| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.974380871050384, | |
| "eval_steps": 500, | |
| "global_step": 2920, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0034158838599487617, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 6.849315068493151e-07, | |
| "loss": 3.0658, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017079419299743808, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 3.4246575342465754e-06, | |
| "loss": 3.0727, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.034158838599487616, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 6.849315068493151e-06, | |
| "loss": 3.0381, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05123825789923143, | |
| "grad_norm": 3.375, | |
| "learning_rate": 1.0273972602739726e-05, | |
| "loss": 2.9796, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06831767719897523, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 1.3698630136986302e-05, | |
| "loss": 2.8478, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08539709649871904, | |
| "grad_norm": 2.375, | |
| "learning_rate": 1.7123287671232875e-05, | |
| "loss": 2.7142, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10247651579846286, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 2.0547945205479453e-05, | |
| "loss": 2.5273, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11955593509820667, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 2.3972602739726026e-05, | |
| "loss": 2.3905, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.13663535439795046, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 2.7397260273972603e-05, | |
| "loss": 2.2615, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1537147736976943, | |
| "grad_norm": 21.875, | |
| "learning_rate": 3.082191780821918e-05, | |
| "loss": 2.1359, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1707941929974381, | |
| "grad_norm": 15.6875, | |
| "learning_rate": 3.424657534246575e-05, | |
| "loss": 2.0159, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18787361229718189, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 3.767123287671233e-05, | |
| "loss": 1.8994, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.2049530315969257, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 4.1095890410958905e-05, | |
| "loss": 1.7873, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2220324508966695, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 4.452054794520548e-05, | |
| "loss": 1.6691, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.23911187019641333, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 4.794520547945205e-05, | |
| "loss": 1.5889, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2561912894961571, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 5.136986301369864e-05, | |
| "loss": 1.5329, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.27327070879590093, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 5.479452054794521e-05, | |
| "loss": 1.4749, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.29035012809564475, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 5.821917808219178e-05, | |
| "loss": 1.438, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3074295473953886, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 6.164383561643835e-05, | |
| "loss": 1.395, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.32450896669513235, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 6.506849315068494e-05, | |
| "loss": 1.3653, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3415883859948762, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 6.84931506849315e-05, | |
| "loss": 1.3329, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.35866780529462, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 7.191780821917809e-05, | |
| "loss": 1.3221, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.37574722459436377, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 7.534246575342466e-05, | |
| "loss": 1.3048, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3928266438941076, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 7.876712328767124e-05, | |
| "loss": 1.2884, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4099060631938514, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 8.219178082191781e-05, | |
| "loss": 1.2687, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4269854824935952, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 8.561643835616438e-05, | |
| "loss": 1.2502, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.444064901793339, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 8.904109589041096e-05, | |
| "loss": 1.2416, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.46114432109308284, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.246575342465755e-05, | |
| "loss": 1.2345, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.47822374039282667, | |
| "grad_norm": 0.7890625, | |
| "learning_rate": 9.58904109589041e-05, | |
| "loss": 1.2381, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.49530315969257044, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 9.931506849315069e-05, | |
| "loss": 1.2236, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5123825789923142, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00010273972602739728, | |
| "loss": 1.2102, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5294619982920581, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 0.00010616438356164384, | |
| "loss": 1.203, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5465414175918019, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00010958904109589041, | |
| "loss": 1.2011, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5636208368915457, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.000113013698630137, | |
| "loss": 1.193, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5807002561912895, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.00011643835616438356, | |
| "loss": 1.1933, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5977796754910333, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00011986301369863014, | |
| "loss": 1.1774, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6148590947907772, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 0.0001232876712328767, | |
| "loss": 1.177, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6319385140905209, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 0.0001267123287671233, | |
| "loss": 1.1705, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6490179333902647, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00013013698630136988, | |
| "loss": 1.1612, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6660973526900086, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00013356164383561644, | |
| "loss": 1.167, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6831767719897524, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.000136986301369863, | |
| "loss": 1.1616, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7002561912894961, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 0.0001404109589041096, | |
| "loss": 1.1553, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.71733561058924, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00014383561643835618, | |
| "loss": 1.1475, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7344150298889838, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00014726027397260274, | |
| "loss": 1.1482, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7514944491887275, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 0.00015068493150684933, | |
| "loss": 1.1427, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7685738684884714, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 0.00015410958904109589, | |
| "loss": 1.1441, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7856532877882152, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.00015753424657534247, | |
| "loss": 1.1336, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.802732707087959, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 0.00016095890410958906, | |
| "loss": 1.1315, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8198121263877028, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 0.00016438356164383562, | |
| "loss": 1.1316, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8368915456874466, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 0.0001678082191780822, | |
| "loss": 1.1381, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8539709649871904, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 0.00017123287671232877, | |
| "loss": 1.1376, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8710503842869343, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 0.00017465753424657536, | |
| "loss": 1.1309, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.888129803586678, | |
| "grad_norm": 1.375, | |
| "learning_rate": 0.00017808219178082192, | |
| "loss": 1.1282, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.9052092228864219, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 0.0001815068493150685, | |
| "loss": 1.1414, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.9222886421861657, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 0.0001849315068493151, | |
| "loss": 1.1403, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9393680614859095, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00018835616438356165, | |
| "loss": 1.1352, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9564474807856533, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.0001917808219178082, | |
| "loss": 1.1201, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9735269000853971, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001952054794520548, | |
| "loss": 1.1145, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9906063193851409, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00019863013698630139, | |
| "loss": 1.1174, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9974380871050385, | |
| "eval_loss": 2.4481546878814697, | |
| "eval_runtime": 0.5643, | |
| "eval_samples_per_second": 17.72, | |
| "eval_steps_per_second": 1.772, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.0076857386848848, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00019999935692582106, | |
| "loss": 1.1057, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.0247651579846284, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00019999542705801296, | |
| "loss": 1.0972, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0418445772843723, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00019998792472605885, | |
| "loss": 1.1012, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.0589239965841162, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00019997685019798912, | |
| "loss": 1.0859, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0760034158838598, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.00019996220386945537, | |
| "loss": 1.0973, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.0930828351836037, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00019994398626371643, | |
| "loss": 1.0961, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.1101622544833476, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.0001999221980316194, | |
| "loss": 1.0901, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.1272416737830913, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00019989683995157677, | |
| "loss": 1.0761, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1443210930828351, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.0001998679129295382, | |
| "loss": 1.082, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.161400512382579, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 0.0001998354179989585, | |
| "loss": 1.0788, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1784799316823227, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00019979935632076048, | |
| "loss": 1.0745, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.1955593509820666, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00019975972918329356, | |
| "loss": 1.0775, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.2126387702818104, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0001997165380022878, | |
| "loss": 1.0761, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.2297181895815543, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00019966978432080316, | |
| "loss": 1.0789, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.246797608881298, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00019961946980917456, | |
| "loss": 1.0762, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.2638770281810419, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.00019956559626495212, | |
| "loss": 1.0748, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.2809564474807855, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00019950816561283685, | |
| "loss": 1.0756, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.2980358667805294, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00019944717990461207, | |
| "loss": 1.0694, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.3151152860802733, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00019938264131907, | |
| "loss": 1.0654, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.332194705380017, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00019931455216193382, | |
| "loss": 1.0645, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.3492741246797608, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 0.00019924291486577559, | |
| "loss": 1.0613, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.3663535439795047, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.000199167731989929, | |
| "loss": 1.0689, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3834329632792486, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00019908900622039822, | |
| "loss": 1.065, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.4005123825789922, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00019900674036976173, | |
| "loss": 1.0668, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.4175918018787361, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.0001989209373770719, | |
| "loss": 1.0628, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.43467122117848, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00019883160030775016, | |
| "loss": 1.0617, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4517506404782237, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00019873873235347719, | |
| "loss": 1.0598, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.4688300597779675, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00019864233683207906, | |
| "loss": 1.0536, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4859094790777114, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0001985424171874087, | |
| "loss": 1.0565, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.5029888983774553, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00019843897698922284, | |
| "loss": 1.0613, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.520068317677199, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.0001983320199330545, | |
| "loss": 1.0517, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.5371477369769426, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00019822154984008088, | |
| "loss": 1.0589, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.5542271562766867, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00019810757065698688, | |
| "loss": 1.0517, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.5713065755764304, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.0001979900864558242, | |
| "loss": 1.0547, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.588385994876174, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.0001978691014338658, | |
| "loss": 1.0537, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.6054654141759181, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00019774461991345577, | |
| "loss": 1.0459, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.6225448334756618, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 0.0001976166463418552, | |
| "loss": 1.0477, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.6396242527754057, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 0.00019748518529108316, | |
| "loss": 1.0472, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.6567036720751496, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.0001973502414577533, | |
| "loss": 1.0521, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.6737830913748932, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00019721181966290613, | |
| "loss": 1.0394, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.690862510674637, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00019706992485183684, | |
| "loss": 1.0328, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.707941929974381, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00019692456209391846, | |
| "loss": 1.0382, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.7250213492741246, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00019677573658242087, | |
| "loss": 1.0418, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.7421007685738685, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.0001966234536343253, | |
| "loss": 1.0416, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.7591801878736124, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0001964677186901342, | |
| "loss": 1.0399, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.776259607173356, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00019630853731367713, | |
| "loss": 1.0404, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.7933390264731, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00019614591519191165, | |
| "loss": 1.0349, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.8104184457728438, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00019597985813472052, | |
| "loss": 1.0303, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.8274978650725875, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00019581037207470382, | |
| "loss": 1.0318, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.8445772843723314, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.0001956374630669672, | |
| "loss": 1.0386, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.8616567036720753, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00019546113728890541, | |
| "loss": 1.0252, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.878736122971819, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00019528140103998177, | |
| "loss": 1.0329, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.8958155422715628, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00019509826074150298, | |
| "loss": 1.0385, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.9128949615713067, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00019491172293638968, | |
| "loss": 1.0322, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.9299743808710503, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00019472179428894288, | |
| "loss": 1.0296, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.9470538001707942, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0001945284815846057, | |
| "loss": 1.0434, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.964133219470538, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00019433179172972102, | |
| "loss": 1.0228, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.9812126387702818, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00019413173175128473, | |
| "loss": 1.0274, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.9982920580700256, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.00019392830879669463, | |
| "loss": 1.0252, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.9982920580700256, | |
| "eval_loss": 2.451388120651245, | |
| "eval_runtime": 0.5458, | |
| "eval_samples_per_second": 18.323, | |
| "eval_steps_per_second": 1.832, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.0153714773697695, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00019372153013349523, | |
| "loss": 1.0051, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.032450896669513, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00019351140314911795, | |
| "loss": 1.0105, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.049530315969257, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00019329793535061723, | |
| "loss": 1.0135, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.066609735269001, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00019308113436440242, | |
| "loss": 1.0062, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.0836891545687446, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 0.0001928610079359652, | |
| "loss": 1.0019, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.1007685738684883, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00019263756392960294, | |
| "loss": 1.0048, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.1178479931682324, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00019241081032813772, | |
| "loss": 1.0094, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.134927412467976, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00019218075523263104, | |
| "loss": 1.0014, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.1520068317677197, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00019194740686209464, | |
| "loss": 1.0085, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.1690862510674638, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.0001917107735531966, | |
| "loss": 1.0014, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.1861656703672074, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 0.0001914708637599636, | |
| "loss": 1.0056, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.203245089666951, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.00019122768605347892, | |
| "loss": 0.998, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.220324508966695, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00019098124912157632, | |
| "loss": 1.0007, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.237403928266439, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 0.00019073156176852935, | |
| "loss": 1.0046, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.2544833475661825, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00019047863291473717, | |
| "loss": 1.0084, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.2715627668659266, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00019022247159640557, | |
| "loss": 1.0006, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.2886421861656703, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.00018996308696522433, | |
| "loss": 1.0057, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.305721605465414, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00018970048828804016, | |
| "loss": 1.0019, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.322801024765158, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.0001894346849465257, | |
| "loss": 1.0054, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.3398804440649017, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.0001891656864368442, | |
| "loss": 1.0021, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.3569598633646454, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00018889350236931055, | |
| "loss": 0.9956, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.3740392826643895, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00018861814246804755, | |
| "loss": 1.0063, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.391118701964133, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00018833961657063885, | |
| "loss": 1.0013, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.408198121263877, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00018805793462777734, | |
| "loss": 0.9951, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.425277540563621, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.0001877731067029096, | |
| "loss": 1.0019, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.4423569598633645, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00018748514297187648, | |
| "loss": 0.995, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.4594363791631086, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00018719405372254948, | |
| "loss": 1.002, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.4765157984628523, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00018689984935446317, | |
| "loss": 0.9942, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.493595217762596, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 1.0012, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.5106746370623396, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00018630213741623383, | |
| "loss": 1.002, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.5277540563620837, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00018599865120011192, | |
| "loss": 0.9975, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.5448334756618274, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00018569209257251026, | |
| "loss": 0.9996, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.561912894961571, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00018538247248562674, | |
| "loss": 1.0001, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.578992314261315, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00018506980200103375, | |
| "loss": 0.9954, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 2.596071733561059, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00018475409228928312, | |
| "loss": 0.9945, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.6131511528608025, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00018443535462950688, | |
| "loss": 0.9918, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 2.6302305721605466, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.0001841136004090144, | |
| "loss": 1.0007, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.6473099914602902, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00018378884112288542, | |
| "loss": 1.0026, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.664389410760034, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00018346108837355972, | |
| "loss": 0.995, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.681468830059778, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.0001831303538704221, | |
| "loss": 0.9916, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 2.6985482493595216, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00018279664942938447, | |
| "loss": 0.9902, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.7156276686592657, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 0.00018245998697246352, | |
| "loss": 1.0003, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 2.7327070879590094, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00018212037852735486, | |
| "loss": 0.9933, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.749786507258753, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00018177783622700327, | |
| "loss": 0.9934, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 2.766865926558497, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.0001814323723091692, | |
| "loss": 0.9887, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.783945345858241, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00018108399911599167, | |
| "loss": 0.995, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 2.8010247651579845, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00018073272909354727, | |
| "loss": 0.9897, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.8181041844577286, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00018037857479140547, | |
| "loss": 0.9923, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.8351836037574722, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00018002154886218033, | |
| "loss": 0.9877, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.852263023057216, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00017966166406107846, | |
| "loss": 0.9936, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.86934244235696, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00017929893324544332, | |
| "loss": 0.9931, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.8864218616567037, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00017893336937429581, | |
| "loss": 0.992, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.9035012809564473, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.00017856498550787144, | |
| "loss": 0.9896, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.9205807002561914, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.0001781937948071536, | |
| "loss": 0.9899, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.937660119555935, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00017781981053340337, | |
| "loss": 0.9869, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.9547395388556787, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00017744304604768588, | |
| "loss": 0.9865, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 2.971818958155423, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00017706351481039284, | |
| "loss": 0.9885, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.9888983774551665, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00017668123038076163, | |
| "loss": 0.988, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.9991460290350127, | |
| "eval_loss": 2.468273639678955, | |
| "eval_runtime": 0.5562, | |
| "eval_samples_per_second": 17.978, | |
| "eval_steps_per_second": 1.798, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 3.00597779675491, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 0.00017629620641639103, | |
| "loss": 0.9741, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.0230572160546543, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00017590845667275312, | |
| "loss": 0.9621, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.040136635354398, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00017551799500270198, | |
| "loss": 0.968, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.0572160546541416, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00017512483535597867, | |
| "loss": 0.9683, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.0742954739538857, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00017472899177871297, | |
| "loss": 0.9671, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.0913748932536294, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.0001743304784129214, | |
| "loss": 0.9563, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.108454312553373, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00017392930949600217, | |
| "loss": 0.9678, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.125533731853117, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.0001735254993602264, | |
| "loss": 0.9594, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.1426131511528608, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00017311906243222614, | |
| "loss": 0.9691, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.1596925704526044, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.0001727100132324789, | |
| "loss": 0.9694, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.1767719897523485, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00017229836637478902, | |
| "loss": 0.9678, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.193851409052092, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00017188413656576534, | |
| "loss": 0.972, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.210930828351836, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00017146733860429612, | |
| "loss": 0.9661, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.22801024765158, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00017104798738101993, | |
| "loss": 0.9567, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.2450896669513236, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00017062609787779403, | |
| "loss": 0.9605, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.2621690862510673, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00017020168516715894, | |
| "loss": 0.9678, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.2792485055508114, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00016977476441179992, | |
| "loss": 0.961, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.296327924850555, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00016934535086400538, | |
| "loss": 0.9657, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.313407344150299, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.0001689134598651219, | |
| "loss": 0.9601, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.330486763450043, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.00016847910684500615, | |
| "loss": 0.9652, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.3475661827497865, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0001680423073214737, | |
| "loss": 0.9755, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.3646456020495306, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.0001676030768997445, | |
| "loss": 0.9641, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.381725021349274, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 0.00016716143127188548, | |
| "loss": 0.9652, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.398804440649018, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.0001667173862162499, | |
| "loss": 0.9647, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 3.415883859948762, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00016627095759691362, | |
| "loss": 0.9689, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.4329632792485056, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.0001658221613631083, | |
| "loss": 0.9656, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 3.4500426985482493, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.0001653710135486518, | |
| "loss": 0.9601, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.4671221178479934, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00016491753027137498, | |
| "loss": 0.9669, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 3.484201537147737, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00016446172773254629, | |
| "loss": 0.9606, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.5012809564474807, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00016400362221629264, | |
| "loss": 0.9693, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.518360375747225, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00016354323008901776, | |
| "loss": 0.9631, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.5354397950469685, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.0001630805677988175, | |
| "loss": 0.9601, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 3.552519214346712, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.0001626156518748922, | |
| "loss": 0.9593, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.5695986336464562, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00016214849892695602, | |
| "loss": 0.9611, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 3.5866780529462, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00016167912564464383, | |
| "loss": 0.966, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.6037574722459436, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00016120754879691464, | |
| "loss": 0.9651, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 3.6208368915456877, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.0001607337852314527, | |
| "loss": 0.9591, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.6379163108454313, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00016025785187406553, | |
| "loss": 0.9578, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 3.654995730145175, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.0001597797657280792, | |
| "loss": 0.9691, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.672075149444919, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00015929954387373103, | |
| "loss": 0.9591, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 3.6891545687446627, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 0.00015881720346755905, | |
| "loss": 0.9629, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.7062339880444064, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00015833276174178937, | |
| "loss": 0.9564, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 3.7233134073441505, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00015784623600372042, | |
| "loss": 0.9644, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.740392826643894, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0001573576436351046, | |
| "loss": 0.9592, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 3.757472245943638, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00015686700209152738, | |
| "loss": 0.9709, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.774551665243382, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 0.00015637432890178353, | |
| "loss": 0.9658, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 3.7916310845431256, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00015587964166725095, | |
| "loss": 0.9621, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.8087105038428692, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.00015538295806126205, | |
| "loss": 0.9648, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 3.8257899231426133, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00015488429582847192, | |
| "loss": 0.9647, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.842869342442357, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0001543836727842248, | |
| "loss": 0.9569, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 3.8599487617421007, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00015388110681391725, | |
| "loss": 0.9615, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.8770281810418448, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00015337661587235953, | |
| "loss": 0.9561, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 3.8941076003415884, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.0001528702179831338, | |
| "loss": 0.9686, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.911187019641332, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00015236193123795041, | |
| "loss": 0.959, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 3.928266438941076, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00015185177379600152, | |
| "loss": 0.9545, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.94534585824082, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00015133976388331227, | |
| "loss": 0.9626, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 3.9624252775405635, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.00015082591979208976, | |
| "loss": 0.9595, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.9795046968403076, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00015031025988006936, | |
| "loss": 0.959, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 3.9965841161400513, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.000149792802569859, | |
| "loss": 0.9741, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.4999709129333496, | |
| "eval_runtime": 0.5563, | |
| "eval_samples_per_second": 17.977, | |
| "eval_steps_per_second": 1.798, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 4.013663535439795, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.00014927356634828094, | |
| "loss": 0.943, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.030742954739539, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00014875256976571135, | |
| "loss": 0.9301, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.047822374039282, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00014822983143541752, | |
| "loss": 0.9339, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.064901793339026, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.0001477053700328929, | |
| "loss": 0.9284, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.0819812126387705, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00014717920429518984, | |
| "loss": 0.9403, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.099060631938514, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00014665135302025035, | |
| "loss": 0.936, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.116140051238258, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00014612183506623432, | |
| "loss": 0.9361, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.133219470538002, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00014559066935084588, | |
| "loss": 0.9354, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.150298889837745, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.0001450578748506576, | |
| "loss": 0.9339, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.167378309137489, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.00014452347060043237, | |
| "loss": 0.9319, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.184457728437233, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00014398747569244354, | |
| "loss": 0.9403, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.2015371477369765, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00014344990927579268, | |
| "loss": 0.9368, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.218616567036721, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00014291079055572554, | |
| "loss": 0.9327, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 4.235695986336465, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.0001423701387929459, | |
| "loss": 0.942, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.252775405636209, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0001418279733029274, | |
| "loss": 0.9416, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 4.269854824935952, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.0001412843134552235, | |
| "loss": 0.9365, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.286934244235696, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.00014073917867277557, | |
| "loss": 0.9334, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 4.304013663535439, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00014019258843121893, | |
| "loss": 0.9374, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.3210930828351835, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.0001396445622581869, | |
| "loss": 0.9309, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 4.3381725021349276, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.0001390951197326134, | |
| "loss": 0.9256, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.355251921434672, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00013854428048403324, | |
| "loss": 0.9336, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 4.372331340734415, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00013799206419188103, | |
| "loss": 0.9441, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.389410760034159, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.00013743849058478808, | |
| "loss": 0.938, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 4.406490179333902, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00013688357943987732, | |
| "loss": 0.9389, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.423569598633646, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00013632735058205706, | |
| "loss": 0.945, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 4.44064901793339, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.0001357698238833126, | |
| "loss": 0.9378, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.4577284372331345, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00013521101926199607, | |
| "loss": 0.9378, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 4.474807856532878, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0001346509566821153, | |
| "loss": 0.9409, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.491887275832622, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00013408965615262008, | |
| "loss": 0.9363, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 4.508966695132365, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00013352713772668765, | |
| "loss": 0.9413, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.526046114432109, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00013296342150100605, | |
| "loss": 0.9509, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 4.543125533731853, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00013239852761505626, | |
| "loss": 0.9361, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.560204953031597, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00013183247625039282, | |
| "loss": 0.9366, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 4.577284372331341, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00013126528762992247, | |
| "loss": 0.9381, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.594363791631085, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.000130696982017182, | |
| "loss": 0.9394, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 4.611443210930828, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00013012757971561415, | |
| "loss": 0.9363, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.628522630230572, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00012955710106784214, | |
| "loss": 0.9323, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 4.645602049530316, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00012898556645494325, | |
| "loss": 0.9387, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.66268146883006, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00012841299629572032, | |
| "loss": 0.935, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 4.679760888129803, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.0001278394110459724, | |
| "loss": 0.9446, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.6968403074295475, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.000127264831197764, | |
| "loss": 0.9372, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 4.713919726729291, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0001266892772786929, | |
| "loss": 0.9363, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.730999146029035, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.00012611276985115678, | |
| "loss": 0.9394, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 4.748078565328779, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.0001255353295116187, | |
| "loss": 0.9438, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.765157984628523, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00012495697688987112, | |
| "loss": 0.942, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 4.782237403928266, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00012437773264829897, | |
| "loss": 0.9436, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.79931682322801, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001237976174811414, | |
| "loss": 0.9403, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 4.816396242527754, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00012321665211375256, | |
| "loss": 0.9361, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.833475661827498, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00012263485730186103, | |
| "loss": 0.9404, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 4.850555081127242, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00012205225383082843, | |
| "loss": 0.9409, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.867634500426986, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.0001214688625149066, | |
| "loss": 0.9351, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 4.884713919726729, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00012088470419649432, | |
| "loss": 0.938, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.901793339026473, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 0.00012029979974539234, | |
| "loss": 0.9425, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 4.918872758326217, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.00011971417005805818, | |
| "loss": 0.9372, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.9359521776259605, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00011912783605685913, | |
| "loss": 0.9399, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 4.953031596925705, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0001185408186893251, | |
| "loss": 0.9385, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.970111016225449, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.0001179531389274001, | |
| "loss": 0.9311, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 4.987190435525192, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.9342, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.997438087105039, | |
| "eval_loss": 2.5202550888061523, | |
| "eval_runtime": 0.5609, | |
| "eval_samples_per_second": 17.829, | |
| "eval_steps_per_second": 1.783, | |
| "step": 1463 | |
| }, | |
| { | |
| "epoch": 5.004269854824936, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00011677587622572763, | |
| "loss": 0.9354, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 5.02134927412468, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00011618633534519141, | |
| "loss": 0.9194, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 5.038428693424423, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00011559621618718414, | |
| "loss": 0.9073, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 5.0555081127241674, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00011500553983446527, | |
| "loss": 0.9146, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 5.0725875320239115, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00011441432738970072, | |
| "loss": 0.9098, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 5.089666951323655, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00011382259997470899, | |
| "loss": 0.9135, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 5.106746370623399, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00011323037872970657, | |
| "loss": 0.9174, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 5.123825789923143, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.00011263768481255264, | |
| "loss": 0.9155, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.140905209222886, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00011204453939799315, | |
| "loss": 0.9115, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 5.15798462852263, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00011145096367690444, | |
| "loss": 0.9112, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 5.175064047822374, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0001108569788555361, | |
| "loss": 0.9142, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 5.192143467122118, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00011026260615475333, | |
| "loss": 0.9116, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 5.209222886421862, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00010966786680927874, | |
| "loss": 0.9141, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 5.226302305721606, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00010907278206693395, | |
| "loss": 0.9168, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 5.243381725021349, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00010847737318788013, | |
| "loss": 0.9216, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 5.260461144321093, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00010788166144385888, | |
| "loss": 0.9167, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 5.277540563620837, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.0001072856681174318, | |
| "loss": 0.9155, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 5.2946199829205804, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00010668941450122055, | |
| "loss": 0.9218, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 5.3116994022203246, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00010609292189714586, | |
| "loss": 0.9132, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 5.328778821520069, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.0001054962116156667, | |
| "loss": 0.9181, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 5.345858240819812, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 0.0001048993049750188, | |
| "loss": 0.9151, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 5.362937660119556, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00010430222330045304, | |
| "loss": 0.9096, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 5.3800170794193, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.0001037049879234737, | |
| "loss": 0.9183, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 5.397096498719043, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.0001031076201810762, | |
| "loss": 0.9151, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 5.414175918018787, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00010251014141498484, | |
| "loss": 0.9205, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.4312553373185315, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00010191257297089052, | |
| "loss": 0.9114, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.448334756618275, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 0.00010131493619768788, | |
| "loss": 0.9148, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 5.465414175918019, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 0.00010071725244671282, | |
| "loss": 0.9202, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.482493595217763, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00010011954307097942, | |
| "loss": 0.9216, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 5.499573014517506, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 9.952182942441733e-05, | |
| "loss": 0.9206, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 5.51665243381725, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 9.892413286110886e-05, | |
| "loss": 0.9193, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 5.533731853116994, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 9.83264747345259e-05, | |
| "loss": 0.9195, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 5.5508112724167376, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.772887639676707e-05, | |
| "loss": 0.9178, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 5.567890691716482, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 9.713135919779515e-05, | |
| "loss": 0.9174, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 5.584970111016226, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.653394448467399e-05, | |
| "loss": 0.9194, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 5.602049530315969, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 9.593665360080599e-05, | |
| "loss": 0.9192, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.619128949615713, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.533950788516974e-05, | |
| "loss": 0.9154, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 5.636208368915457, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 9.474252867155732e-05, | |
| "loss": 0.9142, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 5.6532877882152, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 9.414573728781247e-05, | |
| "loss": 0.9101, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 5.6703672075149445, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.354915505506839e-05, | |
| "loss": 0.9158, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 5.687446626814689, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 9.295280328698604e-05, | |
| "loss": 0.9181, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 5.704526046114432, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.235670328899293e-05, | |
| "loss": 0.9138, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 5.721605465414176, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 9.176087635752156e-05, | |
| "loss": 0.9119, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 5.73868488471392, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 9.116534377924883e-05, | |
| "loss": 0.9213, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 5.755764304013663, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 9.057012683033555e-05, | |
| "loss": 0.9177, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 5.772843723313407, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 8.997524677566627e-05, | |
| "loss": 0.9217, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 5.789923142613151, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 8.938072486808952e-05, | |
| "loss": 0.9167, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 5.807002561912895, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 8.878658234765858e-05, | |
| "loss": 0.9207, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.824081981212639, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 8.81928404408726e-05, | |
| "loss": 0.9173, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 5.841161400512383, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 8.759952035991844e-05, | |
| "loss": 0.9192, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 5.858240819812126, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 8.70066433019125e-05, | |
| "loss": 0.9178, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 5.87532023911187, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 8.641423044814374e-05, | |
| "loss": 0.9246, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 5.892399658411614, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 8.582230296331686e-05, | |
| "loss": 0.9187, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 5.9094790777113575, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 8.5230881994796e-05, | |
| "loss": 0.9211, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 5.926558497011102, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 8.463998867184952e-05, | |
| "loss": 0.9194, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 5.943637916310846, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 8.404964410489485e-05, | |
| "loss": 0.9215, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 5.960717335610589, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 8.34598693847444e-05, | |
| "loss": 0.9144, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 5.977796754910333, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 8.287068558185225e-05, | |
| "loss": 0.9175, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 5.994876174210077, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 8.228211374556103e-05, | |
| "loss": 0.9201, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 5.998292058070025, | |
| "eval_loss": 2.5518593788146973, | |
| "eval_runtime": 0.5597, | |
| "eval_samples_per_second": 17.868, | |
| "eval_steps_per_second": 1.787, | |
| "step": 1756 | |
| }, | |
| { | |
| "epoch": 6.01195559350982, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 8.169417490335007e-05, | |
| "loss": 0.9044, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 6.029035012809564, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 8.110689006008434e-05, | |
| "loss": 0.8914, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 6.0461144321093085, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 8.052028019726371e-05, | |
| "loss": 0.896, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 6.063193851409052, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 7.993436627227368e-05, | |
| "loss": 0.9072, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 6.080273270708796, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 7.934916921763628e-05, | |
| "loss": 0.8999, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 6.09735269000854, | |
| "grad_norm": 0.25, | |
| "learning_rate": 7.876470994026254e-05, | |
| "loss": 0.8991, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 6.114432109308283, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 7.818100932070546e-05, | |
| "loss": 0.9065, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 6.131511528608027, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 7.759808821241406e-05, | |
| "loss": 0.8899, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 6.148590947907771, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 7.701596744098818e-05, | |
| "loss": 0.8956, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 6.165670367207515, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 7.643466780343479e-05, | |
| "loss": 0.8964, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 6.182749786507259, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 7.585421006742463e-05, | |
| "loss": 0.8985, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 6.199829205807003, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 7.527461497055061e-05, | |
| "loss": 0.8979, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 6.216908625106746, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 7.469590321958662e-05, | |
| "loss": 0.9014, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 6.23398804440649, | |
| "grad_norm": 0.25, | |
| "learning_rate": 7.411809548974792e-05, | |
| "loss": 0.9059, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 6.251067463706234, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 7.354121242395254e-05, | |
| "loss": 0.903, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 6.268146883005977, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 7.296527463208358e-05, | |
| "loss": 0.8955, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 6.2852263023057215, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 7.239030269025311e-05, | |
| "loss": 0.8991, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 6.302305721605466, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 7.1816317140067e-05, | |
| "loss": 0.9014, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 6.319385140905209, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 7.124333848789091e-05, | |
| "loss": 0.9015, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 6.336464560204953, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 7.067138720411795e-05, | |
| "loss": 0.9039, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 6.353543979504697, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 7.010048372243698e-05, | |
| "loss": 0.8993, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 6.37062339880444, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 6.953064843910296e-05, | |
| "loss": 0.908, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 6.387702818104184, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 6.8961901712208e-05, | |
| "loss": 0.9021, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 6.4047822374039285, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 6.839426386095425e-05, | |
| "loss": 0.9002, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 6.421861656703672, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 6.782775516492771e-05, | |
| "loss": 0.9007, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 6.438941076003416, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.726239586337408e-05, | |
| "loss": 0.8959, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 6.45602049530316, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 6.669820615447522e-05, | |
| "loss": 0.9078, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 6.473099914602903, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 6.613520619462803e-05, | |
| "loss": 0.8996, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 6.490179333902647, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 6.5573416097724e-05, | |
| "loss": 0.9023, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 6.507258753202391, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 6.50128559344307e-05, | |
| "loss": 0.9004, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 6.5243381725021345, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 6.445354573147484e-05, | |
| "loss": 0.9088, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 6.541417591801879, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 6.389550547092661e-05, | |
| "loss": 0.8937, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 6.558497011101623, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 6.333875508948593e-05, | |
| "loss": 0.906, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 6.575576430401366, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 6.278331447777021e-05, | |
| "loss": 0.9062, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 6.59265584970111, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 6.22292034796035e-05, | |
| "loss": 0.9004, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 6.609735269000854, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 6.167644189130794e-05, | |
| "loss": 0.8995, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 6.626814688300598, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 6.112504946099604e-05, | |
| "loss": 0.9011, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 6.6438941076003415, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 6.057504588786556e-05, | |
| "loss": 0.8957, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 6.660973526900086, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 6.0026450821495536e-05, | |
| "loss": 0.909, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 6.678052946199829, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 5.947928386114428e-05, | |
| "loss": 0.8996, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 6.695132365499573, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 5.8933564555049105e-05, | |
| "loss": 0.9072, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 6.712211784799317, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 5.838931239972824e-05, | |
| "loss": 0.9022, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 6.729291204099061, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 5.784654683928391e-05, | |
| "loss": 0.9009, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 6.746370623398804, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 5.730528726470792e-05, | |
| "loss": 0.8999, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 6.763450042698548, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 5.6765553013188766e-05, | |
| "loss": 0.9002, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 6.780529461998292, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 5.622736336742087e-05, | |
| "loss": 0.8965, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 6.797608881298036, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 5.5690737554915604e-05, | |
| "loss": 0.9015, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 6.81468830059778, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 5.5155694747314504e-05, | |
| "loss": 0.9105, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 6.831767719897524, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 5.462225405970401e-05, | |
| "loss": 0.8978, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.848847139197267, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 5.4090434549933064e-05, | |
| "loss": 0.8999, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 6.865926558497011, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 5.3560255217931785e-05, | |
| "loss": 0.8988, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 6.8830059777967545, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 5.303173500503289e-05, | |
| "loss": 0.9055, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 6.900085397096499, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 5.2504892793295e-05, | |
| "loss": 0.8991, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 6.917164816396243, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 5.197974740482785e-05, | |
| "loss": 0.8997, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 6.934244235695987, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 5.145631760112022e-05, | |
| "loss": 0.8983, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 6.95132365499573, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 5.093462208236931e-05, | |
| "loss": 0.9038, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 6.968403074295474, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 5.041467948681269e-05, | |
| "loss": 0.8978, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 6.985482493595217, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 4.989650839006279e-05, | |
| "loss": 0.9054, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 6.999146029035013, | |
| "eval_loss": 2.5763192176818848, | |
| "eval_runtime": 0.5559, | |
| "eval_samples_per_second": 17.989, | |
| "eval_steps_per_second": 1.799, | |
| "step": 2049 | |
| }, | |
| { | |
| "epoch": 7.002561912894961, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 4.9380127304442634e-05, | |
| "loss": 0.8953, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 7.0196413321947055, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 4.886555467832512e-05, | |
| "loss": 0.893, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 7.036720751494449, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 4.835280889547351e-05, | |
| "loss": 0.8885, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 7.053800170794193, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 4.7841908274384616e-05, | |
| "loss": 0.8916, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 7.070879590093937, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 4.733287106763481e-05, | |
| "loss": 0.8906, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 7.08795900939368, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 4.6825715461227284e-05, | |
| "loss": 0.8876, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 7.105038428693424, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 4.6320459573942856e-05, | |
| "loss": 0.8908, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 7.122117847993168, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 4.581712145669239e-05, | |
| "loss": 0.8887, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 7.1391972672929125, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 4.531571909187197e-05, | |
| "loss": 0.886, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 7.156276686592656, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 4.481627039272056e-05, | |
| "loss": 0.8883, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 7.1733561058924, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 4.431879320267972e-05, | |
| "loss": 0.8922, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 7.190435525192143, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 4.38233052947565e-05, | |
| "loss": 0.8825, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 7.207514944491887, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 4.332982437088825e-05, | |
| "loss": 0.8856, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 7.224594363791631, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 4.2838368061310276e-05, | |
| "loss": 0.8929, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 7.241673783091375, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 4.2348953923925916e-05, | |
| "loss": 0.8977, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 7.2587532023911185, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 4.186159944367936e-05, | |
| "loss": 0.8855, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 7.275832621690863, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 4.137632203193086e-05, | |
| "loss": 0.8837, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 7.292912040990606, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 4.0893139025834806e-05, | |
| "loss": 0.8927, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 7.30999146029035, | |
| "grad_norm": 0.25, | |
| "learning_rate": 4.041206768772022e-05, | |
| "loss": 0.8902, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 7.327070879590094, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 3.993312520447414e-05, | |
| "loss": 0.8904, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 7.344150298889838, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 3.9456328686927525e-05, | |
| "loss": 0.8885, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 7.361229718189581, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 3.898169516924398e-05, | |
| "loss": 0.8945, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 7.3783091374893255, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 3.850924160831115e-05, | |
| "loss": 0.892, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 7.395388556789069, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 3.803898488313501e-05, | |
| "loss": 0.8933, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 7.412467976088813, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 3.757094179423672e-05, | |
| "loss": 0.892, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 7.429547395388557, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 3.710512906305248e-05, | |
| "loss": 0.8905, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 7.446626814688301, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 3.6641563331336125e-05, | |
| "loss": 0.888, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 7.463706233988044, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 3.618026116056456e-05, | |
| "loss": 0.8847, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 7.480785653287788, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 0.8867, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 7.497865072587532, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 3.5264513342831615e-05, | |
| "loss": 0.8894, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 7.514944491887276, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 3.4810100412128747e-05, | |
| "loss": 0.8894, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 7.53202391118702, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 3.435801647371897e-05, | |
| "loss": 0.8922, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 7.549103330486764, | |
| "grad_norm": 0.25, | |
| "learning_rate": 3.3908277678877445e-05, | |
| "loss": 0.8934, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 7.566182749786507, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 3.346090009509613e-05, | |
| "loss": 0.8865, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 7.583262169086251, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 3.3015899705509734e-05, | |
| "loss": 0.8889, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 7.600341588385994, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 3.257329240832454e-05, | |
| "loss": 0.886, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 7.6174210076857385, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 3.21330940162508e-05, | |
| "loss": 0.8875, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 7.634500426985483, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 3.169532025593729e-05, | |
| "loss": 0.8863, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 7.651579846285227, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 3.125998676740987e-05, | |
| "loss": 0.8945, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 7.66865926558497, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 3.0827109103512643e-05, | |
| "loss": 0.888, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 7.685738684884714, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 3.0396702729352023e-05, | |
| "loss": 0.895, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 7.702818104184458, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 2.996878302174472e-05, | |
| "loss": 0.89, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 7.719897523484201, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 2.9543365268667867e-05, | |
| "loss": 0.8868, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 7.736976942783945, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.9120464668713188e-05, | |
| "loss": 0.8944, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 7.7540563620836895, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 2.8700096330544012e-05, | |
| "loss": 0.8946, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 7.771135781383433, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 2.828227527235513e-05, | |
| "loss": 0.8926, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 7.788215200683177, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 2.7867016421336776e-05, | |
| "loss": 0.8984, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 7.805294619982921, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.7454334613140864e-05, | |
| "loss": 0.8874, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 7.822374039282664, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 2.7044244591351232e-05, | |
| "loss": 0.892, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 7.839453458582408, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 2.6636761006956955e-05, | |
| "loss": 0.8936, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 7.856532877882152, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 2.6231898417828603e-05, | |
| "loss": 0.8856, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.873612297181896, | |
| "grad_norm": 0.25, | |
| "learning_rate": 2.582967128819851e-05, | |
| "loss": 0.8886, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 7.89069171648164, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 2.5430093988143778e-05, | |
| "loss": 0.8891, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 7.907771135781384, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 2.5033180793072986e-05, | |
| "loss": 0.8808, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 7.924850555081127, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 2.4638945883216235e-05, | |
| "loss": 0.8868, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 7.941929974380871, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 2.4247403343118335e-05, | |
| "loss": 0.8934, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 7.959009393680615, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 2.385856716113587e-05, | |
| "loss": 0.8878, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 7.976088812980358, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 2.3472451228937253e-05, | |
| "loss": 0.8913, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 7.9931682322801025, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 2.3089069341006565e-05, | |
| "loss": 0.8902, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.592200994491577, | |
| "eval_runtime": 0.5427, | |
| "eval_samples_per_second": 18.425, | |
| "eval_steps_per_second": 1.843, | |
| "step": 2342 | |
| }, | |
| { | |
| "epoch": 8.010247651579846, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 2.2708435194150634e-05, | |
| "loss": 0.8945, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 8.02732707087959, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 2.2330562387009745e-05, | |
| "loss": 0.8833, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 8.044406490179334, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 2.1955464419571782e-05, | |
| "loss": 0.8823, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 8.061485909479078, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 2.1583154692689976e-05, | |
| "loss": 0.8874, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 8.078565328778822, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 2.121364650760408e-05, | |
| "loss": 0.8743, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 8.095644748078564, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 2.08469530654652e-05, | |
| "loss": 0.8872, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 8.112724167378309, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 2.048308746686417e-05, | |
| "loss": 0.8936, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 8.129803586678053, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 2.0122062711363532e-05, | |
| "loss": 0.8818, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 8.146883005977797, | |
| "grad_norm": 0.25, | |
| "learning_rate": 1.9763891697032978e-05, | |
| "loss": 0.887, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 8.163962425277541, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 1.9408587219988805e-05, | |
| "loss": 0.884, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 8.181041844577285, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 1.9056161973936513e-05, | |
| "loss": 0.8892, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 8.198121263877027, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 1.8706628549717452e-05, | |
| "loss": 0.8883, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 8.215200683176771, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 1.835999943485892e-05, | |
| "loss": 0.8802, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 8.232280102476516, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 1.8016287013128018e-05, | |
| "loss": 0.8886, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 8.24935952177626, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 1.767550356408938e-05, | |
| "loss": 0.8784, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 8.266438941076004, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 1.7337661262666294e-05, | |
| "loss": 0.8897, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 8.283518360375748, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 1.7002772178705716e-05, | |
| "loss": 0.8844, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 8.30059777967549, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 1.6670848276547334e-05, | |
| "loss": 0.8856, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 8.317677198975234, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 1.6341901414595705e-05, | |
| "loss": 0.8762, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 8.334756618274978, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 1.601594334489702e-05, | |
| "loss": 0.8802, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 8.351836037574722, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 1.5692985712719e-05, | |
| "loss": 0.8939, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 8.368915456874467, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 1.5373040056134814e-05, | |
| "loss": 0.8804, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 8.38599487617421, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 1.5056117805611115e-05, | |
| "loss": 0.8806, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 8.403074295473953, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 1.474223028359939e-05, | |
| "loss": 0.8856, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 8.420153714773697, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 1.4431388704131632e-05, | |
| "loss": 0.8791, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 8.437233134073441, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 1.4123604172419713e-05, | |
| "loss": 0.8874, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 8.454312553373185, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 1.3818887684458426e-05, | |
| "loss": 0.8827, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 8.47139197267293, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 1.3517250126632986e-05, | |
| "loss": 0.8847, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 8.488471391972674, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 1.321870227532971e-05, | |
| "loss": 0.8826, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 8.505550811272418, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 1.292325479655131e-05, | |
| "loss": 0.89, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 8.52263023057216, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 1.263091824553574e-05, | |
| "loss": 0.8904, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 8.539709649871904, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 1.2341703066379074e-05, | |
| "loss": 0.888, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 8.556789069171648, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 1.205561959166237e-05, | |
| "loss": 0.8841, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 8.573868488471392, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 1.1772678042082607e-05, | |
| "loss": 0.8856, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 8.590947907771136, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 1.149288852608743e-05, | |
| "loss": 0.8871, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 8.608027327070879, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 1.1216261039514087e-05, | |
| "loss": 0.8817, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 8.625106746370623, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 1.094280546523231e-05, | |
| "loss": 0.8825, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 8.642186165670367, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 1.0672531572791178e-05, | |
| "loss": 0.8922, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 8.659265584970111, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 1.0405449018070168e-05, | |
| "loss": 0.8879, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 8.676345004269855, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 1.0141567342934132e-05, | |
| "loss": 0.885, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 8.6934244235696, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 9.880895974892412e-06, | |
| "loss": 0.8886, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 8.710503842869343, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 9.623444226762035e-06, | |
| "loss": 0.8805, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 8.727583262169086, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 9.369221296335006e-06, | |
| "loss": 0.8866, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 8.74466268146883, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 9.118236266049707e-06, | |
| "loss": 0.8811, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 8.761742100768574, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 8.870498102666402e-06, | |
| "loss": 0.8849, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 8.778821520068318, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 8.626015656946895e-06, | |
| "loss": 0.8857, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 8.795900939368062, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 8.384797663338306e-06, | |
| "loss": 0.8833, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 8.812980358667804, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 8.146852739661105e-06, | |
| "loss": 0.885, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 8.830059777967548, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 7.91218938680104e-06, | |
| "loss": 0.8861, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 8.847139197267293, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 7.6808159884057e-06, | |
| "loss": 0.88, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 8.864218616567037, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 7.45274081058478e-06, | |
| "loss": 0.8794, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 8.88129803586678, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 7.2279720016148244e-06, | |
| "loss": 0.8801, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 8.898377455166525, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 7.0065175916482095e-06, | |
| "loss": 0.8818, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 8.915456874466269, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 6.788385492426053e-06, | |
| "loss": 0.8856, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 8.932536293766011, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 6.573583496995816e-06, | |
| "loss": 0.8887, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 8.949615713065755, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 6.36211927943271e-06, | |
| "loss": 0.8778, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 8.9666951323655, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 6.1540003945655286e-06, | |
| "loss": 0.8906, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 8.983774551665244, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 5.949234277706861e-06, | |
| "loss": 0.8818, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 8.997438087105039, | |
| "eval_loss": 2.5981767177581787, | |
| "eval_runtime": 0.5523, | |
| "eval_samples_per_second": 18.106, | |
| "eval_steps_per_second": 1.811, | |
| "step": 2634 | |
| }, | |
| { | |
| "epoch": 9.000853970964988, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 5.74782824438731e-06, | |
| "loss": 0.8865, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 9.017933390264732, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 5.549789490094304e-06, | |
| "loss": 0.8846, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 9.035012809564474, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 5.355125090014845e-06, | |
| "loss": 0.8845, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 9.052092228864218, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 5.163841998782837e-06, | |
| "loss": 0.8852, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 9.069171648163962, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 4.975947050230712e-06, | |
| "loss": 0.8831, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 9.086251067463706, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 4.79144695714504e-06, | |
| "loss": 0.8838, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 9.10333048676345, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 4.610348311026958e-06, | |
| "loss": 0.8892, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 9.120409906063195, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 4.432657581856525e-06, | |
| "loss": 0.882, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 9.137489325362937, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 4.25838111786162e-06, | |
| "loss": 0.8839, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 9.154568744662681, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 4.087525145291204e-06, | |
| "loss": 0.8854, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 9.171648163962425, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 3.920095768192722e-06, | |
| "loss": 0.8823, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 9.18872758326217, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 3.7560989681941992e-06, | |
| "loss": 0.883, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 9.205807002561913, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 3.595540604290437e-06, | |
| "loss": 0.8795, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 9.222886421861658, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 3.4384264126337328e-06, | |
| "loss": 0.8868, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 9.2399658411614, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 3.284762006328945e-06, | |
| "loss": 0.8884, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 9.257045260461144, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 3.1345528752329212e-06, | |
| "loss": 0.8819, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 9.274124679760888, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 2.9878043857584415e-06, | |
| "loss": 0.8893, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 9.291204099060632, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 2.8445217806824077e-06, | |
| "loss": 0.8805, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 9.308283518360376, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 2.704710178958603e-06, | |
| "loss": 0.8796, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 9.32536293766012, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 2.5683745755348044e-06, | |
| "loss": 0.8853, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 9.342442356959863, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 2.435519841174272e-06, | |
| "loss": 0.8844, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 9.359521776259607, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 2.30615072228183e-06, | |
| "loss": 0.8838, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 9.376601195559351, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 2.180271840734216e-06, | |
| "loss": 0.8895, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 9.393680614859095, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 2.057887693714988e-06, | |
| "loss": 0.876, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 9.410760034158839, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 1.9390026535538674e-06, | |
| "loss": 0.8831, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 9.427839453458583, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 1.8236209675705274e-06, | |
| "loss": 0.8851, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 9.444918872758326, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 1.7117467579228053e-06, | |
| "loss": 0.876, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 9.46199829205807, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 1.6033840214595308e-06, | |
| "loss": 0.879, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 9.479077711357814, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 1.4985366295776404e-06, | |
| "loss": 0.8899, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 9.496157130657558, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 1.397208328083921e-06, | |
| "loss": 0.8836, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 9.513236549957302, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 1.2994027370611173e-06, | |
| "loss": 0.8905, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 9.530315969257046, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 1.205123350738746e-06, | |
| "loss": 0.8875, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 9.547395388556788, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 1.114373537368063e-06, | |
| "loss": 0.8838, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 9.564474807856532, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 1.0271565391018922e-06, | |
| "loss": 0.8807, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 9.581554227156277, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 9.434754718787409e-07, | |
| "loss": 0.875, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 9.59863364645602, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 8.633333253113995e-07, | |
| "loss": 0.8845, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 9.615713065755765, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 7.867329625802833e-07, | |
| "loss": 0.88, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 9.632792485055509, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 7.136771203310245e-07, | |
| "loss": 0.8794, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 9.649871904355251, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 6.441684085767396e-07, | |
| "loss": 0.8894, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 9.666951323654995, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 5.782093106048159e-07, | |
| "loss": 0.8803, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 9.68403074295474, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 5.158021828881032e-07, | |
| "loss": 0.8844, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 9.701110162254484, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 4.569492550008603e-07, | |
| "loss": 0.8835, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 9.718189581554228, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 4.016526295389622e-07, | |
| "loss": 0.8832, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 9.735269000853972, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 3.49914282044872e-07, | |
| "loss": 0.879, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 9.752348420153714, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 3.017360609370301e-07, | |
| "loss": 0.8805, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 9.769427839453458, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 2.5711968744382974e-07, | |
| "loss": 0.8853, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 9.786507258753202, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 2.1606675554209922e-07, | |
| "loss": 0.8901, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 9.803586678052946, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 1.7857873190019192e-07, | |
| "loss": 0.8816, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 9.82066609735269, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 1.446569558255395e-07, | |
| "loss": 0.8823, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 9.837745516652435, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 1.143026392168789e-07, | |
| "loss": 0.8837, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 9.854824935952177, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 8.751686652084256e-08, | |
| "loss": 0.8835, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 9.871904355251921, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 6.430059469334504e-08, | |
| "loss": 0.8839, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 9.888983774551665, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 4.465465316529915e-08, | |
| "loss": 0.8863, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 9.90606319385141, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 2.8579743813006432e-08, | |
| "loss": 0.8822, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 9.923142613151153, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 1.6076440933099345e-08, | |
| "loss": 0.8817, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 9.940222032450897, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 7.145191222035497e-09, | |
| "loss": 0.8827, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 9.95730145175064, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 1.7863137600993008e-09, | |
| "loss": 0.881, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 9.974380871050384, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.0, | |
| "loss": 0.8852, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 9.974380871050384, | |
| "eval_loss": 2.5989506244659424, | |
| "eval_runtime": 0.5586, | |
| "eval_samples_per_second": 17.903, | |
| "eval_steps_per_second": 1.79, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 9.974380871050384, | |
| "step": 2920, | |
| "total_flos": 3.4809256003093135e+18, | |
| "train_loss": 0.9919237802289936, | |
| "train_runtime": 34991.5416, | |
| "train_samples_per_second": 8.027, | |
| "train_steps_per_second": 0.083 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 2920, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.4809256003093135e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |