| { | |
| "best_metric": 2.169156074523926, | |
| "best_model_checkpoint": "/home/sunggeunan/data/ICL/outputs/lora/SKIML-ICL_mrqa_nq_v3/Meta-Llama-3-8B-Instruct-unanswerable-1Q-0U-0C-qa_first/checkpoint-297", | |
| "epoch": 0.9970625262274444, | |
| "eval_steps": 500, | |
| "global_step": 297, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003357112882920688, | |
| "grad_norm": 0.39633309841156006, | |
| "learning_rate": 1.111111111111111e-08, | |
| "loss": 2.1607, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006714225765841376, | |
| "grad_norm": 0.4136360287666321, | |
| "learning_rate": 2.222222222222222e-08, | |
| "loss": 2.2063, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010071338648762064, | |
| "grad_norm": 0.40252378582954407, | |
| "learning_rate": 3.3333333333333334e-08, | |
| "loss": 2.1702, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.013428451531682753, | |
| "grad_norm": 0.3657248914241791, | |
| "learning_rate": 4.444444444444444e-08, | |
| "loss": 2.1677, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01678556441460344, | |
| "grad_norm": 0.38506612181663513, | |
| "learning_rate": 5.555555555555555e-08, | |
| "loss": 2.203, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.020142677297524128, | |
| "grad_norm": 0.39267775416374207, | |
| "learning_rate": 6.666666666666667e-08, | |
| "loss": 2.1989, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02349979018044482, | |
| "grad_norm": 0.41893133521080017, | |
| "learning_rate": 7.777777777777778e-08, | |
| "loss": 2.1882, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.026856903063365505, | |
| "grad_norm": 0.363130122423172, | |
| "learning_rate": 8.888888888888888e-08, | |
| "loss": 2.1636, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.030214015946286196, | |
| "grad_norm": 0.43022215366363525, | |
| "learning_rate": 1e-07, | |
| "loss": 2.1881, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03357112882920688, | |
| "grad_norm": 0.43208909034729004, | |
| "learning_rate": 1.111111111111111e-07, | |
| "loss": 2.1748, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03692824171212757, | |
| "grad_norm": 0.4211503267288208, | |
| "learning_rate": 1.2222222222222222e-07, | |
| "loss": 2.2072, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.040285354595048256, | |
| "grad_norm": 0.43464261293411255, | |
| "learning_rate": 1.3333333333333334e-07, | |
| "loss": 2.1711, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.043642467477968946, | |
| "grad_norm": 0.38066577911376953, | |
| "learning_rate": 1.4444444444444442e-07, | |
| "loss": 2.1946, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04699958036088964, | |
| "grad_norm": 0.3847394585609436, | |
| "learning_rate": 1.5555555555555556e-07, | |
| "loss": 2.1638, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05035669324381032, | |
| "grad_norm": 0.40741828083992004, | |
| "learning_rate": 1.6666666666666665e-07, | |
| "loss": 2.2389, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05371380612673101, | |
| "grad_norm": 0.37301868200302124, | |
| "learning_rate": 1.7777777777777776e-07, | |
| "loss": 2.1762, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0570709190096517, | |
| "grad_norm": 0.4193646013736725, | |
| "learning_rate": 1.8888888888888888e-07, | |
| "loss": 2.2245, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06042803189257239, | |
| "grad_norm": 0.4078114330768585, | |
| "learning_rate": 2e-07, | |
| "loss": 2.1756, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06378514477549307, | |
| "grad_norm": 0.40552276372909546, | |
| "learning_rate": 2.111111111111111e-07, | |
| "loss": 2.1302, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06714225765841376, | |
| "grad_norm": 0.40120214223861694, | |
| "learning_rate": 2.222222222222222e-07, | |
| "loss": 2.1546, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07049937054133446, | |
| "grad_norm": 0.3937098979949951, | |
| "learning_rate": 2.3333333333333333e-07, | |
| "loss": 2.1822, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07385648342425515, | |
| "grad_norm": 0.39223670959472656, | |
| "learning_rate": 2.4444444444444445e-07, | |
| "loss": 2.1548, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07721359630717582, | |
| "grad_norm": 0.395595520734787, | |
| "learning_rate": 2.5555555555555553e-07, | |
| "loss": 2.1538, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08057070919009651, | |
| "grad_norm": 0.38706085085868835, | |
| "learning_rate": 2.6666666666666667e-07, | |
| "loss": 2.162, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0839278220730172, | |
| "grad_norm": 0.40628549456596375, | |
| "learning_rate": 2.7777777777777776e-07, | |
| "loss": 2.1493, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08728493495593789, | |
| "grad_norm": 0.3962867259979248, | |
| "learning_rate": 2.8888888888888885e-07, | |
| "loss": 2.1609, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09064204783885858, | |
| "grad_norm": 0.36925041675567627, | |
| "learning_rate": 3e-07, | |
| "loss": 2.096, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.09399916072177927, | |
| "grad_norm": 0.38802072405815125, | |
| "learning_rate": 3.111111111111111e-07, | |
| "loss": 2.2456, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.09735627360469996, | |
| "grad_norm": 0.38850194215774536, | |
| "learning_rate": 3.222222222222222e-07, | |
| "loss": 2.1663, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10071338648762064, | |
| "grad_norm": 0.38965868949890137, | |
| "learning_rate": 3.333333333333333e-07, | |
| "loss": 2.2527, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10407049937054133, | |
| "grad_norm": 1.802207112312317, | |
| "learning_rate": 3.4444444444444444e-07, | |
| "loss": 2.1718, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.10742761225346202, | |
| "grad_norm": 0.41264647245407104, | |
| "learning_rate": 3.5555555555555553e-07, | |
| "loss": 2.216, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.11078472513638271, | |
| "grad_norm": 0.38629451394081116, | |
| "learning_rate": 3.666666666666666e-07, | |
| "loss": 2.1767, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1141418380193034, | |
| "grad_norm": 0.38191673159599304, | |
| "learning_rate": 3.7777777777777775e-07, | |
| "loss": 2.1206, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.11749895090222409, | |
| "grad_norm": 0.3905788064002991, | |
| "learning_rate": 3.888888888888889e-07, | |
| "loss": 2.1053, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12085606378514478, | |
| "grad_norm": 0.4043135941028595, | |
| "learning_rate": 4e-07, | |
| "loss": 2.1955, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.12421317666806546, | |
| "grad_norm": 0.446430504322052, | |
| "learning_rate": 4.1111111111111107e-07, | |
| "loss": 2.2385, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.12757028955098615, | |
| "grad_norm": 0.38461461663246155, | |
| "learning_rate": 4.222222222222222e-07, | |
| "loss": 2.1255, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.13092740243390685, | |
| "grad_norm": 0.4022009074687958, | |
| "learning_rate": 4.3333333333333335e-07, | |
| "loss": 2.1821, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.13428451531682753, | |
| "grad_norm": 0.40789297223091125, | |
| "learning_rate": 4.444444444444444e-07, | |
| "loss": 2.1387, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1376416281997482, | |
| "grad_norm": 0.4071018099784851, | |
| "learning_rate": 4.555555555555555e-07, | |
| "loss": 2.1077, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1409987410826689, | |
| "grad_norm": 0.42578282952308655, | |
| "learning_rate": 4.6666666666666666e-07, | |
| "loss": 2.1956, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1443558539655896, | |
| "grad_norm": 0.4121275842189789, | |
| "learning_rate": 4.777777777777778e-07, | |
| "loss": 2.1491, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1477129668485103, | |
| "grad_norm": 0.3832322657108307, | |
| "learning_rate": 4.888888888888889e-07, | |
| "loss": 2.1465, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.15107007973143097, | |
| "grad_norm": 0.4325246214866638, | |
| "learning_rate": 5e-07, | |
| "loss": 2.2452, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15442719261435164, | |
| "grad_norm": 0.38803404569625854, | |
| "learning_rate": 4.994089834515367e-07, | |
| "loss": 2.1442, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.15778430549727235, | |
| "grad_norm": 0.38622474670410156, | |
| "learning_rate": 4.988179669030732e-07, | |
| "loss": 2.1404, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.16114141838019302, | |
| "grad_norm": 0.365347683429718, | |
| "learning_rate": 4.982269503546099e-07, | |
| "loss": 2.1322, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.16449853126311373, | |
| "grad_norm": 0.3673339784145355, | |
| "learning_rate": 4.976359338061466e-07, | |
| "loss": 2.1242, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1678556441460344, | |
| "grad_norm": 0.3915681838989258, | |
| "learning_rate": 4.970449172576833e-07, | |
| "loss": 2.188, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1712127570289551, | |
| "grad_norm": 0.4330926239490509, | |
| "learning_rate": 4.964539007092198e-07, | |
| "loss": 2.1702, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.17456986991187579, | |
| "grad_norm": 0.40760231018066406, | |
| "learning_rate": 4.958628841607565e-07, | |
| "loss": 2.218, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.17792698279479646, | |
| "grad_norm": 0.432960569858551, | |
| "learning_rate": 4.952718676122931e-07, | |
| "loss": 2.1804, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.18128409567771717, | |
| "grad_norm": 0.38337603211402893, | |
| "learning_rate": 4.946808510638298e-07, | |
| "loss": 2.1611, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.18464120856063784, | |
| "grad_norm": 0.4071826636791229, | |
| "learning_rate": 4.940898345153664e-07, | |
| "loss": 2.1569, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.18799832144355855, | |
| "grad_norm": 0.416966050863266, | |
| "learning_rate": 4.934988179669031e-07, | |
| "loss": 2.1722, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.19135543432647922, | |
| "grad_norm": 0.42446526885032654, | |
| "learning_rate": 4.929078014184397e-07, | |
| "loss": 2.1529, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.19471254720939993, | |
| "grad_norm": 0.41747376322746277, | |
| "learning_rate": 4.923167848699764e-07, | |
| "loss": 2.1191, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1980696600923206, | |
| "grad_norm": 0.44791901111602783, | |
| "learning_rate": 4.917257683215129e-07, | |
| "loss": 2.1183, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.20142677297524128, | |
| "grad_norm": 0.39679446816444397, | |
| "learning_rate": 4.911347517730496e-07, | |
| "loss": 2.161, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.20478388585816198, | |
| "grad_norm": 0.38211897015571594, | |
| "learning_rate": 4.905437352245863e-07, | |
| "loss": 2.1334, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.20814099874108266, | |
| "grad_norm": 0.4393980801105499, | |
| "learning_rate": 4.89952718676123e-07, | |
| "loss": 2.2287, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.21149811162400337, | |
| "grad_norm": 0.40504157543182373, | |
| "learning_rate": 4.893617021276595e-07, | |
| "loss": 2.1986, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.21485522450692404, | |
| "grad_norm": 0.40123313665390015, | |
| "learning_rate": 4.887706855791962e-07, | |
| "loss": 2.2045, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.21821233738984475, | |
| "grad_norm": 0.4357161819934845, | |
| "learning_rate": 4.881796690307328e-07, | |
| "loss": 2.235, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.22156945027276542, | |
| "grad_norm": 0.39656224846839905, | |
| "learning_rate": 4.875886524822695e-07, | |
| "loss": 2.1712, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2249265631556861, | |
| "grad_norm": 0.41355738043785095, | |
| "learning_rate": 4.869976359338061e-07, | |
| "loss": 2.1957, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2282836760386068, | |
| "grad_norm": 0.4384121298789978, | |
| "learning_rate": 4.864066193853428e-07, | |
| "loss": 2.1832, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.23164078892152748, | |
| "grad_norm": 0.4240085184574127, | |
| "learning_rate": 4.858156028368794e-07, | |
| "loss": 2.1331, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.23499790180444818, | |
| "grad_norm": 0.38766977190971375, | |
| "learning_rate": 4.852245862884161e-07, | |
| "loss": 2.1492, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23835501468736886, | |
| "grad_norm": 0.4235953390598297, | |
| "learning_rate": 4.846335697399526e-07, | |
| "loss": 2.2232, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.24171212757028956, | |
| "grad_norm": 0.41447708010673523, | |
| "learning_rate": 4.840425531914893e-07, | |
| "loss": 2.1978, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.24506924045321024, | |
| "grad_norm": 0.4142104685306549, | |
| "learning_rate": 4.83451536643026e-07, | |
| "loss": 2.1532, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.24842635333613092, | |
| "grad_norm": 0.4122621417045593, | |
| "learning_rate": 4.828605200945627e-07, | |
| "loss": 2.2096, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2517834662190516, | |
| "grad_norm": 0.4637957513332367, | |
| "learning_rate": 4.822695035460992e-07, | |
| "loss": 2.2043, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2551405791019723, | |
| "grad_norm": 0.4476231038570404, | |
| "learning_rate": 4.816784869976359e-07, | |
| "loss": 2.2264, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.258497691984893, | |
| "grad_norm": 0.40626445412635803, | |
| "learning_rate": 4.810874704491725e-07, | |
| "loss": 2.2184, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2618548048678137, | |
| "grad_norm": 0.4468678832054138, | |
| "learning_rate": 4.804964539007092e-07, | |
| "loss": 2.2189, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2652119177507344, | |
| "grad_norm": 0.42194902896881104, | |
| "learning_rate": 4.799054373522458e-07, | |
| "loss": 2.1769, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.26856903063365506, | |
| "grad_norm": 0.4420163035392761, | |
| "learning_rate": 4.793144208037825e-07, | |
| "loss": 2.1329, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.27192614351657574, | |
| "grad_norm": 0.4482937455177307, | |
| "learning_rate": 4.787234042553192e-07, | |
| "loss": 2.1954, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2752832563994964, | |
| "grad_norm": 0.41959258913993835, | |
| "learning_rate": 4.781323877068558e-07, | |
| "loss": 2.1776, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.27864036928241714, | |
| "grad_norm": 0.42231133580207825, | |
| "learning_rate": 4.775413711583924e-07, | |
| "loss": 2.139, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.2819974821653378, | |
| "grad_norm": 0.4405987560749054, | |
| "learning_rate": 4.76950354609929e-07, | |
| "loss": 2.2445, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2853545950482585, | |
| "grad_norm": 0.394240140914917, | |
| "learning_rate": 4.7635933806146573e-07, | |
| "loss": 2.1289, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2887117079311792, | |
| "grad_norm": 0.44175001978874207, | |
| "learning_rate": 4.7576832151300236e-07, | |
| "loss": 2.2419, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.29206882081409985, | |
| "grad_norm": 0.41716253757476807, | |
| "learning_rate": 4.75177304964539e-07, | |
| "loss": 2.2557, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2954259336970206, | |
| "grad_norm": 0.41680270433425903, | |
| "learning_rate": 4.745862884160756e-07, | |
| "loss": 2.1866, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.29878304657994126, | |
| "grad_norm": 0.4188416600227356, | |
| "learning_rate": 4.739952718676123e-07, | |
| "loss": 2.1909, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.30214015946286193, | |
| "grad_norm": 0.41669386625289917, | |
| "learning_rate": 4.734042553191489e-07, | |
| "loss": 2.1913, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3054972723457826, | |
| "grad_norm": 0.4323998689651489, | |
| "learning_rate": 4.728132387706856e-07, | |
| "loss": 2.1811, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3088543852287033, | |
| "grad_norm": 0.431393027305603, | |
| "learning_rate": 4.722222222222222e-07, | |
| "loss": 2.1772, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.312211498111624, | |
| "grad_norm": 0.4159488081932068, | |
| "learning_rate": 4.716312056737589e-07, | |
| "loss": 2.1459, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3155686109945447, | |
| "grad_norm": 0.4011417329311371, | |
| "learning_rate": 4.7104018912529545e-07, | |
| "loss": 2.1657, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3189257238774654, | |
| "grad_norm": 0.41295671463012695, | |
| "learning_rate": 4.7044917257683213e-07, | |
| "loss": 2.1217, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.32228283676038605, | |
| "grad_norm": 0.4088380038738251, | |
| "learning_rate": 4.6985815602836876e-07, | |
| "loss": 2.152, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3256399496433068, | |
| "grad_norm": 0.43500083684921265, | |
| "learning_rate": 4.6926713947990543e-07, | |
| "loss": 2.2021, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.32899706252622746, | |
| "grad_norm": 0.4200705587863922, | |
| "learning_rate": 4.6867612293144206e-07, | |
| "loss": 2.1357, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.33235417540914813, | |
| "grad_norm": 0.4516183137893677, | |
| "learning_rate": 4.6808510638297873e-07, | |
| "loss": 2.2323, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3357112882920688, | |
| "grad_norm": 0.49128514528274536, | |
| "learning_rate": 4.674940898345153e-07, | |
| "loss": 2.2364, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3390684011749895, | |
| "grad_norm": 0.4172728657722473, | |
| "learning_rate": 4.66903073286052e-07, | |
| "loss": 2.2085, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.3424255140579102, | |
| "grad_norm": 0.4487544000148773, | |
| "learning_rate": 4.663120567375886e-07, | |
| "loss": 2.1661, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3457826269408309, | |
| "grad_norm": 0.4443681538105011, | |
| "learning_rate": 4.657210401891253e-07, | |
| "loss": 2.2109, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.34913973982375157, | |
| "grad_norm": 0.4674079418182373, | |
| "learning_rate": 4.651300236406619e-07, | |
| "loss": 2.1596, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.35249685270667225, | |
| "grad_norm": 0.43013623356819153, | |
| "learning_rate": 4.645390070921986e-07, | |
| "loss": 2.1658, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3558539655895929, | |
| "grad_norm": 0.43104687333106995, | |
| "learning_rate": 4.6394799054373515e-07, | |
| "loss": 2.1686, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.35921107847251366, | |
| "grad_norm": 0.4218711853027344, | |
| "learning_rate": 4.6335697399527183e-07, | |
| "loss": 2.141, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.36256819135543433, | |
| "grad_norm": 0.45031747221946716, | |
| "learning_rate": 4.6276595744680846e-07, | |
| "loss": 2.1561, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.365925304238355, | |
| "grad_norm": 0.48128026723861694, | |
| "learning_rate": 4.6217494089834513e-07, | |
| "loss": 2.214, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.3692824171212757, | |
| "grad_norm": 0.44868627190589905, | |
| "learning_rate": 4.6158392434988176e-07, | |
| "loss": 2.1488, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3726395300041964, | |
| "grad_norm": 0.44237226247787476, | |
| "learning_rate": 4.6099290780141843e-07, | |
| "loss": 2.1099, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3759966428871171, | |
| "grad_norm": 0.42734286189079285, | |
| "learning_rate": 4.604018912529551e-07, | |
| "loss": 2.2332, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.37935375577003777, | |
| "grad_norm": 0.45235806703567505, | |
| "learning_rate": 4.598108747044917e-07, | |
| "loss": 2.1925, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.38271086865295845, | |
| "grad_norm": 0.4485257863998413, | |
| "learning_rate": 4.5921985815602836e-07, | |
| "loss": 2.1786, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.3860679815358791, | |
| "grad_norm": 0.45567062497138977, | |
| "learning_rate": 4.58628841607565e-07, | |
| "loss": 2.1386, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.38942509441879986, | |
| "grad_norm": 0.45261716842651367, | |
| "learning_rate": 4.5803782505910166e-07, | |
| "loss": 2.136, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.39278220730172053, | |
| "grad_norm": 0.4375866949558258, | |
| "learning_rate": 4.574468085106383e-07, | |
| "loss": 2.1422, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3961393201846412, | |
| "grad_norm": 0.46383175253868103, | |
| "learning_rate": 4.5685579196217496e-07, | |
| "loss": 2.1732, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3994964330675619, | |
| "grad_norm": 0.4010314345359802, | |
| "learning_rate": 4.5626477541371153e-07, | |
| "loss": 2.1329, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.40285354595048256, | |
| "grad_norm": 0.4446873068809509, | |
| "learning_rate": 4.556737588652482e-07, | |
| "loss": 2.1791, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4062106588334033, | |
| "grad_norm": 0.47618600726127625, | |
| "learning_rate": 4.5508274231678483e-07, | |
| "loss": 2.236, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.40956777171632397, | |
| "grad_norm": 0.4493118226528168, | |
| "learning_rate": 4.544917257683215e-07, | |
| "loss": 2.1575, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.41292488459924465, | |
| "grad_norm": 0.4111258387565613, | |
| "learning_rate": 4.5390070921985813e-07, | |
| "loss": 2.24, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.4162819974821653, | |
| "grad_norm": 0.41655582189559937, | |
| "learning_rate": 4.533096926713948e-07, | |
| "loss": 2.0788, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.419639110365086, | |
| "grad_norm": 0.47266441583633423, | |
| "learning_rate": 4.5271867612293143e-07, | |
| "loss": 2.1774, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.42299622324800673, | |
| "grad_norm": 0.464999794960022, | |
| "learning_rate": 4.5212765957446806e-07, | |
| "loss": 2.143, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.4263533361309274, | |
| "grad_norm": 0.44828522205352783, | |
| "learning_rate": 4.515366430260047e-07, | |
| "loss": 2.1363, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.4297104490138481, | |
| "grad_norm": 0.4714733362197876, | |
| "learning_rate": 4.5094562647754136e-07, | |
| "loss": 2.225, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.43306756189676876, | |
| "grad_norm": 0.42666733264923096, | |
| "learning_rate": 4.50354609929078e-07, | |
| "loss": 2.1774, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4364246747796895, | |
| "grad_norm": 0.46839290857315063, | |
| "learning_rate": 4.4976359338061466e-07, | |
| "loss": 2.1427, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.43978178766261017, | |
| "grad_norm": 0.48040419816970825, | |
| "learning_rate": 4.491725768321513e-07, | |
| "loss": 2.1909, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.44313890054553084, | |
| "grad_norm": 0.4932810962200165, | |
| "learning_rate": 4.485815602836879e-07, | |
| "loss": 2.1226, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4464960134284515, | |
| "grad_norm": 0.4730973541736603, | |
| "learning_rate": 4.4799054373522453e-07, | |
| "loss": 2.1844, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4498531263113722, | |
| "grad_norm": 0.44282010197639465, | |
| "learning_rate": 4.473995271867612e-07, | |
| "loss": 2.1503, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.45321023919429293, | |
| "grad_norm": 0.4495702087879181, | |
| "learning_rate": 4.4680851063829783e-07, | |
| "loss": 2.1599, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4565673520772136, | |
| "grad_norm": 0.44728878140449524, | |
| "learning_rate": 4.462174940898345e-07, | |
| "loss": 2.1479, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4599244649601343, | |
| "grad_norm": 0.4495660960674286, | |
| "learning_rate": 4.4562647754137114e-07, | |
| "loss": 2.1272, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.46328157784305496, | |
| "grad_norm": 0.4553879499435425, | |
| "learning_rate": 4.4503546099290776e-07, | |
| "loss": 2.2729, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.46663869072597564, | |
| "grad_norm": 0.46510016918182373, | |
| "learning_rate": 4.444444444444444e-07, | |
| "loss": 2.2367, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.46999580360889637, | |
| "grad_norm": 0.4671325981616974, | |
| "learning_rate": 4.4385342789598106e-07, | |
| "loss": 2.1167, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.47335291649181704, | |
| "grad_norm": 0.4627954661846161, | |
| "learning_rate": 4.432624113475177e-07, | |
| "loss": 2.2521, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.4767100293747377, | |
| "grad_norm": 0.4297815263271332, | |
| "learning_rate": 4.4267139479905436e-07, | |
| "loss": 2.1935, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.4800671422576584, | |
| "grad_norm": 0.4634767770767212, | |
| "learning_rate": 4.4208037825059104e-07, | |
| "loss": 2.1128, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.48342425514057913, | |
| "grad_norm": 0.4689215421676636, | |
| "learning_rate": 4.4148936170212766e-07, | |
| "loss": 2.2286, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4867813680234998, | |
| "grad_norm": 0.4813438355922699, | |
| "learning_rate": 4.408983451536643e-07, | |
| "loss": 2.1433, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4901384809064205, | |
| "grad_norm": 0.45745640993118286, | |
| "learning_rate": 4.403073286052009e-07, | |
| "loss": 2.1949, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.49349559378934116, | |
| "grad_norm": 0.4202418625354767, | |
| "learning_rate": 4.397163120567376e-07, | |
| "loss": 2.1028, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.49685270667226183, | |
| "grad_norm": 0.42282456159591675, | |
| "learning_rate": 4.391252955082742e-07, | |
| "loss": 2.1114, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5002098195551825, | |
| "grad_norm": 0.4623030424118042, | |
| "learning_rate": 4.385342789598109e-07, | |
| "loss": 2.1618, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5035669324381032, | |
| "grad_norm": 0.4584071934223175, | |
| "learning_rate": 4.379432624113475e-07, | |
| "loss": 2.2274, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5069240453210239, | |
| "grad_norm": 0.43828415870666504, | |
| "learning_rate": 4.3735224586288414e-07, | |
| "loss": 2.1807, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5102811582039446, | |
| "grad_norm": 0.4550941288471222, | |
| "learning_rate": 4.3676122931442076e-07, | |
| "loss": 2.1355, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5136382710868653, | |
| "grad_norm": 0.4852266013622284, | |
| "learning_rate": 4.3617021276595744e-07, | |
| "loss": 2.1623, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.516995383969786, | |
| "grad_norm": 0.450320303440094, | |
| "learning_rate": 4.3557919621749406e-07, | |
| "loss": 2.1963, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5203524968527067, | |
| "grad_norm": 0.4544139504432678, | |
| "learning_rate": 4.3498817966903074e-07, | |
| "loss": 2.1413, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5237096097356274, | |
| "grad_norm": 0.4609904885292053, | |
| "learning_rate": 4.3439716312056736e-07, | |
| "loss": 2.2289, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.527066722618548, | |
| "grad_norm": 0.46614569425582886, | |
| "learning_rate": 4.3380614657210404e-07, | |
| "loss": 2.1289, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5304238355014688, | |
| "grad_norm": 0.4586597681045532, | |
| "learning_rate": 4.332151300236406e-07, | |
| "loss": 2.1033, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5337809483843894, | |
| "grad_norm": 0.4757809340953827, | |
| "learning_rate": 4.326241134751773e-07, | |
| "loss": 2.1708, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5371380612673101, | |
| "grad_norm": 0.45364031195640564, | |
| "learning_rate": 4.320330969267139e-07, | |
| "loss": 2.1473, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5404951741502309, | |
| "grad_norm": 0.45321136713027954, | |
| "learning_rate": 4.314420803782506e-07, | |
| "loss": 2.177, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5438522870331515, | |
| "grad_norm": 0.43466734886169434, | |
| "learning_rate": 4.308510638297872e-07, | |
| "loss": 2.1304, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5472093999160722, | |
| "grad_norm": 0.4303533732891083, | |
| "learning_rate": 4.302600472813239e-07, | |
| "loss": 2.0758, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5505665127989928, | |
| "grad_norm": 0.47530239820480347, | |
| "learning_rate": 4.2966903073286046e-07, | |
| "loss": 2.2194, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5539236256819136, | |
| "grad_norm": 0.4379255175590515, | |
| "learning_rate": 4.2907801418439714e-07, | |
| "loss": 2.1497, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5572807385648343, | |
| "grad_norm": 0.4771229922771454, | |
| "learning_rate": 4.2848699763593376e-07, | |
| "loss": 2.1491, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5606378514477549, | |
| "grad_norm": 0.4536450505256653, | |
| "learning_rate": 4.2789598108747044e-07, | |
| "loss": 2.2061, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5639949643306756, | |
| "grad_norm": 0.46324947476387024, | |
| "learning_rate": 4.2730496453900706e-07, | |
| "loss": 2.1859, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.5673520772135963, | |
| "grad_norm": 0.4493923485279083, | |
| "learning_rate": 4.2671394799054374e-07, | |
| "loss": 2.0956, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.570709190096517, | |
| "grad_norm": 0.4963778853416443, | |
| "learning_rate": 4.261229314420803e-07, | |
| "loss": 2.1433, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5740663029794377, | |
| "grad_norm": 0.5063489675521851, | |
| "learning_rate": 4.25531914893617e-07, | |
| "loss": 2.1887, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5774234158623583, | |
| "grad_norm": 0.4580891728401184, | |
| "learning_rate": 4.249408983451536e-07, | |
| "loss": 2.1164, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5807805287452791, | |
| "grad_norm": 0.4890580177307129, | |
| "learning_rate": 4.243498817966903e-07, | |
| "loss": 2.1647, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5841376416281997, | |
| "grad_norm": 0.45317739248275757, | |
| "learning_rate": 4.237588652482269e-07, | |
| "loss": 2.1837, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.5874947545111204, | |
| "grad_norm": 0.4900612533092499, | |
| "learning_rate": 4.231678486997636e-07, | |
| "loss": 2.1558, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5908518673940412, | |
| "grad_norm": 0.47292637825012207, | |
| "learning_rate": 4.2257683215130027e-07, | |
| "loss": 2.163, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.5942089802769618, | |
| "grad_norm": 0.4768417477607727, | |
| "learning_rate": 4.2198581560283684e-07, | |
| "loss": 2.1963, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.5975660931598825, | |
| "grad_norm": 0.4955364465713501, | |
| "learning_rate": 4.213947990543735e-07, | |
| "loss": 2.1402, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6009232060428031, | |
| "grad_norm": 0.46482154726982117, | |
| "learning_rate": 4.2080378250591014e-07, | |
| "loss": 2.238, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6042803189257239, | |
| "grad_norm": 0.47761717438697815, | |
| "learning_rate": 4.202127659574468e-07, | |
| "loss": 2.1089, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6076374318086446, | |
| "grad_norm": 0.48028987646102905, | |
| "learning_rate": 4.1962174940898344e-07, | |
| "loss": 2.1946, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6109945446915652, | |
| "grad_norm": 0.4602825939655304, | |
| "learning_rate": 4.190307328605201e-07, | |
| "loss": 2.1304, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.614351657574486, | |
| "grad_norm": 0.4691866338253021, | |
| "learning_rate": 4.184397163120567e-07, | |
| "loss": 2.1363, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6177087704574066, | |
| "grad_norm": 0.46271318197250366, | |
| "learning_rate": 4.1784869976359336e-07, | |
| "loss": 2.1805, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6210658833403273, | |
| "grad_norm": 0.48010194301605225, | |
| "learning_rate": 4.1725768321513e-07, | |
| "loss": 2.178, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.624422996223248, | |
| "grad_norm": 0.45885005593299866, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "loss": 2.1575, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6277801091061687, | |
| "grad_norm": 0.45524775981903076, | |
| "learning_rate": 4.160756501182033e-07, | |
| "loss": 2.1303, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6311372219890894, | |
| "grad_norm": 0.4570733606815338, | |
| "learning_rate": 4.1548463356973997e-07, | |
| "loss": 2.1628, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.6344943348720101, | |
| "grad_norm": 0.489170640707016, | |
| "learning_rate": 4.148936170212766e-07, | |
| "loss": 2.1663, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6378514477549307, | |
| "grad_norm": 0.47888293862342834, | |
| "learning_rate": 4.143026004728132e-07, | |
| "loss": 2.1347, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6412085606378515, | |
| "grad_norm": 0.4729193449020386, | |
| "learning_rate": 4.1371158392434984e-07, | |
| "loss": 2.1394, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6445656735207721, | |
| "grad_norm": 0.5049130320549011, | |
| "learning_rate": 4.131205673758865e-07, | |
| "loss": 2.2181, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6479227864036928, | |
| "grad_norm": 0.44132182002067566, | |
| "learning_rate": 4.1252955082742314e-07, | |
| "loss": 2.1427, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6512798992866136, | |
| "grad_norm": 0.49706417322158813, | |
| "learning_rate": 4.119385342789598e-07, | |
| "loss": 2.1993, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.6546370121695342, | |
| "grad_norm": 0.46416929364204407, | |
| "learning_rate": 4.1134751773049644e-07, | |
| "loss": 2.1108, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6579941250524549, | |
| "grad_norm": 0.4778405427932739, | |
| "learning_rate": 4.1075650118203306e-07, | |
| "loss": 2.1694, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.6613512379353755, | |
| "grad_norm": 0.46708041429519653, | |
| "learning_rate": 4.101654846335697e-07, | |
| "loss": 2.184, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.6647083508182963, | |
| "grad_norm": 0.48584261536598206, | |
| "learning_rate": 4.0957446808510637e-07, | |
| "loss": 2.1222, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.668065463701217, | |
| "grad_norm": 0.5111873745918274, | |
| "learning_rate": 4.08983451536643e-07, | |
| "loss": 2.1711, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.6714225765841376, | |
| "grad_norm": 0.4958716630935669, | |
| "learning_rate": 4.0839243498817967e-07, | |
| "loss": 2.1523, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6747796894670584, | |
| "grad_norm": 0.48708048462867737, | |
| "learning_rate": 4.078014184397163e-07, | |
| "loss": 2.1921, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.678136802349979, | |
| "grad_norm": 0.47986796498298645, | |
| "learning_rate": 4.0721040189125297e-07, | |
| "loss": 2.1619, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.6814939152328997, | |
| "grad_norm": 0.487250417470932, | |
| "learning_rate": 4.0661938534278954e-07, | |
| "loss": 2.2103, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.6848510281158204, | |
| "grad_norm": 0.5118921995162964, | |
| "learning_rate": 4.060283687943262e-07, | |
| "loss": 2.2549, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.6882081409987411, | |
| "grad_norm": 0.5187731981277466, | |
| "learning_rate": 4.0543735224586284e-07, | |
| "loss": 2.2144, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.6915652538816618, | |
| "grad_norm": 0.4841180145740509, | |
| "learning_rate": 4.048463356973995e-07, | |
| "loss": 2.1528, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.6949223667645824, | |
| "grad_norm": 0.47858700156211853, | |
| "learning_rate": 4.0425531914893614e-07, | |
| "loss": 2.1743, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.6982794796475031, | |
| "grad_norm": 0.47898271679878235, | |
| "learning_rate": 4.036643026004728e-07, | |
| "loss": 2.1268, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.7016365925304239, | |
| "grad_norm": 0.4743264615535736, | |
| "learning_rate": 4.0307328605200944e-07, | |
| "loss": 2.1992, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.7049937054133445, | |
| "grad_norm": 0.5258775353431702, | |
| "learning_rate": 4.0248226950354607e-07, | |
| "loss": 2.1288, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7083508182962652, | |
| "grad_norm": 0.4403035044670105, | |
| "learning_rate": 4.0189125295508274e-07, | |
| "loss": 2.1033, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7117079311791858, | |
| "grad_norm": 0.4601992666721344, | |
| "learning_rate": 4.0130023640661937e-07, | |
| "loss": 2.2051, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7150650440621066, | |
| "grad_norm": 0.48560434579849243, | |
| "learning_rate": 4.0070921985815604e-07, | |
| "loss": 2.1469, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7184221569450273, | |
| "grad_norm": 0.4823721945285797, | |
| "learning_rate": 4.0011820330969267e-07, | |
| "loss": 2.1988, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.7217792698279479, | |
| "grad_norm": 0.48195022344589233, | |
| "learning_rate": 3.995271867612293e-07, | |
| "loss": 2.1211, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7251363827108687, | |
| "grad_norm": 0.5148845314979553, | |
| "learning_rate": 3.989361702127659e-07, | |
| "loss": 2.1803, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.7284934955937893, | |
| "grad_norm": 0.4884459376335144, | |
| "learning_rate": 3.983451536643026e-07, | |
| "loss": 2.1332, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.73185060847671, | |
| "grad_norm": 0.5225220322608948, | |
| "learning_rate": 3.977541371158392e-07, | |
| "loss": 2.1582, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7352077213596308, | |
| "grad_norm": 0.4897938668727875, | |
| "learning_rate": 3.971631205673759e-07, | |
| "loss": 2.1322, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.7385648342425514, | |
| "grad_norm": 0.502916693687439, | |
| "learning_rate": 3.965721040189125e-07, | |
| "loss": 2.1518, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7419219471254721, | |
| "grad_norm": 0.4693153202533722, | |
| "learning_rate": 3.959810874704492e-07, | |
| "loss": 2.1207, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.7452790600083928, | |
| "grad_norm": 0.4866141676902771, | |
| "learning_rate": 3.9539007092198577e-07, | |
| "loss": 2.1424, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.7486361728913135, | |
| "grad_norm": 0.48267892003059387, | |
| "learning_rate": 3.9479905437352244e-07, | |
| "loss": 2.1664, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.7519932857742342, | |
| "grad_norm": 0.505587637424469, | |
| "learning_rate": 3.9420803782505907e-07, | |
| "loss": 2.1723, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.7553503986571548, | |
| "grad_norm": 0.47869905829429626, | |
| "learning_rate": 3.9361702127659574e-07, | |
| "loss": 2.0781, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7587075115400755, | |
| "grad_norm": 0.487474650144577, | |
| "learning_rate": 3.9302600472813237e-07, | |
| "loss": 2.1889, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.7620646244229963, | |
| "grad_norm": 0.5115759968757629, | |
| "learning_rate": 3.9243498817966904e-07, | |
| "loss": 2.2055, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.7654217373059169, | |
| "grad_norm": 0.4802757203578949, | |
| "learning_rate": 3.918439716312056e-07, | |
| "loss": 2.1542, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.7687788501888376, | |
| "grad_norm": 0.48687273263931274, | |
| "learning_rate": 3.912529550827423e-07, | |
| "loss": 2.2195, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.7721359630717582, | |
| "grad_norm": 0.5212287902832031, | |
| "learning_rate": 3.906619385342789e-07, | |
| "loss": 2.2386, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.775493075954679, | |
| "grad_norm": 0.4856519401073456, | |
| "learning_rate": 3.900709219858156e-07, | |
| "loss": 2.1322, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.7788501888375997, | |
| "grad_norm": 0.4821922183036804, | |
| "learning_rate": 3.894799054373522e-07, | |
| "loss": 2.1594, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.7822073017205203, | |
| "grad_norm": 0.46911802887916565, | |
| "learning_rate": 3.888888888888889e-07, | |
| "loss": 2.1279, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.7855644146034411, | |
| "grad_norm": 0.5064778923988342, | |
| "learning_rate": 3.8829787234042547e-07, | |
| "loss": 2.1294, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.7889215274863617, | |
| "grad_norm": 0.5024438500404358, | |
| "learning_rate": 3.8770685579196214e-07, | |
| "loss": 2.1321, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7922786403692824, | |
| "grad_norm": 0.5185412168502808, | |
| "learning_rate": 3.8711583924349877e-07, | |
| "loss": 2.13, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.7956357532522031, | |
| "grad_norm": 0.5049921274185181, | |
| "learning_rate": 3.8652482269503544e-07, | |
| "loss": 2.165, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.7989928661351238, | |
| "grad_norm": 0.5252367258071899, | |
| "learning_rate": 3.8593380614657207e-07, | |
| "loss": 2.0951, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.8023499790180445, | |
| "grad_norm": 0.5152316093444824, | |
| "learning_rate": 3.8534278959810874e-07, | |
| "loss": 2.1517, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.8057070919009651, | |
| "grad_norm": 0.4972199499607086, | |
| "learning_rate": 3.8475177304964537e-07, | |
| "loss": 2.1474, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8090642047838859, | |
| "grad_norm": 0.5103582143783569, | |
| "learning_rate": 3.84160756501182e-07, | |
| "loss": 2.1318, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8124213176668066, | |
| "grad_norm": 0.4988660216331482, | |
| "learning_rate": 3.8356973995271867e-07, | |
| "loss": 2.1228, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8157784305497272, | |
| "grad_norm": 0.5007835030555725, | |
| "learning_rate": 3.829787234042553e-07, | |
| "loss": 2.1176, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8191355434326479, | |
| "grad_norm": 0.4536113440990448, | |
| "learning_rate": 3.8238770685579197e-07, | |
| "loss": 2.1406, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.8224926563155686, | |
| "grad_norm": 0.5342024564743042, | |
| "learning_rate": 3.817966903073286e-07, | |
| "loss": 2.1462, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8258497691984893, | |
| "grad_norm": 0.48217201232910156, | |
| "learning_rate": 3.8120567375886527e-07, | |
| "loss": 2.1259, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.82920688208141, | |
| "grad_norm": 0.5227500200271606, | |
| "learning_rate": 3.8061465721040184e-07, | |
| "loss": 2.1516, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.8325639949643306, | |
| "grad_norm": 0.47303012013435364, | |
| "learning_rate": 3.800236406619385e-07, | |
| "loss": 2.1713, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.8359211078472514, | |
| "grad_norm": 0.512878954410553, | |
| "learning_rate": 3.7943262411347514e-07, | |
| "loss": 2.1799, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.839278220730172, | |
| "grad_norm": 0.5365780591964722, | |
| "learning_rate": 3.788416075650118e-07, | |
| "loss": 2.1795, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8426353336130927, | |
| "grad_norm": 0.5341731905937195, | |
| "learning_rate": 3.7825059101654844e-07, | |
| "loss": 2.2331, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.8459924464960135, | |
| "grad_norm": 0.4720432758331299, | |
| "learning_rate": 3.776595744680851e-07, | |
| "loss": 2.155, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.8493495593789341, | |
| "grad_norm": 0.5171768665313721, | |
| "learning_rate": 3.7706855791962175e-07, | |
| "loss": 2.1395, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.8527066722618548, | |
| "grad_norm": 0.5279157757759094, | |
| "learning_rate": 3.7647754137115837e-07, | |
| "loss": 2.1647, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.8560637851447755, | |
| "grad_norm": 0.5167645812034607, | |
| "learning_rate": 3.75886524822695e-07, | |
| "loss": 2.1915, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8594208980276962, | |
| "grad_norm": 0.4854820668697357, | |
| "learning_rate": 3.7529550827423167e-07, | |
| "loss": 2.1293, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.8627780109106169, | |
| "grad_norm": 0.5053945183753967, | |
| "learning_rate": 3.747044917257683e-07, | |
| "loss": 2.1776, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.8661351237935375, | |
| "grad_norm": 0.5340734720230103, | |
| "learning_rate": 3.7411347517730497e-07, | |
| "loss": 2.2158, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.8694922366764583, | |
| "grad_norm": 0.5089324116706848, | |
| "learning_rate": 3.735224586288416e-07, | |
| "loss": 2.1451, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.872849349559379, | |
| "grad_norm": 0.49475711584091187, | |
| "learning_rate": 3.729314420803782e-07, | |
| "loss": 2.1416, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8762064624422996, | |
| "grad_norm": 0.5191430449485779, | |
| "learning_rate": 3.7234042553191484e-07, | |
| "loss": 2.1815, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.8795635753252203, | |
| "grad_norm": 0.4857535660266876, | |
| "learning_rate": 3.717494089834515e-07, | |
| "loss": 2.1388, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.882920688208141, | |
| "grad_norm": 0.4946460425853729, | |
| "learning_rate": 3.7115839243498815e-07, | |
| "loss": 2.1562, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.8862778010910617, | |
| "grad_norm": 0.4693676233291626, | |
| "learning_rate": 3.705673758865248e-07, | |
| "loss": 2.1189, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.8896349139739824, | |
| "grad_norm": 0.5070816278457642, | |
| "learning_rate": 3.6997635933806145e-07, | |
| "loss": 2.1278, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.892992026856903, | |
| "grad_norm": 0.5286785960197449, | |
| "learning_rate": 3.693853427895981e-07, | |
| "loss": 2.1917, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.8963491397398238, | |
| "grad_norm": 0.48202502727508545, | |
| "learning_rate": 3.687943262411347e-07, | |
| "loss": 2.1224, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.8997062526227444, | |
| "grad_norm": 0.5092111825942993, | |
| "learning_rate": 3.6820330969267137e-07, | |
| "loss": 2.2141, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.9030633655056651, | |
| "grad_norm": 0.5308806300163269, | |
| "learning_rate": 3.67612293144208e-07, | |
| "loss": 2.151, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.9064204783885859, | |
| "grad_norm": 0.5302571058273315, | |
| "learning_rate": 3.6702127659574467e-07, | |
| "loss": 2.1899, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9097775912715065, | |
| "grad_norm": 0.489431768655777, | |
| "learning_rate": 3.664302600472813e-07, | |
| "loss": 2.1448, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.9131347041544272, | |
| "grad_norm": 0.47753775119781494, | |
| "learning_rate": 3.6583924349881797e-07, | |
| "loss": 2.1036, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.9164918170373478, | |
| "grad_norm": 0.49404028058052063, | |
| "learning_rate": 3.652482269503546e-07, | |
| "loss": 2.1422, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9198489299202686, | |
| "grad_norm": 0.5034516453742981, | |
| "learning_rate": 3.646572104018912e-07, | |
| "loss": 2.152, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.9232060428031893, | |
| "grad_norm": 0.5550661683082581, | |
| "learning_rate": 3.640661938534279e-07, | |
| "loss": 2.1861, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9265631556861099, | |
| "grad_norm": 0.4908338487148285, | |
| "learning_rate": 3.634751773049645e-07, | |
| "loss": 2.1026, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.9299202685690307, | |
| "grad_norm": 0.5155569911003113, | |
| "learning_rate": 3.628841607565012e-07, | |
| "loss": 2.1006, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.9332773814519513, | |
| "grad_norm": 0.5384230613708496, | |
| "learning_rate": 3.622931442080378e-07, | |
| "loss": 2.2128, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.936634494334872, | |
| "grad_norm": 0.5264031291007996, | |
| "learning_rate": 3.617021276595745e-07, | |
| "loss": 2.1531, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.9399916072177927, | |
| "grad_norm": 0.5026865601539612, | |
| "learning_rate": 3.6111111111111107e-07, | |
| "loss": 2.1594, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9433487201007134, | |
| "grad_norm": 0.4906868040561676, | |
| "learning_rate": 3.6052009456264775e-07, | |
| "loss": 2.1489, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.9467058329836341, | |
| "grad_norm": 0.5679292678833008, | |
| "learning_rate": 3.5992907801418437e-07, | |
| "loss": 2.1501, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.9500629458665547, | |
| "grad_norm": 0.49988269805908203, | |
| "learning_rate": 3.5933806146572105e-07, | |
| "loss": 2.1413, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.9534200587494754, | |
| "grad_norm": 0.4949737787246704, | |
| "learning_rate": 3.5874704491725767e-07, | |
| "loss": 2.188, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.9567771716323962, | |
| "grad_norm": 0.4845784902572632, | |
| "learning_rate": 3.5815602836879435e-07, | |
| "loss": 2.08, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9601342845153168, | |
| "grad_norm": 0.5556589365005493, | |
| "learning_rate": 3.575650118203309e-07, | |
| "loss": 2.1766, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.9634913973982375, | |
| "grad_norm": 0.5051941871643066, | |
| "learning_rate": 3.569739952718676e-07, | |
| "loss": 2.1159, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.9668485102811583, | |
| "grad_norm": 0.5166348814964294, | |
| "learning_rate": 3.563829787234042e-07, | |
| "loss": 2.2121, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.9702056231640789, | |
| "grad_norm": 0.5659390091896057, | |
| "learning_rate": 3.557919621749409e-07, | |
| "loss": 2.1162, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.9735627360469996, | |
| "grad_norm": 0.5001223683357239, | |
| "learning_rate": 3.552009456264775e-07, | |
| "loss": 2.1424, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9769198489299202, | |
| "grad_norm": 0.4793240427970886, | |
| "learning_rate": 3.546099290780142e-07, | |
| "loss": 2.136, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.980276961812841, | |
| "grad_norm": 0.5031545162200928, | |
| "learning_rate": 3.5401891252955077e-07, | |
| "loss": 2.1264, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.9836340746957617, | |
| "grad_norm": 0.526989221572876, | |
| "learning_rate": 3.5342789598108745e-07, | |
| "loss": 2.2329, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.9869911875786823, | |
| "grad_norm": 0.5093796253204346, | |
| "learning_rate": 3.5283687943262407e-07, | |
| "loss": 2.1477, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.990348300461603, | |
| "grad_norm": 0.5002118945121765, | |
| "learning_rate": 3.5224586288416075e-07, | |
| "loss": 2.195, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.9937054133445237, | |
| "grad_norm": 0.5272600650787354, | |
| "learning_rate": 3.5165484633569737e-07, | |
| "loss": 2.1664, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.9970625262274444, | |
| "grad_norm": 0.48927053809165955, | |
| "learning_rate": 3.5106382978723405e-07, | |
| "loss": 2.1497, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.9970625262274444, | |
| "eval_loss": 2.169156074523926, | |
| "eval_runtime": 360.7188, | |
| "eval_samples_per_second": 1.004, | |
| "eval_steps_per_second": 0.252, | |
| "step": 297 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 891, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.794451043460055e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |