3.6748569011688232, | |
| "learning_rate": 6.279689491368323e-07, | |
| "loss": 1.3889, | |
| "step": 9320 | |
| }, | |
| { | |
| "epoch": 2.92148889583985, | |
| "grad_norm": 3.686070680618286, | |
| "learning_rate": 5.816243772448153e-07, | |
| "loss": 1.4083, | |
| "step": 9340 | |
| }, | |
| { | |
| "epoch": 2.9277447607131686, | |
| "grad_norm": 3.854691505432129, | |
| "learning_rate": 5.352798053527981e-07, | |
| "loss": 1.3814, | |
| "step": 9360 | |
| }, | |
| { | |
| "epoch": 2.934000625586487, | |
| "grad_norm": 4.229208946228027, | |
| "learning_rate": 4.88935233460781e-07, | |
| "loss": 1.3687, | |
| "step": 9380 | |
| }, | |
| { | |
| "epoch": 2.940256490459806, | |
| "grad_norm": 4.185445308685303, | |
| "learning_rate": 4.425906615687638e-07, | |
| "loss": 1.4074, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 2.946512355333125, | |
| "grad_norm": 4.26198673248291, | |
| "learning_rate": 3.9624608967674663e-07, | |
| "loss": 1.3865, | |
| "step": 9420 | |
| }, | |
| { | |
| "epoch": 2.9527682202064436, | |
| "grad_norm": 4.7634735107421875, | |
| "learning_rate": 3.4990151778472947e-07, | |
| "loss": 1.3857, | |
| "step": 9440 | |
| }, | |
| { | |
| "epoch": 2.9590240850797622, | |
| "grad_norm": 3.629460573196411, | |
| "learning_rate": 3.0355694589271236e-07, | |
| "loss": 1.4056, | |
| "step": 9460 | |
| }, | |
| { | |
| "epoch": 2.965279949953081, | |
| "grad_norm": 3.930769443511963, | |
| "learning_rate": 2.572123740006952e-07, | |
| "loss": 1.3554, | |
| "step": 9480 | |
| }, | |
| { | |
| "epoch": 2.9715358148263995, | |
| "grad_norm": 3.6116929054260254, | |
| "learning_rate": 2.1086780210867804e-07, | |
| "loss": 1.3699, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.9777916796997186, | |
| "grad_norm": 4.0823073387146, | |
| "learning_rate": 1.6452323021666087e-07, | |
| "loss": 1.4166, | |
| "step": 9520 | |
| }, | |
| { | |
| "epoch": 2.9840475445730372, | |
| "grad_norm": 3.9738993644714355, | |
| "learning_rate": 1.1817865832464373e-07, | |
| "loss": 1.3666, | |
| "step": 9540 | |
| }, | |
| { | |
| "epoch": 2.990303409446356, | |
| "grad_norm": 4.159948348999023, | |
| "learning_rate": 7.183408643262659e-08, | |
| "loss": 1.3622, | |
| "step": 9560 | |
| }, | |
| { | |
| "epoch": 2.996559274319675, | |
| "grad_norm": 3.861499309539795, | |
| "learning_rate": 2.5489514540609434e-08, | |
| "loss": 1.3759, | |
| "step": 9580 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 9591, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.406598441985901e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |