| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 300, | |
| "global_step": 1098, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00273224043715847, | |
| "grad_norm": 84.5, | |
| "learning_rate": 1e-06, | |
| "loss": 2.8857, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01366120218579235, | |
| "grad_norm": 40.75, | |
| "learning_rate": 1.9999630266778663e-06, | |
| "loss": 2.6209, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0273224043715847, | |
| "grad_norm": 16.875, | |
| "learning_rate": 1.999737088499184e-06, | |
| "loss": 2.1772, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.040983606557377046, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.999305799228933e-06, | |
| "loss": 1.96, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0546448087431694, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 1.998669247456129e-06, | |
| "loss": 1.8595, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06830601092896176, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 1.997827563931747e-06, | |
| "loss": 1.7983, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08196721311475409, | |
| "grad_norm": 3.90625, | |
| "learning_rate": 1.9967809215418726e-06, | |
| "loss": 1.7237, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09562841530054644, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 1.9955295352721854e-06, | |
| "loss": 1.6815, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1092896174863388, | |
| "grad_norm": 2.625, | |
| "learning_rate": 1.9940736621637997e-06, | |
| "loss": 1.6637, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12295081967213115, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 1.992413601260471e-06, | |
| "loss": 1.6501, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1366120218579235, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.990549693547166e-06, | |
| "loss": 1.6316, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15027322404371585, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.988482321880025e-06, | |
| "loss": 1.6138, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.16393442622950818, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.9862119109077224e-06, | |
| "loss": 1.6096, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.17759562841530055, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.983738926984239e-06, | |
| "loss": 1.5883, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.1912568306010929, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.9810638780730727e-06, | |
| "loss": 1.5932, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.20491803278688525, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9781873136428984e-06, | |
| "loss": 1.5829, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2185792349726776, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.975109824554707e-06, | |
| "loss": 1.5835, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23224043715846995, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.971832042940436e-06, | |
| "loss": 1.5746, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2459016393442623, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.968354642073129e-06, | |
| "loss": 1.5765, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.25956284153005466, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.964678336228642e-06, | |
| "loss": 1.5685, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.273224043715847, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.960803880538925e-06, | |
| "loss": 1.564, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.28688524590163933, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.9567320708369176e-06, | |
| "loss": 1.5642, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3005464480874317, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9524637434930776e-06, | |
| "loss": 1.5723, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.31420765027322406, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.9479997752435884e-06, | |
| "loss": 1.5712, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.32786885245901637, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.943341083010272e-06, | |
| "loss": 1.5609, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.34153005464480873, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.9384886237122493e-06, | |
| "loss": 1.5596, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3551912568306011, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.9334433940693826e-06, | |
| "loss": 1.5582, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.36885245901639346, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.928206430397546e-06, | |
| "loss": 1.5554, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3825136612021858, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.9227788083957586e-06, | |
| "loss": 1.5653, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.39617486338797814, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9171616429252344e-06, | |
| "loss": 1.5515, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4098360655737705, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.9113560877803796e-06, | |
| "loss": 1.5509, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.42349726775956287, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.9053633354517988e-06, | |
| "loss": 1.5523, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.4371584699453552, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.8991846168813544e-06, | |
| "loss": 1.5588, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.45081967213114754, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.89282120120932e-06, | |
| "loss": 1.5507, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4644808743169399, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.8862743955136963e-06, | |
| "loss": 1.5467, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4781420765027322, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.8795455445417286e-06, | |
| "loss": 1.5491, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4918032786885246, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.8726360304336893e-06, | |
| "loss": 1.5416, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.505464480874317, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.8655472724389796e-06, | |
| "loss": 1.5483, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.5191256830601093, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.858280726624609e-06, | |
| "loss": 1.5543, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5327868852459017, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.8508378855761095e-06, | |
| "loss": 1.5393, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.546448087431694, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.8432202780909538e-06, | |
| "loss": 1.5435, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5601092896174863, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.83542946886453e-06, | |
| "loss": 1.549, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5737704918032787, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.8274670581687478e-06, | |
| "loss": 1.5422, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.587431693989071, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.819334681523331e-06, | |
| "loss": 1.5413, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.6010928961748634, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.811034009359877e-06, | |
| "loss": 1.5403, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6147540983606558, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.8025667466787391e-06, | |
| "loss": 1.5293, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.6284153005464481, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.7939346326988125e-06, | |
| "loss": 1.5472, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6420765027322405, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.7851394405002884e-06, | |
| "loss": 1.5436, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.6557377049180327, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.7761829766604554e-06, | |
| "loss": 1.5341, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6693989071038251, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.7670670808826191e-06, | |
| "loss": 1.5287, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6830601092896175, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.7577936256182167e-06, | |
| "loss": 1.5389, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6967213114754098, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.7483645156822069e-06, | |
| "loss": 1.5431, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.7103825136612022, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.7387816878618117e-06, | |
| "loss": 1.5424, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7240437158469946, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.729047110518689e-06, | |
| "loss": 1.5294, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.7377049180327869, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.7191627831846222e-06, | |
| "loss": 1.5256, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7513661202185792, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.7091307361508055e-06, | |
| "loss": 1.5359, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7650273224043715, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.6989530300508123e-06, | |
| "loss": 1.5318, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7786885245901639, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.6886317554373302e-06, | |
| "loss": 1.5344, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7923497267759563, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.6781690323527509e-06, | |
| "loss": 1.5272, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.8060109289617486, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.6675670098937032e-06, | |
| "loss": 1.5293, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.819672131147541, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.6568278657696162e-06, | |
| "loss": 1.5401, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.819672131147541, | |
| "eval_loss": 1.533558964729309, | |
| "eval_runtime": 19.6321, | |
| "eval_samples_per_second": 36.063, | |
| "eval_steps_per_second": 1.172, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.6459538058554087e-06, | |
| "loss": 1.5176, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.8469945355191257, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.6349470637383888e-06, | |
| "loss": 1.5312, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.860655737704918, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.6238099002594669e-06, | |
| "loss": 1.5391, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8743169398907104, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.6125446030487642e-06, | |
| "loss": 1.5248, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8879781420765027, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.6011534860557236e-06, | |
| "loss": 1.5381, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.9016393442622951, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.5896388890738127e-06, | |
| "loss": 1.5216, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.9153005464480874, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.578003177259917e-06, | |
| "loss": 1.524, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.9289617486338798, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.566248740648527e-06, | |
| "loss": 1.5255, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9426229508196722, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.5543779936608106e-06, | |
| "loss": 1.5213, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.9562841530054644, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.5423933746086793e-06, | |
| "loss": 1.517, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9699453551912568, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.5302973451939472e-06, | |
| "loss": 1.524, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9836065573770492, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.5180923900026845e-06, | |
| "loss": 1.5224, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9972677595628415, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.5057810159948714e-06, | |
| "loss": 1.5148, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.010928961748634, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.493365751989454e-06, | |
| "loss": 1.5175, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.0245901639344261, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.4808491481449143e-06, | |
| "loss": 1.5084, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.0382513661202186, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.4682337754354533e-06, | |
| "loss": 1.5141, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0519125683060109, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.4555222251228995e-06, | |
| "loss": 1.52, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.0655737704918034, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.442717108224452e-06, | |
| "loss": 1.5167, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0792349726775956, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.4298210549763628e-06, | |
| "loss": 1.5087, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.092896174863388, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.4168367142936734e-06, | |
| "loss": 1.5049, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.1065573770491803, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.4037667532261142e-06, | |
| "loss": 1.5102, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.1202185792349726, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.3906138564102792e-06, | |
| "loss": 1.5156, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.133879781420765, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.3773807255181874e-06, | |
| "loss": 1.5162, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.1475409836065573, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.3640700787023462e-06, | |
| "loss": 1.5099, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.1612021857923498, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.3506846500374283e-06, | |
| "loss": 1.5159, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.174863387978142, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.337227188958679e-06, | |
| "loss": 1.5145, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.1885245901639343, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 1.3237004596971687e-06, | |
| "loss": 1.5066, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.2021857923497268, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.3101072407120055e-06, | |
| "loss": 1.5144, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.215846994535519, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.2964503241196256e-06, | |
| "loss": 1.5124, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.2295081967213115, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.2827325151202782e-06, | |
| "loss": 1.5052, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.2431693989071038, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.2689566314218228e-06, | |
| "loss": 1.5103, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.2568306010928962, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.255125502660958e-06, | |
| "loss": 1.5078, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.2704918032786885, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.2412419698220001e-06, | |
| "loss": 1.5134, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.2841530054644807, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.2273088846533302e-06, | |
| "loss": 1.5064, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.2978142076502732, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.2133291090816295e-06, | |
| "loss": 1.5124, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.3114754098360657, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.1993055146240272e-06, | |
| "loss": 1.5144, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.325136612021858, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.185240981798273e-06, | |
| "loss": 1.5221, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.3387978142076502, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.171138399531068e-06, | |
| "loss": 1.5079, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.3524590163934427, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.1570006645646614e-06, | |
| "loss": 1.5145, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.366120218579235, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.1428306808618454e-06, | |
| "loss": 1.5043, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.3797814207650272, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.1286313590094686e-06, | |
| "loss": 1.5171, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.3934426229508197, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.1144056156205831e-06, | |
| "loss": 1.5114, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.4071038251366121, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.100156372735361e-06, | |
| "loss": 1.5032, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.4207650273224044, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.0858865572208891e-06, | |
| "loss": 1.5151, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.4344262295081966, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.071599100169978e-06, | |
| "loss": 1.5015, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.4480874316939891, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.0572969362990997e-06, | |
| "loss": 1.5075, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.4617486338797814, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 1.042983003345582e-06, | |
| "loss": 1.5117, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.4754098360655736, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.0286602414641815e-06, | |
| "loss": 1.5036, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.489071038251366, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.0143315926231624e-06, | |
| "loss": 1.527, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.5027322404371586, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1e-06, | |
| "loss": 1.5065, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.5163934426229508, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 9.856684073768378e-07, | |
| "loss": 1.4989, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.530054644808743, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 9.713397585358188e-07, | |
| "loss": 1.5066, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.5437158469945356, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 9.57016996654418e-07, | |
| "loss": 1.5024, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.5573770491803278, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 9.427030637009002e-07, | |
| "loss": 1.5095, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.57103825136612, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 9.28400899830022e-07, | |
| "loss": 1.5144, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.5846994535519126, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 9.141134427791109e-07, | |
| "loss": 1.5191, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.598360655737705, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 8.998436272646393e-07, | |
| "loss": 1.5177, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.6120218579234973, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 8.85594384379417e-07, | |
| "loss": 1.5098, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.6256830601092895, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 8.713686409905313e-07, | |
| "loss": 1.503, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.639344262295082, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 8.571693191381544e-07, | |
| "loss": 1.5091, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.639344262295082, | |
| "eval_loss": 1.5182926654815674, | |
| "eval_runtime": 19.6502, | |
| "eval_samples_per_second": 36.03, | |
| "eval_steps_per_second": 1.17, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.6530054644808743, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 8.429993354353388e-07, | |
| "loss": 1.5003, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 1.625, | |
| "learning_rate": 8.288616004689319e-07, | |
| "loss": 1.5068, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.680327868852459, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 8.147590182017269e-07, | |
| "loss": 1.5067, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.6939890710382515, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 8.006944853759732e-07, | |
| "loss": 1.5069, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.7076502732240437, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 7.866708909183702e-07, | |
| "loss": 1.4937, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.721311475409836, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 7.726911153466697e-07, | |
| "loss": 1.5089, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.7349726775956285, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 7.587580301779999e-07, | |
| "loss": 1.511, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.748633879781421, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 7.448744973390422e-07, | |
| "loss": 1.5025, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.762295081967213, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 7.310433685781777e-07, | |
| "loss": 1.5007, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.7759562841530054, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 7.172674848797217e-07, | |
| "loss": 1.4922, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.789617486338798, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 7.035496758803743e-07, | |
| "loss": 1.505, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.8032786885245902, | |
| "grad_norm": 1.625, | |
| "learning_rate": 6.898927592879944e-07, | |
| "loss": 1.5111, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.8169398907103824, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 6.762995403028314e-07, | |
| "loss": 1.5145, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.830601092896175, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 6.627728110413213e-07, | |
| "loss": 1.5097, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.8442622950819674, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 6.493153499625719e-07, | |
| "loss": 1.5087, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.8579234972677594, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 6.359299212976534e-07, | |
| "loss": 1.5044, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.8715846994535519, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 6.226192744818124e-07, | |
| "loss": 1.5105, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.8852459016393444, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 6.093861435897207e-07, | |
| "loss": 1.5109, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.8989071038251366, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 5.962332467738857e-07, | |
| "loss": 1.5103, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.9125683060109289, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 5.83163285706327e-07, | |
| "loss": 1.5023, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.9262295081967213, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 5.701789450236376e-07, | |
| "loss": 1.5084, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.9398907103825138, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 5.57282891775548e-07, | |
| "loss": 1.5063, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.9535519125683058, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 5.444777748771006e-07, | |
| "loss": 1.5074, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.9672131147540983, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 5.317662245645469e-07, | |
| "loss": 1.5037, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.9808743169398908, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 5.191508518550855e-07, | |
| "loss": 1.5085, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.994535519125683, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 5.066342480105459e-07, | |
| "loss": 1.5082, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.0081967213114753, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 4.942189840051287e-07, | |
| "loss": 1.5054, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.021857923497268, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 4.819076099973152e-07, | |
| "loss": 1.4999, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.0355191256830603, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 4.697026548060528e-07, | |
| "loss": 1.5094, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.0491803278688523, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 4.5760662539132077e-07, | |
| "loss": 1.5065, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.0628415300546448, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 4.4562200633918943e-07, | |
| "loss": 1.5013, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 2.0765027322404372, | |
| "grad_norm": 1.625, | |
| "learning_rate": 4.337512593514728e-07, | |
| "loss": 1.5041, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.0901639344262297, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 4.2199682274008255e-07, | |
| "loss": 1.5083, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 2.1038251366120218, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 4.103611109261872e-07, | |
| "loss": 1.4982, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.1174863387978142, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 3.9884651394427625e-07, | |
| "loss": 1.5051, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.1311475409836067, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 3.8745539695123577e-07, | |
| "loss": 1.4924, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.1448087431693987, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 3.761900997405332e-07, | |
| "loss": 1.5113, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 2.158469945355191, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 3.6505293626161127e-07, | |
| "loss": 1.5058, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.1721311475409837, | |
| "grad_norm": 1.625, | |
| "learning_rate": 3.5404619414459147e-07, | |
| "loss": 1.5019, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 2.185792349726776, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 3.4317213423038384e-07, | |
| "loss": 1.4948, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.199453551912568, | |
| "grad_norm": 1.625, | |
| "learning_rate": 3.32432990106297e-07, | |
| "loss": 1.5092, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 2.2131147540983607, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 3.2183096764724914e-07, | |
| "loss": 1.5048, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.226775956284153, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 3.1136824456267006e-07, | |
| "loss": 1.5005, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 2.240437158469945, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 3.01046969949188e-07, | |
| "loss": 1.5097, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.2540983606557377, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 2.908692638491945e-07, | |
| "loss": 1.5054, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.26775956284153, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 2.80837216815378e-07, | |
| "loss": 1.5131, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.281420765027322, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.7095288948131114e-07, | |
| "loss": 1.4999, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.2950819672131146, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 2.6121831213818826e-07, | |
| "loss": 1.4989, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.308743169398907, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.51635484317793e-07, | |
| "loss": 1.495, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.3224043715846996, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.4220637438178313e-07, | |
| "loss": 1.5125, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.3360655737704916, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.3293291911738078e-07, | |
| "loss": 1.5092, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.349726775956284, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 2.2381702333954433e-07, | |
| "loss": 1.4965, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.3633879781420766, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 2.148605594997115e-07, | |
| "loss": 1.497, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 2.3770491803278686, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 2.0606536730118763e-07, | |
| "loss": 1.5081, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.390710382513661, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.9743325332126105e-07, | |
| "loss": 1.5091, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.4043715846994536, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.8896599064012298e-07, | |
| "loss": 1.5045, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.418032786885246, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.8066531847666888e-07, | |
| "loss": 1.5008, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 2.431693989071038, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.7253294183125222e-07, | |
| "loss": 1.511, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.4453551912568305, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.645705311354697e-07, | |
| "loss": 1.4998, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 2.459016393442623, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.5677972190904621e-07, | |
| "loss": 1.5021, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.459016393442623, | |
| "eval_loss": 1.5163270235061646, | |
| "eval_runtime": 19.6134, | |
| "eval_samples_per_second": 36.098, | |
| "eval_steps_per_second": 1.173, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.4726775956284155, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.4916211442389048e-07, | |
| "loss": 1.502, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 2.4863387978142075, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.4171927337539104e-07, | |
| "loss": 1.5012, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.344527275610202e-07, | |
| "loss": 1.5019, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 2.5136612021857925, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.273639695663108e-07, | |
| "loss": 1.5085, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.527322404371585, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.204544554582716e-07, | |
| "loss": 1.4973, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.540983606557377, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.1372560448630375e-07, | |
| "loss": 1.5037, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.5546448087431695, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.0717879879068004e-07, | |
| "loss": 1.5005, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.5683060109289615, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.0081538311864568e-07, | |
| "loss": 1.5024, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.581967213114754, | |
| "grad_norm": 1.625, | |
| "learning_rate": 9.463666454820118e-08, | |
| "loss": 1.4988, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.5956284153005464, | |
| "grad_norm": 1.625, | |
| "learning_rate": 8.864391221962064e-08, | |
| "loss": 1.5053, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.609289617486339, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 8.28383570747655e-08, | |
| "loss": 1.5044, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.6229508196721314, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 7.722119160424112e-08, | |
| "loss": 1.4995, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.6366120218579234, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 7.179356960245409e-08, | |
| "loss": 1.5122, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.650273224043716, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 6.655660593061718e-08, | |
| "loss": 1.5054, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.663934426229508, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 6.151137628775049e-08, | |
| "loss": 1.5108, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.6775956284153004, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 5.665891698972769e-08, | |
| "loss": 1.5003, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.691256830601093, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 5.200022475641153e-08, | |
| "loss": 1.5015, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.7049180327868854, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 4.75362565069225e-08, | |
| "loss": 1.5002, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.718579234972678, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 4.326792916308242e-08, | |
| "loss": 1.5029, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.73224043715847, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 3.919611946107493e-08, | |
| "loss": 1.5068, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.7459016393442623, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 3.532166377135814e-08, | |
| "loss": 1.4961, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.7595628415300544, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 3.164535792687095e-08, | |
| "loss": 1.5, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.773224043715847, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 2.8167957059564095e-08, | |
| "loss": 1.5035, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.7868852459016393, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 2.4890175445293147e-08, | |
| "loss": 1.5007, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.800546448087432, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 2.1812686357101428e-08, | |
| "loss": 1.505, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.8142076502732243, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.8936121926927507e-08, | |
| "loss": 1.5066, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.8278688524590163, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.6261073015761072e-08, | |
| "loss": 1.502, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.841530054644809, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.3788089092277688e-08, | |
| "loss": 1.5016, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.855191256830601, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.1517678119975061e-08, | |
| "loss": 1.5024, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.8688524590163933, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 9.450306452834178e-09, | |
| "loss": 1.5, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.8825136612021858, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.586398739528932e-09, | |
| "loss": 1.4989, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.8961748633879782, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 5.926337836199891e-09, | |
| "loss": 1.507, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.9098360655737707, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 4.470464727814538e-09, | |
| "loss": 1.5119, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.9234972677595628, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 3.219078458127078e-09, | |
| "loss": 1.5072, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.9371584699453552, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 2.172436068252792e-09, | |
| "loss": 1.5074, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.9508196721311473, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.330752543871161e-09, | |
| "loss": 1.4947, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.9644808743169397, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 6.942007710665221e-10, | |
| "loss": 1.5025, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.978142076502732, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 2.6291150081603207e-10, | |
| "loss": 1.5014, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.9918032786885247, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 3.697332213348225e-11, | |
| "loss": 1.4964, | |
| "step": 1095 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1098, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.187967826039144e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |