{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.916014042668107, "eval_steps": 500, "global_step": 5400, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0054010261949770455, "grad_norm": 1.4275089502334595, "learning_rate": 0.00016, "loss": 3.4877567291259766, "step": 10 }, { "epoch": 0.010802052389954091, "grad_norm": 1.4062877893447876, "learning_rate": 0.00019971150378651282, "loss": 3.177476501464844, "step": 20 }, { "epoch": 0.016203078584931137, "grad_norm": 1.4894630908966064, "learning_rate": 0.0001993508835196538, "loss": 2.949508857727051, "step": 30 }, { "epoch": 0.021604104779908182, "grad_norm": 1.1637152433395386, "learning_rate": 0.00019899026325279482, "loss": 3.2568588256835938, "step": 40 }, { "epoch": 0.027005130974885227, "grad_norm": 1.0710796117782593, "learning_rate": 0.00019862964298593581, "loss": 3.033162307739258, "step": 50 }, { "epoch": 0.032406157169862275, "grad_norm": 1.036720633506775, "learning_rate": 0.0001982690227190768, "loss": 2.9209388732910155, "step": 60 }, { "epoch": 0.03780718336483932, "grad_norm": 1.5094584226608276, "learning_rate": 0.00019790840245221782, "loss": 3.199975776672363, "step": 70 }, { "epoch": 0.043208209559816364, "grad_norm": 1.25065016746521, "learning_rate": 0.00019754778218535884, "loss": 3.1586023330688477, "step": 80 }, { "epoch": 0.04860923575479341, "grad_norm": 1.093263864517212, "learning_rate": 0.00019718716191849983, "loss": 3.1673559188842773, "step": 90 }, { "epoch": 0.054010261949770454, "grad_norm": 1.7025214433670044, "learning_rate": 0.00019682654165164082, "loss": 3.0090349197387694, "step": 100 }, { "epoch": 0.059411288144747505, "grad_norm": 1.0379327535629272, "learning_rate": 0.00019646592138478184, "loss": 3.0862411499023437, "step": 110 }, { "epoch": 0.06481231433972455, "grad_norm": 1.0767595767974854, "learning_rate": 0.00019610530111792285, "loss": 2.9345855712890625, "step": 120 }, { "epoch": 0.07021334053470159, "grad_norm": 1.2473753690719604, "learning_rate": 0.00019574468085106384, "loss": 3.168864440917969, "step": 130 }, { "epoch": 0.07561436672967864, "grad_norm": 1.1580232381820679, "learning_rate": 0.00019538406058420483, "loss": 3.0444286346435545, "step": 140 }, { "epoch": 0.08101539292465569, "grad_norm": 1.1562517881393433, "learning_rate": 0.00019502344031734585, "loss": 3.0681676864624023, "step": 150 }, { "epoch": 0.08641641911963273, "grad_norm": 1.2069370746612549, "learning_rate": 0.00019466282005048684, "loss": 3.147779846191406, "step": 160 }, { "epoch": 0.09181744531460978, "grad_norm": 1.1520271301269531, "learning_rate": 0.00019430219978362786, "loss": 3.2159664154052736, "step": 170 }, { "epoch": 0.09721847150958682, "grad_norm": 1.3290963172912598, "learning_rate": 0.00019394157951676885, "loss": 3.17901668548584, "step": 180 }, { "epoch": 0.10261949770456387, "grad_norm": 1.4537458419799805, "learning_rate": 0.00019358095924990984, "loss": 3.076709747314453, "step": 190 }, { "epoch": 0.10802052389954091, "grad_norm": 1.6899892091751099, "learning_rate": 0.00019322033898305085, "loss": 2.9362957000732424, "step": 200 }, { "epoch": 0.11342155009451796, "grad_norm": 1.218902349472046, "learning_rate": 0.00019289578074287774, "loss": 3.0862009048461916, "step": 210 }, { "epoch": 0.11882257628949501, "grad_norm": 1.253646969795227, "learning_rate": 0.00019253516047601876, "loss": 3.0416831970214844, "step": 220 }, { "epoch": 0.12422360248447205, "grad_norm": 1.1816236972808838, "learning_rate": 0.00019217454020915977, "loss": 3.09185791015625, "step": 230 }, { "epoch": 0.1296246286794491, "grad_norm": 1.2364989519119263, "learning_rate": 0.00019181391994230076, "loss": 3.1257530212402345, "step": 240 }, { "epoch": 0.13502565487442614, "grad_norm": 1.3117403984069824, "learning_rate": 0.00019145329967544175, "loss": 3.0078147888183593, "step": 250 }, { "epoch": 0.14042668106940318, "grad_norm": 1.060840368270874, "learning_rate": 0.00019109267940858277, "loss": 3.205386734008789, "step": 260 }, { "epoch": 0.14582770726438024, "grad_norm": 1.0836883783340454, "learning_rate": 0.0001907320591417238, "loss": 3.00787410736084, "step": 270 }, { "epoch": 0.15122873345935728, "grad_norm": 1.1705331802368164, "learning_rate": 0.00019037143887486478, "loss": 3.1415082931518556, "step": 280 }, { "epoch": 0.15662975965433432, "grad_norm": 1.1469744443893433, "learning_rate": 0.00019001081860800577, "loss": 3.0711442947387697, "step": 290 }, { "epoch": 0.16203078584931138, "grad_norm": 1.7230859994888306, "learning_rate": 0.00018965019834114679, "loss": 3.0760629653930662, "step": 300 }, { "epoch": 0.16743181204428842, "grad_norm": 1.3305317163467407, "learning_rate": 0.00018928957807428778, "loss": 3.0109764099121095, "step": 310 }, { "epoch": 0.17283283823926546, "grad_norm": 1.1433296203613281, "learning_rate": 0.0001889289578074288, "loss": 2.900168228149414, "step": 320 }, { "epoch": 0.1782338644342425, "grad_norm": 1.4017837047576904, "learning_rate": 0.00018856833754056978, "loss": 2.9257192611694336, "step": 330 }, { "epoch": 0.18363489062921956, "grad_norm": 1.418676495552063, "learning_rate": 0.00018820771727371077, "loss": 3.1230239868164062, "step": 340 }, { "epoch": 0.1890359168241966, "grad_norm": 1.2229913473129272, "learning_rate": 0.0001878470970068518, "loss": 2.8634975433349608, "step": 350 }, { "epoch": 0.19443694301917364, "grad_norm": 1.2936865091323853, "learning_rate": 0.0001874864767399928, "loss": 3.0415794372558596, "step": 360 }, { "epoch": 0.1998379692141507, "grad_norm": 1.3978519439697266, "learning_rate": 0.0001871258564731338, "loss": 2.9328636169433593, "step": 370 }, { "epoch": 0.20523899540912774, "grad_norm": 1.0732481479644775, "learning_rate": 0.0001867652362062748, "loss": 3.147154426574707, "step": 380 }, { "epoch": 0.21064002160410478, "grad_norm": 1.304332971572876, "learning_rate": 0.0001864046159394158, "loss": 2.98757266998291, "step": 390 }, { "epoch": 0.21604104779908181, "grad_norm": 1.2670519351959229, "learning_rate": 0.00018604399567255682, "loss": 3.0830259323120117, "step": 400 }, { "epoch": 0.22144207399405888, "grad_norm": 1.3472830057144165, "learning_rate": 0.0001856833754056978, "loss": 3.0725728988647463, "step": 410 }, { "epoch": 0.22684310018903592, "grad_norm": 1.5913994312286377, "learning_rate": 0.0001853227551388388, "loss": 2.872740936279297, "step": 420 }, { "epoch": 0.23224412638401296, "grad_norm": 1.4995367527008057, "learning_rate": 0.00018496213487197982, "loss": 3.1985055923461916, "step": 430 }, { "epoch": 0.23764515257899002, "grad_norm": 1.3097314834594727, "learning_rate": 0.0001846015146051208, "loss": 2.9131542205810548, "step": 440 }, { "epoch": 0.24304617877396706, "grad_norm": 1.2168841361999512, "learning_rate": 0.00018424089433826182, "loss": 3.0022405624389648, "step": 450 }, { "epoch": 0.2484472049689441, "grad_norm": 1.2428247928619385, "learning_rate": 0.00018388027407140281, "loss": 3.13458194732666, "step": 460 }, { "epoch": 0.25384823116392113, "grad_norm": 1.0732594728469849, "learning_rate": 0.00018351965380454383, "loss": 2.995112991333008, "step": 470 }, { "epoch": 0.2592492573588982, "grad_norm": 1.0187745094299316, "learning_rate": 0.00018315903353768482, "loss": 2.9637735366821287, "step": 480 }, { "epoch": 0.2646502835538752, "grad_norm": 1.2797402143478394, "learning_rate": 0.00018279841327082584, "loss": 2.9933282852172853, "step": 490 }, { "epoch": 0.2700513097488523, "grad_norm": 0.9948688745498657, "learning_rate": 0.00018243779300396683, "loss": 3.078018569946289, "step": 500 }, { "epoch": 0.27545233594382934, "grad_norm": 1.0406428575515747, "learning_rate": 0.00018207717273710782, "loss": 2.9701290130615234, "step": 510 }, { "epoch": 0.28085336213880635, "grad_norm": 1.184054970741272, "learning_rate": 0.00018171655247024884, "loss": 3.094232368469238, "step": 520 }, { "epoch": 0.2862543883337834, "grad_norm": 1.2439318895339966, "learning_rate": 0.00018135593220338985, "loss": 3.0941631317138674, "step": 530 }, { "epoch": 0.2916554145287605, "grad_norm": 1.3002263307571411, "learning_rate": 0.00018099531193653084, "loss": 2.9608373641967773, "step": 540 }, { "epoch": 0.2970564407237375, "grad_norm": 1.5705729722976685, "learning_rate": 0.00018063469166967183, "loss": 3.026919364929199, "step": 550 }, { "epoch": 0.30245746691871456, "grad_norm": 1.1658886671066284, "learning_rate": 0.00018027407140281285, "loss": 2.9997627258300783, "step": 560 }, { "epoch": 0.3078584931136916, "grad_norm": 1.0013272762298584, "learning_rate": 0.00017991345113595387, "loss": 2.919036865234375, "step": 570 }, { "epoch": 0.31325951930866863, "grad_norm": 1.1810517311096191, "learning_rate": 0.00017955283086909486, "loss": 3.0280284881591797, "step": 580 }, { "epoch": 0.3186605455036457, "grad_norm": 1.4247483015060425, "learning_rate": 0.00017919221060223585, "loss": 2.944463920593262, "step": 590 }, { "epoch": 0.32406157169862276, "grad_norm": 1.1780109405517578, "learning_rate": 0.00017883159033537686, "loss": 3.165457534790039, "step": 600 }, { "epoch": 0.3294625978935998, "grad_norm": 1.082470417022705, "learning_rate": 0.00017847097006851785, "loss": 2.7378583908081056, "step": 610 }, { "epoch": 0.33486362408857684, "grad_norm": 1.1170358657836914, "learning_rate": 0.00017811034980165887, "loss": 3.1393898010253904, "step": 620 }, { "epoch": 0.34026465028355385, "grad_norm": 1.0197583436965942, "learning_rate": 0.00017774972953479986, "loss": 2.8996389389038084, "step": 630 }, { "epoch": 0.3456656764785309, "grad_norm": 0.9194260835647583, "learning_rate": 0.00017738910926794085, "loss": 3.0526611328125, "step": 640 }, { "epoch": 0.351066702673508, "grad_norm": 1.3681086301803589, "learning_rate": 0.00017702848900108187, "loss": 2.85855770111084, "step": 650 }, { "epoch": 0.356467728868485, "grad_norm": 1.3217997550964355, "learning_rate": 0.00017666786873422289, "loss": 2.9244100570678713, "step": 660 }, { "epoch": 0.36186875506346206, "grad_norm": 1.0755029916763306, "learning_rate": 0.00017630724846736388, "loss": 2.989073944091797, "step": 670 }, { "epoch": 0.3672697812584391, "grad_norm": 1.2188295125961304, "learning_rate": 0.00017594662820050487, "loss": 3.005179023742676, "step": 680 }, { "epoch": 0.37267080745341613, "grad_norm": 1.0459349155426025, "learning_rate": 0.00017558600793364588, "loss": 2.9304704666137695, "step": 690 }, { "epoch": 0.3780718336483932, "grad_norm": 1.1236655712127686, "learning_rate": 0.0001752253876667869, "loss": 3.048674964904785, "step": 700 }, { "epoch": 0.38347285984337026, "grad_norm": 1.0007507801055908, "learning_rate": 0.00017486476739992786, "loss": 3.1241172790527343, "step": 710 }, { "epoch": 0.38887388603834727, "grad_norm": 1.2311853170394897, "learning_rate": 0.00017450414713306888, "loss": 2.8489208221435547, "step": 720 }, { "epoch": 0.39427491223332434, "grad_norm": 1.1284152269363403, "learning_rate": 0.0001741435268662099, "loss": 2.889890670776367, "step": 730 }, { "epoch": 0.3996759384283014, "grad_norm": 1.0716314315795898, "learning_rate": 0.0001737829065993509, "loss": 3.019309234619141, "step": 740 }, { "epoch": 0.4050769646232784, "grad_norm": 1.3626065254211426, "learning_rate": 0.00017342228633249188, "loss": 2.8685939788818358, "step": 750 }, { "epoch": 0.4104779908182555, "grad_norm": 1.133290410041809, "learning_rate": 0.0001730616660656329, "loss": 2.813917541503906, "step": 760 }, { "epoch": 0.4158790170132325, "grad_norm": 1.2982553243637085, "learning_rate": 0.0001727010457987739, "loss": 3.066880798339844, "step": 770 }, { "epoch": 0.42128004320820955, "grad_norm": 1.0441523790359497, "learning_rate": 0.0001723404255319149, "loss": 2.8972396850585938, "step": 780 }, { "epoch": 0.4266810694031866, "grad_norm": 1.1574914455413818, "learning_rate": 0.0001719798052650559, "loss": 3.0566661834716795, "step": 790 }, { "epoch": 0.43208209559816363, "grad_norm": 1.146910309791565, "learning_rate": 0.0001716191849981969, "loss": 2.9340593338012697, "step": 800 }, { "epoch": 0.4374831217931407, "grad_norm": 1.1724562644958496, "learning_rate": 0.0001712585647313379, "loss": 2.9801464080810547, "step": 810 }, { "epoch": 0.44288414798811776, "grad_norm": 1.3650254011154175, "learning_rate": 0.00017089794446447892, "loss": 2.9490041732788086, "step": 820 }, { "epoch": 0.44828517418309477, "grad_norm": 1.0346980094909668, "learning_rate": 0.00017053732419761993, "loss": 2.971526336669922, "step": 830 }, { "epoch": 0.45368620037807184, "grad_norm": 0.9564275741577148, "learning_rate": 0.00017017670393076092, "loss": 3.0509788513183596, "step": 840 }, { "epoch": 0.4590872265730489, "grad_norm": 1.4447107315063477, "learning_rate": 0.0001698160836639019, "loss": 2.9528224945068358, "step": 850 }, { "epoch": 0.4644882527680259, "grad_norm": 1.0819096565246582, "learning_rate": 0.00016945546339704293, "loss": 3.0397647857666015, "step": 860 }, { "epoch": 0.469889278963003, "grad_norm": 1.3749688863754272, "learning_rate": 0.00016909484313018395, "loss": 2.9335342407226563, "step": 870 }, { "epoch": 0.47529030515798004, "grad_norm": 0.9845815300941467, "learning_rate": 0.0001687342228633249, "loss": 3.1886215209960938, "step": 880 }, { "epoch": 0.48069133135295705, "grad_norm": 1.0727243423461914, "learning_rate": 0.00016837360259646593, "loss": 3.071747970581055, "step": 890 }, { "epoch": 0.4860923575479341, "grad_norm": 1.0900219678878784, "learning_rate": 0.00016801298232960694, "loss": 2.9294233322143555, "step": 900 }, { "epoch": 0.4914933837429111, "grad_norm": 1.5818015336990356, "learning_rate": 0.00016765236206274793, "loss": 2.7862874984741213, "step": 910 }, { "epoch": 0.4968944099378882, "grad_norm": 1.1314663887023926, "learning_rate": 0.00016729174179588892, "loss": 2.8420650482177736, "step": 920 }, { "epoch": 0.5022954361328652, "grad_norm": 1.2435640096664429, "learning_rate": 0.00016693112152902994, "loss": 2.8855379104614256, "step": 930 }, { "epoch": 0.5076964623278423, "grad_norm": 1.0684075355529785, "learning_rate": 0.00016657050126217096, "loss": 2.908370018005371, "step": 940 }, { "epoch": 0.5130974885228193, "grad_norm": 1.4429422616958618, "learning_rate": 0.00016620988099531195, "loss": 3.0200515747070313, "step": 950 }, { "epoch": 0.5184985147177964, "grad_norm": 1.3540408611297607, "learning_rate": 0.00016584926072845294, "loss": 2.653054428100586, "step": 960 }, { "epoch": 0.5238995409127735, "grad_norm": 1.1349395513534546, "learning_rate": 0.00016548864046159395, "loss": 3.068928909301758, "step": 970 }, { "epoch": 0.5293005671077504, "grad_norm": 1.0107704401016235, "learning_rate": 0.00016512802019473494, "loss": 2.912160110473633, "step": 980 }, { "epoch": 0.5347015933027275, "grad_norm": 1.402188777923584, "learning_rate": 0.00016476739992787596, "loss": 3.3307727813720702, "step": 990 }, { "epoch": 0.5401026194977046, "grad_norm": 1.200325608253479, "learning_rate": 0.00016440677966101695, "loss": 3.0100019454956053, "step": 1000 }, { "epoch": 0.5455036456926816, "grad_norm": 1.1472864151000977, "learning_rate": 0.00016404615939415794, "loss": 2.938519096374512, "step": 1010 }, { "epoch": 0.5509046718876587, "grad_norm": 1.184414267539978, "learning_rate": 0.00016368553912729896, "loss": 2.9431873321533204, "step": 1020 }, { "epoch": 0.5563056980826357, "grad_norm": 1.0453943014144897, "learning_rate": 0.00016332491886043998, "loss": 2.98077335357666, "step": 1030 }, { "epoch": 0.5617067242776127, "grad_norm": 1.1441236734390259, "learning_rate": 0.00016296429859358097, "loss": 3.0176897048950195, "step": 1040 }, { "epoch": 0.5671077504725898, "grad_norm": 0.9775217771530151, "learning_rate": 0.00016260367832672196, "loss": 3.0514453887939452, "step": 1050 }, { "epoch": 0.5725087766675668, "grad_norm": 1.6667460203170776, "learning_rate": 0.00016224305805986297, "loss": 2.9403846740722654, "step": 1060 }, { "epoch": 0.5779098028625439, "grad_norm": 1.2570966482162476, "learning_rate": 0.000161882437793004, "loss": 2.7937477111816404, "step": 1070 }, { "epoch": 0.583310829057521, "grad_norm": 1.3722922801971436, "learning_rate": 0.00016152181752614498, "loss": 2.8590538024902346, "step": 1080 }, { "epoch": 0.588711855252498, "grad_norm": 1.0699330568313599, "learning_rate": 0.00016116119725928597, "loss": 2.8490627288818358, "step": 1090 }, { "epoch": 0.594112881447475, "grad_norm": 1.2045018672943115, "learning_rate": 0.000160800576992427, "loss": 2.920571708679199, "step": 1100 }, { "epoch": 0.599513907642452, "grad_norm": 1.2119923830032349, "learning_rate": 0.00016043995672556798, "loss": 2.8805025100708006, "step": 1110 }, { "epoch": 0.6049149338374291, "grad_norm": 1.1207152605056763, "learning_rate": 0.000160079336458709, "loss": 2.937060737609863, "step": 1120 }, { "epoch": 0.6103159600324062, "grad_norm": 1.3675578832626343, "learning_rate": 0.00015971871619184998, "loss": 2.9657745361328125, "step": 1130 }, { "epoch": 0.6157169862273832, "grad_norm": 1.3562731742858887, "learning_rate": 0.000159358095924991, "loss": 3.0127033233642577, "step": 1140 }, { "epoch": 0.6211180124223602, "grad_norm": 1.0387409925460815, "learning_rate": 0.000158997475658132, "loss": 2.876498794555664, "step": 1150 }, { "epoch": 0.6265190386173373, "grad_norm": 1.2561172246932983, "learning_rate": 0.000158636855391273, "loss": 2.8820684432983397, "step": 1160 }, { "epoch": 0.6319200648123143, "grad_norm": 1.293705701828003, "learning_rate": 0.000158276235124414, "loss": 2.9766061782836912, "step": 1170 }, { "epoch": 0.6373210910072914, "grad_norm": 1.2604966163635254, "learning_rate": 0.000157915614857555, "loss": 3.119212341308594, "step": 1180 }, { "epoch": 0.6427221172022685, "grad_norm": 1.094990611076355, "learning_rate": 0.000157554994590696, "loss": 3.071938133239746, "step": 1190 }, { "epoch": 0.6481231433972455, "grad_norm": 1.1782667636871338, "learning_rate": 0.00015719437432383702, "loss": 3.0528400421142576, "step": 1200 }, { "epoch": 0.6535241695922225, "grad_norm": 1.467441201210022, "learning_rate": 0.000156833754056978, "loss": 2.967136764526367, "step": 1210 }, { "epoch": 0.6589251957871995, "grad_norm": 1.3159500360488892, "learning_rate": 0.000156473133790119, "loss": 2.9709142684936523, "step": 1220 }, { "epoch": 0.6643262219821766, "grad_norm": 1.1980047225952148, "learning_rate": 0.00015611251352326002, "loss": 3.0129560470581054, "step": 1230 }, { "epoch": 0.6697272481771537, "grad_norm": 1.125895619392395, "learning_rate": 0.00015575189325640104, "loss": 2.962136077880859, "step": 1240 }, { "epoch": 0.6751282743721307, "grad_norm": 1.5294994115829468, "learning_rate": 0.00015539127298954203, "loss": 3.139348793029785, "step": 1250 }, { "epoch": 0.6805293005671077, "grad_norm": 1.2642886638641357, "learning_rate": 0.00015503065272268302, "loss": 3.018353080749512, "step": 1260 }, { "epoch": 0.6859303267620848, "grad_norm": 1.0708301067352295, "learning_rate": 0.00015467003245582403, "loss": 2.9507617950439453, "step": 1270 }, { "epoch": 0.6913313529570618, "grad_norm": 1.2057912349700928, "learning_rate": 0.00015430941218896502, "loss": 2.8525314331054688, "step": 1280 }, { "epoch": 0.6967323791520389, "grad_norm": 1.003139615058899, "learning_rate": 0.00015394879192210604, "loss": 2.952615737915039, "step": 1290 }, { "epoch": 0.702133405347016, "grad_norm": 1.3539371490478516, "learning_rate": 0.00015358817165524703, "loss": 2.994022560119629, "step": 1300 }, { "epoch": 0.707534431541993, "grad_norm": 0.9936187863349915, "learning_rate": 0.00015322755138838802, "loss": 3.0847434997558594, "step": 1310 }, { "epoch": 0.71293545773697, "grad_norm": 1.077620267868042, "learning_rate": 0.00015286693112152904, "loss": 3.066029930114746, "step": 1320 }, { "epoch": 0.718336483931947, "grad_norm": 1.4198359251022339, "learning_rate": 0.00015250631085467005, "loss": 2.7741952896118165, "step": 1330 }, { "epoch": 0.7237375101269241, "grad_norm": 1.1443026065826416, "learning_rate": 0.00015214569058781104, "loss": 2.9916961669921873, "step": 1340 }, { "epoch": 0.7291385363219012, "grad_norm": 1.142872929573059, "learning_rate": 0.00015178507032095203, "loss": 2.9340869903564455, "step": 1350 }, { "epoch": 0.7345395625168782, "grad_norm": 1.061112403869629, "learning_rate": 0.00015142445005409305, "loss": 2.909784126281738, "step": 1360 }, { "epoch": 0.7399405887118553, "grad_norm": 1.5785682201385498, "learning_rate": 0.00015106382978723407, "loss": 2.6537490844726563, "step": 1370 }, { "epoch": 0.7453416149068323, "grad_norm": 1.0441385507583618, "learning_rate": 0.00015070320952037503, "loss": 3.0098087310791017, "step": 1380 }, { "epoch": 0.7507426411018093, "grad_norm": 1.1422723531723022, "learning_rate": 0.00015034258925351605, "loss": 3.039506721496582, "step": 1390 }, { "epoch": 0.7561436672967864, "grad_norm": 1.1552327871322632, "learning_rate": 0.00014998196898665707, "loss": 2.951033020019531, "step": 1400 }, { "epoch": 0.7615446934917635, "grad_norm": 1.0944606065750122, "learning_rate": 0.00014962134871979806, "loss": 3.106746482849121, "step": 1410 }, { "epoch": 0.7669457196867405, "grad_norm": 1.4111071825027466, "learning_rate": 0.00014926072845293905, "loss": 2.958788299560547, "step": 1420 }, { "epoch": 0.7723467458817175, "grad_norm": 1.4764567613601685, "learning_rate": 0.00014890010818608006, "loss": 2.830558013916016, "step": 1430 }, { "epoch": 0.7777477720766945, "grad_norm": 1.925668716430664, "learning_rate": 0.00014853948791922108, "loss": 2.8127283096313476, "step": 1440 }, { "epoch": 0.7831487982716716, "grad_norm": 1.1639180183410645, "learning_rate": 0.00014817886765236207, "loss": 2.997245025634766, "step": 1450 }, { "epoch": 0.7885498244666487, "grad_norm": 0.9844375252723694, "learning_rate": 0.00014781824738550306, "loss": 2.9099292755126953, "step": 1460 }, { "epoch": 0.7939508506616257, "grad_norm": 1.1210323572158813, "learning_rate": 0.00014745762711864408, "loss": 3.145793342590332, "step": 1470 }, { "epoch": 0.7993518768566028, "grad_norm": 1.2205151319503784, "learning_rate": 0.000147133068878471, "loss": 2.8995004653930665, "step": 1480 }, { "epoch": 0.8047529030515798, "grad_norm": 1.321806788444519, "learning_rate": 0.00014677244861161198, "loss": 2.916131019592285, "step": 1490 }, { "epoch": 0.8101539292465568, "grad_norm": 1.5386055707931519, "learning_rate": 0.00014641182834475297, "loss": 2.925608253479004, "step": 1500 }, { "epoch": 0.8155549554415339, "grad_norm": 1.0212986469268799, "learning_rate": 0.000146051208077894, "loss": 2.9698795318603515, "step": 1510 }, { "epoch": 0.820955981636511, "grad_norm": 1.1042375564575195, "learning_rate": 0.000145690587811035, "loss": 2.856062126159668, "step": 1520 }, { "epoch": 0.826357007831488, "grad_norm": 1.3728346824645996, "learning_rate": 0.000145329967544176, "loss": 2.9228466033935545, "step": 1530 }, { "epoch": 0.831758034026465, "grad_norm": 1.203946828842163, "learning_rate": 0.00014496934727731698, "loss": 2.8147531509399415, "step": 1540 }, { "epoch": 0.837159060221442, "grad_norm": 1.2796454429626465, "learning_rate": 0.000144608727010458, "loss": 2.8535860061645506, "step": 1550 }, { "epoch": 0.8425600864164191, "grad_norm": 1.708990216255188, "learning_rate": 0.000144248106743599, "loss": 2.9603645324707033, "step": 1560 }, { "epoch": 0.8479611126113962, "grad_norm": 1.1417936086654663, "learning_rate": 0.00014388748647674, "loss": 2.927590560913086, "step": 1570 }, { "epoch": 0.8533621388063732, "grad_norm": 1.289814829826355, "learning_rate": 0.000143526866209881, "loss": 3.0022300720214843, "step": 1580 }, { "epoch": 0.8587631650013503, "grad_norm": 1.4492555856704712, "learning_rate": 0.00014316624594302202, "loss": 2.9754718780517577, "step": 1590 }, { "epoch": 0.8641641911963273, "grad_norm": 1.4755349159240723, "learning_rate": 0.000142805625676163, "loss": 2.852312469482422, "step": 1600 }, { "epoch": 0.8695652173913043, "grad_norm": 1.0230276584625244, "learning_rate": 0.00014244500540930402, "loss": 3.0102075576782226, "step": 1610 }, { "epoch": 0.8749662435862814, "grad_norm": 1.3246142864227295, "learning_rate": 0.000142084385142445, "loss": 2.9929609298706055, "step": 1620 }, { "epoch": 0.8803672697812585, "grad_norm": 1.339630365371704, "learning_rate": 0.000141723764875586, "loss": 3.0588592529296874, "step": 1630 }, { "epoch": 0.8857682959762355, "grad_norm": 1.2536578178405762, "learning_rate": 0.00014136314460872702, "loss": 2.7730195999145506, "step": 1640 }, { "epoch": 0.8911693221712126, "grad_norm": 0.9663991332054138, "learning_rate": 0.00014100252434186804, "loss": 2.8569766998291017, "step": 1650 }, { "epoch": 0.8965703483661895, "grad_norm": 1.0282679796218872, "learning_rate": 0.000140641904075009, "loss": 3.037453460693359, "step": 1660 }, { "epoch": 0.9019713745611666, "grad_norm": 1.2708044052124023, "learning_rate": 0.00014028128380815002, "loss": 2.9699851989746096, "step": 1670 }, { "epoch": 0.9073724007561437, "grad_norm": 1.4541798830032349, "learning_rate": 0.00013992066354129103, "loss": 2.8832717895507813, "step": 1680 }, { "epoch": 0.9127734269511207, "grad_norm": 1.4256030321121216, "learning_rate": 0.00013956004327443205, "loss": 2.948398208618164, "step": 1690 }, { "epoch": 0.9181744531460978, "grad_norm": 1.3157241344451904, "learning_rate": 0.00013919942300757301, "loss": 2.836800384521484, "step": 1700 }, { "epoch": 0.9235754793410748, "grad_norm": 1.3309694528579712, "learning_rate": 0.00013883880274071403, "loss": 2.900766944885254, "step": 1710 }, { "epoch": 0.9289765055360518, "grad_norm": 0.9171543121337891, "learning_rate": 0.00013847818247385505, "loss": 2.8924545288085937, "step": 1720 }, { "epoch": 0.9343775317310289, "grad_norm": 1.0325075387954712, "learning_rate": 0.00013811756220699604, "loss": 2.9084177017211914, "step": 1730 }, { "epoch": 0.939778557926006, "grad_norm": 1.4920449256896973, "learning_rate": 0.00013775694194013703, "loss": 2.810170555114746, "step": 1740 }, { "epoch": 0.945179584120983, "grad_norm": 1.1236705780029297, "learning_rate": 0.00013739632167327804, "loss": 3.0464658737182617, "step": 1750 }, { "epoch": 0.9505806103159601, "grad_norm": 1.1558237075805664, "learning_rate": 0.00013703570140641903, "loss": 3.0324066162109373, "step": 1760 }, { "epoch": 0.955981636510937, "grad_norm": 1.1385915279388428, "learning_rate": 0.00013667508113956005, "loss": 2.9241012573242187, "step": 1770 }, { "epoch": 0.9613826627059141, "grad_norm": 1.3047113418579102, "learning_rate": 0.00013631446087270104, "loss": 2.9352169036865234, "step": 1780 }, { "epoch": 0.9667836889008912, "grad_norm": 1.548905849456787, "learning_rate": 0.00013595384060584206, "loss": 3.0046308517456053, "step": 1790 }, { "epoch": 0.9721847150958682, "grad_norm": 1.1485257148742676, "learning_rate": 0.00013559322033898305, "loss": 2.868555450439453, "step": 1800 }, { "epoch": 0.9775857412908453, "grad_norm": 1.366213083267212, "learning_rate": 0.00013523260007212407, "loss": 3.086391067504883, "step": 1810 }, { "epoch": 0.9829867674858223, "grad_norm": 2.4089717864990234, "learning_rate": 0.00013487197980526506, "loss": 3.0984577178955077, "step": 1820 }, { "epoch": 0.9883877936807993, "grad_norm": 1.1161662340164185, "learning_rate": 0.00013451135953840605, "loss": 2.915675163269043, "step": 1830 }, { "epoch": 0.9937888198757764, "grad_norm": 1.1393717527389526, "learning_rate": 0.00013415073927154706, "loss": 2.879562568664551, "step": 1840 }, { "epoch": 0.9991898460707535, "grad_norm": 1.4226765632629395, "learning_rate": 0.00013379011900468808, "loss": 2.8654153823852537, "step": 1850 }, { "epoch": 1.0043208209559817, "grad_norm": 1.7910537719726562, "learning_rate": 0.00013342949873782907, "loss": 3.015969085693359, "step": 1860 }, { "epoch": 1.0097218471509586, "grad_norm": 1.1190510988235474, "learning_rate": 0.00013306887847097006, "loss": 2.87933235168457, "step": 1870 }, { "epoch": 1.0151228733459357, "grad_norm": 1.1738801002502441, "learning_rate": 0.00013270825820411108, "loss": 2.7941356658935548, "step": 1880 }, { "epoch": 1.0205238995409127, "grad_norm": 1.0437875986099243, "learning_rate": 0.0001323476379372521, "loss": 2.896015167236328, "step": 1890 }, { "epoch": 1.0259249257358898, "grad_norm": 1.4144948720932007, "learning_rate": 0.00013198701767039308, "loss": 2.843356704711914, "step": 1900 }, { "epoch": 1.0313259519308668, "grad_norm": 1.732194423675537, "learning_rate": 0.00013162639740353407, "loss": 2.8961923599243162, "step": 1910 }, { "epoch": 1.036726978125844, "grad_norm": 1.227655291557312, "learning_rate": 0.0001312657771366751, "loss": 2.8469051361083983, "step": 1920 }, { "epoch": 1.042128004320821, "grad_norm": 1.2534716129302979, "learning_rate": 0.00013090515686981608, "loss": 2.8952392578125, "step": 1930 }, { "epoch": 1.047529030515798, "grad_norm": 1.2732499837875366, "learning_rate": 0.0001305445366029571, "loss": 2.9137599945068358, "step": 1940 }, { "epoch": 1.052930056710775, "grad_norm": 1.1516661643981934, "learning_rate": 0.0001301839163360981, "loss": 2.8422595977783205, "step": 1950 }, { "epoch": 1.0583310829057522, "grad_norm": 1.7162693738937378, "learning_rate": 0.00012982329606923908, "loss": 3.0195697784423827, "step": 1960 }, { "epoch": 1.0637321091007292, "grad_norm": 1.1824109554290771, "learning_rate": 0.0001294626758023801, "loss": 2.830906867980957, "step": 1970 }, { "epoch": 1.069133135295706, "grad_norm": 1.246450424194336, "learning_rate": 0.0001291020555355211, "loss": 2.925843048095703, "step": 1980 }, { "epoch": 1.0745341614906831, "grad_norm": 1.919395923614502, "learning_rate": 0.0001287414352686621, "loss": 2.845474433898926, "step": 1990 }, { "epoch": 1.0799351876856602, "grad_norm": 1.2120916843414307, "learning_rate": 0.0001283808150018031, "loss": 2.9256092071533204, "step": 2000 }, { "epoch": 1.0853362138806373, "grad_norm": 1.1324710845947266, "learning_rate": 0.0001280201947349441, "loss": 2.9663860321044924, "step": 2010 }, { "epoch": 1.0907372400756143, "grad_norm": 1.6326500177383423, "learning_rate": 0.00012765957446808513, "loss": 2.7034042358398436, "step": 2020 }, { "epoch": 1.0961382662705914, "grad_norm": 1.3509035110473633, "learning_rate": 0.00012729895420122612, "loss": 2.633696746826172, "step": 2030 }, { "epoch": 1.1015392924655685, "grad_norm": 1.0838254690170288, "learning_rate": 0.0001269383339343671, "loss": 2.8899274826049806, "step": 2040 }, { "epoch": 1.1069403186605455, "grad_norm": 1.6641477346420288, "learning_rate": 0.00012657771366750812, "loss": 2.909597396850586, "step": 2050 }, { "epoch": 1.1123413448555226, "grad_norm": 1.1637628078460693, "learning_rate": 0.00012621709340064914, "loss": 2.804787826538086, "step": 2060 }, { "epoch": 1.1177423710504997, "grad_norm": 1.4898242950439453, "learning_rate": 0.00012585647313379013, "loss": 2.9821834564208984, "step": 2070 }, { "epoch": 1.1231433972454767, "grad_norm": 1.4179521799087524, "learning_rate": 0.00012549585286693112, "loss": 3.029928779602051, "step": 2080 }, { "epoch": 1.1285444234404536, "grad_norm": 1.3157166242599487, "learning_rate": 0.00012513523260007214, "loss": 2.682559585571289, "step": 2090 }, { "epoch": 1.1339454496354306, "grad_norm": 1.190956950187683, "learning_rate": 0.00012477461233321313, "loss": 3.0344194412231444, "step": 2100 }, { "epoch": 1.1393464758304077, "grad_norm": 1.0837123394012451, "learning_rate": 0.00012441399206635415, "loss": 2.693293571472168, "step": 2110 }, { "epoch": 1.1447475020253848, "grad_norm": 1.190393328666687, "learning_rate": 0.00012405337179949514, "loss": 2.816008949279785, "step": 2120 }, { "epoch": 1.1501485282203618, "grad_norm": 1.0684547424316406, "learning_rate": 0.00012369275153263613, "loss": 2.9440284729003907, "step": 2130 }, { "epoch": 1.155549554415339, "grad_norm": 1.1556187868118286, "learning_rate": 0.00012333213126577714, "loss": 3.132794952392578, "step": 2140 }, { "epoch": 1.160950580610316, "grad_norm": 1.0935251712799072, "learning_rate": 0.00012297151099891816, "loss": 2.6665658950805664, "step": 2150 }, { "epoch": 1.166351606805293, "grad_norm": 1.1952039003372192, "learning_rate": 0.00012261089073205915, "loss": 2.918911933898926, "step": 2160 }, { "epoch": 1.17175263300027, "grad_norm": 1.1767187118530273, "learning_rate": 0.00012225027046520014, "loss": 2.9808664321899414, "step": 2170 }, { "epoch": 1.1771536591952472, "grad_norm": 1.0405514240264893, "learning_rate": 0.00012188965019834116, "loss": 2.7352481842041017, "step": 2180 }, { "epoch": 1.1825546853902242, "grad_norm": 1.058395266532898, "learning_rate": 0.00012152902993148216, "loss": 2.8880331039428713, "step": 2190 }, { "epoch": 1.187955711585201, "grad_norm": 1.2822115421295166, "learning_rate": 0.00012116840966462315, "loss": 2.978183937072754, "step": 2200 }, { "epoch": 1.1933567377801781, "grad_norm": 1.2222055196762085, "learning_rate": 0.00012080778939776415, "loss": 3.0284900665283203, "step": 2210 }, { "epoch": 1.1987577639751552, "grad_norm": 1.204604983329773, "learning_rate": 0.00012044716913090517, "loss": 2.789278984069824, "step": 2220 }, { "epoch": 1.2041587901701323, "grad_norm": 1.3971943855285645, "learning_rate": 0.00012008654886404616, "loss": 2.9389053344726563, "step": 2230 }, { "epoch": 1.2095598163651093, "grad_norm": 1.3112043142318726, "learning_rate": 0.00011972592859718716, "loss": 2.8317853927612306, "step": 2240 }, { "epoch": 1.2149608425600864, "grad_norm": 1.3461347818374634, "learning_rate": 0.00011936530833032817, "loss": 2.8396333694458007, "step": 2250 }, { "epoch": 1.2203618687550635, "grad_norm": 1.221990942955017, "learning_rate": 0.00011900468806346918, "loss": 2.9295183181762696, "step": 2260 }, { "epoch": 1.2257628949500405, "grad_norm": 1.1670557260513306, "learning_rate": 0.00011864406779661017, "loss": 2.923117828369141, "step": 2270 }, { "epoch": 1.2311639211450176, "grad_norm": 2.0452191829681396, "learning_rate": 0.00011828344752975118, "loss": 2.8732458114624024, "step": 2280 }, { "epoch": 1.2365649473399947, "grad_norm": 1.1265263557434082, "learning_rate": 0.00011792282726289218, "loss": 3.0273326873779296, "step": 2290 }, { "epoch": 1.2419659735349717, "grad_norm": 1.4754787683486938, "learning_rate": 0.00011756220699603317, "loss": 2.949677085876465, "step": 2300 }, { "epoch": 1.2473669997299486, "grad_norm": 1.2231955528259277, "learning_rate": 0.00011720158672917419, "loss": 2.854171943664551, "step": 2310 }, { "epoch": 1.2527680259249259, "grad_norm": 1.2537592649459839, "learning_rate": 0.00011684096646231519, "loss": 3.004684829711914, "step": 2320 }, { "epoch": 1.2581690521199027, "grad_norm": 1.2691115140914917, "learning_rate": 0.00011648034619545618, "loss": 2.8662010192871095, "step": 2330 }, { "epoch": 1.2635700783148798, "grad_norm": 1.2742242813110352, "learning_rate": 0.00011611972592859719, "loss": 3.037505340576172, "step": 2340 }, { "epoch": 1.2689711045098568, "grad_norm": 1.6218504905700684, "learning_rate": 0.0001157591056617382, "loss": 2.825956916809082, "step": 2350 }, { "epoch": 1.274372130704834, "grad_norm": 1.5230934619903564, "learning_rate": 0.0001153984853948792, "loss": 2.9821327209472654, "step": 2360 }, { "epoch": 1.279773156899811, "grad_norm": 1.3738056421279907, "learning_rate": 0.0001150378651280202, "loss": 2.6793663024902346, "step": 2370 }, { "epoch": 1.285174183094788, "grad_norm": 1.259009599685669, "learning_rate": 0.0001146772448611612, "loss": 2.8002511978149416, "step": 2380 }, { "epoch": 1.290575209289765, "grad_norm": 1.174134612083435, "learning_rate": 0.00011431662459430222, "loss": 2.929630088806152, "step": 2390 }, { "epoch": 1.2959762354847422, "grad_norm": 1.1415808200836182, "learning_rate": 0.0001139560043274432, "loss": 2.8630245208740233, "step": 2400 }, { "epoch": 1.3013772616797192, "grad_norm": 1.1964163780212402, "learning_rate": 0.00011359538406058421, "loss": 2.839987373352051, "step": 2410 }, { "epoch": 1.306778287874696, "grad_norm": 1.6634154319763184, "learning_rate": 0.00011323476379372521, "loss": 2.9448307037353514, "step": 2420 }, { "epoch": 1.3121793140696734, "grad_norm": 1.0702461004257202, "learning_rate": 0.0001128741435268662, "loss": 2.918461799621582, "step": 2430 }, { "epoch": 1.3175803402646502, "grad_norm": 1.5468103885650635, "learning_rate": 0.00011251352326000721, "loss": 2.7711483001708985, "step": 2440 }, { "epoch": 1.3229813664596273, "grad_norm": 1.667710781097412, "learning_rate": 0.00011215290299314822, "loss": 2.8188310623168946, "step": 2450 }, { "epoch": 1.3283823926546043, "grad_norm": 1.4836740493774414, "learning_rate": 0.00011179228272628923, "loss": 2.8318965911865233, "step": 2460 }, { "epoch": 1.3337834188495814, "grad_norm": 1.0029267072677612, "learning_rate": 0.00011143166245943022, "loss": 2.8626220703125, "step": 2470 }, { "epoch": 1.3391844450445585, "grad_norm": 1.4887958765029907, "learning_rate": 0.00011107104219257122, "loss": 2.851857566833496, "step": 2480 }, { "epoch": 1.3445854712395355, "grad_norm": 1.4411993026733398, "learning_rate": 0.00011071042192571224, "loss": 2.7436094284057617, "step": 2490 }, { "epoch": 1.3499864974345126, "grad_norm": 1.478190541267395, "learning_rate": 0.00011034980165885323, "loss": 2.875034523010254, "step": 2500 }, { "epoch": 1.3553875236294897, "grad_norm": 1.5047866106033325, "learning_rate": 0.00010998918139199423, "loss": 2.920446014404297, "step": 2510 }, { "epoch": 1.3607885498244667, "grad_norm": 1.0506335496902466, "learning_rate": 0.00010962856112513525, "loss": 3.1022211074829102, "step": 2520 }, { "epoch": 1.3661895760194436, "grad_norm": 1.2894558906555176, "learning_rate": 0.00010926794085827623, "loss": 2.7895151138305665, "step": 2530 }, { "epoch": 1.3715906022144209, "grad_norm": 1.5136513710021973, "learning_rate": 0.00010890732059141724, "loss": 2.850429344177246, "step": 2540 }, { "epoch": 1.3769916284093977, "grad_norm": 1.5565638542175293, "learning_rate": 0.00010854670032455825, "loss": 2.8842002868652346, "step": 2550 }, { "epoch": 1.3823926546043748, "grad_norm": 1.4687844514846802, "learning_rate": 0.00010818608005769926, "loss": 2.767740821838379, "step": 2560 }, { "epoch": 1.3877936807993518, "grad_norm": 1.247577428817749, "learning_rate": 0.00010782545979084024, "loss": 2.767527961730957, "step": 2570 }, { "epoch": 1.393194706994329, "grad_norm": 1.1225169897079468, "learning_rate": 0.00010746483952398126, "loss": 2.880569267272949, "step": 2580 }, { "epoch": 1.398595733189306, "grad_norm": 1.8888438940048218, "learning_rate": 0.00010710421925712226, "loss": 2.7435197830200195, "step": 2590 }, { "epoch": 1.403996759384283, "grad_norm": 1.1343549489974976, "learning_rate": 0.00010674359899026325, "loss": 2.788307952880859, "step": 2600 }, { "epoch": 1.40939778557926, "grad_norm": 1.075430154800415, "learning_rate": 0.00010638297872340425, "loss": 2.899631881713867, "step": 2610 }, { "epoch": 1.4147988117742372, "grad_norm": 1.4076910018920898, "learning_rate": 0.00010602235845654527, "loss": 2.891121673583984, "step": 2620 }, { "epoch": 1.4201998379692142, "grad_norm": 1.2757014036178589, "learning_rate": 0.00010566173818968625, "loss": 2.963408279418945, "step": 2630 }, { "epoch": 1.425600864164191, "grad_norm": 1.5843652486801147, "learning_rate": 0.00010530111792282726, "loss": 2.938973808288574, "step": 2640 }, { "epoch": 1.4310018903591684, "grad_norm": 1.2759486436843872, "learning_rate": 0.00010494049765596827, "loss": 2.8045913696289064, "step": 2650 }, { "epoch": 1.4364029165541452, "grad_norm": 1.1256285905838013, "learning_rate": 0.00010457987738910929, "loss": 2.7867507934570312, "step": 2660 }, { "epoch": 1.4418039427491223, "grad_norm": 1.369463562965393, "learning_rate": 0.00010421925712225028, "loss": 2.8050779342651366, "step": 2670 }, { "epoch": 1.4472049689440993, "grad_norm": 1.3944294452667236, "learning_rate": 0.00010385863685539128, "loss": 2.8491235733032227, "step": 2680 }, { "epoch": 1.4526059951390764, "grad_norm": 1.236985206604004, "learning_rate": 0.00010349801658853228, "loss": 2.8627124786376954, "step": 2690 }, { "epoch": 1.4580070213340535, "grad_norm": 1.3797553777694702, "learning_rate": 0.00010313739632167327, "loss": 2.936333656311035, "step": 2700 }, { "epoch": 1.4634080475290305, "grad_norm": 1.3635659217834473, "learning_rate": 0.00010277677605481429, "loss": 2.944474983215332, "step": 2710 }, { "epoch": 1.4688090737240076, "grad_norm": 1.476051926612854, "learning_rate": 0.0001024161557879553, "loss": 2.748017120361328, "step": 2720 }, { "epoch": 1.4742100999189847, "grad_norm": 1.3065634965896606, "learning_rate": 0.0001020555355210963, "loss": 2.7723691940307615, "step": 2730 }, { "epoch": 1.4796111261139617, "grad_norm": 1.415992259979248, "learning_rate": 0.00010169491525423729, "loss": 2.8298103332519533, "step": 2740 }, { "epoch": 1.4850121523089386, "grad_norm": 1.5384076833724976, "learning_rate": 0.0001013342949873783, "loss": 2.906233787536621, "step": 2750 }, { "epoch": 1.4904131785039159, "grad_norm": 1.3399168252944946, "learning_rate": 0.00010097367472051931, "loss": 2.9528438568115236, "step": 2760 }, { "epoch": 1.4958142046988927, "grad_norm": 1.3242063522338867, "learning_rate": 0.0001006130544536603, "loss": 2.9394319534301756, "step": 2770 }, { "epoch": 1.50121523089387, "grad_norm": 1.5000715255737305, "learning_rate": 0.0001002524341868013, "loss": 2.883406639099121, "step": 2780 }, { "epoch": 1.5066162570888468, "grad_norm": 1.272521734237671, "learning_rate": 9.98918139199423e-05, "loss": 2.8342365264892577, "step": 2790 }, { "epoch": 1.512017283283824, "grad_norm": 1.1823383569717407, "learning_rate": 9.953119365308331e-05, "loss": 2.7939565658569334, "step": 2800 }, { "epoch": 1.517418309478801, "grad_norm": 1.5179256200790405, "learning_rate": 9.917057338622431e-05, "loss": 2.871601867675781, "step": 2810 }, { "epoch": 1.522819335673778, "grad_norm": 1.7326610088348389, "learning_rate": 9.880995311936532e-05, "loss": 2.96075553894043, "step": 2820 }, { "epoch": 1.528220361868755, "grad_norm": 1.073715329170227, "learning_rate": 9.844933285250632e-05, "loss": 2.952979850769043, "step": 2830 }, { "epoch": 1.5336213880637322, "grad_norm": 1.590678095817566, "learning_rate": 9.808871258564731e-05, "loss": 3.0092388153076173, "step": 2840 }, { "epoch": 1.5390224142587092, "grad_norm": 1.2470190525054932, "learning_rate": 9.772809231878833e-05, "loss": 2.8631858825683594, "step": 2850 }, { "epoch": 1.544423440453686, "grad_norm": 1.2047282457351685, "learning_rate": 9.736747205192932e-05, "loss": 2.8016355514526365, "step": 2860 }, { "epoch": 1.5498244666486634, "grad_norm": 1.467477798461914, "learning_rate": 9.700685178507033e-05, "loss": 2.796140670776367, "step": 2870 }, { "epoch": 1.5552254928436402, "grad_norm": 1.2711927890777588, "learning_rate": 9.664623151821132e-05, "loss": 2.994451332092285, "step": 2880 }, { "epoch": 1.5606265190386175, "grad_norm": 1.808205008506775, "learning_rate": 9.628561125135233e-05, "loss": 2.9458087921142577, "step": 2890 }, { "epoch": 1.5660275452335943, "grad_norm": 1.0437573194503784, "learning_rate": 9.592499098449333e-05, "loss": 2.815134620666504, "step": 2900 }, { "epoch": 1.5714285714285714, "grad_norm": 1.0917341709136963, "learning_rate": 9.556437071763433e-05, "loss": 2.9576305389404296, "step": 2910 }, { "epoch": 1.5768295976235485, "grad_norm": 1.5794211626052856, "learning_rate": 9.520375045077534e-05, "loss": 3.061453437805176, "step": 2920 }, { "epoch": 1.5822306238185255, "grad_norm": 1.0889885425567627, "learning_rate": 9.484313018391634e-05, "loss": 2.8470460891723635, "step": 2930 }, { "epoch": 1.5876316500135026, "grad_norm": 1.1744705438613892, "learning_rate": 9.448250991705734e-05, "loss": 2.74056453704834, "step": 2940 }, { "epoch": 1.5930326762084797, "grad_norm": 1.3190598487854004, "learning_rate": 9.412188965019835e-05, "loss": 2.929494857788086, "step": 2950 }, { "epoch": 1.5984337024034567, "grad_norm": 1.3507816791534424, "learning_rate": 9.376126938333935e-05, "loss": 2.9215248107910154, "step": 2960 }, { "epoch": 1.6038347285984336, "grad_norm": 1.2186241149902344, "learning_rate": 9.340064911648035e-05, "loss": 2.981372833251953, "step": 2970 }, { "epoch": 1.6092357547934109, "grad_norm": 1.780598759651184, "learning_rate": 9.304002884962136e-05, "loss": 2.9505966186523436, "step": 2980 }, { "epoch": 1.6146367809883877, "grad_norm": 1.152493953704834, "learning_rate": 9.267940858276235e-05, "loss": 2.8313037872314455, "step": 2990 }, { "epoch": 1.620037807183365, "grad_norm": 1.4987826347351074, "learning_rate": 9.231878831590337e-05, "loss": 2.811454963684082, "step": 3000 }, { "epoch": 1.6254388333783418, "grad_norm": 1.597718358039856, "learning_rate": 9.195816804904436e-05, "loss": 2.914075469970703, "step": 3010 }, { "epoch": 1.630839859573319, "grad_norm": 1.1017688512802124, "learning_rate": 9.159754778218537e-05, "loss": 2.7891429901123046, "step": 3020 }, { "epoch": 1.636240885768296, "grad_norm": 1.1155205965042114, "learning_rate": 9.123692751532636e-05, "loss": 3.093463325500488, "step": 3030 }, { "epoch": 1.641641911963273, "grad_norm": 1.2455596923828125, "learning_rate": 9.087630724846737e-05, "loss": 2.8561824798583983, "step": 3040 }, { "epoch": 1.64704293815825, "grad_norm": 1.2931147813796997, "learning_rate": 9.051568698160837e-05, "loss": 2.9132923126220702, "step": 3050 }, { "epoch": 1.6524439643532272, "grad_norm": 1.1059184074401855, "learning_rate": 9.015506671474937e-05, "loss": 2.897763252258301, "step": 3060 }, { "epoch": 1.6578449905482042, "grad_norm": 1.7151881456375122, "learning_rate": 8.979444644789038e-05, "loss": 2.8465444564819338, "step": 3070 }, { "epoch": 1.663246016743181, "grad_norm": 1.1424932479858398, "learning_rate": 8.943382618103138e-05, "loss": 2.8876953125, "step": 3080 }, { "epoch": 1.6686470429381584, "grad_norm": 1.2525825500488281, "learning_rate": 8.907320591417238e-05, "loss": 2.834360885620117, "step": 3090 }, { "epoch": 1.6740480691331352, "grad_norm": 1.4049657583236694, "learning_rate": 8.871258564731339e-05, "loss": 2.711612510681152, "step": 3100 }, { "epoch": 1.6794490953281125, "grad_norm": 1.2100424766540527, "learning_rate": 8.835196538045438e-05, "loss": 2.9480472564697267, "step": 3110 }, { "epoch": 1.6848501215230893, "grad_norm": 1.4140290021896362, "learning_rate": 8.79913451135954e-05, "loss": 2.8190261840820314, "step": 3120 }, { "epoch": 1.6902511477180664, "grad_norm": 1.2633049488067627, "learning_rate": 8.763072484673638e-05, "loss": 2.941916656494141, "step": 3130 }, { "epoch": 1.6956521739130435, "grad_norm": 1.1417691707611084, "learning_rate": 8.72701045798774e-05, "loss": 2.9863866806030273, "step": 3140 }, { "epoch": 1.7010532001080205, "grad_norm": 2.104973077774048, "learning_rate": 8.690948431301839e-05, "loss": 2.803839683532715, "step": 3150 }, { "epoch": 1.7064542263029976, "grad_norm": 1.425844669342041, "learning_rate": 8.65488640461594e-05, "loss": 3.081823539733887, "step": 3160 }, { "epoch": 1.7118552524979747, "grad_norm": 1.4708054065704346, "learning_rate": 8.61882437793004e-05, "loss": 2.8608123779296877, "step": 3170 }, { "epoch": 1.7172562786929517, "grad_norm": 1.4438467025756836, "learning_rate": 8.58276235124414e-05, "loss": 2.7762155532836914, "step": 3180 }, { "epoch": 1.7226573048879286, "grad_norm": 1.0802394151687622, "learning_rate": 8.54670032455824e-05, "loss": 2.7581295013427733, "step": 3190 }, { "epoch": 1.7280583310829059, "grad_norm": 1.157166600227356, "learning_rate": 8.510638297872341e-05, "loss": 2.84869270324707, "step": 3200 }, { "epoch": 1.7334593572778827, "grad_norm": 1.4431512355804443, "learning_rate": 8.474576271186441e-05, "loss": 2.9934520721435547, "step": 3210 }, { "epoch": 1.73886038347286, "grad_norm": 1.5134201049804688, "learning_rate": 8.438514244500542e-05, "loss": 2.812991142272949, "step": 3220 }, { "epoch": 1.7442614096678368, "grad_norm": 1.4203664064407349, "learning_rate": 8.402452217814642e-05, "loss": 3.0564746856689453, "step": 3230 }, { "epoch": 1.749662435862814, "grad_norm": 1.1712607145309448, "learning_rate": 8.366390191128742e-05, "loss": 2.9220895767211914, "step": 3240 }, { "epoch": 1.755063462057791, "grad_norm": 1.4347518682479858, "learning_rate": 8.330328164442843e-05, "loss": 2.8819692611694334, "step": 3250 }, { "epoch": 1.760464488252768, "grad_norm": 1.464417576789856, "learning_rate": 8.294266137756942e-05, "loss": 2.9063913345336916, "step": 3260 }, { "epoch": 1.765865514447745, "grad_norm": 1.8630777597427368, "learning_rate": 8.258204111071043e-05, "loss": 2.9152355194091797, "step": 3270 }, { "epoch": 1.7712665406427222, "grad_norm": 1.335752248764038, "learning_rate": 8.222142084385142e-05, "loss": 2.7430774688720705, "step": 3280 }, { "epoch": 1.7766675668376992, "grad_norm": 1.2148716449737549, "learning_rate": 8.186080057699244e-05, "loss": 2.8445466995239257, "step": 3290 }, { "epoch": 1.782068593032676, "grad_norm": 1.905722975730896, "learning_rate": 8.150018031013343e-05, "loss": 2.939733123779297, "step": 3300 }, { "epoch": 1.7874696192276534, "grad_norm": 1.8189131021499634, "learning_rate": 8.113956004327443e-05, "loss": 2.835724639892578, "step": 3310 }, { "epoch": 1.7928706454226302, "grad_norm": 1.1840368509292603, "learning_rate": 8.077893977641544e-05, "loss": 2.795110321044922, "step": 3320 }, { "epoch": 1.7982716716176075, "grad_norm": 1.0759291648864746, "learning_rate": 8.041831950955644e-05, "loss": 2.940400314331055, "step": 3330 }, { "epoch": 1.8036726978125843, "grad_norm": 1.7791543006896973, "learning_rate": 8.005769924269744e-05, "loss": 2.8699357986450194, "step": 3340 }, { "epoch": 1.8090737240075614, "grad_norm": 1.4602655172348022, "learning_rate": 7.969707897583845e-05, "loss": 2.8012218475341797, "step": 3350 }, { "epoch": 1.8144747502025385, "grad_norm": 1.1151624917984009, "learning_rate": 7.933645870897944e-05, "loss": 2.8476186752319337, "step": 3360 }, { "epoch": 1.8198757763975155, "grad_norm": 1.7208023071289062, "learning_rate": 7.897583844212046e-05, "loss": 2.9701114654541017, "step": 3370 }, { "epoch": 1.8252768025924926, "grad_norm": 1.149404764175415, "learning_rate": 7.861521817526145e-05, "loss": 2.98309326171875, "step": 3380 }, { "epoch": 1.8306778287874697, "grad_norm": 1.20687735080719, "learning_rate": 7.825459790840246e-05, "loss": 2.822209358215332, "step": 3390 }, { "epoch": 1.8360788549824467, "grad_norm": 1.4077695608139038, "learning_rate": 7.789397764154345e-05, "loss": 2.994576835632324, "step": 3400 }, { "epoch": 1.8414798811774236, "grad_norm": 1.573683500289917, "learning_rate": 7.753335737468446e-05, "loss": 2.878458595275879, "step": 3410 }, { "epoch": 1.8468809073724008, "grad_norm": 1.4314883947372437, "learning_rate": 7.717273710782546e-05, "loss": 2.96690673828125, "step": 3420 }, { "epoch": 1.8522819335673777, "grad_norm": 1.30421781539917, "learning_rate": 7.681211684096646e-05, "loss": 2.686541938781738, "step": 3430 }, { "epoch": 1.857682959762355, "grad_norm": 1.1393938064575195, "learning_rate": 7.645149657410747e-05, "loss": 2.9077068328857423, "step": 3440 }, { "epoch": 1.8630839859573318, "grad_norm": 1.5559734106063843, "learning_rate": 7.609087630724847e-05, "loss": 2.8635486602783202, "step": 3450 }, { "epoch": 1.8684850121523089, "grad_norm": 0.9626815915107727, "learning_rate": 7.573025604038947e-05, "loss": 2.7763059616088865, "step": 3460 }, { "epoch": 1.873886038347286, "grad_norm": 1.4414341449737549, "learning_rate": 7.536963577353048e-05, "loss": 2.9368614196777343, "step": 3470 }, { "epoch": 1.879287064542263, "grad_norm": 1.253332495689392, "learning_rate": 7.500901550667148e-05, "loss": 3.0294315338134767, "step": 3480 }, { "epoch": 1.88468809073724, "grad_norm": 1.4030373096466064, "learning_rate": 7.464839523981248e-05, "loss": 3.022206687927246, "step": 3490 }, { "epoch": 1.8900891169322172, "grad_norm": 1.2243045568466187, "learning_rate": 7.428777497295349e-05, "loss": 2.762860870361328, "step": 3500 }, { "epoch": 1.8954901431271942, "grad_norm": 1.2897733449935913, "learning_rate": 7.392715470609448e-05, "loss": 3.086939811706543, "step": 3510 }, { "epoch": 1.900891169322171, "grad_norm": 1.369197964668274, "learning_rate": 7.35665344392355e-05, "loss": 2.9227487564086916, "step": 3520 }, { "epoch": 1.9062921955171483, "grad_norm": 1.1837520599365234, "learning_rate": 7.320591417237649e-05, "loss": 2.941070556640625, "step": 3530 }, { "epoch": 1.9116932217121252, "grad_norm": 1.255652666091919, "learning_rate": 7.28452939055175e-05, "loss": 2.9143747329711913, "step": 3540 }, { "epoch": 1.9170942479071025, "grad_norm": 1.4575761556625366, "learning_rate": 7.248467363865849e-05, "loss": 2.69130916595459, "step": 3550 }, { "epoch": 1.9224952741020793, "grad_norm": 1.1948922872543335, "learning_rate": 7.21240533717995e-05, "loss": 2.7779094696044924, "step": 3560 }, { "epoch": 1.9278963002970564, "grad_norm": 1.1845192909240723, "learning_rate": 7.17634331049405e-05, "loss": 2.8299768447875975, "step": 3570 }, { "epoch": 1.9332973264920335, "grad_norm": 1.1422815322875977, "learning_rate": 7.14028128380815e-05, "loss": 2.8594533920288088, "step": 3580 }, { "epoch": 1.9386983526870105, "grad_norm": 1.1020344495773315, "learning_rate": 7.10421925712225e-05, "loss": 2.703672981262207, "step": 3590 }, { "epoch": 1.9440993788819876, "grad_norm": 1.4388782978057861, "learning_rate": 7.068157230436351e-05, "loss": 3.1084011077880858, "step": 3600 }, { "epoch": 1.9495004050769646, "grad_norm": 1.1368095874786377, "learning_rate": 7.03209520375045e-05, "loss": 2.87176456451416, "step": 3610 }, { "epoch": 1.9549014312719417, "grad_norm": 1.01320219039917, "learning_rate": 6.996033177064552e-05, "loss": 2.8748231887817384, "step": 3620 }, { "epoch": 1.9603024574669186, "grad_norm": 1.610155701637268, "learning_rate": 6.959971150378651e-05, "loss": 2.8654943466186524, "step": 3630 }, { "epoch": 1.9657034836618958, "grad_norm": 1.0376965999603271, "learning_rate": 6.923909123692752e-05, "loss": 2.6509654998779295, "step": 3640 }, { "epoch": 1.9711045098568727, "grad_norm": 1.4242373704910278, "learning_rate": 6.887847097006851e-05, "loss": 2.867411994934082, "step": 3650 }, { "epoch": 1.97650553605185, "grad_norm": 1.5325580835342407, "learning_rate": 6.851785070320952e-05, "loss": 2.904957962036133, "step": 3660 }, { "epoch": 1.9819065622468268, "grad_norm": 1.498086929321289, "learning_rate": 6.815723043635052e-05, "loss": 2.8906654357910155, "step": 3670 }, { "epoch": 1.987307588441804, "grad_norm": 1.3114972114562988, "learning_rate": 6.779661016949152e-05, "loss": 2.9065877914428713, "step": 3680 }, { "epoch": 1.992708614636781, "grad_norm": 1.106939673423767, "learning_rate": 6.743598990263253e-05, "loss": 3.0156211853027344, "step": 3690 }, { "epoch": 1.998109640831758, "grad_norm": 1.6266611814498901, "learning_rate": 6.707536963577353e-05, "loss": 2.8312339782714844, "step": 3700 }, { "epoch": 2.003240615716986, "grad_norm": 1.435448408126831, "learning_rate": 6.671474936891454e-05, "loss": 2.837477684020996, "step": 3710 }, { "epoch": 2.0086416419119635, "grad_norm": 1.2783434391021729, "learning_rate": 6.635412910205554e-05, "loss": 2.7862470626831053, "step": 3720 }, { "epoch": 2.0140426681069403, "grad_norm": 1.1178573369979858, "learning_rate": 6.599350883519654e-05, "loss": 2.963496780395508, "step": 3730 }, { "epoch": 2.019443694301917, "grad_norm": 1.3841524124145508, "learning_rate": 6.563288856833755e-05, "loss": 2.7506023406982423, "step": 3740 }, { "epoch": 2.0248447204968945, "grad_norm": 1.5582271814346313, "learning_rate": 6.527226830147855e-05, "loss": 2.8993911743164062, "step": 3750 }, { "epoch": 2.0302457466918713, "grad_norm": 1.4495958089828491, "learning_rate": 6.491164803461954e-05, "loss": 2.712095260620117, "step": 3760 }, { "epoch": 2.0356467728868486, "grad_norm": 1.4813344478607178, "learning_rate": 6.455102776776056e-05, "loss": 2.8489871978759767, "step": 3770 }, { "epoch": 2.0410477990818254, "grad_norm": 1.3987780809402466, "learning_rate": 6.419040750090155e-05, "loss": 2.8386003494262697, "step": 3780 }, { "epoch": 2.0464488252768027, "grad_norm": 1.186592936515808, "learning_rate": 6.382978723404256e-05, "loss": 2.741067123413086, "step": 3790 }, { "epoch": 2.0518498514717796, "grad_norm": 1.0305873155593872, "learning_rate": 6.346916696718355e-05, "loss": 2.867872619628906, "step": 3800 }, { "epoch": 2.057250877666757, "grad_norm": 1.1601252555847168, "learning_rate": 6.310854670032457e-05, "loss": 2.8808584213256836, "step": 3810 }, { "epoch": 2.0626519038617337, "grad_norm": 2.6286604404449463, "learning_rate": 6.274792643346556e-05, "loss": 2.964664077758789, "step": 3820 }, { "epoch": 2.068052930056711, "grad_norm": 1.5530626773834229, "learning_rate": 6.238730616660656e-05, "loss": 2.6725963592529296, "step": 3830 }, { "epoch": 2.073453956251688, "grad_norm": 1.1765906810760498, "learning_rate": 6.202668589974757e-05, "loss": 2.8030033111572266, "step": 3840 }, { "epoch": 2.0788549824466647, "grad_norm": 1.6987247467041016, "learning_rate": 6.166606563288857e-05, "loss": 2.861286163330078, "step": 3850 }, { "epoch": 2.084256008641642, "grad_norm": 2.2734384536743164, "learning_rate": 6.130544536602957e-05, "loss": 2.7523921966552733, "step": 3860 }, { "epoch": 2.089657034836619, "grad_norm": 1.7829551696777344, "learning_rate": 6.094482509917058e-05, "loss": 2.8400415420532226, "step": 3870 }, { "epoch": 2.095058061031596, "grad_norm": 1.6721340417861938, "learning_rate": 6.0584204832311575e-05, "loss": 2.8418838500976564, "step": 3880 }, { "epoch": 2.100459087226573, "grad_norm": 1.4000741243362427, "learning_rate": 6.0223584565452585e-05, "loss": 2.817787933349609, "step": 3890 }, { "epoch": 2.10586011342155, "grad_norm": 1.3177955150604248, "learning_rate": 5.986296429859358e-05, "loss": 2.9000282287597656, "step": 3900 }, { "epoch": 2.111261139616527, "grad_norm": 1.1557340621948242, "learning_rate": 5.950234403173459e-05, "loss": 2.8565101623535156, "step": 3910 }, { "epoch": 2.1166621658115043, "grad_norm": 1.2179324626922607, "learning_rate": 5.914172376487559e-05, "loss": 2.782689094543457, "step": 3920 }, { "epoch": 2.122063192006481, "grad_norm": 1.32815682888031, "learning_rate": 5.8781103498016586e-05, "loss": 3.0159387588500977, "step": 3930 }, { "epoch": 2.1274642182014585, "grad_norm": 1.4391170740127563, "learning_rate": 5.8420483231157596e-05, "loss": 2.914225959777832, "step": 3940 }, { "epoch": 2.1328652443964353, "grad_norm": 1.184972882270813, "learning_rate": 5.805986296429859e-05, "loss": 2.998256874084473, "step": 3950 }, { "epoch": 2.138266270591412, "grad_norm": 1.1208049058914185, "learning_rate": 5.76992426974396e-05, "loss": 2.9700386047363283, "step": 3960 }, { "epoch": 2.1436672967863895, "grad_norm": 1.1472302675247192, "learning_rate": 5.73386224305806e-05, "loss": 2.8869691848754884, "step": 3970 }, { "epoch": 2.1490683229813663, "grad_norm": 1.1823166608810425, "learning_rate": 5.69780021637216e-05, "loss": 2.830540657043457, "step": 3980 }, { "epoch": 2.1544693491763436, "grad_norm": 1.3045367002487183, "learning_rate": 5.661738189686261e-05, "loss": 2.7170568466186524, "step": 3990 }, { "epoch": 2.1598703753713204, "grad_norm": 1.2233638763427734, "learning_rate": 5.6256761630003604e-05, "loss": 2.8441373825073244, "step": 4000 }, { "epoch": 2.1652714015662977, "grad_norm": 1.4851471185684204, "learning_rate": 5.5896141363144614e-05, "loss": 2.812620735168457, "step": 4010 }, { "epoch": 2.1706724277612746, "grad_norm": 1.264336109161377, "learning_rate": 5.553552109628561e-05, "loss": 2.767516326904297, "step": 4020 }, { "epoch": 2.176073453956252, "grad_norm": 1.7260786294937134, "learning_rate": 5.5174900829426614e-05, "loss": 2.8412513732910156, "step": 4030 }, { "epoch": 2.1814744801512287, "grad_norm": 1.5279184579849243, "learning_rate": 5.4814280562567625e-05, "loss": 2.8572755813598634, "step": 4040 }, { "epoch": 2.186875506346206, "grad_norm": 1.7421464920043945, "learning_rate": 5.445366029570862e-05, "loss": 2.9832971572875975, "step": 4050 }, { "epoch": 2.192276532541183, "grad_norm": 1.1378761529922485, "learning_rate": 5.409304002884963e-05, "loss": 2.9112874984741213, "step": 4060 }, { "epoch": 2.1976775587361597, "grad_norm": 1.6481519937515259, "learning_rate": 5.373241976199063e-05, "loss": 2.7275985717773437, "step": 4070 }, { "epoch": 2.203078584931137, "grad_norm": 1.4444156885147095, "learning_rate": 5.3371799495131625e-05, "loss": 2.855612564086914, "step": 4080 }, { "epoch": 2.208479611126114, "grad_norm": 1.5712828636169434, "learning_rate": 5.3011179228272636e-05, "loss": 2.8732295989990235, "step": 4090 }, { "epoch": 2.213880637321091, "grad_norm": 1.4751285314559937, "learning_rate": 5.265055896141363e-05, "loss": 3.008199691772461, "step": 4100 }, { "epoch": 2.219281663516068, "grad_norm": 1.2603405714035034, "learning_rate": 5.228993869455464e-05, "loss": 2.8312009811401366, "step": 4110 }, { "epoch": 2.224682689711045, "grad_norm": 1.2152962684631348, "learning_rate": 5.192931842769564e-05, "loss": 2.8798431396484374, "step": 4120 }, { "epoch": 2.230083715906022, "grad_norm": 1.1346954107284546, "learning_rate": 5.1568698160836636e-05, "loss": 2.9093082427978514, "step": 4130 }, { "epoch": 2.2354847421009993, "grad_norm": 1.3649049997329712, "learning_rate": 5.120807789397765e-05, "loss": 2.766147422790527, "step": 4140 }, { "epoch": 2.240885768295976, "grad_norm": 1.8787068128585815, "learning_rate": 5.0847457627118643e-05, "loss": 3.021650695800781, "step": 4150 }, { "epoch": 2.2462867944909535, "grad_norm": 1.2505935430526733, "learning_rate": 5.0486837360259654e-05, "loss": 2.86655330657959, "step": 4160 }, { "epoch": 2.2516878206859303, "grad_norm": 1.3169344663619995, "learning_rate": 5.012621709340065e-05, "loss": 2.784800148010254, "step": 4170 }, { "epoch": 2.257088846880907, "grad_norm": 1.3133145570755005, "learning_rate": 4.9765596826541654e-05, "loss": 2.744234085083008, "step": 4180 }, { "epoch": 2.2624898730758845, "grad_norm": 1.2438435554504395, "learning_rate": 4.940497655968266e-05, "loss": 2.8312517166137696, "step": 4190 }, { "epoch": 2.2678908992708613, "grad_norm": 1.2213644981384277, "learning_rate": 4.9044356292823654e-05, "loss": 2.8160324096679688, "step": 4200 }, { "epoch": 2.2732919254658386, "grad_norm": 1.2122548818588257, "learning_rate": 4.868373602596466e-05, "loss": 2.963540458679199, "step": 4210 }, { "epoch": 2.2786929516608154, "grad_norm": 1.6764824390411377, "learning_rate": 4.832311575910566e-05, "loss": 3.0543262481689455, "step": 4220 }, { "epoch": 2.2840939778557927, "grad_norm": 1.2248793840408325, "learning_rate": 4.7962495492246665e-05, "loss": 2.9278806686401366, "step": 4230 }, { "epoch": 2.2894950040507696, "grad_norm": 1.4941028356552124, "learning_rate": 4.760187522538767e-05, "loss": 2.8444753646850587, "step": 4240 }, { "epoch": 2.294896030245747, "grad_norm": 1.3570294380187988, "learning_rate": 4.724125495852867e-05, "loss": 2.826213836669922, "step": 4250 }, { "epoch": 2.3002970564407237, "grad_norm": 1.3605459928512573, "learning_rate": 4.6880634691669676e-05, "loss": 2.973318099975586, "step": 4260 }, { "epoch": 2.305698082635701, "grad_norm": 1.3374007940292358, "learning_rate": 4.652001442481068e-05, "loss": 2.795336151123047, "step": 4270 }, { "epoch": 2.311099108830678, "grad_norm": 1.211033821105957, "learning_rate": 4.615939415795168e-05, "loss": 2.8860551834106447, "step": 4280 }, { "epoch": 2.316500135025655, "grad_norm": 1.2097424268722534, "learning_rate": 4.5798773891092686e-05, "loss": 2.866344451904297, "step": 4290 }, { "epoch": 2.321901161220632, "grad_norm": 1.436103105545044, "learning_rate": 4.543815362423368e-05, "loss": 2.8824234008789062, "step": 4300 }, { "epoch": 2.327302187415609, "grad_norm": 1.3415106534957886, "learning_rate": 4.5077533357374686e-05, "loss": 2.9298898696899416, "step": 4310 }, { "epoch": 2.332703213610586, "grad_norm": 1.3115299940109253, "learning_rate": 4.471691309051569e-05, "loss": 2.8047321319580076, "step": 4320 }, { "epoch": 2.338104239805563, "grad_norm": 1.430851697921753, "learning_rate": 4.4356292823656694e-05, "loss": 2.9591575622558595, "step": 4330 }, { "epoch": 2.34350526600054, "grad_norm": 1.1957558393478394, "learning_rate": 4.39956725567977e-05, "loss": 2.8753946304321287, "step": 4340 }, { "epoch": 2.348906292195517, "grad_norm": 1.192518949508667, "learning_rate": 4.36350522899387e-05, "loss": 2.8650138854980467, "step": 4350 }, { "epoch": 2.3543073183904943, "grad_norm": 1.6481420993804932, "learning_rate": 4.32744320230797e-05, "loss": 2.784092903137207, "step": 4360 }, { "epoch": 2.359708344585471, "grad_norm": 1.2017136812210083, "learning_rate": 4.29138117562207e-05, "loss": 2.835672378540039, "step": 4370 }, { "epoch": 2.3651093707804485, "grad_norm": 1.2652090787887573, "learning_rate": 4.2553191489361704e-05, "loss": 2.7191650390625, "step": 4380 }, { "epoch": 2.3705103969754253, "grad_norm": 1.2717392444610596, "learning_rate": 4.219257122250271e-05, "loss": 2.6783191680908205, "step": 4390 }, { "epoch": 2.375911423170402, "grad_norm": 1.0883729457855225, "learning_rate": 4.183195095564371e-05, "loss": 2.8368051528930662, "step": 4400 }, { "epoch": 2.3813124493653794, "grad_norm": 1.378779649734497, "learning_rate": 4.147133068878471e-05, "loss": 2.743060493469238, "step": 4410 }, { "epoch": 2.3867134755603563, "grad_norm": 1.799696683883667, "learning_rate": 4.111071042192571e-05, "loss": 2.705653190612793, "step": 4420 }, { "epoch": 2.3921145017553336, "grad_norm": 1.2935140132904053, "learning_rate": 4.0750090155066715e-05, "loss": 2.797167205810547, "step": 4430 }, { "epoch": 2.3975155279503104, "grad_norm": 1.7609946727752686, "learning_rate": 4.038946988820772e-05, "loss": 2.6935022354125975, "step": 4440 }, { "epoch": 2.4029165541452877, "grad_norm": 1.2063777446746826, "learning_rate": 4.002884962134872e-05, "loss": 2.8422033309936525, "step": 4450 }, { "epoch": 2.4083175803402646, "grad_norm": 1.2928175926208496, "learning_rate": 3.966822935448972e-05, "loss": 2.8214353561401366, "step": 4460 }, { "epoch": 2.413718606535242, "grad_norm": 1.493618130683899, "learning_rate": 3.930760908763072e-05, "loss": 2.9253427505493166, "step": 4470 }, { "epoch": 2.4191196327302187, "grad_norm": 1.629323959350586, "learning_rate": 3.8946988820771726e-05, "loss": 2.848074531555176, "step": 4480 }, { "epoch": 2.424520658925196, "grad_norm": 1.2430917024612427, "learning_rate": 3.858636855391273e-05, "loss": 2.7236486434936524, "step": 4490 }, { "epoch": 2.429921685120173, "grad_norm": 1.1484453678131104, "learning_rate": 3.822574828705373e-05, "loss": 2.8111089706420898, "step": 4500 }, { "epoch": 2.43532271131515, "grad_norm": 1.4623491764068604, "learning_rate": 3.786512802019474e-05, "loss": 2.703995132446289, "step": 4510 }, { "epoch": 2.440723737510127, "grad_norm": 1.573211908340454, "learning_rate": 3.750450775333574e-05, "loss": 2.8535009384155274, "step": 4520 }, { "epoch": 2.446124763705104, "grad_norm": 1.4752460718154907, "learning_rate": 3.7143887486476744e-05, "loss": 2.936534118652344, "step": 4530 }, { "epoch": 2.451525789900081, "grad_norm": 1.3244374990463257, "learning_rate": 3.678326721961775e-05, "loss": 2.7790410995483397, "step": 4540 }, { "epoch": 2.456926816095058, "grad_norm": 1.1254388093948364, "learning_rate": 3.642264695275875e-05, "loss": 2.909018707275391, "step": 4550 }, { "epoch": 2.462327842290035, "grad_norm": 1.5923142433166504, "learning_rate": 3.606202668589975e-05, "loss": 2.7277172088623045, "step": 4560 }, { "epoch": 2.467728868485012, "grad_norm": 1.3873192071914673, "learning_rate": 3.570140641904075e-05, "loss": 2.8443859100341795, "step": 4570 }, { "epoch": 2.4731298946799893, "grad_norm": 1.2413684129714966, "learning_rate": 3.5340786152181755e-05, "loss": 2.710672378540039, "step": 4580 }, { "epoch": 2.478530920874966, "grad_norm": 1.9094045162200928, "learning_rate": 3.498016588532276e-05, "loss": 2.821697807312012, "step": 4590 }, { "epoch": 2.4839319470699435, "grad_norm": 1.1247705221176147, "learning_rate": 3.461954561846376e-05, "loss": 2.814626693725586, "step": 4600 }, { "epoch": 2.4893329732649203, "grad_norm": 1.1954268217086792, "learning_rate": 3.425892535160476e-05, "loss": 2.8749109268188477, "step": 4610 }, { "epoch": 2.494733999459897, "grad_norm": 1.1230069398880005, "learning_rate": 3.389830508474576e-05, "loss": 2.9428171157836913, "step": 4620 }, { "epoch": 2.5001350256548744, "grad_norm": 1.1851457357406616, "learning_rate": 3.3537684817886766e-05, "loss": 2.873911666870117, "step": 4630 }, { "epoch": 2.5055360518498517, "grad_norm": 1.5795778036117554, "learning_rate": 3.317706455102777e-05, "loss": 2.805327033996582, "step": 4640 }, { "epoch": 2.5109370780448286, "grad_norm": 1.2687329053878784, "learning_rate": 3.281644428416877e-05, "loss": 2.7486440658569338, "step": 4650 }, { "epoch": 2.5163381042398054, "grad_norm": 1.0893746614456177, "learning_rate": 3.245582401730977e-05, "loss": 2.835002326965332, "step": 4660 }, { "epoch": 2.5217391304347827, "grad_norm": 1.4868460893630981, "learning_rate": 3.209520375045077e-05, "loss": 2.952839660644531, "step": 4670 }, { "epoch": 2.5271401566297595, "grad_norm": 1.178037166595459, "learning_rate": 3.173458348359178e-05, "loss": 2.716649055480957, "step": 4680 }, { "epoch": 2.532541182824737, "grad_norm": 1.5916475057601929, "learning_rate": 3.137396321673278e-05, "loss": 2.774715614318848, "step": 4690 }, { "epoch": 2.5379422090197137, "grad_norm": 1.1122616529464722, "learning_rate": 3.1013342949873784e-05, "loss": 2.913417625427246, "step": 4700 }, { "epoch": 2.5433432352146905, "grad_norm": 1.394402265548706, "learning_rate": 3.065272268301479e-05, "loss": 2.7905054092407227, "step": 4710 }, { "epoch": 2.548744261409668, "grad_norm": 1.2806884050369263, "learning_rate": 3.0292102416155787e-05, "loss": 2.7806386947631836, "step": 4720 }, { "epoch": 2.554145287604645, "grad_norm": 1.3189789056777954, "learning_rate": 2.993148214929679e-05, "loss": 2.7181257247924804, "step": 4730 }, { "epoch": 2.559546313799622, "grad_norm": 1.5680729150772095, "learning_rate": 2.9570861882437795e-05, "loss": 2.6874979019165037, "step": 4740 }, { "epoch": 2.564947339994599, "grad_norm": 1.3960338830947876, "learning_rate": 2.9210241615578798e-05, "loss": 2.889949607849121, "step": 4750 }, { "epoch": 2.570348366189576, "grad_norm": 1.5850731134414673, "learning_rate": 2.88496213487198e-05, "loss": 2.927565002441406, "step": 4760 }, { "epoch": 2.575749392384553, "grad_norm": 1.5673096179962158, "learning_rate": 2.84890010818608e-05, "loss": 2.874621772766113, "step": 4770 }, { "epoch": 2.58115041857953, "grad_norm": 1.8409554958343506, "learning_rate": 2.8128380815001802e-05, "loss": 2.8448713302612303, "step": 4780 }, { "epoch": 2.586551444774507, "grad_norm": 1.4746390581130981, "learning_rate": 2.7767760548142805e-05, "loss": 2.7815601348876955, "step": 4790 }, { "epoch": 2.5919524709694843, "grad_norm": 1.4501030445098877, "learning_rate": 2.7407140281283812e-05, "loss": 2.792118453979492, "step": 4800 }, { "epoch": 2.597353497164461, "grad_norm": 1.3553701639175415, "learning_rate": 2.7046520014424816e-05, "loss": 2.8748315811157226, "step": 4810 }, { "epoch": 2.6027545233594385, "grad_norm": 2.318094491958618, "learning_rate": 2.6685899747565813e-05, "loss": 2.780918502807617, "step": 4820 }, { "epoch": 2.6081555495544153, "grad_norm": 1.292222023010254, "learning_rate": 2.6325279480706816e-05, "loss": 2.7713043212890627, "step": 4830 }, { "epoch": 2.613556575749392, "grad_norm": 1.2859598398208618, "learning_rate": 2.596465921384782e-05, "loss": 3.0159257888793944, "step": 4840 }, { "epoch": 2.6189576019443694, "grad_norm": 1.7709689140319824, "learning_rate": 2.5604038946988823e-05, "loss": 2.755099296569824, "step": 4850 }, { "epoch": 2.6243586281393467, "grad_norm": 1.1535433530807495, "learning_rate": 2.5243418680129827e-05, "loss": 2.964043617248535, "step": 4860 }, { "epoch": 2.6297596543343236, "grad_norm": 1.7700167894363403, "learning_rate": 2.4882798413270827e-05, "loss": 2.819998550415039, "step": 4870 }, { "epoch": 2.6351606805293004, "grad_norm": 1.37820565700531, "learning_rate": 2.4522178146411827e-05, "loss": 2.5572160720825194, "step": 4880 }, { "epoch": 2.6405617067242777, "grad_norm": 1.4028289318084717, "learning_rate": 2.416155787955283e-05, "loss": 2.896478271484375, "step": 4890 }, { "epoch": 2.6459627329192545, "grad_norm": 1.3253787755966187, "learning_rate": 2.3800937612693834e-05, "loss": 2.813775825500488, "step": 4900 }, { "epoch": 2.651363759114232, "grad_norm": 1.2806957960128784, "learning_rate": 2.3440317345834838e-05, "loss": 2.8163272857666017, "step": 4910 }, { "epoch": 2.6567647853092087, "grad_norm": 1.8627846240997314, "learning_rate": 2.307969707897584e-05, "loss": 2.929352951049805, "step": 4920 }, { "epoch": 2.6621658115041855, "grad_norm": 1.3502531051635742, "learning_rate": 2.271907681211684e-05, "loss": 2.9691762924194336, "step": 4930 }, { "epoch": 2.667566837699163, "grad_norm": 1.4862613677978516, "learning_rate": 2.2358456545257845e-05, "loss": 2.900889778137207, "step": 4940 }, { "epoch": 2.67296786389414, "grad_norm": 1.3858481645584106, "learning_rate": 2.199783627839885e-05, "loss": 2.6516948699951173, "step": 4950 }, { "epoch": 2.678368890089117, "grad_norm": 1.4900953769683838, "learning_rate": 2.163721601153985e-05, "loss": 2.848802185058594, "step": 4960 }, { "epoch": 2.683769916284094, "grad_norm": 1.3861231803894043, "learning_rate": 2.1276595744680852e-05, "loss": 2.78155460357666, "step": 4970 }, { "epoch": 2.689170942479071, "grad_norm": 1.1364151239395142, "learning_rate": 2.0915975477821856e-05, "loss": 2.9016733169555664, "step": 4980 }, { "epoch": 2.694571968674048, "grad_norm": 1.4272339344024658, "learning_rate": 2.0555355210962856e-05, "loss": 2.8002323150634765, "step": 4990 }, { "epoch": 2.699972994869025, "grad_norm": 1.528222918510437, "learning_rate": 2.019473494410386e-05, "loss": 2.655875396728516, "step": 5000 }, { "epoch": 2.705374021064002, "grad_norm": 1.831336259841919, "learning_rate": 1.983411467724486e-05, "loss": 2.7217283248901367, "step": 5010 }, { "epoch": 2.7107750472589793, "grad_norm": 1.775886058807373, "learning_rate": 1.9473494410385863e-05, "loss": 2.7329965591430665, "step": 5020 }, { "epoch": 2.716176073453956, "grad_norm": 1.2464665174484253, "learning_rate": 1.9112874143526867e-05, "loss": 2.853206825256348, "step": 5030 }, { "epoch": 2.7215770996489335, "grad_norm": 1.4410955905914307, "learning_rate": 1.875225387666787e-05, "loss": 2.880515480041504, "step": 5040 }, { "epoch": 2.7269781258439103, "grad_norm": 1.5591436624526978, "learning_rate": 1.8391633609808874e-05, "loss": 2.796125030517578, "step": 5050 }, { "epoch": 2.732379152038887, "grad_norm": 1.9035403728485107, "learning_rate": 1.8031013342949874e-05, "loss": 2.9866916656494142, "step": 5060 }, { "epoch": 2.7377801782338644, "grad_norm": 1.8339112997055054, "learning_rate": 1.7670393076090877e-05, "loss": 2.7200239181518553, "step": 5070 }, { "epoch": 2.7431812044288417, "grad_norm": 1.3336868286132812, "learning_rate": 1.730977280923188e-05, "loss": 2.8392318725585937, "step": 5080 }, { "epoch": 2.7485822306238186, "grad_norm": 1.4816274642944336, "learning_rate": 1.694915254237288e-05, "loss": 2.8393041610717775, "step": 5090 }, { "epoch": 2.7539832568187954, "grad_norm": 1.2817270755767822, "learning_rate": 1.6588532275513885e-05, "loss": 2.90909423828125, "step": 5100 }, { "epoch": 2.7593842830137727, "grad_norm": 1.3823027610778809, "learning_rate": 1.6227912008654885e-05, "loss": 2.8546985626220702, "step": 5110 }, { "epoch": 2.7647853092087495, "grad_norm": 1.2762755155563354, "learning_rate": 1.586729174179589e-05, "loss": 2.765005874633789, "step": 5120 }, { "epoch": 2.770186335403727, "grad_norm": 1.3482586145401, "learning_rate": 1.5506671474936892e-05, "loss": 2.7484037399291994, "step": 5130 }, { "epoch": 2.7755873615987037, "grad_norm": 1.3183389902114868, "learning_rate": 1.5146051208077894e-05, "loss": 2.960574150085449, "step": 5140 }, { "epoch": 2.780988387793681, "grad_norm": 1.3310991525650024, "learning_rate": 1.4785430941218897e-05, "loss": 2.705043411254883, "step": 5150 }, { "epoch": 2.786389413988658, "grad_norm": 1.1508463621139526, "learning_rate": 1.44248106743599e-05, "loss": 2.926662254333496, "step": 5160 }, { "epoch": 2.791790440183635, "grad_norm": 1.1211504936218262, "learning_rate": 1.4064190407500901e-05, "loss": 2.807356071472168, "step": 5170 }, { "epoch": 2.797191466378612, "grad_norm": 1.5991333723068237, "learning_rate": 1.3703570140641906e-05, "loss": 2.8748167037963865, "step": 5180 }, { "epoch": 2.802592492573589, "grad_norm": 1.1993247270584106, "learning_rate": 1.3342949873782906e-05, "loss": 2.7265079498291014, "step": 5190 }, { "epoch": 2.807993518768566, "grad_norm": 1.1483798027038574, "learning_rate": 1.298232960692391e-05, "loss": 2.848540496826172, "step": 5200 }, { "epoch": 2.813394544963543, "grad_norm": 1.4875781536102295, "learning_rate": 1.2621709340064913e-05, "loss": 2.9188844680786135, "step": 5210 }, { "epoch": 2.81879557115852, "grad_norm": 1.5197795629501343, "learning_rate": 1.2261089073205914e-05, "loss": 2.813732719421387, "step": 5220 }, { "epoch": 2.824196597353497, "grad_norm": 1.3780430555343628, "learning_rate": 1.1900468806346917e-05, "loss": 2.848189353942871, "step": 5230 }, { "epoch": 2.8295976235484743, "grad_norm": 1.3100169897079468, "learning_rate": 1.153984853948792e-05, "loss": 2.671047401428223, "step": 5240 }, { "epoch": 2.834998649743451, "grad_norm": 1.2699891328811646, "learning_rate": 1.1179228272628923e-05, "loss": 2.966587257385254, "step": 5250 }, { "epoch": 2.8403996759384285, "grad_norm": 1.6229315996170044, "learning_rate": 1.0818608005769924e-05, "loss": 2.689316749572754, "step": 5260 }, { "epoch": 2.8458007021334053, "grad_norm": 1.2821816205978394, "learning_rate": 1.0457987738910928e-05, "loss": 2.8306648254394533, "step": 5270 }, { "epoch": 2.851201728328382, "grad_norm": 1.7271239757537842, "learning_rate": 1.009736747205193e-05, "loss": 2.834494209289551, "step": 5280 }, { "epoch": 2.8566027545233594, "grad_norm": 1.3066142797470093, "learning_rate": 9.736747205192932e-06, "loss": 2.918941116333008, "step": 5290 }, { "epoch": 2.8620037807183367, "grad_norm": 1.8644407987594604, "learning_rate": 9.376126938333935e-06, "loss": 2.9472114562988283, "step": 5300 }, { "epoch": 2.8674048069133136, "grad_norm": 1.625531792640686, "learning_rate": 9.015506671474937e-06, "loss": 2.7238536834716798, "step": 5310 }, { "epoch": 2.8728058331082904, "grad_norm": 1.3371554613113403, "learning_rate": 8.65488640461594e-06, "loss": 2.518412780761719, "step": 5320 }, { "epoch": 2.8782068593032677, "grad_norm": 1.6209046840667725, "learning_rate": 8.294266137756942e-06, "loss": 2.8124130249023436, "step": 5330 }, { "epoch": 2.8836078854982445, "grad_norm": 1.0688458681106567, "learning_rate": 7.933645870897944e-06, "loss": 2.7795385360717773, "step": 5340 }, { "epoch": 2.889008911693222, "grad_norm": 1.4803920984268188, "learning_rate": 7.573025604038947e-06, "loss": 2.9230424880981447, "step": 5350 }, { "epoch": 2.8944099378881987, "grad_norm": 1.569003701210022, "learning_rate": 7.21240533717995e-06, "loss": 2.811898040771484, "step": 5360 }, { "epoch": 2.899810964083176, "grad_norm": 1.4253180027008057, "learning_rate": 6.851785070320953e-06, "loss": 2.786616325378418, "step": 5370 }, { "epoch": 2.905211990278153, "grad_norm": 1.1237375736236572, "learning_rate": 6.491164803461955e-06, "loss": 2.8173397064208983, "step": 5380 }, { "epoch": 2.91061301647313, "grad_norm": 1.6228185892105103, "learning_rate": 6.130544536602957e-06, "loss": 2.7539085388183593, "step": 5390 }, { "epoch": 2.916014042668107, "grad_norm": 1.693569302558899, "learning_rate": 5.76992426974396e-06, "loss": 2.718755340576172, "step": 5400 } ], "logging_steps": 10, "max_steps": 5556, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 200, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.937265938201088e+17, "train_batch_size": 2, "trial_name": null, "trial_params": null }