{ "best_metric": 0.8874096274375916, "best_model_checkpoint": "/workspace/previous_works/MedBLIP/output/MedBLIP-0007/checkpoint-10000", "epoch": 3.0, "eval_steps": 10000, "global_step": 14319, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0031426775612822125, "grad_norm": 0.0, "learning_rate": 0.0, "loss": 6.4738, "step": 15 }, { "epoch": 0.006285355122564425, "grad_norm": 143.78610229492188, "learning_rate": 1.3966480446927375e-07, "loss": 6.4427, "step": 30 }, { "epoch": 0.009428032683846637, "grad_norm": 125.2808609008789, "learning_rate": 3.3519553072625703e-07, "loss": 6.1729, "step": 45 }, { "epoch": 0.01257071024512885, "grad_norm": 55.64238357543945, "learning_rate": 5.446927374301677e-07, "loss": 5.6704, "step": 60 }, { "epoch": 0.01571338780641106, "grad_norm": 58.784568786621094, "learning_rate": 7.402234636871509e-07, "loss": 4.6903, "step": 75 }, { "epoch": 0.018856065367693273, "grad_norm": 101.14881896972656, "learning_rate": 9.497206703910615e-07, "loss": 4.009, "step": 90 }, { "epoch": 0.02199874292897549, "grad_norm": 40.7690315246582, "learning_rate": 1.159217877094972e-06, "loss": 3.3713, "step": 105 }, { "epoch": 0.0251414204902577, "grad_norm": 22.23933219909668, "learning_rate": 1.3687150837988828e-06, "loss": 2.9785, "step": 120 }, { "epoch": 0.028284098051539912, "grad_norm": 16.25871467590332, "learning_rate": 1.5782122905027933e-06, "loss": 2.8118, "step": 135 }, { "epoch": 0.03142677561282212, "grad_norm": 23.991317749023438, "learning_rate": 1.787709497206704e-06, "loss": 2.6758, "step": 150 }, { "epoch": 0.034569453174104335, "grad_norm": 27.5322265625, "learning_rate": 1.9972067039106146e-06, "loss": 2.5263, "step": 165 }, { "epoch": 0.03771213073538655, "grad_norm": 18.619874954223633, "learning_rate": 2.2067039106145253e-06, "loss": 2.4554, "step": 180 }, { "epoch": 0.04085480829666876, "grad_norm": 17.445724487304688, "learning_rate": 2.416201117318436e-06, "loss": 2.3279, "step": 195 }, { "epoch": 0.04399748585795098, "grad_norm": 19.981834411621094, "learning_rate": 2.6256983240223464e-06, "loss": 2.2568, "step": 210 }, { "epoch": 0.04714016341923319, "grad_norm": 15.571565628051758, "learning_rate": 2.8351955307262576e-06, "loss": 2.1021, "step": 225 }, { "epoch": 0.0502828409805154, "grad_norm": 13.631041526794434, "learning_rate": 3.044692737430168e-06, "loss": 1.9724, "step": 240 }, { "epoch": 0.05342551854179761, "grad_norm": 10.806873321533203, "learning_rate": 3.2541899441340786e-06, "loss": 1.8049, "step": 255 }, { "epoch": 0.056568196103079824, "grad_norm": 14.863541603088379, "learning_rate": 3.4636871508379893e-06, "loss": 1.7358, "step": 270 }, { "epoch": 0.059710873664362035, "grad_norm": 12.929265022277832, "learning_rate": 3.6731843575418996e-06, "loss": 1.6953, "step": 285 }, { "epoch": 0.06285355122564425, "grad_norm": 12.591845512390137, "learning_rate": 3.88268156424581e-06, "loss": 1.6874, "step": 300 }, { "epoch": 0.06599622878692646, "grad_norm": 13.552529335021973, "learning_rate": 4.0921787709497215e-06, "loss": 1.6485, "step": 315 }, { "epoch": 0.06913890634820867, "grad_norm": 9.15485668182373, "learning_rate": 4.301675977653632e-06, "loss": 1.6044, "step": 330 }, { "epoch": 0.07228158390949088, "grad_norm": 7.887852668762207, "learning_rate": 4.511173184357542e-06, "loss": 1.4664, "step": 345 }, { "epoch": 0.0754242614707731, "grad_norm": 12.257076263427734, "learning_rate": 4.7206703910614525e-06, "loss": 1.5338, "step": 360 }, { "epoch": 0.0785669390320553, "grad_norm": 8.181825637817383, "learning_rate": 4.930167597765364e-06, "loss": 1.4954, "step": 375 }, { "epoch": 0.08170961659333752, "grad_norm": 9.838497161865234, "learning_rate": 5.139664804469274e-06, "loss": 1.4217, "step": 390 }, { "epoch": 0.08485229415461974, "grad_norm": 15.192863464355469, "learning_rate": 5.349162011173184e-06, "loss": 1.4725, "step": 405 }, { "epoch": 0.08799497171590195, "grad_norm": 7.013632297515869, "learning_rate": 5.558659217877096e-06, "loss": 1.3989, "step": 420 }, { "epoch": 0.09113764927718417, "grad_norm": 8.598616600036621, "learning_rate": 5.768156424581007e-06, "loss": 1.4828, "step": 435 }, { "epoch": 0.09428032683846638, "grad_norm": 5.625289440155029, "learning_rate": 5.977653631284917e-06, "loss": 1.4513, "step": 450 }, { "epoch": 0.09742300439974859, "grad_norm": 7.280072212219238, "learning_rate": 6.187150837988828e-06, "loss": 1.4092, "step": 465 }, { "epoch": 0.1005656819610308, "grad_norm": 7.608973979949951, "learning_rate": 6.396648044692738e-06, "loss": 1.4124, "step": 480 }, { "epoch": 0.10370835952231301, "grad_norm": 8.430741310119629, "learning_rate": 6.606145251396649e-06, "loss": 1.4466, "step": 495 }, { "epoch": 0.10685103708359522, "grad_norm": 8.642643928527832, "learning_rate": 6.815642458100559e-06, "loss": 1.3558, "step": 510 }, { "epoch": 0.10999371464487744, "grad_norm": 6.632693290710449, "learning_rate": 7.02513966480447e-06, "loss": 1.4129, "step": 525 }, { "epoch": 0.11313639220615965, "grad_norm": 12.695171356201172, "learning_rate": 7.2346368715083805e-06, "loss": 1.3444, "step": 540 }, { "epoch": 0.11627906976744186, "grad_norm": 5.393228054046631, "learning_rate": 7.444134078212291e-06, "loss": 1.3485, "step": 555 }, { "epoch": 0.11942174732872407, "grad_norm": 8.173951148986816, "learning_rate": 7.653631284916202e-06, "loss": 1.4175, "step": 570 }, { "epoch": 0.12256442489000628, "grad_norm": 14.139352798461914, "learning_rate": 7.863128491620112e-06, "loss": 1.3797, "step": 585 }, { "epoch": 0.1257071024512885, "grad_norm": 4.9999613761901855, "learning_rate": 8.072625698324023e-06, "loss": 1.3716, "step": 600 }, { "epoch": 0.12884978001257072, "grad_norm": 4.344189643859863, "learning_rate": 8.282122905027935e-06, "loss": 1.3011, "step": 615 }, { "epoch": 0.13199245757385292, "grad_norm": 11.033843040466309, "learning_rate": 8.491620111731845e-06, "loss": 1.3672, "step": 630 }, { "epoch": 0.13513513513513514, "grad_norm": 4.3888068199157715, "learning_rate": 8.701117318435755e-06, "loss": 1.2773, "step": 645 }, { "epoch": 0.13827781269641734, "grad_norm": 4.085406303405762, "learning_rate": 8.910614525139666e-06, "loss": 1.3057, "step": 660 }, { "epoch": 0.14142049025769957, "grad_norm": 11.730964660644531, "learning_rate": 9.120111731843576e-06, "loss": 1.3752, "step": 675 }, { "epoch": 0.14456316781898176, "grad_norm": 4.986327648162842, "learning_rate": 9.329608938547486e-06, "loss": 1.3055, "step": 690 }, { "epoch": 0.147705845380264, "grad_norm": 5.643145561218262, "learning_rate": 9.539106145251398e-06, "loss": 1.3608, "step": 705 }, { "epoch": 0.1508485229415462, "grad_norm": 12.661754608154297, "learning_rate": 9.748603351955308e-06, "loss": 1.4083, "step": 720 }, { "epoch": 0.1539912005028284, "grad_norm": 6.018293857574463, "learning_rate": 9.958100558659219e-06, "loss": 1.2323, "step": 735 }, { "epoch": 0.1571338780641106, "grad_norm": 4.150609493255615, "learning_rate": 1.0167597765363129e-05, "loss": 1.3068, "step": 750 }, { "epoch": 0.16027655562539284, "grad_norm": 4.229065895080566, "learning_rate": 1.037709497206704e-05, "loss": 1.3257, "step": 765 }, { "epoch": 0.16341923318667503, "grad_norm": 9.254191398620605, "learning_rate": 1.0586592178770951e-05, "loss": 1.2674, "step": 780 }, { "epoch": 0.16656191074795726, "grad_norm": 3.52528977394104, "learning_rate": 1.0796089385474862e-05, "loss": 1.271, "step": 795 }, { "epoch": 0.16970458830923948, "grad_norm": 6.313427448272705, "learning_rate": 1.1005586592178772e-05, "loss": 1.2948, "step": 810 }, { "epoch": 0.17284726587052168, "grad_norm": 3.691848039627075, "learning_rate": 1.1215083798882682e-05, "loss": 1.292, "step": 825 }, { "epoch": 0.1759899434318039, "grad_norm": 3.9371867179870605, "learning_rate": 1.1424581005586593e-05, "loss": 1.3135, "step": 840 }, { "epoch": 0.1791326209930861, "grad_norm": 3.455486297607422, "learning_rate": 1.1634078212290503e-05, "loss": 1.2495, "step": 855 }, { "epoch": 0.18227529855436833, "grad_norm": 10.44939136505127, "learning_rate": 1.1843575418994415e-05, "loss": 1.356, "step": 870 }, { "epoch": 0.18541797611565053, "grad_norm": 3.6111254692077637, "learning_rate": 1.2053072625698325e-05, "loss": 1.3196, "step": 885 }, { "epoch": 0.18856065367693275, "grad_norm": 3.531568765640259, "learning_rate": 1.2262569832402236e-05, "loss": 1.2307, "step": 900 }, { "epoch": 0.19170333123821495, "grad_norm": 3.490081310272217, "learning_rate": 1.2472067039106146e-05, "loss": 1.2445, "step": 915 }, { "epoch": 0.19484600879949718, "grad_norm": 5.011629581451416, "learning_rate": 1.2681564245810056e-05, "loss": 1.3192, "step": 930 }, { "epoch": 0.19798868636077938, "grad_norm": 7.499322414398193, "learning_rate": 1.2891061452513967e-05, "loss": 1.2827, "step": 945 }, { "epoch": 0.2011313639220616, "grad_norm": 3.792930841445923, "learning_rate": 1.3100558659217879e-05, "loss": 1.336, "step": 960 }, { "epoch": 0.2042740414833438, "grad_norm": 4.43565559387207, "learning_rate": 1.3310055865921789e-05, "loss": 1.3371, "step": 975 }, { "epoch": 0.20741671904462602, "grad_norm": 4.2933173179626465, "learning_rate": 1.3519553072625699e-05, "loss": 1.3118, "step": 990 }, { "epoch": 0.21055939660590822, "grad_norm": 3.68387770652771, "learning_rate": 1.372905027932961e-05, "loss": 1.3975, "step": 1005 }, { "epoch": 0.21370207416719045, "grad_norm": 5.9917707443237305, "learning_rate": 1.393854748603352e-05, "loss": 1.2626, "step": 1020 }, { "epoch": 0.21684475172847265, "grad_norm": 3.1892921924591064, "learning_rate": 1.414804469273743e-05, "loss": 1.2781, "step": 1035 }, { "epoch": 0.21998742928975487, "grad_norm": 3.4965710639953613, "learning_rate": 1.4357541899441342e-05, "loss": 1.2557, "step": 1050 }, { "epoch": 0.2231301068510371, "grad_norm": 7.907847881317139, "learning_rate": 1.4567039106145252e-05, "loss": 1.2341, "step": 1065 }, { "epoch": 0.2262727844123193, "grad_norm": 3.6374635696411133, "learning_rate": 1.4776536312849163e-05, "loss": 1.3225, "step": 1080 }, { "epoch": 0.22941546197360152, "grad_norm": 6.52725887298584, "learning_rate": 1.4986033519553073e-05, "loss": 1.2137, "step": 1095 }, { "epoch": 0.23255813953488372, "grad_norm": 4.659281253814697, "learning_rate": 1.5195530726256983e-05, "loss": 1.3373, "step": 1110 }, { "epoch": 0.23570081709616594, "grad_norm": 3.1662540435791016, "learning_rate": 1.5405027932960895e-05, "loss": 1.3348, "step": 1125 }, { "epoch": 0.23884349465744814, "grad_norm": 3.8003756999969482, "learning_rate": 1.5614525139664806e-05, "loss": 1.2948, "step": 1140 }, { "epoch": 0.24198617221873037, "grad_norm": 3.0090460777282715, "learning_rate": 1.5824022346368716e-05, "loss": 1.2724, "step": 1155 }, { "epoch": 0.24512884978001256, "grad_norm": 3.1680350303649902, "learning_rate": 1.6033519553072626e-05, "loss": 1.2402, "step": 1170 }, { "epoch": 0.2482715273412948, "grad_norm": 3.944425106048584, "learning_rate": 1.6243016759776537e-05, "loss": 1.303, "step": 1185 }, { "epoch": 0.251414204902577, "grad_norm": 8.234729766845703, "learning_rate": 1.6452513966480447e-05, "loss": 1.2196, "step": 1200 }, { "epoch": 0.2545568824638592, "grad_norm": 6.725916862487793, "learning_rate": 1.6662011173184357e-05, "loss": 1.3213, "step": 1215 }, { "epoch": 0.25769956002514144, "grad_norm": 3.2696242332458496, "learning_rate": 1.687150837988827e-05, "loss": 1.2611, "step": 1230 }, { "epoch": 0.2608422375864236, "grad_norm": 3.0667693614959717, "learning_rate": 1.708100558659218e-05, "loss": 1.22, "step": 1245 }, { "epoch": 0.26398491514770583, "grad_norm": 3.5260438919067383, "learning_rate": 1.729050279329609e-05, "loss": 1.3142, "step": 1260 }, { "epoch": 0.26712759270898806, "grad_norm": 3.51269793510437, "learning_rate": 1.7500000000000002e-05, "loss": 1.2031, "step": 1275 }, { "epoch": 0.2702702702702703, "grad_norm": 2.9083192348480225, "learning_rate": 1.7709497206703912e-05, "loss": 1.1711, "step": 1290 }, { "epoch": 0.27341294783155246, "grad_norm": 2.915454626083374, "learning_rate": 1.7918994413407822e-05, "loss": 1.2452, "step": 1305 }, { "epoch": 0.2765556253928347, "grad_norm": 4.736166954040527, "learning_rate": 1.8128491620111733e-05, "loss": 1.2251, "step": 1320 }, { "epoch": 0.2796983029541169, "grad_norm": 2.881985664367676, "learning_rate": 1.8337988826815643e-05, "loss": 1.2541, "step": 1335 }, { "epoch": 0.28284098051539913, "grad_norm": 4.29525899887085, "learning_rate": 1.8547486033519553e-05, "loss": 1.2336, "step": 1350 }, { "epoch": 0.28598365807668136, "grad_norm": 3.046929359436035, "learning_rate": 1.8756983240223464e-05, "loss": 1.2576, "step": 1365 }, { "epoch": 0.2891263356379635, "grad_norm": 4.023289680480957, "learning_rate": 1.8966480446927374e-05, "loss": 1.2371, "step": 1380 }, { "epoch": 0.29226901319924575, "grad_norm": 6.221644878387451, "learning_rate": 1.9175977653631284e-05, "loss": 1.1985, "step": 1395 }, { "epoch": 0.295411690760528, "grad_norm": 6.021862030029297, "learning_rate": 1.9385474860335195e-05, "loss": 1.2741, "step": 1410 }, { "epoch": 0.2985543683218102, "grad_norm": 6.475122451782227, "learning_rate": 1.959497206703911e-05, "loss": 1.2405, "step": 1425 }, { "epoch": 0.3016970458830924, "grad_norm": 6.075967788696289, "learning_rate": 1.980446927374302e-05, "loss": 1.1969, "step": 1440 }, { "epoch": 0.3048397234443746, "grad_norm": 2.591923713684082, "learning_rate": 1.9999999702856602e-05, "loss": 1.1979, "step": 1455 }, { "epoch": 0.3079824010056568, "grad_norm": 6.351642608642578, "learning_rate": 1.9999923931385287e-05, "loss": 1.2887, "step": 1470 }, { "epoch": 0.31112507856693905, "grad_norm": 2.5136075019836426, "learning_rate": 1.9999714446548786e-05, "loss": 1.2974, "step": 1485 }, { "epoch": 0.3142677561282212, "grad_norm": 3.683798313140869, "learning_rate": 1.9999371251148214e-05, "loss": 1.2073, "step": 1500 }, { "epoch": 0.31741043368950345, "grad_norm": 6.616783142089844, "learning_rate": 1.9998894349772585e-05, "loss": 1.224, "step": 1515 }, { "epoch": 0.32055311125078567, "grad_norm": 3.386265277862549, "learning_rate": 1.9998283748798757e-05, "loss": 1.2237, "step": 1530 }, { "epoch": 0.3236957888120679, "grad_norm": 3.2038071155548096, "learning_rate": 1.9997539456391343e-05, "loss": 1.2805, "step": 1545 }, { "epoch": 0.32683846637335007, "grad_norm": 5.434456825256348, "learning_rate": 1.9996661482502602e-05, "loss": 1.295, "step": 1560 }, { "epoch": 0.3299811439346323, "grad_norm": 3.0071728229522705, "learning_rate": 1.999564983887231e-05, "loss": 1.2207, "step": 1575 }, { "epoch": 0.3331238214959145, "grad_norm": 4.010280609130859, "learning_rate": 1.9994504539027594e-05, "loss": 1.2005, "step": 1590 }, { "epoch": 0.33626649905719674, "grad_norm": 4.2148237228393555, "learning_rate": 1.9993225598282768e-05, "loss": 1.1608, "step": 1605 }, { "epoch": 0.33940917661847897, "grad_norm": 5.541611194610596, "learning_rate": 1.9991813033739097e-05, "loss": 1.1649, "step": 1620 }, { "epoch": 0.34255185417976114, "grad_norm": 5.37580680847168, "learning_rate": 1.9990266864284607e-05, "loss": 1.2087, "step": 1635 }, { "epoch": 0.34569453174104336, "grad_norm": 3.72402286529541, "learning_rate": 1.998858711059381e-05, "loss": 1.2606, "step": 1650 }, { "epoch": 0.3488372093023256, "grad_norm": 3.596006393432617, "learning_rate": 1.9986773795127425e-05, "loss": 1.2509, "step": 1665 }, { "epoch": 0.3519798868636078, "grad_norm": 2.3336808681488037, "learning_rate": 1.9984826942132087e-05, "loss": 1.2389, "step": 1680 }, { "epoch": 0.35512256442489, "grad_norm": 2.766104221343994, "learning_rate": 1.9982746577640024e-05, "loss": 1.2334, "step": 1695 }, { "epoch": 0.3582652419861722, "grad_norm": 2.6721198558807373, "learning_rate": 1.9980532729468698e-05, "loss": 1.2069, "step": 1710 }, { "epoch": 0.36140791954745444, "grad_norm": 2.4807119369506836, "learning_rate": 1.9978185427220447e-05, "loss": 1.264, "step": 1725 }, { "epoch": 0.36455059710873666, "grad_norm": 2.6071507930755615, "learning_rate": 1.9975704702282074e-05, "loss": 1.2745, "step": 1740 }, { "epoch": 0.36769327467001883, "grad_norm": 2.5577876567840576, "learning_rate": 1.9973090587824443e-05, "loss": 1.2244, "step": 1755 }, { "epoch": 0.37083595223130106, "grad_norm": 5.360003471374512, "learning_rate": 1.997034311880202e-05, "loss": 1.1723, "step": 1770 }, { "epoch": 0.3739786297925833, "grad_norm": 2.606682777404785, "learning_rate": 1.996746233195242e-05, "loss": 1.2349, "step": 1785 }, { "epoch": 0.3771213073538655, "grad_norm": 2.898453712463379, "learning_rate": 1.9964448265795903e-05, "loss": 1.2428, "step": 1800 }, { "epoch": 0.3802639849151477, "grad_norm": 2.8664214611053467, "learning_rate": 1.9961300960634876e-05, "loss": 1.2338, "step": 1815 }, { "epoch": 0.3834066624764299, "grad_norm": 3.170452117919922, "learning_rate": 1.9958020458553326e-05, "loss": 1.1729, "step": 1830 }, { "epoch": 0.38654934003771213, "grad_norm": 2.6928646564483643, "learning_rate": 1.995460680341629e-05, "loss": 1.2791, "step": 1845 }, { "epoch": 0.38969201759899436, "grad_norm": 5.1588287353515625, "learning_rate": 1.9951060040869245e-05, "loss": 1.1712, "step": 1860 }, { "epoch": 0.3928346951602766, "grad_norm": 4.63333797454834, "learning_rate": 1.994738021833751e-05, "loss": 1.285, "step": 1875 }, { "epoch": 0.39597737272155875, "grad_norm": 2.860305070877075, "learning_rate": 1.9943567385025605e-05, "loss": 1.1853, "step": 1890 }, { "epoch": 0.399120050282841, "grad_norm": 3.4676976203918457, "learning_rate": 1.9939621591916587e-05, "loss": 1.2164, "step": 1905 }, { "epoch": 0.4022627278441232, "grad_norm": 4.743997573852539, "learning_rate": 1.9935542891771392e-05, "loss": 1.2182, "step": 1920 }, { "epoch": 0.40540540540540543, "grad_norm": 4.906563758850098, "learning_rate": 1.9931331339128107e-05, "loss": 1.1844, "step": 1935 }, { "epoch": 0.4085480829666876, "grad_norm": 2.863513946533203, "learning_rate": 1.9926986990301244e-05, "loss": 1.2124, "step": 1950 }, { "epoch": 0.4116907605279698, "grad_norm": 2.1002469062805176, "learning_rate": 1.9922509903381e-05, "loss": 1.1589, "step": 1965 }, { "epoch": 0.41483343808925205, "grad_norm": 5.061679840087891, "learning_rate": 1.991790013823246e-05, "loss": 1.2208, "step": 1980 }, { "epoch": 0.4179761156505343, "grad_norm": 5.046812534332275, "learning_rate": 1.9913157756494822e-05, "loss": 1.2122, "step": 1995 }, { "epoch": 0.42111879321181644, "grad_norm": 2.990056037902832, "learning_rate": 1.9908282821580544e-05, "loss": 1.2296, "step": 2010 }, { "epoch": 0.42426147077309867, "grad_norm": 2.477032423019409, "learning_rate": 1.9903275398674523e-05, "loss": 1.2259, "step": 2025 }, { "epoch": 0.4274041483343809, "grad_norm": 2.7786192893981934, "learning_rate": 1.9898135554733198e-05, "loss": 1.2062, "step": 2040 }, { "epoch": 0.4305468258956631, "grad_norm": 2.325221061706543, "learning_rate": 1.989286335848368e-05, "loss": 1.2676, "step": 2055 }, { "epoch": 0.4336895034569453, "grad_norm": 2.26428484916687, "learning_rate": 1.988745888042281e-05, "loss": 1.1832, "step": 2070 }, { "epoch": 0.4368321810182275, "grad_norm": 2.361360788345337, "learning_rate": 1.988192219281624e-05, "loss": 1.1814, "step": 2085 }, { "epoch": 0.43997485857950974, "grad_norm": 2.2548787593841553, "learning_rate": 1.987625336969744e-05, "loss": 1.1371, "step": 2100 }, { "epoch": 0.44311753614079197, "grad_norm": 2.9371745586395264, "learning_rate": 1.9870452486866733e-05, "loss": 1.1185, "step": 2115 }, { "epoch": 0.4462602137020742, "grad_norm": 2.250372886657715, "learning_rate": 1.9864519621890263e-05, "loss": 1.1919, "step": 2130 }, { "epoch": 0.44940289126335636, "grad_norm": 2.822431802749634, "learning_rate": 1.9858454854098968e-05, "loss": 1.1479, "step": 2145 }, { "epoch": 0.4525455688246386, "grad_norm": 5.055064678192139, "learning_rate": 1.985225826458751e-05, "loss": 1.2237, "step": 2160 }, { "epoch": 0.4556882463859208, "grad_norm": 2.3341803550720215, "learning_rate": 1.9845929936213216e-05, "loss": 1.1988, "step": 2175 }, { "epoch": 0.45883092394720304, "grad_norm": 2.5791430473327637, "learning_rate": 1.9839469953594927e-05, "loss": 1.19, "step": 2190 }, { "epoch": 0.4619736015084852, "grad_norm": 3.167426824569702, "learning_rate": 1.9832878403111908e-05, "loss": 1.1852, "step": 2205 }, { "epoch": 0.46511627906976744, "grad_norm": 4.721558570861816, "learning_rate": 1.9826155372902667e-05, "loss": 1.198, "step": 2220 }, { "epoch": 0.46825895663104966, "grad_norm": 2.5482370853424072, "learning_rate": 1.9819300952863783e-05, "loss": 1.1782, "step": 2235 }, { "epoch": 0.4714016341923319, "grad_norm": 3.1147351264953613, "learning_rate": 1.9812315234648714e-05, "loss": 1.2138, "step": 2250 }, { "epoch": 0.47454431175361406, "grad_norm": 4.730167865753174, "learning_rate": 1.980519831166656e-05, "loss": 1.1622, "step": 2265 }, { "epoch": 0.4776869893148963, "grad_norm": 1.9187180995941162, "learning_rate": 1.979795027908082e-05, "loss": 1.1624, "step": 2280 }, { "epoch": 0.4808296668761785, "grad_norm": 2.371293067932129, "learning_rate": 1.9790571233808108e-05, "loss": 1.1823, "step": 2295 }, { "epoch": 0.48397234443746073, "grad_norm": 2.103717803955078, "learning_rate": 1.978306127451687e-05, "loss": 1.2298, "step": 2310 }, { "epoch": 0.4871150219987429, "grad_norm": 3.5898678302764893, "learning_rate": 1.9775420501626073e-05, "loss": 1.2612, "step": 2325 }, { "epoch": 0.49025769956002513, "grad_norm": 2.680027723312378, "learning_rate": 1.976764901730383e-05, "loss": 1.2161, "step": 2340 }, { "epoch": 0.49340037712130735, "grad_norm": 2.514291286468506, "learning_rate": 1.9759746925466073e-05, "loss": 1.1772, "step": 2355 }, { "epoch": 0.4965430546825896, "grad_norm": 2.453972816467285, "learning_rate": 1.9751714331775125e-05, "loss": 1.1982, "step": 2370 }, { "epoch": 0.4996857322438718, "grad_norm": 6.017053604125977, "learning_rate": 1.9743551343638324e-05, "loss": 1.181, "step": 2385 }, { "epoch": 0.502828409805154, "grad_norm": 2.46274733543396, "learning_rate": 1.973525807020656e-05, "loss": 1.1537, "step": 2400 }, { "epoch": 0.5059710873664363, "grad_norm": 2.6295454502105713, "learning_rate": 1.972683462237282e-05, "loss": 1.1961, "step": 2415 }, { "epoch": 0.5091137649277184, "grad_norm": 5.519942760467529, "learning_rate": 1.9718281112770723e-05, "loss": 1.1998, "step": 2430 }, { "epoch": 0.5122564424890006, "grad_norm": 2.5744876861572266, "learning_rate": 1.9709597655772993e-05, "loss": 1.1511, "step": 2445 }, { "epoch": 0.5153991200502829, "grad_norm": 2.621332883834839, "learning_rate": 1.970078436748994e-05, "loss": 1.1398, "step": 2460 }, { "epoch": 0.518541797611565, "grad_norm": 5.679701328277588, "learning_rate": 1.9691841365767904e-05, "loss": 1.1899, "step": 2475 }, { "epoch": 0.5216844751728472, "grad_norm": 2.662759780883789, "learning_rate": 1.9682768770187675e-05, "loss": 1.1442, "step": 2490 }, { "epoch": 0.5248271527341295, "grad_norm": 2.7816829681396484, "learning_rate": 1.967356670206291e-05, "loss": 1.2164, "step": 2505 }, { "epoch": 0.5279698302954117, "grad_norm": 5.280804634094238, "learning_rate": 1.9664235284438495e-05, "loss": 1.186, "step": 2520 }, { "epoch": 0.531112507856694, "grad_norm": 4.3773112297058105, "learning_rate": 1.96547746420889e-05, "loss": 1.2107, "step": 2535 }, { "epoch": 0.5342551854179761, "grad_norm": 2.773582696914673, "learning_rate": 1.9645184901516525e-05, "loss": 1.2216, "step": 2550 }, { "epoch": 0.5373978629792583, "grad_norm": 2.643582820892334, "learning_rate": 1.9635466190949987e-05, "loss": 1.1888, "step": 2565 }, { "epoch": 0.5405405405405406, "grad_norm": 2.277792453765869, "learning_rate": 1.962561864034243e-05, "loss": 1.225, "step": 2580 }, { "epoch": 0.5436832181018227, "grad_norm": 2.580367088317871, "learning_rate": 1.9615642381369765e-05, "loss": 1.2064, "step": 2595 }, { "epoch": 0.5468258956631049, "grad_norm": 4.304893493652344, "learning_rate": 1.9605537547428922e-05, "loss": 1.1652, "step": 2610 }, { "epoch": 0.5499685732243872, "grad_norm": 2.5215561389923096, "learning_rate": 1.959530427363607e-05, "loss": 1.1909, "step": 2625 }, { "epoch": 0.5531112507856694, "grad_norm": 3.4145307540893555, "learning_rate": 1.958494269682479e-05, "loss": 1.1211, "step": 2640 }, { "epoch": 0.5562539283469516, "grad_norm": 4.1698079109191895, "learning_rate": 1.9574452955544274e-05, "loss": 1.1235, "step": 2655 }, { "epoch": 0.5593966059082338, "grad_norm": 4.188419818878174, "learning_rate": 1.9563835190057442e-05, "loss": 1.1925, "step": 2670 }, { "epoch": 0.562539283469516, "grad_norm": 4.869725704193115, "learning_rate": 1.9553089542339093e-05, "loss": 1.172, "step": 2685 }, { "epoch": 0.5656819610307983, "grad_norm": 4.675424575805664, "learning_rate": 1.9542216156073994e-05, "loss": 1.218, "step": 2700 }, { "epoch": 0.5688246385920804, "grad_norm": 2.655273914337158, "learning_rate": 1.9531215176654953e-05, "loss": 1.128, "step": 2715 }, { "epoch": 0.5719673161533627, "grad_norm": 3.345499277114868, "learning_rate": 1.9520086751180894e-05, "loss": 1.2236, "step": 2730 }, { "epoch": 0.5751099937146449, "grad_norm": 2.5058796405792236, "learning_rate": 1.9508831028454862e-05, "loss": 1.1895, "step": 2745 }, { "epoch": 0.578252671275927, "grad_norm": 2.2655258178710938, "learning_rate": 1.9497448158982062e-05, "loss": 1.1486, "step": 2760 }, { "epoch": 0.5813953488372093, "grad_norm": 4.1772894859313965, "learning_rate": 1.948593829496783e-05, "loss": 1.1616, "step": 2775 }, { "epoch": 0.5845380263984915, "grad_norm": 2.290619373321533, "learning_rate": 1.9474301590315605e-05, "loss": 1.1508, "step": 2790 }, { "epoch": 0.5876807039597737, "grad_norm": 2.4885716438293457, "learning_rate": 1.9462538200624855e-05, "loss": 1.1727, "step": 2805 }, { "epoch": 0.590823381521056, "grad_norm": 2.564606189727783, "learning_rate": 1.9450648283189035e-05, "loss": 1.2178, "step": 2820 }, { "epoch": 0.5939660590823381, "grad_norm": 2.756286859512329, "learning_rate": 1.9438631996993424e-05, "loss": 1.2143, "step": 2835 }, { "epoch": 0.5971087366436204, "grad_norm": 4.876669883728027, "learning_rate": 1.9426489502713062e-05, "loss": 1.1908, "step": 2850 }, { "epoch": 0.6002514142049026, "grad_norm": 1.9889987707138062, "learning_rate": 1.9414220962710556e-05, "loss": 1.1404, "step": 2865 }, { "epoch": 0.6033940917661847, "grad_norm": 2.6921072006225586, "learning_rate": 1.9401826541033935e-05, "loss": 1.2159, "step": 2880 }, { "epoch": 0.606536769327467, "grad_norm": 1.9745949506759644, "learning_rate": 1.9389306403414438e-05, "loss": 1.1828, "step": 2895 }, { "epoch": 0.6096794468887492, "grad_norm": 2.012610912322998, "learning_rate": 1.9376660717264316e-05, "loss": 1.1413, "step": 2910 }, { "epoch": 0.6128221244500315, "grad_norm": 3.3427295684814453, "learning_rate": 1.9363889651674575e-05, "loss": 1.1888, "step": 2925 }, { "epoch": 0.6159648020113137, "grad_norm": 3.9471700191497803, "learning_rate": 1.935099337741274e-05, "loss": 1.1882, "step": 2940 }, { "epoch": 0.6191074795725958, "grad_norm": 2.1639480590820312, "learning_rate": 1.9337972066920533e-05, "loss": 1.1518, "step": 2955 }, { "epoch": 0.6222501571338781, "grad_norm": 2.18038272857666, "learning_rate": 1.9324825894311613e-05, "loss": 1.1424, "step": 2970 }, { "epoch": 0.6253928346951603, "grad_norm": 2.4745776653289795, "learning_rate": 1.9311555035369206e-05, "loss": 1.2065, "step": 2985 }, { "epoch": 0.6285355122564424, "grad_norm": 2.4123799800872803, "learning_rate": 1.9298159667543795e-05, "loss": 1.1677, "step": 3000 }, { "epoch": 0.6316781898177247, "grad_norm": 2.8787992000579834, "learning_rate": 1.92846399699507e-05, "loss": 1.2035, "step": 3015 }, { "epoch": 0.6348208673790069, "grad_norm": 2.3905246257781982, "learning_rate": 1.927099612336773e-05, "loss": 1.1933, "step": 3030 }, { "epoch": 0.6379635449402892, "grad_norm": 3.8858561515808105, "learning_rate": 1.9257228310232738e-05, "loss": 1.1913, "step": 3045 }, { "epoch": 0.6411062225015713, "grad_norm": 3.567950963973999, "learning_rate": 1.924333671464118e-05, "loss": 1.2417, "step": 3060 }, { "epoch": 0.6442489000628535, "grad_norm": 1.8170944452285767, "learning_rate": 1.9229321522343677e-05, "loss": 1.1476, "step": 3075 }, { "epoch": 0.6473915776241358, "grad_norm": 3.044593095779419, "learning_rate": 1.9215182920743502e-05, "loss": 1.1939, "step": 3090 }, { "epoch": 0.650534255185418, "grad_norm": 3.13097882270813, "learning_rate": 1.9200921098894094e-05, "loss": 1.2007, "step": 3105 }, { "epoch": 0.6536769327467001, "grad_norm": 2.1468851566314697, "learning_rate": 1.918653624749652e-05, "loss": 1.1587, "step": 3120 }, { "epoch": 0.6568196103079824, "grad_norm": 2.9996323585510254, "learning_rate": 1.9172028558896935e-05, "loss": 1.1762, "step": 3135 }, { "epoch": 0.6599622878692646, "grad_norm": 2.5897107124328613, "learning_rate": 1.9157398227084002e-05, "loss": 1.1334, "step": 3150 }, { "epoch": 0.6631049654305469, "grad_norm": 3.80074143409729, "learning_rate": 1.91426454476863e-05, "loss": 1.2642, "step": 3165 }, { "epoch": 0.666247642991829, "grad_norm": 1.9756152629852295, "learning_rate": 1.9127770417969703e-05, "loss": 1.1418, "step": 3180 }, { "epoch": 0.6693903205531112, "grad_norm": 1.7655247449874878, "learning_rate": 1.9112773336834754e-05, "loss": 1.1605, "step": 3195 }, { "epoch": 0.6725329981143935, "grad_norm": 2.2654428482055664, "learning_rate": 1.9097654404814e-05, "loss": 1.2366, "step": 3210 }, { "epoch": 0.6756756756756757, "grad_norm": 2.2166404724121094, "learning_rate": 1.9082413824069302e-05, "loss": 1.1469, "step": 3225 }, { "epoch": 0.6788183532369579, "grad_norm": 2.409104824066162, "learning_rate": 1.9067051798389148e-05, "loss": 1.1303, "step": 3240 }, { "epoch": 0.6819610307982401, "grad_norm": 2.2395966053009033, "learning_rate": 1.905156853318591e-05, "loss": 1.1507, "step": 3255 }, { "epoch": 0.6851037083595223, "grad_norm": 2.369009017944336, "learning_rate": 1.903596423549312e-05, "loss": 1.1468, "step": 3270 }, { "epoch": 0.6882463859208046, "grad_norm": 4.745908260345459, "learning_rate": 1.9020239113962677e-05, "loss": 1.1675, "step": 3285 }, { "epoch": 0.6913890634820867, "grad_norm": 2.9272751808166504, "learning_rate": 1.900439337886207e-05, "loss": 1.1604, "step": 3300 }, { "epoch": 0.6945317410433689, "grad_norm": 1.923449993133545, "learning_rate": 1.898842724207157e-05, "loss": 1.136, "step": 3315 }, { "epoch": 0.6976744186046512, "grad_norm": 2.657238006591797, "learning_rate": 1.897234091708139e-05, "loss": 1.212, "step": 3330 }, { "epoch": 0.7008170961659334, "grad_norm": 2.484586715698242, "learning_rate": 1.8956134618988828e-05, "loss": 1.1938, "step": 3345 }, { "epoch": 0.7039597737272156, "grad_norm": 2.127962589263916, "learning_rate": 1.89398085644954e-05, "loss": 1.138, "step": 3360 }, { "epoch": 0.7071024512884978, "grad_norm": 4.683898448944092, "learning_rate": 1.8923362971903932e-05, "loss": 1.2393, "step": 3375 }, { "epoch": 0.71024512884978, "grad_norm": 2.090588092803955, "learning_rate": 1.8906798061115654e-05, "loss": 1.1721, "step": 3390 }, { "epoch": 0.7133878064110623, "grad_norm": 2.328129768371582, "learning_rate": 1.8890114053627244e-05, "loss": 1.1351, "step": 3405 }, { "epoch": 0.7165304839723444, "grad_norm": 2.18684458732605, "learning_rate": 1.8873311172527883e-05, "loss": 1.1793, "step": 3420 }, { "epoch": 0.7196731615336267, "grad_norm": 3.865600109100342, "learning_rate": 1.885638964249625e-05, "loss": 1.1605, "step": 3435 }, { "epoch": 0.7228158390949089, "grad_norm": 1.9220081567764282, "learning_rate": 1.8839349689797542e-05, "loss": 1.1056, "step": 3450 }, { "epoch": 0.725958516656191, "grad_norm": 2.822507858276367, "learning_rate": 1.8822191542280433e-05, "loss": 1.2242, "step": 3465 }, { "epoch": 0.7291011942174733, "grad_norm": 2.1855289936065674, "learning_rate": 1.880491542937404e-05, "loss": 1.1582, "step": 3480 }, { "epoch": 0.7322438717787555, "grad_norm": 2.4138951301574707, "learning_rate": 1.8787521582084833e-05, "loss": 1.1889, "step": 3495 }, { "epoch": 0.7353865493400377, "grad_norm": 4.04121732711792, "learning_rate": 1.8770010232993562e-05, "loss": 1.1682, "step": 3510 }, { "epoch": 0.73852922690132, "grad_norm": 2.7555859088897705, "learning_rate": 1.875238161625216e-05, "loss": 1.1242, "step": 3525 }, { "epoch": 0.7416719044626021, "grad_norm": 2.1336309909820557, "learning_rate": 1.873463596758058e-05, "loss": 1.1823, "step": 3540 }, { "epoch": 0.7448145820238844, "grad_norm": 2.1928298473358154, "learning_rate": 1.871677352426367e-05, "loss": 1.1785, "step": 3555 }, { "epoch": 0.7479572595851666, "grad_norm": 1.8896750211715698, "learning_rate": 1.8698794525147988e-05, "loss": 1.1871, "step": 3570 }, { "epoch": 0.7510999371464487, "grad_norm": 2.354185104370117, "learning_rate": 1.8680699210638608e-05, "loss": 1.072, "step": 3585 }, { "epoch": 0.754242614707731, "grad_norm": 2.3728153705596924, "learning_rate": 1.8662487822695906e-05, "loss": 1.1607, "step": 3600 }, { "epoch": 0.7573852922690132, "grad_norm": 1.997661828994751, "learning_rate": 1.8644160604832344e-05, "loss": 1.1824, "step": 3615 }, { "epoch": 0.7605279698302954, "grad_norm": 2.4399378299713135, "learning_rate": 1.8625717802109173e-05, "loss": 1.1794, "step": 3630 }, { "epoch": 0.7636706473915776, "grad_norm": 2.1330130100250244, "learning_rate": 1.860715966113319e-05, "loss": 1.2458, "step": 3645 }, { "epoch": 0.7668133249528598, "grad_norm": 2.0650925636291504, "learning_rate": 1.8588486430053442e-05, "loss": 1.1785, "step": 3660 }, { "epoch": 0.7699560025141421, "grad_norm": 2.570436477661133, "learning_rate": 1.856969835855787e-05, "loss": 1.154, "step": 3675 }, { "epoch": 0.7730986800754243, "grad_norm": 4.064576625823975, "learning_rate": 1.855079569787002e-05, "loss": 1.1833, "step": 3690 }, { "epoch": 0.7762413576367064, "grad_norm": 2.199369430541992, "learning_rate": 1.8531778700745652e-05, "loss": 1.1681, "step": 3705 }, { "epoch": 0.7793840351979887, "grad_norm": 2.3457729816436768, "learning_rate": 1.851264762146937e-05, "loss": 1.1177, "step": 3720 }, { "epoch": 0.7825267127592709, "grad_norm": 1.790176272392273, "learning_rate": 1.849340271585121e-05, "loss": 1.1603, "step": 3735 }, { "epoch": 0.7856693903205532, "grad_norm": 1.7440158128738403, "learning_rate": 1.8474044241223247e-05, "loss": 1.1867, "step": 3750 }, { "epoch": 0.7888120678818353, "grad_norm": 2.208021640777588, "learning_rate": 1.8454572456436126e-05, "loss": 1.1381, "step": 3765 }, { "epoch": 0.7919547454431175, "grad_norm": 2.744593858718872, "learning_rate": 1.8434987621855615e-05, "loss": 1.1186, "step": 3780 }, { "epoch": 0.7950974230043998, "grad_norm": 2.6648318767547607, "learning_rate": 1.841528999935912e-05, "loss": 1.1345, "step": 3795 }, { "epoch": 0.798240100565682, "grad_norm": 2.1822752952575684, "learning_rate": 1.8395479852332185e-05, "loss": 1.2148, "step": 3810 }, { "epoch": 0.8013827781269641, "grad_norm": 2.2398853302001953, "learning_rate": 1.8375557445664962e-05, "loss": 1.1915, "step": 3825 }, { "epoch": 0.8045254556882464, "grad_norm": 3.89241886138916, "learning_rate": 1.835552304574869e-05, "loss": 1.2088, "step": 3840 }, { "epoch": 0.8076681332495286, "grad_norm": 2.0591068267822266, "learning_rate": 1.8335376920472098e-05, "loss": 1.2107, "step": 3855 }, { "epoch": 0.8108108108108109, "grad_norm": 2.421823501586914, "learning_rate": 1.831511933921787e-05, "loss": 1.1696, "step": 3870 }, { "epoch": 0.813953488372093, "grad_norm": 3.9787890911102295, "learning_rate": 1.829475057285899e-05, "loss": 1.1756, "step": 3885 }, { "epoch": 0.8170961659333752, "grad_norm": 2.6043264865875244, "learning_rate": 1.8274270893755166e-05, "loss": 1.1729, "step": 3900 }, { "epoch": 0.8202388434946575, "grad_norm": 2.131624221801758, "learning_rate": 1.825368057574916e-05, "loss": 1.1465, "step": 3915 }, { "epoch": 0.8233815210559396, "grad_norm": 2.2352840900421143, "learning_rate": 1.823297989416314e-05, "loss": 1.1126, "step": 3930 }, { "epoch": 0.8265241986172219, "grad_norm": 1.9127349853515625, "learning_rate": 1.821216912579499e-05, "loss": 1.1081, "step": 3945 }, { "epoch": 0.8296668761785041, "grad_norm": 1.8138941526412964, "learning_rate": 1.819124854891461e-05, "loss": 1.1369, "step": 3960 }, { "epoch": 0.8328095537397863, "grad_norm": 3.4601540565490723, "learning_rate": 1.817021844326021e-05, "loss": 1.1378, "step": 3975 }, { "epoch": 0.8359522313010685, "grad_norm": 2.171710968017578, "learning_rate": 1.814907909003454e-05, "loss": 1.146, "step": 3990 }, { "epoch": 0.8390949088623507, "grad_norm": 2.376568078994751, "learning_rate": 1.8127830771901157e-05, "loss": 1.1759, "step": 4005 }, { "epoch": 0.8422375864236329, "grad_norm": 4.197197437286377, "learning_rate": 1.810647377298064e-05, "loss": 1.1319, "step": 4020 }, { "epoch": 0.8453802639849152, "grad_norm": 1.8338217735290527, "learning_rate": 1.8085008378846777e-05, "loss": 1.126, "step": 4035 }, { "epoch": 0.8485229415461973, "grad_norm": 4.26235818862915, "learning_rate": 1.806343487652276e-05, "loss": 1.1889, "step": 4050 }, { "epoch": 0.8516656191074796, "grad_norm": 2.5566670894622803, "learning_rate": 1.8041753554477344e-05, "loss": 1.1377, "step": 4065 }, { "epoch": 0.8548082966687618, "grad_norm": 3.85239839553833, "learning_rate": 1.8019964702620987e-05, "loss": 1.1859, "step": 4080 }, { "epoch": 0.857950974230044, "grad_norm": 1.832269549369812, "learning_rate": 1.7998068612301976e-05, "loss": 1.1816, "step": 4095 }, { "epoch": 0.8610936517913262, "grad_norm": 2.670729160308838, "learning_rate": 1.7976065576302528e-05, "loss": 1.1686, "step": 4110 }, { "epoch": 0.8642363293526084, "grad_norm": 2.3271324634552, "learning_rate": 1.7953955888834885e-05, "loss": 1.1683, "step": 4125 }, { "epoch": 0.8673790069138906, "grad_norm": 2.2425904273986816, "learning_rate": 1.7931739845537365e-05, "loss": 1.182, "step": 4140 }, { "epoch": 0.8705216844751729, "grad_norm": 1.5735398530960083, "learning_rate": 1.7909417743470415e-05, "loss": 1.0888, "step": 4155 }, { "epoch": 0.873664362036455, "grad_norm": 2.4160172939300537, "learning_rate": 1.7886989881112647e-05, "loss": 1.1943, "step": 4170 }, { "epoch": 0.8768070395977373, "grad_norm": 2.265516519546509, "learning_rate": 1.786445655835684e-05, "loss": 1.1819, "step": 4185 }, { "epoch": 0.8799497171590195, "grad_norm": 3.3854482173919678, "learning_rate": 1.784181807650592e-05, "loss": 1.1299, "step": 4200 }, { "epoch": 0.8830923947203017, "grad_norm": 1.209488868713379, "learning_rate": 1.7819074738268956e-05, "loss": 1.0379, "step": 4215 }, { "epoch": 0.8862350722815839, "grad_norm": 2.617237091064453, "learning_rate": 1.7796226847757084e-05, "loss": 1.2145, "step": 4230 }, { "epoch": 0.8893777498428661, "grad_norm": 4.028698921203613, "learning_rate": 1.7773274710479473e-05, "loss": 1.117, "step": 4245 }, { "epoch": 0.8925204274041484, "grad_norm": 2.382817506790161, "learning_rate": 1.7750218633339194e-05, "loss": 1.1122, "step": 4260 }, { "epoch": 0.8956631049654306, "grad_norm": 2.063629150390625, "learning_rate": 1.7727058924629165e-05, "loss": 1.1553, "step": 4275 }, { "epoch": 0.8988057825267127, "grad_norm": 1.7411893606185913, "learning_rate": 1.7703795894028003e-05, "loss": 1.1544, "step": 4290 }, { "epoch": 0.901948460087995, "grad_norm": 1.9747669696807861, "learning_rate": 1.768042985259588e-05, "loss": 1.1875, "step": 4305 }, { "epoch": 0.9050911376492772, "grad_norm": 2.338653326034546, "learning_rate": 1.765696111277038e-05, "loss": 1.1609, "step": 4320 }, { "epoch": 0.9082338152105593, "grad_norm": 2.3144114017486572, "learning_rate": 1.7633389988362305e-05, "loss": 1.1843, "step": 4335 }, { "epoch": 0.9113764927718416, "grad_norm": 2.072577953338623, "learning_rate": 1.760971679455149e-05, "loss": 1.1109, "step": 4350 }, { "epoch": 0.9145191703331238, "grad_norm": 2.2952466011047363, "learning_rate": 1.758594184788258e-05, "loss": 1.1714, "step": 4365 }, { "epoch": 0.9176618478944061, "grad_norm": 2.254108428955078, "learning_rate": 1.7562065466260816e-05, "loss": 1.1627, "step": 4380 }, { "epoch": 0.9208045254556882, "grad_norm": 2.286694049835205, "learning_rate": 1.753808796894775e-05, "loss": 1.135, "step": 4395 }, { "epoch": 0.9239472030169704, "grad_norm": 3.5377025604248047, "learning_rate": 1.751400967655701e-05, "loss": 1.1172, "step": 4410 }, { "epoch": 0.9270898805782527, "grad_norm": 1.9915852546691895, "learning_rate": 1.7489830911049995e-05, "loss": 1.1235, "step": 4425 }, { "epoch": 0.9302325581395349, "grad_norm": 1.916908860206604, "learning_rate": 1.746555199573158e-05, "loss": 1.1604, "step": 4440 }, { "epoch": 0.933375235700817, "grad_norm": 2.122201442718506, "learning_rate": 1.7441173255245773e-05, "loss": 1.17, "step": 4455 }, { "epoch": 0.9365179132620993, "grad_norm": 2.4088730812072754, "learning_rate": 1.74166950155714e-05, "loss": 1.1222, "step": 4470 }, { "epoch": 0.9396605908233815, "grad_norm": 2.2933411598205566, "learning_rate": 1.7392117604017733e-05, "loss": 1.1322, "step": 4485 }, { "epoch": 0.9428032683846638, "grad_norm": 2.084245204925537, "learning_rate": 1.7367441349220108e-05, "loss": 1.139, "step": 4500 }, { "epoch": 0.9459459459459459, "grad_norm": 2.220381498336792, "learning_rate": 1.734266658113555e-05, "loss": 1.1533, "step": 4515 }, { "epoch": 0.9490886235072281, "grad_norm": 2.4368200302124023, "learning_rate": 1.731779363103833e-05, "loss": 1.1276, "step": 4530 }, { "epoch": 0.9522313010685104, "grad_norm": 3.9131979942321777, "learning_rate": 1.729282283151557e-05, "loss": 1.2134, "step": 4545 }, { "epoch": 0.9553739786297926, "grad_norm": 2.0956265926361084, "learning_rate": 1.7267754516462773e-05, "loss": 1.1336, "step": 4560 }, { "epoch": 0.9585166561910748, "grad_norm": 2.3662939071655273, "learning_rate": 1.724258902107936e-05, "loss": 1.1807, "step": 4575 }, { "epoch": 0.961659333752357, "grad_norm": 2.2511234283447266, "learning_rate": 1.721732668186421e-05, "loss": 1.1407, "step": 4590 }, { "epoch": 0.9648020113136392, "grad_norm": 2.2256953716278076, "learning_rate": 1.7191967836611115e-05, "loss": 1.1797, "step": 4605 }, { "epoch": 0.9679446888749215, "grad_norm": 1.6837149858474731, "learning_rate": 1.716651282440432e-05, "loss": 1.1586, "step": 4620 }, { "epoch": 0.9710873664362036, "grad_norm": 1.941137433052063, "learning_rate": 1.7140961985613952e-05, "loss": 1.1739, "step": 4635 }, { "epoch": 0.9742300439974858, "grad_norm": 3.3290293216705322, "learning_rate": 1.711531566189146e-05, "loss": 1.1511, "step": 4650 }, { "epoch": 0.9773727215587681, "grad_norm": 2.164433002471924, "learning_rate": 1.7089574196165087e-05, "loss": 1.1144, "step": 4665 }, { "epoch": 0.9805153991200503, "grad_norm": 2.117098569869995, "learning_rate": 1.706373793263525e-05, "loss": 1.0804, "step": 4680 }, { "epoch": 0.9836580766813325, "grad_norm": 2.2358627319335938, "learning_rate": 1.7037807216769952e-05, "loss": 1.158, "step": 4695 }, { "epoch": 0.9868007542426147, "grad_norm": 2.1659979820251465, "learning_rate": 1.7011782395300152e-05, "loss": 1.1633, "step": 4710 }, { "epoch": 0.9899434318038969, "grad_norm": 1.9542781114578247, "learning_rate": 1.6985663816215156e-05, "loss": 1.2018, "step": 4725 }, { "epoch": 0.9930861093651792, "grad_norm": 2.3751957416534424, "learning_rate": 1.6959451828757915e-05, "loss": 1.1924, "step": 4740 }, { "epoch": 0.9962287869264613, "grad_norm": 2.5773661136627197, "learning_rate": 1.6933146783420412e-05, "loss": 1.2159, "step": 4755 }, { "epoch": 0.9993714644877436, "grad_norm": 1.986519694328308, "learning_rate": 1.6906749031938932e-05, "loss": 1.1635, "step": 4770 }, { "epoch": 1.0025141420490258, "grad_norm": 2.238046646118164, "learning_rate": 1.688025892728938e-05, "loss": 1.1517, "step": 4785 }, { "epoch": 1.005656819610308, "grad_norm": 2.3779406547546387, "learning_rate": 1.685367682368255e-05, "loss": 1.1341, "step": 4800 }, { "epoch": 1.0087994971715901, "grad_norm": 2.1286470890045166, "learning_rate": 1.6827003076559403e-05, "loss": 1.1293, "step": 4815 }, { "epoch": 1.0119421747328725, "grad_norm": 3.604013442993164, "learning_rate": 1.68002380425863e-05, "loss": 1.1733, "step": 4830 }, { "epoch": 1.0150848522941547, "grad_norm": 1.7821085453033447, "learning_rate": 1.6773382079650237e-05, "loss": 1.1215, "step": 4845 }, { "epoch": 1.0182275298554369, "grad_norm": 2.2310407161712646, "learning_rate": 1.6746435546854068e-05, "loss": 1.1364, "step": 4860 }, { "epoch": 1.021370207416719, "grad_norm": 2.5142269134521484, "learning_rate": 1.6719398804511682e-05, "loss": 1.1585, "step": 4875 }, { "epoch": 1.0245128849780012, "grad_norm": 2.2651455402374268, "learning_rate": 1.6692272214143224e-05, "loss": 1.1408, "step": 4890 }, { "epoch": 1.0276555625392834, "grad_norm": 2.247772693634033, "learning_rate": 1.6665056138470212e-05, "loss": 1.1134, "step": 4905 }, { "epoch": 1.0307982401005658, "grad_norm": 2.3006253242492676, "learning_rate": 1.6637750941410725e-05, "loss": 1.1252, "step": 4920 }, { "epoch": 1.033940917661848, "grad_norm": 1.8263660669326782, "learning_rate": 1.6610356988074517e-05, "loss": 1.1427, "step": 4935 }, { "epoch": 1.03708359522313, "grad_norm": 2.199976921081543, "learning_rate": 1.658287464475814e-05, "loss": 1.1208, "step": 4950 }, { "epoch": 1.0402262727844123, "grad_norm": 2.1440887451171875, "learning_rate": 1.6555304278940053e-05, "loss": 1.1305, "step": 4965 }, { "epoch": 1.0433689503456944, "grad_norm": 2.2581751346588135, "learning_rate": 1.65276462592757e-05, "loss": 1.1683, "step": 4980 }, { "epoch": 1.0465116279069768, "grad_norm": 1.7286102771759033, "learning_rate": 1.6499900955592587e-05, "loss": 1.1857, "step": 4995 }, { "epoch": 1.049654305468259, "grad_norm": 1.8500277996063232, "learning_rate": 1.6472068738885322e-05, "loss": 1.0997, "step": 5010 }, { "epoch": 1.0527969830295412, "grad_norm": 2.4594027996063232, "learning_rate": 1.644414998131067e-05, "loss": 1.0588, "step": 5025 }, { "epoch": 1.0559396605908233, "grad_norm": 2.3058676719665527, "learning_rate": 1.6416145056182577e-05, "loss": 1.1304, "step": 5040 }, { "epoch": 1.0590823381521055, "grad_norm": 1.9614354372024536, "learning_rate": 1.6388054337967173e-05, "loss": 1.1542, "step": 5055 }, { "epoch": 1.062225015713388, "grad_norm": 2.4512290954589844, "learning_rate": 1.635987820227776e-05, "loss": 1.1455, "step": 5070 }, { "epoch": 1.06536769327467, "grad_norm": 3.3408472537994385, "learning_rate": 1.6331617025869787e-05, "loss": 1.1008, "step": 5085 }, { "epoch": 1.0685103708359522, "grad_norm": 3.3527655601501465, "learning_rate": 1.6303271186635837e-05, "loss": 1.1311, "step": 5100 }, { "epoch": 1.0716530483972344, "grad_norm": 2.1265969276428223, "learning_rate": 1.627484106360055e-05, "loss": 1.1328, "step": 5115 }, { "epoch": 1.0747957259585166, "grad_norm": 2.1224098205566406, "learning_rate": 1.624632703691556e-05, "loss": 1.1432, "step": 5130 }, { "epoch": 1.077938403519799, "grad_norm": 2.1932625770568848, "learning_rate": 1.6217729487854404e-05, "loss": 1.1336, "step": 5145 }, { "epoch": 1.0810810810810811, "grad_norm": 1.477196455001831, "learning_rate": 1.6189048798807453e-05, "loss": 1.0739, "step": 5160 }, { "epoch": 1.0842237586423633, "grad_norm": 2.2613472938537598, "learning_rate": 1.616028535327677e-05, "loss": 1.0342, "step": 5175 }, { "epoch": 1.0873664362036455, "grad_norm": 2.0612030029296875, "learning_rate": 1.6131439535870982e-05, "loss": 1.2227, "step": 5190 }, { "epoch": 1.0905091137649277, "grad_norm": 2.0832862854003906, "learning_rate": 1.6102511732300156e-05, "loss": 1.1417, "step": 5205 }, { "epoch": 1.0936517913262098, "grad_norm": 2.326160192489624, "learning_rate": 1.607350232937064e-05, "loss": 1.1104, "step": 5220 }, { "epoch": 1.0967944688874922, "grad_norm": 1.928688883781433, "learning_rate": 1.6044411714979866e-05, "loss": 1.1503, "step": 5235 }, { "epoch": 1.0999371464487744, "grad_norm": 2.209092617034912, "learning_rate": 1.6015240278111184e-05, "loss": 1.0927, "step": 5250 }, { "epoch": 1.1030798240100566, "grad_norm": 2.2068710327148438, "learning_rate": 1.5987941027952896e-05, "loss": 1.097, "step": 5265 }, { "epoch": 1.1062225015713387, "grad_norm": 2.1571450233459473, "learning_rate": 1.5958614441290784e-05, "loss": 1.0841, "step": 5280 }, { "epoch": 1.109365179132621, "grad_norm": 2.010093927383423, "learning_rate": 1.5929208179383714e-05, "loss": 1.1261, "step": 5295 }, { "epoch": 1.1125078566939033, "grad_norm": 2.4886395931243896, "learning_rate": 1.58997226354357e-05, "loss": 1.1357, "step": 5310 }, { "epoch": 1.1156505342551855, "grad_norm": 1.658639907836914, "learning_rate": 1.5870158203710876e-05, "loss": 1.0573, "step": 5325 }, { "epoch": 1.1187932118164676, "grad_norm": 2.1249542236328125, "learning_rate": 1.584051527952821e-05, "loss": 1.1161, "step": 5340 }, { "epoch": 1.1219358893777498, "grad_norm": 2.086566209793091, "learning_rate": 1.5810794259256238e-05, "loss": 1.1282, "step": 5355 }, { "epoch": 1.125078566939032, "grad_norm": 2.5150198936462402, "learning_rate": 1.5780995540307734e-05, "loss": 1.1408, "step": 5370 }, { "epoch": 1.1282212445003144, "grad_norm": 3.7854840755462646, "learning_rate": 1.575111952113444e-05, "loss": 1.1125, "step": 5385 }, { "epoch": 1.1313639220615965, "grad_norm": 2.2000720500946045, "learning_rate": 1.5721166601221697e-05, "loss": 1.099, "step": 5400 }, { "epoch": 1.1345065996228787, "grad_norm": 2.055802345275879, "learning_rate": 1.5691137181083126e-05, "loss": 1.1057, "step": 5415 }, { "epoch": 1.1376492771841609, "grad_norm": 1.9730473756790161, "learning_rate": 1.5661031662255254e-05, "loss": 1.0642, "step": 5430 }, { "epoch": 1.140791954745443, "grad_norm": 2.1444499492645264, "learning_rate": 1.563085044729218e-05, "loss": 1.1095, "step": 5445 }, { "epoch": 1.1439346323067254, "grad_norm": 3.2961201667785645, "learning_rate": 1.5600593939760146e-05, "loss": 1.1613, "step": 5460 }, { "epoch": 1.1470773098680076, "grad_norm": 2.405123710632324, "learning_rate": 1.557026254423218e-05, "loss": 1.1211, "step": 5475 }, { "epoch": 1.1502199874292898, "grad_norm": 3.308258533477783, "learning_rate": 1.5539856666282655e-05, "loss": 1.1381, "step": 5490 }, { "epoch": 1.153362664990572, "grad_norm": 2.2963814735412598, "learning_rate": 1.5509376712481898e-05, "loss": 1.1293, "step": 5505 }, { "epoch": 1.156505342551854, "grad_norm": 1.9797133207321167, "learning_rate": 1.5478823090390726e-05, "loss": 1.0968, "step": 5520 }, { "epoch": 1.1596480201131363, "grad_norm": 2.6032137870788574, "learning_rate": 1.5448196208555014e-05, "loss": 1.1462, "step": 5535 }, { "epoch": 1.1627906976744187, "grad_norm": 1.9197511672973633, "learning_rate": 1.5417496476500212e-05, "loss": 1.0224, "step": 5550 }, { "epoch": 1.1659333752357008, "grad_norm": 2.2676188945770264, "learning_rate": 1.538672430472589e-05, "loss": 1.0625, "step": 5565 }, { "epoch": 1.169076052796983, "grad_norm": 2.4030728340148926, "learning_rate": 1.535588010470026e-05, "loss": 1.1171, "step": 5580 }, { "epoch": 1.1722187303582652, "grad_norm": 1.387372374534607, "learning_rate": 1.5324964288854615e-05, "loss": 1.0836, "step": 5595 }, { "epoch": 1.1753614079195476, "grad_norm": 2.1212260723114014, "learning_rate": 1.5293977270577895e-05, "loss": 1.1069, "step": 5610 }, { "epoch": 1.1785040854808297, "grad_norm": 2.2747840881347656, "learning_rate": 1.526291946421109e-05, "loss": 1.134, "step": 5625 }, { "epoch": 1.181646763042112, "grad_norm": 1.7659993171691895, "learning_rate": 1.5231791285041741e-05, "loss": 1.1512, "step": 5640 }, { "epoch": 1.184789440603394, "grad_norm": 1.815944790840149, "learning_rate": 1.5200593149298375e-05, "loss": 1.0769, "step": 5655 }, { "epoch": 1.1879321181646763, "grad_norm": 2.154893636703491, "learning_rate": 1.5169325474144936e-05, "loss": 1.1198, "step": 5670 }, { "epoch": 1.1910747957259584, "grad_norm": 1.9412492513656616, "learning_rate": 1.5137988677675215e-05, "loss": 1.115, "step": 5685 }, { "epoch": 1.1942174732872408, "grad_norm": 1.8665941953659058, "learning_rate": 1.5106583178907244e-05, "loss": 1.1067, "step": 5700 }, { "epoch": 1.197360150848523, "grad_norm": 1.9800127744674683, "learning_rate": 1.5075109397777714e-05, "loss": 1.0744, "step": 5715 }, { "epoch": 1.2005028284098052, "grad_norm": 3.109827756881714, "learning_rate": 1.5043567755136346e-05, "loss": 1.0699, "step": 5730 }, { "epoch": 1.2036455059710873, "grad_norm": 1.9781428575515747, "learning_rate": 1.5011958672740261e-05, "loss": 1.1205, "step": 5745 }, { "epoch": 1.2067881835323695, "grad_norm": 2.4044976234436035, "learning_rate": 1.498028257324836e-05, "loss": 1.11, "step": 5760 }, { "epoch": 1.2099308610936519, "grad_norm": 2.0579380989074707, "learning_rate": 1.4948539880215642e-05, "loss": 1.1586, "step": 5775 }, { "epoch": 1.213073538654934, "grad_norm": 3.235210418701172, "learning_rate": 1.4916731018087576e-05, "loss": 1.1164, "step": 5790 }, { "epoch": 1.2162162162162162, "grad_norm": 2.258272647857666, "learning_rate": 1.4884856412194386e-05, "loss": 1.1057, "step": 5805 }, { "epoch": 1.2193588937774984, "grad_norm": 3.207092761993408, "learning_rate": 1.4852916488745409e-05, "loss": 1.1112, "step": 5820 }, { "epoch": 1.2225015713387806, "grad_norm": 2.802191972732544, "learning_rate": 1.4820911674823345e-05, "loss": 1.0854, "step": 5835 }, { "epoch": 1.2256442489000627, "grad_norm": 2.0806314945220947, "learning_rate": 1.4788842398378595e-05, "loss": 1.1596, "step": 5850 }, { "epoch": 1.2287869264613451, "grad_norm": 2.3855271339416504, "learning_rate": 1.4756709088223508e-05, "loss": 1.1108, "step": 5865 }, { "epoch": 1.2319296040226273, "grad_norm": 2.3936338424682617, "learning_rate": 1.4724512174026653e-05, "loss": 1.1988, "step": 5880 }, { "epoch": 1.2350722815839095, "grad_norm": 1.9991942644119263, "learning_rate": 1.4692252086307079e-05, "loss": 1.1588, "step": 5895 }, { "epoch": 1.2382149591451916, "grad_norm": 2.2678310871124268, "learning_rate": 1.465992925642856e-05, "loss": 1.084, "step": 5910 }, { "epoch": 1.241357636706474, "grad_norm": 2.767444610595703, "learning_rate": 1.4627544116593816e-05, "loss": 1.1242, "step": 5925 }, { "epoch": 1.2445003142677562, "grad_norm": 1.6328760385513306, "learning_rate": 1.4595097099838747e-05, "loss": 1.087, "step": 5940 }, { "epoch": 1.2476429918290384, "grad_norm": 1.8721544742584229, "learning_rate": 1.4562588640026629e-05, "loss": 1.1078, "step": 5955 }, { "epoch": 1.2507856693903205, "grad_norm": 2.0327460765838623, "learning_rate": 1.453001917184233e-05, "loss": 1.0601, "step": 5970 }, { "epoch": 1.2539283469516027, "grad_norm": 1.8650814294815063, "learning_rate": 1.4497389130786474e-05, "loss": 1.0553, "step": 5985 }, { "epoch": 1.2570710245128849, "grad_norm": 2.416186809539795, "learning_rate": 1.4464698953169649e-05, "loss": 1.1542, "step": 6000 }, { "epoch": 1.260213702074167, "grad_norm": 3.0162692070007324, "learning_rate": 1.443194907610654e-05, "loss": 1.1424, "step": 6015 }, { "epoch": 1.2633563796354494, "grad_norm": 2.833125352859497, "learning_rate": 1.4399139937510107e-05, "loss": 1.1691, "step": 6030 }, { "epoch": 1.2664990571967316, "grad_norm": 1.900730848312378, "learning_rate": 1.4366271976085719e-05, "loss": 1.0463, "step": 6045 }, { "epoch": 1.2696417347580138, "grad_norm": 2.243030071258545, "learning_rate": 1.4333345631325294e-05, "loss": 1.1671, "step": 6060 }, { "epoch": 1.2727844123192962, "grad_norm": 2.186521530151367, "learning_rate": 1.430036134350142e-05, "loss": 1.1653, "step": 6075 }, { "epoch": 1.2759270898805783, "grad_norm": 1.8847732543945312, "learning_rate": 1.4267319553661456e-05, "loss": 1.0842, "step": 6090 }, { "epoch": 1.2790697674418605, "grad_norm": 2.303593158721924, "learning_rate": 1.4234220703621658e-05, "loss": 1.0481, "step": 6105 }, { "epoch": 1.2822124450031427, "grad_norm": 3.213543653488159, "learning_rate": 1.4201065235961258e-05, "loss": 1.1446, "step": 6120 }, { "epoch": 1.2853551225644249, "grad_norm": 2.283177137374878, "learning_rate": 1.4167853594016539e-05, "loss": 1.114, "step": 6135 }, { "epoch": 1.288497800125707, "grad_norm": 3.1280677318573, "learning_rate": 1.4134586221874922e-05, "loss": 1.1568, "step": 6150 }, { "epoch": 1.2916404776869892, "grad_norm": 3.3405239582061768, "learning_rate": 1.410126356436902e-05, "loss": 1.0806, "step": 6165 }, { "epoch": 1.2947831552482716, "grad_norm": 2.9373440742492676, "learning_rate": 1.406788606707069e-05, "loss": 1.128, "step": 6180 }, { "epoch": 1.2979258328095538, "grad_norm": 2.1123921871185303, "learning_rate": 1.4034454176285073e-05, "loss": 1.017, "step": 6195 }, { "epoch": 1.301068510370836, "grad_norm": 2.298046827316284, "learning_rate": 1.4000968339044637e-05, "loss": 1.157, "step": 6210 }, { "epoch": 1.304211187932118, "grad_norm": 3.053687572479248, "learning_rate": 1.3967429003103175e-05, "loss": 1.1575, "step": 6225 }, { "epoch": 1.3073538654934005, "grad_norm": 2.9560487270355225, "learning_rate": 1.3933836616929856e-05, "loss": 1.1414, "step": 6240 }, { "epoch": 1.3104965430546827, "grad_norm": 2.401231050491333, "learning_rate": 1.3900191629703192e-05, "loss": 1.0565, "step": 6255 }, { "epoch": 1.3136392206159648, "grad_norm": 2.4497265815734863, "learning_rate": 1.3866494491305058e-05, "loss": 1.1296, "step": 6270 }, { "epoch": 1.316781898177247, "grad_norm": 2.0227253437042236, "learning_rate": 1.3832745652314652e-05, "loss": 1.1229, "step": 6285 }, { "epoch": 1.3199245757385292, "grad_norm": 1.9386115074157715, "learning_rate": 1.3798945564002493e-05, "loss": 1.0128, "step": 6300 }, { "epoch": 1.3230672532998113, "grad_norm": 1.967246413230896, "learning_rate": 1.376509467832437e-05, "loss": 1.1388, "step": 6315 }, { "epoch": 1.3262099308610937, "grad_norm": 3.249537706375122, "learning_rate": 1.3731193447915316e-05, "loss": 1.0584, "step": 6330 }, { "epoch": 1.329352608422376, "grad_norm": 2.7894890308380127, "learning_rate": 1.3697242326083536e-05, "loss": 1.1075, "step": 6345 }, { "epoch": 1.332495285983658, "grad_norm": 3.2315704822540283, "learning_rate": 1.3663241766804359e-05, "loss": 1.0972, "step": 6360 }, { "epoch": 1.3356379635449402, "grad_norm": 2.295436143875122, "learning_rate": 1.3629192224714163e-05, "loss": 1.1194, "step": 6375 }, { "epoch": 1.3387806411062226, "grad_norm": 1.2325416803359985, "learning_rate": 1.3595094155104297e-05, "loss": 1.099, "step": 6390 }, { "epoch": 1.3419233186675048, "grad_norm": 2.3862216472625732, "learning_rate": 1.3560948013914997e-05, "loss": 1.0684, "step": 6405 }, { "epoch": 1.345065996228787, "grad_norm": 2.1027228832244873, "learning_rate": 1.3526754257729273e-05, "loss": 1.0776, "step": 6420 }, { "epoch": 1.3482086737900691, "grad_norm": 2.421541213989258, "learning_rate": 1.3492513343766831e-05, "loss": 1.1078, "step": 6435 }, { "epoch": 1.3513513513513513, "grad_norm": 2.7943713665008545, "learning_rate": 1.3458225729877933e-05, "loss": 1.1076, "step": 6450 }, { "epoch": 1.3544940289126335, "grad_norm": 2.077890396118164, "learning_rate": 1.3423891874537289e-05, "loss": 1.1609, "step": 6465 }, { "epoch": 1.3576367064739157, "grad_norm": 2.0880138874053955, "learning_rate": 1.3389512236837931e-05, "loss": 1.0667, "step": 6480 }, { "epoch": 1.360779384035198, "grad_norm": 2.1045663356781006, "learning_rate": 1.3355087276485055e-05, "loss": 1.113, "step": 6495 }, { "epoch": 1.3639220615964802, "grad_norm": 2.125664472579956, "learning_rate": 1.3320617453789894e-05, "loss": 1.0619, "step": 6510 }, { "epoch": 1.3670647391577624, "grad_norm": 2.1109349727630615, "learning_rate": 1.3286103229663554e-05, "loss": 1.1235, "step": 6525 }, { "epoch": 1.3702074167190446, "grad_norm": 2.0411548614501953, "learning_rate": 1.3251545065610848e-05, "loss": 1.0872, "step": 6540 }, { "epoch": 1.373350094280327, "grad_norm": 3.083622694015503, "learning_rate": 1.3216943423724135e-05, "loss": 1.117, "step": 6555 }, { "epoch": 1.3764927718416091, "grad_norm": 1.8469839096069336, "learning_rate": 1.3182298766677136e-05, "loss": 1.1404, "step": 6570 }, { "epoch": 1.3796354494028913, "grad_norm": 2.121103525161743, "learning_rate": 1.3147611557718733e-05, "loss": 1.1007, "step": 6585 }, { "epoch": 1.3827781269641735, "grad_norm": 2.183732271194458, "learning_rate": 1.3112882260666805e-05, "loss": 1.1362, "step": 6600 }, { "epoch": 1.3859208045254556, "grad_norm": 2.0963168144226074, "learning_rate": 1.3078111339902004e-05, "loss": 1.1019, "step": 6615 }, { "epoch": 1.3890634820867378, "grad_norm": 3.269718885421753, "learning_rate": 1.3043299260361545e-05, "loss": 1.1073, "step": 6630 }, { "epoch": 1.3922061596480202, "grad_norm": 2.942408323287964, "learning_rate": 1.3008446487532998e-05, "loss": 1.0783, "step": 6645 }, { "epoch": 1.3953488372093024, "grad_norm": 2.6441168785095215, "learning_rate": 1.2973553487448068e-05, "loss": 1.1656, "step": 6660 }, { "epoch": 1.3984915147705845, "grad_norm": 2.102250099182129, "learning_rate": 1.2938620726676345e-05, "loss": 1.1408, "step": 6675 }, { "epoch": 1.4016341923318667, "grad_norm": 3.086590051651001, "learning_rate": 1.2903648672319084e-05, "loss": 1.1059, "step": 6690 }, { "epoch": 1.404776869893149, "grad_norm": 2.3277883529663086, "learning_rate": 1.2868637792002952e-05, "loss": 1.1665, "step": 6705 }, { "epoch": 1.4079195474544313, "grad_norm": 1.6698582172393799, "learning_rate": 1.2833588553873768e-05, "loss": 1.0649, "step": 6720 }, { "epoch": 1.4110622250157134, "grad_norm": 2.26218581199646, "learning_rate": 1.2798501426590262e-05, "loss": 1.1894, "step": 6735 }, { "epoch": 1.4142049025769956, "grad_norm": 1.8945990800857544, "learning_rate": 1.276337687931778e-05, "loss": 1.1507, "step": 6750 }, { "epoch": 1.4173475801382778, "grad_norm": 2.481565475463867, "learning_rate": 1.2728215381722044e-05, "loss": 1.0937, "step": 6765 }, { "epoch": 1.42049025769956, "grad_norm": 3.565307140350342, "learning_rate": 1.2693017403962842e-05, "loss": 1.197, "step": 6780 }, { "epoch": 1.4236329352608421, "grad_norm": 2.697265863418579, "learning_rate": 1.2657783416687763e-05, "loss": 1.0842, "step": 6795 }, { "epoch": 1.4267756128221245, "grad_norm": 2.244889736175537, "learning_rate": 1.2622513891025889e-05, "loss": 1.0942, "step": 6810 }, { "epoch": 1.4299182903834067, "grad_norm": 2.428379774093628, "learning_rate": 1.2587209298581506e-05, "loss": 1.131, "step": 6825 }, { "epoch": 1.4330609679446888, "grad_norm": 2.320680618286133, "learning_rate": 1.2551870111427791e-05, "loss": 1.1279, "step": 6840 }, { "epoch": 1.436203645505971, "grad_norm": 2.004258632659912, "learning_rate": 1.25164968021005e-05, "loss": 1.0718, "step": 6855 }, { "epoch": 1.4393463230672534, "grad_norm": 3.006464719772339, "learning_rate": 1.2481089843591655e-05, "loss": 1.1785, "step": 6870 }, { "epoch": 1.4424890006285356, "grad_norm": 2.481252670288086, "learning_rate": 1.2445649709343217e-05, "loss": 1.1382, "step": 6885 }, { "epoch": 1.4456316781898177, "grad_norm": 2.247004985809326, "learning_rate": 1.2410176873240751e-05, "loss": 1.1179, "step": 6900 }, { "epoch": 1.4487743557511, "grad_norm": 2.7122623920440674, "learning_rate": 1.237467180960709e-05, "loss": 1.0569, "step": 6915 }, { "epoch": 1.451917033312382, "grad_norm": 2.119673013687134, "learning_rate": 1.2339134993196007e-05, "loss": 1.0939, "step": 6930 }, { "epoch": 1.4550597108736643, "grad_norm": 2.4448978900909424, "learning_rate": 1.2303566899185842e-05, "loss": 1.0598, "step": 6945 }, { "epoch": 1.4582023884349467, "grad_norm": 2.211289405822754, "learning_rate": 1.2267968003173166e-05, "loss": 1.1679, "step": 6960 }, { "epoch": 1.4613450659962288, "grad_norm": 1.9416890144348145, "learning_rate": 1.2232338781166422e-05, "loss": 1.0983, "step": 6975 }, { "epoch": 1.464487743557511, "grad_norm": 2.0862271785736084, "learning_rate": 1.2196679709579546e-05, "loss": 1.0911, "step": 6990 }, { "epoch": 1.4676304211187932, "grad_norm": 2.4440407752990723, "learning_rate": 1.2160991265225621e-05, "loss": 1.1073, "step": 7005 }, { "epoch": 1.4707730986800756, "grad_norm": 2.1422340869903564, "learning_rate": 1.2125273925310465e-05, "loss": 1.0904, "step": 7020 }, { "epoch": 1.4739157762413577, "grad_norm": 2.2515573501586914, "learning_rate": 1.2089528167426294e-05, "loss": 1.1244, "step": 7035 }, { "epoch": 1.47705845380264, "grad_norm": 2.149007558822632, "learning_rate": 1.2053754469545291e-05, "loss": 1.0963, "step": 7050 }, { "epoch": 1.480201131363922, "grad_norm": 3.1129860877990723, "learning_rate": 1.2017953310013255e-05, "loss": 1.1655, "step": 7065 }, { "epoch": 1.4833438089252042, "grad_norm": 1.7863973379135132, "learning_rate": 1.1982125167543175e-05, "loss": 1.0844, "step": 7080 }, { "epoch": 1.4864864864864864, "grad_norm": 2.28974986076355, "learning_rate": 1.1946270521208845e-05, "loss": 1.0585, "step": 7095 }, { "epoch": 1.4896291640477686, "grad_norm": 2.3061442375183105, "learning_rate": 1.1910389850438457e-05, "loss": 1.1377, "step": 7110 }, { "epoch": 1.492771841609051, "grad_norm": 1.6387587785720825, "learning_rate": 1.1874483635008183e-05, "loss": 1.1109, "step": 7125 }, { "epoch": 1.4959145191703331, "grad_norm": 1.6554300785064697, "learning_rate": 1.1838552355035761e-05, "loss": 1.0686, "step": 7140 }, { "epoch": 1.4990571967316153, "grad_norm": 2.2356905937194824, "learning_rate": 1.1802596490974088e-05, "loss": 1.0877, "step": 7155 }, { "epoch": 1.5021998742928977, "grad_norm": 2.838385820388794, "learning_rate": 1.1766616523604774e-05, "loss": 1.1015, "step": 7170 }, { "epoch": 1.5053425518541799, "grad_norm": 1.8725031614303589, "learning_rate": 1.1730612934031737e-05, "loss": 1.0806, "step": 7185 }, { "epoch": 1.508485229415462, "grad_norm": 2.853379011154175, "learning_rate": 1.1694586203674742e-05, "loss": 1.0796, "step": 7200 }, { "epoch": 1.5116279069767442, "grad_norm": 2.5570309162139893, "learning_rate": 1.1658536814263e-05, "loss": 1.0773, "step": 7215 }, { "epoch": 1.5147705845380264, "grad_norm": 2.4062445163726807, "learning_rate": 1.1622465247828681e-05, "loss": 0.9971, "step": 7230 }, { "epoch": 1.5179132620993085, "grad_norm": 2.4074044227600098, "learning_rate": 1.1586371986700522e-05, "loss": 1.0774, "step": 7245 }, { "epoch": 1.5210559396605907, "grad_norm": 2.0330376625061035, "learning_rate": 1.155025751349732e-05, "loss": 1.1209, "step": 7260 }, { "epoch": 1.5241986172218729, "grad_norm": 2.229322671890259, "learning_rate": 1.1516531959834856e-05, "loss": 1.1011, "step": 7275 }, { "epoch": 1.5273412947831553, "grad_norm": 2.2459731101989746, "learning_rate": 1.1480377846161259e-05, "loss": 1.095, "step": 7290 }, { "epoch": 1.5304839723444374, "grad_norm": 1.9737571477890015, "learning_rate": 1.1444203937706724e-05, "loss": 1.0764, "step": 7305 }, { "epoch": 1.5336266499057196, "grad_norm": 2.0487782955169678, "learning_rate": 1.1408010718168447e-05, "loss": 1.1322, "step": 7320 }, { "epoch": 1.536769327467002, "grad_norm": 2.7282018661499023, "learning_rate": 1.1371798671501813e-05, "loss": 1.1322, "step": 7335 }, { "epoch": 1.5399120050282842, "grad_norm": 3.496860980987549, "learning_rate": 1.133556828191398e-05, "loss": 1.0829, "step": 7350 }, { "epoch": 1.5430546825895664, "grad_norm": 1.9811838865280151, "learning_rate": 1.1299320033857355e-05, "loss": 1.0813, "step": 7365 }, { "epoch": 1.5461973601508485, "grad_norm": 2.5440211296081543, "learning_rate": 1.1263054412023158e-05, "loss": 1.0347, "step": 7380 }, { "epoch": 1.5493400377121307, "grad_norm": 1.8516343832015991, "learning_rate": 1.1226771901334901e-05, "loss": 1.0719, "step": 7395 }, { "epoch": 1.5524827152734129, "grad_norm": 2.147573709487915, "learning_rate": 1.1190472986941939e-05, "loss": 1.068, "step": 7410 }, { "epoch": 1.555625392834695, "grad_norm": 1.999443531036377, "learning_rate": 1.1154158154212964e-05, "loss": 1.0674, "step": 7425 }, { "epoch": 1.5587680703959774, "grad_norm": 1.5025156736373901, "learning_rate": 1.111782788872952e-05, "loss": 1.1003, "step": 7440 }, { "epoch": 1.5619107479572596, "grad_norm": 2.162722587585449, "learning_rate": 1.1081482676279504e-05, "loss": 1.078, "step": 7455 }, { "epoch": 1.5650534255185418, "grad_norm": 2.049753427505493, "learning_rate": 1.1045123002850683e-05, "loss": 1.1825, "step": 7470 }, { "epoch": 1.5681961030798242, "grad_norm": 2.4640021324157715, "learning_rate": 1.1008749354624181e-05, "loss": 1.0942, "step": 7485 }, { "epoch": 1.5713387806411063, "grad_norm": 2.2149126529693604, "learning_rate": 1.0972362217967998e-05, "loss": 1.1415, "step": 7500 }, { "epoch": 1.5744814582023885, "grad_norm": 1.953383207321167, "learning_rate": 1.0935962079430468e-05, "loss": 1.0875, "step": 7515 }, { "epoch": 1.5776241357636707, "grad_norm": 2.371173620223999, "learning_rate": 1.0899549425733811e-05, "loss": 1.1068, "step": 7530 }, { "epoch": 1.5807668133249528, "grad_norm": 2.135591983795166, "learning_rate": 1.0863124743767566e-05, "loss": 1.1094, "step": 7545 }, { "epoch": 1.583909490886235, "grad_norm": 2.1090261936187744, "learning_rate": 1.0826688520582124e-05, "loss": 1.0459, "step": 7560 }, { "epoch": 1.5870521684475172, "grad_norm": 2.8963959217071533, "learning_rate": 1.0790241243382187e-05, "loss": 1.1019, "step": 7575 }, { "epoch": 1.5901948460087993, "grad_norm": 2.2832436561584473, "learning_rate": 1.0753783399520281e-05, "loss": 1.1125, "step": 7590 }, { "epoch": 1.5933375235700817, "grad_norm": 2.1328582763671875, "learning_rate": 1.0717315476490205e-05, "loss": 1.1088, "step": 7605 }, { "epoch": 1.596480201131364, "grad_norm": 1.7953342199325562, "learning_rate": 1.0680837961920538e-05, "loss": 1.0812, "step": 7620 }, { "epoch": 1.5996228786926463, "grad_norm": 2.0836551189422607, "learning_rate": 1.064435134356812e-05, "loss": 1.1015, "step": 7635 }, { "epoch": 1.6027655562539285, "grad_norm": 3.2132134437561035, "learning_rate": 1.0607856109311508e-05, "loss": 1.1377, "step": 7650 }, { "epoch": 1.6059082338152106, "grad_norm": 2.149661064147949, "learning_rate": 1.0571352747144477e-05, "loss": 1.0525, "step": 7665 }, { "epoch": 1.6090509113764928, "grad_norm": 1.8691046237945557, "learning_rate": 1.053484174516948e-05, "loss": 1.0954, "step": 7680 }, { "epoch": 1.612193588937775, "grad_norm": 1.9283403158187866, "learning_rate": 1.049832359159112e-05, "loss": 1.1199, "step": 7695 }, { "epoch": 1.6153362664990571, "grad_norm": 1.9671255350112915, "learning_rate": 1.0461798774709635e-05, "loss": 1.068, "step": 7710 }, { "epoch": 1.6184789440603393, "grad_norm": 2.721006393432617, "learning_rate": 1.0425267782914359e-05, "loss": 1.1274, "step": 7725 }, { "epoch": 1.6216216216216215, "grad_norm": 2.0059800148010254, "learning_rate": 1.0388731104677191e-05, "loss": 1.0832, "step": 7740 }, { "epoch": 1.6247642991829039, "grad_norm": 2.346212863922119, "learning_rate": 1.0352189228546068e-05, "loss": 1.1092, "step": 7755 }, { "epoch": 1.627906976744186, "grad_norm": 1.6566693782806396, "learning_rate": 1.031564264313843e-05, "loss": 1.033, "step": 7770 }, { "epoch": 1.6310496543054682, "grad_norm": 2.4089715480804443, "learning_rate": 1.0279091837134681e-05, "loss": 1.114, "step": 7785 }, { "epoch": 1.6341923318667506, "grad_norm": 2.9500551223754883, "learning_rate": 1.0242537299271675e-05, "loss": 1.1282, "step": 7800 }, { "epoch": 1.6373350094280328, "grad_norm": 1.8416279554367065, "learning_rate": 1.0205979518336148e-05, "loss": 1.0857, "step": 7815 }, { "epoch": 1.640477686989315, "grad_norm": 2.1354591846466064, "learning_rate": 1.0169418983158218e-05, "loss": 1.0807, "step": 7830 }, { "epoch": 1.6436203645505971, "grad_norm": 1.705767035484314, "learning_rate": 1.013285618260482e-05, "loss": 1.0273, "step": 7845 }, { "epoch": 1.6467630421118793, "grad_norm": 1.9959582090377808, "learning_rate": 1.009629160557318e-05, "loss": 1.0705, "step": 7860 }, { "epoch": 1.6499057196731615, "grad_norm": 2.004734754562378, "learning_rate": 1.0059725740984285e-05, "loss": 1.1216, "step": 7875 }, { "epoch": 1.6530483972344436, "grad_norm": 2.1259968280792236, "learning_rate": 1.0023159077776332e-05, "loss": 1.1618, "step": 7890 }, { "epoch": 1.6561910747957258, "grad_norm": 1.855248212814331, "learning_rate": 9.986592104898202e-06, "loss": 1.1007, "step": 7905 }, { "epoch": 1.6593337523570082, "grad_norm": 2.1948564052581787, "learning_rate": 9.950025311302914e-06, "loss": 1.1203, "step": 7920 }, { "epoch": 1.6624764299182904, "grad_norm": 2.140425205230713, "learning_rate": 9.913459185941085e-06, "loss": 1.0788, "step": 7935 }, { "epoch": 1.6656191074795728, "grad_norm": 3.158501148223877, "learning_rate": 9.87689421775441e-06, "loss": 1.0806, "step": 7950 }, { "epoch": 1.668761785040855, "grad_norm": 1.9126044511795044, "learning_rate": 9.840330895669096e-06, "loss": 1.086, "step": 7965 }, { "epoch": 1.671904462602137, "grad_norm": 2.586374044418335, "learning_rate": 9.803769708589352e-06, "loss": 1.0511, "step": 7980 }, { "epoch": 1.6750471401634193, "grad_norm": 2.192415475845337, "learning_rate": 9.767211145390827e-06, "loss": 1.1149, "step": 7995 }, { "epoch": 1.6781898177247014, "grad_norm": 2.12808895111084, "learning_rate": 9.73065569491409e-06, "loss": 1.1015, "step": 8010 }, { "epoch": 1.6813324952859836, "grad_norm": 2.3615734577178955, "learning_rate": 9.694103845958095e-06, "loss": 1.126, "step": 8025 }, { "epoch": 1.6844751728472658, "grad_norm": 1.6970425844192505, "learning_rate": 9.657556087273632e-06, "loss": 1.0458, "step": 8040 }, { "epoch": 1.687617850408548, "grad_norm": 2.4654476642608643, "learning_rate": 9.621012907556793e-06, "loss": 1.1316, "step": 8055 }, { "epoch": 1.6907605279698303, "grad_norm": 2.2715861797332764, "learning_rate": 9.584474795442454e-06, "loss": 1.0997, "step": 8070 }, { "epoch": 1.6939032055311125, "grad_norm": 2.117677688598633, "learning_rate": 9.547942239497716e-06, "loss": 1.0896, "step": 8085 }, { "epoch": 1.6970458830923947, "grad_norm": 1.9031733274459839, "learning_rate": 9.511415728215407e-06, "loss": 1.0468, "step": 8100 }, { "epoch": 1.700188560653677, "grad_norm": 3.147984504699707, "learning_rate": 9.474895750007508e-06, "loss": 1.0695, "step": 8115 }, { "epoch": 1.7033312382149592, "grad_norm": 2.6544809341430664, "learning_rate": 9.438382793198655e-06, "loss": 1.1083, "step": 8130 }, { "epoch": 1.7064739157762414, "grad_norm": 2.071148633956909, "learning_rate": 9.401877346019587e-06, "loss": 1.1407, "step": 8145 }, { "epoch": 1.7096165933375236, "grad_norm": 2.063387155532837, "learning_rate": 9.36537989660065e-06, "loss": 1.0749, "step": 8160 }, { "epoch": 1.7127592708988058, "grad_norm": 1.9773895740509033, "learning_rate": 9.328890932965225e-06, "loss": 1.0723, "step": 8175 }, { "epoch": 1.715901948460088, "grad_norm": 3.0571677684783936, "learning_rate": 9.29241094302324e-06, "loss": 1.1296, "step": 8190 }, { "epoch": 1.71904462602137, "grad_norm": 2.203171730041504, "learning_rate": 9.255940414564621e-06, "loss": 1.1401, "step": 8205 }, { "epoch": 1.7221873035826523, "grad_norm": 1.9974406957626343, "learning_rate": 9.21947983525279e-06, "loss": 1.1158, "step": 8220 }, { "epoch": 1.7253299811439347, "grad_norm": 2.110973834991455, "learning_rate": 9.183029692618134e-06, "loss": 1.0684, "step": 8235 }, { "epoch": 1.7284726587052168, "grad_norm": 1.8630388975143433, "learning_rate": 9.146590474051483e-06, "loss": 1.0376, "step": 8250 }, { "epoch": 1.7316153362664992, "grad_norm": 2.1341989040374756, "learning_rate": 9.110162666797595e-06, "loss": 1.0955, "step": 8265 }, { "epoch": 1.7347580138277814, "grad_norm": 2.5358388423919678, "learning_rate": 9.076174105269065e-06, "loss": 1.1651, "step": 8280 }, { "epoch": 1.7379006913890636, "grad_norm": 2.8575525283813477, "learning_rate": 9.039769740923183e-06, "loss": 1.1082, "step": 8295 }, { "epoch": 1.7410433689503457, "grad_norm": 2.287961959838867, "learning_rate": 9.003378216236902e-06, "loss": 1.1151, "step": 8310 }, { "epoch": 1.744186046511628, "grad_norm": 1.2058899402618408, "learning_rate": 8.967000017817247e-06, "loss": 1.0793, "step": 8325 }, { "epoch": 1.74732872407291, "grad_norm": 2.0018842220306396, "learning_rate": 8.930635632093046e-06, "loss": 1.0819, "step": 8340 }, { "epoch": 1.7504714016341922, "grad_norm": 1.9017311334609985, "learning_rate": 8.894285545308437e-06, "loss": 1.0658, "step": 8355 }, { "epoch": 1.7536140791954744, "grad_norm": 1.9596132040023804, "learning_rate": 8.857950243516364e-06, "loss": 1.1065, "step": 8370 }, { "epoch": 1.7567567567567568, "grad_norm": 1.979177713394165, "learning_rate": 8.821630212572074e-06, "loss": 1.0362, "step": 8385 }, { "epoch": 1.759899434318039, "grad_norm": 2.1770269870758057, "learning_rate": 8.785325938126608e-06, "loss": 1.0974, "step": 8400 }, { "epoch": 1.7630421118793211, "grad_norm": 2.125962734222412, "learning_rate": 8.749037905620334e-06, "loss": 1.1541, "step": 8415 }, { "epoch": 1.7661847894406035, "grad_norm": 2.0153768062591553, "learning_rate": 8.712766600276435e-06, "loss": 1.1233, "step": 8430 }, { "epoch": 1.7693274670018857, "grad_norm": 2.15311336517334, "learning_rate": 8.676512507094438e-06, "loss": 1.0341, "step": 8445 }, { "epoch": 1.7724701445631679, "grad_norm": 2.8032867908477783, "learning_rate": 8.640276110843702e-06, "loss": 1.0944, "step": 8460 }, { "epoch": 1.77561282212445, "grad_norm": 2.2355639934539795, "learning_rate": 8.604057896056965e-06, "loss": 1.1147, "step": 8475 }, { "epoch": 1.7787554996857322, "grad_norm": 2.222898006439209, "learning_rate": 8.567858347023843e-06, "loss": 1.1063, "step": 8490 }, { "epoch": 1.7818981772470144, "grad_norm": 2.043031692504883, "learning_rate": 8.531677947784383e-06, "loss": 1.1142, "step": 8505 }, { "epoch": 1.7850408548082966, "grad_norm": 2.3380000591278076, "learning_rate": 8.495517182122546e-06, "loss": 1.1379, "step": 8520 }, { "epoch": 1.7881835323695787, "grad_norm": 1.9827522039413452, "learning_rate": 8.459376533559784e-06, "loss": 1.1067, "step": 8535 }, { "epoch": 1.7913262099308611, "grad_norm": 2.2196788787841797, "learning_rate": 8.423256485348543e-06, "loss": 1.0563, "step": 8550 }, { "epoch": 1.7944688874921433, "grad_norm": 2.038891315460205, "learning_rate": 8.387157520465816e-06, "loss": 1.1151, "step": 8565 }, { "epoch": 1.7976115650534257, "grad_norm": 2.597698211669922, "learning_rate": 8.351080121606685e-06, "loss": 1.1025, "step": 8580 }, { "epoch": 1.8007542426147078, "grad_norm": 2.8073418140411377, "learning_rate": 8.315024771177862e-06, "loss": 1.136, "step": 8595 }, { "epoch": 1.80389692017599, "grad_norm": 2.143312692642212, "learning_rate": 8.278991951291234e-06, "loss": 1.0655, "step": 8610 }, { "epoch": 1.8070395977372722, "grad_norm": 1.5993940830230713, "learning_rate": 8.242982143757429e-06, "loss": 1.1023, "step": 8625 }, { "epoch": 1.8101822752985544, "grad_norm": 2.127690553665161, "learning_rate": 8.20699583007936e-06, "loss": 1.104, "step": 8640 }, { "epoch": 1.8133249528598365, "grad_norm": 2.57829213142395, "learning_rate": 8.171033491445806e-06, "loss": 1.0658, "step": 8655 }, { "epoch": 1.8164676304211187, "grad_norm": 2.4844679832458496, "learning_rate": 8.13509560872495e-06, "loss": 1.1207, "step": 8670 }, { "epoch": 1.8196103079824009, "grad_norm": 1.7117849588394165, "learning_rate": 8.099182662457972e-06, "loss": 1.1019, "step": 8685 }, { "epoch": 1.8227529855436833, "grad_norm": 1.6486339569091797, "learning_rate": 8.063295132852616e-06, "loss": 1.1566, "step": 8700 }, { "epoch": 1.8258956631049654, "grad_norm": 2.264791965484619, "learning_rate": 8.027433499776768e-06, "loss": 1.1269, "step": 8715 }, { "epoch": 1.8290383406662476, "grad_norm": 2.0826590061187744, "learning_rate": 7.99159824275204e-06, "loss": 1.1668, "step": 8730 }, { "epoch": 1.83218101822753, "grad_norm": 3.4813215732574463, "learning_rate": 7.955789840947357e-06, "loss": 1.0613, "step": 8745 }, { "epoch": 1.8353236957888122, "grad_norm": 1.950697898864746, "learning_rate": 7.920008773172549e-06, "loss": 1.1151, "step": 8760 }, { "epoch": 1.8384663733500943, "grad_norm": 2.761364459991455, "learning_rate": 7.884255517871955e-06, "loss": 1.0418, "step": 8775 }, { "epoch": 1.8416090509113765, "grad_norm": 2.035162925720215, "learning_rate": 7.848530553118024e-06, "loss": 1.1395, "step": 8790 }, { "epoch": 1.8447517284726587, "grad_norm": 2.061817169189453, "learning_rate": 7.812834356604913e-06, "loss": 1.1337, "step": 8805 }, { "epoch": 1.8478944060339408, "grad_norm": 2.214123010635376, "learning_rate": 7.777167405642109e-06, "loss": 1.09, "step": 8820 }, { "epoch": 1.851037083595223, "grad_norm": 2.1376426219940186, "learning_rate": 7.741530177148041e-06, "loss": 1.0908, "step": 8835 }, { "epoch": 1.8541797611565052, "grad_norm": 1.6116961240768433, "learning_rate": 7.70592314764371e-06, "loss": 1.1011, "step": 8850 }, { "epoch": 1.8573224387177876, "grad_norm": 2.253972291946411, "learning_rate": 7.670346793246317e-06, "loss": 1.0856, "step": 8865 }, { "epoch": 1.8604651162790697, "grad_norm": 1.9321953058242798, "learning_rate": 7.634801589662882e-06, "loss": 1.0851, "step": 8880 }, { "epoch": 1.8636077938403521, "grad_norm": 2.758145332336426, "learning_rate": 7.599288012183897e-06, "loss": 1.0952, "step": 8895 }, { "epoch": 1.8667504714016343, "grad_norm": 3.209368944168091, "learning_rate": 7.563806535676974e-06, "loss": 1.1081, "step": 8910 }, { "epoch": 1.8698931489629165, "grad_norm": 3.290693998336792, "learning_rate": 7.528357634580486e-06, "loss": 1.0403, "step": 8925 }, { "epoch": 1.8730358265241986, "grad_norm": 2.0125486850738525, "learning_rate": 7.4929417828972205e-06, "loss": 1.0801, "step": 8940 }, { "epoch": 1.8761785040854808, "grad_norm": 2.3519296646118164, "learning_rate": 7.457559454188052e-06, "loss": 1.1264, "step": 8955 }, { "epoch": 1.879321181646763, "grad_norm": 2.394660472869873, "learning_rate": 7.422211121565607e-06, "loss": 1.1271, "step": 8970 }, { "epoch": 1.8824638592080452, "grad_norm": 2.127760648727417, "learning_rate": 7.386897257687924e-06, "loss": 1.1308, "step": 8985 }, { "epoch": 1.8856065367693273, "grad_norm": 2.5212738513946533, "learning_rate": 7.3516183347521575e-06, "loss": 1.0758, "step": 9000 }, { "epoch": 1.8887492143306097, "grad_norm": 2.113893508911133, "learning_rate": 7.316374824488247e-06, "loss": 1.1463, "step": 9015 }, { "epoch": 1.8918918918918919, "grad_norm": 1.9831585884094238, "learning_rate": 7.281167198152615e-06, "loss": 1.1409, "step": 9030 }, { "epoch": 1.895034569453174, "grad_norm": 2.290170192718506, "learning_rate": 7.245995926521854e-06, "loss": 1.1156, "step": 9045 }, { "epoch": 1.8981772470144564, "grad_norm": 2.2289717197418213, "learning_rate": 7.2108614798864585e-06, "loss": 1.0908, "step": 9060 }, { "epoch": 1.9013199245757386, "grad_norm": 2.7847399711608887, "learning_rate": 7.17576432804451e-06, "loss": 1.1137, "step": 9075 }, { "epoch": 1.9044626021370208, "grad_norm": 2.2532174587249756, "learning_rate": 7.140704940295404e-06, "loss": 1.0993, "step": 9090 }, { "epoch": 1.907605279698303, "grad_norm": 3.097644329071045, "learning_rate": 7.1056837854335804e-06, "loss": 1.1322, "step": 9105 }, { "epoch": 1.9107479572595851, "grad_norm": 2.270334005355835, "learning_rate": 7.0707013317422446e-06, "loss": 1.0477, "step": 9120 }, { "epoch": 1.9138906348208673, "grad_norm": 2.698101282119751, "learning_rate": 7.035758046987122e-06, "loss": 1.0722, "step": 9135 }, { "epoch": 1.9170333123821495, "grad_norm": 2.2851641178131104, "learning_rate": 7.000854398410182e-06, "loss": 1.1588, "step": 9150 }, { "epoch": 1.9201759899434316, "grad_norm": 2.714940071105957, "learning_rate": 6.965990852723404e-06, "loss": 1.0664, "step": 9165 }, { "epoch": 1.923318667504714, "grad_norm": 2.079648971557617, "learning_rate": 6.9311678761025416e-06, "loss": 1.0701, "step": 9180 }, { "epoch": 1.9264613450659962, "grad_norm": 1.5704066753387451, "learning_rate": 6.896385934180867e-06, "loss": 1.085, "step": 9195 }, { "epoch": 1.9296040226272786, "grad_norm": 2.2583060264587402, "learning_rate": 6.861645492042977e-06, "loss": 1.1032, "step": 9210 }, { "epoch": 1.9327467001885608, "grad_norm": 2.0253028869628906, "learning_rate": 6.826947014218543e-06, "loss": 1.1232, "step": 9225 }, { "epoch": 1.935889377749843, "grad_norm": 3.0294480323791504, "learning_rate": 6.792290964676123e-06, "loss": 1.1583, "step": 9240 }, { "epoch": 1.939032055311125, "grad_norm": 2.480926036834717, "learning_rate": 6.7576778068169335e-06, "loss": 1.126, "step": 9255 }, { "epoch": 1.9421747328724073, "grad_norm": 2.393428087234497, "learning_rate": 6.725411298947324e-06, "loss": 1.0878, "step": 9270 }, { "epoch": 1.9453174104336894, "grad_norm": 2.0476040840148926, "learning_rate": 6.690882376871407e-06, "loss": 1.0794, "step": 9285 }, { "epoch": 1.9484600879949716, "grad_norm": 2.075193166732788, "learning_rate": 6.656397702457396e-06, "loss": 1.0825, "step": 9300 }, { "epoch": 1.9516027655562538, "grad_norm": 2.0196661949157715, "learning_rate": 6.621957736814983e-06, "loss": 1.0755, "step": 9315 }, { "epoch": 1.9547454431175362, "grad_norm": 2.4812753200531006, "learning_rate": 6.58756294045604e-06, "loss": 1.0883, "step": 9330 }, { "epoch": 1.9578881206788183, "grad_norm": 2.0828120708465576, "learning_rate": 6.553213773288465e-06, "loss": 1.1287, "step": 9345 }, { "epoch": 1.9610307982401005, "grad_norm": 2.273803949356079, "learning_rate": 6.518910694610023e-06, "loss": 1.0481, "step": 9360 }, { "epoch": 1.964173475801383, "grad_norm": 3.1020278930664062, "learning_rate": 6.4846541631022155e-06, "loss": 1.078, "step": 9375 }, { "epoch": 1.967316153362665, "grad_norm": 2.7231500148773193, "learning_rate": 6.4504446368241315e-06, "loss": 1.0389, "step": 9390 }, { "epoch": 1.9704588309239472, "grad_norm": 2.4498469829559326, "learning_rate": 6.416282573206341e-06, "loss": 1.1052, "step": 9405 }, { "epoch": 1.9736015084852294, "grad_norm": 2.266944408416748, "learning_rate": 6.382168429044769e-06, "loss": 1.1163, "step": 9420 }, { "epoch": 1.9767441860465116, "grad_norm": 2.877861976623535, "learning_rate": 6.34810266049459e-06, "loss": 1.1106, "step": 9435 }, { "epoch": 1.9798868636077938, "grad_norm": 2.256692886352539, "learning_rate": 6.314085723064118e-06, "loss": 1.0929, "step": 9450 }, { "epoch": 1.983029541169076, "grad_norm": 2.337369203567505, "learning_rate": 6.2801180716087315e-06, "loss": 1.1019, "step": 9465 }, { "epoch": 1.9861722187303583, "grad_norm": 3.803745746612549, "learning_rate": 6.246200160324789e-06, "loss": 1.0641, "step": 9480 }, { "epoch": 1.9893148962916405, "grad_norm": 2.354900598526001, "learning_rate": 6.21233244274354e-06, "loss": 1.1353, "step": 9495 }, { "epoch": 1.9924575738529227, "grad_norm": 2.0362517833709717, "learning_rate": 6.178515371725083e-06, "loss": 1.0638, "step": 9510 }, { "epoch": 1.995600251414205, "grad_norm": 1.7420244216918945, "learning_rate": 6.144749399452294e-06, "loss": 1.0539, "step": 9525 }, { "epoch": 1.9987429289754872, "grad_norm": 1.9186304807662964, "learning_rate": 6.11103497742478e-06, "loss": 1.0553, "step": 9540 }, { "epoch": 2.0018856065367694, "grad_norm": 1.9180148839950562, "learning_rate": 6.077372556452858e-06, "loss": 1.078, "step": 9555 }, { "epoch": 2.0050282840980516, "grad_norm": 1.8619567155838013, "learning_rate": 6.043762586651511e-06, "loss": 1.0238, "step": 9570 }, { "epoch": 2.0081709616593337, "grad_norm": 2.054584503173828, "learning_rate": 6.010205517434373e-06, "loss": 1.0361, "step": 9585 }, { "epoch": 2.011313639220616, "grad_norm": 1.6725273132324219, "learning_rate": 5.97670179750772e-06, "loss": 1.0048, "step": 9600 }, { "epoch": 2.014456316781898, "grad_norm": 2.8712847232818604, "learning_rate": 5.943251874864476e-06, "loss": 1.0199, "step": 9615 }, { "epoch": 2.0175989943431802, "grad_norm": 2.331249237060547, "learning_rate": 5.9098561967782165e-06, "loss": 1.0697, "step": 9630 }, { "epoch": 2.0207416719044624, "grad_norm": 2.359811782836914, "learning_rate": 5.876515209797189e-06, "loss": 1.0052, "step": 9645 }, { "epoch": 2.023884349465745, "grad_norm": 3.5132248401641846, "learning_rate": 5.843229359738336e-06, "loss": 1.0845, "step": 9660 }, { "epoch": 2.027027027027027, "grad_norm": 1.9716084003448486, "learning_rate": 5.809999091681349e-06, "loss": 1.1177, "step": 9675 }, { "epoch": 2.0301697045883094, "grad_norm": 2.3846986293792725, "learning_rate": 5.776824849962706e-06, "loss": 1.0848, "step": 9690 }, { "epoch": 2.0333123821495915, "grad_norm": 1.9413962364196777, "learning_rate": 5.743707078169731e-06, "loss": 1.0824, "step": 9705 }, { "epoch": 2.0364550597108737, "grad_norm": 1.9300832748413086, "learning_rate": 5.710646219134666e-06, "loss": 1.0474, "step": 9720 }, { "epoch": 2.039597737272156, "grad_norm": 2.1981208324432373, "learning_rate": 5.67764271492874e-06, "loss": 1.1053, "step": 9735 }, { "epoch": 2.042740414833438, "grad_norm": 2.165830612182617, "learning_rate": 5.644697006856268e-06, "loss": 1.0203, "step": 9750 }, { "epoch": 2.04588309239472, "grad_norm": 2.567518711090088, "learning_rate": 5.61180953544876e-06, "loss": 1.0509, "step": 9765 }, { "epoch": 2.0490257699560024, "grad_norm": 2.09709095954895, "learning_rate": 5.578980740459009e-06, "loss": 1.0906, "step": 9780 }, { "epoch": 2.0521684475172846, "grad_norm": 1.5784721374511719, "learning_rate": 5.546211060855211e-06, "loss": 1.0694, "step": 9795 }, { "epoch": 2.0553111250785667, "grad_norm": 3.4078545570373535, "learning_rate": 5.513500934815113e-06, "loss": 1.0352, "step": 9810 }, { "epoch": 2.0584538026398493, "grad_norm": 2.0774762630462646, "learning_rate": 5.4808507997201544e-06, "loss": 1.0513, "step": 9825 }, { "epoch": 2.0615964802011315, "grad_norm": 3.19948148727417, "learning_rate": 5.448261092149596e-06, "loss": 1.0432, "step": 9840 }, { "epoch": 2.0647391577624137, "grad_norm": 2.4168691635131836, "learning_rate": 5.415732247874696e-06, "loss": 1.0354, "step": 9855 }, { "epoch": 2.067881835323696, "grad_norm": 1.7410550117492676, "learning_rate": 5.383264701852886e-06, "loss": 1.0389, "step": 9870 }, { "epoch": 2.071024512884978, "grad_norm": 2.2156288623809814, "learning_rate": 5.350858888221957e-06, "loss": 1.0614, "step": 9885 }, { "epoch": 2.07416719044626, "grad_norm": 3.3456056118011475, "learning_rate": 5.318515240294241e-06, "loss": 1.099, "step": 9900 }, { "epoch": 2.0773098680075424, "grad_norm": 3.286379814147949, "learning_rate": 5.28623419055083e-06, "loss": 1.0667, "step": 9915 }, { "epoch": 2.0804525455688245, "grad_norm": 2.086221218109131, "learning_rate": 5.2540161706357855e-06, "loss": 1.0458, "step": 9930 }, { "epoch": 2.0835952231301067, "grad_norm": 3.3937273025512695, "learning_rate": 5.221861611350371e-06, "loss": 1.0696, "step": 9945 }, { "epoch": 2.086737900691389, "grad_norm": 2.162592887878418, "learning_rate": 5.189770942647291e-06, "loss": 1.034, "step": 9960 }, { "epoch": 2.0898805782526715, "grad_norm": 2.144193649291992, "learning_rate": 5.157744593624939e-06, "loss": 1.0997, "step": 9975 }, { "epoch": 2.0930232558139537, "grad_norm": 2.1846065521240234, "learning_rate": 5.125782992521664e-06, "loss": 1.034, "step": 9990 }, { "epoch": 2.0951183741881416, "eval_accuracy": 0.8874096274375916, "eval_loss": 1.1195236444473267, "eval_runtime": 3488.4257, "eval_samples_per_second": 1.372, "eval_steps_per_second": 0.343, "step": 10000 }, { "epoch": 2.096165933375236, "grad_norm": 1.9800641536712646, "learning_rate": 5.09388656671004e-06, "loss": 1.0096, "step": 10005 }, { "epoch": 2.099308610936518, "grad_norm": 2.919948101043701, "learning_rate": 5.062055742691154e-06, "loss": 1.0455, "step": 10020 }, { "epoch": 2.1024512884978, "grad_norm": 2.2051897048950195, "learning_rate": 5.0302909460889045e-06, "loss": 1.0675, "step": 10035 }, { "epoch": 2.1055939660590823, "grad_norm": 3.871448278427124, "learning_rate": 4.998592601644306e-06, "loss": 1.0813, "step": 10050 }, { "epoch": 2.1087366436203645, "grad_norm": 2.2782132625579834, "learning_rate": 4.9669611332098155e-06, "loss": 1.0833, "step": 10065 }, { "epoch": 2.1118793211816467, "grad_norm": 2.384378433227539, "learning_rate": 4.935396963743661e-06, "loss": 1.1207, "step": 10080 }, { "epoch": 2.115021998742929, "grad_norm": 2.561387300491333, "learning_rate": 4.9039005153041755e-06, "loss": 1.0639, "step": 10095 }, { "epoch": 2.118164676304211, "grad_norm": 3.0452916622161865, "learning_rate": 4.872472209044182e-06, "loss": 1.0579, "step": 10110 }, { "epoch": 2.121307353865493, "grad_norm": 2.2836525440216064, "learning_rate": 4.841112465205337e-06, "loss": 1.0984, "step": 10125 }, { "epoch": 2.124450031426776, "grad_norm": 3.4332149028778076, "learning_rate": 4.809821703112518e-06, "loss": 1.0762, "step": 10140 }, { "epoch": 2.127592708988058, "grad_norm": 2.4914424419403076, "learning_rate": 4.778600341168208e-06, "loss": 1.0371, "step": 10155 }, { "epoch": 2.13073538654934, "grad_norm": 3.004932165145874, "learning_rate": 4.747448796846932e-06, "loss": 1.084, "step": 10170 }, { "epoch": 2.1338780641106223, "grad_norm": 2.9469964504241943, "learning_rate": 4.716367486689636e-06, "loss": 1.0769, "step": 10185 }, { "epoch": 2.1370207416719045, "grad_norm": 1.8354699611663818, "learning_rate": 4.685356826298143e-06, "loss": 1.0308, "step": 10200 }, { "epoch": 2.1401634192331866, "grad_norm": 1.9137917757034302, "learning_rate": 4.654417230329575e-06, "loss": 1.0568, "step": 10215 }, { "epoch": 2.143306096794469, "grad_norm": 2.0806703567504883, "learning_rate": 4.6235491124908295e-06, "loss": 1.0535, "step": 10230 }, { "epoch": 2.146448774355751, "grad_norm": 3.0409634113311768, "learning_rate": 4.5927528855330425e-06, "loss": 1.0421, "step": 10245 }, { "epoch": 2.149591451917033, "grad_norm": 2.494769811630249, "learning_rate": 4.562028961246065e-06, "loss": 1.0494, "step": 10260 }, { "epoch": 2.1527341294783153, "grad_norm": 2.0933990478515625, "learning_rate": 4.5334188940857585e-06, "loss": 1.1007, "step": 10275 }, { "epoch": 2.155876807039598, "grad_norm": 2.1855649948120117, "learning_rate": 4.502835919016285e-06, "loss": 1.1062, "step": 10290 }, { "epoch": 2.15901948460088, "grad_norm": 1.9125105142593384, "learning_rate": 4.472326448936779e-06, "loss": 1.0911, "step": 10305 }, { "epoch": 2.1621621621621623, "grad_norm": 2.457319736480713, "learning_rate": 4.441890891802735e-06, "loss": 1.0553, "step": 10320 }, { "epoch": 2.1653048397234445, "grad_norm": 1.8059813976287842, "learning_rate": 4.411529654581325e-06, "loss": 1.0426, "step": 10335 }, { "epoch": 2.1684475172847266, "grad_norm": 1.840040922164917, "learning_rate": 4.381243143245957e-06, "loss": 1.0701, "step": 10350 }, { "epoch": 2.171590194846009, "grad_norm": 2.875662326812744, "learning_rate": 4.351031762770848e-06, "loss": 1.0217, "step": 10365 }, { "epoch": 2.174732872407291, "grad_norm": 2.7019827365875244, "learning_rate": 4.320895917125604e-06, "loss": 1.0767, "step": 10380 }, { "epoch": 2.177875549968573, "grad_norm": 2.331372022628784, "learning_rate": 4.290836009269824e-06, "loss": 1.0414, "step": 10395 }, { "epoch": 2.1810182275298553, "grad_norm": 1.7536449432373047, "learning_rate": 4.260852441147711e-06, "loss": 1.0328, "step": 10410 }, { "epoch": 2.1841609050911375, "grad_norm": 2.233278512954712, "learning_rate": 4.230945613682696e-06, "loss": 1.0464, "step": 10425 }, { "epoch": 2.1873035826524196, "grad_norm": 2.3253679275512695, "learning_rate": 4.2011159267720645e-06, "loss": 1.0734, "step": 10440 }, { "epoch": 2.1904462602137023, "grad_norm": 2.1976492404937744, "learning_rate": 4.171363779281644e-06, "loss": 1.0089, "step": 10455 }, { "epoch": 2.1935889377749844, "grad_norm": 1.6631523370742798, "learning_rate": 4.141689569040433e-06, "loss": 1.0598, "step": 10470 }, { "epoch": 2.1967316153362666, "grad_norm": 2.9074153900146484, "learning_rate": 4.112093692835306e-06, "loss": 1.0564, "step": 10485 }, { "epoch": 2.1998742928975488, "grad_norm": 2.0175225734710693, "learning_rate": 4.082576546405683e-06, "loss": 1.009, "step": 10500 }, { "epoch": 2.203016970458831, "grad_norm": 2.74937105178833, "learning_rate": 4.053138524438268e-06, "loss": 1.0614, "step": 10515 }, { "epoch": 2.206159648020113, "grad_norm": 1.9003492593765259, "learning_rate": 4.0237800205617585e-06, "loss": 1.0904, "step": 10530 }, { "epoch": 2.2093023255813953, "grad_norm": 2.6317954063415527, "learning_rate": 3.994501427341575e-06, "loss": 1.0716, "step": 10545 }, { "epoch": 2.2124450031426774, "grad_norm": 1.7894374132156372, "learning_rate": 3.965303136274614e-06, "loss": 1.1267, "step": 10560 }, { "epoch": 2.2155876807039596, "grad_norm": 2.2193076610565186, "learning_rate": 3.936185537784019e-06, "loss": 1.0509, "step": 10575 }, { "epoch": 2.218730358265242, "grad_norm": 2.435281753540039, "learning_rate": 3.9071490212139665e-06, "loss": 1.0699, "step": 10590 }, { "epoch": 2.2218730358265244, "grad_norm": 2.5646512508392334, "learning_rate": 3.878193974824444e-06, "loss": 1.0761, "step": 10605 }, { "epoch": 2.2250157133878066, "grad_norm": 2.3595800399780273, "learning_rate": 3.849320785786062e-06, "loss": 1.0842, "step": 10620 }, { "epoch": 2.2281583909490887, "grad_norm": 2.8445732593536377, "learning_rate": 3.820529840174885e-06, "loss": 1.0093, "step": 10635 }, { "epoch": 2.231301068510371, "grad_norm": 2.228774309158325, "learning_rate": 3.7918215229672606e-06, "loss": 1.1004, "step": 10650 }, { "epoch": 2.234443746071653, "grad_norm": 2.617530584335327, "learning_rate": 3.7631962180346905e-06, "loss": 1.0769, "step": 10665 }, { "epoch": 2.2375864236329353, "grad_norm": 2.4638450145721436, "learning_rate": 3.73465430813866e-06, "loss": 1.0703, "step": 10680 }, { "epoch": 2.2407291011942174, "grad_norm": 2.0117123126983643, "learning_rate": 3.706196174925558e-06, "loss": 1.0354, "step": 10695 }, { "epoch": 2.2438717787554996, "grad_norm": 2.4193451404571533, "learning_rate": 3.6778221989215537e-06, "loss": 1.0742, "step": 10710 }, { "epoch": 2.2470144563167818, "grad_norm": 2.2464051246643066, "learning_rate": 3.6495327595275223e-06, "loss": 1.0563, "step": 10725 }, { "epoch": 2.250157133878064, "grad_norm": 2.0514557361602783, "learning_rate": 3.6213282350139455e-06, "loss": 1.078, "step": 10740 }, { "epoch": 2.253299811439346, "grad_norm": 3.3520901203155518, "learning_rate": 3.593209002515885e-06, "loss": 1.0656, "step": 10755 }, { "epoch": 2.2564424890006287, "grad_norm": 1.7619084119796753, "learning_rate": 3.56517543802792e-06, "loss": 1.0591, "step": 10770 }, { "epoch": 2.259585166561911, "grad_norm": 2.095613718032837, "learning_rate": 3.5372279163991285e-06, "loss": 1.0567, "step": 10785 }, { "epoch": 2.262727844123193, "grad_norm": 2.4718666076660156, "learning_rate": 3.5093668113280688e-06, "loss": 1.0847, "step": 10800 }, { "epoch": 2.2658705216844752, "grad_norm": 3.0717573165893555, "learning_rate": 3.481592495357786e-06, "loss": 1.125, "step": 10815 }, { "epoch": 2.2690131992457574, "grad_norm": 2.8368752002716064, "learning_rate": 3.4539053398708313e-06, "loss": 1.0832, "step": 10830 }, { "epoch": 2.2721558768070396, "grad_norm": 2.2466132640838623, "learning_rate": 3.4263057150842927e-06, "loss": 1.0572, "step": 10845 }, { "epoch": 2.2752985543683217, "grad_norm": 2.3740978240966797, "learning_rate": 3.398793990044851e-06, "loss": 1.0886, "step": 10860 }, { "epoch": 2.278441231929604, "grad_norm": 2.0145745277404785, "learning_rate": 3.3713705326238354e-06, "loss": 0.9999, "step": 10875 }, { "epoch": 2.281583909490886, "grad_norm": 2.8063840866088867, "learning_rate": 3.3440357095123142e-06, "loss": 1.029, "step": 10890 }, { "epoch": 2.2847265870521687, "grad_norm": 2.4196486473083496, "learning_rate": 3.3167898862161896e-06, "loss": 1.0736, "step": 10905 }, { "epoch": 2.287869264613451, "grad_norm": 2.1821863651275635, "learning_rate": 3.289633427051295e-06, "loss": 1.0106, "step": 10920 }, { "epoch": 2.291011942174733, "grad_norm": 2.0182254314422607, "learning_rate": 3.262566695138557e-06, "loss": 1.0717, "step": 10935 }, { "epoch": 2.294154619736015, "grad_norm": 3.4760282039642334, "learning_rate": 3.2355900523991057e-06, "loss": 1.081, "step": 10950 }, { "epoch": 2.2972972972972974, "grad_norm": 1.5960103273391724, "learning_rate": 3.2087038595494567e-06, "loss": 0.9858, "step": 10965 }, { "epoch": 2.3004399748585795, "grad_norm": 2.3802037239074707, "learning_rate": 3.18190847609667e-06, "loss": 1.1179, "step": 10980 }, { "epoch": 2.3035826524198617, "grad_norm": 2.2108407020568848, "learning_rate": 3.155204260333561e-06, "loss": 1.0281, "step": 10995 }, { "epoch": 2.306725329981144, "grad_norm": 2.4052934646606445, "learning_rate": 3.1285915693339074e-06, "loss": 1.0728, "step": 11010 }, { "epoch": 2.309868007542426, "grad_norm": 2.564016342163086, "learning_rate": 3.1020707589476638e-06, "loss": 1.0539, "step": 11025 }, { "epoch": 2.313010685103708, "grad_norm": 1.9333407878875732, "learning_rate": 3.0756421837962013e-06, "loss": 1.0316, "step": 11040 }, { "epoch": 2.3161533626649904, "grad_norm": 2.317944049835205, "learning_rate": 3.0493061972675808e-06, "loss": 1.0568, "step": 11055 }, { "epoch": 2.3192960402262726, "grad_norm": 1.7732441425323486, "learning_rate": 3.0230631515118236e-06, "loss": 1.0107, "step": 11070 }, { "epoch": 2.322438717787555, "grad_norm": 2.109452724456787, "learning_rate": 2.9969133974361985e-06, "loss": 1.0543, "step": 11085 }, { "epoch": 2.3255813953488373, "grad_norm": 2.198519468307495, "learning_rate": 2.970857284700517e-06, "loss": 1.0958, "step": 11100 }, { "epoch": 2.3287240729101195, "grad_norm": 2.7568089962005615, "learning_rate": 2.9448951617124878e-06, "loss": 1.0141, "step": 11115 }, { "epoch": 2.3318667504714017, "grad_norm": 3.3935024738311768, "learning_rate": 2.919027375623028e-06, "loss": 1.0298, "step": 11130 }, { "epoch": 2.335009428032684, "grad_norm": 2.245539903640747, "learning_rate": 2.893254272321655e-06, "loss": 1.0677, "step": 11145 }, { "epoch": 2.338152105593966, "grad_norm": 3.133040428161621, "learning_rate": 2.8675761964318148e-06, "loss": 1.0253, "step": 11160 }, { "epoch": 2.341294783155248, "grad_norm": 2.1391761302948, "learning_rate": 2.8419934913063175e-06, "loss": 1.0297, "step": 11175 }, { "epoch": 2.3444374607165304, "grad_norm": 3.1094396114349365, "learning_rate": 2.8165064990227255e-06, "loss": 1.0369, "step": 11190 }, { "epoch": 2.3475801382778125, "grad_norm": 3.4849114418029785, "learning_rate": 2.791115560378781e-06, "loss": 1.0666, "step": 11205 }, { "epoch": 2.350722815839095, "grad_norm": 2.6266283988952637, "learning_rate": 2.7658210148878515e-06, "loss": 1.1017, "step": 11220 }, { "epoch": 2.3538654934003773, "grad_norm": 3.290055990219116, "learning_rate": 2.7406232007743892e-06, "loss": 1.017, "step": 11235 }, { "epoch": 2.3570081709616595, "grad_norm": 3.80057954788208, "learning_rate": 2.7155224549694114e-06, "loss": 1.0566, "step": 11250 }, { "epoch": 2.3601508485229417, "grad_norm": 2.253140926361084, "learning_rate": 2.690519113105988e-06, "loss": 1.1236, "step": 11265 }, { "epoch": 2.363293526084224, "grad_norm": 2.4676170349121094, "learning_rate": 2.667270835657341e-06, "loss": 1.0442, "step": 11280 }, { "epoch": 2.366436203645506, "grad_norm": 2.411430597305298, "learning_rate": 2.6424567549411838e-06, "loss": 1.0645, "step": 11295 }, { "epoch": 2.369578881206788, "grad_norm": 2.473489284515381, "learning_rate": 2.6177410551600802e-06, "loss": 1.1046, "step": 11310 }, { "epoch": 2.3727215587680703, "grad_norm": 2.323939800262451, "learning_rate": 2.593124066798477e-06, "loss": 1.0366, "step": 11325 }, { "epoch": 2.3758642363293525, "grad_norm": 2.2324061393737793, "learning_rate": 2.5686061190209134e-06, "loss": 1.036, "step": 11340 }, { "epoch": 2.3790069138906347, "grad_norm": 3.393950939178467, "learning_rate": 2.5441875396676186e-06, "loss": 1.0795, "step": 11355 }, { "epoch": 2.382149591451917, "grad_norm": 2.0376970767974854, "learning_rate": 2.5198686552501196e-06, "loss": 1.0682, "step": 11370 }, { "epoch": 2.385292269013199, "grad_norm": 2.5424187183380127, "learning_rate": 2.4956497909468713e-06, "loss": 1.0599, "step": 11385 }, { "epoch": 2.3884349465744816, "grad_norm": 2.3934738636016846, "learning_rate": 2.4715312705989236e-06, "loss": 1.0725, "step": 11400 }, { "epoch": 2.391577624135764, "grad_norm": 2.5349321365356445, "learning_rate": 2.447513416705579e-06, "loss": 0.977, "step": 11415 }, { "epoch": 2.394720301697046, "grad_norm": 2.4864916801452637, "learning_rate": 2.4235965504200974e-06, "loss": 1.0452, "step": 11430 }, { "epoch": 2.397862979258328, "grad_norm": 2.2827324867248535, "learning_rate": 2.3997809915453686e-06, "loss": 1.0459, "step": 11445 }, { "epoch": 2.4010056568196103, "grad_norm": 2.4698305130004883, "learning_rate": 2.376067058529672e-06, "loss": 1.0619, "step": 11460 }, { "epoch": 2.4041483343808925, "grad_norm": 2.9925546646118164, "learning_rate": 2.3524550684623927e-06, "loss": 1.0541, "step": 11475 }, { "epoch": 2.4072910119421747, "grad_norm": 2.6975173950195312, "learning_rate": 2.328945337069809e-06, "loss": 1.1299, "step": 11490 }, { "epoch": 2.410433689503457, "grad_norm": 2.239131212234497, "learning_rate": 2.305538178710831e-06, "loss": 1.0085, "step": 11505 }, { "epoch": 2.413576367064739, "grad_norm": 2.0249500274658203, "learning_rate": 2.282233906372835e-06, "loss": 1.0513, "step": 11520 }, { "epoch": 2.4167190446260216, "grad_norm": 2.393962860107422, "learning_rate": 2.2590328316674582e-06, "loss": 1.0077, "step": 11535 }, { "epoch": 2.4198617221873038, "grad_norm": 3.019960880279541, "learning_rate": 2.2359352648264387e-06, "loss": 1.0446, "step": 11550 }, { "epoch": 2.423004399748586, "grad_norm": 2.6998488903045654, "learning_rate": 2.2129415146974665e-06, "loss": 1.0773, "step": 11565 }, { "epoch": 2.426147077309868, "grad_norm": 3.3861067295074463, "learning_rate": 2.190051888740048e-06, "loss": 1.103, "step": 11580 }, { "epoch": 2.4292897548711503, "grad_norm": 2.977205514907837, "learning_rate": 2.1672666930214047e-06, "loss": 1.0307, "step": 11595 }, { "epoch": 2.4324324324324325, "grad_norm": 3.0046980381011963, "learning_rate": 2.1445862322123734e-06, "loss": 1.0402, "step": 11610 }, { "epoch": 2.4355751099937146, "grad_norm": 2.431718111038208, "learning_rate": 2.1220108095833345e-06, "loss": 1.0802, "step": 11625 }, { "epoch": 2.438717787554997, "grad_norm": 1.7797611951828003, "learning_rate": 2.099540727000159e-06, "loss": 1.0833, "step": 11640 }, { "epoch": 2.441860465116279, "grad_norm": 2.224626302719116, "learning_rate": 2.077176284920166e-06, "loss": 1.0358, "step": 11655 }, { "epoch": 2.445003142677561, "grad_norm": 2.7776806354522705, "learning_rate": 2.054917782388113e-06, "loss": 1.0251, "step": 11670 }, { "epoch": 2.4481458202388433, "grad_norm": 2.1015563011169434, "learning_rate": 2.0327655170321915e-06, "loss": 1.0326, "step": 11685 }, { "epoch": 2.4512884978001255, "grad_norm": 3.1175501346588135, "learning_rate": 2.010719785060048e-06, "loss": 1.0697, "step": 11700 }, { "epoch": 2.454431175361408, "grad_norm": 2.0512046813964844, "learning_rate": 1.9887808812548272e-06, "loss": 1.063, "step": 11715 }, { "epoch": 2.4575738529226903, "grad_norm": 2.4269354343414307, "learning_rate": 1.966949098971227e-06, "loss": 1.0623, "step": 11730 }, { "epoch": 2.4607165304839724, "grad_norm": 2.026348352432251, "learning_rate": 1.945224730131576e-06, "loss": 1.0249, "step": 11745 }, { "epoch": 2.4638592080452546, "grad_norm": 2.1796414852142334, "learning_rate": 1.9236080652219248e-06, "loss": 1.0627, "step": 11760 }, { "epoch": 2.4670018856065368, "grad_norm": 2.8909108638763428, "learning_rate": 1.9020993932881803e-06, "loss": 0.9954, "step": 11775 }, { "epoch": 2.470144563167819, "grad_norm": 1.939376950263977, "learning_rate": 1.8806990019322169e-06, "loss": 0.9819, "step": 11790 }, { "epoch": 2.473287240729101, "grad_norm": 2.434191942214966, "learning_rate": 1.8594071773080492e-06, "loss": 1.0858, "step": 11805 }, { "epoch": 2.4764299182903833, "grad_norm": 2.3099939823150635, "learning_rate": 1.8382242041179876e-06, "loss": 1.0571, "step": 11820 }, { "epoch": 2.4795725958516655, "grad_norm": 2.7791974544525146, "learning_rate": 1.8171503656088574e-06, "loss": 1.0598, "step": 11835 }, { "epoch": 2.482715273412948, "grad_norm": 2.9822537899017334, "learning_rate": 1.7961859435681882e-06, "loss": 1.029, "step": 11850 }, { "epoch": 2.4858579509742302, "grad_norm": 2.644841194152832, "learning_rate": 1.7753312183204486e-06, "loss": 1.12, "step": 11865 }, { "epoch": 2.4890006285355124, "grad_norm": 2.436312437057495, "learning_rate": 1.7545864687233127e-06, "loss": 1.0353, "step": 11880 }, { "epoch": 2.4921433060967946, "grad_norm": 2.3745834827423096, "learning_rate": 1.7339519721639164e-06, "loss": 1.0758, "step": 11895 }, { "epoch": 2.4952859836580767, "grad_norm": 2.9353575706481934, "learning_rate": 1.713428004555162e-06, "loss": 1.0455, "step": 11910 }, { "epoch": 2.498428661219359, "grad_norm": 2.2262260913848877, "learning_rate": 1.693014840332009e-06, "loss": 1.0345, "step": 11925 }, { "epoch": 2.501571338780641, "grad_norm": 2.1293249130249023, "learning_rate": 1.6727127524478227e-06, "loss": 1.0262, "step": 11940 }, { "epoch": 2.5047140163419233, "grad_norm": 2.169814348220825, "learning_rate": 1.6525220123707176e-06, "loss": 1.0121, "step": 11955 }, { "epoch": 2.5078566939032054, "grad_norm": 2.504615545272827, "learning_rate": 1.6324428900799273e-06, "loss": 1.0202, "step": 11970 }, { "epoch": 2.5109993714644876, "grad_norm": 2.8434998989105225, "learning_rate": 1.6124756540621966e-06, "loss": 1.0042, "step": 11985 }, { "epoch": 2.5141420490257698, "grad_norm": 1.939417839050293, "learning_rate": 1.5926205713081876e-06, "loss": 1.1108, "step": 12000 }, { "epoch": 2.517284726587052, "grad_norm": 2.569380283355713, "learning_rate": 1.572877907308913e-06, "loss": 1.0067, "step": 12015 }, { "epoch": 2.520427404148334, "grad_norm": 2.747483253479004, "learning_rate": 1.5532479260521849e-06, "loss": 1.0509, "step": 12030 }, { "epoch": 2.5235700817096167, "grad_norm": 1.7206883430480957, "learning_rate": 1.533730890019086e-06, "loss": 1.0901, "step": 12045 }, { "epoch": 2.526712759270899, "grad_norm": 2.222593069076538, "learning_rate": 1.5143270601804582e-06, "loss": 1.0182, "step": 12060 }, { "epoch": 2.529855436832181, "grad_norm": 2.6068830490112305, "learning_rate": 1.4950366959934116e-06, "loss": 1.0768, "step": 12075 }, { "epoch": 2.5329981143934632, "grad_norm": 2.674351692199707, "learning_rate": 1.475860055397862e-06, "loss": 1.0114, "step": 12090 }, { "epoch": 2.5361407919547454, "grad_norm": 3.2743771076202393, "learning_rate": 1.4567973948130731e-06, "loss": 1.0805, "step": 12105 }, { "epoch": 2.5392834695160276, "grad_norm": 3.0732719898223877, "learning_rate": 1.4378489691342345e-06, "loss": 1.0802, "step": 12120 }, { "epoch": 2.5424261470773097, "grad_norm": 2.502260446548462, "learning_rate": 1.4190150317290485e-06, "loss": 1.0701, "step": 12135 }, { "epoch": 2.5455688246385924, "grad_norm": 2.1941142082214355, "learning_rate": 1.4002958344343465e-06, "loss": 1.0671, "step": 12150 }, { "epoch": 2.5487115021998745, "grad_norm": 2.5131101608276367, "learning_rate": 1.3816916275527193e-06, "loss": 1.0483, "step": 12165 }, { "epoch": 2.5518541797611567, "grad_norm": 2.770385503768921, "learning_rate": 1.36320265984917e-06, "loss": 1.0708, "step": 12180 }, { "epoch": 2.554996857322439, "grad_norm": 2.3241212368011475, "learning_rate": 1.3448291785477874e-06, "loss": 1.0521, "step": 12195 }, { "epoch": 2.558139534883721, "grad_norm": 3.1057024002075195, "learning_rate": 1.3265714293284414e-06, "loss": 1.0705, "step": 12210 }, { "epoch": 2.561282212445003, "grad_norm": 1.8851613998413086, "learning_rate": 1.3084296563235033e-06, "loss": 1.037, "step": 12225 }, { "epoch": 2.5644248900062854, "grad_norm": 2.893693685531616, "learning_rate": 1.2904041021145597e-06, "loss": 1.0582, "step": 12240 }, { "epoch": 2.5675675675675675, "grad_norm": 2.2626419067382812, "learning_rate": 1.2724950077292042e-06, "loss": 1.0213, "step": 12255 }, { "epoch": 2.5707102451288497, "grad_norm": 2.1927521228790283, "learning_rate": 1.2547026126377849e-06, "loss": 1.0542, "step": 12270 }, { "epoch": 2.573852922690132, "grad_norm": 2.205655097961426, "learning_rate": 1.2382018758160751e-06, "loss": 1.0494, "step": 12285 }, { "epoch": 2.576995600251414, "grad_norm": 2.98274827003479, "learning_rate": 1.2206357725870321e-06, "loss": 1.0926, "step": 12300 }, { "epoch": 2.5801382778126962, "grad_norm": 2.7129874229431152, "learning_rate": 1.2031870620844844e-06, "loss": 1.0396, "step": 12315 }, { "epoch": 2.5832809553739784, "grad_norm": 2.8410415649414062, "learning_rate": 1.1858559776227863e-06, "loss": 1.0353, "step": 12330 }, { "epoch": 2.586423632935261, "grad_norm": 2.896530866622925, "learning_rate": 1.1686427509434617e-06, "loss": 1.1285, "step": 12345 }, { "epoch": 2.589566310496543, "grad_norm": 2.223740339279175, "learning_rate": 1.1515476122121094e-06, "loss": 1.0727, "step": 12360 }, { "epoch": 2.5927089880578253, "grad_norm": 2.6268301010131836, "learning_rate": 1.1345707900153203e-06, "loss": 1.016, "step": 12375 }, { "epoch": 2.5958516656191075, "grad_norm": 2.7400355339050293, "learning_rate": 1.1177125113576237e-06, "loss": 1.0317, "step": 12390 }, { "epoch": 2.5989943431803897, "grad_norm": 2.5901083946228027, "learning_rate": 1.100973001658453e-06, "loss": 1.0526, "step": 12405 }, { "epoch": 2.602137020741672, "grad_norm": 2.201157569885254, "learning_rate": 1.084352484749126e-06, "loss": 1.0881, "step": 12420 }, { "epoch": 2.605279698302954, "grad_norm": 3.0149240493774414, "learning_rate": 1.067851182869859e-06, "loss": 1.02, "step": 12435 }, { "epoch": 2.608422375864236, "grad_norm": 2.6754496097564697, "learning_rate": 1.051469316666791e-06, "loss": 1.0289, "step": 12450 }, { "epoch": 2.611565053425519, "grad_norm": 2.752760648727417, "learning_rate": 1.0352071051890333e-06, "loss": 1.0687, "step": 12465 }, { "epoch": 2.614707730986801, "grad_norm": 2.3588600158691406, "learning_rate": 1.0190647658857433e-06, "loss": 1.0514, "step": 12480 }, { "epoch": 2.617850408548083, "grad_norm": 2.367644786834717, "learning_rate": 1.0030425146032141e-06, "loss": 0.9977, "step": 12495 }, { "epoch": 2.6209930861093653, "grad_norm": 3.080481767654419, "learning_rate": 9.8714056558199e-07, "loss": 1.1113, "step": 12510 }, { "epoch": 2.6241357636706475, "grad_norm": 3.0548975467681885, "learning_rate": 9.713591314539938e-07, "loss": 1.0369, "step": 12525 }, { "epoch": 2.6272784412319297, "grad_norm": 2.131619930267334, "learning_rate": 9.556984232397037e-07, "loss": 1.0072, "step": 12540 }, { "epoch": 2.630421118793212, "grad_norm": 2.963510036468506, "learning_rate": 9.40158650345312e-07, "loss": 1.0475, "step": 12555 }, { "epoch": 2.633563796354494, "grad_norm": 2.7937393188476562, "learning_rate": 9.24740020559931e-07, "loss": 1.0605, "step": 12570 }, { "epoch": 2.636706473915776, "grad_norm": 2.5828826427459717, "learning_rate": 9.094427400528128e-07, "loss": 1.0657, "step": 12585 }, { "epoch": 2.6398491514770583, "grad_norm": 2.3063597679138184, "learning_rate": 8.942670133706044e-07, "loss": 1.015, "step": 12600 }, { "epoch": 2.6429918290383405, "grad_norm": 2.2541069984436035, "learning_rate": 8.792130434345947e-07, "loss": 1.0939, "step": 12615 }, { "epoch": 2.6461345065996227, "grad_norm": 2.0006487369537354, "learning_rate": 8.642810315380168e-07, "loss": 1.0214, "step": 12630 }, { "epoch": 2.649277184160905, "grad_norm": 2.5538716316223145, "learning_rate": 8.494711773433395e-07, "loss": 1.0569, "step": 12645 }, { "epoch": 2.6524198617221875, "grad_norm": 2.3980712890625, "learning_rate": 8.347836788796148e-07, "loss": 1.033, "step": 12660 }, { "epoch": 2.6555625392834696, "grad_norm": 2.408301830291748, "learning_rate": 8.202187325398248e-07, "loss": 1.0013, "step": 12675 }, { "epoch": 2.658705216844752, "grad_norm": 1.9503532648086548, "learning_rate": 8.057765330782486e-07, "loss": 0.9753, "step": 12690 }, { "epoch": 2.661847894406034, "grad_norm": 2.117751359939575, "learning_rate": 7.914572736078618e-07, "loss": 1.091, "step": 12705 }, { "epoch": 2.664990571967316, "grad_norm": 2.651381731033325, "learning_rate": 7.772611455977586e-07, "loss": 1.0366, "step": 12720 }, { "epoch": 2.6681332495285983, "grad_norm": 2.568176031112671, "learning_rate": 7.631883388705897e-07, "loss": 1.0775, "step": 12735 }, { "epoch": 2.6712759270898805, "grad_norm": 2.860661268234253, "learning_rate": 7.492390416000217e-07, "loss": 1.0576, "step": 12750 }, { "epoch": 2.6744186046511627, "grad_norm": 2.9709537029266357, "learning_rate": 7.354134403082191e-07, "loss": 1.0724, "step": 12765 }, { "epoch": 2.6775612822124453, "grad_norm": 1.8460935354232788, "learning_rate": 7.217117198633561e-07, "loss": 1.0227, "step": 12780 }, { "epoch": 2.6807039597737274, "grad_norm": 2.6005988121032715, "learning_rate": 7.081340634771394e-07, "loss": 1.0787, "step": 12795 }, { "epoch": 2.6838466373350096, "grad_norm": 2.098374366760254, "learning_rate": 6.946806527023686e-07, "loss": 1.0402, "step": 12810 }, { "epoch": 2.686989314896292, "grad_norm": 2.3875677585601807, "learning_rate": 6.813516674304888e-07, "loss": 1.0367, "step": 12825 }, { "epoch": 2.690131992457574, "grad_norm": 1.9632166624069214, "learning_rate": 6.681472858892068e-07, "loss": 1.0026, "step": 12840 }, { "epoch": 2.693274670018856, "grad_norm": 2.6119518280029297, "learning_rate": 6.550676846400927e-07, "loss": 1.0865, "step": 12855 }, { "epoch": 2.6964173475801383, "grad_norm": 2.205327272415161, "learning_rate": 6.421130385762275e-07, "loss": 1.0165, "step": 12870 }, { "epoch": 2.6995600251414205, "grad_norm": 3.523000478744507, "learning_rate": 6.292835209198622e-07, "loss": 1.0645, "step": 12885 }, { "epoch": 2.7027027027027026, "grad_norm": 2.800053119659424, "learning_rate": 6.165793032200984e-07, "loss": 1.028, "step": 12900 }, { "epoch": 2.705845380263985, "grad_norm": 2.6920459270477295, "learning_rate": 6.040005553505979e-07, "loss": 1.0499, "step": 12915 }, { "epoch": 2.708988057825267, "grad_norm": 2.7640254497528076, "learning_rate": 5.915474455073111e-07, "loss": 1.0495, "step": 12930 }, { "epoch": 2.712130735386549, "grad_norm": 2.8138866424560547, "learning_rate": 5.792201402062281e-07, "loss": 1.0698, "step": 12945 }, { "epoch": 2.7152734129478313, "grad_norm": 3.2253429889678955, "learning_rate": 5.670188042811498e-07, "loss": 1.0635, "step": 12960 }, { "epoch": 2.718416090509114, "grad_norm": 2.7197041511535645, "learning_rate": 5.549436008814868e-07, "loss": 1.0507, "step": 12975 }, { "epoch": 2.721558768070396, "grad_norm": 2.0629515647888184, "learning_rate": 5.429946914700745e-07, "loss": 1.0665, "step": 12990 }, { "epoch": 2.7247014456316783, "grad_norm": 1.5886244773864746, "learning_rate": 5.311722358210147e-07, "loss": 1.0672, "step": 13005 }, { "epoch": 2.7278441231929604, "grad_norm": 2.3268556594848633, "learning_rate": 5.194763920175461e-07, "loss": 1.0415, "step": 13020 }, { "epoch": 2.7309868007542426, "grad_norm": 2.352407693862915, "learning_rate": 5.079073164499216e-07, "loss": 1.0628, "step": 13035 }, { "epoch": 2.7341294783155248, "grad_norm": 2.2687618732452393, "learning_rate": 4.964651638133211e-07, "loss": 1.1221, "step": 13050 }, { "epoch": 2.737272155876807, "grad_norm": 2.1698989868164062, "learning_rate": 4.851500871057791e-07, "loss": 1.0301, "step": 13065 }, { "epoch": 2.740414833438089, "grad_norm": 2.6248748302459717, "learning_rate": 4.7396223762614545e-07, "loss": 1.1054, "step": 13080 }, { "epoch": 2.7435575109993717, "grad_norm": 2.520930767059326, "learning_rate": 4.629017649720602e-07, "loss": 1.0437, "step": 13095 }, { "epoch": 2.746700188560654, "grad_norm": 3.120213508605957, "learning_rate": 4.519688170379499e-07, "loss": 1.0116, "step": 13110 }, { "epoch": 2.749842866121936, "grad_norm": 2.3425211906433105, "learning_rate": 4.411635400130521e-07, "loss": 1.0608, "step": 13125 }, { "epoch": 2.7529855436832182, "grad_norm": 2.232712745666504, "learning_rate": 4.3048607837945954e-07, "loss": 1.0401, "step": 13140 }, { "epoch": 2.7561282212445004, "grad_norm": 2.2147791385650635, "learning_rate": 4.1993657491019355e-07, "loss": 1.0217, "step": 13155 }, { "epoch": 2.7592708988057826, "grad_norm": 2.526965618133545, "learning_rate": 4.095151706672884e-07, "loss": 1.0915, "step": 13170 }, { "epoch": 2.7624135763670647, "grad_norm": 1.9940340518951416, "learning_rate": 3.9922200499990404e-07, "loss": 1.0553, "step": 13185 }, { "epoch": 2.765556253928347, "grad_norm": 2.973738431930542, "learning_rate": 3.890572155424721e-07, "loss": 1.0615, "step": 13200 }, { "epoch": 2.768698931489629, "grad_norm": 2.3078935146331787, "learning_rate": 3.790209382128451e-07, "loss": 1.0285, "step": 13215 }, { "epoch": 2.7718416090509113, "grad_norm": 2.037052869796753, "learning_rate": 3.691133072104891e-07, "loss": 1.0545, "step": 13230 }, { "epoch": 2.7749842866121934, "grad_norm": 2.9037740230560303, "learning_rate": 3.59334455014676e-07, "loss": 1.0151, "step": 13245 }, { "epoch": 2.7781269641734756, "grad_norm": 2.245582103729248, "learning_rate": 3.4968451238272527e-07, "loss": 1.0715, "step": 13260 }, { "epoch": 2.7812696417347578, "grad_norm": 2.7545788288116455, "learning_rate": 3.4079431818805573e-07, "loss": 1.0315, "step": 13275 }, { "epoch": 2.7844123192960404, "grad_norm": 3.0721166133880615, "learning_rate": 3.313939650727771e-07, "loss": 1.0796, "step": 13290 }, { "epoch": 2.7875549968573226, "grad_norm": 2.1562039852142334, "learning_rate": 3.22122895125897e-07, "loss": 1.0105, "step": 13305 }, { "epoch": 2.7906976744186047, "grad_norm": 2.4602880477905273, "learning_rate": 3.129812323149528e-07, "loss": 1.0805, "step": 13320 }, { "epoch": 2.793840351979887, "grad_norm": 2.485700845718384, "learning_rate": 3.039690988771271e-07, "loss": 1.1113, "step": 13335 }, { "epoch": 2.796983029541169, "grad_norm": 2.2793211936950684, "learning_rate": 2.950866153176024e-07, "loss": 1.0269, "step": 13350 }, { "epoch": 2.8001257071024512, "grad_norm": 2.4900388717651367, "learning_rate": 2.8633390040796014e-07, "loss": 1.0586, "step": 13365 }, { "epoch": 2.8032683846637334, "grad_norm": 2.7282629013061523, "learning_rate": 2.7771107118458874e-07, "loss": 1.0752, "step": 13380 }, { "epoch": 2.8064110622250156, "grad_norm": 2.986842632293701, "learning_rate": 2.692182429471157e-07, "loss": 1.0826, "step": 13395 }, { "epoch": 2.809553739786298, "grad_norm": 1.311810851097107, "learning_rate": 2.608555292568671e-07, "loss": 1.0583, "step": 13410 }, { "epoch": 2.8126964173475804, "grad_norm": 2.533581256866455, "learning_rate": 2.526230419353537e-07, "loss": 1.0604, "step": 13425 }, { "epoch": 2.8158390949088625, "grad_norm": 2.3201465606689453, "learning_rate": 2.445208910627717e-07, "loss": 1.0868, "step": 13440 }, { "epoch": 2.8189817724701447, "grad_norm": 2.0772528648376465, "learning_rate": 2.3654918497653334e-07, "loss": 1.0854, "step": 13455 }, { "epoch": 2.822124450031427, "grad_norm": 2.761883020401001, "learning_rate": 2.287080302698097e-07, "loss": 1.0408, "step": 13470 }, { "epoch": 2.825267127592709, "grad_norm": 2.217202663421631, "learning_rate": 2.2099753179011807e-07, "loss": 1.0712, "step": 13485 }, { "epoch": 2.828409805153991, "grad_norm": 2.955160140991211, "learning_rate": 2.134177926379144e-07, "loss": 1.0678, "step": 13500 }, { "epoch": 2.8315524827152734, "grad_norm": 2.9105584621429443, "learning_rate": 2.0596891416521215e-07, "loss": 1.1113, "step": 13515 }, { "epoch": 2.8346951602765555, "grad_norm": 2.016284704208374, "learning_rate": 1.9865099597422887e-07, "loss": 1.0373, "step": 13530 }, { "epoch": 2.8378378378378377, "grad_norm": 2.6535935401916504, "learning_rate": 1.9146413591605516e-07, "loss": 1.0911, "step": 13545 }, { "epoch": 2.84098051539912, "grad_norm": 3.204658031463623, "learning_rate": 1.844084300893456e-07, "loss": 1.0636, "step": 13560 }, { "epoch": 2.844123192960402, "grad_norm": 2.872605085372925, "learning_rate": 1.7748397283903874e-07, "loss": 1.093, "step": 13575 }, { "epoch": 2.8472658705216842, "grad_norm": 2.8483595848083496, "learning_rate": 1.7069085675508356e-07, "loss": 1.0961, "step": 13590 }, { "epoch": 2.850408548082967, "grad_norm": 2.082428455352783, "learning_rate": 1.6402917267121398e-07, "loss": 1.0593, "step": 13605 }, { "epoch": 2.853551225644249, "grad_norm": 2.1791653633117676, "learning_rate": 1.5749900966372523e-07, "loss": 1.0552, "step": 13620 }, { "epoch": 2.856693903205531, "grad_norm": 1.7146304845809937, "learning_rate": 1.5110045505029036e-07, "loss": 1.0435, "step": 13635 }, { "epoch": 2.8598365807668134, "grad_norm": 2.64682674407959, "learning_rate": 1.4483359438878574e-07, "loss": 1.023, "step": 13650 }, { "epoch": 2.8629792583280955, "grad_norm": 2.390026569366455, "learning_rate": 1.386985114761541e-07, "loss": 1.0559, "step": 13665 }, { "epoch": 2.8661219358893777, "grad_norm": 2.797459602355957, "learning_rate": 1.326952883472743e-07, "loss": 1.1022, "step": 13680 }, { "epoch": 2.86926461345066, "grad_norm": 1.6154651641845703, "learning_rate": 1.2682400527387674e-07, "loss": 1.0415, "step": 13695 }, { "epoch": 2.872407291011942, "grad_norm": 2.7609522342681885, "learning_rate": 1.2108474076345855e-07, "loss": 1.0901, "step": 13710 }, { "epoch": 2.8755499685732246, "grad_norm": 2.618767499923706, "learning_rate": 1.1547757155824124e-07, "loss": 1.0118, "step": 13725 }, { "epoch": 2.878692646134507, "grad_norm": 2.90389347076416, "learning_rate": 1.1000257263414139e-07, "loss": 1.0729, "step": 13740 }, { "epoch": 2.881835323695789, "grad_norm": 2.361311435699463, "learning_rate": 1.0465981719976814e-07, "loss": 1.0597, "step": 13755 }, { "epoch": 2.884978001257071, "grad_norm": 1.77293062210083, "learning_rate": 9.944937669544741e-08, "loss": 1.0155, "step": 13770 }, { "epoch": 2.8881206788183533, "grad_norm": 2.5535075664520264, "learning_rate": 9.437132079226142e-08, "loss": 1.0966, "step": 13785 }, { "epoch": 2.8912633563796355, "grad_norm": 2.4783518314361572, "learning_rate": 8.942571739112171e-08, "loss": 1.0554, "step": 13800 }, { "epoch": 2.8944060339409177, "grad_norm": 1.9884682893753052, "learning_rate": 8.46126326218577e-08, "loss": 0.9958, "step": 13815 }, { "epoch": 2.8975487115022, "grad_norm": 1.811047911643982, "learning_rate": 7.99321308423362e-08, "loss": 1.0913, "step": 13830 }, { "epoch": 2.900691389063482, "grad_norm": 2.5902936458587646, "learning_rate": 7.538427463759768e-08, "loss": 1.0423, "step": 13845 }, { "epoch": 2.903834066624764, "grad_norm": 2.312286853790283, "learning_rate": 7.096912481901919e-08, "loss": 1.0178, "step": 13860 }, { "epoch": 2.9069767441860463, "grad_norm": 1.918196201324463, "learning_rate": 6.668674042350498e-08, "loss": 1.053, "step": 13875 }, { "epoch": 2.9101194217473285, "grad_norm": 1.8703136444091797, "learning_rate": 6.253717871269271e-08, "loss": 1.0538, "step": 13890 }, { "epoch": 2.9132620993086107, "grad_norm": 2.3557560443878174, "learning_rate": 5.8520495172190714e-08, "loss": 1.0793, "step": 13905 }, { "epoch": 2.9164047768698933, "grad_norm": 2.194120407104492, "learning_rate": 5.4636743510835256e-08, "loss": 1.0484, "step": 13920 }, { "epoch": 2.9195474544311755, "grad_norm": 1.7967839241027832, "learning_rate": 5.088597565997111e-08, "loss": 1.0467, "step": 13935 }, { "epoch": 2.9226901319924576, "grad_norm": 1.9517879486083984, "learning_rate": 4.72682417727599e-08, "loss": 1.0425, "step": 13950 }, { "epoch": 2.92583280955374, "grad_norm": 2.5437710285186768, "learning_rate": 4.37835902235062e-08, "loss": 1.0693, "step": 13965 }, { "epoch": 2.928975487115022, "grad_norm": 2.4388108253479004, "learning_rate": 4.043206760701246e-08, "loss": 1.1279, "step": 13980 }, { "epoch": 2.932118164676304, "grad_norm": 2.955836057662964, "learning_rate": 3.7213718737958425e-08, "loss": 1.0869, "step": 13995 }, { "epoch": 2.9352608422375863, "grad_norm": 2.5022730827331543, "learning_rate": 3.412858665029495e-08, "loss": 1.0562, "step": 14010 }, { "epoch": 2.9384035197988685, "grad_norm": 2.614485502243042, "learning_rate": 3.117671259667554e-08, "loss": 1.0316, "step": 14025 }, { "epoch": 2.941546197360151, "grad_norm": 1.9243937730789185, "learning_rate": 2.8358136047902385e-08, "loss": 1.0355, "step": 14040 }, { "epoch": 2.9446888749214333, "grad_norm": 2.475581407546997, "learning_rate": 2.5672894692397865e-08, "loss": 1.0007, "step": 14055 }, { "epoch": 2.9478315524827154, "grad_norm": 2.3144288063049316, "learning_rate": 2.312102443569941e-08, "loss": 1.0889, "step": 14070 }, { "epoch": 2.9509742300439976, "grad_norm": 1.9073032140731812, "learning_rate": 2.0702559399985444e-08, "loss": 1.0502, "step": 14085 }, { "epoch": 2.95411690760528, "grad_norm": 2.616192579269409, "learning_rate": 1.8417531923609066e-08, "loss": 1.0767, "step": 14100 }, { "epoch": 2.957259585166562, "grad_norm": 2.3188884258270264, "learning_rate": 1.6265972560675082e-08, "loss": 1.0091, "step": 14115 }, { "epoch": 2.960402262727844, "grad_norm": 2.5977001190185547, "learning_rate": 1.424791008062587e-08, "loss": 1.0617, "step": 14130 }, { "epoch": 2.9635449402891263, "grad_norm": 2.5720694065093994, "learning_rate": 1.2363371467861706e-08, "loss": 1.0382, "step": 14145 }, { "epoch": 2.9666876178504085, "grad_norm": 2.201608657836914, "learning_rate": 1.061238192137104e-08, "loss": 1.0358, "step": 14160 }, { "epoch": 2.9698302954116906, "grad_norm": 2.547421455383301, "learning_rate": 8.994964854404098e-09, "loss": 1.0699, "step": 14175 }, { "epoch": 2.972972972972973, "grad_norm": 2.6077804565429688, "learning_rate": 7.51114189415314e-09, "loss": 1.0301, "step": 14190 }, { "epoch": 2.976115650534255, "grad_norm": 2.2802724838256836, "learning_rate": 6.16093288146491e-09, "loss": 1.0323, "step": 14205 }, { "epoch": 2.979258328095537, "grad_norm": 3.556347608566284, "learning_rate": 4.9443558705764e-09, "loss": 1.057, "step": 14220 }, { "epoch": 2.9824010056568198, "grad_norm": 2.6890947818756104, "learning_rate": 3.86142712887172e-09, "loss": 1.0276, "step": 14235 }, { "epoch": 2.985543683218102, "grad_norm": 1.9261900186538696, "learning_rate": 2.912161136665592e-09, "loss": 0.9678, "step": 14250 }, { "epoch": 2.988686360779384, "grad_norm": 1.9330463409423828, "learning_rate": 2.096570587010183e-09, "loss": 1.07, "step": 14265 }, { "epoch": 2.9918290383406663, "grad_norm": 2.8647425174713135, "learning_rate": 1.4559673424208165e-09, "loss": 1.0322, "step": 14280 }, { "epoch": 2.9949717159019484, "grad_norm": 2.028707504272461, "learning_rate": 8.98845328231035e-10, "loss": 1.0043, "step": 14295 }, { "epoch": 2.9981143934632306, "grad_norm": 2.955282688140869, "learning_rate": 4.754256775185706e-10, "loss": 1.0309, "step": 14310 }, { "epoch": 3.0, "step": 14319, "total_flos": 3.0631036932776264e+18, "train_loss": 1.1692348350084263, "train_runtime": 44795.2526, "train_samples_per_second": 1.279, "train_steps_per_second": 0.32 } ], "logging_steps": 15, "max_steps": 14319, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.0631036932776264e+18, "train_batch_size": 2, "trial_name": null, "trial_params": null }