diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,158136 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 22586, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 4.4275704813876e-05, + "grad_norm": 6.689538478851318, + "learning_rate": 0.0, + "loss": 1.9457, + "step": 1 + }, + { + "epoch": 8.8551409627752e-05, + "grad_norm": 5.639162063598633, + "learning_rate": 8.857395925597875e-09, + "loss": 1.9569, + "step": 2 + }, + { + "epoch": 0.00013282711444162802, + "grad_norm": 5.6324615478515625, + "learning_rate": 1.771479185119575e-08, + "loss": 1.8497, + "step": 3 + }, + { + "epoch": 0.000177102819255504, + "grad_norm": 6.141303062438965, + "learning_rate": 2.6572187776793625e-08, + "loss": 1.9291, + "step": 4 + }, + { + "epoch": 0.00022137852406938004, + "grad_norm": 6.698505401611328, + "learning_rate": 3.54295837023915e-08, + "loss": 2.0918, + "step": 5 + }, + { + "epoch": 0.00026565422888325603, + "grad_norm": 5.092038631439209, + "learning_rate": 4.428697962798938e-08, + "loss": 1.8204, + "step": 6 + }, + { + "epoch": 0.00030992993369713206, + "grad_norm": 6.834112644195557, + "learning_rate": 5.314437555358725e-08, + "loss": 1.878, + "step": 7 + }, + { + "epoch": 0.000354205638511008, + "grad_norm": 6.683759689331055, + "learning_rate": 6.200177147918512e-08, + "loss": 1.952, + "step": 8 + }, + { + "epoch": 0.00039848134332488405, + "grad_norm": 6.366095542907715, + "learning_rate": 7.0859167404783e-08, + "loss": 1.8992, + "step": 9 + }, + { + "epoch": 0.00044275704813876007, + "grad_norm": 6.126683712005615, + "learning_rate": 7.971656333038087e-08, + "loss": 2.0349, + "step": 10 + }, + { + "epoch": 0.00048703275295263604, + "grad_norm": 5.137956619262695, + "learning_rate": 8.857395925597876e-08, + "loss": 1.8092, + "step": 11 + }, + { + "epoch": 0.0005313084577665121, + "grad_norm": 6.8586554527282715, + "learning_rate": 9.743135518157661e-08, + "loss": 1.8222, + "step": 12 + }, + { + "epoch": 0.000575584162580388, + "grad_norm": 6.3602094650268555, + "learning_rate": 1.062887511071745e-07, + "loss": 1.9728, + "step": 13 + }, + { + "epoch": 0.0006198598673942641, + "grad_norm": 13.514198303222656, + "learning_rate": 1.1514614703277237e-07, + "loss": 2.0464, + "step": 14 + }, + { + "epoch": 0.0006641355722081401, + "grad_norm": 9.353273391723633, + "learning_rate": 1.2400354295837025e-07, + "loss": 1.985, + "step": 15 + }, + { + "epoch": 0.000708411277022016, + "grad_norm": 8.078036308288574, + "learning_rate": 1.3286093888396814e-07, + "loss": 1.9899, + "step": 16 + }, + { + "epoch": 0.0007526869818358921, + "grad_norm": 11.279537200927734, + "learning_rate": 1.41718334809566e-07, + "loss": 1.9052, + "step": 17 + }, + { + "epoch": 0.0007969626866497681, + "grad_norm": 8.455512046813965, + "learning_rate": 1.505757307351639e-07, + "loss": 1.8137, + "step": 18 + }, + { + "epoch": 0.0008412383914636441, + "grad_norm": 11.05827808380127, + "learning_rate": 1.5943312666076175e-07, + "loss": 2.0348, + "step": 19 + }, + { + "epoch": 0.0008855140962775201, + "grad_norm": 9.76033878326416, + "learning_rate": 1.682905225863596e-07, + "loss": 1.9154, + "step": 20 + }, + { + "epoch": 0.0009297898010913961, + "grad_norm": 8.24018383026123, + "learning_rate": 1.7714791851195753e-07, + "loss": 1.8322, + "step": 21 + }, + { + "epoch": 0.0009740655059052721, + "grad_norm": 9.126859664916992, + "learning_rate": 1.8600531443755536e-07, + "loss": 1.8423, + "step": 22 + }, + { + "epoch": 0.0010183412107191482, + "grad_norm": 16.66147804260254, + "learning_rate": 1.9486271036315322e-07, + "loss": 2.0337, + "step": 23 + }, + { + "epoch": 0.0010626169155330241, + "grad_norm": 9.886859893798828, + "learning_rate": 2.0372010628875114e-07, + "loss": 1.9249, + "step": 24 + }, + { + "epoch": 0.0011068926203469, + "grad_norm": 9.413965225219727, + "learning_rate": 2.12577502214349e-07, + "loss": 1.99, + "step": 25 + }, + { + "epoch": 0.001151168325160776, + "grad_norm": 9.782798767089844, + "learning_rate": 2.214348981399469e-07, + "loss": 2.0165, + "step": 26 + }, + { + "epoch": 0.0011954440299746523, + "grad_norm": 6.014155387878418, + "learning_rate": 2.3029229406554475e-07, + "loss": 1.7884, + "step": 27 + }, + { + "epoch": 0.0012397197347885282, + "grad_norm": 9.97746753692627, + "learning_rate": 2.3914968999114264e-07, + "loss": 2.0244, + "step": 28 + }, + { + "epoch": 0.0012839954396024042, + "grad_norm": 12.069507598876953, + "learning_rate": 2.480070859167405e-07, + "loss": 2.1479, + "step": 29 + }, + { + "epoch": 0.0013282711444162802, + "grad_norm": 7.834904193878174, + "learning_rate": 2.5686448184233836e-07, + "loss": 2.0096, + "step": 30 + }, + { + "epoch": 0.0013725468492301561, + "grad_norm": 12.252904891967773, + "learning_rate": 2.657218777679363e-07, + "loss": 1.8648, + "step": 31 + }, + { + "epoch": 0.001416822554044032, + "grad_norm": 17.00823211669922, + "learning_rate": 2.7457927369353414e-07, + "loss": 2.0374, + "step": 32 + }, + { + "epoch": 0.0014610982588579083, + "grad_norm": 10.413047790527344, + "learning_rate": 2.83436669619132e-07, + "loss": 1.8984, + "step": 33 + }, + { + "epoch": 0.0015053739636717843, + "grad_norm": 4.70713472366333, + "learning_rate": 2.9229406554472986e-07, + "loss": 1.836, + "step": 34 + }, + { + "epoch": 0.0015496496684856602, + "grad_norm": 5.282877445220947, + "learning_rate": 3.011514614703278e-07, + "loss": 1.7752, + "step": 35 + }, + { + "epoch": 0.0015939253732995362, + "grad_norm": 8.02734088897705, + "learning_rate": 3.1000885739592564e-07, + "loss": 2.0172, + "step": 36 + }, + { + "epoch": 0.0016382010781134122, + "grad_norm": 6.51701545715332, + "learning_rate": 3.188662533215235e-07, + "loss": 2.0056, + "step": 37 + }, + { + "epoch": 0.0016824767829272881, + "grad_norm": 5.526393413543701, + "learning_rate": 3.2772364924712136e-07, + "loss": 1.9097, + "step": 38 + }, + { + "epoch": 0.0017267524877411643, + "grad_norm": 4.492867946624756, + "learning_rate": 3.365810451727192e-07, + "loss": 1.8216, + "step": 39 + }, + { + "epoch": 0.0017710281925550403, + "grad_norm": 5.679995059967041, + "learning_rate": 3.4543844109831714e-07, + "loss": 1.7382, + "step": 40 + }, + { + "epoch": 0.0018153038973689163, + "grad_norm": 4.702465057373047, + "learning_rate": 3.5429583702391505e-07, + "loss": 2.0016, + "step": 41 + }, + { + "epoch": 0.0018595796021827922, + "grad_norm": 3.9590137004852295, + "learning_rate": 3.6315323294951286e-07, + "loss": 1.8406, + "step": 42 + }, + { + "epoch": 0.0019038553069966682, + "grad_norm": 4.503845691680908, + "learning_rate": 3.720106288751107e-07, + "loss": 1.905, + "step": 43 + }, + { + "epoch": 0.0019481310118105442, + "grad_norm": 4.042447090148926, + "learning_rate": 3.8086802480070864e-07, + "loss": 1.7914, + "step": 44 + }, + { + "epoch": 0.0019924067166244203, + "grad_norm": 5.795334815979004, + "learning_rate": 3.8972542072630645e-07, + "loss": 1.836, + "step": 45 + }, + { + "epoch": 0.0020366824214382963, + "grad_norm": 5.812465190887451, + "learning_rate": 3.9858281665190436e-07, + "loss": 1.8902, + "step": 46 + }, + { + "epoch": 0.0020809581262521723, + "grad_norm": 3.8073978424072266, + "learning_rate": 4.074402125775023e-07, + "loss": 1.7808, + "step": 47 + }, + { + "epoch": 0.0021252338310660483, + "grad_norm": 3.6301112174987793, + "learning_rate": 4.1629760850310014e-07, + "loss": 1.8305, + "step": 48 + }, + { + "epoch": 0.0021695095358799242, + "grad_norm": 3.518317461013794, + "learning_rate": 4.25155004428698e-07, + "loss": 1.8766, + "step": 49 + }, + { + "epoch": 0.0022137852406938, + "grad_norm": 3.0729589462280273, + "learning_rate": 4.3401240035429586e-07, + "loss": 1.7644, + "step": 50 + }, + { + "epoch": 0.002258060945507676, + "grad_norm": 5.09941291809082, + "learning_rate": 4.428697962798938e-07, + "loss": 1.918, + "step": 51 + }, + { + "epoch": 0.002302336650321552, + "grad_norm": 3.7054154872894287, + "learning_rate": 4.517271922054916e-07, + "loss": 1.9058, + "step": 52 + }, + { + "epoch": 0.002346612355135428, + "grad_norm": 3.6708261966705322, + "learning_rate": 4.605845881310895e-07, + "loss": 1.7776, + "step": 53 + }, + { + "epoch": 0.0023908880599493045, + "grad_norm": 3.4701735973358154, + "learning_rate": 4.6944198405668736e-07, + "loss": 1.802, + "step": 54 + }, + { + "epoch": 0.0024351637647631805, + "grad_norm": 2.664015531539917, + "learning_rate": 4.782993799822853e-07, + "loss": 1.7415, + "step": 55 + }, + { + "epoch": 0.0024794394695770564, + "grad_norm": 3.22680401802063, + "learning_rate": 4.871567759078831e-07, + "loss": 1.8513, + "step": 56 + }, + { + "epoch": 0.0025237151743909324, + "grad_norm": 3.0263519287109375, + "learning_rate": 4.96014171833481e-07, + "loss": 1.728, + "step": 57 + }, + { + "epoch": 0.0025679908792048084, + "grad_norm": 2.849757194519043, + "learning_rate": 5.048715677590789e-07, + "loss": 1.8262, + "step": 58 + }, + { + "epoch": 0.0026122665840186843, + "grad_norm": 3.171067714691162, + "learning_rate": 5.137289636846767e-07, + "loss": 1.9171, + "step": 59 + }, + { + "epoch": 0.0026565422888325603, + "grad_norm": 3.428738832473755, + "learning_rate": 5.225863596102746e-07, + "loss": 1.7141, + "step": 60 + }, + { + "epoch": 0.0027008179936464363, + "grad_norm": 2.636751413345337, + "learning_rate": 5.314437555358726e-07, + "loss": 1.6975, + "step": 61 + }, + { + "epoch": 0.0027450936984603123, + "grad_norm": 3.040304660797119, + "learning_rate": 5.403011514614704e-07, + "loss": 1.8896, + "step": 62 + }, + { + "epoch": 0.0027893694032741882, + "grad_norm": 3.384835958480835, + "learning_rate": 5.491585473870683e-07, + "loss": 1.7874, + "step": 63 + }, + { + "epoch": 0.002833645108088064, + "grad_norm": 2.9490742683410645, + "learning_rate": 5.580159433126661e-07, + "loss": 1.7684, + "step": 64 + }, + { + "epoch": 0.0028779208129019406, + "grad_norm": 2.5564839839935303, + "learning_rate": 5.66873339238264e-07, + "loss": 1.643, + "step": 65 + }, + { + "epoch": 0.0029221965177158166, + "grad_norm": 2.9869959354400635, + "learning_rate": 5.757307351638619e-07, + "loss": 1.7488, + "step": 66 + }, + { + "epoch": 0.0029664722225296925, + "grad_norm": 2.8738367557525635, + "learning_rate": 5.845881310894597e-07, + "loss": 1.7568, + "step": 67 + }, + { + "epoch": 0.0030107479273435685, + "grad_norm": 3.073829412460327, + "learning_rate": 5.934455270150576e-07, + "loss": 1.6703, + "step": 68 + }, + { + "epoch": 0.0030550236321574445, + "grad_norm": 2.5102221965789795, + "learning_rate": 6.023029229406556e-07, + "loss": 1.6635, + "step": 69 + }, + { + "epoch": 0.0030992993369713204, + "grad_norm": 2.1908936500549316, + "learning_rate": 6.111603188662533e-07, + "loss": 1.6612, + "step": 70 + }, + { + "epoch": 0.0031435750417851964, + "grad_norm": 2.707554817199707, + "learning_rate": 6.200177147918513e-07, + "loss": 1.736, + "step": 71 + }, + { + "epoch": 0.0031878507465990724, + "grad_norm": 3.2292115688323975, + "learning_rate": 6.288751107174491e-07, + "loss": 1.8121, + "step": 72 + }, + { + "epoch": 0.0032321264514129484, + "grad_norm": 2.6307883262634277, + "learning_rate": 6.37732506643047e-07, + "loss": 1.7507, + "step": 73 + }, + { + "epoch": 0.0032764021562268243, + "grad_norm": 2.127596855163574, + "learning_rate": 6.46589902568645e-07, + "loss": 1.5793, + "step": 74 + }, + { + "epoch": 0.0033206778610407003, + "grad_norm": 2.3516900539398193, + "learning_rate": 6.554472984942427e-07, + "loss": 1.7814, + "step": 75 + }, + { + "epoch": 0.0033649535658545763, + "grad_norm": 2.1595866680145264, + "learning_rate": 6.643046944198406e-07, + "loss": 1.6249, + "step": 76 + }, + { + "epoch": 0.0034092292706684527, + "grad_norm": 2.367553949356079, + "learning_rate": 6.731620903454384e-07, + "loss": 1.5954, + "step": 77 + }, + { + "epoch": 0.0034535049754823286, + "grad_norm": 1.9875465631484985, + "learning_rate": 6.820194862710364e-07, + "loss": 1.5237, + "step": 78 + }, + { + "epoch": 0.0034977806802962046, + "grad_norm": 2.1725237369537354, + "learning_rate": 6.908768821966343e-07, + "loss": 1.6856, + "step": 79 + }, + { + "epoch": 0.0035420563851100806, + "grad_norm": 1.9388563632965088, + "learning_rate": 6.997342781222321e-07, + "loss": 1.5483, + "step": 80 + }, + { + "epoch": 0.0035863320899239565, + "grad_norm": 2.304990768432617, + "learning_rate": 7.085916740478301e-07, + "loss": 1.7093, + "step": 81 + }, + { + "epoch": 0.0036306077947378325, + "grad_norm": 2.146392345428467, + "learning_rate": 7.174490699734279e-07, + "loss": 1.6652, + "step": 82 + }, + { + "epoch": 0.0036748834995517085, + "grad_norm": 2.051173686981201, + "learning_rate": 7.263064658990257e-07, + "loss": 1.5702, + "step": 83 + }, + { + "epoch": 0.0037191592043655844, + "grad_norm": 2.4940009117126465, + "learning_rate": 7.351638618246236e-07, + "loss": 1.6262, + "step": 84 + }, + { + "epoch": 0.0037634349091794604, + "grad_norm": 1.9089691638946533, + "learning_rate": 7.440212577502214e-07, + "loss": 1.4997, + "step": 85 + }, + { + "epoch": 0.0038077106139933364, + "grad_norm": 2.015230655670166, + "learning_rate": 7.528786536758194e-07, + "loss": 1.7023, + "step": 86 + }, + { + "epoch": 0.0038519863188072124, + "grad_norm": 1.9383435249328613, + "learning_rate": 7.617360496014173e-07, + "loss": 1.4895, + "step": 87 + }, + { + "epoch": 0.0038962620236210883, + "grad_norm": 1.9885592460632324, + "learning_rate": 7.705934455270151e-07, + "loss": 1.5534, + "step": 88 + }, + { + "epoch": 0.003940537728434964, + "grad_norm": 2.3666586875915527, + "learning_rate": 7.794508414526129e-07, + "loss": 1.6839, + "step": 89 + }, + { + "epoch": 0.003984813433248841, + "grad_norm": 3.1030075550079346, + "learning_rate": 7.883082373782109e-07, + "loss": 1.6489, + "step": 90 + }, + { + "epoch": 0.004029089138062716, + "grad_norm": 2.44233775138855, + "learning_rate": 7.971656333038087e-07, + "loss": 1.5223, + "step": 91 + }, + { + "epoch": 0.004073364842876593, + "grad_norm": 1.970958948135376, + "learning_rate": 8.060230292294066e-07, + "loss": 1.7268, + "step": 92 + }, + { + "epoch": 0.004117640547690468, + "grad_norm": 1.7055717706680298, + "learning_rate": 8.148804251550045e-07, + "loss": 1.4681, + "step": 93 + }, + { + "epoch": 0.004161916252504345, + "grad_norm": 1.7330037355422974, + "learning_rate": 8.237378210806024e-07, + "loss": 1.4109, + "step": 94 + }, + { + "epoch": 0.004206191957318221, + "grad_norm": 2.134687662124634, + "learning_rate": 8.325952170062003e-07, + "loss": 1.4986, + "step": 95 + }, + { + "epoch": 0.0042504676621320965, + "grad_norm": 3.1047959327697754, + "learning_rate": 8.41452612931798e-07, + "loss": 1.6944, + "step": 96 + }, + { + "epoch": 0.004294743366945973, + "grad_norm": 2.516368865966797, + "learning_rate": 8.50310008857396e-07, + "loss": 1.6457, + "step": 97 + }, + { + "epoch": 0.0043390190717598484, + "grad_norm": 2.0445547103881836, + "learning_rate": 8.591674047829939e-07, + "loss": 1.5158, + "step": 98 + }, + { + "epoch": 0.004383294776573725, + "grad_norm": 2.1044514179229736, + "learning_rate": 8.680248007085917e-07, + "loss": 1.6338, + "step": 99 + }, + { + "epoch": 0.0044275704813876, + "grad_norm": 1.986251950263977, + "learning_rate": 8.768821966341897e-07, + "loss": 1.5485, + "step": 100 + }, + { + "epoch": 0.004471846186201477, + "grad_norm": 2.0721304416656494, + "learning_rate": 8.857395925597875e-07, + "loss": 1.5988, + "step": 101 + }, + { + "epoch": 0.004516121891015352, + "grad_norm": 2.3371362686157227, + "learning_rate": 8.945969884853854e-07, + "loss": 1.5748, + "step": 102 + }, + { + "epoch": 0.004560397595829229, + "grad_norm": 2.0185513496398926, + "learning_rate": 9.034543844109832e-07, + "loss": 1.6374, + "step": 103 + }, + { + "epoch": 0.004604673300643104, + "grad_norm": 1.8404675722122192, + "learning_rate": 9.12311780336581e-07, + "loss": 1.5269, + "step": 104 + }, + { + "epoch": 0.004648949005456981, + "grad_norm": 1.865485668182373, + "learning_rate": 9.21169176262179e-07, + "loss": 1.4941, + "step": 105 + }, + { + "epoch": 0.004693224710270856, + "grad_norm": 1.8353692293167114, + "learning_rate": 9.300265721877769e-07, + "loss": 1.6593, + "step": 106 + }, + { + "epoch": 0.004737500415084733, + "grad_norm": 4.649820804595947, + "learning_rate": 9.388839681133747e-07, + "loss": 1.521, + "step": 107 + }, + { + "epoch": 0.004781776119898609, + "grad_norm": 1.8854844570159912, + "learning_rate": 9.477413640389727e-07, + "loss": 1.4608, + "step": 108 + }, + { + "epoch": 0.0048260518247124845, + "grad_norm": 2.011930227279663, + "learning_rate": 9.565987599645705e-07, + "loss": 1.613, + "step": 109 + }, + { + "epoch": 0.004870327529526361, + "grad_norm": 1.960777759552002, + "learning_rate": 9.654561558901683e-07, + "loss": 1.5534, + "step": 110 + }, + { + "epoch": 0.0049146032343402365, + "grad_norm": 2.1667397022247314, + "learning_rate": 9.743135518157663e-07, + "loss": 1.4826, + "step": 111 + }, + { + "epoch": 0.004958878939154113, + "grad_norm": 1.7090040445327759, + "learning_rate": 9.83170947741364e-07, + "loss": 1.4583, + "step": 112 + }, + { + "epoch": 0.005003154643967988, + "grad_norm": 2.1987860202789307, + "learning_rate": 9.92028343666962e-07, + "loss": 1.5257, + "step": 113 + }, + { + "epoch": 0.005047430348781865, + "grad_norm": 1.7115312814712524, + "learning_rate": 1.00088573959256e-06, + "loss": 1.4096, + "step": 114 + }, + { + "epoch": 0.00509170605359574, + "grad_norm": 2.5126724243164062, + "learning_rate": 1.0097431355181577e-06, + "loss": 1.6183, + "step": 115 + }, + { + "epoch": 0.005135981758409617, + "grad_norm": 1.9702465534210205, + "learning_rate": 1.0186005314437557e-06, + "loss": 1.5559, + "step": 116 + }, + { + "epoch": 0.005180257463223492, + "grad_norm": 2.086444139480591, + "learning_rate": 1.0274579273693534e-06, + "loss": 1.552, + "step": 117 + }, + { + "epoch": 0.005224533168037369, + "grad_norm": 1.7597230672836304, + "learning_rate": 1.0363153232949514e-06, + "loss": 1.4874, + "step": 118 + }, + { + "epoch": 0.005268808872851245, + "grad_norm": 2.296724557876587, + "learning_rate": 1.0451727192205492e-06, + "loss": 1.4674, + "step": 119 + }, + { + "epoch": 0.005313084577665121, + "grad_norm": 1.7965677976608276, + "learning_rate": 1.0540301151461471e-06, + "loss": 1.4509, + "step": 120 + }, + { + "epoch": 0.005357360282478997, + "grad_norm": 2.0141212940216064, + "learning_rate": 1.062887511071745e-06, + "loss": 1.4478, + "step": 121 + }, + { + "epoch": 0.005401635987292873, + "grad_norm": 2.3701186180114746, + "learning_rate": 1.0717449069973429e-06, + "loss": 1.5568, + "step": 122 + }, + { + "epoch": 0.005445911692106749, + "grad_norm": 1.8361740112304688, + "learning_rate": 1.0806023029229408e-06, + "loss": 1.4655, + "step": 123 + }, + { + "epoch": 0.0054901873969206245, + "grad_norm": 1.7711232900619507, + "learning_rate": 1.0894596988485386e-06, + "loss": 1.3507, + "step": 124 + }, + { + "epoch": 0.005534463101734501, + "grad_norm": 1.8630656003952026, + "learning_rate": 1.0983170947741365e-06, + "loss": 1.4483, + "step": 125 + }, + { + "epoch": 0.0055787388065483764, + "grad_norm": 1.9049400091171265, + "learning_rate": 1.1071744906997343e-06, + "loss": 1.5572, + "step": 126 + }, + { + "epoch": 0.005623014511362253, + "grad_norm": 2.0668928623199463, + "learning_rate": 1.1160318866253323e-06, + "loss": 1.4655, + "step": 127 + }, + { + "epoch": 0.005667290216176128, + "grad_norm": 1.956262230873108, + "learning_rate": 1.1248892825509302e-06, + "loss": 1.4876, + "step": 128 + }, + { + "epoch": 0.005711565920990005, + "grad_norm": 2.0680181980133057, + "learning_rate": 1.133746678476528e-06, + "loss": 1.4956, + "step": 129 + }, + { + "epoch": 0.005755841625803881, + "grad_norm": 1.7142980098724365, + "learning_rate": 1.142604074402126e-06, + "loss": 1.4977, + "step": 130 + }, + { + "epoch": 0.005800117330617757, + "grad_norm": 1.7640386819839478, + "learning_rate": 1.1514614703277237e-06, + "loss": 1.4255, + "step": 131 + }, + { + "epoch": 0.005844393035431633, + "grad_norm": 1.6488368511199951, + "learning_rate": 1.1603188662533215e-06, + "loss": 1.3969, + "step": 132 + }, + { + "epoch": 0.005888668740245509, + "grad_norm": 2.095383882522583, + "learning_rate": 1.1691762621789194e-06, + "loss": 1.521, + "step": 133 + }, + { + "epoch": 0.005932944445059385, + "grad_norm": 1.951103687286377, + "learning_rate": 1.1780336581045174e-06, + "loss": 1.5006, + "step": 134 + }, + { + "epoch": 0.005977220149873261, + "grad_norm": 1.9385473728179932, + "learning_rate": 1.1868910540301152e-06, + "loss": 1.4325, + "step": 135 + }, + { + "epoch": 0.006021495854687137, + "grad_norm": 1.84605073928833, + "learning_rate": 1.1957484499557131e-06, + "loss": 1.4817, + "step": 136 + }, + { + "epoch": 0.0060657715595010125, + "grad_norm": 1.8297159671783447, + "learning_rate": 1.204605845881311e-06, + "loss": 1.4178, + "step": 137 + }, + { + "epoch": 0.006110047264314889, + "grad_norm": 1.8058748245239258, + "learning_rate": 1.2134632418069089e-06, + "loss": 1.4815, + "step": 138 + }, + { + "epoch": 0.0061543229691287645, + "grad_norm": 1.7101901769638062, + "learning_rate": 1.2223206377325066e-06, + "loss": 1.4018, + "step": 139 + }, + { + "epoch": 0.006198598673942641, + "grad_norm": 2.053083658218384, + "learning_rate": 1.2311780336581046e-06, + "loss": 1.4321, + "step": 140 + }, + { + "epoch": 0.006242874378756516, + "grad_norm": 1.8737056255340576, + "learning_rate": 1.2400354295837025e-06, + "loss": 1.3486, + "step": 141 + }, + { + "epoch": 0.006287150083570393, + "grad_norm": 2.2152199745178223, + "learning_rate": 1.2488928255093003e-06, + "loss": 1.391, + "step": 142 + }, + { + "epoch": 0.006331425788384269, + "grad_norm": 1.8124629259109497, + "learning_rate": 1.2577502214348983e-06, + "loss": 1.3724, + "step": 143 + }, + { + "epoch": 0.006375701493198145, + "grad_norm": 1.904874324798584, + "learning_rate": 1.266607617360496e-06, + "loss": 1.443, + "step": 144 + }, + { + "epoch": 0.006419977198012021, + "grad_norm": 2.712522029876709, + "learning_rate": 1.275465013286094e-06, + "loss": 1.5496, + "step": 145 + }, + { + "epoch": 0.006464252902825897, + "grad_norm": 1.963131308555603, + "learning_rate": 1.2843224092116918e-06, + "loss": 1.4229, + "step": 146 + }, + { + "epoch": 0.006508528607639773, + "grad_norm": 1.775274395942688, + "learning_rate": 1.29317980513729e-06, + "loss": 1.2448, + "step": 147 + }, + { + "epoch": 0.006552804312453649, + "grad_norm": 2.190737009048462, + "learning_rate": 1.3020372010628877e-06, + "loss": 1.4834, + "step": 148 + }, + { + "epoch": 0.006597080017267525, + "grad_norm": 2.0866715908050537, + "learning_rate": 1.3108945969884854e-06, + "loss": 1.3485, + "step": 149 + }, + { + "epoch": 0.006641355722081401, + "grad_norm": 1.8039039373397827, + "learning_rate": 1.3197519929140834e-06, + "loss": 1.3303, + "step": 150 + }, + { + "epoch": 0.006685631426895277, + "grad_norm": 1.8807085752487183, + "learning_rate": 1.3286093888396812e-06, + "loss": 1.5006, + "step": 151 + }, + { + "epoch": 0.0067299071317091525, + "grad_norm": 1.737419843673706, + "learning_rate": 1.3374667847652791e-06, + "loss": 1.3349, + "step": 152 + }, + { + "epoch": 0.006774182836523029, + "grad_norm": 1.8055393695831299, + "learning_rate": 1.3463241806908769e-06, + "loss": 1.496, + "step": 153 + }, + { + "epoch": 0.006818458541336905, + "grad_norm": 2.0880355834960938, + "learning_rate": 1.355181576616475e-06, + "loss": 1.4904, + "step": 154 + }, + { + "epoch": 0.006862734246150781, + "grad_norm": 2.224071979522705, + "learning_rate": 1.3640389725420728e-06, + "loss": 1.5435, + "step": 155 + }, + { + "epoch": 0.006907009950964657, + "grad_norm": 1.9626963138580322, + "learning_rate": 1.3728963684676706e-06, + "loss": 1.4609, + "step": 156 + }, + { + "epoch": 0.006951285655778533, + "grad_norm": 1.733473300933838, + "learning_rate": 1.3817537643932685e-06, + "loss": 1.3406, + "step": 157 + }, + { + "epoch": 0.006995561360592409, + "grad_norm": 1.6159511804580688, + "learning_rate": 1.3906111603188663e-06, + "loss": 1.3234, + "step": 158 + }, + { + "epoch": 0.007039837065406285, + "grad_norm": 1.865597128868103, + "learning_rate": 1.3994685562444643e-06, + "loss": 1.4272, + "step": 159 + }, + { + "epoch": 0.007084112770220161, + "grad_norm": 2.398378372192383, + "learning_rate": 1.408325952170062e-06, + "loss": 1.4636, + "step": 160 + }, + { + "epoch": 0.007128388475034037, + "grad_norm": 1.992119550704956, + "learning_rate": 1.4171833480956602e-06, + "loss": 1.3702, + "step": 161 + }, + { + "epoch": 0.007172664179847913, + "grad_norm": 2.2376556396484375, + "learning_rate": 1.426040744021258e-06, + "loss": 1.4901, + "step": 162 + }, + { + "epoch": 0.007216939884661789, + "grad_norm": 2.062776565551758, + "learning_rate": 1.4348981399468557e-06, + "loss": 1.3847, + "step": 163 + }, + { + "epoch": 0.007261215589475665, + "grad_norm": 1.8541206121444702, + "learning_rate": 1.4437555358724537e-06, + "loss": 1.4841, + "step": 164 + }, + { + "epoch": 0.0073054912942895405, + "grad_norm": 2.1659841537475586, + "learning_rate": 1.4526129317980514e-06, + "loss": 1.4147, + "step": 165 + }, + { + "epoch": 0.007349766999103417, + "grad_norm": 1.837306022644043, + "learning_rate": 1.4614703277236494e-06, + "loss": 1.4882, + "step": 166 + }, + { + "epoch": 0.007394042703917293, + "grad_norm": 1.6987314224243164, + "learning_rate": 1.4703277236492472e-06, + "loss": 1.3063, + "step": 167 + }, + { + "epoch": 0.007438318408731169, + "grad_norm": 1.7928156852722168, + "learning_rate": 1.4791851195748451e-06, + "loss": 1.3506, + "step": 168 + }, + { + "epoch": 0.007482594113545045, + "grad_norm": 1.8114415407180786, + "learning_rate": 1.4880425155004429e-06, + "loss": 1.4466, + "step": 169 + }, + { + "epoch": 0.007526869818358921, + "grad_norm": 2.122408390045166, + "learning_rate": 1.4968999114260406e-06, + "loss": 1.5224, + "step": 170 + }, + { + "epoch": 0.007571145523172797, + "grad_norm": 1.8953779935836792, + "learning_rate": 1.5057573073516388e-06, + "loss": 1.3424, + "step": 171 + }, + { + "epoch": 0.007615421227986673, + "grad_norm": 1.897600769996643, + "learning_rate": 1.5146147032772366e-06, + "loss": 1.4314, + "step": 172 + }, + { + "epoch": 0.007659696932800549, + "grad_norm": 1.791122317314148, + "learning_rate": 1.5234720992028345e-06, + "loss": 1.3598, + "step": 173 + }, + { + "epoch": 0.007703972637614425, + "grad_norm": 1.6739681959152222, + "learning_rate": 1.5323294951284323e-06, + "loss": 1.3726, + "step": 174 + }, + { + "epoch": 0.007748248342428301, + "grad_norm": 3.0500411987304688, + "learning_rate": 1.5411868910540303e-06, + "loss": 1.5203, + "step": 175 + }, + { + "epoch": 0.007792524047242177, + "grad_norm": 1.6860638856887817, + "learning_rate": 1.550044286979628e-06, + "loss": 1.3269, + "step": 176 + }, + { + "epoch": 0.007836799752056053, + "grad_norm": 1.6990995407104492, + "learning_rate": 1.5589016829052258e-06, + "loss": 1.3325, + "step": 177 + }, + { + "epoch": 0.007881075456869929, + "grad_norm": 1.6639938354492188, + "learning_rate": 1.567759078830824e-06, + "loss": 1.3625, + "step": 178 + }, + { + "epoch": 0.007925351161683806, + "grad_norm": 1.6999707221984863, + "learning_rate": 1.5766164747564217e-06, + "loss": 1.2871, + "step": 179 + }, + { + "epoch": 0.007969626866497681, + "grad_norm": 1.8094635009765625, + "learning_rate": 1.5854738706820197e-06, + "loss": 1.3009, + "step": 180 + }, + { + "epoch": 0.008013902571311557, + "grad_norm": 1.8083000183105469, + "learning_rate": 1.5943312666076174e-06, + "loss": 1.3188, + "step": 181 + }, + { + "epoch": 0.008058178276125432, + "grad_norm": 1.5570943355560303, + "learning_rate": 1.6031886625332154e-06, + "loss": 1.2526, + "step": 182 + }, + { + "epoch": 0.00810245398093931, + "grad_norm": 1.8183670043945312, + "learning_rate": 1.6120460584588132e-06, + "loss": 1.4077, + "step": 183 + }, + { + "epoch": 0.008146729685753185, + "grad_norm": 1.7130922079086304, + "learning_rate": 1.620903454384411e-06, + "loss": 1.3558, + "step": 184 + }, + { + "epoch": 0.00819100539056706, + "grad_norm": 2.0152699947357178, + "learning_rate": 1.629760850310009e-06, + "loss": 1.3923, + "step": 185 + }, + { + "epoch": 0.008235281095380936, + "grad_norm": 2.2459394931793213, + "learning_rate": 1.6386182462356069e-06, + "loss": 1.475, + "step": 186 + }, + { + "epoch": 0.008279556800194814, + "grad_norm": 1.6324098110198975, + "learning_rate": 1.6474756421612048e-06, + "loss": 1.3078, + "step": 187 + }, + { + "epoch": 0.00832383250500869, + "grad_norm": 2.0339138507843018, + "learning_rate": 1.6563330380868026e-06, + "loss": 1.3509, + "step": 188 + }, + { + "epoch": 0.008368108209822565, + "grad_norm": 1.7830489873886108, + "learning_rate": 1.6651904340124005e-06, + "loss": 1.3424, + "step": 189 + }, + { + "epoch": 0.008412383914636442, + "grad_norm": 2.172374725341797, + "learning_rate": 1.6740478299379983e-06, + "loss": 1.3975, + "step": 190 + }, + { + "epoch": 0.008456659619450317, + "grad_norm": 1.7277922630310059, + "learning_rate": 1.682905225863596e-06, + "loss": 1.2614, + "step": 191 + }, + { + "epoch": 0.008500935324264193, + "grad_norm": 2.02286434173584, + "learning_rate": 1.6917626217891942e-06, + "loss": 1.4299, + "step": 192 + }, + { + "epoch": 0.008545211029078069, + "grad_norm": 1.6673126220703125, + "learning_rate": 1.700620017714792e-06, + "loss": 1.2524, + "step": 193 + }, + { + "epoch": 0.008589486733891946, + "grad_norm": 2.1014955043792725, + "learning_rate": 1.70947741364039e-06, + "loss": 1.3524, + "step": 194 + }, + { + "epoch": 0.008633762438705821, + "grad_norm": 1.7165584564208984, + "learning_rate": 1.7183348095659877e-06, + "loss": 1.3463, + "step": 195 + }, + { + "epoch": 0.008678038143519697, + "grad_norm": 1.730338454246521, + "learning_rate": 1.7271922054915857e-06, + "loss": 1.3046, + "step": 196 + }, + { + "epoch": 0.008722313848333572, + "grad_norm": 1.634653091430664, + "learning_rate": 1.7360496014171834e-06, + "loss": 1.2716, + "step": 197 + }, + { + "epoch": 0.00876658955314745, + "grad_norm": 1.7185152769088745, + "learning_rate": 1.7449069973427812e-06, + "loss": 1.1922, + "step": 198 + }, + { + "epoch": 0.008810865257961325, + "grad_norm": 1.8386950492858887, + "learning_rate": 1.7537643932683794e-06, + "loss": 1.3634, + "step": 199 + }, + { + "epoch": 0.0088551409627752, + "grad_norm": 1.7364953756332397, + "learning_rate": 1.7626217891939771e-06, + "loss": 1.3572, + "step": 200 + }, + { + "epoch": 0.008899416667589078, + "grad_norm": 1.9376144409179688, + "learning_rate": 1.771479185119575e-06, + "loss": 1.3435, + "step": 201 + }, + { + "epoch": 0.008943692372402954, + "grad_norm": 2.089015245437622, + "learning_rate": 1.7803365810451729e-06, + "loss": 1.4114, + "step": 202 + }, + { + "epoch": 0.008987968077216829, + "grad_norm": 1.7104213237762451, + "learning_rate": 1.7891939769707708e-06, + "loss": 1.3739, + "step": 203 + }, + { + "epoch": 0.009032243782030705, + "grad_norm": 1.6401504278182983, + "learning_rate": 1.7980513728963686e-06, + "loss": 1.262, + "step": 204 + }, + { + "epoch": 0.009076519486844582, + "grad_norm": 1.6457736492156982, + "learning_rate": 1.8069087688219663e-06, + "loss": 1.3087, + "step": 205 + }, + { + "epoch": 0.009120795191658457, + "grad_norm": 2.0442795753479004, + "learning_rate": 1.8157661647475643e-06, + "loss": 1.4646, + "step": 206 + }, + { + "epoch": 0.009165070896472333, + "grad_norm": 1.9811749458312988, + "learning_rate": 1.824623560673162e-06, + "loss": 1.3774, + "step": 207 + }, + { + "epoch": 0.009209346601286209, + "grad_norm": 1.6526434421539307, + "learning_rate": 1.8334809565987602e-06, + "loss": 1.3144, + "step": 208 + }, + { + "epoch": 0.009253622306100086, + "grad_norm": 2.131859064102173, + "learning_rate": 1.842338352524358e-06, + "loss": 1.4272, + "step": 209 + }, + { + "epoch": 0.009297898010913961, + "grad_norm": 1.8755522966384888, + "learning_rate": 1.851195748449956e-06, + "loss": 1.3175, + "step": 210 + }, + { + "epoch": 0.009342173715727837, + "grad_norm": 1.8867839574813843, + "learning_rate": 1.8600531443755537e-06, + "loss": 1.3493, + "step": 211 + }, + { + "epoch": 0.009386449420541712, + "grad_norm": 1.7255513668060303, + "learning_rate": 1.8689105403011515e-06, + "loss": 1.3646, + "step": 212 + }, + { + "epoch": 0.00943072512535559, + "grad_norm": 1.7379560470581055, + "learning_rate": 1.8777679362267494e-06, + "loss": 1.2939, + "step": 213 + }, + { + "epoch": 0.009475000830169465, + "grad_norm": 5.061211585998535, + "learning_rate": 1.8866253321523472e-06, + "loss": 1.2371, + "step": 214 + }, + { + "epoch": 0.00951927653498334, + "grad_norm": 1.6560953855514526, + "learning_rate": 1.8954827280779454e-06, + "loss": 1.3247, + "step": 215 + }, + { + "epoch": 0.009563552239797218, + "grad_norm": 1.9370836019515991, + "learning_rate": 1.9043401240035431e-06, + "loss": 1.337, + "step": 216 + }, + { + "epoch": 0.009607827944611094, + "grad_norm": 1.8624497652053833, + "learning_rate": 1.913197519929141e-06, + "loss": 1.3978, + "step": 217 + }, + { + "epoch": 0.009652103649424969, + "grad_norm": 2.7053046226501465, + "learning_rate": 1.922054915854739e-06, + "loss": 1.4553, + "step": 218 + }, + { + "epoch": 0.009696379354238845, + "grad_norm": 1.7161411046981812, + "learning_rate": 1.9309123117803366e-06, + "loss": 1.339, + "step": 219 + }, + { + "epoch": 0.009740655059052722, + "grad_norm": 1.7797871828079224, + "learning_rate": 1.9397697077059346e-06, + "loss": 1.3571, + "step": 220 + }, + { + "epoch": 0.009784930763866597, + "grad_norm": 1.9508473873138428, + "learning_rate": 1.9486271036315325e-06, + "loss": 1.4015, + "step": 221 + }, + { + "epoch": 0.009829206468680473, + "grad_norm": 2.084582567214966, + "learning_rate": 1.9574844995571305e-06, + "loss": 1.3766, + "step": 222 + }, + { + "epoch": 0.009873482173494348, + "grad_norm": 1.9474711418151855, + "learning_rate": 1.966341895482728e-06, + "loss": 1.4086, + "step": 223 + }, + { + "epoch": 0.009917757878308226, + "grad_norm": 1.84169340133667, + "learning_rate": 1.975199291408326e-06, + "loss": 1.3156, + "step": 224 + }, + { + "epoch": 0.009962033583122101, + "grad_norm": 1.7759990692138672, + "learning_rate": 1.984056687333924e-06, + "loss": 1.3944, + "step": 225 + }, + { + "epoch": 0.010006309287935977, + "grad_norm": 2.1380062103271484, + "learning_rate": 1.9929140832595215e-06, + "loss": 1.3012, + "step": 226 + }, + { + "epoch": 0.010050584992749854, + "grad_norm": 2.1394989490509033, + "learning_rate": 2.00177147918512e-06, + "loss": 1.4705, + "step": 227 + }, + { + "epoch": 0.01009486069756373, + "grad_norm": 1.9282633066177368, + "learning_rate": 2.0106288751107175e-06, + "loss": 1.3216, + "step": 228 + }, + { + "epoch": 0.010139136402377605, + "grad_norm": 1.817718744277954, + "learning_rate": 2.0194862710363154e-06, + "loss": 1.3032, + "step": 229 + }, + { + "epoch": 0.01018341210719148, + "grad_norm": 2.137608528137207, + "learning_rate": 2.0283436669619134e-06, + "loss": 1.3219, + "step": 230 + }, + { + "epoch": 0.010227687812005358, + "grad_norm": 2.0383975505828857, + "learning_rate": 2.0372010628875114e-06, + "loss": 1.4048, + "step": 231 + }, + { + "epoch": 0.010271963516819234, + "grad_norm": 1.8421354293823242, + "learning_rate": 2.046058458813109e-06, + "loss": 1.2868, + "step": 232 + }, + { + "epoch": 0.010316239221633109, + "grad_norm": 2.066659688949585, + "learning_rate": 2.054915854738707e-06, + "loss": 1.4015, + "step": 233 + }, + { + "epoch": 0.010360514926446985, + "grad_norm": 1.8011773824691772, + "learning_rate": 2.063773250664305e-06, + "loss": 1.3243, + "step": 234 + }, + { + "epoch": 0.010404790631260862, + "grad_norm": 2.3771984577178955, + "learning_rate": 2.072630646589903e-06, + "loss": 1.2778, + "step": 235 + }, + { + "epoch": 0.010449066336074737, + "grad_norm": 2.3346502780914307, + "learning_rate": 2.0814880425155008e-06, + "loss": 1.3865, + "step": 236 + }, + { + "epoch": 0.010493342040888613, + "grad_norm": 1.886533498764038, + "learning_rate": 2.0903454384410983e-06, + "loss": 1.337, + "step": 237 + }, + { + "epoch": 0.01053761774570249, + "grad_norm": 1.6266735792160034, + "learning_rate": 2.0992028343666963e-06, + "loss": 1.2521, + "step": 238 + }, + { + "epoch": 0.010581893450516366, + "grad_norm": 1.768676996231079, + "learning_rate": 2.1080602302922943e-06, + "loss": 1.3646, + "step": 239 + }, + { + "epoch": 0.010626169155330241, + "grad_norm": 2.3775205612182617, + "learning_rate": 2.116917626217892e-06, + "loss": 1.4485, + "step": 240 + }, + { + "epoch": 0.010670444860144117, + "grad_norm": 1.7110388278961182, + "learning_rate": 2.12577502214349e-06, + "loss": 1.2726, + "step": 241 + }, + { + "epoch": 0.010714720564957994, + "grad_norm": 1.8977283239364624, + "learning_rate": 2.1346324180690877e-06, + "loss": 1.3397, + "step": 242 + }, + { + "epoch": 0.01075899626977187, + "grad_norm": 1.9395962953567505, + "learning_rate": 2.1434898139946857e-06, + "loss": 1.2789, + "step": 243 + }, + { + "epoch": 0.010803271974585745, + "grad_norm": 1.4975390434265137, + "learning_rate": 2.1523472099202837e-06, + "loss": 1.1982, + "step": 244 + }, + { + "epoch": 0.01084754767939962, + "grad_norm": 1.6565436124801636, + "learning_rate": 2.1612046058458816e-06, + "loss": 1.3126, + "step": 245 + }, + { + "epoch": 0.010891823384213498, + "grad_norm": 1.7650989294052124, + "learning_rate": 2.170062001771479e-06, + "loss": 1.335, + "step": 246 + }, + { + "epoch": 0.010936099089027373, + "grad_norm": 2.700181484222412, + "learning_rate": 2.178919397697077e-06, + "loss": 1.3825, + "step": 247 + }, + { + "epoch": 0.010980374793841249, + "grad_norm": 2.0590457916259766, + "learning_rate": 2.187776793622675e-06, + "loss": 1.3226, + "step": 248 + }, + { + "epoch": 0.011024650498655126, + "grad_norm": 2.052156925201416, + "learning_rate": 2.196634189548273e-06, + "loss": 1.3429, + "step": 249 + }, + { + "epoch": 0.011068926203469002, + "grad_norm": 1.8762898445129395, + "learning_rate": 2.205491585473871e-06, + "loss": 1.3075, + "step": 250 + }, + { + "epoch": 0.011113201908282877, + "grad_norm": 1.7109332084655762, + "learning_rate": 2.2143489813994686e-06, + "loss": 1.3558, + "step": 251 + }, + { + "epoch": 0.011157477613096753, + "grad_norm": 1.6541163921356201, + "learning_rate": 2.2232063773250666e-06, + "loss": 1.3458, + "step": 252 + }, + { + "epoch": 0.01120175331791063, + "grad_norm": 1.8118610382080078, + "learning_rate": 2.2320637732506645e-06, + "loss": 1.2976, + "step": 253 + }, + { + "epoch": 0.011246029022724506, + "grad_norm": 1.7733922004699707, + "learning_rate": 2.240921169176262e-06, + "loss": 1.3189, + "step": 254 + }, + { + "epoch": 0.011290304727538381, + "grad_norm": 1.6504555940628052, + "learning_rate": 2.2497785651018605e-06, + "loss": 1.2774, + "step": 255 + }, + { + "epoch": 0.011334580432352257, + "grad_norm": 1.809150218963623, + "learning_rate": 2.258635961027458e-06, + "loss": 1.3228, + "step": 256 + }, + { + "epoch": 0.011378856137166134, + "grad_norm": 1.714198112487793, + "learning_rate": 2.267493356953056e-06, + "loss": 1.1939, + "step": 257 + }, + { + "epoch": 0.01142313184198001, + "grad_norm": 1.6355242729187012, + "learning_rate": 2.276350752878654e-06, + "loss": 1.2553, + "step": 258 + }, + { + "epoch": 0.011467407546793885, + "grad_norm": 1.7357940673828125, + "learning_rate": 2.285208148804252e-06, + "loss": 1.2582, + "step": 259 + }, + { + "epoch": 0.011511683251607762, + "grad_norm": 1.6742192506790161, + "learning_rate": 2.2940655447298495e-06, + "loss": 1.2625, + "step": 260 + }, + { + "epoch": 0.011555958956421638, + "grad_norm": 2.0788421630859375, + "learning_rate": 2.3029229406554474e-06, + "loss": 1.4315, + "step": 261 + }, + { + "epoch": 0.011600234661235513, + "grad_norm": 1.9495068788528442, + "learning_rate": 2.3117803365810454e-06, + "loss": 1.2593, + "step": 262 + }, + { + "epoch": 0.011644510366049389, + "grad_norm": 1.8597300052642822, + "learning_rate": 2.320637732506643e-06, + "loss": 1.2947, + "step": 263 + }, + { + "epoch": 0.011688786070863266, + "grad_norm": 2.3137404918670654, + "learning_rate": 2.3294951284322413e-06, + "loss": 1.3343, + "step": 264 + }, + { + "epoch": 0.011733061775677142, + "grad_norm": 2.2233238220214844, + "learning_rate": 2.338352524357839e-06, + "loss": 1.3956, + "step": 265 + }, + { + "epoch": 0.011777337480491017, + "grad_norm": 1.6046983003616333, + "learning_rate": 2.347209920283437e-06, + "loss": 1.2672, + "step": 266 + }, + { + "epoch": 0.011821613185304893, + "grad_norm": 1.7691748142242432, + "learning_rate": 2.356067316209035e-06, + "loss": 1.3285, + "step": 267 + }, + { + "epoch": 0.01186588889011877, + "grad_norm": 2.061913251876831, + "learning_rate": 2.3649247121346324e-06, + "loss": 1.2608, + "step": 268 + }, + { + "epoch": 0.011910164594932646, + "grad_norm": 1.8773752450942993, + "learning_rate": 2.3737821080602303e-06, + "loss": 1.3336, + "step": 269 + }, + { + "epoch": 0.011954440299746521, + "grad_norm": 1.8406442403793335, + "learning_rate": 2.3826395039858283e-06, + "loss": 1.2969, + "step": 270 + }, + { + "epoch": 0.011998716004560397, + "grad_norm": 1.7174017429351807, + "learning_rate": 2.3914968999114263e-06, + "loss": 1.2895, + "step": 271 + }, + { + "epoch": 0.012042991709374274, + "grad_norm": 1.7385094165802002, + "learning_rate": 2.4003542958370242e-06, + "loss": 1.2147, + "step": 272 + }, + { + "epoch": 0.01208726741418815, + "grad_norm": 1.5932230949401855, + "learning_rate": 2.409211691762622e-06, + "loss": 1.2493, + "step": 273 + }, + { + "epoch": 0.012131543119002025, + "grad_norm": 1.5214539766311646, + "learning_rate": 2.4180690876882197e-06, + "loss": 1.2472, + "step": 274 + }, + { + "epoch": 0.012175818823815902, + "grad_norm": 1.886339783668518, + "learning_rate": 2.4269264836138177e-06, + "loss": 1.3094, + "step": 275 + }, + { + "epoch": 0.012220094528629778, + "grad_norm": 1.6930841207504272, + "learning_rate": 2.4357838795394157e-06, + "loss": 1.2477, + "step": 276 + }, + { + "epoch": 0.012264370233443653, + "grad_norm": 1.8042858839035034, + "learning_rate": 2.4446412754650132e-06, + "loss": 1.3022, + "step": 277 + }, + { + "epoch": 0.012308645938257529, + "grad_norm": 1.6968703269958496, + "learning_rate": 2.4534986713906116e-06, + "loss": 1.2593, + "step": 278 + }, + { + "epoch": 0.012352921643071406, + "grad_norm": 2.2480525970458984, + "learning_rate": 2.462356067316209e-06, + "loss": 1.3225, + "step": 279 + }, + { + "epoch": 0.012397197347885282, + "grad_norm": 1.6089892387390137, + "learning_rate": 2.471213463241807e-06, + "loss": 1.2118, + "step": 280 + }, + { + "epoch": 0.012441473052699157, + "grad_norm": 2.8437204360961914, + "learning_rate": 2.480070859167405e-06, + "loss": 1.2626, + "step": 281 + }, + { + "epoch": 0.012485748757513033, + "grad_norm": 1.8030571937561035, + "learning_rate": 2.4889282550930026e-06, + "loss": 1.3426, + "step": 282 + }, + { + "epoch": 0.01253002446232691, + "grad_norm": 1.7981483936309814, + "learning_rate": 2.4977856510186006e-06, + "loss": 1.2507, + "step": 283 + }, + { + "epoch": 0.012574300167140786, + "grad_norm": 1.7347681522369385, + "learning_rate": 2.506643046944199e-06, + "loss": 1.2353, + "step": 284 + }, + { + "epoch": 0.012618575871954661, + "grad_norm": 1.6694273948669434, + "learning_rate": 2.5155004428697965e-06, + "loss": 1.2126, + "step": 285 + }, + { + "epoch": 0.012662851576768538, + "grad_norm": 1.7814375162124634, + "learning_rate": 2.5243578387953945e-06, + "loss": 1.3035, + "step": 286 + }, + { + "epoch": 0.012707127281582414, + "grad_norm": 1.6025785207748413, + "learning_rate": 2.533215234720992e-06, + "loss": 1.2886, + "step": 287 + }, + { + "epoch": 0.01275140298639629, + "grad_norm": 1.890223503112793, + "learning_rate": 2.5420726306465904e-06, + "loss": 1.3593, + "step": 288 + }, + { + "epoch": 0.012795678691210165, + "grad_norm": 1.9576362371444702, + "learning_rate": 2.550930026572188e-06, + "loss": 1.2676, + "step": 289 + }, + { + "epoch": 0.012839954396024042, + "grad_norm": 1.705529808998108, + "learning_rate": 2.559787422497786e-06, + "loss": 1.2335, + "step": 290 + }, + { + "epoch": 0.012884230100837918, + "grad_norm": 1.986191987991333, + "learning_rate": 2.5686448184233835e-06, + "loss": 1.3168, + "step": 291 + }, + { + "epoch": 0.012928505805651793, + "grad_norm": 2.117417335510254, + "learning_rate": 2.5775022143489815e-06, + "loss": 1.3843, + "step": 292 + }, + { + "epoch": 0.012972781510465669, + "grad_norm": 1.8558191061019897, + "learning_rate": 2.58635961027458e-06, + "loss": 1.2385, + "step": 293 + }, + { + "epoch": 0.013017057215279546, + "grad_norm": 1.867425799369812, + "learning_rate": 2.5952170062001774e-06, + "loss": 1.3516, + "step": 294 + }, + { + "epoch": 0.013061332920093422, + "grad_norm": 1.6811012029647827, + "learning_rate": 2.6040744021257754e-06, + "loss": 1.2773, + "step": 295 + }, + { + "epoch": 0.013105608624907297, + "grad_norm": 1.9592818021774292, + "learning_rate": 2.612931798051373e-06, + "loss": 1.3415, + "step": 296 + }, + { + "epoch": 0.013149884329721175, + "grad_norm": 1.6517729759216309, + "learning_rate": 2.621789193976971e-06, + "loss": 1.2686, + "step": 297 + }, + { + "epoch": 0.01319416003453505, + "grad_norm": 1.8636512756347656, + "learning_rate": 2.630646589902569e-06, + "loss": 1.2941, + "step": 298 + }, + { + "epoch": 0.013238435739348926, + "grad_norm": 1.6332558393478394, + "learning_rate": 2.639503985828167e-06, + "loss": 1.2655, + "step": 299 + }, + { + "epoch": 0.013282711444162801, + "grad_norm": 1.6294726133346558, + "learning_rate": 2.6483613817537644e-06, + "loss": 1.1524, + "step": 300 + }, + { + "epoch": 0.013326987148976678, + "grad_norm": 1.5458155870437622, + "learning_rate": 2.6572187776793623e-06, + "loss": 1.1847, + "step": 301 + }, + { + "epoch": 0.013371262853790554, + "grad_norm": 2.6483378410339355, + "learning_rate": 2.6660761736049607e-06, + "loss": 1.3949, + "step": 302 + }, + { + "epoch": 0.01341553855860443, + "grad_norm": 1.973016381263733, + "learning_rate": 2.6749335695305583e-06, + "loss": 1.3797, + "step": 303 + }, + { + "epoch": 0.013459814263418305, + "grad_norm": 1.9744644165039062, + "learning_rate": 2.6837909654561562e-06, + "loss": 1.3192, + "step": 304 + }, + { + "epoch": 0.013504089968232182, + "grad_norm": 2.0532398223876953, + "learning_rate": 2.6926483613817538e-06, + "loss": 1.3066, + "step": 305 + }, + { + "epoch": 0.013548365673046058, + "grad_norm": 1.6579891443252563, + "learning_rate": 2.7015057573073517e-06, + "loss": 1.1986, + "step": 306 + }, + { + "epoch": 0.013592641377859933, + "grad_norm": 1.7669587135314941, + "learning_rate": 2.71036315323295e-06, + "loss": 1.2444, + "step": 307 + }, + { + "epoch": 0.01363691708267381, + "grad_norm": 1.7220371961593628, + "learning_rate": 2.7192205491585477e-06, + "loss": 1.2839, + "step": 308 + }, + { + "epoch": 0.013681192787487686, + "grad_norm": 1.689046859741211, + "learning_rate": 2.7280779450841456e-06, + "loss": 1.1837, + "step": 309 + }, + { + "epoch": 0.013725468492301562, + "grad_norm": 2.1506967544555664, + "learning_rate": 2.736935341009743e-06, + "loss": 1.2346, + "step": 310 + }, + { + "epoch": 0.013769744197115437, + "grad_norm": 2.099219799041748, + "learning_rate": 2.745792736935341e-06, + "loss": 1.2899, + "step": 311 + }, + { + "epoch": 0.013814019901929315, + "grad_norm": 1.5040955543518066, + "learning_rate": 2.754650132860939e-06, + "loss": 1.2142, + "step": 312 + }, + { + "epoch": 0.01385829560674319, + "grad_norm": 1.679699182510376, + "learning_rate": 2.763507528786537e-06, + "loss": 1.2422, + "step": 313 + }, + { + "epoch": 0.013902571311557066, + "grad_norm": 1.9458898305892944, + "learning_rate": 2.7723649247121346e-06, + "loss": 1.2732, + "step": 314 + }, + { + "epoch": 0.013946847016370941, + "grad_norm": 1.6577483415603638, + "learning_rate": 2.7812223206377326e-06, + "loss": 1.2787, + "step": 315 + }, + { + "epoch": 0.013991122721184818, + "grad_norm": 2.1221044063568115, + "learning_rate": 2.790079716563331e-06, + "loss": 1.3872, + "step": 316 + }, + { + "epoch": 0.014035398425998694, + "grad_norm": 1.6810053586959839, + "learning_rate": 2.7989371124889285e-06, + "loss": 1.248, + "step": 317 + }, + { + "epoch": 0.01407967413081257, + "grad_norm": 1.6435317993164062, + "learning_rate": 2.8077945084145265e-06, + "loss": 1.2187, + "step": 318 + }, + { + "epoch": 0.014123949835626447, + "grad_norm": 1.9744353294372559, + "learning_rate": 2.816651904340124e-06, + "loss": 1.2146, + "step": 319 + }, + { + "epoch": 0.014168225540440322, + "grad_norm": 1.6801776885986328, + "learning_rate": 2.825509300265722e-06, + "loss": 1.1643, + "step": 320 + }, + { + "epoch": 0.014212501245254198, + "grad_norm": 1.7196154594421387, + "learning_rate": 2.8343666961913204e-06, + "loss": 1.2104, + "step": 321 + }, + { + "epoch": 0.014256776950068073, + "grad_norm": 1.8392728567123413, + "learning_rate": 2.843224092116918e-06, + "loss": 1.2923, + "step": 322 + }, + { + "epoch": 0.01430105265488195, + "grad_norm": 1.7176883220672607, + "learning_rate": 2.852081488042516e-06, + "loss": 1.201, + "step": 323 + }, + { + "epoch": 0.014345328359695826, + "grad_norm": 1.4937782287597656, + "learning_rate": 2.8609388839681135e-06, + "loss": 1.1013, + "step": 324 + }, + { + "epoch": 0.014389604064509702, + "grad_norm": 1.9761731624603271, + "learning_rate": 2.8697962798937114e-06, + "loss": 1.2455, + "step": 325 + }, + { + "epoch": 0.014433879769323577, + "grad_norm": 1.9415900707244873, + "learning_rate": 2.8786536758193094e-06, + "loss": 1.3067, + "step": 326 + }, + { + "epoch": 0.014478155474137454, + "grad_norm": 1.6141997575759888, + "learning_rate": 2.8875110717449074e-06, + "loss": 1.1287, + "step": 327 + }, + { + "epoch": 0.01452243117895133, + "grad_norm": 1.8949164152145386, + "learning_rate": 2.896368467670505e-06, + "loss": 1.258, + "step": 328 + }, + { + "epoch": 0.014566706883765206, + "grad_norm": 1.9223968982696533, + "learning_rate": 2.905225863596103e-06, + "loss": 1.2674, + "step": 329 + }, + { + "epoch": 0.014610982588579081, + "grad_norm": 1.8842408657073975, + "learning_rate": 2.9140832595217013e-06, + "loss": 1.2788, + "step": 330 + }, + { + "epoch": 0.014655258293392958, + "grad_norm": 2.0673952102661133, + "learning_rate": 2.922940655447299e-06, + "loss": 1.1881, + "step": 331 + }, + { + "epoch": 0.014699533998206834, + "grad_norm": 1.737990140914917, + "learning_rate": 2.9317980513728968e-06, + "loss": 1.2765, + "step": 332 + }, + { + "epoch": 0.01474380970302071, + "grad_norm": 1.7584950923919678, + "learning_rate": 2.9406554472984943e-06, + "loss": 1.2437, + "step": 333 + }, + { + "epoch": 0.014788085407834587, + "grad_norm": 3.177570343017578, + "learning_rate": 2.9495128432240923e-06, + "loss": 1.2301, + "step": 334 + }, + { + "epoch": 0.014832361112648462, + "grad_norm": 1.8012443780899048, + "learning_rate": 2.9583702391496903e-06, + "loss": 1.2646, + "step": 335 + }, + { + "epoch": 0.014876636817462338, + "grad_norm": 2.0184402465820312, + "learning_rate": 2.9672276350752882e-06, + "loss": 1.3388, + "step": 336 + }, + { + "epoch": 0.014920912522276213, + "grad_norm": 1.7653419971466064, + "learning_rate": 2.9760850310008858e-06, + "loss": 1.1911, + "step": 337 + }, + { + "epoch": 0.01496518822709009, + "grad_norm": 1.824842095375061, + "learning_rate": 2.9849424269264837e-06, + "loss": 1.3006, + "step": 338 + }, + { + "epoch": 0.015009463931903966, + "grad_norm": 1.8177319765090942, + "learning_rate": 2.9937998228520813e-06, + "loss": 1.2941, + "step": 339 + }, + { + "epoch": 0.015053739636717842, + "grad_norm": 1.7009533643722534, + "learning_rate": 3.0026572187776797e-06, + "loss": 1.2776, + "step": 340 + }, + { + "epoch": 0.015098015341531717, + "grad_norm": 2.0164287090301514, + "learning_rate": 3.0115146147032776e-06, + "loss": 1.2177, + "step": 341 + }, + { + "epoch": 0.015142291046345594, + "grad_norm": 1.9888083934783936, + "learning_rate": 3.020372010628875e-06, + "loss": 1.2365, + "step": 342 + }, + { + "epoch": 0.01518656675115947, + "grad_norm": 1.8607722520828247, + "learning_rate": 3.029229406554473e-06, + "loss": 1.2499, + "step": 343 + }, + { + "epoch": 0.015230842455973346, + "grad_norm": 1.562185525894165, + "learning_rate": 3.0380868024800715e-06, + "loss": 1.1758, + "step": 344 + }, + { + "epoch": 0.015275118160787223, + "grad_norm": 1.4867573976516724, + "learning_rate": 3.046944198405669e-06, + "loss": 1.162, + "step": 345 + }, + { + "epoch": 0.015319393865601098, + "grad_norm": 1.7262084484100342, + "learning_rate": 3.055801594331267e-06, + "loss": 1.2312, + "step": 346 + }, + { + "epoch": 0.015363669570414974, + "grad_norm": 1.674058198928833, + "learning_rate": 3.0646589902568646e-06, + "loss": 1.2004, + "step": 347 + }, + { + "epoch": 0.01540794527522885, + "grad_norm": 1.8223772048950195, + "learning_rate": 3.0735163861824626e-06, + "loss": 1.2443, + "step": 348 + }, + { + "epoch": 0.015452220980042727, + "grad_norm": 1.8090144395828247, + "learning_rate": 3.0823737821080605e-06, + "loss": 1.2155, + "step": 349 + }, + { + "epoch": 0.015496496684856602, + "grad_norm": 1.7094990015029907, + "learning_rate": 3.0912311780336585e-06, + "loss": 1.3006, + "step": 350 + }, + { + "epoch": 0.015540772389670478, + "grad_norm": 2.0318684577941895, + "learning_rate": 3.100088573959256e-06, + "loss": 1.2646, + "step": 351 + }, + { + "epoch": 0.015585048094484353, + "grad_norm": 2.0940473079681396, + "learning_rate": 3.108945969884854e-06, + "loss": 1.2331, + "step": 352 + }, + { + "epoch": 0.01562932379929823, + "grad_norm": 1.6789638996124268, + "learning_rate": 3.1178033658104516e-06, + "loss": 1.2384, + "step": 353 + }, + { + "epoch": 0.015673599504112106, + "grad_norm": 1.8415131568908691, + "learning_rate": 3.12666076173605e-06, + "loss": 1.2738, + "step": 354 + }, + { + "epoch": 0.015717875208925983, + "grad_norm": 1.6255334615707397, + "learning_rate": 3.135518157661648e-06, + "loss": 1.1871, + "step": 355 + }, + { + "epoch": 0.015762150913739857, + "grad_norm": 2.5051217079162598, + "learning_rate": 3.1443755535872455e-06, + "loss": 1.372, + "step": 356 + }, + { + "epoch": 0.015806426618553734, + "grad_norm": 2.0492818355560303, + "learning_rate": 3.1532329495128434e-06, + "loss": 1.4099, + "step": 357 + }, + { + "epoch": 0.01585070232336761, + "grad_norm": 1.7024933099746704, + "learning_rate": 3.162090345438442e-06, + "loss": 1.1657, + "step": 358 + }, + { + "epoch": 0.015894978028181486, + "grad_norm": 1.8284938335418701, + "learning_rate": 3.1709477413640394e-06, + "loss": 1.2387, + "step": 359 + }, + { + "epoch": 0.015939253732995363, + "grad_norm": 1.972596526145935, + "learning_rate": 3.1798051372896373e-06, + "loss": 1.306, + "step": 360 + }, + { + "epoch": 0.015983529437809237, + "grad_norm": 1.5137332677841187, + "learning_rate": 3.188662533215235e-06, + "loss": 1.1629, + "step": 361 + }, + { + "epoch": 0.016027805142623114, + "grad_norm": 1.6868336200714111, + "learning_rate": 3.197519929140833e-06, + "loss": 1.2312, + "step": 362 + }, + { + "epoch": 0.01607208084743699, + "grad_norm": 2.18588924407959, + "learning_rate": 3.206377325066431e-06, + "loss": 1.4163, + "step": 363 + }, + { + "epoch": 0.016116356552250865, + "grad_norm": 1.695986270904541, + "learning_rate": 3.2152347209920288e-06, + "loss": 1.315, + "step": 364 + }, + { + "epoch": 0.016160632257064742, + "grad_norm": 1.69026517868042, + "learning_rate": 3.2240921169176263e-06, + "loss": 1.2166, + "step": 365 + }, + { + "epoch": 0.01620490796187862, + "grad_norm": 1.9132797718048096, + "learning_rate": 3.2329495128432243e-06, + "loss": 1.2748, + "step": 366 + }, + { + "epoch": 0.016249183666692493, + "grad_norm": 2.1367928981781006, + "learning_rate": 3.241806908768822e-06, + "loss": 1.2476, + "step": 367 + }, + { + "epoch": 0.01629345937150637, + "grad_norm": 2.184598445892334, + "learning_rate": 3.2506643046944202e-06, + "loss": 1.2571, + "step": 368 + }, + { + "epoch": 0.016337735076320248, + "grad_norm": 1.8686652183532715, + "learning_rate": 3.259521700620018e-06, + "loss": 1.2418, + "step": 369 + }, + { + "epoch": 0.01638201078113412, + "grad_norm": 1.8790392875671387, + "learning_rate": 3.2683790965456157e-06, + "loss": 1.2709, + "step": 370 + }, + { + "epoch": 0.016426286485948, + "grad_norm": 1.7831380367279053, + "learning_rate": 3.2772364924712137e-06, + "loss": 1.2344, + "step": 371 + }, + { + "epoch": 0.016470562190761873, + "grad_norm": 1.694366693496704, + "learning_rate": 3.2860938883968117e-06, + "loss": 1.2333, + "step": 372 + }, + { + "epoch": 0.01651483789557575, + "grad_norm": 1.6592434644699097, + "learning_rate": 3.2949512843224096e-06, + "loss": 1.2493, + "step": 373 + }, + { + "epoch": 0.016559113600389627, + "grad_norm": 1.5511507987976074, + "learning_rate": 3.303808680248007e-06, + "loss": 1.0882, + "step": 374 + }, + { + "epoch": 0.0166033893052035, + "grad_norm": 1.9551146030426025, + "learning_rate": 3.312666076173605e-06, + "loss": 1.2804, + "step": 375 + }, + { + "epoch": 0.01664766501001738, + "grad_norm": 1.863979458808899, + "learning_rate": 3.3215234720992027e-06, + "loss": 1.2015, + "step": 376 + }, + { + "epoch": 0.016691940714831256, + "grad_norm": 1.8993819952011108, + "learning_rate": 3.330380868024801e-06, + "loss": 1.1628, + "step": 377 + }, + { + "epoch": 0.01673621641964513, + "grad_norm": 1.836581826210022, + "learning_rate": 3.339238263950399e-06, + "loss": 1.3004, + "step": 378 + }, + { + "epoch": 0.016780492124459007, + "grad_norm": 1.7409298419952393, + "learning_rate": 3.3480956598759966e-06, + "loss": 1.25, + "step": 379 + }, + { + "epoch": 0.016824767829272884, + "grad_norm": 1.8008534908294678, + "learning_rate": 3.3569530558015946e-06, + "loss": 1.1599, + "step": 380 + }, + { + "epoch": 0.016869043534086758, + "grad_norm": 2.0004172325134277, + "learning_rate": 3.365810451727192e-06, + "loss": 1.2798, + "step": 381 + }, + { + "epoch": 0.016913319238900635, + "grad_norm": 1.840151309967041, + "learning_rate": 3.3746678476527905e-06, + "loss": 1.2741, + "step": 382 + }, + { + "epoch": 0.01695759494371451, + "grad_norm": 2.004739284515381, + "learning_rate": 3.3835252435783885e-06, + "loss": 1.2048, + "step": 383 + }, + { + "epoch": 0.017001870648528386, + "grad_norm": 1.7518130540847778, + "learning_rate": 3.392382639503986e-06, + "loss": 1.1427, + "step": 384 + }, + { + "epoch": 0.017046146353342263, + "grad_norm": 1.6033095121383667, + "learning_rate": 3.401240035429584e-06, + "loss": 1.1824, + "step": 385 + }, + { + "epoch": 0.017090422058156137, + "grad_norm": 1.9855581521987915, + "learning_rate": 3.410097431355182e-06, + "loss": 1.2308, + "step": 386 + }, + { + "epoch": 0.017134697762970014, + "grad_norm": 1.7443801164627075, + "learning_rate": 3.41895482728078e-06, + "loss": 1.1913, + "step": 387 + }, + { + "epoch": 0.01717897346778389, + "grad_norm": 2.03080153465271, + "learning_rate": 3.4278122232063775e-06, + "loss": 1.3539, + "step": 388 + }, + { + "epoch": 0.017223249172597765, + "grad_norm": 2.0868656635284424, + "learning_rate": 3.4366696191319754e-06, + "loss": 1.1657, + "step": 389 + }, + { + "epoch": 0.017267524877411643, + "grad_norm": 1.6210265159606934, + "learning_rate": 3.445527015057573e-06, + "loss": 1.1982, + "step": 390 + }, + { + "epoch": 0.01731180058222552, + "grad_norm": 1.644962191581726, + "learning_rate": 3.4543844109831714e-06, + "loss": 1.1372, + "step": 391 + }, + { + "epoch": 0.017356076287039394, + "grad_norm": 1.58273446559906, + "learning_rate": 3.4632418069087693e-06, + "loss": 1.1993, + "step": 392 + }, + { + "epoch": 0.01740035199185327, + "grad_norm": 2.032532215118408, + "learning_rate": 3.472099202834367e-06, + "loss": 1.2084, + "step": 393 + }, + { + "epoch": 0.017444627696667145, + "grad_norm": 1.8565486669540405, + "learning_rate": 3.480956598759965e-06, + "loss": 1.212, + "step": 394 + }, + { + "epoch": 0.017488903401481022, + "grad_norm": 1.5811102390289307, + "learning_rate": 3.4898139946855624e-06, + "loss": 1.196, + "step": 395 + }, + { + "epoch": 0.0175331791062949, + "grad_norm": 1.8889999389648438, + "learning_rate": 3.4986713906111608e-06, + "loss": 1.2244, + "step": 396 + }, + { + "epoch": 0.017577454811108773, + "grad_norm": 2.7975072860717773, + "learning_rate": 3.5075287865367587e-06, + "loss": 1.3666, + "step": 397 + }, + { + "epoch": 0.01762173051592265, + "grad_norm": 1.659697413444519, + "learning_rate": 3.5163861824623563e-06, + "loss": 1.2815, + "step": 398 + }, + { + "epoch": 0.017666006220736528, + "grad_norm": 1.5293773412704468, + "learning_rate": 3.5252435783879543e-06, + "loss": 1.2218, + "step": 399 + }, + { + "epoch": 0.0177102819255504, + "grad_norm": 1.8374724388122559, + "learning_rate": 3.5341009743135522e-06, + "loss": 1.2226, + "step": 400 + }, + { + "epoch": 0.01775455763036428, + "grad_norm": 1.8042172193527222, + "learning_rate": 3.54295837023915e-06, + "loss": 1.2233, + "step": 401 + }, + { + "epoch": 0.017798833335178156, + "grad_norm": 1.5975756645202637, + "learning_rate": 3.5518157661647477e-06, + "loss": 1.1628, + "step": 402 + }, + { + "epoch": 0.01784310903999203, + "grad_norm": 1.6237013339996338, + "learning_rate": 3.5606731620903457e-06, + "loss": 1.1912, + "step": 403 + }, + { + "epoch": 0.017887384744805907, + "grad_norm": 1.6808403730392456, + "learning_rate": 3.5695305580159433e-06, + "loss": 1.2324, + "step": 404 + }, + { + "epoch": 0.01793166044961978, + "grad_norm": 1.9515408277511597, + "learning_rate": 3.5783879539415416e-06, + "loss": 1.2756, + "step": 405 + }, + { + "epoch": 0.017975936154433658, + "grad_norm": 2.4738612174987793, + "learning_rate": 3.5872453498671396e-06, + "loss": 1.3075, + "step": 406 + }, + { + "epoch": 0.018020211859247535, + "grad_norm": 2.6360106468200684, + "learning_rate": 3.596102745792737e-06, + "loss": 1.3406, + "step": 407 + }, + { + "epoch": 0.01806448756406141, + "grad_norm": 1.802252173423767, + "learning_rate": 3.604960141718335e-06, + "loss": 1.2509, + "step": 408 + }, + { + "epoch": 0.018108763268875287, + "grad_norm": 1.7715836763381958, + "learning_rate": 3.6138175376439327e-06, + "loss": 1.234, + "step": 409 + }, + { + "epoch": 0.018153038973689164, + "grad_norm": 1.7379777431488037, + "learning_rate": 3.622674933569531e-06, + "loss": 1.2588, + "step": 410 + }, + { + "epoch": 0.018197314678503038, + "grad_norm": 1.71412193775177, + "learning_rate": 3.6315323294951286e-06, + "loss": 1.2906, + "step": 411 + }, + { + "epoch": 0.018241590383316915, + "grad_norm": 1.5402250289916992, + "learning_rate": 3.6403897254207266e-06, + "loss": 1.0943, + "step": 412 + }, + { + "epoch": 0.01828586608813079, + "grad_norm": 2.0503361225128174, + "learning_rate": 3.649247121346324e-06, + "loss": 1.3088, + "step": 413 + }, + { + "epoch": 0.018330141792944666, + "grad_norm": 1.7081193923950195, + "learning_rate": 3.6581045172719225e-06, + "loss": 1.1448, + "step": 414 + }, + { + "epoch": 0.018374417497758543, + "grad_norm": 1.6444225311279297, + "learning_rate": 3.6669619131975205e-06, + "loss": 1.1643, + "step": 415 + }, + { + "epoch": 0.018418693202572417, + "grad_norm": 1.6321308612823486, + "learning_rate": 3.675819309123118e-06, + "loss": 1.1188, + "step": 416 + }, + { + "epoch": 0.018462968907386294, + "grad_norm": 1.6428037881851196, + "learning_rate": 3.684676705048716e-06, + "loss": 1.2047, + "step": 417 + }, + { + "epoch": 0.01850724461220017, + "grad_norm": 1.7787985801696777, + "learning_rate": 3.6935341009743135e-06, + "loss": 1.2517, + "step": 418 + }, + { + "epoch": 0.018551520317014045, + "grad_norm": 1.7517402172088623, + "learning_rate": 3.702391496899912e-06, + "loss": 1.1902, + "step": 419 + }, + { + "epoch": 0.018595796021827923, + "grad_norm": 1.6731839179992676, + "learning_rate": 3.71124889282551e-06, + "loss": 1.1819, + "step": 420 + }, + { + "epoch": 0.0186400717266418, + "grad_norm": 1.6127426624298096, + "learning_rate": 3.7201062887511074e-06, + "loss": 1.1809, + "step": 421 + }, + { + "epoch": 0.018684347431455674, + "grad_norm": 1.6227391958236694, + "learning_rate": 3.7289636846767054e-06, + "loss": 1.1236, + "step": 422 + }, + { + "epoch": 0.01872862313626955, + "grad_norm": 1.8599915504455566, + "learning_rate": 3.737821080602303e-06, + "loss": 1.1553, + "step": 423 + }, + { + "epoch": 0.018772898841083425, + "grad_norm": 1.848939299583435, + "learning_rate": 3.7466784765279013e-06, + "loss": 1.1608, + "step": 424 + }, + { + "epoch": 0.018817174545897302, + "grad_norm": 1.9512988328933716, + "learning_rate": 3.755535872453499e-06, + "loss": 1.2137, + "step": 425 + }, + { + "epoch": 0.01886145025071118, + "grad_norm": 1.8413097858428955, + "learning_rate": 3.764393268379097e-06, + "loss": 1.2567, + "step": 426 + }, + { + "epoch": 0.018905725955525053, + "grad_norm": 1.7438374757766724, + "learning_rate": 3.7732506643046944e-06, + "loss": 1.2131, + "step": 427 + }, + { + "epoch": 0.01895000166033893, + "grad_norm": 1.7297375202178955, + "learning_rate": 3.7821080602302928e-06, + "loss": 1.2546, + "step": 428 + }, + { + "epoch": 0.018994277365152808, + "grad_norm": 1.8472219705581665, + "learning_rate": 3.7909654561558907e-06, + "loss": 1.1524, + "step": 429 + }, + { + "epoch": 0.01903855306996668, + "grad_norm": 2.041219472885132, + "learning_rate": 3.7998228520814883e-06, + "loss": 1.3234, + "step": 430 + }, + { + "epoch": 0.01908282877478056, + "grad_norm": 1.6864252090454102, + "learning_rate": 3.8086802480070863e-06, + "loss": 1.2179, + "step": 431 + }, + { + "epoch": 0.019127104479594436, + "grad_norm": 1.7797150611877441, + "learning_rate": 3.817537643932684e-06, + "loss": 1.2385, + "step": 432 + }, + { + "epoch": 0.01917138018440831, + "grad_norm": 1.7858232259750366, + "learning_rate": 3.826395039858282e-06, + "loss": 1.2835, + "step": 433 + }, + { + "epoch": 0.019215655889222187, + "grad_norm": 1.8604367971420288, + "learning_rate": 3.83525243578388e-06, + "loss": 1.3248, + "step": 434 + }, + { + "epoch": 0.01925993159403606, + "grad_norm": 4.350935459136963, + "learning_rate": 3.844109831709478e-06, + "loss": 1.2843, + "step": 435 + }, + { + "epoch": 0.019304207298849938, + "grad_norm": 1.929334282875061, + "learning_rate": 3.852967227635076e-06, + "loss": 1.2667, + "step": 436 + }, + { + "epoch": 0.019348483003663815, + "grad_norm": 1.8342339992523193, + "learning_rate": 3.861824623560673e-06, + "loss": 1.2753, + "step": 437 + }, + { + "epoch": 0.01939275870847769, + "grad_norm": 1.7558971643447876, + "learning_rate": 3.870682019486272e-06, + "loss": 1.2203, + "step": 438 + }, + { + "epoch": 0.019437034413291567, + "grad_norm": 1.7841942310333252, + "learning_rate": 3.879539415411869e-06, + "loss": 1.2431, + "step": 439 + }, + { + "epoch": 0.019481310118105444, + "grad_norm": 1.9081562757492065, + "learning_rate": 3.888396811337467e-06, + "loss": 1.2423, + "step": 440 + }, + { + "epoch": 0.019525585822919318, + "grad_norm": 1.6820799112319946, + "learning_rate": 3.897254207263065e-06, + "loss": 1.1298, + "step": 441 + }, + { + "epoch": 0.019569861527733195, + "grad_norm": 1.885330319404602, + "learning_rate": 3.9061116031886635e-06, + "loss": 1.2708, + "step": 442 + }, + { + "epoch": 0.019614137232547072, + "grad_norm": 1.8310706615447998, + "learning_rate": 3.914968999114261e-06, + "loss": 1.33, + "step": 443 + }, + { + "epoch": 0.019658412937360946, + "grad_norm": 1.9289698600769043, + "learning_rate": 3.9238263950398586e-06, + "loss": 1.1914, + "step": 444 + }, + { + "epoch": 0.019702688642174823, + "grad_norm": 1.7773529291152954, + "learning_rate": 3.932683790965456e-06, + "loss": 1.265, + "step": 445 + }, + { + "epoch": 0.019746964346988697, + "grad_norm": 1.7648308277130127, + "learning_rate": 3.9415411868910545e-06, + "loss": 1.1712, + "step": 446 + }, + { + "epoch": 0.019791240051802574, + "grad_norm": 2.129807949066162, + "learning_rate": 3.950398582816652e-06, + "loss": 1.1822, + "step": 447 + }, + { + "epoch": 0.01983551575661645, + "grad_norm": 1.8431291580200195, + "learning_rate": 3.9592559787422504e-06, + "loss": 1.182, + "step": 448 + }, + { + "epoch": 0.019879791461430325, + "grad_norm": 1.6370664834976196, + "learning_rate": 3.968113374667848e-06, + "loss": 1.1244, + "step": 449 + }, + { + "epoch": 0.019924067166244203, + "grad_norm": 1.7030508518218994, + "learning_rate": 3.9769707705934455e-06, + "loss": 1.1879, + "step": 450 + }, + { + "epoch": 0.01996834287105808, + "grad_norm": 1.6088805198669434, + "learning_rate": 3.985828166519043e-06, + "loss": 1.1372, + "step": 451 + }, + { + "epoch": 0.020012618575871954, + "grad_norm": 2.018432855606079, + "learning_rate": 3.9946855624446415e-06, + "loss": 1.2629, + "step": 452 + }, + { + "epoch": 0.02005689428068583, + "grad_norm": 1.6949594020843506, + "learning_rate": 4.00354295837024e-06, + "loss": 1.1956, + "step": 453 + }, + { + "epoch": 0.020101169985499708, + "grad_norm": 1.8518015146255493, + "learning_rate": 4.012400354295837e-06, + "loss": 1.2743, + "step": 454 + }, + { + "epoch": 0.020145445690313582, + "grad_norm": 1.953157901763916, + "learning_rate": 4.021257750221435e-06, + "loss": 1.2423, + "step": 455 + }, + { + "epoch": 0.02018972139512746, + "grad_norm": 1.8100833892822266, + "learning_rate": 4.030115146147033e-06, + "loss": 1.1696, + "step": 456 + }, + { + "epoch": 0.020233997099941333, + "grad_norm": 1.9205970764160156, + "learning_rate": 4.038972542072631e-06, + "loss": 1.2098, + "step": 457 + }, + { + "epoch": 0.02027827280475521, + "grad_norm": 1.6018813848495483, + "learning_rate": 4.047829937998229e-06, + "loss": 1.1548, + "step": 458 + }, + { + "epoch": 0.020322548509569088, + "grad_norm": 1.63569176197052, + "learning_rate": 4.056687333923827e-06, + "loss": 1.1747, + "step": 459 + }, + { + "epoch": 0.02036682421438296, + "grad_norm": 1.905340313911438, + "learning_rate": 4.065544729849424e-06, + "loss": 1.2397, + "step": 460 + }, + { + "epoch": 0.02041109991919684, + "grad_norm": 1.79994535446167, + "learning_rate": 4.074402125775023e-06, + "loss": 1.2513, + "step": 461 + }, + { + "epoch": 0.020455375624010716, + "grad_norm": 3.1559090614318848, + "learning_rate": 4.08325952170062e-06, + "loss": 1.2296, + "step": 462 + }, + { + "epoch": 0.02049965132882459, + "grad_norm": 1.9860984086990356, + "learning_rate": 4.092116917626218e-06, + "loss": 1.2408, + "step": 463 + }, + { + "epoch": 0.020543927033638467, + "grad_norm": 2.0497560501098633, + "learning_rate": 4.100974313551816e-06, + "loss": 1.1872, + "step": 464 + }, + { + "epoch": 0.020588202738452344, + "grad_norm": 1.7650630474090576, + "learning_rate": 4.109831709477414e-06, + "loss": 1.1571, + "step": 465 + }, + { + "epoch": 0.020632478443266218, + "grad_norm": 2.6257388591766357, + "learning_rate": 4.118689105403012e-06, + "loss": 1.2286, + "step": 466 + }, + { + "epoch": 0.020676754148080095, + "grad_norm": 1.6999276876449585, + "learning_rate": 4.12754650132861e-06, + "loss": 1.1521, + "step": 467 + }, + { + "epoch": 0.02072102985289397, + "grad_norm": 1.7056459188461304, + "learning_rate": 4.136403897254207e-06, + "loss": 1.1154, + "step": 468 + }, + { + "epoch": 0.020765305557707846, + "grad_norm": 1.8928890228271484, + "learning_rate": 4.145261293179806e-06, + "loss": 1.2265, + "step": 469 + }, + { + "epoch": 0.020809581262521724, + "grad_norm": 2.1643130779266357, + "learning_rate": 4.154118689105404e-06, + "loss": 1.2377, + "step": 470 + }, + { + "epoch": 0.020853856967335598, + "grad_norm": 1.7573343515396118, + "learning_rate": 4.1629760850310016e-06, + "loss": 1.1639, + "step": 471 + }, + { + "epoch": 0.020898132672149475, + "grad_norm": 1.795003056526184, + "learning_rate": 4.171833480956599e-06, + "loss": 1.1699, + "step": 472 + }, + { + "epoch": 0.020942408376963352, + "grad_norm": 1.6704621315002441, + "learning_rate": 4.180690876882197e-06, + "loss": 1.1863, + "step": 473 + }, + { + "epoch": 0.020986684081777226, + "grad_norm": 1.4344438314437866, + "learning_rate": 4.189548272807795e-06, + "loss": 1.1318, + "step": 474 + }, + { + "epoch": 0.021030959786591103, + "grad_norm": 1.9183332920074463, + "learning_rate": 4.198405668733393e-06, + "loss": 1.2383, + "step": 475 + }, + { + "epoch": 0.02107523549140498, + "grad_norm": 1.8122107982635498, + "learning_rate": 4.207263064658991e-06, + "loss": 1.1904, + "step": 476 + }, + { + "epoch": 0.021119511196218854, + "grad_norm": 1.741397738456726, + "learning_rate": 4.2161204605845885e-06, + "loss": 1.1508, + "step": 477 + }, + { + "epoch": 0.02116378690103273, + "grad_norm": 2.1095621585845947, + "learning_rate": 4.224977856510186e-06, + "loss": 1.2123, + "step": 478 + }, + { + "epoch": 0.021208062605846605, + "grad_norm": 1.628117561340332, + "learning_rate": 4.233835252435784e-06, + "loss": 1.2205, + "step": 479 + }, + { + "epoch": 0.021252338310660483, + "grad_norm": 1.757519245147705, + "learning_rate": 4.242692648361382e-06, + "loss": 1.1492, + "step": 480 + }, + { + "epoch": 0.02129661401547436, + "grad_norm": 1.6781972646713257, + "learning_rate": 4.25155004428698e-06, + "loss": 1.0435, + "step": 481 + }, + { + "epoch": 0.021340889720288234, + "grad_norm": 1.5599050521850586, + "learning_rate": 4.260407440212578e-06, + "loss": 1.1029, + "step": 482 + }, + { + "epoch": 0.02138516542510211, + "grad_norm": 1.783575415611267, + "learning_rate": 4.2692648361381755e-06, + "loss": 1.1457, + "step": 483 + }, + { + "epoch": 0.021429441129915988, + "grad_norm": 1.826880931854248, + "learning_rate": 4.278122232063774e-06, + "loss": 1.1916, + "step": 484 + }, + { + "epoch": 0.021473716834729862, + "grad_norm": 2.1122183799743652, + "learning_rate": 4.2869796279893714e-06, + "loss": 1.2823, + "step": 485 + }, + { + "epoch": 0.02151799253954374, + "grad_norm": 1.672099232673645, + "learning_rate": 4.295837023914969e-06, + "loss": 1.1808, + "step": 486 + }, + { + "epoch": 0.021562268244357616, + "grad_norm": 1.8099795579910278, + "learning_rate": 4.304694419840567e-06, + "loss": 1.1951, + "step": 487 + }, + { + "epoch": 0.02160654394917149, + "grad_norm": 2.019827127456665, + "learning_rate": 4.313551815766165e-06, + "loss": 1.2436, + "step": 488 + }, + { + "epoch": 0.021650819653985368, + "grad_norm": 1.8050892353057861, + "learning_rate": 4.322409211691763e-06, + "loss": 1.1756, + "step": 489 + }, + { + "epoch": 0.02169509535879924, + "grad_norm": 1.9618782997131348, + "learning_rate": 4.331266607617361e-06, + "loss": 1.1942, + "step": 490 + }, + { + "epoch": 0.02173937106361312, + "grad_norm": 1.6199660301208496, + "learning_rate": 4.340124003542958e-06, + "loss": 1.1647, + "step": 491 + }, + { + "epoch": 0.021783646768426996, + "grad_norm": 2.030738592147827, + "learning_rate": 4.348981399468557e-06, + "loss": 1.1951, + "step": 492 + }, + { + "epoch": 0.02182792247324087, + "grad_norm": 1.960539698600769, + "learning_rate": 4.357838795394154e-06, + "loss": 1.3482, + "step": 493 + }, + { + "epoch": 0.021872198178054747, + "grad_norm": 1.767751693725586, + "learning_rate": 4.366696191319753e-06, + "loss": 1.2076, + "step": 494 + }, + { + "epoch": 0.021916473882868624, + "grad_norm": 2.3365938663482666, + "learning_rate": 4.37555358724535e-06, + "loss": 1.2614, + "step": 495 + }, + { + "epoch": 0.021960749587682498, + "grad_norm": 2.5246191024780273, + "learning_rate": 4.384410983170948e-06, + "loss": 1.2827, + "step": 496 + }, + { + "epoch": 0.022005025292496375, + "grad_norm": 2.1244304180145264, + "learning_rate": 4.393268379096546e-06, + "loss": 1.218, + "step": 497 + }, + { + "epoch": 0.022049300997310253, + "grad_norm": 1.6184219121932983, + "learning_rate": 4.402125775022144e-06, + "loss": 1.1689, + "step": 498 + }, + { + "epoch": 0.022093576702124126, + "grad_norm": 1.719342827796936, + "learning_rate": 4.410983170947742e-06, + "loss": 1.2238, + "step": 499 + }, + { + "epoch": 0.022137852406938004, + "grad_norm": 1.8495233058929443, + "learning_rate": 4.41984056687334e-06, + "loss": 1.1468, + "step": 500 + }, + { + "epoch": 0.022182128111751877, + "grad_norm": 1.7275099754333496, + "learning_rate": 4.428697962798937e-06, + "loss": 1.236, + "step": 501 + }, + { + "epoch": 0.022226403816565755, + "grad_norm": 1.9735071659088135, + "learning_rate": 4.437555358724535e-06, + "loss": 1.2763, + "step": 502 + }, + { + "epoch": 0.022270679521379632, + "grad_norm": 2.2084672451019287, + "learning_rate": 4.446412754650133e-06, + "loss": 1.2759, + "step": 503 + }, + { + "epoch": 0.022314955226193506, + "grad_norm": 1.7809523344039917, + "learning_rate": 4.4552701505757315e-06, + "loss": 1.192, + "step": 504 + }, + { + "epoch": 0.022359230931007383, + "grad_norm": 1.569807529449463, + "learning_rate": 4.464127546501329e-06, + "loss": 1.1954, + "step": 505 + }, + { + "epoch": 0.02240350663582126, + "grad_norm": 2.0102994441986084, + "learning_rate": 4.472984942426927e-06, + "loss": 1.1855, + "step": 506 + }, + { + "epoch": 0.022447782340635134, + "grad_norm": 1.5835202932357788, + "learning_rate": 4.481842338352524e-06, + "loss": 1.1282, + "step": 507 + }, + { + "epoch": 0.02249205804544901, + "grad_norm": 1.9046722650527954, + "learning_rate": 4.4906997342781226e-06, + "loss": 1.2971, + "step": 508 + }, + { + "epoch": 0.02253633375026289, + "grad_norm": 1.8120851516723633, + "learning_rate": 4.499557130203721e-06, + "loss": 1.2757, + "step": 509 + }, + { + "epoch": 0.022580609455076762, + "grad_norm": 2.0721306800842285, + "learning_rate": 4.5084145261293185e-06, + "loss": 1.3265, + "step": 510 + }, + { + "epoch": 0.02262488515989064, + "grad_norm": 1.9498564004898071, + "learning_rate": 4.517271922054916e-06, + "loss": 1.2833, + "step": 511 + }, + { + "epoch": 0.022669160864704514, + "grad_norm": 1.936767578125, + "learning_rate": 4.5261293179805144e-06, + "loss": 1.1873, + "step": 512 + }, + { + "epoch": 0.02271343656951839, + "grad_norm": 2.393101692199707, + "learning_rate": 4.534986713906112e-06, + "loss": 1.1273, + "step": 513 + }, + { + "epoch": 0.022757712274332268, + "grad_norm": 1.8263933658599854, + "learning_rate": 4.5438441098317095e-06, + "loss": 1.2932, + "step": 514 + }, + { + "epoch": 0.022801987979146142, + "grad_norm": 1.814591884613037, + "learning_rate": 4.552701505757308e-06, + "loss": 1.2514, + "step": 515 + }, + { + "epoch": 0.02284626368396002, + "grad_norm": 1.5963064432144165, + "learning_rate": 4.5615589016829055e-06, + "loss": 1.1854, + "step": 516 + }, + { + "epoch": 0.022890539388773896, + "grad_norm": 1.7082616090774536, + "learning_rate": 4.570416297608504e-06, + "loss": 1.151, + "step": 517 + }, + { + "epoch": 0.02293481509358777, + "grad_norm": 1.6426870822906494, + "learning_rate": 4.579273693534101e-06, + "loss": 1.243, + "step": 518 + }, + { + "epoch": 0.022979090798401648, + "grad_norm": 1.8850412368774414, + "learning_rate": 4.588131089459699e-06, + "loss": 1.2495, + "step": 519 + }, + { + "epoch": 0.023023366503215525, + "grad_norm": 1.9359650611877441, + "learning_rate": 4.596988485385297e-06, + "loss": 1.2341, + "step": 520 + }, + { + "epoch": 0.0230676422080294, + "grad_norm": 1.7925796508789062, + "learning_rate": 4.605845881310895e-06, + "loss": 1.0923, + "step": 521 + }, + { + "epoch": 0.023111917912843276, + "grad_norm": 1.6918660402297974, + "learning_rate": 4.614703277236493e-06, + "loss": 1.149, + "step": 522 + }, + { + "epoch": 0.02315619361765715, + "grad_norm": 1.5923166275024414, + "learning_rate": 4.623560673162091e-06, + "loss": 1.1533, + "step": 523 + }, + { + "epoch": 0.023200469322471027, + "grad_norm": 2.102856397628784, + "learning_rate": 4.632418069087688e-06, + "loss": 1.3061, + "step": 524 + }, + { + "epoch": 0.023244745027284904, + "grad_norm": 1.9140535593032837, + "learning_rate": 4.641275465013286e-06, + "loss": 1.2325, + "step": 525 + }, + { + "epoch": 0.023289020732098778, + "grad_norm": 1.6876873970031738, + "learning_rate": 4.650132860938884e-06, + "loss": 1.1042, + "step": 526 + }, + { + "epoch": 0.023333296436912655, + "grad_norm": 1.6531200408935547, + "learning_rate": 4.658990256864483e-06, + "loss": 1.1244, + "step": 527 + }, + { + "epoch": 0.023377572141726533, + "grad_norm": 2.0746023654937744, + "learning_rate": 4.66784765279008e-06, + "loss": 1.1781, + "step": 528 + }, + { + "epoch": 0.023421847846540406, + "grad_norm": 1.7426127195358276, + "learning_rate": 4.676705048715678e-06, + "loss": 1.162, + "step": 529 + }, + { + "epoch": 0.023466123551354284, + "grad_norm": 1.5324376821517944, + "learning_rate": 4.685562444641275e-06, + "loss": 1.1235, + "step": 530 + }, + { + "epoch": 0.023510399256168157, + "grad_norm": 1.9309155941009521, + "learning_rate": 4.694419840566874e-06, + "loss": 1.2047, + "step": 531 + }, + { + "epoch": 0.023554674960982035, + "grad_norm": 1.7351802587509155, + "learning_rate": 4.703277236492472e-06, + "loss": 1.2164, + "step": 532 + }, + { + "epoch": 0.023598950665795912, + "grad_norm": 1.7108585834503174, + "learning_rate": 4.71213463241807e-06, + "loss": 1.1868, + "step": 533 + }, + { + "epoch": 0.023643226370609786, + "grad_norm": 1.7044919729232788, + "learning_rate": 4.720992028343667e-06, + "loss": 1.1485, + "step": 534 + }, + { + "epoch": 0.023687502075423663, + "grad_norm": 2.1445953845977783, + "learning_rate": 4.729849424269265e-06, + "loss": 1.2497, + "step": 535 + }, + { + "epoch": 0.02373177778023754, + "grad_norm": 1.9022043943405151, + "learning_rate": 4.738706820194863e-06, + "loss": 1.0906, + "step": 536 + }, + { + "epoch": 0.023776053485051414, + "grad_norm": 1.423737645149231, + "learning_rate": 4.747564216120461e-06, + "loss": 1.1289, + "step": 537 + }, + { + "epoch": 0.02382032918986529, + "grad_norm": 1.533555030822754, + "learning_rate": 4.756421612046059e-06, + "loss": 1.1264, + "step": 538 + }, + { + "epoch": 0.02386460489467917, + "grad_norm": 1.7451767921447754, + "learning_rate": 4.765279007971657e-06, + "loss": 1.234, + "step": 539 + }, + { + "epoch": 0.023908880599493042, + "grad_norm": 2.065422296524048, + "learning_rate": 4.774136403897255e-06, + "loss": 1.2585, + "step": 540 + }, + { + "epoch": 0.02395315630430692, + "grad_norm": 1.8063251972198486, + "learning_rate": 4.7829937998228525e-06, + "loss": 1.2951, + "step": 541 + }, + { + "epoch": 0.023997432009120793, + "grad_norm": 1.593551754951477, + "learning_rate": 4.79185119574845e-06, + "loss": 1.1892, + "step": 542 + }, + { + "epoch": 0.02404170771393467, + "grad_norm": 1.6646034717559814, + "learning_rate": 4.8007085916740485e-06, + "loss": 1.2432, + "step": 543 + }, + { + "epoch": 0.024085983418748548, + "grad_norm": 1.8767844438552856, + "learning_rate": 4.809565987599646e-06, + "loss": 1.25, + "step": 544 + }, + { + "epoch": 0.024130259123562422, + "grad_norm": 2.355994939804077, + "learning_rate": 4.818423383525244e-06, + "loss": 1.1152, + "step": 545 + }, + { + "epoch": 0.0241745348283763, + "grad_norm": 1.6561205387115479, + "learning_rate": 4.827280779450842e-06, + "loss": 1.179, + "step": 546 + }, + { + "epoch": 0.024218810533190176, + "grad_norm": 1.783933162689209, + "learning_rate": 4.8361381753764395e-06, + "loss": 1.1039, + "step": 547 + }, + { + "epoch": 0.02426308623800405, + "grad_norm": 1.4640839099884033, + "learning_rate": 4.844995571302038e-06, + "loss": 1.0059, + "step": 548 + }, + { + "epoch": 0.024307361942817927, + "grad_norm": 1.8851208686828613, + "learning_rate": 4.8538529672276354e-06, + "loss": 1.1307, + "step": 549 + }, + { + "epoch": 0.024351637647631805, + "grad_norm": 1.9281561374664307, + "learning_rate": 4.862710363153234e-06, + "loss": 1.2024, + "step": 550 + }, + { + "epoch": 0.02439591335244568, + "grad_norm": 1.5656177997589111, + "learning_rate": 4.871567759078831e-06, + "loss": 1.112, + "step": 551 + }, + { + "epoch": 0.024440189057259556, + "grad_norm": 1.6560369729995728, + "learning_rate": 4.880425155004429e-06, + "loss": 1.2292, + "step": 552 + }, + { + "epoch": 0.02448446476207343, + "grad_norm": 1.9741536378860474, + "learning_rate": 4.8892825509300264e-06, + "loss": 1.2884, + "step": 553 + }, + { + "epoch": 0.024528740466887307, + "grad_norm": 1.7307358980178833, + "learning_rate": 4.898139946855625e-06, + "loss": 1.2074, + "step": 554 + }, + { + "epoch": 0.024573016171701184, + "grad_norm": 1.7997337579727173, + "learning_rate": 4.906997342781223e-06, + "loss": 1.1938, + "step": 555 + }, + { + "epoch": 0.024617291876515058, + "grad_norm": 1.6764367818832397, + "learning_rate": 4.915854738706821e-06, + "loss": 1.1895, + "step": 556 + }, + { + "epoch": 0.024661567581328935, + "grad_norm": 1.936962604522705, + "learning_rate": 4.924712134632418e-06, + "loss": 1.2839, + "step": 557 + }, + { + "epoch": 0.024705843286142812, + "grad_norm": 1.495977759361267, + "learning_rate": 4.933569530558016e-06, + "loss": 1.0918, + "step": 558 + }, + { + "epoch": 0.024750118990956686, + "grad_norm": 1.7472559213638306, + "learning_rate": 4.942426926483614e-06, + "loss": 1.1202, + "step": 559 + }, + { + "epoch": 0.024794394695770564, + "grad_norm": 1.7280982732772827, + "learning_rate": 4.951284322409212e-06, + "loss": 1.1852, + "step": 560 + }, + { + "epoch": 0.02483867040058444, + "grad_norm": 1.595840573310852, + "learning_rate": 4.96014171833481e-06, + "loss": 1.113, + "step": 561 + }, + { + "epoch": 0.024882946105398315, + "grad_norm": 1.764522910118103, + "learning_rate": 4.968999114260408e-06, + "loss": 1.1466, + "step": 562 + }, + { + "epoch": 0.024927221810212192, + "grad_norm": 1.6819932460784912, + "learning_rate": 4.977856510186005e-06, + "loss": 1.095, + "step": 563 + }, + { + "epoch": 0.024971497515026066, + "grad_norm": 1.690833330154419, + "learning_rate": 4.986713906111604e-06, + "loss": 1.1303, + "step": 564 + }, + { + "epoch": 0.025015773219839943, + "grad_norm": 1.7937945127487183, + "learning_rate": 4.995571302037201e-06, + "loss": 1.1973, + "step": 565 + }, + { + "epoch": 0.02506004892465382, + "grad_norm": 1.8100557327270508, + "learning_rate": 5.0044286979628e-06, + "loss": 1.2075, + "step": 566 + }, + { + "epoch": 0.025104324629467694, + "grad_norm": 1.4266164302825928, + "learning_rate": 5.013286093888398e-06, + "loss": 1.0273, + "step": 567 + }, + { + "epoch": 0.02514860033428157, + "grad_norm": 1.6331342458724976, + "learning_rate": 5.0221434898139955e-06, + "loss": 1.133, + "step": 568 + }, + { + "epoch": 0.02519287603909545, + "grad_norm": 1.826930046081543, + "learning_rate": 5.031000885739593e-06, + "loss": 1.1849, + "step": 569 + }, + { + "epoch": 0.025237151743909322, + "grad_norm": 1.7403701543807983, + "learning_rate": 5.039858281665191e-06, + "loss": 1.1697, + "step": 570 + }, + { + "epoch": 0.0252814274487232, + "grad_norm": 1.7956746816635132, + "learning_rate": 5.048715677590789e-06, + "loss": 1.1819, + "step": 571 + }, + { + "epoch": 0.025325703153537077, + "grad_norm": 1.8329026699066162, + "learning_rate": 5.0575730735163866e-06, + "loss": 1.1137, + "step": 572 + }, + { + "epoch": 0.02536997885835095, + "grad_norm": 1.78904128074646, + "learning_rate": 5.066430469441984e-06, + "loss": 1.2288, + "step": 573 + }, + { + "epoch": 0.025414254563164828, + "grad_norm": 1.9763635396957397, + "learning_rate": 5.075287865367582e-06, + "loss": 1.145, + "step": 574 + }, + { + "epoch": 0.025458530267978702, + "grad_norm": 1.9641051292419434, + "learning_rate": 5.084145261293181e-06, + "loss": 1.2831, + "step": 575 + }, + { + "epoch": 0.02550280597279258, + "grad_norm": 1.8374069929122925, + "learning_rate": 5.0930026572187784e-06, + "loss": 1.2048, + "step": 576 + }, + { + "epoch": 0.025547081677606456, + "grad_norm": 2.150510549545288, + "learning_rate": 5.101860053144376e-06, + "loss": 1.105, + "step": 577 + }, + { + "epoch": 0.02559135738242033, + "grad_norm": 1.7230771780014038, + "learning_rate": 5.110717449069974e-06, + "loss": 1.1294, + "step": 578 + }, + { + "epoch": 0.025635633087234207, + "grad_norm": 1.73488187789917, + "learning_rate": 5.119574844995572e-06, + "loss": 1.1305, + "step": 579 + }, + { + "epoch": 0.025679908792048085, + "grad_norm": 1.5758365392684937, + "learning_rate": 5.1284322409211695e-06, + "loss": 1.1244, + "step": 580 + }, + { + "epoch": 0.02572418449686196, + "grad_norm": 1.752351999282837, + "learning_rate": 5.137289636846767e-06, + "loss": 1.0942, + "step": 581 + }, + { + "epoch": 0.025768460201675836, + "grad_norm": 1.6962940692901611, + "learning_rate": 5.146147032772365e-06, + "loss": 1.1735, + "step": 582 + }, + { + "epoch": 0.025812735906489713, + "grad_norm": 2.0588443279266357, + "learning_rate": 5.155004428697963e-06, + "loss": 1.2229, + "step": 583 + }, + { + "epoch": 0.025857011611303587, + "grad_norm": 1.690758466720581, + "learning_rate": 5.1638618246235605e-06, + "loss": 1.1338, + "step": 584 + }, + { + "epoch": 0.025901287316117464, + "grad_norm": 2.3536369800567627, + "learning_rate": 5.17271922054916e-06, + "loss": 1.1316, + "step": 585 + }, + { + "epoch": 0.025945563020931338, + "grad_norm": 1.9398491382598877, + "learning_rate": 5.181576616474757e-06, + "loss": 1.1603, + "step": 586 + }, + { + "epoch": 0.025989838725745215, + "grad_norm": 1.4992125034332275, + "learning_rate": 5.190434012400355e-06, + "loss": 1.1248, + "step": 587 + }, + { + "epoch": 0.026034114430559092, + "grad_norm": 2.1499311923980713, + "learning_rate": 5.199291408325952e-06, + "loss": 1.2376, + "step": 588 + }, + { + "epoch": 0.026078390135372966, + "grad_norm": 1.7317554950714111, + "learning_rate": 5.208148804251551e-06, + "loss": 1.2047, + "step": 589 + }, + { + "epoch": 0.026122665840186843, + "grad_norm": 1.4345091581344604, + "learning_rate": 5.217006200177148e-06, + "loss": 1.0834, + "step": 590 + }, + { + "epoch": 0.02616694154500072, + "grad_norm": 1.8488110303878784, + "learning_rate": 5.225863596102746e-06, + "loss": 1.1117, + "step": 591 + }, + { + "epoch": 0.026211217249814595, + "grad_norm": 1.9476152658462524, + "learning_rate": 5.234720992028343e-06, + "loss": 1.2497, + "step": 592 + }, + { + "epoch": 0.026255492954628472, + "grad_norm": 2.0874710083007812, + "learning_rate": 5.243578387953942e-06, + "loss": 1.1697, + "step": 593 + }, + { + "epoch": 0.02629976865944235, + "grad_norm": 1.8149152994155884, + "learning_rate": 5.25243578387954e-06, + "loss": 1.2412, + "step": 594 + }, + { + "epoch": 0.026344044364256223, + "grad_norm": 1.6988598108291626, + "learning_rate": 5.261293179805138e-06, + "loss": 1.1737, + "step": 595 + }, + { + "epoch": 0.0263883200690701, + "grad_norm": 2.624433755874634, + "learning_rate": 5.270150575730736e-06, + "loss": 1.3299, + "step": 596 + }, + { + "epoch": 0.026432595773883974, + "grad_norm": 1.791435956954956, + "learning_rate": 5.279007971656334e-06, + "loss": 1.1312, + "step": 597 + }, + { + "epoch": 0.02647687147869785, + "grad_norm": 1.7891743183135986, + "learning_rate": 5.287865367581931e-06, + "loss": 1.1565, + "step": 598 + }, + { + "epoch": 0.02652114718351173, + "grad_norm": 1.5200790166854858, + "learning_rate": 5.296722763507529e-06, + "loss": 1.04, + "step": 599 + }, + { + "epoch": 0.026565422888325602, + "grad_norm": 1.5711637735366821, + "learning_rate": 5.305580159433127e-06, + "loss": 1.1294, + "step": 600 + }, + { + "epoch": 0.02660969859313948, + "grad_norm": 1.8208727836608887, + "learning_rate": 5.314437555358725e-06, + "loss": 1.1081, + "step": 601 + }, + { + "epoch": 0.026653974297953357, + "grad_norm": 1.7482473850250244, + "learning_rate": 5.323294951284322e-06, + "loss": 1.1962, + "step": 602 + }, + { + "epoch": 0.02669825000276723, + "grad_norm": 1.8062106370925903, + "learning_rate": 5.3321523472099214e-06, + "loss": 1.1705, + "step": 603 + }, + { + "epoch": 0.026742525707581108, + "grad_norm": 1.4773896932601929, + "learning_rate": 5.341009743135519e-06, + "loss": 1.1329, + "step": 604 + }, + { + "epoch": 0.026786801412394985, + "grad_norm": 1.7367148399353027, + "learning_rate": 5.3498671390611165e-06, + "loss": 1.2368, + "step": 605 + }, + { + "epoch": 0.02683107711720886, + "grad_norm": 1.7150133848190308, + "learning_rate": 5.358724534986715e-06, + "loss": 1.0972, + "step": 606 + }, + { + "epoch": 0.026875352822022736, + "grad_norm": 1.8515129089355469, + "learning_rate": 5.3675819309123125e-06, + "loss": 1.149, + "step": 607 + }, + { + "epoch": 0.02691962852683661, + "grad_norm": 1.5420953035354614, + "learning_rate": 5.37643932683791e-06, + "loss": 1.059, + "step": 608 + }, + { + "epoch": 0.026963904231650487, + "grad_norm": 1.6444766521453857, + "learning_rate": 5.3852967227635076e-06, + "loss": 1.1773, + "step": 609 + }, + { + "epoch": 0.027008179936464365, + "grad_norm": 1.8091540336608887, + "learning_rate": 5.394154118689106e-06, + "loss": 1.2234, + "step": 610 + }, + { + "epoch": 0.02705245564127824, + "grad_norm": 1.6999022960662842, + "learning_rate": 5.4030115146147035e-06, + "loss": 1.1319, + "step": 611 + }, + { + "epoch": 0.027096731346092116, + "grad_norm": 1.6983284950256348, + "learning_rate": 5.411868910540301e-06, + "loss": 1.067, + "step": 612 + }, + { + "epoch": 0.027141007050905993, + "grad_norm": 1.910929560661316, + "learning_rate": 5.4207263064659e-06, + "loss": 1.2353, + "step": 613 + }, + { + "epoch": 0.027185282755719867, + "grad_norm": 1.8668338060379028, + "learning_rate": 5.429583702391498e-06, + "loss": 1.1085, + "step": 614 + }, + { + "epoch": 0.027229558460533744, + "grad_norm": 1.8587762117385864, + "learning_rate": 5.438441098317095e-06, + "loss": 1.1873, + "step": 615 + }, + { + "epoch": 0.02727383416534762, + "grad_norm": 1.7175986766815186, + "learning_rate": 5.447298494242693e-06, + "loss": 1.146, + "step": 616 + }, + { + "epoch": 0.027318109870161495, + "grad_norm": 1.7718364000320435, + "learning_rate": 5.456155890168291e-06, + "loss": 1.1859, + "step": 617 + }, + { + "epoch": 0.027362385574975372, + "grad_norm": 1.809191107749939, + "learning_rate": 5.465013286093889e-06, + "loss": 1.267, + "step": 618 + }, + { + "epoch": 0.027406661279789246, + "grad_norm": 1.6157780885696411, + "learning_rate": 5.473870682019486e-06, + "loss": 1.1242, + "step": 619 + }, + { + "epoch": 0.027450936984603123, + "grad_norm": 1.6560872793197632, + "learning_rate": 5.482728077945084e-06, + "loss": 1.0881, + "step": 620 + }, + { + "epoch": 0.027495212689417, + "grad_norm": 2.1042487621307373, + "learning_rate": 5.491585473870682e-06, + "loss": 1.192, + "step": 621 + }, + { + "epoch": 0.027539488394230874, + "grad_norm": 1.8719121217727661, + "learning_rate": 5.500442869796281e-06, + "loss": 1.0652, + "step": 622 + }, + { + "epoch": 0.027583764099044752, + "grad_norm": 1.72034752368927, + "learning_rate": 5.509300265721878e-06, + "loss": 1.1655, + "step": 623 + }, + { + "epoch": 0.02762803980385863, + "grad_norm": 1.6227370500564575, + "learning_rate": 5.518157661647477e-06, + "loss": 1.1931, + "step": 624 + }, + { + "epoch": 0.027672315508672503, + "grad_norm": 1.60439932346344, + "learning_rate": 5.527015057573074e-06, + "loss": 1.082, + "step": 625 + }, + { + "epoch": 0.02771659121348638, + "grad_norm": 1.657958745956421, + "learning_rate": 5.535872453498672e-06, + "loss": 1.0892, + "step": 626 + }, + { + "epoch": 0.027760866918300257, + "grad_norm": 1.7443915605545044, + "learning_rate": 5.544729849424269e-06, + "loss": 1.1896, + "step": 627 + }, + { + "epoch": 0.02780514262311413, + "grad_norm": 1.6068094968795776, + "learning_rate": 5.553587245349868e-06, + "loss": 1.1118, + "step": 628 + }, + { + "epoch": 0.02784941832792801, + "grad_norm": 1.7423769235610962, + "learning_rate": 5.562444641275465e-06, + "loss": 1.1617, + "step": 629 + }, + { + "epoch": 0.027893694032741882, + "grad_norm": 1.8724569082260132, + "learning_rate": 5.571302037201063e-06, + "loss": 1.1759, + "step": 630 + }, + { + "epoch": 0.02793796973755576, + "grad_norm": 1.656651258468628, + "learning_rate": 5.580159433126662e-06, + "loss": 1.1345, + "step": 631 + }, + { + "epoch": 0.027982245442369637, + "grad_norm": 2.7768354415893555, + "learning_rate": 5.5890168290522595e-06, + "loss": 1.167, + "step": 632 + }, + { + "epoch": 0.02802652114718351, + "grad_norm": 1.5789172649383545, + "learning_rate": 5.597874224977857e-06, + "loss": 1.1887, + "step": 633 + }, + { + "epoch": 0.028070796851997388, + "grad_norm": 1.6310153007507324, + "learning_rate": 5.606731620903455e-06, + "loss": 1.0734, + "step": 634 + }, + { + "epoch": 0.028115072556811265, + "grad_norm": 1.731374740600586, + "learning_rate": 5.615589016829053e-06, + "loss": 1.21, + "step": 635 + }, + { + "epoch": 0.02815934826162514, + "grad_norm": 1.6740349531173706, + "learning_rate": 5.6244464127546506e-06, + "loss": 1.0426, + "step": 636 + }, + { + "epoch": 0.028203623966439016, + "grad_norm": 1.6519548892974854, + "learning_rate": 5.633303808680248e-06, + "loss": 1.1867, + "step": 637 + }, + { + "epoch": 0.028247899671252893, + "grad_norm": 1.566980242729187, + "learning_rate": 5.642161204605846e-06, + "loss": 1.0952, + "step": 638 + }, + { + "epoch": 0.028292175376066767, + "grad_norm": 1.627239465713501, + "learning_rate": 5.651018600531444e-06, + "loss": 1.0875, + "step": 639 + }, + { + "epoch": 0.028336451080880645, + "grad_norm": 2.0019185543060303, + "learning_rate": 5.659875996457042e-06, + "loss": 1.222, + "step": 640 + }, + { + "epoch": 0.02838072678569452, + "grad_norm": 1.583670973777771, + "learning_rate": 5.668733392382641e-06, + "loss": 1.1599, + "step": 641 + }, + { + "epoch": 0.028425002490508396, + "grad_norm": 1.570052981376648, + "learning_rate": 5.677590788308238e-06, + "loss": 1.1068, + "step": 642 + }, + { + "epoch": 0.028469278195322273, + "grad_norm": 1.696564793586731, + "learning_rate": 5.686448184233836e-06, + "loss": 1.1889, + "step": 643 + }, + { + "epoch": 0.028513553900136147, + "grad_norm": 1.6174219846725464, + "learning_rate": 5.6953055801594335e-06, + "loss": 1.155, + "step": 644 + }, + { + "epoch": 0.028557829604950024, + "grad_norm": 1.7784713506698608, + "learning_rate": 5.704162976085032e-06, + "loss": 1.1324, + "step": 645 + }, + { + "epoch": 0.0286021053097639, + "grad_norm": 2.165430784225464, + "learning_rate": 5.713020372010629e-06, + "loss": 1.234, + "step": 646 + }, + { + "epoch": 0.028646381014577775, + "grad_norm": 1.5870777368545532, + "learning_rate": 5.721877767936227e-06, + "loss": 1.1109, + "step": 647 + }, + { + "epoch": 0.028690656719391652, + "grad_norm": 1.5381951332092285, + "learning_rate": 5.7307351638618245e-06, + "loss": 1.0957, + "step": 648 + }, + { + "epoch": 0.028734932424205526, + "grad_norm": 1.645394206047058, + "learning_rate": 5.739592559787423e-06, + "loss": 1.1653, + "step": 649 + }, + { + "epoch": 0.028779208129019403, + "grad_norm": 1.5593795776367188, + "learning_rate": 5.748449955713021e-06, + "loss": 1.1125, + "step": 650 + }, + { + "epoch": 0.02882348383383328, + "grad_norm": 2.7782111167907715, + "learning_rate": 5.757307351638619e-06, + "loss": 1.1823, + "step": 651 + }, + { + "epoch": 0.028867759538647154, + "grad_norm": 1.602355718612671, + "learning_rate": 5.766164747564217e-06, + "loss": 1.1131, + "step": 652 + }, + { + "epoch": 0.02891203524346103, + "grad_norm": 1.828336477279663, + "learning_rate": 5.775022143489815e-06, + "loss": 1.1872, + "step": 653 + }, + { + "epoch": 0.02895631094827491, + "grad_norm": 1.6302828788757324, + "learning_rate": 5.783879539415412e-06, + "loss": 1.158, + "step": 654 + }, + { + "epoch": 0.029000586653088783, + "grad_norm": 1.5007752180099487, + "learning_rate": 5.79273693534101e-06, + "loss": 1.0801, + "step": 655 + }, + { + "epoch": 0.02904486235790266, + "grad_norm": 1.5570439100265503, + "learning_rate": 5.801594331266608e-06, + "loss": 1.1582, + "step": 656 + }, + { + "epoch": 0.029089138062716537, + "grad_norm": 1.6531339883804321, + "learning_rate": 5.810451727192206e-06, + "loss": 1.2206, + "step": 657 + }, + { + "epoch": 0.02913341376753041, + "grad_norm": 1.666367769241333, + "learning_rate": 5.819309123117803e-06, + "loss": 1.1311, + "step": 658 + }, + { + "epoch": 0.02917768947234429, + "grad_norm": 2.25840163230896, + "learning_rate": 5.8281665190434025e-06, + "loss": 1.2068, + "step": 659 + }, + { + "epoch": 0.029221965177158162, + "grad_norm": 1.5688217878341675, + "learning_rate": 5.837023914969e-06, + "loss": 1.1051, + "step": 660 + }, + { + "epoch": 0.02926624088197204, + "grad_norm": 1.9001628160476685, + "learning_rate": 5.845881310894598e-06, + "loss": 1.0531, + "step": 661 + }, + { + "epoch": 0.029310516586785917, + "grad_norm": 2.0396804809570312, + "learning_rate": 5.854738706820195e-06, + "loss": 1.1288, + "step": 662 + }, + { + "epoch": 0.02935479229159979, + "grad_norm": 1.828851580619812, + "learning_rate": 5.8635961027457936e-06, + "loss": 1.1675, + "step": 663 + }, + { + "epoch": 0.029399067996413668, + "grad_norm": 1.7529685497283936, + "learning_rate": 5.872453498671391e-06, + "loss": 1.1451, + "step": 664 + }, + { + "epoch": 0.029443343701227545, + "grad_norm": 1.7610931396484375, + "learning_rate": 5.881310894596989e-06, + "loss": 1.0785, + "step": 665 + }, + { + "epoch": 0.02948761940604142, + "grad_norm": 1.7451332807540894, + "learning_rate": 5.890168290522586e-06, + "loss": 1.1113, + "step": 666 + }, + { + "epoch": 0.029531895110855296, + "grad_norm": 1.7041691541671753, + "learning_rate": 5.899025686448185e-06, + "loss": 1.1231, + "step": 667 + }, + { + "epoch": 0.029576170815669173, + "grad_norm": 1.5389046669006348, + "learning_rate": 5.907883082373782e-06, + "loss": 1.1428, + "step": 668 + }, + { + "epoch": 0.029620446520483047, + "grad_norm": 1.9029959440231323, + "learning_rate": 5.9167404782993805e-06, + "loss": 1.2224, + "step": 669 + }, + { + "epoch": 0.029664722225296924, + "grad_norm": 2.1075258255004883, + "learning_rate": 5.925597874224979e-06, + "loss": 1.2297, + "step": 670 + }, + { + "epoch": 0.0297089979301108, + "grad_norm": 1.758933424949646, + "learning_rate": 5.9344552701505765e-06, + "loss": 1.1701, + "step": 671 + }, + { + "epoch": 0.029753273634924676, + "grad_norm": 1.9380658864974976, + "learning_rate": 5.943312666076174e-06, + "loss": 1.1666, + "step": 672 + }, + { + "epoch": 0.029797549339738553, + "grad_norm": 1.754888892173767, + "learning_rate": 5.9521700620017716e-06, + "loss": 1.2571, + "step": 673 + }, + { + "epoch": 0.029841825044552427, + "grad_norm": 1.7916693687438965, + "learning_rate": 5.96102745792737e-06, + "loss": 1.1223, + "step": 674 + }, + { + "epoch": 0.029886100749366304, + "grad_norm": 2.278717517852783, + "learning_rate": 5.9698848538529675e-06, + "loss": 1.2196, + "step": 675 + }, + { + "epoch": 0.02993037645418018, + "grad_norm": 1.7907987833023071, + "learning_rate": 5.978742249778565e-06, + "loss": 1.0847, + "step": 676 + }, + { + "epoch": 0.029974652158994055, + "grad_norm": 1.9400681257247925, + "learning_rate": 5.9875996457041626e-06, + "loss": 1.2161, + "step": 677 + }, + { + "epoch": 0.030018927863807932, + "grad_norm": 1.6480571031570435, + "learning_rate": 5.996457041629762e-06, + "loss": 1.1431, + "step": 678 + }, + { + "epoch": 0.03006320356862181, + "grad_norm": 1.823978304862976, + "learning_rate": 6.005314437555359e-06, + "loss": 1.1872, + "step": 679 + }, + { + "epoch": 0.030107479273435683, + "grad_norm": 1.7976018190383911, + "learning_rate": 6.014171833480958e-06, + "loss": 1.2208, + "step": 680 + }, + { + "epoch": 0.03015175497824956, + "grad_norm": 1.7784433364868164, + "learning_rate": 6.023029229406555e-06, + "loss": 1.2013, + "step": 681 + }, + { + "epoch": 0.030196030683063434, + "grad_norm": 1.729435682296753, + "learning_rate": 6.031886625332153e-06, + "loss": 1.129, + "step": 682 + }, + { + "epoch": 0.03024030638787731, + "grad_norm": 2.107863426208496, + "learning_rate": 6.04074402125775e-06, + "loss": 1.1813, + "step": 683 + }, + { + "epoch": 0.03028458209269119, + "grad_norm": 2.7395036220550537, + "learning_rate": 6.049601417183349e-06, + "loss": 1.2303, + "step": 684 + }, + { + "epoch": 0.030328857797505063, + "grad_norm": 1.6568154096603394, + "learning_rate": 6.058458813108946e-06, + "loss": 1.0785, + "step": 685 + }, + { + "epoch": 0.03037313350231894, + "grad_norm": 1.613080382347107, + "learning_rate": 6.067316209034544e-06, + "loss": 1.1598, + "step": 686 + }, + { + "epoch": 0.030417409207132817, + "grad_norm": 1.4660946130752563, + "learning_rate": 6.076173604960143e-06, + "loss": 1.0191, + "step": 687 + }, + { + "epoch": 0.03046168491194669, + "grad_norm": 1.6131199598312378, + "learning_rate": 6.085031000885741e-06, + "loss": 1.1349, + "step": 688 + }, + { + "epoch": 0.03050596061676057, + "grad_norm": 1.6955361366271973, + "learning_rate": 6.093888396811338e-06, + "loss": 1.1516, + "step": 689 + }, + { + "epoch": 0.030550236321574446, + "grad_norm": 1.5831762552261353, + "learning_rate": 6.102745792736936e-06, + "loss": 1.113, + "step": 690 + }, + { + "epoch": 0.03059451202638832, + "grad_norm": 1.5975542068481445, + "learning_rate": 6.111603188662534e-06, + "loss": 1.1067, + "step": 691 + }, + { + "epoch": 0.030638787731202197, + "grad_norm": 1.6597403287887573, + "learning_rate": 6.120460584588132e-06, + "loss": 1.1447, + "step": 692 + }, + { + "epoch": 0.03068306343601607, + "grad_norm": 1.7059966325759888, + "learning_rate": 6.129317980513729e-06, + "loss": 1.0824, + "step": 693 + }, + { + "epoch": 0.030727339140829948, + "grad_norm": 1.6312583684921265, + "learning_rate": 6.138175376439327e-06, + "loss": 1.1101, + "step": 694 + }, + { + "epoch": 0.030771614845643825, + "grad_norm": 2.040283679962158, + "learning_rate": 6.147032772364925e-06, + "loss": 1.1762, + "step": 695 + }, + { + "epoch": 0.0308158905504577, + "grad_norm": 1.52326238155365, + "learning_rate": 6.155890168290523e-06, + "loss": 1.112, + "step": 696 + }, + { + "epoch": 0.030860166255271576, + "grad_norm": 1.842525601387024, + "learning_rate": 6.164747564216121e-06, + "loss": 1.2256, + "step": 697 + }, + { + "epoch": 0.030904441960085453, + "grad_norm": 1.6848838329315186, + "learning_rate": 6.1736049601417195e-06, + "loss": 1.1408, + "step": 698 + }, + { + "epoch": 0.030948717664899327, + "grad_norm": 2.0569376945495605, + "learning_rate": 6.182462356067317e-06, + "loss": 1.1869, + "step": 699 + }, + { + "epoch": 0.030992993369713204, + "grad_norm": 1.6861035823822021, + "learning_rate": 6.1913197519929146e-06, + "loss": 1.1604, + "step": 700 + }, + { + "epoch": 0.03103726907452708, + "grad_norm": 1.8258110284805298, + "learning_rate": 6.200177147918512e-06, + "loss": 1.1888, + "step": 701 + }, + { + "epoch": 0.031081544779340955, + "grad_norm": 2.0443601608276367, + "learning_rate": 6.2090345438441105e-06, + "loss": 1.2575, + "step": 702 + }, + { + "epoch": 0.031125820484154833, + "grad_norm": 1.5925978422164917, + "learning_rate": 6.217891939769708e-06, + "loss": 1.2167, + "step": 703 + }, + { + "epoch": 0.031170096188968707, + "grad_norm": 1.8850388526916504, + "learning_rate": 6.226749335695306e-06, + "loss": 1.2084, + "step": 704 + }, + { + "epoch": 0.031214371893782584, + "grad_norm": 1.881131887435913, + "learning_rate": 6.235606731620903e-06, + "loss": 1.2048, + "step": 705 + }, + { + "epoch": 0.03125864759859646, + "grad_norm": 1.6413209438323975, + "learning_rate": 6.244464127546502e-06, + "loss": 1.0988, + "step": 706 + }, + { + "epoch": 0.031302923303410335, + "grad_norm": 1.6994318962097168, + "learning_rate": 6.2533215234721e-06, + "loss": 1.118, + "step": 707 + }, + { + "epoch": 0.03134719900822421, + "grad_norm": 1.8340460062026978, + "learning_rate": 6.2621789193976975e-06, + "loss": 1.221, + "step": 708 + }, + { + "epoch": 0.03139147471303809, + "grad_norm": 1.8107188940048218, + "learning_rate": 6.271036315323296e-06, + "loss": 1.0701, + "step": 709 + }, + { + "epoch": 0.03143575041785197, + "grad_norm": 1.649435043334961, + "learning_rate": 6.279893711248893e-06, + "loss": 1.1931, + "step": 710 + }, + { + "epoch": 0.03148002612266584, + "grad_norm": 1.9529467821121216, + "learning_rate": 6.288751107174491e-06, + "loss": 1.2797, + "step": 711 + }, + { + "epoch": 0.031524301827479714, + "grad_norm": 1.7687151432037354, + "learning_rate": 6.2976085031000885e-06, + "loss": 1.2081, + "step": 712 + }, + { + "epoch": 0.03156857753229359, + "grad_norm": 1.57136070728302, + "learning_rate": 6.306465899025687e-06, + "loss": 1.115, + "step": 713 + }, + { + "epoch": 0.03161285323710747, + "grad_norm": 1.971044659614563, + "learning_rate": 6.315323294951284e-06, + "loss": 1.243, + "step": 714 + }, + { + "epoch": 0.031657128941921346, + "grad_norm": 1.624457597732544, + "learning_rate": 6.324180690876884e-06, + "loss": 1.1033, + "step": 715 + }, + { + "epoch": 0.03170140464673522, + "grad_norm": 1.5642496347427368, + "learning_rate": 6.333038086802481e-06, + "loss": 1.1128, + "step": 716 + }, + { + "epoch": 0.031745680351549094, + "grad_norm": 1.8848638534545898, + "learning_rate": 6.341895482728079e-06, + "loss": 1.1636, + "step": 717 + }, + { + "epoch": 0.03178995605636297, + "grad_norm": 1.7665469646453857, + "learning_rate": 6.350752878653676e-06, + "loss": 1.1828, + "step": 718 + }, + { + "epoch": 0.03183423176117685, + "grad_norm": 1.8649015426635742, + "learning_rate": 6.359610274579275e-06, + "loss": 1.2366, + "step": 719 + }, + { + "epoch": 0.031878507465990726, + "grad_norm": 1.4919250011444092, + "learning_rate": 6.368467670504872e-06, + "loss": 1.0686, + "step": 720 + }, + { + "epoch": 0.0319227831708046, + "grad_norm": 2.151181697845459, + "learning_rate": 6.37732506643047e-06, + "loss": 1.2632, + "step": 721 + }, + { + "epoch": 0.03196705887561847, + "grad_norm": 1.6256765127182007, + "learning_rate": 6.386182462356067e-06, + "loss": 1.0632, + "step": 722 + }, + { + "epoch": 0.03201133458043235, + "grad_norm": 1.8058379888534546, + "learning_rate": 6.395039858281666e-06, + "loss": 1.1408, + "step": 723 + }, + { + "epoch": 0.03205561028524623, + "grad_norm": 1.7943658828735352, + "learning_rate": 6.403897254207263e-06, + "loss": 1.119, + "step": 724 + }, + { + "epoch": 0.032099885990060105, + "grad_norm": 1.8951923847198486, + "learning_rate": 6.412754650132862e-06, + "loss": 1.1968, + "step": 725 + }, + { + "epoch": 0.03214416169487398, + "grad_norm": 2.011754274368286, + "learning_rate": 6.42161204605846e-06, + "loss": 1.2337, + "step": 726 + }, + { + "epoch": 0.03218843739968786, + "grad_norm": 1.7250877618789673, + "learning_rate": 6.4304694419840576e-06, + "loss": 1.2234, + "step": 727 + }, + { + "epoch": 0.03223271310450173, + "grad_norm": 1.517595648765564, + "learning_rate": 6.439326837909655e-06, + "loss": 1.0529, + "step": 728 + }, + { + "epoch": 0.03227698880931561, + "grad_norm": 1.8736807107925415, + "learning_rate": 6.448184233835253e-06, + "loss": 1.1545, + "step": 729 + }, + { + "epoch": 0.032321264514129484, + "grad_norm": 1.478986382484436, + "learning_rate": 6.457041629760851e-06, + "loss": 1.0382, + "step": 730 + }, + { + "epoch": 0.03236554021894336, + "grad_norm": 1.6561537981033325, + "learning_rate": 6.465899025686449e-06, + "loss": 1.1354, + "step": 731 + }, + { + "epoch": 0.03240981592375724, + "grad_norm": 1.4673439264297485, + "learning_rate": 6.474756421612046e-06, + "loss": 1.1074, + "step": 732 + }, + { + "epoch": 0.03245409162857111, + "grad_norm": 1.8616602420806885, + "learning_rate": 6.483613817537644e-06, + "loss": 1.131, + "step": 733 + }, + { + "epoch": 0.032498367333384987, + "grad_norm": 1.7823584079742432, + "learning_rate": 6.492471213463243e-06, + "loss": 1.1506, + "step": 734 + }, + { + "epoch": 0.032542643038198864, + "grad_norm": 2.3428730964660645, + "learning_rate": 6.5013286093888405e-06, + "loss": 1.2139, + "step": 735 + }, + { + "epoch": 0.03258691874301274, + "grad_norm": 1.7022846937179565, + "learning_rate": 6.510186005314438e-06, + "loss": 1.119, + "step": 736 + }, + { + "epoch": 0.03263119444782662, + "grad_norm": 1.6553999185562134, + "learning_rate": 6.519043401240036e-06, + "loss": 1.0527, + "step": 737 + }, + { + "epoch": 0.032675470152640496, + "grad_norm": 1.6417768001556396, + "learning_rate": 6.527900797165634e-06, + "loss": 1.1135, + "step": 738 + }, + { + "epoch": 0.032719745857454366, + "grad_norm": 1.7112751007080078, + "learning_rate": 6.5367581930912315e-06, + "loss": 1.2122, + "step": 739 + }, + { + "epoch": 0.03276402156226824, + "grad_norm": 1.607801914215088, + "learning_rate": 6.545615589016829e-06, + "loss": 1.141, + "step": 740 + }, + { + "epoch": 0.03280829726708212, + "grad_norm": 1.68310546875, + "learning_rate": 6.554472984942427e-06, + "loss": 1.1209, + "step": 741 + }, + { + "epoch": 0.032852572971896, + "grad_norm": 1.6347533464431763, + "learning_rate": 6.563330380868025e-06, + "loss": 1.2155, + "step": 742 + }, + { + "epoch": 0.032896848676709875, + "grad_norm": 1.5190021991729736, + "learning_rate": 6.572187776793623e-06, + "loss": 1.082, + "step": 743 + }, + { + "epoch": 0.032941124381523745, + "grad_norm": 1.7886120080947876, + "learning_rate": 6.581045172719222e-06, + "loss": 1.1604, + "step": 744 + }, + { + "epoch": 0.03298540008633762, + "grad_norm": 1.7235965728759766, + "learning_rate": 6.589902568644819e-06, + "loss": 1.2094, + "step": 745 + }, + { + "epoch": 0.0330296757911515, + "grad_norm": 1.832850456237793, + "learning_rate": 6.598759964570417e-06, + "loss": 1.1657, + "step": 746 + }, + { + "epoch": 0.03307395149596538, + "grad_norm": 1.775771141052246, + "learning_rate": 6.607617360496014e-06, + "loss": 1.1432, + "step": 747 + }, + { + "epoch": 0.033118227200779254, + "grad_norm": 1.4390367269515991, + "learning_rate": 6.616474756421613e-06, + "loss": 1.1365, + "step": 748 + }, + { + "epoch": 0.03316250290559313, + "grad_norm": 1.5875095129013062, + "learning_rate": 6.62533215234721e-06, + "loss": 1.1217, + "step": 749 + }, + { + "epoch": 0.033206778610407, + "grad_norm": 1.7963145971298218, + "learning_rate": 6.634189548272808e-06, + "loss": 1.1101, + "step": 750 + }, + { + "epoch": 0.03325105431522088, + "grad_norm": 1.4824106693267822, + "learning_rate": 6.643046944198405e-06, + "loss": 1.0684, + "step": 751 + }, + { + "epoch": 0.03329533002003476, + "grad_norm": 1.708621859550476, + "learning_rate": 6.651904340124004e-06, + "loss": 1.0855, + "step": 752 + }, + { + "epoch": 0.033339605724848634, + "grad_norm": 1.7086677551269531, + "learning_rate": 6.660761736049602e-06, + "loss": 1.1365, + "step": 753 + }, + { + "epoch": 0.03338388142966251, + "grad_norm": 1.582942008972168, + "learning_rate": 6.6696191319752006e-06, + "loss": 1.1031, + "step": 754 + }, + { + "epoch": 0.03342815713447638, + "grad_norm": 1.5850480794906616, + "learning_rate": 6.678476527900798e-06, + "loss": 1.1402, + "step": 755 + }, + { + "epoch": 0.03347243283929026, + "grad_norm": 1.553131341934204, + "learning_rate": 6.687333923826396e-06, + "loss": 1.0861, + "step": 756 + }, + { + "epoch": 0.033516708544104136, + "grad_norm": 1.7436864376068115, + "learning_rate": 6.696191319751993e-06, + "loss": 1.1728, + "step": 757 + }, + { + "epoch": 0.03356098424891801, + "grad_norm": 1.7256935834884644, + "learning_rate": 6.705048715677592e-06, + "loss": 1.1378, + "step": 758 + }, + { + "epoch": 0.03360525995373189, + "grad_norm": 1.5427135229110718, + "learning_rate": 6.713906111603189e-06, + "loss": 1.1308, + "step": 759 + }, + { + "epoch": 0.03364953565854577, + "grad_norm": 1.8952510356903076, + "learning_rate": 6.722763507528787e-06, + "loss": 1.1482, + "step": 760 + }, + { + "epoch": 0.03369381136335964, + "grad_norm": 2.1173293590545654, + "learning_rate": 6.731620903454384e-06, + "loss": 1.0945, + "step": 761 + }, + { + "epoch": 0.033738087068173515, + "grad_norm": 1.662434458732605, + "learning_rate": 6.7404782993799835e-06, + "loss": 1.2304, + "step": 762 + }, + { + "epoch": 0.03378236277298739, + "grad_norm": 1.841312289237976, + "learning_rate": 6.749335695305581e-06, + "loss": 1.1917, + "step": 763 + }, + { + "epoch": 0.03382663847780127, + "grad_norm": 1.8312627077102661, + "learning_rate": 6.7581930912311786e-06, + "loss": 1.1475, + "step": 764 + }, + { + "epoch": 0.03387091418261515, + "grad_norm": 1.6673256158828735, + "learning_rate": 6.767050487156777e-06, + "loss": 1.1471, + "step": 765 + }, + { + "epoch": 0.03391518988742902, + "grad_norm": 1.8188673257827759, + "learning_rate": 6.7759078830823745e-06, + "loss": 1.1675, + "step": 766 + }, + { + "epoch": 0.033959465592242895, + "grad_norm": 1.7568005323410034, + "learning_rate": 6.784765279007972e-06, + "loss": 1.1619, + "step": 767 + }, + { + "epoch": 0.03400374129705677, + "grad_norm": 1.6594741344451904, + "learning_rate": 6.79362267493357e-06, + "loss": 1.1896, + "step": 768 + }, + { + "epoch": 0.03404801700187065, + "grad_norm": 2.091128349304199, + "learning_rate": 6.802480070859168e-06, + "loss": 1.2389, + "step": 769 + }, + { + "epoch": 0.03409229270668453, + "grad_norm": 1.6581106185913086, + "learning_rate": 6.8113374667847655e-06, + "loss": 1.1662, + "step": 770 + }, + { + "epoch": 0.034136568411498404, + "grad_norm": 1.6948705911636353, + "learning_rate": 6.820194862710364e-06, + "loss": 1.1473, + "step": 771 + }, + { + "epoch": 0.034180844116312274, + "grad_norm": 1.8065762519836426, + "learning_rate": 6.829052258635962e-06, + "loss": 1.1849, + "step": 772 + }, + { + "epoch": 0.03422511982112615, + "grad_norm": 1.6458922624588013, + "learning_rate": 6.83790965456156e-06, + "loss": 1.1624, + "step": 773 + }, + { + "epoch": 0.03426939552594003, + "grad_norm": 1.9114571809768677, + "learning_rate": 6.846767050487157e-06, + "loss": 1.2564, + "step": 774 + }, + { + "epoch": 0.034313671230753906, + "grad_norm": 1.6135873794555664, + "learning_rate": 6.855624446412755e-06, + "loss": 1.1343, + "step": 775 + }, + { + "epoch": 0.03435794693556778, + "grad_norm": 1.8533742427825928, + "learning_rate": 6.864481842338353e-06, + "loss": 1.2313, + "step": 776 + }, + { + "epoch": 0.034402222640381654, + "grad_norm": 1.4549700021743774, + "learning_rate": 6.873339238263951e-06, + "loss": 1.1275, + "step": 777 + }, + { + "epoch": 0.03444649834519553, + "grad_norm": 1.5908236503601074, + "learning_rate": 6.882196634189548e-06, + "loss": 1.1318, + "step": 778 + }, + { + "epoch": 0.03449077405000941, + "grad_norm": 1.622499942779541, + "learning_rate": 6.891054030115146e-06, + "loss": 1.0703, + "step": 779 + }, + { + "epoch": 0.034535049754823285, + "grad_norm": 1.6085665225982666, + "learning_rate": 6.899911426040744e-06, + "loss": 1.1115, + "step": 780 + }, + { + "epoch": 0.03457932545963716, + "grad_norm": 1.56098473072052, + "learning_rate": 6.908768821966343e-06, + "loss": 1.068, + "step": 781 + }, + { + "epoch": 0.03462360116445104, + "grad_norm": 1.7623769044876099, + "learning_rate": 6.91762621789194e-06, + "loss": 1.1596, + "step": 782 + }, + { + "epoch": 0.03466787686926491, + "grad_norm": 1.7527482509613037, + "learning_rate": 6.926483613817539e-06, + "loss": 1.1635, + "step": 783 + }, + { + "epoch": 0.03471215257407879, + "grad_norm": 1.4360257387161255, + "learning_rate": 6.935341009743136e-06, + "loss": 1.003, + "step": 784 + }, + { + "epoch": 0.034756428278892665, + "grad_norm": 1.4098846912384033, + "learning_rate": 6.944198405668734e-06, + "loss": 1.0045, + "step": 785 + }, + { + "epoch": 0.03480070398370654, + "grad_norm": 1.6578943729400635, + "learning_rate": 6.953055801594331e-06, + "loss": 1.0959, + "step": 786 + }, + { + "epoch": 0.03484497968852042, + "grad_norm": 1.9090783596038818, + "learning_rate": 6.96191319751993e-06, + "loss": 1.1558, + "step": 787 + }, + { + "epoch": 0.03488925539333429, + "grad_norm": 1.802502155303955, + "learning_rate": 6.970770593445527e-06, + "loss": 1.1624, + "step": 788 + }, + { + "epoch": 0.03493353109814817, + "grad_norm": 1.8612244129180908, + "learning_rate": 6.979627989371125e-06, + "loss": 1.234, + "step": 789 + }, + { + "epoch": 0.034977806802962044, + "grad_norm": 1.6443555355072021, + "learning_rate": 6.988485385296724e-06, + "loss": 1.1115, + "step": 790 + }, + { + "epoch": 0.03502208250777592, + "grad_norm": 1.5775914192199707, + "learning_rate": 6.9973427812223216e-06, + "loss": 1.0834, + "step": 791 + }, + { + "epoch": 0.0350663582125898, + "grad_norm": 1.7555649280548096, + "learning_rate": 7.006200177147919e-06, + "loss": 1.0834, + "step": 792 + }, + { + "epoch": 0.035110633917403676, + "grad_norm": 1.6406118869781494, + "learning_rate": 7.0150575730735175e-06, + "loss": 1.0678, + "step": 793 + }, + { + "epoch": 0.035154909622217546, + "grad_norm": 1.708107352256775, + "learning_rate": 7.023914968999115e-06, + "loss": 1.1623, + "step": 794 + }, + { + "epoch": 0.035199185327031424, + "grad_norm": 1.545183539390564, + "learning_rate": 7.032772364924713e-06, + "loss": 1.0754, + "step": 795 + }, + { + "epoch": 0.0352434610318453, + "grad_norm": 1.6484284400939941, + "learning_rate": 7.04162976085031e-06, + "loss": 1.0838, + "step": 796 + }, + { + "epoch": 0.03528773673665918, + "grad_norm": 1.5137097835540771, + "learning_rate": 7.0504871567759085e-06, + "loss": 1.0881, + "step": 797 + }, + { + "epoch": 0.035332012441473055, + "grad_norm": 1.3171130418777466, + "learning_rate": 7.059344552701506e-06, + "loss": 1.041, + "step": 798 + }, + { + "epoch": 0.035376288146286926, + "grad_norm": 4.1901655197143555, + "learning_rate": 7.0682019486271045e-06, + "loss": 1.2005, + "step": 799 + }, + { + "epoch": 0.0354205638511008, + "grad_norm": 2.1004340648651123, + "learning_rate": 7.077059344552703e-06, + "loss": 1.147, + "step": 800 + }, + { + "epoch": 0.03546483955591468, + "grad_norm": 1.6046795845031738, + "learning_rate": 7.0859167404783e-06, + "loss": 1.0321, + "step": 801 + }, + { + "epoch": 0.03550911526072856, + "grad_norm": 1.8237758874893188, + "learning_rate": 7.094774136403898e-06, + "loss": 1.1546, + "step": 802 + }, + { + "epoch": 0.035553390965542435, + "grad_norm": 1.6234028339385986, + "learning_rate": 7.1036315323294955e-06, + "loss": 1.1039, + "step": 803 + }, + { + "epoch": 0.03559766667035631, + "grad_norm": 1.7846301794052124, + "learning_rate": 7.112488928255094e-06, + "loss": 1.2097, + "step": 804 + }, + { + "epoch": 0.03564194237517018, + "grad_norm": 1.8005298376083374, + "learning_rate": 7.121346324180691e-06, + "loss": 1.0107, + "step": 805 + }, + { + "epoch": 0.03568621807998406, + "grad_norm": 1.5168381929397583, + "learning_rate": 7.130203720106289e-06, + "loss": 1.0738, + "step": 806 + }, + { + "epoch": 0.03573049378479794, + "grad_norm": 1.6737844944000244, + "learning_rate": 7.1390611160318865e-06, + "loss": 1.1378, + "step": 807 + }, + { + "epoch": 0.035774769489611814, + "grad_norm": 2.048915386199951, + "learning_rate": 7.147918511957485e-06, + "loss": 1.2546, + "step": 808 + }, + { + "epoch": 0.03581904519442569, + "grad_norm": 1.6504521369934082, + "learning_rate": 7.156775907883083e-06, + "loss": 1.1072, + "step": 809 + }, + { + "epoch": 0.03586332089923956, + "grad_norm": 1.5785285234451294, + "learning_rate": 7.165633303808681e-06, + "loss": 1.0659, + "step": 810 + }, + { + "epoch": 0.03590759660405344, + "grad_norm": 1.4123185873031616, + "learning_rate": 7.174490699734279e-06, + "loss": 1.0671, + "step": 811 + }, + { + "epoch": 0.035951872308867316, + "grad_norm": 1.8265961408615112, + "learning_rate": 7.183348095659877e-06, + "loss": 1.1353, + "step": 812 + }, + { + "epoch": 0.035996148013681194, + "grad_norm": 1.707667589187622, + "learning_rate": 7.192205491585474e-06, + "loss": 1.1468, + "step": 813 + }, + { + "epoch": 0.03604042371849507, + "grad_norm": 1.7779057025909424, + "learning_rate": 7.201062887511072e-06, + "loss": 1.144, + "step": 814 + }, + { + "epoch": 0.03608469942330895, + "grad_norm": 1.5154073238372803, + "learning_rate": 7.20992028343667e-06, + "loss": 1.05, + "step": 815 + }, + { + "epoch": 0.03612897512812282, + "grad_norm": 1.7434487342834473, + "learning_rate": 7.218777679362268e-06, + "loss": 1.2003, + "step": 816 + }, + { + "epoch": 0.036173250832936696, + "grad_norm": 2.1086318492889404, + "learning_rate": 7.227635075287865e-06, + "loss": 1.1902, + "step": 817 + }, + { + "epoch": 0.03621752653775057, + "grad_norm": 1.5708950757980347, + "learning_rate": 7.2364924712134646e-06, + "loss": 1.1052, + "step": 818 + }, + { + "epoch": 0.03626180224256445, + "grad_norm": 1.5362498760223389, + "learning_rate": 7.245349867139062e-06, + "loss": 1.0417, + "step": 819 + }, + { + "epoch": 0.03630607794737833, + "grad_norm": 1.9115464687347412, + "learning_rate": 7.25420726306466e-06, + "loss": 1.19, + "step": 820 + }, + { + "epoch": 0.0363503536521922, + "grad_norm": 1.8449444770812988, + "learning_rate": 7.263064658990257e-06, + "loss": 1.1834, + "step": 821 + }, + { + "epoch": 0.036394629357006075, + "grad_norm": 1.6221888065338135, + "learning_rate": 7.271922054915856e-06, + "loss": 1.1116, + "step": 822 + }, + { + "epoch": 0.03643890506181995, + "grad_norm": 1.5266578197479248, + "learning_rate": 7.280779450841453e-06, + "loss": 1.0742, + "step": 823 + }, + { + "epoch": 0.03648318076663383, + "grad_norm": 1.8828448057174683, + "learning_rate": 7.289636846767051e-06, + "loss": 1.1653, + "step": 824 + }, + { + "epoch": 0.03652745647144771, + "grad_norm": 1.699203610420227, + "learning_rate": 7.298494242692648e-06, + "loss": 1.1477, + "step": 825 + }, + { + "epoch": 0.03657173217626158, + "grad_norm": 1.4252179861068726, + "learning_rate": 7.307351638618247e-06, + "loss": 1.0132, + "step": 826 + }, + { + "epoch": 0.036616007881075455, + "grad_norm": 1.4963362216949463, + "learning_rate": 7.316209034543845e-06, + "loss": 0.9959, + "step": 827 + }, + { + "epoch": 0.03666028358588933, + "grad_norm": 1.546848177909851, + "learning_rate": 7.325066430469443e-06, + "loss": 1.0155, + "step": 828 + }, + { + "epoch": 0.03670455929070321, + "grad_norm": 1.6812400817871094, + "learning_rate": 7.333923826395041e-06, + "loss": 1.143, + "step": 829 + }, + { + "epoch": 0.036748834995517086, + "grad_norm": 1.4345717430114746, + "learning_rate": 7.3427812223206385e-06, + "loss": 1.0399, + "step": 830 + }, + { + "epoch": 0.036793110700330964, + "grad_norm": 1.4869354963302612, + "learning_rate": 7.351638618246236e-06, + "loss": 1.1248, + "step": 831 + }, + { + "epoch": 0.036837386405144834, + "grad_norm": 2.2546300888061523, + "learning_rate": 7.3604960141718344e-06, + "loss": 1.3121, + "step": 832 + }, + { + "epoch": 0.03688166210995871, + "grad_norm": 1.7609663009643555, + "learning_rate": 7.369353410097432e-06, + "loss": 1.1202, + "step": 833 + }, + { + "epoch": 0.03692593781477259, + "grad_norm": 1.9069947004318237, + "learning_rate": 7.3782108060230295e-06, + "loss": 1.1638, + "step": 834 + }, + { + "epoch": 0.036970213519586466, + "grad_norm": 1.6886301040649414, + "learning_rate": 7.387068201948627e-06, + "loss": 1.1504, + "step": 835 + }, + { + "epoch": 0.03701448922440034, + "grad_norm": 1.7252129316329956, + "learning_rate": 7.3959255978742254e-06, + "loss": 1.1224, + "step": 836 + }, + { + "epoch": 0.037058764929214214, + "grad_norm": 2.1480278968811035, + "learning_rate": 7.404782993799824e-06, + "loss": 1.2348, + "step": 837 + }, + { + "epoch": 0.03710304063402809, + "grad_norm": 1.6203397512435913, + "learning_rate": 7.413640389725421e-06, + "loss": 1.0742, + "step": 838 + }, + { + "epoch": 0.03714731633884197, + "grad_norm": 1.7635447978973389, + "learning_rate": 7.42249778565102e-06, + "loss": 1.1113, + "step": 839 + }, + { + "epoch": 0.037191592043655845, + "grad_norm": 1.7146409749984741, + "learning_rate": 7.431355181576617e-06, + "loss": 1.1171, + "step": 840 + }, + { + "epoch": 0.03723586774846972, + "grad_norm": 1.768918752670288, + "learning_rate": 7.440212577502215e-06, + "loss": 1.0958, + "step": 841 + }, + { + "epoch": 0.0372801434532836, + "grad_norm": 1.6969128847122192, + "learning_rate": 7.449069973427812e-06, + "loss": 1.115, + "step": 842 + }, + { + "epoch": 0.03732441915809747, + "grad_norm": 1.6917189359664917, + "learning_rate": 7.457927369353411e-06, + "loss": 1.1782, + "step": 843 + }, + { + "epoch": 0.03736869486291135, + "grad_norm": 1.7696702480316162, + "learning_rate": 7.466784765279008e-06, + "loss": 1.094, + "step": 844 + }, + { + "epoch": 0.037412970567725225, + "grad_norm": 1.9492146968841553, + "learning_rate": 7.475642161204606e-06, + "loss": 1.1651, + "step": 845 + }, + { + "epoch": 0.0374572462725391, + "grad_norm": 1.633669376373291, + "learning_rate": 7.484499557130205e-06, + "loss": 1.0848, + "step": 846 + }, + { + "epoch": 0.03750152197735298, + "grad_norm": 1.5769129991531372, + "learning_rate": 7.493356953055803e-06, + "loss": 1.0908, + "step": 847 + }, + { + "epoch": 0.03754579768216685, + "grad_norm": 1.5971215963363647, + "learning_rate": 7.5022143489814e-06, + "loss": 1.134, + "step": 848 + }, + { + "epoch": 0.03759007338698073, + "grad_norm": 1.6009196043014526, + "learning_rate": 7.511071744906998e-06, + "loss": 1.0328, + "step": 849 + }, + { + "epoch": 0.037634349091794604, + "grad_norm": 2.226170778274536, + "learning_rate": 7.519929140832596e-06, + "loss": 1.3334, + "step": 850 + }, + { + "epoch": 0.03767862479660848, + "grad_norm": 1.8907485008239746, + "learning_rate": 7.528786536758194e-06, + "loss": 1.2335, + "step": 851 + }, + { + "epoch": 0.03772290050142236, + "grad_norm": 2.1591968536376953, + "learning_rate": 7.537643932683791e-06, + "loss": 1.1016, + "step": 852 + }, + { + "epoch": 0.037767176206236236, + "grad_norm": 1.7320398092269897, + "learning_rate": 7.546501328609389e-06, + "loss": 1.1314, + "step": 853 + }, + { + "epoch": 0.037811451911050106, + "grad_norm": 1.7453187704086304, + "learning_rate": 7.555358724534987e-06, + "loss": 1.1647, + "step": 854 + }, + { + "epoch": 0.037855727615863984, + "grad_norm": 1.8602979183197021, + "learning_rate": 7.5642161204605856e-06, + "loss": 1.2024, + "step": 855 + }, + { + "epoch": 0.03790000332067786, + "grad_norm": 1.643903136253357, + "learning_rate": 7.573073516386183e-06, + "loss": 1.0479, + "step": 856 + }, + { + "epoch": 0.03794427902549174, + "grad_norm": 1.6125599145889282, + "learning_rate": 7.5819309123117815e-06, + "loss": 1.0849, + "step": 857 + }, + { + "epoch": 0.037988554730305615, + "grad_norm": 1.6667433977127075, + "learning_rate": 7.590788308237379e-06, + "loss": 1.1148, + "step": 858 + }, + { + "epoch": 0.038032830435119486, + "grad_norm": 1.4210398197174072, + "learning_rate": 7.599645704162977e-06, + "loss": 1.0506, + "step": 859 + }, + { + "epoch": 0.03807710613993336, + "grad_norm": 1.543408989906311, + "learning_rate": 7.608503100088574e-06, + "loss": 1.107, + "step": 860 + }, + { + "epoch": 0.03812138184474724, + "grad_norm": 1.792748212814331, + "learning_rate": 7.6173604960141725e-06, + "loss": 1.1552, + "step": 861 + }, + { + "epoch": 0.03816565754956112, + "grad_norm": 1.5829110145568848, + "learning_rate": 7.62621789193977e-06, + "loss": 1.1764, + "step": 862 + }, + { + "epoch": 0.038209933254374995, + "grad_norm": 1.9046571254730225, + "learning_rate": 7.635075287865368e-06, + "loss": 1.1938, + "step": 863 + }, + { + "epoch": 0.03825420895918887, + "grad_norm": 1.5468449592590332, + "learning_rate": 7.643932683790966e-06, + "loss": 1.1361, + "step": 864 + }, + { + "epoch": 0.03829848466400274, + "grad_norm": 1.6671394109725952, + "learning_rate": 7.652790079716564e-06, + "loss": 1.088, + "step": 865 + }, + { + "epoch": 0.03834276036881662, + "grad_norm": 1.9581550359725952, + "learning_rate": 7.661647475642163e-06, + "loss": 1.2149, + "step": 866 + }, + { + "epoch": 0.0383870360736305, + "grad_norm": 1.6252148151397705, + "learning_rate": 7.67050487156776e-06, + "loss": 1.0821, + "step": 867 + }, + { + "epoch": 0.038431311778444374, + "grad_norm": 1.5016274452209473, + "learning_rate": 7.679362267493358e-06, + "loss": 1.052, + "step": 868 + }, + { + "epoch": 0.03847558748325825, + "grad_norm": 2.3392996788024902, + "learning_rate": 7.688219663418956e-06, + "loss": 1.21, + "step": 869 + }, + { + "epoch": 0.03851986318807212, + "grad_norm": 2.5496106147766113, + "learning_rate": 7.697077059344553e-06, + "loss": 1.2076, + "step": 870 + }, + { + "epoch": 0.038564138892886, + "grad_norm": 1.811383843421936, + "learning_rate": 7.705934455270151e-06, + "loss": 1.1615, + "step": 871 + }, + { + "epoch": 0.038608414597699876, + "grad_norm": 1.8519790172576904, + "learning_rate": 7.714791851195748e-06, + "loss": 1.1719, + "step": 872 + }, + { + "epoch": 0.038652690302513754, + "grad_norm": 1.634089469909668, + "learning_rate": 7.723649247121346e-06, + "loss": 1.1217, + "step": 873 + }, + { + "epoch": 0.03869696600732763, + "grad_norm": 1.6495182514190674, + "learning_rate": 7.732506643046945e-06, + "loss": 1.1615, + "step": 874 + }, + { + "epoch": 0.03874124171214151, + "grad_norm": 1.626312255859375, + "learning_rate": 7.741364038972543e-06, + "loss": 1.0341, + "step": 875 + }, + { + "epoch": 0.03878551741695538, + "grad_norm": 1.6177154779434204, + "learning_rate": 7.750221434898142e-06, + "loss": 1.0753, + "step": 876 + }, + { + "epoch": 0.038829793121769256, + "grad_norm": 1.8013153076171875, + "learning_rate": 7.759078830823738e-06, + "loss": 1.0653, + "step": 877 + }, + { + "epoch": 0.03887406882658313, + "grad_norm": 1.5994617938995361, + "learning_rate": 7.767936226749337e-06, + "loss": 1.1267, + "step": 878 + }, + { + "epoch": 0.03891834453139701, + "grad_norm": 1.773477554321289, + "learning_rate": 7.776793622674933e-06, + "loss": 1.1583, + "step": 879 + }, + { + "epoch": 0.03896262023621089, + "grad_norm": 2.1177449226379395, + "learning_rate": 7.785651018600532e-06, + "loss": 1.1681, + "step": 880 + }, + { + "epoch": 0.03900689594102476, + "grad_norm": 1.5742871761322021, + "learning_rate": 7.79450841452613e-06, + "loss": 1.1551, + "step": 881 + }, + { + "epoch": 0.039051171645838635, + "grad_norm": 1.522483229637146, + "learning_rate": 7.803365810451727e-06, + "loss": 1.0305, + "step": 882 + }, + { + "epoch": 0.03909544735065251, + "grad_norm": 1.973412036895752, + "learning_rate": 7.812223206377327e-06, + "loss": 1.1024, + "step": 883 + }, + { + "epoch": 0.03913972305546639, + "grad_norm": 2.24088191986084, + "learning_rate": 7.821080602302924e-06, + "loss": 1.2822, + "step": 884 + }, + { + "epoch": 0.03918399876028027, + "grad_norm": 1.9012172222137451, + "learning_rate": 7.829937998228522e-06, + "loss": 1.195, + "step": 885 + }, + { + "epoch": 0.039228274465094144, + "grad_norm": 1.6926007270812988, + "learning_rate": 7.838795394154119e-06, + "loss": 1.0958, + "step": 886 + }, + { + "epoch": 0.039272550169908015, + "grad_norm": 1.7655199766159058, + "learning_rate": 7.847652790079717e-06, + "loss": 1.1881, + "step": 887 + }, + { + "epoch": 0.03931682587472189, + "grad_norm": 2.0679731369018555, + "learning_rate": 7.856510186005316e-06, + "loss": 1.2024, + "step": 888 + }, + { + "epoch": 0.03936110157953577, + "grad_norm": 1.720713496208191, + "learning_rate": 7.865367581930912e-06, + "loss": 1.075, + "step": 889 + }, + { + "epoch": 0.039405377284349646, + "grad_norm": 1.992177128791809, + "learning_rate": 7.87422497785651e-06, + "loss": 1.2275, + "step": 890 + }, + { + "epoch": 0.039449652989163524, + "grad_norm": 1.705773949623108, + "learning_rate": 7.883082373782109e-06, + "loss": 1.1833, + "step": 891 + }, + { + "epoch": 0.039493928693977394, + "grad_norm": 1.4013051986694336, + "learning_rate": 7.891939769707706e-06, + "loss": 1.0286, + "step": 892 + }, + { + "epoch": 0.03953820439879127, + "grad_norm": 1.9554784297943115, + "learning_rate": 7.900797165633304e-06, + "loss": 1.1539, + "step": 893 + }, + { + "epoch": 0.03958248010360515, + "grad_norm": 1.733915090560913, + "learning_rate": 7.909654561558902e-06, + "loss": 1.127, + "step": 894 + }, + { + "epoch": 0.039626755808419026, + "grad_norm": 1.5329583883285522, + "learning_rate": 7.918511957484501e-06, + "loss": 1.0615, + "step": 895 + }, + { + "epoch": 0.0396710315132329, + "grad_norm": 1.7228894233703613, + "learning_rate": 7.927369353410098e-06, + "loss": 1.0839, + "step": 896 + }, + { + "epoch": 0.03971530721804678, + "grad_norm": 1.8403364419937134, + "learning_rate": 7.936226749335696e-06, + "loss": 1.191, + "step": 897 + }, + { + "epoch": 0.03975958292286065, + "grad_norm": 2.694002866744995, + "learning_rate": 7.945084145261294e-06, + "loss": 1.2319, + "step": 898 + }, + { + "epoch": 0.03980385862767453, + "grad_norm": 2.1081998348236084, + "learning_rate": 7.953941541186891e-06, + "loss": 1.159, + "step": 899 + }, + { + "epoch": 0.039848134332488405, + "grad_norm": 1.881282925605774, + "learning_rate": 7.96279893711249e-06, + "loss": 1.1293, + "step": 900 + }, + { + "epoch": 0.03989241003730228, + "grad_norm": 1.6685866117477417, + "learning_rate": 7.971656333038086e-06, + "loss": 1.1047, + "step": 901 + }, + { + "epoch": 0.03993668574211616, + "grad_norm": 1.6202077865600586, + "learning_rate": 7.980513728963686e-06, + "loss": 1.0762, + "step": 902 + }, + { + "epoch": 0.03998096144693003, + "grad_norm": 2.0769121646881104, + "learning_rate": 7.989371124889283e-06, + "loss": 1.1596, + "step": 903 + }, + { + "epoch": 0.04002523715174391, + "grad_norm": 1.9200507402420044, + "learning_rate": 7.998228520814881e-06, + "loss": 1.1791, + "step": 904 + }, + { + "epoch": 0.040069512856557785, + "grad_norm": 1.6332710981369019, + "learning_rate": 8.00708591674048e-06, + "loss": 1.1395, + "step": 905 + }, + { + "epoch": 0.04011378856137166, + "grad_norm": 1.7623854875564575, + "learning_rate": 8.015943312666076e-06, + "loss": 0.992, + "step": 906 + }, + { + "epoch": 0.04015806426618554, + "grad_norm": 1.5796409845352173, + "learning_rate": 8.024800708591675e-06, + "loss": 1.0663, + "step": 907 + }, + { + "epoch": 0.040202339970999416, + "grad_norm": 1.6684315204620361, + "learning_rate": 8.033658104517273e-06, + "loss": 1.1277, + "step": 908 + }, + { + "epoch": 0.04024661567581329, + "grad_norm": 1.690132737159729, + "learning_rate": 8.04251550044287e-06, + "loss": 1.2304, + "step": 909 + }, + { + "epoch": 0.040290891380627164, + "grad_norm": 1.8610390424728394, + "learning_rate": 8.051372896368468e-06, + "loss": 1.1109, + "step": 910 + }, + { + "epoch": 0.04033516708544104, + "grad_norm": 1.5676895380020142, + "learning_rate": 8.060230292294067e-06, + "loss": 1.1243, + "step": 911 + }, + { + "epoch": 0.04037944279025492, + "grad_norm": 1.5735645294189453, + "learning_rate": 8.069087688219665e-06, + "loss": 1.0479, + "step": 912 + }, + { + "epoch": 0.040423718495068796, + "grad_norm": 1.8131511211395264, + "learning_rate": 8.077945084145262e-06, + "loss": 1.1498, + "step": 913 + }, + { + "epoch": 0.040467994199882666, + "grad_norm": 1.8414535522460938, + "learning_rate": 8.08680248007086e-06, + "loss": 1.2446, + "step": 914 + }, + { + "epoch": 0.04051226990469654, + "grad_norm": 1.6382564306259155, + "learning_rate": 8.095659875996459e-06, + "loss": 1.0992, + "step": 915 + }, + { + "epoch": 0.04055654560951042, + "grad_norm": 1.8874114751815796, + "learning_rate": 8.104517271922055e-06, + "loss": 1.2234, + "step": 916 + }, + { + "epoch": 0.0406008213143243, + "grad_norm": 1.8353627920150757, + "learning_rate": 8.113374667847654e-06, + "loss": 1.2367, + "step": 917 + }, + { + "epoch": 0.040645097019138175, + "grad_norm": 1.591884732246399, + "learning_rate": 8.12223206377325e-06, + "loss": 1.1317, + "step": 918 + }, + { + "epoch": 0.04068937272395205, + "grad_norm": 2.04508638381958, + "learning_rate": 8.131089459698849e-06, + "loss": 1.1557, + "step": 919 + }, + { + "epoch": 0.04073364842876592, + "grad_norm": 1.785007119178772, + "learning_rate": 8.139946855624447e-06, + "loss": 1.1713, + "step": 920 + }, + { + "epoch": 0.0407779241335798, + "grad_norm": 1.732711911201477, + "learning_rate": 8.148804251550045e-06, + "loss": 1.0315, + "step": 921 + }, + { + "epoch": 0.04082219983839368, + "grad_norm": 1.9773125648498535, + "learning_rate": 8.157661647475644e-06, + "loss": 1.036, + "step": 922 + }, + { + "epoch": 0.040866475543207555, + "grad_norm": 1.745276689529419, + "learning_rate": 8.16651904340124e-06, + "loss": 1.1, + "step": 923 + }, + { + "epoch": 0.04091075124802143, + "grad_norm": 1.8955364227294922, + "learning_rate": 8.175376439326839e-06, + "loss": 1.1542, + "step": 924 + }, + { + "epoch": 0.0409550269528353, + "grad_norm": 1.7107466459274292, + "learning_rate": 8.184233835252436e-06, + "loss": 1.1447, + "step": 925 + }, + { + "epoch": 0.04099930265764918, + "grad_norm": 1.7140098810195923, + "learning_rate": 8.193091231178034e-06, + "loss": 1.1599, + "step": 926 + }, + { + "epoch": 0.04104357836246306, + "grad_norm": 1.9822450876235962, + "learning_rate": 8.201948627103632e-06, + "loss": 1.072, + "step": 927 + }, + { + "epoch": 0.041087854067276934, + "grad_norm": 2.0778377056121826, + "learning_rate": 8.210806023029229e-06, + "loss": 1.1973, + "step": 928 + }, + { + "epoch": 0.04113212977209081, + "grad_norm": 1.5389351844787598, + "learning_rate": 8.219663418954828e-06, + "loss": 1.1016, + "step": 929 + }, + { + "epoch": 0.04117640547690469, + "grad_norm": 1.9504331350326538, + "learning_rate": 8.228520814880426e-06, + "loss": 1.1759, + "step": 930 + }, + { + "epoch": 0.04122068118171856, + "grad_norm": 1.7282137870788574, + "learning_rate": 8.237378210806024e-06, + "loss": 1.0649, + "step": 931 + }, + { + "epoch": 0.041264956886532436, + "grad_norm": 1.7416722774505615, + "learning_rate": 8.246235606731621e-06, + "loss": 1.0661, + "step": 932 + }, + { + "epoch": 0.041309232591346313, + "grad_norm": 1.9226926565170288, + "learning_rate": 8.25509300265722e-06, + "loss": 1.1758, + "step": 933 + }, + { + "epoch": 0.04135350829616019, + "grad_norm": 1.69495689868927, + "learning_rate": 8.263950398582818e-06, + "loss": 1.1354, + "step": 934 + }, + { + "epoch": 0.04139778400097407, + "grad_norm": 1.6997085809707642, + "learning_rate": 8.272807794508414e-06, + "loss": 1.1004, + "step": 935 + }, + { + "epoch": 0.04144205970578794, + "grad_norm": 1.3847051858901978, + "learning_rate": 8.281665190434013e-06, + "loss": 1.0279, + "step": 936 + }, + { + "epoch": 0.041486335410601816, + "grad_norm": 2.0148069858551025, + "learning_rate": 8.290522586359611e-06, + "loss": 1.2441, + "step": 937 + }, + { + "epoch": 0.04153061111541569, + "grad_norm": 1.6693779230117798, + "learning_rate": 8.299379982285208e-06, + "loss": 1.068, + "step": 938 + }, + { + "epoch": 0.04157488682022957, + "grad_norm": 1.5138417482376099, + "learning_rate": 8.308237378210808e-06, + "loss": 1.097, + "step": 939 + }, + { + "epoch": 0.04161916252504345, + "grad_norm": 1.5459898710250854, + "learning_rate": 8.317094774136405e-06, + "loss": 1.1743, + "step": 940 + }, + { + "epoch": 0.041663438229857325, + "grad_norm": 1.6904749870300293, + "learning_rate": 8.325952170062003e-06, + "loss": 1.1613, + "step": 941 + }, + { + "epoch": 0.041707713934671195, + "grad_norm": 1.9775515794754028, + "learning_rate": 8.3348095659876e-06, + "loss": 1.1139, + "step": 942 + }, + { + "epoch": 0.04175198963948507, + "grad_norm": 1.7335728406906128, + "learning_rate": 8.343666961913198e-06, + "loss": 1.1059, + "step": 943 + }, + { + "epoch": 0.04179626534429895, + "grad_norm": 1.5076920986175537, + "learning_rate": 8.352524357838797e-06, + "loss": 1.0955, + "step": 944 + }, + { + "epoch": 0.04184054104911283, + "grad_norm": 1.7618298530578613, + "learning_rate": 8.361381753764393e-06, + "loss": 1.0564, + "step": 945 + }, + { + "epoch": 0.041884816753926704, + "grad_norm": 1.742033839225769, + "learning_rate": 8.370239149689992e-06, + "loss": 1.1236, + "step": 946 + }, + { + "epoch": 0.041929092458740574, + "grad_norm": 1.5819302797317505, + "learning_rate": 8.37909654561559e-06, + "loss": 1.1047, + "step": 947 + }, + { + "epoch": 0.04197336816355445, + "grad_norm": 1.7452354431152344, + "learning_rate": 8.387953941541187e-06, + "loss": 1.1293, + "step": 948 + }, + { + "epoch": 0.04201764386836833, + "grad_norm": 1.5259150266647339, + "learning_rate": 8.396811337466785e-06, + "loss": 1.1085, + "step": 949 + }, + { + "epoch": 0.042061919573182206, + "grad_norm": 1.5744397640228271, + "learning_rate": 8.405668733392384e-06, + "loss": 1.1569, + "step": 950 + }, + { + "epoch": 0.042106195277996084, + "grad_norm": 1.6078009605407715, + "learning_rate": 8.414526129317982e-06, + "loss": 1.1529, + "step": 951 + }, + { + "epoch": 0.04215047098280996, + "grad_norm": 1.5979853868484497, + "learning_rate": 8.423383525243579e-06, + "loss": 1.0057, + "step": 952 + }, + { + "epoch": 0.04219474668762383, + "grad_norm": 2.2535688877105713, + "learning_rate": 8.432240921169177e-06, + "loss": 1.1727, + "step": 953 + }, + { + "epoch": 0.04223902239243771, + "grad_norm": 1.5735183954238892, + "learning_rate": 8.441098317094775e-06, + "loss": 1.0652, + "step": 954 + }, + { + "epoch": 0.042283298097251586, + "grad_norm": 1.3908110857009888, + "learning_rate": 8.449955713020372e-06, + "loss": 1.0849, + "step": 955 + }, + { + "epoch": 0.04232757380206546, + "grad_norm": 1.708431363105774, + "learning_rate": 8.45881310894597e-06, + "loss": 1.1197, + "step": 956 + }, + { + "epoch": 0.04237184950687934, + "grad_norm": 1.527013897895813, + "learning_rate": 8.467670504871567e-06, + "loss": 1.0087, + "step": 957 + }, + { + "epoch": 0.04241612521169321, + "grad_norm": 1.8192415237426758, + "learning_rate": 8.476527900797167e-06, + "loss": 1.0939, + "step": 958 + }, + { + "epoch": 0.04246040091650709, + "grad_norm": 1.982506513595581, + "learning_rate": 8.485385296722764e-06, + "loss": 1.192, + "step": 959 + }, + { + "epoch": 0.042504676621320965, + "grad_norm": 1.767072081565857, + "learning_rate": 8.494242692648362e-06, + "loss": 1.1425, + "step": 960 + }, + { + "epoch": 0.04254895232613484, + "grad_norm": 1.77769136428833, + "learning_rate": 8.50310008857396e-06, + "loss": 1.1917, + "step": 961 + }, + { + "epoch": 0.04259322803094872, + "grad_norm": 1.3744169473648071, + "learning_rate": 8.511957484499558e-06, + "loss": 1.0752, + "step": 962 + }, + { + "epoch": 0.0426375037357626, + "grad_norm": 1.7071926593780518, + "learning_rate": 8.520814880425156e-06, + "loss": 1.0967, + "step": 963 + }, + { + "epoch": 0.04268177944057647, + "grad_norm": 1.5087289810180664, + "learning_rate": 8.529672276350753e-06, + "loss": 1.0154, + "step": 964 + }, + { + "epoch": 0.042726055145390344, + "grad_norm": 1.7988173961639404, + "learning_rate": 8.538529672276351e-06, + "loss": 1.1599, + "step": 965 + }, + { + "epoch": 0.04277033085020422, + "grad_norm": 1.8818655014038086, + "learning_rate": 8.54738706820195e-06, + "loss": 1.1675, + "step": 966 + }, + { + "epoch": 0.0428146065550181, + "grad_norm": 1.7343937158584595, + "learning_rate": 8.556244464127548e-06, + "loss": 1.0601, + "step": 967 + }, + { + "epoch": 0.042858882259831976, + "grad_norm": 1.742428183555603, + "learning_rate": 8.565101860053146e-06, + "loss": 1.1413, + "step": 968 + }, + { + "epoch": 0.04290315796464585, + "grad_norm": 1.555366039276123, + "learning_rate": 8.573959255978743e-06, + "loss": 0.9974, + "step": 969 + }, + { + "epoch": 0.042947433669459724, + "grad_norm": 1.5299072265625, + "learning_rate": 8.582816651904341e-06, + "loss": 1.0615, + "step": 970 + }, + { + "epoch": 0.0429917093742736, + "grad_norm": 2.3410110473632812, + "learning_rate": 8.591674047829938e-06, + "loss": 1.145, + "step": 971 + }, + { + "epoch": 0.04303598507908748, + "grad_norm": 1.8610358238220215, + "learning_rate": 8.600531443755536e-06, + "loss": 1.1579, + "step": 972 + }, + { + "epoch": 0.043080260783901356, + "grad_norm": 1.5331027507781982, + "learning_rate": 8.609388839681135e-06, + "loss": 1.1045, + "step": 973 + }, + { + "epoch": 0.04312453648871523, + "grad_norm": 1.5971720218658447, + "learning_rate": 8.618246235606731e-06, + "loss": 1.0362, + "step": 974 + }, + { + "epoch": 0.0431688121935291, + "grad_norm": 1.5251835584640503, + "learning_rate": 8.62710363153233e-06, + "loss": 1.0891, + "step": 975 + }, + { + "epoch": 0.04321308789834298, + "grad_norm": 1.6426037549972534, + "learning_rate": 8.635961027457928e-06, + "loss": 1.1037, + "step": 976 + }, + { + "epoch": 0.04325736360315686, + "grad_norm": 1.5946534872055054, + "learning_rate": 8.644818423383527e-06, + "loss": 1.0872, + "step": 977 + }, + { + "epoch": 0.043301639307970735, + "grad_norm": 1.5519508123397827, + "learning_rate": 8.653675819309125e-06, + "loss": 1.0549, + "step": 978 + }, + { + "epoch": 0.04334591501278461, + "grad_norm": 2.3440420627593994, + "learning_rate": 8.662533215234722e-06, + "loss": 1.2215, + "step": 979 + }, + { + "epoch": 0.04339019071759848, + "grad_norm": 1.59209144115448, + "learning_rate": 8.67139061116032e-06, + "loss": 1.067, + "step": 980 + }, + { + "epoch": 0.04343446642241236, + "grad_norm": 1.793116807937622, + "learning_rate": 8.680248007085917e-06, + "loss": 1.0888, + "step": 981 + }, + { + "epoch": 0.04347874212722624, + "grad_norm": 1.578392744064331, + "learning_rate": 8.689105403011515e-06, + "loss": 1.0136, + "step": 982 + }, + { + "epoch": 0.043523017832040115, + "grad_norm": 1.8147722482681274, + "learning_rate": 8.697962798937114e-06, + "loss": 1.0686, + "step": 983 + }, + { + "epoch": 0.04356729353685399, + "grad_norm": 1.834503412246704, + "learning_rate": 8.70682019486271e-06, + "loss": 1.1043, + "step": 984 + }, + { + "epoch": 0.04361156924166787, + "grad_norm": 1.6375422477722168, + "learning_rate": 8.715677590788309e-06, + "loss": 1.0875, + "step": 985 + }, + { + "epoch": 0.04365584494648174, + "grad_norm": 1.8177845478057861, + "learning_rate": 8.724534986713907e-06, + "loss": 1.1606, + "step": 986 + }, + { + "epoch": 0.04370012065129562, + "grad_norm": 1.917589545249939, + "learning_rate": 8.733392382639505e-06, + "loss": 1.1768, + "step": 987 + }, + { + "epoch": 0.043744396356109494, + "grad_norm": 1.8295601606369019, + "learning_rate": 8.742249778565102e-06, + "loss": 1.1067, + "step": 988 + }, + { + "epoch": 0.04378867206092337, + "grad_norm": 1.456183671951294, + "learning_rate": 8.7511071744907e-06, + "loss": 1.0171, + "step": 989 + }, + { + "epoch": 0.04383294776573725, + "grad_norm": 1.7272255420684814, + "learning_rate": 8.759964570416299e-06, + "loss": 1.1306, + "step": 990 + }, + { + "epoch": 0.04387722347055112, + "grad_norm": 4.452449798583984, + "learning_rate": 8.768821966341896e-06, + "loss": 1.2139, + "step": 991 + }, + { + "epoch": 0.043921499175364996, + "grad_norm": 1.4130980968475342, + "learning_rate": 8.777679362267494e-06, + "loss": 1.0692, + "step": 992 + }, + { + "epoch": 0.04396577488017887, + "grad_norm": 1.499246597290039, + "learning_rate": 8.786536758193092e-06, + "loss": 1.0233, + "step": 993 + }, + { + "epoch": 0.04401005058499275, + "grad_norm": 1.3889321088790894, + "learning_rate": 8.795394154118689e-06, + "loss": 0.998, + "step": 994 + }, + { + "epoch": 0.04405432628980663, + "grad_norm": 1.5749739408493042, + "learning_rate": 8.804251550044287e-06, + "loss": 1.1281, + "step": 995 + }, + { + "epoch": 0.044098601994620505, + "grad_norm": 1.9032800197601318, + "learning_rate": 8.813108945969886e-06, + "loss": 1.1499, + "step": 996 + }, + { + "epoch": 0.044142877699434376, + "grad_norm": 2.0340700149536133, + "learning_rate": 8.821966341895484e-06, + "loss": 1.0983, + "step": 997 + }, + { + "epoch": 0.04418715340424825, + "grad_norm": 1.5069941282272339, + "learning_rate": 8.830823737821081e-06, + "loss": 1.0757, + "step": 998 + }, + { + "epoch": 0.04423142910906213, + "grad_norm": 1.5489205121994019, + "learning_rate": 8.83968113374668e-06, + "loss": 1.0578, + "step": 999 + }, + { + "epoch": 0.04427570481387601, + "grad_norm": 1.6937538385391235, + "learning_rate": 8.848538529672278e-06, + "loss": 1.1135, + "step": 1000 + }, + { + "epoch": 0.044319980518689885, + "grad_norm": 1.5386772155761719, + "learning_rate": 8.857395925597874e-06, + "loss": 1.103, + "step": 1001 + }, + { + "epoch": 0.044364256223503755, + "grad_norm": 1.4657877683639526, + "learning_rate": 8.866253321523473e-06, + "loss": 0.9922, + "step": 1002 + }, + { + "epoch": 0.04440853192831763, + "grad_norm": 1.4601155519485474, + "learning_rate": 8.87511071744907e-06, + "loss": 1.0747, + "step": 1003 + }, + { + "epoch": 0.04445280763313151, + "grad_norm": 1.8504847288131714, + "learning_rate": 8.883968113374668e-06, + "loss": 1.0677, + "step": 1004 + }, + { + "epoch": 0.04449708333794539, + "grad_norm": 2.0895004272460938, + "learning_rate": 8.892825509300266e-06, + "loss": 1.1687, + "step": 1005 + }, + { + "epoch": 0.044541359042759264, + "grad_norm": 1.8223425149917603, + "learning_rate": 8.901682905225865e-06, + "loss": 1.201, + "step": 1006 + }, + { + "epoch": 0.04458563474757314, + "grad_norm": 1.6917310953140259, + "learning_rate": 8.910540301151463e-06, + "loss": 1.071, + "step": 1007 + }, + { + "epoch": 0.04462991045238701, + "grad_norm": 1.609747290611267, + "learning_rate": 8.91939769707706e-06, + "loss": 1.077, + "step": 1008 + }, + { + "epoch": 0.04467418615720089, + "grad_norm": 1.5878063440322876, + "learning_rate": 8.928255093002658e-06, + "loss": 1.0455, + "step": 1009 + }, + { + "epoch": 0.044718461862014766, + "grad_norm": 1.5580931901931763, + "learning_rate": 8.937112488928255e-06, + "loss": 1.1504, + "step": 1010 + }, + { + "epoch": 0.04476273756682864, + "grad_norm": 1.8168679475784302, + "learning_rate": 8.945969884853853e-06, + "loss": 1.1918, + "step": 1011 + }, + { + "epoch": 0.04480701327164252, + "grad_norm": 1.6424872875213623, + "learning_rate": 8.954827280779452e-06, + "loss": 1.1173, + "step": 1012 + }, + { + "epoch": 0.04485128897645639, + "grad_norm": 1.7415355443954468, + "learning_rate": 8.963684676705048e-06, + "loss": 1.1413, + "step": 1013 + }, + { + "epoch": 0.04489556468127027, + "grad_norm": 1.8271437883377075, + "learning_rate": 8.972542072630648e-06, + "loss": 1.2299, + "step": 1014 + }, + { + "epoch": 0.044939840386084146, + "grad_norm": 1.846982479095459, + "learning_rate": 8.981399468556245e-06, + "loss": 1.1146, + "step": 1015 + }, + { + "epoch": 0.04498411609089802, + "grad_norm": 1.37309992313385, + "learning_rate": 8.990256864481844e-06, + "loss": 1.0021, + "step": 1016 + }, + { + "epoch": 0.0450283917957119, + "grad_norm": 1.6263083219528198, + "learning_rate": 8.999114260407442e-06, + "loss": 1.1614, + "step": 1017 + }, + { + "epoch": 0.04507266750052578, + "grad_norm": 1.686315655708313, + "learning_rate": 9.007971656333039e-06, + "loss": 1.1495, + "step": 1018 + }, + { + "epoch": 0.04511694320533965, + "grad_norm": 2.1023387908935547, + "learning_rate": 9.016829052258637e-06, + "loss": 1.1694, + "step": 1019 + }, + { + "epoch": 0.045161218910153525, + "grad_norm": 1.5684252977371216, + "learning_rate": 9.025686448184234e-06, + "loss": 1.0815, + "step": 1020 + }, + { + "epoch": 0.0452054946149674, + "grad_norm": 1.8234024047851562, + "learning_rate": 9.034543844109832e-06, + "loss": 1.1959, + "step": 1021 + }, + { + "epoch": 0.04524977031978128, + "grad_norm": 1.884421944618225, + "learning_rate": 9.04340124003543e-06, + "loss": 1.2637, + "step": 1022 + }, + { + "epoch": 0.04529404602459516, + "grad_norm": 2.2663888931274414, + "learning_rate": 9.052258635961029e-06, + "loss": 1.1275, + "step": 1023 + }, + { + "epoch": 0.04533832172940903, + "grad_norm": 2.333726644515991, + "learning_rate": 9.061116031886627e-06, + "loss": 1.1974, + "step": 1024 + }, + { + "epoch": 0.045382597434222904, + "grad_norm": 1.6251490116119385, + "learning_rate": 9.069973427812224e-06, + "loss": 0.9995, + "step": 1025 + }, + { + "epoch": 0.04542687313903678, + "grad_norm": 1.5893843173980713, + "learning_rate": 9.078830823737822e-06, + "loss": 1.0806, + "step": 1026 + }, + { + "epoch": 0.04547114884385066, + "grad_norm": 1.974560022354126, + "learning_rate": 9.087688219663419e-06, + "loss": 1.1001, + "step": 1027 + }, + { + "epoch": 0.045515424548664536, + "grad_norm": 1.7836982011795044, + "learning_rate": 9.096545615589017e-06, + "loss": 1.0879, + "step": 1028 + }, + { + "epoch": 0.04555970025347841, + "grad_norm": 1.7974681854248047, + "learning_rate": 9.105403011514616e-06, + "loss": 1.2228, + "step": 1029 + }, + { + "epoch": 0.045603975958292284, + "grad_norm": 1.8234459161758423, + "learning_rate": 9.114260407440213e-06, + "loss": 1.2613, + "step": 1030 + }, + { + "epoch": 0.04564825166310616, + "grad_norm": 1.8301727771759033, + "learning_rate": 9.123117803365811e-06, + "loss": 1.125, + "step": 1031 + }, + { + "epoch": 0.04569252736792004, + "grad_norm": 1.6584638357162476, + "learning_rate": 9.13197519929141e-06, + "loss": 1.033, + "step": 1032 + }, + { + "epoch": 0.045736803072733916, + "grad_norm": 1.5760457515716553, + "learning_rate": 9.140832595217008e-06, + "loss": 1.1106, + "step": 1033 + }, + { + "epoch": 0.04578107877754779, + "grad_norm": 1.5128368139266968, + "learning_rate": 9.149689991142604e-06, + "loss": 1.0235, + "step": 1034 + }, + { + "epoch": 0.04582535448236166, + "grad_norm": 1.7392187118530273, + "learning_rate": 9.158547387068203e-06, + "loss": 1.0571, + "step": 1035 + }, + { + "epoch": 0.04586963018717554, + "grad_norm": 1.5536134243011475, + "learning_rate": 9.167404782993801e-06, + "loss": 1.0469, + "step": 1036 + }, + { + "epoch": 0.04591390589198942, + "grad_norm": 1.5320595502853394, + "learning_rate": 9.176262178919398e-06, + "loss": 1.0185, + "step": 1037 + }, + { + "epoch": 0.045958181596803295, + "grad_norm": 1.6949286460876465, + "learning_rate": 9.185119574844996e-06, + "loss": 1.1134, + "step": 1038 + }, + { + "epoch": 0.04600245730161717, + "grad_norm": 1.850196123123169, + "learning_rate": 9.193976970770595e-06, + "loss": 1.0664, + "step": 1039 + }, + { + "epoch": 0.04604673300643105, + "grad_norm": 2.0023422241210938, + "learning_rate": 9.202834366696191e-06, + "loss": 1.2625, + "step": 1040 + }, + { + "epoch": 0.04609100871124492, + "grad_norm": 1.6546980142593384, + "learning_rate": 9.21169176262179e-06, + "loss": 1.1151, + "step": 1041 + }, + { + "epoch": 0.0461352844160588, + "grad_norm": 1.6241612434387207, + "learning_rate": 9.220549158547388e-06, + "loss": 1.0922, + "step": 1042 + }, + { + "epoch": 0.046179560120872674, + "grad_norm": 1.8197897672653198, + "learning_rate": 9.229406554472987e-06, + "loss": 1.1488, + "step": 1043 + }, + { + "epoch": 0.04622383582568655, + "grad_norm": 1.8616544008255005, + "learning_rate": 9.238263950398583e-06, + "loss": 1.1849, + "step": 1044 + }, + { + "epoch": 0.04626811153050043, + "grad_norm": 1.6244852542877197, + "learning_rate": 9.247121346324182e-06, + "loss": 1.1187, + "step": 1045 + }, + { + "epoch": 0.0463123872353143, + "grad_norm": 1.833425760269165, + "learning_rate": 9.25597874224978e-06, + "loss": 1.1635, + "step": 1046 + }, + { + "epoch": 0.04635666294012818, + "grad_norm": 3.005282163619995, + "learning_rate": 9.264836138175377e-06, + "loss": 1.4117, + "step": 1047 + }, + { + "epoch": 0.046400938644942054, + "grad_norm": 1.8353551626205444, + "learning_rate": 9.273693534100975e-06, + "loss": 1.1674, + "step": 1048 + }, + { + "epoch": 0.04644521434975593, + "grad_norm": 1.5953737497329712, + "learning_rate": 9.282550930026572e-06, + "loss": 1.0994, + "step": 1049 + }, + { + "epoch": 0.04648949005456981, + "grad_norm": 1.5819287300109863, + "learning_rate": 9.29140832595217e-06, + "loss": 1.0535, + "step": 1050 + }, + { + "epoch": 0.046533765759383686, + "grad_norm": 1.7258168458938599, + "learning_rate": 9.300265721877769e-06, + "loss": 1.1261, + "step": 1051 + }, + { + "epoch": 0.046578041464197556, + "grad_norm": 1.7245757579803467, + "learning_rate": 9.309123117803367e-06, + "loss": 1.1346, + "step": 1052 + }, + { + "epoch": 0.04662231716901143, + "grad_norm": 1.6432868242263794, + "learning_rate": 9.317980513728965e-06, + "loss": 1.0852, + "step": 1053 + }, + { + "epoch": 0.04666659287382531, + "grad_norm": 1.8461171388626099, + "learning_rate": 9.326837909654562e-06, + "loss": 1.1111, + "step": 1054 + }, + { + "epoch": 0.04671086857863919, + "grad_norm": 1.5253949165344238, + "learning_rate": 9.33569530558016e-06, + "loss": 1.0724, + "step": 1055 + }, + { + "epoch": 0.046755144283453065, + "grad_norm": 1.684975028038025, + "learning_rate": 9.344552701505759e-06, + "loss": 1.1655, + "step": 1056 + }, + { + "epoch": 0.046799419988266935, + "grad_norm": 1.5839579105377197, + "learning_rate": 9.353410097431356e-06, + "loss": 1.079, + "step": 1057 + }, + { + "epoch": 0.04684369569308081, + "grad_norm": 1.3347371816635132, + "learning_rate": 9.362267493356954e-06, + "loss": 0.9638, + "step": 1058 + }, + { + "epoch": 0.04688797139789469, + "grad_norm": 1.927664041519165, + "learning_rate": 9.37112488928255e-06, + "loss": 1.1946, + "step": 1059 + }, + { + "epoch": 0.04693224710270857, + "grad_norm": 1.5561027526855469, + "learning_rate": 9.379982285208149e-06, + "loss": 1.1501, + "step": 1060 + }, + { + "epoch": 0.046976522807522444, + "grad_norm": 1.740508794784546, + "learning_rate": 9.388839681133747e-06, + "loss": 1.1529, + "step": 1061 + }, + { + "epoch": 0.047020798512336315, + "grad_norm": 1.5281798839569092, + "learning_rate": 9.397697077059346e-06, + "loss": 1.0362, + "step": 1062 + }, + { + "epoch": 0.04706507421715019, + "grad_norm": 1.688083529472351, + "learning_rate": 9.406554472984944e-06, + "loss": 1.0633, + "step": 1063 + }, + { + "epoch": 0.04710934992196407, + "grad_norm": 1.789361596107483, + "learning_rate": 9.415411868910541e-06, + "loss": 1.0614, + "step": 1064 + }, + { + "epoch": 0.04715362562677795, + "grad_norm": 1.7699180841445923, + "learning_rate": 9.42426926483614e-06, + "loss": 1.1738, + "step": 1065 + }, + { + "epoch": 0.047197901331591824, + "grad_norm": 1.6190801858901978, + "learning_rate": 9.433126660761736e-06, + "loss": 1.031, + "step": 1066 + }, + { + "epoch": 0.0472421770364057, + "grad_norm": 1.9315850734710693, + "learning_rate": 9.441984056687334e-06, + "loss": 1.0815, + "step": 1067 + }, + { + "epoch": 0.04728645274121957, + "grad_norm": 1.6093460321426392, + "learning_rate": 9.450841452612933e-06, + "loss": 1.1512, + "step": 1068 + }, + { + "epoch": 0.04733072844603345, + "grad_norm": 1.514045238494873, + "learning_rate": 9.45969884853853e-06, + "loss": 1.1209, + "step": 1069 + }, + { + "epoch": 0.047375004150847326, + "grad_norm": 1.5357089042663574, + "learning_rate": 9.46855624446413e-06, + "loss": 1.0401, + "step": 1070 + }, + { + "epoch": 0.0474192798556612, + "grad_norm": 1.4998518228530884, + "learning_rate": 9.477413640389726e-06, + "loss": 1.0408, + "step": 1071 + }, + { + "epoch": 0.04746355556047508, + "grad_norm": 1.4417043924331665, + "learning_rate": 9.486271036315325e-06, + "loss": 1.0233, + "step": 1072 + }, + { + "epoch": 0.04750783126528895, + "grad_norm": 1.6792681217193604, + "learning_rate": 9.495128432240921e-06, + "loss": 1.134, + "step": 1073 + }, + { + "epoch": 0.04755210697010283, + "grad_norm": 1.8767812252044678, + "learning_rate": 9.50398582816652e-06, + "loss": 1.1705, + "step": 1074 + }, + { + "epoch": 0.047596382674916705, + "grad_norm": 2.4753353595733643, + "learning_rate": 9.512843224092118e-06, + "loss": 1.1911, + "step": 1075 + }, + { + "epoch": 0.04764065837973058, + "grad_norm": 1.503467321395874, + "learning_rate": 9.521700620017715e-06, + "loss": 1.0779, + "step": 1076 + }, + { + "epoch": 0.04768493408454446, + "grad_norm": 1.6474417448043823, + "learning_rate": 9.530558015943313e-06, + "loss": 1.1517, + "step": 1077 + }, + { + "epoch": 0.04772920978935834, + "grad_norm": 1.488194227218628, + "learning_rate": 9.539415411868912e-06, + "loss": 1.0429, + "step": 1078 + }, + { + "epoch": 0.04777348549417221, + "grad_norm": 1.5675454139709473, + "learning_rate": 9.54827280779451e-06, + "loss": 1.0876, + "step": 1079 + }, + { + "epoch": 0.047817761198986085, + "grad_norm": 1.5168653726577759, + "learning_rate": 9.557130203720107e-06, + "loss": 0.9735, + "step": 1080 + }, + { + "epoch": 0.04786203690379996, + "grad_norm": 1.5290552377700806, + "learning_rate": 9.565987599645705e-06, + "loss": 1.0427, + "step": 1081 + }, + { + "epoch": 0.04790631260861384, + "grad_norm": 1.773468017578125, + "learning_rate": 9.574844995571303e-06, + "loss": 1.2123, + "step": 1082 + }, + { + "epoch": 0.04795058831342772, + "grad_norm": 1.6392641067504883, + "learning_rate": 9.5837023914969e-06, + "loss": 1.1498, + "step": 1083 + }, + { + "epoch": 0.04799486401824159, + "grad_norm": 1.338322639465332, + "learning_rate": 9.592559787422499e-06, + "loss": 0.9657, + "step": 1084 + }, + { + "epoch": 0.048039139723055464, + "grad_norm": 1.6504621505737305, + "learning_rate": 9.601417183348097e-06, + "loss": 1.0974, + "step": 1085 + }, + { + "epoch": 0.04808341542786934, + "grad_norm": 1.932027816772461, + "learning_rate": 9.610274579273694e-06, + "loss": 1.2295, + "step": 1086 + }, + { + "epoch": 0.04812769113268322, + "grad_norm": 1.557578206062317, + "learning_rate": 9.619131975199292e-06, + "loss": 1.1011, + "step": 1087 + }, + { + "epoch": 0.048171966837497096, + "grad_norm": 1.5557395219802856, + "learning_rate": 9.627989371124889e-06, + "loss": 1.0801, + "step": 1088 + }, + { + "epoch": 0.04821624254231097, + "grad_norm": 1.598909854888916, + "learning_rate": 9.636846767050489e-06, + "loss": 1.0212, + "step": 1089 + }, + { + "epoch": 0.048260518247124844, + "grad_norm": 1.5231828689575195, + "learning_rate": 9.645704162976086e-06, + "loss": 1.0712, + "step": 1090 + }, + { + "epoch": 0.04830479395193872, + "grad_norm": 1.7507708072662354, + "learning_rate": 9.654561558901684e-06, + "loss": 1.0994, + "step": 1091 + }, + { + "epoch": 0.0483490696567526, + "grad_norm": 1.5904425382614136, + "learning_rate": 9.663418954827282e-06, + "loss": 1.0793, + "step": 1092 + }, + { + "epoch": 0.048393345361566475, + "grad_norm": 1.6247936487197876, + "learning_rate": 9.672276350752879e-06, + "loss": 1.1545, + "step": 1093 + }, + { + "epoch": 0.04843762106638035, + "grad_norm": 1.556182861328125, + "learning_rate": 9.681133746678477e-06, + "loss": 1.1035, + "step": 1094 + }, + { + "epoch": 0.04848189677119422, + "grad_norm": 1.7684487104415894, + "learning_rate": 9.689991142604076e-06, + "loss": 1.1233, + "step": 1095 + }, + { + "epoch": 0.0485261724760081, + "grad_norm": 1.5940029621124268, + "learning_rate": 9.698848538529672e-06, + "loss": 1.1155, + "step": 1096 + }, + { + "epoch": 0.04857044818082198, + "grad_norm": 1.4502018690109253, + "learning_rate": 9.707705934455271e-06, + "loss": 1.0562, + "step": 1097 + }, + { + "epoch": 0.048614723885635855, + "grad_norm": 1.6270554065704346, + "learning_rate": 9.71656333038087e-06, + "loss": 1.1295, + "step": 1098 + }, + { + "epoch": 0.04865899959044973, + "grad_norm": 2.086621046066284, + "learning_rate": 9.725420726306468e-06, + "loss": 1.1508, + "step": 1099 + }, + { + "epoch": 0.04870327529526361, + "grad_norm": 1.6047383546829224, + "learning_rate": 9.734278122232064e-06, + "loss": 1.1764, + "step": 1100 + }, + { + "epoch": 0.04874755100007748, + "grad_norm": 1.6792625188827515, + "learning_rate": 9.743135518157663e-06, + "loss": 1.0721, + "step": 1101 + }, + { + "epoch": 0.04879182670489136, + "grad_norm": 1.5813013315200806, + "learning_rate": 9.751992914083261e-06, + "loss": 1.1579, + "step": 1102 + }, + { + "epoch": 0.048836102409705234, + "grad_norm": 1.569412112236023, + "learning_rate": 9.760850310008858e-06, + "loss": 1.0951, + "step": 1103 + }, + { + "epoch": 0.04888037811451911, + "grad_norm": 1.5470452308654785, + "learning_rate": 9.769707705934456e-06, + "loss": 1.0848, + "step": 1104 + }, + { + "epoch": 0.04892465381933299, + "grad_norm": 1.890821099281311, + "learning_rate": 9.778565101860053e-06, + "loss": 1.1003, + "step": 1105 + }, + { + "epoch": 0.04896892952414686, + "grad_norm": 1.5266520977020264, + "learning_rate": 9.787422497785651e-06, + "loss": 1.0411, + "step": 1106 + }, + { + "epoch": 0.049013205228960736, + "grad_norm": 2.09816575050354, + "learning_rate": 9.79627989371125e-06, + "loss": 1.2277, + "step": 1107 + }, + { + "epoch": 0.049057480933774614, + "grad_norm": 1.7988802194595337, + "learning_rate": 9.805137289636848e-06, + "loss": 1.2661, + "step": 1108 + }, + { + "epoch": 0.04910175663858849, + "grad_norm": 1.5198407173156738, + "learning_rate": 9.813994685562446e-06, + "loss": 1.091, + "step": 1109 + }, + { + "epoch": 0.04914603234340237, + "grad_norm": 1.4994686841964722, + "learning_rate": 9.822852081488043e-06, + "loss": 1.0108, + "step": 1110 + }, + { + "epoch": 0.049190308048216246, + "grad_norm": 2.3539187908172607, + "learning_rate": 9.831709477413642e-06, + "loss": 1.1172, + "step": 1111 + }, + { + "epoch": 0.049234583753030116, + "grad_norm": 1.8550575971603394, + "learning_rate": 9.840566873339238e-06, + "loss": 1.1219, + "step": 1112 + }, + { + "epoch": 0.04927885945784399, + "grad_norm": 1.6826618909835815, + "learning_rate": 9.849424269264837e-06, + "loss": 1.1563, + "step": 1113 + }, + { + "epoch": 0.04932313516265787, + "grad_norm": 1.8055744171142578, + "learning_rate": 9.858281665190435e-06, + "loss": 1.0823, + "step": 1114 + }, + { + "epoch": 0.04936741086747175, + "grad_norm": 1.6710069179534912, + "learning_rate": 9.867139061116032e-06, + "loss": 1.1168, + "step": 1115 + }, + { + "epoch": 0.049411686572285625, + "grad_norm": 1.6637423038482666, + "learning_rate": 9.87599645704163e-06, + "loss": 1.1497, + "step": 1116 + }, + { + "epoch": 0.049455962277099495, + "grad_norm": 1.8386090993881226, + "learning_rate": 9.884853852967229e-06, + "loss": 1.1319, + "step": 1117 + }, + { + "epoch": 0.04950023798191337, + "grad_norm": 1.7437076568603516, + "learning_rate": 9.893711248892827e-06, + "loss": 1.1106, + "step": 1118 + }, + { + "epoch": 0.04954451368672725, + "grad_norm": 1.7933582067489624, + "learning_rate": 9.902568644818424e-06, + "loss": 1.1538, + "step": 1119 + }, + { + "epoch": 0.04958878939154113, + "grad_norm": 1.60869300365448, + "learning_rate": 9.911426040744022e-06, + "loss": 1.0397, + "step": 1120 + }, + { + "epoch": 0.049633065096355004, + "grad_norm": 1.5774224996566772, + "learning_rate": 9.92028343666962e-06, + "loss": 1.0769, + "step": 1121 + }, + { + "epoch": 0.04967734080116888, + "grad_norm": 1.8505640029907227, + "learning_rate": 9.929140832595217e-06, + "loss": 1.1699, + "step": 1122 + }, + { + "epoch": 0.04972161650598275, + "grad_norm": 1.5542787313461304, + "learning_rate": 9.937998228520815e-06, + "loss": 1.0788, + "step": 1123 + }, + { + "epoch": 0.04976589221079663, + "grad_norm": 1.5293821096420288, + "learning_rate": 9.946855624446414e-06, + "loss": 1.0719, + "step": 1124 + }, + { + "epoch": 0.049810167915610506, + "grad_norm": 1.4831794500350952, + "learning_rate": 9.95571302037201e-06, + "loss": 1.0783, + "step": 1125 + }, + { + "epoch": 0.049854443620424384, + "grad_norm": 1.584160566329956, + "learning_rate": 9.964570416297609e-06, + "loss": 1.0593, + "step": 1126 + }, + { + "epoch": 0.04989871932523826, + "grad_norm": 1.466525673866272, + "learning_rate": 9.973427812223207e-06, + "loss": 1.0407, + "step": 1127 + }, + { + "epoch": 0.04994299503005213, + "grad_norm": 1.4694420099258423, + "learning_rate": 9.982285208148806e-06, + "loss": 1.05, + "step": 1128 + }, + { + "epoch": 0.04998727073486601, + "grad_norm": 1.527368187904358, + "learning_rate": 9.991142604074402e-06, + "loss": 1.0693, + "step": 1129 + }, + { + "epoch": 0.050031546439679886, + "grad_norm": 1.9619210958480835, + "learning_rate": 1e-05, + "loss": 1.0662, + "step": 1130 + }, + { + "epoch": 0.05007582214449376, + "grad_norm": 1.4913250207901, + "learning_rate": 1.00088573959256e-05, + "loss": 1.0596, + "step": 1131 + }, + { + "epoch": 0.05012009784930764, + "grad_norm": 1.5689492225646973, + "learning_rate": 1.0017714791851196e-05, + "loss": 1.0958, + "step": 1132 + }, + { + "epoch": 0.05016437355412152, + "grad_norm": 1.6604441404342651, + "learning_rate": 1.0026572187776796e-05, + "loss": 1.115, + "step": 1133 + }, + { + "epoch": 0.05020864925893539, + "grad_norm": 1.5460362434387207, + "learning_rate": 1.0035429583702393e-05, + "loss": 1.0953, + "step": 1134 + }, + { + "epoch": 0.050252924963749265, + "grad_norm": 1.4747941493988037, + "learning_rate": 1.0044286979627991e-05, + "loss": 0.9779, + "step": 1135 + }, + { + "epoch": 0.05029720066856314, + "grad_norm": 1.615526795387268, + "learning_rate": 1.0053144375553588e-05, + "loss": 1.0228, + "step": 1136 + }, + { + "epoch": 0.05034147637337702, + "grad_norm": 2.177618980407715, + "learning_rate": 1.0062001771479186e-05, + "loss": 1.1468, + "step": 1137 + }, + { + "epoch": 0.0503857520781909, + "grad_norm": 1.9319260120391846, + "learning_rate": 1.0070859167404783e-05, + "loss": 1.0597, + "step": 1138 + }, + { + "epoch": 0.05043002778300477, + "grad_norm": 1.618257999420166, + "learning_rate": 1.0079716563330381e-05, + "loss": 1.0597, + "step": 1139 + }, + { + "epoch": 0.050474303487818645, + "grad_norm": 2.007699728012085, + "learning_rate": 1.0088573959255981e-05, + "loss": 1.1605, + "step": 1140 + }, + { + "epoch": 0.05051857919263252, + "grad_norm": 1.4992609024047852, + "learning_rate": 1.0097431355181578e-05, + "loss": 1.0332, + "step": 1141 + }, + { + "epoch": 0.0505628548974464, + "grad_norm": 1.7244611978530884, + "learning_rate": 1.0106288751107176e-05, + "loss": 1.1245, + "step": 1142 + }, + { + "epoch": 0.05060713060226028, + "grad_norm": 1.7947646379470825, + "learning_rate": 1.0115146147032773e-05, + "loss": 1.1465, + "step": 1143 + }, + { + "epoch": 0.050651406307074154, + "grad_norm": 2.1999306678771973, + "learning_rate": 1.0124003542958372e-05, + "loss": 1.1887, + "step": 1144 + }, + { + "epoch": 0.050695682011888024, + "grad_norm": 1.4997544288635254, + "learning_rate": 1.0132860938883968e-05, + "loss": 1.0284, + "step": 1145 + }, + { + "epoch": 0.0507399577167019, + "grad_norm": 2.0180158615112305, + "learning_rate": 1.0141718334809567e-05, + "loss": 1.1078, + "step": 1146 + }, + { + "epoch": 0.05078423342151578, + "grad_norm": 1.7403359413146973, + "learning_rate": 1.0150575730735163e-05, + "loss": 1.1133, + "step": 1147 + }, + { + "epoch": 0.050828509126329656, + "grad_norm": 1.5863028764724731, + "learning_rate": 1.0159433126660763e-05, + "loss": 1.0638, + "step": 1148 + }, + { + "epoch": 0.05087278483114353, + "grad_norm": 1.8284761905670166, + "learning_rate": 1.0168290522586362e-05, + "loss": 1.1313, + "step": 1149 + }, + { + "epoch": 0.050917060535957404, + "grad_norm": 1.5586349964141846, + "learning_rate": 1.0177147918511958e-05, + "loss": 1.104, + "step": 1150 + }, + { + "epoch": 0.05096133624077128, + "grad_norm": 1.840325951576233, + "learning_rate": 1.0186005314437557e-05, + "loss": 1.0146, + "step": 1151 + }, + { + "epoch": 0.05100561194558516, + "grad_norm": 2.1199300289154053, + "learning_rate": 1.0194862710363154e-05, + "loss": 1.1149, + "step": 1152 + }, + { + "epoch": 0.051049887650399035, + "grad_norm": 2.2242796421051025, + "learning_rate": 1.0203720106288752e-05, + "loss": 1.1035, + "step": 1153 + }, + { + "epoch": 0.05109416335521291, + "grad_norm": 2.024542808532715, + "learning_rate": 1.0212577502214349e-05, + "loss": 1.1728, + "step": 1154 + }, + { + "epoch": 0.05113843906002679, + "grad_norm": 1.813781976699829, + "learning_rate": 1.0221434898139949e-05, + "loss": 1.1286, + "step": 1155 + }, + { + "epoch": 0.05118271476484066, + "grad_norm": 1.412493109703064, + "learning_rate": 1.0230292294065545e-05, + "loss": 1.0193, + "step": 1156 + }, + { + "epoch": 0.05122699046965454, + "grad_norm": 1.8096182346343994, + "learning_rate": 1.0239149689991144e-05, + "loss": 1.1697, + "step": 1157 + }, + { + "epoch": 0.051271266174468415, + "grad_norm": 1.8261544704437256, + "learning_rate": 1.0248007085916742e-05, + "loss": 1.041, + "step": 1158 + }, + { + "epoch": 0.05131554187928229, + "grad_norm": 1.7347654104232788, + "learning_rate": 1.0256864481842339e-05, + "loss": 1.054, + "step": 1159 + }, + { + "epoch": 0.05135981758409617, + "grad_norm": 1.6604461669921875, + "learning_rate": 1.0265721877767937e-05, + "loss": 1.121, + "step": 1160 + }, + { + "epoch": 0.05140409328891004, + "grad_norm": 2.0460798740386963, + "learning_rate": 1.0274579273693534e-05, + "loss": 1.0848, + "step": 1161 + }, + { + "epoch": 0.05144836899372392, + "grad_norm": 1.5576598644256592, + "learning_rate": 1.0283436669619134e-05, + "loss": 0.9839, + "step": 1162 + }, + { + "epoch": 0.051492644698537794, + "grad_norm": 2.681201696395874, + "learning_rate": 1.029229406554473e-05, + "loss": 1.1534, + "step": 1163 + }, + { + "epoch": 0.05153692040335167, + "grad_norm": 1.6002683639526367, + "learning_rate": 1.030115146147033e-05, + "loss": 1.1252, + "step": 1164 + }, + { + "epoch": 0.05158119610816555, + "grad_norm": 1.6206294298171997, + "learning_rate": 1.0310008857395926e-05, + "loss": 1.0648, + "step": 1165 + }, + { + "epoch": 0.051625471812979426, + "grad_norm": 1.6261489391326904, + "learning_rate": 1.0318866253321524e-05, + "loss": 1.0638, + "step": 1166 + }, + { + "epoch": 0.051669747517793296, + "grad_norm": 1.6895561218261719, + "learning_rate": 1.0327723649247121e-05, + "loss": 1.1634, + "step": 1167 + }, + { + "epoch": 0.051714023222607174, + "grad_norm": 1.5123425722122192, + "learning_rate": 1.033658104517272e-05, + "loss": 1.0591, + "step": 1168 + }, + { + "epoch": 0.05175829892742105, + "grad_norm": 1.534933090209961, + "learning_rate": 1.034543844109832e-05, + "loss": 1.0852, + "step": 1169 + }, + { + "epoch": 0.05180257463223493, + "grad_norm": 1.4861267805099487, + "learning_rate": 1.0354295837023916e-05, + "loss": 1.0238, + "step": 1170 + }, + { + "epoch": 0.051846850337048805, + "grad_norm": 1.4347182512283325, + "learning_rate": 1.0363153232949515e-05, + "loss": 1.015, + "step": 1171 + }, + { + "epoch": 0.051891126041862676, + "grad_norm": 1.4333715438842773, + "learning_rate": 1.0372010628875111e-05, + "loss": 1.0259, + "step": 1172 + }, + { + "epoch": 0.05193540174667655, + "grad_norm": 1.5809234380722046, + "learning_rate": 1.038086802480071e-05, + "loss": 1.0289, + "step": 1173 + }, + { + "epoch": 0.05197967745149043, + "grad_norm": 1.8880484104156494, + "learning_rate": 1.0389725420726306e-05, + "loss": 1.1462, + "step": 1174 + }, + { + "epoch": 0.05202395315630431, + "grad_norm": 1.7910422086715698, + "learning_rate": 1.0398582816651905e-05, + "loss": 1.0528, + "step": 1175 + }, + { + "epoch": 0.052068228861118185, + "grad_norm": 1.8230780363082886, + "learning_rate": 1.0407440212577501e-05, + "loss": 1.104, + "step": 1176 + }, + { + "epoch": 0.05211250456593206, + "grad_norm": 1.8853957653045654, + "learning_rate": 1.0416297608503101e-05, + "loss": 1.2031, + "step": 1177 + }, + { + "epoch": 0.05215678027074593, + "grad_norm": 1.5286900997161865, + "learning_rate": 1.04251550044287e-05, + "loss": 1.0742, + "step": 1178 + }, + { + "epoch": 0.05220105597555981, + "grad_norm": 1.6563992500305176, + "learning_rate": 1.0434012400354297e-05, + "loss": 1.0926, + "step": 1179 + }, + { + "epoch": 0.05224533168037369, + "grad_norm": 1.901554822921753, + "learning_rate": 1.0442869796279895e-05, + "loss": 1.1267, + "step": 1180 + }, + { + "epoch": 0.052289607385187564, + "grad_norm": 1.4704474210739136, + "learning_rate": 1.0451727192205492e-05, + "loss": 1.004, + "step": 1181 + }, + { + "epoch": 0.05233388309000144, + "grad_norm": 1.9912489652633667, + "learning_rate": 1.046058458813109e-05, + "loss": 1.1144, + "step": 1182 + }, + { + "epoch": 0.05237815879481531, + "grad_norm": 1.7264894247055054, + "learning_rate": 1.0469441984056687e-05, + "loss": 1.1041, + "step": 1183 + }, + { + "epoch": 0.05242243449962919, + "grad_norm": 1.3727269172668457, + "learning_rate": 1.0478299379982287e-05, + "loss": 1.0145, + "step": 1184 + }, + { + "epoch": 0.052466710204443066, + "grad_norm": 1.4917351007461548, + "learning_rate": 1.0487156775907884e-05, + "loss": 1.0323, + "step": 1185 + }, + { + "epoch": 0.052510985909256944, + "grad_norm": 1.4090152978897095, + "learning_rate": 1.0496014171833482e-05, + "loss": 1.0252, + "step": 1186 + }, + { + "epoch": 0.05255526161407082, + "grad_norm": 1.5235600471496582, + "learning_rate": 1.050487156775908e-05, + "loss": 1.0556, + "step": 1187 + }, + { + "epoch": 0.0525995373188847, + "grad_norm": 1.4865820407867432, + "learning_rate": 1.0513728963684677e-05, + "loss": 1.0818, + "step": 1188 + }, + { + "epoch": 0.05264381302369857, + "grad_norm": 1.7223931550979614, + "learning_rate": 1.0522586359610275e-05, + "loss": 1.2052, + "step": 1189 + }, + { + "epoch": 0.052688088728512446, + "grad_norm": 1.6688430309295654, + "learning_rate": 1.0531443755535872e-05, + "loss": 1.0634, + "step": 1190 + }, + { + "epoch": 0.05273236443332632, + "grad_norm": 1.4378167390823364, + "learning_rate": 1.0540301151461472e-05, + "loss": 1.0915, + "step": 1191 + }, + { + "epoch": 0.0527766401381402, + "grad_norm": 1.3975129127502441, + "learning_rate": 1.0549158547387069e-05, + "loss": 1.0895, + "step": 1192 + }, + { + "epoch": 0.05282091584295408, + "grad_norm": 1.547113060951233, + "learning_rate": 1.0558015943312667e-05, + "loss": 1.0873, + "step": 1193 + }, + { + "epoch": 0.05286519154776795, + "grad_norm": 1.5735793113708496, + "learning_rate": 1.0566873339238264e-05, + "loss": 1.0964, + "step": 1194 + }, + { + "epoch": 0.052909467252581825, + "grad_norm": 1.4585212469100952, + "learning_rate": 1.0575730735163862e-05, + "loss": 1.0271, + "step": 1195 + }, + { + "epoch": 0.0529537429573957, + "grad_norm": 1.5806835889816284, + "learning_rate": 1.058458813108946e-05, + "loss": 1.0718, + "step": 1196 + }, + { + "epoch": 0.05299801866220958, + "grad_norm": 1.6112654209136963, + "learning_rate": 1.0593445527015057e-05, + "loss": 1.1035, + "step": 1197 + }, + { + "epoch": 0.05304229436702346, + "grad_norm": 1.4353039264678955, + "learning_rate": 1.0602302922940658e-05, + "loss": 1.0211, + "step": 1198 + }, + { + "epoch": 0.053086570071837334, + "grad_norm": 1.5332859754562378, + "learning_rate": 1.0611160318866254e-05, + "loss": 1.0896, + "step": 1199 + }, + { + "epoch": 0.053130845776651205, + "grad_norm": 2.328212261199951, + "learning_rate": 1.0620017714791853e-05, + "loss": 1.1493, + "step": 1200 + }, + { + "epoch": 0.05317512148146508, + "grad_norm": 1.7472628355026245, + "learning_rate": 1.062887511071745e-05, + "loss": 1.2359, + "step": 1201 + }, + { + "epoch": 0.05321939718627896, + "grad_norm": 1.7721474170684814, + "learning_rate": 1.0637732506643048e-05, + "loss": 1.1783, + "step": 1202 + }, + { + "epoch": 0.053263672891092836, + "grad_norm": 1.7767058610916138, + "learning_rate": 1.0646589902568644e-05, + "loss": 1.1252, + "step": 1203 + }, + { + "epoch": 0.053307948595906714, + "grad_norm": 4.156576633453369, + "learning_rate": 1.0655447298494243e-05, + "loss": 1.0814, + "step": 1204 + }, + { + "epoch": 0.053352224300720584, + "grad_norm": 1.3699251413345337, + "learning_rate": 1.0664304694419843e-05, + "loss": 1.0559, + "step": 1205 + }, + { + "epoch": 0.05339650000553446, + "grad_norm": 1.682627558708191, + "learning_rate": 1.067316209034544e-05, + "loss": 1.0396, + "step": 1206 + }, + { + "epoch": 0.05344077571034834, + "grad_norm": 1.4962096214294434, + "learning_rate": 1.0682019486271038e-05, + "loss": 1.0266, + "step": 1207 + }, + { + "epoch": 0.053485051415162216, + "grad_norm": 1.72304105758667, + "learning_rate": 1.0690876882196635e-05, + "loss": 1.162, + "step": 1208 + }, + { + "epoch": 0.05352932711997609, + "grad_norm": 1.6105518341064453, + "learning_rate": 1.0699734278122233e-05, + "loss": 1.0199, + "step": 1209 + }, + { + "epoch": 0.05357360282478997, + "grad_norm": 1.6197600364685059, + "learning_rate": 1.070859167404783e-05, + "loss": 1.086, + "step": 1210 + }, + { + "epoch": 0.05361787852960384, + "grad_norm": 1.6738337278366089, + "learning_rate": 1.071744906997343e-05, + "loss": 1.1309, + "step": 1211 + }, + { + "epoch": 0.05366215423441772, + "grad_norm": 1.634973406791687, + "learning_rate": 1.0726306465899025e-05, + "loss": 1.1054, + "step": 1212 + }, + { + "epoch": 0.053706429939231595, + "grad_norm": 1.8585455417633057, + "learning_rate": 1.0735163861824625e-05, + "loss": 1.1186, + "step": 1213 + }, + { + "epoch": 0.05375070564404547, + "grad_norm": 1.8953168392181396, + "learning_rate": 1.0744021257750223e-05, + "loss": 1.192, + "step": 1214 + }, + { + "epoch": 0.05379498134885935, + "grad_norm": 1.6528315544128418, + "learning_rate": 1.075287865367582e-05, + "loss": 1.0642, + "step": 1215 + }, + { + "epoch": 0.05383925705367322, + "grad_norm": 1.3511054515838623, + "learning_rate": 1.0761736049601418e-05, + "loss": 0.9599, + "step": 1216 + }, + { + "epoch": 0.0538835327584871, + "grad_norm": 1.6104788780212402, + "learning_rate": 1.0770593445527015e-05, + "loss": 1.0851, + "step": 1217 + }, + { + "epoch": 0.053927808463300975, + "grad_norm": 1.7493832111358643, + "learning_rate": 1.0779450841452615e-05, + "loss": 1.0749, + "step": 1218 + }, + { + "epoch": 0.05397208416811485, + "grad_norm": 2.5303730964660645, + "learning_rate": 1.0788308237378212e-05, + "loss": 1.2625, + "step": 1219 + }, + { + "epoch": 0.05401635987292873, + "grad_norm": 2.061589241027832, + "learning_rate": 1.079716563330381e-05, + "loss": 1.1525, + "step": 1220 + }, + { + "epoch": 0.054060635577742606, + "grad_norm": 1.346318244934082, + "learning_rate": 1.0806023029229407e-05, + "loss": 0.9461, + "step": 1221 + }, + { + "epoch": 0.05410491128255648, + "grad_norm": 2.172135591506958, + "learning_rate": 1.0814880425155005e-05, + "loss": 1.0662, + "step": 1222 + }, + { + "epoch": 0.054149186987370354, + "grad_norm": 1.8328471183776855, + "learning_rate": 1.0823737821080602e-05, + "loss": 1.1165, + "step": 1223 + }, + { + "epoch": 0.05419346269218423, + "grad_norm": 1.6841145753860474, + "learning_rate": 1.08325952170062e-05, + "loss": 1.1175, + "step": 1224 + }, + { + "epoch": 0.05423773839699811, + "grad_norm": 1.8114854097366333, + "learning_rate": 1.08414526129318e-05, + "loss": 1.0769, + "step": 1225 + }, + { + "epoch": 0.054282014101811986, + "grad_norm": 1.8676906824111938, + "learning_rate": 1.0850310008857397e-05, + "loss": 1.0732, + "step": 1226 + }, + { + "epoch": 0.054326289806625856, + "grad_norm": 1.5703176259994507, + "learning_rate": 1.0859167404782996e-05, + "loss": 1.1668, + "step": 1227 + }, + { + "epoch": 0.054370565511439733, + "grad_norm": 1.4564292430877686, + "learning_rate": 1.0868024800708592e-05, + "loss": 1.0425, + "step": 1228 + }, + { + "epoch": 0.05441484121625361, + "grad_norm": 1.7289351224899292, + "learning_rate": 1.087688219663419e-05, + "loss": 1.1638, + "step": 1229 + }, + { + "epoch": 0.05445911692106749, + "grad_norm": 1.790016770362854, + "learning_rate": 1.0885739592559787e-05, + "loss": 1.1464, + "step": 1230 + }, + { + "epoch": 0.054503392625881365, + "grad_norm": 1.6771314144134521, + "learning_rate": 1.0894596988485386e-05, + "loss": 1.0587, + "step": 1231 + }, + { + "epoch": 0.05454766833069524, + "grad_norm": 2.125575065612793, + "learning_rate": 1.0903454384410983e-05, + "loss": 1.2422, + "step": 1232 + }, + { + "epoch": 0.05459194403550911, + "grad_norm": 1.5949898958206177, + "learning_rate": 1.0912311780336583e-05, + "loss": 1.0203, + "step": 1233 + }, + { + "epoch": 0.05463621974032299, + "grad_norm": 1.5426528453826904, + "learning_rate": 1.0921169176262181e-05, + "loss": 1.1271, + "step": 1234 + }, + { + "epoch": 0.05468049544513687, + "grad_norm": 2.551421880722046, + "learning_rate": 1.0930026572187778e-05, + "loss": 1.1441, + "step": 1235 + }, + { + "epoch": 0.054724771149950745, + "grad_norm": 2.28648042678833, + "learning_rate": 1.0938883968113376e-05, + "loss": 1.2354, + "step": 1236 + }, + { + "epoch": 0.05476904685476462, + "grad_norm": 1.6585595607757568, + "learning_rate": 1.0947741364038973e-05, + "loss": 1.0733, + "step": 1237 + }, + { + "epoch": 0.05481332255957849, + "grad_norm": 2.0777359008789062, + "learning_rate": 1.0956598759964571e-05, + "loss": 1.1434, + "step": 1238 + }, + { + "epoch": 0.05485759826439237, + "grad_norm": 1.6178513765335083, + "learning_rate": 1.0965456155890168e-05, + "loss": 1.0904, + "step": 1239 + }, + { + "epoch": 0.05490187396920625, + "grad_norm": 1.6636803150177002, + "learning_rate": 1.0974313551815768e-05, + "loss": 1.1752, + "step": 1240 + }, + { + "epoch": 0.054946149674020124, + "grad_norm": 1.4707154035568237, + "learning_rate": 1.0983170947741365e-05, + "loss": 1.0509, + "step": 1241 + }, + { + "epoch": 0.054990425378834, + "grad_norm": 1.454438328742981, + "learning_rate": 1.0992028343666963e-05, + "loss": 1.0719, + "step": 1242 + }, + { + "epoch": 0.05503470108364788, + "grad_norm": 1.4856648445129395, + "learning_rate": 1.1000885739592561e-05, + "loss": 1.0452, + "step": 1243 + }, + { + "epoch": 0.05507897678846175, + "grad_norm": 1.823286771774292, + "learning_rate": 1.1009743135518158e-05, + "loss": 1.1619, + "step": 1244 + }, + { + "epoch": 0.055123252493275626, + "grad_norm": 1.5845973491668701, + "learning_rate": 1.1018600531443757e-05, + "loss": 1.0234, + "step": 1245 + }, + { + "epoch": 0.055167528198089504, + "grad_norm": 1.5625611543655396, + "learning_rate": 1.1027457927369353e-05, + "loss": 1.0995, + "step": 1246 + }, + { + "epoch": 0.05521180390290338, + "grad_norm": 1.4781581163406372, + "learning_rate": 1.1036315323294953e-05, + "loss": 0.9828, + "step": 1247 + }, + { + "epoch": 0.05525607960771726, + "grad_norm": 1.4482078552246094, + "learning_rate": 1.104517271922055e-05, + "loss": 0.9593, + "step": 1248 + }, + { + "epoch": 0.05530035531253113, + "grad_norm": 1.8199838399887085, + "learning_rate": 1.1054030115146148e-05, + "loss": 1.1592, + "step": 1249 + }, + { + "epoch": 0.055344631017345006, + "grad_norm": 1.6107730865478516, + "learning_rate": 1.1062887511071745e-05, + "loss": 1.0107, + "step": 1250 + }, + { + "epoch": 0.05538890672215888, + "grad_norm": 1.562192678451538, + "learning_rate": 1.1071744906997343e-05, + "loss": 1.1334, + "step": 1251 + }, + { + "epoch": 0.05543318242697276, + "grad_norm": 1.4571642875671387, + "learning_rate": 1.1080602302922942e-05, + "loss": 1.02, + "step": 1252 + }, + { + "epoch": 0.05547745813178664, + "grad_norm": 1.768089771270752, + "learning_rate": 1.1089459698848539e-05, + "loss": 1.1454, + "step": 1253 + }, + { + "epoch": 0.055521733836600515, + "grad_norm": 1.7448464632034302, + "learning_rate": 1.1098317094774139e-05, + "loss": 1.1303, + "step": 1254 + }, + { + "epoch": 0.055566009541414385, + "grad_norm": 1.6780188083648682, + "learning_rate": 1.1107174490699735e-05, + "loss": 1.1794, + "step": 1255 + }, + { + "epoch": 0.05561028524622826, + "grad_norm": 1.4359009265899658, + "learning_rate": 1.1116031886625334e-05, + "loss": 0.9891, + "step": 1256 + }, + { + "epoch": 0.05565456095104214, + "grad_norm": 1.6253383159637451, + "learning_rate": 1.112488928255093e-05, + "loss": 1.1723, + "step": 1257 + }, + { + "epoch": 0.05569883665585602, + "grad_norm": 1.5751296281814575, + "learning_rate": 1.1133746678476529e-05, + "loss": 1.01, + "step": 1258 + }, + { + "epoch": 0.055743112360669894, + "grad_norm": 1.7660369873046875, + "learning_rate": 1.1142604074402126e-05, + "loss": 1.1768, + "step": 1259 + }, + { + "epoch": 0.055787388065483764, + "grad_norm": 1.5835628509521484, + "learning_rate": 1.1151461470327724e-05, + "loss": 1.0933, + "step": 1260 + }, + { + "epoch": 0.05583166377029764, + "grad_norm": 2.088104009628296, + "learning_rate": 1.1160318866253324e-05, + "loss": 1.1596, + "step": 1261 + }, + { + "epoch": 0.05587593947511152, + "grad_norm": 1.745935320854187, + "learning_rate": 1.116917626217892e-05, + "loss": 1.2023, + "step": 1262 + }, + { + "epoch": 0.055920215179925396, + "grad_norm": 1.5283924341201782, + "learning_rate": 1.1178033658104519e-05, + "loss": 1.1023, + "step": 1263 + }, + { + "epoch": 0.055964490884739274, + "grad_norm": 1.6635992527008057, + "learning_rate": 1.1186891054030116e-05, + "loss": 1.1342, + "step": 1264 + }, + { + "epoch": 0.05600876658955315, + "grad_norm": 1.9043351411819458, + "learning_rate": 1.1195748449955714e-05, + "loss": 1.0653, + "step": 1265 + }, + { + "epoch": 0.05605304229436702, + "grad_norm": 1.404150128364563, + "learning_rate": 1.1204605845881311e-05, + "loss": 1.0111, + "step": 1266 + }, + { + "epoch": 0.0560973179991809, + "grad_norm": 1.4505597352981567, + "learning_rate": 1.121346324180691e-05, + "loss": 1.0167, + "step": 1267 + }, + { + "epoch": 0.056141593703994776, + "grad_norm": 1.6431190967559814, + "learning_rate": 1.1222320637732506e-05, + "loss": 1.0176, + "step": 1268 + }, + { + "epoch": 0.05618586940880865, + "grad_norm": 1.5055370330810547, + "learning_rate": 1.1231178033658106e-05, + "loss": 1.0801, + "step": 1269 + }, + { + "epoch": 0.05623014511362253, + "grad_norm": 1.6192317008972168, + "learning_rate": 1.1240035429583704e-05, + "loss": 1.0898, + "step": 1270 + }, + { + "epoch": 0.0562744208184364, + "grad_norm": 3.470925807952881, + "learning_rate": 1.1248892825509301e-05, + "loss": 1.0747, + "step": 1271 + }, + { + "epoch": 0.05631869652325028, + "grad_norm": 1.9860469102859497, + "learning_rate": 1.12577502214349e-05, + "loss": 1.2924, + "step": 1272 + }, + { + "epoch": 0.056362972228064155, + "grad_norm": 1.5796138048171997, + "learning_rate": 1.1266607617360496e-05, + "loss": 1.1042, + "step": 1273 + }, + { + "epoch": 0.05640724793287803, + "grad_norm": 1.514185905456543, + "learning_rate": 1.1275465013286095e-05, + "loss": 1.0523, + "step": 1274 + }, + { + "epoch": 0.05645152363769191, + "grad_norm": 1.6316194534301758, + "learning_rate": 1.1284322409211691e-05, + "loss": 1.09, + "step": 1275 + }, + { + "epoch": 0.05649579934250579, + "grad_norm": 1.5175416469573975, + "learning_rate": 1.1293179805137291e-05, + "loss": 1.0554, + "step": 1276 + }, + { + "epoch": 0.05654007504731966, + "grad_norm": 1.4554872512817383, + "learning_rate": 1.1302037201062888e-05, + "loss": 1.0059, + "step": 1277 + }, + { + "epoch": 0.056584350752133535, + "grad_norm": 1.9476252794265747, + "learning_rate": 1.1310894596988486e-05, + "loss": 1.1754, + "step": 1278 + }, + { + "epoch": 0.05662862645694741, + "grad_norm": 1.5976994037628174, + "learning_rate": 1.1319751992914083e-05, + "loss": 1.0508, + "step": 1279 + }, + { + "epoch": 0.05667290216176129, + "grad_norm": 1.465683937072754, + "learning_rate": 1.1328609388839682e-05, + "loss": 1.044, + "step": 1280 + }, + { + "epoch": 0.056717177866575166, + "grad_norm": 1.6928958892822266, + "learning_rate": 1.1337466784765282e-05, + "loss": 1.0688, + "step": 1281 + }, + { + "epoch": 0.05676145357138904, + "grad_norm": 1.507886290550232, + "learning_rate": 1.1346324180690877e-05, + "loss": 1.1071, + "step": 1282 + }, + { + "epoch": 0.056805729276202914, + "grad_norm": 1.6803295612335205, + "learning_rate": 1.1355181576616477e-05, + "loss": 1.0893, + "step": 1283 + }, + { + "epoch": 0.05685000498101679, + "grad_norm": 1.649227499961853, + "learning_rate": 1.1364038972542073e-05, + "loss": 1.0493, + "step": 1284 + }, + { + "epoch": 0.05689428068583067, + "grad_norm": 1.6837425231933594, + "learning_rate": 1.1372896368467672e-05, + "loss": 1.1126, + "step": 1285 + }, + { + "epoch": 0.056938556390644546, + "grad_norm": 1.9229658842086792, + "learning_rate": 1.1381753764393269e-05, + "loss": 1.1303, + "step": 1286 + }, + { + "epoch": 0.05698283209545842, + "grad_norm": 1.8427404165267944, + "learning_rate": 1.1390611160318867e-05, + "loss": 1.1439, + "step": 1287 + }, + { + "epoch": 0.05702710780027229, + "grad_norm": 1.397782564163208, + "learning_rate": 1.1399468556244464e-05, + "loss": 1.0177, + "step": 1288 + }, + { + "epoch": 0.05707138350508617, + "grad_norm": 1.5426218509674072, + "learning_rate": 1.1408325952170064e-05, + "loss": 1.0563, + "step": 1289 + }, + { + "epoch": 0.05711565920990005, + "grad_norm": 1.5806344747543335, + "learning_rate": 1.1417183348095662e-05, + "loss": 1.1366, + "step": 1290 + }, + { + "epoch": 0.057159934914713925, + "grad_norm": 1.746008038520813, + "learning_rate": 1.1426040744021259e-05, + "loss": 1.1, + "step": 1291 + }, + { + "epoch": 0.0572042106195278, + "grad_norm": 1.5619823932647705, + "learning_rate": 1.1434898139946857e-05, + "loss": 1.0864, + "step": 1292 + }, + { + "epoch": 0.05724848632434167, + "grad_norm": 1.6301517486572266, + "learning_rate": 1.1443755535872454e-05, + "loss": 1.0546, + "step": 1293 + }, + { + "epoch": 0.05729276202915555, + "grad_norm": 1.501312494277954, + "learning_rate": 1.1452612931798052e-05, + "loss": 1.0813, + "step": 1294 + }, + { + "epoch": 0.05733703773396943, + "grad_norm": 1.460442304611206, + "learning_rate": 1.1461470327723649e-05, + "loss": 1.0593, + "step": 1295 + }, + { + "epoch": 0.057381313438783305, + "grad_norm": 1.503791093826294, + "learning_rate": 1.1470327723649249e-05, + "loss": 1.0497, + "step": 1296 + }, + { + "epoch": 0.05742558914359718, + "grad_norm": 1.8304866552352905, + "learning_rate": 1.1479185119574846e-05, + "loss": 1.1372, + "step": 1297 + }, + { + "epoch": 0.05746986484841105, + "grad_norm": 1.7723044157028198, + "learning_rate": 1.1488042515500444e-05, + "loss": 1.056, + "step": 1298 + }, + { + "epoch": 0.05751414055322493, + "grad_norm": 1.6051040887832642, + "learning_rate": 1.1496899911426043e-05, + "loss": 1.0339, + "step": 1299 + }, + { + "epoch": 0.05755841625803881, + "grad_norm": 1.6876146793365479, + "learning_rate": 1.150575730735164e-05, + "loss": 1.0594, + "step": 1300 + }, + { + "epoch": 0.057602691962852684, + "grad_norm": 1.5274029970169067, + "learning_rate": 1.1514614703277238e-05, + "loss": 1.1305, + "step": 1301 + }, + { + "epoch": 0.05764696766766656, + "grad_norm": 1.4382694959640503, + "learning_rate": 1.1523472099202834e-05, + "loss": 1.0681, + "step": 1302 + }, + { + "epoch": 0.05769124337248044, + "grad_norm": 1.614892601966858, + "learning_rate": 1.1532329495128434e-05, + "loss": 1.1106, + "step": 1303 + }, + { + "epoch": 0.05773551907729431, + "grad_norm": 1.709625005722046, + "learning_rate": 1.1541186891054031e-05, + "loss": 1.2023, + "step": 1304 + }, + { + "epoch": 0.057779794782108186, + "grad_norm": 1.5109301805496216, + "learning_rate": 1.155004428697963e-05, + "loss": 0.9833, + "step": 1305 + }, + { + "epoch": 0.05782407048692206, + "grad_norm": 1.681298017501831, + "learning_rate": 1.1558901682905226e-05, + "loss": 1.0418, + "step": 1306 + }, + { + "epoch": 0.05786834619173594, + "grad_norm": 1.7949618101119995, + "learning_rate": 1.1567759078830825e-05, + "loss": 1.1373, + "step": 1307 + }, + { + "epoch": 0.05791262189654982, + "grad_norm": 1.6694037914276123, + "learning_rate": 1.1576616474756423e-05, + "loss": 1.0457, + "step": 1308 + }, + { + "epoch": 0.05795689760136369, + "grad_norm": 1.8219326734542847, + "learning_rate": 1.158547387068202e-05, + "loss": 1.0731, + "step": 1309 + }, + { + "epoch": 0.058001173306177566, + "grad_norm": 1.672319769859314, + "learning_rate": 1.159433126660762e-05, + "loss": 1.085, + "step": 1310 + }, + { + "epoch": 0.05804544901099144, + "grad_norm": 1.9407216310501099, + "learning_rate": 1.1603188662533216e-05, + "loss": 1.0885, + "step": 1311 + }, + { + "epoch": 0.05808972471580532, + "grad_norm": 2.083594799041748, + "learning_rate": 1.1612046058458815e-05, + "loss": 1.1607, + "step": 1312 + }, + { + "epoch": 0.0581340004206192, + "grad_norm": 1.5045324563980103, + "learning_rate": 1.1620903454384412e-05, + "loss": 1.1223, + "step": 1313 + }, + { + "epoch": 0.058178276125433075, + "grad_norm": 1.4567276239395142, + "learning_rate": 1.162976085031001e-05, + "loss": 1.0841, + "step": 1314 + }, + { + "epoch": 0.058222551830246945, + "grad_norm": 1.631119966506958, + "learning_rate": 1.1638618246235607e-05, + "loss": 1.101, + "step": 1315 + }, + { + "epoch": 0.05826682753506082, + "grad_norm": 2.040944814682007, + "learning_rate": 1.1647475642161205e-05, + "loss": 1.1312, + "step": 1316 + }, + { + "epoch": 0.0583111032398747, + "grad_norm": 1.8130592107772827, + "learning_rate": 1.1656333038086805e-05, + "loss": 1.1199, + "step": 1317 + }, + { + "epoch": 0.05835537894468858, + "grad_norm": 1.631821870803833, + "learning_rate": 1.1665190434012402e-05, + "loss": 1.0401, + "step": 1318 + }, + { + "epoch": 0.058399654649502454, + "grad_norm": 1.528732419013977, + "learning_rate": 1.1674047829938e-05, + "loss": 1.22, + "step": 1319 + }, + { + "epoch": 0.058443930354316324, + "grad_norm": 1.5572668313980103, + "learning_rate": 1.1682905225863597e-05, + "loss": 1.081, + "step": 1320 + }, + { + "epoch": 0.0584882060591302, + "grad_norm": 1.312241792678833, + "learning_rate": 1.1691762621789195e-05, + "loss": 0.9545, + "step": 1321 + }, + { + "epoch": 0.05853248176394408, + "grad_norm": 1.474717378616333, + "learning_rate": 1.1700620017714792e-05, + "loss": 1.0816, + "step": 1322 + }, + { + "epoch": 0.058576757468757956, + "grad_norm": 1.604903221130371, + "learning_rate": 1.170947741364039e-05, + "loss": 1.1249, + "step": 1323 + }, + { + "epoch": 0.05862103317357183, + "grad_norm": 1.4435306787490845, + "learning_rate": 1.1718334809565987e-05, + "loss": 1.0811, + "step": 1324 + }, + { + "epoch": 0.05866530887838571, + "grad_norm": 1.5164971351623535, + "learning_rate": 1.1727192205491587e-05, + "loss": 1.0438, + "step": 1325 + }, + { + "epoch": 0.05870958458319958, + "grad_norm": 1.8276606798171997, + "learning_rate": 1.1736049601417186e-05, + "loss": 1.0947, + "step": 1326 + }, + { + "epoch": 0.05875386028801346, + "grad_norm": 1.3034554719924927, + "learning_rate": 1.1744906997342782e-05, + "loss": 0.985, + "step": 1327 + }, + { + "epoch": 0.058798135992827336, + "grad_norm": 1.452500343322754, + "learning_rate": 1.175376439326838e-05, + "loss": 1.013, + "step": 1328 + }, + { + "epoch": 0.05884241169764121, + "grad_norm": 1.3298856019973755, + "learning_rate": 1.1762621789193977e-05, + "loss": 0.8914, + "step": 1329 + }, + { + "epoch": 0.05888668740245509, + "grad_norm": 2.388376235961914, + "learning_rate": 1.1771479185119576e-05, + "loss": 1.0859, + "step": 1330 + }, + { + "epoch": 0.05893096310726896, + "grad_norm": 1.440969705581665, + "learning_rate": 1.1780336581045172e-05, + "loss": 0.9669, + "step": 1331 + }, + { + "epoch": 0.05897523881208284, + "grad_norm": 1.457397699356079, + "learning_rate": 1.1789193976970772e-05, + "loss": 1.0356, + "step": 1332 + }, + { + "epoch": 0.059019514516896715, + "grad_norm": 1.5168023109436035, + "learning_rate": 1.179805137289637e-05, + "loss": 1.0517, + "step": 1333 + }, + { + "epoch": 0.05906379022171059, + "grad_norm": 1.6266270875930786, + "learning_rate": 1.1806908768821968e-05, + "loss": 1.09, + "step": 1334 + }, + { + "epoch": 0.05910806592652447, + "grad_norm": 1.2925313711166382, + "learning_rate": 1.1815766164747564e-05, + "loss": 0.9584, + "step": 1335 + }, + { + "epoch": 0.05915234163133835, + "grad_norm": 2.0800423622131348, + "learning_rate": 1.1824623560673163e-05, + "loss": 1.0822, + "step": 1336 + }, + { + "epoch": 0.05919661733615222, + "grad_norm": 1.4971580505371094, + "learning_rate": 1.1833480956598761e-05, + "loss": 0.9586, + "step": 1337 + }, + { + "epoch": 0.059240893040966094, + "grad_norm": 1.7240842580795288, + "learning_rate": 1.1842338352524358e-05, + "loss": 1.0982, + "step": 1338 + }, + { + "epoch": 0.05928516874577997, + "grad_norm": 1.9927798509597778, + "learning_rate": 1.1851195748449958e-05, + "loss": 1.1928, + "step": 1339 + }, + { + "epoch": 0.05932944445059385, + "grad_norm": 1.5099955797195435, + "learning_rate": 1.1860053144375555e-05, + "loss": 1.0063, + "step": 1340 + }, + { + "epoch": 0.059373720155407726, + "grad_norm": 1.3722714185714722, + "learning_rate": 1.1868910540301153e-05, + "loss": 1.0472, + "step": 1341 + }, + { + "epoch": 0.0594179958602216, + "grad_norm": 1.5654308795928955, + "learning_rate": 1.187776793622675e-05, + "loss": 1.0039, + "step": 1342 + }, + { + "epoch": 0.059462271565035474, + "grad_norm": 1.7315692901611328, + "learning_rate": 1.1886625332152348e-05, + "loss": 1.0668, + "step": 1343 + }, + { + "epoch": 0.05950654726984935, + "grad_norm": 1.6943161487579346, + "learning_rate": 1.1895482728077945e-05, + "loss": 1.1023, + "step": 1344 + }, + { + "epoch": 0.05955082297466323, + "grad_norm": 1.5288302898406982, + "learning_rate": 1.1904340124003543e-05, + "loss": 1.1028, + "step": 1345 + }, + { + "epoch": 0.059595098679477106, + "grad_norm": 1.5954234600067139, + "learning_rate": 1.1913197519929143e-05, + "loss": 1.0999, + "step": 1346 + }, + { + "epoch": 0.05963937438429098, + "grad_norm": 1.8703516721725464, + "learning_rate": 1.192205491585474e-05, + "loss": 1.0742, + "step": 1347 + }, + { + "epoch": 0.05968365008910485, + "grad_norm": 1.7034015655517578, + "learning_rate": 1.1930912311780338e-05, + "loss": 1.0238, + "step": 1348 + }, + { + "epoch": 0.05972792579391873, + "grad_norm": 1.8944480419158936, + "learning_rate": 1.1939769707705935e-05, + "loss": 1.0376, + "step": 1349 + }, + { + "epoch": 0.05977220149873261, + "grad_norm": 1.6045056581497192, + "learning_rate": 1.1948627103631533e-05, + "loss": 1.1428, + "step": 1350 + }, + { + "epoch": 0.059816477203546485, + "grad_norm": 1.5325450897216797, + "learning_rate": 1.195748449955713e-05, + "loss": 1.0152, + "step": 1351 + }, + { + "epoch": 0.05986075290836036, + "grad_norm": 1.4613724946975708, + "learning_rate": 1.1966341895482728e-05, + "loss": 1.0985, + "step": 1352 + }, + { + "epoch": 0.05990502861317423, + "grad_norm": 1.6188009977340698, + "learning_rate": 1.1975199291408325e-05, + "loss": 0.9767, + "step": 1353 + }, + { + "epoch": 0.05994930431798811, + "grad_norm": 1.4226317405700684, + "learning_rate": 1.1984056687333925e-05, + "loss": 1.0825, + "step": 1354 + }, + { + "epoch": 0.05999358002280199, + "grad_norm": 1.6486459970474243, + "learning_rate": 1.1992914083259524e-05, + "loss": 1.0222, + "step": 1355 + }, + { + "epoch": 0.060037855727615864, + "grad_norm": 1.6844593286514282, + "learning_rate": 1.200177147918512e-05, + "loss": 1.0532, + "step": 1356 + }, + { + "epoch": 0.06008213143242974, + "grad_norm": 1.4935146570205688, + "learning_rate": 1.2010628875110719e-05, + "loss": 1.0409, + "step": 1357 + }, + { + "epoch": 0.06012640713724362, + "grad_norm": 1.4328434467315674, + "learning_rate": 1.2019486271036315e-05, + "loss": 1.0349, + "step": 1358 + }, + { + "epoch": 0.06017068284205749, + "grad_norm": 1.9491336345672607, + "learning_rate": 1.2028343666961915e-05, + "loss": 1.0782, + "step": 1359 + }, + { + "epoch": 0.06021495854687137, + "grad_norm": 1.8761646747589111, + "learning_rate": 1.203720106288751e-05, + "loss": 1.2022, + "step": 1360 + }, + { + "epoch": 0.060259234251685244, + "grad_norm": 1.5441831350326538, + "learning_rate": 1.204605845881311e-05, + "loss": 1.057, + "step": 1361 + }, + { + "epoch": 0.06030350995649912, + "grad_norm": 1.593377709388733, + "learning_rate": 1.2054915854738707e-05, + "loss": 1.0657, + "step": 1362 + }, + { + "epoch": 0.060347785661313, + "grad_norm": 1.3703055381774902, + "learning_rate": 1.2063773250664306e-05, + "loss": 1.0067, + "step": 1363 + }, + { + "epoch": 0.06039206136612687, + "grad_norm": 1.932590365409851, + "learning_rate": 1.2072630646589904e-05, + "loss": 1.1212, + "step": 1364 + }, + { + "epoch": 0.060436337070940746, + "grad_norm": 1.6077321767807007, + "learning_rate": 1.20814880425155e-05, + "loss": 1.082, + "step": 1365 + }, + { + "epoch": 0.06048061277575462, + "grad_norm": 1.439178466796875, + "learning_rate": 1.20903454384411e-05, + "loss": 1.0126, + "step": 1366 + }, + { + "epoch": 0.0605248884805685, + "grad_norm": 1.8082295656204224, + "learning_rate": 1.2099202834366698e-05, + "loss": 0.9504, + "step": 1367 + }, + { + "epoch": 0.06056916418538238, + "grad_norm": 1.5100046396255493, + "learning_rate": 1.2108060230292296e-05, + "loss": 1.0122, + "step": 1368 + }, + { + "epoch": 0.060613439890196255, + "grad_norm": 1.4899885654449463, + "learning_rate": 1.2116917626217893e-05, + "loss": 1.0581, + "step": 1369 + }, + { + "epoch": 0.060657715595010125, + "grad_norm": 1.4954253435134888, + "learning_rate": 1.2125775022143491e-05, + "loss": 1.1207, + "step": 1370 + }, + { + "epoch": 0.060701991299824, + "grad_norm": 1.4678019285202026, + "learning_rate": 1.2134632418069088e-05, + "loss": 1.1079, + "step": 1371 + }, + { + "epoch": 0.06074626700463788, + "grad_norm": 1.602581262588501, + "learning_rate": 1.2143489813994686e-05, + "loss": 1.0192, + "step": 1372 + }, + { + "epoch": 0.06079054270945176, + "grad_norm": 1.62953519821167, + "learning_rate": 1.2152347209920286e-05, + "loss": 1.074, + "step": 1373 + }, + { + "epoch": 0.060834818414265635, + "grad_norm": 2.190891742706299, + "learning_rate": 1.2161204605845883e-05, + "loss": 1.1556, + "step": 1374 + }, + { + "epoch": 0.060879094119079505, + "grad_norm": 1.694887399673462, + "learning_rate": 1.2170062001771481e-05, + "loss": 1.0959, + "step": 1375 + }, + { + "epoch": 0.06092336982389338, + "grad_norm": 1.5495601892471313, + "learning_rate": 1.2178919397697078e-05, + "loss": 1.0364, + "step": 1376 + }, + { + "epoch": 0.06096764552870726, + "grad_norm": 1.5721299648284912, + "learning_rate": 1.2187776793622676e-05, + "loss": 1.0769, + "step": 1377 + }, + { + "epoch": 0.06101192123352114, + "grad_norm": 1.4169926643371582, + "learning_rate": 1.2196634189548273e-05, + "loss": 1.0393, + "step": 1378 + }, + { + "epoch": 0.061056196938335014, + "grad_norm": 1.7168926000595093, + "learning_rate": 1.2205491585473871e-05, + "loss": 1.1062, + "step": 1379 + }, + { + "epoch": 0.06110047264314889, + "grad_norm": 1.5431504249572754, + "learning_rate": 1.2214348981399468e-05, + "loss": 1.0921, + "step": 1380 + }, + { + "epoch": 0.06114474834796276, + "grad_norm": 1.564570426940918, + "learning_rate": 1.2223206377325068e-05, + "loss": 1.068, + "step": 1381 + }, + { + "epoch": 0.06118902405277664, + "grad_norm": 1.456314206123352, + "learning_rate": 1.2232063773250667e-05, + "loss": 0.9888, + "step": 1382 + }, + { + "epoch": 0.061233299757590516, + "grad_norm": 2.0186707973480225, + "learning_rate": 1.2240921169176263e-05, + "loss": 1.2995, + "step": 1383 + }, + { + "epoch": 0.06127757546240439, + "grad_norm": 1.4057483673095703, + "learning_rate": 1.2249778565101862e-05, + "loss": 1.0464, + "step": 1384 + }, + { + "epoch": 0.06132185116721827, + "grad_norm": 1.6072666645050049, + "learning_rate": 1.2258635961027458e-05, + "loss": 1.0739, + "step": 1385 + }, + { + "epoch": 0.06136612687203214, + "grad_norm": 1.7555193901062012, + "learning_rate": 1.2267493356953057e-05, + "loss": 1.136, + "step": 1386 + }, + { + "epoch": 0.06141040257684602, + "grad_norm": 1.9024652242660522, + "learning_rate": 1.2276350752878654e-05, + "loss": 1.0888, + "step": 1387 + }, + { + "epoch": 0.061454678281659895, + "grad_norm": 1.7386810779571533, + "learning_rate": 1.2285208148804254e-05, + "loss": 1.1195, + "step": 1388 + }, + { + "epoch": 0.06149895398647377, + "grad_norm": 1.4672174453735352, + "learning_rate": 1.229406554472985e-05, + "loss": 1.0473, + "step": 1389 + }, + { + "epoch": 0.06154322969128765, + "grad_norm": 1.292747139930725, + "learning_rate": 1.2302922940655449e-05, + "loss": 0.9477, + "step": 1390 + }, + { + "epoch": 0.06158750539610153, + "grad_norm": 1.3714005947113037, + "learning_rate": 1.2311780336581045e-05, + "loss": 1.0925, + "step": 1391 + }, + { + "epoch": 0.0616317811009154, + "grad_norm": 1.7280805110931396, + "learning_rate": 1.2320637732506644e-05, + "loss": 1.0866, + "step": 1392 + }, + { + "epoch": 0.061676056805729275, + "grad_norm": 1.512932300567627, + "learning_rate": 1.2329495128432242e-05, + "loss": 1.0325, + "step": 1393 + }, + { + "epoch": 0.06172033251054315, + "grad_norm": 2.1052629947662354, + "learning_rate": 1.2338352524357839e-05, + "loss": 1.1906, + "step": 1394 + }, + { + "epoch": 0.06176460821535703, + "grad_norm": 1.5082319974899292, + "learning_rate": 1.2347209920283439e-05, + "loss": 1.1141, + "step": 1395 + }, + { + "epoch": 0.06180888392017091, + "grad_norm": 1.4090969562530518, + "learning_rate": 1.2356067316209036e-05, + "loss": 1.0923, + "step": 1396 + }, + { + "epoch": 0.06185315962498478, + "grad_norm": 1.2631163597106934, + "learning_rate": 1.2364924712134634e-05, + "loss": 0.9847, + "step": 1397 + }, + { + "epoch": 0.061897435329798654, + "grad_norm": 1.5926192998886108, + "learning_rate": 1.237378210806023e-05, + "loss": 1.1135, + "step": 1398 + }, + { + "epoch": 0.06194171103461253, + "grad_norm": 1.532866358757019, + "learning_rate": 1.2382639503985829e-05, + "loss": 1.0176, + "step": 1399 + }, + { + "epoch": 0.06198598673942641, + "grad_norm": 1.4633570909500122, + "learning_rate": 1.2391496899911426e-05, + "loss": 1.0902, + "step": 1400 + }, + { + "epoch": 0.062030262444240286, + "grad_norm": 2.4157848358154297, + "learning_rate": 1.2400354295837024e-05, + "loss": 1.1649, + "step": 1401 + }, + { + "epoch": 0.06207453814905416, + "grad_norm": 1.3621622323989868, + "learning_rate": 1.2409211691762624e-05, + "loss": 1.0116, + "step": 1402 + }, + { + "epoch": 0.062118813853868034, + "grad_norm": 1.5719881057739258, + "learning_rate": 1.2418069087688221e-05, + "loss": 1.1238, + "step": 1403 + }, + { + "epoch": 0.06216308955868191, + "grad_norm": 1.4771513938903809, + "learning_rate": 1.242692648361382e-05, + "loss": 1.0265, + "step": 1404 + }, + { + "epoch": 0.06220736526349579, + "grad_norm": 1.4918891191482544, + "learning_rate": 1.2435783879539416e-05, + "loss": 1.0762, + "step": 1405 + }, + { + "epoch": 0.062251640968309666, + "grad_norm": 1.4183458089828491, + "learning_rate": 1.2444641275465014e-05, + "loss": 0.9857, + "step": 1406 + }, + { + "epoch": 0.06229591667312354, + "grad_norm": 1.5306382179260254, + "learning_rate": 1.2453498671390611e-05, + "loss": 1.0635, + "step": 1407 + }, + { + "epoch": 0.06234019237793741, + "grad_norm": 1.8659242391586304, + "learning_rate": 1.246235606731621e-05, + "loss": 1.1648, + "step": 1408 + }, + { + "epoch": 0.06238446808275129, + "grad_norm": 1.6804687976837158, + "learning_rate": 1.2471213463241806e-05, + "loss": 1.039, + "step": 1409 + }, + { + "epoch": 0.06242874378756517, + "grad_norm": 2.125248908996582, + "learning_rate": 1.2480070859167406e-05, + "loss": 1.2042, + "step": 1410 + }, + { + "epoch": 0.062473019492379045, + "grad_norm": 1.8912944793701172, + "learning_rate": 1.2488928255093005e-05, + "loss": 1.1002, + "step": 1411 + }, + { + "epoch": 0.06251729519719292, + "grad_norm": 1.505600929260254, + "learning_rate": 1.2497785651018601e-05, + "loss": 1.024, + "step": 1412 + }, + { + "epoch": 0.0625615709020068, + "grad_norm": 1.7283700704574585, + "learning_rate": 1.25066430469442e-05, + "loss": 1.1069, + "step": 1413 + }, + { + "epoch": 0.06260584660682067, + "grad_norm": 1.4707105159759521, + "learning_rate": 1.2515500442869797e-05, + "loss": 1.0955, + "step": 1414 + }, + { + "epoch": 0.06265012231163455, + "grad_norm": 2.329899787902832, + "learning_rate": 1.2524357838795395e-05, + "loss": 1.1167, + "step": 1415 + }, + { + "epoch": 0.06269439801644842, + "grad_norm": 1.4537725448608398, + "learning_rate": 1.2533215234720992e-05, + "loss": 1.0368, + "step": 1416 + }, + { + "epoch": 0.0627386737212623, + "grad_norm": 1.46232008934021, + "learning_rate": 1.2542072630646592e-05, + "loss": 1.0225, + "step": 1417 + }, + { + "epoch": 0.06278294942607618, + "grad_norm": 1.5819770097732544, + "learning_rate": 1.2550930026572188e-05, + "loss": 1.0668, + "step": 1418 + }, + { + "epoch": 0.06282722513089005, + "grad_norm": 1.4244120121002197, + "learning_rate": 1.2559787422497787e-05, + "loss": 1.0288, + "step": 1419 + }, + { + "epoch": 0.06287150083570393, + "grad_norm": 1.54359769821167, + "learning_rate": 1.2568644818423385e-05, + "loss": 1.0601, + "step": 1420 + }, + { + "epoch": 0.0629157765405178, + "grad_norm": 1.4698705673217773, + "learning_rate": 1.2577502214348982e-05, + "loss": 1.0375, + "step": 1421 + }, + { + "epoch": 0.06296005224533167, + "grad_norm": 1.631828784942627, + "learning_rate": 1.258635961027458e-05, + "loss": 1.1083, + "step": 1422 + }, + { + "epoch": 0.06300432795014556, + "grad_norm": 1.3867360353469849, + "learning_rate": 1.2595217006200177e-05, + "loss": 0.9732, + "step": 1423 + }, + { + "epoch": 0.06304860365495943, + "grad_norm": 1.7960799932479858, + "learning_rate": 1.2604074402125777e-05, + "loss": 1.0999, + "step": 1424 + }, + { + "epoch": 0.06309287935977331, + "grad_norm": 1.5308328866958618, + "learning_rate": 1.2612931798051374e-05, + "loss": 1.0785, + "step": 1425 + }, + { + "epoch": 0.06313715506458718, + "grad_norm": 1.616210699081421, + "learning_rate": 1.2621789193976972e-05, + "loss": 1.1423, + "step": 1426 + }, + { + "epoch": 0.06318143076940105, + "grad_norm": 1.5129894018173218, + "learning_rate": 1.2630646589902569e-05, + "loss": 1.0606, + "step": 1427 + }, + { + "epoch": 0.06322570647421494, + "grad_norm": 2.0935394763946533, + "learning_rate": 1.2639503985828167e-05, + "loss": 1.2744, + "step": 1428 + }, + { + "epoch": 0.06326998217902881, + "grad_norm": 1.489134430885315, + "learning_rate": 1.2648361381753767e-05, + "loss": 1.0371, + "step": 1429 + }, + { + "epoch": 0.06331425788384269, + "grad_norm": 1.4862598180770874, + "learning_rate": 1.2657218777679362e-05, + "loss": 1.0638, + "step": 1430 + }, + { + "epoch": 0.06335853358865656, + "grad_norm": 1.859673261642456, + "learning_rate": 1.2666076173604962e-05, + "loss": 1.1824, + "step": 1431 + }, + { + "epoch": 0.06340280929347045, + "grad_norm": 1.5367844104766846, + "learning_rate": 1.2674933569530559e-05, + "loss": 1.014, + "step": 1432 + }, + { + "epoch": 0.06344708499828432, + "grad_norm": 1.6001611948013306, + "learning_rate": 1.2683790965456157e-05, + "loss": 1.0436, + "step": 1433 + }, + { + "epoch": 0.06349136070309819, + "grad_norm": 1.5556440353393555, + "learning_rate": 1.2692648361381754e-05, + "loss": 1.0696, + "step": 1434 + }, + { + "epoch": 0.06353563640791207, + "grad_norm": 1.5624017715454102, + "learning_rate": 1.2701505757307353e-05, + "loss": 1.173, + "step": 1435 + }, + { + "epoch": 0.06357991211272594, + "grad_norm": 1.6394906044006348, + "learning_rate": 1.271036315323295e-05, + "loss": 1.1018, + "step": 1436 + }, + { + "epoch": 0.06362418781753983, + "grad_norm": 1.7680643796920776, + "learning_rate": 1.271922054915855e-05, + "loss": 1.0552, + "step": 1437 + }, + { + "epoch": 0.0636684635223537, + "grad_norm": 1.5041719675064087, + "learning_rate": 1.2728077945084148e-05, + "loss": 0.989, + "step": 1438 + }, + { + "epoch": 0.06371273922716757, + "grad_norm": 1.4844063520431519, + "learning_rate": 1.2736935341009744e-05, + "loss": 1.0216, + "step": 1439 + }, + { + "epoch": 0.06375701493198145, + "grad_norm": 2.5261926651000977, + "learning_rate": 1.2745792736935343e-05, + "loss": 1.0438, + "step": 1440 + }, + { + "epoch": 0.06380129063679532, + "grad_norm": 1.4847722053527832, + "learning_rate": 1.275465013286094e-05, + "loss": 1.005, + "step": 1441 + }, + { + "epoch": 0.0638455663416092, + "grad_norm": 1.764731764793396, + "learning_rate": 1.2763507528786538e-05, + "loss": 1.1466, + "step": 1442 + }, + { + "epoch": 0.06388984204642308, + "grad_norm": 1.7996485233306885, + "learning_rate": 1.2772364924712135e-05, + "loss": 0.982, + "step": 1443 + }, + { + "epoch": 0.06393411775123695, + "grad_norm": 1.8046784400939941, + "learning_rate": 1.2781222320637735e-05, + "loss": 1.1658, + "step": 1444 + }, + { + "epoch": 0.06397839345605083, + "grad_norm": 1.8665281534194946, + "learning_rate": 1.2790079716563331e-05, + "loss": 1.1124, + "step": 1445 + }, + { + "epoch": 0.0640226691608647, + "grad_norm": 1.7176039218902588, + "learning_rate": 1.279893711248893e-05, + "loss": 1.1398, + "step": 1446 + }, + { + "epoch": 0.06406694486567859, + "grad_norm": 1.708836317062378, + "learning_rate": 1.2807794508414526e-05, + "loss": 1.1595, + "step": 1447 + }, + { + "epoch": 0.06411122057049246, + "grad_norm": 1.557928204536438, + "learning_rate": 1.2816651904340125e-05, + "loss": 1.0996, + "step": 1448 + }, + { + "epoch": 0.06415549627530633, + "grad_norm": 1.5590603351593018, + "learning_rate": 1.2825509300265723e-05, + "loss": 1.0823, + "step": 1449 + }, + { + "epoch": 0.06419977198012021, + "grad_norm": 1.516100525856018, + "learning_rate": 1.283436669619132e-05, + "loss": 1.0024, + "step": 1450 + }, + { + "epoch": 0.06424404768493408, + "grad_norm": 1.5389682054519653, + "learning_rate": 1.284322409211692e-05, + "loss": 1.1218, + "step": 1451 + }, + { + "epoch": 0.06428832338974796, + "grad_norm": 1.5561954975128174, + "learning_rate": 1.2852081488042517e-05, + "loss": 1.044, + "step": 1452 + }, + { + "epoch": 0.06433259909456183, + "grad_norm": 1.5878441333770752, + "learning_rate": 1.2860938883968115e-05, + "loss": 1.0254, + "step": 1453 + }, + { + "epoch": 0.06437687479937572, + "grad_norm": 1.485927939414978, + "learning_rate": 1.2869796279893712e-05, + "loss": 1.0699, + "step": 1454 + }, + { + "epoch": 0.06442115050418959, + "grad_norm": 1.4901279211044312, + "learning_rate": 1.287865367581931e-05, + "loss": 1.0049, + "step": 1455 + }, + { + "epoch": 0.06446542620900346, + "grad_norm": 1.3967386484146118, + "learning_rate": 1.2887511071744907e-05, + "loss": 1.0058, + "step": 1456 + }, + { + "epoch": 0.06450970191381734, + "grad_norm": 1.707882046699524, + "learning_rate": 1.2896368467670505e-05, + "loss": 1.1014, + "step": 1457 + }, + { + "epoch": 0.06455397761863121, + "grad_norm": 1.9184542894363403, + "learning_rate": 1.2905225863596105e-05, + "loss": 1.0304, + "step": 1458 + }, + { + "epoch": 0.0645982533234451, + "grad_norm": 1.938835859298706, + "learning_rate": 1.2914083259521702e-05, + "loss": 1.1739, + "step": 1459 + }, + { + "epoch": 0.06464252902825897, + "grad_norm": 1.668752908706665, + "learning_rate": 1.29229406554473e-05, + "loss": 1.0778, + "step": 1460 + }, + { + "epoch": 0.06468680473307284, + "grad_norm": 1.5585358142852783, + "learning_rate": 1.2931798051372897e-05, + "loss": 1.1186, + "step": 1461 + }, + { + "epoch": 0.06473108043788672, + "grad_norm": 1.6872801780700684, + "learning_rate": 1.2940655447298496e-05, + "loss": 1.0186, + "step": 1462 + }, + { + "epoch": 0.0647753561427006, + "grad_norm": 1.5091928243637085, + "learning_rate": 1.2949512843224092e-05, + "loss": 1.0254, + "step": 1463 + }, + { + "epoch": 0.06481963184751448, + "grad_norm": 1.602068543434143, + "learning_rate": 1.295837023914969e-05, + "loss": 1.0649, + "step": 1464 + }, + { + "epoch": 0.06486390755232835, + "grad_norm": 1.7300134897232056, + "learning_rate": 1.2967227635075287e-05, + "loss": 1.1639, + "step": 1465 + }, + { + "epoch": 0.06490818325714222, + "grad_norm": 1.8506338596343994, + "learning_rate": 1.2976085031000887e-05, + "loss": 1.1298, + "step": 1466 + }, + { + "epoch": 0.0649524589619561, + "grad_norm": 1.491620421409607, + "learning_rate": 1.2984942426926486e-05, + "loss": 1.0117, + "step": 1467 + }, + { + "epoch": 0.06499673466676997, + "grad_norm": 1.49282705783844, + "learning_rate": 1.2993799822852083e-05, + "loss": 0.9903, + "step": 1468 + }, + { + "epoch": 0.06504101037158386, + "grad_norm": 1.68382728099823, + "learning_rate": 1.3002657218777681e-05, + "loss": 1.1682, + "step": 1469 + }, + { + "epoch": 0.06508528607639773, + "grad_norm": 1.4225709438323975, + "learning_rate": 1.3011514614703278e-05, + "loss": 0.9858, + "step": 1470 + }, + { + "epoch": 0.0651295617812116, + "grad_norm": 1.3304141759872437, + "learning_rate": 1.3020372010628876e-05, + "loss": 1.0366, + "step": 1471 + }, + { + "epoch": 0.06517383748602548, + "grad_norm": 1.5661872625350952, + "learning_rate": 1.3029229406554473e-05, + "loss": 1.1072, + "step": 1472 + }, + { + "epoch": 0.06521811319083935, + "grad_norm": 1.6847985982894897, + "learning_rate": 1.3038086802480073e-05, + "loss": 1.116, + "step": 1473 + }, + { + "epoch": 0.06526238889565324, + "grad_norm": 1.4183673858642578, + "learning_rate": 1.304694419840567e-05, + "loss": 0.9908, + "step": 1474 + }, + { + "epoch": 0.06530666460046711, + "grad_norm": 1.3990882635116577, + "learning_rate": 1.3055801594331268e-05, + "loss": 0.9502, + "step": 1475 + }, + { + "epoch": 0.06535094030528099, + "grad_norm": 2.0635859966278076, + "learning_rate": 1.3064658990256866e-05, + "loss": 1.0765, + "step": 1476 + }, + { + "epoch": 0.06539521601009486, + "grad_norm": 1.6353355646133423, + "learning_rate": 1.3073516386182463e-05, + "loss": 1.0678, + "step": 1477 + }, + { + "epoch": 0.06543949171490873, + "grad_norm": 1.511740803718567, + "learning_rate": 1.3082373782108061e-05, + "loss": 1.0366, + "step": 1478 + }, + { + "epoch": 0.06548376741972262, + "grad_norm": 1.358681321144104, + "learning_rate": 1.3091231178033658e-05, + "loss": 0.9525, + "step": 1479 + }, + { + "epoch": 0.06552804312453649, + "grad_norm": 1.3766344785690308, + "learning_rate": 1.3100088573959258e-05, + "loss": 1.0077, + "step": 1480 + }, + { + "epoch": 0.06557231882935037, + "grad_norm": 1.587534785270691, + "learning_rate": 1.3108945969884855e-05, + "loss": 1.0991, + "step": 1481 + }, + { + "epoch": 0.06561659453416424, + "grad_norm": 1.4556812047958374, + "learning_rate": 1.3117803365810453e-05, + "loss": 1.043, + "step": 1482 + }, + { + "epoch": 0.06566087023897811, + "grad_norm": 1.7383334636688232, + "learning_rate": 1.312666076173605e-05, + "loss": 1.124, + "step": 1483 + }, + { + "epoch": 0.065705145943792, + "grad_norm": 1.490317463874817, + "learning_rate": 1.3135518157661648e-05, + "loss": 1.0292, + "step": 1484 + }, + { + "epoch": 0.06574942164860587, + "grad_norm": 1.5876322984695435, + "learning_rate": 1.3144375553587247e-05, + "loss": 1.0, + "step": 1485 + }, + { + "epoch": 0.06579369735341975, + "grad_norm": 1.9888255596160889, + "learning_rate": 1.3153232949512843e-05, + "loss": 1.1521, + "step": 1486 + }, + { + "epoch": 0.06583797305823362, + "grad_norm": 1.7220909595489502, + "learning_rate": 1.3162090345438443e-05, + "loss": 0.9441, + "step": 1487 + }, + { + "epoch": 0.06588224876304749, + "grad_norm": 1.5140221118927002, + "learning_rate": 1.317094774136404e-05, + "loss": 1.0561, + "step": 1488 + }, + { + "epoch": 0.06592652446786137, + "grad_norm": 2.0487022399902344, + "learning_rate": 1.3179805137289639e-05, + "loss": 1.1011, + "step": 1489 + }, + { + "epoch": 0.06597080017267525, + "grad_norm": 1.5764296054840088, + "learning_rate": 1.3188662533215235e-05, + "loss": 1.0416, + "step": 1490 + }, + { + "epoch": 0.06601507587748913, + "grad_norm": 1.7998920679092407, + "learning_rate": 1.3197519929140834e-05, + "loss": 1.1428, + "step": 1491 + }, + { + "epoch": 0.066059351582303, + "grad_norm": 1.4220257997512817, + "learning_rate": 1.320637732506643e-05, + "loss": 1.0229, + "step": 1492 + }, + { + "epoch": 0.06610362728711687, + "grad_norm": 1.687604546546936, + "learning_rate": 1.3215234720992029e-05, + "loss": 1.0255, + "step": 1493 + }, + { + "epoch": 0.06614790299193075, + "grad_norm": 1.908704400062561, + "learning_rate": 1.3224092116917629e-05, + "loss": 1.1598, + "step": 1494 + }, + { + "epoch": 0.06619217869674462, + "grad_norm": 1.5895658731460571, + "learning_rate": 1.3232949512843226e-05, + "loss": 1.0873, + "step": 1495 + }, + { + "epoch": 0.06623645440155851, + "grad_norm": 1.4698139429092407, + "learning_rate": 1.3241806908768824e-05, + "loss": 1.0249, + "step": 1496 + }, + { + "epoch": 0.06628073010637238, + "grad_norm": 1.491862177848816, + "learning_rate": 1.325066430469442e-05, + "loss": 1.0506, + "step": 1497 + }, + { + "epoch": 0.06632500581118626, + "grad_norm": 1.482198715209961, + "learning_rate": 1.3259521700620019e-05, + "loss": 0.9954, + "step": 1498 + }, + { + "epoch": 0.06636928151600013, + "grad_norm": 1.730015754699707, + "learning_rate": 1.3268379096545616e-05, + "loss": 1.0506, + "step": 1499 + }, + { + "epoch": 0.066413557220814, + "grad_norm": 1.587395191192627, + "learning_rate": 1.3277236492471214e-05, + "loss": 1.1043, + "step": 1500 + }, + { + "epoch": 0.06645783292562789, + "grad_norm": 1.8524080514907837, + "learning_rate": 1.328609388839681e-05, + "loss": 1.0862, + "step": 1501 + }, + { + "epoch": 0.06650210863044176, + "grad_norm": 2.4448046684265137, + "learning_rate": 1.3294951284322411e-05, + "loss": 1.1636, + "step": 1502 + }, + { + "epoch": 0.06654638433525564, + "grad_norm": 1.4735811948776245, + "learning_rate": 1.3303808680248008e-05, + "loss": 1.0784, + "step": 1503 + }, + { + "epoch": 0.06659066004006951, + "grad_norm": 1.7433375120162964, + "learning_rate": 1.3312666076173606e-05, + "loss": 1.0445, + "step": 1504 + }, + { + "epoch": 0.06663493574488338, + "grad_norm": 1.5542752742767334, + "learning_rate": 1.3321523472099204e-05, + "loss": 1.0709, + "step": 1505 + }, + { + "epoch": 0.06667921144969727, + "grad_norm": 1.606571912765503, + "learning_rate": 1.3330380868024801e-05, + "loss": 1.1064, + "step": 1506 + }, + { + "epoch": 0.06672348715451114, + "grad_norm": 1.4393972158432007, + "learning_rate": 1.3339238263950401e-05, + "loss": 1.0866, + "step": 1507 + }, + { + "epoch": 0.06676776285932502, + "grad_norm": 1.5423721075057983, + "learning_rate": 1.3348095659875996e-05, + "loss": 1.0814, + "step": 1508 + }, + { + "epoch": 0.06681203856413889, + "grad_norm": 1.469772458076477, + "learning_rate": 1.3356953055801596e-05, + "loss": 0.9846, + "step": 1509 + }, + { + "epoch": 0.06685631426895276, + "grad_norm": 1.700380802154541, + "learning_rate": 1.3365810451727193e-05, + "loss": 1.0904, + "step": 1510 + }, + { + "epoch": 0.06690058997376665, + "grad_norm": 1.4453933238983154, + "learning_rate": 1.3374667847652791e-05, + "loss": 1.0228, + "step": 1511 + }, + { + "epoch": 0.06694486567858052, + "grad_norm": 1.5966086387634277, + "learning_rate": 1.3383525243578388e-05, + "loss": 0.9872, + "step": 1512 + }, + { + "epoch": 0.0669891413833944, + "grad_norm": 1.6763293743133545, + "learning_rate": 1.3392382639503986e-05, + "loss": 1.1352, + "step": 1513 + }, + { + "epoch": 0.06703341708820827, + "grad_norm": 1.4437997341156006, + "learning_rate": 1.3401240035429586e-05, + "loss": 1.0078, + "step": 1514 + }, + { + "epoch": 0.06707769279302214, + "grad_norm": 1.7015101909637451, + "learning_rate": 1.3410097431355183e-05, + "loss": 1.0131, + "step": 1515 + }, + { + "epoch": 0.06712196849783603, + "grad_norm": 1.480681300163269, + "learning_rate": 1.3418954827280782e-05, + "loss": 1.0264, + "step": 1516 + }, + { + "epoch": 0.0671662442026499, + "grad_norm": 1.5837786197662354, + "learning_rate": 1.3427812223206378e-05, + "loss": 1.0381, + "step": 1517 + }, + { + "epoch": 0.06721051990746378, + "grad_norm": 1.5647804737091064, + "learning_rate": 1.3436669619131977e-05, + "loss": 1.0328, + "step": 1518 + }, + { + "epoch": 0.06725479561227765, + "grad_norm": 2.0198049545288086, + "learning_rate": 1.3445527015057573e-05, + "loss": 1.1511, + "step": 1519 + }, + { + "epoch": 0.06729907131709154, + "grad_norm": 1.584647297859192, + "learning_rate": 1.3454384410983172e-05, + "loss": 1.1393, + "step": 1520 + }, + { + "epoch": 0.0673433470219054, + "grad_norm": 1.916959285736084, + "learning_rate": 1.3463241806908768e-05, + "loss": 1.1262, + "step": 1521 + }, + { + "epoch": 0.06738762272671928, + "grad_norm": 1.4515711069107056, + "learning_rate": 1.3472099202834369e-05, + "loss": 0.9939, + "step": 1522 + }, + { + "epoch": 0.06743189843153316, + "grad_norm": 1.7591087818145752, + "learning_rate": 1.3480956598759967e-05, + "loss": 1.1886, + "step": 1523 + }, + { + "epoch": 0.06747617413634703, + "grad_norm": 1.5471211671829224, + "learning_rate": 1.3489813994685564e-05, + "loss": 1.089, + "step": 1524 + }, + { + "epoch": 0.06752044984116091, + "grad_norm": 1.3355727195739746, + "learning_rate": 1.3498671390611162e-05, + "loss": 1.0138, + "step": 1525 + }, + { + "epoch": 0.06756472554597479, + "grad_norm": 1.8727229833602905, + "learning_rate": 1.3507528786536759e-05, + "loss": 1.1382, + "step": 1526 + }, + { + "epoch": 0.06760900125078866, + "grad_norm": 1.661893606185913, + "learning_rate": 1.3516386182462357e-05, + "loss": 1.0746, + "step": 1527 + }, + { + "epoch": 0.06765327695560254, + "grad_norm": 1.6987160444259644, + "learning_rate": 1.3525243578387954e-05, + "loss": 1.144, + "step": 1528 + }, + { + "epoch": 0.06769755266041641, + "grad_norm": 1.4835691452026367, + "learning_rate": 1.3534100974313554e-05, + "loss": 1.0124, + "step": 1529 + }, + { + "epoch": 0.0677418283652303, + "grad_norm": 1.793064832687378, + "learning_rate": 1.354295837023915e-05, + "loss": 1.0418, + "step": 1530 + }, + { + "epoch": 0.06778610407004416, + "grad_norm": 1.667426586151123, + "learning_rate": 1.3551815766164749e-05, + "loss": 1.1224, + "step": 1531 + }, + { + "epoch": 0.06783037977485804, + "grad_norm": 1.3876150846481323, + "learning_rate": 1.3560673162090347e-05, + "loss": 1.0219, + "step": 1532 + }, + { + "epoch": 0.06787465547967192, + "grad_norm": 1.4184279441833496, + "learning_rate": 1.3569530558015944e-05, + "loss": 1.0463, + "step": 1533 + }, + { + "epoch": 0.06791893118448579, + "grad_norm": 1.6909418106079102, + "learning_rate": 1.3578387953941542e-05, + "loss": 1.0649, + "step": 1534 + }, + { + "epoch": 0.06796320688929967, + "grad_norm": 1.46371591091156, + "learning_rate": 1.358724534986714e-05, + "loss": 0.9774, + "step": 1535 + }, + { + "epoch": 0.06800748259411354, + "grad_norm": 1.6495407819747925, + "learning_rate": 1.359610274579274e-05, + "loss": 1.065, + "step": 1536 + }, + { + "epoch": 0.06805175829892741, + "grad_norm": 1.717343807220459, + "learning_rate": 1.3604960141718336e-05, + "loss": 1.18, + "step": 1537 + }, + { + "epoch": 0.0680960340037413, + "grad_norm": 1.4548848867416382, + "learning_rate": 1.3613817537643934e-05, + "loss": 1.0573, + "step": 1538 + }, + { + "epoch": 0.06814030970855517, + "grad_norm": 1.6124186515808105, + "learning_rate": 1.3622674933569531e-05, + "loss": 0.9903, + "step": 1539 + }, + { + "epoch": 0.06818458541336905, + "grad_norm": 1.7176713943481445, + "learning_rate": 1.363153232949513e-05, + "loss": 1.1053, + "step": 1540 + }, + { + "epoch": 0.06822886111818292, + "grad_norm": 1.7332658767700195, + "learning_rate": 1.3640389725420728e-05, + "loss": 1.0657, + "step": 1541 + }, + { + "epoch": 0.06827313682299681, + "grad_norm": 1.5798358917236328, + "learning_rate": 1.3649247121346325e-05, + "loss": 1.1306, + "step": 1542 + }, + { + "epoch": 0.06831741252781068, + "grad_norm": 1.3967419862747192, + "learning_rate": 1.3658104517271925e-05, + "loss": 1.0349, + "step": 1543 + }, + { + "epoch": 0.06836168823262455, + "grad_norm": 1.3800606727600098, + "learning_rate": 1.3666961913197521e-05, + "loss": 1.0159, + "step": 1544 + }, + { + "epoch": 0.06840596393743843, + "grad_norm": 1.674945592880249, + "learning_rate": 1.367581930912312e-05, + "loss": 1.1497, + "step": 1545 + }, + { + "epoch": 0.0684502396422523, + "grad_norm": 1.2589294910430908, + "learning_rate": 1.3684676705048716e-05, + "loss": 0.9877, + "step": 1546 + }, + { + "epoch": 0.06849451534706619, + "grad_norm": 1.412672519683838, + "learning_rate": 1.3693534100974315e-05, + "loss": 0.9734, + "step": 1547 + }, + { + "epoch": 0.06853879105188006, + "grad_norm": 1.8876951932907104, + "learning_rate": 1.3702391496899911e-05, + "loss": 1.1827, + "step": 1548 + }, + { + "epoch": 0.06858306675669393, + "grad_norm": 1.7518409490585327, + "learning_rate": 1.371124889282551e-05, + "loss": 1.1427, + "step": 1549 + }, + { + "epoch": 0.06862734246150781, + "grad_norm": 1.5966695547103882, + "learning_rate": 1.372010628875111e-05, + "loss": 1.1461, + "step": 1550 + }, + { + "epoch": 0.06867161816632168, + "grad_norm": 1.4504640102386475, + "learning_rate": 1.3728963684676707e-05, + "loss": 0.9879, + "step": 1551 + }, + { + "epoch": 0.06871589387113557, + "grad_norm": 1.4180325269699097, + "learning_rate": 1.3737821080602305e-05, + "loss": 1.0469, + "step": 1552 + }, + { + "epoch": 0.06876016957594944, + "grad_norm": 1.7166565656661987, + "learning_rate": 1.3746678476527902e-05, + "loss": 1.1207, + "step": 1553 + }, + { + "epoch": 0.06880444528076331, + "grad_norm": 1.476486086845398, + "learning_rate": 1.37555358724535e-05, + "loss": 1.0325, + "step": 1554 + }, + { + "epoch": 0.06884872098557719, + "grad_norm": 2.5504939556121826, + "learning_rate": 1.3764393268379097e-05, + "loss": 1.2031, + "step": 1555 + }, + { + "epoch": 0.06889299669039106, + "grad_norm": 1.3098187446594238, + "learning_rate": 1.3773250664304695e-05, + "loss": 0.9371, + "step": 1556 + }, + { + "epoch": 0.06893727239520495, + "grad_norm": 1.5332980155944824, + "learning_rate": 1.3782108060230292e-05, + "loss": 1.0772, + "step": 1557 + }, + { + "epoch": 0.06898154810001882, + "grad_norm": 1.6609326601028442, + "learning_rate": 1.3790965456155892e-05, + "loss": 1.1204, + "step": 1558 + }, + { + "epoch": 0.06902582380483269, + "grad_norm": 1.4903266429901123, + "learning_rate": 1.3799822852081489e-05, + "loss": 1.0005, + "step": 1559 + }, + { + "epoch": 0.06907009950964657, + "grad_norm": 1.4911597967147827, + "learning_rate": 1.3808680248007087e-05, + "loss": 1.0681, + "step": 1560 + }, + { + "epoch": 0.06911437521446044, + "grad_norm": 1.8017199039459229, + "learning_rate": 1.3817537643932685e-05, + "loss": 1.0866, + "step": 1561 + }, + { + "epoch": 0.06915865091927433, + "grad_norm": 1.256210207939148, + "learning_rate": 1.3826395039858282e-05, + "loss": 0.9931, + "step": 1562 + }, + { + "epoch": 0.0692029266240882, + "grad_norm": 1.6859487295150757, + "learning_rate": 1.383525243578388e-05, + "loss": 1.0502, + "step": 1563 + }, + { + "epoch": 0.06924720232890208, + "grad_norm": 1.5255417823791504, + "learning_rate": 1.3844109831709477e-05, + "loss": 1.1046, + "step": 1564 + }, + { + "epoch": 0.06929147803371595, + "grad_norm": 1.9143704175949097, + "learning_rate": 1.3852967227635077e-05, + "loss": 1.0873, + "step": 1565 + }, + { + "epoch": 0.06933575373852982, + "grad_norm": 1.5773569345474243, + "learning_rate": 1.3861824623560674e-05, + "loss": 1.113, + "step": 1566 + }, + { + "epoch": 0.0693800294433437, + "grad_norm": 1.6392221450805664, + "learning_rate": 1.3870682019486272e-05, + "loss": 1.0488, + "step": 1567 + }, + { + "epoch": 0.06942430514815758, + "grad_norm": 1.6534439325332642, + "learning_rate": 1.3879539415411869e-05, + "loss": 1.0827, + "step": 1568 + }, + { + "epoch": 0.06946858085297146, + "grad_norm": 1.6439112424850464, + "learning_rate": 1.3888396811337468e-05, + "loss": 0.9995, + "step": 1569 + }, + { + "epoch": 0.06951285655778533, + "grad_norm": 1.5048911571502686, + "learning_rate": 1.3897254207263066e-05, + "loss": 1.0388, + "step": 1570 + }, + { + "epoch": 0.0695571322625992, + "grad_norm": 1.7533866167068481, + "learning_rate": 1.3906111603188663e-05, + "loss": 1.0996, + "step": 1571 + }, + { + "epoch": 0.06960140796741308, + "grad_norm": 1.3711292743682861, + "learning_rate": 1.3914968999114263e-05, + "loss": 1.0319, + "step": 1572 + }, + { + "epoch": 0.06964568367222695, + "grad_norm": 1.5682752132415771, + "learning_rate": 1.392382639503986e-05, + "loss": 1.0654, + "step": 1573 + }, + { + "epoch": 0.06968995937704084, + "grad_norm": 1.5783342123031616, + "learning_rate": 1.3932683790965458e-05, + "loss": 1.046, + "step": 1574 + }, + { + "epoch": 0.06973423508185471, + "grad_norm": 1.4988675117492676, + "learning_rate": 1.3941541186891054e-05, + "loss": 1.0681, + "step": 1575 + }, + { + "epoch": 0.06977851078666858, + "grad_norm": 1.502337098121643, + "learning_rate": 1.3950398582816653e-05, + "loss": 1.0638, + "step": 1576 + }, + { + "epoch": 0.06982278649148246, + "grad_norm": 1.7403255701065063, + "learning_rate": 1.395925597874225e-05, + "loss": 1.0569, + "step": 1577 + }, + { + "epoch": 0.06986706219629633, + "grad_norm": 1.5052356719970703, + "learning_rate": 1.3968113374667848e-05, + "loss": 1.0781, + "step": 1578 + }, + { + "epoch": 0.06991133790111022, + "grad_norm": 1.417306661605835, + "learning_rate": 1.3976970770593448e-05, + "loss": 0.9993, + "step": 1579 + }, + { + "epoch": 0.06995561360592409, + "grad_norm": 1.3527570962905884, + "learning_rate": 1.3985828166519045e-05, + "loss": 0.988, + "step": 1580 + }, + { + "epoch": 0.06999988931073796, + "grad_norm": 1.2829350233078003, + "learning_rate": 1.3994685562444643e-05, + "loss": 0.9633, + "step": 1581 + }, + { + "epoch": 0.07004416501555184, + "grad_norm": 1.6342041492462158, + "learning_rate": 1.400354295837024e-05, + "loss": 1.0608, + "step": 1582 + }, + { + "epoch": 0.07008844072036571, + "grad_norm": 1.4861984252929688, + "learning_rate": 1.4012400354295838e-05, + "loss": 1.0095, + "step": 1583 + }, + { + "epoch": 0.0701327164251796, + "grad_norm": 1.5075143575668335, + "learning_rate": 1.4021257750221435e-05, + "loss": 1.0355, + "step": 1584 + }, + { + "epoch": 0.07017699212999347, + "grad_norm": 1.8077731132507324, + "learning_rate": 1.4030115146147035e-05, + "loss": 1.1637, + "step": 1585 + }, + { + "epoch": 0.07022126783480735, + "grad_norm": 1.6136999130249023, + "learning_rate": 1.403897254207263e-05, + "loss": 1.1235, + "step": 1586 + }, + { + "epoch": 0.07026554353962122, + "grad_norm": 1.9260088205337524, + "learning_rate": 1.404782993799823e-05, + "loss": 1.0724, + "step": 1587 + }, + { + "epoch": 0.07030981924443509, + "grad_norm": 1.5247046947479248, + "learning_rate": 1.4056687333923828e-05, + "loss": 1.0249, + "step": 1588 + }, + { + "epoch": 0.07035409494924898, + "grad_norm": 1.6163331270217896, + "learning_rate": 1.4065544729849425e-05, + "loss": 1.0323, + "step": 1589 + }, + { + "epoch": 0.07039837065406285, + "grad_norm": 1.582058310508728, + "learning_rate": 1.4074402125775024e-05, + "loss": 1.1099, + "step": 1590 + }, + { + "epoch": 0.07044264635887673, + "grad_norm": 1.6889022588729858, + "learning_rate": 1.408325952170062e-05, + "loss": 1.1164, + "step": 1591 + }, + { + "epoch": 0.0704869220636906, + "grad_norm": 1.4668583869934082, + "learning_rate": 1.409211691762622e-05, + "loss": 0.984, + "step": 1592 + }, + { + "epoch": 0.07053119776850447, + "grad_norm": 1.4819936752319336, + "learning_rate": 1.4100974313551817e-05, + "loss": 1.0402, + "step": 1593 + }, + { + "epoch": 0.07057547347331836, + "grad_norm": 1.4864133596420288, + "learning_rate": 1.4109831709477415e-05, + "loss": 1.0983, + "step": 1594 + }, + { + "epoch": 0.07061974917813223, + "grad_norm": 1.4680207967758179, + "learning_rate": 1.4118689105403012e-05, + "loss": 1.0734, + "step": 1595 + }, + { + "epoch": 0.07066402488294611, + "grad_norm": 1.914656639099121, + "learning_rate": 1.412754650132861e-05, + "loss": 1.048, + "step": 1596 + }, + { + "epoch": 0.07070830058775998, + "grad_norm": 1.5095821619033813, + "learning_rate": 1.4136403897254209e-05, + "loss": 0.9636, + "step": 1597 + }, + { + "epoch": 0.07075257629257385, + "grad_norm": 1.6365139484405518, + "learning_rate": 1.4145261293179806e-05, + "loss": 1.035, + "step": 1598 + }, + { + "epoch": 0.07079685199738774, + "grad_norm": 1.4234377145767212, + "learning_rate": 1.4154118689105406e-05, + "loss": 1.0365, + "step": 1599 + }, + { + "epoch": 0.0708411277022016, + "grad_norm": 1.531190276145935, + "learning_rate": 1.4162976085031002e-05, + "loss": 1.0565, + "step": 1600 + }, + { + "epoch": 0.07088540340701549, + "grad_norm": 1.4311283826828003, + "learning_rate": 1.41718334809566e-05, + "loss": 1.0101, + "step": 1601 + }, + { + "epoch": 0.07092967911182936, + "grad_norm": 1.656528353691101, + "learning_rate": 1.4180690876882197e-05, + "loss": 1.0917, + "step": 1602 + }, + { + "epoch": 0.07097395481664323, + "grad_norm": 1.3878482580184937, + "learning_rate": 1.4189548272807796e-05, + "loss": 1.0175, + "step": 1603 + }, + { + "epoch": 0.07101823052145712, + "grad_norm": 1.5864996910095215, + "learning_rate": 1.4198405668733393e-05, + "loss": 1.0572, + "step": 1604 + }, + { + "epoch": 0.07106250622627099, + "grad_norm": 1.603859305381775, + "learning_rate": 1.4207263064658991e-05, + "loss": 1.0955, + "step": 1605 + }, + { + "epoch": 0.07110678193108487, + "grad_norm": 1.5458412170410156, + "learning_rate": 1.4216120460584591e-05, + "loss": 1.0301, + "step": 1606 + }, + { + "epoch": 0.07115105763589874, + "grad_norm": 1.432026982307434, + "learning_rate": 1.4224977856510188e-05, + "loss": 1.0213, + "step": 1607 + }, + { + "epoch": 0.07119533334071262, + "grad_norm": 1.671407699584961, + "learning_rate": 1.4233835252435786e-05, + "loss": 1.0423, + "step": 1608 + }, + { + "epoch": 0.0712396090455265, + "grad_norm": 1.5632392168045044, + "learning_rate": 1.4242692648361383e-05, + "loss": 1.0655, + "step": 1609 + }, + { + "epoch": 0.07128388475034036, + "grad_norm": 1.3816169500350952, + "learning_rate": 1.4251550044286981e-05, + "loss": 1.0016, + "step": 1610 + }, + { + "epoch": 0.07132816045515425, + "grad_norm": 1.2813853025436401, + "learning_rate": 1.4260407440212578e-05, + "loss": 1.0439, + "step": 1611 + }, + { + "epoch": 0.07137243615996812, + "grad_norm": 1.665056824684143, + "learning_rate": 1.4269264836138176e-05, + "loss": 1.0826, + "step": 1612 + }, + { + "epoch": 0.071416711864782, + "grad_norm": 1.9416730403900146, + "learning_rate": 1.4278122232063773e-05, + "loss": 1.0811, + "step": 1613 + }, + { + "epoch": 0.07146098756959587, + "grad_norm": 1.367199182510376, + "learning_rate": 1.4286979627989373e-05, + "loss": 1.0049, + "step": 1614 + }, + { + "epoch": 0.07150526327440974, + "grad_norm": 1.6422626972198486, + "learning_rate": 1.429583702391497e-05, + "loss": 1.1099, + "step": 1615 + }, + { + "epoch": 0.07154953897922363, + "grad_norm": 1.4157391786575317, + "learning_rate": 1.4304694419840568e-05, + "loss": 1.0515, + "step": 1616 + }, + { + "epoch": 0.0715938146840375, + "grad_norm": 6.626451015472412, + "learning_rate": 1.4313551815766167e-05, + "loss": 1.0618, + "step": 1617 + }, + { + "epoch": 0.07163809038885138, + "grad_norm": 1.5568190813064575, + "learning_rate": 1.4322409211691763e-05, + "loss": 1.1518, + "step": 1618 + }, + { + "epoch": 0.07168236609366525, + "grad_norm": 1.6714080572128296, + "learning_rate": 1.4331266607617362e-05, + "loss": 1.171, + "step": 1619 + }, + { + "epoch": 0.07172664179847912, + "grad_norm": 1.4817712306976318, + "learning_rate": 1.4340124003542958e-05, + "loss": 1.0044, + "step": 1620 + }, + { + "epoch": 0.07177091750329301, + "grad_norm": 1.4856603145599365, + "learning_rate": 1.4348981399468558e-05, + "loss": 1.0447, + "step": 1621 + }, + { + "epoch": 0.07181519320810688, + "grad_norm": 1.6979044675827026, + "learning_rate": 1.4357838795394155e-05, + "loss": 1.1538, + "step": 1622 + }, + { + "epoch": 0.07185946891292076, + "grad_norm": 1.6284204721450806, + "learning_rate": 1.4366696191319754e-05, + "loss": 1.0594, + "step": 1623 + }, + { + "epoch": 0.07190374461773463, + "grad_norm": 1.3543744087219238, + "learning_rate": 1.437555358724535e-05, + "loss": 1.0238, + "step": 1624 + }, + { + "epoch": 0.0719480203225485, + "grad_norm": 1.841251015663147, + "learning_rate": 1.4384410983170949e-05, + "loss": 1.135, + "step": 1625 + }, + { + "epoch": 0.07199229602736239, + "grad_norm": 2.394878625869751, + "learning_rate": 1.4393268379096547e-05, + "loss": 1.1112, + "step": 1626 + }, + { + "epoch": 0.07203657173217626, + "grad_norm": 1.5137507915496826, + "learning_rate": 1.4402125775022144e-05, + "loss": 1.014, + "step": 1627 + }, + { + "epoch": 0.07208084743699014, + "grad_norm": 1.3639925718307495, + "learning_rate": 1.4410983170947744e-05, + "loss": 1.0074, + "step": 1628 + }, + { + "epoch": 0.07212512314180401, + "grad_norm": 1.239938497543335, + "learning_rate": 1.441984056687334e-05, + "loss": 0.896, + "step": 1629 + }, + { + "epoch": 0.0721693988466179, + "grad_norm": 1.543918490409851, + "learning_rate": 1.4428697962798939e-05, + "loss": 1.0495, + "step": 1630 + }, + { + "epoch": 0.07221367455143177, + "grad_norm": 1.321688175201416, + "learning_rate": 1.4437555358724536e-05, + "loss": 0.9973, + "step": 1631 + }, + { + "epoch": 0.07225795025624564, + "grad_norm": 1.7639715671539307, + "learning_rate": 1.4446412754650134e-05, + "loss": 1.0456, + "step": 1632 + }, + { + "epoch": 0.07230222596105952, + "grad_norm": 1.3042694330215454, + "learning_rate": 1.445527015057573e-05, + "loss": 1.0233, + "step": 1633 + }, + { + "epoch": 0.07234650166587339, + "grad_norm": 1.3966968059539795, + "learning_rate": 1.4464127546501329e-05, + "loss": 1.0148, + "step": 1634 + }, + { + "epoch": 0.07239077737068728, + "grad_norm": 1.4376482963562012, + "learning_rate": 1.4472984942426929e-05, + "loss": 1.0448, + "step": 1635 + }, + { + "epoch": 0.07243505307550115, + "grad_norm": 1.3723431825637817, + "learning_rate": 1.4481842338352526e-05, + "loss": 1.0153, + "step": 1636 + }, + { + "epoch": 0.07247932878031502, + "grad_norm": 1.4084750413894653, + "learning_rate": 1.4490699734278124e-05, + "loss": 1.0671, + "step": 1637 + }, + { + "epoch": 0.0725236044851289, + "grad_norm": 1.6529191732406616, + "learning_rate": 1.4499557130203721e-05, + "loss": 1.1329, + "step": 1638 + }, + { + "epoch": 0.07256788018994277, + "grad_norm": 1.3501306772232056, + "learning_rate": 1.450841452612932e-05, + "loss": 1.0141, + "step": 1639 + }, + { + "epoch": 0.07261215589475666, + "grad_norm": 1.4702658653259277, + "learning_rate": 1.4517271922054916e-05, + "loss": 1.0237, + "step": 1640 + }, + { + "epoch": 0.07265643159957053, + "grad_norm": 1.5581591129302979, + "learning_rate": 1.4526129317980514e-05, + "loss": 1.0544, + "step": 1641 + }, + { + "epoch": 0.0727007073043844, + "grad_norm": 1.6035188436508179, + "learning_rate": 1.4534986713906111e-05, + "loss": 0.9952, + "step": 1642 + }, + { + "epoch": 0.07274498300919828, + "grad_norm": 1.4980268478393555, + "learning_rate": 1.4543844109831711e-05, + "loss": 1.0008, + "step": 1643 + }, + { + "epoch": 0.07278925871401215, + "grad_norm": 1.4160994291305542, + "learning_rate": 1.455270150575731e-05, + "loss": 0.98, + "step": 1644 + }, + { + "epoch": 0.07283353441882603, + "grad_norm": 1.9297282695770264, + "learning_rate": 1.4561558901682906e-05, + "loss": 1.1136, + "step": 1645 + }, + { + "epoch": 0.0728778101236399, + "grad_norm": 1.7883638143539429, + "learning_rate": 1.4570416297608505e-05, + "loss": 1.1447, + "step": 1646 + }, + { + "epoch": 0.07292208582845378, + "grad_norm": 1.497725009918213, + "learning_rate": 1.4579273693534101e-05, + "loss": 1.0171, + "step": 1647 + }, + { + "epoch": 0.07296636153326766, + "grad_norm": 1.4593333005905151, + "learning_rate": 1.45881310894597e-05, + "loss": 1.041, + "step": 1648 + }, + { + "epoch": 0.07301063723808153, + "grad_norm": 1.8100476264953613, + "learning_rate": 1.4596988485385296e-05, + "loss": 1.1269, + "step": 1649 + }, + { + "epoch": 0.07305491294289541, + "grad_norm": 1.546754002571106, + "learning_rate": 1.4605845881310897e-05, + "loss": 1.0723, + "step": 1650 + }, + { + "epoch": 0.07309918864770928, + "grad_norm": 1.360739827156067, + "learning_rate": 1.4614703277236493e-05, + "loss": 1.0094, + "step": 1651 + }, + { + "epoch": 0.07314346435252315, + "grad_norm": 1.4975953102111816, + "learning_rate": 1.4623560673162092e-05, + "loss": 1.073, + "step": 1652 + }, + { + "epoch": 0.07318774005733704, + "grad_norm": 1.650687336921692, + "learning_rate": 1.463241806908769e-05, + "loss": 1.0891, + "step": 1653 + }, + { + "epoch": 0.07323201576215091, + "grad_norm": 1.5827624797821045, + "learning_rate": 1.4641275465013287e-05, + "loss": 1.0061, + "step": 1654 + }, + { + "epoch": 0.0732762914669648, + "grad_norm": 1.6052696704864502, + "learning_rate": 1.4650132860938887e-05, + "loss": 0.9957, + "step": 1655 + }, + { + "epoch": 0.07332056717177866, + "grad_norm": 1.7680610418319702, + "learning_rate": 1.4658990256864482e-05, + "loss": 1.0677, + "step": 1656 + }, + { + "epoch": 0.07336484287659255, + "grad_norm": 1.4452646970748901, + "learning_rate": 1.4667847652790082e-05, + "loss": 1.0442, + "step": 1657 + }, + { + "epoch": 0.07340911858140642, + "grad_norm": 1.4469987154006958, + "learning_rate": 1.4676705048715679e-05, + "loss": 1.0377, + "step": 1658 + }, + { + "epoch": 0.07345339428622029, + "grad_norm": 1.66299307346344, + "learning_rate": 1.4685562444641277e-05, + "loss": 1.1548, + "step": 1659 + }, + { + "epoch": 0.07349766999103417, + "grad_norm": 1.4468265771865845, + "learning_rate": 1.4694419840566874e-05, + "loss": 1.0465, + "step": 1660 + }, + { + "epoch": 0.07354194569584804, + "grad_norm": 1.3819308280944824, + "learning_rate": 1.4703277236492472e-05, + "loss": 1.0563, + "step": 1661 + }, + { + "epoch": 0.07358622140066193, + "grad_norm": 1.4770679473876953, + "learning_rate": 1.4712134632418072e-05, + "loss": 1.0769, + "step": 1662 + }, + { + "epoch": 0.0736304971054758, + "grad_norm": 1.6381686925888062, + "learning_rate": 1.4720992028343669e-05, + "loss": 1.1114, + "step": 1663 + }, + { + "epoch": 0.07367477281028967, + "grad_norm": 1.2295830249786377, + "learning_rate": 1.4729849424269267e-05, + "loss": 0.9016, + "step": 1664 + }, + { + "epoch": 0.07371904851510355, + "grad_norm": 1.4714430570602417, + "learning_rate": 1.4738706820194864e-05, + "loss": 1.0228, + "step": 1665 + }, + { + "epoch": 0.07376332421991742, + "grad_norm": 1.6042460203170776, + "learning_rate": 1.4747564216120462e-05, + "loss": 1.0799, + "step": 1666 + }, + { + "epoch": 0.0738075999247313, + "grad_norm": 1.4031484127044678, + "learning_rate": 1.4756421612046059e-05, + "loss": 1.0789, + "step": 1667 + }, + { + "epoch": 0.07385187562954518, + "grad_norm": 1.3352686166763306, + "learning_rate": 1.4765279007971657e-05, + "loss": 0.9891, + "step": 1668 + }, + { + "epoch": 0.07389615133435905, + "grad_norm": 1.563038945198059, + "learning_rate": 1.4774136403897254e-05, + "loss": 1.0788, + "step": 1669 + }, + { + "epoch": 0.07394042703917293, + "grad_norm": 1.595230221748352, + "learning_rate": 1.4782993799822854e-05, + "loss": 1.0784, + "step": 1670 + }, + { + "epoch": 0.0739847027439868, + "grad_norm": 1.6024200916290283, + "learning_rate": 1.4791851195748451e-05, + "loss": 1.0604, + "step": 1671 + }, + { + "epoch": 0.07402897844880069, + "grad_norm": 1.6210057735443115, + "learning_rate": 1.480070859167405e-05, + "loss": 1.0681, + "step": 1672 + }, + { + "epoch": 0.07407325415361456, + "grad_norm": 1.6676477193832397, + "learning_rate": 1.4809565987599648e-05, + "loss": 0.9693, + "step": 1673 + }, + { + "epoch": 0.07411752985842843, + "grad_norm": 1.2711161375045776, + "learning_rate": 1.4818423383525244e-05, + "loss": 0.9744, + "step": 1674 + }, + { + "epoch": 0.07416180556324231, + "grad_norm": 1.5142714977264404, + "learning_rate": 1.4827280779450843e-05, + "loss": 1.1217, + "step": 1675 + }, + { + "epoch": 0.07420608126805618, + "grad_norm": 1.5437300205230713, + "learning_rate": 1.483613817537644e-05, + "loss": 0.9653, + "step": 1676 + }, + { + "epoch": 0.07425035697287007, + "grad_norm": 1.917041301727295, + "learning_rate": 1.484499557130204e-05, + "loss": 1.1122, + "step": 1677 + }, + { + "epoch": 0.07429463267768394, + "grad_norm": 1.6411012411117554, + "learning_rate": 1.4853852967227636e-05, + "loss": 1.0461, + "step": 1678 + }, + { + "epoch": 0.07433890838249782, + "grad_norm": 1.465583324432373, + "learning_rate": 1.4862710363153235e-05, + "loss": 1.0541, + "step": 1679 + }, + { + "epoch": 0.07438318408731169, + "grad_norm": 1.4219963550567627, + "learning_rate": 1.4871567759078831e-05, + "loss": 1.0166, + "step": 1680 + }, + { + "epoch": 0.07442745979212556, + "grad_norm": 1.540286660194397, + "learning_rate": 1.488042515500443e-05, + "loss": 1.0106, + "step": 1681 + }, + { + "epoch": 0.07447173549693945, + "grad_norm": 1.6775070428848267, + "learning_rate": 1.4889282550930028e-05, + "loss": 1.1396, + "step": 1682 + }, + { + "epoch": 0.07451601120175332, + "grad_norm": 1.3838164806365967, + "learning_rate": 1.4898139946855625e-05, + "loss": 0.9843, + "step": 1683 + }, + { + "epoch": 0.0745602869065672, + "grad_norm": 1.4292405843734741, + "learning_rate": 1.4906997342781225e-05, + "loss": 1.0433, + "step": 1684 + }, + { + "epoch": 0.07460456261138107, + "grad_norm": 1.4233787059783936, + "learning_rate": 1.4915854738706822e-05, + "loss": 1.0785, + "step": 1685 + }, + { + "epoch": 0.07464883831619494, + "grad_norm": 1.6375585794448853, + "learning_rate": 1.492471213463242e-05, + "loss": 1.1053, + "step": 1686 + }, + { + "epoch": 0.07469311402100882, + "grad_norm": 1.449355125427246, + "learning_rate": 1.4933569530558017e-05, + "loss": 1.0563, + "step": 1687 + }, + { + "epoch": 0.0747373897258227, + "grad_norm": 1.5090113878250122, + "learning_rate": 1.4942426926483615e-05, + "loss": 1.0443, + "step": 1688 + }, + { + "epoch": 0.07478166543063658, + "grad_norm": 1.5521750450134277, + "learning_rate": 1.4951284322409212e-05, + "loss": 0.9992, + "step": 1689 + }, + { + "epoch": 0.07482594113545045, + "grad_norm": 1.987420916557312, + "learning_rate": 1.496014171833481e-05, + "loss": 1.0695, + "step": 1690 + }, + { + "epoch": 0.07487021684026432, + "grad_norm": 1.5587513446807861, + "learning_rate": 1.496899911426041e-05, + "loss": 1.0846, + "step": 1691 + }, + { + "epoch": 0.0749144925450782, + "grad_norm": 1.681993007659912, + "learning_rate": 1.4977856510186007e-05, + "loss": 1.1633, + "step": 1692 + }, + { + "epoch": 0.07495876824989207, + "grad_norm": 1.7779338359832764, + "learning_rate": 1.4986713906111605e-05, + "loss": 1.1455, + "step": 1693 + }, + { + "epoch": 0.07500304395470596, + "grad_norm": 1.4325354099273682, + "learning_rate": 1.4995571302037202e-05, + "loss": 1.0908, + "step": 1694 + }, + { + "epoch": 0.07504731965951983, + "grad_norm": 1.4067230224609375, + "learning_rate": 1.50044286979628e-05, + "loss": 0.9948, + "step": 1695 + }, + { + "epoch": 0.0750915953643337, + "grad_norm": 1.649449348449707, + "learning_rate": 1.5013286093888397e-05, + "loss": 1.1515, + "step": 1696 + }, + { + "epoch": 0.07513587106914758, + "grad_norm": 1.59440279006958, + "learning_rate": 1.5022143489813996e-05, + "loss": 1.1062, + "step": 1697 + }, + { + "epoch": 0.07518014677396145, + "grad_norm": 1.5936052799224854, + "learning_rate": 1.5031000885739592e-05, + "loss": 1.1207, + "step": 1698 + }, + { + "epoch": 0.07522442247877534, + "grad_norm": 1.3754980564117432, + "learning_rate": 1.5039858281665192e-05, + "loss": 1.0437, + "step": 1699 + }, + { + "epoch": 0.07526869818358921, + "grad_norm": 1.3941935300827026, + "learning_rate": 1.504871567759079e-05, + "loss": 1.0102, + "step": 1700 + }, + { + "epoch": 0.07531297388840309, + "grad_norm": 1.452244758605957, + "learning_rate": 1.5057573073516387e-05, + "loss": 1.0249, + "step": 1701 + }, + { + "epoch": 0.07535724959321696, + "grad_norm": 1.2789279222488403, + "learning_rate": 1.5066430469441986e-05, + "loss": 1.0029, + "step": 1702 + }, + { + "epoch": 0.07540152529803083, + "grad_norm": 1.496048927307129, + "learning_rate": 1.5075287865367582e-05, + "loss": 1.1081, + "step": 1703 + }, + { + "epoch": 0.07544580100284472, + "grad_norm": 1.6895703077316284, + "learning_rate": 1.5084145261293181e-05, + "loss": 1.06, + "step": 1704 + }, + { + "epoch": 0.07549007670765859, + "grad_norm": 1.7084996700286865, + "learning_rate": 1.5093002657218778e-05, + "loss": 1.1676, + "step": 1705 + }, + { + "epoch": 0.07553435241247247, + "grad_norm": 1.4764420986175537, + "learning_rate": 1.5101860053144378e-05, + "loss": 1.0643, + "step": 1706 + }, + { + "epoch": 0.07557862811728634, + "grad_norm": 1.3393324613571167, + "learning_rate": 1.5110717449069974e-05, + "loss": 0.9636, + "step": 1707 + }, + { + "epoch": 0.07562290382210021, + "grad_norm": 1.6255704164505005, + "learning_rate": 1.5119574844995573e-05, + "loss": 1.1142, + "step": 1708 + }, + { + "epoch": 0.0756671795269141, + "grad_norm": 1.4590530395507812, + "learning_rate": 1.5128432240921171e-05, + "loss": 1.0292, + "step": 1709 + }, + { + "epoch": 0.07571145523172797, + "grad_norm": 1.6565130949020386, + "learning_rate": 1.5137289636846768e-05, + "loss": 1.0746, + "step": 1710 + }, + { + "epoch": 0.07575573093654185, + "grad_norm": 1.668716311454773, + "learning_rate": 1.5146147032772366e-05, + "loss": 1.0384, + "step": 1711 + }, + { + "epoch": 0.07580000664135572, + "grad_norm": 1.326732873916626, + "learning_rate": 1.5155004428697963e-05, + "loss": 0.9882, + "step": 1712 + }, + { + "epoch": 0.07584428234616959, + "grad_norm": 1.4012469053268433, + "learning_rate": 1.5163861824623563e-05, + "loss": 0.9998, + "step": 1713 + }, + { + "epoch": 0.07588855805098348, + "grad_norm": 1.4673748016357422, + "learning_rate": 1.517271922054916e-05, + "loss": 1.121, + "step": 1714 + }, + { + "epoch": 0.07593283375579735, + "grad_norm": 1.4471027851104736, + "learning_rate": 1.5181576616474758e-05, + "loss": 1.0287, + "step": 1715 + }, + { + "epoch": 0.07597710946061123, + "grad_norm": 1.5458118915557861, + "learning_rate": 1.5190434012400355e-05, + "loss": 1.0436, + "step": 1716 + }, + { + "epoch": 0.0760213851654251, + "grad_norm": 1.3714605569839478, + "learning_rate": 1.5199291408325953e-05, + "loss": 1.0253, + "step": 1717 + }, + { + "epoch": 0.07606566087023897, + "grad_norm": 1.765243649482727, + "learning_rate": 1.5208148804251552e-05, + "loss": 1.0491, + "step": 1718 + }, + { + "epoch": 0.07610993657505286, + "grad_norm": 1.7650014162063599, + "learning_rate": 1.5217006200177148e-05, + "loss": 1.1344, + "step": 1719 + }, + { + "epoch": 0.07615421227986673, + "grad_norm": 1.5363519191741943, + "learning_rate": 1.5225863596102748e-05, + "loss": 1.0732, + "step": 1720 + }, + { + "epoch": 0.07619848798468061, + "grad_norm": 1.883934736251831, + "learning_rate": 1.5234720992028345e-05, + "loss": 1.0395, + "step": 1721 + }, + { + "epoch": 0.07624276368949448, + "grad_norm": 1.757368564605713, + "learning_rate": 1.5243578387953943e-05, + "loss": 1.124, + "step": 1722 + }, + { + "epoch": 0.07628703939430836, + "grad_norm": 1.4035903215408325, + "learning_rate": 1.525243578387954e-05, + "loss": 1.0058, + "step": 1723 + }, + { + "epoch": 0.07633131509912224, + "grad_norm": 1.4473637342453003, + "learning_rate": 1.526129317980514e-05, + "loss": 1.037, + "step": 1724 + }, + { + "epoch": 0.0763755908039361, + "grad_norm": 1.4012012481689453, + "learning_rate": 1.5270150575730735e-05, + "loss": 1.0665, + "step": 1725 + }, + { + "epoch": 0.07641986650874999, + "grad_norm": 1.4808578491210938, + "learning_rate": 1.5279007971656335e-05, + "loss": 1.0267, + "step": 1726 + }, + { + "epoch": 0.07646414221356386, + "grad_norm": 1.263554573059082, + "learning_rate": 1.5287865367581932e-05, + "loss": 0.9452, + "step": 1727 + }, + { + "epoch": 0.07650841791837774, + "grad_norm": 1.3169697523117065, + "learning_rate": 1.529672276350753e-05, + "loss": 0.9988, + "step": 1728 + }, + { + "epoch": 0.07655269362319161, + "grad_norm": 1.6422661542892456, + "learning_rate": 1.530558015943313e-05, + "loss": 1.1158, + "step": 1729 + }, + { + "epoch": 0.07659696932800548, + "grad_norm": 1.8252670764923096, + "learning_rate": 1.5314437555358725e-05, + "loss": 1.0797, + "step": 1730 + }, + { + "epoch": 0.07664124503281937, + "grad_norm": 1.4629168510437012, + "learning_rate": 1.5323294951284326e-05, + "loss": 1.0236, + "step": 1731 + }, + { + "epoch": 0.07668552073763324, + "grad_norm": 1.7544677257537842, + "learning_rate": 1.5332152347209922e-05, + "loss": 1.0618, + "step": 1732 + }, + { + "epoch": 0.07672979644244712, + "grad_norm": 1.8256266117095947, + "learning_rate": 1.534100974313552e-05, + "loss": 1.1269, + "step": 1733 + }, + { + "epoch": 0.076774072147261, + "grad_norm": 1.3936131000518799, + "learning_rate": 1.5349867139061116e-05, + "loss": 1.097, + "step": 1734 + }, + { + "epoch": 0.07681834785207486, + "grad_norm": 1.576634407043457, + "learning_rate": 1.5358724534986716e-05, + "loss": 1.0105, + "step": 1735 + }, + { + "epoch": 0.07686262355688875, + "grad_norm": 1.4489758014678955, + "learning_rate": 1.5367581930912312e-05, + "loss": 1.0303, + "step": 1736 + }, + { + "epoch": 0.07690689926170262, + "grad_norm": 1.376660943031311, + "learning_rate": 1.5376439326837913e-05, + "loss": 1.0348, + "step": 1737 + }, + { + "epoch": 0.0769511749665165, + "grad_norm": 1.6158077716827393, + "learning_rate": 1.538529672276351e-05, + "loss": 1.0904, + "step": 1738 + }, + { + "epoch": 0.07699545067133037, + "grad_norm": 1.8747053146362305, + "learning_rate": 1.5394154118689106e-05, + "loss": 1.1128, + "step": 1739 + }, + { + "epoch": 0.07703972637614424, + "grad_norm": 1.7550851106643677, + "learning_rate": 1.5403011514614706e-05, + "loss": 1.0333, + "step": 1740 + }, + { + "epoch": 0.07708400208095813, + "grad_norm": 1.2225261926651, + "learning_rate": 1.5411868910540303e-05, + "loss": 0.9978, + "step": 1741 + }, + { + "epoch": 0.077128277785772, + "grad_norm": 1.4097695350646973, + "learning_rate": 1.54207263064659e-05, + "loss": 0.9343, + "step": 1742 + }, + { + "epoch": 0.07717255349058588, + "grad_norm": 1.5851612091064453, + "learning_rate": 1.5429583702391496e-05, + "loss": 1.0395, + "step": 1743 + }, + { + "epoch": 0.07721682919539975, + "grad_norm": 1.6799805164337158, + "learning_rate": 1.5438441098317096e-05, + "loss": 1.1011, + "step": 1744 + }, + { + "epoch": 0.07726110490021364, + "grad_norm": 1.7235088348388672, + "learning_rate": 1.5447298494242693e-05, + "loss": 1.0532, + "step": 1745 + }, + { + "epoch": 0.07730538060502751, + "grad_norm": 1.5161020755767822, + "learning_rate": 1.5456155890168293e-05, + "loss": 0.9938, + "step": 1746 + }, + { + "epoch": 0.07734965630984138, + "grad_norm": 1.3924858570098877, + "learning_rate": 1.546501328609389e-05, + "loss": 1.0326, + "step": 1747 + }, + { + "epoch": 0.07739393201465526, + "grad_norm": 1.680848240852356, + "learning_rate": 1.5473870682019486e-05, + "loss": 1.1246, + "step": 1748 + }, + { + "epoch": 0.07743820771946913, + "grad_norm": 1.6647335290908813, + "learning_rate": 1.5482728077945086e-05, + "loss": 1.0708, + "step": 1749 + }, + { + "epoch": 0.07748248342428302, + "grad_norm": 1.4185010194778442, + "learning_rate": 1.5491585473870683e-05, + "loss": 1.0704, + "step": 1750 + }, + { + "epoch": 0.07752675912909689, + "grad_norm": 1.3626198768615723, + "learning_rate": 1.5500442869796283e-05, + "loss": 1.0277, + "step": 1751 + }, + { + "epoch": 0.07757103483391076, + "grad_norm": 1.732024073600769, + "learning_rate": 1.550930026572188e-05, + "loss": 1.0786, + "step": 1752 + }, + { + "epoch": 0.07761531053872464, + "grad_norm": 1.6175755262374878, + "learning_rate": 1.5518157661647477e-05, + "loss": 1.0365, + "step": 1753 + }, + { + "epoch": 0.07765958624353851, + "grad_norm": 1.6873019933700562, + "learning_rate": 1.5527015057573073e-05, + "loss": 1.0244, + "step": 1754 + }, + { + "epoch": 0.0777038619483524, + "grad_norm": 1.5191469192504883, + "learning_rate": 1.5535872453498673e-05, + "loss": 1.0116, + "step": 1755 + }, + { + "epoch": 0.07774813765316627, + "grad_norm": 1.724276065826416, + "learning_rate": 1.554472984942427e-05, + "loss": 1.1452, + "step": 1756 + }, + { + "epoch": 0.07779241335798014, + "grad_norm": 1.5313273668289185, + "learning_rate": 1.5553587245349867e-05, + "loss": 1.064, + "step": 1757 + }, + { + "epoch": 0.07783668906279402, + "grad_norm": 1.363182783126831, + "learning_rate": 1.5562444641275467e-05, + "loss": 0.9909, + "step": 1758 + }, + { + "epoch": 0.07788096476760789, + "grad_norm": 1.4487751722335815, + "learning_rate": 1.5571302037201064e-05, + "loss": 0.9705, + "step": 1759 + }, + { + "epoch": 0.07792524047242178, + "grad_norm": 1.7186362743377686, + "learning_rate": 1.5580159433126664e-05, + "loss": 1.1697, + "step": 1760 + }, + { + "epoch": 0.07796951617723565, + "grad_norm": 1.6083180904388428, + "learning_rate": 1.558901682905226e-05, + "loss": 1.0947, + "step": 1761 + }, + { + "epoch": 0.07801379188204952, + "grad_norm": 2.576692581176758, + "learning_rate": 1.5597874224977857e-05, + "loss": 1.0487, + "step": 1762 + }, + { + "epoch": 0.0780580675868634, + "grad_norm": 4.226240634918213, + "learning_rate": 1.5606731620903454e-05, + "loss": 1.0924, + "step": 1763 + }, + { + "epoch": 0.07810234329167727, + "grad_norm": 1.8671499490737915, + "learning_rate": 1.5615589016829054e-05, + "loss": 1.0515, + "step": 1764 + }, + { + "epoch": 0.07814661899649115, + "grad_norm": 1.6236042976379395, + "learning_rate": 1.5624446412754654e-05, + "loss": 1.0493, + "step": 1765 + }, + { + "epoch": 0.07819089470130502, + "grad_norm": 1.596868634223938, + "learning_rate": 1.563330380868025e-05, + "loss": 1.1062, + "step": 1766 + }, + { + "epoch": 0.07823517040611891, + "grad_norm": 1.7652039527893066, + "learning_rate": 1.5642161204605847e-05, + "loss": 1.1196, + "step": 1767 + }, + { + "epoch": 0.07827944611093278, + "grad_norm": 1.5124958753585815, + "learning_rate": 1.5651018600531444e-05, + "loss": 1.0647, + "step": 1768 + }, + { + "epoch": 0.07832372181574665, + "grad_norm": 2.1084845066070557, + "learning_rate": 1.5659875996457044e-05, + "loss": 1.1873, + "step": 1769 + }, + { + "epoch": 0.07836799752056053, + "grad_norm": 1.572322964668274, + "learning_rate": 1.566873339238264e-05, + "loss": 1.0415, + "step": 1770 + }, + { + "epoch": 0.0784122732253744, + "grad_norm": 1.5333049297332764, + "learning_rate": 1.5677590788308237e-05, + "loss": 1.0208, + "step": 1771 + }, + { + "epoch": 0.07845654893018829, + "grad_norm": 1.5850657224655151, + "learning_rate": 1.5686448184233834e-05, + "loss": 1.0998, + "step": 1772 + }, + { + "epoch": 0.07850082463500216, + "grad_norm": 1.512221097946167, + "learning_rate": 1.5695305580159434e-05, + "loss": 1.0886, + "step": 1773 + }, + { + "epoch": 0.07854510033981603, + "grad_norm": 2.3312180042266846, + "learning_rate": 1.5704162976085034e-05, + "loss": 1.1559, + "step": 1774 + }, + { + "epoch": 0.07858937604462991, + "grad_norm": 1.6642481088638306, + "learning_rate": 1.571302037201063e-05, + "loss": 1.0805, + "step": 1775 + }, + { + "epoch": 0.07863365174944378, + "grad_norm": 1.3041901588439941, + "learning_rate": 1.5721877767936228e-05, + "loss": 1.0228, + "step": 1776 + }, + { + "epoch": 0.07867792745425767, + "grad_norm": 1.737097144126892, + "learning_rate": 1.5730735163861824e-05, + "loss": 1.0837, + "step": 1777 + }, + { + "epoch": 0.07872220315907154, + "grad_norm": 1.7831312417984009, + "learning_rate": 1.5739592559787425e-05, + "loss": 1.1223, + "step": 1778 + }, + { + "epoch": 0.07876647886388541, + "grad_norm": 1.484985113143921, + "learning_rate": 1.574844995571302e-05, + "loss": 1.0973, + "step": 1779 + }, + { + "epoch": 0.07881075456869929, + "grad_norm": 1.8360908031463623, + "learning_rate": 1.575730735163862e-05, + "loss": 1.1284, + "step": 1780 + }, + { + "epoch": 0.07885503027351316, + "grad_norm": 1.6712195873260498, + "learning_rate": 1.5766164747564218e-05, + "loss": 1.0517, + "step": 1781 + }, + { + "epoch": 0.07889930597832705, + "grad_norm": 1.7185548543930054, + "learning_rate": 1.5775022143489815e-05, + "loss": 1.097, + "step": 1782 + }, + { + "epoch": 0.07894358168314092, + "grad_norm": 2.017184019088745, + "learning_rate": 1.578387953941541e-05, + "loss": 1.149, + "step": 1783 + }, + { + "epoch": 0.07898785738795479, + "grad_norm": 1.39376699924469, + "learning_rate": 1.579273693534101e-05, + "loss": 0.9786, + "step": 1784 + }, + { + "epoch": 0.07903213309276867, + "grad_norm": 2.337843894958496, + "learning_rate": 1.5801594331266608e-05, + "loss": 1.0611, + "step": 1785 + }, + { + "epoch": 0.07907640879758254, + "grad_norm": 1.359018325805664, + "learning_rate": 1.5810451727192205e-05, + "loss": 1.0126, + "step": 1786 + }, + { + "epoch": 0.07912068450239643, + "grad_norm": 5.649447917938232, + "learning_rate": 1.5819309123117805e-05, + "loss": 1.0695, + "step": 1787 + }, + { + "epoch": 0.0791649602072103, + "grad_norm": 1.253894567489624, + "learning_rate": 1.58281665190434e-05, + "loss": 0.9499, + "step": 1788 + }, + { + "epoch": 0.07920923591202418, + "grad_norm": 1.6381584405899048, + "learning_rate": 1.5837023914969002e-05, + "loss": 1.092, + "step": 1789 + }, + { + "epoch": 0.07925351161683805, + "grad_norm": 1.5871222019195557, + "learning_rate": 1.58458813108946e-05, + "loss": 1.0208, + "step": 1790 + }, + { + "epoch": 0.07929778732165192, + "grad_norm": 1.3530281782150269, + "learning_rate": 1.5854738706820195e-05, + "loss": 0.9647, + "step": 1791 + }, + { + "epoch": 0.0793420630264658, + "grad_norm": 1.5984970331192017, + "learning_rate": 1.5863596102745792e-05, + "loss": 1.0048, + "step": 1792 + }, + { + "epoch": 0.07938633873127968, + "grad_norm": 1.479561686515808, + "learning_rate": 1.5872453498671392e-05, + "loss": 1.1497, + "step": 1793 + }, + { + "epoch": 0.07943061443609356, + "grad_norm": 1.3645261526107788, + "learning_rate": 1.5881310894596992e-05, + "loss": 0.9887, + "step": 1794 + }, + { + "epoch": 0.07947489014090743, + "grad_norm": 1.4581646919250488, + "learning_rate": 1.589016829052259e-05, + "loss": 1.0334, + "step": 1795 + }, + { + "epoch": 0.0795191658457213, + "grad_norm": 1.3505579233169556, + "learning_rate": 1.5899025686448185e-05, + "loss": 1.0188, + "step": 1796 + }, + { + "epoch": 0.07956344155053519, + "grad_norm": 1.4855315685272217, + "learning_rate": 1.5907883082373782e-05, + "loss": 1.0158, + "step": 1797 + }, + { + "epoch": 0.07960771725534906, + "grad_norm": 1.4563299417495728, + "learning_rate": 1.5916740478299382e-05, + "loss": 1.0846, + "step": 1798 + }, + { + "epoch": 0.07965199296016294, + "grad_norm": 1.6134778261184692, + "learning_rate": 1.592559787422498e-05, + "loss": 1.0439, + "step": 1799 + }, + { + "epoch": 0.07969626866497681, + "grad_norm": 1.526430368423462, + "learning_rate": 1.5934455270150576e-05, + "loss": 1.0753, + "step": 1800 + }, + { + "epoch": 0.07974054436979068, + "grad_norm": 1.9763705730438232, + "learning_rate": 1.5943312666076172e-05, + "loss": 1.1417, + "step": 1801 + }, + { + "epoch": 0.07978482007460456, + "grad_norm": 1.441084861755371, + "learning_rate": 1.5952170062001772e-05, + "loss": 1.0271, + "step": 1802 + }, + { + "epoch": 0.07982909577941844, + "grad_norm": 1.61093008518219, + "learning_rate": 1.5961027457927372e-05, + "loss": 1.0254, + "step": 1803 + }, + { + "epoch": 0.07987337148423232, + "grad_norm": 1.581132411956787, + "learning_rate": 1.596988485385297e-05, + "loss": 1.0641, + "step": 1804 + }, + { + "epoch": 0.07991764718904619, + "grad_norm": 1.4161851406097412, + "learning_rate": 1.5978742249778566e-05, + "loss": 1.0661, + "step": 1805 + }, + { + "epoch": 0.07996192289386006, + "grad_norm": 1.3668721914291382, + "learning_rate": 1.5987599645704163e-05, + "loss": 1.0347, + "step": 1806 + }, + { + "epoch": 0.08000619859867394, + "grad_norm": 1.90378737449646, + "learning_rate": 1.5996457041629763e-05, + "loss": 1.1122, + "step": 1807 + }, + { + "epoch": 0.08005047430348781, + "grad_norm": 1.823300838470459, + "learning_rate": 1.600531443755536e-05, + "loss": 1.0079, + "step": 1808 + }, + { + "epoch": 0.0800947500083017, + "grad_norm": 1.5190598964691162, + "learning_rate": 1.601417183348096e-05, + "loss": 1.1127, + "step": 1809 + }, + { + "epoch": 0.08013902571311557, + "grad_norm": 1.4604823589324951, + "learning_rate": 1.6023029229406556e-05, + "loss": 1.0779, + "step": 1810 + }, + { + "epoch": 0.08018330141792945, + "grad_norm": 1.6182410717010498, + "learning_rate": 1.6031886625332153e-05, + "loss": 1.1469, + "step": 1811 + }, + { + "epoch": 0.08022757712274332, + "grad_norm": 1.6850191354751587, + "learning_rate": 1.6040744021257753e-05, + "loss": 1.1207, + "step": 1812 + }, + { + "epoch": 0.0802718528275572, + "grad_norm": 2.2580044269561768, + "learning_rate": 1.604960141718335e-05, + "loss": 1.2263, + "step": 1813 + }, + { + "epoch": 0.08031612853237108, + "grad_norm": 1.4820215702056885, + "learning_rate": 1.605845881310895e-05, + "loss": 1.0458, + "step": 1814 + }, + { + "epoch": 0.08036040423718495, + "grad_norm": 1.9092422723770142, + "learning_rate": 1.6067316209034546e-05, + "loss": 1.1155, + "step": 1815 + }, + { + "epoch": 0.08040467994199883, + "grad_norm": 1.5354830026626587, + "learning_rate": 1.6076173604960143e-05, + "loss": 1.0569, + "step": 1816 + }, + { + "epoch": 0.0804489556468127, + "grad_norm": 1.7948293685913086, + "learning_rate": 1.608503100088574e-05, + "loss": 1.0784, + "step": 1817 + }, + { + "epoch": 0.08049323135162657, + "grad_norm": 1.6625338792800903, + "learning_rate": 1.609388839681134e-05, + "loss": 1.1735, + "step": 1818 + }, + { + "epoch": 0.08053750705644046, + "grad_norm": 1.3722985982894897, + "learning_rate": 1.6102745792736937e-05, + "loss": 1.0385, + "step": 1819 + }, + { + "epoch": 0.08058178276125433, + "grad_norm": 1.454085350036621, + "learning_rate": 1.6111603188662533e-05, + "loss": 1.0229, + "step": 1820 + }, + { + "epoch": 0.08062605846606821, + "grad_norm": 1.597680687904358, + "learning_rate": 1.6120460584588133e-05, + "loss": 1.1648, + "step": 1821 + }, + { + "epoch": 0.08067033417088208, + "grad_norm": 1.5282042026519775, + "learning_rate": 1.612931798051373e-05, + "loss": 1.0555, + "step": 1822 + }, + { + "epoch": 0.08071460987569595, + "grad_norm": 1.5126001834869385, + "learning_rate": 1.613817537643933e-05, + "loss": 1.0959, + "step": 1823 + }, + { + "epoch": 0.08075888558050984, + "grad_norm": 1.4114892482757568, + "learning_rate": 1.6147032772364927e-05, + "loss": 1.0015, + "step": 1824 + }, + { + "epoch": 0.08080316128532371, + "grad_norm": 1.603990912437439, + "learning_rate": 1.6155890168290524e-05, + "loss": 1.1107, + "step": 1825 + }, + { + "epoch": 0.08084743699013759, + "grad_norm": 1.4143683910369873, + "learning_rate": 1.616474756421612e-05, + "loss": 0.9973, + "step": 1826 + }, + { + "epoch": 0.08089171269495146, + "grad_norm": 1.3319597244262695, + "learning_rate": 1.617360496014172e-05, + "loss": 0.9545, + "step": 1827 + }, + { + "epoch": 0.08093598839976533, + "grad_norm": 1.6884897947311401, + "learning_rate": 1.6182462356067317e-05, + "loss": 0.9885, + "step": 1828 + }, + { + "epoch": 0.08098026410457922, + "grad_norm": 1.5565199851989746, + "learning_rate": 1.6191319751992917e-05, + "loss": 1.0292, + "step": 1829 + }, + { + "epoch": 0.08102453980939309, + "grad_norm": 1.6033447980880737, + "learning_rate": 1.6200177147918514e-05, + "loss": 1.0942, + "step": 1830 + }, + { + "epoch": 0.08106881551420697, + "grad_norm": 1.3930888175964355, + "learning_rate": 1.620903454384411e-05, + "loss": 1.1018, + "step": 1831 + }, + { + "epoch": 0.08111309121902084, + "grad_norm": 1.5927528142929077, + "learning_rate": 1.621789193976971e-05, + "loss": 1.123, + "step": 1832 + }, + { + "epoch": 0.08115736692383473, + "grad_norm": 1.5688594579696655, + "learning_rate": 1.6226749335695307e-05, + "loss": 1.0688, + "step": 1833 + }, + { + "epoch": 0.0812016426286486, + "grad_norm": 1.6845061779022217, + "learning_rate": 1.6235606731620904e-05, + "loss": 1.0806, + "step": 1834 + }, + { + "epoch": 0.08124591833346247, + "grad_norm": 1.6765077114105225, + "learning_rate": 1.62444641275465e-05, + "loss": 1.0195, + "step": 1835 + }, + { + "epoch": 0.08129019403827635, + "grad_norm": 1.5421050786972046, + "learning_rate": 1.62533215234721e-05, + "loss": 1.0348, + "step": 1836 + }, + { + "epoch": 0.08133446974309022, + "grad_norm": 1.3686193227767944, + "learning_rate": 1.6262178919397697e-05, + "loss": 0.9555, + "step": 1837 + }, + { + "epoch": 0.0813787454479041, + "grad_norm": 1.5401360988616943, + "learning_rate": 1.6271036315323298e-05, + "loss": 0.9961, + "step": 1838 + }, + { + "epoch": 0.08142302115271798, + "grad_norm": 1.8546544313430786, + "learning_rate": 1.6279893711248894e-05, + "loss": 1.133, + "step": 1839 + }, + { + "epoch": 0.08146729685753185, + "grad_norm": 1.4921624660491943, + "learning_rate": 1.628875110717449e-05, + "loss": 1.0335, + "step": 1840 + }, + { + "epoch": 0.08151157256234573, + "grad_norm": 1.6105842590332031, + "learning_rate": 1.629760850310009e-05, + "loss": 1.0363, + "step": 1841 + }, + { + "epoch": 0.0815558482671596, + "grad_norm": 1.367634654045105, + "learning_rate": 1.6306465899025688e-05, + "loss": 1.032, + "step": 1842 + }, + { + "epoch": 0.08160012397197348, + "grad_norm": 1.7989754676818848, + "learning_rate": 1.6315323294951288e-05, + "loss": 1.1387, + "step": 1843 + }, + { + "epoch": 0.08164439967678735, + "grad_norm": 1.6338562965393066, + "learning_rate": 1.6324180690876884e-05, + "loss": 1.0062, + "step": 1844 + }, + { + "epoch": 0.08168867538160123, + "grad_norm": 1.802771806716919, + "learning_rate": 1.633303808680248e-05, + "loss": 1.1314, + "step": 1845 + }, + { + "epoch": 0.08173295108641511, + "grad_norm": 1.4379384517669678, + "learning_rate": 1.6341895482728078e-05, + "loss": 1.0242, + "step": 1846 + }, + { + "epoch": 0.08177722679122898, + "grad_norm": 1.61032235622406, + "learning_rate": 1.6350752878653678e-05, + "loss": 1.0528, + "step": 1847 + }, + { + "epoch": 0.08182150249604286, + "grad_norm": 1.678940773010254, + "learning_rate": 1.6359610274579275e-05, + "loss": 1.1668, + "step": 1848 + }, + { + "epoch": 0.08186577820085673, + "grad_norm": 1.5305455923080444, + "learning_rate": 1.636846767050487e-05, + "loss": 1.0227, + "step": 1849 + }, + { + "epoch": 0.0819100539056706, + "grad_norm": 1.4719946384429932, + "learning_rate": 1.637732506643047e-05, + "loss": 1.0099, + "step": 1850 + }, + { + "epoch": 0.08195432961048449, + "grad_norm": 1.6524360179901123, + "learning_rate": 1.6386182462356068e-05, + "loss": 1.0113, + "step": 1851 + }, + { + "epoch": 0.08199860531529836, + "grad_norm": 1.3211417198181152, + "learning_rate": 1.6395039858281668e-05, + "loss": 0.9473, + "step": 1852 + }, + { + "epoch": 0.08204288102011224, + "grad_norm": 1.5694506168365479, + "learning_rate": 1.6403897254207265e-05, + "loss": 1.1215, + "step": 1853 + }, + { + "epoch": 0.08208715672492611, + "grad_norm": 1.5245403051376343, + "learning_rate": 1.641275465013286e-05, + "loss": 1.0581, + "step": 1854 + }, + { + "epoch": 0.08213143242974, + "grad_norm": 1.4570610523223877, + "learning_rate": 1.6421612046058458e-05, + "loss": 1.089, + "step": 1855 + }, + { + "epoch": 0.08217570813455387, + "grad_norm": 1.3612236976623535, + "learning_rate": 1.643046944198406e-05, + "loss": 1.0156, + "step": 1856 + }, + { + "epoch": 0.08221998383936774, + "grad_norm": 1.6991050243377686, + "learning_rate": 1.6439326837909655e-05, + "loss": 1.0961, + "step": 1857 + }, + { + "epoch": 0.08226425954418162, + "grad_norm": 1.4560937881469727, + "learning_rate": 1.6448184233835255e-05, + "loss": 1.0595, + "step": 1858 + }, + { + "epoch": 0.08230853524899549, + "grad_norm": 1.5075125694274902, + "learning_rate": 1.6457041629760852e-05, + "loss": 1.0917, + "step": 1859 + }, + { + "epoch": 0.08235281095380938, + "grad_norm": 1.4980109930038452, + "learning_rate": 1.646589902568645e-05, + "loss": 1.0769, + "step": 1860 + }, + { + "epoch": 0.08239708665862325, + "grad_norm": 2.4678916931152344, + "learning_rate": 1.647475642161205e-05, + "loss": 1.1692, + "step": 1861 + }, + { + "epoch": 0.08244136236343712, + "grad_norm": 1.6690195798873901, + "learning_rate": 1.6483613817537645e-05, + "loss": 1.1285, + "step": 1862 + }, + { + "epoch": 0.082485638068251, + "grad_norm": 1.3937263488769531, + "learning_rate": 1.6492471213463242e-05, + "loss": 1.0354, + "step": 1863 + }, + { + "epoch": 0.08252991377306487, + "grad_norm": 1.8939837217330933, + "learning_rate": 1.650132860938884e-05, + "loss": 1.2708, + "step": 1864 + }, + { + "epoch": 0.08257418947787876, + "grad_norm": 1.2960532903671265, + "learning_rate": 1.651018600531444e-05, + "loss": 0.9222, + "step": 1865 + }, + { + "epoch": 0.08261846518269263, + "grad_norm": 1.4921897649765015, + "learning_rate": 1.6519043401240036e-05, + "loss": 1.0212, + "step": 1866 + }, + { + "epoch": 0.0826627408875065, + "grad_norm": 1.4176037311553955, + "learning_rate": 1.6527900797165636e-05, + "loss": 1.032, + "step": 1867 + }, + { + "epoch": 0.08270701659232038, + "grad_norm": 1.500874400138855, + "learning_rate": 1.6536758193091232e-05, + "loss": 1.0797, + "step": 1868 + }, + { + "epoch": 0.08275129229713425, + "grad_norm": 1.4789339303970337, + "learning_rate": 1.654561558901683e-05, + "loss": 1.0845, + "step": 1869 + }, + { + "epoch": 0.08279556800194814, + "grad_norm": 1.4445197582244873, + "learning_rate": 1.655447298494243e-05, + "loss": 1.1025, + "step": 1870 + }, + { + "epoch": 0.082839843706762, + "grad_norm": 1.4537345170974731, + "learning_rate": 1.6563330380868026e-05, + "loss": 0.9768, + "step": 1871 + }, + { + "epoch": 0.08288411941157588, + "grad_norm": 2.7110178470611572, + "learning_rate": 1.6572187776793626e-05, + "loss": 1.2284, + "step": 1872 + }, + { + "epoch": 0.08292839511638976, + "grad_norm": 1.6558266878128052, + "learning_rate": 1.6581045172719223e-05, + "loss": 0.9885, + "step": 1873 + }, + { + "epoch": 0.08297267082120363, + "grad_norm": 1.4615987539291382, + "learning_rate": 1.658990256864482e-05, + "loss": 0.9838, + "step": 1874 + }, + { + "epoch": 0.08301694652601752, + "grad_norm": 1.8572484254837036, + "learning_rate": 1.6598759964570416e-05, + "loss": 1.0632, + "step": 1875 + }, + { + "epoch": 0.08306122223083139, + "grad_norm": 1.9783411026000977, + "learning_rate": 1.6607617360496016e-05, + "loss": 1.1278, + "step": 1876 + }, + { + "epoch": 0.08310549793564527, + "grad_norm": 1.403784990310669, + "learning_rate": 1.6616474756421616e-05, + "loss": 1.0614, + "step": 1877 + }, + { + "epoch": 0.08314977364045914, + "grad_norm": 2.6871814727783203, + "learning_rate": 1.662533215234721e-05, + "loss": 1.1849, + "step": 1878 + }, + { + "epoch": 0.08319404934527301, + "grad_norm": 1.3599735498428345, + "learning_rate": 1.663418954827281e-05, + "loss": 0.9729, + "step": 1879 + }, + { + "epoch": 0.0832383250500869, + "grad_norm": 2.742466926574707, + "learning_rate": 1.6643046944198406e-05, + "loss": 1.0973, + "step": 1880 + }, + { + "epoch": 0.08328260075490077, + "grad_norm": 1.5280909538269043, + "learning_rate": 1.6651904340124006e-05, + "loss": 1.0089, + "step": 1881 + }, + { + "epoch": 0.08332687645971465, + "grad_norm": 1.6493515968322754, + "learning_rate": 1.6660761736049603e-05, + "loss": 1.082, + "step": 1882 + }, + { + "epoch": 0.08337115216452852, + "grad_norm": 1.7493692636489868, + "learning_rate": 1.66696191319752e-05, + "loss": 1.0098, + "step": 1883 + }, + { + "epoch": 0.08341542786934239, + "grad_norm": 3.2923312187194824, + "learning_rate": 1.6678476527900796e-05, + "loss": 1.1075, + "step": 1884 + }, + { + "epoch": 0.08345970357415627, + "grad_norm": 1.547555685043335, + "learning_rate": 1.6687333923826396e-05, + "loss": 1.0715, + "step": 1885 + }, + { + "epoch": 0.08350397927897014, + "grad_norm": 1.6168131828308105, + "learning_rate": 1.6696191319751997e-05, + "loss": 1.0925, + "step": 1886 + }, + { + "epoch": 0.08354825498378403, + "grad_norm": 1.5336639881134033, + "learning_rate": 1.6705048715677593e-05, + "loss": 1.0143, + "step": 1887 + }, + { + "epoch": 0.0835925306885979, + "grad_norm": 1.350684404373169, + "learning_rate": 1.671390611160319e-05, + "loss": 0.9287, + "step": 1888 + }, + { + "epoch": 0.08363680639341177, + "grad_norm": 1.9129910469055176, + "learning_rate": 1.6722763507528787e-05, + "loss": 1.1325, + "step": 1889 + }, + { + "epoch": 0.08368108209822565, + "grad_norm": 1.5541648864746094, + "learning_rate": 1.6731620903454387e-05, + "loss": 1.0243, + "step": 1890 + }, + { + "epoch": 0.08372535780303952, + "grad_norm": 1.4959619045257568, + "learning_rate": 1.6740478299379983e-05, + "loss": 1.052, + "step": 1891 + }, + { + "epoch": 0.08376963350785341, + "grad_norm": 1.5495951175689697, + "learning_rate": 1.6749335695305584e-05, + "loss": 1.0641, + "step": 1892 + }, + { + "epoch": 0.08381390921266728, + "grad_norm": 1.6429373025894165, + "learning_rate": 1.675819309123118e-05, + "loss": 1.0083, + "step": 1893 + }, + { + "epoch": 0.08385818491748115, + "grad_norm": 1.4211843013763428, + "learning_rate": 1.6767050487156777e-05, + "loss": 1.0165, + "step": 1894 + }, + { + "epoch": 0.08390246062229503, + "grad_norm": 1.3802273273468018, + "learning_rate": 1.6775907883082374e-05, + "loss": 0.9675, + "step": 1895 + }, + { + "epoch": 0.0839467363271089, + "grad_norm": 1.506365180015564, + "learning_rate": 1.6784765279007974e-05, + "loss": 1.0396, + "step": 1896 + }, + { + "epoch": 0.08399101203192279, + "grad_norm": 1.375544786453247, + "learning_rate": 1.679362267493357e-05, + "loss": 1.0213, + "step": 1897 + }, + { + "epoch": 0.08403528773673666, + "grad_norm": 1.412147879600525, + "learning_rate": 1.6802480070859167e-05, + "loss": 1.0136, + "step": 1898 + }, + { + "epoch": 0.08407956344155053, + "grad_norm": 1.9242578744888306, + "learning_rate": 1.6811337466784767e-05, + "loss": 1.0959, + "step": 1899 + }, + { + "epoch": 0.08412383914636441, + "grad_norm": 1.5428929328918457, + "learning_rate": 1.6820194862710364e-05, + "loss": 1.0302, + "step": 1900 + }, + { + "epoch": 0.08416811485117828, + "grad_norm": 1.676459789276123, + "learning_rate": 1.6829052258635964e-05, + "loss": 1.0411, + "step": 1901 + }, + { + "epoch": 0.08421239055599217, + "grad_norm": 1.9072264432907104, + "learning_rate": 1.683790965456156e-05, + "loss": 1.1246, + "step": 1902 + }, + { + "epoch": 0.08425666626080604, + "grad_norm": 1.2907075881958008, + "learning_rate": 1.6846767050487157e-05, + "loss": 0.9532, + "step": 1903 + }, + { + "epoch": 0.08430094196561992, + "grad_norm": 1.7712785005569458, + "learning_rate": 1.6855624446412754e-05, + "loss": 1.0301, + "step": 1904 + }, + { + "epoch": 0.08434521767043379, + "grad_norm": 1.455049991607666, + "learning_rate": 1.6864481842338354e-05, + "loss": 1.0789, + "step": 1905 + }, + { + "epoch": 0.08438949337524766, + "grad_norm": 1.417467474937439, + "learning_rate": 1.6873339238263954e-05, + "loss": 0.9383, + "step": 1906 + }, + { + "epoch": 0.08443376908006155, + "grad_norm": 1.3064801692962646, + "learning_rate": 1.688219663418955e-05, + "loss": 0.9475, + "step": 1907 + }, + { + "epoch": 0.08447804478487542, + "grad_norm": 1.6494640111923218, + "learning_rate": 1.6891054030115148e-05, + "loss": 1.0686, + "step": 1908 + }, + { + "epoch": 0.0845223204896893, + "grad_norm": 1.8318480253219604, + "learning_rate": 1.6899911426040744e-05, + "loss": 1.053, + "step": 1909 + }, + { + "epoch": 0.08456659619450317, + "grad_norm": 1.5130870342254639, + "learning_rate": 1.6908768821966344e-05, + "loss": 1.063, + "step": 1910 + }, + { + "epoch": 0.08461087189931704, + "grad_norm": 1.6217632293701172, + "learning_rate": 1.691762621789194e-05, + "loss": 1.0573, + "step": 1911 + }, + { + "epoch": 0.08465514760413093, + "grad_norm": 1.5994192361831665, + "learning_rate": 1.6926483613817538e-05, + "loss": 1.0602, + "step": 1912 + }, + { + "epoch": 0.0846994233089448, + "grad_norm": 1.7475816011428833, + "learning_rate": 1.6935341009743134e-05, + "loss": 1.1672, + "step": 1913 + }, + { + "epoch": 0.08474369901375868, + "grad_norm": 1.4299228191375732, + "learning_rate": 1.6944198405668735e-05, + "loss": 0.987, + "step": 1914 + }, + { + "epoch": 0.08478797471857255, + "grad_norm": 1.465525507926941, + "learning_rate": 1.6953055801594335e-05, + "loss": 1.0069, + "step": 1915 + }, + { + "epoch": 0.08483225042338642, + "grad_norm": 1.4605764150619507, + "learning_rate": 1.696191319751993e-05, + "loss": 1.0002, + "step": 1916 + }, + { + "epoch": 0.0848765261282003, + "grad_norm": 1.491779088973999, + "learning_rate": 1.6970770593445528e-05, + "loss": 1.0515, + "step": 1917 + }, + { + "epoch": 0.08492080183301418, + "grad_norm": 1.370284080505371, + "learning_rate": 1.6979627989371125e-05, + "loss": 1.0571, + "step": 1918 + }, + { + "epoch": 0.08496507753782806, + "grad_norm": 2.0205018520355225, + "learning_rate": 1.6988485385296725e-05, + "loss": 1.1736, + "step": 1919 + }, + { + "epoch": 0.08500935324264193, + "grad_norm": 1.3712730407714844, + "learning_rate": 1.699734278122232e-05, + "loss": 1.0713, + "step": 1920 + }, + { + "epoch": 0.0850536289474558, + "grad_norm": 1.3879395723342896, + "learning_rate": 1.700620017714792e-05, + "loss": 0.9786, + "step": 1921 + }, + { + "epoch": 0.08509790465226968, + "grad_norm": 1.4197362661361694, + "learning_rate": 1.701505757307352e-05, + "loss": 0.9969, + "step": 1922 + }, + { + "epoch": 0.08514218035708356, + "grad_norm": 1.5541247129440308, + "learning_rate": 1.7023914968999115e-05, + "loss": 1.1274, + "step": 1923 + }, + { + "epoch": 0.08518645606189744, + "grad_norm": 1.3581469058990479, + "learning_rate": 1.7032772364924715e-05, + "loss": 0.9754, + "step": 1924 + }, + { + "epoch": 0.08523073176671131, + "grad_norm": 1.4309170246124268, + "learning_rate": 1.7041629760850312e-05, + "loss": 1.0298, + "step": 1925 + }, + { + "epoch": 0.0852750074715252, + "grad_norm": 1.91091787815094, + "learning_rate": 1.705048715677591e-05, + "loss": 1.1231, + "step": 1926 + }, + { + "epoch": 0.08531928317633906, + "grad_norm": 1.823677897453308, + "learning_rate": 1.7059344552701505e-05, + "loss": 1.1453, + "step": 1927 + }, + { + "epoch": 0.08536355888115293, + "grad_norm": 1.4370815753936768, + "learning_rate": 1.7068201948627105e-05, + "loss": 1.0907, + "step": 1928 + }, + { + "epoch": 0.08540783458596682, + "grad_norm": 1.932759165763855, + "learning_rate": 1.7077059344552702e-05, + "loss": 1.0139, + "step": 1929 + }, + { + "epoch": 0.08545211029078069, + "grad_norm": 2.08828067779541, + "learning_rate": 1.7085916740478302e-05, + "loss": 1.2265, + "step": 1930 + }, + { + "epoch": 0.08549638599559457, + "grad_norm": 1.3155213594436646, + "learning_rate": 1.70947741364039e-05, + "loss": 0.9799, + "step": 1931 + }, + { + "epoch": 0.08554066170040844, + "grad_norm": 1.7981046438217163, + "learning_rate": 1.7103631532329495e-05, + "loss": 1.0224, + "step": 1932 + }, + { + "epoch": 0.08558493740522231, + "grad_norm": 1.9474875926971436, + "learning_rate": 1.7112488928255096e-05, + "loss": 1.1578, + "step": 1933 + }, + { + "epoch": 0.0856292131100362, + "grad_norm": 1.5589121580123901, + "learning_rate": 1.7121346324180692e-05, + "loss": 1.0632, + "step": 1934 + }, + { + "epoch": 0.08567348881485007, + "grad_norm": 1.8004553318023682, + "learning_rate": 1.7130203720106292e-05, + "loss": 1.1965, + "step": 1935 + }, + { + "epoch": 0.08571776451966395, + "grad_norm": 1.2739262580871582, + "learning_rate": 1.713906111603189e-05, + "loss": 0.9765, + "step": 1936 + }, + { + "epoch": 0.08576204022447782, + "grad_norm": 1.4683198928833008, + "learning_rate": 1.7147918511957486e-05, + "loss": 1.0803, + "step": 1937 + }, + { + "epoch": 0.0858063159292917, + "grad_norm": 1.3568065166473389, + "learning_rate": 1.7156775907883082e-05, + "loss": 0.9536, + "step": 1938 + }, + { + "epoch": 0.08585059163410558, + "grad_norm": 1.3467473983764648, + "learning_rate": 1.7165633303808682e-05, + "loss": 1.0194, + "step": 1939 + }, + { + "epoch": 0.08589486733891945, + "grad_norm": 1.4719834327697754, + "learning_rate": 1.717449069973428e-05, + "loss": 1.0376, + "step": 1940 + }, + { + "epoch": 0.08593914304373333, + "grad_norm": 1.8419610261917114, + "learning_rate": 1.7183348095659876e-05, + "loss": 1.0618, + "step": 1941 + }, + { + "epoch": 0.0859834187485472, + "grad_norm": 1.7236605882644653, + "learning_rate": 1.7192205491585476e-05, + "loss": 1.1095, + "step": 1942 + }, + { + "epoch": 0.08602769445336107, + "grad_norm": 1.5351401567459106, + "learning_rate": 1.7201062887511073e-05, + "loss": 1.0864, + "step": 1943 + }, + { + "epoch": 0.08607197015817496, + "grad_norm": 1.8349705934524536, + "learning_rate": 1.7209920283436673e-05, + "loss": 1.143, + "step": 1944 + }, + { + "epoch": 0.08611624586298883, + "grad_norm": 1.3783717155456543, + "learning_rate": 1.721877767936227e-05, + "loss": 0.9953, + "step": 1945 + }, + { + "epoch": 0.08616052156780271, + "grad_norm": 1.6862530708312988, + "learning_rate": 1.7227635075287866e-05, + "loss": 1.1135, + "step": 1946 + }, + { + "epoch": 0.08620479727261658, + "grad_norm": 1.3903398513793945, + "learning_rate": 1.7236492471213463e-05, + "loss": 0.9836, + "step": 1947 + }, + { + "epoch": 0.08624907297743047, + "grad_norm": 1.4231846332550049, + "learning_rate": 1.7245349867139063e-05, + "loss": 1.0051, + "step": 1948 + }, + { + "epoch": 0.08629334868224434, + "grad_norm": 1.7122687101364136, + "learning_rate": 1.725420726306466e-05, + "loss": 1.1141, + "step": 1949 + }, + { + "epoch": 0.0863376243870582, + "grad_norm": 1.4830527305603027, + "learning_rate": 1.726306465899026e-05, + "loss": 1.032, + "step": 1950 + }, + { + "epoch": 0.08638190009187209, + "grad_norm": 1.3582262992858887, + "learning_rate": 1.7271922054915856e-05, + "loss": 1.0531, + "step": 1951 + }, + { + "epoch": 0.08642617579668596, + "grad_norm": 1.644674301147461, + "learning_rate": 1.7280779450841453e-05, + "loss": 1.122, + "step": 1952 + }, + { + "epoch": 0.08647045150149985, + "grad_norm": 1.4063063859939575, + "learning_rate": 1.7289636846767053e-05, + "loss": 1.0016, + "step": 1953 + }, + { + "epoch": 0.08651472720631372, + "grad_norm": 1.5522204637527466, + "learning_rate": 1.729849424269265e-05, + "loss": 1.0179, + "step": 1954 + }, + { + "epoch": 0.08655900291112759, + "grad_norm": 1.634239912033081, + "learning_rate": 1.730735163861825e-05, + "loss": 1.1085, + "step": 1955 + }, + { + "epoch": 0.08660327861594147, + "grad_norm": 1.6410367488861084, + "learning_rate": 1.7316209034543843e-05, + "loss": 1.1678, + "step": 1956 + }, + { + "epoch": 0.08664755432075534, + "grad_norm": 1.6255265474319458, + "learning_rate": 1.7325066430469443e-05, + "loss": 1.0738, + "step": 1957 + }, + { + "epoch": 0.08669183002556922, + "grad_norm": 1.692840337753296, + "learning_rate": 1.733392382639504e-05, + "loss": 1.093, + "step": 1958 + }, + { + "epoch": 0.0867361057303831, + "grad_norm": 1.6515614986419678, + "learning_rate": 1.734278122232064e-05, + "loss": 1.1147, + "step": 1959 + }, + { + "epoch": 0.08678038143519697, + "grad_norm": 1.617475152015686, + "learning_rate": 1.7351638618246237e-05, + "loss": 1.1586, + "step": 1960 + }, + { + "epoch": 0.08682465714001085, + "grad_norm": 1.5205026865005493, + "learning_rate": 1.7360496014171834e-05, + "loss": 1.0456, + "step": 1961 + }, + { + "epoch": 0.08686893284482472, + "grad_norm": 1.6654103994369507, + "learning_rate": 1.7369353410097434e-05, + "loss": 1.0174, + "step": 1962 + }, + { + "epoch": 0.0869132085496386, + "grad_norm": 1.5487661361694336, + "learning_rate": 1.737821080602303e-05, + "loss": 1.1209, + "step": 1963 + }, + { + "epoch": 0.08695748425445247, + "grad_norm": 1.4474778175354004, + "learning_rate": 1.738706820194863e-05, + "loss": 1.0215, + "step": 1964 + }, + { + "epoch": 0.08700175995926634, + "grad_norm": 1.398457646369934, + "learning_rate": 1.7395925597874227e-05, + "loss": 1.0399, + "step": 1965 + }, + { + "epoch": 0.08704603566408023, + "grad_norm": 1.5566022396087646, + "learning_rate": 1.7404782993799824e-05, + "loss": 1.1133, + "step": 1966 + }, + { + "epoch": 0.0870903113688941, + "grad_norm": 1.767931342124939, + "learning_rate": 1.741364038972542e-05, + "loss": 1.1568, + "step": 1967 + }, + { + "epoch": 0.08713458707370798, + "grad_norm": 2.301913022994995, + "learning_rate": 1.742249778565102e-05, + "loss": 1.106, + "step": 1968 + }, + { + "epoch": 0.08717886277852185, + "grad_norm": 1.4673497676849365, + "learning_rate": 1.7431355181576617e-05, + "loss": 1.0436, + "step": 1969 + }, + { + "epoch": 0.08722313848333574, + "grad_norm": 1.4829405546188354, + "learning_rate": 1.7440212577502217e-05, + "loss": 1.0143, + "step": 1970 + }, + { + "epoch": 0.08726741418814961, + "grad_norm": 1.5090126991271973, + "learning_rate": 1.7449069973427814e-05, + "loss": 0.9948, + "step": 1971 + }, + { + "epoch": 0.08731168989296348, + "grad_norm": 1.4113390445709229, + "learning_rate": 1.745792736935341e-05, + "loss": 1.0451, + "step": 1972 + }, + { + "epoch": 0.08735596559777736, + "grad_norm": 1.5927375555038452, + "learning_rate": 1.746678476527901e-05, + "loss": 1.1022, + "step": 1973 + }, + { + "epoch": 0.08740024130259123, + "grad_norm": 1.447372555732727, + "learning_rate": 1.7475642161204608e-05, + "loss": 1.0578, + "step": 1974 + }, + { + "epoch": 0.08744451700740512, + "grad_norm": 2.02937650680542, + "learning_rate": 1.7484499557130204e-05, + "loss": 1.1191, + "step": 1975 + }, + { + "epoch": 0.08748879271221899, + "grad_norm": 1.305128812789917, + "learning_rate": 1.74933569530558e-05, + "loss": 0.9649, + "step": 1976 + }, + { + "epoch": 0.08753306841703286, + "grad_norm": 1.3548691272735596, + "learning_rate": 1.75022143489814e-05, + "loss": 0.9647, + "step": 1977 + }, + { + "epoch": 0.08757734412184674, + "grad_norm": 1.3146034479141235, + "learning_rate": 1.7511071744906998e-05, + "loss": 0.9994, + "step": 1978 + }, + { + "epoch": 0.08762161982666061, + "grad_norm": 1.5692322254180908, + "learning_rate": 1.7519929140832598e-05, + "loss": 1.0428, + "step": 1979 + }, + { + "epoch": 0.0876658955314745, + "grad_norm": 1.359392523765564, + "learning_rate": 1.7528786536758195e-05, + "loss": 1.0118, + "step": 1980 + }, + { + "epoch": 0.08771017123628837, + "grad_norm": 1.7068626880645752, + "learning_rate": 1.753764393268379e-05, + "loss": 1.0565, + "step": 1981 + }, + { + "epoch": 0.08775444694110224, + "grad_norm": 1.9845367670059204, + "learning_rate": 1.754650132860939e-05, + "loss": 1.1259, + "step": 1982 + }, + { + "epoch": 0.08779872264591612, + "grad_norm": 1.6934536695480347, + "learning_rate": 1.7555358724534988e-05, + "loss": 1.0997, + "step": 1983 + }, + { + "epoch": 0.08784299835072999, + "grad_norm": 1.4021106958389282, + "learning_rate": 1.7564216120460588e-05, + "loss": 1.0678, + "step": 1984 + }, + { + "epoch": 0.08788727405554388, + "grad_norm": 1.5024065971374512, + "learning_rate": 1.7573073516386185e-05, + "loss": 0.9999, + "step": 1985 + }, + { + "epoch": 0.08793154976035775, + "grad_norm": 2.084974765777588, + "learning_rate": 1.758193091231178e-05, + "loss": 1.1047, + "step": 1986 + }, + { + "epoch": 0.08797582546517162, + "grad_norm": 1.3543307781219482, + "learning_rate": 1.7590788308237378e-05, + "loss": 0.9747, + "step": 1987 + }, + { + "epoch": 0.0880201011699855, + "grad_norm": 1.3590283393859863, + "learning_rate": 1.7599645704162978e-05, + "loss": 0.9411, + "step": 1988 + }, + { + "epoch": 0.08806437687479937, + "grad_norm": 1.5582908391952515, + "learning_rate": 1.7608503100088575e-05, + "loss": 1.1012, + "step": 1989 + }, + { + "epoch": 0.08810865257961326, + "grad_norm": 1.3105236291885376, + "learning_rate": 1.761736049601417e-05, + "loss": 0.9394, + "step": 1990 + }, + { + "epoch": 0.08815292828442713, + "grad_norm": 1.5115370750427246, + "learning_rate": 1.7626217891939772e-05, + "loss": 1.059, + "step": 1991 + }, + { + "epoch": 0.08819720398924101, + "grad_norm": 1.6325860023498535, + "learning_rate": 1.763507528786537e-05, + "loss": 1.05, + "step": 1992 + }, + { + "epoch": 0.08824147969405488, + "grad_norm": 1.4465861320495605, + "learning_rate": 1.764393268379097e-05, + "loss": 1.0449, + "step": 1993 + }, + { + "epoch": 0.08828575539886875, + "grad_norm": 1.2366406917572021, + "learning_rate": 1.7652790079716565e-05, + "loss": 0.9213, + "step": 1994 + }, + { + "epoch": 0.08833003110368264, + "grad_norm": 1.6685118675231934, + "learning_rate": 1.7661647475642162e-05, + "loss": 1.1005, + "step": 1995 + }, + { + "epoch": 0.0883743068084965, + "grad_norm": 1.3515992164611816, + "learning_rate": 1.767050487156776e-05, + "loss": 1.0068, + "step": 1996 + }, + { + "epoch": 0.08841858251331039, + "grad_norm": 1.4223158359527588, + "learning_rate": 1.767936226749336e-05, + "loss": 1.0501, + "step": 1997 + }, + { + "epoch": 0.08846285821812426, + "grad_norm": 1.4145638942718506, + "learning_rate": 1.768821966341896e-05, + "loss": 1.0344, + "step": 1998 + }, + { + "epoch": 0.08850713392293813, + "grad_norm": 1.4626046419143677, + "learning_rate": 1.7697077059344555e-05, + "loss": 1.0598, + "step": 1999 + }, + { + "epoch": 0.08855140962775201, + "grad_norm": 1.5103647708892822, + "learning_rate": 1.7705934455270152e-05, + "loss": 1.0752, + "step": 2000 + }, + { + "epoch": 0.08859568533256588, + "grad_norm": 1.6580262184143066, + "learning_rate": 1.771479185119575e-05, + "loss": 1.1398, + "step": 2001 + }, + { + "epoch": 0.08863996103737977, + "grad_norm": 1.5494259595870972, + "learning_rate": 1.772364924712135e-05, + "loss": 1.0381, + "step": 2002 + }, + { + "epoch": 0.08868423674219364, + "grad_norm": 1.4710466861724854, + "learning_rate": 1.7732506643046946e-05, + "loss": 1.0876, + "step": 2003 + }, + { + "epoch": 0.08872851244700751, + "grad_norm": 2.0629920959472656, + "learning_rate": 1.7741364038972542e-05, + "loss": 1.0075, + "step": 2004 + }, + { + "epoch": 0.0887727881518214, + "grad_norm": 1.6345432996749878, + "learning_rate": 1.775022143489814e-05, + "loss": 1.1774, + "step": 2005 + }, + { + "epoch": 0.08881706385663526, + "grad_norm": 1.367082118988037, + "learning_rate": 1.775907883082374e-05, + "loss": 0.9925, + "step": 2006 + }, + { + "epoch": 0.08886133956144915, + "grad_norm": 1.812800645828247, + "learning_rate": 1.7767936226749336e-05, + "loss": 1.0514, + "step": 2007 + }, + { + "epoch": 0.08890561526626302, + "grad_norm": 1.4069256782531738, + "learning_rate": 1.7776793622674936e-05, + "loss": 1.0187, + "step": 2008 + }, + { + "epoch": 0.08894989097107689, + "grad_norm": 1.2871925830841064, + "learning_rate": 1.7785651018600533e-05, + "loss": 0.9459, + "step": 2009 + }, + { + "epoch": 0.08899416667589077, + "grad_norm": 1.495521903038025, + "learning_rate": 1.779450841452613e-05, + "loss": 1.0771, + "step": 2010 + }, + { + "epoch": 0.08903844238070464, + "grad_norm": 1.22012460231781, + "learning_rate": 1.780336581045173e-05, + "loss": 0.8852, + "step": 2011 + }, + { + "epoch": 0.08908271808551853, + "grad_norm": 1.532533049583435, + "learning_rate": 1.7812223206377326e-05, + "loss": 1.1037, + "step": 2012 + }, + { + "epoch": 0.0891269937903324, + "grad_norm": 1.3866592645645142, + "learning_rate": 1.7821080602302926e-05, + "loss": 1.0078, + "step": 2013 + }, + { + "epoch": 0.08917126949514628, + "grad_norm": 1.6618345975875854, + "learning_rate": 1.7829937998228523e-05, + "loss": 1.0923, + "step": 2014 + }, + { + "epoch": 0.08921554519996015, + "grad_norm": 1.4762448072433472, + "learning_rate": 1.783879539415412e-05, + "loss": 1.0666, + "step": 2015 + }, + { + "epoch": 0.08925982090477402, + "grad_norm": 2.1508891582489014, + "learning_rate": 1.7847652790079716e-05, + "loss": 1.0626, + "step": 2016 + }, + { + "epoch": 0.08930409660958791, + "grad_norm": 2.2149569988250732, + "learning_rate": 1.7856510186005316e-05, + "loss": 0.9637, + "step": 2017 + }, + { + "epoch": 0.08934837231440178, + "grad_norm": 1.4443638324737549, + "learning_rate": 1.7865367581930913e-05, + "loss": 0.9663, + "step": 2018 + }, + { + "epoch": 0.08939264801921566, + "grad_norm": 1.3940927982330322, + "learning_rate": 1.787422497785651e-05, + "loss": 0.9689, + "step": 2019 + }, + { + "epoch": 0.08943692372402953, + "grad_norm": 1.952835202217102, + "learning_rate": 1.788308237378211e-05, + "loss": 1.2276, + "step": 2020 + }, + { + "epoch": 0.0894811994288434, + "grad_norm": 1.577518343925476, + "learning_rate": 1.7891939769707707e-05, + "loss": 0.991, + "step": 2021 + }, + { + "epoch": 0.08952547513365729, + "grad_norm": 1.3546124696731567, + "learning_rate": 1.7900797165633307e-05, + "loss": 1.1089, + "step": 2022 + }, + { + "epoch": 0.08956975083847116, + "grad_norm": 2.587480068206787, + "learning_rate": 1.7909654561558903e-05, + "loss": 1.0968, + "step": 2023 + }, + { + "epoch": 0.08961402654328504, + "grad_norm": 1.64358389377594, + "learning_rate": 1.79185119574845e-05, + "loss": 1.112, + "step": 2024 + }, + { + "epoch": 0.08965830224809891, + "grad_norm": 1.8228888511657715, + "learning_rate": 1.7927369353410097e-05, + "loss": 1.1082, + "step": 2025 + }, + { + "epoch": 0.08970257795291278, + "grad_norm": 1.5170557498931885, + "learning_rate": 1.7936226749335697e-05, + "loss": 1.0959, + "step": 2026 + }, + { + "epoch": 0.08974685365772667, + "grad_norm": 1.4811673164367676, + "learning_rate": 1.7945084145261297e-05, + "loss": 1.0161, + "step": 2027 + }, + { + "epoch": 0.08979112936254054, + "grad_norm": 1.5085155963897705, + "learning_rate": 1.7953941541186894e-05, + "loss": 1.0361, + "step": 2028 + }, + { + "epoch": 0.08983540506735442, + "grad_norm": 1.8553839921951294, + "learning_rate": 1.796279893711249e-05, + "loss": 0.9659, + "step": 2029 + }, + { + "epoch": 0.08987968077216829, + "grad_norm": 1.3896088600158691, + "learning_rate": 1.7971656333038087e-05, + "loss": 1.0503, + "step": 2030 + }, + { + "epoch": 0.08992395647698216, + "grad_norm": 1.4982013702392578, + "learning_rate": 1.7980513728963687e-05, + "loss": 1.1052, + "step": 2031 + }, + { + "epoch": 0.08996823218179605, + "grad_norm": 1.3340582847595215, + "learning_rate": 1.7989371124889284e-05, + "loss": 0.9878, + "step": 2032 + }, + { + "epoch": 0.09001250788660992, + "grad_norm": 1.757428526878357, + "learning_rate": 1.7998228520814884e-05, + "loss": 1.0651, + "step": 2033 + }, + { + "epoch": 0.0900567835914238, + "grad_norm": 1.6204729080200195, + "learning_rate": 1.8007085916740477e-05, + "loss": 1.0779, + "step": 2034 + }, + { + "epoch": 0.09010105929623767, + "grad_norm": 1.8228870630264282, + "learning_rate": 1.8015943312666077e-05, + "loss": 1.0571, + "step": 2035 + }, + { + "epoch": 0.09014533500105155, + "grad_norm": 1.2273081541061401, + "learning_rate": 1.8024800708591677e-05, + "loss": 0.9118, + "step": 2036 + }, + { + "epoch": 0.09018961070586543, + "grad_norm": 1.4448184967041016, + "learning_rate": 1.8033658104517274e-05, + "loss": 0.9616, + "step": 2037 + }, + { + "epoch": 0.0902338864106793, + "grad_norm": 1.541359305381775, + "learning_rate": 1.804251550044287e-05, + "loss": 1.0488, + "step": 2038 + }, + { + "epoch": 0.09027816211549318, + "grad_norm": 1.6689667701721191, + "learning_rate": 1.8051372896368467e-05, + "loss": 1.1025, + "step": 2039 + }, + { + "epoch": 0.09032243782030705, + "grad_norm": 1.3218525648117065, + "learning_rate": 1.8060230292294067e-05, + "loss": 0.9508, + "step": 2040 + }, + { + "epoch": 0.09036671352512093, + "grad_norm": 1.6574819087982178, + "learning_rate": 1.8069087688219664e-05, + "loss": 0.9643, + "step": 2041 + }, + { + "epoch": 0.0904109892299348, + "grad_norm": 1.2876713275909424, + "learning_rate": 1.8077945084145264e-05, + "loss": 0.9647, + "step": 2042 + }, + { + "epoch": 0.09045526493474867, + "grad_norm": 1.6045876741409302, + "learning_rate": 1.808680248007086e-05, + "loss": 1.0359, + "step": 2043 + }, + { + "epoch": 0.09049954063956256, + "grad_norm": 1.9979995489120483, + "learning_rate": 1.8095659875996458e-05, + "loss": 1.0884, + "step": 2044 + }, + { + "epoch": 0.09054381634437643, + "grad_norm": 1.5164240598678589, + "learning_rate": 1.8104517271922058e-05, + "loss": 1.0224, + "step": 2045 + }, + { + "epoch": 0.09058809204919031, + "grad_norm": 1.4299598932266235, + "learning_rate": 1.8113374667847654e-05, + "loss": 1.0589, + "step": 2046 + }, + { + "epoch": 0.09063236775400418, + "grad_norm": 1.6823099851608276, + "learning_rate": 1.8122232063773255e-05, + "loss": 1.1373, + "step": 2047 + }, + { + "epoch": 0.09067664345881805, + "grad_norm": 1.9908900260925293, + "learning_rate": 1.813108945969885e-05, + "loss": 1.1858, + "step": 2048 + }, + { + "epoch": 0.09072091916363194, + "grad_norm": 1.4775347709655762, + "learning_rate": 1.8139946855624448e-05, + "loss": 1.0511, + "step": 2049 + }, + { + "epoch": 0.09076519486844581, + "grad_norm": 1.9358744621276855, + "learning_rate": 1.8148804251550045e-05, + "loss": 1.065, + "step": 2050 + }, + { + "epoch": 0.09080947057325969, + "grad_norm": 1.826063632965088, + "learning_rate": 1.8157661647475645e-05, + "loss": 1.0284, + "step": 2051 + }, + { + "epoch": 0.09085374627807356, + "grad_norm": 1.4802122116088867, + "learning_rate": 1.816651904340124e-05, + "loss": 1.0645, + "step": 2052 + }, + { + "epoch": 0.09089802198288743, + "grad_norm": 1.788594126701355, + "learning_rate": 1.8175376439326838e-05, + "loss": 1.1921, + "step": 2053 + }, + { + "epoch": 0.09094229768770132, + "grad_norm": 1.3070958852767944, + "learning_rate": 1.8184233835252438e-05, + "loss": 0.9484, + "step": 2054 + }, + { + "epoch": 0.09098657339251519, + "grad_norm": 1.2501859664916992, + "learning_rate": 1.8193091231178035e-05, + "loss": 0.9628, + "step": 2055 + }, + { + "epoch": 0.09103084909732907, + "grad_norm": 1.523616075515747, + "learning_rate": 1.8201948627103635e-05, + "loss": 1.0381, + "step": 2056 + }, + { + "epoch": 0.09107512480214294, + "grad_norm": 1.2427067756652832, + "learning_rate": 1.821080602302923e-05, + "loss": 0.9869, + "step": 2057 + }, + { + "epoch": 0.09111940050695683, + "grad_norm": 1.4158169031143188, + "learning_rate": 1.821966341895483e-05, + "loss": 0.9922, + "step": 2058 + }, + { + "epoch": 0.0911636762117707, + "grad_norm": 1.4423305988311768, + "learning_rate": 1.8228520814880425e-05, + "loss": 1.0186, + "step": 2059 + }, + { + "epoch": 0.09120795191658457, + "grad_norm": 1.4308289289474487, + "learning_rate": 1.8237378210806025e-05, + "loss": 1.0393, + "step": 2060 + }, + { + "epoch": 0.09125222762139845, + "grad_norm": 1.384053349494934, + "learning_rate": 1.8246235606731622e-05, + "loss": 1.0186, + "step": 2061 + }, + { + "epoch": 0.09129650332621232, + "grad_norm": 1.568804144859314, + "learning_rate": 1.8255093002657222e-05, + "loss": 1.0394, + "step": 2062 + }, + { + "epoch": 0.0913407790310262, + "grad_norm": 1.4639198780059814, + "learning_rate": 1.826395039858282e-05, + "loss": 1.0725, + "step": 2063 + }, + { + "epoch": 0.09138505473584008, + "grad_norm": 1.3642997741699219, + "learning_rate": 1.8272807794508415e-05, + "loss": 1.0227, + "step": 2064 + }, + { + "epoch": 0.09142933044065395, + "grad_norm": 1.4670689105987549, + "learning_rate": 1.8281665190434015e-05, + "loss": 1.087, + "step": 2065 + }, + { + "epoch": 0.09147360614546783, + "grad_norm": 1.4960373640060425, + "learning_rate": 1.8290522586359612e-05, + "loss": 1.0156, + "step": 2066 + }, + { + "epoch": 0.0915178818502817, + "grad_norm": 1.5868024826049805, + "learning_rate": 1.829937998228521e-05, + "loss": 1.1501, + "step": 2067 + }, + { + "epoch": 0.09156215755509559, + "grad_norm": 1.5230255126953125, + "learning_rate": 1.8308237378210805e-05, + "loss": 1.0737, + "step": 2068 + }, + { + "epoch": 0.09160643325990946, + "grad_norm": 1.351894497871399, + "learning_rate": 1.8317094774136406e-05, + "loss": 0.9688, + "step": 2069 + }, + { + "epoch": 0.09165070896472333, + "grad_norm": 1.5600529909133911, + "learning_rate": 1.8325952170062002e-05, + "loss": 1.0758, + "step": 2070 + }, + { + "epoch": 0.09169498466953721, + "grad_norm": 1.5129055976867676, + "learning_rate": 1.8334809565987602e-05, + "loss": 1.0354, + "step": 2071 + }, + { + "epoch": 0.09173926037435108, + "grad_norm": 1.4646296501159668, + "learning_rate": 1.83436669619132e-05, + "loss": 0.9705, + "step": 2072 + }, + { + "epoch": 0.09178353607916497, + "grad_norm": 1.213607907295227, + "learning_rate": 1.8352524357838796e-05, + "loss": 1.0101, + "step": 2073 + }, + { + "epoch": 0.09182781178397884, + "grad_norm": 1.4091951847076416, + "learning_rate": 1.8361381753764396e-05, + "loss": 1.0441, + "step": 2074 + }, + { + "epoch": 0.0918720874887927, + "grad_norm": 1.582634449005127, + "learning_rate": 1.8370239149689993e-05, + "loss": 1.0269, + "step": 2075 + }, + { + "epoch": 0.09191636319360659, + "grad_norm": 1.4380172491073608, + "learning_rate": 1.8379096545615593e-05, + "loss": 1.1209, + "step": 2076 + }, + { + "epoch": 0.09196063889842046, + "grad_norm": 1.4800496101379395, + "learning_rate": 1.838795394154119e-05, + "loss": 1.0711, + "step": 2077 + }, + { + "epoch": 0.09200491460323434, + "grad_norm": 1.489307165145874, + "learning_rate": 1.8396811337466786e-05, + "loss": 1.0337, + "step": 2078 + }, + { + "epoch": 0.09204919030804821, + "grad_norm": 1.5591264963150024, + "learning_rate": 1.8405668733392383e-05, + "loss": 1.0172, + "step": 2079 + }, + { + "epoch": 0.0920934660128621, + "grad_norm": 1.7535951137542725, + "learning_rate": 1.8414526129317983e-05, + "loss": 1.0804, + "step": 2080 + }, + { + "epoch": 0.09213774171767597, + "grad_norm": 1.459426760673523, + "learning_rate": 1.842338352524358e-05, + "loss": 1.079, + "step": 2081 + }, + { + "epoch": 0.09218201742248984, + "grad_norm": 1.3186744451522827, + "learning_rate": 1.8432240921169176e-05, + "loss": 1.0447, + "step": 2082 + }, + { + "epoch": 0.09222629312730372, + "grad_norm": 1.7146596908569336, + "learning_rate": 1.8441098317094776e-05, + "loss": 1.1357, + "step": 2083 + }, + { + "epoch": 0.0922705688321176, + "grad_norm": 1.4182511568069458, + "learning_rate": 1.8449955713020373e-05, + "loss": 1.0099, + "step": 2084 + }, + { + "epoch": 0.09231484453693148, + "grad_norm": 1.5265717506408691, + "learning_rate": 1.8458813108945973e-05, + "loss": 1.028, + "step": 2085 + }, + { + "epoch": 0.09235912024174535, + "grad_norm": 1.509948968887329, + "learning_rate": 1.846767050487157e-05, + "loss": 1.1233, + "step": 2086 + }, + { + "epoch": 0.09240339594655922, + "grad_norm": 1.3576838970184326, + "learning_rate": 1.8476527900797166e-05, + "loss": 1.0416, + "step": 2087 + }, + { + "epoch": 0.0924476716513731, + "grad_norm": 1.4622658491134644, + "learning_rate": 1.8485385296722763e-05, + "loss": 1.0893, + "step": 2088 + }, + { + "epoch": 0.09249194735618697, + "grad_norm": 1.3663386106491089, + "learning_rate": 1.8494242692648363e-05, + "loss": 1.0072, + "step": 2089 + }, + { + "epoch": 0.09253622306100086, + "grad_norm": 1.3838859796524048, + "learning_rate": 1.850310008857396e-05, + "loss": 0.9613, + "step": 2090 + }, + { + "epoch": 0.09258049876581473, + "grad_norm": 1.6498222351074219, + "learning_rate": 1.851195748449956e-05, + "loss": 1.1815, + "step": 2091 + }, + { + "epoch": 0.0926247744706286, + "grad_norm": 1.3397661447525024, + "learning_rate": 1.8520814880425157e-05, + "loss": 1.0076, + "step": 2092 + }, + { + "epoch": 0.09266905017544248, + "grad_norm": 1.4521360397338867, + "learning_rate": 1.8529672276350753e-05, + "loss": 1.0664, + "step": 2093 + }, + { + "epoch": 0.09271332588025635, + "grad_norm": 1.567938208580017, + "learning_rate": 1.8538529672276353e-05, + "loss": 1.0273, + "step": 2094 + }, + { + "epoch": 0.09275760158507024, + "grad_norm": 1.3229584693908691, + "learning_rate": 1.854738706820195e-05, + "loss": 1.0179, + "step": 2095 + }, + { + "epoch": 0.09280187728988411, + "grad_norm": 1.4814201593399048, + "learning_rate": 1.8556244464127547e-05, + "loss": 0.9977, + "step": 2096 + }, + { + "epoch": 0.09284615299469798, + "grad_norm": 1.1893409490585327, + "learning_rate": 1.8565101860053144e-05, + "loss": 0.9402, + "step": 2097 + }, + { + "epoch": 0.09289042869951186, + "grad_norm": 1.8239481449127197, + "learning_rate": 1.8573959255978744e-05, + "loss": 1.1266, + "step": 2098 + }, + { + "epoch": 0.09293470440432573, + "grad_norm": 1.4471960067749023, + "learning_rate": 1.858281665190434e-05, + "loss": 1.0492, + "step": 2099 + }, + { + "epoch": 0.09297898010913962, + "grad_norm": 1.4769492149353027, + "learning_rate": 1.859167404782994e-05, + "loss": 0.9837, + "step": 2100 + }, + { + "epoch": 0.09302325581395349, + "grad_norm": 1.3406449556350708, + "learning_rate": 1.8600531443755537e-05, + "loss": 1.0907, + "step": 2101 + }, + { + "epoch": 0.09306753151876737, + "grad_norm": 2.8746044635772705, + "learning_rate": 1.8609388839681134e-05, + "loss": 1.1107, + "step": 2102 + }, + { + "epoch": 0.09311180722358124, + "grad_norm": 1.4646848440170288, + "learning_rate": 1.8618246235606734e-05, + "loss": 1.0428, + "step": 2103 + }, + { + "epoch": 0.09315608292839511, + "grad_norm": 1.3142257928848267, + "learning_rate": 1.862710363153233e-05, + "loss": 0.9908, + "step": 2104 + }, + { + "epoch": 0.093200358633209, + "grad_norm": 1.5123794078826904, + "learning_rate": 1.863596102745793e-05, + "loss": 1.0245, + "step": 2105 + }, + { + "epoch": 0.09324463433802287, + "grad_norm": 1.4743882417678833, + "learning_rate": 1.8644818423383527e-05, + "loss": 1.0439, + "step": 2106 + }, + { + "epoch": 0.09328891004283675, + "grad_norm": 1.7135295867919922, + "learning_rate": 1.8653675819309124e-05, + "loss": 1.1174, + "step": 2107 + }, + { + "epoch": 0.09333318574765062, + "grad_norm": 1.6356568336486816, + "learning_rate": 1.866253321523472e-05, + "loss": 1.1484, + "step": 2108 + }, + { + "epoch": 0.09337746145246449, + "grad_norm": 1.4617197513580322, + "learning_rate": 1.867139061116032e-05, + "loss": 1.072, + "step": 2109 + }, + { + "epoch": 0.09342173715727838, + "grad_norm": 1.3200184106826782, + "learning_rate": 1.868024800708592e-05, + "loss": 0.947, + "step": 2110 + }, + { + "epoch": 0.09346601286209225, + "grad_norm": 1.5663905143737793, + "learning_rate": 1.8689105403011518e-05, + "loss": 1.0473, + "step": 2111 + }, + { + "epoch": 0.09351028856690613, + "grad_norm": 1.4201343059539795, + "learning_rate": 1.8697962798937114e-05, + "loss": 0.9936, + "step": 2112 + }, + { + "epoch": 0.09355456427172, + "grad_norm": 1.6461784839630127, + "learning_rate": 1.870682019486271e-05, + "loss": 1.0983, + "step": 2113 + }, + { + "epoch": 0.09359883997653387, + "grad_norm": 1.587364912033081, + "learning_rate": 1.871567759078831e-05, + "loss": 0.9418, + "step": 2114 + }, + { + "epoch": 0.09364311568134775, + "grad_norm": 1.5913439989089966, + "learning_rate": 1.8724534986713908e-05, + "loss": 1.1166, + "step": 2115 + }, + { + "epoch": 0.09368739138616163, + "grad_norm": 1.4353704452514648, + "learning_rate": 1.8733392382639505e-05, + "loss": 0.9577, + "step": 2116 + }, + { + "epoch": 0.09373166709097551, + "grad_norm": 1.6461434364318848, + "learning_rate": 1.87422497785651e-05, + "loss": 1.1748, + "step": 2117 + }, + { + "epoch": 0.09377594279578938, + "grad_norm": 1.6014938354492188, + "learning_rate": 1.87511071744907e-05, + "loss": 1.1098, + "step": 2118 + }, + { + "epoch": 0.09382021850060325, + "grad_norm": 1.434242844581604, + "learning_rate": 1.8759964570416298e-05, + "loss": 1.0561, + "step": 2119 + }, + { + "epoch": 0.09386449420541713, + "grad_norm": 1.4802587032318115, + "learning_rate": 1.8768821966341898e-05, + "loss": 0.9956, + "step": 2120 + }, + { + "epoch": 0.093908769910231, + "grad_norm": 1.702035665512085, + "learning_rate": 1.8777679362267495e-05, + "loss": 1.1225, + "step": 2121 + }, + { + "epoch": 0.09395304561504489, + "grad_norm": 1.4760189056396484, + "learning_rate": 1.878653675819309e-05, + "loss": 0.9916, + "step": 2122 + }, + { + "epoch": 0.09399732131985876, + "grad_norm": 1.6921814680099487, + "learning_rate": 1.879539415411869e-05, + "loss": 1.0876, + "step": 2123 + }, + { + "epoch": 0.09404159702467263, + "grad_norm": 1.6734086275100708, + "learning_rate": 1.8804251550044288e-05, + "loss": 1.1416, + "step": 2124 + }, + { + "epoch": 0.09408587272948651, + "grad_norm": 1.3002443313598633, + "learning_rate": 1.881310894596989e-05, + "loss": 0.982, + "step": 2125 + }, + { + "epoch": 0.09413014843430038, + "grad_norm": 1.5191007852554321, + "learning_rate": 1.8821966341895485e-05, + "loss": 0.9725, + "step": 2126 + }, + { + "epoch": 0.09417442413911427, + "grad_norm": 1.5508909225463867, + "learning_rate": 1.8830823737821082e-05, + "loss": 0.9981, + "step": 2127 + }, + { + "epoch": 0.09421869984392814, + "grad_norm": 1.5914286375045776, + "learning_rate": 1.883968113374668e-05, + "loss": 1.0609, + "step": 2128 + }, + { + "epoch": 0.09426297554874202, + "grad_norm": 2.029332399368286, + "learning_rate": 1.884853852967228e-05, + "loss": 1.142, + "step": 2129 + }, + { + "epoch": 0.0943072512535559, + "grad_norm": 1.534435749053955, + "learning_rate": 1.8857395925597875e-05, + "loss": 1.0361, + "step": 2130 + }, + { + "epoch": 0.09435152695836976, + "grad_norm": 6.065176486968994, + "learning_rate": 1.8866253321523472e-05, + "loss": 1.0245, + "step": 2131 + }, + { + "epoch": 0.09439580266318365, + "grad_norm": 1.6986558437347412, + "learning_rate": 1.8875110717449072e-05, + "loss": 1.1309, + "step": 2132 + }, + { + "epoch": 0.09444007836799752, + "grad_norm": 1.3129065036773682, + "learning_rate": 1.888396811337467e-05, + "loss": 0.9603, + "step": 2133 + }, + { + "epoch": 0.0944843540728114, + "grad_norm": 1.430304765701294, + "learning_rate": 1.889282550930027e-05, + "loss": 1.0814, + "step": 2134 + }, + { + "epoch": 0.09452862977762527, + "grad_norm": 1.4493257999420166, + "learning_rate": 1.8901682905225866e-05, + "loss": 1.0748, + "step": 2135 + }, + { + "epoch": 0.09457290548243914, + "grad_norm": 1.367193579673767, + "learning_rate": 1.8910540301151462e-05, + "loss": 0.9719, + "step": 2136 + }, + { + "epoch": 0.09461718118725303, + "grad_norm": 1.693396806716919, + "learning_rate": 1.891939769707706e-05, + "loss": 1.0575, + "step": 2137 + }, + { + "epoch": 0.0946614568920669, + "grad_norm": 1.5561847686767578, + "learning_rate": 1.892825509300266e-05, + "loss": 1.1391, + "step": 2138 + }, + { + "epoch": 0.09470573259688078, + "grad_norm": 1.531785488128662, + "learning_rate": 1.893711248892826e-05, + "loss": 1.0166, + "step": 2139 + }, + { + "epoch": 0.09475000830169465, + "grad_norm": 1.4653314352035522, + "learning_rate": 1.8945969884853856e-05, + "loss": 1.0247, + "step": 2140 + }, + { + "epoch": 0.09479428400650852, + "grad_norm": 1.282545566558838, + "learning_rate": 1.8954827280779452e-05, + "loss": 1.0029, + "step": 2141 + }, + { + "epoch": 0.0948385597113224, + "grad_norm": 1.9252198934555054, + "learning_rate": 1.896368467670505e-05, + "loss": 1.164, + "step": 2142 + }, + { + "epoch": 0.09488283541613628, + "grad_norm": 1.358757734298706, + "learning_rate": 1.897254207263065e-05, + "loss": 1.0398, + "step": 2143 + }, + { + "epoch": 0.09492711112095016, + "grad_norm": 1.641134262084961, + "learning_rate": 1.8981399468556246e-05, + "loss": 1.0402, + "step": 2144 + }, + { + "epoch": 0.09497138682576403, + "grad_norm": 1.6389106512069702, + "learning_rate": 1.8990256864481843e-05, + "loss": 1.0562, + "step": 2145 + }, + { + "epoch": 0.0950156625305779, + "grad_norm": 1.9159960746765137, + "learning_rate": 1.899911426040744e-05, + "loss": 1.1032, + "step": 2146 + }, + { + "epoch": 0.09505993823539179, + "grad_norm": 1.4810563325881958, + "learning_rate": 1.900797165633304e-05, + "loss": 1.0527, + "step": 2147 + }, + { + "epoch": 0.09510421394020566, + "grad_norm": 1.547391653060913, + "learning_rate": 1.901682905225864e-05, + "loss": 1.0391, + "step": 2148 + }, + { + "epoch": 0.09514848964501954, + "grad_norm": 1.560401439666748, + "learning_rate": 1.9025686448184236e-05, + "loss": 0.9828, + "step": 2149 + }, + { + "epoch": 0.09519276534983341, + "grad_norm": 1.5413395166397095, + "learning_rate": 1.9034543844109833e-05, + "loss": 0.9966, + "step": 2150 + }, + { + "epoch": 0.0952370410546473, + "grad_norm": 1.1648714542388916, + "learning_rate": 1.904340124003543e-05, + "loss": 0.9401, + "step": 2151 + }, + { + "epoch": 0.09528131675946117, + "grad_norm": 1.6696172952651978, + "learning_rate": 1.905225863596103e-05, + "loss": 1.1537, + "step": 2152 + }, + { + "epoch": 0.09532559246427504, + "grad_norm": 1.5441813468933105, + "learning_rate": 1.9061116031886626e-05, + "loss": 1.0106, + "step": 2153 + }, + { + "epoch": 0.09536986816908892, + "grad_norm": 1.5866799354553223, + "learning_rate": 1.9069973427812226e-05, + "loss": 1.0719, + "step": 2154 + }, + { + "epoch": 0.09541414387390279, + "grad_norm": 1.402974009513855, + "learning_rate": 1.9078830823737823e-05, + "loss": 1.1132, + "step": 2155 + }, + { + "epoch": 0.09545841957871667, + "grad_norm": 1.451586365699768, + "learning_rate": 1.908768821966342e-05, + "loss": 1.0706, + "step": 2156 + }, + { + "epoch": 0.09550269528353054, + "grad_norm": 1.4217041730880737, + "learning_rate": 1.909654561558902e-05, + "loss": 1.0171, + "step": 2157 + }, + { + "epoch": 0.09554697098834442, + "grad_norm": 1.4350647926330566, + "learning_rate": 1.9105403011514617e-05, + "loss": 1.0256, + "step": 2158 + }, + { + "epoch": 0.0955912466931583, + "grad_norm": 1.640567660331726, + "learning_rate": 1.9114260407440213e-05, + "loss": 1.0576, + "step": 2159 + }, + { + "epoch": 0.09563552239797217, + "grad_norm": 1.7338967323303223, + "learning_rate": 1.912311780336581e-05, + "loss": 1.0128, + "step": 2160 + }, + { + "epoch": 0.09567979810278605, + "grad_norm": 1.2606996297836304, + "learning_rate": 1.913197519929141e-05, + "loss": 1.0447, + "step": 2161 + }, + { + "epoch": 0.09572407380759992, + "grad_norm": 1.542346715927124, + "learning_rate": 1.9140832595217007e-05, + "loss": 1.085, + "step": 2162 + }, + { + "epoch": 0.0957683495124138, + "grad_norm": 1.6447123289108276, + "learning_rate": 1.9149689991142607e-05, + "loss": 0.9686, + "step": 2163 + }, + { + "epoch": 0.09581262521722768, + "grad_norm": 1.3640934228897095, + "learning_rate": 1.9158547387068204e-05, + "loss": 1.0287, + "step": 2164 + }, + { + "epoch": 0.09585690092204155, + "grad_norm": 1.8631709814071655, + "learning_rate": 1.91674047829938e-05, + "loss": 1.1253, + "step": 2165 + }, + { + "epoch": 0.09590117662685543, + "grad_norm": 1.324567437171936, + "learning_rate": 1.91762621789194e-05, + "loss": 0.9923, + "step": 2166 + }, + { + "epoch": 0.0959454523316693, + "grad_norm": 1.5290313959121704, + "learning_rate": 1.9185119574844997e-05, + "loss": 1.0471, + "step": 2167 + }, + { + "epoch": 0.09598972803648317, + "grad_norm": 1.4305838346481323, + "learning_rate": 1.9193976970770597e-05, + "loss": 0.9677, + "step": 2168 + }, + { + "epoch": 0.09603400374129706, + "grad_norm": 1.488409399986267, + "learning_rate": 1.9202834366696194e-05, + "loss": 1.0197, + "step": 2169 + }, + { + "epoch": 0.09607827944611093, + "grad_norm": 1.780604600906372, + "learning_rate": 1.921169176262179e-05, + "loss": 1.1223, + "step": 2170 + }, + { + "epoch": 0.09612255515092481, + "grad_norm": 1.452416181564331, + "learning_rate": 1.9220549158547387e-05, + "loss": 1.0165, + "step": 2171 + }, + { + "epoch": 0.09616683085573868, + "grad_norm": 1.1904703378677368, + "learning_rate": 1.9229406554472987e-05, + "loss": 0.9322, + "step": 2172 + }, + { + "epoch": 0.09621110656055257, + "grad_norm": 1.4710559844970703, + "learning_rate": 1.9238263950398584e-05, + "loss": 1.0679, + "step": 2173 + }, + { + "epoch": 0.09625538226536644, + "grad_norm": 1.6133294105529785, + "learning_rate": 1.924712134632418e-05, + "loss": 1.081, + "step": 2174 + }, + { + "epoch": 0.09629965797018031, + "grad_norm": 1.4909318685531616, + "learning_rate": 1.9255978742249777e-05, + "loss": 1.0729, + "step": 2175 + }, + { + "epoch": 0.09634393367499419, + "grad_norm": 1.536285400390625, + "learning_rate": 1.9264836138175378e-05, + "loss": 1.0358, + "step": 2176 + }, + { + "epoch": 0.09638820937980806, + "grad_norm": 2.4187183380126953, + "learning_rate": 1.9273693534100978e-05, + "loss": 1.2017, + "step": 2177 + }, + { + "epoch": 0.09643248508462195, + "grad_norm": 1.553173542022705, + "learning_rate": 1.9282550930026574e-05, + "loss": 1.014, + "step": 2178 + }, + { + "epoch": 0.09647676078943582, + "grad_norm": 1.3954498767852783, + "learning_rate": 1.929140832595217e-05, + "loss": 1.0428, + "step": 2179 + }, + { + "epoch": 0.09652103649424969, + "grad_norm": 1.498949646949768, + "learning_rate": 1.9300265721877768e-05, + "loss": 1.022, + "step": 2180 + }, + { + "epoch": 0.09656531219906357, + "grad_norm": 1.4107433557510376, + "learning_rate": 1.9309123117803368e-05, + "loss": 0.9586, + "step": 2181 + }, + { + "epoch": 0.09660958790387744, + "grad_norm": 1.2495551109313965, + "learning_rate": 1.9317980513728964e-05, + "loss": 0.9828, + "step": 2182 + }, + { + "epoch": 0.09665386360869133, + "grad_norm": 1.342616081237793, + "learning_rate": 1.9326837909654565e-05, + "loss": 0.9958, + "step": 2183 + }, + { + "epoch": 0.0966981393135052, + "grad_norm": 1.500708818435669, + "learning_rate": 1.933569530558016e-05, + "loss": 1.0685, + "step": 2184 + }, + { + "epoch": 0.09674241501831907, + "grad_norm": 1.4423940181732178, + "learning_rate": 1.9344552701505758e-05, + "loss": 1.0814, + "step": 2185 + }, + { + "epoch": 0.09678669072313295, + "grad_norm": 1.6186720132827759, + "learning_rate": 1.9353410097431358e-05, + "loss": 1.1646, + "step": 2186 + }, + { + "epoch": 0.09683096642794682, + "grad_norm": 1.3562747240066528, + "learning_rate": 1.9362267493356955e-05, + "loss": 1.0265, + "step": 2187 + }, + { + "epoch": 0.0968752421327607, + "grad_norm": 1.5098917484283447, + "learning_rate": 1.9371124889282555e-05, + "loss": 0.9964, + "step": 2188 + }, + { + "epoch": 0.09691951783757458, + "grad_norm": 1.5044687986373901, + "learning_rate": 1.937998228520815e-05, + "loss": 1.0522, + "step": 2189 + }, + { + "epoch": 0.09696379354238845, + "grad_norm": 1.7266218662261963, + "learning_rate": 1.9388839681133748e-05, + "loss": 1.0251, + "step": 2190 + }, + { + "epoch": 0.09700806924720233, + "grad_norm": 1.500973105430603, + "learning_rate": 1.9397697077059345e-05, + "loss": 1.0044, + "step": 2191 + }, + { + "epoch": 0.0970523449520162, + "grad_norm": 1.256881833076477, + "learning_rate": 1.9406554472984945e-05, + "loss": 0.92, + "step": 2192 + }, + { + "epoch": 0.09709662065683008, + "grad_norm": 1.6155226230621338, + "learning_rate": 1.9415411868910542e-05, + "loss": 0.9887, + "step": 2193 + }, + { + "epoch": 0.09714089636164396, + "grad_norm": 1.3288555145263672, + "learning_rate": 1.942426926483614e-05, + "loss": 0.9556, + "step": 2194 + }, + { + "epoch": 0.09718517206645784, + "grad_norm": 1.3577001094818115, + "learning_rate": 1.943312666076174e-05, + "loss": 1.0673, + "step": 2195 + }, + { + "epoch": 0.09722944777127171, + "grad_norm": 1.3702356815338135, + "learning_rate": 1.9441984056687335e-05, + "loss": 1.0468, + "step": 2196 + }, + { + "epoch": 0.09727372347608558, + "grad_norm": 1.618447184562683, + "learning_rate": 1.9450841452612935e-05, + "loss": 0.9848, + "step": 2197 + }, + { + "epoch": 0.09731799918089946, + "grad_norm": 1.6659398078918457, + "learning_rate": 1.9459698848538532e-05, + "loss": 1.1014, + "step": 2198 + }, + { + "epoch": 0.09736227488571333, + "grad_norm": 1.3806159496307373, + "learning_rate": 1.946855624446413e-05, + "loss": 0.968, + "step": 2199 + }, + { + "epoch": 0.09740655059052722, + "grad_norm": 1.5112415552139282, + "learning_rate": 1.9477413640389725e-05, + "loss": 1.2093, + "step": 2200 + }, + { + "epoch": 0.09745082629534109, + "grad_norm": 1.382082462310791, + "learning_rate": 1.9486271036315325e-05, + "loss": 1.025, + "step": 2201 + }, + { + "epoch": 0.09749510200015496, + "grad_norm": 1.5126522779464722, + "learning_rate": 1.9495128432240922e-05, + "loss": 1.089, + "step": 2202 + }, + { + "epoch": 0.09753937770496884, + "grad_norm": 1.4078032970428467, + "learning_rate": 1.9503985828166522e-05, + "loss": 0.9927, + "step": 2203 + }, + { + "epoch": 0.09758365340978271, + "grad_norm": 1.4523098468780518, + "learning_rate": 1.951284322409212e-05, + "loss": 1.0034, + "step": 2204 + }, + { + "epoch": 0.0976279291145966, + "grad_norm": 1.3839151859283447, + "learning_rate": 1.9521700620017716e-05, + "loss": 1.048, + "step": 2205 + }, + { + "epoch": 0.09767220481941047, + "grad_norm": 1.4571113586425781, + "learning_rate": 1.9530558015943316e-05, + "loss": 0.9471, + "step": 2206 + }, + { + "epoch": 0.09771648052422434, + "grad_norm": 1.4180059432983398, + "learning_rate": 1.9539415411868912e-05, + "loss": 1.0659, + "step": 2207 + }, + { + "epoch": 0.09776075622903822, + "grad_norm": 1.5977367162704468, + "learning_rate": 1.954827280779451e-05, + "loss": 1.1104, + "step": 2208 + }, + { + "epoch": 0.0978050319338521, + "grad_norm": 1.6314010620117188, + "learning_rate": 1.9557130203720106e-05, + "loss": 1.0988, + "step": 2209 + }, + { + "epoch": 0.09784930763866598, + "grad_norm": 1.9311624765396118, + "learning_rate": 1.9565987599645706e-05, + "loss": 1.0915, + "step": 2210 + }, + { + "epoch": 0.09789358334347985, + "grad_norm": 1.4442944526672363, + "learning_rate": 1.9574844995571303e-05, + "loss": 1.0361, + "step": 2211 + }, + { + "epoch": 0.09793785904829372, + "grad_norm": 1.3358856439590454, + "learning_rate": 1.9583702391496903e-05, + "loss": 1.0083, + "step": 2212 + }, + { + "epoch": 0.0979821347531076, + "grad_norm": 1.7417947053909302, + "learning_rate": 1.95925597874225e-05, + "loss": 1.0784, + "step": 2213 + }, + { + "epoch": 0.09802641045792147, + "grad_norm": 1.387622594833374, + "learning_rate": 1.9601417183348096e-05, + "loss": 0.971, + "step": 2214 + }, + { + "epoch": 0.09807068616273536, + "grad_norm": 1.4278717041015625, + "learning_rate": 1.9610274579273696e-05, + "loss": 1.0997, + "step": 2215 + }, + { + "epoch": 0.09811496186754923, + "grad_norm": 1.357359528541565, + "learning_rate": 1.9619131975199293e-05, + "loss": 0.9936, + "step": 2216 + }, + { + "epoch": 0.09815923757236311, + "grad_norm": 1.3371820449829102, + "learning_rate": 1.9627989371124893e-05, + "loss": 1.0493, + "step": 2217 + }, + { + "epoch": 0.09820351327717698, + "grad_norm": 1.5040678977966309, + "learning_rate": 1.963684676705049e-05, + "loss": 1.0378, + "step": 2218 + }, + { + "epoch": 0.09824778898199085, + "grad_norm": 1.386949062347412, + "learning_rate": 1.9645704162976086e-05, + "loss": 1.042, + "step": 2219 + }, + { + "epoch": 0.09829206468680474, + "grad_norm": 1.5234134197235107, + "learning_rate": 1.9654561558901683e-05, + "loss": 1.1294, + "step": 2220 + }, + { + "epoch": 0.0983363403916186, + "grad_norm": 1.2561615705490112, + "learning_rate": 1.9663418954827283e-05, + "loss": 0.9439, + "step": 2221 + }, + { + "epoch": 0.09838061609643249, + "grad_norm": 1.3867509365081787, + "learning_rate": 1.967227635075288e-05, + "loss": 1.0335, + "step": 2222 + }, + { + "epoch": 0.09842489180124636, + "grad_norm": 1.4298372268676758, + "learning_rate": 1.9681133746678476e-05, + "loss": 1.0626, + "step": 2223 + }, + { + "epoch": 0.09846916750606023, + "grad_norm": 1.563646912574768, + "learning_rate": 1.9689991142604077e-05, + "loss": 1.1561, + "step": 2224 + }, + { + "epoch": 0.09851344321087412, + "grad_norm": 1.4655506610870361, + "learning_rate": 1.9698848538529673e-05, + "loss": 1.0706, + "step": 2225 + }, + { + "epoch": 0.09855771891568799, + "grad_norm": 1.5331860780715942, + "learning_rate": 1.9707705934455273e-05, + "loss": 1.091, + "step": 2226 + }, + { + "epoch": 0.09860199462050187, + "grad_norm": 1.7374212741851807, + "learning_rate": 1.971656333038087e-05, + "loss": 1.0809, + "step": 2227 + }, + { + "epoch": 0.09864627032531574, + "grad_norm": 1.6177817583084106, + "learning_rate": 1.9725420726306467e-05, + "loss": 1.1187, + "step": 2228 + }, + { + "epoch": 0.09869054603012961, + "grad_norm": 1.5448552370071411, + "learning_rate": 1.9734278122232063e-05, + "loss": 1.1499, + "step": 2229 + }, + { + "epoch": 0.0987348217349435, + "grad_norm": 1.3742886781692505, + "learning_rate": 1.9743135518157664e-05, + "loss": 1.0194, + "step": 2230 + }, + { + "epoch": 0.09877909743975737, + "grad_norm": 1.5019456148147583, + "learning_rate": 1.975199291408326e-05, + "loss": 1.0177, + "step": 2231 + }, + { + "epoch": 0.09882337314457125, + "grad_norm": 1.8392937183380127, + "learning_rate": 1.976085031000886e-05, + "loss": 1.1311, + "step": 2232 + }, + { + "epoch": 0.09886764884938512, + "grad_norm": 2.0540080070495605, + "learning_rate": 1.9769707705934457e-05, + "loss": 1.1089, + "step": 2233 + }, + { + "epoch": 0.09891192455419899, + "grad_norm": 1.402978777885437, + "learning_rate": 1.9778565101860054e-05, + "loss": 0.9925, + "step": 2234 + }, + { + "epoch": 0.09895620025901287, + "grad_norm": 1.832825779914856, + "learning_rate": 1.9787422497785654e-05, + "loss": 1.1085, + "step": 2235 + }, + { + "epoch": 0.09900047596382675, + "grad_norm": 1.334046483039856, + "learning_rate": 1.979627989371125e-05, + "loss": 0.9723, + "step": 2236 + }, + { + "epoch": 0.09904475166864063, + "grad_norm": 1.379292368888855, + "learning_rate": 1.9805137289636847e-05, + "loss": 0.9848, + "step": 2237 + }, + { + "epoch": 0.0990890273734545, + "grad_norm": 1.4867229461669922, + "learning_rate": 1.9813994685562444e-05, + "loss": 1.0909, + "step": 2238 + }, + { + "epoch": 0.09913330307826838, + "grad_norm": 1.3280084133148193, + "learning_rate": 1.9822852081488044e-05, + "loss": 1.0067, + "step": 2239 + }, + { + "epoch": 0.09917757878308225, + "grad_norm": 1.3559412956237793, + "learning_rate": 1.983170947741364e-05, + "loss": 0.9696, + "step": 2240 + }, + { + "epoch": 0.09922185448789612, + "grad_norm": 1.2786163091659546, + "learning_rate": 1.984056687333924e-05, + "loss": 0.9317, + "step": 2241 + }, + { + "epoch": 0.09926613019271001, + "grad_norm": 1.5347423553466797, + "learning_rate": 1.9849424269264837e-05, + "loss": 1.0884, + "step": 2242 + }, + { + "epoch": 0.09931040589752388, + "grad_norm": 1.466408371925354, + "learning_rate": 1.9858281665190434e-05, + "loss": 1.034, + "step": 2243 + }, + { + "epoch": 0.09935468160233776, + "grad_norm": 2.0072829723358154, + "learning_rate": 1.9867139061116034e-05, + "loss": 1.1774, + "step": 2244 + }, + { + "epoch": 0.09939895730715163, + "grad_norm": 1.4654394388198853, + "learning_rate": 1.987599645704163e-05, + "loss": 1.0678, + "step": 2245 + }, + { + "epoch": 0.0994432330119655, + "grad_norm": 1.6371010541915894, + "learning_rate": 1.988485385296723e-05, + "loss": 1.0683, + "step": 2246 + }, + { + "epoch": 0.09948750871677939, + "grad_norm": 1.4670510292053223, + "learning_rate": 1.9893711248892828e-05, + "loss": 1.0917, + "step": 2247 + }, + { + "epoch": 0.09953178442159326, + "grad_norm": 1.392958164215088, + "learning_rate": 1.9902568644818424e-05, + "loss": 1.0312, + "step": 2248 + }, + { + "epoch": 0.09957606012640714, + "grad_norm": 1.4812285900115967, + "learning_rate": 1.991142604074402e-05, + "loss": 1.0446, + "step": 2249 + }, + { + "epoch": 0.09962033583122101, + "grad_norm": 1.4254117012023926, + "learning_rate": 1.992028343666962e-05, + "loss": 1.0196, + "step": 2250 + }, + { + "epoch": 0.09966461153603488, + "grad_norm": 1.3435696363449097, + "learning_rate": 1.9929140832595218e-05, + "loss": 1.0266, + "step": 2251 + }, + { + "epoch": 0.09970888724084877, + "grad_norm": 1.8269069194793701, + "learning_rate": 1.9937998228520815e-05, + "loss": 1.0543, + "step": 2252 + }, + { + "epoch": 0.09975316294566264, + "grad_norm": 1.3412961959838867, + "learning_rate": 1.9946855624446415e-05, + "loss": 1.0023, + "step": 2253 + }, + { + "epoch": 0.09979743865047652, + "grad_norm": 2.0649795532226562, + "learning_rate": 1.995571302037201e-05, + "loss": 1.1391, + "step": 2254 + }, + { + "epoch": 0.09984171435529039, + "grad_norm": 1.6650198698043823, + "learning_rate": 1.996457041629761e-05, + "loss": 0.9801, + "step": 2255 + }, + { + "epoch": 0.09988599006010426, + "grad_norm": 1.4824658632278442, + "learning_rate": 1.9973427812223208e-05, + "loss": 1.0182, + "step": 2256 + }, + { + "epoch": 0.09993026576491815, + "grad_norm": 1.3869993686676025, + "learning_rate": 1.9982285208148805e-05, + "loss": 1.0787, + "step": 2257 + }, + { + "epoch": 0.09997454146973202, + "grad_norm": 1.7716963291168213, + "learning_rate": 1.99911426040744e-05, + "loss": 1.0083, + "step": 2258 + }, + { + "epoch": 0.1000188171745459, + "grad_norm": 1.4980463981628418, + "learning_rate": 2e-05, + "loss": 1.0977, + "step": 2259 + }, + { + "epoch": 0.10006309287935977, + "grad_norm": 1.6067125797271729, + "learning_rate": 1.9999999880579073e-05, + "loss": 1.1657, + "step": 2260 + }, + { + "epoch": 0.10010736858417366, + "grad_norm": 1.4421621561050415, + "learning_rate": 1.999999952231629e-05, + "loss": 1.0989, + "step": 2261 + }, + { + "epoch": 0.10015164428898753, + "grad_norm": 1.3123998641967773, + "learning_rate": 1.999999892521166e-05, + "loss": 0.9719, + "step": 2262 + }, + { + "epoch": 0.1001959199938014, + "grad_norm": 1.4839640855789185, + "learning_rate": 1.9999998089265197e-05, + "loss": 1.0499, + "step": 2263 + }, + { + "epoch": 0.10024019569861528, + "grad_norm": 1.3681026697158813, + "learning_rate": 1.9999997014476925e-05, + "loss": 1.0738, + "step": 2264 + }, + { + "epoch": 0.10028447140342915, + "grad_norm": 1.2561936378479004, + "learning_rate": 1.9999995700846863e-05, + "loss": 0.9502, + "step": 2265 + }, + { + "epoch": 0.10032874710824304, + "grad_norm": 1.4355113506317139, + "learning_rate": 1.9999994148375048e-05, + "loss": 1.0088, + "step": 2266 + }, + { + "epoch": 0.1003730228130569, + "grad_norm": 2.2005598545074463, + "learning_rate": 1.9999992357061514e-05, + "loss": 1.1566, + "step": 2267 + }, + { + "epoch": 0.10041729851787078, + "grad_norm": 1.412073016166687, + "learning_rate": 1.9999990326906306e-05, + "loss": 1.0887, + "step": 2268 + }, + { + "epoch": 0.10046157422268466, + "grad_norm": 1.5042650699615479, + "learning_rate": 1.999998805790947e-05, + "loss": 1.0698, + "step": 2269 + }, + { + "epoch": 0.10050584992749853, + "grad_norm": 1.4425760507583618, + "learning_rate": 1.9999985550071064e-05, + "loss": 0.9922, + "step": 2270 + }, + { + "epoch": 0.10055012563231241, + "grad_norm": 1.251715898513794, + "learning_rate": 1.999998280339114e-05, + "loss": 0.9304, + "step": 2271 + }, + { + "epoch": 0.10059440133712629, + "grad_norm": 1.3842155933380127, + "learning_rate": 1.9999979817869773e-05, + "loss": 1.0933, + "step": 2272 + }, + { + "epoch": 0.10063867704194016, + "grad_norm": 1.35905921459198, + "learning_rate": 1.999997659350703e-05, + "loss": 0.9863, + "step": 2273 + }, + { + "epoch": 0.10068295274675404, + "grad_norm": 1.1857446432113647, + "learning_rate": 1.999997313030299e-05, + "loss": 0.9222, + "step": 2274 + }, + { + "epoch": 0.10072722845156791, + "grad_norm": 1.5835182666778564, + "learning_rate": 1.9999969428257734e-05, + "loss": 0.9949, + "step": 2275 + }, + { + "epoch": 0.1007715041563818, + "grad_norm": 1.490880012512207, + "learning_rate": 1.999996548737135e-05, + "loss": 1.123, + "step": 2276 + }, + { + "epoch": 0.10081577986119566, + "grad_norm": 1.6733816862106323, + "learning_rate": 1.999996130764393e-05, + "loss": 1.0128, + "step": 2277 + }, + { + "epoch": 0.10086005556600953, + "grad_norm": 1.502406358718872, + "learning_rate": 1.9999956889075575e-05, + "loss": 1.1165, + "step": 2278 + }, + { + "epoch": 0.10090433127082342, + "grad_norm": 1.527185082435608, + "learning_rate": 1.9999952231666398e-05, + "loss": 1.0711, + "step": 2279 + }, + { + "epoch": 0.10094860697563729, + "grad_norm": 1.3896100521087646, + "learning_rate": 1.9999947335416502e-05, + "loss": 1.0139, + "step": 2280 + }, + { + "epoch": 0.10099288268045117, + "grad_norm": 1.6576135158538818, + "learning_rate": 1.9999942200326002e-05, + "loss": 1.0436, + "step": 2281 + }, + { + "epoch": 0.10103715838526504, + "grad_norm": 1.6342912912368774, + "learning_rate": 1.999993682639503e-05, + "loss": 1.1324, + "step": 2282 + }, + { + "epoch": 0.10108143409007893, + "grad_norm": 1.3633143901824951, + "learning_rate": 1.9999931213623708e-05, + "loss": 1.0153, + "step": 2283 + }, + { + "epoch": 0.1011257097948928, + "grad_norm": 10.164758682250977, + "learning_rate": 1.999992536201217e-05, + "loss": 1.0996, + "step": 2284 + }, + { + "epoch": 0.10116998549970667, + "grad_norm": 1.4458473920822144, + "learning_rate": 1.999991927156056e-05, + "loss": 1.0444, + "step": 2285 + }, + { + "epoch": 0.10121426120452055, + "grad_norm": 1.534485101699829, + "learning_rate": 1.9999912942269016e-05, + "loss": 1.0415, + "step": 2286 + }, + { + "epoch": 0.10125853690933442, + "grad_norm": 1.4472886323928833, + "learning_rate": 1.9999906374137693e-05, + "loss": 1.0568, + "step": 2287 + }, + { + "epoch": 0.10130281261414831, + "grad_norm": 2.745159149169922, + "learning_rate": 1.999989956716675e-05, + "loss": 1.0331, + "step": 2288 + }, + { + "epoch": 0.10134708831896218, + "grad_norm": 1.713821291923523, + "learning_rate": 1.999989252135635e-05, + "loss": 1.0127, + "step": 2289 + }, + { + "epoch": 0.10139136402377605, + "grad_norm": 1.3822191953659058, + "learning_rate": 1.999988523670666e-05, + "loss": 1.0439, + "step": 2290 + }, + { + "epoch": 0.10143563972858993, + "grad_norm": 1.50766122341156, + "learning_rate": 1.9999877713217854e-05, + "loss": 1.041, + "step": 2291 + }, + { + "epoch": 0.1014799154334038, + "grad_norm": 1.7370115518569946, + "learning_rate": 1.9999869950890106e-05, + "loss": 1.0342, + "step": 2292 + }, + { + "epoch": 0.10152419113821769, + "grad_norm": 1.339372158050537, + "learning_rate": 1.9999861949723608e-05, + "loss": 0.9581, + "step": 2293 + }, + { + "epoch": 0.10156846684303156, + "grad_norm": 2.1571669578552246, + "learning_rate": 1.9999853709718552e-05, + "loss": 1.1302, + "step": 2294 + }, + { + "epoch": 0.10161274254784543, + "grad_norm": 1.5337092876434326, + "learning_rate": 1.999984523087513e-05, + "loss": 1.0265, + "step": 2295 + }, + { + "epoch": 0.10165701825265931, + "grad_norm": 1.5276633501052856, + "learning_rate": 1.999983651319355e-05, + "loss": 1.0003, + "step": 2296 + }, + { + "epoch": 0.10170129395747318, + "grad_norm": 1.3871146440505981, + "learning_rate": 1.9999827556674015e-05, + "loss": 0.9911, + "step": 2297 + }, + { + "epoch": 0.10174556966228707, + "grad_norm": 1.4464685916900635, + "learning_rate": 1.999981836131674e-05, + "loss": 1.0452, + "step": 2298 + }, + { + "epoch": 0.10178984536710094, + "grad_norm": 1.8641072511672974, + "learning_rate": 1.999980892712195e-05, + "loss": 1.0553, + "step": 2299 + }, + { + "epoch": 0.10183412107191481, + "grad_norm": 1.9464173316955566, + "learning_rate": 1.999979925408986e-05, + "loss": 1.0027, + "step": 2300 + }, + { + "epoch": 0.10187839677672869, + "grad_norm": 1.8112542629241943, + "learning_rate": 1.999978934222071e-05, + "loss": 1.1261, + "step": 2301 + }, + { + "epoch": 0.10192267248154256, + "grad_norm": 1.571408748626709, + "learning_rate": 1.9999779191514733e-05, + "loss": 1.0299, + "step": 2302 + }, + { + "epoch": 0.10196694818635645, + "grad_norm": 1.468188762664795, + "learning_rate": 1.9999768801972174e-05, + "loss": 1.0194, + "step": 2303 + }, + { + "epoch": 0.10201122389117032, + "grad_norm": 1.2375723123550415, + "learning_rate": 1.9999758173593276e-05, + "loss": 1.0253, + "step": 2304 + }, + { + "epoch": 0.1020554995959842, + "grad_norm": 1.512837529182434, + "learning_rate": 1.9999747306378298e-05, + "loss": 1.0468, + "step": 2305 + }, + { + "epoch": 0.10209977530079807, + "grad_norm": 2.1788039207458496, + "learning_rate": 1.9999736200327496e-05, + "loss": 1.132, + "step": 2306 + }, + { + "epoch": 0.10214405100561194, + "grad_norm": 1.7152529954910278, + "learning_rate": 1.999972485544114e-05, + "loss": 1.1476, + "step": 2307 + }, + { + "epoch": 0.10218832671042583, + "grad_norm": 1.622690200805664, + "learning_rate": 1.9999713271719496e-05, + "loss": 1.0898, + "step": 2308 + }, + { + "epoch": 0.1022326024152397, + "grad_norm": 1.2846730947494507, + "learning_rate": 1.999970144916284e-05, + "loss": 1.0066, + "step": 2309 + }, + { + "epoch": 0.10227687812005358, + "grad_norm": 1.3866523504257202, + "learning_rate": 1.999968938777146e-05, + "loss": 1.0537, + "step": 2310 + }, + { + "epoch": 0.10232115382486745, + "grad_norm": 1.2901780605316162, + "learning_rate": 1.999967708754564e-05, + "loss": 1.0183, + "step": 2311 + }, + { + "epoch": 0.10236542952968132, + "grad_norm": 1.5087329149246216, + "learning_rate": 1.9999664548485674e-05, + "loss": 1.0693, + "step": 2312 + }, + { + "epoch": 0.1024097052344952, + "grad_norm": 1.4060916900634766, + "learning_rate": 1.9999651770591866e-05, + "loss": 1.0375, + "step": 2313 + }, + { + "epoch": 0.10245398093930908, + "grad_norm": 1.355509638786316, + "learning_rate": 1.9999638753864515e-05, + "loss": 1.0406, + "step": 2314 + }, + { + "epoch": 0.10249825664412296, + "grad_norm": 1.2887121438980103, + "learning_rate": 1.9999625498303936e-05, + "loss": 0.9344, + "step": 2315 + }, + { + "epoch": 0.10254253234893683, + "grad_norm": 1.4036877155303955, + "learning_rate": 1.9999612003910443e-05, + "loss": 0.9372, + "step": 2316 + }, + { + "epoch": 0.1025868080537507, + "grad_norm": 1.607978105545044, + "learning_rate": 1.9999598270684356e-05, + "loss": 1.0561, + "step": 2317 + }, + { + "epoch": 0.10263108375856458, + "grad_norm": 2.103271245956421, + "learning_rate": 1.9999584298626013e-05, + "loss": 1.1596, + "step": 2318 + }, + { + "epoch": 0.10267535946337845, + "grad_norm": 1.4573172330856323, + "learning_rate": 1.9999570087735732e-05, + "loss": 1.0353, + "step": 2319 + }, + { + "epoch": 0.10271963516819234, + "grad_norm": 1.3218698501586914, + "learning_rate": 1.999955563801387e-05, + "loss": 0.9757, + "step": 2320 + }, + { + "epoch": 0.10276391087300621, + "grad_norm": 1.2930243015289307, + "learning_rate": 1.999954094946076e-05, + "loss": 1.0667, + "step": 2321 + }, + { + "epoch": 0.10280818657782008, + "grad_norm": 1.6171410083770752, + "learning_rate": 1.9999526022076757e-05, + "loss": 1.1023, + "step": 2322 + }, + { + "epoch": 0.10285246228263396, + "grad_norm": 1.3298629522323608, + "learning_rate": 1.9999510855862217e-05, + "loss": 0.9817, + "step": 2323 + }, + { + "epoch": 0.10289673798744783, + "grad_norm": 1.5079083442687988, + "learning_rate": 1.99994954508175e-05, + "loss": 1.1051, + "step": 2324 + }, + { + "epoch": 0.10294101369226172, + "grad_norm": 1.3890464305877686, + "learning_rate": 1.9999479806942976e-05, + "loss": 1.0391, + "step": 2325 + }, + { + "epoch": 0.10298528939707559, + "grad_norm": 1.2797256708145142, + "learning_rate": 1.999946392423902e-05, + "loss": 0.9862, + "step": 2326 + }, + { + "epoch": 0.10302956510188947, + "grad_norm": 1.2320566177368164, + "learning_rate": 1.999944780270601e-05, + "loss": 0.9876, + "step": 2327 + }, + { + "epoch": 0.10307384080670334, + "grad_norm": 1.3979008197784424, + "learning_rate": 1.9999431442344334e-05, + "loss": 0.9871, + "step": 2328 + }, + { + "epoch": 0.10311811651151721, + "grad_norm": 1.5533170700073242, + "learning_rate": 1.9999414843154375e-05, + "loss": 1.0823, + "step": 2329 + }, + { + "epoch": 0.1031623922163311, + "grad_norm": 1.4514776468276978, + "learning_rate": 1.9999398005136538e-05, + "loss": 1.1114, + "step": 2330 + }, + { + "epoch": 0.10320666792114497, + "grad_norm": 1.5508846044540405, + "learning_rate": 1.999938092829122e-05, + "loss": 1.0178, + "step": 2331 + }, + { + "epoch": 0.10325094362595885, + "grad_norm": 1.7962638139724731, + "learning_rate": 1.9999363612618828e-05, + "loss": 1.1428, + "step": 2332 + }, + { + "epoch": 0.10329521933077272, + "grad_norm": 1.4129462242126465, + "learning_rate": 1.999934605811978e-05, + "loss": 0.9976, + "step": 2333 + }, + { + "epoch": 0.10333949503558659, + "grad_norm": 1.9799376726150513, + "learning_rate": 1.9999328264794493e-05, + "loss": 1.0714, + "step": 2334 + }, + { + "epoch": 0.10338377074040048, + "grad_norm": 1.4631011486053467, + "learning_rate": 1.9999310232643393e-05, + "loss": 1.0857, + "step": 2335 + }, + { + "epoch": 0.10342804644521435, + "grad_norm": 1.2872756719589233, + "learning_rate": 1.999929196166691e-05, + "loss": 0.9603, + "step": 2336 + }, + { + "epoch": 0.10347232215002823, + "grad_norm": 1.5279548168182373, + "learning_rate": 1.999927345186548e-05, + "loss": 1.0842, + "step": 2337 + }, + { + "epoch": 0.1035165978548421, + "grad_norm": 1.509270191192627, + "learning_rate": 1.9999254703239544e-05, + "loss": 1.1209, + "step": 2338 + }, + { + "epoch": 0.10356087355965597, + "grad_norm": 1.2846250534057617, + "learning_rate": 1.9999235715789555e-05, + "loss": 0.9805, + "step": 2339 + }, + { + "epoch": 0.10360514926446986, + "grad_norm": 1.5483688116073608, + "learning_rate": 1.9999216489515956e-05, + "loss": 1.0821, + "step": 2340 + }, + { + "epoch": 0.10364942496928373, + "grad_norm": 1.4151114225387573, + "learning_rate": 1.9999197024419216e-05, + "loss": 1.1115, + "step": 2341 + }, + { + "epoch": 0.10369370067409761, + "grad_norm": 1.3469767570495605, + "learning_rate": 1.99991773204998e-05, + "loss": 1.0205, + "step": 2342 + }, + { + "epoch": 0.10373797637891148, + "grad_norm": 1.242255449295044, + "learning_rate": 1.999915737775817e-05, + "loss": 0.9588, + "step": 2343 + }, + { + "epoch": 0.10378225208372535, + "grad_norm": 1.8504829406738281, + "learning_rate": 1.999913719619481e-05, + "loss": 1.0627, + "step": 2344 + }, + { + "epoch": 0.10382652778853924, + "grad_norm": 1.2314857244491577, + "learning_rate": 1.99991167758102e-05, + "loss": 0.9399, + "step": 2345 + }, + { + "epoch": 0.1038708034933531, + "grad_norm": 1.505751609802246, + "learning_rate": 1.9999096116604824e-05, + "loss": 1.1518, + "step": 2346 + }, + { + "epoch": 0.10391507919816699, + "grad_norm": 1.523784875869751, + "learning_rate": 1.9999075218579184e-05, + "loss": 1.0982, + "step": 2347 + }, + { + "epoch": 0.10395935490298086, + "grad_norm": 1.3740154504776, + "learning_rate": 1.9999054081733774e-05, + "loss": 1.0188, + "step": 2348 + }, + { + "epoch": 0.10400363060779473, + "grad_norm": 1.5576860904693604, + "learning_rate": 1.9999032706069098e-05, + "loss": 1.0279, + "step": 2349 + }, + { + "epoch": 0.10404790631260862, + "grad_norm": 1.8976765871047974, + "learning_rate": 1.9999011091585664e-05, + "loss": 1.1546, + "step": 2350 + }, + { + "epoch": 0.10409218201742249, + "grad_norm": 2.0084941387176514, + "learning_rate": 1.999898923828399e-05, + "loss": 1.0168, + "step": 2351 + }, + { + "epoch": 0.10413645772223637, + "grad_norm": 1.9769259691238403, + "learning_rate": 1.9998967146164604e-05, + "loss": 1.0117, + "step": 2352 + }, + { + "epoch": 0.10418073342705024, + "grad_norm": 1.4702633619308472, + "learning_rate": 1.999894481522803e-05, + "loss": 1.0312, + "step": 2353 + }, + { + "epoch": 0.10422500913186412, + "grad_norm": 1.5733815431594849, + "learning_rate": 1.99989222454748e-05, + "loss": 1.0451, + "step": 2354 + }, + { + "epoch": 0.104269284836678, + "grad_norm": 1.4677860736846924, + "learning_rate": 1.999889943690545e-05, + "loss": 1.031, + "step": 2355 + }, + { + "epoch": 0.10431356054149186, + "grad_norm": 1.4642672538757324, + "learning_rate": 1.9998876389520535e-05, + "loss": 1.1305, + "step": 2356 + }, + { + "epoch": 0.10435783624630575, + "grad_norm": 1.3485482931137085, + "learning_rate": 1.9998853103320592e-05, + "loss": 1.0429, + "step": 2357 + }, + { + "epoch": 0.10440211195111962, + "grad_norm": 2.0595085620880127, + "learning_rate": 1.9998829578306187e-05, + "loss": 0.9666, + "step": 2358 + }, + { + "epoch": 0.1044463876559335, + "grad_norm": 1.625392198562622, + "learning_rate": 1.999880581447788e-05, + "loss": 1.088, + "step": 2359 + }, + { + "epoch": 0.10449066336074737, + "grad_norm": 1.2997349500656128, + "learning_rate": 1.9998781811836235e-05, + "loss": 1.0152, + "step": 2360 + }, + { + "epoch": 0.10453493906556124, + "grad_norm": 1.4158066511154175, + "learning_rate": 1.9998757570381827e-05, + "loss": 1.0884, + "step": 2361 + }, + { + "epoch": 0.10457921477037513, + "grad_norm": 1.628106713294983, + "learning_rate": 1.9998733090115238e-05, + "loss": 1.0835, + "step": 2362 + }, + { + "epoch": 0.104623490475189, + "grad_norm": 1.4202401638031006, + "learning_rate": 1.999870837103705e-05, + "loss": 0.9343, + "step": 2363 + }, + { + "epoch": 0.10466776618000288, + "grad_norm": 1.7901251316070557, + "learning_rate": 1.999868341314785e-05, + "loss": 1.036, + "step": 2364 + }, + { + "epoch": 0.10471204188481675, + "grad_norm": 1.692861557006836, + "learning_rate": 1.9998658216448245e-05, + "loss": 1.0541, + "step": 2365 + }, + { + "epoch": 0.10475631758963062, + "grad_norm": 1.473457932472229, + "learning_rate": 1.9998632780938827e-05, + "loss": 1.0325, + "step": 2366 + }, + { + "epoch": 0.10480059329444451, + "grad_norm": 1.316751480102539, + "learning_rate": 1.9998607106620206e-05, + "loss": 1.0158, + "step": 2367 + }, + { + "epoch": 0.10484486899925838, + "grad_norm": 1.3122444152832031, + "learning_rate": 1.999858119349299e-05, + "loss": 0.9888, + "step": 2368 + }, + { + "epoch": 0.10488914470407226, + "grad_norm": 1.74657142162323, + "learning_rate": 1.999855504155781e-05, + "loss": 0.9822, + "step": 2369 + }, + { + "epoch": 0.10493342040888613, + "grad_norm": 1.4364701509475708, + "learning_rate": 1.9998528650815283e-05, + "loss": 1.0264, + "step": 2370 + }, + { + "epoch": 0.1049776961137, + "grad_norm": 1.51537024974823, + "learning_rate": 1.999850202126604e-05, + "loss": 1.0391, + "step": 2371 + }, + { + "epoch": 0.10502197181851389, + "grad_norm": 1.1818716526031494, + "learning_rate": 1.9998475152910718e-05, + "loss": 0.9197, + "step": 2372 + }, + { + "epoch": 0.10506624752332776, + "grad_norm": 1.6789010763168335, + "learning_rate": 1.9998448045749957e-05, + "loss": 1.1599, + "step": 2373 + }, + { + "epoch": 0.10511052322814164, + "grad_norm": 1.5605099201202393, + "learning_rate": 1.9998420699784404e-05, + "loss": 1.0423, + "step": 2374 + }, + { + "epoch": 0.10515479893295551, + "grad_norm": 1.7054100036621094, + "learning_rate": 1.9998393115014713e-05, + "loss": 1.0309, + "step": 2375 + }, + { + "epoch": 0.1051990746377694, + "grad_norm": 1.5389935970306396, + "learning_rate": 1.9998365291441546e-05, + "loss": 1.1153, + "step": 2376 + }, + { + "epoch": 0.10524335034258327, + "grad_norm": 1.8510557413101196, + "learning_rate": 1.9998337229065562e-05, + "loss": 1.0111, + "step": 2377 + }, + { + "epoch": 0.10528762604739714, + "grad_norm": 1.6082279682159424, + "learning_rate": 1.999830892788744e-05, + "loss": 1.0896, + "step": 2378 + }, + { + "epoch": 0.10533190175221102, + "grad_norm": 1.336553692817688, + "learning_rate": 1.9998280387907845e-05, + "loss": 1.0064, + "step": 2379 + }, + { + "epoch": 0.10537617745702489, + "grad_norm": 1.2730766534805298, + "learning_rate": 1.9998251609127465e-05, + "loss": 0.9843, + "step": 2380 + }, + { + "epoch": 0.10542045316183878, + "grad_norm": 1.3207426071166992, + "learning_rate": 1.9998222591546988e-05, + "loss": 0.9495, + "step": 2381 + }, + { + "epoch": 0.10546472886665265, + "grad_norm": 1.7009093761444092, + "learning_rate": 1.9998193335167104e-05, + "loss": 1.1481, + "step": 2382 + }, + { + "epoch": 0.10550900457146652, + "grad_norm": 1.3223627805709839, + "learning_rate": 1.9998163839988513e-05, + "loss": 0.9697, + "step": 2383 + }, + { + "epoch": 0.1055532802762804, + "grad_norm": 1.45418381690979, + "learning_rate": 1.999813410601192e-05, + "loss": 1.0993, + "step": 2384 + }, + { + "epoch": 0.10559755598109427, + "grad_norm": 1.3470534086227417, + "learning_rate": 1.9998104133238034e-05, + "loss": 0.9932, + "step": 2385 + }, + { + "epoch": 0.10564183168590816, + "grad_norm": 1.6557481288909912, + "learning_rate": 1.9998073921667572e-05, + "loss": 1.0385, + "step": 2386 + }, + { + "epoch": 0.10568610739072203, + "grad_norm": 1.5131388902664185, + "learning_rate": 1.9998043471301253e-05, + "loss": 0.9792, + "step": 2387 + }, + { + "epoch": 0.1057303830955359, + "grad_norm": 1.4051510095596313, + "learning_rate": 1.999801278213981e-05, + "loss": 1.0225, + "step": 2388 + }, + { + "epoch": 0.10577465880034978, + "grad_norm": 1.3991988897323608, + "learning_rate": 1.999798185418397e-05, + "loss": 1.0205, + "step": 2389 + }, + { + "epoch": 0.10581893450516365, + "grad_norm": 1.8282272815704346, + "learning_rate": 1.9997950687434475e-05, + "loss": 1.0457, + "step": 2390 + }, + { + "epoch": 0.10586321020997753, + "grad_norm": 1.5668879747390747, + "learning_rate": 1.9997919281892066e-05, + "loss": 1.1115, + "step": 2391 + }, + { + "epoch": 0.1059074859147914, + "grad_norm": 1.767987608909607, + "learning_rate": 1.99978876375575e-05, + "loss": 1.0472, + "step": 2392 + }, + { + "epoch": 0.10595176161960528, + "grad_norm": 1.9041903018951416, + "learning_rate": 1.9997855754431525e-05, + "loss": 1.0931, + "step": 2393 + }, + { + "epoch": 0.10599603732441916, + "grad_norm": 1.5888185501098633, + "learning_rate": 1.9997823632514904e-05, + "loss": 1.0349, + "step": 2394 + }, + { + "epoch": 0.10604031302923303, + "grad_norm": 2.237416982650757, + "learning_rate": 1.999779127180841e-05, + "loss": 1.0876, + "step": 2395 + }, + { + "epoch": 0.10608458873404691, + "grad_norm": 1.771167278289795, + "learning_rate": 1.9997758672312814e-05, + "loss": 1.0762, + "step": 2396 + }, + { + "epoch": 0.10612886443886078, + "grad_norm": 1.3538235425949097, + "learning_rate": 1.9997725834028886e-05, + "loss": 0.9615, + "step": 2397 + }, + { + "epoch": 0.10617314014367467, + "grad_norm": 1.703731894493103, + "learning_rate": 1.9997692756957422e-05, + "loss": 1.084, + "step": 2398 + }, + { + "epoch": 0.10621741584848854, + "grad_norm": 1.5116655826568604, + "learning_rate": 1.9997659441099205e-05, + "loss": 1.0648, + "step": 2399 + }, + { + "epoch": 0.10626169155330241, + "grad_norm": 1.688594937324524, + "learning_rate": 1.999762588645503e-05, + "loss": 1.0413, + "step": 2400 + }, + { + "epoch": 0.1063059672581163, + "grad_norm": 1.4699115753173828, + "learning_rate": 1.9997592093025706e-05, + "loss": 1.0248, + "step": 2401 + }, + { + "epoch": 0.10635024296293016, + "grad_norm": 1.3329883813858032, + "learning_rate": 1.9997558060812034e-05, + "loss": 0.9999, + "step": 2402 + }, + { + "epoch": 0.10639451866774405, + "grad_norm": 1.5953547954559326, + "learning_rate": 1.9997523789814827e-05, + "loss": 1.0937, + "step": 2403 + }, + { + "epoch": 0.10643879437255792, + "grad_norm": 3.684173107147217, + "learning_rate": 1.9997489280034906e-05, + "loss": 1.0458, + "step": 2404 + }, + { + "epoch": 0.10648307007737179, + "grad_norm": 1.5536264181137085, + "learning_rate": 1.999745453147309e-05, + "loss": 1.0231, + "step": 2405 + }, + { + "epoch": 0.10652734578218567, + "grad_norm": 1.552537441253662, + "learning_rate": 1.999741954413022e-05, + "loss": 1.0952, + "step": 2406 + }, + { + "epoch": 0.10657162148699954, + "grad_norm": 1.2781665325164795, + "learning_rate": 1.9997384318007114e-05, + "loss": 1.0068, + "step": 2407 + }, + { + "epoch": 0.10661589719181343, + "grad_norm": 1.827873945236206, + "learning_rate": 1.9997348853104632e-05, + "loss": 1.0541, + "step": 2408 + }, + { + "epoch": 0.1066601728966273, + "grad_norm": 1.4557160139083862, + "learning_rate": 1.999731314942361e-05, + "loss": 1.0284, + "step": 2409 + }, + { + "epoch": 0.10670444860144117, + "grad_norm": 1.5570122003555298, + "learning_rate": 1.9997277206964904e-05, + "loss": 1.0628, + "step": 2410 + }, + { + "epoch": 0.10674872430625505, + "grad_norm": 1.360359787940979, + "learning_rate": 1.9997241025729372e-05, + "loss": 0.9716, + "step": 2411 + }, + { + "epoch": 0.10679300001106892, + "grad_norm": 1.57561457157135, + "learning_rate": 1.999720460571788e-05, + "loss": 1.092, + "step": 2412 + }, + { + "epoch": 0.1068372757158828, + "grad_norm": 1.443453073501587, + "learning_rate": 1.9997167946931293e-05, + "loss": 0.9796, + "step": 2413 + }, + { + "epoch": 0.10688155142069668, + "grad_norm": 1.6888171434402466, + "learning_rate": 1.9997131049370488e-05, + "loss": 1.0772, + "step": 2414 + }, + { + "epoch": 0.10692582712551055, + "grad_norm": 2.2537524700164795, + "learning_rate": 1.9997093913036353e-05, + "loss": 1.1683, + "step": 2415 + }, + { + "epoch": 0.10697010283032443, + "grad_norm": 1.3650128841400146, + "learning_rate": 1.9997056537929766e-05, + "loss": 1.0788, + "step": 2416 + }, + { + "epoch": 0.1070143785351383, + "grad_norm": 1.6113739013671875, + "learning_rate": 1.9997018924051625e-05, + "loss": 1.115, + "step": 2417 + }, + { + "epoch": 0.10705865423995219, + "grad_norm": 1.8891669511795044, + "learning_rate": 1.9996981071402824e-05, + "loss": 1.1002, + "step": 2418 + }, + { + "epoch": 0.10710292994476606, + "grad_norm": 1.4049617052078247, + "learning_rate": 1.9996942979984275e-05, + "loss": 0.9647, + "step": 2419 + }, + { + "epoch": 0.10714720564957994, + "grad_norm": 1.5633448362350464, + "learning_rate": 1.999690464979688e-05, + "loss": 1.0819, + "step": 2420 + }, + { + "epoch": 0.10719148135439381, + "grad_norm": 1.7694166898727417, + "learning_rate": 1.9996866080841558e-05, + "loss": 1.0816, + "step": 2421 + }, + { + "epoch": 0.10723575705920768, + "grad_norm": 1.3258894681930542, + "learning_rate": 1.9996827273119227e-05, + "loss": 1.0204, + "step": 2422 + }, + { + "epoch": 0.10728003276402157, + "grad_norm": 1.583303689956665, + "learning_rate": 1.9996788226630823e-05, + "loss": 1.0418, + "step": 2423 + }, + { + "epoch": 0.10732430846883544, + "grad_norm": 1.7565902471542358, + "learning_rate": 1.9996748941377265e-05, + "loss": 1.089, + "step": 2424 + }, + { + "epoch": 0.10736858417364932, + "grad_norm": 1.522261381149292, + "learning_rate": 1.99967094173595e-05, + "loss": 1.0123, + "step": 2425 + }, + { + "epoch": 0.10741285987846319, + "grad_norm": 1.4322316646575928, + "learning_rate": 1.9996669654578472e-05, + "loss": 1.0302, + "step": 2426 + }, + { + "epoch": 0.10745713558327706, + "grad_norm": 1.3016903400421143, + "learning_rate": 1.9996629653035128e-05, + "loss": 1.0031, + "step": 2427 + }, + { + "epoch": 0.10750141128809095, + "grad_norm": 1.4873008728027344, + "learning_rate": 1.999658941273042e-05, + "loss": 1.0586, + "step": 2428 + }, + { + "epoch": 0.10754568699290482, + "grad_norm": 1.2470347881317139, + "learning_rate": 1.9996548933665318e-05, + "loss": 1.0237, + "step": 2429 + }, + { + "epoch": 0.1075899626977187, + "grad_norm": 1.6610438823699951, + "learning_rate": 1.9996508215840786e-05, + "loss": 1.1121, + "step": 2430 + }, + { + "epoch": 0.10763423840253257, + "grad_norm": 1.465399146080017, + "learning_rate": 1.9996467259257792e-05, + "loss": 1.0845, + "step": 2431 + }, + { + "epoch": 0.10767851410734644, + "grad_norm": 1.5983963012695312, + "learning_rate": 1.9996426063917315e-05, + "loss": 0.9946, + "step": 2432 + }, + { + "epoch": 0.10772278981216032, + "grad_norm": 1.2481963634490967, + "learning_rate": 1.9996384629820343e-05, + "loss": 0.9866, + "step": 2433 + }, + { + "epoch": 0.1077670655169742, + "grad_norm": 1.269118070602417, + "learning_rate": 1.999634295696786e-05, + "loss": 0.9468, + "step": 2434 + }, + { + "epoch": 0.10781134122178808, + "grad_norm": 1.3486300706863403, + "learning_rate": 1.9996301045360874e-05, + "loss": 0.9791, + "step": 2435 + }, + { + "epoch": 0.10785561692660195, + "grad_norm": 1.305200219154358, + "learning_rate": 1.9996258895000368e-05, + "loss": 0.9943, + "step": 2436 + }, + { + "epoch": 0.10789989263141582, + "grad_norm": 2.312880039215088, + "learning_rate": 1.9996216505887362e-05, + "loss": 1.0267, + "step": 2437 + }, + { + "epoch": 0.1079441683362297, + "grad_norm": 2.150965929031372, + "learning_rate": 1.9996173878022864e-05, + "loss": 1.0925, + "step": 2438 + }, + { + "epoch": 0.10798844404104357, + "grad_norm": 1.4657175540924072, + "learning_rate": 1.999613101140789e-05, + "loss": 1.1507, + "step": 2439 + }, + { + "epoch": 0.10803271974585746, + "grad_norm": 1.5247095823287964, + "learning_rate": 1.9996087906043464e-05, + "loss": 1.073, + "step": 2440 + }, + { + "epoch": 0.10807699545067133, + "grad_norm": 1.5723150968551636, + "learning_rate": 1.9996044561930622e-05, + "loss": 1.1275, + "step": 2441 + }, + { + "epoch": 0.10812127115548521, + "grad_norm": 1.5413446426391602, + "learning_rate": 1.999600097907039e-05, + "loss": 1.0469, + "step": 2442 + }, + { + "epoch": 0.10816554686029908, + "grad_norm": 1.558173656463623, + "learning_rate": 1.9995957157463817e-05, + "loss": 1.0298, + "step": 2443 + }, + { + "epoch": 0.10820982256511295, + "grad_norm": 1.684139609336853, + "learning_rate": 1.9995913097111946e-05, + "loss": 1.1014, + "step": 2444 + }, + { + "epoch": 0.10825409826992684, + "grad_norm": 1.7066707611083984, + "learning_rate": 1.999586879801583e-05, + "loss": 1.0957, + "step": 2445 + }, + { + "epoch": 0.10829837397474071, + "grad_norm": 1.7478748559951782, + "learning_rate": 1.9995824260176525e-05, + "loss": 1.0891, + "step": 2446 + }, + { + "epoch": 0.10834264967955459, + "grad_norm": 1.435866355895996, + "learning_rate": 1.9995779483595096e-05, + "loss": 0.9889, + "step": 2447 + }, + { + "epoch": 0.10838692538436846, + "grad_norm": 1.5170499086380005, + "learning_rate": 1.9995734468272616e-05, + "loss": 1.0337, + "step": 2448 + }, + { + "epoch": 0.10843120108918233, + "grad_norm": 1.5129659175872803, + "learning_rate": 1.9995689214210156e-05, + "loss": 1.1504, + "step": 2449 + }, + { + "epoch": 0.10847547679399622, + "grad_norm": 1.589468002319336, + "learning_rate": 1.9995643721408796e-05, + "loss": 1.0764, + "step": 2450 + }, + { + "epoch": 0.10851975249881009, + "grad_norm": 1.6370878219604492, + "learning_rate": 1.9995597989869625e-05, + "loss": 1.0161, + "step": 2451 + }, + { + "epoch": 0.10856402820362397, + "grad_norm": 2.397343158721924, + "learning_rate": 1.9995552019593736e-05, + "loss": 1.0802, + "step": 2452 + }, + { + "epoch": 0.10860830390843784, + "grad_norm": 1.5926728248596191, + "learning_rate": 1.999550581058223e-05, + "loss": 1.1187, + "step": 2453 + }, + { + "epoch": 0.10865257961325171, + "grad_norm": 1.296072244644165, + "learning_rate": 1.99954593628362e-05, + "loss": 0.975, + "step": 2454 + }, + { + "epoch": 0.1086968553180656, + "grad_norm": 2.964416980743408, + "learning_rate": 1.999541267635676e-05, + "loss": 1.1038, + "step": 2455 + }, + { + "epoch": 0.10874113102287947, + "grad_norm": 1.4761054515838623, + "learning_rate": 1.9995365751145032e-05, + "loss": 1.0909, + "step": 2456 + }, + { + "epoch": 0.10878540672769335, + "grad_norm": 1.5763208866119385, + "learning_rate": 1.9995318587202132e-05, + "loss": 1.0755, + "step": 2457 + }, + { + "epoch": 0.10882968243250722, + "grad_norm": 1.5477805137634277, + "learning_rate": 1.9995271184529182e-05, + "loss": 1.1436, + "step": 2458 + }, + { + "epoch": 0.10887395813732109, + "grad_norm": 1.8151732683181763, + "learning_rate": 1.9995223543127322e-05, + "loss": 1.0234, + "step": 2459 + }, + { + "epoch": 0.10891823384213498, + "grad_norm": 1.3253631591796875, + "learning_rate": 1.999517566299768e-05, + "loss": 0.9886, + "step": 2460 + }, + { + "epoch": 0.10896250954694885, + "grad_norm": 1.5635547637939453, + "learning_rate": 1.999512754414141e-05, + "loss": 1.1026, + "step": 2461 + }, + { + "epoch": 0.10900678525176273, + "grad_norm": 1.3734127283096313, + "learning_rate": 1.9995079186559657e-05, + "loss": 1.0992, + "step": 2462 + }, + { + "epoch": 0.1090510609565766, + "grad_norm": 1.5416032075881958, + "learning_rate": 1.9995030590253574e-05, + "loss": 1.0514, + "step": 2463 + }, + { + "epoch": 0.10909533666139049, + "grad_norm": 1.5508729219436646, + "learning_rate": 1.999498175522433e-05, + "loss": 1.0846, + "step": 2464 + }, + { + "epoch": 0.10913961236620436, + "grad_norm": 1.3413375616073608, + "learning_rate": 1.9994932681473078e-05, + "loss": 1.0199, + "step": 2465 + }, + { + "epoch": 0.10918388807101823, + "grad_norm": 1.7543920278549194, + "learning_rate": 1.9994883369001e-05, + "loss": 1.0648, + "step": 2466 + }, + { + "epoch": 0.10922816377583211, + "grad_norm": 1.3556673526763916, + "learning_rate": 1.9994833817809268e-05, + "loss": 1.0379, + "step": 2467 + }, + { + "epoch": 0.10927243948064598, + "grad_norm": 1.7299599647521973, + "learning_rate": 1.9994784027899073e-05, + "loss": 1.0052, + "step": 2468 + }, + { + "epoch": 0.10931671518545986, + "grad_norm": 2.2008440494537354, + "learning_rate": 1.9994733999271596e-05, + "loss": 1.1173, + "step": 2469 + }, + { + "epoch": 0.10936099089027373, + "grad_norm": 1.5933549404144287, + "learning_rate": 1.9994683731928034e-05, + "loss": 1.0613, + "step": 2470 + }, + { + "epoch": 0.1094052665950876, + "grad_norm": 1.3993436098098755, + "learning_rate": 1.9994633225869592e-05, + "loss": 1.0316, + "step": 2471 + }, + { + "epoch": 0.10944954229990149, + "grad_norm": 1.3721487522125244, + "learning_rate": 1.9994582481097473e-05, + "loss": 0.9092, + "step": 2472 + }, + { + "epoch": 0.10949381800471536, + "grad_norm": 1.505555510520935, + "learning_rate": 1.999453149761289e-05, + "loss": 0.9882, + "step": 2473 + }, + { + "epoch": 0.10953809370952924, + "grad_norm": 1.2907356023788452, + "learning_rate": 1.9994480275417057e-05, + "loss": 0.9844, + "step": 2474 + }, + { + "epoch": 0.10958236941434311, + "grad_norm": 1.2372633218765259, + "learning_rate": 1.9994428814511205e-05, + "loss": 0.9777, + "step": 2475 + }, + { + "epoch": 0.10962664511915698, + "grad_norm": 1.4065914154052734, + "learning_rate": 1.9994377114896557e-05, + "loss": 1.0042, + "step": 2476 + }, + { + "epoch": 0.10967092082397087, + "grad_norm": 1.5054274797439575, + "learning_rate": 1.9994325176574346e-05, + "loss": 1.119, + "step": 2477 + }, + { + "epoch": 0.10971519652878474, + "grad_norm": 1.3189042806625366, + "learning_rate": 1.999427299954582e-05, + "loss": 0.9838, + "step": 2478 + }, + { + "epoch": 0.10975947223359862, + "grad_norm": 1.4439691305160522, + "learning_rate": 1.9994220583812218e-05, + "loss": 0.9801, + "step": 2479 + }, + { + "epoch": 0.1098037479384125, + "grad_norm": 1.8704246282577515, + "learning_rate": 1.9994167929374798e-05, + "loss": 1.0305, + "step": 2480 + }, + { + "epoch": 0.10984802364322636, + "grad_norm": 1.3158013820648193, + "learning_rate": 1.9994115036234812e-05, + "loss": 0.9272, + "step": 2481 + }, + { + "epoch": 0.10989229934804025, + "grad_norm": 1.4973936080932617, + "learning_rate": 1.9994061904393526e-05, + "loss": 1.0284, + "step": 2482 + }, + { + "epoch": 0.10993657505285412, + "grad_norm": 1.8071807622909546, + "learning_rate": 1.999400853385221e-05, + "loss": 1.0293, + "step": 2483 + }, + { + "epoch": 0.109980850757668, + "grad_norm": 1.7058292627334595, + "learning_rate": 1.999395492461214e-05, + "loss": 1.1055, + "step": 2484 + }, + { + "epoch": 0.11002512646248187, + "grad_norm": 1.533947467803955, + "learning_rate": 1.999390107667459e-05, + "loss": 1.0924, + "step": 2485 + }, + { + "epoch": 0.11006940216729576, + "grad_norm": 1.3737746477127075, + "learning_rate": 1.999384699004085e-05, + "loss": 0.9462, + "step": 2486 + }, + { + "epoch": 0.11011367787210963, + "grad_norm": 1.430591106414795, + "learning_rate": 1.9993792664712215e-05, + "loss": 1.0768, + "step": 2487 + }, + { + "epoch": 0.1101579535769235, + "grad_norm": 1.8327085971832275, + "learning_rate": 1.9993738100689976e-05, + "loss": 1.0261, + "step": 2488 + }, + { + "epoch": 0.11020222928173738, + "grad_norm": 1.4901454448699951, + "learning_rate": 1.9993683297975442e-05, + "loss": 1.0666, + "step": 2489 + }, + { + "epoch": 0.11024650498655125, + "grad_norm": 1.4043225049972534, + "learning_rate": 1.999362825656992e-05, + "loss": 1.0433, + "step": 2490 + }, + { + "epoch": 0.11029078069136514, + "grad_norm": 1.416422724723816, + "learning_rate": 1.999357297647472e-05, + "loss": 1.0623, + "step": 2491 + }, + { + "epoch": 0.11033505639617901, + "grad_norm": 1.2288448810577393, + "learning_rate": 1.999351745769117e-05, + "loss": 0.9676, + "step": 2492 + }, + { + "epoch": 0.11037933210099288, + "grad_norm": 1.7286303043365479, + "learning_rate": 1.9993461700220593e-05, + "loss": 0.9971, + "step": 2493 + }, + { + "epoch": 0.11042360780580676, + "grad_norm": 1.403350591659546, + "learning_rate": 1.999340570406432e-05, + "loss": 0.978, + "step": 2494 + }, + { + "epoch": 0.11046788351062063, + "grad_norm": 1.3454604148864746, + "learning_rate": 1.9993349469223687e-05, + "loss": 0.9553, + "step": 2495 + }, + { + "epoch": 0.11051215921543452, + "grad_norm": 1.507865071296692, + "learning_rate": 1.999329299570004e-05, + "loss": 1.0557, + "step": 2496 + }, + { + "epoch": 0.11055643492024839, + "grad_norm": 1.207866907119751, + "learning_rate": 1.9993236283494726e-05, + "loss": 0.9241, + "step": 2497 + }, + { + "epoch": 0.11060071062506226, + "grad_norm": 1.7231765985488892, + "learning_rate": 1.9993179332609103e-05, + "loss": 1.1312, + "step": 2498 + }, + { + "epoch": 0.11064498632987614, + "grad_norm": 1.3252167701721191, + "learning_rate": 1.999312214304453e-05, + "loss": 0.9307, + "step": 2499 + }, + { + "epoch": 0.11068926203469001, + "grad_norm": 1.404295802116394, + "learning_rate": 1.9993064714802364e-05, + "loss": 0.9998, + "step": 2500 + }, + { + "epoch": 0.1107335377395039, + "grad_norm": 1.9589730501174927, + "learning_rate": 1.9993007047883988e-05, + "loss": 1.1206, + "step": 2501 + }, + { + "epoch": 0.11077781344431777, + "grad_norm": 1.5650798082351685, + "learning_rate": 1.9992949142290775e-05, + "loss": 1.0945, + "step": 2502 + }, + { + "epoch": 0.11082208914913164, + "grad_norm": 1.6030728816986084, + "learning_rate": 1.9992890998024108e-05, + "loss": 1.0584, + "step": 2503 + }, + { + "epoch": 0.11086636485394552, + "grad_norm": 1.5505644083023071, + "learning_rate": 1.9992832615085372e-05, + "loss": 1.0755, + "step": 2504 + }, + { + "epoch": 0.11091064055875939, + "grad_norm": 1.270106554031372, + "learning_rate": 1.999277399347597e-05, + "loss": 0.9692, + "step": 2505 + }, + { + "epoch": 0.11095491626357327, + "grad_norm": 1.5607467889785767, + "learning_rate": 1.99927151331973e-05, + "loss": 1.0894, + "step": 2506 + }, + { + "epoch": 0.11099919196838715, + "grad_norm": 1.3253096342086792, + "learning_rate": 1.9992656034250762e-05, + "loss": 1.0651, + "step": 2507 + }, + { + "epoch": 0.11104346767320103, + "grad_norm": 1.7485734224319458, + "learning_rate": 1.9992596696637773e-05, + "loss": 1.1455, + "step": 2508 + }, + { + "epoch": 0.1110877433780149, + "grad_norm": 1.4854570627212524, + "learning_rate": 1.9992537120359746e-05, + "loss": 1.0544, + "step": 2509 + }, + { + "epoch": 0.11113201908282877, + "grad_norm": 1.4097332954406738, + "learning_rate": 1.9992477305418108e-05, + "loss": 1.0561, + "step": 2510 + }, + { + "epoch": 0.11117629478764265, + "grad_norm": 1.3784486055374146, + "learning_rate": 1.999241725181428e-05, + "loss": 0.9629, + "step": 2511 + }, + { + "epoch": 0.11122057049245652, + "grad_norm": 1.7519783973693848, + "learning_rate": 1.999235695954971e-05, + "loss": 1.121, + "step": 2512 + }, + { + "epoch": 0.11126484619727041, + "grad_norm": 1.4838166236877441, + "learning_rate": 1.9992296428625825e-05, + "loss": 1.0774, + "step": 2513 + }, + { + "epoch": 0.11130912190208428, + "grad_norm": 1.6040828227996826, + "learning_rate": 1.999223565904408e-05, + "loss": 1.0572, + "step": 2514 + }, + { + "epoch": 0.11135339760689815, + "grad_norm": 1.4483110904693604, + "learning_rate": 1.9992174650805917e-05, + "loss": 1.0524, + "step": 2515 + }, + { + "epoch": 0.11139767331171203, + "grad_norm": 2.452998161315918, + "learning_rate": 1.99921134039128e-05, + "loss": 1.126, + "step": 2516 + }, + { + "epoch": 0.1114419490165259, + "grad_norm": 1.3156818151474, + "learning_rate": 1.9992051918366193e-05, + "loss": 0.9999, + "step": 2517 + }, + { + "epoch": 0.11148622472133979, + "grad_norm": 1.677642583847046, + "learning_rate": 1.9991990194167562e-05, + "loss": 1.1305, + "step": 2518 + }, + { + "epoch": 0.11153050042615366, + "grad_norm": 1.5837061405181885, + "learning_rate": 1.999192823131838e-05, + "loss": 1.1259, + "step": 2519 + }, + { + "epoch": 0.11157477613096753, + "grad_norm": 1.5241878032684326, + "learning_rate": 1.9991866029820125e-05, + "loss": 1.2036, + "step": 2520 + }, + { + "epoch": 0.11161905183578141, + "grad_norm": 1.441146969795227, + "learning_rate": 1.9991803589674287e-05, + "loss": 0.992, + "step": 2521 + }, + { + "epoch": 0.11166332754059528, + "grad_norm": 1.4529260396957397, + "learning_rate": 1.999174091088236e-05, + "loss": 1.0252, + "step": 2522 + }, + { + "epoch": 0.11170760324540917, + "grad_norm": 1.3957198858261108, + "learning_rate": 1.9991677993445832e-05, + "loss": 0.9676, + "step": 2523 + }, + { + "epoch": 0.11175187895022304, + "grad_norm": 1.6099907159805298, + "learning_rate": 1.9991614837366213e-05, + "loss": 1.0686, + "step": 2524 + }, + { + "epoch": 0.11179615465503691, + "grad_norm": 1.4177970886230469, + "learning_rate": 1.9991551442645005e-05, + "loss": 1.0098, + "step": 2525 + }, + { + "epoch": 0.11184043035985079, + "grad_norm": 1.1850587129592896, + "learning_rate": 1.999148780928373e-05, + "loss": 0.9049, + "step": 2526 + }, + { + "epoch": 0.11188470606466466, + "grad_norm": 1.5146645307540894, + "learning_rate": 1.9991423937283904e-05, + "loss": 1.0607, + "step": 2527 + }, + { + "epoch": 0.11192898176947855, + "grad_norm": 1.3493880033493042, + "learning_rate": 1.999135982664705e-05, + "loss": 0.994, + "step": 2528 + }, + { + "epoch": 0.11197325747429242, + "grad_norm": 1.5541740655899048, + "learning_rate": 1.9991295477374707e-05, + "loss": 1.023, + "step": 2529 + }, + { + "epoch": 0.1120175331791063, + "grad_norm": 1.3764736652374268, + "learning_rate": 1.99912308894684e-05, + "loss": 1.0032, + "step": 2530 + }, + { + "epoch": 0.11206180888392017, + "grad_norm": 1.4118553400039673, + "learning_rate": 1.9991166062929682e-05, + "loss": 0.9076, + "step": 2531 + }, + { + "epoch": 0.11210608458873404, + "grad_norm": 1.3498287200927734, + "learning_rate": 1.9991100997760098e-05, + "loss": 1.0019, + "step": 2532 + }, + { + "epoch": 0.11215036029354793, + "grad_norm": 1.4638766050338745, + "learning_rate": 1.9991035693961203e-05, + "loss": 1.0487, + "step": 2533 + }, + { + "epoch": 0.1121946359983618, + "grad_norm": 1.7651712894439697, + "learning_rate": 1.9990970151534554e-05, + "loss": 1.0608, + "step": 2534 + }, + { + "epoch": 0.11223891170317568, + "grad_norm": 2.0269734859466553, + "learning_rate": 1.9990904370481718e-05, + "loss": 1.0598, + "step": 2535 + }, + { + "epoch": 0.11228318740798955, + "grad_norm": 1.2644954919815063, + "learning_rate": 1.9990838350804264e-05, + "loss": 0.9236, + "step": 2536 + }, + { + "epoch": 0.11232746311280342, + "grad_norm": 1.449900507926941, + "learning_rate": 1.9990772092503775e-05, + "loss": 1.0274, + "step": 2537 + }, + { + "epoch": 0.1123717388176173, + "grad_norm": 1.335121989250183, + "learning_rate": 1.9990705595581827e-05, + "loss": 0.9518, + "step": 2538 + }, + { + "epoch": 0.11241601452243118, + "grad_norm": 1.382663607597351, + "learning_rate": 1.9990638860040007e-05, + "loss": 0.9492, + "step": 2539 + }, + { + "epoch": 0.11246029022724506, + "grad_norm": 1.3446359634399414, + "learning_rate": 1.9990571885879917e-05, + "loss": 0.9823, + "step": 2540 + }, + { + "epoch": 0.11250456593205893, + "grad_norm": 1.6887791156768799, + "learning_rate": 1.999050467310315e-05, + "loss": 1.0625, + "step": 2541 + }, + { + "epoch": 0.1125488416368728, + "grad_norm": 1.4777158498764038, + "learning_rate": 1.9990437221711317e-05, + "loss": 0.9688, + "step": 2542 + }, + { + "epoch": 0.11259311734168669, + "grad_norm": 1.893403172492981, + "learning_rate": 1.9990369531706023e-05, + "loss": 1.0328, + "step": 2543 + }, + { + "epoch": 0.11263739304650056, + "grad_norm": 1.4239463806152344, + "learning_rate": 1.9990301603088885e-05, + "loss": 0.9869, + "step": 2544 + }, + { + "epoch": 0.11268166875131444, + "grad_norm": 1.5176217555999756, + "learning_rate": 1.9990233435861532e-05, + "loss": 0.9341, + "step": 2545 + }, + { + "epoch": 0.11272594445612831, + "grad_norm": 1.307322382926941, + "learning_rate": 1.9990165030025585e-05, + "loss": 0.9222, + "step": 2546 + }, + { + "epoch": 0.11277022016094218, + "grad_norm": 2.018644332885742, + "learning_rate": 1.9990096385582682e-05, + "loss": 1.2216, + "step": 2547 + }, + { + "epoch": 0.11281449586575606, + "grad_norm": 1.2714347839355469, + "learning_rate": 1.999002750253446e-05, + "loss": 1.0328, + "step": 2548 + }, + { + "epoch": 0.11285877157056994, + "grad_norm": 1.682835578918457, + "learning_rate": 1.9989958380882566e-05, + "loss": 0.9716, + "step": 2549 + }, + { + "epoch": 0.11290304727538382, + "grad_norm": 1.7822927236557007, + "learning_rate": 1.9989889020628647e-05, + "loss": 1.0864, + "step": 2550 + }, + { + "epoch": 0.11294732298019769, + "grad_norm": 1.5568044185638428, + "learning_rate": 1.998981942177437e-05, + "loss": 1.1482, + "step": 2551 + }, + { + "epoch": 0.11299159868501157, + "grad_norm": 1.38979971408844, + "learning_rate": 1.9989749584321383e-05, + "loss": 1.0658, + "step": 2552 + }, + { + "epoch": 0.11303587438982544, + "grad_norm": 1.3764138221740723, + "learning_rate": 1.998967950827136e-05, + "loss": 1.0531, + "step": 2553 + }, + { + "epoch": 0.11308015009463931, + "grad_norm": 1.3052351474761963, + "learning_rate": 1.9989609193625983e-05, + "loss": 0.9691, + "step": 2554 + }, + { + "epoch": 0.1131244257994532, + "grad_norm": 1.4833660125732422, + "learning_rate": 1.998953864038692e-05, + "loss": 0.8954, + "step": 2555 + }, + { + "epoch": 0.11316870150426707, + "grad_norm": 1.3335565328598022, + "learning_rate": 1.998946784855586e-05, + "loss": 0.9575, + "step": 2556 + }, + { + "epoch": 0.11321297720908095, + "grad_norm": 1.5931386947631836, + "learning_rate": 1.99893968181345e-05, + "loss": 1.1599, + "step": 2557 + }, + { + "epoch": 0.11325725291389482, + "grad_norm": 1.398331642150879, + "learning_rate": 1.9989325549124524e-05, + "loss": 0.9913, + "step": 2558 + }, + { + "epoch": 0.1133015286187087, + "grad_norm": 1.4710299968719482, + "learning_rate": 1.9989254041527643e-05, + "loss": 1.0693, + "step": 2559 + }, + { + "epoch": 0.11334580432352258, + "grad_norm": 1.4541902542114258, + "learning_rate": 1.998918229534556e-05, + "loss": 1.0349, + "step": 2560 + }, + { + "epoch": 0.11339008002833645, + "grad_norm": 1.3429220914840698, + "learning_rate": 1.9989110310579994e-05, + "loss": 0.949, + "step": 2561 + }, + { + "epoch": 0.11343435573315033, + "grad_norm": 1.8265639543533325, + "learning_rate": 1.9989038087232662e-05, + "loss": 0.9867, + "step": 2562 + }, + { + "epoch": 0.1134786314379642, + "grad_norm": 1.4711204767227173, + "learning_rate": 1.998896562530529e-05, + "loss": 1.0539, + "step": 2563 + }, + { + "epoch": 0.11352290714277807, + "grad_norm": 1.2681710720062256, + "learning_rate": 1.9988892924799607e-05, + "loss": 1.0057, + "step": 2564 + }, + { + "epoch": 0.11356718284759196, + "grad_norm": 1.8551278114318848, + "learning_rate": 1.9988819985717346e-05, + "loss": 1.1367, + "step": 2565 + }, + { + "epoch": 0.11361145855240583, + "grad_norm": 1.3148224353790283, + "learning_rate": 1.9988746808060256e-05, + "loss": 1.0146, + "step": 2566 + }, + { + "epoch": 0.11365573425721971, + "grad_norm": 1.3862501382827759, + "learning_rate": 1.9988673391830082e-05, + "loss": 1.0214, + "step": 2567 + }, + { + "epoch": 0.11370000996203358, + "grad_norm": 1.388787031173706, + "learning_rate": 1.9988599737028576e-05, + "loss": 1.0487, + "step": 2568 + }, + { + "epoch": 0.11374428566684745, + "grad_norm": 1.5198116302490234, + "learning_rate": 1.99885258436575e-05, + "loss": 1.0689, + "step": 2569 + }, + { + "epoch": 0.11378856137166134, + "grad_norm": 1.3833308219909668, + "learning_rate": 1.9988451711718613e-05, + "loss": 0.9469, + "step": 2570 + }, + { + "epoch": 0.11383283707647521, + "grad_norm": 1.50824773311615, + "learning_rate": 1.998837734121369e-05, + "loss": 1.0602, + "step": 2571 + }, + { + "epoch": 0.11387711278128909, + "grad_norm": 1.3703669309616089, + "learning_rate": 1.998830273214451e-05, + "loss": 0.9781, + "step": 2572 + }, + { + "epoch": 0.11392138848610296, + "grad_norm": 1.5657354593276978, + "learning_rate": 1.9988227884512854e-05, + "loss": 1.0252, + "step": 2573 + }, + { + "epoch": 0.11396566419091685, + "grad_norm": 1.1740065813064575, + "learning_rate": 1.9988152798320508e-05, + "loss": 0.8804, + "step": 2574 + }, + { + "epoch": 0.11400993989573072, + "grad_norm": 1.1950889825820923, + "learning_rate": 1.998807747356926e-05, + "loss": 0.906, + "step": 2575 + }, + { + "epoch": 0.11405421560054459, + "grad_norm": 1.2038651704788208, + "learning_rate": 1.9988001910260916e-05, + "loss": 0.9801, + "step": 2576 + }, + { + "epoch": 0.11409849130535847, + "grad_norm": 1.1793828010559082, + "learning_rate": 1.9987926108397276e-05, + "loss": 0.9619, + "step": 2577 + }, + { + "epoch": 0.11414276701017234, + "grad_norm": 1.514417290687561, + "learning_rate": 1.998785006798016e-05, + "loss": 1.0052, + "step": 2578 + }, + { + "epoch": 0.11418704271498623, + "grad_norm": 1.2780975103378296, + "learning_rate": 1.9987773789011375e-05, + "loss": 0.9007, + "step": 2579 + }, + { + "epoch": 0.1142313184198001, + "grad_norm": 1.3667060136795044, + "learning_rate": 1.9987697271492742e-05, + "loss": 0.9334, + "step": 2580 + }, + { + "epoch": 0.11427559412461397, + "grad_norm": 1.6794593334197998, + "learning_rate": 1.9987620515426097e-05, + "loss": 1.1083, + "step": 2581 + }, + { + "epoch": 0.11431986982942785, + "grad_norm": 1.512294888496399, + "learning_rate": 1.9987543520813263e-05, + "loss": 1.0614, + "step": 2582 + }, + { + "epoch": 0.11436414553424172, + "grad_norm": 1.8277583122253418, + "learning_rate": 1.9987466287656087e-05, + "loss": 1.0727, + "step": 2583 + }, + { + "epoch": 0.1144084212390556, + "grad_norm": 1.2151459455490112, + "learning_rate": 1.998738881595641e-05, + "loss": 0.8913, + "step": 2584 + }, + { + "epoch": 0.11445269694386948, + "grad_norm": 1.3868359327316284, + "learning_rate": 1.9987311105716086e-05, + "loss": 1.0327, + "step": 2585 + }, + { + "epoch": 0.11449697264868335, + "grad_norm": 1.4154260158538818, + "learning_rate": 1.9987233156936964e-05, + "loss": 0.9883, + "step": 2586 + }, + { + "epoch": 0.11454124835349723, + "grad_norm": 1.7486937046051025, + "learning_rate": 1.9987154969620914e-05, + "loss": 1.164, + "step": 2587 + }, + { + "epoch": 0.1145855240583111, + "grad_norm": 1.3080607652664185, + "learning_rate": 1.9987076543769793e-05, + "loss": 0.9859, + "step": 2588 + }, + { + "epoch": 0.11462979976312498, + "grad_norm": 1.6090738773345947, + "learning_rate": 1.9986997879385488e-05, + "loss": 0.9484, + "step": 2589 + }, + { + "epoch": 0.11467407546793885, + "grad_norm": 1.281241774559021, + "learning_rate": 1.9986918976469867e-05, + "loss": 0.9876, + "step": 2590 + }, + { + "epoch": 0.11471835117275272, + "grad_norm": 1.327552318572998, + "learning_rate": 1.998683983502482e-05, + "loss": 0.9508, + "step": 2591 + }, + { + "epoch": 0.11476262687756661, + "grad_norm": 1.2809948921203613, + "learning_rate": 1.9986760455052232e-05, + "loss": 0.9772, + "step": 2592 + }, + { + "epoch": 0.11480690258238048, + "grad_norm": 1.3647853136062622, + "learning_rate": 1.9986680836554003e-05, + "loss": 1.006, + "step": 2593 + }, + { + "epoch": 0.11485117828719436, + "grad_norm": 1.3846133947372437, + "learning_rate": 1.9986600979532034e-05, + "loss": 1.0796, + "step": 2594 + }, + { + "epoch": 0.11489545399200823, + "grad_norm": 1.3008337020874023, + "learning_rate": 1.9986520883988233e-05, + "loss": 1.0486, + "step": 2595 + }, + { + "epoch": 0.1149397296968221, + "grad_norm": 1.970136046409607, + "learning_rate": 1.998644054992451e-05, + "loss": 1.1284, + "step": 2596 + }, + { + "epoch": 0.11498400540163599, + "grad_norm": 1.5878725051879883, + "learning_rate": 1.9986359977342786e-05, + "loss": 1.1904, + "step": 2597 + }, + { + "epoch": 0.11502828110644986, + "grad_norm": 1.5361979007720947, + "learning_rate": 1.9986279166244986e-05, + "loss": 1.0654, + "step": 2598 + }, + { + "epoch": 0.11507255681126374, + "grad_norm": 1.3426119089126587, + "learning_rate": 1.9986198116633038e-05, + "loss": 0.9439, + "step": 2599 + }, + { + "epoch": 0.11511683251607761, + "grad_norm": 1.3750495910644531, + "learning_rate": 1.9986116828508884e-05, + "loss": 0.981, + "step": 2600 + }, + { + "epoch": 0.1151611082208915, + "grad_norm": 1.478042721748352, + "learning_rate": 1.9986035301874452e-05, + "loss": 1.0444, + "step": 2601 + }, + { + "epoch": 0.11520538392570537, + "grad_norm": 1.5284923315048218, + "learning_rate": 1.9985953536731703e-05, + "loss": 1.06, + "step": 2602 + }, + { + "epoch": 0.11524965963051924, + "grad_norm": 1.2955151796340942, + "learning_rate": 1.998587153308259e-05, + "loss": 1.0009, + "step": 2603 + }, + { + "epoch": 0.11529393533533312, + "grad_norm": 1.571128010749817, + "learning_rate": 1.9985789290929055e-05, + "loss": 1.11, + "step": 2604 + }, + { + "epoch": 0.11533821104014699, + "grad_norm": 1.602333426475525, + "learning_rate": 1.998570681027308e-05, + "loss": 1.0499, + "step": 2605 + }, + { + "epoch": 0.11538248674496088, + "grad_norm": 1.561609148979187, + "learning_rate": 1.9985624091116628e-05, + "loss": 1.0355, + "step": 2606 + }, + { + "epoch": 0.11542676244977475, + "grad_norm": 1.3095769882202148, + "learning_rate": 1.9985541133461675e-05, + "loss": 1.023, + "step": 2607 + }, + { + "epoch": 0.11547103815458862, + "grad_norm": 1.3812525272369385, + "learning_rate": 1.9985457937310203e-05, + "loss": 0.9721, + "step": 2608 + }, + { + "epoch": 0.1155153138594025, + "grad_norm": 1.4816009998321533, + "learning_rate": 1.9985374502664197e-05, + "loss": 1.0568, + "step": 2609 + }, + { + "epoch": 0.11555958956421637, + "grad_norm": 1.6653060913085938, + "learning_rate": 1.9985290829525653e-05, + "loss": 1.108, + "step": 2610 + }, + { + "epoch": 0.11560386526903026, + "grad_norm": 1.4350581169128418, + "learning_rate": 1.9985206917896563e-05, + "loss": 1.044, + "step": 2611 + }, + { + "epoch": 0.11564814097384413, + "grad_norm": 1.4136484861373901, + "learning_rate": 1.9985122767778944e-05, + "loss": 1.0789, + "step": 2612 + }, + { + "epoch": 0.115692416678658, + "grad_norm": 1.6706053018569946, + "learning_rate": 1.9985038379174793e-05, + "loss": 1.013, + "step": 2613 + }, + { + "epoch": 0.11573669238347188, + "grad_norm": 1.188692331314087, + "learning_rate": 1.998495375208613e-05, + "loss": 0.928, + "step": 2614 + }, + { + "epoch": 0.11578096808828575, + "grad_norm": 1.3994988203048706, + "learning_rate": 1.9984868886514977e-05, + "loss": 0.9967, + "step": 2615 + }, + { + "epoch": 0.11582524379309964, + "grad_norm": 1.362666368484497, + "learning_rate": 1.9984783782463364e-05, + "loss": 1.0216, + "step": 2616 + }, + { + "epoch": 0.1158695194979135, + "grad_norm": 1.5770248174667358, + "learning_rate": 1.998469843993332e-05, + "loss": 1.1322, + "step": 2617 + }, + { + "epoch": 0.11591379520272738, + "grad_norm": 1.455720067024231, + "learning_rate": 1.998461285892688e-05, + "loss": 1.0698, + "step": 2618 + }, + { + "epoch": 0.11595807090754126, + "grad_norm": 1.7075735330581665, + "learning_rate": 1.9984527039446093e-05, + "loss": 1.0258, + "step": 2619 + }, + { + "epoch": 0.11600234661235513, + "grad_norm": 1.2749110460281372, + "learning_rate": 1.9984440981493004e-05, + "loss": 0.9837, + "step": 2620 + }, + { + "epoch": 0.11604662231716902, + "grad_norm": 1.435204267501831, + "learning_rate": 1.998435468506968e-05, + "loss": 1.0589, + "step": 2621 + }, + { + "epoch": 0.11609089802198289, + "grad_norm": 1.5850462913513184, + "learning_rate": 1.998426815017817e-05, + "loss": 1.0953, + "step": 2622 + }, + { + "epoch": 0.11613517372679677, + "grad_norm": 1.5259791612625122, + "learning_rate": 1.9984181376820542e-05, + "loss": 1.0279, + "step": 2623 + }, + { + "epoch": 0.11617944943161064, + "grad_norm": 1.361940860748291, + "learning_rate": 1.998409436499887e-05, + "loss": 1.0422, + "step": 2624 + }, + { + "epoch": 0.11622372513642451, + "grad_norm": 1.267163634300232, + "learning_rate": 1.9984007114715237e-05, + "loss": 0.9574, + "step": 2625 + }, + { + "epoch": 0.1162680008412384, + "grad_norm": 1.3305777311325073, + "learning_rate": 1.9983919625971726e-05, + "loss": 0.9762, + "step": 2626 + }, + { + "epoch": 0.11631227654605227, + "grad_norm": 1.5497853755950928, + "learning_rate": 1.998383189877042e-05, + "loss": 1.0593, + "step": 2627 + }, + { + "epoch": 0.11635655225086615, + "grad_norm": 1.2919013500213623, + "learning_rate": 1.998374393311342e-05, + "loss": 1.0239, + "step": 2628 + }, + { + "epoch": 0.11640082795568002, + "grad_norm": 1.6781479120254517, + "learning_rate": 1.9983655729002823e-05, + "loss": 1.1025, + "step": 2629 + }, + { + "epoch": 0.11644510366049389, + "grad_norm": 1.5083775520324707, + "learning_rate": 1.9983567286440738e-05, + "loss": 1.0581, + "step": 2630 + }, + { + "epoch": 0.11648937936530777, + "grad_norm": 1.5113282203674316, + "learning_rate": 1.998347860542928e-05, + "loss": 1.0514, + "step": 2631 + }, + { + "epoch": 0.11653365507012164, + "grad_norm": 1.4215883016586304, + "learning_rate": 1.9983389685970563e-05, + "loss": 0.9696, + "step": 2632 + }, + { + "epoch": 0.11657793077493553, + "grad_norm": 1.2929961681365967, + "learning_rate": 1.9983300528066715e-05, + "loss": 1.0266, + "step": 2633 + }, + { + "epoch": 0.1166222064797494, + "grad_norm": 1.369185447692871, + "learning_rate": 1.9983211131719856e-05, + "loss": 1.044, + "step": 2634 + }, + { + "epoch": 0.11666648218456327, + "grad_norm": 1.1963374614715576, + "learning_rate": 1.9983121496932135e-05, + "loss": 0.9362, + "step": 2635 + }, + { + "epoch": 0.11671075788937715, + "grad_norm": 1.3209497928619385, + "learning_rate": 1.998303162370568e-05, + "loss": 0.952, + "step": 2636 + }, + { + "epoch": 0.11675503359419102, + "grad_norm": 1.532938003540039, + "learning_rate": 1.9982941512042645e-05, + "loss": 1.1242, + "step": 2637 + }, + { + "epoch": 0.11679930929900491, + "grad_norm": 1.518298625946045, + "learning_rate": 1.9982851161945182e-05, + "loss": 1.0216, + "step": 2638 + }, + { + "epoch": 0.11684358500381878, + "grad_norm": 1.4562395811080933, + "learning_rate": 1.9982760573415442e-05, + "loss": 1.0563, + "step": 2639 + }, + { + "epoch": 0.11688786070863265, + "grad_norm": 1.208311676979065, + "learning_rate": 1.99826697464556e-05, + "loss": 0.9204, + "step": 2640 + }, + { + "epoch": 0.11693213641344653, + "grad_norm": 1.7978755235671997, + "learning_rate": 1.9982578681067816e-05, + "loss": 1.0612, + "step": 2641 + }, + { + "epoch": 0.1169764121182604, + "grad_norm": 1.3579381704330444, + "learning_rate": 1.998248737725427e-05, + "loss": 1.0393, + "step": 2642 + }, + { + "epoch": 0.11702068782307429, + "grad_norm": 1.3186849355697632, + "learning_rate": 1.998239583501714e-05, + "loss": 1.0444, + "step": 2643 + }, + { + "epoch": 0.11706496352788816, + "grad_norm": 1.6131112575531006, + "learning_rate": 1.9982304054358615e-05, + "loss": 1.0995, + "step": 2644 + }, + { + "epoch": 0.11710923923270204, + "grad_norm": 1.552789568901062, + "learning_rate": 1.9982212035280883e-05, + "loss": 1.0744, + "step": 2645 + }, + { + "epoch": 0.11715351493751591, + "grad_norm": 1.3831266164779663, + "learning_rate": 1.9982119777786145e-05, + "loss": 1.032, + "step": 2646 + }, + { + "epoch": 0.11719779064232978, + "grad_norm": 1.221823811531067, + "learning_rate": 1.9982027281876607e-05, + "loss": 0.9573, + "step": 2647 + }, + { + "epoch": 0.11724206634714367, + "grad_norm": 1.5483286380767822, + "learning_rate": 1.9981934547554473e-05, + "loss": 0.9857, + "step": 2648 + }, + { + "epoch": 0.11728634205195754, + "grad_norm": 1.6061477661132812, + "learning_rate": 1.998184157482196e-05, + "loss": 1.0281, + "step": 2649 + }, + { + "epoch": 0.11733061775677142, + "grad_norm": 1.4232696294784546, + "learning_rate": 1.9981748363681287e-05, + "loss": 1.0231, + "step": 2650 + }, + { + "epoch": 0.11737489346158529, + "grad_norm": 1.3792834281921387, + "learning_rate": 1.9981654914134684e-05, + "loss": 0.9926, + "step": 2651 + }, + { + "epoch": 0.11741916916639916, + "grad_norm": 1.2600042819976807, + "learning_rate": 1.9981561226184382e-05, + "loss": 0.887, + "step": 2652 + }, + { + "epoch": 0.11746344487121305, + "grad_norm": 1.312341570854187, + "learning_rate": 1.9981467299832612e-05, + "loss": 0.9524, + "step": 2653 + }, + { + "epoch": 0.11750772057602692, + "grad_norm": 1.4759242534637451, + "learning_rate": 1.9981373135081624e-05, + "loss": 1.0203, + "step": 2654 + }, + { + "epoch": 0.1175519962808408, + "grad_norm": 1.4930089712142944, + "learning_rate": 1.998127873193367e-05, + "loss": 1.0173, + "step": 2655 + }, + { + "epoch": 0.11759627198565467, + "grad_norm": 1.658408522605896, + "learning_rate": 1.9981184090391e-05, + "loss": 1.126, + "step": 2656 + }, + { + "epoch": 0.11764054769046854, + "grad_norm": 1.3916960954666138, + "learning_rate": 1.9981089210455874e-05, + "loss": 0.985, + "step": 2657 + }, + { + "epoch": 0.11768482339528243, + "grad_norm": 1.3600283861160278, + "learning_rate": 1.9980994092130558e-05, + "loss": 0.9872, + "step": 2658 + }, + { + "epoch": 0.1177290991000963, + "grad_norm": 1.5451260805130005, + "learning_rate": 1.9980898735417324e-05, + "loss": 1.0042, + "step": 2659 + }, + { + "epoch": 0.11777337480491018, + "grad_norm": 1.4285898208618164, + "learning_rate": 1.998080314031845e-05, + "loss": 1.0313, + "step": 2660 + }, + { + "epoch": 0.11781765050972405, + "grad_norm": 1.641048789024353, + "learning_rate": 1.9980707306836224e-05, + "loss": 1.0551, + "step": 2661 + }, + { + "epoch": 0.11786192621453792, + "grad_norm": 1.4812192916870117, + "learning_rate": 1.9980611234972926e-05, + "loss": 1.1232, + "step": 2662 + }, + { + "epoch": 0.1179062019193518, + "grad_norm": 1.185767412185669, + "learning_rate": 1.9980514924730857e-05, + "loss": 0.9359, + "step": 2663 + }, + { + "epoch": 0.11795047762416568, + "grad_norm": 1.3982157707214355, + "learning_rate": 1.9980418376112317e-05, + "loss": 0.9951, + "step": 2664 + }, + { + "epoch": 0.11799475332897956, + "grad_norm": 1.830815076828003, + "learning_rate": 1.998032158911961e-05, + "loss": 1.0867, + "step": 2665 + }, + { + "epoch": 0.11803902903379343, + "grad_norm": 1.4472845792770386, + "learning_rate": 1.9980224563755044e-05, + "loss": 1.1127, + "step": 2666 + }, + { + "epoch": 0.11808330473860731, + "grad_norm": 1.5835065841674805, + "learning_rate": 1.9980127300020946e-05, + "loss": 1.0813, + "step": 2667 + }, + { + "epoch": 0.11812758044342118, + "grad_norm": 1.7498871088027954, + "learning_rate": 1.998002979791963e-05, + "loss": 1.0274, + "step": 2668 + }, + { + "epoch": 0.11817185614823505, + "grad_norm": 1.2852295637130737, + "learning_rate": 1.9979932057453427e-05, + "loss": 0.9346, + "step": 2669 + }, + { + "epoch": 0.11821613185304894, + "grad_norm": 1.2875243425369263, + "learning_rate": 1.997983407862468e-05, + "loss": 0.8626, + "step": 2670 + }, + { + "epoch": 0.11826040755786281, + "grad_norm": 1.8769123554229736, + "learning_rate": 1.9979735861435716e-05, + "loss": 1.0568, + "step": 2671 + }, + { + "epoch": 0.1183046832626767, + "grad_norm": 1.3867483139038086, + "learning_rate": 1.9979637405888887e-05, + "loss": 1.004, + "step": 2672 + }, + { + "epoch": 0.11834895896749056, + "grad_norm": 2.219402313232422, + "learning_rate": 1.9979538711986545e-05, + "loss": 1.225, + "step": 2673 + }, + { + "epoch": 0.11839323467230443, + "grad_norm": 1.3356231451034546, + "learning_rate": 1.9979439779731047e-05, + "loss": 1.0079, + "step": 2674 + }, + { + "epoch": 0.11843751037711832, + "grad_norm": 1.4950270652770996, + "learning_rate": 1.9979340609124757e-05, + "loss": 1.0692, + "step": 2675 + }, + { + "epoch": 0.11848178608193219, + "grad_norm": 1.5682061910629272, + "learning_rate": 1.997924120017004e-05, + "loss": 1.0801, + "step": 2676 + }, + { + "epoch": 0.11852606178674607, + "grad_norm": 1.4277212619781494, + "learning_rate": 1.9979141552869272e-05, + "loss": 1.083, + "step": 2677 + }, + { + "epoch": 0.11857033749155994, + "grad_norm": 1.2814868688583374, + "learning_rate": 1.9979041667224832e-05, + "loss": 0.9936, + "step": 2678 + }, + { + "epoch": 0.11861461319637381, + "grad_norm": 1.539208173751831, + "learning_rate": 1.997894154323911e-05, + "loss": 1.0646, + "step": 2679 + }, + { + "epoch": 0.1186588889011877, + "grad_norm": 1.2571067810058594, + "learning_rate": 1.9978841180914492e-05, + "loss": 0.9442, + "step": 2680 + }, + { + "epoch": 0.11870316460600157, + "grad_norm": 1.585160255432129, + "learning_rate": 1.997874058025338e-05, + "loss": 1.1311, + "step": 2681 + }, + { + "epoch": 0.11874744031081545, + "grad_norm": 1.5300031900405884, + "learning_rate": 1.9978639741258172e-05, + "loss": 1.1196, + "step": 2682 + }, + { + "epoch": 0.11879171601562932, + "grad_norm": 1.6593998670578003, + "learning_rate": 1.997853866393128e-05, + "loss": 1.1159, + "step": 2683 + }, + { + "epoch": 0.1188359917204432, + "grad_norm": 1.4135868549346924, + "learning_rate": 1.9978437348275116e-05, + "loss": 1.0195, + "step": 2684 + }, + { + "epoch": 0.11888026742525708, + "grad_norm": 1.6101040840148926, + "learning_rate": 1.99783357942921e-05, + "loss": 1.1291, + "step": 2685 + }, + { + "epoch": 0.11892454313007095, + "grad_norm": 1.364884376525879, + "learning_rate": 1.9978234001984664e-05, + "loss": 0.9521, + "step": 2686 + }, + { + "epoch": 0.11896881883488483, + "grad_norm": 1.3073879480361938, + "learning_rate": 1.9978131971355225e-05, + "loss": 1.0676, + "step": 2687 + }, + { + "epoch": 0.1190130945396987, + "grad_norm": 1.5563199520111084, + "learning_rate": 1.9978029702406236e-05, + "loss": 1.0897, + "step": 2688 + }, + { + "epoch": 0.11905737024451259, + "grad_norm": 1.473880410194397, + "learning_rate": 1.9977927195140126e-05, + "loss": 1.0232, + "step": 2689 + }, + { + "epoch": 0.11910164594932646, + "grad_norm": 1.3737858533859253, + "learning_rate": 1.997782444955935e-05, + "loss": 0.9497, + "step": 2690 + }, + { + "epoch": 0.11914592165414033, + "grad_norm": 1.3426365852355957, + "learning_rate": 1.9977721465666365e-05, + "loss": 0.9606, + "step": 2691 + }, + { + "epoch": 0.11919019735895421, + "grad_norm": 1.2767179012298584, + "learning_rate": 1.9977618243463624e-05, + "loss": 1.074, + "step": 2692 + }, + { + "epoch": 0.11923447306376808, + "grad_norm": 1.6839522123336792, + "learning_rate": 1.9977514782953595e-05, + "loss": 1.1162, + "step": 2693 + }, + { + "epoch": 0.11927874876858197, + "grad_norm": 1.2277060747146606, + "learning_rate": 1.997741108413875e-05, + "loss": 0.9764, + "step": 2694 + }, + { + "epoch": 0.11932302447339584, + "grad_norm": 1.3522615432739258, + "learning_rate": 1.9977307147021567e-05, + "loss": 1.0237, + "step": 2695 + }, + { + "epoch": 0.1193673001782097, + "grad_norm": 1.366915225982666, + "learning_rate": 1.9977202971604526e-05, + "loss": 0.9243, + "step": 2696 + }, + { + "epoch": 0.11941157588302359, + "grad_norm": 1.378694772720337, + "learning_rate": 1.9977098557890112e-05, + "loss": 1.0994, + "step": 2697 + }, + { + "epoch": 0.11945585158783746, + "grad_norm": 1.5520734786987305, + "learning_rate": 1.997699390588083e-05, + "loss": 0.9863, + "step": 2698 + }, + { + "epoch": 0.11950012729265135, + "grad_norm": 1.493050217628479, + "learning_rate": 1.9976889015579167e-05, + "loss": 0.9597, + "step": 2699 + }, + { + "epoch": 0.11954440299746522, + "grad_norm": 1.5637205839157104, + "learning_rate": 1.9976783886987632e-05, + "loss": 1.1141, + "step": 2700 + }, + { + "epoch": 0.11958867870227909, + "grad_norm": 2.1109132766723633, + "learning_rate": 1.9976678520108738e-05, + "loss": 1.1528, + "step": 2701 + }, + { + "epoch": 0.11963295440709297, + "grad_norm": 1.5544066429138184, + "learning_rate": 1.9976572914945004e-05, + "loss": 0.9344, + "step": 2702 + }, + { + "epoch": 0.11967723011190684, + "grad_norm": 1.6145509481430054, + "learning_rate": 1.997646707149895e-05, + "loss": 1.0349, + "step": 2703 + }, + { + "epoch": 0.11972150581672072, + "grad_norm": 1.2957905530929565, + "learning_rate": 1.9976360989773098e-05, + "loss": 1.002, + "step": 2704 + }, + { + "epoch": 0.1197657815215346, + "grad_norm": 1.205770492553711, + "learning_rate": 1.9976254669769986e-05, + "loss": 0.9451, + "step": 2705 + }, + { + "epoch": 0.11981005722634847, + "grad_norm": 1.7041782140731812, + "learning_rate": 1.997614811149216e-05, + "loss": 1.1124, + "step": 2706 + }, + { + "epoch": 0.11985433293116235, + "grad_norm": 1.4633457660675049, + "learning_rate": 1.9976041314942156e-05, + "loss": 0.9752, + "step": 2707 + }, + { + "epoch": 0.11989860863597622, + "grad_norm": 1.6639001369476318, + "learning_rate": 1.997593428012253e-05, + "loss": 1.0399, + "step": 2708 + }, + { + "epoch": 0.1199428843407901, + "grad_norm": 1.4545899629592896, + "learning_rate": 1.9975827007035835e-05, + "loss": 1.0347, + "step": 2709 + }, + { + "epoch": 0.11998716004560397, + "grad_norm": 1.6389036178588867, + "learning_rate": 1.9975719495684634e-05, + "loss": 1.0417, + "step": 2710 + }, + { + "epoch": 0.12003143575041786, + "grad_norm": 1.3793872594833374, + "learning_rate": 1.9975611746071495e-05, + "loss": 1.032, + "step": 2711 + }, + { + "epoch": 0.12007571145523173, + "grad_norm": 1.5911815166473389, + "learning_rate": 1.997550375819899e-05, + "loss": 1.1012, + "step": 2712 + }, + { + "epoch": 0.1201199871600456, + "grad_norm": 1.3814756870269775, + "learning_rate": 1.9975395532069707e-05, + "loss": 0.9824, + "step": 2713 + }, + { + "epoch": 0.12016426286485948, + "grad_norm": 1.536200761795044, + "learning_rate": 1.997528706768622e-05, + "loss": 1.0961, + "step": 2714 + }, + { + "epoch": 0.12020853856967335, + "grad_norm": 1.2695395946502686, + "learning_rate": 1.9975178365051123e-05, + "loss": 0.9993, + "step": 2715 + }, + { + "epoch": 0.12025281427448724, + "grad_norm": 1.5605711936950684, + "learning_rate": 1.997506942416701e-05, + "loss": 1.0692, + "step": 2716 + }, + { + "epoch": 0.12029708997930111, + "grad_norm": 1.524031162261963, + "learning_rate": 1.9974960245036496e-05, + "loss": 1.1235, + "step": 2717 + }, + { + "epoch": 0.12034136568411498, + "grad_norm": 1.189652681350708, + "learning_rate": 1.997485082766217e-05, + "loss": 0.9109, + "step": 2718 + }, + { + "epoch": 0.12038564138892886, + "grad_norm": 1.3905912637710571, + "learning_rate": 1.9974741172046655e-05, + "loss": 1.0205, + "step": 2719 + }, + { + "epoch": 0.12042991709374273, + "grad_norm": 1.5714415311813354, + "learning_rate": 1.9974631278192567e-05, + "loss": 1.0866, + "step": 2720 + }, + { + "epoch": 0.12047419279855662, + "grad_norm": 1.3173397779464722, + "learning_rate": 1.9974521146102535e-05, + "loss": 1.0437, + "step": 2721 + }, + { + "epoch": 0.12051846850337049, + "grad_norm": 1.6836339235305786, + "learning_rate": 1.997441077577919e-05, + "loss": 1.0928, + "step": 2722 + }, + { + "epoch": 0.12056274420818436, + "grad_norm": 1.5483604669570923, + "learning_rate": 1.997430016722516e-05, + "loss": 1.0979, + "step": 2723 + }, + { + "epoch": 0.12060701991299824, + "grad_norm": 1.340908408164978, + "learning_rate": 1.9974189320443097e-05, + "loss": 0.9575, + "step": 2724 + }, + { + "epoch": 0.12065129561781211, + "grad_norm": 1.229477047920227, + "learning_rate": 1.9974078235435636e-05, + "loss": 0.9625, + "step": 2725 + }, + { + "epoch": 0.120695571322626, + "grad_norm": 1.381065011024475, + "learning_rate": 1.9973966912205444e-05, + "loss": 1.0039, + "step": 2726 + }, + { + "epoch": 0.12073984702743987, + "grad_norm": 1.3704038858413696, + "learning_rate": 1.997385535075517e-05, + "loss": 1.0285, + "step": 2727 + }, + { + "epoch": 0.12078412273225374, + "grad_norm": 1.6667201519012451, + "learning_rate": 1.9973743551087485e-05, + "loss": 1.0223, + "step": 2728 + }, + { + "epoch": 0.12082839843706762, + "grad_norm": 1.29731023311615, + "learning_rate": 1.997363151320505e-05, + "loss": 0.9822, + "step": 2729 + }, + { + "epoch": 0.12087267414188149, + "grad_norm": 1.8199552297592163, + "learning_rate": 1.9973519237110556e-05, + "loss": 1.0486, + "step": 2730 + }, + { + "epoch": 0.12091694984669538, + "grad_norm": 1.265202283859253, + "learning_rate": 1.997340672280667e-05, + "loss": 0.9499, + "step": 2731 + }, + { + "epoch": 0.12096122555150925, + "grad_norm": 1.4771034717559814, + "learning_rate": 1.9973293970296087e-05, + "loss": 1.0202, + "step": 2732 + }, + { + "epoch": 0.12100550125632313, + "grad_norm": 1.3926784992218018, + "learning_rate": 1.9973180979581497e-05, + "loss": 0.975, + "step": 2733 + }, + { + "epoch": 0.121049776961137, + "grad_norm": 1.2801674604415894, + "learning_rate": 1.99730677506656e-05, + "loss": 0.9951, + "step": 2734 + }, + { + "epoch": 0.12109405266595087, + "grad_norm": 2.116910934448242, + "learning_rate": 1.99729542835511e-05, + "loss": 1.0818, + "step": 2735 + }, + { + "epoch": 0.12113832837076476, + "grad_norm": 1.297335147857666, + "learning_rate": 1.9972840578240705e-05, + "loss": 1.0032, + "step": 2736 + }, + { + "epoch": 0.12118260407557863, + "grad_norm": 1.2667486667633057, + "learning_rate": 1.997272663473714e-05, + "loss": 1.0029, + "step": 2737 + }, + { + "epoch": 0.12122687978039251, + "grad_norm": 1.343455195426941, + "learning_rate": 1.9972612453043112e-05, + "loss": 0.9781, + "step": 2738 + }, + { + "epoch": 0.12127115548520638, + "grad_norm": 1.4334195852279663, + "learning_rate": 1.997249803316136e-05, + "loss": 1.0476, + "step": 2739 + }, + { + "epoch": 0.12131543119002025, + "grad_norm": 1.3709564208984375, + "learning_rate": 1.997238337509461e-05, + "loss": 1.0226, + "step": 2740 + }, + { + "epoch": 0.12135970689483414, + "grad_norm": 2.1071126461029053, + "learning_rate": 1.9972268478845605e-05, + "loss": 1.1488, + "step": 2741 + }, + { + "epoch": 0.121403982599648, + "grad_norm": 1.6716413497924805, + "learning_rate": 1.9972153344417084e-05, + "loss": 0.9823, + "step": 2742 + }, + { + "epoch": 0.12144825830446189, + "grad_norm": 1.5196722745895386, + "learning_rate": 1.9972037971811802e-05, + "loss": 1.0065, + "step": 2743 + }, + { + "epoch": 0.12149253400927576, + "grad_norm": 1.480706810951233, + "learning_rate": 1.9971922361032514e-05, + "loss": 1.0082, + "step": 2744 + }, + { + "epoch": 0.12153680971408963, + "grad_norm": 1.329480528831482, + "learning_rate": 1.9971806512081978e-05, + "loss": 0.9799, + "step": 2745 + }, + { + "epoch": 0.12158108541890351, + "grad_norm": 1.5483464002609253, + "learning_rate": 1.9971690424962965e-05, + "loss": 1.0737, + "step": 2746 + }, + { + "epoch": 0.12162536112371738, + "grad_norm": 3.637451410293579, + "learning_rate": 1.997157409967824e-05, + "loss": 1.0692, + "step": 2747 + }, + { + "epoch": 0.12166963682853127, + "grad_norm": 1.3806312084197998, + "learning_rate": 1.997145753623059e-05, + "loss": 0.9381, + "step": 2748 + }, + { + "epoch": 0.12171391253334514, + "grad_norm": 1.2133750915527344, + "learning_rate": 1.99713407346228e-05, + "loss": 1.0219, + "step": 2749 + }, + { + "epoch": 0.12175818823815901, + "grad_norm": 2.0840325355529785, + "learning_rate": 1.997122369485765e-05, + "loss": 1.13, + "step": 2750 + }, + { + "epoch": 0.1218024639429729, + "grad_norm": 1.735237956047058, + "learning_rate": 1.997110641693794e-05, + "loss": 1.0518, + "step": 2751 + }, + { + "epoch": 0.12184673964778676, + "grad_norm": 1.4134176969528198, + "learning_rate": 1.9970988900866478e-05, + "loss": 1.0668, + "step": 2752 + }, + { + "epoch": 0.12189101535260065, + "grad_norm": 1.4975616931915283, + "learning_rate": 1.997087114664606e-05, + "loss": 1.1601, + "step": 2753 + }, + { + "epoch": 0.12193529105741452, + "grad_norm": 1.480912446975708, + "learning_rate": 1.9970753154279506e-05, + "loss": 1.0569, + "step": 2754 + }, + { + "epoch": 0.1219795667622284, + "grad_norm": 1.427271842956543, + "learning_rate": 1.997063492376963e-05, + "loss": 0.9401, + "step": 2755 + }, + { + "epoch": 0.12202384246704227, + "grad_norm": 1.2318041324615479, + "learning_rate": 1.9970516455119258e-05, + "loss": 0.9548, + "step": 2756 + }, + { + "epoch": 0.12206811817185614, + "grad_norm": 1.4750332832336426, + "learning_rate": 1.9970397748331216e-05, + "loss": 1.0826, + "step": 2757 + }, + { + "epoch": 0.12211239387667003, + "grad_norm": 1.3607022762298584, + "learning_rate": 1.9970278803408342e-05, + "loss": 1.0772, + "step": 2758 + }, + { + "epoch": 0.1221566695814839, + "grad_norm": 1.7440457344055176, + "learning_rate": 1.997015962035348e-05, + "loss": 1.0982, + "step": 2759 + }, + { + "epoch": 0.12220094528629778, + "grad_norm": 1.3774367570877075, + "learning_rate": 1.9970040199169474e-05, + "loss": 1.0279, + "step": 2760 + }, + { + "epoch": 0.12224522099111165, + "grad_norm": 1.6001861095428467, + "learning_rate": 1.9969920539859174e-05, + "loss": 1.0245, + "step": 2761 + }, + { + "epoch": 0.12228949669592552, + "grad_norm": 2.010239362716675, + "learning_rate": 1.996980064242544e-05, + "loss": 1.0507, + "step": 2762 + }, + { + "epoch": 0.12233377240073941, + "grad_norm": 2.4905035495758057, + "learning_rate": 1.9969680506871138e-05, + "loss": 1.1487, + "step": 2763 + }, + { + "epoch": 0.12237804810555328, + "grad_norm": 1.4256467819213867, + "learning_rate": 1.996956013319913e-05, + "loss": 1.008, + "step": 2764 + }, + { + "epoch": 0.12242232381036716, + "grad_norm": 1.3478304147720337, + "learning_rate": 1.9969439521412295e-05, + "loss": 0.9571, + "step": 2765 + }, + { + "epoch": 0.12246659951518103, + "grad_norm": 1.2441749572753906, + "learning_rate": 1.996931867151352e-05, + "loss": 0.9578, + "step": 2766 + }, + { + "epoch": 0.1225108752199949, + "grad_norm": 1.8967821598052979, + "learning_rate": 1.9969197583505683e-05, + "loss": 1.0362, + "step": 2767 + }, + { + "epoch": 0.12255515092480879, + "grad_norm": 1.4026305675506592, + "learning_rate": 1.996907625739168e-05, + "loss": 0.9588, + "step": 2768 + }, + { + "epoch": 0.12259942662962266, + "grad_norm": 1.865900993347168, + "learning_rate": 1.9968954693174405e-05, + "loss": 1.0586, + "step": 2769 + }, + { + "epoch": 0.12264370233443654, + "grad_norm": 1.5405147075653076, + "learning_rate": 1.9968832890856763e-05, + "loss": 0.9433, + "step": 2770 + }, + { + "epoch": 0.12268797803925041, + "grad_norm": 2.285244941711426, + "learning_rate": 1.996871085044167e-05, + "loss": 1.0596, + "step": 2771 + }, + { + "epoch": 0.12273225374406428, + "grad_norm": 1.4318145513534546, + "learning_rate": 1.9968588571932033e-05, + "loss": 1.0673, + "step": 2772 + }, + { + "epoch": 0.12277652944887817, + "grad_norm": 1.3296215534210205, + "learning_rate": 1.9968466055330775e-05, + "loss": 1.0547, + "step": 2773 + }, + { + "epoch": 0.12282080515369204, + "grad_norm": 2.6751797199249268, + "learning_rate": 1.9968343300640817e-05, + "loss": 1.049, + "step": 2774 + }, + { + "epoch": 0.12286508085850592, + "grad_norm": 1.1949316263198853, + "learning_rate": 1.99682203078651e-05, + "loss": 0.9391, + "step": 2775 + }, + { + "epoch": 0.12290935656331979, + "grad_norm": 1.2034579515457153, + "learning_rate": 1.996809707700656e-05, + "loss": 0.9425, + "step": 2776 + }, + { + "epoch": 0.12295363226813368, + "grad_norm": 1.5810045003890991, + "learning_rate": 1.9967973608068133e-05, + "loss": 1.0587, + "step": 2777 + }, + { + "epoch": 0.12299790797294755, + "grad_norm": 1.3988040685653687, + "learning_rate": 1.9967849901052775e-05, + "loss": 0.9953, + "step": 2778 + }, + { + "epoch": 0.12304218367776142, + "grad_norm": 1.4328535795211792, + "learning_rate": 1.9967725955963434e-05, + "loss": 1.0069, + "step": 2779 + }, + { + "epoch": 0.1230864593825753, + "grad_norm": 1.7867209911346436, + "learning_rate": 1.996760177280308e-05, + "loss": 1.1167, + "step": 2780 + }, + { + "epoch": 0.12313073508738917, + "grad_norm": 1.5281561613082886, + "learning_rate": 1.996747735157467e-05, + "loss": 1.0427, + "step": 2781 + }, + { + "epoch": 0.12317501079220305, + "grad_norm": 1.452515721321106, + "learning_rate": 1.9967352692281183e-05, + "loss": 1.0001, + "step": 2782 + }, + { + "epoch": 0.12321928649701692, + "grad_norm": 1.6113648414611816, + "learning_rate": 1.996722779492559e-05, + "loss": 1.0285, + "step": 2783 + }, + { + "epoch": 0.1232635622018308, + "grad_norm": 1.6926652193069458, + "learning_rate": 1.9967102659510877e-05, + "loss": 1.0608, + "step": 2784 + }, + { + "epoch": 0.12330783790664468, + "grad_norm": 1.4686872959136963, + "learning_rate": 1.9966977286040032e-05, + "loss": 1.0743, + "step": 2785 + }, + { + "epoch": 0.12335211361145855, + "grad_norm": 1.2938355207443237, + "learning_rate": 1.9966851674516052e-05, + "loss": 0.9868, + "step": 2786 + }, + { + "epoch": 0.12339638931627243, + "grad_norm": 1.455781102180481, + "learning_rate": 1.9966725824941933e-05, + "loss": 0.9905, + "step": 2787 + }, + { + "epoch": 0.1234406650210863, + "grad_norm": 1.7893480062484741, + "learning_rate": 1.9966599737320683e-05, + "loss": 1.0407, + "step": 2788 + }, + { + "epoch": 0.12348494072590017, + "grad_norm": 1.6877187490463257, + "learning_rate": 1.9966473411655318e-05, + "loss": 0.9775, + "step": 2789 + }, + { + "epoch": 0.12352921643071406, + "grad_norm": 1.3990607261657715, + "learning_rate": 1.9966346847948842e-05, + "loss": 0.9938, + "step": 2790 + }, + { + "epoch": 0.12357349213552793, + "grad_norm": 1.4718372821807861, + "learning_rate": 1.9966220046204295e-05, + "loss": 0.946, + "step": 2791 + }, + { + "epoch": 0.12361776784034181, + "grad_norm": 1.5941723585128784, + "learning_rate": 1.996609300642469e-05, + "loss": 1.0057, + "step": 2792 + }, + { + "epoch": 0.12366204354515568, + "grad_norm": 1.5685548782348633, + "learning_rate": 1.996596572861307e-05, + "loss": 1.0252, + "step": 2793 + }, + { + "epoch": 0.12370631924996955, + "grad_norm": 1.7806034088134766, + "learning_rate": 1.996583821277248e-05, + "loss": 0.9945, + "step": 2794 + }, + { + "epoch": 0.12375059495478344, + "grad_norm": 1.3274998664855957, + "learning_rate": 1.9965710458905952e-05, + "loss": 1.0182, + "step": 2795 + }, + { + "epoch": 0.12379487065959731, + "grad_norm": 1.3495897054672241, + "learning_rate": 1.9965582467016543e-05, + "loss": 1.0265, + "step": 2796 + }, + { + "epoch": 0.12383914636441119, + "grad_norm": 1.22895085811615, + "learning_rate": 1.9965454237107312e-05, + "loss": 0.9257, + "step": 2797 + }, + { + "epoch": 0.12388342206922506, + "grad_norm": 1.2395213842391968, + "learning_rate": 1.9965325769181324e-05, + "loss": 0.9486, + "step": 2798 + }, + { + "epoch": 0.12392769777403895, + "grad_norm": 1.6119784116744995, + "learning_rate": 1.9965197063241643e-05, + "loss": 1.046, + "step": 2799 + }, + { + "epoch": 0.12397197347885282, + "grad_norm": 1.3840001821517944, + "learning_rate": 1.9965068119291345e-05, + "loss": 1.0349, + "step": 2800 + }, + { + "epoch": 0.12401624918366669, + "grad_norm": 1.185795545578003, + "learning_rate": 1.9964938937333507e-05, + "loss": 0.8934, + "step": 2801 + }, + { + "epoch": 0.12406052488848057, + "grad_norm": 1.4427211284637451, + "learning_rate": 1.9964809517371214e-05, + "loss": 1.0452, + "step": 2802 + }, + { + "epoch": 0.12410480059329444, + "grad_norm": 1.3504077196121216, + "learning_rate": 1.9964679859407565e-05, + "loss": 0.9876, + "step": 2803 + }, + { + "epoch": 0.12414907629810833, + "grad_norm": 1.6289540529251099, + "learning_rate": 1.9964549963445644e-05, + "loss": 1.024, + "step": 2804 + }, + { + "epoch": 0.1241933520029222, + "grad_norm": 1.303156852722168, + "learning_rate": 1.9964419829488567e-05, + "loss": 0.9939, + "step": 2805 + }, + { + "epoch": 0.12423762770773607, + "grad_norm": 1.4221343994140625, + "learning_rate": 1.9964289457539434e-05, + "loss": 1.0057, + "step": 2806 + }, + { + "epoch": 0.12428190341254995, + "grad_norm": 1.7984222173690796, + "learning_rate": 1.9964158847601357e-05, + "loss": 1.0093, + "step": 2807 + }, + { + "epoch": 0.12432617911736382, + "grad_norm": 1.3535646200180054, + "learning_rate": 1.996402799967746e-05, + "loss": 0.9738, + "step": 2808 + }, + { + "epoch": 0.1243704548221777, + "grad_norm": 1.3317341804504395, + "learning_rate": 1.996389691377087e-05, + "loss": 0.9719, + "step": 2809 + }, + { + "epoch": 0.12441473052699158, + "grad_norm": 6.409813404083252, + "learning_rate": 1.9963765589884713e-05, + "loss": 0.9837, + "step": 2810 + }, + { + "epoch": 0.12445900623180545, + "grad_norm": 1.3454813957214355, + "learning_rate": 1.9963634028022133e-05, + "loss": 1.0237, + "step": 2811 + }, + { + "epoch": 0.12450328193661933, + "grad_norm": 1.3318979740142822, + "learning_rate": 1.996350222818626e-05, + "loss": 0.9783, + "step": 2812 + }, + { + "epoch": 0.1245475576414332, + "grad_norm": 1.5900192260742188, + "learning_rate": 1.996337019038025e-05, + "loss": 1.0747, + "step": 2813 + }, + { + "epoch": 0.12459183334624709, + "grad_norm": 1.2950631380081177, + "learning_rate": 1.9963237914607257e-05, + "loss": 0.999, + "step": 2814 + }, + { + "epoch": 0.12463610905106096, + "grad_norm": 1.4442540407180786, + "learning_rate": 1.996310540087044e-05, + "loss": 1.0292, + "step": 2815 + }, + { + "epoch": 0.12468038475587483, + "grad_norm": 1.663118839263916, + "learning_rate": 1.996297264917296e-05, + "loss": 1.1141, + "step": 2816 + }, + { + "epoch": 0.12472466046068871, + "grad_norm": 2.083082675933838, + "learning_rate": 1.996283965951799e-05, + "loss": 1.1023, + "step": 2817 + }, + { + "epoch": 0.12476893616550258, + "grad_norm": 1.3199480772018433, + "learning_rate": 1.9962706431908714e-05, + "loss": 0.9729, + "step": 2818 + }, + { + "epoch": 0.12481321187031647, + "grad_norm": 1.6527057886123657, + "learning_rate": 1.99625729663483e-05, + "loss": 1.0591, + "step": 2819 + }, + { + "epoch": 0.12485748757513034, + "grad_norm": 1.4346067905426025, + "learning_rate": 1.996243926283994e-05, + "loss": 1.0689, + "step": 2820 + }, + { + "epoch": 0.1249017632799442, + "grad_norm": 1.4454642534255981, + "learning_rate": 1.9962305321386835e-05, + "loss": 0.934, + "step": 2821 + }, + { + "epoch": 0.12494603898475809, + "grad_norm": 1.6451261043548584, + "learning_rate": 1.996217114199218e-05, + "loss": 1.0738, + "step": 2822 + }, + { + "epoch": 0.12499031468957196, + "grad_norm": 1.6221246719360352, + "learning_rate": 1.9962036724659178e-05, + "loss": 1.1223, + "step": 2823 + }, + { + "epoch": 0.12503459039438583, + "grad_norm": 1.1973432302474976, + "learning_rate": 1.9961902069391038e-05, + "loss": 0.9476, + "step": 2824 + }, + { + "epoch": 0.12507886609919971, + "grad_norm": 1.3902616500854492, + "learning_rate": 1.9961767176190977e-05, + "loss": 1.0364, + "step": 2825 + }, + { + "epoch": 0.1251231418040136, + "grad_norm": 1.7233208417892456, + "learning_rate": 1.9961632045062224e-05, + "loss": 1.0443, + "step": 2826 + }, + { + "epoch": 0.12516741750882746, + "grad_norm": 1.707234501838684, + "learning_rate": 1.9961496676007997e-05, + "loss": 1.0572, + "step": 2827 + }, + { + "epoch": 0.12521169321364134, + "grad_norm": 1.474415898323059, + "learning_rate": 1.996136106903153e-05, + "loss": 1.0944, + "step": 2828 + }, + { + "epoch": 0.12525596891845522, + "grad_norm": 1.2472971677780151, + "learning_rate": 1.9961225224136068e-05, + "loss": 1.0072, + "step": 2829 + }, + { + "epoch": 0.1253002446232691, + "grad_norm": 1.2968947887420654, + "learning_rate": 1.9961089141324853e-05, + "loss": 0.8967, + "step": 2830 + }, + { + "epoch": 0.12534452032808296, + "grad_norm": 1.3232539892196655, + "learning_rate": 1.9960952820601135e-05, + "loss": 1.0149, + "step": 2831 + }, + { + "epoch": 0.12538879603289685, + "grad_norm": 1.4221141338348389, + "learning_rate": 1.996081626196817e-05, + "loss": 1.007, + "step": 2832 + }, + { + "epoch": 0.12543307173771073, + "grad_norm": 1.3634412288665771, + "learning_rate": 1.9960679465429215e-05, + "loss": 1.0509, + "step": 2833 + }, + { + "epoch": 0.1254773474425246, + "grad_norm": 1.726627230644226, + "learning_rate": 1.9960542430987543e-05, + "loss": 1.0847, + "step": 2834 + }, + { + "epoch": 0.12552162314733847, + "grad_norm": 1.4322607517242432, + "learning_rate": 1.9960405158646425e-05, + "loss": 1.0409, + "step": 2835 + }, + { + "epoch": 0.12556589885215236, + "grad_norm": 1.4721981287002563, + "learning_rate": 1.996026764840914e-05, + "loss": 0.9574, + "step": 2836 + }, + { + "epoch": 0.12561017455696624, + "grad_norm": 1.325456142425537, + "learning_rate": 1.996012990027897e-05, + "loss": 0.9912, + "step": 2837 + }, + { + "epoch": 0.1256544502617801, + "grad_norm": 1.2501366138458252, + "learning_rate": 1.9959991914259212e-05, + "loss": 0.9287, + "step": 2838 + }, + { + "epoch": 0.12569872596659398, + "grad_norm": 1.5046064853668213, + "learning_rate": 1.9959853690353156e-05, + "loss": 1.0126, + "step": 2839 + }, + { + "epoch": 0.12574300167140787, + "grad_norm": 1.121896743774414, + "learning_rate": 1.99597152285641e-05, + "loss": 0.9101, + "step": 2840 + }, + { + "epoch": 0.12578727737622172, + "grad_norm": 1.341927170753479, + "learning_rate": 1.9959576528895357e-05, + "loss": 0.9398, + "step": 2841 + }, + { + "epoch": 0.1258315530810356, + "grad_norm": 1.2778605222702026, + "learning_rate": 1.995943759135024e-05, + "loss": 1.0014, + "step": 2842 + }, + { + "epoch": 0.1258758287858495, + "grad_norm": 1.4451007843017578, + "learning_rate": 1.995929841593206e-05, + "loss": 1.0334, + "step": 2843 + }, + { + "epoch": 0.12592010449066335, + "grad_norm": 1.2351019382476807, + "learning_rate": 1.9959159002644154e-05, + "loss": 0.9705, + "step": 2844 + }, + { + "epoch": 0.12596438019547723, + "grad_norm": 1.337778091430664, + "learning_rate": 1.9959019351489842e-05, + "loss": 0.998, + "step": 2845 + }, + { + "epoch": 0.12600865590029112, + "grad_norm": 1.563584566116333, + "learning_rate": 1.995887946247246e-05, + "loss": 1.0094, + "step": 2846 + }, + { + "epoch": 0.126052931605105, + "grad_norm": 1.4733302593231201, + "learning_rate": 1.995873933559535e-05, + "loss": 1.1041, + "step": 2847 + }, + { + "epoch": 0.12609720730991886, + "grad_norm": 1.258105993270874, + "learning_rate": 1.9958598970861863e-05, + "loss": 0.9628, + "step": 2848 + }, + { + "epoch": 0.12614148301473274, + "grad_norm": 1.4503825902938843, + "learning_rate": 1.9958458368275347e-05, + "loss": 1.0046, + "step": 2849 + }, + { + "epoch": 0.12618575871954663, + "grad_norm": 1.2611629962921143, + "learning_rate": 1.995831752783916e-05, + "loss": 1.0049, + "step": 2850 + }, + { + "epoch": 0.12623003442436048, + "grad_norm": 1.4200092554092407, + "learning_rate": 1.995817644955667e-05, + "loss": 1.0045, + "step": 2851 + }, + { + "epoch": 0.12627431012917437, + "grad_norm": 1.204239010810852, + "learning_rate": 1.995803513343124e-05, + "loss": 0.9627, + "step": 2852 + }, + { + "epoch": 0.12631858583398825, + "grad_norm": 1.4170923233032227, + "learning_rate": 1.9957893579466252e-05, + "loss": 1.0549, + "step": 2853 + }, + { + "epoch": 0.1263628615388021, + "grad_norm": 1.5505629777908325, + "learning_rate": 1.9957751787665082e-05, + "loss": 1.0485, + "step": 2854 + }, + { + "epoch": 0.126407137243616, + "grad_norm": 1.7558552026748657, + "learning_rate": 1.9957609758031118e-05, + "loss": 1.0089, + "step": 2855 + }, + { + "epoch": 0.12645141294842988, + "grad_norm": 1.4019681215286255, + "learning_rate": 1.9957467490567754e-05, + "loss": 1.0542, + "step": 2856 + }, + { + "epoch": 0.12649568865324376, + "grad_norm": 1.4855326414108276, + "learning_rate": 1.9957324985278386e-05, + "loss": 1.1004, + "step": 2857 + }, + { + "epoch": 0.12653996435805762, + "grad_norm": 1.2352043390274048, + "learning_rate": 1.9957182242166418e-05, + "loss": 0.9539, + "step": 2858 + }, + { + "epoch": 0.1265842400628715, + "grad_norm": 1.2431749105453491, + "learning_rate": 1.9957039261235263e-05, + "loss": 0.9466, + "step": 2859 + }, + { + "epoch": 0.12662851576768538, + "grad_norm": 1.3561066389083862, + "learning_rate": 1.995689604248833e-05, + "loss": 0.994, + "step": 2860 + }, + { + "epoch": 0.12667279147249924, + "grad_norm": 1.5927268266677856, + "learning_rate": 1.995675258592904e-05, + "loss": 1.046, + "step": 2861 + }, + { + "epoch": 0.12671706717731313, + "grad_norm": 1.4702794551849365, + "learning_rate": 1.9956608891560822e-05, + "loss": 1.0228, + "step": 2862 + }, + { + "epoch": 0.126761342882127, + "grad_norm": 2.010741710662842, + "learning_rate": 1.9956464959387113e-05, + "loss": 1.1003, + "step": 2863 + }, + { + "epoch": 0.1268056185869409, + "grad_norm": 1.8092687129974365, + "learning_rate": 1.9956320789411338e-05, + "loss": 1.0956, + "step": 2864 + }, + { + "epoch": 0.12684989429175475, + "grad_norm": 1.503879427909851, + "learning_rate": 1.995617638163695e-05, + "loss": 1.0619, + "step": 2865 + }, + { + "epoch": 0.12689416999656863, + "grad_norm": 1.101231575012207, + "learning_rate": 1.99560317360674e-05, + "loss": 0.905, + "step": 2866 + }, + { + "epoch": 0.12693844570138252, + "grad_norm": 1.337534785270691, + "learning_rate": 1.9955886852706135e-05, + "loss": 0.9726, + "step": 2867 + }, + { + "epoch": 0.12698272140619637, + "grad_norm": 1.4935433864593506, + "learning_rate": 1.995574173155662e-05, + "loss": 1.0638, + "step": 2868 + }, + { + "epoch": 0.12702699711101026, + "grad_norm": 1.403710126876831, + "learning_rate": 1.9955596372622318e-05, + "loss": 1.0163, + "step": 2869 + }, + { + "epoch": 0.12707127281582414, + "grad_norm": 1.3357632160186768, + "learning_rate": 1.9955450775906705e-05, + "loss": 0.9724, + "step": 2870 + }, + { + "epoch": 0.127115548520638, + "grad_norm": 1.3916068077087402, + "learning_rate": 1.9955304941413257e-05, + "loss": 1.0315, + "step": 2871 + }, + { + "epoch": 0.12715982422545188, + "grad_norm": 1.2041929960250854, + "learning_rate": 1.9955158869145454e-05, + "loss": 1.038, + "step": 2872 + }, + { + "epoch": 0.12720409993026577, + "grad_norm": 1.6568642854690552, + "learning_rate": 1.9955012559106788e-05, + "loss": 1.0059, + "step": 2873 + }, + { + "epoch": 0.12724837563507965, + "grad_norm": 1.710754156112671, + "learning_rate": 1.9954866011300752e-05, + "loss": 1.0376, + "step": 2874 + }, + { + "epoch": 0.1272926513398935, + "grad_norm": 1.6058497428894043, + "learning_rate": 1.9954719225730847e-05, + "loss": 1.052, + "step": 2875 + }, + { + "epoch": 0.1273369270447074, + "grad_norm": 1.3780564069747925, + "learning_rate": 1.995457220240058e-05, + "loss": 0.9402, + "step": 2876 + }, + { + "epoch": 0.12738120274952128, + "grad_norm": 1.1796892881393433, + "learning_rate": 1.995442494131346e-05, + "loss": 0.9418, + "step": 2877 + }, + { + "epoch": 0.12742547845433513, + "grad_norm": 1.2680741548538208, + "learning_rate": 1.9954277442473007e-05, + "loss": 0.9328, + "step": 2878 + }, + { + "epoch": 0.12746975415914902, + "grad_norm": 1.3648381233215332, + "learning_rate": 1.9954129705882742e-05, + "loss": 0.9021, + "step": 2879 + }, + { + "epoch": 0.1275140298639629, + "grad_norm": 1.446608304977417, + "learning_rate": 1.9953981731546192e-05, + "loss": 1.0259, + "step": 2880 + }, + { + "epoch": 0.1275583055687768, + "grad_norm": 1.3321632146835327, + "learning_rate": 1.9953833519466894e-05, + "loss": 1.0, + "step": 2881 + }, + { + "epoch": 0.12760258127359064, + "grad_norm": 1.5635610818862915, + "learning_rate": 1.9953685069648386e-05, + "loss": 1.0404, + "step": 2882 + }, + { + "epoch": 0.12764685697840453, + "grad_norm": 1.8234071731567383, + "learning_rate": 1.9953536382094218e-05, + "loss": 1.072, + "step": 2883 + }, + { + "epoch": 0.1276911326832184, + "grad_norm": 1.2142102718353271, + "learning_rate": 1.9953387456807937e-05, + "loss": 0.9783, + "step": 2884 + }, + { + "epoch": 0.12773540838803227, + "grad_norm": 1.4589364528656006, + "learning_rate": 1.99532382937931e-05, + "loss": 1.0104, + "step": 2885 + }, + { + "epoch": 0.12777968409284615, + "grad_norm": 1.2956701517105103, + "learning_rate": 1.9953088893053268e-05, + "loss": 0.9485, + "step": 2886 + }, + { + "epoch": 0.12782395979766004, + "grad_norm": 1.268857717514038, + "learning_rate": 1.9952939254592015e-05, + "loss": 0.971, + "step": 2887 + }, + { + "epoch": 0.1278682355024739, + "grad_norm": 1.24044668674469, + "learning_rate": 1.9952789378412914e-05, + "loss": 1.0148, + "step": 2888 + }, + { + "epoch": 0.12791251120728778, + "grad_norm": 1.507356882095337, + "learning_rate": 1.995263926451954e-05, + "loss": 1.0087, + "step": 2889 + }, + { + "epoch": 0.12795678691210166, + "grad_norm": 1.3396481275558472, + "learning_rate": 1.995248891291548e-05, + "loss": 1.0055, + "step": 2890 + }, + { + "epoch": 0.12800106261691555, + "grad_norm": 1.220594048500061, + "learning_rate": 1.995233832360433e-05, + "loss": 0.9041, + "step": 2891 + }, + { + "epoch": 0.1280453383217294, + "grad_norm": 1.6981695890426636, + "learning_rate": 1.995218749658968e-05, + "loss": 1.0881, + "step": 2892 + }, + { + "epoch": 0.12808961402654329, + "grad_norm": 2.793910026550293, + "learning_rate": 1.9952036431875134e-05, + "loss": 1.1344, + "step": 2893 + }, + { + "epoch": 0.12813388973135717, + "grad_norm": 1.2789667844772339, + "learning_rate": 1.99518851294643e-05, + "loss": 0.9798, + "step": 2894 + }, + { + "epoch": 0.12817816543617103, + "grad_norm": 1.1507744789123535, + "learning_rate": 1.9951733589360796e-05, + "loss": 0.9005, + "step": 2895 + }, + { + "epoch": 0.1282224411409849, + "grad_norm": 1.6639243364334106, + "learning_rate": 1.9951581811568235e-05, + "loss": 1.0541, + "step": 2896 + }, + { + "epoch": 0.1282667168457988, + "grad_norm": 1.2489320039749146, + "learning_rate": 1.9951429796090247e-05, + "loss": 1.0039, + "step": 2897 + }, + { + "epoch": 0.12831099255061265, + "grad_norm": 1.7036762237548828, + "learning_rate": 1.995127754293046e-05, + "loss": 0.9396, + "step": 2898 + }, + { + "epoch": 0.12835526825542654, + "grad_norm": 1.2591204643249512, + "learning_rate": 1.9951125052092514e-05, + "loss": 0.9941, + "step": 2899 + }, + { + "epoch": 0.12839954396024042, + "grad_norm": 1.5529831647872925, + "learning_rate": 1.9950972323580046e-05, + "loss": 1.1135, + "step": 2900 + }, + { + "epoch": 0.1284438196650543, + "grad_norm": 1.527985692024231, + "learning_rate": 1.995081935739671e-05, + "loss": 1.1068, + "step": 2901 + }, + { + "epoch": 0.12848809536986816, + "grad_norm": 1.8458442687988281, + "learning_rate": 1.995066615354615e-05, + "loss": 1.0234, + "step": 2902 + }, + { + "epoch": 0.12853237107468204, + "grad_norm": 1.3275835514068604, + "learning_rate": 1.9950512712032038e-05, + "loss": 0.983, + "step": 2903 + }, + { + "epoch": 0.12857664677949593, + "grad_norm": 1.4233883619308472, + "learning_rate": 1.995035903285803e-05, + "loss": 0.9894, + "step": 2904 + }, + { + "epoch": 0.12862092248430979, + "grad_norm": 1.4265344142913818, + "learning_rate": 1.9950205116027794e-05, + "loss": 0.9598, + "step": 2905 + }, + { + "epoch": 0.12866519818912367, + "grad_norm": 1.673484206199646, + "learning_rate": 1.9950050961545012e-05, + "loss": 1.0751, + "step": 2906 + }, + { + "epoch": 0.12870947389393755, + "grad_norm": 1.3715072870254517, + "learning_rate": 1.9949896569413368e-05, + "loss": 1.0257, + "step": 2907 + }, + { + "epoch": 0.12875374959875144, + "grad_norm": 1.6252249479293823, + "learning_rate": 1.9949741939636544e-05, + "loss": 1.0583, + "step": 2908 + }, + { + "epoch": 0.1287980253035653, + "grad_norm": 1.6222517490386963, + "learning_rate": 1.9949587072218236e-05, + "loss": 0.9441, + "step": 2909 + }, + { + "epoch": 0.12884230100837918, + "grad_norm": 1.4455382823944092, + "learning_rate": 1.994943196716214e-05, + "loss": 0.9273, + "step": 2910 + }, + { + "epoch": 0.12888657671319306, + "grad_norm": 1.3336821794509888, + "learning_rate": 1.9949276624471964e-05, + "loss": 1.0577, + "step": 2911 + }, + { + "epoch": 0.12893085241800692, + "grad_norm": 1.2947022914886475, + "learning_rate": 1.9949121044151415e-05, + "loss": 0.9482, + "step": 2912 + }, + { + "epoch": 0.1289751281228208, + "grad_norm": 1.4897211790084839, + "learning_rate": 1.9948965226204213e-05, + "loss": 1.047, + "step": 2913 + }, + { + "epoch": 0.1290194038276347, + "grad_norm": 1.21903395652771, + "learning_rate": 1.9948809170634078e-05, + "loss": 0.9677, + "step": 2914 + }, + { + "epoch": 0.12906367953244854, + "grad_norm": 1.1429184675216675, + "learning_rate": 1.9948652877444733e-05, + "loss": 0.9422, + "step": 2915 + }, + { + "epoch": 0.12910795523726243, + "grad_norm": 1.6252493858337402, + "learning_rate": 1.9948496346639915e-05, + "loss": 1.0847, + "step": 2916 + }, + { + "epoch": 0.1291522309420763, + "grad_norm": 1.305435061454773, + "learning_rate": 1.9948339578223365e-05, + "loss": 1.0186, + "step": 2917 + }, + { + "epoch": 0.1291965066468902, + "grad_norm": 1.5561108589172363, + "learning_rate": 1.9948182572198824e-05, + "loss": 1.0128, + "step": 2918 + }, + { + "epoch": 0.12924078235170405, + "grad_norm": 1.3769999742507935, + "learning_rate": 1.994802532857004e-05, + "loss": 0.9919, + "step": 2919 + }, + { + "epoch": 0.12928505805651794, + "grad_norm": 1.322250485420227, + "learning_rate": 1.9947867847340775e-05, + "loss": 0.9709, + "step": 2920 + }, + { + "epoch": 0.12932933376133182, + "grad_norm": 1.253723382949829, + "learning_rate": 1.9947710128514783e-05, + "loss": 0.9381, + "step": 2921 + }, + { + "epoch": 0.12937360946614568, + "grad_norm": 1.1982624530792236, + "learning_rate": 1.9947552172095835e-05, + "loss": 0.9972, + "step": 2922 + }, + { + "epoch": 0.12941788517095956, + "grad_norm": 1.2509695291519165, + "learning_rate": 1.9947393978087703e-05, + "loss": 0.8982, + "step": 2923 + }, + { + "epoch": 0.12946216087577345, + "grad_norm": 1.183538794517517, + "learning_rate": 1.9947235546494165e-05, + "loss": 0.9991, + "step": 2924 + }, + { + "epoch": 0.12950643658058733, + "grad_norm": 1.3627655506134033, + "learning_rate": 1.994707687731901e-05, + "loss": 1.0182, + "step": 2925 + }, + { + "epoch": 0.1295507122854012, + "grad_norm": 1.238484501838684, + "learning_rate": 1.9946917970566015e-05, + "loss": 1.0219, + "step": 2926 + }, + { + "epoch": 0.12959498799021507, + "grad_norm": 1.7582944631576538, + "learning_rate": 1.994675882623899e-05, + "loss": 1.0604, + "step": 2927 + }, + { + "epoch": 0.12963926369502896, + "grad_norm": 1.210707426071167, + "learning_rate": 1.9946599444341724e-05, + "loss": 0.9813, + "step": 2928 + }, + { + "epoch": 0.1296835393998428, + "grad_norm": 1.4822832345962524, + "learning_rate": 1.994643982487803e-05, + "loss": 1.0762, + "step": 2929 + }, + { + "epoch": 0.1297278151046567, + "grad_norm": 1.4959886074066162, + "learning_rate": 1.9946279967851723e-05, + "loss": 1.1114, + "step": 2930 + }, + { + "epoch": 0.12977209080947058, + "grad_norm": 1.4698302745819092, + "learning_rate": 1.9946119873266615e-05, + "loss": 1.0171, + "step": 2931 + }, + { + "epoch": 0.12981636651428444, + "grad_norm": 1.4020278453826904, + "learning_rate": 1.9945959541126534e-05, + "loss": 1.045, + "step": 2932 + }, + { + "epoch": 0.12986064221909832, + "grad_norm": 1.3670045137405396, + "learning_rate": 1.9945798971435307e-05, + "loss": 0.8967, + "step": 2933 + }, + { + "epoch": 0.1299049179239122, + "grad_norm": 1.6978840827941895, + "learning_rate": 1.994563816419677e-05, + "loss": 1.1406, + "step": 2934 + }, + { + "epoch": 0.1299491936287261, + "grad_norm": 1.859772801399231, + "learning_rate": 1.994547711941476e-05, + "loss": 0.931, + "step": 2935 + }, + { + "epoch": 0.12999346933353995, + "grad_norm": 1.3060240745544434, + "learning_rate": 1.994531583709313e-05, + "loss": 1.0432, + "step": 2936 + }, + { + "epoch": 0.13003774503835383, + "grad_norm": 1.7071999311447144, + "learning_rate": 1.994515431723573e-05, + "loss": 1.0398, + "step": 2937 + }, + { + "epoch": 0.13008202074316771, + "grad_norm": 1.2632701396942139, + "learning_rate": 1.9944992559846414e-05, + "loss": 0.9674, + "step": 2938 + }, + { + "epoch": 0.13012629644798157, + "grad_norm": 1.3558447360992432, + "learning_rate": 1.994483056492905e-05, + "loss": 0.9436, + "step": 2939 + }, + { + "epoch": 0.13017057215279546, + "grad_norm": 1.294348955154419, + "learning_rate": 1.9944668332487504e-05, + "loss": 0.9712, + "step": 2940 + }, + { + "epoch": 0.13021484785760934, + "grad_norm": 2.6585514545440674, + "learning_rate": 1.9944505862525653e-05, + "loss": 1.0662, + "step": 2941 + }, + { + "epoch": 0.1302591235624232, + "grad_norm": 1.641510009765625, + "learning_rate": 1.9944343155047378e-05, + "loss": 1.0652, + "step": 2942 + }, + { + "epoch": 0.13030339926723708, + "grad_norm": 1.5026273727416992, + "learning_rate": 1.994418021005656e-05, + "loss": 1.0294, + "step": 2943 + }, + { + "epoch": 0.13034767497205096, + "grad_norm": 1.714943289756775, + "learning_rate": 1.99440170275571e-05, + "loss": 1.1186, + "step": 2944 + }, + { + "epoch": 0.13039195067686485, + "grad_norm": 1.340537190437317, + "learning_rate": 1.9943853607552886e-05, + "loss": 1.0113, + "step": 2945 + }, + { + "epoch": 0.1304362263816787, + "grad_norm": 1.2353461980819702, + "learning_rate": 1.9943689950047827e-05, + "loss": 0.8943, + "step": 2946 + }, + { + "epoch": 0.1304805020864926, + "grad_norm": 1.3341623544692993, + "learning_rate": 1.9943526055045826e-05, + "loss": 1.0305, + "step": 2947 + }, + { + "epoch": 0.13052477779130647, + "grad_norm": 1.5291633605957031, + "learning_rate": 1.9943361922550806e-05, + "loss": 0.9366, + "step": 2948 + }, + { + "epoch": 0.13056905349612033, + "grad_norm": 1.9613091945648193, + "learning_rate": 1.994319755256668e-05, + "loss": 1.0264, + "step": 2949 + }, + { + "epoch": 0.13061332920093421, + "grad_norm": 1.2606271505355835, + "learning_rate": 1.994303294509738e-05, + "loss": 0.9958, + "step": 2950 + }, + { + "epoch": 0.1306576049057481, + "grad_norm": 1.3696470260620117, + "learning_rate": 1.9942868100146834e-05, + "loss": 1.0311, + "step": 2951 + }, + { + "epoch": 0.13070188061056198, + "grad_norm": 1.4110320806503296, + "learning_rate": 1.9942703017718977e-05, + "loss": 0.9871, + "step": 2952 + }, + { + "epoch": 0.13074615631537584, + "grad_norm": 1.3714293241500854, + "learning_rate": 1.9942537697817752e-05, + "loss": 1.0044, + "step": 2953 + }, + { + "epoch": 0.13079043202018972, + "grad_norm": 1.345832347869873, + "learning_rate": 1.994237214044711e-05, + "loss": 1.0234, + "step": 2954 + }, + { + "epoch": 0.1308347077250036, + "grad_norm": 1.3997935056686401, + "learning_rate": 1.9942206345611008e-05, + "loss": 0.933, + "step": 2955 + }, + { + "epoch": 0.13087898342981746, + "grad_norm": 1.3323925733566284, + "learning_rate": 1.9942040313313404e-05, + "loss": 1.0232, + "step": 2956 + }, + { + "epoch": 0.13092325913463135, + "grad_norm": 1.6369227170944214, + "learning_rate": 1.9941874043558254e-05, + "loss": 1.0048, + "step": 2957 + }, + { + "epoch": 0.13096753483944523, + "grad_norm": 2.351524591445923, + "learning_rate": 1.9941707536349544e-05, + "loss": 1.1086, + "step": 2958 + }, + { + "epoch": 0.1310118105442591, + "grad_norm": 1.5710545778274536, + "learning_rate": 1.9941540791691245e-05, + "loss": 1.0126, + "step": 2959 + }, + { + "epoch": 0.13105608624907297, + "grad_norm": 1.2088077068328857, + "learning_rate": 1.9941373809587336e-05, + "loss": 0.9313, + "step": 2960 + }, + { + "epoch": 0.13110036195388686, + "grad_norm": 1.42751145362854, + "learning_rate": 1.994120659004181e-05, + "loss": 1.0608, + "step": 2961 + }, + { + "epoch": 0.13114463765870074, + "grad_norm": 1.7877501249313354, + "learning_rate": 1.994103913305866e-05, + "loss": 1.0931, + "step": 2962 + }, + { + "epoch": 0.1311889133635146, + "grad_norm": 1.2856117486953735, + "learning_rate": 1.994087143864188e-05, + "loss": 0.9745, + "step": 2963 + }, + { + "epoch": 0.13123318906832848, + "grad_norm": 1.4418809413909912, + "learning_rate": 1.9940703506795484e-05, + "loss": 1.0317, + "step": 2964 + }, + { + "epoch": 0.13127746477314237, + "grad_norm": 1.3620256185531616, + "learning_rate": 1.994053533752348e-05, + "loss": 1.0575, + "step": 2965 + }, + { + "epoch": 0.13132174047795622, + "grad_norm": 1.396597981452942, + "learning_rate": 1.994036693082988e-05, + "loss": 1.0276, + "step": 2966 + }, + { + "epoch": 0.1313660161827701, + "grad_norm": 1.8494411706924438, + "learning_rate": 1.994019828671871e-05, + "loss": 0.9997, + "step": 2967 + }, + { + "epoch": 0.131410291887584, + "grad_norm": 1.1719074249267578, + "learning_rate": 1.9940029405193997e-05, + "loss": 0.9889, + "step": 2968 + }, + { + "epoch": 0.13145456759239788, + "grad_norm": 1.3006267547607422, + "learning_rate": 1.9939860286259774e-05, + "loss": 1.0174, + "step": 2969 + }, + { + "epoch": 0.13149884329721173, + "grad_norm": 1.6886475086212158, + "learning_rate": 1.9939690929920082e-05, + "loss": 1.1031, + "step": 2970 + }, + { + "epoch": 0.13154311900202562, + "grad_norm": 1.1701196432113647, + "learning_rate": 1.993952133617897e-05, + "loss": 0.9972, + "step": 2971 + }, + { + "epoch": 0.1315873947068395, + "grad_norm": 1.5265414714813232, + "learning_rate": 1.993935150504048e-05, + "loss": 1.0366, + "step": 2972 + }, + { + "epoch": 0.13163167041165336, + "grad_norm": 1.5602284669876099, + "learning_rate": 1.9939181436508673e-05, + "loss": 1.0407, + "step": 2973 + }, + { + "epoch": 0.13167594611646724, + "grad_norm": 1.3245246410369873, + "learning_rate": 1.993901113058761e-05, + "loss": 0.9674, + "step": 2974 + }, + { + "epoch": 0.13172022182128112, + "grad_norm": 1.2627496719360352, + "learning_rate": 1.993884058728136e-05, + "loss": 1.0032, + "step": 2975 + }, + { + "epoch": 0.13176449752609498, + "grad_norm": 1.8369656801223755, + "learning_rate": 1.9938669806593993e-05, + "loss": 0.944, + "step": 2976 + }, + { + "epoch": 0.13180877323090887, + "grad_norm": 1.4868639707565308, + "learning_rate": 1.993849878852959e-05, + "loss": 0.9464, + "step": 2977 + }, + { + "epoch": 0.13185304893572275, + "grad_norm": 1.2824052572250366, + "learning_rate": 1.993832753309224e-05, + "loss": 0.958, + "step": 2978 + }, + { + "epoch": 0.13189732464053663, + "grad_norm": 1.392106056213379, + "learning_rate": 1.9938156040286027e-05, + "loss": 1.0016, + "step": 2979 + }, + { + "epoch": 0.1319416003453505, + "grad_norm": 1.5755867958068848, + "learning_rate": 1.9937984310115043e-05, + "loss": 1.0402, + "step": 2980 + }, + { + "epoch": 0.13198587605016437, + "grad_norm": 1.4552581310272217, + "learning_rate": 1.9937812342583404e-05, + "loss": 1.0271, + "step": 2981 + }, + { + "epoch": 0.13203015175497826, + "grad_norm": 1.2225202322006226, + "learning_rate": 1.9937640137695204e-05, + "loss": 0.9195, + "step": 2982 + }, + { + "epoch": 0.13207442745979212, + "grad_norm": 1.3228356838226318, + "learning_rate": 1.993746769545456e-05, + "loss": 0.9556, + "step": 2983 + }, + { + "epoch": 0.132118703164606, + "grad_norm": 1.5609976053237915, + "learning_rate": 1.9937295015865594e-05, + "loss": 0.964, + "step": 2984 + }, + { + "epoch": 0.13216297886941988, + "grad_norm": 1.3443586826324463, + "learning_rate": 1.9937122098932428e-05, + "loss": 0.8975, + "step": 2985 + }, + { + "epoch": 0.13220725457423374, + "grad_norm": 1.3625147342681885, + "learning_rate": 1.993694894465919e-05, + "loss": 1.0162, + "step": 2986 + }, + { + "epoch": 0.13225153027904762, + "grad_norm": 1.5467404127120972, + "learning_rate": 1.9936775553050017e-05, + "loss": 1.0251, + "step": 2987 + }, + { + "epoch": 0.1322958059838615, + "grad_norm": 1.6365967988967896, + "learning_rate": 1.9936601924109053e-05, + "loss": 1.0769, + "step": 2988 + }, + { + "epoch": 0.1323400816886754, + "grad_norm": 1.3523818254470825, + "learning_rate": 1.9936428057840442e-05, + "loss": 1.0383, + "step": 2989 + }, + { + "epoch": 0.13238435739348925, + "grad_norm": 1.2609245777130127, + "learning_rate": 1.9936253954248337e-05, + "loss": 0.9066, + "step": 2990 + }, + { + "epoch": 0.13242863309830313, + "grad_norm": 1.366613507270813, + "learning_rate": 1.9936079613336896e-05, + "loss": 0.9924, + "step": 2991 + }, + { + "epoch": 0.13247290880311702, + "grad_norm": 1.4937456846237183, + "learning_rate": 1.9935905035110285e-05, + "loss": 0.9681, + "step": 2992 + }, + { + "epoch": 0.13251718450793087, + "grad_norm": 1.4103736877441406, + "learning_rate": 1.9935730219572673e-05, + "loss": 0.9897, + "step": 2993 + }, + { + "epoch": 0.13256146021274476, + "grad_norm": 1.8491235971450806, + "learning_rate": 1.9935555166728234e-05, + "loss": 1.0948, + "step": 2994 + }, + { + "epoch": 0.13260573591755864, + "grad_norm": 1.6335605382919312, + "learning_rate": 1.9935379876581148e-05, + "loss": 1.0972, + "step": 2995 + }, + { + "epoch": 0.13265001162237253, + "grad_norm": 1.4388456344604492, + "learning_rate": 1.9935204349135606e-05, + "loss": 1.0179, + "step": 2996 + }, + { + "epoch": 0.13269428732718638, + "grad_norm": 1.2781327962875366, + "learning_rate": 1.9935028584395794e-05, + "loss": 0.9403, + "step": 2997 + }, + { + "epoch": 0.13273856303200027, + "grad_norm": 1.276454210281372, + "learning_rate": 1.993485258236592e-05, + "loss": 0.9654, + "step": 2998 + }, + { + "epoch": 0.13278283873681415, + "grad_norm": 1.2718263864517212, + "learning_rate": 1.9934676343050175e-05, + "loss": 0.9592, + "step": 2999 + }, + { + "epoch": 0.132827114441628, + "grad_norm": 1.443642497062683, + "learning_rate": 1.9934499866452778e-05, + "loss": 1.0983, + "step": 3000 + }, + { + "epoch": 0.1328713901464419, + "grad_norm": 1.3674753904342651, + "learning_rate": 1.9934323152577938e-05, + "loss": 1.0308, + "step": 3001 + }, + { + "epoch": 0.13291566585125578, + "grad_norm": 1.2169290781021118, + "learning_rate": 1.993414620142988e-05, + "loss": 1.0006, + "step": 3002 + }, + { + "epoch": 0.13295994155606963, + "grad_norm": 1.4993537664413452, + "learning_rate": 1.993396901301283e-05, + "loss": 1.0059, + "step": 3003 + }, + { + "epoch": 0.13300421726088352, + "grad_norm": 1.3174715042114258, + "learning_rate": 1.9933791587331013e-05, + "loss": 0.9381, + "step": 3004 + }, + { + "epoch": 0.1330484929656974, + "grad_norm": 1.2817127704620361, + "learning_rate": 1.993361392438868e-05, + "loss": 0.947, + "step": 3005 + }, + { + "epoch": 0.13309276867051129, + "grad_norm": 1.5050852298736572, + "learning_rate": 1.993343602419006e-05, + "loss": 0.9306, + "step": 3006 + }, + { + "epoch": 0.13313704437532514, + "grad_norm": 1.3827675580978394, + "learning_rate": 1.9933257886739416e-05, + "loss": 1.0105, + "step": 3007 + }, + { + "epoch": 0.13318132008013903, + "grad_norm": 1.4773184061050415, + "learning_rate": 1.9933079512040988e-05, + "loss": 0.9909, + "step": 3008 + }, + { + "epoch": 0.1332255957849529, + "grad_norm": 1.3753246068954468, + "learning_rate": 1.993290090009905e-05, + "loss": 0.9555, + "step": 3009 + }, + { + "epoch": 0.13326987148976677, + "grad_norm": 1.242503046989441, + "learning_rate": 1.9932722050917858e-05, + "loss": 0.9625, + "step": 3010 + }, + { + "epoch": 0.13331414719458065, + "grad_norm": 1.8431094884872437, + "learning_rate": 1.993254296450169e-05, + "loss": 1.0671, + "step": 3011 + }, + { + "epoch": 0.13335842289939454, + "grad_norm": 1.4263365268707275, + "learning_rate": 1.9932363640854816e-05, + "loss": 0.9589, + "step": 3012 + }, + { + "epoch": 0.13340269860420842, + "grad_norm": 1.4127259254455566, + "learning_rate": 1.9932184079981525e-05, + "loss": 0.9818, + "step": 3013 + }, + { + "epoch": 0.13344697430902228, + "grad_norm": 1.2079987525939941, + "learning_rate": 1.993200428188611e-05, + "loss": 0.9783, + "step": 3014 + }, + { + "epoch": 0.13349125001383616, + "grad_norm": 1.2273774147033691, + "learning_rate": 1.993182424657285e-05, + "loss": 0.9731, + "step": 3015 + }, + { + "epoch": 0.13353552571865004, + "grad_norm": 1.3152109384536743, + "learning_rate": 1.993164397404606e-05, + "loss": 1.0708, + "step": 3016 + }, + { + "epoch": 0.1335798014234639, + "grad_norm": 1.3993093967437744, + "learning_rate": 1.9931463464310044e-05, + "loss": 0.9549, + "step": 3017 + }, + { + "epoch": 0.13362407712827779, + "grad_norm": 1.4479249715805054, + "learning_rate": 1.9931282717369104e-05, + "loss": 0.9926, + "step": 3018 + }, + { + "epoch": 0.13366835283309167, + "grad_norm": 1.3045618534088135, + "learning_rate": 1.9931101733227565e-05, + "loss": 0.927, + "step": 3019 + }, + { + "epoch": 0.13371262853790553, + "grad_norm": 1.275869607925415, + "learning_rate": 1.9930920511889746e-05, + "loss": 0.9808, + "step": 3020 + }, + { + "epoch": 0.1337569042427194, + "grad_norm": 1.523860216140747, + "learning_rate": 1.9930739053359978e-05, + "loss": 0.9494, + "step": 3021 + }, + { + "epoch": 0.1338011799475333, + "grad_norm": 1.4201924800872803, + "learning_rate": 1.9930557357642594e-05, + "loss": 1.0212, + "step": 3022 + }, + { + "epoch": 0.13384545565234718, + "grad_norm": 1.4190548658370972, + "learning_rate": 1.993037542474193e-05, + "loss": 0.9761, + "step": 3023 + }, + { + "epoch": 0.13388973135716103, + "grad_norm": 1.3958848714828491, + "learning_rate": 1.9930193254662338e-05, + "loss": 1.0734, + "step": 3024 + }, + { + "epoch": 0.13393400706197492, + "grad_norm": 1.3107304573059082, + "learning_rate": 1.9930010847408164e-05, + "loss": 0.9754, + "step": 3025 + }, + { + "epoch": 0.1339782827667888, + "grad_norm": 1.2930399179458618, + "learning_rate": 1.992982820298377e-05, + "loss": 1.0429, + "step": 3026 + }, + { + "epoch": 0.13402255847160266, + "grad_norm": 1.4865872859954834, + "learning_rate": 1.9929645321393505e-05, + "loss": 1.0311, + "step": 3027 + }, + { + "epoch": 0.13406683417641654, + "grad_norm": 1.4692133665084839, + "learning_rate": 1.9929462202641756e-05, + "loss": 0.9703, + "step": 3028 + }, + { + "epoch": 0.13411110988123043, + "grad_norm": 1.168082356452942, + "learning_rate": 1.9929278846732883e-05, + "loss": 0.8787, + "step": 3029 + }, + { + "epoch": 0.13415538558604428, + "grad_norm": 1.4641532897949219, + "learning_rate": 1.9929095253671272e-05, + "loss": 0.9292, + "step": 3030 + }, + { + "epoch": 0.13419966129085817, + "grad_norm": 1.3144859075546265, + "learning_rate": 1.9928911423461303e-05, + "loss": 0.9816, + "step": 3031 + }, + { + "epoch": 0.13424393699567205, + "grad_norm": 1.2361633777618408, + "learning_rate": 1.992872735610737e-05, + "loss": 0.9893, + "step": 3032 + }, + { + "epoch": 0.13428821270048594, + "grad_norm": 1.6262050867080688, + "learning_rate": 1.992854305161387e-05, + "loss": 1.0488, + "step": 3033 + }, + { + "epoch": 0.1343324884052998, + "grad_norm": 1.5386159420013428, + "learning_rate": 1.99283585099852e-05, + "loss": 1.0409, + "step": 3034 + }, + { + "epoch": 0.13437676411011368, + "grad_norm": 1.485482096672058, + "learning_rate": 1.9928173731225772e-05, + "loss": 1.0694, + "step": 3035 + }, + { + "epoch": 0.13442103981492756, + "grad_norm": 1.4177336692810059, + "learning_rate": 1.992798871534e-05, + "loss": 0.9253, + "step": 3036 + }, + { + "epoch": 0.13446531551974142, + "grad_norm": 1.337188959121704, + "learning_rate": 1.99278034623323e-05, + "loss": 0.9548, + "step": 3037 + }, + { + "epoch": 0.1345095912245553, + "grad_norm": 1.386951208114624, + "learning_rate": 1.99276179722071e-05, + "loss": 0.9865, + "step": 3038 + }, + { + "epoch": 0.1345538669293692, + "grad_norm": 1.4338268041610718, + "learning_rate": 1.9927432244968826e-05, + "loss": 1.1069, + "step": 3039 + }, + { + "epoch": 0.13459814263418307, + "grad_norm": 1.4789485931396484, + "learning_rate": 1.992724628062191e-05, + "loss": 0.9388, + "step": 3040 + }, + { + "epoch": 0.13464241833899693, + "grad_norm": 1.2465115785598755, + "learning_rate": 1.992706007917081e-05, + "loss": 0.9642, + "step": 3041 + }, + { + "epoch": 0.1346866940438108, + "grad_norm": 1.2892221212387085, + "learning_rate": 1.992687364061996e-05, + "loss": 0.9804, + "step": 3042 + }, + { + "epoch": 0.1347309697486247, + "grad_norm": 1.6762083768844604, + "learning_rate": 1.9926686964973813e-05, + "loss": 1.0481, + "step": 3043 + }, + { + "epoch": 0.13477524545343855, + "grad_norm": 1.2768852710723877, + "learning_rate": 1.9926500052236833e-05, + "loss": 1.0227, + "step": 3044 + }, + { + "epoch": 0.13481952115825244, + "grad_norm": 1.31876540184021, + "learning_rate": 1.9926312902413478e-05, + "loss": 0.9649, + "step": 3045 + }, + { + "epoch": 0.13486379686306632, + "grad_norm": 1.3814290761947632, + "learning_rate": 1.9926125515508223e-05, + "loss": 0.9491, + "step": 3046 + }, + { + "epoch": 0.13490807256788018, + "grad_norm": 1.7055163383483887, + "learning_rate": 1.9925937891525545e-05, + "loss": 1.0731, + "step": 3047 + }, + { + "epoch": 0.13495234827269406, + "grad_norm": 1.9172887802124023, + "learning_rate": 1.992575003046992e-05, + "loss": 1.1256, + "step": 3048 + }, + { + "epoch": 0.13499662397750795, + "grad_norm": 1.3312822580337524, + "learning_rate": 1.992556193234584e-05, + "loss": 1.0062, + "step": 3049 + }, + { + "epoch": 0.13504089968232183, + "grad_norm": 1.3926554918289185, + "learning_rate": 1.992537359715779e-05, + "loss": 0.9614, + "step": 3050 + }, + { + "epoch": 0.1350851753871357, + "grad_norm": 1.6384239196777344, + "learning_rate": 1.992518502491028e-05, + "loss": 1.0152, + "step": 3051 + }, + { + "epoch": 0.13512945109194957, + "grad_norm": 1.6039148569107056, + "learning_rate": 1.9924996215607804e-05, + "loss": 1.0549, + "step": 3052 + }, + { + "epoch": 0.13517372679676345, + "grad_norm": 1.3055596351623535, + "learning_rate": 1.992480716925487e-05, + "loss": 0.9311, + "step": 3053 + }, + { + "epoch": 0.1352180025015773, + "grad_norm": 1.237084150314331, + "learning_rate": 1.9924617885856007e-05, + "loss": 0.9099, + "step": 3054 + }, + { + "epoch": 0.1352622782063912, + "grad_norm": 1.8816684484481812, + "learning_rate": 1.9924428365415717e-05, + "loss": 0.9935, + "step": 3055 + }, + { + "epoch": 0.13530655391120508, + "grad_norm": 1.45123291015625, + "learning_rate": 1.9924238607938544e-05, + "loss": 0.9626, + "step": 3056 + }, + { + "epoch": 0.13535082961601894, + "grad_norm": 1.4692131280899048, + "learning_rate": 1.9924048613429008e-05, + "loss": 1.003, + "step": 3057 + }, + { + "epoch": 0.13539510532083282, + "grad_norm": 1.4461493492126465, + "learning_rate": 1.9923858381891654e-05, + "loss": 1.0196, + "step": 3058 + }, + { + "epoch": 0.1354393810256467, + "grad_norm": 1.2565113306045532, + "learning_rate": 1.992366791333102e-05, + "loss": 0.9749, + "step": 3059 + }, + { + "epoch": 0.1354836567304606, + "grad_norm": 1.3458342552185059, + "learning_rate": 1.992347720775166e-05, + "loss": 0.9077, + "step": 3060 + }, + { + "epoch": 0.13552793243527445, + "grad_norm": 1.2821838855743408, + "learning_rate": 1.9923286265158126e-05, + "loss": 0.9422, + "step": 3061 + }, + { + "epoch": 0.13557220814008833, + "grad_norm": 1.3784856796264648, + "learning_rate": 1.9923095085554978e-05, + "loss": 1.094, + "step": 3062 + }, + { + "epoch": 0.1356164838449022, + "grad_norm": 1.4133551120758057, + "learning_rate": 1.9922903668946788e-05, + "loss": 1.0075, + "step": 3063 + }, + { + "epoch": 0.13566075954971607, + "grad_norm": 1.933391809463501, + "learning_rate": 1.992271201533812e-05, + "loss": 1.1013, + "step": 3064 + }, + { + "epoch": 0.13570503525452995, + "grad_norm": 1.8874796628952026, + "learning_rate": 1.9922520124733556e-05, + "loss": 1.0657, + "step": 3065 + }, + { + "epoch": 0.13574931095934384, + "grad_norm": 1.6544189453125, + "learning_rate": 1.992232799713768e-05, + "loss": 1.0737, + "step": 3066 + }, + { + "epoch": 0.13579358666415772, + "grad_norm": 1.3635128736495972, + "learning_rate": 1.9922135632555073e-05, + "loss": 0.9446, + "step": 3067 + }, + { + "epoch": 0.13583786236897158, + "grad_norm": 1.33771812915802, + "learning_rate": 1.992194303099034e-05, + "loss": 1.0647, + "step": 3068 + }, + { + "epoch": 0.13588213807378546, + "grad_norm": 1.4104636907577515, + "learning_rate": 1.9921750192448077e-05, + "loss": 1.0107, + "step": 3069 + }, + { + "epoch": 0.13592641377859935, + "grad_norm": 1.205270767211914, + "learning_rate": 1.9921557116932886e-05, + "loss": 0.9265, + "step": 3070 + }, + { + "epoch": 0.1359706894834132, + "grad_norm": 1.218802809715271, + "learning_rate": 1.9921363804449383e-05, + "loss": 0.8653, + "step": 3071 + }, + { + "epoch": 0.1360149651882271, + "grad_norm": 1.257977843284607, + "learning_rate": 1.9921170255002185e-05, + "loss": 1.1014, + "step": 3072 + }, + { + "epoch": 0.13605924089304097, + "grad_norm": 1.239561676979065, + "learning_rate": 1.992097646859591e-05, + "loss": 0.9441, + "step": 3073 + }, + { + "epoch": 0.13610351659785483, + "grad_norm": 1.4955929517745972, + "learning_rate": 1.9920782445235196e-05, + "loss": 1.1158, + "step": 3074 + }, + { + "epoch": 0.1361477923026687, + "grad_norm": 1.458895206451416, + "learning_rate": 1.9920588184924664e-05, + "loss": 1.0442, + "step": 3075 + }, + { + "epoch": 0.1361920680074826, + "grad_norm": 1.5233041048049927, + "learning_rate": 1.9920393687668968e-05, + "loss": 1.0949, + "step": 3076 + }, + { + "epoch": 0.13623634371229648, + "grad_norm": 1.6149975061416626, + "learning_rate": 1.992019895347274e-05, + "loss": 0.9525, + "step": 3077 + }, + { + "epoch": 0.13628061941711034, + "grad_norm": 1.3413289785385132, + "learning_rate": 1.992000398234064e-05, + "loss": 1.0205, + "step": 3078 + }, + { + "epoch": 0.13632489512192422, + "grad_norm": 2.4965736865997314, + "learning_rate": 1.9919808774277324e-05, + "loss": 1.0422, + "step": 3079 + }, + { + "epoch": 0.1363691708267381, + "grad_norm": 1.4263696670532227, + "learning_rate": 1.991961332928745e-05, + "loss": 0.963, + "step": 3080 + }, + { + "epoch": 0.13641344653155196, + "grad_norm": 1.5807324647903442, + "learning_rate": 1.9919417647375692e-05, + "loss": 1.0467, + "step": 3081 + }, + { + "epoch": 0.13645772223636585, + "grad_norm": 1.581847071647644, + "learning_rate": 1.9919221728546717e-05, + "loss": 1.1374, + "step": 3082 + }, + { + "epoch": 0.13650199794117973, + "grad_norm": 1.3313593864440918, + "learning_rate": 1.9919025572805206e-05, + "loss": 1.0407, + "step": 3083 + }, + { + "epoch": 0.13654627364599362, + "grad_norm": 1.7568390369415283, + "learning_rate": 1.991882918015585e-05, + "loss": 1.1148, + "step": 3084 + }, + { + "epoch": 0.13659054935080747, + "grad_norm": 1.424733281135559, + "learning_rate": 1.991863255060333e-05, + "loss": 1.072, + "step": 3085 + }, + { + "epoch": 0.13663482505562136, + "grad_norm": 1.2329713106155396, + "learning_rate": 1.991843568415235e-05, + "loss": 0.9444, + "step": 3086 + }, + { + "epoch": 0.13667910076043524, + "grad_norm": 1.4076635837554932, + "learning_rate": 1.991823858080761e-05, + "loss": 1.0385, + "step": 3087 + }, + { + "epoch": 0.1367233764652491, + "grad_norm": 1.225435733795166, + "learning_rate": 1.9918041240573817e-05, + "loss": 0.9748, + "step": 3088 + }, + { + "epoch": 0.13676765217006298, + "grad_norm": 1.3147941827774048, + "learning_rate": 1.9917843663455688e-05, + "loss": 0.9668, + "step": 3089 + }, + { + "epoch": 0.13681192787487687, + "grad_norm": 1.8481767177581787, + "learning_rate": 1.9917645849457932e-05, + "loss": 1.0737, + "step": 3090 + }, + { + "epoch": 0.13685620357969072, + "grad_norm": 1.4347950220108032, + "learning_rate": 1.9917447798585284e-05, + "loss": 1.0594, + "step": 3091 + }, + { + "epoch": 0.1369004792845046, + "grad_norm": 1.2929843664169312, + "learning_rate": 1.991724951084247e-05, + "loss": 0.9835, + "step": 3092 + }, + { + "epoch": 0.1369447549893185, + "grad_norm": 1.219274878501892, + "learning_rate": 1.9917050986234226e-05, + "loss": 0.9596, + "step": 3093 + }, + { + "epoch": 0.13698903069413237, + "grad_norm": 1.243554711341858, + "learning_rate": 1.9916852224765293e-05, + "loss": 0.9888, + "step": 3094 + }, + { + "epoch": 0.13703330639894623, + "grad_norm": 1.2487397193908691, + "learning_rate": 1.991665322644042e-05, + "loss": 0.9839, + "step": 3095 + }, + { + "epoch": 0.13707758210376011, + "grad_norm": 1.525966763496399, + "learning_rate": 1.9916453991264362e-05, + "loss": 0.9705, + "step": 3096 + }, + { + "epoch": 0.137121857808574, + "grad_norm": 1.2300667762756348, + "learning_rate": 1.991625451924187e-05, + "loss": 0.9518, + "step": 3097 + }, + { + "epoch": 0.13716613351338786, + "grad_norm": 1.6984783411026, + "learning_rate": 1.9916054810377713e-05, + "loss": 1.1054, + "step": 3098 + }, + { + "epoch": 0.13721040921820174, + "grad_norm": 1.153236985206604, + "learning_rate": 1.9915854864676665e-05, + "loss": 0.8286, + "step": 3099 + }, + { + "epoch": 0.13725468492301562, + "grad_norm": 1.6467535495758057, + "learning_rate": 1.9915654682143494e-05, + "loss": 0.9223, + "step": 3100 + }, + { + "epoch": 0.13729896062782948, + "grad_norm": 2.7959671020507812, + "learning_rate": 1.9915454262782983e-05, + "loss": 1.0866, + "step": 3101 + }, + { + "epoch": 0.13734323633264336, + "grad_norm": 1.2360347509384155, + "learning_rate": 1.9915253606599928e-05, + "loss": 0.9591, + "step": 3102 + }, + { + "epoch": 0.13738751203745725, + "grad_norm": 1.396435260772705, + "learning_rate": 1.991505271359911e-05, + "loss": 1.0, + "step": 3103 + }, + { + "epoch": 0.13743178774227113, + "grad_norm": 1.3046982288360596, + "learning_rate": 1.9914851583785327e-05, + "loss": 0.9681, + "step": 3104 + }, + { + "epoch": 0.137476063447085, + "grad_norm": 1.5453698635101318, + "learning_rate": 1.991465021716339e-05, + "loss": 1.0168, + "step": 3105 + }, + { + "epoch": 0.13752033915189887, + "grad_norm": 1.361715316772461, + "learning_rate": 1.9914448613738107e-05, + "loss": 0.9475, + "step": 3106 + }, + { + "epoch": 0.13756461485671276, + "grad_norm": 1.503841519355774, + "learning_rate": 1.9914246773514288e-05, + "loss": 1.0484, + "step": 3107 + }, + { + "epoch": 0.13760889056152661, + "grad_norm": 1.537523865699768, + "learning_rate": 1.991404469649676e-05, + "loss": 1.0944, + "step": 3108 + }, + { + "epoch": 0.1376531662663405, + "grad_norm": 1.3347357511520386, + "learning_rate": 1.9913842382690343e-05, + "loss": 0.9373, + "step": 3109 + }, + { + "epoch": 0.13769744197115438, + "grad_norm": 1.230730414390564, + "learning_rate": 1.9913639832099877e-05, + "loss": 1.0144, + "step": 3110 + }, + { + "epoch": 0.13774171767596827, + "grad_norm": 1.3518750667572021, + "learning_rate": 1.9913437044730197e-05, + "loss": 0.984, + "step": 3111 + }, + { + "epoch": 0.13778599338078212, + "grad_norm": 1.3516604900360107, + "learning_rate": 1.991323402058614e-05, + "loss": 0.9314, + "step": 3112 + }, + { + "epoch": 0.137830269085596, + "grad_norm": 1.421565055847168, + "learning_rate": 1.9913030759672564e-05, + "loss": 0.9598, + "step": 3113 + }, + { + "epoch": 0.1378745447904099, + "grad_norm": 1.3767095804214478, + "learning_rate": 1.9912827261994318e-05, + "loss": 0.9538, + "step": 3114 + }, + { + "epoch": 0.13791882049522375, + "grad_norm": 1.457681655883789, + "learning_rate": 1.9912623527556263e-05, + "loss": 0.9107, + "step": 3115 + }, + { + "epoch": 0.13796309620003763, + "grad_norm": 1.228398084640503, + "learning_rate": 1.9912419556363266e-05, + "loss": 0.9217, + "step": 3116 + }, + { + "epoch": 0.13800737190485152, + "grad_norm": 1.2106928825378418, + "learning_rate": 1.99122153484202e-05, + "loss": 0.9409, + "step": 3117 + }, + { + "epoch": 0.13805164760966537, + "grad_norm": 1.7851184606552124, + "learning_rate": 1.9912010903731943e-05, + "loss": 1.0573, + "step": 3118 + }, + { + "epoch": 0.13809592331447926, + "grad_norm": 2.1104061603546143, + "learning_rate": 1.9911806222303375e-05, + "loss": 1.0689, + "step": 3119 + }, + { + "epoch": 0.13814019901929314, + "grad_norm": 1.4039077758789062, + "learning_rate": 1.9911601304139387e-05, + "loss": 0.9793, + "step": 3120 + }, + { + "epoch": 0.13818447472410703, + "grad_norm": 1.299916386604309, + "learning_rate": 1.991139614924487e-05, + "loss": 1.0934, + "step": 3121 + }, + { + "epoch": 0.13822875042892088, + "grad_norm": 1.6770602464675903, + "learning_rate": 1.9911190757624726e-05, + "loss": 1.0001, + "step": 3122 + }, + { + "epoch": 0.13827302613373477, + "grad_norm": 1.4279741048812866, + "learning_rate": 1.991098512928386e-05, + "loss": 0.955, + "step": 3123 + }, + { + "epoch": 0.13831730183854865, + "grad_norm": 3.049245595932007, + "learning_rate": 1.991077926422719e-05, + "loss": 1.0865, + "step": 3124 + }, + { + "epoch": 0.1383615775433625, + "grad_norm": 1.3093351125717163, + "learning_rate": 1.991057316245962e-05, + "loss": 0.9846, + "step": 3125 + }, + { + "epoch": 0.1384058532481764, + "grad_norm": 1.2100462913513184, + "learning_rate": 1.9910366823986084e-05, + "loss": 0.902, + "step": 3126 + }, + { + "epoch": 0.13845012895299028, + "grad_norm": 1.633649230003357, + "learning_rate": 1.9910160248811502e-05, + "loss": 1.039, + "step": 3127 + }, + { + "epoch": 0.13849440465780416, + "grad_norm": 1.3179569244384766, + "learning_rate": 1.9909953436940816e-05, + "loss": 0.9475, + "step": 3128 + }, + { + "epoch": 0.13853868036261802, + "grad_norm": 1.2773412466049194, + "learning_rate": 1.9909746388378958e-05, + "loss": 1.013, + "step": 3129 + }, + { + "epoch": 0.1385829560674319, + "grad_norm": 1.2966227531433105, + "learning_rate": 1.9909539103130876e-05, + "loss": 1.0039, + "step": 3130 + }, + { + "epoch": 0.13862723177224578, + "grad_norm": 3.4740493297576904, + "learning_rate": 1.9909331581201524e-05, + "loss": 1.0147, + "step": 3131 + }, + { + "epoch": 0.13867150747705964, + "grad_norm": 1.406997799873352, + "learning_rate": 1.9909123822595852e-05, + "loss": 1.0403, + "step": 3132 + }, + { + "epoch": 0.13871578318187353, + "grad_norm": 1.2802144289016724, + "learning_rate": 1.9908915827318827e-05, + "loss": 0.9927, + "step": 3133 + }, + { + "epoch": 0.1387600588866874, + "grad_norm": 1.301011562347412, + "learning_rate": 1.9908707595375417e-05, + "loss": 0.9834, + "step": 3134 + }, + { + "epoch": 0.13880433459150127, + "grad_norm": 1.8697792291641235, + "learning_rate": 1.990849912677059e-05, + "loss": 1.0818, + "step": 3135 + }, + { + "epoch": 0.13884861029631515, + "grad_norm": 1.462844967842102, + "learning_rate": 1.9908290421509337e-05, + "loss": 0.9697, + "step": 3136 + }, + { + "epoch": 0.13889288600112903, + "grad_norm": 1.298803448677063, + "learning_rate": 1.990808147959663e-05, + "loss": 1.0075, + "step": 3137 + }, + { + "epoch": 0.13893716170594292, + "grad_norm": 1.8628592491149902, + "learning_rate": 1.9907872301037464e-05, + "loss": 1.1057, + "step": 3138 + }, + { + "epoch": 0.13898143741075678, + "grad_norm": 1.235918641090393, + "learning_rate": 1.9907662885836836e-05, + "loss": 0.9277, + "step": 3139 + }, + { + "epoch": 0.13902571311557066, + "grad_norm": 1.2527436017990112, + "learning_rate": 1.9907453233999747e-05, + "loss": 0.911, + "step": 3140 + }, + { + "epoch": 0.13906998882038454, + "grad_norm": 3.2296738624572754, + "learning_rate": 1.9907243345531206e-05, + "loss": 1.0343, + "step": 3141 + }, + { + "epoch": 0.1391142645251984, + "grad_norm": 1.3928693532943726, + "learning_rate": 1.9907033220436223e-05, + "loss": 0.9349, + "step": 3142 + }, + { + "epoch": 0.13915854023001228, + "grad_norm": 1.674498200416565, + "learning_rate": 1.9906822858719817e-05, + "loss": 1.1099, + "step": 3143 + }, + { + "epoch": 0.13920281593482617, + "grad_norm": 1.598676085472107, + "learning_rate": 1.9906612260387015e-05, + "loss": 0.9949, + "step": 3144 + }, + { + "epoch": 0.13924709163964002, + "grad_norm": 1.540145993232727, + "learning_rate": 1.9906401425442846e-05, + "loss": 0.9948, + "step": 3145 + }, + { + "epoch": 0.1392913673444539, + "grad_norm": 1.4481862783432007, + "learning_rate": 1.9906190353892343e-05, + "loss": 0.9866, + "step": 3146 + }, + { + "epoch": 0.1393356430492678, + "grad_norm": 1.6111299991607666, + "learning_rate": 1.990597904574055e-05, + "loss": 1.0083, + "step": 3147 + }, + { + "epoch": 0.13937991875408168, + "grad_norm": 1.5066123008728027, + "learning_rate": 1.9905767500992515e-05, + "loss": 1.0349, + "step": 3148 + }, + { + "epoch": 0.13942419445889553, + "grad_norm": 1.3807085752487183, + "learning_rate": 1.9905555719653288e-05, + "loss": 1.0011, + "step": 3149 + }, + { + "epoch": 0.13946847016370942, + "grad_norm": 1.7703235149383545, + "learning_rate": 1.9905343701727926e-05, + "loss": 1.0251, + "step": 3150 + }, + { + "epoch": 0.1395127458685233, + "grad_norm": 1.413686990737915, + "learning_rate": 1.99051314472215e-05, + "loss": 1.0245, + "step": 3151 + }, + { + "epoch": 0.13955702157333716, + "grad_norm": 1.2212153673171997, + "learning_rate": 1.9904918956139067e-05, + "loss": 0.9684, + "step": 3152 + }, + { + "epoch": 0.13960129727815104, + "grad_norm": 2.7510597705841064, + "learning_rate": 1.9904706228485714e-05, + "loss": 1.0771, + "step": 3153 + }, + { + "epoch": 0.13964557298296493, + "grad_norm": 1.3370527029037476, + "learning_rate": 1.9904493264266517e-05, + "loss": 0.9916, + "step": 3154 + }, + { + "epoch": 0.1396898486877788, + "grad_norm": 1.294151782989502, + "learning_rate": 1.9904280063486563e-05, + "loss": 0.8954, + "step": 3155 + }, + { + "epoch": 0.13973412439259267, + "grad_norm": 1.2840570211410522, + "learning_rate": 1.9904066626150944e-05, + "loss": 0.9549, + "step": 3156 + }, + { + "epoch": 0.13977840009740655, + "grad_norm": 1.4225800037384033, + "learning_rate": 1.9903852952264756e-05, + "loss": 1.0329, + "step": 3157 + }, + { + "epoch": 0.13982267580222044, + "grad_norm": 1.5674482583999634, + "learning_rate": 1.9903639041833108e-05, + "loss": 1.0051, + "step": 3158 + }, + { + "epoch": 0.1398669515070343, + "grad_norm": 1.2809257507324219, + "learning_rate": 1.9903424894861102e-05, + "loss": 0.9895, + "step": 3159 + }, + { + "epoch": 0.13991122721184818, + "grad_norm": 1.5960781574249268, + "learning_rate": 1.9903210511353858e-05, + "loss": 1.0218, + "step": 3160 + }, + { + "epoch": 0.13995550291666206, + "grad_norm": 1.3067090511322021, + "learning_rate": 1.9902995891316496e-05, + "loss": 0.9693, + "step": 3161 + }, + { + "epoch": 0.13999977862147592, + "grad_norm": 1.2672054767608643, + "learning_rate": 1.9902781034754138e-05, + "loss": 0.9799, + "step": 3162 + }, + { + "epoch": 0.1400440543262898, + "grad_norm": 1.7057702541351318, + "learning_rate": 1.9902565941671917e-05, + "loss": 0.9832, + "step": 3163 + }, + { + "epoch": 0.14008833003110369, + "grad_norm": 1.7592772245407104, + "learning_rate": 1.9902350612074974e-05, + "loss": 1.0218, + "step": 3164 + }, + { + "epoch": 0.14013260573591757, + "grad_norm": 1.4920458793640137, + "learning_rate": 1.9902135045968447e-05, + "loss": 1.0645, + "step": 3165 + }, + { + "epoch": 0.14017688144073143, + "grad_norm": 1.449294090270996, + "learning_rate": 1.9901919243357487e-05, + "loss": 0.9094, + "step": 3166 + }, + { + "epoch": 0.1402211571455453, + "grad_norm": 1.7553473711013794, + "learning_rate": 1.990170320424725e-05, + "loss": 1.0347, + "step": 3167 + }, + { + "epoch": 0.1402654328503592, + "grad_norm": 1.849521279335022, + "learning_rate": 1.9901486928642895e-05, + "loss": 1.0406, + "step": 3168 + }, + { + "epoch": 0.14030970855517305, + "grad_norm": 1.3714783191680908, + "learning_rate": 1.9901270416549584e-05, + "loss": 1.0338, + "step": 3169 + }, + { + "epoch": 0.14035398425998694, + "grad_norm": 1.2439379692077637, + "learning_rate": 1.9901053667972493e-05, + "loss": 0.943, + "step": 3170 + }, + { + "epoch": 0.14039825996480082, + "grad_norm": 1.449491262435913, + "learning_rate": 1.9900836682916796e-05, + "loss": 1.0148, + "step": 3171 + }, + { + "epoch": 0.1404425356696147, + "grad_norm": 1.214705228805542, + "learning_rate": 1.990061946138768e-05, + "loss": 0.9659, + "step": 3172 + }, + { + "epoch": 0.14048681137442856, + "grad_norm": 1.4807687997817993, + "learning_rate": 1.9900402003390326e-05, + "loss": 1.0082, + "step": 3173 + }, + { + "epoch": 0.14053108707924244, + "grad_norm": 1.2961227893829346, + "learning_rate": 1.990018430892993e-05, + "loss": 0.8683, + "step": 3174 + }, + { + "epoch": 0.14057536278405633, + "grad_norm": 1.423746109008789, + "learning_rate": 1.9899966378011695e-05, + "loss": 1.0472, + "step": 3175 + }, + { + "epoch": 0.14061963848887019, + "grad_norm": 1.2947945594787598, + "learning_rate": 1.9899748210640824e-05, + "loss": 0.9494, + "step": 3176 + }, + { + "epoch": 0.14066391419368407, + "grad_norm": 1.376481533050537, + "learning_rate": 1.989952980682253e-05, + "loss": 0.9902, + "step": 3177 + }, + { + "epoch": 0.14070818989849795, + "grad_norm": 1.2777284383773804, + "learning_rate": 1.989931116656202e-05, + "loss": 0.9101, + "step": 3178 + }, + { + "epoch": 0.1407524656033118, + "grad_norm": 1.2259347438812256, + "learning_rate": 1.989909228986453e-05, + "loss": 0.9456, + "step": 3179 + }, + { + "epoch": 0.1407967413081257, + "grad_norm": 1.1344281435012817, + "learning_rate": 1.9898873176735277e-05, + "loss": 0.9092, + "step": 3180 + }, + { + "epoch": 0.14084101701293958, + "grad_norm": 1.6981258392333984, + "learning_rate": 1.98986538271795e-05, + "loss": 0.9855, + "step": 3181 + }, + { + "epoch": 0.14088529271775346, + "grad_norm": 1.6081733703613281, + "learning_rate": 1.9898434241202436e-05, + "loss": 1.0804, + "step": 3182 + }, + { + "epoch": 0.14092956842256732, + "grad_norm": 1.4292097091674805, + "learning_rate": 1.989821441880933e-05, + "loss": 0.8559, + "step": 3183 + }, + { + "epoch": 0.1409738441273812, + "grad_norm": 1.5071345567703247, + "learning_rate": 1.989799436000543e-05, + "loss": 1.0869, + "step": 3184 + }, + { + "epoch": 0.1410181198321951, + "grad_norm": 1.4862840175628662, + "learning_rate": 1.9897774064795994e-05, + "loss": 1.0663, + "step": 3185 + }, + { + "epoch": 0.14106239553700894, + "grad_norm": 1.2824710607528687, + "learning_rate": 1.9897553533186283e-05, + "loss": 0.9478, + "step": 3186 + }, + { + "epoch": 0.14110667124182283, + "grad_norm": 1.4218354225158691, + "learning_rate": 1.989733276518157e-05, + "loss": 0.9741, + "step": 3187 + }, + { + "epoch": 0.1411509469466367, + "grad_norm": 1.5183115005493164, + "learning_rate": 1.989711176078712e-05, + "loss": 1.0065, + "step": 3188 + }, + { + "epoch": 0.14119522265145057, + "grad_norm": 1.5513371229171753, + "learning_rate": 1.9896890520008212e-05, + "loss": 0.9783, + "step": 3189 + }, + { + "epoch": 0.14123949835626445, + "grad_norm": 1.4669305086135864, + "learning_rate": 1.9896669042850134e-05, + "loss": 1.1001, + "step": 3190 + }, + { + "epoch": 0.14128377406107834, + "grad_norm": 1.5580742359161377, + "learning_rate": 1.9896447329318173e-05, + "loss": 1.0154, + "step": 3191 + }, + { + "epoch": 0.14132804976589222, + "grad_norm": 1.4595760107040405, + "learning_rate": 1.9896225379417627e-05, + "loss": 0.9183, + "step": 3192 + }, + { + "epoch": 0.14137232547070608, + "grad_norm": 1.5854064226150513, + "learning_rate": 1.9896003193153794e-05, + "loss": 1.0626, + "step": 3193 + }, + { + "epoch": 0.14141660117551996, + "grad_norm": 1.6509555578231812, + "learning_rate": 1.9895780770531983e-05, + "loss": 1.0236, + "step": 3194 + }, + { + "epoch": 0.14146087688033385, + "grad_norm": 1.2551349401474, + "learning_rate": 1.9895558111557503e-05, + "loss": 0.9921, + "step": 3195 + }, + { + "epoch": 0.1415051525851477, + "grad_norm": 1.5121039152145386, + "learning_rate": 1.989533521623568e-05, + "loss": 0.9992, + "step": 3196 + }, + { + "epoch": 0.1415494282899616, + "grad_norm": 1.3227587938308716, + "learning_rate": 1.989511208457183e-05, + "loss": 0.9894, + "step": 3197 + }, + { + "epoch": 0.14159370399477547, + "grad_norm": 1.4284785985946655, + "learning_rate": 1.9894888716571285e-05, + "loss": 0.9636, + "step": 3198 + }, + { + "epoch": 0.14163797969958936, + "grad_norm": 1.2316288948059082, + "learning_rate": 1.989466511223938e-05, + "loss": 0.9444, + "step": 3199 + }, + { + "epoch": 0.1416822554044032, + "grad_norm": 1.4898545742034912, + "learning_rate": 1.9894441271581456e-05, + "loss": 1.0352, + "step": 3200 + }, + { + "epoch": 0.1417265311092171, + "grad_norm": 1.4130531549453735, + "learning_rate": 1.9894217194602855e-05, + "loss": 0.9849, + "step": 3201 + }, + { + "epoch": 0.14177080681403098, + "grad_norm": 1.2201799154281616, + "learning_rate": 1.9893992881308936e-05, + "loss": 0.9801, + "step": 3202 + }, + { + "epoch": 0.14181508251884484, + "grad_norm": 1.328278660774231, + "learning_rate": 1.9893768331705053e-05, + "loss": 1.0259, + "step": 3203 + }, + { + "epoch": 0.14185935822365872, + "grad_norm": 3.084007501602173, + "learning_rate": 1.9893543545796565e-05, + "loss": 1.0804, + "step": 3204 + }, + { + "epoch": 0.1419036339284726, + "grad_norm": 1.422717809677124, + "learning_rate": 1.989331852358885e-05, + "loss": 0.9918, + "step": 3205 + }, + { + "epoch": 0.14194790963328646, + "grad_norm": 1.4803426265716553, + "learning_rate": 1.9893093265087274e-05, + "loss": 0.9385, + "step": 3206 + }, + { + "epoch": 0.14199218533810035, + "grad_norm": 1.3659437894821167, + "learning_rate": 1.9892867770297224e-05, + "loss": 0.9767, + "step": 3207 + }, + { + "epoch": 0.14203646104291423, + "grad_norm": 1.669659972190857, + "learning_rate": 1.9892642039224078e-05, + "loss": 1.0881, + "step": 3208 + }, + { + "epoch": 0.14208073674772811, + "grad_norm": 1.2557088136672974, + "learning_rate": 1.9892416071873236e-05, + "loss": 0.9133, + "step": 3209 + }, + { + "epoch": 0.14212501245254197, + "grad_norm": 1.2272716760635376, + "learning_rate": 1.989218986825009e-05, + "loss": 0.9356, + "step": 3210 + }, + { + "epoch": 0.14216928815735586, + "grad_norm": 1.4025706052780151, + "learning_rate": 1.9891963428360043e-05, + "loss": 0.9135, + "step": 3211 + }, + { + "epoch": 0.14221356386216974, + "grad_norm": 1.70985746383667, + "learning_rate": 1.9891736752208503e-05, + "loss": 1.0154, + "step": 3212 + }, + { + "epoch": 0.1422578395669836, + "grad_norm": 1.3146405220031738, + "learning_rate": 1.9891509839800887e-05, + "loss": 1.0619, + "step": 3213 + }, + { + "epoch": 0.14230211527179748, + "grad_norm": 1.1605792045593262, + "learning_rate": 1.989128269114261e-05, + "loss": 0.9117, + "step": 3214 + }, + { + "epoch": 0.14234639097661136, + "grad_norm": 1.7683793306350708, + "learning_rate": 1.98910553062391e-05, + "loss": 1.0222, + "step": 3215 + }, + { + "epoch": 0.14239066668142525, + "grad_norm": 1.2334187030792236, + "learning_rate": 1.989082768509579e-05, + "loss": 0.9671, + "step": 3216 + }, + { + "epoch": 0.1424349423862391, + "grad_norm": 1.3069894313812256, + "learning_rate": 1.9890599827718112e-05, + "loss": 1.0271, + "step": 3217 + }, + { + "epoch": 0.142479218091053, + "grad_norm": 1.3184417486190796, + "learning_rate": 1.989037173411151e-05, + "loss": 0.9174, + "step": 3218 + }, + { + "epoch": 0.14252349379586687, + "grad_norm": 1.534306287765503, + "learning_rate": 1.989014340428143e-05, + "loss": 0.9862, + "step": 3219 + }, + { + "epoch": 0.14256776950068073, + "grad_norm": 1.3718746900558472, + "learning_rate": 1.9889914838233334e-05, + "loss": 1.0661, + "step": 3220 + }, + { + "epoch": 0.14261204520549461, + "grad_norm": 1.422856092453003, + "learning_rate": 1.988968603597267e-05, + "loss": 1.024, + "step": 3221 + }, + { + "epoch": 0.1426563209103085, + "grad_norm": 1.4321682453155518, + "learning_rate": 1.9889456997504908e-05, + "loss": 1.0484, + "step": 3222 + }, + { + "epoch": 0.14270059661512235, + "grad_norm": 1.408771276473999, + "learning_rate": 1.988922772283552e-05, + "loss": 0.978, + "step": 3223 + }, + { + "epoch": 0.14274487231993624, + "grad_norm": 1.2254095077514648, + "learning_rate": 1.9888998211969978e-05, + "loss": 0.9834, + "step": 3224 + }, + { + "epoch": 0.14278914802475012, + "grad_norm": 1.3398947715759277, + "learning_rate": 1.9888768464913766e-05, + "loss": 1.0222, + "step": 3225 + }, + { + "epoch": 0.142833423729564, + "grad_norm": 1.4660273790359497, + "learning_rate": 1.9888538481672372e-05, + "loss": 1.0395, + "step": 3226 + }, + { + "epoch": 0.14287769943437786, + "grad_norm": 1.2913868427276611, + "learning_rate": 1.9888308262251286e-05, + "loss": 0.9423, + "step": 3227 + }, + { + "epoch": 0.14292197513919175, + "grad_norm": 1.1080354452133179, + "learning_rate": 1.9888077806656012e-05, + "loss": 0.9266, + "step": 3228 + }, + { + "epoch": 0.14296625084400563, + "grad_norm": 1.7369571924209595, + "learning_rate": 1.988784711489205e-05, + "loss": 1.0324, + "step": 3229 + }, + { + "epoch": 0.1430105265488195, + "grad_norm": 1.3464962244033813, + "learning_rate": 1.9887616186964908e-05, + "loss": 0.9569, + "step": 3230 + }, + { + "epoch": 0.14305480225363337, + "grad_norm": 1.1716970205307007, + "learning_rate": 1.9887385022880102e-05, + "loss": 0.919, + "step": 3231 + }, + { + "epoch": 0.14309907795844726, + "grad_norm": 1.2103705406188965, + "learning_rate": 1.9887153622643162e-05, + "loss": 0.9297, + "step": 3232 + }, + { + "epoch": 0.1431433536632611, + "grad_norm": 1.3542441129684448, + "learning_rate": 1.9886921986259602e-05, + "loss": 1.0525, + "step": 3233 + }, + { + "epoch": 0.143187629368075, + "grad_norm": 1.5274618864059448, + "learning_rate": 1.9886690113734963e-05, + "loss": 1.0268, + "step": 3234 + }, + { + "epoch": 0.14323190507288888, + "grad_norm": 1.5404255390167236, + "learning_rate": 1.9886458005074783e-05, + "loss": 1.0785, + "step": 3235 + }, + { + "epoch": 0.14327618077770277, + "grad_norm": 1.4808034896850586, + "learning_rate": 1.9886225660284603e-05, + "loss": 1.0414, + "step": 3236 + }, + { + "epoch": 0.14332045648251662, + "grad_norm": 3.200537919998169, + "learning_rate": 1.988599307936997e-05, + "loss": 0.9922, + "step": 3237 + }, + { + "epoch": 0.1433647321873305, + "grad_norm": 1.571921467781067, + "learning_rate": 1.988576026233644e-05, + "loss": 0.9959, + "step": 3238 + }, + { + "epoch": 0.1434090078921444, + "grad_norm": 1.4587727785110474, + "learning_rate": 1.9885527209189577e-05, + "loss": 1.0004, + "step": 3239 + }, + { + "epoch": 0.14345328359695825, + "grad_norm": 1.9674805402755737, + "learning_rate": 1.988529391993495e-05, + "loss": 1.1199, + "step": 3240 + }, + { + "epoch": 0.14349755930177213, + "grad_norm": 1.395524501800537, + "learning_rate": 1.9885060394578123e-05, + "loss": 1.0746, + "step": 3241 + }, + { + "epoch": 0.14354183500658602, + "grad_norm": 1.5831294059753418, + "learning_rate": 1.9884826633124677e-05, + "loss": 1.0374, + "step": 3242 + }, + { + "epoch": 0.1435861107113999, + "grad_norm": 1.2818057537078857, + "learning_rate": 1.98845926355802e-05, + "loss": 1.0123, + "step": 3243 + }, + { + "epoch": 0.14363038641621376, + "grad_norm": 1.8786665201187134, + "learning_rate": 1.988435840195027e-05, + "loss": 1.0731, + "step": 3244 + }, + { + "epoch": 0.14367466212102764, + "grad_norm": 1.4583187103271484, + "learning_rate": 1.988412393224049e-05, + "loss": 1.0135, + "step": 3245 + }, + { + "epoch": 0.14371893782584153, + "grad_norm": 1.326584815979004, + "learning_rate": 1.988388922645646e-05, + "loss": 1.0451, + "step": 3246 + }, + { + "epoch": 0.14376321353065538, + "grad_norm": 1.3732391595840454, + "learning_rate": 1.988365428460378e-05, + "loss": 0.9861, + "step": 3247 + }, + { + "epoch": 0.14380748923546927, + "grad_norm": 1.2722654342651367, + "learning_rate": 1.9883419106688067e-05, + "loss": 1.0092, + "step": 3248 + }, + { + "epoch": 0.14385176494028315, + "grad_norm": 1.4343655109405518, + "learning_rate": 1.9883183692714935e-05, + "loss": 1.0803, + "step": 3249 + }, + { + "epoch": 0.143896040645097, + "grad_norm": 1.1765776872634888, + "learning_rate": 1.9882948042690007e-05, + "loss": 0.9464, + "step": 3250 + }, + { + "epoch": 0.1439403163499109, + "grad_norm": 1.3525952100753784, + "learning_rate": 1.9882712156618914e-05, + "loss": 0.9031, + "step": 3251 + }, + { + "epoch": 0.14398459205472477, + "grad_norm": 1.501645803451538, + "learning_rate": 1.9882476034507288e-05, + "loss": 1.0223, + "step": 3252 + }, + { + "epoch": 0.14402886775953866, + "grad_norm": 1.3144184350967407, + "learning_rate": 1.988223967636077e-05, + "loss": 1.0024, + "step": 3253 + }, + { + "epoch": 0.14407314346435252, + "grad_norm": 1.1362714767456055, + "learning_rate": 1.9882003082184997e-05, + "loss": 0.8512, + "step": 3254 + }, + { + "epoch": 0.1441174191691664, + "grad_norm": 1.5105111598968506, + "learning_rate": 1.988176625198563e-05, + "loss": 0.9727, + "step": 3255 + }, + { + "epoch": 0.14416169487398028, + "grad_norm": 1.7246297597885132, + "learning_rate": 1.9881529185768324e-05, + "loss": 1.1234, + "step": 3256 + }, + { + "epoch": 0.14420597057879414, + "grad_norm": 1.1719999313354492, + "learning_rate": 1.988129188353874e-05, + "loss": 0.8769, + "step": 3257 + }, + { + "epoch": 0.14425024628360802, + "grad_norm": 1.4515154361724854, + "learning_rate": 1.9881054345302545e-05, + "loss": 1.0224, + "step": 3258 + }, + { + "epoch": 0.1442945219884219, + "grad_norm": 1.3909660577774048, + "learning_rate": 1.988081657106541e-05, + "loss": 1.01, + "step": 3259 + }, + { + "epoch": 0.1443387976932358, + "grad_norm": 1.4372484683990479, + "learning_rate": 1.9880578560833017e-05, + "loss": 1.008, + "step": 3260 + }, + { + "epoch": 0.14438307339804965, + "grad_norm": 1.323828935623169, + "learning_rate": 1.988034031461105e-05, + "loss": 1.0125, + "step": 3261 + }, + { + "epoch": 0.14442734910286353, + "grad_norm": 1.4204542636871338, + "learning_rate": 1.9880101832405202e-05, + "loss": 1.0594, + "step": 3262 + }, + { + "epoch": 0.14447162480767742, + "grad_norm": 1.6859132051467896, + "learning_rate": 1.9879863114221164e-05, + "loss": 1.0715, + "step": 3263 + }, + { + "epoch": 0.14451590051249127, + "grad_norm": 1.2127848863601685, + "learning_rate": 1.987962416006464e-05, + "loss": 0.9415, + "step": 3264 + }, + { + "epoch": 0.14456017621730516, + "grad_norm": 1.712606430053711, + "learning_rate": 1.987938496994134e-05, + "loss": 1.0669, + "step": 3265 + }, + { + "epoch": 0.14460445192211904, + "grad_norm": 1.2303261756896973, + "learning_rate": 1.987914554385697e-05, + "loss": 0.9168, + "step": 3266 + }, + { + "epoch": 0.1446487276269329, + "grad_norm": 1.3837834596633911, + "learning_rate": 1.9878905881817254e-05, + "loss": 1.0077, + "step": 3267 + }, + { + "epoch": 0.14469300333174678, + "grad_norm": 1.3208847045898438, + "learning_rate": 1.9878665983827915e-05, + "loss": 0.9379, + "step": 3268 + }, + { + "epoch": 0.14473727903656067, + "grad_norm": 1.239286184310913, + "learning_rate": 1.9878425849894683e-05, + "loss": 0.9707, + "step": 3269 + }, + { + "epoch": 0.14478155474137455, + "grad_norm": 1.6080386638641357, + "learning_rate": 1.987818548002329e-05, + "loss": 1.1078, + "step": 3270 + }, + { + "epoch": 0.1448258304461884, + "grad_norm": 1.3761889934539795, + "learning_rate": 1.9877944874219485e-05, + "loss": 0.9955, + "step": 3271 + }, + { + "epoch": 0.1448701061510023, + "grad_norm": 1.4168833494186401, + "learning_rate": 1.9877704032489004e-05, + "loss": 0.9663, + "step": 3272 + }, + { + "epoch": 0.14491438185581618, + "grad_norm": 1.2565066814422607, + "learning_rate": 1.987746295483761e-05, + "loss": 0.8884, + "step": 3273 + }, + { + "epoch": 0.14495865756063003, + "grad_norm": 1.6613863706588745, + "learning_rate": 1.987722164127105e-05, + "loss": 1.069, + "step": 3274 + }, + { + "epoch": 0.14500293326544392, + "grad_norm": 1.3692964315414429, + "learning_rate": 1.98769800917951e-05, + "loss": 1.0599, + "step": 3275 + }, + { + "epoch": 0.1450472089702578, + "grad_norm": 1.1538623571395874, + "learning_rate": 1.9876738306415517e-05, + "loss": 0.9031, + "step": 3276 + }, + { + "epoch": 0.14509148467507166, + "grad_norm": 1.492951512336731, + "learning_rate": 1.9876496285138084e-05, + "loss": 1.0842, + "step": 3277 + }, + { + "epoch": 0.14513576037988554, + "grad_norm": 1.3453644514083862, + "learning_rate": 1.9876254027968583e-05, + "loss": 0.9301, + "step": 3278 + }, + { + "epoch": 0.14518003608469943, + "grad_norm": 1.4969033002853394, + "learning_rate": 1.9876011534912793e-05, + "loss": 0.9192, + "step": 3279 + }, + { + "epoch": 0.1452243117895133, + "grad_norm": 1.4952107667922974, + "learning_rate": 1.9875768805976504e-05, + "loss": 1.0018, + "step": 3280 + }, + { + "epoch": 0.14526858749432717, + "grad_norm": 1.5847448110580444, + "learning_rate": 1.9875525841165524e-05, + "loss": 1.1472, + "step": 3281 + }, + { + "epoch": 0.14531286319914105, + "grad_norm": 1.2164572477340698, + "learning_rate": 1.9875282640485647e-05, + "loss": 0.9433, + "step": 3282 + }, + { + "epoch": 0.14535713890395494, + "grad_norm": 1.2573219537734985, + "learning_rate": 1.9875039203942686e-05, + "loss": 0.9521, + "step": 3283 + }, + { + "epoch": 0.1454014146087688, + "grad_norm": 1.4256696701049805, + "learning_rate": 1.9874795531542456e-05, + "loss": 1.0259, + "step": 3284 + }, + { + "epoch": 0.14544569031358268, + "grad_norm": 1.470821738243103, + "learning_rate": 1.9874551623290773e-05, + "loss": 1.0111, + "step": 3285 + }, + { + "epoch": 0.14548996601839656, + "grad_norm": 1.3553822040557861, + "learning_rate": 1.9874307479193467e-05, + "loss": 0.9595, + "step": 3286 + }, + { + "epoch": 0.14553424172321044, + "grad_norm": 1.6391412019729614, + "learning_rate": 1.9874063099256363e-05, + "loss": 1.0864, + "step": 3287 + }, + { + "epoch": 0.1455785174280243, + "grad_norm": 1.291656255722046, + "learning_rate": 1.9873818483485304e-05, + "loss": 0.9738, + "step": 3288 + }, + { + "epoch": 0.14562279313283819, + "grad_norm": 1.3828213214874268, + "learning_rate": 1.987357363188613e-05, + "loss": 0.9326, + "step": 3289 + }, + { + "epoch": 0.14566706883765207, + "grad_norm": 1.2391945123672485, + "learning_rate": 1.987332854446469e-05, + "loss": 0.9307, + "step": 3290 + }, + { + "epoch": 0.14571134454246593, + "grad_norm": 2.1536567211151123, + "learning_rate": 1.9873083221226833e-05, + "loss": 1.0476, + "step": 3291 + }, + { + "epoch": 0.1457556202472798, + "grad_norm": 1.6402028799057007, + "learning_rate": 1.9872837662178427e-05, + "loss": 1.0567, + "step": 3292 + }, + { + "epoch": 0.1457998959520937, + "grad_norm": 1.2749338150024414, + "learning_rate": 1.987259186732533e-05, + "loss": 0.9865, + "step": 3293 + }, + { + "epoch": 0.14584417165690755, + "grad_norm": 1.2677959203720093, + "learning_rate": 1.987234583667342e-05, + "loss": 0.9536, + "step": 3294 + }, + { + "epoch": 0.14588844736172144, + "grad_norm": 1.2928227186203003, + "learning_rate": 1.9872099570228556e-05, + "loss": 0.9615, + "step": 3295 + }, + { + "epoch": 0.14593272306653532, + "grad_norm": 1.5464645624160767, + "learning_rate": 1.9871853067996644e-05, + "loss": 1.0486, + "step": 3296 + }, + { + "epoch": 0.1459769987713492, + "grad_norm": 1.413997769355774, + "learning_rate": 1.9871606329983553e-05, + "loss": 1.0567, + "step": 3297 + }, + { + "epoch": 0.14602127447616306, + "grad_norm": 1.3342286348342896, + "learning_rate": 1.9871359356195185e-05, + "loss": 0.9074, + "step": 3298 + }, + { + "epoch": 0.14606555018097694, + "grad_norm": 1.4008677005767822, + "learning_rate": 1.9871112146637436e-05, + "loss": 1.0295, + "step": 3299 + }, + { + "epoch": 0.14610982588579083, + "grad_norm": 1.297509789466858, + "learning_rate": 1.987086470131621e-05, + "loss": 0.9207, + "step": 3300 + }, + { + "epoch": 0.14615410159060468, + "grad_norm": 1.2870900630950928, + "learning_rate": 1.987061702023742e-05, + "loss": 0.9601, + "step": 3301 + }, + { + "epoch": 0.14619837729541857, + "grad_norm": 1.501725196838379, + "learning_rate": 1.9870369103406975e-05, + "loss": 0.999, + "step": 3302 + }, + { + "epoch": 0.14624265300023245, + "grad_norm": 1.1599247455596924, + "learning_rate": 1.9870120950830803e-05, + "loss": 0.8632, + "step": 3303 + }, + { + "epoch": 0.1462869287050463, + "grad_norm": 1.2369375228881836, + "learning_rate": 1.986987256251483e-05, + "loss": 0.9204, + "step": 3304 + }, + { + "epoch": 0.1463312044098602, + "grad_norm": 1.2988227605819702, + "learning_rate": 1.9869623938464984e-05, + "loss": 0.8815, + "step": 3305 + }, + { + "epoch": 0.14637548011467408, + "grad_norm": 1.5308904647827148, + "learning_rate": 1.9869375078687207e-05, + "loss": 1.0245, + "step": 3306 + }, + { + "epoch": 0.14641975581948796, + "grad_norm": 1.3032455444335938, + "learning_rate": 1.9869125983187447e-05, + "loss": 0.9723, + "step": 3307 + }, + { + "epoch": 0.14646403152430182, + "grad_norm": 1.1713800430297852, + "learning_rate": 1.9868876651971644e-05, + "loss": 0.9214, + "step": 3308 + }, + { + "epoch": 0.1465083072291157, + "grad_norm": 1.6882659196853638, + "learning_rate": 1.986862708504576e-05, + "loss": 0.9961, + "step": 3309 + }, + { + "epoch": 0.1465525829339296, + "grad_norm": 1.383142352104187, + "learning_rate": 1.9868377282415748e-05, + "loss": 1.0298, + "step": 3310 + }, + { + "epoch": 0.14659685863874344, + "grad_norm": 1.3970059156417847, + "learning_rate": 1.9868127244087585e-05, + "loss": 1.0086, + "step": 3311 + }, + { + "epoch": 0.14664113434355733, + "grad_norm": 1.3439197540283203, + "learning_rate": 1.986787697006724e-05, + "loss": 0.9742, + "step": 3312 + }, + { + "epoch": 0.1466854100483712, + "grad_norm": 1.2532544136047363, + "learning_rate": 1.9867626460360683e-05, + "loss": 0.9376, + "step": 3313 + }, + { + "epoch": 0.1467296857531851, + "grad_norm": 1.3268741369247437, + "learning_rate": 1.9867375714973904e-05, + "loss": 0.9531, + "step": 3314 + }, + { + "epoch": 0.14677396145799895, + "grad_norm": 1.436790108680725, + "learning_rate": 1.986712473391289e-05, + "loss": 1.0132, + "step": 3315 + }, + { + "epoch": 0.14681823716281284, + "grad_norm": 1.2302155494689941, + "learning_rate": 1.9866873517183637e-05, + "loss": 0.9448, + "step": 3316 + }, + { + "epoch": 0.14686251286762672, + "grad_norm": 1.382197380065918, + "learning_rate": 1.9866622064792143e-05, + "loss": 1.0135, + "step": 3317 + }, + { + "epoch": 0.14690678857244058, + "grad_norm": 1.206465244293213, + "learning_rate": 1.9866370376744413e-05, + "loss": 0.9143, + "step": 3318 + }, + { + "epoch": 0.14695106427725446, + "grad_norm": 1.8438947200775146, + "learning_rate": 1.9866118453046467e-05, + "loss": 1.1503, + "step": 3319 + }, + { + "epoch": 0.14699533998206835, + "grad_norm": 1.3546327352523804, + "learning_rate": 1.986586629370431e-05, + "loss": 1.032, + "step": 3320 + }, + { + "epoch": 0.1470396156868822, + "grad_norm": 1.327810525894165, + "learning_rate": 1.9865613898723967e-05, + "loss": 0.9525, + "step": 3321 + }, + { + "epoch": 0.1470838913916961, + "grad_norm": 1.1528043746948242, + "learning_rate": 1.9865361268111472e-05, + "loss": 0.9036, + "step": 3322 + }, + { + "epoch": 0.14712816709650997, + "grad_norm": 1.4081441164016724, + "learning_rate": 1.9865108401872856e-05, + "loss": 1.0134, + "step": 3323 + }, + { + "epoch": 0.14717244280132386, + "grad_norm": 1.8183842897415161, + "learning_rate": 1.9864855300014158e-05, + "loss": 0.9976, + "step": 3324 + }, + { + "epoch": 0.1472167185061377, + "grad_norm": 1.3978830575942993, + "learning_rate": 1.9864601962541426e-05, + "loss": 0.9466, + "step": 3325 + }, + { + "epoch": 0.1472609942109516, + "grad_norm": 1.6629990339279175, + "learning_rate": 1.9864348389460704e-05, + "loss": 1.0088, + "step": 3326 + }, + { + "epoch": 0.14730526991576548, + "grad_norm": 1.4374382495880127, + "learning_rate": 1.9864094580778056e-05, + "loss": 0.9787, + "step": 3327 + }, + { + "epoch": 0.14734954562057934, + "grad_norm": 1.430296540260315, + "learning_rate": 1.9863840536499536e-05, + "loss": 0.972, + "step": 3328 + }, + { + "epoch": 0.14739382132539322, + "grad_norm": 1.590487003326416, + "learning_rate": 1.986358625663122e-05, + "loss": 1.0618, + "step": 3329 + }, + { + "epoch": 0.1474380970302071, + "grad_norm": 1.4397368431091309, + "learning_rate": 1.9863331741179177e-05, + "loss": 0.9798, + "step": 3330 + }, + { + "epoch": 0.147482372735021, + "grad_norm": 1.64410400390625, + "learning_rate": 1.9863076990149488e-05, + "loss": 0.977, + "step": 3331 + }, + { + "epoch": 0.14752664843983485, + "grad_norm": 1.6453086137771606, + "learning_rate": 1.9862822003548234e-05, + "loss": 0.938, + "step": 3332 + }, + { + "epoch": 0.14757092414464873, + "grad_norm": 1.1790401935577393, + "learning_rate": 1.9862566781381506e-05, + "loss": 0.8992, + "step": 3333 + }, + { + "epoch": 0.1476151998494626, + "grad_norm": 1.3916723728179932, + "learning_rate": 1.98623113236554e-05, + "loss": 0.9815, + "step": 3334 + }, + { + "epoch": 0.14765947555427647, + "grad_norm": 1.3783339262008667, + "learning_rate": 1.9862055630376022e-05, + "loss": 1.0331, + "step": 3335 + }, + { + "epoch": 0.14770375125909035, + "grad_norm": 1.3878226280212402, + "learning_rate": 1.9861799701549473e-05, + "loss": 0.9795, + "step": 3336 + }, + { + "epoch": 0.14774802696390424, + "grad_norm": 1.4613933563232422, + "learning_rate": 1.986154353718187e-05, + "loss": 1.003, + "step": 3337 + }, + { + "epoch": 0.1477923026687181, + "grad_norm": 1.351491093635559, + "learning_rate": 1.9861287137279325e-05, + "loss": 0.9354, + "step": 3338 + }, + { + "epoch": 0.14783657837353198, + "grad_norm": 1.6049748659133911, + "learning_rate": 1.986103050184797e-05, + "loss": 1.0284, + "step": 3339 + }, + { + "epoch": 0.14788085407834586, + "grad_norm": 1.5497877597808838, + "learning_rate": 1.9860773630893928e-05, + "loss": 1.0483, + "step": 3340 + }, + { + "epoch": 0.14792512978315975, + "grad_norm": 1.2384023666381836, + "learning_rate": 1.9860516524423337e-05, + "loss": 0.9008, + "step": 3341 + }, + { + "epoch": 0.1479694054879736, + "grad_norm": 1.212498426437378, + "learning_rate": 1.986025918244234e-05, + "loss": 0.979, + "step": 3342 + }, + { + "epoch": 0.1480136811927875, + "grad_norm": 1.3533358573913574, + "learning_rate": 1.9860001604957076e-05, + "loss": 1.0058, + "step": 3343 + }, + { + "epoch": 0.14805795689760137, + "grad_norm": 1.2712312936782837, + "learning_rate": 1.985974379197371e-05, + "loss": 0.9416, + "step": 3344 + }, + { + "epoch": 0.14810223260241523, + "grad_norm": 1.2007743120193481, + "learning_rate": 1.9859485743498383e-05, + "loss": 0.9593, + "step": 3345 + }, + { + "epoch": 0.1481465083072291, + "grad_norm": 1.3028407096862793, + "learning_rate": 1.985922745953727e-05, + "loss": 0.9395, + "step": 3346 + }, + { + "epoch": 0.148190784012043, + "grad_norm": 1.3121365308761597, + "learning_rate": 1.985896894009654e-05, + "loss": 0.9323, + "step": 3347 + }, + { + "epoch": 0.14823505971685685, + "grad_norm": 1.2947473526000977, + "learning_rate": 1.985871018518236e-05, + "loss": 1.0198, + "step": 3348 + }, + { + "epoch": 0.14827933542167074, + "grad_norm": 1.7198106050491333, + "learning_rate": 1.9858451194800917e-05, + "loss": 1.1329, + "step": 3349 + }, + { + "epoch": 0.14832361112648462, + "grad_norm": 1.3016786575317383, + "learning_rate": 1.9858191968958394e-05, + "loss": 0.971, + "step": 3350 + }, + { + "epoch": 0.1483678868312985, + "grad_norm": 1.3083328008651733, + "learning_rate": 1.9857932507660983e-05, + "loss": 0.9241, + "step": 3351 + }, + { + "epoch": 0.14841216253611236, + "grad_norm": 1.4952656030654907, + "learning_rate": 1.9857672810914878e-05, + "loss": 0.9467, + "step": 3352 + }, + { + "epoch": 0.14845643824092625, + "grad_norm": 1.294986605644226, + "learning_rate": 1.9857412878726284e-05, + "loss": 0.9947, + "step": 3353 + }, + { + "epoch": 0.14850071394574013, + "grad_norm": 1.339504599571228, + "learning_rate": 1.9857152711101414e-05, + "loss": 0.9114, + "step": 3354 + }, + { + "epoch": 0.148544989650554, + "grad_norm": 1.4786081314086914, + "learning_rate": 1.9856892308046477e-05, + "loss": 1.0348, + "step": 3355 + }, + { + "epoch": 0.14858926535536787, + "grad_norm": 1.376528263092041, + "learning_rate": 1.9856631669567694e-05, + "loss": 1.0302, + "step": 3356 + }, + { + "epoch": 0.14863354106018176, + "grad_norm": 2.682837963104248, + "learning_rate": 1.985637079567128e-05, + "loss": 1.0129, + "step": 3357 + }, + { + "epoch": 0.14867781676499564, + "grad_norm": 1.7136754989624023, + "learning_rate": 1.9856109686363485e-05, + "loss": 1.0844, + "step": 3358 + }, + { + "epoch": 0.1487220924698095, + "grad_norm": 1.6033118963241577, + "learning_rate": 1.985584834165053e-05, + "loss": 1.0735, + "step": 3359 + }, + { + "epoch": 0.14876636817462338, + "grad_norm": 1.4081751108169556, + "learning_rate": 1.9855586761538665e-05, + "loss": 0.9505, + "step": 3360 + }, + { + "epoch": 0.14881064387943727, + "grad_norm": 1.49469792842865, + "learning_rate": 1.9855324946034136e-05, + "loss": 1.0124, + "step": 3361 + }, + { + "epoch": 0.14885491958425112, + "grad_norm": 2.02915096282959, + "learning_rate": 1.985506289514319e-05, + "loss": 1.1855, + "step": 3362 + }, + { + "epoch": 0.148899195289065, + "grad_norm": 1.3060847520828247, + "learning_rate": 1.9854800608872096e-05, + "loss": 0.9927, + "step": 3363 + }, + { + "epoch": 0.1489434709938789, + "grad_norm": 1.7533951997756958, + "learning_rate": 1.9854538087227112e-05, + "loss": 1.0486, + "step": 3364 + }, + { + "epoch": 0.14898774669869275, + "grad_norm": 1.6699844598770142, + "learning_rate": 1.9854275330214506e-05, + "loss": 0.9567, + "step": 3365 + }, + { + "epoch": 0.14903202240350663, + "grad_norm": 1.2770336866378784, + "learning_rate": 1.985401233784056e-05, + "loss": 0.9752, + "step": 3366 + }, + { + "epoch": 0.14907629810832052, + "grad_norm": 1.5175429582595825, + "learning_rate": 1.985374911011156e-05, + "loss": 1.1094, + "step": 3367 + }, + { + "epoch": 0.1491205738131344, + "grad_norm": 1.5210751295089722, + "learning_rate": 1.9853485647033775e-05, + "loss": 1.0393, + "step": 3368 + }, + { + "epoch": 0.14916484951794826, + "grad_norm": 1.3577975034713745, + "learning_rate": 1.985322194861351e-05, + "loss": 0.9059, + "step": 3369 + }, + { + "epoch": 0.14920912522276214, + "grad_norm": 1.3560481071472168, + "learning_rate": 1.9852958014857066e-05, + "loss": 0.8792, + "step": 3370 + }, + { + "epoch": 0.14925340092757602, + "grad_norm": 1.6186892986297607, + "learning_rate": 1.985269384577074e-05, + "loss": 1.0796, + "step": 3371 + }, + { + "epoch": 0.14929767663238988, + "grad_norm": 2.031906843185425, + "learning_rate": 1.9852429441360845e-05, + "loss": 0.9203, + "step": 3372 + }, + { + "epoch": 0.14934195233720376, + "grad_norm": 1.3817644119262695, + "learning_rate": 1.9852164801633697e-05, + "loss": 1.0322, + "step": 3373 + }, + { + "epoch": 0.14938622804201765, + "grad_norm": 1.4766541719436646, + "learning_rate": 1.9851899926595612e-05, + "loss": 1.0528, + "step": 3374 + }, + { + "epoch": 0.14943050374683153, + "grad_norm": 2.0564889907836914, + "learning_rate": 1.985163481625292e-05, + "loss": 1.0418, + "step": 3375 + }, + { + "epoch": 0.1494747794516454, + "grad_norm": 1.311258316040039, + "learning_rate": 1.985136947061195e-05, + "loss": 0.9759, + "step": 3376 + }, + { + "epoch": 0.14951905515645927, + "grad_norm": 1.5484931468963623, + "learning_rate": 1.985110388967904e-05, + "loss": 1.0247, + "step": 3377 + }, + { + "epoch": 0.14956333086127316, + "grad_norm": 1.8453710079193115, + "learning_rate": 1.9850838073460535e-05, + "loss": 0.976, + "step": 3378 + }, + { + "epoch": 0.14960760656608701, + "grad_norm": 2.0576231479644775, + "learning_rate": 1.9850572021962788e-05, + "loss": 1.0165, + "step": 3379 + }, + { + "epoch": 0.1496518822709009, + "grad_norm": 1.2683364152908325, + "learning_rate": 1.9850305735192146e-05, + "loss": 1.0424, + "step": 3380 + }, + { + "epoch": 0.14969615797571478, + "grad_norm": 1.3144906759262085, + "learning_rate": 1.9850039213154972e-05, + "loss": 0.9201, + "step": 3381 + }, + { + "epoch": 0.14974043368052864, + "grad_norm": 1.4016047716140747, + "learning_rate": 1.9849772455857633e-05, + "loss": 1.0142, + "step": 3382 + }, + { + "epoch": 0.14978470938534252, + "grad_norm": 1.3535481691360474, + "learning_rate": 1.9849505463306494e-05, + "loss": 0.9808, + "step": 3383 + }, + { + "epoch": 0.1498289850901564, + "grad_norm": 1.3480559587478638, + "learning_rate": 1.9849238235507943e-05, + "loss": 0.954, + "step": 3384 + }, + { + "epoch": 0.1498732607949703, + "grad_norm": 1.2324068546295166, + "learning_rate": 1.9848970772468353e-05, + "loss": 0.9735, + "step": 3385 + }, + { + "epoch": 0.14991753649978415, + "grad_norm": 1.502356767654419, + "learning_rate": 1.9848703074194116e-05, + "loss": 0.9526, + "step": 3386 + }, + { + "epoch": 0.14996181220459803, + "grad_norm": 1.8124340772628784, + "learning_rate": 1.9848435140691627e-05, + "loss": 1.0775, + "step": 3387 + }, + { + "epoch": 0.15000608790941192, + "grad_norm": 1.4167232513427734, + "learning_rate": 1.984816697196728e-05, + "loss": 1.0027, + "step": 3388 + }, + { + "epoch": 0.15005036361422577, + "grad_norm": 1.402016282081604, + "learning_rate": 1.9847898568027485e-05, + "loss": 0.944, + "step": 3389 + }, + { + "epoch": 0.15009463931903966, + "grad_norm": 1.7083994150161743, + "learning_rate": 1.9847629928878653e-05, + "loss": 1.0992, + "step": 3390 + }, + { + "epoch": 0.15013891502385354, + "grad_norm": 1.3521844148635864, + "learning_rate": 1.9847361054527196e-05, + "loss": 0.987, + "step": 3391 + }, + { + "epoch": 0.1501831907286674, + "grad_norm": 1.2784754037857056, + "learning_rate": 1.9847091944979538e-05, + "loss": 0.9794, + "step": 3392 + }, + { + "epoch": 0.15022746643348128, + "grad_norm": 1.527724027633667, + "learning_rate": 1.984682260024211e-05, + "loss": 1.121, + "step": 3393 + }, + { + "epoch": 0.15027174213829517, + "grad_norm": 1.4003689289093018, + "learning_rate": 1.984655302032134e-05, + "loss": 0.9702, + "step": 3394 + }, + { + "epoch": 0.15031601784310905, + "grad_norm": 1.2312043905258179, + "learning_rate": 1.984628320522367e-05, + "loss": 0.8316, + "step": 3395 + }, + { + "epoch": 0.1503602935479229, + "grad_norm": 1.248180627822876, + "learning_rate": 1.9846013154955538e-05, + "loss": 1.0064, + "step": 3396 + }, + { + "epoch": 0.1504045692527368, + "grad_norm": 1.4367526769638062, + "learning_rate": 1.98457428695234e-05, + "loss": 1.0573, + "step": 3397 + }, + { + "epoch": 0.15044884495755068, + "grad_norm": 2.0967345237731934, + "learning_rate": 1.9845472348933713e-05, + "loss": 0.9624, + "step": 3398 + }, + { + "epoch": 0.15049312066236453, + "grad_norm": 1.4467813968658447, + "learning_rate": 1.9845201593192935e-05, + "loss": 1.0134, + "step": 3399 + }, + { + "epoch": 0.15053739636717842, + "grad_norm": 1.4448511600494385, + "learning_rate": 1.9844930602307533e-05, + "loss": 1.0068, + "step": 3400 + }, + { + "epoch": 0.1505816720719923, + "grad_norm": 1.5737766027450562, + "learning_rate": 1.984465937628398e-05, + "loss": 0.9607, + "step": 3401 + }, + { + "epoch": 0.15062594777680618, + "grad_norm": 1.2853196859359741, + "learning_rate": 1.9844387915128752e-05, + "loss": 0.9673, + "step": 3402 + }, + { + "epoch": 0.15067022348162004, + "grad_norm": 1.6490720510482788, + "learning_rate": 1.9844116218848335e-05, + "loss": 0.9946, + "step": 3403 + }, + { + "epoch": 0.15071449918643393, + "grad_norm": 1.4229161739349365, + "learning_rate": 1.9843844287449218e-05, + "loss": 1.0202, + "step": 3404 + }, + { + "epoch": 0.1507587748912478, + "grad_norm": 1.4733068943023682, + "learning_rate": 1.9843572120937895e-05, + "loss": 1.0124, + "step": 3405 + }, + { + "epoch": 0.15080305059606167, + "grad_norm": 1.2546294927597046, + "learning_rate": 1.9843299719320864e-05, + "loss": 0.9341, + "step": 3406 + }, + { + "epoch": 0.15084732630087555, + "grad_norm": 1.3871928453445435, + "learning_rate": 1.984302708260464e-05, + "loss": 0.9956, + "step": 3407 + }, + { + "epoch": 0.15089160200568943, + "grad_norm": 1.5332157611846924, + "learning_rate": 1.9842754210795726e-05, + "loss": 1.0268, + "step": 3408 + }, + { + "epoch": 0.1509358777105033, + "grad_norm": 1.4711576700210571, + "learning_rate": 1.984248110390064e-05, + "loss": 1.0206, + "step": 3409 + }, + { + "epoch": 0.15098015341531718, + "grad_norm": 1.6780515909194946, + "learning_rate": 1.9842207761925906e-05, + "loss": 1.0405, + "step": 3410 + }, + { + "epoch": 0.15102442912013106, + "grad_norm": 1.4708987474441528, + "learning_rate": 1.9841934184878056e-05, + "loss": 0.9812, + "step": 3411 + }, + { + "epoch": 0.15106870482494494, + "grad_norm": 1.334082007408142, + "learning_rate": 1.9841660372763622e-05, + "loss": 0.9643, + "step": 3412 + }, + { + "epoch": 0.1511129805297588, + "grad_norm": 1.3965370655059814, + "learning_rate": 1.984138632558914e-05, + "loss": 0.9736, + "step": 3413 + }, + { + "epoch": 0.15115725623457268, + "grad_norm": 1.3082938194274902, + "learning_rate": 1.984111204336116e-05, + "loss": 0.9605, + "step": 3414 + }, + { + "epoch": 0.15120153193938657, + "grad_norm": 1.2623666524887085, + "learning_rate": 1.9840837526086237e-05, + "loss": 0.9186, + "step": 3415 + }, + { + "epoch": 0.15124580764420043, + "grad_norm": 1.5356417894363403, + "learning_rate": 1.9840562773770917e-05, + "loss": 1.0256, + "step": 3416 + }, + { + "epoch": 0.1512900833490143, + "grad_norm": 1.4209651947021484, + "learning_rate": 1.984028778642177e-05, + "loss": 0.9765, + "step": 3417 + }, + { + "epoch": 0.1513343590538282, + "grad_norm": 1.6013786792755127, + "learning_rate": 1.984001256404536e-05, + "loss": 1.0282, + "step": 3418 + }, + { + "epoch": 0.15137863475864208, + "grad_norm": 1.2705068588256836, + "learning_rate": 1.9839737106648262e-05, + "loss": 0.9057, + "step": 3419 + }, + { + "epoch": 0.15142291046345593, + "grad_norm": 1.8138158321380615, + "learning_rate": 1.9839461414237057e-05, + "loss": 1.009, + "step": 3420 + }, + { + "epoch": 0.15146718616826982, + "grad_norm": 1.3437992334365845, + "learning_rate": 1.9839185486818323e-05, + "loss": 1.0377, + "step": 3421 + }, + { + "epoch": 0.1515114618730837, + "grad_norm": 1.2300959825515747, + "learning_rate": 1.9838909324398662e-05, + "loss": 1.0078, + "step": 3422 + }, + { + "epoch": 0.15155573757789756, + "grad_norm": 1.5715999603271484, + "learning_rate": 1.9838632926984656e-05, + "loss": 1.0562, + "step": 3423 + }, + { + "epoch": 0.15160001328271144, + "grad_norm": 1.4024070501327515, + "learning_rate": 1.983835629458292e-05, + "loss": 1.0148, + "step": 3424 + }, + { + "epoch": 0.15164428898752533, + "grad_norm": 1.2487797737121582, + "learning_rate": 1.983807942720005e-05, + "loss": 0.9467, + "step": 3425 + }, + { + "epoch": 0.15168856469233918, + "grad_norm": 1.8362332582473755, + "learning_rate": 1.9837802324842662e-05, + "loss": 0.974, + "step": 3426 + }, + { + "epoch": 0.15173284039715307, + "grad_norm": 1.3269027471542358, + "learning_rate": 1.9837524987517377e-05, + "loss": 0.902, + "step": 3427 + }, + { + "epoch": 0.15177711610196695, + "grad_norm": 1.4545060396194458, + "learning_rate": 1.9837247415230823e-05, + "loss": 0.9407, + "step": 3428 + }, + { + "epoch": 0.15182139180678084, + "grad_norm": 1.286553978919983, + "learning_rate": 1.983696960798962e-05, + "loss": 0.967, + "step": 3429 + }, + { + "epoch": 0.1518656675115947, + "grad_norm": 1.6934270858764648, + "learning_rate": 1.9836691565800405e-05, + "loss": 0.9621, + "step": 3430 + }, + { + "epoch": 0.15190994321640858, + "grad_norm": 1.3160754442214966, + "learning_rate": 1.9836413288669825e-05, + "loss": 0.9744, + "step": 3431 + }, + { + "epoch": 0.15195421892122246, + "grad_norm": 1.2687427997589111, + "learning_rate": 1.983613477660452e-05, + "loss": 0.9473, + "step": 3432 + }, + { + "epoch": 0.15199849462603632, + "grad_norm": 1.3083189725875854, + "learning_rate": 1.9835856029611148e-05, + "loss": 0.9629, + "step": 3433 + }, + { + "epoch": 0.1520427703308502, + "grad_norm": 1.6304832696914673, + "learning_rate": 1.983557704769636e-05, + "loss": 0.9759, + "step": 3434 + }, + { + "epoch": 0.1520870460356641, + "grad_norm": 1.3810713291168213, + "learning_rate": 1.9835297830866827e-05, + "loss": 1.0148, + "step": 3435 + }, + { + "epoch": 0.15213132174047794, + "grad_norm": 1.2898324728012085, + "learning_rate": 1.9835018379129208e-05, + "loss": 0.9644, + "step": 3436 + }, + { + "epoch": 0.15217559744529183, + "grad_norm": 1.449642539024353, + "learning_rate": 1.9834738692490186e-05, + "loss": 0.9637, + "step": 3437 + }, + { + "epoch": 0.1522198731501057, + "grad_norm": 1.3071316480636597, + "learning_rate": 1.983445877095644e-05, + "loss": 0.9476, + "step": 3438 + }, + { + "epoch": 0.1522641488549196, + "grad_norm": 1.7839995622634888, + "learning_rate": 1.9834178614534652e-05, + "loss": 0.9186, + "step": 3439 + }, + { + "epoch": 0.15230842455973345, + "grad_norm": 1.5746586322784424, + "learning_rate": 1.9833898223231513e-05, + "loss": 0.9364, + "step": 3440 + }, + { + "epoch": 0.15235270026454734, + "grad_norm": 1.387505054473877, + "learning_rate": 1.9833617597053723e-05, + "loss": 0.9383, + "step": 3441 + }, + { + "epoch": 0.15239697596936122, + "grad_norm": 1.6346515417099, + "learning_rate": 1.9833336736007986e-05, + "loss": 0.9776, + "step": 3442 + }, + { + "epoch": 0.15244125167417508, + "grad_norm": 1.5517994165420532, + "learning_rate": 1.9833055640101008e-05, + "loss": 1.004, + "step": 3443 + }, + { + "epoch": 0.15248552737898896, + "grad_norm": 1.2625411748886108, + "learning_rate": 1.98327743093395e-05, + "loss": 0.937, + "step": 3444 + }, + { + "epoch": 0.15252980308380285, + "grad_norm": 1.2555125951766968, + "learning_rate": 1.9832492743730183e-05, + "loss": 0.9089, + "step": 3445 + }, + { + "epoch": 0.15257407878861673, + "grad_norm": 1.5157268047332764, + "learning_rate": 1.9832210943279783e-05, + "loss": 0.9767, + "step": 3446 + }, + { + "epoch": 0.15261835449343059, + "grad_norm": 1.5268326997756958, + "learning_rate": 1.9831928907995032e-05, + "loss": 1.0841, + "step": 3447 + }, + { + "epoch": 0.15266263019824447, + "grad_norm": 1.4801182746887207, + "learning_rate": 1.983164663788266e-05, + "loss": 0.9222, + "step": 3448 + }, + { + "epoch": 0.15270690590305835, + "grad_norm": 1.4153834581375122, + "learning_rate": 1.983136413294942e-05, + "loss": 1.0505, + "step": 3449 + }, + { + "epoch": 0.1527511816078722, + "grad_norm": 1.150359034538269, + "learning_rate": 1.983108139320205e-05, + "loss": 0.9109, + "step": 3450 + }, + { + "epoch": 0.1527954573126861, + "grad_norm": 1.2925273180007935, + "learning_rate": 1.9830798418647306e-05, + "loss": 0.9434, + "step": 3451 + }, + { + "epoch": 0.15283973301749998, + "grad_norm": 1.489630103111267, + "learning_rate": 1.983051520929194e-05, + "loss": 0.9975, + "step": 3452 + }, + { + "epoch": 0.15288400872231384, + "grad_norm": 1.2888351678848267, + "learning_rate": 1.9830231765142734e-05, + "loss": 1.0266, + "step": 3453 + }, + { + "epoch": 0.15292828442712772, + "grad_norm": 1.248982548713684, + "learning_rate": 1.982994808620644e-05, + "loss": 0.9179, + "step": 3454 + }, + { + "epoch": 0.1529725601319416, + "grad_norm": 1.4095321893692017, + "learning_rate": 1.982966417248984e-05, + "loss": 0.9937, + "step": 3455 + }, + { + "epoch": 0.1530168358367555, + "grad_norm": 1.5984357595443726, + "learning_rate": 1.9829380023999715e-05, + "loss": 1.0301, + "step": 3456 + }, + { + "epoch": 0.15306111154156934, + "grad_norm": 1.6704068183898926, + "learning_rate": 1.982909564074285e-05, + "loss": 1.0583, + "step": 3457 + }, + { + "epoch": 0.15310538724638323, + "grad_norm": 1.5155227184295654, + "learning_rate": 1.9828811022726042e-05, + "loss": 1.0043, + "step": 3458 + }, + { + "epoch": 0.1531496629511971, + "grad_norm": 1.4249672889709473, + "learning_rate": 1.9828526169956083e-05, + "loss": 1.0173, + "step": 3459 + }, + { + "epoch": 0.15319393865601097, + "grad_norm": 2.0692877769470215, + "learning_rate": 1.9828241082439782e-05, + "loss": 1.1513, + "step": 3460 + }, + { + "epoch": 0.15323821436082485, + "grad_norm": 1.6118801832199097, + "learning_rate": 1.9827955760183943e-05, + "loss": 1.0674, + "step": 3461 + }, + { + "epoch": 0.15328249006563874, + "grad_norm": 1.3885900974273682, + "learning_rate": 1.9827670203195382e-05, + "loss": 1.0249, + "step": 3462 + }, + { + "epoch": 0.15332676577045262, + "grad_norm": 1.6575934886932373, + "learning_rate": 1.9827384411480924e-05, + "loss": 1.0339, + "step": 3463 + }, + { + "epoch": 0.15337104147526648, + "grad_norm": 1.458638310432434, + "learning_rate": 1.9827098385047385e-05, + "loss": 1.005, + "step": 3464 + }, + { + "epoch": 0.15341531718008036, + "grad_norm": 1.4070078134536743, + "learning_rate": 1.9826812123901605e-05, + "loss": 1.0364, + "step": 3465 + }, + { + "epoch": 0.15345959288489425, + "grad_norm": 1.3555742502212524, + "learning_rate": 1.9826525628050423e-05, + "loss": 0.9632, + "step": 3466 + }, + { + "epoch": 0.1535038685897081, + "grad_norm": 1.4674431085586548, + "learning_rate": 1.9826238897500675e-05, + "loss": 1.0244, + "step": 3467 + }, + { + "epoch": 0.153548144294522, + "grad_norm": 1.2449147701263428, + "learning_rate": 1.9825951932259213e-05, + "loss": 0.913, + "step": 3468 + }, + { + "epoch": 0.15359241999933587, + "grad_norm": 1.501749873161316, + "learning_rate": 1.9825664732332886e-05, + "loss": 0.9968, + "step": 3469 + }, + { + "epoch": 0.15363669570414973, + "grad_norm": 1.1858173608779907, + "learning_rate": 1.982537729772856e-05, + "loss": 0.9505, + "step": 3470 + }, + { + "epoch": 0.1536809714089636, + "grad_norm": 1.3504467010498047, + "learning_rate": 1.9825089628453094e-05, + "loss": 1.0216, + "step": 3471 + }, + { + "epoch": 0.1537252471137775, + "grad_norm": 1.586146593093872, + "learning_rate": 1.982480172451336e-05, + "loss": 0.9053, + "step": 3472 + }, + { + "epoch": 0.15376952281859138, + "grad_norm": 1.5771336555480957, + "learning_rate": 1.9824513585916242e-05, + "loss": 1.0795, + "step": 3473 + }, + { + "epoch": 0.15381379852340524, + "grad_norm": 1.8054333925247192, + "learning_rate": 1.982422521266862e-05, + "loss": 0.9397, + "step": 3474 + }, + { + "epoch": 0.15385807422821912, + "grad_norm": 1.3901594877243042, + "learning_rate": 1.982393660477737e-05, + "loss": 0.9666, + "step": 3475 + }, + { + "epoch": 0.153902349933033, + "grad_norm": 1.3439069986343384, + "learning_rate": 1.9823647762249397e-05, + "loss": 1.0366, + "step": 3476 + }, + { + "epoch": 0.15394662563784686, + "grad_norm": 1.2607547044754028, + "learning_rate": 1.9823358685091597e-05, + "loss": 0.948, + "step": 3477 + }, + { + "epoch": 0.15399090134266075, + "grad_norm": 1.5030980110168457, + "learning_rate": 1.982306937331087e-05, + "loss": 1.1112, + "step": 3478 + }, + { + "epoch": 0.15403517704747463, + "grad_norm": 1.6743578910827637, + "learning_rate": 1.982277982691413e-05, + "loss": 1.124, + "step": 3479 + }, + { + "epoch": 0.1540794527522885, + "grad_norm": 1.2031010389328003, + "learning_rate": 1.9822490045908293e-05, + "loss": 0.8977, + "step": 3480 + }, + { + "epoch": 0.15412372845710237, + "grad_norm": 1.353627324104309, + "learning_rate": 1.9822200030300277e-05, + "loss": 0.92, + "step": 3481 + }, + { + "epoch": 0.15416800416191626, + "grad_norm": 1.2714111804962158, + "learning_rate": 1.9821909780097014e-05, + "loss": 0.9759, + "step": 3482 + }, + { + "epoch": 0.15421227986673014, + "grad_norm": 1.160983920097351, + "learning_rate": 1.9821619295305432e-05, + "loss": 0.9137, + "step": 3483 + }, + { + "epoch": 0.154256555571544, + "grad_norm": 1.5575960874557495, + "learning_rate": 1.982132857593247e-05, + "loss": 1.0378, + "step": 3484 + }, + { + "epoch": 0.15430083127635788, + "grad_norm": 1.738916277885437, + "learning_rate": 1.9821037621985072e-05, + "loss": 0.9931, + "step": 3485 + }, + { + "epoch": 0.15434510698117176, + "grad_norm": 1.394728183746338, + "learning_rate": 1.9820746433470185e-05, + "loss": 0.9797, + "step": 3486 + }, + { + "epoch": 0.15438938268598562, + "grad_norm": 1.3671844005584717, + "learning_rate": 1.982045501039477e-05, + "loss": 0.9792, + "step": 3487 + }, + { + "epoch": 0.1544336583907995, + "grad_norm": 1.4759657382965088, + "learning_rate": 1.982016335276578e-05, + "loss": 1.0441, + "step": 3488 + }, + { + "epoch": 0.1544779340956134, + "grad_norm": 1.3215969800949097, + "learning_rate": 1.9819871460590185e-05, + "loss": 1.0498, + "step": 3489 + }, + { + "epoch": 0.15452220980042727, + "grad_norm": 1.2000423669815063, + "learning_rate": 1.9819579333874955e-05, + "loss": 0.8451, + "step": 3490 + }, + { + "epoch": 0.15456648550524113, + "grad_norm": 1.1543692350387573, + "learning_rate": 1.9819286972627066e-05, + "loss": 0.9108, + "step": 3491 + }, + { + "epoch": 0.15461076121005501, + "grad_norm": 1.2803926467895508, + "learning_rate": 1.9818994376853508e-05, + "loss": 0.9019, + "step": 3492 + }, + { + "epoch": 0.1546550369148689, + "grad_norm": 1.3479254245758057, + "learning_rate": 1.981870154656126e-05, + "loss": 0.9346, + "step": 3493 + }, + { + "epoch": 0.15469931261968276, + "grad_norm": 1.3811824321746826, + "learning_rate": 1.9818408481757326e-05, + "loss": 1.0187, + "step": 3494 + }, + { + "epoch": 0.15474358832449664, + "grad_norm": 1.6339575052261353, + "learning_rate": 1.9818115182448695e-05, + "loss": 1.0545, + "step": 3495 + }, + { + "epoch": 0.15478786402931052, + "grad_norm": 1.4689377546310425, + "learning_rate": 1.9817821648642374e-05, + "loss": 0.9515, + "step": 3496 + }, + { + "epoch": 0.15483213973412438, + "grad_norm": 1.426344633102417, + "learning_rate": 1.9817527880345383e-05, + "loss": 1.018, + "step": 3497 + }, + { + "epoch": 0.15487641543893826, + "grad_norm": 1.4475975036621094, + "learning_rate": 1.981723387756473e-05, + "loss": 0.9525, + "step": 3498 + }, + { + "epoch": 0.15492069114375215, + "grad_norm": 1.3254742622375488, + "learning_rate": 1.981693964030744e-05, + "loss": 0.8968, + "step": 3499 + }, + { + "epoch": 0.15496496684856603, + "grad_norm": 1.1850850582122803, + "learning_rate": 1.9816645168580538e-05, + "loss": 0.8958, + "step": 3500 + }, + { + "epoch": 0.1550092425533799, + "grad_norm": 1.4722551107406616, + "learning_rate": 1.981635046239106e-05, + "loss": 0.9693, + "step": 3501 + }, + { + "epoch": 0.15505351825819377, + "grad_norm": 1.412749171257019, + "learning_rate": 1.9816055521746046e-05, + "loss": 0.988, + "step": 3502 + }, + { + "epoch": 0.15509779396300766, + "grad_norm": 1.2641419172286987, + "learning_rate": 1.9815760346652535e-05, + "loss": 0.9209, + "step": 3503 + }, + { + "epoch": 0.1551420696678215, + "grad_norm": 1.423911690711975, + "learning_rate": 1.9815464937117583e-05, + "loss": 0.9948, + "step": 3504 + }, + { + "epoch": 0.1551863453726354, + "grad_norm": 1.3402056694030762, + "learning_rate": 1.9815169293148245e-05, + "loss": 0.9611, + "step": 3505 + }, + { + "epoch": 0.15523062107744928, + "grad_norm": 1.6079139709472656, + "learning_rate": 1.9814873414751575e-05, + "loss": 1.0509, + "step": 3506 + }, + { + "epoch": 0.15527489678226317, + "grad_norm": 1.5217185020446777, + "learning_rate": 1.9814577301934647e-05, + "loss": 1.0183, + "step": 3507 + }, + { + "epoch": 0.15531917248707702, + "grad_norm": 1.6488193273544312, + "learning_rate": 1.9814280954704533e-05, + "loss": 0.9904, + "step": 3508 + }, + { + "epoch": 0.1553634481918909, + "grad_norm": 1.2570346593856812, + "learning_rate": 1.981398437306831e-05, + "loss": 0.9134, + "step": 3509 + }, + { + "epoch": 0.1554077238967048, + "grad_norm": 1.4792110919952393, + "learning_rate": 1.9813687557033058e-05, + "loss": 1.1629, + "step": 3510 + }, + { + "epoch": 0.15545199960151865, + "grad_norm": 1.3192319869995117, + "learning_rate": 1.981339050660587e-05, + "loss": 1.0136, + "step": 3511 + }, + { + "epoch": 0.15549627530633253, + "grad_norm": 1.4547611474990845, + "learning_rate": 1.9813093221793842e-05, + "loss": 0.8915, + "step": 3512 + }, + { + "epoch": 0.15554055101114642, + "grad_norm": 1.2517833709716797, + "learning_rate": 1.9812795702604073e-05, + "loss": 0.9674, + "step": 3513 + }, + { + "epoch": 0.15558482671596027, + "grad_norm": 1.2862908840179443, + "learning_rate": 1.9812497949043665e-05, + "loss": 1.0278, + "step": 3514 + }, + { + "epoch": 0.15562910242077416, + "grad_norm": 1.2271530628204346, + "learning_rate": 1.9812199961119737e-05, + "loss": 0.9713, + "step": 3515 + }, + { + "epoch": 0.15567337812558804, + "grad_norm": 1.49098801612854, + "learning_rate": 1.98119017388394e-05, + "loss": 0.9226, + "step": 3516 + }, + { + "epoch": 0.15571765383040193, + "grad_norm": 1.252315640449524, + "learning_rate": 1.981160328220978e-05, + "loss": 0.9384, + "step": 3517 + }, + { + "epoch": 0.15576192953521578, + "grad_norm": 1.4029978513717651, + "learning_rate": 1.9811304591238e-05, + "loss": 1.0136, + "step": 3518 + }, + { + "epoch": 0.15580620524002967, + "grad_norm": 2.135971784591675, + "learning_rate": 1.9811005665931205e-05, + "loss": 1.1004, + "step": 3519 + }, + { + "epoch": 0.15585048094484355, + "grad_norm": 2.162015199661255, + "learning_rate": 1.9810706506296525e-05, + "loss": 0.9818, + "step": 3520 + }, + { + "epoch": 0.1558947566496574, + "grad_norm": 1.276784062385559, + "learning_rate": 1.981040711234111e-05, + "loss": 0.8697, + "step": 3521 + }, + { + "epoch": 0.1559390323544713, + "grad_norm": 1.2366293668746948, + "learning_rate": 1.981010748407211e-05, + "loss": 0.9478, + "step": 3522 + }, + { + "epoch": 0.15598330805928518, + "grad_norm": 2.2302069664001465, + "learning_rate": 1.9809807621496676e-05, + "loss": 1.0478, + "step": 3523 + }, + { + "epoch": 0.15602758376409903, + "grad_norm": 1.4380892515182495, + "learning_rate": 1.980950752462198e-05, + "loss": 0.9576, + "step": 3524 + }, + { + "epoch": 0.15607185946891292, + "grad_norm": 2.5350918769836426, + "learning_rate": 1.980920719345518e-05, + "loss": 1.1049, + "step": 3525 + }, + { + "epoch": 0.1561161351737268, + "grad_norm": 1.4006221294403076, + "learning_rate": 1.9808906628003453e-05, + "loss": 1.0098, + "step": 3526 + }, + { + "epoch": 0.15616041087854068, + "grad_norm": 1.3108618259429932, + "learning_rate": 1.980860582827398e-05, + "loss": 0.8776, + "step": 3527 + }, + { + "epoch": 0.15620468658335454, + "grad_norm": 1.3947298526763916, + "learning_rate": 1.980830479427394e-05, + "loss": 1.0003, + "step": 3528 + }, + { + "epoch": 0.15624896228816842, + "grad_norm": 1.6251709461212158, + "learning_rate": 1.980800352601053e-05, + "loss": 1.1946, + "step": 3529 + }, + { + "epoch": 0.1562932379929823, + "grad_norm": 1.286418080329895, + "learning_rate": 1.9807702023490936e-05, + "loss": 0.872, + "step": 3530 + }, + { + "epoch": 0.15633751369779617, + "grad_norm": 1.3144131898880005, + "learning_rate": 1.980740028672237e-05, + "loss": 0.9476, + "step": 3531 + }, + { + "epoch": 0.15638178940261005, + "grad_norm": 1.452924132347107, + "learning_rate": 1.9807098315712036e-05, + "loss": 0.9257, + "step": 3532 + }, + { + "epoch": 0.15642606510742393, + "grad_norm": 1.2459052801132202, + "learning_rate": 1.980679611046714e-05, + "loss": 0.9529, + "step": 3533 + }, + { + "epoch": 0.15647034081223782, + "grad_norm": 1.4888513088226318, + "learning_rate": 1.98064936709949e-05, + "loss": 1.0226, + "step": 3534 + }, + { + "epoch": 0.15651461651705167, + "grad_norm": 1.3812158107757568, + "learning_rate": 1.980619099730255e-05, + "loss": 0.9514, + "step": 3535 + }, + { + "epoch": 0.15655889222186556, + "grad_norm": 1.836601734161377, + "learning_rate": 1.980588808939731e-05, + "loss": 1.1129, + "step": 3536 + }, + { + "epoch": 0.15660316792667944, + "grad_norm": 1.353006362915039, + "learning_rate": 1.980558494728642e-05, + "loss": 0.9806, + "step": 3537 + }, + { + "epoch": 0.1566474436314933, + "grad_norm": 26.90633773803711, + "learning_rate": 1.9805281570977117e-05, + "loss": 0.9912, + "step": 3538 + }, + { + "epoch": 0.15669171933630718, + "grad_norm": 1.5277961492538452, + "learning_rate": 1.9804977960476644e-05, + "loss": 0.9399, + "step": 3539 + }, + { + "epoch": 0.15673599504112107, + "grad_norm": 1.292564034461975, + "learning_rate": 1.9804674115792256e-05, + "loss": 0.9499, + "step": 3540 + }, + { + "epoch": 0.15678027074593492, + "grad_norm": 1.431421160697937, + "learning_rate": 1.9804370036931213e-05, + "loss": 1.0548, + "step": 3541 + }, + { + "epoch": 0.1568245464507488, + "grad_norm": 1.6019341945648193, + "learning_rate": 1.980406572390077e-05, + "loss": 0.9937, + "step": 3542 + }, + { + "epoch": 0.1568688221555627, + "grad_norm": 1.3568652868270874, + "learning_rate": 1.9803761176708205e-05, + "loss": 0.9887, + "step": 3543 + }, + { + "epoch": 0.15691309786037658, + "grad_norm": 1.3797261714935303, + "learning_rate": 1.9803456395360784e-05, + "loss": 0.955, + "step": 3544 + }, + { + "epoch": 0.15695737356519043, + "grad_norm": 1.4186615943908691, + "learning_rate": 1.9803151379865793e-05, + "loss": 1.0545, + "step": 3545 + }, + { + "epoch": 0.15700164927000432, + "grad_norm": 1.4574295282363892, + "learning_rate": 1.9802846130230508e-05, + "loss": 1.0598, + "step": 3546 + }, + { + "epoch": 0.1570459249748182, + "grad_norm": 2.620474338531494, + "learning_rate": 1.980254064646223e-05, + "loss": 1.0589, + "step": 3547 + }, + { + "epoch": 0.15709020067963206, + "grad_norm": 1.5303926467895508, + "learning_rate": 1.980223492856825e-05, + "loss": 1.1053, + "step": 3548 + }, + { + "epoch": 0.15713447638444594, + "grad_norm": 1.306378722190857, + "learning_rate": 1.9801928976555863e-05, + "loss": 1.0551, + "step": 3549 + }, + { + "epoch": 0.15717875208925983, + "grad_norm": 1.2312347888946533, + "learning_rate": 1.980162279043239e-05, + "loss": 0.9338, + "step": 3550 + }, + { + "epoch": 0.15722302779407368, + "grad_norm": 1.4535455703735352, + "learning_rate": 1.9801316370205135e-05, + "loss": 0.9771, + "step": 3551 + }, + { + "epoch": 0.15726730349888757, + "grad_norm": 1.3279516696929932, + "learning_rate": 1.980100971588142e-05, + "loss": 0.965, + "step": 3552 + }, + { + "epoch": 0.15731157920370145, + "grad_norm": 1.534794807434082, + "learning_rate": 1.9800702827468567e-05, + "loss": 1.0636, + "step": 3553 + }, + { + "epoch": 0.15735585490851534, + "grad_norm": 1.2935317754745483, + "learning_rate": 1.9800395704973908e-05, + "loss": 0.9445, + "step": 3554 + }, + { + "epoch": 0.1574001306133292, + "grad_norm": 1.3478766679763794, + "learning_rate": 1.9800088348404778e-05, + "loss": 0.9624, + "step": 3555 + }, + { + "epoch": 0.15744440631814308, + "grad_norm": 1.3511642217636108, + "learning_rate": 1.9799780757768517e-05, + "loss": 0.8944, + "step": 3556 + }, + { + "epoch": 0.15748868202295696, + "grad_norm": 1.3152540922164917, + "learning_rate": 1.979947293307247e-05, + "loss": 0.957, + "step": 3557 + }, + { + "epoch": 0.15753295772777082, + "grad_norm": 1.2922983169555664, + "learning_rate": 1.9799164874323993e-05, + "loss": 0.9708, + "step": 3558 + }, + { + "epoch": 0.1575772334325847, + "grad_norm": 1.286444902420044, + "learning_rate": 1.979885658153044e-05, + "loss": 0.9321, + "step": 3559 + }, + { + "epoch": 0.15762150913739859, + "grad_norm": 1.3511890172958374, + "learning_rate": 1.9798548054699177e-05, + "loss": 1.0029, + "step": 3560 + }, + { + "epoch": 0.15766578484221247, + "grad_norm": 1.1937124729156494, + "learning_rate": 1.9798239293837572e-05, + "loss": 0.9233, + "step": 3561 + }, + { + "epoch": 0.15771006054702633, + "grad_norm": 1.3276883363723755, + "learning_rate": 1.9797930298953002e-05, + "loss": 1.0676, + "step": 3562 + }, + { + "epoch": 0.1577543362518402, + "grad_norm": 1.3365267515182495, + "learning_rate": 1.9797621070052843e-05, + "loss": 0.9439, + "step": 3563 + }, + { + "epoch": 0.1577986119566541, + "grad_norm": 1.5902762413024902, + "learning_rate": 1.979731160714448e-05, + "loss": 0.9826, + "step": 3564 + }, + { + "epoch": 0.15784288766146795, + "grad_norm": 1.4557735919952393, + "learning_rate": 1.9797001910235306e-05, + "loss": 0.9361, + "step": 3565 + }, + { + "epoch": 0.15788716336628184, + "grad_norm": 1.677169919013977, + "learning_rate": 1.979669197933272e-05, + "loss": 1.0062, + "step": 3566 + }, + { + "epoch": 0.15793143907109572, + "grad_norm": 1.5147241353988647, + "learning_rate": 1.9796381814444124e-05, + "loss": 1.0034, + "step": 3567 + }, + { + "epoch": 0.15797571477590958, + "grad_norm": 1.4518764019012451, + "learning_rate": 1.9796071415576925e-05, + "loss": 0.9021, + "step": 3568 + }, + { + "epoch": 0.15801999048072346, + "grad_norm": 1.377336025238037, + "learning_rate": 1.9795760782738537e-05, + "loss": 0.992, + "step": 3569 + }, + { + "epoch": 0.15806426618553734, + "grad_norm": 1.322102427482605, + "learning_rate": 1.9795449915936378e-05, + "loss": 0.974, + "step": 3570 + }, + { + "epoch": 0.15810854189035123, + "grad_norm": 1.2625951766967773, + "learning_rate": 1.979513881517787e-05, + "loss": 0.9227, + "step": 3571 + }, + { + "epoch": 0.15815281759516508, + "grad_norm": 1.6371444463729858, + "learning_rate": 1.979482748047045e-05, + "loss": 1.028, + "step": 3572 + }, + { + "epoch": 0.15819709329997897, + "grad_norm": 1.2005677223205566, + "learning_rate": 1.979451591182155e-05, + "loss": 0.9366, + "step": 3573 + }, + { + "epoch": 0.15824136900479285, + "grad_norm": 1.5777026414871216, + "learning_rate": 1.9794204109238616e-05, + "loss": 0.952, + "step": 3574 + }, + { + "epoch": 0.1582856447096067, + "grad_norm": 1.3461027145385742, + "learning_rate": 1.9793892072729087e-05, + "loss": 0.9531, + "step": 3575 + }, + { + "epoch": 0.1583299204144206, + "grad_norm": 1.2640260457992554, + "learning_rate": 1.9793579802300425e-05, + "loss": 0.9557, + "step": 3576 + }, + { + "epoch": 0.15837419611923448, + "grad_norm": 1.3940553665161133, + "learning_rate": 1.9793267297960077e-05, + "loss": 1.006, + "step": 3577 + }, + { + "epoch": 0.15841847182404836, + "grad_norm": 1.2993803024291992, + "learning_rate": 1.979295455971552e-05, + "loss": 0.9515, + "step": 3578 + }, + { + "epoch": 0.15846274752886222, + "grad_norm": 1.322366714477539, + "learning_rate": 1.9792641587574212e-05, + "loss": 1.005, + "step": 3579 + }, + { + "epoch": 0.1585070232336761, + "grad_norm": 1.2910206317901611, + "learning_rate": 1.9792328381543636e-05, + "loss": 0.9521, + "step": 3580 + }, + { + "epoch": 0.15855129893849, + "grad_norm": 1.467437505722046, + "learning_rate": 1.979201494163127e-05, + "loss": 0.9858, + "step": 3581 + }, + { + "epoch": 0.15859557464330384, + "grad_norm": 1.3658877611160278, + "learning_rate": 1.97917012678446e-05, + "loss": 0.971, + "step": 3582 + }, + { + "epoch": 0.15863985034811773, + "grad_norm": 1.3819259405136108, + "learning_rate": 1.9791387360191118e-05, + "loss": 0.9829, + "step": 3583 + }, + { + "epoch": 0.1586841260529316, + "grad_norm": 1.1691044569015503, + "learning_rate": 1.9791073218678317e-05, + "loss": 0.9319, + "step": 3584 + }, + { + "epoch": 0.15872840175774547, + "grad_norm": 1.3969675302505493, + "learning_rate": 1.979075884331371e-05, + "loss": 0.9663, + "step": 3585 + }, + { + "epoch": 0.15877267746255935, + "grad_norm": 1.866310477256775, + "learning_rate": 1.97904442341048e-05, + "loss": 1.0133, + "step": 3586 + }, + { + "epoch": 0.15881695316737324, + "grad_norm": 1.4385560750961304, + "learning_rate": 1.9790129391059098e-05, + "loss": 1.0101, + "step": 3587 + }, + { + "epoch": 0.15886122887218712, + "grad_norm": 1.5650750398635864, + "learning_rate": 1.9789814314184127e-05, + "loss": 0.9676, + "step": 3588 + }, + { + "epoch": 0.15890550457700098, + "grad_norm": 1.2474285364151, + "learning_rate": 1.9789499003487412e-05, + "loss": 1.0251, + "step": 3589 + }, + { + "epoch": 0.15894978028181486, + "grad_norm": 1.4494409561157227, + "learning_rate": 1.9789183458976485e-05, + "loss": 1.0551, + "step": 3590 + }, + { + "epoch": 0.15899405598662875, + "grad_norm": 1.3382316827774048, + "learning_rate": 1.9788867680658884e-05, + "loss": 0.9923, + "step": 3591 + }, + { + "epoch": 0.1590383316914426, + "grad_norm": 1.2720379829406738, + "learning_rate": 1.9788551668542148e-05, + "loss": 0.9039, + "step": 3592 + }, + { + "epoch": 0.1590826073962565, + "grad_norm": 1.6834032535552979, + "learning_rate": 1.9788235422633822e-05, + "loss": 0.992, + "step": 3593 + }, + { + "epoch": 0.15912688310107037, + "grad_norm": 1.4139119386672974, + "learning_rate": 1.9787918942941468e-05, + "loss": 1.0087, + "step": 3594 + }, + { + "epoch": 0.15917115880588423, + "grad_norm": 1.3996572494506836, + "learning_rate": 1.9787602229472636e-05, + "loss": 0.9459, + "step": 3595 + }, + { + "epoch": 0.1592154345106981, + "grad_norm": 1.860203504562378, + "learning_rate": 1.9787285282234893e-05, + "loss": 0.9893, + "step": 3596 + }, + { + "epoch": 0.159259710215512, + "grad_norm": 1.488349199295044, + "learning_rate": 1.9786968101235814e-05, + "loss": 1.018, + "step": 3597 + }, + { + "epoch": 0.15930398592032588, + "grad_norm": 1.6749427318572998, + "learning_rate": 1.978665068648297e-05, + "loss": 1.0565, + "step": 3598 + }, + { + "epoch": 0.15934826162513974, + "grad_norm": 1.331839680671692, + "learning_rate": 1.978633303798394e-05, + "loss": 0.9237, + "step": 3599 + }, + { + "epoch": 0.15939253732995362, + "grad_norm": 1.243267297744751, + "learning_rate": 1.9786015155746317e-05, + "loss": 0.9411, + "step": 3600 + }, + { + "epoch": 0.1594368130347675, + "grad_norm": 1.378248929977417, + "learning_rate": 1.9785697039777688e-05, + "loss": 1.0378, + "step": 3601 + }, + { + "epoch": 0.15948108873958136, + "grad_norm": 1.470024824142456, + "learning_rate": 1.9785378690085655e-05, + "loss": 0.9313, + "step": 3602 + }, + { + "epoch": 0.15952536444439525, + "grad_norm": 1.5375343561172485, + "learning_rate": 1.9785060106677818e-05, + "loss": 1.0877, + "step": 3603 + }, + { + "epoch": 0.15956964014920913, + "grad_norm": 1.631994366645813, + "learning_rate": 1.9784741289561787e-05, + "loss": 1.0536, + "step": 3604 + }, + { + "epoch": 0.15961391585402301, + "grad_norm": 1.5051664113998413, + "learning_rate": 1.9784422238745177e-05, + "loss": 0.9739, + "step": 3605 + }, + { + "epoch": 0.15965819155883687, + "grad_norm": 1.2952591180801392, + "learning_rate": 1.978410295423561e-05, + "loss": 0.9272, + "step": 3606 + }, + { + "epoch": 0.15970246726365075, + "grad_norm": 1.670100450515747, + "learning_rate": 1.978378343604071e-05, + "loss": 1.0311, + "step": 3607 + }, + { + "epoch": 0.15974674296846464, + "grad_norm": 1.2392890453338623, + "learning_rate": 1.9783463684168112e-05, + "loss": 0.9711, + "step": 3608 + }, + { + "epoch": 0.1597910186732785, + "grad_norm": 1.1558740139007568, + "learning_rate": 1.9783143698625444e-05, + "loss": 0.9743, + "step": 3609 + }, + { + "epoch": 0.15983529437809238, + "grad_norm": 1.4122066497802734, + "learning_rate": 1.978282347942036e-05, + "loss": 0.9458, + "step": 3610 + }, + { + "epoch": 0.15987957008290626, + "grad_norm": 1.484259843826294, + "learning_rate": 1.9782503026560497e-05, + "loss": 0.9582, + "step": 3611 + }, + { + "epoch": 0.15992384578772012, + "grad_norm": 1.571476697921753, + "learning_rate": 1.978218234005352e-05, + "loss": 0.9853, + "step": 3612 + }, + { + "epoch": 0.159968121492534, + "grad_norm": 1.6726505756378174, + "learning_rate": 1.978186141990708e-05, + "loss": 0.9536, + "step": 3613 + }, + { + "epoch": 0.1600123971973479, + "grad_norm": 1.3649170398712158, + "learning_rate": 1.9781540266128844e-05, + "loss": 0.9933, + "step": 3614 + }, + { + "epoch": 0.16005667290216177, + "grad_norm": 1.3755007982254028, + "learning_rate": 1.9781218878726486e-05, + "loss": 1.0436, + "step": 3615 + }, + { + "epoch": 0.16010094860697563, + "grad_norm": 1.1838338375091553, + "learning_rate": 1.9780897257707678e-05, + "loss": 0.9553, + "step": 3616 + }, + { + "epoch": 0.1601452243117895, + "grad_norm": 1.4022138118743896, + "learning_rate": 1.9780575403080103e-05, + "loss": 0.9676, + "step": 3617 + }, + { + "epoch": 0.1601895000166034, + "grad_norm": 1.267946720123291, + "learning_rate": 1.9780253314851446e-05, + "loss": 0.9997, + "step": 3618 + }, + { + "epoch": 0.16023377572141725, + "grad_norm": 1.786298155784607, + "learning_rate": 1.9779930993029405e-05, + "loss": 1.0582, + "step": 3619 + }, + { + "epoch": 0.16027805142623114, + "grad_norm": 1.5116819143295288, + "learning_rate": 1.9779608437621674e-05, + "loss": 1.0618, + "step": 3620 + }, + { + "epoch": 0.16032232713104502, + "grad_norm": 1.3394757509231567, + "learning_rate": 1.9779285648635955e-05, + "loss": 0.9182, + "step": 3621 + }, + { + "epoch": 0.1603666028358589, + "grad_norm": 1.2252978086471558, + "learning_rate": 1.977896262607997e-05, + "loss": 0.9621, + "step": 3622 + }, + { + "epoch": 0.16041087854067276, + "grad_norm": 1.2944284677505493, + "learning_rate": 1.9778639369961414e-05, + "loss": 0.9408, + "step": 3623 + }, + { + "epoch": 0.16045515424548665, + "grad_norm": 1.3212810754776, + "learning_rate": 1.9778315880288024e-05, + "loss": 0.9984, + "step": 3624 + }, + { + "epoch": 0.16049942995030053, + "grad_norm": 1.229660153388977, + "learning_rate": 1.9777992157067523e-05, + "loss": 0.9504, + "step": 3625 + }, + { + "epoch": 0.1605437056551144, + "grad_norm": 1.4390956163406372, + "learning_rate": 1.9777668200307637e-05, + "loss": 0.9282, + "step": 3626 + }, + { + "epoch": 0.16058798135992827, + "grad_norm": 1.2254194021224976, + "learning_rate": 1.977734401001611e-05, + "loss": 0.9244, + "step": 3627 + }, + { + "epoch": 0.16063225706474216, + "grad_norm": 1.5931968688964844, + "learning_rate": 1.977701958620068e-05, + "loss": 0.9752, + "step": 3628 + }, + { + "epoch": 0.160676532769556, + "grad_norm": 1.308274507522583, + "learning_rate": 1.9776694928869097e-05, + "loss": 0.9306, + "step": 3629 + }, + { + "epoch": 0.1607208084743699, + "grad_norm": 1.2482022047042847, + "learning_rate": 1.977637003802912e-05, + "loss": 0.9548, + "step": 3630 + }, + { + "epoch": 0.16076508417918378, + "grad_norm": 1.2512905597686768, + "learning_rate": 1.9776044913688503e-05, + "loss": 0.9707, + "step": 3631 + }, + { + "epoch": 0.16080935988399767, + "grad_norm": 1.448624849319458, + "learning_rate": 1.9775719555855014e-05, + "loss": 0.9822, + "step": 3632 + }, + { + "epoch": 0.16085363558881152, + "grad_norm": 1.2715221643447876, + "learning_rate": 1.9775393964536422e-05, + "loss": 0.9661, + "step": 3633 + }, + { + "epoch": 0.1608979112936254, + "grad_norm": 1.2989686727523804, + "learning_rate": 1.9775068139740506e-05, + "loss": 0.8945, + "step": 3634 + }, + { + "epoch": 0.1609421869984393, + "grad_norm": 1.7821307182312012, + "learning_rate": 1.9774742081475046e-05, + "loss": 1.0196, + "step": 3635 + }, + { + "epoch": 0.16098646270325315, + "grad_norm": 1.3674408197402954, + "learning_rate": 1.977441578974783e-05, + "loss": 0.9582, + "step": 3636 + }, + { + "epoch": 0.16103073840806703, + "grad_norm": 2.6759848594665527, + "learning_rate": 1.977408926456665e-05, + "loss": 1.0003, + "step": 3637 + }, + { + "epoch": 0.16107501411288092, + "grad_norm": 1.2681334018707275, + "learning_rate": 1.9773762505939308e-05, + "loss": 0.8691, + "step": 3638 + }, + { + "epoch": 0.16111928981769477, + "grad_norm": 1.5753437280654907, + "learning_rate": 1.9773435513873607e-05, + "loss": 1.0558, + "step": 3639 + }, + { + "epoch": 0.16116356552250866, + "grad_norm": 1.263121247291565, + "learning_rate": 1.9773108288377355e-05, + "loss": 0.9261, + "step": 3640 + }, + { + "epoch": 0.16120784122732254, + "grad_norm": 1.3922456502914429, + "learning_rate": 1.9772780829458372e-05, + "loss": 0.9492, + "step": 3641 + }, + { + "epoch": 0.16125211693213642, + "grad_norm": 1.3582571744918823, + "learning_rate": 1.9772453137124472e-05, + "loss": 1.0467, + "step": 3642 + }, + { + "epoch": 0.16129639263695028, + "grad_norm": 1.3474857807159424, + "learning_rate": 1.9772125211383487e-05, + "loss": 1.0063, + "step": 3643 + }, + { + "epoch": 0.16134066834176417, + "grad_norm": 1.490389347076416, + "learning_rate": 1.9771797052243252e-05, + "loss": 1.032, + "step": 3644 + }, + { + "epoch": 0.16138494404657805, + "grad_norm": 1.247940182685852, + "learning_rate": 1.9771468659711595e-05, + "loss": 0.9123, + "step": 3645 + }, + { + "epoch": 0.1614292197513919, + "grad_norm": 2.195439338684082, + "learning_rate": 1.977114003379637e-05, + "loss": 1.1915, + "step": 3646 + }, + { + "epoch": 0.1614734954562058, + "grad_norm": 1.2186819314956665, + "learning_rate": 1.977081117450542e-05, + "loss": 0.9233, + "step": 3647 + }, + { + "epoch": 0.16151777116101967, + "grad_norm": 1.3175151348114014, + "learning_rate": 1.97704820818466e-05, + "loss": 0.9562, + "step": 3648 + }, + { + "epoch": 0.16156204686583356, + "grad_norm": 1.4395594596862793, + "learning_rate": 1.977015275582777e-05, + "loss": 0.9645, + "step": 3649 + }, + { + "epoch": 0.16160632257064741, + "grad_norm": 1.4476754665374756, + "learning_rate": 1.9769823196456798e-05, + "loss": 1.0075, + "step": 3650 + }, + { + "epoch": 0.1616505982754613, + "grad_norm": 2.263803720474243, + "learning_rate": 1.9769493403741556e-05, + "loss": 1.0494, + "step": 3651 + }, + { + "epoch": 0.16169487398027518, + "grad_norm": 1.4577858448028564, + "learning_rate": 1.9769163377689913e-05, + "loss": 0.9643, + "step": 3652 + }, + { + "epoch": 0.16173914968508904, + "grad_norm": 1.5138251781463623, + "learning_rate": 1.9768833118309764e-05, + "loss": 0.9686, + "step": 3653 + }, + { + "epoch": 0.16178342538990292, + "grad_norm": 1.3445504903793335, + "learning_rate": 1.9768502625608985e-05, + "loss": 0.997, + "step": 3654 + }, + { + "epoch": 0.1618277010947168, + "grad_norm": 1.3871912956237793, + "learning_rate": 1.9768171899595477e-05, + "loss": 0.9743, + "step": 3655 + }, + { + "epoch": 0.16187197679953066, + "grad_norm": 2.0268616676330566, + "learning_rate": 1.9767840940277133e-05, + "loss": 1.0037, + "step": 3656 + }, + { + "epoch": 0.16191625250434455, + "grad_norm": 1.2035211324691772, + "learning_rate": 1.9767509747661866e-05, + "loss": 0.8526, + "step": 3657 + }, + { + "epoch": 0.16196052820915843, + "grad_norm": 1.8210363388061523, + "learning_rate": 1.976717832175758e-05, + "loss": 1.015, + "step": 3658 + }, + { + "epoch": 0.16200480391397232, + "grad_norm": 1.29507577419281, + "learning_rate": 1.976684666257219e-05, + "loss": 0.9645, + "step": 3659 + }, + { + "epoch": 0.16204907961878617, + "grad_norm": 1.356935977935791, + "learning_rate": 1.9766514770113624e-05, + "loss": 1.0188, + "step": 3660 + }, + { + "epoch": 0.16209335532360006, + "grad_norm": 1.2290494441986084, + "learning_rate": 1.9766182644389803e-05, + "loss": 0.9673, + "step": 3661 + }, + { + "epoch": 0.16213763102841394, + "grad_norm": 1.191983699798584, + "learning_rate": 1.9765850285408663e-05, + "loss": 0.9436, + "step": 3662 + }, + { + "epoch": 0.1621819067332278, + "grad_norm": 1.31855046749115, + "learning_rate": 1.976551769317814e-05, + "loss": 0.9608, + "step": 3663 + }, + { + "epoch": 0.16222618243804168, + "grad_norm": 1.3621060848236084, + "learning_rate": 1.9765184867706177e-05, + "loss": 0.9565, + "step": 3664 + }, + { + "epoch": 0.16227045814285557, + "grad_norm": 1.1961873769760132, + "learning_rate": 1.976485180900072e-05, + "loss": 0.922, + "step": 3665 + }, + { + "epoch": 0.16231473384766945, + "grad_norm": 1.3639373779296875, + "learning_rate": 1.9764518517069737e-05, + "loss": 0.9863, + "step": 3666 + }, + { + "epoch": 0.1623590095524833, + "grad_norm": 1.3519928455352783, + "learning_rate": 1.9764184991921178e-05, + "loss": 0.9815, + "step": 3667 + }, + { + "epoch": 0.1624032852572972, + "grad_norm": 1.51594877243042, + "learning_rate": 1.9763851233563008e-05, + "loss": 1.0728, + "step": 3668 + }, + { + "epoch": 0.16244756096211108, + "grad_norm": 3.2801246643066406, + "learning_rate": 1.97635172420032e-05, + "loss": 0.9402, + "step": 3669 + }, + { + "epoch": 0.16249183666692493, + "grad_norm": 1.319985270500183, + "learning_rate": 1.9763183017249735e-05, + "loss": 0.9848, + "step": 3670 + }, + { + "epoch": 0.16253611237173882, + "grad_norm": 1.3260297775268555, + "learning_rate": 1.9762848559310595e-05, + "loss": 0.923, + "step": 3671 + }, + { + "epoch": 0.1625803880765527, + "grad_norm": 1.57757568359375, + "learning_rate": 1.9762513868193763e-05, + "loss": 0.896, + "step": 3672 + }, + { + "epoch": 0.16262466378136656, + "grad_norm": 1.4346979856491089, + "learning_rate": 1.976217894390724e-05, + "loss": 0.9795, + "step": 3673 + }, + { + "epoch": 0.16266893948618044, + "grad_norm": 1.4399349689483643, + "learning_rate": 1.9761843786459018e-05, + "loss": 0.9175, + "step": 3674 + }, + { + "epoch": 0.16271321519099433, + "grad_norm": 1.2782118320465088, + "learning_rate": 1.9761508395857106e-05, + "loss": 0.9031, + "step": 3675 + }, + { + "epoch": 0.1627574908958082, + "grad_norm": 1.4427343606948853, + "learning_rate": 1.9761172772109515e-05, + "loss": 0.9549, + "step": 3676 + }, + { + "epoch": 0.16280176660062207, + "grad_norm": 1.3620644807815552, + "learning_rate": 1.976083691522426e-05, + "loss": 0.9734, + "step": 3677 + }, + { + "epoch": 0.16284604230543595, + "grad_norm": 1.1784497499465942, + "learning_rate": 1.9760500825209362e-05, + "loss": 0.9474, + "step": 3678 + }, + { + "epoch": 0.16289031801024983, + "grad_norm": 1.4410253763198853, + "learning_rate": 1.976016450207285e-05, + "loss": 0.916, + "step": 3679 + }, + { + "epoch": 0.1629345937150637, + "grad_norm": 1.611216425895691, + "learning_rate": 1.9759827945822753e-05, + "loss": 1.0329, + "step": 3680 + }, + { + "epoch": 0.16297886941987758, + "grad_norm": 1.225513219833374, + "learning_rate": 1.9759491156467116e-05, + "loss": 0.9315, + "step": 3681 + }, + { + "epoch": 0.16302314512469146, + "grad_norm": 1.461800217628479, + "learning_rate": 1.9759154134013978e-05, + "loss": 1.0255, + "step": 3682 + }, + { + "epoch": 0.16306742082950532, + "grad_norm": 1.2691916227340698, + "learning_rate": 1.9758816878471393e-05, + "loss": 1.0044, + "step": 3683 + }, + { + "epoch": 0.1631116965343192, + "grad_norm": 1.4455677270889282, + "learning_rate": 1.975847938984741e-05, + "loss": 1.0768, + "step": 3684 + }, + { + "epoch": 0.16315597223913308, + "grad_norm": 1.2754591703414917, + "learning_rate": 1.9758141668150087e-05, + "loss": 0.9431, + "step": 3685 + }, + { + "epoch": 0.16320024794394697, + "grad_norm": 1.2773014307022095, + "learning_rate": 1.9757803713387503e-05, + "loss": 0.9139, + "step": 3686 + }, + { + "epoch": 0.16324452364876083, + "grad_norm": 1.3463026285171509, + "learning_rate": 1.975746552556772e-05, + "loss": 0.9797, + "step": 3687 + }, + { + "epoch": 0.1632887993535747, + "grad_norm": 1.3604570627212524, + "learning_rate": 1.975712710469882e-05, + "loss": 1.0144, + "step": 3688 + }, + { + "epoch": 0.1633330750583886, + "grad_norm": 1.1803772449493408, + "learning_rate": 1.975678845078888e-05, + "loss": 0.9589, + "step": 3689 + }, + { + "epoch": 0.16337735076320245, + "grad_norm": 1.1443294286727905, + "learning_rate": 1.975644956384599e-05, + "loss": 0.8869, + "step": 3690 + }, + { + "epoch": 0.16342162646801633, + "grad_norm": 1.477099061012268, + "learning_rate": 1.9756110443878254e-05, + "loss": 0.908, + "step": 3691 + }, + { + "epoch": 0.16346590217283022, + "grad_norm": 1.455715537071228, + "learning_rate": 1.975577109089376e-05, + "loss": 1.0079, + "step": 3692 + }, + { + "epoch": 0.1635101778776441, + "grad_norm": 1.299608588218689, + "learning_rate": 1.9755431504900614e-05, + "loss": 0.8633, + "step": 3693 + }, + { + "epoch": 0.16355445358245796, + "grad_norm": 1.4202367067337036, + "learning_rate": 1.9755091685906934e-05, + "loss": 0.9117, + "step": 3694 + }, + { + "epoch": 0.16359872928727184, + "grad_norm": 1.410088062286377, + "learning_rate": 1.975475163392083e-05, + "loss": 1.0006, + "step": 3695 + }, + { + "epoch": 0.16364300499208573, + "grad_norm": 1.5187536478042603, + "learning_rate": 1.9754411348950424e-05, + "loss": 0.99, + "step": 3696 + }, + { + "epoch": 0.16368728069689958, + "grad_norm": 1.6049842834472656, + "learning_rate": 1.9754070831003847e-05, + "loss": 1.083, + "step": 3697 + }, + { + "epoch": 0.16373155640171347, + "grad_norm": 1.6087549924850464, + "learning_rate": 1.975373008008923e-05, + "loss": 0.9299, + "step": 3698 + }, + { + "epoch": 0.16377583210652735, + "grad_norm": 1.193049430847168, + "learning_rate": 1.9753389096214716e-05, + "loss": 0.9147, + "step": 3699 + }, + { + "epoch": 0.1638201078113412, + "grad_norm": 1.6158722639083862, + "learning_rate": 1.9753047879388438e-05, + "loss": 1.0326, + "step": 3700 + }, + { + "epoch": 0.1638643835161551, + "grad_norm": 1.2631639242172241, + "learning_rate": 1.9752706429618554e-05, + "loss": 0.995, + "step": 3701 + }, + { + "epoch": 0.16390865922096898, + "grad_norm": 1.2549920082092285, + "learning_rate": 1.975236474691322e-05, + "loss": 0.9486, + "step": 3702 + }, + { + "epoch": 0.16395293492578286, + "grad_norm": 1.2387635707855225, + "learning_rate": 1.9752022831280598e-05, + "loss": 0.8734, + "step": 3703 + }, + { + "epoch": 0.16399721063059672, + "grad_norm": 1.3848333358764648, + "learning_rate": 1.9751680682728845e-05, + "loss": 0.9433, + "step": 3704 + }, + { + "epoch": 0.1640414863354106, + "grad_norm": 1.3360438346862793, + "learning_rate": 1.9751338301266142e-05, + "loss": 1.016, + "step": 3705 + }, + { + "epoch": 0.1640857620402245, + "grad_norm": 1.2609716653823853, + "learning_rate": 1.9750995686900665e-05, + "loss": 0.9677, + "step": 3706 + }, + { + "epoch": 0.16413003774503834, + "grad_norm": 1.267063856124878, + "learning_rate": 1.9750652839640592e-05, + "loss": 0.9872, + "step": 3707 + }, + { + "epoch": 0.16417431344985223, + "grad_norm": 1.2024636268615723, + "learning_rate": 1.9750309759494116e-05, + "loss": 0.9144, + "step": 3708 + }, + { + "epoch": 0.1642185891546661, + "grad_norm": 1.402268648147583, + "learning_rate": 1.974996644646943e-05, + "loss": 0.9278, + "step": 3709 + }, + { + "epoch": 0.16426286485948, + "grad_norm": 1.7092335224151611, + "learning_rate": 1.9749622900574735e-05, + "loss": 0.9752, + "step": 3710 + }, + { + "epoch": 0.16430714056429385, + "grad_norm": 1.686269998550415, + "learning_rate": 1.9749279121818235e-05, + "loss": 1.0158, + "step": 3711 + }, + { + "epoch": 0.16435141626910774, + "grad_norm": 1.3709828853607178, + "learning_rate": 1.9748935110208144e-05, + "loss": 1.0187, + "step": 3712 + }, + { + "epoch": 0.16439569197392162, + "grad_norm": 1.625353455543518, + "learning_rate": 1.9748590865752675e-05, + "loss": 1.0541, + "step": 3713 + }, + { + "epoch": 0.16443996767873548, + "grad_norm": 1.4700891971588135, + "learning_rate": 1.974824638846005e-05, + "loss": 0.9419, + "step": 3714 + }, + { + "epoch": 0.16448424338354936, + "grad_norm": 1.3367762565612793, + "learning_rate": 1.9747901678338496e-05, + "loss": 0.9929, + "step": 3715 + }, + { + "epoch": 0.16452851908836325, + "grad_norm": 1.2438637018203735, + "learning_rate": 1.9747556735396248e-05, + "loss": 1.0069, + "step": 3716 + }, + { + "epoch": 0.1645727947931771, + "grad_norm": 1.3347641229629517, + "learning_rate": 1.9747211559641542e-05, + "loss": 0.9715, + "step": 3717 + }, + { + "epoch": 0.16461707049799099, + "grad_norm": 1.4904946088790894, + "learning_rate": 1.974686615108263e-05, + "loss": 0.9882, + "step": 3718 + }, + { + "epoch": 0.16466134620280487, + "grad_norm": 1.419950246810913, + "learning_rate": 1.9746520509727754e-05, + "loss": 1.0116, + "step": 3719 + }, + { + "epoch": 0.16470562190761875, + "grad_norm": 1.4097691774368286, + "learning_rate": 1.9746174635585168e-05, + "loss": 0.961, + "step": 3720 + }, + { + "epoch": 0.1647498976124326, + "grad_norm": 1.3973675966262817, + "learning_rate": 1.974582852866314e-05, + "loss": 1.0128, + "step": 3721 + }, + { + "epoch": 0.1647941733172465, + "grad_norm": 1.2877700328826904, + "learning_rate": 1.974548218896993e-05, + "loss": 0.918, + "step": 3722 + }, + { + "epoch": 0.16483844902206038, + "grad_norm": 1.6076264381408691, + "learning_rate": 1.974513561651382e-05, + "loss": 1.0753, + "step": 3723 + }, + { + "epoch": 0.16488272472687424, + "grad_norm": 1.3748911619186401, + "learning_rate": 1.9744788811303073e-05, + "loss": 1.0429, + "step": 3724 + }, + { + "epoch": 0.16492700043168812, + "grad_norm": 1.3061274290084839, + "learning_rate": 1.9744441773345982e-05, + "loss": 0.8994, + "step": 3725 + }, + { + "epoch": 0.164971276136502, + "grad_norm": 1.3924262523651123, + "learning_rate": 1.9744094502650835e-05, + "loss": 1.0738, + "step": 3726 + }, + { + "epoch": 0.16501555184131586, + "grad_norm": 1.2081298828125, + "learning_rate": 1.974374699922592e-05, + "loss": 0.8458, + "step": 3727 + }, + { + "epoch": 0.16505982754612974, + "grad_norm": 1.3148553371429443, + "learning_rate": 1.974339926307955e-05, + "loss": 0.8928, + "step": 3728 + }, + { + "epoch": 0.16510410325094363, + "grad_norm": 1.7516052722930908, + "learning_rate": 1.9743051294220015e-05, + "loss": 1.0278, + "step": 3729 + }, + { + "epoch": 0.1651483789557575, + "grad_norm": 1.2743436098098755, + "learning_rate": 1.9742703092655635e-05, + "loss": 0.914, + "step": 3730 + }, + { + "epoch": 0.16519265466057137, + "grad_norm": 1.2131309509277344, + "learning_rate": 1.9742354658394726e-05, + "loss": 0.8766, + "step": 3731 + }, + { + "epoch": 0.16523693036538525, + "grad_norm": 1.4554469585418701, + "learning_rate": 1.9742005991445606e-05, + "loss": 0.9285, + "step": 3732 + }, + { + "epoch": 0.16528120607019914, + "grad_norm": 1.235542893409729, + "learning_rate": 1.9741657091816603e-05, + "loss": 0.9459, + "step": 3733 + }, + { + "epoch": 0.165325481775013, + "grad_norm": 1.2652815580368042, + "learning_rate": 1.974130795951606e-05, + "loss": 0.9665, + "step": 3734 + }, + { + "epoch": 0.16536975747982688, + "grad_norm": 1.3190345764160156, + "learning_rate": 1.97409585945523e-05, + "loss": 1.0189, + "step": 3735 + }, + { + "epoch": 0.16541403318464076, + "grad_norm": 1.2881897687911987, + "learning_rate": 1.974060899693368e-05, + "loss": 0.9702, + "step": 3736 + }, + { + "epoch": 0.16545830888945465, + "grad_norm": 1.7729398012161255, + "learning_rate": 1.9740259166668544e-05, + "loss": 1.0216, + "step": 3737 + }, + { + "epoch": 0.1655025845942685, + "grad_norm": 1.2394661903381348, + "learning_rate": 1.9739909103765248e-05, + "loss": 1.0035, + "step": 3738 + }, + { + "epoch": 0.1655468602990824, + "grad_norm": 1.2513068914413452, + "learning_rate": 1.9739558808232155e-05, + "loss": 0.9836, + "step": 3739 + }, + { + "epoch": 0.16559113600389627, + "grad_norm": 1.4755055904388428, + "learning_rate": 1.9739208280077628e-05, + "loss": 1.0283, + "step": 3740 + }, + { + "epoch": 0.16563541170871013, + "grad_norm": 1.220469355583191, + "learning_rate": 1.9738857519310043e-05, + "loss": 1.0077, + "step": 3741 + }, + { + "epoch": 0.165679687413524, + "grad_norm": 1.2950235605239868, + "learning_rate": 1.9738506525937774e-05, + "loss": 0.9273, + "step": 3742 + }, + { + "epoch": 0.1657239631183379, + "grad_norm": 1.4710314273834229, + "learning_rate": 1.9738155299969207e-05, + "loss": 0.9353, + "step": 3743 + }, + { + "epoch": 0.16576823882315175, + "grad_norm": 1.469647765159607, + "learning_rate": 1.9737803841412732e-05, + "loss": 0.9285, + "step": 3744 + }, + { + "epoch": 0.16581251452796564, + "grad_norm": 1.294064998626709, + "learning_rate": 1.973745215027674e-05, + "loss": 0.9109, + "step": 3745 + }, + { + "epoch": 0.16585679023277952, + "grad_norm": 1.4713554382324219, + "learning_rate": 1.973710022656963e-05, + "loss": 0.9873, + "step": 3746 + }, + { + "epoch": 0.1659010659375934, + "grad_norm": 1.2749003171920776, + "learning_rate": 1.9736748070299813e-05, + "loss": 1.0419, + "step": 3747 + }, + { + "epoch": 0.16594534164240726, + "grad_norm": 1.2758641242980957, + "learning_rate": 1.973639568147569e-05, + "loss": 0.8934, + "step": 3748 + }, + { + "epoch": 0.16598961734722115, + "grad_norm": 1.2671713829040527, + "learning_rate": 1.9736043060105688e-05, + "loss": 0.8721, + "step": 3749 + }, + { + "epoch": 0.16603389305203503, + "grad_norm": 1.3222346305847168, + "learning_rate": 1.9735690206198225e-05, + "loss": 0.9292, + "step": 3750 + }, + { + "epoch": 0.1660781687568489, + "grad_norm": 1.324006199836731, + "learning_rate": 1.973533711976173e-05, + "loss": 0.9108, + "step": 3751 + }, + { + "epoch": 0.16612244446166277, + "grad_norm": 1.7625874280929565, + "learning_rate": 1.973498380080463e-05, + "loss": 0.9376, + "step": 3752 + }, + { + "epoch": 0.16616672016647666, + "grad_norm": 1.3067747354507446, + "learning_rate": 1.973463024933537e-05, + "loss": 0.9142, + "step": 3753 + }, + { + "epoch": 0.16621099587129054, + "grad_norm": 1.1289477348327637, + "learning_rate": 1.9734276465362394e-05, + "loss": 0.8829, + "step": 3754 + }, + { + "epoch": 0.1662552715761044, + "grad_norm": 1.5212132930755615, + "learning_rate": 1.973392244889415e-05, + "loss": 0.9372, + "step": 3755 + }, + { + "epoch": 0.16629954728091828, + "grad_norm": 1.2912580966949463, + "learning_rate": 1.9733568199939098e-05, + "loss": 0.996, + "step": 3756 + }, + { + "epoch": 0.16634382298573216, + "grad_norm": 1.8485795259475708, + "learning_rate": 1.973321371850569e-05, + "loss": 1.0592, + "step": 3757 + }, + { + "epoch": 0.16638809869054602, + "grad_norm": 1.8911112546920776, + "learning_rate": 1.9732859004602397e-05, + "loss": 1.0471, + "step": 3758 + }, + { + "epoch": 0.1664323743953599, + "grad_norm": 1.3730894327163696, + "learning_rate": 1.9732504058237695e-05, + "loss": 0.9152, + "step": 3759 + }, + { + "epoch": 0.1664766501001738, + "grad_norm": 1.5680036544799805, + "learning_rate": 1.9732148879420053e-05, + "loss": 1.0192, + "step": 3760 + }, + { + "epoch": 0.16652092580498765, + "grad_norm": 1.3261680603027344, + "learning_rate": 1.9731793468157963e-05, + "loss": 0.9557, + "step": 3761 + }, + { + "epoch": 0.16656520150980153, + "grad_norm": 1.3485441207885742, + "learning_rate": 1.9731437824459912e-05, + "loss": 0.9706, + "step": 3762 + }, + { + "epoch": 0.16660947721461541, + "grad_norm": 1.3557592630386353, + "learning_rate": 1.9731081948334387e-05, + "loss": 0.9398, + "step": 3763 + }, + { + "epoch": 0.1666537529194293, + "grad_norm": 1.1387295722961426, + "learning_rate": 1.9730725839789896e-05, + "loss": 0.8476, + "step": 3764 + }, + { + "epoch": 0.16669802862424316, + "grad_norm": 1.4038902521133423, + "learning_rate": 1.9730369498834943e-05, + "loss": 1.0021, + "step": 3765 + }, + { + "epoch": 0.16674230432905704, + "grad_norm": 1.5035545825958252, + "learning_rate": 1.9730012925478036e-05, + "loss": 1.0725, + "step": 3766 + }, + { + "epoch": 0.16678658003387092, + "grad_norm": 1.2226016521453857, + "learning_rate": 1.972965611972769e-05, + "loss": 0.9452, + "step": 3767 + }, + { + "epoch": 0.16683085573868478, + "grad_norm": 1.5586224794387817, + "learning_rate": 1.9729299081592432e-05, + "loss": 1.0166, + "step": 3768 + }, + { + "epoch": 0.16687513144349866, + "grad_norm": 1.3183491230010986, + "learning_rate": 1.9728941811080785e-05, + "loss": 0.8944, + "step": 3769 + }, + { + "epoch": 0.16691940714831255, + "grad_norm": 1.1562767028808594, + "learning_rate": 1.972858430820129e-05, + "loss": 0.9392, + "step": 3770 + }, + { + "epoch": 0.1669636828531264, + "grad_norm": 1.3683830499649048, + "learning_rate": 1.9728226572962474e-05, + "loss": 0.9385, + "step": 3771 + }, + { + "epoch": 0.1670079585579403, + "grad_norm": 1.3723303079605103, + "learning_rate": 1.972786860537289e-05, + "loss": 1.0404, + "step": 3772 + }, + { + "epoch": 0.16705223426275417, + "grad_norm": 1.2502199411392212, + "learning_rate": 1.9727510405441086e-05, + "loss": 1.0058, + "step": 3773 + }, + { + "epoch": 0.16709650996756806, + "grad_norm": 2.2809934616088867, + "learning_rate": 1.9727151973175617e-05, + "loss": 1.1986, + "step": 3774 + }, + { + "epoch": 0.16714078567238191, + "grad_norm": 1.5710294246673584, + "learning_rate": 1.972679330858504e-05, + "loss": 1.002, + "step": 3775 + }, + { + "epoch": 0.1671850613771958, + "grad_norm": 1.297098994255066, + "learning_rate": 1.9726434411677925e-05, + "loss": 0.9342, + "step": 3776 + }, + { + "epoch": 0.16722933708200968, + "grad_norm": 1.2391130924224854, + "learning_rate": 1.9726075282462847e-05, + "loss": 0.903, + "step": 3777 + }, + { + "epoch": 0.16727361278682354, + "grad_norm": 1.931756854057312, + "learning_rate": 1.9725715920948377e-05, + "loss": 1.089, + "step": 3778 + }, + { + "epoch": 0.16731788849163742, + "grad_norm": 1.2145509719848633, + "learning_rate": 1.97253563271431e-05, + "loss": 0.9002, + "step": 3779 + }, + { + "epoch": 0.1673621641964513, + "grad_norm": 1.2568856477737427, + "learning_rate": 1.9724996501055612e-05, + "loss": 0.9668, + "step": 3780 + }, + { + "epoch": 0.1674064399012652, + "grad_norm": 1.2557073831558228, + "learning_rate": 1.9724636442694495e-05, + "loss": 0.9702, + "step": 3781 + }, + { + "epoch": 0.16745071560607905, + "grad_norm": 1.8396576642990112, + "learning_rate": 1.9724276152068355e-05, + "loss": 0.9704, + "step": 3782 + }, + { + "epoch": 0.16749499131089293, + "grad_norm": 1.3611079454421997, + "learning_rate": 1.9723915629185798e-05, + "loss": 0.9787, + "step": 3783 + }, + { + "epoch": 0.16753926701570682, + "grad_norm": 1.2250328063964844, + "learning_rate": 1.9723554874055436e-05, + "loss": 0.9129, + "step": 3784 + }, + { + "epoch": 0.16758354272052067, + "grad_norm": 1.5922225713729858, + "learning_rate": 1.9723193886685882e-05, + "loss": 1.116, + "step": 3785 + }, + { + "epoch": 0.16762781842533456, + "grad_norm": 1.191126823425293, + "learning_rate": 1.972283266708576e-05, + "loss": 0.9753, + "step": 3786 + }, + { + "epoch": 0.16767209413014844, + "grad_norm": 1.1597234010696411, + "learning_rate": 1.9722471215263693e-05, + "loss": 0.8639, + "step": 3787 + }, + { + "epoch": 0.1677163698349623, + "grad_norm": 1.2461678981781006, + "learning_rate": 1.972210953122832e-05, + "loss": 0.962, + "step": 3788 + }, + { + "epoch": 0.16776064553977618, + "grad_norm": 1.5927062034606934, + "learning_rate": 1.9721747614988275e-05, + "loss": 1.0169, + "step": 3789 + }, + { + "epoch": 0.16780492124459007, + "grad_norm": 1.6115374565124512, + "learning_rate": 1.9721385466552205e-05, + "loss": 1.0866, + "step": 3790 + }, + { + "epoch": 0.16784919694940395, + "grad_norm": 1.468959093093872, + "learning_rate": 1.9721023085928758e-05, + "loss": 1.0103, + "step": 3791 + }, + { + "epoch": 0.1678934726542178, + "grad_norm": 1.261740803718567, + "learning_rate": 1.972066047312659e-05, + "loss": 0.9687, + "step": 3792 + }, + { + "epoch": 0.1679377483590317, + "grad_norm": 1.1544160842895508, + "learning_rate": 1.9720297628154362e-05, + "loss": 0.959, + "step": 3793 + }, + { + "epoch": 0.16798202406384558, + "grad_norm": 1.3713411092758179, + "learning_rate": 1.971993455102074e-05, + "loss": 0.9767, + "step": 3794 + }, + { + "epoch": 0.16802629976865943, + "grad_norm": 1.6781471967697144, + "learning_rate": 1.9719571241734395e-05, + "loss": 0.9927, + "step": 3795 + }, + { + "epoch": 0.16807057547347332, + "grad_norm": 1.3349144458770752, + "learning_rate": 1.9719207700304003e-05, + "loss": 0.9246, + "step": 3796 + }, + { + "epoch": 0.1681148511782872, + "grad_norm": 1.2932219505310059, + "learning_rate": 1.9718843926738252e-05, + "loss": 0.9792, + "step": 3797 + }, + { + "epoch": 0.16815912688310106, + "grad_norm": 1.386002779006958, + "learning_rate": 1.9718479921045826e-05, + "loss": 0.9299, + "step": 3798 + }, + { + "epoch": 0.16820340258791494, + "grad_norm": 1.3641821146011353, + "learning_rate": 1.9718115683235418e-05, + "loss": 0.9142, + "step": 3799 + }, + { + "epoch": 0.16824767829272883, + "grad_norm": 1.3598268032073975, + "learning_rate": 1.9717751213315733e-05, + "loss": 0.9588, + "step": 3800 + }, + { + "epoch": 0.1682919539975427, + "grad_norm": 1.3548097610473633, + "learning_rate": 1.971738651129547e-05, + "loss": 0.9921, + "step": 3801 + }, + { + "epoch": 0.16833622970235657, + "grad_norm": 1.7544450759887695, + "learning_rate": 1.9717021577183343e-05, + "loss": 0.8731, + "step": 3802 + }, + { + "epoch": 0.16838050540717045, + "grad_norm": 1.2801657915115356, + "learning_rate": 1.971665641098807e-05, + "loss": 0.8652, + "step": 3803 + }, + { + "epoch": 0.16842478111198433, + "grad_norm": 1.1495006084442139, + "learning_rate": 1.9716291012718367e-05, + "loss": 0.8716, + "step": 3804 + }, + { + "epoch": 0.1684690568167982, + "grad_norm": 1.3234705924987793, + "learning_rate": 1.9715925382382965e-05, + "loss": 0.9384, + "step": 3805 + }, + { + "epoch": 0.16851333252161207, + "grad_norm": 1.4090112447738647, + "learning_rate": 1.9715559519990598e-05, + "loss": 1.0562, + "step": 3806 + }, + { + "epoch": 0.16855760822642596, + "grad_norm": 1.5572326183319092, + "learning_rate": 1.9715193425550005e-05, + "loss": 1.0329, + "step": 3807 + }, + { + "epoch": 0.16860188393123984, + "grad_norm": 1.4482096433639526, + "learning_rate": 1.971482709906992e-05, + "loss": 0.9036, + "step": 3808 + }, + { + "epoch": 0.1686461596360537, + "grad_norm": 1.312462568283081, + "learning_rate": 1.9714460540559103e-05, + "loss": 0.9011, + "step": 3809 + }, + { + "epoch": 0.16869043534086758, + "grad_norm": 1.2128915786743164, + "learning_rate": 1.9714093750026308e-05, + "loss": 0.8947, + "step": 3810 + }, + { + "epoch": 0.16873471104568147, + "grad_norm": 1.7021663188934326, + "learning_rate": 1.971372672748029e-05, + "loss": 1.0804, + "step": 3811 + }, + { + "epoch": 0.16877898675049532, + "grad_norm": 1.3325035572052002, + "learning_rate": 1.9713359472929822e-05, + "loss": 0.9612, + "step": 3812 + }, + { + "epoch": 0.1688232624553092, + "grad_norm": 1.4775846004486084, + "learning_rate": 1.971299198638367e-05, + "loss": 1.0073, + "step": 3813 + }, + { + "epoch": 0.1688675381601231, + "grad_norm": 1.4807662963867188, + "learning_rate": 1.9712624267850616e-05, + "loss": 0.967, + "step": 3814 + }, + { + "epoch": 0.16891181386493695, + "grad_norm": 1.2838428020477295, + "learning_rate": 1.9712256317339434e-05, + "loss": 0.9484, + "step": 3815 + }, + { + "epoch": 0.16895608956975083, + "grad_norm": 1.3840479850769043, + "learning_rate": 1.971188813485892e-05, + "loss": 0.9689, + "step": 3816 + }, + { + "epoch": 0.16900036527456472, + "grad_norm": 1.2460050582885742, + "learning_rate": 1.9711519720417867e-05, + "loss": 0.8875, + "step": 3817 + }, + { + "epoch": 0.1690446409793786, + "grad_norm": 1.4075819253921509, + "learning_rate": 1.9711151074025073e-05, + "loss": 0.9665, + "step": 3818 + }, + { + "epoch": 0.16908891668419246, + "grad_norm": 1.318459391593933, + "learning_rate": 1.9710782195689343e-05, + "loss": 0.9408, + "step": 3819 + }, + { + "epoch": 0.16913319238900634, + "grad_norm": 1.3299678564071655, + "learning_rate": 1.9710413085419487e-05, + "loss": 0.9489, + "step": 3820 + }, + { + "epoch": 0.16917746809382023, + "grad_norm": 1.514640212059021, + "learning_rate": 1.971004374322432e-05, + "loss": 0.9819, + "step": 3821 + }, + { + "epoch": 0.16922174379863408, + "grad_norm": 1.3085341453552246, + "learning_rate": 1.9709674169112663e-05, + "loss": 0.8912, + "step": 3822 + }, + { + "epoch": 0.16926601950344797, + "grad_norm": 1.263235092163086, + "learning_rate": 1.9709304363093347e-05, + "loss": 0.9789, + "step": 3823 + }, + { + "epoch": 0.16931029520826185, + "grad_norm": 1.0151721239089966, + "learning_rate": 1.97089343251752e-05, + "loss": 0.8527, + "step": 3824 + }, + { + "epoch": 0.16935457091307574, + "grad_norm": 1.458422303199768, + "learning_rate": 1.9708564055367062e-05, + "loss": 1.011, + "step": 3825 + }, + { + "epoch": 0.1693988466178896, + "grad_norm": 1.4660251140594482, + "learning_rate": 1.9708193553677777e-05, + "loss": 0.9605, + "step": 3826 + }, + { + "epoch": 0.16944312232270348, + "grad_norm": 1.2628549337387085, + "learning_rate": 1.9707822820116193e-05, + "loss": 0.8894, + "step": 3827 + }, + { + "epoch": 0.16948739802751736, + "grad_norm": 1.3906855583190918, + "learning_rate": 1.9707451854691165e-05, + "loss": 0.9466, + "step": 3828 + }, + { + "epoch": 0.16953167373233122, + "grad_norm": 1.5226579904556274, + "learning_rate": 1.9707080657411556e-05, + "loss": 0.9936, + "step": 3829 + }, + { + "epoch": 0.1695759494371451, + "grad_norm": 1.3665187358856201, + "learning_rate": 1.9706709228286228e-05, + "loss": 1.0159, + "step": 3830 + }, + { + "epoch": 0.16962022514195899, + "grad_norm": 1.558207392692566, + "learning_rate": 1.9706337567324056e-05, + "loss": 1.0854, + "step": 3831 + }, + { + "epoch": 0.16966450084677284, + "grad_norm": 1.6076476573944092, + "learning_rate": 1.970596567453391e-05, + "loss": 1.0074, + "step": 3832 + }, + { + "epoch": 0.16970877655158673, + "grad_norm": 1.439255952835083, + "learning_rate": 1.970559354992468e-05, + "loss": 0.8886, + "step": 3833 + }, + { + "epoch": 0.1697530522564006, + "grad_norm": 1.3016945123672485, + "learning_rate": 1.970522119350525e-05, + "loss": 0.9234, + "step": 3834 + }, + { + "epoch": 0.1697973279612145, + "grad_norm": 1.3976402282714844, + "learning_rate": 1.9704848605284512e-05, + "loss": 0.9772, + "step": 3835 + }, + { + "epoch": 0.16984160366602835, + "grad_norm": 1.2586426734924316, + "learning_rate": 1.970447578527137e-05, + "loss": 0.9404, + "step": 3836 + }, + { + "epoch": 0.16988587937084224, + "grad_norm": 1.8490068912506104, + "learning_rate": 1.9704102733474725e-05, + "loss": 1.006, + "step": 3837 + }, + { + "epoch": 0.16993015507565612, + "grad_norm": 1.4199533462524414, + "learning_rate": 1.9703729449903486e-05, + "loss": 1.046, + "step": 3838 + }, + { + "epoch": 0.16997443078046998, + "grad_norm": 1.3374804258346558, + "learning_rate": 1.970335593456657e-05, + "loss": 0.9562, + "step": 3839 + }, + { + "epoch": 0.17001870648528386, + "grad_norm": 1.3663357496261597, + "learning_rate": 1.97029821874729e-05, + "loss": 0.9965, + "step": 3840 + }, + { + "epoch": 0.17006298219009774, + "grad_norm": 1.475867509841919, + "learning_rate": 1.97026082086314e-05, + "loss": 1.0297, + "step": 3841 + }, + { + "epoch": 0.1701072578949116, + "grad_norm": 1.500152587890625, + "learning_rate": 1.9702233998051e-05, + "loss": 1.0086, + "step": 3842 + }, + { + "epoch": 0.17015153359972549, + "grad_norm": 1.2804683446884155, + "learning_rate": 1.9701859555740647e-05, + "loss": 0.9616, + "step": 3843 + }, + { + "epoch": 0.17019580930453937, + "grad_norm": 1.2829291820526123, + "learning_rate": 1.9701484881709274e-05, + "loss": 0.9088, + "step": 3844 + }, + { + "epoch": 0.17024008500935325, + "grad_norm": 1.3940774202346802, + "learning_rate": 1.9701109975965834e-05, + "loss": 0.8958, + "step": 3845 + }, + { + "epoch": 0.1702843607141671, + "grad_norm": 2.1283493041992188, + "learning_rate": 1.9700734838519283e-05, + "loss": 1.0413, + "step": 3846 + }, + { + "epoch": 0.170328636418981, + "grad_norm": 1.1630973815917969, + "learning_rate": 1.9700359469378578e-05, + "loss": 0.8943, + "step": 3847 + }, + { + "epoch": 0.17037291212379488, + "grad_norm": 1.3350697755813599, + "learning_rate": 1.9699983868552685e-05, + "loss": 0.8786, + "step": 3848 + }, + { + "epoch": 0.17041718782860873, + "grad_norm": 1.5230692625045776, + "learning_rate": 1.9699608036050578e-05, + "loss": 0.9457, + "step": 3849 + }, + { + "epoch": 0.17046146353342262, + "grad_norm": 1.3187669515609741, + "learning_rate": 1.9699231971881227e-05, + "loss": 0.8617, + "step": 3850 + }, + { + "epoch": 0.1705057392382365, + "grad_norm": 1.3406383991241455, + "learning_rate": 1.969885567605362e-05, + "loss": 1.0039, + "step": 3851 + }, + { + "epoch": 0.1705500149430504, + "grad_norm": 1.4913395643234253, + "learning_rate": 1.9698479148576742e-05, + "loss": 1.0138, + "step": 3852 + }, + { + "epoch": 0.17059429064786424, + "grad_norm": 1.282739520072937, + "learning_rate": 1.9698102389459585e-05, + "loss": 0.9277, + "step": 3853 + }, + { + "epoch": 0.17063856635267813, + "grad_norm": 2.7811951637268066, + "learning_rate": 1.969772539871115e-05, + "loss": 1.068, + "step": 3854 + }, + { + "epoch": 0.170682842057492, + "grad_norm": 2.64090633392334, + "learning_rate": 1.9697348176340442e-05, + "loss": 0.9358, + "step": 3855 + }, + { + "epoch": 0.17072711776230587, + "grad_norm": 1.7734872102737427, + "learning_rate": 1.9696970722356466e-05, + "loss": 1.0973, + "step": 3856 + }, + { + "epoch": 0.17077139346711975, + "grad_norm": 1.132980227470398, + "learning_rate": 1.9696593036768238e-05, + "loss": 0.9029, + "step": 3857 + }, + { + "epoch": 0.17081566917193364, + "grad_norm": 1.5053818225860596, + "learning_rate": 1.9696215119584783e-05, + "loss": 1.1094, + "step": 3858 + }, + { + "epoch": 0.1708599448767475, + "grad_norm": 1.3356331586837769, + "learning_rate": 1.9695836970815124e-05, + "loss": 0.8924, + "step": 3859 + }, + { + "epoch": 0.17090422058156138, + "grad_norm": 1.3188035488128662, + "learning_rate": 1.9695458590468294e-05, + "loss": 0.9544, + "step": 3860 + }, + { + "epoch": 0.17094849628637526, + "grad_norm": 1.4932596683502197, + "learning_rate": 1.969507997855333e-05, + "loss": 1.0079, + "step": 3861 + }, + { + "epoch": 0.17099277199118915, + "grad_norm": 1.6617395877838135, + "learning_rate": 1.9694701135079277e-05, + "loss": 1.098, + "step": 3862 + }, + { + "epoch": 0.171037047696003, + "grad_norm": 1.2876251935958862, + "learning_rate": 1.969432206005518e-05, + "loss": 0.9281, + "step": 3863 + }, + { + "epoch": 0.1710813234008169, + "grad_norm": 1.2135250568389893, + "learning_rate": 1.969394275349009e-05, + "loss": 0.927, + "step": 3864 + }, + { + "epoch": 0.17112559910563077, + "grad_norm": 1.1833739280700684, + "learning_rate": 1.9693563215393078e-05, + "loss": 0.9502, + "step": 3865 + }, + { + "epoch": 0.17116987481044463, + "grad_norm": 1.379526972770691, + "learning_rate": 1.9693183445773193e-05, + "loss": 0.9488, + "step": 3866 + }, + { + "epoch": 0.1712141505152585, + "grad_norm": 1.218297004699707, + "learning_rate": 1.9692803444639517e-05, + "loss": 0.931, + "step": 3867 + }, + { + "epoch": 0.1712584262200724, + "grad_norm": 1.2822092771530151, + "learning_rate": 1.9692423212001126e-05, + "loss": 0.9694, + "step": 3868 + }, + { + "epoch": 0.17130270192488628, + "grad_norm": 1.595066785812378, + "learning_rate": 1.9692042747867096e-05, + "loss": 0.9528, + "step": 3869 + }, + { + "epoch": 0.17134697762970014, + "grad_norm": 1.4678105115890503, + "learning_rate": 1.9691662052246516e-05, + "loss": 0.9383, + "step": 3870 + }, + { + "epoch": 0.17139125333451402, + "grad_norm": 1.3644766807556152, + "learning_rate": 1.969128112514848e-05, + "loss": 0.9725, + "step": 3871 + }, + { + "epoch": 0.1714355290393279, + "grad_norm": 1.9192603826522827, + "learning_rate": 1.9690899966582085e-05, + "loss": 0.9681, + "step": 3872 + }, + { + "epoch": 0.17147980474414176, + "grad_norm": 1.4553474187850952, + "learning_rate": 1.9690518576556435e-05, + "loss": 1.0178, + "step": 3873 + }, + { + "epoch": 0.17152408044895565, + "grad_norm": 1.227231502532959, + "learning_rate": 1.969013695508064e-05, + "loss": 0.8593, + "step": 3874 + }, + { + "epoch": 0.17156835615376953, + "grad_norm": 1.3779587745666504, + "learning_rate": 1.968975510216381e-05, + "loss": 0.9748, + "step": 3875 + }, + { + "epoch": 0.1716126318585834, + "grad_norm": 1.9820178747177124, + "learning_rate": 1.9689373017815076e-05, + "loss": 1.0733, + "step": 3876 + }, + { + "epoch": 0.17165690756339727, + "grad_norm": 1.6148080825805664, + "learning_rate": 1.9688990702043553e-05, + "loss": 1.0209, + "step": 3877 + }, + { + "epoch": 0.17170118326821115, + "grad_norm": 1.5768307447433472, + "learning_rate": 1.9688608154858377e-05, + "loss": 0.9498, + "step": 3878 + }, + { + "epoch": 0.17174545897302504, + "grad_norm": 1.334086537361145, + "learning_rate": 1.968822537626868e-05, + "loss": 0.9902, + "step": 3879 + }, + { + "epoch": 0.1717897346778389, + "grad_norm": 1.2709451913833618, + "learning_rate": 1.9687842366283612e-05, + "loss": 0.9281, + "step": 3880 + }, + { + "epoch": 0.17183401038265278, + "grad_norm": 1.6905986070632935, + "learning_rate": 1.9687459124912317e-05, + "loss": 1.0257, + "step": 3881 + }, + { + "epoch": 0.17187828608746666, + "grad_norm": 1.2820457220077515, + "learning_rate": 1.968707565216395e-05, + "loss": 0.8964, + "step": 3882 + }, + { + "epoch": 0.17192256179228052, + "grad_norm": 1.7245396375656128, + "learning_rate": 1.9686691948047665e-05, + "loss": 1.0903, + "step": 3883 + }, + { + "epoch": 0.1719668374970944, + "grad_norm": 1.5540771484375, + "learning_rate": 1.968630801257263e-05, + "loss": 1.0271, + "step": 3884 + }, + { + "epoch": 0.1720111132019083, + "grad_norm": 1.5417078733444214, + "learning_rate": 1.968592384574802e-05, + "loss": 1.0315, + "step": 3885 + }, + { + "epoch": 0.17205538890672215, + "grad_norm": 1.3450833559036255, + "learning_rate": 1.9685539447583e-05, + "loss": 0.9939, + "step": 3886 + }, + { + "epoch": 0.17209966461153603, + "grad_norm": 1.4434373378753662, + "learning_rate": 1.968515481808676e-05, + "loss": 0.979, + "step": 3887 + }, + { + "epoch": 0.1721439403163499, + "grad_norm": 1.4243985414505005, + "learning_rate": 1.968476995726848e-05, + "loss": 1.0178, + "step": 3888 + }, + { + "epoch": 0.1721882160211638, + "grad_norm": 1.398943543434143, + "learning_rate": 1.968438486513736e-05, + "loss": 1.0582, + "step": 3889 + }, + { + "epoch": 0.17223249172597765, + "grad_norm": 1.4891490936279297, + "learning_rate": 1.9683999541702588e-05, + "loss": 0.9383, + "step": 3890 + }, + { + "epoch": 0.17227676743079154, + "grad_norm": 1.7357021570205688, + "learning_rate": 1.9683613986973373e-05, + "loss": 1.0196, + "step": 3891 + }, + { + "epoch": 0.17232104313560542, + "grad_norm": 1.506152629852295, + "learning_rate": 1.9683228200958924e-05, + "loss": 0.9801, + "step": 3892 + }, + { + "epoch": 0.17236531884041928, + "grad_norm": 1.1293433904647827, + "learning_rate": 1.9682842183668455e-05, + "loss": 0.8462, + "step": 3893 + }, + { + "epoch": 0.17240959454523316, + "grad_norm": 1.1671260595321655, + "learning_rate": 1.968245593511118e-05, + "loss": 0.9145, + "step": 3894 + }, + { + "epoch": 0.17245387025004705, + "grad_norm": 1.2282613515853882, + "learning_rate": 1.9682069455296334e-05, + "loss": 0.9777, + "step": 3895 + }, + { + "epoch": 0.17249814595486093, + "grad_norm": 1.408359169960022, + "learning_rate": 1.9681682744233138e-05, + "loss": 0.9968, + "step": 3896 + }, + { + "epoch": 0.1725424216596748, + "grad_norm": 2.0696115493774414, + "learning_rate": 1.968129580193084e-05, + "loss": 0.9312, + "step": 3897 + }, + { + "epoch": 0.17258669736448867, + "grad_norm": 1.161241888999939, + "learning_rate": 1.9680908628398665e-05, + "loss": 0.866, + "step": 3898 + }, + { + "epoch": 0.17263097306930256, + "grad_norm": 1.3484125137329102, + "learning_rate": 1.9680521223645876e-05, + "loss": 0.9762, + "step": 3899 + }, + { + "epoch": 0.1726752487741164, + "grad_norm": 1.4899629354476929, + "learning_rate": 1.9680133587681716e-05, + "loss": 0.992, + "step": 3900 + }, + { + "epoch": 0.1727195244789303, + "grad_norm": 1.229486107826233, + "learning_rate": 1.967974572051545e-05, + "loss": 0.9308, + "step": 3901 + }, + { + "epoch": 0.17276380018374418, + "grad_norm": 1.246285080909729, + "learning_rate": 1.9679357622156338e-05, + "loss": 0.9042, + "step": 3902 + }, + { + "epoch": 0.17280807588855804, + "grad_norm": 1.301135778427124, + "learning_rate": 1.967896929261365e-05, + "loss": 1.0353, + "step": 3903 + }, + { + "epoch": 0.17285235159337192, + "grad_norm": 1.2794852256774902, + "learning_rate": 1.967858073189666e-05, + "loss": 0.9526, + "step": 3904 + }, + { + "epoch": 0.1728966272981858, + "grad_norm": 1.1191296577453613, + "learning_rate": 1.9678191940014654e-05, + "loss": 0.8335, + "step": 3905 + }, + { + "epoch": 0.1729409030029997, + "grad_norm": 1.3665210008621216, + "learning_rate": 1.9677802916976915e-05, + "loss": 0.9689, + "step": 3906 + }, + { + "epoch": 0.17298517870781355, + "grad_norm": 1.3067572116851807, + "learning_rate": 1.9677413662792725e-05, + "loss": 0.8988, + "step": 3907 + }, + { + "epoch": 0.17302945441262743, + "grad_norm": 1.3930248022079468, + "learning_rate": 1.9677024177471397e-05, + "loss": 0.8747, + "step": 3908 + }, + { + "epoch": 0.17307373011744132, + "grad_norm": 1.5541502237319946, + "learning_rate": 1.9676634461022222e-05, + "loss": 0.9515, + "step": 3909 + }, + { + "epoch": 0.17311800582225517, + "grad_norm": 1.4261764287948608, + "learning_rate": 1.9676244513454513e-05, + "loss": 0.9797, + "step": 3910 + }, + { + "epoch": 0.17316228152706906, + "grad_norm": 1.3994039297103882, + "learning_rate": 1.9675854334777585e-05, + "loss": 0.9264, + "step": 3911 + }, + { + "epoch": 0.17320655723188294, + "grad_norm": 1.209743618965149, + "learning_rate": 1.9675463925000748e-05, + "loss": 0.9191, + "step": 3912 + }, + { + "epoch": 0.17325083293669682, + "grad_norm": 1.7809704542160034, + "learning_rate": 1.967507328413334e-05, + "loss": 0.9866, + "step": 3913 + }, + { + "epoch": 0.17329510864151068, + "grad_norm": 1.0875924825668335, + "learning_rate": 1.967468241218468e-05, + "loss": 0.8503, + "step": 3914 + }, + { + "epoch": 0.17333938434632457, + "grad_norm": 1.4924367666244507, + "learning_rate": 1.967429130916411e-05, + "loss": 0.984, + "step": 3915 + }, + { + "epoch": 0.17338366005113845, + "grad_norm": 1.6453790664672852, + "learning_rate": 1.967389997508097e-05, + "loss": 1.099, + "step": 3916 + }, + { + "epoch": 0.1734279357559523, + "grad_norm": 1.223598599433899, + "learning_rate": 1.9673508409944603e-05, + "loss": 0.9264, + "step": 3917 + }, + { + "epoch": 0.1734722114607662, + "grad_norm": 1.296152114868164, + "learning_rate": 1.9673116613764363e-05, + "loss": 0.8944, + "step": 3918 + }, + { + "epoch": 0.17351648716558007, + "grad_norm": 1.1511123180389404, + "learning_rate": 1.9672724586549616e-05, + "loss": 0.8878, + "step": 3919 + }, + { + "epoch": 0.17356076287039393, + "grad_norm": 1.3174771070480347, + "learning_rate": 1.9672332328309712e-05, + "loss": 0.9327, + "step": 3920 + }, + { + "epoch": 0.17360503857520782, + "grad_norm": 1.5479462146759033, + "learning_rate": 1.9671939839054026e-05, + "loss": 1.029, + "step": 3921 + }, + { + "epoch": 0.1736493142800217, + "grad_norm": 1.937463641166687, + "learning_rate": 1.9671547118791934e-05, + "loss": 0.8949, + "step": 3922 + }, + { + "epoch": 0.17369358998483558, + "grad_norm": 1.210253119468689, + "learning_rate": 1.9671154167532817e-05, + "loss": 0.9545, + "step": 3923 + }, + { + "epoch": 0.17373786568964944, + "grad_norm": 1.1539480686187744, + "learning_rate": 1.967076098528605e-05, + "loss": 0.9755, + "step": 3924 + }, + { + "epoch": 0.17378214139446332, + "grad_norm": 1.4966180324554443, + "learning_rate": 1.9670367572061035e-05, + "loss": 0.8822, + "step": 3925 + }, + { + "epoch": 0.1738264170992772, + "grad_norm": 1.1946361064910889, + "learning_rate": 1.9669973927867164e-05, + "loss": 0.8934, + "step": 3926 + }, + { + "epoch": 0.17387069280409106, + "grad_norm": 1.560516119003296, + "learning_rate": 1.9669580052713843e-05, + "loss": 0.9482, + "step": 3927 + }, + { + "epoch": 0.17391496850890495, + "grad_norm": 1.313344120979309, + "learning_rate": 1.966918594661047e-05, + "loss": 0.9037, + "step": 3928 + }, + { + "epoch": 0.17395924421371883, + "grad_norm": 1.3238219022750854, + "learning_rate": 1.9668791609566467e-05, + "loss": 0.9327, + "step": 3929 + }, + { + "epoch": 0.1740035199185327, + "grad_norm": 1.2682324647903442, + "learning_rate": 1.9668397041591247e-05, + "loss": 0.9195, + "step": 3930 + }, + { + "epoch": 0.17404779562334657, + "grad_norm": 1.3045765161514282, + "learning_rate": 1.966800224269424e-05, + "loss": 1.0163, + "step": 3931 + }, + { + "epoch": 0.17409207132816046, + "grad_norm": 1.3130868673324585, + "learning_rate": 1.9667607212884866e-05, + "loss": 0.8941, + "step": 3932 + }, + { + "epoch": 0.17413634703297434, + "grad_norm": 1.2575817108154297, + "learning_rate": 1.9667211952172566e-05, + "loss": 0.9408, + "step": 3933 + }, + { + "epoch": 0.1741806227377882, + "grad_norm": 1.2871806621551514, + "learning_rate": 1.9666816460566786e-05, + "loss": 0.9332, + "step": 3934 + }, + { + "epoch": 0.17422489844260208, + "grad_norm": 1.253542184829712, + "learning_rate": 1.9666420738076958e-05, + "loss": 0.9111, + "step": 3935 + }, + { + "epoch": 0.17426917414741597, + "grad_norm": 1.3845033645629883, + "learning_rate": 1.9666024784712546e-05, + "loss": 0.9795, + "step": 3936 + }, + { + "epoch": 0.17431344985222982, + "grad_norm": 1.1955151557922363, + "learning_rate": 1.9665628600483e-05, + "loss": 0.8205, + "step": 3937 + }, + { + "epoch": 0.1743577255570437, + "grad_norm": 1.3341047763824463, + "learning_rate": 1.9665232185397786e-05, + "loss": 0.9546, + "step": 3938 + }, + { + "epoch": 0.1744020012618576, + "grad_norm": 1.7276394367218018, + "learning_rate": 1.966483553946637e-05, + "loss": 1.0929, + "step": 3939 + }, + { + "epoch": 0.17444627696667148, + "grad_norm": 1.1292742490768433, + "learning_rate": 1.9664438662698227e-05, + "loss": 0.8218, + "step": 3940 + }, + { + "epoch": 0.17449055267148533, + "grad_norm": 1.1539900302886963, + "learning_rate": 1.9664041555102838e-05, + "loss": 0.8717, + "step": 3941 + }, + { + "epoch": 0.17453482837629922, + "grad_norm": 1.583809494972229, + "learning_rate": 1.9663644216689683e-05, + "loss": 1.0151, + "step": 3942 + }, + { + "epoch": 0.1745791040811131, + "grad_norm": 1.2666746377944946, + "learning_rate": 1.9663246647468255e-05, + "loss": 0.9494, + "step": 3943 + }, + { + "epoch": 0.17462337978592696, + "grad_norm": 1.5905394554138184, + "learning_rate": 1.966284884744805e-05, + "loss": 1.017, + "step": 3944 + }, + { + "epoch": 0.17466765549074084, + "grad_norm": 1.5185405015945435, + "learning_rate": 1.9662450816638566e-05, + "loss": 0.9762, + "step": 3945 + }, + { + "epoch": 0.17471193119555473, + "grad_norm": 1.626800537109375, + "learning_rate": 1.966205255504931e-05, + "loss": 1.0597, + "step": 3946 + }, + { + "epoch": 0.17475620690036858, + "grad_norm": 1.2627235651016235, + "learning_rate": 1.9661654062689802e-05, + "loss": 0.9423, + "step": 3947 + }, + { + "epoch": 0.17480048260518247, + "grad_norm": 1.283315896987915, + "learning_rate": 1.9661255339569547e-05, + "loss": 0.9275, + "step": 3948 + }, + { + "epoch": 0.17484475830999635, + "grad_norm": 1.5696746110916138, + "learning_rate": 1.966085638569808e-05, + "loss": 0.9952, + "step": 3949 + }, + { + "epoch": 0.17488903401481024, + "grad_norm": 1.121832013130188, + "learning_rate": 1.966045720108492e-05, + "loss": 0.866, + "step": 3950 + }, + { + "epoch": 0.1749333097196241, + "grad_norm": 1.0095183849334717, + "learning_rate": 1.9660057785739608e-05, + "loss": 0.8767, + "step": 3951 + }, + { + "epoch": 0.17497758542443798, + "grad_norm": 1.2075990438461304, + "learning_rate": 1.9659658139671683e-05, + "loss": 0.9561, + "step": 3952 + }, + { + "epoch": 0.17502186112925186, + "grad_norm": 1.1589393615722656, + "learning_rate": 1.9659258262890683e-05, + "loss": 0.9618, + "step": 3953 + }, + { + "epoch": 0.17506613683406572, + "grad_norm": 1.2574384212493896, + "learning_rate": 1.965885815540617e-05, + "loss": 0.893, + "step": 3954 + }, + { + "epoch": 0.1751104125388796, + "grad_norm": 1.376074194908142, + "learning_rate": 1.965845781722769e-05, + "loss": 0.9875, + "step": 3955 + }, + { + "epoch": 0.17515468824369348, + "grad_norm": 1.3853557109832764, + "learning_rate": 1.9658057248364812e-05, + "loss": 0.9831, + "step": 3956 + }, + { + "epoch": 0.17519896394850737, + "grad_norm": 1.3330317735671997, + "learning_rate": 1.96576564488271e-05, + "loss": 0.8943, + "step": 3957 + }, + { + "epoch": 0.17524323965332123, + "grad_norm": 1.321397066116333, + "learning_rate": 1.965725541862413e-05, + "loss": 0.8731, + "step": 3958 + }, + { + "epoch": 0.1752875153581351, + "grad_norm": 1.339850902557373, + "learning_rate": 1.9656854157765472e-05, + "loss": 0.9674, + "step": 3959 + }, + { + "epoch": 0.175331791062949, + "grad_norm": 1.3519511222839355, + "learning_rate": 1.965645266626072e-05, + "loss": 0.9602, + "step": 3960 + }, + { + "epoch": 0.17537606676776285, + "grad_norm": 1.4656167030334473, + "learning_rate": 1.9656050944119457e-05, + "loss": 1.0842, + "step": 3961 + }, + { + "epoch": 0.17542034247257673, + "grad_norm": 1.2803667783737183, + "learning_rate": 1.9655648991351278e-05, + "loss": 0.9287, + "step": 3962 + }, + { + "epoch": 0.17546461817739062, + "grad_norm": 3.0848913192749023, + "learning_rate": 1.9655246807965786e-05, + "loss": 1.0946, + "step": 3963 + }, + { + "epoch": 0.17550889388220448, + "grad_norm": 1.2680708169937134, + "learning_rate": 1.9654844393972582e-05, + "loss": 0.8761, + "step": 3964 + }, + { + "epoch": 0.17555316958701836, + "grad_norm": 1.5726897716522217, + "learning_rate": 1.965444174938129e-05, + "loss": 1.0258, + "step": 3965 + }, + { + "epoch": 0.17559744529183224, + "grad_norm": 1.4175409078598022, + "learning_rate": 1.9654038874201508e-05, + "loss": 0.894, + "step": 3966 + }, + { + "epoch": 0.17564172099664613, + "grad_norm": 1.1628761291503906, + "learning_rate": 1.9653635768442872e-05, + "loss": 0.9365, + "step": 3967 + }, + { + "epoch": 0.17568599670145998, + "grad_norm": 1.7182095050811768, + "learning_rate": 1.9653232432115004e-05, + "loss": 0.9596, + "step": 3968 + }, + { + "epoch": 0.17573027240627387, + "grad_norm": 1.4716570377349854, + "learning_rate": 1.9652828865227546e-05, + "loss": 0.9462, + "step": 3969 + }, + { + "epoch": 0.17577454811108775, + "grad_norm": 1.3689054250717163, + "learning_rate": 1.9652425067790124e-05, + "loss": 0.9454, + "step": 3970 + }, + { + "epoch": 0.1758188238159016, + "grad_norm": 1.3913992643356323, + "learning_rate": 1.965202103981239e-05, + "loss": 0.9059, + "step": 3971 + }, + { + "epoch": 0.1758630995207155, + "grad_norm": 1.1553336381912231, + "learning_rate": 1.9651616781303994e-05, + "loss": 0.8452, + "step": 3972 + }, + { + "epoch": 0.17590737522552938, + "grad_norm": 1.362963318824768, + "learning_rate": 1.965121229227459e-05, + "loss": 0.9927, + "step": 3973 + }, + { + "epoch": 0.17595165093034323, + "grad_norm": 1.475347876548767, + "learning_rate": 1.9650807572733834e-05, + "loss": 0.9103, + "step": 3974 + }, + { + "epoch": 0.17599592663515712, + "grad_norm": 1.3421518802642822, + "learning_rate": 1.96504026226914e-05, + "loss": 0.9185, + "step": 3975 + }, + { + "epoch": 0.176040202339971, + "grad_norm": 1.4606472253799438, + "learning_rate": 1.964999744215696e-05, + "loss": 0.9531, + "step": 3976 + }, + { + "epoch": 0.1760844780447849, + "grad_norm": 1.2462482452392578, + "learning_rate": 1.9649592031140184e-05, + "loss": 0.975, + "step": 3977 + }, + { + "epoch": 0.17612875374959874, + "grad_norm": 1.328901767730713, + "learning_rate": 1.964918638965076e-05, + "loss": 1.0229, + "step": 3978 + }, + { + "epoch": 0.17617302945441263, + "grad_norm": 1.2664239406585693, + "learning_rate": 1.9648780517698375e-05, + "loss": 0.9448, + "step": 3979 + }, + { + "epoch": 0.1762173051592265, + "grad_norm": 1.386365532875061, + "learning_rate": 1.9648374415292728e-05, + "loss": 0.9901, + "step": 3980 + }, + { + "epoch": 0.17626158086404037, + "grad_norm": 1.2557214498519897, + "learning_rate": 1.964796808244351e-05, + "loss": 0.985, + "step": 3981 + }, + { + "epoch": 0.17630585656885425, + "grad_norm": 1.3084949254989624, + "learning_rate": 1.9647561519160434e-05, + "loss": 0.9287, + "step": 3982 + }, + { + "epoch": 0.17635013227366814, + "grad_norm": 1.2524921894073486, + "learning_rate": 1.9647154725453204e-05, + "loss": 0.8634, + "step": 3983 + }, + { + "epoch": 0.17639440797848202, + "grad_norm": 1.620195746421814, + "learning_rate": 1.964674770133154e-05, + "loss": 0.9724, + "step": 3984 + }, + { + "epoch": 0.17643868368329588, + "grad_norm": 1.2579065561294556, + "learning_rate": 1.9646340446805158e-05, + "loss": 0.9807, + "step": 3985 + }, + { + "epoch": 0.17648295938810976, + "grad_norm": 1.2927885055541992, + "learning_rate": 1.964593296188379e-05, + "loss": 0.9842, + "step": 3986 + }, + { + "epoch": 0.17652723509292365, + "grad_norm": 1.3369182348251343, + "learning_rate": 1.9645525246577168e-05, + "loss": 1.0018, + "step": 3987 + }, + { + "epoch": 0.1765715107977375, + "grad_norm": 1.4052664041519165, + "learning_rate": 1.964511730089503e-05, + "loss": 0.9096, + "step": 3988 + }, + { + "epoch": 0.1766157865025514, + "grad_norm": 1.333038330078125, + "learning_rate": 1.964470912484712e-05, + "loss": 0.9878, + "step": 3989 + }, + { + "epoch": 0.17666006220736527, + "grad_norm": 1.3744304180145264, + "learning_rate": 1.9644300718443182e-05, + "loss": 0.9949, + "step": 3990 + }, + { + "epoch": 0.17670433791217913, + "grad_norm": 1.6443371772766113, + "learning_rate": 1.9643892081692975e-05, + "loss": 0.9385, + "step": 3991 + }, + { + "epoch": 0.176748613616993, + "grad_norm": 1.4761216640472412, + "learning_rate": 1.964348321460626e-05, + "loss": 0.9185, + "step": 3992 + }, + { + "epoch": 0.1767928893218069, + "grad_norm": 1.3977159261703491, + "learning_rate": 1.9643074117192796e-05, + "loss": 0.992, + "step": 3993 + }, + { + "epoch": 0.17683716502662078, + "grad_norm": 1.5302420854568481, + "learning_rate": 1.9642664789462363e-05, + "loss": 1.0448, + "step": 3994 + }, + { + "epoch": 0.17688144073143464, + "grad_norm": 1.4035412073135376, + "learning_rate": 1.964225523142473e-05, + "loss": 0.939, + "step": 3995 + }, + { + "epoch": 0.17692571643624852, + "grad_norm": 1.2844513654708862, + "learning_rate": 1.9641845443089686e-05, + "loss": 0.9595, + "step": 3996 + }, + { + "epoch": 0.1769699921410624, + "grad_norm": 1.5389260053634644, + "learning_rate": 1.964143542446701e-05, + "loss": 0.9886, + "step": 3997 + }, + { + "epoch": 0.17701426784587626, + "grad_norm": 1.3377519845962524, + "learning_rate": 1.96410251755665e-05, + "loss": 0.9327, + "step": 3998 + }, + { + "epoch": 0.17705854355069015, + "grad_norm": 1.5633354187011719, + "learning_rate": 1.9640614696397954e-05, + "loss": 1.0805, + "step": 3999 + }, + { + "epoch": 0.17710281925550403, + "grad_norm": 1.2880553007125854, + "learning_rate": 1.9640203986971176e-05, + "loss": 0.9767, + "step": 4000 + }, + { + "epoch": 0.17714709496031789, + "grad_norm": 1.4835090637207031, + "learning_rate": 1.9639793047295974e-05, + "loss": 0.9947, + "step": 4001 + }, + { + "epoch": 0.17719137066513177, + "grad_norm": 1.2663253545761108, + "learning_rate": 1.9639381877382167e-05, + "loss": 0.8597, + "step": 4002 + }, + { + "epoch": 0.17723564636994565, + "grad_norm": 1.5061959028244019, + "learning_rate": 1.963897047723957e-05, + "loss": 0.9186, + "step": 4003 + }, + { + "epoch": 0.17727992207475954, + "grad_norm": 1.368548035621643, + "learning_rate": 1.9638558846878012e-05, + "loss": 0.9369, + "step": 4004 + }, + { + "epoch": 0.1773241977795734, + "grad_norm": 1.2620515823364258, + "learning_rate": 1.9638146986307322e-05, + "loss": 0.921, + "step": 4005 + }, + { + "epoch": 0.17736847348438728, + "grad_norm": 1.5286402702331543, + "learning_rate": 1.9637734895537343e-05, + "loss": 1.0384, + "step": 4006 + }, + { + "epoch": 0.17741274918920116, + "grad_norm": 1.4565833806991577, + "learning_rate": 1.963732257457791e-05, + "loss": 1.0269, + "step": 4007 + }, + { + "epoch": 0.17745702489401502, + "grad_norm": 1.3000234365463257, + "learning_rate": 1.9636910023438878e-05, + "loss": 1.0241, + "step": 4008 + }, + { + "epoch": 0.1775013005988289, + "grad_norm": 1.554237723350525, + "learning_rate": 1.9636497242130093e-05, + "loss": 0.954, + "step": 4009 + }, + { + "epoch": 0.1775455763036428, + "grad_norm": 1.4805611371994019, + "learning_rate": 1.963608423066142e-05, + "loss": 1.0126, + "step": 4010 + }, + { + "epoch": 0.17758985200845667, + "grad_norm": 1.3382056951522827, + "learning_rate": 1.963567098904272e-05, + "loss": 0.9667, + "step": 4011 + }, + { + "epoch": 0.17763412771327053, + "grad_norm": 1.2590429782867432, + "learning_rate": 1.9635257517283862e-05, + "loss": 0.9636, + "step": 4012 + }, + { + "epoch": 0.1776784034180844, + "grad_norm": 1.1831408739089966, + "learning_rate": 1.963484381539473e-05, + "loss": 0.9134, + "step": 4013 + }, + { + "epoch": 0.1777226791228983, + "grad_norm": 1.4771231412887573, + "learning_rate": 1.9634429883385195e-05, + "loss": 1.008, + "step": 4014 + }, + { + "epoch": 0.17776695482771215, + "grad_norm": 1.2990788221359253, + "learning_rate": 1.9634015721265143e-05, + "loss": 0.9717, + "step": 4015 + }, + { + "epoch": 0.17781123053252604, + "grad_norm": 1.24703848361969, + "learning_rate": 1.9633601329044474e-05, + "loss": 0.9139, + "step": 4016 + }, + { + "epoch": 0.17785550623733992, + "grad_norm": 1.2398037910461426, + "learning_rate": 1.963318670673308e-05, + "loss": 0.9429, + "step": 4017 + }, + { + "epoch": 0.17789978194215378, + "grad_norm": 1.2057560682296753, + "learning_rate": 1.9632771854340863e-05, + "loss": 0.8568, + "step": 4018 + }, + { + "epoch": 0.17794405764696766, + "grad_norm": 1.145634651184082, + "learning_rate": 1.9632356771877735e-05, + "loss": 0.9277, + "step": 4019 + }, + { + "epoch": 0.17798833335178155, + "grad_norm": 1.3293319940567017, + "learning_rate": 1.963194145935361e-05, + "loss": 0.9121, + "step": 4020 + }, + { + "epoch": 0.17803260905659543, + "grad_norm": 1.508425235748291, + "learning_rate": 1.9631525916778408e-05, + "loss": 0.9008, + "step": 4021 + }, + { + "epoch": 0.1780768847614093, + "grad_norm": 1.4744364023208618, + "learning_rate": 1.963111014416205e-05, + "loss": 0.875, + "step": 4022 + }, + { + "epoch": 0.17812116046622317, + "grad_norm": 1.763432264328003, + "learning_rate": 1.9630694141514467e-05, + "loss": 1.0265, + "step": 4023 + }, + { + "epoch": 0.17816543617103706, + "grad_norm": 1.2120025157928467, + "learning_rate": 1.9630277908845594e-05, + "loss": 0.9267, + "step": 4024 + }, + { + "epoch": 0.1782097118758509, + "grad_norm": 1.4288564920425415, + "learning_rate": 1.9629861446165376e-05, + "loss": 0.9775, + "step": 4025 + }, + { + "epoch": 0.1782539875806648, + "grad_norm": 2.1139674186706543, + "learning_rate": 1.9629444753483758e-05, + "loss": 1.0293, + "step": 4026 + }, + { + "epoch": 0.17829826328547868, + "grad_norm": 1.3158552646636963, + "learning_rate": 1.9629027830810693e-05, + "loss": 0.9586, + "step": 4027 + }, + { + "epoch": 0.17834253899029257, + "grad_norm": 1.1717948913574219, + "learning_rate": 1.962861067815614e-05, + "loss": 0.9295, + "step": 4028 + }, + { + "epoch": 0.17838681469510642, + "grad_norm": 1.4608871936798096, + "learning_rate": 1.962819329553006e-05, + "loss": 1.0104, + "step": 4029 + }, + { + "epoch": 0.1784310903999203, + "grad_norm": 1.5421268939971924, + "learning_rate": 1.962777568294242e-05, + "loss": 0.8797, + "step": 4030 + }, + { + "epoch": 0.1784753661047342, + "grad_norm": 1.464103102684021, + "learning_rate": 1.9627357840403204e-05, + "loss": 1.0431, + "step": 4031 + }, + { + "epoch": 0.17851964180954805, + "grad_norm": 1.295547366142273, + "learning_rate": 1.9626939767922377e-05, + "loss": 0.947, + "step": 4032 + }, + { + "epoch": 0.17856391751436193, + "grad_norm": 1.2929428815841675, + "learning_rate": 1.9626521465509938e-05, + "loss": 0.9132, + "step": 4033 + }, + { + "epoch": 0.17860819321917581, + "grad_norm": 1.3514461517333984, + "learning_rate": 1.962610293317587e-05, + "loss": 1.0222, + "step": 4034 + }, + { + "epoch": 0.17865246892398967, + "grad_norm": 1.2843061685562134, + "learning_rate": 1.9625684170930172e-05, + "loss": 0.9248, + "step": 4035 + }, + { + "epoch": 0.17869674462880356, + "grad_norm": 1.263821005821228, + "learning_rate": 1.9625265178782842e-05, + "loss": 0.9826, + "step": 4036 + }, + { + "epoch": 0.17874102033361744, + "grad_norm": 1.5948511362075806, + "learning_rate": 1.9624845956743892e-05, + "loss": 0.9427, + "step": 4037 + }, + { + "epoch": 0.17878529603843132, + "grad_norm": 1.4528049230575562, + "learning_rate": 1.962442650482333e-05, + "loss": 1.0699, + "step": 4038 + }, + { + "epoch": 0.17882957174324518, + "grad_norm": 1.3738038539886475, + "learning_rate": 1.9624006823031185e-05, + "loss": 1.0059, + "step": 4039 + }, + { + "epoch": 0.17887384744805906, + "grad_norm": 1.2570117712020874, + "learning_rate": 1.9623586911377468e-05, + "loss": 0.8823, + "step": 4040 + }, + { + "epoch": 0.17891812315287295, + "grad_norm": 1.2750954627990723, + "learning_rate": 1.9623166769872216e-05, + "loss": 0.9534, + "step": 4041 + }, + { + "epoch": 0.1789623988576868, + "grad_norm": 1.2323566675186157, + "learning_rate": 1.962274639852546e-05, + "loss": 0.9934, + "step": 4042 + }, + { + "epoch": 0.1790066745625007, + "grad_norm": 1.3882025480270386, + "learning_rate": 1.962232579734724e-05, + "loss": 0.8966, + "step": 4043 + }, + { + "epoch": 0.17905095026731457, + "grad_norm": 1.338939905166626, + "learning_rate": 1.9621904966347604e-05, + "loss": 0.9129, + "step": 4044 + }, + { + "epoch": 0.17909522597212843, + "grad_norm": 1.3745115995407104, + "learning_rate": 1.9621483905536603e-05, + "loss": 0.9551, + "step": 4045 + }, + { + "epoch": 0.17913950167694231, + "grad_norm": 1.562180519104004, + "learning_rate": 1.9621062614924294e-05, + "loss": 0.918, + "step": 4046 + }, + { + "epoch": 0.1791837773817562, + "grad_norm": 1.5617797374725342, + "learning_rate": 1.9620641094520737e-05, + "loss": 0.9936, + "step": 4047 + }, + { + "epoch": 0.17922805308657008, + "grad_norm": 1.5210992097854614, + "learning_rate": 1.9620219344336e-05, + "loss": 0.9824, + "step": 4048 + }, + { + "epoch": 0.17927232879138394, + "grad_norm": 2.4045324325561523, + "learning_rate": 1.9619797364380157e-05, + "loss": 1.1213, + "step": 4049 + }, + { + "epoch": 0.17931660449619782, + "grad_norm": 1.3631998300552368, + "learning_rate": 1.961937515466329e-05, + "loss": 0.9656, + "step": 4050 + }, + { + "epoch": 0.1793608802010117, + "grad_norm": 1.3842692375183105, + "learning_rate": 1.9618952715195476e-05, + "loss": 0.9106, + "step": 4051 + }, + { + "epoch": 0.17940515590582556, + "grad_norm": 1.1599576473236084, + "learning_rate": 1.961853004598681e-05, + "loss": 0.9176, + "step": 4052 + }, + { + "epoch": 0.17944943161063945, + "grad_norm": 1.4512344598770142, + "learning_rate": 1.9618107147047386e-05, + "loss": 0.9673, + "step": 4053 + }, + { + "epoch": 0.17949370731545333, + "grad_norm": 1.1300277709960938, + "learning_rate": 1.9617684018387303e-05, + "loss": 0.8534, + "step": 4054 + }, + { + "epoch": 0.17953798302026722, + "grad_norm": 1.2351584434509277, + "learning_rate": 1.9617260660016668e-05, + "loss": 0.9337, + "step": 4055 + }, + { + "epoch": 0.17958225872508107, + "grad_norm": 1.3616342544555664, + "learning_rate": 1.9616837071945592e-05, + "loss": 1.0032, + "step": 4056 + }, + { + "epoch": 0.17962653442989496, + "grad_norm": 1.223388433456421, + "learning_rate": 1.9616413254184196e-05, + "loss": 0.9118, + "step": 4057 + }, + { + "epoch": 0.17967081013470884, + "grad_norm": 1.2230339050292969, + "learning_rate": 1.9615989206742595e-05, + "loss": 0.9414, + "step": 4058 + }, + { + "epoch": 0.1797150858395227, + "grad_norm": 1.314932942390442, + "learning_rate": 1.9615564929630925e-05, + "loss": 0.9517, + "step": 4059 + }, + { + "epoch": 0.17975936154433658, + "grad_norm": 1.5434080362319946, + "learning_rate": 1.9615140422859316e-05, + "loss": 1.0145, + "step": 4060 + }, + { + "epoch": 0.17980363724915047, + "grad_norm": 1.4041545391082764, + "learning_rate": 1.9614715686437908e-05, + "loss": 0.9449, + "step": 4061 + }, + { + "epoch": 0.17984791295396432, + "grad_norm": 1.59203040599823, + "learning_rate": 1.961429072037684e-05, + "loss": 0.9188, + "step": 4062 + }, + { + "epoch": 0.1798921886587782, + "grad_norm": 1.3759137392044067, + "learning_rate": 1.961386552468627e-05, + "loss": 1.0116, + "step": 4063 + }, + { + "epoch": 0.1799364643635921, + "grad_norm": 1.3781541585922241, + "learning_rate": 1.961344009937635e-05, + "loss": 0.9318, + "step": 4064 + }, + { + "epoch": 0.17998074006840598, + "grad_norm": 1.5469831228256226, + "learning_rate": 1.9613014444457237e-05, + "loss": 0.9976, + "step": 4065 + }, + { + "epoch": 0.18002501577321983, + "grad_norm": 1.19667649269104, + "learning_rate": 1.9612588559939105e-05, + "loss": 0.869, + "step": 4066 + }, + { + "epoch": 0.18006929147803372, + "grad_norm": 1.3403685092926025, + "learning_rate": 1.961216244583212e-05, + "loss": 0.9079, + "step": 4067 + }, + { + "epoch": 0.1801135671828476, + "grad_norm": 1.32817804813385, + "learning_rate": 1.961173610214646e-05, + "loss": 1.0017, + "step": 4068 + }, + { + "epoch": 0.18015784288766146, + "grad_norm": 1.2945685386657715, + "learning_rate": 1.9611309528892317e-05, + "loss": 1.004, + "step": 4069 + }, + { + "epoch": 0.18020211859247534, + "grad_norm": 1.513136863708496, + "learning_rate": 1.9610882726079864e-05, + "loss": 0.9711, + "step": 4070 + }, + { + "epoch": 0.18024639429728923, + "grad_norm": 1.4237210750579834, + "learning_rate": 1.9610455693719304e-05, + "loss": 0.9167, + "step": 4071 + }, + { + "epoch": 0.1802906700021031, + "grad_norm": 1.4324594736099243, + "learning_rate": 1.9610028431820836e-05, + "loss": 0.9123, + "step": 4072 + }, + { + "epoch": 0.18033494570691697, + "grad_norm": 1.5199270248413086, + "learning_rate": 1.9609600940394663e-05, + "loss": 1.0582, + "step": 4073 + }, + { + "epoch": 0.18037922141173085, + "grad_norm": 1.3569468259811401, + "learning_rate": 1.9609173219450998e-05, + "loss": 0.9866, + "step": 4074 + }, + { + "epoch": 0.18042349711654473, + "grad_norm": 1.2636581659317017, + "learning_rate": 1.960874526900005e-05, + "loss": 0.9045, + "step": 4075 + }, + { + "epoch": 0.1804677728213586, + "grad_norm": 1.6128250360488892, + "learning_rate": 1.960831708905205e-05, + "loss": 0.8917, + "step": 4076 + }, + { + "epoch": 0.18051204852617247, + "grad_norm": 1.315980076789856, + "learning_rate": 1.9607888679617217e-05, + "loss": 0.9595, + "step": 4077 + }, + { + "epoch": 0.18055632423098636, + "grad_norm": 1.316112756729126, + "learning_rate": 1.9607460040705787e-05, + "loss": 0.9885, + "step": 4078 + }, + { + "epoch": 0.18060059993580022, + "grad_norm": 1.418312907218933, + "learning_rate": 1.9607031172327998e-05, + "loss": 0.923, + "step": 4079 + }, + { + "epoch": 0.1806448756406141, + "grad_norm": 1.6693412065505981, + "learning_rate": 1.960660207449409e-05, + "loss": 1.0085, + "step": 4080 + }, + { + "epoch": 0.18068915134542798, + "grad_norm": 1.3515623807907104, + "learning_rate": 1.960617274721431e-05, + "loss": 1.0044, + "step": 4081 + }, + { + "epoch": 0.18073342705024187, + "grad_norm": 1.5743576288223267, + "learning_rate": 1.960574319049892e-05, + "loss": 1.0161, + "step": 4082 + }, + { + "epoch": 0.18077770275505572, + "grad_norm": 1.515518307685852, + "learning_rate": 1.9605313404358176e-05, + "loss": 1.0569, + "step": 4083 + }, + { + "epoch": 0.1808219784598696, + "grad_norm": 1.1642847061157227, + "learning_rate": 1.960488338880234e-05, + "loss": 0.9544, + "step": 4084 + }, + { + "epoch": 0.1808662541646835, + "grad_norm": 1.2992242574691772, + "learning_rate": 1.9604453143841685e-05, + "loss": 0.9351, + "step": 4085 + }, + { + "epoch": 0.18091052986949735, + "grad_norm": 1.2846088409423828, + "learning_rate": 1.9604022669486487e-05, + "loss": 0.9165, + "step": 4086 + }, + { + "epoch": 0.18095480557431123, + "grad_norm": 1.2482500076293945, + "learning_rate": 1.960359196574703e-05, + "loss": 0.9603, + "step": 4087 + }, + { + "epoch": 0.18099908127912512, + "grad_norm": 1.4781709909439087, + "learning_rate": 1.9603161032633597e-05, + "loss": 0.9787, + "step": 4088 + }, + { + "epoch": 0.18104335698393897, + "grad_norm": 1.3269768953323364, + "learning_rate": 1.960272987015648e-05, + "loss": 0.962, + "step": 4089 + }, + { + "epoch": 0.18108763268875286, + "grad_norm": 1.3704954385757446, + "learning_rate": 1.960229847832598e-05, + "loss": 0.9691, + "step": 4090 + }, + { + "epoch": 0.18113190839356674, + "grad_norm": 1.369357705116272, + "learning_rate": 1.9601866857152402e-05, + "loss": 0.9221, + "step": 4091 + }, + { + "epoch": 0.18117618409838063, + "grad_norm": 1.207911729812622, + "learning_rate": 1.9601435006646045e-05, + "loss": 0.8969, + "step": 4092 + }, + { + "epoch": 0.18122045980319448, + "grad_norm": 1.2147279977798462, + "learning_rate": 1.9601002926817237e-05, + "loss": 0.941, + "step": 4093 + }, + { + "epoch": 0.18126473550800837, + "grad_norm": 2.8409130573272705, + "learning_rate": 1.9600570617676288e-05, + "loss": 1.0409, + "step": 4094 + }, + { + "epoch": 0.18130901121282225, + "grad_norm": 1.7589178085327148, + "learning_rate": 1.960013807923353e-05, + "loss": 1.0493, + "step": 4095 + }, + { + "epoch": 0.1813532869176361, + "grad_norm": 1.2667195796966553, + "learning_rate": 1.959970531149929e-05, + "loss": 0.9981, + "step": 4096 + }, + { + "epoch": 0.18139756262245, + "grad_norm": 1.301959753036499, + "learning_rate": 1.9599272314483903e-05, + "loss": 0.9706, + "step": 4097 + }, + { + "epoch": 0.18144183832726388, + "grad_norm": 1.0737658739089966, + "learning_rate": 1.9598839088197715e-05, + "loss": 0.8274, + "step": 4098 + }, + { + "epoch": 0.18148611403207776, + "grad_norm": 1.424047827720642, + "learning_rate": 1.959840563265107e-05, + "loss": 0.9589, + "step": 4099 + }, + { + "epoch": 0.18153038973689162, + "grad_norm": 1.5437260866165161, + "learning_rate": 1.9597971947854325e-05, + "loss": 0.9129, + "step": 4100 + }, + { + "epoch": 0.1815746654417055, + "grad_norm": 1.3564555644989014, + "learning_rate": 1.959753803381783e-05, + "loss": 0.895, + "step": 4101 + }, + { + "epoch": 0.18161894114651939, + "grad_norm": 1.3113588094711304, + "learning_rate": 1.9597103890551956e-05, + "loss": 0.9462, + "step": 4102 + }, + { + "epoch": 0.18166321685133324, + "grad_norm": 1.2593944072723389, + "learning_rate": 1.959666951806707e-05, + "loss": 0.9151, + "step": 4103 + }, + { + "epoch": 0.18170749255614713, + "grad_norm": 1.2386592626571655, + "learning_rate": 1.9596234916373545e-05, + "loss": 0.9259, + "step": 4104 + }, + { + "epoch": 0.181751768260961, + "grad_norm": 1.5820481777191162, + "learning_rate": 1.9595800085481765e-05, + "loss": 0.8753, + "step": 4105 + }, + { + "epoch": 0.18179604396577487, + "grad_norm": 1.314162015914917, + "learning_rate": 1.9595365025402108e-05, + "loss": 0.9051, + "step": 4106 + }, + { + "epoch": 0.18184031967058875, + "grad_norm": 2.774949789047241, + "learning_rate": 1.9594929736144978e-05, + "loss": 0.9797, + "step": 4107 + }, + { + "epoch": 0.18188459537540264, + "grad_norm": 1.2810471057891846, + "learning_rate": 1.959449421772076e-05, + "loss": 0.8996, + "step": 4108 + }, + { + "epoch": 0.18192887108021652, + "grad_norm": 1.8370540142059326, + "learning_rate": 1.9594058470139855e-05, + "loss": 1.0487, + "step": 4109 + }, + { + "epoch": 0.18197314678503038, + "grad_norm": 1.43623948097229, + "learning_rate": 1.9593622493412682e-05, + "loss": 0.983, + "step": 4110 + }, + { + "epoch": 0.18201742248984426, + "grad_norm": 1.2732206583023071, + "learning_rate": 1.9593186287549643e-05, + "loss": 0.9622, + "step": 4111 + }, + { + "epoch": 0.18206169819465814, + "grad_norm": 2.0291993618011475, + "learning_rate": 1.9592749852561162e-05, + "loss": 0.9808, + "step": 4112 + }, + { + "epoch": 0.182105973899472, + "grad_norm": 1.3185358047485352, + "learning_rate": 1.9592313188457662e-05, + "loss": 0.8911, + "step": 4113 + }, + { + "epoch": 0.18215024960428589, + "grad_norm": 1.4454867839813232, + "learning_rate": 1.9591876295249572e-05, + "loss": 0.8956, + "step": 4114 + }, + { + "epoch": 0.18219452530909977, + "grad_norm": 1.5181130170822144, + "learning_rate": 1.9591439172947326e-05, + "loss": 0.8992, + "step": 4115 + }, + { + "epoch": 0.18223880101391365, + "grad_norm": 1.4628562927246094, + "learning_rate": 1.9591001821561364e-05, + "loss": 0.9676, + "step": 4116 + }, + { + "epoch": 0.1822830767187275, + "grad_norm": 1.2766239643096924, + "learning_rate": 1.9590564241102138e-05, + "loss": 0.8779, + "step": 4117 + }, + { + "epoch": 0.1823273524235414, + "grad_norm": 1.3820688724517822, + "learning_rate": 1.9590126431580087e-05, + "loss": 0.9752, + "step": 4118 + }, + { + "epoch": 0.18237162812835528, + "grad_norm": 1.4808826446533203, + "learning_rate": 1.958968839300568e-05, + "loss": 0.8946, + "step": 4119 + }, + { + "epoch": 0.18241590383316914, + "grad_norm": 2.557603597640991, + "learning_rate": 1.958925012538937e-05, + "loss": 0.9138, + "step": 4120 + }, + { + "epoch": 0.18246017953798302, + "grad_norm": 1.8396795988082886, + "learning_rate": 1.958881162874163e-05, + "loss": 1.0066, + "step": 4121 + }, + { + "epoch": 0.1825044552427969, + "grad_norm": 1.481452226638794, + "learning_rate": 1.9588372903072935e-05, + "loss": 0.9606, + "step": 4122 + }, + { + "epoch": 0.18254873094761076, + "grad_norm": 1.4105726480484009, + "learning_rate": 1.9587933948393754e-05, + "loss": 0.9477, + "step": 4123 + }, + { + "epoch": 0.18259300665242464, + "grad_norm": 1.3937454223632812, + "learning_rate": 1.958749476471458e-05, + "loss": 1.052, + "step": 4124 + }, + { + "epoch": 0.18263728235723853, + "grad_norm": 1.4556293487548828, + "learning_rate": 1.95870553520459e-05, + "loss": 0.9968, + "step": 4125 + }, + { + "epoch": 0.1826815580620524, + "grad_norm": 1.5480210781097412, + "learning_rate": 1.9586615710398212e-05, + "loss": 0.9778, + "step": 4126 + }, + { + "epoch": 0.18272583376686627, + "grad_norm": 1.7010788917541504, + "learning_rate": 1.9586175839782014e-05, + "loss": 0.9768, + "step": 4127 + }, + { + "epoch": 0.18277010947168015, + "grad_norm": 1.4716362953186035, + "learning_rate": 1.958573574020781e-05, + "loss": 1.0715, + "step": 4128 + }, + { + "epoch": 0.18281438517649404, + "grad_norm": 1.4654297828674316, + "learning_rate": 1.958529541168611e-05, + "loss": 1.0109, + "step": 4129 + }, + { + "epoch": 0.1828586608813079, + "grad_norm": 1.2214792966842651, + "learning_rate": 1.9584854854227432e-05, + "loss": 0.9368, + "step": 4130 + }, + { + "epoch": 0.18290293658612178, + "grad_norm": 1.7896852493286133, + "learning_rate": 1.9584414067842304e-05, + "loss": 1.0178, + "step": 4131 + }, + { + "epoch": 0.18294721229093566, + "grad_norm": 1.6457622051239014, + "learning_rate": 1.958397305254125e-05, + "loss": 0.9618, + "step": 4132 + }, + { + "epoch": 0.18299148799574952, + "grad_norm": 1.3690160512924194, + "learning_rate": 1.95835318083348e-05, + "loss": 0.8683, + "step": 4133 + }, + { + "epoch": 0.1830357637005634, + "grad_norm": 1.5572563409805298, + "learning_rate": 1.95830903352335e-05, + "loss": 1.0172, + "step": 4134 + }, + { + "epoch": 0.1830800394053773, + "grad_norm": 1.72311270236969, + "learning_rate": 1.958264863324789e-05, + "loss": 0.915, + "step": 4135 + }, + { + "epoch": 0.18312431511019117, + "grad_norm": 1.9488872289657593, + "learning_rate": 1.958220670238852e-05, + "loss": 0.9433, + "step": 4136 + }, + { + "epoch": 0.18316859081500503, + "grad_norm": 1.1369948387145996, + "learning_rate": 1.9581764542665943e-05, + "loss": 0.8762, + "step": 4137 + }, + { + "epoch": 0.1832128665198189, + "grad_norm": 1.5573444366455078, + "learning_rate": 1.958132215409072e-05, + "loss": 1.0048, + "step": 4138 + }, + { + "epoch": 0.1832571422246328, + "grad_norm": 1.5071438550949097, + "learning_rate": 1.958087953667342e-05, + "loss": 0.8614, + "step": 4139 + }, + { + "epoch": 0.18330141792944665, + "grad_norm": 1.4512293338775635, + "learning_rate": 1.9580436690424617e-05, + "loss": 0.9505, + "step": 4140 + }, + { + "epoch": 0.18334569363426054, + "grad_norm": 1.4214322566986084, + "learning_rate": 1.957999361535488e-05, + "loss": 0.9834, + "step": 4141 + }, + { + "epoch": 0.18338996933907442, + "grad_norm": 1.5227714776992798, + "learning_rate": 1.95795503114748e-05, + "loss": 1.0166, + "step": 4142 + }, + { + "epoch": 0.1834342450438883, + "grad_norm": 1.2365552186965942, + "learning_rate": 1.9579106778794957e-05, + "loss": 0.9, + "step": 4143 + }, + { + "epoch": 0.18347852074870216, + "grad_norm": 1.4084205627441406, + "learning_rate": 1.957866301732595e-05, + "loss": 0.966, + "step": 4144 + }, + { + "epoch": 0.18352279645351605, + "grad_norm": 1.2224689722061157, + "learning_rate": 1.9578219027078376e-05, + "loss": 0.9252, + "step": 4145 + }, + { + "epoch": 0.18356707215832993, + "grad_norm": 1.342909812927246, + "learning_rate": 1.9577774808062838e-05, + "loss": 0.9118, + "step": 4146 + }, + { + "epoch": 0.1836113478631438, + "grad_norm": 1.7137248516082764, + "learning_rate": 1.957733036028995e-05, + "loss": 1.0643, + "step": 4147 + }, + { + "epoch": 0.18365562356795767, + "grad_norm": 1.2181411981582642, + "learning_rate": 1.957688568377032e-05, + "loss": 0.9161, + "step": 4148 + }, + { + "epoch": 0.18369989927277156, + "grad_norm": 1.1273655891418457, + "learning_rate": 1.9576440778514576e-05, + "loss": 0.908, + "step": 4149 + }, + { + "epoch": 0.1837441749775854, + "grad_norm": 2.2310826778411865, + "learning_rate": 1.957599564453334e-05, + "loss": 0.8696, + "step": 4150 + }, + { + "epoch": 0.1837884506823993, + "grad_norm": 1.3913973569869995, + "learning_rate": 1.957555028183725e-05, + "loss": 0.9487, + "step": 4151 + }, + { + "epoch": 0.18383272638721318, + "grad_norm": 1.210086703300476, + "learning_rate": 1.957510469043693e-05, + "loss": 0.891, + "step": 4152 + }, + { + "epoch": 0.18387700209202706, + "grad_norm": 1.5342601537704468, + "learning_rate": 1.9574658870343036e-05, + "loss": 0.9766, + "step": 4153 + }, + { + "epoch": 0.18392127779684092, + "grad_norm": 1.325132966041565, + "learning_rate": 1.9574212821566213e-05, + "loss": 0.8963, + "step": 4154 + }, + { + "epoch": 0.1839655535016548, + "grad_norm": 1.3217036724090576, + "learning_rate": 1.957376654411711e-05, + "loss": 0.9288, + "step": 4155 + }, + { + "epoch": 0.1840098292064687, + "grad_norm": 1.7246299982070923, + "learning_rate": 1.9573320038006388e-05, + "loss": 0.9143, + "step": 4156 + }, + { + "epoch": 0.18405410491128255, + "grad_norm": 1.2317532300949097, + "learning_rate": 1.957287330324471e-05, + "loss": 0.9478, + "step": 4157 + }, + { + "epoch": 0.18409838061609643, + "grad_norm": 1.6188347339630127, + "learning_rate": 1.9572426339842748e-05, + "loss": 1.0165, + "step": 4158 + }, + { + "epoch": 0.18414265632091031, + "grad_norm": 1.283898949623108, + "learning_rate": 1.957197914781118e-05, + "loss": 0.9284, + "step": 4159 + }, + { + "epoch": 0.1841869320257242, + "grad_norm": 1.3769893646240234, + "learning_rate": 1.957153172716068e-05, + "loss": 1.036, + "step": 4160 + }, + { + "epoch": 0.18423120773053805, + "grad_norm": 1.5336393117904663, + "learning_rate": 1.957108407790194e-05, + "loss": 1.0298, + "step": 4161 + }, + { + "epoch": 0.18427548343535194, + "grad_norm": 1.3408970832824707, + "learning_rate": 1.9570636200045655e-05, + "loss": 0.986, + "step": 4162 + }, + { + "epoch": 0.18431975914016582, + "grad_norm": 1.3467649221420288, + "learning_rate": 1.9570188093602512e-05, + "loss": 0.9939, + "step": 4163 + }, + { + "epoch": 0.18436403484497968, + "grad_norm": 1.3084639310836792, + "learning_rate": 1.9569739758583216e-05, + "loss": 0.9351, + "step": 4164 + }, + { + "epoch": 0.18440831054979356, + "grad_norm": 1.4206656217575073, + "learning_rate": 1.956929119499848e-05, + "loss": 0.9224, + "step": 4165 + }, + { + "epoch": 0.18445258625460745, + "grad_norm": 1.5154502391815186, + "learning_rate": 1.9568842402859017e-05, + "loss": 1.029, + "step": 4166 + }, + { + "epoch": 0.1844968619594213, + "grad_norm": 1.3837920427322388, + "learning_rate": 1.9568393382175545e-05, + "loss": 0.9942, + "step": 4167 + }, + { + "epoch": 0.1845411376642352, + "grad_norm": 1.312798023223877, + "learning_rate": 1.956794413295879e-05, + "loss": 0.8995, + "step": 4168 + }, + { + "epoch": 0.18458541336904907, + "grad_norm": 1.7064059972763062, + "learning_rate": 1.9567494655219472e-05, + "loss": 1.0503, + "step": 4169 + }, + { + "epoch": 0.18462968907386296, + "grad_norm": 1.311295509338379, + "learning_rate": 1.9567044948968342e-05, + "loss": 0.9013, + "step": 4170 + }, + { + "epoch": 0.1846739647786768, + "grad_norm": 1.4041645526885986, + "learning_rate": 1.956659501421613e-05, + "loss": 0.9798, + "step": 4171 + }, + { + "epoch": 0.1847182404834907, + "grad_norm": 1.685430645942688, + "learning_rate": 1.9566144850973584e-05, + "loss": 0.8769, + "step": 4172 + }, + { + "epoch": 0.18476251618830458, + "grad_norm": 1.3696166276931763, + "learning_rate": 1.9565694459251457e-05, + "loss": 0.9884, + "step": 4173 + }, + { + "epoch": 0.18480679189311844, + "grad_norm": 1.4926763772964478, + "learning_rate": 1.9565243839060512e-05, + "loss": 0.9287, + "step": 4174 + }, + { + "epoch": 0.18485106759793232, + "grad_norm": 1.5343453884124756, + "learning_rate": 1.95647929904115e-05, + "loss": 0.9419, + "step": 4175 + }, + { + "epoch": 0.1848953433027462, + "grad_norm": 1.2696809768676758, + "learning_rate": 1.95643419133152e-05, + "loss": 0.9043, + "step": 4176 + }, + { + "epoch": 0.18493961900756006, + "grad_norm": 1.766892910003662, + "learning_rate": 1.956389060778238e-05, + "loss": 1.1029, + "step": 4177 + }, + { + "epoch": 0.18498389471237395, + "grad_norm": 1.255121111869812, + "learning_rate": 1.9563439073823817e-05, + "loss": 0.9562, + "step": 4178 + }, + { + "epoch": 0.18502817041718783, + "grad_norm": 1.2582018375396729, + "learning_rate": 1.9562987311450298e-05, + "loss": 0.9439, + "step": 4179 + }, + { + "epoch": 0.18507244612200172, + "grad_norm": 1.249509334564209, + "learning_rate": 1.9562535320672617e-05, + "loss": 0.8876, + "step": 4180 + }, + { + "epoch": 0.18511672182681557, + "grad_norm": 1.460715889930725, + "learning_rate": 1.9562083101501564e-05, + "loss": 1.0318, + "step": 4181 + }, + { + "epoch": 0.18516099753162946, + "grad_norm": 1.2974194288253784, + "learning_rate": 1.9561630653947946e-05, + "loss": 0.8415, + "step": 4182 + }, + { + "epoch": 0.18520527323644334, + "grad_norm": 1.2734733819961548, + "learning_rate": 1.9561177978022563e-05, + "loss": 0.9286, + "step": 4183 + }, + { + "epoch": 0.1852495489412572, + "grad_norm": 1.4681172370910645, + "learning_rate": 1.9560725073736226e-05, + "loss": 0.9187, + "step": 4184 + }, + { + "epoch": 0.18529382464607108, + "grad_norm": 1.2967544794082642, + "learning_rate": 1.9560271941099758e-05, + "loss": 0.8598, + "step": 4185 + }, + { + "epoch": 0.18533810035088497, + "grad_norm": 1.5334011316299438, + "learning_rate": 1.9559818580123976e-05, + "loss": 0.9964, + "step": 4186 + }, + { + "epoch": 0.18538237605569885, + "grad_norm": 1.4243597984313965, + "learning_rate": 1.9559364990819717e-05, + "loss": 1.0474, + "step": 4187 + }, + { + "epoch": 0.1854266517605127, + "grad_norm": 1.5692945718765259, + "learning_rate": 1.9558911173197804e-05, + "loss": 0.9458, + "step": 4188 + }, + { + "epoch": 0.1854709274653266, + "grad_norm": 1.7215112447738647, + "learning_rate": 1.955845712726908e-05, + "loss": 1.0795, + "step": 4189 + }, + { + "epoch": 0.18551520317014047, + "grad_norm": 1.223763108253479, + "learning_rate": 1.9558002853044395e-05, + "loss": 0.8075, + "step": 4190 + }, + { + "epoch": 0.18555947887495433, + "grad_norm": 1.410536527633667, + "learning_rate": 1.955754835053459e-05, + "loss": 1.0041, + "step": 4191 + }, + { + "epoch": 0.18560375457976822, + "grad_norm": 1.124098539352417, + "learning_rate": 1.9557093619750528e-05, + "loss": 0.8866, + "step": 4192 + }, + { + "epoch": 0.1856480302845821, + "grad_norm": 1.8469274044036865, + "learning_rate": 1.9556638660703062e-05, + "loss": 1.005, + "step": 4193 + }, + { + "epoch": 0.18569230598939596, + "grad_norm": 1.375740647315979, + "learning_rate": 1.9556183473403066e-05, + "loss": 0.9477, + "step": 4194 + }, + { + "epoch": 0.18573658169420984, + "grad_norm": 1.3599810600280762, + "learning_rate": 1.955572805786141e-05, + "loss": 0.8991, + "step": 4195 + }, + { + "epoch": 0.18578085739902372, + "grad_norm": 1.3157724142074585, + "learning_rate": 1.9555272414088965e-05, + "loss": 0.9292, + "step": 4196 + }, + { + "epoch": 0.1858251331038376, + "grad_norm": 1.6037989854812622, + "learning_rate": 1.9554816542096626e-05, + "loss": 1.0357, + "step": 4197 + }, + { + "epoch": 0.18586940880865147, + "grad_norm": 1.592012643814087, + "learning_rate": 1.9554360441895267e-05, + "loss": 0.8676, + "step": 4198 + }, + { + "epoch": 0.18591368451346535, + "grad_norm": 1.3404552936553955, + "learning_rate": 1.955390411349579e-05, + "loss": 0.8971, + "step": 4199 + }, + { + "epoch": 0.18595796021827923, + "grad_norm": 1.3339509963989258, + "learning_rate": 1.9553447556909092e-05, + "loss": 0.9005, + "step": 4200 + }, + { + "epoch": 0.1860022359230931, + "grad_norm": 1.2520707845687866, + "learning_rate": 1.955299077214608e-05, + "loss": 0.9187, + "step": 4201 + }, + { + "epoch": 0.18604651162790697, + "grad_norm": 1.8370013236999512, + "learning_rate": 1.9552533759217663e-05, + "loss": 0.9138, + "step": 4202 + }, + { + "epoch": 0.18609078733272086, + "grad_norm": 1.712593913078308, + "learning_rate": 1.955207651813475e-05, + "loss": 0.9328, + "step": 4203 + }, + { + "epoch": 0.18613506303753474, + "grad_norm": 1.1595900058746338, + "learning_rate": 1.955161904890827e-05, + "loss": 0.8241, + "step": 4204 + }, + { + "epoch": 0.1861793387423486, + "grad_norm": 1.3777910470962524, + "learning_rate": 1.955116135154915e-05, + "loss": 0.903, + "step": 4205 + }, + { + "epoch": 0.18622361444716248, + "grad_norm": 1.118565320968628, + "learning_rate": 1.955070342606831e-05, + "loss": 0.8238, + "step": 4206 + }, + { + "epoch": 0.18626789015197637, + "grad_norm": 1.5147969722747803, + "learning_rate": 1.95502452724767e-05, + "loss": 0.995, + "step": 4207 + }, + { + "epoch": 0.18631216585679022, + "grad_norm": 1.3320332765579224, + "learning_rate": 1.9549786890785257e-05, + "loss": 0.9527, + "step": 4208 + }, + { + "epoch": 0.1863564415616041, + "grad_norm": 1.378401279449463, + "learning_rate": 1.954932828100493e-05, + "loss": 0.9759, + "step": 4209 + }, + { + "epoch": 0.186400717266418, + "grad_norm": 1.6148654222488403, + "learning_rate": 1.9548869443146673e-05, + "loss": 0.9501, + "step": 4210 + }, + { + "epoch": 0.18644499297123185, + "grad_norm": 1.668394684791565, + "learning_rate": 1.9548410377221442e-05, + "loss": 1.002, + "step": 4211 + }, + { + "epoch": 0.18648926867604573, + "grad_norm": 1.4278069734573364, + "learning_rate": 1.9547951083240204e-05, + "loss": 0.9362, + "step": 4212 + }, + { + "epoch": 0.18653354438085962, + "grad_norm": 1.3220001459121704, + "learning_rate": 1.954749156121393e-05, + "loss": 0.9671, + "step": 4213 + }, + { + "epoch": 0.1865778200856735, + "grad_norm": 1.4530644416809082, + "learning_rate": 1.9547031811153596e-05, + "loss": 1.0466, + "step": 4214 + }, + { + "epoch": 0.18662209579048736, + "grad_norm": 1.2290267944335938, + "learning_rate": 1.954657183307018e-05, + "loss": 0.9456, + "step": 4215 + }, + { + "epoch": 0.18666637149530124, + "grad_norm": 1.4291698932647705, + "learning_rate": 1.9546111626974664e-05, + "loss": 0.9131, + "step": 4216 + }, + { + "epoch": 0.18671064720011513, + "grad_norm": 1.3333709239959717, + "learning_rate": 1.954565119287805e-05, + "loss": 0.8549, + "step": 4217 + }, + { + "epoch": 0.18675492290492898, + "grad_norm": 1.4399021863937378, + "learning_rate": 1.9545190530791326e-05, + "loss": 0.988, + "step": 4218 + }, + { + "epoch": 0.18679919860974287, + "grad_norm": 1.3889105319976807, + "learning_rate": 1.95447296407255e-05, + "loss": 0.975, + "step": 4219 + }, + { + "epoch": 0.18684347431455675, + "grad_norm": 1.5275890827178955, + "learning_rate": 1.9544268522691575e-05, + "loss": 1.018, + "step": 4220 + }, + { + "epoch": 0.1868877500193706, + "grad_norm": 1.423306941986084, + "learning_rate": 1.954380717670057e-05, + "loss": 0.9847, + "step": 4221 + }, + { + "epoch": 0.1869320257241845, + "grad_norm": 1.4155516624450684, + "learning_rate": 1.9543345602763502e-05, + "loss": 0.9956, + "step": 4222 + }, + { + "epoch": 0.18697630142899838, + "grad_norm": 1.3825234174728394, + "learning_rate": 1.9542883800891396e-05, + "loss": 1.0013, + "step": 4223 + }, + { + "epoch": 0.18702057713381226, + "grad_norm": 1.2173717021942139, + "learning_rate": 1.954242177109528e-05, + "loss": 0.8812, + "step": 4224 + }, + { + "epoch": 0.18706485283862612, + "grad_norm": 1.3515745401382446, + "learning_rate": 1.9541959513386186e-05, + "loss": 0.8836, + "step": 4225 + }, + { + "epoch": 0.18710912854344, + "grad_norm": 1.4239128828048706, + "learning_rate": 1.954149702777516e-05, + "loss": 0.9742, + "step": 4226 + }, + { + "epoch": 0.18715340424825389, + "grad_norm": 1.215866208076477, + "learning_rate": 1.954103431427325e-05, + "loss": 0.9206, + "step": 4227 + }, + { + "epoch": 0.18719767995306774, + "grad_norm": 1.4924386739730835, + "learning_rate": 1.95405713728915e-05, + "loss": 1.0051, + "step": 4228 + }, + { + "epoch": 0.18724195565788163, + "grad_norm": 1.2275574207305908, + "learning_rate": 1.9540108203640976e-05, + "loss": 0.8888, + "step": 4229 + }, + { + "epoch": 0.1872862313626955, + "grad_norm": 1.2987384796142578, + "learning_rate": 1.953964480653273e-05, + "loss": 1.0196, + "step": 4230 + }, + { + "epoch": 0.1873305070675094, + "grad_norm": 1.4106584787368774, + "learning_rate": 1.9539181181577835e-05, + "loss": 0.966, + "step": 4231 + }, + { + "epoch": 0.18737478277232325, + "grad_norm": 1.2274088859558105, + "learning_rate": 1.9538717328787368e-05, + "loss": 0.8744, + "step": 4232 + }, + { + "epoch": 0.18741905847713713, + "grad_norm": 1.3131500482559204, + "learning_rate": 1.9538253248172404e-05, + "loss": 0.9879, + "step": 4233 + }, + { + "epoch": 0.18746333418195102, + "grad_norm": 1.1044342517852783, + "learning_rate": 1.9537788939744025e-05, + "loss": 0.8536, + "step": 4234 + }, + { + "epoch": 0.18750760988676488, + "grad_norm": 1.2689162492752075, + "learning_rate": 1.9537324403513325e-05, + "loss": 0.909, + "step": 4235 + }, + { + "epoch": 0.18755188559157876, + "grad_norm": 1.2226046323776245, + "learning_rate": 1.9536859639491397e-05, + "loss": 0.935, + "step": 4236 + }, + { + "epoch": 0.18759616129639264, + "grad_norm": 1.43337082862854, + "learning_rate": 1.9536394647689342e-05, + "loss": 1.01, + "step": 4237 + }, + { + "epoch": 0.1876404370012065, + "grad_norm": 1.5400134325027466, + "learning_rate": 1.9535929428118264e-05, + "loss": 1.019, + "step": 4238 + }, + { + "epoch": 0.18768471270602038, + "grad_norm": 1.5408207178115845, + "learning_rate": 1.9535463980789277e-05, + "loss": 0.961, + "step": 4239 + }, + { + "epoch": 0.18772898841083427, + "grad_norm": 1.3881274461746216, + "learning_rate": 1.95349983057135e-05, + "loss": 0.9582, + "step": 4240 + }, + { + "epoch": 0.18777326411564815, + "grad_norm": 1.083278775215149, + "learning_rate": 1.9534532402902048e-05, + "loss": 0.8582, + "step": 4241 + }, + { + "epoch": 0.187817539820462, + "grad_norm": 1.2809042930603027, + "learning_rate": 1.9534066272366053e-05, + "loss": 0.9717, + "step": 4242 + }, + { + "epoch": 0.1878618155252759, + "grad_norm": 1.1031899452209473, + "learning_rate": 1.9533599914116648e-05, + "loss": 0.8755, + "step": 4243 + }, + { + "epoch": 0.18790609123008978, + "grad_norm": 1.281378149986267, + "learning_rate": 1.9533133328164973e-05, + "loss": 0.9625, + "step": 4244 + }, + { + "epoch": 0.18795036693490363, + "grad_norm": 1.349589467048645, + "learning_rate": 1.9532666514522172e-05, + "loss": 0.9252, + "step": 4245 + }, + { + "epoch": 0.18799464263971752, + "grad_norm": 1.5421642065048218, + "learning_rate": 1.953219947319939e-05, + "loss": 1.082, + "step": 4246 + }, + { + "epoch": 0.1880389183445314, + "grad_norm": 1.2708654403686523, + "learning_rate": 1.9531732204207787e-05, + "loss": 0.8818, + "step": 4247 + }, + { + "epoch": 0.18808319404934526, + "grad_norm": 1.336027979850769, + "learning_rate": 1.9531264707558522e-05, + "loss": 0.9559, + "step": 4248 + }, + { + "epoch": 0.18812746975415914, + "grad_norm": 1.7550179958343506, + "learning_rate": 1.953079698326276e-05, + "loss": 1.0002, + "step": 4249 + }, + { + "epoch": 0.18817174545897303, + "grad_norm": 1.3733514547348022, + "learning_rate": 1.953032903133167e-05, + "loss": 0.984, + "step": 4250 + }, + { + "epoch": 0.1882160211637869, + "grad_norm": 1.3112117052078247, + "learning_rate": 1.952986085177643e-05, + "loss": 0.8963, + "step": 4251 + }, + { + "epoch": 0.18826029686860077, + "grad_norm": 1.3522695302963257, + "learning_rate": 1.9529392444608225e-05, + "loss": 0.8596, + "step": 4252 + }, + { + "epoch": 0.18830457257341465, + "grad_norm": 1.3055187463760376, + "learning_rate": 1.952892380983824e-05, + "loss": 1.0042, + "step": 4253 + }, + { + "epoch": 0.18834884827822854, + "grad_norm": 1.388005018234253, + "learning_rate": 1.9528454947477674e-05, + "loss": 0.8959, + "step": 4254 + }, + { + "epoch": 0.1883931239830424, + "grad_norm": 1.4843909740447998, + "learning_rate": 1.9527985857537713e-05, + "loss": 1.0426, + "step": 4255 + }, + { + "epoch": 0.18843739968785628, + "grad_norm": 1.1777918338775635, + "learning_rate": 1.952751654002957e-05, + "loss": 0.9124, + "step": 4256 + }, + { + "epoch": 0.18848167539267016, + "grad_norm": 1.3612449169158936, + "learning_rate": 1.952704699496445e-05, + "loss": 0.9387, + "step": 4257 + }, + { + "epoch": 0.18852595109748405, + "grad_norm": 1.2322074174880981, + "learning_rate": 1.9526577222353573e-05, + "loss": 0.8723, + "step": 4258 + }, + { + "epoch": 0.1885702268022979, + "grad_norm": 1.215620756149292, + "learning_rate": 1.952610722220816e-05, + "loss": 0.9997, + "step": 4259 + }, + { + "epoch": 0.1886145025071118, + "grad_norm": 1.2898924350738525, + "learning_rate": 1.9525636994539422e-05, + "loss": 0.9227, + "step": 4260 + }, + { + "epoch": 0.18865877821192567, + "grad_norm": 1.397615671157837, + "learning_rate": 1.9525166539358608e-05, + "loss": 1.0267, + "step": 4261 + }, + { + "epoch": 0.18870305391673953, + "grad_norm": 1.4901931285858154, + "learning_rate": 1.9524695856676942e-05, + "loss": 0.9307, + "step": 4262 + }, + { + "epoch": 0.1887473296215534, + "grad_norm": 1.7261799573898315, + "learning_rate": 1.9524224946505675e-05, + "loss": 0.9411, + "step": 4263 + }, + { + "epoch": 0.1887916053263673, + "grad_norm": 1.2903856039047241, + "learning_rate": 1.9523753808856047e-05, + "loss": 0.8293, + "step": 4264 + }, + { + "epoch": 0.18883588103118115, + "grad_norm": 1.224227786064148, + "learning_rate": 1.9523282443739315e-05, + "loss": 0.8133, + "step": 4265 + }, + { + "epoch": 0.18888015673599504, + "grad_norm": 1.3248907327651978, + "learning_rate": 1.9522810851166735e-05, + "loss": 0.8926, + "step": 4266 + }, + { + "epoch": 0.18892443244080892, + "grad_norm": 1.799071192741394, + "learning_rate": 1.9522339031149573e-05, + "loss": 0.9856, + "step": 4267 + }, + { + "epoch": 0.1889687081456228, + "grad_norm": 1.6799441576004028, + "learning_rate": 1.9521866983699098e-05, + "loss": 0.9522, + "step": 4268 + }, + { + "epoch": 0.18901298385043666, + "grad_norm": 1.4426956176757812, + "learning_rate": 1.9521394708826582e-05, + "loss": 0.9606, + "step": 4269 + }, + { + "epoch": 0.18905725955525055, + "grad_norm": 1.2468069791793823, + "learning_rate": 1.9520922206543304e-05, + "loss": 0.9219, + "step": 4270 + }, + { + "epoch": 0.18910153526006443, + "grad_norm": 1.7105740308761597, + "learning_rate": 1.9520449476860553e-05, + "loss": 1.1026, + "step": 4271 + }, + { + "epoch": 0.18914581096487829, + "grad_norm": 1.5898858308792114, + "learning_rate": 1.9519976519789615e-05, + "loss": 1.0328, + "step": 4272 + }, + { + "epoch": 0.18919008666969217, + "grad_norm": 1.378657341003418, + "learning_rate": 1.9519503335341797e-05, + "loss": 0.9487, + "step": 4273 + }, + { + "epoch": 0.18923436237450605, + "grad_norm": 1.4150300025939941, + "learning_rate": 1.9519029923528384e-05, + "loss": 0.8523, + "step": 4274 + }, + { + "epoch": 0.18927863807931994, + "grad_norm": 1.4220314025878906, + "learning_rate": 1.9518556284360696e-05, + "loss": 0.9469, + "step": 4275 + }, + { + "epoch": 0.1893229137841338, + "grad_norm": 1.1722887754440308, + "learning_rate": 1.9518082417850043e-05, + "loss": 0.8623, + "step": 4276 + }, + { + "epoch": 0.18936718948894768, + "grad_norm": 1.2307404279708862, + "learning_rate": 1.951760832400774e-05, + "loss": 0.9203, + "step": 4277 + }, + { + "epoch": 0.18941146519376156, + "grad_norm": 1.6730895042419434, + "learning_rate": 1.9517134002845112e-05, + "loss": 1.0493, + "step": 4278 + }, + { + "epoch": 0.18945574089857542, + "grad_norm": 1.2626392841339111, + "learning_rate": 1.9516659454373484e-05, + "loss": 0.9433, + "step": 4279 + }, + { + "epoch": 0.1895000166033893, + "grad_norm": 1.388687014579773, + "learning_rate": 1.9516184678604197e-05, + "loss": 0.9778, + "step": 4280 + }, + { + "epoch": 0.1895442923082032, + "grad_norm": 1.5867871046066284, + "learning_rate": 1.9515709675548587e-05, + "loss": 0.9719, + "step": 4281 + }, + { + "epoch": 0.18958856801301704, + "grad_norm": 1.1585452556610107, + "learning_rate": 1.9515234445218002e-05, + "loss": 0.8432, + "step": 4282 + }, + { + "epoch": 0.18963284371783093, + "grad_norm": 2.287036657333374, + "learning_rate": 1.9514758987623784e-05, + "loss": 1.0782, + "step": 4283 + }, + { + "epoch": 0.1896771194226448, + "grad_norm": 1.6369024515151978, + "learning_rate": 1.9514283302777298e-05, + "loss": 0.9403, + "step": 4284 + }, + { + "epoch": 0.1897213951274587, + "grad_norm": 1.3030041456222534, + "learning_rate": 1.9513807390689903e-05, + "loss": 0.9007, + "step": 4285 + }, + { + "epoch": 0.18976567083227255, + "grad_norm": 1.7366888523101807, + "learning_rate": 1.9513331251372965e-05, + "loss": 0.9179, + "step": 4286 + }, + { + "epoch": 0.18980994653708644, + "grad_norm": 1.3750731945037842, + "learning_rate": 1.9512854884837856e-05, + "loss": 1.0033, + "step": 4287 + }, + { + "epoch": 0.18985422224190032, + "grad_norm": 1.2658950090408325, + "learning_rate": 1.9512378291095953e-05, + "loss": 0.9036, + "step": 4288 + }, + { + "epoch": 0.18989849794671418, + "grad_norm": 1.1469498872756958, + "learning_rate": 1.951190147015864e-05, + "loss": 0.8861, + "step": 4289 + }, + { + "epoch": 0.18994277365152806, + "grad_norm": 1.2747516632080078, + "learning_rate": 1.9511424422037303e-05, + "loss": 0.8593, + "step": 4290 + }, + { + "epoch": 0.18998704935634195, + "grad_norm": 1.2322778701782227, + "learning_rate": 1.951094714674334e-05, + "loss": 0.938, + "step": 4291 + }, + { + "epoch": 0.1900313250611558, + "grad_norm": 1.4117802381515503, + "learning_rate": 1.9510469644288148e-05, + "loss": 0.9915, + "step": 4292 + }, + { + "epoch": 0.1900756007659697, + "grad_norm": 1.6469818353652954, + "learning_rate": 1.9509991914683134e-05, + "loss": 0.9753, + "step": 4293 + }, + { + "epoch": 0.19011987647078357, + "grad_norm": 1.3718427419662476, + "learning_rate": 1.9509513957939707e-05, + "loss": 0.9751, + "step": 4294 + }, + { + "epoch": 0.19016415217559746, + "grad_norm": 1.272661566734314, + "learning_rate": 1.9509035774069277e-05, + "loss": 0.8899, + "step": 4295 + }, + { + "epoch": 0.1902084278804113, + "grad_norm": 1.6335480213165283, + "learning_rate": 1.9508557363083272e-05, + "loss": 0.9744, + "step": 4296 + }, + { + "epoch": 0.1902527035852252, + "grad_norm": 1.3846832513809204, + "learning_rate": 1.9508078724993118e-05, + "loss": 0.9662, + "step": 4297 + }, + { + "epoch": 0.19029697929003908, + "grad_norm": 1.3603832721710205, + "learning_rate": 1.9507599859810246e-05, + "loss": 0.9259, + "step": 4298 + }, + { + "epoch": 0.19034125499485294, + "grad_norm": 1.2694392204284668, + "learning_rate": 1.950712076754609e-05, + "loss": 0.8651, + "step": 4299 + }, + { + "epoch": 0.19038553069966682, + "grad_norm": 1.1997028589248657, + "learning_rate": 1.95066414482121e-05, + "loss": 0.9275, + "step": 4300 + }, + { + "epoch": 0.1904298064044807, + "grad_norm": 1.6843189001083374, + "learning_rate": 1.950616190181972e-05, + "loss": 0.9958, + "step": 4301 + }, + { + "epoch": 0.1904740821092946, + "grad_norm": 1.1524726152420044, + "learning_rate": 1.95056821283804e-05, + "loss": 0.974, + "step": 4302 + }, + { + "epoch": 0.19051835781410845, + "grad_norm": 1.6249070167541504, + "learning_rate": 1.95052021279056e-05, + "loss": 0.8541, + "step": 4303 + }, + { + "epoch": 0.19056263351892233, + "grad_norm": 1.5284347534179688, + "learning_rate": 1.9504721900406788e-05, + "loss": 0.9289, + "step": 4304 + }, + { + "epoch": 0.19060690922373622, + "grad_norm": 1.6288549900054932, + "learning_rate": 1.9504241445895434e-05, + "loss": 0.974, + "step": 4305 + }, + { + "epoch": 0.19065118492855007, + "grad_norm": 1.4943454265594482, + "learning_rate": 1.950376076438301e-05, + "loss": 0.9471, + "step": 4306 + }, + { + "epoch": 0.19069546063336396, + "grad_norm": 1.163864254951477, + "learning_rate": 1.9503279855881002e-05, + "loss": 0.9704, + "step": 4307 + }, + { + "epoch": 0.19073973633817784, + "grad_norm": 1.7004190683364868, + "learning_rate": 1.9502798720400888e-05, + "loss": 0.9286, + "step": 4308 + }, + { + "epoch": 0.1907840120429917, + "grad_norm": 1.7952994108200073, + "learning_rate": 1.950231735795417e-05, + "loss": 0.9789, + "step": 4309 + }, + { + "epoch": 0.19082828774780558, + "grad_norm": 1.3075436353683472, + "learning_rate": 1.9501835768552332e-05, + "loss": 1.0592, + "step": 4310 + }, + { + "epoch": 0.19087256345261946, + "grad_norm": 1.2196851968765259, + "learning_rate": 1.9501353952206888e-05, + "loss": 0.9257, + "step": 4311 + }, + { + "epoch": 0.19091683915743335, + "grad_norm": 1.537147045135498, + "learning_rate": 1.950087190892934e-05, + "loss": 1.0283, + "step": 4312 + }, + { + "epoch": 0.1909611148622472, + "grad_norm": 2.0402021408081055, + "learning_rate": 1.9500389638731198e-05, + "loss": 1.0462, + "step": 4313 + }, + { + "epoch": 0.1910053905670611, + "grad_norm": 1.200995683670044, + "learning_rate": 1.9499907141623992e-05, + "loss": 0.8602, + "step": 4314 + }, + { + "epoch": 0.19104966627187497, + "grad_norm": 1.2927156686782837, + "learning_rate": 1.9499424417619234e-05, + "loss": 0.951, + "step": 4315 + }, + { + "epoch": 0.19109394197668883, + "grad_norm": 1.3145122528076172, + "learning_rate": 1.9498941466728462e-05, + "loss": 0.838, + "step": 4316 + }, + { + "epoch": 0.19113821768150271, + "grad_norm": 1.2259713411331177, + "learning_rate": 1.9498458288963204e-05, + "loss": 0.8673, + "step": 4317 + }, + { + "epoch": 0.1911824933863166, + "grad_norm": 1.3289045095443726, + "learning_rate": 1.949797488433501e-05, + "loss": 0.9641, + "step": 4318 + }, + { + "epoch": 0.19122676909113048, + "grad_norm": 1.6661314964294434, + "learning_rate": 1.949749125285541e-05, + "loss": 1.028, + "step": 4319 + }, + { + "epoch": 0.19127104479594434, + "grad_norm": 1.6622711420059204, + "learning_rate": 1.9497007394535972e-05, + "loss": 1.0141, + "step": 4320 + }, + { + "epoch": 0.19131532050075822, + "grad_norm": 1.3351212739944458, + "learning_rate": 1.949652330938824e-05, + "loss": 0.8993, + "step": 4321 + }, + { + "epoch": 0.1913595962055721, + "grad_norm": 1.3790476322174072, + "learning_rate": 1.9496038997423784e-05, + "loss": 0.9507, + "step": 4322 + }, + { + "epoch": 0.19140387191038596, + "grad_norm": 1.4873906373977661, + "learning_rate": 1.949555445865417e-05, + "loss": 1.0336, + "step": 4323 + }, + { + "epoch": 0.19144814761519985, + "grad_norm": 1.6448252201080322, + "learning_rate": 1.9495069693090968e-05, + "loss": 1.0898, + "step": 4324 + }, + { + "epoch": 0.19149242332001373, + "grad_norm": 1.1424187421798706, + "learning_rate": 1.9494584700745757e-05, + "loss": 0.8266, + "step": 4325 + }, + { + "epoch": 0.1915366990248276, + "grad_norm": 1.4879472255706787, + "learning_rate": 1.949409948163012e-05, + "loss": 1.0151, + "step": 4326 + }, + { + "epoch": 0.19158097472964147, + "grad_norm": 1.3053029775619507, + "learning_rate": 1.9493614035755647e-05, + "loss": 0.9055, + "step": 4327 + }, + { + "epoch": 0.19162525043445536, + "grad_norm": 1.2269984483718872, + "learning_rate": 1.949312836313394e-05, + "loss": 0.8716, + "step": 4328 + }, + { + "epoch": 0.19166952613926924, + "grad_norm": 1.4211945533752441, + "learning_rate": 1.9492642463776582e-05, + "loss": 0.9559, + "step": 4329 + }, + { + "epoch": 0.1917138018440831, + "grad_norm": 1.203405499458313, + "learning_rate": 1.9492156337695193e-05, + "loss": 0.9098, + "step": 4330 + }, + { + "epoch": 0.19175807754889698, + "grad_norm": 1.4246326684951782, + "learning_rate": 1.9491669984901377e-05, + "loss": 0.9435, + "step": 4331 + }, + { + "epoch": 0.19180235325371087, + "grad_norm": 1.3657959699630737, + "learning_rate": 1.9491183405406754e-05, + "loss": 0.9808, + "step": 4332 + }, + { + "epoch": 0.19184662895852472, + "grad_norm": 1.6354551315307617, + "learning_rate": 1.949069659922294e-05, + "loss": 0.9831, + "step": 4333 + }, + { + "epoch": 0.1918909046633386, + "grad_norm": 1.3361481428146362, + "learning_rate": 1.9490209566361567e-05, + "loss": 0.893, + "step": 4334 + }, + { + "epoch": 0.1919351803681525, + "grad_norm": 1.6010637283325195, + "learning_rate": 1.9489722306834267e-05, + "loss": 0.9582, + "step": 4335 + }, + { + "epoch": 0.19197945607296635, + "grad_norm": 1.2955536842346191, + "learning_rate": 1.9489234820652675e-05, + "loss": 0.9938, + "step": 4336 + }, + { + "epoch": 0.19202373177778023, + "grad_norm": 1.299123764038086, + "learning_rate": 1.9488747107828433e-05, + "loss": 1.0189, + "step": 4337 + }, + { + "epoch": 0.19206800748259412, + "grad_norm": 1.2835674285888672, + "learning_rate": 1.9488259168373198e-05, + "loss": 0.9587, + "step": 4338 + }, + { + "epoch": 0.192112283187408, + "grad_norm": 1.4952843189239502, + "learning_rate": 1.9487771002298618e-05, + "loss": 1.0368, + "step": 4339 + }, + { + "epoch": 0.19215655889222186, + "grad_norm": 1.3922131061553955, + "learning_rate": 1.9487282609616347e-05, + "loss": 0.9343, + "step": 4340 + }, + { + "epoch": 0.19220083459703574, + "grad_norm": 1.3189055919647217, + "learning_rate": 1.948679399033806e-05, + "loss": 0.9571, + "step": 4341 + }, + { + "epoch": 0.19224511030184963, + "grad_norm": 1.3033177852630615, + "learning_rate": 1.9486305144475425e-05, + "loss": 0.964, + "step": 4342 + }, + { + "epoch": 0.19228938600666348, + "grad_norm": 1.7781898975372314, + "learning_rate": 1.948581607204011e-05, + "loss": 1.0252, + "step": 4343 + }, + { + "epoch": 0.19233366171147737, + "grad_norm": 1.5787416696548462, + "learning_rate": 1.9485326773043805e-05, + "loss": 0.9333, + "step": 4344 + }, + { + "epoch": 0.19237793741629125, + "grad_norm": 1.203331470489502, + "learning_rate": 1.9484837247498194e-05, + "loss": 0.9387, + "step": 4345 + }, + { + "epoch": 0.19242221312110513, + "grad_norm": 1.2669693231582642, + "learning_rate": 1.948434749541497e-05, + "loss": 0.8995, + "step": 4346 + }, + { + "epoch": 0.192466488825919, + "grad_norm": 1.2691386938095093, + "learning_rate": 1.9483857516805823e-05, + "loss": 0.9224, + "step": 4347 + }, + { + "epoch": 0.19251076453073288, + "grad_norm": 1.6296329498291016, + "learning_rate": 1.9483367311682463e-05, + "loss": 1.0881, + "step": 4348 + }, + { + "epoch": 0.19255504023554676, + "grad_norm": 1.2704613208770752, + "learning_rate": 1.9482876880056597e-05, + "loss": 0.9167, + "step": 4349 + }, + { + "epoch": 0.19259931594036062, + "grad_norm": 1.293961763381958, + "learning_rate": 1.948238622193994e-05, + "loss": 0.8975, + "step": 4350 + }, + { + "epoch": 0.1926435916451745, + "grad_norm": 1.4592667818069458, + "learning_rate": 1.9481895337344204e-05, + "loss": 1.0041, + "step": 4351 + }, + { + "epoch": 0.19268786734998838, + "grad_norm": 1.2982217073440552, + "learning_rate": 1.9481404226281125e-05, + "loss": 0.9091, + "step": 4352 + }, + { + "epoch": 0.19273214305480224, + "grad_norm": 1.2826967239379883, + "learning_rate": 1.9480912888762423e-05, + "loss": 0.8959, + "step": 4353 + }, + { + "epoch": 0.19277641875961612, + "grad_norm": 3.1292672157287598, + "learning_rate": 1.9480421324799837e-05, + "loss": 0.9561, + "step": 4354 + }, + { + "epoch": 0.19282069446443, + "grad_norm": 1.4827122688293457, + "learning_rate": 1.9479929534405103e-05, + "loss": 0.9352, + "step": 4355 + }, + { + "epoch": 0.1928649701692439, + "grad_norm": 1.2624588012695312, + "learning_rate": 1.9479437517589975e-05, + "loss": 0.8747, + "step": 4356 + }, + { + "epoch": 0.19290924587405775, + "grad_norm": 1.386674165725708, + "learning_rate": 1.94789452743662e-05, + "loss": 0.9621, + "step": 4357 + }, + { + "epoch": 0.19295352157887163, + "grad_norm": 1.2825322151184082, + "learning_rate": 1.9478452804745534e-05, + "loss": 0.8996, + "step": 4358 + }, + { + "epoch": 0.19299779728368552, + "grad_norm": 1.1532162427902222, + "learning_rate": 1.947796010873974e-05, + "loss": 0.8562, + "step": 4359 + }, + { + "epoch": 0.19304207298849937, + "grad_norm": 1.3068726062774658, + "learning_rate": 1.947746718636059e-05, + "loss": 0.9091, + "step": 4360 + }, + { + "epoch": 0.19308634869331326, + "grad_norm": 1.3419169187545776, + "learning_rate": 1.9476974037619846e-05, + "loss": 0.9793, + "step": 4361 + }, + { + "epoch": 0.19313062439812714, + "grad_norm": 1.425368070602417, + "learning_rate": 1.94764806625293e-05, + "loss": 0.8457, + "step": 4362 + }, + { + "epoch": 0.19317490010294103, + "grad_norm": 1.2636938095092773, + "learning_rate": 1.9475987061100724e-05, + "loss": 0.8677, + "step": 4363 + }, + { + "epoch": 0.19321917580775488, + "grad_norm": 1.420400857925415, + "learning_rate": 1.9475493233345916e-05, + "loss": 0.9068, + "step": 4364 + }, + { + "epoch": 0.19326345151256877, + "grad_norm": 1.1936883926391602, + "learning_rate": 1.9474999179276668e-05, + "loss": 0.8048, + "step": 4365 + }, + { + "epoch": 0.19330772721738265, + "grad_norm": 1.4276024103164673, + "learning_rate": 1.947450489890478e-05, + "loss": 0.9859, + "step": 4366 + }, + { + "epoch": 0.1933520029221965, + "grad_norm": 1.355821132659912, + "learning_rate": 1.9474010392242053e-05, + "loss": 0.9804, + "step": 4367 + }, + { + "epoch": 0.1933962786270104, + "grad_norm": 1.7368319034576416, + "learning_rate": 1.9473515659300306e-05, + "loss": 0.9627, + "step": 4368 + }, + { + "epoch": 0.19344055433182428, + "grad_norm": 1.1994789838790894, + "learning_rate": 1.9473020700091348e-05, + "loss": 0.906, + "step": 4369 + }, + { + "epoch": 0.19348483003663813, + "grad_norm": 2.674104928970337, + "learning_rate": 1.9472525514627002e-05, + "loss": 0.9581, + "step": 4370 + }, + { + "epoch": 0.19352910574145202, + "grad_norm": 1.280554175376892, + "learning_rate": 1.9472030102919102e-05, + "loss": 0.939, + "step": 4371 + }, + { + "epoch": 0.1935733814462659, + "grad_norm": 1.1514229774475098, + "learning_rate": 1.9471534464979473e-05, + "loss": 0.8584, + "step": 4372 + }, + { + "epoch": 0.1936176571510798, + "grad_norm": 1.3904298543930054, + "learning_rate": 1.9471038600819956e-05, + "loss": 0.9217, + "step": 4373 + }, + { + "epoch": 0.19366193285589364, + "grad_norm": 1.4596459865570068, + "learning_rate": 1.9470542510452396e-05, + "loss": 0.8502, + "step": 4374 + }, + { + "epoch": 0.19370620856070753, + "grad_norm": 1.3944308757781982, + "learning_rate": 1.9470046193888634e-05, + "loss": 1.0196, + "step": 4375 + }, + { + "epoch": 0.1937504842655214, + "grad_norm": 5.617145538330078, + "learning_rate": 1.946954965114053e-05, + "loss": 0.9237, + "step": 4376 + }, + { + "epoch": 0.19379475997033527, + "grad_norm": 1.404355525970459, + "learning_rate": 1.9469052882219944e-05, + "loss": 0.9685, + "step": 4377 + }, + { + "epoch": 0.19383903567514915, + "grad_norm": 1.216353416442871, + "learning_rate": 1.9468555887138743e-05, + "loss": 0.8973, + "step": 4378 + }, + { + "epoch": 0.19388331137996304, + "grad_norm": 1.5322461128234863, + "learning_rate": 1.946805866590879e-05, + "loss": 0.9372, + "step": 4379 + }, + { + "epoch": 0.1939275870847769, + "grad_norm": 1.2174036502838135, + "learning_rate": 1.9467561218541968e-05, + "loss": 0.9048, + "step": 4380 + }, + { + "epoch": 0.19397186278959078, + "grad_norm": 1.3070671558380127, + "learning_rate": 1.9467063545050154e-05, + "loss": 0.9728, + "step": 4381 + }, + { + "epoch": 0.19401613849440466, + "grad_norm": 1.238576054573059, + "learning_rate": 1.9466565645445237e-05, + "loss": 0.8503, + "step": 4382 + }, + { + "epoch": 0.19406041419921854, + "grad_norm": 1.3119946718215942, + "learning_rate": 1.9466067519739108e-05, + "loss": 0.894, + "step": 4383 + }, + { + "epoch": 0.1941046899040324, + "grad_norm": 1.231610894203186, + "learning_rate": 1.9465569167943663e-05, + "loss": 0.8649, + "step": 4384 + }, + { + "epoch": 0.19414896560884629, + "grad_norm": 1.2992029190063477, + "learning_rate": 1.9465070590070804e-05, + "loss": 0.8899, + "step": 4385 + }, + { + "epoch": 0.19419324131366017, + "grad_norm": 1.1805509328842163, + "learning_rate": 1.946457178613244e-05, + "loss": 0.9119, + "step": 4386 + }, + { + "epoch": 0.19423751701847403, + "grad_norm": 1.5308610200881958, + "learning_rate": 1.9464072756140487e-05, + "loss": 0.9432, + "step": 4387 + }, + { + "epoch": 0.1942817927232879, + "grad_norm": 1.2172480821609497, + "learning_rate": 1.9463573500106864e-05, + "loss": 0.8396, + "step": 4388 + }, + { + "epoch": 0.1943260684281018, + "grad_norm": 1.2522666454315186, + "learning_rate": 1.9463074018043488e-05, + "loss": 0.9646, + "step": 4389 + }, + { + "epoch": 0.19437034413291568, + "grad_norm": 1.4711263179779053, + "learning_rate": 1.9462574309962296e-05, + "loss": 0.9489, + "step": 4390 + }, + { + "epoch": 0.19441461983772954, + "grad_norm": 1.7133673429489136, + "learning_rate": 1.9462074375875224e-05, + "loss": 0.8799, + "step": 4391 + }, + { + "epoch": 0.19445889554254342, + "grad_norm": 1.2710566520690918, + "learning_rate": 1.946157421579421e-05, + "loss": 0.9106, + "step": 4392 + }, + { + "epoch": 0.1945031712473573, + "grad_norm": 1.384620189666748, + "learning_rate": 1.9461073829731197e-05, + "loss": 0.9294, + "step": 4393 + }, + { + "epoch": 0.19454744695217116, + "grad_norm": 1.3864612579345703, + "learning_rate": 1.946057321769814e-05, + "loss": 0.9866, + "step": 4394 + }, + { + "epoch": 0.19459172265698504, + "grad_norm": 1.2355453968048096, + "learning_rate": 1.9460072379706995e-05, + "loss": 0.8893, + "step": 4395 + }, + { + "epoch": 0.19463599836179893, + "grad_norm": 8.829521179199219, + "learning_rate": 1.9459571315769725e-05, + "loss": 1.1369, + "step": 4396 + }, + { + "epoch": 0.19468027406661279, + "grad_norm": 1.3845324516296387, + "learning_rate": 1.9459070025898296e-05, + "loss": 0.9623, + "step": 4397 + }, + { + "epoch": 0.19472454977142667, + "grad_norm": 1.5030831098556519, + "learning_rate": 1.945856851010468e-05, + "loss": 0.9854, + "step": 4398 + }, + { + "epoch": 0.19476882547624055, + "grad_norm": 1.231391191482544, + "learning_rate": 1.9458066768400856e-05, + "loss": 0.9232, + "step": 4399 + }, + { + "epoch": 0.19481310118105444, + "grad_norm": 1.2290353775024414, + "learning_rate": 1.9457564800798812e-05, + "loss": 0.8739, + "step": 4400 + }, + { + "epoch": 0.1948573768858683, + "grad_norm": 1.2889201641082764, + "learning_rate": 1.9457062607310527e-05, + "loss": 0.8663, + "step": 4401 + }, + { + "epoch": 0.19490165259068218, + "grad_norm": 1.8226685523986816, + "learning_rate": 1.945656018794801e-05, + "loss": 1.005, + "step": 4402 + }, + { + "epoch": 0.19494592829549606, + "grad_norm": 1.431657314300537, + "learning_rate": 1.945605754272325e-05, + "loss": 0.9079, + "step": 4403 + }, + { + "epoch": 0.19499020400030992, + "grad_norm": 1.4106839895248413, + "learning_rate": 1.9455554671648248e-05, + "loss": 0.9565, + "step": 4404 + }, + { + "epoch": 0.1950344797051238, + "grad_norm": 1.302459478378296, + "learning_rate": 1.945505157473503e-05, + "loss": 0.8935, + "step": 4405 + }, + { + "epoch": 0.1950787554099377, + "grad_norm": 1.7689543962478638, + "learning_rate": 1.94545482519956e-05, + "loss": 0.9933, + "step": 4406 + }, + { + "epoch": 0.19512303111475157, + "grad_norm": 1.8319867849349976, + "learning_rate": 1.9454044703441985e-05, + "loss": 1.0031, + "step": 4407 + }, + { + "epoch": 0.19516730681956543, + "grad_norm": 1.3297569751739502, + "learning_rate": 1.945354092908621e-05, + "loss": 0.9008, + "step": 4408 + }, + { + "epoch": 0.1952115825243793, + "grad_norm": 2.9140679836273193, + "learning_rate": 1.9453036928940305e-05, + "loss": 1.0097, + "step": 4409 + }, + { + "epoch": 0.1952558582291932, + "grad_norm": 1.6338249444961548, + "learning_rate": 1.945253270301631e-05, + "loss": 1.0293, + "step": 4410 + }, + { + "epoch": 0.19530013393400705, + "grad_norm": 1.687406301498413, + "learning_rate": 1.9452028251326272e-05, + "loss": 0.9909, + "step": 4411 + }, + { + "epoch": 0.19534440963882094, + "grad_norm": 1.4438865184783936, + "learning_rate": 1.945152357388223e-05, + "loss": 1.0517, + "step": 4412 + }, + { + "epoch": 0.19538868534363482, + "grad_norm": 1.3335338830947876, + "learning_rate": 1.945101867069625e-05, + "loss": 0.8503, + "step": 4413 + }, + { + "epoch": 0.19543296104844868, + "grad_norm": 1.3084938526153564, + "learning_rate": 1.9450513541780378e-05, + "loss": 0.9119, + "step": 4414 + }, + { + "epoch": 0.19547723675326256, + "grad_norm": 1.1583749055862427, + "learning_rate": 1.9450008187146685e-05, + "loss": 0.8538, + "step": 4415 + }, + { + "epoch": 0.19552151245807645, + "grad_norm": 1.1336517333984375, + "learning_rate": 1.9449502606807244e-05, + "loss": 0.8515, + "step": 4416 + }, + { + "epoch": 0.19556578816289033, + "grad_norm": 1.385361671447754, + "learning_rate": 1.9448996800774125e-05, + "loss": 0.9995, + "step": 4417 + }, + { + "epoch": 0.1956100638677042, + "grad_norm": 1.1746784448623657, + "learning_rate": 1.9448490769059415e-05, + "loss": 0.8511, + "step": 4418 + }, + { + "epoch": 0.19565433957251807, + "grad_norm": 1.7796000242233276, + "learning_rate": 1.9447984511675192e-05, + "loss": 0.99, + "step": 4419 + }, + { + "epoch": 0.19569861527733196, + "grad_norm": 1.5437349081039429, + "learning_rate": 1.9447478028633555e-05, + "loss": 1.0207, + "step": 4420 + }, + { + "epoch": 0.1957428909821458, + "grad_norm": 1.5567781925201416, + "learning_rate": 1.9446971319946593e-05, + "loss": 1.0245, + "step": 4421 + }, + { + "epoch": 0.1957871666869597, + "grad_norm": 1.5094462633132935, + "learning_rate": 1.944646438562642e-05, + "loss": 1.0148, + "step": 4422 + }, + { + "epoch": 0.19583144239177358, + "grad_norm": 1.3155477046966553, + "learning_rate": 1.944595722568513e-05, + "loss": 0.9581, + "step": 4423 + }, + { + "epoch": 0.19587571809658744, + "grad_norm": 1.617250680923462, + "learning_rate": 1.9445449840134846e-05, + "loss": 1.0311, + "step": 4424 + }, + { + "epoch": 0.19591999380140132, + "grad_norm": 1.1745473146438599, + "learning_rate": 1.9444942228987682e-05, + "loss": 0.9008, + "step": 4425 + }, + { + "epoch": 0.1959642695062152, + "grad_norm": 1.257907509803772, + "learning_rate": 1.9444434392255767e-05, + "loss": 0.9043, + "step": 4426 + }, + { + "epoch": 0.1960085452110291, + "grad_norm": 1.2297190427780151, + "learning_rate": 1.9443926329951225e-05, + "loss": 0.8352, + "step": 4427 + }, + { + "epoch": 0.19605282091584295, + "grad_norm": 1.2663987874984741, + "learning_rate": 1.9443418042086187e-05, + "loss": 0.8646, + "step": 4428 + }, + { + "epoch": 0.19609709662065683, + "grad_norm": 1.311144232749939, + "learning_rate": 1.9442909528672803e-05, + "loss": 0.9173, + "step": 4429 + }, + { + "epoch": 0.19614137232547071, + "grad_norm": 1.5907844305038452, + "learning_rate": 1.9442400789723215e-05, + "loss": 0.9751, + "step": 4430 + }, + { + "epoch": 0.19618564803028457, + "grad_norm": 1.4973158836364746, + "learning_rate": 1.9441891825249567e-05, + "loss": 1.0727, + "step": 4431 + }, + { + "epoch": 0.19622992373509845, + "grad_norm": 1.0927246809005737, + "learning_rate": 1.9441382635264025e-05, + "loss": 0.8103, + "step": 4432 + }, + { + "epoch": 0.19627419943991234, + "grad_norm": 1.3279848098754883, + "learning_rate": 1.9440873219778743e-05, + "loss": 0.9171, + "step": 4433 + }, + { + "epoch": 0.19631847514472622, + "grad_norm": 1.1255714893341064, + "learning_rate": 1.9440363578805893e-05, + "loss": 0.8495, + "step": 4434 + }, + { + "epoch": 0.19636275084954008, + "grad_norm": 1.123224139213562, + "learning_rate": 1.9439853712357648e-05, + "loss": 0.8475, + "step": 4435 + }, + { + "epoch": 0.19640702655435396, + "grad_norm": 1.31903076171875, + "learning_rate": 1.9439343620446182e-05, + "loss": 0.9138, + "step": 4436 + }, + { + "epoch": 0.19645130225916785, + "grad_norm": 1.5420595407485962, + "learning_rate": 1.9438833303083677e-05, + "loss": 0.9254, + "step": 4437 + }, + { + "epoch": 0.1964955779639817, + "grad_norm": 1.737859845161438, + "learning_rate": 1.9438322760282323e-05, + "loss": 0.9549, + "step": 4438 + }, + { + "epoch": 0.1965398536687956, + "grad_norm": 1.535048484802246, + "learning_rate": 1.9437811992054318e-05, + "loss": 0.9857, + "step": 4439 + }, + { + "epoch": 0.19658412937360947, + "grad_norm": 1.457822322845459, + "learning_rate": 1.9437300998411858e-05, + "loss": 0.9554, + "step": 4440 + }, + { + "epoch": 0.19662840507842333, + "grad_norm": 1.4286415576934814, + "learning_rate": 1.9436789779367148e-05, + "loss": 0.9296, + "step": 4441 + }, + { + "epoch": 0.1966726807832372, + "grad_norm": 1.379061222076416, + "learning_rate": 1.9436278334932398e-05, + "loss": 0.9199, + "step": 4442 + }, + { + "epoch": 0.1967169564880511, + "grad_norm": 1.2983615398406982, + "learning_rate": 1.9435766665119823e-05, + "loss": 0.9477, + "step": 4443 + }, + { + "epoch": 0.19676123219286498, + "grad_norm": 1.4261412620544434, + "learning_rate": 1.943525476994164e-05, + "loss": 0.8999, + "step": 4444 + }, + { + "epoch": 0.19680550789767884, + "grad_norm": 1.199999213218689, + "learning_rate": 1.9434742649410085e-05, + "loss": 0.9398, + "step": 4445 + }, + { + "epoch": 0.19684978360249272, + "grad_norm": 1.3863307237625122, + "learning_rate": 1.943423030353738e-05, + "loss": 0.9343, + "step": 4446 + }, + { + "epoch": 0.1968940593073066, + "grad_norm": 1.4531166553497314, + "learning_rate": 1.943371773233577e-05, + "loss": 0.9201, + "step": 4447 + }, + { + "epoch": 0.19693833501212046, + "grad_norm": 1.359067678451538, + "learning_rate": 1.943320493581749e-05, + "loss": 0.8979, + "step": 4448 + }, + { + "epoch": 0.19698261071693435, + "grad_norm": 1.3267124891281128, + "learning_rate": 1.943269191399479e-05, + "loss": 0.8934, + "step": 4449 + }, + { + "epoch": 0.19702688642174823, + "grad_norm": 1.3566006422042847, + "learning_rate": 1.9432178666879925e-05, + "loss": 0.9153, + "step": 4450 + }, + { + "epoch": 0.19707116212656212, + "grad_norm": 1.2703001499176025, + "learning_rate": 1.943166519448515e-05, + "loss": 0.899, + "step": 4451 + }, + { + "epoch": 0.19711543783137597, + "grad_norm": 1.275250792503357, + "learning_rate": 1.9431151496822737e-05, + "loss": 0.9794, + "step": 4452 + }, + { + "epoch": 0.19715971353618986, + "grad_norm": 1.814571738243103, + "learning_rate": 1.9430637573904945e-05, + "loss": 0.9897, + "step": 4453 + }, + { + "epoch": 0.19720398924100374, + "grad_norm": 1.1803466081619263, + "learning_rate": 1.9430123425744057e-05, + "loss": 0.8524, + "step": 4454 + }, + { + "epoch": 0.1972482649458176, + "grad_norm": 1.3553874492645264, + "learning_rate": 1.9429609052352348e-05, + "loss": 0.9649, + "step": 4455 + }, + { + "epoch": 0.19729254065063148, + "grad_norm": 1.0824453830718994, + "learning_rate": 1.9429094453742103e-05, + "loss": 0.7821, + "step": 4456 + }, + { + "epoch": 0.19733681635544537, + "grad_norm": 1.336483359336853, + "learning_rate": 1.9428579629925614e-05, + "loss": 0.8919, + "step": 4457 + }, + { + "epoch": 0.19738109206025922, + "grad_norm": 2.305467367172241, + "learning_rate": 1.942806458091518e-05, + "loss": 1.0175, + "step": 4458 + }, + { + "epoch": 0.1974253677650731, + "grad_norm": 1.6410346031188965, + "learning_rate": 1.94275493067231e-05, + "loss": 0.9449, + "step": 4459 + }, + { + "epoch": 0.197469643469887, + "grad_norm": 1.472652554512024, + "learning_rate": 1.942703380736168e-05, + "loss": 1.0607, + "step": 4460 + }, + { + "epoch": 0.19751391917470087, + "grad_norm": 1.878787636756897, + "learning_rate": 1.9426518082843234e-05, + "loss": 1.0444, + "step": 4461 + }, + { + "epoch": 0.19755819487951473, + "grad_norm": 1.4285987615585327, + "learning_rate": 1.942600213318008e-05, + "loss": 1.0103, + "step": 4462 + }, + { + "epoch": 0.19760247058432862, + "grad_norm": 1.205673098564148, + "learning_rate": 1.9425485958384537e-05, + "loss": 0.8916, + "step": 4463 + }, + { + "epoch": 0.1976467462891425, + "grad_norm": 1.4074021577835083, + "learning_rate": 1.9424969558468937e-05, + "loss": 1.0135, + "step": 4464 + }, + { + "epoch": 0.19769102199395636, + "grad_norm": 1.3776531219482422, + "learning_rate": 1.9424452933445614e-05, + "loss": 0.8649, + "step": 4465 + }, + { + "epoch": 0.19773529769877024, + "grad_norm": 1.4366950988769531, + "learning_rate": 1.9423936083326907e-05, + "loss": 0.9846, + "step": 4466 + }, + { + "epoch": 0.19777957340358412, + "grad_norm": 1.410567045211792, + "learning_rate": 1.942341900812516e-05, + "loss": 0.9814, + "step": 4467 + }, + { + "epoch": 0.19782384910839798, + "grad_norm": 1.31206214427948, + "learning_rate": 1.942290170785272e-05, + "loss": 0.9052, + "step": 4468 + }, + { + "epoch": 0.19786812481321187, + "grad_norm": 1.7758249044418335, + "learning_rate": 1.9422384182521948e-05, + "loss": 1.0247, + "step": 4469 + }, + { + "epoch": 0.19791240051802575, + "grad_norm": 1.268796443939209, + "learning_rate": 1.9421866432145203e-05, + "loss": 0.9296, + "step": 4470 + }, + { + "epoch": 0.19795667622283963, + "grad_norm": 1.2423776388168335, + "learning_rate": 1.9421348456734844e-05, + "loss": 0.854, + "step": 4471 + }, + { + "epoch": 0.1980009519276535, + "grad_norm": 1.1683558225631714, + "learning_rate": 1.9420830256303253e-05, + "loss": 0.8298, + "step": 4472 + }, + { + "epoch": 0.19804522763246737, + "grad_norm": 1.1695877313613892, + "learning_rate": 1.94203118308628e-05, + "loss": 0.8735, + "step": 4473 + }, + { + "epoch": 0.19808950333728126, + "grad_norm": 1.1574383974075317, + "learning_rate": 1.9419793180425876e-05, + "loss": 0.9057, + "step": 4474 + }, + { + "epoch": 0.19813377904209512, + "grad_norm": 1.4262512922286987, + "learning_rate": 1.9419274305004853e-05, + "loss": 0.9748, + "step": 4475 + }, + { + "epoch": 0.198178054746909, + "grad_norm": 1.357711911201477, + "learning_rate": 1.9418755204612136e-05, + "loss": 0.9745, + "step": 4476 + }, + { + "epoch": 0.19822233045172288, + "grad_norm": 1.7625690698623657, + "learning_rate": 1.941823587926012e-05, + "loss": 1.077, + "step": 4477 + }, + { + "epoch": 0.19826660615653677, + "grad_norm": 1.2178940773010254, + "learning_rate": 1.9417716328961206e-05, + "loss": 0.8932, + "step": 4478 + }, + { + "epoch": 0.19831088186135062, + "grad_norm": 1.3830946683883667, + "learning_rate": 1.941719655372781e-05, + "loss": 0.9639, + "step": 4479 + }, + { + "epoch": 0.1983551575661645, + "grad_norm": 1.4635133743286133, + "learning_rate": 1.9416676553572343e-05, + "loss": 1.0377, + "step": 4480 + }, + { + "epoch": 0.1983994332709784, + "grad_norm": 1.2908046245574951, + "learning_rate": 1.941615632850722e-05, + "loss": 0.8972, + "step": 4481 + }, + { + "epoch": 0.19844370897579225, + "grad_norm": 1.297568440437317, + "learning_rate": 1.941563587854487e-05, + "loss": 0.9693, + "step": 4482 + }, + { + "epoch": 0.19848798468060613, + "grad_norm": 1.3217982053756714, + "learning_rate": 1.9415115203697726e-05, + "loss": 0.8872, + "step": 4483 + }, + { + "epoch": 0.19853226038542002, + "grad_norm": 1.4288508892059326, + "learning_rate": 1.941459430397822e-05, + "loss": 0.9528, + "step": 4484 + }, + { + "epoch": 0.19857653609023387, + "grad_norm": 1.3519446849822998, + "learning_rate": 1.9414073179398794e-05, + "loss": 0.9516, + "step": 4485 + }, + { + "epoch": 0.19862081179504776, + "grad_norm": 1.3142735958099365, + "learning_rate": 1.9413551829971894e-05, + "loss": 0.9519, + "step": 4486 + }, + { + "epoch": 0.19866508749986164, + "grad_norm": 1.24962317943573, + "learning_rate": 1.9413030255709978e-05, + "loss": 0.8687, + "step": 4487 + }, + { + "epoch": 0.19870936320467553, + "grad_norm": 1.4832096099853516, + "learning_rate": 1.9412508456625497e-05, + "loss": 0.9667, + "step": 4488 + }, + { + "epoch": 0.19875363890948938, + "grad_norm": 1.2235090732574463, + "learning_rate": 1.9411986432730913e-05, + "loss": 0.9407, + "step": 4489 + }, + { + "epoch": 0.19879791461430327, + "grad_norm": 1.3523613214492798, + "learning_rate": 1.9411464184038698e-05, + "loss": 0.9532, + "step": 4490 + }, + { + "epoch": 0.19884219031911715, + "grad_norm": 1.675278663635254, + "learning_rate": 1.9410941710561323e-05, + "loss": 0.9227, + "step": 4491 + }, + { + "epoch": 0.198886466023931, + "grad_norm": 1.201184630393982, + "learning_rate": 1.941041901231127e-05, + "loss": 0.9281, + "step": 4492 + }, + { + "epoch": 0.1989307417287449, + "grad_norm": 2.140334367752075, + "learning_rate": 1.940989608930102e-05, + "loss": 1.0028, + "step": 4493 + }, + { + "epoch": 0.19897501743355878, + "grad_norm": 1.3061319589614868, + "learning_rate": 1.940937294154306e-05, + "loss": 0.9363, + "step": 4494 + }, + { + "epoch": 0.19901929313837263, + "grad_norm": 1.3983432054519653, + "learning_rate": 1.9408849569049897e-05, + "loss": 0.9542, + "step": 4495 + }, + { + "epoch": 0.19906356884318652, + "grad_norm": 1.1498547792434692, + "learning_rate": 1.940832597183402e-05, + "loss": 0.8509, + "step": 4496 + }, + { + "epoch": 0.1991078445480004, + "grad_norm": 1.3515793085098267, + "learning_rate": 1.9407802149907935e-05, + "loss": 0.9493, + "step": 4497 + }, + { + "epoch": 0.19915212025281429, + "grad_norm": 1.1905440092086792, + "learning_rate": 1.9407278103284158e-05, + "loss": 0.8651, + "step": 4498 + }, + { + "epoch": 0.19919639595762814, + "grad_norm": 1.2844654321670532, + "learning_rate": 1.9406753831975202e-05, + "loss": 0.9519, + "step": 4499 + }, + { + "epoch": 0.19924067166244203, + "grad_norm": 1.3847068548202515, + "learning_rate": 1.940622933599359e-05, + "loss": 0.9324, + "step": 4500 + }, + { + "epoch": 0.1992849473672559, + "grad_norm": 1.562564730644226, + "learning_rate": 1.940570461535185e-05, + "loss": 1.0519, + "step": 4501 + }, + { + "epoch": 0.19932922307206977, + "grad_norm": 1.3736733198165894, + "learning_rate": 1.940517967006251e-05, + "loss": 0.9676, + "step": 4502 + }, + { + "epoch": 0.19937349877688365, + "grad_norm": 1.1532679796218872, + "learning_rate": 1.9404654500138117e-05, + "loss": 0.8501, + "step": 4503 + }, + { + "epoch": 0.19941777448169754, + "grad_norm": 1.2766218185424805, + "learning_rate": 1.9404129105591204e-05, + "loss": 0.9497, + "step": 4504 + }, + { + "epoch": 0.19946205018651142, + "grad_norm": 1.1900922060012817, + "learning_rate": 1.940360348643433e-05, + "loss": 0.8692, + "step": 4505 + }, + { + "epoch": 0.19950632589132528, + "grad_norm": 1.5465339422225952, + "learning_rate": 1.940307764268004e-05, + "loss": 1.0179, + "step": 4506 + }, + { + "epoch": 0.19955060159613916, + "grad_norm": 1.283644437789917, + "learning_rate": 1.9402551574340894e-05, + "loss": 0.9018, + "step": 4507 + }, + { + "epoch": 0.19959487730095304, + "grad_norm": 1.3128621578216553, + "learning_rate": 1.9402025281429465e-05, + "loss": 1.0409, + "step": 4508 + }, + { + "epoch": 0.1996391530057669, + "grad_norm": 1.619580864906311, + "learning_rate": 1.9401498763958316e-05, + "loss": 1.0029, + "step": 4509 + }, + { + "epoch": 0.19968342871058078, + "grad_norm": 1.5466539859771729, + "learning_rate": 1.9400972021940023e-05, + "loss": 0.917, + "step": 4510 + }, + { + "epoch": 0.19972770441539467, + "grad_norm": 1.344245195388794, + "learning_rate": 1.9400445055387167e-05, + "loss": 0.9909, + "step": 4511 + }, + { + "epoch": 0.19977198012020853, + "grad_norm": 1.5254111289978027, + "learning_rate": 1.9399917864312335e-05, + "loss": 0.9812, + "step": 4512 + }, + { + "epoch": 0.1998162558250224, + "grad_norm": 1.2675971984863281, + "learning_rate": 1.939939044872812e-05, + "loss": 0.9025, + "step": 4513 + }, + { + "epoch": 0.1998605315298363, + "grad_norm": 1.3695954084396362, + "learning_rate": 1.939886280864712e-05, + "loss": 0.8886, + "step": 4514 + }, + { + "epoch": 0.19990480723465018, + "grad_norm": 1.2258776426315308, + "learning_rate": 1.939833494408193e-05, + "loss": 0.9104, + "step": 4515 + }, + { + "epoch": 0.19994908293946403, + "grad_norm": 1.2408788204193115, + "learning_rate": 1.9397806855045163e-05, + "loss": 0.9628, + "step": 4516 + }, + { + "epoch": 0.19999335864427792, + "grad_norm": 1.3623008728027344, + "learning_rate": 1.9397278541549432e-05, + "loss": 0.9111, + "step": 4517 + }, + { + "epoch": 0.2000376343490918, + "grad_norm": 1.585384488105774, + "learning_rate": 1.9396750003607358e-05, + "loss": 0.9575, + "step": 4518 + }, + { + "epoch": 0.20008191005390566, + "grad_norm": 1.4993752241134644, + "learning_rate": 1.9396221241231554e-05, + "loss": 0.971, + "step": 4519 + }, + { + "epoch": 0.20012618575871954, + "grad_norm": 1.195469856262207, + "learning_rate": 1.9395692254434663e-05, + "loss": 0.8755, + "step": 4520 + }, + { + "epoch": 0.20017046146353343, + "grad_norm": 1.4633021354675293, + "learning_rate": 1.9395163043229308e-05, + "loss": 0.9536, + "step": 4521 + }, + { + "epoch": 0.2002147371683473, + "grad_norm": 1.1586549282073975, + "learning_rate": 1.9394633607628137e-05, + "loss": 0.8946, + "step": 4522 + }, + { + "epoch": 0.20025901287316117, + "grad_norm": 1.3476362228393555, + "learning_rate": 1.939410394764379e-05, + "loss": 0.8904, + "step": 4523 + }, + { + "epoch": 0.20030328857797505, + "grad_norm": 1.3934242725372314, + "learning_rate": 1.939357406328892e-05, + "loss": 0.8718, + "step": 4524 + }, + { + "epoch": 0.20034756428278894, + "grad_norm": 1.6195253133773804, + "learning_rate": 1.939304395457618e-05, + "loss": 0.9716, + "step": 4525 + }, + { + "epoch": 0.2003918399876028, + "grad_norm": 1.2447673082351685, + "learning_rate": 1.9392513621518235e-05, + "loss": 0.8755, + "step": 4526 + }, + { + "epoch": 0.20043611569241668, + "grad_norm": 1.366417646408081, + "learning_rate": 1.939198306412775e-05, + "loss": 1.0009, + "step": 4527 + }, + { + "epoch": 0.20048039139723056, + "grad_norm": 1.4691141843795776, + "learning_rate": 1.9391452282417398e-05, + "loss": 0.8878, + "step": 4528 + }, + { + "epoch": 0.20052466710204442, + "grad_norm": 1.3162590265274048, + "learning_rate": 1.9390921276399853e-05, + "loss": 0.9562, + "step": 4529 + }, + { + "epoch": 0.2005689428068583, + "grad_norm": 1.6415237188339233, + "learning_rate": 1.93903900460878e-05, + "loss": 1.0239, + "step": 4530 + }, + { + "epoch": 0.2006132185116722, + "grad_norm": 1.2377947568893433, + "learning_rate": 1.9389858591493926e-05, + "loss": 0.9316, + "step": 4531 + }, + { + "epoch": 0.20065749421648607, + "grad_norm": 1.4093852043151855, + "learning_rate": 1.9389326912630925e-05, + "loss": 0.9656, + "step": 4532 + }, + { + "epoch": 0.20070176992129993, + "grad_norm": 1.2957969903945923, + "learning_rate": 1.9388795009511495e-05, + "loss": 0.8957, + "step": 4533 + }, + { + "epoch": 0.2007460456261138, + "grad_norm": 1.573099136352539, + "learning_rate": 1.9388262882148346e-05, + "loss": 1.0354, + "step": 4534 + }, + { + "epoch": 0.2007903213309277, + "grad_norm": 1.4998201131820679, + "learning_rate": 1.938773053055418e-05, + "loss": 0.9315, + "step": 4535 + }, + { + "epoch": 0.20083459703574155, + "grad_norm": 1.4296995401382446, + "learning_rate": 1.938719795474171e-05, + "loss": 1.0088, + "step": 4536 + }, + { + "epoch": 0.20087887274055544, + "grad_norm": 1.3885672092437744, + "learning_rate": 1.9386665154723662e-05, + "loss": 1.0444, + "step": 4537 + }, + { + "epoch": 0.20092314844536932, + "grad_norm": 1.5351718664169312, + "learning_rate": 1.938613213051276e-05, + "loss": 0.9482, + "step": 4538 + }, + { + "epoch": 0.20096742415018318, + "grad_norm": 1.4075876474380493, + "learning_rate": 1.9385598882121735e-05, + "loss": 0.984, + "step": 4539 + }, + { + "epoch": 0.20101169985499706, + "grad_norm": 1.5773333311080933, + "learning_rate": 1.9385065409563323e-05, + "loss": 1.0194, + "step": 4540 + }, + { + "epoch": 0.20105597555981095, + "grad_norm": 1.2765241861343384, + "learning_rate": 1.938453171285027e-05, + "loss": 0.8496, + "step": 4541 + }, + { + "epoch": 0.20110025126462483, + "grad_norm": 2.9711461067199707, + "learning_rate": 1.938399779199531e-05, + "loss": 1.0048, + "step": 4542 + }, + { + "epoch": 0.2011445269694387, + "grad_norm": 1.6548024415969849, + "learning_rate": 1.9383463647011206e-05, + "loss": 1.0171, + "step": 4543 + }, + { + "epoch": 0.20118880267425257, + "grad_norm": 1.669602870941162, + "learning_rate": 1.9382929277910713e-05, + "loss": 1.0198, + "step": 4544 + }, + { + "epoch": 0.20123307837906645, + "grad_norm": 1.2250244617462158, + "learning_rate": 1.9382394684706597e-05, + "loss": 0.9153, + "step": 4545 + }, + { + "epoch": 0.2012773540838803, + "grad_norm": 1.6401073932647705, + "learning_rate": 1.938185986741162e-05, + "loss": 0.9383, + "step": 4546 + }, + { + "epoch": 0.2013216297886942, + "grad_norm": 1.8876312971115112, + "learning_rate": 1.938132482603856e-05, + "loss": 0.9326, + "step": 4547 + }, + { + "epoch": 0.20136590549350808, + "grad_norm": 1.4725708961486816, + "learning_rate": 1.9380789560600197e-05, + "loss": 0.9535, + "step": 4548 + }, + { + "epoch": 0.20141018119832196, + "grad_norm": 1.2604546546936035, + "learning_rate": 1.9380254071109313e-05, + "loss": 0.919, + "step": 4549 + }, + { + "epoch": 0.20145445690313582, + "grad_norm": 1.5762640237808228, + "learning_rate": 1.9379718357578695e-05, + "loss": 0.9636, + "step": 4550 + }, + { + "epoch": 0.2014987326079497, + "grad_norm": 1.1664060354232788, + "learning_rate": 1.9379182420021145e-05, + "loss": 0.8399, + "step": 4551 + }, + { + "epoch": 0.2015430083127636, + "grad_norm": 1.4196114540100098, + "learning_rate": 1.9378646258449454e-05, + "loss": 0.8616, + "step": 4552 + }, + { + "epoch": 0.20158728401757744, + "grad_norm": 1.4968175888061523, + "learning_rate": 1.9378109872876442e-05, + "loss": 0.9842, + "step": 4553 + }, + { + "epoch": 0.20163155972239133, + "grad_norm": 1.4208908081054688, + "learning_rate": 1.9377573263314905e-05, + "loss": 0.9192, + "step": 4554 + }, + { + "epoch": 0.2016758354272052, + "grad_norm": 1.3437392711639404, + "learning_rate": 1.9377036429777673e-05, + "loss": 0.8823, + "step": 4555 + }, + { + "epoch": 0.20172011113201907, + "grad_norm": 1.4074714183807373, + "learning_rate": 1.9376499372277554e-05, + "loss": 0.9385, + "step": 4556 + }, + { + "epoch": 0.20176438683683295, + "grad_norm": 1.3863260746002197, + "learning_rate": 1.9375962090827385e-05, + "loss": 0.9376, + "step": 4557 + }, + { + "epoch": 0.20180866254164684, + "grad_norm": 1.269904613494873, + "learning_rate": 1.9375424585439994e-05, + "loss": 0.9304, + "step": 4558 + }, + { + "epoch": 0.20185293824646072, + "grad_norm": 2.3873989582061768, + "learning_rate": 1.9374886856128223e-05, + "loss": 0.994, + "step": 4559 + }, + { + "epoch": 0.20189721395127458, + "grad_norm": 1.5865572690963745, + "learning_rate": 1.937434890290491e-05, + "loss": 0.9918, + "step": 4560 + }, + { + "epoch": 0.20194148965608846, + "grad_norm": 1.2940629720687866, + "learning_rate": 1.937381072578291e-05, + "loss": 0.929, + "step": 4561 + }, + { + "epoch": 0.20198576536090235, + "grad_norm": 1.2317960262298584, + "learning_rate": 1.937327232477507e-05, + "loss": 0.9387, + "step": 4562 + }, + { + "epoch": 0.2020300410657162, + "grad_norm": 1.272171974182129, + "learning_rate": 1.9372733699894253e-05, + "loss": 0.8682, + "step": 4563 + }, + { + "epoch": 0.2020743167705301, + "grad_norm": 1.4094886779785156, + "learning_rate": 1.9372194851153323e-05, + "loss": 0.9118, + "step": 4564 + }, + { + "epoch": 0.20211859247534397, + "grad_norm": 1.6331595182418823, + "learning_rate": 1.937165577856515e-05, + "loss": 1.0411, + "step": 4565 + }, + { + "epoch": 0.20216286818015786, + "grad_norm": 1.5096663236618042, + "learning_rate": 1.937111648214261e-05, + "loss": 0.9557, + "step": 4566 + }, + { + "epoch": 0.2022071438849717, + "grad_norm": 1.3624464273452759, + "learning_rate": 1.9370576961898585e-05, + "loss": 0.8875, + "step": 4567 + }, + { + "epoch": 0.2022514195897856, + "grad_norm": 1.0960125923156738, + "learning_rate": 1.9370037217845955e-05, + "loss": 0.847, + "step": 4568 + }, + { + "epoch": 0.20229569529459948, + "grad_norm": 1.4074680805206299, + "learning_rate": 1.936949724999762e-05, + "loss": 0.9618, + "step": 4569 + }, + { + "epoch": 0.20233997099941334, + "grad_norm": 1.4325655698776245, + "learning_rate": 1.9368957058366468e-05, + "loss": 0.9998, + "step": 4570 + }, + { + "epoch": 0.20238424670422722, + "grad_norm": 1.4786994457244873, + "learning_rate": 1.936841664296541e-05, + "loss": 1.0432, + "step": 4571 + }, + { + "epoch": 0.2024285224090411, + "grad_norm": 1.6071809530258179, + "learning_rate": 1.9367876003807344e-05, + "loss": 0.9688, + "step": 4572 + }, + { + "epoch": 0.20247279811385496, + "grad_norm": 1.2904430627822876, + "learning_rate": 1.9367335140905187e-05, + "loss": 0.9123, + "step": 4573 + }, + { + "epoch": 0.20251707381866885, + "grad_norm": 1.1297578811645508, + "learning_rate": 1.936679405427186e-05, + "loss": 0.804, + "step": 4574 + }, + { + "epoch": 0.20256134952348273, + "grad_norm": 1.628977656364441, + "learning_rate": 1.9366252743920282e-05, + "loss": 1.0597, + "step": 4575 + }, + { + "epoch": 0.20260562522829662, + "grad_norm": 1.2402862310409546, + "learning_rate": 1.9365711209863387e-05, + "loss": 0.8687, + "step": 4576 + }, + { + "epoch": 0.20264990093311047, + "grad_norm": 1.2053422927856445, + "learning_rate": 1.9365169452114102e-05, + "loss": 0.7884, + "step": 4577 + }, + { + "epoch": 0.20269417663792436, + "grad_norm": 1.3995453119277954, + "learning_rate": 1.936462747068537e-05, + "loss": 0.922, + "step": 4578 + }, + { + "epoch": 0.20273845234273824, + "grad_norm": 1.856151819229126, + "learning_rate": 1.9364085265590137e-05, + "loss": 0.979, + "step": 4579 + }, + { + "epoch": 0.2027827280475521, + "grad_norm": 1.455696702003479, + "learning_rate": 1.9363542836841352e-05, + "loss": 0.96, + "step": 4580 + }, + { + "epoch": 0.20282700375236598, + "grad_norm": 1.2966785430908203, + "learning_rate": 1.936300018445197e-05, + "loss": 0.9493, + "step": 4581 + }, + { + "epoch": 0.20287127945717986, + "grad_norm": 1.5330196619033813, + "learning_rate": 1.936245730843495e-05, + "loss": 1.0307, + "step": 4582 + }, + { + "epoch": 0.20291555516199372, + "grad_norm": 1.5244543552398682, + "learning_rate": 1.9361914208803264e-05, + "loss": 0.927, + "step": 4583 + }, + { + "epoch": 0.2029598308668076, + "grad_norm": 1.169696569442749, + "learning_rate": 1.9361370885569882e-05, + "loss": 0.8592, + "step": 4584 + }, + { + "epoch": 0.2030041065716215, + "grad_norm": 1.3322737216949463, + "learning_rate": 1.9360827338747774e-05, + "loss": 0.8949, + "step": 4585 + }, + { + "epoch": 0.20304838227643537, + "grad_norm": 1.7051382064819336, + "learning_rate": 1.9360283568349928e-05, + "loss": 1.0029, + "step": 4586 + }, + { + "epoch": 0.20309265798124923, + "grad_norm": 1.295699119567871, + "learning_rate": 1.935973957438933e-05, + "loss": 0.9066, + "step": 4587 + }, + { + "epoch": 0.20313693368606311, + "grad_norm": 1.5755949020385742, + "learning_rate": 1.9359195356878974e-05, + "loss": 1.0031, + "step": 4588 + }, + { + "epoch": 0.203181209390877, + "grad_norm": 1.1032729148864746, + "learning_rate": 1.9358650915831858e-05, + "loss": 0.763, + "step": 4589 + }, + { + "epoch": 0.20322548509569086, + "grad_norm": 1.2048834562301636, + "learning_rate": 1.9358106251260984e-05, + "loss": 0.8653, + "step": 4590 + }, + { + "epoch": 0.20326976080050474, + "grad_norm": 1.2894351482391357, + "learning_rate": 1.935756136317936e-05, + "loss": 0.8555, + "step": 4591 + }, + { + "epoch": 0.20331403650531862, + "grad_norm": 1.2169256210327148, + "learning_rate": 1.935701625160001e-05, + "loss": 0.8965, + "step": 4592 + }, + { + "epoch": 0.2033583122101325, + "grad_norm": 1.4562385082244873, + "learning_rate": 1.9356470916535936e-05, + "loss": 1.0096, + "step": 4593 + }, + { + "epoch": 0.20340258791494636, + "grad_norm": 1.5384844541549683, + "learning_rate": 1.9355925358000177e-05, + "loss": 0.938, + "step": 4594 + }, + { + "epoch": 0.20344686361976025, + "grad_norm": 1.1781219244003296, + "learning_rate": 1.935537957600576e-05, + "loss": 0.8099, + "step": 4595 + }, + { + "epoch": 0.20349113932457413, + "grad_norm": 2.0741984844207764, + "learning_rate": 1.9354833570565717e-05, + "loss": 1.0208, + "step": 4596 + }, + { + "epoch": 0.203535415029388, + "grad_norm": 1.7279256582260132, + "learning_rate": 1.935428734169309e-05, + "loss": 0.9418, + "step": 4597 + }, + { + "epoch": 0.20357969073420187, + "grad_norm": 1.3759485483169556, + "learning_rate": 1.9353740889400926e-05, + "loss": 0.9319, + "step": 4598 + }, + { + "epoch": 0.20362396643901576, + "grad_norm": 1.1647943258285522, + "learning_rate": 1.935319421370228e-05, + "loss": 0.9011, + "step": 4599 + }, + { + "epoch": 0.20366824214382961, + "grad_norm": 1.2899317741394043, + "learning_rate": 1.9352647314610205e-05, + "loss": 0.9422, + "step": 4600 + }, + { + "epoch": 0.2037125178486435, + "grad_norm": 1.4199217557907104, + "learning_rate": 1.9352100192137763e-05, + "loss": 0.8924, + "step": 4601 + }, + { + "epoch": 0.20375679355345738, + "grad_norm": 1.633387804031372, + "learning_rate": 1.9351552846298026e-05, + "loss": 0.9411, + "step": 4602 + }, + { + "epoch": 0.20380106925827127, + "grad_norm": 1.6664303541183472, + "learning_rate": 1.935100527710406e-05, + "loss": 0.9919, + "step": 4603 + }, + { + "epoch": 0.20384534496308512, + "grad_norm": 1.9223026037216187, + "learning_rate": 1.9350457484568944e-05, + "loss": 0.9801, + "step": 4604 + }, + { + "epoch": 0.203889620667899, + "grad_norm": 1.209336757659912, + "learning_rate": 1.9349909468705773e-05, + "loss": 0.921, + "step": 4605 + }, + { + "epoch": 0.2039338963727129, + "grad_norm": 1.333080530166626, + "learning_rate": 1.9349361229527617e-05, + "loss": 0.8788, + "step": 4606 + }, + { + "epoch": 0.20397817207752675, + "grad_norm": 1.253993034362793, + "learning_rate": 1.934881276704759e-05, + "loss": 0.989, + "step": 4607 + }, + { + "epoch": 0.20402244778234063, + "grad_norm": 1.775300145149231, + "learning_rate": 1.9348264081278775e-05, + "loss": 0.9829, + "step": 4608 + }, + { + "epoch": 0.20406672348715452, + "grad_norm": 1.4983229637145996, + "learning_rate": 1.934771517223429e-05, + "loss": 0.964, + "step": 4609 + }, + { + "epoch": 0.2041109991919684, + "grad_norm": 1.2758787870407104, + "learning_rate": 1.9347166039927236e-05, + "loss": 0.886, + "step": 4610 + }, + { + "epoch": 0.20415527489678226, + "grad_norm": 1.230561375617981, + "learning_rate": 1.934661668437073e-05, + "loss": 0.8856, + "step": 4611 + }, + { + "epoch": 0.20419955060159614, + "grad_norm": 1.3414909839630127, + "learning_rate": 1.93460671055779e-05, + "loss": 0.9768, + "step": 4612 + }, + { + "epoch": 0.20424382630641003, + "grad_norm": 1.429933786392212, + "learning_rate": 1.9345517303561863e-05, + "loss": 0.9638, + "step": 4613 + }, + { + "epoch": 0.20428810201122388, + "grad_norm": 1.9023687839508057, + "learning_rate": 1.9344967278335757e-05, + "loss": 0.9186, + "step": 4614 + }, + { + "epoch": 0.20433237771603777, + "grad_norm": 1.2832143306732178, + "learning_rate": 1.9344417029912715e-05, + "loss": 0.9404, + "step": 4615 + }, + { + "epoch": 0.20437665342085165, + "grad_norm": 1.401761531829834, + "learning_rate": 1.934386655830588e-05, + "loss": 0.9135, + "step": 4616 + }, + { + "epoch": 0.2044209291256655, + "grad_norm": 2.82155442237854, + "learning_rate": 1.9343315863528405e-05, + "loss": 0.9643, + "step": 4617 + }, + { + "epoch": 0.2044652048304794, + "grad_norm": 1.5212185382843018, + "learning_rate": 1.9342764945593432e-05, + "loss": 0.9799, + "step": 4618 + }, + { + "epoch": 0.20450948053529328, + "grad_norm": 1.5500755310058594, + "learning_rate": 1.934221380451413e-05, + "loss": 1.0169, + "step": 4619 + }, + { + "epoch": 0.20455375624010716, + "grad_norm": 1.1922205686569214, + "learning_rate": 1.9341662440303658e-05, + "loss": 0.9095, + "step": 4620 + }, + { + "epoch": 0.20459803194492102, + "grad_norm": 1.3302569389343262, + "learning_rate": 1.934111085297518e-05, + "loss": 0.9492, + "step": 4621 + }, + { + "epoch": 0.2046423076497349, + "grad_norm": 1.2409707307815552, + "learning_rate": 1.934055904254188e-05, + "loss": 0.929, + "step": 4622 + }, + { + "epoch": 0.20468658335454878, + "grad_norm": 1.4706732034683228, + "learning_rate": 1.934000700901693e-05, + "loss": 0.9176, + "step": 4623 + }, + { + "epoch": 0.20473085905936264, + "grad_norm": 1.2453781366348267, + "learning_rate": 1.9339454752413517e-05, + "loss": 0.8856, + "step": 4624 + }, + { + "epoch": 0.20477513476417653, + "grad_norm": 1.596117377281189, + "learning_rate": 1.9338902272744834e-05, + "loss": 1.0388, + "step": 4625 + }, + { + "epoch": 0.2048194104689904, + "grad_norm": 1.424262523651123, + "learning_rate": 1.9338349570024074e-05, + "loss": 0.9153, + "step": 4626 + }, + { + "epoch": 0.20486368617380427, + "grad_norm": 1.523720145225525, + "learning_rate": 1.9337796644264436e-05, + "loss": 0.886, + "step": 4627 + }, + { + "epoch": 0.20490796187861815, + "grad_norm": 1.257958173751831, + "learning_rate": 1.933724349547913e-05, + "loss": 0.8803, + "step": 4628 + }, + { + "epoch": 0.20495223758343203, + "grad_norm": 1.160413384437561, + "learning_rate": 1.9336690123681366e-05, + "loss": 0.8364, + "step": 4629 + }, + { + "epoch": 0.20499651328824592, + "grad_norm": 1.3592675924301147, + "learning_rate": 1.9336136528884358e-05, + "loss": 0.9655, + "step": 4630 + }, + { + "epoch": 0.20504078899305977, + "grad_norm": 1.246864914894104, + "learning_rate": 1.9335582711101333e-05, + "loss": 0.8959, + "step": 4631 + }, + { + "epoch": 0.20508506469787366, + "grad_norm": 1.1579667329788208, + "learning_rate": 1.9335028670345517e-05, + "loss": 0.8646, + "step": 4632 + }, + { + "epoch": 0.20512934040268754, + "grad_norm": 1.343668818473816, + "learning_rate": 1.9334474406630138e-05, + "loss": 0.9135, + "step": 4633 + }, + { + "epoch": 0.2051736161075014, + "grad_norm": 1.487390160560608, + "learning_rate": 1.9333919919968444e-05, + "loss": 0.9346, + "step": 4634 + }, + { + "epoch": 0.20521789181231528, + "grad_norm": 1.8815960884094238, + "learning_rate": 1.9333365210373668e-05, + "loss": 0.8625, + "step": 4635 + }, + { + "epoch": 0.20526216751712917, + "grad_norm": 1.34503972530365, + "learning_rate": 1.9332810277859068e-05, + "loss": 0.9199, + "step": 4636 + }, + { + "epoch": 0.20530644322194305, + "grad_norm": 1.1461377143859863, + "learning_rate": 1.9332255122437892e-05, + "loss": 0.8731, + "step": 4637 + }, + { + "epoch": 0.2053507189267569, + "grad_norm": 1.3537406921386719, + "learning_rate": 1.9331699744123398e-05, + "loss": 0.9208, + "step": 4638 + }, + { + "epoch": 0.2053949946315708, + "grad_norm": 1.3006452322006226, + "learning_rate": 1.9331144142928853e-05, + "loss": 0.9138, + "step": 4639 + }, + { + "epoch": 0.20543927033638468, + "grad_norm": 1.647498369216919, + "learning_rate": 1.9330588318867533e-05, + "loss": 0.8788, + "step": 4640 + }, + { + "epoch": 0.20548354604119853, + "grad_norm": 1.2651182413101196, + "learning_rate": 1.9330032271952705e-05, + "loss": 0.9377, + "step": 4641 + }, + { + "epoch": 0.20552782174601242, + "grad_norm": 1.1551076173782349, + "learning_rate": 1.9329476002197657e-05, + "loss": 0.8743, + "step": 4642 + }, + { + "epoch": 0.2055720974508263, + "grad_norm": 1.2210084199905396, + "learning_rate": 1.9328919509615664e-05, + "loss": 0.8805, + "step": 4643 + }, + { + "epoch": 0.20561637315564016, + "grad_norm": 1.2701722383499146, + "learning_rate": 1.9328362794220032e-05, + "loss": 1.0171, + "step": 4644 + }, + { + "epoch": 0.20566064886045404, + "grad_norm": 1.2897264957427979, + "learning_rate": 1.932780585602404e-05, + "loss": 0.9323, + "step": 4645 + }, + { + "epoch": 0.20570492456526793, + "grad_norm": 1.155688762664795, + "learning_rate": 1.932724869504101e-05, + "loss": 0.8629, + "step": 4646 + }, + { + "epoch": 0.2057492002700818, + "grad_norm": 1.506894826889038, + "learning_rate": 1.9326691311284237e-05, + "loss": 0.9183, + "step": 4647 + }, + { + "epoch": 0.20579347597489567, + "grad_norm": 1.2780946493148804, + "learning_rate": 1.9326133704767033e-05, + "loss": 0.9654, + "step": 4648 + }, + { + "epoch": 0.20583775167970955, + "grad_norm": 1.1661452054977417, + "learning_rate": 1.9325575875502722e-05, + "loss": 0.8967, + "step": 4649 + }, + { + "epoch": 0.20588202738452344, + "grad_norm": 1.2634133100509644, + "learning_rate": 1.932501782350462e-05, + "loss": 0.9296, + "step": 4650 + }, + { + "epoch": 0.2059263030893373, + "grad_norm": 1.2512867450714111, + "learning_rate": 1.9324459548786066e-05, + "loss": 0.9482, + "step": 4651 + }, + { + "epoch": 0.20597057879415118, + "grad_norm": 1.3290897607803345, + "learning_rate": 1.9323901051360384e-05, + "loss": 0.9865, + "step": 4652 + }, + { + "epoch": 0.20601485449896506, + "grad_norm": 1.341225028038025, + "learning_rate": 1.932334233124092e-05, + "loss": 0.8951, + "step": 4653 + }, + { + "epoch": 0.20605913020377895, + "grad_norm": 1.6781129837036133, + "learning_rate": 1.9322783388441016e-05, + "loss": 1.013, + "step": 4654 + }, + { + "epoch": 0.2061034059085928, + "grad_norm": 1.435558557510376, + "learning_rate": 1.932222422297402e-05, + "loss": 1.0283, + "step": 4655 + }, + { + "epoch": 0.20614768161340669, + "grad_norm": 1.2517050504684448, + "learning_rate": 1.9321664834853288e-05, + "loss": 0.9296, + "step": 4656 + }, + { + "epoch": 0.20619195731822057, + "grad_norm": 1.2596551179885864, + "learning_rate": 1.932110522409218e-05, + "loss": 0.9336, + "step": 4657 + }, + { + "epoch": 0.20623623302303443, + "grad_norm": 1.096766471862793, + "learning_rate": 1.9320545390704066e-05, + "loss": 0.8619, + "step": 4658 + }, + { + "epoch": 0.2062805087278483, + "grad_norm": 1.6179425716400146, + "learning_rate": 1.9319985334702315e-05, + "loss": 0.9691, + "step": 4659 + }, + { + "epoch": 0.2063247844326622, + "grad_norm": 1.1166212558746338, + "learning_rate": 1.9319425056100302e-05, + "loss": 0.8547, + "step": 4660 + }, + { + "epoch": 0.20636906013747605, + "grad_norm": 2.194211483001709, + "learning_rate": 1.931886455491141e-05, + "loss": 0.9765, + "step": 4661 + }, + { + "epoch": 0.20641333584228994, + "grad_norm": 1.2732527256011963, + "learning_rate": 1.9318303831149022e-05, + "loss": 0.913, + "step": 4662 + }, + { + "epoch": 0.20645761154710382, + "grad_norm": 1.357544183731079, + "learning_rate": 1.931774288482654e-05, + "loss": 0.8593, + "step": 4663 + }, + { + "epoch": 0.2065018872519177, + "grad_norm": 1.3088754415512085, + "learning_rate": 1.9317181715957353e-05, + "loss": 0.9087, + "step": 4664 + }, + { + "epoch": 0.20654616295673156, + "grad_norm": 1.3551628589630127, + "learning_rate": 1.9316620324554865e-05, + "loss": 0.9586, + "step": 4665 + }, + { + "epoch": 0.20659043866154544, + "grad_norm": 1.6151413917541504, + "learning_rate": 1.9316058710632492e-05, + "loss": 0.9508, + "step": 4666 + }, + { + "epoch": 0.20663471436635933, + "grad_norm": 1.6320847272872925, + "learning_rate": 1.9315496874203637e-05, + "loss": 0.9222, + "step": 4667 + }, + { + "epoch": 0.20667899007117319, + "grad_norm": 2.049630880355835, + "learning_rate": 1.9314934815281728e-05, + "loss": 0.8996, + "step": 4668 + }, + { + "epoch": 0.20672326577598707, + "grad_norm": 1.278897762298584, + "learning_rate": 1.9314372533880187e-05, + "loss": 0.9262, + "step": 4669 + }, + { + "epoch": 0.20676754148080095, + "grad_norm": 1.2569986581802368, + "learning_rate": 1.9313810030012436e-05, + "loss": 0.8554, + "step": 4670 + }, + { + "epoch": 0.2068118171856148, + "grad_norm": 1.2343969345092773, + "learning_rate": 1.931324730369192e-05, + "loss": 0.8258, + "step": 4671 + }, + { + "epoch": 0.2068560928904287, + "grad_norm": 1.481619119644165, + "learning_rate": 1.9312684354932072e-05, + "loss": 0.9434, + "step": 4672 + }, + { + "epoch": 0.20690036859524258, + "grad_norm": 1.3191308975219727, + "learning_rate": 1.9312121183746344e-05, + "loss": 0.9582, + "step": 4673 + }, + { + "epoch": 0.20694464430005646, + "grad_norm": 1.1740504503250122, + "learning_rate": 1.9311557790148182e-05, + "loss": 0.8864, + "step": 4674 + }, + { + "epoch": 0.20698892000487032, + "grad_norm": 1.2485543489456177, + "learning_rate": 1.9310994174151047e-05, + "loss": 0.8983, + "step": 4675 + }, + { + "epoch": 0.2070331957096842, + "grad_norm": 1.47762930393219, + "learning_rate": 1.9310430335768393e-05, + "loss": 0.9417, + "step": 4676 + }, + { + "epoch": 0.2070774714144981, + "grad_norm": 2.303744316101074, + "learning_rate": 1.9309866275013695e-05, + "loss": 1.0329, + "step": 4677 + }, + { + "epoch": 0.20712174711931194, + "grad_norm": 1.1059849262237549, + "learning_rate": 1.9309301991900417e-05, + "loss": 0.8792, + "step": 4678 + }, + { + "epoch": 0.20716602282412583, + "grad_norm": 1.1322250366210938, + "learning_rate": 1.9308737486442045e-05, + "loss": 0.8144, + "step": 4679 + }, + { + "epoch": 0.2072102985289397, + "grad_norm": 1.3165019750595093, + "learning_rate": 1.9308172758652055e-05, + "loss": 0.8948, + "step": 4680 + }, + { + "epoch": 0.2072545742337536, + "grad_norm": 1.3770502805709839, + "learning_rate": 1.930760780854394e-05, + "loss": 0.9223, + "step": 4681 + }, + { + "epoch": 0.20729884993856745, + "grad_norm": 1.453122615814209, + "learning_rate": 1.930704263613119e-05, + "loss": 0.9408, + "step": 4682 + }, + { + "epoch": 0.20734312564338134, + "grad_norm": 1.2814346551895142, + "learning_rate": 1.9306477241427303e-05, + "loss": 0.8704, + "step": 4683 + }, + { + "epoch": 0.20738740134819522, + "grad_norm": 1.4479471445083618, + "learning_rate": 1.9305911624445785e-05, + "loss": 0.9781, + "step": 4684 + }, + { + "epoch": 0.20743167705300908, + "grad_norm": 1.3769311904907227, + "learning_rate": 1.9305345785200148e-05, + "loss": 0.9022, + "step": 4685 + }, + { + "epoch": 0.20747595275782296, + "grad_norm": 1.522020697593689, + "learning_rate": 1.93047797237039e-05, + "loss": 1.0016, + "step": 4686 + }, + { + "epoch": 0.20752022846263685, + "grad_norm": 1.6156141757965088, + "learning_rate": 1.9304213439970566e-05, + "loss": 1.0113, + "step": 4687 + }, + { + "epoch": 0.2075645041674507, + "grad_norm": 1.1923307180404663, + "learning_rate": 1.930364693401367e-05, + "loss": 0.8862, + "step": 4688 + }, + { + "epoch": 0.2076087798722646, + "grad_norm": 1.2678899765014648, + "learning_rate": 1.9303080205846743e-05, + "loss": 0.9143, + "step": 4689 + }, + { + "epoch": 0.20765305557707847, + "grad_norm": 1.2292463779449463, + "learning_rate": 1.930251325548332e-05, + "loss": 0.9207, + "step": 4690 + }, + { + "epoch": 0.20769733128189236, + "grad_norm": 1.253769040107727, + "learning_rate": 1.930194608293694e-05, + "loss": 0.9106, + "step": 4691 + }, + { + "epoch": 0.2077416069867062, + "grad_norm": 1.1135493516921997, + "learning_rate": 1.9301378688221155e-05, + "loss": 0.9177, + "step": 4692 + }, + { + "epoch": 0.2077858826915201, + "grad_norm": 1.345248818397522, + "learning_rate": 1.9300811071349512e-05, + "loss": 0.9055, + "step": 4693 + }, + { + "epoch": 0.20783015839633398, + "grad_norm": 1.273390293121338, + "learning_rate": 1.930024323233557e-05, + "loss": 0.8576, + "step": 4694 + }, + { + "epoch": 0.20787443410114784, + "grad_norm": 1.394199013710022, + "learning_rate": 1.929967517119289e-05, + "loss": 0.9443, + "step": 4695 + }, + { + "epoch": 0.20791870980596172, + "grad_norm": 1.5720016956329346, + "learning_rate": 1.9299106887935043e-05, + "loss": 0.9681, + "step": 4696 + }, + { + "epoch": 0.2079629855107756, + "grad_norm": 1.3834136724472046, + "learning_rate": 1.9298538382575596e-05, + "loss": 0.9386, + "step": 4697 + }, + { + "epoch": 0.20800726121558946, + "grad_norm": 1.355560064315796, + "learning_rate": 1.9297969655128135e-05, + "loss": 0.9123, + "step": 4698 + }, + { + "epoch": 0.20805153692040335, + "grad_norm": 1.3141402006149292, + "learning_rate": 1.9297400705606234e-05, + "loss": 0.8798, + "step": 4699 + }, + { + "epoch": 0.20809581262521723, + "grad_norm": 1.868922472000122, + "learning_rate": 1.929683153402349e-05, + "loss": 1.1521, + "step": 4700 + }, + { + "epoch": 0.20814008833003111, + "grad_norm": 1.6089187860488892, + "learning_rate": 1.9296262140393498e-05, + "loss": 0.9725, + "step": 4701 + }, + { + "epoch": 0.20818436403484497, + "grad_norm": 1.1785603761672974, + "learning_rate": 1.929569252472985e-05, + "loss": 0.9004, + "step": 4702 + }, + { + "epoch": 0.20822863973965886, + "grad_norm": 1.3636236190795898, + "learning_rate": 1.9295122687046155e-05, + "loss": 1.0201, + "step": 4703 + }, + { + "epoch": 0.20827291544447274, + "grad_norm": 1.1730540990829468, + "learning_rate": 1.929455262735603e-05, + "loss": 0.8784, + "step": 4704 + }, + { + "epoch": 0.2083171911492866, + "grad_norm": 1.3737965822219849, + "learning_rate": 1.929398234567307e-05, + "loss": 0.8845, + "step": 4705 + }, + { + "epoch": 0.20836146685410048, + "grad_norm": 1.3339742422103882, + "learning_rate": 1.9293411842010917e-05, + "loss": 0.8999, + "step": 4706 + }, + { + "epoch": 0.20840574255891436, + "grad_norm": 1.2979501485824585, + "learning_rate": 1.9292841116383192e-05, + "loss": 0.934, + "step": 4707 + }, + { + "epoch": 0.20845001826372825, + "grad_norm": 1.308302879333496, + "learning_rate": 1.929227016880352e-05, + "loss": 0.8768, + "step": 4708 + }, + { + "epoch": 0.2084942939685421, + "grad_norm": 3.1043882369995117, + "learning_rate": 1.929169899928554e-05, + "loss": 0.8838, + "step": 4709 + }, + { + "epoch": 0.208538569673356, + "grad_norm": 1.3669182062149048, + "learning_rate": 1.9291127607842897e-05, + "loss": 0.9218, + "step": 4710 + }, + { + "epoch": 0.20858284537816987, + "grad_norm": 1.2684067487716675, + "learning_rate": 1.9290555994489235e-05, + "loss": 0.8887, + "step": 4711 + }, + { + "epoch": 0.20862712108298373, + "grad_norm": 1.365043044090271, + "learning_rate": 1.928998415923821e-05, + "loss": 0.8939, + "step": 4712 + }, + { + "epoch": 0.20867139678779761, + "grad_norm": 1.3932644128799438, + "learning_rate": 1.9289412102103474e-05, + "loss": 0.873, + "step": 4713 + }, + { + "epoch": 0.2087156724926115, + "grad_norm": 1.302756667137146, + "learning_rate": 1.9288839823098694e-05, + "loss": 0.8403, + "step": 4714 + }, + { + "epoch": 0.20875994819742535, + "grad_norm": 1.3880687952041626, + "learning_rate": 1.928826732223754e-05, + "loss": 0.9506, + "step": 4715 + }, + { + "epoch": 0.20880422390223924, + "grad_norm": 1.5175867080688477, + "learning_rate": 1.9287694599533683e-05, + "loss": 0.9791, + "step": 4716 + }, + { + "epoch": 0.20884849960705312, + "grad_norm": 1.3165363073349, + "learning_rate": 1.92871216550008e-05, + "loss": 0.9379, + "step": 4717 + }, + { + "epoch": 0.208892775311867, + "grad_norm": 1.5763754844665527, + "learning_rate": 1.9286548488652583e-05, + "loss": 0.9351, + "step": 4718 + }, + { + "epoch": 0.20893705101668086, + "grad_norm": 1.26467764377594, + "learning_rate": 1.9285975100502715e-05, + "loss": 0.8966, + "step": 4719 + }, + { + "epoch": 0.20898132672149475, + "grad_norm": 1.4982712268829346, + "learning_rate": 1.9285401490564894e-05, + "loss": 1.0009, + "step": 4720 + }, + { + "epoch": 0.20902560242630863, + "grad_norm": 1.66151762008667, + "learning_rate": 1.9284827658852817e-05, + "loss": 0.9791, + "step": 4721 + }, + { + "epoch": 0.2090698781311225, + "grad_norm": 1.241926670074463, + "learning_rate": 1.928425360538019e-05, + "loss": 0.8709, + "step": 4722 + }, + { + "epoch": 0.20911415383593637, + "grad_norm": 1.291844367980957, + "learning_rate": 1.9283679330160726e-05, + "loss": 0.96, + "step": 4723 + }, + { + "epoch": 0.20915842954075026, + "grad_norm": 1.3936738967895508, + "learning_rate": 1.9283104833208142e-05, + "loss": 0.9711, + "step": 4724 + }, + { + "epoch": 0.20920270524556414, + "grad_norm": 1.7088720798492432, + "learning_rate": 1.9282530114536157e-05, + "loss": 1.0522, + "step": 4725 + }, + { + "epoch": 0.209246980950378, + "grad_norm": 1.3071073293685913, + "learning_rate": 1.92819551741585e-05, + "loss": 0.912, + "step": 4726 + }, + { + "epoch": 0.20929125665519188, + "grad_norm": 1.2698397636413574, + "learning_rate": 1.9281380012088898e-05, + "loss": 0.8686, + "step": 4727 + }, + { + "epoch": 0.20933553236000577, + "grad_norm": 1.146122932434082, + "learning_rate": 1.9280804628341093e-05, + "loss": 0.8741, + "step": 4728 + }, + { + "epoch": 0.20937980806481962, + "grad_norm": 1.218859076499939, + "learning_rate": 1.9280229022928828e-05, + "loss": 0.9132, + "step": 4729 + }, + { + "epoch": 0.2094240837696335, + "grad_norm": 1.1649010181427002, + "learning_rate": 1.9279653195865845e-05, + "loss": 0.7937, + "step": 4730 + }, + { + "epoch": 0.2094683594744474, + "grad_norm": 1.3855290412902832, + "learning_rate": 1.9279077147165903e-05, + "loss": 0.986, + "step": 4731 + }, + { + "epoch": 0.20951263517926125, + "grad_norm": 1.169966459274292, + "learning_rate": 1.927850087684276e-05, + "loss": 0.8312, + "step": 4732 + }, + { + "epoch": 0.20955691088407513, + "grad_norm": 1.4174453020095825, + "learning_rate": 1.9277924384910175e-05, + "loss": 0.931, + "step": 4733 + }, + { + "epoch": 0.20960118658888902, + "grad_norm": 2.1863536834716797, + "learning_rate": 1.9277347671381924e-05, + "loss": 1.0397, + "step": 4734 + }, + { + "epoch": 0.2096454622937029, + "grad_norm": 1.4620493650436401, + "learning_rate": 1.9276770736271778e-05, + "loss": 0.993, + "step": 4735 + }, + { + "epoch": 0.20968973799851676, + "grad_norm": 1.3676944971084595, + "learning_rate": 1.9276193579593513e-05, + "loss": 0.9816, + "step": 4736 + }, + { + "epoch": 0.20973401370333064, + "grad_norm": 1.5567705631256104, + "learning_rate": 1.927561620136092e-05, + "loss": 1.0157, + "step": 4737 + }, + { + "epoch": 0.20977828940814452, + "grad_norm": 1.2730419635772705, + "learning_rate": 1.9275038601587785e-05, + "loss": 0.9251, + "step": 4738 + }, + { + "epoch": 0.20982256511295838, + "grad_norm": 1.3413207530975342, + "learning_rate": 1.9274460780287907e-05, + "loss": 0.9769, + "step": 4739 + }, + { + "epoch": 0.20986684081777227, + "grad_norm": 1.4976998567581177, + "learning_rate": 1.9273882737475086e-05, + "loss": 1.0014, + "step": 4740 + }, + { + "epoch": 0.20991111652258615, + "grad_norm": 1.4204319715499878, + "learning_rate": 1.9273304473163123e-05, + "loss": 0.8691, + "step": 4741 + }, + { + "epoch": 0.2099553922274, + "grad_norm": 1.264279842376709, + "learning_rate": 1.9272725987365833e-05, + "loss": 0.9152, + "step": 4742 + }, + { + "epoch": 0.2099996679322139, + "grad_norm": 1.5233901739120483, + "learning_rate": 1.9272147280097034e-05, + "loss": 0.912, + "step": 4743 + }, + { + "epoch": 0.21004394363702777, + "grad_norm": 2.281977653503418, + "learning_rate": 1.927156835137055e-05, + "loss": 0.978, + "step": 4744 + }, + { + "epoch": 0.21008821934184166, + "grad_norm": 1.6461371183395386, + "learning_rate": 1.9270989201200204e-05, + "loss": 1.0205, + "step": 4745 + }, + { + "epoch": 0.21013249504665552, + "grad_norm": 1.6444107294082642, + "learning_rate": 1.927040982959983e-05, + "loss": 0.9707, + "step": 4746 + }, + { + "epoch": 0.2101767707514694, + "grad_norm": 1.3691812753677368, + "learning_rate": 1.926983023658326e-05, + "loss": 0.9397, + "step": 4747 + }, + { + "epoch": 0.21022104645628328, + "grad_norm": 1.3481910228729248, + "learning_rate": 1.9269250422164347e-05, + "loss": 0.846, + "step": 4748 + }, + { + "epoch": 0.21026532216109714, + "grad_norm": 1.3965197801589966, + "learning_rate": 1.9268670386356934e-05, + "loss": 0.9493, + "step": 4749 + }, + { + "epoch": 0.21030959786591102, + "grad_norm": 1.2783546447753906, + "learning_rate": 1.9268090129174878e-05, + "loss": 0.9431, + "step": 4750 + }, + { + "epoch": 0.2103538735707249, + "grad_norm": 1.8648511171340942, + "learning_rate": 1.926750965063203e-05, + "loss": 0.972, + "step": 4751 + }, + { + "epoch": 0.2103981492755388, + "grad_norm": 1.122151255607605, + "learning_rate": 1.9266928950742265e-05, + "loss": 0.8346, + "step": 4752 + }, + { + "epoch": 0.21044242498035265, + "grad_norm": 1.3613160848617554, + "learning_rate": 1.9266348029519447e-05, + "loss": 0.9656, + "step": 4753 + }, + { + "epoch": 0.21048670068516653, + "grad_norm": 1.3271095752716064, + "learning_rate": 1.9265766886977447e-05, + "loss": 0.882, + "step": 4754 + }, + { + "epoch": 0.21053097638998042, + "grad_norm": 1.4762381315231323, + "learning_rate": 1.9265185523130156e-05, + "loss": 1.0473, + "step": 4755 + }, + { + "epoch": 0.21057525209479427, + "grad_norm": 1.2273818254470825, + "learning_rate": 1.9264603937991445e-05, + "loss": 0.8374, + "step": 4756 + }, + { + "epoch": 0.21061952779960816, + "grad_norm": 1.526962399482727, + "learning_rate": 1.926402213157522e-05, + "loss": 0.9476, + "step": 4757 + }, + { + "epoch": 0.21066380350442204, + "grad_norm": 1.63853919506073, + "learning_rate": 1.9263440103895366e-05, + "loss": 1.0256, + "step": 4758 + }, + { + "epoch": 0.2107080792092359, + "grad_norm": 1.2457820177078247, + "learning_rate": 1.9262857854965787e-05, + "loss": 0.9038, + "step": 4759 + }, + { + "epoch": 0.21075235491404978, + "grad_norm": 1.277458667755127, + "learning_rate": 1.9262275384800392e-05, + "loss": 0.9672, + "step": 4760 + }, + { + "epoch": 0.21079663061886367, + "grad_norm": 1.1752957105636597, + "learning_rate": 1.926169269341309e-05, + "loss": 0.8648, + "step": 4761 + }, + { + "epoch": 0.21084090632367755, + "grad_norm": 1.2311689853668213, + "learning_rate": 1.92611097808178e-05, + "loss": 0.9076, + "step": 4762 + }, + { + "epoch": 0.2108851820284914, + "grad_norm": 1.1748861074447632, + "learning_rate": 1.9260526647028442e-05, + "loss": 0.9135, + "step": 4763 + }, + { + "epoch": 0.2109294577333053, + "grad_norm": 1.6449062824249268, + "learning_rate": 1.925994329205895e-05, + "loss": 0.9706, + "step": 4764 + }, + { + "epoch": 0.21097373343811918, + "grad_norm": 1.417785882949829, + "learning_rate": 1.9259359715923243e-05, + "loss": 0.9569, + "step": 4765 + }, + { + "epoch": 0.21101800914293303, + "grad_norm": 1.3171972036361694, + "learning_rate": 1.9258775918635276e-05, + "loss": 0.9653, + "step": 4766 + }, + { + "epoch": 0.21106228484774692, + "grad_norm": 1.2463606595993042, + "learning_rate": 1.925819190020898e-05, + "loss": 0.8741, + "step": 4767 + }, + { + "epoch": 0.2111065605525608, + "grad_norm": 1.2323369979858398, + "learning_rate": 1.9257607660658312e-05, + "loss": 0.9278, + "step": 4768 + }, + { + "epoch": 0.21115083625737469, + "grad_norm": 1.3012796640396118, + "learning_rate": 1.9257023199997223e-05, + "loss": 0.9598, + "step": 4769 + }, + { + "epoch": 0.21119511196218854, + "grad_norm": 1.440161943435669, + "learning_rate": 1.925643851823967e-05, + "loss": 0.9335, + "step": 4770 + }, + { + "epoch": 0.21123938766700243, + "grad_norm": 1.3875832557678223, + "learning_rate": 1.9255853615399626e-05, + "loss": 0.8483, + "step": 4771 + }, + { + "epoch": 0.2112836633718163, + "grad_norm": 1.3722819089889526, + "learning_rate": 1.9255268491491048e-05, + "loss": 0.9544, + "step": 4772 + }, + { + "epoch": 0.21132793907663017, + "grad_norm": 1.4860475063323975, + "learning_rate": 1.9254683146527922e-05, + "loss": 0.9508, + "step": 4773 + }, + { + "epoch": 0.21137221478144405, + "grad_norm": 1.4294742345809937, + "learning_rate": 1.925409758052422e-05, + "loss": 0.9154, + "step": 4774 + }, + { + "epoch": 0.21141649048625794, + "grad_norm": 1.1229792833328247, + "learning_rate": 1.9253511793493937e-05, + "loss": 0.9649, + "step": 4775 + }, + { + "epoch": 0.2114607661910718, + "grad_norm": 1.3501836061477661, + "learning_rate": 1.9252925785451055e-05, + "loss": 0.9386, + "step": 4776 + }, + { + "epoch": 0.21150504189588568, + "grad_norm": 1.3695952892303467, + "learning_rate": 1.9252339556409582e-05, + "loss": 0.947, + "step": 4777 + }, + { + "epoch": 0.21154931760069956, + "grad_norm": 1.5532783269882202, + "learning_rate": 1.9251753106383505e-05, + "loss": 0.9475, + "step": 4778 + }, + { + "epoch": 0.21159359330551344, + "grad_norm": 1.318625569343567, + "learning_rate": 1.9251166435386837e-05, + "loss": 0.9058, + "step": 4779 + }, + { + "epoch": 0.2116378690103273, + "grad_norm": 1.2267930507659912, + "learning_rate": 1.92505795434336e-05, + "loss": 0.8439, + "step": 4780 + }, + { + "epoch": 0.21168214471514119, + "grad_norm": 1.2354189157485962, + "learning_rate": 1.9249992430537792e-05, + "loss": 0.8833, + "step": 4781 + }, + { + "epoch": 0.21172642041995507, + "grad_norm": 1.273717999458313, + "learning_rate": 1.9249405096713452e-05, + "loss": 0.893, + "step": 4782 + }, + { + "epoch": 0.21177069612476893, + "grad_norm": 1.270385980606079, + "learning_rate": 1.92488175419746e-05, + "loss": 0.9091, + "step": 4783 + }, + { + "epoch": 0.2118149718295828, + "grad_norm": 1.3329205513000488, + "learning_rate": 1.9248229766335277e-05, + "loss": 0.9707, + "step": 4784 + }, + { + "epoch": 0.2118592475343967, + "grad_norm": 1.4316776990890503, + "learning_rate": 1.9247641769809513e-05, + "loss": 0.9377, + "step": 4785 + }, + { + "epoch": 0.21190352323921055, + "grad_norm": 1.2135212421417236, + "learning_rate": 1.9247053552411356e-05, + "loss": 0.9598, + "step": 4786 + }, + { + "epoch": 0.21194779894402443, + "grad_norm": 1.2035739421844482, + "learning_rate": 1.9246465114154854e-05, + "loss": 0.8151, + "step": 4787 + }, + { + "epoch": 0.21199207464883832, + "grad_norm": 1.3578827381134033, + "learning_rate": 1.924587645505406e-05, + "loss": 0.9103, + "step": 4788 + }, + { + "epoch": 0.2120363503536522, + "grad_norm": 1.1165753602981567, + "learning_rate": 1.9245287575123036e-05, + "loss": 0.8203, + "step": 4789 + }, + { + "epoch": 0.21208062605846606, + "grad_norm": 1.7022515535354614, + "learning_rate": 1.9244698474375845e-05, + "loss": 0.9599, + "step": 4790 + }, + { + "epoch": 0.21212490176327994, + "grad_norm": 1.56662118434906, + "learning_rate": 1.924410915282656e-05, + "loss": 0.9043, + "step": 4791 + }, + { + "epoch": 0.21216917746809383, + "grad_norm": 1.3255727291107178, + "learning_rate": 1.924351961048925e-05, + "loss": 0.9267, + "step": 4792 + }, + { + "epoch": 0.21221345317290768, + "grad_norm": 1.3913850784301758, + "learning_rate": 1.9242929847378006e-05, + "loss": 0.9294, + "step": 4793 + }, + { + "epoch": 0.21225772887772157, + "grad_norm": 1.2408397197723389, + "learning_rate": 1.924233986350691e-05, + "loss": 0.8704, + "step": 4794 + }, + { + "epoch": 0.21230200458253545, + "grad_norm": 1.3248975276947021, + "learning_rate": 1.9241749658890046e-05, + "loss": 0.9284, + "step": 4795 + }, + { + "epoch": 0.21234628028734934, + "grad_norm": 2.2473626136779785, + "learning_rate": 1.924115923354152e-05, + "loss": 1.0517, + "step": 4796 + }, + { + "epoch": 0.2123905559921632, + "grad_norm": 1.1716439723968506, + "learning_rate": 1.9240568587475427e-05, + "loss": 0.7924, + "step": 4797 + }, + { + "epoch": 0.21243483169697708, + "grad_norm": 1.6481008529663086, + "learning_rate": 1.923997772070588e-05, + "loss": 1.0883, + "step": 4798 + }, + { + "epoch": 0.21247910740179096, + "grad_norm": 1.7594902515411377, + "learning_rate": 1.9239386633246987e-05, + "loss": 1.0164, + "step": 4799 + }, + { + "epoch": 0.21252338310660482, + "grad_norm": 1.2242647409439087, + "learning_rate": 1.9238795325112867e-05, + "loss": 0.9109, + "step": 4800 + }, + { + "epoch": 0.2125676588114187, + "grad_norm": 1.25222909450531, + "learning_rate": 1.9238203796317646e-05, + "loss": 0.8443, + "step": 4801 + }, + { + "epoch": 0.2126119345162326, + "grad_norm": 1.5516483783721924, + "learning_rate": 1.9237612046875447e-05, + "loss": 0.9831, + "step": 4802 + }, + { + "epoch": 0.21265621022104644, + "grad_norm": 1.2452027797698975, + "learning_rate": 1.9237020076800407e-05, + "loss": 0.9364, + "step": 4803 + }, + { + "epoch": 0.21270048592586033, + "grad_norm": 1.4826180934906006, + "learning_rate": 1.9236427886106663e-05, + "loss": 0.8544, + "step": 4804 + }, + { + "epoch": 0.2127447616306742, + "grad_norm": 1.2688231468200684, + "learning_rate": 1.923583547480836e-05, + "loss": 0.8735, + "step": 4805 + }, + { + "epoch": 0.2127890373354881, + "grad_norm": 1.2161961793899536, + "learning_rate": 1.9235242842919647e-05, + "loss": 0.9089, + "step": 4806 + }, + { + "epoch": 0.21283331304030195, + "grad_norm": 1.674809217453003, + "learning_rate": 1.9234649990454678e-05, + "loss": 0.9974, + "step": 4807 + }, + { + "epoch": 0.21287758874511584, + "grad_norm": 1.2374614477157593, + "learning_rate": 1.923405691742761e-05, + "loss": 0.8954, + "step": 4808 + }, + { + "epoch": 0.21292186444992972, + "grad_norm": 1.9094018936157227, + "learning_rate": 1.9233463623852614e-05, + "loss": 1.0011, + "step": 4809 + }, + { + "epoch": 0.21296614015474358, + "grad_norm": 1.7328572273254395, + "learning_rate": 1.9232870109743857e-05, + "loss": 0.9726, + "step": 4810 + }, + { + "epoch": 0.21301041585955746, + "grad_norm": 2.0800044536590576, + "learning_rate": 1.9232276375115517e-05, + "loss": 0.9416, + "step": 4811 + }, + { + "epoch": 0.21305469156437135, + "grad_norm": 1.6758209466934204, + "learning_rate": 1.9231682419981773e-05, + "loss": 1.0073, + "step": 4812 + }, + { + "epoch": 0.21309896726918523, + "grad_norm": 1.5226719379425049, + "learning_rate": 1.9231088244356808e-05, + "loss": 1.0058, + "step": 4813 + }, + { + "epoch": 0.2131432429739991, + "grad_norm": 1.18743097782135, + "learning_rate": 1.9230493848254815e-05, + "loss": 0.8856, + "step": 4814 + }, + { + "epoch": 0.21318751867881297, + "grad_norm": 1.3680908679962158, + "learning_rate": 1.922989923169e-05, + "loss": 0.9589, + "step": 4815 + }, + { + "epoch": 0.21323179438362685, + "grad_norm": 1.3580946922302246, + "learning_rate": 1.922930439467655e-05, + "loss": 0.9388, + "step": 4816 + }, + { + "epoch": 0.2132760700884407, + "grad_norm": 1.1994277238845825, + "learning_rate": 1.922870933722868e-05, + "loss": 0.8771, + "step": 4817 + }, + { + "epoch": 0.2133203457932546, + "grad_norm": 1.4304322004318237, + "learning_rate": 1.9228114059360606e-05, + "loss": 0.9748, + "step": 4818 + }, + { + "epoch": 0.21336462149806848, + "grad_norm": 1.4884637594223022, + "learning_rate": 1.9227518561086538e-05, + "loss": 0.8339, + "step": 4819 + }, + { + "epoch": 0.21340889720288234, + "grad_norm": 1.358765721321106, + "learning_rate": 1.9226922842420704e-05, + "loss": 0.9395, + "step": 4820 + }, + { + "epoch": 0.21345317290769622, + "grad_norm": 1.3552329540252686, + "learning_rate": 1.9226326903377332e-05, + "loss": 0.9224, + "step": 4821 + }, + { + "epoch": 0.2134974486125101, + "grad_norm": 1.566011905670166, + "learning_rate": 1.922573074397065e-05, + "loss": 0.9034, + "step": 4822 + }, + { + "epoch": 0.213541724317324, + "grad_norm": 1.119022011756897, + "learning_rate": 1.92251343642149e-05, + "loss": 0.7905, + "step": 4823 + }, + { + "epoch": 0.21358600002213785, + "grad_norm": 1.3332347869873047, + "learning_rate": 1.9224537764124332e-05, + "loss": 0.9154, + "step": 4824 + }, + { + "epoch": 0.21363027572695173, + "grad_norm": 1.6798324584960938, + "learning_rate": 1.922394094371319e-05, + "loss": 0.995, + "step": 4825 + }, + { + "epoch": 0.2136745514317656, + "grad_norm": 1.2534911632537842, + "learning_rate": 1.9223343902995727e-05, + "loss": 0.9247, + "step": 4826 + }, + { + "epoch": 0.21371882713657947, + "grad_norm": 1.710676908493042, + "learning_rate": 1.9222746641986207e-05, + "loss": 0.9513, + "step": 4827 + }, + { + "epoch": 0.21376310284139335, + "grad_norm": 1.433467984199524, + "learning_rate": 1.9222149160698892e-05, + "loss": 0.9275, + "step": 4828 + }, + { + "epoch": 0.21380737854620724, + "grad_norm": 1.2275733947753906, + "learning_rate": 1.9221551459148055e-05, + "loss": 0.8155, + "step": 4829 + }, + { + "epoch": 0.2138516542510211, + "grad_norm": 1.4033547639846802, + "learning_rate": 1.9220953537347963e-05, + "loss": 0.9643, + "step": 4830 + }, + { + "epoch": 0.21389592995583498, + "grad_norm": 1.151135802268982, + "learning_rate": 1.922035539531291e-05, + "loss": 0.834, + "step": 4831 + }, + { + "epoch": 0.21394020566064886, + "grad_norm": 1.1769648790359497, + "learning_rate": 1.9219757033057176e-05, + "loss": 0.8885, + "step": 4832 + }, + { + "epoch": 0.21398448136546275, + "grad_norm": 1.4320513010025024, + "learning_rate": 1.921915845059505e-05, + "loss": 0.9975, + "step": 4833 + }, + { + "epoch": 0.2140287570702766, + "grad_norm": 1.2361321449279785, + "learning_rate": 1.921855964794083e-05, + "loss": 0.8952, + "step": 4834 + }, + { + "epoch": 0.2140730327750905, + "grad_norm": 1.2189244031906128, + "learning_rate": 1.921796062510882e-05, + "loss": 0.9416, + "step": 4835 + }, + { + "epoch": 0.21411730847990437, + "grad_norm": 1.405053734779358, + "learning_rate": 1.921736138211333e-05, + "loss": 0.9569, + "step": 4836 + }, + { + "epoch": 0.21416158418471823, + "grad_norm": 1.412741780281067, + "learning_rate": 1.9216761918968662e-05, + "loss": 0.9677, + "step": 4837 + }, + { + "epoch": 0.2142058598895321, + "grad_norm": 1.8506519794464111, + "learning_rate": 1.9216162235689144e-05, + "loss": 0.9557, + "step": 4838 + }, + { + "epoch": 0.214250135594346, + "grad_norm": 1.5457485914230347, + "learning_rate": 1.9215562332289096e-05, + "loss": 1.0488, + "step": 4839 + }, + { + "epoch": 0.21429441129915988, + "grad_norm": 1.3365241289138794, + "learning_rate": 1.921496220878284e-05, + "loss": 0.8373, + "step": 4840 + }, + { + "epoch": 0.21433868700397374, + "grad_norm": 1.200264573097229, + "learning_rate": 1.9214361865184718e-05, + "loss": 0.881, + "step": 4841 + }, + { + "epoch": 0.21438296270878762, + "grad_norm": 1.2409334182739258, + "learning_rate": 1.9213761301509068e-05, + "loss": 0.8784, + "step": 4842 + }, + { + "epoch": 0.2144272384136015, + "grad_norm": 1.2311737537384033, + "learning_rate": 1.921316051777023e-05, + "loss": 0.8206, + "step": 4843 + }, + { + "epoch": 0.21447151411841536, + "grad_norm": 1.174288272857666, + "learning_rate": 1.921255951398255e-05, + "loss": 0.8859, + "step": 4844 + }, + { + "epoch": 0.21451578982322925, + "grad_norm": 1.5121876001358032, + "learning_rate": 1.921195829016039e-05, + "loss": 0.9147, + "step": 4845 + }, + { + "epoch": 0.21456006552804313, + "grad_norm": 1.2072662115097046, + "learning_rate": 1.921135684631811e-05, + "loss": 0.8941, + "step": 4846 + }, + { + "epoch": 0.214604341232857, + "grad_norm": 1.3935925960540771, + "learning_rate": 1.921075518247007e-05, + "loss": 0.9709, + "step": 4847 + }, + { + "epoch": 0.21464861693767087, + "grad_norm": 1.4955029487609863, + "learning_rate": 1.9210153298630643e-05, + "loss": 1.0642, + "step": 4848 + }, + { + "epoch": 0.21469289264248476, + "grad_norm": 1.3318114280700684, + "learning_rate": 1.9209551194814202e-05, + "loss": 0.9813, + "step": 4849 + }, + { + "epoch": 0.21473716834729864, + "grad_norm": 1.5758917331695557, + "learning_rate": 1.9208948871035126e-05, + "loss": 0.9052, + "step": 4850 + }, + { + "epoch": 0.2147814440521125, + "grad_norm": 1.3584266901016235, + "learning_rate": 1.920834632730781e-05, + "loss": 0.9449, + "step": 4851 + }, + { + "epoch": 0.21482571975692638, + "grad_norm": 1.1965311765670776, + "learning_rate": 1.920774356364664e-05, + "loss": 0.8476, + "step": 4852 + }, + { + "epoch": 0.21486999546174027, + "grad_norm": 1.2574927806854248, + "learning_rate": 1.9207140580066008e-05, + "loss": 0.8833, + "step": 4853 + }, + { + "epoch": 0.21491427116655412, + "grad_norm": 1.1702001094818115, + "learning_rate": 1.9206537376580322e-05, + "loss": 0.8619, + "step": 4854 + }, + { + "epoch": 0.214958546871368, + "grad_norm": 1.3981125354766846, + "learning_rate": 1.920593395320399e-05, + "loss": 0.9894, + "step": 4855 + }, + { + "epoch": 0.2150028225761819, + "grad_norm": 1.29432213306427, + "learning_rate": 1.9205330309951415e-05, + "loss": 0.8946, + "step": 4856 + }, + { + "epoch": 0.21504709828099577, + "grad_norm": 1.3018091917037964, + "learning_rate": 1.9204726446837026e-05, + "loss": 0.9086, + "step": 4857 + }, + { + "epoch": 0.21509137398580963, + "grad_norm": 1.1880536079406738, + "learning_rate": 1.920412236387524e-05, + "loss": 0.8876, + "step": 4858 + }, + { + "epoch": 0.21513564969062351, + "grad_norm": 1.4820805788040161, + "learning_rate": 1.920351806108048e-05, + "loss": 0.9775, + "step": 4859 + }, + { + "epoch": 0.2151799253954374, + "grad_norm": 1.2112895250320435, + "learning_rate": 1.920291353846719e-05, + "loss": 0.8652, + "step": 4860 + }, + { + "epoch": 0.21522420110025126, + "grad_norm": 1.2041127681732178, + "learning_rate": 1.9202308796049803e-05, + "loss": 0.8558, + "step": 4861 + }, + { + "epoch": 0.21526847680506514, + "grad_norm": 1.1863569021224976, + "learning_rate": 1.9201703833842765e-05, + "loss": 0.8619, + "step": 4862 + }, + { + "epoch": 0.21531275250987902, + "grad_norm": 1.502156376838684, + "learning_rate": 1.920109865186052e-05, + "loss": 0.97, + "step": 4863 + }, + { + "epoch": 0.21535702821469288, + "grad_norm": 1.2161718606948853, + "learning_rate": 1.9200493250117528e-05, + "loss": 0.8466, + "step": 4864 + }, + { + "epoch": 0.21540130391950676, + "grad_norm": 1.2503747940063477, + "learning_rate": 1.9199887628628243e-05, + "loss": 0.8277, + "step": 4865 + }, + { + "epoch": 0.21544557962432065, + "grad_norm": 1.208402395248413, + "learning_rate": 1.9199281787407136e-05, + "loss": 0.8965, + "step": 4866 + }, + { + "epoch": 0.21548985532913453, + "grad_norm": 1.5025150775909424, + "learning_rate": 1.919867572646867e-05, + "loss": 0.9671, + "step": 4867 + }, + { + "epoch": 0.2155341310339484, + "grad_norm": 1.2678154706954956, + "learning_rate": 1.919806944582733e-05, + "loss": 0.8726, + "step": 4868 + }, + { + "epoch": 0.21557840673876227, + "grad_norm": 1.1642557382583618, + "learning_rate": 1.9197462945497586e-05, + "loss": 0.795, + "step": 4869 + }, + { + "epoch": 0.21562268244357616, + "grad_norm": 1.4584277868270874, + "learning_rate": 1.919685622549393e-05, + "loss": 0.8785, + "step": 4870 + }, + { + "epoch": 0.21566695814839001, + "grad_norm": 2.111119508743286, + "learning_rate": 1.919624928583085e-05, + "loss": 0.9155, + "step": 4871 + }, + { + "epoch": 0.2157112338532039, + "grad_norm": 1.4351180791854858, + "learning_rate": 1.919564212652284e-05, + "loss": 0.9217, + "step": 4872 + }, + { + "epoch": 0.21575550955801778, + "grad_norm": 1.1475348472595215, + "learning_rate": 1.9195034747584413e-05, + "loss": 0.8432, + "step": 4873 + }, + { + "epoch": 0.21579978526283164, + "grad_norm": 1.4833931922912598, + "learning_rate": 1.9194427149030064e-05, + "loss": 0.9644, + "step": 4874 + }, + { + "epoch": 0.21584406096764552, + "grad_norm": 1.5124475955963135, + "learning_rate": 1.919381933087431e-05, + "loss": 1.0318, + "step": 4875 + }, + { + "epoch": 0.2158883366724594, + "grad_norm": 1.319205641746521, + "learning_rate": 1.9193211293131667e-05, + "loss": 0.9505, + "step": 4876 + }, + { + "epoch": 0.2159326123772733, + "grad_norm": 1.3235207796096802, + "learning_rate": 1.9192603035816657e-05, + "loss": 0.88, + "step": 4877 + }, + { + "epoch": 0.21597688808208715, + "grad_norm": 1.2874159812927246, + "learning_rate": 1.919199455894381e-05, + "loss": 0.9406, + "step": 4878 + }, + { + "epoch": 0.21602116378690103, + "grad_norm": 1.3018454313278198, + "learning_rate": 1.9191385862527653e-05, + "loss": 0.8097, + "step": 4879 + }, + { + "epoch": 0.21606543949171492, + "grad_norm": 1.3212019205093384, + "learning_rate": 1.9190776946582734e-05, + "loss": 0.9553, + "step": 4880 + }, + { + "epoch": 0.21610971519652877, + "grad_norm": 1.2430223226547241, + "learning_rate": 1.919016781112359e-05, + "loss": 0.8665, + "step": 4881 + }, + { + "epoch": 0.21615399090134266, + "grad_norm": 1.7005828619003296, + "learning_rate": 1.918955845616477e-05, + "loss": 0.9282, + "step": 4882 + }, + { + "epoch": 0.21619826660615654, + "grad_norm": 1.4559290409088135, + "learning_rate": 1.9188948881720827e-05, + "loss": 0.9937, + "step": 4883 + }, + { + "epoch": 0.21624254231097043, + "grad_norm": 1.2050944566726685, + "learning_rate": 1.9188339087806323e-05, + "loss": 0.8657, + "step": 4884 + }, + { + "epoch": 0.21628681801578428, + "grad_norm": 1.285307765007019, + "learning_rate": 1.918772907443582e-05, + "loss": 0.9096, + "step": 4885 + }, + { + "epoch": 0.21633109372059817, + "grad_norm": 1.518560767173767, + "learning_rate": 1.9187118841623894e-05, + "loss": 0.9869, + "step": 4886 + }, + { + "epoch": 0.21637536942541205, + "grad_norm": 1.4844499826431274, + "learning_rate": 1.9186508389385107e-05, + "loss": 0.9604, + "step": 4887 + }, + { + "epoch": 0.2164196451302259, + "grad_norm": 1.7425487041473389, + "learning_rate": 1.9185897717734053e-05, + "loss": 1.0282, + "step": 4888 + }, + { + "epoch": 0.2164639208350398, + "grad_norm": 1.4980275630950928, + "learning_rate": 1.918528682668531e-05, + "loss": 0.9618, + "step": 4889 + }, + { + "epoch": 0.21650819653985368, + "grad_norm": 1.192579746246338, + "learning_rate": 1.918467571625347e-05, + "loss": 0.8995, + "step": 4890 + }, + { + "epoch": 0.21655247224466753, + "grad_norm": 1.461763620376587, + "learning_rate": 1.9184064386453127e-05, + "loss": 1.0157, + "step": 4891 + }, + { + "epoch": 0.21659674794948142, + "grad_norm": 1.4744319915771484, + "learning_rate": 1.9183452837298888e-05, + "loss": 0.9468, + "step": 4892 + }, + { + "epoch": 0.2166410236542953, + "grad_norm": 1.7796403169631958, + "learning_rate": 1.918284106880535e-05, + "loss": 0.9662, + "step": 4893 + }, + { + "epoch": 0.21668529935910918, + "grad_norm": 1.334115743637085, + "learning_rate": 1.9182229080987134e-05, + "loss": 0.9531, + "step": 4894 + }, + { + "epoch": 0.21672957506392304, + "grad_norm": 1.4483530521392822, + "learning_rate": 1.9181616873858853e-05, + "loss": 0.8936, + "step": 4895 + }, + { + "epoch": 0.21677385076873693, + "grad_norm": 1.5517983436584473, + "learning_rate": 1.918100444743513e-05, + "loss": 0.9821, + "step": 4896 + }, + { + "epoch": 0.2168181264735508, + "grad_norm": 1.57590651512146, + "learning_rate": 1.918039180173059e-05, + "loss": 0.9145, + "step": 4897 + }, + { + "epoch": 0.21686240217836467, + "grad_norm": 1.5186607837677002, + "learning_rate": 1.9179778936759864e-05, + "loss": 0.9611, + "step": 4898 + }, + { + "epoch": 0.21690667788317855, + "grad_norm": 1.3534038066864014, + "learning_rate": 1.9179165852537596e-05, + "loss": 0.8734, + "step": 4899 + }, + { + "epoch": 0.21695095358799243, + "grad_norm": 1.2080063819885254, + "learning_rate": 1.9178552549078424e-05, + "loss": 0.8253, + "step": 4900 + }, + { + "epoch": 0.21699522929280632, + "grad_norm": 2.086030960083008, + "learning_rate": 1.9177939026396997e-05, + "loss": 1.0041, + "step": 4901 + }, + { + "epoch": 0.21703950499762018, + "grad_norm": 1.169480562210083, + "learning_rate": 1.917732528450797e-05, + "loss": 0.8504, + "step": 4902 + }, + { + "epoch": 0.21708378070243406, + "grad_norm": 1.370888113975525, + "learning_rate": 1.9176711323426e-05, + "loss": 0.9432, + "step": 4903 + }, + { + "epoch": 0.21712805640724794, + "grad_norm": 1.542313575744629, + "learning_rate": 1.9176097143165755e-05, + "loss": 1.0007, + "step": 4904 + }, + { + "epoch": 0.2171723321120618, + "grad_norm": 1.25900137424469, + "learning_rate": 1.91754827437419e-05, + "loss": 0.9439, + "step": 4905 + }, + { + "epoch": 0.21721660781687568, + "grad_norm": 1.7270901203155518, + "learning_rate": 1.9174868125169112e-05, + "loss": 0.9345, + "step": 4906 + }, + { + "epoch": 0.21726088352168957, + "grad_norm": 1.9118540287017822, + "learning_rate": 1.9174253287462066e-05, + "loss": 0.9732, + "step": 4907 + }, + { + "epoch": 0.21730515922650342, + "grad_norm": 1.2224055528640747, + "learning_rate": 1.9173638230635453e-05, + "loss": 0.9171, + "step": 4908 + }, + { + "epoch": 0.2173494349313173, + "grad_norm": 1.2689164876937866, + "learning_rate": 1.9173022954703958e-05, + "loss": 0.8414, + "step": 4909 + }, + { + "epoch": 0.2173937106361312, + "grad_norm": 1.266577124595642, + "learning_rate": 1.917240745968228e-05, + "loss": 0.8436, + "step": 4910 + }, + { + "epoch": 0.21743798634094508, + "grad_norm": 1.6893562078475952, + "learning_rate": 1.917179174558512e-05, + "loss": 1.05, + "step": 4911 + }, + { + "epoch": 0.21748226204575893, + "grad_norm": 1.2322721481323242, + "learning_rate": 1.9171175812427178e-05, + "loss": 0.8382, + "step": 4912 + }, + { + "epoch": 0.21752653775057282, + "grad_norm": 1.1416783332824707, + "learning_rate": 1.9170559660223168e-05, + "loss": 0.8773, + "step": 4913 + }, + { + "epoch": 0.2175708134553867, + "grad_norm": 1.348486065864563, + "learning_rate": 1.9169943288987813e-05, + "loss": 0.9294, + "step": 4914 + }, + { + "epoch": 0.21761508916020056, + "grad_norm": 1.4031951427459717, + "learning_rate": 1.9169326698735826e-05, + "loss": 0.9093, + "step": 4915 + }, + { + "epoch": 0.21765936486501444, + "grad_norm": 1.1532578468322754, + "learning_rate": 1.916870988948194e-05, + "loss": 0.855, + "step": 4916 + }, + { + "epoch": 0.21770364056982833, + "grad_norm": 1.2979011535644531, + "learning_rate": 1.916809286124088e-05, + "loss": 0.8754, + "step": 4917 + }, + { + "epoch": 0.21774791627464218, + "grad_norm": 1.2858850955963135, + "learning_rate": 1.9167475614027388e-05, + "loss": 0.8834, + "step": 4918 + }, + { + "epoch": 0.21779219197945607, + "grad_norm": 1.198066234588623, + "learning_rate": 1.9166858147856204e-05, + "loss": 0.928, + "step": 4919 + }, + { + "epoch": 0.21783646768426995, + "grad_norm": 1.4043171405792236, + "learning_rate": 1.9166240462742077e-05, + "loss": 0.8999, + "step": 4920 + }, + { + "epoch": 0.21788074338908384, + "grad_norm": 1.400800347328186, + "learning_rate": 1.9165622558699763e-05, + "loss": 0.9298, + "step": 4921 + }, + { + "epoch": 0.2179250190938977, + "grad_norm": 1.5102285146713257, + "learning_rate": 1.9165004435744016e-05, + "loss": 0.9578, + "step": 4922 + }, + { + "epoch": 0.21796929479871158, + "grad_norm": 1.4941868782043457, + "learning_rate": 1.9164386093889598e-05, + "loss": 0.9275, + "step": 4923 + }, + { + "epoch": 0.21801357050352546, + "grad_norm": 1.4418941736221313, + "learning_rate": 1.9163767533151283e-05, + "loss": 0.9177, + "step": 4924 + }, + { + "epoch": 0.21805784620833932, + "grad_norm": 1.4524827003479004, + "learning_rate": 1.916314875354384e-05, + "loss": 0.9366, + "step": 4925 + }, + { + "epoch": 0.2181021219131532, + "grad_norm": 1.1797544956207275, + "learning_rate": 1.9162529755082053e-05, + "loss": 0.8476, + "step": 4926 + }, + { + "epoch": 0.21814639761796709, + "grad_norm": 1.586363434791565, + "learning_rate": 1.91619105377807e-05, + "loss": 1.0076, + "step": 4927 + }, + { + "epoch": 0.21819067332278097, + "grad_norm": 1.4873239994049072, + "learning_rate": 1.9161291101654577e-05, + "loss": 0.9572, + "step": 4928 + }, + { + "epoch": 0.21823494902759483, + "grad_norm": 1.5043814182281494, + "learning_rate": 1.9160671446718473e-05, + "loss": 0.9378, + "step": 4929 + }, + { + "epoch": 0.2182792247324087, + "grad_norm": 1.4860200881958008, + "learning_rate": 1.916005157298719e-05, + "loss": 0.9014, + "step": 4930 + }, + { + "epoch": 0.2183235004372226, + "grad_norm": 1.4906350374221802, + "learning_rate": 1.9159431480475536e-05, + "loss": 1.0097, + "step": 4931 + }, + { + "epoch": 0.21836777614203645, + "grad_norm": 1.3602005243301392, + "learning_rate": 1.9158811169198315e-05, + "loss": 0.9525, + "step": 4932 + }, + { + "epoch": 0.21841205184685034, + "grad_norm": 1.3976465463638306, + "learning_rate": 1.915819063917035e-05, + "loss": 0.9155, + "step": 4933 + }, + { + "epoch": 0.21845632755166422, + "grad_norm": 1.367238163948059, + "learning_rate": 1.9157569890406458e-05, + "loss": 0.8909, + "step": 4934 + }, + { + "epoch": 0.21850060325647808, + "grad_norm": 1.2511775493621826, + "learning_rate": 1.9156948922921462e-05, + "loss": 0.9129, + "step": 4935 + }, + { + "epoch": 0.21854487896129196, + "grad_norm": 1.4235483407974243, + "learning_rate": 1.91563277367302e-05, + "loss": 0.9138, + "step": 4936 + }, + { + "epoch": 0.21858915466610584, + "grad_norm": 1.186693787574768, + "learning_rate": 1.915570633184751e-05, + "loss": 0.8685, + "step": 4937 + }, + { + "epoch": 0.21863343037091973, + "grad_norm": 1.3838087320327759, + "learning_rate": 1.915508470828822e-05, + "loss": 0.9126, + "step": 4938 + }, + { + "epoch": 0.21867770607573359, + "grad_norm": 1.6838819980621338, + "learning_rate": 1.915446286606719e-05, + "loss": 0.939, + "step": 4939 + }, + { + "epoch": 0.21872198178054747, + "grad_norm": 1.1543478965759277, + "learning_rate": 1.915384080519927e-05, + "loss": 0.8524, + "step": 4940 + }, + { + "epoch": 0.21876625748536135, + "grad_norm": 1.1294023990631104, + "learning_rate": 1.9153218525699314e-05, + "loss": 0.8022, + "step": 4941 + }, + { + "epoch": 0.2188105331901752, + "grad_norm": 1.2118191719055176, + "learning_rate": 1.9152596027582187e-05, + "loss": 0.8196, + "step": 4942 + }, + { + "epoch": 0.2188548088949891, + "grad_norm": 1.6485927104949951, + "learning_rate": 1.9151973310862754e-05, + "loss": 0.9359, + "step": 4943 + }, + { + "epoch": 0.21889908459980298, + "grad_norm": 1.3664677143096924, + "learning_rate": 1.9151350375555894e-05, + "loss": 0.969, + "step": 4944 + }, + { + "epoch": 0.21894336030461684, + "grad_norm": 1.1700775623321533, + "learning_rate": 1.915072722167648e-05, + "loss": 0.8225, + "step": 4945 + }, + { + "epoch": 0.21898763600943072, + "grad_norm": 1.169533371925354, + "learning_rate": 1.9150103849239397e-05, + "loss": 0.856, + "step": 4946 + }, + { + "epoch": 0.2190319117142446, + "grad_norm": 1.4450026750564575, + "learning_rate": 1.9149480258259535e-05, + "loss": 0.9454, + "step": 4947 + }, + { + "epoch": 0.2190761874190585, + "grad_norm": 1.3468612432479858, + "learning_rate": 1.9148856448751783e-05, + "loss": 0.9609, + "step": 4948 + }, + { + "epoch": 0.21912046312387234, + "grad_norm": 1.1379830837249756, + "learning_rate": 1.9148232420731048e-05, + "loss": 0.8988, + "step": 4949 + }, + { + "epoch": 0.21916473882868623, + "grad_norm": 1.2355908155441284, + "learning_rate": 1.9147608174212225e-05, + "loss": 0.862, + "step": 4950 + }, + { + "epoch": 0.2192090145335001, + "grad_norm": 1.2307604551315308, + "learning_rate": 1.9146983709210236e-05, + "loss": 0.8677, + "step": 4951 + }, + { + "epoch": 0.21925329023831397, + "grad_norm": 1.5835951566696167, + "learning_rate": 1.9146359025739984e-05, + "loss": 0.9851, + "step": 4952 + }, + { + "epoch": 0.21929756594312785, + "grad_norm": 1.2875361442565918, + "learning_rate": 1.9145734123816398e-05, + "loss": 0.917, + "step": 4953 + }, + { + "epoch": 0.21934184164794174, + "grad_norm": 1.2302128076553345, + "learning_rate": 1.9145109003454396e-05, + "loss": 0.8722, + "step": 4954 + }, + { + "epoch": 0.21938611735275562, + "grad_norm": 1.5346194505691528, + "learning_rate": 1.9144483664668915e-05, + "loss": 1.0719, + "step": 4955 + }, + { + "epoch": 0.21943039305756948, + "grad_norm": 1.0555052757263184, + "learning_rate": 1.914385810747488e-05, + "loss": 0.8042, + "step": 4956 + }, + { + "epoch": 0.21947466876238336, + "grad_norm": 1.3880212306976318, + "learning_rate": 1.9143232331887245e-05, + "loss": 0.8827, + "step": 4957 + }, + { + "epoch": 0.21951894446719725, + "grad_norm": 1.372010588645935, + "learning_rate": 1.914260633792095e-05, + "loss": 0.8605, + "step": 4958 + }, + { + "epoch": 0.2195632201720111, + "grad_norm": 1.71210515499115, + "learning_rate": 1.9141980125590947e-05, + "loss": 1.0622, + "step": 4959 + }, + { + "epoch": 0.219607495876825, + "grad_norm": 2.136420965194702, + "learning_rate": 1.9141353694912194e-05, + "loss": 0.916, + "step": 4960 + }, + { + "epoch": 0.21965177158163887, + "grad_norm": 1.378610610961914, + "learning_rate": 1.9140727045899647e-05, + "loss": 0.8725, + "step": 4961 + }, + { + "epoch": 0.21969604728645273, + "grad_norm": 1.3863431215286255, + "learning_rate": 1.914010017856828e-05, + "loss": 0.8427, + "step": 4962 + }, + { + "epoch": 0.2197403229912666, + "grad_norm": 1.14789879322052, + "learning_rate": 1.913947309293306e-05, + "loss": 0.8335, + "step": 4963 + }, + { + "epoch": 0.2197845986960805, + "grad_norm": 1.1938093900680542, + "learning_rate": 1.913884578900897e-05, + "loss": 0.8966, + "step": 4964 + }, + { + "epoch": 0.21982887440089438, + "grad_norm": 1.4934966564178467, + "learning_rate": 1.913821826681099e-05, + "loss": 0.9353, + "step": 4965 + }, + { + "epoch": 0.21987315010570824, + "grad_norm": 1.5734477043151855, + "learning_rate": 1.91375905263541e-05, + "loss": 0.8908, + "step": 4966 + }, + { + "epoch": 0.21991742581052212, + "grad_norm": 1.308604121208191, + "learning_rate": 1.9136962567653308e-05, + "loss": 0.9459, + "step": 4967 + }, + { + "epoch": 0.219961701515336, + "grad_norm": 1.558056354522705, + "learning_rate": 1.9136334390723605e-05, + "loss": 0.9685, + "step": 4968 + }, + { + "epoch": 0.22000597722014986, + "grad_norm": 1.3907989263534546, + "learning_rate": 1.913570599557999e-05, + "loss": 0.9574, + "step": 4969 + }, + { + "epoch": 0.22005025292496375, + "grad_norm": 1.130076289176941, + "learning_rate": 1.9135077382237478e-05, + "loss": 0.8011, + "step": 4970 + }, + { + "epoch": 0.22009452862977763, + "grad_norm": 1.3160218000411987, + "learning_rate": 1.9134448550711077e-05, + "loss": 0.8618, + "step": 4971 + }, + { + "epoch": 0.22013880433459151, + "grad_norm": 1.2824852466583252, + "learning_rate": 1.9133819501015812e-05, + "loss": 0.8099, + "step": 4972 + }, + { + "epoch": 0.22018308003940537, + "grad_norm": 1.270920991897583, + "learning_rate": 1.9133190233166708e-05, + "loss": 0.8575, + "step": 4973 + }, + { + "epoch": 0.22022735574421926, + "grad_norm": 1.298993468284607, + "learning_rate": 1.913256074717879e-05, + "loss": 0.872, + "step": 4974 + }, + { + "epoch": 0.22027163144903314, + "grad_norm": 1.1265701055526733, + "learning_rate": 1.9131931043067092e-05, + "loss": 0.8266, + "step": 4975 + }, + { + "epoch": 0.220315907153847, + "grad_norm": 1.4772926568984985, + "learning_rate": 1.9131301120846662e-05, + "loss": 0.9181, + "step": 4976 + }, + { + "epoch": 0.22036018285866088, + "grad_norm": 2.1059350967407227, + "learning_rate": 1.9130670980532534e-05, + "loss": 1.0663, + "step": 4977 + }, + { + "epoch": 0.22040445856347476, + "grad_norm": 1.3335330486297607, + "learning_rate": 1.9130040622139768e-05, + "loss": 0.9481, + "step": 4978 + }, + { + "epoch": 0.22044873426828862, + "grad_norm": 1.3644626140594482, + "learning_rate": 1.9129410045683415e-05, + "loss": 0.8737, + "step": 4979 + }, + { + "epoch": 0.2204930099731025, + "grad_norm": 1.4287179708480835, + "learning_rate": 1.9128779251178535e-05, + "loss": 0.8982, + "step": 4980 + }, + { + "epoch": 0.2205372856779164, + "grad_norm": 1.4031825065612793, + "learning_rate": 1.9128148238640198e-05, + "loss": 0.962, + "step": 4981 + }, + { + "epoch": 0.22058156138273027, + "grad_norm": 1.3845314979553223, + "learning_rate": 1.9127517008083472e-05, + "loss": 0.8815, + "step": 4982 + }, + { + "epoch": 0.22062583708754413, + "grad_norm": 1.2579550743103027, + "learning_rate": 1.9126885559523433e-05, + "loss": 0.8338, + "step": 4983 + }, + { + "epoch": 0.22067011279235801, + "grad_norm": 1.4209535121917725, + "learning_rate": 1.9126253892975166e-05, + "loss": 0.9188, + "step": 4984 + }, + { + "epoch": 0.2207143884971719, + "grad_norm": 1.492522954940796, + "learning_rate": 1.912562200845375e-05, + "loss": 0.973, + "step": 4985 + }, + { + "epoch": 0.22075866420198575, + "grad_norm": 1.4222341775894165, + "learning_rate": 1.9124989905974288e-05, + "loss": 0.9091, + "step": 4986 + }, + { + "epoch": 0.22080293990679964, + "grad_norm": 1.2416774034500122, + "learning_rate": 1.9124357585551872e-05, + "loss": 0.9323, + "step": 4987 + }, + { + "epoch": 0.22084721561161352, + "grad_norm": 1.6292576789855957, + "learning_rate": 1.91237250472016e-05, + "loss": 0.9294, + "step": 4988 + }, + { + "epoch": 0.22089149131642738, + "grad_norm": 1.6704680919647217, + "learning_rate": 1.9123092290938587e-05, + "loss": 0.9115, + "step": 4989 + }, + { + "epoch": 0.22093576702124126, + "grad_norm": 1.3857992887496948, + "learning_rate": 1.9122459316777945e-05, + "loss": 0.8938, + "step": 4990 + }, + { + "epoch": 0.22098004272605515, + "grad_norm": 1.935835599899292, + "learning_rate": 1.9121826124734788e-05, + "loss": 0.9454, + "step": 4991 + }, + { + "epoch": 0.22102431843086903, + "grad_norm": 1.2310274839401245, + "learning_rate": 1.912119271482424e-05, + "loss": 0.8469, + "step": 4992 + }, + { + "epoch": 0.2210685941356829, + "grad_norm": 1.332218885421753, + "learning_rate": 1.9120559087061433e-05, + "loss": 0.8821, + "step": 4993 + }, + { + "epoch": 0.22111286984049677, + "grad_norm": 1.238726019859314, + "learning_rate": 1.9119925241461495e-05, + "loss": 0.8382, + "step": 4994 + }, + { + "epoch": 0.22115714554531066, + "grad_norm": 1.314442753791809, + "learning_rate": 1.9119291178039573e-05, + "loss": 0.9083, + "step": 4995 + }, + { + "epoch": 0.2212014212501245, + "grad_norm": 1.3145815134048462, + "learning_rate": 1.9118656896810804e-05, + "loss": 0.9306, + "step": 4996 + }, + { + "epoch": 0.2212456969549384, + "grad_norm": 1.2545710802078247, + "learning_rate": 1.911802239779034e-05, + "loss": 0.9177, + "step": 4997 + }, + { + "epoch": 0.22128997265975228, + "grad_norm": 1.3196890354156494, + "learning_rate": 1.9117387680993333e-05, + "loss": 0.9331, + "step": 4998 + }, + { + "epoch": 0.22133424836456617, + "grad_norm": 1.5304967164993286, + "learning_rate": 1.911675274643495e-05, + "loss": 0.9476, + "step": 4999 + }, + { + "epoch": 0.22137852406938002, + "grad_norm": 1.3487215042114258, + "learning_rate": 1.9116117594130347e-05, + "loss": 0.97, + "step": 5000 + }, + { + "epoch": 0.2214227997741939, + "grad_norm": 1.389307975769043, + "learning_rate": 1.91154822240947e-05, + "loss": 0.9374, + "step": 5001 + }, + { + "epoch": 0.2214670754790078, + "grad_norm": 1.259012222290039, + "learning_rate": 1.911484663634318e-05, + "loss": 0.8945, + "step": 5002 + }, + { + "epoch": 0.22151135118382165, + "grad_norm": 1.0939329862594604, + "learning_rate": 1.911421083089097e-05, + "loss": 0.8452, + "step": 5003 + }, + { + "epoch": 0.22155562688863553, + "grad_norm": 1.6178028583526611, + "learning_rate": 1.9113574807753253e-05, + "loss": 0.9502, + "step": 5004 + }, + { + "epoch": 0.22159990259344942, + "grad_norm": 1.1114355325698853, + "learning_rate": 1.911293856694523e-05, + "loss": 0.8058, + "step": 5005 + }, + { + "epoch": 0.22164417829826327, + "grad_norm": 1.3384088277816772, + "learning_rate": 1.911230210848208e-05, + "loss": 0.868, + "step": 5006 + }, + { + "epoch": 0.22168845400307716, + "grad_norm": 1.536072850227356, + "learning_rate": 1.911166543237902e-05, + "loss": 0.9732, + "step": 5007 + }, + { + "epoch": 0.22173272970789104, + "grad_norm": 1.3119730949401855, + "learning_rate": 1.9111028538651247e-05, + "loss": 0.8684, + "step": 5008 + }, + { + "epoch": 0.22177700541270493, + "grad_norm": 1.575348138809204, + "learning_rate": 1.9110391427313978e-05, + "loss": 0.9386, + "step": 5009 + }, + { + "epoch": 0.22182128111751878, + "grad_norm": 1.4422820806503296, + "learning_rate": 1.9109754098382428e-05, + "loss": 0.9226, + "step": 5010 + }, + { + "epoch": 0.22186555682233267, + "grad_norm": 1.316945195198059, + "learning_rate": 1.9109116551871815e-05, + "loss": 0.9455, + "step": 5011 + }, + { + "epoch": 0.22190983252714655, + "grad_norm": 1.380028486251831, + "learning_rate": 1.9108478787797373e-05, + "loss": 0.8698, + "step": 5012 + }, + { + "epoch": 0.2219541082319604, + "grad_norm": 1.4922051429748535, + "learning_rate": 1.910784080617433e-05, + "loss": 0.8802, + "step": 5013 + }, + { + "epoch": 0.2219983839367743, + "grad_norm": 1.5324145555496216, + "learning_rate": 1.9107202607017926e-05, + "loss": 0.9339, + "step": 5014 + }, + { + "epoch": 0.22204265964158817, + "grad_norm": 1.2921909093856812, + "learning_rate": 1.9106564190343407e-05, + "loss": 0.8625, + "step": 5015 + }, + { + "epoch": 0.22208693534640206, + "grad_norm": 1.209600567817688, + "learning_rate": 1.9105925556166012e-05, + "loss": 0.866, + "step": 5016 + }, + { + "epoch": 0.22213121105121592, + "grad_norm": 1.4367918968200684, + "learning_rate": 1.9105286704501002e-05, + "loss": 0.9215, + "step": 5017 + }, + { + "epoch": 0.2221754867560298, + "grad_norm": 1.5478525161743164, + "learning_rate": 1.910464763536363e-05, + "loss": 0.8585, + "step": 5018 + }, + { + "epoch": 0.22221976246084368, + "grad_norm": 1.4685921669006348, + "learning_rate": 1.9104008348769164e-05, + "loss": 0.9294, + "step": 5019 + }, + { + "epoch": 0.22226403816565754, + "grad_norm": 1.4636294841766357, + "learning_rate": 1.9103368844732873e-05, + "loss": 0.958, + "step": 5020 + }, + { + "epoch": 0.22230831387047142, + "grad_norm": 1.1701767444610596, + "learning_rate": 1.9102729123270028e-05, + "loss": 0.8137, + "step": 5021 + }, + { + "epoch": 0.2223525895752853, + "grad_norm": 1.3869459629058838, + "learning_rate": 1.9102089184395913e-05, + "loss": 0.8636, + "step": 5022 + }, + { + "epoch": 0.22239686528009917, + "grad_norm": 1.9001775979995728, + "learning_rate": 1.9101449028125804e-05, + "loss": 0.9983, + "step": 5023 + }, + { + "epoch": 0.22244114098491305, + "grad_norm": 1.2562111616134644, + "learning_rate": 1.9100808654474996e-05, + "loss": 0.8779, + "step": 5024 + }, + { + "epoch": 0.22248541668972693, + "grad_norm": 1.5678197145462036, + "learning_rate": 1.9100168063458783e-05, + "loss": 0.9157, + "step": 5025 + }, + { + "epoch": 0.22252969239454082, + "grad_norm": 1.4283322095870972, + "learning_rate": 1.909952725509247e-05, + "loss": 0.901, + "step": 5026 + }, + { + "epoch": 0.22257396809935467, + "grad_norm": 1.3055113554000854, + "learning_rate": 1.9098886229391355e-05, + "loss": 0.9118, + "step": 5027 + }, + { + "epoch": 0.22261824380416856, + "grad_norm": 1.4441523551940918, + "learning_rate": 1.9098244986370748e-05, + "loss": 0.8749, + "step": 5028 + }, + { + "epoch": 0.22266251950898244, + "grad_norm": 1.1290143728256226, + "learning_rate": 1.9097603526045973e-05, + "loss": 0.8296, + "step": 5029 + }, + { + "epoch": 0.2227067952137963, + "grad_norm": 1.4649734497070312, + "learning_rate": 1.9096961848432344e-05, + "loss": 0.9075, + "step": 5030 + }, + { + "epoch": 0.22275107091861018, + "grad_norm": 1.3834267854690552, + "learning_rate": 1.9096319953545186e-05, + "loss": 0.9516, + "step": 5031 + }, + { + "epoch": 0.22279534662342407, + "grad_norm": 1.1952214241027832, + "learning_rate": 1.9095677841399834e-05, + "loss": 0.8737, + "step": 5032 + }, + { + "epoch": 0.22283962232823792, + "grad_norm": 1.5818959474563599, + "learning_rate": 1.909503551201162e-05, + "loss": 0.8671, + "step": 5033 + }, + { + "epoch": 0.2228838980330518, + "grad_norm": 1.3219703435897827, + "learning_rate": 1.909439296539589e-05, + "loss": 0.8848, + "step": 5034 + }, + { + "epoch": 0.2229281737378657, + "grad_norm": 1.4755418300628662, + "learning_rate": 1.909375020156799e-05, + "loss": 0.8667, + "step": 5035 + }, + { + "epoch": 0.22297244944267958, + "grad_norm": 1.2417954206466675, + "learning_rate": 1.909310722054327e-05, + "loss": 0.8654, + "step": 5036 + }, + { + "epoch": 0.22301672514749343, + "grad_norm": 1.2266273498535156, + "learning_rate": 1.909246402233709e-05, + "loss": 0.8979, + "step": 5037 + }, + { + "epoch": 0.22306100085230732, + "grad_norm": 1.5139442682266235, + "learning_rate": 1.909182060696481e-05, + "loss": 1.0285, + "step": 5038 + }, + { + "epoch": 0.2231052765571212, + "grad_norm": 1.844444990158081, + "learning_rate": 1.9091176974441795e-05, + "loss": 1.0304, + "step": 5039 + }, + { + "epoch": 0.22314955226193506, + "grad_norm": 1.3503825664520264, + "learning_rate": 1.909053312478342e-05, + "loss": 0.8851, + "step": 5040 + }, + { + "epoch": 0.22319382796674894, + "grad_norm": 1.4319671392440796, + "learning_rate": 1.9089889058005065e-05, + "loss": 0.9899, + "step": 5041 + }, + { + "epoch": 0.22323810367156283, + "grad_norm": 1.3837947845458984, + "learning_rate": 1.908924477412211e-05, + "loss": 0.964, + "step": 5042 + }, + { + "epoch": 0.2232823793763767, + "grad_norm": 1.4769147634506226, + "learning_rate": 1.9088600273149947e-05, + "loss": 0.9476, + "step": 5043 + }, + { + "epoch": 0.22332665508119057, + "grad_norm": 1.1939996480941772, + "learning_rate": 1.908795555510396e-05, + "loss": 0.8173, + "step": 5044 + }, + { + "epoch": 0.22337093078600445, + "grad_norm": 1.1212612390518188, + "learning_rate": 1.908731061999956e-05, + "loss": 0.8981, + "step": 5045 + }, + { + "epoch": 0.22341520649081834, + "grad_norm": 1.2176543474197388, + "learning_rate": 1.9086665467852142e-05, + "loss": 0.8536, + "step": 5046 + }, + { + "epoch": 0.2234594821956322, + "grad_norm": 1.3139201402664185, + "learning_rate": 1.9086020098677124e-05, + "loss": 0.8847, + "step": 5047 + }, + { + "epoch": 0.22350375790044608, + "grad_norm": 1.5061500072479248, + "learning_rate": 1.9085374512489906e-05, + "loss": 0.8876, + "step": 5048 + }, + { + "epoch": 0.22354803360525996, + "grad_norm": 1.118698239326477, + "learning_rate": 1.9084728709305922e-05, + "loss": 0.8635, + "step": 5049 + }, + { + "epoch": 0.22359230931007382, + "grad_norm": 1.461500644683838, + "learning_rate": 1.9084082689140585e-05, + "loss": 0.9619, + "step": 5050 + }, + { + "epoch": 0.2236365850148877, + "grad_norm": 1.2573264837265015, + "learning_rate": 1.9083436452009333e-05, + "loss": 0.8998, + "step": 5051 + }, + { + "epoch": 0.22368086071970159, + "grad_norm": 1.4597175121307373, + "learning_rate": 1.9082789997927594e-05, + "loss": 0.8996, + "step": 5052 + }, + { + "epoch": 0.22372513642451547, + "grad_norm": 1.3170478343963623, + "learning_rate": 1.9082143326910817e-05, + "loss": 0.8995, + "step": 5053 + }, + { + "epoch": 0.22376941212932933, + "grad_norm": 1.179357886314392, + "learning_rate": 1.9081496438974435e-05, + "loss": 0.9047, + "step": 5054 + }, + { + "epoch": 0.2238136878341432, + "grad_norm": 1.1022522449493408, + "learning_rate": 1.9080849334133907e-05, + "loss": 0.863, + "step": 5055 + }, + { + "epoch": 0.2238579635389571, + "grad_norm": 1.2501612901687622, + "learning_rate": 1.908020201240469e-05, + "loss": 0.8701, + "step": 5056 + }, + { + "epoch": 0.22390223924377095, + "grad_norm": 1.5318909883499146, + "learning_rate": 1.9079554473802238e-05, + "loss": 0.9384, + "step": 5057 + }, + { + "epoch": 0.22394651494858483, + "grad_norm": 1.7026548385620117, + "learning_rate": 1.907890671834202e-05, + "loss": 0.9958, + "step": 5058 + }, + { + "epoch": 0.22399079065339872, + "grad_norm": 1.107724666595459, + "learning_rate": 1.907825874603951e-05, + "loss": 0.8425, + "step": 5059 + }, + { + "epoch": 0.2240350663582126, + "grad_norm": 1.570472240447998, + "learning_rate": 1.907761055691018e-05, + "loss": 0.9912, + "step": 5060 + }, + { + "epoch": 0.22407934206302646, + "grad_norm": 1.3418275117874146, + "learning_rate": 1.907696215096951e-05, + "loss": 0.8376, + "step": 5061 + }, + { + "epoch": 0.22412361776784034, + "grad_norm": 1.2727888822555542, + "learning_rate": 1.9076313528232993e-05, + "loss": 0.9763, + "step": 5062 + }, + { + "epoch": 0.22416789347265423, + "grad_norm": 1.1851110458374023, + "learning_rate": 1.907566468871612e-05, + "loss": 0.8751, + "step": 5063 + }, + { + "epoch": 0.22421216917746808, + "grad_norm": 1.4147400856018066, + "learning_rate": 1.9075015632434382e-05, + "loss": 0.9144, + "step": 5064 + }, + { + "epoch": 0.22425644488228197, + "grad_norm": 1.2872461080551147, + "learning_rate": 1.9074366359403285e-05, + "loss": 0.8986, + "step": 5065 + }, + { + "epoch": 0.22430072058709585, + "grad_norm": 1.2890602350234985, + "learning_rate": 1.9073716869638336e-05, + "loss": 0.8834, + "step": 5066 + }, + { + "epoch": 0.2243449962919097, + "grad_norm": 1.7596583366394043, + "learning_rate": 1.907306716315505e-05, + "loss": 0.9783, + "step": 5067 + }, + { + "epoch": 0.2243892719967236, + "grad_norm": 1.3165465593338013, + "learning_rate": 1.907241723996894e-05, + "loss": 0.9187, + "step": 5068 + }, + { + "epoch": 0.22443354770153748, + "grad_norm": 1.1902996301651, + "learning_rate": 1.9071767100095532e-05, + "loss": 0.86, + "step": 5069 + }, + { + "epoch": 0.22447782340635136, + "grad_norm": 1.3680509328842163, + "learning_rate": 1.9071116743550354e-05, + "loss": 0.9458, + "step": 5070 + }, + { + "epoch": 0.22452209911116522, + "grad_norm": 1.383436679840088, + "learning_rate": 1.9070466170348937e-05, + "loss": 0.9334, + "step": 5071 + }, + { + "epoch": 0.2245663748159791, + "grad_norm": 1.2760928869247437, + "learning_rate": 1.9069815380506823e-05, + "loss": 0.8746, + "step": 5072 + }, + { + "epoch": 0.224610650520793, + "grad_norm": 1.184049129486084, + "learning_rate": 1.9069164374039553e-05, + "loss": 0.8396, + "step": 5073 + }, + { + "epoch": 0.22465492622560684, + "grad_norm": 2.80513072013855, + "learning_rate": 1.9068513150962675e-05, + "loss": 0.9981, + "step": 5074 + }, + { + "epoch": 0.22469920193042073, + "grad_norm": 1.2141797542572021, + "learning_rate": 1.9067861711291744e-05, + "loss": 0.9119, + "step": 5075 + }, + { + "epoch": 0.2247434776352346, + "grad_norm": 1.2155766487121582, + "learning_rate": 1.9067210055042322e-05, + "loss": 0.8369, + "step": 5076 + }, + { + "epoch": 0.22478775334004847, + "grad_norm": 1.5324474573135376, + "learning_rate": 1.9066558182229967e-05, + "loss": 1.0191, + "step": 5077 + }, + { + "epoch": 0.22483202904486235, + "grad_norm": 1.3146710395812988, + "learning_rate": 1.9065906092870254e-05, + "loss": 0.8823, + "step": 5078 + }, + { + "epoch": 0.22487630474967624, + "grad_norm": 1.1620105504989624, + "learning_rate": 1.906525378697876e-05, + "loss": 0.8249, + "step": 5079 + }, + { + "epoch": 0.22492058045449012, + "grad_norm": 1.5614207983016968, + "learning_rate": 1.9064601264571056e-05, + "loss": 0.9996, + "step": 5080 + }, + { + "epoch": 0.22496485615930398, + "grad_norm": 1.0854743719100952, + "learning_rate": 1.906394852566273e-05, + "loss": 0.8297, + "step": 5081 + }, + { + "epoch": 0.22500913186411786, + "grad_norm": 1.4179250001907349, + "learning_rate": 1.9063295570269374e-05, + "loss": 0.9737, + "step": 5082 + }, + { + "epoch": 0.22505340756893175, + "grad_norm": 1.5309364795684814, + "learning_rate": 1.9062642398406586e-05, + "loss": 0.9113, + "step": 5083 + }, + { + "epoch": 0.2250976832737456, + "grad_norm": 1.3228442668914795, + "learning_rate": 1.9061989010089963e-05, + "loss": 0.9105, + "step": 5084 + }, + { + "epoch": 0.2251419589785595, + "grad_norm": 1.7115803956985474, + "learning_rate": 1.9061335405335107e-05, + "loss": 0.9894, + "step": 5085 + }, + { + "epoch": 0.22518623468337337, + "grad_norm": 1.4631452560424805, + "learning_rate": 1.9060681584157637e-05, + "loss": 0.9257, + "step": 5086 + }, + { + "epoch": 0.22523051038818725, + "grad_norm": 2.6701836585998535, + "learning_rate": 1.9060027546573164e-05, + "loss": 1.0276, + "step": 5087 + }, + { + "epoch": 0.2252747860930011, + "grad_norm": 1.3627490997314453, + "learning_rate": 1.9059373292597308e-05, + "loss": 0.9584, + "step": 5088 + }, + { + "epoch": 0.225319061797815, + "grad_norm": 1.3377997875213623, + "learning_rate": 1.9058718822245697e-05, + "loss": 0.9618, + "step": 5089 + }, + { + "epoch": 0.22536333750262888, + "grad_norm": 1.1506739854812622, + "learning_rate": 1.9058064135533967e-05, + "loss": 0.8072, + "step": 5090 + }, + { + "epoch": 0.22540761320744274, + "grad_norm": 1.3974182605743408, + "learning_rate": 1.905740923247775e-05, + "loss": 0.873, + "step": 5091 + }, + { + "epoch": 0.22545188891225662, + "grad_norm": 1.441145658493042, + "learning_rate": 1.9056754113092686e-05, + "loss": 0.9676, + "step": 5092 + }, + { + "epoch": 0.2254961646170705, + "grad_norm": 1.3777313232421875, + "learning_rate": 1.905609877739442e-05, + "loss": 0.8996, + "step": 5093 + }, + { + "epoch": 0.22554044032188436, + "grad_norm": 1.497605800628662, + "learning_rate": 1.9055443225398616e-05, + "loss": 0.9857, + "step": 5094 + }, + { + "epoch": 0.22558471602669825, + "grad_norm": 1.4579293727874756, + "learning_rate": 1.9054787457120924e-05, + "loss": 0.9344, + "step": 5095 + }, + { + "epoch": 0.22562899173151213, + "grad_norm": 1.3696850538253784, + "learning_rate": 1.9054131472577e-05, + "loss": 0.9538, + "step": 5096 + }, + { + "epoch": 0.225673267436326, + "grad_norm": 1.4937787055969238, + "learning_rate": 1.9053475271782523e-05, + "loss": 0.9392, + "step": 5097 + }, + { + "epoch": 0.22571754314113987, + "grad_norm": 1.2697705030441284, + "learning_rate": 1.905281885475316e-05, + "loss": 0.8992, + "step": 5098 + }, + { + "epoch": 0.22576181884595375, + "grad_norm": 1.2714961767196655, + "learning_rate": 1.9052162221504584e-05, + "loss": 0.8702, + "step": 5099 + }, + { + "epoch": 0.22580609455076764, + "grad_norm": 1.2832086086273193, + "learning_rate": 1.905150537205249e-05, + "loss": 0.9244, + "step": 5100 + }, + { + "epoch": 0.2258503702555815, + "grad_norm": 1.236154317855835, + "learning_rate": 1.9050848306412558e-05, + "loss": 0.886, + "step": 5101 + }, + { + "epoch": 0.22589464596039538, + "grad_norm": 1.47946298122406, + "learning_rate": 1.9050191024600488e-05, + "loss": 0.9383, + "step": 5102 + }, + { + "epoch": 0.22593892166520926, + "grad_norm": 1.17635178565979, + "learning_rate": 1.904953352663197e-05, + "loss": 0.8685, + "step": 5103 + }, + { + "epoch": 0.22598319737002315, + "grad_norm": 1.9023499488830566, + "learning_rate": 1.9048875812522712e-05, + "loss": 0.9019, + "step": 5104 + }, + { + "epoch": 0.226027473074837, + "grad_norm": 1.643486499786377, + "learning_rate": 1.9048217882288422e-05, + "loss": 0.9594, + "step": 5105 + }, + { + "epoch": 0.2260717487796509, + "grad_norm": 1.3203202486038208, + "learning_rate": 1.904755973594482e-05, + "loss": 0.8661, + "step": 5106 + }, + { + "epoch": 0.22611602448446477, + "grad_norm": 1.217321515083313, + "learning_rate": 1.9046901373507616e-05, + "loss": 0.823, + "step": 5107 + }, + { + "epoch": 0.22616030018927863, + "grad_norm": 1.270387887954712, + "learning_rate": 1.904624279499254e-05, + "loss": 0.9334, + "step": 5108 + }, + { + "epoch": 0.2262045758940925, + "grad_norm": 1.2784682512283325, + "learning_rate": 1.9045584000415323e-05, + "loss": 0.885, + "step": 5109 + }, + { + "epoch": 0.2262488515989064, + "grad_norm": 1.3199244737625122, + "learning_rate": 1.9044924989791694e-05, + "loss": 0.8853, + "step": 5110 + }, + { + "epoch": 0.22629312730372025, + "grad_norm": 1.3673410415649414, + "learning_rate": 1.9044265763137398e-05, + "loss": 0.9224, + "step": 5111 + }, + { + "epoch": 0.22633740300853414, + "grad_norm": 1.0903065204620361, + "learning_rate": 1.9043606320468178e-05, + "loss": 0.8104, + "step": 5112 + }, + { + "epoch": 0.22638167871334802, + "grad_norm": 1.2207218408584595, + "learning_rate": 1.9042946661799784e-05, + "loss": 0.8808, + "step": 5113 + }, + { + "epoch": 0.2264259544181619, + "grad_norm": 1.2682080268859863, + "learning_rate": 1.9042286787147973e-05, + "loss": 0.9211, + "step": 5114 + }, + { + "epoch": 0.22647023012297576, + "grad_norm": 1.4172494411468506, + "learning_rate": 1.9041626696528503e-05, + "loss": 0.9801, + "step": 5115 + }, + { + "epoch": 0.22651450582778965, + "grad_norm": 1.4804730415344238, + "learning_rate": 1.9040966389957143e-05, + "loss": 0.9308, + "step": 5116 + }, + { + "epoch": 0.22655878153260353, + "grad_norm": 1.2464962005615234, + "learning_rate": 1.904030586744966e-05, + "loss": 0.9545, + "step": 5117 + }, + { + "epoch": 0.2266030572374174, + "grad_norm": 1.1917253732681274, + "learning_rate": 1.9039645129021837e-05, + "loss": 0.8883, + "step": 5118 + }, + { + "epoch": 0.22664733294223127, + "grad_norm": 1.2884515523910522, + "learning_rate": 1.9038984174689447e-05, + "loss": 0.8568, + "step": 5119 + }, + { + "epoch": 0.22669160864704516, + "grad_norm": 1.533509373664856, + "learning_rate": 1.9038323004468278e-05, + "loss": 0.8858, + "step": 5120 + }, + { + "epoch": 0.226735884351859, + "grad_norm": 1.2600173950195312, + "learning_rate": 1.903766161837413e-05, + "loss": 0.9102, + "step": 5121 + }, + { + "epoch": 0.2267801600566729, + "grad_norm": 1.2659631967544556, + "learning_rate": 1.9037000016422787e-05, + "loss": 0.9607, + "step": 5122 + }, + { + "epoch": 0.22682443576148678, + "grad_norm": 2.288602113723755, + "learning_rate": 1.9036338198630056e-05, + "loss": 0.9432, + "step": 5123 + }, + { + "epoch": 0.22686871146630067, + "grad_norm": 1.2525213956832886, + "learning_rate": 1.9035676165011748e-05, + "loss": 0.8653, + "step": 5124 + }, + { + "epoch": 0.22691298717111452, + "grad_norm": 1.1866847276687622, + "learning_rate": 1.903501391558367e-05, + "loss": 0.8294, + "step": 5125 + }, + { + "epoch": 0.2269572628759284, + "grad_norm": 1.2823954820632935, + "learning_rate": 1.9034351450361646e-05, + "loss": 0.8729, + "step": 5126 + }, + { + "epoch": 0.2270015385807423, + "grad_norm": 1.6194099187850952, + "learning_rate": 1.903368876936149e-05, + "loss": 0.9918, + "step": 5127 + }, + { + "epoch": 0.22704581428555615, + "grad_norm": 1.3126195669174194, + "learning_rate": 1.903302587259903e-05, + "loss": 0.8274, + "step": 5128 + }, + { + "epoch": 0.22709008999037003, + "grad_norm": 1.4921704530715942, + "learning_rate": 1.9032362760090105e-05, + "loss": 0.9099, + "step": 5129 + }, + { + "epoch": 0.22713436569518392, + "grad_norm": 2.0048630237579346, + "learning_rate": 1.9031699431850553e-05, + "loss": 0.967, + "step": 5130 + }, + { + "epoch": 0.2271786413999978, + "grad_norm": 2.600107192993164, + "learning_rate": 1.903103588789621e-05, + "loss": 1.0275, + "step": 5131 + }, + { + "epoch": 0.22722291710481166, + "grad_norm": 1.3109596967697144, + "learning_rate": 1.903037212824293e-05, + "loss": 0.9173, + "step": 5132 + }, + { + "epoch": 0.22726719280962554, + "grad_norm": 1.2412179708480835, + "learning_rate": 1.9029708152906564e-05, + "loss": 0.8279, + "step": 5133 + }, + { + "epoch": 0.22731146851443942, + "grad_norm": 1.3276728391647339, + "learning_rate": 1.9029043961902973e-05, + "loss": 0.8638, + "step": 5134 + }, + { + "epoch": 0.22735574421925328, + "grad_norm": 1.1209092140197754, + "learning_rate": 1.9028379555248015e-05, + "loss": 0.8718, + "step": 5135 + }, + { + "epoch": 0.22740001992406716, + "grad_norm": 1.3638668060302734, + "learning_rate": 1.9027714932957564e-05, + "loss": 0.8846, + "step": 5136 + }, + { + "epoch": 0.22744429562888105, + "grad_norm": 1.6510499715805054, + "learning_rate": 1.902705009504749e-05, + "loss": 0.9125, + "step": 5137 + }, + { + "epoch": 0.2274885713336949, + "grad_norm": 1.3340400457382202, + "learning_rate": 1.9026385041533678e-05, + "loss": 0.8968, + "step": 5138 + }, + { + "epoch": 0.2275328470385088, + "grad_norm": 1.1381337642669678, + "learning_rate": 1.9025719772432006e-05, + "loss": 0.8068, + "step": 5139 + }, + { + "epoch": 0.22757712274332267, + "grad_norm": 1.8133176565170288, + "learning_rate": 1.902505428775837e-05, + "loss": 0.9529, + "step": 5140 + }, + { + "epoch": 0.22762139844813656, + "grad_norm": 1.316295862197876, + "learning_rate": 1.902438858752866e-05, + "loss": 0.8249, + "step": 5141 + }, + { + "epoch": 0.22766567415295041, + "grad_norm": 1.228957176208496, + "learning_rate": 1.9023722671758773e-05, + "loss": 0.8999, + "step": 5142 + }, + { + "epoch": 0.2277099498577643, + "grad_norm": 1.3095654249191284, + "learning_rate": 1.9023056540464622e-05, + "loss": 0.8724, + "step": 5143 + }, + { + "epoch": 0.22775422556257818, + "grad_norm": 1.4168752431869507, + "learning_rate": 1.902239019366211e-05, + "loss": 0.9249, + "step": 5144 + }, + { + "epoch": 0.22779850126739204, + "grad_norm": 1.5980256795883179, + "learning_rate": 1.9021723631367152e-05, + "loss": 0.9481, + "step": 5145 + }, + { + "epoch": 0.22784277697220592, + "grad_norm": 1.3806617259979248, + "learning_rate": 1.9021056853595677e-05, + "loss": 0.8705, + "step": 5146 + }, + { + "epoch": 0.2278870526770198, + "grad_norm": 1.3287632465362549, + "learning_rate": 1.90203898603636e-05, + "loss": 0.9255, + "step": 5147 + }, + { + "epoch": 0.2279313283818337, + "grad_norm": 1.3001713752746582, + "learning_rate": 1.9019722651686855e-05, + "loss": 0.897, + "step": 5148 + }, + { + "epoch": 0.22797560408664755, + "grad_norm": 1.7785463333129883, + "learning_rate": 1.901905522758138e-05, + "loss": 0.9701, + "step": 5149 + }, + { + "epoch": 0.22801987979146143, + "grad_norm": 1.3930246829986572, + "learning_rate": 1.9018387588063115e-05, + "loss": 0.8613, + "step": 5150 + }, + { + "epoch": 0.22806415549627532, + "grad_norm": 1.2086262702941895, + "learning_rate": 1.9017719733148007e-05, + "loss": 0.8548, + "step": 5151 + }, + { + "epoch": 0.22810843120108917, + "grad_norm": 1.3215687274932861, + "learning_rate": 1.9017051662852003e-05, + "loss": 0.8886, + "step": 5152 + }, + { + "epoch": 0.22815270690590306, + "grad_norm": 1.0631356239318848, + "learning_rate": 1.9016383377191063e-05, + "loss": 0.7584, + "step": 5153 + }, + { + "epoch": 0.22819698261071694, + "grad_norm": 1.288852334022522, + "learning_rate": 1.9015714876181147e-05, + "loss": 0.8706, + "step": 5154 + }, + { + "epoch": 0.2282412583155308, + "grad_norm": 1.3160369396209717, + "learning_rate": 1.9015046159838223e-05, + "loss": 0.9648, + "step": 5155 + }, + { + "epoch": 0.22828553402034468, + "grad_norm": 1.4317609071731567, + "learning_rate": 1.901437722817826e-05, + "loss": 0.9226, + "step": 5156 + }, + { + "epoch": 0.22832980972515857, + "grad_norm": 1.2695119380950928, + "learning_rate": 1.9013708081217238e-05, + "loss": 0.911, + "step": 5157 + }, + { + "epoch": 0.22837408542997245, + "grad_norm": 1.4704692363739014, + "learning_rate": 1.9013038718971137e-05, + "loss": 0.9767, + "step": 5158 + }, + { + "epoch": 0.2284183611347863, + "grad_norm": 1.356365442276001, + "learning_rate": 1.9012369141455945e-05, + "loss": 0.9117, + "step": 5159 + }, + { + "epoch": 0.2284626368396002, + "grad_norm": 1.364768624305725, + "learning_rate": 1.9011699348687655e-05, + "loss": 0.8558, + "step": 5160 + }, + { + "epoch": 0.22850691254441408, + "grad_norm": 1.625410556793213, + "learning_rate": 1.9011029340682263e-05, + "loss": 0.9101, + "step": 5161 + }, + { + "epoch": 0.22855118824922793, + "grad_norm": 1.3005496263504028, + "learning_rate": 1.9010359117455773e-05, + "loss": 0.9043, + "step": 5162 + }, + { + "epoch": 0.22859546395404182, + "grad_norm": 1.2577897310256958, + "learning_rate": 1.900968867902419e-05, + "loss": 0.9226, + "step": 5163 + }, + { + "epoch": 0.2286397396588557, + "grad_norm": 1.7577576637268066, + "learning_rate": 1.9009018025403534e-05, + "loss": 0.9946, + "step": 5164 + }, + { + "epoch": 0.22868401536366956, + "grad_norm": 1.2383086681365967, + "learning_rate": 1.9008347156609813e-05, + "loss": 0.8883, + "step": 5165 + }, + { + "epoch": 0.22872829106848344, + "grad_norm": 1.704055666923523, + "learning_rate": 1.9007676072659057e-05, + "loss": 1.0093, + "step": 5166 + }, + { + "epoch": 0.22877256677329733, + "grad_norm": 1.053716778755188, + "learning_rate": 1.900700477356729e-05, + "loss": 0.7894, + "step": 5167 + }, + { + "epoch": 0.2288168424781112, + "grad_norm": 1.2873475551605225, + "learning_rate": 1.900633325935055e-05, + "loss": 0.8555, + "step": 5168 + }, + { + "epoch": 0.22886111818292507, + "grad_norm": 1.3395696878433228, + "learning_rate": 1.9005661530024873e-05, + "loss": 0.9555, + "step": 5169 + }, + { + "epoch": 0.22890539388773895, + "grad_norm": 1.5119415521621704, + "learning_rate": 1.9004989585606303e-05, + "loss": 0.9685, + "step": 5170 + }, + { + "epoch": 0.22894966959255283, + "grad_norm": 1.3111943006515503, + "learning_rate": 1.9004317426110888e-05, + "loss": 0.8723, + "step": 5171 + }, + { + "epoch": 0.2289939452973667, + "grad_norm": 1.4776949882507324, + "learning_rate": 1.9003645051554686e-05, + "loss": 0.9605, + "step": 5172 + }, + { + "epoch": 0.22903822100218058, + "grad_norm": 1.6467608213424683, + "learning_rate": 1.900297246195375e-05, + "loss": 0.9032, + "step": 5173 + }, + { + "epoch": 0.22908249670699446, + "grad_norm": 1.2516909837722778, + "learning_rate": 1.9002299657324147e-05, + "loss": 0.8493, + "step": 5174 + }, + { + "epoch": 0.22912677241180834, + "grad_norm": 1.2489492893218994, + "learning_rate": 1.9001626637681947e-05, + "loss": 0.926, + "step": 5175 + }, + { + "epoch": 0.2291710481166222, + "grad_norm": 1.3694690465927124, + "learning_rate": 1.9000953403043224e-05, + "loss": 0.906, + "step": 5176 + }, + { + "epoch": 0.22921532382143608, + "grad_norm": 1.3039517402648926, + "learning_rate": 1.9000279953424057e-05, + "loss": 0.8962, + "step": 5177 + }, + { + "epoch": 0.22925959952624997, + "grad_norm": 1.7286896705627441, + "learning_rate": 1.8999606288840535e-05, + "loss": 1.0006, + "step": 5178 + }, + { + "epoch": 0.22930387523106383, + "grad_norm": 1.1065798997879028, + "learning_rate": 1.8998932409308742e-05, + "loss": 0.7553, + "step": 5179 + }, + { + "epoch": 0.2293481509358777, + "grad_norm": 2.1604857444763184, + "learning_rate": 1.8998258314844778e-05, + "loss": 0.8967, + "step": 5180 + }, + { + "epoch": 0.2293924266406916, + "grad_norm": 1.3239548206329346, + "learning_rate": 1.8997584005464737e-05, + "loss": 0.9089, + "step": 5181 + }, + { + "epoch": 0.22943670234550545, + "grad_norm": 1.161272644996643, + "learning_rate": 1.8996909481184732e-05, + "loss": 0.8347, + "step": 5182 + }, + { + "epoch": 0.22948097805031933, + "grad_norm": 1.339529275894165, + "learning_rate": 1.8996234742020868e-05, + "loss": 0.9137, + "step": 5183 + }, + { + "epoch": 0.22952525375513322, + "grad_norm": 1.61497962474823, + "learning_rate": 1.8995559787989258e-05, + "loss": 0.957, + "step": 5184 + }, + { + "epoch": 0.2295695294599471, + "grad_norm": 1.2405729293823242, + "learning_rate": 1.8994884619106034e-05, + "loss": 0.8395, + "step": 5185 + }, + { + "epoch": 0.22961380516476096, + "grad_norm": 1.200392246246338, + "learning_rate": 1.899420923538731e-05, + "loss": 0.8789, + "step": 5186 + }, + { + "epoch": 0.22965808086957484, + "grad_norm": 2.1910200119018555, + "learning_rate": 1.8993533636849223e-05, + "loss": 1.0264, + "step": 5187 + }, + { + "epoch": 0.22970235657438873, + "grad_norm": 1.789176344871521, + "learning_rate": 1.899285782350791e-05, + "loss": 0.9026, + "step": 5188 + }, + { + "epoch": 0.22974663227920258, + "grad_norm": 1.3305119276046753, + "learning_rate": 1.8992181795379508e-05, + "loss": 0.9257, + "step": 5189 + }, + { + "epoch": 0.22979090798401647, + "grad_norm": 1.5219388008117676, + "learning_rate": 1.8991505552480164e-05, + "loss": 0.9982, + "step": 5190 + }, + { + "epoch": 0.22983518368883035, + "grad_norm": 1.63774573802948, + "learning_rate": 1.8990829094826037e-05, + "loss": 0.9866, + "step": 5191 + }, + { + "epoch": 0.2298794593936442, + "grad_norm": 1.235924243927002, + "learning_rate": 1.8990152422433273e-05, + "loss": 0.8693, + "step": 5192 + }, + { + "epoch": 0.2299237350984581, + "grad_norm": 1.2568391561508179, + "learning_rate": 1.898947553531804e-05, + "loss": 0.9086, + "step": 5193 + }, + { + "epoch": 0.22996801080327198, + "grad_norm": 1.156771183013916, + "learning_rate": 1.8988798433496502e-05, + "loss": 0.8661, + "step": 5194 + }, + { + "epoch": 0.23001228650808586, + "grad_norm": 1.2632883787155151, + "learning_rate": 1.898812111698483e-05, + "loss": 0.8761, + "step": 5195 + }, + { + "epoch": 0.23005656221289972, + "grad_norm": 1.2647854089736938, + "learning_rate": 1.898744358579921e-05, + "loss": 0.8791, + "step": 5196 + }, + { + "epoch": 0.2301008379177136, + "grad_norm": 1.3653732538223267, + "learning_rate": 1.898676583995581e-05, + "loss": 0.9056, + "step": 5197 + }, + { + "epoch": 0.2301451136225275, + "grad_norm": 1.3149309158325195, + "learning_rate": 1.8986087879470832e-05, + "loss": 0.8884, + "step": 5198 + }, + { + "epoch": 0.23018938932734134, + "grad_norm": 1.9736725091934204, + "learning_rate": 1.8985409704360457e-05, + "loss": 0.9636, + "step": 5199 + }, + { + "epoch": 0.23023366503215523, + "grad_norm": 1.1537365913391113, + "learning_rate": 1.898473131464089e-05, + "loss": 0.8495, + "step": 5200 + }, + { + "epoch": 0.2302779407369691, + "grad_norm": 1.625519037246704, + "learning_rate": 1.898405271032833e-05, + "loss": 0.9104, + "step": 5201 + }, + { + "epoch": 0.230322216441783, + "grad_norm": 1.349205493927002, + "learning_rate": 1.8983373891438986e-05, + "loss": 0.8994, + "step": 5202 + }, + { + "epoch": 0.23036649214659685, + "grad_norm": 1.5719791650772095, + "learning_rate": 1.8982694857989072e-05, + "loss": 0.8541, + "step": 5203 + }, + { + "epoch": 0.23041076785141074, + "grad_norm": 1.5044500827789307, + "learning_rate": 1.8982015609994805e-05, + "loss": 0.8838, + "step": 5204 + }, + { + "epoch": 0.23045504355622462, + "grad_norm": 1.4648360013961792, + "learning_rate": 1.898133614747241e-05, + "loss": 0.9469, + "step": 5205 + }, + { + "epoch": 0.23049931926103848, + "grad_norm": 1.3555916547775269, + "learning_rate": 1.8980656470438113e-05, + "loss": 0.8701, + "step": 5206 + }, + { + "epoch": 0.23054359496585236, + "grad_norm": 1.5300021171569824, + "learning_rate": 1.8979976578908146e-05, + "loss": 0.9802, + "step": 5207 + }, + { + "epoch": 0.23058787067066625, + "grad_norm": 1.3705517053604126, + "learning_rate": 1.8979296472898754e-05, + "loss": 0.8517, + "step": 5208 + }, + { + "epoch": 0.2306321463754801, + "grad_norm": 1.2494651079177856, + "learning_rate": 1.8978616152426176e-05, + "loss": 0.8876, + "step": 5209 + }, + { + "epoch": 0.23067642208029399, + "grad_norm": 1.2257295846939087, + "learning_rate": 1.8977935617506662e-05, + "loss": 0.8687, + "step": 5210 + }, + { + "epoch": 0.23072069778510787, + "grad_norm": 1.2866287231445312, + "learning_rate": 1.8977254868156465e-05, + "loss": 0.8266, + "step": 5211 + }, + { + "epoch": 0.23076497348992175, + "grad_norm": 1.5728501081466675, + "learning_rate": 1.8976573904391843e-05, + "loss": 0.9431, + "step": 5212 + }, + { + "epoch": 0.2308092491947356, + "grad_norm": 1.5566426515579224, + "learning_rate": 1.8975892726229065e-05, + "loss": 0.8671, + "step": 5213 + }, + { + "epoch": 0.2308535248995495, + "grad_norm": 1.4948347806930542, + "learning_rate": 1.89752113336844e-05, + "loss": 0.9507, + "step": 5214 + }, + { + "epoch": 0.23089780060436338, + "grad_norm": 1.6382529735565186, + "learning_rate": 1.8974529726774115e-05, + "loss": 0.9363, + "step": 5215 + }, + { + "epoch": 0.23094207630917724, + "grad_norm": 1.4050706624984741, + "learning_rate": 1.8973847905514494e-05, + "loss": 0.956, + "step": 5216 + }, + { + "epoch": 0.23098635201399112, + "grad_norm": 1.2959574460983276, + "learning_rate": 1.897316586992183e-05, + "loss": 0.9139, + "step": 5217 + }, + { + "epoch": 0.231030627718805, + "grad_norm": 1.2945271730422974, + "learning_rate": 1.8972483620012402e-05, + "loss": 0.8568, + "step": 5218 + }, + { + "epoch": 0.2310749034236189, + "grad_norm": 1.5599126815795898, + "learning_rate": 1.8971801155802505e-05, + "loss": 0.9978, + "step": 5219 + }, + { + "epoch": 0.23111917912843274, + "grad_norm": 1.3586519956588745, + "learning_rate": 1.8971118477308444e-05, + "loss": 0.9042, + "step": 5220 + }, + { + "epoch": 0.23116345483324663, + "grad_norm": 1.4531582593917847, + "learning_rate": 1.8970435584546523e-05, + "loss": 1.0067, + "step": 5221 + }, + { + "epoch": 0.2312077305380605, + "grad_norm": 1.1337209939956665, + "learning_rate": 1.8969752477533053e-05, + "loss": 0.8731, + "step": 5222 + }, + { + "epoch": 0.23125200624287437, + "grad_norm": 1.5066648721694946, + "learning_rate": 1.896906915628435e-05, + "loss": 0.9662, + "step": 5223 + }, + { + "epoch": 0.23129628194768825, + "grad_norm": 1.5370553731918335, + "learning_rate": 1.896838562081673e-05, + "loss": 0.9288, + "step": 5224 + }, + { + "epoch": 0.23134055765250214, + "grad_norm": 1.2162461280822754, + "learning_rate": 1.8967701871146526e-05, + "loss": 0.8963, + "step": 5225 + }, + { + "epoch": 0.231384833357316, + "grad_norm": 1.2196216583251953, + "learning_rate": 1.896701790729006e-05, + "loss": 0.8363, + "step": 5226 + }, + { + "epoch": 0.23142910906212988, + "grad_norm": 1.4559327363967896, + "learning_rate": 1.8966333729263674e-05, + "loss": 0.9121, + "step": 5227 + }, + { + "epoch": 0.23147338476694376, + "grad_norm": 1.5675824880599976, + "learning_rate": 1.8965649337083708e-05, + "loss": 1.0221, + "step": 5228 + }, + { + "epoch": 0.23151766047175765, + "grad_norm": 1.3079928159713745, + "learning_rate": 1.896496473076651e-05, + "loss": 0.9127, + "step": 5229 + }, + { + "epoch": 0.2315619361765715, + "grad_norm": 1.2378171682357788, + "learning_rate": 1.8964279910328426e-05, + "loss": 0.9163, + "step": 5230 + }, + { + "epoch": 0.2316062118813854, + "grad_norm": 1.3269230127334595, + "learning_rate": 1.8963594875785815e-05, + "loss": 0.8551, + "step": 5231 + }, + { + "epoch": 0.23165048758619927, + "grad_norm": 1.3891148567199707, + "learning_rate": 1.8962909627155038e-05, + "loss": 0.9113, + "step": 5232 + }, + { + "epoch": 0.23169476329101313, + "grad_norm": 1.2383720874786377, + "learning_rate": 1.8962224164452465e-05, + "loss": 0.8746, + "step": 5233 + }, + { + "epoch": 0.231739038995827, + "grad_norm": 1.2047159671783447, + "learning_rate": 1.8961538487694465e-05, + "loss": 0.9212, + "step": 5234 + }, + { + "epoch": 0.2317833147006409, + "grad_norm": 1.3713067770004272, + "learning_rate": 1.896085259689741e-05, + "loss": 0.9809, + "step": 5235 + }, + { + "epoch": 0.23182759040545475, + "grad_norm": 1.4811244010925293, + "learning_rate": 1.8960166492077694e-05, + "loss": 0.8329, + "step": 5236 + }, + { + "epoch": 0.23187186611026864, + "grad_norm": 1.2644866704940796, + "learning_rate": 1.8959480173251694e-05, + "loss": 0.928, + "step": 5237 + }, + { + "epoch": 0.23191614181508252, + "grad_norm": 1.5575133562088013, + "learning_rate": 1.8958793640435807e-05, + "loss": 0.957, + "step": 5238 + }, + { + "epoch": 0.2319604175198964, + "grad_norm": 1.221234917640686, + "learning_rate": 1.8958106893646424e-05, + "loss": 0.837, + "step": 5239 + }, + { + "epoch": 0.23200469322471026, + "grad_norm": 1.096923828125, + "learning_rate": 1.8957419932899955e-05, + "loss": 0.7512, + "step": 5240 + }, + { + "epoch": 0.23204896892952415, + "grad_norm": 2.1606900691986084, + "learning_rate": 1.8956732758212803e-05, + "loss": 1.0032, + "step": 5241 + }, + { + "epoch": 0.23209324463433803, + "grad_norm": 1.2637393474578857, + "learning_rate": 1.8956045369601382e-05, + "loss": 0.8907, + "step": 5242 + }, + { + "epoch": 0.2321375203391519, + "grad_norm": 1.5567717552185059, + "learning_rate": 1.895535776708211e-05, + "loss": 0.9654, + "step": 5243 + }, + { + "epoch": 0.23218179604396577, + "grad_norm": 1.2050455808639526, + "learning_rate": 1.895466995067141e-05, + "loss": 0.8407, + "step": 5244 + }, + { + "epoch": 0.23222607174877966, + "grad_norm": 2.000636577606201, + "learning_rate": 1.8953981920385712e-05, + "loss": 0.9307, + "step": 5245 + }, + { + "epoch": 0.23227034745359354, + "grad_norm": 1.3366131782531738, + "learning_rate": 1.8953293676241443e-05, + "loss": 0.9058, + "step": 5246 + }, + { + "epoch": 0.2323146231584074, + "grad_norm": 1.625659704208374, + "learning_rate": 1.8952605218255043e-05, + "loss": 0.949, + "step": 5247 + }, + { + "epoch": 0.23235889886322128, + "grad_norm": 1.337778091430664, + "learning_rate": 1.8951916546442958e-05, + "loss": 0.8695, + "step": 5248 + }, + { + "epoch": 0.23240317456803516, + "grad_norm": 1.25348699092865, + "learning_rate": 1.8951227660821637e-05, + "loss": 0.806, + "step": 5249 + }, + { + "epoch": 0.23244745027284902, + "grad_norm": 1.356162667274475, + "learning_rate": 1.8950538561407533e-05, + "loss": 0.7832, + "step": 5250 + }, + { + "epoch": 0.2324917259776629, + "grad_norm": 1.6161940097808838, + "learning_rate": 1.89498492482171e-05, + "loss": 0.9845, + "step": 5251 + }, + { + "epoch": 0.2325360016824768, + "grad_norm": 1.3914809226989746, + "learning_rate": 1.894915972126681e-05, + "loss": 0.8944, + "step": 5252 + }, + { + "epoch": 0.23258027738729065, + "grad_norm": 1.3544535636901855, + "learning_rate": 1.8948469980573118e-05, + "loss": 0.9694, + "step": 5253 + }, + { + "epoch": 0.23262455309210453, + "grad_norm": 1.3837902545928955, + "learning_rate": 1.8947780026152512e-05, + "loss": 0.9187, + "step": 5254 + }, + { + "epoch": 0.23266882879691841, + "grad_norm": 1.2237861156463623, + "learning_rate": 1.8947089858021465e-05, + "loss": 0.8617, + "step": 5255 + }, + { + "epoch": 0.2327131045017323, + "grad_norm": 1.6758546829223633, + "learning_rate": 1.894639947619646e-05, + "loss": 0.9184, + "step": 5256 + }, + { + "epoch": 0.23275738020654616, + "grad_norm": 1.4390602111816406, + "learning_rate": 1.8945708880693995e-05, + "loss": 0.9604, + "step": 5257 + }, + { + "epoch": 0.23280165591136004, + "grad_norm": 1.351305365562439, + "learning_rate": 1.8945018071530548e-05, + "loss": 0.8268, + "step": 5258 + }, + { + "epoch": 0.23284593161617392, + "grad_norm": 1.3264589309692383, + "learning_rate": 1.8944327048722634e-05, + "loss": 0.8055, + "step": 5259 + }, + { + "epoch": 0.23289020732098778, + "grad_norm": 1.5970689058303833, + "learning_rate": 1.894363581228675e-05, + "loss": 0.9308, + "step": 5260 + }, + { + "epoch": 0.23293448302580166, + "grad_norm": 1.2769170999526978, + "learning_rate": 1.89429443622394e-05, + "loss": 0.8503, + "step": 5261 + }, + { + "epoch": 0.23297875873061555, + "grad_norm": 1.2286643981933594, + "learning_rate": 1.8942252698597113e-05, + "loss": 0.9001, + "step": 5262 + }, + { + "epoch": 0.23302303443542943, + "grad_norm": 1.2870253324508667, + "learning_rate": 1.8941560821376403e-05, + "loss": 0.9255, + "step": 5263 + }, + { + "epoch": 0.2330673101402433, + "grad_norm": 1.3538293838500977, + "learning_rate": 1.894086873059379e-05, + "loss": 0.8376, + "step": 5264 + }, + { + "epoch": 0.23311158584505717, + "grad_norm": 1.41509211063385, + "learning_rate": 1.8940176426265803e-05, + "loss": 0.9393, + "step": 5265 + }, + { + "epoch": 0.23315586154987106, + "grad_norm": 1.4712135791778564, + "learning_rate": 1.8939483908408986e-05, + "loss": 1.0542, + "step": 5266 + }, + { + "epoch": 0.2332001372546849, + "grad_norm": 1.2264456748962402, + "learning_rate": 1.8938791177039875e-05, + "loss": 0.8348, + "step": 5267 + }, + { + "epoch": 0.2332444129594988, + "grad_norm": 3.8693647384643555, + "learning_rate": 1.8938098232175016e-05, + "loss": 0.8685, + "step": 5268 + }, + { + "epoch": 0.23328868866431268, + "grad_norm": 1.0214793682098389, + "learning_rate": 1.8937405073830956e-05, + "loss": 0.8202, + "step": 5269 + }, + { + "epoch": 0.23333296436912654, + "grad_norm": 1.3939775228500366, + "learning_rate": 1.8936711702024252e-05, + "loss": 0.8806, + "step": 5270 + }, + { + "epoch": 0.23337724007394042, + "grad_norm": 1.3465098142623901, + "learning_rate": 1.893601811677147e-05, + "loss": 0.8391, + "step": 5271 + }, + { + "epoch": 0.2334215157787543, + "grad_norm": 1.4015400409698486, + "learning_rate": 1.8935324318089166e-05, + "loss": 0.948, + "step": 5272 + }, + { + "epoch": 0.2334657914835682, + "grad_norm": 1.400002360343933, + "learning_rate": 1.8934630305993917e-05, + "loss": 0.9993, + "step": 5273 + }, + { + "epoch": 0.23351006718838205, + "grad_norm": 1.34634530544281, + "learning_rate": 1.89339360805023e-05, + "loss": 0.9641, + "step": 5274 + }, + { + "epoch": 0.23355434289319593, + "grad_norm": 1.2335535287857056, + "learning_rate": 1.8933241641630892e-05, + "loss": 0.8501, + "step": 5275 + }, + { + "epoch": 0.23359861859800982, + "grad_norm": 1.3753697872161865, + "learning_rate": 1.8932546989396283e-05, + "loss": 0.9364, + "step": 5276 + }, + { + "epoch": 0.23364289430282367, + "grad_norm": 1.277371883392334, + "learning_rate": 1.893185212381506e-05, + "loss": 0.8678, + "step": 5277 + }, + { + "epoch": 0.23368717000763756, + "grad_norm": 1.911144733428955, + "learning_rate": 1.8931157044903824e-05, + "loss": 0.8373, + "step": 5278 + }, + { + "epoch": 0.23373144571245144, + "grad_norm": 1.4244263172149658, + "learning_rate": 1.8930461752679173e-05, + "loss": 0.9365, + "step": 5279 + }, + { + "epoch": 0.2337757214172653, + "grad_norm": 1.326552152633667, + "learning_rate": 1.8929766247157715e-05, + "loss": 0.8979, + "step": 5280 + }, + { + "epoch": 0.23381999712207918, + "grad_norm": 2.0670278072357178, + "learning_rate": 1.8929070528356065e-05, + "loss": 0.9769, + "step": 5281 + }, + { + "epoch": 0.23386427282689307, + "grad_norm": 1.2531201839447021, + "learning_rate": 1.892837459629083e-05, + "loss": 0.9346, + "step": 5282 + }, + { + "epoch": 0.23390854853170695, + "grad_norm": 1.6602524518966675, + "learning_rate": 1.892767845097864e-05, + "loss": 0.8734, + "step": 5283 + }, + { + "epoch": 0.2339528242365208, + "grad_norm": 1.401930570602417, + "learning_rate": 1.8926982092436117e-05, + "loss": 0.964, + "step": 5284 + }, + { + "epoch": 0.2339970999413347, + "grad_norm": 1.1616531610488892, + "learning_rate": 1.8926285520679895e-05, + "loss": 0.8603, + "step": 5285 + }, + { + "epoch": 0.23404137564614858, + "grad_norm": 1.5412122011184692, + "learning_rate": 1.8925588735726615e-05, + "loss": 0.8444, + "step": 5286 + }, + { + "epoch": 0.23408565135096243, + "grad_norm": 1.5005730390548706, + "learning_rate": 1.8924891737592914e-05, + "loss": 0.9675, + "step": 5287 + }, + { + "epoch": 0.23412992705577632, + "grad_norm": 1.2388795614242554, + "learning_rate": 1.8924194526295442e-05, + "loss": 0.9508, + "step": 5288 + }, + { + "epoch": 0.2341742027605902, + "grad_norm": 1.2930326461791992, + "learning_rate": 1.8923497101850848e-05, + "loss": 0.8834, + "step": 5289 + }, + { + "epoch": 0.23421847846540408, + "grad_norm": 1.8093433380126953, + "learning_rate": 1.892279946427579e-05, + "loss": 1.0239, + "step": 5290 + }, + { + "epoch": 0.23426275417021794, + "grad_norm": 1.372978687286377, + "learning_rate": 1.8922101613586936e-05, + "loss": 0.9072, + "step": 5291 + }, + { + "epoch": 0.23430702987503182, + "grad_norm": 1.4257298707962036, + "learning_rate": 1.8921403549800947e-05, + "loss": 0.9612, + "step": 5292 + }, + { + "epoch": 0.2343513055798457, + "grad_norm": 1.5907325744628906, + "learning_rate": 1.8920705272934497e-05, + "loss": 0.9606, + "step": 5293 + }, + { + "epoch": 0.23439558128465957, + "grad_norm": 1.294299602508545, + "learning_rate": 1.892000678300427e-05, + "loss": 0.884, + "step": 5294 + }, + { + "epoch": 0.23443985698947345, + "grad_norm": 1.530411958694458, + "learning_rate": 1.891930808002694e-05, + "loss": 0.9259, + "step": 5295 + }, + { + "epoch": 0.23448413269428733, + "grad_norm": 1.4561703205108643, + "learning_rate": 1.89186091640192e-05, + "loss": 0.8705, + "step": 5296 + }, + { + "epoch": 0.2345284083991012, + "grad_norm": 1.4572205543518066, + "learning_rate": 1.8917910034997742e-05, + "loss": 0.9672, + "step": 5297 + }, + { + "epoch": 0.23457268410391507, + "grad_norm": 1.4505029916763306, + "learning_rate": 1.8917210692979263e-05, + "loss": 0.9092, + "step": 5298 + }, + { + "epoch": 0.23461695980872896, + "grad_norm": 1.2935696840286255, + "learning_rate": 1.891651113798047e-05, + "loss": 0.8458, + "step": 5299 + }, + { + "epoch": 0.23466123551354284, + "grad_norm": 1.444312572479248, + "learning_rate": 1.8915811370018065e-05, + "loss": 0.9675, + "step": 5300 + }, + { + "epoch": 0.2347055112183567, + "grad_norm": 1.6718733310699463, + "learning_rate": 1.8915111389108767e-05, + "loss": 0.8988, + "step": 5301 + }, + { + "epoch": 0.23474978692317058, + "grad_norm": 1.3969287872314453, + "learning_rate": 1.8914411195269294e-05, + "loss": 0.8944, + "step": 5302 + }, + { + "epoch": 0.23479406262798447, + "grad_norm": 1.5410226583480835, + "learning_rate": 1.8913710788516363e-05, + "loss": 0.9222, + "step": 5303 + }, + { + "epoch": 0.23483833833279832, + "grad_norm": 1.2105464935302734, + "learning_rate": 1.8913010168866712e-05, + "loss": 0.9241, + "step": 5304 + }, + { + "epoch": 0.2348826140376122, + "grad_norm": 1.3564140796661377, + "learning_rate": 1.8912309336337068e-05, + "loss": 0.8884, + "step": 5305 + }, + { + "epoch": 0.2349268897424261, + "grad_norm": 1.187113881111145, + "learning_rate": 1.8911608290944175e-05, + "loss": 0.8413, + "step": 5306 + }, + { + "epoch": 0.23497116544723998, + "grad_norm": 1.2441874742507935, + "learning_rate": 1.891090703270477e-05, + "loss": 0.8566, + "step": 5307 + }, + { + "epoch": 0.23501544115205383, + "grad_norm": 1.511705994606018, + "learning_rate": 1.891020556163561e-05, + "loss": 0.8747, + "step": 5308 + }, + { + "epoch": 0.23505971685686772, + "grad_norm": 1.5710601806640625, + "learning_rate": 1.8909503877753444e-05, + "loss": 0.898, + "step": 5309 + }, + { + "epoch": 0.2351039925616816, + "grad_norm": 1.305084466934204, + "learning_rate": 1.890880198107503e-05, + "loss": 0.9042, + "step": 5310 + }, + { + "epoch": 0.23514826826649546, + "grad_norm": 1.3358362913131714, + "learning_rate": 1.8908099871617137e-05, + "loss": 0.8494, + "step": 5311 + }, + { + "epoch": 0.23519254397130934, + "grad_norm": 1.442428469657898, + "learning_rate": 1.890739754939653e-05, + "loss": 0.9036, + "step": 5312 + }, + { + "epoch": 0.23523681967612323, + "grad_norm": 1.6870274543762207, + "learning_rate": 1.890669501442999e-05, + "loss": 0.8955, + "step": 5313 + }, + { + "epoch": 0.23528109538093708, + "grad_norm": 1.272101640701294, + "learning_rate": 1.8905992266734287e-05, + "loss": 0.8555, + "step": 5314 + }, + { + "epoch": 0.23532537108575097, + "grad_norm": 1.2846474647521973, + "learning_rate": 1.890528930632621e-05, + "loss": 0.8428, + "step": 5315 + }, + { + "epoch": 0.23536964679056485, + "grad_norm": 1.291521668434143, + "learning_rate": 1.890458613322255e-05, + "loss": 0.8311, + "step": 5316 + }, + { + "epoch": 0.23541392249537874, + "grad_norm": 1.4055473804473877, + "learning_rate": 1.8903882747440102e-05, + "loss": 0.9194, + "step": 5317 + }, + { + "epoch": 0.2354581982001926, + "grad_norm": 1.268355369567871, + "learning_rate": 1.890317914899566e-05, + "loss": 0.8135, + "step": 5318 + }, + { + "epoch": 0.23550247390500648, + "grad_norm": 1.2012999057769775, + "learning_rate": 1.8902475337906036e-05, + "loss": 0.8833, + "step": 5319 + }, + { + "epoch": 0.23554674960982036, + "grad_norm": 1.5031739473342896, + "learning_rate": 1.890177131418804e-05, + "loss": 0.984, + "step": 5320 + }, + { + "epoch": 0.23559102531463422, + "grad_norm": 1.7445427179336548, + "learning_rate": 1.890106707785848e-05, + "loss": 0.8928, + "step": 5321 + }, + { + "epoch": 0.2356353010194481, + "grad_norm": 1.3347842693328857, + "learning_rate": 1.890036262893418e-05, + "loss": 0.9521, + "step": 5322 + }, + { + "epoch": 0.23567957672426199, + "grad_norm": 1.1971577405929565, + "learning_rate": 1.889965796743197e-05, + "loss": 0.8797, + "step": 5323 + }, + { + "epoch": 0.23572385242907584, + "grad_norm": 1.3402583599090576, + "learning_rate": 1.8898953093368673e-05, + "loss": 0.8725, + "step": 5324 + }, + { + "epoch": 0.23576812813388973, + "grad_norm": 1.3166635036468506, + "learning_rate": 1.8898248006761128e-05, + "loss": 0.8393, + "step": 5325 + }, + { + "epoch": 0.2358124038387036, + "grad_norm": 1.9586000442504883, + "learning_rate": 1.8897542707626172e-05, + "loss": 0.9338, + "step": 5326 + }, + { + "epoch": 0.2358566795435175, + "grad_norm": 1.2208542823791504, + "learning_rate": 1.889683719598066e-05, + "loss": 0.9142, + "step": 5327 + }, + { + "epoch": 0.23590095524833135, + "grad_norm": 1.1788591146469116, + "learning_rate": 1.889613147184143e-05, + "loss": 0.8851, + "step": 5328 + }, + { + "epoch": 0.23594523095314524, + "grad_norm": 1.2132554054260254, + "learning_rate": 1.889542553522534e-05, + "loss": 0.8806, + "step": 5329 + }, + { + "epoch": 0.23598950665795912, + "grad_norm": 1.2438689470291138, + "learning_rate": 1.8894719386149262e-05, + "loss": 0.863, + "step": 5330 + }, + { + "epoch": 0.23603378236277298, + "grad_norm": 1.426612138748169, + "learning_rate": 1.889401302463005e-05, + "loss": 0.8836, + "step": 5331 + }, + { + "epoch": 0.23607805806758686, + "grad_norm": 1.2046877145767212, + "learning_rate": 1.8893306450684575e-05, + "loss": 0.8327, + "step": 5332 + }, + { + "epoch": 0.23612233377240074, + "grad_norm": 1.198697805404663, + "learning_rate": 1.8892599664329723e-05, + "loss": 0.847, + "step": 5333 + }, + { + "epoch": 0.23616660947721463, + "grad_norm": 2.1245791912078857, + "learning_rate": 1.8891892665582365e-05, + "loss": 1.0071, + "step": 5334 + }, + { + "epoch": 0.23621088518202848, + "grad_norm": 1.4008736610412598, + "learning_rate": 1.8891185454459392e-05, + "loss": 0.9744, + "step": 5335 + }, + { + "epoch": 0.23625516088684237, + "grad_norm": 1.8471083641052246, + "learning_rate": 1.8890478030977694e-05, + "loss": 0.9192, + "step": 5336 + }, + { + "epoch": 0.23629943659165625, + "grad_norm": 1.5379846096038818, + "learning_rate": 1.8889770395154166e-05, + "loss": 0.9719, + "step": 5337 + }, + { + "epoch": 0.2363437122964701, + "grad_norm": 1.6103289127349854, + "learning_rate": 1.8889062547005713e-05, + "loss": 0.9253, + "step": 5338 + }, + { + "epoch": 0.236387988001284, + "grad_norm": 1.3550955057144165, + "learning_rate": 1.8888354486549238e-05, + "loss": 0.9048, + "step": 5339 + }, + { + "epoch": 0.23643226370609788, + "grad_norm": 1.1834492683410645, + "learning_rate": 1.888764621380165e-05, + "loss": 0.8865, + "step": 5340 + }, + { + "epoch": 0.23647653941091173, + "grad_norm": 1.5271570682525635, + "learning_rate": 1.8886937728779872e-05, + "loss": 0.8481, + "step": 5341 + }, + { + "epoch": 0.23652081511572562, + "grad_norm": 1.3750687837600708, + "learning_rate": 1.888622903150082e-05, + "loss": 0.9071, + "step": 5342 + }, + { + "epoch": 0.2365650908205395, + "grad_norm": 1.2579668760299683, + "learning_rate": 1.8885520121981424e-05, + "loss": 0.9105, + "step": 5343 + }, + { + "epoch": 0.2366093665253534, + "grad_norm": 1.1791268587112427, + "learning_rate": 1.8884811000238615e-05, + "loss": 0.8637, + "step": 5344 + }, + { + "epoch": 0.23665364223016724, + "grad_norm": 1.3872711658477783, + "learning_rate": 1.888410166628933e-05, + "loss": 0.9226, + "step": 5345 + }, + { + "epoch": 0.23669791793498113, + "grad_norm": 1.1466455459594727, + "learning_rate": 1.8883392120150513e-05, + "loss": 0.881, + "step": 5346 + }, + { + "epoch": 0.236742193639795, + "grad_norm": 1.4609133005142212, + "learning_rate": 1.8882682361839104e-05, + "loss": 0.9019, + "step": 5347 + }, + { + "epoch": 0.23678646934460887, + "grad_norm": 1.296045184135437, + "learning_rate": 1.888197239137206e-05, + "loss": 0.9092, + "step": 5348 + }, + { + "epoch": 0.23683074504942275, + "grad_norm": 1.336520791053772, + "learning_rate": 1.8881262208766338e-05, + "loss": 0.9011, + "step": 5349 + }, + { + "epoch": 0.23687502075423664, + "grad_norm": 1.28643798828125, + "learning_rate": 1.88805518140389e-05, + "loss": 0.9196, + "step": 5350 + }, + { + "epoch": 0.23691929645905052, + "grad_norm": 1.3789966106414795, + "learning_rate": 1.8879841207206713e-05, + "loss": 0.9417, + "step": 5351 + }, + { + "epoch": 0.23696357216386438, + "grad_norm": 1.3259540796279907, + "learning_rate": 1.887913038828675e-05, + "loss": 0.8397, + "step": 5352 + }, + { + "epoch": 0.23700784786867826, + "grad_norm": 1.1184723377227783, + "learning_rate": 1.8878419357295984e-05, + "loss": 0.846, + "step": 5353 + }, + { + "epoch": 0.23705212357349215, + "grad_norm": 1.4770727157592773, + "learning_rate": 1.8877708114251405e-05, + "loss": 0.9252, + "step": 5354 + }, + { + "epoch": 0.237096399278306, + "grad_norm": 1.474499225616455, + "learning_rate": 1.887699665916999e-05, + "loss": 0.9198, + "step": 5355 + }, + { + "epoch": 0.2371406749831199, + "grad_norm": 1.3322035074234009, + "learning_rate": 1.8876284992068742e-05, + "loss": 0.9202, + "step": 5356 + }, + { + "epoch": 0.23718495068793377, + "grad_norm": 1.5840412378311157, + "learning_rate": 1.8875573112964653e-05, + "loss": 0.9896, + "step": 5357 + }, + { + "epoch": 0.23722922639274763, + "grad_norm": 1.694737195968628, + "learning_rate": 1.887486102187473e-05, + "loss": 0.9317, + "step": 5358 + }, + { + "epoch": 0.2372735020975615, + "grad_norm": 1.9489666223526, + "learning_rate": 1.8874148718815975e-05, + "loss": 0.9359, + "step": 5359 + }, + { + "epoch": 0.2373177778023754, + "grad_norm": 1.5512104034423828, + "learning_rate": 1.8873436203805404e-05, + "loss": 0.9083, + "step": 5360 + }, + { + "epoch": 0.23736205350718928, + "grad_norm": 1.6191130876541138, + "learning_rate": 1.8872723476860033e-05, + "loss": 0.9536, + "step": 5361 + }, + { + "epoch": 0.23740632921200314, + "grad_norm": 1.3986979722976685, + "learning_rate": 1.887201053799689e-05, + "loss": 0.8994, + "step": 5362 + }, + { + "epoch": 0.23745060491681702, + "grad_norm": 1.1958949565887451, + "learning_rate": 1.8871297387232993e-05, + "loss": 0.789, + "step": 5363 + }, + { + "epoch": 0.2374948806216309, + "grad_norm": 1.4798383712768555, + "learning_rate": 1.8870584024585387e-05, + "loss": 0.9119, + "step": 5364 + }, + { + "epoch": 0.23753915632644476, + "grad_norm": 1.3513518571853638, + "learning_rate": 1.88698704500711e-05, + "loss": 0.9062, + "step": 5365 + }, + { + "epoch": 0.23758343203125865, + "grad_norm": 1.4632681608200073, + "learning_rate": 1.8869156663707182e-05, + "loss": 0.9068, + "step": 5366 + }, + { + "epoch": 0.23762770773607253, + "grad_norm": 1.6099445819854736, + "learning_rate": 1.886844266551068e-05, + "loss": 1.0145, + "step": 5367 + }, + { + "epoch": 0.2376719834408864, + "grad_norm": 1.196574330329895, + "learning_rate": 1.886772845549864e-05, + "loss": 0.8512, + "step": 5368 + }, + { + "epoch": 0.23771625914570027, + "grad_norm": 1.3517439365386963, + "learning_rate": 1.8867014033688132e-05, + "loss": 0.8613, + "step": 5369 + }, + { + "epoch": 0.23776053485051415, + "grad_norm": 1.3034930229187012, + "learning_rate": 1.8866299400096212e-05, + "loss": 0.9929, + "step": 5370 + }, + { + "epoch": 0.23780481055532804, + "grad_norm": 1.382245421409607, + "learning_rate": 1.886558455473995e-05, + "loss": 0.8608, + "step": 5371 + }, + { + "epoch": 0.2378490862601419, + "grad_norm": 1.371338963508606, + "learning_rate": 1.886486949763642e-05, + "loss": 0.9254, + "step": 5372 + }, + { + "epoch": 0.23789336196495578, + "grad_norm": 1.2018704414367676, + "learning_rate": 1.8864154228802697e-05, + "loss": 0.8806, + "step": 5373 + }, + { + "epoch": 0.23793763766976966, + "grad_norm": 1.5048704147338867, + "learning_rate": 1.886343874825587e-05, + "loss": 0.9512, + "step": 5374 + }, + { + "epoch": 0.23798191337458352, + "grad_norm": 1.7345715761184692, + "learning_rate": 1.8862723056013026e-05, + "loss": 0.8969, + "step": 5375 + }, + { + "epoch": 0.2380261890793974, + "grad_norm": 1.2476418018341064, + "learning_rate": 1.8862007152091256e-05, + "loss": 0.875, + "step": 5376 + }, + { + "epoch": 0.2380704647842113, + "grad_norm": 1.2565762996673584, + "learning_rate": 1.8861291036507663e-05, + "loss": 0.8564, + "step": 5377 + }, + { + "epoch": 0.23811474048902517, + "grad_norm": 1.6230976581573486, + "learning_rate": 1.8860574709279348e-05, + "loss": 0.8956, + "step": 5378 + }, + { + "epoch": 0.23815901619383903, + "grad_norm": 1.4828884601593018, + "learning_rate": 1.8859858170423423e-05, + "loss": 0.9152, + "step": 5379 + }, + { + "epoch": 0.2382032918986529, + "grad_norm": 1.2547988891601562, + "learning_rate": 1.8859141419956993e-05, + "loss": 0.8395, + "step": 5380 + }, + { + "epoch": 0.2382475676034668, + "grad_norm": 1.5029754638671875, + "learning_rate": 1.885842445789719e-05, + "loss": 0.9348, + "step": 5381 + }, + { + "epoch": 0.23829184330828065, + "grad_norm": 1.2427215576171875, + "learning_rate": 1.885770728426113e-05, + "loss": 0.8616, + "step": 5382 + }, + { + "epoch": 0.23833611901309454, + "grad_norm": 1.3745825290679932, + "learning_rate": 1.8856989899065943e-05, + "loss": 0.8681, + "step": 5383 + }, + { + "epoch": 0.23838039471790842, + "grad_norm": 1.4175262451171875, + "learning_rate": 1.8856272302328764e-05, + "loss": 0.9461, + "step": 5384 + }, + { + "epoch": 0.23842467042272228, + "grad_norm": 1.2170445919036865, + "learning_rate": 1.8855554494066732e-05, + "loss": 0.881, + "step": 5385 + }, + { + "epoch": 0.23846894612753616, + "grad_norm": 1.384169101715088, + "learning_rate": 1.8854836474296993e-05, + "loss": 0.9604, + "step": 5386 + }, + { + "epoch": 0.23851322183235005, + "grad_norm": 1.4675403833389282, + "learning_rate": 1.8854118243036692e-05, + "loss": 0.894, + "step": 5387 + }, + { + "epoch": 0.23855749753716393, + "grad_norm": 1.3969709873199463, + "learning_rate": 1.885339980030299e-05, + "loss": 0.8765, + "step": 5388 + }, + { + "epoch": 0.2386017732419778, + "grad_norm": 1.298197627067566, + "learning_rate": 1.8852681146113036e-05, + "loss": 0.8012, + "step": 5389 + }, + { + "epoch": 0.23864604894679167, + "grad_norm": 1.2201826572418213, + "learning_rate": 1.8851962280484008e-05, + "loss": 0.8162, + "step": 5390 + }, + { + "epoch": 0.23869032465160556, + "grad_norm": 1.1641645431518555, + "learning_rate": 1.8851243203433063e-05, + "loss": 0.8556, + "step": 5391 + }, + { + "epoch": 0.2387346003564194, + "grad_norm": 1.3203614950180054, + "learning_rate": 1.8850523914977385e-05, + "loss": 0.8755, + "step": 5392 + }, + { + "epoch": 0.2387788760612333, + "grad_norm": 1.5572172403335571, + "learning_rate": 1.8849804415134145e-05, + "loss": 0.9085, + "step": 5393 + }, + { + "epoch": 0.23882315176604718, + "grad_norm": 1.0766026973724365, + "learning_rate": 1.8849084703920536e-05, + "loss": 0.8129, + "step": 5394 + }, + { + "epoch": 0.23886742747086107, + "grad_norm": 1.330385684967041, + "learning_rate": 1.8848364781353744e-05, + "loss": 0.8563, + "step": 5395 + }, + { + "epoch": 0.23891170317567492, + "grad_norm": 1.0530914068222046, + "learning_rate": 1.884764464745096e-05, + "loss": 0.8201, + "step": 5396 + }, + { + "epoch": 0.2389559788804888, + "grad_norm": 1.4869024753570557, + "learning_rate": 1.8846924302229395e-05, + "loss": 0.9729, + "step": 5397 + }, + { + "epoch": 0.2390002545853027, + "grad_norm": 1.2796354293823242, + "learning_rate": 1.884620374570624e-05, + "loss": 0.9417, + "step": 5398 + }, + { + "epoch": 0.23904453029011655, + "grad_norm": 1.3732060194015503, + "learning_rate": 1.8845482977898713e-05, + "loss": 0.8794, + "step": 5399 + }, + { + "epoch": 0.23908880599493043, + "grad_norm": 1.487134575843811, + "learning_rate": 1.884476199882403e-05, + "loss": 0.98, + "step": 5400 + }, + { + "epoch": 0.23913308169974432, + "grad_norm": 1.39290452003479, + "learning_rate": 1.8844040808499407e-05, + "loss": 0.9759, + "step": 5401 + }, + { + "epoch": 0.23917735740455817, + "grad_norm": 1.2077431678771973, + "learning_rate": 1.8843319406942067e-05, + "loss": 0.8733, + "step": 5402 + }, + { + "epoch": 0.23922163310937206, + "grad_norm": 1.3098149299621582, + "learning_rate": 1.8842597794169245e-05, + "loss": 0.8759, + "step": 5403 + }, + { + "epoch": 0.23926590881418594, + "grad_norm": 1.3315995931625366, + "learning_rate": 1.8841875970198177e-05, + "loss": 0.9165, + "step": 5404 + }, + { + "epoch": 0.23931018451899982, + "grad_norm": 1.347147822380066, + "learning_rate": 1.8841153935046098e-05, + "loss": 0.8733, + "step": 5405 + }, + { + "epoch": 0.23935446022381368, + "grad_norm": 1.606221318244934, + "learning_rate": 1.884043168873026e-05, + "loss": 0.9023, + "step": 5406 + }, + { + "epoch": 0.23939873592862757, + "grad_norm": 1.281843900680542, + "learning_rate": 1.8839709231267906e-05, + "loss": 0.8394, + "step": 5407 + }, + { + "epoch": 0.23944301163344145, + "grad_norm": 1.6235840320587158, + "learning_rate": 1.8838986562676295e-05, + "loss": 0.9287, + "step": 5408 + }, + { + "epoch": 0.2394872873382553, + "grad_norm": 1.2459466457366943, + "learning_rate": 1.883826368297269e-05, + "loss": 0.8582, + "step": 5409 + }, + { + "epoch": 0.2395315630430692, + "grad_norm": 1.376512050628662, + "learning_rate": 1.8837540592174347e-05, + "loss": 0.9266, + "step": 5410 + }, + { + "epoch": 0.23957583874788307, + "grad_norm": 1.373671054840088, + "learning_rate": 1.8836817290298546e-05, + "loss": 0.8897, + "step": 5411 + }, + { + "epoch": 0.23962011445269693, + "grad_norm": 1.2579745054244995, + "learning_rate": 1.883609377736256e-05, + "loss": 0.8955, + "step": 5412 + }, + { + "epoch": 0.23966439015751081, + "grad_norm": 1.4865708351135254, + "learning_rate": 1.883537005338367e-05, + "loss": 0.8651, + "step": 5413 + }, + { + "epoch": 0.2397086658623247, + "grad_norm": 1.676271677017212, + "learning_rate": 1.8834646118379156e-05, + "loss": 0.8829, + "step": 5414 + }, + { + "epoch": 0.23975294156713858, + "grad_norm": 1.371504783630371, + "learning_rate": 1.8833921972366315e-05, + "loss": 0.9038, + "step": 5415 + }, + { + "epoch": 0.23979721727195244, + "grad_norm": 1.598697304725647, + "learning_rate": 1.8833197615362443e-05, + "loss": 0.9141, + "step": 5416 + }, + { + "epoch": 0.23984149297676632, + "grad_norm": 1.2727214097976685, + "learning_rate": 1.8832473047384835e-05, + "loss": 0.8593, + "step": 5417 + }, + { + "epoch": 0.2398857686815802, + "grad_norm": 1.1885191202163696, + "learning_rate": 1.8831748268450802e-05, + "loss": 0.8651, + "step": 5418 + }, + { + "epoch": 0.23993004438639406, + "grad_norm": 1.2731068134307861, + "learning_rate": 1.883102327857765e-05, + "loss": 0.8393, + "step": 5419 + }, + { + "epoch": 0.23997432009120795, + "grad_norm": 1.4829720258712769, + "learning_rate": 1.88302980777827e-05, + "loss": 0.9179, + "step": 5420 + }, + { + "epoch": 0.24001859579602183, + "grad_norm": 1.2425363063812256, + "learning_rate": 1.8829572666083267e-05, + "loss": 0.885, + "step": 5421 + }, + { + "epoch": 0.24006287150083572, + "grad_norm": 1.3910021781921387, + "learning_rate": 1.8828847043496684e-05, + "loss": 0.8837, + "step": 5422 + }, + { + "epoch": 0.24010714720564957, + "grad_norm": 1.4874260425567627, + "learning_rate": 1.882812121004028e-05, + "loss": 0.9508, + "step": 5423 + }, + { + "epoch": 0.24015142291046346, + "grad_norm": 1.158772349357605, + "learning_rate": 1.8827395165731383e-05, + "loss": 0.8289, + "step": 5424 + }, + { + "epoch": 0.24019569861527734, + "grad_norm": 1.4298152923583984, + "learning_rate": 1.8826668910587343e-05, + "loss": 0.9198, + "step": 5425 + }, + { + "epoch": 0.2402399743200912, + "grad_norm": 1.408066987991333, + "learning_rate": 1.8825942444625504e-05, + "loss": 0.9334, + "step": 5426 + }, + { + "epoch": 0.24028425002490508, + "grad_norm": 1.5478103160858154, + "learning_rate": 1.8825215767863215e-05, + "loss": 0.9313, + "step": 5427 + }, + { + "epoch": 0.24032852572971897, + "grad_norm": 1.4319044351577759, + "learning_rate": 1.882448888031783e-05, + "loss": 0.9335, + "step": 5428 + }, + { + "epoch": 0.24037280143453282, + "grad_norm": 1.7025667428970337, + "learning_rate": 1.882376178200672e-05, + "loss": 0.8857, + "step": 5429 + }, + { + "epoch": 0.2404170771393467, + "grad_norm": 1.1868433952331543, + "learning_rate": 1.882303447294724e-05, + "loss": 0.8432, + "step": 5430 + }, + { + "epoch": 0.2404613528441606, + "grad_norm": 1.3331570625305176, + "learning_rate": 1.8822306953156767e-05, + "loss": 0.8522, + "step": 5431 + }, + { + "epoch": 0.24050562854897448, + "grad_norm": 1.0532822608947754, + "learning_rate": 1.882157922265267e-05, + "loss": 0.7987, + "step": 5432 + }, + { + "epoch": 0.24054990425378833, + "grad_norm": 1.4016190767288208, + "learning_rate": 1.8820851281452346e-05, + "loss": 0.8917, + "step": 5433 + }, + { + "epoch": 0.24059417995860222, + "grad_norm": 1.3950377702713013, + "learning_rate": 1.8820123129573163e-05, + "loss": 0.9286, + "step": 5434 + }, + { + "epoch": 0.2406384556634161, + "grad_norm": 1.42722487449646, + "learning_rate": 1.8819394767032527e-05, + "loss": 0.9255, + "step": 5435 + }, + { + "epoch": 0.24068273136822996, + "grad_norm": 1.3498705625534058, + "learning_rate": 1.8818666193847824e-05, + "loss": 0.9066, + "step": 5436 + }, + { + "epoch": 0.24072700707304384, + "grad_norm": 1.5075592994689941, + "learning_rate": 1.8817937410036462e-05, + "loss": 0.8528, + "step": 5437 + }, + { + "epoch": 0.24077128277785773, + "grad_norm": 1.4671014547348022, + "learning_rate": 1.8817208415615844e-05, + "loss": 0.8955, + "step": 5438 + }, + { + "epoch": 0.24081555848267158, + "grad_norm": 1.4903132915496826, + "learning_rate": 1.881647921060338e-05, + "loss": 0.9828, + "step": 5439 + }, + { + "epoch": 0.24085983418748547, + "grad_norm": 1.3224470615386963, + "learning_rate": 1.881574979501649e-05, + "loss": 0.9118, + "step": 5440 + }, + { + "epoch": 0.24090410989229935, + "grad_norm": 1.4487454891204834, + "learning_rate": 1.8815020168872596e-05, + "loss": 0.9377, + "step": 5441 + }, + { + "epoch": 0.24094838559711323, + "grad_norm": 1.3132247924804688, + "learning_rate": 1.881429033218912e-05, + "loss": 0.9581, + "step": 5442 + }, + { + "epoch": 0.2409926613019271, + "grad_norm": 1.1937912702560425, + "learning_rate": 1.88135602849835e-05, + "loss": 0.8042, + "step": 5443 + }, + { + "epoch": 0.24103693700674098, + "grad_norm": 1.2696601152420044, + "learning_rate": 1.8812830027273162e-05, + "loss": 0.9126, + "step": 5444 + }, + { + "epoch": 0.24108121271155486, + "grad_norm": 1.3609426021575928, + "learning_rate": 1.8812099559075562e-05, + "loss": 0.8886, + "step": 5445 + }, + { + "epoch": 0.24112548841636872, + "grad_norm": 1.7033510208129883, + "learning_rate": 1.8811368880408132e-05, + "loss": 0.9499, + "step": 5446 + }, + { + "epoch": 0.2411697641211826, + "grad_norm": 1.8397588729858398, + "learning_rate": 1.8810637991288337e-05, + "loss": 1.0142, + "step": 5447 + }, + { + "epoch": 0.24121403982599648, + "grad_norm": 1.3040134906768799, + "learning_rate": 1.8809906891733624e-05, + "loss": 0.8129, + "step": 5448 + }, + { + "epoch": 0.24125831553081037, + "grad_norm": 1.1768141984939575, + "learning_rate": 1.880917558176146e-05, + "loss": 0.8599, + "step": 5449 + }, + { + "epoch": 0.24130259123562423, + "grad_norm": 1.5717641115188599, + "learning_rate": 1.8808444061389307e-05, + "loss": 0.8922, + "step": 5450 + }, + { + "epoch": 0.2413468669404381, + "grad_norm": 1.4029009342193604, + "learning_rate": 1.8807712330634645e-05, + "loss": 0.9785, + "step": 5451 + }, + { + "epoch": 0.241391142645252, + "grad_norm": 1.3506639003753662, + "learning_rate": 1.880698038951494e-05, + "loss": 0.8484, + "step": 5452 + }, + { + "epoch": 0.24143541835006585, + "grad_norm": 1.4245901107788086, + "learning_rate": 1.8806248238047684e-05, + "loss": 0.8265, + "step": 5453 + }, + { + "epoch": 0.24147969405487973, + "grad_norm": 1.5294753313064575, + "learning_rate": 1.8805515876250355e-05, + "loss": 0.9875, + "step": 5454 + }, + { + "epoch": 0.24152396975969362, + "grad_norm": 1.2345746755599976, + "learning_rate": 1.880478330414045e-05, + "loss": 0.8556, + "step": 5455 + }, + { + "epoch": 0.24156824546450748, + "grad_norm": 1.118394136428833, + "learning_rate": 1.8804050521735468e-05, + "loss": 0.8218, + "step": 5456 + }, + { + "epoch": 0.24161252116932136, + "grad_norm": 1.3183356523513794, + "learning_rate": 1.8803317529052906e-05, + "loss": 0.8712, + "step": 5457 + }, + { + "epoch": 0.24165679687413524, + "grad_norm": 1.295123815536499, + "learning_rate": 1.880258432611027e-05, + "loss": 0.9183, + "step": 5458 + }, + { + "epoch": 0.24170107257894913, + "grad_norm": 1.2996774911880493, + "learning_rate": 1.880185091292508e-05, + "loss": 0.8532, + "step": 5459 + }, + { + "epoch": 0.24174534828376298, + "grad_norm": 1.2602605819702148, + "learning_rate": 1.8801117289514847e-05, + "loss": 0.9128, + "step": 5460 + }, + { + "epoch": 0.24178962398857687, + "grad_norm": 1.2616904973983765, + "learning_rate": 1.880038345589709e-05, + "loss": 0.8557, + "step": 5461 + }, + { + "epoch": 0.24183389969339075, + "grad_norm": 1.1662648916244507, + "learning_rate": 1.8799649412089338e-05, + "loss": 0.8064, + "step": 5462 + }, + { + "epoch": 0.2418781753982046, + "grad_norm": 1.2450040578842163, + "learning_rate": 1.879891515810913e-05, + "loss": 0.8554, + "step": 5463 + }, + { + "epoch": 0.2419224511030185, + "grad_norm": 3.703963279724121, + "learning_rate": 1.8798180693973996e-05, + "loss": 0.8969, + "step": 5464 + }, + { + "epoch": 0.24196672680783238, + "grad_norm": 1.590989112854004, + "learning_rate": 1.8797446019701478e-05, + "loss": 0.9181, + "step": 5465 + }, + { + "epoch": 0.24201100251264626, + "grad_norm": 1.221187949180603, + "learning_rate": 1.8796711135309128e-05, + "loss": 0.8992, + "step": 5466 + }, + { + "epoch": 0.24205527821746012, + "grad_norm": 1.34467351436615, + "learning_rate": 1.879597604081449e-05, + "loss": 0.8624, + "step": 5467 + }, + { + "epoch": 0.242099553922274, + "grad_norm": 1.604002833366394, + "learning_rate": 1.879524073623513e-05, + "loss": 0.9056, + "step": 5468 + }, + { + "epoch": 0.2421438296270879, + "grad_norm": 1.2776485681533813, + "learning_rate": 1.8794505221588608e-05, + "loss": 0.8599, + "step": 5469 + }, + { + "epoch": 0.24218810533190174, + "grad_norm": 1.3394274711608887, + "learning_rate": 1.879376949689249e-05, + "loss": 0.8578, + "step": 5470 + }, + { + "epoch": 0.24223238103671563, + "grad_norm": 1.5644880533218384, + "learning_rate": 1.8793033562164344e-05, + "loss": 1.0302, + "step": 5471 + }, + { + "epoch": 0.2422766567415295, + "grad_norm": 1.254629135131836, + "learning_rate": 1.879229741742175e-05, + "loss": 0.8709, + "step": 5472 + }, + { + "epoch": 0.24232093244634337, + "grad_norm": 1.2931427955627441, + "learning_rate": 1.8791561062682296e-05, + "loss": 0.9332, + "step": 5473 + }, + { + "epoch": 0.24236520815115725, + "grad_norm": 1.2772544622421265, + "learning_rate": 1.8790824497963563e-05, + "loss": 0.8867, + "step": 5474 + }, + { + "epoch": 0.24240948385597114, + "grad_norm": 3.5095055103302, + "learning_rate": 1.879008772328314e-05, + "loss": 0.9627, + "step": 5475 + }, + { + "epoch": 0.24245375956078502, + "grad_norm": 1.498239278793335, + "learning_rate": 1.8789350738658632e-05, + "loss": 0.9218, + "step": 5476 + }, + { + "epoch": 0.24249803526559888, + "grad_norm": 1.1294090747833252, + "learning_rate": 1.878861354410764e-05, + "loss": 0.7899, + "step": 5477 + }, + { + "epoch": 0.24254231097041276, + "grad_norm": 1.5670725107192993, + "learning_rate": 1.878787613964777e-05, + "loss": 0.9293, + "step": 5478 + }, + { + "epoch": 0.24258658667522665, + "grad_norm": 2.0981554985046387, + "learning_rate": 1.878713852529663e-05, + "loss": 0.8981, + "step": 5479 + }, + { + "epoch": 0.2426308623800405, + "grad_norm": 1.5367798805236816, + "learning_rate": 1.878640070107184e-05, + "loss": 0.8858, + "step": 5480 + }, + { + "epoch": 0.24267513808485439, + "grad_norm": 1.3428047895431519, + "learning_rate": 1.878566266699103e-05, + "loss": 0.8632, + "step": 5481 + }, + { + "epoch": 0.24271941378966827, + "grad_norm": 1.1907494068145752, + "learning_rate": 1.8784924423071815e-05, + "loss": 0.8825, + "step": 5482 + }, + { + "epoch": 0.24276368949448213, + "grad_norm": 1.5268021821975708, + "learning_rate": 1.8784185969331838e-05, + "loss": 0.9274, + "step": 5483 + }, + { + "epoch": 0.242807965199296, + "grad_norm": 1.2936689853668213, + "learning_rate": 1.8783447305788725e-05, + "loss": 0.9527, + "step": 5484 + }, + { + "epoch": 0.2428522409041099, + "grad_norm": 1.1859891414642334, + "learning_rate": 1.878270843246013e-05, + "loss": 0.8879, + "step": 5485 + }, + { + "epoch": 0.24289651660892378, + "grad_norm": 1.539891004562378, + "learning_rate": 1.8781969349363692e-05, + "loss": 0.9213, + "step": 5486 + }, + { + "epoch": 0.24294079231373764, + "grad_norm": 1.266261339187622, + "learning_rate": 1.878123005651707e-05, + "loss": 0.8608, + "step": 5487 + }, + { + "epoch": 0.24298506801855152, + "grad_norm": 1.5865510702133179, + "learning_rate": 1.8780490553937914e-05, + "loss": 0.9399, + "step": 5488 + }, + { + "epoch": 0.2430293437233654, + "grad_norm": 1.484094262123108, + "learning_rate": 1.8779750841643896e-05, + "loss": 0.9974, + "step": 5489 + }, + { + "epoch": 0.24307361942817926, + "grad_norm": 1.4194488525390625, + "learning_rate": 1.877901091965267e-05, + "loss": 0.9864, + "step": 5490 + }, + { + "epoch": 0.24311789513299314, + "grad_norm": 1.073968768119812, + "learning_rate": 1.8778270787981924e-05, + "loss": 0.8292, + "step": 5491 + }, + { + "epoch": 0.24316217083780703, + "grad_norm": 1.3187583684921265, + "learning_rate": 1.8777530446649326e-05, + "loss": 0.8979, + "step": 5492 + }, + { + "epoch": 0.2432064465426209, + "grad_norm": 1.4318989515304565, + "learning_rate": 1.8776789895672557e-05, + "loss": 0.9137, + "step": 5493 + }, + { + "epoch": 0.24325072224743477, + "grad_norm": 1.4915639162063599, + "learning_rate": 1.877604913506931e-05, + "loss": 0.8934, + "step": 5494 + }, + { + "epoch": 0.24329499795224865, + "grad_norm": 1.285317301750183, + "learning_rate": 1.8775308164857275e-05, + "loss": 0.9406, + "step": 5495 + }, + { + "epoch": 0.24333927365706254, + "grad_norm": 1.5945188999176025, + "learning_rate": 1.877456698505415e-05, + "loss": 0.9828, + "step": 5496 + }, + { + "epoch": 0.2433835493618764, + "grad_norm": 1.2713721990585327, + "learning_rate": 1.8773825595677636e-05, + "loss": 0.8613, + "step": 5497 + }, + { + "epoch": 0.24342782506669028, + "grad_norm": 1.328177571296692, + "learning_rate": 1.8773083996745443e-05, + "loss": 0.8976, + "step": 5498 + }, + { + "epoch": 0.24347210077150416, + "grad_norm": 1.31527841091156, + "learning_rate": 1.877234218827528e-05, + "loss": 0.9534, + "step": 5499 + }, + { + "epoch": 0.24351637647631802, + "grad_norm": 1.154832124710083, + "learning_rate": 1.8771600170284866e-05, + "loss": 0.846, + "step": 5500 + }, + { + "epoch": 0.2435606521811319, + "grad_norm": 1.2307281494140625, + "learning_rate": 1.8770857942791925e-05, + "loss": 0.8524, + "step": 5501 + }, + { + "epoch": 0.2436049278859458, + "grad_norm": 1.1245557069778442, + "learning_rate": 1.8770115505814187e-05, + "loss": 0.8492, + "step": 5502 + }, + { + "epoch": 0.24364920359075967, + "grad_norm": 1.358476996421814, + "learning_rate": 1.8769372859369376e-05, + "loss": 0.9518, + "step": 5503 + }, + { + "epoch": 0.24369347929557353, + "grad_norm": 1.2299115657806396, + "learning_rate": 1.8768630003475238e-05, + "loss": 0.8648, + "step": 5504 + }, + { + "epoch": 0.2437377550003874, + "grad_norm": 2.0737102031707764, + "learning_rate": 1.876788693814951e-05, + "loss": 1.0319, + "step": 5505 + }, + { + "epoch": 0.2437820307052013, + "grad_norm": 1.4181534051895142, + "learning_rate": 1.8767143663409943e-05, + "loss": 1.0195, + "step": 5506 + }, + { + "epoch": 0.24382630641001515, + "grad_norm": 1.1955653429031372, + "learning_rate": 1.8766400179274287e-05, + "loss": 0.8348, + "step": 5507 + }, + { + "epoch": 0.24387058211482904, + "grad_norm": 1.5717049837112427, + "learning_rate": 1.8765656485760302e-05, + "loss": 0.9701, + "step": 5508 + }, + { + "epoch": 0.24391485781964292, + "grad_norm": 1.6039457321166992, + "learning_rate": 1.876491258288575e-05, + "loss": 0.9394, + "step": 5509 + }, + { + "epoch": 0.2439591335244568, + "grad_norm": 1.2653124332427979, + "learning_rate": 1.8764168470668394e-05, + "loss": 0.8943, + "step": 5510 + }, + { + "epoch": 0.24400340922927066, + "grad_norm": 1.2664978504180908, + "learning_rate": 1.876342414912601e-05, + "loss": 0.882, + "step": 5511 + }, + { + "epoch": 0.24404768493408455, + "grad_norm": 1.4475153684616089, + "learning_rate": 1.8762679618276377e-05, + "loss": 0.9169, + "step": 5512 + }, + { + "epoch": 0.24409196063889843, + "grad_norm": 1.375946044921875, + "learning_rate": 1.8761934878137278e-05, + "loss": 0.9564, + "step": 5513 + }, + { + "epoch": 0.2441362363437123, + "grad_norm": 1.134019374847412, + "learning_rate": 1.87611899287265e-05, + "loss": 0.7685, + "step": 5514 + }, + { + "epoch": 0.24418051204852617, + "grad_norm": 1.3772790431976318, + "learning_rate": 1.876044477006183e-05, + "loss": 0.8161, + "step": 5515 + }, + { + "epoch": 0.24422478775334006, + "grad_norm": 1.368388056755066, + "learning_rate": 1.875969940216107e-05, + "loss": 0.9562, + "step": 5516 + }, + { + "epoch": 0.2442690634581539, + "grad_norm": 1.2103474140167236, + "learning_rate": 1.8758953825042023e-05, + "loss": 0.8899, + "step": 5517 + }, + { + "epoch": 0.2443133391629678, + "grad_norm": 1.5161137580871582, + "learning_rate": 1.8758208038722495e-05, + "loss": 0.9513, + "step": 5518 + }, + { + "epoch": 0.24435761486778168, + "grad_norm": 1.5566890239715576, + "learning_rate": 1.87574620432203e-05, + "loss": 0.9773, + "step": 5519 + }, + { + "epoch": 0.24440189057259556, + "grad_norm": 1.2510409355163574, + "learning_rate": 1.8756715838553257e-05, + "loss": 0.8137, + "step": 5520 + }, + { + "epoch": 0.24444616627740942, + "grad_norm": 1.510562539100647, + "learning_rate": 1.8755969424739186e-05, + "loss": 0.8673, + "step": 5521 + }, + { + "epoch": 0.2444904419822233, + "grad_norm": 1.4746830463409424, + "learning_rate": 1.875522280179591e-05, + "loss": 0.9145, + "step": 5522 + }, + { + "epoch": 0.2445347176870372, + "grad_norm": 1.2879024744033813, + "learning_rate": 1.8754475969741272e-05, + "loss": 0.8635, + "step": 5523 + }, + { + "epoch": 0.24457899339185105, + "grad_norm": 1.4386500120162964, + "learning_rate": 1.87537289285931e-05, + "loss": 0.9104, + "step": 5524 + }, + { + "epoch": 0.24462326909666493, + "grad_norm": 1.160139799118042, + "learning_rate": 1.875298167836924e-05, + "loss": 0.8185, + "step": 5525 + }, + { + "epoch": 0.24466754480147881, + "grad_norm": 1.3937889337539673, + "learning_rate": 1.8752234219087538e-05, + "loss": 0.9451, + "step": 5526 + }, + { + "epoch": 0.24471182050629267, + "grad_norm": 1.61875319480896, + "learning_rate": 1.875148655076585e-05, + "loss": 0.9572, + "step": 5527 + }, + { + "epoch": 0.24475609621110656, + "grad_norm": 1.6648863554000854, + "learning_rate": 1.8750738673422032e-05, + "loss": 0.9264, + "step": 5528 + }, + { + "epoch": 0.24480037191592044, + "grad_norm": 1.1195507049560547, + "learning_rate": 1.8749990587073946e-05, + "loss": 0.812, + "step": 5529 + }, + { + "epoch": 0.24484464762073432, + "grad_norm": 1.318750023841858, + "learning_rate": 1.8749242291739457e-05, + "loss": 0.8936, + "step": 5530 + }, + { + "epoch": 0.24488892332554818, + "grad_norm": 1.23310124874115, + "learning_rate": 1.8748493787436444e-05, + "loss": 0.8308, + "step": 5531 + }, + { + "epoch": 0.24493319903036206, + "grad_norm": 1.4644505977630615, + "learning_rate": 1.8747745074182776e-05, + "loss": 0.9127, + "step": 5532 + }, + { + "epoch": 0.24497747473517595, + "grad_norm": 1.0928723812103271, + "learning_rate": 1.8746996151996343e-05, + "loss": 0.7369, + "step": 5533 + }, + { + "epoch": 0.2450217504399898, + "grad_norm": 1.1252658367156982, + "learning_rate": 1.8746247020895028e-05, + "loss": 0.8822, + "step": 5534 + }, + { + "epoch": 0.2450660261448037, + "grad_norm": 1.222531795501709, + "learning_rate": 1.8745497680896722e-05, + "loss": 0.8589, + "step": 5535 + }, + { + "epoch": 0.24511030184961757, + "grad_norm": 1.2129267454147339, + "learning_rate": 1.874474813201933e-05, + "loss": 0.876, + "step": 5536 + }, + { + "epoch": 0.24515457755443146, + "grad_norm": 1.3622854948043823, + "learning_rate": 1.8743998374280742e-05, + "loss": 0.948, + "step": 5537 + }, + { + "epoch": 0.24519885325924531, + "grad_norm": 1.2708170413970947, + "learning_rate": 1.8743248407698877e-05, + "loss": 0.9128, + "step": 5538 + }, + { + "epoch": 0.2452431289640592, + "grad_norm": 1.4730424880981445, + "learning_rate": 1.874249823229164e-05, + "loss": 0.9485, + "step": 5539 + }, + { + "epoch": 0.24528740466887308, + "grad_norm": 1.2979480028152466, + "learning_rate": 1.8741747848076955e-05, + "loss": 0.8474, + "step": 5540 + }, + { + "epoch": 0.24533168037368694, + "grad_norm": 1.5388669967651367, + "learning_rate": 1.8740997255072738e-05, + "loss": 0.8742, + "step": 5541 + }, + { + "epoch": 0.24537595607850082, + "grad_norm": 2.414715051651001, + "learning_rate": 1.874024645329692e-05, + "loss": 1.0375, + "step": 5542 + }, + { + "epoch": 0.2454202317833147, + "grad_norm": 1.414658784866333, + "learning_rate": 1.873949544276743e-05, + "loss": 0.848, + "step": 5543 + }, + { + "epoch": 0.24546450748812856, + "grad_norm": 1.178588628768921, + "learning_rate": 1.873874422350221e-05, + "loss": 0.8727, + "step": 5544 + }, + { + "epoch": 0.24550878319294245, + "grad_norm": 1.2581353187561035, + "learning_rate": 1.87379927955192e-05, + "loss": 0.8191, + "step": 5545 + }, + { + "epoch": 0.24555305889775633, + "grad_norm": 1.526882529258728, + "learning_rate": 1.873724115883634e-05, + "loss": 0.8366, + "step": 5546 + }, + { + "epoch": 0.24559733460257022, + "grad_norm": 1.4916722774505615, + "learning_rate": 1.87364893134716e-05, + "loss": 0.8516, + "step": 5547 + }, + { + "epoch": 0.24564161030738407, + "grad_norm": 1.2978159189224243, + "learning_rate": 1.8735737259442918e-05, + "loss": 0.8299, + "step": 5548 + }, + { + "epoch": 0.24568588601219796, + "grad_norm": 1.2256486415863037, + "learning_rate": 1.8734984996768263e-05, + "loss": 0.7451, + "step": 5549 + }, + { + "epoch": 0.24573016171701184, + "grad_norm": 1.3795967102050781, + "learning_rate": 1.873423252546561e-05, + "loss": 0.8504, + "step": 5550 + }, + { + "epoch": 0.2457744374218257, + "grad_norm": 1.465959906578064, + "learning_rate": 1.873347984555292e-05, + "loss": 0.9694, + "step": 5551 + }, + { + "epoch": 0.24581871312663958, + "grad_norm": 1.3098052740097046, + "learning_rate": 1.873272695704818e-05, + "loss": 0.9279, + "step": 5552 + }, + { + "epoch": 0.24586298883145347, + "grad_norm": 1.6010338068008423, + "learning_rate": 1.8731973859969362e-05, + "loss": 0.9045, + "step": 5553 + }, + { + "epoch": 0.24590726453626735, + "grad_norm": 1.6311736106872559, + "learning_rate": 1.873122055433446e-05, + "loss": 1.0065, + "step": 5554 + }, + { + "epoch": 0.2459515402410812, + "grad_norm": 1.4032611846923828, + "learning_rate": 1.8730467040161463e-05, + "loss": 0.8989, + "step": 5555 + }, + { + "epoch": 0.2459958159458951, + "grad_norm": 1.7431726455688477, + "learning_rate": 1.8729713317468376e-05, + "loss": 0.978, + "step": 5556 + }, + { + "epoch": 0.24604009165070898, + "grad_norm": 1.2744030952453613, + "learning_rate": 1.8728959386273188e-05, + "loss": 0.8235, + "step": 5557 + }, + { + "epoch": 0.24608436735552283, + "grad_norm": 1.2518837451934814, + "learning_rate": 1.8728205246593912e-05, + "loss": 0.8068, + "step": 5558 + }, + { + "epoch": 0.24612864306033672, + "grad_norm": 1.6524147987365723, + "learning_rate": 1.8727450898448562e-05, + "loss": 0.903, + "step": 5559 + }, + { + "epoch": 0.2461729187651506, + "grad_norm": 1.5566856861114502, + "learning_rate": 1.8726696341855156e-05, + "loss": 0.9063, + "step": 5560 + }, + { + "epoch": 0.24621719446996446, + "grad_norm": 1.3770010471343994, + "learning_rate": 1.872594157683171e-05, + "loss": 0.9318, + "step": 5561 + }, + { + "epoch": 0.24626147017477834, + "grad_norm": 1.2830363512039185, + "learning_rate": 1.8725186603396256e-05, + "loss": 0.8803, + "step": 5562 + }, + { + "epoch": 0.24630574587959222, + "grad_norm": 1.2841920852661133, + "learning_rate": 1.8724431421566822e-05, + "loss": 0.9125, + "step": 5563 + }, + { + "epoch": 0.2463500215844061, + "grad_norm": 1.3535367250442505, + "learning_rate": 1.8723676031361452e-05, + "loss": 0.868, + "step": 5564 + }, + { + "epoch": 0.24639429728921997, + "grad_norm": 1.19920814037323, + "learning_rate": 1.872292043279818e-05, + "loss": 0.8328, + "step": 5565 + }, + { + "epoch": 0.24643857299403385, + "grad_norm": 1.241165041923523, + "learning_rate": 1.8722164625895054e-05, + "loss": 0.8894, + "step": 5566 + }, + { + "epoch": 0.24648284869884773, + "grad_norm": 1.3281179666519165, + "learning_rate": 1.872140861067013e-05, + "loss": 0.8919, + "step": 5567 + }, + { + "epoch": 0.2465271244036616, + "grad_norm": 1.2432259321212769, + "learning_rate": 1.872065238714146e-05, + "loss": 0.7831, + "step": 5568 + }, + { + "epoch": 0.24657140010847547, + "grad_norm": 1.4220435619354248, + "learning_rate": 1.8719895955327116e-05, + "loss": 0.828, + "step": 5569 + }, + { + "epoch": 0.24661567581328936, + "grad_norm": 1.4338685274124146, + "learning_rate": 1.8719139315245146e-05, + "loss": 0.8531, + "step": 5570 + }, + { + "epoch": 0.24665995151810322, + "grad_norm": 1.400243878364563, + "learning_rate": 1.8718382466913642e-05, + "loss": 0.8989, + "step": 5571 + }, + { + "epoch": 0.2467042272229171, + "grad_norm": 1.5053541660308838, + "learning_rate": 1.8717625410350665e-05, + "loss": 0.8395, + "step": 5572 + }, + { + "epoch": 0.24674850292773098, + "grad_norm": 1.4289237260818481, + "learning_rate": 1.871686814557431e-05, + "loss": 0.8261, + "step": 5573 + }, + { + "epoch": 0.24679277863254487, + "grad_norm": 2.5641684532165527, + "learning_rate": 1.8716110672602652e-05, + "loss": 0.9816, + "step": 5574 + }, + { + "epoch": 0.24683705433735872, + "grad_norm": 1.8272267580032349, + "learning_rate": 1.8715352991453792e-05, + "loss": 0.9961, + "step": 5575 + }, + { + "epoch": 0.2468813300421726, + "grad_norm": 1.2177685499191284, + "learning_rate": 1.8714595102145816e-05, + "loss": 0.7918, + "step": 5576 + }, + { + "epoch": 0.2469256057469865, + "grad_norm": 1.1591870784759521, + "learning_rate": 1.8713837004696836e-05, + "loss": 0.8821, + "step": 5577 + }, + { + "epoch": 0.24696988145180035, + "grad_norm": 1.384597659111023, + "learning_rate": 1.871307869912495e-05, + "loss": 0.9136, + "step": 5578 + }, + { + "epoch": 0.24701415715661423, + "grad_norm": 1.5664054155349731, + "learning_rate": 1.8712320185448277e-05, + "loss": 0.9264, + "step": 5579 + }, + { + "epoch": 0.24705843286142812, + "grad_norm": 1.2018578052520752, + "learning_rate": 1.871156146368493e-05, + "loss": 0.8358, + "step": 5580 + }, + { + "epoch": 0.247102708566242, + "grad_norm": 1.2404515743255615, + "learning_rate": 1.871080253385303e-05, + "loss": 0.8432, + "step": 5581 + }, + { + "epoch": 0.24714698427105586, + "grad_norm": 1.5160179138183594, + "learning_rate": 1.8710043395970704e-05, + "loss": 0.8786, + "step": 5582 + }, + { + "epoch": 0.24719125997586974, + "grad_norm": 1.415476679801941, + "learning_rate": 1.8709284050056083e-05, + "loss": 0.9342, + "step": 5583 + }, + { + "epoch": 0.24723553568068363, + "grad_norm": 1.409756064414978, + "learning_rate": 1.8708524496127307e-05, + "loss": 0.9304, + "step": 5584 + }, + { + "epoch": 0.24727981138549748, + "grad_norm": 1.3170130252838135, + "learning_rate": 1.8707764734202513e-05, + "loss": 0.8944, + "step": 5585 + }, + { + "epoch": 0.24732408709031137, + "grad_norm": 1.4918876886367798, + "learning_rate": 1.8707004764299848e-05, + "loss": 0.8482, + "step": 5586 + }, + { + "epoch": 0.24736836279512525, + "grad_norm": 1.220565676689148, + "learning_rate": 1.870624458643746e-05, + "loss": 0.8766, + "step": 5587 + }, + { + "epoch": 0.2474126384999391, + "grad_norm": 1.53025484085083, + "learning_rate": 1.870548420063351e-05, + "loss": 0.9401, + "step": 5588 + }, + { + "epoch": 0.247456914204753, + "grad_norm": 1.245559573173523, + "learning_rate": 1.8704723606906164e-05, + "loss": 0.9156, + "step": 5589 + }, + { + "epoch": 0.24750118990956688, + "grad_norm": 1.152451753616333, + "learning_rate": 1.8703962805273578e-05, + "loss": 0.7831, + "step": 5590 + }, + { + "epoch": 0.24754546561438076, + "grad_norm": 1.4936046600341797, + "learning_rate": 1.870320179575393e-05, + "loss": 0.9182, + "step": 5591 + }, + { + "epoch": 0.24758974131919462, + "grad_norm": 1.5192357301712036, + "learning_rate": 1.8702440578365387e-05, + "loss": 0.9121, + "step": 5592 + }, + { + "epoch": 0.2476340170240085, + "grad_norm": 1.3432013988494873, + "learning_rate": 1.8701679153126142e-05, + "loss": 0.8777, + "step": 5593 + }, + { + "epoch": 0.24767829272882239, + "grad_norm": 1.4293946027755737, + "learning_rate": 1.8700917520054376e-05, + "loss": 0.8612, + "step": 5594 + }, + { + "epoch": 0.24772256843363624, + "grad_norm": 1.36894690990448, + "learning_rate": 1.8700155679168277e-05, + "loss": 0.8898, + "step": 5595 + }, + { + "epoch": 0.24776684413845013, + "grad_norm": 1.3128297328948975, + "learning_rate": 1.8699393630486044e-05, + "loss": 0.9657, + "step": 5596 + }, + { + "epoch": 0.247811119843264, + "grad_norm": 1.3849544525146484, + "learning_rate": 1.8698631374025876e-05, + "loss": 0.9004, + "step": 5597 + }, + { + "epoch": 0.2478553955480779, + "grad_norm": 1.4186006784439087, + "learning_rate": 1.8697868909805985e-05, + "loss": 0.8395, + "step": 5598 + }, + { + "epoch": 0.24789967125289175, + "grad_norm": 1.2069127559661865, + "learning_rate": 1.8697106237844575e-05, + "loss": 0.8354, + "step": 5599 + }, + { + "epoch": 0.24794394695770564, + "grad_norm": 1.351199746131897, + "learning_rate": 1.8696343358159862e-05, + "loss": 0.8741, + "step": 5600 + }, + { + "epoch": 0.24798822266251952, + "grad_norm": 2.144535541534424, + "learning_rate": 1.869558027077007e-05, + "loss": 0.983, + "step": 5601 + }, + { + "epoch": 0.24803249836733338, + "grad_norm": 1.3346548080444336, + "learning_rate": 1.8694816975693426e-05, + "loss": 0.8752, + "step": 5602 + }, + { + "epoch": 0.24807677407214726, + "grad_norm": 1.4238533973693848, + "learning_rate": 1.8694053472948154e-05, + "loss": 0.9201, + "step": 5603 + }, + { + "epoch": 0.24812104977696114, + "grad_norm": 1.7061443328857422, + "learning_rate": 1.8693289762552496e-05, + "loss": 1.0041, + "step": 5604 + }, + { + "epoch": 0.248165325481775, + "grad_norm": 1.4090805053710938, + "learning_rate": 1.8692525844524692e-05, + "loss": 0.8006, + "step": 5605 + }, + { + "epoch": 0.24820960118658889, + "grad_norm": 1.2789844274520874, + "learning_rate": 1.8691761718882984e-05, + "loss": 0.8751, + "step": 5606 + }, + { + "epoch": 0.24825387689140277, + "grad_norm": 1.4672917127609253, + "learning_rate": 1.8690997385645623e-05, + "loss": 0.8825, + "step": 5607 + }, + { + "epoch": 0.24829815259621665, + "grad_norm": 1.456363320350647, + "learning_rate": 1.8690232844830865e-05, + "loss": 0.8909, + "step": 5608 + }, + { + "epoch": 0.2483424283010305, + "grad_norm": 1.3945118188858032, + "learning_rate": 1.8689468096456975e-05, + "loss": 0.8202, + "step": 5609 + }, + { + "epoch": 0.2483867040058444, + "grad_norm": 1.1570650339126587, + "learning_rate": 1.868870314054221e-05, + "loss": 0.8546, + "step": 5610 + }, + { + "epoch": 0.24843097971065828, + "grad_norm": 1.1994364261627197, + "learning_rate": 1.8687937977104854e-05, + "loss": 0.8446, + "step": 5611 + }, + { + "epoch": 0.24847525541547213, + "grad_norm": 1.462160587310791, + "learning_rate": 1.8687172606163167e-05, + "loss": 0.8116, + "step": 5612 + }, + { + "epoch": 0.24851953112028602, + "grad_norm": 1.5699912309646606, + "learning_rate": 1.8686407027735437e-05, + "loss": 0.9216, + "step": 5613 + }, + { + "epoch": 0.2485638068250999, + "grad_norm": 1.5999423265457153, + "learning_rate": 1.8685641241839946e-05, + "loss": 0.9617, + "step": 5614 + }, + { + "epoch": 0.24860808252991376, + "grad_norm": 1.334886074066162, + "learning_rate": 1.8684875248494987e-05, + "loss": 0.8639, + "step": 5615 + }, + { + "epoch": 0.24865235823472764, + "grad_norm": 1.790226697921753, + "learning_rate": 1.8684109047718856e-05, + "loss": 0.9992, + "step": 5616 + }, + { + "epoch": 0.24869663393954153, + "grad_norm": 1.3139082193374634, + "learning_rate": 1.8683342639529848e-05, + "loss": 0.8436, + "step": 5617 + }, + { + "epoch": 0.2487409096443554, + "grad_norm": 1.4855775833129883, + "learning_rate": 1.8682576023946276e-05, + "loss": 0.8947, + "step": 5618 + }, + { + "epoch": 0.24878518534916927, + "grad_norm": 1.2437812089920044, + "learning_rate": 1.868180920098644e-05, + "loss": 0.8579, + "step": 5619 + }, + { + "epoch": 0.24882946105398315, + "grad_norm": 1.2560912370681763, + "learning_rate": 1.8681042170668663e-05, + "loss": 0.8683, + "step": 5620 + }, + { + "epoch": 0.24887373675879704, + "grad_norm": 1.076005220413208, + "learning_rate": 1.868027493301126e-05, + "loss": 0.7753, + "step": 5621 + }, + { + "epoch": 0.2489180124636109, + "grad_norm": 1.5277652740478516, + "learning_rate": 1.867950748803256e-05, + "loss": 0.8976, + "step": 5622 + }, + { + "epoch": 0.24896228816842478, + "grad_norm": 1.3115270137786865, + "learning_rate": 1.8678739835750893e-05, + "loss": 0.8323, + "step": 5623 + }, + { + "epoch": 0.24900656387323866, + "grad_norm": 1.2204571962356567, + "learning_rate": 1.8677971976184588e-05, + "loss": 0.8242, + "step": 5624 + }, + { + "epoch": 0.24905083957805255, + "grad_norm": 1.6212791204452515, + "learning_rate": 1.867720390935199e-05, + "loss": 0.9413, + "step": 5625 + }, + { + "epoch": 0.2490951152828664, + "grad_norm": 1.330186367034912, + "learning_rate": 1.867643563527144e-05, + "loss": 0.8787, + "step": 5626 + }, + { + "epoch": 0.2491393909876803, + "grad_norm": 1.3577759265899658, + "learning_rate": 1.8675667153961294e-05, + "loss": 0.9053, + "step": 5627 + }, + { + "epoch": 0.24918366669249417, + "grad_norm": 1.3159905672073364, + "learning_rate": 1.8674898465439896e-05, + "loss": 0.8982, + "step": 5628 + }, + { + "epoch": 0.24922794239730803, + "grad_norm": 1.6352983713150024, + "learning_rate": 1.8674129569725613e-05, + "loss": 0.9882, + "step": 5629 + }, + { + "epoch": 0.2492722181021219, + "grad_norm": 1.2144192457199097, + "learning_rate": 1.867336046683681e-05, + "loss": 0.854, + "step": 5630 + }, + { + "epoch": 0.2493164938069358, + "grad_norm": 1.4255361557006836, + "learning_rate": 1.8672591156791853e-05, + "loss": 0.9656, + "step": 5631 + }, + { + "epoch": 0.24936076951174965, + "grad_norm": 1.3541587591171265, + "learning_rate": 1.8671821639609117e-05, + "loss": 0.8742, + "step": 5632 + }, + { + "epoch": 0.24940504521656354, + "grad_norm": 1.9205738306045532, + "learning_rate": 1.8671051915306985e-05, + "loss": 0.9642, + "step": 5633 + }, + { + "epoch": 0.24944932092137742, + "grad_norm": 1.4328969717025757, + "learning_rate": 1.8670281983903837e-05, + "loss": 0.8655, + "step": 5634 + }, + { + "epoch": 0.2494935966261913, + "grad_norm": 1.1626969575881958, + "learning_rate": 1.8669511845418064e-05, + "loss": 0.8825, + "step": 5635 + }, + { + "epoch": 0.24953787233100516, + "grad_norm": 1.3180460929870605, + "learning_rate": 1.8668741499868056e-05, + "loss": 0.903, + "step": 5636 + }, + { + "epoch": 0.24958214803581905, + "grad_norm": 1.2742854356765747, + "learning_rate": 1.8667970947272218e-05, + "loss": 0.9083, + "step": 5637 + }, + { + "epoch": 0.24962642374063293, + "grad_norm": 1.323523759841919, + "learning_rate": 1.866720018764895e-05, + "loss": 0.9524, + "step": 5638 + }, + { + "epoch": 0.2496706994454468, + "grad_norm": 1.2796639204025269, + "learning_rate": 1.8666429221016663e-05, + "loss": 0.8504, + "step": 5639 + }, + { + "epoch": 0.24971497515026067, + "grad_norm": 1.529518961906433, + "learning_rate": 1.8665658047393772e-05, + "loss": 0.919, + "step": 5640 + }, + { + "epoch": 0.24975925085507455, + "grad_norm": 1.6288871765136719, + "learning_rate": 1.8664886666798695e-05, + "loss": 0.8442, + "step": 5641 + }, + { + "epoch": 0.2498035265598884, + "grad_norm": 1.4641773700714111, + "learning_rate": 1.8664115079249857e-05, + "loss": 0.8537, + "step": 5642 + }, + { + "epoch": 0.2498478022647023, + "grad_norm": 1.1230168342590332, + "learning_rate": 1.866334328476568e-05, + "loss": 0.7722, + "step": 5643 + }, + { + "epoch": 0.24989207796951618, + "grad_norm": 1.6828469038009644, + "learning_rate": 1.8662571283364607e-05, + "loss": 0.9154, + "step": 5644 + }, + { + "epoch": 0.24993635367433006, + "grad_norm": 1.3561460971832275, + "learning_rate": 1.8661799075065068e-05, + "loss": 0.9018, + "step": 5645 + }, + { + "epoch": 0.24998062937914392, + "grad_norm": 1.4203301668167114, + "learning_rate": 1.866102665988551e-05, + "loss": 0.9504, + "step": 5646 + }, + { + "epoch": 0.2500249050839578, + "grad_norm": 1.3007944822311401, + "learning_rate": 1.866025403784439e-05, + "loss": 0.8718, + "step": 5647 + }, + { + "epoch": 0.25006918078877166, + "grad_norm": 1.2187782526016235, + "learning_rate": 1.8659481208960146e-05, + "loss": 0.8107, + "step": 5648 + }, + { + "epoch": 0.2501134564935856, + "grad_norm": 1.5222147703170776, + "learning_rate": 1.8658708173251248e-05, + "loss": 0.9136, + "step": 5649 + }, + { + "epoch": 0.25015773219839943, + "grad_norm": 1.1973085403442383, + "learning_rate": 1.8657934930736153e-05, + "loss": 0.8293, + "step": 5650 + }, + { + "epoch": 0.2502020079032133, + "grad_norm": 1.157028079032898, + "learning_rate": 1.8657161481433335e-05, + "loss": 0.7899, + "step": 5651 + }, + { + "epoch": 0.2502462836080272, + "grad_norm": 1.3304545879364014, + "learning_rate": 1.865638782536126e-05, + "loss": 0.9114, + "step": 5652 + }, + { + "epoch": 0.25029055931284105, + "grad_norm": 1.3167660236358643, + "learning_rate": 1.8655613962538416e-05, + "loss": 0.8765, + "step": 5653 + }, + { + "epoch": 0.2503348350176549, + "grad_norm": 1.3678014278411865, + "learning_rate": 1.8654839892983275e-05, + "loss": 0.8929, + "step": 5654 + }, + { + "epoch": 0.2503791107224688, + "grad_norm": 1.313942551612854, + "learning_rate": 1.865406561671433e-05, + "loss": 0.8677, + "step": 5655 + }, + { + "epoch": 0.2504233864272827, + "grad_norm": 1.5914925336837769, + "learning_rate": 1.865329113375008e-05, + "loss": 0.9612, + "step": 5656 + }, + { + "epoch": 0.2504676621320966, + "grad_norm": 1.3211755752563477, + "learning_rate": 1.865251644410901e-05, + "loss": 0.8999, + "step": 5657 + }, + { + "epoch": 0.25051193783691045, + "grad_norm": 1.7358293533325195, + "learning_rate": 1.8651741547809633e-05, + "loss": 0.9384, + "step": 5658 + }, + { + "epoch": 0.2505562135417243, + "grad_norm": 1.4405884742736816, + "learning_rate": 1.865096644487045e-05, + "loss": 0.8468, + "step": 5659 + }, + { + "epoch": 0.2506004892465382, + "grad_norm": 1.465789794921875, + "learning_rate": 1.865019113530998e-05, + "loss": 0.9333, + "step": 5660 + }, + { + "epoch": 0.2506447649513521, + "grad_norm": 1.3457386493682861, + "learning_rate": 1.864941561914674e-05, + "loss": 0.8096, + "step": 5661 + }, + { + "epoch": 0.25068904065616593, + "grad_norm": 1.6130213737487793, + "learning_rate": 1.864863989639925e-05, + "loss": 0.9462, + "step": 5662 + }, + { + "epoch": 0.25073331636097984, + "grad_norm": 1.409856915473938, + "learning_rate": 1.864786396708604e-05, + "loss": 0.9369, + "step": 5663 + }, + { + "epoch": 0.2507775920657937, + "grad_norm": 1.118776559829712, + "learning_rate": 1.8647087831225636e-05, + "loss": 0.7966, + "step": 5664 + }, + { + "epoch": 0.25082186777060755, + "grad_norm": 1.2374554872512817, + "learning_rate": 1.8646311488836583e-05, + "loss": 0.8116, + "step": 5665 + }, + { + "epoch": 0.25086614347542147, + "grad_norm": 1.3567155599594116, + "learning_rate": 1.8645534939937414e-05, + "loss": 0.8688, + "step": 5666 + }, + { + "epoch": 0.2509104191802353, + "grad_norm": 1.219396710395813, + "learning_rate": 1.864475818454669e-05, + "loss": 0.9004, + "step": 5667 + }, + { + "epoch": 0.2509546948850492, + "grad_norm": 1.3005173206329346, + "learning_rate": 1.864398122268295e-05, + "loss": 0.8768, + "step": 5668 + }, + { + "epoch": 0.2509989705898631, + "grad_norm": 1.374121069908142, + "learning_rate": 1.8643204054364762e-05, + "loss": 0.8363, + "step": 5669 + }, + { + "epoch": 0.25104324629467695, + "grad_norm": 1.253731608390808, + "learning_rate": 1.864242667961068e-05, + "loss": 0.8413, + "step": 5670 + }, + { + "epoch": 0.2510875219994908, + "grad_norm": 1.649098515510559, + "learning_rate": 1.8641649098439274e-05, + "loss": 1.0197, + "step": 5671 + }, + { + "epoch": 0.2511317977043047, + "grad_norm": 1.5793780088424683, + "learning_rate": 1.864087131086912e-05, + "loss": 0.8758, + "step": 5672 + }, + { + "epoch": 0.25117607340911857, + "grad_norm": 1.9894038438796997, + "learning_rate": 1.8640093316918782e-05, + "loss": 0.9567, + "step": 5673 + }, + { + "epoch": 0.2512203491139325, + "grad_norm": 1.394850492477417, + "learning_rate": 1.8639315116606852e-05, + "loss": 0.8041, + "step": 5674 + }, + { + "epoch": 0.25126462481874634, + "grad_norm": 1.2769454717636108, + "learning_rate": 1.8638536709951916e-05, + "loss": 0.8906, + "step": 5675 + }, + { + "epoch": 0.2513089005235602, + "grad_norm": 1.3259468078613281, + "learning_rate": 1.863775809697257e-05, + "loss": 0.8612, + "step": 5676 + }, + { + "epoch": 0.2513531762283741, + "grad_norm": 1.2525876760482788, + "learning_rate": 1.86369792776874e-05, + "loss": 0.8281, + "step": 5677 + }, + { + "epoch": 0.25139745193318797, + "grad_norm": 1.3517212867736816, + "learning_rate": 1.863620025211501e-05, + "loss": 0.9809, + "step": 5678 + }, + { + "epoch": 0.2514417276380018, + "grad_norm": 1.4961018562316895, + "learning_rate": 1.863542102027401e-05, + "loss": 0.9349, + "step": 5679 + }, + { + "epoch": 0.25148600334281573, + "grad_norm": 1.2540650367736816, + "learning_rate": 1.8634641582183012e-05, + "loss": 0.8232, + "step": 5680 + }, + { + "epoch": 0.2515302790476296, + "grad_norm": 1.3224296569824219, + "learning_rate": 1.863386193786063e-05, + "loss": 0.9594, + "step": 5681 + }, + { + "epoch": 0.25157455475244345, + "grad_norm": 1.2763997316360474, + "learning_rate": 1.8633082087325485e-05, + "loss": 0.8919, + "step": 5682 + }, + { + "epoch": 0.25161883045725736, + "grad_norm": 1.4542990922927856, + "learning_rate": 1.8632302030596206e-05, + "loss": 0.9305, + "step": 5683 + }, + { + "epoch": 0.2516631061620712, + "grad_norm": 1.5976742506027222, + "learning_rate": 1.8631521767691422e-05, + "loss": 0.9165, + "step": 5684 + }, + { + "epoch": 0.25170738186688507, + "grad_norm": 1.6560921669006348, + "learning_rate": 1.8630741298629763e-05, + "loss": 1.026, + "step": 5685 + }, + { + "epoch": 0.251751657571699, + "grad_norm": 1.2774962186813354, + "learning_rate": 1.862996062342988e-05, + "loss": 0.9179, + "step": 5686 + }, + { + "epoch": 0.25179593327651284, + "grad_norm": 1.3157198429107666, + "learning_rate": 1.8629179742110414e-05, + "loss": 0.8975, + "step": 5687 + }, + { + "epoch": 0.2518402089813267, + "grad_norm": 1.1948115825653076, + "learning_rate": 1.8628398654690017e-05, + "loss": 0.835, + "step": 5688 + }, + { + "epoch": 0.2518844846861406, + "grad_norm": 1.9791810512542725, + "learning_rate": 1.862761736118734e-05, + "loss": 0.9739, + "step": 5689 + }, + { + "epoch": 0.25192876039095446, + "grad_norm": 1.284496545791626, + "learning_rate": 1.862683586162105e-05, + "loss": 0.8688, + "step": 5690 + }, + { + "epoch": 0.2519730360957684, + "grad_norm": 1.2254952192306519, + "learning_rate": 1.8626054156009807e-05, + "loss": 0.8578, + "step": 5691 + }, + { + "epoch": 0.25201731180058223, + "grad_norm": 1.3164132833480835, + "learning_rate": 1.8625272244372288e-05, + "loss": 0.9351, + "step": 5692 + }, + { + "epoch": 0.2520615875053961, + "grad_norm": 1.5291541814804077, + "learning_rate": 1.862449012672716e-05, + "loss": 0.8332, + "step": 5693 + }, + { + "epoch": 0.25210586321021, + "grad_norm": 1.167758822441101, + "learning_rate": 1.862370780309311e-05, + "loss": 0.7694, + "step": 5694 + }, + { + "epoch": 0.25215013891502386, + "grad_norm": 1.3968133926391602, + "learning_rate": 1.862292527348882e-05, + "loss": 0.848, + "step": 5695 + }, + { + "epoch": 0.2521944146198377, + "grad_norm": 1.422768473625183, + "learning_rate": 1.8622142537932984e-05, + "loss": 0.9028, + "step": 5696 + }, + { + "epoch": 0.2522386903246516, + "grad_norm": 1.4985507726669312, + "learning_rate": 1.8621359596444293e-05, + "loss": 0.8892, + "step": 5697 + }, + { + "epoch": 0.2522829660294655, + "grad_norm": 1.2884414196014404, + "learning_rate": 1.8620576449041446e-05, + "loss": 0.8904, + "step": 5698 + }, + { + "epoch": 0.25232724173427934, + "grad_norm": 1.6130330562591553, + "learning_rate": 1.861979309574315e-05, + "loss": 0.9181, + "step": 5699 + }, + { + "epoch": 0.25237151743909325, + "grad_norm": 1.4045289754867554, + "learning_rate": 1.8619009536568115e-05, + "loss": 0.8807, + "step": 5700 + }, + { + "epoch": 0.2524157931439071, + "grad_norm": 1.2760512828826904, + "learning_rate": 1.8618225771535053e-05, + "loss": 0.8356, + "step": 5701 + }, + { + "epoch": 0.25246006884872096, + "grad_norm": 4.3345417976379395, + "learning_rate": 1.861744180066269e-05, + "loss": 0.9793, + "step": 5702 + }, + { + "epoch": 0.2525043445535349, + "grad_norm": 1.2271267175674438, + "learning_rate": 1.861665762396974e-05, + "loss": 0.8327, + "step": 5703 + }, + { + "epoch": 0.25254862025834873, + "grad_norm": 1.3877437114715576, + "learning_rate": 1.8615873241474943e-05, + "loss": 0.9446, + "step": 5704 + }, + { + "epoch": 0.2525928959631626, + "grad_norm": 1.4465810060501099, + "learning_rate": 1.8615088653197028e-05, + "loss": 0.8288, + "step": 5705 + }, + { + "epoch": 0.2526371716679765, + "grad_norm": 1.2904102802276611, + "learning_rate": 1.8614303859154735e-05, + "loss": 0.8521, + "step": 5706 + }, + { + "epoch": 0.25268144737279036, + "grad_norm": 1.1166133880615234, + "learning_rate": 1.8613518859366807e-05, + "loss": 0.8184, + "step": 5707 + }, + { + "epoch": 0.2527257230776042, + "grad_norm": 1.1129066944122314, + "learning_rate": 1.8612733653852e-05, + "loss": 0.7889, + "step": 5708 + }, + { + "epoch": 0.2527699987824181, + "grad_norm": 1.55813729763031, + "learning_rate": 1.8611948242629056e-05, + "loss": 0.8951, + "step": 5709 + }, + { + "epoch": 0.252814274487232, + "grad_norm": 1.418284296989441, + "learning_rate": 1.8611162625716745e-05, + "loss": 0.9281, + "step": 5710 + }, + { + "epoch": 0.2528585501920459, + "grad_norm": 1.5835895538330078, + "learning_rate": 1.861037680313382e-05, + "loss": 0.9476, + "step": 5711 + }, + { + "epoch": 0.25290282589685975, + "grad_norm": 1.7175620794296265, + "learning_rate": 1.860959077489906e-05, + "loss": 0.9428, + "step": 5712 + }, + { + "epoch": 0.2529471016016736, + "grad_norm": 1.222442388534546, + "learning_rate": 1.8608804541031234e-05, + "loss": 0.8252, + "step": 5713 + }, + { + "epoch": 0.2529913773064875, + "grad_norm": 1.3747237920761108, + "learning_rate": 1.860801810154912e-05, + "loss": 0.9101, + "step": 5714 + }, + { + "epoch": 0.2530356530113014, + "grad_norm": 2.110184907913208, + "learning_rate": 1.8607231456471505e-05, + "loss": 0.928, + "step": 5715 + }, + { + "epoch": 0.25307992871611523, + "grad_norm": 1.2618827819824219, + "learning_rate": 1.8606444605817177e-05, + "loss": 0.8521, + "step": 5716 + }, + { + "epoch": 0.25312420442092914, + "grad_norm": 1.521604299545288, + "learning_rate": 1.860565754960492e-05, + "loss": 0.9884, + "step": 5717 + }, + { + "epoch": 0.253168480125743, + "grad_norm": 1.2455790042877197, + "learning_rate": 1.8604870287853546e-05, + "loss": 0.8617, + "step": 5718 + }, + { + "epoch": 0.25321275583055686, + "grad_norm": 1.3348970413208008, + "learning_rate": 1.8604082820581848e-05, + "loss": 0.8813, + "step": 5719 + }, + { + "epoch": 0.25325703153537077, + "grad_norm": 1.5162453651428223, + "learning_rate": 1.8603295147808637e-05, + "loss": 0.8625, + "step": 5720 + }, + { + "epoch": 0.2533013072401846, + "grad_norm": 1.1870118379592896, + "learning_rate": 1.8602507269552727e-05, + "loss": 0.8427, + "step": 5721 + }, + { + "epoch": 0.2533455829449985, + "grad_norm": 1.2975002527236938, + "learning_rate": 1.8601719185832937e-05, + "loss": 0.8595, + "step": 5722 + }, + { + "epoch": 0.2533898586498124, + "grad_norm": 1.389440655708313, + "learning_rate": 1.8600930896668087e-05, + "loss": 0.9486, + "step": 5723 + }, + { + "epoch": 0.25343413435462625, + "grad_norm": 1.146743655204773, + "learning_rate": 1.8600142402077006e-05, + "loss": 0.759, + "step": 5724 + }, + { + "epoch": 0.2534784100594401, + "grad_norm": 1.4653546810150146, + "learning_rate": 1.8599353702078527e-05, + "loss": 0.9101, + "step": 5725 + }, + { + "epoch": 0.253522685764254, + "grad_norm": 1.2509572505950928, + "learning_rate": 1.859856479669149e-05, + "loss": 0.8339, + "step": 5726 + }, + { + "epoch": 0.2535669614690679, + "grad_norm": 1.403869867324829, + "learning_rate": 1.8597775685934726e-05, + "loss": 0.938, + "step": 5727 + }, + { + "epoch": 0.2536112371738818, + "grad_norm": 1.3068441152572632, + "learning_rate": 1.8596986369827094e-05, + "loss": 0.8264, + "step": 5728 + }, + { + "epoch": 0.25365551287869564, + "grad_norm": 1.2144157886505127, + "learning_rate": 1.8596196848387443e-05, + "loss": 0.922, + "step": 5729 + }, + { + "epoch": 0.2536997885835095, + "grad_norm": 1.641617774963379, + "learning_rate": 1.8595407121634626e-05, + "loss": 0.9274, + "step": 5730 + }, + { + "epoch": 0.2537440642883234, + "grad_norm": 1.1704354286193848, + "learning_rate": 1.8594617189587515e-05, + "loss": 0.832, + "step": 5731 + }, + { + "epoch": 0.25378833999313727, + "grad_norm": 1.3959052562713623, + "learning_rate": 1.8593827052264967e-05, + "loss": 0.842, + "step": 5732 + }, + { + "epoch": 0.2538326156979511, + "grad_norm": 1.2229405641555786, + "learning_rate": 1.8593036709685855e-05, + "loss": 0.8637, + "step": 5733 + }, + { + "epoch": 0.25387689140276504, + "grad_norm": 1.253859519958496, + "learning_rate": 1.8592246161869064e-05, + "loss": 0.8167, + "step": 5734 + }, + { + "epoch": 0.2539211671075789, + "grad_norm": 1.595505714416504, + "learning_rate": 1.8591455408833462e-05, + "loss": 0.922, + "step": 5735 + }, + { + "epoch": 0.25396544281239275, + "grad_norm": 1.154792070388794, + "learning_rate": 1.859066445059795e-05, + "loss": 0.7721, + "step": 5736 + }, + { + "epoch": 0.25400971851720666, + "grad_norm": 1.538986086845398, + "learning_rate": 1.8589873287181407e-05, + "loss": 0.9316, + "step": 5737 + }, + { + "epoch": 0.2540539942220205, + "grad_norm": 1.4476380348205566, + "learning_rate": 1.8589081918602735e-05, + "loss": 0.9008, + "step": 5738 + }, + { + "epoch": 0.2540982699268344, + "grad_norm": 1.2498610019683838, + "learning_rate": 1.858829034488084e-05, + "loss": 0.9234, + "step": 5739 + }, + { + "epoch": 0.2541425456316483, + "grad_norm": 1.4372631311416626, + "learning_rate": 1.8587498566034618e-05, + "loss": 0.9175, + "step": 5740 + }, + { + "epoch": 0.25418682133646214, + "grad_norm": 1.2072076797485352, + "learning_rate": 1.858670658208299e-05, + "loss": 0.828, + "step": 5741 + }, + { + "epoch": 0.254231097041276, + "grad_norm": 1.4489762783050537, + "learning_rate": 1.858591439304486e-05, + "loss": 0.9119, + "step": 5742 + }, + { + "epoch": 0.2542753727460899, + "grad_norm": 1.6810669898986816, + "learning_rate": 1.858512199893916e-05, + "loss": 0.9616, + "step": 5743 + }, + { + "epoch": 0.25431964845090377, + "grad_norm": 1.3519551753997803, + "learning_rate": 1.858432939978481e-05, + "loss": 0.8685, + "step": 5744 + }, + { + "epoch": 0.2543639241557177, + "grad_norm": 1.7134782075881958, + "learning_rate": 1.8583536595600742e-05, + "loss": 0.9425, + "step": 5745 + }, + { + "epoch": 0.25440819986053154, + "grad_norm": 1.1526470184326172, + "learning_rate": 1.8582743586405892e-05, + "loss": 0.7759, + "step": 5746 + }, + { + "epoch": 0.2544524755653454, + "grad_norm": 1.3604373931884766, + "learning_rate": 1.85819503722192e-05, + "loss": 0.8477, + "step": 5747 + }, + { + "epoch": 0.2544967512701593, + "grad_norm": 1.2933661937713623, + "learning_rate": 1.858115695305961e-05, + "loss": 0.7856, + "step": 5748 + }, + { + "epoch": 0.25454102697497316, + "grad_norm": 1.2787350416183472, + "learning_rate": 1.8580363328946073e-05, + "loss": 0.8536, + "step": 5749 + }, + { + "epoch": 0.254585302679787, + "grad_norm": 1.1345670223236084, + "learning_rate": 1.8579569499897544e-05, + "loss": 0.8008, + "step": 5750 + }, + { + "epoch": 0.25462957838460093, + "grad_norm": 21.226364135742188, + "learning_rate": 1.8578775465932984e-05, + "loss": 0.8543, + "step": 5751 + }, + { + "epoch": 0.2546738540894148, + "grad_norm": 1.348341703414917, + "learning_rate": 1.857798122707136e-05, + "loss": 0.8946, + "step": 5752 + }, + { + "epoch": 0.25471812979422864, + "grad_norm": 1.2244808673858643, + "learning_rate": 1.8577186783331632e-05, + "loss": 0.8069, + "step": 5753 + }, + { + "epoch": 0.25476240549904255, + "grad_norm": 1.367935299873352, + "learning_rate": 1.857639213473279e-05, + "loss": 0.8551, + "step": 5754 + }, + { + "epoch": 0.2548066812038564, + "grad_norm": 1.3476488590240479, + "learning_rate": 1.8575597281293798e-05, + "loss": 0.969, + "step": 5755 + }, + { + "epoch": 0.25485095690867027, + "grad_norm": 1.5570714473724365, + "learning_rate": 1.857480222303365e-05, + "loss": 0.9275, + "step": 5756 + }, + { + "epoch": 0.2548952326134842, + "grad_norm": 1.338659644126892, + "learning_rate": 1.8574006959971335e-05, + "loss": 0.8401, + "step": 5757 + }, + { + "epoch": 0.25493950831829804, + "grad_norm": 1.5890847444534302, + "learning_rate": 1.857321149212584e-05, + "loss": 0.8831, + "step": 5758 + }, + { + "epoch": 0.2549837840231119, + "grad_norm": 1.3286490440368652, + "learning_rate": 1.8572415819516174e-05, + "loss": 0.872, + "step": 5759 + }, + { + "epoch": 0.2550280597279258, + "grad_norm": 1.3206454515457153, + "learning_rate": 1.8571619942161333e-05, + "loss": 0.804, + "step": 5760 + }, + { + "epoch": 0.25507233543273966, + "grad_norm": 1.3332782983779907, + "learning_rate": 1.857082386008033e-05, + "loss": 0.9088, + "step": 5761 + }, + { + "epoch": 0.2551166111375536, + "grad_norm": 1.7048468589782715, + "learning_rate": 1.8570027573292177e-05, + "loss": 0.9553, + "step": 5762 + }, + { + "epoch": 0.25516088684236743, + "grad_norm": 1.4513729810714722, + "learning_rate": 1.8569231081815895e-05, + "loss": 0.8878, + "step": 5763 + }, + { + "epoch": 0.2552051625471813, + "grad_norm": 1.4663496017456055, + "learning_rate": 1.8568434385670502e-05, + "loss": 0.9539, + "step": 5764 + }, + { + "epoch": 0.2552494382519952, + "grad_norm": 1.8887193202972412, + "learning_rate": 1.8567637484875035e-05, + "loss": 0.8573, + "step": 5765 + }, + { + "epoch": 0.25529371395680905, + "grad_norm": 1.7283910512924194, + "learning_rate": 1.8566840379448523e-05, + "loss": 0.9897, + "step": 5766 + }, + { + "epoch": 0.2553379896616229, + "grad_norm": 1.182548999786377, + "learning_rate": 1.8566043069410002e-05, + "loss": 0.8455, + "step": 5767 + }, + { + "epoch": 0.2553822653664368, + "grad_norm": 1.5442720651626587, + "learning_rate": 1.8565245554778516e-05, + "loss": 0.9526, + "step": 5768 + }, + { + "epoch": 0.2554265410712507, + "grad_norm": 1.3897863626480103, + "learning_rate": 1.8564447835573113e-05, + "loss": 0.8916, + "step": 5769 + }, + { + "epoch": 0.25547081677606454, + "grad_norm": 1.814794898033142, + "learning_rate": 1.8563649911812852e-05, + "loss": 0.9059, + "step": 5770 + }, + { + "epoch": 0.25551509248087845, + "grad_norm": 1.6039122343063354, + "learning_rate": 1.8562851783516783e-05, + "loss": 0.9826, + "step": 5771 + }, + { + "epoch": 0.2555593681856923, + "grad_norm": 1.3283084630966187, + "learning_rate": 1.856205345070397e-05, + "loss": 0.8106, + "step": 5772 + }, + { + "epoch": 0.25560364389050616, + "grad_norm": 1.2216511964797974, + "learning_rate": 1.856125491339348e-05, + "loss": 0.8668, + "step": 5773 + }, + { + "epoch": 0.2556479195953201, + "grad_norm": 1.3618682622909546, + "learning_rate": 1.8560456171604392e-05, + "loss": 0.8618, + "step": 5774 + }, + { + "epoch": 0.25569219530013393, + "grad_norm": 1.552557349205017, + "learning_rate": 1.8559657225355776e-05, + "loss": 0.8629, + "step": 5775 + }, + { + "epoch": 0.2557364710049478, + "grad_norm": 1.187432885169983, + "learning_rate": 1.8558858074666718e-05, + "loss": 0.8133, + "step": 5776 + }, + { + "epoch": 0.2557807467097617, + "grad_norm": 1.277757167816162, + "learning_rate": 1.85580587195563e-05, + "loss": 0.8953, + "step": 5777 + }, + { + "epoch": 0.25582502241457555, + "grad_norm": 1.2902731895446777, + "learning_rate": 1.8557259160043623e-05, + "loss": 0.8247, + "step": 5778 + }, + { + "epoch": 0.25586929811938947, + "grad_norm": 1.2566088438034058, + "learning_rate": 1.8556459396147777e-05, + "loss": 0.8263, + "step": 5779 + }, + { + "epoch": 0.2559135738242033, + "grad_norm": 1.207153081893921, + "learning_rate": 1.8555659427887862e-05, + "loss": 0.8616, + "step": 5780 + }, + { + "epoch": 0.2559578495290172, + "grad_norm": 1.3003437519073486, + "learning_rate": 1.855485925528299e-05, + "loss": 0.8423, + "step": 5781 + }, + { + "epoch": 0.2560021252338311, + "grad_norm": 1.3240762948989868, + "learning_rate": 1.855405887835227e-05, + "loss": 0.9174, + "step": 5782 + }, + { + "epoch": 0.25604640093864495, + "grad_norm": 2.1207187175750732, + "learning_rate": 1.855325829711482e-05, + "loss": 0.8429, + "step": 5783 + }, + { + "epoch": 0.2560906766434588, + "grad_norm": 1.4148595333099365, + "learning_rate": 1.855245751158976e-05, + "loss": 0.8974, + "step": 5784 + }, + { + "epoch": 0.2561349523482727, + "grad_norm": 1.362252116203308, + "learning_rate": 1.8551656521796216e-05, + "loss": 0.8629, + "step": 5785 + }, + { + "epoch": 0.25617922805308657, + "grad_norm": 1.410872220993042, + "learning_rate": 1.8550855327753316e-05, + "loss": 0.9786, + "step": 5786 + }, + { + "epoch": 0.25622350375790043, + "grad_norm": 1.464064598083496, + "learning_rate": 1.8550053929480202e-05, + "loss": 0.9018, + "step": 5787 + }, + { + "epoch": 0.25626777946271434, + "grad_norm": 1.424623966217041, + "learning_rate": 1.854925232699601e-05, + "loss": 0.9814, + "step": 5788 + }, + { + "epoch": 0.2563120551675282, + "grad_norm": 1.561025857925415, + "learning_rate": 1.854845052031989e-05, + "loss": 0.9603, + "step": 5789 + }, + { + "epoch": 0.25635633087234205, + "grad_norm": 1.3193039894104004, + "learning_rate": 1.8547648509470985e-05, + "loss": 0.848, + "step": 5790 + }, + { + "epoch": 0.25640060657715597, + "grad_norm": 1.467512845993042, + "learning_rate": 1.8546846294468455e-05, + "loss": 0.878, + "step": 5791 + }, + { + "epoch": 0.2564448822819698, + "grad_norm": 1.2591890096664429, + "learning_rate": 1.8546043875331467e-05, + "loss": 0.9094, + "step": 5792 + }, + { + "epoch": 0.2564891579867837, + "grad_norm": 1.350967288017273, + "learning_rate": 1.8545241252079172e-05, + "loss": 0.9038, + "step": 5793 + }, + { + "epoch": 0.2565334336915976, + "grad_norm": 1.2934725284576416, + "learning_rate": 1.8544438424730752e-05, + "loss": 0.8468, + "step": 5794 + }, + { + "epoch": 0.25657770939641145, + "grad_norm": 1.1878886222839355, + "learning_rate": 1.8543635393305375e-05, + "loss": 0.8556, + "step": 5795 + }, + { + "epoch": 0.2566219851012253, + "grad_norm": 1.2698872089385986, + "learning_rate": 1.8542832157822227e-05, + "loss": 0.8472, + "step": 5796 + }, + { + "epoch": 0.2566662608060392, + "grad_norm": 1.3925766944885254, + "learning_rate": 1.8542028718300487e-05, + "loss": 0.9212, + "step": 5797 + }, + { + "epoch": 0.25671053651085307, + "grad_norm": 1.423671007156372, + "learning_rate": 1.8541225074759344e-05, + "loss": 0.871, + "step": 5798 + }, + { + "epoch": 0.256754812215667, + "grad_norm": 1.2644869089126587, + "learning_rate": 1.8540421227217998e-05, + "loss": 0.7975, + "step": 5799 + }, + { + "epoch": 0.25679908792048084, + "grad_norm": 1.2749555110931396, + "learning_rate": 1.853961717569564e-05, + "loss": 0.7829, + "step": 5800 + }, + { + "epoch": 0.2568433636252947, + "grad_norm": 1.5849910974502563, + "learning_rate": 1.8538812920211484e-05, + "loss": 0.9453, + "step": 5801 + }, + { + "epoch": 0.2568876393301086, + "grad_norm": 1.1467809677124023, + "learning_rate": 1.853800846078473e-05, + "loss": 0.7533, + "step": 5802 + }, + { + "epoch": 0.25693191503492246, + "grad_norm": 1.3679531812667847, + "learning_rate": 1.85372037974346e-05, + "loss": 0.8631, + "step": 5803 + }, + { + "epoch": 0.2569761907397363, + "grad_norm": 1.3639369010925293, + "learning_rate": 1.8536398930180306e-05, + "loss": 0.9373, + "step": 5804 + }, + { + "epoch": 0.25702046644455023, + "grad_norm": 1.5153638124465942, + "learning_rate": 1.8535593859041073e-05, + "loss": 0.8874, + "step": 5805 + }, + { + "epoch": 0.2570647421493641, + "grad_norm": 1.3813772201538086, + "learning_rate": 1.853478858403613e-05, + "loss": 0.8417, + "step": 5806 + }, + { + "epoch": 0.25710901785417795, + "grad_norm": 1.3425966501235962, + "learning_rate": 1.8533983105184713e-05, + "loss": 0.8131, + "step": 5807 + }, + { + "epoch": 0.25715329355899186, + "grad_norm": 1.4976979494094849, + "learning_rate": 1.853317742250606e-05, + "loss": 0.9304, + "step": 5808 + }, + { + "epoch": 0.2571975692638057, + "grad_norm": 1.099625825881958, + "learning_rate": 1.853237153601941e-05, + "loss": 0.7255, + "step": 5809 + }, + { + "epoch": 0.25724184496861957, + "grad_norm": 1.3541643619537354, + "learning_rate": 1.8531565445744012e-05, + "loss": 0.9244, + "step": 5810 + }, + { + "epoch": 0.2572861206734335, + "grad_norm": 1.2896709442138672, + "learning_rate": 1.8530759151699122e-05, + "loss": 0.9077, + "step": 5811 + }, + { + "epoch": 0.25733039637824734, + "grad_norm": 1.2341387271881104, + "learning_rate": 1.8529952653903996e-05, + "loss": 0.8992, + "step": 5812 + }, + { + "epoch": 0.2573746720830612, + "grad_norm": 1.2103739976882935, + "learning_rate": 1.8529145952377894e-05, + "loss": 0.8923, + "step": 5813 + }, + { + "epoch": 0.2574189477878751, + "grad_norm": 1.286213994026184, + "learning_rate": 1.8528339047140087e-05, + "loss": 0.9364, + "step": 5814 + }, + { + "epoch": 0.25746322349268896, + "grad_norm": 1.2944709062576294, + "learning_rate": 1.8527531938209847e-05, + "loss": 0.902, + "step": 5815 + }, + { + "epoch": 0.2575074991975029, + "grad_norm": 1.6220824718475342, + "learning_rate": 1.852672462560645e-05, + "loss": 0.9318, + "step": 5816 + }, + { + "epoch": 0.25755177490231673, + "grad_norm": 1.8014236688613892, + "learning_rate": 1.8525917109349173e-05, + "loss": 1.0048, + "step": 5817 + }, + { + "epoch": 0.2575960506071306, + "grad_norm": 1.544314980506897, + "learning_rate": 1.8525109389457314e-05, + "loss": 0.8759, + "step": 5818 + }, + { + "epoch": 0.2576403263119445, + "grad_norm": 1.4525291919708252, + "learning_rate": 1.8524301465950155e-05, + "loss": 0.9093, + "step": 5819 + }, + { + "epoch": 0.25768460201675836, + "grad_norm": 1.5453176498413086, + "learning_rate": 1.8523493338847e-05, + "loss": 0.8822, + "step": 5820 + }, + { + "epoch": 0.2577288777215722, + "grad_norm": 1.5269855260849, + "learning_rate": 1.8522685008167143e-05, + "loss": 0.9105, + "step": 5821 + }, + { + "epoch": 0.2577731534263861, + "grad_norm": 1.7312147617340088, + "learning_rate": 1.8521876473929893e-05, + "loss": 0.961, + "step": 5822 + }, + { + "epoch": 0.2578174291312, + "grad_norm": 1.3069220781326294, + "learning_rate": 1.8521067736154567e-05, + "loss": 0.8467, + "step": 5823 + }, + { + "epoch": 0.25786170483601384, + "grad_norm": 1.380174160003662, + "learning_rate": 1.8520258794860476e-05, + "loss": 0.806, + "step": 5824 + }, + { + "epoch": 0.25790598054082775, + "grad_norm": 1.4262217283248901, + "learning_rate": 1.8519449650066936e-05, + "loss": 0.9027, + "step": 5825 + }, + { + "epoch": 0.2579502562456416, + "grad_norm": 1.1472558975219727, + "learning_rate": 1.851864030179328e-05, + "loss": 0.7877, + "step": 5826 + }, + { + "epoch": 0.25799453195045546, + "grad_norm": 1.1925983428955078, + "learning_rate": 1.8517830750058838e-05, + "loss": 0.8232, + "step": 5827 + }, + { + "epoch": 0.2580388076552694, + "grad_norm": 1.3396753072738647, + "learning_rate": 1.8517020994882945e-05, + "loss": 0.8745, + "step": 5828 + }, + { + "epoch": 0.25808308336008323, + "grad_norm": 1.5728338956832886, + "learning_rate": 1.8516211036284937e-05, + "loss": 1.0203, + "step": 5829 + }, + { + "epoch": 0.2581273590648971, + "grad_norm": 1.3541690111160278, + "learning_rate": 1.8515400874284168e-05, + "loss": 0.8638, + "step": 5830 + }, + { + "epoch": 0.258171634769711, + "grad_norm": 1.2458082437515259, + "learning_rate": 1.8514590508899973e-05, + "loss": 0.879, + "step": 5831 + }, + { + "epoch": 0.25821591047452486, + "grad_norm": 1.639639973640442, + "learning_rate": 1.8513779940151722e-05, + "loss": 0.9413, + "step": 5832 + }, + { + "epoch": 0.25826018617933877, + "grad_norm": 1.3272849321365356, + "learning_rate": 1.8512969168058773e-05, + "loss": 0.88, + "step": 5833 + }, + { + "epoch": 0.2583044618841526, + "grad_norm": 1.5078239440917969, + "learning_rate": 1.851215819264048e-05, + "loss": 0.8493, + "step": 5834 + }, + { + "epoch": 0.2583487375889665, + "grad_norm": 1.3424535989761353, + "learning_rate": 1.8511347013916228e-05, + "loss": 0.8304, + "step": 5835 + }, + { + "epoch": 0.2583930132937804, + "grad_norm": 1.194052815437317, + "learning_rate": 1.8510535631905377e-05, + "loss": 0.8242, + "step": 5836 + }, + { + "epoch": 0.25843728899859425, + "grad_norm": 1.0500543117523193, + "learning_rate": 1.850972404662731e-05, + "loss": 0.8084, + "step": 5837 + }, + { + "epoch": 0.2584815647034081, + "grad_norm": 1.4298957586288452, + "learning_rate": 1.8508912258101416e-05, + "loss": 0.9083, + "step": 5838 + }, + { + "epoch": 0.258525840408222, + "grad_norm": 1.6336724758148193, + "learning_rate": 1.850810026634708e-05, + "loss": 0.8415, + "step": 5839 + }, + { + "epoch": 0.2585701161130359, + "grad_norm": 1.3196438550949097, + "learning_rate": 1.85072880713837e-05, + "loss": 0.9063, + "step": 5840 + }, + { + "epoch": 0.25861439181784973, + "grad_norm": 3.790052890777588, + "learning_rate": 1.8506475673230668e-05, + "loss": 0.9775, + "step": 5841 + }, + { + "epoch": 0.25865866752266364, + "grad_norm": 1.6322948932647705, + "learning_rate": 1.8505663071907392e-05, + "loss": 0.9266, + "step": 5842 + }, + { + "epoch": 0.2587029432274775, + "grad_norm": 1.3850915431976318, + "learning_rate": 1.8504850267433278e-05, + "loss": 0.8482, + "step": 5843 + }, + { + "epoch": 0.25874721893229136, + "grad_norm": 1.1562507152557373, + "learning_rate": 1.8504037259827742e-05, + "loss": 0.8173, + "step": 5844 + }, + { + "epoch": 0.25879149463710527, + "grad_norm": 1.3467628955841064, + "learning_rate": 1.85032240491102e-05, + "loss": 0.9251, + "step": 5845 + }, + { + "epoch": 0.2588357703419191, + "grad_norm": 1.3337770700454712, + "learning_rate": 1.8502410635300072e-05, + "loss": 0.9077, + "step": 5846 + }, + { + "epoch": 0.258880046046733, + "grad_norm": 1.3984267711639404, + "learning_rate": 1.850159701841679e-05, + "loss": 0.898, + "step": 5847 + }, + { + "epoch": 0.2589243217515469, + "grad_norm": 1.1638199090957642, + "learning_rate": 1.8500783198479787e-05, + "loss": 0.8581, + "step": 5848 + }, + { + "epoch": 0.25896859745636075, + "grad_norm": 1.3513116836547852, + "learning_rate": 1.84999691755085e-05, + "loss": 0.9241, + "step": 5849 + }, + { + "epoch": 0.25901287316117466, + "grad_norm": 1.7133744955062866, + "learning_rate": 1.849915494952237e-05, + "loss": 0.795, + "step": 5850 + }, + { + "epoch": 0.2590571488659885, + "grad_norm": 1.3165208101272583, + "learning_rate": 1.8498340520540843e-05, + "loss": 0.8563, + "step": 5851 + }, + { + "epoch": 0.2591014245708024, + "grad_norm": 1.58371901512146, + "learning_rate": 1.849752588858337e-05, + "loss": 0.8547, + "step": 5852 + }, + { + "epoch": 0.2591457002756163, + "grad_norm": 1.0243264436721802, + "learning_rate": 1.8496711053669412e-05, + "loss": 0.7142, + "step": 5853 + }, + { + "epoch": 0.25918997598043014, + "grad_norm": 1.4082748889923096, + "learning_rate": 1.849589601581843e-05, + "loss": 0.91, + "step": 5854 + }, + { + "epoch": 0.259234251685244, + "grad_norm": 1.4278249740600586, + "learning_rate": 1.849508077504989e-05, + "loss": 0.8308, + "step": 5855 + }, + { + "epoch": 0.2592785273900579, + "grad_norm": 1.2918137311935425, + "learning_rate": 1.849426533138326e-05, + "loss": 0.8298, + "step": 5856 + }, + { + "epoch": 0.25932280309487177, + "grad_norm": 1.572596788406372, + "learning_rate": 1.8493449684838022e-05, + "loss": 0.8748, + "step": 5857 + }, + { + "epoch": 0.2593670787996856, + "grad_norm": 1.2510454654693604, + "learning_rate": 1.849263383543365e-05, + "loss": 0.8303, + "step": 5858 + }, + { + "epoch": 0.25941135450449954, + "grad_norm": 1.2310609817504883, + "learning_rate": 1.8491817783189636e-05, + "loss": 0.8759, + "step": 5859 + }, + { + "epoch": 0.2594556302093134, + "grad_norm": 1.6870639324188232, + "learning_rate": 1.849100152812547e-05, + "loss": 0.8894, + "step": 5860 + }, + { + "epoch": 0.25949990591412725, + "grad_norm": 1.9843871593475342, + "learning_rate": 1.8490185070260645e-05, + "loss": 0.8436, + "step": 5861 + }, + { + "epoch": 0.25954418161894116, + "grad_norm": 1.2835584878921509, + "learning_rate": 1.848936840961466e-05, + "loss": 0.8736, + "step": 5862 + }, + { + "epoch": 0.259588457323755, + "grad_norm": 1.2887132167816162, + "learning_rate": 1.8488551546207025e-05, + "loss": 0.8992, + "step": 5863 + }, + { + "epoch": 0.2596327330285689, + "grad_norm": 1.3122646808624268, + "learning_rate": 1.8487734480057246e-05, + "loss": 0.8606, + "step": 5864 + }, + { + "epoch": 0.2596770087333828, + "grad_norm": 1.3687529563903809, + "learning_rate": 1.8486917211184844e-05, + "loss": 0.8576, + "step": 5865 + }, + { + "epoch": 0.25972128443819664, + "grad_norm": 1.1738619804382324, + "learning_rate": 1.8486099739609332e-05, + "loss": 0.8154, + "step": 5866 + }, + { + "epoch": 0.2597655601430105, + "grad_norm": 1.2281677722930908, + "learning_rate": 1.8485282065350237e-05, + "loss": 0.7899, + "step": 5867 + }, + { + "epoch": 0.2598098358478244, + "grad_norm": 1.2777386903762817, + "learning_rate": 1.848446418842709e-05, + "loss": 0.9277, + "step": 5868 + }, + { + "epoch": 0.25985411155263827, + "grad_norm": 1.3979045152664185, + "learning_rate": 1.848364610885942e-05, + "loss": 0.9265, + "step": 5869 + }, + { + "epoch": 0.2598983872574522, + "grad_norm": 1.1108322143554688, + "learning_rate": 1.8482827826666774e-05, + "loss": 0.7217, + "step": 5870 + }, + { + "epoch": 0.25994266296226604, + "grad_norm": 1.6299302577972412, + "learning_rate": 1.8482009341868696e-05, + "loss": 0.854, + "step": 5871 + }, + { + "epoch": 0.2599869386670799, + "grad_norm": 1.5467569828033447, + "learning_rate": 1.8481190654484728e-05, + "loss": 0.9093, + "step": 5872 + }, + { + "epoch": 0.2600312143718938, + "grad_norm": 1.268049716949463, + "learning_rate": 1.8480371764534422e-05, + "loss": 0.8509, + "step": 5873 + }, + { + "epoch": 0.26007549007670766, + "grad_norm": 1.562674641609192, + "learning_rate": 1.847955267203735e-05, + "loss": 0.9383, + "step": 5874 + }, + { + "epoch": 0.2601197657815215, + "grad_norm": 1.3173712491989136, + "learning_rate": 1.847873337701306e-05, + "loss": 0.8887, + "step": 5875 + }, + { + "epoch": 0.26016404148633543, + "grad_norm": 1.2473399639129639, + "learning_rate": 1.847791387948113e-05, + "loss": 0.8576, + "step": 5876 + }, + { + "epoch": 0.2602083171911493, + "grad_norm": 1.3500100374221802, + "learning_rate": 1.847709417946113e-05, + "loss": 0.8779, + "step": 5877 + }, + { + "epoch": 0.26025259289596314, + "grad_norm": 1.4309605360031128, + "learning_rate": 1.8476274276972635e-05, + "loss": 0.9198, + "step": 5878 + }, + { + "epoch": 0.26029686860077705, + "grad_norm": 1.3421162366867065, + "learning_rate": 1.8475454172035238e-05, + "loss": 0.8938, + "step": 5879 + }, + { + "epoch": 0.2603411443055909, + "grad_norm": 1.215354323387146, + "learning_rate": 1.847463386466851e-05, + "loss": 0.8152, + "step": 5880 + }, + { + "epoch": 0.26038542001040477, + "grad_norm": 1.2564997673034668, + "learning_rate": 1.8473813354892058e-05, + "loss": 0.8302, + "step": 5881 + }, + { + "epoch": 0.2604296957152187, + "grad_norm": 1.5017403364181519, + "learning_rate": 1.8472992642725472e-05, + "loss": 0.9514, + "step": 5882 + }, + { + "epoch": 0.26047397142003254, + "grad_norm": 1.2718685865402222, + "learning_rate": 1.8472171728188356e-05, + "loss": 0.8363, + "step": 5883 + }, + { + "epoch": 0.2605182471248464, + "grad_norm": 1.3696740865707397, + "learning_rate": 1.8471350611300318e-05, + "loss": 0.9204, + "step": 5884 + }, + { + "epoch": 0.2605625228296603, + "grad_norm": 1.2936042547225952, + "learning_rate": 1.8470529292080963e-05, + "loss": 0.8466, + "step": 5885 + }, + { + "epoch": 0.26060679853447416, + "grad_norm": 1.171306848526001, + "learning_rate": 1.8469707770549917e-05, + "loss": 0.8219, + "step": 5886 + }, + { + "epoch": 0.26065107423928807, + "grad_norm": 1.213206171989441, + "learning_rate": 1.8468886046726797e-05, + "loss": 0.7934, + "step": 5887 + }, + { + "epoch": 0.26069534994410193, + "grad_norm": 1.3081125020980835, + "learning_rate": 1.846806412063123e-05, + "loss": 0.8337, + "step": 5888 + }, + { + "epoch": 0.2607396256489158, + "grad_norm": 1.4142347574234009, + "learning_rate": 1.8467241992282842e-05, + "loss": 0.9337, + "step": 5889 + }, + { + "epoch": 0.2607839013537297, + "grad_norm": 1.5187071561813354, + "learning_rate": 1.846641966170128e-05, + "loss": 0.9036, + "step": 5890 + }, + { + "epoch": 0.26082817705854355, + "grad_norm": 1.5402506589889526, + "learning_rate": 1.8465597128906173e-05, + "loss": 0.8473, + "step": 5891 + }, + { + "epoch": 0.2608724527633574, + "grad_norm": 1.6417911052703857, + "learning_rate": 1.8464774393917174e-05, + "loss": 0.9519, + "step": 5892 + }, + { + "epoch": 0.2609167284681713, + "grad_norm": 1.2889267206192017, + "learning_rate": 1.846395145675393e-05, + "loss": 0.8704, + "step": 5893 + }, + { + "epoch": 0.2609610041729852, + "grad_norm": 1.4071050882339478, + "learning_rate": 1.8463128317436093e-05, + "loss": 0.9021, + "step": 5894 + }, + { + "epoch": 0.26100527987779903, + "grad_norm": 1.1630135774612427, + "learning_rate": 1.8462304975983334e-05, + "loss": 0.7721, + "step": 5895 + }, + { + "epoch": 0.26104955558261295, + "grad_norm": 1.3293492794036865, + "learning_rate": 1.8461481432415308e-05, + "loss": 0.8575, + "step": 5896 + }, + { + "epoch": 0.2610938312874268, + "grad_norm": 1.3646389245986938, + "learning_rate": 1.8460657686751688e-05, + "loss": 0.8792, + "step": 5897 + }, + { + "epoch": 0.26113810699224066, + "grad_norm": 1.6144193410873413, + "learning_rate": 1.8459833739012146e-05, + "loss": 0.9896, + "step": 5898 + }, + { + "epoch": 0.26118238269705457, + "grad_norm": 1.4388794898986816, + "learning_rate": 1.8459009589216364e-05, + "loss": 0.9003, + "step": 5899 + }, + { + "epoch": 0.26122665840186843, + "grad_norm": 1.204961895942688, + "learning_rate": 1.845818523738403e-05, + "loss": 0.8353, + "step": 5900 + }, + { + "epoch": 0.2612709341066823, + "grad_norm": 1.5997463464736938, + "learning_rate": 1.8457360683534824e-05, + "loss": 0.8049, + "step": 5901 + }, + { + "epoch": 0.2613152098114962, + "grad_norm": 1.1841741800308228, + "learning_rate": 1.8456535927688445e-05, + "loss": 0.798, + "step": 5902 + }, + { + "epoch": 0.26135948551631005, + "grad_norm": 1.601127028465271, + "learning_rate": 1.845571096986459e-05, + "loss": 0.8965, + "step": 5903 + }, + { + "epoch": 0.26140376122112396, + "grad_norm": 1.29355788230896, + "learning_rate": 1.845488581008297e-05, + "loss": 0.8417, + "step": 5904 + }, + { + "epoch": 0.2614480369259378, + "grad_norm": 1.3566256761550903, + "learning_rate": 1.8454060448363278e-05, + "loss": 0.8311, + "step": 5905 + }, + { + "epoch": 0.2614923126307517, + "grad_norm": 1.61092209815979, + "learning_rate": 1.845323488472524e-05, + "loss": 0.943, + "step": 5906 + }, + { + "epoch": 0.2615365883355656, + "grad_norm": 1.4278560876846313, + "learning_rate": 1.845240911918857e-05, + "loss": 0.9328, + "step": 5907 + }, + { + "epoch": 0.26158086404037945, + "grad_norm": 1.355852723121643, + "learning_rate": 1.8451583151772988e-05, + "loss": 0.8224, + "step": 5908 + }, + { + "epoch": 0.2616251397451933, + "grad_norm": 1.362702488899231, + "learning_rate": 1.8450756982498224e-05, + "loss": 0.8283, + "step": 5909 + }, + { + "epoch": 0.2616694154500072, + "grad_norm": 1.3816314935684204, + "learning_rate": 1.8449930611384014e-05, + "loss": 0.819, + "step": 5910 + }, + { + "epoch": 0.26171369115482107, + "grad_norm": 1.8677376508712769, + "learning_rate": 1.844910403845009e-05, + "loss": 0.9225, + "step": 5911 + }, + { + "epoch": 0.2617579668596349, + "grad_norm": 1.3282057046890259, + "learning_rate": 1.844827726371619e-05, + "loss": 0.8677, + "step": 5912 + }, + { + "epoch": 0.26180224256444884, + "grad_norm": 1.380537509918213, + "learning_rate": 1.8447450287202075e-05, + "loss": 0.9053, + "step": 5913 + }, + { + "epoch": 0.2618465182692627, + "grad_norm": 1.1882619857788086, + "learning_rate": 1.844662310892748e-05, + "loss": 0.8492, + "step": 5914 + }, + { + "epoch": 0.26189079397407655, + "grad_norm": 1.5291812419891357, + "learning_rate": 1.844579572891218e-05, + "loss": 0.9452, + "step": 5915 + }, + { + "epoch": 0.26193506967889046, + "grad_norm": 1.225730299949646, + "learning_rate": 1.8444968147175915e-05, + "loss": 0.9408, + "step": 5916 + }, + { + "epoch": 0.2619793453837043, + "grad_norm": 1.2429602146148682, + "learning_rate": 1.8444140363738466e-05, + "loss": 0.8256, + "step": 5917 + }, + { + "epoch": 0.2620236210885182, + "grad_norm": 1.2987174987792969, + "learning_rate": 1.84433123786196e-05, + "loss": 0.7818, + "step": 5918 + }, + { + "epoch": 0.2620678967933321, + "grad_norm": 1.850854516029358, + "learning_rate": 1.8442484191839093e-05, + "loss": 0.8602, + "step": 5919 + }, + { + "epoch": 0.26211217249814595, + "grad_norm": 1.3694976568222046, + "learning_rate": 1.8441655803416724e-05, + "loss": 0.9255, + "step": 5920 + }, + { + "epoch": 0.26215644820295986, + "grad_norm": 1.2662080526351929, + "learning_rate": 1.844082721337228e-05, + "loss": 0.9044, + "step": 5921 + }, + { + "epoch": 0.2622007239077737, + "grad_norm": 1.41648268699646, + "learning_rate": 1.8439998421725555e-05, + "loss": 0.8961, + "step": 5922 + }, + { + "epoch": 0.26224499961258757, + "grad_norm": 1.3676776885986328, + "learning_rate": 1.8439169428496336e-05, + "loss": 0.8749, + "step": 5923 + }, + { + "epoch": 0.2622892753174015, + "grad_norm": 1.4341684579849243, + "learning_rate": 1.8438340233704425e-05, + "loss": 0.9237, + "step": 5924 + }, + { + "epoch": 0.26233355102221534, + "grad_norm": 1.2643781900405884, + "learning_rate": 1.843751083736963e-05, + "loss": 0.8676, + "step": 5925 + }, + { + "epoch": 0.2623778267270292, + "grad_norm": 1.237579584121704, + "learning_rate": 1.8436681239511763e-05, + "loss": 0.7822, + "step": 5926 + }, + { + "epoch": 0.2624221024318431, + "grad_norm": 1.2913718223571777, + "learning_rate": 1.843585144015063e-05, + "loss": 0.892, + "step": 5927 + }, + { + "epoch": 0.26246637813665696, + "grad_norm": 1.157782793045044, + "learning_rate": 1.843502143930605e-05, + "loss": 0.8896, + "step": 5928 + }, + { + "epoch": 0.2625106538414708, + "grad_norm": 1.096404790878296, + "learning_rate": 1.8434191236997856e-05, + "loss": 0.818, + "step": 5929 + }, + { + "epoch": 0.26255492954628473, + "grad_norm": 1.4453538656234741, + "learning_rate": 1.8433360833245873e-05, + "loss": 0.8198, + "step": 5930 + }, + { + "epoch": 0.2625992052510986, + "grad_norm": 1.4524837732315063, + "learning_rate": 1.843253022806993e-05, + "loss": 0.9148, + "step": 5931 + }, + { + "epoch": 0.26264348095591245, + "grad_norm": 1.1550700664520264, + "learning_rate": 1.843169942148987e-05, + "loss": 0.8504, + "step": 5932 + }, + { + "epoch": 0.26268775666072636, + "grad_norm": 1.331457495689392, + "learning_rate": 1.8430868413525533e-05, + "loss": 0.899, + "step": 5933 + }, + { + "epoch": 0.2627320323655402, + "grad_norm": 1.1575108766555786, + "learning_rate": 1.843003720419677e-05, + "loss": 0.8028, + "step": 5934 + }, + { + "epoch": 0.26277630807035407, + "grad_norm": 1.4180632829666138, + "learning_rate": 1.842920579352343e-05, + "loss": 0.9141, + "step": 5935 + }, + { + "epoch": 0.262820583775168, + "grad_norm": 1.4587786197662354, + "learning_rate": 1.8428374181525375e-05, + "loss": 0.8171, + "step": 5936 + }, + { + "epoch": 0.26286485947998184, + "grad_norm": 1.6088213920593262, + "learning_rate": 1.8427542368222465e-05, + "loss": 0.9771, + "step": 5937 + }, + { + "epoch": 0.26290913518479575, + "grad_norm": 1.2111810445785522, + "learning_rate": 1.842671035363457e-05, + "loss": 0.8179, + "step": 5938 + }, + { + "epoch": 0.2629534108896096, + "grad_norm": 1.7709400653839111, + "learning_rate": 1.8425878137781556e-05, + "loss": 0.9176, + "step": 5939 + }, + { + "epoch": 0.26299768659442346, + "grad_norm": 1.390366792678833, + "learning_rate": 1.8425045720683306e-05, + "loss": 0.8884, + "step": 5940 + }, + { + "epoch": 0.2630419622992374, + "grad_norm": 1.537272572517395, + "learning_rate": 1.8424213102359698e-05, + "loss": 0.9845, + "step": 5941 + }, + { + "epoch": 0.26308623800405123, + "grad_norm": 1.4900668859481812, + "learning_rate": 1.8423380282830616e-05, + "loss": 0.8886, + "step": 5942 + }, + { + "epoch": 0.2631305137088651, + "grad_norm": 1.3694180250167847, + "learning_rate": 1.842254726211596e-05, + "loss": 0.8342, + "step": 5943 + }, + { + "epoch": 0.263174789413679, + "grad_norm": 1.2850300073623657, + "learning_rate": 1.8421714040235615e-05, + "loss": 0.8707, + "step": 5944 + }, + { + "epoch": 0.26321906511849286, + "grad_norm": 1.3161813020706177, + "learning_rate": 1.842088061720949e-05, + "loss": 0.8802, + "step": 5945 + }, + { + "epoch": 0.2632633408233067, + "grad_norm": 1.649587631225586, + "learning_rate": 1.842004699305749e-05, + "loss": 0.9443, + "step": 5946 + }, + { + "epoch": 0.2633076165281206, + "grad_norm": 1.3290201425552368, + "learning_rate": 1.8419213167799518e-05, + "loss": 0.8776, + "step": 5947 + }, + { + "epoch": 0.2633518922329345, + "grad_norm": 1.5497734546661377, + "learning_rate": 1.84183791414555e-05, + "loss": 0.78, + "step": 5948 + }, + { + "epoch": 0.26339616793774834, + "grad_norm": 1.3169342279434204, + "learning_rate": 1.841754491404535e-05, + "loss": 0.8441, + "step": 5949 + }, + { + "epoch": 0.26344044364256225, + "grad_norm": 1.3395451307296753, + "learning_rate": 1.841671048558899e-05, + "loss": 0.8153, + "step": 5950 + }, + { + "epoch": 0.2634847193473761, + "grad_norm": 1.4381259679794312, + "learning_rate": 1.8415875856106357e-05, + "loss": 0.9158, + "step": 5951 + }, + { + "epoch": 0.26352899505218996, + "grad_norm": 1.726354956626892, + "learning_rate": 1.841504102561738e-05, + "loss": 0.9988, + "step": 5952 + }, + { + "epoch": 0.2635732707570039, + "grad_norm": 1.3728145360946655, + "learning_rate": 1.8414205994142002e-05, + "loss": 0.8278, + "step": 5953 + }, + { + "epoch": 0.26361754646181773, + "grad_norm": 1.437577247619629, + "learning_rate": 1.841337076170016e-05, + "loss": 0.8205, + "step": 5954 + }, + { + "epoch": 0.2636618221666316, + "grad_norm": 1.2720633745193481, + "learning_rate": 1.8412535328311813e-05, + "loss": 0.8505, + "step": 5955 + }, + { + "epoch": 0.2637060978714455, + "grad_norm": 1.5520761013031006, + "learning_rate": 1.841169969399691e-05, + "loss": 0.8691, + "step": 5956 + }, + { + "epoch": 0.26375037357625936, + "grad_norm": 1.423830270767212, + "learning_rate": 1.8410863858775405e-05, + "loss": 0.8921, + "step": 5957 + }, + { + "epoch": 0.26379464928107327, + "grad_norm": 1.4933242797851562, + "learning_rate": 1.8410027822667267e-05, + "loss": 0.8528, + "step": 5958 + }, + { + "epoch": 0.2638389249858871, + "grad_norm": 2.1198623180389404, + "learning_rate": 1.840919158569246e-05, + "loss": 0.8898, + "step": 5959 + }, + { + "epoch": 0.263883200690701, + "grad_norm": 1.106949806213379, + "learning_rate": 1.840835514787096e-05, + "loss": 0.8186, + "step": 5960 + }, + { + "epoch": 0.2639274763955149, + "grad_norm": 1.1304720640182495, + "learning_rate": 1.8407518509222742e-05, + "loss": 0.7372, + "step": 5961 + }, + { + "epoch": 0.26397175210032875, + "grad_norm": 1.4521783590316772, + "learning_rate": 1.8406681669767795e-05, + "loss": 0.8665, + "step": 5962 + }, + { + "epoch": 0.2640160278051426, + "grad_norm": 1.422884464263916, + "learning_rate": 1.8405844629526097e-05, + "loss": 0.7543, + "step": 5963 + }, + { + "epoch": 0.2640603035099565, + "grad_norm": 2.7377965450286865, + "learning_rate": 1.8405007388517645e-05, + "loss": 0.8618, + "step": 5964 + }, + { + "epoch": 0.2641045792147704, + "grad_norm": 1.4528193473815918, + "learning_rate": 1.8404169946762435e-05, + "loss": 0.9026, + "step": 5965 + }, + { + "epoch": 0.26414885491958423, + "grad_norm": 1.354536771774292, + "learning_rate": 1.8403332304280474e-05, + "loss": 0.9074, + "step": 5966 + }, + { + "epoch": 0.26419313062439814, + "grad_norm": 1.1873544454574585, + "learning_rate": 1.8402494461091755e-05, + "loss": 0.8564, + "step": 5967 + }, + { + "epoch": 0.264237406329212, + "grad_norm": 1.5926930904388428, + "learning_rate": 1.8401656417216304e-05, + "loss": 0.9553, + "step": 5968 + }, + { + "epoch": 0.26428168203402586, + "grad_norm": 1.3836791515350342, + "learning_rate": 1.8400818172674125e-05, + "loss": 0.8976, + "step": 5969 + }, + { + "epoch": 0.26432595773883977, + "grad_norm": 1.1704432964324951, + "learning_rate": 1.8399979727485247e-05, + "loss": 0.7457, + "step": 5970 + }, + { + "epoch": 0.2643702334436536, + "grad_norm": 1.52936851978302, + "learning_rate": 1.8399141081669694e-05, + "loss": 0.9631, + "step": 5971 + }, + { + "epoch": 0.2644145091484675, + "grad_norm": 1.6140645742416382, + "learning_rate": 1.8398302235247492e-05, + "loss": 0.9326, + "step": 5972 + }, + { + "epoch": 0.2644587848532814, + "grad_norm": 1.2897623777389526, + "learning_rate": 1.8397463188238683e-05, + "loss": 0.9078, + "step": 5973 + }, + { + "epoch": 0.26450306055809525, + "grad_norm": 1.3157912492752075, + "learning_rate": 1.83966239406633e-05, + "loss": 0.8578, + "step": 5974 + }, + { + "epoch": 0.26454733626290916, + "grad_norm": 1.6416995525360107, + "learning_rate": 1.8395784492541392e-05, + "loss": 0.8383, + "step": 5975 + }, + { + "epoch": 0.264591611967723, + "grad_norm": 1.294754147529602, + "learning_rate": 1.8394944843893006e-05, + "loss": 0.8968, + "step": 5976 + }, + { + "epoch": 0.2646358876725369, + "grad_norm": 1.341489553451538, + "learning_rate": 1.83941049947382e-05, + "loss": 0.862, + "step": 5977 + }, + { + "epoch": 0.2646801633773508, + "grad_norm": 1.4054518938064575, + "learning_rate": 1.839326494509703e-05, + "loss": 0.8087, + "step": 5978 + }, + { + "epoch": 0.26472443908216464, + "grad_norm": 1.451720952987671, + "learning_rate": 1.839242469498956e-05, + "loss": 0.9896, + "step": 5979 + }, + { + "epoch": 0.2647687147869785, + "grad_norm": 1.294365644454956, + "learning_rate": 1.839158424443586e-05, + "loss": 0.8855, + "step": 5980 + }, + { + "epoch": 0.2648129904917924, + "grad_norm": 1.2268235683441162, + "learning_rate": 1.8390743593456007e-05, + "loss": 0.8205, + "step": 5981 + }, + { + "epoch": 0.26485726619660627, + "grad_norm": 1.3232208490371704, + "learning_rate": 1.8389902742070068e-05, + "loss": 0.9088, + "step": 5982 + }, + { + "epoch": 0.2649015419014201, + "grad_norm": 1.1271445751190186, + "learning_rate": 1.838906169029814e-05, + "loss": 0.8294, + "step": 5983 + }, + { + "epoch": 0.26494581760623404, + "grad_norm": 1.4250056743621826, + "learning_rate": 1.83882204381603e-05, + "loss": 0.9359, + "step": 5984 + }, + { + "epoch": 0.2649900933110479, + "grad_norm": 1.5126270055770874, + "learning_rate": 1.8387378985676648e-05, + "loss": 0.8971, + "step": 5985 + }, + { + "epoch": 0.26503436901586175, + "grad_norm": 1.3187401294708252, + "learning_rate": 1.8386537332867274e-05, + "loss": 0.8087, + "step": 5986 + }, + { + "epoch": 0.26507864472067566, + "grad_norm": 1.3069994449615479, + "learning_rate": 1.8385695479752284e-05, + "loss": 0.813, + "step": 5987 + }, + { + "epoch": 0.2651229204254895, + "grad_norm": 1.315000295639038, + "learning_rate": 1.8384853426351793e-05, + "loss": 0.8807, + "step": 5988 + }, + { + "epoch": 0.2651671961303034, + "grad_norm": 1.0911411046981812, + "learning_rate": 1.8384011172685897e-05, + "loss": 0.8307, + "step": 5989 + }, + { + "epoch": 0.2652114718351173, + "grad_norm": 1.3469657897949219, + "learning_rate": 1.8383168718774723e-05, + "loss": 0.8423, + "step": 5990 + }, + { + "epoch": 0.26525574753993114, + "grad_norm": 1.3595324754714966, + "learning_rate": 1.8382326064638395e-05, + "loss": 0.8703, + "step": 5991 + }, + { + "epoch": 0.26530002324474505, + "grad_norm": 1.6372472047805786, + "learning_rate": 1.8381483210297027e-05, + "loss": 0.8165, + "step": 5992 + }, + { + "epoch": 0.2653442989495589, + "grad_norm": 1.5584386587142944, + "learning_rate": 1.8380640155770763e-05, + "loss": 1.0098, + "step": 5993 + }, + { + "epoch": 0.26538857465437277, + "grad_norm": 1.4306252002716064, + "learning_rate": 1.8379796901079726e-05, + "loss": 0.9097, + "step": 5994 + }, + { + "epoch": 0.2654328503591867, + "grad_norm": 1.2622581720352173, + "learning_rate": 1.8378953446244068e-05, + "loss": 0.916, + "step": 5995 + }, + { + "epoch": 0.26547712606400053, + "grad_norm": 2.9827163219451904, + "learning_rate": 1.8378109791283926e-05, + "loss": 0.9, + "step": 5996 + }, + { + "epoch": 0.2655214017688144, + "grad_norm": 1.385874629020691, + "learning_rate": 1.8377265936219457e-05, + "loss": 0.8844, + "step": 5997 + }, + { + "epoch": 0.2655656774736283, + "grad_norm": 1.5398478507995605, + "learning_rate": 1.837642188107081e-05, + "loss": 0.874, + "step": 5998 + }, + { + "epoch": 0.26560995317844216, + "grad_norm": 1.9834108352661133, + "learning_rate": 1.8375577625858148e-05, + "loss": 0.8142, + "step": 5999 + }, + { + "epoch": 0.265654228883256, + "grad_norm": 1.2038333415985107, + "learning_rate": 1.8374733170601633e-05, + "loss": 0.8465, + "step": 6000 + }, + { + "epoch": 0.26569850458806993, + "grad_norm": 1.5987657308578491, + "learning_rate": 1.8373888515321436e-05, + "loss": 0.8813, + "step": 6001 + }, + { + "epoch": 0.2657427802928838, + "grad_norm": 1.3323988914489746, + "learning_rate": 1.8373043660037728e-05, + "loss": 0.8555, + "step": 6002 + }, + { + "epoch": 0.26578705599769764, + "grad_norm": 1.4662667512893677, + "learning_rate": 1.837219860477069e-05, + "loss": 0.9578, + "step": 6003 + }, + { + "epoch": 0.26583133170251155, + "grad_norm": 1.6132569313049316, + "learning_rate": 1.837135334954051e-05, + "loss": 0.9167, + "step": 6004 + }, + { + "epoch": 0.2658756074073254, + "grad_norm": 1.6478800773620605, + "learning_rate": 1.8370507894367366e-05, + "loss": 0.9356, + "step": 6005 + }, + { + "epoch": 0.26591988311213927, + "grad_norm": 1.3219707012176514, + "learning_rate": 1.8369662239271457e-05, + "loss": 0.8146, + "step": 6006 + }, + { + "epoch": 0.2659641588169532, + "grad_norm": 1.2653261423110962, + "learning_rate": 1.8368816384272983e-05, + "loss": 0.8459, + "step": 6007 + }, + { + "epoch": 0.26600843452176703, + "grad_norm": 1.3102774620056152, + "learning_rate": 1.8367970329392142e-05, + "loss": 0.8319, + "step": 6008 + }, + { + "epoch": 0.26605271022658095, + "grad_norm": 1.4108034372329712, + "learning_rate": 1.8367124074649144e-05, + "loss": 0.9275, + "step": 6009 + }, + { + "epoch": 0.2660969859313948, + "grad_norm": 1.349257469177246, + "learning_rate": 1.83662776200642e-05, + "loss": 0.9201, + "step": 6010 + }, + { + "epoch": 0.26614126163620866, + "grad_norm": 1.2246309518814087, + "learning_rate": 1.8365430965657527e-05, + "loss": 0.819, + "step": 6011 + }, + { + "epoch": 0.26618553734102257, + "grad_norm": 1.2133084535598755, + "learning_rate": 1.8364584111449346e-05, + "loss": 0.7933, + "step": 6012 + }, + { + "epoch": 0.2662298130458364, + "grad_norm": 1.256778597831726, + "learning_rate": 1.8363737057459886e-05, + "loss": 0.7929, + "step": 6013 + }, + { + "epoch": 0.2662740887506503, + "grad_norm": 1.532798171043396, + "learning_rate": 1.8362889803709374e-05, + "loss": 0.965, + "step": 6014 + }, + { + "epoch": 0.2663183644554642, + "grad_norm": 1.4369920492172241, + "learning_rate": 1.8362042350218047e-05, + "loss": 0.9407, + "step": 6015 + }, + { + "epoch": 0.26636264016027805, + "grad_norm": 1.3510459661483765, + "learning_rate": 1.836119469700615e-05, + "loss": 0.9142, + "step": 6016 + }, + { + "epoch": 0.2664069158650919, + "grad_norm": 1.2696195840835571, + "learning_rate": 1.8360346844093926e-05, + "loss": 0.8732, + "step": 6017 + }, + { + "epoch": 0.2664511915699058, + "grad_norm": 1.3598313331604004, + "learning_rate": 1.8359498791501624e-05, + "loss": 0.7957, + "step": 6018 + }, + { + "epoch": 0.2664954672747197, + "grad_norm": 1.1299959421157837, + "learning_rate": 1.83586505392495e-05, + "loss": 0.7657, + "step": 6019 + }, + { + "epoch": 0.26653974297953353, + "grad_norm": 1.3559153079986572, + "learning_rate": 1.8357802087357816e-05, + "loss": 0.8816, + "step": 6020 + }, + { + "epoch": 0.26658401868434745, + "grad_norm": 1.450807809829712, + "learning_rate": 1.835695343584683e-05, + "loss": 0.9287, + "step": 6021 + }, + { + "epoch": 0.2666282943891613, + "grad_norm": 1.377292513847351, + "learning_rate": 1.8356104584736818e-05, + "loss": 0.9103, + "step": 6022 + }, + { + "epoch": 0.26667257009397516, + "grad_norm": 1.1805576086044312, + "learning_rate": 1.8355255534048053e-05, + "loss": 0.8139, + "step": 6023 + }, + { + "epoch": 0.26671684579878907, + "grad_norm": 1.4039820432662964, + "learning_rate": 1.8354406283800813e-05, + "loss": 0.8379, + "step": 6024 + }, + { + "epoch": 0.2667611215036029, + "grad_norm": 1.3049671649932861, + "learning_rate": 1.8353556834015375e-05, + "loss": 0.8176, + "step": 6025 + }, + { + "epoch": 0.26680539720841684, + "grad_norm": 1.3867892026901245, + "learning_rate": 1.8352707184712038e-05, + "loss": 0.9002, + "step": 6026 + }, + { + "epoch": 0.2668496729132307, + "grad_norm": 1.4089486598968506, + "learning_rate": 1.8351857335911094e-05, + "loss": 0.898, + "step": 6027 + }, + { + "epoch": 0.26689394861804455, + "grad_norm": 1.4915204048156738, + "learning_rate": 1.8351007287632833e-05, + "loss": 0.9221, + "step": 6028 + }, + { + "epoch": 0.26693822432285846, + "grad_norm": 1.3701924085617065, + "learning_rate": 1.8350157039897568e-05, + "loss": 0.872, + "step": 6029 + }, + { + "epoch": 0.2669825000276723, + "grad_norm": 1.1886346340179443, + "learning_rate": 1.8349306592725596e-05, + "loss": 0.8356, + "step": 6030 + }, + { + "epoch": 0.2670267757324862, + "grad_norm": 1.0505841970443726, + "learning_rate": 1.8348455946137234e-05, + "loss": 0.7549, + "step": 6031 + }, + { + "epoch": 0.2670710514373001, + "grad_norm": 1.1945358514785767, + "learning_rate": 1.8347605100152804e-05, + "loss": 0.8876, + "step": 6032 + }, + { + "epoch": 0.26711532714211395, + "grad_norm": 1.5785008668899536, + "learning_rate": 1.834675405479262e-05, + "loss": 0.9215, + "step": 6033 + }, + { + "epoch": 0.2671596028469278, + "grad_norm": 1.1514856815338135, + "learning_rate": 1.834590281007701e-05, + "loss": 0.8185, + "step": 6034 + }, + { + "epoch": 0.2672038785517417, + "grad_norm": 1.457423448562622, + "learning_rate": 1.834505136602631e-05, + "loss": 0.8895, + "step": 6035 + }, + { + "epoch": 0.26724815425655557, + "grad_norm": 1.6537073850631714, + "learning_rate": 1.834419972266085e-05, + "loss": 0.8974, + "step": 6036 + }, + { + "epoch": 0.2672924299613694, + "grad_norm": 1.2129484415054321, + "learning_rate": 1.834334788000098e-05, + "loss": 0.8195, + "step": 6037 + }, + { + "epoch": 0.26733670566618334, + "grad_norm": 1.3637232780456543, + "learning_rate": 1.8342495838067035e-05, + "loss": 0.8827, + "step": 6038 + }, + { + "epoch": 0.2673809813709972, + "grad_norm": 1.1518232822418213, + "learning_rate": 1.834164359687937e-05, + "loss": 0.7325, + "step": 6039 + }, + { + "epoch": 0.26742525707581105, + "grad_norm": 1.7608956098556519, + "learning_rate": 1.8340791156458337e-05, + "loss": 0.9541, + "step": 6040 + }, + { + "epoch": 0.26746953278062496, + "grad_norm": 1.1778202056884766, + "learning_rate": 1.83399385168243e-05, + "loss": 0.7803, + "step": 6041 + }, + { + "epoch": 0.2675138084854388, + "grad_norm": 1.4080256223678589, + "learning_rate": 1.833908567799762e-05, + "loss": 0.8924, + "step": 6042 + }, + { + "epoch": 0.2675580841902527, + "grad_norm": 1.296614408493042, + "learning_rate": 1.8338232639998672e-05, + "loss": 0.8456, + "step": 6043 + }, + { + "epoch": 0.2676023598950666, + "grad_norm": 1.4299254417419434, + "learning_rate": 1.8337379402847827e-05, + "loss": 0.8686, + "step": 6044 + }, + { + "epoch": 0.26764663559988044, + "grad_norm": 1.319920539855957, + "learning_rate": 1.8336525966565462e-05, + "loss": 0.8584, + "step": 6045 + }, + { + "epoch": 0.26769091130469436, + "grad_norm": 1.1327378749847412, + "learning_rate": 1.8335672331171964e-05, + "loss": 0.8373, + "step": 6046 + }, + { + "epoch": 0.2677351870095082, + "grad_norm": 1.2553349733352661, + "learning_rate": 1.8334818496687718e-05, + "loss": 0.8017, + "step": 6047 + }, + { + "epoch": 0.26777946271432207, + "grad_norm": 1.40995454788208, + "learning_rate": 1.833396446313312e-05, + "loss": 0.8012, + "step": 6048 + }, + { + "epoch": 0.267823738419136, + "grad_norm": 1.6288636922836304, + "learning_rate": 1.8333110230528565e-05, + "loss": 0.9552, + "step": 6049 + }, + { + "epoch": 0.26786801412394984, + "grad_norm": 1.2771302461624146, + "learning_rate": 1.8332255798894456e-05, + "loss": 0.8706, + "step": 6050 + }, + { + "epoch": 0.2679122898287637, + "grad_norm": 1.1638520956039429, + "learning_rate": 1.83314011682512e-05, + "loss": 0.8282, + "step": 6051 + }, + { + "epoch": 0.2679565655335776, + "grad_norm": 1.1771738529205322, + "learning_rate": 1.8330546338619218e-05, + "loss": 0.848, + "step": 6052 + }, + { + "epoch": 0.26800084123839146, + "grad_norm": 1.4006285667419434, + "learning_rate": 1.8329691310018914e-05, + "loss": 0.8795, + "step": 6053 + }, + { + "epoch": 0.2680451169432053, + "grad_norm": 1.5653859376907349, + "learning_rate": 1.8328836082470717e-05, + "loss": 0.9275, + "step": 6054 + }, + { + "epoch": 0.26808939264801923, + "grad_norm": 1.40843665599823, + "learning_rate": 1.8327980655995054e-05, + "loss": 0.9529, + "step": 6055 + }, + { + "epoch": 0.2681336683528331, + "grad_norm": 1.4725487232208252, + "learning_rate": 1.8327125030612348e-05, + "loss": 0.8851, + "step": 6056 + }, + { + "epoch": 0.26817794405764694, + "grad_norm": 1.6546911001205444, + "learning_rate": 1.8326269206343046e-05, + "loss": 0.9413, + "step": 6057 + }, + { + "epoch": 0.26822221976246086, + "grad_norm": 1.3674513101577759, + "learning_rate": 1.8325413183207583e-05, + "loss": 0.9106, + "step": 6058 + }, + { + "epoch": 0.2682664954672747, + "grad_norm": 1.5107524394989014, + "learning_rate": 1.8324556961226406e-05, + "loss": 0.8374, + "step": 6059 + }, + { + "epoch": 0.26831077117208857, + "grad_norm": 1.4195363521575928, + "learning_rate": 1.832370054041996e-05, + "loss": 0.883, + "step": 6060 + }, + { + "epoch": 0.2683550468769025, + "grad_norm": 1.1654108762741089, + "learning_rate": 1.8322843920808705e-05, + "loss": 0.867, + "step": 6061 + }, + { + "epoch": 0.26839932258171634, + "grad_norm": 1.3710378408432007, + "learning_rate": 1.83219871024131e-05, + "loss": 0.8923, + "step": 6062 + }, + { + "epoch": 0.26844359828653025, + "grad_norm": 1.4670029878616333, + "learning_rate": 1.832113008525361e-05, + "loss": 0.8781, + "step": 6063 + }, + { + "epoch": 0.2684878739913441, + "grad_norm": 1.2073038816452026, + "learning_rate": 1.8320272869350705e-05, + "loss": 0.7695, + "step": 6064 + }, + { + "epoch": 0.26853214969615796, + "grad_norm": 1.3695435523986816, + "learning_rate": 1.8319415454724858e-05, + "loss": 0.8452, + "step": 6065 + }, + { + "epoch": 0.2685764254009719, + "grad_norm": 1.6030850410461426, + "learning_rate": 1.831855784139654e-05, + "loss": 0.8698, + "step": 6066 + }, + { + "epoch": 0.26862070110578573, + "grad_norm": 1.0975784063339233, + "learning_rate": 1.8317700029386245e-05, + "loss": 0.7919, + "step": 6067 + }, + { + "epoch": 0.2686649768105996, + "grad_norm": 1.3199127912521362, + "learning_rate": 1.8316842018714456e-05, + "loss": 0.8359, + "step": 6068 + }, + { + "epoch": 0.2687092525154135, + "grad_norm": 1.3805556297302246, + "learning_rate": 1.831598380940167e-05, + "loss": 0.886, + "step": 6069 + }, + { + "epoch": 0.26875352822022736, + "grad_norm": 1.2170711755752563, + "learning_rate": 1.831512540146838e-05, + "loss": 0.8522, + "step": 6070 + }, + { + "epoch": 0.2687978039250412, + "grad_norm": 1.2080605030059814, + "learning_rate": 1.831426679493509e-05, + "loss": 0.7662, + "step": 6071 + }, + { + "epoch": 0.2688420796298551, + "grad_norm": 1.234630823135376, + "learning_rate": 1.8313407989822308e-05, + "loss": 0.8021, + "step": 6072 + }, + { + "epoch": 0.268886355334669, + "grad_norm": 1.269970417022705, + "learning_rate": 1.8312548986150546e-05, + "loss": 0.8674, + "step": 6073 + }, + { + "epoch": 0.26893063103948284, + "grad_norm": 1.8657011985778809, + "learning_rate": 1.831168978394032e-05, + "loss": 0.994, + "step": 6074 + }, + { + "epoch": 0.26897490674429675, + "grad_norm": 1.2708783149719238, + "learning_rate": 1.831083038321215e-05, + "loss": 0.8499, + "step": 6075 + }, + { + "epoch": 0.2690191824491106, + "grad_norm": 1.3472795486450195, + "learning_rate": 1.830997078398656e-05, + "loss": 0.8197, + "step": 6076 + }, + { + "epoch": 0.26906345815392446, + "grad_norm": 1.7504832744598389, + "learning_rate": 1.830911098628409e-05, + "loss": 0.7867, + "step": 6077 + }, + { + "epoch": 0.2691077338587384, + "grad_norm": 1.554781436920166, + "learning_rate": 1.8308250990125268e-05, + "loss": 0.9428, + "step": 6078 + }, + { + "epoch": 0.26915200956355223, + "grad_norm": 1.308061957359314, + "learning_rate": 1.8307390795530633e-05, + "loss": 0.8355, + "step": 6079 + }, + { + "epoch": 0.26919628526836614, + "grad_norm": 1.27203369140625, + "learning_rate": 1.8306530402520736e-05, + "loss": 0.8967, + "step": 6080 + }, + { + "epoch": 0.26924056097318, + "grad_norm": 1.5260175466537476, + "learning_rate": 1.8305669811116123e-05, + "loss": 0.8843, + "step": 6081 + }, + { + "epoch": 0.26928483667799386, + "grad_norm": 1.4904371500015259, + "learning_rate": 1.830480902133735e-05, + "loss": 0.8217, + "step": 6082 + }, + { + "epoch": 0.26932911238280777, + "grad_norm": 1.4967069625854492, + "learning_rate": 1.8303948033204973e-05, + "loss": 0.8883, + "step": 6083 + }, + { + "epoch": 0.2693733880876216, + "grad_norm": 1.3883800506591797, + "learning_rate": 1.8303086846739558e-05, + "loss": 0.824, + "step": 6084 + }, + { + "epoch": 0.2694176637924355, + "grad_norm": 1.4938170909881592, + "learning_rate": 1.8302225461961677e-05, + "loss": 0.9288, + "step": 6085 + }, + { + "epoch": 0.2694619394972494, + "grad_norm": 1.3407258987426758, + "learning_rate": 1.8301363878891902e-05, + "loss": 0.8428, + "step": 6086 + }, + { + "epoch": 0.26950621520206325, + "grad_norm": 1.5578688383102417, + "learning_rate": 1.8300502097550807e-05, + "loss": 0.915, + "step": 6087 + }, + { + "epoch": 0.2695504909068771, + "grad_norm": 1.3305729627609253, + "learning_rate": 1.829964011795898e-05, + "loss": 0.849, + "step": 6088 + }, + { + "epoch": 0.269594766611691, + "grad_norm": 1.2333577871322632, + "learning_rate": 1.8298777940137005e-05, + "loss": 0.849, + "step": 6089 + }, + { + "epoch": 0.2696390423165049, + "grad_norm": 1.429779052734375, + "learning_rate": 1.829791556410548e-05, + "loss": 0.8621, + "step": 6090 + }, + { + "epoch": 0.26968331802131873, + "grad_norm": 1.5593122243881226, + "learning_rate": 1.8297052989884993e-05, + "loss": 0.8997, + "step": 6091 + }, + { + "epoch": 0.26972759372613264, + "grad_norm": 1.6332781314849854, + "learning_rate": 1.8296190217496152e-05, + "loss": 0.9526, + "step": 6092 + }, + { + "epoch": 0.2697718694309465, + "grad_norm": 1.7650158405303955, + "learning_rate": 1.8295327246959562e-05, + "loss": 0.8777, + "step": 6093 + }, + { + "epoch": 0.26981614513576035, + "grad_norm": 1.3257074356079102, + "learning_rate": 1.8294464078295837e-05, + "loss": 0.8461, + "step": 6094 + }, + { + "epoch": 0.26986042084057427, + "grad_norm": 1.3217819929122925, + "learning_rate": 1.829360071152559e-05, + "loss": 0.7947, + "step": 6095 + }, + { + "epoch": 0.2699046965453881, + "grad_norm": 1.4739676713943481, + "learning_rate": 1.8292737146669447e-05, + "loss": 0.8947, + "step": 6096 + }, + { + "epoch": 0.26994897225020204, + "grad_norm": 1.4214210510253906, + "learning_rate": 1.8291873383748022e-05, + "loss": 0.8514, + "step": 6097 + }, + { + "epoch": 0.2699932479550159, + "grad_norm": 1.3245526552200317, + "learning_rate": 1.8291009422781957e-05, + "loss": 0.8803, + "step": 6098 + }, + { + "epoch": 0.27003752365982975, + "grad_norm": 1.6443147659301758, + "learning_rate": 1.8290145263791883e-05, + "loss": 0.9153, + "step": 6099 + }, + { + "epoch": 0.27008179936464366, + "grad_norm": 1.4748003482818604, + "learning_rate": 1.828928090679844e-05, + "loss": 0.8668, + "step": 6100 + }, + { + "epoch": 0.2701260750694575, + "grad_norm": 1.2783461809158325, + "learning_rate": 1.828841635182227e-05, + "loss": 0.8036, + "step": 6101 + }, + { + "epoch": 0.2701703507742714, + "grad_norm": 1.4888437986373901, + "learning_rate": 1.8287551598884024e-05, + "loss": 0.835, + "step": 6102 + }, + { + "epoch": 0.2702146264790853, + "grad_norm": 1.2248027324676514, + "learning_rate": 1.8286686648004356e-05, + "loss": 0.8305, + "step": 6103 + }, + { + "epoch": 0.27025890218389914, + "grad_norm": 1.3574550151824951, + "learning_rate": 1.8285821499203927e-05, + "loss": 0.8552, + "step": 6104 + }, + { + "epoch": 0.270303177888713, + "grad_norm": 1.2821521759033203, + "learning_rate": 1.8284956152503398e-05, + "loss": 0.8399, + "step": 6105 + }, + { + "epoch": 0.2703474535935269, + "grad_norm": 1.6072217226028442, + "learning_rate": 1.8284090607923436e-05, + "loss": 0.8711, + "step": 6106 + }, + { + "epoch": 0.27039172929834077, + "grad_norm": 1.4386216402053833, + "learning_rate": 1.8283224865484713e-05, + "loss": 0.9005, + "step": 6107 + }, + { + "epoch": 0.2704360050031546, + "grad_norm": 1.4085040092468262, + "learning_rate": 1.828235892520791e-05, + "loss": 0.8247, + "step": 6108 + }, + { + "epoch": 0.27048028070796853, + "grad_norm": 1.2709457874298096, + "learning_rate": 1.8281492787113707e-05, + "loss": 0.871, + "step": 6109 + }, + { + "epoch": 0.2705245564127824, + "grad_norm": 1.5796983242034912, + "learning_rate": 1.8280626451222796e-05, + "loss": 0.9151, + "step": 6110 + }, + { + "epoch": 0.27056883211759625, + "grad_norm": 1.472212553024292, + "learning_rate": 1.8279759917555862e-05, + "loss": 0.9008, + "step": 6111 + }, + { + "epoch": 0.27061310782241016, + "grad_norm": 1.233484148979187, + "learning_rate": 1.8278893186133603e-05, + "loss": 0.7721, + "step": 6112 + }, + { + "epoch": 0.270657383527224, + "grad_norm": 1.3337613344192505, + "learning_rate": 1.8278026256976723e-05, + "loss": 0.8403, + "step": 6113 + }, + { + "epoch": 0.2707016592320379, + "grad_norm": 1.2236988544464111, + "learning_rate": 1.8277159130105924e-05, + "loss": 0.819, + "step": 6114 + }, + { + "epoch": 0.2707459349368518, + "grad_norm": 1.4303427934646606, + "learning_rate": 1.8276291805541918e-05, + "loss": 0.8946, + "step": 6115 + }, + { + "epoch": 0.27079021064166564, + "grad_norm": 1.557070255279541, + "learning_rate": 1.8275424283305426e-05, + "loss": 0.9356, + "step": 6116 + }, + { + "epoch": 0.27083448634647955, + "grad_norm": 1.209783911705017, + "learning_rate": 1.8274556563417157e-05, + "loss": 0.8812, + "step": 6117 + }, + { + "epoch": 0.2708787620512934, + "grad_norm": 1.4481412172317505, + "learning_rate": 1.8273688645897843e-05, + "loss": 0.9922, + "step": 6118 + }, + { + "epoch": 0.27092303775610727, + "grad_norm": 1.4358017444610596, + "learning_rate": 1.8272820530768216e-05, + "loss": 0.9383, + "step": 6119 + }, + { + "epoch": 0.2709673134609212, + "grad_norm": 1.345977544784546, + "learning_rate": 1.8271952218049004e-05, + "loss": 0.9195, + "step": 6120 + }, + { + "epoch": 0.27101158916573503, + "grad_norm": 1.3885903358459473, + "learning_rate": 1.827108370776095e-05, + "loss": 0.9053, + "step": 6121 + }, + { + "epoch": 0.2710558648705489, + "grad_norm": 1.3605822324752808, + "learning_rate": 1.8270214999924795e-05, + "loss": 0.8758, + "step": 6122 + }, + { + "epoch": 0.2711001405753628, + "grad_norm": 1.1207191944122314, + "learning_rate": 1.826934609456129e-05, + "loss": 0.8006, + "step": 6123 + }, + { + "epoch": 0.27114441628017666, + "grad_norm": 1.2559435367584229, + "learning_rate": 1.8268476991691183e-05, + "loss": 0.8307, + "step": 6124 + }, + { + "epoch": 0.2711886919849905, + "grad_norm": 1.2913538217544556, + "learning_rate": 1.8267607691335237e-05, + "loss": 0.8318, + "step": 6125 + }, + { + "epoch": 0.2712329676898044, + "grad_norm": 1.3536993265151978, + "learning_rate": 1.8266738193514214e-05, + "loss": 0.8531, + "step": 6126 + }, + { + "epoch": 0.2712772433946183, + "grad_norm": 1.6877880096435547, + "learning_rate": 1.8265868498248883e-05, + "loss": 0.8139, + "step": 6127 + }, + { + "epoch": 0.27132151909943214, + "grad_norm": 1.258682131767273, + "learning_rate": 1.826499860556001e-05, + "loss": 0.7843, + "step": 6128 + }, + { + "epoch": 0.27136579480424605, + "grad_norm": 1.3425500392913818, + "learning_rate": 1.8264128515468375e-05, + "loss": 0.9863, + "step": 6129 + }, + { + "epoch": 0.2714100705090599, + "grad_norm": 1.378913164138794, + "learning_rate": 1.826325822799476e-05, + "loss": 0.9211, + "step": 6130 + }, + { + "epoch": 0.27145434621387377, + "grad_norm": 1.4593122005462646, + "learning_rate": 1.826238774315995e-05, + "loss": 0.9289, + "step": 6131 + }, + { + "epoch": 0.2714986219186877, + "grad_norm": 1.2498902082443237, + "learning_rate": 1.8261517060984737e-05, + "loss": 0.8323, + "step": 6132 + }, + { + "epoch": 0.27154289762350153, + "grad_norm": 1.343980073928833, + "learning_rate": 1.8260646181489913e-05, + "loss": 0.834, + "step": 6133 + }, + { + "epoch": 0.27158717332831545, + "grad_norm": 1.3817938566207886, + "learning_rate": 1.8259775104696283e-05, + "loss": 0.8476, + "step": 6134 + }, + { + "epoch": 0.2716314490331293, + "grad_norm": 1.135468602180481, + "learning_rate": 1.8258903830624652e-05, + "loss": 0.7646, + "step": 6135 + }, + { + "epoch": 0.27167572473794316, + "grad_norm": 1.6978728771209717, + "learning_rate": 1.8258032359295826e-05, + "loss": 0.9264, + "step": 6136 + }, + { + "epoch": 0.27172000044275707, + "grad_norm": 1.296384334564209, + "learning_rate": 1.825716069073062e-05, + "loss": 0.8227, + "step": 6137 + }, + { + "epoch": 0.2717642761475709, + "grad_norm": 1.3872227668762207, + "learning_rate": 1.8256288824949857e-05, + "loss": 0.8002, + "step": 6138 + }, + { + "epoch": 0.2718085518523848, + "grad_norm": 1.3274558782577515, + "learning_rate": 1.8255416761974355e-05, + "loss": 0.922, + "step": 6139 + }, + { + "epoch": 0.2718528275571987, + "grad_norm": 1.5375782251358032, + "learning_rate": 1.8254544501824947e-05, + "loss": 0.9415, + "step": 6140 + }, + { + "epoch": 0.27189710326201255, + "grad_norm": 1.6258333921432495, + "learning_rate": 1.8253672044522465e-05, + "loss": 0.9026, + "step": 6141 + }, + { + "epoch": 0.2719413789668264, + "grad_norm": 1.8301050662994385, + "learning_rate": 1.8252799390087745e-05, + "loss": 0.9645, + "step": 6142 + }, + { + "epoch": 0.2719856546716403, + "grad_norm": 1.3833962678909302, + "learning_rate": 1.8251926538541632e-05, + "loss": 0.8206, + "step": 6143 + }, + { + "epoch": 0.2720299303764542, + "grad_norm": 1.7109832763671875, + "learning_rate": 1.825105348990497e-05, + "loss": 0.8595, + "step": 6144 + }, + { + "epoch": 0.27207420608126803, + "grad_norm": 1.1918771266937256, + "learning_rate": 1.825018024419862e-05, + "loss": 0.8073, + "step": 6145 + }, + { + "epoch": 0.27211848178608194, + "grad_norm": 1.1430107355117798, + "learning_rate": 1.8249306801443426e-05, + "loss": 0.7484, + "step": 6146 + }, + { + "epoch": 0.2721627574908958, + "grad_norm": 2.391176462173462, + "learning_rate": 1.824843316166026e-05, + "loss": 0.9928, + "step": 6147 + }, + { + "epoch": 0.27220703319570966, + "grad_norm": 1.2315481901168823, + "learning_rate": 1.8247559324869982e-05, + "loss": 0.88, + "step": 6148 + }, + { + "epoch": 0.27225130890052357, + "grad_norm": 1.1365951299667358, + "learning_rate": 1.8246685291093465e-05, + "loss": 0.8366, + "step": 6149 + }, + { + "epoch": 0.2722955846053374, + "grad_norm": 1.178484320640564, + "learning_rate": 1.8245811060351586e-05, + "loss": 0.8266, + "step": 6150 + }, + { + "epoch": 0.27233986031015134, + "grad_norm": 1.51921808719635, + "learning_rate": 1.8244936632665223e-05, + "loss": 1.0005, + "step": 6151 + }, + { + "epoch": 0.2723841360149652, + "grad_norm": 1.4298666715621948, + "learning_rate": 1.8244062008055263e-05, + "loss": 0.891, + "step": 6152 + }, + { + "epoch": 0.27242841171977905, + "grad_norm": 1.2057454586029053, + "learning_rate": 1.8243187186542594e-05, + "loss": 0.8494, + "step": 6153 + }, + { + "epoch": 0.27247268742459296, + "grad_norm": 1.6589018106460571, + "learning_rate": 1.8242312168148108e-05, + "loss": 0.9233, + "step": 6154 + }, + { + "epoch": 0.2725169631294068, + "grad_norm": 1.4095057249069214, + "learning_rate": 1.824143695289271e-05, + "loss": 0.933, + "step": 6155 + }, + { + "epoch": 0.2725612388342207, + "grad_norm": 1.2114487886428833, + "learning_rate": 1.8240561540797295e-05, + "loss": 0.8176, + "step": 6156 + }, + { + "epoch": 0.2726055145390346, + "grad_norm": 1.3998326063156128, + "learning_rate": 1.8239685931882785e-05, + "loss": 0.8872, + "step": 6157 + }, + { + "epoch": 0.27264979024384844, + "grad_norm": 1.2317270040512085, + "learning_rate": 1.823881012617008e-05, + "loss": 0.8432, + "step": 6158 + }, + { + "epoch": 0.2726940659486623, + "grad_norm": 1.5382384061813354, + "learning_rate": 1.823793412368011e-05, + "loss": 0.9033, + "step": 6159 + }, + { + "epoch": 0.2727383416534762, + "grad_norm": 1.6582227945327759, + "learning_rate": 1.8237057924433786e-05, + "loss": 0.907, + "step": 6160 + }, + { + "epoch": 0.27278261735829007, + "grad_norm": 1.7307542562484741, + "learning_rate": 1.8236181528452043e-05, + "loss": 0.874, + "step": 6161 + }, + { + "epoch": 0.2728268930631039, + "grad_norm": 1.3858067989349365, + "learning_rate": 1.823530493575581e-05, + "loss": 0.9535, + "step": 6162 + }, + { + "epoch": 0.27287116876791784, + "grad_norm": 1.2268123626708984, + "learning_rate": 1.823442814636602e-05, + "loss": 0.8255, + "step": 6163 + }, + { + "epoch": 0.2729154444727317, + "grad_norm": 1.1970254182815552, + "learning_rate": 1.823355116030362e-05, + "loss": 0.826, + "step": 6164 + }, + { + "epoch": 0.27295972017754555, + "grad_norm": 1.2453930377960205, + "learning_rate": 1.8232673977589556e-05, + "loss": 0.8143, + "step": 6165 + }, + { + "epoch": 0.27300399588235946, + "grad_norm": 1.4379454851150513, + "learning_rate": 1.823179659824478e-05, + "loss": 0.8665, + "step": 6166 + }, + { + "epoch": 0.2730482715871733, + "grad_norm": 1.2928705215454102, + "learning_rate": 1.8230919022290242e-05, + "loss": 0.8062, + "step": 6167 + }, + { + "epoch": 0.27309254729198723, + "grad_norm": 2.3550901412963867, + "learning_rate": 1.8230041249746904e-05, + "loss": 0.9736, + "step": 6168 + }, + { + "epoch": 0.2731368229968011, + "grad_norm": 1.5501388311386108, + "learning_rate": 1.8229163280635734e-05, + "loss": 0.8341, + "step": 6169 + }, + { + "epoch": 0.27318109870161494, + "grad_norm": 1.1931732892990112, + "learning_rate": 1.8228285114977698e-05, + "loss": 0.7977, + "step": 6170 + }, + { + "epoch": 0.27322537440642886, + "grad_norm": 1.3564655780792236, + "learning_rate": 1.822740675279377e-05, + "loss": 0.8662, + "step": 6171 + }, + { + "epoch": 0.2732696501112427, + "grad_norm": 1.4779132604599, + "learning_rate": 1.8226528194104934e-05, + "loss": 0.8359, + "step": 6172 + }, + { + "epoch": 0.27331392581605657, + "grad_norm": 1.438344955444336, + "learning_rate": 1.822564943893217e-05, + "loss": 0.8997, + "step": 6173 + }, + { + "epoch": 0.2733582015208705, + "grad_norm": 1.2729716300964355, + "learning_rate": 1.8224770487296466e-05, + "loss": 0.8541, + "step": 6174 + }, + { + "epoch": 0.27340247722568434, + "grad_norm": 1.2422716617584229, + "learning_rate": 1.8223891339218815e-05, + "loss": 0.8786, + "step": 6175 + }, + { + "epoch": 0.2734467529304982, + "grad_norm": 1.5359243154525757, + "learning_rate": 1.8223011994720218e-05, + "loss": 0.8603, + "step": 6176 + }, + { + "epoch": 0.2734910286353121, + "grad_norm": 1.4189411401748657, + "learning_rate": 1.8222132453821673e-05, + "loss": 0.8628, + "step": 6177 + }, + { + "epoch": 0.27353530434012596, + "grad_norm": 1.5459789037704468, + "learning_rate": 1.8221252716544188e-05, + "loss": 0.8439, + "step": 6178 + }, + { + "epoch": 0.2735795800449398, + "grad_norm": 1.3877326250076294, + "learning_rate": 1.8220372782908778e-05, + "loss": 0.8823, + "step": 6179 + }, + { + "epoch": 0.27362385574975373, + "grad_norm": 1.2636698484420776, + "learning_rate": 1.8219492652936454e-05, + "loss": 0.8245, + "step": 6180 + }, + { + "epoch": 0.2736681314545676, + "grad_norm": 1.1811583042144775, + "learning_rate": 1.8218612326648246e-05, + "loss": 0.7994, + "step": 6181 + }, + { + "epoch": 0.27371240715938144, + "grad_norm": 1.1989004611968994, + "learning_rate": 1.821773180406517e-05, + "loss": 0.7947, + "step": 6182 + }, + { + "epoch": 0.27375668286419536, + "grad_norm": 1.2301181554794312, + "learning_rate": 1.8216851085208263e-05, + "loss": 0.8583, + "step": 6183 + }, + { + "epoch": 0.2738009585690092, + "grad_norm": 1.1872954368591309, + "learning_rate": 1.8215970170098555e-05, + "loss": 0.891, + "step": 6184 + }, + { + "epoch": 0.2738452342738231, + "grad_norm": 1.3078317642211914, + "learning_rate": 1.8215089058757094e-05, + "loss": 0.8751, + "step": 6185 + }, + { + "epoch": 0.273889509978637, + "grad_norm": 1.469635009765625, + "learning_rate": 1.8214207751204917e-05, + "loss": 0.891, + "step": 6186 + }, + { + "epoch": 0.27393378568345084, + "grad_norm": 1.2530300617218018, + "learning_rate": 1.8213326247463076e-05, + "loss": 0.8443, + "step": 6187 + }, + { + "epoch": 0.27397806138826475, + "grad_norm": 1.2770932912826538, + "learning_rate": 1.8212444547552627e-05, + "loss": 0.8351, + "step": 6188 + }, + { + "epoch": 0.2740223370930786, + "grad_norm": 1.3777247667312622, + "learning_rate": 1.8211562651494625e-05, + "loss": 0.8122, + "step": 6189 + }, + { + "epoch": 0.27406661279789246, + "grad_norm": 1.4943552017211914, + "learning_rate": 1.8210680559310136e-05, + "loss": 0.9086, + "step": 6190 + }, + { + "epoch": 0.2741108885027064, + "grad_norm": 1.3272377252578735, + "learning_rate": 1.8209798271020227e-05, + "loss": 0.7558, + "step": 6191 + }, + { + "epoch": 0.27415516420752023, + "grad_norm": 1.498897671699524, + "learning_rate": 1.820891578664597e-05, + "loss": 0.8728, + "step": 6192 + }, + { + "epoch": 0.2741994399123341, + "grad_norm": 1.4367777109146118, + "learning_rate": 1.8208033106208444e-05, + "loss": 0.9397, + "step": 6193 + }, + { + "epoch": 0.274243715617148, + "grad_norm": 1.2935473918914795, + "learning_rate": 1.8207150229728733e-05, + "loss": 0.8325, + "step": 6194 + }, + { + "epoch": 0.27428799132196185, + "grad_norm": 1.3675352334976196, + "learning_rate": 1.8206267157227918e-05, + "loss": 0.8383, + "step": 6195 + }, + { + "epoch": 0.2743322670267757, + "grad_norm": 1.2634061574935913, + "learning_rate": 1.8205383888727094e-05, + "loss": 0.854, + "step": 6196 + }, + { + "epoch": 0.2743765427315896, + "grad_norm": 1.4547953605651855, + "learning_rate": 1.820450042424736e-05, + "loss": 0.8792, + "step": 6197 + }, + { + "epoch": 0.2744208184364035, + "grad_norm": 1.2858740091323853, + "learning_rate": 1.820361676380981e-05, + "loss": 0.8616, + "step": 6198 + }, + { + "epoch": 0.27446509414121734, + "grad_norm": 1.3609930276870728, + "learning_rate": 1.8202732907435553e-05, + "loss": 0.8742, + "step": 6199 + }, + { + "epoch": 0.27450936984603125, + "grad_norm": 1.5586704015731812, + "learning_rate": 1.8201848855145704e-05, + "loss": 0.965, + "step": 6200 + }, + { + "epoch": 0.2745536455508451, + "grad_norm": 1.427504539489746, + "learning_rate": 1.820096460696137e-05, + "loss": 0.7994, + "step": 6201 + }, + { + "epoch": 0.27459792125565896, + "grad_norm": 1.1892222166061401, + "learning_rate": 1.8200080162903674e-05, + "loss": 0.8437, + "step": 6202 + }, + { + "epoch": 0.2746421969604729, + "grad_norm": 1.232505202293396, + "learning_rate": 1.819919552299374e-05, + "loss": 0.7935, + "step": 6203 + }, + { + "epoch": 0.27468647266528673, + "grad_norm": 1.2643743753433228, + "learning_rate": 1.81983106872527e-05, + "loss": 0.8227, + "step": 6204 + }, + { + "epoch": 0.27473074837010064, + "grad_norm": 1.403668761253357, + "learning_rate": 1.819742565570168e-05, + "loss": 0.8662, + "step": 6205 + }, + { + "epoch": 0.2747750240749145, + "grad_norm": 1.225851058959961, + "learning_rate": 1.8196540428361828e-05, + "loss": 0.8632, + "step": 6206 + }, + { + "epoch": 0.27481929977972835, + "grad_norm": 1.5746732950210571, + "learning_rate": 1.8195655005254274e-05, + "loss": 0.8303, + "step": 6207 + }, + { + "epoch": 0.27486357548454227, + "grad_norm": 1.5607260465621948, + "learning_rate": 1.8194769386400177e-05, + "loss": 0.9906, + "step": 6208 + }, + { + "epoch": 0.2749078511893561, + "grad_norm": 1.1763452291488647, + "learning_rate": 1.819388357182069e-05, + "loss": 0.858, + "step": 6209 + }, + { + "epoch": 0.27495212689417, + "grad_norm": 1.1927772760391235, + "learning_rate": 1.819299756153696e-05, + "loss": 0.8613, + "step": 6210 + }, + { + "epoch": 0.2749964025989839, + "grad_norm": 1.2961671352386475, + "learning_rate": 1.8192111355570155e-05, + "loss": 0.851, + "step": 6211 + }, + { + "epoch": 0.27504067830379775, + "grad_norm": 1.4492143392562866, + "learning_rate": 1.8191224953941443e-05, + "loss": 0.9633, + "step": 6212 + }, + { + "epoch": 0.2750849540086116, + "grad_norm": 1.5691770315170288, + "learning_rate": 1.8190338356671993e-05, + "loss": 0.9258, + "step": 6213 + }, + { + "epoch": 0.2751292297134255, + "grad_norm": 1.5443296432495117, + "learning_rate": 1.8189451563782977e-05, + "loss": 0.9322, + "step": 6214 + }, + { + "epoch": 0.2751735054182394, + "grad_norm": 1.248410701751709, + "learning_rate": 1.818856457529558e-05, + "loss": 0.8766, + "step": 6215 + }, + { + "epoch": 0.27521778112305323, + "grad_norm": 1.3123887777328491, + "learning_rate": 1.8187677391230985e-05, + "loss": 0.8473, + "step": 6216 + }, + { + "epoch": 0.27526205682786714, + "grad_norm": 1.2962006330490112, + "learning_rate": 1.818679001161038e-05, + "loss": 0.8742, + "step": 6217 + }, + { + "epoch": 0.275306332532681, + "grad_norm": 1.349287509918213, + "learning_rate": 1.8185902436454967e-05, + "loss": 0.8068, + "step": 6218 + }, + { + "epoch": 0.27535060823749485, + "grad_norm": 1.3431658744812012, + "learning_rate": 1.8185014665785936e-05, + "loss": 0.8587, + "step": 6219 + }, + { + "epoch": 0.27539488394230877, + "grad_norm": 1.5408574342727661, + "learning_rate": 1.8184126699624497e-05, + "loss": 0.9009, + "step": 6220 + }, + { + "epoch": 0.2754391596471226, + "grad_norm": 1.3855340480804443, + "learning_rate": 1.818323853799185e-05, + "loss": 0.8718, + "step": 6221 + }, + { + "epoch": 0.27548343535193653, + "grad_norm": 1.4604923725128174, + "learning_rate": 1.818235018090922e-05, + "loss": 0.9159, + "step": 6222 + }, + { + "epoch": 0.2755277110567504, + "grad_norm": 1.0868926048278809, + "learning_rate": 1.8181461628397815e-05, + "loss": 0.7689, + "step": 6223 + }, + { + "epoch": 0.27557198676156425, + "grad_norm": 1.5997895002365112, + "learning_rate": 1.8180572880478856e-05, + "loss": 0.8795, + "step": 6224 + }, + { + "epoch": 0.27561626246637816, + "grad_norm": 1.2377684116363525, + "learning_rate": 1.8179683937173582e-05, + "loss": 0.8909, + "step": 6225 + }, + { + "epoch": 0.275660538171192, + "grad_norm": 1.1804779767990112, + "learning_rate": 1.8178794798503214e-05, + "loss": 0.8207, + "step": 6226 + }, + { + "epoch": 0.27570481387600587, + "grad_norm": 1.3355721235275269, + "learning_rate": 1.8177905464488993e-05, + "loss": 0.8985, + "step": 6227 + }, + { + "epoch": 0.2757490895808198, + "grad_norm": 1.380383014678955, + "learning_rate": 1.8177015935152156e-05, + "loss": 0.8548, + "step": 6228 + }, + { + "epoch": 0.27579336528563364, + "grad_norm": 1.1957629919052124, + "learning_rate": 1.8176126210513957e-05, + "loss": 0.8024, + "step": 6229 + }, + { + "epoch": 0.2758376409904475, + "grad_norm": 1.2470964193344116, + "learning_rate": 1.8175236290595635e-05, + "loss": 0.7863, + "step": 6230 + }, + { + "epoch": 0.2758819166952614, + "grad_norm": 1.5630552768707275, + "learning_rate": 1.8174346175418452e-05, + "loss": 0.925, + "step": 6231 + }, + { + "epoch": 0.27592619240007527, + "grad_norm": 1.348807454109192, + "learning_rate": 1.817345586500367e-05, + "loss": 0.9265, + "step": 6232 + }, + { + "epoch": 0.2759704681048891, + "grad_norm": 1.4124587774276733, + "learning_rate": 1.8172565359372546e-05, + "loss": 0.8528, + "step": 6233 + }, + { + "epoch": 0.27601474380970303, + "grad_norm": 1.546409249305725, + "learning_rate": 1.8171674658546354e-05, + "loss": 0.9085, + "step": 6234 + }, + { + "epoch": 0.2760590195145169, + "grad_norm": 1.1572372913360596, + "learning_rate": 1.8170783762546363e-05, + "loss": 0.8513, + "step": 6235 + }, + { + "epoch": 0.27610329521933075, + "grad_norm": 1.1863172054290771, + "learning_rate": 1.816989267139386e-05, + "loss": 0.8303, + "step": 6236 + }, + { + "epoch": 0.27614757092414466, + "grad_norm": 1.294666051864624, + "learning_rate": 1.816900138511012e-05, + "loss": 0.8636, + "step": 6237 + }, + { + "epoch": 0.2761918466289585, + "grad_norm": 1.236519694328308, + "learning_rate": 1.8168109903716434e-05, + "loss": 0.7966, + "step": 6238 + }, + { + "epoch": 0.2762361223337724, + "grad_norm": 1.3835053443908691, + "learning_rate": 1.8167218227234095e-05, + "loss": 0.9991, + "step": 6239 + }, + { + "epoch": 0.2762803980385863, + "grad_norm": 1.5853430032730103, + "learning_rate": 1.8166326355684395e-05, + "loss": 0.8974, + "step": 6240 + }, + { + "epoch": 0.27632467374340014, + "grad_norm": 1.3915185928344727, + "learning_rate": 1.8165434289088645e-05, + "loss": 0.9468, + "step": 6241 + }, + { + "epoch": 0.27636894944821405, + "grad_norm": 1.6219708919525146, + "learning_rate": 1.816454202746814e-05, + "loss": 0.8649, + "step": 6242 + }, + { + "epoch": 0.2764132251530279, + "grad_norm": 1.183007001876831, + "learning_rate": 1.8163649570844198e-05, + "loss": 0.8683, + "step": 6243 + }, + { + "epoch": 0.27645750085784176, + "grad_norm": 1.1778254508972168, + "learning_rate": 1.8162756919238137e-05, + "loss": 0.8043, + "step": 6244 + }, + { + "epoch": 0.2765017765626557, + "grad_norm": 1.1875580549240112, + "learning_rate": 1.8161864072671267e-05, + "loss": 0.7862, + "step": 6245 + }, + { + "epoch": 0.27654605226746953, + "grad_norm": 2.391568899154663, + "learning_rate": 1.8160971031164926e-05, + "loss": 0.8609, + "step": 6246 + }, + { + "epoch": 0.2765903279722834, + "grad_norm": 1.2840981483459473, + "learning_rate": 1.816007779474043e-05, + "loss": 0.8311, + "step": 6247 + }, + { + "epoch": 0.2766346036770973, + "grad_norm": 1.3107925653457642, + "learning_rate": 1.8159184363419124e-05, + "loss": 0.8585, + "step": 6248 + }, + { + "epoch": 0.27667887938191116, + "grad_norm": 1.3383466005325317, + "learning_rate": 1.815829073722234e-05, + "loss": 0.8676, + "step": 6249 + }, + { + "epoch": 0.276723155086725, + "grad_norm": 1.2711457014083862, + "learning_rate": 1.8157396916171424e-05, + "loss": 0.804, + "step": 6250 + }, + { + "epoch": 0.2767674307915389, + "grad_norm": 1.302825689315796, + "learning_rate": 1.815650290028773e-05, + "loss": 0.8193, + "step": 6251 + }, + { + "epoch": 0.2768117064963528, + "grad_norm": 1.5663390159606934, + "learning_rate": 1.8155608689592604e-05, + "loss": 0.8857, + "step": 6252 + }, + { + "epoch": 0.27685598220116664, + "grad_norm": 1.4084155559539795, + "learning_rate": 1.8154714284107403e-05, + "loss": 0.8612, + "step": 6253 + }, + { + "epoch": 0.27690025790598055, + "grad_norm": 1.2036056518554688, + "learning_rate": 1.8153819683853494e-05, + "loss": 0.7758, + "step": 6254 + }, + { + "epoch": 0.2769445336107944, + "grad_norm": 1.3247499465942383, + "learning_rate": 1.8152924888852234e-05, + "loss": 0.828, + "step": 6255 + }, + { + "epoch": 0.2769888093156083, + "grad_norm": 1.686437726020813, + "learning_rate": 1.8152029899125007e-05, + "loss": 0.8544, + "step": 6256 + }, + { + "epoch": 0.2770330850204222, + "grad_norm": 1.7599738836288452, + "learning_rate": 1.8151134714693183e-05, + "loss": 0.9769, + "step": 6257 + }, + { + "epoch": 0.27707736072523603, + "grad_norm": 1.2286477088928223, + "learning_rate": 1.8150239335578144e-05, + "loss": 0.8637, + "step": 6258 + }, + { + "epoch": 0.27712163643004994, + "grad_norm": 1.5197747945785522, + "learning_rate": 1.8149343761801274e-05, + "loss": 0.8756, + "step": 6259 + }, + { + "epoch": 0.2771659121348638, + "grad_norm": 1.309995174407959, + "learning_rate": 1.8148447993383963e-05, + "loss": 0.8798, + "step": 6260 + }, + { + "epoch": 0.27721018783967766, + "grad_norm": 1.2312465906143188, + "learning_rate": 1.8147552030347606e-05, + "loss": 0.7763, + "step": 6261 + }, + { + "epoch": 0.27725446354449157, + "grad_norm": 1.5615193843841553, + "learning_rate": 1.8146655872713604e-05, + "loss": 0.8747, + "step": 6262 + }, + { + "epoch": 0.2772987392493054, + "grad_norm": 1.6423746347427368, + "learning_rate": 1.814575952050336e-05, + "loss": 0.8925, + "step": 6263 + }, + { + "epoch": 0.2773430149541193, + "grad_norm": 1.4960439205169678, + "learning_rate": 1.814486297373828e-05, + "loss": 0.8746, + "step": 6264 + }, + { + "epoch": 0.2773872906589332, + "grad_norm": 1.5178704261779785, + "learning_rate": 1.814396623243978e-05, + "loss": 0.8949, + "step": 6265 + }, + { + "epoch": 0.27743156636374705, + "grad_norm": 1.5302770137786865, + "learning_rate": 1.814306929662928e-05, + "loss": 0.8705, + "step": 6266 + }, + { + "epoch": 0.2774758420685609, + "grad_norm": 1.3429925441741943, + "learning_rate": 1.8142172166328198e-05, + "loss": 0.8656, + "step": 6267 + }, + { + "epoch": 0.2775201177733748, + "grad_norm": 1.3514283895492554, + "learning_rate": 1.8141274841557964e-05, + "loss": 0.7793, + "step": 6268 + }, + { + "epoch": 0.2775643934781887, + "grad_norm": 1.301773190498352, + "learning_rate": 1.814037732234001e-05, + "loss": 0.8505, + "step": 6269 + }, + { + "epoch": 0.27760866918300253, + "grad_norm": 1.297656774520874, + "learning_rate": 1.8139479608695773e-05, + "loss": 0.8817, + "step": 6270 + }, + { + "epoch": 0.27765294488781644, + "grad_norm": 1.4738876819610596, + "learning_rate": 1.8138581700646688e-05, + "loss": 0.91, + "step": 6271 + }, + { + "epoch": 0.2776972205926303, + "grad_norm": 1.6565659046173096, + "learning_rate": 1.8137683598214215e-05, + "loss": 0.9326, + "step": 6272 + }, + { + "epoch": 0.2777414962974442, + "grad_norm": 1.2645124197006226, + "learning_rate": 1.8136785301419786e-05, + "loss": 0.794, + "step": 6273 + }, + { + "epoch": 0.27778577200225807, + "grad_norm": 1.3967677354812622, + "learning_rate": 1.813588681028487e-05, + "loss": 0.9006, + "step": 6274 + }, + { + "epoch": 0.2778300477070719, + "grad_norm": 1.3915923833847046, + "learning_rate": 1.813498812483092e-05, + "loss": 0.8202, + "step": 6275 + }, + { + "epoch": 0.27787432341188584, + "grad_norm": 1.4607360363006592, + "learning_rate": 1.81340892450794e-05, + "loss": 0.8275, + "step": 6276 + }, + { + "epoch": 0.2779185991166997, + "grad_norm": 1.290507197380066, + "learning_rate": 1.8133190171051785e-05, + "loss": 0.8221, + "step": 6277 + }, + { + "epoch": 0.27796287482151355, + "grad_norm": 1.6867916584014893, + "learning_rate": 1.8132290902769547e-05, + "loss": 0.8853, + "step": 6278 + }, + { + "epoch": 0.27800715052632746, + "grad_norm": 1.2930970191955566, + "learning_rate": 1.813139144025416e-05, + "loss": 0.8683, + "step": 6279 + }, + { + "epoch": 0.2780514262311413, + "grad_norm": 1.3728101253509521, + "learning_rate": 1.813049178352711e-05, + "loss": 0.9107, + "step": 6280 + }, + { + "epoch": 0.2780957019359552, + "grad_norm": 1.1973010301589966, + "learning_rate": 1.812959193260988e-05, + "loss": 0.786, + "step": 6281 + }, + { + "epoch": 0.2781399776407691, + "grad_norm": 1.484223484992981, + "learning_rate": 1.812869188752397e-05, + "loss": 0.9401, + "step": 6282 + }, + { + "epoch": 0.27818425334558294, + "grad_norm": 1.2985258102416992, + "learning_rate": 1.812779164829087e-05, + "loss": 0.8747, + "step": 6283 + }, + { + "epoch": 0.2782285290503968, + "grad_norm": 1.4572806358337402, + "learning_rate": 1.8126891214932086e-05, + "loss": 0.967, + "step": 6284 + }, + { + "epoch": 0.2782728047552107, + "grad_norm": 1.5556015968322754, + "learning_rate": 1.8125990587469125e-05, + "loss": 0.8814, + "step": 6285 + }, + { + "epoch": 0.27831708046002457, + "grad_norm": 1.2106071710586548, + "learning_rate": 1.812508976592349e-05, + "loss": 0.757, + "step": 6286 + }, + { + "epoch": 0.2783613561648384, + "grad_norm": 1.2936218976974487, + "learning_rate": 1.8124188750316708e-05, + "loss": 0.8366, + "step": 6287 + }, + { + "epoch": 0.27840563186965234, + "grad_norm": 1.6805546283721924, + "learning_rate": 1.812328754067029e-05, + "loss": 0.9232, + "step": 6288 + }, + { + "epoch": 0.2784499075744662, + "grad_norm": 1.2951008081436157, + "learning_rate": 1.812238613700576e-05, + "loss": 0.794, + "step": 6289 + }, + { + "epoch": 0.27849418327928005, + "grad_norm": 1.455409049987793, + "learning_rate": 1.8121484539344653e-05, + "loss": 0.8505, + "step": 6290 + }, + { + "epoch": 0.27853845898409396, + "grad_norm": 1.3005601167678833, + "learning_rate": 1.8120582747708503e-05, + "loss": 0.7924, + "step": 6291 + }, + { + "epoch": 0.2785827346889078, + "grad_norm": 1.3488134145736694, + "learning_rate": 1.8119680762118844e-05, + "loss": 0.8101, + "step": 6292 + }, + { + "epoch": 0.27862701039372173, + "grad_norm": 1.1867269277572632, + "learning_rate": 1.811877858259722e-05, + "loss": 0.818, + "step": 6293 + }, + { + "epoch": 0.2786712860985356, + "grad_norm": 1.2302249670028687, + "learning_rate": 1.811787620916518e-05, + "loss": 0.8165, + "step": 6294 + }, + { + "epoch": 0.27871556180334944, + "grad_norm": 1.3362959623336792, + "learning_rate": 1.8116973641844277e-05, + "loss": 0.8465, + "step": 6295 + }, + { + "epoch": 0.27875983750816336, + "grad_norm": 1.3892056941986084, + "learning_rate": 1.8116070880656068e-05, + "loss": 0.8969, + "step": 6296 + }, + { + "epoch": 0.2788041132129772, + "grad_norm": 1.2665610313415527, + "learning_rate": 1.8115167925622116e-05, + "loss": 0.7696, + "step": 6297 + }, + { + "epoch": 0.27884838891779107, + "grad_norm": 1.3251874446868896, + "learning_rate": 1.8114264776763987e-05, + "loss": 0.9103, + "step": 6298 + }, + { + "epoch": 0.278892664622605, + "grad_norm": 1.30780029296875, + "learning_rate": 1.8113361434103245e-05, + "loss": 0.8432, + "step": 6299 + }, + { + "epoch": 0.27893694032741884, + "grad_norm": 1.289343237876892, + "learning_rate": 1.8112457897661478e-05, + "loss": 0.8167, + "step": 6300 + }, + { + "epoch": 0.2789812160322327, + "grad_norm": 1.476423740386963, + "learning_rate": 1.8111554167460252e-05, + "loss": 0.902, + "step": 6301 + }, + { + "epoch": 0.2790254917370466, + "grad_norm": 1.3283315896987915, + "learning_rate": 1.8110650243521165e-05, + "loss": 0.7875, + "step": 6302 + }, + { + "epoch": 0.27906976744186046, + "grad_norm": 1.4628552198410034, + "learning_rate": 1.81097461258658e-05, + "loss": 0.8298, + "step": 6303 + }, + { + "epoch": 0.2791140431466743, + "grad_norm": 1.1978219747543335, + "learning_rate": 1.810884181451575e-05, + "loss": 0.8306, + "step": 6304 + }, + { + "epoch": 0.27915831885148823, + "grad_norm": 1.397111415863037, + "learning_rate": 1.8107937309492616e-05, + "loss": 0.8801, + "step": 6305 + }, + { + "epoch": 0.2792025945563021, + "grad_norm": 1.157480239868164, + "learning_rate": 1.8107032610818005e-05, + "loss": 0.8505, + "step": 6306 + }, + { + "epoch": 0.27924687026111594, + "grad_norm": 1.9943366050720215, + "learning_rate": 1.810612771851352e-05, + "loss": 0.8882, + "step": 6307 + }, + { + "epoch": 0.27929114596592985, + "grad_norm": 1.1444827318191528, + "learning_rate": 1.810522263260077e-05, + "loss": 0.6907, + "step": 6308 + }, + { + "epoch": 0.2793354216707437, + "grad_norm": 1.4624370336532593, + "learning_rate": 1.810431735310138e-05, + "loss": 0.8589, + "step": 6309 + }, + { + "epoch": 0.2793796973755576, + "grad_norm": 1.4633867740631104, + "learning_rate": 1.8103411880036966e-05, + "loss": 0.8731, + "step": 6310 + }, + { + "epoch": 0.2794239730803715, + "grad_norm": 1.4658931493759155, + "learning_rate": 1.8102506213429163e-05, + "loss": 0.8726, + "step": 6311 + }, + { + "epoch": 0.27946824878518534, + "grad_norm": 1.3842878341674805, + "learning_rate": 1.8101600353299593e-05, + "loss": 0.8772, + "step": 6312 + }, + { + "epoch": 0.27951252448999925, + "grad_norm": 1.8455804586410522, + "learning_rate": 1.8100694299669894e-05, + "loss": 0.8525, + "step": 6313 + }, + { + "epoch": 0.2795568001948131, + "grad_norm": 1.227271318435669, + "learning_rate": 1.8099788052561708e-05, + "loss": 0.8204, + "step": 6314 + }, + { + "epoch": 0.27960107589962696, + "grad_norm": 1.4074442386627197, + "learning_rate": 1.809888161199668e-05, + "loss": 0.8803, + "step": 6315 + }, + { + "epoch": 0.2796453516044409, + "grad_norm": 1.5468473434448242, + "learning_rate": 1.8097974977996463e-05, + "loss": 0.8304, + "step": 6316 + }, + { + "epoch": 0.27968962730925473, + "grad_norm": 1.2553178071975708, + "learning_rate": 1.80970681505827e-05, + "loss": 0.8545, + "step": 6317 + }, + { + "epoch": 0.2797339030140686, + "grad_norm": 1.4289804697036743, + "learning_rate": 1.809616112977706e-05, + "loss": 0.866, + "step": 6318 + }, + { + "epoch": 0.2797781787188825, + "grad_norm": 1.6237845420837402, + "learning_rate": 1.8095253915601207e-05, + "loss": 0.9056, + "step": 6319 + }, + { + "epoch": 0.27982245442369635, + "grad_norm": 2.000379800796509, + "learning_rate": 1.8094346508076802e-05, + "loss": 0.8888, + "step": 6320 + }, + { + "epoch": 0.2798667301285102, + "grad_norm": 1.3446398973464966, + "learning_rate": 1.8093438907225523e-05, + "loss": 0.8878, + "step": 6321 + }, + { + "epoch": 0.2799110058333241, + "grad_norm": 1.9097660779953003, + "learning_rate": 1.8092531113069044e-05, + "loss": 0.9475, + "step": 6322 + }, + { + "epoch": 0.279955281538138, + "grad_norm": 1.2114169597625732, + "learning_rate": 1.809162312562905e-05, + "loss": 0.8041, + "step": 6323 + }, + { + "epoch": 0.27999955724295184, + "grad_norm": 1.3884012699127197, + "learning_rate": 1.8090714944927226e-05, + "loss": 0.8243, + "step": 6324 + }, + { + "epoch": 0.28004383294776575, + "grad_norm": 1.389933466911316, + "learning_rate": 1.8089806570985263e-05, + "loss": 0.8465, + "step": 6325 + }, + { + "epoch": 0.2800881086525796, + "grad_norm": 1.2356574535369873, + "learning_rate": 1.8088898003824855e-05, + "loss": 0.8982, + "step": 6326 + }, + { + "epoch": 0.2801323843573935, + "grad_norm": 1.4203449487686157, + "learning_rate": 1.8087989243467708e-05, + "loss": 0.8455, + "step": 6327 + }, + { + "epoch": 0.28017666006220737, + "grad_norm": 1.466176152229309, + "learning_rate": 1.808708028993552e-05, + "loss": 0.8251, + "step": 6328 + }, + { + "epoch": 0.28022093576702123, + "grad_norm": 1.424252986907959, + "learning_rate": 1.8086171143250006e-05, + "loss": 0.9466, + "step": 6329 + }, + { + "epoch": 0.28026521147183514, + "grad_norm": 1.3997935056686401, + "learning_rate": 1.808526180343288e-05, + "loss": 0.8039, + "step": 6330 + }, + { + "epoch": 0.280309487176649, + "grad_norm": 1.5041078329086304, + "learning_rate": 1.8084352270505855e-05, + "loss": 0.8964, + "step": 6331 + }, + { + "epoch": 0.28035376288146285, + "grad_norm": 1.41476309299469, + "learning_rate": 1.808344254449066e-05, + "loss": 0.8957, + "step": 6332 + }, + { + "epoch": 0.28039803858627677, + "grad_norm": 1.080274224281311, + "learning_rate": 1.8082532625409022e-05, + "loss": 0.7419, + "step": 6333 + }, + { + "epoch": 0.2804423142910906, + "grad_norm": 1.8355680704116821, + "learning_rate": 1.8081622513282674e-05, + "loss": 0.8527, + "step": 6334 + }, + { + "epoch": 0.2804865899959045, + "grad_norm": 1.3616458177566528, + "learning_rate": 1.8080712208133353e-05, + "loss": 0.809, + "step": 6335 + }, + { + "epoch": 0.2805308657007184, + "grad_norm": 1.333123803138733, + "learning_rate": 1.8079801709982798e-05, + "loss": 0.8595, + "step": 6336 + }, + { + "epoch": 0.28057514140553225, + "grad_norm": 1.4029933214187622, + "learning_rate": 1.807889101885276e-05, + "loss": 0.8814, + "step": 6337 + }, + { + "epoch": 0.2806194171103461, + "grad_norm": 2.092027425765991, + "learning_rate": 1.807798013476499e-05, + "loss": 1.0141, + "step": 6338 + }, + { + "epoch": 0.28066369281516, + "grad_norm": 1.3307863473892212, + "learning_rate": 1.8077069057741235e-05, + "loss": 0.8609, + "step": 6339 + }, + { + "epoch": 0.28070796851997387, + "grad_norm": 1.7557185888290405, + "learning_rate": 1.8076157787803268e-05, + "loss": 1.0391, + "step": 6340 + }, + { + "epoch": 0.28075224422478773, + "grad_norm": 1.2459981441497803, + "learning_rate": 1.8075246324972845e-05, + "loss": 0.8869, + "step": 6341 + }, + { + "epoch": 0.28079651992960164, + "grad_norm": 1.3890830278396606, + "learning_rate": 1.807433466927174e-05, + "loss": 0.8878, + "step": 6342 + }, + { + "epoch": 0.2808407956344155, + "grad_norm": 1.4272469282150269, + "learning_rate": 1.8073422820721728e-05, + "loss": 0.9098, + "step": 6343 + }, + { + "epoch": 0.2808850713392294, + "grad_norm": 1.7246814966201782, + "learning_rate": 1.8072510779344584e-05, + "loss": 0.9136, + "step": 6344 + }, + { + "epoch": 0.28092934704404326, + "grad_norm": 1.4220246076583862, + "learning_rate": 1.8071598545162093e-05, + "loss": 0.8683, + "step": 6345 + }, + { + "epoch": 0.2809736227488571, + "grad_norm": 1.3309375047683716, + "learning_rate": 1.8070686118196043e-05, + "loss": 0.8934, + "step": 6346 + }, + { + "epoch": 0.28101789845367103, + "grad_norm": 1.4011127948760986, + "learning_rate": 1.8069773498468224e-05, + "loss": 0.8006, + "step": 6347 + }, + { + "epoch": 0.2810621741584849, + "grad_norm": 1.4370752573013306, + "learning_rate": 1.8068860686000437e-05, + "loss": 0.8496, + "step": 6348 + }, + { + "epoch": 0.28110644986329875, + "grad_norm": 1.2711220979690552, + "learning_rate": 1.8067947680814484e-05, + "loss": 0.8422, + "step": 6349 + }, + { + "epoch": 0.28115072556811266, + "grad_norm": 1.4167003631591797, + "learning_rate": 1.806703448293217e-05, + "loss": 0.7934, + "step": 6350 + }, + { + "epoch": 0.2811950012729265, + "grad_norm": 1.4665132761001587, + "learning_rate": 1.8066121092375303e-05, + "loss": 0.824, + "step": 6351 + }, + { + "epoch": 0.28123927697774037, + "grad_norm": 1.2672141790390015, + "learning_rate": 1.8065207509165705e-05, + "loss": 0.9032, + "step": 6352 + }, + { + "epoch": 0.2812835526825543, + "grad_norm": 1.382145643234253, + "learning_rate": 1.8064293733325188e-05, + "loss": 0.8001, + "step": 6353 + }, + { + "epoch": 0.28132782838736814, + "grad_norm": 1.205336570739746, + "learning_rate": 1.8063379764875583e-05, + "loss": 0.825, + "step": 6354 + }, + { + "epoch": 0.281372104092182, + "grad_norm": 1.2778880596160889, + "learning_rate": 1.806246560383872e-05, + "loss": 0.737, + "step": 6355 + }, + { + "epoch": 0.2814163797969959, + "grad_norm": 1.5497167110443115, + "learning_rate": 1.8061551250236426e-05, + "loss": 0.8225, + "step": 6356 + }, + { + "epoch": 0.28146065550180976, + "grad_norm": 1.241515874862671, + "learning_rate": 1.806063670409055e-05, + "loss": 0.9057, + "step": 6357 + }, + { + "epoch": 0.2815049312066236, + "grad_norm": 1.2797759771347046, + "learning_rate": 1.805972196542293e-05, + "loss": 0.8895, + "step": 6358 + }, + { + "epoch": 0.28154920691143753, + "grad_norm": 1.1822313070297241, + "learning_rate": 1.8058807034255408e-05, + "loss": 0.8124, + "step": 6359 + }, + { + "epoch": 0.2815934826162514, + "grad_norm": 1.6208600997924805, + "learning_rate": 1.8057891910609845e-05, + "loss": 0.8485, + "step": 6360 + }, + { + "epoch": 0.28163775832106525, + "grad_norm": 1.202398657798767, + "learning_rate": 1.8056976594508093e-05, + "loss": 0.8774, + "step": 6361 + }, + { + "epoch": 0.28168203402587916, + "grad_norm": 1.4969931840896606, + "learning_rate": 1.8056061085972016e-05, + "loss": 0.8505, + "step": 6362 + }, + { + "epoch": 0.281726309730693, + "grad_norm": 1.5446858406066895, + "learning_rate": 1.8055145385023477e-05, + "loss": 0.7807, + "step": 6363 + }, + { + "epoch": 0.2817705854355069, + "grad_norm": 1.3071941137313843, + "learning_rate": 1.8054229491684357e-05, + "loss": 0.8365, + "step": 6364 + }, + { + "epoch": 0.2818148611403208, + "grad_norm": 1.2011324167251587, + "learning_rate": 1.8053313405976516e-05, + "loss": 0.7898, + "step": 6365 + }, + { + "epoch": 0.28185913684513464, + "grad_norm": 1.5059524774551392, + "learning_rate": 1.8052397127921845e-05, + "loss": 0.7663, + "step": 6366 + }, + { + "epoch": 0.28190341254994855, + "grad_norm": 1.8921513557434082, + "learning_rate": 1.805148065754222e-05, + "loss": 0.8084, + "step": 6367 + }, + { + "epoch": 0.2819476882547624, + "grad_norm": 1.301504135131836, + "learning_rate": 1.805056399485954e-05, + "loss": 0.8563, + "step": 6368 + }, + { + "epoch": 0.28199196395957626, + "grad_norm": 1.2122983932495117, + "learning_rate": 1.8049647139895697e-05, + "loss": 0.7787, + "step": 6369 + }, + { + "epoch": 0.2820362396643902, + "grad_norm": 1.3234639167785645, + "learning_rate": 1.804873009267258e-05, + "loss": 0.7761, + "step": 6370 + }, + { + "epoch": 0.28208051536920403, + "grad_norm": 1.2183668613433838, + "learning_rate": 1.80478128532121e-05, + "loss": 0.8541, + "step": 6371 + }, + { + "epoch": 0.2821247910740179, + "grad_norm": 1.4537153244018555, + "learning_rate": 1.8046895421536166e-05, + "loss": 0.9388, + "step": 6372 + }, + { + "epoch": 0.2821690667788318, + "grad_norm": 1.2101459503173828, + "learning_rate": 1.8045977797666685e-05, + "loss": 0.8681, + "step": 6373 + }, + { + "epoch": 0.28221334248364566, + "grad_norm": 1.403905987739563, + "learning_rate": 1.8045059981625573e-05, + "loss": 0.8677, + "step": 6374 + }, + { + "epoch": 0.2822576181884595, + "grad_norm": 1.2443536520004272, + "learning_rate": 1.804414197343476e-05, + "loss": 0.8052, + "step": 6375 + }, + { + "epoch": 0.2823018938932734, + "grad_norm": 1.255934715270996, + "learning_rate": 1.804322377311616e-05, + "loss": 0.782, + "step": 6376 + }, + { + "epoch": 0.2823461695980873, + "grad_norm": 1.612573266029358, + "learning_rate": 1.8042305380691713e-05, + "loss": 0.9121, + "step": 6377 + }, + { + "epoch": 0.28239044530290114, + "grad_norm": 1.4187378883361816, + "learning_rate": 1.8041386796183345e-05, + "loss": 0.8549, + "step": 6378 + }, + { + "epoch": 0.28243472100771505, + "grad_norm": 1.6414906978607178, + "learning_rate": 1.8040468019613008e-05, + "loss": 0.9127, + "step": 6379 + }, + { + "epoch": 0.2824789967125289, + "grad_norm": 1.5386807918548584, + "learning_rate": 1.8039549051002635e-05, + "loss": 0.8967, + "step": 6380 + }, + { + "epoch": 0.2825232724173428, + "grad_norm": 1.3973931074142456, + "learning_rate": 1.8038629890374176e-05, + "loss": 0.8785, + "step": 6381 + }, + { + "epoch": 0.2825675481221567, + "grad_norm": 1.4551808834075928, + "learning_rate": 1.8037710537749593e-05, + "loss": 0.9005, + "step": 6382 + }, + { + "epoch": 0.28261182382697053, + "grad_norm": 1.271666407585144, + "learning_rate": 1.8036790993150836e-05, + "loss": 0.8314, + "step": 6383 + }, + { + "epoch": 0.28265609953178444, + "grad_norm": 1.550487995147705, + "learning_rate": 1.8035871256599868e-05, + "loss": 0.9734, + "step": 6384 + }, + { + "epoch": 0.2827003752365983, + "grad_norm": 1.0923281908035278, + "learning_rate": 1.803495132811866e-05, + "loss": 0.7765, + "step": 6385 + }, + { + "epoch": 0.28274465094141216, + "grad_norm": 1.2890644073486328, + "learning_rate": 1.8034031207729184e-05, + "loss": 0.8349, + "step": 6386 + }, + { + "epoch": 0.28278892664622607, + "grad_norm": 1.349563479423523, + "learning_rate": 1.803311089545341e-05, + "loss": 0.8527, + "step": 6387 + }, + { + "epoch": 0.2828332023510399, + "grad_norm": 1.1080052852630615, + "learning_rate": 1.8032190391313324e-05, + "loss": 0.7352, + "step": 6388 + }, + { + "epoch": 0.2828774780558538, + "grad_norm": 1.3031573295593262, + "learning_rate": 1.803126969533091e-05, + "loss": 0.7978, + "step": 6389 + }, + { + "epoch": 0.2829217537606677, + "grad_norm": 1.3958282470703125, + "learning_rate": 1.8030348807528165e-05, + "loss": 0.8634, + "step": 6390 + }, + { + "epoch": 0.28296602946548155, + "grad_norm": 1.4727071523666382, + "learning_rate": 1.802942772792707e-05, + "loss": 0.9608, + "step": 6391 + }, + { + "epoch": 0.2830103051702954, + "grad_norm": 1.3643712997436523, + "learning_rate": 1.8028506456549635e-05, + "loss": 0.8972, + "step": 6392 + }, + { + "epoch": 0.2830545808751093, + "grad_norm": 1.7919116020202637, + "learning_rate": 1.802758499341786e-05, + "loss": 0.7939, + "step": 6393 + }, + { + "epoch": 0.2830988565799232, + "grad_norm": 1.6634392738342285, + "learning_rate": 1.8026663338553753e-05, + "loss": 0.9501, + "step": 6394 + }, + { + "epoch": 0.28314313228473703, + "grad_norm": 1.4282859563827515, + "learning_rate": 1.8025741491979326e-05, + "loss": 0.9325, + "step": 6395 + }, + { + "epoch": 0.28318740798955094, + "grad_norm": 1.479287028312683, + "learning_rate": 1.8024819453716603e-05, + "loss": 0.9549, + "step": 6396 + }, + { + "epoch": 0.2832316836943648, + "grad_norm": 1.5341323614120483, + "learning_rate": 1.80238972237876e-05, + "loss": 0.9322, + "step": 6397 + }, + { + "epoch": 0.2832759593991787, + "grad_norm": 1.378528356552124, + "learning_rate": 1.8022974802214342e-05, + "loss": 0.8469, + "step": 6398 + }, + { + "epoch": 0.28332023510399257, + "grad_norm": 2.0283687114715576, + "learning_rate": 1.8022052189018865e-05, + "loss": 0.9401, + "step": 6399 + }, + { + "epoch": 0.2833645108088064, + "grad_norm": 1.5791782140731812, + "learning_rate": 1.80211293842232e-05, + "loss": 0.843, + "step": 6400 + }, + { + "epoch": 0.28340878651362034, + "grad_norm": 1.3310763835906982, + "learning_rate": 1.8020206387849394e-05, + "loss": 0.8236, + "step": 6401 + }, + { + "epoch": 0.2834530622184342, + "grad_norm": 1.554817795753479, + "learning_rate": 1.801928319991949e-05, + "loss": 0.8757, + "step": 6402 + }, + { + "epoch": 0.28349733792324805, + "grad_norm": 1.194237470626831, + "learning_rate": 1.8018359820455535e-05, + "loss": 0.8234, + "step": 6403 + }, + { + "epoch": 0.28354161362806196, + "grad_norm": 1.729659080505371, + "learning_rate": 1.8017436249479585e-05, + "loss": 0.9277, + "step": 6404 + }, + { + "epoch": 0.2835858893328758, + "grad_norm": 1.183583378791809, + "learning_rate": 1.80165124870137e-05, + "loss": 0.8127, + "step": 6405 + }, + { + "epoch": 0.2836301650376897, + "grad_norm": 1.362159252166748, + "learning_rate": 1.8015588533079936e-05, + "loss": 0.9744, + "step": 6406 + }, + { + "epoch": 0.2836744407425036, + "grad_norm": 1.5769869089126587, + "learning_rate": 1.801466438770037e-05, + "loss": 0.9263, + "step": 6407 + }, + { + "epoch": 0.28371871644731744, + "grad_norm": 1.281004786491394, + "learning_rate": 1.801374005089707e-05, + "loss": 0.813, + "step": 6408 + }, + { + "epoch": 0.2837629921521313, + "grad_norm": 1.363815426826477, + "learning_rate": 1.8012815522692118e-05, + "loss": 0.7916, + "step": 6409 + }, + { + "epoch": 0.2838072678569452, + "grad_norm": 1.4725513458251953, + "learning_rate": 1.8011890803107586e-05, + "loss": 0.8692, + "step": 6410 + }, + { + "epoch": 0.28385154356175907, + "grad_norm": 1.1848002672195435, + "learning_rate": 1.8010965892165568e-05, + "loss": 0.7916, + "step": 6411 + }, + { + "epoch": 0.2838958192665729, + "grad_norm": 1.3587181568145752, + "learning_rate": 1.8010040789888152e-05, + "loss": 0.8942, + "step": 6412 + }, + { + "epoch": 0.28394009497138684, + "grad_norm": 1.57933509349823, + "learning_rate": 1.8009115496297435e-05, + "loss": 0.921, + "step": 6413 + }, + { + "epoch": 0.2839843706762007, + "grad_norm": 1.4411448240280151, + "learning_rate": 1.8008190011415515e-05, + "loss": 0.9448, + "step": 6414 + }, + { + "epoch": 0.2840286463810146, + "grad_norm": 1.5054004192352295, + "learning_rate": 1.8007264335264495e-05, + "loss": 0.8196, + "step": 6415 + }, + { + "epoch": 0.28407292208582846, + "grad_norm": 1.3159645795822144, + "learning_rate": 1.800633846786649e-05, + "loss": 0.8308, + "step": 6416 + }, + { + "epoch": 0.2841171977906423, + "grad_norm": 1.4467970132827759, + "learning_rate": 1.8005412409243604e-05, + "loss": 0.8804, + "step": 6417 + }, + { + "epoch": 0.28416147349545623, + "grad_norm": 1.3733056783676147, + "learning_rate": 1.8004486159417965e-05, + "loss": 0.8833, + "step": 6418 + }, + { + "epoch": 0.2842057492002701, + "grad_norm": 7.4761199951171875, + "learning_rate": 1.800355971841169e-05, + "loss": 0.766, + "step": 6419 + }, + { + "epoch": 0.28425002490508394, + "grad_norm": 1.19186270236969, + "learning_rate": 1.800263308624691e-05, + "loss": 0.7613, + "step": 6420 + }, + { + "epoch": 0.28429430060989785, + "grad_norm": 1.4636648893356323, + "learning_rate": 1.8001706262945756e-05, + "loss": 0.8266, + "step": 6421 + }, + { + "epoch": 0.2843385763147117, + "grad_norm": 1.3189318180084229, + "learning_rate": 1.8000779248530358e-05, + "loss": 0.8227, + "step": 6422 + }, + { + "epoch": 0.28438285201952557, + "grad_norm": 1.307102084159851, + "learning_rate": 1.7999852043022866e-05, + "loss": 0.7611, + "step": 6423 + }, + { + "epoch": 0.2844271277243395, + "grad_norm": 1.328163504600525, + "learning_rate": 1.799892464644542e-05, + "loss": 0.718, + "step": 6424 + }, + { + "epoch": 0.28447140342915334, + "grad_norm": 1.415635585784912, + "learning_rate": 1.799799705882017e-05, + "loss": 0.8547, + "step": 6425 + }, + { + "epoch": 0.2845156791339672, + "grad_norm": 1.3075798749923706, + "learning_rate": 1.7997069280169276e-05, + "loss": 0.8974, + "step": 6426 + }, + { + "epoch": 0.2845599548387811, + "grad_norm": 1.3331753015518188, + "learning_rate": 1.7996141310514893e-05, + "loss": 0.8102, + "step": 6427 + }, + { + "epoch": 0.28460423054359496, + "grad_norm": 1.703302264213562, + "learning_rate": 1.7995213149879186e-05, + "loss": 0.8593, + "step": 6428 + }, + { + "epoch": 0.2846485062484088, + "grad_norm": 1.3051105737686157, + "learning_rate": 1.7994284798284322e-05, + "loss": 0.8844, + "step": 6429 + }, + { + "epoch": 0.28469278195322273, + "grad_norm": 1.7690422534942627, + "learning_rate": 1.7993356255752473e-05, + "loss": 0.9065, + "step": 6430 + }, + { + "epoch": 0.2847370576580366, + "grad_norm": 1.281847596168518, + "learning_rate": 1.799242752230582e-05, + "loss": 0.8668, + "step": 6431 + }, + { + "epoch": 0.2847813333628505, + "grad_norm": 1.4671112298965454, + "learning_rate": 1.7991498597966545e-05, + "loss": 0.8862, + "step": 6432 + }, + { + "epoch": 0.28482560906766435, + "grad_norm": 1.5579185485839844, + "learning_rate": 1.799056948275683e-05, + "loss": 0.9461, + "step": 6433 + }, + { + "epoch": 0.2848698847724782, + "grad_norm": 1.3216581344604492, + "learning_rate": 1.798964017669887e-05, + "loss": 0.8201, + "step": 6434 + }, + { + "epoch": 0.2849141604772921, + "grad_norm": 1.401404857635498, + "learning_rate": 1.7988710679814857e-05, + "loss": 0.9222, + "step": 6435 + }, + { + "epoch": 0.284958436182106, + "grad_norm": 1.5277012586593628, + "learning_rate": 1.7987780992126997e-05, + "loss": 0.8928, + "step": 6436 + }, + { + "epoch": 0.28500271188691984, + "grad_norm": 1.275283932685852, + "learning_rate": 1.7986851113657495e-05, + "loss": 0.8011, + "step": 6437 + }, + { + "epoch": 0.28504698759173375, + "grad_norm": 1.275215744972229, + "learning_rate": 1.7985921044428553e-05, + "loss": 0.8458, + "step": 6438 + }, + { + "epoch": 0.2850912632965476, + "grad_norm": 1.6411139965057373, + "learning_rate": 1.798499078446239e-05, + "loss": 0.9626, + "step": 6439 + }, + { + "epoch": 0.28513553900136146, + "grad_norm": 1.351850152015686, + "learning_rate": 1.798406033378122e-05, + "loss": 0.878, + "step": 6440 + }, + { + "epoch": 0.28517981470617537, + "grad_norm": 1.2070766687393188, + "learning_rate": 1.7983129692407276e-05, + "loss": 0.766, + "step": 6441 + }, + { + "epoch": 0.28522409041098923, + "grad_norm": 1.44735586643219, + "learning_rate": 1.798219886036278e-05, + "loss": 0.8592, + "step": 6442 + }, + { + "epoch": 0.2852683661158031, + "grad_norm": 1.2663447856903076, + "learning_rate": 1.7981267837669957e-05, + "loss": 0.8206, + "step": 6443 + }, + { + "epoch": 0.285312641820617, + "grad_norm": 1.3943921327590942, + "learning_rate": 1.7980336624351055e-05, + "loss": 0.886, + "step": 6444 + }, + { + "epoch": 0.28535691752543085, + "grad_norm": 1.4260313510894775, + "learning_rate": 1.797940522042831e-05, + "loss": 0.9058, + "step": 6445 + }, + { + "epoch": 0.2854011932302447, + "grad_norm": 1.138166904449463, + "learning_rate": 1.797847362592397e-05, + "loss": 0.7902, + "step": 6446 + }, + { + "epoch": 0.2854454689350586, + "grad_norm": 1.3411757946014404, + "learning_rate": 1.7977541840860283e-05, + "loss": 0.8213, + "step": 6447 + }, + { + "epoch": 0.2854897446398725, + "grad_norm": 1.2881920337677002, + "learning_rate": 1.79766098652595e-05, + "loss": 0.7491, + "step": 6448 + }, + { + "epoch": 0.28553402034468633, + "grad_norm": 1.3297803401947021, + "learning_rate": 1.797567769914389e-05, + "loss": 0.85, + "step": 6449 + }, + { + "epoch": 0.28557829604950025, + "grad_norm": 1.1488428115844727, + "learning_rate": 1.797474534253571e-05, + "loss": 0.759, + "step": 6450 + }, + { + "epoch": 0.2856225717543141, + "grad_norm": 1.3874659538269043, + "learning_rate": 1.7973812795457233e-05, + "loss": 0.8634, + "step": 6451 + }, + { + "epoch": 0.285666847459128, + "grad_norm": 1.4688776731491089, + "learning_rate": 1.797288005793073e-05, + "loss": 0.8976, + "step": 6452 + }, + { + "epoch": 0.28571112316394187, + "grad_norm": 1.59812593460083, + "learning_rate": 1.7971947129978475e-05, + "loss": 0.8457, + "step": 6453 + }, + { + "epoch": 0.28575539886875573, + "grad_norm": 1.2382086515426636, + "learning_rate": 1.7971014011622755e-05, + "loss": 0.8426, + "step": 6454 + }, + { + "epoch": 0.28579967457356964, + "grad_norm": 1.1939420700073242, + "learning_rate": 1.7970080702885857e-05, + "loss": 0.7999, + "step": 6455 + }, + { + "epoch": 0.2858439502783835, + "grad_norm": 1.3255635499954224, + "learning_rate": 1.7969147203790068e-05, + "loss": 0.7828, + "step": 6456 + }, + { + "epoch": 0.28588822598319735, + "grad_norm": 1.1960134506225586, + "learning_rate": 1.7968213514357692e-05, + "loss": 0.8105, + "step": 6457 + }, + { + "epoch": 0.28593250168801126, + "grad_norm": 1.2697200775146484, + "learning_rate": 1.796727963461102e-05, + "loss": 0.781, + "step": 6458 + }, + { + "epoch": 0.2859767773928251, + "grad_norm": 1.3090418577194214, + "learning_rate": 1.796634556457236e-05, + "loss": 0.819, + "step": 6459 + }, + { + "epoch": 0.286021053097639, + "grad_norm": 1.432347059249878, + "learning_rate": 1.7965411304264026e-05, + "loss": 0.8915, + "step": 6460 + }, + { + "epoch": 0.2860653288024529, + "grad_norm": 1.3387421369552612, + "learning_rate": 1.7964476853708327e-05, + "loss": 0.7873, + "step": 6461 + }, + { + "epoch": 0.28610960450726675, + "grad_norm": 1.4517830610275269, + "learning_rate": 1.7963542212927586e-05, + "loss": 0.8322, + "step": 6462 + }, + { + "epoch": 0.2861538802120806, + "grad_norm": 1.5050922632217407, + "learning_rate": 1.7962607381944123e-05, + "loss": 0.9287, + "step": 6463 + }, + { + "epoch": 0.2861981559168945, + "grad_norm": 1.4617327451705933, + "learning_rate": 1.796167236078026e-05, + "loss": 0.9203, + "step": 6464 + }, + { + "epoch": 0.28624243162170837, + "grad_norm": 1.7925220727920532, + "learning_rate": 1.7960737149458344e-05, + "loss": 0.8771, + "step": 6465 + }, + { + "epoch": 0.2862867073265222, + "grad_norm": 1.3744361400604248, + "learning_rate": 1.7959801748000697e-05, + "loss": 0.9041, + "step": 6466 + }, + { + "epoch": 0.28633098303133614, + "grad_norm": 1.3320344686508179, + "learning_rate": 1.7958866156429667e-05, + "loss": 0.8136, + "step": 6467 + }, + { + "epoch": 0.28637525873615, + "grad_norm": 1.2777577638626099, + "learning_rate": 1.79579303747676e-05, + "loss": 0.8724, + "step": 6468 + }, + { + "epoch": 0.2864195344409639, + "grad_norm": 1.4876738786697388, + "learning_rate": 1.7956994403036848e-05, + "loss": 0.8256, + "step": 6469 + }, + { + "epoch": 0.28646381014577776, + "grad_norm": 1.353580355644226, + "learning_rate": 1.795605824125976e-05, + "loss": 0.7612, + "step": 6470 + }, + { + "epoch": 0.2865080858505916, + "grad_norm": 1.1406835317611694, + "learning_rate": 1.79551218894587e-05, + "loss": 0.8224, + "step": 6471 + }, + { + "epoch": 0.28655236155540553, + "grad_norm": 1.4606519937515259, + "learning_rate": 1.795418534765603e-05, + "loss": 0.8344, + "step": 6472 + }, + { + "epoch": 0.2865966372602194, + "grad_norm": 1.2383480072021484, + "learning_rate": 1.7953248615874123e-05, + "loss": 0.7612, + "step": 6473 + }, + { + "epoch": 0.28664091296503325, + "grad_norm": 1.3006361722946167, + "learning_rate": 1.7952311694135346e-05, + "loss": 0.8422, + "step": 6474 + }, + { + "epoch": 0.28668518866984716, + "grad_norm": 1.4747169017791748, + "learning_rate": 1.7951374582462082e-05, + "loss": 0.8285, + "step": 6475 + }, + { + "epoch": 0.286729464374661, + "grad_norm": 1.4060360193252563, + "learning_rate": 1.7950437280876706e-05, + "loss": 0.8545, + "step": 6476 + }, + { + "epoch": 0.28677374007947487, + "grad_norm": 1.4231971502304077, + "learning_rate": 1.7949499789401614e-05, + "loss": 0.8756, + "step": 6477 + }, + { + "epoch": 0.2868180157842888, + "grad_norm": 1.2181951999664307, + "learning_rate": 1.7948562108059187e-05, + "loss": 0.7632, + "step": 6478 + }, + { + "epoch": 0.28686229148910264, + "grad_norm": 1.286086082458496, + "learning_rate": 1.7947624236871828e-05, + "loss": 0.8146, + "step": 6479 + }, + { + "epoch": 0.2869065671939165, + "grad_norm": 1.397870421409607, + "learning_rate": 1.7946686175861936e-05, + "loss": 0.8575, + "step": 6480 + }, + { + "epoch": 0.2869508428987304, + "grad_norm": 1.7850943803787231, + "learning_rate": 1.7945747925051914e-05, + "loss": 0.7912, + "step": 6481 + }, + { + "epoch": 0.28699511860354426, + "grad_norm": 1.8932955265045166, + "learning_rate": 1.7944809484464173e-05, + "loss": 0.9247, + "step": 6482 + }, + { + "epoch": 0.2870393943083581, + "grad_norm": 1.5976983308792114, + "learning_rate": 1.7943870854121126e-05, + "loss": 0.8992, + "step": 6483 + }, + { + "epoch": 0.28708367001317203, + "grad_norm": 1.4696595668792725, + "learning_rate": 1.7942932034045193e-05, + "loss": 0.8186, + "step": 6484 + }, + { + "epoch": 0.2871279457179859, + "grad_norm": 1.5832287073135376, + "learning_rate": 1.7941993024258796e-05, + "loss": 0.8372, + "step": 6485 + }, + { + "epoch": 0.2871722214227998, + "grad_norm": 1.1969281435012817, + "learning_rate": 1.794105382478436e-05, + "loss": 0.7893, + "step": 6486 + }, + { + "epoch": 0.28721649712761366, + "grad_norm": 1.5229817628860474, + "learning_rate": 1.794011443564432e-05, + "loss": 0.9275, + "step": 6487 + }, + { + "epoch": 0.2872607728324275, + "grad_norm": 1.9190984964370728, + "learning_rate": 1.7939174856861116e-05, + "loss": 1.0249, + "step": 6488 + }, + { + "epoch": 0.2873050485372414, + "grad_norm": 1.332187533378601, + "learning_rate": 1.7938235088457178e-05, + "loss": 0.7955, + "step": 6489 + }, + { + "epoch": 0.2873493242420553, + "grad_norm": 1.358031988143921, + "learning_rate": 1.7937295130454964e-05, + "loss": 0.8623, + "step": 6490 + }, + { + "epoch": 0.28739359994686914, + "grad_norm": 1.1948117017745972, + "learning_rate": 1.7936354982876914e-05, + "loss": 0.8258, + "step": 6491 + }, + { + "epoch": 0.28743787565168305, + "grad_norm": 1.4520134925842285, + "learning_rate": 1.793541464574549e-05, + "loss": 0.8567, + "step": 6492 + }, + { + "epoch": 0.2874821513564969, + "grad_norm": 1.6296933889389038, + "learning_rate": 1.793447411908315e-05, + "loss": 0.8581, + "step": 6493 + }, + { + "epoch": 0.28752642706131076, + "grad_norm": 1.5922417640686035, + "learning_rate": 1.7933533402912354e-05, + "loss": 0.8906, + "step": 6494 + }, + { + "epoch": 0.2875707027661247, + "grad_norm": 1.3955620527267456, + "learning_rate": 1.793259249725557e-05, + "loss": 0.8305, + "step": 6495 + }, + { + "epoch": 0.28761497847093853, + "grad_norm": 1.2467457056045532, + "learning_rate": 1.7931651402135277e-05, + "loss": 0.7742, + "step": 6496 + }, + { + "epoch": 0.2876592541757524, + "grad_norm": 1.4867291450500488, + "learning_rate": 1.7930710117573945e-05, + "loss": 0.8578, + "step": 6497 + }, + { + "epoch": 0.2877035298805663, + "grad_norm": 1.3852418661117554, + "learning_rate": 1.7929768643594063e-05, + "loss": 0.9398, + "step": 6498 + }, + { + "epoch": 0.28774780558538016, + "grad_norm": 1.5269279479980469, + "learning_rate": 1.7928826980218112e-05, + "loss": 0.916, + "step": 6499 + }, + { + "epoch": 0.287792081290194, + "grad_norm": 1.1889393329620361, + "learning_rate": 1.7927885127468586e-05, + "loss": 0.7109, + "step": 6500 + }, + { + "epoch": 0.2878363569950079, + "grad_norm": 1.9951242208480835, + "learning_rate": 1.7926943085367975e-05, + "loss": 0.8146, + "step": 6501 + }, + { + "epoch": 0.2878806326998218, + "grad_norm": 1.5464977025985718, + "learning_rate": 1.7926000853938784e-05, + "loss": 0.8753, + "step": 6502 + }, + { + "epoch": 0.2879249084046357, + "grad_norm": 3.2209107875823975, + "learning_rate": 1.7925058433203523e-05, + "loss": 0.8862, + "step": 6503 + }, + { + "epoch": 0.28796918410944955, + "grad_norm": 1.4017646312713623, + "learning_rate": 1.7924115823184688e-05, + "loss": 0.8051, + "step": 6504 + }, + { + "epoch": 0.2880134598142634, + "grad_norm": 1.549252986907959, + "learning_rate": 1.7923173023904795e-05, + "loss": 0.9602, + "step": 6505 + }, + { + "epoch": 0.2880577355190773, + "grad_norm": 1.5458720922470093, + "learning_rate": 1.792223003538637e-05, + "loss": 0.877, + "step": 6506 + }, + { + "epoch": 0.2881020112238912, + "grad_norm": 1.4746556282043457, + "learning_rate": 1.792128685765193e-05, + "loss": 0.9152, + "step": 6507 + }, + { + "epoch": 0.28814628692870503, + "grad_norm": 1.6812831163406372, + "learning_rate": 1.7920343490724008e-05, + "loss": 0.8471, + "step": 6508 + }, + { + "epoch": 0.28819056263351894, + "grad_norm": 1.5135631561279297, + "learning_rate": 1.7919399934625126e-05, + "loss": 0.9234, + "step": 6509 + }, + { + "epoch": 0.2882348383383328, + "grad_norm": 1.2279512882232666, + "learning_rate": 1.7918456189377825e-05, + "loss": 0.8124, + "step": 6510 + }, + { + "epoch": 0.28827911404314666, + "grad_norm": 1.1365904808044434, + "learning_rate": 1.7917512255004644e-05, + "loss": 0.7603, + "step": 6511 + }, + { + "epoch": 0.28832338974796057, + "grad_norm": 1.2249051332473755, + "learning_rate": 1.7916568131528133e-05, + "loss": 0.8282, + "step": 6512 + }, + { + "epoch": 0.2883676654527744, + "grad_norm": 1.2147574424743652, + "learning_rate": 1.7915623818970835e-05, + "loss": 0.8168, + "step": 6513 + }, + { + "epoch": 0.2884119411575883, + "grad_norm": 1.4438644647598267, + "learning_rate": 1.791467931735531e-05, + "loss": 0.9093, + "step": 6514 + }, + { + "epoch": 0.2884562168624022, + "grad_norm": 1.9196289777755737, + "learning_rate": 1.791373462670411e-05, + "loss": 0.7916, + "step": 6515 + }, + { + "epoch": 0.28850049256721605, + "grad_norm": 1.378641963005066, + "learning_rate": 1.7912789747039805e-05, + "loss": 0.8973, + "step": 6516 + }, + { + "epoch": 0.2885447682720299, + "grad_norm": 1.3889812231063843, + "learning_rate": 1.791184467838496e-05, + "loss": 0.8208, + "step": 6517 + }, + { + "epoch": 0.2885890439768438, + "grad_norm": 1.3861018419265747, + "learning_rate": 1.7910899420762148e-05, + "loss": 0.9182, + "step": 6518 + }, + { + "epoch": 0.2886333196816577, + "grad_norm": 1.3905457258224487, + "learning_rate": 1.7909953974193942e-05, + "loss": 0.8698, + "step": 6519 + }, + { + "epoch": 0.2886775953864716, + "grad_norm": 1.3325575590133667, + "learning_rate": 1.790900833870293e-05, + "loss": 0.8867, + "step": 6520 + }, + { + "epoch": 0.28872187109128544, + "grad_norm": 1.392702341079712, + "learning_rate": 1.7908062514311685e-05, + "loss": 0.96, + "step": 6521 + }, + { + "epoch": 0.2887661467960993, + "grad_norm": 1.4429328441619873, + "learning_rate": 1.790711650104281e-05, + "loss": 0.9216, + "step": 6522 + }, + { + "epoch": 0.2888104225009132, + "grad_norm": 1.78301203250885, + "learning_rate": 1.79061702989189e-05, + "loss": 1.0211, + "step": 6523 + }, + { + "epoch": 0.28885469820572707, + "grad_norm": 1.2760611772537231, + "learning_rate": 1.7905223907962543e-05, + "loss": 0.832, + "step": 6524 + }, + { + "epoch": 0.2888989739105409, + "grad_norm": 1.5356110334396362, + "learning_rate": 1.7904277328196357e-05, + "loss": 0.8657, + "step": 6525 + }, + { + "epoch": 0.28894324961535484, + "grad_norm": 1.3007959127426147, + "learning_rate": 1.7903330559642936e-05, + "loss": 0.8011, + "step": 6526 + }, + { + "epoch": 0.2889875253201687, + "grad_norm": 1.1607351303100586, + "learning_rate": 1.7902383602324905e-05, + "loss": 0.7767, + "step": 6527 + }, + { + "epoch": 0.28903180102498255, + "grad_norm": 1.2976102828979492, + "learning_rate": 1.790143645626487e-05, + "loss": 0.7669, + "step": 6528 + }, + { + "epoch": 0.28907607672979646, + "grad_norm": 1.1460410356521606, + "learning_rate": 1.7900489121485466e-05, + "loss": 0.814, + "step": 6529 + }, + { + "epoch": 0.2891203524346103, + "grad_norm": 1.7718168497085571, + "learning_rate": 1.7899541598009305e-05, + "loss": 0.9394, + "step": 6530 + }, + { + "epoch": 0.2891646281394242, + "grad_norm": 1.5130455493927002, + "learning_rate": 1.789859388585903e-05, + "loss": 0.8761, + "step": 6531 + }, + { + "epoch": 0.2892089038442381, + "grad_norm": 1.3073232173919678, + "learning_rate": 1.7897645985057274e-05, + "loss": 0.8982, + "step": 6532 + }, + { + "epoch": 0.28925317954905194, + "grad_norm": 1.509218454360962, + "learning_rate": 1.7896697895626668e-05, + "loss": 0.8977, + "step": 6533 + }, + { + "epoch": 0.2892974552538658, + "grad_norm": 1.418595790863037, + "learning_rate": 1.7895749617589865e-05, + "loss": 0.8716, + "step": 6534 + }, + { + "epoch": 0.2893417309586797, + "grad_norm": 1.4601387977600098, + "learning_rate": 1.789480115096951e-05, + "loss": 0.879, + "step": 6535 + }, + { + "epoch": 0.28938600666349357, + "grad_norm": 1.1770604848861694, + "learning_rate": 1.789385249578826e-05, + "loss": 0.7969, + "step": 6536 + }, + { + "epoch": 0.2894302823683074, + "grad_norm": 1.6801031827926636, + "learning_rate": 1.7892903652068766e-05, + "loss": 0.8565, + "step": 6537 + }, + { + "epoch": 0.28947455807312134, + "grad_norm": 1.3300706148147583, + "learning_rate": 1.78919546198337e-05, + "loss": 0.8673, + "step": 6538 + }, + { + "epoch": 0.2895188337779352, + "grad_norm": 1.2618533372879028, + "learning_rate": 1.789100539910572e-05, + "loss": 0.8396, + "step": 6539 + }, + { + "epoch": 0.2895631094827491, + "grad_norm": 1.460092306137085, + "learning_rate": 1.7890055989907504e-05, + "loss": 0.9349, + "step": 6540 + }, + { + "epoch": 0.28960738518756296, + "grad_norm": 1.1634434461593628, + "learning_rate": 1.7889106392261725e-05, + "loss": 0.745, + "step": 6541 + }, + { + "epoch": 0.2896516608923768, + "grad_norm": 1.215687870979309, + "learning_rate": 1.7888156606191064e-05, + "loss": 0.8279, + "step": 6542 + }, + { + "epoch": 0.28969593659719073, + "grad_norm": 1.5560652017593384, + "learning_rate": 1.7887206631718202e-05, + "loss": 0.8401, + "step": 6543 + }, + { + "epoch": 0.2897402123020046, + "grad_norm": 1.6919785737991333, + "learning_rate": 1.7886256468865833e-05, + "loss": 0.8563, + "step": 6544 + }, + { + "epoch": 0.28978448800681844, + "grad_norm": 1.4419811964035034, + "learning_rate": 1.788530611765665e-05, + "loss": 0.7955, + "step": 6545 + }, + { + "epoch": 0.28982876371163235, + "grad_norm": 1.5359998941421509, + "learning_rate": 1.788435557811335e-05, + "loss": 0.9278, + "step": 6546 + }, + { + "epoch": 0.2898730394164462, + "grad_norm": 1.2929645776748657, + "learning_rate": 1.788340485025864e-05, + "loss": 0.8689, + "step": 6547 + }, + { + "epoch": 0.28991731512126007, + "grad_norm": 1.219557762145996, + "learning_rate": 1.788245393411522e-05, + "loss": 0.8118, + "step": 6548 + }, + { + "epoch": 0.289961590826074, + "grad_norm": 1.7398416996002197, + "learning_rate": 1.7881502829705808e-05, + "loss": 0.9714, + "step": 6549 + }, + { + "epoch": 0.29000586653088783, + "grad_norm": 1.611107587814331, + "learning_rate": 1.7880551537053118e-05, + "loss": 0.913, + "step": 6550 + }, + { + "epoch": 0.2900501422357017, + "grad_norm": 1.5659347772598267, + "learning_rate": 1.7879600056179868e-05, + "loss": 0.9024, + "step": 6551 + }, + { + "epoch": 0.2900944179405156, + "grad_norm": 1.3783994913101196, + "learning_rate": 1.7878648387108792e-05, + "loss": 0.87, + "step": 6552 + }, + { + "epoch": 0.29013869364532946, + "grad_norm": 1.0545761585235596, + "learning_rate": 1.7877696529862608e-05, + "loss": 0.7819, + "step": 6553 + }, + { + "epoch": 0.2901829693501433, + "grad_norm": 1.1280931234359741, + "learning_rate": 1.787674448446406e-05, + "loss": 0.7739, + "step": 6554 + }, + { + "epoch": 0.29022724505495723, + "grad_norm": 1.4357470273971558, + "learning_rate": 1.7875792250935883e-05, + "loss": 0.9115, + "step": 6555 + }, + { + "epoch": 0.2902715207597711, + "grad_norm": 1.2698441743850708, + "learning_rate": 1.7874839829300824e-05, + "loss": 0.889, + "step": 6556 + }, + { + "epoch": 0.290315796464585, + "grad_norm": 1.606002926826477, + "learning_rate": 1.7873887219581626e-05, + "loss": 0.9333, + "step": 6557 + }, + { + "epoch": 0.29036007216939885, + "grad_norm": 1.1404885053634644, + "learning_rate": 1.787293442180104e-05, + "loss": 0.759, + "step": 6558 + }, + { + "epoch": 0.2904043478742127, + "grad_norm": 1.3754891157150269, + "learning_rate": 1.787198143598183e-05, + "loss": 0.8825, + "step": 6559 + }, + { + "epoch": 0.2904486235790266, + "grad_norm": 1.233973741531372, + "learning_rate": 1.787102826214675e-05, + "loss": 0.836, + "step": 6560 + }, + { + "epoch": 0.2904928992838405, + "grad_norm": 1.2980512380599976, + "learning_rate": 1.787007490031857e-05, + "loss": 0.7901, + "step": 6561 + }, + { + "epoch": 0.29053717498865433, + "grad_norm": 1.2419970035552979, + "learning_rate": 1.7869121350520057e-05, + "loss": 0.7883, + "step": 6562 + }, + { + "epoch": 0.29058145069346825, + "grad_norm": 1.535402774810791, + "learning_rate": 1.786816761277399e-05, + "loss": 0.9491, + "step": 6563 + }, + { + "epoch": 0.2906257263982821, + "grad_norm": 1.3280842304229736, + "learning_rate": 1.7867213687103145e-05, + "loss": 0.8419, + "step": 6564 + }, + { + "epoch": 0.29067000210309596, + "grad_norm": 1.232158899307251, + "learning_rate": 1.7866259573530308e-05, + "loss": 0.8191, + "step": 6565 + }, + { + "epoch": 0.29071427780790987, + "grad_norm": 1.23395836353302, + "learning_rate": 1.7865305272078265e-05, + "loss": 0.7594, + "step": 6566 + }, + { + "epoch": 0.2907585535127237, + "grad_norm": 1.574276328086853, + "learning_rate": 1.786435078276981e-05, + "loss": 0.881, + "step": 6567 + }, + { + "epoch": 0.2908028292175376, + "grad_norm": 1.8300648927688599, + "learning_rate": 1.7863396105627737e-05, + "loss": 0.9814, + "step": 6568 + }, + { + "epoch": 0.2908471049223515, + "grad_norm": 1.3673807382583618, + "learning_rate": 1.7862441240674857e-05, + "loss": 0.8736, + "step": 6569 + }, + { + "epoch": 0.29089138062716535, + "grad_norm": 1.3313480615615845, + "learning_rate": 1.786148618793396e-05, + "loss": 0.8232, + "step": 6570 + }, + { + "epoch": 0.2909356563319792, + "grad_norm": 1.4674886465072632, + "learning_rate": 1.7860530947427878e-05, + "loss": 0.8171, + "step": 6571 + }, + { + "epoch": 0.2909799320367931, + "grad_norm": 1.45835542678833, + "learning_rate": 1.7859575519179405e-05, + "loss": 0.9021, + "step": 6572 + }, + { + "epoch": 0.291024207741607, + "grad_norm": 1.8599165678024292, + "learning_rate": 1.7858619903211378e-05, + "loss": 0.9281, + "step": 6573 + }, + { + "epoch": 0.2910684834464209, + "grad_norm": 1.4234367609024048, + "learning_rate": 1.7857664099546607e-05, + "loss": 0.8312, + "step": 6574 + }, + { + "epoch": 0.29111275915123475, + "grad_norm": 1.2706257104873657, + "learning_rate": 1.785670810820793e-05, + "loss": 0.7921, + "step": 6575 + }, + { + "epoch": 0.2911570348560486, + "grad_norm": 1.1215506792068481, + "learning_rate": 1.7855751929218178e-05, + "loss": 0.7593, + "step": 6576 + }, + { + "epoch": 0.2912013105608625, + "grad_norm": 1.4441052675247192, + "learning_rate": 1.785479556260019e-05, + "loss": 0.8555, + "step": 6577 + }, + { + "epoch": 0.29124558626567637, + "grad_norm": 1.2029845714569092, + "learning_rate": 1.7853839008376802e-05, + "loss": 0.7781, + "step": 6578 + }, + { + "epoch": 0.2912898619704902, + "grad_norm": 1.8112080097198486, + "learning_rate": 1.785288226657086e-05, + "loss": 1.039, + "step": 6579 + }, + { + "epoch": 0.29133413767530414, + "grad_norm": 1.3477922677993774, + "learning_rate": 1.7851925337205225e-05, + "loss": 0.8752, + "step": 6580 + }, + { + "epoch": 0.291378413380118, + "grad_norm": 1.1033658981323242, + "learning_rate": 1.7850968220302744e-05, + "loss": 0.7604, + "step": 6581 + }, + { + "epoch": 0.29142268908493185, + "grad_norm": 1.1764358282089233, + "learning_rate": 1.785001091588628e-05, + "loss": 0.7802, + "step": 6582 + }, + { + "epoch": 0.29146696478974576, + "grad_norm": 1.3028903007507324, + "learning_rate": 1.78490534239787e-05, + "loss": 0.8086, + "step": 6583 + }, + { + "epoch": 0.2915112404945596, + "grad_norm": 1.366547703742981, + "learning_rate": 1.7848095744602868e-05, + "loss": 0.7975, + "step": 6584 + }, + { + "epoch": 0.2915555161993735, + "grad_norm": 1.3412799835205078, + "learning_rate": 1.7847137877781655e-05, + "loss": 0.809, + "step": 6585 + }, + { + "epoch": 0.2915997919041874, + "grad_norm": 1.4987796545028687, + "learning_rate": 1.784617982353795e-05, + "loss": 0.9168, + "step": 6586 + }, + { + "epoch": 0.29164406760900125, + "grad_norm": 1.8540438413619995, + "learning_rate": 1.7845221581894624e-05, + "loss": 0.8713, + "step": 6587 + }, + { + "epoch": 0.2916883433138151, + "grad_norm": 1.3980026245117188, + "learning_rate": 1.784426315287457e-05, + "loss": 0.8557, + "step": 6588 + }, + { + "epoch": 0.291732619018629, + "grad_norm": 1.2200849056243896, + "learning_rate": 1.7843304536500677e-05, + "loss": 0.7628, + "step": 6589 + }, + { + "epoch": 0.29177689472344287, + "grad_norm": 1.1193214654922485, + "learning_rate": 1.784234573279584e-05, + "loss": 0.7266, + "step": 6590 + }, + { + "epoch": 0.2918211704282568, + "grad_norm": 1.4470714330673218, + "learning_rate": 1.784138674178296e-05, + "loss": 0.9074, + "step": 6591 + }, + { + "epoch": 0.29186544613307064, + "grad_norm": 1.3329110145568848, + "learning_rate": 1.7840427563484948e-05, + "loss": 0.8399, + "step": 6592 + }, + { + "epoch": 0.2919097218378845, + "grad_norm": 1.4021633863449097, + "learning_rate": 1.78394681979247e-05, + "loss": 0.8805, + "step": 6593 + }, + { + "epoch": 0.2919539975426984, + "grad_norm": 1.3458735942840576, + "learning_rate": 1.7838508645125142e-05, + "loss": 0.8365, + "step": 6594 + }, + { + "epoch": 0.29199827324751226, + "grad_norm": 1.3415687084197998, + "learning_rate": 1.783754890510919e-05, + "loss": 0.9192, + "step": 6595 + }, + { + "epoch": 0.2920425489523261, + "grad_norm": 1.3734452724456787, + "learning_rate": 1.783658897789976e-05, + "loss": 0.864, + "step": 6596 + }, + { + "epoch": 0.29208682465714003, + "grad_norm": 1.734406590461731, + "learning_rate": 1.7835628863519785e-05, + "loss": 0.8733, + "step": 6597 + }, + { + "epoch": 0.2921311003619539, + "grad_norm": 1.3465677499771118, + "learning_rate": 1.783466856199219e-05, + "loss": 0.8058, + "step": 6598 + }, + { + "epoch": 0.29217537606676774, + "grad_norm": 1.3133130073547363, + "learning_rate": 1.7833708073339924e-05, + "loss": 0.8695, + "step": 6599 + }, + { + "epoch": 0.29221965177158166, + "grad_norm": 1.446542501449585, + "learning_rate": 1.7832747397585913e-05, + "loss": 0.8454, + "step": 6600 + }, + { + "epoch": 0.2922639274763955, + "grad_norm": 1.458886742591858, + "learning_rate": 1.783178653475311e-05, + "loss": 0.8595, + "step": 6601 + }, + { + "epoch": 0.29230820318120937, + "grad_norm": 1.6459448337554932, + "learning_rate": 1.783082548486446e-05, + "loss": 0.8177, + "step": 6602 + }, + { + "epoch": 0.2923524788860233, + "grad_norm": 1.343145728111267, + "learning_rate": 1.782986424794292e-05, + "loss": 0.8372, + "step": 6603 + }, + { + "epoch": 0.29239675459083714, + "grad_norm": 1.1539469957351685, + "learning_rate": 1.782890282401145e-05, + "loss": 0.7765, + "step": 6604 + }, + { + "epoch": 0.292441030295651, + "grad_norm": 1.3577816486358643, + "learning_rate": 1.782794121309301e-05, + "loss": 0.8747, + "step": 6605 + }, + { + "epoch": 0.2924853060004649, + "grad_norm": 1.3887529373168945, + "learning_rate": 1.7826979415210567e-05, + "loss": 0.8427, + "step": 6606 + }, + { + "epoch": 0.29252958170527876, + "grad_norm": 1.1936099529266357, + "learning_rate": 1.7826017430387095e-05, + "loss": 0.781, + "step": 6607 + }, + { + "epoch": 0.2925738574100926, + "grad_norm": 1.726426601409912, + "learning_rate": 1.7825055258645566e-05, + "loss": 0.894, + "step": 6608 + }, + { + "epoch": 0.29261813311490653, + "grad_norm": 1.1785260438919067, + "learning_rate": 1.7824092900008965e-05, + "loss": 0.8393, + "step": 6609 + }, + { + "epoch": 0.2926624088197204, + "grad_norm": 1.683465838432312, + "learning_rate": 1.7823130354500277e-05, + "loss": 0.8263, + "step": 6610 + }, + { + "epoch": 0.2927066845245343, + "grad_norm": 1.3537431955337524, + "learning_rate": 1.7822167622142488e-05, + "loss": 0.8092, + "step": 6611 + }, + { + "epoch": 0.29275096022934816, + "grad_norm": 1.7608052492141724, + "learning_rate": 1.7821204702958598e-05, + "loss": 0.8143, + "step": 6612 + }, + { + "epoch": 0.292795235934162, + "grad_norm": 1.36228609085083, + "learning_rate": 1.78202415969716e-05, + "loss": 0.8448, + "step": 6613 + }, + { + "epoch": 0.2928395116389759, + "grad_norm": 1.5764185190200806, + "learning_rate": 1.7819278304204497e-05, + "loss": 0.8741, + "step": 6614 + }, + { + "epoch": 0.2928837873437898, + "grad_norm": 1.4322444200515747, + "learning_rate": 1.78183148246803e-05, + "loss": 0.9697, + "step": 6615 + }, + { + "epoch": 0.29292806304860364, + "grad_norm": 1.2924330234527588, + "learning_rate": 1.781735115842202e-05, + "loss": 0.881, + "step": 6616 + }, + { + "epoch": 0.29297233875341755, + "grad_norm": 1.2616850137710571, + "learning_rate": 1.781638730545267e-05, + "loss": 0.8446, + "step": 6617 + }, + { + "epoch": 0.2930166144582314, + "grad_norm": 1.2788403034210205, + "learning_rate": 1.7815423265795275e-05, + "loss": 0.8519, + "step": 6618 + }, + { + "epoch": 0.29306089016304526, + "grad_norm": 1.9088025093078613, + "learning_rate": 1.781445903947286e-05, + "loss": 0.9558, + "step": 6619 + }, + { + "epoch": 0.2931051658678592, + "grad_norm": 1.3524526357650757, + "learning_rate": 1.7813494626508453e-05, + "loss": 0.8196, + "step": 6620 + }, + { + "epoch": 0.29314944157267303, + "grad_norm": 1.1984972953796387, + "learning_rate": 1.781253002692509e-05, + "loss": 0.8501, + "step": 6621 + }, + { + "epoch": 0.2931937172774869, + "grad_norm": 1.3836582899093628, + "learning_rate": 1.7811565240745807e-05, + "loss": 0.8344, + "step": 6622 + }, + { + "epoch": 0.2932379929823008, + "grad_norm": 1.4451950788497925, + "learning_rate": 1.781060026799365e-05, + "loss": 0.8196, + "step": 6623 + }, + { + "epoch": 0.29328226868711466, + "grad_norm": 1.1711866855621338, + "learning_rate": 1.7809635108691663e-05, + "loss": 0.7819, + "step": 6624 + }, + { + "epoch": 0.2933265443919285, + "grad_norm": 1.0778621435165405, + "learning_rate": 1.7808669762862903e-05, + "loss": 0.7713, + "step": 6625 + }, + { + "epoch": 0.2933708200967424, + "grad_norm": 1.332505702972412, + "learning_rate": 1.780770423053042e-05, + "loss": 0.9047, + "step": 6626 + }, + { + "epoch": 0.2934150958015563, + "grad_norm": 1.2120169401168823, + "learning_rate": 1.780673851171728e-05, + "loss": 0.8236, + "step": 6627 + }, + { + "epoch": 0.2934593715063702, + "grad_norm": 1.8183274269104004, + "learning_rate": 1.780577260644655e-05, + "loss": 0.8768, + "step": 6628 + }, + { + "epoch": 0.29350364721118405, + "grad_norm": 1.4242660999298096, + "learning_rate": 1.7804806514741294e-05, + "loss": 0.8826, + "step": 6629 + }, + { + "epoch": 0.2935479229159979, + "grad_norm": 1.3818581104278564, + "learning_rate": 1.7803840236624593e-05, + "loss": 0.8589, + "step": 6630 + }, + { + "epoch": 0.2935921986208118, + "grad_norm": 1.5211251974105835, + "learning_rate": 1.780287377211952e-05, + "loss": 0.8271, + "step": 6631 + }, + { + "epoch": 0.2936364743256257, + "grad_norm": 1.218927025794983, + "learning_rate": 1.7801907121249157e-05, + "loss": 0.7684, + "step": 6632 + }, + { + "epoch": 0.29368075003043953, + "grad_norm": 1.3582806587219238, + "learning_rate": 1.78009402840366e-05, + "loss": 0.8169, + "step": 6633 + }, + { + "epoch": 0.29372502573525344, + "grad_norm": 1.370643138885498, + "learning_rate": 1.7799973260504934e-05, + "loss": 0.8349, + "step": 6634 + }, + { + "epoch": 0.2937693014400673, + "grad_norm": 1.4669443368911743, + "learning_rate": 1.7799006050677256e-05, + "loss": 0.7927, + "step": 6635 + }, + { + "epoch": 0.29381357714488116, + "grad_norm": 1.5534758567810059, + "learning_rate": 1.779803865457667e-05, + "loss": 0.8803, + "step": 6636 + }, + { + "epoch": 0.29385785284969507, + "grad_norm": 1.394559621810913, + "learning_rate": 1.779707107222628e-05, + "loss": 0.8539, + "step": 6637 + }, + { + "epoch": 0.2939021285545089, + "grad_norm": 2.076479196548462, + "learning_rate": 1.7796103303649194e-05, + "loss": 0.9868, + "step": 6638 + }, + { + "epoch": 0.2939464042593228, + "grad_norm": 1.3398234844207764, + "learning_rate": 1.779513534886853e-05, + "loss": 0.8051, + "step": 6639 + }, + { + "epoch": 0.2939906799641367, + "grad_norm": 1.4024293422698975, + "learning_rate": 1.7794167207907408e-05, + "loss": 0.8907, + "step": 6640 + }, + { + "epoch": 0.29403495566895055, + "grad_norm": 1.135053038597107, + "learning_rate": 1.7793198880788947e-05, + "loss": 0.7999, + "step": 6641 + }, + { + "epoch": 0.2940792313737644, + "grad_norm": 1.5491163730621338, + "learning_rate": 1.7792230367536277e-05, + "loss": 0.8374, + "step": 6642 + }, + { + "epoch": 0.2941235070785783, + "grad_norm": 1.4976543188095093, + "learning_rate": 1.779126166817253e-05, + "loss": 0.8391, + "step": 6643 + }, + { + "epoch": 0.2941677827833922, + "grad_norm": 1.4184013605117798, + "learning_rate": 1.779029278272084e-05, + "loss": 0.9322, + "step": 6644 + }, + { + "epoch": 0.2942120584882061, + "grad_norm": 1.3169615268707275, + "learning_rate": 1.778932371120435e-05, + "loss": 0.8293, + "step": 6645 + }, + { + "epoch": 0.29425633419301994, + "grad_norm": 1.3973324298858643, + "learning_rate": 1.7788354453646203e-05, + "loss": 0.9055, + "step": 6646 + }, + { + "epoch": 0.2943006098978338, + "grad_norm": 1.3677942752838135, + "learning_rate": 1.7787385010069552e-05, + "loss": 0.8394, + "step": 6647 + }, + { + "epoch": 0.2943448856026477, + "grad_norm": 1.22247314453125, + "learning_rate": 1.778641538049755e-05, + "loss": 0.8086, + "step": 6648 + }, + { + "epoch": 0.29438916130746157, + "grad_norm": 1.4698187112808228, + "learning_rate": 1.778544556495336e-05, + "loss": 0.8618, + "step": 6649 + }, + { + "epoch": 0.2944334370122754, + "grad_norm": 1.2679855823516846, + "learning_rate": 1.778447556346014e-05, + "loss": 0.8459, + "step": 6650 + }, + { + "epoch": 0.29447771271708933, + "grad_norm": 1.5302324295043945, + "learning_rate": 1.7783505376041063e-05, + "loss": 0.8712, + "step": 6651 + }, + { + "epoch": 0.2945219884219032, + "grad_norm": 1.6123117208480835, + "learning_rate": 1.7782535002719298e-05, + "loss": 0.8998, + "step": 6652 + }, + { + "epoch": 0.29456626412671705, + "grad_norm": 1.21523118019104, + "learning_rate": 1.7781564443518016e-05, + "loss": 0.8042, + "step": 6653 + }, + { + "epoch": 0.29461053983153096, + "grad_norm": 1.436232328414917, + "learning_rate": 1.7780593698460405e-05, + "loss": 0.8295, + "step": 6654 + }, + { + "epoch": 0.2946548155363448, + "grad_norm": 1.3444247245788574, + "learning_rate": 1.777962276756965e-05, + "loss": 0.8408, + "step": 6655 + }, + { + "epoch": 0.2946990912411587, + "grad_norm": 1.2771906852722168, + "learning_rate": 1.777865165086894e-05, + "loss": 0.833, + "step": 6656 + }, + { + "epoch": 0.2947433669459726, + "grad_norm": 1.2633644342422485, + "learning_rate": 1.7777680348381466e-05, + "loss": 0.7965, + "step": 6657 + }, + { + "epoch": 0.29478764265078644, + "grad_norm": 1.618053674697876, + "learning_rate": 1.7776708860130432e-05, + "loss": 0.9191, + "step": 6658 + }, + { + "epoch": 0.2948319183556003, + "grad_norm": 1.3877936601638794, + "learning_rate": 1.777573718613904e-05, + "loss": 0.7914, + "step": 6659 + }, + { + "epoch": 0.2948761940604142, + "grad_norm": 1.4852423667907715, + "learning_rate": 1.7774765326430495e-05, + "loss": 0.8347, + "step": 6660 + }, + { + "epoch": 0.29492046976522807, + "grad_norm": 1.3301794528961182, + "learning_rate": 1.777379328102801e-05, + "loss": 0.8746, + "step": 6661 + }, + { + "epoch": 0.294964745470042, + "grad_norm": 1.3249154090881348, + "learning_rate": 1.77728210499548e-05, + "loss": 0.7667, + "step": 6662 + }, + { + "epoch": 0.29500902117485583, + "grad_norm": 1.4860776662826538, + "learning_rate": 1.7771848633234092e-05, + "loss": 0.9253, + "step": 6663 + }, + { + "epoch": 0.2950532968796697, + "grad_norm": 1.21734619140625, + "learning_rate": 1.7770876030889107e-05, + "loss": 0.782, + "step": 6664 + }, + { + "epoch": 0.2950975725844836, + "grad_norm": 1.7139674425125122, + "learning_rate": 1.7769903242943075e-05, + "loss": 0.8598, + "step": 6665 + }, + { + "epoch": 0.29514184828929746, + "grad_norm": 1.3033925294876099, + "learning_rate": 1.776893026941923e-05, + "loss": 0.8473, + "step": 6666 + }, + { + "epoch": 0.2951861239941113, + "grad_norm": 1.3675897121429443, + "learning_rate": 1.776795711034081e-05, + "loss": 0.88, + "step": 6667 + }, + { + "epoch": 0.2952303996989252, + "grad_norm": 1.7836382389068604, + "learning_rate": 1.776698376573106e-05, + "loss": 0.898, + "step": 6668 + }, + { + "epoch": 0.2952746754037391, + "grad_norm": 1.383327841758728, + "learning_rate": 1.7766010235613225e-05, + "loss": 0.8383, + "step": 6669 + }, + { + "epoch": 0.29531895110855294, + "grad_norm": 1.202959656715393, + "learning_rate": 1.7765036520010567e-05, + "loss": 0.7995, + "step": 6670 + }, + { + "epoch": 0.29536322681336685, + "grad_norm": 1.299691915512085, + "learning_rate": 1.7764062618946325e-05, + "loss": 0.853, + "step": 6671 + }, + { + "epoch": 0.2954075025181807, + "grad_norm": 1.211104154586792, + "learning_rate": 1.7763088532443772e-05, + "loss": 0.8564, + "step": 6672 + }, + { + "epoch": 0.29545177822299457, + "grad_norm": 1.1943554878234863, + "learning_rate": 1.7762114260526175e-05, + "loss": 0.8205, + "step": 6673 + }, + { + "epoch": 0.2954960539278085, + "grad_norm": 1.7359728813171387, + "learning_rate": 1.7761139803216795e-05, + "loss": 0.9023, + "step": 6674 + }, + { + "epoch": 0.29554032963262233, + "grad_norm": 1.3507044315338135, + "learning_rate": 1.7760165160538907e-05, + "loss": 0.8095, + "step": 6675 + }, + { + "epoch": 0.2955846053374362, + "grad_norm": 1.160443902015686, + "learning_rate": 1.7759190332515798e-05, + "loss": 0.7378, + "step": 6676 + }, + { + "epoch": 0.2956288810422501, + "grad_norm": 1.4607064723968506, + "learning_rate": 1.7758215319170742e-05, + "loss": 0.8781, + "step": 6677 + }, + { + "epoch": 0.29567315674706396, + "grad_norm": 1.3609389066696167, + "learning_rate": 1.775724012052703e-05, + "loss": 0.8864, + "step": 6678 + }, + { + "epoch": 0.29571743245187787, + "grad_norm": 1.4253273010253906, + "learning_rate": 1.775626473660796e-05, + "loss": 0.8685, + "step": 6679 + }, + { + "epoch": 0.2957617081566917, + "grad_norm": 1.3559401035308838, + "learning_rate": 1.7755289167436817e-05, + "loss": 0.8181, + "step": 6680 + }, + { + "epoch": 0.2958059838615056, + "grad_norm": 1.3648980855941772, + "learning_rate": 1.7754313413036907e-05, + "loss": 0.8534, + "step": 6681 + }, + { + "epoch": 0.2958502595663195, + "grad_norm": 2.177100896835327, + "learning_rate": 1.7753337473431535e-05, + "loss": 0.8779, + "step": 6682 + }, + { + "epoch": 0.29589453527113335, + "grad_norm": 1.3720964193344116, + "learning_rate": 1.7752361348644012e-05, + "loss": 0.8011, + "step": 6683 + }, + { + "epoch": 0.2959388109759472, + "grad_norm": 1.266248345375061, + "learning_rate": 1.775138503869765e-05, + "loss": 0.8111, + "step": 6684 + }, + { + "epoch": 0.2959830866807611, + "grad_norm": 1.2320220470428467, + "learning_rate": 1.7750408543615765e-05, + "loss": 0.8895, + "step": 6685 + }, + { + "epoch": 0.296027362385575, + "grad_norm": 1.6491750478744507, + "learning_rate": 1.7749431863421684e-05, + "loss": 0.8999, + "step": 6686 + }, + { + "epoch": 0.29607163809038883, + "grad_norm": 1.4483332633972168, + "learning_rate": 1.7748454998138735e-05, + "loss": 0.8449, + "step": 6687 + }, + { + "epoch": 0.29611591379520275, + "grad_norm": 1.1338322162628174, + "learning_rate": 1.7747477947790247e-05, + "loss": 0.8146, + "step": 6688 + }, + { + "epoch": 0.2961601895000166, + "grad_norm": 1.1137652397155762, + "learning_rate": 1.7746500712399554e-05, + "loss": 0.7258, + "step": 6689 + }, + { + "epoch": 0.29620446520483046, + "grad_norm": 1.551538348197937, + "learning_rate": 1.7745523291989998e-05, + "loss": 0.9256, + "step": 6690 + }, + { + "epoch": 0.29624874090964437, + "grad_norm": 1.3331279754638672, + "learning_rate": 1.774454568658493e-05, + "loss": 0.8241, + "step": 6691 + }, + { + "epoch": 0.2962930166144582, + "grad_norm": 1.5268898010253906, + "learning_rate": 1.774356789620769e-05, + "loss": 0.8292, + "step": 6692 + }, + { + "epoch": 0.2963372923192721, + "grad_norm": 1.3479098081588745, + "learning_rate": 1.7742589920881635e-05, + "loss": 0.8737, + "step": 6693 + }, + { + "epoch": 0.296381568024086, + "grad_norm": 1.3204083442687988, + "learning_rate": 1.7741611760630125e-05, + "loss": 0.84, + "step": 6694 + }, + { + "epoch": 0.29642584372889985, + "grad_norm": 1.4705307483673096, + "learning_rate": 1.7740633415476526e-05, + "loss": 0.9009, + "step": 6695 + }, + { + "epoch": 0.2964701194337137, + "grad_norm": 1.5991425514221191, + "learning_rate": 1.7739654885444194e-05, + "loss": 0.8888, + "step": 6696 + }, + { + "epoch": 0.2965143951385276, + "grad_norm": 1.5977237224578857, + "learning_rate": 1.773867617055651e-05, + "loss": 0.8951, + "step": 6697 + }, + { + "epoch": 0.2965586708433415, + "grad_norm": 2.0432589054107666, + "learning_rate": 1.7737697270836847e-05, + "loss": 1.0615, + "step": 6698 + }, + { + "epoch": 0.2966029465481554, + "grad_norm": 1.3200429677963257, + "learning_rate": 1.7736718186308585e-05, + "loss": 0.8091, + "step": 6699 + }, + { + "epoch": 0.29664722225296924, + "grad_norm": 1.2412668466567993, + "learning_rate": 1.7735738916995108e-05, + "loss": 0.8242, + "step": 6700 + }, + { + "epoch": 0.2966914979577831, + "grad_norm": 1.5922112464904785, + "learning_rate": 1.7734759462919805e-05, + "loss": 0.9376, + "step": 6701 + }, + { + "epoch": 0.296735773662597, + "grad_norm": 1.324497938156128, + "learning_rate": 1.773377982410607e-05, + "loss": 0.8499, + "step": 6702 + }, + { + "epoch": 0.29678004936741087, + "grad_norm": 1.1403626203536987, + "learning_rate": 1.7732800000577304e-05, + "loss": 0.8131, + "step": 6703 + }, + { + "epoch": 0.2968243250722247, + "grad_norm": 1.2313108444213867, + "learning_rate": 1.7731819992356903e-05, + "loss": 0.8388, + "step": 6704 + }, + { + "epoch": 0.29686860077703864, + "grad_norm": 1.2961174249649048, + "learning_rate": 1.773083979946828e-05, + "loss": 0.8223, + "step": 6705 + }, + { + "epoch": 0.2969128764818525, + "grad_norm": 1.6537913084030151, + "learning_rate": 1.7729859421934842e-05, + "loss": 0.8578, + "step": 6706 + }, + { + "epoch": 0.29695715218666635, + "grad_norm": 1.5264997482299805, + "learning_rate": 1.7728878859780002e-05, + "loss": 0.9178, + "step": 6707 + }, + { + "epoch": 0.29700142789148026, + "grad_norm": 1.654850959777832, + "learning_rate": 1.772789811302719e-05, + "loss": 1.0096, + "step": 6708 + }, + { + "epoch": 0.2970457035962941, + "grad_norm": 1.493537425994873, + "learning_rate": 1.772691718169982e-05, + "loss": 0.9657, + "step": 6709 + }, + { + "epoch": 0.297089979301108, + "grad_norm": 1.560776948928833, + "learning_rate": 1.7725936065821322e-05, + "loss": 0.8194, + "step": 6710 + }, + { + "epoch": 0.2971342550059219, + "grad_norm": 1.3507440090179443, + "learning_rate": 1.7724954765415137e-05, + "loss": 0.8795, + "step": 6711 + }, + { + "epoch": 0.29717853071073574, + "grad_norm": 1.2422932386398315, + "learning_rate": 1.7723973280504692e-05, + "loss": 0.8242, + "step": 6712 + }, + { + "epoch": 0.2972228064155496, + "grad_norm": 1.3756352663040161, + "learning_rate": 1.772299161111344e-05, + "loss": 0.8277, + "step": 6713 + }, + { + "epoch": 0.2972670821203635, + "grad_norm": 1.380137324333191, + "learning_rate": 1.7722009757264818e-05, + "loss": 0.8565, + "step": 6714 + }, + { + "epoch": 0.29731135782517737, + "grad_norm": 1.2982251644134521, + "learning_rate": 1.772102771898228e-05, + "loss": 0.8849, + "step": 6715 + }, + { + "epoch": 0.2973556335299913, + "grad_norm": 1.4835810661315918, + "learning_rate": 1.772004549628928e-05, + "loss": 0.891, + "step": 6716 + }, + { + "epoch": 0.29739990923480514, + "grad_norm": 1.3297474384307861, + "learning_rate": 1.7719063089209286e-05, + "loss": 0.8201, + "step": 6717 + }, + { + "epoch": 0.297444184939619, + "grad_norm": 1.1675490140914917, + "learning_rate": 1.771808049776575e-05, + "loss": 0.7798, + "step": 6718 + }, + { + "epoch": 0.2974884606444329, + "grad_norm": 1.4382200241088867, + "learning_rate": 1.7717097721982148e-05, + "loss": 0.9243, + "step": 6719 + }, + { + "epoch": 0.29753273634924676, + "grad_norm": 1.400828242301941, + "learning_rate": 1.7716114761881948e-05, + "loss": 0.8258, + "step": 6720 + }, + { + "epoch": 0.2975770120540606, + "grad_norm": 1.3652667999267578, + "learning_rate": 1.771513161748863e-05, + "loss": 0.9146, + "step": 6721 + }, + { + "epoch": 0.29762128775887453, + "grad_norm": 1.1137341260910034, + "learning_rate": 1.7714148288825677e-05, + "loss": 0.749, + "step": 6722 + }, + { + "epoch": 0.2976655634636884, + "grad_norm": 1.3830324411392212, + "learning_rate": 1.7713164775916573e-05, + "loss": 0.7896, + "step": 6723 + }, + { + "epoch": 0.29770983916850224, + "grad_norm": 1.6256170272827148, + "learning_rate": 1.7712181078784805e-05, + "loss": 0.8637, + "step": 6724 + }, + { + "epoch": 0.29775411487331616, + "grad_norm": 1.7425994873046875, + "learning_rate": 1.771119719745388e-05, + "loss": 0.9411, + "step": 6725 + }, + { + "epoch": 0.29779839057813, + "grad_norm": 1.6018116474151611, + "learning_rate": 1.7710213131947283e-05, + "loss": 0.9174, + "step": 6726 + }, + { + "epoch": 0.29784266628294387, + "grad_norm": 1.3870370388031006, + "learning_rate": 1.770922888228852e-05, + "loss": 0.8159, + "step": 6727 + }, + { + "epoch": 0.2978869419877578, + "grad_norm": 1.5025852918624878, + "learning_rate": 1.7708244448501108e-05, + "loss": 0.8951, + "step": 6728 + }, + { + "epoch": 0.29793121769257164, + "grad_norm": 2.396303415298462, + "learning_rate": 1.770725983060855e-05, + "loss": 0.8264, + "step": 6729 + }, + { + "epoch": 0.2979754933973855, + "grad_norm": 1.2830113172531128, + "learning_rate": 1.7706275028634364e-05, + "loss": 0.8222, + "step": 6730 + }, + { + "epoch": 0.2980197691021994, + "grad_norm": 1.3302887678146362, + "learning_rate": 1.7705290042602082e-05, + "loss": 0.7921, + "step": 6731 + }, + { + "epoch": 0.29806404480701326, + "grad_norm": 1.328735113143921, + "learning_rate": 1.7704304872535215e-05, + "loss": 0.8577, + "step": 6732 + }, + { + "epoch": 0.2981083205118272, + "grad_norm": 1.6555663347244263, + "learning_rate": 1.77033195184573e-05, + "loss": 0.8973, + "step": 6733 + }, + { + "epoch": 0.29815259621664103, + "grad_norm": 1.2568602561950684, + "learning_rate": 1.7702333980391873e-05, + "loss": 0.8417, + "step": 6734 + }, + { + "epoch": 0.2981968719214549, + "grad_norm": 1.3736213445663452, + "learning_rate": 1.7701348258362468e-05, + "loss": 0.9476, + "step": 6735 + }, + { + "epoch": 0.2982411476262688, + "grad_norm": 1.1567707061767578, + "learning_rate": 1.7700362352392632e-05, + "loss": 0.8055, + "step": 6736 + }, + { + "epoch": 0.29828542333108266, + "grad_norm": 1.3465687036514282, + "learning_rate": 1.7699376262505908e-05, + "loss": 0.8258, + "step": 6737 + }, + { + "epoch": 0.2983296990358965, + "grad_norm": 1.502518653869629, + "learning_rate": 1.769838998872585e-05, + "loss": 0.8715, + "step": 6738 + }, + { + "epoch": 0.2983739747407104, + "grad_norm": 1.483480453491211, + "learning_rate": 1.769740353107602e-05, + "loss": 0.9406, + "step": 6739 + }, + { + "epoch": 0.2984182504455243, + "grad_norm": 1.2292813062667847, + "learning_rate": 1.7696416889579974e-05, + "loss": 0.79, + "step": 6740 + }, + { + "epoch": 0.29846252615033814, + "grad_norm": 1.157128930091858, + "learning_rate": 1.7695430064261275e-05, + "loss": 0.7956, + "step": 6741 + }, + { + "epoch": 0.29850680185515205, + "grad_norm": 1.3442991971969604, + "learning_rate": 1.7694443055143494e-05, + "loss": 0.883, + "step": 6742 + }, + { + "epoch": 0.2985510775599659, + "grad_norm": 1.4837833642959595, + "learning_rate": 1.7693455862250212e-05, + "loss": 0.8219, + "step": 6743 + }, + { + "epoch": 0.29859535326477976, + "grad_norm": 2.125265598297119, + "learning_rate": 1.7692468485604994e-05, + "loss": 0.9752, + "step": 6744 + }, + { + "epoch": 0.2986396289695937, + "grad_norm": 1.2831847667694092, + "learning_rate": 1.7691480925231432e-05, + "loss": 0.8596, + "step": 6745 + }, + { + "epoch": 0.29868390467440753, + "grad_norm": 1.2299433946609497, + "learning_rate": 1.769049318115311e-05, + "loss": 0.847, + "step": 6746 + }, + { + "epoch": 0.2987281803792214, + "grad_norm": 1.2970467805862427, + "learning_rate": 1.768950525339362e-05, + "loss": 0.8024, + "step": 6747 + }, + { + "epoch": 0.2987724560840353, + "grad_norm": 1.2888127565383911, + "learning_rate": 1.768851714197656e-05, + "loss": 0.8284, + "step": 6748 + }, + { + "epoch": 0.29881673178884915, + "grad_norm": 1.656705379486084, + "learning_rate": 1.768752884692553e-05, + "loss": 0.8652, + "step": 6749 + }, + { + "epoch": 0.29886100749366307, + "grad_norm": 1.6047589778900146, + "learning_rate": 1.768654036826413e-05, + "loss": 0.9731, + "step": 6750 + }, + { + "epoch": 0.2989052831984769, + "grad_norm": 1.3673373460769653, + "learning_rate": 1.768555170601597e-05, + "loss": 0.8711, + "step": 6751 + }, + { + "epoch": 0.2989495589032908, + "grad_norm": 1.4118971824645996, + "learning_rate": 1.7684562860204668e-05, + "loss": 0.803, + "step": 6752 + }, + { + "epoch": 0.2989938346081047, + "grad_norm": 1.286191701889038, + "learning_rate": 1.7683573830853836e-05, + "loss": 0.8575, + "step": 6753 + }, + { + "epoch": 0.29903811031291855, + "grad_norm": 1.3998035192489624, + "learning_rate": 1.7682584617987105e-05, + "loss": 0.8809, + "step": 6754 + }, + { + "epoch": 0.2990823860177324, + "grad_norm": 1.302734613418579, + "learning_rate": 1.768159522162809e-05, + "loss": 0.7756, + "step": 6755 + }, + { + "epoch": 0.2991266617225463, + "grad_norm": 1.4294548034667969, + "learning_rate": 1.768060564180043e-05, + "loss": 0.8853, + "step": 6756 + }, + { + "epoch": 0.2991709374273602, + "grad_norm": 1.5261937379837036, + "learning_rate": 1.767961587852776e-05, + "loss": 0.8244, + "step": 6757 + }, + { + "epoch": 0.29921521313217403, + "grad_norm": 1.4285119771957397, + "learning_rate": 1.7678625931833714e-05, + "loss": 0.8169, + "step": 6758 + }, + { + "epoch": 0.29925948883698794, + "grad_norm": 1.5364272594451904, + "learning_rate": 1.7677635801741944e-05, + "loss": 0.7789, + "step": 6759 + }, + { + "epoch": 0.2993037645418018, + "grad_norm": 1.5984282493591309, + "learning_rate": 1.7676645488276092e-05, + "loss": 0.8625, + "step": 6760 + }, + { + "epoch": 0.29934804024661565, + "grad_norm": 1.5240147113800049, + "learning_rate": 1.7675654991459812e-05, + "loss": 0.9846, + "step": 6761 + }, + { + "epoch": 0.29939231595142957, + "grad_norm": 1.2120758295059204, + "learning_rate": 1.7674664311316765e-05, + "loss": 0.859, + "step": 6762 + }, + { + "epoch": 0.2994365916562434, + "grad_norm": 1.2719286680221558, + "learning_rate": 1.7673673447870608e-05, + "loss": 0.7977, + "step": 6763 + }, + { + "epoch": 0.2994808673610573, + "grad_norm": 1.3457499742507935, + "learning_rate": 1.767268240114501e-05, + "loss": 0.8697, + "step": 6764 + }, + { + "epoch": 0.2995251430658712, + "grad_norm": 1.2443400621414185, + "learning_rate": 1.7671691171163638e-05, + "loss": 0.7524, + "step": 6765 + }, + { + "epoch": 0.29956941877068505, + "grad_norm": 1.3091254234313965, + "learning_rate": 1.767069975795017e-05, + "loss": 0.8146, + "step": 6766 + }, + { + "epoch": 0.29961369447549896, + "grad_norm": 1.2902132272720337, + "learning_rate": 1.766970816152828e-05, + "loss": 0.8493, + "step": 6767 + }, + { + "epoch": 0.2996579701803128, + "grad_norm": 1.757283091545105, + "learning_rate": 1.7668716381921663e-05, + "loss": 1.0163, + "step": 6768 + }, + { + "epoch": 0.2997022458851267, + "grad_norm": 1.2956970930099487, + "learning_rate": 1.7667724419153996e-05, + "loss": 0.7739, + "step": 6769 + }, + { + "epoch": 0.2997465215899406, + "grad_norm": 1.3747271299362183, + "learning_rate": 1.7666732273248975e-05, + "loss": 0.9011, + "step": 6770 + }, + { + "epoch": 0.29979079729475444, + "grad_norm": 1.4892226457595825, + "learning_rate": 1.7665739944230296e-05, + "loss": 0.8568, + "step": 6771 + }, + { + "epoch": 0.2998350729995683, + "grad_norm": 1.213323712348938, + "learning_rate": 1.766474743212166e-05, + "loss": 0.8668, + "step": 6772 + }, + { + "epoch": 0.2998793487043822, + "grad_norm": 1.2325702905654907, + "learning_rate": 1.766375473694677e-05, + "loss": 0.815, + "step": 6773 + }, + { + "epoch": 0.29992362440919607, + "grad_norm": 1.2704994678497314, + "learning_rate": 1.766276185872934e-05, + "loss": 0.7867, + "step": 6774 + }, + { + "epoch": 0.2999679001140099, + "grad_norm": 1.4098165035247803, + "learning_rate": 1.7661768797493084e-05, + "loss": 0.8908, + "step": 6775 + }, + { + "epoch": 0.30001217581882383, + "grad_norm": 1.4279991388320923, + "learning_rate": 1.7660775553261718e-05, + "loss": 0.8943, + "step": 6776 + }, + { + "epoch": 0.3000564515236377, + "grad_norm": 1.215089201927185, + "learning_rate": 1.7659782126058964e-05, + "loss": 0.8486, + "step": 6777 + }, + { + "epoch": 0.30010072722845155, + "grad_norm": 1.3818869590759277, + "learning_rate": 1.7658788515908552e-05, + "loss": 0.8948, + "step": 6778 + }, + { + "epoch": 0.30014500293326546, + "grad_norm": 1.4177712202072144, + "learning_rate": 1.7657794722834213e-05, + "loss": 0.9393, + "step": 6779 + }, + { + "epoch": 0.3001892786380793, + "grad_norm": 1.477515459060669, + "learning_rate": 1.765680074685968e-05, + "loss": 0.8287, + "step": 6780 + }, + { + "epoch": 0.30023355434289317, + "grad_norm": 1.475205659866333, + "learning_rate": 1.7655806588008694e-05, + "loss": 0.8562, + "step": 6781 + }, + { + "epoch": 0.3002778300477071, + "grad_norm": 1.3250675201416016, + "learning_rate": 1.7654812246305006e-05, + "loss": 0.8557, + "step": 6782 + }, + { + "epoch": 0.30032210575252094, + "grad_norm": 1.3866194486618042, + "learning_rate": 1.765381772177236e-05, + "loss": 0.7891, + "step": 6783 + }, + { + "epoch": 0.3003663814573348, + "grad_norm": 1.4389568567276, + "learning_rate": 1.7652823014434506e-05, + "loss": 0.876, + "step": 6784 + }, + { + "epoch": 0.3004106571621487, + "grad_norm": 1.2945560216903687, + "learning_rate": 1.7651828124315206e-05, + "loss": 0.8864, + "step": 6785 + }, + { + "epoch": 0.30045493286696257, + "grad_norm": 1.4166789054870605, + "learning_rate": 1.7650833051438225e-05, + "loss": 0.8201, + "step": 6786 + }, + { + "epoch": 0.3004992085717765, + "grad_norm": 1.4301280975341797, + "learning_rate": 1.764983779582732e-05, + "loss": 0.899, + "step": 6787 + }, + { + "epoch": 0.30054348427659033, + "grad_norm": 1.285046100616455, + "learning_rate": 1.7648842357506276e-05, + "loss": 0.8422, + "step": 6788 + }, + { + "epoch": 0.3005877599814042, + "grad_norm": 1.4222626686096191, + "learning_rate": 1.7647846736498854e-05, + "loss": 0.8247, + "step": 6789 + }, + { + "epoch": 0.3006320356862181, + "grad_norm": 1.4890373945236206, + "learning_rate": 1.7646850932828842e-05, + "loss": 0.872, + "step": 6790 + }, + { + "epoch": 0.30067631139103196, + "grad_norm": 1.244326114654541, + "learning_rate": 1.7645854946520024e-05, + "loss": 0.7734, + "step": 6791 + }, + { + "epoch": 0.3007205870958458, + "grad_norm": 1.4544017314910889, + "learning_rate": 1.7644858777596184e-05, + "loss": 0.8231, + "step": 6792 + }, + { + "epoch": 0.3007648628006597, + "grad_norm": 1.949092149734497, + "learning_rate": 1.7643862426081118e-05, + "loss": 0.9397, + "step": 6793 + }, + { + "epoch": 0.3008091385054736, + "grad_norm": 1.1722086668014526, + "learning_rate": 1.764286589199862e-05, + "loss": 0.7968, + "step": 6794 + }, + { + "epoch": 0.30085341421028744, + "grad_norm": 1.7244189977645874, + "learning_rate": 1.7641869175372493e-05, + "loss": 0.921, + "step": 6795 + }, + { + "epoch": 0.30089768991510135, + "grad_norm": 1.1217330694198608, + "learning_rate": 1.7640872276226544e-05, + "loss": 0.7182, + "step": 6796 + }, + { + "epoch": 0.3009419656199152, + "grad_norm": 1.2854832410812378, + "learning_rate": 1.763987519458458e-05, + "loss": 0.7634, + "step": 6797 + }, + { + "epoch": 0.30098624132472906, + "grad_norm": 1.489015817642212, + "learning_rate": 1.763887793047042e-05, + "loss": 0.8825, + "step": 6798 + }, + { + "epoch": 0.301030517029543, + "grad_norm": 1.4072896242141724, + "learning_rate": 1.763788048390788e-05, + "loss": 0.836, + "step": 6799 + }, + { + "epoch": 0.30107479273435683, + "grad_norm": 1.5537320375442505, + "learning_rate": 1.7636882854920784e-05, + "loss": 0.8909, + "step": 6800 + }, + { + "epoch": 0.3011190684391707, + "grad_norm": 1.094791293144226, + "learning_rate": 1.7635885043532957e-05, + "loss": 0.7948, + "step": 6801 + }, + { + "epoch": 0.3011633441439846, + "grad_norm": 2.1753907203674316, + "learning_rate": 1.7634887049768238e-05, + "loss": 0.8439, + "step": 6802 + }, + { + "epoch": 0.30120761984879846, + "grad_norm": 1.6879346370697021, + "learning_rate": 1.7633888873650454e-05, + "loss": 0.8536, + "step": 6803 + }, + { + "epoch": 0.30125189555361237, + "grad_norm": 1.6278752088546753, + "learning_rate": 1.763289051520345e-05, + "loss": 0.9572, + "step": 6804 + }, + { + "epoch": 0.3012961712584262, + "grad_norm": 1.258034586906433, + "learning_rate": 1.7631891974451072e-05, + "loss": 0.7992, + "step": 6805 + }, + { + "epoch": 0.3013404469632401, + "grad_norm": 1.7075021266937256, + "learning_rate": 1.763089325141717e-05, + "loss": 0.883, + "step": 6806 + }, + { + "epoch": 0.301384722668054, + "grad_norm": 1.5086475610733032, + "learning_rate": 1.7629894346125596e-05, + "loss": 0.8496, + "step": 6807 + }, + { + "epoch": 0.30142899837286785, + "grad_norm": 1.295518159866333, + "learning_rate": 1.7628895258600205e-05, + "loss": 0.77, + "step": 6808 + }, + { + "epoch": 0.3014732740776817, + "grad_norm": 1.3502098321914673, + "learning_rate": 1.7627895988864864e-05, + "loss": 0.9116, + "step": 6809 + }, + { + "epoch": 0.3015175497824956, + "grad_norm": 1.6763867139816284, + "learning_rate": 1.7626896536943436e-05, + "loss": 0.8925, + "step": 6810 + }, + { + "epoch": 0.3015618254873095, + "grad_norm": 1.8865067958831787, + "learning_rate": 1.7625896902859797e-05, + "loss": 0.8353, + "step": 6811 + }, + { + "epoch": 0.30160610119212333, + "grad_norm": 1.2808740139007568, + "learning_rate": 1.762489708663782e-05, + "loss": 0.8561, + "step": 6812 + }, + { + "epoch": 0.30165037689693724, + "grad_norm": 1.4301825761795044, + "learning_rate": 1.7623897088301387e-05, + "loss": 0.8338, + "step": 6813 + }, + { + "epoch": 0.3016946526017511, + "grad_norm": 1.2397558689117432, + "learning_rate": 1.762289690787437e-05, + "loss": 0.8511, + "step": 6814 + }, + { + "epoch": 0.30173892830656496, + "grad_norm": 1.2345656156539917, + "learning_rate": 1.7621896545380676e-05, + "loss": 0.7269, + "step": 6815 + }, + { + "epoch": 0.30178320401137887, + "grad_norm": 1.478870153427124, + "learning_rate": 1.762089600084419e-05, + "loss": 0.8881, + "step": 6816 + }, + { + "epoch": 0.3018274797161927, + "grad_norm": 1.4921767711639404, + "learning_rate": 1.76198952742888e-05, + "loss": 0.9026, + "step": 6817 + }, + { + "epoch": 0.3018717554210066, + "grad_norm": 1.3603852987289429, + "learning_rate": 1.761889436573842e-05, + "loss": 0.8476, + "step": 6818 + }, + { + "epoch": 0.3019160311258205, + "grad_norm": 1.5872467756271362, + "learning_rate": 1.7617893275216953e-05, + "loss": 0.9356, + "step": 6819 + }, + { + "epoch": 0.30196030683063435, + "grad_norm": 1.5780199766159058, + "learning_rate": 1.7616892002748306e-05, + "loss": 0.7722, + "step": 6820 + }, + { + "epoch": 0.30200458253544826, + "grad_norm": 1.3309472799301147, + "learning_rate": 1.7615890548356395e-05, + "loss": 0.8611, + "step": 6821 + }, + { + "epoch": 0.3020488582402621, + "grad_norm": 1.1894580125808716, + "learning_rate": 1.7614888912065142e-05, + "loss": 0.7855, + "step": 6822 + }, + { + "epoch": 0.302093133945076, + "grad_norm": 1.5306090116500854, + "learning_rate": 1.7613887093898466e-05, + "loss": 0.8955, + "step": 6823 + }, + { + "epoch": 0.3021374096498899, + "grad_norm": 1.7691128253936768, + "learning_rate": 1.7612885093880298e-05, + "loss": 0.8505, + "step": 6824 + }, + { + "epoch": 0.30218168535470374, + "grad_norm": 1.1326508522033691, + "learning_rate": 1.7611882912034565e-05, + "loss": 0.8633, + "step": 6825 + }, + { + "epoch": 0.3022259610595176, + "grad_norm": 1.2731788158416748, + "learning_rate": 1.7610880548385206e-05, + "loss": 0.7949, + "step": 6826 + }, + { + "epoch": 0.3022702367643315, + "grad_norm": 1.5352267026901245, + "learning_rate": 1.7609878002956167e-05, + "loss": 0.974, + "step": 6827 + }, + { + "epoch": 0.30231451246914537, + "grad_norm": 1.2793201208114624, + "learning_rate": 1.7608875275771385e-05, + "loss": 0.8576, + "step": 6828 + }, + { + "epoch": 0.3023587881739592, + "grad_norm": 1.4971888065338135, + "learning_rate": 1.7607872366854812e-05, + "loss": 0.7478, + "step": 6829 + }, + { + "epoch": 0.30240306387877314, + "grad_norm": 1.4539790153503418, + "learning_rate": 1.76068692762304e-05, + "loss": 0.9011, + "step": 6830 + }, + { + "epoch": 0.302447339583587, + "grad_norm": 1.212644338607788, + "learning_rate": 1.7605866003922116e-05, + "loss": 0.8231, + "step": 6831 + }, + { + "epoch": 0.30249161528840085, + "grad_norm": 1.5450841188430786, + "learning_rate": 1.760486254995391e-05, + "loss": 0.9422, + "step": 6832 + }, + { + "epoch": 0.30253589099321476, + "grad_norm": 1.373781442642212, + "learning_rate": 1.7603858914349757e-05, + "loss": 0.8816, + "step": 6833 + }, + { + "epoch": 0.3025801666980286, + "grad_norm": 1.1650892496109009, + "learning_rate": 1.760285509713362e-05, + "loss": 0.7642, + "step": 6834 + }, + { + "epoch": 0.3026244424028425, + "grad_norm": 1.1265777349472046, + "learning_rate": 1.7601851098329484e-05, + "loss": 0.714, + "step": 6835 + }, + { + "epoch": 0.3026687181076564, + "grad_norm": 1.3430957794189453, + "learning_rate": 1.7600846917961323e-05, + "loss": 0.7737, + "step": 6836 + }, + { + "epoch": 0.30271299381247024, + "grad_norm": 1.3011302947998047, + "learning_rate": 1.759984255605312e-05, + "loss": 0.8626, + "step": 6837 + }, + { + "epoch": 0.30275726951728416, + "grad_norm": 1.3072720766067505, + "learning_rate": 1.759883801262887e-05, + "loss": 0.8307, + "step": 6838 + }, + { + "epoch": 0.302801545222098, + "grad_norm": 1.136294960975647, + "learning_rate": 1.759783328771256e-05, + "loss": 0.7421, + "step": 6839 + }, + { + "epoch": 0.30284582092691187, + "grad_norm": 1.214360237121582, + "learning_rate": 1.7596828381328186e-05, + "loss": 0.7963, + "step": 6840 + }, + { + "epoch": 0.3028900966317258, + "grad_norm": 1.310942530632019, + "learning_rate": 1.7595823293499754e-05, + "loss": 0.8272, + "step": 6841 + }, + { + "epoch": 0.30293437233653964, + "grad_norm": 1.2375743389129639, + "learning_rate": 1.7594818024251267e-05, + "loss": 0.7876, + "step": 6842 + }, + { + "epoch": 0.3029786480413535, + "grad_norm": 1.5848801136016846, + "learning_rate": 1.759381257360673e-05, + "loss": 0.8791, + "step": 6843 + }, + { + "epoch": 0.3030229237461674, + "grad_norm": 1.257680892944336, + "learning_rate": 1.759280694159017e-05, + "loss": 0.7879, + "step": 6844 + }, + { + "epoch": 0.30306719945098126, + "grad_norm": 1.3024078607559204, + "learning_rate": 1.7591801128225595e-05, + "loss": 0.7122, + "step": 6845 + }, + { + "epoch": 0.3031114751557951, + "grad_norm": 1.3705633878707886, + "learning_rate": 1.7590795133537035e-05, + "loss": 0.8509, + "step": 6846 + }, + { + "epoch": 0.30315575086060903, + "grad_norm": 1.365535855293274, + "learning_rate": 1.7589788957548513e-05, + "loss": 0.8552, + "step": 6847 + }, + { + "epoch": 0.3032000265654229, + "grad_norm": 1.3725857734680176, + "learning_rate": 1.7588782600284058e-05, + "loss": 0.8079, + "step": 6848 + }, + { + "epoch": 0.30324430227023674, + "grad_norm": 1.60563063621521, + "learning_rate": 1.7587776061767715e-05, + "loss": 0.8702, + "step": 6849 + }, + { + "epoch": 0.30328857797505065, + "grad_norm": 1.4629684686660767, + "learning_rate": 1.7586769342023513e-05, + "loss": 0.8755, + "step": 6850 + }, + { + "epoch": 0.3033328536798645, + "grad_norm": 1.1598492860794067, + "learning_rate": 1.7585762441075504e-05, + "loss": 0.7756, + "step": 6851 + }, + { + "epoch": 0.30337712938467837, + "grad_norm": 1.1606236696243286, + "learning_rate": 1.758475535894774e-05, + "loss": 0.7836, + "step": 6852 + }, + { + "epoch": 0.3034214050894923, + "grad_norm": 1.12003493309021, + "learning_rate": 1.7583748095664267e-05, + "loss": 0.7472, + "step": 6853 + }, + { + "epoch": 0.30346568079430614, + "grad_norm": 1.2304280996322632, + "learning_rate": 1.7582740651249145e-05, + "loss": 0.8135, + "step": 6854 + }, + { + "epoch": 0.30350995649912, + "grad_norm": 1.312349796295166, + "learning_rate": 1.758173302572644e-05, + "loss": 0.8497, + "step": 6855 + }, + { + "epoch": 0.3035542322039339, + "grad_norm": 1.1989688873291016, + "learning_rate": 1.7580725219120208e-05, + "loss": 0.8447, + "step": 6856 + }, + { + "epoch": 0.30359850790874776, + "grad_norm": 1.546468734741211, + "learning_rate": 1.757971723145453e-05, + "loss": 0.9179, + "step": 6857 + }, + { + "epoch": 0.3036427836135617, + "grad_norm": 1.6295902729034424, + "learning_rate": 1.757870906275348e-05, + "loss": 0.8733, + "step": 6858 + }, + { + "epoch": 0.30368705931837553, + "grad_norm": 1.7540931701660156, + "learning_rate": 1.7577700713041133e-05, + "loss": 0.9896, + "step": 6859 + }, + { + "epoch": 0.3037313350231894, + "grad_norm": 1.2540178298950195, + "learning_rate": 1.7576692182341574e-05, + "loss": 0.8535, + "step": 6860 + }, + { + "epoch": 0.3037756107280033, + "grad_norm": 3.4869155883789062, + "learning_rate": 1.7575683470678894e-05, + "loss": 0.9241, + "step": 6861 + }, + { + "epoch": 0.30381988643281715, + "grad_norm": 1.2964133024215698, + "learning_rate": 1.7574674578077176e-05, + "loss": 0.8497, + "step": 6862 + }, + { + "epoch": 0.303864162137631, + "grad_norm": 1.4358011484146118, + "learning_rate": 1.757366550456053e-05, + "loss": 0.8762, + "step": 6863 + }, + { + "epoch": 0.3039084378424449, + "grad_norm": 1.348090410232544, + "learning_rate": 1.757265625015305e-05, + "loss": 0.8838, + "step": 6864 + }, + { + "epoch": 0.3039527135472588, + "grad_norm": 1.4085652828216553, + "learning_rate": 1.757164681487884e-05, + "loss": 0.9416, + "step": 6865 + }, + { + "epoch": 0.30399698925207264, + "grad_norm": 1.3486182689666748, + "learning_rate": 1.757063719876201e-05, + "loss": 0.7776, + "step": 6866 + }, + { + "epoch": 0.30404126495688655, + "grad_norm": 1.4918869733810425, + "learning_rate": 1.7569627401826673e-05, + "loss": 0.8672, + "step": 6867 + }, + { + "epoch": 0.3040855406617004, + "grad_norm": 1.7263541221618652, + "learning_rate": 1.7568617424096955e-05, + "loss": 0.9169, + "step": 6868 + }, + { + "epoch": 0.30412981636651426, + "grad_norm": 1.3925985097885132, + "learning_rate": 1.7567607265596966e-05, + "loss": 0.9134, + "step": 6869 + }, + { + "epoch": 0.3041740920713282, + "grad_norm": 1.419423222541809, + "learning_rate": 1.756659692635084e-05, + "loss": 0.8507, + "step": 6870 + }, + { + "epoch": 0.30421836777614203, + "grad_norm": 1.2268187999725342, + "learning_rate": 1.756558640638271e-05, + "loss": 0.7769, + "step": 6871 + }, + { + "epoch": 0.3042626434809559, + "grad_norm": 1.4246900081634521, + "learning_rate": 1.7564575705716707e-05, + "loss": 0.8516, + "step": 6872 + }, + { + "epoch": 0.3043069191857698, + "grad_norm": 1.4656916856765747, + "learning_rate": 1.7563564824376975e-05, + "loss": 0.8192, + "step": 6873 + }, + { + "epoch": 0.30435119489058365, + "grad_norm": 1.3859388828277588, + "learning_rate": 1.7562553762387656e-05, + "loss": 0.8576, + "step": 6874 + }, + { + "epoch": 0.30439547059539757, + "grad_norm": 1.1565364599227905, + "learning_rate": 1.7561542519772894e-05, + "loss": 0.7411, + "step": 6875 + }, + { + "epoch": 0.3044397463002114, + "grad_norm": 1.599513292312622, + "learning_rate": 1.7560531096556847e-05, + "loss": 0.8516, + "step": 6876 + }, + { + "epoch": 0.3044840220050253, + "grad_norm": 1.2602206468582153, + "learning_rate": 1.7559519492763676e-05, + "loss": 0.8474, + "step": 6877 + }, + { + "epoch": 0.3045282977098392, + "grad_norm": 1.2166714668273926, + "learning_rate": 1.755850770841753e-05, + "loss": 0.7332, + "step": 6878 + }, + { + "epoch": 0.30457257341465305, + "grad_norm": 1.2365343570709229, + "learning_rate": 1.7557495743542586e-05, + "loss": 0.8209, + "step": 6879 + }, + { + "epoch": 0.3046168491194669, + "grad_norm": 1.4406846761703491, + "learning_rate": 1.755648359816301e-05, + "loss": 0.7985, + "step": 6880 + }, + { + "epoch": 0.3046611248242808, + "grad_norm": 1.2823805809020996, + "learning_rate": 1.755547127230297e-05, + "loss": 0.9044, + "step": 6881 + }, + { + "epoch": 0.30470540052909467, + "grad_norm": 1.512709140777588, + "learning_rate": 1.7554458765986657e-05, + "loss": 0.8733, + "step": 6882 + }, + { + "epoch": 0.30474967623390853, + "grad_norm": 1.3649194240570068, + "learning_rate": 1.7553446079238244e-05, + "loss": 0.8299, + "step": 6883 + }, + { + "epoch": 0.30479395193872244, + "grad_norm": 1.4411065578460693, + "learning_rate": 1.7552433212081924e-05, + "loss": 0.936, + "step": 6884 + }, + { + "epoch": 0.3048382276435363, + "grad_norm": 1.3150634765625, + "learning_rate": 1.7551420164541883e-05, + "loss": 0.7956, + "step": 6885 + }, + { + "epoch": 0.30488250334835015, + "grad_norm": 1.6178420782089233, + "learning_rate": 1.755040693664232e-05, + "loss": 0.8703, + "step": 6886 + }, + { + "epoch": 0.30492677905316407, + "grad_norm": 1.3311686515808105, + "learning_rate": 1.7549393528407443e-05, + "loss": 0.8203, + "step": 6887 + }, + { + "epoch": 0.3049710547579779, + "grad_norm": 1.2891817092895508, + "learning_rate": 1.7548379939861438e-05, + "loss": 0.7839, + "step": 6888 + }, + { + "epoch": 0.3050153304627918, + "grad_norm": 1.2340887784957886, + "learning_rate": 1.754736617102853e-05, + "loss": 0.8506, + "step": 6889 + }, + { + "epoch": 0.3050596061676057, + "grad_norm": 1.2473686933517456, + "learning_rate": 1.7546352221932926e-05, + "loss": 0.8397, + "step": 6890 + }, + { + "epoch": 0.30510388187241955, + "grad_norm": 1.3002341985702515, + "learning_rate": 1.754533809259884e-05, + "loss": 0.8375, + "step": 6891 + }, + { + "epoch": 0.30514815757723346, + "grad_norm": 1.8705066442489624, + "learning_rate": 1.75443237830505e-05, + "loss": 0.9793, + "step": 6892 + }, + { + "epoch": 0.3051924332820473, + "grad_norm": 1.279632806777954, + "learning_rate": 1.754330929331213e-05, + "loss": 0.8253, + "step": 6893 + }, + { + "epoch": 0.30523670898686117, + "grad_norm": 1.3382844924926758, + "learning_rate": 1.7542294623407962e-05, + "loss": 0.855, + "step": 6894 + }, + { + "epoch": 0.3052809846916751, + "grad_norm": 1.3298585414886475, + "learning_rate": 1.7541279773362223e-05, + "loss": 0.8453, + "step": 6895 + }, + { + "epoch": 0.30532526039648894, + "grad_norm": 1.4216043949127197, + "learning_rate": 1.754026474319916e-05, + "loss": 0.8104, + "step": 6896 + }, + { + "epoch": 0.3053695361013028, + "grad_norm": 1.242172122001648, + "learning_rate": 1.753924953294301e-05, + "loss": 0.8405, + "step": 6897 + }, + { + "epoch": 0.3054138118061167, + "grad_norm": 1.4429987668991089, + "learning_rate": 1.7538234142618028e-05, + "loss": 0.8663, + "step": 6898 + }, + { + "epoch": 0.30545808751093056, + "grad_norm": 1.5413358211517334, + "learning_rate": 1.7537218572248457e-05, + "loss": 0.9349, + "step": 6899 + }, + { + "epoch": 0.3055023632157444, + "grad_norm": 1.2892720699310303, + "learning_rate": 1.7536202821858558e-05, + "loss": 0.7922, + "step": 6900 + }, + { + "epoch": 0.30554663892055833, + "grad_norm": 1.177533507347107, + "learning_rate": 1.7535186891472595e-05, + "loss": 0.7559, + "step": 6901 + }, + { + "epoch": 0.3055909146253722, + "grad_norm": 1.7709358930587769, + "learning_rate": 1.7534170781114824e-05, + "loss": 0.9563, + "step": 6902 + }, + { + "epoch": 0.30563519033018605, + "grad_norm": 1.1954538822174072, + "learning_rate": 1.7533154490809522e-05, + "loss": 0.8224, + "step": 6903 + }, + { + "epoch": 0.30567946603499996, + "grad_norm": 1.4175950288772583, + "learning_rate": 1.7532138020580954e-05, + "loss": 0.9024, + "step": 6904 + }, + { + "epoch": 0.3057237417398138, + "grad_norm": 1.2611678838729858, + "learning_rate": 1.7531121370453407e-05, + "loss": 0.9053, + "step": 6905 + }, + { + "epoch": 0.30576801744462767, + "grad_norm": 1.3001062870025635, + "learning_rate": 1.7530104540451155e-05, + "loss": 0.8436, + "step": 6906 + }, + { + "epoch": 0.3058122931494416, + "grad_norm": 1.1551189422607422, + "learning_rate": 1.752908753059849e-05, + "loss": 0.8049, + "step": 6907 + }, + { + "epoch": 0.30585656885425544, + "grad_norm": 1.403051733970642, + "learning_rate": 1.75280703409197e-05, + "loss": 0.8248, + "step": 6908 + }, + { + "epoch": 0.30590084455906935, + "grad_norm": 1.1354787349700928, + "learning_rate": 1.7527052971439077e-05, + "loss": 0.7579, + "step": 6909 + }, + { + "epoch": 0.3059451202638832, + "grad_norm": 1.41193425655365, + "learning_rate": 1.7526035422180923e-05, + "loss": 0.8739, + "step": 6910 + }, + { + "epoch": 0.30598939596869706, + "grad_norm": 1.7443479299545288, + "learning_rate": 1.752501769316954e-05, + "loss": 0.9176, + "step": 6911 + }, + { + "epoch": 0.306033671673511, + "grad_norm": 1.6533403396606445, + "learning_rate": 1.7523999784429238e-05, + "loss": 0.8395, + "step": 6912 + }, + { + "epoch": 0.30607794737832483, + "grad_norm": 1.397294044494629, + "learning_rate": 1.7522981695984326e-05, + "loss": 0.954, + "step": 6913 + }, + { + "epoch": 0.3061222230831387, + "grad_norm": 1.6384214162826538, + "learning_rate": 1.7521963427859125e-05, + "loss": 0.9193, + "step": 6914 + }, + { + "epoch": 0.3061664987879526, + "grad_norm": 1.2233448028564453, + "learning_rate": 1.7520944980077948e-05, + "loss": 0.8289, + "step": 6915 + }, + { + "epoch": 0.30621077449276646, + "grad_norm": 1.3585491180419922, + "learning_rate": 1.7519926352665125e-05, + "loss": 0.8096, + "step": 6916 + }, + { + "epoch": 0.3062550501975803, + "grad_norm": 1.4846858978271484, + "learning_rate": 1.7518907545644986e-05, + "loss": 0.7697, + "step": 6917 + }, + { + "epoch": 0.3062993259023942, + "grad_norm": 1.4166384935379028, + "learning_rate": 1.751788855904186e-05, + "loss": 0.8455, + "step": 6918 + }, + { + "epoch": 0.3063436016072081, + "grad_norm": 2.047729969024658, + "learning_rate": 1.7516869392880088e-05, + "loss": 0.8347, + "step": 6919 + }, + { + "epoch": 0.30638787731202194, + "grad_norm": 1.2571741342544556, + "learning_rate": 1.7515850047184014e-05, + "loss": 0.8143, + "step": 6920 + }, + { + "epoch": 0.30643215301683585, + "grad_norm": 1.3111402988433838, + "learning_rate": 1.7514830521977978e-05, + "loss": 0.8483, + "step": 6921 + }, + { + "epoch": 0.3064764287216497, + "grad_norm": 1.462913990020752, + "learning_rate": 1.751381081728633e-05, + "loss": 0.8684, + "step": 6922 + }, + { + "epoch": 0.30652070442646356, + "grad_norm": 1.209158182144165, + "learning_rate": 1.7512790933133435e-05, + "loss": 0.7856, + "step": 6923 + }, + { + "epoch": 0.3065649801312775, + "grad_norm": 1.2694560289382935, + "learning_rate": 1.7511770869543646e-05, + "loss": 0.7797, + "step": 6924 + }, + { + "epoch": 0.30660925583609133, + "grad_norm": 1.3237348794937134, + "learning_rate": 1.7510750626541325e-05, + "loss": 0.7845, + "step": 6925 + }, + { + "epoch": 0.30665353154090524, + "grad_norm": 1.422525405883789, + "learning_rate": 1.7509730204150836e-05, + "loss": 0.8313, + "step": 6926 + }, + { + "epoch": 0.3066978072457191, + "grad_norm": 1.2345964908599854, + "learning_rate": 1.750870960239656e-05, + "loss": 0.7765, + "step": 6927 + }, + { + "epoch": 0.30674208295053296, + "grad_norm": 1.3431333303451538, + "learning_rate": 1.7507688821302867e-05, + "loss": 0.8416, + "step": 6928 + }, + { + "epoch": 0.30678635865534687, + "grad_norm": 1.2039741277694702, + "learning_rate": 1.7506667860894143e-05, + "loss": 0.8107, + "step": 6929 + }, + { + "epoch": 0.3068306343601607, + "grad_norm": 1.499225378036499, + "learning_rate": 1.7505646721194764e-05, + "loss": 0.842, + "step": 6930 + }, + { + "epoch": 0.3068749100649746, + "grad_norm": 1.4079499244689941, + "learning_rate": 1.750462540222913e-05, + "loss": 0.8812, + "step": 6931 + }, + { + "epoch": 0.3069191857697885, + "grad_norm": 1.3805956840515137, + "learning_rate": 1.7503603904021623e-05, + "loss": 0.8725, + "step": 6932 + }, + { + "epoch": 0.30696346147460235, + "grad_norm": 1.391442060470581, + "learning_rate": 1.7502582226596653e-05, + "loss": 0.8474, + "step": 6933 + }, + { + "epoch": 0.3070077371794162, + "grad_norm": 1.3289158344268799, + "learning_rate": 1.750156036997861e-05, + "loss": 0.8561, + "step": 6934 + }, + { + "epoch": 0.3070520128842301, + "grad_norm": 1.5858221054077148, + "learning_rate": 1.7500538334191906e-05, + "loss": 0.9049, + "step": 6935 + }, + { + "epoch": 0.307096288589044, + "grad_norm": 1.313481330871582, + "learning_rate": 1.7499516119260956e-05, + "loss": 0.7504, + "step": 6936 + }, + { + "epoch": 0.30714056429385783, + "grad_norm": 1.2912646532058716, + "learning_rate": 1.7498493725210164e-05, + "loss": 0.847, + "step": 6937 + }, + { + "epoch": 0.30718483999867174, + "grad_norm": 1.7881507873535156, + "learning_rate": 1.749747115206396e-05, + "loss": 0.9855, + "step": 6938 + }, + { + "epoch": 0.3072291157034856, + "grad_norm": 1.5216654539108276, + "learning_rate": 1.7496448399846757e-05, + "loss": 0.8502, + "step": 6939 + }, + { + "epoch": 0.30727339140829946, + "grad_norm": 1.6347788572311401, + "learning_rate": 1.749542546858299e-05, + "loss": 0.9075, + "step": 6940 + }, + { + "epoch": 0.30731766711311337, + "grad_norm": 1.4688782691955566, + "learning_rate": 1.749440235829709e-05, + "loss": 0.8114, + "step": 6941 + }, + { + "epoch": 0.3073619428179272, + "grad_norm": 1.232789397239685, + "learning_rate": 1.749337906901349e-05, + "loss": 0.818, + "step": 6942 + }, + { + "epoch": 0.3074062185227411, + "grad_norm": 1.4300719499588013, + "learning_rate": 1.7492355600756633e-05, + "loss": 0.8502, + "step": 6943 + }, + { + "epoch": 0.307450494227555, + "grad_norm": 1.3771501779556274, + "learning_rate": 1.7491331953550963e-05, + "loss": 0.8271, + "step": 6944 + }, + { + "epoch": 0.30749476993236885, + "grad_norm": 1.7458148002624512, + "learning_rate": 1.7490308127420928e-05, + "loss": 0.9244, + "step": 6945 + }, + { + "epoch": 0.30753904563718276, + "grad_norm": 1.4490405321121216, + "learning_rate": 1.7489284122390984e-05, + "loss": 0.8649, + "step": 6946 + }, + { + "epoch": 0.3075833213419966, + "grad_norm": 1.322090983390808, + "learning_rate": 1.7488259938485585e-05, + "loss": 0.7706, + "step": 6947 + }, + { + "epoch": 0.3076275970468105, + "grad_norm": 1.1547839641571045, + "learning_rate": 1.7487235575729195e-05, + "loss": 0.7616, + "step": 6948 + }, + { + "epoch": 0.3076718727516244, + "grad_norm": 1.5388787984848022, + "learning_rate": 1.7486211034146282e-05, + "loss": 0.9266, + "step": 6949 + }, + { + "epoch": 0.30771614845643824, + "grad_norm": 1.7047598361968994, + "learning_rate": 1.748518631376131e-05, + "loss": 0.9083, + "step": 6950 + }, + { + "epoch": 0.3077604241612521, + "grad_norm": 1.4142884016036987, + "learning_rate": 1.7484161414598763e-05, + "loss": 0.9093, + "step": 6951 + }, + { + "epoch": 0.307804699866066, + "grad_norm": 1.2342602014541626, + "learning_rate": 1.748313633668311e-05, + "loss": 0.833, + "step": 6952 + }, + { + "epoch": 0.30784897557087987, + "grad_norm": 1.178303599357605, + "learning_rate": 1.7482111080038837e-05, + "loss": 0.7927, + "step": 6953 + }, + { + "epoch": 0.3078932512756937, + "grad_norm": 1.5011367797851562, + "learning_rate": 1.7481085644690437e-05, + "loss": 0.8677, + "step": 6954 + }, + { + "epoch": 0.30793752698050764, + "grad_norm": 1.3305352926254272, + "learning_rate": 1.7480060030662396e-05, + "loss": 0.875, + "step": 6955 + }, + { + "epoch": 0.3079818026853215, + "grad_norm": 1.3654584884643555, + "learning_rate": 1.7479034237979212e-05, + "loss": 0.914, + "step": 6956 + }, + { + "epoch": 0.30802607839013535, + "grad_norm": 1.1769157648086548, + "learning_rate": 1.7478008266665382e-05, + "loss": 0.8267, + "step": 6957 + }, + { + "epoch": 0.30807035409494926, + "grad_norm": 1.3214162588119507, + "learning_rate": 1.747698211674542e-05, + "loss": 0.792, + "step": 6958 + }, + { + "epoch": 0.3081146297997631, + "grad_norm": 1.3047165870666504, + "learning_rate": 1.7475955788243823e-05, + "loss": 0.8536, + "step": 6959 + }, + { + "epoch": 0.308158905504577, + "grad_norm": 1.2780067920684814, + "learning_rate": 1.747492928118511e-05, + "loss": 0.7872, + "step": 6960 + }, + { + "epoch": 0.3082031812093909, + "grad_norm": 1.2011650800704956, + "learning_rate": 1.7473902595593796e-05, + "loss": 0.7998, + "step": 6961 + }, + { + "epoch": 0.30824745691420474, + "grad_norm": 1.4652032852172852, + "learning_rate": 1.747287573149441e-05, + "loss": 0.815, + "step": 6962 + }, + { + "epoch": 0.30829173261901865, + "grad_norm": 1.174364447593689, + "learning_rate": 1.7471848688911465e-05, + "loss": 0.8037, + "step": 6963 + }, + { + "epoch": 0.3083360083238325, + "grad_norm": 1.2562544345855713, + "learning_rate": 1.74708214678695e-05, + "loss": 0.8605, + "step": 6964 + }, + { + "epoch": 0.30838028402864637, + "grad_norm": 1.2268035411834717, + "learning_rate": 1.746979406839305e-05, + "loss": 0.7994, + "step": 6965 + }, + { + "epoch": 0.3084245597334603, + "grad_norm": 1.4530082941055298, + "learning_rate": 1.7468766490506647e-05, + "loss": 0.8983, + "step": 6966 + }, + { + "epoch": 0.30846883543827414, + "grad_norm": 1.3749922513961792, + "learning_rate": 1.746773873423484e-05, + "loss": 0.7672, + "step": 6967 + }, + { + "epoch": 0.308513111143088, + "grad_norm": 1.2404863834381104, + "learning_rate": 1.7466710799602173e-05, + "loss": 0.7964, + "step": 6968 + }, + { + "epoch": 0.3085573868479019, + "grad_norm": 1.5043383836746216, + "learning_rate": 1.7465682686633197e-05, + "loss": 0.8308, + "step": 6969 + }, + { + "epoch": 0.30860166255271576, + "grad_norm": 1.1258307695388794, + "learning_rate": 1.7464654395352472e-05, + "loss": 0.7723, + "step": 6970 + }, + { + "epoch": 0.3086459382575296, + "grad_norm": 1.324117660522461, + "learning_rate": 1.7463625925784553e-05, + "loss": 0.803, + "step": 6971 + }, + { + "epoch": 0.30869021396234353, + "grad_norm": 1.6276990175247192, + "learning_rate": 1.7462597277954004e-05, + "loss": 0.8213, + "step": 6972 + }, + { + "epoch": 0.3087344896671574, + "grad_norm": 1.1581532955169678, + "learning_rate": 1.7461568451885397e-05, + "loss": 0.8201, + "step": 6973 + }, + { + "epoch": 0.30877876537197124, + "grad_norm": 1.2879937887191772, + "learning_rate": 1.7460539447603302e-05, + "loss": 0.8362, + "step": 6974 + }, + { + "epoch": 0.30882304107678515, + "grad_norm": 1.3363020420074463, + "learning_rate": 1.74595102651323e-05, + "loss": 0.8181, + "step": 6975 + }, + { + "epoch": 0.308867316781599, + "grad_norm": 1.4648569822311401, + "learning_rate": 1.7458480904496964e-05, + "loss": 0.9116, + "step": 6976 + }, + { + "epoch": 0.30891159248641287, + "grad_norm": 1.4985312223434448, + "learning_rate": 1.7457451365721887e-05, + "loss": 0.8178, + "step": 6977 + }, + { + "epoch": 0.3089558681912268, + "grad_norm": 1.5270849466323853, + "learning_rate": 1.7456421648831658e-05, + "loss": 0.8018, + "step": 6978 + }, + { + "epoch": 0.30900014389604064, + "grad_norm": 1.2440768480300903, + "learning_rate": 1.7455391753850867e-05, + "loss": 0.8033, + "step": 6979 + }, + { + "epoch": 0.30904441960085455, + "grad_norm": 1.5847251415252686, + "learning_rate": 1.7454361680804117e-05, + "loss": 0.8823, + "step": 6980 + }, + { + "epoch": 0.3090886953056684, + "grad_norm": 1.5415537357330322, + "learning_rate": 1.7453331429716005e-05, + "loss": 0.8334, + "step": 6981 + }, + { + "epoch": 0.30913297101048226, + "grad_norm": 1.1903917789459229, + "learning_rate": 1.745230100061114e-05, + "loss": 0.7868, + "step": 6982 + }, + { + "epoch": 0.3091772467152962, + "grad_norm": 1.2610552310943604, + "learning_rate": 1.7451270393514134e-05, + "loss": 0.8384, + "step": 6983 + }, + { + "epoch": 0.30922152242011003, + "grad_norm": 1.335951328277588, + "learning_rate": 1.7450239608449605e-05, + "loss": 0.8589, + "step": 6984 + }, + { + "epoch": 0.3092657981249239, + "grad_norm": 1.4276992082595825, + "learning_rate": 1.7449208645442166e-05, + "loss": 0.8278, + "step": 6985 + }, + { + "epoch": 0.3093100738297378, + "grad_norm": 1.4713876247406006, + "learning_rate": 1.7448177504516443e-05, + "loss": 0.8903, + "step": 6986 + }, + { + "epoch": 0.30935434953455165, + "grad_norm": 1.2186692953109741, + "learning_rate": 1.744714618569707e-05, + "loss": 0.7333, + "step": 6987 + }, + { + "epoch": 0.3093986252393655, + "grad_norm": 1.570881962776184, + "learning_rate": 1.744611468900867e-05, + "loss": 0.8661, + "step": 6988 + }, + { + "epoch": 0.3094429009441794, + "grad_norm": 1.3780124187469482, + "learning_rate": 1.744508301447588e-05, + "loss": 0.7752, + "step": 6989 + }, + { + "epoch": 0.3094871766489933, + "grad_norm": 1.3864538669586182, + "learning_rate": 1.744405116212335e-05, + "loss": 0.8763, + "step": 6990 + }, + { + "epoch": 0.30953145235380713, + "grad_norm": 1.6349421739578247, + "learning_rate": 1.7443019131975716e-05, + "loss": 1.0189, + "step": 6991 + }, + { + "epoch": 0.30957572805862105, + "grad_norm": 1.2193180322647095, + "learning_rate": 1.7441986924057636e-05, + "loss": 0.7823, + "step": 6992 + }, + { + "epoch": 0.3096200037634349, + "grad_norm": 1.396511435508728, + "learning_rate": 1.744095453839375e-05, + "loss": 0.9044, + "step": 6993 + }, + { + "epoch": 0.30966427946824876, + "grad_norm": 1.3510899543762207, + "learning_rate": 1.743992197500873e-05, + "loss": 0.7776, + "step": 6994 + }, + { + "epoch": 0.30970855517306267, + "grad_norm": 1.3325271606445312, + "learning_rate": 1.7438889233927227e-05, + "loss": 0.8068, + "step": 6995 + }, + { + "epoch": 0.30975283087787653, + "grad_norm": 1.418204426765442, + "learning_rate": 1.7437856315173913e-05, + "loss": 0.8767, + "step": 6996 + }, + { + "epoch": 0.30979710658269044, + "grad_norm": 1.9577536582946777, + "learning_rate": 1.7436823218773457e-05, + "loss": 0.868, + "step": 6997 + }, + { + "epoch": 0.3098413822875043, + "grad_norm": 1.4470056295394897, + "learning_rate": 1.7435789944750534e-05, + "loss": 0.9077, + "step": 6998 + }, + { + "epoch": 0.30988565799231815, + "grad_norm": 1.2943084239959717, + "learning_rate": 1.7434756493129823e-05, + "loss": 0.8996, + "step": 6999 + }, + { + "epoch": 0.30992993369713207, + "grad_norm": 1.404892921447754, + "learning_rate": 1.7433722863936004e-05, + "loss": 0.9155, + "step": 7000 + }, + { + "epoch": 0.3099742094019459, + "grad_norm": 1.2950677871704102, + "learning_rate": 1.743268905719377e-05, + "loss": 0.8384, + "step": 7001 + }, + { + "epoch": 0.3100184851067598, + "grad_norm": 1.6337449550628662, + "learning_rate": 1.743165507292781e-05, + "loss": 0.8794, + "step": 7002 + }, + { + "epoch": 0.3100627608115737, + "grad_norm": 1.343301773071289, + "learning_rate": 1.7430620911162818e-05, + "loss": 0.8909, + "step": 7003 + }, + { + "epoch": 0.31010703651638755, + "grad_norm": 1.2838108539581299, + "learning_rate": 1.74295865719235e-05, + "loss": 0.7949, + "step": 7004 + }, + { + "epoch": 0.3101513122212014, + "grad_norm": 1.2539423704147339, + "learning_rate": 1.7428552055234547e-05, + "loss": 0.7903, + "step": 7005 + }, + { + "epoch": 0.3101955879260153, + "grad_norm": 1.3117761611938477, + "learning_rate": 1.7427517361120685e-05, + "loss": 0.8622, + "step": 7006 + }, + { + "epoch": 0.31023986363082917, + "grad_norm": 1.1182966232299805, + "learning_rate": 1.7426482489606616e-05, + "loss": 0.7464, + "step": 7007 + }, + { + "epoch": 0.310284139335643, + "grad_norm": 1.4100964069366455, + "learning_rate": 1.742544744071706e-05, + "loss": 0.8523, + "step": 7008 + }, + { + "epoch": 0.31032841504045694, + "grad_norm": 1.2728583812713623, + "learning_rate": 1.7424412214476736e-05, + "loss": 0.7863, + "step": 7009 + }, + { + "epoch": 0.3103726907452708, + "grad_norm": 1.2920482158660889, + "learning_rate": 1.7423376810910374e-05, + "loss": 0.8197, + "step": 7010 + }, + { + "epoch": 0.31041696645008465, + "grad_norm": 1.5344823598861694, + "learning_rate": 1.74223412300427e-05, + "loss": 0.8793, + "step": 7011 + }, + { + "epoch": 0.31046124215489856, + "grad_norm": 1.390397310256958, + "learning_rate": 1.7421305471898453e-05, + "loss": 0.8495, + "step": 7012 + }, + { + "epoch": 0.3105055178597124, + "grad_norm": 1.6375701427459717, + "learning_rate": 1.742026953650236e-05, + "loss": 0.8648, + "step": 7013 + }, + { + "epoch": 0.31054979356452633, + "grad_norm": 1.3266037702560425, + "learning_rate": 1.7419233423879178e-05, + "loss": 0.7856, + "step": 7014 + }, + { + "epoch": 0.3105940692693402, + "grad_norm": 1.8611969947814941, + "learning_rate": 1.741819713405364e-05, + "loss": 0.8953, + "step": 7015 + }, + { + "epoch": 0.31063834497415405, + "grad_norm": 1.198357105255127, + "learning_rate": 1.7417160667050508e-05, + "loss": 0.7626, + "step": 7016 + }, + { + "epoch": 0.31068262067896796, + "grad_norm": 1.688772201538086, + "learning_rate": 1.741612402289453e-05, + "loss": 0.8437, + "step": 7017 + }, + { + "epoch": 0.3107268963837818, + "grad_norm": 1.5196536779403687, + "learning_rate": 1.741508720161047e-05, + "loss": 0.7919, + "step": 7018 + }, + { + "epoch": 0.31077117208859567, + "grad_norm": 1.3305319547653198, + "learning_rate": 1.7414050203223092e-05, + "loss": 0.7307, + "step": 7019 + }, + { + "epoch": 0.3108154477934096, + "grad_norm": 1.183409333229065, + "learning_rate": 1.7413013027757158e-05, + "loss": 0.7417, + "step": 7020 + }, + { + "epoch": 0.31085972349822344, + "grad_norm": 1.2990010976791382, + "learning_rate": 1.7411975675237447e-05, + "loss": 0.8033, + "step": 7021 + }, + { + "epoch": 0.3109039992030373, + "grad_norm": 2.123326301574707, + "learning_rate": 1.741093814568873e-05, + "loss": 0.9267, + "step": 7022 + }, + { + "epoch": 0.3109482749078512, + "grad_norm": 1.3673449754714966, + "learning_rate": 1.7409900439135786e-05, + "loss": 0.8279, + "step": 7023 + }, + { + "epoch": 0.31099255061266506, + "grad_norm": 1.1979877948760986, + "learning_rate": 1.7408862555603406e-05, + "loss": 0.7817, + "step": 7024 + }, + { + "epoch": 0.3110368263174789, + "grad_norm": 1.2723642587661743, + "learning_rate": 1.740782449511638e-05, + "loss": 0.8156, + "step": 7025 + }, + { + "epoch": 0.31108110202229283, + "grad_norm": 1.5817279815673828, + "learning_rate": 1.7406786257699494e-05, + "loss": 0.7963, + "step": 7026 + }, + { + "epoch": 0.3111253777271067, + "grad_norm": 1.3846439123153687, + "learning_rate": 1.7405747843377553e-05, + "loss": 0.7954, + "step": 7027 + }, + { + "epoch": 0.31116965343192055, + "grad_norm": 1.5148597955703735, + "learning_rate": 1.740470925217535e-05, + "loss": 0.8079, + "step": 7028 + }, + { + "epoch": 0.31121392913673446, + "grad_norm": 1.3025214672088623, + "learning_rate": 1.74036704841177e-05, + "loss": 0.8213, + "step": 7029 + }, + { + "epoch": 0.3112582048415483, + "grad_norm": 1.4195072650909424, + "learning_rate": 1.7402631539229404e-05, + "loss": 0.853, + "step": 7030 + }, + { + "epoch": 0.31130248054636217, + "grad_norm": 1.5654531717300415, + "learning_rate": 1.7401592417535284e-05, + "loss": 0.7955, + "step": 7031 + }, + { + "epoch": 0.3113467562511761, + "grad_norm": 1.332371473312378, + "learning_rate": 1.7400553119060156e-05, + "loss": 0.8221, + "step": 7032 + }, + { + "epoch": 0.31139103195598994, + "grad_norm": 1.1916953325271606, + "learning_rate": 1.739951364382884e-05, + "loss": 0.724, + "step": 7033 + }, + { + "epoch": 0.31143530766080385, + "grad_norm": 1.5196905136108398, + "learning_rate": 1.7398473991866168e-05, + "loss": 0.9286, + "step": 7034 + }, + { + "epoch": 0.3114795833656177, + "grad_norm": 1.4659228324890137, + "learning_rate": 1.7397434163196967e-05, + "loss": 0.8939, + "step": 7035 + }, + { + "epoch": 0.31152385907043156, + "grad_norm": 1.4941614866256714, + "learning_rate": 1.7396394157846076e-05, + "loss": 0.8972, + "step": 7036 + }, + { + "epoch": 0.3115681347752455, + "grad_norm": 1.4496551752090454, + "learning_rate": 1.739535397583833e-05, + "loss": 0.8176, + "step": 7037 + }, + { + "epoch": 0.31161241048005933, + "grad_norm": 1.2217143774032593, + "learning_rate": 1.739431361719858e-05, + "loss": 0.8169, + "step": 7038 + }, + { + "epoch": 0.3116566861848732, + "grad_norm": 1.457975149154663, + "learning_rate": 1.739327308195167e-05, + "loss": 0.8416, + "step": 7039 + }, + { + "epoch": 0.3117009618896871, + "grad_norm": 1.3944175243377686, + "learning_rate": 1.7392232370122448e-05, + "loss": 0.8657, + "step": 7040 + }, + { + "epoch": 0.31174523759450096, + "grad_norm": 2.408571481704712, + "learning_rate": 1.7391191481735775e-05, + "loss": 0.912, + "step": 7041 + }, + { + "epoch": 0.3117895132993148, + "grad_norm": 1.3254762887954712, + "learning_rate": 1.7390150416816512e-05, + "loss": 0.8256, + "step": 7042 + }, + { + "epoch": 0.3118337890041287, + "grad_norm": 1.330393671989441, + "learning_rate": 1.7389109175389526e-05, + "loss": 0.8384, + "step": 7043 + }, + { + "epoch": 0.3118780647089426, + "grad_norm": 1.4739115238189697, + "learning_rate": 1.7388067757479684e-05, + "loss": 0.8052, + "step": 7044 + }, + { + "epoch": 0.31192234041375644, + "grad_norm": 1.2066047191619873, + "learning_rate": 1.7387026163111856e-05, + "loss": 0.7754, + "step": 7045 + }, + { + "epoch": 0.31196661611857035, + "grad_norm": 1.3209937810897827, + "learning_rate": 1.7385984392310925e-05, + "loss": 0.7964, + "step": 7046 + }, + { + "epoch": 0.3120108918233842, + "grad_norm": 1.5869431495666504, + "learning_rate": 1.7384942445101772e-05, + "loss": 0.8503, + "step": 7047 + }, + { + "epoch": 0.31205516752819806, + "grad_norm": 1.4035097360610962, + "learning_rate": 1.7383900321509283e-05, + "loss": 0.8624, + "step": 7048 + }, + { + "epoch": 0.312099443233012, + "grad_norm": 1.355278730392456, + "learning_rate": 1.738285802155834e-05, + "loss": 0.8254, + "step": 7049 + }, + { + "epoch": 0.31214371893782583, + "grad_norm": 1.2607733011245728, + "learning_rate": 1.738181554527385e-05, + "loss": 0.7663, + "step": 7050 + }, + { + "epoch": 0.31218799464263974, + "grad_norm": 1.1028807163238525, + "learning_rate": 1.7380772892680706e-05, + "loss": 0.8126, + "step": 7051 + }, + { + "epoch": 0.3122322703474536, + "grad_norm": 1.15080988407135, + "learning_rate": 1.737973006380381e-05, + "loss": 0.7371, + "step": 7052 + }, + { + "epoch": 0.31227654605226746, + "grad_norm": 1.4153308868408203, + "learning_rate": 1.7378687058668068e-05, + "loss": 0.8232, + "step": 7053 + }, + { + "epoch": 0.31232082175708137, + "grad_norm": 1.3160332441329956, + "learning_rate": 1.73776438772984e-05, + "loss": 0.8434, + "step": 7054 + }, + { + "epoch": 0.3123650974618952, + "grad_norm": 1.1414819955825806, + "learning_rate": 1.737660051971971e-05, + "loss": 0.722, + "step": 7055 + }, + { + "epoch": 0.3124093731667091, + "grad_norm": 2.0065484046936035, + "learning_rate": 1.737555698595692e-05, + "loss": 0.9686, + "step": 7056 + }, + { + "epoch": 0.312453648871523, + "grad_norm": 1.3942793607711792, + "learning_rate": 1.7374513276034963e-05, + "loss": 0.8785, + "step": 7057 + }, + { + "epoch": 0.31249792457633685, + "grad_norm": 1.2178195714950562, + "learning_rate": 1.7373469389978757e-05, + "loss": 0.7816, + "step": 7058 + }, + { + "epoch": 0.3125422002811507, + "grad_norm": 1.6935782432556152, + "learning_rate": 1.737242532781324e-05, + "loss": 0.9921, + "step": 7059 + }, + { + "epoch": 0.3125864759859646, + "grad_norm": 1.476381540298462, + "learning_rate": 1.7371381089563345e-05, + "loss": 0.8465, + "step": 7060 + }, + { + "epoch": 0.3126307516907785, + "grad_norm": 1.365927815437317, + "learning_rate": 1.737033667525402e-05, + "loss": 0.874, + "step": 7061 + }, + { + "epoch": 0.31267502739559233, + "grad_norm": 1.4243098497390747, + "learning_rate": 1.7369292084910203e-05, + "loss": 0.8925, + "step": 7062 + }, + { + "epoch": 0.31271930310040624, + "grad_norm": 1.1513237953186035, + "learning_rate": 1.7368247318556843e-05, + "loss": 0.7911, + "step": 7063 + }, + { + "epoch": 0.3127635788052201, + "grad_norm": 1.5979931354522705, + "learning_rate": 1.736720237621889e-05, + "loss": 0.9032, + "step": 7064 + }, + { + "epoch": 0.31280785451003396, + "grad_norm": 1.2164489030838013, + "learning_rate": 1.7366157257921314e-05, + "loss": 0.8407, + "step": 7065 + }, + { + "epoch": 0.31285213021484787, + "grad_norm": 1.118281602859497, + "learning_rate": 1.7365111963689067e-05, + "loss": 0.6603, + "step": 7066 + }, + { + "epoch": 0.3128964059196617, + "grad_norm": 1.2219336032867432, + "learning_rate": 1.736406649354712e-05, + "loss": 0.823, + "step": 7067 + }, + { + "epoch": 0.31294068162447564, + "grad_norm": 1.3293087482452393, + "learning_rate": 1.736302084752044e-05, + "loss": 0.7881, + "step": 7068 + }, + { + "epoch": 0.3129849573292895, + "grad_norm": 1.172745704650879, + "learning_rate": 1.7361975025633998e-05, + "loss": 0.8272, + "step": 7069 + }, + { + "epoch": 0.31302923303410335, + "grad_norm": 1.2532252073287964, + "learning_rate": 1.736092902791278e-05, + "loss": 0.7833, + "step": 7070 + }, + { + "epoch": 0.31307350873891726, + "grad_norm": 1.3288607597351074, + "learning_rate": 1.7359882854381768e-05, + "loss": 0.808, + "step": 7071 + }, + { + "epoch": 0.3131177844437311, + "grad_norm": 1.1636545658111572, + "learning_rate": 1.7358836505065945e-05, + "loss": 0.8061, + "step": 7072 + }, + { + "epoch": 0.313162060148545, + "grad_norm": 1.2179797887802124, + "learning_rate": 1.73577899799903e-05, + "loss": 0.814, + "step": 7073 + }, + { + "epoch": 0.3132063358533589, + "grad_norm": 1.2367507219314575, + "learning_rate": 1.735674327917984e-05, + "loss": 0.8341, + "step": 7074 + }, + { + "epoch": 0.31325061155817274, + "grad_norm": 2.0952255725860596, + "learning_rate": 1.735569640265955e-05, + "loss": 0.8119, + "step": 7075 + }, + { + "epoch": 0.3132948872629866, + "grad_norm": 1.4367469549179077, + "learning_rate": 1.7354649350454443e-05, + "loss": 0.8539, + "step": 7076 + }, + { + "epoch": 0.3133391629678005, + "grad_norm": 1.5468454360961914, + "learning_rate": 1.7353602122589528e-05, + "loss": 0.846, + "step": 7077 + }, + { + "epoch": 0.31338343867261437, + "grad_norm": 1.7960126399993896, + "learning_rate": 1.7352554719089808e-05, + "loss": 0.7442, + "step": 7078 + }, + { + "epoch": 0.3134277143774282, + "grad_norm": 1.4947305917739868, + "learning_rate": 1.735150713998031e-05, + "loss": 0.9002, + "step": 7079 + }, + { + "epoch": 0.31347199008224214, + "grad_norm": 1.2431079149246216, + "learning_rate": 1.735045938528605e-05, + "loss": 0.8374, + "step": 7080 + }, + { + "epoch": 0.313516265787056, + "grad_norm": 1.3448652029037476, + "learning_rate": 1.734941145503205e-05, + "loss": 0.8121, + "step": 7081 + }, + { + "epoch": 0.31356054149186985, + "grad_norm": 1.5632284879684448, + "learning_rate": 1.7348363349243336e-05, + "loss": 0.8883, + "step": 7082 + }, + { + "epoch": 0.31360481719668376, + "grad_norm": 1.3132052421569824, + "learning_rate": 1.7347315067944955e-05, + "loss": 0.8318, + "step": 7083 + }, + { + "epoch": 0.3136490929014976, + "grad_norm": 1.3461002111434937, + "learning_rate": 1.7346266611161933e-05, + "loss": 0.831, + "step": 7084 + }, + { + "epoch": 0.31369336860631153, + "grad_norm": 1.3216980695724487, + "learning_rate": 1.7345217978919312e-05, + "loss": 0.7747, + "step": 7085 + }, + { + "epoch": 0.3137376443111254, + "grad_norm": 1.4937502145767212, + "learning_rate": 1.7344169171242148e-05, + "loss": 0.8348, + "step": 7086 + }, + { + "epoch": 0.31378192001593924, + "grad_norm": 1.4963432550430298, + "learning_rate": 1.7343120188155475e-05, + "loss": 0.9184, + "step": 7087 + }, + { + "epoch": 0.31382619572075315, + "grad_norm": 1.175302505493164, + "learning_rate": 1.734207102968436e-05, + "loss": 0.7393, + "step": 7088 + }, + { + "epoch": 0.313870471425567, + "grad_norm": 1.2902568578720093, + "learning_rate": 1.7341021695853856e-05, + "loss": 0.8408, + "step": 7089 + }, + { + "epoch": 0.31391474713038087, + "grad_norm": 1.426615834236145, + "learning_rate": 1.7339972186689026e-05, + "loss": 0.8414, + "step": 7090 + }, + { + "epoch": 0.3139590228351948, + "grad_norm": 1.413749098777771, + "learning_rate": 1.7338922502214933e-05, + "loss": 0.8538, + "step": 7091 + }, + { + "epoch": 0.31400329854000864, + "grad_norm": 1.361387848854065, + "learning_rate": 1.7337872642456658e-05, + "loss": 0.8412, + "step": 7092 + }, + { + "epoch": 0.3140475742448225, + "grad_norm": 1.3179305791854858, + "learning_rate": 1.7336822607439266e-05, + "loss": 0.7889, + "step": 7093 + }, + { + "epoch": 0.3140918499496364, + "grad_norm": 1.3575971126556396, + "learning_rate": 1.733577239718784e-05, + "loss": 0.8496, + "step": 7094 + }, + { + "epoch": 0.31413612565445026, + "grad_norm": 1.4075387716293335, + "learning_rate": 1.733472201172746e-05, + "loss": 0.7729, + "step": 7095 + }, + { + "epoch": 0.3141804013592641, + "grad_norm": 1.48075270652771, + "learning_rate": 1.7333671451083222e-05, + "loss": 0.8496, + "step": 7096 + }, + { + "epoch": 0.31422467706407803, + "grad_norm": 1.2658358812332153, + "learning_rate": 1.7332620715280213e-05, + "loss": 0.8247, + "step": 7097 + }, + { + "epoch": 0.3142689527688919, + "grad_norm": 1.731605052947998, + "learning_rate": 1.7331569804343527e-05, + "loss": 0.8847, + "step": 7098 + }, + { + "epoch": 0.31431322847370574, + "grad_norm": 2.367382764816284, + "learning_rate": 1.7330518718298263e-05, + "loss": 0.7928, + "step": 7099 + }, + { + "epoch": 0.31435750417851965, + "grad_norm": 1.4744669198989868, + "learning_rate": 1.7329467457169533e-05, + "loss": 0.873, + "step": 7100 + }, + { + "epoch": 0.3144017798833335, + "grad_norm": 1.4044712781906128, + "learning_rate": 1.7328416020982438e-05, + "loss": 0.8039, + "step": 7101 + }, + { + "epoch": 0.31444605558814737, + "grad_norm": 1.4635109901428223, + "learning_rate": 1.7327364409762096e-05, + "loss": 0.8318, + "step": 7102 + }, + { + "epoch": 0.3144903312929613, + "grad_norm": 1.38943612575531, + "learning_rate": 1.7326312623533617e-05, + "loss": 0.764, + "step": 7103 + }, + { + "epoch": 0.31453460699777513, + "grad_norm": 1.7031463384628296, + "learning_rate": 1.7325260662322128e-05, + "loss": 0.9153, + "step": 7104 + }, + { + "epoch": 0.31457888270258905, + "grad_norm": 1.4407875537872314, + "learning_rate": 1.732420852615275e-05, + "loss": 0.8163, + "step": 7105 + }, + { + "epoch": 0.3146231584074029, + "grad_norm": 1.3256462812423706, + "learning_rate": 1.732315621505062e-05, + "loss": 0.7702, + "step": 7106 + }, + { + "epoch": 0.31466743411221676, + "grad_norm": 1.4754304885864258, + "learning_rate": 1.7322103729040868e-05, + "loss": 0.8237, + "step": 7107 + }, + { + "epoch": 0.31471170981703067, + "grad_norm": 1.3241024017333984, + "learning_rate": 1.7321051068148624e-05, + "loss": 0.7613, + "step": 7108 + }, + { + "epoch": 0.31475598552184453, + "grad_norm": 1.326882243156433, + "learning_rate": 1.731999823239904e-05, + "loss": 0.7974, + "step": 7109 + }, + { + "epoch": 0.3148002612266584, + "grad_norm": 1.6439173221588135, + "learning_rate": 1.7318945221817255e-05, + "loss": 0.9222, + "step": 7110 + }, + { + "epoch": 0.3148445369314723, + "grad_norm": 1.371450662612915, + "learning_rate": 1.7317892036428424e-05, + "loss": 0.8797, + "step": 7111 + }, + { + "epoch": 0.31488881263628615, + "grad_norm": 1.6139357089996338, + "learning_rate": 1.7316838676257702e-05, + "loss": 0.882, + "step": 7112 + }, + { + "epoch": 0.3149330883411, + "grad_norm": 1.5952727794647217, + "learning_rate": 1.7315785141330246e-05, + "loss": 0.8468, + "step": 7113 + }, + { + "epoch": 0.3149773640459139, + "grad_norm": 1.243935227394104, + "learning_rate": 1.7314731431671216e-05, + "loss": 0.6871, + "step": 7114 + }, + { + "epoch": 0.3150216397507278, + "grad_norm": 1.6440726518630981, + "learning_rate": 1.7313677547305782e-05, + "loss": 0.8634, + "step": 7115 + }, + { + "epoch": 0.31506591545554163, + "grad_norm": 1.1119557619094849, + "learning_rate": 1.731262348825912e-05, + "loss": 0.7163, + "step": 7116 + }, + { + "epoch": 0.31511019116035555, + "grad_norm": 1.2752147912979126, + "learning_rate": 1.7311569254556397e-05, + "loss": 0.8423, + "step": 7117 + }, + { + "epoch": 0.3151544668651694, + "grad_norm": 1.3567795753479004, + "learning_rate": 1.7310514846222792e-05, + "loss": 0.764, + "step": 7118 + }, + { + "epoch": 0.31519874256998326, + "grad_norm": 1.3223648071289062, + "learning_rate": 1.7309460263283496e-05, + "loss": 0.7572, + "step": 7119 + }, + { + "epoch": 0.31524301827479717, + "grad_norm": 1.329563021659851, + "learning_rate": 1.7308405505763693e-05, + "loss": 0.8042, + "step": 7120 + }, + { + "epoch": 0.315287293979611, + "grad_norm": 1.2671059370040894, + "learning_rate": 1.7307350573688574e-05, + "loss": 0.7879, + "step": 7121 + }, + { + "epoch": 0.31533156968442494, + "grad_norm": 1.1281383037567139, + "learning_rate": 1.7306295467083336e-05, + "loss": 0.7143, + "step": 7122 + }, + { + "epoch": 0.3153758453892388, + "grad_norm": 1.2415934801101685, + "learning_rate": 1.730524018597318e-05, + "loss": 0.7333, + "step": 7123 + }, + { + "epoch": 0.31542012109405265, + "grad_norm": 1.3087191581726074, + "learning_rate": 1.7304184730383313e-05, + "loss": 0.7783, + "step": 7124 + }, + { + "epoch": 0.31546439679886656, + "grad_norm": 1.7209445238113403, + "learning_rate": 1.7303129100338937e-05, + "loss": 0.8841, + "step": 7125 + }, + { + "epoch": 0.3155086725036804, + "grad_norm": 1.5430493354797363, + "learning_rate": 1.7302073295865272e-05, + "loss": 0.8945, + "step": 7126 + }, + { + "epoch": 0.3155529482084943, + "grad_norm": 1.4380494356155396, + "learning_rate": 1.730101731698753e-05, + "loss": 0.7432, + "step": 7127 + }, + { + "epoch": 0.3155972239133082, + "grad_norm": 1.6033010482788086, + "learning_rate": 1.7299961163730934e-05, + "loss": 0.8521, + "step": 7128 + }, + { + "epoch": 0.31564149961812205, + "grad_norm": 1.2412363290786743, + "learning_rate": 1.7298904836120712e-05, + "loss": 0.8172, + "step": 7129 + }, + { + "epoch": 0.3156857753229359, + "grad_norm": 1.431544303894043, + "learning_rate": 1.7297848334182088e-05, + "loss": 0.816, + "step": 7130 + }, + { + "epoch": 0.3157300510277498, + "grad_norm": 1.5901875495910645, + "learning_rate": 1.72967916579403e-05, + "loss": 0.9156, + "step": 7131 + }, + { + "epoch": 0.31577432673256367, + "grad_norm": 1.4280177354812622, + "learning_rate": 1.7295734807420585e-05, + "loss": 0.8023, + "step": 7132 + }, + { + "epoch": 0.3158186024373775, + "grad_norm": 1.3922380208969116, + "learning_rate": 1.7294677782648187e-05, + "loss": 0.8302, + "step": 7133 + }, + { + "epoch": 0.31586287814219144, + "grad_norm": 1.3252233266830444, + "learning_rate": 1.7293620583648345e-05, + "loss": 0.8341, + "step": 7134 + }, + { + "epoch": 0.3159071538470053, + "grad_norm": 1.4112776517868042, + "learning_rate": 1.7292563210446315e-05, + "loss": 0.8416, + "step": 7135 + }, + { + "epoch": 0.31595142955181915, + "grad_norm": 1.1899363994598389, + "learning_rate": 1.729150566306735e-05, + "loss": 0.824, + "step": 7136 + }, + { + "epoch": 0.31599570525663306, + "grad_norm": 1.1705518960952759, + "learning_rate": 1.7290447941536713e-05, + "loss": 0.7683, + "step": 7137 + }, + { + "epoch": 0.3160399809614469, + "grad_norm": 1.717146396636963, + "learning_rate": 1.7289390045879662e-05, + "loss": 0.8662, + "step": 7138 + }, + { + "epoch": 0.31608425666626083, + "grad_norm": 1.6452571153640747, + "learning_rate": 1.7288331976121463e-05, + "loss": 0.9387, + "step": 7139 + }, + { + "epoch": 0.3161285323710747, + "grad_norm": 1.5230746269226074, + "learning_rate": 1.728727373228739e-05, + "loss": 0.8257, + "step": 7140 + }, + { + "epoch": 0.31617280807588855, + "grad_norm": 1.5816128253936768, + "learning_rate": 1.7286215314402714e-05, + "loss": 0.938, + "step": 7141 + }, + { + "epoch": 0.31621708378070246, + "grad_norm": 1.8655107021331787, + "learning_rate": 1.7285156722492722e-05, + "loss": 0.8304, + "step": 7142 + }, + { + "epoch": 0.3162613594855163, + "grad_norm": 1.210970401763916, + "learning_rate": 1.7284097956582694e-05, + "loss": 0.8142, + "step": 7143 + }, + { + "epoch": 0.31630563519033017, + "grad_norm": 1.5841728448867798, + "learning_rate": 1.7283039016697913e-05, + "loss": 0.8759, + "step": 7144 + }, + { + "epoch": 0.3163499108951441, + "grad_norm": 1.5479685068130493, + "learning_rate": 1.7281979902863678e-05, + "loss": 0.9248, + "step": 7145 + }, + { + "epoch": 0.31639418659995794, + "grad_norm": 1.1750324964523315, + "learning_rate": 1.7280920615105284e-05, + "loss": 0.7338, + "step": 7146 + }, + { + "epoch": 0.3164384623047718, + "grad_norm": 1.1426446437835693, + "learning_rate": 1.7279861153448028e-05, + "loss": 0.809, + "step": 7147 + }, + { + "epoch": 0.3164827380095857, + "grad_norm": 1.0948379039764404, + "learning_rate": 1.7278801517917216e-05, + "loss": 0.7311, + "step": 7148 + }, + { + "epoch": 0.31652701371439956, + "grad_norm": 1.335402488708496, + "learning_rate": 1.7277741708538155e-05, + "loss": 0.8788, + "step": 7149 + }, + { + "epoch": 0.3165712894192134, + "grad_norm": 1.3137726783752441, + "learning_rate": 1.727668172533616e-05, + "loss": 0.8576, + "step": 7150 + }, + { + "epoch": 0.31661556512402733, + "grad_norm": 1.4171946048736572, + "learning_rate": 1.7275621568336545e-05, + "loss": 0.8803, + "step": 7151 + }, + { + "epoch": 0.3166598408288412, + "grad_norm": 1.2898954153060913, + "learning_rate": 1.7274561237564636e-05, + "loss": 0.8636, + "step": 7152 + }, + { + "epoch": 0.31670411653365504, + "grad_norm": 1.4495925903320312, + "learning_rate": 1.7273500733045755e-05, + "loss": 0.7929, + "step": 7153 + }, + { + "epoch": 0.31674839223846896, + "grad_norm": 1.5184746980667114, + "learning_rate": 1.7272440054805227e-05, + "loss": 0.7176, + "step": 7154 + }, + { + "epoch": 0.3167926679432828, + "grad_norm": 1.2007290124893188, + "learning_rate": 1.7271379202868394e-05, + "loss": 0.796, + "step": 7155 + }, + { + "epoch": 0.3168369436480967, + "grad_norm": 1.390836477279663, + "learning_rate": 1.7270318177260586e-05, + "loss": 0.9217, + "step": 7156 + }, + { + "epoch": 0.3168812193529106, + "grad_norm": 1.3314883708953857, + "learning_rate": 1.7269256978007152e-05, + "loss": 0.8173, + "step": 7157 + }, + { + "epoch": 0.31692549505772444, + "grad_norm": 1.5901907682418823, + "learning_rate": 1.726819560513343e-05, + "loss": 0.7778, + "step": 7158 + }, + { + "epoch": 0.31696977076253835, + "grad_norm": 1.4191272258758545, + "learning_rate": 1.7267134058664776e-05, + "loss": 0.8991, + "step": 7159 + }, + { + "epoch": 0.3170140464673522, + "grad_norm": 1.1554393768310547, + "learning_rate": 1.7266072338626544e-05, + "loss": 0.7694, + "step": 7160 + }, + { + "epoch": 0.31705832217216606, + "grad_norm": 1.2749624252319336, + "learning_rate": 1.7265010445044086e-05, + "loss": 0.8234, + "step": 7161 + }, + { + "epoch": 0.31710259787698, + "grad_norm": 1.158814787864685, + "learning_rate": 1.726394837794277e-05, + "loss": 0.7281, + "step": 7162 + }, + { + "epoch": 0.31714687358179383, + "grad_norm": 1.154296875, + "learning_rate": 1.7262886137347967e-05, + "loss": 0.8383, + "step": 7163 + }, + { + "epoch": 0.3171911492866077, + "grad_norm": 1.626664638519287, + "learning_rate": 1.7261823723285036e-05, + "loss": 0.9262, + "step": 7164 + }, + { + "epoch": 0.3172354249914216, + "grad_norm": 1.2259101867675781, + "learning_rate": 1.726076113577936e-05, + "loss": 0.809, + "step": 7165 + }, + { + "epoch": 0.31727970069623546, + "grad_norm": 1.16962730884552, + "learning_rate": 1.7259698374856316e-05, + "loss": 0.7981, + "step": 7166 + }, + { + "epoch": 0.3173239764010493, + "grad_norm": 1.4812049865722656, + "learning_rate": 1.725863544054129e-05, + "loss": 0.9009, + "step": 7167 + }, + { + "epoch": 0.3173682521058632, + "grad_norm": 1.590348482131958, + "learning_rate": 1.7257572332859664e-05, + "loss": 0.9109, + "step": 7168 + }, + { + "epoch": 0.3174125278106771, + "grad_norm": 1.4184048175811768, + "learning_rate": 1.7256509051836832e-05, + "loss": 0.8631, + "step": 7169 + }, + { + "epoch": 0.31745680351549094, + "grad_norm": 1.1220240592956543, + "learning_rate": 1.725544559749819e-05, + "loss": 0.7109, + "step": 7170 + }, + { + "epoch": 0.31750107922030485, + "grad_norm": 1.2746623754501343, + "learning_rate": 1.7254381969869137e-05, + "loss": 0.8246, + "step": 7171 + }, + { + "epoch": 0.3175453549251187, + "grad_norm": 1.240759253501892, + "learning_rate": 1.725331816897508e-05, + "loss": 0.8305, + "step": 7172 + }, + { + "epoch": 0.3175896306299326, + "grad_norm": 1.3655860424041748, + "learning_rate": 1.7252254194841422e-05, + "loss": 0.801, + "step": 7173 + }, + { + "epoch": 0.3176339063347465, + "grad_norm": 1.4120384454727173, + "learning_rate": 1.725119004749358e-05, + "loss": 0.877, + "step": 7174 + }, + { + "epoch": 0.31767818203956033, + "grad_norm": 1.1430584192276, + "learning_rate": 1.7250125726956965e-05, + "loss": 0.8194, + "step": 7175 + }, + { + "epoch": 0.31772245774437424, + "grad_norm": 3.370302438735962, + "learning_rate": 1.7249061233257004e-05, + "loss": 0.7868, + "step": 7176 + }, + { + "epoch": 0.3177667334491881, + "grad_norm": 1.2181577682495117, + "learning_rate": 1.724799656641911e-05, + "loss": 0.7642, + "step": 7177 + }, + { + "epoch": 0.31781100915400196, + "grad_norm": 1.5651158094406128, + "learning_rate": 1.7246931726468728e-05, + "loss": 0.8279, + "step": 7178 + }, + { + "epoch": 0.31785528485881587, + "grad_norm": 1.4477275609970093, + "learning_rate": 1.7245866713431278e-05, + "loss": 0.845, + "step": 7179 + }, + { + "epoch": 0.3178995605636297, + "grad_norm": 1.194959044456482, + "learning_rate": 1.7244801527332204e-05, + "loss": 0.7975, + "step": 7180 + }, + { + "epoch": 0.3179438362684436, + "grad_norm": 1.5146942138671875, + "learning_rate": 1.7243736168196942e-05, + "loss": 0.8869, + "step": 7181 + }, + { + "epoch": 0.3179881119732575, + "grad_norm": 1.2709707021713257, + "learning_rate": 1.724267063605094e-05, + "loss": 0.7282, + "step": 7182 + }, + { + "epoch": 0.31803238767807135, + "grad_norm": 1.5357189178466797, + "learning_rate": 1.724160493091965e-05, + "loss": 0.8601, + "step": 7183 + }, + { + "epoch": 0.3180766633828852, + "grad_norm": 1.3611700534820557, + "learning_rate": 1.724053905282852e-05, + "loss": 0.8905, + "step": 7184 + }, + { + "epoch": 0.3181209390876991, + "grad_norm": 1.4519884586334229, + "learning_rate": 1.723947300180301e-05, + "loss": 0.8749, + "step": 7185 + }, + { + "epoch": 0.318165214792513, + "grad_norm": 1.3912204504013062, + "learning_rate": 1.7238406777868584e-05, + "loss": 0.8377, + "step": 7186 + }, + { + "epoch": 0.31820949049732683, + "grad_norm": 1.6683474779129028, + "learning_rate": 1.72373403810507e-05, + "loss": 0.8753, + "step": 7187 + }, + { + "epoch": 0.31825376620214074, + "grad_norm": 1.5310077667236328, + "learning_rate": 1.723627381137484e-05, + "loss": 0.8314, + "step": 7188 + }, + { + "epoch": 0.3182980419069546, + "grad_norm": 1.5266822576522827, + "learning_rate": 1.7235207068866472e-05, + "loss": 0.9252, + "step": 7189 + }, + { + "epoch": 0.31834231761176845, + "grad_norm": 1.2665961980819702, + "learning_rate": 1.7234140153551072e-05, + "loss": 0.8241, + "step": 7190 + }, + { + "epoch": 0.31838659331658237, + "grad_norm": 1.3392069339752197, + "learning_rate": 1.7233073065454124e-05, + "loss": 0.9013, + "step": 7191 + }, + { + "epoch": 0.3184308690213962, + "grad_norm": 1.981190800666809, + "learning_rate": 1.7232005804601117e-05, + "loss": 0.863, + "step": 7192 + }, + { + "epoch": 0.31847514472621014, + "grad_norm": 1.4701117277145386, + "learning_rate": 1.723093837101754e-05, + "loss": 0.8575, + "step": 7193 + }, + { + "epoch": 0.318519420431024, + "grad_norm": 1.483885407447815, + "learning_rate": 1.7229870764728887e-05, + "loss": 0.9029, + "step": 7194 + }, + { + "epoch": 0.31856369613583785, + "grad_norm": 1.4433789253234863, + "learning_rate": 1.7228802985760656e-05, + "loss": 0.8186, + "step": 7195 + }, + { + "epoch": 0.31860797184065176, + "grad_norm": 1.436229944229126, + "learning_rate": 1.7227735034138352e-05, + "loss": 0.8195, + "step": 7196 + }, + { + "epoch": 0.3186522475454656, + "grad_norm": 1.9048079252243042, + "learning_rate": 1.7226666909887484e-05, + "loss": 0.8711, + "step": 7197 + }, + { + "epoch": 0.3186965232502795, + "grad_norm": 1.3915164470672607, + "learning_rate": 1.7225598613033564e-05, + "loss": 0.7723, + "step": 7198 + }, + { + "epoch": 0.3187407989550934, + "grad_norm": 1.2631703615188599, + "learning_rate": 1.72245301436021e-05, + "loss": 0.8091, + "step": 7199 + }, + { + "epoch": 0.31878507465990724, + "grad_norm": 1.4622514247894287, + "learning_rate": 1.7223461501618616e-05, + "loss": 0.8476, + "step": 7200 + }, + { + "epoch": 0.3188293503647211, + "grad_norm": 1.3006302118301392, + "learning_rate": 1.7222392687108635e-05, + "loss": 0.7999, + "step": 7201 + }, + { + "epoch": 0.318873626069535, + "grad_norm": 1.3845516443252563, + "learning_rate": 1.7221323700097685e-05, + "loss": 0.9125, + "step": 7202 + }, + { + "epoch": 0.31891790177434887, + "grad_norm": 1.2620104551315308, + "learning_rate": 1.72202545406113e-05, + "loss": 0.8073, + "step": 7203 + }, + { + "epoch": 0.3189621774791627, + "grad_norm": 1.4101300239562988, + "learning_rate": 1.7219185208675015e-05, + "loss": 0.9204, + "step": 7204 + }, + { + "epoch": 0.31900645318397663, + "grad_norm": 1.4872733354568481, + "learning_rate": 1.7218115704314365e-05, + "loss": 0.817, + "step": 7205 + }, + { + "epoch": 0.3190507288887905, + "grad_norm": 1.4714317321777344, + "learning_rate": 1.7217046027554904e-05, + "loss": 0.8791, + "step": 7206 + }, + { + "epoch": 0.31909500459360435, + "grad_norm": 1.3000028133392334, + "learning_rate": 1.721597617842217e-05, + "loss": 0.7797, + "step": 7207 + }, + { + "epoch": 0.31913928029841826, + "grad_norm": 1.5709012746810913, + "learning_rate": 1.7214906156941724e-05, + "loss": 0.8113, + "step": 7208 + }, + { + "epoch": 0.3191835560032321, + "grad_norm": 1.3519845008850098, + "learning_rate": 1.721383596313912e-05, + "loss": 0.8208, + "step": 7209 + }, + { + "epoch": 0.31922783170804603, + "grad_norm": 1.2910875082015991, + "learning_rate": 1.7212765597039914e-05, + "loss": 0.8276, + "step": 7210 + }, + { + "epoch": 0.3192721074128599, + "grad_norm": 1.1766778230667114, + "learning_rate": 1.721169505866968e-05, + "loss": 0.7288, + "step": 7211 + }, + { + "epoch": 0.31931638311767374, + "grad_norm": 1.5079233646392822, + "learning_rate": 1.7210624348053976e-05, + "loss": 0.8144, + "step": 7212 + }, + { + "epoch": 0.31936065882248765, + "grad_norm": 1.6283429861068726, + "learning_rate": 1.7209553465218382e-05, + "loss": 0.9161, + "step": 7213 + }, + { + "epoch": 0.3194049345273015, + "grad_norm": 1.3621643781661987, + "learning_rate": 1.7208482410188475e-05, + "loss": 0.8191, + "step": 7214 + }, + { + "epoch": 0.31944921023211537, + "grad_norm": 1.2630293369293213, + "learning_rate": 1.7207411182989834e-05, + "loss": 0.7703, + "step": 7215 + }, + { + "epoch": 0.3194934859369293, + "grad_norm": 1.217865228652954, + "learning_rate": 1.7206339783648044e-05, + "loss": 0.8627, + "step": 7216 + }, + { + "epoch": 0.31953776164174313, + "grad_norm": 1.4506688117980957, + "learning_rate": 1.72052682121887e-05, + "loss": 0.886, + "step": 7217 + }, + { + "epoch": 0.319582037346557, + "grad_norm": 1.7173677682876587, + "learning_rate": 1.720419646863739e-05, + "loss": 0.9183, + "step": 7218 + }, + { + "epoch": 0.3196263130513709, + "grad_norm": 1.554879903793335, + "learning_rate": 1.720312455301971e-05, + "loss": 0.8565, + "step": 7219 + }, + { + "epoch": 0.31967058875618476, + "grad_norm": 1.493577480316162, + "learning_rate": 1.7202052465361268e-05, + "loss": 0.6804, + "step": 7220 + }, + { + "epoch": 0.3197148644609986, + "grad_norm": 1.3949191570281982, + "learning_rate": 1.7200980205687666e-05, + "loss": 0.9004, + "step": 7221 + }, + { + "epoch": 0.3197591401658125, + "grad_norm": 1.4794938564300537, + "learning_rate": 1.7199907774024515e-05, + "loss": 0.8945, + "step": 7222 + }, + { + "epoch": 0.3198034158706264, + "grad_norm": 1.4069125652313232, + "learning_rate": 1.719883517039743e-05, + "loss": 0.7798, + "step": 7223 + }, + { + "epoch": 0.31984769157544024, + "grad_norm": 1.1234755516052246, + "learning_rate": 1.7197762394832027e-05, + "loss": 0.7386, + "step": 7224 + }, + { + "epoch": 0.31989196728025415, + "grad_norm": 1.3332021236419678, + "learning_rate": 1.719668944735393e-05, + "loss": 0.7935, + "step": 7225 + }, + { + "epoch": 0.319936242985068, + "grad_norm": 1.5446974039077759, + "learning_rate": 1.7195616327988768e-05, + "loss": 0.8596, + "step": 7226 + }, + { + "epoch": 0.3199805186898819, + "grad_norm": 1.3337887525558472, + "learning_rate": 1.7194543036762165e-05, + "loss": 0.8312, + "step": 7227 + }, + { + "epoch": 0.3200247943946958, + "grad_norm": 1.3946410417556763, + "learning_rate": 1.719346957369976e-05, + "loss": 0.8176, + "step": 7228 + }, + { + "epoch": 0.32006907009950963, + "grad_norm": 1.1915836334228516, + "learning_rate": 1.7192395938827197e-05, + "loss": 0.7972, + "step": 7229 + }, + { + "epoch": 0.32011334580432355, + "grad_norm": 1.5430870056152344, + "learning_rate": 1.719132213217011e-05, + "loss": 0.8345, + "step": 7230 + }, + { + "epoch": 0.3201576215091374, + "grad_norm": 1.339293122291565, + "learning_rate": 1.7190248153754147e-05, + "loss": 0.7913, + "step": 7231 + }, + { + "epoch": 0.32020189721395126, + "grad_norm": 1.2510302066802979, + "learning_rate": 1.7189174003604962e-05, + "loss": 0.7592, + "step": 7232 + }, + { + "epoch": 0.32024617291876517, + "grad_norm": 1.3902045488357544, + "learning_rate": 1.7188099681748213e-05, + "loss": 0.8067, + "step": 7233 + }, + { + "epoch": 0.320290448623579, + "grad_norm": 1.3263161182403564, + "learning_rate": 1.718702518820955e-05, + "loss": 0.8381, + "step": 7234 + }, + { + "epoch": 0.3203347243283929, + "grad_norm": 1.6005219221115112, + "learning_rate": 1.7185950523014647e-05, + "loss": 0.8755, + "step": 7235 + }, + { + "epoch": 0.3203790000332068, + "grad_norm": 1.587299108505249, + "learning_rate": 1.7184875686189167e-05, + "loss": 1.0095, + "step": 7236 + }, + { + "epoch": 0.32042327573802065, + "grad_norm": 1.1739128828048706, + "learning_rate": 1.7183800677758776e-05, + "loss": 0.8116, + "step": 7237 + }, + { + "epoch": 0.3204675514428345, + "grad_norm": 1.242738127708435, + "learning_rate": 1.718272549774916e-05, + "loss": 0.8308, + "step": 7238 + }, + { + "epoch": 0.3205118271476484, + "grad_norm": 1.3740226030349731, + "learning_rate": 1.7181650146185993e-05, + "loss": 0.8296, + "step": 7239 + }, + { + "epoch": 0.3205561028524623, + "grad_norm": 1.5133506059646606, + "learning_rate": 1.718057462309496e-05, + "loss": 0.7956, + "step": 7240 + }, + { + "epoch": 0.32060037855727613, + "grad_norm": 1.4509769678115845, + "learning_rate": 1.7179498928501747e-05, + "loss": 0.7989, + "step": 7241 + }, + { + "epoch": 0.32064465426209005, + "grad_norm": 1.084594964981079, + "learning_rate": 1.7178423062432052e-05, + "loss": 0.705, + "step": 7242 + }, + { + "epoch": 0.3206889299669039, + "grad_norm": 1.1560089588165283, + "learning_rate": 1.7177347024911562e-05, + "loss": 0.7836, + "step": 7243 + }, + { + "epoch": 0.3207332056717178, + "grad_norm": 1.505751609802246, + "learning_rate": 1.7176270815965986e-05, + "loss": 0.8764, + "step": 7244 + }, + { + "epoch": 0.32077748137653167, + "grad_norm": 1.3988537788391113, + "learning_rate": 1.717519443562102e-05, + "loss": 0.9478, + "step": 7245 + }, + { + "epoch": 0.3208217570813455, + "grad_norm": 1.4169471263885498, + "learning_rate": 1.7174117883902383e-05, + "loss": 0.8755, + "step": 7246 + }, + { + "epoch": 0.32086603278615944, + "grad_norm": 1.3144056797027588, + "learning_rate": 1.717304116083578e-05, + "loss": 0.798, + "step": 7247 + }, + { + "epoch": 0.3209103084909733, + "grad_norm": 1.2158596515655518, + "learning_rate": 1.7171964266446925e-05, + "loss": 0.7579, + "step": 7248 + }, + { + "epoch": 0.32095458419578715, + "grad_norm": 1.33452570438385, + "learning_rate": 1.7170887200761547e-05, + "loss": 0.835, + "step": 7249 + }, + { + "epoch": 0.32099885990060106, + "grad_norm": 1.6148357391357422, + "learning_rate": 1.7169809963805368e-05, + "loss": 0.8596, + "step": 7250 + }, + { + "epoch": 0.3210431356054149, + "grad_norm": 1.3204612731933594, + "learning_rate": 1.7168732555604114e-05, + "loss": 0.807, + "step": 7251 + }, + { + "epoch": 0.3210874113102288, + "grad_norm": 1.4089269638061523, + "learning_rate": 1.7167654976183523e-05, + "loss": 0.8049, + "step": 7252 + }, + { + "epoch": 0.3211316870150427, + "grad_norm": 2.3663017749786377, + "learning_rate": 1.7166577225569322e-05, + "loss": 0.8578, + "step": 7253 + }, + { + "epoch": 0.32117596271985654, + "grad_norm": 1.2471197843551636, + "learning_rate": 1.7165499303787266e-05, + "loss": 0.8407, + "step": 7254 + }, + { + "epoch": 0.3212202384246704, + "grad_norm": 1.3207298517227173, + "learning_rate": 1.7164421210863088e-05, + "loss": 0.8521, + "step": 7255 + }, + { + "epoch": 0.3212645141294843, + "grad_norm": 1.3107860088348389, + "learning_rate": 1.716334294682255e-05, + "loss": 0.7927, + "step": 7256 + }, + { + "epoch": 0.32130878983429817, + "grad_norm": 1.415189504623413, + "learning_rate": 1.7162264511691392e-05, + "loss": 0.8852, + "step": 7257 + }, + { + "epoch": 0.321353065539112, + "grad_norm": 1.533716082572937, + "learning_rate": 1.716118590549538e-05, + "loss": 0.8483, + "step": 7258 + }, + { + "epoch": 0.32139734124392594, + "grad_norm": 1.5684494972229004, + "learning_rate": 1.7160107128260274e-05, + "loss": 0.9119, + "step": 7259 + }, + { + "epoch": 0.3214416169487398, + "grad_norm": 1.2638524770736694, + "learning_rate": 1.715902818001184e-05, + "loss": 0.7814, + "step": 7260 + }, + { + "epoch": 0.3214858926535537, + "grad_norm": 1.3364264965057373, + "learning_rate": 1.7157949060775845e-05, + "loss": 0.828, + "step": 7261 + }, + { + "epoch": 0.32153016835836756, + "grad_norm": 1.327391505241394, + "learning_rate": 1.7156869770578066e-05, + "loss": 0.8533, + "step": 7262 + }, + { + "epoch": 0.3215744440631814, + "grad_norm": 1.2475676536560059, + "learning_rate": 1.715579030944428e-05, + "loss": 0.8771, + "step": 7263 + }, + { + "epoch": 0.32161871976799533, + "grad_norm": 1.4518910646438599, + "learning_rate": 1.7154710677400265e-05, + "loss": 0.8321, + "step": 7264 + }, + { + "epoch": 0.3216629954728092, + "grad_norm": 1.4308758974075317, + "learning_rate": 1.7153630874471817e-05, + "loss": 0.9049, + "step": 7265 + }, + { + "epoch": 0.32170727117762304, + "grad_norm": 1.3324151039123535, + "learning_rate": 1.7152550900684718e-05, + "loss": 0.8164, + "step": 7266 + }, + { + "epoch": 0.32175154688243696, + "grad_norm": 1.2463713884353638, + "learning_rate": 1.7151470756064764e-05, + "loss": 0.8969, + "step": 7267 + }, + { + "epoch": 0.3217958225872508, + "grad_norm": 1.1369363069534302, + "learning_rate": 1.7150390440637754e-05, + "loss": 0.8139, + "step": 7268 + }, + { + "epoch": 0.32184009829206467, + "grad_norm": 1.2647911310195923, + "learning_rate": 1.714930995442949e-05, + "loss": 0.8605, + "step": 7269 + }, + { + "epoch": 0.3218843739968786, + "grad_norm": 1.3366501331329346, + "learning_rate": 1.7148229297465783e-05, + "loss": 0.8411, + "step": 7270 + }, + { + "epoch": 0.32192864970169244, + "grad_norm": 1.3040554523468018, + "learning_rate": 1.7147148469772433e-05, + "loss": 0.8093, + "step": 7271 + }, + { + "epoch": 0.3219729254065063, + "grad_norm": 1.513083577156067, + "learning_rate": 1.7146067471375266e-05, + "loss": 0.8843, + "step": 7272 + }, + { + "epoch": 0.3220172011113202, + "grad_norm": 1.3500787019729614, + "learning_rate": 1.7144986302300093e-05, + "loss": 0.8115, + "step": 7273 + }, + { + "epoch": 0.32206147681613406, + "grad_norm": 1.346457839012146, + "learning_rate": 1.7143904962572745e-05, + "loss": 0.8715, + "step": 7274 + }, + { + "epoch": 0.3221057525209479, + "grad_norm": 1.2278536558151245, + "learning_rate": 1.7142823452219036e-05, + "loss": 0.8211, + "step": 7275 + }, + { + "epoch": 0.32215002822576183, + "grad_norm": 1.3610148429870605, + "learning_rate": 1.714174177126481e-05, + "loss": 0.8062, + "step": 7276 + }, + { + "epoch": 0.3221943039305757, + "grad_norm": 1.2214164733886719, + "learning_rate": 1.7140659919735897e-05, + "loss": 0.8224, + "step": 7277 + }, + { + "epoch": 0.32223857963538954, + "grad_norm": 1.4003413915634155, + "learning_rate": 1.7139577897658133e-05, + "loss": 0.8679, + "step": 7278 + }, + { + "epoch": 0.32228285534020346, + "grad_norm": 1.3298951387405396, + "learning_rate": 1.713849570505737e-05, + "loss": 0.8251, + "step": 7279 + }, + { + "epoch": 0.3223271310450173, + "grad_norm": 1.2468388080596924, + "learning_rate": 1.713741334195944e-05, + "loss": 0.7265, + "step": 7280 + }, + { + "epoch": 0.3223714067498312, + "grad_norm": 1.2434903383255005, + "learning_rate": 1.7136330808390213e-05, + "loss": 0.7119, + "step": 7281 + }, + { + "epoch": 0.3224156824546451, + "grad_norm": 1.4776283502578735, + "learning_rate": 1.713524810437553e-05, + "loss": 0.8552, + "step": 7282 + }, + { + "epoch": 0.32245995815945894, + "grad_norm": 1.2218413352966309, + "learning_rate": 1.7134165229941256e-05, + "loss": 0.7612, + "step": 7283 + }, + { + "epoch": 0.32250423386427285, + "grad_norm": 1.2347058057785034, + "learning_rate": 1.7133082185113255e-05, + "loss": 0.7333, + "step": 7284 + }, + { + "epoch": 0.3225485095690867, + "grad_norm": 1.672696828842163, + "learning_rate": 1.7131998969917393e-05, + "loss": 0.8811, + "step": 7285 + }, + { + "epoch": 0.32259278527390056, + "grad_norm": 1.2329914569854736, + "learning_rate": 1.7130915584379547e-05, + "loss": 0.7554, + "step": 7286 + }, + { + "epoch": 0.3226370609787145, + "grad_norm": 1.4199113845825195, + "learning_rate": 1.7129832028525584e-05, + "loss": 0.8382, + "step": 7287 + }, + { + "epoch": 0.32268133668352833, + "grad_norm": 1.1917273998260498, + "learning_rate": 1.712874830238139e-05, + "loss": 0.7854, + "step": 7288 + }, + { + "epoch": 0.3227256123883422, + "grad_norm": 1.4547984600067139, + "learning_rate": 1.7127664405972846e-05, + "loss": 0.8804, + "step": 7289 + }, + { + "epoch": 0.3227698880931561, + "grad_norm": 1.301003098487854, + "learning_rate": 1.7126580339325842e-05, + "loss": 0.7768, + "step": 7290 + }, + { + "epoch": 0.32281416379796996, + "grad_norm": 1.3610395193099976, + "learning_rate": 1.712549610246627e-05, + "loss": 0.7967, + "step": 7291 + }, + { + "epoch": 0.3228584395027838, + "grad_norm": 1.322603702545166, + "learning_rate": 1.7124411695420025e-05, + "loss": 0.7639, + "step": 7292 + }, + { + "epoch": 0.3229027152075977, + "grad_norm": 1.980635166168213, + "learning_rate": 1.7123327118213005e-05, + "loss": 1.0224, + "step": 7293 + }, + { + "epoch": 0.3229469909124116, + "grad_norm": 1.5464003086090088, + "learning_rate": 1.712224237087112e-05, + "loss": 0.8928, + "step": 7294 + }, + { + "epoch": 0.32299126661722544, + "grad_norm": 1.3384279012680054, + "learning_rate": 1.7121157453420278e-05, + "loss": 0.8305, + "step": 7295 + }, + { + "epoch": 0.32303554232203935, + "grad_norm": 1.4662028551101685, + "learning_rate": 1.7120072365886384e-05, + "loss": 0.8756, + "step": 7296 + }, + { + "epoch": 0.3230798180268532, + "grad_norm": 1.8285239934921265, + "learning_rate": 1.711898710829536e-05, + "loss": 0.9022, + "step": 7297 + }, + { + "epoch": 0.3231240937316671, + "grad_norm": 1.4296773672103882, + "learning_rate": 1.7117901680673122e-05, + "loss": 0.806, + "step": 7298 + }, + { + "epoch": 0.323168369436481, + "grad_norm": 1.5446701049804688, + "learning_rate": 1.7116816083045603e-05, + "loss": 0.726, + "step": 7299 + }, + { + "epoch": 0.32321264514129483, + "grad_norm": 1.2703229188919067, + "learning_rate": 1.7115730315438727e-05, + "loss": 0.8945, + "step": 7300 + }, + { + "epoch": 0.32325692084610874, + "grad_norm": 1.797262191772461, + "learning_rate": 1.711464437787842e-05, + "loss": 0.9474, + "step": 7301 + }, + { + "epoch": 0.3233011965509226, + "grad_norm": 1.9210163354873657, + "learning_rate": 1.711355827039063e-05, + "loss": 0.906, + "step": 7302 + }, + { + "epoch": 0.32334547225573645, + "grad_norm": 1.2842482328414917, + "learning_rate": 1.711247199300129e-05, + "loss": 0.809, + "step": 7303 + }, + { + "epoch": 0.32338974796055037, + "grad_norm": 1.3203551769256592, + "learning_rate": 1.711138554573635e-05, + "loss": 0.7866, + "step": 7304 + }, + { + "epoch": 0.3234340236653642, + "grad_norm": 1.453065276145935, + "learning_rate": 1.711029892862176e-05, + "loss": 0.8335, + "step": 7305 + }, + { + "epoch": 0.3234782993701781, + "grad_norm": 1.2501161098480225, + "learning_rate": 1.7109212141683464e-05, + "loss": 0.7617, + "step": 7306 + }, + { + "epoch": 0.323522575074992, + "grad_norm": 1.584562063217163, + "learning_rate": 1.7108125184947426e-05, + "loss": 0.7995, + "step": 7307 + }, + { + "epoch": 0.32356685077980585, + "grad_norm": 1.5026010274887085, + "learning_rate": 1.7107038058439606e-05, + "loss": 0.8049, + "step": 7308 + }, + { + "epoch": 0.3236111264846197, + "grad_norm": 1.120032787322998, + "learning_rate": 1.7105950762185967e-05, + "loss": 0.7502, + "step": 7309 + }, + { + "epoch": 0.3236554021894336, + "grad_norm": 1.240535020828247, + "learning_rate": 1.7104863296212482e-05, + "loss": 0.7555, + "step": 7310 + }, + { + "epoch": 0.3236996778942475, + "grad_norm": 1.572295904159546, + "learning_rate": 1.710377566054512e-05, + "loss": 0.8201, + "step": 7311 + }, + { + "epoch": 0.32374395359906133, + "grad_norm": 1.323527216911316, + "learning_rate": 1.7102687855209864e-05, + "loss": 0.8063, + "step": 7312 + }, + { + "epoch": 0.32378822930387524, + "grad_norm": 1.5040068626403809, + "learning_rate": 1.710159988023269e-05, + "loss": 0.8612, + "step": 7313 + }, + { + "epoch": 0.3238325050086891, + "grad_norm": 1.4128185510635376, + "learning_rate": 1.7100511735639583e-05, + "loss": 0.8217, + "step": 7314 + }, + { + "epoch": 0.323876780713503, + "grad_norm": 1.4815335273742676, + "learning_rate": 1.7099423421456537e-05, + "loss": 0.8405, + "step": 7315 + }, + { + "epoch": 0.32392105641831687, + "grad_norm": 1.4095598459243774, + "learning_rate": 1.7098334937709544e-05, + "loss": 0.8705, + "step": 7316 + }, + { + "epoch": 0.3239653321231307, + "grad_norm": 1.5649327039718628, + "learning_rate": 1.70972462844246e-05, + "loss": 0.8793, + "step": 7317 + }, + { + "epoch": 0.32400960782794463, + "grad_norm": 1.5590054988861084, + "learning_rate": 1.7096157461627708e-05, + "loss": 0.8929, + "step": 7318 + }, + { + "epoch": 0.3240538835327585, + "grad_norm": 1.1327966451644897, + "learning_rate": 1.709506846934487e-05, + "loss": 0.7885, + "step": 7319 + }, + { + "epoch": 0.32409815923757235, + "grad_norm": 1.4060312509536743, + "learning_rate": 1.70939793076021e-05, + "loss": 0.8527, + "step": 7320 + }, + { + "epoch": 0.32414243494238626, + "grad_norm": 1.3664722442626953, + "learning_rate": 1.7092889976425412e-05, + "loss": 0.784, + "step": 7321 + }, + { + "epoch": 0.3241867106472001, + "grad_norm": 1.3189502954483032, + "learning_rate": 1.709180047584082e-05, + "loss": 0.8904, + "step": 7322 + }, + { + "epoch": 0.324230986352014, + "grad_norm": 1.2291473150253296, + "learning_rate": 1.709071080587435e-05, + "loss": 0.8105, + "step": 7323 + }, + { + "epoch": 0.3242752620568279, + "grad_norm": 1.5900404453277588, + "learning_rate": 1.708962096655202e-05, + "loss": 0.9136, + "step": 7324 + }, + { + "epoch": 0.32431953776164174, + "grad_norm": 1.4365116357803345, + "learning_rate": 1.7088530957899873e-05, + "loss": 0.8295, + "step": 7325 + }, + { + "epoch": 0.3243638134664556, + "grad_norm": 1.4834684133529663, + "learning_rate": 1.7087440779943934e-05, + "loss": 0.8712, + "step": 7326 + }, + { + "epoch": 0.3244080891712695, + "grad_norm": 1.297985315322876, + "learning_rate": 1.7086350432710243e-05, + "loss": 0.8112, + "step": 7327 + }, + { + "epoch": 0.32445236487608337, + "grad_norm": 1.2010265588760376, + "learning_rate": 1.708525991622484e-05, + "loss": 0.8008, + "step": 7328 + }, + { + "epoch": 0.3244966405808972, + "grad_norm": 1.4724326133728027, + "learning_rate": 1.7084169230513776e-05, + "loss": 0.8144, + "step": 7329 + }, + { + "epoch": 0.32454091628571113, + "grad_norm": 1.2874224185943604, + "learning_rate": 1.708307837560309e-05, + "loss": 0.8171, + "step": 7330 + }, + { + "epoch": 0.324585191990525, + "grad_norm": 1.5348626375198364, + "learning_rate": 1.7081987351518854e-05, + "loss": 0.8096, + "step": 7331 + }, + { + "epoch": 0.3246294676953389, + "grad_norm": 1.2248069047927856, + "learning_rate": 1.7080896158287114e-05, + "loss": 0.7747, + "step": 7332 + }, + { + "epoch": 0.32467374340015276, + "grad_norm": 1.2237067222595215, + "learning_rate": 1.7079804795933933e-05, + "loss": 0.8307, + "step": 7333 + }, + { + "epoch": 0.3247180191049666, + "grad_norm": 1.288082480430603, + "learning_rate": 1.7078713264485376e-05, + "loss": 0.8402, + "step": 7334 + }, + { + "epoch": 0.3247622948097805, + "grad_norm": 1.538727879524231, + "learning_rate": 1.707762156396752e-05, + "loss": 0.8713, + "step": 7335 + }, + { + "epoch": 0.3248065705145944, + "grad_norm": 1.6346604824066162, + "learning_rate": 1.7076529694406437e-05, + "loss": 0.9302, + "step": 7336 + }, + { + "epoch": 0.32485084621940824, + "grad_norm": 1.5717763900756836, + "learning_rate": 1.7075437655828202e-05, + "loss": 0.7456, + "step": 7337 + }, + { + "epoch": 0.32489512192422215, + "grad_norm": 1.993767499923706, + "learning_rate": 1.7074345448258902e-05, + "loss": 0.9088, + "step": 7338 + }, + { + "epoch": 0.324939397629036, + "grad_norm": 1.6272436380386353, + "learning_rate": 1.7073253071724618e-05, + "loss": 0.8289, + "step": 7339 + }, + { + "epoch": 0.32498367333384987, + "grad_norm": 1.9171780347824097, + "learning_rate": 1.7072160526251446e-05, + "loss": 0.8502, + "step": 7340 + }, + { + "epoch": 0.3250279490386638, + "grad_norm": 1.232522964477539, + "learning_rate": 1.7071067811865477e-05, + "loss": 0.7819, + "step": 7341 + }, + { + "epoch": 0.32507222474347763, + "grad_norm": 1.4296936988830566, + "learning_rate": 1.706997492859281e-05, + "loss": 0.8983, + "step": 7342 + }, + { + "epoch": 0.3251165004482915, + "grad_norm": 1.4624817371368408, + "learning_rate": 1.706888187645955e-05, + "loss": 0.9081, + "step": 7343 + }, + { + "epoch": 0.3251607761531054, + "grad_norm": 1.4549198150634766, + "learning_rate": 1.70677886554918e-05, + "loss": 0.9554, + "step": 7344 + }, + { + "epoch": 0.32520505185791926, + "grad_norm": 1.7234915494918823, + "learning_rate": 1.7066695265715676e-05, + "loss": 0.856, + "step": 7345 + }, + { + "epoch": 0.3252493275627331, + "grad_norm": 1.6462996006011963, + "learning_rate": 1.7065601707157288e-05, + "loss": 0.8362, + "step": 7346 + }, + { + "epoch": 0.325293603267547, + "grad_norm": 1.1879620552062988, + "learning_rate": 1.7064507979842755e-05, + "loss": 0.8414, + "step": 7347 + }, + { + "epoch": 0.3253378789723609, + "grad_norm": 1.1507492065429688, + "learning_rate": 1.7063414083798204e-05, + "loss": 0.7962, + "step": 7348 + }, + { + "epoch": 0.32538215467717474, + "grad_norm": 1.3799375295639038, + "learning_rate": 1.7062320019049753e-05, + "loss": 0.8519, + "step": 7349 + }, + { + "epoch": 0.32542643038198865, + "grad_norm": 1.4829596281051636, + "learning_rate": 1.7061225785623547e-05, + "loss": 0.7362, + "step": 7350 + }, + { + "epoch": 0.3254707060868025, + "grad_norm": 1.4486945867538452, + "learning_rate": 1.7060131383545705e-05, + "loss": 0.9193, + "step": 7351 + }, + { + "epoch": 0.3255149817916164, + "grad_norm": 1.6252039670944214, + "learning_rate": 1.7059036812842377e-05, + "loss": 0.8659, + "step": 7352 + }, + { + "epoch": 0.3255592574964303, + "grad_norm": 1.6225509643554688, + "learning_rate": 1.7057942073539705e-05, + "loss": 0.8136, + "step": 7353 + }, + { + "epoch": 0.32560353320124413, + "grad_norm": 1.4269565343856812, + "learning_rate": 1.705684716566383e-05, + "loss": 0.8114, + "step": 7354 + }, + { + "epoch": 0.32564780890605804, + "grad_norm": 1.1453968286514282, + "learning_rate": 1.7055752089240907e-05, + "loss": 0.7743, + "step": 7355 + }, + { + "epoch": 0.3256920846108719, + "grad_norm": 1.1541613340377808, + "learning_rate": 1.705465684429709e-05, + "loss": 0.7756, + "step": 7356 + }, + { + "epoch": 0.32573636031568576, + "grad_norm": 1.2080155611038208, + "learning_rate": 1.7053561430858537e-05, + "loss": 0.7747, + "step": 7357 + }, + { + "epoch": 0.32578063602049967, + "grad_norm": 1.288564682006836, + "learning_rate": 1.7052465848951417e-05, + "loss": 0.8331, + "step": 7358 + }, + { + "epoch": 0.3258249117253135, + "grad_norm": 1.2036019563674927, + "learning_rate": 1.7051370098601888e-05, + "loss": 0.7695, + "step": 7359 + }, + { + "epoch": 0.3258691874301274, + "grad_norm": 1.322928547859192, + "learning_rate": 1.7050274179836127e-05, + "loss": 0.8376, + "step": 7360 + }, + { + "epoch": 0.3259134631349413, + "grad_norm": 1.5332164764404297, + "learning_rate": 1.7049178092680308e-05, + "loss": 0.845, + "step": 7361 + }, + { + "epoch": 0.32595773883975515, + "grad_norm": 1.3660197257995605, + "learning_rate": 1.7048081837160607e-05, + "loss": 0.8518, + "step": 7362 + }, + { + "epoch": 0.326002014544569, + "grad_norm": 1.2622414827346802, + "learning_rate": 1.7046985413303215e-05, + "loss": 0.8284, + "step": 7363 + }, + { + "epoch": 0.3260462902493829, + "grad_norm": 1.573508381843567, + "learning_rate": 1.704588882113431e-05, + "loss": 0.8138, + "step": 7364 + }, + { + "epoch": 0.3260905659541968, + "grad_norm": 1.1732144355773926, + "learning_rate": 1.7044792060680093e-05, + "loss": 0.788, + "step": 7365 + }, + { + "epoch": 0.32613484165901063, + "grad_norm": 1.3995548486709595, + "learning_rate": 1.7043695131966748e-05, + "loss": 0.8467, + "step": 7366 + }, + { + "epoch": 0.32617911736382454, + "grad_norm": 1.4789396524429321, + "learning_rate": 1.7042598035020482e-05, + "loss": 0.8102, + "step": 7367 + }, + { + "epoch": 0.3262233930686384, + "grad_norm": 1.2716459035873413, + "learning_rate": 1.70415007698675e-05, + "loss": 0.8304, + "step": 7368 + }, + { + "epoch": 0.3262676687734523, + "grad_norm": 1.4143521785736084, + "learning_rate": 1.7040403336534e-05, + "loss": 0.7931, + "step": 7369 + }, + { + "epoch": 0.32631194447826617, + "grad_norm": 1.1930341720581055, + "learning_rate": 1.70393057350462e-05, + "loss": 0.7767, + "step": 7370 + }, + { + "epoch": 0.32635622018308, + "grad_norm": 1.3078781366348267, + "learning_rate": 1.7038207965430316e-05, + "loss": 0.8563, + "step": 7371 + }, + { + "epoch": 0.32640049588789394, + "grad_norm": 1.107177734375, + "learning_rate": 1.7037110027712564e-05, + "loss": 0.7286, + "step": 7372 + }, + { + "epoch": 0.3264447715927078, + "grad_norm": 1.1779568195343018, + "learning_rate": 1.7036011921919172e-05, + "loss": 0.7769, + "step": 7373 + }, + { + "epoch": 0.32648904729752165, + "grad_norm": 1.0569331645965576, + "learning_rate": 1.7034913648076357e-05, + "loss": 0.7299, + "step": 7374 + }, + { + "epoch": 0.32653332300233556, + "grad_norm": 1.2770634889602661, + "learning_rate": 1.7033815206210365e-05, + "loss": 0.8285, + "step": 7375 + }, + { + "epoch": 0.3265775987071494, + "grad_norm": 1.135247826576233, + "learning_rate": 1.703271659634742e-05, + "loss": 0.7385, + "step": 7376 + }, + { + "epoch": 0.3266218744119633, + "grad_norm": 1.2574903964996338, + "learning_rate": 1.703161781851377e-05, + "loss": 0.8163, + "step": 7377 + }, + { + "epoch": 0.3266661501167772, + "grad_norm": 1.5366809368133545, + "learning_rate": 1.703051887273565e-05, + "loss": 0.8345, + "step": 7378 + }, + { + "epoch": 0.32671042582159104, + "grad_norm": 1.2408115863800049, + "learning_rate": 1.7029419759039308e-05, + "loss": 0.7516, + "step": 7379 + }, + { + "epoch": 0.3267547015264049, + "grad_norm": 1.576052188873291, + "learning_rate": 1.7028320477451004e-05, + "loss": 0.8107, + "step": 7380 + }, + { + "epoch": 0.3267989772312188, + "grad_norm": 1.39120352268219, + "learning_rate": 1.7027221027996986e-05, + "loss": 0.8338, + "step": 7381 + }, + { + "epoch": 0.32684325293603267, + "grad_norm": 1.3114758729934692, + "learning_rate": 1.7026121410703515e-05, + "loss": 0.7914, + "step": 7382 + }, + { + "epoch": 0.3268875286408465, + "grad_norm": 1.4675835371017456, + "learning_rate": 1.7025021625596852e-05, + "loss": 0.8865, + "step": 7383 + }, + { + "epoch": 0.32693180434566044, + "grad_norm": 1.3184345960617065, + "learning_rate": 1.7023921672703272e-05, + "loss": 0.7653, + "step": 7384 + }, + { + "epoch": 0.3269760800504743, + "grad_norm": 1.2796704769134521, + "learning_rate": 1.7022821552049044e-05, + "loss": 0.7457, + "step": 7385 + }, + { + "epoch": 0.3270203557552882, + "grad_norm": 1.3930468559265137, + "learning_rate": 1.7021721263660436e-05, + "loss": 0.8425, + "step": 7386 + }, + { + "epoch": 0.32706463146010206, + "grad_norm": 1.0591548681259155, + "learning_rate": 1.7020620807563735e-05, + "loss": 0.6611, + "step": 7387 + }, + { + "epoch": 0.3271089071649159, + "grad_norm": 1.3668079376220703, + "learning_rate": 1.7019520183785224e-05, + "loss": 0.8347, + "step": 7388 + }, + { + "epoch": 0.32715318286972983, + "grad_norm": 1.4076184034347534, + "learning_rate": 1.7018419392351185e-05, + "loss": 0.8144, + "step": 7389 + }, + { + "epoch": 0.3271974585745437, + "grad_norm": 1.3280736207962036, + "learning_rate": 1.701731843328792e-05, + "loss": 0.7655, + "step": 7390 + }, + { + "epoch": 0.32724173427935754, + "grad_norm": 1.3760590553283691, + "learning_rate": 1.7016217306621713e-05, + "loss": 0.8811, + "step": 7391 + }, + { + "epoch": 0.32728600998417146, + "grad_norm": 1.4416614770889282, + "learning_rate": 1.7015116012378868e-05, + "loss": 0.8172, + "step": 7392 + }, + { + "epoch": 0.3273302856889853, + "grad_norm": 1.2580076456069946, + "learning_rate": 1.701401455058569e-05, + "loss": 0.7983, + "step": 7393 + }, + { + "epoch": 0.32737456139379917, + "grad_norm": 1.4385305643081665, + "learning_rate": 1.7012912921268487e-05, + "loss": 0.8103, + "step": 7394 + }, + { + "epoch": 0.3274188370986131, + "grad_norm": 1.319301724433899, + "learning_rate": 1.7011811124453567e-05, + "loss": 0.8213, + "step": 7395 + }, + { + "epoch": 0.32746311280342694, + "grad_norm": 1.7028146982192993, + "learning_rate": 1.701070916016725e-05, + "loss": 0.8724, + "step": 7396 + }, + { + "epoch": 0.3275073885082408, + "grad_norm": 1.3042470216751099, + "learning_rate": 1.700960702843585e-05, + "loss": 0.81, + "step": 7397 + }, + { + "epoch": 0.3275516642130547, + "grad_norm": 1.4638844728469849, + "learning_rate": 1.7008504729285694e-05, + "loss": 0.7698, + "step": 7398 + }, + { + "epoch": 0.32759593991786856, + "grad_norm": 1.3554693460464478, + "learning_rate": 1.700740226274311e-05, + "loss": 0.7613, + "step": 7399 + }, + { + "epoch": 0.3276402156226824, + "grad_norm": 1.6337361335754395, + "learning_rate": 1.700629962883443e-05, + "loss": 0.9024, + "step": 7400 + }, + { + "epoch": 0.32768449132749633, + "grad_norm": 1.7637696266174316, + "learning_rate": 1.7005196827585986e-05, + "loss": 0.8987, + "step": 7401 + }, + { + "epoch": 0.3277287670323102, + "grad_norm": 1.3782052993774414, + "learning_rate": 1.700409385902412e-05, + "loss": 0.8073, + "step": 7402 + }, + { + "epoch": 0.3277730427371241, + "grad_norm": 1.3280452489852905, + "learning_rate": 1.7002990723175177e-05, + "loss": 0.8023, + "step": 7403 + }, + { + "epoch": 0.32781731844193795, + "grad_norm": 1.2783184051513672, + "learning_rate": 1.70018874200655e-05, + "loss": 0.7647, + "step": 7404 + }, + { + "epoch": 0.3278615941467518, + "grad_norm": 1.3111615180969238, + "learning_rate": 1.7000783949721443e-05, + "loss": 0.8208, + "step": 7405 + }, + { + "epoch": 0.3279058698515657, + "grad_norm": 1.401497721672058, + "learning_rate": 1.6999680312169366e-05, + "loss": 0.8634, + "step": 7406 + }, + { + "epoch": 0.3279501455563796, + "grad_norm": 1.3543587923049927, + "learning_rate": 1.699857650743562e-05, + "loss": 0.8652, + "step": 7407 + }, + { + "epoch": 0.32799442126119344, + "grad_norm": 1.4967528581619263, + "learning_rate": 1.6997472535546574e-05, + "loss": 0.9478, + "step": 7408 + }, + { + "epoch": 0.32803869696600735, + "grad_norm": 1.3758563995361328, + "learning_rate": 1.6996368396528592e-05, + "loss": 0.8871, + "step": 7409 + }, + { + "epoch": 0.3280829726708212, + "grad_norm": 1.3011524677276611, + "learning_rate": 1.699526409040805e-05, + "loss": 0.7507, + "step": 7410 + }, + { + "epoch": 0.32812724837563506, + "grad_norm": 1.5295088291168213, + "learning_rate": 1.6994159617211318e-05, + "loss": 0.806, + "step": 7411 + }, + { + "epoch": 0.328171524080449, + "grad_norm": 1.4548689126968384, + "learning_rate": 1.699305497696478e-05, + "loss": 0.7839, + "step": 7412 + }, + { + "epoch": 0.32821579978526283, + "grad_norm": 1.3446835279464722, + "learning_rate": 1.699195016969482e-05, + "loss": 0.7682, + "step": 7413 + }, + { + "epoch": 0.3282600754900767, + "grad_norm": 1.3621588945388794, + "learning_rate": 1.699084519542782e-05, + "loss": 0.8517, + "step": 7414 + }, + { + "epoch": 0.3283043511948906, + "grad_norm": 1.4089829921722412, + "learning_rate": 1.6989740054190177e-05, + "loss": 0.7676, + "step": 7415 + }, + { + "epoch": 0.32834862689970445, + "grad_norm": 1.4997520446777344, + "learning_rate": 1.6988634746008285e-05, + "loss": 0.8849, + "step": 7416 + }, + { + "epoch": 0.3283929026045183, + "grad_norm": 1.2908412218093872, + "learning_rate": 1.6987529270908542e-05, + "loss": 0.7923, + "step": 7417 + }, + { + "epoch": 0.3284371783093322, + "grad_norm": 1.3250147104263306, + "learning_rate": 1.698642362891735e-05, + "loss": 0.808, + "step": 7418 + }, + { + "epoch": 0.3284814540141461, + "grad_norm": 1.188542366027832, + "learning_rate": 1.6985317820061116e-05, + "loss": 0.7815, + "step": 7419 + }, + { + "epoch": 0.32852572971896, + "grad_norm": 1.2065892219543457, + "learning_rate": 1.698421184436626e-05, + "loss": 0.7871, + "step": 7420 + }, + { + "epoch": 0.32857000542377385, + "grad_norm": 1.0771762132644653, + "learning_rate": 1.6983105701859185e-05, + "loss": 0.7776, + "step": 7421 + }, + { + "epoch": 0.3286142811285877, + "grad_norm": 1.3366254568099976, + "learning_rate": 1.6981999392566318e-05, + "loss": 0.7619, + "step": 7422 + }, + { + "epoch": 0.3286585568334016, + "grad_norm": 1.3686519861221313, + "learning_rate": 1.698089291651408e-05, + "loss": 0.8234, + "step": 7423 + }, + { + "epoch": 0.3287028325382155, + "grad_norm": 1.4544909000396729, + "learning_rate": 1.6979786273728897e-05, + "loss": 0.8784, + "step": 7424 + }, + { + "epoch": 0.32874710824302933, + "grad_norm": 1.5296529531478882, + "learning_rate": 1.697867946423721e-05, + "loss": 0.9256, + "step": 7425 + }, + { + "epoch": 0.32879138394784324, + "grad_norm": 1.3884782791137695, + "learning_rate": 1.6977572488065438e-05, + "loss": 0.8103, + "step": 7426 + }, + { + "epoch": 0.3288356596526571, + "grad_norm": 1.5181918144226074, + "learning_rate": 1.697646534524003e-05, + "loss": 0.866, + "step": 7427 + }, + { + "epoch": 0.32887993535747095, + "grad_norm": 1.347997784614563, + "learning_rate": 1.6975358035787432e-05, + "loss": 0.8061, + "step": 7428 + }, + { + "epoch": 0.32892421106228487, + "grad_norm": 1.4110503196716309, + "learning_rate": 1.6974250559734084e-05, + "loss": 0.8634, + "step": 7429 + }, + { + "epoch": 0.3289684867670987, + "grad_norm": 1.1564234495162964, + "learning_rate": 1.6973142917106438e-05, + "loss": 0.712, + "step": 7430 + }, + { + "epoch": 0.3290127624719126, + "grad_norm": 1.441724181175232, + "learning_rate": 1.697203510793095e-05, + "loss": 0.8323, + "step": 7431 + }, + { + "epoch": 0.3290570381767265, + "grad_norm": 1.3260993957519531, + "learning_rate": 1.6970927132234085e-05, + "loss": 0.8546, + "step": 7432 + }, + { + "epoch": 0.32910131388154035, + "grad_norm": 1.7636560201644897, + "learning_rate": 1.69698189900423e-05, + "loss": 0.876, + "step": 7433 + }, + { + "epoch": 0.3291455895863542, + "grad_norm": 1.431725263595581, + "learning_rate": 1.696871068138206e-05, + "loss": 0.8254, + "step": 7434 + }, + { + "epoch": 0.3291898652911681, + "grad_norm": 1.2048423290252686, + "learning_rate": 1.6967602206279847e-05, + "loss": 0.7054, + "step": 7435 + }, + { + "epoch": 0.32923414099598197, + "grad_norm": 1.5248329639434814, + "learning_rate": 1.696649356476212e-05, + "loss": 0.7946, + "step": 7436 + }, + { + "epoch": 0.32927841670079583, + "grad_norm": 1.3973934650421143, + "learning_rate": 1.696538475685537e-05, + "loss": 0.7712, + "step": 7437 + }, + { + "epoch": 0.32932269240560974, + "grad_norm": 1.7632638216018677, + "learning_rate": 1.696427578258608e-05, + "loss": 0.7446, + "step": 7438 + }, + { + "epoch": 0.3293669681104236, + "grad_norm": 1.6174572706222534, + "learning_rate": 1.6963166641980732e-05, + "loss": 0.8296, + "step": 7439 + }, + { + "epoch": 0.3294112438152375, + "grad_norm": 1.47213613986969, + "learning_rate": 1.6962057335065814e-05, + "loss": 0.897, + "step": 7440 + }, + { + "epoch": 0.32945551952005137, + "grad_norm": 1.6455615758895874, + "learning_rate": 1.696094786186783e-05, + "loss": 0.8822, + "step": 7441 + }, + { + "epoch": 0.3294997952248652, + "grad_norm": 1.5410211086273193, + "learning_rate": 1.695983822241327e-05, + "loss": 0.8913, + "step": 7442 + }, + { + "epoch": 0.32954407092967913, + "grad_norm": 1.530085802078247, + "learning_rate": 1.6958728416728644e-05, + "loss": 0.841, + "step": 7443 + }, + { + "epoch": 0.329588346634493, + "grad_norm": 1.2628543376922607, + "learning_rate": 1.6957618444840455e-05, + "loss": 0.6681, + "step": 7444 + }, + { + "epoch": 0.32963262233930685, + "grad_norm": 1.2673767805099487, + "learning_rate": 1.6956508306775216e-05, + "loss": 0.8103, + "step": 7445 + }, + { + "epoch": 0.32967689804412076, + "grad_norm": 1.5161632299423218, + "learning_rate": 1.6955398002559437e-05, + "loss": 0.8762, + "step": 7446 + }, + { + "epoch": 0.3297211737489346, + "grad_norm": 1.5669723749160767, + "learning_rate": 1.6954287532219643e-05, + "loss": 0.8578, + "step": 7447 + }, + { + "epoch": 0.32976544945374847, + "grad_norm": 1.3903629779815674, + "learning_rate": 1.6953176895782353e-05, + "loss": 0.8054, + "step": 7448 + }, + { + "epoch": 0.3298097251585624, + "grad_norm": 1.6755257844924927, + "learning_rate": 1.6952066093274093e-05, + "loss": 0.9054, + "step": 7449 + }, + { + "epoch": 0.32985400086337624, + "grad_norm": 1.5322197675704956, + "learning_rate": 1.6950955124721398e-05, + "loss": 0.7899, + "step": 7450 + }, + { + "epoch": 0.3298982765681901, + "grad_norm": 1.2841358184814453, + "learning_rate": 1.6949843990150798e-05, + "loss": 0.8822, + "step": 7451 + }, + { + "epoch": 0.329942552273004, + "grad_norm": 1.4452728033065796, + "learning_rate": 1.694873268958883e-05, + "loss": 0.8097, + "step": 7452 + }, + { + "epoch": 0.32998682797781786, + "grad_norm": 1.235308051109314, + "learning_rate": 1.694762122306204e-05, + "loss": 0.8171, + "step": 7453 + }, + { + "epoch": 0.3300311036826317, + "grad_norm": 1.2873505353927612, + "learning_rate": 1.6946509590596978e-05, + "loss": 0.7841, + "step": 7454 + }, + { + "epoch": 0.33007537938744563, + "grad_norm": 1.54270601272583, + "learning_rate": 1.694539779222019e-05, + "loss": 0.8439, + "step": 7455 + }, + { + "epoch": 0.3301196550922595, + "grad_norm": 1.162192702293396, + "learning_rate": 1.6944285827958228e-05, + "loss": 0.7256, + "step": 7456 + }, + { + "epoch": 0.3301639307970734, + "grad_norm": 1.3096842765808105, + "learning_rate": 1.6943173697837655e-05, + "loss": 0.812, + "step": 7457 + }, + { + "epoch": 0.33020820650188726, + "grad_norm": 1.1248857975006104, + "learning_rate": 1.694206140188503e-05, + "loss": 0.7338, + "step": 7458 + }, + { + "epoch": 0.3302524822067011, + "grad_norm": 1.5115416049957275, + "learning_rate": 1.6940948940126924e-05, + "loss": 0.8381, + "step": 7459 + }, + { + "epoch": 0.330296757911515, + "grad_norm": 1.2501376867294312, + "learning_rate": 1.69398363125899e-05, + "loss": 0.7907, + "step": 7460 + }, + { + "epoch": 0.3303410336163289, + "grad_norm": 1.3223323822021484, + "learning_rate": 1.693872351930054e-05, + "loss": 0.8528, + "step": 7461 + }, + { + "epoch": 0.33038530932114274, + "grad_norm": 2.1228606700897217, + "learning_rate": 1.693761056028542e-05, + "loss": 0.9678, + "step": 7462 + }, + { + "epoch": 0.33042958502595665, + "grad_norm": 1.3695911169052124, + "learning_rate": 1.6936497435571116e-05, + "loss": 0.8229, + "step": 7463 + }, + { + "epoch": 0.3304738607307705, + "grad_norm": 1.4214926958084106, + "learning_rate": 1.693538414518422e-05, + "loss": 0.7937, + "step": 7464 + }, + { + "epoch": 0.33051813643558436, + "grad_norm": 1.5559600591659546, + "learning_rate": 1.693427068915132e-05, + "loss": 0.8757, + "step": 7465 + }, + { + "epoch": 0.3305624121403983, + "grad_norm": 1.6517255306243896, + "learning_rate": 1.693315706749901e-05, + "loss": 0.8397, + "step": 7466 + }, + { + "epoch": 0.33060668784521213, + "grad_norm": 1.2746158838272095, + "learning_rate": 1.6932043280253892e-05, + "loss": 0.7903, + "step": 7467 + }, + { + "epoch": 0.330650963550026, + "grad_norm": 1.1600621938705444, + "learning_rate": 1.693092932744256e-05, + "loss": 0.798, + "step": 7468 + }, + { + "epoch": 0.3306952392548399, + "grad_norm": 1.6120377779006958, + "learning_rate": 1.6929815209091627e-05, + "loss": 0.8278, + "step": 7469 + }, + { + "epoch": 0.33073951495965376, + "grad_norm": 1.350064992904663, + "learning_rate": 1.6928700925227705e-05, + "loss": 0.8816, + "step": 7470 + }, + { + "epoch": 0.3307837906644676, + "grad_norm": 1.4124186038970947, + "learning_rate": 1.6927586475877396e-05, + "loss": 0.8025, + "step": 7471 + }, + { + "epoch": 0.3308280663692815, + "grad_norm": 1.3349770307540894, + "learning_rate": 1.692647186106733e-05, + "loss": 0.9081, + "step": 7472 + }, + { + "epoch": 0.3308723420740954, + "grad_norm": 1.3073515892028809, + "learning_rate": 1.692535708082412e-05, + "loss": 0.7478, + "step": 7473 + }, + { + "epoch": 0.3309166177789093, + "grad_norm": 1.6710039377212524, + "learning_rate": 1.6924242135174395e-05, + "loss": 0.8269, + "step": 7474 + }, + { + "epoch": 0.33096089348372315, + "grad_norm": 1.3665014505386353, + "learning_rate": 1.6923127024144784e-05, + "loss": 0.7888, + "step": 7475 + }, + { + "epoch": 0.331005169188537, + "grad_norm": 1.1986429691314697, + "learning_rate": 1.692201174776192e-05, + "loss": 0.7469, + "step": 7476 + }, + { + "epoch": 0.3310494448933509, + "grad_norm": 1.465261459350586, + "learning_rate": 1.6920896306052443e-05, + "loss": 0.7889, + "step": 7477 + }, + { + "epoch": 0.3310937205981648, + "grad_norm": 1.1977415084838867, + "learning_rate": 1.6919780699042995e-05, + "loss": 0.7527, + "step": 7478 + }, + { + "epoch": 0.33113799630297863, + "grad_norm": 1.2346283197402954, + "learning_rate": 1.6918664926760214e-05, + "loss": 0.7842, + "step": 7479 + }, + { + "epoch": 0.33118227200779254, + "grad_norm": 1.5497318506240845, + "learning_rate": 1.691754898923076e-05, + "loss": 0.8574, + "step": 7480 + }, + { + "epoch": 0.3312265477126064, + "grad_norm": 1.5054839849472046, + "learning_rate": 1.691643288648128e-05, + "loss": 0.8757, + "step": 7481 + }, + { + "epoch": 0.33127082341742026, + "grad_norm": 1.5156209468841553, + "learning_rate": 1.6915316618538435e-05, + "loss": 0.8403, + "step": 7482 + }, + { + "epoch": 0.33131509912223417, + "grad_norm": 1.5308997631072998, + "learning_rate": 1.6914200185428876e-05, + "loss": 0.8745, + "step": 7483 + }, + { + "epoch": 0.331359374827048, + "grad_norm": 1.3439064025878906, + "learning_rate": 1.6913083587179278e-05, + "loss": 0.7769, + "step": 7484 + }, + { + "epoch": 0.3314036505318619, + "grad_norm": 1.3147286176681519, + "learning_rate": 1.691196682381631e-05, + "loss": 0.7596, + "step": 7485 + }, + { + "epoch": 0.3314479262366758, + "grad_norm": 1.6330205202102661, + "learning_rate": 1.6910849895366638e-05, + "loss": 0.8045, + "step": 7486 + }, + { + "epoch": 0.33149220194148965, + "grad_norm": 1.7193313837051392, + "learning_rate": 1.690973280185695e-05, + "loss": 0.8991, + "step": 7487 + }, + { + "epoch": 0.3315364776463035, + "grad_norm": 1.7932099103927612, + "learning_rate": 1.6908615543313912e-05, + "loss": 0.9096, + "step": 7488 + }, + { + "epoch": 0.3315807533511174, + "grad_norm": 1.8831549882888794, + "learning_rate": 1.690749811976422e-05, + "loss": 0.9072, + "step": 7489 + }, + { + "epoch": 0.3316250290559313, + "grad_norm": 1.5201784372329712, + "learning_rate": 1.690638053123456e-05, + "loss": 0.9216, + "step": 7490 + }, + { + "epoch": 0.3316693047607452, + "grad_norm": 1.1671618223190308, + "learning_rate": 1.6905262777751627e-05, + "loss": 0.757, + "step": 7491 + }, + { + "epoch": 0.33171358046555904, + "grad_norm": 1.3653337955474854, + "learning_rate": 1.6904144859342115e-05, + "loss": 0.7209, + "step": 7492 + }, + { + "epoch": 0.3317578561703729, + "grad_norm": 1.2578438520431519, + "learning_rate": 1.690302677603272e-05, + "loss": 0.8368, + "step": 7493 + }, + { + "epoch": 0.3318021318751868, + "grad_norm": 1.842938780784607, + "learning_rate": 1.6901908527850153e-05, + "loss": 0.8453, + "step": 7494 + }, + { + "epoch": 0.33184640758000067, + "grad_norm": 1.112287998199463, + "learning_rate": 1.6900790114821122e-05, + "loss": 0.7489, + "step": 7495 + }, + { + "epoch": 0.3318906832848145, + "grad_norm": 2.153784990310669, + "learning_rate": 1.6899671536972336e-05, + "loss": 0.9595, + "step": 7496 + }, + { + "epoch": 0.33193495898962844, + "grad_norm": 1.22135329246521, + "learning_rate": 1.6898552794330516e-05, + "loss": 0.777, + "step": 7497 + }, + { + "epoch": 0.3319792346944423, + "grad_norm": 1.4721734523773193, + "learning_rate": 1.6897433886922377e-05, + "loss": 0.8572, + "step": 7498 + }, + { + "epoch": 0.33202351039925615, + "grad_norm": 1.3644955158233643, + "learning_rate": 1.6896314814774647e-05, + "loss": 0.8922, + "step": 7499 + }, + { + "epoch": 0.33206778610407006, + "grad_norm": 1.304502010345459, + "learning_rate": 1.6895195577914053e-05, + "loss": 0.7742, + "step": 7500 + }, + { + "epoch": 0.3321120618088839, + "grad_norm": 1.4714593887329102, + "learning_rate": 1.6894076176367326e-05, + "loss": 0.7545, + "step": 7501 + }, + { + "epoch": 0.3321563375136978, + "grad_norm": 1.8457558155059814, + "learning_rate": 1.6892956610161202e-05, + "loss": 0.8503, + "step": 7502 + }, + { + "epoch": 0.3322006132185117, + "grad_norm": 1.65219247341156, + "learning_rate": 1.6891836879322424e-05, + "loss": 0.921, + "step": 7503 + }, + { + "epoch": 0.33224488892332554, + "grad_norm": 1.5827112197875977, + "learning_rate": 1.689071698387773e-05, + "loss": 0.8707, + "step": 7504 + }, + { + "epoch": 0.3322891646281394, + "grad_norm": 1.2963484525680542, + "learning_rate": 1.6889596923853876e-05, + "loss": 0.7619, + "step": 7505 + }, + { + "epoch": 0.3323334403329533, + "grad_norm": 1.2719203233718872, + "learning_rate": 1.6888476699277608e-05, + "loss": 0.7868, + "step": 7506 + }, + { + "epoch": 0.33237771603776717, + "grad_norm": 1.282610297203064, + "learning_rate": 1.6887356310175683e-05, + "loss": 0.8013, + "step": 7507 + }, + { + "epoch": 0.3324219917425811, + "grad_norm": 1.349393606185913, + "learning_rate": 1.6886235756574857e-05, + "loss": 0.7548, + "step": 7508 + }, + { + "epoch": 0.33246626744739494, + "grad_norm": 1.2989813089370728, + "learning_rate": 1.6885115038501903e-05, + "loss": 0.8627, + "step": 7509 + }, + { + "epoch": 0.3325105431522088, + "grad_norm": 1.333779215812683, + "learning_rate": 1.6883994155983577e-05, + "loss": 0.8054, + "step": 7510 + }, + { + "epoch": 0.3325548188570227, + "grad_norm": 1.3252205848693848, + "learning_rate": 1.6882873109046656e-05, + "loss": 0.7534, + "step": 7511 + }, + { + "epoch": 0.33259909456183656, + "grad_norm": 1.4173035621643066, + "learning_rate": 1.688175189771792e-05, + "loss": 0.797, + "step": 7512 + }, + { + "epoch": 0.3326433702666504, + "grad_norm": 1.5207312107086182, + "learning_rate": 1.6880630522024136e-05, + "loss": 0.8402, + "step": 7513 + }, + { + "epoch": 0.33268764597146433, + "grad_norm": 1.3101507425308228, + "learning_rate": 1.68795089819921e-05, + "loss": 0.7931, + "step": 7514 + }, + { + "epoch": 0.3327319216762782, + "grad_norm": 1.263494849205017, + "learning_rate": 1.6878387277648593e-05, + "loss": 0.7922, + "step": 7515 + }, + { + "epoch": 0.33277619738109204, + "grad_norm": 1.3410987854003906, + "learning_rate": 1.6877265409020405e-05, + "loss": 0.7928, + "step": 7516 + }, + { + "epoch": 0.33282047308590595, + "grad_norm": 1.4748988151550293, + "learning_rate": 1.687614337613433e-05, + "loss": 0.9108, + "step": 7517 + }, + { + "epoch": 0.3328647487907198, + "grad_norm": 1.328528881072998, + "learning_rate": 1.6875021179017175e-05, + "loss": 0.7625, + "step": 7518 + }, + { + "epoch": 0.33290902449553367, + "grad_norm": 1.3422828912734985, + "learning_rate": 1.6873898817695737e-05, + "loss": 0.8433, + "step": 7519 + }, + { + "epoch": 0.3329533002003476, + "grad_norm": 1.5144901275634766, + "learning_rate": 1.687277629219682e-05, + "loss": 0.8162, + "step": 7520 + }, + { + "epoch": 0.33299757590516144, + "grad_norm": 1.2199119329452515, + "learning_rate": 1.6871653602547232e-05, + "loss": 0.7584, + "step": 7521 + }, + { + "epoch": 0.3330418516099753, + "grad_norm": 1.3600910902023315, + "learning_rate": 1.6870530748773797e-05, + "loss": 0.7288, + "step": 7522 + }, + { + "epoch": 0.3330861273147892, + "grad_norm": 1.5731537342071533, + "learning_rate": 1.686940773090333e-05, + "loss": 0.9454, + "step": 7523 + }, + { + "epoch": 0.33313040301960306, + "grad_norm": 1.3468552827835083, + "learning_rate": 1.6868284548962655e-05, + "loss": 0.7402, + "step": 7524 + }, + { + "epoch": 0.3331746787244169, + "grad_norm": 1.2086071968078613, + "learning_rate": 1.686716120297859e-05, + "loss": 0.7822, + "step": 7525 + }, + { + "epoch": 0.33321895442923083, + "grad_norm": 1.3220460414886475, + "learning_rate": 1.6866037692977978e-05, + "loss": 0.8186, + "step": 7526 + }, + { + "epoch": 0.3332632301340447, + "grad_norm": 1.5590291023254395, + "learning_rate": 1.6864914018987643e-05, + "loss": 0.8862, + "step": 7527 + }, + { + "epoch": 0.3333075058388586, + "grad_norm": 1.4929335117340088, + "learning_rate": 1.6863790181034426e-05, + "loss": 0.9006, + "step": 7528 + }, + { + "epoch": 0.33335178154367245, + "grad_norm": 1.2245851755142212, + "learning_rate": 1.686266617914517e-05, + "loss": 0.8033, + "step": 7529 + }, + { + "epoch": 0.3333960572484863, + "grad_norm": 1.4690771102905273, + "learning_rate": 1.6861542013346718e-05, + "loss": 0.8942, + "step": 7530 + }, + { + "epoch": 0.3334403329533002, + "grad_norm": 1.9125034809112549, + "learning_rate": 1.686041768366592e-05, + "loss": 0.7652, + "step": 7531 + }, + { + "epoch": 0.3334846086581141, + "grad_norm": 1.46519136428833, + "learning_rate": 1.685929319012964e-05, + "loss": 0.8579, + "step": 7532 + }, + { + "epoch": 0.33352888436292794, + "grad_norm": 1.3297661542892456, + "learning_rate": 1.685816853276472e-05, + "loss": 0.7457, + "step": 7533 + }, + { + "epoch": 0.33357316006774185, + "grad_norm": 1.1724802255630493, + "learning_rate": 1.6857043711598033e-05, + "loss": 0.7222, + "step": 7534 + }, + { + "epoch": 0.3336174357725557, + "grad_norm": 1.3716590404510498, + "learning_rate": 1.6855918726656435e-05, + "loss": 0.8239, + "step": 7535 + }, + { + "epoch": 0.33366171147736956, + "grad_norm": 1.1419984102249146, + "learning_rate": 1.6854793577966803e-05, + "loss": 0.7186, + "step": 7536 + }, + { + "epoch": 0.33370598718218347, + "grad_norm": 1.311996340751648, + "learning_rate": 1.685366826555601e-05, + "loss": 0.7303, + "step": 7537 + }, + { + "epoch": 0.33375026288699733, + "grad_norm": 1.2634481191635132, + "learning_rate": 1.685254278945093e-05, + "loss": 0.749, + "step": 7538 + }, + { + "epoch": 0.3337945385918112, + "grad_norm": 1.218651294708252, + "learning_rate": 1.6851417149678442e-05, + "loss": 0.7839, + "step": 7539 + }, + { + "epoch": 0.3338388142966251, + "grad_norm": 1.4097427129745483, + "learning_rate": 1.6850291346265438e-05, + "loss": 0.8861, + "step": 7540 + }, + { + "epoch": 0.33388309000143895, + "grad_norm": 1.3399943113327026, + "learning_rate": 1.68491653792388e-05, + "loss": 0.8151, + "step": 7541 + }, + { + "epoch": 0.3339273657062528, + "grad_norm": 1.5216357707977295, + "learning_rate": 1.6848039248625426e-05, + "loss": 0.8294, + "step": 7542 + }, + { + "epoch": 0.3339716414110667, + "grad_norm": 1.5087261199951172, + "learning_rate": 1.6846912954452208e-05, + "loss": 0.7578, + "step": 7543 + }, + { + "epoch": 0.3340159171158806, + "grad_norm": 2.1539864540100098, + "learning_rate": 1.6845786496746052e-05, + "loss": 0.8619, + "step": 7544 + }, + { + "epoch": 0.3340601928206945, + "grad_norm": 1.9672552347183228, + "learning_rate": 1.6844659875533857e-05, + "loss": 0.9172, + "step": 7545 + }, + { + "epoch": 0.33410446852550835, + "grad_norm": 1.4930106401443481, + "learning_rate": 1.6843533090842535e-05, + "loss": 0.8157, + "step": 7546 + }, + { + "epoch": 0.3341487442303222, + "grad_norm": 1.9822261333465576, + "learning_rate": 1.6842406142698997e-05, + "loss": 0.9066, + "step": 7547 + }, + { + "epoch": 0.3341930199351361, + "grad_norm": 1.2524222135543823, + "learning_rate": 1.684127903113016e-05, + "loss": 0.8209, + "step": 7548 + }, + { + "epoch": 0.33423729563994997, + "grad_norm": 1.3587791919708252, + "learning_rate": 1.6840151756162945e-05, + "loss": 0.7579, + "step": 7549 + }, + { + "epoch": 0.33428157134476383, + "grad_norm": 1.2564219236373901, + "learning_rate": 1.6839024317824273e-05, + "loss": 0.7211, + "step": 7550 + }, + { + "epoch": 0.33432584704957774, + "grad_norm": 1.2762508392333984, + "learning_rate": 1.683789671614107e-05, + "loss": 0.7947, + "step": 7551 + }, + { + "epoch": 0.3343701227543916, + "grad_norm": 3.150327205657959, + "learning_rate": 1.683676895114028e-05, + "loss": 0.8567, + "step": 7552 + }, + { + "epoch": 0.33441439845920545, + "grad_norm": 1.2344555854797363, + "learning_rate": 1.6835641022848826e-05, + "loss": 0.7876, + "step": 7553 + }, + { + "epoch": 0.33445867416401937, + "grad_norm": 1.3867144584655762, + "learning_rate": 1.6834512931293653e-05, + "loss": 0.836, + "step": 7554 + }, + { + "epoch": 0.3345029498688332, + "grad_norm": 1.5230357646942139, + "learning_rate": 1.6833384676501698e-05, + "loss": 0.8151, + "step": 7555 + }, + { + "epoch": 0.3345472255736471, + "grad_norm": 1.4961333274841309, + "learning_rate": 1.683225625849992e-05, + "loss": 0.7955, + "step": 7556 + }, + { + "epoch": 0.334591501278461, + "grad_norm": 1.2425552606582642, + "learning_rate": 1.6831127677315266e-05, + "loss": 0.8158, + "step": 7557 + }, + { + "epoch": 0.33463577698327485, + "grad_norm": 1.331959843635559, + "learning_rate": 1.6829998932974683e-05, + "loss": 0.7975, + "step": 7558 + }, + { + "epoch": 0.3346800526880887, + "grad_norm": 1.4569683074951172, + "learning_rate": 1.682887002550514e-05, + "loss": 0.8814, + "step": 7559 + }, + { + "epoch": 0.3347243283929026, + "grad_norm": 1.4042682647705078, + "learning_rate": 1.6827740954933595e-05, + "loss": 0.8676, + "step": 7560 + }, + { + "epoch": 0.33476860409771647, + "grad_norm": 1.1769428253173828, + "learning_rate": 1.682661172128702e-05, + "loss": 0.7791, + "step": 7561 + }, + { + "epoch": 0.3348128798025304, + "grad_norm": 1.2960513830184937, + "learning_rate": 1.682548232459238e-05, + "loss": 0.8345, + "step": 7562 + }, + { + "epoch": 0.33485715550734424, + "grad_norm": 1.4648325443267822, + "learning_rate": 1.6824352764876653e-05, + "loss": 0.8973, + "step": 7563 + }, + { + "epoch": 0.3349014312121581, + "grad_norm": 1.3621810674667358, + "learning_rate": 1.682322304216682e-05, + "loss": 0.7843, + "step": 7564 + }, + { + "epoch": 0.334945706916972, + "grad_norm": 1.5943681001663208, + "learning_rate": 1.6822093156489852e-05, + "loss": 0.8398, + "step": 7565 + }, + { + "epoch": 0.33498998262178586, + "grad_norm": 1.4787846803665161, + "learning_rate": 1.682096310787275e-05, + "loss": 0.8845, + "step": 7566 + }, + { + "epoch": 0.3350342583265997, + "grad_norm": 1.6685972213745117, + "learning_rate": 1.6819832896342498e-05, + "loss": 0.7882, + "step": 7567 + }, + { + "epoch": 0.33507853403141363, + "grad_norm": 1.5661472082138062, + "learning_rate": 1.6818702521926087e-05, + "loss": 0.8012, + "step": 7568 + }, + { + "epoch": 0.3351228097362275, + "grad_norm": 1.3440135717391968, + "learning_rate": 1.6817571984650523e-05, + "loss": 0.8523, + "step": 7569 + }, + { + "epoch": 0.33516708544104135, + "grad_norm": 1.7610728740692139, + "learning_rate": 1.6816441284542798e-05, + "loss": 0.8263, + "step": 7570 + }, + { + "epoch": 0.33521136114585526, + "grad_norm": 1.397165298461914, + "learning_rate": 1.6815310421629923e-05, + "loss": 0.8114, + "step": 7571 + }, + { + "epoch": 0.3352556368506691, + "grad_norm": 1.7972151041030884, + "learning_rate": 1.6814179395938915e-05, + "loss": 0.8771, + "step": 7572 + }, + { + "epoch": 0.33529991255548297, + "grad_norm": 1.3257167339324951, + "learning_rate": 1.6813048207496775e-05, + "loss": 0.902, + "step": 7573 + }, + { + "epoch": 0.3353441882602969, + "grad_norm": 1.4074180126190186, + "learning_rate": 1.6811916856330524e-05, + "loss": 0.7908, + "step": 7574 + }, + { + "epoch": 0.33538846396511074, + "grad_norm": 1.3747456073760986, + "learning_rate": 1.6810785342467188e-05, + "loss": 0.8449, + "step": 7575 + }, + { + "epoch": 0.3354327396699246, + "grad_norm": 1.2434481382369995, + "learning_rate": 1.680965366593379e-05, + "loss": 0.7082, + "step": 7576 + }, + { + "epoch": 0.3354770153747385, + "grad_norm": 1.497312307357788, + "learning_rate": 1.6808521826757358e-05, + "loss": 0.923, + "step": 7577 + }, + { + "epoch": 0.33552129107955236, + "grad_norm": 1.1045039892196655, + "learning_rate": 1.6807389824964924e-05, + "loss": 0.7837, + "step": 7578 + }, + { + "epoch": 0.3355655667843663, + "grad_norm": 1.4117001295089722, + "learning_rate": 1.6806257660583534e-05, + "loss": 0.8622, + "step": 7579 + }, + { + "epoch": 0.33560984248918013, + "grad_norm": 1.4485981464385986, + "learning_rate": 1.6805125333640214e-05, + "loss": 0.8952, + "step": 7580 + }, + { + "epoch": 0.335654118193994, + "grad_norm": 1.3379300832748413, + "learning_rate": 1.680399284416202e-05, + "loss": 0.7148, + "step": 7581 + }, + { + "epoch": 0.3356983938988079, + "grad_norm": 1.3157031536102295, + "learning_rate": 1.6802860192175996e-05, + "loss": 0.7823, + "step": 7582 + }, + { + "epoch": 0.33574266960362176, + "grad_norm": 1.212856650352478, + "learning_rate": 1.6801727377709195e-05, + "loss": 0.7606, + "step": 7583 + }, + { + "epoch": 0.3357869453084356, + "grad_norm": 1.3789383172988892, + "learning_rate": 1.6800594400788675e-05, + "loss": 0.805, + "step": 7584 + }, + { + "epoch": 0.3358312210132495, + "grad_norm": 1.1967880725860596, + "learning_rate": 1.6799461261441494e-05, + "loss": 0.7625, + "step": 7585 + }, + { + "epoch": 0.3358754967180634, + "grad_norm": 1.576682686805725, + "learning_rate": 1.6798327959694714e-05, + "loss": 0.8338, + "step": 7586 + }, + { + "epoch": 0.33591977242287724, + "grad_norm": 1.5506829023361206, + "learning_rate": 1.6797194495575412e-05, + "loss": 0.9032, + "step": 7587 + }, + { + "epoch": 0.33596404812769115, + "grad_norm": 1.7252154350280762, + "learning_rate": 1.679606086911065e-05, + "loss": 0.9288, + "step": 7588 + }, + { + "epoch": 0.336008323832505, + "grad_norm": 1.3804634809494019, + "learning_rate": 1.679492708032751e-05, + "loss": 0.7949, + "step": 7589 + }, + { + "epoch": 0.33605259953731886, + "grad_norm": 1.4347710609436035, + "learning_rate": 1.679379312925307e-05, + "loss": 0.7968, + "step": 7590 + }, + { + "epoch": 0.3360968752421328, + "grad_norm": 1.4334120750427246, + "learning_rate": 1.6792659015914408e-05, + "loss": 0.8108, + "step": 7591 + }, + { + "epoch": 0.33614115094694663, + "grad_norm": 1.269039273262024, + "learning_rate": 1.6791524740338623e-05, + "loss": 0.7127, + "step": 7592 + }, + { + "epoch": 0.3361854266517605, + "grad_norm": 2.3683793544769287, + "learning_rate": 1.6790390302552793e-05, + "loss": 0.8478, + "step": 7593 + }, + { + "epoch": 0.3362297023565744, + "grad_norm": 1.572774052619934, + "learning_rate": 1.6789255702584023e-05, + "loss": 0.8441, + "step": 7594 + }, + { + "epoch": 0.33627397806138826, + "grad_norm": 1.5685288906097412, + "learning_rate": 1.6788120940459408e-05, + "loss": 0.9329, + "step": 7595 + }, + { + "epoch": 0.3363182537662021, + "grad_norm": 1.226014256477356, + "learning_rate": 1.6786986016206053e-05, + "loss": 0.8563, + "step": 7596 + }, + { + "epoch": 0.336362529471016, + "grad_norm": 1.3535124063491821, + "learning_rate": 1.678585092985106e-05, + "loss": 0.8, + "step": 7597 + }, + { + "epoch": 0.3364068051758299, + "grad_norm": 1.2818573713302612, + "learning_rate": 1.6784715681421543e-05, + "loss": 0.7942, + "step": 7598 + }, + { + "epoch": 0.3364510808806438, + "grad_norm": 1.6229157447814941, + "learning_rate": 1.678358027094462e-05, + "loss": 0.8332, + "step": 7599 + }, + { + "epoch": 0.33649535658545765, + "grad_norm": 1.6372263431549072, + "learning_rate": 1.67824446984474e-05, + "loss": 0.7859, + "step": 7600 + }, + { + "epoch": 0.3365396322902715, + "grad_norm": 1.518452525138855, + "learning_rate": 1.6781308963957016e-05, + "loss": 0.7804, + "step": 7601 + }, + { + "epoch": 0.3365839079950854, + "grad_norm": 1.36318039894104, + "learning_rate": 1.678017306750059e-05, + "loss": 0.821, + "step": 7602 + }, + { + "epoch": 0.3366281836998993, + "grad_norm": 1.669381856918335, + "learning_rate": 1.6779037009105246e-05, + "loss": 0.8458, + "step": 7603 + }, + { + "epoch": 0.33667245940471313, + "grad_norm": 1.2564047574996948, + "learning_rate": 1.6777900788798125e-05, + "loss": 0.8731, + "step": 7604 + }, + { + "epoch": 0.33671673510952704, + "grad_norm": 1.209505319595337, + "learning_rate": 1.6776764406606362e-05, + "loss": 0.754, + "step": 7605 + }, + { + "epoch": 0.3367610108143409, + "grad_norm": 1.2462718486785889, + "learning_rate": 1.67756278625571e-05, + "loss": 0.7983, + "step": 7606 + }, + { + "epoch": 0.33680528651915476, + "grad_norm": 1.5621734857559204, + "learning_rate": 1.6774491156677482e-05, + "loss": 0.8229, + "step": 7607 + }, + { + "epoch": 0.33684956222396867, + "grad_norm": 1.2643733024597168, + "learning_rate": 1.6773354288994657e-05, + "loss": 0.7619, + "step": 7608 + }, + { + "epoch": 0.3368938379287825, + "grad_norm": 1.2658867835998535, + "learning_rate": 1.6772217259535782e-05, + "loss": 0.7687, + "step": 7609 + }, + { + "epoch": 0.3369381136335964, + "grad_norm": 1.2685924768447876, + "learning_rate": 1.677108006832801e-05, + "loss": 0.8341, + "step": 7610 + }, + { + "epoch": 0.3369823893384103, + "grad_norm": 1.4897655248641968, + "learning_rate": 1.676994271539851e-05, + "loss": 0.8353, + "step": 7611 + }, + { + "epoch": 0.33702666504322415, + "grad_norm": 1.4707988500595093, + "learning_rate": 1.6768805200774437e-05, + "loss": 0.8291, + "step": 7612 + }, + { + "epoch": 0.337070940748038, + "grad_norm": 1.47547447681427, + "learning_rate": 1.676766752448296e-05, + "loss": 0.8669, + "step": 7613 + }, + { + "epoch": 0.3371152164528519, + "grad_norm": 1.1821610927581787, + "learning_rate": 1.6766529686551258e-05, + "loss": 0.6862, + "step": 7614 + }, + { + "epoch": 0.3371594921576658, + "grad_norm": 1.3971072435379028, + "learning_rate": 1.67653916870065e-05, + "loss": 0.8241, + "step": 7615 + }, + { + "epoch": 0.3372037678624797, + "grad_norm": 1.4249505996704102, + "learning_rate": 1.6764253525875875e-05, + "loss": 0.9309, + "step": 7616 + }, + { + "epoch": 0.33724804356729354, + "grad_norm": 1.120739221572876, + "learning_rate": 1.6763115203186558e-05, + "loss": 0.8184, + "step": 7617 + }, + { + "epoch": 0.3372923192721074, + "grad_norm": 1.2723344564437866, + "learning_rate": 1.6761976718965747e-05, + "loss": 0.8654, + "step": 7618 + }, + { + "epoch": 0.3373365949769213, + "grad_norm": 1.4267926216125488, + "learning_rate": 1.6760838073240623e-05, + "loss": 0.8172, + "step": 7619 + }, + { + "epoch": 0.33738087068173517, + "grad_norm": 1.2598670721054077, + "learning_rate": 1.675969926603839e-05, + "loss": 0.7802, + "step": 7620 + }, + { + "epoch": 0.337425146386549, + "grad_norm": 1.4750906229019165, + "learning_rate": 1.675856029738624e-05, + "loss": 1.0009, + "step": 7621 + }, + { + "epoch": 0.33746942209136294, + "grad_norm": 1.4371520280838013, + "learning_rate": 1.6757421167311383e-05, + "loss": 0.8911, + "step": 7622 + }, + { + "epoch": 0.3375136977961768, + "grad_norm": 1.18712317943573, + "learning_rate": 1.6756281875841025e-05, + "loss": 0.7741, + "step": 7623 + }, + { + "epoch": 0.33755797350099065, + "grad_norm": 1.218395709991455, + "learning_rate": 1.6755142423002372e-05, + "loss": 0.7443, + "step": 7624 + }, + { + "epoch": 0.33760224920580456, + "grad_norm": 1.1920702457427979, + "learning_rate": 1.6754002808822642e-05, + "loss": 0.7248, + "step": 7625 + }, + { + "epoch": 0.3376465249106184, + "grad_norm": 1.3502906560897827, + "learning_rate": 1.6752863033329058e-05, + "loss": 0.7974, + "step": 7626 + }, + { + "epoch": 0.3376908006154323, + "grad_norm": 1.4865221977233887, + "learning_rate": 1.6751723096548834e-05, + "loss": 0.8733, + "step": 7627 + }, + { + "epoch": 0.3377350763202462, + "grad_norm": 1.608349084854126, + "learning_rate": 1.6750582998509204e-05, + "loss": 0.9027, + "step": 7628 + }, + { + "epoch": 0.33777935202506004, + "grad_norm": 1.3546900749206543, + "learning_rate": 1.6749442739237397e-05, + "loss": 0.7599, + "step": 7629 + }, + { + "epoch": 0.3378236277298739, + "grad_norm": 1.4713953733444214, + "learning_rate": 1.6748302318760642e-05, + "loss": 0.9187, + "step": 7630 + }, + { + "epoch": 0.3378679034346878, + "grad_norm": 1.508314847946167, + "learning_rate": 1.6747161737106185e-05, + "loss": 0.8397, + "step": 7631 + }, + { + "epoch": 0.33791217913950167, + "grad_norm": 1.1963261365890503, + "learning_rate": 1.6746020994301263e-05, + "loss": 0.7759, + "step": 7632 + }, + { + "epoch": 0.3379564548443156, + "grad_norm": 1.3685075044631958, + "learning_rate": 1.6744880090373118e-05, + "loss": 0.7557, + "step": 7633 + }, + { + "epoch": 0.33800073054912944, + "grad_norm": 1.2460112571716309, + "learning_rate": 1.674373902534901e-05, + "loss": 0.7688, + "step": 7634 + }, + { + "epoch": 0.3380450062539433, + "grad_norm": 1.2501626014709473, + "learning_rate": 1.6742597799256182e-05, + "loss": 0.7926, + "step": 7635 + }, + { + "epoch": 0.3380892819587572, + "grad_norm": 1.3907264471054077, + "learning_rate": 1.6741456412121897e-05, + "loss": 0.7954, + "step": 7636 + }, + { + "epoch": 0.33813355766357106, + "grad_norm": 1.6681722402572632, + "learning_rate": 1.6740314863973414e-05, + "loss": 0.9261, + "step": 7637 + }, + { + "epoch": 0.3381778333683849, + "grad_norm": 1.4578484296798706, + "learning_rate": 1.6739173154838e-05, + "loss": 0.8539, + "step": 7638 + }, + { + "epoch": 0.33822210907319883, + "grad_norm": 1.2253552675247192, + "learning_rate": 1.673803128474292e-05, + "loss": 0.7731, + "step": 7639 + }, + { + "epoch": 0.3382663847780127, + "grad_norm": 1.6918684244155884, + "learning_rate": 1.673688925371545e-05, + "loss": 0.8426, + "step": 7640 + }, + { + "epoch": 0.33831066048282654, + "grad_norm": 1.38008713722229, + "learning_rate": 1.673574706178287e-05, + "loss": 0.7595, + "step": 7641 + }, + { + "epoch": 0.33835493618764045, + "grad_norm": 1.8047305345535278, + "learning_rate": 1.6734604708972454e-05, + "loss": 0.8041, + "step": 7642 + }, + { + "epoch": 0.3383992118924543, + "grad_norm": 1.5900933742523193, + "learning_rate": 1.6733462195311484e-05, + "loss": 0.8819, + "step": 7643 + }, + { + "epoch": 0.33844348759726817, + "grad_norm": 1.3171730041503906, + "learning_rate": 1.6732319520827255e-05, + "loss": 0.824, + "step": 7644 + }, + { + "epoch": 0.3384877633020821, + "grad_norm": 1.2530434131622314, + "learning_rate": 1.6731176685547055e-05, + "loss": 0.7744, + "step": 7645 + }, + { + "epoch": 0.33853203900689594, + "grad_norm": 1.2686129808425903, + "learning_rate": 1.673003368949818e-05, + "loss": 0.7715, + "step": 7646 + }, + { + "epoch": 0.3385763147117098, + "grad_norm": 1.3317571878433228, + "learning_rate": 1.6728890532707932e-05, + "loss": 0.805, + "step": 7647 + }, + { + "epoch": 0.3386205904165237, + "grad_norm": 1.2294014692306519, + "learning_rate": 1.6727747215203612e-05, + "loss": 0.7378, + "step": 7648 + }, + { + "epoch": 0.33866486612133756, + "grad_norm": 1.2971482276916504, + "learning_rate": 1.6726603737012527e-05, + "loss": 0.7986, + "step": 7649 + }, + { + "epoch": 0.33870914182615147, + "grad_norm": 1.3812811374664307, + "learning_rate": 1.6725460098161986e-05, + "loss": 0.8189, + "step": 7650 + }, + { + "epoch": 0.33875341753096533, + "grad_norm": 1.7303650379180908, + "learning_rate": 1.6724316298679313e-05, + "loss": 0.8194, + "step": 7651 + }, + { + "epoch": 0.3387976932357792, + "grad_norm": 1.2153695821762085, + "learning_rate": 1.6723172338591815e-05, + "loss": 0.7834, + "step": 7652 + }, + { + "epoch": 0.3388419689405931, + "grad_norm": 1.2251763343811035, + "learning_rate": 1.6722028217926822e-05, + "loss": 0.7853, + "step": 7653 + }, + { + "epoch": 0.33888624464540695, + "grad_norm": 1.234270691871643, + "learning_rate": 1.672088393671166e-05, + "loss": 0.7807, + "step": 7654 + }, + { + "epoch": 0.3389305203502208, + "grad_norm": 1.506890892982483, + "learning_rate": 1.6719739494973656e-05, + "loss": 0.8288, + "step": 7655 + }, + { + "epoch": 0.3389747960550347, + "grad_norm": 1.5689693689346313, + "learning_rate": 1.6718594892740146e-05, + "loss": 0.7956, + "step": 7656 + }, + { + "epoch": 0.3390190717598486, + "grad_norm": 1.1928316354751587, + "learning_rate": 1.6717450130038467e-05, + "loss": 0.7238, + "step": 7657 + }, + { + "epoch": 0.33906334746466243, + "grad_norm": 1.3328251838684082, + "learning_rate": 1.671630520689596e-05, + "loss": 0.8223, + "step": 7658 + }, + { + "epoch": 0.33910762316947635, + "grad_norm": 1.3051022291183472, + "learning_rate": 1.671516012333997e-05, + "loss": 0.7736, + "step": 7659 + }, + { + "epoch": 0.3391518988742902, + "grad_norm": 1.3085981607437134, + "learning_rate": 1.671401487939785e-05, + "loss": 0.7122, + "step": 7660 + }, + { + "epoch": 0.33919617457910406, + "grad_norm": 1.3680729866027832, + "learning_rate": 1.6712869475096954e-05, + "loss": 0.8578, + "step": 7661 + }, + { + "epoch": 0.33924045028391797, + "grad_norm": 1.2252033948898315, + "learning_rate": 1.6711723910464633e-05, + "loss": 0.7456, + "step": 7662 + }, + { + "epoch": 0.33928472598873183, + "grad_norm": 1.5190820693969727, + "learning_rate": 1.6710578185528254e-05, + "loss": 0.8457, + "step": 7663 + }, + { + "epoch": 0.3393290016935457, + "grad_norm": 1.3342643976211548, + "learning_rate": 1.670943230031518e-05, + "loss": 0.7527, + "step": 7664 + }, + { + "epoch": 0.3393732773983596, + "grad_norm": 1.4291871786117554, + "learning_rate": 1.6708286254852775e-05, + "loss": 0.7971, + "step": 7665 + }, + { + "epoch": 0.33941755310317345, + "grad_norm": 1.4776047468185425, + "learning_rate": 1.6707140049168417e-05, + "loss": 0.7928, + "step": 7666 + }, + { + "epoch": 0.33946182880798736, + "grad_norm": 1.31255304813385, + "learning_rate": 1.670599368328948e-05, + "loss": 0.7936, + "step": 7667 + }, + { + "epoch": 0.3395061045128012, + "grad_norm": 1.5473929643630981, + "learning_rate": 1.6704847157243344e-05, + "loss": 0.8065, + "step": 7668 + }, + { + "epoch": 0.3395503802176151, + "grad_norm": 1.7511097192764282, + "learning_rate": 1.670370047105739e-05, + "loss": 0.8407, + "step": 7669 + }, + { + "epoch": 0.339594655922429, + "grad_norm": 1.507092833518982, + "learning_rate": 1.670255362475901e-05, + "loss": 0.82, + "step": 7670 + }, + { + "epoch": 0.33963893162724285, + "grad_norm": 1.2827316522598267, + "learning_rate": 1.67014066183756e-05, + "loss": 0.8069, + "step": 7671 + }, + { + "epoch": 0.3396832073320567, + "grad_norm": 1.8155943155288696, + "learning_rate": 1.6700259451934542e-05, + "loss": 0.903, + "step": 7672 + }, + { + "epoch": 0.3397274830368706, + "grad_norm": 1.4177358150482178, + "learning_rate": 1.6699112125463245e-05, + "loss": 0.8412, + "step": 7673 + }, + { + "epoch": 0.33977175874168447, + "grad_norm": 1.4543297290802002, + "learning_rate": 1.669796463898911e-05, + "loss": 0.7856, + "step": 7674 + }, + { + "epoch": 0.3398160344464983, + "grad_norm": 1.5540300607681274, + "learning_rate": 1.6696816992539543e-05, + "loss": 0.8803, + "step": 7675 + }, + { + "epoch": 0.33986031015131224, + "grad_norm": 1.2317192554473877, + "learning_rate": 1.6695669186141953e-05, + "loss": 0.7161, + "step": 7676 + }, + { + "epoch": 0.3399045858561261, + "grad_norm": 1.318260908126831, + "learning_rate": 1.6694521219823757e-05, + "loss": 0.7782, + "step": 7677 + }, + { + "epoch": 0.33994886156093995, + "grad_norm": 1.546655535697937, + "learning_rate": 1.669337309361237e-05, + "loss": 0.7936, + "step": 7678 + }, + { + "epoch": 0.33999313726575386, + "grad_norm": 1.7034658193588257, + "learning_rate": 1.6692224807535223e-05, + "loss": 0.7984, + "step": 7679 + }, + { + "epoch": 0.3400374129705677, + "grad_norm": 1.4035747051239014, + "learning_rate": 1.6691076361619732e-05, + "loss": 0.8246, + "step": 7680 + }, + { + "epoch": 0.3400816886753816, + "grad_norm": 1.512602686882019, + "learning_rate": 1.6689927755893335e-05, + "loss": 0.8041, + "step": 7681 + }, + { + "epoch": 0.3401259643801955, + "grad_norm": 1.4374176263809204, + "learning_rate": 1.6688778990383453e-05, + "loss": 0.791, + "step": 7682 + }, + { + "epoch": 0.34017024008500935, + "grad_norm": 1.2373439073562622, + "learning_rate": 1.6687630065117536e-05, + "loss": 0.754, + "step": 7683 + }, + { + "epoch": 0.3402145157898232, + "grad_norm": 1.2827702760696411, + "learning_rate": 1.668648098012302e-05, + "loss": 0.7735, + "step": 7684 + }, + { + "epoch": 0.3402587914946371, + "grad_norm": 1.5299397706985474, + "learning_rate": 1.6685331735427347e-05, + "loss": 0.9201, + "step": 7685 + }, + { + "epoch": 0.34030306719945097, + "grad_norm": 1.660993218421936, + "learning_rate": 1.668418233105797e-05, + "loss": 0.8942, + "step": 7686 + }, + { + "epoch": 0.3403473429042649, + "grad_norm": 1.2635821104049683, + "learning_rate": 1.668303276704234e-05, + "loss": 0.7664, + "step": 7687 + }, + { + "epoch": 0.34039161860907874, + "grad_norm": 1.3041313886642456, + "learning_rate": 1.6681883043407918e-05, + "loss": 0.8154, + "step": 7688 + }, + { + "epoch": 0.3404358943138926, + "grad_norm": 1.4618732929229736, + "learning_rate": 1.6680733160182158e-05, + "loss": 0.8517, + "step": 7689 + }, + { + "epoch": 0.3404801700187065, + "grad_norm": 1.2307223081588745, + "learning_rate": 1.6679583117392524e-05, + "loss": 0.7946, + "step": 7690 + }, + { + "epoch": 0.34052444572352036, + "grad_norm": 1.3852424621582031, + "learning_rate": 1.6678432915066488e-05, + "loss": 0.9057, + "step": 7691 + }, + { + "epoch": 0.3405687214283342, + "grad_norm": 1.8500068187713623, + "learning_rate": 1.667728255323152e-05, + "loss": 0.8368, + "step": 7692 + }, + { + "epoch": 0.34061299713314813, + "grad_norm": 1.6501723527908325, + "learning_rate": 1.6676132031915092e-05, + "loss": 0.9245, + "step": 7693 + }, + { + "epoch": 0.340657272837962, + "grad_norm": 1.232061743736267, + "learning_rate": 1.667498135114469e-05, + "loss": 0.7578, + "step": 7694 + }, + { + "epoch": 0.34070154854277584, + "grad_norm": 1.2424262762069702, + "learning_rate": 1.6673830510947792e-05, + "loss": 0.813, + "step": 7695 + }, + { + "epoch": 0.34074582424758976, + "grad_norm": 2.3189051151275635, + "learning_rate": 1.6672679511351886e-05, + "loss": 0.7755, + "step": 7696 + }, + { + "epoch": 0.3407900999524036, + "grad_norm": 1.4585039615631104, + "learning_rate": 1.6671528352384464e-05, + "loss": 0.8337, + "step": 7697 + }, + { + "epoch": 0.34083437565721747, + "grad_norm": 1.3202682733535767, + "learning_rate": 1.6670377034073014e-05, + "loss": 0.8503, + "step": 7698 + }, + { + "epoch": 0.3408786513620314, + "grad_norm": 1.0848219394683838, + "learning_rate": 1.666922555644505e-05, + "loss": 0.7116, + "step": 7699 + }, + { + "epoch": 0.34092292706684524, + "grad_norm": 1.200721025466919, + "learning_rate": 1.6668073919528054e-05, + "loss": 0.7535, + "step": 7700 + }, + { + "epoch": 0.3409672027716591, + "grad_norm": 1.4882142543792725, + "learning_rate": 1.6666922123349547e-05, + "loss": 0.8763, + "step": 7701 + }, + { + "epoch": 0.341011478476473, + "grad_norm": 1.224509358406067, + "learning_rate": 1.666577016793703e-05, + "loss": 0.6999, + "step": 7702 + }, + { + "epoch": 0.34105575418128686, + "grad_norm": 1.6814601421356201, + "learning_rate": 1.6664618053318023e-05, + "loss": 0.8231, + "step": 7703 + }, + { + "epoch": 0.3411000298861008, + "grad_norm": 1.316429853439331, + "learning_rate": 1.6663465779520042e-05, + "loss": 0.8006, + "step": 7704 + }, + { + "epoch": 0.34114430559091463, + "grad_norm": 1.696314811706543, + "learning_rate": 1.66623133465706e-05, + "loss": 0.9601, + "step": 7705 + }, + { + "epoch": 0.3411885812957285, + "grad_norm": 1.331041932106018, + "learning_rate": 1.6661160754497237e-05, + "loss": 0.6871, + "step": 7706 + }, + { + "epoch": 0.3412328570005424, + "grad_norm": 1.217002272605896, + "learning_rate": 1.666000800332747e-05, + "loss": 0.7676, + "step": 7707 + }, + { + "epoch": 0.34127713270535626, + "grad_norm": 1.4325275421142578, + "learning_rate": 1.665885509308883e-05, + "loss": 0.8575, + "step": 7708 + }, + { + "epoch": 0.3413214084101701, + "grad_norm": 1.25831139087677, + "learning_rate": 1.6657702023808863e-05, + "loss": 0.7853, + "step": 7709 + }, + { + "epoch": 0.341365684114984, + "grad_norm": 1.4118678569793701, + "learning_rate": 1.66565487955151e-05, + "loss": 0.8228, + "step": 7710 + }, + { + "epoch": 0.3414099598197979, + "grad_norm": 1.7105439901351929, + "learning_rate": 1.665539540823509e-05, + "loss": 0.9136, + "step": 7711 + }, + { + "epoch": 0.34145423552461174, + "grad_norm": 1.4197109937667847, + "learning_rate": 1.6654241861996383e-05, + "loss": 0.8766, + "step": 7712 + }, + { + "epoch": 0.34149851122942565, + "grad_norm": 1.7069635391235352, + "learning_rate": 1.6653088156826526e-05, + "loss": 0.9075, + "step": 7713 + }, + { + "epoch": 0.3415427869342395, + "grad_norm": 1.318257451057434, + "learning_rate": 1.6651934292753074e-05, + "loss": 0.733, + "step": 7714 + }, + { + "epoch": 0.34158706263905336, + "grad_norm": 1.399388074874878, + "learning_rate": 1.6650780269803587e-05, + "loss": 0.7844, + "step": 7715 + }, + { + "epoch": 0.3416313383438673, + "grad_norm": 1.274424433708191, + "learning_rate": 1.664962608800563e-05, + "loss": 0.8512, + "step": 7716 + }, + { + "epoch": 0.34167561404868113, + "grad_norm": 1.516879677772522, + "learning_rate": 1.664847174738677e-05, + "loss": 0.8399, + "step": 7717 + }, + { + "epoch": 0.341719889753495, + "grad_norm": 1.4953784942626953, + "learning_rate": 1.664731724797457e-05, + "loss": 0.8449, + "step": 7718 + }, + { + "epoch": 0.3417641654583089, + "grad_norm": 1.3829392194747925, + "learning_rate": 1.6646162589796616e-05, + "loss": 0.7677, + "step": 7719 + }, + { + "epoch": 0.34180844116312276, + "grad_norm": 1.6593583822250366, + "learning_rate": 1.6645007772880478e-05, + "loss": 0.8808, + "step": 7720 + }, + { + "epoch": 0.34185271686793667, + "grad_norm": 1.528205156326294, + "learning_rate": 1.664385279725374e-05, + "loss": 0.7953, + "step": 7721 + }, + { + "epoch": 0.3418969925727505, + "grad_norm": 1.2263082265853882, + "learning_rate": 1.6642697662943988e-05, + "loss": 0.8004, + "step": 7722 + }, + { + "epoch": 0.3419412682775644, + "grad_norm": 1.347137212753296, + "learning_rate": 1.664154236997881e-05, + "loss": 0.7961, + "step": 7723 + }, + { + "epoch": 0.3419855439823783, + "grad_norm": 1.2470675706863403, + "learning_rate": 1.66403869183858e-05, + "loss": 0.7252, + "step": 7724 + }, + { + "epoch": 0.34202981968719215, + "grad_norm": 1.6212177276611328, + "learning_rate": 1.6639231308192555e-05, + "loss": 0.9287, + "step": 7725 + }, + { + "epoch": 0.342074095392006, + "grad_norm": 1.5406992435455322, + "learning_rate": 1.6638075539426675e-05, + "loss": 0.8643, + "step": 7726 + }, + { + "epoch": 0.3421183710968199, + "grad_norm": 1.449472188949585, + "learning_rate": 1.6636919612115768e-05, + "loss": 0.796, + "step": 7727 + }, + { + "epoch": 0.3421626468016338, + "grad_norm": 1.2023670673370361, + "learning_rate": 1.663576352628744e-05, + "loss": 0.7808, + "step": 7728 + }, + { + "epoch": 0.34220692250644763, + "grad_norm": 1.8037265539169312, + "learning_rate": 1.66346072819693e-05, + "loss": 0.9026, + "step": 7729 + }, + { + "epoch": 0.34225119821126154, + "grad_norm": 1.2854043245315552, + "learning_rate": 1.6633450879188973e-05, + "loss": 0.7041, + "step": 7730 + }, + { + "epoch": 0.3422954739160754, + "grad_norm": 1.1924458742141724, + "learning_rate": 1.663229431797407e-05, + "loss": 0.8369, + "step": 7731 + }, + { + "epoch": 0.34233974962088926, + "grad_norm": 1.3253679275512695, + "learning_rate": 1.6631137598352218e-05, + "loss": 0.7592, + "step": 7732 + }, + { + "epoch": 0.34238402532570317, + "grad_norm": 1.2428333759307861, + "learning_rate": 1.6629980720351042e-05, + "loss": 0.8062, + "step": 7733 + }, + { + "epoch": 0.342428301030517, + "grad_norm": 1.5835933685302734, + "learning_rate": 1.6628823683998175e-05, + "loss": 0.8064, + "step": 7734 + }, + { + "epoch": 0.3424725767353309, + "grad_norm": 1.476981282234192, + "learning_rate": 1.6627666489321254e-05, + "loss": 0.8695, + "step": 7735 + }, + { + "epoch": 0.3425168524401448, + "grad_norm": 2.7509915828704834, + "learning_rate": 1.6626509136347912e-05, + "loss": 0.9451, + "step": 7736 + }, + { + "epoch": 0.34256112814495865, + "grad_norm": 1.3285914659500122, + "learning_rate": 1.6625351625105796e-05, + "loss": 0.7567, + "step": 7737 + }, + { + "epoch": 0.34260540384977256, + "grad_norm": 1.4411569833755493, + "learning_rate": 1.6624193955622554e-05, + "loss": 0.8242, + "step": 7738 + }, + { + "epoch": 0.3426496795545864, + "grad_norm": 1.3197031021118164, + "learning_rate": 1.662303612792583e-05, + "loss": 0.7782, + "step": 7739 + }, + { + "epoch": 0.3426939552594003, + "grad_norm": 1.3806235790252686, + "learning_rate": 1.662187814204328e-05, + "loss": 0.8314, + "step": 7740 + }, + { + "epoch": 0.3427382309642142, + "grad_norm": 1.5386730432510376, + "learning_rate": 1.6620719998002567e-05, + "loss": 0.8005, + "step": 7741 + }, + { + "epoch": 0.34278250666902804, + "grad_norm": 1.3587325811386108, + "learning_rate": 1.661956169583134e-05, + "loss": 0.7795, + "step": 7742 + }, + { + "epoch": 0.3428267823738419, + "grad_norm": 1.297294020652771, + "learning_rate": 1.661840323555728e-05, + "loss": 0.7632, + "step": 7743 + }, + { + "epoch": 0.3428710580786558, + "grad_norm": 1.9778048992156982, + "learning_rate": 1.661724461720804e-05, + "loss": 0.7885, + "step": 7744 + }, + { + "epoch": 0.34291533378346967, + "grad_norm": 1.159245491027832, + "learning_rate": 1.6616085840811305e-05, + "loss": 0.749, + "step": 7745 + }, + { + "epoch": 0.3429596094882835, + "grad_norm": 1.3594768047332764, + "learning_rate": 1.6614926906394744e-05, + "loss": 0.8028, + "step": 7746 + }, + { + "epoch": 0.34300388519309744, + "grad_norm": 1.5706802606582642, + "learning_rate": 1.6613767813986045e-05, + "loss": 0.8975, + "step": 7747 + }, + { + "epoch": 0.3430481608979113, + "grad_norm": 1.34628427028656, + "learning_rate": 1.6612608563612883e-05, + "loss": 0.8147, + "step": 7748 + }, + { + "epoch": 0.34309243660272515, + "grad_norm": 1.4763554334640503, + "learning_rate": 1.661144915530295e-05, + "loss": 0.8657, + "step": 7749 + }, + { + "epoch": 0.34313671230753906, + "grad_norm": 1.1483739614486694, + "learning_rate": 1.6610289589083936e-05, + "loss": 0.7384, + "step": 7750 + }, + { + "epoch": 0.3431809880123529, + "grad_norm": 1.8103375434875488, + "learning_rate": 1.660912986498354e-05, + "loss": 1.0251, + "step": 7751 + }, + { + "epoch": 0.3432252637171668, + "grad_norm": 1.51897394657135, + "learning_rate": 1.6607969983029457e-05, + "loss": 0.9109, + "step": 7752 + }, + { + "epoch": 0.3432695394219807, + "grad_norm": 1.2856616973876953, + "learning_rate": 1.660680994324939e-05, + "loss": 0.7829, + "step": 7753 + }, + { + "epoch": 0.34331381512679454, + "grad_norm": 1.3061031103134155, + "learning_rate": 1.660564974567105e-05, + "loss": 0.7752, + "step": 7754 + }, + { + "epoch": 0.3433580908316084, + "grad_norm": 1.2448488473892212, + "learning_rate": 1.660448939032214e-05, + "loss": 0.7239, + "step": 7755 + }, + { + "epoch": 0.3434023665364223, + "grad_norm": 1.56528639793396, + "learning_rate": 1.6603328877230384e-05, + "loss": 0.9514, + "step": 7756 + }, + { + "epoch": 0.34344664224123617, + "grad_norm": 1.5789200067520142, + "learning_rate": 1.660216820642349e-05, + "loss": 0.8311, + "step": 7757 + }, + { + "epoch": 0.3434909179460501, + "grad_norm": 1.4651923179626465, + "learning_rate": 1.660100737792919e-05, + "loss": 0.9494, + "step": 7758 + }, + { + "epoch": 0.34353519365086393, + "grad_norm": 1.3539552688598633, + "learning_rate": 1.6599846391775197e-05, + "loss": 0.757, + "step": 7759 + }, + { + "epoch": 0.3435794693556778, + "grad_norm": 1.2221370935440063, + "learning_rate": 1.6598685247989248e-05, + "loss": 0.7688, + "step": 7760 + }, + { + "epoch": 0.3436237450604917, + "grad_norm": 1.6941617727279663, + "learning_rate": 1.6597523946599076e-05, + "loss": 0.7942, + "step": 7761 + }, + { + "epoch": 0.34366802076530556, + "grad_norm": 1.4326268434524536, + "learning_rate": 1.6596362487632414e-05, + "loss": 0.8381, + "step": 7762 + }, + { + "epoch": 0.3437122964701194, + "grad_norm": 1.1673697233200073, + "learning_rate": 1.659520087111701e-05, + "loss": 0.7314, + "step": 7763 + }, + { + "epoch": 0.34375657217493333, + "grad_norm": 1.1763802766799927, + "learning_rate": 1.6594039097080598e-05, + "loss": 0.8007, + "step": 7764 + }, + { + "epoch": 0.3438008478797472, + "grad_norm": 1.2948426008224487, + "learning_rate": 1.6592877165550935e-05, + "loss": 0.78, + "step": 7765 + }, + { + "epoch": 0.34384512358456104, + "grad_norm": 1.7222973108291626, + "learning_rate": 1.659171507655577e-05, + "loss": 0.832, + "step": 7766 + }, + { + "epoch": 0.34388939928937495, + "grad_norm": 1.4896104335784912, + "learning_rate": 1.659055283012285e-05, + "loss": 0.7647, + "step": 7767 + }, + { + "epoch": 0.3439336749941888, + "grad_norm": 1.5303854942321777, + "learning_rate": 1.6589390426279953e-05, + "loss": 0.8688, + "step": 7768 + }, + { + "epoch": 0.34397795069900267, + "grad_norm": 1.5289671421051025, + "learning_rate": 1.6588227865054822e-05, + "loss": 0.7316, + "step": 7769 + }, + { + "epoch": 0.3440222264038166, + "grad_norm": 1.4675337076187134, + "learning_rate": 1.6587065146475232e-05, + "loss": 0.7526, + "step": 7770 + }, + { + "epoch": 0.34406650210863043, + "grad_norm": 1.3842235803604126, + "learning_rate": 1.6585902270568963e-05, + "loss": 0.8114, + "step": 7771 + }, + { + "epoch": 0.3441107778134443, + "grad_norm": 1.3965092897415161, + "learning_rate": 1.6584739237363776e-05, + "loss": 0.8747, + "step": 7772 + }, + { + "epoch": 0.3441550535182582, + "grad_norm": 1.8459765911102295, + "learning_rate": 1.6583576046887454e-05, + "loss": 0.9001, + "step": 7773 + }, + { + "epoch": 0.34419932922307206, + "grad_norm": 1.2621206045150757, + "learning_rate": 1.658241269916778e-05, + "loss": 0.7772, + "step": 7774 + }, + { + "epoch": 0.34424360492788597, + "grad_norm": 1.4796113967895508, + "learning_rate": 1.6581249194232533e-05, + "loss": 0.7849, + "step": 7775 + }, + { + "epoch": 0.3442878806326998, + "grad_norm": 1.4034348726272583, + "learning_rate": 1.6580085532109512e-05, + "loss": 0.7903, + "step": 7776 + }, + { + "epoch": 0.3443321563375137, + "grad_norm": 2.442828893661499, + "learning_rate": 1.6578921712826505e-05, + "loss": 0.8533, + "step": 7777 + }, + { + "epoch": 0.3443764320423276, + "grad_norm": 1.4069186449050903, + "learning_rate": 1.6577757736411313e-05, + "loss": 0.8183, + "step": 7778 + }, + { + "epoch": 0.34442070774714145, + "grad_norm": 1.4036312103271484, + "learning_rate": 1.6576593602891727e-05, + "loss": 0.7737, + "step": 7779 + }, + { + "epoch": 0.3444649834519553, + "grad_norm": 1.2630153894424438, + "learning_rate": 1.6575429312295562e-05, + "loss": 0.7551, + "step": 7780 + }, + { + "epoch": 0.3445092591567692, + "grad_norm": 1.4078779220581055, + "learning_rate": 1.657426486465062e-05, + "loss": 0.7679, + "step": 7781 + }, + { + "epoch": 0.3445535348615831, + "grad_norm": 1.5155022144317627, + "learning_rate": 1.6573100259984714e-05, + "loss": 0.7717, + "step": 7782 + }, + { + "epoch": 0.34459781056639693, + "grad_norm": 1.2012172937393188, + "learning_rate": 1.6571935498325663e-05, + "loss": 0.7468, + "step": 7783 + }, + { + "epoch": 0.34464208627121085, + "grad_norm": 1.4201045036315918, + "learning_rate": 1.657077057970128e-05, + "loss": 0.7927, + "step": 7784 + }, + { + "epoch": 0.3446863619760247, + "grad_norm": 1.8066872358322144, + "learning_rate": 1.6569605504139396e-05, + "loss": 0.8828, + "step": 7785 + }, + { + "epoch": 0.34473063768083856, + "grad_norm": 1.49653160572052, + "learning_rate": 1.6568440271667832e-05, + "loss": 0.8565, + "step": 7786 + }, + { + "epoch": 0.34477491338565247, + "grad_norm": 1.3111941814422607, + "learning_rate": 1.656727488231442e-05, + "loss": 0.7519, + "step": 7787 + }, + { + "epoch": 0.3448191890904663, + "grad_norm": 1.2978360652923584, + "learning_rate": 1.6566109336106992e-05, + "loss": 0.8151, + "step": 7788 + }, + { + "epoch": 0.3448634647952802, + "grad_norm": 1.219835877418518, + "learning_rate": 1.6564943633073393e-05, + "loss": 0.793, + "step": 7789 + }, + { + "epoch": 0.3449077405000941, + "grad_norm": 1.3102035522460938, + "learning_rate": 1.656377777324146e-05, + "loss": 0.752, + "step": 7790 + }, + { + "epoch": 0.34495201620490795, + "grad_norm": 1.495429515838623, + "learning_rate": 1.6562611756639038e-05, + "loss": 0.7439, + "step": 7791 + }, + { + "epoch": 0.34499629190972186, + "grad_norm": 1.7242366075515747, + "learning_rate": 1.656144558329398e-05, + "loss": 0.886, + "step": 7792 + }, + { + "epoch": 0.3450405676145357, + "grad_norm": 1.2569668292999268, + "learning_rate": 1.6560279253234134e-05, + "loss": 0.8042, + "step": 7793 + }, + { + "epoch": 0.3450848433193496, + "grad_norm": 1.415568232536316, + "learning_rate": 1.655911276648736e-05, + "loss": 0.7722, + "step": 7794 + }, + { + "epoch": 0.3451291190241635, + "grad_norm": 1.3053661584854126, + "learning_rate": 1.6557946123081516e-05, + "loss": 0.8204, + "step": 7795 + }, + { + "epoch": 0.34517339472897735, + "grad_norm": 1.5029655694961548, + "learning_rate": 1.6556779323044474e-05, + "loss": 0.8585, + "step": 7796 + }, + { + "epoch": 0.3452176704337912, + "grad_norm": 1.3812241554260254, + "learning_rate": 1.6555612366404092e-05, + "loss": 0.8051, + "step": 7797 + }, + { + "epoch": 0.3452619461386051, + "grad_norm": 1.3276530504226685, + "learning_rate": 1.6554445253188253e-05, + "loss": 0.9049, + "step": 7798 + }, + { + "epoch": 0.34530622184341897, + "grad_norm": 1.4478970766067505, + "learning_rate": 1.6553277983424818e-05, + "loss": 0.7953, + "step": 7799 + }, + { + "epoch": 0.3453504975482328, + "grad_norm": 1.3337955474853516, + "learning_rate": 1.6552110557141675e-05, + "loss": 0.803, + "step": 7800 + }, + { + "epoch": 0.34539477325304674, + "grad_norm": 1.3180912733078003, + "learning_rate": 1.655094297436671e-05, + "loss": 0.7938, + "step": 7801 + }, + { + "epoch": 0.3454390489578606, + "grad_norm": 1.8178472518920898, + "learning_rate": 1.6549775235127805e-05, + "loss": 0.8198, + "step": 7802 + }, + { + "epoch": 0.34548332466267445, + "grad_norm": 1.3400263786315918, + "learning_rate": 1.6548607339452853e-05, + "loss": 0.7825, + "step": 7803 + }, + { + "epoch": 0.34552760036748836, + "grad_norm": 1.3216338157653809, + "learning_rate": 1.6547439287369744e-05, + "loss": 0.7967, + "step": 7804 + }, + { + "epoch": 0.3455718760723022, + "grad_norm": 1.4908419847488403, + "learning_rate": 1.654627107890638e-05, + "loss": 0.7979, + "step": 7805 + }, + { + "epoch": 0.3456161517771161, + "grad_norm": 1.1774893999099731, + "learning_rate": 1.6545102714090663e-05, + "loss": 0.7623, + "step": 7806 + }, + { + "epoch": 0.34566042748193, + "grad_norm": 1.3484128713607788, + "learning_rate": 1.6543934192950493e-05, + "loss": 0.7912, + "step": 7807 + }, + { + "epoch": 0.34570470318674384, + "grad_norm": 1.5961015224456787, + "learning_rate": 1.6542765515513787e-05, + "loss": 0.8766, + "step": 7808 + }, + { + "epoch": 0.34574897889155776, + "grad_norm": 1.3014973402023315, + "learning_rate": 1.654159668180845e-05, + "loss": 0.8029, + "step": 7809 + }, + { + "epoch": 0.3457932545963716, + "grad_norm": 1.4592390060424805, + "learning_rate": 1.6540427691862406e-05, + "loss": 0.8212, + "step": 7810 + }, + { + "epoch": 0.34583753030118547, + "grad_norm": 1.2144830226898193, + "learning_rate": 1.6539258545703573e-05, + "loss": 0.8034, + "step": 7811 + }, + { + "epoch": 0.3458818060059994, + "grad_norm": 1.451241135597229, + "learning_rate": 1.653808924335987e-05, + "loss": 0.7869, + "step": 7812 + }, + { + "epoch": 0.34592608171081324, + "grad_norm": 1.258782982826233, + "learning_rate": 1.653691978485923e-05, + "loss": 0.6924, + "step": 7813 + }, + { + "epoch": 0.3459703574156271, + "grad_norm": 1.4194676876068115, + "learning_rate": 1.6535750170229585e-05, + "loss": 0.7675, + "step": 7814 + }, + { + "epoch": 0.346014633120441, + "grad_norm": 1.2747730016708374, + "learning_rate": 1.6534580399498866e-05, + "loss": 0.7406, + "step": 7815 + }, + { + "epoch": 0.34605890882525486, + "grad_norm": 1.571212887763977, + "learning_rate": 1.653341047269502e-05, + "loss": 0.9417, + "step": 7816 + }, + { + "epoch": 0.3461031845300687, + "grad_norm": 1.4651658535003662, + "learning_rate": 1.6532240389845982e-05, + "loss": 0.8451, + "step": 7817 + }, + { + "epoch": 0.34614746023488263, + "grad_norm": 1.6412798166275024, + "learning_rate": 1.65310701509797e-05, + "loss": 0.8311, + "step": 7818 + }, + { + "epoch": 0.3461917359396965, + "grad_norm": 1.3944470882415771, + "learning_rate": 1.6529899756124126e-05, + "loss": 0.8261, + "step": 7819 + }, + { + "epoch": 0.34623601164451034, + "grad_norm": 1.429722547531128, + "learning_rate": 1.6528729205307213e-05, + "loss": 0.7889, + "step": 7820 + }, + { + "epoch": 0.34628028734932426, + "grad_norm": 1.364776611328125, + "learning_rate": 1.6527558498556917e-05, + "loss": 0.7119, + "step": 7821 + }, + { + "epoch": 0.3463245630541381, + "grad_norm": 1.5079345703125, + "learning_rate": 1.6526387635901206e-05, + "loss": 0.9063, + "step": 7822 + }, + { + "epoch": 0.34636883875895197, + "grad_norm": 1.2589858770370483, + "learning_rate": 1.6525216617368037e-05, + "loss": 0.7743, + "step": 7823 + }, + { + "epoch": 0.3464131144637659, + "grad_norm": 1.2694774866104126, + "learning_rate": 1.6524045442985383e-05, + "loss": 0.6933, + "step": 7824 + }, + { + "epoch": 0.34645739016857974, + "grad_norm": 1.381962776184082, + "learning_rate": 1.6522874112781213e-05, + "loss": 0.8729, + "step": 7825 + }, + { + "epoch": 0.34650166587339365, + "grad_norm": 1.2746189832687378, + "learning_rate": 1.6521702626783507e-05, + "loss": 0.7764, + "step": 7826 + }, + { + "epoch": 0.3465459415782075, + "grad_norm": 1.698897361755371, + "learning_rate": 1.6520530985020243e-05, + "loss": 0.8395, + "step": 7827 + }, + { + "epoch": 0.34659021728302136, + "grad_norm": 1.3240861892700195, + "learning_rate": 1.6519359187519414e-05, + "loss": 0.8231, + "step": 7828 + }, + { + "epoch": 0.3466344929878353, + "grad_norm": 1.4825098514556885, + "learning_rate": 1.651818723430899e-05, + "loss": 0.8261, + "step": 7829 + }, + { + "epoch": 0.34667876869264913, + "grad_norm": 1.5279457569122314, + "learning_rate": 1.6517015125416972e-05, + "loss": 0.8136, + "step": 7830 + }, + { + "epoch": 0.346723044397463, + "grad_norm": 1.314659833908081, + "learning_rate": 1.6515842860871355e-05, + "loss": 0.7887, + "step": 7831 + }, + { + "epoch": 0.3467673201022769, + "grad_norm": 1.8760603666305542, + "learning_rate": 1.651467044070014e-05, + "loss": 0.8781, + "step": 7832 + }, + { + "epoch": 0.34681159580709076, + "grad_norm": 1.4693100452423096, + "learning_rate": 1.651349786493132e-05, + "loss": 0.9123, + "step": 7833 + }, + { + "epoch": 0.3468558715119046, + "grad_norm": 1.7166435718536377, + "learning_rate": 1.6512325133592914e-05, + "loss": 0.8907, + "step": 7834 + }, + { + "epoch": 0.3469001472167185, + "grad_norm": 1.3355021476745605, + "learning_rate": 1.651115224671292e-05, + "loss": 0.801, + "step": 7835 + }, + { + "epoch": 0.3469444229215324, + "grad_norm": 1.2176283597946167, + "learning_rate": 1.650997920431936e-05, + "loss": 0.7277, + "step": 7836 + }, + { + "epoch": 0.34698869862634624, + "grad_norm": 1.2408461570739746, + "learning_rate": 1.6508806006440242e-05, + "loss": 0.8021, + "step": 7837 + }, + { + "epoch": 0.34703297433116015, + "grad_norm": 1.6368674039840698, + "learning_rate": 1.6507632653103597e-05, + "loss": 0.9078, + "step": 7838 + }, + { + "epoch": 0.347077250035974, + "grad_norm": 1.185441017150879, + "learning_rate": 1.6506459144337443e-05, + "loss": 0.7734, + "step": 7839 + }, + { + "epoch": 0.34712152574078786, + "grad_norm": 1.5959924459457397, + "learning_rate": 1.650528548016981e-05, + "loss": 0.8682, + "step": 7840 + }, + { + "epoch": 0.3471658014456018, + "grad_norm": 1.2548270225524902, + "learning_rate": 1.650411166062873e-05, + "loss": 0.7071, + "step": 7841 + }, + { + "epoch": 0.34721007715041563, + "grad_norm": 1.2759827375411987, + "learning_rate": 1.6502937685742237e-05, + "loss": 0.7973, + "step": 7842 + }, + { + "epoch": 0.3472543528552295, + "grad_norm": 1.3691350221633911, + "learning_rate": 1.6501763555538373e-05, + "loss": 0.8176, + "step": 7843 + }, + { + "epoch": 0.3472986285600434, + "grad_norm": 1.1511130332946777, + "learning_rate": 1.6500589270045185e-05, + "loss": 0.7094, + "step": 7844 + }, + { + "epoch": 0.34734290426485726, + "grad_norm": 1.142452359199524, + "learning_rate": 1.6499414829290712e-05, + "loss": 0.6699, + "step": 7845 + }, + { + "epoch": 0.34738717996967117, + "grad_norm": 1.4127434492111206, + "learning_rate": 1.6498240233303008e-05, + "loss": 0.8451, + "step": 7846 + }, + { + "epoch": 0.347431455674485, + "grad_norm": 1.293294072151184, + "learning_rate": 1.6497065482110123e-05, + "loss": 0.7929, + "step": 7847 + }, + { + "epoch": 0.3474757313792989, + "grad_norm": 1.356817364692688, + "learning_rate": 1.6495890575740124e-05, + "loss": 0.7824, + "step": 7848 + }, + { + "epoch": 0.3475200070841128, + "grad_norm": 1.4044843912124634, + "learning_rate": 1.6494715514221065e-05, + "loss": 0.8181, + "step": 7849 + }, + { + "epoch": 0.34756428278892665, + "grad_norm": 1.4092987775802612, + "learning_rate": 1.6493540297581018e-05, + "loss": 0.8704, + "step": 7850 + }, + { + "epoch": 0.3476085584937405, + "grad_norm": 1.1594897508621216, + "learning_rate": 1.6492364925848045e-05, + "loss": 0.7313, + "step": 7851 + }, + { + "epoch": 0.3476528341985544, + "grad_norm": 1.1248115301132202, + "learning_rate": 1.6491189399050224e-05, + "loss": 0.7794, + "step": 7852 + }, + { + "epoch": 0.3476971099033683, + "grad_norm": 1.416374921798706, + "learning_rate": 1.6490013717215626e-05, + "loss": 0.7652, + "step": 7853 + }, + { + "epoch": 0.34774138560818213, + "grad_norm": 1.3132200241088867, + "learning_rate": 1.648883788037234e-05, + "loss": 0.753, + "step": 7854 + }, + { + "epoch": 0.34778566131299604, + "grad_norm": 1.6509382724761963, + "learning_rate": 1.6487661888548442e-05, + "loss": 0.7901, + "step": 7855 + }, + { + "epoch": 0.3478299370178099, + "grad_norm": 1.447628378868103, + "learning_rate": 1.6486485741772022e-05, + "loss": 0.8564, + "step": 7856 + }, + { + "epoch": 0.34787421272262375, + "grad_norm": 1.2873458862304688, + "learning_rate": 1.6485309440071168e-05, + "loss": 0.8033, + "step": 7857 + }, + { + "epoch": 0.34791848842743767, + "grad_norm": 1.3532516956329346, + "learning_rate": 1.6484132983473985e-05, + "loss": 0.7436, + "step": 7858 + }, + { + "epoch": 0.3479627641322515, + "grad_norm": 1.154820442199707, + "learning_rate": 1.648295637200856e-05, + "loss": 0.677, + "step": 7859 + }, + { + "epoch": 0.3480070398370654, + "grad_norm": 1.2694441080093384, + "learning_rate": 1.6481779605703003e-05, + "loss": 0.7157, + "step": 7860 + }, + { + "epoch": 0.3480513155418793, + "grad_norm": 2.2017223834991455, + "learning_rate": 1.648060268458542e-05, + "loss": 0.8335, + "step": 7861 + }, + { + "epoch": 0.34809559124669315, + "grad_norm": 3.0592942237854004, + "learning_rate": 1.6479425608683915e-05, + "loss": 0.8512, + "step": 7862 + }, + { + "epoch": 0.34813986695150706, + "grad_norm": 1.5219547748565674, + "learning_rate": 1.6478248378026605e-05, + "loss": 0.8999, + "step": 7863 + }, + { + "epoch": 0.3481841426563209, + "grad_norm": 1.1723636388778687, + "learning_rate": 1.647707099264161e-05, + "loss": 0.74, + "step": 7864 + }, + { + "epoch": 0.3482284183611348, + "grad_norm": 1.4484609365463257, + "learning_rate": 1.647589345255705e-05, + "loss": 0.8405, + "step": 7865 + }, + { + "epoch": 0.3482726940659487, + "grad_norm": 1.3655006885528564, + "learning_rate": 1.6474715757801043e-05, + "loss": 0.7845, + "step": 7866 + }, + { + "epoch": 0.34831696977076254, + "grad_norm": 1.327143669128418, + "learning_rate": 1.6473537908401723e-05, + "loss": 0.7943, + "step": 7867 + }, + { + "epoch": 0.3483612454755764, + "grad_norm": 1.3373734951019287, + "learning_rate": 1.647235990438722e-05, + "loss": 0.7541, + "step": 7868 + }, + { + "epoch": 0.3484055211803903, + "grad_norm": 1.2998151779174805, + "learning_rate": 1.6471181745785673e-05, + "loss": 0.7804, + "step": 7869 + }, + { + "epoch": 0.34844979688520417, + "grad_norm": 1.4302774667739868, + "learning_rate": 1.647000343262522e-05, + "loss": 0.8018, + "step": 7870 + }, + { + "epoch": 0.348494072590018, + "grad_norm": 1.2562506198883057, + "learning_rate": 1.6468824964934e-05, + "loss": 0.773, + "step": 7871 + }, + { + "epoch": 0.34853834829483193, + "grad_norm": 1.1875308752059937, + "learning_rate": 1.6467646342740166e-05, + "loss": 0.8276, + "step": 7872 + }, + { + "epoch": 0.3485826239996458, + "grad_norm": 1.1792690753936768, + "learning_rate": 1.6466467566071863e-05, + "loss": 0.6767, + "step": 7873 + }, + { + "epoch": 0.34862689970445965, + "grad_norm": 1.6487278938293457, + "learning_rate": 1.646528863495725e-05, + "loss": 0.8373, + "step": 7874 + }, + { + "epoch": 0.34867117540927356, + "grad_norm": 1.304760217666626, + "learning_rate": 1.6464109549424477e-05, + "loss": 0.7907, + "step": 7875 + }, + { + "epoch": 0.3487154511140874, + "grad_norm": 6.17778205871582, + "learning_rate": 1.6462930309501716e-05, + "loss": 0.8851, + "step": 7876 + }, + { + "epoch": 0.3487597268189013, + "grad_norm": 1.2085703611373901, + "learning_rate": 1.6461750915217126e-05, + "loss": 0.7168, + "step": 7877 + }, + { + "epoch": 0.3488040025237152, + "grad_norm": 1.2640976905822754, + "learning_rate": 1.6460571366598875e-05, + "loss": 0.7819, + "step": 7878 + }, + { + "epoch": 0.34884827822852904, + "grad_norm": 1.1726758480072021, + "learning_rate": 1.6459391663675135e-05, + "loss": 0.7565, + "step": 7879 + }, + { + "epoch": 0.34889255393334295, + "grad_norm": 1.6472845077514648, + "learning_rate": 1.645821180647409e-05, + "loss": 0.8754, + "step": 7880 + }, + { + "epoch": 0.3489368296381568, + "grad_norm": 1.5696115493774414, + "learning_rate": 1.645703179502391e-05, + "loss": 0.8478, + "step": 7881 + }, + { + "epoch": 0.34898110534297067, + "grad_norm": 1.2804434299468994, + "learning_rate": 1.6455851629352783e-05, + "loss": 0.7777, + "step": 7882 + }, + { + "epoch": 0.3490253810477846, + "grad_norm": 1.9272339344024658, + "learning_rate": 1.6454671309488903e-05, + "loss": 0.9651, + "step": 7883 + }, + { + "epoch": 0.34906965675259843, + "grad_norm": 1.398087978363037, + "learning_rate": 1.6453490835460448e-05, + "loss": 0.8125, + "step": 7884 + }, + { + "epoch": 0.3491139324574123, + "grad_norm": 1.3856810331344604, + "learning_rate": 1.6452310207295622e-05, + "loss": 0.8383, + "step": 7885 + }, + { + "epoch": 0.3491582081622262, + "grad_norm": 1.686789631843567, + "learning_rate": 1.645112942502262e-05, + "loss": 0.8137, + "step": 7886 + }, + { + "epoch": 0.34920248386704006, + "grad_norm": 1.487289309501648, + "learning_rate": 1.644994848866964e-05, + "loss": 0.8374, + "step": 7887 + }, + { + "epoch": 0.3492467595718539, + "grad_norm": 1.487831711769104, + "learning_rate": 1.6448767398264894e-05, + "loss": 0.8671, + "step": 7888 + }, + { + "epoch": 0.3492910352766678, + "grad_norm": 1.5312796831130981, + "learning_rate": 1.6447586153836592e-05, + "loss": 0.803, + "step": 7889 + }, + { + "epoch": 0.3493353109814817, + "grad_norm": 1.4995743036270142, + "learning_rate": 1.6446404755412944e-05, + "loss": 0.8411, + "step": 7890 + }, + { + "epoch": 0.34937958668629554, + "grad_norm": 1.275422215461731, + "learning_rate": 1.644522320302217e-05, + "loss": 0.7428, + "step": 7891 + }, + { + "epoch": 0.34942386239110945, + "grad_norm": 1.4084545373916626, + "learning_rate": 1.644404149669248e-05, + "loss": 0.9012, + "step": 7892 + }, + { + "epoch": 0.3494681380959233, + "grad_norm": 1.2061692476272583, + "learning_rate": 1.6442859636452113e-05, + "loss": 0.7418, + "step": 7893 + }, + { + "epoch": 0.34951241380073717, + "grad_norm": 1.498417615890503, + "learning_rate": 1.6441677622329285e-05, + "loss": 0.8326, + "step": 7894 + }, + { + "epoch": 0.3495566895055511, + "grad_norm": 1.3993568420410156, + "learning_rate": 1.6440495454352233e-05, + "loss": 0.8361, + "step": 7895 + }, + { + "epoch": 0.34960096521036493, + "grad_norm": 1.4732277393341064, + "learning_rate": 1.6439313132549193e-05, + "loss": 0.8068, + "step": 7896 + }, + { + "epoch": 0.34964524091517885, + "grad_norm": 1.248186469078064, + "learning_rate": 1.64381306569484e-05, + "loss": 0.7844, + "step": 7897 + }, + { + "epoch": 0.3496895166199927, + "grad_norm": 1.4845284223556519, + "learning_rate": 1.64369480275781e-05, + "loss": 0.817, + "step": 7898 + }, + { + "epoch": 0.34973379232480656, + "grad_norm": 1.8631538152694702, + "learning_rate": 1.6435765244466536e-05, + "loss": 0.8475, + "step": 7899 + }, + { + "epoch": 0.34977806802962047, + "grad_norm": 1.239668607711792, + "learning_rate": 1.643458230764196e-05, + "loss": 0.7646, + "step": 7900 + }, + { + "epoch": 0.3498223437344343, + "grad_norm": 1.2503348588943481, + "learning_rate": 1.6433399217132625e-05, + "loss": 0.8109, + "step": 7901 + }, + { + "epoch": 0.3498666194392482, + "grad_norm": 1.4843989610671997, + "learning_rate": 1.643221597296679e-05, + "loss": 0.8698, + "step": 7902 + }, + { + "epoch": 0.3499108951440621, + "grad_norm": 1.2677087783813477, + "learning_rate": 1.6431032575172712e-05, + "loss": 0.7979, + "step": 7903 + }, + { + "epoch": 0.34995517084887595, + "grad_norm": 1.1050395965576172, + "learning_rate": 1.6429849023778655e-05, + "loss": 0.7694, + "step": 7904 + }, + { + "epoch": 0.3499994465536898, + "grad_norm": 1.400595784187317, + "learning_rate": 1.6428665318812893e-05, + "loss": 0.8768, + "step": 7905 + }, + { + "epoch": 0.3500437222585037, + "grad_norm": 1.5841585397720337, + "learning_rate": 1.642748146030369e-05, + "loss": 0.8196, + "step": 7906 + }, + { + "epoch": 0.3500879979633176, + "grad_norm": 1.2800110578536987, + "learning_rate": 1.642629744827933e-05, + "loss": 0.7786, + "step": 7907 + }, + { + "epoch": 0.35013227366813143, + "grad_norm": 1.2780296802520752, + "learning_rate": 1.6425113282768087e-05, + "loss": 0.7193, + "step": 7908 + }, + { + "epoch": 0.35017654937294534, + "grad_norm": 1.620883822441101, + "learning_rate": 1.6423928963798242e-05, + "loss": 0.8116, + "step": 7909 + }, + { + "epoch": 0.3502208250777592, + "grad_norm": 1.1348628997802734, + "learning_rate": 1.642274449139809e-05, + "loss": 0.7313, + "step": 7910 + }, + { + "epoch": 0.35026510078257306, + "grad_norm": 1.4660817384719849, + "learning_rate": 1.6421559865595908e-05, + "loss": 0.7762, + "step": 7911 + }, + { + "epoch": 0.35030937648738697, + "grad_norm": 1.2580304145812988, + "learning_rate": 1.642037508642e-05, + "loss": 0.7269, + "step": 7912 + }, + { + "epoch": 0.3503536521922008, + "grad_norm": 1.4279916286468506, + "learning_rate": 1.6419190153898666e-05, + "loss": 0.7339, + "step": 7913 + }, + { + "epoch": 0.35039792789701474, + "grad_norm": 1.4342105388641357, + "learning_rate": 1.6418005068060193e-05, + "loss": 0.8566, + "step": 7914 + }, + { + "epoch": 0.3504422036018286, + "grad_norm": 1.5311472415924072, + "learning_rate": 1.64168198289329e-05, + "loss": 0.8243, + "step": 7915 + }, + { + "epoch": 0.35048647930664245, + "grad_norm": 1.7062389850616455, + "learning_rate": 1.6415634436545093e-05, + "loss": 0.8155, + "step": 7916 + }, + { + "epoch": 0.35053075501145636, + "grad_norm": 1.3578333854675293, + "learning_rate": 1.6414448890925078e-05, + "loss": 0.8786, + "step": 7917 + }, + { + "epoch": 0.3505750307162702, + "grad_norm": 1.3207573890686035, + "learning_rate": 1.6413263192101173e-05, + "loss": 0.7412, + "step": 7918 + }, + { + "epoch": 0.3506193064210841, + "grad_norm": 1.6567434072494507, + "learning_rate": 1.6412077340101698e-05, + "loss": 0.8052, + "step": 7919 + }, + { + "epoch": 0.350663582125898, + "grad_norm": 1.2725210189819336, + "learning_rate": 1.641089133495498e-05, + "loss": 0.7214, + "step": 7920 + }, + { + "epoch": 0.35070785783071184, + "grad_norm": 1.383104920387268, + "learning_rate": 1.6409705176689338e-05, + "loss": 0.7507, + "step": 7921 + }, + { + "epoch": 0.3507521335355257, + "grad_norm": 1.3512295484542847, + "learning_rate": 1.640851886533311e-05, + "loss": 0.7802, + "step": 7922 + }, + { + "epoch": 0.3507964092403396, + "grad_norm": 1.322507381439209, + "learning_rate": 1.6407332400914625e-05, + "loss": 0.7899, + "step": 7923 + }, + { + "epoch": 0.35084068494515347, + "grad_norm": 1.305808663368225, + "learning_rate": 1.6406145783462225e-05, + "loss": 0.7636, + "step": 7924 + }, + { + "epoch": 0.3508849606499673, + "grad_norm": 1.2946444749832153, + "learning_rate": 1.6404959013004247e-05, + "loss": 0.8014, + "step": 7925 + }, + { + "epoch": 0.35092923635478124, + "grad_norm": 1.3417189121246338, + "learning_rate": 1.640377208956904e-05, + "loss": 0.7785, + "step": 7926 + }, + { + "epoch": 0.3509735120595951, + "grad_norm": 1.5918939113616943, + "learning_rate": 1.640258501318495e-05, + "loss": 0.8311, + "step": 7927 + }, + { + "epoch": 0.35101778776440895, + "grad_norm": 1.216675043106079, + "learning_rate": 1.640139778388033e-05, + "loss": 0.7346, + "step": 7928 + }, + { + "epoch": 0.35106206346922286, + "grad_norm": 1.4207748174667358, + "learning_rate": 1.6400210401683535e-05, + "loss": 0.8703, + "step": 7929 + }, + { + "epoch": 0.3511063391740367, + "grad_norm": 1.359339952468872, + "learning_rate": 1.639902286662293e-05, + "loss": 0.7735, + "step": 7930 + }, + { + "epoch": 0.3511506148788506, + "grad_norm": 1.5784361362457275, + "learning_rate": 1.6397835178726866e-05, + "loss": 0.819, + "step": 7931 + }, + { + "epoch": 0.3511948905836645, + "grad_norm": 2.0560622215270996, + "learning_rate": 1.6396647338023724e-05, + "loss": 0.8646, + "step": 7932 + }, + { + "epoch": 0.35123916628847834, + "grad_norm": 1.032544732093811, + "learning_rate": 1.639545934454187e-05, + "loss": 0.677, + "step": 7933 + }, + { + "epoch": 0.35128344199329226, + "grad_norm": 1.1959617137908936, + "learning_rate": 1.639427119830967e-05, + "loss": 0.6954, + "step": 7934 + }, + { + "epoch": 0.3513277176981061, + "grad_norm": 1.3286598920822144, + "learning_rate": 1.6393082899355516e-05, + "loss": 0.7438, + "step": 7935 + }, + { + "epoch": 0.35137199340291997, + "grad_norm": 1.423883318901062, + "learning_rate": 1.639189444770778e-05, + "loss": 0.8954, + "step": 7936 + }, + { + "epoch": 0.3514162691077339, + "grad_norm": 1.154246211051941, + "learning_rate": 1.639070584339485e-05, + "loss": 0.745, + "step": 7937 + }, + { + "epoch": 0.35146054481254774, + "grad_norm": 1.3321608304977417, + "learning_rate": 1.6389517086445112e-05, + "loss": 0.7561, + "step": 7938 + }, + { + "epoch": 0.3515048205173616, + "grad_norm": 1.3928954601287842, + "learning_rate": 1.638832817688696e-05, + "loss": 0.8061, + "step": 7939 + }, + { + "epoch": 0.3515490962221755, + "grad_norm": 1.7207728624343872, + "learning_rate": 1.6387139114748798e-05, + "loss": 0.8358, + "step": 7940 + }, + { + "epoch": 0.35159337192698936, + "grad_norm": 1.4111961126327515, + "learning_rate": 1.6385949900059015e-05, + "loss": 0.7911, + "step": 7941 + }, + { + "epoch": 0.3516376476318032, + "grad_norm": 1.5751533508300781, + "learning_rate": 1.6384760532846015e-05, + "loss": 0.7742, + "step": 7942 + }, + { + "epoch": 0.35168192333661713, + "grad_norm": 1.4949922561645508, + "learning_rate": 1.6383571013138214e-05, + "loss": 0.7809, + "step": 7943 + }, + { + "epoch": 0.351726199041431, + "grad_norm": 1.4444891214370728, + "learning_rate": 1.6382381340964015e-05, + "loss": 0.7274, + "step": 7944 + }, + { + "epoch": 0.35177047474624484, + "grad_norm": 1.5396959781646729, + "learning_rate": 1.6381191516351834e-05, + "loss": 0.8061, + "step": 7945 + }, + { + "epoch": 0.35181475045105876, + "grad_norm": 1.3347769975662231, + "learning_rate": 1.6380001539330088e-05, + "loss": 0.7475, + "step": 7946 + }, + { + "epoch": 0.3518590261558726, + "grad_norm": 1.4210381507873535, + "learning_rate": 1.63788114099272e-05, + "loss": 0.8708, + "step": 7947 + }, + { + "epoch": 0.35190330186068647, + "grad_norm": 2.194369077682495, + "learning_rate": 1.6377621128171597e-05, + "loss": 0.8297, + "step": 7948 + }, + { + "epoch": 0.3519475775655004, + "grad_norm": 1.2074575424194336, + "learning_rate": 1.6376430694091706e-05, + "loss": 0.7097, + "step": 7949 + }, + { + "epoch": 0.35199185327031424, + "grad_norm": 1.2196102142333984, + "learning_rate": 1.637524010771596e-05, + "loss": 0.784, + "step": 7950 + }, + { + "epoch": 0.35203612897512815, + "grad_norm": 1.280902624130249, + "learning_rate": 1.6374049369072794e-05, + "loss": 0.6684, + "step": 7951 + }, + { + "epoch": 0.352080404679942, + "grad_norm": 1.459488868713379, + "learning_rate": 1.6372858478190647e-05, + "loss": 0.8109, + "step": 7952 + }, + { + "epoch": 0.35212468038475586, + "grad_norm": 1.5782136917114258, + "learning_rate": 1.637166743509797e-05, + "loss": 0.8615, + "step": 7953 + }, + { + "epoch": 0.3521689560895698, + "grad_norm": 1.506011962890625, + "learning_rate": 1.63704762398232e-05, + "loss": 0.9232, + "step": 7954 + }, + { + "epoch": 0.35221323179438363, + "grad_norm": 1.4043432474136353, + "learning_rate": 1.6369284892394793e-05, + "loss": 0.8924, + "step": 7955 + }, + { + "epoch": 0.3522575074991975, + "grad_norm": 1.6156901121139526, + "learning_rate": 1.63680933928412e-05, + "loss": 0.7729, + "step": 7956 + }, + { + "epoch": 0.3523017832040114, + "grad_norm": 1.505988597869873, + "learning_rate": 1.6366901741190885e-05, + "loss": 0.7872, + "step": 7957 + }, + { + "epoch": 0.35234605890882525, + "grad_norm": 1.7989951372146606, + "learning_rate": 1.63657099374723e-05, + "loss": 0.9489, + "step": 7958 + }, + { + "epoch": 0.3523903346136391, + "grad_norm": 1.8556029796600342, + "learning_rate": 1.6364517981713922e-05, + "loss": 0.9118, + "step": 7959 + }, + { + "epoch": 0.352434610318453, + "grad_norm": 1.8683180809020996, + "learning_rate": 1.6363325873944215e-05, + "loss": 0.8957, + "step": 7960 + }, + { + "epoch": 0.3524788860232669, + "grad_norm": 1.2974073886871338, + "learning_rate": 1.6362133614191648e-05, + "loss": 0.8064, + "step": 7961 + }, + { + "epoch": 0.35252316172808074, + "grad_norm": 1.312600016593933, + "learning_rate": 1.6360941202484703e-05, + "loss": 0.8043, + "step": 7962 + }, + { + "epoch": 0.35256743743289465, + "grad_norm": 1.2253625392913818, + "learning_rate": 1.6359748638851852e-05, + "loss": 0.7728, + "step": 7963 + }, + { + "epoch": 0.3526117131377085, + "grad_norm": 1.5400182008743286, + "learning_rate": 1.6358555923321586e-05, + "loss": 0.8082, + "step": 7964 + }, + { + "epoch": 0.35265598884252236, + "grad_norm": 1.487064242362976, + "learning_rate": 1.6357363055922388e-05, + "loss": 0.852, + "step": 7965 + }, + { + "epoch": 0.3527002645473363, + "grad_norm": 1.2978235483169556, + "learning_rate": 1.635617003668275e-05, + "loss": 0.7212, + "step": 7966 + }, + { + "epoch": 0.35274454025215013, + "grad_norm": 2.5576791763305664, + "learning_rate": 1.6354976865631168e-05, + "loss": 0.8538, + "step": 7967 + }, + { + "epoch": 0.35278881595696404, + "grad_norm": 1.5380021333694458, + "learning_rate": 1.6353783542796137e-05, + "loss": 0.8165, + "step": 7968 + }, + { + "epoch": 0.3528330916617779, + "grad_norm": 1.3270601034164429, + "learning_rate": 1.635259006820616e-05, + "loss": 0.7836, + "step": 7969 + }, + { + "epoch": 0.35287736736659175, + "grad_norm": 1.3848474025726318, + "learning_rate": 1.6351396441889736e-05, + "loss": 0.7977, + "step": 7970 + }, + { + "epoch": 0.35292164307140567, + "grad_norm": 1.3057535886764526, + "learning_rate": 1.6350202663875385e-05, + "loss": 0.8276, + "step": 7971 + }, + { + "epoch": 0.3529659187762195, + "grad_norm": 1.288891077041626, + "learning_rate": 1.6349008734191615e-05, + "loss": 0.7496, + "step": 7972 + }, + { + "epoch": 0.3530101944810334, + "grad_norm": 1.5478456020355225, + "learning_rate": 1.6347814652866935e-05, + "loss": 0.8585, + "step": 7973 + }, + { + "epoch": 0.3530544701858473, + "grad_norm": 1.3147170543670654, + "learning_rate": 1.6346620419929874e-05, + "loss": 0.7264, + "step": 7974 + }, + { + "epoch": 0.35309874589066115, + "grad_norm": 1.6905359029769897, + "learning_rate": 1.6345426035408954e-05, + "loss": 0.8996, + "step": 7975 + }, + { + "epoch": 0.353143021595475, + "grad_norm": 1.334947109222412, + "learning_rate": 1.63442314993327e-05, + "loss": 0.8359, + "step": 7976 + }, + { + "epoch": 0.3531872973002889, + "grad_norm": 1.3981199264526367, + "learning_rate": 1.634303681172964e-05, + "loss": 0.8015, + "step": 7977 + }, + { + "epoch": 0.3532315730051028, + "grad_norm": 1.3199427127838135, + "learning_rate": 1.6341841972628308e-05, + "loss": 0.7431, + "step": 7978 + }, + { + "epoch": 0.35327584870991663, + "grad_norm": 1.7362853288650513, + "learning_rate": 1.634064698205725e-05, + "loss": 0.7988, + "step": 7979 + }, + { + "epoch": 0.35332012441473054, + "grad_norm": 1.4047669172286987, + "learning_rate": 1.6339451840044994e-05, + "loss": 0.7849, + "step": 7980 + }, + { + "epoch": 0.3533644001195444, + "grad_norm": 1.649375081062317, + "learning_rate": 1.63382565466201e-05, + "loss": 0.8785, + "step": 7981 + }, + { + "epoch": 0.35340867582435825, + "grad_norm": 1.2651022672653198, + "learning_rate": 1.6337061101811107e-05, + "loss": 0.8053, + "step": 7982 + }, + { + "epoch": 0.35345295152917217, + "grad_norm": 1.358751893043518, + "learning_rate": 1.633586550564657e-05, + "loss": 0.8128, + "step": 7983 + }, + { + "epoch": 0.353497227233986, + "grad_norm": 1.4007718563079834, + "learning_rate": 1.633466975815504e-05, + "loss": 0.8044, + "step": 7984 + }, + { + "epoch": 0.35354150293879993, + "grad_norm": 1.1600713729858398, + "learning_rate": 1.6333473859365084e-05, + "loss": 0.7393, + "step": 7985 + }, + { + "epoch": 0.3535857786436138, + "grad_norm": 1.2179274559020996, + "learning_rate": 1.6332277809305263e-05, + "loss": 0.7736, + "step": 7986 + }, + { + "epoch": 0.35363005434842765, + "grad_norm": 1.3347338438034058, + "learning_rate": 1.633108160800414e-05, + "loss": 0.7478, + "step": 7987 + }, + { + "epoch": 0.35367433005324156, + "grad_norm": 1.4288980960845947, + "learning_rate": 1.632988525549029e-05, + "loss": 0.7987, + "step": 7988 + }, + { + "epoch": 0.3537186057580554, + "grad_norm": 1.6513638496398926, + "learning_rate": 1.632868875179228e-05, + "loss": 0.8045, + "step": 7989 + }, + { + "epoch": 0.35376288146286927, + "grad_norm": 1.2073994874954224, + "learning_rate": 1.6327492096938696e-05, + "loss": 0.6991, + "step": 7990 + }, + { + "epoch": 0.3538071571676832, + "grad_norm": 1.383103847503662, + "learning_rate": 1.6326295290958114e-05, + "loss": 0.7919, + "step": 7991 + }, + { + "epoch": 0.35385143287249704, + "grad_norm": 1.3898340463638306, + "learning_rate": 1.6325098333879122e-05, + "loss": 0.8177, + "step": 7992 + }, + { + "epoch": 0.3538957085773109, + "grad_norm": 1.1619924306869507, + "learning_rate": 1.6323901225730307e-05, + "loss": 0.661, + "step": 7993 + }, + { + "epoch": 0.3539399842821248, + "grad_norm": 1.179352045059204, + "learning_rate": 1.632270396654026e-05, + "loss": 0.7137, + "step": 7994 + }, + { + "epoch": 0.35398425998693867, + "grad_norm": 1.4235682487487793, + "learning_rate": 1.6321506556337575e-05, + "loss": 0.8735, + "step": 7995 + }, + { + "epoch": 0.3540285356917525, + "grad_norm": 1.289264440536499, + "learning_rate": 1.632030899515085e-05, + "loss": 0.7748, + "step": 7996 + }, + { + "epoch": 0.35407281139656643, + "grad_norm": 1.8694519996643066, + "learning_rate": 1.6319111283008698e-05, + "loss": 0.8793, + "step": 7997 + }, + { + "epoch": 0.3541170871013803, + "grad_norm": 1.115776777267456, + "learning_rate": 1.631791341993971e-05, + "loss": 0.7624, + "step": 7998 + }, + { + "epoch": 0.35416136280619415, + "grad_norm": 1.454532265663147, + "learning_rate": 1.631671540597251e-05, + "loss": 0.8659, + "step": 7999 + }, + { + "epoch": 0.35420563851100806, + "grad_norm": 1.4836877584457397, + "learning_rate": 1.6315517241135704e-05, + "loss": 0.8022, + "step": 8000 + }, + { + "epoch": 0.3542499142158219, + "grad_norm": 1.3743783235549927, + "learning_rate": 1.631431892545791e-05, + "loss": 0.7442, + "step": 8001 + }, + { + "epoch": 0.35429418992063577, + "grad_norm": 1.2244354486465454, + "learning_rate": 1.6313120458967756e-05, + "loss": 0.7483, + "step": 8002 + }, + { + "epoch": 0.3543384656254497, + "grad_norm": 1.5302643775939941, + "learning_rate": 1.631192184169385e-05, + "loss": 0.8805, + "step": 8003 + }, + { + "epoch": 0.35438274133026354, + "grad_norm": 2.175809144973755, + "learning_rate": 1.6310723073664842e-05, + "loss": 0.9258, + "step": 8004 + }, + { + "epoch": 0.35442701703507745, + "grad_norm": 1.4114333391189575, + "learning_rate": 1.6309524154909345e-05, + "loss": 0.7367, + "step": 8005 + }, + { + "epoch": 0.3544712927398913, + "grad_norm": 1.1522080898284912, + "learning_rate": 1.6308325085455998e-05, + "loss": 0.7079, + "step": 8006 + }, + { + "epoch": 0.35451556844470516, + "grad_norm": 1.269271731376648, + "learning_rate": 1.6307125865333447e-05, + "loss": 0.7902, + "step": 8007 + }, + { + "epoch": 0.3545598441495191, + "grad_norm": 1.7226719856262207, + "learning_rate": 1.6305926494570327e-05, + "loss": 0.9461, + "step": 8008 + }, + { + "epoch": 0.35460411985433293, + "grad_norm": 1.2334890365600586, + "learning_rate": 1.630472697319529e-05, + "loss": 0.7812, + "step": 8009 + }, + { + "epoch": 0.3546483955591468, + "grad_norm": 1.3928526639938354, + "learning_rate": 1.6303527301236983e-05, + "loss": 0.7969, + "step": 8010 + }, + { + "epoch": 0.3546926712639607, + "grad_norm": 1.3444491624832153, + "learning_rate": 1.6302327478724057e-05, + "loss": 0.749, + "step": 8011 + }, + { + "epoch": 0.35473694696877456, + "grad_norm": 1.8696399927139282, + "learning_rate": 1.6301127505685174e-05, + "loss": 0.7883, + "step": 8012 + }, + { + "epoch": 0.3547812226735884, + "grad_norm": 1.2132292985916138, + "learning_rate": 1.6299927382148987e-05, + "loss": 0.7495, + "step": 8013 + }, + { + "epoch": 0.3548254983784023, + "grad_norm": 1.4114296436309814, + "learning_rate": 1.6298727108144167e-05, + "loss": 0.8139, + "step": 8014 + }, + { + "epoch": 0.3548697740832162, + "grad_norm": 1.7492780685424805, + "learning_rate": 1.629752668369938e-05, + "loss": 0.83, + "step": 8015 + }, + { + "epoch": 0.35491404978803004, + "grad_norm": 1.2406059503555298, + "learning_rate": 1.629632610884329e-05, + "loss": 0.7317, + "step": 8016 + }, + { + "epoch": 0.35495832549284395, + "grad_norm": 1.239777684211731, + "learning_rate": 1.629512538360458e-05, + "loss": 0.7366, + "step": 8017 + }, + { + "epoch": 0.3550026011976578, + "grad_norm": 1.260096549987793, + "learning_rate": 1.6293924508011933e-05, + "loss": 0.711, + "step": 8018 + }, + { + "epoch": 0.35504687690247166, + "grad_norm": 1.5551785230636597, + "learning_rate": 1.6292723482094013e-05, + "loss": 0.8675, + "step": 8019 + }, + { + "epoch": 0.3550911526072856, + "grad_norm": 1.3840670585632324, + "learning_rate": 1.6291522305879523e-05, + "loss": 0.7833, + "step": 8020 + }, + { + "epoch": 0.35513542831209943, + "grad_norm": 1.7006906270980835, + "learning_rate": 1.6290320979397143e-05, + "loss": 0.9111, + "step": 8021 + }, + { + "epoch": 0.35517970401691334, + "grad_norm": 1.249631404876709, + "learning_rate": 1.6289119502675567e-05, + "loss": 0.7548, + "step": 8022 + }, + { + "epoch": 0.3552239797217272, + "grad_norm": 1.3226467370986938, + "learning_rate": 1.6287917875743495e-05, + "loss": 0.7572, + "step": 8023 + }, + { + "epoch": 0.35526825542654106, + "grad_norm": 1.2150039672851562, + "learning_rate": 1.6286716098629622e-05, + "loss": 0.7588, + "step": 8024 + }, + { + "epoch": 0.35531253113135497, + "grad_norm": 1.3687206506729126, + "learning_rate": 1.628551417136266e-05, + "loss": 0.7528, + "step": 8025 + }, + { + "epoch": 0.3553568068361688, + "grad_norm": 1.2992719411849976, + "learning_rate": 1.62843120939713e-05, + "loss": 0.8067, + "step": 8026 + }, + { + "epoch": 0.3554010825409827, + "grad_norm": 1.3106892108917236, + "learning_rate": 1.628310986648427e-05, + "loss": 0.7596, + "step": 8027 + }, + { + "epoch": 0.3554453582457966, + "grad_norm": 1.347802758216858, + "learning_rate": 1.6281907488930277e-05, + "loss": 0.7837, + "step": 8028 + }, + { + "epoch": 0.35548963395061045, + "grad_norm": 1.2351744174957275, + "learning_rate": 1.6280704961338033e-05, + "loss": 0.7456, + "step": 8029 + }, + { + "epoch": 0.3555339096554243, + "grad_norm": 1.2482551336288452, + "learning_rate": 1.6279502283736272e-05, + "loss": 0.7253, + "step": 8030 + }, + { + "epoch": 0.3555781853602382, + "grad_norm": 1.4903032779693604, + "learning_rate": 1.6278299456153704e-05, + "loss": 0.7545, + "step": 8031 + }, + { + "epoch": 0.3556224610650521, + "grad_norm": 1.5743294954299927, + "learning_rate": 1.6277096478619073e-05, + "loss": 0.8713, + "step": 8032 + }, + { + "epoch": 0.35566673676986593, + "grad_norm": 1.4390536546707153, + "learning_rate": 1.6275893351161095e-05, + "loss": 0.8142, + "step": 8033 + }, + { + "epoch": 0.35571101247467984, + "grad_norm": 1.0595612525939941, + "learning_rate": 1.627469007380852e-05, + "loss": 0.6811, + "step": 8034 + }, + { + "epoch": 0.3557552881794937, + "grad_norm": 1.459186315536499, + "learning_rate": 1.6273486646590082e-05, + "loss": 0.7688, + "step": 8035 + }, + { + "epoch": 0.35579956388430756, + "grad_norm": 1.5565544366836548, + "learning_rate": 1.6272283069534524e-05, + "loss": 0.8409, + "step": 8036 + }, + { + "epoch": 0.35584383958912147, + "grad_norm": 1.5456488132476807, + "learning_rate": 1.627107934267059e-05, + "loss": 0.8023, + "step": 8037 + }, + { + "epoch": 0.3558881152939353, + "grad_norm": 1.2253851890563965, + "learning_rate": 1.626987546602703e-05, + "loss": 0.7132, + "step": 8038 + }, + { + "epoch": 0.35593239099874924, + "grad_norm": 1.4320714473724365, + "learning_rate": 1.6268671439632603e-05, + "loss": 0.7941, + "step": 8039 + }, + { + "epoch": 0.3559766667035631, + "grad_norm": 1.2874741554260254, + "learning_rate": 1.6267467263516062e-05, + "loss": 0.7452, + "step": 8040 + }, + { + "epoch": 0.35602094240837695, + "grad_norm": 1.2981704473495483, + "learning_rate": 1.6266262937706167e-05, + "loss": 0.7719, + "step": 8041 + }, + { + "epoch": 0.35606521811319086, + "grad_norm": 1.415604591369629, + "learning_rate": 1.626505846223169e-05, + "loss": 0.8261, + "step": 8042 + }, + { + "epoch": 0.3561094938180047, + "grad_norm": 1.3363844156265259, + "learning_rate": 1.6263853837121384e-05, + "loss": 0.7983, + "step": 8043 + }, + { + "epoch": 0.3561537695228186, + "grad_norm": 1.3574875593185425, + "learning_rate": 1.626264906240403e-05, + "loss": 0.7988, + "step": 8044 + }, + { + "epoch": 0.3561980452276325, + "grad_norm": 1.2887520790100098, + "learning_rate": 1.6261444138108407e-05, + "loss": 0.8294, + "step": 8045 + }, + { + "epoch": 0.35624232093244634, + "grad_norm": 1.3422966003417969, + "learning_rate": 1.6260239064263293e-05, + "loss": 0.7698, + "step": 8046 + }, + { + "epoch": 0.3562865966372602, + "grad_norm": 1.3712433576583862, + "learning_rate": 1.625903384089746e-05, + "loss": 0.8336, + "step": 8047 + }, + { + "epoch": 0.3563308723420741, + "grad_norm": 1.2354923486709595, + "learning_rate": 1.62578284680397e-05, + "loss": 0.7574, + "step": 8048 + }, + { + "epoch": 0.35637514804688797, + "grad_norm": 1.5291774272918701, + "learning_rate": 1.6256622945718804e-05, + "loss": 0.8711, + "step": 8049 + }, + { + "epoch": 0.3564194237517018, + "grad_norm": 1.537562370300293, + "learning_rate": 1.6255417273963564e-05, + "loss": 0.8607, + "step": 8050 + }, + { + "epoch": 0.35646369945651574, + "grad_norm": 1.4636999368667603, + "learning_rate": 1.6254211452802774e-05, + "loss": 0.846, + "step": 8051 + }, + { + "epoch": 0.3565079751613296, + "grad_norm": 1.3585398197174072, + "learning_rate": 1.6253005482265238e-05, + "loss": 0.7853, + "step": 8052 + }, + { + "epoch": 0.35655225086614345, + "grad_norm": 1.345672607421875, + "learning_rate": 1.6251799362379753e-05, + "loss": 0.7723, + "step": 8053 + }, + { + "epoch": 0.35659652657095736, + "grad_norm": 1.423431396484375, + "learning_rate": 1.6250593093175133e-05, + "loss": 0.8399, + "step": 8054 + }, + { + "epoch": 0.3566408022757712, + "grad_norm": 1.2889976501464844, + "learning_rate": 1.6249386674680186e-05, + "loss": 0.7733, + "step": 8055 + }, + { + "epoch": 0.35668507798058513, + "grad_norm": 1.3596192598342896, + "learning_rate": 1.6248180106923727e-05, + "loss": 0.7814, + "step": 8056 + }, + { + "epoch": 0.356729353685399, + "grad_norm": 1.6053828001022339, + "learning_rate": 1.6246973389934573e-05, + "loss": 0.7991, + "step": 8057 + }, + { + "epoch": 0.35677362939021284, + "grad_norm": 2.176833391189575, + "learning_rate": 1.6245766523741552e-05, + "loss": 0.9373, + "step": 8058 + }, + { + "epoch": 0.35681790509502676, + "grad_norm": 1.609686255455017, + "learning_rate": 1.6244559508373476e-05, + "loss": 0.7847, + "step": 8059 + }, + { + "epoch": 0.3568621807998406, + "grad_norm": 1.3202471733093262, + "learning_rate": 1.624335234385918e-05, + "loss": 0.7935, + "step": 8060 + }, + { + "epoch": 0.35690645650465447, + "grad_norm": 1.5112398862838745, + "learning_rate": 1.62421450302275e-05, + "loss": 0.812, + "step": 8061 + }, + { + "epoch": 0.3569507322094684, + "grad_norm": 1.3885953426361084, + "learning_rate": 1.624093756750727e-05, + "loss": 0.8186, + "step": 8062 + }, + { + "epoch": 0.35699500791428224, + "grad_norm": 1.239903211593628, + "learning_rate": 1.6239729955727324e-05, + "loss": 0.7786, + "step": 8063 + }, + { + "epoch": 0.3570392836190961, + "grad_norm": 1.418108344078064, + "learning_rate": 1.6238522194916512e-05, + "loss": 0.8044, + "step": 8064 + }, + { + "epoch": 0.35708355932391, + "grad_norm": 1.4426801204681396, + "learning_rate": 1.6237314285103676e-05, + "loss": 0.8128, + "step": 8065 + }, + { + "epoch": 0.35712783502872386, + "grad_norm": 1.367108941078186, + "learning_rate": 1.6236106226317664e-05, + "loss": 0.8315, + "step": 8066 + }, + { + "epoch": 0.3571721107335377, + "grad_norm": 1.3789616823196411, + "learning_rate": 1.6234898018587336e-05, + "loss": 0.7291, + "step": 8067 + }, + { + "epoch": 0.35721638643835163, + "grad_norm": 1.421579122543335, + "learning_rate": 1.6233689661941544e-05, + "loss": 0.8184, + "step": 8068 + }, + { + "epoch": 0.3572606621431655, + "grad_norm": 1.7073066234588623, + "learning_rate": 1.6232481156409152e-05, + "loss": 0.8625, + "step": 8069 + }, + { + "epoch": 0.35730493784797934, + "grad_norm": 1.5103750228881836, + "learning_rate": 1.623127250201902e-05, + "loss": 0.8306, + "step": 8070 + }, + { + "epoch": 0.35734921355279325, + "grad_norm": 1.1831074953079224, + "learning_rate": 1.6230063698800018e-05, + "loss": 0.7605, + "step": 8071 + }, + { + "epoch": 0.3573934892576071, + "grad_norm": 1.4695440530776978, + "learning_rate": 1.6228854746781018e-05, + "loss": 0.7803, + "step": 8072 + }, + { + "epoch": 0.357437764962421, + "grad_norm": 1.7091248035430908, + "learning_rate": 1.6227645645990894e-05, + "loss": 0.7851, + "step": 8073 + }, + { + "epoch": 0.3574820406672349, + "grad_norm": 1.3298730850219727, + "learning_rate": 1.6226436396458524e-05, + "loss": 0.8091, + "step": 8074 + }, + { + "epoch": 0.35752631637204874, + "grad_norm": 1.3185148239135742, + "learning_rate": 1.6225226998212793e-05, + "loss": 0.7949, + "step": 8075 + }, + { + "epoch": 0.35757059207686265, + "grad_norm": 1.0868055820465088, + "learning_rate": 1.622401745128258e-05, + "loss": 0.7731, + "step": 8076 + }, + { + "epoch": 0.3576148677816765, + "grad_norm": 1.4609490633010864, + "learning_rate": 1.622280775569678e-05, + "loss": 0.7797, + "step": 8077 + }, + { + "epoch": 0.35765914348649036, + "grad_norm": 1.0799692869186401, + "learning_rate": 1.6221597911484283e-05, + "loss": 0.6695, + "step": 8078 + }, + { + "epoch": 0.3577034191913043, + "grad_norm": 1.551161289215088, + "learning_rate": 1.6220387918673987e-05, + "loss": 0.8175, + "step": 8079 + }, + { + "epoch": 0.35774769489611813, + "grad_norm": 1.721927285194397, + "learning_rate": 1.621917777729479e-05, + "loss": 0.9453, + "step": 8080 + }, + { + "epoch": 0.357791970600932, + "grad_norm": 1.1939970254898071, + "learning_rate": 1.6217967487375595e-05, + "loss": 0.7609, + "step": 8081 + }, + { + "epoch": 0.3578362463057459, + "grad_norm": 1.245951533317566, + "learning_rate": 1.6216757048945307e-05, + "loss": 0.7826, + "step": 8082 + }, + { + "epoch": 0.35788052201055975, + "grad_norm": 1.7116056680679321, + "learning_rate": 1.621554646203284e-05, + "loss": 0.797, + "step": 8083 + }, + { + "epoch": 0.3579247977153736, + "grad_norm": 1.5515925884246826, + "learning_rate": 1.6214335726667113e-05, + "loss": 0.8039, + "step": 8084 + }, + { + "epoch": 0.3579690734201875, + "grad_norm": 1.442363977432251, + "learning_rate": 1.621312484287703e-05, + "loss": 0.7438, + "step": 8085 + }, + { + "epoch": 0.3580133491250014, + "grad_norm": 1.2912806272506714, + "learning_rate": 1.621191381069152e-05, + "loss": 0.7705, + "step": 8086 + }, + { + "epoch": 0.35805762482981524, + "grad_norm": 1.5530662536621094, + "learning_rate": 1.6210702630139508e-05, + "loss": 0.7384, + "step": 8087 + }, + { + "epoch": 0.35810190053462915, + "grad_norm": 1.817305564880371, + "learning_rate": 1.6209491301249917e-05, + "loss": 0.7434, + "step": 8088 + }, + { + "epoch": 0.358146176239443, + "grad_norm": 1.2740733623504639, + "learning_rate": 1.6208279824051688e-05, + "loss": 0.7856, + "step": 8089 + }, + { + "epoch": 0.35819045194425686, + "grad_norm": 1.27066969871521, + "learning_rate": 1.6207068198573747e-05, + "loss": 0.7527, + "step": 8090 + }, + { + "epoch": 0.35823472764907077, + "grad_norm": 1.4569323062896729, + "learning_rate": 1.6205856424845038e-05, + "loss": 0.8732, + "step": 8091 + }, + { + "epoch": 0.35827900335388463, + "grad_norm": 1.1560674905776978, + "learning_rate": 1.62046445028945e-05, + "loss": 0.7427, + "step": 8092 + }, + { + "epoch": 0.35832327905869854, + "grad_norm": 1.7017707824707031, + "learning_rate": 1.620343243275108e-05, + "loss": 0.9213, + "step": 8093 + }, + { + "epoch": 0.3583675547635124, + "grad_norm": 1.2750312089920044, + "learning_rate": 1.6202220214443728e-05, + "loss": 0.7505, + "step": 8094 + }, + { + "epoch": 0.35841183046832625, + "grad_norm": 1.541149616241455, + "learning_rate": 1.6201007848001393e-05, + "loss": 0.7866, + "step": 8095 + }, + { + "epoch": 0.35845610617314017, + "grad_norm": 1.6074984073638916, + "learning_rate": 1.6199795333453036e-05, + "loss": 0.8606, + "step": 8096 + }, + { + "epoch": 0.358500381877954, + "grad_norm": 1.2156727313995361, + "learning_rate": 1.619858267082762e-05, + "loss": 0.7472, + "step": 8097 + }, + { + "epoch": 0.3585446575827679, + "grad_norm": 1.5062017440795898, + "learning_rate": 1.61973698601541e-05, + "loss": 0.7689, + "step": 8098 + }, + { + "epoch": 0.3585889332875818, + "grad_norm": 1.3550207614898682, + "learning_rate": 1.6196156901461447e-05, + "loss": 0.8481, + "step": 8099 + }, + { + "epoch": 0.35863320899239565, + "grad_norm": 1.2890058755874634, + "learning_rate": 1.619494379477863e-05, + "loss": 0.7926, + "step": 8100 + }, + { + "epoch": 0.3586774846972095, + "grad_norm": 1.3741525411605835, + "learning_rate": 1.6193730540134627e-05, + "loss": 0.7879, + "step": 8101 + }, + { + "epoch": 0.3587217604020234, + "grad_norm": 1.356177568435669, + "learning_rate": 1.6192517137558416e-05, + "loss": 0.8007, + "step": 8102 + }, + { + "epoch": 0.35876603610683727, + "grad_norm": 1.1556452512741089, + "learning_rate": 1.619130358707897e-05, + "loss": 0.7245, + "step": 8103 + }, + { + "epoch": 0.35881031181165113, + "grad_norm": 1.2347935438156128, + "learning_rate": 1.6190089888725284e-05, + "loss": 0.7783, + "step": 8104 + }, + { + "epoch": 0.35885458751646504, + "grad_norm": 1.378588080406189, + "learning_rate": 1.6188876042526337e-05, + "loss": 0.8057, + "step": 8105 + }, + { + "epoch": 0.3588988632212789, + "grad_norm": 1.2523136138916016, + "learning_rate": 1.6187662048511126e-05, + "loss": 0.7936, + "step": 8106 + }, + { + "epoch": 0.35894313892609275, + "grad_norm": 1.3450359106063843, + "learning_rate": 1.6186447906708648e-05, + "loss": 0.7346, + "step": 8107 + }, + { + "epoch": 0.35898741463090666, + "grad_norm": 1.3720269203186035, + "learning_rate": 1.6185233617147896e-05, + "loss": 0.8357, + "step": 8108 + }, + { + "epoch": 0.3590316903357205, + "grad_norm": 1.6323248147964478, + "learning_rate": 1.6184019179857876e-05, + "loss": 0.8599, + "step": 8109 + }, + { + "epoch": 0.35907596604053443, + "grad_norm": 1.2123826742172241, + "learning_rate": 1.6182804594867588e-05, + "loss": 0.7369, + "step": 8110 + }, + { + "epoch": 0.3591202417453483, + "grad_norm": 1.312221884727478, + "learning_rate": 1.6181589862206053e-05, + "loss": 0.7719, + "step": 8111 + }, + { + "epoch": 0.35916451745016215, + "grad_norm": 1.3069208860397339, + "learning_rate": 1.6180374981902277e-05, + "loss": 0.6992, + "step": 8112 + }, + { + "epoch": 0.35920879315497606, + "grad_norm": 1.3303951025009155, + "learning_rate": 1.6179159953985274e-05, + "loss": 0.6951, + "step": 8113 + }, + { + "epoch": 0.3592530688597899, + "grad_norm": 1.7109755277633667, + "learning_rate": 1.617794477848407e-05, + "loss": 0.7328, + "step": 8114 + }, + { + "epoch": 0.35929734456460377, + "grad_norm": 1.2806267738342285, + "learning_rate": 1.617672945542768e-05, + "loss": 0.7976, + "step": 8115 + }, + { + "epoch": 0.3593416202694177, + "grad_norm": 1.5980830192565918, + "learning_rate": 1.617551398484514e-05, + "loss": 0.9364, + "step": 8116 + }, + { + "epoch": 0.35938589597423154, + "grad_norm": 1.4582116603851318, + "learning_rate": 1.6174298366765476e-05, + "loss": 0.7952, + "step": 8117 + }, + { + "epoch": 0.3594301716790454, + "grad_norm": 1.2371770143508911, + "learning_rate": 1.6173082601217728e-05, + "loss": 0.7839, + "step": 8118 + }, + { + "epoch": 0.3594744473838593, + "grad_norm": 1.3892501592636108, + "learning_rate": 1.6171866688230923e-05, + "loss": 0.8219, + "step": 8119 + }, + { + "epoch": 0.35951872308867316, + "grad_norm": 1.4049891233444214, + "learning_rate": 1.6170650627834106e-05, + "loss": 0.8012, + "step": 8120 + }, + { + "epoch": 0.359562998793487, + "grad_norm": 1.287325382232666, + "learning_rate": 1.6169434420056325e-05, + "loss": 0.7761, + "step": 8121 + }, + { + "epoch": 0.35960727449830093, + "grad_norm": 1.9471502304077148, + "learning_rate": 1.616821806492663e-05, + "loss": 0.8328, + "step": 8122 + }, + { + "epoch": 0.3596515502031148, + "grad_norm": 1.3298827409744263, + "learning_rate": 1.616700156247406e-05, + "loss": 0.8035, + "step": 8123 + }, + { + "epoch": 0.35969582590792865, + "grad_norm": 1.2406495809555054, + "learning_rate": 1.6165784912727687e-05, + "loss": 0.7865, + "step": 8124 + }, + { + "epoch": 0.35974010161274256, + "grad_norm": 1.6167261600494385, + "learning_rate": 1.6164568115716556e-05, + "loss": 0.7578, + "step": 8125 + }, + { + "epoch": 0.3597843773175564, + "grad_norm": 1.433491826057434, + "learning_rate": 1.616335117146974e-05, + "loss": 0.7974, + "step": 8126 + }, + { + "epoch": 0.3598286530223703, + "grad_norm": 1.5235509872436523, + "learning_rate": 1.6162134080016298e-05, + "loss": 0.8581, + "step": 8127 + }, + { + "epoch": 0.3598729287271842, + "grad_norm": 1.2067501544952393, + "learning_rate": 1.6160916841385298e-05, + "loss": 0.7268, + "step": 8128 + }, + { + "epoch": 0.35991720443199804, + "grad_norm": 1.4227397441864014, + "learning_rate": 1.6159699455605823e-05, + "loss": 0.7254, + "step": 8129 + }, + { + "epoch": 0.35996148013681195, + "grad_norm": 1.3936012983322144, + "learning_rate": 1.6158481922706934e-05, + "loss": 0.7733, + "step": 8130 + }, + { + "epoch": 0.3600057558416258, + "grad_norm": 1.4067434072494507, + "learning_rate": 1.615726424271772e-05, + "loss": 0.8295, + "step": 8131 + }, + { + "epoch": 0.36005003154643966, + "grad_norm": 1.1525524854660034, + "learning_rate": 1.615604641566727e-05, + "loss": 0.7231, + "step": 8132 + }, + { + "epoch": 0.3600943072512536, + "grad_norm": 1.81107759475708, + "learning_rate": 1.6154828441584655e-05, + "loss": 0.8296, + "step": 8133 + }, + { + "epoch": 0.36013858295606743, + "grad_norm": 1.1892160177230835, + "learning_rate": 1.615361032049898e-05, + "loss": 0.7062, + "step": 8134 + }, + { + "epoch": 0.3601828586608813, + "grad_norm": 1.4777796268463135, + "learning_rate": 1.6152392052439328e-05, + "loss": 0.7908, + "step": 8135 + }, + { + "epoch": 0.3602271343656952, + "grad_norm": 1.277106523513794, + "learning_rate": 1.6151173637434805e-05, + "loss": 0.7726, + "step": 8136 + }, + { + "epoch": 0.36027141007050906, + "grad_norm": 1.2686641216278076, + "learning_rate": 1.6149955075514508e-05, + "loss": 0.7477, + "step": 8137 + }, + { + "epoch": 0.3603156857753229, + "grad_norm": 1.3400236368179321, + "learning_rate": 1.6148736366707538e-05, + "loss": 0.7831, + "step": 8138 + }, + { + "epoch": 0.3603599614801368, + "grad_norm": 1.4771078824996948, + "learning_rate": 1.614751751104301e-05, + "loss": 0.777, + "step": 8139 + }, + { + "epoch": 0.3604042371849507, + "grad_norm": 1.5640714168548584, + "learning_rate": 1.6146298508550033e-05, + "loss": 0.8786, + "step": 8140 + }, + { + "epoch": 0.36044851288976454, + "grad_norm": 1.6073477268218994, + "learning_rate": 1.6145079359257717e-05, + "loss": 0.8397, + "step": 8141 + }, + { + "epoch": 0.36049278859457845, + "grad_norm": 1.2710037231445312, + "learning_rate": 1.6143860063195184e-05, + "loss": 0.7349, + "step": 8142 + }, + { + "epoch": 0.3605370642993923, + "grad_norm": 1.4074714183807373, + "learning_rate": 1.6142640620391555e-05, + "loss": 0.808, + "step": 8143 + }, + { + "epoch": 0.3605813400042062, + "grad_norm": 1.402702808380127, + "learning_rate": 1.6141421030875962e-05, + "loss": 0.8274, + "step": 8144 + }, + { + "epoch": 0.3606256157090201, + "grad_norm": 1.63419771194458, + "learning_rate": 1.6140201294677525e-05, + "loss": 0.8823, + "step": 8145 + }, + { + "epoch": 0.36066989141383393, + "grad_norm": 1.4937920570373535, + "learning_rate": 1.6138981411825376e-05, + "loss": 0.7576, + "step": 8146 + }, + { + "epoch": 0.36071416711864784, + "grad_norm": 1.4463658332824707, + "learning_rate": 1.613776138234866e-05, + "loss": 0.8242, + "step": 8147 + }, + { + "epoch": 0.3607584428234617, + "grad_norm": 1.2663465738296509, + "learning_rate": 1.613654120627651e-05, + "loss": 0.7538, + "step": 8148 + }, + { + "epoch": 0.36080271852827556, + "grad_norm": 1.273527979850769, + "learning_rate": 1.6135320883638068e-05, + "loss": 0.7576, + "step": 8149 + }, + { + "epoch": 0.36084699423308947, + "grad_norm": 1.2764897346496582, + "learning_rate": 1.6134100414462483e-05, + "loss": 0.7984, + "step": 8150 + }, + { + "epoch": 0.3608912699379033, + "grad_norm": 1.7798196077346802, + "learning_rate": 1.6132879798778902e-05, + "loss": 0.7401, + "step": 8151 + }, + { + "epoch": 0.3609355456427172, + "grad_norm": 1.358007550239563, + "learning_rate": 1.6131659036616483e-05, + "loss": 0.7676, + "step": 8152 + }, + { + "epoch": 0.3609798213475311, + "grad_norm": 1.4905942678451538, + "learning_rate": 1.613043812800438e-05, + "loss": 0.8782, + "step": 8153 + }, + { + "epoch": 0.36102409705234495, + "grad_norm": 1.2816334962844849, + "learning_rate": 1.6129217072971757e-05, + "loss": 0.8064, + "step": 8154 + }, + { + "epoch": 0.3610683727571588, + "grad_norm": 1.4858546257019043, + "learning_rate": 1.612799587154777e-05, + "loss": 0.8715, + "step": 8155 + }, + { + "epoch": 0.3611126484619727, + "grad_norm": 1.6442618370056152, + "learning_rate": 1.6126774523761595e-05, + "loss": 0.7736, + "step": 8156 + }, + { + "epoch": 0.3611569241667866, + "grad_norm": 1.3171747922897339, + "learning_rate": 1.6125553029642394e-05, + "loss": 0.7386, + "step": 8157 + }, + { + "epoch": 0.36120119987160043, + "grad_norm": 1.3156126737594604, + "learning_rate": 1.6124331389219353e-05, + "loss": 0.8343, + "step": 8158 + }, + { + "epoch": 0.36124547557641434, + "grad_norm": 1.7433298826217651, + "learning_rate": 1.612310960252164e-05, + "loss": 0.8896, + "step": 8159 + }, + { + "epoch": 0.3612897512812282, + "grad_norm": 1.4318102598190308, + "learning_rate": 1.6121887669578437e-05, + "loss": 0.7918, + "step": 8160 + }, + { + "epoch": 0.3613340269860421, + "grad_norm": 1.2742235660552979, + "learning_rate": 1.6120665590418936e-05, + "loss": 0.7314, + "step": 8161 + }, + { + "epoch": 0.36137830269085597, + "grad_norm": 1.3448305130004883, + "learning_rate": 1.611944336507232e-05, + "loss": 0.8095, + "step": 8162 + }, + { + "epoch": 0.3614225783956698, + "grad_norm": 1.1604037284851074, + "learning_rate": 1.6118220993567783e-05, + "loss": 0.7773, + "step": 8163 + }, + { + "epoch": 0.36146685410048374, + "grad_norm": 1.3131881952285767, + "learning_rate": 1.6116998475934514e-05, + "loss": 0.7605, + "step": 8164 + }, + { + "epoch": 0.3615111298052976, + "grad_norm": 1.5154868364334106, + "learning_rate": 1.6115775812201723e-05, + "loss": 0.8218, + "step": 8165 + }, + { + "epoch": 0.36155540551011145, + "grad_norm": 1.2597771883010864, + "learning_rate": 1.6114553002398602e-05, + "loss": 0.8329, + "step": 8166 + }, + { + "epoch": 0.36159968121492536, + "grad_norm": 1.1463526487350464, + "learning_rate": 1.6113330046554363e-05, + "loss": 0.7038, + "step": 8167 + }, + { + "epoch": 0.3616439569197392, + "grad_norm": 1.226076364517212, + "learning_rate": 1.6112106944698215e-05, + "loss": 0.734, + "step": 8168 + }, + { + "epoch": 0.3616882326245531, + "grad_norm": 1.2183666229248047, + "learning_rate": 1.6110883696859368e-05, + "loss": 0.6986, + "step": 8169 + }, + { + "epoch": 0.361732508329367, + "grad_norm": 1.3265033960342407, + "learning_rate": 1.610966030306704e-05, + "loss": 0.8162, + "step": 8170 + }, + { + "epoch": 0.36177678403418084, + "grad_norm": 1.3199522495269775, + "learning_rate": 1.610843676335045e-05, + "loss": 0.7781, + "step": 8171 + }, + { + "epoch": 0.3618210597389947, + "grad_norm": 1.5634945631027222, + "learning_rate": 1.610721307773882e-05, + "loss": 0.7793, + "step": 8172 + }, + { + "epoch": 0.3618653354438086, + "grad_norm": 1.2043817043304443, + "learning_rate": 1.610598924626138e-05, + "loss": 0.6936, + "step": 8173 + }, + { + "epoch": 0.36190961114862247, + "grad_norm": 1.201546311378479, + "learning_rate": 1.6104765268947357e-05, + "loss": 0.7582, + "step": 8174 + }, + { + "epoch": 0.3619538868534363, + "grad_norm": 1.4544953107833862, + "learning_rate": 1.610354114582599e-05, + "loss": 0.752, + "step": 8175 + }, + { + "epoch": 0.36199816255825024, + "grad_norm": 1.5580369234085083, + "learning_rate": 1.610231687692651e-05, + "loss": 0.8531, + "step": 8176 + }, + { + "epoch": 0.3620424382630641, + "grad_norm": 1.3383806943893433, + "learning_rate": 1.6101092462278157e-05, + "loss": 0.7293, + "step": 8177 + }, + { + "epoch": 0.36208671396787795, + "grad_norm": 1.4851853847503662, + "learning_rate": 1.6099867901910184e-05, + "loss": 0.8544, + "step": 8178 + }, + { + "epoch": 0.36213098967269186, + "grad_norm": 1.3335448503494263, + "learning_rate": 1.609864319585183e-05, + "loss": 0.7906, + "step": 8179 + }, + { + "epoch": 0.3621752653775057, + "grad_norm": 1.5814539194107056, + "learning_rate": 1.6097418344132348e-05, + "loss": 0.8924, + "step": 8180 + }, + { + "epoch": 0.36221954108231963, + "grad_norm": 1.1843925714492798, + "learning_rate": 1.6096193346780995e-05, + "loss": 0.7411, + "step": 8181 + }, + { + "epoch": 0.3622638167871335, + "grad_norm": 1.3378396034240723, + "learning_rate": 1.6094968203827027e-05, + "loss": 0.7399, + "step": 8182 + }, + { + "epoch": 0.36230809249194734, + "grad_norm": 1.3002960681915283, + "learning_rate": 1.6093742915299706e-05, + "loss": 0.6709, + "step": 8183 + }, + { + "epoch": 0.36235236819676125, + "grad_norm": 1.6533864736557007, + "learning_rate": 1.6092517481228296e-05, + "loss": 0.7419, + "step": 8184 + }, + { + "epoch": 0.3623966439015751, + "grad_norm": 1.4615248441696167, + "learning_rate": 1.609129190164207e-05, + "loss": 0.7447, + "step": 8185 + }, + { + "epoch": 0.36244091960638897, + "grad_norm": 1.2358534336090088, + "learning_rate": 1.6090066176570297e-05, + "loss": 0.7143, + "step": 8186 + }, + { + "epoch": 0.3624851953112029, + "grad_norm": 1.216903805732727, + "learning_rate": 1.6088840306042247e-05, + "loss": 0.7983, + "step": 8187 + }, + { + "epoch": 0.36252947101601674, + "grad_norm": 1.4008229970932007, + "learning_rate": 1.608761429008721e-05, + "loss": 0.8515, + "step": 8188 + }, + { + "epoch": 0.3625737467208306, + "grad_norm": 1.3270437717437744, + "learning_rate": 1.6086388128734455e-05, + "loss": 0.7366, + "step": 8189 + }, + { + "epoch": 0.3626180224256445, + "grad_norm": 1.529910683631897, + "learning_rate": 1.608516182201328e-05, + "loss": 0.7911, + "step": 8190 + }, + { + "epoch": 0.36266229813045836, + "grad_norm": 1.6157618761062622, + "learning_rate": 1.6083935369952972e-05, + "loss": 0.8063, + "step": 8191 + }, + { + "epoch": 0.3627065738352722, + "grad_norm": 1.1423237323760986, + "learning_rate": 1.6082708772582817e-05, + "loss": 0.741, + "step": 8192 + }, + { + "epoch": 0.36275084954008613, + "grad_norm": 1.3853157758712769, + "learning_rate": 1.6081482029932117e-05, + "loss": 0.8083, + "step": 8193 + }, + { + "epoch": 0.3627951252449, + "grad_norm": 1.6209667921066284, + "learning_rate": 1.6080255142030172e-05, + "loss": 0.8718, + "step": 8194 + }, + { + "epoch": 0.36283940094971384, + "grad_norm": 1.4786852598190308, + "learning_rate": 1.607902810890628e-05, + "loss": 0.839, + "step": 8195 + }, + { + "epoch": 0.36288367665452775, + "grad_norm": 1.1565271615982056, + "learning_rate": 1.6077800930589756e-05, + "loss": 0.6908, + "step": 8196 + }, + { + "epoch": 0.3629279523593416, + "grad_norm": 1.4349820613861084, + "learning_rate": 1.6076573607109903e-05, + "loss": 0.7809, + "step": 8197 + }, + { + "epoch": 0.3629722280641555, + "grad_norm": 1.2507562637329102, + "learning_rate": 1.6075346138496037e-05, + "loss": 0.7567, + "step": 8198 + }, + { + "epoch": 0.3630165037689694, + "grad_norm": 1.411590576171875, + "learning_rate": 1.607411852477748e-05, + "loss": 0.7691, + "step": 8199 + }, + { + "epoch": 0.36306077947378324, + "grad_norm": 1.3410998582839966, + "learning_rate": 1.607289076598354e-05, + "loss": 0.7396, + "step": 8200 + }, + { + "epoch": 0.36310505517859715, + "grad_norm": 1.1919294595718384, + "learning_rate": 1.6071662862143554e-05, + "loss": 0.7429, + "step": 8201 + }, + { + "epoch": 0.363149330883411, + "grad_norm": 1.357591152191162, + "learning_rate": 1.607043481328684e-05, + "loss": 0.8414, + "step": 8202 + }, + { + "epoch": 0.36319360658822486, + "grad_norm": 1.2571029663085938, + "learning_rate": 1.6069206619442738e-05, + "loss": 0.6977, + "step": 8203 + }, + { + "epoch": 0.36323788229303877, + "grad_norm": 1.3154840469360352, + "learning_rate": 1.6067978280640575e-05, + "loss": 0.7875, + "step": 8204 + }, + { + "epoch": 0.36328215799785263, + "grad_norm": 1.333236813545227, + "learning_rate": 1.606674979690969e-05, + "loss": 0.7451, + "step": 8205 + }, + { + "epoch": 0.3633264337026665, + "grad_norm": 2.0313026905059814, + "learning_rate": 1.6065521168279427e-05, + "loss": 0.8952, + "step": 8206 + }, + { + "epoch": 0.3633707094074804, + "grad_norm": 1.493260383605957, + "learning_rate": 1.6064292394779132e-05, + "loss": 0.8236, + "step": 8207 + }, + { + "epoch": 0.36341498511229425, + "grad_norm": 1.4197542667388916, + "learning_rate": 1.6063063476438148e-05, + "loss": 0.7892, + "step": 8208 + }, + { + "epoch": 0.3634592608171081, + "grad_norm": 1.8426390886306763, + "learning_rate": 1.6061834413285832e-05, + "loss": 0.7868, + "step": 8209 + }, + { + "epoch": 0.363503536521922, + "grad_norm": 1.4592417478561401, + "learning_rate": 1.6060605205351534e-05, + "loss": 0.7814, + "step": 8210 + }, + { + "epoch": 0.3635478122267359, + "grad_norm": 1.2787669897079468, + "learning_rate": 1.605937585266462e-05, + "loss": 0.7768, + "step": 8211 + }, + { + "epoch": 0.36359208793154973, + "grad_norm": 1.3236883878707886, + "learning_rate": 1.605814635525444e-05, + "loss": 0.8083, + "step": 8212 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 1.4087035655975342, + "learning_rate": 1.6056916713150373e-05, + "loss": 0.8064, + "step": 8213 + }, + { + "epoch": 0.3636806393411775, + "grad_norm": 1.5325640439987183, + "learning_rate": 1.6055686926381782e-05, + "loss": 0.8312, + "step": 8214 + }, + { + "epoch": 0.3637249150459914, + "grad_norm": 1.194460391998291, + "learning_rate": 1.605445699497804e-05, + "loss": 0.703, + "step": 8215 + }, + { + "epoch": 0.36376919075080527, + "grad_norm": 1.2928537130355835, + "learning_rate": 1.605322691896852e-05, + "loss": 0.7591, + "step": 8216 + }, + { + "epoch": 0.36381346645561913, + "grad_norm": 1.3743640184402466, + "learning_rate": 1.6051996698382603e-05, + "loss": 0.7758, + "step": 8217 + }, + { + "epoch": 0.36385774216043304, + "grad_norm": 1.340847373008728, + "learning_rate": 1.6050766333249673e-05, + "loss": 0.8031, + "step": 8218 + }, + { + "epoch": 0.3639020178652469, + "grad_norm": 1.3335868120193481, + "learning_rate": 1.6049535823599115e-05, + "loss": 0.8043, + "step": 8219 + }, + { + "epoch": 0.36394629357006075, + "grad_norm": 1.4377565383911133, + "learning_rate": 1.6048305169460322e-05, + "loss": 0.693, + "step": 8220 + }, + { + "epoch": 0.36399056927487466, + "grad_norm": 1.675707459449768, + "learning_rate": 1.6047074370862686e-05, + "loss": 0.8185, + "step": 8221 + }, + { + "epoch": 0.3640348449796885, + "grad_norm": 1.1082814931869507, + "learning_rate": 1.60458434278356e-05, + "loss": 0.735, + "step": 8222 + }, + { + "epoch": 0.3640791206845024, + "grad_norm": 1.6171334981918335, + "learning_rate": 1.6044612340408466e-05, + "loss": 0.7927, + "step": 8223 + }, + { + "epoch": 0.3641233963893163, + "grad_norm": 1.26100754737854, + "learning_rate": 1.604338110861069e-05, + "loss": 0.7484, + "step": 8224 + }, + { + "epoch": 0.36416767209413015, + "grad_norm": 1.1434353590011597, + "learning_rate": 1.604214973247167e-05, + "loss": 0.6977, + "step": 8225 + }, + { + "epoch": 0.364211947798944, + "grad_norm": 1.3396331071853638, + "learning_rate": 1.6040918212020834e-05, + "loss": 0.7827, + "step": 8226 + }, + { + "epoch": 0.3642562235037579, + "grad_norm": 1.4178837537765503, + "learning_rate": 1.6039686547287577e-05, + "loss": 0.7755, + "step": 8227 + }, + { + "epoch": 0.36430049920857177, + "grad_norm": 1.5196478366851807, + "learning_rate": 1.6038454738301327e-05, + "loss": 0.8067, + "step": 8228 + }, + { + "epoch": 0.3643447749133856, + "grad_norm": 1.3075569868087769, + "learning_rate": 1.6037222785091504e-05, + "loss": 0.7728, + "step": 8229 + }, + { + "epoch": 0.36438905061819954, + "grad_norm": 1.915911078453064, + "learning_rate": 1.603599068768753e-05, + "loss": 0.8453, + "step": 8230 + }, + { + "epoch": 0.3644333263230134, + "grad_norm": 1.2451835870742798, + "learning_rate": 1.6034758446118832e-05, + "loss": 0.708, + "step": 8231 + }, + { + "epoch": 0.3644776020278273, + "grad_norm": 1.2187579870224, + "learning_rate": 1.6033526060414845e-05, + "loss": 0.731, + "step": 8232 + }, + { + "epoch": 0.36452187773264116, + "grad_norm": 1.2518293857574463, + "learning_rate": 1.6032293530604996e-05, + "loss": 0.7193, + "step": 8233 + }, + { + "epoch": 0.364566153437455, + "grad_norm": 1.45266854763031, + "learning_rate": 1.603106085671873e-05, + "loss": 0.819, + "step": 8234 + }, + { + "epoch": 0.36461042914226893, + "grad_norm": 1.4179540872573853, + "learning_rate": 1.6029828038785486e-05, + "loss": 0.8537, + "step": 8235 + }, + { + "epoch": 0.3646547048470828, + "grad_norm": 1.1173360347747803, + "learning_rate": 1.6028595076834708e-05, + "loss": 0.6485, + "step": 8236 + }, + { + "epoch": 0.36469898055189665, + "grad_norm": 1.2692245244979858, + "learning_rate": 1.6027361970895847e-05, + "loss": 0.7239, + "step": 8237 + }, + { + "epoch": 0.36474325625671056, + "grad_norm": 1.7366825342178345, + "learning_rate": 1.6026128720998353e-05, + "loss": 0.808, + "step": 8238 + }, + { + "epoch": 0.3647875319615244, + "grad_norm": 1.3850489854812622, + "learning_rate": 1.6024895327171683e-05, + "loss": 0.7757, + "step": 8239 + }, + { + "epoch": 0.36483180766633827, + "grad_norm": 1.2786715030670166, + "learning_rate": 1.6023661789445285e-05, + "loss": 0.7544, + "step": 8240 + }, + { + "epoch": 0.3648760833711522, + "grad_norm": 1.428541898727417, + "learning_rate": 1.602242810784864e-05, + "loss": 0.7838, + "step": 8241 + }, + { + "epoch": 0.36492035907596604, + "grad_norm": 1.4853310585021973, + "learning_rate": 1.60211942824112e-05, + "loss": 0.7563, + "step": 8242 + }, + { + "epoch": 0.3649646347807799, + "grad_norm": 1.417077660560608, + "learning_rate": 1.6019960313162436e-05, + "loss": 0.79, + "step": 8243 + }, + { + "epoch": 0.3650089104855938, + "grad_norm": 1.6956562995910645, + "learning_rate": 1.6018726200131823e-05, + "loss": 0.8139, + "step": 8244 + }, + { + "epoch": 0.36505318619040766, + "grad_norm": 1.309878945350647, + "learning_rate": 1.6017491943348832e-05, + "loss": 0.7152, + "step": 8245 + }, + { + "epoch": 0.3650974618952215, + "grad_norm": 1.1846978664398193, + "learning_rate": 1.6016257542842952e-05, + "loss": 0.749, + "step": 8246 + }, + { + "epoch": 0.36514173760003543, + "grad_norm": 1.3410643339157104, + "learning_rate": 1.601502299864365e-05, + "loss": 0.7756, + "step": 8247 + }, + { + "epoch": 0.3651860133048493, + "grad_norm": 1.5146138668060303, + "learning_rate": 1.601378831078043e-05, + "loss": 0.8411, + "step": 8248 + }, + { + "epoch": 0.36523028900966314, + "grad_norm": 1.4699333906173706, + "learning_rate": 1.601255347928277e-05, + "loss": 0.7915, + "step": 8249 + }, + { + "epoch": 0.36527456471447706, + "grad_norm": 1.2063543796539307, + "learning_rate": 1.6011318504180163e-05, + "loss": 0.7381, + "step": 8250 + }, + { + "epoch": 0.3653188404192909, + "grad_norm": 1.3477612733840942, + "learning_rate": 1.601008338550211e-05, + "loss": 0.7788, + "step": 8251 + }, + { + "epoch": 0.3653631161241048, + "grad_norm": 1.1617770195007324, + "learning_rate": 1.600884812327811e-05, + "loss": 0.7039, + "step": 8252 + }, + { + "epoch": 0.3654073918289187, + "grad_norm": 1.5190054178237915, + "learning_rate": 1.600761271753766e-05, + "loss": 0.7881, + "step": 8253 + }, + { + "epoch": 0.36545166753373254, + "grad_norm": 1.5813579559326172, + "learning_rate": 1.6006377168310278e-05, + "loss": 0.8026, + "step": 8254 + }, + { + "epoch": 0.36549594323854645, + "grad_norm": 2.726883888244629, + "learning_rate": 1.600514147562547e-05, + "loss": 0.8662, + "step": 8255 + }, + { + "epoch": 0.3655402189433603, + "grad_norm": 1.2335785627365112, + "learning_rate": 1.6003905639512743e-05, + "loss": 0.7193, + "step": 8256 + }, + { + "epoch": 0.36558449464817416, + "grad_norm": 1.5159600973129272, + "learning_rate": 1.600266966000162e-05, + "loss": 0.7737, + "step": 8257 + }, + { + "epoch": 0.3656287703529881, + "grad_norm": 1.5527132749557495, + "learning_rate": 1.6001433537121616e-05, + "loss": 0.8227, + "step": 8258 + }, + { + "epoch": 0.36567304605780193, + "grad_norm": 1.2167106866836548, + "learning_rate": 1.600019727090226e-05, + "loss": 0.7099, + "step": 8259 + }, + { + "epoch": 0.3657173217626158, + "grad_norm": 1.1880736351013184, + "learning_rate": 1.5998960861373077e-05, + "loss": 0.6996, + "step": 8260 + }, + { + "epoch": 0.3657615974674297, + "grad_norm": 1.274370551109314, + "learning_rate": 1.5997724308563603e-05, + "loss": 0.7832, + "step": 8261 + }, + { + "epoch": 0.36580587317224356, + "grad_norm": 1.2815240621566772, + "learning_rate": 1.5996487612503364e-05, + "loss": 0.7882, + "step": 8262 + }, + { + "epoch": 0.3658501488770574, + "grad_norm": 1.2192277908325195, + "learning_rate": 1.59952507732219e-05, + "loss": 0.703, + "step": 8263 + }, + { + "epoch": 0.3658944245818713, + "grad_norm": 1.2698662281036377, + "learning_rate": 1.5994013790748755e-05, + "loss": 0.7628, + "step": 8264 + }, + { + "epoch": 0.3659387002866852, + "grad_norm": 1.874612808227539, + "learning_rate": 1.599277666511347e-05, + "loss": 0.8644, + "step": 8265 + }, + { + "epoch": 0.36598297599149904, + "grad_norm": 1.532187819480896, + "learning_rate": 1.5991539396345596e-05, + "loss": 0.8401, + "step": 8266 + }, + { + "epoch": 0.36602725169631295, + "grad_norm": 1.2911431789398193, + "learning_rate": 1.599030198447468e-05, + "loss": 0.6958, + "step": 8267 + }, + { + "epoch": 0.3660715274011268, + "grad_norm": 1.1835063695907593, + "learning_rate": 1.5989064429530276e-05, + "loss": 0.7018, + "step": 8268 + }, + { + "epoch": 0.3661158031059407, + "grad_norm": 1.4575438499450684, + "learning_rate": 1.598782673154195e-05, + "loss": 0.7482, + "step": 8269 + }, + { + "epoch": 0.3661600788107546, + "grad_norm": 1.4767049551010132, + "learning_rate": 1.5986588890539254e-05, + "loss": 0.7833, + "step": 8270 + }, + { + "epoch": 0.36620435451556843, + "grad_norm": 1.132590651512146, + "learning_rate": 1.598535090655176e-05, + "loss": 0.7127, + "step": 8271 + }, + { + "epoch": 0.36624863022038234, + "grad_norm": 1.2930978536605835, + "learning_rate": 1.5984112779609028e-05, + "loss": 0.796, + "step": 8272 + }, + { + "epoch": 0.3662929059251962, + "grad_norm": 1.5179455280303955, + "learning_rate": 1.598287450974064e-05, + "loss": 0.7788, + "step": 8273 + }, + { + "epoch": 0.36633718163001006, + "grad_norm": 1.3500198125839233, + "learning_rate": 1.5981636096976167e-05, + "loss": 0.8209, + "step": 8274 + }, + { + "epoch": 0.36638145733482397, + "grad_norm": 1.389875888824463, + "learning_rate": 1.598039754134518e-05, + "loss": 0.794, + "step": 8275 + }, + { + "epoch": 0.3664257330396378, + "grad_norm": 1.4443773031234741, + "learning_rate": 1.597915884287727e-05, + "loss": 0.7934, + "step": 8276 + }, + { + "epoch": 0.3664700087444517, + "grad_norm": 1.2731624841690063, + "learning_rate": 1.597792000160202e-05, + "loss": 0.7564, + "step": 8277 + }, + { + "epoch": 0.3665142844492656, + "grad_norm": 1.6260346174240112, + "learning_rate": 1.5976681017549015e-05, + "loss": 0.8026, + "step": 8278 + }, + { + "epoch": 0.36655856015407945, + "grad_norm": 1.6104180812835693, + "learning_rate": 1.5975441890747855e-05, + "loss": 0.8883, + "step": 8279 + }, + { + "epoch": 0.3666028358588933, + "grad_norm": 1.1674847602844238, + "learning_rate": 1.597420262122813e-05, + "loss": 0.6588, + "step": 8280 + }, + { + "epoch": 0.3666471115637072, + "grad_norm": 1.2808877229690552, + "learning_rate": 1.597296320901944e-05, + "loss": 0.7547, + "step": 8281 + }, + { + "epoch": 0.3666913872685211, + "grad_norm": 1.3753193616867065, + "learning_rate": 1.5971723654151383e-05, + "loss": 0.7559, + "step": 8282 + }, + { + "epoch": 0.36673566297333493, + "grad_norm": 1.5696816444396973, + "learning_rate": 1.5970483956653572e-05, + "loss": 0.8107, + "step": 8283 + }, + { + "epoch": 0.36677993867814884, + "grad_norm": 1.4736155271530151, + "learning_rate": 1.5969244116555613e-05, + "loss": 0.8046, + "step": 8284 + }, + { + "epoch": 0.3668242143829627, + "grad_norm": 1.4759951829910278, + "learning_rate": 1.5968004133887118e-05, + "loss": 0.8212, + "step": 8285 + }, + { + "epoch": 0.3668684900877766, + "grad_norm": 1.3534282445907593, + "learning_rate": 1.5966764008677703e-05, + "loss": 0.8274, + "step": 8286 + }, + { + "epoch": 0.36691276579259047, + "grad_norm": 1.4823739528656006, + "learning_rate": 1.596552374095699e-05, + "loss": 0.7954, + "step": 8287 + }, + { + "epoch": 0.3669570414974043, + "grad_norm": 1.4273244142532349, + "learning_rate": 1.5964283330754598e-05, + "loss": 0.7882, + "step": 8288 + }, + { + "epoch": 0.36700131720221824, + "grad_norm": 1.4036105871200562, + "learning_rate": 1.5963042778100153e-05, + "loss": 0.6912, + "step": 8289 + }, + { + "epoch": 0.3670455929070321, + "grad_norm": 1.3346213102340698, + "learning_rate": 1.5961802083023292e-05, + "loss": 0.7305, + "step": 8290 + }, + { + "epoch": 0.36708986861184595, + "grad_norm": 1.502434492111206, + "learning_rate": 1.596056124555364e-05, + "loss": 0.7527, + "step": 8291 + }, + { + "epoch": 0.36713414431665986, + "grad_norm": 1.2242058515548706, + "learning_rate": 1.5959320265720835e-05, + "loss": 0.6929, + "step": 8292 + }, + { + "epoch": 0.3671784200214737, + "grad_norm": 1.5103343725204468, + "learning_rate": 1.5958079143554515e-05, + "loss": 0.7944, + "step": 8293 + }, + { + "epoch": 0.3672226957262876, + "grad_norm": 1.1525171995162964, + "learning_rate": 1.5956837879084332e-05, + "loss": 0.763, + "step": 8294 + }, + { + "epoch": 0.3672669714311015, + "grad_norm": 1.3614033460617065, + "learning_rate": 1.5955596472339926e-05, + "loss": 0.8031, + "step": 8295 + }, + { + "epoch": 0.36731124713591534, + "grad_norm": 1.8351707458496094, + "learning_rate": 1.5954354923350943e-05, + "loss": 0.7567, + "step": 8296 + }, + { + "epoch": 0.3673555228407292, + "grad_norm": 1.6526700258255005, + "learning_rate": 1.5953113232147042e-05, + "loss": 0.8378, + "step": 8297 + }, + { + "epoch": 0.3673997985455431, + "grad_norm": 4.930476188659668, + "learning_rate": 1.595187139875788e-05, + "loss": 0.8468, + "step": 8298 + }, + { + "epoch": 0.36744407425035697, + "grad_norm": 1.245984673500061, + "learning_rate": 1.5950629423213116e-05, + "loss": 0.7157, + "step": 8299 + }, + { + "epoch": 0.3674883499551708, + "grad_norm": 1.3311774730682373, + "learning_rate": 1.594938730554241e-05, + "loss": 0.7886, + "step": 8300 + }, + { + "epoch": 0.36753262565998474, + "grad_norm": 1.2931747436523438, + "learning_rate": 1.594814504577544e-05, + "loss": 0.8157, + "step": 8301 + }, + { + "epoch": 0.3675769013647986, + "grad_norm": 1.4019392728805542, + "learning_rate": 1.594690264394186e-05, + "loss": 0.7661, + "step": 8302 + }, + { + "epoch": 0.3676211770696125, + "grad_norm": 1.5424996614456177, + "learning_rate": 1.5945660100071362e-05, + "loss": 0.815, + "step": 8303 + }, + { + "epoch": 0.36766545277442636, + "grad_norm": 1.3000988960266113, + "learning_rate": 1.5944417414193605e-05, + "loss": 0.7722, + "step": 8304 + }, + { + "epoch": 0.3677097284792402, + "grad_norm": 1.2829185724258423, + "learning_rate": 1.5943174586338282e-05, + "loss": 0.7393, + "step": 8305 + }, + { + "epoch": 0.36775400418405413, + "grad_norm": 1.4384238719940186, + "learning_rate": 1.5941931616535074e-05, + "loss": 0.7978, + "step": 8306 + }, + { + "epoch": 0.367798279888868, + "grad_norm": 1.307381510734558, + "learning_rate": 1.5940688504813664e-05, + "loss": 0.7448, + "step": 8307 + }, + { + "epoch": 0.36784255559368184, + "grad_norm": 1.358407974243164, + "learning_rate": 1.593944525120375e-05, + "loss": 0.7932, + "step": 8308 + }, + { + "epoch": 0.36788683129849575, + "grad_norm": 1.3396421670913696, + "learning_rate": 1.5938201855735017e-05, + "loss": 0.7177, + "step": 8309 + }, + { + "epoch": 0.3679311070033096, + "grad_norm": 1.6457126140594482, + "learning_rate": 1.593695831843717e-05, + "loss": 0.7287, + "step": 8310 + }, + { + "epoch": 0.36797538270812347, + "grad_norm": 1.2643929719924927, + "learning_rate": 1.5935714639339912e-05, + "loss": 0.6639, + "step": 8311 + }, + { + "epoch": 0.3680196584129374, + "grad_norm": 1.201440453529358, + "learning_rate": 1.5934470818472937e-05, + "loss": 0.6631, + "step": 8312 + }, + { + "epoch": 0.36806393411775123, + "grad_norm": 1.854164481163025, + "learning_rate": 1.5933226855865962e-05, + "loss": 0.8667, + "step": 8313 + }, + { + "epoch": 0.3681082098225651, + "grad_norm": 1.089116096496582, + "learning_rate": 1.5931982751548695e-05, + "loss": 0.6346, + "step": 8314 + }, + { + "epoch": 0.368152485527379, + "grad_norm": 1.1953552961349487, + "learning_rate": 1.5930738505550846e-05, + "loss": 0.7953, + "step": 8315 + }, + { + "epoch": 0.36819676123219286, + "grad_norm": 1.2983486652374268, + "learning_rate": 1.592949411790214e-05, + "loss": 0.7414, + "step": 8316 + }, + { + "epoch": 0.3682410369370067, + "grad_norm": 1.30595862865448, + "learning_rate": 1.5928249588632294e-05, + "loss": 0.7375, + "step": 8317 + }, + { + "epoch": 0.36828531264182063, + "grad_norm": 1.2522937059402466, + "learning_rate": 1.5927004917771033e-05, + "loss": 0.8169, + "step": 8318 + }, + { + "epoch": 0.3683295883466345, + "grad_norm": 1.4350602626800537, + "learning_rate": 1.592576010534809e-05, + "loss": 0.792, + "step": 8319 + }, + { + "epoch": 0.3683738640514484, + "grad_norm": 1.3681347370147705, + "learning_rate": 1.5924515151393184e-05, + "loss": 0.8257, + "step": 8320 + }, + { + "epoch": 0.36841813975626225, + "grad_norm": 1.5278677940368652, + "learning_rate": 1.5923270055936065e-05, + "loss": 0.7551, + "step": 8321 + }, + { + "epoch": 0.3684624154610761, + "grad_norm": 1.5001965761184692, + "learning_rate": 1.592202481900646e-05, + "loss": 0.8162, + "step": 8322 + }, + { + "epoch": 0.36850669116589, + "grad_norm": 1.331443190574646, + "learning_rate": 1.5920779440634113e-05, + "loss": 0.8009, + "step": 8323 + }, + { + "epoch": 0.3685509668707039, + "grad_norm": 1.9788190126419067, + "learning_rate": 1.5919533920848774e-05, + "loss": 0.8074, + "step": 8324 + }, + { + "epoch": 0.36859524257551773, + "grad_norm": 1.1212029457092285, + "learning_rate": 1.591828825968018e-05, + "loss": 0.7004, + "step": 8325 + }, + { + "epoch": 0.36863951828033165, + "grad_norm": 1.2539544105529785, + "learning_rate": 1.5917042457158097e-05, + "loss": 0.709, + "step": 8326 + }, + { + "epoch": 0.3686837939851455, + "grad_norm": 1.224495530128479, + "learning_rate": 1.591579651331227e-05, + "loss": 0.7202, + "step": 8327 + }, + { + "epoch": 0.36872806968995936, + "grad_norm": 1.3219884634017944, + "learning_rate": 1.5914550428172463e-05, + "loss": 0.8109, + "step": 8328 + }, + { + "epoch": 0.36877234539477327, + "grad_norm": 1.2831380367279053, + "learning_rate": 1.591330420176843e-05, + "loss": 0.7158, + "step": 8329 + }, + { + "epoch": 0.3688166210995871, + "grad_norm": 1.3586726188659668, + "learning_rate": 1.591205783412994e-05, + "loss": 0.7816, + "step": 8330 + }, + { + "epoch": 0.368860896804401, + "grad_norm": 1.312790870666504, + "learning_rate": 1.5910811325286768e-05, + "loss": 0.7852, + "step": 8331 + }, + { + "epoch": 0.3689051725092149, + "grad_norm": 1.4249672889709473, + "learning_rate": 1.5909564675268675e-05, + "loss": 0.8093, + "step": 8332 + }, + { + "epoch": 0.36894944821402875, + "grad_norm": 1.264148235321045, + "learning_rate": 1.5908317884105446e-05, + "loss": 0.7471, + "step": 8333 + }, + { + "epoch": 0.3689937239188426, + "grad_norm": 1.581563949584961, + "learning_rate": 1.5907070951826854e-05, + "loss": 0.8783, + "step": 8334 + }, + { + "epoch": 0.3690379996236565, + "grad_norm": 1.4094328880310059, + "learning_rate": 1.590582387846268e-05, + "loss": 0.8275, + "step": 8335 + }, + { + "epoch": 0.3690822753284704, + "grad_norm": 1.425469994544983, + "learning_rate": 1.5904576664042712e-05, + "loss": 0.7704, + "step": 8336 + }, + { + "epoch": 0.36912655103328423, + "grad_norm": 1.1518577337265015, + "learning_rate": 1.590332930859674e-05, + "loss": 0.7428, + "step": 8337 + }, + { + "epoch": 0.36917082673809815, + "grad_norm": 1.5191189050674438, + "learning_rate": 1.5902081812154552e-05, + "loss": 0.9179, + "step": 8338 + }, + { + "epoch": 0.369215102442912, + "grad_norm": 1.5827276706695557, + "learning_rate": 1.5900834174745945e-05, + "loss": 0.813, + "step": 8339 + }, + { + "epoch": 0.3692593781477259, + "grad_norm": 1.2080708742141724, + "learning_rate": 1.589958639640072e-05, + "loss": 0.7737, + "step": 8340 + }, + { + "epoch": 0.36930365385253977, + "grad_norm": 1.2176222801208496, + "learning_rate": 1.589833847714868e-05, + "loss": 0.7355, + "step": 8341 + }, + { + "epoch": 0.3693479295573536, + "grad_norm": 1.2157410383224487, + "learning_rate": 1.5897090417019622e-05, + "loss": 0.7192, + "step": 8342 + }, + { + "epoch": 0.36939220526216754, + "grad_norm": 1.2414754629135132, + "learning_rate": 1.5895842216043366e-05, + "loss": 0.7497, + "step": 8343 + }, + { + "epoch": 0.3694364809669814, + "grad_norm": 1.4390242099761963, + "learning_rate": 1.5894593874249714e-05, + "loss": 0.8677, + "step": 8344 + }, + { + "epoch": 0.36948075667179525, + "grad_norm": 1.307461142539978, + "learning_rate": 1.589334539166849e-05, + "loss": 0.6804, + "step": 8345 + }, + { + "epoch": 0.36952503237660916, + "grad_norm": 1.2184101343154907, + "learning_rate": 1.5892096768329514e-05, + "loss": 0.7238, + "step": 8346 + }, + { + "epoch": 0.369569308081423, + "grad_norm": 1.1927845478057861, + "learning_rate": 1.5890848004262597e-05, + "loss": 0.7008, + "step": 8347 + }, + { + "epoch": 0.3696135837862369, + "grad_norm": 1.3062580823898315, + "learning_rate": 1.588959909949758e-05, + "loss": 0.7657, + "step": 8348 + }, + { + "epoch": 0.3696578594910508, + "grad_norm": 1.2131972312927246, + "learning_rate": 1.5888350054064275e-05, + "loss": 0.7459, + "step": 8349 + }, + { + "epoch": 0.36970213519586465, + "grad_norm": 1.4193309545516968, + "learning_rate": 1.588710086799253e-05, + "loss": 0.7898, + "step": 8350 + }, + { + "epoch": 0.3697464109006785, + "grad_norm": 1.315590739250183, + "learning_rate": 1.5885851541312172e-05, + "loss": 0.8239, + "step": 8351 + }, + { + "epoch": 0.3697906866054924, + "grad_norm": 1.357221007347107, + "learning_rate": 1.588460207405304e-05, + "loss": 0.7471, + "step": 8352 + }, + { + "epoch": 0.36983496231030627, + "grad_norm": 1.569053053855896, + "learning_rate": 1.5883352466244986e-05, + "loss": 0.8952, + "step": 8353 + }, + { + "epoch": 0.3698792380151201, + "grad_norm": 1.3760932683944702, + "learning_rate": 1.5882102717917842e-05, + "loss": 0.8028, + "step": 8354 + }, + { + "epoch": 0.36992351371993404, + "grad_norm": 1.6110163927078247, + "learning_rate": 1.5880852829101464e-05, + "loss": 0.8442, + "step": 8355 + }, + { + "epoch": 0.3699677894247479, + "grad_norm": 1.3052668571472168, + "learning_rate": 1.587960279982571e-05, + "loss": 0.8149, + "step": 8356 + }, + { + "epoch": 0.3700120651295618, + "grad_norm": 1.2415344715118408, + "learning_rate": 1.5878352630120426e-05, + "loss": 0.7029, + "step": 8357 + }, + { + "epoch": 0.37005634083437566, + "grad_norm": 1.380140781402588, + "learning_rate": 1.5877102320015478e-05, + "loss": 0.728, + "step": 8358 + }, + { + "epoch": 0.3701006165391895, + "grad_norm": 1.4750089645385742, + "learning_rate": 1.587585186954072e-05, + "loss": 0.8137, + "step": 8359 + }, + { + "epoch": 0.37014489224400343, + "grad_norm": 1.3947290182113647, + "learning_rate": 1.5874601278726032e-05, + "loss": 0.8158, + "step": 8360 + }, + { + "epoch": 0.3701891679488173, + "grad_norm": 1.5825214385986328, + "learning_rate": 1.5873350547601274e-05, + "loss": 0.7974, + "step": 8361 + }, + { + "epoch": 0.37023344365363114, + "grad_norm": 1.3478925228118896, + "learning_rate": 1.5872099676196316e-05, + "loss": 0.808, + "step": 8362 + }, + { + "epoch": 0.37027771935844506, + "grad_norm": 1.2142976522445679, + "learning_rate": 1.5870848664541046e-05, + "loss": 0.7718, + "step": 8363 + }, + { + "epoch": 0.3703219950632589, + "grad_norm": 1.4010093212127686, + "learning_rate": 1.5869597512665332e-05, + "loss": 0.7828, + "step": 8364 + }, + { + "epoch": 0.37036627076807277, + "grad_norm": 1.143093228340149, + "learning_rate": 1.586834622059906e-05, + "loss": 0.7037, + "step": 8365 + }, + { + "epoch": 0.3704105464728867, + "grad_norm": 1.145228624343872, + "learning_rate": 1.5867094788372117e-05, + "loss": 0.7095, + "step": 8366 + }, + { + "epoch": 0.37045482217770054, + "grad_norm": 1.525976538658142, + "learning_rate": 1.5865843216014392e-05, + "loss": 0.8596, + "step": 8367 + }, + { + "epoch": 0.3704990978825144, + "grad_norm": 1.228470802307129, + "learning_rate": 1.586459150355578e-05, + "loss": 0.7909, + "step": 8368 + }, + { + "epoch": 0.3705433735873283, + "grad_norm": 1.353715419769287, + "learning_rate": 1.5863339651026173e-05, + "loss": 0.7704, + "step": 8369 + }, + { + "epoch": 0.37058764929214216, + "grad_norm": 1.1194992065429688, + "learning_rate": 1.586208765845547e-05, + "loss": 0.7451, + "step": 8370 + }, + { + "epoch": 0.370631924996956, + "grad_norm": 1.3580278158187866, + "learning_rate": 1.586083552587358e-05, + "loss": 0.7366, + "step": 8371 + }, + { + "epoch": 0.37067620070176993, + "grad_norm": 1.287277340888977, + "learning_rate": 1.5859583253310407e-05, + "loss": 0.7674, + "step": 8372 + }, + { + "epoch": 0.3707204764065838, + "grad_norm": 1.3791072368621826, + "learning_rate": 1.5858330840795856e-05, + "loss": 0.7659, + "step": 8373 + }, + { + "epoch": 0.3707647521113977, + "grad_norm": 1.60735285282135, + "learning_rate": 1.5857078288359845e-05, + "loss": 0.8576, + "step": 8374 + }, + { + "epoch": 0.37080902781621156, + "grad_norm": 1.3611531257629395, + "learning_rate": 1.5855825596032287e-05, + "loss": 0.7813, + "step": 8375 + }, + { + "epoch": 0.3708533035210254, + "grad_norm": 1.4741801023483276, + "learning_rate": 1.5854572763843103e-05, + "loss": 0.8728, + "step": 8376 + }, + { + "epoch": 0.3708975792258393, + "grad_norm": 1.8147022724151611, + "learning_rate": 1.5853319791822216e-05, + "loss": 0.8234, + "step": 8377 + }, + { + "epoch": 0.3709418549306532, + "grad_norm": 1.7485296726226807, + "learning_rate": 1.5852066679999548e-05, + "loss": 0.857, + "step": 8378 + }, + { + "epoch": 0.37098613063546704, + "grad_norm": 1.3382200002670288, + "learning_rate": 1.5850813428405036e-05, + "loss": 0.7772, + "step": 8379 + }, + { + "epoch": 0.37103040634028095, + "grad_norm": 1.2928757667541504, + "learning_rate": 1.5849560037068607e-05, + "loss": 0.788, + "step": 8380 + }, + { + "epoch": 0.3710746820450948, + "grad_norm": 1.685642123222351, + "learning_rate": 1.5848306506020205e-05, + "loss": 0.8169, + "step": 8381 + }, + { + "epoch": 0.37111895774990866, + "grad_norm": 1.7154251337051392, + "learning_rate": 1.584705283528976e-05, + "loss": 0.8011, + "step": 8382 + }, + { + "epoch": 0.3711632334547226, + "grad_norm": 1.3986506462097168, + "learning_rate": 1.5845799024907214e-05, + "loss": 0.7342, + "step": 8383 + }, + { + "epoch": 0.37120750915953643, + "grad_norm": 1.2122150659561157, + "learning_rate": 1.5844545074902524e-05, + "loss": 0.7535, + "step": 8384 + }, + { + "epoch": 0.3712517848643503, + "grad_norm": 1.4064075946807861, + "learning_rate": 1.5843290985305632e-05, + "loss": 0.7194, + "step": 8385 + }, + { + "epoch": 0.3712960605691642, + "grad_norm": 1.3392841815948486, + "learning_rate": 1.5842036756146492e-05, + "loss": 0.8085, + "step": 8386 + }, + { + "epoch": 0.37134033627397806, + "grad_norm": 1.5471115112304688, + "learning_rate": 1.5840782387455066e-05, + "loss": 0.7843, + "step": 8387 + }, + { + "epoch": 0.3713846119787919, + "grad_norm": 1.2740387916564941, + "learning_rate": 1.5839527879261304e-05, + "loss": 0.7586, + "step": 8388 + }, + { + "epoch": 0.3714288876836058, + "grad_norm": 1.8093715906143188, + "learning_rate": 1.5838273231595172e-05, + "loss": 0.8319, + "step": 8389 + }, + { + "epoch": 0.3714731633884197, + "grad_norm": 1.5449129343032837, + "learning_rate": 1.5837018444486638e-05, + "loss": 0.8586, + "step": 8390 + }, + { + "epoch": 0.3715174390932336, + "grad_norm": 1.5077475309371948, + "learning_rate": 1.5835763517965676e-05, + "loss": 0.7957, + "step": 8391 + }, + { + "epoch": 0.37156171479804745, + "grad_norm": 1.3519705533981323, + "learning_rate": 1.5834508452062247e-05, + "loss": 0.7601, + "step": 8392 + }, + { + "epoch": 0.3716059905028613, + "grad_norm": 1.564413070678711, + "learning_rate": 1.583325324680634e-05, + "loss": 0.7687, + "step": 8393 + }, + { + "epoch": 0.3716502662076752, + "grad_norm": 1.7381504774093628, + "learning_rate": 1.5831997902227923e-05, + "loss": 0.8301, + "step": 8394 + }, + { + "epoch": 0.3716945419124891, + "grad_norm": 1.1419227123260498, + "learning_rate": 1.583074241835699e-05, + "loss": 0.7312, + "step": 8395 + }, + { + "epoch": 0.37173881761730293, + "grad_norm": 1.162247896194458, + "learning_rate": 1.582948679522352e-05, + "loss": 0.7352, + "step": 8396 + }, + { + "epoch": 0.37178309332211684, + "grad_norm": 1.3820626735687256, + "learning_rate": 1.5828231032857503e-05, + "loss": 0.7591, + "step": 8397 + }, + { + "epoch": 0.3718273690269307, + "grad_norm": 1.582988977432251, + "learning_rate": 1.5826975131288936e-05, + "loss": 0.7834, + "step": 8398 + }, + { + "epoch": 0.37187164473174456, + "grad_norm": 1.2938326597213745, + "learning_rate": 1.582571909054781e-05, + "loss": 0.8024, + "step": 8399 + }, + { + "epoch": 0.37191592043655847, + "grad_norm": 1.6508210897445679, + "learning_rate": 1.582446291066413e-05, + "loss": 0.8978, + "step": 8400 + }, + { + "epoch": 0.3719601961413723, + "grad_norm": 1.3116881847381592, + "learning_rate": 1.5823206591667893e-05, + "loss": 0.7721, + "step": 8401 + }, + { + "epoch": 0.3720044718461862, + "grad_norm": 1.6118375062942505, + "learning_rate": 1.582195013358911e-05, + "loss": 0.7367, + "step": 8402 + }, + { + "epoch": 0.3720487475510001, + "grad_norm": 1.3993898630142212, + "learning_rate": 1.5820693536457787e-05, + "loss": 0.7141, + "step": 8403 + }, + { + "epoch": 0.37209302325581395, + "grad_norm": 1.2403615713119507, + "learning_rate": 1.5819436800303937e-05, + "loss": 0.7706, + "step": 8404 + }, + { + "epoch": 0.3721372989606278, + "grad_norm": 1.1587105989456177, + "learning_rate": 1.581817992515758e-05, + "loss": 0.6827, + "step": 8405 + }, + { + "epoch": 0.3721815746654417, + "grad_norm": 1.410178542137146, + "learning_rate": 1.581692291104873e-05, + "loss": 0.7908, + "step": 8406 + }, + { + "epoch": 0.3722258503702556, + "grad_norm": 1.7119903564453125, + "learning_rate": 1.5815665758007415e-05, + "loss": 0.8086, + "step": 8407 + }, + { + "epoch": 0.3722701260750695, + "grad_norm": 1.2748355865478516, + "learning_rate": 1.581440846606366e-05, + "loss": 0.7431, + "step": 8408 + }, + { + "epoch": 0.37231440177988334, + "grad_norm": 1.4133018255233765, + "learning_rate": 1.581315103524749e-05, + "loss": 0.876, + "step": 8409 + }, + { + "epoch": 0.3723586774846972, + "grad_norm": 1.2740628719329834, + "learning_rate": 1.5811893465588943e-05, + "loss": 0.7177, + "step": 8410 + }, + { + "epoch": 0.3724029531895111, + "grad_norm": 1.53304922580719, + "learning_rate": 1.5810635757118052e-05, + "loss": 0.7952, + "step": 8411 + }, + { + "epoch": 0.37244722889432497, + "grad_norm": 1.3034719228744507, + "learning_rate": 1.5809377909864857e-05, + "loss": 0.852, + "step": 8412 + }, + { + "epoch": 0.3724915045991388, + "grad_norm": 1.4391027688980103, + "learning_rate": 1.58081199238594e-05, + "loss": 0.8244, + "step": 8413 + }, + { + "epoch": 0.37253578030395273, + "grad_norm": 1.3305230140686035, + "learning_rate": 1.580686179913173e-05, + "loss": 0.7019, + "step": 8414 + }, + { + "epoch": 0.3725800560087666, + "grad_norm": 1.374311923980713, + "learning_rate": 1.580560353571189e-05, + "loss": 0.6966, + "step": 8415 + }, + { + "epoch": 0.37262433171358045, + "grad_norm": 1.4013007879257202, + "learning_rate": 1.580434513362994e-05, + "loss": 0.7426, + "step": 8416 + }, + { + "epoch": 0.37266860741839436, + "grad_norm": 1.2782270908355713, + "learning_rate": 1.580308659291593e-05, + "loss": 0.7419, + "step": 8417 + }, + { + "epoch": 0.3727128831232082, + "grad_norm": 1.446852684020996, + "learning_rate": 1.5801827913599924e-05, + "loss": 0.8125, + "step": 8418 + }, + { + "epoch": 0.3727571588280221, + "grad_norm": 2.163313865661621, + "learning_rate": 1.5800569095711983e-05, + "loss": 0.8915, + "step": 8419 + }, + { + "epoch": 0.372801434532836, + "grad_norm": 1.7213433980941772, + "learning_rate": 1.579931013928217e-05, + "loss": 0.8139, + "step": 8420 + }, + { + "epoch": 0.37284571023764984, + "grad_norm": 1.6336934566497803, + "learning_rate": 1.579805104434056e-05, + "loss": 0.8641, + "step": 8421 + }, + { + "epoch": 0.3728899859424637, + "grad_norm": 1.3138762712478638, + "learning_rate": 1.5796791810917217e-05, + "loss": 0.7349, + "step": 8422 + }, + { + "epoch": 0.3729342616472776, + "grad_norm": 1.4557299613952637, + "learning_rate": 1.5795532439042224e-05, + "loss": 0.7775, + "step": 8423 + }, + { + "epoch": 0.37297853735209147, + "grad_norm": 1.4873580932617188, + "learning_rate": 1.5794272928745656e-05, + "loss": 0.8038, + "step": 8424 + }, + { + "epoch": 0.3730228130569053, + "grad_norm": 1.2806943655014038, + "learning_rate": 1.5793013280057602e-05, + "loss": 0.7721, + "step": 8425 + }, + { + "epoch": 0.37306708876171923, + "grad_norm": 1.2338478565216064, + "learning_rate": 1.5791753493008136e-05, + "loss": 0.7344, + "step": 8426 + }, + { + "epoch": 0.3731113644665331, + "grad_norm": 1.474474549293518, + "learning_rate": 1.5790493567627357e-05, + "loss": 0.6766, + "step": 8427 + }, + { + "epoch": 0.373155640171347, + "grad_norm": 1.4515329599380493, + "learning_rate": 1.5789233503945354e-05, + "loss": 0.7577, + "step": 8428 + }, + { + "epoch": 0.37319991587616086, + "grad_norm": 2.0173888206481934, + "learning_rate": 1.578797330199222e-05, + "loss": 0.8173, + "step": 8429 + }, + { + "epoch": 0.3732441915809747, + "grad_norm": 1.336864709854126, + "learning_rate": 1.578671296179806e-05, + "loss": 0.7558, + "step": 8430 + }, + { + "epoch": 0.3732884672857886, + "grad_norm": 1.3095312118530273, + "learning_rate": 1.578545248339297e-05, + "loss": 0.7224, + "step": 8431 + }, + { + "epoch": 0.3733327429906025, + "grad_norm": 1.3829165697097778, + "learning_rate": 1.578419186680706e-05, + "loss": 0.7754, + "step": 8432 + }, + { + "epoch": 0.37337701869541634, + "grad_norm": 1.7836450338363647, + "learning_rate": 1.5782931112070437e-05, + "loss": 0.8954, + "step": 8433 + }, + { + "epoch": 0.37342129440023025, + "grad_norm": 1.3444204330444336, + "learning_rate": 1.578167021921321e-05, + "loss": 0.7821, + "step": 8434 + }, + { + "epoch": 0.3734655701050441, + "grad_norm": 1.7948873043060303, + "learning_rate": 1.57804091882655e-05, + "loss": 0.8796, + "step": 8435 + }, + { + "epoch": 0.37350984580985797, + "grad_norm": 1.655774712562561, + "learning_rate": 1.5779148019257424e-05, + "loss": 0.7794, + "step": 8436 + }, + { + "epoch": 0.3735541215146719, + "grad_norm": 1.5499197244644165, + "learning_rate": 1.57778867122191e-05, + "loss": 0.7827, + "step": 8437 + }, + { + "epoch": 0.37359839721948573, + "grad_norm": 1.5945626497268677, + "learning_rate": 1.577662526718066e-05, + "loss": 0.8508, + "step": 8438 + }, + { + "epoch": 0.3736426729242996, + "grad_norm": 1.8212482929229736, + "learning_rate": 1.5775363684172225e-05, + "loss": 0.7906, + "step": 8439 + }, + { + "epoch": 0.3736869486291135, + "grad_norm": 1.2117310762405396, + "learning_rate": 1.5774101963223934e-05, + "loss": 0.7076, + "step": 8440 + }, + { + "epoch": 0.37373122433392736, + "grad_norm": 2.05936861038208, + "learning_rate": 1.577284010436592e-05, + "loss": 0.7893, + "step": 8441 + }, + { + "epoch": 0.3737755000387412, + "grad_norm": 1.3226990699768066, + "learning_rate": 1.5771578107628312e-05, + "loss": 0.7782, + "step": 8442 + }, + { + "epoch": 0.3738197757435551, + "grad_norm": 1.525501012802124, + "learning_rate": 1.577031597304127e-05, + "loss": 0.7717, + "step": 8443 + }, + { + "epoch": 0.373864051448369, + "grad_norm": 1.3732315301895142, + "learning_rate": 1.5769053700634923e-05, + "loss": 0.8345, + "step": 8444 + }, + { + "epoch": 0.3739083271531829, + "grad_norm": 1.5303223133087158, + "learning_rate": 1.5767791290439433e-05, + "loss": 0.7948, + "step": 8445 + }, + { + "epoch": 0.37395260285799675, + "grad_norm": 1.285198450088501, + "learning_rate": 1.576652874248494e-05, + "loss": 0.7564, + "step": 8446 + }, + { + "epoch": 0.3739968785628106, + "grad_norm": 1.5195027589797974, + "learning_rate": 1.5765266056801603e-05, + "loss": 0.7426, + "step": 8447 + }, + { + "epoch": 0.3740411542676245, + "grad_norm": 1.3927022218704224, + "learning_rate": 1.576400323341958e-05, + "loss": 0.9143, + "step": 8448 + }, + { + "epoch": 0.3740854299724384, + "grad_norm": 1.603652834892273, + "learning_rate": 1.576274027236903e-05, + "loss": 0.7875, + "step": 8449 + }, + { + "epoch": 0.37412970567725223, + "grad_norm": 1.328068494796753, + "learning_rate": 1.576147717368013e-05, + "loss": 0.7921, + "step": 8450 + }, + { + "epoch": 0.37417398138206615, + "grad_norm": 1.550974726676941, + "learning_rate": 1.5760213937383032e-05, + "loss": 0.7728, + "step": 8451 + }, + { + "epoch": 0.37421825708688, + "grad_norm": 1.5922682285308838, + "learning_rate": 1.5758950563507915e-05, + "loss": 0.8843, + "step": 8452 + }, + { + "epoch": 0.37426253279169386, + "grad_norm": 1.385585069656372, + "learning_rate": 1.5757687052084955e-05, + "loss": 0.7083, + "step": 8453 + }, + { + "epoch": 0.37430680849650777, + "grad_norm": 1.2086981534957886, + "learning_rate": 1.5756423403144325e-05, + "loss": 0.7581, + "step": 8454 + }, + { + "epoch": 0.3743510842013216, + "grad_norm": 1.354966640472412, + "learning_rate": 1.5755159616716214e-05, + "loss": 0.7414, + "step": 8455 + }, + { + "epoch": 0.3743953599061355, + "grad_norm": 1.1387536525726318, + "learning_rate": 1.5753895692830797e-05, + "loss": 0.7009, + "step": 8456 + }, + { + "epoch": 0.3744396356109494, + "grad_norm": 1.6789597272872925, + "learning_rate": 1.5752631631518265e-05, + "loss": 0.8151, + "step": 8457 + }, + { + "epoch": 0.37448391131576325, + "grad_norm": 1.3639494180679321, + "learning_rate": 1.5751367432808814e-05, + "loss": 0.8117, + "step": 8458 + }, + { + "epoch": 0.3745281870205771, + "grad_norm": 1.25895357131958, + "learning_rate": 1.5750103096732635e-05, + "loss": 0.769, + "step": 8459 + }, + { + "epoch": 0.374572462725391, + "grad_norm": 1.2177733182907104, + "learning_rate": 1.5748838623319923e-05, + "loss": 0.7074, + "step": 8460 + }, + { + "epoch": 0.3746167384302049, + "grad_norm": 1.7742966413497925, + "learning_rate": 1.5747574012600883e-05, + "loss": 0.881, + "step": 8461 + }, + { + "epoch": 0.3746610141350188, + "grad_norm": 1.5622756481170654, + "learning_rate": 1.574630926460572e-05, + "loss": 0.8148, + "step": 8462 + }, + { + "epoch": 0.37470528983983264, + "grad_norm": 1.3826309442520142, + "learning_rate": 1.5745044379364637e-05, + "loss": 0.7901, + "step": 8463 + }, + { + "epoch": 0.3747495655446465, + "grad_norm": 1.2077821493148804, + "learning_rate": 1.574377935690784e-05, + "loss": 0.7605, + "step": 8464 + }, + { + "epoch": 0.3747938412494604, + "grad_norm": 1.2901101112365723, + "learning_rate": 1.5742514197265563e-05, + "loss": 0.8073, + "step": 8465 + }, + { + "epoch": 0.37483811695427427, + "grad_norm": 1.2006222009658813, + "learning_rate": 1.5741248900468e-05, + "loss": 0.7237, + "step": 8466 + }, + { + "epoch": 0.3748823926590881, + "grad_norm": 1.2436474561691284, + "learning_rate": 1.5739983466545387e-05, + "loss": 0.7684, + "step": 8467 + }, + { + "epoch": 0.37492666836390204, + "grad_norm": 1.486411213874817, + "learning_rate": 1.573871789552794e-05, + "loss": 0.8537, + "step": 8468 + }, + { + "epoch": 0.3749709440687159, + "grad_norm": 1.2154860496520996, + "learning_rate": 1.5737452187445886e-05, + "loss": 0.6753, + "step": 8469 + }, + { + "epoch": 0.37501521977352975, + "grad_norm": 1.3414620161056519, + "learning_rate": 1.5736186342329462e-05, + "loss": 0.8072, + "step": 8470 + }, + { + "epoch": 0.37505949547834366, + "grad_norm": 1.3429539203643799, + "learning_rate": 1.5734920360208898e-05, + "loss": 0.8065, + "step": 8471 + }, + { + "epoch": 0.3751037711831575, + "grad_norm": 1.3684203624725342, + "learning_rate": 1.573365424111443e-05, + "loss": 0.7997, + "step": 8472 + }, + { + "epoch": 0.3751480468879714, + "grad_norm": 1.4180487394332886, + "learning_rate": 1.57323879850763e-05, + "loss": 0.7768, + "step": 8473 + }, + { + "epoch": 0.3751923225927853, + "grad_norm": 1.3622990846633911, + "learning_rate": 1.5731121592124745e-05, + "loss": 0.7697, + "step": 8474 + }, + { + "epoch": 0.37523659829759914, + "grad_norm": 1.4280518293380737, + "learning_rate": 1.5729855062290024e-05, + "loss": 0.7428, + "step": 8475 + }, + { + "epoch": 0.375280874002413, + "grad_norm": 1.9392445087432861, + "learning_rate": 1.5728588395602375e-05, + "loss": 0.7658, + "step": 8476 + }, + { + "epoch": 0.3753251497072269, + "grad_norm": 1.4761958122253418, + "learning_rate": 1.572732159209206e-05, + "loss": 0.8259, + "step": 8477 + }, + { + "epoch": 0.37536942541204077, + "grad_norm": 1.3148066997528076, + "learning_rate": 1.572605465178933e-05, + "loss": 0.8054, + "step": 8478 + }, + { + "epoch": 0.3754137011168547, + "grad_norm": 1.3076516389846802, + "learning_rate": 1.572478757472445e-05, + "loss": 0.7392, + "step": 8479 + }, + { + "epoch": 0.37545797682166854, + "grad_norm": 1.32835054397583, + "learning_rate": 1.572352036092768e-05, + "loss": 0.7339, + "step": 8480 + }, + { + "epoch": 0.3755022525264824, + "grad_norm": 1.3420668840408325, + "learning_rate": 1.572225301042928e-05, + "loss": 0.8216, + "step": 8481 + }, + { + "epoch": 0.3755465282312963, + "grad_norm": 1.7459123134613037, + "learning_rate": 1.5720985523259533e-05, + "loss": 0.7945, + "step": 8482 + }, + { + "epoch": 0.37559080393611016, + "grad_norm": 1.237511157989502, + "learning_rate": 1.5719717899448702e-05, + "loss": 0.7311, + "step": 8483 + }, + { + "epoch": 0.375635079640924, + "grad_norm": 1.514832854270935, + "learning_rate": 1.5718450139027067e-05, + "loss": 0.8961, + "step": 8484 + }, + { + "epoch": 0.37567935534573793, + "grad_norm": 1.529245376586914, + "learning_rate": 1.5717182242024906e-05, + "loss": 0.808, + "step": 8485 + }, + { + "epoch": 0.3757236310505518, + "grad_norm": 1.476830005645752, + "learning_rate": 1.5715914208472503e-05, + "loss": 0.784, + "step": 8486 + }, + { + "epoch": 0.37576790675536564, + "grad_norm": 1.4552651643753052, + "learning_rate": 1.5714646038400143e-05, + "loss": 0.8104, + "step": 8487 + }, + { + "epoch": 0.37581218246017956, + "grad_norm": 2.147230386734009, + "learning_rate": 1.5713377731838115e-05, + "loss": 0.7918, + "step": 8488 + }, + { + "epoch": 0.3758564581649934, + "grad_norm": 1.576159954071045, + "learning_rate": 1.5712109288816713e-05, + "loss": 0.8327, + "step": 8489 + }, + { + "epoch": 0.37590073386980727, + "grad_norm": 1.2377007007598877, + "learning_rate": 1.5710840709366232e-05, + "loss": 0.7459, + "step": 8490 + }, + { + "epoch": 0.3759450095746212, + "grad_norm": 1.2748514413833618, + "learning_rate": 1.570957199351697e-05, + "loss": 0.6938, + "step": 8491 + }, + { + "epoch": 0.37598928527943504, + "grad_norm": 1.5056613683700562, + "learning_rate": 1.5708303141299226e-05, + "loss": 0.7725, + "step": 8492 + }, + { + "epoch": 0.3760335609842489, + "grad_norm": 1.7130346298217773, + "learning_rate": 1.5707034152743314e-05, + "loss": 0.9211, + "step": 8493 + }, + { + "epoch": 0.3760778366890628, + "grad_norm": 1.3752155303955078, + "learning_rate": 1.5705765027879536e-05, + "loss": 0.8229, + "step": 8494 + }, + { + "epoch": 0.37612211239387666, + "grad_norm": 1.452157974243164, + "learning_rate": 1.5704495766738206e-05, + "loss": 0.7217, + "step": 8495 + }, + { + "epoch": 0.3761663880986905, + "grad_norm": 1.3217277526855469, + "learning_rate": 1.5703226369349642e-05, + "loss": 0.7514, + "step": 8496 + }, + { + "epoch": 0.37621066380350443, + "grad_norm": 1.3390449285507202, + "learning_rate": 1.570195683574416e-05, + "loss": 0.7878, + "step": 8497 + }, + { + "epoch": 0.3762549395083183, + "grad_norm": 2.9014458656311035, + "learning_rate": 1.570068716595208e-05, + "loss": 0.7494, + "step": 8498 + }, + { + "epoch": 0.3762992152131322, + "grad_norm": 1.5087465047836304, + "learning_rate": 1.5699417360003725e-05, + "loss": 0.8176, + "step": 8499 + }, + { + "epoch": 0.37634349091794606, + "grad_norm": 1.2986621856689453, + "learning_rate": 1.5698147417929434e-05, + "loss": 0.8031, + "step": 8500 + }, + { + "epoch": 0.3763877666227599, + "grad_norm": 1.2764720916748047, + "learning_rate": 1.5696877339759527e-05, + "loss": 0.7568, + "step": 8501 + }, + { + "epoch": 0.3764320423275738, + "grad_norm": 1.4225757122039795, + "learning_rate": 1.5695607125524345e-05, + "loss": 0.7782, + "step": 8502 + }, + { + "epoch": 0.3764763180323877, + "grad_norm": 1.2532140016555786, + "learning_rate": 1.569433677525422e-05, + "loss": 0.7715, + "step": 8503 + }, + { + "epoch": 0.37652059373720154, + "grad_norm": 1.3433959484100342, + "learning_rate": 1.5693066288979504e-05, + "loss": 0.816, + "step": 8504 + }, + { + "epoch": 0.37656486944201545, + "grad_norm": 1.5928879976272583, + "learning_rate": 1.569179566673053e-05, + "loss": 0.773, + "step": 8505 + }, + { + "epoch": 0.3766091451468293, + "grad_norm": 1.2963597774505615, + "learning_rate": 1.5690524908537653e-05, + "loss": 0.8091, + "step": 8506 + }, + { + "epoch": 0.37665342085164316, + "grad_norm": 1.1425484418869019, + "learning_rate": 1.5689254014431225e-05, + "loss": 0.6784, + "step": 8507 + }, + { + "epoch": 0.3766976965564571, + "grad_norm": 1.5848701000213623, + "learning_rate": 1.5687982984441594e-05, + "loss": 0.7237, + "step": 8508 + }, + { + "epoch": 0.37674197226127093, + "grad_norm": 1.3954906463623047, + "learning_rate": 1.5686711818599122e-05, + "loss": 0.7787, + "step": 8509 + }, + { + "epoch": 0.3767862479660848, + "grad_norm": 1.2062890529632568, + "learning_rate": 1.568544051693417e-05, + "loss": 0.7426, + "step": 8510 + }, + { + "epoch": 0.3768305236708987, + "grad_norm": 1.404437780380249, + "learning_rate": 1.5684169079477095e-05, + "loss": 0.8878, + "step": 8511 + }, + { + "epoch": 0.37687479937571255, + "grad_norm": 2.0216147899627686, + "learning_rate": 1.5682897506258274e-05, + "loss": 0.898, + "step": 8512 + }, + { + "epoch": 0.3769190750805264, + "grad_norm": 1.2149152755737305, + "learning_rate": 1.5681625797308072e-05, + "loss": 0.7667, + "step": 8513 + }, + { + "epoch": 0.3769633507853403, + "grad_norm": 1.2297004461288452, + "learning_rate": 1.5680353952656863e-05, + "loss": 0.7676, + "step": 8514 + }, + { + "epoch": 0.3770076264901542, + "grad_norm": 1.7523260116577148, + "learning_rate": 1.5679081972335028e-05, + "loss": 0.8723, + "step": 8515 + }, + { + "epoch": 0.3770519021949681, + "grad_norm": 1.258960485458374, + "learning_rate": 1.567780985637294e-05, + "loss": 0.7015, + "step": 8516 + }, + { + "epoch": 0.37709617789978195, + "grad_norm": 1.331887125968933, + "learning_rate": 1.5676537604800992e-05, + "loss": 0.7491, + "step": 8517 + }, + { + "epoch": 0.3771404536045958, + "grad_norm": 1.2209539413452148, + "learning_rate": 1.567526521764956e-05, + "loss": 0.7491, + "step": 8518 + }, + { + "epoch": 0.3771847293094097, + "grad_norm": 1.3721891641616821, + "learning_rate": 1.5673992694949043e-05, + "loss": 0.7213, + "step": 8519 + }, + { + "epoch": 0.3772290050142236, + "grad_norm": 1.5105817317962646, + "learning_rate": 1.567272003672983e-05, + "loss": 0.8031, + "step": 8520 + }, + { + "epoch": 0.37727328071903743, + "grad_norm": 1.2883164882659912, + "learning_rate": 1.5671447243022317e-05, + "loss": 0.7611, + "step": 8521 + }, + { + "epoch": 0.37731755642385134, + "grad_norm": 1.5082136392593384, + "learning_rate": 1.56701743138569e-05, + "loss": 0.8064, + "step": 8522 + }, + { + "epoch": 0.3773618321286652, + "grad_norm": 1.2118890285491943, + "learning_rate": 1.5668901249263996e-05, + "loss": 0.7422, + "step": 8523 + }, + { + "epoch": 0.37740610783347905, + "grad_norm": 1.4195486307144165, + "learning_rate": 1.5667628049273992e-05, + "loss": 0.8634, + "step": 8524 + }, + { + "epoch": 0.37745038353829297, + "grad_norm": 1.3365899324417114, + "learning_rate": 1.5666354713917312e-05, + "loss": 0.7522, + "step": 8525 + }, + { + "epoch": 0.3774946592431068, + "grad_norm": 1.3021641969680786, + "learning_rate": 1.5665081243224364e-05, + "loss": 0.7733, + "step": 8526 + }, + { + "epoch": 0.3775389349479207, + "grad_norm": 1.5242875814437866, + "learning_rate": 1.566380763722556e-05, + "loss": 0.8487, + "step": 8527 + }, + { + "epoch": 0.3775832106527346, + "grad_norm": 1.676377296447754, + "learning_rate": 1.5662533895951324e-05, + "loss": 0.815, + "step": 8528 + }, + { + "epoch": 0.37762748635754845, + "grad_norm": 1.493114709854126, + "learning_rate": 1.5661260019432077e-05, + "loss": 0.7575, + "step": 8529 + }, + { + "epoch": 0.3776717620623623, + "grad_norm": 1.1515663862228394, + "learning_rate": 1.5659986007698243e-05, + "loss": 0.7265, + "step": 8530 + }, + { + "epoch": 0.3777160377671762, + "grad_norm": 1.5338307619094849, + "learning_rate": 1.565871186078025e-05, + "loss": 0.8405, + "step": 8531 + }, + { + "epoch": 0.3777603134719901, + "grad_norm": 1.4886823892593384, + "learning_rate": 1.5657437578708537e-05, + "loss": 0.8159, + "step": 8532 + }, + { + "epoch": 0.377804589176804, + "grad_norm": 1.3022526502609253, + "learning_rate": 1.565616316151353e-05, + "loss": 0.7655, + "step": 8533 + }, + { + "epoch": 0.37784886488161784, + "grad_norm": 1.5720016956329346, + "learning_rate": 1.565488860922567e-05, + "loss": 0.7572, + "step": 8534 + }, + { + "epoch": 0.3778931405864317, + "grad_norm": 1.8392574787139893, + "learning_rate": 1.5653613921875404e-05, + "loss": 0.8478, + "step": 8535 + }, + { + "epoch": 0.3779374162912456, + "grad_norm": 1.638364315032959, + "learning_rate": 1.565233909949317e-05, + "loss": 0.8739, + "step": 8536 + }, + { + "epoch": 0.37798169199605947, + "grad_norm": 1.4458814859390259, + "learning_rate": 1.565106414210942e-05, + "loss": 0.6986, + "step": 8537 + }, + { + "epoch": 0.3780259677008733, + "grad_norm": 1.2285727262496948, + "learning_rate": 1.5649789049754604e-05, + "loss": 0.7022, + "step": 8538 + }, + { + "epoch": 0.37807024340568723, + "grad_norm": 1.6144605875015259, + "learning_rate": 1.5648513822459176e-05, + "loss": 0.8154, + "step": 8539 + }, + { + "epoch": 0.3781145191105011, + "grad_norm": 1.35726797580719, + "learning_rate": 1.5647238460253596e-05, + "loss": 0.7522, + "step": 8540 + }, + { + "epoch": 0.37815879481531495, + "grad_norm": 1.4657657146453857, + "learning_rate": 1.5645962963168322e-05, + "loss": 0.8003, + "step": 8541 + }, + { + "epoch": 0.37820307052012886, + "grad_norm": 1.4255303144454956, + "learning_rate": 1.564468733123382e-05, + "loss": 0.7693, + "step": 8542 + }, + { + "epoch": 0.3782473462249427, + "grad_norm": 1.535044550895691, + "learning_rate": 1.5643411564480557e-05, + "loss": 0.7611, + "step": 8543 + }, + { + "epoch": 0.37829162192975657, + "grad_norm": 1.3487048149108887, + "learning_rate": 1.5642135662939003e-05, + "loss": 0.7301, + "step": 8544 + }, + { + "epoch": 0.3783358976345705, + "grad_norm": 1.7885851860046387, + "learning_rate": 1.564085962663964e-05, + "loss": 0.861, + "step": 8545 + }, + { + "epoch": 0.37838017333938434, + "grad_norm": 1.446026086807251, + "learning_rate": 1.563958345561293e-05, + "loss": 0.839, + "step": 8546 + }, + { + "epoch": 0.3784244490441982, + "grad_norm": 1.4006695747375488, + "learning_rate": 1.563830714988936e-05, + "loss": 0.84, + "step": 8547 + }, + { + "epoch": 0.3784687247490121, + "grad_norm": 1.3747284412384033, + "learning_rate": 1.563703070949942e-05, + "loss": 0.7654, + "step": 8548 + }, + { + "epoch": 0.37851300045382597, + "grad_norm": 1.27386474609375, + "learning_rate": 1.5635754134473586e-05, + "loss": 0.7068, + "step": 8549 + }, + { + "epoch": 0.3785572761586399, + "grad_norm": 1.8542990684509277, + "learning_rate": 1.563447742484236e-05, + "loss": 0.8055, + "step": 8550 + }, + { + "epoch": 0.37860155186345373, + "grad_norm": 1.5176299810409546, + "learning_rate": 1.563320058063622e-05, + "loss": 0.7939, + "step": 8551 + }, + { + "epoch": 0.3786458275682676, + "grad_norm": 1.4782929420471191, + "learning_rate": 1.563192360188568e-05, + "loss": 0.7783, + "step": 8552 + }, + { + "epoch": 0.3786901032730815, + "grad_norm": 1.7094868421554565, + "learning_rate": 1.5630646488621223e-05, + "loss": 0.8689, + "step": 8553 + }, + { + "epoch": 0.37873437897789536, + "grad_norm": 1.543811321258545, + "learning_rate": 1.5629369240873363e-05, + "loss": 0.7827, + "step": 8554 + }, + { + "epoch": 0.3787786546827092, + "grad_norm": 1.4335076808929443, + "learning_rate": 1.5628091858672602e-05, + "loss": 0.8077, + "step": 8555 + }, + { + "epoch": 0.3788229303875231, + "grad_norm": 1.2583765983581543, + "learning_rate": 1.5626814342049447e-05, + "loss": 0.6932, + "step": 8556 + }, + { + "epoch": 0.378867206092337, + "grad_norm": 1.589635968208313, + "learning_rate": 1.5625536691034415e-05, + "loss": 0.7295, + "step": 8557 + }, + { + "epoch": 0.37891148179715084, + "grad_norm": 1.325075626373291, + "learning_rate": 1.562425890565802e-05, + "loss": 0.7101, + "step": 8558 + }, + { + "epoch": 0.37895575750196475, + "grad_norm": 1.308150291442871, + "learning_rate": 1.562298098595078e-05, + "loss": 0.7322, + "step": 8559 + }, + { + "epoch": 0.3790000332067786, + "grad_norm": 1.5089199542999268, + "learning_rate": 1.5621702931943218e-05, + "loss": 0.7889, + "step": 8560 + }, + { + "epoch": 0.37904430891159246, + "grad_norm": 1.334579586982727, + "learning_rate": 1.5620424743665857e-05, + "loss": 0.7742, + "step": 8561 + }, + { + "epoch": 0.3790885846164064, + "grad_norm": 1.1753666400909424, + "learning_rate": 1.561914642114923e-05, + "loss": 0.7099, + "step": 8562 + }, + { + "epoch": 0.37913286032122023, + "grad_norm": 1.4461842775344849, + "learning_rate": 1.5617867964423865e-05, + "loss": 0.7875, + "step": 8563 + }, + { + "epoch": 0.3791771360260341, + "grad_norm": 1.398747205734253, + "learning_rate": 1.5616589373520295e-05, + "loss": 0.7039, + "step": 8564 + }, + { + "epoch": 0.379221411730848, + "grad_norm": 1.369334101676941, + "learning_rate": 1.5615310648469068e-05, + "loss": 0.7471, + "step": 8565 + }, + { + "epoch": 0.37926568743566186, + "grad_norm": 1.2549877166748047, + "learning_rate": 1.561403178930071e-05, + "loss": 0.7193, + "step": 8566 + }, + { + "epoch": 0.37930996314047577, + "grad_norm": 1.6458178758621216, + "learning_rate": 1.561275279604578e-05, + "loss": 0.7635, + "step": 8567 + }, + { + "epoch": 0.3793542388452896, + "grad_norm": 1.1188507080078125, + "learning_rate": 1.5611473668734816e-05, + "loss": 0.7276, + "step": 8568 + }, + { + "epoch": 0.3793985145501035, + "grad_norm": 1.1652333736419678, + "learning_rate": 1.5610194407398377e-05, + "loss": 0.7751, + "step": 8569 + }, + { + "epoch": 0.3794427902549174, + "grad_norm": 1.2833322286605835, + "learning_rate": 1.5608915012067008e-05, + "loss": 0.7301, + "step": 8570 + }, + { + "epoch": 0.37948706595973125, + "grad_norm": 1.150266408920288, + "learning_rate": 1.5607635482771272e-05, + "loss": 0.6905, + "step": 8571 + }, + { + "epoch": 0.3795313416645451, + "grad_norm": 1.318505048751831, + "learning_rate": 1.560635581954173e-05, + "loss": 0.7949, + "step": 8572 + }, + { + "epoch": 0.379575617369359, + "grad_norm": 1.3816852569580078, + "learning_rate": 1.560507602240894e-05, + "loss": 0.751, + "step": 8573 + }, + { + "epoch": 0.3796198930741729, + "grad_norm": 1.1740158796310425, + "learning_rate": 1.5603796091403478e-05, + "loss": 0.6955, + "step": 8574 + }, + { + "epoch": 0.37966416877898673, + "grad_norm": 1.5775688886642456, + "learning_rate": 1.5602516026555905e-05, + "loss": 0.8334, + "step": 8575 + }, + { + "epoch": 0.37970844448380064, + "grad_norm": 1.1838819980621338, + "learning_rate": 1.56012358278968e-05, + "loss": 0.7174, + "step": 8576 + }, + { + "epoch": 0.3797527201886145, + "grad_norm": 1.4140383005142212, + "learning_rate": 1.559995549545674e-05, + "loss": 0.738, + "step": 8577 + }, + { + "epoch": 0.37979699589342836, + "grad_norm": 1.315751314163208, + "learning_rate": 1.55986750292663e-05, + "loss": 0.7604, + "step": 8578 + }, + { + "epoch": 0.37984127159824227, + "grad_norm": 1.2920773029327393, + "learning_rate": 1.5597394429356066e-05, + "loss": 0.7281, + "step": 8579 + }, + { + "epoch": 0.3798855473030561, + "grad_norm": 1.296282172203064, + "learning_rate": 1.5596113695756624e-05, + "loss": 0.7218, + "step": 8580 + }, + { + "epoch": 0.37992982300787, + "grad_norm": 2.0068438053131104, + "learning_rate": 1.559483282849856e-05, + "loss": 0.8065, + "step": 8581 + }, + { + "epoch": 0.3799740987126839, + "grad_norm": 1.2269831895828247, + "learning_rate": 1.5593551827612476e-05, + "loss": 0.7286, + "step": 8582 + }, + { + "epoch": 0.38001837441749775, + "grad_norm": 1.5854369401931763, + "learning_rate": 1.559227069312896e-05, + "loss": 0.8299, + "step": 8583 + }, + { + "epoch": 0.3800626501223116, + "grad_norm": 1.3708117008209229, + "learning_rate": 1.559098942507861e-05, + "loss": 0.7726, + "step": 8584 + }, + { + "epoch": 0.3801069258271255, + "grad_norm": 1.3205851316452026, + "learning_rate": 1.5589708023492028e-05, + "loss": 0.7842, + "step": 8585 + }, + { + "epoch": 0.3801512015319394, + "grad_norm": 1.344873309135437, + "learning_rate": 1.558842648839982e-05, + "loss": 0.6945, + "step": 8586 + }, + { + "epoch": 0.3801954772367533, + "grad_norm": 1.588883638381958, + "learning_rate": 1.55871448198326e-05, + "loss": 0.7353, + "step": 8587 + }, + { + "epoch": 0.38023975294156714, + "grad_norm": 1.2208036184310913, + "learning_rate": 1.558586301782097e-05, + "loss": 0.7514, + "step": 8588 + }, + { + "epoch": 0.380284028646381, + "grad_norm": 1.3011720180511475, + "learning_rate": 1.5584581082395555e-05, + "loss": 0.7564, + "step": 8589 + }, + { + "epoch": 0.3803283043511949, + "grad_norm": 1.454485535621643, + "learning_rate": 1.558329901358697e-05, + "loss": 0.7559, + "step": 8590 + }, + { + "epoch": 0.38037258005600877, + "grad_norm": 1.1346988677978516, + "learning_rate": 1.5582016811425828e-05, + "loss": 0.675, + "step": 8591 + }, + { + "epoch": 0.3804168557608226, + "grad_norm": 1.5189658403396606, + "learning_rate": 1.5580734475942762e-05, + "loss": 0.8131, + "step": 8592 + }, + { + "epoch": 0.38046113146563654, + "grad_norm": 1.1915535926818848, + "learning_rate": 1.5579452007168397e-05, + "loss": 0.7277, + "step": 8593 + }, + { + "epoch": 0.3805054071704504, + "grad_norm": 1.4543628692626953, + "learning_rate": 1.5578169405133363e-05, + "loss": 0.7889, + "step": 8594 + }, + { + "epoch": 0.38054968287526425, + "grad_norm": 1.3205080032348633, + "learning_rate": 1.5576886669868297e-05, + "loss": 0.7077, + "step": 8595 + }, + { + "epoch": 0.38059395858007816, + "grad_norm": 1.3038145303726196, + "learning_rate": 1.557560380140383e-05, + "loss": 0.7546, + "step": 8596 + }, + { + "epoch": 0.380638234284892, + "grad_norm": 1.10756254196167, + "learning_rate": 1.5574320799770612e-05, + "loss": 0.7586, + "step": 8597 + }, + { + "epoch": 0.3806825099897059, + "grad_norm": 1.2916353940963745, + "learning_rate": 1.5573037664999274e-05, + "loss": 0.767, + "step": 8598 + }, + { + "epoch": 0.3807267856945198, + "grad_norm": 1.4131616353988647, + "learning_rate": 1.557175439712047e-05, + "loss": 0.797, + "step": 8599 + }, + { + "epoch": 0.38077106139933364, + "grad_norm": 1.2720515727996826, + "learning_rate": 1.5570470996164852e-05, + "loss": 0.7349, + "step": 8600 + }, + { + "epoch": 0.3808153371041475, + "grad_norm": 2.1703197956085205, + "learning_rate": 1.5569187462163068e-05, + "loss": 0.8984, + "step": 8601 + }, + { + "epoch": 0.3808596128089614, + "grad_norm": 5.373884677886963, + "learning_rate": 1.5567903795145778e-05, + "loss": 0.7062, + "step": 8602 + }, + { + "epoch": 0.38090388851377527, + "grad_norm": 1.2517110109329224, + "learning_rate": 1.5566619995143637e-05, + "loss": 0.7976, + "step": 8603 + }, + { + "epoch": 0.3809481642185892, + "grad_norm": 1.2053766250610352, + "learning_rate": 1.556533606218731e-05, + "loss": 0.6952, + "step": 8604 + }, + { + "epoch": 0.38099243992340304, + "grad_norm": 1.2190603017807007, + "learning_rate": 1.5564051996307466e-05, + "loss": 0.7639, + "step": 8605 + }, + { + "epoch": 0.3810367156282169, + "grad_norm": 1.49390709400177, + "learning_rate": 1.5562767797534767e-05, + "loss": 0.7566, + "step": 8606 + }, + { + "epoch": 0.3810809913330308, + "grad_norm": 3.2021145820617676, + "learning_rate": 1.556148346589989e-05, + "loss": 0.86, + "step": 8607 + }, + { + "epoch": 0.38112526703784466, + "grad_norm": 1.23550546169281, + "learning_rate": 1.5560199001433506e-05, + "loss": 0.6941, + "step": 8608 + }, + { + "epoch": 0.3811695427426585, + "grad_norm": 1.4031312465667725, + "learning_rate": 1.55589144041663e-05, + "loss": 0.7575, + "step": 8609 + }, + { + "epoch": 0.38121381844747243, + "grad_norm": 1.2992898225784302, + "learning_rate": 1.555762967412895e-05, + "loss": 0.7157, + "step": 8610 + }, + { + "epoch": 0.3812580941522863, + "grad_norm": 1.3762168884277344, + "learning_rate": 1.5556344811352134e-05, + "loss": 0.8123, + "step": 8611 + }, + { + "epoch": 0.38130236985710014, + "grad_norm": 1.6936553716659546, + "learning_rate": 1.5555059815866554e-05, + "loss": 0.8141, + "step": 8612 + }, + { + "epoch": 0.38134664556191405, + "grad_norm": 1.422463059425354, + "learning_rate": 1.555377468770289e-05, + "loss": 0.7354, + "step": 8613 + }, + { + "epoch": 0.3813909212667279, + "grad_norm": 1.2344293594360352, + "learning_rate": 1.5552489426891843e-05, + "loss": 0.7394, + "step": 8614 + }, + { + "epoch": 0.38143519697154177, + "grad_norm": 1.3914538621902466, + "learning_rate": 1.5551204033464102e-05, + "loss": 0.7804, + "step": 8615 + }, + { + "epoch": 0.3814794726763557, + "grad_norm": 1.2834880352020264, + "learning_rate": 1.5549918507450375e-05, + "loss": 0.6843, + "step": 8616 + }, + { + "epoch": 0.38152374838116954, + "grad_norm": 1.6514958143234253, + "learning_rate": 1.5548632848881363e-05, + "loss": 0.8165, + "step": 8617 + }, + { + "epoch": 0.3815680240859834, + "grad_norm": 1.5249017477035522, + "learning_rate": 1.554734705778777e-05, + "loss": 0.8609, + "step": 8618 + }, + { + "epoch": 0.3816122997907973, + "grad_norm": 1.1804935932159424, + "learning_rate": 1.5546061134200316e-05, + "loss": 0.7301, + "step": 8619 + }, + { + "epoch": 0.38165657549561116, + "grad_norm": 1.4833351373672485, + "learning_rate": 1.55447750781497e-05, + "loss": 0.7866, + "step": 8620 + }, + { + "epoch": 0.3817008512004251, + "grad_norm": 1.282959222793579, + "learning_rate": 1.5543488889666655e-05, + "loss": 0.7772, + "step": 8621 + }, + { + "epoch": 0.38174512690523893, + "grad_norm": 2.021221160888672, + "learning_rate": 1.5542202568781886e-05, + "loss": 0.7395, + "step": 8622 + }, + { + "epoch": 0.3817894026100528, + "grad_norm": 1.5752314329147339, + "learning_rate": 1.5540916115526118e-05, + "loss": 0.739, + "step": 8623 + }, + { + "epoch": 0.3818336783148667, + "grad_norm": 1.6671797037124634, + "learning_rate": 1.5539629529930085e-05, + "loss": 0.7353, + "step": 8624 + }, + { + "epoch": 0.38187795401968055, + "grad_norm": 1.292027473449707, + "learning_rate": 1.5538342812024515e-05, + "loss": 0.7088, + "step": 8625 + }, + { + "epoch": 0.3819222297244944, + "grad_norm": 1.3070356845855713, + "learning_rate": 1.553705596184013e-05, + "loss": 0.7334, + "step": 8626 + }, + { + "epoch": 0.3819665054293083, + "grad_norm": 1.2934448719024658, + "learning_rate": 1.5535768979407673e-05, + "loss": 0.688, + "step": 8627 + }, + { + "epoch": 0.3820107811341222, + "grad_norm": 1.0822155475616455, + "learning_rate": 1.553448186475788e-05, + "loss": 0.6794, + "step": 8628 + }, + { + "epoch": 0.38205505683893604, + "grad_norm": 1.4907959699630737, + "learning_rate": 1.5533194617921498e-05, + "loss": 0.7363, + "step": 8629 + }, + { + "epoch": 0.38209933254374995, + "grad_norm": 1.5218069553375244, + "learning_rate": 1.5531907238929263e-05, + "loss": 0.8034, + "step": 8630 + }, + { + "epoch": 0.3821436082485638, + "grad_norm": 1.299111008644104, + "learning_rate": 1.553061972781193e-05, + "loss": 0.7854, + "step": 8631 + }, + { + "epoch": 0.38218788395337766, + "grad_norm": 1.806548833847046, + "learning_rate": 1.5529332084600253e-05, + "loss": 0.8407, + "step": 8632 + }, + { + "epoch": 0.3822321596581916, + "grad_norm": 2.0359647274017334, + "learning_rate": 1.5528044309324975e-05, + "loss": 0.8547, + "step": 8633 + }, + { + "epoch": 0.38227643536300543, + "grad_norm": 1.2254196405410767, + "learning_rate": 1.552675640201686e-05, + "loss": 0.7571, + "step": 8634 + }, + { + "epoch": 0.3823207110678193, + "grad_norm": 1.6898128986358643, + "learning_rate": 1.5525468362706672e-05, + "loss": 0.7753, + "step": 8635 + }, + { + "epoch": 0.3823649867726332, + "grad_norm": 1.7125006914138794, + "learning_rate": 1.552418019142517e-05, + "loss": 0.7726, + "step": 8636 + }, + { + "epoch": 0.38240926247744705, + "grad_norm": 1.2474859952926636, + "learning_rate": 1.552289188820312e-05, + "loss": 0.7346, + "step": 8637 + }, + { + "epoch": 0.38245353818226097, + "grad_norm": 1.2148469686508179, + "learning_rate": 1.5521603453071295e-05, + "loss": 0.7727, + "step": 8638 + }, + { + "epoch": 0.3824978138870748, + "grad_norm": 1.507218360900879, + "learning_rate": 1.5520314886060467e-05, + "loss": 0.7567, + "step": 8639 + }, + { + "epoch": 0.3825420895918887, + "grad_norm": 1.3655073642730713, + "learning_rate": 1.5519026187201416e-05, + "loss": 0.7031, + "step": 8640 + }, + { + "epoch": 0.3825863652967026, + "grad_norm": 1.0782442092895508, + "learning_rate": 1.5517737356524912e-05, + "loss": 0.6699, + "step": 8641 + }, + { + "epoch": 0.38263064100151645, + "grad_norm": 1.259630560874939, + "learning_rate": 1.551644839406175e-05, + "loss": 0.7277, + "step": 8642 + }, + { + "epoch": 0.3826749167063303, + "grad_norm": 1.3478407859802246, + "learning_rate": 1.551515929984271e-05, + "loss": 0.7088, + "step": 8643 + }, + { + "epoch": 0.3827191924111442, + "grad_norm": 1.2807362079620361, + "learning_rate": 1.5513870073898578e-05, + "loss": 0.7409, + "step": 8644 + }, + { + "epoch": 0.38276346811595807, + "grad_norm": 1.2350767850875854, + "learning_rate": 1.5512580716260147e-05, + "loss": 0.7502, + "step": 8645 + }, + { + "epoch": 0.38280774382077193, + "grad_norm": 1.3050730228424072, + "learning_rate": 1.5511291226958216e-05, + "loss": 0.7831, + "step": 8646 + }, + { + "epoch": 0.38285201952558584, + "grad_norm": 1.3602478504180908, + "learning_rate": 1.5510001606023585e-05, + "loss": 0.7713, + "step": 8647 + }, + { + "epoch": 0.3828962952303997, + "grad_norm": 1.420418381690979, + "learning_rate": 1.5508711853487046e-05, + "loss": 0.7291, + "step": 8648 + }, + { + "epoch": 0.38294057093521355, + "grad_norm": 1.7902579307556152, + "learning_rate": 1.5507421969379414e-05, + "loss": 0.7915, + "step": 8649 + }, + { + "epoch": 0.38298484664002747, + "grad_norm": 1.597590684890747, + "learning_rate": 1.550613195373149e-05, + "loss": 0.7478, + "step": 8650 + }, + { + "epoch": 0.3830291223448413, + "grad_norm": 1.551945686340332, + "learning_rate": 1.5504841806574086e-05, + "loss": 0.8116, + "step": 8651 + }, + { + "epoch": 0.3830733980496552, + "grad_norm": 1.3141437768936157, + "learning_rate": 1.5503551527938022e-05, + "loss": 0.7967, + "step": 8652 + }, + { + "epoch": 0.3831176737544691, + "grad_norm": 1.2941021919250488, + "learning_rate": 1.550226111785411e-05, + "loss": 0.6777, + "step": 8653 + }, + { + "epoch": 0.38316194945928295, + "grad_norm": 1.3879190683364868, + "learning_rate": 1.550097057635317e-05, + "loss": 0.7061, + "step": 8654 + }, + { + "epoch": 0.38320622516409686, + "grad_norm": 1.507630705833435, + "learning_rate": 1.549967990346603e-05, + "loss": 0.8207, + "step": 8655 + }, + { + "epoch": 0.3832505008689107, + "grad_norm": 1.3567031621932983, + "learning_rate": 1.5498389099223507e-05, + "loss": 0.8013, + "step": 8656 + }, + { + "epoch": 0.38329477657372457, + "grad_norm": 1.227010726928711, + "learning_rate": 1.5497098163656444e-05, + "loss": 0.7941, + "step": 8657 + }, + { + "epoch": 0.3833390522785385, + "grad_norm": 1.2053943872451782, + "learning_rate": 1.5495807096795663e-05, + "loss": 0.7686, + "step": 8658 + }, + { + "epoch": 0.38338332798335234, + "grad_norm": 1.3333353996276855, + "learning_rate": 1.5494515898672006e-05, + "loss": 0.7073, + "step": 8659 + }, + { + "epoch": 0.3834276036881662, + "grad_norm": 1.619550108909607, + "learning_rate": 1.549322456931631e-05, + "loss": 0.7938, + "step": 8660 + }, + { + "epoch": 0.3834718793929801, + "grad_norm": 1.5406745672225952, + "learning_rate": 1.5491933108759417e-05, + "loss": 0.7904, + "step": 8661 + }, + { + "epoch": 0.38351615509779396, + "grad_norm": 1.29420804977417, + "learning_rate": 1.5490641517032176e-05, + "loss": 0.7726, + "step": 8662 + }, + { + "epoch": 0.3835604308026078, + "grad_norm": 1.3756695985794067, + "learning_rate": 1.548934979416543e-05, + "loss": 0.7777, + "step": 8663 + }, + { + "epoch": 0.38360470650742173, + "grad_norm": 1.3849220275878906, + "learning_rate": 1.5488057940190038e-05, + "loss": 0.7575, + "step": 8664 + }, + { + "epoch": 0.3836489822122356, + "grad_norm": 1.3444364070892334, + "learning_rate": 1.5486765955136845e-05, + "loss": 0.7457, + "step": 8665 + }, + { + "epoch": 0.38369325791704945, + "grad_norm": 1.9308668375015259, + "learning_rate": 1.548547383903672e-05, + "loss": 0.7545, + "step": 8666 + }, + { + "epoch": 0.38373753362186336, + "grad_norm": 1.4060662984848022, + "learning_rate": 1.5484181591920516e-05, + "loss": 0.8255, + "step": 8667 + }, + { + "epoch": 0.3837818093266772, + "grad_norm": 1.2451534271240234, + "learning_rate": 1.5482889213819102e-05, + "loss": 0.8113, + "step": 8668 + }, + { + "epoch": 0.38382608503149107, + "grad_norm": 1.6730797290802002, + "learning_rate": 1.5481596704763342e-05, + "loss": 0.782, + "step": 8669 + }, + { + "epoch": 0.383870360736305, + "grad_norm": 1.2197513580322266, + "learning_rate": 1.5480304064784108e-05, + "loss": 0.6939, + "step": 8670 + }, + { + "epoch": 0.38391463644111884, + "grad_norm": 1.2078745365142822, + "learning_rate": 1.5479011293912273e-05, + "loss": 0.7141, + "step": 8671 + }, + { + "epoch": 0.3839589121459327, + "grad_norm": 1.3755522966384888, + "learning_rate": 1.5477718392178716e-05, + "loss": 0.7634, + "step": 8672 + }, + { + "epoch": 0.3840031878507466, + "grad_norm": 1.3723050355911255, + "learning_rate": 1.5476425359614314e-05, + "loss": 0.7285, + "step": 8673 + }, + { + "epoch": 0.38404746355556046, + "grad_norm": 2.1236252784729004, + "learning_rate": 1.5475132196249952e-05, + "loss": 0.7951, + "step": 8674 + }, + { + "epoch": 0.3840917392603744, + "grad_norm": 1.6817899942398071, + "learning_rate": 1.5473838902116514e-05, + "loss": 0.8257, + "step": 8675 + }, + { + "epoch": 0.38413601496518823, + "grad_norm": 1.5427312850952148, + "learning_rate": 1.5472545477244893e-05, + "loss": 0.7668, + "step": 8676 + }, + { + "epoch": 0.3841802906700021, + "grad_norm": 1.3700670003890991, + "learning_rate": 1.5471251921665977e-05, + "loss": 0.775, + "step": 8677 + }, + { + "epoch": 0.384224566374816, + "grad_norm": 1.5091320276260376, + "learning_rate": 1.5469958235410664e-05, + "loss": 0.7886, + "step": 8678 + }, + { + "epoch": 0.38426884207962986, + "grad_norm": 1.4363832473754883, + "learning_rate": 1.5468664418509856e-05, + "loss": 0.7952, + "step": 8679 + }, + { + "epoch": 0.3843131177844437, + "grad_norm": 1.432965636253357, + "learning_rate": 1.5467370470994447e-05, + "loss": 0.8355, + "step": 8680 + }, + { + "epoch": 0.3843573934892576, + "grad_norm": 1.490180253982544, + "learning_rate": 1.5466076392895347e-05, + "loss": 0.6951, + "step": 8681 + }, + { + "epoch": 0.3844016691940715, + "grad_norm": 1.1795728206634521, + "learning_rate": 1.5464782184243465e-05, + "loss": 0.7346, + "step": 8682 + }, + { + "epoch": 0.38444594489888534, + "grad_norm": 1.185348629951477, + "learning_rate": 1.5463487845069708e-05, + "loss": 0.7428, + "step": 8683 + }, + { + "epoch": 0.38449022060369925, + "grad_norm": 1.1279083490371704, + "learning_rate": 1.5462193375404995e-05, + "loss": 0.7023, + "step": 8684 + }, + { + "epoch": 0.3845344963085131, + "grad_norm": 1.4174636602401733, + "learning_rate": 1.5460898775280235e-05, + "loss": 0.7646, + "step": 8685 + }, + { + "epoch": 0.38457877201332696, + "grad_norm": 1.6182160377502441, + "learning_rate": 1.545960404472636e-05, + "loss": 0.7798, + "step": 8686 + }, + { + "epoch": 0.3846230477181409, + "grad_norm": 1.388048529624939, + "learning_rate": 1.5458309183774286e-05, + "loss": 0.6785, + "step": 8687 + }, + { + "epoch": 0.38466732342295473, + "grad_norm": 1.4664504528045654, + "learning_rate": 1.545701419245494e-05, + "loss": 0.7793, + "step": 8688 + }, + { + "epoch": 0.3847115991277686, + "grad_norm": 1.1243447065353394, + "learning_rate": 1.5455719070799256e-05, + "loss": 0.693, + "step": 8689 + }, + { + "epoch": 0.3847558748325825, + "grad_norm": 1.9877926111221313, + "learning_rate": 1.5454423818838163e-05, + "loss": 0.9286, + "step": 8690 + }, + { + "epoch": 0.38480015053739636, + "grad_norm": 1.308517575263977, + "learning_rate": 1.5453128436602597e-05, + "loss": 0.7826, + "step": 8691 + }, + { + "epoch": 0.38484442624221027, + "grad_norm": 1.58762788772583, + "learning_rate": 1.5451832924123497e-05, + "loss": 0.796, + "step": 8692 + }, + { + "epoch": 0.3848887019470241, + "grad_norm": 1.4881932735443115, + "learning_rate": 1.545053728143181e-05, + "loss": 0.7139, + "step": 8693 + }, + { + "epoch": 0.384932977651838, + "grad_norm": 1.2226117849349976, + "learning_rate": 1.544924150855848e-05, + "loss": 0.7249, + "step": 8694 + }, + { + "epoch": 0.3849772533566519, + "grad_norm": 1.7913904190063477, + "learning_rate": 1.5447945605534447e-05, + "loss": 0.7984, + "step": 8695 + }, + { + "epoch": 0.38502152906146575, + "grad_norm": 1.4316864013671875, + "learning_rate": 1.5446649572390672e-05, + "loss": 0.7907, + "step": 8696 + }, + { + "epoch": 0.3850658047662796, + "grad_norm": 1.4510387182235718, + "learning_rate": 1.5445353409158107e-05, + "loss": 0.7755, + "step": 8697 + }, + { + "epoch": 0.3851100804710935, + "grad_norm": 1.2850208282470703, + "learning_rate": 1.5444057115867713e-05, + "loss": 0.7498, + "step": 8698 + }, + { + "epoch": 0.3851543561759074, + "grad_norm": 2.0134589672088623, + "learning_rate": 1.5442760692550443e-05, + "loss": 0.855, + "step": 8699 + }, + { + "epoch": 0.38519863188072123, + "grad_norm": 1.274789810180664, + "learning_rate": 1.5441464139237264e-05, + "loss": 0.817, + "step": 8700 + }, + { + "epoch": 0.38524290758553514, + "grad_norm": 1.5923986434936523, + "learning_rate": 1.5440167455959148e-05, + "loss": 0.7932, + "step": 8701 + }, + { + "epoch": 0.385287183290349, + "grad_norm": 1.6788382530212402, + "learning_rate": 1.5438870642747062e-05, + "loss": 0.8651, + "step": 8702 + }, + { + "epoch": 0.38533145899516286, + "grad_norm": 1.3487790822982788, + "learning_rate": 1.5437573699631978e-05, + "loss": 0.747, + "step": 8703 + }, + { + "epoch": 0.38537573469997677, + "grad_norm": 1.1609450578689575, + "learning_rate": 1.543627662664488e-05, + "loss": 0.6975, + "step": 8704 + }, + { + "epoch": 0.3854200104047906, + "grad_norm": 1.8246710300445557, + "learning_rate": 1.5434979423816733e-05, + "loss": 0.7823, + "step": 8705 + }, + { + "epoch": 0.3854642861096045, + "grad_norm": 1.5814297199249268, + "learning_rate": 1.5433682091178525e-05, + "loss": 0.8538, + "step": 8706 + }, + { + "epoch": 0.3855085618144184, + "grad_norm": 1.302675724029541, + "learning_rate": 1.5432384628761252e-05, + "loss": 0.7133, + "step": 8707 + }, + { + "epoch": 0.38555283751923225, + "grad_norm": 1.5564666986465454, + "learning_rate": 1.5431087036595892e-05, + "loss": 0.8072, + "step": 8708 + }, + { + "epoch": 0.38559711322404616, + "grad_norm": 1.384438395500183, + "learning_rate": 1.5429789314713444e-05, + "loss": 0.7576, + "step": 8709 + }, + { + "epoch": 0.38564138892886, + "grad_norm": 1.6042158603668213, + "learning_rate": 1.5428491463144895e-05, + "loss": 0.886, + "step": 8710 + }, + { + "epoch": 0.3856856646336739, + "grad_norm": 1.3558839559555054, + "learning_rate": 1.542719348192125e-05, + "loss": 0.7739, + "step": 8711 + }, + { + "epoch": 0.3857299403384878, + "grad_norm": 1.388683557510376, + "learning_rate": 1.5425895371073507e-05, + "loss": 0.7549, + "step": 8712 + }, + { + "epoch": 0.38577421604330164, + "grad_norm": 1.383737325668335, + "learning_rate": 1.5424597130632667e-05, + "loss": 0.7488, + "step": 8713 + }, + { + "epoch": 0.3858184917481155, + "grad_norm": 1.4942970275878906, + "learning_rate": 1.542329876062975e-05, + "loss": 0.7646, + "step": 8714 + }, + { + "epoch": 0.3858627674529294, + "grad_norm": 1.3073391914367676, + "learning_rate": 1.542200026109575e-05, + "loss": 0.7494, + "step": 8715 + }, + { + "epoch": 0.38590704315774327, + "grad_norm": 1.4752143621444702, + "learning_rate": 1.5420701632061693e-05, + "loss": 0.8312, + "step": 8716 + }, + { + "epoch": 0.3859513188625571, + "grad_norm": 1.291363000869751, + "learning_rate": 1.5419402873558593e-05, + "loss": 0.7414, + "step": 8717 + }, + { + "epoch": 0.38599559456737104, + "grad_norm": 1.5842158794403076, + "learning_rate": 1.541810398561746e-05, + "loss": 0.8485, + "step": 8718 + }, + { + "epoch": 0.3860398702721849, + "grad_norm": 1.5004888772964478, + "learning_rate": 1.541680496826933e-05, + "loss": 0.8109, + "step": 8719 + }, + { + "epoch": 0.38608414597699875, + "grad_norm": 1.3828115463256836, + "learning_rate": 1.5415505821545223e-05, + "loss": 0.7139, + "step": 8720 + }, + { + "epoch": 0.38612842168181266, + "grad_norm": 1.4365181922912598, + "learning_rate": 1.5414206545476172e-05, + "loss": 0.7803, + "step": 8721 + }, + { + "epoch": 0.3861726973866265, + "grad_norm": 1.5534685850143433, + "learning_rate": 1.5412907140093204e-05, + "loss": 0.7704, + "step": 8722 + }, + { + "epoch": 0.3862169730914404, + "grad_norm": 1.3077318668365479, + "learning_rate": 1.5411607605427356e-05, + "loss": 0.7646, + "step": 8723 + }, + { + "epoch": 0.3862612487962543, + "grad_norm": 1.178511142730713, + "learning_rate": 1.5410307941509667e-05, + "loss": 0.6963, + "step": 8724 + }, + { + "epoch": 0.38630552450106814, + "grad_norm": 1.529165267944336, + "learning_rate": 1.5409008148371176e-05, + "loss": 0.842, + "step": 8725 + }, + { + "epoch": 0.38634980020588205, + "grad_norm": 1.4871275424957275, + "learning_rate": 1.540770822604293e-05, + "loss": 0.8881, + "step": 8726 + }, + { + "epoch": 0.3863940759106959, + "grad_norm": 1.4654697179794312, + "learning_rate": 1.5406408174555978e-05, + "loss": 0.7888, + "step": 8727 + }, + { + "epoch": 0.38643835161550977, + "grad_norm": 1.1219663619995117, + "learning_rate": 1.5405107993941368e-05, + "loss": 0.7002, + "step": 8728 + }, + { + "epoch": 0.3864826273203237, + "grad_norm": 1.215195894241333, + "learning_rate": 1.5403807684230154e-05, + "loss": 0.774, + "step": 8729 + }, + { + "epoch": 0.38652690302513754, + "grad_norm": 1.3769718408584595, + "learning_rate": 1.5402507245453393e-05, + "loss": 0.8156, + "step": 8730 + }, + { + "epoch": 0.3865711787299514, + "grad_norm": 1.5768159627914429, + "learning_rate": 1.5401206677642147e-05, + "loss": 0.7574, + "step": 8731 + }, + { + "epoch": 0.3866154544347653, + "grad_norm": 1.5350812673568726, + "learning_rate": 1.5399905980827478e-05, + "loss": 0.7962, + "step": 8732 + }, + { + "epoch": 0.38665973013957916, + "grad_norm": 1.2551393508911133, + "learning_rate": 1.539860515504045e-05, + "loss": 0.7185, + "step": 8733 + }, + { + "epoch": 0.386704005844393, + "grad_norm": 1.3064415454864502, + "learning_rate": 1.539730420031213e-05, + "loss": 0.7597, + "step": 8734 + }, + { + "epoch": 0.38674828154920693, + "grad_norm": 1.7335941791534424, + "learning_rate": 1.53960031166736e-05, + "loss": 0.7302, + "step": 8735 + }, + { + "epoch": 0.3867925572540208, + "grad_norm": 1.4143723249435425, + "learning_rate": 1.5394701904155927e-05, + "loss": 0.7853, + "step": 8736 + }, + { + "epoch": 0.38683683295883464, + "grad_norm": 1.2006158828735352, + "learning_rate": 1.5393400562790192e-05, + "loss": 0.7091, + "step": 8737 + }, + { + "epoch": 0.38688110866364855, + "grad_norm": 1.4969245195388794, + "learning_rate": 1.5392099092607473e-05, + "loss": 0.7924, + "step": 8738 + }, + { + "epoch": 0.3869253843684624, + "grad_norm": 1.5995018482208252, + "learning_rate": 1.5390797493638862e-05, + "loss": 0.7812, + "step": 8739 + }, + { + "epoch": 0.38696966007327627, + "grad_norm": 1.5284044742584229, + "learning_rate": 1.5389495765915436e-05, + "loss": 0.832, + "step": 8740 + }, + { + "epoch": 0.3870139357780902, + "grad_norm": 1.3213638067245483, + "learning_rate": 1.53881939094683e-05, + "loss": 0.7191, + "step": 8741 + }, + { + "epoch": 0.38705821148290404, + "grad_norm": 1.5537904500961304, + "learning_rate": 1.5386891924328534e-05, + "loss": 0.8142, + "step": 8742 + }, + { + "epoch": 0.3871024871877179, + "grad_norm": 1.3096046447753906, + "learning_rate": 1.538558981052724e-05, + "loss": 0.7212, + "step": 8743 + }, + { + "epoch": 0.3871467628925318, + "grad_norm": 1.488181710243225, + "learning_rate": 1.5384287568095523e-05, + "loss": 0.7379, + "step": 8744 + }, + { + "epoch": 0.38719103859734566, + "grad_norm": 1.228494644165039, + "learning_rate": 1.538298519706448e-05, + "loss": 0.7351, + "step": 8745 + }, + { + "epoch": 0.3872353143021596, + "grad_norm": 1.298608422279358, + "learning_rate": 1.5381682697465217e-05, + "loss": 0.723, + "step": 8746 + }, + { + "epoch": 0.38727959000697343, + "grad_norm": 1.2725671529769897, + "learning_rate": 1.5380380069328845e-05, + "loss": 0.7233, + "step": 8747 + }, + { + "epoch": 0.3873238657117873, + "grad_norm": 1.5097017288208008, + "learning_rate": 1.537907731268648e-05, + "loss": 0.7681, + "step": 8748 + }, + { + "epoch": 0.3873681414166012, + "grad_norm": 1.7127715349197388, + "learning_rate": 1.5377774427569224e-05, + "loss": 0.782, + "step": 8749 + }, + { + "epoch": 0.38741241712141505, + "grad_norm": 1.3592910766601562, + "learning_rate": 1.5376471414008212e-05, + "loss": 0.7382, + "step": 8750 + }, + { + "epoch": 0.3874566928262289, + "grad_norm": 1.168443202972412, + "learning_rate": 1.537516827203456e-05, + "loss": 0.7261, + "step": 8751 + }, + { + "epoch": 0.3875009685310428, + "grad_norm": 1.2826076745986938, + "learning_rate": 1.5373865001679385e-05, + "loss": 0.7502, + "step": 8752 + }, + { + "epoch": 0.3875452442358567, + "grad_norm": 1.3474020957946777, + "learning_rate": 1.537256160297382e-05, + "loss": 0.7409, + "step": 8753 + }, + { + "epoch": 0.38758951994067053, + "grad_norm": 1.357485055923462, + "learning_rate": 1.5371258075948997e-05, + "loss": 0.7566, + "step": 8754 + }, + { + "epoch": 0.38763379564548445, + "grad_norm": 1.1451092958450317, + "learning_rate": 1.5369954420636048e-05, + "loss": 0.7479, + "step": 8755 + }, + { + "epoch": 0.3876780713502983, + "grad_norm": 1.5856064558029175, + "learning_rate": 1.536865063706611e-05, + "loss": 0.7869, + "step": 8756 + }, + { + "epoch": 0.38772234705511216, + "grad_norm": 1.5633898973464966, + "learning_rate": 1.5367346725270325e-05, + "loss": 0.7513, + "step": 8757 + }, + { + "epoch": 0.38776662275992607, + "grad_norm": 1.315793514251709, + "learning_rate": 1.536604268527983e-05, + "loss": 0.759, + "step": 8758 + }, + { + "epoch": 0.38781089846473993, + "grad_norm": 1.2807085514068604, + "learning_rate": 1.5364738517125777e-05, + "loss": 0.8192, + "step": 8759 + }, + { + "epoch": 0.3878551741695538, + "grad_norm": 1.5971518754959106, + "learning_rate": 1.5363434220839313e-05, + "loss": 0.7582, + "step": 8760 + }, + { + "epoch": 0.3878994498743677, + "grad_norm": 1.272247314453125, + "learning_rate": 1.5362129796451588e-05, + "loss": 0.8017, + "step": 8761 + }, + { + "epoch": 0.38794372557918155, + "grad_norm": 1.4147590398788452, + "learning_rate": 1.536082524399376e-05, + "loss": 0.7303, + "step": 8762 + }, + { + "epoch": 0.38798800128399547, + "grad_norm": 1.3226407766342163, + "learning_rate": 1.5359520563496985e-05, + "loss": 0.7034, + "step": 8763 + }, + { + "epoch": 0.3880322769888093, + "grad_norm": 1.350923776626587, + "learning_rate": 1.5358215754992425e-05, + "loss": 0.7842, + "step": 8764 + }, + { + "epoch": 0.3880765526936232, + "grad_norm": 1.5464261770248413, + "learning_rate": 1.5356910818511247e-05, + "loss": 0.8093, + "step": 8765 + }, + { + "epoch": 0.3881208283984371, + "grad_norm": 1.3644371032714844, + "learning_rate": 1.5355605754084616e-05, + "loss": 0.7316, + "step": 8766 + }, + { + "epoch": 0.38816510410325095, + "grad_norm": 1.2377872467041016, + "learning_rate": 1.53543005617437e-05, + "loss": 0.7947, + "step": 8767 + }, + { + "epoch": 0.3882093798080648, + "grad_norm": 1.444298267364502, + "learning_rate": 1.5352995241519674e-05, + "loss": 0.8205, + "step": 8768 + }, + { + "epoch": 0.3882536555128787, + "grad_norm": 1.5201395750045776, + "learning_rate": 1.5351689793443714e-05, + "loss": 0.7993, + "step": 8769 + }, + { + "epoch": 0.38829793121769257, + "grad_norm": 1.2985341548919678, + "learning_rate": 1.5350384217547007e-05, + "loss": 0.7964, + "step": 8770 + }, + { + "epoch": 0.3883422069225064, + "grad_norm": 1.5805015563964844, + "learning_rate": 1.5349078513860728e-05, + "loss": 0.7802, + "step": 8771 + }, + { + "epoch": 0.38838648262732034, + "grad_norm": 1.6271581649780273, + "learning_rate": 1.534777268241606e-05, + "loss": 0.7776, + "step": 8772 + }, + { + "epoch": 0.3884307583321342, + "grad_norm": 1.2289763689041138, + "learning_rate": 1.5346466723244195e-05, + "loss": 0.6703, + "step": 8773 + }, + { + "epoch": 0.38847503403694805, + "grad_norm": 1.4551959037780762, + "learning_rate": 1.534516063637633e-05, + "loss": 0.7813, + "step": 8774 + }, + { + "epoch": 0.38851930974176196, + "grad_norm": 1.495491862297058, + "learning_rate": 1.5343854421843656e-05, + "loss": 0.7208, + "step": 8775 + }, + { + "epoch": 0.3885635854465758, + "grad_norm": 1.4254769086837769, + "learning_rate": 1.5342548079677366e-05, + "loss": 0.6795, + "step": 8776 + }, + { + "epoch": 0.3886078611513897, + "grad_norm": 1.2823727130889893, + "learning_rate": 1.534124160990867e-05, + "loss": 0.678, + "step": 8777 + }, + { + "epoch": 0.3886521368562036, + "grad_norm": 1.3238259553909302, + "learning_rate": 1.5339935012568757e-05, + "loss": 0.7602, + "step": 8778 + }, + { + "epoch": 0.38869641256101745, + "grad_norm": 1.3327635526657104, + "learning_rate": 1.5338628287688858e-05, + "loss": 0.7745, + "step": 8779 + }, + { + "epoch": 0.38874068826583136, + "grad_norm": 1.5541810989379883, + "learning_rate": 1.533732143530016e-05, + "loss": 0.7627, + "step": 8780 + }, + { + "epoch": 0.3887849639706452, + "grad_norm": 1.5906546115875244, + "learning_rate": 1.533601445543389e-05, + "loss": 0.8139, + "step": 8781 + }, + { + "epoch": 0.38882923967545907, + "grad_norm": 1.1996194124221802, + "learning_rate": 1.5334707348121254e-05, + "loss": 0.7546, + "step": 8782 + }, + { + "epoch": 0.388873515380273, + "grad_norm": 1.3422938585281372, + "learning_rate": 1.533340011339348e-05, + "loss": 0.7243, + "step": 8783 + }, + { + "epoch": 0.38891779108508684, + "grad_norm": 1.399420976638794, + "learning_rate": 1.5332092751281788e-05, + "loss": 0.7844, + "step": 8784 + }, + { + "epoch": 0.3889620667899007, + "grad_norm": 1.6047255992889404, + "learning_rate": 1.53307852618174e-05, + "loss": 0.7341, + "step": 8785 + }, + { + "epoch": 0.3890063424947146, + "grad_norm": 1.430673599243164, + "learning_rate": 1.532947764503155e-05, + "loss": 0.7912, + "step": 8786 + }, + { + "epoch": 0.38905061819952846, + "grad_norm": 1.4904683828353882, + "learning_rate": 1.5328169900955463e-05, + "loss": 0.7506, + "step": 8787 + }, + { + "epoch": 0.3890948939043423, + "grad_norm": 1.5190396308898926, + "learning_rate": 1.5326862029620376e-05, + "loss": 0.7868, + "step": 8788 + }, + { + "epoch": 0.38913916960915623, + "grad_norm": 1.628784418106079, + "learning_rate": 1.532555403105753e-05, + "loss": 0.7627, + "step": 8789 + }, + { + "epoch": 0.3891834453139701, + "grad_norm": 1.3592407703399658, + "learning_rate": 1.532424590529816e-05, + "loss": 0.7656, + "step": 8790 + }, + { + "epoch": 0.38922772101878395, + "grad_norm": 1.6790724992752075, + "learning_rate": 1.532293765237351e-05, + "loss": 0.8078, + "step": 8791 + }, + { + "epoch": 0.38927199672359786, + "grad_norm": 1.331309199333191, + "learning_rate": 1.532162927231483e-05, + "loss": 0.7707, + "step": 8792 + }, + { + "epoch": 0.3893162724284117, + "grad_norm": 1.8774347305297852, + "learning_rate": 1.5320320765153367e-05, + "loss": 0.7513, + "step": 8793 + }, + { + "epoch": 0.38936054813322557, + "grad_norm": 1.360850214958191, + "learning_rate": 1.5319012130920377e-05, + "loss": 0.7516, + "step": 8794 + }, + { + "epoch": 0.3894048238380395, + "grad_norm": 1.3590251207351685, + "learning_rate": 1.531770336964711e-05, + "loss": 0.7397, + "step": 8795 + }, + { + "epoch": 0.38944909954285334, + "grad_norm": 1.2672795057296753, + "learning_rate": 1.5316394481364827e-05, + "loss": 0.755, + "step": 8796 + }, + { + "epoch": 0.38949337524766725, + "grad_norm": 1.3536124229431152, + "learning_rate": 1.5315085466104793e-05, + "loss": 0.7565, + "step": 8797 + }, + { + "epoch": 0.3895376509524811, + "grad_norm": 1.0764883756637573, + "learning_rate": 1.5313776323898274e-05, + "loss": 0.6801, + "step": 8798 + }, + { + "epoch": 0.38958192665729496, + "grad_norm": 1.460363507270813, + "learning_rate": 1.531246705477653e-05, + "loss": 0.811, + "step": 8799 + }, + { + "epoch": 0.3896262023621089, + "grad_norm": 1.8307738304138184, + "learning_rate": 1.5311157658770834e-05, + "loss": 0.8166, + "step": 8800 + }, + { + "epoch": 0.38967047806692273, + "grad_norm": 1.2960742712020874, + "learning_rate": 1.5309848135912466e-05, + "loss": 0.7255, + "step": 8801 + }, + { + "epoch": 0.3897147537717366, + "grad_norm": 1.304714560508728, + "learning_rate": 1.5308538486232692e-05, + "loss": 0.7008, + "step": 8802 + }, + { + "epoch": 0.3897590294765505, + "grad_norm": 1.6517177820205688, + "learning_rate": 1.5307228709762806e-05, + "loss": 0.7982, + "step": 8803 + }, + { + "epoch": 0.38980330518136436, + "grad_norm": 1.093601942062378, + "learning_rate": 1.530591880653408e-05, + "loss": 0.6442, + "step": 8804 + }, + { + "epoch": 0.3898475808861782, + "grad_norm": 1.758380651473999, + "learning_rate": 1.5304608776577805e-05, + "loss": 0.9014, + "step": 8805 + }, + { + "epoch": 0.3898918565909921, + "grad_norm": 1.4021443128585815, + "learning_rate": 1.5303298619925268e-05, + "loss": 0.6736, + "step": 8806 + }, + { + "epoch": 0.389936132295806, + "grad_norm": 1.2644522190093994, + "learning_rate": 1.530198833660776e-05, + "loss": 0.7021, + "step": 8807 + }, + { + "epoch": 0.38998040800061984, + "grad_norm": 1.6459105014801025, + "learning_rate": 1.530067792665658e-05, + "loss": 0.872, + "step": 8808 + }, + { + "epoch": 0.39002468370543375, + "grad_norm": 1.4251848459243774, + "learning_rate": 1.5299367390103024e-05, + "loss": 0.7685, + "step": 8809 + }, + { + "epoch": 0.3900689594102476, + "grad_norm": 1.4230945110321045, + "learning_rate": 1.529805672697839e-05, + "loss": 0.7719, + "step": 8810 + }, + { + "epoch": 0.39011323511506146, + "grad_norm": 1.432364583015442, + "learning_rate": 1.529674593731399e-05, + "loss": 0.7675, + "step": 8811 + }, + { + "epoch": 0.3901575108198754, + "grad_norm": 1.450720191001892, + "learning_rate": 1.529543502114112e-05, + "loss": 0.8489, + "step": 8812 + }, + { + "epoch": 0.39020178652468923, + "grad_norm": 1.4480552673339844, + "learning_rate": 1.5294123978491098e-05, + "loss": 0.7938, + "step": 8813 + }, + { + "epoch": 0.39024606222950314, + "grad_norm": 1.4631098508834839, + "learning_rate": 1.5292812809395238e-05, + "loss": 0.7605, + "step": 8814 + }, + { + "epoch": 0.390290337934317, + "grad_norm": 1.3991341590881348, + "learning_rate": 1.529150151388485e-05, + "loss": 0.7917, + "step": 8815 + }, + { + "epoch": 0.39033461363913086, + "grad_norm": 1.2184325456619263, + "learning_rate": 1.529019009199126e-05, + "loss": 0.721, + "step": 8816 + }, + { + "epoch": 0.39037888934394477, + "grad_norm": 1.3056368827819824, + "learning_rate": 1.5288878543745785e-05, + "loss": 0.7434, + "step": 8817 + }, + { + "epoch": 0.3904231650487586, + "grad_norm": 1.3111191987991333, + "learning_rate": 1.528756686917975e-05, + "loss": 0.8777, + "step": 8818 + }, + { + "epoch": 0.3904674407535725, + "grad_norm": 1.5120835304260254, + "learning_rate": 1.5286255068324493e-05, + "loss": 0.8171, + "step": 8819 + }, + { + "epoch": 0.3905117164583864, + "grad_norm": 1.2091304063796997, + "learning_rate": 1.528494314121133e-05, + "loss": 0.6949, + "step": 8820 + }, + { + "epoch": 0.39055599216320025, + "grad_norm": 1.3377971649169922, + "learning_rate": 1.528363108787161e-05, + "loss": 0.7599, + "step": 8821 + }, + { + "epoch": 0.3906002678680141, + "grad_norm": 1.7142341136932373, + "learning_rate": 1.528231890833666e-05, + "loss": 0.8453, + "step": 8822 + }, + { + "epoch": 0.390644543572828, + "grad_norm": 1.3573311567306519, + "learning_rate": 1.528100660263782e-05, + "loss": 0.8391, + "step": 8823 + }, + { + "epoch": 0.3906888192776419, + "grad_norm": 1.268314242362976, + "learning_rate": 1.5279694170806446e-05, + "loss": 0.7144, + "step": 8824 + }, + { + "epoch": 0.39073309498245573, + "grad_norm": 1.1062123775482178, + "learning_rate": 1.527838161287387e-05, + "loss": 0.6385, + "step": 8825 + }, + { + "epoch": 0.39077737068726964, + "grad_norm": 1.5102956295013428, + "learning_rate": 1.527706892887145e-05, + "loss": 0.8093, + "step": 8826 + }, + { + "epoch": 0.3908216463920835, + "grad_norm": 1.4653749465942383, + "learning_rate": 1.5275756118830533e-05, + "loss": 0.7423, + "step": 8827 + }, + { + "epoch": 0.39086592209689736, + "grad_norm": 1.2517688274383545, + "learning_rate": 1.5274443182782478e-05, + "loss": 0.6853, + "step": 8828 + }, + { + "epoch": 0.39091019780171127, + "grad_norm": 1.5496047735214233, + "learning_rate": 1.527313012075864e-05, + "loss": 0.7572, + "step": 8829 + }, + { + "epoch": 0.3909544735065251, + "grad_norm": 1.3767956495285034, + "learning_rate": 1.5271816932790383e-05, + "loss": 0.7483, + "step": 8830 + }, + { + "epoch": 0.390998749211339, + "grad_norm": 1.6976697444915771, + "learning_rate": 1.5270503618909076e-05, + "loss": 0.7827, + "step": 8831 + }, + { + "epoch": 0.3910430249161529, + "grad_norm": 1.417134165763855, + "learning_rate": 1.5269190179146075e-05, + "loss": 0.8181, + "step": 8832 + }, + { + "epoch": 0.39108730062096675, + "grad_norm": 1.3439570665359497, + "learning_rate": 1.526787661353276e-05, + "loss": 0.7141, + "step": 8833 + }, + { + "epoch": 0.39113157632578066, + "grad_norm": 1.540478229522705, + "learning_rate": 1.52665629221005e-05, + "loss": 0.7761, + "step": 8834 + }, + { + "epoch": 0.3911758520305945, + "grad_norm": 1.308811902999878, + "learning_rate": 1.5265249104880675e-05, + "loss": 0.7553, + "step": 8835 + }, + { + "epoch": 0.3912201277354084, + "grad_norm": 1.7091279029846191, + "learning_rate": 1.5263935161904663e-05, + "loss": 0.8259, + "step": 8836 + }, + { + "epoch": 0.3912644034402223, + "grad_norm": 1.1411898136138916, + "learning_rate": 1.5262621093203842e-05, + "loss": 0.652, + "step": 8837 + }, + { + "epoch": 0.39130867914503614, + "grad_norm": 1.5467047691345215, + "learning_rate": 1.5261306898809604e-05, + "loss": 0.779, + "step": 8838 + }, + { + "epoch": 0.39135295484985, + "grad_norm": 1.4761229753494263, + "learning_rate": 1.5259992578753335e-05, + "loss": 0.7105, + "step": 8839 + }, + { + "epoch": 0.3913972305546639, + "grad_norm": 1.2987052202224731, + "learning_rate": 1.5258678133066423e-05, + "loss": 0.7635, + "step": 8840 + }, + { + "epoch": 0.39144150625947777, + "grad_norm": 1.5808290243148804, + "learning_rate": 1.5257363561780267e-05, + "loss": 0.7097, + "step": 8841 + }, + { + "epoch": 0.3914857819642916, + "grad_norm": 1.441309928894043, + "learning_rate": 1.5256048864926263e-05, + "loss": 0.8118, + "step": 8842 + }, + { + "epoch": 0.39153005766910554, + "grad_norm": 1.2291090488433838, + "learning_rate": 1.5254734042535812e-05, + "loss": 0.7212, + "step": 8843 + }, + { + "epoch": 0.3915743333739194, + "grad_norm": 1.4544953107833862, + "learning_rate": 1.5253419094640317e-05, + "loss": 0.7863, + "step": 8844 + }, + { + "epoch": 0.39161860907873325, + "grad_norm": 1.4659342765808105, + "learning_rate": 1.5252104021271183e-05, + "loss": 0.8293, + "step": 8845 + }, + { + "epoch": 0.39166288478354716, + "grad_norm": 1.4013690948486328, + "learning_rate": 1.5250788822459824e-05, + "loss": 0.7716, + "step": 8846 + }, + { + "epoch": 0.391707160488361, + "grad_norm": 1.788530707359314, + "learning_rate": 1.5249473498237646e-05, + "loss": 0.8712, + "step": 8847 + }, + { + "epoch": 0.3917514361931749, + "grad_norm": 1.4859411716461182, + "learning_rate": 1.5248158048636069e-05, + "loss": 0.7715, + "step": 8848 + }, + { + "epoch": 0.3917957118979888, + "grad_norm": 1.3641321659088135, + "learning_rate": 1.5246842473686509e-05, + "loss": 0.7154, + "step": 8849 + }, + { + "epoch": 0.39183998760280264, + "grad_norm": 1.5552067756652832, + "learning_rate": 1.524552677342039e-05, + "loss": 0.7664, + "step": 8850 + }, + { + "epoch": 0.39188426330761655, + "grad_norm": 1.539684772491455, + "learning_rate": 1.5244210947869138e-05, + "loss": 0.8391, + "step": 8851 + }, + { + "epoch": 0.3919285390124304, + "grad_norm": 1.7921994924545288, + "learning_rate": 1.5242894997064173e-05, + "loss": 0.8749, + "step": 8852 + }, + { + "epoch": 0.39197281471724427, + "grad_norm": 1.2951842546463013, + "learning_rate": 1.524157892103693e-05, + "loss": 0.7386, + "step": 8853 + }, + { + "epoch": 0.3920170904220582, + "grad_norm": 1.7452791929244995, + "learning_rate": 1.5240262719818846e-05, + "loss": 0.8159, + "step": 8854 + }, + { + "epoch": 0.39206136612687204, + "grad_norm": 1.6667073965072632, + "learning_rate": 1.5238946393441349e-05, + "loss": 0.7655, + "step": 8855 + }, + { + "epoch": 0.3921056418316859, + "grad_norm": 1.9142067432403564, + "learning_rate": 1.5237629941935888e-05, + "loss": 0.8103, + "step": 8856 + }, + { + "epoch": 0.3921499175364998, + "grad_norm": 1.3534795045852661, + "learning_rate": 1.5236313365333896e-05, + "loss": 0.8291, + "step": 8857 + }, + { + "epoch": 0.39219419324131366, + "grad_norm": 1.1771740913391113, + "learning_rate": 1.5234996663666826e-05, + "loss": 0.691, + "step": 8858 + }, + { + "epoch": 0.3922384689461275, + "grad_norm": 1.6646310091018677, + "learning_rate": 1.5233679836966122e-05, + "loss": 0.7901, + "step": 8859 + }, + { + "epoch": 0.39228274465094143, + "grad_norm": 1.4739404916763306, + "learning_rate": 1.5232362885263237e-05, + "loss": 0.7536, + "step": 8860 + }, + { + "epoch": 0.3923270203557553, + "grad_norm": 1.3558391332626343, + "learning_rate": 1.5231045808589624e-05, + "loss": 0.7143, + "step": 8861 + }, + { + "epoch": 0.39237129606056914, + "grad_norm": 1.734178066253662, + "learning_rate": 1.522972860697674e-05, + "loss": 0.8209, + "step": 8862 + }, + { + "epoch": 0.39241557176538305, + "grad_norm": 1.6269195079803467, + "learning_rate": 1.5228411280456051e-05, + "loss": 0.8683, + "step": 8863 + }, + { + "epoch": 0.3924598474701969, + "grad_norm": 1.9192724227905273, + "learning_rate": 1.5227093829059013e-05, + "loss": 0.6781, + "step": 8864 + }, + { + "epoch": 0.39250412317501077, + "grad_norm": 1.8269028663635254, + "learning_rate": 1.5225776252817094e-05, + "loss": 0.8177, + "step": 8865 + }, + { + "epoch": 0.3925483988798247, + "grad_norm": 1.5200202465057373, + "learning_rate": 1.5224458551761764e-05, + "loss": 0.7565, + "step": 8866 + }, + { + "epoch": 0.39259267458463853, + "grad_norm": 1.459656000137329, + "learning_rate": 1.5223140725924494e-05, + "loss": 0.7586, + "step": 8867 + }, + { + "epoch": 0.39263695028945245, + "grad_norm": 1.845550537109375, + "learning_rate": 1.5221822775336766e-05, + "loss": 0.808, + "step": 8868 + }, + { + "epoch": 0.3926812259942663, + "grad_norm": 1.3107253313064575, + "learning_rate": 1.5220504700030048e-05, + "loss": 0.7614, + "step": 8869 + }, + { + "epoch": 0.39272550169908016, + "grad_norm": 1.6450228691101074, + "learning_rate": 1.5219186500035827e-05, + "loss": 0.7857, + "step": 8870 + }, + { + "epoch": 0.39276977740389407, + "grad_norm": 1.1951653957366943, + "learning_rate": 1.5217868175385588e-05, + "loss": 0.7417, + "step": 8871 + }, + { + "epoch": 0.39281405310870793, + "grad_norm": 1.0946706533432007, + "learning_rate": 1.5216549726110815e-05, + "loss": 0.658, + "step": 8872 + }, + { + "epoch": 0.3928583288135218, + "grad_norm": 1.4991422891616821, + "learning_rate": 1.5215231152242998e-05, + "loss": 0.7334, + "step": 8873 + }, + { + "epoch": 0.3929026045183357, + "grad_norm": 1.4774565696716309, + "learning_rate": 1.5213912453813632e-05, + "loss": 0.7522, + "step": 8874 + }, + { + "epoch": 0.39294688022314955, + "grad_norm": 1.3187588453292847, + "learning_rate": 1.5212593630854211e-05, + "loss": 0.7322, + "step": 8875 + }, + { + "epoch": 0.3929911559279634, + "grad_norm": 1.2268197536468506, + "learning_rate": 1.521127468339624e-05, + "loss": 0.7187, + "step": 8876 + }, + { + "epoch": 0.3930354316327773, + "grad_norm": 1.6066285371780396, + "learning_rate": 1.520995561147121e-05, + "loss": 0.7363, + "step": 8877 + }, + { + "epoch": 0.3930797073375912, + "grad_norm": 1.2303699254989624, + "learning_rate": 1.520863641511064e-05, + "loss": 0.7099, + "step": 8878 + }, + { + "epoch": 0.39312398304240503, + "grad_norm": 1.6499167680740356, + "learning_rate": 1.5207317094346024e-05, + "loss": 0.7982, + "step": 8879 + }, + { + "epoch": 0.39316825874721895, + "grad_norm": 1.3539340496063232, + "learning_rate": 1.5205997649208882e-05, + "loss": 0.8242, + "step": 8880 + }, + { + "epoch": 0.3932125344520328, + "grad_norm": 1.4091705083847046, + "learning_rate": 1.5204678079730724e-05, + "loss": 0.7246, + "step": 8881 + }, + { + "epoch": 0.39325681015684666, + "grad_norm": 1.5650956630706787, + "learning_rate": 1.5203358385943067e-05, + "loss": 0.8619, + "step": 8882 + }, + { + "epoch": 0.39330108586166057, + "grad_norm": 1.3654104471206665, + "learning_rate": 1.5202038567877436e-05, + "loss": 0.7201, + "step": 8883 + }, + { + "epoch": 0.3933453615664744, + "grad_norm": 1.3112372159957886, + "learning_rate": 1.5200718625565345e-05, + "loss": 0.7555, + "step": 8884 + }, + { + "epoch": 0.39338963727128834, + "grad_norm": 1.3077857494354248, + "learning_rate": 1.5199398559038323e-05, + "loss": 0.7468, + "step": 8885 + }, + { + "epoch": 0.3934339129761022, + "grad_norm": 1.548938512802124, + "learning_rate": 1.5198078368327904e-05, + "loss": 0.7443, + "step": 8886 + }, + { + "epoch": 0.39347818868091605, + "grad_norm": 1.338797926902771, + "learning_rate": 1.5196758053465614e-05, + "loss": 0.7436, + "step": 8887 + }, + { + "epoch": 0.39352246438572996, + "grad_norm": 1.452283263206482, + "learning_rate": 1.519543761448299e-05, + "loss": 0.7742, + "step": 8888 + }, + { + "epoch": 0.3935667400905438, + "grad_norm": 1.4352562427520752, + "learning_rate": 1.5194117051411568e-05, + "loss": 0.7689, + "step": 8889 + }, + { + "epoch": 0.3936110157953577, + "grad_norm": 1.3694418668746948, + "learning_rate": 1.5192796364282889e-05, + "loss": 0.6909, + "step": 8890 + }, + { + "epoch": 0.3936552915001716, + "grad_norm": 1.580637812614441, + "learning_rate": 1.5191475553128498e-05, + "loss": 0.7946, + "step": 8891 + }, + { + "epoch": 0.39369956720498545, + "grad_norm": 1.317723035812378, + "learning_rate": 1.5190154617979939e-05, + "loss": 0.6961, + "step": 8892 + }, + { + "epoch": 0.3937438429097993, + "grad_norm": 1.4302260875701904, + "learning_rate": 1.5188833558868765e-05, + "loss": 0.715, + "step": 8893 + }, + { + "epoch": 0.3937881186146132, + "grad_norm": 1.3760651350021362, + "learning_rate": 1.5187512375826522e-05, + "loss": 0.7082, + "step": 8894 + }, + { + "epoch": 0.39383239431942707, + "grad_norm": 1.352253794670105, + "learning_rate": 1.5186191068884774e-05, + "loss": 0.7264, + "step": 8895 + }, + { + "epoch": 0.3938766700242409, + "grad_norm": 1.695732593536377, + "learning_rate": 1.5184869638075075e-05, + "loss": 0.8734, + "step": 8896 + }, + { + "epoch": 0.39392094572905484, + "grad_norm": 1.5683223009109497, + "learning_rate": 1.5183548083428985e-05, + "loss": 0.7719, + "step": 8897 + }, + { + "epoch": 0.3939652214338687, + "grad_norm": 1.6778737306594849, + "learning_rate": 1.5182226404978071e-05, + "loss": 0.7311, + "step": 8898 + }, + { + "epoch": 0.39400949713868255, + "grad_norm": 1.4000518321990967, + "learning_rate": 1.5180904602753898e-05, + "loss": 0.7547, + "step": 8899 + }, + { + "epoch": 0.39405377284349646, + "grad_norm": 1.4272509813308716, + "learning_rate": 1.5179582676788036e-05, + "loss": 0.8113, + "step": 8900 + }, + { + "epoch": 0.3940980485483103, + "grad_norm": 1.2004371881484985, + "learning_rate": 1.517826062711206e-05, + "loss": 0.7231, + "step": 8901 + }, + { + "epoch": 0.39414232425312423, + "grad_norm": 1.4911943674087524, + "learning_rate": 1.5176938453757548e-05, + "loss": 0.8187, + "step": 8902 + }, + { + "epoch": 0.3941865999579381, + "grad_norm": 1.3616747856140137, + "learning_rate": 1.5175616156756072e-05, + "loss": 0.7519, + "step": 8903 + }, + { + "epoch": 0.39423087566275195, + "grad_norm": 1.6370781660079956, + "learning_rate": 1.517429373613922e-05, + "loss": 0.8235, + "step": 8904 + }, + { + "epoch": 0.39427515136756586, + "grad_norm": 1.3051271438598633, + "learning_rate": 1.5172971191938573e-05, + "loss": 0.7556, + "step": 8905 + }, + { + "epoch": 0.3943194270723797, + "grad_norm": 1.2209219932556152, + "learning_rate": 1.5171648524185725e-05, + "loss": 0.7411, + "step": 8906 + }, + { + "epoch": 0.39436370277719357, + "grad_norm": 1.4425849914550781, + "learning_rate": 1.517032573291226e-05, + "loss": 0.7448, + "step": 8907 + }, + { + "epoch": 0.3944079784820075, + "grad_norm": 1.5991522073745728, + "learning_rate": 1.5169002818149778e-05, + "loss": 0.8167, + "step": 8908 + }, + { + "epoch": 0.39445225418682134, + "grad_norm": 1.8803446292877197, + "learning_rate": 1.5167679779929869e-05, + "loss": 0.825, + "step": 8909 + }, + { + "epoch": 0.3944965298916352, + "grad_norm": 2.1308486461639404, + "learning_rate": 1.5166356618284136e-05, + "loss": 0.8, + "step": 8910 + }, + { + "epoch": 0.3945408055964491, + "grad_norm": 1.4300001859664917, + "learning_rate": 1.5165033333244186e-05, + "loss": 0.7755, + "step": 8911 + }, + { + "epoch": 0.39458508130126296, + "grad_norm": 2.0223593711853027, + "learning_rate": 1.5163709924841614e-05, + "loss": 0.8392, + "step": 8912 + }, + { + "epoch": 0.3946293570060768, + "grad_norm": 1.368380069732666, + "learning_rate": 1.5162386393108039e-05, + "loss": 0.6958, + "step": 8913 + }, + { + "epoch": 0.39467363271089073, + "grad_norm": 1.8304100036621094, + "learning_rate": 1.5161062738075068e-05, + "loss": 0.8076, + "step": 8914 + }, + { + "epoch": 0.3947179084157046, + "grad_norm": 2.234062910079956, + "learning_rate": 1.5159738959774315e-05, + "loss": 0.8674, + "step": 8915 + }, + { + "epoch": 0.39476218412051844, + "grad_norm": 1.421363115310669, + "learning_rate": 1.51584150582374e-05, + "loss": 0.7106, + "step": 8916 + }, + { + "epoch": 0.39480645982533236, + "grad_norm": 1.8264024257659912, + "learning_rate": 1.5157091033495938e-05, + "loss": 0.7862, + "step": 8917 + }, + { + "epoch": 0.3948507355301462, + "grad_norm": 1.203309178352356, + "learning_rate": 1.5155766885581557e-05, + "loss": 0.6869, + "step": 8918 + }, + { + "epoch": 0.39489501123496007, + "grad_norm": 1.4055290222167969, + "learning_rate": 1.5154442614525885e-05, + "loss": 0.8052, + "step": 8919 + }, + { + "epoch": 0.394939286939774, + "grad_norm": 1.52849543094635, + "learning_rate": 1.5153118220360546e-05, + "loss": 0.7826, + "step": 8920 + }, + { + "epoch": 0.39498356264458784, + "grad_norm": 1.3751673698425293, + "learning_rate": 1.5151793703117173e-05, + "loss": 0.682, + "step": 8921 + }, + { + "epoch": 0.39502783834940175, + "grad_norm": 1.911401391029358, + "learning_rate": 1.5150469062827404e-05, + "loss": 0.7881, + "step": 8922 + }, + { + "epoch": 0.3950721140542156, + "grad_norm": 1.4177143573760986, + "learning_rate": 1.5149144299522874e-05, + "loss": 0.7652, + "step": 8923 + }, + { + "epoch": 0.39511638975902946, + "grad_norm": 1.5201929807662964, + "learning_rate": 1.5147819413235223e-05, + "loss": 0.8081, + "step": 8924 + }, + { + "epoch": 0.3951606654638434, + "grad_norm": 1.4445068836212158, + "learning_rate": 1.5146494403996101e-05, + "loss": 0.7602, + "step": 8925 + }, + { + "epoch": 0.39520494116865723, + "grad_norm": 1.16824209690094, + "learning_rate": 1.5145169271837148e-05, + "loss": 0.7154, + "step": 8926 + }, + { + "epoch": 0.3952492168734711, + "grad_norm": 1.3449350595474243, + "learning_rate": 1.5143844016790015e-05, + "loss": 0.7627, + "step": 8927 + }, + { + "epoch": 0.395293492578285, + "grad_norm": 1.4315003156661987, + "learning_rate": 1.5142518638886361e-05, + "loss": 0.801, + "step": 8928 + }, + { + "epoch": 0.39533776828309886, + "grad_norm": 1.5284947156906128, + "learning_rate": 1.514119313815783e-05, + "loss": 0.8416, + "step": 8929 + }, + { + "epoch": 0.3953820439879127, + "grad_norm": 1.3649622201919556, + "learning_rate": 1.5139867514636092e-05, + "loss": 0.7506, + "step": 8930 + }, + { + "epoch": 0.3954263196927266, + "grad_norm": 1.3347861766815186, + "learning_rate": 1.5138541768352802e-05, + "loss": 0.7645, + "step": 8931 + }, + { + "epoch": 0.3954705953975405, + "grad_norm": 1.2179580926895142, + "learning_rate": 1.5137215899339626e-05, + "loss": 0.7585, + "step": 8932 + }, + { + "epoch": 0.39551487110235434, + "grad_norm": 1.434403657913208, + "learning_rate": 1.513588990762823e-05, + "loss": 0.7441, + "step": 8933 + }, + { + "epoch": 0.39555914680716825, + "grad_norm": 1.2719037532806396, + "learning_rate": 1.5134563793250285e-05, + "loss": 0.7401, + "step": 8934 + }, + { + "epoch": 0.3956034225119821, + "grad_norm": 1.214235782623291, + "learning_rate": 1.5133237556237465e-05, + "loss": 0.7285, + "step": 8935 + }, + { + "epoch": 0.39564769821679596, + "grad_norm": 1.4901264905929565, + "learning_rate": 1.5131911196621446e-05, + "loss": 0.766, + "step": 8936 + }, + { + "epoch": 0.3956919739216099, + "grad_norm": 1.6595970392227173, + "learning_rate": 1.5130584714433907e-05, + "loss": 0.7131, + "step": 8937 + }, + { + "epoch": 0.39573624962642373, + "grad_norm": 1.3232924938201904, + "learning_rate": 1.5129258109706531e-05, + "loss": 0.7449, + "step": 8938 + }, + { + "epoch": 0.39578052533123764, + "grad_norm": 1.7512216567993164, + "learning_rate": 1.5127931382471e-05, + "loss": 0.7766, + "step": 8939 + }, + { + "epoch": 0.3958248010360515, + "grad_norm": 1.3783475160598755, + "learning_rate": 1.5126604532759003e-05, + "loss": 0.8105, + "step": 8940 + }, + { + "epoch": 0.39586907674086536, + "grad_norm": 1.3761792182922363, + "learning_rate": 1.512527756060223e-05, + "loss": 0.7622, + "step": 8941 + }, + { + "epoch": 0.39591335244567927, + "grad_norm": 1.7325009107589722, + "learning_rate": 1.5123950466032375e-05, + "loss": 0.8012, + "step": 8942 + }, + { + "epoch": 0.3959576281504931, + "grad_norm": 1.3887627124786377, + "learning_rate": 1.5122623249081139e-05, + "loss": 0.7784, + "step": 8943 + }, + { + "epoch": 0.396001903855307, + "grad_norm": 1.261317253112793, + "learning_rate": 1.5121295909780216e-05, + "loss": 0.6485, + "step": 8944 + }, + { + "epoch": 0.3960461795601209, + "grad_norm": 1.4365452527999878, + "learning_rate": 1.511996844816131e-05, + "loss": 0.7495, + "step": 8945 + }, + { + "epoch": 0.39609045526493475, + "grad_norm": 1.2371770143508911, + "learning_rate": 1.5118640864256126e-05, + "loss": 0.7352, + "step": 8946 + }, + { + "epoch": 0.3961347309697486, + "grad_norm": 1.3926554918289185, + "learning_rate": 1.5117313158096371e-05, + "loss": 0.7905, + "step": 8947 + }, + { + "epoch": 0.3961790066745625, + "grad_norm": 1.4977169036865234, + "learning_rate": 1.5115985329713761e-05, + "loss": 0.7022, + "step": 8948 + }, + { + "epoch": 0.3962232823793764, + "grad_norm": 1.2049425840377808, + "learning_rate": 1.5114657379140006e-05, + "loss": 0.7186, + "step": 8949 + }, + { + "epoch": 0.39626755808419023, + "grad_norm": 1.3711451292037964, + "learning_rate": 1.5113329306406824e-05, + "loss": 0.7344, + "step": 8950 + }, + { + "epoch": 0.39631183378900414, + "grad_norm": 1.3050603866577148, + "learning_rate": 1.5112001111545933e-05, + "loss": 0.7333, + "step": 8951 + }, + { + "epoch": 0.396356109493818, + "grad_norm": 2.4426257610321045, + "learning_rate": 1.5110672794589059e-05, + "loss": 0.8962, + "step": 8952 + }, + { + "epoch": 0.39640038519863185, + "grad_norm": 1.6113762855529785, + "learning_rate": 1.510934435556793e-05, + "loss": 0.7476, + "step": 8953 + }, + { + "epoch": 0.39644466090344577, + "grad_norm": 1.2891387939453125, + "learning_rate": 1.5108015794514268e-05, + "loss": 0.7309, + "step": 8954 + }, + { + "epoch": 0.3964889366082596, + "grad_norm": 1.473245620727539, + "learning_rate": 1.5106687111459809e-05, + "loss": 0.8278, + "step": 8955 + }, + { + "epoch": 0.39653321231307354, + "grad_norm": 2.2626986503601074, + "learning_rate": 1.5105358306436283e-05, + "loss": 0.9462, + "step": 8956 + }, + { + "epoch": 0.3965774880178874, + "grad_norm": 1.4427696466445923, + "learning_rate": 1.5104029379475433e-05, + "loss": 0.7796, + "step": 8957 + }, + { + "epoch": 0.39662176372270125, + "grad_norm": 1.3433544635772705, + "learning_rate": 1.5102700330609e-05, + "loss": 0.7422, + "step": 8958 + }, + { + "epoch": 0.39666603942751516, + "grad_norm": 1.6125022172927856, + "learning_rate": 1.510137115986872e-05, + "loss": 0.805, + "step": 8959 + }, + { + "epoch": 0.396710315132329, + "grad_norm": 1.5321037769317627, + "learning_rate": 1.5100041867286345e-05, + "loss": 0.8815, + "step": 8960 + }, + { + "epoch": 0.3967545908371429, + "grad_norm": 1.4427874088287354, + "learning_rate": 1.5098712452893622e-05, + "loss": 0.7445, + "step": 8961 + }, + { + "epoch": 0.3967988665419568, + "grad_norm": 1.272094488143921, + "learning_rate": 1.5097382916722305e-05, + "loss": 0.7547, + "step": 8962 + }, + { + "epoch": 0.39684314224677064, + "grad_norm": 1.616758108139038, + "learning_rate": 1.5096053258804148e-05, + "loss": 0.7347, + "step": 8963 + }, + { + "epoch": 0.3968874179515845, + "grad_norm": 1.3466202020645142, + "learning_rate": 1.5094723479170905e-05, + "loss": 0.7361, + "step": 8964 + }, + { + "epoch": 0.3969316936563984, + "grad_norm": 1.344995379447937, + "learning_rate": 1.5093393577854345e-05, + "loss": 0.7665, + "step": 8965 + }, + { + "epoch": 0.39697596936121227, + "grad_norm": 1.3014339208602905, + "learning_rate": 1.509206355488622e-05, + "loss": 0.7071, + "step": 8966 + }, + { + "epoch": 0.3970202450660261, + "grad_norm": 1.3839869499206543, + "learning_rate": 1.509073341029831e-05, + "loss": 0.7551, + "step": 8967 + }, + { + "epoch": 0.39706452077084003, + "grad_norm": 1.451768398284912, + "learning_rate": 1.5089403144122375e-05, + "loss": 0.7022, + "step": 8968 + }, + { + "epoch": 0.3971087964756539, + "grad_norm": 1.6901942491531372, + "learning_rate": 1.5088072756390189e-05, + "loss": 0.7737, + "step": 8969 + }, + { + "epoch": 0.39715307218046775, + "grad_norm": 1.4083280563354492, + "learning_rate": 1.5086742247133528e-05, + "loss": 0.7962, + "step": 8970 + }, + { + "epoch": 0.39719734788528166, + "grad_norm": 1.4492682218551636, + "learning_rate": 1.508541161638417e-05, + "loss": 0.7759, + "step": 8971 + }, + { + "epoch": 0.3972416235900955, + "grad_norm": 1.323449730873108, + "learning_rate": 1.5084080864173896e-05, + "loss": 0.7117, + "step": 8972 + }, + { + "epoch": 0.39728589929490943, + "grad_norm": 1.9788073301315308, + "learning_rate": 1.508274999053449e-05, + "loss": 0.7735, + "step": 8973 + }, + { + "epoch": 0.3973301749997233, + "grad_norm": 1.3528575897216797, + "learning_rate": 1.508141899549774e-05, + "loss": 0.8323, + "step": 8974 + }, + { + "epoch": 0.39737445070453714, + "grad_norm": 1.3217705488204956, + "learning_rate": 1.5080087879095437e-05, + "loss": 0.7472, + "step": 8975 + }, + { + "epoch": 0.39741872640935105, + "grad_norm": 1.4695738554000854, + "learning_rate": 1.5078756641359368e-05, + "loss": 0.8669, + "step": 8976 + }, + { + "epoch": 0.3974630021141649, + "grad_norm": 1.6299800872802734, + "learning_rate": 1.5077425282321332e-05, + "loss": 0.7381, + "step": 8977 + }, + { + "epoch": 0.39750727781897877, + "grad_norm": 1.6567798852920532, + "learning_rate": 1.507609380201313e-05, + "loss": 0.7618, + "step": 8978 + }, + { + "epoch": 0.3975515535237927, + "grad_norm": 1.6677742004394531, + "learning_rate": 1.5074762200466557e-05, + "loss": 0.8141, + "step": 8979 + }, + { + "epoch": 0.39759582922860653, + "grad_norm": 1.683702826499939, + "learning_rate": 1.5073430477713424e-05, + "loss": 0.818, + "step": 8980 + }, + { + "epoch": 0.3976401049334204, + "grad_norm": 1.4862730503082275, + "learning_rate": 1.5072098633785534e-05, + "loss": 0.8125, + "step": 8981 + }, + { + "epoch": 0.3976843806382343, + "grad_norm": 1.689511775970459, + "learning_rate": 1.5070766668714699e-05, + "loss": 0.8158, + "step": 8982 + }, + { + "epoch": 0.39772865634304816, + "grad_norm": 1.3331763744354248, + "learning_rate": 1.506943458253273e-05, + "loss": 0.7045, + "step": 8983 + }, + { + "epoch": 0.397772932047862, + "grad_norm": 1.7361642122268677, + "learning_rate": 1.506810237527144e-05, + "loss": 0.9186, + "step": 8984 + }, + { + "epoch": 0.3978172077526759, + "grad_norm": 1.5133572816848755, + "learning_rate": 1.5066770046962655e-05, + "loss": 0.6718, + "step": 8985 + }, + { + "epoch": 0.3978614834574898, + "grad_norm": 1.3194553852081299, + "learning_rate": 1.5065437597638195e-05, + "loss": 0.7143, + "step": 8986 + }, + { + "epoch": 0.39790575916230364, + "grad_norm": 1.4403005838394165, + "learning_rate": 1.5064105027329878e-05, + "loss": 0.8064, + "step": 8987 + }, + { + "epoch": 0.39795003486711755, + "grad_norm": 1.5433237552642822, + "learning_rate": 1.506277233606954e-05, + "loss": 0.8082, + "step": 8988 + }, + { + "epoch": 0.3979943105719314, + "grad_norm": 1.1861114501953125, + "learning_rate": 1.5061439523889003e-05, + "loss": 0.721, + "step": 8989 + }, + { + "epoch": 0.39803858627674527, + "grad_norm": 1.7338835000991821, + "learning_rate": 1.5060106590820107e-05, + "loss": 0.8021, + "step": 8990 + }, + { + "epoch": 0.3980828619815592, + "grad_norm": 1.4034250974655151, + "learning_rate": 1.5058773536894685e-05, + "loss": 0.709, + "step": 8991 + }, + { + "epoch": 0.39812713768637303, + "grad_norm": 1.467728853225708, + "learning_rate": 1.5057440362144573e-05, + "loss": 0.815, + "step": 8992 + }, + { + "epoch": 0.39817141339118695, + "grad_norm": 1.375423789024353, + "learning_rate": 1.505610706660162e-05, + "loss": 0.7781, + "step": 8993 + }, + { + "epoch": 0.3982156890960008, + "grad_norm": 1.2883599996566772, + "learning_rate": 1.5054773650297663e-05, + "loss": 0.7647, + "step": 8994 + }, + { + "epoch": 0.39825996480081466, + "grad_norm": 1.4960416555404663, + "learning_rate": 1.5053440113264555e-05, + "loss": 0.7333, + "step": 8995 + }, + { + "epoch": 0.39830424050562857, + "grad_norm": 1.3257611989974976, + "learning_rate": 1.5052106455534145e-05, + "loss": 0.8261, + "step": 8996 + }, + { + "epoch": 0.3983485162104424, + "grad_norm": 1.5329803228378296, + "learning_rate": 1.5050772677138286e-05, + "loss": 0.7633, + "step": 8997 + }, + { + "epoch": 0.3983927919152563, + "grad_norm": 1.4149508476257324, + "learning_rate": 1.5049438778108832e-05, + "loss": 0.7122, + "step": 8998 + }, + { + "epoch": 0.3984370676200702, + "grad_norm": 1.6347317695617676, + "learning_rate": 1.5048104758477643e-05, + "loss": 0.8117, + "step": 8999 + }, + { + "epoch": 0.39848134332488405, + "grad_norm": 1.4464261531829834, + "learning_rate": 1.5046770618276586e-05, + "loss": 0.8078, + "step": 9000 + }, + { + "epoch": 0.3985256190296979, + "grad_norm": 1.3875949382781982, + "learning_rate": 1.5045436357537523e-05, + "loss": 0.7605, + "step": 9001 + }, + { + "epoch": 0.3985698947345118, + "grad_norm": 1.4316608905792236, + "learning_rate": 1.5044101976292318e-05, + "loss": 0.7769, + "step": 9002 + }, + { + "epoch": 0.3986141704393257, + "grad_norm": 1.608772873878479, + "learning_rate": 1.5042767474572846e-05, + "loss": 0.8081, + "step": 9003 + }, + { + "epoch": 0.39865844614413953, + "grad_norm": 1.3724044561386108, + "learning_rate": 1.5041432852410975e-05, + "loss": 0.7819, + "step": 9004 + }, + { + "epoch": 0.39870272184895345, + "grad_norm": 1.2877321243286133, + "learning_rate": 1.5040098109838592e-05, + "loss": 0.7503, + "step": 9005 + }, + { + "epoch": 0.3987469975537673, + "grad_norm": 1.2064025402069092, + "learning_rate": 1.5038763246887564e-05, + "loss": 0.7301, + "step": 9006 + }, + { + "epoch": 0.39879127325858116, + "grad_norm": 1.3958494663238525, + "learning_rate": 1.5037428263589778e-05, + "loss": 0.7596, + "step": 9007 + }, + { + "epoch": 0.39883554896339507, + "grad_norm": 1.4982362985610962, + "learning_rate": 1.5036093159977121e-05, + "loss": 0.8501, + "step": 9008 + }, + { + "epoch": 0.3988798246682089, + "grad_norm": 1.4599016904830933, + "learning_rate": 1.5034757936081482e-05, + "loss": 0.7698, + "step": 9009 + }, + { + "epoch": 0.39892410037302284, + "grad_norm": 1.5350570678710938, + "learning_rate": 1.5033422591934748e-05, + "loss": 0.7828, + "step": 9010 + }, + { + "epoch": 0.3989683760778367, + "grad_norm": 1.3289203643798828, + "learning_rate": 1.503208712756881e-05, + "loss": 0.769, + "step": 9011 + }, + { + "epoch": 0.39901265178265055, + "grad_norm": 1.278213620185852, + "learning_rate": 1.503075154301557e-05, + "loss": 0.7516, + "step": 9012 + }, + { + "epoch": 0.39905692748746446, + "grad_norm": 1.2765978574752808, + "learning_rate": 1.5029415838306927e-05, + "loss": 0.7555, + "step": 9013 + }, + { + "epoch": 0.3991012031922783, + "grad_norm": 1.635905146598816, + "learning_rate": 1.5028080013474781e-05, + "loss": 0.7303, + "step": 9014 + }, + { + "epoch": 0.3991454788970922, + "grad_norm": 1.481055498123169, + "learning_rate": 1.5026744068551042e-05, + "loss": 0.7436, + "step": 9015 + }, + { + "epoch": 0.3991897546019061, + "grad_norm": 1.4223445653915405, + "learning_rate": 1.5025408003567608e-05, + "loss": 0.8051, + "step": 9016 + }, + { + "epoch": 0.39923403030671994, + "grad_norm": 1.346092939376831, + "learning_rate": 1.5024071818556396e-05, + "loss": 0.6996, + "step": 9017 + }, + { + "epoch": 0.3992783060115338, + "grad_norm": 1.386825442314148, + "learning_rate": 1.5022735513549322e-05, + "loss": 0.767, + "step": 9018 + }, + { + "epoch": 0.3993225817163477, + "grad_norm": 1.765293002128601, + "learning_rate": 1.5021399088578302e-05, + "loss": 0.8691, + "step": 9019 + }, + { + "epoch": 0.39936685742116157, + "grad_norm": 1.410964846611023, + "learning_rate": 1.502006254367525e-05, + "loss": 0.7282, + "step": 9020 + }, + { + "epoch": 0.3994111331259754, + "grad_norm": 1.8482284545898438, + "learning_rate": 1.501872587887209e-05, + "loss": 0.7768, + "step": 9021 + }, + { + "epoch": 0.39945540883078934, + "grad_norm": 1.5525872707366943, + "learning_rate": 1.501738909420075e-05, + "loss": 0.7521, + "step": 9022 + }, + { + "epoch": 0.3994996845356032, + "grad_norm": 1.4746099710464478, + "learning_rate": 1.5016052189693158e-05, + "loss": 0.8921, + "step": 9023 + }, + { + "epoch": 0.39954396024041705, + "grad_norm": 1.2292423248291016, + "learning_rate": 1.5014715165381245e-05, + "loss": 0.6789, + "step": 9024 + }, + { + "epoch": 0.39958823594523096, + "grad_norm": 2.0675017833709717, + "learning_rate": 1.5013378021296943e-05, + "loss": 0.8537, + "step": 9025 + }, + { + "epoch": 0.3996325116500448, + "grad_norm": 1.6120831966400146, + "learning_rate": 1.5012040757472184e-05, + "loss": 0.7994, + "step": 9026 + }, + { + "epoch": 0.39967678735485873, + "grad_norm": 1.4086042642593384, + "learning_rate": 1.5010703373938915e-05, + "loss": 0.7209, + "step": 9027 + }, + { + "epoch": 0.3997210630596726, + "grad_norm": 1.6862680912017822, + "learning_rate": 1.500936587072908e-05, + "loss": 0.8371, + "step": 9028 + }, + { + "epoch": 0.39976533876448644, + "grad_norm": 1.3127212524414062, + "learning_rate": 1.5008028247874618e-05, + "loss": 0.6731, + "step": 9029 + }, + { + "epoch": 0.39980961446930036, + "grad_norm": 1.252052664756775, + "learning_rate": 1.5006690505407477e-05, + "loss": 0.7572, + "step": 9030 + }, + { + "epoch": 0.3998538901741142, + "grad_norm": 1.271614909172058, + "learning_rate": 1.500535264335961e-05, + "loss": 0.79, + "step": 9031 + }, + { + "epoch": 0.39989816587892807, + "grad_norm": 1.236358642578125, + "learning_rate": 1.500401466176297e-05, + "loss": 0.6851, + "step": 9032 + }, + { + "epoch": 0.399942441583742, + "grad_norm": 1.27528715133667, + "learning_rate": 1.5002676560649518e-05, + "loss": 0.8296, + "step": 9033 + }, + { + "epoch": 0.39998671728855584, + "grad_norm": 1.376345157623291, + "learning_rate": 1.5001338340051206e-05, + "loss": 0.748, + "step": 9034 + }, + { + "epoch": 0.4000309929933697, + "grad_norm": 1.305248737335205, + "learning_rate": 1.5000000000000002e-05, + "loss": 0.7302, + "step": 9035 + }, + { + "epoch": 0.4000752686981836, + "grad_norm": 1.5199120044708252, + "learning_rate": 1.4998661540527867e-05, + "loss": 0.8078, + "step": 9036 + }, + { + "epoch": 0.40011954440299746, + "grad_norm": 1.581479549407959, + "learning_rate": 1.4997322961666774e-05, + "loss": 0.7661, + "step": 9037 + }, + { + "epoch": 0.4001638201078113, + "grad_norm": 1.3072501420974731, + "learning_rate": 1.4995984263448691e-05, + "loss": 0.76, + "step": 9038 + }, + { + "epoch": 0.40020809581262523, + "grad_norm": 1.666640281677246, + "learning_rate": 1.499464544590559e-05, + "loss": 0.7954, + "step": 9039 + }, + { + "epoch": 0.4002523715174391, + "grad_norm": 1.4546326398849487, + "learning_rate": 1.499330650906945e-05, + "loss": 0.7525, + "step": 9040 + }, + { + "epoch": 0.40029664722225294, + "grad_norm": 1.4124119281768799, + "learning_rate": 1.4991967452972247e-05, + "loss": 0.7817, + "step": 9041 + }, + { + "epoch": 0.40034092292706686, + "grad_norm": 1.3050099611282349, + "learning_rate": 1.4990628277645969e-05, + "loss": 0.7229, + "step": 9042 + }, + { + "epoch": 0.4003851986318807, + "grad_norm": 1.3276090621948242, + "learning_rate": 1.4989288983122598e-05, + "loss": 0.6821, + "step": 9043 + }, + { + "epoch": 0.4004294743366946, + "grad_norm": 1.5392223596572876, + "learning_rate": 1.4987949569434121e-05, + "loss": 0.7375, + "step": 9044 + }, + { + "epoch": 0.4004737500415085, + "grad_norm": 1.2941659688949585, + "learning_rate": 1.498661003661253e-05, + "loss": 0.7307, + "step": 9045 + }, + { + "epoch": 0.40051802574632234, + "grad_norm": 1.255847692489624, + "learning_rate": 1.4985270384689818e-05, + "loss": 0.6373, + "step": 9046 + }, + { + "epoch": 0.40056230145113625, + "grad_norm": 1.3299976587295532, + "learning_rate": 1.4983930613697984e-05, + "loss": 0.7176, + "step": 9047 + }, + { + "epoch": 0.4006065771559501, + "grad_norm": 1.4057130813598633, + "learning_rate": 1.4982590723669026e-05, + "loss": 0.6791, + "step": 9048 + }, + { + "epoch": 0.40065085286076396, + "grad_norm": 1.7761623859405518, + "learning_rate": 1.4981250714634943e-05, + "loss": 0.7954, + "step": 9049 + }, + { + "epoch": 0.4006951285655779, + "grad_norm": 1.5957354307174683, + "learning_rate": 1.4979910586627744e-05, + "loss": 0.7165, + "step": 9050 + }, + { + "epoch": 0.40073940427039173, + "grad_norm": 1.309624195098877, + "learning_rate": 1.4978570339679435e-05, + "loss": 0.7907, + "step": 9051 + }, + { + "epoch": 0.4007836799752056, + "grad_norm": 1.4418123960494995, + "learning_rate": 1.4977229973822027e-05, + "loss": 0.7675, + "step": 9052 + }, + { + "epoch": 0.4008279556800195, + "grad_norm": 1.460964322090149, + "learning_rate": 1.4975889489087534e-05, + "loss": 0.7189, + "step": 9053 + }, + { + "epoch": 0.40087223138483336, + "grad_norm": 1.1945828199386597, + "learning_rate": 1.4974548885507969e-05, + "loss": 0.6933, + "step": 9054 + }, + { + "epoch": 0.4009165070896472, + "grad_norm": 1.0997003316879272, + "learning_rate": 1.497320816311536e-05, + "loss": 0.6443, + "step": 9055 + }, + { + "epoch": 0.4009607827944611, + "grad_norm": 1.2551653385162354, + "learning_rate": 1.4971867321941717e-05, + "loss": 0.7431, + "step": 9056 + }, + { + "epoch": 0.401005058499275, + "grad_norm": 1.3800534009933472, + "learning_rate": 1.4970526362019077e-05, + "loss": 0.7524, + "step": 9057 + }, + { + "epoch": 0.40104933420408884, + "grad_norm": 1.1968246698379517, + "learning_rate": 1.496918528337946e-05, + "loss": 0.7026, + "step": 9058 + }, + { + "epoch": 0.40109360990890275, + "grad_norm": 1.3885337114334106, + "learning_rate": 1.4967844086054896e-05, + "loss": 0.8068, + "step": 9059 + }, + { + "epoch": 0.4011378856137166, + "grad_norm": 1.3351384401321411, + "learning_rate": 1.4966502770077424e-05, + "loss": 0.7236, + "step": 9060 + }, + { + "epoch": 0.4011821613185305, + "grad_norm": 1.5328104496002197, + "learning_rate": 1.496516133547908e-05, + "loss": 0.7122, + "step": 9061 + }, + { + "epoch": 0.4012264370233444, + "grad_norm": 1.2422343492507935, + "learning_rate": 1.4963819782291897e-05, + "loss": 0.8032, + "step": 9062 + }, + { + "epoch": 0.40127071272815823, + "grad_norm": 1.126015067100525, + "learning_rate": 1.4962478110547918e-05, + "loss": 0.6982, + "step": 9063 + }, + { + "epoch": 0.40131498843297214, + "grad_norm": 1.2699003219604492, + "learning_rate": 1.496113632027919e-05, + "loss": 0.7288, + "step": 9064 + }, + { + "epoch": 0.401359264137786, + "grad_norm": 1.5280927419662476, + "learning_rate": 1.4959794411517763e-05, + "loss": 0.7471, + "step": 9065 + }, + { + "epoch": 0.40140353984259985, + "grad_norm": 1.446340560913086, + "learning_rate": 1.4958452384295685e-05, + "loss": 0.7809, + "step": 9066 + }, + { + "epoch": 0.40144781554741377, + "grad_norm": 1.3706530332565308, + "learning_rate": 1.495711023864501e-05, + "loss": 0.8439, + "step": 9067 + }, + { + "epoch": 0.4014920912522276, + "grad_norm": 1.3657548427581787, + "learning_rate": 1.495576797459779e-05, + "loss": 0.7583, + "step": 9068 + }, + { + "epoch": 0.4015363669570415, + "grad_norm": 1.5263185501098633, + "learning_rate": 1.4954425592186089e-05, + "loss": 0.7471, + "step": 9069 + }, + { + "epoch": 0.4015806426618554, + "grad_norm": 1.8063349723815918, + "learning_rate": 1.4953083091441966e-05, + "loss": 0.7649, + "step": 9070 + }, + { + "epoch": 0.40162491836666925, + "grad_norm": 1.448663353919983, + "learning_rate": 1.4951740472397487e-05, + "loss": 0.8167, + "step": 9071 + }, + { + "epoch": 0.4016691940714831, + "grad_norm": 1.275307536125183, + "learning_rate": 1.4950397735084718e-05, + "loss": 0.7322, + "step": 9072 + }, + { + "epoch": 0.401713469776297, + "grad_norm": 1.5464117527008057, + "learning_rate": 1.4949054879535732e-05, + "loss": 0.8578, + "step": 9073 + }, + { + "epoch": 0.4017577454811109, + "grad_norm": 1.2816016674041748, + "learning_rate": 1.4947711905782596e-05, + "loss": 0.7026, + "step": 9074 + }, + { + "epoch": 0.40180202118592473, + "grad_norm": 1.226401448249817, + "learning_rate": 1.4946368813857393e-05, + "loss": 0.7419, + "step": 9075 + }, + { + "epoch": 0.40184629689073864, + "grad_norm": 1.2845784425735474, + "learning_rate": 1.4945025603792198e-05, + "loss": 0.7364, + "step": 9076 + }, + { + "epoch": 0.4018905725955525, + "grad_norm": 1.203636646270752, + "learning_rate": 1.4943682275619091e-05, + "loss": 0.6975, + "step": 9077 + }, + { + "epoch": 0.40193484830036635, + "grad_norm": 1.7760032415390015, + "learning_rate": 1.494233882937016e-05, + "loss": 0.7712, + "step": 9078 + }, + { + "epoch": 0.40197912400518027, + "grad_norm": 1.4585942029953003, + "learning_rate": 1.4940995265077486e-05, + "loss": 0.7636, + "step": 9079 + }, + { + "epoch": 0.4020233997099941, + "grad_norm": 1.807771921157837, + "learning_rate": 1.4939651582773172e-05, + "loss": 0.832, + "step": 9080 + }, + { + "epoch": 0.40206767541480803, + "grad_norm": 2.039616346359253, + "learning_rate": 1.4938307782489294e-05, + "loss": 0.7322, + "step": 9081 + }, + { + "epoch": 0.4021119511196219, + "grad_norm": 1.5026532411575317, + "learning_rate": 1.4936963864257958e-05, + "loss": 0.7993, + "step": 9082 + }, + { + "epoch": 0.40215622682443575, + "grad_norm": 1.1771471500396729, + "learning_rate": 1.4935619828111261e-05, + "loss": 0.7757, + "step": 9083 + }, + { + "epoch": 0.40220050252924966, + "grad_norm": 1.438299536705017, + "learning_rate": 1.49342756740813e-05, + "loss": 0.6926, + "step": 9084 + }, + { + "epoch": 0.4022447782340635, + "grad_norm": 1.4247442483901978, + "learning_rate": 1.4932931402200189e-05, + "loss": 0.7338, + "step": 9085 + }, + { + "epoch": 0.4022890539388774, + "grad_norm": 1.496221661567688, + "learning_rate": 1.493158701250002e-05, + "loss": 0.7971, + "step": 9086 + }, + { + "epoch": 0.4023333296436913, + "grad_norm": 1.4677778482437134, + "learning_rate": 1.4930242505012915e-05, + "loss": 0.7761, + "step": 9087 + }, + { + "epoch": 0.40237760534850514, + "grad_norm": 1.1953272819519043, + "learning_rate": 1.4928897879770981e-05, + "loss": 0.6721, + "step": 9088 + }, + { + "epoch": 0.402421881053319, + "grad_norm": 1.91411554813385, + "learning_rate": 1.4927553136806335e-05, + "loss": 0.8382, + "step": 9089 + }, + { + "epoch": 0.4024661567581329, + "grad_norm": 1.18524169921875, + "learning_rate": 1.4926208276151096e-05, + "loss": 0.6952, + "step": 9090 + }, + { + "epoch": 0.40251043246294677, + "grad_norm": 1.47810959815979, + "learning_rate": 1.4924863297837378e-05, + "loss": 0.7266, + "step": 9091 + }, + { + "epoch": 0.4025547081677606, + "grad_norm": 1.42555570602417, + "learning_rate": 1.4923518201897316e-05, + "loss": 0.7723, + "step": 9092 + }, + { + "epoch": 0.40259898387257453, + "grad_norm": 1.817112684249878, + "learning_rate": 1.4922172988363026e-05, + "loss": 0.7913, + "step": 9093 + }, + { + "epoch": 0.4026432595773884, + "grad_norm": 1.2957414388656616, + "learning_rate": 1.4920827657266644e-05, + "loss": 0.7017, + "step": 9094 + }, + { + "epoch": 0.40268753528220225, + "grad_norm": 1.338269829750061, + "learning_rate": 1.4919482208640301e-05, + "loss": 0.7991, + "step": 9095 + }, + { + "epoch": 0.40273181098701616, + "grad_norm": 1.4350124597549438, + "learning_rate": 1.491813664251613e-05, + "loss": 0.7565, + "step": 9096 + }, + { + "epoch": 0.40277608669183, + "grad_norm": 1.612248182296753, + "learning_rate": 1.491679095892627e-05, + "loss": 0.8064, + "step": 9097 + }, + { + "epoch": 0.4028203623966439, + "grad_norm": 1.507407546043396, + "learning_rate": 1.491544515790286e-05, + "loss": 0.8241, + "step": 9098 + }, + { + "epoch": 0.4028646381014578, + "grad_norm": 1.5251175165176392, + "learning_rate": 1.4914099239478046e-05, + "loss": 0.6969, + "step": 9099 + }, + { + "epoch": 0.40290891380627164, + "grad_norm": 1.1606979370117188, + "learning_rate": 1.4912753203683974e-05, + "loss": 0.67, + "step": 9100 + }, + { + "epoch": 0.40295318951108555, + "grad_norm": 1.4329441785812378, + "learning_rate": 1.4911407050552792e-05, + "loss": 0.7473, + "step": 9101 + }, + { + "epoch": 0.4029974652158994, + "grad_norm": 1.1798949241638184, + "learning_rate": 1.4910060780116649e-05, + "loss": 0.6125, + "step": 9102 + }, + { + "epoch": 0.40304174092071327, + "grad_norm": 1.5107582807540894, + "learning_rate": 1.4908714392407703e-05, + "loss": 0.8109, + "step": 9103 + }, + { + "epoch": 0.4030860166255272, + "grad_norm": 1.4065176248550415, + "learning_rate": 1.490736788745811e-05, + "loss": 0.7578, + "step": 9104 + }, + { + "epoch": 0.40313029233034103, + "grad_norm": 1.4495997428894043, + "learning_rate": 1.4906021265300035e-05, + "loss": 0.7355, + "step": 9105 + }, + { + "epoch": 0.4031745680351549, + "grad_norm": 1.4266319274902344, + "learning_rate": 1.4904674525965632e-05, + "loss": 0.707, + "step": 9106 + }, + { + "epoch": 0.4032188437399688, + "grad_norm": 1.5946249961853027, + "learning_rate": 1.4903327669487073e-05, + "loss": 0.8102, + "step": 9107 + }, + { + "epoch": 0.40326311944478266, + "grad_norm": 1.3599108457565308, + "learning_rate": 1.4901980695896527e-05, + "loss": 0.7946, + "step": 9108 + }, + { + "epoch": 0.4033073951495965, + "grad_norm": 1.4408568143844604, + "learning_rate": 1.490063360522616e-05, + "loss": 0.7099, + "step": 9109 + }, + { + "epoch": 0.4033516708544104, + "grad_norm": 1.494365930557251, + "learning_rate": 1.4899286397508152e-05, + "loss": 0.702, + "step": 9110 + }, + { + "epoch": 0.4033959465592243, + "grad_norm": 1.684946894645691, + "learning_rate": 1.4897939072774677e-05, + "loss": 0.8053, + "step": 9111 + }, + { + "epoch": 0.40344022226403814, + "grad_norm": 1.2194079160690308, + "learning_rate": 1.4896591631057912e-05, + "loss": 0.6923, + "step": 9112 + }, + { + "epoch": 0.40348449796885205, + "grad_norm": 1.122656226158142, + "learning_rate": 1.489524407239005e-05, + "loss": 0.7312, + "step": 9113 + }, + { + "epoch": 0.4035287736736659, + "grad_norm": 1.5019418001174927, + "learning_rate": 1.4893896396803262e-05, + "loss": 0.7977, + "step": 9114 + }, + { + "epoch": 0.4035730493784798, + "grad_norm": 1.3194819688796997, + "learning_rate": 1.4892548604329746e-05, + "loss": 0.6936, + "step": 9115 + }, + { + "epoch": 0.4036173250832937, + "grad_norm": 1.559305191040039, + "learning_rate": 1.4891200695001691e-05, + "loss": 0.7533, + "step": 9116 + }, + { + "epoch": 0.40366160078810753, + "grad_norm": 1.3968696594238281, + "learning_rate": 1.488985266885129e-05, + "loss": 0.6669, + "step": 9117 + }, + { + "epoch": 0.40370587649292144, + "grad_norm": 1.5508536100387573, + "learning_rate": 1.488850452591074e-05, + "loss": 0.7765, + "step": 9118 + }, + { + "epoch": 0.4037501521977353, + "grad_norm": 1.2599766254425049, + "learning_rate": 1.4887156266212237e-05, + "loss": 0.7287, + "step": 9119 + }, + { + "epoch": 0.40379442790254916, + "grad_norm": 1.4762035608291626, + "learning_rate": 1.4885807889787989e-05, + "loss": 0.7146, + "step": 9120 + }, + { + "epoch": 0.40383870360736307, + "grad_norm": 1.702641248703003, + "learning_rate": 1.4884459396670194e-05, + "loss": 0.8213, + "step": 9121 + }, + { + "epoch": 0.4038829793121769, + "grad_norm": 1.5104793310165405, + "learning_rate": 1.4883110786891069e-05, + "loss": 0.7847, + "step": 9122 + }, + { + "epoch": 0.4039272550169908, + "grad_norm": 1.2072298526763916, + "learning_rate": 1.4881762060482814e-05, + "loss": 0.6644, + "step": 9123 + }, + { + "epoch": 0.4039715307218047, + "grad_norm": 1.5650206804275513, + "learning_rate": 1.488041321747765e-05, + "loss": 0.7523, + "step": 9124 + }, + { + "epoch": 0.40401580642661855, + "grad_norm": 1.415351390838623, + "learning_rate": 1.487906425790779e-05, + "loss": 0.7297, + "step": 9125 + }, + { + "epoch": 0.4040600821314324, + "grad_norm": 1.3343560695648193, + "learning_rate": 1.4877715181805448e-05, + "loss": 0.6998, + "step": 9126 + }, + { + "epoch": 0.4041043578362463, + "grad_norm": 1.3196191787719727, + "learning_rate": 1.4876365989202855e-05, + "loss": 0.7486, + "step": 9127 + }, + { + "epoch": 0.4041486335410602, + "grad_norm": 1.313088059425354, + "learning_rate": 1.487501668013223e-05, + "loss": 0.7479, + "step": 9128 + }, + { + "epoch": 0.40419290924587403, + "grad_norm": 1.6312763690948486, + "learning_rate": 1.4873667254625801e-05, + "loss": 0.8093, + "step": 9129 + }, + { + "epoch": 0.40423718495068794, + "grad_norm": 1.3005305528640747, + "learning_rate": 1.4872317712715799e-05, + "loss": 0.7857, + "step": 9130 + }, + { + "epoch": 0.4042814606555018, + "grad_norm": 1.2749288082122803, + "learning_rate": 1.4870968054434452e-05, + "loss": 0.7398, + "step": 9131 + }, + { + "epoch": 0.4043257363603157, + "grad_norm": 1.1646963357925415, + "learning_rate": 1.4869618279814004e-05, + "loss": 0.7022, + "step": 9132 + }, + { + "epoch": 0.40437001206512957, + "grad_norm": 1.539525032043457, + "learning_rate": 1.4868268388886686e-05, + "loss": 0.8463, + "step": 9133 + }, + { + "epoch": 0.4044142877699434, + "grad_norm": 1.6369861364364624, + "learning_rate": 1.4866918381684738e-05, + "loss": 0.6755, + "step": 9134 + }, + { + "epoch": 0.40445856347475734, + "grad_norm": 1.4434431791305542, + "learning_rate": 1.4865568258240413e-05, + "loss": 0.7487, + "step": 9135 + }, + { + "epoch": 0.4045028391795712, + "grad_norm": 1.360253095626831, + "learning_rate": 1.486421801858595e-05, + "loss": 0.713, + "step": 9136 + }, + { + "epoch": 0.40454711488438505, + "grad_norm": 1.4392540454864502, + "learning_rate": 1.4862867662753599e-05, + "loss": 0.7439, + "step": 9137 + }, + { + "epoch": 0.40459139058919896, + "grad_norm": 1.3959475755691528, + "learning_rate": 1.4861517190775614e-05, + "loss": 0.7507, + "step": 9138 + }, + { + "epoch": 0.4046356662940128, + "grad_norm": 1.4118640422821045, + "learning_rate": 1.4860166602684248e-05, + "loss": 0.7704, + "step": 9139 + }, + { + "epoch": 0.4046799419988267, + "grad_norm": 1.289539098739624, + "learning_rate": 1.4858815898511763e-05, + "loss": 0.7303, + "step": 9140 + }, + { + "epoch": 0.4047242177036406, + "grad_norm": 1.5404688119888306, + "learning_rate": 1.4857465078290414e-05, + "loss": 0.7189, + "step": 9141 + }, + { + "epoch": 0.40476849340845444, + "grad_norm": 1.3439202308654785, + "learning_rate": 1.4856114142052468e-05, + "loss": 0.7367, + "step": 9142 + }, + { + "epoch": 0.4048127691132683, + "grad_norm": 1.2301403284072876, + "learning_rate": 1.4854763089830188e-05, + "loss": 0.7573, + "step": 9143 + }, + { + "epoch": 0.4048570448180822, + "grad_norm": 1.2456550598144531, + "learning_rate": 1.4853411921655844e-05, + "loss": 0.7377, + "step": 9144 + }, + { + "epoch": 0.40490132052289607, + "grad_norm": 1.3997679948806763, + "learning_rate": 1.4852060637561711e-05, + "loss": 0.7567, + "step": 9145 + }, + { + "epoch": 0.4049455962277099, + "grad_norm": 1.4307094812393188, + "learning_rate": 1.4850709237580061e-05, + "loss": 0.8157, + "step": 9146 + }, + { + "epoch": 0.40498987193252384, + "grad_norm": 1.4900704622268677, + "learning_rate": 1.4849357721743169e-05, + "loss": 0.7985, + "step": 9147 + }, + { + "epoch": 0.4050341476373377, + "grad_norm": 1.4848381280899048, + "learning_rate": 1.4848006090083313e-05, + "loss": 0.7233, + "step": 9148 + }, + { + "epoch": 0.4050784233421516, + "grad_norm": 1.3782860040664673, + "learning_rate": 1.4846654342632783e-05, + "loss": 0.7519, + "step": 9149 + }, + { + "epoch": 0.40512269904696546, + "grad_norm": 1.4693028926849365, + "learning_rate": 1.4845302479423857e-05, + "loss": 0.7564, + "step": 9150 + }, + { + "epoch": 0.4051669747517793, + "grad_norm": 1.212257981300354, + "learning_rate": 1.4843950500488831e-05, + "loss": 0.7019, + "step": 9151 + }, + { + "epoch": 0.40521125045659323, + "grad_norm": 1.470763087272644, + "learning_rate": 1.484259840585999e-05, + "loss": 0.7039, + "step": 9152 + }, + { + "epoch": 0.4052555261614071, + "grad_norm": 1.408071517944336, + "learning_rate": 1.4841246195569628e-05, + "loss": 0.6984, + "step": 9153 + }, + { + "epoch": 0.40529980186622094, + "grad_norm": 2.323901891708374, + "learning_rate": 1.4839893869650039e-05, + "loss": 0.802, + "step": 9154 + }, + { + "epoch": 0.40534407757103486, + "grad_norm": 1.3420042991638184, + "learning_rate": 1.483854142813353e-05, + "loss": 0.7065, + "step": 9155 + }, + { + "epoch": 0.4053883532758487, + "grad_norm": 1.373279333114624, + "learning_rate": 1.4837188871052399e-05, + "loss": 0.7671, + "step": 9156 + }, + { + "epoch": 0.40543262898066257, + "grad_norm": 1.438672661781311, + "learning_rate": 1.4835836198438952e-05, + "loss": 0.82, + "step": 9157 + }, + { + "epoch": 0.4054769046854765, + "grad_norm": 1.5257078409194946, + "learning_rate": 1.483448341032549e-05, + "loss": 0.7211, + "step": 9158 + }, + { + "epoch": 0.40552118039029034, + "grad_norm": 1.44582200050354, + "learning_rate": 1.483313050674433e-05, + "loss": 0.7243, + "step": 9159 + }, + { + "epoch": 0.4055654560951042, + "grad_norm": 1.4563649892807007, + "learning_rate": 1.4831777487727786e-05, + "loss": 0.6878, + "step": 9160 + }, + { + "epoch": 0.4056097317999181, + "grad_norm": 1.6078689098358154, + "learning_rate": 1.4830424353308166e-05, + "loss": 0.851, + "step": 9161 + }, + { + "epoch": 0.40565400750473196, + "grad_norm": 1.5983978509902954, + "learning_rate": 1.4829071103517796e-05, + "loss": 0.814, + "step": 9162 + }, + { + "epoch": 0.4056982832095458, + "grad_norm": 1.5282135009765625, + "learning_rate": 1.4827717738388993e-05, + "loss": 0.7219, + "step": 9163 + }, + { + "epoch": 0.40574255891435973, + "grad_norm": 1.3716665506362915, + "learning_rate": 1.4826364257954086e-05, + "loss": 0.7255, + "step": 9164 + }, + { + "epoch": 0.4057868346191736, + "grad_norm": 1.5497764348983765, + "learning_rate": 1.4825010662245397e-05, + "loss": 0.8466, + "step": 9165 + }, + { + "epoch": 0.40583111032398744, + "grad_norm": 1.3631377220153809, + "learning_rate": 1.4823656951295256e-05, + "loss": 0.7798, + "step": 9166 + }, + { + "epoch": 0.40587538602880135, + "grad_norm": 1.5182486772537231, + "learning_rate": 1.4822303125135998e-05, + "loss": 0.8265, + "step": 9167 + }, + { + "epoch": 0.4059196617336152, + "grad_norm": 1.494674801826477, + "learning_rate": 1.4820949183799952e-05, + "loss": 0.7158, + "step": 9168 + }, + { + "epoch": 0.4059639374384291, + "grad_norm": 1.6419485807418823, + "learning_rate": 1.4819595127319465e-05, + "loss": 0.7934, + "step": 9169 + }, + { + "epoch": 0.406008213143243, + "grad_norm": 1.4878249168395996, + "learning_rate": 1.4818240955726872e-05, + "loss": 0.8006, + "step": 9170 + }, + { + "epoch": 0.40605248884805684, + "grad_norm": 1.1383999586105347, + "learning_rate": 1.4816886669054514e-05, + "loss": 0.6295, + "step": 9171 + }, + { + "epoch": 0.40609676455287075, + "grad_norm": 1.25419020652771, + "learning_rate": 1.4815532267334742e-05, + "loss": 0.7723, + "step": 9172 + }, + { + "epoch": 0.4061410402576846, + "grad_norm": 1.452587604522705, + "learning_rate": 1.4814177750599901e-05, + "loss": 0.7879, + "step": 9173 + }, + { + "epoch": 0.40618531596249846, + "grad_norm": 1.3199667930603027, + "learning_rate": 1.4812823118882348e-05, + "loss": 0.7283, + "step": 9174 + }, + { + "epoch": 0.4062295916673124, + "grad_norm": 1.4207755327224731, + "learning_rate": 1.4811468372214432e-05, + "loss": 0.7349, + "step": 9175 + }, + { + "epoch": 0.40627386737212623, + "grad_norm": 1.4507869482040405, + "learning_rate": 1.481011351062851e-05, + "loss": 0.7538, + "step": 9176 + }, + { + "epoch": 0.4063181430769401, + "grad_norm": 1.6286399364471436, + "learning_rate": 1.4808758534156943e-05, + "loss": 0.7946, + "step": 9177 + }, + { + "epoch": 0.406362418781754, + "grad_norm": 1.3140710592269897, + "learning_rate": 1.4807403442832096e-05, + "loss": 0.7218, + "step": 9178 + }, + { + "epoch": 0.40640669448656785, + "grad_norm": 1.2280751466751099, + "learning_rate": 1.4806048236686335e-05, + "loss": 0.7325, + "step": 9179 + }, + { + "epoch": 0.4064509701913817, + "grad_norm": 1.319811224937439, + "learning_rate": 1.4804692915752018e-05, + "loss": 0.7774, + "step": 9180 + }, + { + "epoch": 0.4064952458961956, + "grad_norm": 1.3596099615097046, + "learning_rate": 1.4803337480061526e-05, + "loss": 0.7819, + "step": 9181 + }, + { + "epoch": 0.4065395216010095, + "grad_norm": 1.4783939123153687, + "learning_rate": 1.480198192964723e-05, + "loss": 0.7461, + "step": 9182 + }, + { + "epoch": 0.40658379730582334, + "grad_norm": 1.8651976585388184, + "learning_rate": 1.4800626264541503e-05, + "loss": 0.8089, + "step": 9183 + }, + { + "epoch": 0.40662807301063725, + "grad_norm": 1.2543485164642334, + "learning_rate": 1.4799270484776731e-05, + "loss": 0.7458, + "step": 9184 + }, + { + "epoch": 0.4066723487154511, + "grad_norm": 1.1535875797271729, + "learning_rate": 1.4797914590385287e-05, + "loss": 0.6504, + "step": 9185 + }, + { + "epoch": 0.406716624420265, + "grad_norm": 1.1949362754821777, + "learning_rate": 1.4796558581399561e-05, + "loss": 0.7044, + "step": 9186 + }, + { + "epoch": 0.4067609001250789, + "grad_norm": 1.3635280132293701, + "learning_rate": 1.479520245785194e-05, + "loss": 0.7552, + "step": 9187 + }, + { + "epoch": 0.40680517582989273, + "grad_norm": 1.2790868282318115, + "learning_rate": 1.4793846219774812e-05, + "loss": 0.7418, + "step": 9188 + }, + { + "epoch": 0.40684945153470664, + "grad_norm": 1.336472749710083, + "learning_rate": 1.479248986720057e-05, + "loss": 0.7457, + "step": 9189 + }, + { + "epoch": 0.4068937272395205, + "grad_norm": 1.350765585899353, + "learning_rate": 1.479113340016161e-05, + "loss": 0.7523, + "step": 9190 + }, + { + "epoch": 0.40693800294433435, + "grad_norm": 1.3341583013534546, + "learning_rate": 1.4789776818690328e-05, + "loss": 0.7093, + "step": 9191 + }, + { + "epoch": 0.40698227864914827, + "grad_norm": 1.3156726360321045, + "learning_rate": 1.4788420122819129e-05, + "loss": 0.729, + "step": 9192 + }, + { + "epoch": 0.4070265543539621, + "grad_norm": 1.2055000066757202, + "learning_rate": 1.4787063312580414e-05, + "loss": 0.7045, + "step": 9193 + }, + { + "epoch": 0.407070830058776, + "grad_norm": 1.3470605611801147, + "learning_rate": 1.478570638800659e-05, + "loss": 0.6915, + "step": 9194 + }, + { + "epoch": 0.4071151057635899, + "grad_norm": 1.3699824810028076, + "learning_rate": 1.4784349349130063e-05, + "loss": 0.7014, + "step": 9195 + }, + { + "epoch": 0.40715938146840375, + "grad_norm": 1.3516494035720825, + "learning_rate": 1.4782992195983248e-05, + "loss": 0.7212, + "step": 9196 + }, + { + "epoch": 0.4072036571732176, + "grad_norm": 1.4899139404296875, + "learning_rate": 1.478163492859856e-05, + "loss": 0.7469, + "step": 9197 + }, + { + "epoch": 0.4072479328780315, + "grad_norm": 1.2790284156799316, + "learning_rate": 1.4780277547008418e-05, + "loss": 0.7373, + "step": 9198 + }, + { + "epoch": 0.40729220858284537, + "grad_norm": 1.4432055950164795, + "learning_rate": 1.4778920051245236e-05, + "loss": 0.7591, + "step": 9199 + }, + { + "epoch": 0.40733648428765923, + "grad_norm": 1.966966986656189, + "learning_rate": 1.477756244134144e-05, + "loss": 0.8502, + "step": 9200 + }, + { + "epoch": 0.40738075999247314, + "grad_norm": 1.378496766090393, + "learning_rate": 1.4776204717329452e-05, + "loss": 0.7702, + "step": 9201 + }, + { + "epoch": 0.407425035697287, + "grad_norm": 1.4933606386184692, + "learning_rate": 1.4774846879241706e-05, + "loss": 0.774, + "step": 9202 + }, + { + "epoch": 0.4074693114021009, + "grad_norm": 1.407063603401184, + "learning_rate": 1.4773488927110633e-05, + "loss": 0.7729, + "step": 9203 + }, + { + "epoch": 0.40751358710691477, + "grad_norm": 1.463051676750183, + "learning_rate": 1.4772130860968661e-05, + "loss": 0.8067, + "step": 9204 + }, + { + "epoch": 0.4075578628117286, + "grad_norm": 1.6881937980651855, + "learning_rate": 1.4770772680848228e-05, + "loss": 0.7756, + "step": 9205 + }, + { + "epoch": 0.40760213851654253, + "grad_norm": 1.625974178314209, + "learning_rate": 1.4769414386781772e-05, + "loss": 0.7701, + "step": 9206 + }, + { + "epoch": 0.4076464142213564, + "grad_norm": 1.2834782600402832, + "learning_rate": 1.4768055978801743e-05, + "loss": 0.7105, + "step": 9207 + }, + { + "epoch": 0.40769068992617025, + "grad_norm": 1.40479576587677, + "learning_rate": 1.4766697456940576e-05, + "loss": 0.7413, + "step": 9208 + }, + { + "epoch": 0.40773496563098416, + "grad_norm": 1.8681453466415405, + "learning_rate": 1.476533882123072e-05, + "loss": 0.8093, + "step": 9209 + }, + { + "epoch": 0.407779241335798, + "grad_norm": 1.3206533193588257, + "learning_rate": 1.4763980071704629e-05, + "loss": 0.7882, + "step": 9210 + }, + { + "epoch": 0.40782351704061187, + "grad_norm": 1.5151149034500122, + "learning_rate": 1.476262120839475e-05, + "loss": 0.7993, + "step": 9211 + }, + { + "epoch": 0.4078677927454258, + "grad_norm": 1.3174047470092773, + "learning_rate": 1.4761262231333546e-05, + "loss": 0.6878, + "step": 9212 + }, + { + "epoch": 0.40791206845023964, + "grad_norm": 1.4923244714736938, + "learning_rate": 1.4759903140553466e-05, + "loss": 0.7188, + "step": 9213 + }, + { + "epoch": 0.4079563441550535, + "grad_norm": 1.3889150619506836, + "learning_rate": 1.4758543936086978e-05, + "loss": 0.7866, + "step": 9214 + }, + { + "epoch": 0.4080006198598674, + "grad_norm": 1.3892394304275513, + "learning_rate": 1.475718461796654e-05, + "loss": 0.7924, + "step": 9215 + }, + { + "epoch": 0.40804489556468126, + "grad_norm": 1.3123078346252441, + "learning_rate": 1.4755825186224622e-05, + "loss": 0.6629, + "step": 9216 + }, + { + "epoch": 0.4080891712694951, + "grad_norm": 1.2599067687988281, + "learning_rate": 1.4754465640893692e-05, + "loss": 0.7378, + "step": 9217 + }, + { + "epoch": 0.40813344697430903, + "grad_norm": 1.4077621698379517, + "learning_rate": 1.4753105982006219e-05, + "loss": 0.7773, + "step": 9218 + }, + { + "epoch": 0.4081777226791229, + "grad_norm": 1.2767733335494995, + "learning_rate": 1.4751746209594683e-05, + "loss": 0.7419, + "step": 9219 + }, + { + "epoch": 0.4082219983839368, + "grad_norm": 1.3364917039871216, + "learning_rate": 1.4750386323691557e-05, + "loss": 0.7237, + "step": 9220 + }, + { + "epoch": 0.40826627408875066, + "grad_norm": 1.3205146789550781, + "learning_rate": 1.474902632432932e-05, + "loss": 0.7085, + "step": 9221 + }, + { + "epoch": 0.4083105497935645, + "grad_norm": 1.5271246433258057, + "learning_rate": 1.474766621154046e-05, + "loss": 0.8037, + "step": 9222 + }, + { + "epoch": 0.4083548254983784, + "grad_norm": 1.6633539199829102, + "learning_rate": 1.4746305985357454e-05, + "loss": 0.7097, + "step": 9223 + }, + { + "epoch": 0.4083991012031923, + "grad_norm": 1.3596577644348145, + "learning_rate": 1.4744945645812796e-05, + "loss": 0.6843, + "step": 9224 + }, + { + "epoch": 0.40844337690800614, + "grad_norm": 1.1889673471450806, + "learning_rate": 1.4743585192938974e-05, + "loss": 0.7807, + "step": 9225 + }, + { + "epoch": 0.40848765261282005, + "grad_norm": 1.4373257160186768, + "learning_rate": 1.4742224626768482e-05, + "loss": 0.6897, + "step": 9226 + }, + { + "epoch": 0.4085319283176339, + "grad_norm": 1.4371908903121948, + "learning_rate": 1.4740863947333819e-05, + "loss": 0.7428, + "step": 9227 + }, + { + "epoch": 0.40857620402244776, + "grad_norm": 1.2015137672424316, + "learning_rate": 1.4739503154667476e-05, + "loss": 0.6769, + "step": 9228 + }, + { + "epoch": 0.4086204797272617, + "grad_norm": 1.4475504159927368, + "learning_rate": 1.4738142248801963e-05, + "loss": 0.7463, + "step": 9229 + }, + { + "epoch": 0.40866475543207553, + "grad_norm": 1.5239801406860352, + "learning_rate": 1.4736781229769776e-05, + "loss": 0.7722, + "step": 9230 + }, + { + "epoch": 0.4087090311368894, + "grad_norm": 1.5445612668991089, + "learning_rate": 1.473542009760343e-05, + "loss": 0.7656, + "step": 9231 + }, + { + "epoch": 0.4087533068417033, + "grad_norm": 1.7701752185821533, + "learning_rate": 1.4734058852335431e-05, + "loss": 0.7528, + "step": 9232 + }, + { + "epoch": 0.40879758254651716, + "grad_norm": 1.3645824193954468, + "learning_rate": 1.4732697493998287e-05, + "loss": 0.7598, + "step": 9233 + }, + { + "epoch": 0.408841858251331, + "grad_norm": 1.4726423025131226, + "learning_rate": 1.473133602262452e-05, + "loss": 0.7968, + "step": 9234 + }, + { + "epoch": 0.4088861339561449, + "grad_norm": 1.7276277542114258, + "learning_rate": 1.4729974438246646e-05, + "loss": 0.7365, + "step": 9235 + }, + { + "epoch": 0.4089304096609588, + "grad_norm": 1.3452750444412231, + "learning_rate": 1.472861274089718e-05, + "loss": 0.6998, + "step": 9236 + }, + { + "epoch": 0.40897468536577264, + "grad_norm": 1.2524527311325073, + "learning_rate": 1.4727250930608652e-05, + "loss": 0.6705, + "step": 9237 + }, + { + "epoch": 0.40901896107058655, + "grad_norm": 1.1729844808578491, + "learning_rate": 1.472588900741358e-05, + "loss": 0.6359, + "step": 9238 + }, + { + "epoch": 0.4090632367754004, + "grad_norm": 1.1819114685058594, + "learning_rate": 1.4724526971344504e-05, + "loss": 0.7033, + "step": 9239 + }, + { + "epoch": 0.4091075124802143, + "grad_norm": 1.4329897165298462, + "learning_rate": 1.4723164822433945e-05, + "loss": 0.7572, + "step": 9240 + }, + { + "epoch": 0.4091517881850282, + "grad_norm": 1.5975769758224487, + "learning_rate": 1.4721802560714438e-05, + "loss": 0.7688, + "step": 9241 + }, + { + "epoch": 0.40919606388984203, + "grad_norm": 1.8331325054168701, + "learning_rate": 1.4720440186218523e-05, + "loss": 0.8362, + "step": 9242 + }, + { + "epoch": 0.40924033959465594, + "grad_norm": 1.4227253198623657, + "learning_rate": 1.4719077698978737e-05, + "loss": 0.7927, + "step": 9243 + }, + { + "epoch": 0.4092846152994698, + "grad_norm": 1.6389914751052856, + "learning_rate": 1.4717715099027622e-05, + "loss": 0.7286, + "step": 9244 + }, + { + "epoch": 0.40932889100428366, + "grad_norm": 1.403790831565857, + "learning_rate": 1.4716352386397725e-05, + "loss": 0.7834, + "step": 9245 + }, + { + "epoch": 0.40937316670909757, + "grad_norm": 1.5675890445709229, + "learning_rate": 1.4714989561121591e-05, + "loss": 0.8265, + "step": 9246 + }, + { + "epoch": 0.4094174424139114, + "grad_norm": 1.3137142658233643, + "learning_rate": 1.4713626623231769e-05, + "loss": 0.7297, + "step": 9247 + }, + { + "epoch": 0.4094617181187253, + "grad_norm": 1.6034715175628662, + "learning_rate": 1.4712263572760814e-05, + "loss": 0.7096, + "step": 9248 + }, + { + "epoch": 0.4095059938235392, + "grad_norm": 1.54391348361969, + "learning_rate": 1.4710900409741282e-05, + "loss": 0.7576, + "step": 9249 + }, + { + "epoch": 0.40955026952835305, + "grad_norm": 1.3956035375595093, + "learning_rate": 1.4709537134205726e-05, + "loss": 0.7277, + "step": 9250 + }, + { + "epoch": 0.4095945452331669, + "grad_norm": 1.386070728302002, + "learning_rate": 1.4708173746186714e-05, + "loss": 0.6816, + "step": 9251 + }, + { + "epoch": 0.4096388209379808, + "grad_norm": 1.1480906009674072, + "learning_rate": 1.4706810245716803e-05, + "loss": 0.6821, + "step": 9252 + }, + { + "epoch": 0.4096830966427947, + "grad_norm": 1.4128977060317993, + "learning_rate": 1.4705446632828561e-05, + "loss": 0.659, + "step": 9253 + }, + { + "epoch": 0.40972737234760853, + "grad_norm": 1.4848486185073853, + "learning_rate": 1.4704082907554562e-05, + "loss": 0.8604, + "step": 9254 + }, + { + "epoch": 0.40977164805242244, + "grad_norm": 1.470308542251587, + "learning_rate": 1.4702719069927368e-05, + "loss": 0.7302, + "step": 9255 + }, + { + "epoch": 0.4098159237572363, + "grad_norm": 1.3295981884002686, + "learning_rate": 1.470135511997956e-05, + "loss": 0.6608, + "step": 9256 + }, + { + "epoch": 0.4098601994620502, + "grad_norm": 1.43038809299469, + "learning_rate": 1.4699991057743715e-05, + "loss": 0.7489, + "step": 9257 + }, + { + "epoch": 0.40990447516686407, + "grad_norm": 1.336987853050232, + "learning_rate": 1.4698626883252408e-05, + "loss": 0.715, + "step": 9258 + }, + { + "epoch": 0.4099487508716779, + "grad_norm": 1.3719228506088257, + "learning_rate": 1.4697262596538227e-05, + "loss": 0.7329, + "step": 9259 + }, + { + "epoch": 0.40999302657649184, + "grad_norm": 1.758709192276001, + "learning_rate": 1.469589819763375e-05, + "loss": 0.8056, + "step": 9260 + }, + { + "epoch": 0.4100373022813057, + "grad_norm": 1.4118609428405762, + "learning_rate": 1.4694533686571566e-05, + "loss": 0.6677, + "step": 9261 + }, + { + "epoch": 0.41008157798611955, + "grad_norm": 1.2450523376464844, + "learning_rate": 1.4693169063384272e-05, + "loss": 0.7041, + "step": 9262 + }, + { + "epoch": 0.41012585369093346, + "grad_norm": 1.1852284669876099, + "learning_rate": 1.4691804328104455e-05, + "loss": 0.6912, + "step": 9263 + }, + { + "epoch": 0.4101701293957473, + "grad_norm": 1.6295396089553833, + "learning_rate": 1.469043948076471e-05, + "loss": 0.8548, + "step": 9264 + }, + { + "epoch": 0.4102144051005612, + "grad_norm": 1.6880449056625366, + "learning_rate": 1.4689074521397638e-05, + "loss": 0.8385, + "step": 9265 + }, + { + "epoch": 0.4102586808053751, + "grad_norm": 1.8085243701934814, + "learning_rate": 1.4687709450035838e-05, + "loss": 0.7865, + "step": 9266 + }, + { + "epoch": 0.41030295651018894, + "grad_norm": 1.475990653038025, + "learning_rate": 1.4686344266711916e-05, + "loss": 0.7856, + "step": 9267 + }, + { + "epoch": 0.4103472322150028, + "grad_norm": 1.7689447402954102, + "learning_rate": 1.468497897145848e-05, + "loss": 0.8399, + "step": 9268 + }, + { + "epoch": 0.4103915079198167, + "grad_norm": 1.3006056547164917, + "learning_rate": 1.4683613564308132e-05, + "loss": 0.7361, + "step": 9269 + }, + { + "epoch": 0.41043578362463057, + "grad_norm": 1.0504025220870972, + "learning_rate": 1.4682248045293487e-05, + "loss": 0.6326, + "step": 9270 + }, + { + "epoch": 0.4104800593294444, + "grad_norm": 1.472428560256958, + "learning_rate": 1.4680882414447163e-05, + "loss": 0.7466, + "step": 9271 + }, + { + "epoch": 0.41052433503425834, + "grad_norm": 1.2303354740142822, + "learning_rate": 1.4679516671801772e-05, + "loss": 0.7169, + "step": 9272 + }, + { + "epoch": 0.4105686107390722, + "grad_norm": 1.216652750968933, + "learning_rate": 1.4678150817389936e-05, + "loss": 0.7171, + "step": 9273 + }, + { + "epoch": 0.4106128864438861, + "grad_norm": 2.0185344219207764, + "learning_rate": 1.4676784851244278e-05, + "loss": 0.8071, + "step": 9274 + }, + { + "epoch": 0.41065716214869996, + "grad_norm": 1.2357239723205566, + "learning_rate": 1.4675418773397418e-05, + "loss": 0.7456, + "step": 9275 + }, + { + "epoch": 0.4107014378535138, + "grad_norm": 1.7705687284469604, + "learning_rate": 1.467405258388199e-05, + "loss": 0.7686, + "step": 9276 + }, + { + "epoch": 0.41074571355832773, + "grad_norm": 1.6603997945785522, + "learning_rate": 1.4672686282730622e-05, + "loss": 0.84, + "step": 9277 + }, + { + "epoch": 0.4107899892631416, + "grad_norm": 1.970823884010315, + "learning_rate": 1.4671319869975946e-05, + "loss": 0.8303, + "step": 9278 + }, + { + "epoch": 0.41083426496795544, + "grad_norm": 1.5788861513137817, + "learning_rate": 1.46699533456506e-05, + "loss": 0.7641, + "step": 9279 + }, + { + "epoch": 0.41087854067276935, + "grad_norm": 1.5520226955413818, + "learning_rate": 1.4668586709787218e-05, + "loss": 0.7236, + "step": 9280 + }, + { + "epoch": 0.4109228163775832, + "grad_norm": 1.336156964302063, + "learning_rate": 1.4667219962418446e-05, + "loss": 0.7459, + "step": 9281 + }, + { + "epoch": 0.41096709208239707, + "grad_norm": 1.3544940948486328, + "learning_rate": 1.4665853103576923e-05, + "loss": 0.7175, + "step": 9282 + }, + { + "epoch": 0.411011367787211, + "grad_norm": 1.24762761592865, + "learning_rate": 1.4664486133295301e-05, + "loss": 0.7232, + "step": 9283 + }, + { + "epoch": 0.41105564349202484, + "grad_norm": 1.5685454607009888, + "learning_rate": 1.4663119051606225e-05, + "loss": 0.7694, + "step": 9284 + }, + { + "epoch": 0.4110999191968387, + "grad_norm": 1.3633307218551636, + "learning_rate": 1.4661751858542343e-05, + "loss": 0.7067, + "step": 9285 + }, + { + "epoch": 0.4111441949016526, + "grad_norm": 1.4671646356582642, + "learning_rate": 1.466038455413632e-05, + "loss": 0.7042, + "step": 9286 + }, + { + "epoch": 0.41118847060646646, + "grad_norm": 1.2822002172470093, + "learning_rate": 1.4659017138420804e-05, + "loss": 0.712, + "step": 9287 + }, + { + "epoch": 0.4112327463112803, + "grad_norm": 1.5210273265838623, + "learning_rate": 1.4657649611428455e-05, + "loss": 0.7569, + "step": 9288 + }, + { + "epoch": 0.41127702201609423, + "grad_norm": 1.4246249198913574, + "learning_rate": 1.4656281973191939e-05, + "loss": 0.7104, + "step": 9289 + }, + { + "epoch": 0.4113212977209081, + "grad_norm": 1.838313102722168, + "learning_rate": 1.4654914223743918e-05, + "loss": 0.8281, + "step": 9290 + }, + { + "epoch": 0.411365573425722, + "grad_norm": 1.6298917531967163, + "learning_rate": 1.4653546363117063e-05, + "loss": 0.761, + "step": 9291 + }, + { + "epoch": 0.41140984913053585, + "grad_norm": 1.2644232511520386, + "learning_rate": 1.4652178391344043e-05, + "loss": 0.7053, + "step": 9292 + }, + { + "epoch": 0.4114541248353497, + "grad_norm": 1.3922113180160522, + "learning_rate": 1.4650810308457528e-05, + "loss": 0.7499, + "step": 9293 + }, + { + "epoch": 0.4114984005401636, + "grad_norm": 1.3344957828521729, + "learning_rate": 1.4649442114490198e-05, + "loss": 0.8058, + "step": 9294 + }, + { + "epoch": 0.4115426762449775, + "grad_norm": 1.6651067733764648, + "learning_rate": 1.4648073809474723e-05, + "loss": 0.8359, + "step": 9295 + }, + { + "epoch": 0.41158695194979134, + "grad_norm": 1.280013084411621, + "learning_rate": 1.4646705393443796e-05, + "loss": 0.7036, + "step": 9296 + }, + { + "epoch": 0.41163122765460525, + "grad_norm": 1.1940277814865112, + "learning_rate": 1.4645336866430094e-05, + "loss": 0.6736, + "step": 9297 + }, + { + "epoch": 0.4116755033594191, + "grad_norm": 1.2782539129257202, + "learning_rate": 1.4643968228466299e-05, + "loss": 0.7592, + "step": 9298 + }, + { + "epoch": 0.41171977906423296, + "grad_norm": 1.5126423835754395, + "learning_rate": 1.4642599479585106e-05, + "loss": 0.8216, + "step": 9299 + }, + { + "epoch": 0.41176405476904687, + "grad_norm": 1.3593430519104004, + "learning_rate": 1.4641230619819205e-05, + "loss": 0.7546, + "step": 9300 + }, + { + "epoch": 0.41180833047386073, + "grad_norm": 1.2499706745147705, + "learning_rate": 1.4639861649201291e-05, + "loss": 0.7462, + "step": 9301 + }, + { + "epoch": 0.4118526061786746, + "grad_norm": 1.4637782573699951, + "learning_rate": 1.4638492567764057e-05, + "loss": 0.8292, + "step": 9302 + }, + { + "epoch": 0.4118968818834885, + "grad_norm": 1.2965548038482666, + "learning_rate": 1.4637123375540206e-05, + "loss": 0.7264, + "step": 9303 + }, + { + "epoch": 0.41194115758830235, + "grad_norm": 1.5373820066452026, + "learning_rate": 1.4635754072562438e-05, + "loss": 0.7336, + "step": 9304 + }, + { + "epoch": 0.4119854332931162, + "grad_norm": 1.3242628574371338, + "learning_rate": 1.4634384658863458e-05, + "loss": 0.7603, + "step": 9305 + }, + { + "epoch": 0.4120297089979301, + "grad_norm": 1.211000680923462, + "learning_rate": 1.4633015134475977e-05, + "loss": 0.6847, + "step": 9306 + }, + { + "epoch": 0.412073984702744, + "grad_norm": 1.1768969297409058, + "learning_rate": 1.46316454994327e-05, + "loss": 0.7088, + "step": 9307 + }, + { + "epoch": 0.4121182604075579, + "grad_norm": 1.4720360040664673, + "learning_rate": 1.463027575376634e-05, + "loss": 0.7219, + "step": 9308 + }, + { + "epoch": 0.41216253611237175, + "grad_norm": 1.2243818044662476, + "learning_rate": 1.4628905897509616e-05, + "loss": 0.7198, + "step": 9309 + }, + { + "epoch": 0.4122068118171856, + "grad_norm": 1.5975701808929443, + "learning_rate": 1.4627535930695241e-05, + "loss": 0.7842, + "step": 9310 + }, + { + "epoch": 0.4122510875219995, + "grad_norm": 1.4891507625579834, + "learning_rate": 1.4626165853355943e-05, + "loss": 0.7611, + "step": 9311 + }, + { + "epoch": 0.41229536322681337, + "grad_norm": 1.2543717622756958, + "learning_rate": 1.4624795665524436e-05, + "loss": 0.7459, + "step": 9312 + }, + { + "epoch": 0.41233963893162723, + "grad_norm": 1.3817397356033325, + "learning_rate": 1.462342536723345e-05, + "loss": 0.7196, + "step": 9313 + }, + { + "epoch": 0.41238391463644114, + "grad_norm": 1.5082262754440308, + "learning_rate": 1.4622054958515716e-05, + "loss": 0.7277, + "step": 9314 + }, + { + "epoch": 0.412428190341255, + "grad_norm": 1.3344491720199585, + "learning_rate": 1.4620684439403962e-05, + "loss": 0.7401, + "step": 9315 + }, + { + "epoch": 0.41247246604606885, + "grad_norm": 1.4395898580551147, + "learning_rate": 1.4619313809930923e-05, + "loss": 0.8191, + "step": 9316 + }, + { + "epoch": 0.41251674175088276, + "grad_norm": 1.4598947763442993, + "learning_rate": 1.4617943070129334e-05, + "loss": 0.7624, + "step": 9317 + }, + { + "epoch": 0.4125610174556966, + "grad_norm": 1.4126399755477905, + "learning_rate": 1.4616572220031934e-05, + "loss": 0.767, + "step": 9318 + }, + { + "epoch": 0.4126052931605105, + "grad_norm": 1.5701894760131836, + "learning_rate": 1.4615201259671468e-05, + "loss": 0.7515, + "step": 9319 + }, + { + "epoch": 0.4126495688653244, + "grad_norm": 1.4185236692428589, + "learning_rate": 1.4613830189080676e-05, + "loss": 0.7807, + "step": 9320 + }, + { + "epoch": 0.41269384457013825, + "grad_norm": 1.3237857818603516, + "learning_rate": 1.4612459008292307e-05, + "loss": 0.7837, + "step": 9321 + }, + { + "epoch": 0.4127381202749521, + "grad_norm": 1.1873035430908203, + "learning_rate": 1.461108771733911e-05, + "loss": 0.6578, + "step": 9322 + }, + { + "epoch": 0.412782395979766, + "grad_norm": 1.134878396987915, + "learning_rate": 1.4609716316253834e-05, + "loss": 0.6837, + "step": 9323 + }, + { + "epoch": 0.41282667168457987, + "grad_norm": 1.7331178188323975, + "learning_rate": 1.4608344805069243e-05, + "loss": 0.8025, + "step": 9324 + }, + { + "epoch": 0.4128709473893937, + "grad_norm": 1.8611873388290405, + "learning_rate": 1.4606973183818088e-05, + "loss": 0.7786, + "step": 9325 + }, + { + "epoch": 0.41291522309420764, + "grad_norm": 1.3608399629592896, + "learning_rate": 1.4605601452533126e-05, + "loss": 0.7635, + "step": 9326 + }, + { + "epoch": 0.4129594987990215, + "grad_norm": 1.3258837461471558, + "learning_rate": 1.4604229611247124e-05, + "loss": 0.7052, + "step": 9327 + }, + { + "epoch": 0.4130037745038354, + "grad_norm": 1.34709632396698, + "learning_rate": 1.4602857659992849e-05, + "loss": 0.7635, + "step": 9328 + }, + { + "epoch": 0.41304805020864926, + "grad_norm": 1.3900715112686157, + "learning_rate": 1.4601485598803064e-05, + "loss": 0.7484, + "step": 9329 + }, + { + "epoch": 0.4130923259134631, + "grad_norm": 1.6058106422424316, + "learning_rate": 1.4600113427710545e-05, + "loss": 0.7863, + "step": 9330 + }, + { + "epoch": 0.41313660161827703, + "grad_norm": 1.4760783910751343, + "learning_rate": 1.4598741146748064e-05, + "loss": 0.7357, + "step": 9331 + }, + { + "epoch": 0.4131808773230909, + "grad_norm": 1.5959656238555908, + "learning_rate": 1.4597368755948389e-05, + "loss": 0.8154, + "step": 9332 + }, + { + "epoch": 0.41322515302790475, + "grad_norm": 1.4751445055007935, + "learning_rate": 1.4595996255344307e-05, + "loss": 0.7592, + "step": 9333 + }, + { + "epoch": 0.41326942873271866, + "grad_norm": 1.45149827003479, + "learning_rate": 1.45946236449686e-05, + "loss": 0.7491, + "step": 9334 + }, + { + "epoch": 0.4133137044375325, + "grad_norm": 1.5176653861999512, + "learning_rate": 1.4593250924854043e-05, + "loss": 0.8248, + "step": 9335 + }, + { + "epoch": 0.41335798014234637, + "grad_norm": 1.3623247146606445, + "learning_rate": 1.4591878095033432e-05, + "loss": 0.7259, + "step": 9336 + }, + { + "epoch": 0.4134022558471603, + "grad_norm": 1.348128318786621, + "learning_rate": 1.4590505155539548e-05, + "loss": 0.7684, + "step": 9337 + }, + { + "epoch": 0.41344653155197414, + "grad_norm": 1.6254829168319702, + "learning_rate": 1.4589132106405187e-05, + "loss": 0.7919, + "step": 9338 + }, + { + "epoch": 0.413490807256788, + "grad_norm": 1.410631775856018, + "learning_rate": 1.4587758947663146e-05, + "loss": 0.7535, + "step": 9339 + }, + { + "epoch": 0.4135350829616019, + "grad_norm": 1.2436485290527344, + "learning_rate": 1.4586385679346213e-05, + "loss": 0.8207, + "step": 9340 + }, + { + "epoch": 0.41357935866641576, + "grad_norm": 1.4038572311401367, + "learning_rate": 1.4585012301487196e-05, + "loss": 0.7849, + "step": 9341 + }, + { + "epoch": 0.4136236343712296, + "grad_norm": 1.7022454738616943, + "learning_rate": 1.458363881411889e-05, + "loss": 0.8774, + "step": 9342 + }, + { + "epoch": 0.41366791007604353, + "grad_norm": 1.8734742403030396, + "learning_rate": 1.4582265217274105e-05, + "loss": 0.6787, + "step": 9343 + }, + { + "epoch": 0.4137121857808574, + "grad_norm": 1.3994382619857788, + "learning_rate": 1.458089151098565e-05, + "loss": 0.7293, + "step": 9344 + }, + { + "epoch": 0.4137564614856713, + "grad_norm": 1.3728110790252686, + "learning_rate": 1.4579517695286325e-05, + "loss": 0.7857, + "step": 9345 + }, + { + "epoch": 0.41380073719048516, + "grad_norm": 1.5053372383117676, + "learning_rate": 1.457814377020895e-05, + "loss": 0.7487, + "step": 9346 + }, + { + "epoch": 0.413845012895299, + "grad_norm": 1.5125635862350464, + "learning_rate": 1.457676973578634e-05, + "loss": 0.7649, + "step": 9347 + }, + { + "epoch": 0.4138892886001129, + "grad_norm": 1.4106428623199463, + "learning_rate": 1.4575395592051311e-05, + "loss": 0.723, + "step": 9348 + }, + { + "epoch": 0.4139335643049268, + "grad_norm": 1.206018328666687, + "learning_rate": 1.4574021339036686e-05, + "loss": 0.7328, + "step": 9349 + }, + { + "epoch": 0.41397784000974064, + "grad_norm": 1.2420048713684082, + "learning_rate": 1.4572646976775282e-05, + "loss": 0.6815, + "step": 9350 + }, + { + "epoch": 0.41402211571455455, + "grad_norm": 1.272362232208252, + "learning_rate": 1.457127250529993e-05, + "loss": 0.6816, + "step": 9351 + }, + { + "epoch": 0.4140663914193684, + "grad_norm": 1.3667128086090088, + "learning_rate": 1.4569897924643455e-05, + "loss": 0.6699, + "step": 9352 + }, + { + "epoch": 0.41411066712418226, + "grad_norm": 1.4514902830123901, + "learning_rate": 1.456852323483869e-05, + "loss": 0.7452, + "step": 9353 + }, + { + "epoch": 0.4141549428289962, + "grad_norm": 1.4549741744995117, + "learning_rate": 1.4567148435918469e-05, + "loss": 0.8055, + "step": 9354 + }, + { + "epoch": 0.41419921853381003, + "grad_norm": 1.734034776687622, + "learning_rate": 1.4565773527915624e-05, + "loss": 0.7922, + "step": 9355 + }, + { + "epoch": 0.4142434942386239, + "grad_norm": 1.2682008743286133, + "learning_rate": 1.4564398510862996e-05, + "loss": 0.697, + "step": 9356 + }, + { + "epoch": 0.4142877699434378, + "grad_norm": 1.4519774913787842, + "learning_rate": 1.4563023384793425e-05, + "loss": 0.7208, + "step": 9357 + }, + { + "epoch": 0.41433204564825166, + "grad_norm": 1.3388019800186157, + "learning_rate": 1.4561648149739758e-05, + "loss": 0.7805, + "step": 9358 + }, + { + "epoch": 0.4143763213530655, + "grad_norm": 1.5286844968795776, + "learning_rate": 1.4560272805734838e-05, + "loss": 0.7279, + "step": 9359 + }, + { + "epoch": 0.4144205970578794, + "grad_norm": 1.4150117635726929, + "learning_rate": 1.4558897352811517e-05, + "loss": 0.776, + "step": 9360 + }, + { + "epoch": 0.4144648727626933, + "grad_norm": 1.418273687362671, + "learning_rate": 1.4557521791002644e-05, + "loss": 0.7394, + "step": 9361 + }, + { + "epoch": 0.4145091484675072, + "grad_norm": 1.324633240699768, + "learning_rate": 1.4556146120341073e-05, + "loss": 0.6689, + "step": 9362 + }, + { + "epoch": 0.41455342417232105, + "grad_norm": 1.3782097101211548, + "learning_rate": 1.4554770340859661e-05, + "loss": 0.7799, + "step": 9363 + }, + { + "epoch": 0.4145976998771349, + "grad_norm": 1.582763910293579, + "learning_rate": 1.455339445259127e-05, + "loss": 0.7597, + "step": 9364 + }, + { + "epoch": 0.4146419755819488, + "grad_norm": 1.2427445650100708, + "learning_rate": 1.4552018455568758e-05, + "loss": 0.6315, + "step": 9365 + }, + { + "epoch": 0.4146862512867627, + "grad_norm": 1.4555063247680664, + "learning_rate": 1.4550642349824993e-05, + "loss": 0.7233, + "step": 9366 + }, + { + "epoch": 0.41473052699157653, + "grad_norm": 1.494210124015808, + "learning_rate": 1.4549266135392841e-05, + "loss": 0.6941, + "step": 9367 + }, + { + "epoch": 0.41477480269639044, + "grad_norm": 1.935046911239624, + "learning_rate": 1.4547889812305169e-05, + "loss": 0.868, + "step": 9368 + }, + { + "epoch": 0.4148190784012043, + "grad_norm": 1.2085731029510498, + "learning_rate": 1.4546513380594855e-05, + "loss": 0.6654, + "step": 9369 + }, + { + "epoch": 0.41486335410601816, + "grad_norm": 1.5353659391403198, + "learning_rate": 1.4545136840294769e-05, + "loss": 0.7591, + "step": 9370 + }, + { + "epoch": 0.41490762981083207, + "grad_norm": 1.4631444215774536, + "learning_rate": 1.454376019143779e-05, + "loss": 0.7791, + "step": 9371 + }, + { + "epoch": 0.4149519055156459, + "grad_norm": 1.719908356666565, + "learning_rate": 1.4542383434056801e-05, + "loss": 0.8108, + "step": 9372 + }, + { + "epoch": 0.4149961812204598, + "grad_norm": 1.5108165740966797, + "learning_rate": 1.454100656818468e-05, + "loss": 0.7593, + "step": 9373 + }, + { + "epoch": 0.4150404569252737, + "grad_norm": 1.3758198022842407, + "learning_rate": 1.4539629593854316e-05, + "loss": 0.7228, + "step": 9374 + }, + { + "epoch": 0.41508473263008755, + "grad_norm": 1.274202585220337, + "learning_rate": 1.4538252511098594e-05, + "loss": 0.7375, + "step": 9375 + }, + { + "epoch": 0.4151290083349014, + "grad_norm": 1.5479289293289185, + "learning_rate": 1.4536875319950405e-05, + "loss": 0.8156, + "step": 9376 + }, + { + "epoch": 0.4151732840397153, + "grad_norm": 1.437859296798706, + "learning_rate": 1.4535498020442647e-05, + "loss": 0.7576, + "step": 9377 + }, + { + "epoch": 0.4152175597445292, + "grad_norm": 1.5752174854278564, + "learning_rate": 1.4534120612608207e-05, + "loss": 0.8311, + "step": 9378 + }, + { + "epoch": 0.4152618354493431, + "grad_norm": 1.310562014579773, + "learning_rate": 1.4532743096479994e-05, + "loss": 0.7015, + "step": 9379 + }, + { + "epoch": 0.41530611115415694, + "grad_norm": 1.6997791528701782, + "learning_rate": 1.4531365472090895e-05, + "loss": 0.8325, + "step": 9380 + }, + { + "epoch": 0.4153503868589708, + "grad_norm": 1.4223767518997192, + "learning_rate": 1.4529987739473828e-05, + "loss": 0.8388, + "step": 9381 + }, + { + "epoch": 0.4153946625637847, + "grad_norm": 1.1563066244125366, + "learning_rate": 1.452860989866169e-05, + "loss": 0.6536, + "step": 9382 + }, + { + "epoch": 0.41543893826859857, + "grad_norm": 1.6235555410385132, + "learning_rate": 1.4527231949687391e-05, + "loss": 0.7589, + "step": 9383 + }, + { + "epoch": 0.4154832139734124, + "grad_norm": 1.5839245319366455, + "learning_rate": 1.4525853892583844e-05, + "loss": 0.7656, + "step": 9384 + }, + { + "epoch": 0.41552748967822634, + "grad_norm": 1.4004229307174683, + "learning_rate": 1.452447572738396e-05, + "loss": 0.7071, + "step": 9385 + }, + { + "epoch": 0.4155717653830402, + "grad_norm": 2.0784058570861816, + "learning_rate": 1.452309745412066e-05, + "loss": 0.7931, + "step": 9386 + }, + { + "epoch": 0.41561604108785405, + "grad_norm": 1.2720427513122559, + "learning_rate": 1.4521719072826858e-05, + "loss": 0.7256, + "step": 9387 + }, + { + "epoch": 0.41566031679266796, + "grad_norm": 1.2838422060012817, + "learning_rate": 1.4520340583535482e-05, + "loss": 0.6688, + "step": 9388 + }, + { + "epoch": 0.4157045924974818, + "grad_norm": 1.3992773294448853, + "learning_rate": 1.4518961986279446e-05, + "loss": 0.7164, + "step": 9389 + }, + { + "epoch": 0.4157488682022957, + "grad_norm": 1.6588019132614136, + "learning_rate": 1.4517583281091685e-05, + "loss": 0.7398, + "step": 9390 + }, + { + "epoch": 0.4157931439071096, + "grad_norm": 1.415342926979065, + "learning_rate": 1.4516204468005127e-05, + "loss": 0.7501, + "step": 9391 + }, + { + "epoch": 0.41583741961192344, + "grad_norm": 1.9235761165618896, + "learning_rate": 1.4514825547052697e-05, + "loss": 0.8069, + "step": 9392 + }, + { + "epoch": 0.4158816953167373, + "grad_norm": 1.766950011253357, + "learning_rate": 1.4513446518267342e-05, + "loss": 0.7722, + "step": 9393 + }, + { + "epoch": 0.4159259710215512, + "grad_norm": 1.2788960933685303, + "learning_rate": 1.4512067381681987e-05, + "loss": 0.6593, + "step": 9394 + }, + { + "epoch": 0.41597024672636507, + "grad_norm": 1.126993179321289, + "learning_rate": 1.4510688137329577e-05, + "loss": 0.623, + "step": 9395 + }, + { + "epoch": 0.4160145224311789, + "grad_norm": 1.4990689754486084, + "learning_rate": 1.450930878524306e-05, + "loss": 0.7327, + "step": 9396 + }, + { + "epoch": 0.41605879813599284, + "grad_norm": 1.5137358903884888, + "learning_rate": 1.4507929325455366e-05, + "loss": 0.7128, + "step": 9397 + }, + { + "epoch": 0.4161030738408067, + "grad_norm": 1.1578360795974731, + "learning_rate": 1.4506549757999456e-05, + "loss": 0.7087, + "step": 9398 + }, + { + "epoch": 0.4161473495456206, + "grad_norm": 1.4257982969284058, + "learning_rate": 1.4505170082908269e-05, + "loss": 0.7347, + "step": 9399 + }, + { + "epoch": 0.41619162525043446, + "grad_norm": 1.3313300609588623, + "learning_rate": 1.4503790300214766e-05, + "loss": 0.6963, + "step": 9400 + }, + { + "epoch": 0.4162359009552483, + "grad_norm": 1.4525152444839478, + "learning_rate": 1.4502410409951901e-05, + "loss": 0.8116, + "step": 9401 + }, + { + "epoch": 0.41628017666006223, + "grad_norm": 1.5300644636154175, + "learning_rate": 1.4501030412152626e-05, + "loss": 0.7841, + "step": 9402 + }, + { + "epoch": 0.4163244523648761, + "grad_norm": 1.3727450370788574, + "learning_rate": 1.4499650306849904e-05, + "loss": 0.7243, + "step": 9403 + }, + { + "epoch": 0.41636872806968994, + "grad_norm": 1.2241950035095215, + "learning_rate": 1.4498270094076698e-05, + "loss": 0.7289, + "step": 9404 + }, + { + "epoch": 0.41641300377450385, + "grad_norm": 1.171284794807434, + "learning_rate": 1.4496889773865977e-05, + "loss": 0.6953, + "step": 9405 + }, + { + "epoch": 0.4164572794793177, + "grad_norm": 1.4431147575378418, + "learning_rate": 1.4495509346250704e-05, + "loss": 0.7969, + "step": 9406 + }, + { + "epoch": 0.41650155518413157, + "grad_norm": 1.9299566745758057, + "learning_rate": 1.449412881126385e-05, + "loss": 0.8596, + "step": 9407 + }, + { + "epoch": 0.4165458308889455, + "grad_norm": 1.6925077438354492, + "learning_rate": 1.449274816893839e-05, + "loss": 0.8456, + "step": 9408 + }, + { + "epoch": 0.41659010659375934, + "grad_norm": 1.6048030853271484, + "learning_rate": 1.4491367419307294e-05, + "loss": 0.7701, + "step": 9409 + }, + { + "epoch": 0.4166343822985732, + "grad_norm": 1.407173752784729, + "learning_rate": 1.4489986562403548e-05, + "loss": 0.7362, + "step": 9410 + }, + { + "epoch": 0.4166786580033871, + "grad_norm": 1.6985541582107544, + "learning_rate": 1.4488605598260129e-05, + "loss": 0.8388, + "step": 9411 + }, + { + "epoch": 0.41672293370820096, + "grad_norm": 1.448301076889038, + "learning_rate": 1.4487224526910018e-05, + "loss": 0.7139, + "step": 9412 + }, + { + "epoch": 0.4167672094130148, + "grad_norm": 1.410222053527832, + "learning_rate": 1.4485843348386203e-05, + "loss": 0.7464, + "step": 9413 + }, + { + "epoch": 0.41681148511782873, + "grad_norm": 1.3952780961990356, + "learning_rate": 1.4484462062721674e-05, + "loss": 0.7277, + "step": 9414 + }, + { + "epoch": 0.4168557608226426, + "grad_norm": 1.2563300132751465, + "learning_rate": 1.4483080669949416e-05, + "loss": 0.7472, + "step": 9415 + }, + { + "epoch": 0.4169000365274565, + "grad_norm": 1.4795944690704346, + "learning_rate": 1.4481699170102432e-05, + "loss": 0.8029, + "step": 9416 + }, + { + "epoch": 0.41694431223227035, + "grad_norm": 1.2250903844833374, + "learning_rate": 1.4480317563213708e-05, + "loss": 0.6822, + "step": 9417 + }, + { + "epoch": 0.4169885879370842, + "grad_norm": 1.3287010192871094, + "learning_rate": 1.4478935849316249e-05, + "loss": 0.7465, + "step": 9418 + }, + { + "epoch": 0.4170328636418981, + "grad_norm": 1.1326764822006226, + "learning_rate": 1.4477554028443056e-05, + "loss": 0.6843, + "step": 9419 + }, + { + "epoch": 0.417077139346712, + "grad_norm": 1.3332093954086304, + "learning_rate": 1.4476172100627127e-05, + "loss": 0.7021, + "step": 9420 + }, + { + "epoch": 0.41712141505152583, + "grad_norm": 1.6987513303756714, + "learning_rate": 1.4474790065901474e-05, + "loss": 0.7569, + "step": 9421 + }, + { + "epoch": 0.41716569075633975, + "grad_norm": 1.2480409145355225, + "learning_rate": 1.4473407924299103e-05, + "loss": 0.6943, + "step": 9422 + }, + { + "epoch": 0.4172099664611536, + "grad_norm": 2.7702200412750244, + "learning_rate": 1.4472025675853028e-05, + "loss": 0.7783, + "step": 9423 + }, + { + "epoch": 0.41725424216596746, + "grad_norm": 1.3964287042617798, + "learning_rate": 1.4470643320596257e-05, + "loss": 0.7458, + "step": 9424 + }, + { + "epoch": 0.41729851787078137, + "grad_norm": 1.6661131381988525, + "learning_rate": 1.4469260858561814e-05, + "loss": 0.8272, + "step": 9425 + }, + { + "epoch": 0.41734279357559523, + "grad_norm": 1.1214834451675415, + "learning_rate": 1.4467878289782713e-05, + "loss": 0.6823, + "step": 9426 + }, + { + "epoch": 0.4173870692804091, + "grad_norm": 2.1821787357330322, + "learning_rate": 1.4466495614291977e-05, + "loss": 0.7848, + "step": 9427 + }, + { + "epoch": 0.417431344985223, + "grad_norm": 1.7198374271392822, + "learning_rate": 1.4465112832122632e-05, + "loss": 0.7483, + "step": 9428 + }, + { + "epoch": 0.41747562069003685, + "grad_norm": 1.2518868446350098, + "learning_rate": 1.44637299433077e-05, + "loss": 0.738, + "step": 9429 + }, + { + "epoch": 0.4175198963948507, + "grad_norm": 1.2312915325164795, + "learning_rate": 1.4462346947880212e-05, + "loss": 0.6833, + "step": 9430 + }, + { + "epoch": 0.4175641720996646, + "grad_norm": 1.2878366708755493, + "learning_rate": 1.4460963845873203e-05, + "loss": 0.7513, + "step": 9431 + }, + { + "epoch": 0.4176084478044785, + "grad_norm": 1.5731016397476196, + "learning_rate": 1.4459580637319704e-05, + "loss": 0.8037, + "step": 9432 + }, + { + "epoch": 0.4176527235092924, + "grad_norm": 1.328303575515747, + "learning_rate": 1.4458197322252753e-05, + "loss": 0.6882, + "step": 9433 + }, + { + "epoch": 0.41769699921410625, + "grad_norm": 1.3198541402816772, + "learning_rate": 1.445681390070539e-05, + "loss": 0.6513, + "step": 9434 + }, + { + "epoch": 0.4177412749189201, + "grad_norm": 1.1172353029251099, + "learning_rate": 1.4455430372710652e-05, + "loss": 0.6382, + "step": 9435 + }, + { + "epoch": 0.417785550623734, + "grad_norm": 1.4735010862350464, + "learning_rate": 1.4454046738301593e-05, + "loss": 0.7602, + "step": 9436 + }, + { + "epoch": 0.41782982632854787, + "grad_norm": 1.175261378288269, + "learning_rate": 1.4452662997511248e-05, + "loss": 0.6683, + "step": 9437 + }, + { + "epoch": 0.4178741020333617, + "grad_norm": 1.359058141708374, + "learning_rate": 1.445127915037268e-05, + "loss": 0.6701, + "step": 9438 + }, + { + "epoch": 0.41791837773817564, + "grad_norm": 1.3640750646591187, + "learning_rate": 1.4449895196918928e-05, + "loss": 0.734, + "step": 9439 + }, + { + "epoch": 0.4179626534429895, + "grad_norm": 1.3307708501815796, + "learning_rate": 1.4448511137183053e-05, + "loss": 0.7314, + "step": 9440 + }, + { + "epoch": 0.41800692914780335, + "grad_norm": 1.3688716888427734, + "learning_rate": 1.4447126971198113e-05, + "loss": 0.7685, + "step": 9441 + }, + { + "epoch": 0.41805120485261726, + "grad_norm": 1.4248062372207642, + "learning_rate": 1.4445742698997168e-05, + "loss": 0.6656, + "step": 9442 + }, + { + "epoch": 0.4180954805574311, + "grad_norm": 1.5150752067565918, + "learning_rate": 1.4444358320613274e-05, + "loss": 0.6906, + "step": 9443 + }, + { + "epoch": 0.418139756262245, + "grad_norm": 1.372743010520935, + "learning_rate": 1.4442973836079503e-05, + "loss": 0.7357, + "step": 9444 + }, + { + "epoch": 0.4181840319670589, + "grad_norm": 1.4895501136779785, + "learning_rate": 1.4441589245428917e-05, + "loss": 0.752, + "step": 9445 + }, + { + "epoch": 0.41822830767187275, + "grad_norm": 1.4457226991653442, + "learning_rate": 1.4440204548694593e-05, + "loss": 0.8231, + "step": 9446 + }, + { + "epoch": 0.4182725833766866, + "grad_norm": 1.4779802560806274, + "learning_rate": 1.4438819745909599e-05, + "loss": 0.7443, + "step": 9447 + }, + { + "epoch": 0.4183168590815005, + "grad_norm": 1.8113640546798706, + "learning_rate": 1.4437434837107005e-05, + "loss": 0.8804, + "step": 9448 + }, + { + "epoch": 0.41836113478631437, + "grad_norm": 1.2608965635299683, + "learning_rate": 1.4436049822319893e-05, + "loss": 0.7167, + "step": 9449 + }, + { + "epoch": 0.4184054104911283, + "grad_norm": 1.4676841497421265, + "learning_rate": 1.4434664701581345e-05, + "loss": 0.7404, + "step": 9450 + }, + { + "epoch": 0.41844968619594214, + "grad_norm": 1.2077345848083496, + "learning_rate": 1.4433279474924442e-05, + "loss": 0.7086, + "step": 9451 + }, + { + "epoch": 0.418493961900756, + "grad_norm": 1.3140555620193481, + "learning_rate": 1.4431894142382268e-05, + "loss": 0.7012, + "step": 9452 + }, + { + "epoch": 0.4185382376055699, + "grad_norm": 1.3711696863174438, + "learning_rate": 1.4430508703987914e-05, + "loss": 0.7353, + "step": 9453 + }, + { + "epoch": 0.41858251331038376, + "grad_norm": 1.2719058990478516, + "learning_rate": 1.4429123159774461e-05, + "loss": 0.6315, + "step": 9454 + }, + { + "epoch": 0.4186267890151976, + "grad_norm": 1.5343574285507202, + "learning_rate": 1.4427737509775008e-05, + "loss": 0.7776, + "step": 9455 + }, + { + "epoch": 0.41867106472001153, + "grad_norm": 1.5636714696884155, + "learning_rate": 1.4426351754022654e-05, + "loss": 0.8017, + "step": 9456 + }, + { + "epoch": 0.4187153404248254, + "grad_norm": 1.3389345407485962, + "learning_rate": 1.4424965892550491e-05, + "loss": 0.6797, + "step": 9457 + }, + { + "epoch": 0.41875961612963924, + "grad_norm": 1.7389001846313477, + "learning_rate": 1.4423579925391622e-05, + "loss": 0.7733, + "step": 9458 + }, + { + "epoch": 0.41880389183445316, + "grad_norm": 1.3160600662231445, + "learning_rate": 1.4422193852579144e-05, + "loss": 0.6926, + "step": 9459 + }, + { + "epoch": 0.418848167539267, + "grad_norm": 1.6067414283752441, + "learning_rate": 1.4420807674146169e-05, + "loss": 0.8373, + "step": 9460 + }, + { + "epoch": 0.41889244324408087, + "grad_norm": 1.37533700466156, + "learning_rate": 1.4419421390125803e-05, + "loss": 0.6885, + "step": 9461 + }, + { + "epoch": 0.4189367189488948, + "grad_norm": 1.2444734573364258, + "learning_rate": 1.4418035000551156e-05, + "loss": 0.7205, + "step": 9462 + }, + { + "epoch": 0.41898099465370864, + "grad_norm": 1.3060002326965332, + "learning_rate": 1.4416648505455339e-05, + "loss": 0.6692, + "step": 9463 + }, + { + "epoch": 0.4190252703585225, + "grad_norm": 1.9354219436645508, + "learning_rate": 1.4415261904871466e-05, + "loss": 0.821, + "step": 9464 + }, + { + "epoch": 0.4190695460633364, + "grad_norm": 1.6844391822814941, + "learning_rate": 1.441387519883266e-05, + "loss": 0.7964, + "step": 9465 + }, + { + "epoch": 0.41911382176815026, + "grad_norm": 1.4405170679092407, + "learning_rate": 1.4412488387372043e-05, + "loss": 0.7082, + "step": 9466 + }, + { + "epoch": 0.4191580974729642, + "grad_norm": 1.268865704536438, + "learning_rate": 1.4411101470522727e-05, + "loss": 0.6978, + "step": 9467 + }, + { + "epoch": 0.41920237317777803, + "grad_norm": 1.2896230220794678, + "learning_rate": 1.440971444831785e-05, + "loss": 0.7325, + "step": 9468 + }, + { + "epoch": 0.4192466488825919, + "grad_norm": 1.6190671920776367, + "learning_rate": 1.440832732079053e-05, + "loss": 0.7527, + "step": 9469 + }, + { + "epoch": 0.4192909245874058, + "grad_norm": 1.2459495067596436, + "learning_rate": 1.4406940087973901e-05, + "loss": 0.6752, + "step": 9470 + }, + { + "epoch": 0.41933520029221966, + "grad_norm": 1.4851689338684082, + "learning_rate": 1.4405552749901098e-05, + "loss": 0.7887, + "step": 9471 + }, + { + "epoch": 0.4193794759970335, + "grad_norm": 1.325053095817566, + "learning_rate": 1.4404165306605254e-05, + "loss": 0.6652, + "step": 9472 + }, + { + "epoch": 0.4194237517018474, + "grad_norm": 1.2617664337158203, + "learning_rate": 1.4402777758119508e-05, + "loss": 0.7667, + "step": 9473 + }, + { + "epoch": 0.4194680274066613, + "grad_norm": 1.4227776527404785, + "learning_rate": 1.4401390104476998e-05, + "loss": 0.7723, + "step": 9474 + }, + { + "epoch": 0.41951230311147514, + "grad_norm": 1.712301254272461, + "learning_rate": 1.4400002345710871e-05, + "loss": 0.8282, + "step": 9475 + }, + { + "epoch": 0.41955657881628905, + "grad_norm": 1.5240604877471924, + "learning_rate": 1.4398614481854272e-05, + "loss": 0.7946, + "step": 9476 + }, + { + "epoch": 0.4196008545211029, + "grad_norm": 1.3079230785369873, + "learning_rate": 1.4397226512940345e-05, + "loss": 0.7011, + "step": 9477 + }, + { + "epoch": 0.41964513022591676, + "grad_norm": 1.3083269596099854, + "learning_rate": 1.4395838439002243e-05, + "loss": 0.6561, + "step": 9478 + }, + { + "epoch": 0.4196894059307307, + "grad_norm": 1.4142543077468872, + "learning_rate": 1.439445026007312e-05, + "loss": 0.805, + "step": 9479 + }, + { + "epoch": 0.41973368163554453, + "grad_norm": 1.5422464609146118, + "learning_rate": 1.4393061976186132e-05, + "loss": 0.801, + "step": 9480 + }, + { + "epoch": 0.4197779573403584, + "grad_norm": 1.6364165544509888, + "learning_rate": 1.4391673587374436e-05, + "loss": 0.8215, + "step": 9481 + }, + { + "epoch": 0.4198222330451723, + "grad_norm": 1.2916823625564575, + "learning_rate": 1.4390285093671187e-05, + "loss": 0.7208, + "step": 9482 + }, + { + "epoch": 0.41986650874998616, + "grad_norm": 1.451481819152832, + "learning_rate": 1.438889649510956e-05, + "loss": 0.7704, + "step": 9483 + }, + { + "epoch": 0.4199107844548, + "grad_norm": 1.538954496383667, + "learning_rate": 1.4387507791722709e-05, + "loss": 0.7393, + "step": 9484 + }, + { + "epoch": 0.4199550601596139, + "grad_norm": 1.2920634746551514, + "learning_rate": 1.4386118983543808e-05, + "loss": 0.7848, + "step": 9485 + }, + { + "epoch": 0.4199993358644278, + "grad_norm": 2.2508273124694824, + "learning_rate": 1.4384730070606027e-05, + "loss": 0.6646, + "step": 9486 + }, + { + "epoch": 0.4200436115692417, + "grad_norm": 1.4250682592391968, + "learning_rate": 1.438334105294254e-05, + "loss": 0.7212, + "step": 9487 + }, + { + "epoch": 0.42008788727405555, + "grad_norm": 1.292410969734192, + "learning_rate": 1.438195193058652e-05, + "loss": 0.6733, + "step": 9488 + }, + { + "epoch": 0.4201321629788694, + "grad_norm": 1.626665711402893, + "learning_rate": 1.4380562703571147e-05, + "loss": 0.7988, + "step": 9489 + }, + { + "epoch": 0.4201764386836833, + "grad_norm": 2.1712779998779297, + "learning_rate": 1.4379173371929596e-05, + "loss": 0.8633, + "step": 9490 + }, + { + "epoch": 0.4202207143884972, + "grad_norm": 1.4837287664413452, + "learning_rate": 1.4377783935695057e-05, + "loss": 0.7558, + "step": 9491 + }, + { + "epoch": 0.42026499009331103, + "grad_norm": 1.498714566230774, + "learning_rate": 1.4376394394900714e-05, + "loss": 0.7793, + "step": 9492 + }, + { + "epoch": 0.42030926579812494, + "grad_norm": 1.3697571754455566, + "learning_rate": 1.4375004749579755e-05, + "loss": 0.7719, + "step": 9493 + }, + { + "epoch": 0.4203535415029388, + "grad_norm": 1.3794184923171997, + "learning_rate": 1.4373614999765369e-05, + "loss": 0.7906, + "step": 9494 + }, + { + "epoch": 0.42039781720775266, + "grad_norm": 1.3033816814422607, + "learning_rate": 1.437222514549075e-05, + "loss": 0.6872, + "step": 9495 + }, + { + "epoch": 0.42044209291256657, + "grad_norm": 1.4255443811416626, + "learning_rate": 1.4370835186789093e-05, + "loss": 0.6475, + "step": 9496 + }, + { + "epoch": 0.4204863686173804, + "grad_norm": 1.4510656595230103, + "learning_rate": 1.4369445123693595e-05, + "loss": 0.7253, + "step": 9497 + }, + { + "epoch": 0.4205306443221943, + "grad_norm": 1.8164006471633911, + "learning_rate": 1.4368054956237463e-05, + "loss": 0.7554, + "step": 9498 + }, + { + "epoch": 0.4205749200270082, + "grad_norm": 1.4028077125549316, + "learning_rate": 1.436666468445389e-05, + "loss": 0.8123, + "step": 9499 + }, + { + "epoch": 0.42061919573182205, + "grad_norm": 1.3418965339660645, + "learning_rate": 1.4365274308376088e-05, + "loss": 0.73, + "step": 9500 + }, + { + "epoch": 0.4206634714366359, + "grad_norm": 1.4884065389633179, + "learning_rate": 1.4363883828037264e-05, + "loss": 0.7862, + "step": 9501 + }, + { + "epoch": 0.4207077471414498, + "grad_norm": 1.7962623834609985, + "learning_rate": 1.4362493243470628e-05, + "loss": 0.7767, + "step": 9502 + }, + { + "epoch": 0.4207520228462637, + "grad_norm": 1.443573236465454, + "learning_rate": 1.4361102554709391e-05, + "loss": 0.8068, + "step": 9503 + }, + { + "epoch": 0.4207962985510776, + "grad_norm": 1.371424674987793, + "learning_rate": 1.4359711761786775e-05, + "loss": 0.7326, + "step": 9504 + }, + { + "epoch": 0.42084057425589144, + "grad_norm": 1.9679982662200928, + "learning_rate": 1.4358320864735991e-05, + "loss": 0.8238, + "step": 9505 + }, + { + "epoch": 0.4208848499607053, + "grad_norm": 1.6757408380508423, + "learning_rate": 1.4356929863590258e-05, + "loss": 0.8295, + "step": 9506 + }, + { + "epoch": 0.4209291256655192, + "grad_norm": 1.6233283281326294, + "learning_rate": 1.4355538758382805e-05, + "loss": 0.7382, + "step": 9507 + }, + { + "epoch": 0.42097340137033307, + "grad_norm": 1.6748086214065552, + "learning_rate": 1.4354147549146858e-05, + "loss": 0.8996, + "step": 9508 + }, + { + "epoch": 0.4210176770751469, + "grad_norm": 1.4266514778137207, + "learning_rate": 1.435275623591564e-05, + "loss": 0.7784, + "step": 9509 + }, + { + "epoch": 0.42106195277996084, + "grad_norm": 1.249657392501831, + "learning_rate": 1.4351364818722385e-05, + "loss": 0.7205, + "step": 9510 + }, + { + "epoch": 0.4211062284847747, + "grad_norm": 1.4630063772201538, + "learning_rate": 1.4349973297600321e-05, + "loss": 0.7233, + "step": 9511 + }, + { + "epoch": 0.42115050418958855, + "grad_norm": 1.2762079238891602, + "learning_rate": 1.4348581672582688e-05, + "loss": 0.7143, + "step": 9512 + }, + { + "epoch": 0.42119477989440246, + "grad_norm": 1.4706730842590332, + "learning_rate": 1.4347189943702726e-05, + "loss": 0.7426, + "step": 9513 + }, + { + "epoch": 0.4212390555992163, + "grad_norm": 1.758023738861084, + "learning_rate": 1.4345798110993668e-05, + "loss": 0.7802, + "step": 9514 + }, + { + "epoch": 0.4212833313040302, + "grad_norm": 1.5120902061462402, + "learning_rate": 1.4344406174488761e-05, + "loss": 0.7468, + "step": 9515 + }, + { + "epoch": 0.4213276070088441, + "grad_norm": 1.8859517574310303, + "learning_rate": 1.4343014134221249e-05, + "loss": 0.823, + "step": 9516 + }, + { + "epoch": 0.42137188271365794, + "grad_norm": 1.739593744277954, + "learning_rate": 1.4341621990224383e-05, + "loss": 0.7596, + "step": 9517 + }, + { + "epoch": 0.4214161584184718, + "grad_norm": 1.3795077800750732, + "learning_rate": 1.434022974253141e-05, + "loss": 0.7371, + "step": 9518 + }, + { + "epoch": 0.4214604341232857, + "grad_norm": 1.5226778984069824, + "learning_rate": 1.4338837391175582e-05, + "loss": 0.7513, + "step": 9519 + }, + { + "epoch": 0.42150470982809957, + "grad_norm": 1.4886906147003174, + "learning_rate": 1.433744493619016e-05, + "loss": 0.7425, + "step": 9520 + }, + { + "epoch": 0.4215489855329135, + "grad_norm": 1.0761622190475464, + "learning_rate": 1.4336052377608391e-05, + "loss": 0.6398, + "step": 9521 + }, + { + "epoch": 0.42159326123772733, + "grad_norm": 1.3034147024154663, + "learning_rate": 1.4334659715463546e-05, + "loss": 0.6919, + "step": 9522 + }, + { + "epoch": 0.4216375369425412, + "grad_norm": 1.199337124824524, + "learning_rate": 1.4333266949788885e-05, + "loss": 0.6677, + "step": 9523 + }, + { + "epoch": 0.4216818126473551, + "grad_norm": 1.3433769941329956, + "learning_rate": 1.4331874080617665e-05, + "loss": 0.6809, + "step": 9524 + }, + { + "epoch": 0.42172608835216896, + "grad_norm": 1.3418081998825073, + "learning_rate": 1.4330481107983163e-05, + "loss": 0.7451, + "step": 9525 + }, + { + "epoch": 0.4217703640569828, + "grad_norm": 1.4611907005310059, + "learning_rate": 1.4329088031918645e-05, + "loss": 0.8121, + "step": 9526 + }, + { + "epoch": 0.42181463976179673, + "grad_norm": 1.356440544128418, + "learning_rate": 1.4327694852457387e-05, + "loss": 0.6966, + "step": 9527 + }, + { + "epoch": 0.4218589154666106, + "grad_norm": 1.6140084266662598, + "learning_rate": 1.4326301569632659e-05, + "loss": 0.7088, + "step": 9528 + }, + { + "epoch": 0.42190319117142444, + "grad_norm": 1.5014350414276123, + "learning_rate": 1.4324908183477739e-05, + "loss": 0.8161, + "step": 9529 + }, + { + "epoch": 0.42194746687623835, + "grad_norm": 1.36347234249115, + "learning_rate": 1.4323514694025913e-05, + "loss": 0.7218, + "step": 9530 + }, + { + "epoch": 0.4219917425810522, + "grad_norm": 1.2825123071670532, + "learning_rate": 1.4322121101310454e-05, + "loss": 0.6953, + "step": 9531 + }, + { + "epoch": 0.42203601828586607, + "grad_norm": 1.4942846298217773, + "learning_rate": 1.4320727405364655e-05, + "loss": 0.703, + "step": 9532 + }, + { + "epoch": 0.42208029399068, + "grad_norm": 1.277251958847046, + "learning_rate": 1.43193336062218e-05, + "loss": 0.7145, + "step": 9533 + }, + { + "epoch": 0.42212456969549383, + "grad_norm": 1.6313397884368896, + "learning_rate": 1.4317939703915176e-05, + "loss": 0.8377, + "step": 9534 + }, + { + "epoch": 0.4221688454003077, + "grad_norm": 1.9654786586761475, + "learning_rate": 1.4316545698478079e-05, + "loss": 0.7669, + "step": 9535 + }, + { + "epoch": 0.4222131211051216, + "grad_norm": 1.5768903493881226, + "learning_rate": 1.4315151589943803e-05, + "loss": 0.7192, + "step": 9536 + }, + { + "epoch": 0.42225739680993546, + "grad_norm": 1.233923077583313, + "learning_rate": 1.4313757378345646e-05, + "loss": 0.7236, + "step": 9537 + }, + { + "epoch": 0.42230167251474937, + "grad_norm": 1.635024070739746, + "learning_rate": 1.4312363063716906e-05, + "loss": 0.8008, + "step": 9538 + }, + { + "epoch": 0.4223459482195632, + "grad_norm": 1.7595797777175903, + "learning_rate": 1.4310968646090884e-05, + "loss": 0.7632, + "step": 9539 + }, + { + "epoch": 0.4223902239243771, + "grad_norm": 1.4952093362808228, + "learning_rate": 1.4309574125500887e-05, + "loss": 0.7616, + "step": 9540 + }, + { + "epoch": 0.422434499629191, + "grad_norm": 1.2485822439193726, + "learning_rate": 1.4308179501980221e-05, + "loss": 0.6614, + "step": 9541 + }, + { + "epoch": 0.42247877533400485, + "grad_norm": 1.5566130876541138, + "learning_rate": 1.4306784775562194e-05, + "loss": 0.787, + "step": 9542 + }, + { + "epoch": 0.4225230510388187, + "grad_norm": 1.2513079643249512, + "learning_rate": 1.4305389946280121e-05, + "loss": 0.6805, + "step": 9543 + }, + { + "epoch": 0.4225673267436326, + "grad_norm": 1.638992428779602, + "learning_rate": 1.4303995014167314e-05, + "loss": 0.7661, + "step": 9544 + }, + { + "epoch": 0.4226116024484465, + "grad_norm": 1.1764541864395142, + "learning_rate": 1.4302599979257092e-05, + "loss": 0.6543, + "step": 9545 + }, + { + "epoch": 0.42265587815326033, + "grad_norm": 1.34342622756958, + "learning_rate": 1.4301204841582772e-05, + "loss": 0.701, + "step": 9546 + }, + { + "epoch": 0.42270015385807425, + "grad_norm": 1.3319107294082642, + "learning_rate": 1.4299809601177676e-05, + "loss": 0.6699, + "step": 9547 + }, + { + "epoch": 0.4227444295628881, + "grad_norm": 1.545508861541748, + "learning_rate": 1.429841425807513e-05, + "loss": 0.8623, + "step": 9548 + }, + { + "epoch": 0.42278870526770196, + "grad_norm": 1.9688246250152588, + "learning_rate": 1.4297018812308457e-05, + "loss": 0.7273, + "step": 9549 + }, + { + "epoch": 0.42283298097251587, + "grad_norm": 1.4318705797195435, + "learning_rate": 1.429562326391099e-05, + "loss": 0.6602, + "step": 9550 + }, + { + "epoch": 0.4228772566773297, + "grad_norm": 1.3217462301254272, + "learning_rate": 1.4294227612916061e-05, + "loss": 0.7149, + "step": 9551 + }, + { + "epoch": 0.4229215323821436, + "grad_norm": 1.4613137245178223, + "learning_rate": 1.4292831859356997e-05, + "loss": 0.8186, + "step": 9552 + }, + { + "epoch": 0.4229658080869575, + "grad_norm": 1.3941375017166138, + "learning_rate": 1.4291436003267144e-05, + "loss": 0.7405, + "step": 9553 + }, + { + "epoch": 0.42301008379177135, + "grad_norm": 1.5570168495178223, + "learning_rate": 1.429004004467983e-05, + "loss": 0.7568, + "step": 9554 + }, + { + "epoch": 0.42305435949658526, + "grad_norm": 1.3871524333953857, + "learning_rate": 1.428864398362841e-05, + "loss": 0.6761, + "step": 9555 + }, + { + "epoch": 0.4230986352013991, + "grad_norm": 1.9935773611068726, + "learning_rate": 1.4287247820146217e-05, + "loss": 0.714, + "step": 9556 + }, + { + "epoch": 0.423142910906213, + "grad_norm": 1.6360758543014526, + "learning_rate": 1.4285851554266601e-05, + "loss": 0.7787, + "step": 9557 + }, + { + "epoch": 0.4231871866110269, + "grad_norm": 1.1942423582077026, + "learning_rate": 1.428445518602291e-05, + "loss": 0.6892, + "step": 9558 + }, + { + "epoch": 0.42323146231584075, + "grad_norm": 1.5251610279083252, + "learning_rate": 1.4283058715448494e-05, + "loss": 0.7794, + "step": 9559 + }, + { + "epoch": 0.4232757380206546, + "grad_norm": 1.5290498733520508, + "learning_rate": 1.4281662142576711e-05, + "loss": 0.7948, + "step": 9560 + }, + { + "epoch": 0.4233200137254685, + "grad_norm": 1.3353737592697144, + "learning_rate": 1.4280265467440914e-05, + "loss": 0.7299, + "step": 9561 + }, + { + "epoch": 0.42336428943028237, + "grad_norm": 1.5008615255355835, + "learning_rate": 1.427886869007446e-05, + "loss": 0.788, + "step": 9562 + }, + { + "epoch": 0.4234085651350962, + "grad_norm": 1.3510602712631226, + "learning_rate": 1.427747181051071e-05, + "loss": 0.7094, + "step": 9563 + }, + { + "epoch": 0.42345284083991014, + "grad_norm": 1.5122997760772705, + "learning_rate": 1.4276074828783029e-05, + "loss": 0.7336, + "step": 9564 + }, + { + "epoch": 0.423497116544724, + "grad_norm": 1.3080672025680542, + "learning_rate": 1.4274677744924787e-05, + "loss": 0.7193, + "step": 9565 + }, + { + "epoch": 0.42354139224953785, + "grad_norm": 1.1972676515579224, + "learning_rate": 1.4273280558969342e-05, + "loss": 0.7033, + "step": 9566 + }, + { + "epoch": 0.42358566795435176, + "grad_norm": 1.2901078462600708, + "learning_rate": 1.4271883270950073e-05, + "loss": 0.7357, + "step": 9567 + }, + { + "epoch": 0.4236299436591656, + "grad_norm": 1.3422905206680298, + "learning_rate": 1.427048588090035e-05, + "loss": 0.6563, + "step": 9568 + }, + { + "epoch": 0.4236742193639795, + "grad_norm": 1.4261878728866577, + "learning_rate": 1.4269088388853548e-05, + "loss": 0.7755, + "step": 9569 + }, + { + "epoch": 0.4237184950687934, + "grad_norm": 1.3504012823104858, + "learning_rate": 1.4267690794843048e-05, + "loss": 0.7301, + "step": 9570 + }, + { + "epoch": 0.42376277077360724, + "grad_norm": 1.529634714126587, + "learning_rate": 1.4266293098902227e-05, + "loss": 0.8307, + "step": 9571 + }, + { + "epoch": 0.4238070464784211, + "grad_norm": 1.4351030588150024, + "learning_rate": 1.4264895301064465e-05, + "loss": 0.7929, + "step": 9572 + }, + { + "epoch": 0.423851322183235, + "grad_norm": 1.3068411350250244, + "learning_rate": 1.4263497401363157e-05, + "loss": 0.7412, + "step": 9573 + }, + { + "epoch": 0.42389559788804887, + "grad_norm": 1.6386146545410156, + "learning_rate": 1.4262099399831683e-05, + "loss": 0.7077, + "step": 9574 + }, + { + "epoch": 0.4239398735928628, + "grad_norm": 1.6583160161972046, + "learning_rate": 1.4260701296503436e-05, + "loss": 0.791, + "step": 9575 + }, + { + "epoch": 0.42398414929767664, + "grad_norm": 1.583565592765808, + "learning_rate": 1.4259303091411807e-05, + "loss": 0.7021, + "step": 9576 + }, + { + "epoch": 0.4240284250024905, + "grad_norm": 1.3206086158752441, + "learning_rate": 1.425790478459019e-05, + "loss": 0.8242, + "step": 9577 + }, + { + "epoch": 0.4240727007073044, + "grad_norm": 1.1918920278549194, + "learning_rate": 1.4256506376071988e-05, + "loss": 0.6626, + "step": 9578 + }, + { + "epoch": 0.42411697641211826, + "grad_norm": 1.191408395767212, + "learning_rate": 1.4255107865890597e-05, + "loss": 0.6356, + "step": 9579 + }, + { + "epoch": 0.4241612521169321, + "grad_norm": 1.262261152267456, + "learning_rate": 1.4253709254079415e-05, + "loss": 0.6613, + "step": 9580 + }, + { + "epoch": 0.42420552782174603, + "grad_norm": 1.658327579498291, + "learning_rate": 1.4252310540671853e-05, + "loss": 0.8603, + "step": 9581 + }, + { + "epoch": 0.4242498035265599, + "grad_norm": 1.2279492616653442, + "learning_rate": 1.4250911725701315e-05, + "loss": 0.7008, + "step": 9582 + }, + { + "epoch": 0.42429407923137374, + "grad_norm": 1.3585604429244995, + "learning_rate": 1.4249512809201214e-05, + "loss": 0.7562, + "step": 9583 + }, + { + "epoch": 0.42433835493618766, + "grad_norm": 1.3868974447250366, + "learning_rate": 1.424811379120496e-05, + "loss": 0.7465, + "step": 9584 + }, + { + "epoch": 0.4243826306410015, + "grad_norm": 1.554848313331604, + "learning_rate": 1.4246714671745964e-05, + "loss": 0.7617, + "step": 9585 + }, + { + "epoch": 0.42442690634581537, + "grad_norm": 1.7930630445480347, + "learning_rate": 1.4245315450857649e-05, + "loss": 0.912, + "step": 9586 + }, + { + "epoch": 0.4244711820506293, + "grad_norm": 1.2873561382293701, + "learning_rate": 1.4243916128573428e-05, + "loss": 0.777, + "step": 9587 + }, + { + "epoch": 0.42451545775544314, + "grad_norm": 1.5105968713760376, + "learning_rate": 1.4242516704926728e-05, + "loss": 0.7277, + "step": 9588 + }, + { + "epoch": 0.424559733460257, + "grad_norm": 2.2706973552703857, + "learning_rate": 1.4241117179950972e-05, + "loss": 0.9781, + "step": 9589 + }, + { + "epoch": 0.4246040091650709, + "grad_norm": 1.4080618619918823, + "learning_rate": 1.4239717553679585e-05, + "loss": 0.7319, + "step": 9590 + }, + { + "epoch": 0.42464828486988476, + "grad_norm": 1.3995047807693481, + "learning_rate": 1.4238317826145993e-05, + "loss": 0.7767, + "step": 9591 + }, + { + "epoch": 0.4246925605746987, + "grad_norm": 1.3242563009262085, + "learning_rate": 1.423691799738363e-05, + "loss": 0.7438, + "step": 9592 + }, + { + "epoch": 0.42473683627951253, + "grad_norm": 1.2074899673461914, + "learning_rate": 1.4235518067425936e-05, + "loss": 0.6198, + "step": 9593 + }, + { + "epoch": 0.4247811119843264, + "grad_norm": 1.5039777755737305, + "learning_rate": 1.4234118036306338e-05, + "loss": 0.656, + "step": 9594 + }, + { + "epoch": 0.4248253876891403, + "grad_norm": 1.3484790325164795, + "learning_rate": 1.423271790405828e-05, + "loss": 0.685, + "step": 9595 + }, + { + "epoch": 0.42486966339395416, + "grad_norm": 1.4840627908706665, + "learning_rate": 1.42313176707152e-05, + "loss": 0.7165, + "step": 9596 + }, + { + "epoch": 0.424913939098768, + "grad_norm": 1.4792664051055908, + "learning_rate": 1.4229917336310544e-05, + "loss": 0.7416, + "step": 9597 + }, + { + "epoch": 0.4249582148035819, + "grad_norm": 1.335139274597168, + "learning_rate": 1.4228516900877756e-05, + "loss": 0.7307, + "step": 9598 + }, + { + "epoch": 0.4250024905083958, + "grad_norm": 1.3589903116226196, + "learning_rate": 1.4227116364450281e-05, + "loss": 0.6901, + "step": 9599 + }, + { + "epoch": 0.42504676621320964, + "grad_norm": 1.5203819274902344, + "learning_rate": 1.4225715727061577e-05, + "loss": 0.7299, + "step": 9600 + }, + { + "epoch": 0.42509104191802355, + "grad_norm": 1.5456629991531372, + "learning_rate": 1.4224314988745093e-05, + "loss": 0.8089, + "step": 9601 + }, + { + "epoch": 0.4251353176228374, + "grad_norm": 1.72113037109375, + "learning_rate": 1.4222914149534286e-05, + "loss": 0.7787, + "step": 9602 + }, + { + "epoch": 0.42517959332765126, + "grad_norm": 1.4128905534744263, + "learning_rate": 1.4221513209462615e-05, + "loss": 0.7322, + "step": 9603 + }, + { + "epoch": 0.4252238690324652, + "grad_norm": 1.2983486652374268, + "learning_rate": 1.4220112168563533e-05, + "loss": 0.669, + "step": 9604 + }, + { + "epoch": 0.42526814473727903, + "grad_norm": 1.1477751731872559, + "learning_rate": 1.4218711026870513e-05, + "loss": 0.6726, + "step": 9605 + }, + { + "epoch": 0.4253124204420929, + "grad_norm": 1.6059513092041016, + "learning_rate": 1.421730978441701e-05, + "loss": 0.6681, + "step": 9606 + }, + { + "epoch": 0.4253566961469068, + "grad_norm": 1.31948721408844, + "learning_rate": 1.4215908441236502e-05, + "loss": 0.7462, + "step": 9607 + }, + { + "epoch": 0.42540097185172066, + "grad_norm": 1.699472188949585, + "learning_rate": 1.4214506997362451e-05, + "loss": 0.7447, + "step": 9608 + }, + { + "epoch": 0.42544524755653457, + "grad_norm": 1.2931164503097534, + "learning_rate": 1.4213105452828332e-05, + "loss": 0.661, + "step": 9609 + }, + { + "epoch": 0.4254895232613484, + "grad_norm": 1.4902456998825073, + "learning_rate": 1.4211703807667618e-05, + "loss": 0.7459, + "step": 9610 + }, + { + "epoch": 0.4255337989661623, + "grad_norm": 1.2087889909744263, + "learning_rate": 1.421030206191379e-05, + "loss": 0.7033, + "step": 9611 + }, + { + "epoch": 0.4255780746709762, + "grad_norm": 1.4981327056884766, + "learning_rate": 1.4208900215600328e-05, + "loss": 0.7268, + "step": 9612 + }, + { + "epoch": 0.42562235037579005, + "grad_norm": 1.4334452152252197, + "learning_rate": 1.4207498268760707e-05, + "loss": 0.7333, + "step": 9613 + }, + { + "epoch": 0.4256666260806039, + "grad_norm": 1.4604178667068481, + "learning_rate": 1.4206096221428416e-05, + "loss": 0.8229, + "step": 9614 + }, + { + "epoch": 0.4257109017854178, + "grad_norm": 1.5251424312591553, + "learning_rate": 1.4204694073636941e-05, + "loss": 0.6846, + "step": 9615 + }, + { + "epoch": 0.4257551774902317, + "grad_norm": 1.2821511030197144, + "learning_rate": 1.4203291825419774e-05, + "loss": 0.6789, + "step": 9616 + }, + { + "epoch": 0.42579945319504553, + "grad_norm": 1.652622103691101, + "learning_rate": 1.4201889476810406e-05, + "loss": 0.7623, + "step": 9617 + }, + { + "epoch": 0.42584372889985944, + "grad_norm": 1.395677089691162, + "learning_rate": 1.4200487027842324e-05, + "loss": 0.7248, + "step": 9618 + }, + { + "epoch": 0.4258880046046733, + "grad_norm": 1.2794325351715088, + "learning_rate": 1.4199084478549031e-05, + "loss": 0.7196, + "step": 9619 + }, + { + "epoch": 0.42593228030948715, + "grad_norm": 1.4324102401733398, + "learning_rate": 1.4197681828964028e-05, + "loss": 0.7305, + "step": 9620 + }, + { + "epoch": 0.42597655601430107, + "grad_norm": 1.5608282089233398, + "learning_rate": 1.419627907912081e-05, + "loss": 0.7609, + "step": 9621 + }, + { + "epoch": 0.4260208317191149, + "grad_norm": 1.5477588176727295, + "learning_rate": 1.4194876229052882e-05, + "loss": 0.8166, + "step": 9622 + }, + { + "epoch": 0.4260651074239288, + "grad_norm": 1.3831413984298706, + "learning_rate": 1.419347327879375e-05, + "loss": 0.7101, + "step": 9623 + }, + { + "epoch": 0.4261093831287427, + "grad_norm": 1.3892735242843628, + "learning_rate": 1.4192070228376924e-05, + "loss": 0.7141, + "step": 9624 + }, + { + "epoch": 0.42615365883355655, + "grad_norm": 1.4403778314590454, + "learning_rate": 1.4190667077835915e-05, + "loss": 0.7191, + "step": 9625 + }, + { + "epoch": 0.42619793453837046, + "grad_norm": 1.4169925451278687, + "learning_rate": 1.4189263827204237e-05, + "loss": 0.6676, + "step": 9626 + }, + { + "epoch": 0.4262422102431843, + "grad_norm": 1.4389687776565552, + "learning_rate": 1.41878604765154e-05, + "loss": 0.7737, + "step": 9627 + }, + { + "epoch": 0.4262864859479982, + "grad_norm": 1.3597241640090942, + "learning_rate": 1.4186457025802927e-05, + "loss": 0.68, + "step": 9628 + }, + { + "epoch": 0.4263307616528121, + "grad_norm": 1.4856566190719604, + "learning_rate": 1.4185053475100335e-05, + "loss": 0.7823, + "step": 9629 + }, + { + "epoch": 0.42637503735762594, + "grad_norm": 1.3776354789733887, + "learning_rate": 1.418364982444115e-05, + "loss": 0.7902, + "step": 9630 + }, + { + "epoch": 0.4264193130624398, + "grad_norm": 1.665885329246521, + "learning_rate": 1.4182246073858897e-05, + "loss": 0.8244, + "step": 9631 + }, + { + "epoch": 0.4264635887672537, + "grad_norm": 1.2899328470230103, + "learning_rate": 1.41808422233871e-05, + "loss": 0.7199, + "step": 9632 + }, + { + "epoch": 0.42650786447206757, + "grad_norm": 1.489464521408081, + "learning_rate": 1.417943827305929e-05, + "loss": 0.7841, + "step": 9633 + }, + { + "epoch": 0.4265521401768814, + "grad_norm": 1.1475530862808228, + "learning_rate": 1.4178034222908999e-05, + "loss": 0.7684, + "step": 9634 + }, + { + "epoch": 0.42659641588169533, + "grad_norm": 1.6427351236343384, + "learning_rate": 1.4176630072969764e-05, + "loss": 0.7636, + "step": 9635 + }, + { + "epoch": 0.4266406915865092, + "grad_norm": 1.281045913696289, + "learning_rate": 1.4175225823275121e-05, + "loss": 0.6654, + "step": 9636 + }, + { + "epoch": 0.42668496729132305, + "grad_norm": 1.4898033142089844, + "learning_rate": 1.4173821473858609e-05, + "loss": 0.7028, + "step": 9637 + }, + { + "epoch": 0.42672924299613696, + "grad_norm": 1.482749581336975, + "learning_rate": 1.4172417024753767e-05, + "loss": 0.7279, + "step": 9638 + }, + { + "epoch": 0.4267735187009508, + "grad_norm": 1.5773578882217407, + "learning_rate": 1.4171012475994142e-05, + "loss": 0.7814, + "step": 9639 + }, + { + "epoch": 0.42681779440576467, + "grad_norm": 1.326355218887329, + "learning_rate": 1.4169607827613284e-05, + "loss": 0.6881, + "step": 9640 + }, + { + "epoch": 0.4268620701105786, + "grad_norm": 1.433185338973999, + "learning_rate": 1.4168203079644734e-05, + "loss": 0.6961, + "step": 9641 + }, + { + "epoch": 0.42690634581539244, + "grad_norm": 1.3887817859649658, + "learning_rate": 1.416679823212205e-05, + "loss": 0.7884, + "step": 9642 + }, + { + "epoch": 0.4269506215202063, + "grad_norm": 1.2795411348342896, + "learning_rate": 1.4165393285078782e-05, + "loss": 0.6844, + "step": 9643 + }, + { + "epoch": 0.4269948972250202, + "grad_norm": 1.7521713972091675, + "learning_rate": 1.4163988238548485e-05, + "loss": 0.7956, + "step": 9644 + }, + { + "epoch": 0.42703917292983407, + "grad_norm": 1.3697268962860107, + "learning_rate": 1.4162583092564722e-05, + "loss": 0.7467, + "step": 9645 + }, + { + "epoch": 0.427083448634648, + "grad_norm": 1.5827829837799072, + "learning_rate": 1.4161177847161049e-05, + "loss": 0.7944, + "step": 9646 + }, + { + "epoch": 0.42712772433946183, + "grad_norm": 1.3816648721694946, + "learning_rate": 1.4159772502371034e-05, + "loss": 0.7091, + "step": 9647 + }, + { + "epoch": 0.4271720000442757, + "grad_norm": 1.366586685180664, + "learning_rate": 1.4158367058228238e-05, + "loss": 0.6654, + "step": 9648 + }, + { + "epoch": 0.4272162757490896, + "grad_norm": 1.114200472831726, + "learning_rate": 1.415696151476623e-05, + "loss": 0.6748, + "step": 9649 + }, + { + "epoch": 0.42726055145390346, + "grad_norm": 1.3651783466339111, + "learning_rate": 1.4155555872018582e-05, + "loss": 0.7625, + "step": 9650 + }, + { + "epoch": 0.4273048271587173, + "grad_norm": 1.5640294551849365, + "learning_rate": 1.4154150130018867e-05, + "loss": 0.7095, + "step": 9651 + }, + { + "epoch": 0.4273491028635312, + "grad_norm": 1.5925734043121338, + "learning_rate": 1.4152744288800657e-05, + "loss": 0.7245, + "step": 9652 + }, + { + "epoch": 0.4273933785683451, + "grad_norm": 1.4534186124801636, + "learning_rate": 1.415133834839753e-05, + "loss": 0.7708, + "step": 9653 + }, + { + "epoch": 0.42743765427315894, + "grad_norm": 1.6454464197158813, + "learning_rate": 1.4149932308843069e-05, + "loss": 0.8294, + "step": 9654 + }, + { + "epoch": 0.42748192997797285, + "grad_norm": 1.5169200897216797, + "learning_rate": 1.4148526170170853e-05, + "loss": 0.7521, + "step": 9655 + }, + { + "epoch": 0.4275262056827867, + "grad_norm": 1.4906413555145264, + "learning_rate": 1.4147119932414465e-05, + "loss": 0.7737, + "step": 9656 + }, + { + "epoch": 0.42757048138760056, + "grad_norm": 1.2885764837265015, + "learning_rate": 1.4145713595607498e-05, + "loss": 0.6952, + "step": 9657 + }, + { + "epoch": 0.4276147570924145, + "grad_norm": 1.5301231145858765, + "learning_rate": 1.4144307159783533e-05, + "loss": 0.751, + "step": 9658 + }, + { + "epoch": 0.42765903279722833, + "grad_norm": 1.4668809175491333, + "learning_rate": 1.4142900624976168e-05, + "loss": 0.7169, + "step": 9659 + }, + { + "epoch": 0.4277033085020422, + "grad_norm": 1.3550399541854858, + "learning_rate": 1.4141493991218998e-05, + "loss": 0.7638, + "step": 9660 + }, + { + "epoch": 0.4277475842068561, + "grad_norm": 1.3618800640106201, + "learning_rate": 1.414008725854561e-05, + "loss": 0.7086, + "step": 9661 + }, + { + "epoch": 0.42779185991166996, + "grad_norm": 1.3426297903060913, + "learning_rate": 1.4138680426989613e-05, + "loss": 0.7348, + "step": 9662 + }, + { + "epoch": 0.42783613561648387, + "grad_norm": 1.478782057762146, + "learning_rate": 1.4137273496584602e-05, + "loss": 0.7273, + "step": 9663 + }, + { + "epoch": 0.4278804113212977, + "grad_norm": 1.3195672035217285, + "learning_rate": 1.4135866467364183e-05, + "loss": 0.7061, + "step": 9664 + }, + { + "epoch": 0.4279246870261116, + "grad_norm": 1.6544640064239502, + "learning_rate": 1.4134459339361963e-05, + "loss": 0.6762, + "step": 9665 + }, + { + "epoch": 0.4279689627309255, + "grad_norm": 2.0522420406341553, + "learning_rate": 1.4133052112611545e-05, + "loss": 0.8653, + "step": 9666 + }, + { + "epoch": 0.42801323843573935, + "grad_norm": 1.957617998123169, + "learning_rate": 1.4131644787146543e-05, + "loss": 0.7951, + "step": 9667 + }, + { + "epoch": 0.4280575141405532, + "grad_norm": 1.29119074344635, + "learning_rate": 1.4130237363000573e-05, + "loss": 0.6731, + "step": 9668 + }, + { + "epoch": 0.4281017898453671, + "grad_norm": 1.3532984256744385, + "learning_rate": 1.4128829840207241e-05, + "loss": 0.786, + "step": 9669 + }, + { + "epoch": 0.428146065550181, + "grad_norm": 1.5009592771530151, + "learning_rate": 1.4127422218800174e-05, + "loss": 0.7148, + "step": 9670 + }, + { + "epoch": 0.42819034125499483, + "grad_norm": 1.425110936164856, + "learning_rate": 1.4126014498812986e-05, + "loss": 0.8065, + "step": 9671 + }, + { + "epoch": 0.42823461695980874, + "grad_norm": 1.3736364841461182, + "learning_rate": 1.4124606680279302e-05, + "loss": 0.7073, + "step": 9672 + }, + { + "epoch": 0.4282788926646226, + "grad_norm": 1.476682186126709, + "learning_rate": 1.4123198763232746e-05, + "loss": 0.8116, + "step": 9673 + }, + { + "epoch": 0.42832316836943646, + "grad_norm": 1.2268604040145874, + "learning_rate": 1.4121790747706943e-05, + "loss": 0.7122, + "step": 9674 + }, + { + "epoch": 0.42836744407425037, + "grad_norm": 1.3765395879745483, + "learning_rate": 1.4120382633735528e-05, + "loss": 0.6924, + "step": 9675 + }, + { + "epoch": 0.4284117197790642, + "grad_norm": 1.7284780740737915, + "learning_rate": 1.4118974421352124e-05, + "loss": 0.7184, + "step": 9676 + }, + { + "epoch": 0.4284559954838781, + "grad_norm": 1.2849189043045044, + "learning_rate": 1.4117566110590372e-05, + "loss": 0.6905, + "step": 9677 + }, + { + "epoch": 0.428500271188692, + "grad_norm": 1.498203158378601, + "learning_rate": 1.4116157701483908e-05, + "loss": 0.8345, + "step": 9678 + }, + { + "epoch": 0.42854454689350585, + "grad_norm": 1.6701890230178833, + "learning_rate": 1.4114749194066364e-05, + "loss": 0.8448, + "step": 9679 + }, + { + "epoch": 0.42858882259831976, + "grad_norm": 1.3915913105010986, + "learning_rate": 1.411334058837139e-05, + "loss": 0.7323, + "step": 9680 + }, + { + "epoch": 0.4286330983031336, + "grad_norm": 1.2486196756362915, + "learning_rate": 1.4111931884432624e-05, + "loss": 0.7161, + "step": 9681 + }, + { + "epoch": 0.4286773740079475, + "grad_norm": 1.237999439239502, + "learning_rate": 1.4110523082283714e-05, + "loss": 0.7276, + "step": 9682 + }, + { + "epoch": 0.4287216497127614, + "grad_norm": 1.4530820846557617, + "learning_rate": 1.4109114181958305e-05, + "loss": 0.7792, + "step": 9683 + }, + { + "epoch": 0.42876592541757524, + "grad_norm": 1.5207276344299316, + "learning_rate": 1.410770518349005e-05, + "loss": 0.8151, + "step": 9684 + }, + { + "epoch": 0.4288102011223891, + "grad_norm": 1.3643943071365356, + "learning_rate": 1.4106296086912604e-05, + "loss": 0.7161, + "step": 9685 + }, + { + "epoch": 0.428854476827203, + "grad_norm": 1.1546785831451416, + "learning_rate": 1.4104886892259615e-05, + "loss": 0.6369, + "step": 9686 + }, + { + "epoch": 0.42889875253201687, + "grad_norm": 1.5330350399017334, + "learning_rate": 1.4103477599564749e-05, + "loss": 0.7199, + "step": 9687 + }, + { + "epoch": 0.4289430282368307, + "grad_norm": 1.4765928983688354, + "learning_rate": 1.410206820886166e-05, + "loss": 0.6924, + "step": 9688 + }, + { + "epoch": 0.42898730394164464, + "grad_norm": 1.6215201616287231, + "learning_rate": 1.410065872018401e-05, + "loss": 0.8097, + "step": 9689 + }, + { + "epoch": 0.4290315796464585, + "grad_norm": 1.4072539806365967, + "learning_rate": 1.4099249133565469e-05, + "loss": 0.7211, + "step": 9690 + }, + { + "epoch": 0.42907585535127235, + "grad_norm": 1.5651023387908936, + "learning_rate": 1.4097839449039696e-05, + "loss": 0.7089, + "step": 9691 + }, + { + "epoch": 0.42912013105608626, + "grad_norm": 1.5740708112716675, + "learning_rate": 1.409642966664037e-05, + "loss": 0.7559, + "step": 9692 + }, + { + "epoch": 0.4291644067609001, + "grad_norm": 1.448853611946106, + "learning_rate": 1.4095019786401153e-05, + "loss": 0.6969, + "step": 9693 + }, + { + "epoch": 0.429208682465714, + "grad_norm": 1.5987131595611572, + "learning_rate": 1.4093609808355722e-05, + "loss": 0.8456, + "step": 9694 + }, + { + "epoch": 0.4292529581705279, + "grad_norm": 1.3241935968399048, + "learning_rate": 1.4092199732537756e-05, + "loss": 0.7255, + "step": 9695 + }, + { + "epoch": 0.42929723387534174, + "grad_norm": 1.5737922191619873, + "learning_rate": 1.409078955898093e-05, + "loss": 0.7459, + "step": 9696 + }, + { + "epoch": 0.42934150958015566, + "grad_norm": 1.5595207214355469, + "learning_rate": 1.4089379287718927e-05, + "loss": 0.7152, + "step": 9697 + }, + { + "epoch": 0.4293857852849695, + "grad_norm": 1.1442680358886719, + "learning_rate": 1.4087968918785429e-05, + "loss": 0.6323, + "step": 9698 + }, + { + "epoch": 0.42943006098978337, + "grad_norm": 1.8989514112472534, + "learning_rate": 1.4086558452214121e-05, + "loss": 0.7942, + "step": 9699 + }, + { + "epoch": 0.4294743366945973, + "grad_norm": 1.4316555261611938, + "learning_rate": 1.4085147888038693e-05, + "loss": 0.7375, + "step": 9700 + }, + { + "epoch": 0.42951861239941114, + "grad_norm": 1.4497705698013306, + "learning_rate": 1.4083737226292836e-05, + "loss": 0.7829, + "step": 9701 + }, + { + "epoch": 0.429562888104225, + "grad_norm": 1.4125462770462036, + "learning_rate": 1.4082326467010238e-05, + "loss": 0.7004, + "step": 9702 + }, + { + "epoch": 0.4296071638090389, + "grad_norm": 1.3709765672683716, + "learning_rate": 1.4080915610224594e-05, + "loss": 0.6821, + "step": 9703 + }, + { + "epoch": 0.42965143951385276, + "grad_norm": 1.394547462463379, + "learning_rate": 1.4079504655969606e-05, + "loss": 0.7084, + "step": 9704 + }, + { + "epoch": 0.4296957152186666, + "grad_norm": 1.6634234189987183, + "learning_rate": 1.4078093604278972e-05, + "loss": 0.7418, + "step": 9705 + }, + { + "epoch": 0.42973999092348053, + "grad_norm": 1.4566121101379395, + "learning_rate": 1.4076682455186393e-05, + "loss": 0.7346, + "step": 9706 + }, + { + "epoch": 0.4297842666282944, + "grad_norm": 1.5383449792861938, + "learning_rate": 1.4075271208725572e-05, + "loss": 0.7514, + "step": 9707 + }, + { + "epoch": 0.42982854233310824, + "grad_norm": 1.4973310232162476, + "learning_rate": 1.4073859864930215e-05, + "loss": 0.7767, + "step": 9708 + }, + { + "epoch": 0.42987281803792216, + "grad_norm": 1.3729490041732788, + "learning_rate": 1.407244842383403e-05, + "loss": 0.6738, + "step": 9709 + }, + { + "epoch": 0.429917093742736, + "grad_norm": 1.2560805082321167, + "learning_rate": 1.4071036885470736e-05, + "loss": 0.7574, + "step": 9710 + }, + { + "epoch": 0.42996136944754987, + "grad_norm": 1.79560387134552, + "learning_rate": 1.406962524987404e-05, + "loss": 0.6707, + "step": 9711 + }, + { + "epoch": 0.4300056451523638, + "grad_norm": 1.422835111618042, + "learning_rate": 1.4068213517077655e-05, + "loss": 0.8131, + "step": 9712 + }, + { + "epoch": 0.43004992085717764, + "grad_norm": 1.295342206954956, + "learning_rate": 1.4066801687115305e-05, + "loss": 0.6549, + "step": 9713 + }, + { + "epoch": 0.43009419656199155, + "grad_norm": 1.4352304935455322, + "learning_rate": 1.4065389760020708e-05, + "loss": 0.7278, + "step": 9714 + }, + { + "epoch": 0.4301384722668054, + "grad_norm": 1.3998298645019531, + "learning_rate": 1.4063977735827586e-05, + "loss": 0.762, + "step": 9715 + }, + { + "epoch": 0.43018274797161926, + "grad_norm": 1.3399012088775635, + "learning_rate": 1.4062565614569665e-05, + "loss": 0.7074, + "step": 9716 + }, + { + "epoch": 0.4302270236764332, + "grad_norm": 1.4527186155319214, + "learning_rate": 1.4061153396280673e-05, + "loss": 0.7538, + "step": 9717 + }, + { + "epoch": 0.43027129938124703, + "grad_norm": 1.4188430309295654, + "learning_rate": 1.4059741080994336e-05, + "loss": 0.7349, + "step": 9718 + }, + { + "epoch": 0.4303155750860609, + "grad_norm": 1.4477918148040771, + "learning_rate": 1.4058328668744393e-05, + "loss": 0.7301, + "step": 9719 + }, + { + "epoch": 0.4303598507908748, + "grad_norm": 1.5079805850982666, + "learning_rate": 1.4056916159564571e-05, + "loss": 0.7507, + "step": 9720 + }, + { + "epoch": 0.43040412649568865, + "grad_norm": 1.2587199211120605, + "learning_rate": 1.405550355348861e-05, + "loss": 0.7433, + "step": 9721 + }, + { + "epoch": 0.4304484022005025, + "grad_norm": 1.255156397819519, + "learning_rate": 1.405409085055025e-05, + "loss": 0.6603, + "step": 9722 + }, + { + "epoch": 0.4304926779053164, + "grad_norm": 1.3418325185775757, + "learning_rate": 1.405267805078323e-05, + "loss": 0.73, + "step": 9723 + }, + { + "epoch": 0.4305369536101303, + "grad_norm": 1.5810160636901855, + "learning_rate": 1.4051265154221296e-05, + "loss": 0.7535, + "step": 9724 + }, + { + "epoch": 0.43058122931494414, + "grad_norm": 1.5550785064697266, + "learning_rate": 1.4049852160898192e-05, + "loss": 0.7291, + "step": 9725 + }, + { + "epoch": 0.43062550501975805, + "grad_norm": 1.3464924097061157, + "learning_rate": 1.4048439070847666e-05, + "loss": 0.7831, + "step": 9726 + }, + { + "epoch": 0.4306697807245719, + "grad_norm": 1.205530047416687, + "learning_rate": 1.4047025884103469e-05, + "loss": 0.6657, + "step": 9727 + }, + { + "epoch": 0.43071405642938576, + "grad_norm": 1.3458179235458374, + "learning_rate": 1.4045612600699352e-05, + "loss": 0.7377, + "step": 9728 + }, + { + "epoch": 0.4307583321341997, + "grad_norm": 1.449180245399475, + "learning_rate": 1.4044199220669077e-05, + "loss": 0.7688, + "step": 9729 + }, + { + "epoch": 0.43080260783901353, + "grad_norm": 1.1753172874450684, + "learning_rate": 1.4042785744046395e-05, + "loss": 0.6209, + "step": 9730 + }, + { + "epoch": 0.4308468835438274, + "grad_norm": 1.5089768171310425, + "learning_rate": 1.4041372170865064e-05, + "loss": 0.7654, + "step": 9731 + }, + { + "epoch": 0.4308911592486413, + "grad_norm": 1.5794011354446411, + "learning_rate": 1.4039958501158852e-05, + "loss": 0.786, + "step": 9732 + }, + { + "epoch": 0.43093543495345515, + "grad_norm": 1.5804603099822998, + "learning_rate": 1.4038544734961518e-05, + "loss": 0.7168, + "step": 9733 + }, + { + "epoch": 0.43097971065826907, + "grad_norm": 1.4222533702850342, + "learning_rate": 1.4037130872306835e-05, + "loss": 0.6947, + "step": 9734 + }, + { + "epoch": 0.4310239863630829, + "grad_norm": 1.3983219861984253, + "learning_rate": 1.4035716913228568e-05, + "loss": 0.7362, + "step": 9735 + }, + { + "epoch": 0.4310682620678968, + "grad_norm": 1.4760092496871948, + "learning_rate": 1.4034302857760485e-05, + "loss": 0.7601, + "step": 9736 + }, + { + "epoch": 0.4311125377727107, + "grad_norm": 1.900982141494751, + "learning_rate": 1.4032888705936367e-05, + "loss": 0.7362, + "step": 9737 + }, + { + "epoch": 0.43115681347752455, + "grad_norm": 1.3986819982528687, + "learning_rate": 1.4031474457789981e-05, + "loss": 0.7206, + "step": 9738 + }, + { + "epoch": 0.4312010891823384, + "grad_norm": 1.3857643604278564, + "learning_rate": 1.4030060113355118e-05, + "loss": 0.7726, + "step": 9739 + }, + { + "epoch": 0.4312453648871523, + "grad_norm": 1.4525339603424072, + "learning_rate": 1.4028645672665545e-05, + "loss": 0.7935, + "step": 9740 + }, + { + "epoch": 0.4312896405919662, + "grad_norm": 1.7094053030014038, + "learning_rate": 1.4027231135755051e-05, + "loss": 0.748, + "step": 9741 + }, + { + "epoch": 0.43133391629678003, + "grad_norm": 1.3289005756378174, + "learning_rate": 1.4025816502657419e-05, + "loss": 0.727, + "step": 9742 + }, + { + "epoch": 0.43137819200159394, + "grad_norm": 1.3409343957901, + "learning_rate": 1.402440177340644e-05, + "loss": 0.7547, + "step": 9743 + }, + { + "epoch": 0.4314224677064078, + "grad_norm": 1.8786797523498535, + "learning_rate": 1.4022986948035903e-05, + "loss": 0.7953, + "step": 9744 + }, + { + "epoch": 0.43146674341122165, + "grad_norm": 1.3319830894470215, + "learning_rate": 1.4021572026579596e-05, + "loss": 0.6674, + "step": 9745 + }, + { + "epoch": 0.43151101911603557, + "grad_norm": 1.4605257511138916, + "learning_rate": 1.4020157009071314e-05, + "loss": 0.7987, + "step": 9746 + }, + { + "epoch": 0.4315552948208494, + "grad_norm": 1.1734631061553955, + "learning_rate": 1.401874189554486e-05, + "loss": 0.6512, + "step": 9747 + }, + { + "epoch": 0.4315995705256633, + "grad_norm": 1.3413044214248657, + "learning_rate": 1.4017326686034025e-05, + "loss": 0.7076, + "step": 9748 + }, + { + "epoch": 0.4316438462304772, + "grad_norm": 1.3359653949737549, + "learning_rate": 1.4015911380572614e-05, + "loss": 0.7401, + "step": 9749 + }, + { + "epoch": 0.43168812193529105, + "grad_norm": 1.5246859788894653, + "learning_rate": 1.4014495979194426e-05, + "loss": 0.7513, + "step": 9750 + }, + { + "epoch": 0.43173239764010496, + "grad_norm": 1.5362645387649536, + "learning_rate": 1.4013080481933271e-05, + "loss": 0.6863, + "step": 9751 + }, + { + "epoch": 0.4317766733449188, + "grad_norm": 1.3982954025268555, + "learning_rate": 1.401166488882296e-05, + "loss": 0.6852, + "step": 9752 + }, + { + "epoch": 0.43182094904973267, + "grad_norm": 1.4437159299850464, + "learning_rate": 1.40102491998973e-05, + "loss": 0.7353, + "step": 9753 + }, + { + "epoch": 0.4318652247545466, + "grad_norm": 1.2097982168197632, + "learning_rate": 1.4008833415190102e-05, + "loss": 0.6426, + "step": 9754 + }, + { + "epoch": 0.43190950045936044, + "grad_norm": 1.4474126100540161, + "learning_rate": 1.400741753473518e-05, + "loss": 0.7546, + "step": 9755 + }, + { + "epoch": 0.4319537761641743, + "grad_norm": 1.5535904169082642, + "learning_rate": 1.4006001558566352e-05, + "loss": 0.6921, + "step": 9756 + }, + { + "epoch": 0.4319980518689882, + "grad_norm": 1.365431547164917, + "learning_rate": 1.4004585486717441e-05, + "loss": 0.7686, + "step": 9757 + }, + { + "epoch": 0.43204232757380207, + "grad_norm": 1.3718552589416504, + "learning_rate": 1.4003169319222268e-05, + "loss": 0.7985, + "step": 9758 + }, + { + "epoch": 0.4320866032786159, + "grad_norm": 1.7463123798370361, + "learning_rate": 1.4001753056114654e-05, + "loss": 0.7453, + "step": 9759 + }, + { + "epoch": 0.43213087898342983, + "grad_norm": 1.3753399848937988, + "learning_rate": 1.4000336697428423e-05, + "loss": 0.7248, + "step": 9760 + }, + { + "epoch": 0.4321751546882437, + "grad_norm": 1.5641309022903442, + "learning_rate": 1.3998920243197408e-05, + "loss": 0.7313, + "step": 9761 + }, + { + "epoch": 0.43221943039305755, + "grad_norm": 1.927849292755127, + "learning_rate": 1.3997503693455443e-05, + "loss": 0.7097, + "step": 9762 + }, + { + "epoch": 0.43226370609787146, + "grad_norm": 1.3966529369354248, + "learning_rate": 1.3996087048236357e-05, + "loss": 0.7089, + "step": 9763 + }, + { + "epoch": 0.4323079818026853, + "grad_norm": 1.3659945726394653, + "learning_rate": 1.3994670307573985e-05, + "loss": 0.6729, + "step": 9764 + }, + { + "epoch": 0.43235225750749917, + "grad_norm": 1.3366763591766357, + "learning_rate": 1.3993253471502164e-05, + "loss": 0.7404, + "step": 9765 + }, + { + "epoch": 0.4323965332123131, + "grad_norm": 1.4470751285552979, + "learning_rate": 1.3991836540054733e-05, + "loss": 0.7287, + "step": 9766 + }, + { + "epoch": 0.43244080891712694, + "grad_norm": 1.2287516593933105, + "learning_rate": 1.3990419513265541e-05, + "loss": 0.6278, + "step": 9767 + }, + { + "epoch": 0.43248508462194085, + "grad_norm": 1.5643130540847778, + "learning_rate": 1.3989002391168428e-05, + "loss": 0.7635, + "step": 9768 + }, + { + "epoch": 0.4325293603267547, + "grad_norm": 1.3064956665039062, + "learning_rate": 1.3987585173797238e-05, + "loss": 0.6763, + "step": 9769 + }, + { + "epoch": 0.43257363603156856, + "grad_norm": 1.6334642171859741, + "learning_rate": 1.3986167861185822e-05, + "loss": 0.7728, + "step": 9770 + }, + { + "epoch": 0.4326179117363825, + "grad_norm": 1.5583992004394531, + "learning_rate": 1.3984750453368033e-05, + "loss": 0.6953, + "step": 9771 + }, + { + "epoch": 0.43266218744119633, + "grad_norm": 1.3459376096725464, + "learning_rate": 1.3983332950377727e-05, + "loss": 0.6899, + "step": 9772 + }, + { + "epoch": 0.4327064631460102, + "grad_norm": 1.2932591438293457, + "learning_rate": 1.3981915352248755e-05, + "loss": 0.6785, + "step": 9773 + }, + { + "epoch": 0.4327507388508241, + "grad_norm": 1.2116252183914185, + "learning_rate": 1.3980497659014978e-05, + "loss": 0.6128, + "step": 9774 + }, + { + "epoch": 0.43279501455563796, + "grad_norm": 1.4354016780853271, + "learning_rate": 1.3979079870710253e-05, + "loss": 0.7409, + "step": 9775 + }, + { + "epoch": 0.4328392902604518, + "grad_norm": 1.3790243864059448, + "learning_rate": 1.3977661987368448e-05, + "loss": 0.7709, + "step": 9776 + }, + { + "epoch": 0.4328835659652657, + "grad_norm": 1.577452540397644, + "learning_rate": 1.3976244009023424e-05, + "loss": 0.7909, + "step": 9777 + }, + { + "epoch": 0.4329278416700796, + "grad_norm": 1.4040334224700928, + "learning_rate": 1.3974825935709047e-05, + "loss": 0.6676, + "step": 9778 + }, + { + "epoch": 0.43297211737489344, + "grad_norm": 1.4602333307266235, + "learning_rate": 1.397340776745919e-05, + "loss": 0.788, + "step": 9779 + }, + { + "epoch": 0.43301639307970735, + "grad_norm": 1.6131590604782104, + "learning_rate": 1.3971989504307726e-05, + "loss": 0.7103, + "step": 9780 + }, + { + "epoch": 0.4330606687845212, + "grad_norm": 1.692136287689209, + "learning_rate": 1.3970571146288525e-05, + "loss": 0.782, + "step": 9781 + }, + { + "epoch": 0.43310494448933506, + "grad_norm": 3.5146517753601074, + "learning_rate": 1.3969152693435466e-05, + "loss": 0.6773, + "step": 9782 + }, + { + "epoch": 0.433149220194149, + "grad_norm": 1.4159971475601196, + "learning_rate": 1.3967734145782425e-05, + "loss": 0.7544, + "step": 9783 + }, + { + "epoch": 0.43319349589896283, + "grad_norm": 2.613917350769043, + "learning_rate": 1.3966315503363288e-05, + "loss": 0.8014, + "step": 9784 + }, + { + "epoch": 0.43323777160377674, + "grad_norm": 1.636494517326355, + "learning_rate": 1.3964896766211933e-05, + "loss": 0.7811, + "step": 9785 + }, + { + "epoch": 0.4332820473085906, + "grad_norm": 1.7855547666549683, + "learning_rate": 1.3963477934362246e-05, + "loss": 0.7004, + "step": 9786 + }, + { + "epoch": 0.43332632301340446, + "grad_norm": 1.4465582370758057, + "learning_rate": 1.3962059007848119e-05, + "loss": 0.719, + "step": 9787 + }, + { + "epoch": 0.43337059871821837, + "grad_norm": 1.8871821165084839, + "learning_rate": 1.3960639986703432e-05, + "loss": 0.866, + "step": 9788 + }, + { + "epoch": 0.4334148744230322, + "grad_norm": 1.3541635274887085, + "learning_rate": 1.3959220870962092e-05, + "loss": 0.7587, + "step": 9789 + }, + { + "epoch": 0.4334591501278461, + "grad_norm": 1.2662566900253296, + "learning_rate": 1.3957801660657979e-05, + "loss": 0.7126, + "step": 9790 + }, + { + "epoch": 0.43350342583266, + "grad_norm": 1.3648115396499634, + "learning_rate": 1.3956382355824999e-05, + "loss": 0.6878, + "step": 9791 + }, + { + "epoch": 0.43354770153747385, + "grad_norm": 1.3429782390594482, + "learning_rate": 1.3954962956497047e-05, + "loss": 0.7152, + "step": 9792 + }, + { + "epoch": 0.4335919772422877, + "grad_norm": 5.3651533126831055, + "learning_rate": 1.3953543462708023e-05, + "loss": 0.6532, + "step": 9793 + }, + { + "epoch": 0.4336362529471016, + "grad_norm": 1.2303071022033691, + "learning_rate": 1.3952123874491837e-05, + "loss": 0.7307, + "step": 9794 + }, + { + "epoch": 0.4336805286519155, + "grad_norm": 1.50037682056427, + "learning_rate": 1.3950704191882388e-05, + "loss": 0.721, + "step": 9795 + }, + { + "epoch": 0.43372480435672933, + "grad_norm": 1.3243356943130493, + "learning_rate": 1.3949284414913583e-05, + "loss": 0.7237, + "step": 9796 + }, + { + "epoch": 0.43376908006154324, + "grad_norm": 1.3383588790893555, + "learning_rate": 1.3947864543619337e-05, + "loss": 0.7403, + "step": 9797 + }, + { + "epoch": 0.4338133557663571, + "grad_norm": 1.347542643547058, + "learning_rate": 1.3946444578033562e-05, + "loss": 0.765, + "step": 9798 + }, + { + "epoch": 0.43385763147117096, + "grad_norm": 1.2758816480636597, + "learning_rate": 1.3945024518190173e-05, + "loss": 0.6629, + "step": 9799 + }, + { + "epoch": 0.43390190717598487, + "grad_norm": 1.427477240562439, + "learning_rate": 1.3943604364123085e-05, + "loss": 0.7538, + "step": 9800 + }, + { + "epoch": 0.4339461828807987, + "grad_norm": 3.462354898452759, + "learning_rate": 1.3942184115866216e-05, + "loss": 0.7478, + "step": 9801 + }, + { + "epoch": 0.43399045858561264, + "grad_norm": 1.4133020639419556, + "learning_rate": 1.3940763773453491e-05, + "loss": 0.6518, + "step": 9802 + }, + { + "epoch": 0.4340347342904265, + "grad_norm": 1.7638473510742188, + "learning_rate": 1.3939343336918828e-05, + "loss": 0.7552, + "step": 9803 + }, + { + "epoch": 0.43407900999524035, + "grad_norm": 1.3082976341247559, + "learning_rate": 1.3937922806296162e-05, + "loss": 0.7355, + "step": 9804 + }, + { + "epoch": 0.43412328570005426, + "grad_norm": 1.4558058977127075, + "learning_rate": 1.3936502181619415e-05, + "loss": 0.6888, + "step": 9805 + }, + { + "epoch": 0.4341675614048681, + "grad_norm": 1.3029924631118774, + "learning_rate": 1.3935081462922518e-05, + "loss": 0.6552, + "step": 9806 + }, + { + "epoch": 0.434211837109682, + "grad_norm": 1.3260201215744019, + "learning_rate": 1.3933660650239405e-05, + "loss": 0.6479, + "step": 9807 + }, + { + "epoch": 0.4342561128144959, + "grad_norm": 1.3050265312194824, + "learning_rate": 1.3932239743604007e-05, + "loss": 0.6756, + "step": 9808 + }, + { + "epoch": 0.43430038851930974, + "grad_norm": 1.2999800443649292, + "learning_rate": 1.3930818743050268e-05, + "loss": 0.7177, + "step": 9809 + }, + { + "epoch": 0.4343446642241236, + "grad_norm": 1.4145445823669434, + "learning_rate": 1.3929397648612124e-05, + "loss": 0.7191, + "step": 9810 + }, + { + "epoch": 0.4343889399289375, + "grad_norm": 1.3038396835327148, + "learning_rate": 1.3927976460323513e-05, + "loss": 0.6791, + "step": 9811 + }, + { + "epoch": 0.43443321563375137, + "grad_norm": 1.3475815057754517, + "learning_rate": 1.3926555178218387e-05, + "loss": 0.7219, + "step": 9812 + }, + { + "epoch": 0.4344774913385652, + "grad_norm": 1.335325837135315, + "learning_rate": 1.3925133802330684e-05, + "loss": 0.6577, + "step": 9813 + }, + { + "epoch": 0.43452176704337914, + "grad_norm": 1.2424763441085815, + "learning_rate": 1.3923712332694357e-05, + "loss": 0.6963, + "step": 9814 + }, + { + "epoch": 0.434566042748193, + "grad_norm": 1.960193395614624, + "learning_rate": 1.3922290769343358e-05, + "loss": 0.8252, + "step": 9815 + }, + { + "epoch": 0.43461031845300685, + "grad_norm": 1.338397741317749, + "learning_rate": 1.3920869112311635e-05, + "loss": 0.6578, + "step": 9816 + }, + { + "epoch": 0.43465459415782076, + "grad_norm": 1.2871264219284058, + "learning_rate": 1.3919447361633147e-05, + "loss": 0.7424, + "step": 9817 + }, + { + "epoch": 0.4346988698626346, + "grad_norm": 1.4953123331069946, + "learning_rate": 1.3918025517341849e-05, + "loss": 0.8438, + "step": 9818 + }, + { + "epoch": 0.4347431455674485, + "grad_norm": 1.38509202003479, + "learning_rate": 1.3916603579471705e-05, + "loss": 0.7737, + "step": 9819 + }, + { + "epoch": 0.4347874212722624, + "grad_norm": 2.1076581478118896, + "learning_rate": 1.3915181548056671e-05, + "loss": 0.766, + "step": 9820 + }, + { + "epoch": 0.43483169697707624, + "grad_norm": 1.3969824314117432, + "learning_rate": 1.3913759423130711e-05, + "loss": 0.7277, + "step": 9821 + }, + { + "epoch": 0.43487597268189015, + "grad_norm": 1.4517912864685059, + "learning_rate": 1.3912337204727797e-05, + "loss": 0.7543, + "step": 9822 + }, + { + "epoch": 0.434920248386704, + "grad_norm": 1.5068165063858032, + "learning_rate": 1.3910914892881895e-05, + "loss": 0.7624, + "step": 9823 + }, + { + "epoch": 0.43496452409151787, + "grad_norm": 1.5362451076507568, + "learning_rate": 1.3909492487626977e-05, + "loss": 0.6933, + "step": 9824 + }, + { + "epoch": 0.4350087997963318, + "grad_norm": 1.4841383695602417, + "learning_rate": 1.3908069988997008e-05, + "loss": 0.6748, + "step": 9825 + }, + { + "epoch": 0.43505307550114564, + "grad_norm": 1.4825894832611084, + "learning_rate": 1.3906647397025972e-05, + "loss": 0.6829, + "step": 9826 + }, + { + "epoch": 0.4350973512059595, + "grad_norm": 3.631559371948242, + "learning_rate": 1.3905224711747844e-05, + "loss": 0.8125, + "step": 9827 + }, + { + "epoch": 0.4351416269107734, + "grad_norm": 2.116194248199463, + "learning_rate": 1.3903801933196606e-05, + "loss": 0.797, + "step": 9828 + }, + { + "epoch": 0.43518590261558726, + "grad_norm": 1.6091033220291138, + "learning_rate": 1.3902379061406235e-05, + "loss": 0.7954, + "step": 9829 + }, + { + "epoch": 0.4352301783204011, + "grad_norm": 1.6925642490386963, + "learning_rate": 1.3900956096410719e-05, + "loss": 0.7437, + "step": 9830 + }, + { + "epoch": 0.43527445402521503, + "grad_norm": 1.611742377281189, + "learning_rate": 1.3899533038244042e-05, + "loss": 0.7824, + "step": 9831 + }, + { + "epoch": 0.4353187297300289, + "grad_norm": 1.5524632930755615, + "learning_rate": 1.3898109886940193e-05, + "loss": 0.6632, + "step": 9832 + }, + { + "epoch": 0.43536300543484274, + "grad_norm": 1.234873652458191, + "learning_rate": 1.3896686642533163e-05, + "loss": 0.6791, + "step": 9833 + }, + { + "epoch": 0.43540728113965665, + "grad_norm": 1.3446931838989258, + "learning_rate": 1.3895263305056947e-05, + "loss": 0.7016, + "step": 9834 + }, + { + "epoch": 0.4354515568444705, + "grad_norm": 1.5209194421768188, + "learning_rate": 1.3893839874545535e-05, + "loss": 0.7813, + "step": 9835 + }, + { + "epoch": 0.43549583254928437, + "grad_norm": 1.9487038850784302, + "learning_rate": 1.3892416351032933e-05, + "loss": 0.7401, + "step": 9836 + }, + { + "epoch": 0.4355401082540983, + "grad_norm": 1.645630121231079, + "learning_rate": 1.3890992734553131e-05, + "loss": 0.8311, + "step": 9837 + }, + { + "epoch": 0.43558438395891214, + "grad_norm": 1.1897313594818115, + "learning_rate": 1.388956902514014e-05, + "loss": 0.6768, + "step": 9838 + }, + { + "epoch": 0.43562865966372605, + "grad_norm": 1.155145525932312, + "learning_rate": 1.3888145222827958e-05, + "loss": 0.6322, + "step": 9839 + }, + { + "epoch": 0.4356729353685399, + "grad_norm": 1.3338656425476074, + "learning_rate": 1.3886721327650592e-05, + "loss": 0.7872, + "step": 9840 + }, + { + "epoch": 0.43571721107335376, + "grad_norm": 1.3287408351898193, + "learning_rate": 1.3885297339642054e-05, + "loss": 0.7023, + "step": 9841 + }, + { + "epoch": 0.4357614867781677, + "grad_norm": 1.320206642150879, + "learning_rate": 1.388387325883635e-05, + "loss": 0.7354, + "step": 9842 + }, + { + "epoch": 0.43580576248298153, + "grad_norm": 1.2607417106628418, + "learning_rate": 1.3882449085267497e-05, + "loss": 0.712, + "step": 9843 + }, + { + "epoch": 0.4358500381877954, + "grad_norm": 1.32608163356781, + "learning_rate": 1.3881024818969511e-05, + "loss": 0.6879, + "step": 9844 + }, + { + "epoch": 0.4358943138926093, + "grad_norm": 1.5568485260009766, + "learning_rate": 1.3879600459976404e-05, + "loss": 0.7883, + "step": 9845 + }, + { + "epoch": 0.43593858959742315, + "grad_norm": 1.2779927253723145, + "learning_rate": 1.3878176008322202e-05, + "loss": 0.6767, + "step": 9846 + }, + { + "epoch": 0.435982865302237, + "grad_norm": 1.4009371995925903, + "learning_rate": 1.3876751464040924e-05, + "loss": 0.7042, + "step": 9847 + }, + { + "epoch": 0.4360271410070509, + "grad_norm": 1.8739750385284424, + "learning_rate": 1.387532682716659e-05, + "loss": 0.7612, + "step": 9848 + }, + { + "epoch": 0.4360714167118648, + "grad_norm": 1.3514505624771118, + "learning_rate": 1.3873902097733231e-05, + "loss": 0.6852, + "step": 9849 + }, + { + "epoch": 0.43611569241667864, + "grad_norm": 1.2939900159835815, + "learning_rate": 1.3872477275774875e-05, + "loss": 0.7041, + "step": 9850 + }, + { + "epoch": 0.43615996812149255, + "grad_norm": 1.680558681488037, + "learning_rate": 1.3871052361325552e-05, + "loss": 0.7638, + "step": 9851 + }, + { + "epoch": 0.4362042438263064, + "grad_norm": 1.5645381212234497, + "learning_rate": 1.3869627354419297e-05, + "loss": 0.6961, + "step": 9852 + }, + { + "epoch": 0.43624851953112026, + "grad_norm": 1.4880088567733765, + "learning_rate": 1.3868202255090141e-05, + "loss": 0.7094, + "step": 9853 + }, + { + "epoch": 0.43629279523593417, + "grad_norm": 1.6548792123794556, + "learning_rate": 1.3866777063372128e-05, + "loss": 0.7553, + "step": 9854 + }, + { + "epoch": 0.43633707094074803, + "grad_norm": 1.3749639987945557, + "learning_rate": 1.386535177929929e-05, + "loss": 0.716, + "step": 9855 + }, + { + "epoch": 0.43638134664556194, + "grad_norm": 1.4025352001190186, + "learning_rate": 1.3863926402905672e-05, + "loss": 0.6973, + "step": 9856 + }, + { + "epoch": 0.4364256223503758, + "grad_norm": 1.2883710861206055, + "learning_rate": 1.3862500934225321e-05, + "loss": 0.6687, + "step": 9857 + }, + { + "epoch": 0.43646989805518965, + "grad_norm": 1.4540512561798096, + "learning_rate": 1.3861075373292273e-05, + "loss": 0.7082, + "step": 9858 + }, + { + "epoch": 0.43651417376000357, + "grad_norm": 1.720098853111267, + "learning_rate": 1.385964972014059e-05, + "loss": 0.7335, + "step": 9859 + }, + { + "epoch": 0.4365584494648174, + "grad_norm": 1.4924458265304565, + "learning_rate": 1.3858223974804315e-05, + "loss": 0.7277, + "step": 9860 + }, + { + "epoch": 0.4366027251696313, + "grad_norm": 1.4335683584213257, + "learning_rate": 1.3856798137317502e-05, + "loss": 0.7197, + "step": 9861 + }, + { + "epoch": 0.4366470008744452, + "grad_norm": 1.6103994846343994, + "learning_rate": 1.3855372207714205e-05, + "loss": 0.7294, + "step": 9862 + }, + { + "epoch": 0.43669127657925905, + "grad_norm": 1.5455901622772217, + "learning_rate": 1.385394618602848e-05, + "loss": 0.7432, + "step": 9863 + }, + { + "epoch": 0.4367355522840729, + "grad_norm": 1.6165471076965332, + "learning_rate": 1.385252007229439e-05, + "loss": 0.7771, + "step": 9864 + }, + { + "epoch": 0.4367798279888868, + "grad_norm": 1.1926532983779907, + "learning_rate": 1.3851093866545996e-05, + "loss": 0.6964, + "step": 9865 + }, + { + "epoch": 0.43682410369370067, + "grad_norm": 1.6635220050811768, + "learning_rate": 1.3849667568817363e-05, + "loss": 0.7749, + "step": 9866 + }, + { + "epoch": 0.43686837939851453, + "grad_norm": 1.5594998598098755, + "learning_rate": 1.384824117914255e-05, + "loss": 0.7749, + "step": 9867 + }, + { + "epoch": 0.43691265510332844, + "grad_norm": 1.4303371906280518, + "learning_rate": 1.3846814697555631e-05, + "loss": 0.7066, + "step": 9868 + }, + { + "epoch": 0.4369569308081423, + "grad_norm": 1.7678449153900146, + "learning_rate": 1.3845388124090676e-05, + "loss": 0.7007, + "step": 9869 + }, + { + "epoch": 0.43700120651295615, + "grad_norm": 1.3626314401626587, + "learning_rate": 1.3843961458781753e-05, + "loss": 0.7446, + "step": 9870 + }, + { + "epoch": 0.43704548221777006, + "grad_norm": 1.5935068130493164, + "learning_rate": 1.3842534701662947e-05, + "loss": 0.7097, + "step": 9871 + }, + { + "epoch": 0.4370897579225839, + "grad_norm": 1.7448647022247314, + "learning_rate": 1.3841107852768324e-05, + "loss": 0.7903, + "step": 9872 + }, + { + "epoch": 0.43713403362739783, + "grad_norm": 1.5964354276657104, + "learning_rate": 1.3839680912131967e-05, + "loss": 0.7955, + "step": 9873 + }, + { + "epoch": 0.4371783093322117, + "grad_norm": 1.6173633337020874, + "learning_rate": 1.383825387978796e-05, + "loss": 0.6664, + "step": 9874 + }, + { + "epoch": 0.43722258503702555, + "grad_norm": 1.4115121364593506, + "learning_rate": 1.3836826755770386e-05, + "loss": 0.6714, + "step": 9875 + }, + { + "epoch": 0.43726686074183946, + "grad_norm": 1.302412509918213, + "learning_rate": 1.3835399540113326e-05, + "loss": 0.7358, + "step": 9876 + }, + { + "epoch": 0.4373111364466533, + "grad_norm": 1.6174399852752686, + "learning_rate": 1.383397223285087e-05, + "loss": 0.7585, + "step": 9877 + }, + { + "epoch": 0.43735541215146717, + "grad_norm": 1.3511711359024048, + "learning_rate": 1.3832544834017109e-05, + "loss": 0.7637, + "step": 9878 + }, + { + "epoch": 0.4373996878562811, + "grad_norm": 1.392557978630066, + "learning_rate": 1.3831117343646136e-05, + "loss": 0.6929, + "step": 9879 + }, + { + "epoch": 0.43744396356109494, + "grad_norm": 1.6874220371246338, + "learning_rate": 1.3829689761772046e-05, + "loss": 0.7927, + "step": 9880 + }, + { + "epoch": 0.4374882392659088, + "grad_norm": 1.3093600273132324, + "learning_rate": 1.3828262088428936e-05, + "loss": 0.664, + "step": 9881 + }, + { + "epoch": 0.4375325149707227, + "grad_norm": 1.6946275234222412, + "learning_rate": 1.3826834323650899e-05, + "loss": 0.7837, + "step": 9882 + }, + { + "epoch": 0.43757679067553656, + "grad_norm": 1.6584004163742065, + "learning_rate": 1.382540646747204e-05, + "loss": 0.7972, + "step": 9883 + }, + { + "epoch": 0.4376210663803504, + "grad_norm": 1.1658457517623901, + "learning_rate": 1.3823978519926465e-05, + "loss": 0.5989, + "step": 9884 + }, + { + "epoch": 0.43766534208516433, + "grad_norm": 1.430213451385498, + "learning_rate": 1.3822550481048276e-05, + "loss": 0.7295, + "step": 9885 + }, + { + "epoch": 0.4377096177899782, + "grad_norm": 2.2485432624816895, + "learning_rate": 1.3821122350871583e-05, + "loss": 0.7246, + "step": 9886 + }, + { + "epoch": 0.43775389349479205, + "grad_norm": 1.3215746879577637, + "learning_rate": 1.381969412943049e-05, + "loss": 0.6738, + "step": 9887 + }, + { + "epoch": 0.43779816919960596, + "grad_norm": 1.3432481288909912, + "learning_rate": 1.3818265816759113e-05, + "loss": 0.7003, + "step": 9888 + }, + { + "epoch": 0.4378424449044198, + "grad_norm": 1.3665817975997925, + "learning_rate": 1.3816837412891568e-05, + "loss": 0.6944, + "step": 9889 + }, + { + "epoch": 0.43788672060923367, + "grad_norm": 1.2391668558120728, + "learning_rate": 1.381540891786197e-05, + "loss": 0.7551, + "step": 9890 + }, + { + "epoch": 0.4379309963140476, + "grad_norm": 1.312087893486023, + "learning_rate": 1.3813980331704437e-05, + "loss": 0.7109, + "step": 9891 + }, + { + "epoch": 0.43797527201886144, + "grad_norm": 1.9278348684310913, + "learning_rate": 1.3812551654453086e-05, + "loss": 0.6563, + "step": 9892 + }, + { + "epoch": 0.43801954772367535, + "grad_norm": 1.9719737768173218, + "learning_rate": 1.3811122886142043e-05, + "loss": 0.7465, + "step": 9893 + }, + { + "epoch": 0.4380638234284892, + "grad_norm": 1.7661559581756592, + "learning_rate": 1.3809694026805438e-05, + "loss": 0.7304, + "step": 9894 + }, + { + "epoch": 0.43810809913330306, + "grad_norm": 1.334453821182251, + "learning_rate": 1.3808265076477389e-05, + "loss": 0.733, + "step": 9895 + }, + { + "epoch": 0.438152374838117, + "grad_norm": 1.6918495893478394, + "learning_rate": 1.380683603519203e-05, + "loss": 0.7689, + "step": 9896 + }, + { + "epoch": 0.43819665054293083, + "grad_norm": 1.3341208696365356, + "learning_rate": 1.3805406902983491e-05, + "loss": 0.8122, + "step": 9897 + }, + { + "epoch": 0.4382409262477447, + "grad_norm": 1.4752894639968872, + "learning_rate": 1.3803977679885907e-05, + "loss": 0.7356, + "step": 9898 + }, + { + "epoch": 0.4382852019525586, + "grad_norm": 1.4176545143127441, + "learning_rate": 1.3802548365933417e-05, + "loss": 0.7538, + "step": 9899 + }, + { + "epoch": 0.43832947765737246, + "grad_norm": 1.4382671117782593, + "learning_rate": 1.3801118961160152e-05, + "loss": 0.7278, + "step": 9900 + }, + { + "epoch": 0.4383737533621863, + "grad_norm": 1.5472773313522339, + "learning_rate": 1.3799689465600256e-05, + "loss": 0.7199, + "step": 9901 + }, + { + "epoch": 0.4384180290670002, + "grad_norm": 1.1604896783828735, + "learning_rate": 1.379825987928787e-05, + "loss": 0.7177, + "step": 9902 + }, + { + "epoch": 0.4384623047718141, + "grad_norm": 1.5765975713729858, + "learning_rate": 1.3796830202257141e-05, + "loss": 0.7378, + "step": 9903 + }, + { + "epoch": 0.43850658047662794, + "grad_norm": 1.3035142421722412, + "learning_rate": 1.3795400434542213e-05, + "loss": 0.6881, + "step": 9904 + }, + { + "epoch": 0.43855085618144185, + "grad_norm": 3.4389281272888184, + "learning_rate": 1.3793970576177235e-05, + "loss": 0.7415, + "step": 9905 + }, + { + "epoch": 0.4385951318862557, + "grad_norm": 1.2576414346694946, + "learning_rate": 1.3792540627196362e-05, + "loss": 0.6751, + "step": 9906 + }, + { + "epoch": 0.43863940759106956, + "grad_norm": 1.5349948406219482, + "learning_rate": 1.3791110587633742e-05, + "loss": 0.7289, + "step": 9907 + }, + { + "epoch": 0.4386836832958835, + "grad_norm": 1.500684142112732, + "learning_rate": 1.3789680457523535e-05, + "loss": 0.7186, + "step": 9908 + }, + { + "epoch": 0.43872795900069733, + "grad_norm": 1.3421380519866943, + "learning_rate": 1.3788250236899894e-05, + "loss": 0.6912, + "step": 9909 + }, + { + "epoch": 0.43877223470551124, + "grad_norm": 1.5805822610855103, + "learning_rate": 1.378681992579698e-05, + "loss": 0.7669, + "step": 9910 + }, + { + "epoch": 0.4388165104103251, + "grad_norm": 1.154802918434143, + "learning_rate": 1.3785389524248957e-05, + "loss": 0.6032, + "step": 9911 + }, + { + "epoch": 0.43886078611513896, + "grad_norm": 1.2942637205123901, + "learning_rate": 1.3783959032289987e-05, + "loss": 0.7141, + "step": 9912 + }, + { + "epoch": 0.43890506181995287, + "grad_norm": 1.433035135269165, + "learning_rate": 1.3782528449954236e-05, + "loss": 0.6917, + "step": 9913 + }, + { + "epoch": 0.4389493375247667, + "grad_norm": 1.386824131011963, + "learning_rate": 1.3781097777275872e-05, + "loss": 0.723, + "step": 9914 + }, + { + "epoch": 0.4389936132295806, + "grad_norm": 1.285420536994934, + "learning_rate": 1.3779667014289067e-05, + "loss": 0.6729, + "step": 9915 + }, + { + "epoch": 0.4390378889343945, + "grad_norm": 1.7916226387023926, + "learning_rate": 1.3778236161027993e-05, + "loss": 0.739, + "step": 9916 + }, + { + "epoch": 0.43908216463920835, + "grad_norm": 1.5950961112976074, + "learning_rate": 1.3776805217526823e-05, + "loss": 0.7948, + "step": 9917 + }, + { + "epoch": 0.4391264403440222, + "grad_norm": 1.5409173965454102, + "learning_rate": 1.3775374183819738e-05, + "loss": 0.7719, + "step": 9918 + }, + { + "epoch": 0.4391707160488361, + "grad_norm": 1.3304952383041382, + "learning_rate": 1.3773943059940915e-05, + "loss": 0.7002, + "step": 9919 + }, + { + "epoch": 0.43921499175365, + "grad_norm": 1.4970966577529907, + "learning_rate": 1.3772511845924533e-05, + "loss": 0.7261, + "step": 9920 + }, + { + "epoch": 0.43925926745846383, + "grad_norm": 1.3294650316238403, + "learning_rate": 1.3771080541804778e-05, + "loss": 0.7089, + "step": 9921 + }, + { + "epoch": 0.43930354316327774, + "grad_norm": 1.3823987245559692, + "learning_rate": 1.3769649147615835e-05, + "loss": 0.709, + "step": 9922 + }, + { + "epoch": 0.4393478188680916, + "grad_norm": 1.4054893255233765, + "learning_rate": 1.3768217663391892e-05, + "loss": 0.7359, + "step": 9923 + }, + { + "epoch": 0.43939209457290546, + "grad_norm": 1.6701395511627197, + "learning_rate": 1.3766786089167136e-05, + "loss": 0.8136, + "step": 9924 + }, + { + "epoch": 0.43943637027771937, + "grad_norm": 1.2462406158447266, + "learning_rate": 1.3765354424975764e-05, + "loss": 0.6953, + "step": 9925 + }, + { + "epoch": 0.4394806459825332, + "grad_norm": 1.5989874601364136, + "learning_rate": 1.3763922670851965e-05, + "loss": 0.8034, + "step": 9926 + }, + { + "epoch": 0.43952492168734714, + "grad_norm": 1.5594005584716797, + "learning_rate": 1.3762490826829941e-05, + "loss": 0.7816, + "step": 9927 + }, + { + "epoch": 0.439569197392161, + "grad_norm": 1.3075799942016602, + "learning_rate": 1.3761058892943884e-05, + "loss": 0.6461, + "step": 9928 + }, + { + "epoch": 0.43961347309697485, + "grad_norm": 1.562628149986267, + "learning_rate": 1.3759626869228e-05, + "loss": 0.7203, + "step": 9929 + }, + { + "epoch": 0.43965774880178876, + "grad_norm": 1.4121731519699097, + "learning_rate": 1.3758194755716487e-05, + "loss": 0.7749, + "step": 9930 + }, + { + "epoch": 0.4397020245066026, + "grad_norm": 1.662299633026123, + "learning_rate": 1.3756762552443555e-05, + "loss": 0.8605, + "step": 9931 + }, + { + "epoch": 0.4397463002114165, + "grad_norm": 1.3809162378311157, + "learning_rate": 1.3755330259443407e-05, + "loss": 0.7587, + "step": 9932 + }, + { + "epoch": 0.4397905759162304, + "grad_norm": 1.1623356342315674, + "learning_rate": 1.375389787675025e-05, + "loss": 0.6941, + "step": 9933 + }, + { + "epoch": 0.43983485162104424, + "grad_norm": 1.782736897468567, + "learning_rate": 1.3752465404398305e-05, + "loss": 0.7294, + "step": 9934 + }, + { + "epoch": 0.4398791273258581, + "grad_norm": 1.4798212051391602, + "learning_rate": 1.3751032842421775e-05, + "loss": 0.7208, + "step": 9935 + }, + { + "epoch": 0.439923403030672, + "grad_norm": 1.3665083646774292, + "learning_rate": 1.3749600190854881e-05, + "loss": 0.7179, + "step": 9936 + }, + { + "epoch": 0.43996767873548587, + "grad_norm": 1.706634521484375, + "learning_rate": 1.3748167449731841e-05, + "loss": 0.7134, + "step": 9937 + }, + { + "epoch": 0.4400119544402997, + "grad_norm": 1.4324419498443604, + "learning_rate": 1.3746734619086872e-05, + "loss": 0.736, + "step": 9938 + }, + { + "epoch": 0.44005623014511364, + "grad_norm": 1.3085771799087524, + "learning_rate": 1.3745301698954197e-05, + "loss": 0.6775, + "step": 9939 + }, + { + "epoch": 0.4401005058499275, + "grad_norm": 2.003185272216797, + "learning_rate": 1.3743868689368038e-05, + "loss": 0.7891, + "step": 9940 + }, + { + "epoch": 0.44014478155474135, + "grad_norm": 1.3773577213287354, + "learning_rate": 1.3742435590362628e-05, + "loss": 0.709, + "step": 9941 + }, + { + "epoch": 0.44018905725955526, + "grad_norm": 1.5482213497161865, + "learning_rate": 1.374100240197219e-05, + "loss": 0.737, + "step": 9942 + }, + { + "epoch": 0.4402333329643691, + "grad_norm": 1.5929497480392456, + "learning_rate": 1.3739569124230956e-05, + "loss": 0.7232, + "step": 9943 + }, + { + "epoch": 0.44027760866918303, + "grad_norm": 1.3740171194076538, + "learning_rate": 1.3738135757173157e-05, + "loss": 0.7368, + "step": 9944 + }, + { + "epoch": 0.4403218843739969, + "grad_norm": 1.6254225969314575, + "learning_rate": 1.3736702300833027e-05, + "loss": 0.7954, + "step": 9945 + }, + { + "epoch": 0.44036616007881074, + "grad_norm": 1.2710915803909302, + "learning_rate": 1.3735268755244811e-05, + "loss": 0.799, + "step": 9946 + }, + { + "epoch": 0.44041043578362465, + "grad_norm": 1.3862805366516113, + "learning_rate": 1.3733835120442738e-05, + "loss": 0.8003, + "step": 9947 + }, + { + "epoch": 0.4404547114884385, + "grad_norm": 1.2927395105361938, + "learning_rate": 1.3732401396461054e-05, + "loss": 0.6487, + "step": 9948 + }, + { + "epoch": 0.44049898719325237, + "grad_norm": 1.5568206310272217, + "learning_rate": 1.3730967583334e-05, + "loss": 0.7458, + "step": 9949 + }, + { + "epoch": 0.4405432628980663, + "grad_norm": 1.3401846885681152, + "learning_rate": 1.3729533681095828e-05, + "loss": 0.6813, + "step": 9950 + }, + { + "epoch": 0.44058753860288014, + "grad_norm": 1.1844408512115479, + "learning_rate": 1.3728099689780778e-05, + "loss": 0.6599, + "step": 9951 + }, + { + "epoch": 0.440631814307694, + "grad_norm": 1.992436170578003, + "learning_rate": 1.37266656094231e-05, + "loss": 0.8518, + "step": 9952 + }, + { + "epoch": 0.4406760900125079, + "grad_norm": 1.7516257762908936, + "learning_rate": 1.3725231440057052e-05, + "loss": 0.7844, + "step": 9953 + }, + { + "epoch": 0.44072036571732176, + "grad_norm": 1.7451508045196533, + "learning_rate": 1.3723797181716881e-05, + "loss": 0.6551, + "step": 9954 + }, + { + "epoch": 0.4407646414221356, + "grad_norm": 1.4150609970092773, + "learning_rate": 1.372236283443685e-05, + "loss": 0.7751, + "step": 9955 + }, + { + "epoch": 0.44080891712694953, + "grad_norm": 1.7081245183944702, + "learning_rate": 1.3720928398251211e-05, + "loss": 0.8669, + "step": 9956 + }, + { + "epoch": 0.4408531928317634, + "grad_norm": 1.3312033414840698, + "learning_rate": 1.3719493873194228e-05, + "loss": 0.6825, + "step": 9957 + }, + { + "epoch": 0.44089746853657724, + "grad_norm": 1.3397457599639893, + "learning_rate": 1.3718059259300161e-05, + "loss": 0.7028, + "step": 9958 + }, + { + "epoch": 0.44094174424139115, + "grad_norm": 1.689238429069519, + "learning_rate": 1.3716624556603275e-05, + "loss": 0.8091, + "step": 9959 + }, + { + "epoch": 0.440986019946205, + "grad_norm": 1.6183098554611206, + "learning_rate": 1.3715189765137842e-05, + "loss": 0.7241, + "step": 9960 + }, + { + "epoch": 0.4410302956510189, + "grad_norm": 1.6283270120620728, + "learning_rate": 1.3713754884938124e-05, + "loss": 0.809, + "step": 9961 + }, + { + "epoch": 0.4410745713558328, + "grad_norm": 1.3671597242355347, + "learning_rate": 1.3712319916038395e-05, + "loss": 0.6936, + "step": 9962 + }, + { + "epoch": 0.44111884706064663, + "grad_norm": 1.8856709003448486, + "learning_rate": 1.3710884858472926e-05, + "loss": 0.7119, + "step": 9963 + }, + { + "epoch": 0.44116312276546055, + "grad_norm": 1.8435850143432617, + "learning_rate": 1.3709449712275992e-05, + "loss": 0.7805, + "step": 9964 + }, + { + "epoch": 0.4412073984702744, + "grad_norm": 1.388100028038025, + "learning_rate": 1.3708014477481878e-05, + "loss": 0.7314, + "step": 9965 + }, + { + "epoch": 0.44125167417508826, + "grad_norm": 1.2354092597961426, + "learning_rate": 1.3706579154124852e-05, + "loss": 0.6376, + "step": 9966 + }, + { + "epoch": 0.44129594987990217, + "grad_norm": 1.6174732446670532, + "learning_rate": 1.3705143742239205e-05, + "loss": 0.737, + "step": 9967 + }, + { + "epoch": 0.44134022558471603, + "grad_norm": 1.5451523065567017, + "learning_rate": 1.3703708241859212e-05, + "loss": 0.6686, + "step": 9968 + }, + { + "epoch": 0.4413845012895299, + "grad_norm": 1.5710676908493042, + "learning_rate": 1.3702272653019164e-05, + "loss": 0.7226, + "step": 9969 + }, + { + "epoch": 0.4414287769943438, + "grad_norm": 1.3064525127410889, + "learning_rate": 1.370083697575335e-05, + "loss": 0.6108, + "step": 9970 + }, + { + "epoch": 0.44147305269915765, + "grad_norm": 1.2563313245773315, + "learning_rate": 1.3699401210096062e-05, + "loss": 0.6838, + "step": 9971 + }, + { + "epoch": 0.4415173284039715, + "grad_norm": 1.3848857879638672, + "learning_rate": 1.3697965356081581e-05, + "loss": 0.7305, + "step": 9972 + }, + { + "epoch": 0.4415616041087854, + "grad_norm": 1.3132909536361694, + "learning_rate": 1.3696529413744213e-05, + "loss": 0.701, + "step": 9973 + }, + { + "epoch": 0.4416058798135993, + "grad_norm": 1.6376891136169434, + "learning_rate": 1.369509338311825e-05, + "loss": 0.7656, + "step": 9974 + }, + { + "epoch": 0.44165015551841313, + "grad_norm": 1.514769434928894, + "learning_rate": 1.3693657264237987e-05, + "loss": 0.7025, + "step": 9975 + }, + { + "epoch": 0.44169443122322705, + "grad_norm": 1.8584805727005005, + "learning_rate": 1.369222105713773e-05, + "loss": 0.6987, + "step": 9976 + }, + { + "epoch": 0.4417387069280409, + "grad_norm": 1.4349150657653809, + "learning_rate": 1.369078476185178e-05, + "loss": 0.7608, + "step": 9977 + }, + { + "epoch": 0.44178298263285476, + "grad_norm": 1.6952760219573975, + "learning_rate": 1.3689348378414438e-05, + "loss": 0.8092, + "step": 9978 + }, + { + "epoch": 0.44182725833766867, + "grad_norm": 1.43545663356781, + "learning_rate": 1.3687911906860016e-05, + "loss": 0.7054, + "step": 9979 + }, + { + "epoch": 0.4418715340424825, + "grad_norm": 1.3325519561767578, + "learning_rate": 1.368647534722282e-05, + "loss": 0.7559, + "step": 9980 + }, + { + "epoch": 0.44191580974729644, + "grad_norm": 2.045051097869873, + "learning_rate": 1.3685038699537165e-05, + "loss": 0.8676, + "step": 9981 + }, + { + "epoch": 0.4419600854521103, + "grad_norm": 1.6868027448654175, + "learning_rate": 1.3683601963837356e-05, + "loss": 0.7003, + "step": 9982 + }, + { + "epoch": 0.44200436115692415, + "grad_norm": 1.4938615560531616, + "learning_rate": 1.3682165140157719e-05, + "loss": 0.7289, + "step": 9983 + }, + { + "epoch": 0.44204863686173806, + "grad_norm": 1.6501798629760742, + "learning_rate": 1.3680728228532565e-05, + "loss": 0.8794, + "step": 9984 + }, + { + "epoch": 0.4420929125665519, + "grad_norm": 1.2935532331466675, + "learning_rate": 1.367929122899621e-05, + "loss": 0.7229, + "step": 9985 + }, + { + "epoch": 0.4421371882713658, + "grad_norm": 1.5729010105133057, + "learning_rate": 1.3677854141582983e-05, + "loss": 0.7567, + "step": 9986 + }, + { + "epoch": 0.4421814639761797, + "grad_norm": 1.277747631072998, + "learning_rate": 1.3676416966327201e-05, + "loss": 0.7274, + "step": 9987 + }, + { + "epoch": 0.44222573968099355, + "grad_norm": 1.6903001070022583, + "learning_rate": 1.3674979703263195e-05, + "loss": 0.8128, + "step": 9988 + }, + { + "epoch": 0.4422700153858074, + "grad_norm": 1.1063705682754517, + "learning_rate": 1.3673542352425295e-05, + "loss": 0.6616, + "step": 9989 + }, + { + "epoch": 0.4423142910906213, + "grad_norm": 1.9163310527801514, + "learning_rate": 1.3672104913847823e-05, + "loss": 0.7688, + "step": 9990 + }, + { + "epoch": 0.44235856679543517, + "grad_norm": 1.37859046459198, + "learning_rate": 1.3670667387565115e-05, + "loss": 0.7485, + "step": 9991 + }, + { + "epoch": 0.442402842500249, + "grad_norm": 1.3694748878479004, + "learning_rate": 1.3669229773611505e-05, + "loss": 0.7055, + "step": 9992 + }, + { + "epoch": 0.44244711820506294, + "grad_norm": 1.1577401161193848, + "learning_rate": 1.3667792072021336e-05, + "loss": 0.6244, + "step": 9993 + }, + { + "epoch": 0.4424913939098768, + "grad_norm": 1.3260235786437988, + "learning_rate": 1.366635428282893e-05, + "loss": 0.7128, + "step": 9994 + }, + { + "epoch": 0.44253566961469065, + "grad_norm": 1.3530856370925903, + "learning_rate": 1.3664916406068643e-05, + "loss": 0.6978, + "step": 9995 + }, + { + "epoch": 0.44257994531950456, + "grad_norm": 1.421428918838501, + "learning_rate": 1.3663478441774811e-05, + "loss": 0.7717, + "step": 9996 + }, + { + "epoch": 0.4426242210243184, + "grad_norm": 1.3234000205993652, + "learning_rate": 1.3662040389981777e-05, + "loss": 0.7075, + "step": 9997 + }, + { + "epoch": 0.44266849672913233, + "grad_norm": 1.1772339344024658, + "learning_rate": 1.3660602250723893e-05, + "loss": 0.679, + "step": 9998 + }, + { + "epoch": 0.4427127724339462, + "grad_norm": 1.3877283334732056, + "learning_rate": 1.3659164024035503e-05, + "loss": 0.6331, + "step": 9999 + }, + { + "epoch": 0.44275704813876005, + "grad_norm": 1.4331083297729492, + "learning_rate": 1.365772570995096e-05, + "loss": 0.674, + "step": 10000 + }, + { + "epoch": 0.44280132384357396, + "grad_norm": 1.457988977432251, + "learning_rate": 1.3656287308504617e-05, + "loss": 0.775, + "step": 10001 + }, + { + "epoch": 0.4428455995483878, + "grad_norm": 1.5481538772583008, + "learning_rate": 1.3654848819730829e-05, + "loss": 0.7582, + "step": 10002 + }, + { + "epoch": 0.44288987525320167, + "grad_norm": 1.648998498916626, + "learning_rate": 1.3653410243663953e-05, + "loss": 0.8372, + "step": 10003 + }, + { + "epoch": 0.4429341509580156, + "grad_norm": 1.4704283475875854, + "learning_rate": 1.3651971580338345e-05, + "loss": 0.6918, + "step": 10004 + }, + { + "epoch": 0.44297842666282944, + "grad_norm": 1.1622096300125122, + "learning_rate": 1.3650532829788373e-05, + "loss": 0.7073, + "step": 10005 + }, + { + "epoch": 0.4430227023676433, + "grad_norm": 1.6799211502075195, + "learning_rate": 1.3649093992048395e-05, + "loss": 0.6897, + "step": 10006 + }, + { + "epoch": 0.4430669780724572, + "grad_norm": 1.320387840270996, + "learning_rate": 1.364765506715278e-05, + "loss": 0.6426, + "step": 10007 + }, + { + "epoch": 0.44311125377727106, + "grad_norm": 1.8738224506378174, + "learning_rate": 1.3646216055135892e-05, + "loss": 0.7263, + "step": 10008 + }, + { + "epoch": 0.4431555294820849, + "grad_norm": 1.5932374000549316, + "learning_rate": 1.3644776956032103e-05, + "loss": 0.8123, + "step": 10009 + }, + { + "epoch": 0.44319980518689883, + "grad_norm": 1.6338651180267334, + "learning_rate": 1.3643337769875783e-05, + "loss": 0.8217, + "step": 10010 + }, + { + "epoch": 0.4432440808917127, + "grad_norm": 1.2901115417480469, + "learning_rate": 1.3641898496701306e-05, + "loss": 0.7121, + "step": 10011 + }, + { + "epoch": 0.44328835659652654, + "grad_norm": 1.4319652318954468, + "learning_rate": 1.3640459136543052e-05, + "loss": 0.6944, + "step": 10012 + }, + { + "epoch": 0.44333263230134046, + "grad_norm": 1.4463109970092773, + "learning_rate": 1.3639019689435396e-05, + "loss": 0.6731, + "step": 10013 + }, + { + "epoch": 0.4433769080061543, + "grad_norm": 1.590764045715332, + "learning_rate": 1.3637580155412716e-05, + "loss": 0.7337, + "step": 10014 + }, + { + "epoch": 0.4434211837109682, + "grad_norm": 1.342868685722351, + "learning_rate": 1.3636140534509392e-05, + "loss": 0.6489, + "step": 10015 + }, + { + "epoch": 0.4434654594157821, + "grad_norm": 1.3685388565063477, + "learning_rate": 1.3634700826759817e-05, + "loss": 0.752, + "step": 10016 + }, + { + "epoch": 0.44350973512059594, + "grad_norm": 1.8318302631378174, + "learning_rate": 1.3633261032198372e-05, + "loss": 0.8154, + "step": 10017 + }, + { + "epoch": 0.44355401082540985, + "grad_norm": 1.5087206363677979, + "learning_rate": 1.3631821150859446e-05, + "loss": 0.7533, + "step": 10018 + }, + { + "epoch": 0.4435982865302237, + "grad_norm": 1.4491652250289917, + "learning_rate": 1.3630381182777425e-05, + "loss": 0.743, + "step": 10019 + }, + { + "epoch": 0.44364256223503756, + "grad_norm": 1.3892097473144531, + "learning_rate": 1.3628941127986709e-05, + "loss": 0.7835, + "step": 10020 + }, + { + "epoch": 0.4436868379398515, + "grad_norm": 1.400850772857666, + "learning_rate": 1.3627500986521689e-05, + "loss": 0.6954, + "step": 10021 + }, + { + "epoch": 0.44373111364466533, + "grad_norm": 1.3734930753707886, + "learning_rate": 1.3626060758416759e-05, + "loss": 0.668, + "step": 10022 + }, + { + "epoch": 0.4437753893494792, + "grad_norm": 1.3285279273986816, + "learning_rate": 1.3624620443706322e-05, + "loss": 0.7932, + "step": 10023 + }, + { + "epoch": 0.4438196650542931, + "grad_norm": 1.5292749404907227, + "learning_rate": 1.3623180042424777e-05, + "loss": 0.7408, + "step": 10024 + }, + { + "epoch": 0.44386394075910696, + "grad_norm": 1.1338807344436646, + "learning_rate": 1.3621739554606524e-05, + "loss": 0.6871, + "step": 10025 + }, + { + "epoch": 0.4439082164639208, + "grad_norm": 1.4328861236572266, + "learning_rate": 1.3620298980285975e-05, + "loss": 0.7274, + "step": 10026 + }, + { + "epoch": 0.4439524921687347, + "grad_norm": 1.7755688428878784, + "learning_rate": 1.3618858319497532e-05, + "loss": 0.8108, + "step": 10027 + }, + { + "epoch": 0.4439967678735486, + "grad_norm": 1.4528805017471313, + "learning_rate": 1.3617417572275603e-05, + "loss": 0.6867, + "step": 10028 + }, + { + "epoch": 0.44404104357836244, + "grad_norm": 1.4840772151947021, + "learning_rate": 1.36159767386546e-05, + "loss": 0.7277, + "step": 10029 + }, + { + "epoch": 0.44408531928317635, + "grad_norm": 1.5385897159576416, + "learning_rate": 1.3614535818668938e-05, + "loss": 0.713, + "step": 10030 + }, + { + "epoch": 0.4441295949879902, + "grad_norm": 1.6043835878372192, + "learning_rate": 1.3613094812353032e-05, + "loss": 0.806, + "step": 10031 + }, + { + "epoch": 0.4441738706928041, + "grad_norm": 1.6964846849441528, + "learning_rate": 1.3611653719741295e-05, + "loss": 0.7327, + "step": 10032 + }, + { + "epoch": 0.444218146397618, + "grad_norm": 1.5707340240478516, + "learning_rate": 1.3610212540868154e-05, + "loss": 0.8153, + "step": 10033 + }, + { + "epoch": 0.44426242210243183, + "grad_norm": 1.336843729019165, + "learning_rate": 1.3608771275768022e-05, + "loss": 0.6879, + "step": 10034 + }, + { + "epoch": 0.44430669780724574, + "grad_norm": 1.297886848449707, + "learning_rate": 1.360732992447533e-05, + "loss": 0.7016, + "step": 10035 + }, + { + "epoch": 0.4443509735120596, + "grad_norm": 1.423335075378418, + "learning_rate": 1.3605888487024499e-05, + "loss": 0.7251, + "step": 10036 + }, + { + "epoch": 0.44439524921687346, + "grad_norm": 1.7257788181304932, + "learning_rate": 1.3604446963449956e-05, + "loss": 0.8405, + "step": 10037 + }, + { + "epoch": 0.44443952492168737, + "grad_norm": 1.4387471675872803, + "learning_rate": 1.3603005353786132e-05, + "loss": 0.7393, + "step": 10038 + }, + { + "epoch": 0.4444838006265012, + "grad_norm": 1.8394674062728882, + "learning_rate": 1.3601563658067458e-05, + "loss": 0.6399, + "step": 10039 + }, + { + "epoch": 0.4445280763313151, + "grad_norm": 1.258230209350586, + "learning_rate": 1.360012187632837e-05, + "loss": 0.6531, + "step": 10040 + }, + { + "epoch": 0.444572352036129, + "grad_norm": 1.3823086023330688, + "learning_rate": 1.3598680008603305e-05, + "loss": 0.7588, + "step": 10041 + }, + { + "epoch": 0.44461662774094285, + "grad_norm": 1.1675641536712646, + "learning_rate": 1.3597238054926693e-05, + "loss": 0.6625, + "step": 10042 + }, + { + "epoch": 0.4446609034457567, + "grad_norm": 1.765419840812683, + "learning_rate": 1.3595796015332986e-05, + "loss": 0.7383, + "step": 10043 + }, + { + "epoch": 0.4447051791505706, + "grad_norm": 1.3491932153701782, + "learning_rate": 1.359435388985661e-05, + "loss": 0.7061, + "step": 10044 + }, + { + "epoch": 0.4447494548553845, + "grad_norm": 1.1112314462661743, + "learning_rate": 1.3592911678532024e-05, + "loss": 0.6051, + "step": 10045 + }, + { + "epoch": 0.44479373056019833, + "grad_norm": 1.4798657894134521, + "learning_rate": 1.3591469381393668e-05, + "loss": 0.75, + "step": 10046 + }, + { + "epoch": 0.44483800626501224, + "grad_norm": 1.225071907043457, + "learning_rate": 1.3590026998475987e-05, + "loss": 0.6978, + "step": 10047 + }, + { + "epoch": 0.4448822819698261, + "grad_norm": 1.344746470451355, + "learning_rate": 1.3588584529813437e-05, + "loss": 0.7075, + "step": 10048 + }, + { + "epoch": 0.44492655767464, + "grad_norm": 1.348178744316101, + "learning_rate": 1.3587141975440468e-05, + "loss": 0.6942, + "step": 10049 + }, + { + "epoch": 0.44497083337945387, + "grad_norm": 1.5722272396087646, + "learning_rate": 1.3585699335391533e-05, + "loss": 0.6743, + "step": 10050 + }, + { + "epoch": 0.4450151090842677, + "grad_norm": 1.5054056644439697, + "learning_rate": 1.3584256609701086e-05, + "loss": 0.758, + "step": 10051 + }, + { + "epoch": 0.44505938478908164, + "grad_norm": 1.56282639503479, + "learning_rate": 1.3582813798403588e-05, + "loss": 0.7031, + "step": 10052 + }, + { + "epoch": 0.4451036604938955, + "grad_norm": 2.2077722549438477, + "learning_rate": 1.3581370901533503e-05, + "loss": 0.861, + "step": 10053 + }, + { + "epoch": 0.44514793619870935, + "grad_norm": 1.5952240228652954, + "learning_rate": 1.3579927919125291e-05, + "loss": 0.7071, + "step": 10054 + }, + { + "epoch": 0.44519221190352326, + "grad_norm": 1.383238673210144, + "learning_rate": 1.3578484851213416e-05, + "loss": 0.7432, + "step": 10055 + }, + { + "epoch": 0.4452364876083371, + "grad_norm": 1.2825151681900024, + "learning_rate": 1.3577041697832338e-05, + "loss": 0.6918, + "step": 10056 + }, + { + "epoch": 0.445280763313151, + "grad_norm": 1.2466986179351807, + "learning_rate": 1.3575598459016535e-05, + "loss": 0.6896, + "step": 10057 + }, + { + "epoch": 0.4453250390179649, + "grad_norm": 1.274483561515808, + "learning_rate": 1.3574155134800475e-05, + "loss": 0.6668, + "step": 10058 + }, + { + "epoch": 0.44536931472277874, + "grad_norm": 1.2456578016281128, + "learning_rate": 1.357271172521863e-05, + "loss": 0.7176, + "step": 10059 + }, + { + "epoch": 0.4454135904275926, + "grad_norm": 1.7903081178665161, + "learning_rate": 1.3571268230305474e-05, + "loss": 0.7617, + "step": 10060 + }, + { + "epoch": 0.4454578661324065, + "grad_norm": 1.3872051239013672, + "learning_rate": 1.3569824650095484e-05, + "loss": 0.6669, + "step": 10061 + }, + { + "epoch": 0.44550214183722037, + "grad_norm": 1.6247347593307495, + "learning_rate": 1.3568380984623137e-05, + "loss": 0.8991, + "step": 10062 + }, + { + "epoch": 0.4455464175420342, + "grad_norm": 1.1645886898040771, + "learning_rate": 1.3566937233922919e-05, + "loss": 0.6274, + "step": 10063 + }, + { + "epoch": 0.44559069324684814, + "grad_norm": 1.0876860618591309, + "learning_rate": 1.3565493398029308e-05, + "loss": 0.6086, + "step": 10064 + }, + { + "epoch": 0.445634968951662, + "grad_norm": 1.863520860671997, + "learning_rate": 1.356404947697679e-05, + "loss": 0.7116, + "step": 10065 + }, + { + "epoch": 0.44567924465647585, + "grad_norm": 1.558968424797058, + "learning_rate": 1.3562605470799852e-05, + "loss": 0.7313, + "step": 10066 + }, + { + "epoch": 0.44572352036128976, + "grad_norm": 1.3403043746948242, + "learning_rate": 1.3561161379532983e-05, + "loss": 0.7235, + "step": 10067 + }, + { + "epoch": 0.4457677960661036, + "grad_norm": 1.2984689474105835, + "learning_rate": 1.3559717203210675e-05, + "loss": 0.6713, + "step": 10068 + }, + { + "epoch": 0.44581207177091753, + "grad_norm": 1.2388896942138672, + "learning_rate": 1.355827294186742e-05, + "loss": 0.6621, + "step": 10069 + }, + { + "epoch": 0.4458563474757314, + "grad_norm": 1.588395595550537, + "learning_rate": 1.3556828595537712e-05, + "loss": 0.7347, + "step": 10070 + }, + { + "epoch": 0.44590062318054524, + "grad_norm": 1.339876651763916, + "learning_rate": 1.3555384164256048e-05, + "loss": 0.6766, + "step": 10071 + }, + { + "epoch": 0.44594489888535915, + "grad_norm": 1.4628074169158936, + "learning_rate": 1.3553939648056931e-05, + "loss": 0.7479, + "step": 10072 + }, + { + "epoch": 0.445989174590173, + "grad_norm": 1.2133389711380005, + "learning_rate": 1.355249504697486e-05, + "loss": 0.6582, + "step": 10073 + }, + { + "epoch": 0.44603345029498687, + "grad_norm": 1.3359993696212769, + "learning_rate": 1.3551050361044334e-05, + "loss": 0.7398, + "step": 10074 + }, + { + "epoch": 0.4460777259998008, + "grad_norm": 1.619859218597412, + "learning_rate": 1.3549605590299863e-05, + "loss": 0.7827, + "step": 10075 + }, + { + "epoch": 0.44612200170461463, + "grad_norm": 1.3739351034164429, + "learning_rate": 1.3548160734775953e-05, + "loss": 0.6769, + "step": 10076 + }, + { + "epoch": 0.4461662774094285, + "grad_norm": 1.4442965984344482, + "learning_rate": 1.3546715794507112e-05, + "loss": 0.7698, + "step": 10077 + }, + { + "epoch": 0.4462105531142424, + "grad_norm": 1.5291433334350586, + "learning_rate": 1.3545270769527854e-05, + "loss": 0.7854, + "step": 10078 + }, + { + "epoch": 0.44625482881905626, + "grad_norm": 1.3745616674423218, + "learning_rate": 1.3543825659872689e-05, + "loss": 0.7549, + "step": 10079 + }, + { + "epoch": 0.4462991045238701, + "grad_norm": 1.3922719955444336, + "learning_rate": 1.3542380465576133e-05, + "loss": 0.745, + "step": 10080 + }, + { + "epoch": 0.44634338022868403, + "grad_norm": 1.9322741031646729, + "learning_rate": 1.3540935186672704e-05, + "loss": 0.7949, + "step": 10081 + }, + { + "epoch": 0.4463876559334979, + "grad_norm": 1.7528311014175415, + "learning_rate": 1.3539489823196923e-05, + "loss": 0.7787, + "step": 10082 + }, + { + "epoch": 0.44643193163831174, + "grad_norm": 1.477750539779663, + "learning_rate": 1.3538044375183308e-05, + "loss": 0.8197, + "step": 10083 + }, + { + "epoch": 0.44647620734312565, + "grad_norm": 1.3690465688705444, + "learning_rate": 1.3536598842666382e-05, + "loss": 0.6679, + "step": 10084 + }, + { + "epoch": 0.4465204830479395, + "grad_norm": 1.7166988849639893, + "learning_rate": 1.3535153225680675e-05, + "loss": 0.7715, + "step": 10085 + }, + { + "epoch": 0.4465647587527534, + "grad_norm": 1.4822016954421997, + "learning_rate": 1.3533707524260709e-05, + "loss": 0.6645, + "step": 10086 + }, + { + "epoch": 0.4466090344575673, + "grad_norm": 1.2864702939987183, + "learning_rate": 1.3532261738441019e-05, + "loss": 0.7027, + "step": 10087 + }, + { + "epoch": 0.44665331016238113, + "grad_norm": 1.36102294921875, + "learning_rate": 1.3530815868256132e-05, + "loss": 0.6925, + "step": 10088 + }, + { + "epoch": 0.44669758586719505, + "grad_norm": 1.4895687103271484, + "learning_rate": 1.3529369913740585e-05, + "loss": 0.7662, + "step": 10089 + }, + { + "epoch": 0.4467418615720089, + "grad_norm": 1.500710368156433, + "learning_rate": 1.352792387492891e-05, + "loss": 0.7212, + "step": 10090 + }, + { + "epoch": 0.44678613727682276, + "grad_norm": 1.9391429424285889, + "learning_rate": 1.3526477751855645e-05, + "loss": 0.8194, + "step": 10091 + }, + { + "epoch": 0.44683041298163667, + "grad_norm": 1.497672200202942, + "learning_rate": 1.3525031544555329e-05, + "loss": 0.7379, + "step": 10092 + }, + { + "epoch": 0.4468746886864505, + "grad_norm": 1.302285075187683, + "learning_rate": 1.3523585253062507e-05, + "loss": 0.6455, + "step": 10093 + }, + { + "epoch": 0.4469189643912644, + "grad_norm": 1.2821680307388306, + "learning_rate": 1.3522138877411719e-05, + "loss": 0.6384, + "step": 10094 + }, + { + "epoch": 0.4469632400960783, + "grad_norm": 1.8289417028427124, + "learning_rate": 1.3520692417637513e-05, + "loss": 0.8494, + "step": 10095 + }, + { + "epoch": 0.44700751580089215, + "grad_norm": 1.2010775804519653, + "learning_rate": 1.3519245873774434e-05, + "loss": 0.6854, + "step": 10096 + }, + { + "epoch": 0.447051791505706, + "grad_norm": 1.505423665046692, + "learning_rate": 1.3517799245857032e-05, + "loss": 0.7265, + "step": 10097 + }, + { + "epoch": 0.4470960672105199, + "grad_norm": 1.3928362131118774, + "learning_rate": 1.3516352533919863e-05, + "loss": 0.6604, + "step": 10098 + }, + { + "epoch": 0.4471403429153338, + "grad_norm": 1.3145569562911987, + "learning_rate": 1.3514905737997474e-05, + "loss": 0.694, + "step": 10099 + }, + { + "epoch": 0.44718461862014763, + "grad_norm": 1.2359851598739624, + "learning_rate": 1.3513458858124424e-05, + "loss": 0.6983, + "step": 10100 + }, + { + "epoch": 0.44722889432496155, + "grad_norm": 1.4159899950027466, + "learning_rate": 1.351201189433527e-05, + "loss": 0.744, + "step": 10101 + }, + { + "epoch": 0.4472731700297754, + "grad_norm": 1.2745859622955322, + "learning_rate": 1.3510564846664571e-05, + "loss": 0.6499, + "step": 10102 + }, + { + "epoch": 0.4473174457345893, + "grad_norm": 1.598929762840271, + "learning_rate": 1.350911771514689e-05, + "loss": 0.7998, + "step": 10103 + }, + { + "epoch": 0.44736172143940317, + "grad_norm": 1.4421381950378418, + "learning_rate": 1.3507670499816787e-05, + "loss": 0.7589, + "step": 10104 + }, + { + "epoch": 0.447405997144217, + "grad_norm": 1.4426953792572021, + "learning_rate": 1.3506223200708835e-05, + "loss": 0.6603, + "step": 10105 + }, + { + "epoch": 0.44745027284903094, + "grad_norm": 1.3018057346343994, + "learning_rate": 1.3504775817857593e-05, + "loss": 0.7239, + "step": 10106 + }, + { + "epoch": 0.4474945485538448, + "grad_norm": 1.4562023878097534, + "learning_rate": 1.3503328351297635e-05, + "loss": 0.7215, + "step": 10107 + }, + { + "epoch": 0.44753882425865865, + "grad_norm": 1.3596692085266113, + "learning_rate": 1.3501880801063535e-05, + "loss": 0.7213, + "step": 10108 + }, + { + "epoch": 0.44758309996347256, + "grad_norm": 1.3288582563400269, + "learning_rate": 1.3500433167189857e-05, + "loss": 0.6797, + "step": 10109 + }, + { + "epoch": 0.4476273756682864, + "grad_norm": 1.275994062423706, + "learning_rate": 1.3498985449711186e-05, + "loss": 0.6377, + "step": 10110 + }, + { + "epoch": 0.4476716513731003, + "grad_norm": 1.3392523527145386, + "learning_rate": 1.3497537648662099e-05, + "loss": 0.6946, + "step": 10111 + }, + { + "epoch": 0.4477159270779142, + "grad_norm": 1.2918559312820435, + "learning_rate": 1.3496089764077169e-05, + "loss": 0.6899, + "step": 10112 + }, + { + "epoch": 0.44776020278272805, + "grad_norm": 1.5382176637649536, + "learning_rate": 1.3494641795990986e-05, + "loss": 0.6808, + "step": 10113 + }, + { + "epoch": 0.4478044784875419, + "grad_norm": 1.4710500240325928, + "learning_rate": 1.3493193744438125e-05, + "loss": 0.6993, + "step": 10114 + }, + { + "epoch": 0.4478487541923558, + "grad_norm": 1.562615990638733, + "learning_rate": 1.3491745609453177e-05, + "loss": 0.7911, + "step": 10115 + }, + { + "epoch": 0.44789302989716967, + "grad_norm": 1.438083529472351, + "learning_rate": 1.349029739107073e-05, + "loss": 0.7463, + "step": 10116 + }, + { + "epoch": 0.4479373056019835, + "grad_norm": 1.541189193725586, + "learning_rate": 1.3488849089325368e-05, + "loss": 0.7348, + "step": 10117 + }, + { + "epoch": 0.44798158130679744, + "grad_norm": 1.1712809801101685, + "learning_rate": 1.3487400704251692e-05, + "loss": 0.6849, + "step": 10118 + }, + { + "epoch": 0.4480258570116113, + "grad_norm": 1.2146780490875244, + "learning_rate": 1.3485952235884283e-05, + "loss": 0.6621, + "step": 10119 + }, + { + "epoch": 0.4480701327164252, + "grad_norm": 1.3919377326965332, + "learning_rate": 1.3484503684257748e-05, + "loss": 0.6513, + "step": 10120 + }, + { + "epoch": 0.44811440842123906, + "grad_norm": 1.2347688674926758, + "learning_rate": 1.348305504940668e-05, + "loss": 0.6854, + "step": 10121 + }, + { + "epoch": 0.4481586841260529, + "grad_norm": 1.603385090827942, + "learning_rate": 1.3481606331365673e-05, + "loss": 0.6841, + "step": 10122 + }, + { + "epoch": 0.44820295983086683, + "grad_norm": 1.865370750427246, + "learning_rate": 1.348015753016934e-05, + "loss": 0.7199, + "step": 10123 + }, + { + "epoch": 0.4482472355356807, + "grad_norm": 1.2596955299377441, + "learning_rate": 1.3478708645852272e-05, + "loss": 0.6426, + "step": 10124 + }, + { + "epoch": 0.44829151124049454, + "grad_norm": 1.271903157234192, + "learning_rate": 1.3477259678449086e-05, + "loss": 0.7134, + "step": 10125 + }, + { + "epoch": 0.44833578694530846, + "grad_norm": 1.9332306385040283, + "learning_rate": 1.347581062799438e-05, + "loss": 0.8745, + "step": 10126 + }, + { + "epoch": 0.4483800626501223, + "grad_norm": 1.6885590553283691, + "learning_rate": 1.3474361494522769e-05, + "loss": 0.7086, + "step": 10127 + }, + { + "epoch": 0.44842433835493617, + "grad_norm": 1.7121641635894775, + "learning_rate": 1.3472912278068862e-05, + "loss": 0.8113, + "step": 10128 + }, + { + "epoch": 0.4484686140597501, + "grad_norm": 1.556481122970581, + "learning_rate": 1.3471462978667274e-05, + "loss": 0.7331, + "step": 10129 + }, + { + "epoch": 0.44851288976456394, + "grad_norm": 1.3894188404083252, + "learning_rate": 1.347001359635262e-05, + "loss": 0.7528, + "step": 10130 + }, + { + "epoch": 0.4485571654693778, + "grad_norm": 1.202946424484253, + "learning_rate": 1.3468564131159515e-05, + "loss": 0.6329, + "step": 10131 + }, + { + "epoch": 0.4486014411741917, + "grad_norm": 1.6174514293670654, + "learning_rate": 1.3467114583122579e-05, + "loss": 0.6902, + "step": 10132 + }, + { + "epoch": 0.44864571687900556, + "grad_norm": 1.3923066854476929, + "learning_rate": 1.3465664952276434e-05, + "loss": 0.7128, + "step": 10133 + }, + { + "epoch": 0.4486899925838194, + "grad_norm": 1.2268953323364258, + "learning_rate": 1.3464215238655707e-05, + "loss": 0.6614, + "step": 10134 + }, + { + "epoch": 0.44873426828863333, + "grad_norm": 1.5378879308700562, + "learning_rate": 1.3462765442295018e-05, + "loss": 0.7562, + "step": 10135 + }, + { + "epoch": 0.4487785439934472, + "grad_norm": 1.47900390625, + "learning_rate": 1.3461315563228994e-05, + "loss": 0.6599, + "step": 10136 + }, + { + "epoch": 0.44882281969826104, + "grad_norm": 1.5913437604904175, + "learning_rate": 1.3459865601492266e-05, + "loss": 0.7003, + "step": 10137 + }, + { + "epoch": 0.44886709540307496, + "grad_norm": 1.3819721937179565, + "learning_rate": 1.3458415557119469e-05, + "loss": 0.7091, + "step": 10138 + }, + { + "epoch": 0.4489113711078888, + "grad_norm": 1.2901501655578613, + "learning_rate": 1.345696543014523e-05, + "loss": 0.6626, + "step": 10139 + }, + { + "epoch": 0.4489556468127027, + "grad_norm": 2.4339542388916016, + "learning_rate": 1.3455515220604188e-05, + "loss": 0.8158, + "step": 10140 + }, + { + "epoch": 0.4489999225175166, + "grad_norm": 1.8413537740707397, + "learning_rate": 1.3454064928530975e-05, + "loss": 0.7978, + "step": 10141 + }, + { + "epoch": 0.44904419822233044, + "grad_norm": 1.4337681531906128, + "learning_rate": 1.3452614553960234e-05, + "loss": 0.7257, + "step": 10142 + }, + { + "epoch": 0.44908847392714435, + "grad_norm": 1.7886359691619873, + "learning_rate": 1.3451164096926608e-05, + "loss": 0.8271, + "step": 10143 + }, + { + "epoch": 0.4491327496319582, + "grad_norm": 1.2886658906936646, + "learning_rate": 1.3449713557464739e-05, + "loss": 0.6825, + "step": 10144 + }, + { + "epoch": 0.44917702533677206, + "grad_norm": 1.2875211238861084, + "learning_rate": 1.3448262935609267e-05, + "loss": 0.6697, + "step": 10145 + }, + { + "epoch": 0.449221301041586, + "grad_norm": 1.7731763124465942, + "learning_rate": 1.3446812231394843e-05, + "loss": 0.7147, + "step": 10146 + }, + { + "epoch": 0.44926557674639983, + "grad_norm": 1.31821870803833, + "learning_rate": 1.3445361444856115e-05, + "loss": 0.6926, + "step": 10147 + }, + { + "epoch": 0.4493098524512137, + "grad_norm": 1.406180739402771, + "learning_rate": 1.3443910576027738e-05, + "loss": 0.6512, + "step": 10148 + }, + { + "epoch": 0.4493541281560276, + "grad_norm": 1.2766387462615967, + "learning_rate": 1.3442459624944358e-05, + "loss": 0.647, + "step": 10149 + }, + { + "epoch": 0.44939840386084146, + "grad_norm": 1.5304533243179321, + "learning_rate": 1.3441008591640634e-05, + "loss": 0.708, + "step": 10150 + }, + { + "epoch": 0.4494426795656553, + "grad_norm": 1.5352840423583984, + "learning_rate": 1.343955747615122e-05, + "loss": 0.7976, + "step": 10151 + }, + { + "epoch": 0.4494869552704692, + "grad_norm": 1.2824618816375732, + "learning_rate": 1.3438106278510775e-05, + "loss": 0.7106, + "step": 10152 + }, + { + "epoch": 0.4495312309752831, + "grad_norm": 1.3551350831985474, + "learning_rate": 1.3436654998753967e-05, + "loss": 0.72, + "step": 10153 + }, + { + "epoch": 0.44957550668009694, + "grad_norm": 1.302741289138794, + "learning_rate": 1.3435203636915447e-05, + "loss": 0.6599, + "step": 10154 + }, + { + "epoch": 0.44961978238491085, + "grad_norm": 1.4393669366836548, + "learning_rate": 1.3433752193029888e-05, + "loss": 0.7648, + "step": 10155 + }, + { + "epoch": 0.4496640580897247, + "grad_norm": 1.8930089473724365, + "learning_rate": 1.343230066713195e-05, + "loss": 0.7652, + "step": 10156 + }, + { + "epoch": 0.4497083337945386, + "grad_norm": 1.557442307472229, + "learning_rate": 1.3430849059256308e-05, + "loss": 0.7564, + "step": 10157 + }, + { + "epoch": 0.4497526094993525, + "grad_norm": 1.5315836668014526, + "learning_rate": 1.342939736943763e-05, + "loss": 0.7391, + "step": 10158 + }, + { + "epoch": 0.44979688520416633, + "grad_norm": 1.5374250411987305, + "learning_rate": 1.3427945597710587e-05, + "loss": 0.7814, + "step": 10159 + }, + { + "epoch": 0.44984116090898024, + "grad_norm": 1.3502610921859741, + "learning_rate": 1.3426493744109857e-05, + "loss": 0.7162, + "step": 10160 + }, + { + "epoch": 0.4498854366137941, + "grad_norm": 1.5854393243789673, + "learning_rate": 1.342504180867011e-05, + "loss": 0.7207, + "step": 10161 + }, + { + "epoch": 0.44992971231860796, + "grad_norm": 1.4847643375396729, + "learning_rate": 1.342358979142603e-05, + "loss": 0.7553, + "step": 10162 + }, + { + "epoch": 0.44997398802342187, + "grad_norm": 1.6250859498977661, + "learning_rate": 1.3422137692412298e-05, + "loss": 0.7164, + "step": 10163 + }, + { + "epoch": 0.4500182637282357, + "grad_norm": 1.8353222608566284, + "learning_rate": 1.342068551166359e-05, + "loss": 0.7495, + "step": 10164 + }, + { + "epoch": 0.4500625394330496, + "grad_norm": 1.3105323314666748, + "learning_rate": 1.3419233249214597e-05, + "loss": 0.7493, + "step": 10165 + }, + { + "epoch": 0.4501068151378635, + "grad_norm": 1.6867294311523438, + "learning_rate": 1.3417780905099998e-05, + "loss": 0.7847, + "step": 10166 + }, + { + "epoch": 0.45015109084267735, + "grad_norm": 1.4291940927505493, + "learning_rate": 1.3416328479354489e-05, + "loss": 0.6388, + "step": 10167 + }, + { + "epoch": 0.4501953665474912, + "grad_norm": 1.2204666137695312, + "learning_rate": 1.3414875972012753e-05, + "loss": 0.6709, + "step": 10168 + }, + { + "epoch": 0.4502396422523051, + "grad_norm": 1.6160190105438232, + "learning_rate": 1.3413423383109482e-05, + "loss": 0.8561, + "step": 10169 + }, + { + "epoch": 0.450283917957119, + "grad_norm": 1.190807580947876, + "learning_rate": 1.341197071267938e-05, + "loss": 0.6696, + "step": 10170 + }, + { + "epoch": 0.45032819366193283, + "grad_norm": 1.294741153717041, + "learning_rate": 1.341051796075713e-05, + "loss": 0.6333, + "step": 10171 + }, + { + "epoch": 0.45037246936674674, + "grad_norm": 1.5911943912506104, + "learning_rate": 1.340906512737744e-05, + "loss": 0.7558, + "step": 10172 + }, + { + "epoch": 0.4504167450715606, + "grad_norm": 1.5426390171051025, + "learning_rate": 1.3407612212575003e-05, + "loss": 0.7333, + "step": 10173 + }, + { + "epoch": 0.4504610207763745, + "grad_norm": 1.3850563764572144, + "learning_rate": 1.340615921638452e-05, + "loss": 0.6417, + "step": 10174 + }, + { + "epoch": 0.45050529648118837, + "grad_norm": 1.52497398853302, + "learning_rate": 1.3404706138840701e-05, + "loss": 0.6479, + "step": 10175 + }, + { + "epoch": 0.4505495721860022, + "grad_norm": 1.3633078336715698, + "learning_rate": 1.340325297997825e-05, + "loss": 0.6869, + "step": 10176 + }, + { + "epoch": 0.45059384789081613, + "grad_norm": 1.3317747116088867, + "learning_rate": 1.3401799739831871e-05, + "loss": 0.6012, + "step": 10177 + }, + { + "epoch": 0.45063812359563, + "grad_norm": 1.5704543590545654, + "learning_rate": 1.3400346418436272e-05, + "loss": 0.7903, + "step": 10178 + }, + { + "epoch": 0.45068239930044385, + "grad_norm": 1.64057195186615, + "learning_rate": 1.3398893015826166e-05, + "loss": 0.8194, + "step": 10179 + }, + { + "epoch": 0.45072667500525776, + "grad_norm": 1.466870903968811, + "learning_rate": 1.3397439532036273e-05, + "loss": 0.7307, + "step": 10180 + }, + { + "epoch": 0.4507709507100716, + "grad_norm": 1.3247699737548828, + "learning_rate": 1.3395985967101303e-05, + "loss": 0.738, + "step": 10181 + }, + { + "epoch": 0.4508152264148855, + "grad_norm": 1.3541868925094604, + "learning_rate": 1.3394532321055973e-05, + "loss": 0.7037, + "step": 10182 + }, + { + "epoch": 0.4508595021196994, + "grad_norm": 1.3450870513916016, + "learning_rate": 1.3393078593934998e-05, + "loss": 0.6537, + "step": 10183 + }, + { + "epoch": 0.45090377782451324, + "grad_norm": 1.8161870241165161, + "learning_rate": 1.3391624785773107e-05, + "loss": 0.7401, + "step": 10184 + }, + { + "epoch": 0.4509480535293271, + "grad_norm": 1.8270663022994995, + "learning_rate": 1.3390170896605019e-05, + "loss": 0.7053, + "step": 10185 + }, + { + "epoch": 0.450992329234141, + "grad_norm": 1.541831612586975, + "learning_rate": 1.338871692646546e-05, + "loss": 0.667, + "step": 10186 + }, + { + "epoch": 0.45103660493895487, + "grad_norm": 1.4046567678451538, + "learning_rate": 1.3387262875389158e-05, + "loss": 0.7294, + "step": 10187 + }, + { + "epoch": 0.4510808806437687, + "grad_norm": 1.3569835424423218, + "learning_rate": 1.3385808743410838e-05, + "loss": 0.7257, + "step": 10188 + }, + { + "epoch": 0.45112515634858263, + "grad_norm": 1.5142457485198975, + "learning_rate": 1.338435453056523e-05, + "loss": 0.7486, + "step": 10189 + }, + { + "epoch": 0.4511694320533965, + "grad_norm": 1.381493091583252, + "learning_rate": 1.3382900236887075e-05, + "loss": 0.6624, + "step": 10190 + }, + { + "epoch": 0.4512137077582104, + "grad_norm": 1.7058360576629639, + "learning_rate": 1.3381445862411102e-05, + "loss": 0.7399, + "step": 10191 + }, + { + "epoch": 0.45125798346302426, + "grad_norm": 1.424787163734436, + "learning_rate": 1.3379991407172046e-05, + "loss": 0.7658, + "step": 10192 + }, + { + "epoch": 0.4513022591678381, + "grad_norm": 1.549532175064087, + "learning_rate": 1.3378536871204646e-05, + "loss": 0.7261, + "step": 10193 + }, + { + "epoch": 0.451346534872652, + "grad_norm": 1.6029212474822998, + "learning_rate": 1.3377082254543643e-05, + "loss": 0.7023, + "step": 10194 + }, + { + "epoch": 0.4513908105774659, + "grad_norm": 1.0907613039016724, + "learning_rate": 1.3375627557223785e-05, + "loss": 0.6427, + "step": 10195 + }, + { + "epoch": 0.45143508628227974, + "grad_norm": 1.3756452798843384, + "learning_rate": 1.3374172779279808e-05, + "loss": 0.696, + "step": 10196 + }, + { + "epoch": 0.45147936198709365, + "grad_norm": 1.530282974243164, + "learning_rate": 1.3372717920746464e-05, + "loss": 0.7194, + "step": 10197 + }, + { + "epoch": 0.4515236376919075, + "grad_norm": 1.4119677543640137, + "learning_rate": 1.3371262981658495e-05, + "loss": 0.686, + "step": 10198 + }, + { + "epoch": 0.45156791339672137, + "grad_norm": 1.6966429948806763, + "learning_rate": 1.3369807962050656e-05, + "loss": 0.7953, + "step": 10199 + }, + { + "epoch": 0.4516121891015353, + "grad_norm": 1.1837644577026367, + "learning_rate": 1.3368352861957701e-05, + "loss": 0.7065, + "step": 10200 + }, + { + "epoch": 0.45165646480634913, + "grad_norm": 1.332646131515503, + "learning_rate": 1.3366897681414375e-05, + "loss": 0.6659, + "step": 10201 + }, + { + "epoch": 0.451700740511163, + "grad_norm": 1.790825605392456, + "learning_rate": 1.3365442420455443e-05, + "loss": 0.6936, + "step": 10202 + }, + { + "epoch": 0.4517450162159769, + "grad_norm": 1.580080509185791, + "learning_rate": 1.3363987079115656e-05, + "loss": 0.7418, + "step": 10203 + }, + { + "epoch": 0.45178929192079076, + "grad_norm": 1.1495587825775146, + "learning_rate": 1.336253165742978e-05, + "loss": 0.6547, + "step": 10204 + }, + { + "epoch": 0.4518335676256046, + "grad_norm": 1.53677499294281, + "learning_rate": 1.3361076155432573e-05, + "loss": 0.7746, + "step": 10205 + }, + { + "epoch": 0.4518778433304185, + "grad_norm": 1.4668713808059692, + "learning_rate": 1.3359620573158796e-05, + "loss": 0.7957, + "step": 10206 + }, + { + "epoch": 0.4519221190352324, + "grad_norm": 1.3159751892089844, + "learning_rate": 1.3358164910643221e-05, + "loss": 0.6899, + "step": 10207 + }, + { + "epoch": 0.4519663947400463, + "grad_norm": 1.5594964027404785, + "learning_rate": 1.3356709167920606e-05, + "loss": 0.692, + "step": 10208 + }, + { + "epoch": 0.45201067044486015, + "grad_norm": 1.2186338901519775, + "learning_rate": 1.3355253345025729e-05, + "loss": 0.6316, + "step": 10209 + }, + { + "epoch": 0.452054946149674, + "grad_norm": 1.4004911184310913, + "learning_rate": 1.335379744199336e-05, + "loss": 0.687, + "step": 10210 + }, + { + "epoch": 0.4520992218544879, + "grad_norm": 1.5958389043807983, + "learning_rate": 1.3352341458858264e-05, + "loss": 0.7169, + "step": 10211 + }, + { + "epoch": 0.4521434975593018, + "grad_norm": 1.6883795261383057, + "learning_rate": 1.3350885395655229e-05, + "loss": 0.7718, + "step": 10212 + }, + { + "epoch": 0.45218777326411563, + "grad_norm": 1.6826391220092773, + "learning_rate": 1.3349429252419017e-05, + "loss": 0.7639, + "step": 10213 + }, + { + "epoch": 0.45223204896892955, + "grad_norm": 1.4084779024124146, + "learning_rate": 1.334797302918442e-05, + "loss": 0.7272, + "step": 10214 + }, + { + "epoch": 0.4522763246737434, + "grad_norm": 1.5380076169967651, + "learning_rate": 1.3346516725986214e-05, + "loss": 0.6803, + "step": 10215 + }, + { + "epoch": 0.45232060037855726, + "grad_norm": 1.583751916885376, + "learning_rate": 1.3345060342859175e-05, + "loss": 0.7849, + "step": 10216 + }, + { + "epoch": 0.45236487608337117, + "grad_norm": 2.0822348594665527, + "learning_rate": 1.3343603879838098e-05, + "loss": 0.7647, + "step": 10217 + }, + { + "epoch": 0.452409151788185, + "grad_norm": 1.3330342769622803, + "learning_rate": 1.3342147336957764e-05, + "loss": 0.6829, + "step": 10218 + }, + { + "epoch": 0.4524534274929989, + "grad_norm": 1.5493342876434326, + "learning_rate": 1.3340690714252962e-05, + "loss": 0.8257, + "step": 10219 + }, + { + "epoch": 0.4524977031978128, + "grad_norm": 2.2605481147766113, + "learning_rate": 1.3339234011758482e-05, + "loss": 0.7161, + "step": 10220 + }, + { + "epoch": 0.45254197890262665, + "grad_norm": 1.3877662420272827, + "learning_rate": 1.3337777229509117e-05, + "loss": 0.7156, + "step": 10221 + }, + { + "epoch": 0.4525862546074405, + "grad_norm": 1.2890002727508545, + "learning_rate": 1.3336320367539662e-05, + "loss": 0.6916, + "step": 10222 + }, + { + "epoch": 0.4526305303122544, + "grad_norm": 1.2995233535766602, + "learning_rate": 1.3334863425884907e-05, + "loss": 0.6403, + "step": 10223 + }, + { + "epoch": 0.4526748060170683, + "grad_norm": 1.4916585683822632, + "learning_rate": 1.3333406404579658e-05, + "loss": 0.668, + "step": 10224 + }, + { + "epoch": 0.45271908172188213, + "grad_norm": 1.4272212982177734, + "learning_rate": 1.333194930365871e-05, + "loss": 0.6858, + "step": 10225 + }, + { + "epoch": 0.45276335742669604, + "grad_norm": 1.3174021244049072, + "learning_rate": 1.3330492123156866e-05, + "loss": 0.6624, + "step": 10226 + }, + { + "epoch": 0.4528076331315099, + "grad_norm": 1.247139573097229, + "learning_rate": 1.3329034863108932e-05, + "loss": 0.6848, + "step": 10227 + }, + { + "epoch": 0.4528519088363238, + "grad_norm": 1.5930757522583008, + "learning_rate": 1.3327577523549709e-05, + "loss": 0.7433, + "step": 10228 + }, + { + "epoch": 0.45289618454113767, + "grad_norm": 1.2477867603302002, + "learning_rate": 1.3326120104514004e-05, + "loss": 0.6583, + "step": 10229 + }, + { + "epoch": 0.4529404602459515, + "grad_norm": 1.5357078313827515, + "learning_rate": 1.3324662606036631e-05, + "loss": 0.6753, + "step": 10230 + }, + { + "epoch": 0.45298473595076544, + "grad_norm": 1.7165840864181519, + "learning_rate": 1.3323205028152397e-05, + "loss": 0.7372, + "step": 10231 + }, + { + "epoch": 0.4530290116555793, + "grad_norm": 1.4032477140426636, + "learning_rate": 1.3321747370896119e-05, + "loss": 0.7197, + "step": 10232 + }, + { + "epoch": 0.45307328736039315, + "grad_norm": 1.3813692331314087, + "learning_rate": 1.332028963430261e-05, + "loss": 0.6368, + "step": 10233 + }, + { + "epoch": 0.45311756306520706, + "grad_norm": 1.3950471878051758, + "learning_rate": 1.3318831818406683e-05, + "loss": 0.7078, + "step": 10234 + }, + { + "epoch": 0.4531618387700209, + "grad_norm": 1.2620528936386108, + "learning_rate": 1.3317373923243165e-05, + "loss": 0.6569, + "step": 10235 + }, + { + "epoch": 0.4532061144748348, + "grad_norm": 1.8726505041122437, + "learning_rate": 1.3315915948846867e-05, + "loss": 0.7191, + "step": 10236 + }, + { + "epoch": 0.4532503901796487, + "grad_norm": 1.8869917392730713, + "learning_rate": 1.3314457895252619e-05, + "loss": 0.776, + "step": 10237 + }, + { + "epoch": 0.45329466588446254, + "grad_norm": 1.1993114948272705, + "learning_rate": 1.3312999762495244e-05, + "loss": 0.6793, + "step": 10238 + }, + { + "epoch": 0.4533389415892764, + "grad_norm": 1.4057892560958862, + "learning_rate": 1.3311541550609566e-05, + "loss": 0.6748, + "step": 10239 + }, + { + "epoch": 0.4533832172940903, + "grad_norm": 1.297257661819458, + "learning_rate": 1.3310083259630415e-05, + "loss": 0.7255, + "step": 10240 + }, + { + "epoch": 0.45342749299890417, + "grad_norm": 1.281509280204773, + "learning_rate": 1.3308624889592618e-05, + "loss": 0.737, + "step": 10241 + }, + { + "epoch": 0.453471768703718, + "grad_norm": 1.4740822315216064, + "learning_rate": 1.3307166440531011e-05, + "loss": 0.7359, + "step": 10242 + }, + { + "epoch": 0.45351604440853194, + "grad_norm": 1.5648603439331055, + "learning_rate": 1.330570791248043e-05, + "loss": 0.7668, + "step": 10243 + }, + { + "epoch": 0.4535603201133458, + "grad_norm": 1.5362285375595093, + "learning_rate": 1.33042493054757e-05, + "loss": 0.7119, + "step": 10244 + }, + { + "epoch": 0.4536045958181597, + "grad_norm": 1.5336148738861084, + "learning_rate": 1.3302790619551673e-05, + "loss": 0.7187, + "step": 10245 + }, + { + "epoch": 0.45364887152297356, + "grad_norm": 1.1315327882766724, + "learning_rate": 1.3301331854743175e-05, + "loss": 0.6304, + "step": 10246 + }, + { + "epoch": 0.4536931472277874, + "grad_norm": 2.130587339401245, + "learning_rate": 1.3299873011085058e-05, + "loss": 0.7028, + "step": 10247 + }, + { + "epoch": 0.45373742293260133, + "grad_norm": 1.3339720964431763, + "learning_rate": 1.3298414088612162e-05, + "loss": 0.8194, + "step": 10248 + }, + { + "epoch": 0.4537816986374152, + "grad_norm": 1.6453368663787842, + "learning_rate": 1.3296955087359326e-05, + "loss": 0.8146, + "step": 10249 + }, + { + "epoch": 0.45382597434222904, + "grad_norm": 1.4845929145812988, + "learning_rate": 1.3295496007361407e-05, + "loss": 0.7036, + "step": 10250 + }, + { + "epoch": 0.45387025004704296, + "grad_norm": 1.3466299772262573, + "learning_rate": 1.3294036848653247e-05, + "loss": 0.7526, + "step": 10251 + }, + { + "epoch": 0.4539145257518568, + "grad_norm": 1.1652086973190308, + "learning_rate": 1.3292577611269701e-05, + "loss": 0.6671, + "step": 10252 + }, + { + "epoch": 0.45395880145667067, + "grad_norm": 1.5753170251846313, + "learning_rate": 1.3291118295245617e-05, + "loss": 0.7705, + "step": 10253 + }, + { + "epoch": 0.4540030771614846, + "grad_norm": 1.3919167518615723, + "learning_rate": 1.3289658900615851e-05, + "loss": 0.6846, + "step": 10254 + }, + { + "epoch": 0.45404735286629844, + "grad_norm": 1.3863402605056763, + "learning_rate": 1.3288199427415264e-05, + "loss": 0.7059, + "step": 10255 + }, + { + "epoch": 0.4540916285711123, + "grad_norm": 1.2848527431488037, + "learning_rate": 1.328673987567871e-05, + "loss": 0.7108, + "step": 10256 + }, + { + "epoch": 0.4541359042759262, + "grad_norm": 1.457365870475769, + "learning_rate": 1.328528024544105e-05, + "loss": 0.7016, + "step": 10257 + }, + { + "epoch": 0.45418017998074006, + "grad_norm": 1.331117868423462, + "learning_rate": 1.3283820536737146e-05, + "loss": 0.6645, + "step": 10258 + }, + { + "epoch": 0.4542244556855539, + "grad_norm": 1.5413633584976196, + "learning_rate": 1.3282360749601862e-05, + "loss": 0.7345, + "step": 10259 + }, + { + "epoch": 0.45426873139036783, + "grad_norm": 1.787350058555603, + "learning_rate": 1.3280900884070064e-05, + "loss": 0.7812, + "step": 10260 + }, + { + "epoch": 0.4543130070951817, + "grad_norm": 1.3858741521835327, + "learning_rate": 1.3279440940176622e-05, + "loss": 0.6673, + "step": 10261 + }, + { + "epoch": 0.4543572827999956, + "grad_norm": 1.5197137594223022, + "learning_rate": 1.3277980917956404e-05, + "loss": 0.7103, + "step": 10262 + }, + { + "epoch": 0.45440155850480946, + "grad_norm": 1.3866139650344849, + "learning_rate": 1.3276520817444276e-05, + "loss": 0.6963, + "step": 10263 + }, + { + "epoch": 0.4544458342096233, + "grad_norm": 1.5870773792266846, + "learning_rate": 1.3275060638675119e-05, + "loss": 0.7458, + "step": 10264 + }, + { + "epoch": 0.4544901099144372, + "grad_norm": 1.314540982246399, + "learning_rate": 1.3273600381683804e-05, + "loss": 0.671, + "step": 10265 + }, + { + "epoch": 0.4545343856192511, + "grad_norm": 1.396725058555603, + "learning_rate": 1.3272140046505213e-05, + "loss": 0.7455, + "step": 10266 + }, + { + "epoch": 0.45457866132406494, + "grad_norm": 1.245913028717041, + "learning_rate": 1.3270679633174219e-05, + "loss": 0.6318, + "step": 10267 + }, + { + "epoch": 0.45462293702887885, + "grad_norm": 1.6341913938522339, + "learning_rate": 1.3269219141725702e-05, + "loss": 0.721, + "step": 10268 + }, + { + "epoch": 0.4546672127336927, + "grad_norm": 1.8037762641906738, + "learning_rate": 1.3267758572194553e-05, + "loss": 0.7595, + "step": 10269 + }, + { + "epoch": 0.45471148843850656, + "grad_norm": 1.3772956132888794, + "learning_rate": 1.3266297924615648e-05, + "loss": 0.6842, + "step": 10270 + }, + { + "epoch": 0.4547557641433205, + "grad_norm": 1.275477409362793, + "learning_rate": 1.3264837199023877e-05, + "loss": 0.6473, + "step": 10271 + }, + { + "epoch": 0.45480003984813433, + "grad_norm": 1.6613107919692993, + "learning_rate": 1.3263376395454128e-05, + "loss": 0.7509, + "step": 10272 + }, + { + "epoch": 0.4548443155529482, + "grad_norm": 1.7237402200698853, + "learning_rate": 1.326191551394129e-05, + "loss": 0.8291, + "step": 10273 + }, + { + "epoch": 0.4548885912577621, + "grad_norm": 1.48918616771698, + "learning_rate": 1.3260454554520258e-05, + "loss": 0.7468, + "step": 10274 + }, + { + "epoch": 0.45493286696257595, + "grad_norm": 1.2126104831695557, + "learning_rate": 1.3258993517225923e-05, + "loss": 0.6393, + "step": 10275 + }, + { + "epoch": 0.4549771426673898, + "grad_norm": 1.6192996501922607, + "learning_rate": 1.3257532402093181e-05, + "loss": 0.7417, + "step": 10276 + }, + { + "epoch": 0.4550214183722037, + "grad_norm": 1.5627412796020508, + "learning_rate": 1.3256071209156932e-05, + "loss": 0.6889, + "step": 10277 + }, + { + "epoch": 0.4550656940770176, + "grad_norm": 1.4916646480560303, + "learning_rate": 1.3254609938452068e-05, + "loss": 0.7925, + "step": 10278 + }, + { + "epoch": 0.4551099697818315, + "grad_norm": 1.6965210437774658, + "learning_rate": 1.32531485900135e-05, + "loss": 0.7778, + "step": 10279 + }, + { + "epoch": 0.45515424548664535, + "grad_norm": 1.3401916027069092, + "learning_rate": 1.3251687163876128e-05, + "loss": 0.7199, + "step": 10280 + }, + { + "epoch": 0.4551985211914592, + "grad_norm": 1.7475531101226807, + "learning_rate": 1.3250225660074852e-05, + "loss": 0.7866, + "step": 10281 + }, + { + "epoch": 0.4552427968962731, + "grad_norm": 1.5763798952102661, + "learning_rate": 1.3248764078644584e-05, + "loss": 0.7076, + "step": 10282 + }, + { + "epoch": 0.455287072601087, + "grad_norm": 1.5827463865280151, + "learning_rate": 1.3247302419620232e-05, + "loss": 0.6901, + "step": 10283 + }, + { + "epoch": 0.45533134830590083, + "grad_norm": 1.3496516942977905, + "learning_rate": 1.3245840683036703e-05, + "loss": 0.7317, + "step": 10284 + }, + { + "epoch": 0.45537562401071474, + "grad_norm": 1.4881787300109863, + "learning_rate": 1.3244378868928916e-05, + "loss": 0.8173, + "step": 10285 + }, + { + "epoch": 0.4554198997155286, + "grad_norm": 1.2625430822372437, + "learning_rate": 1.3242916977331779e-05, + "loss": 0.6674, + "step": 10286 + }, + { + "epoch": 0.45546417542034245, + "grad_norm": 1.1977286338806152, + "learning_rate": 1.324145500828021e-05, + "loss": 0.6429, + "step": 10287 + }, + { + "epoch": 0.45550845112515637, + "grad_norm": 1.248264193534851, + "learning_rate": 1.3239992961809129e-05, + "loss": 0.6288, + "step": 10288 + }, + { + "epoch": 0.4555527268299702, + "grad_norm": 1.405349612236023, + "learning_rate": 1.3238530837953455e-05, + "loss": 0.6803, + "step": 10289 + }, + { + "epoch": 0.4555970025347841, + "grad_norm": 1.157559871673584, + "learning_rate": 1.3237068636748107e-05, + "loss": 0.663, + "step": 10290 + }, + { + "epoch": 0.455641278239598, + "grad_norm": 1.1708743572235107, + "learning_rate": 1.323560635822801e-05, + "loss": 0.614, + "step": 10291 + }, + { + "epoch": 0.45568555394441185, + "grad_norm": 1.643283724784851, + "learning_rate": 1.323414400242809e-05, + "loss": 0.7133, + "step": 10292 + }, + { + "epoch": 0.4557298296492257, + "grad_norm": 1.2977118492126465, + "learning_rate": 1.3232681569383275e-05, + "loss": 0.6873, + "step": 10293 + }, + { + "epoch": 0.4557741053540396, + "grad_norm": 2.0737998485565186, + "learning_rate": 1.3231219059128493e-05, + "loss": 0.6514, + "step": 10294 + }, + { + "epoch": 0.4558183810588535, + "grad_norm": 1.4940627813339233, + "learning_rate": 1.3229756471698674e-05, + "loss": 0.7122, + "step": 10295 + }, + { + "epoch": 0.4558626567636674, + "grad_norm": 1.5074554681777954, + "learning_rate": 1.322829380712875e-05, + "loss": 0.7614, + "step": 10296 + }, + { + "epoch": 0.45590693246848124, + "grad_norm": 1.2181053161621094, + "learning_rate": 1.3226831065453658e-05, + "loss": 0.6694, + "step": 10297 + }, + { + "epoch": 0.4559512081732951, + "grad_norm": 1.3387560844421387, + "learning_rate": 1.3225368246708334e-05, + "loss": 0.7565, + "step": 10298 + }, + { + "epoch": 0.455995483878109, + "grad_norm": 1.3668186664581299, + "learning_rate": 1.3223905350927718e-05, + "loss": 0.6356, + "step": 10299 + }, + { + "epoch": 0.45603975958292287, + "grad_norm": 1.3456594944000244, + "learning_rate": 1.3222442378146743e-05, + "loss": 0.6501, + "step": 10300 + }, + { + "epoch": 0.4560840352877367, + "grad_norm": 1.5249112844467163, + "learning_rate": 1.3220979328400356e-05, + "loss": 0.7737, + "step": 10301 + }, + { + "epoch": 0.45612831099255063, + "grad_norm": 1.9182595014572144, + "learning_rate": 1.3219516201723503e-05, + "loss": 0.7882, + "step": 10302 + }, + { + "epoch": 0.4561725866973645, + "grad_norm": 1.5639927387237549, + "learning_rate": 1.3218052998151128e-05, + "loss": 0.7491, + "step": 10303 + }, + { + "epoch": 0.45621686240217835, + "grad_norm": 1.1803417205810547, + "learning_rate": 1.3216589717718175e-05, + "loss": 0.6187, + "step": 10304 + }, + { + "epoch": 0.45626113810699226, + "grad_norm": 1.3222867250442505, + "learning_rate": 1.3215126360459593e-05, + "loss": 0.7821, + "step": 10305 + }, + { + "epoch": 0.4563054138118061, + "grad_norm": 1.3805031776428223, + "learning_rate": 1.321366292641034e-05, + "loss": 0.7397, + "step": 10306 + }, + { + "epoch": 0.45634968951661997, + "grad_norm": 1.2463958263397217, + "learning_rate": 1.3212199415605365e-05, + "loss": 0.6393, + "step": 10307 + }, + { + "epoch": 0.4563939652214339, + "grad_norm": 1.3398406505584717, + "learning_rate": 1.3210735828079619e-05, + "loss": 0.631, + "step": 10308 + }, + { + "epoch": 0.45643824092624774, + "grad_norm": 1.601682186126709, + "learning_rate": 1.3209272163868065e-05, + "loss": 0.7236, + "step": 10309 + }, + { + "epoch": 0.4564825166310616, + "grad_norm": 1.9015223979949951, + "learning_rate": 1.3207808423005656e-05, + "loss": 0.7463, + "step": 10310 + }, + { + "epoch": 0.4565267923358755, + "grad_norm": 1.8386850357055664, + "learning_rate": 1.3206344605527355e-05, + "loss": 0.7751, + "step": 10311 + }, + { + "epoch": 0.45657106804068937, + "grad_norm": 1.3823535442352295, + "learning_rate": 1.3204880711468126e-05, + "loss": 0.7054, + "step": 10312 + }, + { + "epoch": 0.4566153437455032, + "grad_norm": 3.014951705932617, + "learning_rate": 1.320341674086293e-05, + "loss": 0.7152, + "step": 10313 + }, + { + "epoch": 0.45665961945031713, + "grad_norm": 1.4156782627105713, + "learning_rate": 1.3201952693746733e-05, + "loss": 0.7613, + "step": 10314 + }, + { + "epoch": 0.456703895155131, + "grad_norm": 2.277973175048828, + "learning_rate": 1.3200488570154501e-05, + "loss": 0.8, + "step": 10315 + }, + { + "epoch": 0.4567481708599449, + "grad_norm": 1.417496681213379, + "learning_rate": 1.3199024370121206e-05, + "loss": 0.7256, + "step": 10316 + }, + { + "epoch": 0.45679244656475876, + "grad_norm": 1.6155054569244385, + "learning_rate": 1.3197560093681823e-05, + "loss": 0.7244, + "step": 10317 + }, + { + "epoch": 0.4568367222695726, + "grad_norm": 1.4105112552642822, + "learning_rate": 1.319609574087132e-05, + "loss": 0.6641, + "step": 10318 + }, + { + "epoch": 0.4568809979743865, + "grad_norm": 1.281198501586914, + "learning_rate": 1.3194631311724671e-05, + "loss": 0.6683, + "step": 10319 + }, + { + "epoch": 0.4569252736792004, + "grad_norm": 1.234262466430664, + "learning_rate": 1.3193166806276852e-05, + "loss": 0.6413, + "step": 10320 + }, + { + "epoch": 0.45696954938401424, + "grad_norm": 1.2841253280639648, + "learning_rate": 1.3191702224562844e-05, + "loss": 0.7305, + "step": 10321 + }, + { + "epoch": 0.45701382508882815, + "grad_norm": 1.3875542879104614, + "learning_rate": 1.3190237566617627e-05, + "loss": 0.6464, + "step": 10322 + }, + { + "epoch": 0.457058100793642, + "grad_norm": 1.5230085849761963, + "learning_rate": 1.318877283247619e-05, + "loss": 0.6759, + "step": 10323 + }, + { + "epoch": 0.45710237649845586, + "grad_norm": 1.3817932605743408, + "learning_rate": 1.3187308022173503e-05, + "loss": 0.6872, + "step": 10324 + }, + { + "epoch": 0.4571466522032698, + "grad_norm": 1.2488782405853271, + "learning_rate": 1.3185843135744561e-05, + "loss": 0.6694, + "step": 10325 + }, + { + "epoch": 0.45719092790808363, + "grad_norm": 1.397422432899475, + "learning_rate": 1.3184378173224348e-05, + "loss": 0.668, + "step": 10326 + }, + { + "epoch": 0.4572352036128975, + "grad_norm": 1.4200066328048706, + "learning_rate": 1.3182913134647862e-05, + "loss": 0.625, + "step": 10327 + }, + { + "epoch": 0.4572794793177114, + "grad_norm": 1.5204936265945435, + "learning_rate": 1.318144802005008e-05, + "loss": 0.6827, + "step": 10328 + }, + { + "epoch": 0.45732375502252526, + "grad_norm": 1.3894261121749878, + "learning_rate": 1.3179982829466006e-05, + "loss": 0.7436, + "step": 10329 + }, + { + "epoch": 0.4573680307273391, + "grad_norm": 1.8920987844467163, + "learning_rate": 1.3178517562930628e-05, + "loss": 0.655, + "step": 10330 + }, + { + "epoch": 0.457412306432153, + "grad_norm": 1.6167833805084229, + "learning_rate": 1.317705222047895e-05, + "loss": 0.6964, + "step": 10331 + }, + { + "epoch": 0.4574565821369669, + "grad_norm": 1.3412330150604248, + "learning_rate": 1.3175586802145966e-05, + "loss": 0.6597, + "step": 10332 + }, + { + "epoch": 0.4575008578417808, + "grad_norm": 1.1703203916549683, + "learning_rate": 1.3174121307966674e-05, + "loss": 0.6498, + "step": 10333 + }, + { + "epoch": 0.45754513354659465, + "grad_norm": 1.275383710861206, + "learning_rate": 1.3172655737976082e-05, + "loss": 0.6586, + "step": 10334 + }, + { + "epoch": 0.4575894092514085, + "grad_norm": 1.2581597566604614, + "learning_rate": 1.317119009220919e-05, + "loss": 0.7206, + "step": 10335 + }, + { + "epoch": 0.4576336849562224, + "grad_norm": 1.4288442134857178, + "learning_rate": 1.3169724370701004e-05, + "loss": 0.6818, + "step": 10336 + }, + { + "epoch": 0.4576779606610363, + "grad_norm": 1.7521634101867676, + "learning_rate": 1.3168258573486534e-05, + "loss": 0.7523, + "step": 10337 + }, + { + "epoch": 0.45772223636585013, + "grad_norm": 1.3895673751831055, + "learning_rate": 1.3166792700600785e-05, + "loss": 0.7256, + "step": 10338 + }, + { + "epoch": 0.45776651207066404, + "grad_norm": 1.3555983304977417, + "learning_rate": 1.3165326752078774e-05, + "loss": 0.6886, + "step": 10339 + }, + { + "epoch": 0.4578107877754779, + "grad_norm": 1.440750002861023, + "learning_rate": 1.3163860727955507e-05, + "loss": 0.7126, + "step": 10340 + }, + { + "epoch": 0.45785506348029176, + "grad_norm": 1.5995327234268188, + "learning_rate": 1.3162394628266007e-05, + "loss": 0.7457, + "step": 10341 + }, + { + "epoch": 0.45789933918510567, + "grad_norm": 1.666754961013794, + "learning_rate": 1.3160928453045285e-05, + "loss": 0.7194, + "step": 10342 + }, + { + "epoch": 0.4579436148899195, + "grad_norm": 1.3463140726089478, + "learning_rate": 1.3159462202328357e-05, + "loss": 0.7383, + "step": 10343 + }, + { + "epoch": 0.4579878905947334, + "grad_norm": 1.3200440406799316, + "learning_rate": 1.3157995876150252e-05, + "loss": 0.6176, + "step": 10344 + }, + { + "epoch": 0.4580321662995473, + "grad_norm": 1.30582594871521, + "learning_rate": 1.3156529474545983e-05, + "loss": 0.6713, + "step": 10345 + }, + { + "epoch": 0.45807644200436115, + "grad_norm": 1.2776869535446167, + "learning_rate": 1.3155062997550578e-05, + "loss": 0.7011, + "step": 10346 + }, + { + "epoch": 0.458120717709175, + "grad_norm": 1.5919946432113647, + "learning_rate": 1.3153596445199063e-05, + "loss": 0.6969, + "step": 10347 + }, + { + "epoch": 0.4581649934139889, + "grad_norm": 1.4018738269805908, + "learning_rate": 1.3152129817526466e-05, + "loss": 0.6323, + "step": 10348 + }, + { + "epoch": 0.4582092691188028, + "grad_norm": 1.2730135917663574, + "learning_rate": 1.3150663114567814e-05, + "loss": 0.6995, + "step": 10349 + }, + { + "epoch": 0.4582535448236167, + "grad_norm": 1.5988656282424927, + "learning_rate": 1.3149196336358139e-05, + "loss": 0.7546, + "step": 10350 + }, + { + "epoch": 0.45829782052843054, + "grad_norm": 1.3338900804519653, + "learning_rate": 1.3147729482932473e-05, + "loss": 0.6652, + "step": 10351 + }, + { + "epoch": 0.4583420962332444, + "grad_norm": 1.2349849939346313, + "learning_rate": 1.3146262554325853e-05, + "loss": 0.6513, + "step": 10352 + }, + { + "epoch": 0.4583863719380583, + "grad_norm": 3.246396064758301, + "learning_rate": 1.3144795550573311e-05, + "loss": 0.8683, + "step": 10353 + }, + { + "epoch": 0.45843064764287217, + "grad_norm": 1.726176142692566, + "learning_rate": 1.3143328471709889e-05, + "loss": 0.7256, + "step": 10354 + }, + { + "epoch": 0.458474923347686, + "grad_norm": 1.6575971841812134, + "learning_rate": 1.3141861317770628e-05, + "loss": 0.7375, + "step": 10355 + }, + { + "epoch": 0.45851919905249994, + "grad_norm": 1.3351171016693115, + "learning_rate": 1.3140394088790565e-05, + "loss": 0.6649, + "step": 10356 + }, + { + "epoch": 0.4585634747573138, + "grad_norm": 1.4582291841506958, + "learning_rate": 1.3138926784804747e-05, + "loss": 0.7045, + "step": 10357 + }, + { + "epoch": 0.45860775046212765, + "grad_norm": 1.3534677028656006, + "learning_rate": 1.3137459405848217e-05, + "loss": 0.801, + "step": 10358 + }, + { + "epoch": 0.45865202616694156, + "grad_norm": 1.3255864381790161, + "learning_rate": 1.3135991951956024e-05, + "loss": 0.6792, + "step": 10359 + }, + { + "epoch": 0.4586963018717554, + "grad_norm": 1.3153164386749268, + "learning_rate": 1.313452442316322e-05, + "loss": 0.7271, + "step": 10360 + }, + { + "epoch": 0.4587405775765693, + "grad_norm": 1.3171727657318115, + "learning_rate": 1.3133056819504848e-05, + "loss": 0.7051, + "step": 10361 + }, + { + "epoch": 0.4587848532813832, + "grad_norm": 1.7144299745559692, + "learning_rate": 1.3131589141015968e-05, + "loss": 0.767, + "step": 10362 + }, + { + "epoch": 0.45882912898619704, + "grad_norm": 1.4247685670852661, + "learning_rate": 1.313012138773163e-05, + "loss": 0.6993, + "step": 10363 + }, + { + "epoch": 0.4588734046910109, + "grad_norm": 1.2528226375579834, + "learning_rate": 1.3128653559686893e-05, + "loss": 0.637, + "step": 10364 + }, + { + "epoch": 0.4589176803958248, + "grad_norm": 1.233093023300171, + "learning_rate": 1.3127185656916811e-05, + "loss": 0.6624, + "step": 10365 + }, + { + "epoch": 0.45896195610063867, + "grad_norm": 1.3259577751159668, + "learning_rate": 1.3125717679456447e-05, + "loss": 0.7264, + "step": 10366 + }, + { + "epoch": 0.4590062318054526, + "grad_norm": 2.033651828765869, + "learning_rate": 1.3124249627340859e-05, + "loss": 0.699, + "step": 10367 + }, + { + "epoch": 0.45905050751026644, + "grad_norm": 1.6539981365203857, + "learning_rate": 1.3122781500605115e-05, + "loss": 0.7321, + "step": 10368 + }, + { + "epoch": 0.4590947832150803, + "grad_norm": 1.2449681758880615, + "learning_rate": 1.3121313299284277e-05, + "loss": 0.6447, + "step": 10369 + }, + { + "epoch": 0.4591390589198942, + "grad_norm": 1.5145424604415894, + "learning_rate": 1.3119845023413413e-05, + "loss": 0.705, + "step": 10370 + }, + { + "epoch": 0.45918333462470806, + "grad_norm": 1.472519874572754, + "learning_rate": 1.3118376673027588e-05, + "loss": 0.8204, + "step": 10371 + }, + { + "epoch": 0.4592276103295219, + "grad_norm": 1.3607197999954224, + "learning_rate": 1.311690824816188e-05, + "loss": 0.709, + "step": 10372 + }, + { + "epoch": 0.45927188603433583, + "grad_norm": 1.2389758825302124, + "learning_rate": 1.311543974885135e-05, + "loss": 0.6409, + "step": 10373 + }, + { + "epoch": 0.4593161617391497, + "grad_norm": 2.0883147716522217, + "learning_rate": 1.3113971175131085e-05, + "loss": 0.8043, + "step": 10374 + }, + { + "epoch": 0.45936043744396354, + "grad_norm": 1.691667079925537, + "learning_rate": 1.311250252703615e-05, + "loss": 0.7718, + "step": 10375 + }, + { + "epoch": 0.45940471314877745, + "grad_norm": 1.3553402423858643, + "learning_rate": 1.3111033804601627e-05, + "loss": 0.7047, + "step": 10376 + }, + { + "epoch": 0.4594489888535913, + "grad_norm": 1.4808924198150635, + "learning_rate": 1.3109565007862597e-05, + "loss": 0.7951, + "step": 10377 + }, + { + "epoch": 0.45949326455840517, + "grad_norm": 1.2799824476242065, + "learning_rate": 1.3108096136854134e-05, + "loss": 0.6551, + "step": 10378 + }, + { + "epoch": 0.4595375402632191, + "grad_norm": 1.396735429763794, + "learning_rate": 1.3106627191611333e-05, + "loss": 0.6876, + "step": 10379 + }, + { + "epoch": 0.45958181596803294, + "grad_norm": 1.4472023248672485, + "learning_rate": 1.3105158172169263e-05, + "loss": 0.7468, + "step": 10380 + }, + { + "epoch": 0.4596260916728468, + "grad_norm": 1.369658350944519, + "learning_rate": 1.3103689078563022e-05, + "loss": 0.679, + "step": 10381 + }, + { + "epoch": 0.4596703673776607, + "grad_norm": 1.7498005628585815, + "learning_rate": 1.3102219910827694e-05, + "loss": 0.793, + "step": 10382 + }, + { + "epoch": 0.45971464308247456, + "grad_norm": 1.1736677885055542, + "learning_rate": 1.310075066899837e-05, + "loss": 0.6329, + "step": 10383 + }, + { + "epoch": 0.4597589187872884, + "grad_norm": 1.603232979774475, + "learning_rate": 1.3099281353110141e-05, + "loss": 0.7026, + "step": 10384 + }, + { + "epoch": 0.45980319449210233, + "grad_norm": 1.5781534910202026, + "learning_rate": 1.3097811963198099e-05, + "loss": 0.675, + "step": 10385 + }, + { + "epoch": 0.4598474701969162, + "grad_norm": 1.3566545248031616, + "learning_rate": 1.309634249929734e-05, + "loss": 0.68, + "step": 10386 + }, + { + "epoch": 0.4598917459017301, + "grad_norm": 1.4395148754119873, + "learning_rate": 1.3094872961442964e-05, + "loss": 0.684, + "step": 10387 + }, + { + "epoch": 0.45993602160654395, + "grad_norm": 1.41295325756073, + "learning_rate": 1.3093403349670067e-05, + "loss": 0.8025, + "step": 10388 + }, + { + "epoch": 0.4599802973113578, + "grad_norm": 1.5551238059997559, + "learning_rate": 1.309193366401375e-05, + "loss": 0.631, + "step": 10389 + }, + { + "epoch": 0.4600245730161717, + "grad_norm": 1.4140344858169556, + "learning_rate": 1.3090463904509111e-05, + "loss": 0.6722, + "step": 10390 + }, + { + "epoch": 0.4600688487209856, + "grad_norm": 1.4415384531021118, + "learning_rate": 1.3088994071191264e-05, + "loss": 0.7331, + "step": 10391 + }, + { + "epoch": 0.46011312442579944, + "grad_norm": 1.13728928565979, + "learning_rate": 1.3087524164095302e-05, + "loss": 0.5904, + "step": 10392 + }, + { + "epoch": 0.46015740013061335, + "grad_norm": 1.3159310817718506, + "learning_rate": 1.3086054183256345e-05, + "loss": 0.7133, + "step": 10393 + }, + { + "epoch": 0.4602016758354272, + "grad_norm": 1.3899544477462769, + "learning_rate": 1.3084584128709494e-05, + "loss": 0.7409, + "step": 10394 + }, + { + "epoch": 0.46024595154024106, + "grad_norm": 1.3967143297195435, + "learning_rate": 1.3083114000489863e-05, + "loss": 0.6588, + "step": 10395 + }, + { + "epoch": 0.460290227245055, + "grad_norm": 1.5482286214828491, + "learning_rate": 1.3081643798632566e-05, + "loss": 0.7178, + "step": 10396 + }, + { + "epoch": 0.46033450294986883, + "grad_norm": 1.2986723184585571, + "learning_rate": 1.3080173523172712e-05, + "loss": 0.7009, + "step": 10397 + }, + { + "epoch": 0.4603787786546827, + "grad_norm": 1.464083194732666, + "learning_rate": 1.3078703174145426e-05, + "loss": 0.6636, + "step": 10398 + }, + { + "epoch": 0.4604230543594966, + "grad_norm": 1.2282931804656982, + "learning_rate": 1.307723275158582e-05, + "loss": 0.6491, + "step": 10399 + }, + { + "epoch": 0.46046733006431045, + "grad_norm": 1.2179691791534424, + "learning_rate": 1.3075762255529014e-05, + "loss": 0.6416, + "step": 10400 + }, + { + "epoch": 0.4605116057691243, + "grad_norm": 1.385947823524475, + "learning_rate": 1.3074291686010132e-05, + "loss": 0.7293, + "step": 10401 + }, + { + "epoch": 0.4605558814739382, + "grad_norm": 1.4594266414642334, + "learning_rate": 1.3072821043064297e-05, + "loss": 0.7511, + "step": 10402 + }, + { + "epoch": 0.4606001571787521, + "grad_norm": 1.7038506269454956, + "learning_rate": 1.307135032672663e-05, + "loss": 0.7342, + "step": 10403 + }, + { + "epoch": 0.460644432883566, + "grad_norm": 1.6906054019927979, + "learning_rate": 1.3069879537032266e-05, + "loss": 0.702, + "step": 10404 + }, + { + "epoch": 0.46068870858837985, + "grad_norm": 1.5809489488601685, + "learning_rate": 1.3068408674016326e-05, + "loss": 0.6925, + "step": 10405 + }, + { + "epoch": 0.4607329842931937, + "grad_norm": 1.4595632553100586, + "learning_rate": 1.3066937737713943e-05, + "loss": 0.7193, + "step": 10406 + }, + { + "epoch": 0.4607772599980076, + "grad_norm": 1.3284634351730347, + "learning_rate": 1.3065466728160253e-05, + "loss": 0.6876, + "step": 10407 + }, + { + "epoch": 0.46082153570282147, + "grad_norm": 1.4580302238464355, + "learning_rate": 1.3063995645390381e-05, + "loss": 0.734, + "step": 10408 + }, + { + "epoch": 0.46086581140763533, + "grad_norm": 1.2792847156524658, + "learning_rate": 1.306252448943947e-05, + "loss": 0.6986, + "step": 10409 + }, + { + "epoch": 0.46091008711244924, + "grad_norm": 1.2432085275650024, + "learning_rate": 1.3061053260342655e-05, + "loss": 0.6644, + "step": 10410 + }, + { + "epoch": 0.4609543628172631, + "grad_norm": 1.4058305025100708, + "learning_rate": 1.3059581958135075e-05, + "loss": 0.6589, + "step": 10411 + }, + { + "epoch": 0.46099863852207695, + "grad_norm": 1.427805781364441, + "learning_rate": 1.3058110582851872e-05, + "loss": 0.7057, + "step": 10412 + }, + { + "epoch": 0.46104291422689087, + "grad_norm": 1.4512540102005005, + "learning_rate": 1.3056639134528186e-05, + "loss": 0.6967, + "step": 10413 + }, + { + "epoch": 0.4610871899317047, + "grad_norm": 1.7754058837890625, + "learning_rate": 1.3055167613199166e-05, + "loss": 0.8672, + "step": 10414 + }, + { + "epoch": 0.4611314656365186, + "grad_norm": 1.3364624977111816, + "learning_rate": 1.3053696018899954e-05, + "loss": 0.6447, + "step": 10415 + }, + { + "epoch": 0.4611757413413325, + "grad_norm": 1.417125940322876, + "learning_rate": 1.3052224351665697e-05, + "loss": 0.6628, + "step": 10416 + }, + { + "epoch": 0.46122001704614635, + "grad_norm": 1.2921714782714844, + "learning_rate": 1.305075261153155e-05, + "loss": 0.6291, + "step": 10417 + }, + { + "epoch": 0.4612642927509602, + "grad_norm": 1.4851382970809937, + "learning_rate": 1.3049280798532658e-05, + "loss": 0.7337, + "step": 10418 + }, + { + "epoch": 0.4613085684557741, + "grad_norm": 1.2266451120376587, + "learning_rate": 1.3047808912704178e-05, + "loss": 0.6531, + "step": 10419 + }, + { + "epoch": 0.46135284416058797, + "grad_norm": 1.5207133293151855, + "learning_rate": 1.3046336954081262e-05, + "loss": 0.7267, + "step": 10420 + }, + { + "epoch": 0.4613971198654019, + "grad_norm": 1.4326913356781006, + "learning_rate": 1.3044864922699072e-05, + "loss": 0.7249, + "step": 10421 + }, + { + "epoch": 0.46144139557021574, + "grad_norm": 1.1687110662460327, + "learning_rate": 1.3043392818592762e-05, + "loss": 0.7054, + "step": 10422 + }, + { + "epoch": 0.4614856712750296, + "grad_norm": 1.5407317876815796, + "learning_rate": 1.3041920641797487e-05, + "loss": 0.8329, + "step": 10423 + }, + { + "epoch": 0.4615299469798435, + "grad_norm": 1.5698257684707642, + "learning_rate": 1.3040448392348422e-05, + "loss": 0.7244, + "step": 10424 + }, + { + "epoch": 0.46157422268465736, + "grad_norm": 1.349911093711853, + "learning_rate": 1.3038976070280715e-05, + "loss": 0.6663, + "step": 10425 + }, + { + "epoch": 0.4616184983894712, + "grad_norm": 1.5211836099624634, + "learning_rate": 1.3037503675629549e-05, + "loss": 0.728, + "step": 10426 + }, + { + "epoch": 0.46166277409428513, + "grad_norm": 1.3354718685150146, + "learning_rate": 1.3036031208430073e-05, + "loss": 0.6604, + "step": 10427 + }, + { + "epoch": 0.461707049799099, + "grad_norm": 1.4180840253829956, + "learning_rate": 1.3034558668717466e-05, + "loss": 0.7071, + "step": 10428 + }, + { + "epoch": 0.46175132550391285, + "grad_norm": 1.3584414720535278, + "learning_rate": 1.3033086056526893e-05, + "loss": 0.6601, + "step": 10429 + }, + { + "epoch": 0.46179560120872676, + "grad_norm": 1.5874030590057373, + "learning_rate": 1.3031613371893532e-05, + "loss": 0.7546, + "step": 10430 + }, + { + "epoch": 0.4618398769135406, + "grad_norm": 1.463793396949768, + "learning_rate": 1.3030140614852558e-05, + "loss": 0.6999, + "step": 10431 + }, + { + "epoch": 0.46188415261835447, + "grad_norm": 1.2914243936538696, + "learning_rate": 1.3028667785439138e-05, + "loss": 0.6252, + "step": 10432 + }, + { + "epoch": 0.4619284283231684, + "grad_norm": 1.1700831651687622, + "learning_rate": 1.3027194883688452e-05, + "loss": 0.6032, + "step": 10433 + }, + { + "epoch": 0.46197270402798224, + "grad_norm": 1.4482375383377075, + "learning_rate": 1.3025721909635684e-05, + "loss": 0.7006, + "step": 10434 + }, + { + "epoch": 0.4620169797327961, + "grad_norm": 1.438731074333191, + "learning_rate": 1.3024248863316012e-05, + "loss": 0.7901, + "step": 10435 + }, + { + "epoch": 0.46206125543761, + "grad_norm": 1.507980227470398, + "learning_rate": 1.3022775744764617e-05, + "loss": 0.8229, + "step": 10436 + }, + { + "epoch": 0.46210553114242386, + "grad_norm": 1.5570002794265747, + "learning_rate": 1.3021302554016686e-05, + "loss": 0.7041, + "step": 10437 + }, + { + "epoch": 0.4621498068472378, + "grad_norm": 1.5414080619812012, + "learning_rate": 1.3019829291107402e-05, + "loss": 0.6593, + "step": 10438 + }, + { + "epoch": 0.46219408255205163, + "grad_norm": 1.486786127090454, + "learning_rate": 1.3018355956071955e-05, + "loss": 0.6969, + "step": 10439 + }, + { + "epoch": 0.4622383582568655, + "grad_norm": 1.2395378351211548, + "learning_rate": 1.3016882548945536e-05, + "loss": 0.6315, + "step": 10440 + }, + { + "epoch": 0.4622826339616794, + "grad_norm": 1.2701753377914429, + "learning_rate": 1.3015409069763331e-05, + "loss": 0.6716, + "step": 10441 + }, + { + "epoch": 0.46232690966649326, + "grad_norm": 1.4680867195129395, + "learning_rate": 1.3013935518560535e-05, + "loss": 0.6719, + "step": 10442 + }, + { + "epoch": 0.4623711853713071, + "grad_norm": 1.472811222076416, + "learning_rate": 1.3012461895372343e-05, + "loss": 0.6811, + "step": 10443 + }, + { + "epoch": 0.462415461076121, + "grad_norm": 1.4472707509994507, + "learning_rate": 1.3010988200233954e-05, + "loss": 0.709, + "step": 10444 + }, + { + "epoch": 0.4624597367809349, + "grad_norm": 1.5402601957321167, + "learning_rate": 1.3009514433180562e-05, + "loss": 0.6548, + "step": 10445 + }, + { + "epoch": 0.46250401248574874, + "grad_norm": 1.3952523469924927, + "learning_rate": 1.3008040594247368e-05, + "loss": 0.6896, + "step": 10446 + }, + { + "epoch": 0.46254828819056265, + "grad_norm": 1.3409696817398071, + "learning_rate": 1.3006566683469574e-05, + "loss": 0.7262, + "step": 10447 + }, + { + "epoch": 0.4625925638953765, + "grad_norm": 1.3494395017623901, + "learning_rate": 1.3005092700882383e-05, + "loss": 0.6645, + "step": 10448 + }, + { + "epoch": 0.46263683960019036, + "grad_norm": 1.492202877998352, + "learning_rate": 1.3003618646520999e-05, + "loss": 0.6931, + "step": 10449 + }, + { + "epoch": 0.4626811153050043, + "grad_norm": 1.2827225923538208, + "learning_rate": 1.300214452042063e-05, + "loss": 0.6898, + "step": 10450 + }, + { + "epoch": 0.46272539100981813, + "grad_norm": 1.6407498121261597, + "learning_rate": 1.3000670322616485e-05, + "loss": 0.7241, + "step": 10451 + }, + { + "epoch": 0.462769666714632, + "grad_norm": 1.3472360372543335, + "learning_rate": 1.299919605314377e-05, + "loss": 0.6655, + "step": 10452 + }, + { + "epoch": 0.4628139424194459, + "grad_norm": 1.4314459562301636, + "learning_rate": 1.2997721712037698e-05, + "loss": 0.7298, + "step": 10453 + }, + { + "epoch": 0.46285821812425976, + "grad_norm": 2.1214637756347656, + "learning_rate": 1.299624729933349e-05, + "loss": 0.8662, + "step": 10454 + }, + { + "epoch": 0.46290249382907367, + "grad_norm": 1.5832154750823975, + "learning_rate": 1.299477281506635e-05, + "loss": 0.7576, + "step": 10455 + }, + { + "epoch": 0.4629467695338875, + "grad_norm": 1.5790425539016724, + "learning_rate": 1.29932982592715e-05, + "loss": 0.7096, + "step": 10456 + }, + { + "epoch": 0.4629910452387014, + "grad_norm": 1.3477730751037598, + "learning_rate": 1.2991823631984158e-05, + "loss": 0.723, + "step": 10457 + }, + { + "epoch": 0.4630353209435153, + "grad_norm": 1.4397201538085938, + "learning_rate": 1.2990348933239547e-05, + "loss": 0.647, + "step": 10458 + }, + { + "epoch": 0.46307959664832915, + "grad_norm": 1.3689227104187012, + "learning_rate": 1.2988874163072884e-05, + "loss": 0.7061, + "step": 10459 + }, + { + "epoch": 0.463123872353143, + "grad_norm": 1.358031988143921, + "learning_rate": 1.2987399321519396e-05, + "loss": 0.6898, + "step": 10460 + }, + { + "epoch": 0.4631681480579569, + "grad_norm": 1.3718303442001343, + "learning_rate": 1.2985924408614307e-05, + "loss": 0.6704, + "step": 10461 + }, + { + "epoch": 0.4632124237627708, + "grad_norm": 1.3734304904937744, + "learning_rate": 1.2984449424392846e-05, + "loss": 0.7313, + "step": 10462 + }, + { + "epoch": 0.46325669946758463, + "grad_norm": 1.488847017288208, + "learning_rate": 1.2982974368890243e-05, + "loss": 0.7017, + "step": 10463 + }, + { + "epoch": 0.46330097517239854, + "grad_norm": 1.4786895513534546, + "learning_rate": 1.2981499242141722e-05, + "loss": 0.7097, + "step": 10464 + }, + { + "epoch": 0.4633452508772124, + "grad_norm": 1.5475246906280518, + "learning_rate": 1.298002404418252e-05, + "loss": 0.6908, + "step": 10465 + }, + { + "epoch": 0.46338952658202626, + "grad_norm": 1.394376516342163, + "learning_rate": 1.2978548775047872e-05, + "loss": 0.6443, + "step": 10466 + }, + { + "epoch": 0.46343380228684017, + "grad_norm": 1.2783335447311401, + "learning_rate": 1.2977073434773009e-05, + "loss": 0.6712, + "step": 10467 + }, + { + "epoch": 0.463478077991654, + "grad_norm": 1.5650750398635864, + "learning_rate": 1.2975598023393172e-05, + "loss": 0.7406, + "step": 10468 + }, + { + "epoch": 0.4635223536964679, + "grad_norm": 1.793637990951538, + "learning_rate": 1.29741225409436e-05, + "loss": 0.7, + "step": 10469 + }, + { + "epoch": 0.4635666294012818, + "grad_norm": 1.2227967977523804, + "learning_rate": 1.2972646987459532e-05, + "loss": 0.6608, + "step": 10470 + }, + { + "epoch": 0.46361090510609565, + "grad_norm": 1.556649923324585, + "learning_rate": 1.2971171362976211e-05, + "loss": 0.8338, + "step": 10471 + }, + { + "epoch": 0.4636551808109095, + "grad_norm": 1.3126394748687744, + "learning_rate": 1.2969695667528882e-05, + "loss": 0.7084, + "step": 10472 + }, + { + "epoch": 0.4636994565157234, + "grad_norm": 1.4904770851135254, + "learning_rate": 1.2968219901152787e-05, + "loss": 0.7256, + "step": 10473 + }, + { + "epoch": 0.4637437322205373, + "grad_norm": 1.1862497329711914, + "learning_rate": 1.296674406388318e-05, + "loss": 0.6574, + "step": 10474 + }, + { + "epoch": 0.4637880079253512, + "grad_norm": 1.5035802125930786, + "learning_rate": 1.2965268155755307e-05, + "loss": 0.6643, + "step": 10475 + }, + { + "epoch": 0.46383228363016504, + "grad_norm": 1.1997134685516357, + "learning_rate": 1.2963792176804417e-05, + "loss": 0.6604, + "step": 10476 + }, + { + "epoch": 0.4638765593349789, + "grad_norm": 1.3552314043045044, + "learning_rate": 1.2962316127065762e-05, + "loss": 0.6552, + "step": 10477 + }, + { + "epoch": 0.4639208350397928, + "grad_norm": 1.7323176860809326, + "learning_rate": 1.2960840006574603e-05, + "loss": 0.669, + "step": 10478 + }, + { + "epoch": 0.46396511074460667, + "grad_norm": 1.2503464221954346, + "learning_rate": 1.2959363815366189e-05, + "loss": 0.6456, + "step": 10479 + }, + { + "epoch": 0.4640093864494205, + "grad_norm": 1.4497195482254028, + "learning_rate": 1.295788755347578e-05, + "loss": 0.7231, + "step": 10480 + }, + { + "epoch": 0.46405366215423444, + "grad_norm": 1.4809893369674683, + "learning_rate": 1.2956411220938634e-05, + "loss": 0.6924, + "step": 10481 + }, + { + "epoch": 0.4640979378590483, + "grad_norm": 1.5598260164260864, + "learning_rate": 1.2954934817790014e-05, + "loss": 0.6952, + "step": 10482 + }, + { + "epoch": 0.46414221356386215, + "grad_norm": 2.3977582454681396, + "learning_rate": 1.295345834406518e-05, + "loss": 0.7988, + "step": 10483 + }, + { + "epoch": 0.46418648926867606, + "grad_norm": 1.297837734222412, + "learning_rate": 1.29519817997994e-05, + "loss": 0.7189, + "step": 10484 + }, + { + "epoch": 0.4642307649734899, + "grad_norm": 1.7217069864273071, + "learning_rate": 1.295050518502794e-05, + "loss": 0.7713, + "step": 10485 + }, + { + "epoch": 0.4642750406783038, + "grad_norm": 1.649266242980957, + "learning_rate": 1.2949028499786062e-05, + "loss": 0.6718, + "step": 10486 + }, + { + "epoch": 0.4643193163831177, + "grad_norm": 1.5012110471725464, + "learning_rate": 1.2947551744109044e-05, + "loss": 0.697, + "step": 10487 + }, + { + "epoch": 0.46436359208793154, + "grad_norm": 2.5004122257232666, + "learning_rate": 1.294607491803215e-05, + "loss": 0.7824, + "step": 10488 + }, + { + "epoch": 0.4644078677927454, + "grad_norm": 1.5879377126693726, + "learning_rate": 1.2944598021590657e-05, + "loss": 0.7548, + "step": 10489 + }, + { + "epoch": 0.4644521434975593, + "grad_norm": 1.686014175415039, + "learning_rate": 1.2943121054819834e-05, + "loss": 0.7275, + "step": 10490 + }, + { + "epoch": 0.46449641920237317, + "grad_norm": 1.32965886592865, + "learning_rate": 1.2941644017754964e-05, + "loss": 0.6666, + "step": 10491 + }, + { + "epoch": 0.4645406949071871, + "grad_norm": 1.4618594646453857, + "learning_rate": 1.2940166910431323e-05, + "loss": 0.7103, + "step": 10492 + }, + { + "epoch": 0.46458497061200094, + "grad_norm": 1.317990779876709, + "learning_rate": 1.2938689732884189e-05, + "loss": 0.656, + "step": 10493 + }, + { + "epoch": 0.4646292463168148, + "grad_norm": 1.3802028894424438, + "learning_rate": 1.293721248514884e-05, + "loss": 0.643, + "step": 10494 + }, + { + "epoch": 0.4646735220216287, + "grad_norm": 1.6268984079360962, + "learning_rate": 1.2935735167260565e-05, + "loss": 0.7842, + "step": 10495 + }, + { + "epoch": 0.46471779772644256, + "grad_norm": 1.5251240730285645, + "learning_rate": 1.2934257779254646e-05, + "loss": 0.7454, + "step": 10496 + }, + { + "epoch": 0.4647620734312564, + "grad_norm": 1.6520212888717651, + "learning_rate": 1.293278032116637e-05, + "loss": 0.7788, + "step": 10497 + }, + { + "epoch": 0.46480634913607033, + "grad_norm": 1.4817190170288086, + "learning_rate": 1.2931302793031022e-05, + "loss": 0.7225, + "step": 10498 + }, + { + "epoch": 0.4648506248408842, + "grad_norm": 1.343042016029358, + "learning_rate": 1.2929825194883893e-05, + "loss": 0.6995, + "step": 10499 + }, + { + "epoch": 0.46489490054569804, + "grad_norm": 1.3599040508270264, + "learning_rate": 1.2928347526760274e-05, + "loss": 0.6281, + "step": 10500 + }, + { + "epoch": 0.46493917625051195, + "grad_norm": 1.7797784805297852, + "learning_rate": 1.2926869788695462e-05, + "loss": 0.7388, + "step": 10501 + }, + { + "epoch": 0.4649834519553258, + "grad_norm": 1.1971813440322876, + "learning_rate": 1.2925391980724749e-05, + "loss": 0.5866, + "step": 10502 + }, + { + "epoch": 0.46502772766013967, + "grad_norm": 1.3780944347381592, + "learning_rate": 1.2923914102883429e-05, + "loss": 0.7292, + "step": 10503 + }, + { + "epoch": 0.4650720033649536, + "grad_norm": 1.4702470302581787, + "learning_rate": 1.2922436155206799e-05, + "loss": 0.5985, + "step": 10504 + }, + { + "epoch": 0.46511627906976744, + "grad_norm": 1.118541955947876, + "learning_rate": 1.2920958137730162e-05, + "loss": 0.6278, + "step": 10505 + }, + { + "epoch": 0.4651605547745813, + "grad_norm": 1.3722811937332153, + "learning_rate": 1.2919480050488821e-05, + "loss": 0.6506, + "step": 10506 + }, + { + "epoch": 0.4652048304793952, + "grad_norm": 1.421803593635559, + "learning_rate": 1.2918001893518073e-05, + "loss": 0.7379, + "step": 10507 + }, + { + "epoch": 0.46524910618420906, + "grad_norm": 1.4537222385406494, + "learning_rate": 1.2916523666853227e-05, + "loss": 0.6864, + "step": 10508 + }, + { + "epoch": 0.46529338188902297, + "grad_norm": 1.4387751817703247, + "learning_rate": 1.2915045370529586e-05, + "loss": 0.6911, + "step": 10509 + }, + { + "epoch": 0.46533765759383683, + "grad_norm": 1.3715252876281738, + "learning_rate": 1.2913567004582463e-05, + "loss": 0.6555, + "step": 10510 + }, + { + "epoch": 0.4653819332986507, + "grad_norm": 1.334947109222412, + "learning_rate": 1.2912088569047163e-05, + "loss": 0.674, + "step": 10511 + }, + { + "epoch": 0.4654262090034646, + "grad_norm": 1.3741225004196167, + "learning_rate": 1.2910610063958999e-05, + "loss": 0.6667, + "step": 10512 + }, + { + "epoch": 0.46547048470827845, + "grad_norm": 1.6331127882003784, + "learning_rate": 1.2909131489353283e-05, + "loss": 0.671, + "step": 10513 + }, + { + "epoch": 0.4655147604130923, + "grad_norm": 1.4572962522506714, + "learning_rate": 1.290765284526533e-05, + "loss": 0.6814, + "step": 10514 + }, + { + "epoch": 0.4655590361179062, + "grad_norm": 1.4843212366104126, + "learning_rate": 1.2906174131730458e-05, + "loss": 0.6981, + "step": 10515 + }, + { + "epoch": 0.4656033118227201, + "grad_norm": 1.3642957210540771, + "learning_rate": 1.290469534878398e-05, + "loss": 0.7073, + "step": 10516 + }, + { + "epoch": 0.46564758752753393, + "grad_norm": 1.4604943990707397, + "learning_rate": 1.2903216496461219e-05, + "loss": 0.7306, + "step": 10517 + }, + { + "epoch": 0.46569186323234785, + "grad_norm": 1.5409705638885498, + "learning_rate": 1.29017375747975e-05, + "loss": 0.7375, + "step": 10518 + }, + { + "epoch": 0.4657361389371617, + "grad_norm": 1.3043878078460693, + "learning_rate": 1.2900258583828138e-05, + "loss": 0.7287, + "step": 10519 + }, + { + "epoch": 0.46578041464197556, + "grad_norm": 1.173130989074707, + "learning_rate": 1.2898779523588463e-05, + "loss": 0.6361, + "step": 10520 + }, + { + "epoch": 0.46582469034678947, + "grad_norm": 1.2888634204864502, + "learning_rate": 1.28973003941138e-05, + "loss": 0.7112, + "step": 10521 + }, + { + "epoch": 0.46586896605160333, + "grad_norm": 1.21556556224823, + "learning_rate": 1.2895821195439471e-05, + "loss": 0.6298, + "step": 10522 + }, + { + "epoch": 0.4659132417564172, + "grad_norm": 1.3251163959503174, + "learning_rate": 1.2894341927600815e-05, + "loss": 0.7144, + "step": 10523 + }, + { + "epoch": 0.4659575174612311, + "grad_norm": 1.5632623434066772, + "learning_rate": 1.2892862590633158e-05, + "loss": 0.6705, + "step": 10524 + }, + { + "epoch": 0.46600179316604495, + "grad_norm": 1.2832746505737305, + "learning_rate": 1.2891383184571834e-05, + "loss": 0.6842, + "step": 10525 + }, + { + "epoch": 0.46604606887085887, + "grad_norm": 1.6355763673782349, + "learning_rate": 1.2889903709452176e-05, + "loss": 0.7351, + "step": 10526 + }, + { + "epoch": 0.4660903445756727, + "grad_norm": 1.5618853569030762, + "learning_rate": 1.288842416530952e-05, + "loss": 0.7308, + "step": 10527 + }, + { + "epoch": 0.4661346202804866, + "grad_norm": 1.2208219766616821, + "learning_rate": 1.2886944552179209e-05, + "loss": 0.6238, + "step": 10528 + }, + { + "epoch": 0.4661788959853005, + "grad_norm": 1.5024317502975464, + "learning_rate": 1.288546487009657e-05, + "loss": 0.7272, + "step": 10529 + }, + { + "epoch": 0.46622317169011435, + "grad_norm": 1.3754422664642334, + "learning_rate": 1.288398511909696e-05, + "loss": 0.729, + "step": 10530 + }, + { + "epoch": 0.4662674473949282, + "grad_norm": 1.3726868629455566, + "learning_rate": 1.2882505299215711e-05, + "loss": 0.6796, + "step": 10531 + }, + { + "epoch": 0.4663117230997421, + "grad_norm": 1.6521689891815186, + "learning_rate": 1.2881025410488168e-05, + "loss": 0.7389, + "step": 10532 + }, + { + "epoch": 0.46635599880455597, + "grad_norm": 1.2943849563598633, + "learning_rate": 1.287954545294968e-05, + "loss": 0.6476, + "step": 10533 + }, + { + "epoch": 0.4664002745093698, + "grad_norm": 1.3107486963272095, + "learning_rate": 1.2878065426635595e-05, + "loss": 0.6325, + "step": 10534 + }, + { + "epoch": 0.46644455021418374, + "grad_norm": 1.3663803339004517, + "learning_rate": 1.287658533158126e-05, + "loss": 0.6213, + "step": 10535 + }, + { + "epoch": 0.4664888259189976, + "grad_norm": 2.0322279930114746, + "learning_rate": 1.2875105167822026e-05, + "loss": 0.7118, + "step": 10536 + }, + { + "epoch": 0.46653310162381145, + "grad_norm": 1.7200442552566528, + "learning_rate": 1.2873624935393245e-05, + "loss": 0.6925, + "step": 10537 + }, + { + "epoch": 0.46657737732862536, + "grad_norm": 1.2877432107925415, + "learning_rate": 1.2872144634330274e-05, + "loss": 0.6948, + "step": 10538 + }, + { + "epoch": 0.4666216530334392, + "grad_norm": 1.5379022359848022, + "learning_rate": 1.2870664264668467e-05, + "loss": 0.7291, + "step": 10539 + }, + { + "epoch": 0.4666659287382531, + "grad_norm": 1.5449705123901367, + "learning_rate": 1.2869183826443181e-05, + "loss": 0.7281, + "step": 10540 + }, + { + "epoch": 0.466710204443067, + "grad_norm": 1.4772365093231201, + "learning_rate": 1.2867703319689776e-05, + "loss": 0.7686, + "step": 10541 + }, + { + "epoch": 0.46675448014788085, + "grad_norm": 1.447332739830017, + "learning_rate": 1.286622274444361e-05, + "loss": 0.7077, + "step": 10542 + }, + { + "epoch": 0.46679875585269476, + "grad_norm": 1.2890268564224243, + "learning_rate": 1.2864742100740048e-05, + "loss": 0.612, + "step": 10543 + }, + { + "epoch": 0.4668430315575086, + "grad_norm": 1.3220735788345337, + "learning_rate": 1.2863261388614457e-05, + "loss": 0.6866, + "step": 10544 + }, + { + "epoch": 0.46688730726232247, + "grad_norm": 1.7110110521316528, + "learning_rate": 1.2861780608102195e-05, + "loss": 0.7115, + "step": 10545 + }, + { + "epoch": 0.4669315829671364, + "grad_norm": 1.3656927347183228, + "learning_rate": 1.2860299759238636e-05, + "loss": 0.66, + "step": 10546 + }, + { + "epoch": 0.46697585867195024, + "grad_norm": 1.9268141984939575, + "learning_rate": 1.2858818842059145e-05, + "loss": 0.791, + "step": 10547 + }, + { + "epoch": 0.4670201343767641, + "grad_norm": 1.1864538192749023, + "learning_rate": 1.2857337856599093e-05, + "loss": 0.5847, + "step": 10548 + }, + { + "epoch": 0.467064410081578, + "grad_norm": 1.1766941547393799, + "learning_rate": 1.2855856802893855e-05, + "loss": 0.6702, + "step": 10549 + }, + { + "epoch": 0.46710868578639186, + "grad_norm": 1.489682912826538, + "learning_rate": 1.2854375680978799e-05, + "loss": 0.7489, + "step": 10550 + }, + { + "epoch": 0.4671529614912057, + "grad_norm": 1.4296000003814697, + "learning_rate": 1.2852894490889307e-05, + "loss": 0.6623, + "step": 10551 + }, + { + "epoch": 0.46719723719601963, + "grad_norm": 1.4780356884002686, + "learning_rate": 1.2851413232660753e-05, + "loss": 0.6976, + "step": 10552 + }, + { + "epoch": 0.4672415129008335, + "grad_norm": 1.447644591331482, + "learning_rate": 1.2849931906328517e-05, + "loss": 0.647, + "step": 10553 + }, + { + "epoch": 0.46728578860564735, + "grad_norm": 1.3208872079849243, + "learning_rate": 1.2848450511927974e-05, + "loss": 0.6987, + "step": 10554 + }, + { + "epoch": 0.46733006431046126, + "grad_norm": 1.450739860534668, + "learning_rate": 1.284696904949451e-05, + "loss": 0.6765, + "step": 10555 + }, + { + "epoch": 0.4673743400152751, + "grad_norm": 1.595287561416626, + "learning_rate": 1.2845487519063512e-05, + "loss": 0.6881, + "step": 10556 + }, + { + "epoch": 0.46741861572008897, + "grad_norm": 1.2957675457000732, + "learning_rate": 1.2844005920670358e-05, + "loss": 0.6635, + "step": 10557 + }, + { + "epoch": 0.4674628914249029, + "grad_norm": 1.5206269025802612, + "learning_rate": 1.2842524254350442e-05, + "loss": 0.6934, + "step": 10558 + }, + { + "epoch": 0.46750716712971674, + "grad_norm": 1.3564120531082153, + "learning_rate": 1.2841042520139144e-05, + "loss": 0.6512, + "step": 10559 + }, + { + "epoch": 0.4675514428345306, + "grad_norm": 1.6425795555114746, + "learning_rate": 1.283956071807186e-05, + "loss": 0.7008, + "step": 10560 + }, + { + "epoch": 0.4675957185393445, + "grad_norm": 1.77897047996521, + "learning_rate": 1.2838078848183982e-05, + "loss": 0.6852, + "step": 10561 + }, + { + "epoch": 0.46763999424415836, + "grad_norm": 1.787919282913208, + "learning_rate": 1.2836596910510902e-05, + "loss": 0.817, + "step": 10562 + }, + { + "epoch": 0.4676842699489723, + "grad_norm": 1.2929861545562744, + "learning_rate": 1.2835114905088013e-05, + "loss": 0.7413, + "step": 10563 + }, + { + "epoch": 0.46772854565378613, + "grad_norm": 1.63895845413208, + "learning_rate": 1.2833632831950714e-05, + "loss": 0.8043, + "step": 10564 + }, + { + "epoch": 0.4677728213586, + "grad_norm": 1.6198445558547974, + "learning_rate": 1.2832150691134398e-05, + "loss": 0.7617, + "step": 10565 + }, + { + "epoch": 0.4678170970634139, + "grad_norm": 2.0600779056549072, + "learning_rate": 1.2830668482674472e-05, + "loss": 0.7477, + "step": 10566 + }, + { + "epoch": 0.46786137276822776, + "grad_norm": 1.3904476165771484, + "learning_rate": 1.2829186206606336e-05, + "loss": 0.7118, + "step": 10567 + }, + { + "epoch": 0.4679056484730416, + "grad_norm": 1.364341139793396, + "learning_rate": 1.282770386296539e-05, + "loss": 0.6736, + "step": 10568 + }, + { + "epoch": 0.4679499241778555, + "grad_norm": 1.5288608074188232, + "learning_rate": 1.2826221451787037e-05, + "loss": 0.7543, + "step": 10569 + }, + { + "epoch": 0.4679941998826694, + "grad_norm": 1.4909189939498901, + "learning_rate": 1.2824738973106688e-05, + "loss": 0.6789, + "step": 10570 + }, + { + "epoch": 0.46803847558748324, + "grad_norm": 1.5355504751205444, + "learning_rate": 1.282325642695975e-05, + "loss": 0.7806, + "step": 10571 + }, + { + "epoch": 0.46808275129229715, + "grad_norm": 1.5496242046356201, + "learning_rate": 1.2821773813381629e-05, + "loss": 0.7354, + "step": 10572 + }, + { + "epoch": 0.468127026997111, + "grad_norm": 1.2880477905273438, + "learning_rate": 1.2820291132407738e-05, + "loss": 0.7156, + "step": 10573 + }, + { + "epoch": 0.46817130270192486, + "grad_norm": 1.5370872020721436, + "learning_rate": 1.2818808384073488e-05, + "loss": 0.789, + "step": 10574 + }, + { + "epoch": 0.4682155784067388, + "grad_norm": 1.4310928583145142, + "learning_rate": 1.2817325568414299e-05, + "loss": 0.7355, + "step": 10575 + }, + { + "epoch": 0.46825985411155263, + "grad_norm": 1.407225251197815, + "learning_rate": 1.281584268546558e-05, + "loss": 0.6738, + "step": 10576 + }, + { + "epoch": 0.4683041298163665, + "grad_norm": 1.5757230520248413, + "learning_rate": 1.2814359735262752e-05, + "loss": 0.7249, + "step": 10577 + }, + { + "epoch": 0.4683484055211804, + "grad_norm": 1.639306902885437, + "learning_rate": 1.2812876717841234e-05, + "loss": 0.744, + "step": 10578 + }, + { + "epoch": 0.46839268122599426, + "grad_norm": 1.5413583517074585, + "learning_rate": 1.2811393633236444e-05, + "loss": 0.6766, + "step": 10579 + }, + { + "epoch": 0.46843695693080817, + "grad_norm": 1.2798235416412354, + "learning_rate": 1.2809910481483805e-05, + "loss": 0.6851, + "step": 10580 + }, + { + "epoch": 0.468481232635622, + "grad_norm": 2.3889482021331787, + "learning_rate": 1.2808427262618748e-05, + "loss": 0.9131, + "step": 10581 + }, + { + "epoch": 0.4685255083404359, + "grad_norm": 1.3680919408798218, + "learning_rate": 1.2806943976676687e-05, + "loss": 0.6861, + "step": 10582 + }, + { + "epoch": 0.4685697840452498, + "grad_norm": 1.3573534488677979, + "learning_rate": 1.2805460623693058e-05, + "loss": 0.6615, + "step": 10583 + }, + { + "epoch": 0.46861405975006365, + "grad_norm": 1.4964079856872559, + "learning_rate": 1.2803977203703284e-05, + "loss": 0.6952, + "step": 10584 + }, + { + "epoch": 0.4686583354548775, + "grad_norm": 1.6716504096984863, + "learning_rate": 1.2802493716742796e-05, + "loss": 0.8352, + "step": 10585 + }, + { + "epoch": 0.4687026111596914, + "grad_norm": 1.4506856203079224, + "learning_rate": 1.2801010162847031e-05, + "loss": 0.6845, + "step": 10586 + }, + { + "epoch": 0.4687468868645053, + "grad_norm": 1.4376450777053833, + "learning_rate": 1.2799526542051419e-05, + "loss": 0.7173, + "step": 10587 + }, + { + "epoch": 0.46879116256931913, + "grad_norm": 1.2852689027786255, + "learning_rate": 1.2798042854391395e-05, + "loss": 0.6966, + "step": 10588 + }, + { + "epoch": 0.46883543827413304, + "grad_norm": 1.715907335281372, + "learning_rate": 1.2796559099902393e-05, + "loss": 0.7029, + "step": 10589 + }, + { + "epoch": 0.4688797139789469, + "grad_norm": 1.4804633855819702, + "learning_rate": 1.2795075278619855e-05, + "loss": 0.695, + "step": 10590 + }, + { + "epoch": 0.46892398968376076, + "grad_norm": 1.2763924598693848, + "learning_rate": 1.279359139057922e-05, + "loss": 0.6901, + "step": 10591 + }, + { + "epoch": 0.46896826538857467, + "grad_norm": 1.6593799591064453, + "learning_rate": 1.2792107435815929e-05, + "loss": 0.745, + "step": 10592 + }, + { + "epoch": 0.4690125410933885, + "grad_norm": 1.1976803541183472, + "learning_rate": 1.2790623414365427e-05, + "loss": 0.7019, + "step": 10593 + }, + { + "epoch": 0.4690568167982024, + "grad_norm": 1.2456586360931396, + "learning_rate": 1.2789139326263153e-05, + "loss": 0.6621, + "step": 10594 + }, + { + "epoch": 0.4691010925030163, + "grad_norm": 1.4110127687454224, + "learning_rate": 1.2787655171544561e-05, + "loss": 0.7016, + "step": 10595 + }, + { + "epoch": 0.46914536820783015, + "grad_norm": 1.6847290992736816, + "learning_rate": 1.2786170950245095e-05, + "loss": 0.7606, + "step": 10596 + }, + { + "epoch": 0.46918964391264406, + "grad_norm": 1.7991951704025269, + "learning_rate": 1.2784686662400203e-05, + "loss": 0.8478, + "step": 10597 + }, + { + "epoch": 0.4692339196174579, + "grad_norm": 1.5614116191864014, + "learning_rate": 1.2783202308045338e-05, + "loss": 0.8295, + "step": 10598 + }, + { + "epoch": 0.4692781953222718, + "grad_norm": 1.5408225059509277, + "learning_rate": 1.278171788721595e-05, + "loss": 0.714, + "step": 10599 + }, + { + "epoch": 0.4693224710270857, + "grad_norm": 1.144385814666748, + "learning_rate": 1.27802333999475e-05, + "loss": 0.5947, + "step": 10600 + }, + { + "epoch": 0.46936674673189954, + "grad_norm": 1.5967427492141724, + "learning_rate": 1.2778748846275435e-05, + "loss": 0.7515, + "step": 10601 + }, + { + "epoch": 0.4694110224367134, + "grad_norm": 1.6992892026901245, + "learning_rate": 1.2777264226235216e-05, + "loss": 0.8259, + "step": 10602 + }, + { + "epoch": 0.4694552981415273, + "grad_norm": 1.3848633766174316, + "learning_rate": 1.2775779539862305e-05, + "loss": 0.7419, + "step": 10603 + }, + { + "epoch": 0.46949957384634117, + "grad_norm": 1.4498424530029297, + "learning_rate": 1.2774294787192156e-05, + "loss": 0.6682, + "step": 10604 + }, + { + "epoch": 0.469543849551155, + "grad_norm": 1.4927616119384766, + "learning_rate": 1.277280996826024e-05, + "loss": 0.6705, + "step": 10605 + }, + { + "epoch": 0.46958812525596894, + "grad_norm": 1.4921019077301025, + "learning_rate": 1.2771325083102014e-05, + "loss": 0.7457, + "step": 10606 + }, + { + "epoch": 0.4696324009607828, + "grad_norm": 1.362391710281372, + "learning_rate": 1.276984013175294e-05, + "loss": 0.6899, + "step": 10607 + }, + { + "epoch": 0.46967667666559665, + "grad_norm": 1.5160409212112427, + "learning_rate": 1.2768355114248493e-05, + "loss": 0.683, + "step": 10608 + }, + { + "epoch": 0.46972095237041056, + "grad_norm": 1.5097283124923706, + "learning_rate": 1.2766870030624143e-05, + "loss": 0.8048, + "step": 10609 + }, + { + "epoch": 0.4697652280752244, + "grad_norm": 1.529893159866333, + "learning_rate": 1.276538488091535e-05, + "loss": 0.6824, + "step": 10610 + }, + { + "epoch": 0.4698095037800383, + "grad_norm": 1.4343949556350708, + "learning_rate": 1.2763899665157591e-05, + "loss": 0.6977, + "step": 10611 + }, + { + "epoch": 0.4698537794848522, + "grad_norm": 2.1957054138183594, + "learning_rate": 1.276241438338634e-05, + "loss": 0.7747, + "step": 10612 + }, + { + "epoch": 0.46989805518966604, + "grad_norm": 1.6432459354400635, + "learning_rate": 1.276092903563707e-05, + "loss": 0.7428, + "step": 10613 + }, + { + "epoch": 0.46994233089447995, + "grad_norm": 1.2893314361572266, + "learning_rate": 1.275944362194526e-05, + "loss": 0.6811, + "step": 10614 + }, + { + "epoch": 0.4699866065992938, + "grad_norm": 1.5219906568527222, + "learning_rate": 1.2757958142346384e-05, + "loss": 0.6999, + "step": 10615 + }, + { + "epoch": 0.47003088230410767, + "grad_norm": 2.4585697650909424, + "learning_rate": 1.2756472596875923e-05, + "loss": 0.7328, + "step": 10616 + }, + { + "epoch": 0.4700751580089216, + "grad_norm": 1.6186186075210571, + "learning_rate": 1.2754986985569357e-05, + "loss": 0.6207, + "step": 10617 + }, + { + "epoch": 0.47011943371373544, + "grad_norm": 1.4629912376403809, + "learning_rate": 1.2753501308462173e-05, + "loss": 0.659, + "step": 10618 + }, + { + "epoch": 0.4701637094185493, + "grad_norm": 1.3836479187011719, + "learning_rate": 1.2752015565589852e-05, + "loss": 0.6964, + "step": 10619 + }, + { + "epoch": 0.4702079851233632, + "grad_norm": 1.2463756799697876, + "learning_rate": 1.2750529756987882e-05, + "loss": 0.7168, + "step": 10620 + }, + { + "epoch": 0.47025226082817706, + "grad_norm": 1.7629213333129883, + "learning_rate": 1.2749043882691742e-05, + "loss": 0.7477, + "step": 10621 + }, + { + "epoch": 0.4702965365329909, + "grad_norm": 1.6545659303665161, + "learning_rate": 1.274755794273693e-05, + "loss": 0.6939, + "step": 10622 + }, + { + "epoch": 0.47034081223780483, + "grad_norm": 1.284580111503601, + "learning_rate": 1.2746071937158937e-05, + "loss": 0.7134, + "step": 10623 + }, + { + "epoch": 0.4703850879426187, + "grad_norm": 1.795272946357727, + "learning_rate": 1.2744585865993248e-05, + "loss": 0.6597, + "step": 10624 + }, + { + "epoch": 0.47042936364743254, + "grad_norm": 1.2854803800582886, + "learning_rate": 1.2743099729275363e-05, + "loss": 0.5899, + "step": 10625 + }, + { + "epoch": 0.47047363935224645, + "grad_norm": 1.5101879835128784, + "learning_rate": 1.2741613527040773e-05, + "loss": 0.6997, + "step": 10626 + }, + { + "epoch": 0.4705179150570603, + "grad_norm": 1.6151551008224487, + "learning_rate": 1.2740127259324975e-05, + "loss": 0.6911, + "step": 10627 + }, + { + "epoch": 0.47056219076187417, + "grad_norm": 1.5141303539276123, + "learning_rate": 1.273864092616347e-05, + "loss": 0.7036, + "step": 10628 + }, + { + "epoch": 0.4706064664666881, + "grad_norm": 1.6134333610534668, + "learning_rate": 1.2737154527591757e-05, + "loss": 0.7493, + "step": 10629 + }, + { + "epoch": 0.47065074217150193, + "grad_norm": 1.5247973203659058, + "learning_rate": 1.2735668063645337e-05, + "loss": 0.7058, + "step": 10630 + }, + { + "epoch": 0.4706950178763158, + "grad_norm": 1.3785600662231445, + "learning_rate": 1.273418153435971e-05, + "loss": 0.6698, + "step": 10631 + }, + { + "epoch": 0.4707392935811297, + "grad_norm": 1.2975183725357056, + "learning_rate": 1.2732694939770385e-05, + "loss": 0.7178, + "step": 10632 + }, + { + "epoch": 0.47078356928594356, + "grad_norm": 1.4789841175079346, + "learning_rate": 1.2731208279912867e-05, + "loss": 0.6367, + "step": 10633 + }, + { + "epoch": 0.47082784499075747, + "grad_norm": 1.3812167644500732, + "learning_rate": 1.2729721554822662e-05, + "loss": 0.7315, + "step": 10634 + }, + { + "epoch": 0.47087212069557133, + "grad_norm": 1.439225435256958, + "learning_rate": 1.2728234764535283e-05, + "loss": 0.7004, + "step": 10635 + }, + { + "epoch": 0.4709163964003852, + "grad_norm": 1.4132590293884277, + "learning_rate": 1.2726747909086235e-05, + "loss": 0.7367, + "step": 10636 + }, + { + "epoch": 0.4709606721051991, + "grad_norm": 1.327354073524475, + "learning_rate": 1.2725260988511035e-05, + "loss": 0.6659, + "step": 10637 + }, + { + "epoch": 0.47100494781001295, + "grad_norm": 1.6799830198287964, + "learning_rate": 1.2723774002845194e-05, + "loss": 0.6408, + "step": 10638 + }, + { + "epoch": 0.4710492235148268, + "grad_norm": 1.364248514175415, + "learning_rate": 1.272228695212423e-05, + "loss": 0.6996, + "step": 10639 + }, + { + "epoch": 0.4710934992196407, + "grad_norm": 1.6336065530776978, + "learning_rate": 1.2720799836383657e-05, + "loss": 0.7646, + "step": 10640 + }, + { + "epoch": 0.4711377749244546, + "grad_norm": 1.7335619926452637, + "learning_rate": 1.2719312655658995e-05, + "loss": 0.7743, + "step": 10641 + }, + { + "epoch": 0.47118205062926843, + "grad_norm": 1.3861005306243896, + "learning_rate": 1.2717825409985767e-05, + "loss": 0.6721, + "step": 10642 + }, + { + "epoch": 0.47122632633408235, + "grad_norm": 1.3570612668991089, + "learning_rate": 1.2716338099399492e-05, + "loss": 0.655, + "step": 10643 + }, + { + "epoch": 0.4712706020388962, + "grad_norm": 1.3522465229034424, + "learning_rate": 1.271485072393569e-05, + "loss": 0.6875, + "step": 10644 + }, + { + "epoch": 0.47131487774371006, + "grad_norm": 1.4378575086593628, + "learning_rate": 1.2713363283629892e-05, + "loss": 0.7399, + "step": 10645 + }, + { + "epoch": 0.47135915344852397, + "grad_norm": 1.4186325073242188, + "learning_rate": 1.2711875778517619e-05, + "loss": 0.6744, + "step": 10646 + }, + { + "epoch": 0.4714034291533378, + "grad_norm": 1.8003554344177246, + "learning_rate": 1.2710388208634403e-05, + "loss": 0.6988, + "step": 10647 + }, + { + "epoch": 0.4714477048581517, + "grad_norm": 1.5997439622879028, + "learning_rate": 1.270890057401577e-05, + "loss": 0.6731, + "step": 10648 + }, + { + "epoch": 0.4714919805629656, + "grad_norm": 1.3395261764526367, + "learning_rate": 1.2707412874697254e-05, + "loss": 0.6833, + "step": 10649 + }, + { + "epoch": 0.47153625626777945, + "grad_norm": 1.4204744100570679, + "learning_rate": 1.2705925110714385e-05, + "loss": 0.7063, + "step": 10650 + }, + { + "epoch": 0.47158053197259336, + "grad_norm": 1.3374730348587036, + "learning_rate": 1.2704437282102698e-05, + "loss": 0.6811, + "step": 10651 + }, + { + "epoch": 0.4716248076774072, + "grad_norm": 1.6983829736709595, + "learning_rate": 1.270294938889773e-05, + "loss": 0.7098, + "step": 10652 + }, + { + "epoch": 0.4716690833822211, + "grad_norm": 1.7642492055892944, + "learning_rate": 1.2701461431135013e-05, + "loss": 0.7364, + "step": 10653 + }, + { + "epoch": 0.471713359087035, + "grad_norm": 1.4186856746673584, + "learning_rate": 1.2699973408850093e-05, + "loss": 0.7282, + "step": 10654 + }, + { + "epoch": 0.47175763479184885, + "grad_norm": 1.3594552278518677, + "learning_rate": 1.2698485322078504e-05, + "loss": 0.677, + "step": 10655 + }, + { + "epoch": 0.4718019104966627, + "grad_norm": 1.6258394718170166, + "learning_rate": 1.2696997170855792e-05, + "loss": 0.7676, + "step": 10656 + }, + { + "epoch": 0.4718461862014766, + "grad_norm": 1.354600429534912, + "learning_rate": 1.2695508955217497e-05, + "loss": 0.6714, + "step": 10657 + }, + { + "epoch": 0.47189046190629047, + "grad_norm": 1.3763829469680786, + "learning_rate": 1.2694020675199168e-05, + "loss": 0.6039, + "step": 10658 + }, + { + "epoch": 0.4719347376111043, + "grad_norm": 1.300813913345337, + "learning_rate": 1.2692532330836346e-05, + "loss": 0.6646, + "step": 10659 + }, + { + "epoch": 0.47197901331591824, + "grad_norm": 1.5679845809936523, + "learning_rate": 1.2691043922164584e-05, + "loss": 0.7256, + "step": 10660 + }, + { + "epoch": 0.4720232890207321, + "grad_norm": 1.3697164058685303, + "learning_rate": 1.2689555449219429e-05, + "loss": 0.6175, + "step": 10661 + }, + { + "epoch": 0.47206756472554595, + "grad_norm": 1.4225924015045166, + "learning_rate": 1.2688066912036429e-05, + "loss": 0.6875, + "step": 10662 + }, + { + "epoch": 0.47211184043035986, + "grad_norm": 1.531127691268921, + "learning_rate": 1.268657831065114e-05, + "loss": 0.7404, + "step": 10663 + }, + { + "epoch": 0.4721561161351737, + "grad_norm": 1.3993465900421143, + "learning_rate": 1.2685089645099118e-05, + "loss": 0.7246, + "step": 10664 + }, + { + "epoch": 0.4722003918399876, + "grad_norm": 1.1991300582885742, + "learning_rate": 1.2683600915415914e-05, + "loss": 0.6507, + "step": 10665 + }, + { + "epoch": 0.4722446675448015, + "grad_norm": 1.6154757738113403, + "learning_rate": 1.2682112121637089e-05, + "loss": 0.7524, + "step": 10666 + }, + { + "epoch": 0.47228894324961535, + "grad_norm": 1.436078429222107, + "learning_rate": 1.2680623263798198e-05, + "loss": 0.7343, + "step": 10667 + }, + { + "epoch": 0.47233321895442926, + "grad_norm": 1.3740676641464233, + "learning_rate": 1.2679134341934805e-05, + "loss": 0.7261, + "step": 10668 + }, + { + "epoch": 0.4723774946592431, + "grad_norm": 1.4611495733261108, + "learning_rate": 1.2677645356082468e-05, + "loss": 0.6955, + "step": 10669 + }, + { + "epoch": 0.47242177036405697, + "grad_norm": 1.9587905406951904, + "learning_rate": 1.2676156306276751e-05, + "loss": 0.8806, + "step": 10670 + }, + { + "epoch": 0.4724660460688709, + "grad_norm": 1.3139841556549072, + "learning_rate": 1.2674667192553226e-05, + "loss": 0.6843, + "step": 10671 + }, + { + "epoch": 0.47251032177368474, + "grad_norm": 1.8281488418579102, + "learning_rate": 1.2673178014947445e-05, + "loss": 0.7622, + "step": 10672 + }, + { + "epoch": 0.4725545974784986, + "grad_norm": 1.3131773471832275, + "learning_rate": 1.2671688773494988e-05, + "loss": 0.683, + "step": 10673 + }, + { + "epoch": 0.4725988731833125, + "grad_norm": 1.6816803216934204, + "learning_rate": 1.2670199468231417e-05, + "loss": 0.6897, + "step": 10674 + }, + { + "epoch": 0.47264314888812636, + "grad_norm": 1.707349419593811, + "learning_rate": 1.266871009919231e-05, + "loss": 0.6781, + "step": 10675 + }, + { + "epoch": 0.4726874245929402, + "grad_norm": 1.682820200920105, + "learning_rate": 1.2667220666413233e-05, + "loss": 0.7376, + "step": 10676 + }, + { + "epoch": 0.47273170029775413, + "grad_norm": 1.5605478286743164, + "learning_rate": 1.266573116992976e-05, + "loss": 0.7524, + "step": 10677 + }, + { + "epoch": 0.472775976002568, + "grad_norm": 1.580176830291748, + "learning_rate": 1.266424160977747e-05, + "loss": 0.7146, + "step": 10678 + }, + { + "epoch": 0.47282025170738184, + "grad_norm": 1.319318175315857, + "learning_rate": 1.2662751985991939e-05, + "loss": 0.6382, + "step": 10679 + }, + { + "epoch": 0.47286452741219576, + "grad_norm": 1.1911662817001343, + "learning_rate": 1.2661262298608748e-05, + "loss": 0.6799, + "step": 10680 + }, + { + "epoch": 0.4729088031170096, + "grad_norm": 1.9670658111572266, + "learning_rate": 1.2659772547663469e-05, + "loss": 0.7568, + "step": 10681 + }, + { + "epoch": 0.47295307882182347, + "grad_norm": 1.3915529251098633, + "learning_rate": 1.265828273319169e-05, + "loss": 0.6625, + "step": 10682 + }, + { + "epoch": 0.4729973545266374, + "grad_norm": 1.4717386960983276, + "learning_rate": 1.2656792855228993e-05, + "loss": 0.7231, + "step": 10683 + }, + { + "epoch": 0.47304163023145124, + "grad_norm": 1.2413538694381714, + "learning_rate": 1.265530291381096e-05, + "loss": 0.6529, + "step": 10684 + }, + { + "epoch": 0.47308590593626515, + "grad_norm": 1.4971740245819092, + "learning_rate": 1.2653812908973182e-05, + "loss": 0.6589, + "step": 10685 + }, + { + "epoch": 0.473130181641079, + "grad_norm": 1.5757651329040527, + "learning_rate": 1.2652322840751242e-05, + "loss": 0.6868, + "step": 10686 + }, + { + "epoch": 0.47317445734589286, + "grad_norm": 1.349043369293213, + "learning_rate": 1.2650832709180727e-05, + "loss": 0.6887, + "step": 10687 + }, + { + "epoch": 0.4732187330507068, + "grad_norm": 1.238922357559204, + "learning_rate": 1.2649342514297236e-05, + "loss": 0.6379, + "step": 10688 + }, + { + "epoch": 0.47326300875552063, + "grad_norm": 1.5249048471450806, + "learning_rate": 1.2647852256136356e-05, + "loss": 0.6978, + "step": 10689 + }, + { + "epoch": 0.4733072844603345, + "grad_norm": 1.4145399332046509, + "learning_rate": 1.264636193473368e-05, + "loss": 0.6425, + "step": 10690 + }, + { + "epoch": 0.4733515601651484, + "grad_norm": 1.3702038526535034, + "learning_rate": 1.2644871550124803e-05, + "loss": 0.7041, + "step": 10691 + }, + { + "epoch": 0.47339583586996226, + "grad_norm": 1.8299440145492554, + "learning_rate": 1.2643381102345322e-05, + "loss": 0.7976, + "step": 10692 + }, + { + "epoch": 0.4734401115747761, + "grad_norm": 1.346610188484192, + "learning_rate": 1.2641890591430836e-05, + "loss": 0.6758, + "step": 10693 + }, + { + "epoch": 0.47348438727959, + "grad_norm": 1.510079026222229, + "learning_rate": 1.2640400017416947e-05, + "loss": 0.6819, + "step": 10694 + }, + { + "epoch": 0.4735286629844039, + "grad_norm": 1.3674442768096924, + "learning_rate": 1.2638909380339252e-05, + "loss": 0.7035, + "step": 10695 + }, + { + "epoch": 0.47357293868921774, + "grad_norm": 1.3825222253799438, + "learning_rate": 1.2637418680233354e-05, + "loss": 0.6832, + "step": 10696 + }, + { + "epoch": 0.47361721439403165, + "grad_norm": 1.312540888786316, + "learning_rate": 1.2635927917134855e-05, + "loss": 0.7049, + "step": 10697 + }, + { + "epoch": 0.4736614900988455, + "grad_norm": 1.4526357650756836, + "learning_rate": 1.2634437091079371e-05, + "loss": 0.6785, + "step": 10698 + }, + { + "epoch": 0.47370576580365936, + "grad_norm": 1.3798587322235107, + "learning_rate": 1.2632946202102501e-05, + "loss": 0.5811, + "step": 10699 + }, + { + "epoch": 0.4737500415084733, + "grad_norm": 1.4939104318618774, + "learning_rate": 1.2631455250239853e-05, + "loss": 0.6508, + "step": 10700 + }, + { + "epoch": 0.47379431721328713, + "grad_norm": 1.326471209526062, + "learning_rate": 1.2629964235527038e-05, + "loss": 0.6595, + "step": 10701 + }, + { + "epoch": 0.47383859291810104, + "grad_norm": 1.2845430374145508, + "learning_rate": 1.2628473157999665e-05, + "loss": 0.6684, + "step": 10702 + }, + { + "epoch": 0.4738828686229149, + "grad_norm": 1.4886664152145386, + "learning_rate": 1.2626982017693358e-05, + "loss": 0.6771, + "step": 10703 + }, + { + "epoch": 0.47392714432772876, + "grad_norm": 1.3518213033676147, + "learning_rate": 1.2625490814643721e-05, + "loss": 0.6734, + "step": 10704 + }, + { + "epoch": 0.47397142003254267, + "grad_norm": 1.463276982307434, + "learning_rate": 1.2623999548886376e-05, + "loss": 0.6923, + "step": 10705 + }, + { + "epoch": 0.4740156957373565, + "grad_norm": 1.4562633037567139, + "learning_rate": 1.2622508220456937e-05, + "loss": 0.6919, + "step": 10706 + }, + { + "epoch": 0.4740599714421704, + "grad_norm": 1.339483618736267, + "learning_rate": 1.2621016829391022e-05, + "loss": 0.6273, + "step": 10707 + }, + { + "epoch": 0.4741042471469843, + "grad_norm": 2.1415796279907227, + "learning_rate": 1.2619525375724258e-05, + "loss": 0.8239, + "step": 10708 + }, + { + "epoch": 0.47414852285179815, + "grad_norm": 1.4183175563812256, + "learning_rate": 1.261803385949226e-05, + "loss": 0.73, + "step": 10709 + }, + { + "epoch": 0.474192798556612, + "grad_norm": 1.3395205736160278, + "learning_rate": 1.261654228073066e-05, + "loss": 0.6395, + "step": 10710 + }, + { + "epoch": 0.4742370742614259, + "grad_norm": 1.504334807395935, + "learning_rate": 1.2615050639475071e-05, + "loss": 0.6835, + "step": 10711 + }, + { + "epoch": 0.4742813499662398, + "grad_norm": 1.6094465255737305, + "learning_rate": 1.2613558935761132e-05, + "loss": 0.7292, + "step": 10712 + }, + { + "epoch": 0.47432562567105363, + "grad_norm": 1.5514612197875977, + "learning_rate": 1.2612067169624464e-05, + "loss": 0.6893, + "step": 10713 + }, + { + "epoch": 0.47436990137586754, + "grad_norm": 1.1438541412353516, + "learning_rate": 1.2610575341100697e-05, + "loss": 0.5911, + "step": 10714 + }, + { + "epoch": 0.4744141770806814, + "grad_norm": 1.4682468175888062, + "learning_rate": 1.2609083450225468e-05, + "loss": 0.7072, + "step": 10715 + }, + { + "epoch": 0.47445845278549525, + "grad_norm": 1.2730817794799805, + "learning_rate": 1.2607591497034402e-05, + "loss": 0.6625, + "step": 10716 + }, + { + "epoch": 0.47450272849030917, + "grad_norm": 1.1633048057556152, + "learning_rate": 1.2606099481563137e-05, + "loss": 0.6508, + "step": 10717 + }, + { + "epoch": 0.474547004195123, + "grad_norm": 1.1944977045059204, + "learning_rate": 1.2604607403847307e-05, + "loss": 0.6202, + "step": 10718 + }, + { + "epoch": 0.4745912798999369, + "grad_norm": 1.508711814880371, + "learning_rate": 1.260311526392255e-05, + "loss": 0.682, + "step": 10719 + }, + { + "epoch": 0.4746355556047508, + "grad_norm": 1.3474442958831787, + "learning_rate": 1.2601623061824504e-05, + "loss": 0.6721, + "step": 10720 + }, + { + "epoch": 0.47467983130956465, + "grad_norm": 1.3602277040481567, + "learning_rate": 1.260013079758881e-05, + "loss": 0.6395, + "step": 10721 + }, + { + "epoch": 0.47472410701437856, + "grad_norm": 1.5054044723510742, + "learning_rate": 1.259863847125111e-05, + "loss": 0.7394, + "step": 10722 + }, + { + "epoch": 0.4747683827191924, + "grad_norm": 1.546456217765808, + "learning_rate": 1.2597146082847047e-05, + "loss": 0.6567, + "step": 10723 + }, + { + "epoch": 0.4748126584240063, + "grad_norm": 1.5989086627960205, + "learning_rate": 1.2595653632412261e-05, + "loss": 0.6611, + "step": 10724 + }, + { + "epoch": 0.4748569341288202, + "grad_norm": 1.5532552003860474, + "learning_rate": 1.2594161119982404e-05, + "loss": 0.7337, + "step": 10725 + }, + { + "epoch": 0.47490120983363404, + "grad_norm": 1.2678166627883911, + "learning_rate": 1.2592668545593118e-05, + "loss": 0.5892, + "step": 10726 + }, + { + "epoch": 0.4749454855384479, + "grad_norm": 2.0177347660064697, + "learning_rate": 1.2591175909280061e-05, + "loss": 0.797, + "step": 10727 + }, + { + "epoch": 0.4749897612432618, + "grad_norm": 2.1010780334472656, + "learning_rate": 1.2589683211078873e-05, + "loss": 0.7776, + "step": 10728 + }, + { + "epoch": 0.47503403694807567, + "grad_norm": 1.715610146522522, + "learning_rate": 1.2588190451025209e-05, + "loss": 0.6617, + "step": 10729 + }, + { + "epoch": 0.4750783126528895, + "grad_norm": 1.6343423128128052, + "learning_rate": 1.2586697629154726e-05, + "loss": 0.7194, + "step": 10730 + }, + { + "epoch": 0.47512258835770343, + "grad_norm": 1.326344609260559, + "learning_rate": 1.2585204745503072e-05, + "loss": 0.6198, + "step": 10731 + }, + { + "epoch": 0.4751668640625173, + "grad_norm": 1.7037577629089355, + "learning_rate": 1.2583711800105914e-05, + "loss": 0.7761, + "step": 10732 + }, + { + "epoch": 0.47521113976733115, + "grad_norm": 1.457409381866455, + "learning_rate": 1.25822187929989e-05, + "loss": 0.6404, + "step": 10733 + }, + { + "epoch": 0.47525541547214506, + "grad_norm": 1.3232929706573486, + "learning_rate": 1.2580725724217691e-05, + "loss": 0.6245, + "step": 10734 + }, + { + "epoch": 0.4752996911769589, + "grad_norm": 1.4833061695098877, + "learning_rate": 1.2579232593797952e-05, + "loss": 0.7121, + "step": 10735 + }, + { + "epoch": 0.4753439668817728, + "grad_norm": 2.111541271209717, + "learning_rate": 1.2577739401775342e-05, + "loss": 0.7421, + "step": 10736 + }, + { + "epoch": 0.4753882425865867, + "grad_norm": 1.1263068914413452, + "learning_rate": 1.2576246148185527e-05, + "loss": 0.609, + "step": 10737 + }, + { + "epoch": 0.47543251829140054, + "grad_norm": 1.7455241680145264, + "learning_rate": 1.2574752833064165e-05, + "loss": 0.7631, + "step": 10738 + }, + { + "epoch": 0.47547679399621445, + "grad_norm": 1.3263957500457764, + "learning_rate": 1.2573259456446931e-05, + "loss": 0.6849, + "step": 10739 + }, + { + "epoch": 0.4755210697010283, + "grad_norm": 1.420837163925171, + "learning_rate": 1.257176601836949e-05, + "loss": 0.6801, + "step": 10740 + }, + { + "epoch": 0.47556534540584217, + "grad_norm": 1.6042816638946533, + "learning_rate": 1.2570272518867515e-05, + "loss": 0.7333, + "step": 10741 + }, + { + "epoch": 0.4756096211106561, + "grad_norm": 1.1700105667114258, + "learning_rate": 1.2568778957976671e-05, + "loss": 0.6793, + "step": 10742 + }, + { + "epoch": 0.47565389681546993, + "grad_norm": 1.2309640645980835, + "learning_rate": 1.2567285335732633e-05, + "loss": 0.6912, + "step": 10743 + }, + { + "epoch": 0.4756981725202838, + "grad_norm": 1.391334056854248, + "learning_rate": 1.2565791652171073e-05, + "loss": 0.7545, + "step": 10744 + }, + { + "epoch": 0.4757424482250977, + "grad_norm": 1.728377342224121, + "learning_rate": 1.2564297907327672e-05, + "loss": 0.7744, + "step": 10745 + }, + { + "epoch": 0.47578672392991156, + "grad_norm": 1.4831045866012573, + "learning_rate": 1.2562804101238105e-05, + "loss": 0.6836, + "step": 10746 + }, + { + "epoch": 0.4758309996347254, + "grad_norm": 1.1643264293670654, + "learning_rate": 1.2561310233938044e-05, + "loss": 0.575, + "step": 10747 + }, + { + "epoch": 0.4758752753395393, + "grad_norm": 1.6726291179656982, + "learning_rate": 1.2559816305463174e-05, + "loss": 0.7093, + "step": 10748 + }, + { + "epoch": 0.4759195510443532, + "grad_norm": 1.5357704162597656, + "learning_rate": 1.2558322315849177e-05, + "loss": 0.7165, + "step": 10749 + }, + { + "epoch": 0.47596382674916704, + "grad_norm": 1.7604588270187378, + "learning_rate": 1.2556828265131736e-05, + "loss": 0.771, + "step": 10750 + }, + { + "epoch": 0.47600810245398095, + "grad_norm": 1.2058436870574951, + "learning_rate": 1.2555334153346533e-05, + "loss": 0.5834, + "step": 10751 + }, + { + "epoch": 0.4760523781587948, + "grad_norm": 1.545794129371643, + "learning_rate": 1.2553839980529255e-05, + "loss": 0.7487, + "step": 10752 + }, + { + "epoch": 0.47609665386360867, + "grad_norm": 1.2114651203155518, + "learning_rate": 1.2552345746715586e-05, + "loss": 0.5824, + "step": 10753 + }, + { + "epoch": 0.4761409295684226, + "grad_norm": 1.5059913396835327, + "learning_rate": 1.2550851451941217e-05, + "loss": 0.7439, + "step": 10754 + }, + { + "epoch": 0.47618520527323643, + "grad_norm": 1.4513999223709106, + "learning_rate": 1.2549357096241841e-05, + "loss": 0.6981, + "step": 10755 + }, + { + "epoch": 0.47622948097805035, + "grad_norm": 1.385939359664917, + "learning_rate": 1.2547862679653145e-05, + "loss": 0.7356, + "step": 10756 + }, + { + "epoch": 0.4762737566828642, + "grad_norm": 1.6814823150634766, + "learning_rate": 1.2546368202210823e-05, + "loss": 0.7597, + "step": 10757 + }, + { + "epoch": 0.47631803238767806, + "grad_norm": 1.429049015045166, + "learning_rate": 1.2544873663950569e-05, + "loss": 0.665, + "step": 10758 + }, + { + "epoch": 0.47636230809249197, + "grad_norm": 1.4594637155532837, + "learning_rate": 1.254337906490808e-05, + "loss": 0.6596, + "step": 10759 + }, + { + "epoch": 0.4764065837973058, + "grad_norm": 1.362610936164856, + "learning_rate": 1.2541884405119057e-05, + "loss": 0.715, + "step": 10760 + }, + { + "epoch": 0.4764508595021197, + "grad_norm": 1.364950180053711, + "learning_rate": 1.2540389684619192e-05, + "loss": 0.6895, + "step": 10761 + }, + { + "epoch": 0.4764951352069336, + "grad_norm": 1.446300983428955, + "learning_rate": 1.2538894903444188e-05, + "loss": 0.6596, + "step": 10762 + }, + { + "epoch": 0.47653941091174745, + "grad_norm": 1.5610817670822144, + "learning_rate": 1.2537400061629742e-05, + "loss": 0.67, + "step": 10763 + }, + { + "epoch": 0.4765836866165613, + "grad_norm": 1.2059566974639893, + "learning_rate": 1.2535905159211567e-05, + "loss": 0.6224, + "step": 10764 + }, + { + "epoch": 0.4766279623213752, + "grad_norm": 1.7970631122589111, + "learning_rate": 1.2534410196225362e-05, + "loss": 0.725, + "step": 10765 + }, + { + "epoch": 0.4766722380261891, + "grad_norm": 1.2427327632904053, + "learning_rate": 1.253291517270683e-05, + "loss": 0.623, + "step": 10766 + }, + { + "epoch": 0.47671651373100293, + "grad_norm": 1.4524625539779663, + "learning_rate": 1.2531420088691683e-05, + "loss": 0.6797, + "step": 10767 + }, + { + "epoch": 0.47676078943581685, + "grad_norm": 2.024853229522705, + "learning_rate": 1.2529924944215628e-05, + "loss": 0.7887, + "step": 10768 + }, + { + "epoch": 0.4768050651406307, + "grad_norm": 1.6920424699783325, + "learning_rate": 1.2528429739314376e-05, + "loss": 0.7727, + "step": 10769 + }, + { + "epoch": 0.47684934084544456, + "grad_norm": 1.5059340000152588, + "learning_rate": 1.252693447402364e-05, + "loss": 0.6945, + "step": 10770 + }, + { + "epoch": 0.47689361655025847, + "grad_norm": 1.169180989265442, + "learning_rate": 1.2525439148379127e-05, + "loss": 0.6098, + "step": 10771 + }, + { + "epoch": 0.4769378922550723, + "grad_norm": 2.0360348224639893, + "learning_rate": 1.2523943762416561e-05, + "loss": 0.7668, + "step": 10772 + }, + { + "epoch": 0.47698216795988624, + "grad_norm": 1.9283493757247925, + "learning_rate": 1.2522448316171651e-05, + "loss": 0.7555, + "step": 10773 + }, + { + "epoch": 0.4770264436647001, + "grad_norm": 1.8100135326385498, + "learning_rate": 1.2520952809680119e-05, + "loss": 0.7013, + "step": 10774 + }, + { + "epoch": 0.47707071936951395, + "grad_norm": 1.217002272605896, + "learning_rate": 1.251945724297768e-05, + "loss": 0.6105, + "step": 10775 + }, + { + "epoch": 0.47711499507432786, + "grad_norm": 1.7201792001724243, + "learning_rate": 1.2517961616100056e-05, + "loss": 0.733, + "step": 10776 + }, + { + "epoch": 0.4771592707791417, + "grad_norm": 1.7515192031860352, + "learning_rate": 1.2516465929082968e-05, + "loss": 0.7254, + "step": 10777 + }, + { + "epoch": 0.4772035464839556, + "grad_norm": 1.398599624633789, + "learning_rate": 1.2514970181962141e-05, + "loss": 0.6246, + "step": 10778 + }, + { + "epoch": 0.4772478221887695, + "grad_norm": 1.3637577295303345, + "learning_rate": 1.25134743747733e-05, + "loss": 0.6984, + "step": 10779 + }, + { + "epoch": 0.47729209789358334, + "grad_norm": 1.4334604740142822, + "learning_rate": 1.2511978507552172e-05, + "loss": 0.6377, + "step": 10780 + }, + { + "epoch": 0.4773363735983972, + "grad_norm": 1.3388240337371826, + "learning_rate": 1.2510482580334477e-05, + "loss": 0.6433, + "step": 10781 + }, + { + "epoch": 0.4773806493032111, + "grad_norm": 1.2575783729553223, + "learning_rate": 1.2508986593155955e-05, + "loss": 0.6372, + "step": 10782 + }, + { + "epoch": 0.47742492500802497, + "grad_norm": 1.3932738304138184, + "learning_rate": 1.2507490546052328e-05, + "loss": 0.605, + "step": 10783 + }, + { + "epoch": 0.4774692007128388, + "grad_norm": 1.4574029445648193, + "learning_rate": 1.2505994439059332e-05, + "loss": 0.678, + "step": 10784 + }, + { + "epoch": 0.47751347641765274, + "grad_norm": 1.2843064069747925, + "learning_rate": 1.2504498272212701e-05, + "loss": 0.7165, + "step": 10785 + }, + { + "epoch": 0.4775577521224666, + "grad_norm": 1.3831541538238525, + "learning_rate": 1.2503002045548164e-05, + "loss": 0.6717, + "step": 10786 + }, + { + "epoch": 0.47760202782728045, + "grad_norm": 1.2369564771652222, + "learning_rate": 1.2501505759101465e-05, + "loss": 0.5974, + "step": 10787 + }, + { + "epoch": 0.47764630353209436, + "grad_norm": 1.3641971349716187, + "learning_rate": 1.2500009412908338e-05, + "loss": 0.7051, + "step": 10788 + }, + { + "epoch": 0.4776905792369082, + "grad_norm": 1.9469393491744995, + "learning_rate": 1.2498513007004518e-05, + "loss": 0.6891, + "step": 10789 + }, + { + "epoch": 0.47773485494172213, + "grad_norm": 1.8116111755371094, + "learning_rate": 1.2497016541425753e-05, + "loss": 0.7719, + "step": 10790 + }, + { + "epoch": 0.477779130646536, + "grad_norm": 1.328803539276123, + "learning_rate": 1.2495520016207778e-05, + "loss": 0.6835, + "step": 10791 + }, + { + "epoch": 0.47782340635134984, + "grad_norm": 1.3714492321014404, + "learning_rate": 1.249402343138634e-05, + "loss": 0.6856, + "step": 10792 + }, + { + "epoch": 0.47786768205616376, + "grad_norm": 2.0145301818847656, + "learning_rate": 1.2492526786997185e-05, + "loss": 0.789, + "step": 10793 + }, + { + "epoch": 0.4779119577609776, + "grad_norm": 2.2556581497192383, + "learning_rate": 1.2491030083076055e-05, + "loss": 0.7738, + "step": 10794 + }, + { + "epoch": 0.47795623346579147, + "grad_norm": 1.3079004287719727, + "learning_rate": 1.2489533319658703e-05, + "loss": 0.6984, + "step": 10795 + }, + { + "epoch": 0.4780005091706054, + "grad_norm": 1.2743555307388306, + "learning_rate": 1.248803649678087e-05, + "loss": 0.6456, + "step": 10796 + }, + { + "epoch": 0.47804478487541924, + "grad_norm": 1.3768033981323242, + "learning_rate": 1.2486539614478312e-05, + "loss": 0.6901, + "step": 10797 + }, + { + "epoch": 0.4780890605802331, + "grad_norm": 1.544523000717163, + "learning_rate": 1.2485042672786784e-05, + "loss": 0.7786, + "step": 10798 + }, + { + "epoch": 0.478133336285047, + "grad_norm": 1.3465452194213867, + "learning_rate": 1.248354567174203e-05, + "loss": 0.6464, + "step": 10799 + }, + { + "epoch": 0.47817761198986086, + "grad_norm": 1.4291722774505615, + "learning_rate": 1.2482048611379814e-05, + "loss": 0.7671, + "step": 10800 + }, + { + "epoch": 0.4782218876946747, + "grad_norm": 2.0361616611480713, + "learning_rate": 1.2480551491735883e-05, + "loss": 0.6745, + "step": 10801 + }, + { + "epoch": 0.47826616339948863, + "grad_norm": 1.5991261005401611, + "learning_rate": 1.2479054312846005e-05, + "loss": 0.7427, + "step": 10802 + }, + { + "epoch": 0.4783104391043025, + "grad_norm": 1.372277021408081, + "learning_rate": 1.2477557074745932e-05, + "loss": 0.7268, + "step": 10803 + }, + { + "epoch": 0.47835471480911634, + "grad_norm": 1.5796802043914795, + "learning_rate": 1.2476059777471424e-05, + "loss": 0.6804, + "step": 10804 + }, + { + "epoch": 0.47839899051393026, + "grad_norm": 1.4113576412200928, + "learning_rate": 1.2474562421058245e-05, + "loss": 0.7083, + "step": 10805 + }, + { + "epoch": 0.4784432662187441, + "grad_norm": 1.5430424213409424, + "learning_rate": 1.2473065005542155e-05, + "loss": 0.7129, + "step": 10806 + }, + { + "epoch": 0.47848754192355797, + "grad_norm": 1.6991671323776245, + "learning_rate": 1.2471567530958926e-05, + "loss": 0.6682, + "step": 10807 + }, + { + "epoch": 0.4785318176283719, + "grad_norm": 1.2590442895889282, + "learning_rate": 1.2470069997344317e-05, + "loss": 0.6876, + "step": 10808 + }, + { + "epoch": 0.47857609333318574, + "grad_norm": 1.2776470184326172, + "learning_rate": 1.2468572404734094e-05, + "loss": 0.6407, + "step": 10809 + }, + { + "epoch": 0.47862036903799965, + "grad_norm": 1.3882009983062744, + "learning_rate": 1.246707475316403e-05, + "loss": 0.7309, + "step": 10810 + }, + { + "epoch": 0.4786646447428135, + "grad_norm": 1.4551010131835938, + "learning_rate": 1.2465577042669896e-05, + "loss": 0.7224, + "step": 10811 + }, + { + "epoch": 0.47870892044762736, + "grad_norm": 1.5436522960662842, + "learning_rate": 1.2464079273287463e-05, + "loss": 0.7401, + "step": 10812 + }, + { + "epoch": 0.4787531961524413, + "grad_norm": 1.846874713897705, + "learning_rate": 1.2462581445052501e-05, + "loss": 0.7386, + "step": 10813 + }, + { + "epoch": 0.47879747185725513, + "grad_norm": 1.266257882118225, + "learning_rate": 1.2461083558000786e-05, + "loss": 0.6255, + "step": 10814 + }, + { + "epoch": 0.478841747562069, + "grad_norm": 1.3717319965362549, + "learning_rate": 1.2459585612168095e-05, + "loss": 0.6341, + "step": 10815 + }, + { + "epoch": 0.4788860232668829, + "grad_norm": 1.6491330862045288, + "learning_rate": 1.2458087607590206e-05, + "loss": 0.7365, + "step": 10816 + }, + { + "epoch": 0.47893029897169676, + "grad_norm": 1.556686282157898, + "learning_rate": 1.2456589544302894e-05, + "loss": 0.6524, + "step": 10817 + }, + { + "epoch": 0.4789745746765106, + "grad_norm": 1.4602354764938354, + "learning_rate": 1.2455091422341941e-05, + "loss": 0.7154, + "step": 10818 + }, + { + "epoch": 0.4790188503813245, + "grad_norm": 1.4416990280151367, + "learning_rate": 1.2453593241743127e-05, + "loss": 0.6915, + "step": 10819 + }, + { + "epoch": 0.4790631260861384, + "grad_norm": 1.9276155233383179, + "learning_rate": 1.245209500254224e-05, + "loss": 0.7588, + "step": 10820 + }, + { + "epoch": 0.47910740179095224, + "grad_norm": 1.3733934164047241, + "learning_rate": 1.2450596704775058e-05, + "loss": 0.6657, + "step": 10821 + }, + { + "epoch": 0.47915167749576615, + "grad_norm": 1.3707170486450195, + "learning_rate": 1.244909834847737e-05, + "loss": 0.5758, + "step": 10822 + }, + { + "epoch": 0.47919595320058, + "grad_norm": 1.5665799379348755, + "learning_rate": 1.244759993368496e-05, + "loss": 0.6979, + "step": 10823 + }, + { + "epoch": 0.47924022890539386, + "grad_norm": 1.3134695291519165, + "learning_rate": 1.2446101460433619e-05, + "loss": 0.656, + "step": 10824 + }, + { + "epoch": 0.4792845046102078, + "grad_norm": 1.580674648284912, + "learning_rate": 1.2444602928759138e-05, + "loss": 0.7561, + "step": 10825 + }, + { + "epoch": 0.47932878031502163, + "grad_norm": 1.6291817426681519, + "learning_rate": 1.2443104338697305e-05, + "loss": 0.7588, + "step": 10826 + }, + { + "epoch": 0.47937305601983554, + "grad_norm": 1.3389320373535156, + "learning_rate": 1.2441605690283915e-05, + "loss": 0.6269, + "step": 10827 + }, + { + "epoch": 0.4794173317246494, + "grad_norm": 1.2776156663894653, + "learning_rate": 1.2440106983554761e-05, + "loss": 0.6195, + "step": 10828 + }, + { + "epoch": 0.47946160742946325, + "grad_norm": 1.3773744106292725, + "learning_rate": 1.2438608218545639e-05, + "loss": 0.6957, + "step": 10829 + }, + { + "epoch": 0.47950588313427717, + "grad_norm": 1.6886751651763916, + "learning_rate": 1.2437109395292344e-05, + "loss": 0.6871, + "step": 10830 + }, + { + "epoch": 0.479550158839091, + "grad_norm": 1.3635154962539673, + "learning_rate": 1.2435610513830676e-05, + "loss": 0.6746, + "step": 10831 + }, + { + "epoch": 0.4795944345439049, + "grad_norm": 1.2063273191452026, + "learning_rate": 1.2434111574196435e-05, + "loss": 0.6853, + "step": 10832 + }, + { + "epoch": 0.4796387102487188, + "grad_norm": 1.703765869140625, + "learning_rate": 1.2432612576425417e-05, + "loss": 0.8235, + "step": 10833 + }, + { + "epoch": 0.47968298595353265, + "grad_norm": 1.4420619010925293, + "learning_rate": 1.2431113520553435e-05, + "loss": 0.7423, + "step": 10834 + }, + { + "epoch": 0.4797272616583465, + "grad_norm": 1.8260775804519653, + "learning_rate": 1.2429614406616281e-05, + "loss": 0.7181, + "step": 10835 + }, + { + "epoch": 0.4797715373631604, + "grad_norm": 1.2914692163467407, + "learning_rate": 1.2428115234649766e-05, + "loss": 0.6984, + "step": 10836 + }, + { + "epoch": 0.4798158130679743, + "grad_norm": 1.5541356801986694, + "learning_rate": 1.2426616004689697e-05, + "loss": 0.6696, + "step": 10837 + }, + { + "epoch": 0.47986008877278813, + "grad_norm": 1.3602488040924072, + "learning_rate": 1.2425116716771878e-05, + "loss": 0.7038, + "step": 10838 + }, + { + "epoch": 0.47990436447760204, + "grad_norm": 1.391021966934204, + "learning_rate": 1.2423617370932127e-05, + "loss": 0.721, + "step": 10839 + }, + { + "epoch": 0.4799486401824159, + "grad_norm": 1.2546638250350952, + "learning_rate": 1.2422117967206244e-05, + "loss": 0.6471, + "step": 10840 + }, + { + "epoch": 0.47999291588722975, + "grad_norm": 1.315182089805603, + "learning_rate": 1.2420618505630045e-05, + "loss": 0.6407, + "step": 10841 + }, + { + "epoch": 0.48003719159204367, + "grad_norm": 1.1824008226394653, + "learning_rate": 1.2419118986239345e-05, + "loss": 0.633, + "step": 10842 + }, + { + "epoch": 0.4800814672968575, + "grad_norm": 1.2899338006973267, + "learning_rate": 1.2417619409069954e-05, + "loss": 0.6539, + "step": 10843 + }, + { + "epoch": 0.48012574300167143, + "grad_norm": 1.4269570112228394, + "learning_rate": 1.2416119774157698e-05, + "loss": 0.6804, + "step": 10844 + }, + { + "epoch": 0.4801700187064853, + "grad_norm": 1.4761669635772705, + "learning_rate": 1.2414620081538384e-05, + "loss": 0.6753, + "step": 10845 + }, + { + "epoch": 0.48021429441129915, + "grad_norm": 1.4375391006469727, + "learning_rate": 1.2413120331247836e-05, + "loss": 0.6598, + "step": 10846 + }, + { + "epoch": 0.48025857011611306, + "grad_norm": 1.472309947013855, + "learning_rate": 1.2411620523321878e-05, + "loss": 0.7147, + "step": 10847 + }, + { + "epoch": 0.4803028458209269, + "grad_norm": 1.282094955444336, + "learning_rate": 1.2410120657796323e-05, + "loss": 0.646, + "step": 10848 + }, + { + "epoch": 0.4803471215257408, + "grad_norm": 1.4053771495819092, + "learning_rate": 1.2408620734707e-05, + "loss": 0.6247, + "step": 10849 + }, + { + "epoch": 0.4803913972305547, + "grad_norm": 1.3197696208953857, + "learning_rate": 1.2407120754089733e-05, + "loss": 0.6868, + "step": 10850 + }, + { + "epoch": 0.48043567293536854, + "grad_norm": 1.4281494617462158, + "learning_rate": 1.2405620715980345e-05, + "loss": 0.6618, + "step": 10851 + }, + { + "epoch": 0.4804799486401824, + "grad_norm": 1.1929285526275635, + "learning_rate": 1.2404120620414664e-05, + "loss": 0.6249, + "step": 10852 + }, + { + "epoch": 0.4805242243449963, + "grad_norm": 1.430849313735962, + "learning_rate": 1.2402620467428518e-05, + "loss": 0.6573, + "step": 10853 + }, + { + "epoch": 0.48056850004981017, + "grad_norm": 1.4551537036895752, + "learning_rate": 1.2401120257057744e-05, + "loss": 0.6784, + "step": 10854 + }, + { + "epoch": 0.480612775754624, + "grad_norm": 1.427275538444519, + "learning_rate": 1.2399619989338165e-05, + "loss": 0.7317, + "step": 10855 + }, + { + "epoch": 0.48065705145943793, + "grad_norm": 1.528073787689209, + "learning_rate": 1.2398119664305617e-05, + "loss": 0.7129, + "step": 10856 + }, + { + "epoch": 0.4807013271642518, + "grad_norm": 1.5097854137420654, + "learning_rate": 1.2396619281995931e-05, + "loss": 0.6878, + "step": 10857 + }, + { + "epoch": 0.48074560286906565, + "grad_norm": 1.2865521907806396, + "learning_rate": 1.2395118842444946e-05, + "loss": 0.6383, + "step": 10858 + }, + { + "epoch": 0.48078987857387956, + "grad_norm": 1.4045348167419434, + "learning_rate": 1.2393618345688503e-05, + "loss": 0.7169, + "step": 10859 + }, + { + "epoch": 0.4808341542786934, + "grad_norm": 1.2297905683517456, + "learning_rate": 1.2392117791762427e-05, + "loss": 0.5857, + "step": 10860 + }, + { + "epoch": 0.4808784299835073, + "grad_norm": 1.3424701690673828, + "learning_rate": 1.239061718070257e-05, + "loss": 0.676, + "step": 10861 + }, + { + "epoch": 0.4809227056883212, + "grad_norm": 1.6767505407333374, + "learning_rate": 1.2389116512544766e-05, + "loss": 0.7474, + "step": 10862 + }, + { + "epoch": 0.48096698139313504, + "grad_norm": 1.4468175172805786, + "learning_rate": 1.238761578732486e-05, + "loss": 0.7295, + "step": 10863 + }, + { + "epoch": 0.48101125709794895, + "grad_norm": 1.4181365966796875, + "learning_rate": 1.2386115005078695e-05, + "loss": 0.7006, + "step": 10864 + }, + { + "epoch": 0.4810555328027628, + "grad_norm": 1.7065008878707886, + "learning_rate": 1.2384614165842116e-05, + "loss": 0.6992, + "step": 10865 + }, + { + "epoch": 0.48109980850757667, + "grad_norm": 1.5447617769241333, + "learning_rate": 1.2383113269650968e-05, + "loss": 0.7021, + "step": 10866 + }, + { + "epoch": 0.4811440842123906, + "grad_norm": 1.7512832880020142, + "learning_rate": 1.2381612316541101e-05, + "loss": 0.7735, + "step": 10867 + }, + { + "epoch": 0.48118835991720443, + "grad_norm": 1.3361051082611084, + "learning_rate": 1.2380111306548363e-05, + "loss": 0.6799, + "step": 10868 + }, + { + "epoch": 0.4812326356220183, + "grad_norm": 1.6397958993911743, + "learning_rate": 1.2378610239708606e-05, + "loss": 0.7734, + "step": 10869 + }, + { + "epoch": 0.4812769113268322, + "grad_norm": 1.410319447517395, + "learning_rate": 1.2377109116057678e-05, + "loss": 0.714, + "step": 10870 + }, + { + "epoch": 0.48132118703164606, + "grad_norm": 1.8139582872390747, + "learning_rate": 1.2375607935631433e-05, + "loss": 0.7495, + "step": 10871 + }, + { + "epoch": 0.4813654627364599, + "grad_norm": 1.2840195894241333, + "learning_rate": 1.237410669846573e-05, + "loss": 0.635, + "step": 10872 + }, + { + "epoch": 0.4814097384412738, + "grad_norm": 1.3408933877944946, + "learning_rate": 1.237260540459642e-05, + "loss": 0.6613, + "step": 10873 + }, + { + "epoch": 0.4814540141460877, + "grad_norm": 1.3139567375183105, + "learning_rate": 1.2371104054059364e-05, + "loss": 0.6719, + "step": 10874 + }, + { + "epoch": 0.48149828985090154, + "grad_norm": 1.4514782428741455, + "learning_rate": 1.2369602646890415e-05, + "loss": 0.6526, + "step": 10875 + }, + { + "epoch": 0.48154256555571545, + "grad_norm": 1.6137747764587402, + "learning_rate": 1.2368101183125435e-05, + "loss": 0.7472, + "step": 10876 + }, + { + "epoch": 0.4815868412605293, + "grad_norm": 1.4547462463378906, + "learning_rate": 1.236659966280029e-05, + "loss": 0.6982, + "step": 10877 + }, + { + "epoch": 0.48163111696534316, + "grad_norm": 1.4519411325454712, + "learning_rate": 1.236509808595084e-05, + "loss": 0.6879, + "step": 10878 + }, + { + "epoch": 0.4816753926701571, + "grad_norm": 1.4933243989944458, + "learning_rate": 1.2363596452612945e-05, + "loss": 0.6983, + "step": 10879 + }, + { + "epoch": 0.48171966837497093, + "grad_norm": 1.4790345430374146, + "learning_rate": 1.2362094762822475e-05, + "loss": 0.6497, + "step": 10880 + }, + { + "epoch": 0.48176394407978484, + "grad_norm": 1.4131397008895874, + "learning_rate": 1.2360593016615292e-05, + "loss": 0.7188, + "step": 10881 + }, + { + "epoch": 0.4818082197845987, + "grad_norm": 1.6528065204620361, + "learning_rate": 1.235909121402727e-05, + "loss": 0.6546, + "step": 10882 + }, + { + "epoch": 0.48185249548941256, + "grad_norm": 1.606486439704895, + "learning_rate": 1.2357589355094275e-05, + "loss": 0.6547, + "step": 10883 + }, + { + "epoch": 0.48189677119422647, + "grad_norm": 1.4068716764450073, + "learning_rate": 1.2356087439852176e-05, + "loss": 0.6646, + "step": 10884 + }, + { + "epoch": 0.4819410468990403, + "grad_norm": 1.4478362798690796, + "learning_rate": 1.2354585468336847e-05, + "loss": 0.7425, + "step": 10885 + }, + { + "epoch": 0.4819853226038542, + "grad_norm": 1.6625193357467651, + "learning_rate": 1.2353083440584161e-05, + "loss": 0.7079, + "step": 10886 + }, + { + "epoch": 0.4820295983086681, + "grad_norm": 1.455107569694519, + "learning_rate": 1.2351581356629999e-05, + "loss": 0.6281, + "step": 10887 + }, + { + "epoch": 0.48207387401348195, + "grad_norm": 1.3881820440292358, + "learning_rate": 1.2350079216510222e-05, + "loss": 0.7564, + "step": 10888 + }, + { + "epoch": 0.4821181497182958, + "grad_norm": 1.4411213397979736, + "learning_rate": 1.2348577020260721e-05, + "loss": 0.7441, + "step": 10889 + }, + { + "epoch": 0.4821624254231097, + "grad_norm": 1.5943952798843384, + "learning_rate": 1.234707476791737e-05, + "loss": 0.714, + "step": 10890 + }, + { + "epoch": 0.4822067011279236, + "grad_norm": 1.5166382789611816, + "learning_rate": 1.2345572459516049e-05, + "loss": 0.7264, + "step": 10891 + }, + { + "epoch": 0.48225097683273743, + "grad_norm": 1.327554702758789, + "learning_rate": 1.2344070095092639e-05, + "loss": 0.6221, + "step": 10892 + }, + { + "epoch": 0.48229525253755134, + "grad_norm": 1.5963574647903442, + "learning_rate": 1.2342567674683024e-05, + "loss": 0.6768, + "step": 10893 + }, + { + "epoch": 0.4823395282423652, + "grad_norm": 1.6046626567840576, + "learning_rate": 1.2341065198323088e-05, + "loss": 0.6237, + "step": 10894 + }, + { + "epoch": 0.48238380394717906, + "grad_norm": 1.5229483842849731, + "learning_rate": 1.2339562666048714e-05, + "loss": 0.5996, + "step": 10895 + }, + { + "epoch": 0.48242807965199297, + "grad_norm": 1.4899077415466309, + "learning_rate": 1.2338060077895793e-05, + "loss": 0.7271, + "step": 10896 + }, + { + "epoch": 0.4824723553568068, + "grad_norm": 1.2970637083053589, + "learning_rate": 1.2336557433900209e-05, + "loss": 0.6484, + "step": 10897 + }, + { + "epoch": 0.48251663106162074, + "grad_norm": 1.775162696838379, + "learning_rate": 1.2335054734097854e-05, + "loss": 0.6933, + "step": 10898 + }, + { + "epoch": 0.4825609067664346, + "grad_norm": 1.327305793762207, + "learning_rate": 1.2333551978524618e-05, + "loss": 0.6575, + "step": 10899 + }, + { + "epoch": 0.48260518247124845, + "grad_norm": 1.4322036504745483, + "learning_rate": 1.2332049167216393e-05, + "loss": 0.7138, + "step": 10900 + }, + { + "epoch": 0.48264945817606236, + "grad_norm": 1.897782564163208, + "learning_rate": 1.2330546300209072e-05, + "loss": 0.7539, + "step": 10901 + }, + { + "epoch": 0.4826937338808762, + "grad_norm": 1.4515063762664795, + "learning_rate": 1.2329043377538552e-05, + "loss": 0.693, + "step": 10902 + }, + { + "epoch": 0.4827380095856901, + "grad_norm": 1.455723762512207, + "learning_rate": 1.2327540399240726e-05, + "loss": 0.7123, + "step": 10903 + }, + { + "epoch": 0.482782285290504, + "grad_norm": 1.4006695747375488, + "learning_rate": 1.2326037365351492e-05, + "loss": 0.7684, + "step": 10904 + }, + { + "epoch": 0.48282656099531784, + "grad_norm": 1.4530028104782104, + "learning_rate": 1.2324534275906747e-05, + "loss": 0.6563, + "step": 10905 + }, + { + "epoch": 0.4828708367001317, + "grad_norm": 1.2148282527923584, + "learning_rate": 1.2323031130942398e-05, + "loss": 0.6178, + "step": 10906 + }, + { + "epoch": 0.4829151124049456, + "grad_norm": 1.3439923524856567, + "learning_rate": 1.2321527930494342e-05, + "loss": 0.6512, + "step": 10907 + }, + { + "epoch": 0.48295938810975947, + "grad_norm": 2.0698914527893066, + "learning_rate": 1.2320024674598479e-05, + "loss": 0.7834, + "step": 10908 + }, + { + "epoch": 0.4830036638145733, + "grad_norm": 1.704406976699829, + "learning_rate": 1.2318521363290718e-05, + "loss": 0.7407, + "step": 10909 + }, + { + "epoch": 0.48304793951938724, + "grad_norm": 1.3349182605743408, + "learning_rate": 1.2317017996606959e-05, + "loss": 0.6581, + "step": 10910 + }, + { + "epoch": 0.4830922152242011, + "grad_norm": 1.486767292022705, + "learning_rate": 1.2315514574583113e-05, + "loss": 0.6834, + "step": 10911 + }, + { + "epoch": 0.48313649092901495, + "grad_norm": 1.563464879989624, + "learning_rate": 1.2314011097255089e-05, + "loss": 0.7036, + "step": 10912 + }, + { + "epoch": 0.48318076663382886, + "grad_norm": 1.5697537660598755, + "learning_rate": 1.231250756465879e-05, + "loss": 0.6674, + "step": 10913 + }, + { + "epoch": 0.4832250423386427, + "grad_norm": 1.5086147785186768, + "learning_rate": 1.2311003976830136e-05, + "loss": 0.6845, + "step": 10914 + }, + { + "epoch": 0.48326931804345663, + "grad_norm": 1.2759486436843872, + "learning_rate": 1.2309500333805032e-05, + "loss": 0.659, + "step": 10915 + }, + { + "epoch": 0.4833135937482705, + "grad_norm": 1.3506550788879395, + "learning_rate": 1.230799663561939e-05, + "loss": 0.6146, + "step": 10916 + }, + { + "epoch": 0.48335786945308434, + "grad_norm": 1.407881736755371, + "learning_rate": 1.2306492882309132e-05, + "loss": 0.6827, + "step": 10917 + }, + { + "epoch": 0.48340214515789826, + "grad_norm": 1.426707148551941, + "learning_rate": 1.2304989073910165e-05, + "loss": 0.6744, + "step": 10918 + }, + { + "epoch": 0.4834464208627121, + "grad_norm": 1.5290310382843018, + "learning_rate": 1.2303485210458413e-05, + "loss": 0.6949, + "step": 10919 + }, + { + "epoch": 0.48349069656752597, + "grad_norm": 1.2622826099395752, + "learning_rate": 1.2301981291989794e-05, + "loss": 0.6484, + "step": 10920 + }, + { + "epoch": 0.4835349722723399, + "grad_norm": 1.7142421007156372, + "learning_rate": 1.2300477318540222e-05, + "loss": 0.7917, + "step": 10921 + }, + { + "epoch": 0.48357924797715374, + "grad_norm": 1.8070143461227417, + "learning_rate": 1.2298973290145627e-05, + "loss": 0.7631, + "step": 10922 + }, + { + "epoch": 0.4836235236819676, + "grad_norm": 1.3478984832763672, + "learning_rate": 1.2297469206841921e-05, + "loss": 0.704, + "step": 10923 + }, + { + "epoch": 0.4836677993867815, + "grad_norm": 1.5058518648147583, + "learning_rate": 1.2295965068665039e-05, + "loss": 0.6948, + "step": 10924 + }, + { + "epoch": 0.48371207509159536, + "grad_norm": 1.4440590143203735, + "learning_rate": 1.2294460875650897e-05, + "loss": 0.7324, + "step": 10925 + }, + { + "epoch": 0.4837563507964092, + "grad_norm": 1.3035472631454468, + "learning_rate": 1.2292956627835429e-05, + "loss": 0.7096, + "step": 10926 + }, + { + "epoch": 0.48380062650122313, + "grad_norm": 1.3369377851486206, + "learning_rate": 1.2291452325254556e-05, + "loss": 0.7033, + "step": 10927 + }, + { + "epoch": 0.483844902206037, + "grad_norm": 1.4541033506393433, + "learning_rate": 1.228994796794421e-05, + "loss": 0.7324, + "step": 10928 + }, + { + "epoch": 0.48388917791085084, + "grad_norm": 1.0876225233078003, + "learning_rate": 1.2288443555940324e-05, + "loss": 0.6774, + "step": 10929 + }, + { + "epoch": 0.48393345361566475, + "grad_norm": 1.2502132654190063, + "learning_rate": 1.2286939089278827e-05, + "loss": 0.5675, + "step": 10930 + }, + { + "epoch": 0.4839777293204786, + "grad_norm": 1.5219991207122803, + "learning_rate": 1.2285434567995646e-05, + "loss": 0.6226, + "step": 10931 + }, + { + "epoch": 0.4840220050252925, + "grad_norm": 1.2829033136367798, + "learning_rate": 1.2283929992126729e-05, + "loss": 0.6348, + "step": 10932 + }, + { + "epoch": 0.4840662807301064, + "grad_norm": 1.5460190773010254, + "learning_rate": 1.2282425361707996e-05, + "loss": 0.6915, + "step": 10933 + }, + { + "epoch": 0.48411055643492024, + "grad_norm": 1.4897249937057495, + "learning_rate": 1.22809206767754e-05, + "loss": 0.7072, + "step": 10934 + }, + { + "epoch": 0.48415483213973415, + "grad_norm": 1.7946709394454956, + "learning_rate": 1.2279415937364866e-05, + "loss": 0.8062, + "step": 10935 + }, + { + "epoch": 0.484199107844548, + "grad_norm": 1.7356599569320679, + "learning_rate": 1.227791114351234e-05, + "loss": 0.6469, + "step": 10936 + }, + { + "epoch": 0.48424338354936186, + "grad_norm": 1.28611421585083, + "learning_rate": 1.227640629525376e-05, + "loss": 0.6559, + "step": 10937 + }, + { + "epoch": 0.4842876592541758, + "grad_norm": 1.824358582496643, + "learning_rate": 1.2274901392625068e-05, + "loss": 0.6569, + "step": 10938 + }, + { + "epoch": 0.48433193495898963, + "grad_norm": 1.301817536354065, + "learning_rate": 1.2273396435662212e-05, + "loss": 0.5845, + "step": 10939 + }, + { + "epoch": 0.4843762106638035, + "grad_norm": 1.3262099027633667, + "learning_rate": 1.227189142440113e-05, + "loss": 0.6834, + "step": 10940 + }, + { + "epoch": 0.4844204863686174, + "grad_norm": 1.6950191259384155, + "learning_rate": 1.2270386358877772e-05, + "loss": 0.6633, + "step": 10941 + }, + { + "epoch": 0.48446476207343125, + "grad_norm": 1.2334429025650024, + "learning_rate": 1.2268881239128084e-05, + "loss": 0.6106, + "step": 10942 + }, + { + "epoch": 0.4845090377782451, + "grad_norm": 2.045060634613037, + "learning_rate": 1.2267376065188016e-05, + "loss": 0.7845, + "step": 10943 + }, + { + "epoch": 0.484553313483059, + "grad_norm": 1.4415793418884277, + "learning_rate": 1.2265870837093519e-05, + "loss": 0.6844, + "step": 10944 + }, + { + "epoch": 0.4845975891878729, + "grad_norm": 1.429697036743164, + "learning_rate": 1.2264365554880538e-05, + "loss": 0.6808, + "step": 10945 + }, + { + "epoch": 0.48464186489268674, + "grad_norm": 1.4632775783538818, + "learning_rate": 1.226286021858503e-05, + "loss": 0.6454, + "step": 10946 + }, + { + "epoch": 0.48468614059750065, + "grad_norm": 1.9280329942703247, + "learning_rate": 1.2261354828242948e-05, + "loss": 0.7021, + "step": 10947 + }, + { + "epoch": 0.4847304163023145, + "grad_norm": 1.1972436904907227, + "learning_rate": 1.225984938389025e-05, + "loss": 0.6037, + "step": 10948 + }, + { + "epoch": 0.4847746920071284, + "grad_norm": 1.480891227722168, + "learning_rate": 1.2258343885562889e-05, + "loss": 0.7237, + "step": 10949 + }, + { + "epoch": 0.4848189677119423, + "grad_norm": 1.3677870035171509, + "learning_rate": 1.2256838333296819e-05, + "loss": 0.6329, + "step": 10950 + }, + { + "epoch": 0.48486324341675613, + "grad_norm": 1.3878884315490723, + "learning_rate": 1.2255332727128007e-05, + "loss": 0.6914, + "step": 10951 + }, + { + "epoch": 0.48490751912157004, + "grad_norm": 1.429428219795227, + "learning_rate": 1.2253827067092407e-05, + "loss": 0.6932, + "step": 10952 + }, + { + "epoch": 0.4849517948263839, + "grad_norm": 1.361233115196228, + "learning_rate": 1.2252321353225984e-05, + "loss": 0.722, + "step": 10953 + }, + { + "epoch": 0.48499607053119775, + "grad_norm": 1.5852957963943481, + "learning_rate": 1.2250815585564702e-05, + "loss": 0.7261, + "step": 10954 + }, + { + "epoch": 0.48504034623601167, + "grad_norm": 1.408943772315979, + "learning_rate": 1.2249309764144517e-05, + "loss": 0.6458, + "step": 10955 + }, + { + "epoch": 0.4850846219408255, + "grad_norm": 1.291866660118103, + "learning_rate": 1.2247803889001402e-05, + "loss": 0.6586, + "step": 10956 + }, + { + "epoch": 0.4851288976456394, + "grad_norm": 1.628018856048584, + "learning_rate": 1.2246297960171321e-05, + "loss": 0.692, + "step": 10957 + }, + { + "epoch": 0.4851731733504533, + "grad_norm": 1.5509881973266602, + "learning_rate": 1.2244791977690245e-05, + "loss": 0.6332, + "step": 10958 + }, + { + "epoch": 0.48521744905526715, + "grad_norm": 1.5591275691986084, + "learning_rate": 1.224328594159414e-05, + "loss": 0.7056, + "step": 10959 + }, + { + "epoch": 0.485261724760081, + "grad_norm": 1.5171414613723755, + "learning_rate": 1.2241779851918972e-05, + "loss": 0.6672, + "step": 10960 + }, + { + "epoch": 0.4853060004648949, + "grad_norm": 1.7211285829544067, + "learning_rate": 1.224027370870072e-05, + "loss": 0.7467, + "step": 10961 + }, + { + "epoch": 0.48535027616970877, + "grad_norm": 1.345004677772522, + "learning_rate": 1.2238767511975358e-05, + "loss": 0.6349, + "step": 10962 + }, + { + "epoch": 0.48539455187452263, + "grad_norm": 1.47917902469635, + "learning_rate": 1.2237261261778852e-05, + "loss": 0.7112, + "step": 10963 + }, + { + "epoch": 0.48543882757933654, + "grad_norm": 1.4391283988952637, + "learning_rate": 1.2235754958147186e-05, + "loss": 0.7123, + "step": 10964 + }, + { + "epoch": 0.4854831032841504, + "grad_norm": 1.5855157375335693, + "learning_rate": 1.223424860111633e-05, + "loss": 0.6842, + "step": 10965 + }, + { + "epoch": 0.48552737898896425, + "grad_norm": 1.3102853298187256, + "learning_rate": 1.2232742190722267e-05, + "loss": 0.6494, + "step": 10966 + }, + { + "epoch": 0.48557165469377817, + "grad_norm": 1.447176218032837, + "learning_rate": 1.2231235727000977e-05, + "loss": 0.6991, + "step": 10967 + }, + { + "epoch": 0.485615930398592, + "grad_norm": 1.4982562065124512, + "learning_rate": 1.2229729209988434e-05, + "loss": 0.7083, + "step": 10968 + }, + { + "epoch": 0.48566020610340593, + "grad_norm": 1.319154977798462, + "learning_rate": 1.2228222639720627e-05, + "loss": 0.6734, + "step": 10969 + }, + { + "epoch": 0.4857044818082198, + "grad_norm": 1.7387633323669434, + "learning_rate": 1.2226716016233535e-05, + "loss": 0.8032, + "step": 10970 + }, + { + "epoch": 0.48574875751303365, + "grad_norm": 1.244421124458313, + "learning_rate": 1.2225209339563144e-05, + "loss": 0.6496, + "step": 10971 + }, + { + "epoch": 0.48579303321784756, + "grad_norm": 1.461966872215271, + "learning_rate": 1.2223702609745442e-05, + "loss": 0.671, + "step": 10972 + }, + { + "epoch": 0.4858373089226614, + "grad_norm": 1.4718005657196045, + "learning_rate": 1.2222195826816411e-05, + "loss": 0.6801, + "step": 10973 + }, + { + "epoch": 0.48588158462747527, + "grad_norm": 1.3631882667541504, + "learning_rate": 1.2220688990812047e-05, + "loss": 0.6779, + "step": 10974 + }, + { + "epoch": 0.4859258603322892, + "grad_norm": 1.626217246055603, + "learning_rate": 1.221918210176833e-05, + "loss": 0.7622, + "step": 10975 + }, + { + "epoch": 0.48597013603710304, + "grad_norm": 1.516545057296753, + "learning_rate": 1.2217675159721255e-05, + "loss": 0.7141, + "step": 10976 + }, + { + "epoch": 0.4860144117419169, + "grad_norm": 1.3637663125991821, + "learning_rate": 1.2216168164706818e-05, + "loss": 0.6913, + "step": 10977 + }, + { + "epoch": 0.4860586874467308, + "grad_norm": 1.8623509407043457, + "learning_rate": 1.2214661116761007e-05, + "loss": 0.7869, + "step": 10978 + }, + { + "epoch": 0.48610296315154466, + "grad_norm": 1.4831130504608154, + "learning_rate": 1.221315401591982e-05, + "loss": 0.6246, + "step": 10979 + }, + { + "epoch": 0.4861472388563585, + "grad_norm": 1.1653385162353516, + "learning_rate": 1.2211646862219246e-05, + "loss": 0.6713, + "step": 10980 + }, + { + "epoch": 0.48619151456117243, + "grad_norm": 1.444596290588379, + "learning_rate": 1.2210139655695293e-05, + "loss": 0.6735, + "step": 10981 + }, + { + "epoch": 0.4862357902659863, + "grad_norm": 1.362349271774292, + "learning_rate": 1.2208632396383953e-05, + "loss": 0.6538, + "step": 10982 + }, + { + "epoch": 0.48628006597080015, + "grad_norm": 2.1254334449768066, + "learning_rate": 1.2207125084321226e-05, + "loss": 0.8624, + "step": 10983 + }, + { + "epoch": 0.48632434167561406, + "grad_norm": 1.41339111328125, + "learning_rate": 1.2205617719543112e-05, + "loss": 0.6761, + "step": 10984 + }, + { + "epoch": 0.4863686173804279, + "grad_norm": 1.7071295976638794, + "learning_rate": 1.2204110302085615e-05, + "loss": 0.6998, + "step": 10985 + }, + { + "epoch": 0.4864128930852418, + "grad_norm": 1.2987412214279175, + "learning_rate": 1.2202602831984742e-05, + "loss": 0.6116, + "step": 10986 + }, + { + "epoch": 0.4864571687900557, + "grad_norm": 1.437924861907959, + "learning_rate": 1.220109530927649e-05, + "loss": 0.6565, + "step": 10987 + }, + { + "epoch": 0.48650144449486954, + "grad_norm": 1.5049830675125122, + "learning_rate": 1.2199587733996867e-05, + "loss": 0.6914, + "step": 10988 + }, + { + "epoch": 0.48654572019968345, + "grad_norm": 1.4903546571731567, + "learning_rate": 1.2198080106181887e-05, + "loss": 0.7531, + "step": 10989 + }, + { + "epoch": 0.4865899959044973, + "grad_norm": 1.430260181427002, + "learning_rate": 1.219657242586755e-05, + "loss": 0.646, + "step": 10990 + }, + { + "epoch": 0.48663427160931116, + "grad_norm": 1.479001522064209, + "learning_rate": 1.219506469308987e-05, + "loss": 0.7714, + "step": 10991 + }, + { + "epoch": 0.4866785473141251, + "grad_norm": 1.2624022960662842, + "learning_rate": 1.2193556907884857e-05, + "loss": 0.6475, + "step": 10992 + }, + { + "epoch": 0.48672282301893893, + "grad_norm": 1.436708688735962, + "learning_rate": 1.2192049070288524e-05, + "loss": 0.6931, + "step": 10993 + }, + { + "epoch": 0.4867670987237528, + "grad_norm": 1.440485954284668, + "learning_rate": 1.219054118033688e-05, + "loss": 0.6267, + "step": 10994 + }, + { + "epoch": 0.4868113744285667, + "grad_norm": 1.7706165313720703, + "learning_rate": 1.218903323806595e-05, + "loss": 0.7362, + "step": 10995 + }, + { + "epoch": 0.48685565013338056, + "grad_norm": 1.3953278064727783, + "learning_rate": 1.218752524351174e-05, + "loss": 0.6627, + "step": 10996 + }, + { + "epoch": 0.4868999258381944, + "grad_norm": 1.3922263383865356, + "learning_rate": 1.218601719671027e-05, + "loss": 0.6668, + "step": 10997 + }, + { + "epoch": 0.4869442015430083, + "grad_norm": 1.361008644104004, + "learning_rate": 1.2184509097697559e-05, + "loss": 0.6627, + "step": 10998 + }, + { + "epoch": 0.4869884772478222, + "grad_norm": 1.2240437269210815, + "learning_rate": 1.2183000946509628e-05, + "loss": 0.6317, + "step": 10999 + }, + { + "epoch": 0.48703275295263604, + "grad_norm": 1.8537510633468628, + "learning_rate": 1.21814927431825e-05, + "loss": 0.6729, + "step": 11000 + }, + { + "epoch": 0.48707702865744995, + "grad_norm": 1.211398720741272, + "learning_rate": 1.217998448775219e-05, + "loss": 0.6855, + "step": 11001 + }, + { + "epoch": 0.4871213043622638, + "grad_norm": 1.7472360134124756, + "learning_rate": 1.2178476180254726e-05, + "loss": 0.7028, + "step": 11002 + }, + { + "epoch": 0.4871655800670777, + "grad_norm": 1.655534029006958, + "learning_rate": 1.217696782072613e-05, + "loss": 0.7301, + "step": 11003 + }, + { + "epoch": 0.4872098557718916, + "grad_norm": 1.340939998626709, + "learning_rate": 1.2175459409202435e-05, + "loss": 0.6042, + "step": 11004 + }, + { + "epoch": 0.48725413147670543, + "grad_norm": 1.5009568929672241, + "learning_rate": 1.2173950945719662e-05, + "loss": 0.6985, + "step": 11005 + }, + { + "epoch": 0.48729840718151934, + "grad_norm": 1.737157940864563, + "learning_rate": 1.217244243031384e-05, + "loss": 0.7411, + "step": 11006 + }, + { + "epoch": 0.4873426828863332, + "grad_norm": 1.3248847723007202, + "learning_rate": 1.2170933863020997e-05, + "loss": 0.6676, + "step": 11007 + }, + { + "epoch": 0.48738695859114706, + "grad_norm": 1.6923590898513794, + "learning_rate": 1.2169425243877168e-05, + "loss": 0.75, + "step": 11008 + }, + { + "epoch": 0.48743123429596097, + "grad_norm": 1.432895302772522, + "learning_rate": 1.2167916572918386e-05, + "loss": 0.6577, + "step": 11009 + }, + { + "epoch": 0.4874755100007748, + "grad_norm": 1.8475594520568848, + "learning_rate": 1.216640785018068e-05, + "loss": 0.7227, + "step": 11010 + }, + { + "epoch": 0.4875197857055887, + "grad_norm": 1.602759599685669, + "learning_rate": 1.2164899075700087e-05, + "loss": 0.7719, + "step": 11011 + }, + { + "epoch": 0.4875640614104026, + "grad_norm": 1.3451228141784668, + "learning_rate": 1.216339024951264e-05, + "loss": 0.6967, + "step": 11012 + }, + { + "epoch": 0.48760833711521645, + "grad_norm": 1.355528712272644, + "learning_rate": 1.2161881371654379e-05, + "loss": 0.6213, + "step": 11013 + }, + { + "epoch": 0.4876526128200303, + "grad_norm": 1.3799368143081665, + "learning_rate": 1.2160372442161346e-05, + "loss": 0.6476, + "step": 11014 + }, + { + "epoch": 0.4876968885248442, + "grad_norm": 1.3394198417663574, + "learning_rate": 1.2158863461069573e-05, + "loss": 0.6814, + "step": 11015 + }, + { + "epoch": 0.4877411642296581, + "grad_norm": 1.480440616607666, + "learning_rate": 1.2157354428415103e-05, + "loss": 0.7375, + "step": 11016 + }, + { + "epoch": 0.48778543993447193, + "grad_norm": 1.5758607387542725, + "learning_rate": 1.215584534423398e-05, + "loss": 0.7041, + "step": 11017 + }, + { + "epoch": 0.48782971563928584, + "grad_norm": 2.0358469486236572, + "learning_rate": 1.2154336208562248e-05, + "loss": 0.8203, + "step": 11018 + }, + { + "epoch": 0.4878739913440997, + "grad_norm": 1.5252543687820435, + "learning_rate": 1.2152827021435946e-05, + "loss": 0.68, + "step": 11019 + }, + { + "epoch": 0.4879182670489136, + "grad_norm": 1.345829725265503, + "learning_rate": 1.2151317782891125e-05, + "loss": 0.6491, + "step": 11020 + }, + { + "epoch": 0.48796254275372747, + "grad_norm": 1.3434754610061646, + "learning_rate": 1.2149808492963833e-05, + "loss": 0.6387, + "step": 11021 + }, + { + "epoch": 0.4880068184585413, + "grad_norm": 1.343508005142212, + "learning_rate": 1.2148299151690114e-05, + "loss": 0.7086, + "step": 11022 + }, + { + "epoch": 0.48805109416335524, + "grad_norm": 1.719545602798462, + "learning_rate": 1.2146789759106016e-05, + "loss": 0.7944, + "step": 11023 + }, + { + "epoch": 0.4880953698681691, + "grad_norm": 1.7814671993255615, + "learning_rate": 1.2145280315247598e-05, + "loss": 0.7688, + "step": 11024 + }, + { + "epoch": 0.48813964557298295, + "grad_norm": 1.3122644424438477, + "learning_rate": 1.2143770820150902e-05, + "loss": 0.6725, + "step": 11025 + }, + { + "epoch": 0.48818392127779686, + "grad_norm": 1.305016040802002, + "learning_rate": 1.2142261273851987e-05, + "loss": 0.6368, + "step": 11026 + }, + { + "epoch": 0.4882281969826107, + "grad_norm": 1.4909764528274536, + "learning_rate": 1.2140751676386906e-05, + "loss": 0.6544, + "step": 11027 + }, + { + "epoch": 0.4882724726874246, + "grad_norm": 1.6858546733856201, + "learning_rate": 1.2139242027791716e-05, + "loss": 0.7086, + "step": 11028 + }, + { + "epoch": 0.4883167483922385, + "grad_norm": 1.5920261144638062, + "learning_rate": 1.213773232810247e-05, + "loss": 0.6923, + "step": 11029 + }, + { + "epoch": 0.48836102409705234, + "grad_norm": 1.448889970779419, + "learning_rate": 1.2136222577355225e-05, + "loss": 0.6363, + "step": 11030 + }, + { + "epoch": 0.4884052998018662, + "grad_norm": 1.705683708190918, + "learning_rate": 1.2134712775586049e-05, + "loss": 0.8385, + "step": 11031 + }, + { + "epoch": 0.4884495755066801, + "grad_norm": 1.619309902191162, + "learning_rate": 1.2133202922830991e-05, + "loss": 0.77, + "step": 11032 + }, + { + "epoch": 0.48849385121149397, + "grad_norm": 1.4363102912902832, + "learning_rate": 1.2131693019126123e-05, + "loss": 0.7095, + "step": 11033 + }, + { + "epoch": 0.4885381269163078, + "grad_norm": 1.5641834735870361, + "learning_rate": 1.2130183064507501e-05, + "loss": 0.7078, + "step": 11034 + }, + { + "epoch": 0.48858240262112174, + "grad_norm": 1.508681297302246, + "learning_rate": 1.2128673059011189e-05, + "loss": 0.6598, + "step": 11035 + }, + { + "epoch": 0.4886266783259356, + "grad_norm": 1.5403108596801758, + "learning_rate": 1.2127163002673256e-05, + "loss": 0.6841, + "step": 11036 + }, + { + "epoch": 0.48867095403074945, + "grad_norm": 1.540047526359558, + "learning_rate": 1.2125652895529766e-05, + "loss": 0.6269, + "step": 11037 + }, + { + "epoch": 0.48871522973556336, + "grad_norm": 1.4854575395584106, + "learning_rate": 1.2124142737616792e-05, + "loss": 0.6476, + "step": 11038 + }, + { + "epoch": 0.4887595054403772, + "grad_norm": 1.4753185510635376, + "learning_rate": 1.2122632528970392e-05, + "loss": 0.6568, + "step": 11039 + }, + { + "epoch": 0.48880378114519113, + "grad_norm": 1.2675501108169556, + "learning_rate": 1.2121122269626644e-05, + "loss": 0.6003, + "step": 11040 + }, + { + "epoch": 0.488848056850005, + "grad_norm": 1.7295247316360474, + "learning_rate": 1.211961195962162e-05, + "loss": 0.7618, + "step": 11041 + }, + { + "epoch": 0.48889233255481884, + "grad_norm": 1.3092372417449951, + "learning_rate": 1.211810159899139e-05, + "loss": 0.6375, + "step": 11042 + }, + { + "epoch": 0.48893660825963275, + "grad_norm": 1.6534686088562012, + "learning_rate": 1.2116591187772026e-05, + "loss": 0.7835, + "step": 11043 + }, + { + "epoch": 0.4889808839644466, + "grad_norm": 1.522250771522522, + "learning_rate": 1.2115080725999604e-05, + "loss": 0.7161, + "step": 11044 + }, + { + "epoch": 0.48902515966926047, + "grad_norm": 1.264897346496582, + "learning_rate": 1.2113570213710202e-05, + "loss": 0.706, + "step": 11045 + }, + { + "epoch": 0.4890694353740744, + "grad_norm": 1.2691586017608643, + "learning_rate": 1.2112059650939895e-05, + "loss": 0.6118, + "step": 11046 + }, + { + "epoch": 0.48911371107888824, + "grad_norm": 1.4622406959533691, + "learning_rate": 1.2110549037724768e-05, + "loss": 0.7114, + "step": 11047 + }, + { + "epoch": 0.4891579867837021, + "grad_norm": 1.3381636142730713, + "learning_rate": 1.2109038374100893e-05, + "loss": 0.6843, + "step": 11048 + }, + { + "epoch": 0.489202262488516, + "grad_norm": 1.7120550870895386, + "learning_rate": 1.210752766010435e-05, + "loss": 0.7274, + "step": 11049 + }, + { + "epoch": 0.48924653819332986, + "grad_norm": 1.5003384351730347, + "learning_rate": 1.2106016895771226e-05, + "loss": 0.6763, + "step": 11050 + }, + { + "epoch": 0.4892908138981437, + "grad_norm": 1.9383898973464966, + "learning_rate": 1.2104506081137608e-05, + "loss": 0.7347, + "step": 11051 + }, + { + "epoch": 0.48933508960295763, + "grad_norm": 1.4190139770507812, + "learning_rate": 1.2102995216239572e-05, + "loss": 0.6977, + "step": 11052 + }, + { + "epoch": 0.4893793653077715, + "grad_norm": 1.3617409467697144, + "learning_rate": 1.2101484301113212e-05, + "loss": 0.6974, + "step": 11053 + }, + { + "epoch": 0.48942364101258534, + "grad_norm": 1.414503574371338, + "learning_rate": 1.2099973335794603e-05, + "loss": 0.7104, + "step": 11054 + }, + { + "epoch": 0.48946791671739925, + "grad_norm": 1.7073941230773926, + "learning_rate": 1.2098462320319846e-05, + "loss": 0.7152, + "step": 11055 + }, + { + "epoch": 0.4895121924222131, + "grad_norm": 1.3232721090316772, + "learning_rate": 1.2096951254725024e-05, + "loss": 0.6828, + "step": 11056 + }, + { + "epoch": 0.489556468127027, + "grad_norm": 1.7877907752990723, + "learning_rate": 1.2095440139046229e-05, + "loss": 0.7082, + "step": 11057 + }, + { + "epoch": 0.4896007438318409, + "grad_norm": 1.3622798919677734, + "learning_rate": 1.2093928973319554e-05, + "loss": 0.658, + "step": 11058 + }, + { + "epoch": 0.48964501953665474, + "grad_norm": 1.7267482280731201, + "learning_rate": 1.2092417757581085e-05, + "loss": 0.6984, + "step": 11059 + }, + { + "epoch": 0.48968929524146865, + "grad_norm": 1.3450320959091187, + "learning_rate": 1.2090906491866924e-05, + "loss": 0.7394, + "step": 11060 + }, + { + "epoch": 0.4897335709462825, + "grad_norm": 1.5801889896392822, + "learning_rate": 1.2089395176213166e-05, + "loss": 0.7545, + "step": 11061 + }, + { + "epoch": 0.48977784665109636, + "grad_norm": 1.8452879190444946, + "learning_rate": 1.2087883810655904e-05, + "loss": 0.715, + "step": 11062 + }, + { + "epoch": 0.48982212235591027, + "grad_norm": 1.6441291570663452, + "learning_rate": 1.208637239523124e-05, + "loss": 0.7742, + "step": 11063 + }, + { + "epoch": 0.48986639806072413, + "grad_norm": 1.1780983209609985, + "learning_rate": 1.2084860929975266e-05, + "loss": 0.5787, + "step": 11064 + }, + { + "epoch": 0.489910673765538, + "grad_norm": 1.4825395345687866, + "learning_rate": 1.2083349414924087e-05, + "loss": 0.6753, + "step": 11065 + }, + { + "epoch": 0.4899549494703519, + "grad_norm": 1.5838654041290283, + "learning_rate": 1.2081837850113809e-05, + "loss": 0.733, + "step": 11066 + }, + { + "epoch": 0.48999922517516575, + "grad_norm": 1.5545625686645508, + "learning_rate": 1.2080326235580521e-05, + "loss": 0.581, + "step": 11067 + }, + { + "epoch": 0.4900435008799796, + "grad_norm": 1.6769548654556274, + "learning_rate": 1.207881457136034e-05, + "loss": 0.6888, + "step": 11068 + }, + { + "epoch": 0.4900877765847935, + "grad_norm": 1.6768262386322021, + "learning_rate": 1.2077302857489363e-05, + "loss": 0.69, + "step": 11069 + }, + { + "epoch": 0.4901320522896074, + "grad_norm": 1.3816745281219482, + "learning_rate": 1.2075791094003702e-05, + "loss": 0.6283, + "step": 11070 + }, + { + "epoch": 0.49017632799442123, + "grad_norm": 1.4925966262817383, + "learning_rate": 1.2074279280939458e-05, + "loss": 0.6549, + "step": 11071 + }, + { + "epoch": 0.49022060369923515, + "grad_norm": 1.4943132400512695, + "learning_rate": 1.207276741833274e-05, + "loss": 0.7222, + "step": 11072 + }, + { + "epoch": 0.490264879404049, + "grad_norm": 1.4496577978134155, + "learning_rate": 1.2071255506219664e-05, + "loss": 0.6873, + "step": 11073 + }, + { + "epoch": 0.4903091551088629, + "grad_norm": 1.6271473169326782, + "learning_rate": 1.2069743544636334e-05, + "loss": 0.6936, + "step": 11074 + }, + { + "epoch": 0.49035343081367677, + "grad_norm": 1.559660792350769, + "learning_rate": 1.2068231533618868e-05, + "loss": 0.7279, + "step": 11075 + }, + { + "epoch": 0.49039770651849063, + "grad_norm": 1.3947638273239136, + "learning_rate": 1.2066719473203375e-05, + "loss": 0.6909, + "step": 11076 + }, + { + "epoch": 0.49044198222330454, + "grad_norm": 1.5859638452529907, + "learning_rate": 1.2065207363425966e-05, + "loss": 0.6991, + "step": 11077 + }, + { + "epoch": 0.4904862579281184, + "grad_norm": 1.2961310148239136, + "learning_rate": 1.2063695204322765e-05, + "loss": 0.6593, + "step": 11078 + }, + { + "epoch": 0.49053053363293225, + "grad_norm": 1.4035735130310059, + "learning_rate": 1.2062182995929883e-05, + "loss": 0.6692, + "step": 11079 + }, + { + "epoch": 0.49057480933774616, + "grad_norm": 1.6519317626953125, + "learning_rate": 1.206067073828344e-05, + "loss": 0.7302, + "step": 11080 + }, + { + "epoch": 0.49061908504256, + "grad_norm": 1.4721640348434448, + "learning_rate": 1.2059158431419555e-05, + "loss": 0.6283, + "step": 11081 + }, + { + "epoch": 0.4906633607473739, + "grad_norm": 1.3067294359207153, + "learning_rate": 1.2057646075374345e-05, + "loss": 0.6899, + "step": 11082 + }, + { + "epoch": 0.4907076364521878, + "grad_norm": 1.783348798751831, + "learning_rate": 1.2056133670183937e-05, + "loss": 0.7205, + "step": 11083 + }, + { + "epoch": 0.49075191215700165, + "grad_norm": 1.4286653995513916, + "learning_rate": 1.2054621215884445e-05, + "loss": 0.7016, + "step": 11084 + }, + { + "epoch": 0.4907961878618155, + "grad_norm": 1.266202449798584, + "learning_rate": 1.2053108712512003e-05, + "loss": 0.6907, + "step": 11085 + }, + { + "epoch": 0.4908404635666294, + "grad_norm": 1.2418440580368042, + "learning_rate": 1.205159616010273e-05, + "loss": 0.6215, + "step": 11086 + }, + { + "epoch": 0.49088473927144327, + "grad_norm": 1.532219648361206, + "learning_rate": 1.2050083558692751e-05, + "loss": 0.6519, + "step": 11087 + }, + { + "epoch": 0.4909290149762571, + "grad_norm": 1.4396618604660034, + "learning_rate": 1.20485709083182e-05, + "loss": 0.659, + "step": 11088 + }, + { + "epoch": 0.49097329068107104, + "grad_norm": 1.189086675643921, + "learning_rate": 1.2047058209015199e-05, + "loss": 0.5767, + "step": 11089 + }, + { + "epoch": 0.4910175663858849, + "grad_norm": 1.8234339952468872, + "learning_rate": 1.2045545460819876e-05, + "loss": 0.7361, + "step": 11090 + }, + { + "epoch": 0.4910618420906988, + "grad_norm": 1.2050917148590088, + "learning_rate": 1.204403266376837e-05, + "loss": 0.6428, + "step": 11091 + }, + { + "epoch": 0.49110611779551266, + "grad_norm": 1.3095495700836182, + "learning_rate": 1.2042519817896805e-05, + "loss": 0.6453, + "step": 11092 + }, + { + "epoch": 0.4911503935003265, + "grad_norm": 1.5505340099334717, + "learning_rate": 1.2041006923241318e-05, + "loss": 0.7473, + "step": 11093 + }, + { + "epoch": 0.49119466920514043, + "grad_norm": 1.2356418371200562, + "learning_rate": 1.2039493979838043e-05, + "loss": 0.6606, + "step": 11094 + }, + { + "epoch": 0.4912389449099543, + "grad_norm": 1.532431960105896, + "learning_rate": 1.2037980987723115e-05, + "loss": 0.6127, + "step": 11095 + }, + { + "epoch": 0.49128322061476815, + "grad_norm": 1.594609260559082, + "learning_rate": 1.203646794693267e-05, + "loss": 0.691, + "step": 11096 + }, + { + "epoch": 0.49132749631958206, + "grad_norm": 1.2294527292251587, + "learning_rate": 1.2034954857502845e-05, + "loss": 0.6031, + "step": 11097 + }, + { + "epoch": 0.4913717720243959, + "grad_norm": 1.210140585899353, + "learning_rate": 1.2033441719469783e-05, + "loss": 0.6544, + "step": 11098 + }, + { + "epoch": 0.49141604772920977, + "grad_norm": 1.6988205909729004, + "learning_rate": 1.203192853286962e-05, + "loss": 0.7665, + "step": 11099 + }, + { + "epoch": 0.4914603234340237, + "grad_norm": 1.6489765644073486, + "learning_rate": 1.2030415297738498e-05, + "loss": 0.6862, + "step": 11100 + }, + { + "epoch": 0.49150459913883754, + "grad_norm": 1.3115158081054688, + "learning_rate": 1.202890201411256e-05, + "loss": 0.6213, + "step": 11101 + }, + { + "epoch": 0.4915488748436514, + "grad_norm": 1.3662418127059937, + "learning_rate": 1.202738868202795e-05, + "loss": 0.7268, + "step": 11102 + }, + { + "epoch": 0.4915931505484653, + "grad_norm": 1.3520123958587646, + "learning_rate": 1.2025875301520811e-05, + "loss": 0.6964, + "step": 11103 + }, + { + "epoch": 0.49163742625327916, + "grad_norm": 1.299118995666504, + "learning_rate": 1.202436187262729e-05, + "loss": 0.6027, + "step": 11104 + }, + { + "epoch": 0.491681701958093, + "grad_norm": 1.2952295541763306, + "learning_rate": 1.2022848395383535e-05, + "loss": 0.6128, + "step": 11105 + }, + { + "epoch": 0.49172597766290693, + "grad_norm": 1.5648798942565918, + "learning_rate": 1.2021334869825693e-05, + "loss": 0.7114, + "step": 11106 + }, + { + "epoch": 0.4917702533677208, + "grad_norm": 1.3782716989517212, + "learning_rate": 1.2019821295989913e-05, + "loss": 0.6465, + "step": 11107 + }, + { + "epoch": 0.4918145290725347, + "grad_norm": 1.5358617305755615, + "learning_rate": 1.2018307673912346e-05, + "loss": 0.6395, + "step": 11108 + }, + { + "epoch": 0.49185880477734856, + "grad_norm": 1.6083437204360962, + "learning_rate": 1.2016794003629145e-05, + "loss": 0.6894, + "step": 11109 + }, + { + "epoch": 0.4919030804821624, + "grad_norm": 2.072875499725342, + "learning_rate": 1.2015280285176458e-05, + "loss": 0.7973, + "step": 11110 + }, + { + "epoch": 0.4919473561869763, + "grad_norm": 1.4779411554336548, + "learning_rate": 1.2013766518590444e-05, + "loss": 0.6038, + "step": 11111 + }, + { + "epoch": 0.4919916318917902, + "grad_norm": 1.4559698104858398, + "learning_rate": 1.2012252703907256e-05, + "loss": 0.6881, + "step": 11112 + }, + { + "epoch": 0.49203590759660404, + "grad_norm": 1.276094675064087, + "learning_rate": 1.2010738841163054e-05, + "loss": 0.6303, + "step": 11113 + }, + { + "epoch": 0.49208018330141795, + "grad_norm": 1.4970303773880005, + "learning_rate": 1.2009224930393989e-05, + "loss": 0.7066, + "step": 11114 + }, + { + "epoch": 0.4921244590062318, + "grad_norm": 1.2026348114013672, + "learning_rate": 1.2007710971636221e-05, + "loss": 0.6059, + "step": 11115 + }, + { + "epoch": 0.49216873471104566, + "grad_norm": 1.7819055318832397, + "learning_rate": 1.2006196964925913e-05, + "loss": 0.6625, + "step": 11116 + }, + { + "epoch": 0.4922130104158596, + "grad_norm": 1.322530746459961, + "learning_rate": 1.2004682910299225e-05, + "loss": 0.6776, + "step": 11117 + }, + { + "epoch": 0.49225728612067343, + "grad_norm": 1.6054123640060425, + "learning_rate": 1.2003168807792318e-05, + "loss": 0.6743, + "step": 11118 + }, + { + "epoch": 0.4923015618254873, + "grad_norm": 1.3730368614196777, + "learning_rate": 1.2001654657441352e-05, + "loss": 0.6033, + "step": 11119 + }, + { + "epoch": 0.4923458375303012, + "grad_norm": 1.3639901876449585, + "learning_rate": 1.2000140459282497e-05, + "loss": 0.6292, + "step": 11120 + }, + { + "epoch": 0.49239011323511506, + "grad_norm": 1.2923401594161987, + "learning_rate": 1.1998626213351914e-05, + "loss": 0.6455, + "step": 11121 + }, + { + "epoch": 0.4924343889399289, + "grad_norm": 1.510201096534729, + "learning_rate": 1.1997111919685773e-05, + "loss": 0.6957, + "step": 11122 + }, + { + "epoch": 0.4924786646447428, + "grad_norm": 2.22163987159729, + "learning_rate": 1.1995597578320241e-05, + "loss": 0.7105, + "step": 11123 + }, + { + "epoch": 0.4925229403495567, + "grad_norm": 1.4059134721755981, + "learning_rate": 1.1994083189291482e-05, + "loss": 0.6595, + "step": 11124 + }, + { + "epoch": 0.49256721605437054, + "grad_norm": 1.3572779893875122, + "learning_rate": 1.1992568752635669e-05, + "loss": 0.6478, + "step": 11125 + }, + { + "epoch": 0.49261149175918445, + "grad_norm": 1.4793689250946045, + "learning_rate": 1.1991054268388978e-05, + "loss": 0.6302, + "step": 11126 + }, + { + "epoch": 0.4926557674639983, + "grad_norm": 1.5070784091949463, + "learning_rate": 1.1989539736587577e-05, + "loss": 0.6973, + "step": 11127 + }, + { + "epoch": 0.4927000431688122, + "grad_norm": 1.5575292110443115, + "learning_rate": 1.1988025157267638e-05, + "loss": 0.6154, + "step": 11128 + }, + { + "epoch": 0.4927443188736261, + "grad_norm": 1.330863118171692, + "learning_rate": 1.1986510530465334e-05, + "loss": 0.6019, + "step": 11129 + }, + { + "epoch": 0.49278859457843993, + "grad_norm": 1.7554699182510376, + "learning_rate": 1.1984995856216845e-05, + "loss": 0.7086, + "step": 11130 + }, + { + "epoch": 0.49283287028325384, + "grad_norm": 1.4323457479476929, + "learning_rate": 1.1983481134558347e-05, + "loss": 0.6393, + "step": 11131 + }, + { + "epoch": 0.4928771459880677, + "grad_norm": 1.1860276460647583, + "learning_rate": 1.1981966365526018e-05, + "loss": 0.592, + "step": 11132 + }, + { + "epoch": 0.49292142169288156, + "grad_norm": 1.3418675661087036, + "learning_rate": 1.1980451549156038e-05, + "loss": 0.7486, + "step": 11133 + }, + { + "epoch": 0.49296569739769547, + "grad_norm": 1.7256768941879272, + "learning_rate": 1.197893668548458e-05, + "loss": 0.7201, + "step": 11134 + }, + { + "epoch": 0.4930099731025093, + "grad_norm": 1.7259352207183838, + "learning_rate": 1.1977421774547832e-05, + "loss": 0.6825, + "step": 11135 + }, + { + "epoch": 0.4930542488073232, + "grad_norm": 1.404464602470398, + "learning_rate": 1.1975906816381977e-05, + "loss": 0.778, + "step": 11136 + }, + { + "epoch": 0.4930985245121371, + "grad_norm": 1.8058772087097168, + "learning_rate": 1.1974391811023195e-05, + "loss": 0.7293, + "step": 11137 + }, + { + "epoch": 0.49314280021695095, + "grad_norm": 1.5169446468353271, + "learning_rate": 1.1972876758507677e-05, + "loss": 0.7237, + "step": 11138 + }, + { + "epoch": 0.4931870759217648, + "grad_norm": 1.3228049278259277, + "learning_rate": 1.1971361658871597e-05, + "loss": 0.6579, + "step": 11139 + }, + { + "epoch": 0.4932313516265787, + "grad_norm": 1.3085277080535889, + "learning_rate": 1.196984651215115e-05, + "loss": 0.6961, + "step": 11140 + }, + { + "epoch": 0.4932756273313926, + "grad_norm": 1.4010469913482666, + "learning_rate": 1.1968331318382529e-05, + "loss": 0.6945, + "step": 11141 + }, + { + "epoch": 0.49331990303620643, + "grad_norm": 1.3488960266113281, + "learning_rate": 1.196681607760191e-05, + "loss": 0.6263, + "step": 11142 + }, + { + "epoch": 0.49336417874102034, + "grad_norm": 1.4341307878494263, + "learning_rate": 1.1965300789845495e-05, + "loss": 0.7239, + "step": 11143 + }, + { + "epoch": 0.4934084544458342, + "grad_norm": 1.1933797597885132, + "learning_rate": 1.1963785455149467e-05, + "loss": 0.5986, + "step": 11144 + }, + { + "epoch": 0.4934527301506481, + "grad_norm": 1.3359758853912354, + "learning_rate": 1.1962270073550026e-05, + "loss": 0.677, + "step": 11145 + }, + { + "epoch": 0.49349700585546197, + "grad_norm": 1.3211884498596191, + "learning_rate": 1.1960754645083364e-05, + "loss": 0.6777, + "step": 11146 + }, + { + "epoch": 0.4935412815602758, + "grad_norm": 1.4036550521850586, + "learning_rate": 1.1959239169785668e-05, + "loss": 0.6399, + "step": 11147 + }, + { + "epoch": 0.49358555726508974, + "grad_norm": 1.5944393873214722, + "learning_rate": 1.1957723647693142e-05, + "loss": 0.662, + "step": 11148 + }, + { + "epoch": 0.4936298329699036, + "grad_norm": 1.313044786453247, + "learning_rate": 1.195620807884198e-05, + "loss": 0.629, + "step": 11149 + }, + { + "epoch": 0.49367410867471745, + "grad_norm": 2.1592471599578857, + "learning_rate": 1.1954692463268383e-05, + "loss": 0.7631, + "step": 11150 + }, + { + "epoch": 0.49371838437953136, + "grad_norm": 1.2845693826675415, + "learning_rate": 1.1953176801008545e-05, + "loss": 0.6539, + "step": 11151 + }, + { + "epoch": 0.4937626600843452, + "grad_norm": 1.5209035873413086, + "learning_rate": 1.1951661092098673e-05, + "loss": 0.6821, + "step": 11152 + }, + { + "epoch": 0.4938069357891591, + "grad_norm": 1.272153377532959, + "learning_rate": 1.1950145336574963e-05, + "loss": 0.6112, + "step": 11153 + }, + { + "epoch": 0.493851211493973, + "grad_norm": 1.5331164598464966, + "learning_rate": 1.1948629534473618e-05, + "loss": 0.7172, + "step": 11154 + }, + { + "epoch": 0.49389548719878684, + "grad_norm": 1.2053776979446411, + "learning_rate": 1.1947113685830844e-05, + "loss": 0.6115, + "step": 11155 + }, + { + "epoch": 0.4939397629036007, + "grad_norm": 1.7214435338974, + "learning_rate": 1.1945597790682849e-05, + "loss": 0.7341, + "step": 11156 + }, + { + "epoch": 0.4939840386084146, + "grad_norm": 1.4369006156921387, + "learning_rate": 1.1944081849065829e-05, + "loss": 0.6964, + "step": 11157 + }, + { + "epoch": 0.49402831431322847, + "grad_norm": 1.3473283052444458, + "learning_rate": 1.1942565861016002e-05, + "loss": 0.6295, + "step": 11158 + }, + { + "epoch": 0.4940725900180423, + "grad_norm": 1.788926124572754, + "learning_rate": 1.1941049826569567e-05, + "loss": 0.7539, + "step": 11159 + }, + { + "epoch": 0.49411686572285624, + "grad_norm": 1.1657373905181885, + "learning_rate": 1.1939533745762738e-05, + "loss": 0.5792, + "step": 11160 + }, + { + "epoch": 0.4941611414276701, + "grad_norm": 1.5722205638885498, + "learning_rate": 1.1938017618631727e-05, + "loss": 0.6544, + "step": 11161 + }, + { + "epoch": 0.494205417132484, + "grad_norm": 1.307061791419983, + "learning_rate": 1.1936501445212743e-05, + "loss": 0.6883, + "step": 11162 + }, + { + "epoch": 0.49424969283729786, + "grad_norm": 1.254435658454895, + "learning_rate": 1.1934985225541998e-05, + "loss": 0.634, + "step": 11163 + }, + { + "epoch": 0.4942939685421117, + "grad_norm": 1.54322350025177, + "learning_rate": 1.1933468959655706e-05, + "loss": 0.6856, + "step": 11164 + }, + { + "epoch": 0.49433824424692563, + "grad_norm": 1.6941096782684326, + "learning_rate": 1.1931952647590084e-05, + "loss": 0.6995, + "step": 11165 + }, + { + "epoch": 0.4943825199517395, + "grad_norm": 1.238797664642334, + "learning_rate": 1.1930436289381344e-05, + "loss": 0.7112, + "step": 11166 + }, + { + "epoch": 0.49442679565655334, + "grad_norm": 1.3712682723999023, + "learning_rate": 1.1928919885065705e-05, + "loss": 0.7011, + "step": 11167 + }, + { + "epoch": 0.49447107136136725, + "grad_norm": 1.853389859199524, + "learning_rate": 1.1927403434679386e-05, + "loss": 0.7228, + "step": 11168 + }, + { + "epoch": 0.4945153470661811, + "grad_norm": 1.3562673330307007, + "learning_rate": 1.1925886938258606e-05, + "loss": 0.6855, + "step": 11169 + }, + { + "epoch": 0.49455962277099497, + "grad_norm": 1.6448864936828613, + "learning_rate": 1.1924370395839584e-05, + "loss": 0.7304, + "step": 11170 + }, + { + "epoch": 0.4946038984758089, + "grad_norm": 1.441683292388916, + "learning_rate": 1.1922853807458542e-05, + "loss": 0.683, + "step": 11171 + }, + { + "epoch": 0.49464817418062274, + "grad_norm": 1.783807635307312, + "learning_rate": 1.19213371731517e-05, + "loss": 0.7575, + "step": 11172 + }, + { + "epoch": 0.4946924498854366, + "grad_norm": 1.667982578277588, + "learning_rate": 1.1919820492955285e-05, + "loss": 0.7855, + "step": 11173 + }, + { + "epoch": 0.4947367255902505, + "grad_norm": 1.4708991050720215, + "learning_rate": 1.1918303766905523e-05, + "loss": 0.6584, + "step": 11174 + }, + { + "epoch": 0.49478100129506436, + "grad_norm": 1.2771227359771729, + "learning_rate": 1.1916786995038637e-05, + "loss": 0.6532, + "step": 11175 + }, + { + "epoch": 0.4948252769998782, + "grad_norm": 1.445719838142395, + "learning_rate": 1.1915270177390851e-05, + "loss": 0.7306, + "step": 11176 + }, + { + "epoch": 0.49486955270469213, + "grad_norm": 1.2895139455795288, + "learning_rate": 1.1913753313998399e-05, + "loss": 0.6473, + "step": 11177 + }, + { + "epoch": 0.494913828409506, + "grad_norm": 1.3171521425247192, + "learning_rate": 1.1912236404897508e-05, + "loss": 0.6752, + "step": 11178 + }, + { + "epoch": 0.4949581041143199, + "grad_norm": 1.3487266302108765, + "learning_rate": 1.1910719450124406e-05, + "loss": 0.653, + "step": 11179 + }, + { + "epoch": 0.49500237981913375, + "grad_norm": 1.4639028310775757, + "learning_rate": 1.190920244971533e-05, + "loss": 0.6348, + "step": 11180 + }, + { + "epoch": 0.4950466555239476, + "grad_norm": 1.5013989210128784, + "learning_rate": 1.1907685403706501e-05, + "loss": 0.7339, + "step": 11181 + }, + { + "epoch": 0.4950909312287615, + "grad_norm": 1.5827412605285645, + "learning_rate": 1.190616831213416e-05, + "loss": 0.6845, + "step": 11182 + }, + { + "epoch": 0.4951352069335754, + "grad_norm": 1.5800142288208008, + "learning_rate": 1.1904651175034548e-05, + "loss": 0.6981, + "step": 11183 + }, + { + "epoch": 0.49517948263838923, + "grad_norm": 1.4703830480575562, + "learning_rate": 1.190313399244389e-05, + "loss": 0.6813, + "step": 11184 + }, + { + "epoch": 0.49522375834320315, + "grad_norm": 1.284538984298706, + "learning_rate": 1.1901616764398426e-05, + "loss": 0.724, + "step": 11185 + }, + { + "epoch": 0.495268034048017, + "grad_norm": 1.4772627353668213, + "learning_rate": 1.190009949093439e-05, + "loss": 0.6233, + "step": 11186 + }, + { + "epoch": 0.49531230975283086, + "grad_norm": 1.3069664239883423, + "learning_rate": 1.1898582172088027e-05, + "loss": 0.6294, + "step": 11187 + }, + { + "epoch": 0.49535658545764477, + "grad_norm": 1.474084496498108, + "learning_rate": 1.1897064807895578e-05, + "loss": 0.7637, + "step": 11188 + }, + { + "epoch": 0.49540086116245863, + "grad_norm": 1.3204482793807983, + "learning_rate": 1.1895547398393278e-05, + "loss": 0.6037, + "step": 11189 + }, + { + "epoch": 0.4954451368672725, + "grad_norm": 1.4737660884857178, + "learning_rate": 1.1894029943617374e-05, + "loss": 0.7168, + "step": 11190 + }, + { + "epoch": 0.4954894125720864, + "grad_norm": 1.463172197341919, + "learning_rate": 1.1892512443604103e-05, + "loss": 0.6705, + "step": 11191 + }, + { + "epoch": 0.49553368827690025, + "grad_norm": 1.5304679870605469, + "learning_rate": 1.1890994898389714e-05, + "loss": 0.6897, + "step": 11192 + }, + { + "epoch": 0.4955779639817141, + "grad_norm": 1.5164003372192383, + "learning_rate": 1.1889477308010457e-05, + "loss": 0.7054, + "step": 11193 + }, + { + "epoch": 0.495622239686528, + "grad_norm": 1.394006371498108, + "learning_rate": 1.1887959672502567e-05, + "loss": 0.6061, + "step": 11194 + }, + { + "epoch": 0.4956665153913419, + "grad_norm": 1.4956448078155518, + "learning_rate": 1.18864419919023e-05, + "loss": 0.5928, + "step": 11195 + }, + { + "epoch": 0.4957107910961558, + "grad_norm": 1.329388976097107, + "learning_rate": 1.18849242662459e-05, + "loss": 0.6667, + "step": 11196 + }, + { + "epoch": 0.49575506680096965, + "grad_norm": 1.377658486366272, + "learning_rate": 1.1883406495569623e-05, + "loss": 0.6913, + "step": 11197 + }, + { + "epoch": 0.4957993425057835, + "grad_norm": 1.4799540042877197, + "learning_rate": 1.1881888679909712e-05, + "loss": 0.6606, + "step": 11198 + }, + { + "epoch": 0.4958436182105974, + "grad_norm": 1.510178565979004, + "learning_rate": 1.1880370819302422e-05, + "loss": 0.7316, + "step": 11199 + }, + { + "epoch": 0.49588789391541127, + "grad_norm": 1.707234263420105, + "learning_rate": 1.1878852913784007e-05, + "loss": 0.7003, + "step": 11200 + }, + { + "epoch": 0.4959321696202251, + "grad_norm": 1.7218449115753174, + "learning_rate": 1.1877334963390719e-05, + "loss": 0.7421, + "step": 11201 + }, + { + "epoch": 0.49597644532503904, + "grad_norm": 1.2512702941894531, + "learning_rate": 1.1875816968158816e-05, + "loss": 0.6369, + "step": 11202 + }, + { + "epoch": 0.4960207210298529, + "grad_norm": 1.6164172887802124, + "learning_rate": 1.187429892812455e-05, + "loss": 0.7743, + "step": 11203 + }, + { + "epoch": 0.49606499673466675, + "grad_norm": 1.4298489093780518, + "learning_rate": 1.187278084332418e-05, + "loss": 0.6275, + "step": 11204 + }, + { + "epoch": 0.49610927243948066, + "grad_norm": 1.3506580591201782, + "learning_rate": 1.1871262713793967e-05, + "loss": 0.6723, + "step": 11205 + }, + { + "epoch": 0.4961535481442945, + "grad_norm": 1.5990207195281982, + "learning_rate": 1.1869744539570165e-05, + "loss": 0.6923, + "step": 11206 + }, + { + "epoch": 0.4961978238491084, + "grad_norm": 1.2214490175247192, + "learning_rate": 1.1868226320689037e-05, + "loss": 0.5805, + "step": 11207 + }, + { + "epoch": 0.4962420995539223, + "grad_norm": 1.2817440032958984, + "learning_rate": 1.1866708057186848e-05, + "loss": 0.6331, + "step": 11208 + }, + { + "epoch": 0.49628637525873615, + "grad_norm": 1.6033917665481567, + "learning_rate": 1.1865189749099853e-05, + "loss": 0.7022, + "step": 11209 + }, + { + "epoch": 0.49633065096355, + "grad_norm": 1.5461493730545044, + "learning_rate": 1.1863671396464322e-05, + "loss": 0.7185, + "step": 11210 + }, + { + "epoch": 0.4963749266683639, + "grad_norm": 1.331790566444397, + "learning_rate": 1.1862152999316517e-05, + "loss": 0.5916, + "step": 11211 + }, + { + "epoch": 0.49641920237317777, + "grad_norm": 1.4210941791534424, + "learning_rate": 1.1860634557692702e-05, + "loss": 0.6904, + "step": 11212 + }, + { + "epoch": 0.4964634780779916, + "grad_norm": 1.5451323986053467, + "learning_rate": 1.1859116071629148e-05, + "loss": 0.7032, + "step": 11213 + }, + { + "epoch": 0.49650775378280554, + "grad_norm": 1.4856297969818115, + "learning_rate": 1.1857597541162118e-05, + "loss": 0.7519, + "step": 11214 + }, + { + "epoch": 0.4965520294876194, + "grad_norm": 1.3494956493377686, + "learning_rate": 1.1856078966327886e-05, + "loss": 0.6397, + "step": 11215 + }, + { + "epoch": 0.4965963051924333, + "grad_norm": 1.7137831449508667, + "learning_rate": 1.1854560347162715e-05, + "loss": 0.7455, + "step": 11216 + }, + { + "epoch": 0.49664058089724716, + "grad_norm": 1.3781477212905884, + "learning_rate": 1.1853041683702884e-05, + "loss": 0.6971, + "step": 11217 + }, + { + "epoch": 0.496684856602061, + "grad_norm": 1.8324980735778809, + "learning_rate": 1.1851522975984659e-05, + "loss": 0.7178, + "step": 11218 + }, + { + "epoch": 0.49672913230687493, + "grad_norm": 1.9677364826202393, + "learning_rate": 1.1850004224044315e-05, + "loss": 0.7375, + "step": 11219 + }, + { + "epoch": 0.4967734080116888, + "grad_norm": 1.299181580543518, + "learning_rate": 1.184848542791813e-05, + "loss": 0.6869, + "step": 11220 + }, + { + "epoch": 0.49681768371650264, + "grad_norm": 1.298330307006836, + "learning_rate": 1.1846966587642374e-05, + "loss": 0.6819, + "step": 11221 + }, + { + "epoch": 0.49686195942131656, + "grad_norm": 1.220902681350708, + "learning_rate": 1.1845447703253323e-05, + "loss": 0.6615, + "step": 11222 + }, + { + "epoch": 0.4969062351261304, + "grad_norm": 1.5015885829925537, + "learning_rate": 1.184392877478726e-05, + "loss": 0.7387, + "step": 11223 + }, + { + "epoch": 0.49695051083094427, + "grad_norm": 1.4432040452957153, + "learning_rate": 1.1842409802280454e-05, + "loss": 0.6234, + "step": 11224 + }, + { + "epoch": 0.4969947865357582, + "grad_norm": 1.6271353960037231, + "learning_rate": 1.1840890785769195e-05, + "loss": 0.6995, + "step": 11225 + }, + { + "epoch": 0.49703906224057204, + "grad_norm": 1.7718931436538696, + "learning_rate": 1.1839371725289757e-05, + "loss": 0.7407, + "step": 11226 + }, + { + "epoch": 0.4970833379453859, + "grad_norm": 1.1742362976074219, + "learning_rate": 1.1837852620878421e-05, + "loss": 0.5703, + "step": 11227 + }, + { + "epoch": 0.4971276136501998, + "grad_norm": 1.2883955240249634, + "learning_rate": 1.1836333472571473e-05, + "loss": 0.6178, + "step": 11228 + }, + { + "epoch": 0.49717188935501366, + "grad_norm": 1.6414800882339478, + "learning_rate": 1.1834814280405194e-05, + "loss": 0.6711, + "step": 11229 + }, + { + "epoch": 0.4972161650598275, + "grad_norm": 1.7451586723327637, + "learning_rate": 1.1833295044415874e-05, + "loss": 0.7478, + "step": 11230 + }, + { + "epoch": 0.49726044076464143, + "grad_norm": 1.2230169773101807, + "learning_rate": 1.183177576463979e-05, + "loss": 0.6963, + "step": 11231 + }, + { + "epoch": 0.4973047164694553, + "grad_norm": 1.3788111209869385, + "learning_rate": 1.1830256441113235e-05, + "loss": 0.6625, + "step": 11232 + }, + { + "epoch": 0.4973489921742692, + "grad_norm": 1.3003816604614258, + "learning_rate": 1.1828737073872497e-05, + "loss": 0.6627, + "step": 11233 + }, + { + "epoch": 0.49739326787908306, + "grad_norm": 1.3063591718673706, + "learning_rate": 1.1827217662953859e-05, + "loss": 0.7013, + "step": 11234 + }, + { + "epoch": 0.4974375435838969, + "grad_norm": 1.4145621061325073, + "learning_rate": 1.182569820839362e-05, + "loss": 0.6838, + "step": 11235 + }, + { + "epoch": 0.4974818192887108, + "grad_norm": 1.370713710784912, + "learning_rate": 1.1824178710228062e-05, + "loss": 0.7142, + "step": 11236 + }, + { + "epoch": 0.4975260949935247, + "grad_norm": 1.5120998620986938, + "learning_rate": 1.182265916849348e-05, + "loss": 0.6449, + "step": 11237 + }, + { + "epoch": 0.49757037069833854, + "grad_norm": 1.4274941682815552, + "learning_rate": 1.1821139583226172e-05, + "loss": 0.6736, + "step": 11238 + }, + { + "epoch": 0.49761464640315245, + "grad_norm": 1.2488887310028076, + "learning_rate": 1.1819619954462422e-05, + "loss": 0.5879, + "step": 11239 + }, + { + "epoch": 0.4976589221079663, + "grad_norm": 1.3047295808792114, + "learning_rate": 1.1818100282238538e-05, + "loss": 0.6816, + "step": 11240 + }, + { + "epoch": 0.49770319781278016, + "grad_norm": 1.6141351461410522, + "learning_rate": 1.1816580566590804e-05, + "loss": 0.7086, + "step": 11241 + }, + { + "epoch": 0.4977474735175941, + "grad_norm": 1.5891239643096924, + "learning_rate": 1.1815060807555524e-05, + "loss": 0.6314, + "step": 11242 + }, + { + "epoch": 0.49779174922240793, + "grad_norm": 1.8705439567565918, + "learning_rate": 1.1813541005168993e-05, + "loss": 0.6947, + "step": 11243 + }, + { + "epoch": 0.4978360249272218, + "grad_norm": 1.3482939004898071, + "learning_rate": 1.1812021159467516e-05, + "loss": 0.6287, + "step": 11244 + }, + { + "epoch": 0.4978803006320357, + "grad_norm": 1.426479458808899, + "learning_rate": 1.1810501270487387e-05, + "loss": 0.6228, + "step": 11245 + }, + { + "epoch": 0.49792457633684956, + "grad_norm": 1.3964952230453491, + "learning_rate": 1.1808981338264906e-05, + "loss": 0.6555, + "step": 11246 + }, + { + "epoch": 0.4979688520416634, + "grad_norm": 1.3052576780319214, + "learning_rate": 1.1807461362836382e-05, + "loss": 0.6143, + "step": 11247 + }, + { + "epoch": 0.4980131277464773, + "grad_norm": 1.6799366474151611, + "learning_rate": 1.1805941344238114e-05, + "loss": 0.7461, + "step": 11248 + }, + { + "epoch": 0.4980574034512912, + "grad_norm": 1.4357030391693115, + "learning_rate": 1.1804421282506409e-05, + "loss": 0.6106, + "step": 11249 + }, + { + "epoch": 0.4981016791561051, + "grad_norm": 2.088606119155884, + "learning_rate": 1.1802901177677568e-05, + "loss": 0.7895, + "step": 11250 + }, + { + "epoch": 0.49814595486091895, + "grad_norm": 1.2842084169387817, + "learning_rate": 1.1801381029787902e-05, + "loss": 0.6161, + "step": 11251 + }, + { + "epoch": 0.4981902305657328, + "grad_norm": 1.29091215133667, + "learning_rate": 1.1799860838873715e-05, + "loss": 0.6899, + "step": 11252 + }, + { + "epoch": 0.4982345062705467, + "grad_norm": 1.3578169345855713, + "learning_rate": 1.1798340604971319e-05, + "loss": 0.6733, + "step": 11253 + }, + { + "epoch": 0.4982787819753606, + "grad_norm": 1.3479546308517456, + "learning_rate": 1.1796820328117025e-05, + "loss": 0.6911, + "step": 11254 + }, + { + "epoch": 0.49832305768017443, + "grad_norm": 1.3395129442214966, + "learning_rate": 1.1795300008347137e-05, + "loss": 0.7037, + "step": 11255 + }, + { + "epoch": 0.49836733338498834, + "grad_norm": 1.3716864585876465, + "learning_rate": 1.1793779645697969e-05, + "loss": 0.7208, + "step": 11256 + }, + { + "epoch": 0.4984116090898022, + "grad_norm": 1.6612902879714966, + "learning_rate": 1.1792259240205835e-05, + "loss": 0.7134, + "step": 11257 + }, + { + "epoch": 0.49845588479461606, + "grad_norm": 1.4796733856201172, + "learning_rate": 1.179073879190705e-05, + "loss": 0.685, + "step": 11258 + }, + { + "epoch": 0.49850016049942997, + "grad_norm": 1.405747890472412, + "learning_rate": 1.1789218300837929e-05, + "loss": 0.6785, + "step": 11259 + }, + { + "epoch": 0.4985444362042438, + "grad_norm": 1.3999369144439697, + "learning_rate": 1.1787697767034782e-05, + "loss": 0.6713, + "step": 11260 + }, + { + "epoch": 0.4985887119090577, + "grad_norm": 1.5363396406173706, + "learning_rate": 1.178617719053393e-05, + "loss": 0.6912, + "step": 11261 + }, + { + "epoch": 0.4986329876138716, + "grad_norm": 1.298639178276062, + "learning_rate": 1.1784656571371691e-05, + "loss": 0.6137, + "step": 11262 + }, + { + "epoch": 0.49867726331868545, + "grad_norm": 1.77785062789917, + "learning_rate": 1.1783135909584382e-05, + "loss": 0.739, + "step": 11263 + }, + { + "epoch": 0.4987215390234993, + "grad_norm": 1.5600050687789917, + "learning_rate": 1.1781615205208326e-05, + "loss": 0.6848, + "step": 11264 + }, + { + "epoch": 0.4987658147283132, + "grad_norm": 1.6392433643341064, + "learning_rate": 1.178009445827984e-05, + "loss": 0.7251, + "step": 11265 + }, + { + "epoch": 0.4988100904331271, + "grad_norm": 1.5624786615371704, + "learning_rate": 1.1778573668835247e-05, + "loss": 0.6893, + "step": 11266 + }, + { + "epoch": 0.498854366137941, + "grad_norm": 1.2067238092422485, + "learning_rate": 1.1777052836910871e-05, + "loss": 0.6562, + "step": 11267 + }, + { + "epoch": 0.49889864184275484, + "grad_norm": 1.3519648313522339, + "learning_rate": 1.1775531962543036e-05, + "loss": 0.6523, + "step": 11268 + }, + { + "epoch": 0.4989429175475687, + "grad_norm": 1.1905698776245117, + "learning_rate": 1.1774011045768061e-05, + "loss": 0.5931, + "step": 11269 + }, + { + "epoch": 0.4989871932523826, + "grad_norm": 1.4784016609191895, + "learning_rate": 1.177249008662228e-05, + "loss": 0.6535, + "step": 11270 + }, + { + "epoch": 0.49903146895719647, + "grad_norm": 1.4918577671051025, + "learning_rate": 1.1770969085142017e-05, + "loss": 0.6948, + "step": 11271 + }, + { + "epoch": 0.4990757446620103, + "grad_norm": 1.322144865989685, + "learning_rate": 1.1769448041363598e-05, + "loss": 0.6044, + "step": 11272 + }, + { + "epoch": 0.49912002036682424, + "grad_norm": 1.2411726713180542, + "learning_rate": 1.1767926955323357e-05, + "loss": 0.6083, + "step": 11273 + }, + { + "epoch": 0.4991642960716381, + "grad_norm": 1.6634104251861572, + "learning_rate": 1.1766405827057615e-05, + "loss": 0.6967, + "step": 11274 + }, + { + "epoch": 0.49920857177645195, + "grad_norm": 1.3200162649154663, + "learning_rate": 1.1764884656602711e-05, + "loss": 0.6313, + "step": 11275 + }, + { + "epoch": 0.49925284748126586, + "grad_norm": 1.9087179899215698, + "learning_rate": 1.1763363443994971e-05, + "loss": 0.8066, + "step": 11276 + }, + { + "epoch": 0.4992971231860797, + "grad_norm": 1.2970441579818726, + "learning_rate": 1.1761842189270733e-05, + "loss": 0.6574, + "step": 11277 + }, + { + "epoch": 0.4993413988908936, + "grad_norm": 1.523232102394104, + "learning_rate": 1.176032089246633e-05, + "loss": 0.7146, + "step": 11278 + }, + { + "epoch": 0.4993856745957075, + "grad_norm": 1.2682088613510132, + "learning_rate": 1.1758799553618094e-05, + "loss": 0.6179, + "step": 11279 + }, + { + "epoch": 0.49942995030052134, + "grad_norm": 1.3620917797088623, + "learning_rate": 1.1757278172762363e-05, + "loss": 0.6901, + "step": 11280 + }, + { + "epoch": 0.4994742260053352, + "grad_norm": 1.3241266012191772, + "learning_rate": 1.1755756749935475e-05, + "loss": 0.6592, + "step": 11281 + }, + { + "epoch": 0.4995185017101491, + "grad_norm": 1.4706083536148071, + "learning_rate": 1.1754235285173765e-05, + "loss": 0.6202, + "step": 11282 + }, + { + "epoch": 0.49956277741496297, + "grad_norm": 1.4105656147003174, + "learning_rate": 1.1752713778513576e-05, + "loss": 0.7102, + "step": 11283 + }, + { + "epoch": 0.4996070531197768, + "grad_norm": 1.7700014114379883, + "learning_rate": 1.1751192229991241e-05, + "loss": 0.6057, + "step": 11284 + }, + { + "epoch": 0.49965132882459073, + "grad_norm": 1.3499608039855957, + "learning_rate": 1.174967063964311e-05, + "loss": 0.6243, + "step": 11285 + }, + { + "epoch": 0.4996956045294046, + "grad_norm": 1.2489773035049438, + "learning_rate": 1.1748149007505516e-05, + "loss": 0.6338, + "step": 11286 + }, + { + "epoch": 0.4997398802342185, + "grad_norm": 1.3465847969055176, + "learning_rate": 1.1746627333614812e-05, + "loss": 0.6595, + "step": 11287 + }, + { + "epoch": 0.49978415593903236, + "grad_norm": 1.3367478847503662, + "learning_rate": 1.1745105618007333e-05, + "loss": 0.6067, + "step": 11288 + }, + { + "epoch": 0.4998284316438462, + "grad_norm": 1.5012727975845337, + "learning_rate": 1.1743583860719425e-05, + "loss": 0.711, + "step": 11289 + }, + { + "epoch": 0.49987270734866013, + "grad_norm": 1.379895806312561, + "learning_rate": 1.174206206178744e-05, + "loss": 0.645, + "step": 11290 + }, + { + "epoch": 0.499916983053474, + "grad_norm": 2.4057278633117676, + "learning_rate": 1.1740540221247718e-05, + "loss": 0.639, + "step": 11291 + }, + { + "epoch": 0.49996125875828784, + "grad_norm": 1.445935606956482, + "learning_rate": 1.1739018339136617e-05, + "loss": 0.706, + "step": 11292 + }, + { + "epoch": 0.5000055344631017, + "grad_norm": 1.8198286294937134, + "learning_rate": 1.173749641549047e-05, + "loss": 0.7654, + "step": 11293 + }, + { + "epoch": 0.5000498101679156, + "grad_norm": 1.9200795888900757, + "learning_rate": 1.1735974450345639e-05, + "loss": 0.6996, + "step": 11294 + }, + { + "epoch": 0.5000940858727295, + "grad_norm": 1.286611795425415, + "learning_rate": 1.1734452443738473e-05, + "loss": 0.7145, + "step": 11295 + }, + { + "epoch": 0.5001383615775433, + "grad_norm": 1.4974104166030884, + "learning_rate": 1.1732930395705324e-05, + "loss": 0.6949, + "step": 11296 + }, + { + "epoch": 0.5001826372823572, + "grad_norm": 1.3904379606246948, + "learning_rate": 1.173140830628254e-05, + "loss": 0.6735, + "step": 11297 + }, + { + "epoch": 0.5002269129871711, + "grad_norm": 1.4573743343353271, + "learning_rate": 1.172988617550648e-05, + "loss": 0.5824, + "step": 11298 + }, + { + "epoch": 0.500271188691985, + "grad_norm": 1.6313562393188477, + "learning_rate": 1.1728364003413495e-05, + "loss": 0.6652, + "step": 11299 + }, + { + "epoch": 0.5003154643967989, + "grad_norm": 1.3683545589447021, + "learning_rate": 1.1726841790039947e-05, + "loss": 0.6882, + "step": 11300 + }, + { + "epoch": 0.5003597401016128, + "grad_norm": 1.4834072589874268, + "learning_rate": 1.1725319535422189e-05, + "loss": 0.6583, + "step": 11301 + }, + { + "epoch": 0.5004040158064266, + "grad_norm": 2.05535626411438, + "learning_rate": 1.1723797239596576e-05, + "loss": 0.866, + "step": 11302 + }, + { + "epoch": 0.5004482915112405, + "grad_norm": 1.3308162689208984, + "learning_rate": 1.1722274902599469e-05, + "loss": 0.6977, + "step": 11303 + }, + { + "epoch": 0.5004925672160544, + "grad_norm": 1.3318175077438354, + "learning_rate": 1.1720752524467227e-05, + "loss": 0.6676, + "step": 11304 + }, + { + "epoch": 0.5005368429208682, + "grad_norm": 1.5141149759292603, + "learning_rate": 1.1719230105236217e-05, + "loss": 0.7028, + "step": 11305 + }, + { + "epoch": 0.5005811186256821, + "grad_norm": 1.2215362787246704, + "learning_rate": 1.1717707644942792e-05, + "loss": 0.6048, + "step": 11306 + }, + { + "epoch": 0.500625394330496, + "grad_norm": 1.2569303512573242, + "learning_rate": 1.1716185143623322e-05, + "loss": 0.689, + "step": 11307 + }, + { + "epoch": 0.5006696700353098, + "grad_norm": 1.3665599822998047, + "learning_rate": 1.1714662601314164e-05, + "loss": 0.6564, + "step": 11308 + }, + { + "epoch": 0.5007139457401237, + "grad_norm": 1.0857458114624023, + "learning_rate": 1.1713140018051686e-05, + "loss": 0.5772, + "step": 11309 + }, + { + "epoch": 0.5007582214449376, + "grad_norm": 1.4631843566894531, + "learning_rate": 1.1711617393872255e-05, + "loss": 0.6509, + "step": 11310 + }, + { + "epoch": 0.5008024971497516, + "grad_norm": 1.4963921308517456, + "learning_rate": 1.1710094728812235e-05, + "loss": 0.7303, + "step": 11311 + }, + { + "epoch": 0.5008467728545654, + "grad_norm": 1.5483735799789429, + "learning_rate": 1.1708572022907996e-05, + "loss": 0.7469, + "step": 11312 + }, + { + "epoch": 0.5008910485593793, + "grad_norm": 1.5777170658111572, + "learning_rate": 1.1707049276195903e-05, + "loss": 0.6288, + "step": 11313 + }, + { + "epoch": 0.5009353242641932, + "grad_norm": 1.3359837532043457, + "learning_rate": 1.1705526488712328e-05, + "loss": 0.6577, + "step": 11314 + }, + { + "epoch": 0.500979599969007, + "grad_norm": 1.4692134857177734, + "learning_rate": 1.1704003660493646e-05, + "loss": 0.72, + "step": 11315 + }, + { + "epoch": 0.5010238756738209, + "grad_norm": 1.5004873275756836, + "learning_rate": 1.1702480791576223e-05, + "loss": 0.6503, + "step": 11316 + }, + { + "epoch": 0.5010681513786348, + "grad_norm": 1.7483633756637573, + "learning_rate": 1.170095788199643e-05, + "loss": 0.704, + "step": 11317 + }, + { + "epoch": 0.5011124270834486, + "grad_norm": 1.239375352859497, + "learning_rate": 1.1699434931790643e-05, + "loss": 0.6061, + "step": 11318 + }, + { + "epoch": 0.5011567027882625, + "grad_norm": 1.5569826364517212, + "learning_rate": 1.1697911940995237e-05, + "loss": 0.7067, + "step": 11319 + }, + { + "epoch": 0.5012009784930764, + "grad_norm": 1.3597155809402466, + "learning_rate": 1.169638890964659e-05, + "loss": 0.6214, + "step": 11320 + }, + { + "epoch": 0.5012452541978902, + "grad_norm": 1.5645146369934082, + "learning_rate": 1.1694865837781069e-05, + "loss": 0.6706, + "step": 11321 + }, + { + "epoch": 0.5012895299027041, + "grad_norm": 1.530102252960205, + "learning_rate": 1.1693342725435063e-05, + "loss": 0.6998, + "step": 11322 + }, + { + "epoch": 0.5013338056075181, + "grad_norm": 1.4192783832550049, + "learning_rate": 1.1691819572644941e-05, + "loss": 0.7523, + "step": 11323 + }, + { + "epoch": 0.5013780813123319, + "grad_norm": 1.6238218545913696, + "learning_rate": 1.1690296379447089e-05, + "loss": 0.714, + "step": 11324 + }, + { + "epoch": 0.5014223570171458, + "grad_norm": 1.2854558229446411, + "learning_rate": 1.1688773145877883e-05, + "loss": 0.6284, + "step": 11325 + }, + { + "epoch": 0.5014666327219597, + "grad_norm": 1.2866581678390503, + "learning_rate": 1.1687249871973703e-05, + "loss": 0.6544, + "step": 11326 + }, + { + "epoch": 0.5015109084267735, + "grad_norm": 1.2665165662765503, + "learning_rate": 1.1685726557770935e-05, + "loss": 0.5624, + "step": 11327 + }, + { + "epoch": 0.5015551841315874, + "grad_norm": 1.277992844581604, + "learning_rate": 1.1684203203305962e-05, + "loss": 0.6243, + "step": 11328 + }, + { + "epoch": 0.5015994598364013, + "grad_norm": 1.8367902040481567, + "learning_rate": 1.1682679808615167e-05, + "loss": 0.6283, + "step": 11329 + }, + { + "epoch": 0.5016437355412151, + "grad_norm": 1.5666745901107788, + "learning_rate": 1.1681156373734935e-05, + "loss": 0.645, + "step": 11330 + }, + { + "epoch": 0.501688011246029, + "grad_norm": 1.4726313352584839, + "learning_rate": 1.1679632898701649e-05, + "loss": 0.6889, + "step": 11331 + }, + { + "epoch": 0.5017322869508429, + "grad_norm": 1.5826473236083984, + "learning_rate": 1.1678109383551703e-05, + "loss": 0.7615, + "step": 11332 + }, + { + "epoch": 0.5017765626556567, + "grad_norm": 1.529409646987915, + "learning_rate": 1.1676585828321476e-05, + "loss": 0.6994, + "step": 11333 + }, + { + "epoch": 0.5018208383604706, + "grad_norm": 1.654344916343689, + "learning_rate": 1.1675062233047365e-05, + "loss": 0.7645, + "step": 11334 + }, + { + "epoch": 0.5018651140652846, + "grad_norm": 1.2378591299057007, + "learning_rate": 1.1673538597765757e-05, + "loss": 0.6085, + "step": 11335 + }, + { + "epoch": 0.5019093897700984, + "grad_norm": 1.6093817949295044, + "learning_rate": 1.167201492251304e-05, + "loss": 0.6956, + "step": 11336 + }, + { + "epoch": 0.5019536654749123, + "grad_norm": 1.3996819257736206, + "learning_rate": 1.167049120732561e-05, + "loss": 0.7419, + "step": 11337 + }, + { + "epoch": 0.5019979411797262, + "grad_norm": 1.1352440118789673, + "learning_rate": 1.1668967452239855e-05, + "loss": 0.5934, + "step": 11338 + }, + { + "epoch": 0.50204221688454, + "grad_norm": 1.7350614070892334, + "learning_rate": 1.1667443657292172e-05, + "loss": 0.7927, + "step": 11339 + }, + { + "epoch": 0.5020864925893539, + "grad_norm": 1.6021201610565186, + "learning_rate": 1.1665919822518958e-05, + "loss": 0.7122, + "step": 11340 + }, + { + "epoch": 0.5021307682941678, + "grad_norm": 1.302453637123108, + "learning_rate": 1.1664395947956602e-05, + "loss": 0.648, + "step": 11341 + }, + { + "epoch": 0.5021750439989816, + "grad_norm": 1.3401386737823486, + "learning_rate": 1.1662872033641506e-05, + "loss": 0.6545, + "step": 11342 + }, + { + "epoch": 0.5022193197037955, + "grad_norm": 1.5152758359909058, + "learning_rate": 1.1661348079610065e-05, + "loss": 0.6041, + "step": 11343 + }, + { + "epoch": 0.5022635954086094, + "grad_norm": 1.6925569772720337, + "learning_rate": 1.1659824085898677e-05, + "loss": 0.7512, + "step": 11344 + }, + { + "epoch": 0.5023078711134232, + "grad_norm": 1.866481065750122, + "learning_rate": 1.1658300052543742e-05, + "loss": 0.7266, + "step": 11345 + }, + { + "epoch": 0.5023521468182371, + "grad_norm": 1.2466758489608765, + "learning_rate": 1.165677597958166e-05, + "loss": 0.5622, + "step": 11346 + }, + { + "epoch": 0.5023964225230511, + "grad_norm": 1.4878103733062744, + "learning_rate": 1.1655251867048837e-05, + "loss": 0.6594, + "step": 11347 + }, + { + "epoch": 0.502440698227865, + "grad_norm": 1.7616771459579468, + "learning_rate": 1.1653727714981668e-05, + "loss": 0.764, + "step": 11348 + }, + { + "epoch": 0.5024849739326788, + "grad_norm": 1.4681845903396606, + "learning_rate": 1.1652203523416556e-05, + "loss": 0.6196, + "step": 11349 + }, + { + "epoch": 0.5025292496374927, + "grad_norm": 1.2831004858016968, + "learning_rate": 1.1650679292389913e-05, + "loss": 0.5991, + "step": 11350 + }, + { + "epoch": 0.5025735253423066, + "grad_norm": 1.85275137424469, + "learning_rate": 1.1649155021938137e-05, + "loss": 0.7501, + "step": 11351 + }, + { + "epoch": 0.5026178010471204, + "grad_norm": 1.4467164278030396, + "learning_rate": 1.1647630712097638e-05, + "loss": 0.6376, + "step": 11352 + }, + { + "epoch": 0.5026620767519343, + "grad_norm": 1.2676602602005005, + "learning_rate": 1.1646106362904819e-05, + "loss": 0.6036, + "step": 11353 + }, + { + "epoch": 0.5027063524567482, + "grad_norm": 1.8851401805877686, + "learning_rate": 1.1644581974396087e-05, + "loss": 0.7299, + "step": 11354 + }, + { + "epoch": 0.502750628161562, + "grad_norm": 1.471122145652771, + "learning_rate": 1.1643057546607858e-05, + "loss": 0.6969, + "step": 11355 + }, + { + "epoch": 0.5027949038663759, + "grad_norm": 1.4795578718185425, + "learning_rate": 1.1641533079576535e-05, + "loss": 0.6521, + "step": 11356 + }, + { + "epoch": 0.5028391795711898, + "grad_norm": 1.5682929754257202, + "learning_rate": 1.1640008573338532e-05, + "loss": 0.6882, + "step": 11357 + }, + { + "epoch": 0.5028834552760036, + "grad_norm": 1.926499366760254, + "learning_rate": 1.1638484027930261e-05, + "loss": 0.7348, + "step": 11358 + }, + { + "epoch": 0.5029277309808176, + "grad_norm": 1.5236490964889526, + "learning_rate": 1.1636959443388131e-05, + "loss": 0.7119, + "step": 11359 + }, + { + "epoch": 0.5029720066856315, + "grad_norm": 1.8577215671539307, + "learning_rate": 1.163543481974856e-05, + "loss": 0.8061, + "step": 11360 + }, + { + "epoch": 0.5030162823904453, + "grad_norm": 1.3048933744430542, + "learning_rate": 1.1633910157047956e-05, + "loss": 0.6416, + "step": 11361 + }, + { + "epoch": 0.5030605580952592, + "grad_norm": 1.459000587463379, + "learning_rate": 1.1632385455322742e-05, + "loss": 0.699, + "step": 11362 + }, + { + "epoch": 0.5031048338000731, + "grad_norm": 1.408940315246582, + "learning_rate": 1.1630860714609329e-05, + "loss": 0.5751, + "step": 11363 + }, + { + "epoch": 0.5031491095048869, + "grad_norm": 1.7142940759658813, + "learning_rate": 1.1629335934944135e-05, + "loss": 0.6436, + "step": 11364 + }, + { + "epoch": 0.5031933852097008, + "grad_norm": 1.5631407499313354, + "learning_rate": 1.162781111636358e-05, + "loss": 0.7552, + "step": 11365 + }, + { + "epoch": 0.5032376609145147, + "grad_norm": 1.6298960447311401, + "learning_rate": 1.162628625890408e-05, + "loss": 0.6435, + "step": 11366 + }, + { + "epoch": 0.5032819366193285, + "grad_norm": 1.7342222929000854, + "learning_rate": 1.1624761362602061e-05, + "loss": 0.7501, + "step": 11367 + }, + { + "epoch": 0.5033262123241424, + "grad_norm": 1.4235899448394775, + "learning_rate": 1.1623236427493936e-05, + "loss": 0.6621, + "step": 11368 + }, + { + "epoch": 0.5033704880289563, + "grad_norm": 1.4384818077087402, + "learning_rate": 1.162171145361613e-05, + "loss": 0.699, + "step": 11369 + }, + { + "epoch": 0.5034147637337701, + "grad_norm": 1.7350363731384277, + "learning_rate": 1.162018644100507e-05, + "loss": 0.7347, + "step": 11370 + }, + { + "epoch": 0.503459039438584, + "grad_norm": 1.5487663745880127, + "learning_rate": 1.1618661389697171e-05, + "loss": 0.6939, + "step": 11371 + }, + { + "epoch": 0.503503315143398, + "grad_norm": 1.3693267107009888, + "learning_rate": 1.161713629972887e-05, + "loss": 0.6273, + "step": 11372 + }, + { + "epoch": 0.5035475908482118, + "grad_norm": 1.4210174083709717, + "learning_rate": 1.1615611171136581e-05, + "loss": 0.6657, + "step": 11373 + }, + { + "epoch": 0.5035918665530257, + "grad_norm": 1.460663080215454, + "learning_rate": 1.1614086003956734e-05, + "loss": 0.706, + "step": 11374 + }, + { + "epoch": 0.5036361422578396, + "grad_norm": 1.418978214263916, + "learning_rate": 1.1612560798225755e-05, + "loss": 0.7445, + "step": 11375 + }, + { + "epoch": 0.5036804179626534, + "grad_norm": 1.3981654644012451, + "learning_rate": 1.161103555398008e-05, + "loss": 0.6298, + "step": 11376 + }, + { + "epoch": 0.5037246936674673, + "grad_norm": 1.6448209285736084, + "learning_rate": 1.160951027125613e-05, + "loss": 0.6908, + "step": 11377 + }, + { + "epoch": 0.5037689693722812, + "grad_norm": 1.7703050374984741, + "learning_rate": 1.1607984950090336e-05, + "loss": 0.6937, + "step": 11378 + }, + { + "epoch": 0.503813245077095, + "grad_norm": 1.4256430864334106, + "learning_rate": 1.1606459590519132e-05, + "loss": 0.6517, + "step": 11379 + }, + { + "epoch": 0.5038575207819089, + "grad_norm": 1.3750056028366089, + "learning_rate": 1.1604934192578949e-05, + "loss": 0.6969, + "step": 11380 + }, + { + "epoch": 0.5039017964867228, + "grad_norm": 1.5878973007202148, + "learning_rate": 1.160340875630622e-05, + "loss": 0.6722, + "step": 11381 + }, + { + "epoch": 0.5039460721915368, + "grad_norm": 1.5007375478744507, + "learning_rate": 1.1601883281737378e-05, + "loss": 0.7244, + "step": 11382 + }, + { + "epoch": 0.5039903478963506, + "grad_norm": 1.392344355583191, + "learning_rate": 1.1600357768908856e-05, + "loss": 0.6568, + "step": 11383 + }, + { + "epoch": 0.5040346236011645, + "grad_norm": 1.3379155397415161, + "learning_rate": 1.1598832217857095e-05, + "loss": 0.669, + "step": 11384 + }, + { + "epoch": 0.5040788993059784, + "grad_norm": 1.5343806743621826, + "learning_rate": 1.1597306628618525e-05, + "loss": 0.632, + "step": 11385 + }, + { + "epoch": 0.5041231750107922, + "grad_norm": 1.3465629816055298, + "learning_rate": 1.1595781001229588e-05, + "loss": 0.6668, + "step": 11386 + }, + { + "epoch": 0.5041674507156061, + "grad_norm": 1.440867304801941, + "learning_rate": 1.1594255335726725e-05, + "loss": 0.6545, + "step": 11387 + }, + { + "epoch": 0.50421172642042, + "grad_norm": 1.278724193572998, + "learning_rate": 1.1592729632146366e-05, + "loss": 0.6144, + "step": 11388 + }, + { + "epoch": 0.5042560021252338, + "grad_norm": 1.2453885078430176, + "learning_rate": 1.159120389052496e-05, + "loss": 0.6447, + "step": 11389 + }, + { + "epoch": 0.5043002778300477, + "grad_norm": 1.5762085914611816, + "learning_rate": 1.158967811089894e-05, + "loss": 0.6795, + "step": 11390 + }, + { + "epoch": 0.5043445535348616, + "grad_norm": 1.3087997436523438, + "learning_rate": 1.1588152293304759e-05, + "loss": 0.6459, + "step": 11391 + }, + { + "epoch": 0.5043888292396754, + "grad_norm": 1.368186354637146, + "learning_rate": 1.1586626437778851e-05, + "loss": 0.6662, + "step": 11392 + }, + { + "epoch": 0.5044331049444893, + "grad_norm": 1.358919620513916, + "learning_rate": 1.1585100544357662e-05, + "loss": 0.66, + "step": 11393 + }, + { + "epoch": 0.5044773806493033, + "grad_norm": 1.2128503322601318, + "learning_rate": 1.1583574613077637e-05, + "loss": 0.6017, + "step": 11394 + }, + { + "epoch": 0.504521656354117, + "grad_norm": 1.741942286491394, + "learning_rate": 1.1582048643975223e-05, + "loss": 0.7039, + "step": 11395 + }, + { + "epoch": 0.504565932058931, + "grad_norm": 1.2912846803665161, + "learning_rate": 1.1580522637086863e-05, + "loss": 0.626, + "step": 11396 + }, + { + "epoch": 0.5046102077637449, + "grad_norm": 1.3961355686187744, + "learning_rate": 1.157899659244901e-05, + "loss": 0.6462, + "step": 11397 + }, + { + "epoch": 0.5046544834685587, + "grad_norm": 1.9501460790634155, + "learning_rate": 1.1577470510098105e-05, + "loss": 0.7628, + "step": 11398 + }, + { + "epoch": 0.5046987591733726, + "grad_norm": 1.3373197317123413, + "learning_rate": 1.1575944390070604e-05, + "loss": 0.6422, + "step": 11399 + }, + { + "epoch": 0.5047430348781865, + "grad_norm": 1.5770728588104248, + "learning_rate": 1.1574418232402955e-05, + "loss": 0.6485, + "step": 11400 + }, + { + "epoch": 0.5047873105830003, + "grad_norm": 1.2508645057678223, + "learning_rate": 1.1572892037131607e-05, + "loss": 0.6741, + "step": 11401 + }, + { + "epoch": 0.5048315862878142, + "grad_norm": 1.6947463750839233, + "learning_rate": 1.1571365804293013e-05, + "loss": 0.7127, + "step": 11402 + }, + { + "epoch": 0.5048758619926281, + "grad_norm": 1.2889949083328247, + "learning_rate": 1.1569839533923626e-05, + "loss": 0.6104, + "step": 11403 + }, + { + "epoch": 0.5049201376974419, + "grad_norm": 1.3548498153686523, + "learning_rate": 1.1568313226059902e-05, + "loss": 0.6712, + "step": 11404 + }, + { + "epoch": 0.5049644134022558, + "grad_norm": 1.3508310317993164, + "learning_rate": 1.1566786880738291e-05, + "loss": 0.6698, + "step": 11405 + }, + { + "epoch": 0.5050086891070698, + "grad_norm": 1.305282711982727, + "learning_rate": 1.1565260497995252e-05, + "loss": 0.7013, + "step": 11406 + }, + { + "epoch": 0.5050529648118836, + "grad_norm": 1.5529462099075317, + "learning_rate": 1.1563734077867243e-05, + "loss": 0.6315, + "step": 11407 + }, + { + "epoch": 0.5050972405166975, + "grad_norm": 1.4113668203353882, + "learning_rate": 1.1562207620390715e-05, + "loss": 0.6856, + "step": 11408 + }, + { + "epoch": 0.5051415162215114, + "grad_norm": 1.283090591430664, + "learning_rate": 1.1560681125602134e-05, + "loss": 0.6079, + "step": 11409 + }, + { + "epoch": 0.5051857919263252, + "grad_norm": 1.4261142015457153, + "learning_rate": 1.1559154593537953e-05, + "loss": 0.6522, + "step": 11410 + }, + { + "epoch": 0.5052300676311391, + "grad_norm": 1.7406154870986938, + "learning_rate": 1.155762802423463e-05, + "loss": 0.8339, + "step": 11411 + }, + { + "epoch": 0.505274343335953, + "grad_norm": 1.570157527923584, + "learning_rate": 1.1556101417728635e-05, + "loss": 0.7197, + "step": 11412 + }, + { + "epoch": 0.5053186190407668, + "grad_norm": 1.5543079376220703, + "learning_rate": 1.1554574774056424e-05, + "loss": 0.7324, + "step": 11413 + }, + { + "epoch": 0.5053628947455807, + "grad_norm": 1.1885179281234741, + "learning_rate": 1.155304809325446e-05, + "loss": 0.5887, + "step": 11414 + }, + { + "epoch": 0.5054071704503946, + "grad_norm": 1.2340586185455322, + "learning_rate": 1.1551521375359207e-05, + "loss": 0.6253, + "step": 11415 + }, + { + "epoch": 0.5054514461552084, + "grad_norm": 1.3382195234298706, + "learning_rate": 1.1549994620407127e-05, + "loss": 0.6357, + "step": 11416 + }, + { + "epoch": 0.5054957218600223, + "grad_norm": 1.5310797691345215, + "learning_rate": 1.154846782843469e-05, + "loss": 0.7553, + "step": 11417 + }, + { + "epoch": 0.5055399975648363, + "grad_norm": 1.5680863857269287, + "learning_rate": 1.1546940999478358e-05, + "loss": 0.6992, + "step": 11418 + }, + { + "epoch": 0.5055842732696502, + "grad_norm": 1.2423760890960693, + "learning_rate": 1.1545414133574603e-05, + "loss": 0.6136, + "step": 11419 + }, + { + "epoch": 0.505628548974464, + "grad_norm": 1.4528824090957642, + "learning_rate": 1.1543887230759887e-05, + "loss": 0.6842, + "step": 11420 + }, + { + "epoch": 0.5056728246792779, + "grad_norm": 1.496889591217041, + "learning_rate": 1.1542360291070682e-05, + "loss": 0.6854, + "step": 11421 + }, + { + "epoch": 0.5057171003840918, + "grad_norm": 1.3294576406478882, + "learning_rate": 1.1540833314543459e-05, + "loss": 0.6247, + "step": 11422 + }, + { + "epoch": 0.5057613760889056, + "grad_norm": 1.613968014717102, + "learning_rate": 1.1539306301214685e-05, + "loss": 0.6794, + "step": 11423 + }, + { + "epoch": 0.5058056517937195, + "grad_norm": 1.347561240196228, + "learning_rate": 1.1537779251120836e-05, + "loss": 0.6448, + "step": 11424 + }, + { + "epoch": 0.5058499274985334, + "grad_norm": 1.6348118782043457, + "learning_rate": 1.1536252164298376e-05, + "loss": 0.6317, + "step": 11425 + }, + { + "epoch": 0.5058942032033472, + "grad_norm": 1.3638339042663574, + "learning_rate": 1.1534725040783788e-05, + "loss": 0.6089, + "step": 11426 + }, + { + "epoch": 0.5059384789081611, + "grad_norm": 1.2156867980957031, + "learning_rate": 1.153319788061354e-05, + "loss": 0.6159, + "step": 11427 + }, + { + "epoch": 0.505982754612975, + "grad_norm": 1.344395399093628, + "learning_rate": 1.1531670683824113e-05, + "loss": 0.6312, + "step": 11428 + }, + { + "epoch": 0.5060270303177888, + "grad_norm": 2.0847184658050537, + "learning_rate": 1.1530143450451977e-05, + "loss": 0.679, + "step": 11429 + }, + { + "epoch": 0.5060713060226028, + "grad_norm": 1.4828130006790161, + "learning_rate": 1.1528616180533608e-05, + "loss": 0.6267, + "step": 11430 + }, + { + "epoch": 0.5061155817274167, + "grad_norm": 1.5313547849655151, + "learning_rate": 1.1527088874105485e-05, + "loss": 0.6505, + "step": 11431 + }, + { + "epoch": 0.5061598574322305, + "grad_norm": 1.5901780128479004, + "learning_rate": 1.1525561531204092e-05, + "loss": 0.6579, + "step": 11432 + }, + { + "epoch": 0.5062041331370444, + "grad_norm": 1.6439781188964844, + "learning_rate": 1.1524034151865904e-05, + "loss": 0.7488, + "step": 11433 + }, + { + "epoch": 0.5062484088418583, + "grad_norm": 1.3818069696426392, + "learning_rate": 1.15225067361274e-05, + "loss": 0.6291, + "step": 11434 + }, + { + "epoch": 0.5062926845466721, + "grad_norm": 1.4210407733917236, + "learning_rate": 1.152097928402506e-05, + "loss": 0.6895, + "step": 11435 + }, + { + "epoch": 0.506336960251486, + "grad_norm": 1.7110354900360107, + "learning_rate": 1.1519451795595369e-05, + "loss": 0.6629, + "step": 11436 + }, + { + "epoch": 0.5063812359562999, + "grad_norm": 1.5056060552597046, + "learning_rate": 1.1517924270874812e-05, + "loss": 0.7525, + "step": 11437 + }, + { + "epoch": 0.5064255116611137, + "grad_norm": 1.6305463314056396, + "learning_rate": 1.151639670989987e-05, + "loss": 0.7026, + "step": 11438 + }, + { + "epoch": 0.5064697873659276, + "grad_norm": 1.4242311716079712, + "learning_rate": 1.1514869112707026e-05, + "loss": 0.6179, + "step": 11439 + }, + { + "epoch": 0.5065140630707415, + "grad_norm": 1.3967299461364746, + "learning_rate": 1.1513341479332763e-05, + "loss": 0.6651, + "step": 11440 + }, + { + "epoch": 0.5065583387755553, + "grad_norm": 1.177188515663147, + "learning_rate": 1.1511813809813572e-05, + "loss": 0.5523, + "step": 11441 + }, + { + "epoch": 0.5066026144803693, + "grad_norm": 1.341940999031067, + "learning_rate": 1.1510286104185944e-05, + "loss": 0.698, + "step": 11442 + }, + { + "epoch": 0.5066468901851832, + "grad_norm": 1.586513876914978, + "learning_rate": 1.1508758362486358e-05, + "loss": 0.7677, + "step": 11443 + }, + { + "epoch": 0.506691165889997, + "grad_norm": 1.2515536546707153, + "learning_rate": 1.150723058475131e-05, + "loss": 0.633, + "step": 11444 + }, + { + "epoch": 0.5067354415948109, + "grad_norm": 1.1261287927627563, + "learning_rate": 1.1505702771017284e-05, + "loss": 0.5862, + "step": 11445 + }, + { + "epoch": 0.5067797172996248, + "grad_norm": 1.3936827182769775, + "learning_rate": 1.1504174921320772e-05, + "loss": 0.6614, + "step": 11446 + }, + { + "epoch": 0.5068239930044386, + "grad_norm": 1.2763690948486328, + "learning_rate": 1.1502647035698274e-05, + "loss": 0.6344, + "step": 11447 + }, + { + "epoch": 0.5068682687092525, + "grad_norm": 1.5025291442871094, + "learning_rate": 1.1501119114186268e-05, + "loss": 0.717, + "step": 11448 + }, + { + "epoch": 0.5069125444140664, + "grad_norm": 1.2315019369125366, + "learning_rate": 1.1499591156821257e-05, + "loss": 0.641, + "step": 11449 + }, + { + "epoch": 0.5069568201188802, + "grad_norm": 1.7367112636566162, + "learning_rate": 1.149806316363973e-05, + "loss": 0.6293, + "step": 11450 + }, + { + "epoch": 0.5070010958236941, + "grad_norm": 1.5070286989212036, + "learning_rate": 1.1496535134678186e-05, + "loss": 0.7324, + "step": 11451 + }, + { + "epoch": 0.507045371528508, + "grad_norm": 1.7126703262329102, + "learning_rate": 1.1495007069973119e-05, + "loss": 0.7354, + "step": 11452 + }, + { + "epoch": 0.507089647233322, + "grad_norm": 1.383655071258545, + "learning_rate": 1.1493478969561023e-05, + "loss": 0.6823, + "step": 11453 + }, + { + "epoch": 0.5071339229381358, + "grad_norm": 1.2815971374511719, + "learning_rate": 1.14919508334784e-05, + "loss": 0.6276, + "step": 11454 + }, + { + "epoch": 0.5071781986429497, + "grad_norm": 1.481336236000061, + "learning_rate": 1.1490422661761744e-05, + "loss": 0.6156, + "step": 11455 + }, + { + "epoch": 0.5072224743477636, + "grad_norm": 1.5848608016967773, + "learning_rate": 1.1488894454447558e-05, + "loss": 0.649, + "step": 11456 + }, + { + "epoch": 0.5072667500525774, + "grad_norm": 1.4253119230270386, + "learning_rate": 1.1487366211572343e-05, + "loss": 0.6194, + "step": 11457 + }, + { + "epoch": 0.5073110257573913, + "grad_norm": 1.8770205974578857, + "learning_rate": 1.1485837933172594e-05, + "loss": 0.72, + "step": 11458 + }, + { + "epoch": 0.5073553014622052, + "grad_norm": 2.638293981552124, + "learning_rate": 1.1484309619284816e-05, + "loss": 0.7204, + "step": 11459 + }, + { + "epoch": 0.507399577167019, + "grad_norm": 2.1410434246063232, + "learning_rate": 1.1482781269945512e-05, + "loss": 0.7098, + "step": 11460 + }, + { + "epoch": 0.5074438528718329, + "grad_norm": 1.9282855987548828, + "learning_rate": 1.1481252885191184e-05, + "loss": 0.7516, + "step": 11461 + }, + { + "epoch": 0.5074881285766468, + "grad_norm": 1.718105673789978, + "learning_rate": 1.1479724465058337e-05, + "loss": 0.6665, + "step": 11462 + }, + { + "epoch": 0.5075324042814606, + "grad_norm": 2.3872947692871094, + "learning_rate": 1.1478196009583475e-05, + "loss": 0.6397, + "step": 11463 + }, + { + "epoch": 0.5075766799862745, + "grad_norm": 1.4361170530319214, + "learning_rate": 1.1476667518803106e-05, + "loss": 0.6265, + "step": 11464 + }, + { + "epoch": 0.5076209556910884, + "grad_norm": 1.5128532648086548, + "learning_rate": 1.1475138992753736e-05, + "loss": 0.6843, + "step": 11465 + }, + { + "epoch": 0.5076652313959023, + "grad_norm": 1.3248310089111328, + "learning_rate": 1.1473610431471874e-05, + "loss": 0.6018, + "step": 11466 + }, + { + "epoch": 0.5077095071007162, + "grad_norm": 1.5394225120544434, + "learning_rate": 1.1472081834994025e-05, + "loss": 0.7338, + "step": 11467 + }, + { + "epoch": 0.5077537828055301, + "grad_norm": 1.269643783569336, + "learning_rate": 1.1470553203356696e-05, + "loss": 0.6126, + "step": 11468 + }, + { + "epoch": 0.5077980585103439, + "grad_norm": 1.3986071348190308, + "learning_rate": 1.1469024536596408e-05, + "loss": 0.7248, + "step": 11469 + }, + { + "epoch": 0.5078423342151578, + "grad_norm": 1.4013954401016235, + "learning_rate": 1.146749583474966e-05, + "loss": 0.7504, + "step": 11470 + }, + { + "epoch": 0.5078866099199717, + "grad_norm": 1.251082181930542, + "learning_rate": 1.1465967097852971e-05, + "loss": 0.576, + "step": 11471 + }, + { + "epoch": 0.5079308856247855, + "grad_norm": 1.2423666715621948, + "learning_rate": 1.1464438325942853e-05, + "loss": 0.6096, + "step": 11472 + }, + { + "epoch": 0.5079751613295994, + "grad_norm": 1.3763850927352905, + "learning_rate": 1.1462909519055813e-05, + "loss": 0.6708, + "step": 11473 + }, + { + "epoch": 0.5080194370344133, + "grad_norm": 1.5981711149215698, + "learning_rate": 1.1461380677228375e-05, + "loss": 0.7217, + "step": 11474 + }, + { + "epoch": 0.5080637127392271, + "grad_norm": 1.6356706619262695, + "learning_rate": 1.1459851800497048e-05, + "loss": 0.7026, + "step": 11475 + }, + { + "epoch": 0.508107988444041, + "grad_norm": 1.3185620307922363, + "learning_rate": 1.1458322888898347e-05, + "loss": 0.6465, + "step": 11476 + }, + { + "epoch": 0.508152264148855, + "grad_norm": 1.3397701978683472, + "learning_rate": 1.1456793942468796e-05, + "loss": 0.6865, + "step": 11477 + }, + { + "epoch": 0.5081965398536687, + "grad_norm": 1.5289959907531738, + "learning_rate": 1.1455264961244903e-05, + "loss": 0.665, + "step": 11478 + }, + { + "epoch": 0.5082408155584827, + "grad_norm": 1.4062567949295044, + "learning_rate": 1.1453735945263196e-05, + "loss": 0.6075, + "step": 11479 + }, + { + "epoch": 0.5082850912632966, + "grad_norm": 1.4132245779037476, + "learning_rate": 1.1452206894560188e-05, + "loss": 0.6251, + "step": 11480 + }, + { + "epoch": 0.5083293669681104, + "grad_norm": 1.3269871473312378, + "learning_rate": 1.1450677809172396e-05, + "loss": 0.6429, + "step": 11481 + }, + { + "epoch": 0.5083736426729243, + "grad_norm": 1.404740333557129, + "learning_rate": 1.1449148689136351e-05, + "loss": 0.646, + "step": 11482 + }, + { + "epoch": 0.5084179183777382, + "grad_norm": 1.3245350122451782, + "learning_rate": 1.1447619534488568e-05, + "loss": 0.6492, + "step": 11483 + }, + { + "epoch": 0.508462194082552, + "grad_norm": 1.8002358675003052, + "learning_rate": 1.1446090345265572e-05, + "loss": 0.7519, + "step": 11484 + }, + { + "epoch": 0.5085064697873659, + "grad_norm": 1.3070787191390991, + "learning_rate": 1.1444561121503885e-05, + "loss": 0.6202, + "step": 11485 + }, + { + "epoch": 0.5085507454921798, + "grad_norm": 1.3280328512191772, + "learning_rate": 1.1443031863240033e-05, + "loss": 0.6371, + "step": 11486 + }, + { + "epoch": 0.5085950211969937, + "grad_norm": 1.437011957168579, + "learning_rate": 1.1441502570510537e-05, + "loss": 0.6976, + "step": 11487 + }, + { + "epoch": 0.5086392969018075, + "grad_norm": 1.2287933826446533, + "learning_rate": 1.1439973243351927e-05, + "loss": 0.6076, + "step": 11488 + }, + { + "epoch": 0.5086835726066214, + "grad_norm": 1.4179816246032715, + "learning_rate": 1.1438443881800729e-05, + "loss": 0.7016, + "step": 11489 + }, + { + "epoch": 0.5087278483114354, + "grad_norm": 1.6549794673919678, + "learning_rate": 1.1436914485893471e-05, + "loss": 0.7348, + "step": 11490 + }, + { + "epoch": 0.5087721240162492, + "grad_norm": 1.2471126317977905, + "learning_rate": 1.1435385055666681e-05, + "loss": 0.6115, + "step": 11491 + }, + { + "epoch": 0.5088163997210631, + "grad_norm": 1.2605210542678833, + "learning_rate": 1.1433855591156883e-05, + "loss": 0.6828, + "step": 11492 + }, + { + "epoch": 0.508860675425877, + "grad_norm": 1.3996436595916748, + "learning_rate": 1.1432326092400617e-05, + "loss": 0.6487, + "step": 11493 + }, + { + "epoch": 0.5089049511306908, + "grad_norm": 1.5455925464630127, + "learning_rate": 1.1430796559434407e-05, + "loss": 0.7133, + "step": 11494 + }, + { + "epoch": 0.5089492268355047, + "grad_norm": 1.6417750120162964, + "learning_rate": 1.1429266992294786e-05, + "loss": 0.698, + "step": 11495 + }, + { + "epoch": 0.5089935025403186, + "grad_norm": 1.458786129951477, + "learning_rate": 1.1427737391018288e-05, + "loss": 0.6571, + "step": 11496 + }, + { + "epoch": 0.5090377782451324, + "grad_norm": 1.3312668800354004, + "learning_rate": 1.1426207755641441e-05, + "loss": 0.6516, + "step": 11497 + }, + { + "epoch": 0.5090820539499463, + "grad_norm": 1.469655990600586, + "learning_rate": 1.1424678086200782e-05, + "loss": 0.7128, + "step": 11498 + }, + { + "epoch": 0.5091263296547602, + "grad_norm": 1.9870213270187378, + "learning_rate": 1.1423148382732854e-05, + "loss": 0.7285, + "step": 11499 + }, + { + "epoch": 0.509170605359574, + "grad_norm": 1.391963243484497, + "learning_rate": 1.142161864527418e-05, + "loss": 0.6269, + "step": 11500 + }, + { + "epoch": 0.509214881064388, + "grad_norm": 1.5422941446304321, + "learning_rate": 1.1420088873861306e-05, + "loss": 0.6991, + "step": 11501 + }, + { + "epoch": 0.5092591567692019, + "grad_norm": 1.9655894041061401, + "learning_rate": 1.1418559068530761e-05, + "loss": 0.6916, + "step": 11502 + }, + { + "epoch": 0.5093034324740157, + "grad_norm": 1.6635282039642334, + "learning_rate": 1.1417029229319092e-05, + "loss": 0.6726, + "step": 11503 + }, + { + "epoch": 0.5093477081788296, + "grad_norm": 1.4764150381088257, + "learning_rate": 1.1415499356262832e-05, + "loss": 0.685, + "step": 11504 + }, + { + "epoch": 0.5093919838836435, + "grad_norm": 1.380761981010437, + "learning_rate": 1.141396944939852e-05, + "loss": 0.6767, + "step": 11505 + }, + { + "epoch": 0.5094362595884573, + "grad_norm": 1.5109217166900635, + "learning_rate": 1.1412439508762702e-05, + "loss": 0.6768, + "step": 11506 + }, + { + "epoch": 0.5094805352932712, + "grad_norm": 1.3198373317718506, + "learning_rate": 1.1410909534391917e-05, + "loss": 0.6358, + "step": 11507 + }, + { + "epoch": 0.5095248109980851, + "grad_norm": 1.3369660377502441, + "learning_rate": 1.1409379526322706e-05, + "loss": 0.6101, + "step": 11508 + }, + { + "epoch": 0.5095690867028989, + "grad_norm": 1.2569999694824219, + "learning_rate": 1.1407849484591612e-05, + "loss": 0.6492, + "step": 11509 + }, + { + "epoch": 0.5096133624077128, + "grad_norm": 1.5432074069976807, + "learning_rate": 1.1406319409235178e-05, + "loss": 0.7231, + "step": 11510 + }, + { + "epoch": 0.5096576381125267, + "grad_norm": 1.8393023014068604, + "learning_rate": 1.140478930028995e-05, + "loss": 0.7054, + "step": 11511 + }, + { + "epoch": 0.5097019138173405, + "grad_norm": 1.7339564561843872, + "learning_rate": 1.1403259157792474e-05, + "loss": 0.7025, + "step": 11512 + }, + { + "epoch": 0.5097461895221544, + "grad_norm": 1.2525014877319336, + "learning_rate": 1.1401728981779297e-05, + "loss": 0.5873, + "step": 11513 + }, + { + "epoch": 0.5097904652269684, + "grad_norm": 1.485167384147644, + "learning_rate": 1.1400198772286966e-05, + "loss": 0.6459, + "step": 11514 + }, + { + "epoch": 0.5098347409317822, + "grad_norm": 1.3484406471252441, + "learning_rate": 1.1398668529352021e-05, + "loss": 0.6168, + "step": 11515 + }, + { + "epoch": 0.5098790166365961, + "grad_norm": 1.6765379905700684, + "learning_rate": 1.1397138253011024e-05, + "loss": 0.6431, + "step": 11516 + }, + { + "epoch": 0.50992329234141, + "grad_norm": 1.5181565284729004, + "learning_rate": 1.1395607943300512e-05, + "loss": 0.6255, + "step": 11517 + }, + { + "epoch": 0.5099675680462238, + "grad_norm": 1.6675364971160889, + "learning_rate": 1.1394077600257046e-05, + "loss": 0.7145, + "step": 11518 + }, + { + "epoch": 0.5100118437510377, + "grad_norm": 2.3286640644073486, + "learning_rate": 1.139254722391717e-05, + "loss": 0.6887, + "step": 11519 + }, + { + "epoch": 0.5100561194558516, + "grad_norm": 3.10336971282959, + "learning_rate": 1.1391016814317433e-05, + "loss": 0.7408, + "step": 11520 + }, + { + "epoch": 0.5101003951606654, + "grad_norm": 1.5783284902572632, + "learning_rate": 1.1389486371494397e-05, + "loss": 0.7164, + "step": 11521 + }, + { + "epoch": 0.5101446708654793, + "grad_norm": 1.3558974266052246, + "learning_rate": 1.1387955895484608e-05, + "loss": 0.6396, + "step": 11522 + }, + { + "epoch": 0.5101889465702932, + "grad_norm": 1.4038251638412476, + "learning_rate": 1.1386425386324622e-05, + "loss": 0.6217, + "step": 11523 + }, + { + "epoch": 0.5102332222751071, + "grad_norm": 1.6430299282073975, + "learning_rate": 1.1384894844050997e-05, + "loss": 0.7566, + "step": 11524 + }, + { + "epoch": 0.510277497979921, + "grad_norm": 1.4493085145950317, + "learning_rate": 1.1383364268700282e-05, + "loss": 0.6794, + "step": 11525 + }, + { + "epoch": 0.5103217736847349, + "grad_norm": 1.3778111934661865, + "learning_rate": 1.1381833660309042e-05, + "loss": 0.6438, + "step": 11526 + }, + { + "epoch": 0.5103660493895488, + "grad_norm": 1.5155048370361328, + "learning_rate": 1.1380303018913832e-05, + "loss": 0.6678, + "step": 11527 + }, + { + "epoch": 0.5104103250943626, + "grad_norm": 1.3654402494430542, + "learning_rate": 1.1378772344551204e-05, + "loss": 0.6662, + "step": 11528 + }, + { + "epoch": 0.5104546007991765, + "grad_norm": 1.3159489631652832, + "learning_rate": 1.137724163725772e-05, + "loss": 0.6494, + "step": 11529 + }, + { + "epoch": 0.5104988765039904, + "grad_norm": 1.4943736791610718, + "learning_rate": 1.1375710897069944e-05, + "loss": 0.6568, + "step": 11530 + }, + { + "epoch": 0.5105431522088042, + "grad_norm": 1.3032621145248413, + "learning_rate": 1.1374180124024435e-05, + "loss": 0.6889, + "step": 11531 + }, + { + "epoch": 0.5105874279136181, + "grad_norm": 1.522539496421814, + "learning_rate": 1.137264931815775e-05, + "loss": 0.6457, + "step": 11532 + }, + { + "epoch": 0.510631703618432, + "grad_norm": 1.2084811925888062, + "learning_rate": 1.1371118479506454e-05, + "loss": 0.6321, + "step": 11533 + }, + { + "epoch": 0.5106759793232458, + "grad_norm": 1.5053398609161377, + "learning_rate": 1.136958760810711e-05, + "loss": 0.6635, + "step": 11534 + }, + { + "epoch": 0.5107202550280597, + "grad_norm": 1.189738392829895, + "learning_rate": 1.1368056703996277e-05, + "loss": 0.5775, + "step": 11535 + }, + { + "epoch": 0.5107645307328736, + "grad_norm": 1.3513509035110474, + "learning_rate": 1.136652576721053e-05, + "loss": 0.628, + "step": 11536 + }, + { + "epoch": 0.5108088064376874, + "grad_norm": 1.5168777704238892, + "learning_rate": 1.1364994797786424e-05, + "loss": 0.6266, + "step": 11537 + }, + { + "epoch": 0.5108530821425014, + "grad_norm": 1.3871300220489502, + "learning_rate": 1.136346379576053e-05, + "loss": 0.6525, + "step": 11538 + }, + { + "epoch": 0.5108973578473153, + "grad_norm": 1.2936040163040161, + "learning_rate": 1.1361932761169412e-05, + "loss": 0.6088, + "step": 11539 + }, + { + "epoch": 0.5109416335521291, + "grad_norm": 1.6713919639587402, + "learning_rate": 1.1360401694049639e-05, + "loss": 0.6448, + "step": 11540 + }, + { + "epoch": 0.510985909256943, + "grad_norm": 1.3552366495132446, + "learning_rate": 1.135887059443778e-05, + "loss": 0.6567, + "step": 11541 + }, + { + "epoch": 0.5110301849617569, + "grad_norm": 1.2586519718170166, + "learning_rate": 1.1357339462370404e-05, + "loss": 0.6142, + "step": 11542 + }, + { + "epoch": 0.5110744606665707, + "grad_norm": 1.7445894479751587, + "learning_rate": 1.1355808297884078e-05, + "loss": 0.6568, + "step": 11543 + }, + { + "epoch": 0.5111187363713846, + "grad_norm": 1.7005624771118164, + "learning_rate": 1.1354277101015378e-05, + "loss": 0.7771, + "step": 11544 + }, + { + "epoch": 0.5111630120761985, + "grad_norm": 1.4568887948989868, + "learning_rate": 1.135274587180087e-05, + "loss": 0.6236, + "step": 11545 + }, + { + "epoch": 0.5112072877810123, + "grad_norm": 1.4631476402282715, + "learning_rate": 1.135121461027713e-05, + "loss": 0.6368, + "step": 11546 + }, + { + "epoch": 0.5112515634858262, + "grad_norm": 1.386405110359192, + "learning_rate": 1.134968331648073e-05, + "loss": 0.6175, + "step": 11547 + }, + { + "epoch": 0.5112958391906401, + "grad_norm": 1.491098165512085, + "learning_rate": 1.1348151990448238e-05, + "loss": 0.6979, + "step": 11548 + }, + { + "epoch": 0.511340114895454, + "grad_norm": 1.3466613292694092, + "learning_rate": 1.1346620632216236e-05, + "loss": 0.588, + "step": 11549 + }, + { + "epoch": 0.5113843906002679, + "grad_norm": 1.5600416660308838, + "learning_rate": 1.1345089241821299e-05, + "loss": 0.7118, + "step": 11550 + }, + { + "epoch": 0.5114286663050818, + "grad_norm": 1.3515576124191284, + "learning_rate": 1.13435578193e-05, + "loss": 0.6444, + "step": 11551 + }, + { + "epoch": 0.5114729420098956, + "grad_norm": 1.9194862842559814, + "learning_rate": 1.1342026364688911e-05, + "loss": 0.6697, + "step": 11552 + }, + { + "epoch": 0.5115172177147095, + "grad_norm": 1.6931233406066895, + "learning_rate": 1.1340494878024619e-05, + "loss": 0.7014, + "step": 11553 + }, + { + "epoch": 0.5115614934195234, + "grad_norm": 1.4275392293930054, + "learning_rate": 1.13389633593437e-05, + "loss": 0.6069, + "step": 11554 + }, + { + "epoch": 0.5116057691243372, + "grad_norm": 1.563923954963684, + "learning_rate": 1.133743180868273e-05, + "loss": 0.6589, + "step": 11555 + }, + { + "epoch": 0.5116500448291511, + "grad_norm": 1.677785873413086, + "learning_rate": 1.1335900226078292e-05, + "loss": 0.6913, + "step": 11556 + }, + { + "epoch": 0.511694320533965, + "grad_norm": 1.55360746383667, + "learning_rate": 1.133436861156696e-05, + "loss": 0.6572, + "step": 11557 + }, + { + "epoch": 0.5117385962387789, + "grad_norm": 1.5330334901809692, + "learning_rate": 1.133283696518532e-05, + "loss": 0.7111, + "step": 11558 + }, + { + "epoch": 0.5117828719435927, + "grad_norm": 1.5717159509658813, + "learning_rate": 1.133130528696996e-05, + "loss": 0.6278, + "step": 11559 + }, + { + "epoch": 0.5118271476484066, + "grad_norm": 1.3696086406707764, + "learning_rate": 1.1329773576957455e-05, + "loss": 0.6038, + "step": 11560 + }, + { + "epoch": 0.5118714233532206, + "grad_norm": 1.3915032148361206, + "learning_rate": 1.1328241835184393e-05, + "loss": 0.6827, + "step": 11561 + }, + { + "epoch": 0.5119156990580344, + "grad_norm": 1.4680129289627075, + "learning_rate": 1.1326710061687351e-05, + "loss": 0.638, + "step": 11562 + }, + { + "epoch": 0.5119599747628483, + "grad_norm": 1.4001085758209229, + "learning_rate": 1.132517825650292e-05, + "loss": 0.648, + "step": 11563 + }, + { + "epoch": 0.5120042504676622, + "grad_norm": 1.4953320026397705, + "learning_rate": 1.1323646419667688e-05, + "loss": 0.7184, + "step": 11564 + }, + { + "epoch": 0.512048526172476, + "grad_norm": 1.4274731874465942, + "learning_rate": 1.132211455121824e-05, + "loss": 0.6876, + "step": 11565 + }, + { + "epoch": 0.5120928018772899, + "grad_norm": 1.308622121810913, + "learning_rate": 1.1320582651191163e-05, + "loss": 0.6555, + "step": 11566 + }, + { + "epoch": 0.5121370775821038, + "grad_norm": 1.6483042240142822, + "learning_rate": 1.1319050719623039e-05, + "loss": 0.6944, + "step": 11567 + }, + { + "epoch": 0.5121813532869176, + "grad_norm": 1.4354846477508545, + "learning_rate": 1.1317518756550466e-05, + "loss": 0.6604, + "step": 11568 + }, + { + "epoch": 0.5122256289917315, + "grad_norm": 1.3685623407363892, + "learning_rate": 1.1315986762010031e-05, + "loss": 0.6244, + "step": 11569 + }, + { + "epoch": 0.5122699046965454, + "grad_norm": 1.2624232769012451, + "learning_rate": 1.1314454736038325e-05, + "loss": 0.6017, + "step": 11570 + }, + { + "epoch": 0.5123141804013592, + "grad_norm": 1.4906928539276123, + "learning_rate": 1.1312922678671935e-05, + "loss": 0.6371, + "step": 11571 + }, + { + "epoch": 0.5123584561061731, + "grad_norm": 1.5471922159194946, + "learning_rate": 1.1311390589947457e-05, + "loss": 0.6973, + "step": 11572 + }, + { + "epoch": 0.512402731810987, + "grad_norm": 1.3508433103561401, + "learning_rate": 1.130985846990148e-05, + "loss": 0.7265, + "step": 11573 + }, + { + "epoch": 0.5124470075158009, + "grad_norm": 1.7567849159240723, + "learning_rate": 1.1308326318570605e-05, + "loss": 0.7407, + "step": 11574 + }, + { + "epoch": 0.5124912832206148, + "grad_norm": 1.552870512008667, + "learning_rate": 1.1306794135991418e-05, + "loss": 0.7089, + "step": 11575 + }, + { + "epoch": 0.5125355589254287, + "grad_norm": 1.642683506011963, + "learning_rate": 1.130526192220052e-05, + "loss": 0.6715, + "step": 11576 + }, + { + "epoch": 0.5125798346302425, + "grad_norm": 1.8284807205200195, + "learning_rate": 1.13037296772345e-05, + "loss": 0.6899, + "step": 11577 + }, + { + "epoch": 0.5126241103350564, + "grad_norm": 1.5538830757141113, + "learning_rate": 1.1302197401129957e-05, + "loss": 0.683, + "step": 11578 + }, + { + "epoch": 0.5126683860398703, + "grad_norm": 1.62272310256958, + "learning_rate": 1.1300665093923495e-05, + "loss": 0.7694, + "step": 11579 + }, + { + "epoch": 0.5127126617446841, + "grad_norm": 2.0821378231048584, + "learning_rate": 1.12991327556517e-05, + "loss": 0.703, + "step": 11580 + }, + { + "epoch": 0.512756937449498, + "grad_norm": 1.4279050827026367, + "learning_rate": 1.1297600386351182e-05, + "loss": 0.6303, + "step": 11581 + }, + { + "epoch": 0.5128012131543119, + "grad_norm": 1.3237731456756592, + "learning_rate": 1.1296067986058532e-05, + "loss": 0.6585, + "step": 11582 + }, + { + "epoch": 0.5128454888591257, + "grad_norm": 1.6392399072647095, + "learning_rate": 1.1294535554810356e-05, + "loss": 0.7045, + "step": 11583 + }, + { + "epoch": 0.5128897645639396, + "grad_norm": 1.410245418548584, + "learning_rate": 1.1293003092643251e-05, + "loss": 0.6748, + "step": 11584 + }, + { + "epoch": 0.5129340402687536, + "grad_norm": 1.312209129333496, + "learning_rate": 1.1291470599593818e-05, + "loss": 0.6515, + "step": 11585 + }, + { + "epoch": 0.5129783159735674, + "grad_norm": 1.5737876892089844, + "learning_rate": 1.1289938075698666e-05, + "loss": 0.7102, + "step": 11586 + }, + { + "epoch": 0.5130225916783813, + "grad_norm": 1.6714764833450317, + "learning_rate": 1.128840552099439e-05, + "loss": 0.7039, + "step": 11587 + }, + { + "epoch": 0.5130668673831952, + "grad_norm": 1.751283049583435, + "learning_rate": 1.1286872935517599e-05, + "loss": 0.7273, + "step": 11588 + }, + { + "epoch": 0.513111143088009, + "grad_norm": 1.4850677251815796, + "learning_rate": 1.1285340319304893e-05, + "loss": 0.6552, + "step": 11589 + }, + { + "epoch": 0.5131554187928229, + "grad_norm": 1.3378106355667114, + "learning_rate": 1.1283807672392881e-05, + "loss": 0.5826, + "step": 11590 + }, + { + "epoch": 0.5131996944976368, + "grad_norm": 1.4349443912506104, + "learning_rate": 1.128227499481817e-05, + "loss": 0.6988, + "step": 11591 + }, + { + "epoch": 0.5132439702024506, + "grad_norm": 1.56508469581604, + "learning_rate": 1.1280742286617364e-05, + "loss": 0.7065, + "step": 11592 + }, + { + "epoch": 0.5132882459072645, + "grad_norm": 1.47749924659729, + "learning_rate": 1.127920954782707e-05, + "loss": 0.6944, + "step": 11593 + }, + { + "epoch": 0.5133325216120784, + "grad_norm": 1.3980011940002441, + "learning_rate": 1.1277676778483899e-05, + "loss": 0.6663, + "step": 11594 + }, + { + "epoch": 0.5133767973168923, + "grad_norm": 1.5011610984802246, + "learning_rate": 1.1276143978624457e-05, + "loss": 0.6625, + "step": 11595 + }, + { + "epoch": 0.5134210730217061, + "grad_norm": 1.62265145778656, + "learning_rate": 1.1274611148285358e-05, + "loss": 0.6731, + "step": 11596 + }, + { + "epoch": 0.51346534872652, + "grad_norm": 2.054542064666748, + "learning_rate": 1.1273078287503205e-05, + "loss": 0.7764, + "step": 11597 + }, + { + "epoch": 0.513509624431334, + "grad_norm": 1.4640841484069824, + "learning_rate": 1.127154539631462e-05, + "loss": 0.7107, + "step": 11598 + }, + { + "epoch": 0.5135539001361478, + "grad_norm": 1.9743250608444214, + "learning_rate": 1.1270012474756204e-05, + "loss": 0.7429, + "step": 11599 + }, + { + "epoch": 0.5135981758409617, + "grad_norm": 1.4806995391845703, + "learning_rate": 1.1268479522864573e-05, + "loss": 0.6421, + "step": 11600 + }, + { + "epoch": 0.5136424515457756, + "grad_norm": 1.5264524221420288, + "learning_rate": 1.1266946540676344e-05, + "loss": 0.6639, + "step": 11601 + }, + { + "epoch": 0.5136867272505894, + "grad_norm": 1.4579051733016968, + "learning_rate": 1.1265413528228128e-05, + "loss": 0.6504, + "step": 11602 + }, + { + "epoch": 0.5137310029554033, + "grad_norm": 1.405828595161438, + "learning_rate": 1.1263880485556542e-05, + "loss": 0.6179, + "step": 11603 + }, + { + "epoch": 0.5137752786602172, + "grad_norm": 1.3887701034545898, + "learning_rate": 1.1262347412698195e-05, + "loss": 0.6185, + "step": 11604 + }, + { + "epoch": 0.513819554365031, + "grad_norm": 1.216700553894043, + "learning_rate": 1.1260814309689712e-05, + "loss": 0.6133, + "step": 11605 + }, + { + "epoch": 0.5138638300698449, + "grad_norm": 1.405989646911621, + "learning_rate": 1.1259281176567705e-05, + "loss": 0.7182, + "step": 11606 + }, + { + "epoch": 0.5139081057746588, + "grad_norm": 1.5158694982528687, + "learning_rate": 1.1257748013368792e-05, + "loss": 0.6702, + "step": 11607 + }, + { + "epoch": 0.5139523814794726, + "grad_norm": 1.568871259689331, + "learning_rate": 1.1256214820129593e-05, + "loss": 0.6933, + "step": 11608 + }, + { + "epoch": 0.5139966571842866, + "grad_norm": 1.6674076318740845, + "learning_rate": 1.1254681596886722e-05, + "loss": 0.7499, + "step": 11609 + }, + { + "epoch": 0.5140409328891005, + "grad_norm": 1.5274021625518799, + "learning_rate": 1.1253148343676805e-05, + "loss": 0.6755, + "step": 11610 + }, + { + "epoch": 0.5140852085939143, + "grad_norm": 1.2817610502243042, + "learning_rate": 1.125161506053646e-05, + "loss": 0.6304, + "step": 11611 + }, + { + "epoch": 0.5141294842987282, + "grad_norm": 2.035673141479492, + "learning_rate": 1.1250081747502311e-05, + "loss": 0.7575, + "step": 11612 + }, + { + "epoch": 0.5141737600035421, + "grad_norm": 1.3845691680908203, + "learning_rate": 1.1248548404610975e-05, + "loss": 0.643, + "step": 11613 + }, + { + "epoch": 0.5142180357083559, + "grad_norm": 1.7219301462173462, + "learning_rate": 1.1247015031899076e-05, + "loss": 0.6492, + "step": 11614 + }, + { + "epoch": 0.5142623114131698, + "grad_norm": 1.6632463932037354, + "learning_rate": 1.1245481629403238e-05, + "loss": 0.7559, + "step": 11615 + }, + { + "epoch": 0.5143065871179837, + "grad_norm": 1.492261290550232, + "learning_rate": 1.1243948197160087e-05, + "loss": 0.6036, + "step": 11616 + }, + { + "epoch": 0.5143508628227975, + "grad_norm": 1.3883832693099976, + "learning_rate": 1.1242414735206246e-05, + "loss": 0.6142, + "step": 11617 + }, + { + "epoch": 0.5143951385276114, + "grad_norm": 1.4498165845870972, + "learning_rate": 1.1240881243578343e-05, + "loss": 0.7068, + "step": 11618 + }, + { + "epoch": 0.5144394142324253, + "grad_norm": 1.7952404022216797, + "learning_rate": 1.1239347722312997e-05, + "loss": 0.7423, + "step": 11619 + }, + { + "epoch": 0.5144836899372391, + "grad_norm": 1.3363102674484253, + "learning_rate": 1.1237814171446842e-05, + "loss": 0.6275, + "step": 11620 + }, + { + "epoch": 0.514527965642053, + "grad_norm": 1.495914340019226, + "learning_rate": 1.1236280591016507e-05, + "loss": 0.7038, + "step": 11621 + }, + { + "epoch": 0.514572241346867, + "grad_norm": 1.6635069847106934, + "learning_rate": 1.1234746981058612e-05, + "loss": 0.7589, + "step": 11622 + }, + { + "epoch": 0.5146165170516808, + "grad_norm": 1.5332770347595215, + "learning_rate": 1.1233213341609792e-05, + "loss": 0.6373, + "step": 11623 + }, + { + "epoch": 0.5146607927564947, + "grad_norm": 1.8391327857971191, + "learning_rate": 1.1231679672706674e-05, + "loss": 0.6298, + "step": 11624 + }, + { + "epoch": 0.5147050684613086, + "grad_norm": 1.3446892499923706, + "learning_rate": 1.1230145974385892e-05, + "loss": 0.637, + "step": 11625 + }, + { + "epoch": 0.5147493441661224, + "grad_norm": 1.5413689613342285, + "learning_rate": 1.1228612246684076e-05, + "loss": 0.7041, + "step": 11626 + }, + { + "epoch": 0.5147936198709363, + "grad_norm": 1.6107324361801147, + "learning_rate": 1.1227078489637852e-05, + "loss": 0.7551, + "step": 11627 + }, + { + "epoch": 0.5148378955757502, + "grad_norm": 1.6828436851501465, + "learning_rate": 1.1225544703283863e-05, + "loss": 0.7087, + "step": 11628 + }, + { + "epoch": 0.5148821712805641, + "grad_norm": 1.5729972124099731, + "learning_rate": 1.1224010887658731e-05, + "loss": 0.6935, + "step": 11629 + }, + { + "epoch": 0.5149264469853779, + "grad_norm": 1.4926331043243408, + "learning_rate": 1.1222477042799103e-05, + "loss": 0.7105, + "step": 11630 + }, + { + "epoch": 0.5149707226901918, + "grad_norm": 1.4800541400909424, + "learning_rate": 1.12209431687416e-05, + "loss": 0.7197, + "step": 11631 + }, + { + "epoch": 0.5150149983950058, + "grad_norm": 1.7775336503982544, + "learning_rate": 1.1219409265522865e-05, + "loss": 0.7292, + "step": 11632 + }, + { + "epoch": 0.5150592740998196, + "grad_norm": 1.8829894065856934, + "learning_rate": 1.1217875333179535e-05, + "loss": 0.7379, + "step": 11633 + }, + { + "epoch": 0.5151035498046335, + "grad_norm": 1.468564510345459, + "learning_rate": 1.1216341371748241e-05, + "loss": 0.7044, + "step": 11634 + }, + { + "epoch": 0.5151478255094474, + "grad_norm": 1.5168241262435913, + "learning_rate": 1.1214807381265627e-05, + "loss": 0.657, + "step": 11635 + }, + { + "epoch": 0.5151921012142612, + "grad_norm": 1.2495695352554321, + "learning_rate": 1.1213273361768327e-05, + "loss": 0.6406, + "step": 11636 + }, + { + "epoch": 0.5152363769190751, + "grad_norm": 1.429600477218628, + "learning_rate": 1.1211739313292977e-05, + "loss": 0.6907, + "step": 11637 + }, + { + "epoch": 0.515280652623889, + "grad_norm": 1.495569109916687, + "learning_rate": 1.1210205235876223e-05, + "loss": 0.6606, + "step": 11638 + }, + { + "epoch": 0.5153249283287028, + "grad_norm": 1.6317988634109497, + "learning_rate": 1.1208671129554703e-05, + "loss": 0.7488, + "step": 11639 + }, + { + "epoch": 0.5153692040335167, + "grad_norm": 1.537854790687561, + "learning_rate": 1.1207136994365055e-05, + "loss": 0.7149, + "step": 11640 + }, + { + "epoch": 0.5154134797383306, + "grad_norm": 1.1587034463882446, + "learning_rate": 1.1205602830343925e-05, + "loss": 0.5582, + "step": 11641 + }, + { + "epoch": 0.5154577554431444, + "grad_norm": 1.5808647871017456, + "learning_rate": 1.1204068637527951e-05, + "loss": 0.7138, + "step": 11642 + }, + { + "epoch": 0.5155020311479583, + "grad_norm": 1.4305375814437866, + "learning_rate": 1.120253441595378e-05, + "loss": 0.6, + "step": 11643 + }, + { + "epoch": 0.5155463068527723, + "grad_norm": 1.3685295581817627, + "learning_rate": 1.1201000165658052e-05, + "loss": 0.652, + "step": 11644 + }, + { + "epoch": 0.515590582557586, + "grad_norm": 1.2340278625488281, + "learning_rate": 1.1199465886677414e-05, + "loss": 0.5861, + "step": 11645 + }, + { + "epoch": 0.5156348582624, + "grad_norm": 1.3938294649124146, + "learning_rate": 1.119793157904851e-05, + "loss": 0.6914, + "step": 11646 + }, + { + "epoch": 0.5156791339672139, + "grad_norm": 1.496156930923462, + "learning_rate": 1.1196397242807983e-05, + "loss": 0.6253, + "step": 11647 + }, + { + "epoch": 0.5157234096720277, + "grad_norm": 1.3396446704864502, + "learning_rate": 1.1194862877992487e-05, + "loss": 0.686, + "step": 11648 + }, + { + "epoch": 0.5157676853768416, + "grad_norm": 1.540418028831482, + "learning_rate": 1.119332848463866e-05, + "loss": 0.7088, + "step": 11649 + }, + { + "epoch": 0.5158119610816555, + "grad_norm": 1.5409024953842163, + "learning_rate": 1.1191794062783155e-05, + "loss": 0.7165, + "step": 11650 + }, + { + "epoch": 0.5158562367864693, + "grad_norm": 2.3038222789764404, + "learning_rate": 1.1190259612462622e-05, + "loss": 0.7091, + "step": 11651 + }, + { + "epoch": 0.5159005124912832, + "grad_norm": 1.6797140836715698, + "learning_rate": 1.1188725133713702e-05, + "loss": 0.7832, + "step": 11652 + }, + { + "epoch": 0.5159447881960971, + "grad_norm": 1.3026119470596313, + "learning_rate": 1.1187190626573052e-05, + "loss": 0.619, + "step": 11653 + }, + { + "epoch": 0.5159890639009109, + "grad_norm": 1.7193310260772705, + "learning_rate": 1.1185656091077324e-05, + "loss": 0.6858, + "step": 11654 + }, + { + "epoch": 0.5160333396057248, + "grad_norm": 1.4053672552108765, + "learning_rate": 1.118412152726316e-05, + "loss": 0.6563, + "step": 11655 + }, + { + "epoch": 0.5160776153105388, + "grad_norm": 1.4197138547897339, + "learning_rate": 1.1182586935167224e-05, + "loss": 0.6345, + "step": 11656 + }, + { + "epoch": 0.5161218910153526, + "grad_norm": 1.8539339303970337, + "learning_rate": 1.1181052314826155e-05, + "loss": 0.7687, + "step": 11657 + }, + { + "epoch": 0.5161661667201665, + "grad_norm": 1.6178027391433716, + "learning_rate": 1.1179517666276618e-05, + "loss": 0.6499, + "step": 11658 + }, + { + "epoch": 0.5162104424249804, + "grad_norm": 1.2446800470352173, + "learning_rate": 1.1177982989555261e-05, + "loss": 0.6209, + "step": 11659 + }, + { + "epoch": 0.5162547181297942, + "grad_norm": 1.9729024171829224, + "learning_rate": 1.1176448284698737e-05, + "loss": 0.6617, + "step": 11660 + }, + { + "epoch": 0.5162989938346081, + "grad_norm": 1.2420188188552856, + "learning_rate": 1.1174913551743707e-05, + "loss": 0.6068, + "step": 11661 + }, + { + "epoch": 0.516343269539422, + "grad_norm": 1.4535282850265503, + "learning_rate": 1.1173378790726819e-05, + "loss": 0.6605, + "step": 11662 + }, + { + "epoch": 0.5163875452442358, + "grad_norm": 1.5245150327682495, + "learning_rate": 1.1171844001684739e-05, + "loss": 0.6842, + "step": 11663 + }, + { + "epoch": 0.5164318209490497, + "grad_norm": 1.631760597229004, + "learning_rate": 1.1170309184654118e-05, + "loss": 0.7709, + "step": 11664 + }, + { + "epoch": 0.5164760966538636, + "grad_norm": 1.7663904428482056, + "learning_rate": 1.116877433967161e-05, + "loss": 0.6367, + "step": 11665 + }, + { + "epoch": 0.5165203723586775, + "grad_norm": 1.5381970405578613, + "learning_rate": 1.1167239466773884e-05, + "loss": 0.6164, + "step": 11666 + }, + { + "epoch": 0.5165646480634913, + "grad_norm": 1.4795875549316406, + "learning_rate": 1.1165704565997593e-05, + "loss": 0.6871, + "step": 11667 + }, + { + "epoch": 0.5166089237683053, + "grad_norm": 1.6389538049697876, + "learning_rate": 1.1164169637379397e-05, + "loss": 0.6748, + "step": 11668 + }, + { + "epoch": 0.5166531994731192, + "grad_norm": 1.464144229888916, + "learning_rate": 1.1162634680955956e-05, + "loss": 0.6577, + "step": 11669 + }, + { + "epoch": 0.516697475177933, + "grad_norm": 1.5031448602676392, + "learning_rate": 1.1161099696763933e-05, + "loss": 0.7293, + "step": 11670 + }, + { + "epoch": 0.5167417508827469, + "grad_norm": 1.358372688293457, + "learning_rate": 1.1159564684839992e-05, + "loss": 0.6551, + "step": 11671 + }, + { + "epoch": 0.5167860265875608, + "grad_norm": 1.3674906492233276, + "learning_rate": 1.1158029645220786e-05, + "loss": 0.6283, + "step": 11672 + }, + { + "epoch": 0.5168303022923746, + "grad_norm": 1.2778143882751465, + "learning_rate": 1.1156494577942993e-05, + "loss": 0.6666, + "step": 11673 + }, + { + "epoch": 0.5168745779971885, + "grad_norm": 1.4425837993621826, + "learning_rate": 1.1154959483043264e-05, + "loss": 0.6495, + "step": 11674 + }, + { + "epoch": 0.5169188537020024, + "grad_norm": 1.3570117950439453, + "learning_rate": 1.1153424360558268e-05, + "loss": 0.6439, + "step": 11675 + }, + { + "epoch": 0.5169631294068162, + "grad_norm": 1.582749366760254, + "learning_rate": 1.1151889210524671e-05, + "loss": 0.6054, + "step": 11676 + }, + { + "epoch": 0.5170074051116301, + "grad_norm": 1.6249362230300903, + "learning_rate": 1.115035403297914e-05, + "loss": 0.6031, + "step": 11677 + }, + { + "epoch": 0.517051680816444, + "grad_norm": 1.5584673881530762, + "learning_rate": 1.114881882795834e-05, + "loss": 0.6951, + "step": 11678 + }, + { + "epoch": 0.5170959565212578, + "grad_norm": 1.5050699710845947, + "learning_rate": 1.1147283595498932e-05, + "loss": 0.6763, + "step": 11679 + }, + { + "epoch": 0.5171402322260717, + "grad_norm": 1.3029991388320923, + "learning_rate": 1.1145748335637592e-05, + "loss": 0.6299, + "step": 11680 + }, + { + "epoch": 0.5171845079308857, + "grad_norm": 1.5150299072265625, + "learning_rate": 1.1144213048410987e-05, + "loss": 0.6675, + "step": 11681 + }, + { + "epoch": 0.5172287836356995, + "grad_norm": 1.2707011699676514, + "learning_rate": 1.1142677733855788e-05, + "loss": 0.62, + "step": 11682 + }, + { + "epoch": 0.5172730593405134, + "grad_norm": 1.462509036064148, + "learning_rate": 1.1141142392008656e-05, + "loss": 0.6536, + "step": 11683 + }, + { + "epoch": 0.5173173350453273, + "grad_norm": 1.4874868392944336, + "learning_rate": 1.113960702290627e-05, + "loss": 0.6478, + "step": 11684 + }, + { + "epoch": 0.5173616107501411, + "grad_norm": 1.2640159130096436, + "learning_rate": 1.1138071626585296e-05, + "loss": 0.6396, + "step": 11685 + }, + { + "epoch": 0.517405886454955, + "grad_norm": 1.5277149677276611, + "learning_rate": 1.1136536203082411e-05, + "loss": 0.7562, + "step": 11686 + }, + { + "epoch": 0.5174501621597689, + "grad_norm": 1.4293116331100464, + "learning_rate": 1.1135000752434284e-05, + "loss": 0.6489, + "step": 11687 + }, + { + "epoch": 0.5174944378645827, + "grad_norm": 1.3121750354766846, + "learning_rate": 1.1133465274677588e-05, + "loss": 0.5867, + "step": 11688 + }, + { + "epoch": 0.5175387135693966, + "grad_norm": 1.6740578413009644, + "learning_rate": 1.1131929769848995e-05, + "loss": 0.7581, + "step": 11689 + }, + { + "epoch": 0.5175829892742105, + "grad_norm": 1.7754234075546265, + "learning_rate": 1.113039423798518e-05, + "loss": 0.6797, + "step": 11690 + }, + { + "epoch": 0.5176272649790243, + "grad_norm": 1.448740005493164, + "learning_rate": 1.1128858679122822e-05, + "loss": 0.6354, + "step": 11691 + }, + { + "epoch": 0.5176715406838382, + "grad_norm": 1.6703087091445923, + "learning_rate": 1.1127323093298594e-05, + "loss": 0.7082, + "step": 11692 + }, + { + "epoch": 0.5177158163886522, + "grad_norm": 1.4765616655349731, + "learning_rate": 1.1125787480549171e-05, + "loss": 0.6502, + "step": 11693 + }, + { + "epoch": 0.517760092093466, + "grad_norm": 1.3713829517364502, + "learning_rate": 1.1124251840911229e-05, + "loss": 0.6554, + "step": 11694 + }, + { + "epoch": 0.5178043677982799, + "grad_norm": 1.4227129220962524, + "learning_rate": 1.1122716174421446e-05, + "loss": 0.6277, + "step": 11695 + }, + { + "epoch": 0.5178486435030938, + "grad_norm": 1.441056728363037, + "learning_rate": 1.1121180481116504e-05, + "loss": 0.6996, + "step": 11696 + }, + { + "epoch": 0.5178929192079076, + "grad_norm": 1.6890250444412231, + "learning_rate": 1.1119644761033079e-05, + "loss": 0.7557, + "step": 11697 + }, + { + "epoch": 0.5179371949127215, + "grad_norm": 1.5999670028686523, + "learning_rate": 1.1118109014207852e-05, + "loss": 0.6957, + "step": 11698 + }, + { + "epoch": 0.5179814706175354, + "grad_norm": 1.3906099796295166, + "learning_rate": 1.11165732406775e-05, + "loss": 0.6305, + "step": 11699 + }, + { + "epoch": 0.5180257463223493, + "grad_norm": 1.426774263381958, + "learning_rate": 1.1115037440478702e-05, + "loss": 0.6276, + "step": 11700 + }, + { + "epoch": 0.5180700220271631, + "grad_norm": 1.6134060621261597, + "learning_rate": 1.111350161364815e-05, + "loss": 0.6736, + "step": 11701 + }, + { + "epoch": 0.518114297731977, + "grad_norm": 1.4901777505874634, + "learning_rate": 1.1111965760222511e-05, + "loss": 0.624, + "step": 11702 + }, + { + "epoch": 0.518158573436791, + "grad_norm": 1.432221531867981, + "learning_rate": 1.111042988023848e-05, + "loss": 0.6921, + "step": 11703 + }, + { + "epoch": 0.5182028491416047, + "grad_norm": 1.5144537687301636, + "learning_rate": 1.1108893973732732e-05, + "loss": 0.6144, + "step": 11704 + }, + { + "epoch": 0.5182471248464187, + "grad_norm": 1.3859821557998657, + "learning_rate": 1.1107358040741958e-05, + "loss": 0.6344, + "step": 11705 + }, + { + "epoch": 0.5182914005512326, + "grad_norm": 1.3099645376205444, + "learning_rate": 1.110582208130284e-05, + "loss": 0.643, + "step": 11706 + }, + { + "epoch": 0.5183356762560464, + "grad_norm": 1.8142086267471313, + "learning_rate": 1.1104286095452058e-05, + "loss": 0.7755, + "step": 11707 + }, + { + "epoch": 0.5183799519608603, + "grad_norm": 1.9044147729873657, + "learning_rate": 1.1102750083226301e-05, + "loss": 0.6921, + "step": 11708 + }, + { + "epoch": 0.5184242276656742, + "grad_norm": 2.8645544052124023, + "learning_rate": 1.1101214044662258e-05, + "loss": 0.6492, + "step": 11709 + }, + { + "epoch": 0.518468503370488, + "grad_norm": 1.3292542695999146, + "learning_rate": 1.1099677979796615e-05, + "loss": 0.5746, + "step": 11710 + }, + { + "epoch": 0.5185127790753019, + "grad_norm": 1.43280029296875, + "learning_rate": 1.1098141888666058e-05, + "loss": 0.6679, + "step": 11711 + }, + { + "epoch": 0.5185570547801158, + "grad_norm": 1.4012224674224854, + "learning_rate": 1.1096605771307277e-05, + "loss": 0.6597, + "step": 11712 + }, + { + "epoch": 0.5186013304849296, + "grad_norm": 1.7400810718536377, + "learning_rate": 1.1095069627756958e-05, + "loss": 0.7113, + "step": 11713 + }, + { + "epoch": 0.5186456061897435, + "grad_norm": 1.4096384048461914, + "learning_rate": 1.1093533458051792e-05, + "loss": 0.6503, + "step": 11714 + }, + { + "epoch": 0.5186898818945574, + "grad_norm": 1.3041211366653442, + "learning_rate": 1.1091997262228473e-05, + "loss": 0.5983, + "step": 11715 + }, + { + "epoch": 0.5187341575993712, + "grad_norm": 2.022596836090088, + "learning_rate": 1.1090461040323688e-05, + "loss": 0.7272, + "step": 11716 + }, + { + "epoch": 0.5187784333041852, + "grad_norm": 1.2813764810562134, + "learning_rate": 1.1088924792374127e-05, + "loss": 0.6139, + "step": 11717 + }, + { + "epoch": 0.5188227090089991, + "grad_norm": 1.4083434343338013, + "learning_rate": 1.1087388518416487e-05, + "loss": 0.6446, + "step": 11718 + }, + { + "epoch": 0.5188669847138129, + "grad_norm": 1.580336332321167, + "learning_rate": 1.1085852218487453e-05, + "loss": 0.7343, + "step": 11719 + }, + { + "epoch": 0.5189112604186268, + "grad_norm": 1.851520299911499, + "learning_rate": 1.108431589262373e-05, + "loss": 0.6958, + "step": 11720 + }, + { + "epoch": 0.5189555361234407, + "grad_norm": 1.297010064125061, + "learning_rate": 1.1082779540861998e-05, + "loss": 0.5991, + "step": 11721 + }, + { + "epoch": 0.5189998118282545, + "grad_norm": 1.5987024307250977, + "learning_rate": 1.1081243163238959e-05, + "loss": 0.7464, + "step": 11722 + }, + { + "epoch": 0.5190440875330684, + "grad_norm": 1.5140891075134277, + "learning_rate": 1.1079706759791311e-05, + "loss": 0.6463, + "step": 11723 + }, + { + "epoch": 0.5190883632378823, + "grad_norm": 1.3999086618423462, + "learning_rate": 1.1078170330555742e-05, + "loss": 0.6633, + "step": 11724 + }, + { + "epoch": 0.5191326389426961, + "grad_norm": 1.4053682088851929, + "learning_rate": 1.1076633875568959e-05, + "loss": 0.5884, + "step": 11725 + }, + { + "epoch": 0.51917691464751, + "grad_norm": 1.4945143461227417, + "learning_rate": 1.1075097394867646e-05, + "loss": 0.7332, + "step": 11726 + }, + { + "epoch": 0.519221190352324, + "grad_norm": 1.3802578449249268, + "learning_rate": 1.1073560888488507e-05, + "loss": 0.6059, + "step": 11727 + }, + { + "epoch": 0.5192654660571377, + "grad_norm": 1.4774810075759888, + "learning_rate": 1.1072024356468245e-05, + "loss": 0.6949, + "step": 11728 + }, + { + "epoch": 0.5193097417619517, + "grad_norm": 1.3623243570327759, + "learning_rate": 1.1070487798843552e-05, + "loss": 0.6249, + "step": 11729 + }, + { + "epoch": 0.5193540174667656, + "grad_norm": 1.7934246063232422, + "learning_rate": 1.1068951215651132e-05, + "loss": 0.6728, + "step": 11730 + }, + { + "epoch": 0.5193982931715794, + "grad_norm": 1.4772374629974365, + "learning_rate": 1.1067414606927677e-05, + "loss": 0.6942, + "step": 11731 + }, + { + "epoch": 0.5194425688763933, + "grad_norm": 1.616437315940857, + "learning_rate": 1.1065877972709894e-05, + "loss": 0.622, + "step": 11732 + }, + { + "epoch": 0.5194868445812072, + "grad_norm": 1.4161911010742188, + "learning_rate": 1.1064341313034487e-05, + "loss": 0.6198, + "step": 11733 + }, + { + "epoch": 0.519531120286021, + "grad_norm": 1.604422926902771, + "learning_rate": 1.1062804627938154e-05, + "loss": 0.7447, + "step": 11734 + }, + { + "epoch": 0.5195753959908349, + "grad_norm": 1.4683283567428589, + "learning_rate": 1.10612679174576e-05, + "loss": 0.6528, + "step": 11735 + }, + { + "epoch": 0.5196196716956488, + "grad_norm": 1.6432029008865356, + "learning_rate": 1.105973118162952e-05, + "loss": 0.6315, + "step": 11736 + }, + { + "epoch": 0.5196639474004627, + "grad_norm": 1.50845468044281, + "learning_rate": 1.1058194420490625e-05, + "loss": 0.643, + "step": 11737 + }, + { + "epoch": 0.5197082231052765, + "grad_norm": 1.4520697593688965, + "learning_rate": 1.1056657634077622e-05, + "loss": 0.6409, + "step": 11738 + }, + { + "epoch": 0.5197524988100904, + "grad_norm": 1.3062515258789062, + "learning_rate": 1.1055120822427208e-05, + "loss": 0.5844, + "step": 11739 + }, + { + "epoch": 0.5197967745149044, + "grad_norm": 1.5247150659561157, + "learning_rate": 1.1053583985576094e-05, + "loss": 0.6147, + "step": 11740 + }, + { + "epoch": 0.5198410502197182, + "grad_norm": 1.2939927577972412, + "learning_rate": 1.1052047123560982e-05, + "loss": 0.6549, + "step": 11741 + }, + { + "epoch": 0.5198853259245321, + "grad_norm": 1.182342529296875, + "learning_rate": 1.105051023641858e-05, + "loss": 0.6105, + "step": 11742 + }, + { + "epoch": 0.519929601629346, + "grad_norm": 1.4221031665802002, + "learning_rate": 1.10489733241856e-05, + "loss": 0.6727, + "step": 11743 + }, + { + "epoch": 0.5199738773341598, + "grad_norm": 1.6135693788528442, + "learning_rate": 1.1047436386898748e-05, + "loss": 0.6724, + "step": 11744 + }, + { + "epoch": 0.5200181530389737, + "grad_norm": 1.4574705362319946, + "learning_rate": 1.1045899424594729e-05, + "loss": 0.6728, + "step": 11745 + }, + { + "epoch": 0.5200624287437876, + "grad_norm": 1.6232516765594482, + "learning_rate": 1.1044362437310253e-05, + "loss": 0.7198, + "step": 11746 + }, + { + "epoch": 0.5201067044486014, + "grad_norm": 1.425757646560669, + "learning_rate": 1.1042825425082026e-05, + "loss": 0.5853, + "step": 11747 + }, + { + "epoch": 0.5201509801534153, + "grad_norm": 1.405544400215149, + "learning_rate": 1.1041288387946771e-05, + "loss": 0.6538, + "step": 11748 + }, + { + "epoch": 0.5201952558582292, + "grad_norm": 1.6532096862792969, + "learning_rate": 1.1039751325941187e-05, + "loss": 0.6899, + "step": 11749 + }, + { + "epoch": 0.520239531563043, + "grad_norm": 1.3440983295440674, + "learning_rate": 1.103821423910199e-05, + "loss": 0.6356, + "step": 11750 + }, + { + "epoch": 0.520283807267857, + "grad_norm": 1.4047926664352417, + "learning_rate": 1.103667712746589e-05, + "loss": 0.6878, + "step": 11751 + }, + { + "epoch": 0.5203280829726709, + "grad_norm": 1.2092562913894653, + "learning_rate": 1.10351399910696e-05, + "loss": 0.5982, + "step": 11752 + }, + { + "epoch": 0.5203723586774847, + "grad_norm": 1.6401880979537964, + "learning_rate": 1.1033602829949837e-05, + "loss": 0.6757, + "step": 11753 + }, + { + "epoch": 0.5204166343822986, + "grad_norm": 1.5771336555480957, + "learning_rate": 1.1032065644143308e-05, + "loss": 0.6744, + "step": 11754 + }, + { + "epoch": 0.5204609100871125, + "grad_norm": 1.46547532081604, + "learning_rate": 1.1030528433686736e-05, + "loss": 0.68, + "step": 11755 + }, + { + "epoch": 0.5205051857919263, + "grad_norm": 1.1565446853637695, + "learning_rate": 1.1028991198616828e-05, + "loss": 0.6417, + "step": 11756 + }, + { + "epoch": 0.5205494614967402, + "grad_norm": 1.3278324604034424, + "learning_rate": 1.1027453938970305e-05, + "loss": 0.6432, + "step": 11757 + }, + { + "epoch": 0.5205937372015541, + "grad_norm": 1.5359965562820435, + "learning_rate": 1.1025916654783881e-05, + "loss": 0.6644, + "step": 11758 + }, + { + "epoch": 0.5206380129063679, + "grad_norm": 1.4311134815216064, + "learning_rate": 1.1024379346094272e-05, + "loss": 0.6383, + "step": 11759 + }, + { + "epoch": 0.5206822886111818, + "grad_norm": 1.7201671600341797, + "learning_rate": 1.1022842012938197e-05, + "loss": 0.6929, + "step": 11760 + }, + { + "epoch": 0.5207265643159957, + "grad_norm": 1.6104761362075806, + "learning_rate": 1.1021304655352374e-05, + "loss": 0.5939, + "step": 11761 + }, + { + "epoch": 0.5207708400208095, + "grad_norm": 1.5156694650650024, + "learning_rate": 1.1019767273373521e-05, + "loss": 0.6361, + "step": 11762 + }, + { + "epoch": 0.5208151157256234, + "grad_norm": 2.5669679641723633, + "learning_rate": 1.1018229867038358e-05, + "loss": 0.6777, + "step": 11763 + }, + { + "epoch": 0.5208593914304374, + "grad_norm": 1.3728314638137817, + "learning_rate": 1.1016692436383598e-05, + "loss": 0.6038, + "step": 11764 + }, + { + "epoch": 0.5209036671352512, + "grad_norm": 1.5317469835281372, + "learning_rate": 1.1015154981445973e-05, + "loss": 0.6488, + "step": 11765 + }, + { + "epoch": 0.5209479428400651, + "grad_norm": 1.1504250764846802, + "learning_rate": 1.1013617502262195e-05, + "loss": 0.5892, + "step": 11766 + }, + { + "epoch": 0.520992218544879, + "grad_norm": 1.435773491859436, + "learning_rate": 1.101207999886899e-05, + "loss": 0.6669, + "step": 11767 + }, + { + "epoch": 0.5210364942496928, + "grad_norm": 1.2311978340148926, + "learning_rate": 1.1010542471303078e-05, + "loss": 0.677, + "step": 11768 + }, + { + "epoch": 0.5210807699545067, + "grad_norm": 1.4411795139312744, + "learning_rate": 1.1009004919601179e-05, + "loss": 0.6555, + "step": 11769 + }, + { + "epoch": 0.5211250456593206, + "grad_norm": 1.5938053131103516, + "learning_rate": 1.1007467343800022e-05, + "loss": 0.7441, + "step": 11770 + }, + { + "epoch": 0.5211693213641345, + "grad_norm": 1.2679715156555176, + "learning_rate": 1.1005929743936325e-05, + "loss": 0.5598, + "step": 11771 + }, + { + "epoch": 0.5212135970689483, + "grad_norm": 1.492379903793335, + "learning_rate": 1.1004392120046821e-05, + "loss": 0.7214, + "step": 11772 + }, + { + "epoch": 0.5212578727737622, + "grad_norm": 1.4794150590896606, + "learning_rate": 1.1002854472168225e-05, + "loss": 0.6728, + "step": 11773 + }, + { + "epoch": 0.5213021484785761, + "grad_norm": 1.5849659442901611, + "learning_rate": 1.1001316800337266e-05, + "loss": 0.6853, + "step": 11774 + }, + { + "epoch": 0.52134642418339, + "grad_norm": 1.6991935968399048, + "learning_rate": 1.0999779104590671e-05, + "loss": 0.6538, + "step": 11775 + }, + { + "epoch": 0.5213906998882039, + "grad_norm": 1.4327696561813354, + "learning_rate": 1.0998241384965166e-05, + "loss": 0.6882, + "step": 11776 + }, + { + "epoch": 0.5214349755930178, + "grad_norm": 1.3765686750411987, + "learning_rate": 1.099670364149748e-05, + "loss": 0.6174, + "step": 11777 + }, + { + "epoch": 0.5214792512978316, + "grad_norm": 1.2880592346191406, + "learning_rate": 1.0995165874224338e-05, + "loss": 0.6269, + "step": 11778 + }, + { + "epoch": 0.5215235270026455, + "grad_norm": 1.5275593996047974, + "learning_rate": 1.0993628083182468e-05, + "loss": 0.7126, + "step": 11779 + }, + { + "epoch": 0.5215678027074594, + "grad_norm": 1.470699429512024, + "learning_rate": 1.0992090268408602e-05, + "loss": 0.7216, + "step": 11780 + }, + { + "epoch": 0.5216120784122732, + "grad_norm": 1.2888554334640503, + "learning_rate": 1.0990552429939467e-05, + "loss": 0.6062, + "step": 11781 + }, + { + "epoch": 0.5216563541170871, + "grad_norm": 1.4649697542190552, + "learning_rate": 1.0989014567811793e-05, + "loss": 0.6731, + "step": 11782 + }, + { + "epoch": 0.521700629821901, + "grad_norm": 1.2922815084457397, + "learning_rate": 1.0987476682062314e-05, + "loss": 0.6008, + "step": 11783 + }, + { + "epoch": 0.5217449055267148, + "grad_norm": 1.5565404891967773, + "learning_rate": 1.0985938772727756e-05, + "loss": 0.7128, + "step": 11784 + }, + { + "epoch": 0.5217891812315287, + "grad_norm": 1.302502155303955, + "learning_rate": 1.0984400839844854e-05, + "loss": 0.5796, + "step": 11785 + }, + { + "epoch": 0.5218334569363426, + "grad_norm": 1.524362325668335, + "learning_rate": 1.098286288345034e-05, + "loss": 0.6868, + "step": 11786 + }, + { + "epoch": 0.5218777326411564, + "grad_norm": 1.5589778423309326, + "learning_rate": 1.0981324903580945e-05, + "loss": 0.6727, + "step": 11787 + }, + { + "epoch": 0.5219220083459704, + "grad_norm": 1.3769586086273193, + "learning_rate": 1.0979786900273406e-05, + "loss": 0.6542, + "step": 11788 + }, + { + "epoch": 0.5219662840507843, + "grad_norm": 1.4553258419036865, + "learning_rate": 1.0978248873564449e-05, + "loss": 0.6827, + "step": 11789 + }, + { + "epoch": 0.5220105597555981, + "grad_norm": 1.4648460149765015, + "learning_rate": 1.0976710823490822e-05, + "loss": 0.6846, + "step": 11790 + }, + { + "epoch": 0.522054835460412, + "grad_norm": 1.2846765518188477, + "learning_rate": 1.0975172750089248e-05, + "loss": 0.6727, + "step": 11791 + }, + { + "epoch": 0.5220991111652259, + "grad_norm": 1.1634769439697266, + "learning_rate": 1.0973634653396466e-05, + "loss": 0.6231, + "step": 11792 + }, + { + "epoch": 0.5221433868700397, + "grad_norm": 1.8264515399932861, + "learning_rate": 1.0972096533449218e-05, + "loss": 0.721, + "step": 11793 + }, + { + "epoch": 0.5221876625748536, + "grad_norm": 1.4495466947555542, + "learning_rate": 1.097055839028423e-05, + "loss": 0.6422, + "step": 11794 + }, + { + "epoch": 0.5222319382796675, + "grad_norm": 1.9231013059616089, + "learning_rate": 1.0969020223938246e-05, + "loss": 0.7028, + "step": 11795 + }, + { + "epoch": 0.5222762139844813, + "grad_norm": 1.6391535997390747, + "learning_rate": 1.0967482034448007e-05, + "loss": 0.6966, + "step": 11796 + }, + { + "epoch": 0.5223204896892952, + "grad_norm": 1.3084321022033691, + "learning_rate": 1.0965943821850243e-05, + "loss": 0.6284, + "step": 11797 + }, + { + "epoch": 0.5223647653941091, + "grad_norm": 1.6522653102874756, + "learning_rate": 1.0964405586181699e-05, + "loss": 0.6941, + "step": 11798 + }, + { + "epoch": 0.5224090410989229, + "grad_norm": 1.209638237953186, + "learning_rate": 1.0962867327479112e-05, + "loss": 0.5861, + "step": 11799 + }, + { + "epoch": 0.5224533168037369, + "grad_norm": 1.6568135023117065, + "learning_rate": 1.0961329045779228e-05, + "loss": 0.6508, + "step": 11800 + }, + { + "epoch": 0.5224975925085508, + "grad_norm": 1.5455269813537598, + "learning_rate": 1.0959790741118776e-05, + "loss": 0.6468, + "step": 11801 + }, + { + "epoch": 0.5225418682133646, + "grad_norm": 1.4432421922683716, + "learning_rate": 1.0958252413534506e-05, + "loss": 0.6821, + "step": 11802 + }, + { + "epoch": 0.5225861439181785, + "grad_norm": 1.3851113319396973, + "learning_rate": 1.0956714063063156e-05, + "loss": 0.5862, + "step": 11803 + }, + { + "epoch": 0.5226304196229924, + "grad_norm": 1.379199743270874, + "learning_rate": 1.0955175689741473e-05, + "loss": 0.6897, + "step": 11804 + }, + { + "epoch": 0.5226746953278063, + "grad_norm": 1.5045255422592163, + "learning_rate": 1.0953637293606194e-05, + "loss": 0.6155, + "step": 11805 + }, + { + "epoch": 0.5227189710326201, + "grad_norm": 1.534310221672058, + "learning_rate": 1.0952098874694064e-05, + "loss": 0.6971, + "step": 11806 + }, + { + "epoch": 0.522763246737434, + "grad_norm": 1.4338397979736328, + "learning_rate": 1.0950560433041825e-05, + "loss": 0.662, + "step": 11807 + }, + { + "epoch": 0.5228075224422479, + "grad_norm": 1.9567313194274902, + "learning_rate": 1.094902196868623e-05, + "loss": 0.7589, + "step": 11808 + }, + { + "epoch": 0.5228517981470617, + "grad_norm": 1.4919811487197876, + "learning_rate": 1.0947483481664017e-05, + "loss": 0.667, + "step": 11809 + }, + { + "epoch": 0.5228960738518756, + "grad_norm": 1.3973950147628784, + "learning_rate": 1.094594497201193e-05, + "loss": 0.6595, + "step": 11810 + }, + { + "epoch": 0.5229403495566896, + "grad_norm": 1.2258182764053345, + "learning_rate": 1.0944406439766719e-05, + "loss": 0.5605, + "step": 11811 + }, + { + "epoch": 0.5229846252615034, + "grad_norm": 1.439727783203125, + "learning_rate": 1.0942867884965127e-05, + "loss": 0.6973, + "step": 11812 + }, + { + "epoch": 0.5230289009663173, + "grad_norm": 1.538669466972351, + "learning_rate": 1.0941329307643907e-05, + "loss": 0.6521, + "step": 11813 + }, + { + "epoch": 0.5230731766711312, + "grad_norm": 1.3507128953933716, + "learning_rate": 1.0939790707839802e-05, + "loss": 0.6868, + "step": 11814 + }, + { + "epoch": 0.523117452375945, + "grad_norm": 1.59223473072052, + "learning_rate": 1.0938252085589561e-05, + "loss": 0.6548, + "step": 11815 + }, + { + "epoch": 0.5231617280807589, + "grad_norm": 1.8829236030578613, + "learning_rate": 1.093671344092993e-05, + "loss": 0.6146, + "step": 11816 + }, + { + "epoch": 0.5232060037855728, + "grad_norm": 1.2620705366134644, + "learning_rate": 1.0935174773897663e-05, + "loss": 0.5898, + "step": 11817 + }, + { + "epoch": 0.5232502794903866, + "grad_norm": 1.3672446012496948, + "learning_rate": 1.0933636084529507e-05, + "loss": 0.6476, + "step": 11818 + }, + { + "epoch": 0.5232945551952005, + "grad_norm": 1.4700225591659546, + "learning_rate": 1.0932097372862216e-05, + "loss": 0.7277, + "step": 11819 + }, + { + "epoch": 0.5233388309000144, + "grad_norm": 1.4720591306686401, + "learning_rate": 1.0930558638932538e-05, + "loss": 0.6561, + "step": 11820 + }, + { + "epoch": 0.5233831066048282, + "grad_norm": 1.2788987159729004, + "learning_rate": 1.0929019882777221e-05, + "loss": 0.577, + "step": 11821 + }, + { + "epoch": 0.5234273823096421, + "grad_norm": 1.7938501834869385, + "learning_rate": 1.0927481104433024e-05, + "loss": 0.7142, + "step": 11822 + }, + { + "epoch": 0.523471658014456, + "grad_norm": 1.5359361171722412, + "learning_rate": 1.0925942303936694e-05, + "loss": 0.6719, + "step": 11823 + }, + { + "epoch": 0.5235159337192699, + "grad_norm": 1.4636907577514648, + "learning_rate": 1.0924403481324988e-05, + "loss": 0.6628, + "step": 11824 + }, + { + "epoch": 0.5235602094240838, + "grad_norm": 1.3872921466827393, + "learning_rate": 1.0922864636634656e-05, + "loss": 0.5985, + "step": 11825 + }, + { + "epoch": 0.5236044851288977, + "grad_norm": 1.524816632270813, + "learning_rate": 1.0921325769902454e-05, + "loss": 0.7013, + "step": 11826 + }, + { + "epoch": 0.5236487608337115, + "grad_norm": 1.3464263677597046, + "learning_rate": 1.0919786881165136e-05, + "loss": 0.7065, + "step": 11827 + }, + { + "epoch": 0.5236930365385254, + "grad_norm": 1.6238372325897217, + "learning_rate": 1.0918247970459456e-05, + "loss": 0.7341, + "step": 11828 + }, + { + "epoch": 0.5237373122433393, + "grad_norm": 1.4158178567886353, + "learning_rate": 1.0916709037822172e-05, + "loss": 0.6213, + "step": 11829 + }, + { + "epoch": 0.5237815879481531, + "grad_norm": 1.4835504293441772, + "learning_rate": 1.091517008329004e-05, + "loss": 0.6801, + "step": 11830 + }, + { + "epoch": 0.523825863652967, + "grad_norm": 1.3989990949630737, + "learning_rate": 1.0913631106899813e-05, + "loss": 0.6162, + "step": 11831 + }, + { + "epoch": 0.5238701393577809, + "grad_norm": 1.476447343826294, + "learning_rate": 1.0912092108688254e-05, + "loss": 0.6685, + "step": 11832 + }, + { + "epoch": 0.5239144150625947, + "grad_norm": 1.4024529457092285, + "learning_rate": 1.091055308869212e-05, + "loss": 0.6472, + "step": 11833 + }, + { + "epoch": 0.5239586907674086, + "grad_norm": 1.33005690574646, + "learning_rate": 1.0909014046948161e-05, + "loss": 0.6037, + "step": 11834 + }, + { + "epoch": 0.5240029664722226, + "grad_norm": 1.8155648708343506, + "learning_rate": 1.0907474983493144e-05, + "loss": 0.691, + "step": 11835 + }, + { + "epoch": 0.5240472421770364, + "grad_norm": 1.4291224479675293, + "learning_rate": 1.0905935898363824e-05, + "loss": 0.6668, + "step": 11836 + }, + { + "epoch": 0.5240915178818503, + "grad_norm": 1.4157774448394775, + "learning_rate": 1.0904396791596968e-05, + "loss": 0.6693, + "step": 11837 + }, + { + "epoch": 0.5241357935866642, + "grad_norm": 1.4600698947906494, + "learning_rate": 1.0902857663229326e-05, + "loss": 0.6136, + "step": 11838 + }, + { + "epoch": 0.524180069291478, + "grad_norm": 1.2713509798049927, + "learning_rate": 1.0901318513297666e-05, + "loss": 0.6505, + "step": 11839 + }, + { + "epoch": 0.5242243449962919, + "grad_norm": 1.2980859279632568, + "learning_rate": 1.0899779341838748e-05, + "loss": 0.596, + "step": 11840 + }, + { + "epoch": 0.5242686207011058, + "grad_norm": 1.1726796627044678, + "learning_rate": 1.089824014888933e-05, + "loss": 0.5547, + "step": 11841 + }, + { + "epoch": 0.5243128964059197, + "grad_norm": 1.392203450202942, + "learning_rate": 1.089670093448618e-05, + "loss": 0.5884, + "step": 11842 + }, + { + "epoch": 0.5243571721107335, + "grad_norm": 1.371381163597107, + "learning_rate": 1.0895161698666057e-05, + "loss": 0.6214, + "step": 11843 + }, + { + "epoch": 0.5244014478155474, + "grad_norm": 1.5917842388153076, + "learning_rate": 1.0893622441465726e-05, + "loss": 0.6434, + "step": 11844 + }, + { + "epoch": 0.5244457235203613, + "grad_norm": 1.4186530113220215, + "learning_rate": 1.0892083162921951e-05, + "loss": 0.6463, + "step": 11845 + }, + { + "epoch": 0.5244899992251751, + "grad_norm": 1.1953752040863037, + "learning_rate": 1.0890543863071494e-05, + "loss": 0.5674, + "step": 11846 + }, + { + "epoch": 0.524534274929989, + "grad_norm": 1.2140522003173828, + "learning_rate": 1.0889004541951125e-05, + "loss": 0.5676, + "step": 11847 + }, + { + "epoch": 0.524578550634803, + "grad_norm": 1.866661548614502, + "learning_rate": 1.0887465199597606e-05, + "loss": 0.7539, + "step": 11848 + }, + { + "epoch": 0.5246228263396168, + "grad_norm": 1.5325897932052612, + "learning_rate": 1.0885925836047701e-05, + "loss": 0.6696, + "step": 11849 + }, + { + "epoch": 0.5246671020444307, + "grad_norm": 1.4797511100769043, + "learning_rate": 1.0884386451338181e-05, + "loss": 0.6622, + "step": 11850 + }, + { + "epoch": 0.5247113777492446, + "grad_norm": 1.4702019691467285, + "learning_rate": 1.0882847045505809e-05, + "loss": 0.6769, + "step": 11851 + }, + { + "epoch": 0.5247556534540584, + "grad_norm": 1.481934666633606, + "learning_rate": 1.0881307618587358e-05, + "loss": 0.7209, + "step": 11852 + }, + { + "epoch": 0.5247999291588723, + "grad_norm": 1.2864620685577393, + "learning_rate": 1.087976817061959e-05, + "loss": 0.6261, + "step": 11853 + }, + { + "epoch": 0.5248442048636862, + "grad_norm": 1.4153391122817993, + "learning_rate": 1.0878228701639271e-05, + "loss": 0.6239, + "step": 11854 + }, + { + "epoch": 0.5248884805685, + "grad_norm": 1.3812471628189087, + "learning_rate": 1.087668921168318e-05, + "loss": 0.6307, + "step": 11855 + }, + { + "epoch": 0.5249327562733139, + "grad_norm": 1.1014883518218994, + "learning_rate": 1.0875149700788078e-05, + "loss": 0.5644, + "step": 11856 + }, + { + "epoch": 0.5249770319781278, + "grad_norm": 1.4385117292404175, + "learning_rate": 1.0873610168990743e-05, + "loss": 0.6639, + "step": 11857 + }, + { + "epoch": 0.5250213076829416, + "grad_norm": 1.2171871662139893, + "learning_rate": 1.0872070616327934e-05, + "loss": 0.5789, + "step": 11858 + }, + { + "epoch": 0.5250655833877556, + "grad_norm": 1.5274296998977661, + "learning_rate": 1.087053104283643e-05, + "loss": 0.6402, + "step": 11859 + }, + { + "epoch": 0.5251098590925695, + "grad_norm": 1.4284390211105347, + "learning_rate": 1.0868991448553003e-05, + "loss": 0.6787, + "step": 11860 + }, + { + "epoch": 0.5251541347973833, + "grad_norm": 1.1126381158828735, + "learning_rate": 1.0867451833514422e-05, + "loss": 0.5473, + "step": 11861 + }, + { + "epoch": 0.5251984105021972, + "grad_norm": 1.3190830945968628, + "learning_rate": 1.086591219775746e-05, + "loss": 0.6107, + "step": 11862 + }, + { + "epoch": 0.5252426862070111, + "grad_norm": 1.3196038007736206, + "learning_rate": 1.0864372541318891e-05, + "loss": 0.6582, + "step": 11863 + }, + { + "epoch": 0.5252869619118249, + "grad_norm": 1.3454301357269287, + "learning_rate": 1.0862832864235485e-05, + "loss": 0.5848, + "step": 11864 + }, + { + "epoch": 0.5253312376166388, + "grad_norm": 1.5217702388763428, + "learning_rate": 1.0861293166544021e-05, + "loss": 0.6333, + "step": 11865 + }, + { + "epoch": 0.5253755133214527, + "grad_norm": 1.3559012413024902, + "learning_rate": 1.0859753448281271e-05, + "loss": 0.6549, + "step": 11866 + }, + { + "epoch": 0.5254197890262665, + "grad_norm": 1.4481984376907349, + "learning_rate": 1.0858213709484012e-05, + "loss": 0.6628, + "step": 11867 + }, + { + "epoch": 0.5254640647310804, + "grad_norm": 1.479689121246338, + "learning_rate": 1.0856673950189012e-05, + "loss": 0.5868, + "step": 11868 + }, + { + "epoch": 0.5255083404358943, + "grad_norm": 1.789029836654663, + "learning_rate": 1.0855134170433053e-05, + "loss": 0.6944, + "step": 11869 + }, + { + "epoch": 0.5255526161407081, + "grad_norm": 1.5839449167251587, + "learning_rate": 1.0853594370252913e-05, + "loss": 0.744, + "step": 11870 + }, + { + "epoch": 0.525596891845522, + "grad_norm": 1.6264665126800537, + "learning_rate": 1.0852054549685367e-05, + "loss": 0.6568, + "step": 11871 + }, + { + "epoch": 0.525641167550336, + "grad_norm": 1.484681487083435, + "learning_rate": 1.0850514708767193e-05, + "loss": 0.6524, + "step": 11872 + }, + { + "epoch": 0.5256854432551498, + "grad_norm": 1.5737186670303345, + "learning_rate": 1.0848974847535163e-05, + "loss": 0.6447, + "step": 11873 + }, + { + "epoch": 0.5257297189599637, + "grad_norm": 1.242061972618103, + "learning_rate": 1.0847434966026062e-05, + "loss": 0.6341, + "step": 11874 + }, + { + "epoch": 0.5257739946647776, + "grad_norm": 1.3963611125946045, + "learning_rate": 1.084589506427667e-05, + "loss": 0.6308, + "step": 11875 + }, + { + "epoch": 0.5258182703695915, + "grad_norm": 1.7348158359527588, + "learning_rate": 1.0844355142323759e-05, + "loss": 0.7489, + "step": 11876 + }, + { + "epoch": 0.5258625460744053, + "grad_norm": 1.2768748998641968, + "learning_rate": 1.0842815200204116e-05, + "loss": 0.6187, + "step": 11877 + }, + { + "epoch": 0.5259068217792192, + "grad_norm": 1.3930720090866089, + "learning_rate": 1.0841275237954517e-05, + "loss": 0.6738, + "step": 11878 + }, + { + "epoch": 0.5259510974840331, + "grad_norm": 1.2181358337402344, + "learning_rate": 1.0839735255611741e-05, + "loss": 0.6015, + "step": 11879 + }, + { + "epoch": 0.5259953731888469, + "grad_norm": 1.6075475215911865, + "learning_rate": 1.0838195253212578e-05, + "loss": 0.6299, + "step": 11880 + }, + { + "epoch": 0.5260396488936608, + "grad_norm": 1.9151233434677124, + "learning_rate": 1.0836655230793797e-05, + "loss": 0.7682, + "step": 11881 + }, + { + "epoch": 0.5260839245984748, + "grad_norm": 1.2770448923110962, + "learning_rate": 1.0835115188392193e-05, + "loss": 0.6245, + "step": 11882 + }, + { + "epoch": 0.5261282003032886, + "grad_norm": 1.8762903213500977, + "learning_rate": 1.083357512604454e-05, + "loss": 0.6861, + "step": 11883 + }, + { + "epoch": 0.5261724760081025, + "grad_norm": 1.41744863986969, + "learning_rate": 1.0832035043787625e-05, + "loss": 0.6288, + "step": 11884 + }, + { + "epoch": 0.5262167517129164, + "grad_norm": 1.4322755336761475, + "learning_rate": 1.083049494165823e-05, + "loss": 0.6097, + "step": 11885 + }, + { + "epoch": 0.5262610274177302, + "grad_norm": 1.4294975996017456, + "learning_rate": 1.0828954819693138e-05, + "loss": 0.6514, + "step": 11886 + }, + { + "epoch": 0.5263053031225441, + "grad_norm": 1.6767593622207642, + "learning_rate": 1.0827414677929137e-05, + "loss": 0.6513, + "step": 11887 + }, + { + "epoch": 0.526349578827358, + "grad_norm": 1.8391308784484863, + "learning_rate": 1.0825874516403008e-05, + "loss": 0.725, + "step": 11888 + }, + { + "epoch": 0.5263938545321718, + "grad_norm": 1.2976958751678467, + "learning_rate": 1.082433433515154e-05, + "loss": 0.6502, + "step": 11889 + }, + { + "epoch": 0.5264381302369857, + "grad_norm": 1.5117130279541016, + "learning_rate": 1.0822794134211521e-05, + "loss": 0.6195, + "step": 11890 + }, + { + "epoch": 0.5264824059417996, + "grad_norm": 1.3599592447280884, + "learning_rate": 1.0821253913619727e-05, + "loss": 0.6423, + "step": 11891 + }, + { + "epoch": 0.5265266816466134, + "grad_norm": 1.3282684087753296, + "learning_rate": 1.0819713673412956e-05, + "loss": 0.6815, + "step": 11892 + }, + { + "epoch": 0.5265709573514273, + "grad_norm": 1.3934141397476196, + "learning_rate": 1.0818173413627991e-05, + "loss": 0.5965, + "step": 11893 + }, + { + "epoch": 0.5266152330562412, + "grad_norm": 1.5457264184951782, + "learning_rate": 1.0816633134301622e-05, + "loss": 0.6238, + "step": 11894 + }, + { + "epoch": 0.526659508761055, + "grad_norm": 1.2584037780761719, + "learning_rate": 1.0815092835470633e-05, + "loss": 0.6803, + "step": 11895 + }, + { + "epoch": 0.526703784465869, + "grad_norm": 1.5759936571121216, + "learning_rate": 1.0813552517171815e-05, + "loss": 0.6648, + "step": 11896 + }, + { + "epoch": 0.5267480601706829, + "grad_norm": 1.3617255687713623, + "learning_rate": 1.081201217944196e-05, + "loss": 0.6056, + "step": 11897 + }, + { + "epoch": 0.5267923358754967, + "grad_norm": 1.953980803489685, + "learning_rate": 1.0810471822317852e-05, + "loss": 0.66, + "step": 11898 + }, + { + "epoch": 0.5268366115803106, + "grad_norm": 1.4800457954406738, + "learning_rate": 1.0808931445836289e-05, + "loss": 0.7443, + "step": 11899 + }, + { + "epoch": 0.5268808872851245, + "grad_norm": 1.520614743232727, + "learning_rate": 1.0807391050034053e-05, + "loss": 0.7017, + "step": 11900 + }, + { + "epoch": 0.5269251629899383, + "grad_norm": 1.1988798379898071, + "learning_rate": 1.0805850634947941e-05, + "loss": 0.5751, + "step": 11901 + }, + { + "epoch": 0.5269694386947522, + "grad_norm": 1.4805827140808105, + "learning_rate": 1.0804310200614742e-05, + "loss": 0.6624, + "step": 11902 + }, + { + "epoch": 0.5270137143995661, + "grad_norm": 2.2548556327819824, + "learning_rate": 1.080276974707125e-05, + "loss": 0.7408, + "step": 11903 + }, + { + "epoch": 0.5270579901043799, + "grad_norm": 1.5733259916305542, + "learning_rate": 1.0801229274354255e-05, + "loss": 0.6298, + "step": 11904 + }, + { + "epoch": 0.5271022658091938, + "grad_norm": 1.4442901611328125, + "learning_rate": 1.0799688782500551e-05, + "loss": 0.618, + "step": 11905 + }, + { + "epoch": 0.5271465415140077, + "grad_norm": 1.4688562154769897, + "learning_rate": 1.079814827154693e-05, + "loss": 0.5785, + "step": 11906 + }, + { + "epoch": 0.5271908172188216, + "grad_norm": 1.299869418144226, + "learning_rate": 1.0796607741530191e-05, + "loss": 0.6555, + "step": 11907 + }, + { + "epoch": 0.5272350929236355, + "grad_norm": 1.8309592008590698, + "learning_rate": 1.0795067192487123e-05, + "loss": 0.7317, + "step": 11908 + }, + { + "epoch": 0.5272793686284494, + "grad_norm": 1.9684691429138184, + "learning_rate": 1.0793526624454522e-05, + "loss": 0.6372, + "step": 11909 + }, + { + "epoch": 0.5273236443332632, + "grad_norm": 1.401977300643921, + "learning_rate": 1.0791986037469186e-05, + "loss": 0.6372, + "step": 11910 + }, + { + "epoch": 0.5273679200380771, + "grad_norm": 1.504442572593689, + "learning_rate": 1.0790445431567904e-05, + "loss": 0.6589, + "step": 11911 + }, + { + "epoch": 0.527412195742891, + "grad_norm": 1.3314367532730103, + "learning_rate": 1.0788904806787481e-05, + "loss": 0.617, + "step": 11912 + }, + { + "epoch": 0.5274564714477049, + "grad_norm": 1.3580714464187622, + "learning_rate": 1.0787364163164708e-05, + "loss": 0.6474, + "step": 11913 + }, + { + "epoch": 0.5275007471525187, + "grad_norm": 1.6526412963867188, + "learning_rate": 1.078582350073638e-05, + "loss": 0.5895, + "step": 11914 + }, + { + "epoch": 0.5275450228573326, + "grad_norm": 1.464591383934021, + "learning_rate": 1.07842828195393e-05, + "loss": 0.6701, + "step": 11915 + }, + { + "epoch": 0.5275892985621465, + "grad_norm": 1.9698736667633057, + "learning_rate": 1.0782742119610264e-05, + "loss": 0.7187, + "step": 11916 + }, + { + "epoch": 0.5276335742669603, + "grad_norm": 1.4520970582962036, + "learning_rate": 1.0781201400986069e-05, + "loss": 0.6469, + "step": 11917 + }, + { + "epoch": 0.5276778499717742, + "grad_norm": 1.6260628700256348, + "learning_rate": 1.0779660663703515e-05, + "loss": 0.6771, + "step": 11918 + }, + { + "epoch": 0.5277221256765882, + "grad_norm": 1.3388117551803589, + "learning_rate": 1.0778119907799399e-05, + "loss": 0.6818, + "step": 11919 + }, + { + "epoch": 0.527766401381402, + "grad_norm": 1.515285849571228, + "learning_rate": 1.0776579133310525e-05, + "loss": 0.6981, + "step": 11920 + }, + { + "epoch": 0.5278106770862159, + "grad_norm": 1.1363879442214966, + "learning_rate": 1.0775038340273689e-05, + "loss": 0.6088, + "step": 11921 + }, + { + "epoch": 0.5278549527910298, + "grad_norm": 1.9291138648986816, + "learning_rate": 1.0773497528725694e-05, + "loss": 0.7095, + "step": 11922 + }, + { + "epoch": 0.5278992284958436, + "grad_norm": 1.6596437692642212, + "learning_rate": 1.077195669870334e-05, + "loss": 0.7278, + "step": 11923 + }, + { + "epoch": 0.5279435042006575, + "grad_norm": 1.5605671405792236, + "learning_rate": 1.077041585024343e-05, + "loss": 0.7216, + "step": 11924 + }, + { + "epoch": 0.5279877799054714, + "grad_norm": 1.4633581638336182, + "learning_rate": 1.0768874983382765e-05, + "loss": 0.5406, + "step": 11925 + }, + { + "epoch": 0.5280320556102852, + "grad_norm": 1.534359335899353, + "learning_rate": 1.0767334098158144e-05, + "loss": 0.7101, + "step": 11926 + }, + { + "epoch": 0.5280763313150991, + "grad_norm": 1.6963615417480469, + "learning_rate": 1.0765793194606377e-05, + "loss": 0.654, + "step": 11927 + }, + { + "epoch": 0.528120607019913, + "grad_norm": 1.4994679689407349, + "learning_rate": 1.076425227276426e-05, + "loss": 0.621, + "step": 11928 + }, + { + "epoch": 0.5281648827247268, + "grad_norm": 1.2640838623046875, + "learning_rate": 1.07627113326686e-05, + "loss": 0.6229, + "step": 11929 + }, + { + "epoch": 0.5282091584295407, + "grad_norm": 1.8194153308868408, + "learning_rate": 1.0761170374356202e-05, + "loss": 0.7702, + "step": 11930 + }, + { + "epoch": 0.5282534341343547, + "grad_norm": 1.2856167554855347, + "learning_rate": 1.075962939786387e-05, + "loss": 0.6361, + "step": 11931 + }, + { + "epoch": 0.5282977098391685, + "grad_norm": 1.3779045343399048, + "learning_rate": 1.0758088403228408e-05, + "loss": 0.6421, + "step": 11932 + }, + { + "epoch": 0.5283419855439824, + "grad_norm": 1.3926119804382324, + "learning_rate": 1.0756547390486618e-05, + "loss": 0.6445, + "step": 11933 + }, + { + "epoch": 0.5283862612487963, + "grad_norm": 1.764419674873352, + "learning_rate": 1.0755006359675315e-05, + "loss": 0.6669, + "step": 11934 + }, + { + "epoch": 0.5284305369536101, + "grad_norm": 1.2606645822525024, + "learning_rate": 1.0753465310831295e-05, + "loss": 0.6521, + "step": 11935 + }, + { + "epoch": 0.528474812658424, + "grad_norm": 1.8838636875152588, + "learning_rate": 1.0751924243991374e-05, + "loss": 0.7572, + "step": 11936 + }, + { + "epoch": 0.5285190883632379, + "grad_norm": 1.4137710332870483, + "learning_rate": 1.0750383159192353e-05, + "loss": 0.6865, + "step": 11937 + }, + { + "epoch": 0.5285633640680517, + "grad_norm": 1.3768141269683838, + "learning_rate": 1.0748842056471039e-05, + "loss": 0.6551, + "step": 11938 + }, + { + "epoch": 0.5286076397728656, + "grad_norm": 1.331069827079773, + "learning_rate": 1.0747300935864245e-05, + "loss": 0.6235, + "step": 11939 + }, + { + "epoch": 0.5286519154776795, + "grad_norm": 1.4737977981567383, + "learning_rate": 1.0745759797408774e-05, + "loss": 0.6505, + "step": 11940 + }, + { + "epoch": 0.5286961911824933, + "grad_norm": 1.6929885149002075, + "learning_rate": 1.0744218641141439e-05, + "loss": 0.6485, + "step": 11941 + }, + { + "epoch": 0.5287404668873072, + "grad_norm": 1.5082039833068848, + "learning_rate": 1.074267746709905e-05, + "loss": 0.668, + "step": 11942 + }, + { + "epoch": 0.5287847425921212, + "grad_norm": 1.413593053817749, + "learning_rate": 1.0741136275318411e-05, + "loss": 0.6453, + "step": 11943 + }, + { + "epoch": 0.528829018296935, + "grad_norm": 1.5804792642593384, + "learning_rate": 1.0739595065836336e-05, + "loss": 0.6425, + "step": 11944 + }, + { + "epoch": 0.5288732940017489, + "grad_norm": 1.41796875, + "learning_rate": 1.0738053838689635e-05, + "loss": 0.5797, + "step": 11945 + }, + { + "epoch": 0.5289175697065628, + "grad_norm": 1.410764455795288, + "learning_rate": 1.073651259391512e-05, + "loss": 0.6604, + "step": 11946 + }, + { + "epoch": 0.5289618454113767, + "grad_norm": 1.316035270690918, + "learning_rate": 1.0734971331549604e-05, + "loss": 0.6298, + "step": 11947 + }, + { + "epoch": 0.5290061211161905, + "grad_norm": 1.3787444829940796, + "learning_rate": 1.0733430051629892e-05, + "loss": 0.6418, + "step": 11948 + }, + { + "epoch": 0.5290503968210044, + "grad_norm": 1.5639969110488892, + "learning_rate": 1.0731888754192804e-05, + "loss": 0.6133, + "step": 11949 + }, + { + "epoch": 0.5290946725258183, + "grad_norm": 1.5291576385498047, + "learning_rate": 1.0730347439275144e-05, + "loss": 0.6439, + "step": 11950 + }, + { + "epoch": 0.5291389482306321, + "grad_norm": 1.6715794801712036, + "learning_rate": 1.0728806106913735e-05, + "loss": 0.6737, + "step": 11951 + }, + { + "epoch": 0.529183223935446, + "grad_norm": 1.4731899499893188, + "learning_rate": 1.0727264757145382e-05, + "loss": 0.6217, + "step": 11952 + }, + { + "epoch": 0.52922749964026, + "grad_norm": 1.5911158323287964, + "learning_rate": 1.0725723390006904e-05, + "loss": 0.6753, + "step": 11953 + }, + { + "epoch": 0.5292717753450737, + "grad_norm": 1.326931118965149, + "learning_rate": 1.0724182005535113e-05, + "loss": 0.6204, + "step": 11954 + }, + { + "epoch": 0.5293160510498877, + "grad_norm": 1.4910372495651245, + "learning_rate": 1.0722640603766825e-05, + "loss": 0.632, + "step": 11955 + }, + { + "epoch": 0.5293603267547016, + "grad_norm": 1.3294917345046997, + "learning_rate": 1.0721099184738855e-05, + "loss": 0.6311, + "step": 11956 + }, + { + "epoch": 0.5294046024595154, + "grad_norm": 1.751880407333374, + "learning_rate": 1.0719557748488018e-05, + "loss": 0.6823, + "step": 11957 + }, + { + "epoch": 0.5294488781643293, + "grad_norm": 1.6723213195800781, + "learning_rate": 1.0718016295051129e-05, + "loss": 0.6843, + "step": 11958 + }, + { + "epoch": 0.5294931538691432, + "grad_norm": 1.310228705406189, + "learning_rate": 1.0716474824465007e-05, + "loss": 0.6154, + "step": 11959 + }, + { + "epoch": 0.529537429573957, + "grad_norm": 1.7458163499832153, + "learning_rate": 1.0714933336766466e-05, + "loss": 0.7567, + "step": 11960 + }, + { + "epoch": 0.5295817052787709, + "grad_norm": 1.1425490379333496, + "learning_rate": 1.0713391831992324e-05, + "loss": 0.5867, + "step": 11961 + }, + { + "epoch": 0.5296259809835848, + "grad_norm": 1.221977710723877, + "learning_rate": 1.0711850310179401e-05, + "loss": 0.5803, + "step": 11962 + }, + { + "epoch": 0.5296702566883986, + "grad_norm": 1.7445569038391113, + "learning_rate": 1.071030877136451e-05, + "loss": 0.7184, + "step": 11963 + }, + { + "epoch": 0.5297145323932125, + "grad_norm": 1.6015560626983643, + "learning_rate": 1.0708767215584474e-05, + "loss": 0.6897, + "step": 11964 + }, + { + "epoch": 0.5297588080980264, + "grad_norm": 1.340035319328308, + "learning_rate": 1.0707225642876113e-05, + "loss": 0.6673, + "step": 11965 + }, + { + "epoch": 0.5298030838028402, + "grad_norm": 1.3298856019973755, + "learning_rate": 1.0705684053276239e-05, + "loss": 0.6259, + "step": 11966 + }, + { + "epoch": 0.5298473595076542, + "grad_norm": 1.7124907970428467, + "learning_rate": 1.0704142446821677e-05, + "loss": 0.6923, + "step": 11967 + }, + { + "epoch": 0.5298916352124681, + "grad_norm": 1.8367605209350586, + "learning_rate": 1.0702600823549245e-05, + "loss": 0.6808, + "step": 11968 + }, + { + "epoch": 0.5299359109172819, + "grad_norm": 1.3963388204574585, + "learning_rate": 1.0701059183495767e-05, + "loss": 0.6953, + "step": 11969 + }, + { + "epoch": 0.5299801866220958, + "grad_norm": 1.4057649374008179, + "learning_rate": 1.0699517526698063e-05, + "loss": 0.6384, + "step": 11970 + }, + { + "epoch": 0.5300244623269097, + "grad_norm": 1.6483875513076782, + "learning_rate": 1.0697975853192948e-05, + "loss": 0.7307, + "step": 11971 + }, + { + "epoch": 0.5300687380317235, + "grad_norm": 1.2430301904678345, + "learning_rate": 1.069643416301725e-05, + "loss": 0.5913, + "step": 11972 + }, + { + "epoch": 0.5301130137365374, + "grad_norm": 1.227330207824707, + "learning_rate": 1.0694892456207788e-05, + "loss": 0.5966, + "step": 11973 + }, + { + "epoch": 0.5301572894413513, + "grad_norm": 1.3078557252883911, + "learning_rate": 1.0693350732801386e-05, + "loss": 0.6047, + "step": 11974 + }, + { + "epoch": 0.5302015651461651, + "grad_norm": 1.5062494277954102, + "learning_rate": 1.0691808992834866e-05, + "loss": 0.5974, + "step": 11975 + }, + { + "epoch": 0.530245840850979, + "grad_norm": 1.4669440984725952, + "learning_rate": 1.0690267236345053e-05, + "loss": 0.6331, + "step": 11976 + }, + { + "epoch": 0.530290116555793, + "grad_norm": 1.2679914236068726, + "learning_rate": 1.0688725463368767e-05, + "loss": 0.609, + "step": 11977 + }, + { + "epoch": 0.5303343922606067, + "grad_norm": 1.5464627742767334, + "learning_rate": 1.0687183673942835e-05, + "loss": 0.6079, + "step": 11978 + }, + { + "epoch": 0.5303786679654207, + "grad_norm": 1.4918781518936157, + "learning_rate": 1.0685641868104085e-05, + "loss": 0.6579, + "step": 11979 + }, + { + "epoch": 0.5304229436702346, + "grad_norm": 1.4416342973709106, + "learning_rate": 1.068410004588933e-05, + "loss": 0.6406, + "step": 11980 + }, + { + "epoch": 0.5304672193750484, + "grad_norm": 1.4848450422286987, + "learning_rate": 1.0682558207335405e-05, + "loss": 0.6837, + "step": 11981 + }, + { + "epoch": 0.5305114950798623, + "grad_norm": 1.5827581882476807, + "learning_rate": 1.0681016352479134e-05, + "loss": 0.6515, + "step": 11982 + }, + { + "epoch": 0.5305557707846762, + "grad_norm": 2.0958993434906006, + "learning_rate": 1.0679474481357342e-05, + "loss": 0.6881, + "step": 11983 + }, + { + "epoch": 0.5306000464894901, + "grad_norm": 1.3599098920822144, + "learning_rate": 1.0677932594006854e-05, + "loss": 0.664, + "step": 11984 + }, + { + "epoch": 0.5306443221943039, + "grad_norm": 1.4424721002578735, + "learning_rate": 1.0676390690464498e-05, + "loss": 0.7096, + "step": 11985 + }, + { + "epoch": 0.5306885978991178, + "grad_norm": 1.4426329135894775, + "learning_rate": 1.0674848770767101e-05, + "loss": 0.6193, + "step": 11986 + }, + { + "epoch": 0.5307328736039317, + "grad_norm": 1.2220711708068848, + "learning_rate": 1.067330683495149e-05, + "loss": 0.6034, + "step": 11987 + }, + { + "epoch": 0.5307771493087455, + "grad_norm": 1.8364102840423584, + "learning_rate": 1.0671764883054499e-05, + "loss": 0.6641, + "step": 11988 + }, + { + "epoch": 0.5308214250135594, + "grad_norm": 1.336848497390747, + "learning_rate": 1.0670222915112946e-05, + "loss": 0.5899, + "step": 11989 + }, + { + "epoch": 0.5308657007183734, + "grad_norm": 1.5633002519607544, + "learning_rate": 1.0668680931163663e-05, + "loss": 0.6835, + "step": 11990 + }, + { + "epoch": 0.5309099764231872, + "grad_norm": 1.357750654220581, + "learning_rate": 1.0667138931243481e-05, + "loss": 0.6059, + "step": 11991 + }, + { + "epoch": 0.5309542521280011, + "grad_norm": 1.4012209177017212, + "learning_rate": 1.0665596915389232e-05, + "loss": 0.6314, + "step": 11992 + }, + { + "epoch": 0.530998527832815, + "grad_norm": 1.5219241380691528, + "learning_rate": 1.066405488363774e-05, + "loss": 0.6858, + "step": 11993 + }, + { + "epoch": 0.5310428035376288, + "grad_norm": 1.138401746749878, + "learning_rate": 1.066251283602584e-05, + "loss": 0.6045, + "step": 11994 + }, + { + "epoch": 0.5310870792424427, + "grad_norm": 1.827368974685669, + "learning_rate": 1.0660970772590357e-05, + "loss": 0.7293, + "step": 11995 + }, + { + "epoch": 0.5311313549472566, + "grad_norm": 2.166163206100464, + "learning_rate": 1.0659428693368126e-05, + "loss": 0.6348, + "step": 11996 + }, + { + "epoch": 0.5311756306520704, + "grad_norm": 1.4813802242279053, + "learning_rate": 1.0657886598395979e-05, + "loss": 0.6612, + "step": 11997 + }, + { + "epoch": 0.5312199063568843, + "grad_norm": 1.4139869213104248, + "learning_rate": 1.0656344487710748e-05, + "loss": 0.6575, + "step": 11998 + }, + { + "epoch": 0.5312641820616982, + "grad_norm": 1.2311313152313232, + "learning_rate": 1.0654802361349262e-05, + "loss": 0.5782, + "step": 11999 + }, + { + "epoch": 0.531308457766512, + "grad_norm": 1.5383994579315186, + "learning_rate": 1.065326021934835e-05, + "loss": 0.6389, + "step": 12000 + }, + { + "epoch": 0.5313527334713259, + "grad_norm": 1.3850805759429932, + "learning_rate": 1.0651718061744854e-05, + "loss": 0.6288, + "step": 12001 + }, + { + "epoch": 0.5313970091761399, + "grad_norm": 1.281023383140564, + "learning_rate": 1.0650175888575605e-05, + "loss": 0.6276, + "step": 12002 + }, + { + "epoch": 0.5314412848809537, + "grad_norm": 1.4210796356201172, + "learning_rate": 1.064863369987743e-05, + "loss": 0.6334, + "step": 12003 + }, + { + "epoch": 0.5314855605857676, + "grad_norm": 1.3016201257705688, + "learning_rate": 1.0647091495687171e-05, + "loss": 0.7578, + "step": 12004 + }, + { + "epoch": 0.5315298362905815, + "grad_norm": 1.1642905473709106, + "learning_rate": 1.0645549276041655e-05, + "loss": 0.5677, + "step": 12005 + }, + { + "epoch": 0.5315741119953953, + "grad_norm": 1.3580701351165771, + "learning_rate": 1.064400704097772e-05, + "loss": 0.6255, + "step": 12006 + }, + { + "epoch": 0.5316183877002092, + "grad_norm": 1.4351836442947388, + "learning_rate": 1.0642464790532206e-05, + "loss": 0.6537, + "step": 12007 + }, + { + "epoch": 0.5316626634050231, + "grad_norm": 1.5152360200881958, + "learning_rate": 1.0640922524741939e-05, + "loss": 0.693, + "step": 12008 + }, + { + "epoch": 0.5317069391098369, + "grad_norm": 1.2782191038131714, + "learning_rate": 1.0639380243643763e-05, + "loss": 0.6349, + "step": 12009 + }, + { + "epoch": 0.5317512148146508, + "grad_norm": 1.5888278484344482, + "learning_rate": 1.0637837947274505e-05, + "loss": 0.6634, + "step": 12010 + }, + { + "epoch": 0.5317954905194647, + "grad_norm": 1.3306001424789429, + "learning_rate": 1.0636295635671012e-05, + "loss": 0.6081, + "step": 12011 + }, + { + "epoch": 0.5318397662242785, + "grad_norm": 1.2901164293289185, + "learning_rate": 1.0634753308870114e-05, + "loss": 0.5892, + "step": 12012 + }, + { + "epoch": 0.5318840419290924, + "grad_norm": 1.4000765085220337, + "learning_rate": 1.0633210966908649e-05, + "loss": 0.5995, + "step": 12013 + }, + { + "epoch": 0.5319283176339064, + "grad_norm": 1.5189762115478516, + "learning_rate": 1.0631668609823458e-05, + "loss": 0.6911, + "step": 12014 + }, + { + "epoch": 0.5319725933387202, + "grad_norm": 1.4748153686523438, + "learning_rate": 1.0630126237651373e-05, + "loss": 0.668, + "step": 12015 + }, + { + "epoch": 0.5320168690435341, + "grad_norm": 1.4506839513778687, + "learning_rate": 1.062858385042924e-05, + "loss": 0.6767, + "step": 12016 + }, + { + "epoch": 0.532061144748348, + "grad_norm": 1.1582618951797485, + "learning_rate": 1.0627041448193891e-05, + "loss": 0.6256, + "step": 12017 + }, + { + "epoch": 0.5321054204531619, + "grad_norm": 1.7704997062683105, + "learning_rate": 1.0625499030982167e-05, + "loss": 0.7002, + "step": 12018 + }, + { + "epoch": 0.5321496961579757, + "grad_norm": 1.5965831279754639, + "learning_rate": 1.0623956598830911e-05, + "loss": 0.6445, + "step": 12019 + }, + { + "epoch": 0.5321939718627896, + "grad_norm": 1.629710078239441, + "learning_rate": 1.0622414151776956e-05, + "loss": 0.645, + "step": 12020 + }, + { + "epoch": 0.5322382475676035, + "grad_norm": 1.2766672372817993, + "learning_rate": 1.0620871689857149e-05, + "loss": 0.6256, + "step": 12021 + }, + { + "epoch": 0.5322825232724173, + "grad_norm": 1.372301459312439, + "learning_rate": 1.0619329213108327e-05, + "loss": 0.6165, + "step": 12022 + }, + { + "epoch": 0.5323267989772312, + "grad_norm": 1.4300453662872314, + "learning_rate": 1.0617786721567327e-05, + "loss": 0.6535, + "step": 12023 + }, + { + "epoch": 0.5323710746820451, + "grad_norm": 1.4914582967758179, + "learning_rate": 1.0616244215270998e-05, + "loss": 0.639, + "step": 12024 + }, + { + "epoch": 0.5324153503868589, + "grad_norm": 1.4921681880950928, + "learning_rate": 1.0614701694256176e-05, + "loss": 0.6808, + "step": 12025 + }, + { + "epoch": 0.5324596260916729, + "grad_norm": 1.2306368350982666, + "learning_rate": 1.0613159158559707e-05, + "loss": 0.5994, + "step": 12026 + }, + { + "epoch": 0.5325039017964868, + "grad_norm": 1.2992966175079346, + "learning_rate": 1.0611616608218429e-05, + "loss": 0.6605, + "step": 12027 + }, + { + "epoch": 0.5325481775013006, + "grad_norm": 1.2559765577316284, + "learning_rate": 1.0610074043269185e-05, + "loss": 0.5962, + "step": 12028 + }, + { + "epoch": 0.5325924532061145, + "grad_norm": 1.413941740989685, + "learning_rate": 1.060853146374882e-05, + "loss": 0.6378, + "step": 12029 + }, + { + "epoch": 0.5326367289109284, + "grad_norm": 1.9796005487442017, + "learning_rate": 1.060698886969418e-05, + "loss": 0.6447, + "step": 12030 + }, + { + "epoch": 0.5326810046157422, + "grad_norm": 1.9916374683380127, + "learning_rate": 1.06054462611421e-05, + "loss": 0.6717, + "step": 12031 + }, + { + "epoch": 0.5327252803205561, + "grad_norm": 1.3732025623321533, + "learning_rate": 1.0603903638129432e-05, + "loss": 0.6189, + "step": 12032 + }, + { + "epoch": 0.53276955602537, + "grad_norm": 1.4305851459503174, + "learning_rate": 1.0602361000693016e-05, + "loss": 0.6677, + "step": 12033 + }, + { + "epoch": 0.5328138317301838, + "grad_norm": 1.4381895065307617, + "learning_rate": 1.0600818348869699e-05, + "loss": 0.6769, + "step": 12034 + }, + { + "epoch": 0.5328581074349977, + "grad_norm": 1.67184579372406, + "learning_rate": 1.0599275682696325e-05, + "loss": 0.612, + "step": 12035 + }, + { + "epoch": 0.5329023831398116, + "grad_norm": 1.6566842794418335, + "learning_rate": 1.0597733002209738e-05, + "loss": 0.6977, + "step": 12036 + }, + { + "epoch": 0.5329466588446254, + "grad_norm": 1.6331077814102173, + "learning_rate": 1.0596190307446786e-05, + "loss": 0.6586, + "step": 12037 + }, + { + "epoch": 0.5329909345494394, + "grad_norm": 1.330213189125061, + "learning_rate": 1.0594647598444312e-05, + "loss": 0.6467, + "step": 12038 + }, + { + "epoch": 0.5330352102542533, + "grad_norm": 1.4410916566848755, + "learning_rate": 1.0593104875239166e-05, + "loss": 0.5918, + "step": 12039 + }, + { + "epoch": 0.5330794859590671, + "grad_norm": 1.3447576761245728, + "learning_rate": 1.0591562137868195e-05, + "loss": 0.6392, + "step": 12040 + }, + { + "epoch": 0.533123761663881, + "grad_norm": 1.643802285194397, + "learning_rate": 1.0590019386368242e-05, + "loss": 0.6965, + "step": 12041 + }, + { + "epoch": 0.5331680373686949, + "grad_norm": 1.5280275344848633, + "learning_rate": 1.0588476620776154e-05, + "loss": 0.6216, + "step": 12042 + }, + { + "epoch": 0.5332123130735087, + "grad_norm": 1.398976445198059, + "learning_rate": 1.058693384112878e-05, + "loss": 0.6554, + "step": 12043 + }, + { + "epoch": 0.5332565887783226, + "grad_norm": 1.7713803052902222, + "learning_rate": 1.0585391047462977e-05, + "loss": 0.689, + "step": 12044 + }, + { + "epoch": 0.5333008644831365, + "grad_norm": 1.5511410236358643, + "learning_rate": 1.058384823981558e-05, + "loss": 0.6399, + "step": 12045 + }, + { + "epoch": 0.5333451401879503, + "grad_norm": 1.7150131464004517, + "learning_rate": 1.0582305418223443e-05, + "loss": 0.6757, + "step": 12046 + }, + { + "epoch": 0.5333894158927642, + "grad_norm": 1.4655895233154297, + "learning_rate": 1.0580762582723416e-05, + "loss": 0.6581, + "step": 12047 + }, + { + "epoch": 0.5334336915975781, + "grad_norm": 1.7288923263549805, + "learning_rate": 1.0579219733352346e-05, + "loss": 0.6828, + "step": 12048 + }, + { + "epoch": 0.5334779673023919, + "grad_norm": 1.4277836084365845, + "learning_rate": 1.0577676870147089e-05, + "loss": 0.7059, + "step": 12049 + }, + { + "epoch": 0.5335222430072059, + "grad_norm": 1.2217093706130981, + "learning_rate": 1.0576133993144486e-05, + "loss": 0.5668, + "step": 12050 + }, + { + "epoch": 0.5335665187120198, + "grad_norm": 1.3962029218673706, + "learning_rate": 1.0574591102381395e-05, + "loss": 0.6639, + "step": 12051 + }, + { + "epoch": 0.5336107944168337, + "grad_norm": 1.365127444267273, + "learning_rate": 1.057304819789466e-05, + "loss": 0.6251, + "step": 12052 + }, + { + "epoch": 0.5336550701216475, + "grad_norm": 1.5500181913375854, + "learning_rate": 1.0571505279721136e-05, + "loss": 0.6294, + "step": 12053 + }, + { + "epoch": 0.5336993458264614, + "grad_norm": 1.405389666557312, + "learning_rate": 1.0569962347897678e-05, + "loss": 0.6353, + "step": 12054 + }, + { + "epoch": 0.5337436215312753, + "grad_norm": 1.210343599319458, + "learning_rate": 1.0568419402461127e-05, + "loss": 0.6635, + "step": 12055 + }, + { + "epoch": 0.5337878972360891, + "grad_norm": 1.2291996479034424, + "learning_rate": 1.0566876443448346e-05, + "loss": 0.6584, + "step": 12056 + }, + { + "epoch": 0.533832172940903, + "grad_norm": 1.6486258506774902, + "learning_rate": 1.056533347089618e-05, + "loss": 0.7405, + "step": 12057 + }, + { + "epoch": 0.5338764486457169, + "grad_norm": 1.6163641214370728, + "learning_rate": 1.0563790484841486e-05, + "loss": 0.7003, + "step": 12058 + }, + { + "epoch": 0.5339207243505307, + "grad_norm": 1.3655072450637817, + "learning_rate": 1.0562247485321116e-05, + "loss": 0.621, + "step": 12059 + }, + { + "epoch": 0.5339650000553446, + "grad_norm": 1.557108759880066, + "learning_rate": 1.0560704472371919e-05, + "loss": 0.6384, + "step": 12060 + }, + { + "epoch": 0.5340092757601586, + "grad_norm": 1.3236833810806274, + "learning_rate": 1.0559161446030757e-05, + "loss": 0.5711, + "step": 12061 + }, + { + "epoch": 0.5340535514649724, + "grad_norm": 1.3776895999908447, + "learning_rate": 1.0557618406334473e-05, + "loss": 0.5799, + "step": 12062 + }, + { + "epoch": 0.5340978271697863, + "grad_norm": 1.734104871749878, + "learning_rate": 1.0556075353319931e-05, + "loss": 0.7959, + "step": 12063 + }, + { + "epoch": 0.5341421028746002, + "grad_norm": 1.345329761505127, + "learning_rate": 1.0554532287023985e-05, + "loss": 0.6373, + "step": 12064 + }, + { + "epoch": 0.534186378579414, + "grad_norm": 1.3416171073913574, + "learning_rate": 1.0552989207483482e-05, + "loss": 0.5804, + "step": 12065 + }, + { + "epoch": 0.5342306542842279, + "grad_norm": 1.4944485425949097, + "learning_rate": 1.0551446114735286e-05, + "loss": 0.6705, + "step": 12066 + }, + { + "epoch": 0.5342749299890418, + "grad_norm": 1.2605633735656738, + "learning_rate": 1.0549903008816244e-05, + "loss": 0.6176, + "step": 12067 + }, + { + "epoch": 0.5343192056938556, + "grad_norm": 1.6800084114074707, + "learning_rate": 1.054835988976322e-05, + "loss": 0.6852, + "step": 12068 + }, + { + "epoch": 0.5343634813986695, + "grad_norm": 1.1276766061782837, + "learning_rate": 1.0546816757613067e-05, + "loss": 0.5621, + "step": 12069 + }, + { + "epoch": 0.5344077571034834, + "grad_norm": 1.3515881299972534, + "learning_rate": 1.0545273612402637e-05, + "loss": 0.6393, + "step": 12070 + }, + { + "epoch": 0.5344520328082972, + "grad_norm": 1.3791553974151611, + "learning_rate": 1.0543730454168794e-05, + "loss": 0.6194, + "step": 12071 + }, + { + "epoch": 0.5344963085131111, + "grad_norm": 1.798172950744629, + "learning_rate": 1.0542187282948391e-05, + "loss": 0.6266, + "step": 12072 + }, + { + "epoch": 0.534540584217925, + "grad_norm": 1.3140782117843628, + "learning_rate": 1.0540644098778286e-05, + "loss": 0.5784, + "step": 12073 + }, + { + "epoch": 0.5345848599227389, + "grad_norm": 1.4975038766860962, + "learning_rate": 1.0539100901695338e-05, + "loss": 0.6862, + "step": 12074 + }, + { + "epoch": 0.5346291356275528, + "grad_norm": 1.38222074508667, + "learning_rate": 1.0537557691736402e-05, + "loss": 0.5944, + "step": 12075 + }, + { + "epoch": 0.5346734113323667, + "grad_norm": 1.4769326448440552, + "learning_rate": 1.053601446893834e-05, + "loss": 0.719, + "step": 12076 + }, + { + "epoch": 0.5347176870371805, + "grad_norm": 1.4021387100219727, + "learning_rate": 1.0534471233338007e-05, + "loss": 0.6256, + "step": 12077 + }, + { + "epoch": 0.5347619627419944, + "grad_norm": 1.4822819232940674, + "learning_rate": 1.0532927984972265e-05, + "loss": 0.6293, + "step": 12078 + }, + { + "epoch": 0.5348062384468083, + "grad_norm": 1.5634729862213135, + "learning_rate": 1.0531384723877974e-05, + "loss": 0.6371, + "step": 12079 + }, + { + "epoch": 0.5348505141516221, + "grad_norm": 1.4581537246704102, + "learning_rate": 1.0529841450091986e-05, + "loss": 0.652, + "step": 12080 + }, + { + "epoch": 0.534894789856436, + "grad_norm": 1.8828282356262207, + "learning_rate": 1.0528298163651172e-05, + "loss": 0.7148, + "step": 12081 + }, + { + "epoch": 0.5349390655612499, + "grad_norm": 1.366141676902771, + "learning_rate": 1.0526754864592386e-05, + "loss": 0.6553, + "step": 12082 + }, + { + "epoch": 0.5349833412660637, + "grad_norm": 1.565434217453003, + "learning_rate": 1.0525211552952485e-05, + "loss": 0.6705, + "step": 12083 + }, + { + "epoch": 0.5350276169708776, + "grad_norm": 1.7126327753067017, + "learning_rate": 1.0523668228768335e-05, + "loss": 0.7176, + "step": 12084 + }, + { + "epoch": 0.5350718926756916, + "grad_norm": 1.3019012212753296, + "learning_rate": 1.0522124892076797e-05, + "loss": 0.5983, + "step": 12085 + }, + { + "epoch": 0.5351161683805054, + "grad_norm": 1.3643008470535278, + "learning_rate": 1.0520581542914731e-05, + "loss": 0.6496, + "step": 12086 + }, + { + "epoch": 0.5351604440853193, + "grad_norm": 1.3629101514816284, + "learning_rate": 1.0519038181319e-05, + "loss": 0.5985, + "step": 12087 + }, + { + "epoch": 0.5352047197901332, + "grad_norm": 1.5546491146087646, + "learning_rate": 1.0517494807326463e-05, + "loss": 0.7175, + "step": 12088 + }, + { + "epoch": 0.5352489954949471, + "grad_norm": 1.4325445890426636, + "learning_rate": 1.0515951420973984e-05, + "loss": 0.6065, + "step": 12089 + }, + { + "epoch": 0.5352932711997609, + "grad_norm": 1.6225723028182983, + "learning_rate": 1.0514408022298424e-05, + "loss": 0.658, + "step": 12090 + }, + { + "epoch": 0.5353375469045748, + "grad_norm": 1.4532558917999268, + "learning_rate": 1.0512864611336648e-05, + "loss": 0.6717, + "step": 12091 + }, + { + "epoch": 0.5353818226093887, + "grad_norm": 1.4447420835494995, + "learning_rate": 1.051132118812552e-05, + "loss": 0.7049, + "step": 12092 + }, + { + "epoch": 0.5354260983142025, + "grad_norm": 1.4706865549087524, + "learning_rate": 1.05097777527019e-05, + "loss": 0.6465, + "step": 12093 + }, + { + "epoch": 0.5354703740190164, + "grad_norm": 1.7588481903076172, + "learning_rate": 1.0508234305102655e-05, + "loss": 0.7276, + "step": 12094 + }, + { + "epoch": 0.5355146497238303, + "grad_norm": 1.6357606649398804, + "learning_rate": 1.0506690845364646e-05, + "loss": 0.7119, + "step": 12095 + }, + { + "epoch": 0.5355589254286441, + "grad_norm": 1.6277445554733276, + "learning_rate": 1.050514737352474e-05, + "loss": 0.6238, + "step": 12096 + }, + { + "epoch": 0.535603201133458, + "grad_norm": 1.3638689517974854, + "learning_rate": 1.05036038896198e-05, + "loss": 0.6603, + "step": 12097 + }, + { + "epoch": 0.535647476838272, + "grad_norm": 1.7517656087875366, + "learning_rate": 1.050206039368669e-05, + "loss": 0.665, + "step": 12098 + }, + { + "epoch": 0.5356917525430858, + "grad_norm": 1.3912723064422607, + "learning_rate": 1.0500516885762278e-05, + "loss": 0.7112, + "step": 12099 + }, + { + "epoch": 0.5357360282478997, + "grad_norm": 1.75332772731781, + "learning_rate": 1.0498973365883425e-05, + "loss": 0.756, + "step": 12100 + }, + { + "epoch": 0.5357803039527136, + "grad_norm": 1.3340740203857422, + "learning_rate": 1.0497429834087002e-05, + "loss": 0.6077, + "step": 12101 + }, + { + "epoch": 0.5358245796575274, + "grad_norm": 1.5156662464141846, + "learning_rate": 1.0495886290409873e-05, + "loss": 0.7122, + "step": 12102 + }, + { + "epoch": 0.5358688553623413, + "grad_norm": 1.2305854558944702, + "learning_rate": 1.04943427348889e-05, + "loss": 0.5982, + "step": 12103 + }, + { + "epoch": 0.5359131310671552, + "grad_norm": 1.4992789030075073, + "learning_rate": 1.0492799167560958e-05, + "loss": 0.6502, + "step": 12104 + }, + { + "epoch": 0.535957406771969, + "grad_norm": 1.3339757919311523, + "learning_rate": 1.0491255588462905e-05, + "loss": 0.6644, + "step": 12105 + }, + { + "epoch": 0.5360016824767829, + "grad_norm": 1.821886658668518, + "learning_rate": 1.0489711997631618e-05, + "loss": 0.7036, + "step": 12106 + }, + { + "epoch": 0.5360459581815968, + "grad_norm": 1.7236480712890625, + "learning_rate": 1.0488168395103953e-05, + "loss": 0.7135, + "step": 12107 + }, + { + "epoch": 0.5360902338864106, + "grad_norm": 1.4383796453475952, + "learning_rate": 1.0486624780916784e-05, + "loss": 0.6749, + "step": 12108 + }, + { + "epoch": 0.5361345095912246, + "grad_norm": 1.399996280670166, + "learning_rate": 1.048508115510698e-05, + "loss": 0.6558, + "step": 12109 + }, + { + "epoch": 0.5361787852960385, + "grad_norm": 1.6212854385375977, + "learning_rate": 1.0483537517711408e-05, + "loss": 0.676, + "step": 12110 + }, + { + "epoch": 0.5362230610008523, + "grad_norm": 1.3732123374938965, + "learning_rate": 1.0481993868766935e-05, + "loss": 0.6685, + "step": 12111 + }, + { + "epoch": 0.5362673367056662, + "grad_norm": 1.4022932052612305, + "learning_rate": 1.048045020831043e-05, + "loss": 0.5827, + "step": 12112 + }, + { + "epoch": 0.5363116124104801, + "grad_norm": 1.2603158950805664, + "learning_rate": 1.0478906536378764e-05, + "loss": 0.6014, + "step": 12113 + }, + { + "epoch": 0.5363558881152939, + "grad_norm": 1.4752181768417358, + "learning_rate": 1.0477362853008805e-05, + "loss": 0.7392, + "step": 12114 + }, + { + "epoch": 0.5364001638201078, + "grad_norm": 1.8112332820892334, + "learning_rate": 1.0475819158237426e-05, + "loss": 0.7397, + "step": 12115 + }, + { + "epoch": 0.5364444395249217, + "grad_norm": 1.6194044351577759, + "learning_rate": 1.047427545210149e-05, + "loss": 0.6976, + "step": 12116 + }, + { + "epoch": 0.5364887152297355, + "grad_norm": 1.2569571733474731, + "learning_rate": 1.047273173463787e-05, + "loss": 0.615, + "step": 12117 + }, + { + "epoch": 0.5365329909345494, + "grad_norm": 1.3281068801879883, + "learning_rate": 1.047118800588344e-05, + "loss": 0.6557, + "step": 12118 + }, + { + "epoch": 0.5365772666393633, + "grad_norm": 1.2068705558776855, + "learning_rate": 1.0469644265875068e-05, + "loss": 0.6058, + "step": 12119 + }, + { + "epoch": 0.5366215423441771, + "grad_norm": 1.4072571992874146, + "learning_rate": 1.0468100514649627e-05, + "loss": 0.6051, + "step": 12120 + }, + { + "epoch": 0.536665818048991, + "grad_norm": 1.2370634078979492, + "learning_rate": 1.0466556752243984e-05, + "loss": 0.6135, + "step": 12121 + }, + { + "epoch": 0.536710093753805, + "grad_norm": 1.4238719940185547, + "learning_rate": 1.0465012978695013e-05, + "loss": 0.6347, + "step": 12122 + }, + { + "epoch": 0.5367543694586189, + "grad_norm": 1.542504072189331, + "learning_rate": 1.0463469194039584e-05, + "loss": 0.6834, + "step": 12123 + }, + { + "epoch": 0.5367986451634327, + "grad_norm": 1.5288550853729248, + "learning_rate": 1.0461925398314573e-05, + "loss": 0.6384, + "step": 12124 + }, + { + "epoch": 0.5368429208682466, + "grad_norm": 1.6488670110702515, + "learning_rate": 1.046038159155685e-05, + "loss": 0.6973, + "step": 12125 + }, + { + "epoch": 0.5368871965730605, + "grad_norm": 1.2988996505737305, + "learning_rate": 1.0458837773803288e-05, + "loss": 0.6222, + "step": 12126 + }, + { + "epoch": 0.5369314722778743, + "grad_norm": 1.4473105669021606, + "learning_rate": 1.0457293945090756e-05, + "loss": 0.6282, + "step": 12127 + }, + { + "epoch": 0.5369757479826882, + "grad_norm": 1.5608320236206055, + "learning_rate": 1.0455750105456131e-05, + "loss": 0.6254, + "step": 12128 + }, + { + "epoch": 0.5370200236875021, + "grad_norm": 1.3461607694625854, + "learning_rate": 1.0454206254936287e-05, + "loss": 0.6468, + "step": 12129 + }, + { + "epoch": 0.5370642993923159, + "grad_norm": 1.3952348232269287, + "learning_rate": 1.0452662393568098e-05, + "loss": 0.6457, + "step": 12130 + }, + { + "epoch": 0.5371085750971298, + "grad_norm": 2.304730176925659, + "learning_rate": 1.0451118521388437e-05, + "loss": 0.7905, + "step": 12131 + }, + { + "epoch": 0.5371528508019437, + "grad_norm": 1.5102802515029907, + "learning_rate": 1.0449574638434175e-05, + "loss": 0.608, + "step": 12132 + }, + { + "epoch": 0.5371971265067575, + "grad_norm": 1.4479360580444336, + "learning_rate": 1.0448030744742188e-05, + "loss": 0.6743, + "step": 12133 + }, + { + "epoch": 0.5372414022115715, + "grad_norm": 1.2573050260543823, + "learning_rate": 1.0446486840349355e-05, + "loss": 0.6041, + "step": 12134 + }, + { + "epoch": 0.5372856779163854, + "grad_norm": 1.4909144639968872, + "learning_rate": 1.0444942925292544e-05, + "loss": 0.7108, + "step": 12135 + }, + { + "epoch": 0.5373299536211992, + "grad_norm": 1.3380472660064697, + "learning_rate": 1.0443398999608636e-05, + "loss": 0.5651, + "step": 12136 + }, + { + "epoch": 0.5373742293260131, + "grad_norm": 1.4428424835205078, + "learning_rate": 1.0441855063334501e-05, + "loss": 0.6586, + "step": 12137 + }, + { + "epoch": 0.537418505030827, + "grad_norm": 1.2631901502609253, + "learning_rate": 1.0440311116507021e-05, + "loss": 0.6399, + "step": 12138 + }, + { + "epoch": 0.5374627807356408, + "grad_norm": 1.4598451852798462, + "learning_rate": 1.0438767159163068e-05, + "loss": 0.6549, + "step": 12139 + }, + { + "epoch": 0.5375070564404547, + "grad_norm": 1.5799988508224487, + "learning_rate": 1.0437223191339517e-05, + "loss": 0.6959, + "step": 12140 + }, + { + "epoch": 0.5375513321452686, + "grad_norm": 1.3597006797790527, + "learning_rate": 1.0435679213073247e-05, + "loss": 0.6204, + "step": 12141 + }, + { + "epoch": 0.5375956078500824, + "grad_norm": 1.4468939304351807, + "learning_rate": 1.0434135224401132e-05, + "loss": 0.6683, + "step": 12142 + }, + { + "epoch": 0.5376398835548963, + "grad_norm": 1.8083703517913818, + "learning_rate": 1.0432591225360052e-05, + "loss": 0.7369, + "step": 12143 + }, + { + "epoch": 0.5376841592597102, + "grad_norm": 1.440837025642395, + "learning_rate": 1.0431047215986882e-05, + "loss": 0.6584, + "step": 12144 + }, + { + "epoch": 0.537728434964524, + "grad_norm": 1.6730111837387085, + "learning_rate": 1.04295031963185e-05, + "loss": 0.6789, + "step": 12145 + }, + { + "epoch": 0.537772710669338, + "grad_norm": 1.2880998849868774, + "learning_rate": 1.0427959166391786e-05, + "loss": 0.6645, + "step": 12146 + }, + { + "epoch": 0.5378169863741519, + "grad_norm": 1.387115478515625, + "learning_rate": 1.0426415126243615e-05, + "loss": 0.7153, + "step": 12147 + }, + { + "epoch": 0.5378612620789657, + "grad_norm": 2.6469767093658447, + "learning_rate": 1.0424871075910865e-05, + "loss": 0.7447, + "step": 12148 + }, + { + "epoch": 0.5379055377837796, + "grad_norm": 1.5306065082550049, + "learning_rate": 1.0423327015430417e-05, + "loss": 0.6827, + "step": 12149 + }, + { + "epoch": 0.5379498134885935, + "grad_norm": 1.4874951839447021, + "learning_rate": 1.0421782944839145e-05, + "loss": 0.691, + "step": 12150 + }, + { + "epoch": 0.5379940891934073, + "grad_norm": 1.4097319841384888, + "learning_rate": 1.0420238864173932e-05, + "loss": 0.6008, + "step": 12151 + }, + { + "epoch": 0.5380383648982212, + "grad_norm": 1.4299849271774292, + "learning_rate": 1.0418694773471655e-05, + "loss": 0.6831, + "step": 12152 + }, + { + "epoch": 0.5380826406030351, + "grad_norm": 1.4559818506240845, + "learning_rate": 1.04171506727692e-05, + "loss": 0.5698, + "step": 12153 + }, + { + "epoch": 0.5381269163078489, + "grad_norm": 1.5825556516647339, + "learning_rate": 1.0415606562103436e-05, + "loss": 0.6811, + "step": 12154 + }, + { + "epoch": 0.5381711920126628, + "grad_norm": 1.3272428512573242, + "learning_rate": 1.0414062441511245e-05, + "loss": 0.7242, + "step": 12155 + }, + { + "epoch": 0.5382154677174767, + "grad_norm": 1.2678401470184326, + "learning_rate": 1.0412518311029516e-05, + "loss": 0.5968, + "step": 12156 + }, + { + "epoch": 0.5382597434222905, + "grad_norm": 1.7578775882720947, + "learning_rate": 1.0410974170695118e-05, + "loss": 0.6655, + "step": 12157 + }, + { + "epoch": 0.5383040191271045, + "grad_norm": 1.3304462432861328, + "learning_rate": 1.0409430020544941e-05, + "loss": 0.6057, + "step": 12158 + }, + { + "epoch": 0.5383482948319184, + "grad_norm": 1.2640763521194458, + "learning_rate": 1.0407885860615859e-05, + "loss": 0.5933, + "step": 12159 + }, + { + "epoch": 0.5383925705367323, + "grad_norm": 1.6246830224990845, + "learning_rate": 1.0406341690944753e-05, + "loss": 0.6937, + "step": 12160 + }, + { + "epoch": 0.5384368462415461, + "grad_norm": 1.3038078546524048, + "learning_rate": 1.0404797511568509e-05, + "loss": 0.5554, + "step": 12161 + }, + { + "epoch": 0.53848112194636, + "grad_norm": 1.5884100198745728, + "learning_rate": 1.0403253322524007e-05, + "loss": 0.6601, + "step": 12162 + }, + { + "epoch": 0.5385253976511739, + "grad_norm": 1.643009901046753, + "learning_rate": 1.0401709123848127e-05, + "loss": 0.7472, + "step": 12163 + }, + { + "epoch": 0.5385696733559877, + "grad_norm": 1.3349995613098145, + "learning_rate": 1.0400164915577751e-05, + "loss": 0.6386, + "step": 12164 + }, + { + "epoch": 0.5386139490608016, + "grad_norm": 1.4319260120391846, + "learning_rate": 1.0398620697749759e-05, + "loss": 0.6165, + "step": 12165 + }, + { + "epoch": 0.5386582247656155, + "grad_norm": 1.3521530628204346, + "learning_rate": 1.039707647040104e-05, + "loss": 0.619, + "step": 12166 + }, + { + "epoch": 0.5387025004704293, + "grad_norm": 1.822536587715149, + "learning_rate": 1.0395532233568473e-05, + "loss": 0.8193, + "step": 12167 + }, + { + "epoch": 0.5387467761752432, + "grad_norm": 1.2847377061843872, + "learning_rate": 1.039398798728894e-05, + "loss": 0.6024, + "step": 12168 + }, + { + "epoch": 0.5387910518800572, + "grad_norm": 1.3715426921844482, + "learning_rate": 1.0392443731599324e-05, + "loss": 0.5843, + "step": 12169 + }, + { + "epoch": 0.538835327584871, + "grad_norm": 1.5718156099319458, + "learning_rate": 1.0390899466536508e-05, + "loss": 0.6433, + "step": 12170 + }, + { + "epoch": 0.5388796032896849, + "grad_norm": 1.3555572032928467, + "learning_rate": 1.0389355192137379e-05, + "loss": 0.6421, + "step": 12171 + }, + { + "epoch": 0.5389238789944988, + "grad_norm": 1.5881699323654175, + "learning_rate": 1.0387810908438817e-05, + "loss": 0.6171, + "step": 12172 + }, + { + "epoch": 0.5389681546993126, + "grad_norm": 1.5796325206756592, + "learning_rate": 1.0386266615477708e-05, + "loss": 0.6158, + "step": 12173 + }, + { + "epoch": 0.5390124304041265, + "grad_norm": 1.5446451902389526, + "learning_rate": 1.0384722313290934e-05, + "loss": 0.5783, + "step": 12174 + }, + { + "epoch": 0.5390567061089404, + "grad_norm": 2.043137311935425, + "learning_rate": 1.038317800191538e-05, + "loss": 0.6687, + "step": 12175 + }, + { + "epoch": 0.5391009818137542, + "grad_norm": 1.3160879611968994, + "learning_rate": 1.0381633681387936e-05, + "loss": 0.6061, + "step": 12176 + }, + { + "epoch": 0.5391452575185681, + "grad_norm": 1.2608741521835327, + "learning_rate": 1.0380089351745477e-05, + "loss": 0.5351, + "step": 12177 + }, + { + "epoch": 0.539189533223382, + "grad_norm": 1.3041770458221436, + "learning_rate": 1.0378545013024897e-05, + "loss": 0.6038, + "step": 12178 + }, + { + "epoch": 0.5392338089281958, + "grad_norm": 1.4618151187896729, + "learning_rate": 1.0377000665263076e-05, + "loss": 0.6456, + "step": 12179 + }, + { + "epoch": 0.5392780846330097, + "grad_norm": 1.6311955451965332, + "learning_rate": 1.0375456308496898e-05, + "loss": 0.6959, + "step": 12180 + }, + { + "epoch": 0.5393223603378237, + "grad_norm": 1.449647307395935, + "learning_rate": 1.037391194276326e-05, + "loss": 0.6831, + "step": 12181 + }, + { + "epoch": 0.5393666360426375, + "grad_norm": 1.4602147340774536, + "learning_rate": 1.0372367568099033e-05, + "loss": 0.6728, + "step": 12182 + }, + { + "epoch": 0.5394109117474514, + "grad_norm": 1.3450500965118408, + "learning_rate": 1.0370823184541112e-05, + "loss": 0.661, + "step": 12183 + }, + { + "epoch": 0.5394551874522653, + "grad_norm": 2.018796443939209, + "learning_rate": 1.0369278792126378e-05, + "loss": 0.6927, + "step": 12184 + }, + { + "epoch": 0.5394994631570791, + "grad_norm": 1.5197850465774536, + "learning_rate": 1.0367734390891726e-05, + "loss": 0.7065, + "step": 12185 + }, + { + "epoch": 0.539543738861893, + "grad_norm": 1.343627691268921, + "learning_rate": 1.0366189980874035e-05, + "loss": 0.5738, + "step": 12186 + }, + { + "epoch": 0.5395880145667069, + "grad_norm": 1.3446691036224365, + "learning_rate": 1.0364645562110194e-05, + "loss": 0.6007, + "step": 12187 + }, + { + "epoch": 0.5396322902715207, + "grad_norm": 1.4561954736709595, + "learning_rate": 1.036310113463709e-05, + "loss": 0.7332, + "step": 12188 + }, + { + "epoch": 0.5396765659763346, + "grad_norm": 1.434659719467163, + "learning_rate": 1.0361556698491612e-05, + "loss": 0.6135, + "step": 12189 + }, + { + "epoch": 0.5397208416811485, + "grad_norm": 1.6207371950149536, + "learning_rate": 1.0360012253710647e-05, + "loss": 0.6675, + "step": 12190 + }, + { + "epoch": 0.5397651173859623, + "grad_norm": 1.4014620780944824, + "learning_rate": 1.0358467800331085e-05, + "loss": 0.6533, + "step": 12191 + }, + { + "epoch": 0.5398093930907762, + "grad_norm": 2.683605670928955, + "learning_rate": 1.0356923338389807e-05, + "loss": 0.7227, + "step": 12192 + }, + { + "epoch": 0.5398536687955902, + "grad_norm": 1.5275864601135254, + "learning_rate": 1.0355378867923706e-05, + "loss": 0.6109, + "step": 12193 + }, + { + "epoch": 0.5398979445004041, + "grad_norm": 1.7660645246505737, + "learning_rate": 1.035383438896967e-05, + "loss": 0.7324, + "step": 12194 + }, + { + "epoch": 0.5399422202052179, + "grad_norm": 1.4027019739151, + "learning_rate": 1.0352289901564592e-05, + "loss": 0.6066, + "step": 12195 + }, + { + "epoch": 0.5399864959100318, + "grad_norm": 1.547992467880249, + "learning_rate": 1.0350745405745352e-05, + "loss": 0.6776, + "step": 12196 + }, + { + "epoch": 0.5400307716148457, + "grad_norm": 1.497904896736145, + "learning_rate": 1.0349200901548846e-05, + "loss": 0.6718, + "step": 12197 + }, + { + "epoch": 0.5400750473196595, + "grad_norm": 1.3674520254135132, + "learning_rate": 1.034765638901196e-05, + "loss": 0.599, + "step": 12198 + }, + { + "epoch": 0.5401193230244734, + "grad_norm": 1.4242920875549316, + "learning_rate": 1.0346111868171584e-05, + "loss": 0.5989, + "step": 12199 + }, + { + "epoch": 0.5401635987292873, + "grad_norm": 1.6298811435699463, + "learning_rate": 1.0344567339064608e-05, + "loss": 0.7276, + "step": 12200 + }, + { + "epoch": 0.5402078744341011, + "grad_norm": 1.8792893886566162, + "learning_rate": 1.0343022801727922e-05, + "loss": 0.7075, + "step": 12201 + }, + { + "epoch": 0.540252150138915, + "grad_norm": 1.8870272636413574, + "learning_rate": 1.0341478256198416e-05, + "loss": 0.7045, + "step": 12202 + }, + { + "epoch": 0.540296425843729, + "grad_norm": 1.2910103797912598, + "learning_rate": 1.0339933702512978e-05, + "loss": 0.6088, + "step": 12203 + }, + { + "epoch": 0.5403407015485427, + "grad_norm": 1.5899285078048706, + "learning_rate": 1.0338389140708502e-05, + "loss": 0.6528, + "step": 12204 + }, + { + "epoch": 0.5403849772533567, + "grad_norm": 1.5452193021774292, + "learning_rate": 1.0336844570821877e-05, + "loss": 0.6466, + "step": 12205 + }, + { + "epoch": 0.5404292529581706, + "grad_norm": 1.606076955795288, + "learning_rate": 1.0335299992889993e-05, + "loss": 0.6052, + "step": 12206 + }, + { + "epoch": 0.5404735286629844, + "grad_norm": 1.4177968502044678, + "learning_rate": 1.0333755406949742e-05, + "loss": 0.6201, + "step": 12207 + }, + { + "epoch": 0.5405178043677983, + "grad_norm": 1.3755183219909668, + "learning_rate": 1.0332210813038015e-05, + "loss": 0.6236, + "step": 12208 + }, + { + "epoch": 0.5405620800726122, + "grad_norm": 1.2491859197616577, + "learning_rate": 1.0330666211191705e-05, + "loss": 0.5867, + "step": 12209 + }, + { + "epoch": 0.540606355777426, + "grad_norm": 1.5268696546554565, + "learning_rate": 1.0329121601447697e-05, + "loss": 0.6134, + "step": 12210 + }, + { + "epoch": 0.5406506314822399, + "grad_norm": 1.3333650827407837, + "learning_rate": 1.0327576983842894e-05, + "loss": 0.6293, + "step": 12211 + }, + { + "epoch": 0.5406949071870538, + "grad_norm": 1.4379500150680542, + "learning_rate": 1.0326032358414175e-05, + "loss": 0.625, + "step": 12212 + }, + { + "epoch": 0.5407391828918676, + "grad_norm": 1.2361433506011963, + "learning_rate": 1.0324487725198442e-05, + "loss": 0.6653, + "step": 12213 + }, + { + "epoch": 0.5407834585966815, + "grad_norm": 1.7376219034194946, + "learning_rate": 1.0322943084232583e-05, + "loss": 0.633, + "step": 12214 + }, + { + "epoch": 0.5408277343014954, + "grad_norm": 1.5511773824691772, + "learning_rate": 1.0321398435553491e-05, + "loss": 0.6276, + "step": 12215 + }, + { + "epoch": 0.5408720100063092, + "grad_norm": 1.340893030166626, + "learning_rate": 1.0319853779198062e-05, + "loss": 0.5724, + "step": 12216 + }, + { + "epoch": 0.5409162857111232, + "grad_norm": 1.5747405290603638, + "learning_rate": 1.031830911520318e-05, + "loss": 0.6789, + "step": 12217 + }, + { + "epoch": 0.5409605614159371, + "grad_norm": 1.3648310899734497, + "learning_rate": 1.031676444360575e-05, + "loss": 0.6076, + "step": 12218 + }, + { + "epoch": 0.5410048371207509, + "grad_norm": 1.668897271156311, + "learning_rate": 1.0315219764442657e-05, + "loss": 0.6467, + "step": 12219 + }, + { + "epoch": 0.5410491128255648, + "grad_norm": 1.451593279838562, + "learning_rate": 1.0313675077750795e-05, + "loss": 0.6785, + "step": 12220 + }, + { + "epoch": 0.5410933885303787, + "grad_norm": 1.4511317014694214, + "learning_rate": 1.031213038356706e-05, + "loss": 0.5933, + "step": 12221 + }, + { + "epoch": 0.5411376642351925, + "grad_norm": 1.3872401714324951, + "learning_rate": 1.0310585681928346e-05, + "loss": 0.664, + "step": 12222 + }, + { + "epoch": 0.5411819399400064, + "grad_norm": 1.4888523817062378, + "learning_rate": 1.0309040972871547e-05, + "loss": 0.6504, + "step": 12223 + }, + { + "epoch": 0.5412262156448203, + "grad_norm": 1.5486315488815308, + "learning_rate": 1.0307496256433553e-05, + "loss": 0.6262, + "step": 12224 + }, + { + "epoch": 0.5412704913496341, + "grad_norm": 1.376081943511963, + "learning_rate": 1.0305951532651262e-05, + "loss": 0.6747, + "step": 12225 + }, + { + "epoch": 0.541314767054448, + "grad_norm": 1.814189076423645, + "learning_rate": 1.0304406801561571e-05, + "loss": 0.7143, + "step": 12226 + }, + { + "epoch": 0.5413590427592619, + "grad_norm": 1.5110467672348022, + "learning_rate": 1.0302862063201367e-05, + "loss": 0.5751, + "step": 12227 + }, + { + "epoch": 0.5414033184640757, + "grad_norm": 1.3275262117385864, + "learning_rate": 1.030131731760755e-05, + "loss": 0.633, + "step": 12228 + }, + { + "epoch": 0.5414475941688897, + "grad_norm": 1.4901607036590576, + "learning_rate": 1.0299772564817017e-05, + "loss": 0.608, + "step": 12229 + }, + { + "epoch": 0.5414918698737036, + "grad_norm": 1.2665998935699463, + "learning_rate": 1.0298227804866659e-05, + "loss": 0.578, + "step": 12230 + }, + { + "epoch": 0.5415361455785175, + "grad_norm": 1.3873761892318726, + "learning_rate": 1.0296683037793371e-05, + "loss": 0.5756, + "step": 12231 + }, + { + "epoch": 0.5415804212833313, + "grad_norm": 1.4443074464797974, + "learning_rate": 1.0295138263634051e-05, + "loss": 0.5908, + "step": 12232 + }, + { + "epoch": 0.5416246969881452, + "grad_norm": 1.1918988227844238, + "learning_rate": 1.0293593482425598e-05, + "loss": 0.6032, + "step": 12233 + }, + { + "epoch": 0.5416689726929591, + "grad_norm": 1.747511386871338, + "learning_rate": 1.0292048694204897e-05, + "loss": 0.6513, + "step": 12234 + }, + { + "epoch": 0.5417132483977729, + "grad_norm": 1.5362039804458618, + "learning_rate": 1.0290503899008853e-05, + "loss": 0.7007, + "step": 12235 + }, + { + "epoch": 0.5417575241025868, + "grad_norm": 1.548987865447998, + "learning_rate": 1.0288959096874362e-05, + "loss": 0.6401, + "step": 12236 + }, + { + "epoch": 0.5418017998074007, + "grad_norm": 1.2895324230194092, + "learning_rate": 1.0287414287838318e-05, + "loss": 0.6192, + "step": 12237 + }, + { + "epoch": 0.5418460755122145, + "grad_norm": 2.062448024749756, + "learning_rate": 1.0285869471937618e-05, + "loss": 0.7205, + "step": 12238 + }, + { + "epoch": 0.5418903512170284, + "grad_norm": 1.3411821126937866, + "learning_rate": 1.0284324649209156e-05, + "loss": 0.6255, + "step": 12239 + }, + { + "epoch": 0.5419346269218424, + "grad_norm": 1.664275050163269, + "learning_rate": 1.028277981968983e-05, + "loss": 0.6612, + "step": 12240 + }, + { + "epoch": 0.5419789026266562, + "grad_norm": 1.4730252027511597, + "learning_rate": 1.0281234983416541e-05, + "loss": 0.5905, + "step": 12241 + }, + { + "epoch": 0.5420231783314701, + "grad_norm": 1.6512304544448853, + "learning_rate": 1.0279690140426184e-05, + "loss": 0.6709, + "step": 12242 + }, + { + "epoch": 0.542067454036284, + "grad_norm": 1.3819917440414429, + "learning_rate": 1.0278145290755657e-05, + "loss": 0.5909, + "step": 12243 + }, + { + "epoch": 0.5421117297410978, + "grad_norm": 1.566502571105957, + "learning_rate": 1.0276600434441853e-05, + "loss": 0.6537, + "step": 12244 + }, + { + "epoch": 0.5421560054459117, + "grad_norm": 1.3249473571777344, + "learning_rate": 1.0275055571521672e-05, + "loss": 0.6212, + "step": 12245 + }, + { + "epoch": 0.5422002811507256, + "grad_norm": 1.2014142274856567, + "learning_rate": 1.0273510702032015e-05, + "loss": 0.6076, + "step": 12246 + }, + { + "epoch": 0.5422445568555394, + "grad_norm": 1.554898738861084, + "learning_rate": 1.0271965826009778e-05, + "loss": 0.6057, + "step": 12247 + }, + { + "epoch": 0.5422888325603533, + "grad_norm": 1.4256668090820312, + "learning_rate": 1.0270420943491859e-05, + "loss": 0.6112, + "step": 12248 + }, + { + "epoch": 0.5423331082651672, + "grad_norm": 1.173440933227539, + "learning_rate": 1.0268876054515153e-05, + "loss": 0.5593, + "step": 12249 + }, + { + "epoch": 0.542377383969981, + "grad_norm": 1.3393714427947998, + "learning_rate": 1.0267331159116563e-05, + "loss": 0.6303, + "step": 12250 + }, + { + "epoch": 0.5424216596747949, + "grad_norm": 1.509735345840454, + "learning_rate": 1.0265786257332986e-05, + "loss": 0.6464, + "step": 12251 + }, + { + "epoch": 0.5424659353796089, + "grad_norm": 1.4837285280227661, + "learning_rate": 1.0264241349201323e-05, + "loss": 0.6587, + "step": 12252 + }, + { + "epoch": 0.5425102110844227, + "grad_norm": 2.0366768836975098, + "learning_rate": 1.0262696434758468e-05, + "loss": 0.6433, + "step": 12253 + }, + { + "epoch": 0.5425544867892366, + "grad_norm": 1.8520727157592773, + "learning_rate": 1.0261151514041321e-05, + "loss": 0.6747, + "step": 12254 + }, + { + "epoch": 0.5425987624940505, + "grad_norm": 1.4818469285964966, + "learning_rate": 1.0259606587086783e-05, + "loss": 0.5754, + "step": 12255 + }, + { + "epoch": 0.5426430381988643, + "grad_norm": 1.9224284887313843, + "learning_rate": 1.0258061653931757e-05, + "loss": 0.7233, + "step": 12256 + }, + { + "epoch": 0.5426873139036782, + "grad_norm": 1.5214344263076782, + "learning_rate": 1.0256516714613137e-05, + "loss": 0.6925, + "step": 12257 + }, + { + "epoch": 0.5427315896084921, + "grad_norm": 1.298893928527832, + "learning_rate": 1.0254971769167822e-05, + "loss": 0.5958, + "step": 12258 + }, + { + "epoch": 0.5427758653133059, + "grad_norm": 1.3399757146835327, + "learning_rate": 1.0253426817632715e-05, + "loss": 0.6471, + "step": 12259 + }, + { + "epoch": 0.5428201410181198, + "grad_norm": 1.2553974390029907, + "learning_rate": 1.0251881860044714e-05, + "loss": 0.5844, + "step": 12260 + }, + { + "epoch": 0.5428644167229337, + "grad_norm": 1.7847660779953003, + "learning_rate": 1.0250336896440722e-05, + "loss": 0.6354, + "step": 12261 + }, + { + "epoch": 0.5429086924277475, + "grad_norm": 1.374800205230713, + "learning_rate": 1.0248791926857635e-05, + "loss": 0.6047, + "step": 12262 + }, + { + "epoch": 0.5429529681325614, + "grad_norm": 1.3848185539245605, + "learning_rate": 1.0247246951332358e-05, + "loss": 0.6138, + "step": 12263 + }, + { + "epoch": 0.5429972438373754, + "grad_norm": 1.378342866897583, + "learning_rate": 1.0245701969901785e-05, + "loss": 0.5988, + "step": 12264 + }, + { + "epoch": 0.5430415195421893, + "grad_norm": 1.500969409942627, + "learning_rate": 1.0244156982602824e-05, + "loss": 0.6349, + "step": 12265 + }, + { + "epoch": 0.5430857952470031, + "grad_norm": 1.5407607555389404, + "learning_rate": 1.024261198947237e-05, + "loss": 0.6067, + "step": 12266 + }, + { + "epoch": 0.543130070951817, + "grad_norm": 1.3441396951675415, + "learning_rate": 1.0241066990547328e-05, + "loss": 0.6147, + "step": 12267 + }, + { + "epoch": 0.5431743466566309, + "grad_norm": 1.8570032119750977, + "learning_rate": 1.0239521985864595e-05, + "loss": 0.667, + "step": 12268 + }, + { + "epoch": 0.5432186223614447, + "grad_norm": 1.4073007106781006, + "learning_rate": 1.0237976975461074e-05, + "loss": 0.6095, + "step": 12269 + }, + { + "epoch": 0.5432628980662586, + "grad_norm": 1.654547095298767, + "learning_rate": 1.0236431959373672e-05, + "loss": 0.6147, + "step": 12270 + }, + { + "epoch": 0.5433071737710725, + "grad_norm": 1.4257192611694336, + "learning_rate": 1.0234886937639282e-05, + "loss": 0.6879, + "step": 12271 + }, + { + "epoch": 0.5433514494758863, + "grad_norm": 1.421154260635376, + "learning_rate": 1.0233341910294807e-05, + "loss": 0.617, + "step": 12272 + }, + { + "epoch": 0.5433957251807002, + "grad_norm": 1.4327428340911865, + "learning_rate": 1.0231796877377152e-05, + "loss": 0.5793, + "step": 12273 + }, + { + "epoch": 0.5434400008855141, + "grad_norm": 1.6863164901733398, + "learning_rate": 1.0230251838923215e-05, + "loss": 0.6475, + "step": 12274 + }, + { + "epoch": 0.5434842765903279, + "grad_norm": 1.333282709121704, + "learning_rate": 1.0228706794969905e-05, + "loss": 0.6039, + "step": 12275 + }, + { + "epoch": 0.5435285522951419, + "grad_norm": 1.602460503578186, + "learning_rate": 1.0227161745554117e-05, + "loss": 0.6753, + "step": 12276 + }, + { + "epoch": 0.5435728279999558, + "grad_norm": 1.5843873023986816, + "learning_rate": 1.0225616690712754e-05, + "loss": 0.6624, + "step": 12277 + }, + { + "epoch": 0.5436171037047696, + "grad_norm": 1.5685911178588867, + "learning_rate": 1.0224071630482721e-05, + "loss": 0.6565, + "step": 12278 + }, + { + "epoch": 0.5436613794095835, + "grad_norm": 1.6743192672729492, + "learning_rate": 1.0222526564900921e-05, + "loss": 0.6516, + "step": 12279 + }, + { + "epoch": 0.5437056551143974, + "grad_norm": 1.4355511665344238, + "learning_rate": 1.0220981494004256e-05, + "loss": 0.6098, + "step": 12280 + }, + { + "epoch": 0.5437499308192112, + "grad_norm": 1.3829749822616577, + "learning_rate": 1.0219436417829624e-05, + "loss": 0.5688, + "step": 12281 + }, + { + "epoch": 0.5437942065240251, + "grad_norm": 1.462625503540039, + "learning_rate": 1.0217891336413934e-05, + "loss": 0.6344, + "step": 12282 + }, + { + "epoch": 0.543838482228839, + "grad_norm": 2.2273917198181152, + "learning_rate": 1.0216346249794087e-05, + "loss": 0.6916, + "step": 12283 + }, + { + "epoch": 0.5438827579336528, + "grad_norm": 1.3156962394714355, + "learning_rate": 1.0214801158006983e-05, + "loss": 0.5729, + "step": 12284 + }, + { + "epoch": 0.5439270336384667, + "grad_norm": 1.6500033140182495, + "learning_rate": 1.0213256061089536e-05, + "loss": 0.7189, + "step": 12285 + }, + { + "epoch": 0.5439713093432806, + "grad_norm": 1.7192072868347168, + "learning_rate": 1.0211710959078636e-05, + "loss": 0.6562, + "step": 12286 + }, + { + "epoch": 0.5440155850480944, + "grad_norm": 1.623430848121643, + "learning_rate": 1.021016585201119e-05, + "loss": 0.6554, + "step": 12287 + }, + { + "epoch": 0.5440598607529084, + "grad_norm": 1.4011316299438477, + "learning_rate": 1.0208620739924109e-05, + "loss": 0.6217, + "step": 12288 + }, + { + "epoch": 0.5441041364577223, + "grad_norm": 1.3037832975387573, + "learning_rate": 1.020707562285429e-05, + "loss": 0.6182, + "step": 12289 + }, + { + "epoch": 0.5441484121625361, + "grad_norm": 1.4867607355117798, + "learning_rate": 1.0205530500838638e-05, + "loss": 0.6791, + "step": 12290 + }, + { + "epoch": 0.54419268786735, + "grad_norm": 1.390525221824646, + "learning_rate": 1.0203985373914056e-05, + "loss": 0.6333, + "step": 12291 + }, + { + "epoch": 0.5442369635721639, + "grad_norm": 1.4780892133712769, + "learning_rate": 1.0202440242117448e-05, + "loss": 0.6442, + "step": 12292 + }, + { + "epoch": 0.5442812392769777, + "grad_norm": 1.883169174194336, + "learning_rate": 1.0200895105485721e-05, + "loss": 0.7329, + "step": 12293 + }, + { + "epoch": 0.5443255149817916, + "grad_norm": 1.325120210647583, + "learning_rate": 1.0199349964055782e-05, + "loss": 0.6297, + "step": 12294 + }, + { + "epoch": 0.5443697906866055, + "grad_norm": 2.1630518436431885, + "learning_rate": 1.019780481786453e-05, + "loss": 0.6942, + "step": 12295 + }, + { + "epoch": 0.5444140663914193, + "grad_norm": 1.3820220232009888, + "learning_rate": 1.0196259666948866e-05, + "loss": 0.5556, + "step": 12296 + }, + { + "epoch": 0.5444583420962332, + "grad_norm": 1.6289417743682861, + "learning_rate": 1.0194714511345702e-05, + "loss": 0.6139, + "step": 12297 + }, + { + "epoch": 0.5445026178010471, + "grad_norm": 1.4614344835281372, + "learning_rate": 1.0193169351091942e-05, + "loss": 0.6528, + "step": 12298 + }, + { + "epoch": 0.544546893505861, + "grad_norm": 1.5007750988006592, + "learning_rate": 1.019162418622449e-05, + "loss": 0.6389, + "step": 12299 + }, + { + "epoch": 0.5445911692106749, + "grad_norm": 1.5229147672653198, + "learning_rate": 1.019007901678025e-05, + "loss": 0.6521, + "step": 12300 + }, + { + "epoch": 0.5446354449154888, + "grad_norm": 1.9239418506622314, + "learning_rate": 1.0188533842796125e-05, + "loss": 0.7463, + "step": 12301 + }, + { + "epoch": 0.5446797206203027, + "grad_norm": 1.658151626586914, + "learning_rate": 1.0186988664309023e-05, + "loss": 0.695, + "step": 12302 + }, + { + "epoch": 0.5447239963251165, + "grad_norm": 1.7787811756134033, + "learning_rate": 1.018544348135585e-05, + "loss": 0.6592, + "step": 12303 + }, + { + "epoch": 0.5447682720299304, + "grad_norm": 1.4966180324554443, + "learning_rate": 1.0183898293973513e-05, + "loss": 0.5885, + "step": 12304 + }, + { + "epoch": 0.5448125477347443, + "grad_norm": 1.4346539974212646, + "learning_rate": 1.0182353102198912e-05, + "loss": 0.6307, + "step": 12305 + }, + { + "epoch": 0.5448568234395581, + "grad_norm": 1.328056812286377, + "learning_rate": 1.0180807906068955e-05, + "loss": 0.6018, + "step": 12306 + }, + { + "epoch": 0.544901099144372, + "grad_norm": 1.3321142196655273, + "learning_rate": 1.017926270562055e-05, + "loss": 0.5871, + "step": 12307 + }, + { + "epoch": 0.5449453748491859, + "grad_norm": 1.657820463180542, + "learning_rate": 1.0177717500890603e-05, + "loss": 0.6572, + "step": 12308 + }, + { + "epoch": 0.5449896505539997, + "grad_norm": 1.375063419342041, + "learning_rate": 1.0176172291916013e-05, + "loss": 0.6284, + "step": 12309 + }, + { + "epoch": 0.5450339262588136, + "grad_norm": 1.4016032218933105, + "learning_rate": 1.0174627078733696e-05, + "loss": 0.6261, + "step": 12310 + }, + { + "epoch": 0.5450782019636276, + "grad_norm": 1.54049813747406, + "learning_rate": 1.0173081861380551e-05, + "loss": 0.6621, + "step": 12311 + }, + { + "epoch": 0.5451224776684414, + "grad_norm": 1.6638399362564087, + "learning_rate": 1.0171536639893487e-05, + "loss": 0.6913, + "step": 12312 + }, + { + "epoch": 0.5451667533732553, + "grad_norm": 1.3357274532318115, + "learning_rate": 1.0169991414309413e-05, + "loss": 0.6198, + "step": 12313 + }, + { + "epoch": 0.5452110290780692, + "grad_norm": 1.297810435295105, + "learning_rate": 1.0168446184665229e-05, + "loss": 0.5617, + "step": 12314 + }, + { + "epoch": 0.545255304782883, + "grad_norm": 1.622435450553894, + "learning_rate": 1.0166900950997845e-05, + "loss": 0.6596, + "step": 12315 + }, + { + "epoch": 0.5452995804876969, + "grad_norm": 1.531810998916626, + "learning_rate": 1.0165355713344168e-05, + "loss": 0.6767, + "step": 12316 + }, + { + "epoch": 0.5453438561925108, + "grad_norm": 1.5658458471298218, + "learning_rate": 1.0163810471741105e-05, + "loss": 0.6297, + "step": 12317 + }, + { + "epoch": 0.5453881318973246, + "grad_norm": 1.556850552558899, + "learning_rate": 1.0162265226225563e-05, + "loss": 0.7213, + "step": 12318 + }, + { + "epoch": 0.5454324076021385, + "grad_norm": 1.244708776473999, + "learning_rate": 1.0160719976834447e-05, + "loss": 0.6035, + "step": 12319 + }, + { + "epoch": 0.5454766833069524, + "grad_norm": 1.7170921564102173, + "learning_rate": 1.0159174723604666e-05, + "loss": 0.7343, + "step": 12320 + }, + { + "epoch": 0.5455209590117662, + "grad_norm": 1.357033610343933, + "learning_rate": 1.0157629466573125e-05, + "loss": 0.6533, + "step": 12321 + }, + { + "epoch": 0.5455652347165801, + "grad_norm": 1.4419729709625244, + "learning_rate": 1.0156084205776735e-05, + "loss": 0.6081, + "step": 12322 + }, + { + "epoch": 0.545609510421394, + "grad_norm": 1.4642380475997925, + "learning_rate": 1.01545389412524e-05, + "loss": 0.6067, + "step": 12323 + }, + { + "epoch": 0.5456537861262079, + "grad_norm": 1.223731517791748, + "learning_rate": 1.0152993673037027e-05, + "loss": 0.5814, + "step": 12324 + }, + { + "epoch": 0.5456980618310218, + "grad_norm": 1.2543245553970337, + "learning_rate": 1.0151448401167525e-05, + "loss": 0.5894, + "step": 12325 + }, + { + "epoch": 0.5457423375358357, + "grad_norm": 1.4242331981658936, + "learning_rate": 1.01499031256808e-05, + "loss": 0.639, + "step": 12326 + }, + { + "epoch": 0.5457866132406495, + "grad_norm": 1.4638681411743164, + "learning_rate": 1.0148357846613762e-05, + "loss": 0.6205, + "step": 12327 + }, + { + "epoch": 0.5458308889454634, + "grad_norm": 1.3455266952514648, + "learning_rate": 1.014681256400332e-05, + "loss": 0.5709, + "step": 12328 + }, + { + "epoch": 0.5458751646502773, + "grad_norm": 1.6833152770996094, + "learning_rate": 1.0145267277886377e-05, + "loss": 0.688, + "step": 12329 + }, + { + "epoch": 0.5459194403550911, + "grad_norm": 1.4853310585021973, + "learning_rate": 1.0143721988299843e-05, + "loss": 0.5955, + "step": 12330 + }, + { + "epoch": 0.545963716059905, + "grad_norm": 1.6531833410263062, + "learning_rate": 1.0142176695280627e-05, + "loss": 0.6733, + "step": 12331 + }, + { + "epoch": 0.5460079917647189, + "grad_norm": 1.474862813949585, + "learning_rate": 1.0140631398865637e-05, + "loss": 0.5971, + "step": 12332 + }, + { + "epoch": 0.5460522674695327, + "grad_norm": 1.6692123413085938, + "learning_rate": 1.0139086099091782e-05, + "loss": 0.6593, + "step": 12333 + }, + { + "epoch": 0.5460965431743466, + "grad_norm": 1.5973126888275146, + "learning_rate": 1.0137540795995967e-05, + "loss": 0.6801, + "step": 12334 + }, + { + "epoch": 0.5461408188791606, + "grad_norm": 1.3180909156799316, + "learning_rate": 1.01359954896151e-05, + "loss": 0.6219, + "step": 12335 + }, + { + "epoch": 0.5461850945839745, + "grad_norm": 1.291370153427124, + "learning_rate": 1.0134450179986095e-05, + "loss": 0.5925, + "step": 12336 + }, + { + "epoch": 0.5462293702887883, + "grad_norm": 1.3704079389572144, + "learning_rate": 1.0132904867145855e-05, + "loss": 0.591, + "step": 12337 + }, + { + "epoch": 0.5462736459936022, + "grad_norm": 1.5659122467041016, + "learning_rate": 1.013135955113129e-05, + "loss": 0.6616, + "step": 12338 + }, + { + "epoch": 0.5463179216984161, + "grad_norm": 1.762817144393921, + "learning_rate": 1.012981423197931e-05, + "loss": 0.7052, + "step": 12339 + }, + { + "epoch": 0.5463621974032299, + "grad_norm": 1.405571699142456, + "learning_rate": 1.0128268909726824e-05, + "loss": 0.6426, + "step": 12340 + }, + { + "epoch": 0.5464064731080438, + "grad_norm": 1.4808292388916016, + "learning_rate": 1.0126723584410738e-05, + "loss": 0.5774, + "step": 12341 + }, + { + "epoch": 0.5464507488128577, + "grad_norm": 1.320870280265808, + "learning_rate": 1.0125178256067962e-05, + "loss": 0.6174, + "step": 12342 + }, + { + "epoch": 0.5464950245176715, + "grad_norm": 1.4068576097488403, + "learning_rate": 1.0123632924735406e-05, + "loss": 0.6422, + "step": 12343 + }, + { + "epoch": 0.5465393002224854, + "grad_norm": 1.4301327466964722, + "learning_rate": 1.0122087590449976e-05, + "loss": 0.6717, + "step": 12344 + }, + { + "epoch": 0.5465835759272993, + "grad_norm": 1.4876443147659302, + "learning_rate": 1.0120542253248586e-05, + "loss": 0.6549, + "step": 12345 + }, + { + "epoch": 0.5466278516321131, + "grad_norm": 1.6483111381530762, + "learning_rate": 1.0118996913168145e-05, + "loss": 0.6971, + "step": 12346 + }, + { + "epoch": 0.546672127336927, + "grad_norm": 1.3661303520202637, + "learning_rate": 1.0117451570245554e-05, + "loss": 0.6978, + "step": 12347 + }, + { + "epoch": 0.546716403041741, + "grad_norm": 1.7673660516738892, + "learning_rate": 1.0115906224517732e-05, + "loss": 0.6572, + "step": 12348 + }, + { + "epoch": 0.5467606787465548, + "grad_norm": 1.4482507705688477, + "learning_rate": 1.011436087602158e-05, + "loss": 0.5925, + "step": 12349 + }, + { + "epoch": 0.5468049544513687, + "grad_norm": 1.4043527841567993, + "learning_rate": 1.0112815524794015e-05, + "loss": 0.6729, + "step": 12350 + }, + { + "epoch": 0.5468492301561826, + "grad_norm": 1.2332751750946045, + "learning_rate": 1.0111270170871943e-05, + "loss": 0.4984, + "step": 12351 + }, + { + "epoch": 0.5468935058609964, + "grad_norm": 2.8857457637786865, + "learning_rate": 1.0109724814292272e-05, + "loss": 0.6336, + "step": 12352 + }, + { + "epoch": 0.5469377815658103, + "grad_norm": 1.4214009046554565, + "learning_rate": 1.0108179455091916e-05, + "loss": 0.6494, + "step": 12353 + }, + { + "epoch": 0.5469820572706242, + "grad_norm": 1.4033904075622559, + "learning_rate": 1.0106634093307776e-05, + "loss": 0.6193, + "step": 12354 + }, + { + "epoch": 0.547026332975438, + "grad_norm": 1.4455195665359497, + "learning_rate": 1.0105088728976773e-05, + "loss": 0.5935, + "step": 12355 + }, + { + "epoch": 0.5470706086802519, + "grad_norm": 1.3694947957992554, + "learning_rate": 1.0103543362135811e-05, + "loss": 0.6319, + "step": 12356 + }, + { + "epoch": 0.5471148843850658, + "grad_norm": 1.649929165840149, + "learning_rate": 1.0101997992821798e-05, + "loss": 0.662, + "step": 12357 + }, + { + "epoch": 0.5471591600898796, + "grad_norm": 1.3656543493270874, + "learning_rate": 1.0100452621071646e-05, + "loss": 0.5977, + "step": 12358 + }, + { + "epoch": 0.5472034357946935, + "grad_norm": 1.5687905550003052, + "learning_rate": 1.0098907246922264e-05, + "loss": 0.6611, + "step": 12359 + }, + { + "epoch": 0.5472477114995075, + "grad_norm": 1.665593147277832, + "learning_rate": 1.0097361870410566e-05, + "loss": 0.6085, + "step": 12360 + }, + { + "epoch": 0.5472919872043213, + "grad_norm": 1.488530158996582, + "learning_rate": 1.0095816491573458e-05, + "loss": 0.6702, + "step": 12361 + }, + { + "epoch": 0.5473362629091352, + "grad_norm": 1.452130675315857, + "learning_rate": 1.0094271110447848e-05, + "loss": 0.6303, + "step": 12362 + }, + { + "epoch": 0.5473805386139491, + "grad_norm": 1.7207942008972168, + "learning_rate": 1.0092725727070653e-05, + "loss": 0.638, + "step": 12363 + }, + { + "epoch": 0.5474248143187629, + "grad_norm": 1.717151403427124, + "learning_rate": 1.0091180341478781e-05, + "loss": 0.6731, + "step": 12364 + }, + { + "epoch": 0.5474690900235768, + "grad_norm": 1.3778265714645386, + "learning_rate": 1.0089634953709138e-05, + "loss": 0.6176, + "step": 12365 + }, + { + "epoch": 0.5475133657283907, + "grad_norm": 1.5798742771148682, + "learning_rate": 1.0088089563798636e-05, + "loss": 0.6542, + "step": 12366 + }, + { + "epoch": 0.5475576414332045, + "grad_norm": 1.5806639194488525, + "learning_rate": 1.0086544171784187e-05, + "loss": 0.7166, + "step": 12367 + }, + { + "epoch": 0.5476019171380184, + "grad_norm": 1.5725334882736206, + "learning_rate": 1.0084998777702701e-05, + "loss": 0.7095, + "step": 12368 + }, + { + "epoch": 0.5476461928428323, + "grad_norm": 1.5926940441131592, + "learning_rate": 1.0083453381591091e-05, + "loss": 0.704, + "step": 12369 + }, + { + "epoch": 0.5476904685476462, + "grad_norm": 1.557098150253296, + "learning_rate": 1.0081907983486264e-05, + "loss": 0.6675, + "step": 12370 + }, + { + "epoch": 0.54773474425246, + "grad_norm": 1.6034600734710693, + "learning_rate": 1.0080362583425129e-05, + "loss": 0.6841, + "step": 12371 + }, + { + "epoch": 0.547779019957274, + "grad_norm": 1.30519700050354, + "learning_rate": 1.0078817181444602e-05, + "loss": 0.617, + "step": 12372 + }, + { + "epoch": 0.5478232956620879, + "grad_norm": 1.2179533243179321, + "learning_rate": 1.0077271777581587e-05, + "loss": 0.5888, + "step": 12373 + }, + { + "epoch": 0.5478675713669017, + "grad_norm": 1.598786473274231, + "learning_rate": 1.0075726371873002e-05, + "loss": 0.6567, + "step": 12374 + }, + { + "epoch": 0.5479118470717156, + "grad_norm": 1.31406831741333, + "learning_rate": 1.0074180964355755e-05, + "loss": 0.6343, + "step": 12375 + }, + { + "epoch": 0.5479561227765295, + "grad_norm": 1.5689252614974976, + "learning_rate": 1.0072635555066752e-05, + "loss": 0.7392, + "step": 12376 + }, + { + "epoch": 0.5480003984813433, + "grad_norm": 1.5235446691513062, + "learning_rate": 1.0071090144042911e-05, + "loss": 0.6652, + "step": 12377 + }, + { + "epoch": 0.5480446741861572, + "grad_norm": 1.4791274070739746, + "learning_rate": 1.006954473132114e-05, + "loss": 0.6859, + "step": 12378 + }, + { + "epoch": 0.5480889498909711, + "grad_norm": 1.4193508625030518, + "learning_rate": 1.0067999316938348e-05, + "loss": 0.6497, + "step": 12379 + }, + { + "epoch": 0.5481332255957849, + "grad_norm": 1.1783781051635742, + "learning_rate": 1.006645390093145e-05, + "loss": 0.6018, + "step": 12380 + }, + { + "epoch": 0.5481775013005988, + "grad_norm": 1.2396284341812134, + "learning_rate": 1.0064908483337352e-05, + "loss": 0.6045, + "step": 12381 + }, + { + "epoch": 0.5482217770054127, + "grad_norm": 1.8690108060836792, + "learning_rate": 1.0063363064192968e-05, + "loss": 0.7339, + "step": 12382 + }, + { + "epoch": 0.5482660527102265, + "grad_norm": 1.6466809511184692, + "learning_rate": 1.006181764353521e-05, + "loss": 0.6834, + "step": 12383 + }, + { + "epoch": 0.5483103284150405, + "grad_norm": 1.4987802505493164, + "learning_rate": 1.0060272221400988e-05, + "loss": 0.65, + "step": 12384 + }, + { + "epoch": 0.5483546041198544, + "grad_norm": 1.9719194173812866, + "learning_rate": 1.0058726797827215e-05, + "loss": 0.6942, + "step": 12385 + }, + { + "epoch": 0.5483988798246682, + "grad_norm": 1.4208351373672485, + "learning_rate": 1.0057181372850795e-05, + "loss": 0.6431, + "step": 12386 + }, + { + "epoch": 0.5484431555294821, + "grad_norm": 1.5202149152755737, + "learning_rate": 1.0055635946508649e-05, + "loss": 0.6291, + "step": 12387 + }, + { + "epoch": 0.548487431234296, + "grad_norm": 1.4542113542556763, + "learning_rate": 1.0054090518837682e-05, + "loss": 0.5809, + "step": 12388 + }, + { + "epoch": 0.5485317069391098, + "grad_norm": 1.6102110147476196, + "learning_rate": 1.0052545089874806e-05, + "loss": 0.6209, + "step": 12389 + }, + { + "epoch": 0.5485759826439237, + "grad_norm": 1.3451093435287476, + "learning_rate": 1.0050999659656935e-05, + "loss": 0.6381, + "step": 12390 + }, + { + "epoch": 0.5486202583487376, + "grad_norm": 1.2318495512008667, + "learning_rate": 1.0049454228220976e-05, + "loss": 0.5583, + "step": 12391 + }, + { + "epoch": 0.5486645340535514, + "grad_norm": 1.4974730014801025, + "learning_rate": 1.0047908795603847e-05, + "loss": 0.6606, + "step": 12392 + }, + { + "epoch": 0.5487088097583653, + "grad_norm": 1.4449647665023804, + "learning_rate": 1.0046363361842453e-05, + "loss": 0.6285, + "step": 12393 + }, + { + "epoch": 0.5487530854631792, + "grad_norm": 1.5441690683364868, + "learning_rate": 1.0044817926973707e-05, + "loss": 0.6749, + "step": 12394 + }, + { + "epoch": 0.548797361167993, + "grad_norm": 1.3034125566482544, + "learning_rate": 1.0043272491034523e-05, + "loss": 0.6354, + "step": 12395 + }, + { + "epoch": 0.548841636872807, + "grad_norm": 1.4228664636611938, + "learning_rate": 1.004172705406181e-05, + "loss": 0.5792, + "step": 12396 + }, + { + "epoch": 0.5488859125776209, + "grad_norm": 1.3773435354232788, + "learning_rate": 1.0040181616092481e-05, + "loss": 0.5934, + "step": 12397 + }, + { + "epoch": 0.5489301882824347, + "grad_norm": 1.6442575454711914, + "learning_rate": 1.0038636177163447e-05, + "loss": 0.6666, + "step": 12398 + }, + { + "epoch": 0.5489744639872486, + "grad_norm": 1.5362012386322021, + "learning_rate": 1.0037090737311618e-05, + "loss": 0.6074, + "step": 12399 + }, + { + "epoch": 0.5490187396920625, + "grad_norm": 1.312021017074585, + "learning_rate": 1.003554529657391e-05, + "loss": 0.6575, + "step": 12400 + }, + { + "epoch": 0.5490630153968763, + "grad_norm": 1.4868366718292236, + "learning_rate": 1.0033999854987227e-05, + "loss": 0.6178, + "step": 12401 + }, + { + "epoch": 0.5491072911016902, + "grad_norm": 1.366531252861023, + "learning_rate": 1.0032454412588488e-05, + "loss": 0.6143, + "step": 12402 + }, + { + "epoch": 0.5491515668065041, + "grad_norm": 1.257454514503479, + "learning_rate": 1.0030908969414603e-05, + "loss": 0.537, + "step": 12403 + }, + { + "epoch": 0.5491958425113179, + "grad_norm": 1.319171667098999, + "learning_rate": 1.0029363525502478e-05, + "loss": 0.5755, + "step": 12404 + }, + { + "epoch": 0.5492401182161318, + "grad_norm": 1.405344843864441, + "learning_rate": 1.0027818080889034e-05, + "loss": 0.6547, + "step": 12405 + }, + { + "epoch": 0.5492843939209457, + "grad_norm": 1.3424787521362305, + "learning_rate": 1.0026272635611173e-05, + "loss": 0.6177, + "step": 12406 + }, + { + "epoch": 0.5493286696257597, + "grad_norm": 1.3077940940856934, + "learning_rate": 1.0024727189705818e-05, + "loss": 0.5401, + "step": 12407 + }, + { + "epoch": 0.5493729453305735, + "grad_norm": 1.525731086730957, + "learning_rate": 1.0023181743209868e-05, + "loss": 0.6182, + "step": 12408 + }, + { + "epoch": 0.5494172210353874, + "grad_norm": 1.6968477964401245, + "learning_rate": 1.0021636296160243e-05, + "loss": 0.6797, + "step": 12409 + }, + { + "epoch": 0.5494614967402013, + "grad_norm": 1.7207084894180298, + "learning_rate": 1.0020090848593853e-05, + "loss": 0.6861, + "step": 12410 + }, + { + "epoch": 0.5495057724450151, + "grad_norm": 1.646845817565918, + "learning_rate": 1.0018545400547609e-05, + "loss": 0.6423, + "step": 12411 + }, + { + "epoch": 0.549550048149829, + "grad_norm": 1.5958971977233887, + "learning_rate": 1.0016999952058426e-05, + "loss": 0.6116, + "step": 12412 + }, + { + "epoch": 0.5495943238546429, + "grad_norm": 1.3430598974227905, + "learning_rate": 1.0015454503163209e-05, + "loss": 0.6232, + "step": 12413 + }, + { + "epoch": 0.5496385995594567, + "grad_norm": 1.4943037033081055, + "learning_rate": 1.0013909053898874e-05, + "loss": 0.6569, + "step": 12414 + }, + { + "epoch": 0.5496828752642706, + "grad_norm": 1.4160939455032349, + "learning_rate": 1.0012363604302335e-05, + "loss": 0.5655, + "step": 12415 + }, + { + "epoch": 0.5497271509690845, + "grad_norm": 1.3634406328201294, + "learning_rate": 1.0010818154410501e-05, + "loss": 0.618, + "step": 12416 + }, + { + "epoch": 0.5497714266738983, + "grad_norm": 1.5668734312057495, + "learning_rate": 1.0009272704260285e-05, + "loss": 0.6439, + "step": 12417 + }, + { + "epoch": 0.5498157023787122, + "grad_norm": 1.5198748111724854, + "learning_rate": 1.0007727253888595e-05, + "loss": 0.6695, + "step": 12418 + }, + { + "epoch": 0.5498599780835262, + "grad_norm": 1.445271372795105, + "learning_rate": 1.0006181803332346e-05, + "loss": 0.6226, + "step": 12419 + }, + { + "epoch": 0.54990425378834, + "grad_norm": 1.6568360328674316, + "learning_rate": 1.0004636352628451e-05, + "loss": 0.5909, + "step": 12420 + }, + { + "epoch": 0.5499485294931539, + "grad_norm": 1.495185136795044, + "learning_rate": 1.000309090181382e-05, + "loss": 0.6081, + "step": 12421 + }, + { + "epoch": 0.5499928051979678, + "grad_norm": 1.848172664642334, + "learning_rate": 1.000154545092537e-05, + "loss": 0.6706, + "step": 12422 + }, + { + "epoch": 0.5500370809027816, + "grad_norm": 1.4411182403564453, + "learning_rate": 1e-05, + "loss": 0.6377, + "step": 12423 + }, + { + "epoch": 0.5500813566075955, + "grad_norm": 1.3759945631027222, + "learning_rate": 9.998454549074636e-06, + "loss": 0.6724, + "step": 12424 + }, + { + "epoch": 0.5501256323124094, + "grad_norm": 1.537758708000183, + "learning_rate": 9.996909098186181e-06, + "loss": 0.6356, + "step": 12425 + }, + { + "epoch": 0.5501699080172232, + "grad_norm": 1.3226559162139893, + "learning_rate": 9.99536364737155e-06, + "loss": 0.6179, + "step": 12426 + }, + { + "epoch": 0.5502141837220371, + "grad_norm": 1.5552359819412231, + "learning_rate": 9.993818196667657e-06, + "loss": 0.5655, + "step": 12427 + }, + { + "epoch": 0.550258459426851, + "grad_norm": 1.8257476091384888, + "learning_rate": 9.99227274611141e-06, + "loss": 0.6499, + "step": 12428 + }, + { + "epoch": 0.5503027351316648, + "grad_norm": 1.3616623878479004, + "learning_rate": 9.990727295739718e-06, + "loss": 0.5745, + "step": 12429 + }, + { + "epoch": 0.5503470108364787, + "grad_norm": 1.3949365615844727, + "learning_rate": 9.989181845589502e-06, + "loss": 0.6055, + "step": 12430 + }, + { + "epoch": 0.5503912865412927, + "grad_norm": 1.21440851688385, + "learning_rate": 9.98763639569767e-06, + "loss": 0.5997, + "step": 12431 + }, + { + "epoch": 0.5504355622461065, + "grad_norm": 1.4212722778320312, + "learning_rate": 9.986090946101126e-06, + "loss": 0.6872, + "step": 12432 + }, + { + "epoch": 0.5504798379509204, + "grad_norm": 1.866866111755371, + "learning_rate": 9.984545496836793e-06, + "loss": 0.7493, + "step": 12433 + }, + { + "epoch": 0.5505241136557343, + "grad_norm": 1.3736618757247925, + "learning_rate": 9.98300004794158e-06, + "loss": 0.5725, + "step": 12434 + }, + { + "epoch": 0.5505683893605481, + "grad_norm": 1.5210094451904297, + "learning_rate": 9.981454599452391e-06, + "loss": 0.6386, + "step": 12435 + }, + { + "epoch": 0.550612665065362, + "grad_norm": 1.4416195154190063, + "learning_rate": 9.979909151406149e-06, + "loss": 0.6537, + "step": 12436 + }, + { + "epoch": 0.5506569407701759, + "grad_norm": 1.4847118854522705, + "learning_rate": 9.978363703839759e-06, + "loss": 0.6299, + "step": 12437 + }, + { + "epoch": 0.5507012164749897, + "grad_norm": 1.4528536796569824, + "learning_rate": 9.976818256790137e-06, + "loss": 0.6296, + "step": 12438 + }, + { + "epoch": 0.5507454921798036, + "grad_norm": 1.811938762664795, + "learning_rate": 9.975272810294185e-06, + "loss": 0.6581, + "step": 12439 + }, + { + "epoch": 0.5507897678846175, + "grad_norm": 1.1531277894973755, + "learning_rate": 9.973727364388828e-06, + "loss": 0.6014, + "step": 12440 + }, + { + "epoch": 0.5508340435894314, + "grad_norm": 1.3136839866638184, + "learning_rate": 9.97218191911097e-06, + "loss": 0.591, + "step": 12441 + }, + { + "epoch": 0.5508783192942452, + "grad_norm": 1.447188138961792, + "learning_rate": 9.97063647449752e-06, + "loss": 0.656, + "step": 12442 + }, + { + "epoch": 0.5509225949990592, + "grad_norm": 1.3953546285629272, + "learning_rate": 9.9690910305854e-06, + "loss": 0.5569, + "step": 12443 + }, + { + "epoch": 0.5509668707038731, + "grad_norm": 1.2087078094482422, + "learning_rate": 9.967545587411515e-06, + "loss": 0.5993, + "step": 12444 + }, + { + "epoch": 0.5510111464086869, + "grad_norm": 1.611862063407898, + "learning_rate": 9.966000145012773e-06, + "loss": 0.6761, + "step": 12445 + }, + { + "epoch": 0.5510554221135008, + "grad_norm": 1.3534605503082275, + "learning_rate": 9.964454703426094e-06, + "loss": 0.6287, + "step": 12446 + }, + { + "epoch": 0.5510996978183147, + "grad_norm": 1.7575228214263916, + "learning_rate": 9.962909262688384e-06, + "loss": 0.6479, + "step": 12447 + }, + { + "epoch": 0.5511439735231285, + "grad_norm": 1.3520015478134155, + "learning_rate": 9.961363822836558e-06, + "loss": 0.5888, + "step": 12448 + }, + { + "epoch": 0.5511882492279424, + "grad_norm": 1.4295849800109863, + "learning_rate": 9.95981838390752e-06, + "loss": 0.6803, + "step": 12449 + }, + { + "epoch": 0.5512325249327563, + "grad_norm": 1.3123433589935303, + "learning_rate": 9.958272945938194e-06, + "loss": 0.6428, + "step": 12450 + }, + { + "epoch": 0.5512768006375701, + "grad_norm": 1.3302236795425415, + "learning_rate": 9.956727508965482e-06, + "loss": 0.6052, + "step": 12451 + }, + { + "epoch": 0.551321076342384, + "grad_norm": 1.8712791204452515, + "learning_rate": 9.955182073026293e-06, + "loss": 0.707, + "step": 12452 + }, + { + "epoch": 0.5513653520471979, + "grad_norm": 1.3789565563201904, + "learning_rate": 9.953636638157549e-06, + "loss": 0.6019, + "step": 12453 + }, + { + "epoch": 0.5514096277520117, + "grad_norm": 1.7146110534667969, + "learning_rate": 9.952091204396156e-06, + "loss": 0.6953, + "step": 12454 + }, + { + "epoch": 0.5514539034568257, + "grad_norm": 1.6598719358444214, + "learning_rate": 9.950545771779024e-06, + "loss": 0.6955, + "step": 12455 + }, + { + "epoch": 0.5514981791616396, + "grad_norm": 1.450282096862793, + "learning_rate": 9.949000340343069e-06, + "loss": 0.5888, + "step": 12456 + }, + { + "epoch": 0.5515424548664534, + "grad_norm": 1.4189730882644653, + "learning_rate": 9.947454910125198e-06, + "loss": 0.587, + "step": 12457 + }, + { + "epoch": 0.5515867305712673, + "grad_norm": 1.3020497560501099, + "learning_rate": 9.945909481162323e-06, + "loss": 0.6028, + "step": 12458 + }, + { + "epoch": 0.5516310062760812, + "grad_norm": 1.5524383783340454, + "learning_rate": 9.944364053491356e-06, + "loss": 0.5904, + "step": 12459 + }, + { + "epoch": 0.551675281980895, + "grad_norm": 1.2694565057754517, + "learning_rate": 9.942818627149208e-06, + "loss": 0.5575, + "step": 12460 + }, + { + "epoch": 0.5517195576857089, + "grad_norm": 1.981475591659546, + "learning_rate": 9.941273202172792e-06, + "loss": 0.7307, + "step": 12461 + }, + { + "epoch": 0.5517638333905228, + "grad_norm": 1.350615382194519, + "learning_rate": 9.939727778599014e-06, + "loss": 0.5844, + "step": 12462 + }, + { + "epoch": 0.5518081090953366, + "grad_norm": 1.518746256828308, + "learning_rate": 9.938182356464793e-06, + "loss": 0.6131, + "step": 12463 + }, + { + "epoch": 0.5518523848001505, + "grad_norm": 1.3110612630844116, + "learning_rate": 9.936636935807036e-06, + "loss": 0.5447, + "step": 12464 + }, + { + "epoch": 0.5518966605049644, + "grad_norm": 1.278048038482666, + "learning_rate": 9.93509151666265e-06, + "loss": 0.687, + "step": 12465 + }, + { + "epoch": 0.5519409362097782, + "grad_norm": 1.509049892425537, + "learning_rate": 9.933546099068553e-06, + "loss": 0.6038, + "step": 12466 + }, + { + "epoch": 0.5519852119145922, + "grad_norm": 1.5184051990509033, + "learning_rate": 9.932000683061654e-06, + "loss": 0.683, + "step": 12467 + }, + { + "epoch": 0.5520294876194061, + "grad_norm": 1.5356374979019165, + "learning_rate": 9.930455268678864e-06, + "loss": 0.6918, + "step": 12468 + }, + { + "epoch": 0.5520737633242199, + "grad_norm": 1.3079859018325806, + "learning_rate": 9.92890985595709e-06, + "loss": 0.5646, + "step": 12469 + }, + { + "epoch": 0.5521180390290338, + "grad_norm": 1.6296038627624512, + "learning_rate": 9.92736444493325e-06, + "loss": 0.6587, + "step": 12470 + }, + { + "epoch": 0.5521623147338477, + "grad_norm": 1.6976293325424194, + "learning_rate": 9.92581903564425e-06, + "loss": 0.643, + "step": 12471 + }, + { + "epoch": 0.5522065904386615, + "grad_norm": 1.5536081790924072, + "learning_rate": 9.924273628127e-06, + "loss": 0.6598, + "step": 12472 + }, + { + "epoch": 0.5522508661434754, + "grad_norm": 1.563403844833374, + "learning_rate": 9.922728222418415e-06, + "loss": 0.609, + "step": 12473 + }, + { + "epoch": 0.5522951418482893, + "grad_norm": 1.5703434944152832, + "learning_rate": 9.921182818555403e-06, + "loss": 0.6406, + "step": 12474 + }, + { + "epoch": 0.5523394175531031, + "grad_norm": 1.8096349239349365, + "learning_rate": 9.919637416574871e-06, + "loss": 0.6872, + "step": 12475 + }, + { + "epoch": 0.552383693257917, + "grad_norm": 1.803613543510437, + "learning_rate": 9.918092016513738e-06, + "loss": 0.6533, + "step": 12476 + }, + { + "epoch": 0.5524279689627309, + "grad_norm": 1.6659648418426514, + "learning_rate": 9.916546618408912e-06, + "loss": 0.6781, + "step": 12477 + }, + { + "epoch": 0.5524722446675449, + "grad_norm": 2.043555974960327, + "learning_rate": 9.9150012222973e-06, + "loss": 0.6597, + "step": 12478 + }, + { + "epoch": 0.5525165203723587, + "grad_norm": 1.360729694366455, + "learning_rate": 9.913455828215815e-06, + "loss": 0.5675, + "step": 12479 + }, + { + "epoch": 0.5525607960771726, + "grad_norm": 1.449683666229248, + "learning_rate": 9.911910436201366e-06, + "loss": 0.6362, + "step": 12480 + }, + { + "epoch": 0.5526050717819865, + "grad_norm": 1.4168246984481812, + "learning_rate": 9.910365046290866e-06, + "loss": 0.6411, + "step": 12481 + }, + { + "epoch": 0.5526493474868003, + "grad_norm": 1.3673949241638184, + "learning_rate": 9.908819658521222e-06, + "loss": 0.6248, + "step": 12482 + }, + { + "epoch": 0.5526936231916142, + "grad_norm": 1.3864426612854004, + "learning_rate": 9.90727427292935e-06, + "loss": 0.5807, + "step": 12483 + }, + { + "epoch": 0.5527378988964281, + "grad_norm": 1.3677773475646973, + "learning_rate": 9.905728889552154e-06, + "loss": 0.5816, + "step": 12484 + }, + { + "epoch": 0.5527821746012419, + "grad_norm": 1.5013107061386108, + "learning_rate": 9.904183508426542e-06, + "loss": 0.6009, + "step": 12485 + }, + { + "epoch": 0.5528264503060558, + "grad_norm": 1.3091905117034912, + "learning_rate": 9.902638129589435e-06, + "loss": 0.5851, + "step": 12486 + }, + { + "epoch": 0.5528707260108697, + "grad_norm": 1.374711275100708, + "learning_rate": 9.901092753077738e-06, + "loss": 0.5974, + "step": 12487 + }, + { + "epoch": 0.5529150017156835, + "grad_norm": 1.4404525756835938, + "learning_rate": 9.899547378928359e-06, + "loss": 0.7149, + "step": 12488 + }, + { + "epoch": 0.5529592774204974, + "grad_norm": 1.425755500793457, + "learning_rate": 9.898002007178204e-06, + "loss": 0.619, + "step": 12489 + }, + { + "epoch": 0.5530035531253114, + "grad_norm": 1.2997856140136719, + "learning_rate": 9.896456637864192e-06, + "loss": 0.6384, + "step": 12490 + }, + { + "epoch": 0.5530478288301252, + "grad_norm": 1.3248106241226196, + "learning_rate": 9.89491127102323e-06, + "loss": 0.6174, + "step": 12491 + }, + { + "epoch": 0.5530921045349391, + "grad_norm": 1.767928123474121, + "learning_rate": 9.893365906692224e-06, + "loss": 0.6369, + "step": 12492 + }, + { + "epoch": 0.553136380239753, + "grad_norm": 1.3854272365570068, + "learning_rate": 9.891820544908088e-06, + "loss": 0.5786, + "step": 12493 + }, + { + "epoch": 0.5531806559445668, + "grad_norm": 1.388922929763794, + "learning_rate": 9.89027518570773e-06, + "loss": 0.5964, + "step": 12494 + }, + { + "epoch": 0.5532249316493807, + "grad_norm": 1.6180140972137451, + "learning_rate": 9.88872982912806e-06, + "loss": 0.6525, + "step": 12495 + }, + { + "epoch": 0.5532692073541946, + "grad_norm": 1.3673112392425537, + "learning_rate": 9.887184475205987e-06, + "loss": 0.6376, + "step": 12496 + }, + { + "epoch": 0.5533134830590084, + "grad_norm": 2.062218427658081, + "learning_rate": 9.885639123978423e-06, + "loss": 0.6631, + "step": 12497 + }, + { + "epoch": 0.5533577587638223, + "grad_norm": 1.6020395755767822, + "learning_rate": 9.884093775482273e-06, + "loss": 0.6299, + "step": 12498 + }, + { + "epoch": 0.5534020344686362, + "grad_norm": 1.5316784381866455, + "learning_rate": 9.882548429754446e-06, + "loss": 0.5884, + "step": 12499 + }, + { + "epoch": 0.55344631017345, + "grad_norm": 1.4409044981002808, + "learning_rate": 9.881003086831859e-06, + "loss": 0.61, + "step": 12500 + }, + { + "epoch": 0.5534905858782639, + "grad_norm": 1.3661777973175049, + "learning_rate": 9.879457746751415e-06, + "loss": 0.5617, + "step": 12501 + }, + { + "epoch": 0.5535348615830779, + "grad_norm": 1.7233277559280396, + "learning_rate": 9.877912409550024e-06, + "loss": 0.6383, + "step": 12502 + }, + { + "epoch": 0.5535791372878917, + "grad_norm": 1.7712451219558716, + "learning_rate": 9.876367075264597e-06, + "loss": 0.6918, + "step": 12503 + }, + { + "epoch": 0.5536234129927056, + "grad_norm": 1.5005700588226318, + "learning_rate": 9.87482174393204e-06, + "loss": 0.6592, + "step": 12504 + }, + { + "epoch": 0.5536676886975195, + "grad_norm": 1.6575021743774414, + "learning_rate": 9.873276415589265e-06, + "loss": 0.7546, + "step": 12505 + }, + { + "epoch": 0.5537119644023333, + "grad_norm": 1.6232539415359497, + "learning_rate": 9.87173109027318e-06, + "loss": 0.564, + "step": 12506 + }, + { + "epoch": 0.5537562401071472, + "grad_norm": 1.295854091644287, + "learning_rate": 9.870185768020694e-06, + "loss": 0.5852, + "step": 12507 + }, + { + "epoch": 0.5538005158119611, + "grad_norm": 1.473242163658142, + "learning_rate": 9.868640448868714e-06, + "loss": 0.6153, + "step": 12508 + }, + { + "epoch": 0.5538447915167749, + "grad_norm": 1.678976058959961, + "learning_rate": 9.867095132854147e-06, + "loss": 0.6493, + "step": 12509 + }, + { + "epoch": 0.5538890672215888, + "grad_norm": 1.7492904663085938, + "learning_rate": 9.865549820013907e-06, + "loss": 0.6673, + "step": 12510 + }, + { + "epoch": 0.5539333429264027, + "grad_norm": 1.3985849618911743, + "learning_rate": 9.864004510384903e-06, + "loss": 0.6576, + "step": 12511 + }, + { + "epoch": 0.5539776186312166, + "grad_norm": 1.757063627243042, + "learning_rate": 9.862459204004035e-06, + "loss": 0.6128, + "step": 12512 + }, + { + "epoch": 0.5540218943360304, + "grad_norm": 1.81006920337677, + "learning_rate": 9.860913900908222e-06, + "loss": 0.7027, + "step": 12513 + }, + { + "epoch": 0.5540661700408444, + "grad_norm": 1.3545197248458862, + "learning_rate": 9.859368601134365e-06, + "loss": 0.6019, + "step": 12514 + }, + { + "epoch": 0.5541104457456583, + "grad_norm": 1.2639944553375244, + "learning_rate": 9.857823304719375e-06, + "loss": 0.6265, + "step": 12515 + }, + { + "epoch": 0.5541547214504721, + "grad_norm": 1.2954148054122925, + "learning_rate": 9.856278011700158e-06, + "loss": 0.5934, + "step": 12516 + }, + { + "epoch": 0.554198997155286, + "grad_norm": 1.4734630584716797, + "learning_rate": 9.854732722113628e-06, + "loss": 0.5766, + "step": 12517 + }, + { + "epoch": 0.5542432728600999, + "grad_norm": 1.5815399885177612, + "learning_rate": 9.853187435996684e-06, + "loss": 0.6594, + "step": 12518 + }, + { + "epoch": 0.5542875485649137, + "grad_norm": 1.6559760570526123, + "learning_rate": 9.851642153386239e-06, + "loss": 0.6719, + "step": 12519 + }, + { + "epoch": 0.5543318242697276, + "grad_norm": 1.3331239223480225, + "learning_rate": 9.850096874319202e-06, + "loss": 0.5874, + "step": 12520 + }, + { + "epoch": 0.5543760999745415, + "grad_norm": 1.32712984085083, + "learning_rate": 9.84855159883248e-06, + "loss": 0.6311, + "step": 12521 + }, + { + "epoch": 0.5544203756793553, + "grad_norm": 1.562124252319336, + "learning_rate": 9.847006326962975e-06, + "loss": 0.6631, + "step": 12522 + }, + { + "epoch": 0.5544646513841692, + "grad_norm": 1.4039998054504395, + "learning_rate": 9.845461058747602e-06, + "loss": 0.59, + "step": 12523 + }, + { + "epoch": 0.5545089270889831, + "grad_norm": 1.1835529804229736, + "learning_rate": 9.843915794223269e-06, + "loss": 0.5912, + "step": 12524 + }, + { + "epoch": 0.5545532027937969, + "grad_norm": 1.5671848058700562, + "learning_rate": 9.842370533426876e-06, + "loss": 0.6198, + "step": 12525 + }, + { + "epoch": 0.5545974784986109, + "grad_norm": 1.4337823390960693, + "learning_rate": 9.840825276395335e-06, + "loss": 0.5903, + "step": 12526 + }, + { + "epoch": 0.5546417542034248, + "grad_norm": 1.2929457426071167, + "learning_rate": 9.839280023165556e-06, + "loss": 0.5873, + "step": 12527 + }, + { + "epoch": 0.5546860299082386, + "grad_norm": 1.4495114088058472, + "learning_rate": 9.837734773774442e-06, + "loss": 0.6282, + "step": 12528 + }, + { + "epoch": 0.5547303056130525, + "grad_norm": 1.8730708360671997, + "learning_rate": 9.836189528258897e-06, + "loss": 0.7355, + "step": 12529 + }, + { + "epoch": 0.5547745813178664, + "grad_norm": 1.3101109266281128, + "learning_rate": 9.834644286655835e-06, + "loss": 0.5953, + "step": 12530 + }, + { + "epoch": 0.5548188570226802, + "grad_norm": 1.2858730554580688, + "learning_rate": 9.83309904900216e-06, + "loss": 0.6048, + "step": 12531 + }, + { + "epoch": 0.5548631327274941, + "grad_norm": 1.489431619644165, + "learning_rate": 9.831553815334773e-06, + "loss": 0.635, + "step": 12532 + }, + { + "epoch": 0.554907408432308, + "grad_norm": 1.9333817958831787, + "learning_rate": 9.83000858569059e-06, + "loss": 0.7861, + "step": 12533 + }, + { + "epoch": 0.5549516841371218, + "grad_norm": 1.213619351387024, + "learning_rate": 9.828463360106518e-06, + "loss": 0.5604, + "step": 12534 + }, + { + "epoch": 0.5549959598419357, + "grad_norm": 1.3179258108139038, + "learning_rate": 9.826918138619454e-06, + "loss": 0.633, + "step": 12535 + }, + { + "epoch": 0.5550402355467496, + "grad_norm": 1.1802947521209717, + "learning_rate": 9.825372921266306e-06, + "loss": 0.6617, + "step": 12536 + }, + { + "epoch": 0.5550845112515634, + "grad_norm": 1.6858298778533936, + "learning_rate": 9.823827708083988e-06, + "loss": 0.6981, + "step": 12537 + }, + { + "epoch": 0.5551287869563774, + "grad_norm": 1.4363633394241333, + "learning_rate": 9.822282499109402e-06, + "loss": 0.6731, + "step": 12538 + }, + { + "epoch": 0.5551730626611913, + "grad_norm": 1.3477473258972168, + "learning_rate": 9.820737294379452e-06, + "loss": 0.5518, + "step": 12539 + }, + { + "epoch": 0.5552173383660051, + "grad_norm": 1.8550268411636353, + "learning_rate": 9.819192093931047e-06, + "loss": 0.6603, + "step": 12540 + }, + { + "epoch": 0.555261614070819, + "grad_norm": 1.710944414138794, + "learning_rate": 9.817646897801093e-06, + "loss": 0.67, + "step": 12541 + }, + { + "epoch": 0.5553058897756329, + "grad_norm": 1.2830222845077515, + "learning_rate": 9.816101706026488e-06, + "loss": 0.6293, + "step": 12542 + }, + { + "epoch": 0.5553501654804467, + "grad_norm": 1.3291014432907104, + "learning_rate": 9.814556518644153e-06, + "loss": 0.6124, + "step": 12543 + }, + { + "epoch": 0.5553944411852606, + "grad_norm": 1.7722384929656982, + "learning_rate": 9.81301133569098e-06, + "loss": 0.6953, + "step": 12544 + }, + { + "epoch": 0.5554387168900745, + "grad_norm": 1.3150488138198853, + "learning_rate": 9.81146615720388e-06, + "loss": 0.6277, + "step": 12545 + }, + { + "epoch": 0.5554829925948884, + "grad_norm": 1.8632557392120361, + "learning_rate": 9.809920983219754e-06, + "loss": 0.6501, + "step": 12546 + }, + { + "epoch": 0.5555272682997022, + "grad_norm": 1.7722768783569336, + "learning_rate": 9.808375813775512e-06, + "loss": 0.7251, + "step": 12547 + }, + { + "epoch": 0.5555715440045161, + "grad_norm": 1.2925491333007812, + "learning_rate": 9.806830648908061e-06, + "loss": 0.6308, + "step": 12548 + }, + { + "epoch": 0.55561581970933, + "grad_norm": 1.546216607093811, + "learning_rate": 9.805285488654298e-06, + "loss": 0.6503, + "step": 12549 + }, + { + "epoch": 0.5556600954141439, + "grad_norm": 1.5454275608062744, + "learning_rate": 9.803740333051136e-06, + "loss": 0.6077, + "step": 12550 + }, + { + "epoch": 0.5557043711189578, + "grad_norm": 1.6911473274230957, + "learning_rate": 9.802195182135477e-06, + "loss": 0.6441, + "step": 12551 + }, + { + "epoch": 0.5557486468237717, + "grad_norm": 1.5238057374954224, + "learning_rate": 9.800650035944222e-06, + "loss": 0.6477, + "step": 12552 + }, + { + "epoch": 0.5557929225285855, + "grad_norm": 2.1594579219818115, + "learning_rate": 9.79910489451428e-06, + "loss": 0.7009, + "step": 12553 + }, + { + "epoch": 0.5558371982333994, + "grad_norm": 1.4700067043304443, + "learning_rate": 9.797559757882554e-06, + "loss": 0.5378, + "step": 12554 + }, + { + "epoch": 0.5558814739382133, + "grad_norm": 1.568808913230896, + "learning_rate": 9.79601462608595e-06, + "loss": 0.6491, + "step": 12555 + }, + { + "epoch": 0.5559257496430271, + "grad_norm": 1.6073977947235107, + "learning_rate": 9.794469499161365e-06, + "loss": 0.697, + "step": 12556 + }, + { + "epoch": 0.555970025347841, + "grad_norm": 1.429648518562317, + "learning_rate": 9.792924377145714e-06, + "loss": 0.5625, + "step": 12557 + }, + { + "epoch": 0.5560143010526549, + "grad_norm": 1.389016032218933, + "learning_rate": 9.791379260075896e-06, + "loss": 0.6521, + "step": 12558 + }, + { + "epoch": 0.5560585767574687, + "grad_norm": 1.5836857557296753, + "learning_rate": 9.78983414798881e-06, + "loss": 0.7344, + "step": 12559 + }, + { + "epoch": 0.5561028524622826, + "grad_norm": 2.1740970611572266, + "learning_rate": 9.788289040921367e-06, + "loss": 0.6947, + "step": 12560 + }, + { + "epoch": 0.5561471281670965, + "grad_norm": 2.341158628463745, + "learning_rate": 9.786743938910469e-06, + "loss": 0.6685, + "step": 12561 + }, + { + "epoch": 0.5561914038719104, + "grad_norm": 1.4250397682189941, + "learning_rate": 9.785198841993015e-06, + "loss": 0.5437, + "step": 12562 + }, + { + "epoch": 0.5562356795767243, + "grad_norm": 1.6189510822296143, + "learning_rate": 9.783653750205916e-06, + "loss": 0.5991, + "step": 12563 + }, + { + "epoch": 0.5562799552815382, + "grad_norm": 1.805523157119751, + "learning_rate": 9.78210866358607e-06, + "loss": 0.6416, + "step": 12564 + }, + { + "epoch": 0.556324230986352, + "grad_norm": 1.1006627082824707, + "learning_rate": 9.780563582170381e-06, + "loss": 0.5136, + "step": 12565 + }, + { + "epoch": 0.5563685066911659, + "grad_norm": 1.5752156972885132, + "learning_rate": 9.779018505995746e-06, + "loss": 0.7325, + "step": 12566 + }, + { + "epoch": 0.5564127823959798, + "grad_norm": 1.4170746803283691, + "learning_rate": 9.777473435099084e-06, + "loss": 0.6665, + "step": 12567 + }, + { + "epoch": 0.5564570581007936, + "grad_norm": 1.5852934122085571, + "learning_rate": 9.775928369517282e-06, + "loss": 0.6608, + "step": 12568 + }, + { + "epoch": 0.5565013338056075, + "grad_norm": 1.9726277589797974, + "learning_rate": 9.774383309287246e-06, + "loss": 0.6734, + "step": 12569 + }, + { + "epoch": 0.5565456095104214, + "grad_norm": 1.8341642618179321, + "learning_rate": 9.772838254445887e-06, + "loss": 0.7312, + "step": 12570 + }, + { + "epoch": 0.5565898852152352, + "grad_norm": 1.8736472129821777, + "learning_rate": 9.771293205030098e-06, + "loss": 0.7038, + "step": 12571 + }, + { + "epoch": 0.5566341609200491, + "grad_norm": 1.55894136428833, + "learning_rate": 9.769748161076785e-06, + "loss": 0.6233, + "step": 12572 + }, + { + "epoch": 0.556678436624863, + "grad_norm": 1.6444048881530762, + "learning_rate": 9.768203122622851e-06, + "loss": 0.6698, + "step": 12573 + }, + { + "epoch": 0.5567227123296768, + "grad_norm": 1.6942299604415894, + "learning_rate": 9.766658089705196e-06, + "loss": 0.7455, + "step": 12574 + }, + { + "epoch": 0.5567669880344908, + "grad_norm": 1.5242359638214111, + "learning_rate": 9.765113062360723e-06, + "loss": 0.637, + "step": 12575 + }, + { + "epoch": 0.5568112637393047, + "grad_norm": 1.3670575618743896, + "learning_rate": 9.763568040626333e-06, + "loss": 0.602, + "step": 12576 + }, + { + "epoch": 0.5568555394441185, + "grad_norm": 1.9641071557998657, + "learning_rate": 9.762023024538928e-06, + "loss": 0.6445, + "step": 12577 + }, + { + "epoch": 0.5568998151489324, + "grad_norm": 2.5868587493896484, + "learning_rate": 9.760478014135408e-06, + "loss": 0.7441, + "step": 12578 + }, + { + "epoch": 0.5569440908537463, + "grad_norm": 1.5872265100479126, + "learning_rate": 9.758933009452674e-06, + "loss": 0.62, + "step": 12579 + }, + { + "epoch": 0.5569883665585601, + "grad_norm": 1.4922454357147217, + "learning_rate": 9.757388010527632e-06, + "loss": 0.6151, + "step": 12580 + }, + { + "epoch": 0.557032642263374, + "grad_norm": 1.4519261121749878, + "learning_rate": 9.75584301739718e-06, + "loss": 0.5644, + "step": 12581 + }, + { + "epoch": 0.5570769179681879, + "grad_norm": 1.561537742614746, + "learning_rate": 9.754298030098215e-06, + "loss": 0.6193, + "step": 12582 + }, + { + "epoch": 0.5571211936730018, + "grad_norm": 1.4710184335708618, + "learning_rate": 9.752753048667646e-06, + "loss": 0.6506, + "step": 12583 + }, + { + "epoch": 0.5571654693778156, + "grad_norm": 1.6062687635421753, + "learning_rate": 9.751208073142368e-06, + "loss": 0.7124, + "step": 12584 + }, + { + "epoch": 0.5572097450826295, + "grad_norm": 1.2772936820983887, + "learning_rate": 9.749663103559283e-06, + "loss": 0.5439, + "step": 12585 + }, + { + "epoch": 0.5572540207874435, + "grad_norm": 1.7488462924957275, + "learning_rate": 9.748118139955288e-06, + "loss": 0.6692, + "step": 12586 + }, + { + "epoch": 0.5572982964922573, + "grad_norm": 1.9158207178115845, + "learning_rate": 9.746573182367288e-06, + "loss": 0.7129, + "step": 12587 + }, + { + "epoch": 0.5573425721970712, + "grad_norm": 1.6074774265289307, + "learning_rate": 9.745028230832183e-06, + "loss": 0.6315, + "step": 12588 + }, + { + "epoch": 0.5573868479018851, + "grad_norm": 1.7180815935134888, + "learning_rate": 9.743483285386865e-06, + "loss": 0.6427, + "step": 12589 + }, + { + "epoch": 0.5574311236066989, + "grad_norm": 1.0946093797683716, + "learning_rate": 9.741938346068248e-06, + "loss": 0.556, + "step": 12590 + }, + { + "epoch": 0.5574753993115128, + "grad_norm": 1.4679876565933228, + "learning_rate": 9.740393412913219e-06, + "loss": 0.6175, + "step": 12591 + }, + { + "epoch": 0.5575196750163267, + "grad_norm": 1.413659930229187, + "learning_rate": 9.738848485958679e-06, + "loss": 0.5861, + "step": 12592 + }, + { + "epoch": 0.5575639507211405, + "grad_norm": 1.595180869102478, + "learning_rate": 9.737303565241534e-06, + "loss": 0.6864, + "step": 12593 + }, + { + "epoch": 0.5576082264259544, + "grad_norm": 1.4235401153564453, + "learning_rate": 9.73575865079868e-06, + "loss": 0.5858, + "step": 12594 + }, + { + "epoch": 0.5576525021307683, + "grad_norm": 1.4657179117202759, + "learning_rate": 9.734213742667017e-06, + "loss": 0.5893, + "step": 12595 + }, + { + "epoch": 0.5576967778355821, + "grad_norm": 1.7203850746154785, + "learning_rate": 9.732668840883437e-06, + "loss": 0.7887, + "step": 12596 + }, + { + "epoch": 0.557741053540396, + "grad_norm": 1.7407809495925903, + "learning_rate": 9.731123945484849e-06, + "loss": 0.6633, + "step": 12597 + }, + { + "epoch": 0.55778532924521, + "grad_norm": 1.4268827438354492, + "learning_rate": 9.729579056508146e-06, + "loss": 0.7157, + "step": 12598 + }, + { + "epoch": 0.5578296049500238, + "grad_norm": 1.3743783235549927, + "learning_rate": 9.728034173990224e-06, + "loss": 0.6338, + "step": 12599 + }, + { + "epoch": 0.5578738806548377, + "grad_norm": 1.6554553508758545, + "learning_rate": 9.726489297967987e-06, + "loss": 0.7065, + "step": 12600 + }, + { + "epoch": 0.5579181563596516, + "grad_norm": 1.8015364408493042, + "learning_rate": 9.724944428478331e-06, + "loss": 0.6428, + "step": 12601 + }, + { + "epoch": 0.5579624320644654, + "grad_norm": 1.342557430267334, + "learning_rate": 9.723399565558149e-06, + "loss": 0.5983, + "step": 12602 + }, + { + "epoch": 0.5580067077692793, + "grad_norm": 1.3038017749786377, + "learning_rate": 9.721854709244346e-06, + "loss": 0.6005, + "step": 12603 + }, + { + "epoch": 0.5580509834740932, + "grad_norm": 1.6196767091751099, + "learning_rate": 9.720309859573818e-06, + "loss": 0.6186, + "step": 12604 + }, + { + "epoch": 0.558095259178907, + "grad_norm": 1.2186840772628784, + "learning_rate": 9.718765016583462e-06, + "loss": 0.5168, + "step": 12605 + }, + { + "epoch": 0.5581395348837209, + "grad_norm": 1.381540298461914, + "learning_rate": 9.71722018031017e-06, + "loss": 0.6094, + "step": 12606 + }, + { + "epoch": 0.5581838105885348, + "grad_norm": 1.3844562768936157, + "learning_rate": 9.715675350790847e-06, + "loss": 0.557, + "step": 12607 + }, + { + "epoch": 0.5582280862933486, + "grad_norm": 1.602551817893982, + "learning_rate": 9.714130528062387e-06, + "loss": 0.6656, + "step": 12608 + }, + { + "epoch": 0.5582723619981625, + "grad_norm": 1.3493483066558838, + "learning_rate": 9.712585712161685e-06, + "loss": 0.6371, + "step": 12609 + }, + { + "epoch": 0.5583166377029765, + "grad_norm": 1.1417334079742432, + "learning_rate": 9.711040903125641e-06, + "loss": 0.5539, + "step": 12610 + }, + { + "epoch": 0.5583609134077903, + "grad_norm": 1.823417067527771, + "learning_rate": 9.709496100991149e-06, + "loss": 0.7069, + "step": 12611 + }, + { + "epoch": 0.5584051891126042, + "grad_norm": 1.2452234029769897, + "learning_rate": 9.707951305795103e-06, + "loss": 0.5485, + "step": 12612 + }, + { + "epoch": 0.5584494648174181, + "grad_norm": 1.387486219406128, + "learning_rate": 9.706406517574406e-06, + "loss": 0.5638, + "step": 12613 + }, + { + "epoch": 0.5584937405222319, + "grad_norm": 3.419969081878662, + "learning_rate": 9.704861736365952e-06, + "loss": 0.632, + "step": 12614 + }, + { + "epoch": 0.5585380162270458, + "grad_norm": 1.9115869998931885, + "learning_rate": 9.703316962206634e-06, + "loss": 0.7237, + "step": 12615 + }, + { + "epoch": 0.5585822919318597, + "grad_norm": 1.6967910528182983, + "learning_rate": 9.701772195133343e-06, + "loss": 0.612, + "step": 12616 + }, + { + "epoch": 0.5586265676366736, + "grad_norm": 1.5481114387512207, + "learning_rate": 9.700227435182987e-06, + "loss": 0.6389, + "step": 12617 + }, + { + "epoch": 0.5586708433414874, + "grad_norm": 1.5490325689315796, + "learning_rate": 9.698682682392451e-06, + "loss": 0.6032, + "step": 12618 + }, + { + "epoch": 0.5587151190463013, + "grad_norm": 1.5546443462371826, + "learning_rate": 9.697137936798635e-06, + "loss": 0.6169, + "step": 12619 + }, + { + "epoch": 0.5587593947511152, + "grad_norm": 1.8600645065307617, + "learning_rate": 9.695593198438434e-06, + "loss": 0.6894, + "step": 12620 + }, + { + "epoch": 0.558803670455929, + "grad_norm": 1.5453861951828003, + "learning_rate": 9.69404846734874e-06, + "loss": 0.6546, + "step": 12621 + }, + { + "epoch": 0.558847946160743, + "grad_norm": 1.2600165605545044, + "learning_rate": 9.692503743566445e-06, + "loss": 0.5925, + "step": 12622 + }, + { + "epoch": 0.5588922218655569, + "grad_norm": 1.3115870952606201, + "learning_rate": 9.690959027128457e-06, + "loss": 0.6264, + "step": 12623 + }, + { + "epoch": 0.5589364975703707, + "grad_norm": 1.468769907951355, + "learning_rate": 9.689414318071657e-06, + "loss": 0.6497, + "step": 12624 + }, + { + "epoch": 0.5589807732751846, + "grad_norm": 1.267412781715393, + "learning_rate": 9.687869616432943e-06, + "loss": 0.6284, + "step": 12625 + }, + { + "epoch": 0.5590250489799985, + "grad_norm": 1.4714957475662231, + "learning_rate": 9.686324922249205e-06, + "loss": 0.658, + "step": 12626 + }, + { + "epoch": 0.5590693246848123, + "grad_norm": 1.4836153984069824, + "learning_rate": 9.684780235557346e-06, + "loss": 0.5354, + "step": 12627 + }, + { + "epoch": 0.5591136003896262, + "grad_norm": 1.3625630140304565, + "learning_rate": 9.683235556394254e-06, + "loss": 0.6333, + "step": 12628 + }, + { + "epoch": 0.5591578760944401, + "grad_norm": 1.2896748781204224, + "learning_rate": 9.681690884796819e-06, + "loss": 0.5621, + "step": 12629 + }, + { + "epoch": 0.5592021517992539, + "grad_norm": 1.5092378854751587, + "learning_rate": 9.680146220801943e-06, + "loss": 0.5753, + "step": 12630 + }, + { + "epoch": 0.5592464275040678, + "grad_norm": 1.5109241008758545, + "learning_rate": 9.678601564446512e-06, + "loss": 0.6567, + "step": 12631 + }, + { + "epoch": 0.5592907032088817, + "grad_norm": 1.5372142791748047, + "learning_rate": 9.677056915767418e-06, + "loss": 0.6111, + "step": 12632 + }, + { + "epoch": 0.5593349789136955, + "grad_norm": 1.4643646478652954, + "learning_rate": 9.675512274801561e-06, + "loss": 0.627, + "step": 12633 + }, + { + "epoch": 0.5593792546185095, + "grad_norm": 1.420595645904541, + "learning_rate": 9.673967641585828e-06, + "loss": 0.5748, + "step": 12634 + }, + { + "epoch": 0.5594235303233234, + "grad_norm": 1.522446870803833, + "learning_rate": 9.672423016157113e-06, + "loss": 0.6495, + "step": 12635 + }, + { + "epoch": 0.5594678060281372, + "grad_norm": 1.2893234491348267, + "learning_rate": 9.670878398552303e-06, + "loss": 0.6005, + "step": 12636 + }, + { + "epoch": 0.5595120817329511, + "grad_norm": 1.6192922592163086, + "learning_rate": 9.669333788808298e-06, + "loss": 0.6703, + "step": 12637 + }, + { + "epoch": 0.559556357437765, + "grad_norm": 1.2289706468582153, + "learning_rate": 9.66778918696199e-06, + "loss": 0.5541, + "step": 12638 + }, + { + "epoch": 0.5596006331425788, + "grad_norm": 1.326249599456787, + "learning_rate": 9.66624459305026e-06, + "loss": 0.5841, + "step": 12639 + }, + { + "epoch": 0.5596449088473927, + "grad_norm": 1.3834227323532104, + "learning_rate": 9.664700007110009e-06, + "loss": 0.6454, + "step": 12640 + }, + { + "epoch": 0.5596891845522066, + "grad_norm": 1.7361489534378052, + "learning_rate": 9.663155429178126e-06, + "loss": 0.6657, + "step": 12641 + }, + { + "epoch": 0.5597334602570204, + "grad_norm": 1.7970212697982788, + "learning_rate": 9.6616108592915e-06, + "loss": 0.6687, + "step": 12642 + }, + { + "epoch": 0.5597777359618343, + "grad_norm": 1.2836912870407104, + "learning_rate": 9.660066297487024e-06, + "loss": 0.5614, + "step": 12643 + }, + { + "epoch": 0.5598220116666482, + "grad_norm": 1.3236401081085205, + "learning_rate": 9.658521743801588e-06, + "loss": 0.6571, + "step": 12644 + }, + { + "epoch": 0.559866287371462, + "grad_norm": 1.346632957458496, + "learning_rate": 9.656977198272081e-06, + "loss": 0.5788, + "step": 12645 + }, + { + "epoch": 0.559910563076276, + "grad_norm": 1.775147795677185, + "learning_rate": 9.655432660935394e-06, + "loss": 0.743, + "step": 12646 + }, + { + "epoch": 0.5599548387810899, + "grad_norm": 1.6365703344345093, + "learning_rate": 9.65388813182842e-06, + "loss": 0.6773, + "step": 12647 + }, + { + "epoch": 0.5599991144859037, + "grad_norm": 1.2160427570343018, + "learning_rate": 9.652343610988045e-06, + "loss": 0.6671, + "step": 12648 + }, + { + "epoch": 0.5600433901907176, + "grad_norm": 1.6762638092041016, + "learning_rate": 9.650799098451154e-06, + "loss": 0.6401, + "step": 12649 + }, + { + "epoch": 0.5600876658955315, + "grad_norm": 1.743434190750122, + "learning_rate": 9.64925459425465e-06, + "loss": 0.7009, + "step": 12650 + }, + { + "epoch": 0.5601319416003453, + "grad_norm": 1.5283482074737549, + "learning_rate": 9.647710098435413e-06, + "loss": 0.6557, + "step": 12651 + }, + { + "epoch": 0.5601762173051592, + "grad_norm": 1.6306732892990112, + "learning_rate": 9.646165611030332e-06, + "loss": 0.6001, + "step": 12652 + }, + { + "epoch": 0.5602204930099731, + "grad_norm": 1.5422203540802002, + "learning_rate": 9.644621132076296e-06, + "loss": 0.5924, + "step": 12653 + }, + { + "epoch": 0.560264768714787, + "grad_norm": 1.5639387369155884, + "learning_rate": 9.643076661610197e-06, + "loss": 0.6722, + "step": 12654 + }, + { + "epoch": 0.5603090444196008, + "grad_norm": 1.3389099836349487, + "learning_rate": 9.641532199668922e-06, + "loss": 0.5936, + "step": 12655 + }, + { + "epoch": 0.5603533201244147, + "grad_norm": 1.64303719997406, + "learning_rate": 9.639987746289356e-06, + "loss": 0.5874, + "step": 12656 + }, + { + "epoch": 0.5603975958292287, + "grad_norm": 2.76340651512146, + "learning_rate": 9.638443301508391e-06, + "loss": 0.7407, + "step": 12657 + }, + { + "epoch": 0.5604418715340425, + "grad_norm": 1.313862919807434, + "learning_rate": 9.636898865362913e-06, + "loss": 0.5401, + "step": 12658 + }, + { + "epoch": 0.5604861472388564, + "grad_norm": 1.557345986366272, + "learning_rate": 9.635354437889808e-06, + "loss": 0.6671, + "step": 12659 + }, + { + "epoch": 0.5605304229436703, + "grad_norm": 1.5993916988372803, + "learning_rate": 9.633810019125968e-06, + "loss": 0.5847, + "step": 12660 + }, + { + "epoch": 0.5605746986484841, + "grad_norm": 1.4244674444198608, + "learning_rate": 9.632265609108277e-06, + "loss": 0.6035, + "step": 12661 + }, + { + "epoch": 0.560618974353298, + "grad_norm": 1.3254210948944092, + "learning_rate": 9.630721207873625e-06, + "loss": 0.5988, + "step": 12662 + }, + { + "epoch": 0.5606632500581119, + "grad_norm": 1.6443537473678589, + "learning_rate": 9.629176815458891e-06, + "loss": 0.5913, + "step": 12663 + }, + { + "epoch": 0.5607075257629257, + "grad_norm": 1.3335850238800049, + "learning_rate": 9.62763243190097e-06, + "loss": 0.5704, + "step": 12664 + }, + { + "epoch": 0.5607518014677396, + "grad_norm": 1.452990174293518, + "learning_rate": 9.626088057236745e-06, + "loss": 0.6361, + "step": 12665 + }, + { + "epoch": 0.5607960771725535, + "grad_norm": 1.581752896308899, + "learning_rate": 9.624543691503102e-06, + "loss": 0.6132, + "step": 12666 + }, + { + "epoch": 0.5608403528773673, + "grad_norm": 1.5146443843841553, + "learning_rate": 9.622999334736927e-06, + "loss": 0.6057, + "step": 12667 + }, + { + "epoch": 0.5608846285821812, + "grad_norm": 1.2745506763458252, + "learning_rate": 9.621454986975108e-06, + "loss": 0.5656, + "step": 12668 + }, + { + "epoch": 0.5609289042869952, + "grad_norm": 1.2004342079162598, + "learning_rate": 9.619910648254523e-06, + "loss": 0.5811, + "step": 12669 + }, + { + "epoch": 0.560973179991809, + "grad_norm": 1.6424005031585693, + "learning_rate": 9.61836631861207e-06, + "loss": 0.6164, + "step": 12670 + }, + { + "epoch": 0.5610174556966229, + "grad_norm": 1.5392334461212158, + "learning_rate": 9.616821998084622e-06, + "loss": 0.6304, + "step": 12671 + }, + { + "epoch": 0.5610617314014368, + "grad_norm": 1.5460495948791504, + "learning_rate": 9.615277686709071e-06, + "loss": 0.705, + "step": 12672 + }, + { + "epoch": 0.5611060071062506, + "grad_norm": 1.4014607667922974, + "learning_rate": 9.613733384522295e-06, + "loss": 0.7337, + "step": 12673 + }, + { + "epoch": 0.5611502828110645, + "grad_norm": 1.4060232639312744, + "learning_rate": 9.612189091561186e-06, + "loss": 0.6714, + "step": 12674 + }, + { + "epoch": 0.5611945585158784, + "grad_norm": 1.74529230594635, + "learning_rate": 9.610644807862625e-06, + "loss": 0.6653, + "step": 12675 + }, + { + "epoch": 0.5612388342206922, + "grad_norm": 1.2564266920089722, + "learning_rate": 9.609100533463494e-06, + "loss": 0.5793, + "step": 12676 + }, + { + "epoch": 0.5612831099255061, + "grad_norm": 1.3970832824707031, + "learning_rate": 9.607556268400679e-06, + "loss": 0.6346, + "step": 12677 + }, + { + "epoch": 0.56132738563032, + "grad_norm": 1.4482860565185547, + "learning_rate": 9.606012012711065e-06, + "loss": 0.7203, + "step": 12678 + }, + { + "epoch": 0.5613716613351338, + "grad_norm": 1.480644702911377, + "learning_rate": 9.60446776643153e-06, + "loss": 0.6456, + "step": 12679 + }, + { + "epoch": 0.5614159370399477, + "grad_norm": 1.2415529489517212, + "learning_rate": 9.602923529598963e-06, + "loss": 0.6341, + "step": 12680 + }, + { + "epoch": 0.5614602127447617, + "grad_norm": 1.2844922542572021, + "learning_rate": 9.601379302250243e-06, + "loss": 0.5892, + "step": 12681 + }, + { + "epoch": 0.5615044884495755, + "grad_norm": 1.511383295059204, + "learning_rate": 9.599835084422256e-06, + "loss": 0.5924, + "step": 12682 + }, + { + "epoch": 0.5615487641543894, + "grad_norm": 1.3886792659759521, + "learning_rate": 9.598290876151875e-06, + "loss": 0.5696, + "step": 12683 + }, + { + "epoch": 0.5615930398592033, + "grad_norm": 1.5508838891983032, + "learning_rate": 9.596746677475996e-06, + "loss": 0.7073, + "step": 12684 + }, + { + "epoch": 0.5616373155640171, + "grad_norm": 1.4246312379837036, + "learning_rate": 9.595202488431496e-06, + "loss": 0.6516, + "step": 12685 + }, + { + "epoch": 0.561681591268831, + "grad_norm": 1.4857655763626099, + "learning_rate": 9.593658309055249e-06, + "loss": 0.5813, + "step": 12686 + }, + { + "epoch": 0.5617258669736449, + "grad_norm": 1.2960702180862427, + "learning_rate": 9.592114139384146e-06, + "loss": 0.6065, + "step": 12687 + }, + { + "epoch": 0.5617701426784588, + "grad_norm": 1.4056082963943481, + "learning_rate": 9.590569979455064e-06, + "loss": 0.5636, + "step": 12688 + }, + { + "epoch": 0.5618144183832726, + "grad_norm": 1.242808222770691, + "learning_rate": 9.589025829304884e-06, + "loss": 0.5536, + "step": 12689 + }, + { + "epoch": 0.5618586940880865, + "grad_norm": 1.3342311382293701, + "learning_rate": 9.587481688970488e-06, + "loss": 0.6487, + "step": 12690 + }, + { + "epoch": 0.5619029697929004, + "grad_norm": 1.5215177536010742, + "learning_rate": 9.585937558488757e-06, + "loss": 0.6303, + "step": 12691 + }, + { + "epoch": 0.5619472454977142, + "grad_norm": 1.3859424591064453, + "learning_rate": 9.584393437896569e-06, + "loss": 0.6369, + "step": 12692 + }, + { + "epoch": 0.5619915212025282, + "grad_norm": 1.7783981561660767, + "learning_rate": 9.582849327230803e-06, + "loss": 0.6645, + "step": 12693 + }, + { + "epoch": 0.5620357969073421, + "grad_norm": 1.5649783611297607, + "learning_rate": 9.581305226528346e-06, + "loss": 0.6159, + "step": 12694 + }, + { + "epoch": 0.5620800726121559, + "grad_norm": 1.7949821949005127, + "learning_rate": 9.579761135826071e-06, + "loss": 0.6393, + "step": 12695 + }, + { + "epoch": 0.5621243483169698, + "grad_norm": 2.05828857421875, + "learning_rate": 9.578217055160855e-06, + "loss": 0.7128, + "step": 12696 + }, + { + "epoch": 0.5621686240217837, + "grad_norm": 1.4286540746688843, + "learning_rate": 9.576672984569586e-06, + "loss": 0.5903, + "step": 12697 + }, + { + "epoch": 0.5622128997265975, + "grad_norm": 1.1838221549987793, + "learning_rate": 9.575128924089138e-06, + "loss": 0.5541, + "step": 12698 + }, + { + "epoch": 0.5622571754314114, + "grad_norm": 1.6516318321228027, + "learning_rate": 9.573584873756387e-06, + "loss": 0.7159, + "step": 12699 + }, + { + "epoch": 0.5623014511362253, + "grad_norm": 1.7512235641479492, + "learning_rate": 9.572040833608216e-06, + "loss": 0.6711, + "step": 12700 + }, + { + "epoch": 0.5623457268410391, + "grad_norm": 1.4735121726989746, + "learning_rate": 9.570496803681501e-06, + "loss": 0.6562, + "step": 12701 + }, + { + "epoch": 0.562390002545853, + "grad_norm": 1.5579382181167603, + "learning_rate": 9.568952784013121e-06, + "loss": 0.6196, + "step": 12702 + }, + { + "epoch": 0.5624342782506669, + "grad_norm": 1.5572727918624878, + "learning_rate": 9.567408774639951e-06, + "loss": 0.647, + "step": 12703 + }, + { + "epoch": 0.5624785539554807, + "grad_norm": 1.607917308807373, + "learning_rate": 9.565864775598872e-06, + "loss": 0.6534, + "step": 12704 + }, + { + "epoch": 0.5625228296602947, + "grad_norm": 1.287939190864563, + "learning_rate": 9.56432078692676e-06, + "loss": 0.5906, + "step": 12705 + }, + { + "epoch": 0.5625671053651086, + "grad_norm": 1.4982588291168213, + "learning_rate": 9.562776808660485e-06, + "loss": 0.645, + "step": 12706 + }, + { + "epoch": 0.5626113810699224, + "grad_norm": 1.4272058010101318, + "learning_rate": 9.561232840836935e-06, + "loss": 0.6214, + "step": 12707 + }, + { + "epoch": 0.5626556567747363, + "grad_norm": 1.3805441856384277, + "learning_rate": 9.559688883492982e-06, + "loss": 0.6351, + "step": 12708 + }, + { + "epoch": 0.5626999324795502, + "grad_norm": 1.4594463109970093, + "learning_rate": 9.558144936665499e-06, + "loss": 0.6217, + "step": 12709 + }, + { + "epoch": 0.562744208184364, + "grad_norm": 1.3182307481765747, + "learning_rate": 9.556601000391367e-06, + "loss": 0.597, + "step": 12710 + }, + { + "epoch": 0.5627884838891779, + "grad_norm": 1.556503176689148, + "learning_rate": 9.555057074707457e-06, + "loss": 0.5927, + "step": 12711 + }, + { + "epoch": 0.5628327595939918, + "grad_norm": 1.3814945220947266, + "learning_rate": 9.553513159650648e-06, + "loss": 0.5334, + "step": 12712 + }, + { + "epoch": 0.5628770352988056, + "grad_norm": 1.2376457452774048, + "learning_rate": 9.551969255257813e-06, + "loss": 0.6282, + "step": 12713 + }, + { + "epoch": 0.5629213110036195, + "grad_norm": 1.3975790739059448, + "learning_rate": 9.550425361565828e-06, + "loss": 0.674, + "step": 12714 + }, + { + "epoch": 0.5629655867084334, + "grad_norm": 1.438036322593689, + "learning_rate": 9.548881478611568e-06, + "loss": 0.5725, + "step": 12715 + }, + { + "epoch": 0.5630098624132472, + "grad_norm": 1.457696557044983, + "learning_rate": 9.5473376064319e-06, + "loss": 0.5543, + "step": 12716 + }, + { + "epoch": 0.5630541381180612, + "grad_norm": 1.2692103385925293, + "learning_rate": 9.545793745063714e-06, + "loss": 0.5947, + "step": 12717 + }, + { + "epoch": 0.5630984138228751, + "grad_norm": 1.6971399784088135, + "learning_rate": 9.544249894543872e-06, + "loss": 0.6233, + "step": 12718 + }, + { + "epoch": 0.5631426895276889, + "grad_norm": 1.9214359521865845, + "learning_rate": 9.542706054909245e-06, + "loss": 0.6762, + "step": 12719 + }, + { + "epoch": 0.5631869652325028, + "grad_norm": 1.2365233898162842, + "learning_rate": 9.541162226196717e-06, + "loss": 0.608, + "step": 12720 + }, + { + "epoch": 0.5632312409373167, + "grad_norm": 1.862368106842041, + "learning_rate": 9.539618408443153e-06, + "loss": 0.6432, + "step": 12721 + }, + { + "epoch": 0.5632755166421305, + "grad_norm": 1.2184276580810547, + "learning_rate": 9.53807460168543e-06, + "loss": 0.5673, + "step": 12722 + }, + { + "epoch": 0.5633197923469444, + "grad_norm": 1.4686341285705566, + "learning_rate": 9.536530805960418e-06, + "loss": 0.6424, + "step": 12723 + }, + { + "epoch": 0.5633640680517583, + "grad_norm": 1.4090168476104736, + "learning_rate": 9.534987021304992e-06, + "loss": 0.5964, + "step": 12724 + }, + { + "epoch": 0.5634083437565722, + "grad_norm": 1.4244016408920288, + "learning_rate": 9.533443247756021e-06, + "loss": 0.5815, + "step": 12725 + }, + { + "epoch": 0.563452619461386, + "grad_norm": 1.8690459728240967, + "learning_rate": 9.531899485350377e-06, + "loss": 0.7008, + "step": 12726 + }, + { + "epoch": 0.5634968951661999, + "grad_norm": 1.4382418394088745, + "learning_rate": 9.530355734124934e-06, + "loss": 0.5821, + "step": 12727 + }, + { + "epoch": 0.5635411708710139, + "grad_norm": 1.334553599357605, + "learning_rate": 9.528811994116562e-06, + "loss": 0.5794, + "step": 12728 + }, + { + "epoch": 0.5635854465758277, + "grad_norm": 1.5172796249389648, + "learning_rate": 9.52726826536213e-06, + "loss": 0.6082, + "step": 12729 + }, + { + "epoch": 0.5636297222806416, + "grad_norm": 1.2462823390960693, + "learning_rate": 9.525724547898512e-06, + "loss": 0.5081, + "step": 12730 + }, + { + "epoch": 0.5636739979854555, + "grad_norm": 1.4454319477081299, + "learning_rate": 9.524180841762577e-06, + "loss": 0.6983, + "step": 12731 + }, + { + "epoch": 0.5637182736902693, + "grad_norm": 1.4056198596954346, + "learning_rate": 9.522637146991198e-06, + "loss": 0.6563, + "step": 12732 + }, + { + "epoch": 0.5637625493950832, + "grad_norm": 1.6052864789962769, + "learning_rate": 9.521093463621238e-06, + "loss": 0.6567, + "step": 12733 + }, + { + "epoch": 0.5638068250998971, + "grad_norm": 1.4927852153778076, + "learning_rate": 9.519549791689571e-06, + "loss": 0.6544, + "step": 12734 + }, + { + "epoch": 0.5638511008047109, + "grad_norm": 1.5088578462600708, + "learning_rate": 9.518006131233068e-06, + "loss": 0.6342, + "step": 12735 + }, + { + "epoch": 0.5638953765095248, + "grad_norm": 1.3718817234039307, + "learning_rate": 9.516462482288594e-06, + "loss": 0.6098, + "step": 12736 + }, + { + "epoch": 0.5639396522143387, + "grad_norm": 1.557552456855774, + "learning_rate": 9.514918844893024e-06, + "loss": 0.6651, + "step": 12737 + }, + { + "epoch": 0.5639839279191525, + "grad_norm": 1.6103101968765259, + "learning_rate": 9.513375219083218e-06, + "loss": 0.7044, + "step": 12738 + }, + { + "epoch": 0.5640282036239664, + "grad_norm": 1.4791477918624878, + "learning_rate": 9.511831604896047e-06, + "loss": 0.5887, + "step": 12739 + }, + { + "epoch": 0.5640724793287804, + "grad_norm": 1.4443469047546387, + "learning_rate": 9.510288002368385e-06, + "loss": 0.6705, + "step": 12740 + }, + { + "epoch": 0.5641167550335942, + "grad_norm": 2.365593910217285, + "learning_rate": 9.508744411537098e-06, + "loss": 0.6371, + "step": 12741 + }, + { + "epoch": 0.5641610307384081, + "grad_norm": 1.7257499694824219, + "learning_rate": 9.507200832439047e-06, + "loss": 0.6698, + "step": 12742 + }, + { + "epoch": 0.564205306443222, + "grad_norm": 1.6385070085525513, + "learning_rate": 9.5056572651111e-06, + "loss": 0.7135, + "step": 12743 + }, + { + "epoch": 0.5642495821480358, + "grad_norm": 1.074476718902588, + "learning_rate": 9.504113709590129e-06, + "loss": 0.57, + "step": 12744 + }, + { + "epoch": 0.5642938578528497, + "grad_norm": 1.2752379179000854, + "learning_rate": 9.502570165913e-06, + "loss": 0.6428, + "step": 12745 + }, + { + "epoch": 0.5643381335576636, + "grad_norm": 1.2057424783706665, + "learning_rate": 9.501026634116575e-06, + "loss": 0.5228, + "step": 12746 + }, + { + "epoch": 0.5643824092624774, + "grad_norm": 1.2834010124206543, + "learning_rate": 9.499483114237726e-06, + "loss": 0.6032, + "step": 12747 + }, + { + "epoch": 0.5644266849672913, + "grad_norm": 1.4831278324127197, + "learning_rate": 9.497939606313313e-06, + "loss": 0.6252, + "step": 12748 + }, + { + "epoch": 0.5644709606721052, + "grad_norm": 1.5695629119873047, + "learning_rate": 9.4963961103802e-06, + "loss": 0.6411, + "step": 12749 + }, + { + "epoch": 0.564515236376919, + "grad_norm": 1.148321509361267, + "learning_rate": 9.494852626475263e-06, + "loss": 0.5071, + "step": 12750 + }, + { + "epoch": 0.5645595120817329, + "grad_norm": 1.6860405206680298, + "learning_rate": 9.493309154635357e-06, + "loss": 0.6741, + "step": 12751 + }, + { + "epoch": 0.5646037877865469, + "grad_norm": 1.319362759590149, + "learning_rate": 9.49176569489735e-06, + "loss": 0.6256, + "step": 12752 + }, + { + "epoch": 0.5646480634913607, + "grad_norm": 1.3747252225875854, + "learning_rate": 9.4902222472981e-06, + "loss": 0.5943, + "step": 12753 + }, + { + "epoch": 0.5646923391961746, + "grad_norm": 1.512394905090332, + "learning_rate": 9.488678811874483e-06, + "loss": 0.7676, + "step": 12754 + }, + { + "epoch": 0.5647366149009885, + "grad_norm": 1.4699854850769043, + "learning_rate": 9.487135388663353e-06, + "loss": 0.6322, + "step": 12755 + }, + { + "epoch": 0.5647808906058023, + "grad_norm": 1.4859869480133057, + "learning_rate": 9.485591977701577e-06, + "loss": 0.5854, + "step": 12756 + }, + { + "epoch": 0.5648251663106162, + "grad_norm": 1.5373352766036987, + "learning_rate": 9.48404857902602e-06, + "loss": 0.6737, + "step": 12757 + }, + { + "epoch": 0.5648694420154301, + "grad_norm": 1.3847107887268066, + "learning_rate": 9.48250519267354e-06, + "loss": 0.5527, + "step": 12758 + }, + { + "epoch": 0.564913717720244, + "grad_norm": 1.5348966121673584, + "learning_rate": 9.480961818681004e-06, + "loss": 0.6656, + "step": 12759 + }, + { + "epoch": 0.5649579934250578, + "grad_norm": 1.2530879974365234, + "learning_rate": 9.47941845708527e-06, + "loss": 0.5731, + "step": 12760 + }, + { + "epoch": 0.5650022691298717, + "grad_norm": 1.280592918395996, + "learning_rate": 9.477875107923205e-06, + "loss": 0.589, + "step": 12761 + }, + { + "epoch": 0.5650465448346856, + "grad_norm": 1.5150909423828125, + "learning_rate": 9.476331771231668e-06, + "loss": 0.6601, + "step": 12762 + }, + { + "epoch": 0.5650908205394994, + "grad_norm": 1.4760104417800903, + "learning_rate": 9.474788447047516e-06, + "loss": 0.5942, + "step": 12763 + }, + { + "epoch": 0.5651350962443134, + "grad_norm": 1.5597110986709595, + "learning_rate": 9.473245135407618e-06, + "loss": 0.6272, + "step": 12764 + }, + { + "epoch": 0.5651793719491273, + "grad_norm": 1.514718770980835, + "learning_rate": 9.471701836348833e-06, + "loss": 0.6684, + "step": 12765 + }, + { + "epoch": 0.5652236476539411, + "grad_norm": 1.1823867559432983, + "learning_rate": 9.470158549908014e-06, + "loss": 0.6051, + "step": 12766 + }, + { + "epoch": 0.565267923358755, + "grad_norm": 2.0614733695983887, + "learning_rate": 9.46861527612203e-06, + "loss": 0.6932, + "step": 12767 + }, + { + "epoch": 0.5653121990635689, + "grad_norm": 1.8113305568695068, + "learning_rate": 9.467072015027738e-06, + "loss": 0.7244, + "step": 12768 + }, + { + "epoch": 0.5653564747683827, + "grad_norm": 1.526769995689392, + "learning_rate": 9.465528766661997e-06, + "loss": 0.6315, + "step": 12769 + }, + { + "epoch": 0.5654007504731966, + "grad_norm": 1.5534899234771729, + "learning_rate": 9.463985531061664e-06, + "loss": 0.6471, + "step": 12770 + }, + { + "epoch": 0.5654450261780105, + "grad_norm": 1.3699885606765747, + "learning_rate": 9.4624423082636e-06, + "loss": 0.5746, + "step": 12771 + }, + { + "epoch": 0.5654893018828243, + "grad_norm": 1.7135579586029053, + "learning_rate": 9.460899098304667e-06, + "loss": 0.6777, + "step": 12772 + }, + { + "epoch": 0.5655335775876382, + "grad_norm": 1.5763922929763794, + "learning_rate": 9.459355901221716e-06, + "loss": 0.6134, + "step": 12773 + }, + { + "epoch": 0.5655778532924521, + "grad_norm": 1.5036200284957886, + "learning_rate": 9.457812717051614e-06, + "loss": 0.6435, + "step": 12774 + }, + { + "epoch": 0.5656221289972659, + "grad_norm": 1.6792316436767578, + "learning_rate": 9.45626954583121e-06, + "loss": 0.7233, + "step": 12775 + }, + { + "epoch": 0.5656664047020799, + "grad_norm": 1.4701688289642334, + "learning_rate": 9.454726387597363e-06, + "loss": 0.5708, + "step": 12776 + }, + { + "epoch": 0.5657106804068938, + "grad_norm": 1.4064650535583496, + "learning_rate": 9.453183242386936e-06, + "loss": 0.5692, + "step": 12777 + }, + { + "epoch": 0.5657549561117076, + "grad_norm": 1.6583386659622192, + "learning_rate": 9.451640110236783e-06, + "loss": 0.6924, + "step": 12778 + }, + { + "epoch": 0.5657992318165215, + "grad_norm": 1.6250265836715698, + "learning_rate": 9.450096991183759e-06, + "loss": 0.6076, + "step": 12779 + }, + { + "epoch": 0.5658435075213354, + "grad_norm": 1.3058782815933228, + "learning_rate": 9.448553885264717e-06, + "loss": 0.6628, + "step": 12780 + }, + { + "epoch": 0.5658877832261492, + "grad_norm": 1.3696297407150269, + "learning_rate": 9.44701079251652e-06, + "loss": 0.6262, + "step": 12781 + }, + { + "epoch": 0.5659320589309631, + "grad_norm": 1.8190665245056152, + "learning_rate": 9.44546771297602e-06, + "loss": 0.6238, + "step": 12782 + }, + { + "epoch": 0.565976334635777, + "grad_norm": 1.513519048690796, + "learning_rate": 9.44392464668007e-06, + "loss": 0.631, + "step": 12783 + }, + { + "epoch": 0.5660206103405908, + "grad_norm": 1.8398102521896362, + "learning_rate": 9.442381593665529e-06, + "loss": 0.6215, + "step": 12784 + }, + { + "epoch": 0.5660648860454047, + "grad_norm": 1.4814625978469849, + "learning_rate": 9.44083855396925e-06, + "loss": 0.611, + "step": 12785 + }, + { + "epoch": 0.5661091617502186, + "grad_norm": 1.49530029296875, + "learning_rate": 9.439295527628083e-06, + "loss": 0.6389, + "step": 12786 + }, + { + "epoch": 0.5661534374550324, + "grad_norm": 1.640135645866394, + "learning_rate": 9.437752514678888e-06, + "loss": 0.6344, + "step": 12787 + }, + { + "epoch": 0.5661977131598463, + "grad_norm": 1.5923309326171875, + "learning_rate": 9.436209515158517e-06, + "loss": 0.6434, + "step": 12788 + }, + { + "epoch": 0.5662419888646603, + "grad_norm": 1.3512502908706665, + "learning_rate": 9.434666529103825e-06, + "loss": 0.566, + "step": 12789 + }, + { + "epoch": 0.5662862645694741, + "grad_norm": 1.618605613708496, + "learning_rate": 9.433123556551657e-06, + "loss": 0.6307, + "step": 12790 + }, + { + "epoch": 0.566330540274288, + "grad_norm": 2.3124382495880127, + "learning_rate": 9.431580597538876e-06, + "loss": 0.749, + "step": 12791 + }, + { + "epoch": 0.5663748159791019, + "grad_norm": 1.705568552017212, + "learning_rate": 9.430037652102328e-06, + "loss": 0.6555, + "step": 12792 + }, + { + "epoch": 0.5664190916839158, + "grad_norm": 1.594280481338501, + "learning_rate": 9.428494720278866e-06, + "loss": 0.6519, + "step": 12793 + }, + { + "epoch": 0.5664633673887296, + "grad_norm": 1.5409767627716064, + "learning_rate": 9.426951802105343e-06, + "loss": 0.6085, + "step": 12794 + }, + { + "epoch": 0.5665076430935435, + "grad_norm": 1.708676815032959, + "learning_rate": 9.42540889761861e-06, + "loss": 0.6431, + "step": 12795 + }, + { + "epoch": 0.5665519187983574, + "grad_norm": 2.1826834678649902, + "learning_rate": 9.423866006855514e-06, + "loss": 0.685, + "step": 12796 + }, + { + "epoch": 0.5665961945031712, + "grad_norm": 1.276574969291687, + "learning_rate": 9.422323129852915e-06, + "loss": 0.5911, + "step": 12797 + }, + { + "epoch": 0.5666404702079851, + "grad_norm": 1.9153786897659302, + "learning_rate": 9.420780266647656e-06, + "loss": 0.6851, + "step": 12798 + }, + { + "epoch": 0.566684745912799, + "grad_norm": 1.4965895414352417, + "learning_rate": 9.41923741727659e-06, + "loss": 0.6134, + "step": 12799 + }, + { + "epoch": 0.5667290216176128, + "grad_norm": 1.8016082048416138, + "learning_rate": 9.417694581776558e-06, + "loss": 0.6563, + "step": 12800 + }, + { + "epoch": 0.5667732973224268, + "grad_norm": 1.4937080144882202, + "learning_rate": 9.416151760184424e-06, + "loss": 0.6586, + "step": 12801 + }, + { + "epoch": 0.5668175730272407, + "grad_norm": 1.9144271612167358, + "learning_rate": 9.414608952537028e-06, + "loss": 0.7461, + "step": 12802 + }, + { + "epoch": 0.5668618487320545, + "grad_norm": 1.4195939302444458, + "learning_rate": 9.413066158871219e-06, + "loss": 0.5809, + "step": 12803 + }, + { + "epoch": 0.5669061244368684, + "grad_norm": 1.3131210803985596, + "learning_rate": 9.41152337922385e-06, + "loss": 0.6093, + "step": 12804 + }, + { + "epoch": 0.5669504001416823, + "grad_norm": 1.900409460067749, + "learning_rate": 9.409980613631763e-06, + "loss": 0.7271, + "step": 12805 + }, + { + "epoch": 0.5669946758464961, + "grad_norm": 1.9114702939987183, + "learning_rate": 9.40843786213181e-06, + "loss": 0.6562, + "step": 12806 + }, + { + "epoch": 0.56703895155131, + "grad_norm": 1.3284372091293335, + "learning_rate": 9.406895124760836e-06, + "loss": 0.6248, + "step": 12807 + }, + { + "epoch": 0.5670832272561239, + "grad_norm": 1.7020941972732544, + "learning_rate": 9.405352401555691e-06, + "loss": 0.593, + "step": 12808 + }, + { + "epoch": 0.5671275029609377, + "grad_norm": 1.352380394935608, + "learning_rate": 9.403809692553219e-06, + "loss": 0.566, + "step": 12809 + }, + { + "epoch": 0.5671717786657516, + "grad_norm": 1.2981656789779663, + "learning_rate": 9.402266997790263e-06, + "loss": 0.6177, + "step": 12810 + }, + { + "epoch": 0.5672160543705655, + "grad_norm": 1.431074857711792, + "learning_rate": 9.400724317303677e-06, + "loss": 0.6796, + "step": 12811 + }, + { + "epoch": 0.5672603300753793, + "grad_norm": 1.738670825958252, + "learning_rate": 9.399181651130308e-06, + "loss": 0.6884, + "step": 12812 + }, + { + "epoch": 0.5673046057801933, + "grad_norm": 1.4215552806854248, + "learning_rate": 9.397638999306984e-06, + "loss": 0.629, + "step": 12813 + }, + { + "epoch": 0.5673488814850072, + "grad_norm": 1.698531985282898, + "learning_rate": 9.39609636187057e-06, + "loss": 0.7411, + "step": 12814 + }, + { + "epoch": 0.567393157189821, + "grad_norm": 1.3078655004501343, + "learning_rate": 9.394553738857902e-06, + "loss": 0.6452, + "step": 12815 + }, + { + "epoch": 0.5674374328946349, + "grad_norm": 1.2948108911514282, + "learning_rate": 9.393011130305824e-06, + "loss": 0.5851, + "step": 12816 + }, + { + "epoch": 0.5674817085994488, + "grad_norm": 1.3257979154586792, + "learning_rate": 9.39146853625118e-06, + "loss": 0.605, + "step": 12817 + }, + { + "epoch": 0.5675259843042626, + "grad_norm": 1.2664695978164673, + "learning_rate": 9.389925956730817e-06, + "loss": 0.5666, + "step": 12818 + }, + { + "epoch": 0.5675702600090765, + "grad_norm": 1.3458672761917114, + "learning_rate": 9.388383391781576e-06, + "loss": 0.6102, + "step": 12819 + }, + { + "epoch": 0.5676145357138904, + "grad_norm": 1.714717984199524, + "learning_rate": 9.386840841440294e-06, + "loss": 0.6528, + "step": 12820 + }, + { + "epoch": 0.5676588114187042, + "grad_norm": 1.522823691368103, + "learning_rate": 9.385298305743826e-06, + "loss": 0.6537, + "step": 12821 + }, + { + "epoch": 0.5677030871235181, + "grad_norm": 1.3799017667770386, + "learning_rate": 9.383755784729007e-06, + "loss": 0.5668, + "step": 12822 + }, + { + "epoch": 0.567747362828332, + "grad_norm": 1.3969324827194214, + "learning_rate": 9.382213278432673e-06, + "loss": 0.628, + "step": 12823 + }, + { + "epoch": 0.5677916385331458, + "grad_norm": 1.590987205505371, + "learning_rate": 9.380670786891676e-06, + "loss": 0.6695, + "step": 12824 + }, + { + "epoch": 0.5678359142379598, + "grad_norm": 1.4859473705291748, + "learning_rate": 9.379128310142855e-06, + "loss": 0.5947, + "step": 12825 + }, + { + "epoch": 0.5678801899427737, + "grad_norm": 1.5047223567962646, + "learning_rate": 9.377585848223044e-06, + "loss": 0.6627, + "step": 12826 + }, + { + "epoch": 0.5679244656475875, + "grad_norm": 1.7171083688735962, + "learning_rate": 9.376043401169092e-06, + "loss": 0.5814, + "step": 12827 + }, + { + "epoch": 0.5679687413524014, + "grad_norm": 1.806448221206665, + "learning_rate": 9.374500969017836e-06, + "loss": 0.6204, + "step": 12828 + }, + { + "epoch": 0.5680130170572153, + "grad_norm": 1.3779219388961792, + "learning_rate": 9.372958551806114e-06, + "loss": 0.6384, + "step": 12829 + }, + { + "epoch": 0.5680572927620292, + "grad_norm": 1.461574912071228, + "learning_rate": 9.371416149570764e-06, + "loss": 0.6324, + "step": 12830 + }, + { + "epoch": 0.568101568466843, + "grad_norm": 1.5329580307006836, + "learning_rate": 9.369873762348629e-06, + "loss": 0.5967, + "step": 12831 + }, + { + "epoch": 0.5681458441716569, + "grad_norm": 1.5077295303344727, + "learning_rate": 9.368331390176548e-06, + "loss": 0.5901, + "step": 12832 + }, + { + "epoch": 0.5681901198764708, + "grad_norm": 1.3670063018798828, + "learning_rate": 9.366789033091353e-06, + "loss": 0.5704, + "step": 12833 + }, + { + "epoch": 0.5682343955812846, + "grad_norm": 1.4665131568908691, + "learning_rate": 9.36524669112989e-06, + "loss": 0.647, + "step": 12834 + }, + { + "epoch": 0.5682786712860985, + "grad_norm": 1.504347324371338, + "learning_rate": 9.363704364328992e-06, + "loss": 0.6333, + "step": 12835 + }, + { + "epoch": 0.5683229469909125, + "grad_norm": 1.4862293004989624, + "learning_rate": 9.362162052725495e-06, + "loss": 0.612, + "step": 12836 + }, + { + "epoch": 0.5683672226957263, + "grad_norm": 1.423564076423645, + "learning_rate": 9.360619756356242e-06, + "loss": 0.6149, + "step": 12837 + }, + { + "epoch": 0.5684114984005402, + "grad_norm": 1.185184121131897, + "learning_rate": 9.359077475258064e-06, + "loss": 0.5918, + "step": 12838 + }, + { + "epoch": 0.5684557741053541, + "grad_norm": 1.5378353595733643, + "learning_rate": 9.3575352094678e-06, + "loss": 0.6156, + "step": 12839 + }, + { + "epoch": 0.5685000498101679, + "grad_norm": 1.4397040605545044, + "learning_rate": 9.355992959022281e-06, + "loss": 0.6268, + "step": 12840 + }, + { + "epoch": 0.5685443255149818, + "grad_norm": 1.9614055156707764, + "learning_rate": 9.354450723958348e-06, + "loss": 0.7425, + "step": 12841 + }, + { + "epoch": 0.5685886012197957, + "grad_norm": 1.4644051790237427, + "learning_rate": 9.352908504312834e-06, + "loss": 0.6054, + "step": 12842 + }, + { + "epoch": 0.5686328769246095, + "grad_norm": 1.4341751337051392, + "learning_rate": 9.351366300122569e-06, + "loss": 0.5904, + "step": 12843 + }, + { + "epoch": 0.5686771526294234, + "grad_norm": 1.4622681140899658, + "learning_rate": 9.3498241114244e-06, + "loss": 0.6378, + "step": 12844 + }, + { + "epoch": 0.5687214283342373, + "grad_norm": 1.7814728021621704, + "learning_rate": 9.348281938255148e-06, + "loss": 0.6379, + "step": 12845 + }, + { + "epoch": 0.5687657040390511, + "grad_norm": 1.7064993381500244, + "learning_rate": 9.346739780651648e-06, + "loss": 0.6725, + "step": 12846 + }, + { + "epoch": 0.568809979743865, + "grad_norm": 1.9283905029296875, + "learning_rate": 9.345197638650742e-06, + "loss": 0.7295, + "step": 12847 + }, + { + "epoch": 0.568854255448679, + "grad_norm": 1.5577468872070312, + "learning_rate": 9.343655512289255e-06, + "loss": 0.6035, + "step": 12848 + }, + { + "epoch": 0.5688985311534928, + "grad_norm": 1.7606405019760132, + "learning_rate": 9.342113401604023e-06, + "loss": 0.7305, + "step": 12849 + }, + { + "epoch": 0.5689428068583067, + "grad_norm": 1.5896672010421753, + "learning_rate": 9.340571306631875e-06, + "loss": 0.6417, + "step": 12850 + }, + { + "epoch": 0.5689870825631206, + "grad_norm": 1.5430402755737305, + "learning_rate": 9.339029227409646e-06, + "loss": 0.6457, + "step": 12851 + }, + { + "epoch": 0.5690313582679344, + "grad_norm": 2.3431999683380127, + "learning_rate": 9.337487163974165e-06, + "loss": 0.6671, + "step": 12852 + }, + { + "epoch": 0.5690756339727483, + "grad_norm": 1.4273821115493774, + "learning_rate": 9.335945116362262e-06, + "loss": 0.6168, + "step": 12853 + }, + { + "epoch": 0.5691199096775622, + "grad_norm": 1.3431798219680786, + "learning_rate": 9.334403084610773e-06, + "loss": 0.6256, + "step": 12854 + }, + { + "epoch": 0.569164185382376, + "grad_norm": 1.2124515771865845, + "learning_rate": 9.332861068756522e-06, + "loss": 0.5651, + "step": 12855 + }, + { + "epoch": 0.5692084610871899, + "grad_norm": 1.7649351358413696, + "learning_rate": 9.331319068836338e-06, + "loss": 0.6292, + "step": 12856 + }, + { + "epoch": 0.5692527367920038, + "grad_norm": 1.6563688516616821, + "learning_rate": 9.329777084887057e-06, + "loss": 0.6584, + "step": 12857 + }, + { + "epoch": 0.5692970124968176, + "grad_norm": 1.8047300577163696, + "learning_rate": 9.328235116945506e-06, + "loss": 0.6984, + "step": 12858 + }, + { + "epoch": 0.5693412882016315, + "grad_norm": 1.4395489692687988, + "learning_rate": 9.326693165048511e-06, + "loss": 0.6158, + "step": 12859 + }, + { + "epoch": 0.5693855639064455, + "grad_norm": 1.683053731918335, + "learning_rate": 9.325151229232899e-06, + "loss": 0.6222, + "step": 12860 + }, + { + "epoch": 0.5694298396112593, + "grad_norm": 1.5520637035369873, + "learning_rate": 9.323609309535503e-06, + "loss": 0.6418, + "step": 12861 + }, + { + "epoch": 0.5694741153160732, + "grad_norm": 1.4981144666671753, + "learning_rate": 9.322067405993149e-06, + "loss": 0.6209, + "step": 12862 + }, + { + "epoch": 0.5695183910208871, + "grad_norm": 1.613255262374878, + "learning_rate": 9.320525518642661e-06, + "loss": 0.6086, + "step": 12863 + }, + { + "epoch": 0.569562666725701, + "grad_norm": 1.6442065238952637, + "learning_rate": 9.318983647520869e-06, + "loss": 0.6175, + "step": 12864 + }, + { + "epoch": 0.5696069424305148, + "grad_norm": 1.3816368579864502, + "learning_rate": 9.317441792664598e-06, + "loss": 0.6467, + "step": 12865 + }, + { + "epoch": 0.5696512181353287, + "grad_norm": 1.291847586631775, + "learning_rate": 9.315899954110674e-06, + "loss": 0.5776, + "step": 12866 + }, + { + "epoch": 0.5696954938401426, + "grad_norm": 1.3010421991348267, + "learning_rate": 9.314358131895919e-06, + "loss": 0.5749, + "step": 12867 + }, + { + "epoch": 0.5697397695449564, + "grad_norm": 1.910098671913147, + "learning_rate": 9.312816326057168e-06, + "loss": 0.6994, + "step": 12868 + }, + { + "epoch": 0.5697840452497703, + "grad_norm": 1.731392741203308, + "learning_rate": 9.311274536631236e-06, + "loss": 0.5887, + "step": 12869 + }, + { + "epoch": 0.5698283209545842, + "grad_norm": 1.2645727396011353, + "learning_rate": 9.309732763654948e-06, + "loss": 0.5748, + "step": 12870 + }, + { + "epoch": 0.569872596659398, + "grad_norm": 1.4479954242706299, + "learning_rate": 9.308191007165135e-06, + "loss": 0.5733, + "step": 12871 + }, + { + "epoch": 0.569916872364212, + "grad_norm": 1.2863188982009888, + "learning_rate": 9.306649267198616e-06, + "loss": 0.5642, + "step": 12872 + }, + { + "epoch": 0.5699611480690259, + "grad_norm": 1.590202808380127, + "learning_rate": 9.305107543792214e-06, + "loss": 0.6216, + "step": 12873 + }, + { + "epoch": 0.5700054237738397, + "grad_norm": 1.4748305082321167, + "learning_rate": 9.303565836982753e-06, + "loss": 0.5881, + "step": 12874 + }, + { + "epoch": 0.5700496994786536, + "grad_norm": 1.3591831922531128, + "learning_rate": 9.302024146807057e-06, + "loss": 0.592, + "step": 12875 + }, + { + "epoch": 0.5700939751834675, + "grad_norm": 1.5150197744369507, + "learning_rate": 9.300482473301942e-06, + "loss": 0.625, + "step": 12876 + }, + { + "epoch": 0.5701382508882813, + "grad_norm": 1.6208873987197876, + "learning_rate": 9.298940816504234e-06, + "loss": 0.6692, + "step": 12877 + }, + { + "epoch": 0.5701825265930952, + "grad_norm": 1.5707024335861206, + "learning_rate": 9.297399176450757e-06, + "loss": 0.5809, + "step": 12878 + }, + { + "epoch": 0.5702268022979091, + "grad_norm": 1.6407138109207153, + "learning_rate": 9.295857553178327e-06, + "loss": 0.7272, + "step": 12879 + }, + { + "epoch": 0.5702710780027229, + "grad_norm": 1.7007657289505005, + "learning_rate": 9.294315946723763e-06, + "loss": 0.6997, + "step": 12880 + }, + { + "epoch": 0.5703153537075368, + "grad_norm": 1.4542807340621948, + "learning_rate": 9.29277435712389e-06, + "loss": 0.6232, + "step": 12881 + }, + { + "epoch": 0.5703596294123507, + "grad_norm": 1.4891436100006104, + "learning_rate": 9.291232784415528e-06, + "loss": 0.6568, + "step": 12882 + }, + { + "epoch": 0.5704039051171645, + "grad_norm": 1.2922106981277466, + "learning_rate": 9.28969122863549e-06, + "loss": 0.5618, + "step": 12883 + }, + { + "epoch": 0.5704481808219785, + "grad_norm": 1.8047114610671997, + "learning_rate": 9.2881496898206e-06, + "loss": 0.6435, + "step": 12884 + }, + { + "epoch": 0.5704924565267924, + "grad_norm": 1.4228847026824951, + "learning_rate": 9.286608168007678e-06, + "loss": 0.5782, + "step": 12885 + }, + { + "epoch": 0.5705367322316062, + "grad_norm": 1.5015859603881836, + "learning_rate": 9.285066663233539e-06, + "loss": 0.6189, + "step": 12886 + }, + { + "epoch": 0.5705810079364201, + "grad_norm": 1.5474802255630493, + "learning_rate": 9.283525175534997e-06, + "loss": 0.6367, + "step": 12887 + }, + { + "epoch": 0.570625283641234, + "grad_norm": 1.3618531227111816, + "learning_rate": 9.281983704948873e-06, + "loss": 0.5823, + "step": 12888 + }, + { + "epoch": 0.5706695593460478, + "grad_norm": 1.4585577249526978, + "learning_rate": 9.280442251511987e-06, + "loss": 0.6087, + "step": 12889 + }, + { + "epoch": 0.5707138350508617, + "grad_norm": 1.4816840887069702, + "learning_rate": 9.278900815261147e-06, + "loss": 0.6187, + "step": 12890 + }, + { + "epoch": 0.5707581107556756, + "grad_norm": 1.2018808126449585, + "learning_rate": 9.277359396233176e-06, + "loss": 0.6085, + "step": 12891 + }, + { + "epoch": 0.5708023864604894, + "grad_norm": 1.412441611289978, + "learning_rate": 9.275817994464892e-06, + "loss": 0.6313, + "step": 12892 + }, + { + "epoch": 0.5708466621653033, + "grad_norm": 1.4236749410629272, + "learning_rate": 9.274276609993097e-06, + "loss": 0.6265, + "step": 12893 + }, + { + "epoch": 0.5708909378701172, + "grad_norm": 1.5316011905670166, + "learning_rate": 9.27273524285462e-06, + "loss": 0.6082, + "step": 12894 + }, + { + "epoch": 0.570935213574931, + "grad_norm": 1.3609436750411987, + "learning_rate": 9.27119389308627e-06, + "loss": 0.5666, + "step": 12895 + }, + { + "epoch": 0.570979489279745, + "grad_norm": 1.3883178234100342, + "learning_rate": 9.26965256072486e-06, + "loss": 0.5574, + "step": 12896 + }, + { + "epoch": 0.5710237649845589, + "grad_norm": 1.2627094984054565, + "learning_rate": 9.268111245807202e-06, + "loss": 0.6601, + "step": 12897 + }, + { + "epoch": 0.5710680406893727, + "grad_norm": 1.4342647790908813, + "learning_rate": 9.266569948370113e-06, + "loss": 0.6419, + "step": 12898 + }, + { + "epoch": 0.5711123163941866, + "grad_norm": 1.4199631214141846, + "learning_rate": 9.265028668450403e-06, + "loss": 0.6259, + "step": 12899 + }, + { + "epoch": 0.5711565920990005, + "grad_norm": 1.6168733835220337, + "learning_rate": 9.263487406084881e-06, + "loss": 0.6678, + "step": 12900 + }, + { + "epoch": 0.5712008678038144, + "grad_norm": 1.2871179580688477, + "learning_rate": 9.261946161310367e-06, + "loss": 0.5701, + "step": 12901 + }, + { + "epoch": 0.5712451435086282, + "grad_norm": 1.4786403179168701, + "learning_rate": 9.260404934163669e-06, + "loss": 0.5693, + "step": 12902 + }, + { + "epoch": 0.5712894192134421, + "grad_norm": 1.8381379842758179, + "learning_rate": 9.25886372468159e-06, + "loss": 0.6277, + "step": 12903 + }, + { + "epoch": 0.571333694918256, + "grad_norm": 1.5029376745224, + "learning_rate": 9.257322532900954e-06, + "loss": 0.5968, + "step": 12904 + }, + { + "epoch": 0.5713779706230698, + "grad_norm": 1.4169971942901611, + "learning_rate": 9.255781358858563e-06, + "loss": 0.563, + "step": 12905 + }, + { + "epoch": 0.5714222463278837, + "grad_norm": 2.6421103477478027, + "learning_rate": 9.25424020259123e-06, + "loss": 0.6715, + "step": 12906 + }, + { + "epoch": 0.5714665220326977, + "grad_norm": 1.430828332901001, + "learning_rate": 9.252699064135759e-06, + "loss": 0.6923, + "step": 12907 + }, + { + "epoch": 0.5715107977375115, + "grad_norm": 2.151855707168579, + "learning_rate": 9.251157943528965e-06, + "loss": 0.7649, + "step": 12908 + }, + { + "epoch": 0.5715550734423254, + "grad_norm": 1.5581470727920532, + "learning_rate": 9.249616840807652e-06, + "loss": 0.6335, + "step": 12909 + }, + { + "epoch": 0.5715993491471393, + "grad_norm": 1.509814739227295, + "learning_rate": 9.24807575600863e-06, + "loss": 0.6017, + "step": 12910 + }, + { + "epoch": 0.5716436248519531, + "grad_norm": 1.4070249795913696, + "learning_rate": 9.246534689168708e-06, + "loss": 0.6233, + "step": 12911 + }, + { + "epoch": 0.571687900556767, + "grad_norm": 1.3152034282684326, + "learning_rate": 9.244993640324692e-06, + "loss": 0.534, + "step": 12912 + }, + { + "epoch": 0.5717321762615809, + "grad_norm": 1.4791839122772217, + "learning_rate": 9.243452609513382e-06, + "loss": 0.6379, + "step": 12913 + }, + { + "epoch": 0.5717764519663947, + "grad_norm": 1.5016884803771973, + "learning_rate": 9.241911596771595e-06, + "loss": 0.5312, + "step": 12914 + }, + { + "epoch": 0.5718207276712086, + "grad_norm": 1.3422855138778687, + "learning_rate": 9.240370602136134e-06, + "loss": 0.6002, + "step": 12915 + }, + { + "epoch": 0.5718650033760225, + "grad_norm": 1.5546140670776367, + "learning_rate": 9.238829625643803e-06, + "loss": 0.6849, + "step": 12916 + }, + { + "epoch": 0.5719092790808363, + "grad_norm": 1.4228452444076538, + "learning_rate": 9.237288667331402e-06, + "loss": 0.6593, + "step": 12917 + }, + { + "epoch": 0.5719535547856502, + "grad_norm": 1.2675436735153198, + "learning_rate": 9.235747727235743e-06, + "loss": 0.6143, + "step": 12918 + }, + { + "epoch": 0.5719978304904642, + "grad_norm": 1.7218687534332275, + "learning_rate": 9.234206805393628e-06, + "loss": 0.6589, + "step": 12919 + }, + { + "epoch": 0.572042106195278, + "grad_norm": 1.2617460489273071, + "learning_rate": 9.232665901841857e-06, + "loss": 0.5639, + "step": 12920 + }, + { + "epoch": 0.5720863819000919, + "grad_norm": 1.5583053827285767, + "learning_rate": 9.231125016617239e-06, + "loss": 0.5798, + "step": 12921 + }, + { + "epoch": 0.5721306576049058, + "grad_norm": 1.2391467094421387, + "learning_rate": 9.229584149756574e-06, + "loss": 0.5928, + "step": 12922 + }, + { + "epoch": 0.5721749333097196, + "grad_norm": 1.683331847190857, + "learning_rate": 9.228043301296658e-06, + "loss": 0.6797, + "step": 12923 + }, + { + "epoch": 0.5722192090145335, + "grad_norm": 1.965591549873352, + "learning_rate": 9.22650247127431e-06, + "loss": 0.6988, + "step": 12924 + }, + { + "epoch": 0.5722634847193474, + "grad_norm": 1.2968974113464355, + "learning_rate": 9.224961659726314e-06, + "loss": 0.5331, + "step": 12925 + }, + { + "epoch": 0.5723077604241612, + "grad_norm": 1.4065178632736206, + "learning_rate": 9.22342086668948e-06, + "loss": 0.5375, + "step": 12926 + }, + { + "epoch": 0.5723520361289751, + "grad_norm": 1.639005422592163, + "learning_rate": 9.221880092200601e-06, + "loss": 0.6437, + "step": 12927 + }, + { + "epoch": 0.572396311833789, + "grad_norm": 1.4118082523345947, + "learning_rate": 9.220339336296487e-06, + "loss": 0.6338, + "step": 12928 + }, + { + "epoch": 0.5724405875386028, + "grad_norm": 1.4155217409133911, + "learning_rate": 9.218798599013933e-06, + "loss": 0.6026, + "step": 12929 + }, + { + "epoch": 0.5724848632434167, + "grad_norm": 1.5497517585754395, + "learning_rate": 9.217257880389738e-06, + "loss": 0.5925, + "step": 12930 + }, + { + "epoch": 0.5725291389482307, + "grad_norm": 1.5361318588256836, + "learning_rate": 9.215717180460702e-06, + "loss": 0.6501, + "step": 12931 + }, + { + "epoch": 0.5725734146530445, + "grad_norm": 1.3400022983551025, + "learning_rate": 9.214176499263623e-06, + "loss": 0.6346, + "step": 12932 + }, + { + "epoch": 0.5726176903578584, + "grad_norm": 1.7168344259262085, + "learning_rate": 9.212635836835295e-06, + "loss": 0.6896, + "step": 12933 + }, + { + "epoch": 0.5726619660626723, + "grad_norm": 1.3532369136810303, + "learning_rate": 9.211095193212522e-06, + "loss": 0.6361, + "step": 12934 + }, + { + "epoch": 0.5727062417674862, + "grad_norm": 1.512664556503296, + "learning_rate": 9.209554568432098e-06, + "loss": 0.6674, + "step": 12935 + }, + { + "epoch": 0.5727505174723, + "grad_norm": 1.5058399438858032, + "learning_rate": 9.208013962530819e-06, + "loss": 0.629, + "step": 12936 + }, + { + "epoch": 0.5727947931771139, + "grad_norm": 1.8753907680511475, + "learning_rate": 9.206473375545478e-06, + "loss": 0.6866, + "step": 12937 + }, + { + "epoch": 0.5728390688819278, + "grad_norm": 1.6162338256835938, + "learning_rate": 9.204932807512878e-06, + "loss": 0.6793, + "step": 12938 + }, + { + "epoch": 0.5728833445867416, + "grad_norm": 1.3230456113815308, + "learning_rate": 9.203392258469814e-06, + "loss": 0.5853, + "step": 12939 + }, + { + "epoch": 0.5729276202915555, + "grad_norm": 1.6499086618423462, + "learning_rate": 9.20185172845307e-06, + "loss": 0.6161, + "step": 12940 + }, + { + "epoch": 0.5729718959963694, + "grad_norm": 1.427748203277588, + "learning_rate": 9.20031121749945e-06, + "loss": 0.5826, + "step": 12941 + }, + { + "epoch": 0.5730161717011832, + "grad_norm": 1.4193191528320312, + "learning_rate": 9.198770725645749e-06, + "loss": 0.6096, + "step": 12942 + }, + { + "epoch": 0.5730604474059972, + "grad_norm": 1.4688289165496826, + "learning_rate": 9.197230252928752e-06, + "loss": 0.627, + "step": 12943 + }, + { + "epoch": 0.5731047231108111, + "grad_norm": 1.4896148443222046, + "learning_rate": 9.195689799385261e-06, + "loss": 0.6707, + "step": 12944 + }, + { + "epoch": 0.5731489988156249, + "grad_norm": 1.5959043502807617, + "learning_rate": 9.194149365052064e-06, + "loss": 0.6748, + "step": 12945 + }, + { + "epoch": 0.5731932745204388, + "grad_norm": 1.6442391872406006, + "learning_rate": 9.19260894996595e-06, + "loss": 0.5638, + "step": 12946 + }, + { + "epoch": 0.5732375502252527, + "grad_norm": 1.3059121370315552, + "learning_rate": 9.191068554163713e-06, + "loss": 0.5769, + "step": 12947 + }, + { + "epoch": 0.5732818259300665, + "grad_norm": 1.2838257551193237, + "learning_rate": 9.18952817768215e-06, + "loss": 0.604, + "step": 12948 + }, + { + "epoch": 0.5733261016348804, + "grad_norm": 1.4987307786941528, + "learning_rate": 9.187987820558045e-06, + "loss": 0.5688, + "step": 12949 + }, + { + "epoch": 0.5733703773396943, + "grad_norm": 2.0598795413970947, + "learning_rate": 9.186447482828185e-06, + "loss": 0.6114, + "step": 12950 + }, + { + "epoch": 0.5734146530445081, + "grad_norm": 1.7547708749771118, + "learning_rate": 9.184907164529369e-06, + "loss": 0.696, + "step": 12951 + }, + { + "epoch": 0.573458928749322, + "grad_norm": 1.390694260597229, + "learning_rate": 9.183366865698382e-06, + "loss": 0.6239, + "step": 12952 + }, + { + "epoch": 0.5735032044541359, + "grad_norm": 1.269986867904663, + "learning_rate": 9.181826586372009e-06, + "loss": 0.5418, + "step": 12953 + }, + { + "epoch": 0.5735474801589497, + "grad_norm": 1.7420028448104858, + "learning_rate": 9.180286326587045e-06, + "loss": 0.6647, + "step": 12954 + }, + { + "epoch": 0.5735917558637637, + "grad_norm": 1.6082826852798462, + "learning_rate": 9.178746086380274e-06, + "loss": 0.6468, + "step": 12955 + }, + { + "epoch": 0.5736360315685776, + "grad_norm": 1.6573823690414429, + "learning_rate": 9.177205865788486e-06, + "loss": 0.6823, + "step": 12956 + }, + { + "epoch": 0.5736803072733914, + "grad_norm": 1.859466791152954, + "learning_rate": 9.175665664848463e-06, + "loss": 0.7404, + "step": 12957 + }, + { + "epoch": 0.5737245829782053, + "grad_norm": 1.7341190576553345, + "learning_rate": 9.174125483596996e-06, + "loss": 0.7014, + "step": 12958 + }, + { + "epoch": 0.5737688586830192, + "grad_norm": 1.4926135540008545, + "learning_rate": 9.172585322070868e-06, + "loss": 0.6611, + "step": 12959 + }, + { + "epoch": 0.573813134387833, + "grad_norm": 1.388173222541809, + "learning_rate": 9.171045180306862e-06, + "loss": 0.6209, + "step": 12960 + }, + { + "epoch": 0.5738574100926469, + "grad_norm": 1.3682011365890503, + "learning_rate": 9.169505058341774e-06, + "loss": 0.656, + "step": 12961 + }, + { + "epoch": 0.5739016857974608, + "grad_norm": 1.7675085067749023, + "learning_rate": 9.167964956212378e-06, + "loss": 0.6457, + "step": 12962 + }, + { + "epoch": 0.5739459615022746, + "grad_norm": 1.6596685647964478, + "learning_rate": 9.166424873955462e-06, + "loss": 0.6741, + "step": 12963 + }, + { + "epoch": 0.5739902372070885, + "grad_norm": 1.3061822652816772, + "learning_rate": 9.164884811607809e-06, + "loss": 0.607, + "step": 12964 + }, + { + "epoch": 0.5740345129119024, + "grad_norm": 1.8070247173309326, + "learning_rate": 9.163344769206204e-06, + "loss": 0.6359, + "step": 12965 + }, + { + "epoch": 0.5740787886167162, + "grad_norm": 1.6643848419189453, + "learning_rate": 9.161804746787428e-06, + "loss": 0.5927, + "step": 12966 + }, + { + "epoch": 0.5741230643215302, + "grad_norm": 1.8477213382720947, + "learning_rate": 9.160264744388259e-06, + "loss": 0.6492, + "step": 12967 + }, + { + "epoch": 0.5741673400263441, + "grad_norm": 1.4713948965072632, + "learning_rate": 9.158724762045488e-06, + "loss": 0.5844, + "step": 12968 + }, + { + "epoch": 0.5742116157311579, + "grad_norm": 1.4814162254333496, + "learning_rate": 9.157184799795889e-06, + "loss": 0.7001, + "step": 12969 + }, + { + "epoch": 0.5742558914359718, + "grad_norm": 1.3544273376464844, + "learning_rate": 9.155644857676241e-06, + "loss": 0.61, + "step": 12970 + }, + { + "epoch": 0.5743001671407857, + "grad_norm": 1.3483556509017944, + "learning_rate": 9.154104935723333e-06, + "loss": 0.5899, + "step": 12971 + }, + { + "epoch": 0.5743444428455996, + "grad_norm": 1.3115733861923218, + "learning_rate": 9.152565033973941e-06, + "loss": 0.6071, + "step": 12972 + }, + { + "epoch": 0.5743887185504134, + "grad_norm": 1.3074126243591309, + "learning_rate": 9.151025152464837e-06, + "loss": 0.6141, + "step": 12973 + }, + { + "epoch": 0.5744329942552273, + "grad_norm": 1.2681292295455933, + "learning_rate": 9.14948529123281e-06, + "loss": 0.5516, + "step": 12974 + }, + { + "epoch": 0.5744772699600412, + "grad_norm": 1.7999056577682495, + "learning_rate": 9.147945450314635e-06, + "loss": 0.6603, + "step": 12975 + }, + { + "epoch": 0.574521545664855, + "grad_norm": 1.3281766176223755, + "learning_rate": 9.14640562974709e-06, + "loss": 0.6555, + "step": 12976 + }, + { + "epoch": 0.5745658213696689, + "grad_norm": 1.3763784170150757, + "learning_rate": 9.144865829566949e-06, + "loss": 0.6216, + "step": 12977 + }, + { + "epoch": 0.5746100970744829, + "grad_norm": 1.3652786016464233, + "learning_rate": 9.143326049810991e-06, + "loss": 0.5679, + "step": 12978 + }, + { + "epoch": 0.5746543727792967, + "grad_norm": 1.4049042463302612, + "learning_rate": 9.141786290515995e-06, + "loss": 0.6461, + "step": 12979 + }, + { + "epoch": 0.5746986484841106, + "grad_norm": 1.2624675035476685, + "learning_rate": 9.140246551718732e-06, + "loss": 0.575, + "step": 12980 + }, + { + "epoch": 0.5747429241889245, + "grad_norm": 1.501707911491394, + "learning_rate": 9.138706833455982e-06, + "loss": 0.6117, + "step": 12981 + }, + { + "epoch": 0.5747871998937383, + "grad_norm": 1.5375003814697266, + "learning_rate": 9.137167135764519e-06, + "loss": 0.6967, + "step": 12982 + }, + { + "epoch": 0.5748314755985522, + "grad_norm": 1.3032175302505493, + "learning_rate": 9.135627458681116e-06, + "loss": 0.5665, + "step": 12983 + }, + { + "epoch": 0.5748757513033661, + "grad_norm": 1.666012167930603, + "learning_rate": 9.134087802242542e-06, + "loss": 0.626, + "step": 12984 + }, + { + "epoch": 0.5749200270081799, + "grad_norm": 1.300734519958496, + "learning_rate": 9.13254816648558e-06, + "loss": 0.6154, + "step": 12985 + }, + { + "epoch": 0.5749643027129938, + "grad_norm": 1.7366615533828735, + "learning_rate": 9.131008551447e-06, + "loss": 0.6654, + "step": 12986 + }, + { + "epoch": 0.5750085784178077, + "grad_norm": 1.8364951610565186, + "learning_rate": 9.12946895716357e-06, + "loss": 0.7485, + "step": 12987 + }, + { + "epoch": 0.5750528541226215, + "grad_norm": 1.9312790632247925, + "learning_rate": 9.12792938367207e-06, + "loss": 0.6226, + "step": 12988 + }, + { + "epoch": 0.5750971298274354, + "grad_norm": 1.3675960302352905, + "learning_rate": 9.126389831009264e-06, + "loss": 0.5869, + "step": 12989 + }, + { + "epoch": 0.5751414055322494, + "grad_norm": 1.3495409488677979, + "learning_rate": 9.124850299211923e-06, + "loss": 0.6611, + "step": 12990 + }, + { + "epoch": 0.5751856812370632, + "grad_norm": 1.9349077939987183, + "learning_rate": 9.123310788316823e-06, + "loss": 0.6807, + "step": 12991 + }, + { + "epoch": 0.5752299569418771, + "grad_norm": 1.5378926992416382, + "learning_rate": 9.121771298360732e-06, + "loss": 0.6146, + "step": 12992 + }, + { + "epoch": 0.575274232646691, + "grad_norm": 1.6545835733413696, + "learning_rate": 9.120231829380417e-06, + "loss": 0.5989, + "step": 12993 + }, + { + "epoch": 0.5753185083515048, + "grad_norm": 1.2857496738433838, + "learning_rate": 9.118692381412645e-06, + "loss": 0.5391, + "step": 12994 + }, + { + "epoch": 0.5753627840563187, + "grad_norm": 1.6192481517791748, + "learning_rate": 9.117152954494195e-06, + "loss": 0.6075, + "step": 12995 + }, + { + "epoch": 0.5754070597611326, + "grad_norm": 1.4912887811660767, + "learning_rate": 9.115613548661824e-06, + "loss": 0.5846, + "step": 12996 + }, + { + "epoch": 0.5754513354659464, + "grad_norm": 1.9437514543533325, + "learning_rate": 9.114074163952299e-06, + "loss": 0.7268, + "step": 12997 + }, + { + "epoch": 0.5754956111707603, + "grad_norm": 1.467390775680542, + "learning_rate": 9.112534800402395e-06, + "loss": 0.648, + "step": 12998 + }, + { + "epoch": 0.5755398868755742, + "grad_norm": 1.6511809825897217, + "learning_rate": 9.110995458048878e-06, + "loss": 0.6483, + "step": 12999 + }, + { + "epoch": 0.575584162580388, + "grad_norm": 2.0003721714019775, + "learning_rate": 9.109456136928506e-06, + "loss": 0.7038, + "step": 13000 + }, + { + "epoch": 0.5756284382852019, + "grad_norm": 1.8954105377197266, + "learning_rate": 9.107916837078052e-06, + "loss": 0.7568, + "step": 13001 + }, + { + "epoch": 0.5756727139900158, + "grad_norm": 1.6583589315414429, + "learning_rate": 9.106377558534278e-06, + "loss": 0.6751, + "step": 13002 + }, + { + "epoch": 0.5757169896948297, + "grad_norm": 1.6065483093261719, + "learning_rate": 9.104838301333947e-06, + "loss": 0.7177, + "step": 13003 + }, + { + "epoch": 0.5757612653996436, + "grad_norm": 1.6545655727386475, + "learning_rate": 9.103299065513823e-06, + "loss": 0.6614, + "step": 13004 + }, + { + "epoch": 0.5758055411044575, + "grad_norm": 1.3868733644485474, + "learning_rate": 9.101759851110673e-06, + "loss": 0.6066, + "step": 13005 + }, + { + "epoch": 0.5758498168092714, + "grad_norm": 4.868565082550049, + "learning_rate": 9.100220658161259e-06, + "loss": 0.5874, + "step": 13006 + }, + { + "epoch": 0.5758940925140852, + "grad_norm": 1.4287296533584595, + "learning_rate": 9.098681486702336e-06, + "loss": 0.6046, + "step": 13007 + }, + { + "epoch": 0.5759383682188991, + "grad_norm": 1.5437954664230347, + "learning_rate": 9.097142336770675e-06, + "loss": 0.5816, + "step": 13008 + }, + { + "epoch": 0.575982643923713, + "grad_norm": 2.3735296726226807, + "learning_rate": 9.095603208403037e-06, + "loss": 0.8605, + "step": 13009 + }, + { + "epoch": 0.5760269196285268, + "grad_norm": 1.5764559507369995, + "learning_rate": 9.094064101636176e-06, + "loss": 0.6581, + "step": 13010 + }, + { + "epoch": 0.5760711953333407, + "grad_norm": 1.738555908203125, + "learning_rate": 9.092525016506858e-06, + "loss": 0.6376, + "step": 13011 + }, + { + "epoch": 0.5761154710381546, + "grad_norm": 1.3199900388717651, + "learning_rate": 9.090985953051842e-06, + "loss": 0.6216, + "step": 13012 + }, + { + "epoch": 0.5761597467429684, + "grad_norm": 1.4930819272994995, + "learning_rate": 9.089446911307887e-06, + "loss": 0.6051, + "step": 13013 + }, + { + "epoch": 0.5762040224477823, + "grad_norm": 1.2497471570968628, + "learning_rate": 9.087907891311748e-06, + "loss": 0.5912, + "step": 13014 + }, + { + "epoch": 0.5762482981525963, + "grad_norm": 1.7000865936279297, + "learning_rate": 9.086368893100188e-06, + "loss": 0.6392, + "step": 13015 + }, + { + "epoch": 0.5762925738574101, + "grad_norm": 1.360482096672058, + "learning_rate": 9.084829916709965e-06, + "loss": 0.5776, + "step": 13016 + }, + { + "epoch": 0.576336849562224, + "grad_norm": 1.5933341979980469, + "learning_rate": 9.083290962177828e-06, + "loss": 0.5637, + "step": 13017 + }, + { + "epoch": 0.5763811252670379, + "grad_norm": 1.4407920837402344, + "learning_rate": 9.081752029540545e-06, + "loss": 0.5544, + "step": 13018 + }, + { + "epoch": 0.5764254009718517, + "grad_norm": 1.465406894683838, + "learning_rate": 9.080213118834869e-06, + "loss": 0.6222, + "step": 13019 + }, + { + "epoch": 0.5764696766766656, + "grad_norm": 1.3940247297286987, + "learning_rate": 9.078674230097548e-06, + "loss": 0.5994, + "step": 13020 + }, + { + "epoch": 0.5765139523814795, + "grad_norm": 1.4798409938812256, + "learning_rate": 9.077135363365348e-06, + "loss": 0.6478, + "step": 13021 + }, + { + "epoch": 0.5765582280862933, + "grad_norm": 1.4733072519302368, + "learning_rate": 9.075596518675017e-06, + "loss": 0.6533, + "step": 13022 + }, + { + "epoch": 0.5766025037911072, + "grad_norm": 1.4862772226333618, + "learning_rate": 9.07405769606331e-06, + "loss": 0.611, + "step": 13023 + }, + { + "epoch": 0.5766467794959211, + "grad_norm": 1.671903133392334, + "learning_rate": 9.072518895566979e-06, + "loss": 0.6436, + "step": 13024 + }, + { + "epoch": 0.5766910552007349, + "grad_norm": 1.7265504598617554, + "learning_rate": 9.070980117222782e-06, + "loss": 0.611, + "step": 13025 + }, + { + "epoch": 0.5767353309055488, + "grad_norm": 1.500184178352356, + "learning_rate": 9.069441361067467e-06, + "loss": 0.6297, + "step": 13026 + }, + { + "epoch": 0.5767796066103628, + "grad_norm": 1.5747920274734497, + "learning_rate": 9.067902627137787e-06, + "loss": 0.6965, + "step": 13027 + }, + { + "epoch": 0.5768238823151766, + "grad_norm": 1.3298423290252686, + "learning_rate": 9.066363915470494e-06, + "loss": 0.6116, + "step": 13028 + }, + { + "epoch": 0.5768681580199905, + "grad_norm": 1.8033857345581055, + "learning_rate": 9.064825226102342e-06, + "loss": 0.7014, + "step": 13029 + }, + { + "epoch": 0.5769124337248044, + "grad_norm": 1.5618178844451904, + "learning_rate": 9.063286559070072e-06, + "loss": 0.6622, + "step": 13030 + }, + { + "epoch": 0.5769567094296182, + "grad_norm": 1.9916712045669556, + "learning_rate": 9.061747914410442e-06, + "loss": 0.766, + "step": 13031 + }, + { + "epoch": 0.5770009851344321, + "grad_norm": 1.275769591331482, + "learning_rate": 9.060209292160203e-06, + "loss": 0.5793, + "step": 13032 + }, + { + "epoch": 0.577045260839246, + "grad_norm": 1.3491172790527344, + "learning_rate": 9.058670692356096e-06, + "loss": 0.5785, + "step": 13033 + }, + { + "epoch": 0.5770895365440598, + "grad_norm": 1.8211349248886108, + "learning_rate": 9.057132115034873e-06, + "loss": 0.6369, + "step": 13034 + }, + { + "epoch": 0.5771338122488737, + "grad_norm": 1.2153708934783936, + "learning_rate": 9.055593560233284e-06, + "loss": 0.5581, + "step": 13035 + }, + { + "epoch": 0.5771780879536876, + "grad_norm": 1.7208727598190308, + "learning_rate": 9.054055027988073e-06, + "loss": 0.6075, + "step": 13036 + }, + { + "epoch": 0.5772223636585014, + "grad_norm": 1.398547887802124, + "learning_rate": 9.052516518335987e-06, + "loss": 0.6192, + "step": 13037 + }, + { + "epoch": 0.5772666393633153, + "grad_norm": 1.61117684841156, + "learning_rate": 9.050978031313773e-06, + "loss": 0.6197, + "step": 13038 + }, + { + "epoch": 0.5773109150681293, + "grad_norm": 1.5558221340179443, + "learning_rate": 9.049439566958176e-06, + "loss": 0.6552, + "step": 13039 + }, + { + "epoch": 0.5773551907729432, + "grad_norm": 1.3535493612289429, + "learning_rate": 9.047901125305937e-06, + "loss": 0.5615, + "step": 13040 + }, + { + "epoch": 0.577399466477757, + "grad_norm": 1.6288870573043823, + "learning_rate": 9.046362706393809e-06, + "loss": 0.6773, + "step": 13041 + }, + { + "epoch": 0.5774437421825709, + "grad_norm": 1.393554925918579, + "learning_rate": 9.04482431025853e-06, + "loss": 0.6715, + "step": 13042 + }, + { + "epoch": 0.5774880178873848, + "grad_norm": 1.3876557350158691, + "learning_rate": 9.043285936936849e-06, + "loss": 0.5821, + "step": 13043 + }, + { + "epoch": 0.5775322935921986, + "grad_norm": 1.2264001369476318, + "learning_rate": 9.041747586465496e-06, + "loss": 0.5419, + "step": 13044 + }, + { + "epoch": 0.5775765692970125, + "grad_norm": 1.5895297527313232, + "learning_rate": 9.040209258881227e-06, + "loss": 0.5591, + "step": 13045 + }, + { + "epoch": 0.5776208450018264, + "grad_norm": 1.5574382543563843, + "learning_rate": 9.038670954220779e-06, + "loss": 0.6308, + "step": 13046 + }, + { + "epoch": 0.5776651207066402, + "grad_norm": 1.692688226699829, + "learning_rate": 9.037132672520888e-06, + "loss": 0.6712, + "step": 13047 + }, + { + "epoch": 0.5777093964114541, + "grad_norm": 1.6411017179489136, + "learning_rate": 9.035594413818303e-06, + "loss": 0.6625, + "step": 13048 + }, + { + "epoch": 0.577753672116268, + "grad_norm": 1.282365083694458, + "learning_rate": 9.03405617814976e-06, + "loss": 0.611, + "step": 13049 + }, + { + "epoch": 0.5777979478210818, + "grad_norm": 1.2841073274612427, + "learning_rate": 9.032517965551995e-06, + "loss": 0.5403, + "step": 13050 + }, + { + "epoch": 0.5778422235258958, + "grad_norm": 1.5775947570800781, + "learning_rate": 9.030979776061755e-06, + "loss": 0.6202, + "step": 13051 + }, + { + "epoch": 0.5778864992307097, + "grad_norm": 2.602780818939209, + "learning_rate": 9.029441609715775e-06, + "loss": 0.7022, + "step": 13052 + }, + { + "epoch": 0.5779307749355235, + "grad_norm": 1.3930097818374634, + "learning_rate": 9.02790346655079e-06, + "loss": 0.6012, + "step": 13053 + }, + { + "epoch": 0.5779750506403374, + "grad_norm": 1.8605477809906006, + "learning_rate": 9.026365346603534e-06, + "loss": 0.7409, + "step": 13054 + }, + { + "epoch": 0.5780193263451513, + "grad_norm": 2.165872573852539, + "learning_rate": 9.024827249910755e-06, + "loss": 0.6013, + "step": 13055 + }, + { + "epoch": 0.5780636020499651, + "grad_norm": 1.5475496053695679, + "learning_rate": 9.023289176509181e-06, + "loss": 0.6895, + "step": 13056 + }, + { + "epoch": 0.578107877754779, + "grad_norm": 1.5818406343460083, + "learning_rate": 9.021751126435549e-06, + "loss": 0.6548, + "step": 13057 + }, + { + "epoch": 0.5781521534595929, + "grad_norm": 1.5629149675369263, + "learning_rate": 9.020213099726597e-06, + "loss": 0.6227, + "step": 13058 + }, + { + "epoch": 0.5781964291644067, + "grad_norm": 1.221298336982727, + "learning_rate": 9.018675096419059e-06, + "loss": 0.5648, + "step": 13059 + }, + { + "epoch": 0.5782407048692206, + "grad_norm": 1.1813774108886719, + "learning_rate": 9.017137116549662e-06, + "loss": 0.5319, + "step": 13060 + }, + { + "epoch": 0.5782849805740345, + "grad_norm": 1.4376121759414673, + "learning_rate": 9.015599160155148e-06, + "loss": 0.593, + "step": 13061 + }, + { + "epoch": 0.5783292562788483, + "grad_norm": 1.5260059833526611, + "learning_rate": 9.014061227272247e-06, + "loss": 0.5886, + "step": 13062 + }, + { + "epoch": 0.5783735319836623, + "grad_norm": 1.3645128011703491, + "learning_rate": 9.012523317937691e-06, + "loss": 0.6418, + "step": 13063 + }, + { + "epoch": 0.5784178076884762, + "grad_norm": 1.2791576385498047, + "learning_rate": 9.010985432188207e-06, + "loss": 0.5414, + "step": 13064 + }, + { + "epoch": 0.57846208339329, + "grad_norm": 1.5892150402069092, + "learning_rate": 9.009447570060535e-06, + "loss": 0.669, + "step": 13065 + }, + { + "epoch": 0.5785063590981039, + "grad_norm": 1.3407044410705566, + "learning_rate": 9.007909731591403e-06, + "loss": 0.5607, + "step": 13066 + }, + { + "epoch": 0.5785506348029178, + "grad_norm": 1.2544053792953491, + "learning_rate": 9.006371916817533e-06, + "loss": 0.5831, + "step": 13067 + }, + { + "epoch": 0.5785949105077316, + "grad_norm": 1.7243643999099731, + "learning_rate": 9.004834125775665e-06, + "loss": 0.669, + "step": 13068 + }, + { + "epoch": 0.5786391862125455, + "grad_norm": 1.421553373336792, + "learning_rate": 9.003296358502524e-06, + "loss": 0.639, + "step": 13069 + }, + { + "epoch": 0.5786834619173594, + "grad_norm": 1.4800076484680176, + "learning_rate": 9.001758615034835e-06, + "loss": 0.6007, + "step": 13070 + }, + { + "epoch": 0.5787277376221732, + "grad_norm": 1.5275639295578003, + "learning_rate": 9.000220895409332e-06, + "loss": 0.6382, + "step": 13071 + }, + { + "epoch": 0.5787720133269871, + "grad_norm": 1.7114166021347046, + "learning_rate": 8.998683199662738e-06, + "loss": 0.5919, + "step": 13072 + }, + { + "epoch": 0.578816289031801, + "grad_norm": 1.4913572072982788, + "learning_rate": 8.99714552783178e-06, + "loss": 0.6817, + "step": 13073 + }, + { + "epoch": 0.5788605647366148, + "grad_norm": 1.5027064085006714, + "learning_rate": 8.99560787995318e-06, + "loss": 0.6373, + "step": 13074 + }, + { + "epoch": 0.5789048404414288, + "grad_norm": 1.3207985162734985, + "learning_rate": 8.994070256063676e-06, + "loss": 0.6035, + "step": 13075 + }, + { + "epoch": 0.5789491161462427, + "grad_norm": 1.4752967357635498, + "learning_rate": 8.992532656199983e-06, + "loss": 0.592, + "step": 13076 + }, + { + "epoch": 0.5789933918510566, + "grad_norm": 1.7134162187576294, + "learning_rate": 8.990995080398821e-06, + "loss": 0.5917, + "step": 13077 + }, + { + "epoch": 0.5790376675558704, + "grad_norm": 1.8435091972351074, + "learning_rate": 8.989457528696925e-06, + "loss": 0.7034, + "step": 13078 + }, + { + "epoch": 0.5790819432606843, + "grad_norm": 1.3296443223953247, + "learning_rate": 8.987920001131015e-06, + "loss": 0.5625, + "step": 13079 + }, + { + "epoch": 0.5791262189654982, + "grad_norm": 1.8387528657913208, + "learning_rate": 8.986382497737807e-06, + "loss": 0.5908, + "step": 13080 + }, + { + "epoch": 0.579170494670312, + "grad_norm": 1.5803003311157227, + "learning_rate": 8.98484501855403e-06, + "loss": 0.6721, + "step": 13081 + }, + { + "epoch": 0.5792147703751259, + "grad_norm": 1.5163973569869995, + "learning_rate": 8.983307563616404e-06, + "loss": 0.6441, + "step": 13082 + }, + { + "epoch": 0.5792590460799398, + "grad_norm": 1.2958382368087769, + "learning_rate": 8.981770132961649e-06, + "loss": 0.6009, + "step": 13083 + }, + { + "epoch": 0.5793033217847536, + "grad_norm": 1.6793524026870728, + "learning_rate": 8.980232726626484e-06, + "loss": 0.6067, + "step": 13084 + }, + { + "epoch": 0.5793475974895675, + "grad_norm": 1.3648276329040527, + "learning_rate": 8.97869534464763e-06, + "loss": 0.5972, + "step": 13085 + }, + { + "epoch": 0.5793918731943815, + "grad_norm": 1.5177586078643799, + "learning_rate": 8.977157987061807e-06, + "loss": 0.634, + "step": 13086 + }, + { + "epoch": 0.5794361488991953, + "grad_norm": 1.3885868787765503, + "learning_rate": 8.97562065390573e-06, + "loss": 0.5645, + "step": 13087 + }, + { + "epoch": 0.5794804246040092, + "grad_norm": 1.3572815656661987, + "learning_rate": 8.97408334521612e-06, + "loss": 0.586, + "step": 13088 + }, + { + "epoch": 0.5795247003088231, + "grad_norm": 1.5327447652816772, + "learning_rate": 8.972546061029698e-06, + "loss": 0.6823, + "step": 13089 + }, + { + "epoch": 0.5795689760136369, + "grad_norm": 1.6003614664077759, + "learning_rate": 8.971008801383177e-06, + "loss": 0.5972, + "step": 13090 + }, + { + "epoch": 0.5796132517184508, + "grad_norm": 1.6070010662078857, + "learning_rate": 8.969471566313267e-06, + "loss": 0.6127, + "step": 13091 + }, + { + "epoch": 0.5796575274232647, + "grad_norm": 1.5117642879486084, + "learning_rate": 8.967934355856694e-06, + "loss": 0.6369, + "step": 13092 + }, + { + "epoch": 0.5797018031280785, + "grad_norm": 1.9963839054107666, + "learning_rate": 8.966397170050168e-06, + "loss": 0.7371, + "step": 13093 + }, + { + "epoch": 0.5797460788328924, + "grad_norm": 1.5561758279800415, + "learning_rate": 8.964860008930402e-06, + "loss": 0.6512, + "step": 13094 + }, + { + "epoch": 0.5797903545377063, + "grad_norm": 1.7963920831680298, + "learning_rate": 8.963322872534115e-06, + "loss": 0.6003, + "step": 13095 + }, + { + "epoch": 0.5798346302425201, + "grad_norm": 1.2754020690917969, + "learning_rate": 8.961785760898015e-06, + "loss": 0.6064, + "step": 13096 + }, + { + "epoch": 0.579878905947334, + "grad_norm": 1.4623758792877197, + "learning_rate": 8.960248674058815e-06, + "loss": 0.542, + "step": 13097 + }, + { + "epoch": 0.579923181652148, + "grad_norm": 1.5797785520553589, + "learning_rate": 8.95871161205323e-06, + "loss": 0.6362, + "step": 13098 + }, + { + "epoch": 0.5799674573569618, + "grad_norm": 1.1334384679794312, + "learning_rate": 8.957174574917975e-06, + "loss": 0.5296, + "step": 13099 + }, + { + "epoch": 0.5800117330617757, + "grad_norm": 1.8722305297851562, + "learning_rate": 8.955637562689754e-06, + "loss": 0.6583, + "step": 13100 + }, + { + "epoch": 0.5800560087665896, + "grad_norm": 1.4880105257034302, + "learning_rate": 8.954100575405274e-06, + "loss": 0.664, + "step": 13101 + }, + { + "epoch": 0.5801002844714034, + "grad_norm": 1.3860747814178467, + "learning_rate": 8.952563613101255e-06, + "loss": 0.66, + "step": 13102 + }, + { + "epoch": 0.5801445601762173, + "grad_norm": 1.3983983993530273, + "learning_rate": 8.9510266758144e-06, + "loss": 0.6021, + "step": 13103 + }, + { + "epoch": 0.5801888358810312, + "grad_norm": 1.506502389907837, + "learning_rate": 8.94948976358142e-06, + "loss": 0.6137, + "step": 13104 + }, + { + "epoch": 0.580233111585845, + "grad_norm": 1.4695260524749756, + "learning_rate": 8.947952876439021e-06, + "loss": 0.5953, + "step": 13105 + }, + { + "epoch": 0.5802773872906589, + "grad_norm": 1.612281084060669, + "learning_rate": 8.94641601442391e-06, + "loss": 0.6127, + "step": 13106 + }, + { + "epoch": 0.5803216629954728, + "grad_norm": 1.514201283454895, + "learning_rate": 8.944879177572796e-06, + "loss": 0.7127, + "step": 13107 + }, + { + "epoch": 0.5803659387002866, + "grad_norm": 1.3738343715667725, + "learning_rate": 8.943342365922383e-06, + "loss": 0.6329, + "step": 13108 + }, + { + "epoch": 0.5804102144051005, + "grad_norm": 1.9042459726333618, + "learning_rate": 8.941805579509378e-06, + "loss": 0.6966, + "step": 13109 + }, + { + "epoch": 0.5804544901099145, + "grad_norm": 1.2828470468521118, + "learning_rate": 8.940268818370486e-06, + "loss": 0.5439, + "step": 13110 + }, + { + "epoch": 0.5804987658147284, + "grad_norm": 2.1793246269226074, + "learning_rate": 8.938732082542404e-06, + "loss": 0.7169, + "step": 13111 + }, + { + "epoch": 0.5805430415195422, + "grad_norm": 1.977035403251648, + "learning_rate": 8.937195372061847e-06, + "loss": 0.6713, + "step": 13112 + }, + { + "epoch": 0.5805873172243561, + "grad_norm": 1.6243282556533813, + "learning_rate": 8.935658686965514e-06, + "loss": 0.6237, + "step": 13113 + }, + { + "epoch": 0.58063159292917, + "grad_norm": 1.3479952812194824, + "learning_rate": 8.934122027290106e-06, + "loss": 0.6363, + "step": 13114 + }, + { + "epoch": 0.5806758686339838, + "grad_norm": 1.636398434638977, + "learning_rate": 8.932585393072324e-06, + "loss": 0.6623, + "step": 13115 + }, + { + "epoch": 0.5807201443387977, + "grad_norm": 1.4565078020095825, + "learning_rate": 8.931048784348875e-06, + "loss": 0.6018, + "step": 13116 + }, + { + "epoch": 0.5807644200436116, + "grad_norm": 1.511370301246643, + "learning_rate": 8.92951220115645e-06, + "loss": 0.606, + "step": 13117 + }, + { + "epoch": 0.5808086957484254, + "grad_norm": 1.4731100797653198, + "learning_rate": 8.927975643531759e-06, + "loss": 0.5863, + "step": 13118 + }, + { + "epoch": 0.5808529714532393, + "grad_norm": 1.3318727016448975, + "learning_rate": 8.926439111511494e-06, + "loss": 0.626, + "step": 13119 + }, + { + "epoch": 0.5808972471580532, + "grad_norm": 1.316207766532898, + "learning_rate": 8.924902605132359e-06, + "loss": 0.5763, + "step": 13120 + }, + { + "epoch": 0.580941522862867, + "grad_norm": 1.3264102935791016, + "learning_rate": 8.923366124431044e-06, + "loss": 0.6227, + "step": 13121 + }, + { + "epoch": 0.580985798567681, + "grad_norm": 1.2233303785324097, + "learning_rate": 8.92182966944426e-06, + "loss": 0.554, + "step": 13122 + }, + { + "epoch": 0.5810300742724949, + "grad_norm": 1.5233092308044434, + "learning_rate": 8.920293240208694e-06, + "loss": 0.6123, + "step": 13123 + }, + { + "epoch": 0.5810743499773087, + "grad_norm": 1.4257477521896362, + "learning_rate": 8.918756836761041e-06, + "loss": 0.5707, + "step": 13124 + }, + { + "epoch": 0.5811186256821226, + "grad_norm": 1.2242052555084229, + "learning_rate": 8.917220459138003e-06, + "loss": 0.5102, + "step": 13125 + }, + { + "epoch": 0.5811629013869365, + "grad_norm": 1.4129749536514282, + "learning_rate": 8.915684107376276e-06, + "loss": 0.5567, + "step": 13126 + }, + { + "epoch": 0.5812071770917503, + "grad_norm": 1.5420128107070923, + "learning_rate": 8.914147781512547e-06, + "loss": 0.7274, + "step": 13127 + }, + { + "epoch": 0.5812514527965642, + "grad_norm": 1.523575782775879, + "learning_rate": 8.912611481583516e-06, + "loss": 0.6327, + "step": 13128 + }, + { + "epoch": 0.5812957285013781, + "grad_norm": 1.3570834398269653, + "learning_rate": 8.911075207625875e-06, + "loss": 0.5792, + "step": 13129 + }, + { + "epoch": 0.5813400042061919, + "grad_norm": 1.57833731174469, + "learning_rate": 8.909538959676316e-06, + "loss": 0.6363, + "step": 13130 + }, + { + "epoch": 0.5813842799110058, + "grad_norm": 1.6157708168029785, + "learning_rate": 8.908002737771529e-06, + "loss": 0.6873, + "step": 13131 + }, + { + "epoch": 0.5814285556158197, + "grad_norm": 1.2989892959594727, + "learning_rate": 8.90646654194821e-06, + "loss": 0.6334, + "step": 13132 + }, + { + "epoch": 0.5814728313206335, + "grad_norm": 1.350054144859314, + "learning_rate": 8.904930372243046e-06, + "loss": 0.6178, + "step": 13133 + }, + { + "epoch": 0.5815171070254475, + "grad_norm": 1.5264434814453125, + "learning_rate": 8.903394228692725e-06, + "loss": 0.6194, + "step": 13134 + }, + { + "epoch": 0.5815613827302614, + "grad_norm": 1.3575284481048584, + "learning_rate": 8.901858111333945e-06, + "loss": 0.6, + "step": 13135 + }, + { + "epoch": 0.5816056584350752, + "grad_norm": 1.542508840560913, + "learning_rate": 8.900322020203388e-06, + "loss": 0.607, + "step": 13136 + }, + { + "epoch": 0.5816499341398891, + "grad_norm": 1.3484529256820679, + "learning_rate": 8.898785955337744e-06, + "loss": 0.5773, + "step": 13137 + }, + { + "epoch": 0.581694209844703, + "grad_norm": 1.382043480873108, + "learning_rate": 8.8972499167737e-06, + "loss": 0.59, + "step": 13138 + }, + { + "epoch": 0.5817384855495168, + "grad_norm": 1.3828247785568237, + "learning_rate": 8.895713904547947e-06, + "loss": 0.6103, + "step": 13139 + }, + { + "epoch": 0.5817827612543307, + "grad_norm": 1.4772205352783203, + "learning_rate": 8.894177918697167e-06, + "loss": 0.6884, + "step": 13140 + }, + { + "epoch": 0.5818270369591446, + "grad_norm": 1.4508485794067383, + "learning_rate": 8.892641959258044e-06, + "loss": 0.6302, + "step": 13141 + }, + { + "epoch": 0.5818713126639584, + "grad_norm": 1.4863677024841309, + "learning_rate": 8.89110602626727e-06, + "loss": 0.642, + "step": 13142 + }, + { + "epoch": 0.5819155883687723, + "grad_norm": 1.6154758930206299, + "learning_rate": 8.889570119761524e-06, + "loss": 0.6109, + "step": 13143 + }, + { + "epoch": 0.5819598640735862, + "grad_norm": 1.215941309928894, + "learning_rate": 8.888034239777489e-06, + "loss": 0.5641, + "step": 13144 + }, + { + "epoch": 0.5820041397784, + "grad_norm": 1.6782007217407227, + "learning_rate": 8.886498386351854e-06, + "loss": 0.6046, + "step": 13145 + }, + { + "epoch": 0.582048415483214, + "grad_norm": 1.7012783288955688, + "learning_rate": 8.884962559521301e-06, + "loss": 0.6696, + "step": 13146 + }, + { + "epoch": 0.5820926911880279, + "grad_norm": 1.598649263381958, + "learning_rate": 8.883426759322502e-06, + "loss": 0.6018, + "step": 13147 + }, + { + "epoch": 0.5821369668928418, + "grad_norm": 1.3995004892349243, + "learning_rate": 8.881890985792151e-06, + "loss": 0.5913, + "step": 13148 + }, + { + "epoch": 0.5821812425976556, + "grad_norm": 1.5422762632369995, + "learning_rate": 8.880355238966923e-06, + "loss": 0.6415, + "step": 13149 + }, + { + "epoch": 0.5822255183024695, + "grad_norm": 1.3658347129821777, + "learning_rate": 8.878819518883499e-06, + "loss": 0.5938, + "step": 13150 + }, + { + "epoch": 0.5822697940072834, + "grad_norm": 1.4924439191818237, + "learning_rate": 8.877283825578554e-06, + "loss": 0.6036, + "step": 13151 + }, + { + "epoch": 0.5823140697120972, + "grad_norm": 1.345291018486023, + "learning_rate": 8.875748159088774e-06, + "loss": 0.5668, + "step": 13152 + }, + { + "epoch": 0.5823583454169111, + "grad_norm": 1.4185748100280762, + "learning_rate": 8.874212519450834e-06, + "loss": 0.5935, + "step": 13153 + }, + { + "epoch": 0.582402621121725, + "grad_norm": 1.8865220546722412, + "learning_rate": 8.872676906701409e-06, + "loss": 0.6525, + "step": 13154 + }, + { + "epoch": 0.5824468968265388, + "grad_norm": 1.4729849100112915, + "learning_rate": 8.871141320877181e-06, + "loss": 0.6439, + "step": 13155 + }, + { + "epoch": 0.5824911725313527, + "grad_norm": 1.447235107421875, + "learning_rate": 8.869605762014823e-06, + "loss": 0.6088, + "step": 13156 + }, + { + "epoch": 0.5825354482361667, + "grad_norm": 1.1993489265441895, + "learning_rate": 8.868070230151007e-06, + "loss": 0.5677, + "step": 13157 + }, + { + "epoch": 0.5825797239409805, + "grad_norm": 1.2465652227401733, + "learning_rate": 8.866534725322416e-06, + "loss": 0.5261, + "step": 13158 + }, + { + "epoch": 0.5826239996457944, + "grad_norm": 1.7079758644104004, + "learning_rate": 8.864999247565719e-06, + "loss": 0.6473, + "step": 13159 + }, + { + "epoch": 0.5826682753506083, + "grad_norm": 1.417886734008789, + "learning_rate": 8.863463796917592e-06, + "loss": 0.5711, + "step": 13160 + }, + { + "epoch": 0.5827125510554221, + "grad_norm": 1.7446434497833252, + "learning_rate": 8.861928373414704e-06, + "loss": 0.6148, + "step": 13161 + }, + { + "epoch": 0.582756826760236, + "grad_norm": 1.4070796966552734, + "learning_rate": 8.860392977093732e-06, + "loss": 0.5497, + "step": 13162 + }, + { + "epoch": 0.5828011024650499, + "grad_norm": 1.7070411443710327, + "learning_rate": 8.858857607991347e-06, + "loss": 0.6544, + "step": 13163 + }, + { + "epoch": 0.5828453781698637, + "grad_norm": 1.5067923069000244, + "learning_rate": 8.857322266144217e-06, + "loss": 0.6039, + "step": 13164 + }, + { + "epoch": 0.5828896538746776, + "grad_norm": 1.9803587198257446, + "learning_rate": 8.855786951589015e-06, + "loss": 0.7081, + "step": 13165 + }, + { + "epoch": 0.5829339295794915, + "grad_norm": 1.9890851974487305, + "learning_rate": 8.854251664362412e-06, + "loss": 0.6727, + "step": 13166 + }, + { + "epoch": 0.5829782052843053, + "grad_norm": 1.3339691162109375, + "learning_rate": 8.852716404501068e-06, + "loss": 0.5653, + "step": 13167 + }, + { + "epoch": 0.5830224809891192, + "grad_norm": 1.4364559650421143, + "learning_rate": 8.851181172041664e-06, + "loss": 0.5915, + "step": 13168 + }, + { + "epoch": 0.5830667566939332, + "grad_norm": 1.8115755319595337, + "learning_rate": 8.849645967020864e-06, + "loss": 0.6345, + "step": 13169 + }, + { + "epoch": 0.583111032398747, + "grad_norm": 1.9354355335235596, + "learning_rate": 8.848110789475334e-06, + "loss": 0.6425, + "step": 13170 + }, + { + "epoch": 0.5831553081035609, + "grad_norm": 1.3607434034347534, + "learning_rate": 8.846575639441732e-06, + "loss": 0.6227, + "step": 13171 + }, + { + "epoch": 0.5831995838083748, + "grad_norm": 1.4183201789855957, + "learning_rate": 8.845040516956739e-06, + "loss": 0.5818, + "step": 13172 + }, + { + "epoch": 0.5832438595131886, + "grad_norm": 1.4408196210861206, + "learning_rate": 8.84350542205701e-06, + "loss": 0.6556, + "step": 13173 + }, + { + "epoch": 0.5832881352180025, + "grad_norm": 1.5602456331253052, + "learning_rate": 8.841970354779212e-06, + "loss": 0.6008, + "step": 13174 + }, + { + "epoch": 0.5833324109228164, + "grad_norm": 1.3000339269638062, + "learning_rate": 8.840435315160013e-06, + "loss": 0.5304, + "step": 13175 + }, + { + "epoch": 0.5833766866276302, + "grad_norm": 1.4703969955444336, + "learning_rate": 8.838900303236068e-06, + "loss": 0.6299, + "step": 13176 + }, + { + "epoch": 0.5834209623324441, + "grad_norm": 1.366344690322876, + "learning_rate": 8.837365319044042e-06, + "loss": 0.6047, + "step": 13177 + }, + { + "epoch": 0.583465238037258, + "grad_norm": 1.2608832120895386, + "learning_rate": 8.835830362620605e-06, + "loss": 0.5435, + "step": 13178 + }, + { + "epoch": 0.5835095137420718, + "grad_norm": 1.5734745264053345, + "learning_rate": 8.83429543400241e-06, + "loss": 0.6449, + "step": 13179 + }, + { + "epoch": 0.5835537894468857, + "grad_norm": 1.4135453701019287, + "learning_rate": 8.832760533226119e-06, + "loss": 0.6668, + "step": 13180 + }, + { + "epoch": 0.5835980651516997, + "grad_norm": 1.5288808345794678, + "learning_rate": 8.83122566032839e-06, + "loss": 0.656, + "step": 13181 + }, + { + "epoch": 0.5836423408565136, + "grad_norm": 1.5116422176361084, + "learning_rate": 8.829690815345886e-06, + "loss": 0.6203, + "step": 13182 + }, + { + "epoch": 0.5836866165613274, + "grad_norm": 1.3517637252807617, + "learning_rate": 8.828155998315264e-06, + "loss": 0.565, + "step": 13183 + }, + { + "epoch": 0.5837308922661413, + "grad_norm": 1.505966305732727, + "learning_rate": 8.826621209273181e-06, + "loss": 0.5639, + "step": 13184 + }, + { + "epoch": 0.5837751679709552, + "grad_norm": 1.3515914678573608, + "learning_rate": 8.825086448256296e-06, + "loss": 0.6053, + "step": 13185 + }, + { + "epoch": 0.583819443675769, + "grad_norm": 1.8222055435180664, + "learning_rate": 8.823551715301266e-06, + "loss": 0.6643, + "step": 13186 + }, + { + "epoch": 0.5838637193805829, + "grad_norm": 1.4780696630477905, + "learning_rate": 8.822017010444742e-06, + "loss": 0.5894, + "step": 13187 + }, + { + "epoch": 0.5839079950853968, + "grad_norm": 1.371974229812622, + "learning_rate": 8.820482333723385e-06, + "loss": 0.6238, + "step": 13188 + }, + { + "epoch": 0.5839522707902106, + "grad_norm": 1.425650954246521, + "learning_rate": 8.818947685173848e-06, + "loss": 0.6017, + "step": 13189 + }, + { + "epoch": 0.5839965464950245, + "grad_norm": 1.472849726676941, + "learning_rate": 8.817413064832783e-06, + "loss": 0.603, + "step": 13190 + }, + { + "epoch": 0.5840408221998384, + "grad_norm": 1.7085812091827393, + "learning_rate": 8.81587847273684e-06, + "loss": 0.6601, + "step": 13191 + }, + { + "epoch": 0.5840850979046522, + "grad_norm": 1.4230555295944214, + "learning_rate": 8.814343908922679e-06, + "loss": 0.6384, + "step": 13192 + }, + { + "epoch": 0.5841293736094662, + "grad_norm": 1.3525696992874146, + "learning_rate": 8.812809373426951e-06, + "loss": 0.6333, + "step": 13193 + }, + { + "epoch": 0.5841736493142801, + "grad_norm": 1.797103762626648, + "learning_rate": 8.8112748662863e-06, + "loss": 0.6477, + "step": 13194 + }, + { + "epoch": 0.5842179250190939, + "grad_norm": 1.8155416250228882, + "learning_rate": 8.809740387537383e-06, + "loss": 0.6596, + "step": 13195 + }, + { + "epoch": 0.5842622007239078, + "grad_norm": 1.8847864866256714, + "learning_rate": 8.808205937216847e-06, + "loss": 0.676, + "step": 13196 + }, + { + "epoch": 0.5843064764287217, + "grad_norm": 1.556467056274414, + "learning_rate": 8.806671515361344e-06, + "loss": 0.66, + "step": 13197 + }, + { + "epoch": 0.5843507521335355, + "grad_norm": 1.476359486579895, + "learning_rate": 8.805137122007516e-06, + "loss": 0.6079, + "step": 13198 + }, + { + "epoch": 0.5843950278383494, + "grad_norm": 1.4174928665161133, + "learning_rate": 8.803602757192019e-06, + "loss": 0.562, + "step": 13199 + }, + { + "epoch": 0.5844393035431633, + "grad_norm": 1.719051718711853, + "learning_rate": 8.802068420951494e-06, + "loss": 0.6403, + "step": 13200 + }, + { + "epoch": 0.5844835792479771, + "grad_norm": 1.3382364511489868, + "learning_rate": 8.800534113322586e-06, + "loss": 0.5349, + "step": 13201 + }, + { + "epoch": 0.584527854952791, + "grad_norm": 1.3208459615707397, + "learning_rate": 8.798999834341951e-06, + "loss": 0.5897, + "step": 13202 + }, + { + "epoch": 0.5845721306576049, + "grad_norm": 1.308498740196228, + "learning_rate": 8.797465584046225e-06, + "loss": 0.5101, + "step": 13203 + }, + { + "epoch": 0.5846164063624187, + "grad_norm": 1.5370163917541504, + "learning_rate": 8.795931362472049e-06, + "loss": 0.6524, + "step": 13204 + }, + { + "epoch": 0.5846606820672327, + "grad_norm": 1.5662952661514282, + "learning_rate": 8.794397169656076e-06, + "loss": 0.6321, + "step": 13205 + }, + { + "epoch": 0.5847049577720466, + "grad_norm": 1.363456130027771, + "learning_rate": 8.792863005634947e-06, + "loss": 0.6223, + "step": 13206 + }, + { + "epoch": 0.5847492334768604, + "grad_norm": 1.4922822713851929, + "learning_rate": 8.791328870445302e-06, + "loss": 0.5081, + "step": 13207 + }, + { + "epoch": 0.5847935091816743, + "grad_norm": 1.5204694271087646, + "learning_rate": 8.789794764123779e-06, + "loss": 0.6123, + "step": 13208 + }, + { + "epoch": 0.5848377848864882, + "grad_norm": 1.5744246244430542, + "learning_rate": 8.788260686707026e-06, + "loss": 0.5993, + "step": 13209 + }, + { + "epoch": 0.584882060591302, + "grad_norm": 1.4955954551696777, + "learning_rate": 8.78672663823168e-06, + "loss": 0.571, + "step": 13210 + }, + { + "epoch": 0.5849263362961159, + "grad_norm": 1.9473668336868286, + "learning_rate": 8.785192618734378e-06, + "loss": 0.6619, + "step": 13211 + }, + { + "epoch": 0.5849706120009298, + "grad_norm": 1.4204903841018677, + "learning_rate": 8.783658628251762e-06, + "loss": 0.6483, + "step": 13212 + }, + { + "epoch": 0.5850148877057436, + "grad_norm": 1.4736413955688477, + "learning_rate": 8.782124666820471e-06, + "loss": 0.6096, + "step": 13213 + }, + { + "epoch": 0.5850591634105575, + "grad_norm": 1.3558390140533447, + "learning_rate": 8.780590734477135e-06, + "loss": 0.5605, + "step": 13214 + }, + { + "epoch": 0.5851034391153714, + "grad_norm": 1.4665788412094116, + "learning_rate": 8.779056831258402e-06, + "loss": 0.6303, + "step": 13215 + }, + { + "epoch": 0.5851477148201852, + "grad_norm": 1.4787907600402832, + "learning_rate": 8.7775229572009e-06, + "loss": 0.6354, + "step": 13216 + }, + { + "epoch": 0.5851919905249992, + "grad_norm": 1.237047553062439, + "learning_rate": 8.775989112341272e-06, + "loss": 0.618, + "step": 13217 + }, + { + "epoch": 0.5852362662298131, + "grad_norm": 1.7858725786209106, + "learning_rate": 8.77445529671614e-06, + "loss": 0.6095, + "step": 13218 + }, + { + "epoch": 0.585280541934627, + "grad_norm": 1.3011952638626099, + "learning_rate": 8.77292151036215e-06, + "loss": 0.6673, + "step": 13219 + }, + { + "epoch": 0.5853248176394408, + "grad_norm": 1.7233179807662964, + "learning_rate": 8.77138775331593e-06, + "loss": 0.6649, + "step": 13220 + }, + { + "epoch": 0.5853690933442547, + "grad_norm": 1.5156539678573608, + "learning_rate": 8.76985402561411e-06, + "loss": 0.6577, + "step": 13221 + }, + { + "epoch": 0.5854133690490686, + "grad_norm": 1.4741168022155762, + "learning_rate": 8.76832032729333e-06, + "loss": 0.577, + "step": 13222 + }, + { + "epoch": 0.5854576447538824, + "grad_norm": 1.3644695281982422, + "learning_rate": 8.766786658390213e-06, + "loss": 0.6045, + "step": 13223 + }, + { + "epoch": 0.5855019204586963, + "grad_norm": 1.330167531967163, + "learning_rate": 8.76525301894139e-06, + "loss": 0.6143, + "step": 13224 + }, + { + "epoch": 0.5855461961635102, + "grad_norm": 1.264654278755188, + "learning_rate": 8.763719408983496e-06, + "loss": 0.5879, + "step": 13225 + }, + { + "epoch": 0.585590471868324, + "grad_norm": 1.356258511543274, + "learning_rate": 8.762185828553161e-06, + "loss": 0.6136, + "step": 13226 + }, + { + "epoch": 0.5856347475731379, + "grad_norm": 1.2927534580230713, + "learning_rate": 8.760652277687007e-06, + "loss": 0.6028, + "step": 13227 + }, + { + "epoch": 0.5856790232779518, + "grad_norm": 1.1868504285812378, + "learning_rate": 8.75911875642166e-06, + "loss": 0.5282, + "step": 13228 + }, + { + "epoch": 0.5857232989827656, + "grad_norm": 1.4605907201766968, + "learning_rate": 8.757585264793756e-06, + "loss": 0.5696, + "step": 13229 + }, + { + "epoch": 0.5857675746875796, + "grad_norm": 1.4496837854385376, + "learning_rate": 8.756051802839915e-06, + "loss": 0.5892, + "step": 13230 + }, + { + "epoch": 0.5858118503923935, + "grad_norm": 1.4187638759613037, + "learning_rate": 8.754518370596764e-06, + "loss": 0.5745, + "step": 13231 + }, + { + "epoch": 0.5858561260972073, + "grad_norm": 1.39894437789917, + "learning_rate": 8.752984968100927e-06, + "loss": 0.5939, + "step": 13232 + }, + { + "epoch": 0.5859004018020212, + "grad_norm": 1.3007960319519043, + "learning_rate": 8.75145159538903e-06, + "loss": 0.565, + "step": 13233 + }, + { + "epoch": 0.5859446775068351, + "grad_norm": 1.3327585458755493, + "learning_rate": 8.749918252497694e-06, + "loss": 0.5712, + "step": 13234 + }, + { + "epoch": 0.5859889532116489, + "grad_norm": 1.503256916999817, + "learning_rate": 8.748384939463543e-06, + "loss": 0.6101, + "step": 13235 + }, + { + "epoch": 0.5860332289164628, + "grad_norm": 1.700270414352417, + "learning_rate": 8.746851656323198e-06, + "loss": 0.6333, + "step": 13236 + }, + { + "epoch": 0.5860775046212767, + "grad_norm": 1.7562134265899658, + "learning_rate": 8.745318403113283e-06, + "loss": 0.6665, + "step": 13237 + }, + { + "epoch": 0.5861217803260905, + "grad_norm": 1.760711908340454, + "learning_rate": 8.74378517987041e-06, + "loss": 0.636, + "step": 13238 + }, + { + "epoch": 0.5861660560309044, + "grad_norm": 1.3278449773788452, + "learning_rate": 8.742251986631211e-06, + "loss": 0.5777, + "step": 13239 + }, + { + "epoch": 0.5862103317357183, + "grad_norm": 1.7170659303665161, + "learning_rate": 8.740718823432299e-06, + "loss": 0.683, + "step": 13240 + }, + { + "epoch": 0.5862546074405321, + "grad_norm": 1.3641538619995117, + "learning_rate": 8.73918569031029e-06, + "loss": 0.6035, + "step": 13241 + }, + { + "epoch": 0.5862988831453461, + "grad_norm": 1.4771746397018433, + "learning_rate": 8.737652587301807e-06, + "loss": 0.615, + "step": 13242 + }, + { + "epoch": 0.58634315885016, + "grad_norm": 1.2411051988601685, + "learning_rate": 8.736119514443463e-06, + "loss": 0.537, + "step": 13243 + }, + { + "epoch": 0.5863874345549738, + "grad_norm": 1.7127724885940552, + "learning_rate": 8.734586471771874e-06, + "loss": 0.6946, + "step": 13244 + }, + { + "epoch": 0.5864317102597877, + "grad_norm": 1.256176233291626, + "learning_rate": 8.733053459323659e-06, + "loss": 0.5059, + "step": 13245 + }, + { + "epoch": 0.5864759859646016, + "grad_norm": 1.5021133422851562, + "learning_rate": 8.73152047713543e-06, + "loss": 0.6119, + "step": 13246 + }, + { + "epoch": 0.5865202616694154, + "grad_norm": 1.6103489398956299, + "learning_rate": 8.729987525243801e-06, + "loss": 0.627, + "step": 13247 + }, + { + "epoch": 0.5865645373742293, + "grad_norm": 1.4167790412902832, + "learning_rate": 8.728454603685383e-06, + "loss": 0.5181, + "step": 13248 + }, + { + "epoch": 0.5866088130790432, + "grad_norm": 1.6142970323562622, + "learning_rate": 8.726921712496797e-06, + "loss": 0.5811, + "step": 13249 + }, + { + "epoch": 0.586653088783857, + "grad_norm": 1.8565386533737183, + "learning_rate": 8.725388851714647e-06, + "loss": 0.624, + "step": 13250 + }, + { + "epoch": 0.5866973644886709, + "grad_norm": 1.161916732788086, + "learning_rate": 8.723856021375543e-06, + "loss": 0.5253, + "step": 13251 + }, + { + "epoch": 0.5867416401934848, + "grad_norm": 1.4844125509262085, + "learning_rate": 8.722323221516103e-06, + "loss": 0.5482, + "step": 13252 + }, + { + "epoch": 0.5867859158982988, + "grad_norm": 1.7404093742370605, + "learning_rate": 8.720790452172932e-06, + "loss": 0.5445, + "step": 13253 + }, + { + "epoch": 0.5868301916031126, + "grad_norm": 1.7621300220489502, + "learning_rate": 8.719257713382638e-06, + "loss": 0.6404, + "step": 13254 + }, + { + "epoch": 0.5868744673079265, + "grad_norm": 1.4834460020065308, + "learning_rate": 8.717725005181832e-06, + "loss": 0.642, + "step": 13255 + }, + { + "epoch": 0.5869187430127404, + "grad_norm": 1.5918219089508057, + "learning_rate": 8.71619232760712e-06, + "loss": 0.6315, + "step": 13256 + }, + { + "epoch": 0.5869630187175542, + "grad_norm": 1.4435555934906006, + "learning_rate": 8.71465968069511e-06, + "loss": 0.5471, + "step": 13257 + }, + { + "epoch": 0.5870072944223681, + "grad_norm": 1.43893563747406, + "learning_rate": 8.713127064482404e-06, + "loss": 0.6038, + "step": 13258 + }, + { + "epoch": 0.587051570127182, + "grad_norm": 1.4877849817276, + "learning_rate": 8.711594479005614e-06, + "loss": 0.6323, + "step": 13259 + }, + { + "epoch": 0.5870958458319958, + "grad_norm": 1.2861372232437134, + "learning_rate": 8.710061924301339e-06, + "loss": 0.5892, + "step": 13260 + }, + { + "epoch": 0.5871401215368097, + "grad_norm": 3.4765007495880127, + "learning_rate": 8.70852940040618e-06, + "loss": 0.5921, + "step": 13261 + }, + { + "epoch": 0.5871843972416236, + "grad_norm": 1.8376836776733398, + "learning_rate": 8.70699690735675e-06, + "loss": 0.6228, + "step": 13262 + }, + { + "epoch": 0.5872286729464374, + "grad_norm": 2.484348773956299, + "learning_rate": 8.705464445189646e-06, + "loss": 0.6891, + "step": 13263 + }, + { + "epoch": 0.5872729486512513, + "grad_norm": 1.8402206897735596, + "learning_rate": 8.703932013941468e-06, + "loss": 0.7159, + "step": 13264 + }, + { + "epoch": 0.5873172243560653, + "grad_norm": 1.4585034847259521, + "learning_rate": 8.702399613648821e-06, + "loss": 0.6333, + "step": 13265 + }, + { + "epoch": 0.5873615000608791, + "grad_norm": 1.4917296171188354, + "learning_rate": 8.700867244348302e-06, + "loss": 0.5922, + "step": 13266 + }, + { + "epoch": 0.587405775765693, + "grad_norm": 1.5492068529129028, + "learning_rate": 8.699334906076512e-06, + "loss": 0.6559, + "step": 13267 + }, + { + "epoch": 0.5874500514705069, + "grad_norm": 1.3074393272399902, + "learning_rate": 8.697802598870045e-06, + "loss": 0.6003, + "step": 13268 + }, + { + "epoch": 0.5874943271753207, + "grad_norm": 1.3364219665527344, + "learning_rate": 8.696270322765506e-06, + "loss": 0.628, + "step": 13269 + }, + { + "epoch": 0.5875386028801346, + "grad_norm": 1.6190991401672363, + "learning_rate": 8.694738077799487e-06, + "loss": 0.6578, + "step": 13270 + }, + { + "epoch": 0.5875828785849485, + "grad_norm": 1.5323177576065063, + "learning_rate": 8.693205864008582e-06, + "loss": 0.6578, + "step": 13271 + }, + { + "epoch": 0.5876271542897623, + "grad_norm": 1.355013132095337, + "learning_rate": 8.691673681429398e-06, + "loss": 0.5894, + "step": 13272 + }, + { + "epoch": 0.5876714299945762, + "grad_norm": 1.4893543720245361, + "learning_rate": 8.690141530098523e-06, + "loss": 0.5932, + "step": 13273 + }, + { + "epoch": 0.5877157056993901, + "grad_norm": 1.5944048166275024, + "learning_rate": 8.688609410052544e-06, + "loss": 0.5421, + "step": 13274 + }, + { + "epoch": 0.5877599814042039, + "grad_norm": 1.6024911403656006, + "learning_rate": 8.687077321328066e-06, + "loss": 0.6401, + "step": 13275 + }, + { + "epoch": 0.5878042571090178, + "grad_norm": 1.6185994148254395, + "learning_rate": 8.685545263961679e-06, + "loss": 0.6634, + "step": 13276 + }, + { + "epoch": 0.5878485328138318, + "grad_norm": 1.4538410902023315, + "learning_rate": 8.684013237989972e-06, + "loss": 0.604, + "step": 13277 + }, + { + "epoch": 0.5878928085186456, + "grad_norm": 1.2678265571594238, + "learning_rate": 8.682481243449534e-06, + "loss": 0.5854, + "step": 13278 + }, + { + "epoch": 0.5879370842234595, + "grad_norm": 1.5510492324829102, + "learning_rate": 8.680949280376963e-06, + "loss": 0.6211, + "step": 13279 + }, + { + "epoch": 0.5879813599282734, + "grad_norm": 1.4019986391067505, + "learning_rate": 8.679417348808844e-06, + "loss": 0.5801, + "step": 13280 + }, + { + "epoch": 0.5880256356330872, + "grad_norm": 1.6290791034698486, + "learning_rate": 8.67788544878176e-06, + "loss": 0.6022, + "step": 13281 + }, + { + "epoch": 0.5880699113379011, + "grad_norm": 1.3982237577438354, + "learning_rate": 8.676353580332313e-06, + "loss": 0.5987, + "step": 13282 + }, + { + "epoch": 0.588114187042715, + "grad_norm": 1.3463807106018066, + "learning_rate": 8.674821743497082e-06, + "loss": 0.5862, + "step": 13283 + }, + { + "epoch": 0.5881584627475288, + "grad_norm": 1.394331455230713, + "learning_rate": 8.67328993831265e-06, + "loss": 0.5527, + "step": 13284 + }, + { + "epoch": 0.5882027384523427, + "grad_norm": 2.020116090774536, + "learning_rate": 8.67175816481561e-06, + "loss": 0.6386, + "step": 13285 + }, + { + "epoch": 0.5882470141571566, + "grad_norm": 1.1822421550750732, + "learning_rate": 8.670226423042547e-06, + "loss": 0.5449, + "step": 13286 + }, + { + "epoch": 0.5882912898619705, + "grad_norm": 1.5283366441726685, + "learning_rate": 8.668694713030042e-06, + "loss": 0.5887, + "step": 13287 + }, + { + "epoch": 0.5883355655667843, + "grad_norm": 1.859570026397705, + "learning_rate": 8.667163034814678e-06, + "loss": 0.6656, + "step": 13288 + }, + { + "epoch": 0.5883798412715983, + "grad_norm": 1.5287702083587646, + "learning_rate": 8.665631388433042e-06, + "loss": 0.6009, + "step": 13289 + }, + { + "epoch": 0.5884241169764122, + "grad_norm": 1.3127562999725342, + "learning_rate": 8.664099773921715e-06, + "loss": 0.5614, + "step": 13290 + }, + { + "epoch": 0.588468392681226, + "grad_norm": 1.4258865118026733, + "learning_rate": 8.662568191317273e-06, + "loss": 0.6133, + "step": 13291 + }, + { + "epoch": 0.5885126683860399, + "grad_norm": 1.4295663833618164, + "learning_rate": 8.661036640656305e-06, + "loss": 0.5781, + "step": 13292 + }, + { + "epoch": 0.5885569440908538, + "grad_norm": 1.2703931331634521, + "learning_rate": 8.659505121975383e-06, + "loss": 0.5396, + "step": 13293 + }, + { + "epoch": 0.5886012197956676, + "grad_norm": 1.4078437089920044, + "learning_rate": 8.657973635311087e-06, + "loss": 0.6328, + "step": 13294 + }, + { + "epoch": 0.5886454955004815, + "grad_norm": 1.5792920589447021, + "learning_rate": 8.656442180700004e-06, + "loss": 0.628, + "step": 13295 + }, + { + "epoch": 0.5886897712052954, + "grad_norm": 1.4401915073394775, + "learning_rate": 8.654910758178705e-06, + "loss": 0.6192, + "step": 13296 + }, + { + "epoch": 0.5887340469101092, + "grad_norm": 1.473022699356079, + "learning_rate": 8.653379367783769e-06, + "loss": 0.6045, + "step": 13297 + }, + { + "epoch": 0.5887783226149231, + "grad_norm": 1.3319095373153687, + "learning_rate": 8.651848009551764e-06, + "loss": 0.615, + "step": 13298 + }, + { + "epoch": 0.588822598319737, + "grad_norm": 1.461938500404358, + "learning_rate": 8.650316683519276e-06, + "loss": 0.5989, + "step": 13299 + }, + { + "epoch": 0.5888668740245508, + "grad_norm": 1.3109315633773804, + "learning_rate": 8.648785389722873e-06, + "loss": 0.5679, + "step": 13300 + }, + { + "epoch": 0.5889111497293648, + "grad_norm": 1.6621705293655396, + "learning_rate": 8.647254128199131e-06, + "loss": 0.6913, + "step": 13301 + }, + { + "epoch": 0.5889554254341787, + "grad_norm": 1.4628276824951172, + "learning_rate": 8.645722898984625e-06, + "loss": 0.5589, + "step": 13302 + }, + { + "epoch": 0.5889997011389925, + "grad_norm": 1.6000179052352905, + "learning_rate": 8.644191702115923e-06, + "loss": 0.6605, + "step": 13303 + }, + { + "epoch": 0.5890439768438064, + "grad_norm": 1.3575396537780762, + "learning_rate": 8.642660537629596e-06, + "loss": 0.6239, + "step": 13304 + }, + { + "epoch": 0.5890882525486203, + "grad_norm": 1.5408638715744019, + "learning_rate": 8.641129405562221e-06, + "loss": 0.5508, + "step": 13305 + }, + { + "epoch": 0.5891325282534341, + "grad_norm": 1.2816914319992065, + "learning_rate": 8.639598305950365e-06, + "loss": 0.5982, + "step": 13306 + }, + { + "epoch": 0.589176803958248, + "grad_norm": 1.6369043588638306, + "learning_rate": 8.638067238830591e-06, + "loss": 0.5869, + "step": 13307 + }, + { + "epoch": 0.5892210796630619, + "grad_norm": 1.4462522268295288, + "learning_rate": 8.636536204239472e-06, + "loss": 0.5582, + "step": 13308 + }, + { + "epoch": 0.5892653553678757, + "grad_norm": 1.2787866592407227, + "learning_rate": 8.635005202213578e-06, + "loss": 0.5663, + "step": 13309 + }, + { + "epoch": 0.5893096310726896, + "grad_norm": 1.2585536241531372, + "learning_rate": 8.633474232789474e-06, + "loss": 0.5703, + "step": 13310 + }, + { + "epoch": 0.5893539067775035, + "grad_norm": 1.4621613025665283, + "learning_rate": 8.631943296003723e-06, + "loss": 0.5665, + "step": 13311 + }, + { + "epoch": 0.5893981824823173, + "grad_norm": 1.342457890510559, + "learning_rate": 8.630412391892894e-06, + "loss": 0.6233, + "step": 13312 + }, + { + "epoch": 0.5894424581871313, + "grad_norm": 1.219280481338501, + "learning_rate": 8.628881520493551e-06, + "loss": 0.583, + "step": 13313 + }, + { + "epoch": 0.5894867338919452, + "grad_norm": 1.6394953727722168, + "learning_rate": 8.627350681842256e-06, + "loss": 0.6254, + "step": 13314 + }, + { + "epoch": 0.589531009596759, + "grad_norm": 1.5128486156463623, + "learning_rate": 8.62581987597557e-06, + "loss": 0.6151, + "step": 13315 + }, + { + "epoch": 0.5895752853015729, + "grad_norm": 1.4743598699569702, + "learning_rate": 8.62428910293006e-06, + "loss": 0.565, + "step": 13316 + }, + { + "epoch": 0.5896195610063868, + "grad_norm": 1.4665560722351074, + "learning_rate": 8.622758362742283e-06, + "loss": 0.5369, + "step": 13317 + }, + { + "epoch": 0.5896638367112006, + "grad_norm": 1.4052354097366333, + "learning_rate": 8.621227655448798e-06, + "loss": 0.6002, + "step": 13318 + }, + { + "epoch": 0.5897081124160145, + "grad_norm": 1.5094408988952637, + "learning_rate": 8.619696981086173e-06, + "loss": 0.6039, + "step": 13319 + }, + { + "epoch": 0.5897523881208284, + "grad_norm": 1.4708952903747559, + "learning_rate": 8.618166339690963e-06, + "loss": 0.587, + "step": 13320 + }, + { + "epoch": 0.5897966638256422, + "grad_norm": 1.4433423280715942, + "learning_rate": 8.616635731299716e-06, + "loss": 0.5775, + "step": 13321 + }, + { + "epoch": 0.5898409395304561, + "grad_norm": 1.452024221420288, + "learning_rate": 8.615105155949006e-06, + "loss": 0.5975, + "step": 13322 + }, + { + "epoch": 0.58988521523527, + "grad_norm": 1.5635851621627808, + "learning_rate": 8.61357461367538e-06, + "loss": 0.6414, + "step": 13323 + }, + { + "epoch": 0.589929490940084, + "grad_norm": 1.437382459640503, + "learning_rate": 8.612044104515396e-06, + "loss": 0.607, + "step": 13324 + }, + { + "epoch": 0.5899737666448978, + "grad_norm": 1.2282578945159912, + "learning_rate": 8.610513628505605e-06, + "loss": 0.5771, + "step": 13325 + }, + { + "epoch": 0.5900180423497117, + "grad_norm": 1.7110910415649414, + "learning_rate": 8.608983185682568e-06, + "loss": 0.7024, + "step": 13326 + }, + { + "epoch": 0.5900623180545256, + "grad_norm": 1.8964961767196655, + "learning_rate": 8.607452776082836e-06, + "loss": 0.6439, + "step": 13327 + }, + { + "epoch": 0.5901065937593394, + "grad_norm": 1.3838508129119873, + "learning_rate": 8.605922399742956e-06, + "loss": 0.573, + "step": 13328 + }, + { + "epoch": 0.5901508694641533, + "grad_norm": 1.649639368057251, + "learning_rate": 8.60439205669949e-06, + "loss": 0.6332, + "step": 13329 + }, + { + "epoch": 0.5901951451689672, + "grad_norm": 1.7069340944290161, + "learning_rate": 8.602861746988981e-06, + "loss": 0.7067, + "step": 13330 + }, + { + "epoch": 0.590239420873781, + "grad_norm": 1.6908648014068604, + "learning_rate": 8.601331470647977e-06, + "loss": 0.602, + "step": 13331 + }, + { + "epoch": 0.5902836965785949, + "grad_norm": 1.8294904232025146, + "learning_rate": 8.599801227713039e-06, + "loss": 0.6616, + "step": 13332 + }, + { + "epoch": 0.5903279722834088, + "grad_norm": 2.064796209335327, + "learning_rate": 8.598271018220705e-06, + "loss": 0.7131, + "step": 13333 + }, + { + "epoch": 0.5903722479882226, + "grad_norm": 1.5873299837112427, + "learning_rate": 8.596740842207529e-06, + "loss": 0.6645, + "step": 13334 + }, + { + "epoch": 0.5904165236930365, + "grad_norm": 1.499461054801941, + "learning_rate": 8.595210699710051e-06, + "loss": 0.6836, + "step": 13335 + }, + { + "epoch": 0.5904607993978505, + "grad_norm": 1.5446051359176636, + "learning_rate": 8.593680590764826e-06, + "loss": 0.6847, + "step": 13336 + }, + { + "epoch": 0.5905050751026643, + "grad_norm": 1.2574443817138672, + "learning_rate": 8.592150515408393e-06, + "loss": 0.559, + "step": 13337 + }, + { + "epoch": 0.5905493508074782, + "grad_norm": 1.541948914527893, + "learning_rate": 8.590620473677297e-06, + "loss": 0.6434, + "step": 13338 + }, + { + "epoch": 0.5905936265122921, + "grad_norm": 1.4245654344558716, + "learning_rate": 8.589090465608087e-06, + "loss": 0.584, + "step": 13339 + }, + { + "epoch": 0.5906379022171059, + "grad_norm": 1.6084539890289307, + "learning_rate": 8.587560491237301e-06, + "loss": 0.5964, + "step": 13340 + }, + { + "epoch": 0.5906821779219198, + "grad_norm": 1.6092265844345093, + "learning_rate": 8.58603055060148e-06, + "loss": 0.6203, + "step": 13341 + }, + { + "epoch": 0.5907264536267337, + "grad_norm": 1.4264512062072754, + "learning_rate": 8.58450064373717e-06, + "loss": 0.5947, + "step": 13342 + }, + { + "epoch": 0.5907707293315475, + "grad_norm": 1.5033457279205322, + "learning_rate": 8.582970770680911e-06, + "loss": 0.5898, + "step": 13343 + }, + { + "epoch": 0.5908150050363614, + "grad_norm": 1.5071601867675781, + "learning_rate": 8.581440931469244e-06, + "loss": 0.7034, + "step": 13344 + }, + { + "epoch": 0.5908592807411753, + "grad_norm": 1.7884621620178223, + "learning_rate": 8.579911126138697e-06, + "loss": 0.622, + "step": 13345 + }, + { + "epoch": 0.5909035564459891, + "grad_norm": 1.5851311683654785, + "learning_rate": 8.578381354725821e-06, + "loss": 0.6954, + "step": 13346 + }, + { + "epoch": 0.590947832150803, + "grad_norm": 1.3916656970977783, + "learning_rate": 8.576851617267151e-06, + "loss": 0.5864, + "step": 13347 + }, + { + "epoch": 0.590992107855617, + "grad_norm": 1.3278307914733887, + "learning_rate": 8.575321913799218e-06, + "loss": 0.5935, + "step": 13348 + }, + { + "epoch": 0.5910363835604308, + "grad_norm": 1.297935962677002, + "learning_rate": 8.573792244358564e-06, + "loss": 0.5864, + "step": 13349 + }, + { + "epoch": 0.5910806592652447, + "grad_norm": 1.1974154710769653, + "learning_rate": 8.572262608981719e-06, + "loss": 0.535, + "step": 13350 + }, + { + "epoch": 0.5911249349700586, + "grad_norm": 1.3683770895004272, + "learning_rate": 8.570733007705216e-06, + "loss": 0.5482, + "step": 13351 + }, + { + "epoch": 0.5911692106748724, + "grad_norm": 1.7783727645874023, + "learning_rate": 8.569203440565594e-06, + "loss": 0.6342, + "step": 13352 + }, + { + "epoch": 0.5912134863796863, + "grad_norm": 1.5388498306274414, + "learning_rate": 8.567673907599388e-06, + "loss": 0.5438, + "step": 13353 + }, + { + "epoch": 0.5912577620845002, + "grad_norm": 2.393815279006958, + "learning_rate": 8.56614440884312e-06, + "loss": 0.7644, + "step": 13354 + }, + { + "epoch": 0.591302037789314, + "grad_norm": 1.4932668209075928, + "learning_rate": 8.564614944333322e-06, + "loss": 0.6014, + "step": 13355 + }, + { + "epoch": 0.5913463134941279, + "grad_norm": 1.4797065258026123, + "learning_rate": 8.563085514106532e-06, + "loss": 0.573, + "step": 13356 + }, + { + "epoch": 0.5913905891989418, + "grad_norm": 1.3840810060501099, + "learning_rate": 8.561556118199273e-06, + "loss": 0.5616, + "step": 13357 + }, + { + "epoch": 0.5914348649037557, + "grad_norm": 1.1428771018981934, + "learning_rate": 8.560026756648075e-06, + "loss": 0.5229, + "step": 13358 + }, + { + "epoch": 0.5914791406085695, + "grad_norm": 1.4160683155059814, + "learning_rate": 8.558497429489465e-06, + "loss": 0.5622, + "step": 13359 + }, + { + "epoch": 0.5915234163133835, + "grad_norm": 1.6767252683639526, + "learning_rate": 8.556968136759972e-06, + "loss": 0.5993, + "step": 13360 + }, + { + "epoch": 0.5915676920181974, + "grad_norm": 1.7979999780654907, + "learning_rate": 8.555438878496117e-06, + "loss": 0.6836, + "step": 13361 + }, + { + "epoch": 0.5916119677230112, + "grad_norm": 1.6436702013015747, + "learning_rate": 8.553909654734431e-06, + "loss": 0.5914, + "step": 13362 + }, + { + "epoch": 0.5916562434278251, + "grad_norm": 1.2792023420333862, + "learning_rate": 8.552380465511435e-06, + "loss": 0.5436, + "step": 13363 + }, + { + "epoch": 0.591700519132639, + "grad_norm": 1.4714996814727783, + "learning_rate": 8.550851310863652e-06, + "loss": 0.6009, + "step": 13364 + }, + { + "epoch": 0.5917447948374528, + "grad_norm": 1.577863335609436, + "learning_rate": 8.549322190827602e-06, + "loss": 0.6242, + "step": 13365 + }, + { + "epoch": 0.5917890705422667, + "grad_norm": 1.4359034299850464, + "learning_rate": 8.547793105439816e-06, + "loss": 0.6115, + "step": 13366 + }, + { + "epoch": 0.5918333462470806, + "grad_norm": 1.434256672859192, + "learning_rate": 8.546264054736809e-06, + "loss": 0.6112, + "step": 13367 + }, + { + "epoch": 0.5918776219518944, + "grad_norm": 1.582689881324768, + "learning_rate": 8.544735038755097e-06, + "loss": 0.5616, + "step": 13368 + }, + { + "epoch": 0.5919218976567083, + "grad_norm": 1.4376215934753418, + "learning_rate": 8.543206057531207e-06, + "loss": 0.583, + "step": 13369 + }, + { + "epoch": 0.5919661733615222, + "grad_norm": 1.2748019695281982, + "learning_rate": 8.541677111101655e-06, + "loss": 0.5783, + "step": 13370 + }, + { + "epoch": 0.592010449066336, + "grad_norm": 1.3859789371490479, + "learning_rate": 8.540148199502955e-06, + "loss": 0.6375, + "step": 13371 + }, + { + "epoch": 0.59205472477115, + "grad_norm": 1.5128406286239624, + "learning_rate": 8.538619322771629e-06, + "loss": 0.5418, + "step": 13372 + }, + { + "epoch": 0.5920990004759639, + "grad_norm": 1.696852207183838, + "learning_rate": 8.53709048094419e-06, + "loss": 0.635, + "step": 13373 + }, + { + "epoch": 0.5921432761807777, + "grad_norm": 1.430746078491211, + "learning_rate": 8.535561674057154e-06, + "loss": 0.5892, + "step": 13374 + }, + { + "epoch": 0.5921875518855916, + "grad_norm": 1.431660771369934, + "learning_rate": 8.53403290214703e-06, + "loss": 0.6227, + "step": 13375 + }, + { + "epoch": 0.5922318275904055, + "grad_norm": 1.4800037145614624, + "learning_rate": 8.532504165250344e-06, + "loss": 0.5938, + "step": 13376 + }, + { + "epoch": 0.5922761032952193, + "grad_norm": 1.6891716718673706, + "learning_rate": 8.530975463403599e-06, + "loss": 0.6422, + "step": 13377 + }, + { + "epoch": 0.5923203790000332, + "grad_norm": 1.5658115148544312, + "learning_rate": 8.529446796643304e-06, + "loss": 0.6461, + "step": 13378 + }, + { + "epoch": 0.5923646547048471, + "grad_norm": 1.5697335004806519, + "learning_rate": 8.52791816500598e-06, + "loss": 0.5794, + "step": 13379 + }, + { + "epoch": 0.5924089304096609, + "grad_norm": 1.6504191160202026, + "learning_rate": 8.52638956852813e-06, + "loss": 0.6528, + "step": 13380 + }, + { + "epoch": 0.5924532061144748, + "grad_norm": 1.5509105920791626, + "learning_rate": 8.524861007246266e-06, + "loss": 0.5926, + "step": 13381 + }, + { + "epoch": 0.5924974818192887, + "grad_norm": 1.5537577867507935, + "learning_rate": 8.523332481196896e-06, + "loss": 0.5991, + "step": 13382 + }, + { + "epoch": 0.5925417575241025, + "grad_norm": 2.2748637199401855, + "learning_rate": 8.521803990416526e-06, + "loss": 0.7316, + "step": 13383 + }, + { + "epoch": 0.5925860332289165, + "grad_norm": 1.4436898231506348, + "learning_rate": 8.520275534941668e-06, + "loss": 0.6249, + "step": 13384 + }, + { + "epoch": 0.5926303089337304, + "grad_norm": 4.033997058868408, + "learning_rate": 8.51874711480882e-06, + "loss": 0.6071, + "step": 13385 + }, + { + "epoch": 0.5926745846385442, + "grad_norm": 1.5732519626617432, + "learning_rate": 8.517218730054493e-06, + "loss": 0.7162, + "step": 13386 + }, + { + "epoch": 0.5927188603433581, + "grad_norm": 1.3703923225402832, + "learning_rate": 8.515690380715189e-06, + "loss": 0.5079, + "step": 13387 + }, + { + "epoch": 0.592763136048172, + "grad_norm": 1.8015029430389404, + "learning_rate": 8.514162066827408e-06, + "loss": 0.6251, + "step": 13388 + }, + { + "epoch": 0.5928074117529858, + "grad_norm": 1.5618311166763306, + "learning_rate": 8.51263378842766e-06, + "loss": 0.6296, + "step": 13389 + }, + { + "epoch": 0.5928516874577997, + "grad_norm": 1.6642348766326904, + "learning_rate": 8.511105545552443e-06, + "loss": 0.6522, + "step": 13390 + }, + { + "epoch": 0.5928959631626136, + "grad_norm": 1.4205622673034668, + "learning_rate": 8.509577338238255e-06, + "loss": 0.6378, + "step": 13391 + }, + { + "epoch": 0.5929402388674274, + "grad_norm": 1.4393742084503174, + "learning_rate": 8.508049166521602e-06, + "loss": 0.6322, + "step": 13392 + }, + { + "epoch": 0.5929845145722413, + "grad_norm": 1.3573229312896729, + "learning_rate": 8.506521030438979e-06, + "loss": 0.5996, + "step": 13393 + }, + { + "epoch": 0.5930287902770552, + "grad_norm": 1.7943642139434814, + "learning_rate": 8.504992930026886e-06, + "loss": 0.6499, + "step": 13394 + }, + { + "epoch": 0.5930730659818692, + "grad_norm": 1.5693391561508179, + "learning_rate": 8.503464865321817e-06, + "loss": 0.6131, + "step": 13395 + }, + { + "epoch": 0.593117341686683, + "grad_norm": 1.338470697402954, + "learning_rate": 8.501936836360274e-06, + "loss": 0.5938, + "step": 13396 + }, + { + "epoch": 0.5931616173914969, + "grad_norm": 1.6500588655471802, + "learning_rate": 8.500408843178748e-06, + "loss": 0.6018, + "step": 13397 + }, + { + "epoch": 0.5932058930963108, + "grad_norm": 1.442760705947876, + "learning_rate": 8.498880885813734e-06, + "loss": 0.6428, + "step": 13398 + }, + { + "epoch": 0.5932501688011246, + "grad_norm": 1.506725549697876, + "learning_rate": 8.497352964301731e-06, + "loss": 0.6411, + "step": 13399 + }, + { + "epoch": 0.5932944445059385, + "grad_norm": 1.416445255279541, + "learning_rate": 8.495825078679231e-06, + "loss": 0.6607, + "step": 13400 + }, + { + "epoch": 0.5933387202107524, + "grad_norm": 1.2542777061462402, + "learning_rate": 8.494297228982717e-06, + "loss": 0.5318, + "step": 13401 + }, + { + "epoch": 0.5933829959155662, + "grad_norm": 1.4573396444320679, + "learning_rate": 8.492769415248693e-06, + "loss": 0.5989, + "step": 13402 + }, + { + "epoch": 0.5934272716203801, + "grad_norm": 1.5120307207107544, + "learning_rate": 8.491241637513644e-06, + "loss": 0.6537, + "step": 13403 + }, + { + "epoch": 0.593471547325194, + "grad_norm": 1.5232187509536743, + "learning_rate": 8.48971389581406e-06, + "loss": 0.5948, + "step": 13404 + }, + { + "epoch": 0.5935158230300078, + "grad_norm": 1.512032151222229, + "learning_rate": 8.488186190186426e-06, + "loss": 0.6139, + "step": 13405 + }, + { + "epoch": 0.5935600987348217, + "grad_norm": 1.3890460729599, + "learning_rate": 8.486658520667239e-06, + "loss": 0.5945, + "step": 13406 + }, + { + "epoch": 0.5936043744396357, + "grad_norm": 1.5135571956634521, + "learning_rate": 8.48513088729298e-06, + "loss": 0.6079, + "step": 13407 + }, + { + "epoch": 0.5936486501444495, + "grad_norm": 1.462067723274231, + "learning_rate": 8.483603290100132e-06, + "loss": 0.6373, + "step": 13408 + }, + { + "epoch": 0.5936929258492634, + "grad_norm": 1.7201440334320068, + "learning_rate": 8.482075729125191e-06, + "loss": 0.6141, + "step": 13409 + }, + { + "epoch": 0.5937372015540773, + "grad_norm": 1.563437819480896, + "learning_rate": 8.480548204404633e-06, + "loss": 0.6184, + "step": 13410 + }, + { + "epoch": 0.5937814772588911, + "grad_norm": 1.75059974193573, + "learning_rate": 8.479020715974939e-06, + "loss": 0.6508, + "step": 13411 + }, + { + "epoch": 0.593825752963705, + "grad_norm": 1.763826847076416, + "learning_rate": 8.477493263872602e-06, + "loss": 0.7302, + "step": 13412 + }, + { + "epoch": 0.5938700286685189, + "grad_norm": 1.577243685722351, + "learning_rate": 8.4759658481341e-06, + "loss": 0.6196, + "step": 13413 + }, + { + "epoch": 0.5939143043733327, + "grad_norm": 1.653693675994873, + "learning_rate": 8.474438468795911e-06, + "loss": 0.6504, + "step": 13414 + }, + { + "epoch": 0.5939585800781466, + "grad_norm": 1.5287785530090332, + "learning_rate": 8.472911125894515e-06, + "loss": 0.6004, + "step": 13415 + }, + { + "epoch": 0.5940028557829605, + "grad_norm": 1.5918102264404297, + "learning_rate": 8.471383819466396e-06, + "loss": 0.5567, + "step": 13416 + }, + { + "epoch": 0.5940471314877743, + "grad_norm": 1.8123544454574585, + "learning_rate": 8.469856549548028e-06, + "loss": 0.4985, + "step": 13417 + }, + { + "epoch": 0.5940914071925882, + "grad_norm": 1.5454721450805664, + "learning_rate": 8.468329316175892e-06, + "loss": 0.7279, + "step": 13418 + }, + { + "epoch": 0.5941356828974022, + "grad_norm": 1.2815110683441162, + "learning_rate": 8.466802119386462e-06, + "loss": 0.5671, + "step": 13419 + }, + { + "epoch": 0.594179958602216, + "grad_norm": 1.6881685256958008, + "learning_rate": 8.465274959216216e-06, + "loss": 0.6383, + "step": 13420 + }, + { + "epoch": 0.5942242343070299, + "grad_norm": 1.8333245515823364, + "learning_rate": 8.463747835701629e-06, + "loss": 0.7224, + "step": 13421 + }, + { + "epoch": 0.5942685100118438, + "grad_norm": 1.4908136129379272, + "learning_rate": 8.462220748879169e-06, + "loss": 0.6289, + "step": 13422 + }, + { + "epoch": 0.5943127857166576, + "grad_norm": 1.5861279964447021, + "learning_rate": 8.460693698785318e-06, + "loss": 0.6994, + "step": 13423 + }, + { + "epoch": 0.5943570614214715, + "grad_norm": 1.460831880569458, + "learning_rate": 8.459166685456546e-06, + "loss": 0.5753, + "step": 13424 + }, + { + "epoch": 0.5944013371262854, + "grad_norm": 1.1753953695297241, + "learning_rate": 8.45763970892932e-06, + "loss": 0.5528, + "step": 13425 + }, + { + "epoch": 0.5944456128310992, + "grad_norm": 1.3885174989700317, + "learning_rate": 8.456112769240115e-06, + "loss": 0.6098, + "step": 13426 + }, + { + "epoch": 0.5944898885359131, + "grad_norm": 1.5164071321487427, + "learning_rate": 8.4545858664254e-06, + "loss": 0.6398, + "step": 13427 + }, + { + "epoch": 0.594534164240727, + "grad_norm": 1.3744072914123535, + "learning_rate": 8.453059000521642e-06, + "loss": 0.5637, + "step": 13428 + }, + { + "epoch": 0.5945784399455409, + "grad_norm": 1.9224162101745605, + "learning_rate": 8.451532171565312e-06, + "loss": 0.7129, + "step": 13429 + }, + { + "epoch": 0.5946227156503547, + "grad_norm": 1.39631187915802, + "learning_rate": 8.450005379592874e-06, + "loss": 0.6492, + "step": 13430 + }, + { + "epoch": 0.5946669913551687, + "grad_norm": 1.5315476655960083, + "learning_rate": 8.448478624640798e-06, + "loss": 0.6524, + "step": 13431 + }, + { + "epoch": 0.5947112670599826, + "grad_norm": 2.3917884826660156, + "learning_rate": 8.446951906745543e-06, + "loss": 0.6742, + "step": 13432 + }, + { + "epoch": 0.5947555427647964, + "grad_norm": 1.7099353075027466, + "learning_rate": 8.44542522594358e-06, + "loss": 0.6936, + "step": 13433 + }, + { + "epoch": 0.5947998184696103, + "grad_norm": 1.9998762607574463, + "learning_rate": 8.443898582271368e-06, + "loss": 0.6917, + "step": 13434 + }, + { + "epoch": 0.5948440941744242, + "grad_norm": 1.6479003429412842, + "learning_rate": 8.442371975765368e-06, + "loss": 0.632, + "step": 13435 + }, + { + "epoch": 0.594888369879238, + "grad_norm": 1.6371140480041504, + "learning_rate": 8.44084540646205e-06, + "loss": 0.6088, + "step": 13436 + }, + { + "epoch": 0.5949326455840519, + "grad_norm": 1.4066264629364014, + "learning_rate": 8.439318874397871e-06, + "loss": 0.6241, + "step": 13437 + }, + { + "epoch": 0.5949769212888658, + "grad_norm": 1.4177817106246948, + "learning_rate": 8.437792379609285e-06, + "loss": 0.5954, + "step": 13438 + }, + { + "epoch": 0.5950211969936796, + "grad_norm": 1.4649314880371094, + "learning_rate": 8.43626592213276e-06, + "loss": 0.5606, + "step": 13439 + }, + { + "epoch": 0.5950654726984935, + "grad_norm": 1.4321770668029785, + "learning_rate": 8.43473950200475e-06, + "loss": 0.6144, + "step": 13440 + }, + { + "epoch": 0.5951097484033074, + "grad_norm": 1.5385395288467407, + "learning_rate": 8.433213119261712e-06, + "loss": 0.6354, + "step": 13441 + }, + { + "epoch": 0.5951540241081212, + "grad_norm": 1.7365226745605469, + "learning_rate": 8.4316867739401e-06, + "loss": 0.728, + "step": 13442 + }, + { + "epoch": 0.5951982998129351, + "grad_norm": 2.3041064739227295, + "learning_rate": 8.430160466076378e-06, + "loss": 0.7419, + "step": 13443 + }, + { + "epoch": 0.5952425755177491, + "grad_norm": 1.6377791166305542, + "learning_rate": 8.428634195706992e-06, + "loss": 0.627, + "step": 13444 + }, + { + "epoch": 0.5952868512225629, + "grad_norm": 1.3195805549621582, + "learning_rate": 8.427107962868394e-06, + "loss": 0.594, + "step": 13445 + }, + { + "epoch": 0.5953311269273768, + "grad_norm": 1.564540147781372, + "learning_rate": 8.425581767597048e-06, + "loss": 0.6278, + "step": 13446 + }, + { + "epoch": 0.5953754026321907, + "grad_norm": 1.3131719827651978, + "learning_rate": 8.424055609929397e-06, + "loss": 0.5676, + "step": 13447 + }, + { + "epoch": 0.5954196783370045, + "grad_norm": 1.305509328842163, + "learning_rate": 8.422529489901897e-06, + "loss": 0.6273, + "step": 13448 + }, + { + "epoch": 0.5954639540418184, + "grad_norm": 1.369576096534729, + "learning_rate": 8.421003407550995e-06, + "loss": 0.573, + "step": 13449 + }, + { + "epoch": 0.5955082297466323, + "grad_norm": 1.3543429374694824, + "learning_rate": 8.41947736291314e-06, + "loss": 0.5928, + "step": 13450 + }, + { + "epoch": 0.5955525054514461, + "grad_norm": 1.5478359460830688, + "learning_rate": 8.417951356024782e-06, + "loss": 0.6451, + "step": 13451 + }, + { + "epoch": 0.59559678115626, + "grad_norm": 1.6065226793289185, + "learning_rate": 8.416425386922366e-06, + "loss": 0.5884, + "step": 13452 + }, + { + "epoch": 0.5956410568610739, + "grad_norm": 1.507157802581787, + "learning_rate": 8.414899455642341e-06, + "loss": 0.5783, + "step": 13453 + }, + { + "epoch": 0.5956853325658877, + "grad_norm": 1.4330521821975708, + "learning_rate": 8.413373562221154e-06, + "loss": 0.6657, + "step": 13454 + }, + { + "epoch": 0.5957296082707016, + "grad_norm": 1.4748088121414185, + "learning_rate": 8.411847706695243e-06, + "loss": 0.6461, + "step": 13455 + }, + { + "epoch": 0.5957738839755156, + "grad_norm": 1.6042475700378418, + "learning_rate": 8.410321889101061e-06, + "loss": 0.6789, + "step": 13456 + }, + { + "epoch": 0.5958181596803294, + "grad_norm": 1.1895827054977417, + "learning_rate": 8.408796109475045e-06, + "loss": 0.5045, + "step": 13457 + }, + { + "epoch": 0.5958624353851433, + "grad_norm": 1.4857655763626099, + "learning_rate": 8.407270367853634e-06, + "loss": 0.6539, + "step": 13458 + }, + { + "epoch": 0.5959067110899572, + "grad_norm": 1.5797057151794434, + "learning_rate": 8.405744664273278e-06, + "loss": 0.5857, + "step": 13459 + }, + { + "epoch": 0.595950986794771, + "grad_norm": 1.416746973991394, + "learning_rate": 8.404218998770413e-06, + "loss": 0.5739, + "step": 13460 + }, + { + "epoch": 0.5959952624995849, + "grad_norm": 1.4382591247558594, + "learning_rate": 8.402693371381478e-06, + "loss": 0.7086, + "step": 13461 + }, + { + "epoch": 0.5960395382043988, + "grad_norm": 1.5656378269195557, + "learning_rate": 8.401167782142908e-06, + "loss": 0.6303, + "step": 13462 + }, + { + "epoch": 0.5960838139092126, + "grad_norm": 1.6107197999954224, + "learning_rate": 8.399642231091147e-06, + "loss": 0.6581, + "step": 13463 + }, + { + "epoch": 0.5961280896140265, + "grad_norm": 1.2128095626831055, + "learning_rate": 8.398116718262627e-06, + "loss": 0.5304, + "step": 13464 + }, + { + "epoch": 0.5961723653188404, + "grad_norm": 1.2912719249725342, + "learning_rate": 8.396591243693783e-06, + "loss": 0.5163, + "step": 13465 + }, + { + "epoch": 0.5962166410236543, + "grad_norm": 1.651719570159912, + "learning_rate": 8.395065807421055e-06, + "loss": 0.6523, + "step": 13466 + }, + { + "epoch": 0.5962609167284681, + "grad_norm": 1.631407380104065, + "learning_rate": 8.393540409480873e-06, + "loss": 0.6505, + "step": 13467 + }, + { + "epoch": 0.5963051924332821, + "grad_norm": 1.2421506643295288, + "learning_rate": 8.392015049909666e-06, + "loss": 0.5689, + "step": 13468 + }, + { + "epoch": 0.596349468138096, + "grad_norm": 1.5409740209579468, + "learning_rate": 8.390489728743873e-06, + "loss": 0.6388, + "step": 13469 + }, + { + "epoch": 0.5963937438429098, + "grad_norm": 1.2557541131973267, + "learning_rate": 8.388964446019922e-06, + "loss": 0.5499, + "step": 13470 + }, + { + "epoch": 0.5964380195477237, + "grad_norm": 1.7507376670837402, + "learning_rate": 8.387439201774248e-06, + "loss": 0.5988, + "step": 13471 + }, + { + "epoch": 0.5964822952525376, + "grad_norm": 1.9122188091278076, + "learning_rate": 8.385913996043269e-06, + "loss": 0.7027, + "step": 13472 + }, + { + "epoch": 0.5965265709573514, + "grad_norm": 1.4687578678131104, + "learning_rate": 8.384388828863424e-06, + "loss": 0.5946, + "step": 13473 + }, + { + "epoch": 0.5965708466621653, + "grad_norm": 1.5324761867523193, + "learning_rate": 8.382863700271136e-06, + "loss": 0.6166, + "step": 13474 + }, + { + "epoch": 0.5966151223669792, + "grad_norm": 1.2375693321228027, + "learning_rate": 8.381338610302828e-06, + "loss": 0.5839, + "step": 13475 + }, + { + "epoch": 0.596659398071793, + "grad_norm": 1.5795679092407227, + "learning_rate": 8.379813558994934e-06, + "loss": 0.6603, + "step": 13476 + }, + { + "epoch": 0.5967036737766069, + "grad_norm": 1.4482420682907104, + "learning_rate": 8.378288546383872e-06, + "loss": 0.6415, + "step": 13477 + }, + { + "epoch": 0.5967479494814208, + "grad_norm": 1.569852352142334, + "learning_rate": 8.376763572506065e-06, + "loss": 0.6436, + "step": 13478 + }, + { + "epoch": 0.5967922251862346, + "grad_norm": 1.5054019689559937, + "learning_rate": 8.375238637397942e-06, + "loss": 0.6086, + "step": 13479 + }, + { + "epoch": 0.5968365008910486, + "grad_norm": 1.7318942546844482, + "learning_rate": 8.373713741095924e-06, + "loss": 0.6265, + "step": 13480 + }, + { + "epoch": 0.5968807765958625, + "grad_norm": 1.5966415405273438, + "learning_rate": 8.372188883636425e-06, + "loss": 0.6169, + "step": 13481 + }, + { + "epoch": 0.5969250523006763, + "grad_norm": 1.4738056659698486, + "learning_rate": 8.370664065055867e-06, + "loss": 0.5612, + "step": 13482 + }, + { + "epoch": 0.5969693280054902, + "grad_norm": 1.3725824356079102, + "learning_rate": 8.369139285390675e-06, + "loss": 0.5704, + "step": 13483 + }, + { + "epoch": 0.5970136037103041, + "grad_norm": 1.3687242269515991, + "learning_rate": 8.367614544677262e-06, + "loss": 0.6422, + "step": 13484 + }, + { + "epoch": 0.5970578794151179, + "grad_norm": 1.3146586418151855, + "learning_rate": 8.366089842952045e-06, + "loss": 0.5128, + "step": 13485 + }, + { + "epoch": 0.5971021551199318, + "grad_norm": 1.4496440887451172, + "learning_rate": 8.364565180251444e-06, + "loss": 0.5863, + "step": 13486 + }, + { + "epoch": 0.5971464308247457, + "grad_norm": 1.7481377124786377, + "learning_rate": 8.363040556611872e-06, + "loss": 0.6461, + "step": 13487 + }, + { + "epoch": 0.5971907065295595, + "grad_norm": 1.4693535566329956, + "learning_rate": 8.36151597206974e-06, + "loss": 0.5627, + "step": 13488 + }, + { + "epoch": 0.5972349822343734, + "grad_norm": 1.8160678148269653, + "learning_rate": 8.35999142666147e-06, + "loss": 0.6377, + "step": 13489 + }, + { + "epoch": 0.5972792579391873, + "grad_norm": 1.4760490655899048, + "learning_rate": 8.358466920423466e-06, + "loss": 0.553, + "step": 13490 + }, + { + "epoch": 0.5973235336440011, + "grad_norm": 1.3287280797958374, + "learning_rate": 8.356942453392147e-06, + "loss": 0.6415, + "step": 13491 + }, + { + "epoch": 0.5973678093488151, + "grad_norm": 2.0881690979003906, + "learning_rate": 8.355418025603913e-06, + "loss": 0.7421, + "step": 13492 + }, + { + "epoch": 0.597412085053629, + "grad_norm": 1.4417996406555176, + "learning_rate": 8.353893637095185e-06, + "loss": 0.6599, + "step": 13493 + }, + { + "epoch": 0.5974563607584428, + "grad_norm": 1.6396936178207397, + "learning_rate": 8.352369287902367e-06, + "loss": 0.5776, + "step": 13494 + }, + { + "epoch": 0.5975006364632567, + "grad_norm": 1.4639698266983032, + "learning_rate": 8.350844978061865e-06, + "loss": 0.6191, + "step": 13495 + }, + { + "epoch": 0.5975449121680706, + "grad_norm": 1.6858406066894531, + "learning_rate": 8.349320707610089e-06, + "loss": 0.7385, + "step": 13496 + }, + { + "epoch": 0.5975891878728844, + "grad_norm": 1.6037651300430298, + "learning_rate": 8.347796476583446e-06, + "loss": 0.603, + "step": 13497 + }, + { + "epoch": 0.5976334635776983, + "grad_norm": 1.3865244388580322, + "learning_rate": 8.346272285018335e-06, + "loss": 0.5686, + "step": 13498 + }, + { + "epoch": 0.5976777392825122, + "grad_norm": 1.5956690311431885, + "learning_rate": 8.344748132951167e-06, + "loss": 0.6821, + "step": 13499 + }, + { + "epoch": 0.5977220149873261, + "grad_norm": 1.542273759841919, + "learning_rate": 8.343224020418341e-06, + "loss": 0.6019, + "step": 13500 + }, + { + "epoch": 0.5977662906921399, + "grad_norm": 1.5975936651229858, + "learning_rate": 8.341699947456261e-06, + "loss": 0.6735, + "step": 13501 + }, + { + "epoch": 0.5978105663969538, + "grad_norm": 1.392975091934204, + "learning_rate": 8.340175914101324e-06, + "loss": 0.5865, + "step": 13502 + }, + { + "epoch": 0.5978548421017678, + "grad_norm": 1.7340188026428223, + "learning_rate": 8.33865192038994e-06, + "loss": 0.6143, + "step": 13503 + }, + { + "epoch": 0.5978991178065816, + "grad_norm": 1.3592907190322876, + "learning_rate": 8.3371279663585e-06, + "loss": 0.5358, + "step": 13504 + }, + { + "epoch": 0.5979433935113955, + "grad_norm": 1.5601248741149902, + "learning_rate": 8.3356040520434e-06, + "loss": 0.6835, + "step": 13505 + }, + { + "epoch": 0.5979876692162094, + "grad_norm": 1.6687908172607422, + "learning_rate": 8.334080177481045e-06, + "loss": 0.5659, + "step": 13506 + }, + { + "epoch": 0.5980319449210232, + "grad_norm": 1.2183971405029297, + "learning_rate": 8.33255634270783e-06, + "loss": 0.5684, + "step": 13507 + }, + { + "epoch": 0.5980762206258371, + "grad_norm": 1.5981236696243286, + "learning_rate": 8.331032547760147e-06, + "loss": 0.662, + "step": 13508 + }, + { + "epoch": 0.598120496330651, + "grad_norm": 1.400639533996582, + "learning_rate": 8.329508792674393e-06, + "loss": 0.542, + "step": 13509 + }, + { + "epoch": 0.5981647720354648, + "grad_norm": 1.428950548171997, + "learning_rate": 8.327985077486963e-06, + "loss": 0.5873, + "step": 13510 + }, + { + "epoch": 0.5982090477402787, + "grad_norm": 1.436372995376587, + "learning_rate": 8.326461402234248e-06, + "loss": 0.6192, + "step": 13511 + }, + { + "epoch": 0.5982533234450926, + "grad_norm": 1.7967280149459839, + "learning_rate": 8.324937766952638e-06, + "loss": 0.641, + "step": 13512 + }, + { + "epoch": 0.5982975991499064, + "grad_norm": 1.1846691370010376, + "learning_rate": 8.323414171678527e-06, + "loss": 0.5214, + "step": 13513 + }, + { + "epoch": 0.5983418748547203, + "grad_norm": 1.5641940832138062, + "learning_rate": 8.321890616448302e-06, + "loss": 0.7048, + "step": 13514 + }, + { + "epoch": 0.5983861505595343, + "grad_norm": 1.7213448286056519, + "learning_rate": 8.320367101298351e-06, + "loss": 0.6895, + "step": 13515 + }, + { + "epoch": 0.5984304262643481, + "grad_norm": 1.358388900756836, + "learning_rate": 8.318843626265067e-06, + "loss": 0.6292, + "step": 13516 + }, + { + "epoch": 0.598474701969162, + "grad_norm": 1.3810817003250122, + "learning_rate": 8.317320191384835e-06, + "loss": 0.5775, + "step": 13517 + }, + { + "epoch": 0.5985189776739759, + "grad_norm": 1.4956034421920776, + "learning_rate": 8.315796796694038e-06, + "loss": 0.6417, + "step": 13518 + }, + { + "epoch": 0.5985632533787897, + "grad_norm": 1.450722098350525, + "learning_rate": 8.314273442229067e-06, + "loss": 0.5493, + "step": 13519 + }, + { + "epoch": 0.5986075290836036, + "grad_norm": 1.442068099975586, + "learning_rate": 8.3127501280263e-06, + "loss": 0.599, + "step": 13520 + }, + { + "epoch": 0.5986518047884175, + "grad_norm": 1.7051808834075928, + "learning_rate": 8.311226854122124e-06, + "loss": 0.6756, + "step": 13521 + }, + { + "epoch": 0.5986960804932313, + "grad_norm": 1.4180431365966797, + "learning_rate": 8.309703620552916e-06, + "loss": 0.5862, + "step": 13522 + }, + { + "epoch": 0.5987403561980452, + "grad_norm": 1.3966248035430908, + "learning_rate": 8.308180427355062e-06, + "loss": 0.5888, + "step": 13523 + }, + { + "epoch": 0.5987846319028591, + "grad_norm": 1.3587948083877563, + "learning_rate": 8.306657274564942e-06, + "loss": 0.5772, + "step": 13524 + }, + { + "epoch": 0.5988289076076729, + "grad_norm": 1.7535326480865479, + "learning_rate": 8.305134162218931e-06, + "loss": 0.6433, + "step": 13525 + }, + { + "epoch": 0.5988731833124868, + "grad_norm": 1.300018548965454, + "learning_rate": 8.303611090353414e-06, + "loss": 0.6063, + "step": 13526 + }, + { + "epoch": 0.5989174590173008, + "grad_norm": 2.101036787033081, + "learning_rate": 8.302088059004768e-06, + "loss": 0.6822, + "step": 13527 + }, + { + "epoch": 0.5989617347221146, + "grad_norm": 1.6371737718582153, + "learning_rate": 8.300565068209362e-06, + "loss": 0.6173, + "step": 13528 + }, + { + "epoch": 0.5990060104269285, + "grad_norm": 1.7657088041305542, + "learning_rate": 8.299042118003571e-06, + "loss": 0.6731, + "step": 13529 + }, + { + "epoch": 0.5990502861317424, + "grad_norm": 1.2840683460235596, + "learning_rate": 8.297519208423782e-06, + "loss": 0.5926, + "step": 13530 + }, + { + "epoch": 0.5990945618365562, + "grad_norm": 1.4654130935668945, + "learning_rate": 8.295996339506357e-06, + "loss": 0.6482, + "step": 13531 + }, + { + "epoch": 0.5991388375413701, + "grad_norm": 1.9291973114013672, + "learning_rate": 8.29447351128767e-06, + "loss": 0.6093, + "step": 13532 + }, + { + "epoch": 0.599183113246184, + "grad_norm": 1.4548311233520508, + "learning_rate": 8.292950723804099e-06, + "loss": 0.5688, + "step": 13533 + }, + { + "epoch": 0.5992273889509979, + "grad_norm": 1.298063039779663, + "learning_rate": 8.291427977092007e-06, + "loss": 0.598, + "step": 13534 + }, + { + "epoch": 0.5992716646558117, + "grad_norm": 1.5948432683944702, + "learning_rate": 8.289905271187766e-06, + "loss": 0.5859, + "step": 13535 + }, + { + "epoch": 0.5993159403606256, + "grad_norm": 1.6432751417160034, + "learning_rate": 8.28838260612775e-06, + "loss": 0.6105, + "step": 13536 + }, + { + "epoch": 0.5993602160654395, + "grad_norm": 1.3231178522109985, + "learning_rate": 8.28685998194832e-06, + "loss": 0.6132, + "step": 13537 + }, + { + "epoch": 0.5994044917702533, + "grad_norm": 1.2590157985687256, + "learning_rate": 8.285337398685841e-06, + "loss": 0.6588, + "step": 13538 + }, + { + "epoch": 0.5994487674750673, + "grad_norm": 1.4774980545043945, + "learning_rate": 8.283814856376681e-06, + "loss": 0.5789, + "step": 13539 + }, + { + "epoch": 0.5994930431798812, + "grad_norm": 1.4455317258834839, + "learning_rate": 8.28229235505721e-06, + "loss": 0.6047, + "step": 13540 + }, + { + "epoch": 0.599537318884695, + "grad_norm": 1.3606165647506714, + "learning_rate": 8.280769894763786e-06, + "loss": 0.595, + "step": 13541 + }, + { + "epoch": 0.5995815945895089, + "grad_norm": 1.500425934791565, + "learning_rate": 8.279247475532774e-06, + "loss": 0.6238, + "step": 13542 + }, + { + "epoch": 0.5996258702943228, + "grad_norm": 1.4784331321716309, + "learning_rate": 8.277725097400536e-06, + "loss": 0.6191, + "step": 13543 + }, + { + "epoch": 0.5996701459991366, + "grad_norm": 1.2927262783050537, + "learning_rate": 8.27620276040343e-06, + "loss": 0.5747, + "step": 13544 + }, + { + "epoch": 0.5997144217039505, + "grad_norm": 1.8988897800445557, + "learning_rate": 8.274680464577816e-06, + "loss": 0.624, + "step": 13545 + }, + { + "epoch": 0.5997586974087644, + "grad_norm": 1.401950478553772, + "learning_rate": 8.273158209960058e-06, + "loss": 0.6187, + "step": 13546 + }, + { + "epoch": 0.5998029731135782, + "grad_norm": 1.2800002098083496, + "learning_rate": 8.271635996586507e-06, + "loss": 0.5092, + "step": 13547 + }, + { + "epoch": 0.5998472488183921, + "grad_norm": 1.7898919582366943, + "learning_rate": 8.270113824493525e-06, + "loss": 0.6599, + "step": 13548 + }, + { + "epoch": 0.599891524523206, + "grad_norm": 1.3509953022003174, + "learning_rate": 8.26859169371746e-06, + "loss": 0.5995, + "step": 13549 + }, + { + "epoch": 0.5999358002280198, + "grad_norm": 1.7119733095169067, + "learning_rate": 8.26706960429468e-06, + "loss": 0.6452, + "step": 13550 + }, + { + "epoch": 0.5999800759328338, + "grad_norm": 1.288983702659607, + "learning_rate": 8.265547556261532e-06, + "loss": 0.5407, + "step": 13551 + }, + { + "epoch": 0.6000243516376477, + "grad_norm": 1.661858320236206, + "learning_rate": 8.264025549654363e-06, + "loss": 0.63, + "step": 13552 + }, + { + "epoch": 0.6000686273424615, + "grad_norm": 1.1771515607833862, + "learning_rate": 8.262503584509531e-06, + "loss": 0.4955, + "step": 13553 + }, + { + "epoch": 0.6001129030472754, + "grad_norm": 1.2737828493118286, + "learning_rate": 8.26098166086339e-06, + "loss": 0.5103, + "step": 13554 + }, + { + "epoch": 0.6001571787520893, + "grad_norm": 1.5442777872085571, + "learning_rate": 8.259459778752282e-06, + "loss": 0.6947, + "step": 13555 + }, + { + "epoch": 0.6002014544569031, + "grad_norm": 2.071622610092163, + "learning_rate": 8.257937938212562e-06, + "loss": 0.7373, + "step": 13556 + }, + { + "epoch": 0.600245730161717, + "grad_norm": 1.511322021484375, + "learning_rate": 8.256416139280577e-06, + "loss": 0.5954, + "step": 13557 + }, + { + "epoch": 0.6002900058665309, + "grad_norm": 1.3288823366165161, + "learning_rate": 8.254894381992672e-06, + "loss": 0.5914, + "step": 13558 + }, + { + "epoch": 0.6003342815713447, + "grad_norm": 1.771730899810791, + "learning_rate": 8.253372666385192e-06, + "loss": 0.6338, + "step": 13559 + }, + { + "epoch": 0.6003785572761586, + "grad_norm": 1.3827346563339233, + "learning_rate": 8.251850992494486e-06, + "loss": 0.557, + "step": 13560 + }, + { + "epoch": 0.6004228329809725, + "grad_norm": 1.29026198387146, + "learning_rate": 8.250329360356895e-06, + "loss": 0.544, + "step": 13561 + }, + { + "epoch": 0.6004671086857863, + "grad_norm": 1.4396860599517822, + "learning_rate": 8.248807770008759e-06, + "loss": 0.5895, + "step": 13562 + }, + { + "epoch": 0.6005113843906003, + "grad_norm": 1.5300796031951904, + "learning_rate": 8.247286221486429e-06, + "loss": 0.6686, + "step": 13563 + }, + { + "epoch": 0.6005556600954142, + "grad_norm": 1.5100603103637695, + "learning_rate": 8.245764714826237e-06, + "loss": 0.5471, + "step": 13564 + }, + { + "epoch": 0.600599935800228, + "grad_norm": 1.8630836009979248, + "learning_rate": 8.244243250064527e-06, + "loss": 0.6186, + "step": 13565 + }, + { + "epoch": 0.6006442115050419, + "grad_norm": 1.5679104328155518, + "learning_rate": 8.242721827237639e-06, + "loss": 0.6405, + "step": 13566 + }, + { + "epoch": 0.6006884872098558, + "grad_norm": 1.3641122579574585, + "learning_rate": 8.24120044638191e-06, + "loss": 0.566, + "step": 13567 + }, + { + "epoch": 0.6007327629146696, + "grad_norm": 1.5214112997055054, + "learning_rate": 8.239679107533675e-06, + "loss": 0.6773, + "step": 13568 + }, + { + "epoch": 0.6007770386194835, + "grad_norm": 1.3938087224960327, + "learning_rate": 8.238157810729268e-06, + "loss": 0.5845, + "step": 13569 + }, + { + "epoch": 0.6008213143242974, + "grad_norm": 1.7612197399139404, + "learning_rate": 8.236636556005032e-06, + "loss": 0.6694, + "step": 13570 + }, + { + "epoch": 0.6008655900291113, + "grad_norm": 1.3100882768630981, + "learning_rate": 8.235115343397295e-06, + "loss": 0.5763, + "step": 13571 + }, + { + "epoch": 0.6009098657339251, + "grad_norm": 1.6890869140625, + "learning_rate": 8.233594172942387e-06, + "loss": 0.6226, + "step": 13572 + }, + { + "epoch": 0.600954141438739, + "grad_norm": 1.6174371242523193, + "learning_rate": 8.232073044676648e-06, + "loss": 0.6039, + "step": 13573 + }, + { + "epoch": 0.600998417143553, + "grad_norm": 2.013681411743164, + "learning_rate": 8.230551958636404e-06, + "loss": 0.6077, + "step": 13574 + }, + { + "epoch": 0.6010426928483668, + "grad_norm": 1.5150539875030518, + "learning_rate": 8.229030914857985e-06, + "loss": 0.6172, + "step": 13575 + }, + { + "epoch": 0.6010869685531807, + "grad_norm": 1.4436914920806885, + "learning_rate": 8.227509913377721e-06, + "loss": 0.5904, + "step": 13576 + }, + { + "epoch": 0.6011312442579946, + "grad_norm": 1.4191428422927856, + "learning_rate": 8.22598895423194e-06, + "loss": 0.6418, + "step": 13577 + }, + { + "epoch": 0.6011755199628084, + "grad_norm": 1.8563923835754395, + "learning_rate": 8.224468037456969e-06, + "loss": 0.6742, + "step": 13578 + }, + { + "epoch": 0.6012197956676223, + "grad_norm": 1.5519533157348633, + "learning_rate": 8.222947163089132e-06, + "loss": 0.5818, + "step": 13579 + }, + { + "epoch": 0.6012640713724362, + "grad_norm": 1.4545446634292603, + "learning_rate": 8.221426331164756e-06, + "loss": 0.6048, + "step": 13580 + }, + { + "epoch": 0.60130834707725, + "grad_norm": 1.811958909034729, + "learning_rate": 8.219905541720163e-06, + "loss": 0.6288, + "step": 13581 + }, + { + "epoch": 0.6013526227820639, + "grad_norm": 1.244286298751831, + "learning_rate": 8.218384794791674e-06, + "loss": 0.541, + "step": 13582 + }, + { + "epoch": 0.6013968984868778, + "grad_norm": 1.6364229917526245, + "learning_rate": 8.21686409041562e-06, + "loss": 0.6157, + "step": 13583 + }, + { + "epoch": 0.6014411741916916, + "grad_norm": 1.5260435342788696, + "learning_rate": 8.215343428628312e-06, + "loss": 0.6483, + "step": 13584 + }, + { + "epoch": 0.6014854498965055, + "grad_norm": 1.3049923181533813, + "learning_rate": 8.21382280946607e-06, + "loss": 0.5566, + "step": 13585 + }, + { + "epoch": 0.6015297256013195, + "grad_norm": 2.3044023513793945, + "learning_rate": 8.21230223296522e-06, + "loss": 0.6085, + "step": 13586 + }, + { + "epoch": 0.6015740013061333, + "grad_norm": 1.485332727432251, + "learning_rate": 8.210781699162075e-06, + "loss": 0.6249, + "step": 13587 + }, + { + "epoch": 0.6016182770109472, + "grad_norm": 1.3958059549331665, + "learning_rate": 8.209261208092952e-06, + "loss": 0.5692, + "step": 13588 + }, + { + "epoch": 0.6016625527157611, + "grad_norm": 1.2695338726043701, + "learning_rate": 8.207740759794165e-06, + "loss": 0.5105, + "step": 13589 + }, + { + "epoch": 0.6017068284205749, + "grad_norm": 1.33749258518219, + "learning_rate": 8.206220354302034e-06, + "loss": 0.604, + "step": 13590 + }, + { + "epoch": 0.6017511041253888, + "grad_norm": 1.279305100440979, + "learning_rate": 8.204699991652868e-06, + "loss": 0.6083, + "step": 13591 + }, + { + "epoch": 0.6017953798302027, + "grad_norm": 1.4476814270019531, + "learning_rate": 8.203179671882979e-06, + "loss": 0.5416, + "step": 13592 + }, + { + "epoch": 0.6018396555350165, + "grad_norm": 1.5022599697113037, + "learning_rate": 8.201659395028683e-06, + "loss": 0.5948, + "step": 13593 + }, + { + "epoch": 0.6018839312398304, + "grad_norm": 1.6620275974273682, + "learning_rate": 8.200139161126288e-06, + "loss": 0.6307, + "step": 13594 + }, + { + "epoch": 0.6019282069446443, + "grad_norm": 1.2459701299667358, + "learning_rate": 8.1986189702121e-06, + "loss": 0.5224, + "step": 13595 + }, + { + "epoch": 0.6019724826494581, + "grad_norm": 1.5029078722000122, + "learning_rate": 8.197098822322434e-06, + "loss": 0.5754, + "step": 13596 + }, + { + "epoch": 0.602016758354272, + "grad_norm": 1.502830982208252, + "learning_rate": 8.195578717493595e-06, + "loss": 0.6434, + "step": 13597 + }, + { + "epoch": 0.602061034059086, + "grad_norm": 1.399571418762207, + "learning_rate": 8.194058655761891e-06, + "loss": 0.5869, + "step": 13598 + }, + { + "epoch": 0.6021053097638998, + "grad_norm": 1.634529709815979, + "learning_rate": 8.19253863716362e-06, + "loss": 0.6246, + "step": 13599 + }, + { + "epoch": 0.6021495854687137, + "grad_norm": 2.1209628582000732, + "learning_rate": 8.191018661735096e-06, + "loss": 0.6704, + "step": 13600 + }, + { + "epoch": 0.6021938611735276, + "grad_norm": 1.7286128997802734, + "learning_rate": 8.189498729512618e-06, + "loss": 0.6042, + "step": 13601 + }, + { + "epoch": 0.6022381368783414, + "grad_norm": 1.186802625656128, + "learning_rate": 8.187978840532486e-06, + "loss": 0.5263, + "step": 13602 + }, + { + "epoch": 0.6022824125831553, + "grad_norm": 1.241341233253479, + "learning_rate": 8.186458994831008e-06, + "loss": 0.5405, + "step": 13603 + }, + { + "epoch": 0.6023266882879692, + "grad_norm": 1.4973034858703613, + "learning_rate": 8.18493919244448e-06, + "loss": 0.5723, + "step": 13604 + }, + { + "epoch": 0.6023709639927831, + "grad_norm": 1.3906611204147339, + "learning_rate": 8.183419433409197e-06, + "loss": 0.6002, + "step": 13605 + }, + { + "epoch": 0.6024152396975969, + "grad_norm": 1.2068548202514648, + "learning_rate": 8.181899717761465e-06, + "loss": 0.5312, + "step": 13606 + }, + { + "epoch": 0.6024595154024108, + "grad_norm": 1.7133738994598389, + "learning_rate": 8.18038004553758e-06, + "loss": 0.6231, + "step": 13607 + }, + { + "epoch": 0.6025037911072247, + "grad_norm": 1.255627155303955, + "learning_rate": 8.178860416773833e-06, + "loss": 0.5795, + "step": 13608 + }, + { + "epoch": 0.6025480668120385, + "grad_norm": 1.4841021299362183, + "learning_rate": 8.17734083150652e-06, + "loss": 0.5833, + "step": 13609 + }, + { + "epoch": 0.6025923425168525, + "grad_norm": 1.7127512693405151, + "learning_rate": 8.175821289771941e-06, + "loss": 0.604, + "step": 13610 + }, + { + "epoch": 0.6026366182216664, + "grad_norm": 1.6765857934951782, + "learning_rate": 8.174301791606384e-06, + "loss": 0.6537, + "step": 13611 + }, + { + "epoch": 0.6026808939264802, + "grad_norm": 1.3980811834335327, + "learning_rate": 8.172782337046141e-06, + "loss": 0.6357, + "step": 13612 + }, + { + "epoch": 0.6027251696312941, + "grad_norm": 1.7527058124542236, + "learning_rate": 8.171262926127507e-06, + "loss": 0.6642, + "step": 13613 + }, + { + "epoch": 0.602769445336108, + "grad_norm": 1.5008195638656616, + "learning_rate": 8.169743558886769e-06, + "loss": 0.6239, + "step": 13614 + }, + { + "epoch": 0.6028137210409218, + "grad_norm": 1.4247727394104004, + "learning_rate": 8.168224235360212e-06, + "loss": 0.5408, + "step": 13615 + }, + { + "epoch": 0.6028579967457357, + "grad_norm": 1.3726931810379028, + "learning_rate": 8.166704955584131e-06, + "loss": 0.5929, + "step": 13616 + }, + { + "epoch": 0.6029022724505496, + "grad_norm": 1.550247073173523, + "learning_rate": 8.165185719594807e-06, + "loss": 0.6265, + "step": 13617 + }, + { + "epoch": 0.6029465481553634, + "grad_norm": 1.4802895784378052, + "learning_rate": 8.163666527428532e-06, + "loss": 0.6449, + "step": 13618 + }, + { + "epoch": 0.6029908238601773, + "grad_norm": 1.0417417287826538, + "learning_rate": 8.16214737912158e-06, + "loss": 0.4259, + "step": 13619 + }, + { + "epoch": 0.6030350995649912, + "grad_norm": 1.3591805696487427, + "learning_rate": 8.160628274710247e-06, + "loss": 0.5788, + "step": 13620 + }, + { + "epoch": 0.603079375269805, + "grad_norm": 1.567033052444458, + "learning_rate": 8.15910921423081e-06, + "loss": 0.5571, + "step": 13621 + }, + { + "epoch": 0.603123650974619, + "grad_norm": 1.3169395923614502, + "learning_rate": 8.157590197719547e-06, + "loss": 0.5553, + "step": 13622 + }, + { + "epoch": 0.6031679266794329, + "grad_norm": 1.4776160717010498, + "learning_rate": 8.156071225212746e-06, + "loss": 0.548, + "step": 13623 + }, + { + "epoch": 0.6032122023842467, + "grad_norm": 1.7411317825317383, + "learning_rate": 8.15455229674668e-06, + "loss": 0.6469, + "step": 13624 + }, + { + "epoch": 0.6032564780890606, + "grad_norm": 1.5354578495025635, + "learning_rate": 8.15303341235763e-06, + "loss": 0.6325, + "step": 13625 + }, + { + "epoch": 0.6033007537938745, + "grad_norm": 1.5027261972427368, + "learning_rate": 8.151514572081873e-06, + "loss": 0.628, + "step": 13626 + }, + { + "epoch": 0.6033450294986883, + "grad_norm": 1.5411523580551147, + "learning_rate": 8.149995775955686e-06, + "loss": 0.6346, + "step": 13627 + }, + { + "epoch": 0.6033893052035022, + "grad_norm": 1.505144715309143, + "learning_rate": 8.148477024015345e-06, + "loss": 0.5378, + "step": 13628 + }, + { + "epoch": 0.6034335809083161, + "grad_norm": 1.5728110074996948, + "learning_rate": 8.146958316297117e-06, + "loss": 0.5987, + "step": 13629 + }, + { + "epoch": 0.6034778566131299, + "grad_norm": 1.4420217275619507, + "learning_rate": 8.145439652837288e-06, + "loss": 0.5898, + "step": 13630 + }, + { + "epoch": 0.6035221323179438, + "grad_norm": 1.6983227729797363, + "learning_rate": 8.14392103367212e-06, + "loss": 0.6945, + "step": 13631 + }, + { + "epoch": 0.6035664080227577, + "grad_norm": 1.2762588262557983, + "learning_rate": 8.142402458837882e-06, + "loss": 0.5782, + "step": 13632 + }, + { + "epoch": 0.6036106837275715, + "grad_norm": 1.2882843017578125, + "learning_rate": 8.140883928370855e-06, + "loss": 0.5915, + "step": 13633 + }, + { + "epoch": 0.6036549594323855, + "grad_norm": 1.6903365850448608, + "learning_rate": 8.139365442307301e-06, + "loss": 0.599, + "step": 13634 + }, + { + "epoch": 0.6036992351371994, + "grad_norm": 1.4162791967391968, + "learning_rate": 8.137847000683485e-06, + "loss": 0.6175, + "step": 13635 + }, + { + "epoch": 0.6037435108420132, + "grad_norm": 1.3121670484542847, + "learning_rate": 8.13632860353568e-06, + "loss": 0.5441, + "step": 13636 + }, + { + "epoch": 0.6037877865468271, + "grad_norm": 1.405349850654602, + "learning_rate": 8.134810250900149e-06, + "loss": 0.5944, + "step": 13637 + }, + { + "epoch": 0.603832062251641, + "grad_norm": 1.2248564958572388, + "learning_rate": 8.133291942813157e-06, + "loss": 0.5978, + "step": 13638 + }, + { + "epoch": 0.6038763379564548, + "grad_norm": 1.5895743370056152, + "learning_rate": 8.131773679310964e-06, + "loss": 0.6303, + "step": 13639 + }, + { + "epoch": 0.6039206136612687, + "grad_norm": 1.347292423248291, + "learning_rate": 8.130255460429838e-06, + "loss": 0.5646, + "step": 13640 + }, + { + "epoch": 0.6039648893660826, + "grad_norm": 1.2032274007797241, + "learning_rate": 8.128737286206038e-06, + "loss": 0.5381, + "step": 13641 + }, + { + "epoch": 0.6040091650708965, + "grad_norm": 1.5428415536880493, + "learning_rate": 8.12721915667582e-06, + "loss": 0.5831, + "step": 13642 + }, + { + "epoch": 0.6040534407757103, + "grad_norm": 1.6338658332824707, + "learning_rate": 8.125701071875452e-06, + "loss": 0.6239, + "step": 13643 + }, + { + "epoch": 0.6040977164805242, + "grad_norm": 1.6225873231887817, + "learning_rate": 8.12418303184119e-06, + "loss": 0.6707, + "step": 13644 + }, + { + "epoch": 0.6041419921853382, + "grad_norm": 1.4362558126449585, + "learning_rate": 8.122665036609284e-06, + "loss": 0.5958, + "step": 13645 + }, + { + "epoch": 0.604186267890152, + "grad_norm": 1.3177536725997925, + "learning_rate": 8.121147086215995e-06, + "loss": 0.5938, + "step": 13646 + }, + { + "epoch": 0.6042305435949659, + "grad_norm": 1.5254554748535156, + "learning_rate": 8.119629180697581e-06, + "loss": 0.589, + "step": 13647 + }, + { + "epoch": 0.6042748192997798, + "grad_norm": 1.336473822593689, + "learning_rate": 8.118111320090293e-06, + "loss": 0.545, + "step": 13648 + }, + { + "epoch": 0.6043190950045936, + "grad_norm": 1.543782114982605, + "learning_rate": 8.11659350443038e-06, + "loss": 0.612, + "step": 13649 + }, + { + "epoch": 0.6043633707094075, + "grad_norm": 1.570693016052246, + "learning_rate": 8.115075733754101e-06, + "loss": 0.6128, + "step": 13650 + }, + { + "epoch": 0.6044076464142214, + "grad_norm": 1.6818293333053589, + "learning_rate": 8.113558008097704e-06, + "loss": 0.6362, + "step": 13651 + }, + { + "epoch": 0.6044519221190352, + "grad_norm": 1.3182638883590698, + "learning_rate": 8.112040327497434e-06, + "loss": 0.5155, + "step": 13652 + }, + { + "epoch": 0.6044961978238491, + "grad_norm": 1.6018441915512085, + "learning_rate": 8.110522691989546e-06, + "loss": 0.6759, + "step": 13653 + }, + { + "epoch": 0.604540473528663, + "grad_norm": 1.494657278060913, + "learning_rate": 8.10900510161029e-06, + "loss": 0.5942, + "step": 13654 + }, + { + "epoch": 0.6045847492334768, + "grad_norm": 1.77886164188385, + "learning_rate": 8.107487556395902e-06, + "loss": 0.6314, + "step": 13655 + }, + { + "epoch": 0.6046290249382907, + "grad_norm": 1.560325026512146, + "learning_rate": 8.105970056382629e-06, + "loss": 0.5647, + "step": 13656 + }, + { + "epoch": 0.6046733006431046, + "grad_norm": 1.4416745901107788, + "learning_rate": 8.104452601606725e-06, + "loss": 0.5649, + "step": 13657 + }, + { + "epoch": 0.6047175763479185, + "grad_norm": 1.537463665008545, + "learning_rate": 8.102935192104427e-06, + "loss": 0.5886, + "step": 13658 + }, + { + "epoch": 0.6047618520527324, + "grad_norm": 1.577884554862976, + "learning_rate": 8.101417827911975e-06, + "loss": 0.6606, + "step": 13659 + }, + { + "epoch": 0.6048061277575463, + "grad_norm": 1.8057271242141724, + "learning_rate": 8.099900509065612e-06, + "loss": 0.6745, + "step": 13660 + }, + { + "epoch": 0.6048504034623601, + "grad_norm": 1.676031231880188, + "learning_rate": 8.098383235601581e-06, + "loss": 0.6225, + "step": 13661 + }, + { + "epoch": 0.604894679167174, + "grad_norm": 1.3368768692016602, + "learning_rate": 8.096866007556112e-06, + "loss": 0.5081, + "step": 13662 + }, + { + "epoch": 0.6049389548719879, + "grad_norm": 1.3945250511169434, + "learning_rate": 8.095348824965457e-06, + "loss": 0.582, + "step": 13663 + }, + { + "epoch": 0.6049832305768017, + "grad_norm": 1.6018130779266357, + "learning_rate": 8.093831687865841e-06, + "loss": 0.5821, + "step": 13664 + }, + { + "epoch": 0.6050275062816156, + "grad_norm": 1.4687198400497437, + "learning_rate": 8.092314596293504e-06, + "loss": 0.5862, + "step": 13665 + }, + { + "epoch": 0.6050717819864295, + "grad_norm": 1.2825862169265747, + "learning_rate": 8.090797550284675e-06, + "loss": 0.5391, + "step": 13666 + }, + { + "epoch": 0.6051160576912433, + "grad_norm": 1.3529386520385742, + "learning_rate": 8.089280549875596e-06, + "loss": 0.6222, + "step": 13667 + }, + { + "epoch": 0.6051603333960572, + "grad_norm": 1.4622963666915894, + "learning_rate": 8.087763595102495e-06, + "loss": 0.6481, + "step": 13668 + }, + { + "epoch": 0.6052046091008711, + "grad_norm": 2.5154941082000732, + "learning_rate": 8.086246686001603e-06, + "loss": 0.5651, + "step": 13669 + }, + { + "epoch": 0.605248884805685, + "grad_norm": 1.3078358173370361, + "learning_rate": 8.08472982260915e-06, + "loss": 0.5293, + "step": 13670 + }, + { + "epoch": 0.6052931605104989, + "grad_norm": 1.5182071924209595, + "learning_rate": 8.083213004961368e-06, + "loss": 0.653, + "step": 13671 + }, + { + "epoch": 0.6053374362153128, + "grad_norm": 1.5397365093231201, + "learning_rate": 8.08169623309448e-06, + "loss": 0.5991, + "step": 13672 + }, + { + "epoch": 0.6053817119201266, + "grad_norm": 1.5177769660949707, + "learning_rate": 8.080179507044717e-06, + "loss": 0.6249, + "step": 13673 + }, + { + "epoch": 0.6054259876249405, + "grad_norm": 1.478628396987915, + "learning_rate": 8.078662826848305e-06, + "loss": 0.5827, + "step": 13674 + }, + { + "epoch": 0.6054702633297544, + "grad_norm": 1.3185442686080933, + "learning_rate": 8.077146192541465e-06, + "loss": 0.5728, + "step": 13675 + }, + { + "epoch": 0.6055145390345683, + "grad_norm": 1.6623823642730713, + "learning_rate": 8.07562960416042e-06, + "loss": 0.6118, + "step": 13676 + }, + { + "epoch": 0.6055588147393821, + "grad_norm": 1.376840591430664, + "learning_rate": 8.074113061741397e-06, + "loss": 0.5687, + "step": 13677 + }, + { + "epoch": 0.605603090444196, + "grad_norm": 1.2988303899765015, + "learning_rate": 8.07259656532062e-06, + "loss": 0.5137, + "step": 13678 + }, + { + "epoch": 0.6056473661490099, + "grad_norm": 1.6938999891281128, + "learning_rate": 8.071080114934297e-06, + "loss": 0.605, + "step": 13679 + }, + { + "epoch": 0.6056916418538237, + "grad_norm": 1.4112756252288818, + "learning_rate": 8.069563710618659e-06, + "loss": 0.6273, + "step": 13680 + }, + { + "epoch": 0.6057359175586376, + "grad_norm": 1.3137922286987305, + "learning_rate": 8.06804735240992e-06, + "loss": 0.5722, + "step": 13681 + }, + { + "epoch": 0.6057801932634516, + "grad_norm": 1.5593212842941284, + "learning_rate": 8.066531040344296e-06, + "loss": 0.6431, + "step": 13682 + }, + { + "epoch": 0.6058244689682654, + "grad_norm": 1.597817301750183, + "learning_rate": 8.065014774458004e-06, + "loss": 0.5935, + "step": 13683 + }, + { + "epoch": 0.6058687446730793, + "grad_norm": 1.4684187173843384, + "learning_rate": 8.06349855478726e-06, + "loss": 0.5575, + "step": 13684 + }, + { + "epoch": 0.6059130203778932, + "grad_norm": 1.4142396450042725, + "learning_rate": 8.061982381368276e-06, + "loss": 0.593, + "step": 13685 + }, + { + "epoch": 0.605957296082707, + "grad_norm": 1.4740052223205566, + "learning_rate": 8.060466254237263e-06, + "loss": 0.6344, + "step": 13686 + }, + { + "epoch": 0.6060015717875209, + "grad_norm": 1.3865442276000977, + "learning_rate": 8.058950173430435e-06, + "loss": 0.6059, + "step": 13687 + }, + { + "epoch": 0.6060458474923348, + "grad_norm": 1.4687073230743408, + "learning_rate": 8.057434138984003e-06, + "loss": 0.5759, + "step": 13688 + }, + { + "epoch": 0.6060901231971486, + "grad_norm": 1.465097188949585, + "learning_rate": 8.05591815093417e-06, + "loss": 0.5828, + "step": 13689 + }, + { + "epoch": 0.6061343989019625, + "grad_norm": 1.468062162399292, + "learning_rate": 8.054402209317155e-06, + "loss": 0.6056, + "step": 13690 + }, + { + "epoch": 0.6061786746067764, + "grad_norm": 1.540112853050232, + "learning_rate": 8.052886314169157e-06, + "loss": 0.6153, + "step": 13691 + }, + { + "epoch": 0.6062229503115902, + "grad_norm": 1.709082007408142, + "learning_rate": 8.051370465526382e-06, + "loss": 0.6064, + "step": 13692 + }, + { + "epoch": 0.6062672260164041, + "grad_norm": 1.388682246208191, + "learning_rate": 8.04985466342504e-06, + "loss": 0.5364, + "step": 13693 + }, + { + "epoch": 0.6063115017212181, + "grad_norm": 1.5449072122573853, + "learning_rate": 8.04833890790133e-06, + "loss": 0.5213, + "step": 13694 + }, + { + "epoch": 0.6063557774260319, + "grad_norm": 1.5611374378204346, + "learning_rate": 8.046823198991458e-06, + "loss": 0.6161, + "step": 13695 + }, + { + "epoch": 0.6064000531308458, + "grad_norm": 1.9261609315872192, + "learning_rate": 8.04530753673162e-06, + "loss": 0.6511, + "step": 13696 + }, + { + "epoch": 0.6064443288356597, + "grad_norm": 1.4862948656082153, + "learning_rate": 8.043791921158023e-06, + "loss": 0.644, + "step": 13697 + }, + { + "epoch": 0.6064886045404735, + "grad_norm": 1.3382279872894287, + "learning_rate": 8.042276352306863e-06, + "loss": 0.565, + "step": 13698 + }, + { + "epoch": 0.6065328802452874, + "grad_norm": 1.4012246131896973, + "learning_rate": 8.040760830214334e-06, + "loss": 0.5946, + "step": 13699 + }, + { + "epoch": 0.6065771559501013, + "grad_norm": 1.5461444854736328, + "learning_rate": 8.039245354916641e-06, + "loss": 0.6211, + "step": 13700 + }, + { + "epoch": 0.6066214316549151, + "grad_norm": 1.604701280593872, + "learning_rate": 8.037729926449976e-06, + "loss": 0.6242, + "step": 13701 + }, + { + "epoch": 0.606665707359729, + "grad_norm": 1.5555493831634521, + "learning_rate": 8.036214544850533e-06, + "loss": 0.65, + "step": 13702 + }, + { + "epoch": 0.6067099830645429, + "grad_norm": 1.2665202617645264, + "learning_rate": 8.034699210154507e-06, + "loss": 0.5255, + "step": 13703 + }, + { + "epoch": 0.6067542587693567, + "grad_norm": 1.7460734844207764, + "learning_rate": 8.033183922398091e-06, + "loss": 0.5791, + "step": 13704 + }, + { + "epoch": 0.6067985344741706, + "grad_norm": 1.4932427406311035, + "learning_rate": 8.031668681617476e-06, + "loss": 0.6206, + "step": 13705 + }, + { + "epoch": 0.6068428101789846, + "grad_norm": 1.3416492938995361, + "learning_rate": 8.030153487848849e-06, + "loss": 0.5981, + "step": 13706 + }, + { + "epoch": 0.6068870858837984, + "grad_norm": 1.4886806011199951, + "learning_rate": 8.028638341128405e-06, + "loss": 0.6262, + "step": 13707 + }, + { + "epoch": 0.6069313615886123, + "grad_norm": 1.5095014572143555, + "learning_rate": 8.02712324149233e-06, + "loss": 0.5914, + "step": 13708 + }, + { + "epoch": 0.6069756372934262, + "grad_norm": 1.1895372867584229, + "learning_rate": 8.025608188976803e-06, + "loss": 0.5134, + "step": 13709 + }, + { + "epoch": 0.60701991299824, + "grad_norm": 1.6061151027679443, + "learning_rate": 8.024093183618024e-06, + "loss": 0.6009, + "step": 13710 + }, + { + "epoch": 0.6070641887030539, + "grad_norm": 1.5471789836883545, + "learning_rate": 8.02257822545217e-06, + "loss": 0.5697, + "step": 13711 + }, + { + "epoch": 0.6071084644078678, + "grad_norm": 1.8290566205978394, + "learning_rate": 8.021063314515421e-06, + "loss": 0.6621, + "step": 13712 + }, + { + "epoch": 0.6071527401126817, + "grad_norm": 1.3868989944458008, + "learning_rate": 8.019548450843965e-06, + "loss": 0.5898, + "step": 13713 + }, + { + "epoch": 0.6071970158174955, + "grad_norm": 1.5547995567321777, + "learning_rate": 8.018033634473984e-06, + "loss": 0.5836, + "step": 13714 + }, + { + "epoch": 0.6072412915223094, + "grad_norm": 1.5038827657699585, + "learning_rate": 8.016518865441654e-06, + "loss": 0.5971, + "step": 13715 + }, + { + "epoch": 0.6072855672271233, + "grad_norm": 1.3271961212158203, + "learning_rate": 8.015004143783156e-06, + "loss": 0.5608, + "step": 13716 + }, + { + "epoch": 0.6073298429319371, + "grad_norm": 1.4227750301361084, + "learning_rate": 8.013489469534668e-06, + "loss": 0.6506, + "step": 13717 + }, + { + "epoch": 0.6073741186367511, + "grad_norm": 1.519685983657837, + "learning_rate": 8.011974842732367e-06, + "loss": 0.5806, + "step": 13718 + }, + { + "epoch": 0.607418394341565, + "grad_norm": 2.0420801639556885, + "learning_rate": 8.010460263412426e-06, + "loss": 0.7344, + "step": 13719 + }, + { + "epoch": 0.6074626700463788, + "grad_norm": 1.4606519937515259, + "learning_rate": 8.008945731611024e-06, + "loss": 0.5808, + "step": 13720 + }, + { + "epoch": 0.6075069457511927, + "grad_norm": 1.4146133661270142, + "learning_rate": 8.007431247364333e-06, + "loss": 0.5732, + "step": 13721 + }, + { + "epoch": 0.6075512214560066, + "grad_norm": 1.5536617040634155, + "learning_rate": 8.005916810708518e-06, + "loss": 0.6451, + "step": 13722 + }, + { + "epoch": 0.6075954971608204, + "grad_norm": 1.3455883264541626, + "learning_rate": 8.004402421679762e-06, + "loss": 0.5612, + "step": 13723 + }, + { + "epoch": 0.6076397728656343, + "grad_norm": 1.5152673721313477, + "learning_rate": 8.002888080314229e-06, + "loss": 0.607, + "step": 13724 + }, + { + "epoch": 0.6076840485704482, + "grad_norm": 1.8204761743545532, + "learning_rate": 8.001373786648091e-06, + "loss": 0.6035, + "step": 13725 + }, + { + "epoch": 0.607728324275262, + "grad_norm": 1.3812823295593262, + "learning_rate": 7.999859540717504e-06, + "loss": 0.5608, + "step": 13726 + }, + { + "epoch": 0.6077725999800759, + "grad_norm": 1.5285322666168213, + "learning_rate": 7.99834534255865e-06, + "loss": 0.5406, + "step": 13727 + }, + { + "epoch": 0.6078168756848898, + "grad_norm": 1.2711267471313477, + "learning_rate": 7.996831192207687e-06, + "loss": 0.5631, + "step": 13728 + }, + { + "epoch": 0.6078611513897036, + "grad_norm": 1.6336668729782104, + "learning_rate": 7.995317089700778e-06, + "loss": 0.6003, + "step": 13729 + }, + { + "epoch": 0.6079054270945176, + "grad_norm": 2.0377869606018066, + "learning_rate": 7.993803035074089e-06, + "loss": 0.658, + "step": 13730 + }, + { + "epoch": 0.6079497027993315, + "grad_norm": 1.6245347261428833, + "learning_rate": 7.992289028363782e-06, + "loss": 0.631, + "step": 13731 + }, + { + "epoch": 0.6079939785041453, + "grad_norm": 1.895277738571167, + "learning_rate": 7.990775069606011e-06, + "loss": 0.6266, + "step": 13732 + }, + { + "epoch": 0.6080382542089592, + "grad_norm": 1.6745625734329224, + "learning_rate": 7.989261158836948e-06, + "loss": 0.6291, + "step": 13733 + }, + { + "epoch": 0.6080825299137731, + "grad_norm": 1.441329002380371, + "learning_rate": 7.987747296092747e-06, + "loss": 0.6467, + "step": 13734 + }, + { + "epoch": 0.6081268056185869, + "grad_norm": 1.5509700775146484, + "learning_rate": 7.98623348140956e-06, + "loss": 0.6153, + "step": 13735 + }, + { + "epoch": 0.6081710813234008, + "grad_norm": 1.5430752038955688, + "learning_rate": 7.984719714823544e-06, + "loss": 0.5464, + "step": 13736 + }, + { + "epoch": 0.6082153570282147, + "grad_norm": 1.5181165933609009, + "learning_rate": 7.983205996370858e-06, + "loss": 0.6145, + "step": 13737 + }, + { + "epoch": 0.6082596327330285, + "grad_norm": 1.9005497694015503, + "learning_rate": 7.981692326087657e-06, + "loss": 0.6723, + "step": 13738 + }, + { + "epoch": 0.6083039084378424, + "grad_norm": 1.4395387172698975, + "learning_rate": 7.980178704010089e-06, + "loss": 0.5726, + "step": 13739 + }, + { + "epoch": 0.6083481841426563, + "grad_norm": 1.235862135887146, + "learning_rate": 7.97866513017431e-06, + "loss": 0.5394, + "step": 13740 + }, + { + "epoch": 0.6083924598474701, + "grad_norm": 1.5662838220596313, + "learning_rate": 7.97715160461647e-06, + "loss": 0.6373, + "step": 13741 + }, + { + "epoch": 0.6084367355522841, + "grad_norm": 1.4223114252090454, + "learning_rate": 7.975638127372711e-06, + "loss": 0.577, + "step": 13742 + }, + { + "epoch": 0.608481011257098, + "grad_norm": 1.4929022789001465, + "learning_rate": 7.974124698479192e-06, + "loss": 0.6557, + "step": 13743 + }, + { + "epoch": 0.6085252869619118, + "grad_norm": 1.314064860343933, + "learning_rate": 7.972611317972054e-06, + "loss": 0.5704, + "step": 13744 + }, + { + "epoch": 0.6085695626667257, + "grad_norm": 1.236666202545166, + "learning_rate": 7.971097985887445e-06, + "loss": 0.5283, + "step": 13745 + }, + { + "epoch": 0.6086138383715396, + "grad_norm": 1.5846588611602783, + "learning_rate": 7.969584702261503e-06, + "loss": 0.6358, + "step": 13746 + }, + { + "epoch": 0.6086581140763535, + "grad_norm": 1.942147970199585, + "learning_rate": 7.968071467130383e-06, + "loss": 0.6103, + "step": 13747 + }, + { + "epoch": 0.6087023897811673, + "grad_norm": 1.5349841117858887, + "learning_rate": 7.96655828053022e-06, + "loss": 0.587, + "step": 13748 + }, + { + "epoch": 0.6087466654859812, + "grad_norm": 1.465452790260315, + "learning_rate": 7.965045142497155e-06, + "loss": 0.6134, + "step": 13749 + }, + { + "epoch": 0.6087909411907951, + "grad_norm": 1.5180740356445312, + "learning_rate": 7.963532053067333e-06, + "loss": 0.581, + "step": 13750 + }, + { + "epoch": 0.6088352168956089, + "grad_norm": 1.5707027912139893, + "learning_rate": 7.962019012276888e-06, + "loss": 0.5496, + "step": 13751 + }, + { + "epoch": 0.6088794926004228, + "grad_norm": 1.3121124505996704, + "learning_rate": 7.960506020161962e-06, + "loss": 0.5521, + "step": 13752 + }, + { + "epoch": 0.6089237683052368, + "grad_norm": 1.6196976900100708, + "learning_rate": 7.958993076758685e-06, + "loss": 0.5895, + "step": 13753 + }, + { + "epoch": 0.6089680440100506, + "grad_norm": 1.6456660032272339, + "learning_rate": 7.957480182103198e-06, + "loss": 0.5769, + "step": 13754 + }, + { + "epoch": 0.6090123197148645, + "grad_norm": 1.4919987916946411, + "learning_rate": 7.955967336231635e-06, + "loss": 0.5606, + "step": 13755 + }, + { + "epoch": 0.6090565954196784, + "grad_norm": 1.2088959217071533, + "learning_rate": 7.954454539180124e-06, + "loss": 0.5041, + "step": 13756 + }, + { + "epoch": 0.6091008711244922, + "grad_norm": 1.4494131803512573, + "learning_rate": 7.952941790984804e-06, + "loss": 0.6387, + "step": 13757 + }, + { + "epoch": 0.6091451468293061, + "grad_norm": 1.435080885887146, + "learning_rate": 7.951429091681805e-06, + "loss": 0.5788, + "step": 13758 + }, + { + "epoch": 0.60918942253412, + "grad_norm": 1.4552536010742188, + "learning_rate": 7.949916441307249e-06, + "loss": 0.6343, + "step": 13759 + }, + { + "epoch": 0.6092336982389338, + "grad_norm": 1.5460307598114014, + "learning_rate": 7.948403839897272e-06, + "loss": 0.5754, + "step": 13760 + }, + { + "epoch": 0.6092779739437477, + "grad_norm": 1.3079460859298706, + "learning_rate": 7.946891287488e-06, + "loss": 0.5441, + "step": 13761 + }, + { + "epoch": 0.6093222496485616, + "grad_norm": 1.444394588470459, + "learning_rate": 7.945378784115558e-06, + "loss": 0.5558, + "step": 13762 + }, + { + "epoch": 0.6093665253533754, + "grad_norm": 1.7398860454559326, + "learning_rate": 7.943866329816068e-06, + "loss": 0.6457, + "step": 13763 + }, + { + "epoch": 0.6094108010581893, + "grad_norm": 1.4367363452911377, + "learning_rate": 7.942353924625658e-06, + "loss": 0.6183, + "step": 13764 + }, + { + "epoch": 0.6094550767630033, + "grad_norm": 1.5071922540664673, + "learning_rate": 7.94084156858045e-06, + "loss": 0.5958, + "step": 13765 + }, + { + "epoch": 0.6094993524678171, + "grad_norm": 1.4634664058685303, + "learning_rate": 7.939329261716563e-06, + "loss": 0.5555, + "step": 13766 + }, + { + "epoch": 0.609543628172631, + "grad_norm": 1.63973867893219, + "learning_rate": 7.93781700407012e-06, + "loss": 0.6122, + "step": 13767 + }, + { + "epoch": 0.6095879038774449, + "grad_norm": 1.433742880821228, + "learning_rate": 7.93630479567724e-06, + "loss": 0.5734, + "step": 13768 + }, + { + "epoch": 0.6096321795822587, + "grad_norm": 1.4739160537719727, + "learning_rate": 7.934792636574034e-06, + "loss": 0.601, + "step": 13769 + }, + { + "epoch": 0.6096764552870726, + "grad_norm": 1.3576853275299072, + "learning_rate": 7.933280526796628e-06, + "loss": 0.5387, + "step": 13770 + }, + { + "epoch": 0.6097207309918865, + "grad_norm": 1.3615803718566895, + "learning_rate": 7.931768466381135e-06, + "loss": 0.5367, + "step": 13771 + }, + { + "epoch": 0.6097650066967003, + "grad_norm": 1.6685326099395752, + "learning_rate": 7.930256455363668e-06, + "loss": 0.6683, + "step": 13772 + }, + { + "epoch": 0.6098092824015142, + "grad_norm": 1.2992939949035645, + "learning_rate": 7.928744493780338e-06, + "loss": 0.5814, + "step": 13773 + }, + { + "epoch": 0.6098535581063281, + "grad_norm": 1.3265942335128784, + "learning_rate": 7.92723258166726e-06, + "loss": 0.5689, + "step": 13774 + }, + { + "epoch": 0.6098978338111419, + "grad_norm": 1.9067931175231934, + "learning_rate": 7.925720719060547e-06, + "loss": 0.6835, + "step": 13775 + }, + { + "epoch": 0.6099421095159558, + "grad_norm": 1.6901003122329712, + "learning_rate": 7.924208905996303e-06, + "loss": 0.6367, + "step": 13776 + }, + { + "epoch": 0.6099863852207698, + "grad_norm": 1.5081433057785034, + "learning_rate": 7.92269714251064e-06, + "loss": 0.6277, + "step": 13777 + }, + { + "epoch": 0.6100306609255836, + "grad_norm": 1.4852533340454102, + "learning_rate": 7.921185428639665e-06, + "loss": 0.5467, + "step": 13778 + }, + { + "epoch": 0.6100749366303975, + "grad_norm": 1.7185187339782715, + "learning_rate": 7.919673764419479e-06, + "loss": 0.6268, + "step": 13779 + }, + { + "epoch": 0.6101192123352114, + "grad_norm": 1.3698134422302246, + "learning_rate": 7.918162149886194e-06, + "loss": 0.5973, + "step": 13780 + }, + { + "epoch": 0.6101634880400253, + "grad_norm": 1.4340406656265259, + "learning_rate": 7.916650585075916e-06, + "loss": 0.5796, + "step": 13781 + }, + { + "epoch": 0.6102077637448391, + "grad_norm": 1.3800591230392456, + "learning_rate": 7.915139070024737e-06, + "loss": 0.549, + "step": 13782 + }, + { + "epoch": 0.610252039449653, + "grad_norm": 1.3801231384277344, + "learning_rate": 7.913627604768762e-06, + "loss": 0.5985, + "step": 13783 + }, + { + "epoch": 0.6102963151544669, + "grad_norm": 1.412155270576477, + "learning_rate": 7.912116189344097e-06, + "loss": 0.5743, + "step": 13784 + }, + { + "epoch": 0.6103405908592807, + "grad_norm": 1.1924632787704468, + "learning_rate": 7.910604823786836e-06, + "loss": 0.5687, + "step": 13785 + }, + { + "epoch": 0.6103848665640946, + "grad_norm": 1.2649422883987427, + "learning_rate": 7.909093508133076e-06, + "loss": 0.5917, + "step": 13786 + }, + { + "epoch": 0.6104291422689085, + "grad_norm": 1.5090370178222656, + "learning_rate": 7.907582242418916e-06, + "loss": 0.5877, + "step": 13787 + }, + { + "epoch": 0.6104734179737223, + "grad_norm": 2.06901216506958, + "learning_rate": 7.906071026680451e-06, + "loss": 0.6245, + "step": 13788 + }, + { + "epoch": 0.6105176936785363, + "grad_norm": 1.8554847240447998, + "learning_rate": 7.904559860953771e-06, + "loss": 0.6153, + "step": 13789 + }, + { + "epoch": 0.6105619693833502, + "grad_norm": 1.7057170867919922, + "learning_rate": 7.90304874527498e-06, + "loss": 0.6721, + "step": 13790 + }, + { + "epoch": 0.610606245088164, + "grad_norm": 1.2346882820129395, + "learning_rate": 7.901537679680159e-06, + "loss": 0.5279, + "step": 13791 + }, + { + "epoch": 0.6106505207929779, + "grad_norm": 1.896769642829895, + "learning_rate": 7.9000266642054e-06, + "loss": 0.5509, + "step": 13792 + }, + { + "epoch": 0.6106947964977918, + "grad_norm": 1.4419783353805542, + "learning_rate": 7.898515698886792e-06, + "loss": 0.5978, + "step": 13793 + }, + { + "epoch": 0.6107390722026056, + "grad_norm": 1.8187085390090942, + "learning_rate": 7.89700478376043e-06, + "loss": 0.6593, + "step": 13794 + }, + { + "epoch": 0.6107833479074195, + "grad_norm": 2.101515293121338, + "learning_rate": 7.895493918862395e-06, + "loss": 0.6805, + "step": 13795 + }, + { + "epoch": 0.6108276236122334, + "grad_norm": 1.3925433158874512, + "learning_rate": 7.893983104228772e-06, + "loss": 0.5938, + "step": 13796 + }, + { + "epoch": 0.6108718993170472, + "grad_norm": 1.4108123779296875, + "learning_rate": 7.892472339895651e-06, + "loss": 0.5396, + "step": 13797 + }, + { + "epoch": 0.6109161750218611, + "grad_norm": 1.3482117652893066, + "learning_rate": 7.890961625899112e-06, + "loss": 0.6424, + "step": 13798 + }, + { + "epoch": 0.610960450726675, + "grad_norm": 1.449331283569336, + "learning_rate": 7.889450962275235e-06, + "loss": 0.5997, + "step": 13799 + }, + { + "epoch": 0.6110047264314888, + "grad_norm": 1.50096595287323, + "learning_rate": 7.887940349060106e-06, + "loss": 0.6868, + "step": 13800 + }, + { + "epoch": 0.6110490021363028, + "grad_norm": 2.0311038494110107, + "learning_rate": 7.886429786289803e-06, + "loss": 0.6263, + "step": 13801 + }, + { + "epoch": 0.6110932778411167, + "grad_norm": 2.3835458755493164, + "learning_rate": 7.884919274000401e-06, + "loss": 0.6433, + "step": 13802 + }, + { + "epoch": 0.6111375535459305, + "grad_norm": 1.5800964832305908, + "learning_rate": 7.883408812227977e-06, + "loss": 0.5922, + "step": 13803 + }, + { + "epoch": 0.6111818292507444, + "grad_norm": 1.3178280591964722, + "learning_rate": 7.881898401008614e-06, + "loss": 0.5458, + "step": 13804 + }, + { + "epoch": 0.6112261049555583, + "grad_norm": 1.5735441446304321, + "learning_rate": 7.880388040378386e-06, + "loss": 0.6108, + "step": 13805 + }, + { + "epoch": 0.6112703806603721, + "grad_norm": 1.3796297311782837, + "learning_rate": 7.878877730373356e-06, + "loss": 0.5476, + "step": 13806 + }, + { + "epoch": 0.611314656365186, + "grad_norm": 1.5007996559143066, + "learning_rate": 7.87736747102961e-06, + "loss": 0.5969, + "step": 13807 + }, + { + "epoch": 0.6113589320699999, + "grad_norm": 1.6073479652404785, + "learning_rate": 7.875857262383213e-06, + "loss": 0.6454, + "step": 13808 + }, + { + "epoch": 0.6114032077748137, + "grad_norm": 1.3369874954223633, + "learning_rate": 7.874347104470234e-06, + "loss": 0.5543, + "step": 13809 + }, + { + "epoch": 0.6114474834796276, + "grad_norm": 1.24754798412323, + "learning_rate": 7.872836997326745e-06, + "loss": 0.4945, + "step": 13810 + }, + { + "epoch": 0.6114917591844415, + "grad_norm": 1.3963758945465088, + "learning_rate": 7.871326940988813e-06, + "loss": 0.5556, + "step": 13811 + }, + { + "epoch": 0.6115360348892553, + "grad_norm": 1.2300927639007568, + "learning_rate": 7.869816935492504e-06, + "loss": 0.5481, + "step": 13812 + }, + { + "epoch": 0.6115803105940693, + "grad_norm": 1.4556227922439575, + "learning_rate": 7.86830698087388e-06, + "loss": 0.6065, + "step": 13813 + }, + { + "epoch": 0.6116245862988832, + "grad_norm": 1.3541637659072876, + "learning_rate": 7.86679707716901e-06, + "loss": 0.5799, + "step": 13814 + }, + { + "epoch": 0.611668862003697, + "grad_norm": 1.515470266342163, + "learning_rate": 7.865287224413956e-06, + "loss": 0.5405, + "step": 13815 + }, + { + "epoch": 0.6117131377085109, + "grad_norm": 1.3641586303710938, + "learning_rate": 7.863777422644775e-06, + "loss": 0.5999, + "step": 13816 + }, + { + "epoch": 0.6117574134133248, + "grad_norm": 1.4108604192733765, + "learning_rate": 7.862267671897532e-06, + "loss": 0.609, + "step": 13817 + }, + { + "epoch": 0.6118016891181387, + "grad_norm": 1.402082920074463, + "learning_rate": 7.860757972208288e-06, + "loss": 0.5267, + "step": 13818 + }, + { + "epoch": 0.6118459648229525, + "grad_norm": 1.5493576526641846, + "learning_rate": 7.859248323613095e-06, + "loss": 0.6637, + "step": 13819 + }, + { + "epoch": 0.6118902405277664, + "grad_norm": 1.589438796043396, + "learning_rate": 7.857738726148015e-06, + "loss": 0.5941, + "step": 13820 + }, + { + "epoch": 0.6119345162325803, + "grad_norm": 1.7684451341629028, + "learning_rate": 7.8562291798491e-06, + "loss": 0.6942, + "step": 13821 + }, + { + "epoch": 0.6119787919373941, + "grad_norm": 1.3531008958816528, + "learning_rate": 7.854719684752407e-06, + "loss": 0.5036, + "step": 13822 + }, + { + "epoch": 0.612023067642208, + "grad_norm": 1.5257632732391357, + "learning_rate": 7.853210240893985e-06, + "loss": 0.5716, + "step": 13823 + }, + { + "epoch": 0.612067343347022, + "grad_norm": 1.3927836418151855, + "learning_rate": 7.851700848309891e-06, + "loss": 0.5707, + "step": 13824 + }, + { + "epoch": 0.6121116190518358, + "grad_norm": 1.8676105737686157, + "learning_rate": 7.850191507036172e-06, + "loss": 0.5811, + "step": 13825 + }, + { + "epoch": 0.6121558947566497, + "grad_norm": 1.3104548454284668, + "learning_rate": 7.848682217108875e-06, + "loss": 0.5664, + "step": 13826 + }, + { + "epoch": 0.6122001704614636, + "grad_norm": 1.4082896709442139, + "learning_rate": 7.847172978564055e-06, + "loss": 0.5556, + "step": 13827 + }, + { + "epoch": 0.6122444461662774, + "grad_norm": 1.3786875009536743, + "learning_rate": 7.845663791437756e-06, + "loss": 0.5707, + "step": 13828 + }, + { + "epoch": 0.6122887218710913, + "grad_norm": 1.4611563682556152, + "learning_rate": 7.844154655766022e-06, + "loss": 0.5865, + "step": 13829 + }, + { + "epoch": 0.6123329975759052, + "grad_norm": 1.528395652770996, + "learning_rate": 7.842645571584899e-06, + "loss": 0.6094, + "step": 13830 + }, + { + "epoch": 0.612377273280719, + "grad_norm": 1.3095595836639404, + "learning_rate": 7.841136538930432e-06, + "loss": 0.5515, + "step": 13831 + }, + { + "epoch": 0.6124215489855329, + "grad_norm": 1.4329683780670166, + "learning_rate": 7.83962755783866e-06, + "loss": 0.5722, + "step": 13832 + }, + { + "epoch": 0.6124658246903468, + "grad_norm": 1.499833106994629, + "learning_rate": 7.838118628345621e-06, + "loss": 0.6042, + "step": 13833 + }, + { + "epoch": 0.6125101003951606, + "grad_norm": 1.332349419593811, + "learning_rate": 7.836609750487363e-06, + "loss": 0.5868, + "step": 13834 + }, + { + "epoch": 0.6125543760999745, + "grad_norm": 1.522413969039917, + "learning_rate": 7.835100924299918e-06, + "loss": 0.5403, + "step": 13835 + }, + { + "epoch": 0.6125986518047885, + "grad_norm": 1.7250531911849976, + "learning_rate": 7.83359214981932e-06, + "loss": 0.6056, + "step": 13836 + }, + { + "epoch": 0.6126429275096023, + "grad_norm": 1.3657952547073364, + "learning_rate": 7.832083427081618e-06, + "loss": 0.5812, + "step": 13837 + }, + { + "epoch": 0.6126872032144162, + "grad_norm": 1.917238712310791, + "learning_rate": 7.830574756122835e-06, + "loss": 0.5624, + "step": 13838 + }, + { + "epoch": 0.6127314789192301, + "grad_norm": 1.6776174306869507, + "learning_rate": 7.829066136979003e-06, + "loss": 0.6262, + "step": 13839 + }, + { + "epoch": 0.6127757546240439, + "grad_norm": 1.2408112287521362, + "learning_rate": 7.827557569686163e-06, + "loss": 0.553, + "step": 13840 + }, + { + "epoch": 0.6128200303288578, + "grad_norm": 1.60777747631073, + "learning_rate": 7.826049054280341e-06, + "loss": 0.5747, + "step": 13841 + }, + { + "epoch": 0.6128643060336717, + "grad_norm": 2.2350175380706787, + "learning_rate": 7.82454059079757e-06, + "loss": 0.595, + "step": 13842 + }, + { + "epoch": 0.6129085817384855, + "grad_norm": 1.5873013734817505, + "learning_rate": 7.823032179273871e-06, + "loss": 0.6271, + "step": 13843 + }, + { + "epoch": 0.6129528574432994, + "grad_norm": 1.6286180019378662, + "learning_rate": 7.821523819745279e-06, + "loss": 0.6357, + "step": 13844 + }, + { + "epoch": 0.6129971331481133, + "grad_norm": 1.4314587116241455, + "learning_rate": 7.820015512247816e-06, + "loss": 0.5958, + "step": 13845 + }, + { + "epoch": 0.6130414088529271, + "grad_norm": 2.3332560062408447, + "learning_rate": 7.818507256817505e-06, + "loss": 0.7045, + "step": 13846 + }, + { + "epoch": 0.613085684557741, + "grad_norm": 1.355483889579773, + "learning_rate": 7.816999053490373e-06, + "loss": 0.5562, + "step": 13847 + }, + { + "epoch": 0.613129960262555, + "grad_norm": 1.2380772829055786, + "learning_rate": 7.815490902302444e-06, + "loss": 0.5415, + "step": 13848 + }, + { + "epoch": 0.6131742359673688, + "grad_norm": 1.6062356233596802, + "learning_rate": 7.81398280328973e-06, + "loss": 0.6998, + "step": 13849 + }, + { + "epoch": 0.6132185116721827, + "grad_norm": 1.4435930252075195, + "learning_rate": 7.812474756488262e-06, + "loss": 0.6193, + "step": 13850 + }, + { + "epoch": 0.6132627873769966, + "grad_norm": 1.6659724712371826, + "learning_rate": 7.810966761934053e-06, + "loss": 0.6003, + "step": 13851 + }, + { + "epoch": 0.6133070630818105, + "grad_norm": 1.2657113075256348, + "learning_rate": 7.809458819663123e-06, + "loss": 0.5314, + "step": 13852 + }, + { + "epoch": 0.6133513387866243, + "grad_norm": 1.6145573854446411, + "learning_rate": 7.807950929711478e-06, + "loss": 0.656, + "step": 13853 + }, + { + "epoch": 0.6133956144914382, + "grad_norm": 1.6336379051208496, + "learning_rate": 7.806443092115146e-06, + "loss": 0.6586, + "step": 13854 + }, + { + "epoch": 0.6134398901962521, + "grad_norm": 1.6277869939804077, + "learning_rate": 7.804935306910133e-06, + "loss": 0.6244, + "step": 13855 + }, + { + "epoch": 0.6134841659010659, + "grad_norm": 1.6248372793197632, + "learning_rate": 7.803427574132451e-06, + "loss": 0.6218, + "step": 13856 + }, + { + "epoch": 0.6135284416058798, + "grad_norm": 1.6312110424041748, + "learning_rate": 7.801919893818117e-06, + "loss": 0.617, + "step": 13857 + }, + { + "epoch": 0.6135727173106937, + "grad_norm": 1.4375038146972656, + "learning_rate": 7.800412266003135e-06, + "loss": 0.5265, + "step": 13858 + }, + { + "epoch": 0.6136169930155075, + "grad_norm": 1.3967188596725464, + "learning_rate": 7.798904690723512e-06, + "loss": 0.577, + "step": 13859 + }, + { + "epoch": 0.6136612687203215, + "grad_norm": 1.7198604345321655, + "learning_rate": 7.797397168015261e-06, + "loss": 0.6112, + "step": 13860 + }, + { + "epoch": 0.6137055444251354, + "grad_norm": 1.5414515733718872, + "learning_rate": 7.795889697914387e-06, + "loss": 0.6674, + "step": 13861 + }, + { + "epoch": 0.6137498201299492, + "grad_norm": 1.3446364402770996, + "learning_rate": 7.794382280456892e-06, + "loss": 0.5412, + "step": 13862 + }, + { + "epoch": 0.6137940958347631, + "grad_norm": 1.4328681230545044, + "learning_rate": 7.792874915678776e-06, + "loss": 0.5547, + "step": 13863 + }, + { + "epoch": 0.613838371539577, + "grad_norm": 1.504551649093628, + "learning_rate": 7.791367603616049e-06, + "loss": 0.5909, + "step": 13864 + }, + { + "epoch": 0.6138826472443908, + "grad_norm": 1.501152515411377, + "learning_rate": 7.789860344304709e-06, + "loss": 0.6027, + "step": 13865 + }, + { + "epoch": 0.6139269229492047, + "grad_norm": 1.576012372970581, + "learning_rate": 7.788353137780754e-06, + "loss": 0.5672, + "step": 13866 + }, + { + "epoch": 0.6139711986540186, + "grad_norm": 1.359770655632019, + "learning_rate": 7.786845984080184e-06, + "loss": 0.5308, + "step": 13867 + }, + { + "epoch": 0.6140154743588324, + "grad_norm": 1.5767699480056763, + "learning_rate": 7.785338883238997e-06, + "loss": 0.589, + "step": 13868 + }, + { + "epoch": 0.6140597500636463, + "grad_norm": 1.4736326932907104, + "learning_rate": 7.783831835293186e-06, + "loss": 0.5995, + "step": 13869 + }, + { + "epoch": 0.6141040257684602, + "grad_norm": 1.7289800643920898, + "learning_rate": 7.782324840278746e-06, + "loss": 0.6658, + "step": 13870 + }, + { + "epoch": 0.614148301473274, + "grad_norm": 1.440232515335083, + "learning_rate": 7.780817898231674e-06, + "loss": 0.5366, + "step": 13871 + }, + { + "epoch": 0.614192577178088, + "grad_norm": 1.6171190738677979, + "learning_rate": 7.779311009187958e-06, + "loss": 0.6673, + "step": 13872 + }, + { + "epoch": 0.6142368528829019, + "grad_norm": 1.4405261278152466, + "learning_rate": 7.777804173183589e-06, + "loss": 0.5713, + "step": 13873 + }, + { + "epoch": 0.6142811285877157, + "grad_norm": 1.4994256496429443, + "learning_rate": 7.77629739025456e-06, + "loss": 0.6432, + "step": 13874 + }, + { + "epoch": 0.6143254042925296, + "grad_norm": 1.6683006286621094, + "learning_rate": 7.774790660436857e-06, + "loss": 0.6705, + "step": 13875 + }, + { + "epoch": 0.6143696799973435, + "grad_norm": 1.5538467168807983, + "learning_rate": 7.773283983766466e-06, + "loss": 0.6311, + "step": 13876 + }, + { + "epoch": 0.6144139557021573, + "grad_norm": 1.3379100561141968, + "learning_rate": 7.771777360279375e-06, + "loss": 0.5405, + "step": 13877 + }, + { + "epoch": 0.6144582314069712, + "grad_norm": 1.4817575216293335, + "learning_rate": 7.770270790011569e-06, + "loss": 0.5984, + "step": 13878 + }, + { + "epoch": 0.6145025071117851, + "grad_norm": 2.182821273803711, + "learning_rate": 7.76876427299903e-06, + "loss": 0.6311, + "step": 13879 + }, + { + "epoch": 0.6145467828165989, + "grad_norm": 2.0060675144195557, + "learning_rate": 7.767257809277734e-06, + "loss": 0.6121, + "step": 13880 + }, + { + "epoch": 0.6145910585214128, + "grad_norm": 1.2370619773864746, + "learning_rate": 7.765751398883673e-06, + "loss": 0.5689, + "step": 13881 + }, + { + "epoch": 0.6146353342262267, + "grad_norm": 1.4379429817199707, + "learning_rate": 7.76424504185282e-06, + "loss": 0.6008, + "step": 13882 + }, + { + "epoch": 0.6146796099310405, + "grad_norm": 1.5192674398422241, + "learning_rate": 7.76273873822115e-06, + "loss": 0.5948, + "step": 13883 + }, + { + "epoch": 0.6147238856358545, + "grad_norm": 1.5344936847686768, + "learning_rate": 7.761232488024647e-06, + "loss": 0.6182, + "step": 13884 + }, + { + "epoch": 0.6147681613406684, + "grad_norm": 1.5146898031234741, + "learning_rate": 7.759726291299285e-06, + "loss": 0.6373, + "step": 13885 + }, + { + "epoch": 0.6148124370454822, + "grad_norm": 1.4266963005065918, + "learning_rate": 7.75822014808103e-06, + "loss": 0.5755, + "step": 13886 + }, + { + "epoch": 0.6148567127502961, + "grad_norm": 1.2865757942199707, + "learning_rate": 7.756714058405866e-06, + "loss": 0.5839, + "step": 13887 + }, + { + "epoch": 0.61490098845511, + "grad_norm": 1.3977091312408447, + "learning_rate": 7.755208022309759e-06, + "loss": 0.5953, + "step": 13888 + }, + { + "epoch": 0.6149452641599239, + "grad_norm": 1.4134076833724976, + "learning_rate": 7.75370203982868e-06, + "loss": 0.5708, + "step": 13889 + }, + { + "epoch": 0.6149895398647377, + "grad_norm": 1.3866386413574219, + "learning_rate": 7.7521961109986e-06, + "loss": 0.5675, + "step": 13890 + }, + { + "epoch": 0.6150338155695516, + "grad_norm": 1.441605806350708, + "learning_rate": 7.750690235855485e-06, + "loss": 0.5932, + "step": 13891 + }, + { + "epoch": 0.6150780912743655, + "grad_norm": 1.3816955089569092, + "learning_rate": 7.749184414435304e-06, + "loss": 0.5747, + "step": 13892 + }, + { + "epoch": 0.6151223669791793, + "grad_norm": 1.496647596359253, + "learning_rate": 7.747678646774018e-06, + "loss": 0.6123, + "step": 13893 + }, + { + "epoch": 0.6151666426839932, + "grad_norm": 1.3464932441711426, + "learning_rate": 7.746172932907596e-06, + "loss": 0.5789, + "step": 13894 + }, + { + "epoch": 0.6152109183888071, + "grad_norm": 1.9201409816741943, + "learning_rate": 7.744667272871998e-06, + "loss": 0.6293, + "step": 13895 + }, + { + "epoch": 0.615255194093621, + "grad_norm": 1.4514826536178589, + "learning_rate": 7.74316166670318e-06, + "loss": 0.5792, + "step": 13896 + }, + { + "epoch": 0.6152994697984349, + "grad_norm": 1.304382562637329, + "learning_rate": 7.741656114437115e-06, + "loss": 0.5426, + "step": 13897 + }, + { + "epoch": 0.6153437455032488, + "grad_norm": 1.664570927619934, + "learning_rate": 7.740150616109754e-06, + "loss": 0.5919, + "step": 13898 + }, + { + "epoch": 0.6153880212080626, + "grad_norm": 1.5175319910049438, + "learning_rate": 7.738645171757054e-06, + "loss": 0.6275, + "step": 13899 + }, + { + "epoch": 0.6154322969128765, + "grad_norm": 1.753216028213501, + "learning_rate": 7.737139781414973e-06, + "loss": 0.6003, + "step": 13900 + }, + { + "epoch": 0.6154765726176904, + "grad_norm": 2.2456793785095215, + "learning_rate": 7.735634445119467e-06, + "loss": 0.6326, + "step": 13901 + }, + { + "epoch": 0.6155208483225042, + "grad_norm": 1.5164670944213867, + "learning_rate": 7.734129162906488e-06, + "loss": 0.5859, + "step": 13902 + }, + { + "epoch": 0.6155651240273181, + "grad_norm": 1.4450408220291138, + "learning_rate": 7.732623934811987e-06, + "loss": 0.5352, + "step": 13903 + }, + { + "epoch": 0.615609399732132, + "grad_norm": 1.5896327495574951, + "learning_rate": 7.731118760871918e-06, + "loss": 0.5793, + "step": 13904 + }, + { + "epoch": 0.6156536754369458, + "grad_norm": 1.2647181749343872, + "learning_rate": 7.729613641122231e-06, + "loss": 0.5587, + "step": 13905 + }, + { + "epoch": 0.6156979511417597, + "grad_norm": 1.3482741117477417, + "learning_rate": 7.728108575598871e-06, + "loss": 0.544, + "step": 13906 + }, + { + "epoch": 0.6157422268465736, + "grad_norm": 1.4476319551467896, + "learning_rate": 7.726603564337791e-06, + "loss": 0.5852, + "step": 13907 + }, + { + "epoch": 0.6157865025513874, + "grad_norm": 1.5800567865371704, + "learning_rate": 7.725098607374935e-06, + "loss": 0.5973, + "step": 13908 + }, + { + "epoch": 0.6158307782562014, + "grad_norm": 1.464361548423767, + "learning_rate": 7.723593704746246e-06, + "loss": 0.5971, + "step": 13909 + }, + { + "epoch": 0.6158750539610153, + "grad_norm": 1.6071388721466064, + "learning_rate": 7.722088856487664e-06, + "loss": 0.6748, + "step": 13910 + }, + { + "epoch": 0.6159193296658291, + "grad_norm": 1.4882341623306274, + "learning_rate": 7.720584062635137e-06, + "loss": 0.6036, + "step": 13911 + }, + { + "epoch": 0.615963605370643, + "grad_norm": 1.4607118368148804, + "learning_rate": 7.719079323224604e-06, + "loss": 0.5583, + "step": 13912 + }, + { + "epoch": 0.6160078810754569, + "grad_norm": 1.7952735424041748, + "learning_rate": 7.717574638292002e-06, + "loss": 0.6259, + "step": 13913 + }, + { + "epoch": 0.6160521567802707, + "grad_norm": 1.4131748676300049, + "learning_rate": 7.716070007873275e-06, + "loss": 0.6464, + "step": 13914 + }, + { + "epoch": 0.6160964324850846, + "grad_norm": 1.5448065996170044, + "learning_rate": 7.714565432004355e-06, + "loss": 0.5705, + "step": 13915 + }, + { + "epoch": 0.6161407081898985, + "grad_norm": 1.735840082168579, + "learning_rate": 7.713060910721175e-06, + "loss": 0.6595, + "step": 13916 + }, + { + "epoch": 0.6161849838947123, + "grad_norm": 1.6014646291732788, + "learning_rate": 7.71155644405968e-06, + "loss": 0.6228, + "step": 13917 + }, + { + "epoch": 0.6162292595995262, + "grad_norm": 1.499517798423767, + "learning_rate": 7.710052032055792e-06, + "loss": 0.6205, + "step": 13918 + }, + { + "epoch": 0.6162735353043401, + "grad_norm": 1.5367493629455566, + "learning_rate": 7.708547674745447e-06, + "loss": 0.6229, + "step": 13919 + }, + { + "epoch": 0.616317811009154, + "grad_norm": 1.2806873321533203, + "learning_rate": 7.707043372164573e-06, + "loss": 0.525, + "step": 13920 + }, + { + "epoch": 0.6163620867139679, + "grad_norm": 1.400887131690979, + "learning_rate": 7.705539124349105e-06, + "loss": 0.6002, + "step": 13921 + }, + { + "epoch": 0.6164063624187818, + "grad_norm": 1.3901515007019043, + "learning_rate": 7.704034931334964e-06, + "loss": 0.5496, + "step": 13922 + }, + { + "epoch": 0.6164506381235957, + "grad_norm": 1.3436524868011475, + "learning_rate": 7.702530793158079e-06, + "loss": 0.565, + "step": 13923 + }, + { + "epoch": 0.6164949138284095, + "grad_norm": 1.5767595767974854, + "learning_rate": 7.701026709854378e-06, + "loss": 0.5984, + "step": 13924 + }, + { + "epoch": 0.6165391895332234, + "grad_norm": 1.3163044452667236, + "learning_rate": 7.699522681459781e-06, + "loss": 0.5727, + "step": 13925 + }, + { + "epoch": 0.6165834652380373, + "grad_norm": 1.5863256454467773, + "learning_rate": 7.698018708010211e-06, + "loss": 0.6114, + "step": 13926 + }, + { + "epoch": 0.6166277409428511, + "grad_norm": 1.681341290473938, + "learning_rate": 7.69651478954159e-06, + "loss": 0.6302, + "step": 13927 + }, + { + "epoch": 0.616672016647665, + "grad_norm": 1.6522369384765625, + "learning_rate": 7.695010926089838e-06, + "loss": 0.6687, + "step": 13928 + }, + { + "epoch": 0.6167162923524789, + "grad_norm": 1.3484108448028564, + "learning_rate": 7.693507117690875e-06, + "loss": 0.6027, + "step": 13929 + }, + { + "epoch": 0.6167605680572927, + "grad_norm": 1.2463284730911255, + "learning_rate": 7.692003364380611e-06, + "loss": 0.5363, + "step": 13930 + }, + { + "epoch": 0.6168048437621066, + "grad_norm": 1.3944991827011108, + "learning_rate": 7.690499666194971e-06, + "loss": 0.5635, + "step": 13931 + }, + { + "epoch": 0.6168491194669206, + "grad_norm": 1.5140398740768433, + "learning_rate": 7.688996023169869e-06, + "loss": 0.5813, + "step": 13932 + }, + { + "epoch": 0.6168933951717344, + "grad_norm": 1.5596272945404053, + "learning_rate": 7.687492435341208e-06, + "loss": 0.5698, + "step": 13933 + }, + { + "epoch": 0.6169376708765483, + "grad_norm": 1.391408920288086, + "learning_rate": 7.685988902744912e-06, + "loss": 0.5861, + "step": 13934 + }, + { + "epoch": 0.6169819465813622, + "grad_norm": 1.9556896686553955, + "learning_rate": 7.684485425416888e-06, + "loss": 0.6162, + "step": 13935 + }, + { + "epoch": 0.617026222286176, + "grad_norm": 1.4631619453430176, + "learning_rate": 7.682982003393041e-06, + "loss": 0.5896, + "step": 13936 + }, + { + "epoch": 0.6170704979909899, + "grad_norm": 1.522993564605713, + "learning_rate": 7.681478636709285e-06, + "loss": 0.6134, + "step": 13937 + }, + { + "epoch": 0.6171147736958038, + "grad_norm": 1.7617056369781494, + "learning_rate": 7.679975325401524e-06, + "loss": 0.6197, + "step": 13938 + }, + { + "epoch": 0.6171590494006176, + "grad_norm": 1.2425838708877563, + "learning_rate": 7.678472069505663e-06, + "loss": 0.5485, + "step": 13939 + }, + { + "epoch": 0.6172033251054315, + "grad_norm": 1.6155329942703247, + "learning_rate": 7.676968869057603e-06, + "loss": 0.5366, + "step": 13940 + }, + { + "epoch": 0.6172476008102454, + "grad_norm": 1.4852111339569092, + "learning_rate": 7.675465724093254e-06, + "loss": 0.5945, + "step": 13941 + }, + { + "epoch": 0.6172918765150592, + "grad_norm": 1.4535616636276245, + "learning_rate": 7.673962634648513e-06, + "loss": 0.5388, + "step": 13942 + }, + { + "epoch": 0.6173361522198731, + "grad_norm": 1.5327730178833008, + "learning_rate": 7.672459600759276e-06, + "loss": 0.5745, + "step": 13943 + }, + { + "epoch": 0.6173804279246871, + "grad_norm": 1.2995591163635254, + "learning_rate": 7.670956622461451e-06, + "loss": 0.6042, + "step": 13944 + }, + { + "epoch": 0.6174247036295009, + "grad_norm": 1.6839189529418945, + "learning_rate": 7.669453699790931e-06, + "loss": 0.5781, + "step": 13945 + }, + { + "epoch": 0.6174689793343148, + "grad_norm": 1.75608229637146, + "learning_rate": 7.667950832783608e-06, + "loss": 0.6506, + "step": 13946 + }, + { + "epoch": 0.6175132550391287, + "grad_norm": 1.1570082902908325, + "learning_rate": 7.666448021475385e-06, + "loss": 0.4602, + "step": 13947 + }, + { + "epoch": 0.6175575307439425, + "grad_norm": 1.4688307046890259, + "learning_rate": 7.664945265902148e-06, + "loss": 0.6523, + "step": 13948 + }, + { + "epoch": 0.6176018064487564, + "grad_norm": 1.2529314756393433, + "learning_rate": 7.663442566099795e-06, + "loss": 0.5976, + "step": 13949 + }, + { + "epoch": 0.6176460821535703, + "grad_norm": 1.4874268770217896, + "learning_rate": 7.66193992210421e-06, + "loss": 0.6048, + "step": 13950 + }, + { + "epoch": 0.6176903578583841, + "grad_norm": 1.750862717628479, + "learning_rate": 7.66043733395129e-06, + "loss": 0.634, + "step": 13951 + }, + { + "epoch": 0.617734633563198, + "grad_norm": 1.4612067937850952, + "learning_rate": 7.658934801676917e-06, + "loss": 0.5574, + "step": 13952 + }, + { + "epoch": 0.6177789092680119, + "grad_norm": 1.498268485069275, + "learning_rate": 7.657432325316978e-06, + "loss": 0.5835, + "step": 13953 + }, + { + "epoch": 0.6178231849728257, + "grad_norm": 1.5322202444076538, + "learning_rate": 7.655929904907363e-06, + "loss": 0.5775, + "step": 13954 + }, + { + "epoch": 0.6178674606776396, + "grad_norm": 1.4835284948349, + "learning_rate": 7.654427540483954e-06, + "loss": 0.5637, + "step": 13955 + }, + { + "epoch": 0.6179117363824536, + "grad_norm": 1.481630563735962, + "learning_rate": 7.65292523208263e-06, + "loss": 0.6124, + "step": 13956 + }, + { + "epoch": 0.6179560120872674, + "grad_norm": 1.6551724672317505, + "learning_rate": 7.65142297973928e-06, + "loss": 0.5849, + "step": 13957 + }, + { + "epoch": 0.6180002877920813, + "grad_norm": 1.6576677560806274, + "learning_rate": 7.64992078348978e-06, + "loss": 0.647, + "step": 13958 + }, + { + "epoch": 0.6180445634968952, + "grad_norm": 1.3977264165878296, + "learning_rate": 7.648418643370008e-06, + "loss": 0.5336, + "step": 13959 + }, + { + "epoch": 0.6180888392017091, + "grad_norm": 1.9765249490737915, + "learning_rate": 7.64691655941584e-06, + "loss": 0.7044, + "step": 13960 + }, + { + "epoch": 0.6181331149065229, + "grad_norm": 1.508791208267212, + "learning_rate": 7.645414531663155e-06, + "loss": 0.6055, + "step": 13961 + }, + { + "epoch": 0.6181773906113368, + "grad_norm": 1.2396758794784546, + "learning_rate": 7.643912560147829e-06, + "loss": 0.5127, + "step": 13962 + }, + { + "epoch": 0.6182216663161507, + "grad_norm": 1.5797836780548096, + "learning_rate": 7.642410644905726e-06, + "loss": 0.6454, + "step": 13963 + }, + { + "epoch": 0.6182659420209645, + "grad_norm": 1.4177628755569458, + "learning_rate": 7.640908785972734e-06, + "loss": 0.6006, + "step": 13964 + }, + { + "epoch": 0.6183102177257784, + "grad_norm": 1.4427436590194702, + "learning_rate": 7.639406983384711e-06, + "loss": 0.518, + "step": 13965 + }, + { + "epoch": 0.6183544934305923, + "grad_norm": 1.4201170206069946, + "learning_rate": 7.637905237177527e-06, + "loss": 0.6225, + "step": 13966 + }, + { + "epoch": 0.6183987691354061, + "grad_norm": 1.5601396560668945, + "learning_rate": 7.636403547387056e-06, + "loss": 0.5946, + "step": 13967 + }, + { + "epoch": 0.6184430448402201, + "grad_norm": 1.4245378971099854, + "learning_rate": 7.634901914049164e-06, + "loss": 0.5407, + "step": 13968 + }, + { + "epoch": 0.618487320545034, + "grad_norm": 1.5424860715866089, + "learning_rate": 7.633400337199712e-06, + "loss": 0.622, + "step": 13969 + }, + { + "epoch": 0.6185315962498478, + "grad_norm": 1.7296644449234009, + "learning_rate": 7.631898816874565e-06, + "loss": 0.5836, + "step": 13970 + }, + { + "epoch": 0.6185758719546617, + "grad_norm": 1.665192723274231, + "learning_rate": 7.630397353109588e-06, + "loss": 0.6205, + "step": 13971 + }, + { + "epoch": 0.6186201476594756, + "grad_norm": 1.6264948844909668, + "learning_rate": 7.628895945940641e-06, + "loss": 0.6436, + "step": 13972 + }, + { + "epoch": 0.6186644233642894, + "grad_norm": 1.5824122428894043, + "learning_rate": 7.627394595403582e-06, + "loss": 0.6097, + "step": 13973 + }, + { + "epoch": 0.6187086990691033, + "grad_norm": 1.415450096130371, + "learning_rate": 7.625893301534273e-06, + "loss": 0.5716, + "step": 13974 + }, + { + "epoch": 0.6187529747739172, + "grad_norm": 1.7701116800308228, + "learning_rate": 7.62439206436857e-06, + "loss": 0.58, + "step": 13975 + }, + { + "epoch": 0.618797250478731, + "grad_norm": 1.932145595550537, + "learning_rate": 7.622890883942326e-06, + "loss": 0.6467, + "step": 13976 + }, + { + "epoch": 0.6188415261835449, + "grad_norm": 1.5477652549743652, + "learning_rate": 7.621389760291397e-06, + "loss": 0.5726, + "step": 13977 + }, + { + "epoch": 0.6188858018883588, + "grad_norm": 1.4957318305969238, + "learning_rate": 7.619888693451638e-06, + "loss": 0.6577, + "step": 13978 + }, + { + "epoch": 0.6189300775931726, + "grad_norm": 1.4727365970611572, + "learning_rate": 7.618387683458903e-06, + "loss": 0.6157, + "step": 13979 + }, + { + "epoch": 0.6189743532979866, + "grad_norm": 1.2964524030685425, + "learning_rate": 7.616886730349033e-06, + "loss": 0.629, + "step": 13980 + }, + { + "epoch": 0.6190186290028005, + "grad_norm": 1.4969183206558228, + "learning_rate": 7.615385834157887e-06, + "loss": 0.5864, + "step": 13981 + }, + { + "epoch": 0.6190629047076143, + "grad_norm": 1.5818220376968384, + "learning_rate": 7.613884994921309e-06, + "loss": 0.5998, + "step": 13982 + }, + { + "epoch": 0.6191071804124282, + "grad_norm": 1.2731572389602661, + "learning_rate": 7.612384212675142e-06, + "loss": 0.5314, + "step": 13983 + }, + { + "epoch": 0.6191514561172421, + "grad_norm": 1.4741429090499878, + "learning_rate": 7.610883487455237e-06, + "loss": 0.5612, + "step": 13984 + }, + { + "epoch": 0.6191957318220559, + "grad_norm": 1.2412644624710083, + "learning_rate": 7.609382819297435e-06, + "loss": 0.5516, + "step": 13985 + }, + { + "epoch": 0.6192400075268698, + "grad_norm": 1.2783021926879883, + "learning_rate": 7.607882208237577e-06, + "loss": 0.5778, + "step": 13986 + }, + { + "epoch": 0.6192842832316837, + "grad_norm": 1.3550838232040405, + "learning_rate": 7.606381654311501e-06, + "loss": 0.5979, + "step": 13987 + }, + { + "epoch": 0.6193285589364975, + "grad_norm": 1.6164700984954834, + "learning_rate": 7.6048811575550556e-06, + "loss": 0.5858, + "step": 13988 + }, + { + "epoch": 0.6193728346413114, + "grad_norm": 1.4327746629714966, + "learning_rate": 7.603380718004072e-06, + "loss": 0.5852, + "step": 13989 + }, + { + "epoch": 0.6194171103461253, + "grad_norm": 1.277834177017212, + "learning_rate": 7.601880335694385e-06, + "loss": 0.5246, + "step": 13990 + }, + { + "epoch": 0.6194613860509391, + "grad_norm": 1.385108470916748, + "learning_rate": 7.600380010661836e-06, + "loss": 0.5624, + "step": 13991 + }, + { + "epoch": 0.6195056617557531, + "grad_norm": 1.7178274393081665, + "learning_rate": 7.598879742942259e-06, + "loss": 0.6296, + "step": 13992 + }, + { + "epoch": 0.619549937460567, + "grad_norm": 1.31501305103302, + "learning_rate": 7.597379532571481e-06, + "loss": 0.5025, + "step": 13993 + }, + { + "epoch": 0.6195942131653809, + "grad_norm": 1.7926769256591797, + "learning_rate": 7.595879379585338e-06, + "loss": 0.6375, + "step": 13994 + }, + { + "epoch": 0.6196384888701947, + "grad_norm": 1.5867335796356201, + "learning_rate": 7.594379284019659e-06, + "loss": 0.6134, + "step": 13995 + }, + { + "epoch": 0.6196827645750086, + "grad_norm": 1.3177566528320312, + "learning_rate": 7.592879245910273e-06, + "loss": 0.55, + "step": 13996 + }, + { + "epoch": 0.6197270402798225, + "grad_norm": 1.315426230430603, + "learning_rate": 7.591379265293003e-06, + "loss": 0.569, + "step": 13997 + }, + { + "epoch": 0.6197713159846363, + "grad_norm": 1.3267816305160522, + "learning_rate": 7.589879342203681e-06, + "loss": 0.5918, + "step": 13998 + }, + { + "epoch": 0.6198155916894502, + "grad_norm": 1.55797278881073, + "learning_rate": 7.5883794766781275e-06, + "loss": 0.6241, + "step": 13999 + }, + { + "epoch": 0.6198598673942641, + "grad_norm": 1.2827304601669312, + "learning_rate": 7.586879668752163e-06, + "loss": 0.5865, + "step": 14000 + }, + { + "epoch": 0.6199041430990779, + "grad_norm": 1.5087753534317017, + "learning_rate": 7.585379918461617e-06, + "loss": 0.6366, + "step": 14001 + }, + { + "epoch": 0.6199484188038918, + "grad_norm": 1.5463616847991943, + "learning_rate": 7.583880225842306e-06, + "loss": 0.605, + "step": 14002 + }, + { + "epoch": 0.6199926945087058, + "grad_norm": 1.4418433904647827, + "learning_rate": 7.582380590930045e-06, + "loss": 0.5906, + "step": 14003 + }, + { + "epoch": 0.6200369702135196, + "grad_norm": 1.43027925491333, + "learning_rate": 7.5808810137606595e-06, + "loss": 0.5938, + "step": 14004 + }, + { + "epoch": 0.6200812459183335, + "grad_norm": 1.3882780075073242, + "learning_rate": 7.579381494369959e-06, + "loss": 0.5829, + "step": 14005 + }, + { + "epoch": 0.6201255216231474, + "grad_norm": 1.428924560546875, + "learning_rate": 7.577882032793762e-06, + "loss": 0.5646, + "step": 14006 + }, + { + "epoch": 0.6201697973279612, + "grad_norm": 1.4607247114181519, + "learning_rate": 7.576382629067877e-06, + "loss": 0.6407, + "step": 14007 + }, + { + "epoch": 0.6202140730327751, + "grad_norm": 1.7693980932235718, + "learning_rate": 7.574883283228123e-06, + "loss": 0.6113, + "step": 14008 + }, + { + "epoch": 0.620258348737589, + "grad_norm": 1.4805549383163452, + "learning_rate": 7.573383995310307e-06, + "loss": 0.5388, + "step": 14009 + }, + { + "epoch": 0.6203026244424028, + "grad_norm": 1.8390387296676636, + "learning_rate": 7.571884765350234e-06, + "loss": 0.6378, + "step": 14010 + }, + { + "epoch": 0.6203469001472167, + "grad_norm": 1.2738935947418213, + "learning_rate": 7.5703855933837204e-06, + "loss": 0.5682, + "step": 14011 + }, + { + "epoch": 0.6203911758520306, + "grad_norm": 1.364106297492981, + "learning_rate": 7.568886479446571e-06, + "loss": 0.5558, + "step": 14012 + }, + { + "epoch": 0.6204354515568444, + "grad_norm": 1.492828369140625, + "learning_rate": 7.567387423574582e-06, + "loss": 0.6154, + "step": 14013 + }, + { + "epoch": 0.6204797272616583, + "grad_norm": 1.358870267868042, + "learning_rate": 7.565888425803568e-06, + "loss": 0.5346, + "step": 14014 + }, + { + "epoch": 0.6205240029664723, + "grad_norm": 1.5894203186035156, + "learning_rate": 7.564389486169327e-06, + "loss": 0.6004, + "step": 14015 + }, + { + "epoch": 0.620568278671286, + "grad_norm": 1.5550730228424072, + "learning_rate": 7.56289060470766e-06, + "loss": 0.6277, + "step": 14016 + }, + { + "epoch": 0.6206125543761, + "grad_norm": 1.9775104522705078, + "learning_rate": 7.561391781454364e-06, + "loss": 0.6103, + "step": 14017 + }, + { + "epoch": 0.6206568300809139, + "grad_norm": 1.5904433727264404, + "learning_rate": 7.559893016445242e-06, + "loss": 0.62, + "step": 14018 + }, + { + "epoch": 0.6207011057857277, + "grad_norm": 1.7555832862854004, + "learning_rate": 7.558394309716088e-06, + "loss": 0.6951, + "step": 14019 + }, + { + "epoch": 0.6207453814905416, + "grad_norm": 1.375096321105957, + "learning_rate": 7.556895661302696e-06, + "loss": 0.5868, + "step": 14020 + }, + { + "epoch": 0.6207896571953555, + "grad_norm": 1.3465169668197632, + "learning_rate": 7.555397071240866e-06, + "loss": 0.5671, + "step": 14021 + }, + { + "epoch": 0.6208339329001693, + "grad_norm": 1.6322258710861206, + "learning_rate": 7.553898539566384e-06, + "loss": 0.6154, + "step": 14022 + }, + { + "epoch": 0.6208782086049832, + "grad_norm": 1.3724186420440674, + "learning_rate": 7.552400066315041e-06, + "loss": 0.5728, + "step": 14023 + }, + { + "epoch": 0.6209224843097971, + "grad_norm": 1.6196730136871338, + "learning_rate": 7.550901651522633e-06, + "loss": 0.6299, + "step": 14024 + }, + { + "epoch": 0.6209667600146109, + "grad_norm": 1.6022897958755493, + "learning_rate": 7.5494032952249456e-06, + "loss": 0.6061, + "step": 14025 + }, + { + "epoch": 0.6210110357194248, + "grad_norm": 1.2207883596420288, + "learning_rate": 7.547904997457764e-06, + "loss": 0.4919, + "step": 14026 + }, + { + "epoch": 0.6210553114242388, + "grad_norm": 1.4765101671218872, + "learning_rate": 7.546406758256873e-06, + "loss": 0.6173, + "step": 14027 + }, + { + "epoch": 0.6210995871290527, + "grad_norm": 1.578371286392212, + "learning_rate": 7.544908577658062e-06, + "loss": 0.6261, + "step": 14028 + }, + { + "epoch": 0.6211438628338665, + "grad_norm": 1.3865617513656616, + "learning_rate": 7.54341045569711e-06, + "loss": 0.5695, + "step": 14029 + }, + { + "epoch": 0.6211881385386804, + "grad_norm": 1.7782831192016602, + "learning_rate": 7.541912392409797e-06, + "loss": 0.6022, + "step": 14030 + }, + { + "epoch": 0.6212324142434943, + "grad_norm": 1.616211175918579, + "learning_rate": 7.540414387831907e-06, + "loss": 0.5702, + "step": 14031 + }, + { + "epoch": 0.6212766899483081, + "grad_norm": 1.2590482234954834, + "learning_rate": 7.5389164419992166e-06, + "loss": 0.5087, + "step": 14032 + }, + { + "epoch": 0.621320965653122, + "grad_norm": 1.456663966178894, + "learning_rate": 7.537418554947499e-06, + "loss": 0.5364, + "step": 14033 + }, + { + "epoch": 0.6213652413579359, + "grad_norm": 1.5042589902877808, + "learning_rate": 7.535920726712538e-06, + "loss": 0.5372, + "step": 14034 + }, + { + "epoch": 0.6214095170627497, + "grad_norm": 1.609182357788086, + "learning_rate": 7.534422957330108e-06, + "loss": 0.6203, + "step": 14035 + }, + { + "epoch": 0.6214537927675636, + "grad_norm": 1.8579798936843872, + "learning_rate": 7.532925246835974e-06, + "loss": 0.6533, + "step": 14036 + }, + { + "epoch": 0.6214980684723775, + "grad_norm": 1.6422572135925293, + "learning_rate": 7.531427595265908e-06, + "loss": 0.5694, + "step": 14037 + }, + { + "epoch": 0.6215423441771913, + "grad_norm": 1.33191978931427, + "learning_rate": 7.529930002655687e-06, + "loss": 0.5711, + "step": 14038 + }, + { + "epoch": 0.6215866198820053, + "grad_norm": 1.4775530099868774, + "learning_rate": 7.528432469041078e-06, + "loss": 0.6142, + "step": 14039 + }, + { + "epoch": 0.6216308955868192, + "grad_norm": 1.4843974113464355, + "learning_rate": 7.5269349944578454e-06, + "loss": 0.5864, + "step": 14040 + }, + { + "epoch": 0.621675171291633, + "grad_norm": 1.2624937295913696, + "learning_rate": 7.525437578941759e-06, + "loss": 0.5341, + "step": 14041 + }, + { + "epoch": 0.6217194469964469, + "grad_norm": 1.4446042776107788, + "learning_rate": 7.5239402225285796e-06, + "loss": 0.6249, + "step": 14042 + }, + { + "epoch": 0.6217637227012608, + "grad_norm": 1.3786425590515137, + "learning_rate": 7.522442925254068e-06, + "loss": 0.5482, + "step": 14043 + }, + { + "epoch": 0.6218079984060746, + "grad_norm": 1.3252142667770386, + "learning_rate": 7.520945687153998e-06, + "loss": 0.5583, + "step": 14044 + }, + { + "epoch": 0.6218522741108885, + "grad_norm": 1.639007806777954, + "learning_rate": 7.519448508264118e-06, + "loss": 0.5708, + "step": 14045 + }, + { + "epoch": 0.6218965498157024, + "grad_norm": 1.499841332435608, + "learning_rate": 7.517951388620191e-06, + "loss": 0.5236, + "step": 14046 + }, + { + "epoch": 0.6219408255205162, + "grad_norm": 1.384661078453064, + "learning_rate": 7.516454328257969e-06, + "loss": 0.6199, + "step": 14047 + }, + { + "epoch": 0.6219851012253301, + "grad_norm": 1.4347411394119263, + "learning_rate": 7.514957327213219e-06, + "loss": 0.5748, + "step": 14048 + }, + { + "epoch": 0.622029376930144, + "grad_norm": 1.8000800609588623, + "learning_rate": 7.513460385521689e-06, + "loss": 0.6461, + "step": 14049 + }, + { + "epoch": 0.6220736526349578, + "grad_norm": 1.4990302324295044, + "learning_rate": 7.511963503219132e-06, + "loss": 0.6001, + "step": 14050 + }, + { + "epoch": 0.6221179283397718, + "grad_norm": 1.440036416053772, + "learning_rate": 7.5104666803413015e-06, + "loss": 0.607, + "step": 14051 + }, + { + "epoch": 0.6221622040445857, + "grad_norm": 1.5249024629592896, + "learning_rate": 7.508969916923948e-06, + "loss": 0.6065, + "step": 14052 + }, + { + "epoch": 0.6222064797493995, + "grad_norm": 1.4187688827514648, + "learning_rate": 7.507473213002818e-06, + "loss": 0.551, + "step": 14053 + }, + { + "epoch": 0.6222507554542134, + "grad_norm": 1.3829562664031982, + "learning_rate": 7.505976568613662e-06, + "loss": 0.6156, + "step": 14054 + }, + { + "epoch": 0.6222950311590273, + "grad_norm": 1.3796706199645996, + "learning_rate": 7.5044799837922255e-06, + "loss": 0.6261, + "step": 14055 + }, + { + "epoch": 0.6223393068638411, + "grad_norm": 1.7511626482009888, + "learning_rate": 7.5029834585742535e-06, + "loss": 0.6744, + "step": 14056 + }, + { + "epoch": 0.622383582568655, + "grad_norm": 1.482853889465332, + "learning_rate": 7.501486992995483e-06, + "loss": 0.59, + "step": 14057 + }, + { + "epoch": 0.6224278582734689, + "grad_norm": 1.5083972215652466, + "learning_rate": 7.499990587091666e-06, + "loss": 0.5583, + "step": 14058 + }, + { + "epoch": 0.6224721339782827, + "grad_norm": 1.2986114025115967, + "learning_rate": 7.49849424089854e-06, + "loss": 0.5236, + "step": 14059 + }, + { + "epoch": 0.6225164096830966, + "grad_norm": 1.3693526983261108, + "learning_rate": 7.496997954451836e-06, + "loss": 0.6056, + "step": 14060 + }, + { + "epoch": 0.6225606853879105, + "grad_norm": 1.3021475076675415, + "learning_rate": 7.495501727787302e-06, + "loss": 0.5767, + "step": 14061 + }, + { + "epoch": 0.6226049610927243, + "grad_norm": 1.4338053464889526, + "learning_rate": 7.49400556094067e-06, + "loss": 0.5201, + "step": 14062 + }, + { + "epoch": 0.6226492367975383, + "grad_norm": 1.559958815574646, + "learning_rate": 7.4925094539476725e-06, + "loss": 0.5815, + "step": 14063 + }, + { + "epoch": 0.6226935125023522, + "grad_norm": 1.5249749422073364, + "learning_rate": 7.491013406844048e-06, + "loss": 0.5793, + "step": 14064 + }, + { + "epoch": 0.6227377882071661, + "grad_norm": 1.3425246477127075, + "learning_rate": 7.489517419665524e-06, + "loss": 0.5487, + "step": 14065 + }, + { + "epoch": 0.6227820639119799, + "grad_norm": 1.5657517910003662, + "learning_rate": 7.4880214924478345e-06, + "loss": 0.6169, + "step": 14066 + }, + { + "epoch": 0.6228263396167938, + "grad_norm": 1.393420934677124, + "learning_rate": 7.486525625226699e-06, + "loss": 0.6366, + "step": 14067 + }, + { + "epoch": 0.6228706153216077, + "grad_norm": 1.5185636281967163, + "learning_rate": 7.485029818037861e-06, + "loss": 0.5836, + "step": 14068 + }, + { + "epoch": 0.6229148910264215, + "grad_norm": 1.3585420846939087, + "learning_rate": 7.483534070917036e-06, + "loss": 0.6151, + "step": 14069 + }, + { + "epoch": 0.6229591667312354, + "grad_norm": 1.3470981121063232, + "learning_rate": 7.482038383899946e-06, + "loss": 0.5937, + "step": 14070 + }, + { + "epoch": 0.6230034424360493, + "grad_norm": 1.2262245416641235, + "learning_rate": 7.4805427570223225e-06, + "loss": 0.5177, + "step": 14071 + }, + { + "epoch": 0.6230477181408631, + "grad_norm": 1.566728949546814, + "learning_rate": 7.479047190319885e-06, + "loss": 0.6195, + "step": 14072 + }, + { + "epoch": 0.623091993845677, + "grad_norm": 1.861238956451416, + "learning_rate": 7.47755168382835e-06, + "loss": 0.5902, + "step": 14073 + }, + { + "epoch": 0.623136269550491, + "grad_norm": 1.6538954973220825, + "learning_rate": 7.476056237583441e-06, + "loss": 0.6211, + "step": 14074 + }, + { + "epoch": 0.6231805452553048, + "grad_norm": 1.5881907939910889, + "learning_rate": 7.474560851620873e-06, + "loss": 0.5762, + "step": 14075 + }, + { + "epoch": 0.6232248209601187, + "grad_norm": 1.3080954551696777, + "learning_rate": 7.473065525976365e-06, + "loss": 0.5633, + "step": 14076 + }, + { + "epoch": 0.6232690966649326, + "grad_norm": 1.3722792863845825, + "learning_rate": 7.471570260685626e-06, + "loss": 0.5567, + "step": 14077 + }, + { + "epoch": 0.6233133723697464, + "grad_norm": 1.5170453786849976, + "learning_rate": 7.470075055784375e-06, + "loss": 0.6087, + "step": 14078 + }, + { + "epoch": 0.6233576480745603, + "grad_norm": 2.4897425174713135, + "learning_rate": 7.468579911308321e-06, + "loss": 0.6066, + "step": 14079 + }, + { + "epoch": 0.6234019237793742, + "grad_norm": 1.4845138788223267, + "learning_rate": 7.467084827293171e-06, + "loss": 0.5168, + "step": 14080 + }, + { + "epoch": 0.623446199484188, + "grad_norm": 1.4782028198242188, + "learning_rate": 7.465589803774641e-06, + "loss": 0.5859, + "step": 14081 + }, + { + "epoch": 0.6234904751890019, + "grad_norm": 1.3611912727355957, + "learning_rate": 7.464094840788435e-06, + "loss": 0.5595, + "step": 14082 + }, + { + "epoch": 0.6235347508938158, + "grad_norm": 1.5961359739303589, + "learning_rate": 7.462599938370262e-06, + "loss": 0.5962, + "step": 14083 + }, + { + "epoch": 0.6235790265986296, + "grad_norm": 1.3920421600341797, + "learning_rate": 7.4611050965558165e-06, + "loss": 0.5975, + "step": 14084 + }, + { + "epoch": 0.6236233023034435, + "grad_norm": 1.5533055067062378, + "learning_rate": 7.4596103153808116e-06, + "loss": 0.5473, + "step": 14085 + }, + { + "epoch": 0.6236675780082575, + "grad_norm": 1.5523167848587036, + "learning_rate": 7.458115594880948e-06, + "loss": 0.6027, + "step": 14086 + }, + { + "epoch": 0.6237118537130713, + "grad_norm": 1.756529450416565, + "learning_rate": 7.45662093509192e-06, + "loss": 0.581, + "step": 14087 + }, + { + "epoch": 0.6237561294178852, + "grad_norm": 1.3865967988967896, + "learning_rate": 7.455126336049433e-06, + "loss": 0.5918, + "step": 14088 + }, + { + "epoch": 0.6238004051226991, + "grad_norm": 1.3376915454864502, + "learning_rate": 7.453631797789181e-06, + "loss": 0.577, + "step": 14089 + }, + { + "epoch": 0.6238446808275129, + "grad_norm": 1.5323299169540405, + "learning_rate": 7.452137320346856e-06, + "loss": 0.5535, + "step": 14090 + }, + { + "epoch": 0.6238889565323268, + "grad_norm": 1.4256576299667358, + "learning_rate": 7.450642903758163e-06, + "loss": 0.6001, + "step": 14091 + }, + { + "epoch": 0.6239332322371407, + "grad_norm": 1.279486894607544, + "learning_rate": 7.4491485480587865e-06, + "loss": 0.5763, + "step": 14092 + }, + { + "epoch": 0.6239775079419545, + "grad_norm": 1.3106672763824463, + "learning_rate": 7.44765425328442e-06, + "loss": 0.5842, + "step": 14093 + }, + { + "epoch": 0.6240217836467684, + "grad_norm": 1.405827522277832, + "learning_rate": 7.446160019470749e-06, + "loss": 0.5351, + "step": 14094 + }, + { + "epoch": 0.6240660593515823, + "grad_norm": 1.3903676271438599, + "learning_rate": 7.444665846653471e-06, + "loss": 0.5525, + "step": 14095 + }, + { + "epoch": 0.6241103350563961, + "grad_norm": 1.6226859092712402, + "learning_rate": 7.4431717348682675e-06, + "loss": 0.5759, + "step": 14096 + }, + { + "epoch": 0.62415461076121, + "grad_norm": 1.5090364217758179, + "learning_rate": 7.441677684150825e-06, + "loss": 0.6653, + "step": 14097 + }, + { + "epoch": 0.624198886466024, + "grad_norm": 1.4300628900527954, + "learning_rate": 7.440183694536829e-06, + "loss": 0.5262, + "step": 14098 + }, + { + "epoch": 0.6242431621708379, + "grad_norm": 1.4444506168365479, + "learning_rate": 7.4386897660619604e-06, + "loss": 0.6032, + "step": 14099 + }, + { + "epoch": 0.6242874378756517, + "grad_norm": 1.528990387916565, + "learning_rate": 7.4371958987619e-06, + "loss": 0.5674, + "step": 14100 + }, + { + "epoch": 0.6243317135804656, + "grad_norm": 1.48125422000885, + "learning_rate": 7.435702092672331e-06, + "loss": 0.5662, + "step": 14101 + }, + { + "epoch": 0.6243759892852795, + "grad_norm": 1.3480857610702515, + "learning_rate": 7.434208347828929e-06, + "loss": 0.6313, + "step": 14102 + }, + { + "epoch": 0.6244202649900933, + "grad_norm": 1.332633376121521, + "learning_rate": 7.432714664267373e-06, + "loss": 0.5077, + "step": 14103 + }, + { + "epoch": 0.6244645406949072, + "grad_norm": 1.2553372383117676, + "learning_rate": 7.431221042023331e-06, + "loss": 0.5538, + "step": 14104 + }, + { + "epoch": 0.6245088163997211, + "grad_norm": 1.6696311235427856, + "learning_rate": 7.429727481132487e-06, + "loss": 0.5922, + "step": 14105 + }, + { + "epoch": 0.6245530921045349, + "grad_norm": 1.7767051458358765, + "learning_rate": 7.4282339816305125e-06, + "loss": 0.6443, + "step": 14106 + }, + { + "epoch": 0.6245973678093488, + "grad_norm": 1.6923259496688843, + "learning_rate": 7.426740543553069e-06, + "loss": 0.6406, + "step": 14107 + }, + { + "epoch": 0.6246416435141627, + "grad_norm": 1.4270373582839966, + "learning_rate": 7.425247166935836e-06, + "loss": 0.5616, + "step": 14108 + }, + { + "epoch": 0.6246859192189765, + "grad_norm": 1.7237335443496704, + "learning_rate": 7.423753851814479e-06, + "loss": 0.539, + "step": 14109 + }, + { + "epoch": 0.6247301949237904, + "grad_norm": 1.2142369747161865, + "learning_rate": 7.42226059822466e-06, + "loss": 0.5196, + "step": 14110 + }, + { + "epoch": 0.6247744706286044, + "grad_norm": 1.603846788406372, + "learning_rate": 7.420767406202051e-06, + "loss": 0.5735, + "step": 14111 + }, + { + "epoch": 0.6248187463334182, + "grad_norm": 1.5065186023712158, + "learning_rate": 7.419274275782311e-06, + "loss": 0.5711, + "step": 14112 + }, + { + "epoch": 0.6248630220382321, + "grad_norm": 1.538373589515686, + "learning_rate": 7.417781207001106e-06, + "loss": 0.5832, + "step": 14113 + }, + { + "epoch": 0.624907297743046, + "grad_norm": 1.5688433647155762, + "learning_rate": 7.416288199894089e-06, + "loss": 0.6359, + "step": 14114 + }, + { + "epoch": 0.6249515734478598, + "grad_norm": 1.5461188554763794, + "learning_rate": 7.414795254496929e-06, + "loss": 0.6383, + "step": 14115 + }, + { + "epoch": 0.6249958491526737, + "grad_norm": 1.7100330591201782, + "learning_rate": 7.41330237084528e-06, + "loss": 0.7079, + "step": 14116 + }, + { + "epoch": 0.6250401248574876, + "grad_norm": 1.3253318071365356, + "learning_rate": 7.411809548974792e-06, + "loss": 0.5896, + "step": 14117 + }, + { + "epoch": 0.6250844005623014, + "grad_norm": 1.5014938116073608, + "learning_rate": 7.41031678892113e-06, + "loss": 0.6274, + "step": 14118 + }, + { + "epoch": 0.6251286762671153, + "grad_norm": 1.2620922327041626, + "learning_rate": 7.408824090719944e-06, + "loss": 0.5353, + "step": 14119 + }, + { + "epoch": 0.6251729519719292, + "grad_norm": 1.470517873764038, + "learning_rate": 7.407331454406881e-06, + "loss": 0.5587, + "step": 14120 + }, + { + "epoch": 0.625217227676743, + "grad_norm": 1.3526543378829956, + "learning_rate": 7.405838880017599e-06, + "loss": 0.6246, + "step": 14121 + }, + { + "epoch": 0.625261503381557, + "grad_norm": 1.338597059249878, + "learning_rate": 7.4043463675877415e-06, + "loss": 0.5809, + "step": 14122 + }, + { + "epoch": 0.6253057790863709, + "grad_norm": 1.5045579671859741, + "learning_rate": 7.402853917152959e-06, + "loss": 0.5801, + "step": 14123 + }, + { + "epoch": 0.6253500547911847, + "grad_norm": 1.5043715238571167, + "learning_rate": 7.401361528748892e-06, + "loss": 0.6336, + "step": 14124 + }, + { + "epoch": 0.6253943304959986, + "grad_norm": 1.5553828477859497, + "learning_rate": 7.399869202411193e-06, + "loss": 0.6626, + "step": 14125 + }, + { + "epoch": 0.6254386062008125, + "grad_norm": 1.4130494594573975, + "learning_rate": 7.3983769381755e-06, + "loss": 0.5979, + "step": 14126 + }, + { + "epoch": 0.6254828819056263, + "grad_norm": 1.392865777015686, + "learning_rate": 7.396884736077452e-06, + "loss": 0.5383, + "step": 14127 + }, + { + "epoch": 0.6255271576104402, + "grad_norm": 1.4995282888412476, + "learning_rate": 7.395392596152695e-06, + "loss": 0.6172, + "step": 14128 + }, + { + "epoch": 0.6255714333152541, + "grad_norm": 1.275138020515442, + "learning_rate": 7.3939005184368674e-06, + "loss": 0.5616, + "step": 14129 + }, + { + "epoch": 0.6256157090200679, + "grad_norm": 1.5304832458496094, + "learning_rate": 7.3924085029656e-06, + "loss": 0.6218, + "step": 14130 + }, + { + "epoch": 0.6256599847248818, + "grad_norm": 1.3135240077972412, + "learning_rate": 7.390916549774536e-06, + "loss": 0.5675, + "step": 14131 + }, + { + "epoch": 0.6257042604296957, + "grad_norm": 1.3739694356918335, + "learning_rate": 7.389424658899304e-06, + "loss": 0.5126, + "step": 14132 + }, + { + "epoch": 0.6257485361345095, + "grad_norm": 2.19402813911438, + "learning_rate": 7.3879328303755404e-06, + "loss": 0.7362, + "step": 14133 + }, + { + "epoch": 0.6257928118393234, + "grad_norm": 1.7116785049438477, + "learning_rate": 7.386441064238872e-06, + "loss": 0.5764, + "step": 14134 + }, + { + "epoch": 0.6258370875441374, + "grad_norm": 1.3249101638793945, + "learning_rate": 7.3849493605249314e-06, + "loss": 0.5698, + "step": 14135 + }, + { + "epoch": 0.6258813632489513, + "grad_norm": 1.3284918069839478, + "learning_rate": 7.3834577192693465e-06, + "loss": 0.5589, + "step": 14136 + }, + { + "epoch": 0.6259256389537651, + "grad_norm": 1.5170519351959229, + "learning_rate": 7.381966140507739e-06, + "loss": 0.646, + "step": 14137 + }, + { + "epoch": 0.625969914658579, + "grad_norm": 1.5956146717071533, + "learning_rate": 7.380474624275744e-06, + "loss": 0.622, + "step": 14138 + }, + { + "epoch": 0.6260141903633929, + "grad_norm": 1.4804679155349731, + "learning_rate": 7.378983170608982e-06, + "loss": 0.6183, + "step": 14139 + }, + { + "epoch": 0.6260584660682067, + "grad_norm": 1.33692467212677, + "learning_rate": 7.377491779543064e-06, + "loss": 0.6019, + "step": 14140 + }, + { + "epoch": 0.6261027417730206, + "grad_norm": 1.953468680381775, + "learning_rate": 7.376000451113626e-06, + "loss": 0.5924, + "step": 14141 + }, + { + "epoch": 0.6261470174778345, + "grad_norm": 1.6018489599227905, + "learning_rate": 7.37450918535628e-06, + "loss": 0.5771, + "step": 14142 + }, + { + "epoch": 0.6261912931826483, + "grad_norm": 1.6315943002700806, + "learning_rate": 7.373017982306645e-06, + "loss": 0.6133, + "step": 14143 + }, + { + "epoch": 0.6262355688874622, + "grad_norm": 1.758859395980835, + "learning_rate": 7.371526842000334e-06, + "loss": 0.6542, + "step": 14144 + }, + { + "epoch": 0.6262798445922761, + "grad_norm": 1.3938294649124146, + "learning_rate": 7.370035764472966e-06, + "loss": 0.5706, + "step": 14145 + }, + { + "epoch": 0.62632412029709, + "grad_norm": 1.3030930757522583, + "learning_rate": 7.368544749760152e-06, + "loss": 0.5753, + "step": 14146 + }, + { + "epoch": 0.6263683960019039, + "grad_norm": 1.5887765884399414, + "learning_rate": 7.3670537978975036e-06, + "loss": 0.5483, + "step": 14147 + }, + { + "epoch": 0.6264126717067178, + "grad_norm": 1.617301106452942, + "learning_rate": 7.365562908920632e-06, + "loss": 0.5531, + "step": 14148 + }, + { + "epoch": 0.6264569474115316, + "grad_norm": 1.7741258144378662, + "learning_rate": 7.3640720828651456e-06, + "loss": 0.613, + "step": 14149 + }, + { + "epoch": 0.6265012231163455, + "grad_norm": 1.2847259044647217, + "learning_rate": 7.362581319766647e-06, + "loss": 0.532, + "step": 14150 + }, + { + "epoch": 0.6265454988211594, + "grad_norm": 1.4714280366897583, + "learning_rate": 7.3610906196607514e-06, + "loss": 0.5843, + "step": 14151 + }, + { + "epoch": 0.6265897745259732, + "grad_norm": 1.503427267074585, + "learning_rate": 7.359599982583057e-06, + "loss": 0.5627, + "step": 14152 + }, + { + "epoch": 0.6266340502307871, + "grad_norm": 1.433271884918213, + "learning_rate": 7.358109408569165e-06, + "loss": 0.5842, + "step": 14153 + }, + { + "epoch": 0.626678325935601, + "grad_norm": 1.528367042541504, + "learning_rate": 7.35661889765468e-06, + "loss": 0.669, + "step": 14154 + }, + { + "epoch": 0.6267226016404148, + "grad_norm": 1.5795683860778809, + "learning_rate": 7.3551284498752e-06, + "loss": 0.6699, + "step": 14155 + }, + { + "epoch": 0.6267668773452287, + "grad_norm": 1.6418126821517944, + "learning_rate": 7.353638065266324e-06, + "loss": 0.619, + "step": 14156 + }, + { + "epoch": 0.6268111530500426, + "grad_norm": 1.6505299806594849, + "learning_rate": 7.352147743863647e-06, + "loss": 0.5512, + "step": 14157 + }, + { + "epoch": 0.6268554287548564, + "grad_norm": 1.9223010540008545, + "learning_rate": 7.350657485702766e-06, + "loss": 0.6376, + "step": 14158 + }, + { + "epoch": 0.6268997044596704, + "grad_norm": 1.48261296749115, + "learning_rate": 7.349167290819274e-06, + "loss": 0.5758, + "step": 14159 + }, + { + "epoch": 0.6269439801644843, + "grad_norm": 2.049830436706543, + "learning_rate": 7.34767715924876e-06, + "loss": 0.6438, + "step": 14160 + }, + { + "epoch": 0.6269882558692981, + "grad_norm": 1.557085633277893, + "learning_rate": 7.3461870910268195e-06, + "loss": 0.6003, + "step": 14161 + }, + { + "epoch": 0.627032531574112, + "grad_norm": 1.286691427230835, + "learning_rate": 7.344697086189043e-06, + "loss": 0.5328, + "step": 14162 + }, + { + "epoch": 0.6270768072789259, + "grad_norm": 1.6355323791503906, + "learning_rate": 7.343207144771012e-06, + "loss": 0.6375, + "step": 14163 + }, + { + "epoch": 0.6271210829837397, + "grad_norm": 1.4177254438400269, + "learning_rate": 7.341717266808312e-06, + "loss": 0.5867, + "step": 14164 + }, + { + "epoch": 0.6271653586885536, + "grad_norm": 1.429273009300232, + "learning_rate": 7.3402274523365335e-06, + "loss": 0.6095, + "step": 14165 + }, + { + "epoch": 0.6272096343933675, + "grad_norm": 1.651408076286316, + "learning_rate": 7.338737701391256e-06, + "loss": 0.68, + "step": 14166 + }, + { + "epoch": 0.6272539100981813, + "grad_norm": 1.3902063369750977, + "learning_rate": 7.337248014008062e-06, + "loss": 0.5476, + "step": 14167 + }, + { + "epoch": 0.6272981858029952, + "grad_norm": 1.4613494873046875, + "learning_rate": 7.335758390222532e-06, + "loss": 0.5656, + "step": 14168 + }, + { + "epoch": 0.6273424615078091, + "grad_norm": 1.3753159046173096, + "learning_rate": 7.334268830070244e-06, + "loss": 0.5799, + "step": 14169 + }, + { + "epoch": 0.6273867372126231, + "grad_norm": 1.7994040250778198, + "learning_rate": 7.332779333586768e-06, + "loss": 0.6573, + "step": 14170 + }, + { + "epoch": 0.6274310129174369, + "grad_norm": 1.6656358242034912, + "learning_rate": 7.331289900807694e-06, + "loss": 0.6164, + "step": 14171 + }, + { + "epoch": 0.6274752886222508, + "grad_norm": 1.2356486320495605, + "learning_rate": 7.329800531768585e-06, + "loss": 0.546, + "step": 14172 + }, + { + "epoch": 0.6275195643270647, + "grad_norm": 1.6656278371810913, + "learning_rate": 7.328311226505017e-06, + "loss": 0.5782, + "step": 14173 + }, + { + "epoch": 0.6275638400318785, + "grad_norm": 1.4787825345993042, + "learning_rate": 7.326821985052555e-06, + "loss": 0.5386, + "step": 14174 + }, + { + "epoch": 0.6276081157366924, + "grad_norm": 1.4505255222320557, + "learning_rate": 7.325332807446779e-06, + "loss": 0.6248, + "step": 14175 + }, + { + "epoch": 0.6276523914415063, + "grad_norm": 1.5676878690719604, + "learning_rate": 7.323843693723249e-06, + "loss": 0.5724, + "step": 14176 + }, + { + "epoch": 0.6276966671463201, + "grad_norm": 1.5162473917007446, + "learning_rate": 7.322354643917533e-06, + "loss": 0.6173, + "step": 14177 + }, + { + "epoch": 0.627740942851134, + "grad_norm": 1.2777631282806396, + "learning_rate": 7.320865658065197e-06, + "loss": 0.5594, + "step": 14178 + }, + { + "epoch": 0.6277852185559479, + "grad_norm": 1.5530736446380615, + "learning_rate": 7.319376736201804e-06, + "loss": 0.6097, + "step": 14179 + }, + { + "epoch": 0.6278294942607617, + "grad_norm": 1.3150192499160767, + "learning_rate": 7.317887878362912e-06, + "loss": 0.5328, + "step": 14180 + }, + { + "epoch": 0.6278737699655756, + "grad_norm": 1.639051079750061, + "learning_rate": 7.316399084584087e-06, + "loss": 0.57, + "step": 14181 + }, + { + "epoch": 0.6279180456703896, + "grad_norm": 1.5550074577331543, + "learning_rate": 7.314910354900886e-06, + "loss": 0.6051, + "step": 14182 + }, + { + "epoch": 0.6279623213752034, + "grad_norm": 1.1732714176177979, + "learning_rate": 7.313421689348863e-06, + "loss": 0.5065, + "step": 14183 + }, + { + "epoch": 0.6280065970800173, + "grad_norm": 1.5544906854629517, + "learning_rate": 7.311933087963572e-06, + "loss": 0.5999, + "step": 14184 + }, + { + "epoch": 0.6280508727848312, + "grad_norm": 2.1165101528167725, + "learning_rate": 7.310444550780575e-06, + "loss": 0.6225, + "step": 14185 + }, + { + "epoch": 0.628095148489645, + "grad_norm": 1.4389442205429077, + "learning_rate": 7.308956077835421e-06, + "loss": 0.5526, + "step": 14186 + }, + { + "epoch": 0.6281394241944589, + "grad_norm": 1.3240745067596436, + "learning_rate": 7.307467669163655e-06, + "loss": 0.4999, + "step": 14187 + }, + { + "epoch": 0.6281836998992728, + "grad_norm": 1.4471995830535889, + "learning_rate": 7.305979324800835e-06, + "loss": 0.6158, + "step": 14188 + }, + { + "epoch": 0.6282279756040866, + "grad_norm": 1.2092350721359253, + "learning_rate": 7.3044910447825045e-06, + "loss": 0.5329, + "step": 14189 + }, + { + "epoch": 0.6282722513089005, + "grad_norm": 1.3304678201675415, + "learning_rate": 7.30300282914421e-06, + "loss": 0.5251, + "step": 14190 + }, + { + "epoch": 0.6283165270137144, + "grad_norm": 1.3142999410629272, + "learning_rate": 7.301514677921497e-06, + "loss": 0.5265, + "step": 14191 + }, + { + "epoch": 0.6283608027185282, + "grad_norm": 1.6477128267288208, + "learning_rate": 7.300026591149911e-06, + "loss": 0.6086, + "step": 14192 + }, + { + "epoch": 0.6284050784233421, + "grad_norm": 1.776842713356018, + "learning_rate": 7.29853856886499e-06, + "loss": 0.5202, + "step": 14193 + }, + { + "epoch": 0.6284493541281561, + "grad_norm": 1.2647422552108765, + "learning_rate": 7.297050611102272e-06, + "loss": 0.5653, + "step": 14194 + }, + { + "epoch": 0.6284936298329699, + "grad_norm": 1.335610032081604, + "learning_rate": 7.295562717897306e-06, + "loss": 0.5523, + "step": 14195 + }, + { + "epoch": 0.6285379055377838, + "grad_norm": 1.4787788391113281, + "learning_rate": 7.29407488928562e-06, + "loss": 0.6558, + "step": 14196 + }, + { + "epoch": 0.6285821812425977, + "grad_norm": 1.3294535875320435, + "learning_rate": 7.292587125302748e-06, + "loss": 0.5761, + "step": 14197 + }, + { + "epoch": 0.6286264569474115, + "grad_norm": 2.0569677352905273, + "learning_rate": 7.291099425984231e-06, + "loss": 0.703, + "step": 14198 + }, + { + "epoch": 0.6286707326522254, + "grad_norm": 1.5953757762908936, + "learning_rate": 7.2896117913656e-06, + "loss": 0.5254, + "step": 14199 + }, + { + "epoch": 0.6287150083570393, + "grad_norm": 1.4932224750518799, + "learning_rate": 7.288124221482385e-06, + "loss": 0.6385, + "step": 14200 + }, + { + "epoch": 0.6287592840618531, + "grad_norm": 1.7724770307540894, + "learning_rate": 7.2866367163701115e-06, + "loss": 0.5764, + "step": 14201 + }, + { + "epoch": 0.628803559766667, + "grad_norm": 1.3554285764694214, + "learning_rate": 7.285149276064313e-06, + "loss": 0.5782, + "step": 14202 + }, + { + "epoch": 0.6288478354714809, + "grad_norm": 1.4464956521987915, + "learning_rate": 7.283661900600514e-06, + "loss": 0.5653, + "step": 14203 + }, + { + "epoch": 0.6288921111762947, + "grad_norm": 1.531536340713501, + "learning_rate": 7.282174590014236e-06, + "loss": 0.6191, + "step": 14204 + }, + { + "epoch": 0.6289363868811086, + "grad_norm": 1.419683814048767, + "learning_rate": 7.280687344341006e-06, + "loss": 0.5284, + "step": 14205 + }, + { + "epoch": 0.6289806625859226, + "grad_norm": 1.7227829694747925, + "learning_rate": 7.2792001636163465e-06, + "loss": 0.6507, + "step": 14206 + }, + { + "epoch": 0.6290249382907365, + "grad_norm": 1.4514999389648438, + "learning_rate": 7.277713047875772e-06, + "loss": 0.5762, + "step": 14207 + }, + { + "epoch": 0.6290692139955503, + "grad_norm": 1.5281480550765991, + "learning_rate": 7.276225997154808e-06, + "loss": 0.5896, + "step": 14208 + }, + { + "epoch": 0.6291134897003642, + "grad_norm": 1.188242793083191, + "learning_rate": 7.2747390114889686e-06, + "loss": 0.5239, + "step": 14209 + }, + { + "epoch": 0.6291577654051781, + "grad_norm": 1.4908771514892578, + "learning_rate": 7.273252090913771e-06, + "loss": 0.5977, + "step": 14210 + }, + { + "epoch": 0.6292020411099919, + "grad_norm": 1.5558241605758667, + "learning_rate": 7.27176523546472e-06, + "loss": 0.6147, + "step": 14211 + }, + { + "epoch": 0.6292463168148058, + "grad_norm": 1.3883056640625, + "learning_rate": 7.270278445177339e-06, + "loss": 0.5886, + "step": 14212 + }, + { + "epoch": 0.6292905925196197, + "grad_norm": 1.4124222993850708, + "learning_rate": 7.268791720087136e-06, + "loss": 0.5717, + "step": 14213 + }, + { + "epoch": 0.6293348682244335, + "grad_norm": 1.9214980602264404, + "learning_rate": 7.267305060229617e-06, + "loss": 0.6211, + "step": 14214 + }, + { + "epoch": 0.6293791439292474, + "grad_norm": 1.618125319480896, + "learning_rate": 7.265818465640292e-06, + "loss": 0.6367, + "step": 14215 + }, + { + "epoch": 0.6294234196340613, + "grad_norm": 1.381786823272705, + "learning_rate": 7.264331936354667e-06, + "loss": 0.5554, + "step": 14216 + }, + { + "epoch": 0.6294676953388751, + "grad_norm": 1.4864870309829712, + "learning_rate": 7.262845472408243e-06, + "loss": 0.5018, + "step": 14217 + }, + { + "epoch": 0.6295119710436891, + "grad_norm": 1.4392292499542236, + "learning_rate": 7.261359073836533e-06, + "loss": 0.6101, + "step": 14218 + }, + { + "epoch": 0.629556246748503, + "grad_norm": 1.2943755388259888, + "learning_rate": 7.259872740675028e-06, + "loss": 0.5608, + "step": 14219 + }, + { + "epoch": 0.6296005224533168, + "grad_norm": 1.4980783462524414, + "learning_rate": 7.2583864729592325e-06, + "loss": 0.6298, + "step": 14220 + }, + { + "epoch": 0.6296447981581307, + "grad_norm": 1.5837618112564087, + "learning_rate": 7.256900270724639e-06, + "loss": 0.605, + "step": 14221 + }, + { + "epoch": 0.6296890738629446, + "grad_norm": 1.5966222286224365, + "learning_rate": 7.255414134006753e-06, + "loss": 0.595, + "step": 14222 + }, + { + "epoch": 0.6297333495677584, + "grad_norm": 1.3426443338394165, + "learning_rate": 7.253928062841067e-06, + "loss": 0.6135, + "step": 14223 + }, + { + "epoch": 0.6297776252725723, + "grad_norm": 1.4603960514068604, + "learning_rate": 7.25244205726307e-06, + "loss": 0.5983, + "step": 14224 + }, + { + "epoch": 0.6298219009773862, + "grad_norm": 1.3180352449417114, + "learning_rate": 7.250956117308259e-06, + "loss": 0.47, + "step": 14225 + }, + { + "epoch": 0.6298661766822, + "grad_norm": 1.6393381357192993, + "learning_rate": 7.249470243012125e-06, + "loss": 0.6136, + "step": 14226 + }, + { + "epoch": 0.6299104523870139, + "grad_norm": 1.8574119806289673, + "learning_rate": 7.24798443441015e-06, + "loss": 0.7541, + "step": 14227 + }, + { + "epoch": 0.6299547280918278, + "grad_norm": 1.8201671838760376, + "learning_rate": 7.246498691537829e-06, + "loss": 0.7057, + "step": 14228 + }, + { + "epoch": 0.6299990037966416, + "grad_norm": 1.6351401805877686, + "learning_rate": 7.245013014430645e-06, + "loss": 0.6433, + "step": 14229 + }, + { + "epoch": 0.6300432795014556, + "grad_norm": 1.3952901363372803, + "learning_rate": 7.2435274031240825e-06, + "loss": 0.6341, + "step": 14230 + }, + { + "epoch": 0.6300875552062695, + "grad_norm": 1.3446335792541504, + "learning_rate": 7.242041857653619e-06, + "loss": 0.5783, + "step": 14231 + }, + { + "epoch": 0.6301318309110833, + "grad_norm": 1.461302399635315, + "learning_rate": 7.240556378054744e-06, + "loss": 0.579, + "step": 14232 + }, + { + "epoch": 0.6301761066158972, + "grad_norm": 1.2901182174682617, + "learning_rate": 7.239070964362933e-06, + "loss": 0.5588, + "step": 14233 + }, + { + "epoch": 0.6302203823207111, + "grad_norm": 1.57439386844635, + "learning_rate": 7.237585616613663e-06, + "loss": 0.6473, + "step": 14234 + }, + { + "epoch": 0.6302646580255249, + "grad_norm": 1.798414707183838, + "learning_rate": 7.2361003348424105e-06, + "loss": 0.5608, + "step": 14235 + }, + { + "epoch": 0.6303089337303388, + "grad_norm": 1.2788877487182617, + "learning_rate": 7.234615119084654e-06, + "loss": 0.514, + "step": 14236 + }, + { + "epoch": 0.6303532094351527, + "grad_norm": 1.5832217931747437, + "learning_rate": 7.233129969375861e-06, + "loss": 0.6548, + "step": 14237 + }, + { + "epoch": 0.6303974851399665, + "grad_norm": 1.6149526834487915, + "learning_rate": 7.2316448857515076e-06, + "loss": 0.6343, + "step": 14238 + }, + { + "epoch": 0.6304417608447804, + "grad_norm": 1.9514663219451904, + "learning_rate": 7.230159868247062e-06, + "loss": 0.6577, + "step": 14239 + }, + { + "epoch": 0.6304860365495943, + "grad_norm": 1.6996079683303833, + "learning_rate": 7.228674916897993e-06, + "loss": 0.6312, + "step": 14240 + }, + { + "epoch": 0.6305303122544083, + "grad_norm": 1.5151268243789673, + "learning_rate": 7.227190031739761e-06, + "loss": 0.582, + "step": 14241 + }, + { + "epoch": 0.630574587959222, + "grad_norm": 1.6009000539779663, + "learning_rate": 7.225705212807847e-06, + "loss": 0.5771, + "step": 14242 + }, + { + "epoch": 0.630618863664036, + "grad_norm": 1.3411179780960083, + "learning_rate": 7.224220460137701e-06, + "loss": 0.5482, + "step": 14243 + }, + { + "epoch": 0.6306631393688499, + "grad_norm": 1.3751347064971924, + "learning_rate": 7.222735773764784e-06, + "loss": 0.5322, + "step": 14244 + }, + { + "epoch": 0.6307074150736637, + "grad_norm": 1.5797086954116821, + "learning_rate": 7.221251153724569e-06, + "loss": 0.5525, + "step": 14245 + }, + { + "epoch": 0.6307516907784776, + "grad_norm": 1.4933792352676392, + "learning_rate": 7.219766600052505e-06, + "loss": 0.5466, + "step": 14246 + }, + { + "epoch": 0.6307959664832915, + "grad_norm": 1.4954144954681396, + "learning_rate": 7.2182821127840504e-06, + "loss": 0.6206, + "step": 14247 + }, + { + "epoch": 0.6308402421881053, + "grad_norm": 1.1920933723449707, + "learning_rate": 7.216797691954665e-06, + "loss": 0.4965, + "step": 14248 + }, + { + "epoch": 0.6308845178929192, + "grad_norm": 1.5617713928222656, + "learning_rate": 7.215313337599801e-06, + "loss": 0.5387, + "step": 14249 + }, + { + "epoch": 0.6309287935977331, + "grad_norm": 1.57517409324646, + "learning_rate": 7.21382904975491e-06, + "loss": 0.5674, + "step": 14250 + }, + { + "epoch": 0.6309730693025469, + "grad_norm": 1.5690172910690308, + "learning_rate": 7.212344828455442e-06, + "loss": 0.5965, + "step": 14251 + }, + { + "epoch": 0.6310173450073608, + "grad_norm": 1.2886159420013428, + "learning_rate": 7.210860673736849e-06, + "loss": 0.5502, + "step": 14252 + }, + { + "epoch": 0.6310616207121748, + "grad_norm": 2.040283441543579, + "learning_rate": 7.2093765856345785e-06, + "loss": 0.684, + "step": 14253 + }, + { + "epoch": 0.6311058964169886, + "grad_norm": 1.5665762424468994, + "learning_rate": 7.207892564184071e-06, + "loss": 0.5609, + "step": 14254 + }, + { + "epoch": 0.6311501721218025, + "grad_norm": 1.7917084693908691, + "learning_rate": 7.206408609420782e-06, + "loss": 0.6259, + "step": 14255 + }, + { + "epoch": 0.6311944478266164, + "grad_norm": 1.630314826965332, + "learning_rate": 7.204924721380148e-06, + "loss": 0.6012, + "step": 14256 + }, + { + "epoch": 0.6312387235314302, + "grad_norm": 1.4564465284347534, + "learning_rate": 7.203440900097609e-06, + "loss": 0.5893, + "step": 14257 + }, + { + "epoch": 0.6312829992362441, + "grad_norm": 1.7674596309661865, + "learning_rate": 7.201957145608609e-06, + "loss": 0.6405, + "step": 14258 + }, + { + "epoch": 0.631327274941058, + "grad_norm": 1.6458539962768555, + "learning_rate": 7.200473457948584e-06, + "loss": 0.5937, + "step": 14259 + }, + { + "epoch": 0.6313715506458718, + "grad_norm": 1.5632710456848145, + "learning_rate": 7.198989837152972e-06, + "loss": 0.636, + "step": 14260 + }, + { + "epoch": 0.6314158263506857, + "grad_norm": 1.4204797744750977, + "learning_rate": 7.197506283257205e-06, + "loss": 0.5093, + "step": 14261 + }, + { + "epoch": 0.6314601020554996, + "grad_norm": 1.6205942630767822, + "learning_rate": 7.19602279629672e-06, + "loss": 0.6356, + "step": 14262 + }, + { + "epoch": 0.6315043777603134, + "grad_norm": 1.412465214729309, + "learning_rate": 7.194539376306947e-06, + "loss": 0.5694, + "step": 14263 + }, + { + "epoch": 0.6315486534651273, + "grad_norm": 1.521551489830017, + "learning_rate": 7.1930560233233125e-06, + "loss": 0.5684, + "step": 14264 + }, + { + "epoch": 0.6315929291699413, + "grad_norm": 1.6477328538894653, + "learning_rate": 7.191572737381255e-06, + "loss": 0.5635, + "step": 14265 + }, + { + "epoch": 0.631637204874755, + "grad_norm": 1.8510971069335938, + "learning_rate": 7.1900895185161966e-06, + "loss": 0.6167, + "step": 14266 + }, + { + "epoch": 0.631681480579569, + "grad_norm": 1.3681104183197021, + "learning_rate": 7.1886063667635575e-06, + "loss": 0.5723, + "step": 14267 + }, + { + "epoch": 0.6317257562843829, + "grad_norm": 1.2195602655410767, + "learning_rate": 7.187123282158769e-06, + "loss": 0.5752, + "step": 14268 + }, + { + "epoch": 0.6317700319891967, + "grad_norm": 1.3825780153274536, + "learning_rate": 7.1856402647372505e-06, + "loss": 0.5287, + "step": 14269 + }, + { + "epoch": 0.6318143076940106, + "grad_norm": 1.9845291376113892, + "learning_rate": 7.184157314534423e-06, + "loss": 0.5723, + "step": 14270 + }, + { + "epoch": 0.6318585833988245, + "grad_norm": 1.4206609725952148, + "learning_rate": 7.182674431585703e-06, + "loss": 0.5959, + "step": 14271 + }, + { + "epoch": 0.6319028591036383, + "grad_norm": 2.0509955883026123, + "learning_rate": 7.181191615926512e-06, + "loss": 0.6914, + "step": 14272 + }, + { + "epoch": 0.6319471348084522, + "grad_norm": 1.6519980430603027, + "learning_rate": 7.179708867592266e-06, + "loss": 0.6073, + "step": 14273 + }, + { + "epoch": 0.6319914105132661, + "grad_norm": 1.4684795141220093, + "learning_rate": 7.178226186618375e-06, + "loss": 0.5958, + "step": 14274 + }, + { + "epoch": 0.63203568621808, + "grad_norm": 1.5019958019256592, + "learning_rate": 7.176743573040254e-06, + "loss": 0.5597, + "step": 14275 + }, + { + "epoch": 0.6320799619228938, + "grad_norm": 1.6168593168258667, + "learning_rate": 7.175261026893315e-06, + "loss": 0.6133, + "step": 14276 + }, + { + "epoch": 0.6321242376277078, + "grad_norm": 1.6922476291656494, + "learning_rate": 7.173778548212963e-06, + "loss": 0.6031, + "step": 14277 + }, + { + "epoch": 0.6321685133325217, + "grad_norm": 1.573683261871338, + "learning_rate": 7.172296137034612e-06, + "loss": 0.5938, + "step": 14278 + }, + { + "epoch": 0.6322127890373355, + "grad_norm": 2.2948782444000244, + "learning_rate": 7.170813793393666e-06, + "loss": 0.6084, + "step": 14279 + }, + { + "epoch": 0.6322570647421494, + "grad_norm": 1.4305750131607056, + "learning_rate": 7.169331517325528e-06, + "loss": 0.5502, + "step": 14280 + }, + { + "epoch": 0.6323013404469633, + "grad_norm": 1.3421218395233154, + "learning_rate": 7.167849308865602e-06, + "loss": 0.5463, + "step": 14281 + }, + { + "epoch": 0.6323456161517771, + "grad_norm": 1.4430197477340698, + "learning_rate": 7.1663671680492906e-06, + "loss": 0.5869, + "step": 14282 + }, + { + "epoch": 0.632389891856591, + "grad_norm": 1.522955298423767, + "learning_rate": 7.164885094911991e-06, + "loss": 0.6264, + "step": 14283 + }, + { + "epoch": 0.6324341675614049, + "grad_norm": 1.5998163223266602, + "learning_rate": 7.1634030894891e-06, + "loss": 0.5953, + "step": 14284 + }, + { + "epoch": 0.6324784432662187, + "grad_norm": 1.5406337976455688, + "learning_rate": 7.161921151816021e-06, + "loss": 0.5407, + "step": 14285 + }, + { + "epoch": 0.6325227189710326, + "grad_norm": 1.3959300518035889, + "learning_rate": 7.160439281928142e-06, + "loss": 0.5481, + "step": 14286 + }, + { + "epoch": 0.6325669946758465, + "grad_norm": 1.3000969886779785, + "learning_rate": 7.1589574798608565e-06, + "loss": 0.5849, + "step": 14287 + }, + { + "epoch": 0.6326112703806603, + "grad_norm": 1.5062230825424194, + "learning_rate": 7.157475745649562e-06, + "loss": 0.5947, + "step": 14288 + }, + { + "epoch": 0.6326555460854743, + "grad_norm": 1.3715386390686035, + "learning_rate": 7.1559940793296466e-06, + "loss": 0.5909, + "step": 14289 + }, + { + "epoch": 0.6326998217902882, + "grad_norm": 1.4627498388290405, + "learning_rate": 7.154512480936495e-06, + "loss": 0.585, + "step": 14290 + }, + { + "epoch": 0.632744097495102, + "grad_norm": 2.1609413623809814, + "learning_rate": 7.15303095050549e-06, + "loss": 0.617, + "step": 14291 + }, + { + "epoch": 0.6327883731999159, + "grad_norm": 1.894351601600647, + "learning_rate": 7.151549488072029e-06, + "loss": 0.6575, + "step": 14292 + }, + { + "epoch": 0.6328326489047298, + "grad_norm": 1.700987458229065, + "learning_rate": 7.150068093671489e-06, + "loss": 0.5923, + "step": 14293 + }, + { + "epoch": 0.6328769246095436, + "grad_norm": 1.434325098991394, + "learning_rate": 7.14858676733925e-06, + "loss": 0.555, + "step": 14294 + }, + { + "epoch": 0.6329212003143575, + "grad_norm": 1.4363681077957153, + "learning_rate": 7.1471055091106955e-06, + "loss": 0.5645, + "step": 14295 + }, + { + "epoch": 0.6329654760191714, + "grad_norm": 1.4297106266021729, + "learning_rate": 7.1456243190212026e-06, + "loss": 0.529, + "step": 14296 + }, + { + "epoch": 0.6330097517239852, + "grad_norm": 1.4739718437194824, + "learning_rate": 7.144143197106146e-06, + "loss": 0.6343, + "step": 14297 + }, + { + "epoch": 0.6330540274287991, + "grad_norm": 1.3465111255645752, + "learning_rate": 7.142662143400909e-06, + "loss": 0.5507, + "step": 14298 + }, + { + "epoch": 0.633098303133613, + "grad_norm": 1.48284912109375, + "learning_rate": 7.141181157940859e-06, + "loss": 0.5727, + "step": 14299 + }, + { + "epoch": 0.6331425788384268, + "grad_norm": 1.4625779390335083, + "learning_rate": 7.139700240761369e-06, + "loss": 0.5454, + "step": 14300 + }, + { + "epoch": 0.6331868545432408, + "grad_norm": 1.5093787908554077, + "learning_rate": 7.138219391897806e-06, + "loss": 0.5571, + "step": 14301 + }, + { + "epoch": 0.6332311302480547, + "grad_norm": 1.5219917297363281, + "learning_rate": 7.136738611385547e-06, + "loss": 0.5591, + "step": 14302 + }, + { + "epoch": 0.6332754059528685, + "grad_norm": 1.190652847290039, + "learning_rate": 7.135257899259953e-06, + "loss": 0.5419, + "step": 14303 + }, + { + "epoch": 0.6333196816576824, + "grad_norm": 1.5588366985321045, + "learning_rate": 7.133777255556391e-06, + "loss": 0.563, + "step": 14304 + }, + { + "epoch": 0.6333639573624963, + "grad_norm": 1.609256386756897, + "learning_rate": 7.132296680310227e-06, + "loss": 0.576, + "step": 14305 + }, + { + "epoch": 0.6334082330673101, + "grad_norm": 1.5865634679794312, + "learning_rate": 7.130816173556823e-06, + "loss": 0.5665, + "step": 14306 + }, + { + "epoch": 0.633452508772124, + "grad_norm": 2.215254306793213, + "learning_rate": 7.129335735331537e-06, + "loss": 0.5876, + "step": 14307 + }, + { + "epoch": 0.6334967844769379, + "grad_norm": 1.5941367149353027, + "learning_rate": 7.127855365669729e-06, + "loss": 0.6208, + "step": 14308 + }, + { + "epoch": 0.6335410601817517, + "grad_norm": 1.2603468894958496, + "learning_rate": 7.126375064606758e-06, + "loss": 0.488, + "step": 14309 + }, + { + "epoch": 0.6335853358865656, + "grad_norm": 1.4978471994400024, + "learning_rate": 7.124894832177979e-06, + "loss": 0.5402, + "step": 14310 + }, + { + "epoch": 0.6336296115913795, + "grad_norm": 1.756944179534912, + "learning_rate": 7.123414668418742e-06, + "loss": 0.6853, + "step": 14311 + }, + { + "epoch": 0.6336738872961934, + "grad_norm": 1.5747804641723633, + "learning_rate": 7.121934573364407e-06, + "loss": 0.5825, + "step": 14312 + }, + { + "epoch": 0.6337181630010073, + "grad_norm": 1.2591204643249512, + "learning_rate": 7.120454547050324e-06, + "loss": 0.5217, + "step": 14313 + }, + { + "epoch": 0.6337624387058212, + "grad_norm": 1.415848731994629, + "learning_rate": 7.118974589511833e-06, + "loss": 0.5681, + "step": 14314 + }, + { + "epoch": 0.6338067144106351, + "grad_norm": 1.4183669090270996, + "learning_rate": 7.117494700784292e-06, + "loss": 0.5385, + "step": 14315 + }, + { + "epoch": 0.6338509901154489, + "grad_norm": 1.5130786895751953, + "learning_rate": 7.116014880903044e-06, + "loss": 0.5803, + "step": 14316 + }, + { + "epoch": 0.6338952658202628, + "grad_norm": 1.2466241121292114, + "learning_rate": 7.114535129903432e-06, + "loss": 0.5288, + "step": 14317 + }, + { + "epoch": 0.6339395415250767, + "grad_norm": 1.1954772472381592, + "learning_rate": 7.113055447820796e-06, + "loss": 0.572, + "step": 14318 + }, + { + "epoch": 0.6339838172298905, + "grad_norm": 1.4195704460144043, + "learning_rate": 7.111575834690483e-06, + "loss": 0.5561, + "step": 14319 + }, + { + "epoch": 0.6340280929347044, + "grad_norm": 1.5642777681350708, + "learning_rate": 7.1100962905478285e-06, + "loss": 0.606, + "step": 14320 + }, + { + "epoch": 0.6340723686395183, + "grad_norm": 1.6890536546707153, + "learning_rate": 7.108616815428167e-06, + "loss": 0.6064, + "step": 14321 + }, + { + "epoch": 0.6341166443443321, + "grad_norm": 1.7081085443496704, + "learning_rate": 7.107137409366846e-06, + "loss": 0.613, + "step": 14322 + }, + { + "epoch": 0.634160920049146, + "grad_norm": 1.6643015146255493, + "learning_rate": 7.10565807239919e-06, + "loss": 0.6644, + "step": 14323 + }, + { + "epoch": 0.63420519575396, + "grad_norm": 1.2969894409179688, + "learning_rate": 7.1041788045605294e-06, + "loss": 0.5343, + "step": 14324 + }, + { + "epoch": 0.6342494714587738, + "grad_norm": 1.3579556941986084, + "learning_rate": 7.102699605886206e-06, + "loss": 0.5873, + "step": 14325 + }, + { + "epoch": 0.6342937471635877, + "grad_norm": 1.3395134210586548, + "learning_rate": 7.101220476411541e-06, + "loss": 0.5579, + "step": 14326 + }, + { + "epoch": 0.6343380228684016, + "grad_norm": 1.453991174697876, + "learning_rate": 7.099741416171866e-06, + "loss": 0.5921, + "step": 14327 + }, + { + "epoch": 0.6343822985732154, + "grad_norm": 1.407620906829834, + "learning_rate": 7.098262425202504e-06, + "loss": 0.5315, + "step": 14328 + }, + { + "epoch": 0.6344265742780293, + "grad_norm": 2.145089626312256, + "learning_rate": 7.096783503538782e-06, + "loss": 0.7512, + "step": 14329 + }, + { + "epoch": 0.6344708499828432, + "grad_norm": 1.4975963830947876, + "learning_rate": 7.0953046512160236e-06, + "loss": 0.6006, + "step": 14330 + }, + { + "epoch": 0.634515125687657, + "grad_norm": 1.681778073310852, + "learning_rate": 7.093825868269546e-06, + "loss": 0.6155, + "step": 14331 + }, + { + "epoch": 0.6345594013924709, + "grad_norm": 1.6091580390930176, + "learning_rate": 7.092347154734674e-06, + "loss": 0.6016, + "step": 14332 + }, + { + "epoch": 0.6346036770972848, + "grad_norm": 1.9383554458618164, + "learning_rate": 7.090868510646722e-06, + "loss": 0.6756, + "step": 14333 + }, + { + "epoch": 0.6346479528020986, + "grad_norm": 1.4756219387054443, + "learning_rate": 7.0893899360410025e-06, + "loss": 0.6038, + "step": 14334 + }, + { + "epoch": 0.6346922285069125, + "grad_norm": 1.6179550886154175, + "learning_rate": 7.087911430952839e-06, + "loss": 0.6096, + "step": 14335 + }, + { + "epoch": 0.6347365042117264, + "grad_norm": 1.3856887817382812, + "learning_rate": 7.086432995417539e-06, + "loss": 0.5288, + "step": 14336 + }, + { + "epoch": 0.6347807799165402, + "grad_norm": 1.3181949853897095, + "learning_rate": 7.084954629470417e-06, + "loss": 0.5723, + "step": 14337 + }, + { + "epoch": 0.6348250556213542, + "grad_norm": 1.5464808940887451, + "learning_rate": 7.083476333146775e-06, + "loss": 0.5723, + "step": 14338 + }, + { + "epoch": 0.6348693313261681, + "grad_norm": 1.391822099685669, + "learning_rate": 7.08199810648193e-06, + "loss": 0.5916, + "step": 14339 + }, + { + "epoch": 0.6349136070309819, + "grad_norm": 1.689384937286377, + "learning_rate": 7.080519949511184e-06, + "loss": 0.5671, + "step": 14340 + }, + { + "epoch": 0.6349578827357958, + "grad_norm": 1.8075573444366455, + "learning_rate": 7.07904186226984e-06, + "loss": 0.6118, + "step": 14341 + }, + { + "epoch": 0.6350021584406097, + "grad_norm": 1.721794605255127, + "learning_rate": 7.0775638447932035e-06, + "loss": 0.5952, + "step": 14342 + }, + { + "epoch": 0.6350464341454235, + "grad_norm": 1.4043077230453491, + "learning_rate": 7.076085897116576e-06, + "loss": 0.534, + "step": 14343 + }, + { + "epoch": 0.6350907098502374, + "grad_norm": 1.5759118795394897, + "learning_rate": 7.074608019275253e-06, + "loss": 0.5517, + "step": 14344 + }, + { + "epoch": 0.6351349855550513, + "grad_norm": 1.372104287147522, + "learning_rate": 7.073130211304539e-06, + "loss": 0.5474, + "step": 14345 + }, + { + "epoch": 0.6351792612598652, + "grad_norm": 1.6061903238296509, + "learning_rate": 7.0716524732397275e-06, + "loss": 0.6605, + "step": 14346 + }, + { + "epoch": 0.635223536964679, + "grad_norm": 1.2809408903121948, + "learning_rate": 7.070174805116111e-06, + "loss": 0.5602, + "step": 14347 + }, + { + "epoch": 0.635267812669493, + "grad_norm": 1.3643101453781128, + "learning_rate": 7.0686972069689794e-06, + "loss": 0.5547, + "step": 14348 + }, + { + "epoch": 0.6353120883743069, + "grad_norm": 1.6489460468292236, + "learning_rate": 7.067219678833633e-06, + "loss": 0.5908, + "step": 14349 + }, + { + "epoch": 0.6353563640791207, + "grad_norm": 1.3752429485321045, + "learning_rate": 7.065742220745357e-06, + "loss": 0.59, + "step": 14350 + }, + { + "epoch": 0.6354006397839346, + "grad_norm": 1.6100635528564453, + "learning_rate": 7.064264832739436e-06, + "loss": 0.5835, + "step": 14351 + }, + { + "epoch": 0.6354449154887485, + "grad_norm": 1.7550588846206665, + "learning_rate": 7.062787514851161e-06, + "loss": 0.5045, + "step": 14352 + }, + { + "epoch": 0.6354891911935623, + "grad_norm": 1.862053394317627, + "learning_rate": 7.061310267115815e-06, + "loss": 0.6045, + "step": 14353 + }, + { + "epoch": 0.6355334668983762, + "grad_norm": 1.484343409538269, + "learning_rate": 7.05983308956868e-06, + "loss": 0.6168, + "step": 14354 + }, + { + "epoch": 0.6355777426031901, + "grad_norm": 1.2807220220565796, + "learning_rate": 7.058355982245038e-06, + "loss": 0.5245, + "step": 14355 + }, + { + "epoch": 0.6356220183080039, + "grad_norm": 1.506859302520752, + "learning_rate": 7.056878945180168e-06, + "loss": 0.5665, + "step": 14356 + }, + { + "epoch": 0.6356662940128178, + "grad_norm": 1.426611304283142, + "learning_rate": 7.055401978409349e-06, + "loss": 0.5658, + "step": 14357 + }, + { + "epoch": 0.6357105697176317, + "grad_norm": 1.5117539167404175, + "learning_rate": 7.053925081967851e-06, + "loss": 0.6364, + "step": 14358 + }, + { + "epoch": 0.6357548454224455, + "grad_norm": 1.2973992824554443, + "learning_rate": 7.052448255890958e-06, + "loss": 0.4645, + "step": 14359 + }, + { + "epoch": 0.6357991211272594, + "grad_norm": 1.5320369005203247, + "learning_rate": 7.050971500213939e-06, + "loss": 0.5588, + "step": 14360 + }, + { + "epoch": 0.6358433968320734, + "grad_norm": 1.4301061630249023, + "learning_rate": 7.049494814972062e-06, + "loss": 0.6103, + "step": 14361 + }, + { + "epoch": 0.6358876725368872, + "grad_norm": 1.528235912322998, + "learning_rate": 7.048018200200601e-06, + "loss": 0.6137, + "step": 14362 + }, + { + "epoch": 0.6359319482417011, + "grad_norm": 1.4752156734466553, + "learning_rate": 7.046541655934822e-06, + "loss": 0.5652, + "step": 14363 + }, + { + "epoch": 0.635976223946515, + "grad_norm": 1.6100119352340698, + "learning_rate": 7.0450651822099894e-06, + "loss": 0.5904, + "step": 14364 + }, + { + "epoch": 0.6360204996513288, + "grad_norm": 1.3774073123931885, + "learning_rate": 7.04358877906137e-06, + "loss": 0.5714, + "step": 14365 + }, + { + "epoch": 0.6360647753561427, + "grad_norm": 2.0250742435455322, + "learning_rate": 7.042112446524225e-06, + "loss": 0.5481, + "step": 14366 + }, + { + "epoch": 0.6361090510609566, + "grad_norm": 1.3276647329330444, + "learning_rate": 7.040636184633817e-06, + "loss": 0.5596, + "step": 14367 + }, + { + "epoch": 0.6361533267657704, + "grad_norm": 1.4750794172286987, + "learning_rate": 7.0391599934254e-06, + "loss": 0.557, + "step": 14368 + }, + { + "epoch": 0.6361976024705843, + "grad_norm": 1.5406107902526855, + "learning_rate": 7.03768387293424e-06, + "loss": 0.5399, + "step": 14369 + }, + { + "epoch": 0.6362418781753982, + "grad_norm": 1.3952741622924805, + "learning_rate": 7.036207823195589e-06, + "loss": 0.5233, + "step": 14370 + }, + { + "epoch": 0.636286153880212, + "grad_norm": 1.220847725868225, + "learning_rate": 7.034731844244696e-06, + "loss": 0.5469, + "step": 14371 + }, + { + "epoch": 0.636330429585026, + "grad_norm": 1.3310449123382568, + "learning_rate": 7.0332559361168206e-06, + "loss": 0.522, + "step": 14372 + }, + { + "epoch": 0.6363747052898399, + "grad_norm": 1.5304503440856934, + "learning_rate": 7.031780098847214e-06, + "loss": 0.5775, + "step": 14373 + }, + { + "epoch": 0.6364189809946537, + "grad_norm": 1.359144687652588, + "learning_rate": 7.030304332471121e-06, + "loss": 0.5873, + "step": 14374 + }, + { + "epoch": 0.6364632566994676, + "grad_norm": 1.5402374267578125, + "learning_rate": 7.028828637023792e-06, + "loss": 0.5699, + "step": 14375 + }, + { + "epoch": 0.6365075324042815, + "grad_norm": 1.6008297204971313, + "learning_rate": 7.027353012540472e-06, + "loss": 0.608, + "step": 14376 + }, + { + "epoch": 0.6365518081090953, + "grad_norm": 1.7061582803726196, + "learning_rate": 7.025877459056405e-06, + "loss": 0.6557, + "step": 14377 + }, + { + "epoch": 0.6365960838139092, + "grad_norm": 1.5467337369918823, + "learning_rate": 7.024401976606831e-06, + "loss": 0.5733, + "step": 14378 + }, + { + "epoch": 0.6366403595187231, + "grad_norm": 1.4598335027694702, + "learning_rate": 7.022926565226995e-06, + "loss": 0.5399, + "step": 14379 + }, + { + "epoch": 0.6366846352235369, + "grad_norm": 1.271821141242981, + "learning_rate": 7.021451224952134e-06, + "loss": 0.5472, + "step": 14380 + }, + { + "epoch": 0.6367289109283508, + "grad_norm": 1.6312766075134277, + "learning_rate": 7.0199759558174815e-06, + "loss": 0.5289, + "step": 14381 + }, + { + "epoch": 0.6367731866331647, + "grad_norm": 1.5958975553512573, + "learning_rate": 7.01850075785828e-06, + "loss": 0.607, + "step": 14382 + }, + { + "epoch": 0.6368174623379786, + "grad_norm": 1.3803871870040894, + "learning_rate": 7.017025631109762e-06, + "loss": 0.6073, + "step": 14383 + }, + { + "epoch": 0.6368617380427924, + "grad_norm": 2.584747314453125, + "learning_rate": 7.015550575607155e-06, + "loss": 0.5869, + "step": 14384 + }, + { + "epoch": 0.6369060137476064, + "grad_norm": 1.7026578187942505, + "learning_rate": 7.014075591385694e-06, + "loss": 0.6563, + "step": 14385 + }, + { + "epoch": 0.6369502894524203, + "grad_norm": 1.4371026754379272, + "learning_rate": 7.012600678480607e-06, + "loss": 0.6027, + "step": 14386 + }, + { + "epoch": 0.6369945651572341, + "grad_norm": 1.4487696886062622, + "learning_rate": 7.01112583692712e-06, + "loss": 0.586, + "step": 14387 + }, + { + "epoch": 0.637038840862048, + "grad_norm": 2.564138412475586, + "learning_rate": 7.009651066760457e-06, + "loss": 0.6576, + "step": 14388 + }, + { + "epoch": 0.6370831165668619, + "grad_norm": 1.737318992614746, + "learning_rate": 7.0081763680158445e-06, + "loss": 0.5712, + "step": 14389 + }, + { + "epoch": 0.6371273922716757, + "grad_norm": 1.4739058017730713, + "learning_rate": 7.006701740728505e-06, + "loss": 0.6462, + "step": 14390 + }, + { + "epoch": 0.6371716679764896, + "grad_norm": 1.5494529008865356, + "learning_rate": 7.005227184933652e-06, + "loss": 0.6192, + "step": 14391 + }, + { + "epoch": 0.6372159436813035, + "grad_norm": 1.4375274181365967, + "learning_rate": 7.003752700666513e-06, + "loss": 0.5785, + "step": 14392 + }, + { + "epoch": 0.6372602193861173, + "grad_norm": 1.6230684518814087, + "learning_rate": 7.002278287962305e-06, + "loss": 0.6532, + "step": 14393 + }, + { + "epoch": 0.6373044950909312, + "grad_norm": 1.7099690437316895, + "learning_rate": 7.000803946856231e-06, + "loss": 0.5858, + "step": 14394 + }, + { + "epoch": 0.6373487707957451, + "grad_norm": 1.4100440740585327, + "learning_rate": 6.999329677383517e-06, + "loss": 0.5957, + "step": 14395 + }, + { + "epoch": 0.637393046500559, + "grad_norm": 1.2750933170318604, + "learning_rate": 6.997855479579372e-06, + "loss": 0.5272, + "step": 14396 + }, + { + "epoch": 0.6374373222053729, + "grad_norm": 1.756731390953064, + "learning_rate": 6.996381353479004e-06, + "loss": 0.6541, + "step": 14397 + }, + { + "epoch": 0.6374815979101868, + "grad_norm": 1.432393193244934, + "learning_rate": 6.994907299117619e-06, + "loss": 0.4986, + "step": 14398 + }, + { + "epoch": 0.6375258736150006, + "grad_norm": 1.328538179397583, + "learning_rate": 6.993433316530429e-06, + "loss": 0.5695, + "step": 14399 + }, + { + "epoch": 0.6375701493198145, + "grad_norm": 1.45182466506958, + "learning_rate": 6.9919594057526355e-06, + "loss": 0.5299, + "step": 14400 + }, + { + "epoch": 0.6376144250246284, + "grad_norm": 1.3166735172271729, + "learning_rate": 6.990485566819441e-06, + "loss": 0.4842, + "step": 14401 + }, + { + "epoch": 0.6376587007294422, + "grad_norm": 1.802911400794983, + "learning_rate": 6.98901179976605e-06, + "loss": 0.5948, + "step": 14402 + }, + { + "epoch": 0.6377029764342561, + "grad_norm": 1.5386319160461426, + "learning_rate": 6.9875381046276605e-06, + "loss": 0.561, + "step": 14403 + }, + { + "epoch": 0.63774725213907, + "grad_norm": 1.3369613885879517, + "learning_rate": 6.986064481439465e-06, + "loss": 0.5812, + "step": 14404 + }, + { + "epoch": 0.6377915278438838, + "grad_norm": 1.7376302480697632, + "learning_rate": 6.984590930236672e-06, + "loss": 0.5166, + "step": 14405 + }, + { + "epoch": 0.6378358035486977, + "grad_norm": 1.5797780752182007, + "learning_rate": 6.9831174510544684e-06, + "loss": 0.6298, + "step": 14406 + }, + { + "epoch": 0.6378800792535116, + "grad_norm": 1.4882636070251465, + "learning_rate": 6.981644043928047e-06, + "loss": 0.5758, + "step": 14407 + }, + { + "epoch": 0.6379243549583254, + "grad_norm": 1.3820979595184326, + "learning_rate": 6.9801707088926e-06, + "loss": 0.5266, + "step": 14408 + }, + { + "epoch": 0.6379686306631394, + "grad_norm": 1.702542781829834, + "learning_rate": 6.978697445983317e-06, + "loss": 0.5596, + "step": 14409 + }, + { + "epoch": 0.6380129063679533, + "grad_norm": 1.5394067764282227, + "learning_rate": 6.9772242552353865e-06, + "loss": 0.6182, + "step": 14410 + }, + { + "epoch": 0.6380571820727671, + "grad_norm": 1.5525683164596558, + "learning_rate": 6.97575113668399e-06, + "loss": 0.6031, + "step": 14411 + }, + { + "epoch": 0.638101457777581, + "grad_norm": 1.5586175918579102, + "learning_rate": 6.974278090364319e-06, + "loss": 0.6522, + "step": 14412 + }, + { + "epoch": 0.6381457334823949, + "grad_norm": 1.3306769132614136, + "learning_rate": 6.972805116311552e-06, + "loss": 0.5661, + "step": 14413 + }, + { + "epoch": 0.6381900091872087, + "grad_norm": 1.39335298538208, + "learning_rate": 6.971332214560869e-06, + "loss": 0.5953, + "step": 14414 + }, + { + "epoch": 0.6382342848920226, + "grad_norm": 2.0309672355651855, + "learning_rate": 6.969859385147446e-06, + "loss": 0.5723, + "step": 14415 + }, + { + "epoch": 0.6382785605968365, + "grad_norm": 1.4812983274459839, + "learning_rate": 6.968386628106471e-06, + "loss": 0.5205, + "step": 14416 + }, + { + "epoch": 0.6383228363016504, + "grad_norm": 1.4172496795654297, + "learning_rate": 6.966913943473111e-06, + "loss": 0.562, + "step": 14417 + }, + { + "epoch": 0.6383671120064642, + "grad_norm": 1.7841135263442993, + "learning_rate": 6.965441331282537e-06, + "loss": 0.6181, + "step": 14418 + }, + { + "epoch": 0.6384113877112781, + "grad_norm": 1.3207110166549683, + "learning_rate": 6.963968791569931e-06, + "loss": 0.5416, + "step": 14419 + }, + { + "epoch": 0.6384556634160921, + "grad_norm": 1.6612058877944946, + "learning_rate": 6.962496324370458e-06, + "loss": 0.6454, + "step": 14420 + }, + { + "epoch": 0.6384999391209059, + "grad_norm": 1.3309221267700195, + "learning_rate": 6.961023929719284e-06, + "loss": 0.5361, + "step": 14421 + }, + { + "epoch": 0.6385442148257198, + "grad_norm": 1.312882423400879, + "learning_rate": 6.959551607651582e-06, + "loss": 0.5566, + "step": 14422 + }, + { + "epoch": 0.6385884905305337, + "grad_norm": 1.6064802408218384, + "learning_rate": 6.958079358202514e-06, + "loss": 0.5931, + "step": 14423 + }, + { + "epoch": 0.6386327662353475, + "grad_norm": 1.659730076789856, + "learning_rate": 6.956607181407244e-06, + "loss": 0.6232, + "step": 14424 + }, + { + "epoch": 0.6386770419401614, + "grad_norm": 1.3787813186645508, + "learning_rate": 6.955135077300932e-06, + "loss": 0.5466, + "step": 14425 + }, + { + "epoch": 0.6387213176449753, + "grad_norm": 1.4111331701278687, + "learning_rate": 6.953663045918741e-06, + "loss": 0.6068, + "step": 14426 + }, + { + "epoch": 0.6387655933497891, + "grad_norm": 1.3648061752319336, + "learning_rate": 6.952191087295827e-06, + "loss": 0.5006, + "step": 14427 + }, + { + "epoch": 0.638809869054603, + "grad_norm": 1.5364350080490112, + "learning_rate": 6.950719201467343e-06, + "loss": 0.5511, + "step": 14428 + }, + { + "epoch": 0.6388541447594169, + "grad_norm": 1.4309470653533936, + "learning_rate": 6.949247388468453e-06, + "loss": 0.621, + "step": 14429 + }, + { + "epoch": 0.6388984204642307, + "grad_norm": 1.5046708583831787, + "learning_rate": 6.947775648334305e-06, + "loss": 0.5678, + "step": 14430 + }, + { + "epoch": 0.6389426961690446, + "grad_norm": 1.7091783285140991, + "learning_rate": 6.946303981100049e-06, + "loss": 0.5991, + "step": 14431 + }, + { + "epoch": 0.6389869718738586, + "grad_norm": 1.5192636251449585, + "learning_rate": 6.944832386800837e-06, + "loss": 0.5978, + "step": 14432 + }, + { + "epoch": 0.6390312475786724, + "grad_norm": 1.5150096416473389, + "learning_rate": 6.943360865471815e-06, + "loss": 0.5906, + "step": 14433 + }, + { + "epoch": 0.6390755232834863, + "grad_norm": 1.6341742277145386, + "learning_rate": 6.941889417148132e-06, + "loss": 0.5564, + "step": 14434 + }, + { + "epoch": 0.6391197989883002, + "grad_norm": 1.450478434562683, + "learning_rate": 6.940418041864927e-06, + "loss": 0.5956, + "step": 14435 + }, + { + "epoch": 0.639164074693114, + "grad_norm": 1.4250844717025757, + "learning_rate": 6.938946739657348e-06, + "loss": 0.6359, + "step": 14436 + }, + { + "epoch": 0.6392083503979279, + "grad_norm": 1.2956852912902832, + "learning_rate": 6.937475510560534e-06, + "loss": 0.5713, + "step": 14437 + }, + { + "epoch": 0.6392526261027418, + "grad_norm": 1.7488784790039062, + "learning_rate": 6.93600435460962e-06, + "loss": 0.6208, + "step": 14438 + }, + { + "epoch": 0.6392969018075556, + "grad_norm": 1.922297477722168, + "learning_rate": 6.934533271839751e-06, + "loss": 0.5887, + "step": 14439 + }, + { + "epoch": 0.6393411775123695, + "grad_norm": 1.4710655212402344, + "learning_rate": 6.933062262286061e-06, + "loss": 0.606, + "step": 14440 + }, + { + "epoch": 0.6393854532171834, + "grad_norm": 1.458121657371521, + "learning_rate": 6.931591325983675e-06, + "loss": 0.5885, + "step": 14441 + }, + { + "epoch": 0.6394297289219972, + "grad_norm": 1.5064549446105957, + "learning_rate": 6.9301204629677375e-06, + "loss": 0.5694, + "step": 14442 + }, + { + "epoch": 0.6394740046268111, + "grad_norm": 1.5210394859313965, + "learning_rate": 6.928649673273371e-06, + "loss": 0.6197, + "step": 14443 + }, + { + "epoch": 0.639518280331625, + "grad_norm": 1.3165658712387085, + "learning_rate": 6.927178956935707e-06, + "loss": 0.6108, + "step": 14444 + }, + { + "epoch": 0.6395625560364389, + "grad_norm": 1.2852879762649536, + "learning_rate": 6.925708313989869e-06, + "loss": 0.5613, + "step": 14445 + }, + { + "epoch": 0.6396068317412528, + "grad_norm": 1.4602556228637695, + "learning_rate": 6.9242377444709895e-06, + "loss": 0.5399, + "step": 14446 + }, + { + "epoch": 0.6396511074460667, + "grad_norm": 1.4834855794906616, + "learning_rate": 6.922767248414184e-06, + "loss": 0.5852, + "step": 14447 + }, + { + "epoch": 0.6396953831508805, + "grad_norm": 1.2206485271453857, + "learning_rate": 6.921296825854576e-06, + "loss": 0.5506, + "step": 14448 + }, + { + "epoch": 0.6397396588556944, + "grad_norm": 1.6938046216964722, + "learning_rate": 6.91982647682729e-06, + "loss": 0.6377, + "step": 14449 + }, + { + "epoch": 0.6397839345605083, + "grad_norm": 1.7019717693328857, + "learning_rate": 6.91835620136744e-06, + "loss": 0.594, + "step": 14450 + }, + { + "epoch": 0.6398282102653221, + "grad_norm": 1.8815168142318726, + "learning_rate": 6.916885999510137e-06, + "loss": 0.6122, + "step": 14451 + }, + { + "epoch": 0.639872485970136, + "grad_norm": 1.4413186311721802, + "learning_rate": 6.915415871290507e-06, + "loss": 0.5649, + "step": 14452 + }, + { + "epoch": 0.6399167616749499, + "grad_norm": 1.9638601541519165, + "learning_rate": 6.913945816743658e-06, + "loss": 0.5827, + "step": 14453 + }, + { + "epoch": 0.6399610373797638, + "grad_norm": 1.334263563156128, + "learning_rate": 6.9124758359047e-06, + "loss": 0.5703, + "step": 14454 + }, + { + "epoch": 0.6400053130845776, + "grad_norm": 1.4447197914123535, + "learning_rate": 6.91100592880874e-06, + "loss": 0.6159, + "step": 14455 + }, + { + "epoch": 0.6400495887893916, + "grad_norm": 1.3387748003005981, + "learning_rate": 6.909536095490891e-06, + "loss": 0.522, + "step": 14456 + }, + { + "epoch": 0.6400938644942055, + "grad_norm": 1.4885410070419312, + "learning_rate": 6.908066335986255e-06, + "loss": 0.5458, + "step": 14457 + }, + { + "epoch": 0.6401381401990193, + "grad_norm": 1.451887845993042, + "learning_rate": 6.906596650329937e-06, + "loss": 0.6337, + "step": 14458 + }, + { + "epoch": 0.6401824159038332, + "grad_norm": 1.3359110355377197, + "learning_rate": 6.90512703855704e-06, + "loss": 0.5918, + "step": 14459 + }, + { + "epoch": 0.6402266916086471, + "grad_norm": 1.516592025756836, + "learning_rate": 6.903657500702663e-06, + "loss": 0.576, + "step": 14460 + }, + { + "epoch": 0.6402709673134609, + "grad_norm": 1.399263858795166, + "learning_rate": 6.902188036801902e-06, + "loss": 0.5868, + "step": 14461 + }, + { + "epoch": 0.6403152430182748, + "grad_norm": 1.4182132482528687, + "learning_rate": 6.900718646889861e-06, + "loss": 0.5334, + "step": 14462 + }, + { + "epoch": 0.6403595187230887, + "grad_norm": 1.8396952152252197, + "learning_rate": 6.8992493310016325e-06, + "loss": 0.6727, + "step": 14463 + }, + { + "epoch": 0.6404037944279025, + "grad_norm": 1.5435049533843994, + "learning_rate": 6.897780089172311e-06, + "loss": 0.5801, + "step": 14464 + }, + { + "epoch": 0.6404480701327164, + "grad_norm": 1.2577468156814575, + "learning_rate": 6.89631092143698e-06, + "loss": 0.5147, + "step": 14465 + }, + { + "epoch": 0.6404923458375303, + "grad_norm": 1.6692293882369995, + "learning_rate": 6.894841827830739e-06, + "loss": 0.6169, + "step": 14466 + }, + { + "epoch": 0.6405366215423441, + "grad_norm": 1.6234126091003418, + "learning_rate": 6.893372808388674e-06, + "loss": 0.5648, + "step": 14467 + }, + { + "epoch": 0.640580897247158, + "grad_norm": 1.4912391901016235, + "learning_rate": 6.891903863145866e-06, + "loss": 0.5423, + "step": 14468 + }, + { + "epoch": 0.640625172951972, + "grad_norm": 1.3242695331573486, + "learning_rate": 6.8904349921374065e-06, + "loss": 0.5259, + "step": 14469 + }, + { + "epoch": 0.6406694486567858, + "grad_norm": 1.3429702520370483, + "learning_rate": 6.8889661953983764e-06, + "loss": 0.567, + "step": 14470 + }, + { + "epoch": 0.6407137243615997, + "grad_norm": 1.8029048442840576, + "learning_rate": 6.88749747296385e-06, + "loss": 0.5727, + "step": 14471 + }, + { + "epoch": 0.6407580000664136, + "grad_norm": 1.3026286363601685, + "learning_rate": 6.886028824868919e-06, + "loss": 0.5352, + "step": 14472 + }, + { + "epoch": 0.6408022757712274, + "grad_norm": 1.7032195329666138, + "learning_rate": 6.884560251148651e-06, + "loss": 0.5782, + "step": 14473 + }, + { + "epoch": 0.6408465514760413, + "grad_norm": 1.4606188535690308, + "learning_rate": 6.883091751838126e-06, + "loss": 0.5097, + "step": 14474 + }, + { + "epoch": 0.6408908271808552, + "grad_norm": 1.3880255222320557, + "learning_rate": 6.881623326972412e-06, + "loss": 0.531, + "step": 14475 + }, + { + "epoch": 0.640935102885669, + "grad_norm": 1.420458197593689, + "learning_rate": 6.88015497658659e-06, + "loss": 0.6248, + "step": 14476 + }, + { + "epoch": 0.6409793785904829, + "grad_norm": 2.253793478012085, + "learning_rate": 6.878686700715725e-06, + "loss": 0.6712, + "step": 14477 + }, + { + "epoch": 0.6410236542952968, + "grad_norm": 1.574403166770935, + "learning_rate": 6.877218499394886e-06, + "loss": 0.5777, + "step": 14478 + }, + { + "epoch": 0.6410679300001106, + "grad_norm": 1.6409000158309937, + "learning_rate": 6.875750372659143e-06, + "loss": 0.6005, + "step": 14479 + }, + { + "epoch": 0.6411122057049246, + "grad_norm": 1.833975076675415, + "learning_rate": 6.874282320543557e-06, + "loss": 0.6741, + "step": 14480 + }, + { + "epoch": 0.6411564814097385, + "grad_norm": 1.318845272064209, + "learning_rate": 6.872814343083192e-06, + "loss": 0.5605, + "step": 14481 + }, + { + "epoch": 0.6412007571145523, + "grad_norm": 1.6034351587295532, + "learning_rate": 6.871346440313111e-06, + "loss": 0.5701, + "step": 14482 + }, + { + "epoch": 0.6412450328193662, + "grad_norm": 1.466886043548584, + "learning_rate": 6.869878612268372e-06, + "loss": 0.5663, + "step": 14483 + }, + { + "epoch": 0.6412893085241801, + "grad_norm": 1.4183391332626343, + "learning_rate": 6.868410858984036e-06, + "loss": 0.5387, + "step": 14484 + }, + { + "epoch": 0.6413335842289939, + "grad_norm": 1.5721986293792725, + "learning_rate": 6.8669431804951515e-06, + "loss": 0.5857, + "step": 14485 + }, + { + "epoch": 0.6413778599338078, + "grad_norm": 1.4546411037445068, + "learning_rate": 6.865475576836782e-06, + "loss": 0.6163, + "step": 14486 + }, + { + "epoch": 0.6414221356386217, + "grad_norm": 1.3606560230255127, + "learning_rate": 6.864008048043977e-06, + "loss": 0.5471, + "step": 14487 + }, + { + "epoch": 0.6414664113434356, + "grad_norm": 1.489121913909912, + "learning_rate": 6.8625405941517845e-06, + "loss": 0.582, + "step": 14488 + }, + { + "epoch": 0.6415106870482494, + "grad_norm": 1.462296485900879, + "learning_rate": 6.8610732151952565e-06, + "loss": 0.5665, + "step": 14489 + }, + { + "epoch": 0.6415549627530633, + "grad_norm": 1.6020880937576294, + "learning_rate": 6.859605911209439e-06, + "loss": 0.6403, + "step": 14490 + }, + { + "epoch": 0.6415992384578773, + "grad_norm": 1.5664427280426025, + "learning_rate": 6.8581386822293765e-06, + "loss": 0.6277, + "step": 14491 + }, + { + "epoch": 0.641643514162691, + "grad_norm": 1.2880685329437256, + "learning_rate": 6.856671528290113e-06, + "loss": 0.5426, + "step": 14492 + }, + { + "epoch": 0.641687789867505, + "grad_norm": 1.6096059083938599, + "learning_rate": 6.855204449426692e-06, + "loss": 0.5921, + "step": 14493 + }, + { + "epoch": 0.6417320655723189, + "grad_norm": 1.5612355470657349, + "learning_rate": 6.853737445674152e-06, + "loss": 0.5986, + "step": 14494 + }, + { + "epoch": 0.6417763412771327, + "grad_norm": 1.6728838682174683, + "learning_rate": 6.852270517067527e-06, + "loss": 0.579, + "step": 14495 + }, + { + "epoch": 0.6418206169819466, + "grad_norm": 1.2583261728286743, + "learning_rate": 6.8508036636418654e-06, + "loss": 0.4948, + "step": 14496 + }, + { + "epoch": 0.6418648926867605, + "grad_norm": 1.7183825969696045, + "learning_rate": 6.849336885432191e-06, + "loss": 0.65, + "step": 14497 + }, + { + "epoch": 0.6419091683915743, + "grad_norm": 1.718714952468872, + "learning_rate": 6.847870182473535e-06, + "loss": 0.5709, + "step": 14498 + }, + { + "epoch": 0.6419534440963882, + "grad_norm": 1.565813422203064, + "learning_rate": 6.846403554800938e-06, + "loss": 0.5869, + "step": 14499 + }, + { + "epoch": 0.6419977198012021, + "grad_norm": 1.5000594854354858, + "learning_rate": 6.8449370024494235e-06, + "loss": 0.5625, + "step": 14500 + }, + { + "epoch": 0.6420419955060159, + "grad_norm": 1.5133941173553467, + "learning_rate": 6.843470525454019e-06, + "loss": 0.5946, + "step": 14501 + }, + { + "epoch": 0.6420862712108298, + "grad_norm": 1.4730145931243896, + "learning_rate": 6.8420041238497525e-06, + "loss": 0.5028, + "step": 14502 + }, + { + "epoch": 0.6421305469156438, + "grad_norm": 1.368621587753296, + "learning_rate": 6.840537797671644e-06, + "loss": 0.5982, + "step": 14503 + }, + { + "epoch": 0.6421748226204576, + "grad_norm": 1.4145885705947876, + "learning_rate": 6.839071546954721e-06, + "loss": 0.5085, + "step": 14504 + }, + { + "epoch": 0.6422190983252715, + "grad_norm": 1.465781807899475, + "learning_rate": 6.837605371733996e-06, + "loss": 0.5752, + "step": 14505 + }, + { + "epoch": 0.6422633740300854, + "grad_norm": 1.6596341133117676, + "learning_rate": 6.836139272044495e-06, + "loss": 0.5666, + "step": 14506 + }, + { + "epoch": 0.6423076497348992, + "grad_norm": 1.3899720907211304, + "learning_rate": 6.8346732479212305e-06, + "loss": 0.5028, + "step": 14507 + }, + { + "epoch": 0.6423519254397131, + "grad_norm": 1.3916035890579224, + "learning_rate": 6.833207299399216e-06, + "loss": 0.5222, + "step": 14508 + }, + { + "epoch": 0.642396201144527, + "grad_norm": 1.4360874891281128, + "learning_rate": 6.83174142651347e-06, + "loss": 0.549, + "step": 14509 + }, + { + "epoch": 0.6424404768493408, + "grad_norm": 2.4220969676971436, + "learning_rate": 6.830275629298999e-06, + "loss": 0.6433, + "step": 14510 + }, + { + "epoch": 0.6424847525541547, + "grad_norm": 1.6691060066223145, + "learning_rate": 6.828809907790812e-06, + "loss": 0.601, + "step": 14511 + }, + { + "epoch": 0.6425290282589686, + "grad_norm": 1.5593996047973633, + "learning_rate": 6.827344262023921e-06, + "loss": 0.5446, + "step": 14512 + }, + { + "epoch": 0.6425733039637824, + "grad_norm": 1.6525895595550537, + "learning_rate": 6.8258786920333286e-06, + "loss": 0.6024, + "step": 14513 + }, + { + "epoch": 0.6426175796685963, + "grad_norm": 1.3531924486160278, + "learning_rate": 6.8244131978540385e-06, + "loss": 0.5143, + "step": 14514 + }, + { + "epoch": 0.6426618553734103, + "grad_norm": 1.469194769859314, + "learning_rate": 6.822947779521053e-06, + "loss": 0.5813, + "step": 14515 + }, + { + "epoch": 0.642706131078224, + "grad_norm": 1.5040478706359863, + "learning_rate": 6.821482437069374e-06, + "loss": 0.5664, + "step": 14516 + }, + { + "epoch": 0.642750406783038, + "grad_norm": 1.4316412210464478, + "learning_rate": 6.820017170533999e-06, + "loss": 0.5692, + "step": 14517 + }, + { + "epoch": 0.6427946824878519, + "grad_norm": 1.4872572422027588, + "learning_rate": 6.818551979949921e-06, + "loss": 0.5918, + "step": 14518 + }, + { + "epoch": 0.6428389581926657, + "grad_norm": 1.7606476545333862, + "learning_rate": 6.8170868653521425e-06, + "loss": 0.618, + "step": 14519 + }, + { + "epoch": 0.6428832338974796, + "grad_norm": 1.2582069635391235, + "learning_rate": 6.815621826775655e-06, + "loss": 0.5411, + "step": 14520 + }, + { + "epoch": 0.6429275096022935, + "grad_norm": 1.4278922080993652, + "learning_rate": 6.81415686425544e-06, + "loss": 0.5897, + "step": 14521 + }, + { + "epoch": 0.6429717853071074, + "grad_norm": 1.6020902395248413, + "learning_rate": 6.8126919778265e-06, + "loss": 0.5749, + "step": 14522 + }, + { + "epoch": 0.6430160610119212, + "grad_norm": 1.4990959167480469, + "learning_rate": 6.8112271675238154e-06, + "loss": 0.6121, + "step": 14523 + }, + { + "epoch": 0.6430603367167351, + "grad_norm": 1.570425033569336, + "learning_rate": 6.809762433382374e-06, + "loss": 0.6237, + "step": 14524 + }, + { + "epoch": 0.643104612421549, + "grad_norm": 1.46918785572052, + "learning_rate": 6.8082977754371585e-06, + "loss": 0.5822, + "step": 14525 + }, + { + "epoch": 0.6431488881263628, + "grad_norm": 1.761460542678833, + "learning_rate": 6.806833193723152e-06, + "loss": 0.5727, + "step": 14526 + }, + { + "epoch": 0.6431931638311768, + "grad_norm": 1.3530057668685913, + "learning_rate": 6.805368688275336e-06, + "loss": 0.5024, + "step": 14527 + }, + { + "epoch": 0.6432374395359907, + "grad_norm": 1.5776678323745728, + "learning_rate": 6.803904259128684e-06, + "loss": 0.6338, + "step": 14528 + }, + { + "epoch": 0.6432817152408045, + "grad_norm": 1.388753890991211, + "learning_rate": 6.80243990631818e-06, + "loss": 0.5789, + "step": 14529 + }, + { + "epoch": 0.6433259909456184, + "grad_norm": 1.3539596796035767, + "learning_rate": 6.800975629878794e-06, + "loss": 0.5477, + "step": 14530 + }, + { + "epoch": 0.6433702666504323, + "grad_norm": 1.4017138481140137, + "learning_rate": 6.799511429845502e-06, + "loss": 0.5504, + "step": 14531 + }, + { + "epoch": 0.6434145423552461, + "grad_norm": 1.5596096515655518, + "learning_rate": 6.798047306253268e-06, + "loss": 0.5762, + "step": 14532 + }, + { + "epoch": 0.64345881806006, + "grad_norm": 2.156463623046875, + "learning_rate": 6.796583259137073e-06, + "loss": 0.6819, + "step": 14533 + }, + { + "epoch": 0.6435030937648739, + "grad_norm": 1.7357566356658936, + "learning_rate": 6.7951192885318774e-06, + "loss": 0.6068, + "step": 14534 + }, + { + "epoch": 0.6435473694696877, + "grad_norm": 1.3875657320022583, + "learning_rate": 6.793655394472645e-06, + "loss": 0.5301, + "step": 14535 + }, + { + "epoch": 0.6435916451745016, + "grad_norm": 1.3578840494155884, + "learning_rate": 6.792191576994345e-06, + "loss": 0.565, + "step": 14536 + }, + { + "epoch": 0.6436359208793155, + "grad_norm": 1.7859289646148682, + "learning_rate": 6.790727836131939e-06, + "loss": 0.5976, + "step": 14537 + }, + { + "epoch": 0.6436801965841293, + "grad_norm": 1.3349807262420654, + "learning_rate": 6.789264171920382e-06, + "loss": 0.5532, + "step": 14538 + }, + { + "epoch": 0.6437244722889433, + "grad_norm": 1.3888856172561646, + "learning_rate": 6.7878005843946395e-06, + "loss": 0.5393, + "step": 14539 + }, + { + "epoch": 0.6437687479937572, + "grad_norm": 1.4353402853012085, + "learning_rate": 6.7863370735896635e-06, + "loss": 0.5528, + "step": 14540 + }, + { + "epoch": 0.643813023698571, + "grad_norm": 1.3455649614334106, + "learning_rate": 6.78487363954041e-06, + "loss": 0.5726, + "step": 14541 + }, + { + "epoch": 0.6438572994033849, + "grad_norm": 1.5775394439697266, + "learning_rate": 6.783410282281828e-06, + "loss": 0.5823, + "step": 14542 + }, + { + "epoch": 0.6439015751081988, + "grad_norm": 1.7864867448806763, + "learning_rate": 6.781947001848876e-06, + "loss": 0.6367, + "step": 14543 + }, + { + "epoch": 0.6439458508130126, + "grad_norm": 1.547073245048523, + "learning_rate": 6.780483798276502e-06, + "loss": 0.586, + "step": 14544 + }, + { + "epoch": 0.6439901265178265, + "grad_norm": 1.3315178155899048, + "learning_rate": 6.779020671599644e-06, + "loss": 0.5819, + "step": 14545 + }, + { + "epoch": 0.6440344022226404, + "grad_norm": 1.6379454135894775, + "learning_rate": 6.777557621853259e-06, + "loss": 0.5682, + "step": 14546 + }, + { + "epoch": 0.6440786779274542, + "grad_norm": 1.2270371913909912, + "learning_rate": 6.776094649072286e-06, + "loss": 0.4915, + "step": 14547 + }, + { + "epoch": 0.6441229536322681, + "grad_norm": 1.2514557838439941, + "learning_rate": 6.774631753291667e-06, + "loss": 0.5474, + "step": 14548 + }, + { + "epoch": 0.644167229337082, + "grad_norm": 1.5404940843582153, + "learning_rate": 6.773168934546344e-06, + "loss": 0.6478, + "step": 14549 + }, + { + "epoch": 0.6442115050418958, + "grad_norm": 1.589066505432129, + "learning_rate": 6.7717061928712525e-06, + "loss": 0.6118, + "step": 14550 + }, + { + "epoch": 0.6442557807467097, + "grad_norm": 1.5681121349334717, + "learning_rate": 6.7702435283013315e-06, + "loss": 0.6377, + "step": 14551 + }, + { + "epoch": 0.6443000564515237, + "grad_norm": 1.2701433897018433, + "learning_rate": 6.768780940871511e-06, + "loss": 0.5324, + "step": 14552 + }, + { + "epoch": 0.6443443321563375, + "grad_norm": 1.384637475013733, + "learning_rate": 6.7673184306167295e-06, + "loss": 0.5805, + "step": 14553 + }, + { + "epoch": 0.6443886078611514, + "grad_norm": 1.264418363571167, + "learning_rate": 6.765855997571914e-06, + "loss": 0.5532, + "step": 14554 + }, + { + "epoch": 0.6444328835659653, + "grad_norm": 1.454188585281372, + "learning_rate": 6.764393641771991e-06, + "loss": 0.6136, + "step": 14555 + }, + { + "epoch": 0.6444771592707791, + "grad_norm": 1.6260621547698975, + "learning_rate": 6.762931363251895e-06, + "loss": 0.5816, + "step": 14556 + }, + { + "epoch": 0.644521434975593, + "grad_norm": 1.5806430578231812, + "learning_rate": 6.761469162046549e-06, + "loss": 0.6574, + "step": 14557 + }, + { + "epoch": 0.6445657106804069, + "grad_norm": 1.5290745496749878, + "learning_rate": 6.760007038190872e-06, + "loss": 0.5555, + "step": 14558 + }, + { + "epoch": 0.6446099863852208, + "grad_norm": 1.5447978973388672, + "learning_rate": 6.758544991719791e-06, + "loss": 0.5632, + "step": 14559 + }, + { + "epoch": 0.6446542620900346, + "grad_norm": 1.3416147232055664, + "learning_rate": 6.757083022668225e-06, + "loss": 0.5742, + "step": 14560 + }, + { + "epoch": 0.6446985377948485, + "grad_norm": 1.2937358617782593, + "learning_rate": 6.755621131071089e-06, + "loss": 0.5129, + "step": 14561 + }, + { + "epoch": 0.6447428134996624, + "grad_norm": 1.467116117477417, + "learning_rate": 6.754159316963298e-06, + "loss": 0.5433, + "step": 14562 + }, + { + "epoch": 0.6447870892044762, + "grad_norm": 1.941380262374878, + "learning_rate": 6.752697580379772e-06, + "loss": 0.5764, + "step": 14563 + }, + { + "epoch": 0.6448313649092902, + "grad_norm": 1.384731650352478, + "learning_rate": 6.751235921355421e-06, + "loss": 0.5286, + "step": 14564 + }, + { + "epoch": 0.6448756406141041, + "grad_norm": 1.5235439538955688, + "learning_rate": 6.749774339925149e-06, + "loss": 0.5034, + "step": 14565 + }, + { + "epoch": 0.6449199163189179, + "grad_norm": 1.7086334228515625, + "learning_rate": 6.748312836123876e-06, + "loss": 0.5239, + "step": 14566 + }, + { + "epoch": 0.6449641920237318, + "grad_norm": 1.2389130592346191, + "learning_rate": 6.746851409986504e-06, + "loss": 0.4867, + "step": 14567 + }, + { + "epoch": 0.6450084677285457, + "grad_norm": 1.3820635080337524, + "learning_rate": 6.745390061547932e-06, + "loss": 0.5405, + "step": 14568 + }, + { + "epoch": 0.6450527434333595, + "grad_norm": 1.6073334217071533, + "learning_rate": 6.743928790843073e-06, + "loss": 0.5959, + "step": 14569 + }, + { + "epoch": 0.6450970191381734, + "grad_norm": 1.541360855102539, + "learning_rate": 6.742467597906822e-06, + "loss": 0.6643, + "step": 14570 + }, + { + "epoch": 0.6451412948429873, + "grad_norm": 1.4145952463150024, + "learning_rate": 6.7410064827740805e-06, + "loss": 0.5783, + "step": 14571 + }, + { + "epoch": 0.6451855705478011, + "grad_norm": 1.6020405292510986, + "learning_rate": 6.739545445479745e-06, + "loss": 0.6104, + "step": 14572 + }, + { + "epoch": 0.645229846252615, + "grad_norm": 1.3547691106796265, + "learning_rate": 6.738084486058711e-06, + "loss": 0.5664, + "step": 14573 + }, + { + "epoch": 0.645274121957429, + "grad_norm": 1.468117356300354, + "learning_rate": 6.736623604545875e-06, + "loss": 0.6001, + "step": 14574 + }, + { + "epoch": 0.6453183976622427, + "grad_norm": 1.4244123697280884, + "learning_rate": 6.7351628009761225e-06, + "loss": 0.5799, + "step": 14575 + }, + { + "epoch": 0.6453626733670567, + "grad_norm": 1.6858361959457397, + "learning_rate": 6.733702075384356e-06, + "loss": 0.6148, + "step": 14576 + }, + { + "epoch": 0.6454069490718706, + "grad_norm": 1.4932764768600464, + "learning_rate": 6.732241427805452e-06, + "loss": 0.5348, + "step": 14577 + }, + { + "epoch": 0.6454512247766844, + "grad_norm": 1.9771231412887573, + "learning_rate": 6.730780858274297e-06, + "loss": 0.6515, + "step": 14578 + }, + { + "epoch": 0.6454955004814983, + "grad_norm": 1.4087045192718506, + "learning_rate": 6.729320366825785e-06, + "loss": 0.5991, + "step": 14579 + }, + { + "epoch": 0.6455397761863122, + "grad_norm": 1.454856276512146, + "learning_rate": 6.727859953494791e-06, + "loss": 0.5737, + "step": 14580 + }, + { + "epoch": 0.645584051891126, + "grad_norm": 1.6426526308059692, + "learning_rate": 6.726399618316197e-06, + "loss": 0.6318, + "step": 14581 + }, + { + "epoch": 0.6456283275959399, + "grad_norm": 1.5107824802398682, + "learning_rate": 6.724939361324883e-06, + "loss": 0.5967, + "step": 14582 + }, + { + "epoch": 0.6456726033007538, + "grad_norm": 1.7492352724075317, + "learning_rate": 6.723479182555726e-06, + "loss": 0.5721, + "step": 14583 + }, + { + "epoch": 0.6457168790055676, + "grad_norm": 1.6671148538589478, + "learning_rate": 6.722019082043602e-06, + "loss": 0.6279, + "step": 14584 + }, + { + "epoch": 0.6457611547103815, + "grad_norm": 1.577748417854309, + "learning_rate": 6.720559059823381e-06, + "loss": 0.5939, + "step": 14585 + }, + { + "epoch": 0.6458054304151954, + "grad_norm": 1.356003999710083, + "learning_rate": 6.719099115929938e-06, + "loss": 0.5624, + "step": 14586 + }, + { + "epoch": 0.6458497061200092, + "grad_norm": 1.822615623474121, + "learning_rate": 6.717639250398142e-06, + "loss": 0.6536, + "step": 14587 + }, + { + "epoch": 0.6458939818248232, + "grad_norm": 1.7476316690444946, + "learning_rate": 6.716179463262855e-06, + "loss": 0.5875, + "step": 14588 + }, + { + "epoch": 0.6459382575296371, + "grad_norm": 1.4733057022094727, + "learning_rate": 6.714719754558952e-06, + "loss": 0.6129, + "step": 14589 + }, + { + "epoch": 0.6459825332344509, + "grad_norm": 1.5545936822891235, + "learning_rate": 6.713260124321293e-06, + "loss": 0.5333, + "step": 14590 + }, + { + "epoch": 0.6460268089392648, + "grad_norm": 2.2382993698120117, + "learning_rate": 6.711800572584741e-06, + "loss": 0.5688, + "step": 14591 + }, + { + "epoch": 0.6460710846440787, + "grad_norm": 1.574196219444275, + "learning_rate": 6.71034109938415e-06, + "loss": 0.6402, + "step": 14592 + }, + { + "epoch": 0.6461153603488926, + "grad_norm": 1.4495881795883179, + "learning_rate": 6.708881704754386e-06, + "loss": 0.519, + "step": 14593 + }, + { + "epoch": 0.6461596360537064, + "grad_norm": 2.009071111679077, + "learning_rate": 6.707422388730305e-06, + "loss": 0.5693, + "step": 14594 + }, + { + "epoch": 0.6462039117585203, + "grad_norm": 1.4272445440292358, + "learning_rate": 6.705963151346755e-06, + "loss": 0.5795, + "step": 14595 + }, + { + "epoch": 0.6462481874633342, + "grad_norm": 1.5422208309173584, + "learning_rate": 6.704503992638596e-06, + "loss": 0.5667, + "step": 14596 + }, + { + "epoch": 0.646292463168148, + "grad_norm": 1.4441884756088257, + "learning_rate": 6.703044912640676e-06, + "loss": 0.5449, + "step": 14597 + }, + { + "epoch": 0.646336738872962, + "grad_norm": 1.5400888919830322, + "learning_rate": 6.70158591138784e-06, + "loss": 0.6273, + "step": 14598 + }, + { + "epoch": 0.6463810145777759, + "grad_norm": 1.6687812805175781, + "learning_rate": 6.7001269889149435e-06, + "loss": 0.6014, + "step": 14599 + }, + { + "epoch": 0.6464252902825897, + "grad_norm": 1.5451385974884033, + "learning_rate": 6.698668145256827e-06, + "loss": 0.5955, + "step": 14600 + }, + { + "epoch": 0.6464695659874036, + "grad_norm": 1.7355269193649292, + "learning_rate": 6.697209380448333e-06, + "loss": 0.667, + "step": 14601 + }, + { + "epoch": 0.6465138416922175, + "grad_norm": 1.560637354850769, + "learning_rate": 6.6957506945242986e-06, + "loss": 0.5777, + "step": 14602 + }, + { + "epoch": 0.6465581173970313, + "grad_norm": 1.452582597732544, + "learning_rate": 6.694292087519574e-06, + "loss": 0.617, + "step": 14603 + }, + { + "epoch": 0.6466023931018452, + "grad_norm": 1.4574036598205566, + "learning_rate": 6.692833559468989e-06, + "loss": 0.5404, + "step": 14604 + }, + { + "epoch": 0.6466466688066591, + "grad_norm": 1.4272024631500244, + "learning_rate": 6.691375110407384e-06, + "loss": 0.5803, + "step": 14605 + }, + { + "epoch": 0.6466909445114729, + "grad_norm": 1.5403728485107422, + "learning_rate": 6.689916740369589e-06, + "loss": 0.5882, + "step": 14606 + }, + { + "epoch": 0.6467352202162868, + "grad_norm": 1.5147581100463867, + "learning_rate": 6.688458449390438e-06, + "loss": 0.5533, + "step": 14607 + }, + { + "epoch": 0.6467794959211007, + "grad_norm": 1.6185405254364014, + "learning_rate": 6.687000237504758e-06, + "loss": 0.5881, + "step": 14608 + }, + { + "epoch": 0.6468237716259145, + "grad_norm": 1.447087287902832, + "learning_rate": 6.685542104747382e-06, + "loss": 0.6055, + "step": 14609 + }, + { + "epoch": 0.6468680473307284, + "grad_norm": 1.3361302614212036, + "learning_rate": 6.684084051153136e-06, + "loss": 0.5886, + "step": 14610 + }, + { + "epoch": 0.6469123230355424, + "grad_norm": 1.7403749227523804, + "learning_rate": 6.68262607675684e-06, + "loss": 0.535, + "step": 14611 + }, + { + "epoch": 0.6469565987403562, + "grad_norm": 1.484480857849121, + "learning_rate": 6.681168181593317e-06, + "loss": 0.5572, + "step": 14612 + }, + { + "epoch": 0.6470008744451701, + "grad_norm": 1.4779818058013916, + "learning_rate": 6.6797103656973935e-06, + "loss": 0.5332, + "step": 14613 + }, + { + "epoch": 0.647045150149984, + "grad_norm": 1.2774089574813843, + "learning_rate": 6.678252629103883e-06, + "loss": 0.52, + "step": 14614 + }, + { + "epoch": 0.6470894258547978, + "grad_norm": 1.3503116369247437, + "learning_rate": 6.676794971847604e-06, + "loss": 0.507, + "step": 14615 + }, + { + "epoch": 0.6471337015596117, + "grad_norm": 1.226357340812683, + "learning_rate": 6.67533739396337e-06, + "loss": 0.5352, + "step": 14616 + }, + { + "epoch": 0.6471779772644256, + "grad_norm": 1.4233479499816895, + "learning_rate": 6.673879895485998e-06, + "loss": 0.5659, + "step": 14617 + }, + { + "epoch": 0.6472222529692394, + "grad_norm": 1.51346755027771, + "learning_rate": 6.672422476450294e-06, + "loss": 0.5503, + "step": 14618 + }, + { + "epoch": 0.6472665286740533, + "grad_norm": 1.4902716875076294, + "learning_rate": 6.670965136891072e-06, + "loss": 0.5443, + "step": 14619 + }, + { + "epoch": 0.6473108043788672, + "grad_norm": 1.5281325578689575, + "learning_rate": 6.669507876843135e-06, + "loss": 0.5224, + "step": 14620 + }, + { + "epoch": 0.647355080083681, + "grad_norm": 1.9249337911605835, + "learning_rate": 6.668050696341293e-06, + "loss": 0.6295, + "step": 14621 + }, + { + "epoch": 0.647399355788495, + "grad_norm": 1.2524116039276123, + "learning_rate": 6.666593595420343e-06, + "loss": 0.469, + "step": 14622 + }, + { + "epoch": 0.6474436314933089, + "grad_norm": 1.437065839767456, + "learning_rate": 6.6651365741150955e-06, + "loss": 0.5287, + "step": 14623 + }, + { + "epoch": 0.6474879071981227, + "grad_norm": 1.2667347192764282, + "learning_rate": 6.663679632460345e-06, + "loss": 0.5109, + "step": 14624 + }, + { + "epoch": 0.6475321829029366, + "grad_norm": 1.4616758823394775, + "learning_rate": 6.662222770490885e-06, + "loss": 0.5833, + "step": 14625 + }, + { + "epoch": 0.6475764586077505, + "grad_norm": 1.6134207248687744, + "learning_rate": 6.66076598824152e-06, + "loss": 0.5729, + "step": 14626 + }, + { + "epoch": 0.6476207343125643, + "grad_norm": 1.500300645828247, + "learning_rate": 6.659309285747041e-06, + "loss": 0.6072, + "step": 14627 + }, + { + "epoch": 0.6476650100173782, + "grad_norm": 1.4852603673934937, + "learning_rate": 6.657852663042237e-06, + "loss": 0.5799, + "step": 14628 + }, + { + "epoch": 0.6477092857221921, + "grad_norm": 1.4746389389038086, + "learning_rate": 6.656396120161903e-06, + "loss": 0.5429, + "step": 14629 + }, + { + "epoch": 0.647753561427006, + "grad_norm": 1.5792033672332764, + "learning_rate": 6.654939657140827e-06, + "loss": 0.5791, + "step": 14630 + }, + { + "epoch": 0.6477978371318198, + "grad_norm": 1.71841299533844, + "learning_rate": 6.653483274013792e-06, + "loss": 0.5351, + "step": 14631 + }, + { + "epoch": 0.6478421128366337, + "grad_norm": 1.3007570505142212, + "learning_rate": 6.652026970815582e-06, + "loss": 0.5172, + "step": 14632 + }, + { + "epoch": 0.6478863885414476, + "grad_norm": 1.8905816078186035, + "learning_rate": 6.650570747580984e-06, + "loss": 0.5739, + "step": 14633 + }, + { + "epoch": 0.6479306642462614, + "grad_norm": 1.5577523708343506, + "learning_rate": 6.649114604344778e-06, + "loss": 0.5122, + "step": 14634 + }, + { + "epoch": 0.6479749399510754, + "grad_norm": 1.456283450126648, + "learning_rate": 6.647658541141735e-06, + "loss": 0.5675, + "step": 14635 + }, + { + "epoch": 0.6480192156558893, + "grad_norm": 1.2820780277252197, + "learning_rate": 6.6462025580066434e-06, + "loss": 0.4958, + "step": 14636 + }, + { + "epoch": 0.6480634913607031, + "grad_norm": 1.406593680381775, + "learning_rate": 6.644746654974273e-06, + "loss": 0.561, + "step": 14637 + }, + { + "epoch": 0.648107767065517, + "grad_norm": 1.266990303993225, + "learning_rate": 6.6432908320793985e-06, + "loss": 0.5361, + "step": 14638 + }, + { + "epoch": 0.6481520427703309, + "grad_norm": 1.4887672662734985, + "learning_rate": 6.641835089356783e-06, + "loss": 0.6169, + "step": 14639 + }, + { + "epoch": 0.6481963184751447, + "grad_norm": 1.348410964012146, + "learning_rate": 6.640379426841206e-06, + "loss": 0.5636, + "step": 14640 + }, + { + "epoch": 0.6482405941799586, + "grad_norm": 1.3775843381881714, + "learning_rate": 6.638923844567432e-06, + "loss": 0.5528, + "step": 14641 + }, + { + "epoch": 0.6482848698847725, + "grad_norm": 1.6288069486618042, + "learning_rate": 6.637468342570223e-06, + "loss": 0.555, + "step": 14642 + }, + { + "epoch": 0.6483291455895863, + "grad_norm": 1.9330662488937378, + "learning_rate": 6.636012920884346e-06, + "loss": 0.7024, + "step": 14643 + }, + { + "epoch": 0.6483734212944002, + "grad_norm": 1.5242475271224976, + "learning_rate": 6.634557579544561e-06, + "loss": 0.678, + "step": 14644 + }, + { + "epoch": 0.6484176969992141, + "grad_norm": 1.3540071249008179, + "learning_rate": 6.633102318585624e-06, + "loss": 0.5708, + "step": 14645 + }, + { + "epoch": 0.648461972704028, + "grad_norm": 1.6502152681350708, + "learning_rate": 6.631647138042303e-06, + "loss": 0.5331, + "step": 14646 + }, + { + "epoch": 0.6485062484088419, + "grad_norm": 1.4997930526733398, + "learning_rate": 6.630192037949348e-06, + "loss": 0.6156, + "step": 14647 + }, + { + "epoch": 0.6485505241136558, + "grad_norm": 2.183955669403076, + "learning_rate": 6.628737018341509e-06, + "loss": 0.6293, + "step": 14648 + }, + { + "epoch": 0.6485947998184696, + "grad_norm": 1.4790605306625366, + "learning_rate": 6.627282079253539e-06, + "loss": 0.5705, + "step": 14649 + }, + { + "epoch": 0.6486390755232835, + "grad_norm": 1.4094868898391724, + "learning_rate": 6.625827220720193e-06, + "loss": 0.5217, + "step": 14650 + }, + { + "epoch": 0.6486833512280974, + "grad_norm": 1.3596107959747314, + "learning_rate": 6.624372442776219e-06, + "loss": 0.584, + "step": 14651 + }, + { + "epoch": 0.6487276269329112, + "grad_norm": 1.4369722604751587, + "learning_rate": 6.622917745456357e-06, + "loss": 0.5675, + "step": 14652 + }, + { + "epoch": 0.6487719026377251, + "grad_norm": 1.781313419342041, + "learning_rate": 6.6214631287953564e-06, + "loss": 0.5743, + "step": 14653 + }, + { + "epoch": 0.648816178342539, + "grad_norm": 1.754240870475769, + "learning_rate": 6.620008592827959e-06, + "loss": 0.5645, + "step": 14654 + }, + { + "epoch": 0.6488604540473528, + "grad_norm": 1.4720110893249512, + "learning_rate": 6.618554137588902e-06, + "loss": 0.6168, + "step": 14655 + }, + { + "epoch": 0.6489047297521667, + "grad_norm": 1.6396675109863281, + "learning_rate": 6.617099763112929e-06, + "loss": 0.5941, + "step": 14656 + }, + { + "epoch": 0.6489490054569806, + "grad_norm": 2.1513264179229736, + "learning_rate": 6.615645469434771e-06, + "loss": 0.6225, + "step": 14657 + }, + { + "epoch": 0.6489932811617944, + "grad_norm": 1.4471557140350342, + "learning_rate": 6.614191256589168e-06, + "loss": 0.5436, + "step": 14658 + }, + { + "epoch": 0.6490375568666084, + "grad_norm": 1.3482613563537598, + "learning_rate": 6.612737124610844e-06, + "loss": 0.5447, + "step": 14659 + }, + { + "epoch": 0.6490818325714223, + "grad_norm": 1.494715929031372, + "learning_rate": 6.611283073534543e-06, + "loss": 0.5663, + "step": 14660 + }, + { + "epoch": 0.6491261082762361, + "grad_norm": 1.3064043521881104, + "learning_rate": 6.6098291033949825e-06, + "loss": 0.535, + "step": 14661 + }, + { + "epoch": 0.64917038398105, + "grad_norm": 1.7646321058273315, + "learning_rate": 6.608375214226894e-06, + "loss": 0.6486, + "step": 14662 + }, + { + "epoch": 0.6492146596858639, + "grad_norm": 1.3031511306762695, + "learning_rate": 6.606921406065003e-06, + "loss": 0.5043, + "step": 14663 + }, + { + "epoch": 0.6492589353906778, + "grad_norm": 1.6478955745697021, + "learning_rate": 6.605467678944033e-06, + "loss": 0.5121, + "step": 14664 + }, + { + "epoch": 0.6493032110954916, + "grad_norm": 1.5470753908157349, + "learning_rate": 6.6040140328987e-06, + "loss": 0.5998, + "step": 14665 + }, + { + "epoch": 0.6493474868003055, + "grad_norm": 1.57319974899292, + "learning_rate": 6.602560467963729e-06, + "loss": 0.5707, + "step": 14666 + }, + { + "epoch": 0.6493917625051194, + "grad_norm": 1.3912509679794312, + "learning_rate": 6.601106984173835e-06, + "loss": 0.5121, + "step": 14667 + }, + { + "epoch": 0.6494360382099332, + "grad_norm": 1.6391807794570923, + "learning_rate": 6.599653581563733e-06, + "loss": 0.5478, + "step": 14668 + }, + { + "epoch": 0.6494803139147471, + "grad_norm": 1.367270588874817, + "learning_rate": 6.598200260168133e-06, + "loss": 0.5601, + "step": 14669 + }, + { + "epoch": 0.649524589619561, + "grad_norm": 1.4466617107391357, + "learning_rate": 6.596747020021754e-06, + "loss": 0.6035, + "step": 14670 + }, + { + "epoch": 0.6495688653243749, + "grad_norm": 1.6492195129394531, + "learning_rate": 6.595293861159302e-06, + "loss": 0.5854, + "step": 14671 + }, + { + "epoch": 0.6496131410291888, + "grad_norm": 1.5222622156143188, + "learning_rate": 6.593840783615479e-06, + "loss": 0.6181, + "step": 14672 + }, + { + "epoch": 0.6496574167340027, + "grad_norm": 1.4749352931976318, + "learning_rate": 6.592387787425001e-06, + "loss": 0.5932, + "step": 14673 + }, + { + "epoch": 0.6497016924388165, + "grad_norm": 1.4020404815673828, + "learning_rate": 6.590934872622564e-06, + "loss": 0.5355, + "step": 14674 + }, + { + "epoch": 0.6497459681436304, + "grad_norm": 2.1055479049682617, + "learning_rate": 6.5894820392428705e-06, + "loss": 0.6281, + "step": 14675 + }, + { + "epoch": 0.6497902438484443, + "grad_norm": 1.869187831878662, + "learning_rate": 6.5880292873206235e-06, + "loss": 0.6186, + "step": 14676 + }, + { + "epoch": 0.6498345195532581, + "grad_norm": 1.308517575263977, + "learning_rate": 6.586576616890518e-06, + "loss": 0.5228, + "step": 14677 + }, + { + "epoch": 0.649878795258072, + "grad_norm": 1.5714913606643677, + "learning_rate": 6.585124027987253e-06, + "loss": 0.565, + "step": 14678 + }, + { + "epoch": 0.6499230709628859, + "grad_norm": 1.3592826128005981, + "learning_rate": 6.5836715206455155e-06, + "loss": 0.6119, + "step": 14679 + }, + { + "epoch": 0.6499673466676997, + "grad_norm": 1.5215736627578735, + "learning_rate": 6.582219094900005e-06, + "loss": 0.5465, + "step": 14680 + }, + { + "epoch": 0.6500116223725136, + "grad_norm": 1.2910656929016113, + "learning_rate": 6.58076675078541e-06, + "loss": 0.5034, + "step": 14681 + }, + { + "epoch": 0.6500558980773276, + "grad_norm": 1.8733959197998047, + "learning_rate": 6.57931448833641e-06, + "loss": 0.6354, + "step": 14682 + }, + { + "epoch": 0.6501001737821414, + "grad_norm": 1.3682775497436523, + "learning_rate": 6.577862307587706e-06, + "loss": 0.5271, + "step": 14683 + }, + { + "epoch": 0.6501444494869553, + "grad_norm": 1.508547067642212, + "learning_rate": 6.5764102085739716e-06, + "loss": 0.5919, + "step": 14684 + }, + { + "epoch": 0.6501887251917692, + "grad_norm": 1.4421751499176025, + "learning_rate": 6.57495819132989e-06, + "loss": 0.5442, + "step": 14685 + }, + { + "epoch": 0.650233000896583, + "grad_norm": 1.6661986112594604, + "learning_rate": 6.573506255890146e-06, + "loss": 0.541, + "step": 14686 + }, + { + "epoch": 0.6502772766013969, + "grad_norm": 1.5056476593017578, + "learning_rate": 6.5720544022894165e-06, + "loss": 0.5647, + "step": 14687 + }, + { + "epoch": 0.6503215523062108, + "grad_norm": 1.426053762435913, + "learning_rate": 6.570602630562374e-06, + "loss": 0.6236, + "step": 14688 + }, + { + "epoch": 0.6503658280110246, + "grad_norm": 1.5832792520523071, + "learning_rate": 6.569150940743694e-06, + "loss": 0.5589, + "step": 14689 + }, + { + "epoch": 0.6504101037158385, + "grad_norm": 1.3666119575500488, + "learning_rate": 6.5676993328680524e-06, + "loss": 0.5572, + "step": 14690 + }, + { + "epoch": 0.6504543794206524, + "grad_norm": 1.4752817153930664, + "learning_rate": 6.566247806970119e-06, + "loss": 0.5672, + "step": 14691 + }, + { + "epoch": 0.6504986551254662, + "grad_norm": 1.5187300443649292, + "learning_rate": 6.564796363084555e-06, + "loss": 0.5445, + "step": 14692 + }, + { + "epoch": 0.6505429308302801, + "grad_norm": 1.8755972385406494, + "learning_rate": 6.563345001246037e-06, + "loss": 0.6744, + "step": 14693 + }, + { + "epoch": 0.650587206535094, + "grad_norm": 1.5451990365982056, + "learning_rate": 6.561893721489228e-06, + "loss": 0.5593, + "step": 14694 + }, + { + "epoch": 0.6506314822399079, + "grad_norm": 1.7100393772125244, + "learning_rate": 6.560442523848781e-06, + "loss": 0.5763, + "step": 14695 + }, + { + "epoch": 0.6506757579447218, + "grad_norm": 1.659555435180664, + "learning_rate": 6.558991408359369e-06, + "loss": 0.5752, + "step": 14696 + }, + { + "epoch": 0.6507200336495357, + "grad_norm": 1.3173807859420776, + "learning_rate": 6.557540375055646e-06, + "loss": 0.5837, + "step": 14697 + }, + { + "epoch": 0.6507643093543495, + "grad_norm": 1.5795108079910278, + "learning_rate": 6.556089423972267e-06, + "loss": 0.5537, + "step": 14698 + }, + { + "epoch": 0.6508085850591634, + "grad_norm": 1.473317265510559, + "learning_rate": 6.5546385551438864e-06, + "loss": 0.5059, + "step": 14699 + }, + { + "epoch": 0.6508528607639773, + "grad_norm": 1.4363059997558594, + "learning_rate": 6.55318776860516e-06, + "loss": 0.5994, + "step": 14700 + }, + { + "epoch": 0.6508971364687912, + "grad_norm": 1.3679629564285278, + "learning_rate": 6.551737064390737e-06, + "loss": 0.5169, + "step": 14701 + }, + { + "epoch": 0.650941412173605, + "grad_norm": 1.441089391708374, + "learning_rate": 6.550286442535263e-06, + "loss": 0.5235, + "step": 14702 + }, + { + "epoch": 0.6509856878784189, + "grad_norm": 1.6137604713439941, + "learning_rate": 6.548835903073395e-06, + "loss": 0.6074, + "step": 14703 + }, + { + "epoch": 0.6510299635832328, + "grad_norm": 1.4249675273895264, + "learning_rate": 6.547385446039769e-06, + "loss": 0.5819, + "step": 14704 + }, + { + "epoch": 0.6510742392880466, + "grad_norm": 1.8534444570541382, + "learning_rate": 6.545935071469026e-06, + "loss": 0.6714, + "step": 14705 + }, + { + "epoch": 0.6511185149928606, + "grad_norm": 1.4256880283355713, + "learning_rate": 6.5444847793958165e-06, + "loss": 0.5765, + "step": 14706 + }, + { + "epoch": 0.6511627906976745, + "grad_norm": 1.9402081966400146, + "learning_rate": 6.543034569854773e-06, + "loss": 0.6005, + "step": 14707 + }, + { + "epoch": 0.6512070664024883, + "grad_norm": 1.5629525184631348, + "learning_rate": 6.541584442880535e-06, + "loss": 0.5411, + "step": 14708 + }, + { + "epoch": 0.6512513421073022, + "grad_norm": 1.313098669052124, + "learning_rate": 6.540134398507735e-06, + "loss": 0.5092, + "step": 14709 + }, + { + "epoch": 0.6512956178121161, + "grad_norm": 1.3806418180465698, + "learning_rate": 6.538684436771009e-06, + "loss": 0.5406, + "step": 14710 + }, + { + "epoch": 0.6513398935169299, + "grad_norm": 1.8918230533599854, + "learning_rate": 6.537234557704988e-06, + "loss": 0.6162, + "step": 14711 + }, + { + "epoch": 0.6513841692217438, + "grad_norm": 1.390316128730774, + "learning_rate": 6.535784761344297e-06, + "loss": 0.5397, + "step": 14712 + }, + { + "epoch": 0.6514284449265577, + "grad_norm": 1.596096158027649, + "learning_rate": 6.534335047723568e-06, + "loss": 0.6073, + "step": 14713 + }, + { + "epoch": 0.6514727206313715, + "grad_norm": 1.4187724590301514, + "learning_rate": 6.5328854168774255e-06, + "loss": 0.502, + "step": 14714 + }, + { + "epoch": 0.6515169963361854, + "grad_norm": 1.2453936338424683, + "learning_rate": 6.531435868840488e-06, + "loss": 0.5455, + "step": 14715 + }, + { + "epoch": 0.6515612720409993, + "grad_norm": 1.4622660875320435, + "learning_rate": 6.529986403647384e-06, + "loss": 0.6205, + "step": 14716 + }, + { + "epoch": 0.6516055477458131, + "grad_norm": 2.050715923309326, + "learning_rate": 6.528537021332728e-06, + "loss": 0.5873, + "step": 14717 + }, + { + "epoch": 0.651649823450627, + "grad_norm": 1.2915743589401245, + "learning_rate": 6.527087721931142e-06, + "loss": 0.4792, + "step": 14718 + }, + { + "epoch": 0.651694099155441, + "grad_norm": 1.4010132551193237, + "learning_rate": 6.525638505477232e-06, + "loss": 0.571, + "step": 14719 + }, + { + "epoch": 0.6517383748602548, + "grad_norm": 1.5137450695037842, + "learning_rate": 6.524189372005622e-06, + "loss": 0.6063, + "step": 14720 + }, + { + "epoch": 0.6517826505650687, + "grad_norm": 1.5015314817428589, + "learning_rate": 6.522740321550918e-06, + "loss": 0.5817, + "step": 14721 + }, + { + "epoch": 0.6518269262698826, + "grad_norm": 1.5835471153259277, + "learning_rate": 6.521291354147727e-06, + "loss": 0.5901, + "step": 14722 + }, + { + "epoch": 0.6518712019746964, + "grad_norm": 1.1250489950180054, + "learning_rate": 6.519842469830664e-06, + "loss": 0.487, + "step": 14723 + }, + { + "epoch": 0.6519154776795103, + "grad_norm": 1.260254979133606, + "learning_rate": 6.518393668634328e-06, + "loss": 0.5085, + "step": 14724 + }, + { + "epoch": 0.6519597533843242, + "grad_norm": 1.6705008745193481, + "learning_rate": 6.516944950593322e-06, + "loss": 0.5945, + "step": 14725 + }, + { + "epoch": 0.652004029089138, + "grad_norm": 1.5147908926010132, + "learning_rate": 6.515496315742255e-06, + "loss": 0.5535, + "step": 14726 + }, + { + "epoch": 0.6520483047939519, + "grad_norm": 1.7645485401153564, + "learning_rate": 6.514047764115718e-06, + "loss": 0.5707, + "step": 14727 + }, + { + "epoch": 0.6520925804987658, + "grad_norm": 1.3023930788040161, + "learning_rate": 6.5125992957483144e-06, + "loss": 0.4931, + "step": 14728 + }, + { + "epoch": 0.6521368562035796, + "grad_norm": 1.8350871801376343, + "learning_rate": 6.511150910674631e-06, + "loss": 0.5848, + "step": 14729 + }, + { + "epoch": 0.6521811319083936, + "grad_norm": 1.4163203239440918, + "learning_rate": 6.509702608929273e-06, + "loss": 0.5533, + "step": 14730 + }, + { + "epoch": 0.6522254076132075, + "grad_norm": 1.5254507064819336, + "learning_rate": 6.508254390546824e-06, + "loss": 0.5972, + "step": 14731 + }, + { + "epoch": 0.6522696833180213, + "grad_norm": 1.6699861288070679, + "learning_rate": 6.506806255561876e-06, + "loss": 0.5654, + "step": 14732 + }, + { + "epoch": 0.6523139590228352, + "grad_norm": 1.6104906797409058, + "learning_rate": 6.505358204009018e-06, + "loss": 0.5806, + "step": 14733 + }, + { + "epoch": 0.6523582347276491, + "grad_norm": 1.408323884010315, + "learning_rate": 6.503910235922833e-06, + "loss": 0.5737, + "step": 14734 + }, + { + "epoch": 0.652402510432463, + "grad_norm": 1.3921198844909668, + "learning_rate": 6.502462351337904e-06, + "loss": 0.5861, + "step": 14735 + }, + { + "epoch": 0.6524467861372768, + "grad_norm": 1.7459797859191895, + "learning_rate": 6.501014550288816e-06, + "loss": 0.4924, + "step": 14736 + }, + { + "epoch": 0.6524910618420907, + "grad_norm": 1.4861074686050415, + "learning_rate": 6.499566832810146e-06, + "loss": 0.5636, + "step": 14737 + }, + { + "epoch": 0.6525353375469046, + "grad_norm": 1.8292838335037231, + "learning_rate": 6.498119198936472e-06, + "loss": 0.5996, + "step": 14738 + }, + { + "epoch": 0.6525796132517184, + "grad_norm": 1.497559666633606, + "learning_rate": 6.496671648702366e-06, + "loss": 0.5486, + "step": 14739 + }, + { + "epoch": 0.6526238889565323, + "grad_norm": 1.4751040935516357, + "learning_rate": 6.495224182142409e-06, + "loss": 0.5934, + "step": 14740 + }, + { + "epoch": 0.6526681646613463, + "grad_norm": 1.733479619026184, + "learning_rate": 6.493776799291168e-06, + "loss": 0.6139, + "step": 14741 + }, + { + "epoch": 0.65271244036616, + "grad_norm": 1.5661265850067139, + "learning_rate": 6.492329500183212e-06, + "loss": 0.5523, + "step": 14742 + }, + { + "epoch": 0.652756716070974, + "grad_norm": 1.491076111793518, + "learning_rate": 6.490882284853113e-06, + "loss": 0.5538, + "step": 14743 + }, + { + "epoch": 0.6528009917757879, + "grad_norm": 1.3615937232971191, + "learning_rate": 6.489435153335432e-06, + "loss": 0.5358, + "step": 14744 + }, + { + "epoch": 0.6528452674806017, + "grad_norm": 1.4886584281921387, + "learning_rate": 6.487988105664733e-06, + "loss": 0.6181, + "step": 14745 + }, + { + "epoch": 0.6528895431854156, + "grad_norm": 1.5118423700332642, + "learning_rate": 6.486541141875578e-06, + "loss": 0.5636, + "step": 14746 + }, + { + "epoch": 0.6529338188902295, + "grad_norm": 1.4003593921661377, + "learning_rate": 6.485094262002529e-06, + "loss": 0.5193, + "step": 14747 + }, + { + "epoch": 0.6529780945950433, + "grad_norm": 1.3335391283035278, + "learning_rate": 6.48364746608014e-06, + "loss": 0.5704, + "step": 14748 + }, + { + "epoch": 0.6530223702998572, + "grad_norm": 1.885949730873108, + "learning_rate": 6.482200754142967e-06, + "loss": 0.5569, + "step": 14749 + }, + { + "epoch": 0.6530666460046711, + "grad_norm": 1.6809663772583008, + "learning_rate": 6.480754126225569e-06, + "loss": 0.6136, + "step": 14750 + }, + { + "epoch": 0.6531109217094849, + "grad_norm": 1.5948870182037354, + "learning_rate": 6.479307582362491e-06, + "loss": 0.5048, + "step": 14751 + }, + { + "epoch": 0.6531551974142988, + "grad_norm": 1.716183066368103, + "learning_rate": 6.477861122588282e-06, + "loss": 0.6354, + "step": 14752 + }, + { + "epoch": 0.6531994731191128, + "grad_norm": 2.1102170944213867, + "learning_rate": 6.476414746937495e-06, + "loss": 0.7272, + "step": 14753 + }, + { + "epoch": 0.6532437488239266, + "grad_norm": 1.3419362306594849, + "learning_rate": 6.474968455444672e-06, + "loss": 0.598, + "step": 14754 + }, + { + "epoch": 0.6532880245287405, + "grad_norm": 1.4120951890945435, + "learning_rate": 6.473522248144359e-06, + "loss": 0.576, + "step": 14755 + }, + { + "epoch": 0.6533323002335544, + "grad_norm": 1.4930262565612793, + "learning_rate": 6.472076125071094e-06, + "loss": 0.5769, + "step": 14756 + }, + { + "epoch": 0.6533765759383682, + "grad_norm": 1.3878260850906372, + "learning_rate": 6.4706300862594195e-06, + "loss": 0.5948, + "step": 14757 + }, + { + "epoch": 0.6534208516431821, + "grad_norm": 1.2566412687301636, + "learning_rate": 6.469184131743871e-06, + "loss": 0.522, + "step": 14758 + }, + { + "epoch": 0.653465127347996, + "grad_norm": 1.5388903617858887, + "learning_rate": 6.467738261558982e-06, + "loss": 0.5593, + "step": 14759 + }, + { + "epoch": 0.6535094030528098, + "grad_norm": 1.8858060836791992, + "learning_rate": 6.466292475739293e-06, + "loss": 0.646, + "step": 14760 + }, + { + "epoch": 0.6535536787576237, + "grad_norm": 1.4737317562103271, + "learning_rate": 6.464846774319329e-06, + "loss": 0.5414, + "step": 14761 + }, + { + "epoch": 0.6535979544624376, + "grad_norm": 1.6187050342559814, + "learning_rate": 6.463401157333619e-06, + "loss": 0.5649, + "step": 14762 + }, + { + "epoch": 0.6536422301672514, + "grad_norm": 1.5096958875656128, + "learning_rate": 6.461955624816696e-06, + "loss": 0.5589, + "step": 14763 + }, + { + "epoch": 0.6536865058720653, + "grad_norm": 1.6151487827301025, + "learning_rate": 6.460510176803081e-06, + "loss": 0.5492, + "step": 14764 + }, + { + "epoch": 0.6537307815768792, + "grad_norm": 1.3915379047393799, + "learning_rate": 6.459064813327302e-06, + "loss": 0.576, + "step": 14765 + }, + { + "epoch": 0.653775057281693, + "grad_norm": 1.4720170497894287, + "learning_rate": 6.457619534423869e-06, + "loss": 0.5511, + "step": 14766 + }, + { + "epoch": 0.653819332986507, + "grad_norm": 1.6536723375320435, + "learning_rate": 6.456174340127315e-06, + "loss": 0.5897, + "step": 14767 + }, + { + "epoch": 0.6538636086913209, + "grad_norm": 1.372849702835083, + "learning_rate": 6.4547292304721495e-06, + "loss": 0.5377, + "step": 14768 + }, + { + "epoch": 0.6539078843961348, + "grad_norm": 1.8746931552886963, + "learning_rate": 6.45328420549289e-06, + "loss": 0.5931, + "step": 14769 + }, + { + "epoch": 0.6539521601009486, + "grad_norm": 1.5863860845565796, + "learning_rate": 6.451839265224051e-06, + "loss": 0.5722, + "step": 14770 + }, + { + "epoch": 0.6539964358057625, + "grad_norm": 1.5526732206344604, + "learning_rate": 6.45039440970014e-06, + "loss": 0.5734, + "step": 14771 + }, + { + "epoch": 0.6540407115105764, + "grad_norm": 1.898236870765686, + "learning_rate": 6.448949638955666e-06, + "loss": 0.6182, + "step": 14772 + }, + { + "epoch": 0.6540849872153902, + "grad_norm": 1.7117103338241577, + "learning_rate": 6.447504953025142e-06, + "loss": 0.6204, + "step": 14773 + }, + { + "epoch": 0.6541292629202041, + "grad_norm": 1.5843583345413208, + "learning_rate": 6.446060351943073e-06, + "loss": 0.5934, + "step": 14774 + }, + { + "epoch": 0.654173538625018, + "grad_norm": 1.8291200399398804, + "learning_rate": 6.444615835743955e-06, + "loss": 0.6105, + "step": 14775 + }, + { + "epoch": 0.6542178143298318, + "grad_norm": 1.5820982456207275, + "learning_rate": 6.44317140446229e-06, + "loss": 0.5216, + "step": 14776 + }, + { + "epoch": 0.6542620900346457, + "grad_norm": 1.7223076820373535, + "learning_rate": 6.441727058132583e-06, + "loss": 0.5659, + "step": 14777 + }, + { + "epoch": 0.6543063657394597, + "grad_norm": 1.8034858703613281, + "learning_rate": 6.440282796789328e-06, + "loss": 0.5574, + "step": 14778 + }, + { + "epoch": 0.6543506414442735, + "grad_norm": 1.5349969863891602, + "learning_rate": 6.438838620467018e-06, + "loss": 0.5474, + "step": 14779 + }, + { + "epoch": 0.6543949171490874, + "grad_norm": 1.3105868101119995, + "learning_rate": 6.437394529200151e-06, + "loss": 0.5471, + "step": 14780 + }, + { + "epoch": 0.6544391928539013, + "grad_norm": 2.0385074615478516, + "learning_rate": 6.435950523023213e-06, + "loss": 0.6516, + "step": 14781 + }, + { + "epoch": 0.6544834685587151, + "grad_norm": 1.4911093711853027, + "learning_rate": 6.434506601970694e-06, + "loss": 0.5959, + "step": 14782 + }, + { + "epoch": 0.654527744263529, + "grad_norm": 1.635871171951294, + "learning_rate": 6.433062766077085e-06, + "loss": 0.6564, + "step": 14783 + }, + { + "epoch": 0.6545720199683429, + "grad_norm": 1.3563120365142822, + "learning_rate": 6.4316190153768666e-06, + "loss": 0.4965, + "step": 14784 + }, + { + "epoch": 0.6546162956731567, + "grad_norm": 1.7040352821350098, + "learning_rate": 6.430175349904522e-06, + "loss": 0.5201, + "step": 14785 + }, + { + "epoch": 0.6546605713779706, + "grad_norm": 1.4728198051452637, + "learning_rate": 6.4287317696945275e-06, + "loss": 0.6044, + "step": 14786 + }, + { + "epoch": 0.6547048470827845, + "grad_norm": 1.291035532951355, + "learning_rate": 6.427288274781372e-06, + "loss": 0.5625, + "step": 14787 + }, + { + "epoch": 0.6547491227875983, + "grad_norm": 1.6532410383224487, + "learning_rate": 6.425844865199526e-06, + "loss": 0.6319, + "step": 14788 + }, + { + "epoch": 0.6547933984924122, + "grad_norm": 1.389882206916809, + "learning_rate": 6.4244015409834645e-06, + "loss": 0.5544, + "step": 14789 + }, + { + "epoch": 0.6548376741972262, + "grad_norm": 1.4909985065460205, + "learning_rate": 6.422958302167663e-06, + "loss": 0.6214, + "step": 14790 + }, + { + "epoch": 0.65488194990204, + "grad_norm": 1.3899167776107788, + "learning_rate": 6.42151514878659e-06, + "loss": 0.6252, + "step": 14791 + }, + { + "epoch": 0.6549262256068539, + "grad_norm": 2.176861524581909, + "learning_rate": 6.420072080874711e-06, + "loss": 0.642, + "step": 14792 + }, + { + "epoch": 0.6549705013116678, + "grad_norm": 1.7240090370178223, + "learning_rate": 6.418629098466498e-06, + "loss": 0.5666, + "step": 14793 + }, + { + "epoch": 0.6550147770164816, + "grad_norm": 1.795798897743225, + "learning_rate": 6.417186201596412e-06, + "loss": 0.6049, + "step": 14794 + }, + { + "epoch": 0.6550590527212955, + "grad_norm": 1.8454947471618652, + "learning_rate": 6.415743390298918e-06, + "loss": 0.6459, + "step": 14795 + }, + { + "epoch": 0.6551033284261094, + "grad_norm": 1.3259526491165161, + "learning_rate": 6.414300664608471e-06, + "loss": 0.5373, + "step": 14796 + }, + { + "epoch": 0.6551476041309232, + "grad_norm": 1.5808025598526, + "learning_rate": 6.412858024559535e-06, + "loss": 0.5996, + "step": 14797 + }, + { + "epoch": 0.6551918798357371, + "grad_norm": 1.5042849779129028, + "learning_rate": 6.411415470186567e-06, + "loss": 0.5641, + "step": 14798 + }, + { + "epoch": 0.655236155540551, + "grad_norm": 1.447249174118042, + "learning_rate": 6.409973001524012e-06, + "loss": 0.5918, + "step": 14799 + }, + { + "epoch": 0.6552804312453648, + "grad_norm": 1.6646746397018433, + "learning_rate": 6.408530618606335e-06, + "loss": 0.5325, + "step": 14800 + }, + { + "epoch": 0.6553247069501787, + "grad_norm": 1.6103931665420532, + "learning_rate": 6.4070883214679776e-06, + "loss": 0.5779, + "step": 14801 + }, + { + "epoch": 0.6553689826549927, + "grad_norm": 1.2358553409576416, + "learning_rate": 6.405646110143389e-06, + "loss": 0.5802, + "step": 14802 + }, + { + "epoch": 0.6554132583598065, + "grad_norm": 1.5028092861175537, + "learning_rate": 6.404203984667019e-06, + "loss": 0.5297, + "step": 14803 + }, + { + "epoch": 0.6554575340646204, + "grad_norm": 1.4210309982299805, + "learning_rate": 6.402761945073308e-06, + "loss": 0.579, + "step": 14804 + }, + { + "epoch": 0.6555018097694343, + "grad_norm": 2.4705734252929688, + "learning_rate": 6.4013199913967e-06, + "loss": 0.6567, + "step": 14805 + }, + { + "epoch": 0.6555460854742482, + "grad_norm": 1.2934651374816895, + "learning_rate": 6.399878123671631e-06, + "loss": 0.5383, + "step": 14806 + }, + { + "epoch": 0.655590361179062, + "grad_norm": 1.4511653184890747, + "learning_rate": 6.3984363419325435e-06, + "loss": 0.5578, + "step": 14807 + }, + { + "epoch": 0.6556346368838759, + "grad_norm": 1.477077603340149, + "learning_rate": 6.396994646213872e-06, + "loss": 0.6149, + "step": 14808 + }, + { + "epoch": 0.6556789125886898, + "grad_norm": 1.3378225564956665, + "learning_rate": 6.395553036550046e-06, + "loss": 0.5395, + "step": 14809 + }, + { + "epoch": 0.6557231882935036, + "grad_norm": 1.5213700532913208, + "learning_rate": 6.394111512975505e-06, + "loss": 0.5363, + "step": 14810 + }, + { + "epoch": 0.6557674639983175, + "grad_norm": 1.3644944429397583, + "learning_rate": 6.392670075524674e-06, + "loss": 0.5051, + "step": 14811 + }, + { + "epoch": 0.6558117397031314, + "grad_norm": 1.6654897928237915, + "learning_rate": 6.391228724231979e-06, + "loss": 0.6092, + "step": 14812 + }, + { + "epoch": 0.6558560154079452, + "grad_norm": 1.8344372510910034, + "learning_rate": 6.389787459131849e-06, + "loss": 0.5515, + "step": 14813 + }, + { + "epoch": 0.6559002911127592, + "grad_norm": 1.516391634941101, + "learning_rate": 6.388346280258706e-06, + "loss": 0.5775, + "step": 14814 + }, + { + "epoch": 0.6559445668175731, + "grad_norm": 1.4643304347991943, + "learning_rate": 6.386905187646972e-06, + "loss": 0.5326, + "step": 14815 + }, + { + "epoch": 0.6559888425223869, + "grad_norm": 1.3586779832839966, + "learning_rate": 6.385464181331065e-06, + "loss": 0.5006, + "step": 14816 + }, + { + "epoch": 0.6560331182272008, + "grad_norm": 1.1877851486206055, + "learning_rate": 6.384023261345402e-06, + "loss": 0.4887, + "step": 14817 + }, + { + "epoch": 0.6560773939320147, + "grad_norm": 1.5976359844207764, + "learning_rate": 6.382582427724402e-06, + "loss": 0.587, + "step": 14818 + }, + { + "epoch": 0.6561216696368285, + "grad_norm": 1.5110446214675903, + "learning_rate": 6.3811416805024695e-06, + "loss": 0.5438, + "step": 14819 + }, + { + "epoch": 0.6561659453416424, + "grad_norm": 1.359719157218933, + "learning_rate": 6.379701019714026e-06, + "loss": 0.5759, + "step": 14820 + }, + { + "epoch": 0.6562102210464563, + "grad_norm": 1.7656549215316772, + "learning_rate": 6.378260445393478e-06, + "loss": 0.5985, + "step": 14821 + }, + { + "epoch": 0.6562544967512701, + "grad_norm": 1.5379393100738525, + "learning_rate": 6.376819957575224e-06, + "loss": 0.565, + "step": 14822 + }, + { + "epoch": 0.656298772456084, + "grad_norm": 1.7849969863891602, + "learning_rate": 6.375379556293679e-06, + "loss": 0.6505, + "step": 14823 + }, + { + "epoch": 0.656343048160898, + "grad_norm": 1.4735746383666992, + "learning_rate": 6.373939241583244e-06, + "loss": 0.5399, + "step": 14824 + }, + { + "epoch": 0.6563873238657117, + "grad_norm": 1.6387605667114258, + "learning_rate": 6.3724990134783145e-06, + "loss": 0.5537, + "step": 14825 + }, + { + "epoch": 0.6564315995705257, + "grad_norm": 1.484185814857483, + "learning_rate": 6.371058872013294e-06, + "loss": 0.5492, + "step": 14826 + }, + { + "epoch": 0.6564758752753396, + "grad_norm": 1.2625062465667725, + "learning_rate": 6.369618817222577e-06, + "loss": 0.4754, + "step": 14827 + }, + { + "epoch": 0.6565201509801534, + "grad_norm": 1.438990831375122, + "learning_rate": 6.368178849140559e-06, + "loss": 0.5291, + "step": 14828 + }, + { + "epoch": 0.6565644266849673, + "grad_norm": 1.5651588439941406, + "learning_rate": 6.3667389678016286e-06, + "loss": 0.5714, + "step": 14829 + }, + { + "epoch": 0.6566087023897812, + "grad_norm": 1.5413492918014526, + "learning_rate": 6.365299173240185e-06, + "loss": 0.5642, + "step": 14830 + }, + { + "epoch": 0.656652978094595, + "grad_norm": 1.444462776184082, + "learning_rate": 6.363859465490609e-06, + "loss": 0.6077, + "step": 14831 + }, + { + "epoch": 0.6566972537994089, + "grad_norm": 1.4956547021865845, + "learning_rate": 6.362419844587287e-06, + "loss": 0.5116, + "step": 14832 + }, + { + "epoch": 0.6567415295042228, + "grad_norm": 1.3741068840026855, + "learning_rate": 6.3609803105646064e-06, + "loss": 0.5456, + "step": 14833 + }, + { + "epoch": 0.6567858052090366, + "grad_norm": 1.406200885772705, + "learning_rate": 6.35954086345695e-06, + "loss": 0.6058, + "step": 14834 + }, + { + "epoch": 0.6568300809138505, + "grad_norm": 1.5379880666732788, + "learning_rate": 6.3581015032986945e-06, + "loss": 0.5453, + "step": 14835 + }, + { + "epoch": 0.6568743566186644, + "grad_norm": 1.3574533462524414, + "learning_rate": 6.356662230124218e-06, + "loss": 0.5332, + "step": 14836 + }, + { + "epoch": 0.6569186323234782, + "grad_norm": 1.5456793308258057, + "learning_rate": 6.355223043967901e-06, + "loss": 0.534, + "step": 14837 + }, + { + "epoch": 0.6569629080282922, + "grad_norm": 1.713536024093628, + "learning_rate": 6.353783944864111e-06, + "loss": 0.6174, + "step": 14838 + }, + { + "epoch": 0.6570071837331061, + "grad_norm": 2.3661351203918457, + "learning_rate": 6.352344932847223e-06, + "loss": 0.5731, + "step": 14839 + }, + { + "epoch": 0.65705145943792, + "grad_norm": 1.490537166595459, + "learning_rate": 6.3509060079516076e-06, + "loss": 0.5047, + "step": 14840 + }, + { + "epoch": 0.6570957351427338, + "grad_norm": 2.2472898960113525, + "learning_rate": 6.349467170211632e-06, + "loss": 0.6228, + "step": 14841 + }, + { + "epoch": 0.6571400108475477, + "grad_norm": 1.4193552732467651, + "learning_rate": 6.348028419661654e-06, + "loss": 0.5404, + "step": 14842 + }, + { + "epoch": 0.6571842865523616, + "grad_norm": 1.4618326425552368, + "learning_rate": 6.34658975633605e-06, + "loss": 0.5473, + "step": 14843 + }, + { + "epoch": 0.6572285622571754, + "grad_norm": 1.3003203868865967, + "learning_rate": 6.345151180269174e-06, + "loss": 0.5646, + "step": 14844 + }, + { + "epoch": 0.6572728379619893, + "grad_norm": 1.7151590585708618, + "learning_rate": 6.343712691495388e-06, + "loss": 0.6409, + "step": 14845 + }, + { + "epoch": 0.6573171136668032, + "grad_norm": 1.556732177734375, + "learning_rate": 6.342274290049042e-06, + "loss": 0.5798, + "step": 14846 + }, + { + "epoch": 0.657361389371617, + "grad_norm": 1.4134252071380615, + "learning_rate": 6.3408359759644996e-06, + "loss": 0.5831, + "step": 14847 + }, + { + "epoch": 0.657405665076431, + "grad_norm": 1.6218985319137573, + "learning_rate": 6.339397749276111e-06, + "loss": 0.5439, + "step": 14848 + }, + { + "epoch": 0.6574499407812449, + "grad_norm": 1.3225985765457153, + "learning_rate": 6.337959610018225e-06, + "loss": 0.5413, + "step": 14849 + }, + { + "epoch": 0.6574942164860587, + "grad_norm": 1.6365673542022705, + "learning_rate": 6.336521558225193e-06, + "loss": 0.5864, + "step": 14850 + }, + { + "epoch": 0.6575384921908726, + "grad_norm": 1.3818186521530151, + "learning_rate": 6.335083593931362e-06, + "loss": 0.5638, + "step": 14851 + }, + { + "epoch": 0.6575827678956865, + "grad_norm": 1.876692295074463, + "learning_rate": 6.333645717171069e-06, + "loss": 0.5547, + "step": 14852 + }, + { + "epoch": 0.6576270436005003, + "grad_norm": 1.8845148086547852, + "learning_rate": 6.332207927978668e-06, + "loss": 0.6614, + "step": 14853 + }, + { + "epoch": 0.6576713193053142, + "grad_norm": 1.3400017023086548, + "learning_rate": 6.3307702263884965e-06, + "loss": 0.5521, + "step": 14854 + }, + { + "epoch": 0.6577155950101281, + "grad_norm": 1.883762001991272, + "learning_rate": 6.3293326124348886e-06, + "loss": 0.6107, + "step": 14855 + }, + { + "epoch": 0.6577598707149419, + "grad_norm": 1.4679133892059326, + "learning_rate": 6.327895086152178e-06, + "loss": 0.5819, + "step": 14856 + }, + { + "epoch": 0.6578041464197558, + "grad_norm": 1.2842521667480469, + "learning_rate": 6.326457647574709e-06, + "loss": 0.5109, + "step": 14857 + }, + { + "epoch": 0.6578484221245697, + "grad_norm": 1.8163293600082397, + "learning_rate": 6.3250202967368055e-06, + "loss": 0.5588, + "step": 14858 + }, + { + "epoch": 0.6578926978293835, + "grad_norm": 1.3567572832107544, + "learning_rate": 6.323583033672799e-06, + "loss": 0.5951, + "step": 14859 + }, + { + "epoch": 0.6579369735341974, + "grad_norm": 1.467477798461914, + "learning_rate": 6.322145858417021e-06, + "loss": 0.5453, + "step": 14860 + }, + { + "epoch": 0.6579812492390114, + "grad_norm": 1.4599857330322266, + "learning_rate": 6.320708771003794e-06, + "loss": 0.6017, + "step": 14861 + }, + { + "epoch": 0.6580255249438252, + "grad_norm": 1.4221636056900024, + "learning_rate": 6.319271771467441e-06, + "loss": 0.4978, + "step": 14862 + }, + { + "epoch": 0.6580698006486391, + "grad_norm": 1.5566743612289429, + "learning_rate": 6.317834859842285e-06, + "loss": 0.5306, + "step": 14863 + }, + { + "epoch": 0.658114076353453, + "grad_norm": 1.556012511253357, + "learning_rate": 6.316398036162645e-06, + "loss": 0.5589, + "step": 14864 + }, + { + "epoch": 0.6581583520582668, + "grad_norm": 2.0418431758880615, + "learning_rate": 6.314961300462841e-06, + "loss": 0.5808, + "step": 14865 + }, + { + "epoch": 0.6582026277630807, + "grad_norm": 1.4519994258880615, + "learning_rate": 6.313524652777179e-06, + "loss": 0.6155, + "step": 14866 + }, + { + "epoch": 0.6582469034678946, + "grad_norm": 1.3718912601470947, + "learning_rate": 6.312088093139985e-06, + "loss": 0.579, + "step": 14867 + }, + { + "epoch": 0.6582911791727084, + "grad_norm": 1.5477702617645264, + "learning_rate": 6.310651621585564e-06, + "loss": 0.5928, + "step": 14868 + }, + { + "epoch": 0.6583354548775223, + "grad_norm": 1.3426728248596191, + "learning_rate": 6.309215238148223e-06, + "loss": 0.5515, + "step": 14869 + }, + { + "epoch": 0.6583797305823362, + "grad_norm": 1.339745283126831, + "learning_rate": 6.307778942862273e-06, + "loss": 0.5578, + "step": 14870 + }, + { + "epoch": 0.65842400628715, + "grad_norm": 1.8780282735824585, + "learning_rate": 6.306342735762015e-06, + "loss": 0.558, + "step": 14871 + }, + { + "epoch": 0.6584682819919639, + "grad_norm": 1.382071852684021, + "learning_rate": 6.304906616881755e-06, + "loss": 0.6314, + "step": 14872 + }, + { + "epoch": 0.6585125576967779, + "grad_norm": 1.2730135917663574, + "learning_rate": 6.303470586255788e-06, + "loss": 0.4877, + "step": 14873 + }, + { + "epoch": 0.6585568334015917, + "grad_norm": 1.2037771940231323, + "learning_rate": 6.302034643918421e-06, + "loss": 0.457, + "step": 14874 + }, + { + "epoch": 0.6586011091064056, + "grad_norm": 1.551220417022705, + "learning_rate": 6.300598789903944e-06, + "loss": 0.5445, + "step": 14875 + }, + { + "epoch": 0.6586453848112195, + "grad_norm": 2.808696985244751, + "learning_rate": 6.299163024246649e-06, + "loss": 0.5525, + "step": 14876 + }, + { + "epoch": 0.6586896605160334, + "grad_norm": 1.5395538806915283, + "learning_rate": 6.297727346980838e-06, + "loss": 0.6218, + "step": 14877 + }, + { + "epoch": 0.6587339362208472, + "grad_norm": 1.6573361158370972, + "learning_rate": 6.296291758140791e-06, + "loss": 0.6107, + "step": 14878 + }, + { + "epoch": 0.6587782119256611, + "grad_norm": 1.8030756711959839, + "learning_rate": 6.294856257760797e-06, + "loss": 0.6353, + "step": 14879 + }, + { + "epoch": 0.658822487630475, + "grad_norm": 1.4994460344314575, + "learning_rate": 6.293420845875149e-06, + "loss": 0.4838, + "step": 14880 + }, + { + "epoch": 0.6588667633352888, + "grad_norm": 1.4093647003173828, + "learning_rate": 6.291985522518127e-06, + "loss": 0.5777, + "step": 14881 + }, + { + "epoch": 0.6589110390401027, + "grad_norm": 1.3532907962799072, + "learning_rate": 6.290550287724009e-06, + "loss": 0.5371, + "step": 14882 + }, + { + "epoch": 0.6589553147449166, + "grad_norm": 1.763224720954895, + "learning_rate": 6.289115141527077e-06, + "loss": 0.6179, + "step": 14883 + }, + { + "epoch": 0.6589995904497304, + "grad_norm": 1.2660868167877197, + "learning_rate": 6.287680083961609e-06, + "loss": 0.4563, + "step": 14884 + }, + { + "epoch": 0.6590438661545444, + "grad_norm": 1.6031430959701538, + "learning_rate": 6.286245115061881e-06, + "loss": 0.5695, + "step": 14885 + }, + { + "epoch": 0.6590881418593583, + "grad_norm": 1.4268155097961426, + "learning_rate": 6.284810234862162e-06, + "loss": 0.5434, + "step": 14886 + }, + { + "epoch": 0.6591324175641721, + "grad_norm": 1.5683270692825317, + "learning_rate": 6.283375443396726e-06, + "loss": 0.516, + "step": 14887 + }, + { + "epoch": 0.659176693268986, + "grad_norm": 1.4393093585968018, + "learning_rate": 6.281940740699843e-06, + "loss": 0.6506, + "step": 14888 + }, + { + "epoch": 0.6592209689737999, + "grad_norm": 1.7910611629486084, + "learning_rate": 6.280506126805775e-06, + "loss": 0.6182, + "step": 14889 + }, + { + "epoch": 0.6592652446786137, + "grad_norm": 1.3408983945846558, + "learning_rate": 6.279071601748792e-06, + "loss": 0.4837, + "step": 14890 + }, + { + "epoch": 0.6593095203834276, + "grad_norm": 1.4481326341629028, + "learning_rate": 6.277637165563153e-06, + "loss": 0.4967, + "step": 14891 + }, + { + "epoch": 0.6593537960882415, + "grad_norm": 1.8050189018249512, + "learning_rate": 6.276202818283122e-06, + "loss": 0.533, + "step": 14892 + }, + { + "epoch": 0.6593980717930553, + "grad_norm": 1.2060165405273438, + "learning_rate": 6.2747685599429506e-06, + "loss": 0.4215, + "step": 14893 + }, + { + "epoch": 0.6594423474978692, + "grad_norm": 1.4740747213363647, + "learning_rate": 6.273334390576902e-06, + "loss": 0.5965, + "step": 14894 + }, + { + "epoch": 0.6594866232026831, + "grad_norm": 1.3552743196487427, + "learning_rate": 6.271900310219228e-06, + "loss": 0.4748, + "step": 14895 + }, + { + "epoch": 0.6595308989074969, + "grad_norm": 1.6976878643035889, + "learning_rate": 6.270466318904176e-06, + "loss": 0.5929, + "step": 14896 + }, + { + "epoch": 0.6595751746123109, + "grad_norm": 1.864740014076233, + "learning_rate": 6.269032416666001e-06, + "loss": 0.6106, + "step": 14897 + }, + { + "epoch": 0.6596194503171248, + "grad_norm": 1.679245948791504, + "learning_rate": 6.26759860353895e-06, + "loss": 0.5787, + "step": 14898 + }, + { + "epoch": 0.6596637260219386, + "grad_norm": 1.3022093772888184, + "learning_rate": 6.266164879557263e-06, + "loss": 0.5044, + "step": 14899 + }, + { + "epoch": 0.6597080017267525, + "grad_norm": 1.8046061992645264, + "learning_rate": 6.2647312447551915e-06, + "loss": 0.6173, + "step": 14900 + }, + { + "epoch": 0.6597522774315664, + "grad_norm": 1.3802684545516968, + "learning_rate": 6.263297699166975e-06, + "loss": 0.5499, + "step": 14901 + }, + { + "epoch": 0.6597965531363802, + "grad_norm": 1.4402321577072144, + "learning_rate": 6.26186424282685e-06, + "loss": 0.5563, + "step": 14902 + }, + { + "epoch": 0.6598408288411941, + "grad_norm": 1.5418874025344849, + "learning_rate": 6.260430875769047e-06, + "loss": 0.5223, + "step": 14903 + }, + { + "epoch": 0.659885104546008, + "grad_norm": 1.4291728734970093, + "learning_rate": 6.258997598027813e-06, + "loss": 0.5385, + "step": 14904 + }, + { + "epoch": 0.6599293802508218, + "grad_norm": 1.5583183765411377, + "learning_rate": 6.257564409637375e-06, + "loss": 0.6094, + "step": 14905 + }, + { + "epoch": 0.6599736559556357, + "grad_norm": 1.4761719703674316, + "learning_rate": 6.256131310631962e-06, + "loss": 0.5538, + "step": 14906 + }, + { + "epoch": 0.6600179316604496, + "grad_norm": 1.5439300537109375, + "learning_rate": 6.254698301045806e-06, + "loss": 0.5908, + "step": 14907 + }, + { + "epoch": 0.6600622073652634, + "grad_norm": 1.6136195659637451, + "learning_rate": 6.253265380913133e-06, + "loss": 0.5682, + "step": 14908 + }, + { + "epoch": 0.6601064830700774, + "grad_norm": 2.0618040561676025, + "learning_rate": 6.251832550268162e-06, + "loss": 0.6979, + "step": 14909 + }, + { + "epoch": 0.6601507587748913, + "grad_norm": 1.501548171043396, + "learning_rate": 6.250399809145121e-06, + "loss": 0.5399, + "step": 14910 + }, + { + "epoch": 0.6601950344797052, + "grad_norm": 2.015805959701538, + "learning_rate": 6.248967157578227e-06, + "loss": 0.6132, + "step": 14911 + }, + { + "epoch": 0.660239310184519, + "grad_norm": 1.4672006368637085, + "learning_rate": 6.247534595601701e-06, + "loss": 0.6096, + "step": 14912 + }, + { + "epoch": 0.6602835858893329, + "grad_norm": 1.570701003074646, + "learning_rate": 6.246102123249749e-06, + "loss": 0.5613, + "step": 14913 + }, + { + "epoch": 0.6603278615941468, + "grad_norm": 1.7147767543792725, + "learning_rate": 6.244669740556596e-06, + "loss": 0.5577, + "step": 14914 + }, + { + "epoch": 0.6603721372989606, + "grad_norm": 1.3905645608901978, + "learning_rate": 6.24323744755645e-06, + "loss": 0.5521, + "step": 14915 + }, + { + "epoch": 0.6604164130037745, + "grad_norm": 1.7792139053344727, + "learning_rate": 6.241805244283514e-06, + "loss": 0.5878, + "step": 14916 + }, + { + "epoch": 0.6604606887085884, + "grad_norm": 1.4146288633346558, + "learning_rate": 6.240373130772003e-06, + "loss": 0.5841, + "step": 14917 + }, + { + "epoch": 0.6605049644134022, + "grad_norm": 1.9231128692626953, + "learning_rate": 6.238941107056118e-06, + "loss": 0.6071, + "step": 14918 + }, + { + "epoch": 0.6605492401182161, + "grad_norm": 1.6618574857711792, + "learning_rate": 6.237509173170062e-06, + "loss": 0.6053, + "step": 14919 + }, + { + "epoch": 0.66059351582303, + "grad_norm": 1.7472277879714966, + "learning_rate": 6.236077329148036e-06, + "loss": 0.7003, + "step": 14920 + }, + { + "epoch": 0.6606377915278439, + "grad_norm": 1.7173491716384888, + "learning_rate": 6.23464557502424e-06, + "loss": 0.6631, + "step": 14921 + }, + { + "epoch": 0.6606820672326578, + "grad_norm": 1.4438081979751587, + "learning_rate": 6.233213910832867e-06, + "loss": 0.5484, + "step": 14922 + }, + { + "epoch": 0.6607263429374717, + "grad_norm": 1.3325616121292114, + "learning_rate": 6.23178233660811e-06, + "loss": 0.4712, + "step": 14923 + }, + { + "epoch": 0.6607706186422855, + "grad_norm": 1.5377978086471558, + "learning_rate": 6.230350852384166e-06, + "loss": 0.5887, + "step": 14924 + }, + { + "epoch": 0.6608148943470994, + "grad_norm": 1.4109337329864502, + "learning_rate": 6.2289194581952265e-06, + "loss": 0.5448, + "step": 14925 + }, + { + "epoch": 0.6608591700519133, + "grad_norm": 1.3626781702041626, + "learning_rate": 6.227488154075467e-06, + "loss": 0.576, + "step": 14926 + }, + { + "epoch": 0.6609034457567271, + "grad_norm": 1.463372826576233, + "learning_rate": 6.226056940059086e-06, + "loss": 0.5915, + "step": 14927 + }, + { + "epoch": 0.660947721461541, + "grad_norm": 1.3216285705566406, + "learning_rate": 6.224625816180264e-06, + "loss": 0.5006, + "step": 14928 + }, + { + "epoch": 0.6609919971663549, + "grad_norm": 1.4411150217056274, + "learning_rate": 6.223194782473176e-06, + "loss": 0.5183, + "step": 14929 + }, + { + "epoch": 0.6610362728711687, + "grad_norm": 1.4224622249603271, + "learning_rate": 6.221763838972009e-06, + "loss": 0.5395, + "step": 14930 + }, + { + "epoch": 0.6610805485759826, + "grad_norm": 1.7986568212509155, + "learning_rate": 6.220332985710936e-06, + "loss": 0.5674, + "step": 14931 + }, + { + "epoch": 0.6611248242807966, + "grad_norm": 1.839781641960144, + "learning_rate": 6.218902222724132e-06, + "loss": 0.6136, + "step": 14932 + }, + { + "epoch": 0.6611690999856104, + "grad_norm": 1.2658488750457764, + "learning_rate": 6.217471550045768e-06, + "loss": 0.5374, + "step": 14933 + }, + { + "epoch": 0.6612133756904243, + "grad_norm": 1.324569821357727, + "learning_rate": 6.216040967710017e-06, + "loss": 0.5017, + "step": 14934 + }, + { + "epoch": 0.6612576513952382, + "grad_norm": 1.4662126302719116, + "learning_rate": 6.214610475751049e-06, + "loss": 0.6136, + "step": 14935 + }, + { + "epoch": 0.661301927100052, + "grad_norm": 1.6190651655197144, + "learning_rate": 6.21318007420302e-06, + "loss": 0.5869, + "step": 14936 + }, + { + "epoch": 0.6613462028048659, + "grad_norm": 1.7365044355392456, + "learning_rate": 6.211749763100108e-06, + "loss": 0.541, + "step": 14937 + }, + { + "epoch": 0.6613904785096798, + "grad_norm": 1.4142967462539673, + "learning_rate": 6.210319542476468e-06, + "loss": 0.5252, + "step": 14938 + }, + { + "epoch": 0.6614347542144936, + "grad_norm": 1.479924201965332, + "learning_rate": 6.208889412366258e-06, + "loss": 0.5312, + "step": 14939 + }, + { + "epoch": 0.6614790299193075, + "grad_norm": 1.701209545135498, + "learning_rate": 6.207459372803639e-06, + "loss": 0.5373, + "step": 14940 + }, + { + "epoch": 0.6615233056241214, + "grad_norm": 1.5799999237060547, + "learning_rate": 6.2060294238227655e-06, + "loss": 0.5789, + "step": 14941 + }, + { + "epoch": 0.6615675813289352, + "grad_norm": 1.7355401515960693, + "learning_rate": 6.20459956545779e-06, + "loss": 0.5749, + "step": 14942 + }, + { + "epoch": 0.6616118570337491, + "grad_norm": 1.282417893409729, + "learning_rate": 6.203169797742862e-06, + "loss": 0.5193, + "step": 14943 + }, + { + "epoch": 0.661656132738563, + "grad_norm": 1.592777132987976, + "learning_rate": 6.201740120712133e-06, + "loss": 0.6061, + "step": 14944 + }, + { + "epoch": 0.6617004084433769, + "grad_norm": 1.404665470123291, + "learning_rate": 6.20031053439975e-06, + "loss": 0.5907, + "step": 14945 + }, + { + "epoch": 0.6617446841481908, + "grad_norm": 1.592754602432251, + "learning_rate": 6.19888103883985e-06, + "loss": 0.5983, + "step": 14946 + }, + { + "epoch": 0.6617889598530047, + "grad_norm": 1.5749974250793457, + "learning_rate": 6.197451634066587e-06, + "loss": 0.5575, + "step": 14947 + }, + { + "epoch": 0.6618332355578186, + "grad_norm": 1.35866117477417, + "learning_rate": 6.196022320114097e-06, + "loss": 0.5108, + "step": 14948 + }, + { + "epoch": 0.6618775112626324, + "grad_norm": 1.3021049499511719, + "learning_rate": 6.19459309701651e-06, + "loss": 0.578, + "step": 14949 + }, + { + "epoch": 0.6619217869674463, + "grad_norm": 1.5418519973754883, + "learning_rate": 6.193163964807973e-06, + "loss": 0.5181, + "step": 14950 + }, + { + "epoch": 0.6619660626722602, + "grad_norm": 1.3600258827209473, + "learning_rate": 6.191734923522614e-06, + "loss": 0.5192, + "step": 14951 + }, + { + "epoch": 0.662010338377074, + "grad_norm": 1.3303248882293701, + "learning_rate": 6.190305973194567e-06, + "loss": 0.5405, + "step": 14952 + }, + { + "epoch": 0.6620546140818879, + "grad_norm": 1.3824130296707153, + "learning_rate": 6.188877113857958e-06, + "loss": 0.5485, + "step": 14953 + }, + { + "epoch": 0.6620988897867018, + "grad_norm": 1.5310697555541992, + "learning_rate": 6.187448345546917e-06, + "loss": 0.5738, + "step": 14954 + }, + { + "epoch": 0.6621431654915156, + "grad_norm": 1.3877207040786743, + "learning_rate": 6.186019668295568e-06, + "loss": 0.5326, + "step": 14955 + }, + { + "epoch": 0.6621874411963296, + "grad_norm": 1.5212244987487793, + "learning_rate": 6.18459108213803e-06, + "loss": 0.5513, + "step": 14956 + }, + { + "epoch": 0.6622317169011435, + "grad_norm": 1.391077995300293, + "learning_rate": 6.183162587108433e-06, + "loss": 0.5068, + "step": 14957 + }, + { + "epoch": 0.6622759926059573, + "grad_norm": 1.775696873664856, + "learning_rate": 6.181734183240888e-06, + "loss": 0.6642, + "step": 14958 + }, + { + "epoch": 0.6623202683107712, + "grad_norm": 1.2956076860427856, + "learning_rate": 6.1803058705695105e-06, + "loss": 0.5372, + "step": 14959 + }, + { + "epoch": 0.6623645440155851, + "grad_norm": 1.376723051071167, + "learning_rate": 6.178877649128421e-06, + "loss": 0.5443, + "step": 14960 + }, + { + "epoch": 0.6624088197203989, + "grad_norm": 1.8822218179702759, + "learning_rate": 6.177449518951726e-06, + "loss": 0.5923, + "step": 14961 + }, + { + "epoch": 0.6624530954252128, + "grad_norm": 1.623343825340271, + "learning_rate": 6.1760214800735375e-06, + "loss": 0.5246, + "step": 14962 + }, + { + "epoch": 0.6624973711300267, + "grad_norm": 1.5396947860717773, + "learning_rate": 6.174593532527959e-06, + "loss": 0.5421, + "step": 14963 + }, + { + "epoch": 0.6625416468348405, + "grad_norm": 1.4146758317947388, + "learning_rate": 6.173165676349103e-06, + "loss": 0.5477, + "step": 14964 + }, + { + "epoch": 0.6625859225396544, + "grad_norm": 1.356776237487793, + "learning_rate": 6.171737911571069e-06, + "loss": 0.5547, + "step": 14965 + }, + { + "epoch": 0.6626301982444683, + "grad_norm": 1.4869298934936523, + "learning_rate": 6.1703102382279545e-06, + "loss": 0.5678, + "step": 14966 + }, + { + "epoch": 0.6626744739492821, + "grad_norm": 1.6058748960494995, + "learning_rate": 6.168882656353865e-06, + "loss": 0.6028, + "step": 14967 + }, + { + "epoch": 0.662718749654096, + "grad_norm": 1.4539422988891602, + "learning_rate": 6.167455165982894e-06, + "loss": 0.5925, + "step": 14968 + }, + { + "epoch": 0.66276302535891, + "grad_norm": 1.4525612592697144, + "learning_rate": 6.166027767149134e-06, + "loss": 0.5407, + "step": 14969 + }, + { + "epoch": 0.6628073010637238, + "grad_norm": 1.4188324213027954, + "learning_rate": 6.164600459886677e-06, + "loss": 0.5414, + "step": 14970 + }, + { + "epoch": 0.6628515767685377, + "grad_norm": 1.4024616479873657, + "learning_rate": 6.163173244229618e-06, + "loss": 0.5684, + "step": 14971 + }, + { + "epoch": 0.6628958524733516, + "grad_norm": 1.2430979013442993, + "learning_rate": 6.161746120212044e-06, + "loss": 0.5506, + "step": 14972 + }, + { + "epoch": 0.6629401281781654, + "grad_norm": 2.021353244781494, + "learning_rate": 6.160319087868034e-06, + "loss": 0.5851, + "step": 14973 + }, + { + "epoch": 0.6629844038829793, + "grad_norm": 1.3834236860275269, + "learning_rate": 6.158892147231679e-06, + "loss": 0.5082, + "step": 14974 + }, + { + "epoch": 0.6630286795877932, + "grad_norm": 1.6345710754394531, + "learning_rate": 6.157465298337057e-06, + "loss": 0.563, + "step": 14975 + }, + { + "epoch": 0.663072955292607, + "grad_norm": 1.5470530986785889, + "learning_rate": 6.1560385412182465e-06, + "loss": 0.5534, + "step": 14976 + }, + { + "epoch": 0.6631172309974209, + "grad_norm": 1.6129820346832275, + "learning_rate": 6.1546118759093275e-06, + "loss": 0.568, + "step": 14977 + }, + { + "epoch": 0.6631615067022348, + "grad_norm": 1.4459137916564941, + "learning_rate": 6.153185302444373e-06, + "loss": 0.5664, + "step": 14978 + }, + { + "epoch": 0.6632057824070486, + "grad_norm": 1.3448699712753296, + "learning_rate": 6.151758820857455e-06, + "loss": 0.5635, + "step": 14979 + }, + { + "epoch": 0.6632500581118626, + "grad_norm": 1.8126373291015625, + "learning_rate": 6.1503324311826404e-06, + "loss": 0.5798, + "step": 14980 + }, + { + "epoch": 0.6632943338166765, + "grad_norm": 1.329362154006958, + "learning_rate": 6.148906133454007e-06, + "loss": 0.5392, + "step": 14981 + }, + { + "epoch": 0.6633386095214904, + "grad_norm": 1.7422497272491455, + "learning_rate": 6.147479927705613e-06, + "loss": 0.5855, + "step": 14982 + }, + { + "epoch": 0.6633828852263042, + "grad_norm": 1.4514484405517578, + "learning_rate": 6.14605381397152e-06, + "loss": 0.573, + "step": 14983 + }, + { + "epoch": 0.6634271609311181, + "grad_norm": 1.6608151197433472, + "learning_rate": 6.144627792285798e-06, + "loss": 0.558, + "step": 14984 + }, + { + "epoch": 0.663471436635932, + "grad_norm": 1.4355708360671997, + "learning_rate": 6.143201862682502e-06, + "loss": 0.4877, + "step": 14985 + }, + { + "epoch": 0.6635157123407458, + "grad_norm": 1.4679007530212402, + "learning_rate": 6.1417760251956874e-06, + "loss": 0.5567, + "step": 14986 + }, + { + "epoch": 0.6635599880455597, + "grad_norm": 1.6197255849838257, + "learning_rate": 6.140350279859413e-06, + "loss": 0.5769, + "step": 14987 + }, + { + "epoch": 0.6636042637503736, + "grad_norm": 1.7492338418960571, + "learning_rate": 6.138924626707727e-06, + "loss": 0.6419, + "step": 14988 + }, + { + "epoch": 0.6636485394551874, + "grad_norm": 1.542171597480774, + "learning_rate": 6.137499065774686e-06, + "loss": 0.5795, + "step": 14989 + }, + { + "epoch": 0.6636928151600013, + "grad_norm": 1.4735996723175049, + "learning_rate": 6.136073597094331e-06, + "loss": 0.5586, + "step": 14990 + }, + { + "epoch": 0.6637370908648152, + "grad_norm": 1.4272624254226685, + "learning_rate": 6.1346482207007145e-06, + "loss": 0.5774, + "step": 14991 + }, + { + "epoch": 0.663781366569629, + "grad_norm": 1.3499045372009277, + "learning_rate": 6.133222936627879e-06, + "loss": 0.5832, + "step": 14992 + }, + { + "epoch": 0.663825642274443, + "grad_norm": 1.5495792627334595, + "learning_rate": 6.131797744909859e-06, + "loss": 0.5896, + "step": 14993 + }, + { + "epoch": 0.6638699179792569, + "grad_norm": 1.769442081451416, + "learning_rate": 6.130372645580704e-06, + "loss": 0.6087, + "step": 14994 + }, + { + "epoch": 0.6639141936840707, + "grad_norm": 1.5633403062820435, + "learning_rate": 6.128947638674449e-06, + "loss": 0.5322, + "step": 14995 + }, + { + "epoch": 0.6639584693888846, + "grad_norm": 1.467334270477295, + "learning_rate": 6.127522724225127e-06, + "loss": 0.5904, + "step": 14996 + }, + { + "epoch": 0.6640027450936985, + "grad_norm": 1.4059829711914062, + "learning_rate": 6.126097902266772e-06, + "loss": 0.5605, + "step": 14997 + }, + { + "epoch": 0.6640470207985123, + "grad_norm": 1.599859356880188, + "learning_rate": 6.124673172833415e-06, + "loss": 0.5563, + "step": 14998 + }, + { + "epoch": 0.6640912965033262, + "grad_norm": 1.5219777822494507, + "learning_rate": 6.123248535959083e-06, + "loss": 0.5665, + "step": 14999 + }, + { + "epoch": 0.6641355722081401, + "grad_norm": 1.6545122861862183, + "learning_rate": 6.121823991677802e-06, + "loss": 0.6129, + "step": 15000 + }, + { + "epoch": 0.6641798479129539, + "grad_norm": 1.4104931354522705, + "learning_rate": 6.120399540023599e-06, + "loss": 0.5999, + "step": 15001 + }, + { + "epoch": 0.6642241236177678, + "grad_norm": 1.6321048736572266, + "learning_rate": 6.118975181030493e-06, + "loss": 0.6286, + "step": 15002 + }, + { + "epoch": 0.6642683993225817, + "grad_norm": 1.4897464513778687, + "learning_rate": 6.1175509147325015e-06, + "loss": 0.5102, + "step": 15003 + }, + { + "epoch": 0.6643126750273955, + "grad_norm": 1.6264410018920898, + "learning_rate": 6.116126741163653e-06, + "loss": 0.6016, + "step": 15004 + }, + { + "epoch": 0.6643569507322095, + "grad_norm": 1.4490127563476562, + "learning_rate": 6.114702660357951e-06, + "loss": 0.5281, + "step": 15005 + }, + { + "epoch": 0.6644012264370234, + "grad_norm": 1.5686899423599243, + "learning_rate": 6.113278672349407e-06, + "loss": 0.5461, + "step": 15006 + }, + { + "epoch": 0.6644455021418372, + "grad_norm": 1.644210934638977, + "learning_rate": 6.111854777172044e-06, + "loss": 0.5528, + "step": 15007 + }, + { + "epoch": 0.6644897778466511, + "grad_norm": 1.2996469736099243, + "learning_rate": 6.110430974859862e-06, + "loss": 0.5306, + "step": 15008 + }, + { + "epoch": 0.664534053551465, + "grad_norm": 1.676393985748291, + "learning_rate": 6.1090072654468716e-06, + "loss": 0.6446, + "step": 15009 + }, + { + "epoch": 0.6645783292562788, + "grad_norm": 1.566981554031372, + "learning_rate": 6.10758364896707e-06, + "loss": 0.504, + "step": 15010 + }, + { + "epoch": 0.6646226049610927, + "grad_norm": 1.6145461797714233, + "learning_rate": 6.106160125454466e-06, + "loss": 0.5964, + "step": 15011 + }, + { + "epoch": 0.6646668806659066, + "grad_norm": 1.7479143142700195, + "learning_rate": 6.104736694943059e-06, + "loss": 0.5651, + "step": 15012 + }, + { + "epoch": 0.6647111563707204, + "grad_norm": 1.5633728504180908, + "learning_rate": 6.10331335746684e-06, + "loss": 0.6269, + "step": 15013 + }, + { + "epoch": 0.6647554320755343, + "grad_norm": 1.328465223312378, + "learning_rate": 6.1018901130598115e-06, + "loss": 0.5495, + "step": 15014 + }, + { + "epoch": 0.6647997077803482, + "grad_norm": 1.3768631219863892, + "learning_rate": 6.100466961755964e-06, + "loss": 0.5024, + "step": 15015 + }, + { + "epoch": 0.6648439834851622, + "grad_norm": 1.4049642086029053, + "learning_rate": 6.0990439035892835e-06, + "loss": 0.5384, + "step": 15016 + }, + { + "epoch": 0.664888259189976, + "grad_norm": 1.5645381212234497, + "learning_rate": 6.097620938593767e-06, + "loss": 0.5713, + "step": 15017 + }, + { + "epoch": 0.6649325348947899, + "grad_norm": 1.4138526916503906, + "learning_rate": 6.096198066803397e-06, + "loss": 0.5652, + "step": 15018 + }, + { + "epoch": 0.6649768105996038, + "grad_norm": 1.5322370529174805, + "learning_rate": 6.094775288252157e-06, + "loss": 0.5688, + "step": 15019 + }, + { + "epoch": 0.6650210863044176, + "grad_norm": 1.9525701999664307, + "learning_rate": 6.09335260297403e-06, + "loss": 0.5887, + "step": 15020 + }, + { + "epoch": 0.6650653620092315, + "grad_norm": 1.5543627738952637, + "learning_rate": 6.091930011002994e-06, + "loss": 0.5641, + "step": 15021 + }, + { + "epoch": 0.6651096377140454, + "grad_norm": 1.3599425554275513, + "learning_rate": 6.09050751237303e-06, + "loss": 0.4861, + "step": 15022 + }, + { + "epoch": 0.6651539134188592, + "grad_norm": 1.6747008562088013, + "learning_rate": 6.089085107118107e-06, + "loss": 0.5995, + "step": 15023 + }, + { + "epoch": 0.6651981891236731, + "grad_norm": 1.2186927795410156, + "learning_rate": 6.087662795272204e-06, + "loss": 0.4955, + "step": 15024 + }, + { + "epoch": 0.665242464828487, + "grad_norm": 1.3941946029663086, + "learning_rate": 6.086240576869291e-06, + "loss": 0.5214, + "step": 15025 + }, + { + "epoch": 0.6652867405333008, + "grad_norm": 1.5123244524002075, + "learning_rate": 6.08481845194333e-06, + "loss": 0.5792, + "step": 15026 + }, + { + "epoch": 0.6653310162381147, + "grad_norm": 1.3066481351852417, + "learning_rate": 6.083396420528298e-06, + "loss": 0.5347, + "step": 15027 + }, + { + "epoch": 0.6653752919429287, + "grad_norm": 1.599711298942566, + "learning_rate": 6.081974482658154e-06, + "loss": 0.5944, + "step": 15028 + }, + { + "epoch": 0.6654195676477425, + "grad_norm": 1.6751173734664917, + "learning_rate": 6.080552638366858e-06, + "loss": 0.6011, + "step": 15029 + }, + { + "epoch": 0.6654638433525564, + "grad_norm": 1.336474895477295, + "learning_rate": 6.079130887688367e-06, + "loss": 0.5201, + "step": 15030 + }, + { + "epoch": 0.6655081190573703, + "grad_norm": 1.666589379310608, + "learning_rate": 6.0777092306566454e-06, + "loss": 0.6036, + "step": 15031 + }, + { + "epoch": 0.6655523947621841, + "grad_norm": 1.5514914989471436, + "learning_rate": 6.076287667305645e-06, + "loss": 0.5576, + "step": 15032 + }, + { + "epoch": 0.665596670466998, + "grad_norm": 1.383111596107483, + "learning_rate": 6.074866197669318e-06, + "loss": 0.5568, + "step": 15033 + }, + { + "epoch": 0.6656409461718119, + "grad_norm": 1.2803502082824707, + "learning_rate": 6.073444821781617e-06, + "loss": 0.512, + "step": 15034 + }, + { + "epoch": 0.6656852218766257, + "grad_norm": 1.2707055807113647, + "learning_rate": 6.072023539676489e-06, + "loss": 0.4994, + "step": 15035 + }, + { + "epoch": 0.6657294975814396, + "grad_norm": 1.3571929931640625, + "learning_rate": 6.07060235138788e-06, + "loss": 0.5035, + "step": 15036 + }, + { + "epoch": 0.6657737732862535, + "grad_norm": 1.5504980087280273, + "learning_rate": 6.069181256949734e-06, + "loss": 0.6308, + "step": 15037 + }, + { + "epoch": 0.6658180489910673, + "grad_norm": 1.4525485038757324, + "learning_rate": 6.067760256395996e-06, + "loss": 0.591, + "step": 15038 + }, + { + "epoch": 0.6658623246958812, + "grad_norm": 1.8279482126235962, + "learning_rate": 6.066339349760602e-06, + "loss": 0.6514, + "step": 15039 + }, + { + "epoch": 0.6659066004006952, + "grad_norm": 1.2818961143493652, + "learning_rate": 6.064918537077484e-06, + "loss": 0.4865, + "step": 15040 + }, + { + "epoch": 0.665950876105509, + "grad_norm": 1.7356617450714111, + "learning_rate": 6.063497818380587e-06, + "loss": 0.6098, + "step": 15041 + }, + { + "epoch": 0.6659951518103229, + "grad_norm": 1.407783031463623, + "learning_rate": 6.06207719370384e-06, + "loss": 0.6008, + "step": 15042 + }, + { + "epoch": 0.6660394275151368, + "grad_norm": 1.3966344594955444, + "learning_rate": 6.060656663081171e-06, + "loss": 0.5299, + "step": 15043 + }, + { + "epoch": 0.6660837032199506, + "grad_norm": 1.4133449792861938, + "learning_rate": 6.059236226546512e-06, + "loss": 0.5884, + "step": 15044 + }, + { + "epoch": 0.6661279789247645, + "grad_norm": 1.6183050870895386, + "learning_rate": 6.057815884133787e-06, + "loss": 0.566, + "step": 15045 + }, + { + "epoch": 0.6661722546295784, + "grad_norm": 1.6108181476593018, + "learning_rate": 6.056395635876917e-06, + "loss": 0.5754, + "step": 15046 + }, + { + "epoch": 0.6662165303343922, + "grad_norm": 1.359800100326538, + "learning_rate": 6.0549754818098295e-06, + "loss": 0.5051, + "step": 15047 + }, + { + "epoch": 0.6662608060392061, + "grad_norm": 1.4751476049423218, + "learning_rate": 6.053555421966439e-06, + "loss": 0.5653, + "step": 15048 + }, + { + "epoch": 0.66630508174402, + "grad_norm": 1.3896501064300537, + "learning_rate": 6.052135456380666e-06, + "loss": 0.5369, + "step": 15049 + }, + { + "epoch": 0.6663493574488338, + "grad_norm": 1.452993392944336, + "learning_rate": 6.050715585086417e-06, + "loss": 0.5435, + "step": 15050 + }, + { + "epoch": 0.6663936331536477, + "grad_norm": 1.4275636672973633, + "learning_rate": 6.0492958081176155e-06, + "loss": 0.5987, + "step": 15051 + }, + { + "epoch": 0.6664379088584617, + "grad_norm": 2.3099794387817383, + "learning_rate": 6.04787612550817e-06, + "loss": 0.6197, + "step": 15052 + }, + { + "epoch": 0.6664821845632756, + "grad_norm": 1.7352896928787231, + "learning_rate": 6.046456537291978e-06, + "loss": 0.6151, + "step": 15053 + }, + { + "epoch": 0.6665264602680894, + "grad_norm": 1.1928513050079346, + "learning_rate": 6.045037043502956e-06, + "loss": 0.5453, + "step": 15054 + }, + { + "epoch": 0.6665707359729033, + "grad_norm": 1.8506969213485718, + "learning_rate": 6.043617644175005e-06, + "loss": 0.6211, + "step": 15055 + }, + { + "epoch": 0.6666150116777172, + "grad_norm": 1.3767731189727783, + "learning_rate": 6.042198339342022e-06, + "loss": 0.5449, + "step": 15056 + }, + { + "epoch": 0.666659287382531, + "grad_norm": 1.3474560976028442, + "learning_rate": 6.040779129037912e-06, + "loss": 0.5758, + "step": 15057 + }, + { + "epoch": 0.6667035630873449, + "grad_norm": 1.2595596313476562, + "learning_rate": 6.039360013296569e-06, + "loss": 0.4899, + "step": 15058 + }, + { + "epoch": 0.6667478387921588, + "grad_norm": 1.4840055704116821, + "learning_rate": 6.037940992151887e-06, + "loss": 0.5654, + "step": 15059 + }, + { + "epoch": 0.6667921144969726, + "grad_norm": 1.5047091245651245, + "learning_rate": 6.036522065637757e-06, + "loss": 0.5803, + "step": 15060 + }, + { + "epoch": 0.6668363902017865, + "grad_norm": 1.8687341213226318, + "learning_rate": 6.035103233788071e-06, + "loss": 0.6071, + "step": 15061 + }, + { + "epoch": 0.6668806659066004, + "grad_norm": 1.3918404579162598, + "learning_rate": 6.0336844966367155e-06, + "loss": 0.5685, + "step": 15062 + }, + { + "epoch": 0.6669249416114142, + "grad_norm": 1.3158055543899536, + "learning_rate": 6.032265854217574e-06, + "loss": 0.5378, + "step": 15063 + }, + { + "epoch": 0.6669692173162282, + "grad_norm": 1.554380178451538, + "learning_rate": 6.030847306564536e-06, + "loss": 0.5437, + "step": 15064 + }, + { + "epoch": 0.6670134930210421, + "grad_norm": 1.3568040132522583, + "learning_rate": 6.029428853711476e-06, + "loss": 0.5767, + "step": 15065 + }, + { + "epoch": 0.6670577687258559, + "grad_norm": 1.6061744689941406, + "learning_rate": 6.028010495692276e-06, + "loss": 0.6235, + "step": 15066 + }, + { + "epoch": 0.6671020444306698, + "grad_norm": 1.7293288707733154, + "learning_rate": 6.026592232540811e-06, + "loss": 0.6428, + "step": 15067 + }, + { + "epoch": 0.6671463201354837, + "grad_norm": 1.5386192798614502, + "learning_rate": 6.025174064290956e-06, + "loss": 0.6123, + "step": 15068 + }, + { + "epoch": 0.6671905958402975, + "grad_norm": 1.7472199201583862, + "learning_rate": 6.023755990976582e-06, + "loss": 0.6312, + "step": 15069 + }, + { + "epoch": 0.6672348715451114, + "grad_norm": 1.7952029705047607, + "learning_rate": 6.022338012631556e-06, + "loss": 0.6449, + "step": 15070 + }, + { + "epoch": 0.6672791472499253, + "grad_norm": 1.2975610494613647, + "learning_rate": 6.020920129289751e-06, + "loss": 0.4693, + "step": 15071 + }, + { + "epoch": 0.6673234229547391, + "grad_norm": 1.3561360836029053, + "learning_rate": 6.019502340985028e-06, + "loss": 0.5313, + "step": 15072 + }, + { + "epoch": 0.667367698659553, + "grad_norm": 1.5180612802505493, + "learning_rate": 6.018084647751247e-06, + "loss": 0.6129, + "step": 15073 + }, + { + "epoch": 0.6674119743643669, + "grad_norm": 1.3271937370300293, + "learning_rate": 6.016667049622276e-06, + "loss": 0.5157, + "step": 15074 + }, + { + "epoch": 0.6674562500691807, + "grad_norm": 1.7311723232269287, + "learning_rate": 6.015249546631969e-06, + "loss": 0.5711, + "step": 15075 + }, + { + "epoch": 0.6675005257739947, + "grad_norm": 1.5985543727874756, + "learning_rate": 6.013832138814183e-06, + "loss": 0.595, + "step": 15076 + }, + { + "epoch": 0.6675448014788086, + "grad_norm": 1.4449108839035034, + "learning_rate": 6.012414826202766e-06, + "loss": 0.4746, + "step": 15077 + }, + { + "epoch": 0.6675890771836224, + "grad_norm": 1.3731467723846436, + "learning_rate": 6.010997608831577e-06, + "loss": 0.5478, + "step": 15078 + }, + { + "epoch": 0.6676333528884363, + "grad_norm": 1.5427714586257935, + "learning_rate": 6.009580486734463e-06, + "loss": 0.5831, + "step": 15079 + }, + { + "epoch": 0.6676776285932502, + "grad_norm": 1.469681978225708, + "learning_rate": 6.008163459945268e-06, + "loss": 0.5845, + "step": 15080 + }, + { + "epoch": 0.667721904298064, + "grad_norm": 1.8106974363327026, + "learning_rate": 6.00674652849784e-06, + "loss": 0.5491, + "step": 15081 + }, + { + "epoch": 0.6677661800028779, + "grad_norm": 1.678726315498352, + "learning_rate": 6.005329692426021e-06, + "loss": 0.603, + "step": 15082 + }, + { + "epoch": 0.6678104557076918, + "grad_norm": 1.1568403244018555, + "learning_rate": 6.003912951763644e-06, + "loss": 0.5491, + "step": 15083 + }, + { + "epoch": 0.6678547314125056, + "grad_norm": 1.441680669784546, + "learning_rate": 6.002496306544559e-06, + "loss": 0.5593, + "step": 15084 + }, + { + "epoch": 0.6678990071173195, + "grad_norm": 1.3605234622955322, + "learning_rate": 6.001079756802592e-06, + "loss": 0.5551, + "step": 15085 + }, + { + "epoch": 0.6679432828221334, + "grad_norm": 1.4056583642959595, + "learning_rate": 5.999663302571581e-06, + "loss": 0.5595, + "step": 15086 + }, + { + "epoch": 0.6679875585269474, + "grad_norm": 1.6742500066757202, + "learning_rate": 5.9982469438853495e-06, + "loss": 0.5757, + "step": 15087 + }, + { + "epoch": 0.6680318342317612, + "grad_norm": 1.5689196586608887, + "learning_rate": 5.996830680777736e-06, + "loss": 0.6087, + "step": 15088 + }, + { + "epoch": 0.6680761099365751, + "grad_norm": 2.028470039367676, + "learning_rate": 5.9954145132825616e-06, + "loss": 0.6219, + "step": 15089 + }, + { + "epoch": 0.668120385641389, + "grad_norm": 1.4214179515838623, + "learning_rate": 5.993998441433648e-06, + "loss": 0.5259, + "step": 15090 + }, + { + "epoch": 0.6681646613462028, + "grad_norm": 1.6850980520248413, + "learning_rate": 5.992582465264824e-06, + "loss": 0.5795, + "step": 15091 + }, + { + "epoch": 0.6682089370510167, + "grad_norm": 1.6409673690795898, + "learning_rate": 5.991166584809904e-06, + "loss": 0.5655, + "step": 15092 + }, + { + "epoch": 0.6682532127558306, + "grad_norm": 1.714451789855957, + "learning_rate": 5.989750800102704e-06, + "loss": 0.5824, + "step": 15093 + }, + { + "epoch": 0.6682974884606444, + "grad_norm": 1.5930758714675903, + "learning_rate": 5.988335111177042e-06, + "loss": 0.5663, + "step": 15094 + }, + { + "epoch": 0.6683417641654583, + "grad_norm": 1.4041249752044678, + "learning_rate": 5.98691951806673e-06, + "loss": 0.5493, + "step": 15095 + }, + { + "epoch": 0.6683860398702722, + "grad_norm": 1.3267848491668701, + "learning_rate": 5.985504020805578e-06, + "loss": 0.5256, + "step": 15096 + }, + { + "epoch": 0.668430315575086, + "grad_norm": 1.5776311159133911, + "learning_rate": 5.98408861942739e-06, + "loss": 0.618, + "step": 15097 + }, + { + "epoch": 0.6684745912798999, + "grad_norm": 1.4303884506225586, + "learning_rate": 5.982673313965979e-06, + "loss": 0.5727, + "step": 15098 + }, + { + "epoch": 0.6685188669847139, + "grad_norm": 1.8221546411514282, + "learning_rate": 5.9812581044551475e-06, + "loss": 0.5184, + "step": 15099 + }, + { + "epoch": 0.6685631426895277, + "grad_norm": 1.5941426753997803, + "learning_rate": 5.979842990928687e-06, + "loss": 0.6181, + "step": 15100 + }, + { + "epoch": 0.6686074183943416, + "grad_norm": 1.7204521894454956, + "learning_rate": 5.978427973420408e-06, + "loss": 0.619, + "step": 15101 + }, + { + "epoch": 0.6686516940991555, + "grad_norm": 1.736650824546814, + "learning_rate": 5.977013051964102e-06, + "loss": 0.5608, + "step": 15102 + }, + { + "epoch": 0.6686959698039693, + "grad_norm": 1.6278702020645142, + "learning_rate": 5.975598226593562e-06, + "loss": 0.5466, + "step": 15103 + }, + { + "epoch": 0.6687402455087832, + "grad_norm": 1.2956211566925049, + "learning_rate": 5.9741834973425826e-06, + "loss": 0.5211, + "step": 15104 + }, + { + "epoch": 0.6687845212135971, + "grad_norm": 1.691956639289856, + "learning_rate": 5.972768864244953e-06, + "loss": 0.6105, + "step": 15105 + }, + { + "epoch": 0.6688287969184109, + "grad_norm": 1.475021481513977, + "learning_rate": 5.9713543273344596e-06, + "loss": 0.541, + "step": 15106 + }, + { + "epoch": 0.6688730726232248, + "grad_norm": 1.4621200561523438, + "learning_rate": 5.9699398866448846e-06, + "loss": 0.5861, + "step": 15107 + }, + { + "epoch": 0.6689173483280387, + "grad_norm": 1.4242380857467651, + "learning_rate": 5.9685255422100195e-06, + "loss": 0.5345, + "step": 15108 + }, + { + "epoch": 0.6689616240328525, + "grad_norm": 1.7099944353103638, + "learning_rate": 5.967111294063639e-06, + "loss": 0.6419, + "step": 15109 + }, + { + "epoch": 0.6690058997376664, + "grad_norm": 1.4126304388046265, + "learning_rate": 5.965697142239515e-06, + "loss": 0.5406, + "step": 15110 + }, + { + "epoch": 0.6690501754424804, + "grad_norm": 1.3745225667953491, + "learning_rate": 5.964283086771435e-06, + "loss": 0.576, + "step": 15111 + }, + { + "epoch": 0.6690944511472942, + "grad_norm": 1.394494891166687, + "learning_rate": 5.962869127693167e-06, + "loss": 0.5308, + "step": 15112 + }, + { + "epoch": 0.6691387268521081, + "grad_norm": 1.6122153997421265, + "learning_rate": 5.961455265038483e-06, + "loss": 0.5206, + "step": 15113 + }, + { + "epoch": 0.669183002556922, + "grad_norm": 1.8352241516113281, + "learning_rate": 5.960041498841151e-06, + "loss": 0.682, + "step": 15114 + }, + { + "epoch": 0.6692272782617358, + "grad_norm": 1.509008526802063, + "learning_rate": 5.95862782913494e-06, + "loss": 0.5951, + "step": 15115 + }, + { + "epoch": 0.6692715539665497, + "grad_norm": 1.4792755842208862, + "learning_rate": 5.957214255953612e-06, + "loss": 0.5397, + "step": 15116 + }, + { + "epoch": 0.6693158296713636, + "grad_norm": 1.3863353729248047, + "learning_rate": 5.955800779330927e-06, + "loss": 0.5358, + "step": 15117 + }, + { + "epoch": 0.6693601053761774, + "grad_norm": 1.3883895874023438, + "learning_rate": 5.95438739930065e-06, + "loss": 0.5763, + "step": 15118 + }, + { + "epoch": 0.6694043810809913, + "grad_norm": 1.6769638061523438, + "learning_rate": 5.952974115896536e-06, + "loss": 0.6196, + "step": 15119 + }, + { + "epoch": 0.6694486567858052, + "grad_norm": 1.3047255277633667, + "learning_rate": 5.9515609291523354e-06, + "loss": 0.5464, + "step": 15120 + }, + { + "epoch": 0.669492932490619, + "grad_norm": 1.6382755041122437, + "learning_rate": 5.95014783910181e-06, + "loss": 0.6409, + "step": 15121 + }, + { + "epoch": 0.6695372081954329, + "grad_norm": 1.5361336469650269, + "learning_rate": 5.9487348457787074e-06, + "loss": 0.5635, + "step": 15122 + }, + { + "epoch": 0.6695814839002469, + "grad_norm": 1.454388976097107, + "learning_rate": 5.947321949216771e-06, + "loss": 0.5448, + "step": 15123 + }, + { + "epoch": 0.6696257596050608, + "grad_norm": 1.437341570854187, + "learning_rate": 5.945909149449753e-06, + "loss": 0.494, + "step": 15124 + }, + { + "epoch": 0.6696700353098746, + "grad_norm": 1.2873882055282593, + "learning_rate": 5.9444964465113915e-06, + "loss": 0.5282, + "step": 15125 + }, + { + "epoch": 0.6697143110146885, + "grad_norm": 1.3822135925292969, + "learning_rate": 5.943083840435433e-06, + "loss": 0.5236, + "step": 15126 + }, + { + "epoch": 0.6697585867195024, + "grad_norm": 1.4623596668243408, + "learning_rate": 5.9416713312556105e-06, + "loss": 0.5054, + "step": 15127 + }, + { + "epoch": 0.6698028624243162, + "grad_norm": 1.8380801677703857, + "learning_rate": 5.940258919005666e-06, + "loss": 0.6121, + "step": 15128 + }, + { + "epoch": 0.6698471381291301, + "grad_norm": 1.594805121421814, + "learning_rate": 5.938846603719333e-06, + "loss": 0.5925, + "step": 15129 + }, + { + "epoch": 0.669891413833944, + "grad_norm": 1.556962490081787, + "learning_rate": 5.937434385430337e-06, + "loss": 0.5917, + "step": 15130 + }, + { + "epoch": 0.6699356895387578, + "grad_norm": 1.5555680990219116, + "learning_rate": 5.936022264172418e-06, + "loss": 0.5655, + "step": 15131 + }, + { + "epoch": 0.6699799652435717, + "grad_norm": 1.3195167779922485, + "learning_rate": 5.9346102399792974e-06, + "loss": 0.5201, + "step": 15132 + }, + { + "epoch": 0.6700242409483856, + "grad_norm": 2.049401044845581, + "learning_rate": 5.933198312884696e-06, + "loss": 0.6264, + "step": 15133 + }, + { + "epoch": 0.6700685166531994, + "grad_norm": 1.4427080154418945, + "learning_rate": 5.931786482922346e-06, + "loss": 0.5636, + "step": 15134 + }, + { + "epoch": 0.6701127923580134, + "grad_norm": 2.087480306625366, + "learning_rate": 5.930374750125963e-06, + "loss": 0.5998, + "step": 15135 + }, + { + "epoch": 0.6701570680628273, + "grad_norm": 2.0160975456237793, + "learning_rate": 5.928963114529267e-06, + "loss": 0.5985, + "step": 15136 + }, + { + "epoch": 0.6702013437676411, + "grad_norm": 1.386653184890747, + "learning_rate": 5.927551576165969e-06, + "loss": 0.5922, + "step": 15137 + }, + { + "epoch": 0.670245619472455, + "grad_norm": 1.7906525135040283, + "learning_rate": 5.926140135069788e-06, + "loss": 0.6647, + "step": 15138 + }, + { + "epoch": 0.6702898951772689, + "grad_norm": 1.4827725887298584, + "learning_rate": 5.924728791274432e-06, + "loss": 0.5261, + "step": 15139 + }, + { + "epoch": 0.6703341708820827, + "grad_norm": 1.7601649761199951, + "learning_rate": 5.92331754481361e-06, + "loss": 0.5915, + "step": 15140 + }, + { + "epoch": 0.6703784465868966, + "grad_norm": 1.2677634954452515, + "learning_rate": 5.921906395721031e-06, + "loss": 0.4884, + "step": 15141 + }, + { + "epoch": 0.6704227222917105, + "grad_norm": 1.527854323387146, + "learning_rate": 5.920495344030397e-06, + "loss": 0.5108, + "step": 15142 + }, + { + "epoch": 0.6704669979965243, + "grad_norm": 1.6923466920852661, + "learning_rate": 5.919084389775404e-06, + "loss": 0.5843, + "step": 15143 + }, + { + "epoch": 0.6705112737013382, + "grad_norm": 1.618701457977295, + "learning_rate": 5.917673532989765e-06, + "loss": 0.6392, + "step": 15144 + }, + { + "epoch": 0.6705555494061521, + "grad_norm": 1.4322707653045654, + "learning_rate": 5.916262773707167e-06, + "loss": 0.5169, + "step": 15145 + }, + { + "epoch": 0.6705998251109659, + "grad_norm": 1.4987642765045166, + "learning_rate": 5.914852111961309e-06, + "loss": 0.5535, + "step": 15146 + }, + { + "epoch": 0.6706441008157799, + "grad_norm": 1.4610273838043213, + "learning_rate": 5.913441547785879e-06, + "loss": 0.5534, + "step": 15147 + }, + { + "epoch": 0.6706883765205938, + "grad_norm": 1.3181931972503662, + "learning_rate": 5.912031081214573e-06, + "loss": 0.5399, + "step": 15148 + }, + { + "epoch": 0.6707326522254076, + "grad_norm": 1.3206037282943726, + "learning_rate": 5.910620712281076e-06, + "loss": 0.558, + "step": 15149 + }, + { + "epoch": 0.6707769279302215, + "grad_norm": 1.410154938697815, + "learning_rate": 5.909210441019071e-06, + "loss": 0.556, + "step": 15150 + }, + { + "epoch": 0.6708212036350354, + "grad_norm": 1.2987322807312012, + "learning_rate": 5.907800267462247e-06, + "loss": 0.4708, + "step": 15151 + }, + { + "epoch": 0.6708654793398492, + "grad_norm": 1.8967047929763794, + "learning_rate": 5.906390191644281e-06, + "loss": 0.6732, + "step": 15152 + }, + { + "epoch": 0.6709097550446631, + "grad_norm": 1.3050512075424194, + "learning_rate": 5.9049802135988475e-06, + "loss": 0.544, + "step": 15153 + }, + { + "epoch": 0.670954030749477, + "grad_norm": 1.377052664756775, + "learning_rate": 5.903570333359632e-06, + "loss": 0.562, + "step": 15154 + }, + { + "epoch": 0.6709983064542908, + "grad_norm": 1.6057846546173096, + "learning_rate": 5.902160550960305e-06, + "loss": 0.5835, + "step": 15155 + }, + { + "epoch": 0.6710425821591047, + "grad_norm": 1.467689871788025, + "learning_rate": 5.900750866434536e-06, + "loss": 0.5994, + "step": 15156 + }, + { + "epoch": 0.6710868578639186, + "grad_norm": 1.3377991914749146, + "learning_rate": 5.89934127981599e-06, + "loss": 0.5412, + "step": 15157 + }, + { + "epoch": 0.6711311335687326, + "grad_norm": 1.5052695274353027, + "learning_rate": 5.897931791138343e-06, + "loss": 0.5692, + "step": 15158 + }, + { + "epoch": 0.6711754092735464, + "grad_norm": 1.3724509477615356, + "learning_rate": 5.896522400435254e-06, + "loss": 0.5077, + "step": 15159 + }, + { + "epoch": 0.6712196849783603, + "grad_norm": 1.47940194606781, + "learning_rate": 5.8951131077403855e-06, + "loss": 0.5522, + "step": 15160 + }, + { + "epoch": 0.6712639606831742, + "grad_norm": 1.7006052732467651, + "learning_rate": 5.8937039130873995e-06, + "loss": 0.619, + "step": 15161 + }, + { + "epoch": 0.671308236387988, + "grad_norm": 1.304555892944336, + "learning_rate": 5.892294816509952e-06, + "loss": 0.5255, + "step": 15162 + }, + { + "epoch": 0.6713525120928019, + "grad_norm": 1.5008994340896606, + "learning_rate": 5.890885818041694e-06, + "loss": 0.6001, + "step": 15163 + }, + { + "epoch": 0.6713967877976158, + "grad_norm": 1.5718480348587036, + "learning_rate": 5.88947691771629e-06, + "loss": 0.5424, + "step": 15164 + }, + { + "epoch": 0.6714410635024296, + "grad_norm": 1.2704708576202393, + "learning_rate": 5.88806811556738e-06, + "loss": 0.4903, + "step": 15165 + }, + { + "epoch": 0.6714853392072435, + "grad_norm": 1.6645983457565308, + "learning_rate": 5.886659411628615e-06, + "loss": 0.5587, + "step": 15166 + }, + { + "epoch": 0.6715296149120574, + "grad_norm": 1.568035364151001, + "learning_rate": 5.885250805933636e-06, + "loss": 0.538, + "step": 15167 + }, + { + "epoch": 0.6715738906168712, + "grad_norm": 1.274755835533142, + "learning_rate": 5.883842298516096e-06, + "loss": 0.4622, + "step": 15168 + }, + { + "epoch": 0.6716181663216851, + "grad_norm": 1.8589776754379272, + "learning_rate": 5.882433889409631e-06, + "loss": 0.6676, + "step": 15169 + }, + { + "epoch": 0.671662442026499, + "grad_norm": 1.4449235200881958, + "learning_rate": 5.881025578647876e-06, + "loss": 0.549, + "step": 15170 + }, + { + "epoch": 0.6717067177313129, + "grad_norm": 1.5228685140609741, + "learning_rate": 5.879617366264476e-06, + "loss": 0.4814, + "step": 15171 + }, + { + "epoch": 0.6717509934361268, + "grad_norm": 1.431426763534546, + "learning_rate": 5.878209252293059e-06, + "loss": 0.523, + "step": 15172 + }, + { + "epoch": 0.6717952691409407, + "grad_norm": 1.7172726392745972, + "learning_rate": 5.876801236767257e-06, + "loss": 0.5947, + "step": 15173 + }, + { + "epoch": 0.6718395448457545, + "grad_norm": 1.5508623123168945, + "learning_rate": 5.875393319720701e-06, + "loss": 0.537, + "step": 15174 + }, + { + "epoch": 0.6718838205505684, + "grad_norm": 1.5009835958480835, + "learning_rate": 5.873985501187019e-06, + "loss": 0.5843, + "step": 15175 + }, + { + "epoch": 0.6719280962553823, + "grad_norm": 1.8321064710617065, + "learning_rate": 5.87257778119983e-06, + "loss": 0.6424, + "step": 15176 + }, + { + "epoch": 0.6719723719601961, + "grad_norm": 1.5142242908477783, + "learning_rate": 5.87117015979276e-06, + "loss": 0.5269, + "step": 15177 + }, + { + "epoch": 0.67201664766501, + "grad_norm": 1.6040335893630981, + "learning_rate": 5.869762636999431e-06, + "loss": 0.5244, + "step": 15178 + }, + { + "epoch": 0.6720609233698239, + "grad_norm": 1.4995884895324707, + "learning_rate": 5.86835521285346e-06, + "loss": 0.501, + "step": 15179 + }, + { + "epoch": 0.6721051990746377, + "grad_norm": 1.4891525506973267, + "learning_rate": 5.866947887388456e-06, + "loss": 0.4665, + "step": 15180 + }, + { + "epoch": 0.6721494747794516, + "grad_norm": 1.4530730247497559, + "learning_rate": 5.865540660638041e-06, + "loss": 0.5224, + "step": 15181 + }, + { + "epoch": 0.6721937504842656, + "grad_norm": 2.1815125942230225, + "learning_rate": 5.864133532635818e-06, + "loss": 0.661, + "step": 15182 + }, + { + "epoch": 0.6722380261890794, + "grad_norm": 1.9882503747940063, + "learning_rate": 5.862726503415399e-06, + "loss": 0.5908, + "step": 15183 + }, + { + "epoch": 0.6722823018938933, + "grad_norm": 1.3851460218429565, + "learning_rate": 5.861319573010389e-06, + "loss": 0.5033, + "step": 15184 + }, + { + "epoch": 0.6723265775987072, + "grad_norm": 1.5901199579238892, + "learning_rate": 5.8599127414543925e-06, + "loss": 0.554, + "step": 15185 + }, + { + "epoch": 0.672370853303521, + "grad_norm": 1.6795538663864136, + "learning_rate": 5.858506008781009e-06, + "loss": 0.6376, + "step": 15186 + }, + { + "epoch": 0.6724151290083349, + "grad_norm": 1.5778751373291016, + "learning_rate": 5.857099375023834e-06, + "loss": 0.5658, + "step": 15187 + }, + { + "epoch": 0.6724594047131488, + "grad_norm": 1.4843294620513916, + "learning_rate": 5.85569284021647e-06, + "loss": 0.5614, + "step": 15188 + }, + { + "epoch": 0.6725036804179626, + "grad_norm": 1.7525882720947266, + "learning_rate": 5.8542864043925085e-06, + "loss": 0.5731, + "step": 15189 + }, + { + "epoch": 0.6725479561227765, + "grad_norm": 1.8635855913162231, + "learning_rate": 5.852880067585535e-06, + "loss": 0.5749, + "step": 15190 + }, + { + "epoch": 0.6725922318275904, + "grad_norm": 1.5193952322006226, + "learning_rate": 5.851473829829151e-06, + "loss": 0.5669, + "step": 15191 + }, + { + "epoch": 0.6726365075324042, + "grad_norm": 1.9280920028686523, + "learning_rate": 5.850067691156936e-06, + "loss": 0.6273, + "step": 15192 + }, + { + "epoch": 0.6726807832372181, + "grad_norm": 1.3331784009933472, + "learning_rate": 5.848661651602474e-06, + "loss": 0.5164, + "step": 15193 + }, + { + "epoch": 0.672725058942032, + "grad_norm": 1.3698341846466064, + "learning_rate": 5.8472557111993465e-06, + "loss": 0.5162, + "step": 15194 + }, + { + "epoch": 0.672769334646846, + "grad_norm": 1.4262348413467407, + "learning_rate": 5.845849869981137e-06, + "loss": 0.5791, + "step": 15195 + }, + { + "epoch": 0.6728136103516598, + "grad_norm": 1.5519170761108398, + "learning_rate": 5.8444441279814215e-06, + "loss": 0.571, + "step": 15196 + }, + { + "epoch": 0.6728578860564737, + "grad_norm": 1.438503384590149, + "learning_rate": 5.843038485233772e-06, + "loss": 0.556, + "step": 15197 + }, + { + "epoch": 0.6729021617612876, + "grad_norm": 1.7326146364212036, + "learning_rate": 5.8416329417717665e-06, + "loss": 0.5723, + "step": 15198 + }, + { + "epoch": 0.6729464374661014, + "grad_norm": 1.474118947982788, + "learning_rate": 5.840227497628971e-06, + "loss": 0.5692, + "step": 15199 + }, + { + "epoch": 0.6729907131709153, + "grad_norm": 1.3935095071792603, + "learning_rate": 5.838822152838951e-06, + "loss": 0.5553, + "step": 15200 + }, + { + "epoch": 0.6730349888757292, + "grad_norm": 1.3830699920654297, + "learning_rate": 5.837416907435281e-06, + "loss": 0.6013, + "step": 15201 + }, + { + "epoch": 0.673079264580543, + "grad_norm": 1.2211490869522095, + "learning_rate": 5.836011761451516e-06, + "loss": 0.4901, + "step": 15202 + }, + { + "epoch": 0.6731235402853569, + "grad_norm": 1.6713135242462158, + "learning_rate": 5.8346067149212225e-06, + "loss": 0.5258, + "step": 15203 + }, + { + "epoch": 0.6731678159901708, + "grad_norm": 1.5626665353775024, + "learning_rate": 5.833201767877955e-06, + "loss": 0.5953, + "step": 15204 + }, + { + "epoch": 0.6732120916949846, + "grad_norm": 1.26217520236969, + "learning_rate": 5.831796920355271e-06, + "loss": 0.5019, + "step": 15205 + }, + { + "epoch": 0.6732563673997985, + "grad_norm": 1.6772321462631226, + "learning_rate": 5.830392172386723e-06, + "loss": 0.5409, + "step": 15206 + }, + { + "epoch": 0.6733006431046125, + "grad_norm": 1.6313221454620361, + "learning_rate": 5.828987524005859e-06, + "loss": 0.5689, + "step": 15207 + }, + { + "epoch": 0.6733449188094263, + "grad_norm": 1.6110131740570068, + "learning_rate": 5.827582975246236e-06, + "loss": 0.5777, + "step": 15208 + }, + { + "epoch": 0.6733891945142402, + "grad_norm": 1.4100052118301392, + "learning_rate": 5.826178526141397e-06, + "loss": 0.5564, + "step": 15209 + }, + { + "epoch": 0.6734334702190541, + "grad_norm": 1.4320495128631592, + "learning_rate": 5.8247741767248815e-06, + "loss": 0.569, + "step": 15210 + }, + { + "epoch": 0.6734777459238679, + "grad_norm": 1.4757370948791504, + "learning_rate": 5.823369927030238e-06, + "loss": 0.5609, + "step": 15211 + }, + { + "epoch": 0.6735220216286818, + "grad_norm": 1.9971671104431152, + "learning_rate": 5.821965777091003e-06, + "loss": 0.703, + "step": 15212 + }, + { + "epoch": 0.6735662973334957, + "grad_norm": 1.4073811769485474, + "learning_rate": 5.820561726940714e-06, + "loss": 0.5425, + "step": 15213 + }, + { + "epoch": 0.6736105730383095, + "grad_norm": 1.3905646800994873, + "learning_rate": 5.819157776612905e-06, + "loss": 0.5398, + "step": 15214 + }, + { + "epoch": 0.6736548487431234, + "grad_norm": 1.7847602367401123, + "learning_rate": 5.81775392614111e-06, + "loss": 0.6375, + "step": 15215 + }, + { + "epoch": 0.6736991244479373, + "grad_norm": 1.428514003753662, + "learning_rate": 5.816350175558855e-06, + "loss": 0.5438, + "step": 15216 + }, + { + "epoch": 0.6737434001527511, + "grad_norm": 1.9247610569000244, + "learning_rate": 5.8149465248996665e-06, + "loss": 0.5692, + "step": 15217 + }, + { + "epoch": 0.673787675857565, + "grad_norm": 1.7260315418243408, + "learning_rate": 5.813542974197076e-06, + "loss": 0.5717, + "step": 15218 + }, + { + "epoch": 0.673831951562379, + "grad_norm": 1.3414534330368042, + "learning_rate": 5.812139523484604e-06, + "loss": 0.5381, + "step": 15219 + }, + { + "epoch": 0.6738762272671928, + "grad_norm": 1.5378141403198242, + "learning_rate": 5.810736172795763e-06, + "loss": 0.5818, + "step": 15220 + }, + { + "epoch": 0.6739205029720067, + "grad_norm": 1.31472909450531, + "learning_rate": 5.809332922164084e-06, + "loss": 0.5275, + "step": 15221 + }, + { + "epoch": 0.6739647786768206, + "grad_norm": 1.7396117448806763, + "learning_rate": 5.807929771623075e-06, + "loss": 0.6059, + "step": 15222 + }, + { + "epoch": 0.6740090543816344, + "grad_norm": 1.3961830139160156, + "learning_rate": 5.806526721206252e-06, + "loss": 0.4882, + "step": 15223 + }, + { + "epoch": 0.6740533300864483, + "grad_norm": 1.3402912616729736, + "learning_rate": 5.805123770947121e-06, + "loss": 0.4705, + "step": 15224 + }, + { + "epoch": 0.6740976057912622, + "grad_norm": 1.5386089086532593, + "learning_rate": 5.8037209208791945e-06, + "loss": 0.6347, + "step": 15225 + }, + { + "epoch": 0.674141881496076, + "grad_norm": 1.6927073001861572, + "learning_rate": 5.802318171035978e-06, + "loss": 0.5791, + "step": 15226 + }, + { + "epoch": 0.6741861572008899, + "grad_norm": 1.3655084371566772, + "learning_rate": 5.800915521450969e-06, + "loss": 0.5174, + "step": 15227 + }, + { + "epoch": 0.6742304329057038, + "grad_norm": 1.4785429239273071, + "learning_rate": 5.799512972157677e-06, + "loss": 0.5883, + "step": 15228 + }, + { + "epoch": 0.6742747086105177, + "grad_norm": 1.689092755317688, + "learning_rate": 5.7981105231896e-06, + "loss": 0.5949, + "step": 15229 + }, + { + "epoch": 0.6743189843153315, + "grad_norm": 1.8164716958999634, + "learning_rate": 5.796708174580225e-06, + "loss": 0.588, + "step": 15230 + }, + { + "epoch": 0.6743632600201455, + "grad_norm": 1.611254096031189, + "learning_rate": 5.795305926363058e-06, + "loss": 0.5385, + "step": 15231 + }, + { + "epoch": 0.6744075357249594, + "grad_norm": 1.6025584936141968, + "learning_rate": 5.793903778571586e-06, + "loss": 0.5698, + "step": 15232 + }, + { + "epoch": 0.6744518114297732, + "grad_norm": 1.423726201057434, + "learning_rate": 5.792501731239296e-06, + "loss": 0.5445, + "step": 15233 + }, + { + "epoch": 0.6744960871345871, + "grad_norm": 1.4846179485321045, + "learning_rate": 5.791099784399677e-06, + "loss": 0.5821, + "step": 15234 + }, + { + "epoch": 0.674540362839401, + "grad_norm": 1.4157296419143677, + "learning_rate": 5.7896979380862124e-06, + "loss": 0.5231, + "step": 15235 + }, + { + "epoch": 0.6745846385442148, + "grad_norm": 1.4708739519119263, + "learning_rate": 5.788296192332386e-06, + "loss": 0.5716, + "step": 15236 + }, + { + "epoch": 0.6746289142490287, + "grad_norm": 2.3596510887145996, + "learning_rate": 5.78689454717167e-06, + "loss": 0.5762, + "step": 15237 + }, + { + "epoch": 0.6746731899538426, + "grad_norm": 1.4135870933532715, + "learning_rate": 5.785493002637552e-06, + "loss": 0.5277, + "step": 15238 + }, + { + "epoch": 0.6747174656586564, + "grad_norm": 1.568067193031311, + "learning_rate": 5.784091558763501e-06, + "loss": 0.6487, + "step": 15239 + }, + { + "epoch": 0.6747617413634703, + "grad_norm": 1.4808794260025024, + "learning_rate": 5.782690215582988e-06, + "loss": 0.5459, + "step": 15240 + }, + { + "epoch": 0.6748060170682842, + "grad_norm": 1.3905776739120483, + "learning_rate": 5.781288973129488e-06, + "loss": 0.5243, + "step": 15241 + }, + { + "epoch": 0.674850292773098, + "grad_norm": 1.3626669645309448, + "learning_rate": 5.77988783143647e-06, + "loss": 0.5647, + "step": 15242 + }, + { + "epoch": 0.674894568477912, + "grad_norm": 1.3866089582443237, + "learning_rate": 5.778486790537392e-06, + "loss": 0.5355, + "step": 15243 + }, + { + "epoch": 0.6749388441827259, + "grad_norm": 2.150564670562744, + "learning_rate": 5.777085850465716e-06, + "loss": 0.7122, + "step": 15244 + }, + { + "epoch": 0.6749831198875397, + "grad_norm": 1.6905869245529175, + "learning_rate": 5.775685011254909e-06, + "loss": 0.5443, + "step": 15245 + }, + { + "epoch": 0.6750273955923536, + "grad_norm": 1.9180586338043213, + "learning_rate": 5.774284272938427e-06, + "loss": 0.6511, + "step": 15246 + }, + { + "epoch": 0.6750716712971675, + "grad_norm": 1.4225648641586304, + "learning_rate": 5.772883635549719e-06, + "loss": 0.5161, + "step": 15247 + }, + { + "epoch": 0.6751159470019813, + "grad_norm": 1.572765827178955, + "learning_rate": 5.771483099122248e-06, + "loss": 0.5535, + "step": 15248 + }, + { + "epoch": 0.6751602227067952, + "grad_norm": 1.450382947921753, + "learning_rate": 5.770082663689461e-06, + "loss": 0.5842, + "step": 15249 + }, + { + "epoch": 0.6752044984116091, + "grad_norm": 1.5493087768554688, + "learning_rate": 5.7686823292848005e-06, + "loss": 0.5732, + "step": 15250 + }, + { + "epoch": 0.6752487741164229, + "grad_norm": 1.5916568040847778, + "learning_rate": 5.767282095941725e-06, + "loss": 0.5883, + "step": 15251 + }, + { + "epoch": 0.6752930498212368, + "grad_norm": 1.3845492601394653, + "learning_rate": 5.765881963693667e-06, + "loss": 0.5004, + "step": 15252 + }, + { + "epoch": 0.6753373255260507, + "grad_norm": 1.3519972562789917, + "learning_rate": 5.764481932574071e-06, + "loss": 0.5319, + "step": 15253 + }, + { + "epoch": 0.6753816012308645, + "grad_norm": 1.3024780750274658, + "learning_rate": 5.76308200261637e-06, + "loss": 0.4843, + "step": 15254 + }, + { + "epoch": 0.6754258769356785, + "grad_norm": 1.6962507963180542, + "learning_rate": 5.761682173854011e-06, + "loss": 0.5683, + "step": 15255 + }, + { + "epoch": 0.6754701526404924, + "grad_norm": 1.5987318754196167, + "learning_rate": 5.760282446320422e-06, + "loss": 0.5597, + "step": 15256 + }, + { + "epoch": 0.6755144283453062, + "grad_norm": 1.7112414836883545, + "learning_rate": 5.75888282004903e-06, + "loss": 0.5623, + "step": 15257 + }, + { + "epoch": 0.6755587040501201, + "grad_norm": 1.448896884918213, + "learning_rate": 5.757483295073273e-06, + "loss": 0.5411, + "step": 15258 + }, + { + "epoch": 0.675602979754934, + "grad_norm": 1.9848065376281738, + "learning_rate": 5.756083871426573e-06, + "loss": 0.6725, + "step": 15259 + }, + { + "epoch": 0.6756472554597478, + "grad_norm": 1.4702093601226807, + "learning_rate": 5.754684549142356e-06, + "loss": 0.5717, + "step": 15260 + }, + { + "epoch": 0.6756915311645617, + "grad_norm": 1.4852362871170044, + "learning_rate": 5.753285328254039e-06, + "loss": 0.5281, + "step": 15261 + }, + { + "epoch": 0.6757358068693756, + "grad_norm": 1.5683002471923828, + "learning_rate": 5.751886208795045e-06, + "loss": 0.537, + "step": 15262 + }, + { + "epoch": 0.6757800825741894, + "grad_norm": 1.4534832239151, + "learning_rate": 5.75048719079879e-06, + "loss": 0.5639, + "step": 15263 + }, + { + "epoch": 0.6758243582790033, + "grad_norm": 1.7426987886428833, + "learning_rate": 5.749088274298685e-06, + "loss": 0.5436, + "step": 15264 + }, + { + "epoch": 0.6758686339838172, + "grad_norm": 1.5294872522354126, + "learning_rate": 5.74768945932815e-06, + "loss": 0.5731, + "step": 15265 + }, + { + "epoch": 0.6759129096886312, + "grad_norm": 1.4926385879516602, + "learning_rate": 5.746290745920589e-06, + "loss": 0.5735, + "step": 15266 + }, + { + "epoch": 0.675957185393445, + "grad_norm": 1.7112656831741333, + "learning_rate": 5.744892134109406e-06, + "loss": 0.5944, + "step": 15267 + }, + { + "epoch": 0.6760014610982589, + "grad_norm": 1.7244597673416138, + "learning_rate": 5.743493623928014e-06, + "loss": 0.6255, + "step": 15268 + }, + { + "epoch": 0.6760457368030728, + "grad_norm": 1.8094443082809448, + "learning_rate": 5.7420952154098095e-06, + "loss": 0.6347, + "step": 15269 + }, + { + "epoch": 0.6760900125078866, + "grad_norm": 1.7497427463531494, + "learning_rate": 5.740696908588195e-06, + "loss": 0.6429, + "step": 15270 + }, + { + "epoch": 0.6761342882127005, + "grad_norm": 1.9395021200180054, + "learning_rate": 5.7392987034965675e-06, + "loss": 0.5591, + "step": 15271 + }, + { + "epoch": 0.6761785639175144, + "grad_norm": 1.5457392930984497, + "learning_rate": 5.73790060016832e-06, + "loss": 0.5131, + "step": 15272 + }, + { + "epoch": 0.6762228396223282, + "grad_norm": 1.335838794708252, + "learning_rate": 5.736502598636847e-06, + "loss": 0.4918, + "step": 15273 + }, + { + "epoch": 0.6762671153271421, + "grad_norm": 1.4164384603500366, + "learning_rate": 5.735104698935534e-06, + "loss": 0.5276, + "step": 15274 + }, + { + "epoch": 0.676311391031956, + "grad_norm": 1.5707954168319702, + "learning_rate": 5.7337069010977775e-06, + "loss": 0.5474, + "step": 15275 + }, + { + "epoch": 0.6763556667367698, + "grad_norm": 1.508384346961975, + "learning_rate": 5.7323092051569565e-06, + "loss": 0.5556, + "step": 15276 + }, + { + "epoch": 0.6763999424415837, + "grad_norm": 1.6305650472640991, + "learning_rate": 5.730911611146452e-06, + "loss": 0.5945, + "step": 15277 + }, + { + "epoch": 0.6764442181463977, + "grad_norm": 1.650717854499817, + "learning_rate": 5.729514119099651e-06, + "loss": 0.576, + "step": 15278 + }, + { + "epoch": 0.6764884938512115, + "grad_norm": 1.5173652172088623, + "learning_rate": 5.728116729049929e-06, + "loss": 0.564, + "step": 15279 + }, + { + "epoch": 0.6765327695560254, + "grad_norm": 1.4077869653701782, + "learning_rate": 5.7267194410306595e-06, + "loss": 0.5474, + "step": 15280 + }, + { + "epoch": 0.6765770452608393, + "grad_norm": 1.5721142292022705, + "learning_rate": 5.725322255075217e-06, + "loss": 0.5315, + "step": 15281 + }, + { + "epoch": 0.6766213209656531, + "grad_norm": 1.5141113996505737, + "learning_rate": 5.723925171216973e-06, + "loss": 0.581, + "step": 15282 + }, + { + "epoch": 0.676665596670467, + "grad_norm": 1.6756740808486938, + "learning_rate": 5.722528189489294e-06, + "loss": 0.5761, + "step": 15283 + }, + { + "epoch": 0.6767098723752809, + "grad_norm": 1.5706287622451782, + "learning_rate": 5.721131309925542e-06, + "loss": 0.5441, + "step": 15284 + }, + { + "epoch": 0.6767541480800947, + "grad_norm": 1.55744469165802, + "learning_rate": 5.719734532559089e-06, + "loss": 0.5953, + "step": 15285 + }, + { + "epoch": 0.6767984237849086, + "grad_norm": 1.3209072351455688, + "learning_rate": 5.718337857423291e-06, + "loss": 0.4928, + "step": 15286 + }, + { + "epoch": 0.6768426994897225, + "grad_norm": 1.7437843084335327, + "learning_rate": 5.716941284551504e-06, + "loss": 0.5948, + "step": 15287 + }, + { + "epoch": 0.6768869751945363, + "grad_norm": 1.6680628061294556, + "learning_rate": 5.715544813977089e-06, + "loss": 0.5438, + "step": 15288 + }, + { + "epoch": 0.6769312508993502, + "grad_norm": 2.0708999633789062, + "learning_rate": 5.7141484457334004e-06, + "loss": 0.6137, + "step": 15289 + }, + { + "epoch": 0.6769755266041642, + "grad_norm": 1.7134881019592285, + "learning_rate": 5.712752179853785e-06, + "loss": 0.6371, + "step": 15290 + }, + { + "epoch": 0.677019802308978, + "grad_norm": 1.4341514110565186, + "learning_rate": 5.711356016371593e-06, + "loss": 0.5434, + "step": 15291 + }, + { + "epoch": 0.6770640780137919, + "grad_norm": 1.2633678913116455, + "learning_rate": 5.70995995532017e-06, + "loss": 0.4945, + "step": 15292 + }, + { + "epoch": 0.6771083537186058, + "grad_norm": 1.4414781332015991, + "learning_rate": 5.708563996732862e-06, + "loss": 0.5119, + "step": 15293 + }, + { + "epoch": 0.6771526294234196, + "grad_norm": 1.2807708978652954, + "learning_rate": 5.707168140643003e-06, + "loss": 0.4742, + "step": 15294 + }, + { + "epoch": 0.6771969051282335, + "grad_norm": 1.528180480003357, + "learning_rate": 5.705772387083943e-06, + "loss": 0.5775, + "step": 15295 + }, + { + "epoch": 0.6772411808330474, + "grad_norm": 1.6295958757400513, + "learning_rate": 5.7043767360890124e-06, + "loss": 0.6065, + "step": 15296 + }, + { + "epoch": 0.6772854565378612, + "grad_norm": 1.4933888912200928, + "learning_rate": 5.702981187691542e-06, + "loss": 0.569, + "step": 15297 + }, + { + "epoch": 0.6773297322426751, + "grad_norm": 1.4665247201919556, + "learning_rate": 5.701585741924876e-06, + "loss": 0.559, + "step": 15298 + }, + { + "epoch": 0.677374007947489, + "grad_norm": 1.6234475374221802, + "learning_rate": 5.700190398822329e-06, + "loss": 0.5538, + "step": 15299 + }, + { + "epoch": 0.6774182836523029, + "grad_norm": 1.5887778997421265, + "learning_rate": 5.698795158417234e-06, + "loss": 0.646, + "step": 15300 + }, + { + "epoch": 0.6774625593571167, + "grad_norm": 1.6343743801116943, + "learning_rate": 5.69740002074291e-06, + "loss": 0.5577, + "step": 15301 + }, + { + "epoch": 0.6775068350619307, + "grad_norm": 1.2183090448379517, + "learning_rate": 5.696004985832688e-06, + "loss": 0.4658, + "step": 15302 + }, + { + "epoch": 0.6775511107667446, + "grad_norm": 1.6567821502685547, + "learning_rate": 5.694610053719882e-06, + "loss": 0.5504, + "step": 15303 + }, + { + "epoch": 0.6775953864715584, + "grad_norm": 1.2424999475479126, + "learning_rate": 5.693215224437806e-06, + "loss": 0.4764, + "step": 15304 + }, + { + "epoch": 0.6776396621763723, + "grad_norm": 1.480638027191162, + "learning_rate": 5.6918204980197815e-06, + "loss": 0.5831, + "step": 15305 + }, + { + "epoch": 0.6776839378811862, + "grad_norm": 1.6924716234207153, + "learning_rate": 5.690425874499116e-06, + "loss": 0.5508, + "step": 15306 + }, + { + "epoch": 0.677728213586, + "grad_norm": 1.5184006690979004, + "learning_rate": 5.68903135390912e-06, + "loss": 0.586, + "step": 15307 + }, + { + "epoch": 0.6777724892908139, + "grad_norm": 1.2230875492095947, + "learning_rate": 5.6876369362830984e-06, + "loss": 0.4999, + "step": 15308 + }, + { + "epoch": 0.6778167649956278, + "grad_norm": 1.4601898193359375, + "learning_rate": 5.6862426216543586e-06, + "loss": 0.5184, + "step": 15309 + }, + { + "epoch": 0.6778610407004416, + "grad_norm": 1.3186839818954468, + "learning_rate": 5.684848410056202e-06, + "loss": 0.5536, + "step": 15310 + }, + { + "epoch": 0.6779053164052555, + "grad_norm": 1.5712499618530273, + "learning_rate": 5.683454301521922e-06, + "loss": 0.6161, + "step": 15311 + }, + { + "epoch": 0.6779495921100694, + "grad_norm": 1.3167634010314941, + "learning_rate": 5.682060296084828e-06, + "loss": 0.5175, + "step": 15312 + }, + { + "epoch": 0.6779938678148832, + "grad_norm": 1.4415329694747925, + "learning_rate": 5.680666393778206e-06, + "loss": 0.5306, + "step": 15313 + }, + { + "epoch": 0.6780381435196972, + "grad_norm": 1.989030361175537, + "learning_rate": 5.679272594635345e-06, + "loss": 0.6637, + "step": 15314 + }, + { + "epoch": 0.6780824192245111, + "grad_norm": 1.7213501930236816, + "learning_rate": 5.6778788986895464e-06, + "loss": 0.5752, + "step": 15315 + }, + { + "epoch": 0.6781266949293249, + "grad_norm": 1.4392763376235962, + "learning_rate": 5.676485305974091e-06, + "loss": 0.5553, + "step": 15316 + }, + { + "epoch": 0.6781709706341388, + "grad_norm": 1.9672993421554565, + "learning_rate": 5.675091816522263e-06, + "loss": 0.6471, + "step": 15317 + }, + { + "epoch": 0.6782152463389527, + "grad_norm": 1.5464341640472412, + "learning_rate": 5.673698430367345e-06, + "loss": 0.5708, + "step": 15318 + }, + { + "epoch": 0.6782595220437665, + "grad_norm": 1.767786979675293, + "learning_rate": 5.672305147542619e-06, + "loss": 0.6533, + "step": 15319 + }, + { + "epoch": 0.6783037977485804, + "grad_norm": 4.31387186050415, + "learning_rate": 5.670911968081359e-06, + "loss": 0.5499, + "step": 15320 + }, + { + "epoch": 0.6783480734533943, + "grad_norm": 3.1259543895721436, + "learning_rate": 5.669518892016839e-06, + "loss": 0.5797, + "step": 15321 + }, + { + "epoch": 0.6783923491582081, + "grad_norm": 1.6246131658554077, + "learning_rate": 5.668125919382337e-06, + "loss": 0.5332, + "step": 15322 + }, + { + "epoch": 0.678436624863022, + "grad_norm": 1.832258939743042, + "learning_rate": 5.666733050211122e-06, + "loss": 0.5368, + "step": 15323 + }, + { + "epoch": 0.6784809005678359, + "grad_norm": 1.3701518774032593, + "learning_rate": 5.665340284536454e-06, + "loss": 0.5122, + "step": 15324 + }, + { + "epoch": 0.6785251762726497, + "grad_norm": 1.386339545249939, + "learning_rate": 5.663947622391609e-06, + "loss": 0.4817, + "step": 15325 + }, + { + "epoch": 0.6785694519774637, + "grad_norm": 1.384982705116272, + "learning_rate": 5.662555063809845e-06, + "loss": 0.5162, + "step": 15326 + }, + { + "epoch": 0.6786137276822776, + "grad_norm": 1.5528409481048584, + "learning_rate": 5.66116260882442e-06, + "loss": 0.5708, + "step": 15327 + }, + { + "epoch": 0.6786580033870914, + "grad_norm": 1.7070895433425903, + "learning_rate": 5.659770257468594e-06, + "loss": 0.5673, + "step": 15328 + }, + { + "epoch": 0.6787022790919053, + "grad_norm": 1.8631131649017334, + "learning_rate": 5.658378009775622e-06, + "loss": 0.6426, + "step": 15329 + }, + { + "epoch": 0.6787465547967192, + "grad_norm": 1.4428507089614868, + "learning_rate": 5.6569858657787546e-06, + "loss": 0.5572, + "step": 15330 + }, + { + "epoch": 0.678790830501533, + "grad_norm": 1.5841608047485352, + "learning_rate": 5.655593825511241e-06, + "loss": 0.5585, + "step": 15331 + }, + { + "epoch": 0.6788351062063469, + "grad_norm": 1.5170166492462158, + "learning_rate": 5.654201889006336e-06, + "loss": 0.5218, + "step": 15332 + }, + { + "epoch": 0.6788793819111608, + "grad_norm": 1.3268946409225464, + "learning_rate": 5.65281005629728e-06, + "loss": 0.55, + "step": 15333 + }, + { + "epoch": 0.6789236576159747, + "grad_norm": 2.057424783706665, + "learning_rate": 5.651418327417311e-06, + "loss": 0.6749, + "step": 15334 + }, + { + "epoch": 0.6789679333207885, + "grad_norm": 1.4873390197753906, + "learning_rate": 5.65002670239968e-06, + "loss": 0.6011, + "step": 15335 + }, + { + "epoch": 0.6790122090256024, + "grad_norm": 1.737207293510437, + "learning_rate": 5.6486351812776185e-06, + "loss": 0.6035, + "step": 15336 + }, + { + "epoch": 0.6790564847304164, + "grad_norm": 1.3854156732559204, + "learning_rate": 5.647243764084363e-06, + "loss": 0.5867, + "step": 15337 + }, + { + "epoch": 0.6791007604352302, + "grad_norm": 1.5774027109146118, + "learning_rate": 5.645852450853146e-06, + "loss": 0.5876, + "step": 15338 + }, + { + "epoch": 0.6791450361400441, + "grad_norm": 1.5505237579345703, + "learning_rate": 5.6444612416171976e-06, + "loss": 0.4844, + "step": 15339 + }, + { + "epoch": 0.679189311844858, + "grad_norm": 1.8915166854858398, + "learning_rate": 5.643070136409747e-06, + "loss": 0.5989, + "step": 15340 + }, + { + "epoch": 0.6792335875496718, + "grad_norm": 1.517367959022522, + "learning_rate": 5.641679135264013e-06, + "loss": 0.5989, + "step": 15341 + }, + { + "epoch": 0.6792778632544857, + "grad_norm": 1.9259254932403564, + "learning_rate": 5.6402882382132295e-06, + "loss": 0.5557, + "step": 15342 + }, + { + "epoch": 0.6793221389592996, + "grad_norm": 1.4456912279129028, + "learning_rate": 5.6388974452906105e-06, + "loss": 0.573, + "step": 15343 + }, + { + "epoch": 0.6793664146641134, + "grad_norm": 1.4506839513778687, + "learning_rate": 5.637506756529372e-06, + "loss": 0.4982, + "step": 15344 + }, + { + "epoch": 0.6794106903689273, + "grad_norm": 1.4327763319015503, + "learning_rate": 5.636116171962736e-06, + "loss": 0.567, + "step": 15345 + }, + { + "epoch": 0.6794549660737412, + "grad_norm": 2.027172088623047, + "learning_rate": 5.6347256916239166e-06, + "loss": 0.6315, + "step": 15346 + }, + { + "epoch": 0.679499241778555, + "grad_norm": 1.4929122924804688, + "learning_rate": 5.633335315546111e-06, + "loss": 0.5568, + "step": 15347 + }, + { + "epoch": 0.6795435174833689, + "grad_norm": 1.8572957515716553, + "learning_rate": 5.631945043762541e-06, + "loss": 0.6282, + "step": 15348 + }, + { + "epoch": 0.6795877931881829, + "grad_norm": 1.2486077547073364, + "learning_rate": 5.630554876306407e-06, + "loss": 0.5135, + "step": 15349 + }, + { + "epoch": 0.6796320688929967, + "grad_norm": 1.525130271911621, + "learning_rate": 5.629164813210911e-06, + "loss": 0.5924, + "step": 15350 + }, + { + "epoch": 0.6796763445978106, + "grad_norm": 1.3092836141586304, + "learning_rate": 5.627774854509252e-06, + "loss": 0.4895, + "step": 15351 + }, + { + "epoch": 0.6797206203026245, + "grad_norm": 1.5041577816009521, + "learning_rate": 5.626385000234633e-06, + "loss": 0.5212, + "step": 15352 + }, + { + "epoch": 0.6797648960074383, + "grad_norm": 1.5597918033599854, + "learning_rate": 5.624995250420248e-06, + "loss": 0.5482, + "step": 15353 + }, + { + "epoch": 0.6798091717122522, + "grad_norm": 1.5981675386428833, + "learning_rate": 5.623605605099289e-06, + "loss": 0.6142, + "step": 15354 + }, + { + "epoch": 0.6798534474170661, + "grad_norm": 1.5681726932525635, + "learning_rate": 5.622216064304946e-06, + "loss": 0.5185, + "step": 15355 + }, + { + "epoch": 0.6798977231218799, + "grad_norm": 1.3954253196716309, + "learning_rate": 5.620826628070408e-06, + "loss": 0.4996, + "step": 15356 + }, + { + "epoch": 0.6799419988266938, + "grad_norm": 1.3574150800704956, + "learning_rate": 5.6194372964288576e-06, + "loss": 0.5524, + "step": 15357 + }, + { + "epoch": 0.6799862745315077, + "grad_norm": 1.5235220193862915, + "learning_rate": 5.618048069413484e-06, + "loss": 0.5706, + "step": 15358 + }, + { + "epoch": 0.6800305502363215, + "grad_norm": 1.4703911542892456, + "learning_rate": 5.616658947057464e-06, + "loss": 0.5974, + "step": 15359 + }, + { + "epoch": 0.6800748259411354, + "grad_norm": 1.685002326965332, + "learning_rate": 5.615269929393976e-06, + "loss": 0.5976, + "step": 15360 + }, + { + "epoch": 0.6801191016459494, + "grad_norm": 1.6582427024841309, + "learning_rate": 5.613881016456192e-06, + "loss": 0.6139, + "step": 15361 + }, + { + "epoch": 0.6801633773507632, + "grad_norm": 1.5752754211425781, + "learning_rate": 5.612492208277291e-06, + "loss": 0.4981, + "step": 15362 + }, + { + "epoch": 0.6802076530555771, + "grad_norm": 1.5855075120925903, + "learning_rate": 5.611103504890444e-06, + "loss": 0.5614, + "step": 15363 + }, + { + "epoch": 0.680251928760391, + "grad_norm": 1.420351505279541, + "learning_rate": 5.6097149063288134e-06, + "loss": 0.5058, + "step": 15364 + }, + { + "epoch": 0.6802962044652048, + "grad_norm": 1.6419038772583008, + "learning_rate": 5.60832641262557e-06, + "loss": 0.5995, + "step": 15365 + }, + { + "epoch": 0.6803404801700187, + "grad_norm": 1.464226245880127, + "learning_rate": 5.6069380238138725e-06, + "loss": 0.5501, + "step": 15366 + }, + { + "epoch": 0.6803847558748326, + "grad_norm": 1.4757822751998901, + "learning_rate": 5.605549739926881e-06, + "loss": 0.5482, + "step": 15367 + }, + { + "epoch": 0.6804290315796464, + "grad_norm": 1.6411893367767334, + "learning_rate": 5.604161560997758e-06, + "loss": 0.5467, + "step": 15368 + }, + { + "epoch": 0.6804733072844603, + "grad_norm": 1.7787127494812012, + "learning_rate": 5.602773487059658e-06, + "loss": 0.5801, + "step": 15369 + }, + { + "epoch": 0.6805175829892742, + "grad_norm": 1.3905514478683472, + "learning_rate": 5.601385518145733e-06, + "loss": 0.5056, + "step": 15370 + }, + { + "epoch": 0.6805618586940881, + "grad_norm": 1.742815613746643, + "learning_rate": 5.599997654289129e-06, + "loss": 0.6536, + "step": 15371 + }, + { + "epoch": 0.6806061343989019, + "grad_norm": 1.3417651653289795, + "learning_rate": 5.598609895523004e-06, + "loss": 0.5107, + "step": 15372 + }, + { + "epoch": 0.6806504101037159, + "grad_norm": 1.4988090991973877, + "learning_rate": 5.597222241880496e-06, + "loss": 0.5215, + "step": 15373 + }, + { + "epoch": 0.6806946858085298, + "grad_norm": 1.3921921253204346, + "learning_rate": 5.595834693394751e-06, + "loss": 0.5426, + "step": 15374 + }, + { + "epoch": 0.6807389615133436, + "grad_norm": 1.4520806074142456, + "learning_rate": 5.594447250098907e-06, + "loss": 0.5435, + "step": 15375 + }, + { + "epoch": 0.6807832372181575, + "grad_norm": 1.5354729890823364, + "learning_rate": 5.593059912026104e-06, + "loss": 0.4963, + "step": 15376 + }, + { + "epoch": 0.6808275129229714, + "grad_norm": 1.4605516195297241, + "learning_rate": 5.591672679209472e-06, + "loss": 0.5396, + "step": 15377 + }, + { + "epoch": 0.6808717886277852, + "grad_norm": 1.523492455482483, + "learning_rate": 5.590285551682154e-06, + "loss": 0.5665, + "step": 15378 + }, + { + "epoch": 0.6809160643325991, + "grad_norm": 1.324037790298462, + "learning_rate": 5.588898529477274e-06, + "loss": 0.5232, + "step": 15379 + }, + { + "epoch": 0.680960340037413, + "grad_norm": 1.4587452411651611, + "learning_rate": 5.587511612627962e-06, + "loss": 0.5513, + "step": 15380 + }, + { + "epoch": 0.6810046157422268, + "grad_norm": 1.62688148021698, + "learning_rate": 5.5861248011673375e-06, + "loss": 0.5429, + "step": 15381 + }, + { + "epoch": 0.6810488914470407, + "grad_norm": 1.519810676574707, + "learning_rate": 5.584738095128532e-06, + "loss": 0.4873, + "step": 15382 + }, + { + "epoch": 0.6810931671518546, + "grad_norm": 1.8144357204437256, + "learning_rate": 5.5833514945446636e-06, + "loss": 0.5707, + "step": 15383 + }, + { + "epoch": 0.6811374428566684, + "grad_norm": 2.2623836994171143, + "learning_rate": 5.581964999448848e-06, + "loss": 0.5226, + "step": 15384 + }, + { + "epoch": 0.6811817185614824, + "grad_norm": 1.5875211954116821, + "learning_rate": 5.5805786098742e-06, + "loss": 0.614, + "step": 15385 + }, + { + "epoch": 0.6812259942662963, + "grad_norm": 1.3900398015975952, + "learning_rate": 5.579192325853834e-06, + "loss": 0.5437, + "step": 15386 + }, + { + "epoch": 0.6812702699711101, + "grad_norm": 1.3184335231781006, + "learning_rate": 5.5778061474208565e-06, + "loss": 0.5093, + "step": 15387 + }, + { + "epoch": 0.681314545675924, + "grad_norm": 1.42803955078125, + "learning_rate": 5.576420074608382e-06, + "loss": 0.5097, + "step": 15388 + }, + { + "epoch": 0.6813588213807379, + "grad_norm": 1.527855634689331, + "learning_rate": 5.57503410744951e-06, + "loss": 0.538, + "step": 15389 + }, + { + "epoch": 0.6814030970855517, + "grad_norm": 1.5590497255325317, + "learning_rate": 5.573648245977349e-06, + "loss": 0.5178, + "step": 15390 + }, + { + "epoch": 0.6814473727903656, + "grad_norm": 1.7587202787399292, + "learning_rate": 5.57226249022499e-06, + "loss": 0.567, + "step": 15391 + }, + { + "epoch": 0.6814916484951795, + "grad_norm": 1.6545071601867676, + "learning_rate": 5.5708768402255405e-06, + "loss": 0.5747, + "step": 15392 + }, + { + "epoch": 0.6815359241999933, + "grad_norm": 1.6246832609176636, + "learning_rate": 5.569491296012095e-06, + "loss": 0.5067, + "step": 15393 + }, + { + "epoch": 0.6815801999048072, + "grad_norm": 1.2735004425048828, + "learning_rate": 5.568105857617734e-06, + "loss": 0.5115, + "step": 15394 + }, + { + "epoch": 0.6816244756096211, + "grad_norm": 1.4001569747924805, + "learning_rate": 5.566720525075561e-06, + "loss": 0.5494, + "step": 15395 + }, + { + "epoch": 0.6816687513144349, + "grad_norm": 1.9049732685089111, + "learning_rate": 5.5653352984186586e-06, + "loss": 0.6487, + "step": 15396 + }, + { + "epoch": 0.6817130270192489, + "grad_norm": 1.5507651567459106, + "learning_rate": 5.563950177680106e-06, + "loss": 0.5704, + "step": 15397 + }, + { + "epoch": 0.6817573027240628, + "grad_norm": 1.5282752513885498, + "learning_rate": 5.562565162892997e-06, + "loss": 0.5838, + "step": 15398 + }, + { + "epoch": 0.6818015784288766, + "grad_norm": 1.834660291671753, + "learning_rate": 5.5611802540904055e-06, + "loss": 0.5688, + "step": 15399 + }, + { + "epoch": 0.6818458541336905, + "grad_norm": 1.3722891807556152, + "learning_rate": 5.55979545130541e-06, + "loss": 0.5641, + "step": 15400 + }, + { + "epoch": 0.6818901298385044, + "grad_norm": 1.6220935583114624, + "learning_rate": 5.55841075457108e-06, + "loss": 0.5611, + "step": 15401 + }, + { + "epoch": 0.6819344055433182, + "grad_norm": 1.5490505695343018, + "learning_rate": 5.557026163920502e-06, + "loss": 0.5522, + "step": 15402 + }, + { + "epoch": 0.6819786812481321, + "grad_norm": 1.8607995510101318, + "learning_rate": 5.5556416793867314e-06, + "loss": 0.6363, + "step": 15403 + }, + { + "epoch": 0.682022956952946, + "grad_norm": 1.808989405632019, + "learning_rate": 5.554257301002837e-06, + "loss": 0.6379, + "step": 15404 + }, + { + "epoch": 0.6820672326577599, + "grad_norm": 2.1256117820739746, + "learning_rate": 5.552873028801891e-06, + "loss": 0.6046, + "step": 15405 + }, + { + "epoch": 0.6821115083625737, + "grad_norm": 1.4399968385696411, + "learning_rate": 5.55148886281695e-06, + "loss": 0.5671, + "step": 15406 + }, + { + "epoch": 0.6821557840673876, + "grad_norm": 1.3026683330535889, + "learning_rate": 5.550104803081078e-06, + "loss": 0.5538, + "step": 15407 + }, + { + "epoch": 0.6822000597722016, + "grad_norm": 1.4401787519454956, + "learning_rate": 5.548720849627325e-06, + "loss": 0.5363, + "step": 15408 + }, + { + "epoch": 0.6822443354770154, + "grad_norm": 1.4489924907684326, + "learning_rate": 5.547337002488753e-06, + "loss": 0.5228, + "step": 15409 + }, + { + "epoch": 0.6822886111818293, + "grad_norm": 1.9392437934875488, + "learning_rate": 5.545953261698411e-06, + "loss": 0.6896, + "step": 15410 + }, + { + "epoch": 0.6823328868866432, + "grad_norm": 1.3914780616760254, + "learning_rate": 5.54456962728935e-06, + "loss": 0.5361, + "step": 15411 + }, + { + "epoch": 0.682377162591457, + "grad_norm": 1.3106539249420166, + "learning_rate": 5.543186099294615e-06, + "loss": 0.5239, + "step": 15412 + }, + { + "epoch": 0.6824214382962709, + "grad_norm": 1.7774602174758911, + "learning_rate": 5.541802677747252e-06, + "loss": 0.658, + "step": 15413 + }, + { + "epoch": 0.6824657140010848, + "grad_norm": 1.5358037948608398, + "learning_rate": 5.540419362680297e-06, + "loss": 0.512, + "step": 15414 + }, + { + "epoch": 0.6825099897058986, + "grad_norm": 1.3211125135421753, + "learning_rate": 5.539036154126799e-06, + "loss": 0.5185, + "step": 15415 + }, + { + "epoch": 0.6825542654107125, + "grad_norm": 1.321682095527649, + "learning_rate": 5.537653052119789e-06, + "loss": 0.494, + "step": 15416 + }, + { + "epoch": 0.6825985411155264, + "grad_norm": 1.4089765548706055, + "learning_rate": 5.536270056692304e-06, + "loss": 0.5739, + "step": 15417 + }, + { + "epoch": 0.6826428168203402, + "grad_norm": 1.2381223440170288, + "learning_rate": 5.534887167877369e-06, + "loss": 0.4428, + "step": 15418 + }, + { + "epoch": 0.6826870925251541, + "grad_norm": 1.3701037168502808, + "learning_rate": 5.533504385708024e-06, + "loss": 0.5266, + "step": 15419 + }, + { + "epoch": 0.682731368229968, + "grad_norm": 1.8694154024124146, + "learning_rate": 5.532121710217289e-06, + "loss": 0.6019, + "step": 15420 + }, + { + "epoch": 0.6827756439347819, + "grad_norm": 1.358048677444458, + "learning_rate": 5.530739141438189e-06, + "loss": 0.6052, + "step": 15421 + }, + { + "epoch": 0.6828199196395958, + "grad_norm": 1.3472990989685059, + "learning_rate": 5.529356679403746e-06, + "loss": 0.5288, + "step": 15422 + }, + { + "epoch": 0.6828641953444097, + "grad_norm": 1.7322958707809448, + "learning_rate": 5.527974324146979e-06, + "loss": 0.6432, + "step": 15423 + }, + { + "epoch": 0.6829084710492235, + "grad_norm": 1.4441766738891602, + "learning_rate": 5.5265920757008985e-06, + "loss": 0.5445, + "step": 15424 + }, + { + "epoch": 0.6829527467540374, + "grad_norm": 1.3966495990753174, + "learning_rate": 5.5252099340985284e-06, + "loss": 0.5302, + "step": 15425 + }, + { + "epoch": 0.6829970224588513, + "grad_norm": 1.4754550457000732, + "learning_rate": 5.523827899372876e-06, + "loss": 0.5196, + "step": 15426 + }, + { + "epoch": 0.6830412981636651, + "grad_norm": 1.3285385370254517, + "learning_rate": 5.5224459715569486e-06, + "loss": 0.5316, + "step": 15427 + }, + { + "epoch": 0.683085573868479, + "grad_norm": 1.7131946086883545, + "learning_rate": 5.5210641506837506e-06, + "loss": 0.6245, + "step": 15428 + }, + { + "epoch": 0.6831298495732929, + "grad_norm": 1.3385717868804932, + "learning_rate": 5.519682436786292e-06, + "loss": 0.4931, + "step": 15429 + }, + { + "epoch": 0.6831741252781067, + "grad_norm": 1.4297640323638916, + "learning_rate": 5.518300829897572e-06, + "loss": 0.554, + "step": 15430 + }, + { + "epoch": 0.6832184009829206, + "grad_norm": 1.8365787267684937, + "learning_rate": 5.516919330050585e-06, + "loss": 0.6519, + "step": 15431 + }, + { + "epoch": 0.6832626766877345, + "grad_norm": 1.380129098892212, + "learning_rate": 5.515537937278331e-06, + "loss": 0.5344, + "step": 15432 + }, + { + "epoch": 0.6833069523925484, + "grad_norm": 1.6344929933547974, + "learning_rate": 5.514156651613801e-06, + "loss": 0.5576, + "step": 15433 + }, + { + "epoch": 0.6833512280973623, + "grad_norm": 1.42838454246521, + "learning_rate": 5.512775473089984e-06, + "loss": 0.5846, + "step": 15434 + }, + { + "epoch": 0.6833955038021762, + "grad_norm": 1.3445217609405518, + "learning_rate": 5.511394401739874e-06, + "loss": 0.4991, + "step": 15435 + }, + { + "epoch": 0.68343977950699, + "grad_norm": 1.3986529111862183, + "learning_rate": 5.510013437596455e-06, + "loss": 0.5526, + "step": 15436 + }, + { + "epoch": 0.6834840552118039, + "grad_norm": 1.8218505382537842, + "learning_rate": 5.508632580692708e-06, + "loss": 0.5704, + "step": 15437 + }, + { + "epoch": 0.6835283309166178, + "grad_norm": 1.6211055517196655, + "learning_rate": 5.5072518310616115e-06, + "loss": 0.5037, + "step": 15438 + }, + { + "epoch": 0.6835726066214316, + "grad_norm": 1.4091339111328125, + "learning_rate": 5.5058711887361514e-06, + "loss": 0.5632, + "step": 15439 + }, + { + "epoch": 0.6836168823262455, + "grad_norm": 1.4303117990493774, + "learning_rate": 5.504490653749298e-06, + "loss": 0.5173, + "step": 15440 + }, + { + "epoch": 0.6836611580310594, + "grad_norm": 1.4690678119659424, + "learning_rate": 5.5031102261340255e-06, + "loss": 0.5723, + "step": 15441 + }, + { + "epoch": 0.6837054337358733, + "grad_norm": 1.6086760759353638, + "learning_rate": 5.501729905923303e-06, + "loss": 0.5879, + "step": 15442 + }, + { + "epoch": 0.6837497094406871, + "grad_norm": 1.2059699296951294, + "learning_rate": 5.5003496931500996e-06, + "loss": 0.4894, + "step": 15443 + }, + { + "epoch": 0.683793985145501, + "grad_norm": 1.3072705268859863, + "learning_rate": 5.498969587847376e-06, + "loss": 0.5062, + "step": 15444 + }, + { + "epoch": 0.683838260850315, + "grad_norm": 1.6894420385360718, + "learning_rate": 5.497589590048102e-06, + "loss": 0.5792, + "step": 15445 + }, + { + "epoch": 0.6838825365551288, + "grad_norm": 1.3493424654006958, + "learning_rate": 5.496209699785236e-06, + "loss": 0.501, + "step": 15446 + }, + { + "epoch": 0.6839268122599427, + "grad_norm": 1.4190053939819336, + "learning_rate": 5.494829917091733e-06, + "loss": 0.4965, + "step": 15447 + }, + { + "epoch": 0.6839710879647566, + "grad_norm": 1.4797239303588867, + "learning_rate": 5.493450242000546e-06, + "loss": 0.5686, + "step": 15448 + }, + { + "epoch": 0.6840153636695704, + "grad_norm": 1.2126235961914062, + "learning_rate": 5.492070674544638e-06, + "loss": 0.5182, + "step": 15449 + }, + { + "epoch": 0.6840596393743843, + "grad_norm": 1.6543632745742798, + "learning_rate": 5.490691214756949e-06, + "loss": 0.5831, + "step": 15450 + }, + { + "epoch": 0.6841039150791982, + "grad_norm": 1.4145126342773438, + "learning_rate": 5.489311862670423e-06, + "loss": 0.536, + "step": 15451 + }, + { + "epoch": 0.684148190784012, + "grad_norm": 1.4999924898147583, + "learning_rate": 5.487932618318015e-06, + "loss": 0.5043, + "step": 15452 + }, + { + "epoch": 0.6841924664888259, + "grad_norm": 1.4788529872894287, + "learning_rate": 5.486553481732663e-06, + "loss": 0.4778, + "step": 15453 + }, + { + "epoch": 0.6842367421936398, + "grad_norm": 1.4329302310943604, + "learning_rate": 5.485174452947302e-06, + "loss": 0.5282, + "step": 15454 + }, + { + "epoch": 0.6842810178984536, + "grad_norm": 1.752669334411621, + "learning_rate": 5.4837955319948765e-06, + "loss": 0.6118, + "step": 15455 + }, + { + "epoch": 0.6843252936032675, + "grad_norm": 1.5202594995498657, + "learning_rate": 5.4824167189083176e-06, + "loss": 0.613, + "step": 15456 + }, + { + "epoch": 0.6843695693080815, + "grad_norm": 1.4282865524291992, + "learning_rate": 5.481038013720557e-06, + "loss": 0.507, + "step": 15457 + }, + { + "epoch": 0.6844138450128953, + "grad_norm": 1.6231108903884888, + "learning_rate": 5.479659416464524e-06, + "loss": 0.5403, + "step": 15458 + }, + { + "epoch": 0.6844581207177092, + "grad_norm": 1.2820547819137573, + "learning_rate": 5.478280927173145e-06, + "loss": 0.5461, + "step": 15459 + }, + { + "epoch": 0.6845023964225231, + "grad_norm": 1.3877705335617065, + "learning_rate": 5.476902545879346e-06, + "loss": 0.4796, + "step": 15460 + }, + { + "epoch": 0.6845466721273369, + "grad_norm": 1.4665307998657227, + "learning_rate": 5.4755242726160414e-06, + "loss": 0.5281, + "step": 15461 + }, + { + "epoch": 0.6845909478321508, + "grad_norm": 1.7178890705108643, + "learning_rate": 5.4741461074161595e-06, + "loss": 0.5989, + "step": 15462 + }, + { + "epoch": 0.6846352235369647, + "grad_norm": 1.4623231887817383, + "learning_rate": 5.472768050312614e-06, + "loss": 0.5736, + "step": 15463 + }, + { + "epoch": 0.6846794992417785, + "grad_norm": 1.462838053703308, + "learning_rate": 5.471390101338312e-06, + "loss": 0.5336, + "step": 15464 + }, + { + "epoch": 0.6847237749465924, + "grad_norm": 1.523715853691101, + "learning_rate": 5.4700122605261745e-06, + "loss": 0.5595, + "step": 15465 + }, + { + "epoch": 0.6847680506514063, + "grad_norm": 1.4026316404342651, + "learning_rate": 5.468634527909104e-06, + "loss": 0.5718, + "step": 15466 + }, + { + "epoch": 0.6848123263562201, + "grad_norm": 1.485840916633606, + "learning_rate": 5.467256903520011e-06, + "loss": 0.5356, + "step": 15467 + }, + { + "epoch": 0.684856602061034, + "grad_norm": 1.261031150817871, + "learning_rate": 5.465879387391795e-06, + "loss": 0.54, + "step": 15468 + }, + { + "epoch": 0.684900877765848, + "grad_norm": 1.4376368522644043, + "learning_rate": 5.4645019795573574e-06, + "loss": 0.4787, + "step": 15469 + }, + { + "epoch": 0.6849451534706618, + "grad_norm": 1.4356331825256348, + "learning_rate": 5.463124680049599e-06, + "loss": 0.468, + "step": 15470 + }, + { + "epoch": 0.6849894291754757, + "grad_norm": 1.4713082313537598, + "learning_rate": 5.4617474889014074e-06, + "loss": 0.4785, + "step": 15471 + }, + { + "epoch": 0.6850337048802896, + "grad_norm": 1.5368903875350952, + "learning_rate": 5.460370406145688e-06, + "loss": 0.588, + "step": 15472 + }, + { + "epoch": 0.6850779805851034, + "grad_norm": 1.759045124053955, + "learning_rate": 5.458993431815323e-06, + "loss": 0.6032, + "step": 15473 + }, + { + "epoch": 0.6851222562899173, + "grad_norm": 1.4032766819000244, + "learning_rate": 5.4576165659432e-06, + "loss": 0.5511, + "step": 15474 + }, + { + "epoch": 0.6851665319947312, + "grad_norm": 2.156440496444702, + "learning_rate": 5.45623980856221e-06, + "loss": 0.6384, + "step": 15475 + }, + { + "epoch": 0.6852108076995451, + "grad_norm": 1.32929527759552, + "learning_rate": 5.454863159705233e-06, + "loss": 0.5436, + "step": 15476 + }, + { + "epoch": 0.6852550834043589, + "grad_norm": 1.2694376707077026, + "learning_rate": 5.453486619405147e-06, + "loss": 0.5282, + "step": 15477 + }, + { + "epoch": 0.6852993591091728, + "grad_norm": 1.4184519052505493, + "learning_rate": 5.452110187694833e-06, + "loss": 0.5322, + "step": 15478 + }, + { + "epoch": 0.6853436348139867, + "grad_norm": 1.6499097347259521, + "learning_rate": 5.450733864607163e-06, + "loss": 0.5253, + "step": 15479 + }, + { + "epoch": 0.6853879105188005, + "grad_norm": 1.321997880935669, + "learning_rate": 5.449357650175011e-06, + "loss": 0.5012, + "step": 15480 + }, + { + "epoch": 0.6854321862236145, + "grad_norm": 1.5175211429595947, + "learning_rate": 5.447981544431242e-06, + "loss": 0.5086, + "step": 15481 + }, + { + "epoch": 0.6854764619284284, + "grad_norm": 1.4873569011688232, + "learning_rate": 5.446605547408732e-06, + "loss": 0.5479, + "step": 15482 + }, + { + "epoch": 0.6855207376332422, + "grad_norm": 1.638076663017273, + "learning_rate": 5.445229659140341e-06, + "loss": 0.6276, + "step": 15483 + }, + { + "epoch": 0.6855650133380561, + "grad_norm": 1.7875405550003052, + "learning_rate": 5.443853879658927e-06, + "loss": 0.5661, + "step": 15484 + }, + { + "epoch": 0.68560928904287, + "grad_norm": 1.3108985424041748, + "learning_rate": 5.442478208997357e-06, + "loss": 0.5645, + "step": 15485 + }, + { + "epoch": 0.6856535647476838, + "grad_norm": 1.3983428478240967, + "learning_rate": 5.4411026471884854e-06, + "loss": 0.5284, + "step": 15486 + }, + { + "epoch": 0.6856978404524977, + "grad_norm": 1.5409163236618042, + "learning_rate": 5.439727194265163e-06, + "loss": 0.5438, + "step": 15487 + }, + { + "epoch": 0.6857421161573116, + "grad_norm": 1.7248413562774658, + "learning_rate": 5.438351850260244e-06, + "loss": 0.5885, + "step": 15488 + }, + { + "epoch": 0.6857863918621254, + "grad_norm": 1.5629830360412598, + "learning_rate": 5.436976615206577e-06, + "loss": 0.5884, + "step": 15489 + }, + { + "epoch": 0.6858306675669393, + "grad_norm": 1.5375210046768188, + "learning_rate": 5.435601489137009e-06, + "loss": 0.5576, + "step": 15490 + }, + { + "epoch": 0.6858749432717532, + "grad_norm": 1.605114221572876, + "learning_rate": 5.434226472084379e-06, + "loss": 0.5199, + "step": 15491 + }, + { + "epoch": 0.685919218976567, + "grad_norm": 1.3710891008377075, + "learning_rate": 5.432851564081534e-06, + "loss": 0.5274, + "step": 15492 + }, + { + "epoch": 0.685963494681381, + "grad_norm": 1.3888299465179443, + "learning_rate": 5.431476765161312e-06, + "loss": 0.5889, + "step": 15493 + }, + { + "epoch": 0.6860077703861949, + "grad_norm": 1.4555641412734985, + "learning_rate": 5.430102075356545e-06, + "loss": 0.5143, + "step": 15494 + }, + { + "epoch": 0.6860520460910087, + "grad_norm": 1.2195147275924683, + "learning_rate": 5.428727494700071e-06, + "loss": 0.4741, + "step": 15495 + }, + { + "epoch": 0.6860963217958226, + "grad_norm": 1.8029981851577759, + "learning_rate": 5.427353023224724e-06, + "loss": 0.5479, + "step": 15496 + }, + { + "epoch": 0.6861405975006365, + "grad_norm": 1.388878583908081, + "learning_rate": 5.425978660963322e-06, + "loss": 0.5263, + "step": 15497 + }, + { + "epoch": 0.6861848732054503, + "grad_norm": 1.357477068901062, + "learning_rate": 5.424604407948691e-06, + "loss": 0.4676, + "step": 15498 + }, + { + "epoch": 0.6862291489102642, + "grad_norm": 1.4073091745376587, + "learning_rate": 5.423230264213664e-06, + "loss": 0.5755, + "step": 15499 + }, + { + "epoch": 0.6862734246150781, + "grad_norm": 1.689139723777771, + "learning_rate": 5.4218562297910535e-06, + "loss": 0.5947, + "step": 15500 + }, + { + "epoch": 0.6863177003198919, + "grad_norm": 2.03369402885437, + "learning_rate": 5.420482304713677e-06, + "loss": 0.623, + "step": 15501 + }, + { + "epoch": 0.6863619760247058, + "grad_norm": 1.4935582876205444, + "learning_rate": 5.419108489014356e-06, + "loss": 0.5636, + "step": 15502 + }, + { + "epoch": 0.6864062517295197, + "grad_norm": 1.7387328147888184, + "learning_rate": 5.417734782725896e-06, + "loss": 0.5813, + "step": 15503 + }, + { + "epoch": 0.6864505274343335, + "grad_norm": 1.4258488416671753, + "learning_rate": 5.4163611858811096e-06, + "loss": 0.5406, + "step": 15504 + }, + { + "epoch": 0.6864948031391475, + "grad_norm": 1.6190787553787231, + "learning_rate": 5.414987698512809e-06, + "loss": 0.5379, + "step": 15505 + }, + { + "epoch": 0.6865390788439614, + "grad_norm": 1.7256799936294556, + "learning_rate": 5.413614320653791e-06, + "loss": 0.5935, + "step": 15506 + }, + { + "epoch": 0.6865833545487752, + "grad_norm": 1.528877854347229, + "learning_rate": 5.4122410523368615e-06, + "loss": 0.5985, + "step": 15507 + }, + { + "epoch": 0.6866276302535891, + "grad_norm": 1.4292213916778564, + "learning_rate": 5.410867893594814e-06, + "loss": 0.5645, + "step": 15508 + }, + { + "epoch": 0.686671905958403, + "grad_norm": 1.3683985471725464, + "learning_rate": 5.409494844460455e-06, + "loss": 0.5461, + "step": 15509 + }, + { + "epoch": 0.6867161816632168, + "grad_norm": 1.6849385499954224, + "learning_rate": 5.408121904966574e-06, + "loss": 0.5826, + "step": 15510 + }, + { + "epoch": 0.6867604573680307, + "grad_norm": 1.8506348133087158, + "learning_rate": 5.406749075145957e-06, + "loss": 0.6166, + "step": 15511 + }, + { + "epoch": 0.6868047330728446, + "grad_norm": 1.6958519220352173, + "learning_rate": 5.4053763550314044e-06, + "loss": 0.6118, + "step": 15512 + }, + { + "epoch": 0.6868490087776585, + "grad_norm": 1.6348512172698975, + "learning_rate": 5.4040037446556944e-06, + "loss": 0.5799, + "step": 15513 + }, + { + "epoch": 0.6868932844824723, + "grad_norm": 1.2702279090881348, + "learning_rate": 5.402631244051614e-06, + "loss": 0.5567, + "step": 15514 + }, + { + "epoch": 0.6869375601872862, + "grad_norm": 1.2136067152023315, + "learning_rate": 5.401258853251943e-06, + "loss": 0.4879, + "step": 15515 + }, + { + "epoch": 0.6869818358921002, + "grad_norm": 1.542225956916809, + "learning_rate": 5.399886572289459e-06, + "loss": 0.6224, + "step": 15516 + }, + { + "epoch": 0.687026111596914, + "grad_norm": 1.5522500276565552, + "learning_rate": 5.3985144011969395e-06, + "loss": 0.5732, + "step": 15517 + }, + { + "epoch": 0.6870703873017279, + "grad_norm": 1.5086591243743896, + "learning_rate": 5.3971423400071534e-06, + "loss": 0.5372, + "step": 15518 + }, + { + "epoch": 0.6871146630065418, + "grad_norm": 1.517048954963684, + "learning_rate": 5.3957703887528766e-06, + "loss": 0.5532, + "step": 15519 + }, + { + "epoch": 0.6871589387113556, + "grad_norm": 1.4811363220214844, + "learning_rate": 5.394398547466877e-06, + "loss": 0.4868, + "step": 15520 + }, + { + "epoch": 0.6872032144161695, + "grad_norm": 2.2262065410614014, + "learning_rate": 5.393026816181914e-06, + "loss": 0.6146, + "step": 15521 + }, + { + "epoch": 0.6872474901209834, + "grad_norm": 1.7823072671890259, + "learning_rate": 5.391655194930758e-06, + "loss": 0.6245, + "step": 15522 + }, + { + "epoch": 0.6872917658257972, + "grad_norm": 1.411363124847412, + "learning_rate": 5.390283683746165e-06, + "loss": 0.5158, + "step": 15523 + }, + { + "epoch": 0.6873360415306111, + "grad_norm": 1.4341620206832886, + "learning_rate": 5.388912282660893e-06, + "loss": 0.5516, + "step": 15524 + }, + { + "epoch": 0.687380317235425, + "grad_norm": 1.4360820055007935, + "learning_rate": 5.387540991707697e-06, + "loss": 0.5507, + "step": 15525 + }, + { + "epoch": 0.6874245929402388, + "grad_norm": 1.5284534692764282, + "learning_rate": 5.386169810919329e-06, + "loss": 0.4885, + "step": 15526 + }, + { + "epoch": 0.6874688686450527, + "grad_norm": 1.5658986568450928, + "learning_rate": 5.384798740328538e-06, + "loss": 0.5521, + "step": 15527 + }, + { + "epoch": 0.6875131443498667, + "grad_norm": 1.4455546140670776, + "learning_rate": 5.383427779968066e-06, + "loss": 0.5552, + "step": 15528 + }, + { + "epoch": 0.6875574200546805, + "grad_norm": 1.4548590183258057, + "learning_rate": 5.382056929870669e-06, + "loss": 0.5645, + "step": 15529 + }, + { + "epoch": 0.6876016957594944, + "grad_norm": 1.5347042083740234, + "learning_rate": 5.38068619006908e-06, + "loss": 0.5667, + "step": 15530 + }, + { + "epoch": 0.6876459714643083, + "grad_norm": 1.5029903650283813, + "learning_rate": 5.379315560596038e-06, + "loss": 0.4948, + "step": 15531 + }, + { + "epoch": 0.6876902471691221, + "grad_norm": 1.6091458797454834, + "learning_rate": 5.377945041484285e-06, + "loss": 0.5968, + "step": 15532 + }, + { + "epoch": 0.687734522873936, + "grad_norm": 1.3530964851379395, + "learning_rate": 5.37657463276655e-06, + "loss": 0.4677, + "step": 15533 + }, + { + "epoch": 0.6877787985787499, + "grad_norm": 1.8048210144042969, + "learning_rate": 5.3752043344755654e-06, + "loss": 0.6014, + "step": 15534 + }, + { + "epoch": 0.6878230742835637, + "grad_norm": 1.8942731618881226, + "learning_rate": 5.373834146644061e-06, + "loss": 0.5891, + "step": 15535 + }, + { + "epoch": 0.6878673499883776, + "grad_norm": 1.2961093187332153, + "learning_rate": 5.372464069304761e-06, + "loss": 0.545, + "step": 15536 + }, + { + "epoch": 0.6879116256931915, + "grad_norm": 1.5141639709472656, + "learning_rate": 5.371094102490388e-06, + "loss": 0.5111, + "step": 15537 + }, + { + "epoch": 0.6879559013980053, + "grad_norm": 1.4096553325653076, + "learning_rate": 5.3697242462336605e-06, + "loss": 0.5381, + "step": 15538 + }, + { + "epoch": 0.6880001771028192, + "grad_norm": 1.2530953884124756, + "learning_rate": 5.3683545005673035e-06, + "loss": 0.5, + "step": 15539 + }, + { + "epoch": 0.6880444528076332, + "grad_norm": 1.47003173828125, + "learning_rate": 5.366984865524026e-06, + "loss": 0.5338, + "step": 15540 + }, + { + "epoch": 0.688088728512447, + "grad_norm": 1.4836552143096924, + "learning_rate": 5.3656153411365404e-06, + "loss": 0.5457, + "step": 15541 + }, + { + "epoch": 0.6881330042172609, + "grad_norm": 1.7352921962738037, + "learning_rate": 5.3642459274375625e-06, + "loss": 0.5494, + "step": 15542 + }, + { + "epoch": 0.6881772799220748, + "grad_norm": 2.0073392391204834, + "learning_rate": 5.3628766244597964e-06, + "loss": 0.6117, + "step": 15543 + }, + { + "epoch": 0.6882215556268886, + "grad_norm": 1.7008743286132812, + "learning_rate": 5.36150743223595e-06, + "loss": 0.5623, + "step": 15544 + }, + { + "epoch": 0.6882658313317025, + "grad_norm": 1.5689812898635864, + "learning_rate": 5.360138350798714e-06, + "loss": 0.5403, + "step": 15545 + }, + { + "epoch": 0.6883101070365164, + "grad_norm": 1.397831916809082, + "learning_rate": 5.358769380180798e-06, + "loss": 0.5174, + "step": 15546 + }, + { + "epoch": 0.6883543827413303, + "grad_norm": 1.3996343612670898, + "learning_rate": 5.357400520414898e-06, + "loss": 0.539, + "step": 15547 + }, + { + "epoch": 0.6883986584461441, + "grad_norm": 1.6675186157226562, + "learning_rate": 5.356031771533703e-06, + "loss": 0.5747, + "step": 15548 + }, + { + "epoch": 0.688442934150958, + "grad_norm": 1.8912817239761353, + "learning_rate": 5.354663133569911e-06, + "loss": 0.5958, + "step": 15549 + }, + { + "epoch": 0.6884872098557719, + "grad_norm": 1.795344591140747, + "learning_rate": 5.353294606556208e-06, + "loss": 0.6117, + "step": 15550 + }, + { + "epoch": 0.6885314855605857, + "grad_norm": 1.412593126296997, + "learning_rate": 5.351926190525274e-06, + "loss": 0.5568, + "step": 15551 + }, + { + "epoch": 0.6885757612653997, + "grad_norm": 1.367200493812561, + "learning_rate": 5.3505578855098085e-06, + "loss": 0.5664, + "step": 15552 + }, + { + "epoch": 0.6886200369702136, + "grad_norm": 1.8240649700164795, + "learning_rate": 5.3491896915424775e-06, + "loss": 0.6302, + "step": 15553 + }, + { + "epoch": 0.6886643126750274, + "grad_norm": 1.5514538288116455, + "learning_rate": 5.347821608655963e-06, + "loss": 0.5834, + "step": 15554 + }, + { + "epoch": 0.6887085883798413, + "grad_norm": 1.6249632835388184, + "learning_rate": 5.346453636882939e-06, + "loss": 0.568, + "step": 15555 + }, + { + "epoch": 0.6887528640846552, + "grad_norm": 1.8153194189071655, + "learning_rate": 5.345085776256083e-06, + "loss": 0.5833, + "step": 15556 + }, + { + "epoch": 0.688797139789469, + "grad_norm": 1.7857078313827515, + "learning_rate": 5.3437180268080645e-06, + "loss": 0.6251, + "step": 15557 + }, + { + "epoch": 0.6888414154942829, + "grad_norm": 1.3983300924301147, + "learning_rate": 5.342350388571546e-06, + "loss": 0.5591, + "step": 15558 + }, + { + "epoch": 0.6888856911990968, + "grad_norm": 1.6671024560928345, + "learning_rate": 5.340982861579199e-06, + "loss": 0.5935, + "step": 15559 + }, + { + "epoch": 0.6889299669039106, + "grad_norm": 1.4672024250030518, + "learning_rate": 5.339615445863684e-06, + "loss": 0.5104, + "step": 15560 + }, + { + "epoch": 0.6889742426087245, + "grad_norm": 1.3488627672195435, + "learning_rate": 5.338248141457658e-06, + "loss": 0.4935, + "step": 15561 + }, + { + "epoch": 0.6890185183135384, + "grad_norm": 1.4746218919754028, + "learning_rate": 5.336880948393781e-06, + "loss": 0.5605, + "step": 15562 + }, + { + "epoch": 0.6890627940183522, + "grad_norm": 1.4167495965957642, + "learning_rate": 5.335513866704704e-06, + "loss": 0.4854, + "step": 15563 + }, + { + "epoch": 0.6891070697231662, + "grad_norm": 1.514102816581726, + "learning_rate": 5.334146896423081e-06, + "loss": 0.541, + "step": 15564 + }, + { + "epoch": 0.6891513454279801, + "grad_norm": 1.5508313179016113, + "learning_rate": 5.332780037581558e-06, + "loss": 0.5374, + "step": 15565 + }, + { + "epoch": 0.6891956211327939, + "grad_norm": 1.7595915794372559, + "learning_rate": 5.331413290212786e-06, + "loss": 0.6229, + "step": 15566 + }, + { + "epoch": 0.6892398968376078, + "grad_norm": 1.4182922840118408, + "learning_rate": 5.330046654349406e-06, + "loss": 0.5421, + "step": 15567 + }, + { + "epoch": 0.6892841725424217, + "grad_norm": 1.5399893522262573, + "learning_rate": 5.328680130024056e-06, + "loss": 0.5725, + "step": 15568 + }, + { + "epoch": 0.6893284482472355, + "grad_norm": 1.4106730222702026, + "learning_rate": 5.32731371726938e-06, + "loss": 0.5572, + "step": 15569 + }, + { + "epoch": 0.6893727239520494, + "grad_norm": 1.4113719463348389, + "learning_rate": 5.325947416118012e-06, + "loss": 0.5402, + "step": 15570 + }, + { + "epoch": 0.6894169996568633, + "grad_norm": 2.1701157093048096, + "learning_rate": 5.324581226602585e-06, + "loss": 0.5916, + "step": 15571 + }, + { + "epoch": 0.6894612753616771, + "grad_norm": 1.5733730792999268, + "learning_rate": 5.3232151487557275e-06, + "loss": 0.5261, + "step": 15572 + }, + { + "epoch": 0.689505551066491, + "grad_norm": 2.1702141761779785, + "learning_rate": 5.321849182610069e-06, + "loss": 0.5613, + "step": 15573 + }, + { + "epoch": 0.6895498267713049, + "grad_norm": 1.2429190874099731, + "learning_rate": 5.3204833281982335e-06, + "loss": 0.4986, + "step": 15574 + }, + { + "epoch": 0.6895941024761187, + "grad_norm": 2.1571857929229736, + "learning_rate": 5.31911758555284e-06, + "loss": 0.6032, + "step": 15575 + }, + { + "epoch": 0.6896383781809327, + "grad_norm": 1.3126940727233887, + "learning_rate": 5.3177519547065146e-06, + "loss": 0.4928, + "step": 15576 + }, + { + "epoch": 0.6896826538857466, + "grad_norm": 1.5741878747940063, + "learning_rate": 5.316386435691871e-06, + "loss": 0.5625, + "step": 15577 + }, + { + "epoch": 0.6897269295905604, + "grad_norm": 1.4714666604995728, + "learning_rate": 5.315021028541521e-06, + "loss": 0.513, + "step": 15578 + }, + { + "epoch": 0.6897712052953743, + "grad_norm": 2.031662702560425, + "learning_rate": 5.313655733288083e-06, + "loss": 0.5523, + "step": 15579 + }, + { + "epoch": 0.6898154810001882, + "grad_norm": 1.9014949798583984, + "learning_rate": 5.312290549964162e-06, + "loss": 0.6063, + "step": 15580 + }, + { + "epoch": 0.6898597567050021, + "grad_norm": 1.5539295673370361, + "learning_rate": 5.310925478602365e-06, + "loss": 0.5793, + "step": 15581 + }, + { + "epoch": 0.6899040324098159, + "grad_norm": 1.5621600151062012, + "learning_rate": 5.309560519235293e-06, + "loss": 0.57, + "step": 15582 + }, + { + "epoch": 0.6899483081146298, + "grad_norm": 1.5036287307739258, + "learning_rate": 5.3081956718955505e-06, + "loss": 0.5398, + "step": 15583 + }, + { + "epoch": 0.6899925838194437, + "grad_norm": 1.7233103513717651, + "learning_rate": 5.306830936615733e-06, + "loss": 0.5155, + "step": 15584 + }, + { + "epoch": 0.6900368595242575, + "grad_norm": 1.4100137948989868, + "learning_rate": 5.305466313428435e-06, + "loss": 0.5079, + "step": 15585 + }, + { + "epoch": 0.6900811352290714, + "grad_norm": 1.2972490787506104, + "learning_rate": 5.304101802366255e-06, + "loss": 0.5245, + "step": 15586 + }, + { + "epoch": 0.6901254109338854, + "grad_norm": 1.7951425313949585, + "learning_rate": 5.302737403461778e-06, + "loss": 0.5747, + "step": 15587 + }, + { + "epoch": 0.6901696866386992, + "grad_norm": 1.387732982635498, + "learning_rate": 5.301373116747592e-06, + "loss": 0.5059, + "step": 15588 + }, + { + "epoch": 0.6902139623435131, + "grad_norm": 1.375525712966919, + "learning_rate": 5.3000089422562865e-06, + "loss": 0.5319, + "step": 15589 + }, + { + "epoch": 0.690258238048327, + "grad_norm": 1.2533782720565796, + "learning_rate": 5.29864488002044e-06, + "loss": 0.528, + "step": 15590 + }, + { + "epoch": 0.6903025137531408, + "grad_norm": 1.6336480379104614, + "learning_rate": 5.297280930072632e-06, + "loss": 0.5497, + "step": 15591 + }, + { + "epoch": 0.6903467894579547, + "grad_norm": 1.6604152917861938, + "learning_rate": 5.295917092445442e-06, + "loss": 0.6219, + "step": 15592 + }, + { + "epoch": 0.6903910651627686, + "grad_norm": 1.7463628053665161, + "learning_rate": 5.29455336717144e-06, + "loss": 0.6051, + "step": 15593 + }, + { + "epoch": 0.6904353408675824, + "grad_norm": 1.3430161476135254, + "learning_rate": 5.2931897542832015e-06, + "loss": 0.4923, + "step": 15594 + }, + { + "epoch": 0.6904796165723963, + "grad_norm": 1.2330340147018433, + "learning_rate": 5.291826253813288e-06, + "loss": 0.5025, + "step": 15595 + }, + { + "epoch": 0.6905238922772102, + "grad_norm": 1.650152564048767, + "learning_rate": 5.2904628657942745e-06, + "loss": 0.5005, + "step": 15596 + }, + { + "epoch": 0.690568167982024, + "grad_norm": 1.423224925994873, + "learning_rate": 5.2890995902587214e-06, + "loss": 0.5898, + "step": 15597 + }, + { + "epoch": 0.6906124436868379, + "grad_norm": 1.576090931892395, + "learning_rate": 5.2877364272391855e-06, + "loss": 0.5492, + "step": 15598 + }, + { + "epoch": 0.6906567193916519, + "grad_norm": 1.475226879119873, + "learning_rate": 5.28637337676823e-06, + "loss": 0.5658, + "step": 15599 + }, + { + "epoch": 0.6907009950964657, + "grad_norm": 1.580161213874817, + "learning_rate": 5.285010438878415e-06, + "loss": 0.6064, + "step": 15600 + }, + { + "epoch": 0.6907452708012796, + "grad_norm": 1.6924268007278442, + "learning_rate": 5.283647613602276e-06, + "loss": 0.5705, + "step": 15601 + }, + { + "epoch": 0.6907895465060935, + "grad_norm": 1.321662425994873, + "learning_rate": 5.2822849009723795e-06, + "loss": 0.4484, + "step": 15602 + }, + { + "epoch": 0.6908338222109073, + "grad_norm": 2.2547433376312256, + "learning_rate": 5.280922301021267e-06, + "loss": 0.5235, + "step": 15603 + }, + { + "epoch": 0.6908780979157212, + "grad_norm": 1.48110830783844, + "learning_rate": 5.279559813781482e-06, + "loss": 0.5208, + "step": 15604 + }, + { + "epoch": 0.6909223736205351, + "grad_norm": 1.3966628313064575, + "learning_rate": 5.278197439285564e-06, + "loss": 0.5198, + "step": 15605 + }, + { + "epoch": 0.6909666493253489, + "grad_norm": 1.691296100616455, + "learning_rate": 5.276835177566059e-06, + "loss": 0.546, + "step": 15606 + }, + { + "epoch": 0.6910109250301628, + "grad_norm": 1.8324025869369507, + "learning_rate": 5.2754730286555e-06, + "loss": 0.5891, + "step": 15607 + }, + { + "epoch": 0.6910552007349767, + "grad_norm": 1.3666497468948364, + "learning_rate": 5.2741109925864205e-06, + "loss": 0.5367, + "step": 15608 + }, + { + "epoch": 0.6910994764397905, + "grad_norm": 1.396309494972229, + "learning_rate": 5.272749069391353e-06, + "loss": 0.6048, + "step": 15609 + }, + { + "epoch": 0.6911437521446044, + "grad_norm": 1.40304434299469, + "learning_rate": 5.271387259102824e-06, + "loss": 0.4871, + "step": 15610 + }, + { + "epoch": 0.6911880278494184, + "grad_norm": 1.7482026815414429, + "learning_rate": 5.270025561753357e-06, + "loss": 0.6321, + "step": 15611 + }, + { + "epoch": 0.6912323035542322, + "grad_norm": 1.6531503200531006, + "learning_rate": 5.268663977375481e-06, + "loss": 0.5952, + "step": 15612 + }, + { + "epoch": 0.6912765792590461, + "grad_norm": 1.5198476314544678, + "learning_rate": 5.267302506001714e-06, + "loss": 0.5401, + "step": 15613 + }, + { + "epoch": 0.69132085496386, + "grad_norm": 1.6056023836135864, + "learning_rate": 5.265941147664574e-06, + "loss": 0.5717, + "step": 15614 + }, + { + "epoch": 0.6913651306686738, + "grad_norm": 1.5687026977539062, + "learning_rate": 5.26457990239657e-06, + "loss": 0.5732, + "step": 15615 + }, + { + "epoch": 0.6914094063734877, + "grad_norm": 1.7269070148468018, + "learning_rate": 5.263218770230224e-06, + "loss": 0.5763, + "step": 15616 + }, + { + "epoch": 0.6914536820783016, + "grad_norm": 1.519404649734497, + "learning_rate": 5.261857751198041e-06, + "loss": 0.5666, + "step": 15617 + }, + { + "epoch": 0.6914979577831155, + "grad_norm": 1.328827977180481, + "learning_rate": 5.260496845332527e-06, + "loss": 0.4917, + "step": 15618 + }, + { + "epoch": 0.6915422334879293, + "grad_norm": 1.1811615228652954, + "learning_rate": 5.259136052666187e-06, + "loss": 0.4722, + "step": 15619 + }, + { + "epoch": 0.6915865091927432, + "grad_norm": 1.3780415058135986, + "learning_rate": 5.2577753732315214e-06, + "loss": 0.5117, + "step": 15620 + }, + { + "epoch": 0.6916307848975571, + "grad_norm": 1.569734811782837, + "learning_rate": 5.256414807061028e-06, + "loss": 0.5657, + "step": 15621 + }, + { + "epoch": 0.6916750606023709, + "grad_norm": 1.94149911403656, + "learning_rate": 5.255054354187207e-06, + "loss": 0.5496, + "step": 15622 + }, + { + "epoch": 0.6917193363071849, + "grad_norm": 1.6618030071258545, + "learning_rate": 5.25369401464255e-06, + "loss": 0.6277, + "step": 15623 + }, + { + "epoch": 0.6917636120119988, + "grad_norm": 1.4185174703598022, + "learning_rate": 5.252333788459545e-06, + "loss": 0.4963, + "step": 15624 + }, + { + "epoch": 0.6918078877168126, + "grad_norm": 1.7147784233093262, + "learning_rate": 5.250973675670679e-06, + "loss": 0.5367, + "step": 15625 + }, + { + "epoch": 0.6918521634216265, + "grad_norm": 1.5170235633850098, + "learning_rate": 5.249613676308445e-06, + "loss": 0.5268, + "step": 15626 + }, + { + "epoch": 0.6918964391264404, + "grad_norm": 1.347515344619751, + "learning_rate": 5.2482537904053185e-06, + "loss": 0.4968, + "step": 15627 + }, + { + "epoch": 0.6919407148312542, + "grad_norm": 1.4954109191894531, + "learning_rate": 5.246894017993783e-06, + "loss": 0.6478, + "step": 15628 + }, + { + "epoch": 0.6919849905360681, + "grad_norm": 1.5349297523498535, + "learning_rate": 5.245534359106313e-06, + "loss": 0.5297, + "step": 15629 + }, + { + "epoch": 0.692029266240882, + "grad_norm": 1.4106789827346802, + "learning_rate": 5.244174813775382e-06, + "loss": 0.532, + "step": 15630 + }, + { + "epoch": 0.6920735419456958, + "grad_norm": 1.7317250967025757, + "learning_rate": 5.2428153820334655e-06, + "loss": 0.5962, + "step": 15631 + }, + { + "epoch": 0.6921178176505097, + "grad_norm": 1.2355482578277588, + "learning_rate": 5.241456063913026e-06, + "loss": 0.4907, + "step": 15632 + }, + { + "epoch": 0.6921620933553236, + "grad_norm": 1.4627482891082764, + "learning_rate": 5.240096859446538e-06, + "loss": 0.5213, + "step": 15633 + }, + { + "epoch": 0.6922063690601374, + "grad_norm": 1.4964808225631714, + "learning_rate": 5.23873776866646e-06, + "loss": 0.5964, + "step": 15634 + }, + { + "epoch": 0.6922506447649514, + "grad_norm": 1.4903197288513184, + "learning_rate": 5.237378791605249e-06, + "loss": 0.5305, + "step": 15635 + }, + { + "epoch": 0.6922949204697653, + "grad_norm": 1.8265193700790405, + "learning_rate": 5.236019928295373e-06, + "loss": 0.6289, + "step": 15636 + }, + { + "epoch": 0.6923391961745791, + "grad_norm": 1.326485276222229, + "learning_rate": 5.23466117876928e-06, + "loss": 0.4815, + "step": 15637 + }, + { + "epoch": 0.692383471879393, + "grad_norm": 1.5044208765029907, + "learning_rate": 5.233302543059426e-06, + "loss": 0.5634, + "step": 15638 + }, + { + "epoch": 0.6924277475842069, + "grad_norm": 1.7524173259735107, + "learning_rate": 5.231944021198261e-06, + "loss": 0.5794, + "step": 15639 + }, + { + "epoch": 0.6924720232890207, + "grad_norm": 1.295620322227478, + "learning_rate": 5.2305856132182285e-06, + "loss": 0.5051, + "step": 15640 + }, + { + "epoch": 0.6925162989938346, + "grad_norm": 1.638225793838501, + "learning_rate": 5.229227319151778e-06, + "loss": 0.6214, + "step": 15641 + }, + { + "epoch": 0.6925605746986485, + "grad_norm": 1.4901931285858154, + "learning_rate": 5.227869139031342e-06, + "loss": 0.5404, + "step": 15642 + }, + { + "epoch": 0.6926048504034623, + "grad_norm": 1.4489307403564453, + "learning_rate": 5.226511072889371e-06, + "loss": 0.5626, + "step": 15643 + }, + { + "epoch": 0.6926491261082762, + "grad_norm": 1.4096027612686157, + "learning_rate": 5.225153120758295e-06, + "loss": 0.4561, + "step": 15644 + }, + { + "epoch": 0.6926934018130901, + "grad_norm": 1.644326090812683, + "learning_rate": 5.223795282670546e-06, + "loss": 0.6023, + "step": 15645 + }, + { + "epoch": 0.6927376775179039, + "grad_norm": 1.3880168199539185, + "learning_rate": 5.222437558658562e-06, + "loss": 0.5525, + "step": 15646 + }, + { + "epoch": 0.6927819532227179, + "grad_norm": 1.5644580125808716, + "learning_rate": 5.22107994875477e-06, + "loss": 0.5143, + "step": 15647 + }, + { + "epoch": 0.6928262289275318, + "grad_norm": 1.4267734289169312, + "learning_rate": 5.219722452991585e-06, + "loss": 0.498, + "step": 15648 + }, + { + "epoch": 0.6928705046323456, + "grad_norm": 1.3690309524536133, + "learning_rate": 5.218365071401441e-06, + "loss": 0.5056, + "step": 15649 + }, + { + "epoch": 0.6929147803371595, + "grad_norm": 1.4311405420303345, + "learning_rate": 5.217007804016754e-06, + "loss": 0.5429, + "step": 15650 + }, + { + "epoch": 0.6929590560419734, + "grad_norm": 1.5060060024261475, + "learning_rate": 5.215650650869941e-06, + "loss": 0.5262, + "step": 15651 + }, + { + "epoch": 0.6930033317467873, + "grad_norm": 1.3613200187683105, + "learning_rate": 5.214293611993413e-06, + "loss": 0.5087, + "step": 15652 + }, + { + "epoch": 0.6930476074516011, + "grad_norm": 1.28973388671875, + "learning_rate": 5.212936687419587e-06, + "loss": 0.5147, + "step": 15653 + }, + { + "epoch": 0.693091883156415, + "grad_norm": 1.4322377443313599, + "learning_rate": 5.211579877180874e-06, + "loss": 0.5012, + "step": 15654 + }, + { + "epoch": 0.6931361588612289, + "grad_norm": 1.3184208869934082, + "learning_rate": 5.210223181309671e-06, + "loss": 0.5576, + "step": 15655 + }, + { + "epoch": 0.6931804345660427, + "grad_norm": 1.3571702241897583, + "learning_rate": 5.208866599838395e-06, + "loss": 0.5005, + "step": 15656 + }, + { + "epoch": 0.6932247102708566, + "grad_norm": 2.368947744369507, + "learning_rate": 5.207510132799436e-06, + "loss": 0.593, + "step": 15657 + }, + { + "epoch": 0.6932689859756705, + "grad_norm": 1.6367251873016357, + "learning_rate": 5.20615378022519e-06, + "loss": 0.6193, + "step": 15658 + }, + { + "epoch": 0.6933132616804843, + "grad_norm": 1.520111322402954, + "learning_rate": 5.204797542148063e-06, + "loss": 0.5704, + "step": 15659 + }, + { + "epoch": 0.6933575373852983, + "grad_norm": 1.3923532962799072, + "learning_rate": 5.2034414186004415e-06, + "loss": 0.5166, + "step": 15660 + }, + { + "epoch": 0.6934018130901122, + "grad_norm": 1.388649582862854, + "learning_rate": 5.202085409614717e-06, + "loss": 0.536, + "step": 15661 + }, + { + "epoch": 0.693446088794926, + "grad_norm": 1.2281274795532227, + "learning_rate": 5.200729515223271e-06, + "loss": 0.4867, + "step": 15662 + }, + { + "epoch": 0.6934903644997399, + "grad_norm": 1.3099905252456665, + "learning_rate": 5.199373735458497e-06, + "loss": 0.5167, + "step": 15663 + }, + { + "epoch": 0.6935346402045538, + "grad_norm": 1.3551404476165771, + "learning_rate": 5.198018070352773e-06, + "loss": 0.4746, + "step": 15664 + }, + { + "epoch": 0.6935789159093676, + "grad_norm": 1.484626293182373, + "learning_rate": 5.196662519938478e-06, + "loss": 0.5463, + "step": 15665 + }, + { + "epoch": 0.6936231916141815, + "grad_norm": 1.76791512966156, + "learning_rate": 5.195307084247987e-06, + "loss": 0.5469, + "step": 15666 + }, + { + "epoch": 0.6936674673189954, + "grad_norm": 1.6954611539840698, + "learning_rate": 5.1939517633136734e-06, + "loss": 0.5685, + "step": 15667 + }, + { + "epoch": 0.6937117430238092, + "grad_norm": 1.4016027450561523, + "learning_rate": 5.192596557167906e-06, + "loss": 0.5446, + "step": 15668 + }, + { + "epoch": 0.6937560187286231, + "grad_norm": 1.4814534187316895, + "learning_rate": 5.191241465843057e-06, + "loss": 0.5171, + "step": 15669 + }, + { + "epoch": 0.693800294433437, + "grad_norm": 1.297721028327942, + "learning_rate": 5.189886489371494e-06, + "loss": 0.4936, + "step": 15670 + }, + { + "epoch": 0.6938445701382508, + "grad_norm": 1.6319189071655273, + "learning_rate": 5.188531627785573e-06, + "loss": 0.5891, + "step": 15671 + }, + { + "epoch": 0.6938888458430648, + "grad_norm": 1.7651567459106445, + "learning_rate": 5.187176881117652e-06, + "loss": 0.5405, + "step": 15672 + }, + { + "epoch": 0.6939331215478787, + "grad_norm": 1.2065879106521606, + "learning_rate": 5.185822249400099e-06, + "loss": 0.4647, + "step": 15673 + }, + { + "epoch": 0.6939773972526925, + "grad_norm": 1.9002728462219238, + "learning_rate": 5.18446773266526e-06, + "loss": 0.5587, + "step": 15674 + }, + { + "epoch": 0.6940216729575064, + "grad_norm": 1.4440373182296753, + "learning_rate": 5.183113330945488e-06, + "loss": 0.5414, + "step": 15675 + }, + { + "epoch": 0.6940659486623203, + "grad_norm": 2.3743069171905518, + "learning_rate": 5.1817590442731336e-06, + "loss": 0.6401, + "step": 15676 + }, + { + "epoch": 0.6941102243671341, + "grad_norm": 2.8852102756500244, + "learning_rate": 5.1804048726805404e-06, + "loss": 0.5728, + "step": 15677 + }, + { + "epoch": 0.694154500071948, + "grad_norm": 1.3785908222198486, + "learning_rate": 5.179050816200048e-06, + "loss": 0.4997, + "step": 15678 + }, + { + "epoch": 0.6941987757767619, + "grad_norm": 1.483933925628662, + "learning_rate": 5.1776968748640065e-06, + "loss": 0.5555, + "step": 15679 + }, + { + "epoch": 0.6942430514815757, + "grad_norm": 2.189072608947754, + "learning_rate": 5.1763430487047485e-06, + "loss": 0.7398, + "step": 15680 + }, + { + "epoch": 0.6942873271863896, + "grad_norm": 1.4391556978225708, + "learning_rate": 5.174989337754608e-06, + "loss": 0.5261, + "step": 15681 + }, + { + "epoch": 0.6943316028912035, + "grad_norm": 1.8098101615905762, + "learning_rate": 5.173635742045915e-06, + "loss": 0.5181, + "step": 15682 + }, + { + "epoch": 0.6943758785960173, + "grad_norm": 1.3915570974349976, + "learning_rate": 5.1722822616110076e-06, + "loss": 0.5226, + "step": 15683 + }, + { + "epoch": 0.6944201543008313, + "grad_norm": 1.6784729957580566, + "learning_rate": 5.170928896482205e-06, + "loss": 0.5746, + "step": 15684 + }, + { + "epoch": 0.6944644300056452, + "grad_norm": 1.333666443824768, + "learning_rate": 5.169575646691837e-06, + "loss": 0.4499, + "step": 15685 + }, + { + "epoch": 0.694508705710459, + "grad_norm": 1.5155811309814453, + "learning_rate": 5.1682225122722206e-06, + "loss": 0.5557, + "step": 15686 + }, + { + "epoch": 0.6945529814152729, + "grad_norm": 1.5666711330413818, + "learning_rate": 5.166869493255674e-06, + "loss": 0.5923, + "step": 15687 + }, + { + "epoch": 0.6945972571200868, + "grad_norm": 2.0030486583709717, + "learning_rate": 5.165516589674511e-06, + "loss": 0.5698, + "step": 15688 + }, + { + "epoch": 0.6946415328249007, + "grad_norm": 2.066255807876587, + "learning_rate": 5.1641638015610525e-06, + "loss": 0.5934, + "step": 15689 + }, + { + "epoch": 0.6946858085297145, + "grad_norm": 1.574074149131775, + "learning_rate": 5.1628111289476025e-06, + "loss": 0.534, + "step": 15690 + }, + { + "epoch": 0.6947300842345284, + "grad_norm": 1.702372670173645, + "learning_rate": 5.161458571866471e-06, + "loss": 0.6072, + "step": 15691 + }, + { + "epoch": 0.6947743599393423, + "grad_norm": 1.6164392232894897, + "learning_rate": 5.160106130349959e-06, + "loss": 0.5642, + "step": 15692 + }, + { + "epoch": 0.6948186356441561, + "grad_norm": 1.6762758493423462, + "learning_rate": 5.158753804430374e-06, + "loss": 0.5593, + "step": 15693 + }, + { + "epoch": 0.69486291134897, + "grad_norm": 1.5236858129501343, + "learning_rate": 5.1574015941400124e-06, + "loss": 0.5793, + "step": 15694 + }, + { + "epoch": 0.694907187053784, + "grad_norm": 1.269821047782898, + "learning_rate": 5.156049499511171e-06, + "loss": 0.5027, + "step": 15695 + }, + { + "epoch": 0.6949514627585978, + "grad_norm": 1.308812141418457, + "learning_rate": 5.154697520576144e-06, + "loss": 0.4822, + "step": 15696 + }, + { + "epoch": 0.6949957384634117, + "grad_norm": 1.5907752513885498, + "learning_rate": 5.153345657367221e-06, + "loss": 0.5911, + "step": 15697 + }, + { + "epoch": 0.6950400141682256, + "grad_norm": 1.4019588232040405, + "learning_rate": 5.151993909916686e-06, + "loss": 0.5514, + "step": 15698 + }, + { + "epoch": 0.6950842898730394, + "grad_norm": 1.4974435567855835, + "learning_rate": 5.1506422782568345e-06, + "loss": 0.5775, + "step": 15699 + }, + { + "epoch": 0.6951285655778533, + "grad_norm": 1.2963656187057495, + "learning_rate": 5.149290762419943e-06, + "loss": 0.4971, + "step": 15700 + }, + { + "epoch": 0.6951728412826672, + "grad_norm": 1.9450725317001343, + "learning_rate": 5.147939362438291e-06, + "loss": 0.6071, + "step": 15701 + }, + { + "epoch": 0.695217116987481, + "grad_norm": 1.8125452995300293, + "learning_rate": 5.146588078344154e-06, + "loss": 0.5323, + "step": 15702 + }, + { + "epoch": 0.6952613926922949, + "grad_norm": 1.3838974237442017, + "learning_rate": 5.1452369101698154e-06, + "loss": 0.5186, + "step": 15703 + }, + { + "epoch": 0.6953056683971088, + "grad_norm": 1.5699909925460815, + "learning_rate": 5.143885857947537e-06, + "loss": 0.5765, + "step": 15704 + }, + { + "epoch": 0.6953499441019226, + "grad_norm": 1.6809853315353394, + "learning_rate": 5.142534921709588e-06, + "loss": 0.5865, + "step": 15705 + }, + { + "epoch": 0.6953942198067365, + "grad_norm": 1.5989723205566406, + "learning_rate": 5.1411841014882415e-06, + "loss": 0.5796, + "step": 15706 + }, + { + "epoch": 0.6954384955115505, + "grad_norm": 1.2872751951217651, + "learning_rate": 5.139833397315755e-06, + "loss": 0.4855, + "step": 15707 + }, + { + "epoch": 0.6954827712163643, + "grad_norm": 1.5235542058944702, + "learning_rate": 5.1384828092243865e-06, + "loss": 0.4951, + "step": 15708 + }, + { + "epoch": 0.6955270469211782, + "grad_norm": 1.519469976425171, + "learning_rate": 5.137132337246402e-06, + "loss": 0.5518, + "step": 15709 + }, + { + "epoch": 0.6955713226259921, + "grad_norm": 1.4105076789855957, + "learning_rate": 5.135781981414053e-06, + "loss": 0.521, + "step": 15710 + }, + { + "epoch": 0.6956155983308059, + "grad_norm": 1.4998072385787964, + "learning_rate": 5.1344317417595905e-06, + "loss": 0.5479, + "step": 15711 + }, + { + "epoch": 0.6956598740356198, + "grad_norm": 1.968817949295044, + "learning_rate": 5.133081618315263e-06, + "loss": 0.6136, + "step": 15712 + }, + { + "epoch": 0.6957041497404337, + "grad_norm": 1.343618392944336, + "learning_rate": 5.131731611113319e-06, + "loss": 0.5405, + "step": 15713 + }, + { + "epoch": 0.6957484254452475, + "grad_norm": 1.497374176979065, + "learning_rate": 5.1303817201860015e-06, + "loss": 0.5219, + "step": 15714 + }, + { + "epoch": 0.6957927011500614, + "grad_norm": 1.4999641180038452, + "learning_rate": 5.129031945565548e-06, + "loss": 0.5085, + "step": 15715 + }, + { + "epoch": 0.6958369768548753, + "grad_norm": 1.3014239072799683, + "learning_rate": 5.127682287284205e-06, + "loss": 0.5232, + "step": 15716 + }, + { + "epoch": 0.6958812525596891, + "grad_norm": 1.3183404207229614, + "learning_rate": 5.126332745374203e-06, + "loss": 0.4942, + "step": 15717 + }, + { + "epoch": 0.695925528264503, + "grad_norm": 1.405535340309143, + "learning_rate": 5.12498331986777e-06, + "loss": 0.5296, + "step": 15718 + }, + { + "epoch": 0.695969803969317, + "grad_norm": 2.3769445419311523, + "learning_rate": 5.123634010797145e-06, + "loss": 0.5184, + "step": 15719 + }, + { + "epoch": 0.6960140796741308, + "grad_norm": 1.4613840579986572, + "learning_rate": 5.1222848181945515e-06, + "loss": 0.5583, + "step": 15720 + }, + { + "epoch": 0.6960583553789447, + "grad_norm": 1.8441016674041748, + "learning_rate": 5.120935742092215e-06, + "loss": 0.637, + "step": 15721 + }, + { + "epoch": 0.6961026310837586, + "grad_norm": 1.4473024606704712, + "learning_rate": 5.119586782522355e-06, + "loss": 0.522, + "step": 15722 + }, + { + "epoch": 0.6961469067885725, + "grad_norm": 1.2831788063049316, + "learning_rate": 5.11823793951719e-06, + "loss": 0.4917, + "step": 15723 + }, + { + "epoch": 0.6961911824933863, + "grad_norm": 1.5258867740631104, + "learning_rate": 5.1168892131089376e-06, + "loss": 0.5024, + "step": 15724 + }, + { + "epoch": 0.6962354581982002, + "grad_norm": 1.5071461200714111, + "learning_rate": 5.1155406033298046e-06, + "loss": 0.5263, + "step": 15725 + }, + { + "epoch": 0.6962797339030141, + "grad_norm": 1.4315414428710938, + "learning_rate": 5.114192110212013e-06, + "loss": 0.5754, + "step": 15726 + }, + { + "epoch": 0.6963240096078279, + "grad_norm": 1.4236350059509277, + "learning_rate": 5.112843733787765e-06, + "loss": 0.5345, + "step": 15727 + }, + { + "epoch": 0.6963682853126418, + "grad_norm": 1.631818413734436, + "learning_rate": 5.111495474089261e-06, + "loss": 0.5409, + "step": 15728 + }, + { + "epoch": 0.6964125610174557, + "grad_norm": 1.5132622718811035, + "learning_rate": 5.1101473311487105e-06, + "loss": 0.5238, + "step": 15729 + }, + { + "epoch": 0.6964568367222695, + "grad_norm": 1.4925342798233032, + "learning_rate": 5.10879930499831e-06, + "loss": 0.5191, + "step": 15730 + }, + { + "epoch": 0.6965011124270835, + "grad_norm": 1.5131648778915405, + "learning_rate": 5.107451395670255e-06, + "loss": 0.5457, + "step": 15731 + }, + { + "epoch": 0.6965453881318974, + "grad_norm": 1.4816757440567017, + "learning_rate": 5.10610360319674e-06, + "loss": 0.547, + "step": 15732 + }, + { + "epoch": 0.6965896638367112, + "grad_norm": 1.5363825559616089, + "learning_rate": 5.104755927609957e-06, + "loss": 0.5247, + "step": 15733 + }, + { + "epoch": 0.6966339395415251, + "grad_norm": 1.4123963117599487, + "learning_rate": 5.103408368942091e-06, + "loss": 0.5457, + "step": 15734 + }, + { + "epoch": 0.696678215246339, + "grad_norm": 1.574959635734558, + "learning_rate": 5.102060927225325e-06, + "loss": 0.5234, + "step": 15735 + }, + { + "epoch": 0.6967224909511528, + "grad_norm": 1.4626859426498413, + "learning_rate": 5.100713602491851e-06, + "loss": 0.5522, + "step": 15736 + }, + { + "epoch": 0.6967667666559667, + "grad_norm": 1.780216097831726, + "learning_rate": 5.099366394773842e-06, + "loss": 0.5218, + "step": 15737 + }, + { + "epoch": 0.6968110423607806, + "grad_norm": 1.4003828763961792, + "learning_rate": 5.0980193041034785e-06, + "loss": 0.5486, + "step": 15738 + }, + { + "epoch": 0.6968553180655944, + "grad_norm": 1.8872603178024292, + "learning_rate": 5.096672330512926e-06, + "loss": 0.6482, + "step": 15739 + }, + { + "epoch": 0.6968995937704083, + "grad_norm": 1.6730589866638184, + "learning_rate": 5.095325474034369e-06, + "loss": 0.5736, + "step": 15740 + }, + { + "epoch": 0.6969438694752222, + "grad_norm": 1.4495360851287842, + "learning_rate": 5.093978734699968e-06, + "loss": 0.5166, + "step": 15741 + }, + { + "epoch": 0.696988145180036, + "grad_norm": 1.550349235534668, + "learning_rate": 5.092632112541891e-06, + "loss": 0.5427, + "step": 15742 + }, + { + "epoch": 0.69703242088485, + "grad_norm": 1.2289305925369263, + "learning_rate": 5.0912856075923e-06, + "loss": 0.4911, + "step": 15743 + }, + { + "epoch": 0.6970766965896639, + "grad_norm": 1.5526134967803955, + "learning_rate": 5.089939219883356e-06, + "loss": 0.5941, + "step": 15744 + }, + { + "epoch": 0.6971209722944777, + "grad_norm": 1.4433146715164185, + "learning_rate": 5.088592949447211e-06, + "loss": 0.4821, + "step": 15745 + }, + { + "epoch": 0.6971652479992916, + "grad_norm": 1.5615551471710205, + "learning_rate": 5.087246796316028e-06, + "loss": 0.5465, + "step": 15746 + }, + { + "epoch": 0.6972095237041055, + "grad_norm": 1.7221934795379639, + "learning_rate": 5.085900760521955e-06, + "loss": 0.5978, + "step": 15747 + }, + { + "epoch": 0.6972537994089193, + "grad_norm": 1.5759952068328857, + "learning_rate": 5.084554842097143e-06, + "loss": 0.5088, + "step": 15748 + }, + { + "epoch": 0.6972980751137332, + "grad_norm": 1.3684953451156616, + "learning_rate": 5.083209041073731e-06, + "loss": 0.5264, + "step": 15749 + }, + { + "epoch": 0.6973423508185471, + "grad_norm": 1.5929911136627197, + "learning_rate": 5.081863357483871e-06, + "loss": 0.4791, + "step": 15750 + }, + { + "epoch": 0.6973866265233609, + "grad_norm": 1.4660481214523315, + "learning_rate": 5.080517791359705e-06, + "loss": 0.5805, + "step": 15751 + }, + { + "epoch": 0.6974309022281748, + "grad_norm": 1.5849734544754028, + "learning_rate": 5.079172342733358e-06, + "loss": 0.5441, + "step": 15752 + }, + { + "epoch": 0.6974751779329887, + "grad_norm": 1.8832054138183594, + "learning_rate": 5.077827011636978e-06, + "loss": 0.6149, + "step": 15753 + }, + { + "epoch": 0.6975194536378025, + "grad_norm": 1.477095365524292, + "learning_rate": 5.076481798102691e-06, + "loss": 0.5382, + "step": 15754 + }, + { + "epoch": 0.6975637293426165, + "grad_norm": 1.6752263307571411, + "learning_rate": 5.075136702162622e-06, + "loss": 0.6019, + "step": 15755 + }, + { + "epoch": 0.6976080050474304, + "grad_norm": 1.6925266981124878, + "learning_rate": 5.0737917238489085e-06, + "loss": 0.5726, + "step": 15756 + }, + { + "epoch": 0.6976522807522442, + "grad_norm": 1.210157036781311, + "learning_rate": 5.072446863193668e-06, + "loss": 0.4919, + "step": 15757 + }, + { + "epoch": 0.6976965564570581, + "grad_norm": 1.7149620056152344, + "learning_rate": 5.071102120229023e-06, + "loss": 0.5621, + "step": 15758 + }, + { + "epoch": 0.697740832161872, + "grad_norm": 1.7413694858551025, + "learning_rate": 5.06975749498709e-06, + "loss": 0.6531, + "step": 15759 + }, + { + "epoch": 0.6977851078666859, + "grad_norm": 1.29939603805542, + "learning_rate": 5.068412987499984e-06, + "loss": 0.4947, + "step": 15760 + }, + { + "epoch": 0.6978293835714997, + "grad_norm": 1.566940188407898, + "learning_rate": 5.06706859779982e-06, + "loss": 0.5231, + "step": 15761 + }, + { + "epoch": 0.6978736592763136, + "grad_norm": 1.5203698873519897, + "learning_rate": 5.065724325918702e-06, + "loss": 0.5883, + "step": 15762 + }, + { + "epoch": 0.6979179349811275, + "grad_norm": 1.4038656949996948, + "learning_rate": 5.064380171888743e-06, + "loss": 0.5058, + "step": 15763 + }, + { + "epoch": 0.6979622106859413, + "grad_norm": 1.303095817565918, + "learning_rate": 5.063036135742046e-06, + "loss": 0.5585, + "step": 15764 + }, + { + "epoch": 0.6980064863907552, + "grad_norm": 1.4490734338760376, + "learning_rate": 5.061692217510708e-06, + "loss": 0.4919, + "step": 15765 + }, + { + "epoch": 0.6980507620955692, + "grad_norm": 1.361039161682129, + "learning_rate": 5.060348417226832e-06, + "loss": 0.4665, + "step": 15766 + }, + { + "epoch": 0.698095037800383, + "grad_norm": 1.3276739120483398, + "learning_rate": 5.0590047349225135e-06, + "loss": 0.4961, + "step": 15767 + }, + { + "epoch": 0.6981393135051969, + "grad_norm": 1.816831111907959, + "learning_rate": 5.057661170629844e-06, + "loss": 0.5389, + "step": 15768 + }, + { + "epoch": 0.6981835892100108, + "grad_norm": 1.8793326616287231, + "learning_rate": 5.056317724380912e-06, + "loss": 0.5898, + "step": 15769 + }, + { + "epoch": 0.6982278649148246, + "grad_norm": 1.712854266166687, + "learning_rate": 5.054974396207808e-06, + "loss": 0.5958, + "step": 15770 + }, + { + "epoch": 0.6982721406196385, + "grad_norm": 1.4240944385528564, + "learning_rate": 5.053631186142612e-06, + "loss": 0.55, + "step": 15771 + }, + { + "epoch": 0.6983164163244524, + "grad_norm": 1.7254890203475952, + "learning_rate": 5.0522880942174055e-06, + "loss": 0.5316, + "step": 15772 + }, + { + "epoch": 0.6983606920292662, + "grad_norm": 1.48475182056427, + "learning_rate": 5.050945120464273e-06, + "loss": 0.5534, + "step": 15773 + }, + { + "epoch": 0.6984049677340801, + "grad_norm": 1.5415862798690796, + "learning_rate": 5.049602264915286e-06, + "loss": 0.5215, + "step": 15774 + }, + { + "epoch": 0.698449243438894, + "grad_norm": 1.66244375705719, + "learning_rate": 5.048259527602513e-06, + "loss": 0.576, + "step": 15775 + }, + { + "epoch": 0.6984935191437078, + "grad_norm": 1.3537846803665161, + "learning_rate": 5.046916908558035e-06, + "loss": 0.5012, + "step": 15776 + }, + { + "epoch": 0.6985377948485217, + "grad_norm": 1.625072717666626, + "learning_rate": 5.0455744078139136e-06, + "loss": 0.5715, + "step": 15777 + }, + { + "epoch": 0.6985820705533357, + "grad_norm": 1.3564364910125732, + "learning_rate": 5.044232025402212e-06, + "loss": 0.5294, + "step": 15778 + }, + { + "epoch": 0.6986263462581495, + "grad_norm": 1.466951608657837, + "learning_rate": 5.042889761354994e-06, + "loss": 0.5785, + "step": 15779 + }, + { + "epoch": 0.6986706219629634, + "grad_norm": 1.57114577293396, + "learning_rate": 5.0415476157043184e-06, + "loss": 0.599, + "step": 15780 + }, + { + "epoch": 0.6987148976677773, + "grad_norm": 1.6345072984695435, + "learning_rate": 5.040205588482241e-06, + "loss": 0.575, + "step": 15781 + }, + { + "epoch": 0.6987591733725911, + "grad_norm": 1.4383713006973267, + "learning_rate": 5.03886367972081e-06, + "loss": 0.5135, + "step": 15782 + }, + { + "epoch": 0.698803449077405, + "grad_norm": 1.3678697347640991, + "learning_rate": 5.037521889452084e-06, + "loss": 0.4844, + "step": 15783 + }, + { + "epoch": 0.6988477247822189, + "grad_norm": 1.2382572889328003, + "learning_rate": 5.036180217708108e-06, + "loss": 0.5423, + "step": 15784 + }, + { + "epoch": 0.6988920004870327, + "grad_norm": 1.5111392736434937, + "learning_rate": 5.034838664520922e-06, + "loss": 0.5267, + "step": 15785 + }, + { + "epoch": 0.6989362761918466, + "grad_norm": 1.2375696897506714, + "learning_rate": 5.033497229922575e-06, + "loss": 0.4854, + "step": 15786 + }, + { + "epoch": 0.6989805518966605, + "grad_norm": 1.5282680988311768, + "learning_rate": 5.032155913945103e-06, + "loss": 0.5749, + "step": 15787 + }, + { + "epoch": 0.6990248276014743, + "grad_norm": 1.5604277849197388, + "learning_rate": 5.030814716620542e-06, + "loss": 0.4939, + "step": 15788 + }, + { + "epoch": 0.6990691033062882, + "grad_norm": 1.352829098701477, + "learning_rate": 5.029473637980926e-06, + "loss": 0.5453, + "step": 15789 + }, + { + "epoch": 0.6991133790111022, + "grad_norm": 1.323204517364502, + "learning_rate": 5.028132678058284e-06, + "loss": 0.5029, + "step": 15790 + }, + { + "epoch": 0.699157654715916, + "grad_norm": 1.6194021701812744, + "learning_rate": 5.026791836884646e-06, + "loss": 0.527, + "step": 15791 + }, + { + "epoch": 0.6992019304207299, + "grad_norm": 1.66828191280365, + "learning_rate": 5.02545111449203e-06, + "loss": 0.5896, + "step": 15792 + }, + { + "epoch": 0.6992462061255438, + "grad_norm": 1.3825011253356934, + "learning_rate": 5.024110510912469e-06, + "loss": 0.5293, + "step": 15793 + }, + { + "epoch": 0.6992904818303577, + "grad_norm": 1.6101021766662598, + "learning_rate": 5.022770026177977e-06, + "loss": 0.5353, + "step": 15794 + }, + { + "epoch": 0.6993347575351715, + "grad_norm": 1.7029091119766235, + "learning_rate": 5.021429660320565e-06, + "loss": 0.5789, + "step": 15795 + }, + { + "epoch": 0.6993790332399854, + "grad_norm": 1.6560757160186768, + "learning_rate": 5.0200894133722575e-06, + "loss": 0.6144, + "step": 15796 + }, + { + "epoch": 0.6994233089447993, + "grad_norm": 1.534859299659729, + "learning_rate": 5.018749285365059e-06, + "loss": 0.5185, + "step": 15797 + }, + { + "epoch": 0.6994675846496131, + "grad_norm": 1.424492359161377, + "learning_rate": 5.0174092763309805e-06, + "loss": 0.458, + "step": 15798 + }, + { + "epoch": 0.699511860354427, + "grad_norm": 1.483574628829956, + "learning_rate": 5.016069386302018e-06, + "loss": 0.5608, + "step": 15799 + }, + { + "epoch": 0.6995561360592409, + "grad_norm": 1.4146219491958618, + "learning_rate": 5.0147296153101835e-06, + "loss": 0.5214, + "step": 15800 + }, + { + "epoch": 0.6996004117640547, + "grad_norm": 1.4182941913604736, + "learning_rate": 5.013389963387474e-06, + "loss": 0.4895, + "step": 15801 + }, + { + "epoch": 0.6996446874688687, + "grad_norm": 1.442213773727417, + "learning_rate": 5.0120504305658805e-06, + "loss": 0.5444, + "step": 15802 + }, + { + "epoch": 0.6996889631736826, + "grad_norm": 1.755697250366211, + "learning_rate": 5.0107110168774045e-06, + "loss": 0.519, + "step": 15803 + }, + { + "epoch": 0.6997332388784964, + "grad_norm": 1.4004333019256592, + "learning_rate": 5.009371722354033e-06, + "loss": 0.5107, + "step": 15804 + }, + { + "epoch": 0.6997775145833103, + "grad_norm": 1.2747999429702759, + "learning_rate": 5.008032547027753e-06, + "loss": 0.4839, + "step": 15805 + }, + { + "epoch": 0.6998217902881242, + "grad_norm": 1.4277483224868774, + "learning_rate": 5.006693490930556e-06, + "loss": 0.5335, + "step": 15806 + }, + { + "epoch": 0.699866065992938, + "grad_norm": 1.2461636066436768, + "learning_rate": 5.005354554094417e-06, + "loss": 0.4657, + "step": 15807 + }, + { + "epoch": 0.6999103416977519, + "grad_norm": 1.281046986579895, + "learning_rate": 5.004015736551316e-06, + "loss": 0.5027, + "step": 15808 + }, + { + "epoch": 0.6999546174025658, + "grad_norm": 1.5257623195648193, + "learning_rate": 5.002677038333229e-06, + "loss": 0.5591, + "step": 15809 + }, + { + "epoch": 0.6999988931073796, + "grad_norm": 1.621433973312378, + "learning_rate": 5.001338459472136e-06, + "loss": 0.5403, + "step": 15810 + }, + { + "epoch": 0.7000431688121935, + "grad_norm": 1.4680052995681763, + "learning_rate": 5.000000000000003e-06, + "loss": 0.5464, + "step": 15811 + }, + { + "epoch": 0.7000874445170074, + "grad_norm": 1.6715435981750488, + "learning_rate": 4.998661659948795e-06, + "loss": 0.5921, + "step": 15812 + }, + { + "epoch": 0.7001317202218212, + "grad_norm": 1.496622085571289, + "learning_rate": 4.997323439350484e-06, + "loss": 0.4885, + "step": 15813 + }, + { + "epoch": 0.7001759959266352, + "grad_norm": 1.3818907737731934, + "learning_rate": 4.99598533823703e-06, + "loss": 0.504, + "step": 15814 + }, + { + "epoch": 0.7002202716314491, + "grad_norm": 1.5073723793029785, + "learning_rate": 4.994647356640393e-06, + "loss": 0.5427, + "step": 15815 + }, + { + "epoch": 0.7002645473362629, + "grad_norm": 1.247498631477356, + "learning_rate": 4.9933094945925275e-06, + "loss": 0.4913, + "step": 15816 + }, + { + "epoch": 0.7003088230410768, + "grad_norm": 1.5345827341079712, + "learning_rate": 4.991971752125389e-06, + "loss": 0.4901, + "step": 15817 + }, + { + "epoch": 0.7003530987458907, + "grad_norm": 1.3444013595581055, + "learning_rate": 4.990634129270926e-06, + "loss": 0.4988, + "step": 15818 + }, + { + "epoch": 0.7003973744507045, + "grad_norm": 1.3935449123382568, + "learning_rate": 4.989296626061084e-06, + "loss": 0.5598, + "step": 15819 + }, + { + "epoch": 0.7004416501555184, + "grad_norm": 1.556530475616455, + "learning_rate": 4.987959242527818e-06, + "loss": 0.5846, + "step": 15820 + }, + { + "epoch": 0.7004859258603323, + "grad_norm": 2.0615787506103516, + "learning_rate": 4.986621978703063e-06, + "loss": 0.606, + "step": 15821 + }, + { + "epoch": 0.7005302015651461, + "grad_norm": 1.2489209175109863, + "learning_rate": 4.985284834618757e-06, + "loss": 0.4817, + "step": 15822 + }, + { + "epoch": 0.70057447726996, + "grad_norm": 1.5671664476394653, + "learning_rate": 4.983947810306843e-06, + "loss": 0.5425, + "step": 15823 + }, + { + "epoch": 0.7006187529747739, + "grad_norm": 1.528930902481079, + "learning_rate": 4.982610905799251e-06, + "loss": 0.5896, + "step": 15824 + }, + { + "epoch": 0.7006630286795877, + "grad_norm": 1.5541343688964844, + "learning_rate": 4.981274121127913e-06, + "loss": 0.5502, + "step": 15825 + }, + { + "epoch": 0.7007073043844017, + "grad_norm": 1.3407803773880005, + "learning_rate": 4.979937456324755e-06, + "loss": 0.499, + "step": 15826 + }, + { + "epoch": 0.7007515800892156, + "grad_norm": 1.5984535217285156, + "learning_rate": 4.978600911421705e-06, + "loss": 0.462, + "step": 15827 + }, + { + "epoch": 0.7007958557940295, + "grad_norm": 1.4014699459075928, + "learning_rate": 4.977264486450682e-06, + "loss": 0.5688, + "step": 15828 + }, + { + "epoch": 0.7008401314988433, + "grad_norm": 1.2888439893722534, + "learning_rate": 4.975928181443604e-06, + "loss": 0.4864, + "step": 15829 + }, + { + "epoch": 0.7008844072036572, + "grad_norm": 1.6733131408691406, + "learning_rate": 4.974591996432395e-06, + "loss": 0.6108, + "step": 15830 + }, + { + "epoch": 0.7009286829084711, + "grad_norm": 1.5424448251724243, + "learning_rate": 4.973255931448964e-06, + "loss": 0.5975, + "step": 15831 + }, + { + "epoch": 0.7009729586132849, + "grad_norm": 1.5412110090255737, + "learning_rate": 4.971919986525218e-06, + "loss": 0.4677, + "step": 15832 + }, + { + "epoch": 0.7010172343180988, + "grad_norm": 1.693177342414856, + "learning_rate": 4.970584161693073e-06, + "loss": 0.6033, + "step": 15833 + }, + { + "epoch": 0.7010615100229127, + "grad_norm": 1.4767309427261353, + "learning_rate": 4.96924845698443e-06, + "loss": 0.4929, + "step": 15834 + }, + { + "epoch": 0.7011057857277265, + "grad_norm": 1.5090547800064087, + "learning_rate": 4.967912872431192e-06, + "loss": 0.5322, + "step": 15835 + }, + { + "epoch": 0.7011500614325404, + "grad_norm": 1.3277474641799927, + "learning_rate": 4.966577408065257e-06, + "loss": 0.484, + "step": 15836 + }, + { + "epoch": 0.7011943371373544, + "grad_norm": 1.6139353513717651, + "learning_rate": 4.965242063918523e-06, + "loss": 0.5979, + "step": 15837 + }, + { + "epoch": 0.7012386128421682, + "grad_norm": 1.4392169713974, + "learning_rate": 4.963906840022883e-06, + "loss": 0.486, + "step": 15838 + }, + { + "epoch": 0.7012828885469821, + "grad_norm": 1.438751220703125, + "learning_rate": 4.962571736410224e-06, + "loss": 0.554, + "step": 15839 + }, + { + "epoch": 0.701327164251796, + "grad_norm": 1.7187496423721313, + "learning_rate": 4.961236753112439e-06, + "loss": 0.5569, + "step": 15840 + }, + { + "epoch": 0.7013714399566098, + "grad_norm": 1.229101300239563, + "learning_rate": 4.959901890161414e-06, + "loss": 0.4533, + "step": 15841 + }, + { + "epoch": 0.7014157156614237, + "grad_norm": 1.5090885162353516, + "learning_rate": 4.958567147589023e-06, + "loss": 0.5289, + "step": 15842 + }, + { + "epoch": 0.7014599913662376, + "grad_norm": 1.3152265548706055, + "learning_rate": 4.957232525427156e-06, + "loss": 0.497, + "step": 15843 + }, + { + "epoch": 0.7015042670710514, + "grad_norm": 2.014512300491333, + "learning_rate": 4.955898023707684e-06, + "loss": 0.6127, + "step": 15844 + }, + { + "epoch": 0.7015485427758653, + "grad_norm": 1.483862042427063, + "learning_rate": 4.95456364246248e-06, + "loss": 0.5258, + "step": 15845 + }, + { + "epoch": 0.7015928184806792, + "grad_norm": 1.5796269178390503, + "learning_rate": 4.9532293817234144e-06, + "loss": 0.5072, + "step": 15846 + }, + { + "epoch": 0.701637094185493, + "grad_norm": 1.4253658056259155, + "learning_rate": 4.9518952415223576e-06, + "loss": 0.5115, + "step": 15847 + }, + { + "epoch": 0.7016813698903069, + "grad_norm": 1.5485152006149292, + "learning_rate": 4.950561221891171e-06, + "loss": 0.5784, + "step": 15848 + }, + { + "epoch": 0.7017256455951209, + "grad_norm": 1.399440884590149, + "learning_rate": 4.949227322861716e-06, + "loss": 0.5011, + "step": 15849 + }, + { + "epoch": 0.7017699212999347, + "grad_norm": 1.664344310760498, + "learning_rate": 4.947893544465857e-06, + "loss": 0.5356, + "step": 15850 + }, + { + "epoch": 0.7018141970047486, + "grad_norm": 2.017242193222046, + "learning_rate": 4.946559886735447e-06, + "loss": 0.6371, + "step": 15851 + }, + { + "epoch": 0.7018584727095625, + "grad_norm": 1.6018918752670288, + "learning_rate": 4.9452263497023354e-06, + "loss": 0.5423, + "step": 15852 + }, + { + "epoch": 0.7019027484143763, + "grad_norm": 1.5410966873168945, + "learning_rate": 4.943892933398381e-06, + "loss": 0.5671, + "step": 15853 + }, + { + "epoch": 0.7019470241191902, + "grad_norm": 1.5383903980255127, + "learning_rate": 4.942559637855431e-06, + "loss": 0.5059, + "step": 15854 + }, + { + "epoch": 0.7019912998240041, + "grad_norm": 1.3176348209381104, + "learning_rate": 4.941226463105322e-06, + "loss": 0.5391, + "step": 15855 + }, + { + "epoch": 0.7020355755288179, + "grad_norm": 1.3785427808761597, + "learning_rate": 4.9398934091798954e-06, + "loss": 0.567, + "step": 15856 + }, + { + "epoch": 0.7020798512336318, + "grad_norm": 1.5641008615493774, + "learning_rate": 4.938560476110999e-06, + "loss": 0.5549, + "step": 15857 + }, + { + "epoch": 0.7021241269384457, + "grad_norm": 1.494704008102417, + "learning_rate": 4.937227663930465e-06, + "loss": 0.5753, + "step": 15858 + }, + { + "epoch": 0.7021684026432595, + "grad_norm": 1.3631869554519653, + "learning_rate": 4.935894972670122e-06, + "loss": 0.5568, + "step": 15859 + }, + { + "epoch": 0.7022126783480734, + "grad_norm": 1.2991483211517334, + "learning_rate": 4.9345624023618075e-06, + "loss": 0.5549, + "step": 15860 + }, + { + "epoch": 0.7022569540528873, + "grad_norm": 1.7282694578170776, + "learning_rate": 4.933229953037346e-06, + "loss": 0.6085, + "step": 15861 + }, + { + "epoch": 0.7023012297577012, + "grad_norm": 1.3870214223861694, + "learning_rate": 4.931897624728562e-06, + "loss": 0.5288, + "step": 15862 + }, + { + "epoch": 0.7023455054625151, + "grad_norm": 1.3980871438980103, + "learning_rate": 4.930565417467276e-06, + "loss": 0.5187, + "step": 15863 + }, + { + "epoch": 0.702389781167329, + "grad_norm": 1.6007455587387085, + "learning_rate": 4.929233331285307e-06, + "loss": 0.62, + "step": 15864 + }, + { + "epoch": 0.7024340568721429, + "grad_norm": 1.4002939462661743, + "learning_rate": 4.927901366214471e-06, + "loss": 0.5063, + "step": 15865 + }, + { + "epoch": 0.7024783325769567, + "grad_norm": 1.6388803720474243, + "learning_rate": 4.926569522286578e-06, + "loss": 0.5446, + "step": 15866 + }, + { + "epoch": 0.7025226082817706, + "grad_norm": 1.4275692701339722, + "learning_rate": 4.925237799533445e-06, + "loss": 0.4993, + "step": 15867 + }, + { + "epoch": 0.7025668839865845, + "grad_norm": 1.436074137687683, + "learning_rate": 4.923906197986874e-06, + "loss": 0.5151, + "step": 15868 + }, + { + "epoch": 0.7026111596913983, + "grad_norm": 1.4419386386871338, + "learning_rate": 4.922574717678668e-06, + "loss": 0.5541, + "step": 15869 + }, + { + "epoch": 0.7026554353962122, + "grad_norm": 1.4139539003372192, + "learning_rate": 4.921243358640634e-06, + "loss": 0.5409, + "step": 15870 + }, + { + "epoch": 0.7026997111010261, + "grad_norm": 1.3208026885986328, + "learning_rate": 4.919912120904566e-06, + "loss": 0.479, + "step": 15871 + }, + { + "epoch": 0.7027439868058399, + "grad_norm": 1.8442600965499878, + "learning_rate": 4.918581004502262e-06, + "loss": 0.5594, + "step": 15872 + }, + { + "epoch": 0.7027882625106538, + "grad_norm": 1.4718093872070312, + "learning_rate": 4.917250009465513e-06, + "loss": 0.5594, + "step": 15873 + }, + { + "epoch": 0.7028325382154678, + "grad_norm": 1.368801236152649, + "learning_rate": 4.915919135826108e-06, + "loss": 0.5171, + "step": 15874 + }, + { + "epoch": 0.7028768139202816, + "grad_norm": 1.790523648262024, + "learning_rate": 4.9145883836158355e-06, + "loss": 0.5456, + "step": 15875 + }, + { + "epoch": 0.7029210896250955, + "grad_norm": 1.596336841583252, + "learning_rate": 4.913257752866475e-06, + "loss": 0.5727, + "step": 15876 + }, + { + "epoch": 0.7029653653299094, + "grad_norm": 1.2601364850997925, + "learning_rate": 4.911927243609816e-06, + "loss": 0.5087, + "step": 15877 + }, + { + "epoch": 0.7030096410347232, + "grad_norm": 1.4421308040618896, + "learning_rate": 4.91059685587763e-06, + "loss": 0.4694, + "step": 15878 + }, + { + "epoch": 0.7030539167395371, + "grad_norm": 1.544121503829956, + "learning_rate": 4.909266589701691e-06, + "loss": 0.5587, + "step": 15879 + }, + { + "epoch": 0.703098192444351, + "grad_norm": 1.6023361682891846, + "learning_rate": 4.9079364451137786e-06, + "loss": 0.5401, + "step": 15880 + }, + { + "epoch": 0.7031424681491648, + "grad_norm": 1.4762418270111084, + "learning_rate": 4.906606422145659e-06, + "loss": 0.5419, + "step": 15881 + }, + { + "epoch": 0.7031867438539787, + "grad_norm": 1.6057285070419312, + "learning_rate": 4.905276520829097e-06, + "loss": 0.57, + "step": 15882 + }, + { + "epoch": 0.7032310195587926, + "grad_norm": 1.3019142150878906, + "learning_rate": 4.903946741195857e-06, + "loss": 0.5105, + "step": 15883 + }, + { + "epoch": 0.7032752952636064, + "grad_norm": 1.8767696619033813, + "learning_rate": 4.902617083277699e-06, + "loss": 0.5985, + "step": 15884 + }, + { + "epoch": 0.7033195709684203, + "grad_norm": 1.5669760704040527, + "learning_rate": 4.901287547106383e-06, + "loss": 0.5711, + "step": 15885 + }, + { + "epoch": 0.7033638466732343, + "grad_norm": 1.4658880233764648, + "learning_rate": 4.899958132713657e-06, + "loss": 0.5026, + "step": 15886 + }, + { + "epoch": 0.7034081223780481, + "grad_norm": 1.505686640739441, + "learning_rate": 4.898628840131283e-06, + "loss": 0.5694, + "step": 15887 + }, + { + "epoch": 0.703452398082862, + "grad_norm": 1.6428731679916382, + "learning_rate": 4.897299669391006e-06, + "loss": 0.6057, + "step": 15888 + }, + { + "epoch": 0.7034966737876759, + "grad_norm": 1.5623756647109985, + "learning_rate": 4.8959706205245675e-06, + "loss": 0.521, + "step": 15889 + }, + { + "epoch": 0.7035409494924897, + "grad_norm": 1.6505897045135498, + "learning_rate": 4.894641693563717e-06, + "loss": 0.572, + "step": 15890 + }, + { + "epoch": 0.7035852251973036, + "grad_norm": 1.8713997602462769, + "learning_rate": 4.893312888540195e-06, + "loss": 0.6303, + "step": 15891 + }, + { + "epoch": 0.7036295009021175, + "grad_norm": 1.4941858053207397, + "learning_rate": 4.891984205485736e-06, + "loss": 0.511, + "step": 15892 + }, + { + "epoch": 0.7036737766069313, + "grad_norm": 1.8203768730163574, + "learning_rate": 4.890655644432075e-06, + "loss": 0.6145, + "step": 15893 + }, + { + "epoch": 0.7037180523117452, + "grad_norm": 1.472123146057129, + "learning_rate": 4.889327205410943e-06, + "loss": 0.5626, + "step": 15894 + }, + { + "epoch": 0.7037623280165591, + "grad_norm": 1.1966761350631714, + "learning_rate": 4.8879988884540705e-06, + "loss": 0.4809, + "step": 15895 + }, + { + "epoch": 0.7038066037213729, + "grad_norm": 1.4631291627883911, + "learning_rate": 4.8866706935931785e-06, + "loss": 0.487, + "step": 15896 + }, + { + "epoch": 0.7038508794261868, + "grad_norm": 1.478615641593933, + "learning_rate": 4.8853426208599964e-06, + "loss": 0.5487, + "step": 15897 + }, + { + "epoch": 0.7038951551310008, + "grad_norm": 1.5704760551452637, + "learning_rate": 4.884014670286242e-06, + "loss": 0.5663, + "step": 15898 + }, + { + "epoch": 0.7039394308358147, + "grad_norm": 1.5903363227844238, + "learning_rate": 4.882686841903627e-06, + "loss": 0.5035, + "step": 15899 + }, + { + "epoch": 0.7039837065406285, + "grad_norm": 1.6019246578216553, + "learning_rate": 4.8813591357438746e-06, + "loss": 0.5131, + "step": 15900 + }, + { + "epoch": 0.7040279822454424, + "grad_norm": 1.5058836936950684, + "learning_rate": 4.880031551838695e-06, + "loss": 0.5703, + "step": 15901 + }, + { + "epoch": 0.7040722579502563, + "grad_norm": 1.4637047052383423, + "learning_rate": 4.878704090219787e-06, + "loss": 0.4989, + "step": 15902 + }, + { + "epoch": 0.7041165336550701, + "grad_norm": 1.4967114925384521, + "learning_rate": 4.877376750918864e-06, + "loss": 0.5263, + "step": 15903 + }, + { + "epoch": 0.704160809359884, + "grad_norm": 1.4100100994110107, + "learning_rate": 4.876049533967626e-06, + "loss": 0.5407, + "step": 15904 + }, + { + "epoch": 0.7042050850646979, + "grad_norm": 1.648513913154602, + "learning_rate": 4.874722439397773e-06, + "loss": 0.5672, + "step": 15905 + }, + { + "epoch": 0.7042493607695117, + "grad_norm": 1.1878377199172974, + "learning_rate": 4.873395467240999e-06, + "loss": 0.4502, + "step": 15906 + }, + { + "epoch": 0.7042936364743256, + "grad_norm": 1.6320661306381226, + "learning_rate": 4.872068617529002e-06, + "loss": 0.5774, + "step": 15907 + }, + { + "epoch": 0.7043379121791395, + "grad_norm": 1.4903160333633423, + "learning_rate": 4.870741890293472e-06, + "loss": 0.5218, + "step": 15908 + }, + { + "epoch": 0.7043821878839533, + "grad_norm": 1.465631365776062, + "learning_rate": 4.869415285566091e-06, + "loss": 0.5889, + "step": 15909 + }, + { + "epoch": 0.7044264635887673, + "grad_norm": 1.2554367780685425, + "learning_rate": 4.868088803378557e-06, + "loss": 0.5005, + "step": 15910 + }, + { + "epoch": 0.7044707392935812, + "grad_norm": 1.4261573553085327, + "learning_rate": 4.866762443762538e-06, + "loss": 0.5148, + "step": 15911 + }, + { + "epoch": 0.704515014998395, + "grad_norm": 1.7213677167892456, + "learning_rate": 4.865436206749716e-06, + "loss": 0.6023, + "step": 15912 + }, + { + "epoch": 0.7045592907032089, + "grad_norm": 1.9157911539077759, + "learning_rate": 4.864110092371772e-06, + "loss": 0.5189, + "step": 15913 + }, + { + "epoch": 0.7046035664080228, + "grad_norm": 1.4750683307647705, + "learning_rate": 4.862784100660378e-06, + "loss": 0.5126, + "step": 15914 + }, + { + "epoch": 0.7046478421128366, + "grad_norm": 1.5071132183074951, + "learning_rate": 4.861458231647202e-06, + "loss": 0.5159, + "step": 15915 + }, + { + "epoch": 0.7046921178176505, + "grad_norm": 1.215834379196167, + "learning_rate": 4.8601324853639085e-06, + "loss": 0.5008, + "step": 15916 + }, + { + "epoch": 0.7047363935224644, + "grad_norm": 1.5862658023834229, + "learning_rate": 4.85880686184217e-06, + "loss": 0.5272, + "step": 15917 + }, + { + "epoch": 0.7047806692272782, + "grad_norm": 1.6399202346801758, + "learning_rate": 4.857481361113643e-06, + "loss": 0.595, + "step": 15918 + }, + { + "epoch": 0.7048249449320921, + "grad_norm": 1.2704038619995117, + "learning_rate": 4.8561559832099865e-06, + "loss": 0.5236, + "step": 15919 + }, + { + "epoch": 0.704869220636906, + "grad_norm": 1.3988142013549805, + "learning_rate": 4.854830728162856e-06, + "loss": 0.5496, + "step": 15920 + }, + { + "epoch": 0.7049134963417198, + "grad_norm": 1.3177756071090698, + "learning_rate": 4.8535055960039046e-06, + "loss": 0.4756, + "step": 15921 + }, + { + "epoch": 0.7049577720465338, + "grad_norm": 1.128255844116211, + "learning_rate": 4.852180586764777e-06, + "loss": 0.4563, + "step": 15922 + }, + { + "epoch": 0.7050020477513477, + "grad_norm": 1.905324101448059, + "learning_rate": 4.85085570047713e-06, + "loss": 0.5861, + "step": 15923 + }, + { + "epoch": 0.7050463234561615, + "grad_norm": 1.4715684652328491, + "learning_rate": 4.8495309371726e-06, + "loss": 0.5, + "step": 15924 + }, + { + "epoch": 0.7050905991609754, + "grad_norm": 1.6106165647506714, + "learning_rate": 4.848206296882831e-06, + "loss": 0.5185, + "step": 15925 + }, + { + "epoch": 0.7051348748657893, + "grad_norm": 1.6927403211593628, + "learning_rate": 4.846881779639455e-06, + "loss": 0.5105, + "step": 15926 + }, + { + "epoch": 0.7051791505706031, + "grad_norm": 1.4595636129379272, + "learning_rate": 4.845557385474117e-06, + "loss": 0.5663, + "step": 15927 + }, + { + "epoch": 0.705223426275417, + "grad_norm": 1.6531428098678589, + "learning_rate": 4.844233114418444e-06, + "loss": 0.6043, + "step": 15928 + }, + { + "epoch": 0.7052677019802309, + "grad_norm": 1.319790244102478, + "learning_rate": 4.842908966504064e-06, + "loss": 0.5173, + "step": 15929 + }, + { + "epoch": 0.7053119776850447, + "grad_norm": 1.7871415615081787, + "learning_rate": 4.841584941762605e-06, + "loss": 0.6492, + "step": 15930 + }, + { + "epoch": 0.7053562533898586, + "grad_norm": 1.6066217422485352, + "learning_rate": 4.840261040225689e-06, + "loss": 0.5349, + "step": 15931 + }, + { + "epoch": 0.7054005290946725, + "grad_norm": 1.7316724061965942, + "learning_rate": 4.838937261924933e-06, + "loss": 0.6258, + "step": 15932 + }, + { + "epoch": 0.7054448047994863, + "grad_norm": 1.4834619760513306, + "learning_rate": 4.8376136068919635e-06, + "loss": 0.5467, + "step": 15933 + }, + { + "epoch": 0.7054890805043003, + "grad_norm": 1.4507540464401245, + "learning_rate": 4.836290075158388e-06, + "loss": 0.5557, + "step": 15934 + }, + { + "epoch": 0.7055333562091142, + "grad_norm": 1.644511342048645, + "learning_rate": 4.8349666667558205e-06, + "loss": 0.5003, + "step": 15935 + }, + { + "epoch": 0.7055776319139281, + "grad_norm": 1.312915563583374, + "learning_rate": 4.8336433817158634e-06, + "loss": 0.485, + "step": 15936 + }, + { + "epoch": 0.7056219076187419, + "grad_norm": 1.4385883808135986, + "learning_rate": 4.832320220070132e-06, + "loss": 0.531, + "step": 15937 + }, + { + "epoch": 0.7056661833235558, + "grad_norm": 1.5639588832855225, + "learning_rate": 4.830997181850225e-06, + "loss": 0.5629, + "step": 15938 + }, + { + "epoch": 0.7057104590283697, + "grad_norm": 1.7952789068222046, + "learning_rate": 4.829674267087742e-06, + "loss": 0.536, + "step": 15939 + }, + { + "epoch": 0.7057547347331835, + "grad_norm": 1.5052380561828613, + "learning_rate": 4.8283514758142795e-06, + "loss": 0.4779, + "step": 15940 + }, + { + "epoch": 0.7057990104379974, + "grad_norm": 1.4877253770828247, + "learning_rate": 4.827028808061429e-06, + "loss": 0.491, + "step": 15941 + }, + { + "epoch": 0.7058432861428113, + "grad_norm": 1.5523715019226074, + "learning_rate": 4.825706263860781e-06, + "loss": 0.5357, + "step": 15942 + }, + { + "epoch": 0.7058875618476251, + "grad_norm": 1.4239901304244995, + "learning_rate": 4.824383843243929e-06, + "loss": 0.5653, + "step": 15943 + }, + { + "epoch": 0.705931837552439, + "grad_norm": 1.6776721477508545, + "learning_rate": 4.823061546242456e-06, + "loss": 0.5843, + "step": 15944 + }, + { + "epoch": 0.705976113257253, + "grad_norm": 1.7482558488845825, + "learning_rate": 4.821739372887942e-06, + "loss": 0.5551, + "step": 15945 + }, + { + "epoch": 0.7060203889620668, + "grad_norm": 1.6283018589019775, + "learning_rate": 4.820417323211963e-06, + "loss": 0.5761, + "step": 15946 + }, + { + "epoch": 0.7060646646668807, + "grad_norm": 1.3287949562072754, + "learning_rate": 4.819095397246103e-06, + "loss": 0.5375, + "step": 15947 + }, + { + "epoch": 0.7061089403716946, + "grad_norm": 1.4224954843521118, + "learning_rate": 4.81777359502193e-06, + "loss": 0.509, + "step": 15948 + }, + { + "epoch": 0.7061532160765084, + "grad_norm": 1.3182624578475952, + "learning_rate": 4.816451916571016e-06, + "loss": 0.4958, + "step": 15949 + }, + { + "epoch": 0.7061974917813223, + "grad_norm": 1.6303313970565796, + "learning_rate": 4.815130361924928e-06, + "loss": 0.5418, + "step": 15950 + }, + { + "epoch": 0.7062417674861362, + "grad_norm": 1.4945943355560303, + "learning_rate": 4.813808931115228e-06, + "loss": 0.4927, + "step": 15951 + }, + { + "epoch": 0.70628604319095, + "grad_norm": 1.4053928852081299, + "learning_rate": 4.812487624173477e-06, + "loss": 0.5118, + "step": 15952 + }, + { + "epoch": 0.7063303188957639, + "grad_norm": 1.570692539215088, + "learning_rate": 4.811166441131238e-06, + "loss": 0.5659, + "step": 15953 + }, + { + "epoch": 0.7063745946005778, + "grad_norm": 1.9347809553146362, + "learning_rate": 4.8098453820200634e-06, + "loss": 0.6599, + "step": 15954 + }, + { + "epoch": 0.7064188703053916, + "grad_norm": 1.8980191946029663, + "learning_rate": 4.808524446871505e-06, + "loss": 0.658, + "step": 15955 + }, + { + "epoch": 0.7064631460102055, + "grad_norm": 1.5705946683883667, + "learning_rate": 4.807203635717112e-06, + "loss": 0.583, + "step": 15956 + }, + { + "epoch": 0.7065074217150195, + "grad_norm": 1.3099801540374756, + "learning_rate": 4.805882948588438e-06, + "loss": 0.5301, + "step": 15957 + }, + { + "epoch": 0.7065516974198333, + "grad_norm": 1.4736177921295166, + "learning_rate": 4.804562385517016e-06, + "loss": 0.4875, + "step": 15958 + }, + { + "epoch": 0.7065959731246472, + "grad_norm": 1.9552888870239258, + "learning_rate": 4.803241946534388e-06, + "loss": 0.5758, + "step": 15959 + }, + { + "epoch": 0.7066402488294611, + "grad_norm": 2.036933183670044, + "learning_rate": 4.8019216316720986e-06, + "loss": 0.5647, + "step": 15960 + }, + { + "epoch": 0.7066845245342749, + "grad_norm": 1.4915671348571777, + "learning_rate": 4.800601440961678e-06, + "loss": 0.6071, + "step": 15961 + }, + { + "epoch": 0.7067288002390888, + "grad_norm": 1.9425313472747803, + "learning_rate": 4.79928137443466e-06, + "loss": 0.5621, + "step": 15962 + }, + { + "epoch": 0.7067730759439027, + "grad_norm": 1.7923332452774048, + "learning_rate": 4.797961432122568e-06, + "loss": 0.5731, + "step": 15963 + }, + { + "epoch": 0.7068173516487165, + "grad_norm": 1.5305500030517578, + "learning_rate": 4.796641614056934e-06, + "loss": 0.5394, + "step": 15964 + }, + { + "epoch": 0.7068616273535304, + "grad_norm": 1.7643516063690186, + "learning_rate": 4.795321920269279e-06, + "loss": 0.5669, + "step": 15965 + }, + { + "epoch": 0.7069059030583443, + "grad_norm": 1.3751713037490845, + "learning_rate": 4.794002350791122e-06, + "loss": 0.5009, + "step": 15966 + }, + { + "epoch": 0.7069501787631581, + "grad_norm": 1.561269998550415, + "learning_rate": 4.792682905653979e-06, + "loss": 0.5294, + "step": 15967 + }, + { + "epoch": 0.706994454467972, + "grad_norm": 1.6242766380310059, + "learning_rate": 4.791363584889367e-06, + "loss": 0.5649, + "step": 15968 + }, + { + "epoch": 0.707038730172786, + "grad_norm": 1.3498876094818115, + "learning_rate": 4.7900443885287895e-06, + "loss": 0.4766, + "step": 15969 + }, + { + "epoch": 0.7070830058775999, + "grad_norm": 1.5985106229782104, + "learning_rate": 4.788725316603764e-06, + "loss": 0.5506, + "step": 15970 + }, + { + "epoch": 0.7071272815824137, + "grad_norm": 1.5166079998016357, + "learning_rate": 4.78740636914579e-06, + "loss": 0.5647, + "step": 15971 + }, + { + "epoch": 0.7071715572872276, + "grad_norm": 1.6644902229309082, + "learning_rate": 4.786087546186372e-06, + "loss": 0.5833, + "step": 15972 + }, + { + "epoch": 0.7072158329920415, + "grad_norm": 1.3354190587997437, + "learning_rate": 4.784768847757003e-06, + "loss": 0.4883, + "step": 15973 + }, + { + "epoch": 0.7072601086968553, + "grad_norm": 1.6606900691986084, + "learning_rate": 4.783450273889188e-06, + "loss": 0.5509, + "step": 15974 + }, + { + "epoch": 0.7073043844016692, + "grad_norm": 1.4430346488952637, + "learning_rate": 4.782131824614415e-06, + "loss": 0.5019, + "step": 15975 + }, + { + "epoch": 0.7073486601064831, + "grad_norm": 1.2805774211883545, + "learning_rate": 4.780813499964175e-06, + "loss": 0.5141, + "step": 15976 + }, + { + "epoch": 0.7073929358112969, + "grad_norm": 1.6168063879013062, + "learning_rate": 4.779495299969956e-06, + "loss": 0.604, + "step": 15977 + }, + { + "epoch": 0.7074372115161108, + "grad_norm": 1.6311564445495605, + "learning_rate": 4.778177224663241e-06, + "loss": 0.5582, + "step": 15978 + }, + { + "epoch": 0.7074814872209247, + "grad_norm": 1.3479818105697632, + "learning_rate": 4.776859274075506e-06, + "loss": 0.5197, + "step": 15979 + }, + { + "epoch": 0.7075257629257385, + "grad_norm": 1.8490451574325562, + "learning_rate": 4.775541448238239e-06, + "loss": 0.587, + "step": 15980 + }, + { + "epoch": 0.7075700386305525, + "grad_norm": 1.437228798866272, + "learning_rate": 4.77422374718291e-06, + "loss": 0.544, + "step": 15981 + }, + { + "epoch": 0.7076143143353664, + "grad_norm": 1.8628565073013306, + "learning_rate": 4.772906170940993e-06, + "loss": 0.6547, + "step": 15982 + }, + { + "epoch": 0.7076585900401802, + "grad_norm": 1.4000383615493774, + "learning_rate": 4.7715887195439515e-06, + "loss": 0.5425, + "step": 15983 + }, + { + "epoch": 0.7077028657449941, + "grad_norm": 1.574824571609497, + "learning_rate": 4.77027139302326e-06, + "loss": 0.5667, + "step": 15984 + }, + { + "epoch": 0.707747141449808, + "grad_norm": 1.3640737533569336, + "learning_rate": 4.768954191410379e-06, + "loss": 0.5286, + "step": 15985 + }, + { + "epoch": 0.7077914171546218, + "grad_norm": 1.4440340995788574, + "learning_rate": 4.767637114736766e-06, + "loss": 0.4697, + "step": 15986 + }, + { + "epoch": 0.7078356928594357, + "grad_norm": 1.5710816383361816, + "learning_rate": 4.766320163033882e-06, + "loss": 0.5354, + "step": 15987 + }, + { + "epoch": 0.7078799685642496, + "grad_norm": 1.3350062370300293, + "learning_rate": 4.765003336333178e-06, + "loss": 0.5106, + "step": 15988 + }, + { + "epoch": 0.7079242442690634, + "grad_norm": 1.5399117469787598, + "learning_rate": 4.763686634666104e-06, + "loss": 0.5247, + "step": 15989 + }, + { + "epoch": 0.7079685199738773, + "grad_norm": 1.3825711011886597, + "learning_rate": 4.762370058064115e-06, + "loss": 0.5191, + "step": 15990 + }, + { + "epoch": 0.7080127956786912, + "grad_norm": 1.4525526762008667, + "learning_rate": 4.7610536065586535e-06, + "loss": 0.5278, + "step": 15991 + }, + { + "epoch": 0.708057071383505, + "grad_norm": 1.397711157798767, + "learning_rate": 4.759737280181159e-06, + "loss": 0.5427, + "step": 15992 + }, + { + "epoch": 0.708101347088319, + "grad_norm": 1.5690319538116455, + "learning_rate": 4.758421078963069e-06, + "loss": 0.578, + "step": 15993 + }, + { + "epoch": 0.7081456227931329, + "grad_norm": 1.318231225013733, + "learning_rate": 4.757105002935829e-06, + "loss": 0.4852, + "step": 15994 + }, + { + "epoch": 0.7081898984979467, + "grad_norm": 1.4485944509506226, + "learning_rate": 4.755789052130866e-06, + "loss": 0.4997, + "step": 15995 + }, + { + "epoch": 0.7082341742027606, + "grad_norm": 1.556602954864502, + "learning_rate": 4.754473226579612e-06, + "loss": 0.5442, + "step": 15996 + }, + { + "epoch": 0.7082784499075745, + "grad_norm": 1.6162844896316528, + "learning_rate": 4.753157526313494e-06, + "loss": 0.5457, + "step": 15997 + }, + { + "epoch": 0.7083227256123883, + "grad_norm": 1.368924856185913, + "learning_rate": 4.7518419513639355e-06, + "loss": 0.514, + "step": 15998 + }, + { + "epoch": 0.7083670013172022, + "grad_norm": 1.427196741104126, + "learning_rate": 4.750526501762356e-06, + "loss": 0.4754, + "step": 15999 + }, + { + "epoch": 0.7084112770220161, + "grad_norm": 1.5682581663131714, + "learning_rate": 4.7492111775401796e-06, + "loss": 0.5249, + "step": 16000 + }, + { + "epoch": 0.7084555527268299, + "grad_norm": 2.306577205657959, + "learning_rate": 4.747895978728819e-06, + "loss": 0.6339, + "step": 16001 + }, + { + "epoch": 0.7084998284316438, + "grad_norm": 1.2563891410827637, + "learning_rate": 4.746580905359687e-06, + "loss": 0.492, + "step": 16002 + }, + { + "epoch": 0.7085441041364577, + "grad_norm": 1.4395265579223633, + "learning_rate": 4.745265957464189e-06, + "loss": 0.4966, + "step": 16003 + }, + { + "epoch": 0.7085883798412715, + "grad_norm": 1.3542197942733765, + "learning_rate": 4.743951135073738e-06, + "loss": 0.5363, + "step": 16004 + }, + { + "epoch": 0.7086326555460855, + "grad_norm": 1.6266268491744995, + "learning_rate": 4.7426364382197385e-06, + "loss": 0.5525, + "step": 16005 + }, + { + "epoch": 0.7086769312508994, + "grad_norm": 1.6622190475463867, + "learning_rate": 4.74132186693358e-06, + "loss": 0.5341, + "step": 16006 + }, + { + "epoch": 0.7087212069557133, + "grad_norm": 1.316123366355896, + "learning_rate": 4.7400074212466705e-06, + "loss": 0.5147, + "step": 16007 + }, + { + "epoch": 0.7087654826605271, + "grad_norm": 2.2575008869171143, + "learning_rate": 4.738693101190401e-06, + "loss": 0.6612, + "step": 16008 + }, + { + "epoch": 0.708809758365341, + "grad_norm": 1.4151259660720825, + "learning_rate": 4.7373789067961585e-06, + "loss": 0.498, + "step": 16009 + }, + { + "epoch": 0.7088540340701549, + "grad_norm": 1.7348740100860596, + "learning_rate": 4.73606483809534e-06, + "loss": 0.6182, + "step": 16010 + }, + { + "epoch": 0.7088983097749687, + "grad_norm": 1.6436846256256104, + "learning_rate": 4.734750895119327e-06, + "loss": 0.6007, + "step": 16011 + }, + { + "epoch": 0.7089425854797826, + "grad_norm": 1.6042150259017944, + "learning_rate": 4.733437077899501e-06, + "loss": 0.5809, + "step": 16012 + }, + { + "epoch": 0.7089868611845965, + "grad_norm": 1.2759222984313965, + "learning_rate": 4.732123386467242e-06, + "loss": 0.464, + "step": 16013 + }, + { + "epoch": 0.7090311368894103, + "grad_norm": 1.5648828744888306, + "learning_rate": 4.730809820853928e-06, + "loss": 0.4973, + "step": 16014 + }, + { + "epoch": 0.7090754125942242, + "grad_norm": 1.4328644275665283, + "learning_rate": 4.72949638109093e-06, + "loss": 0.5506, + "step": 16015 + }, + { + "epoch": 0.7091196882990382, + "grad_norm": 1.6139910221099854, + "learning_rate": 4.7281830672096165e-06, + "loss": 0.6327, + "step": 16016 + }, + { + "epoch": 0.709163964003852, + "grad_norm": 1.4764220714569092, + "learning_rate": 4.726869879241361e-06, + "loss": 0.5785, + "step": 16017 + }, + { + "epoch": 0.7092082397086659, + "grad_norm": 1.6346147060394287, + "learning_rate": 4.725556817217526e-06, + "loss": 0.5521, + "step": 16018 + }, + { + "epoch": 0.7092525154134798, + "grad_norm": 1.404022216796875, + "learning_rate": 4.7242438811694666e-06, + "loss": 0.5544, + "step": 16019 + }, + { + "epoch": 0.7092967911182936, + "grad_norm": 1.5042402744293213, + "learning_rate": 4.7229310711285515e-06, + "loss": 0.5611, + "step": 16020 + }, + { + "epoch": 0.7093410668231075, + "grad_norm": 1.4955933094024658, + "learning_rate": 4.72161838712613e-06, + "loss": 0.5472, + "step": 16021 + }, + { + "epoch": 0.7093853425279214, + "grad_norm": 1.475273609161377, + "learning_rate": 4.7203058291935566e-06, + "loss": 0.5395, + "step": 16022 + }, + { + "epoch": 0.7094296182327352, + "grad_norm": 1.494710922241211, + "learning_rate": 4.718993397362179e-06, + "loss": 0.5356, + "step": 16023 + }, + { + "epoch": 0.7094738939375491, + "grad_norm": 1.4214550256729126, + "learning_rate": 4.717681091663344e-06, + "loss": 0.4807, + "step": 16024 + }, + { + "epoch": 0.709518169642363, + "grad_norm": 1.6859171390533447, + "learning_rate": 4.716368912128395e-06, + "loss": 0.5709, + "step": 16025 + }, + { + "epoch": 0.7095624453471768, + "grad_norm": 1.3532177209854126, + "learning_rate": 4.715056858788669e-06, + "loss": 0.5425, + "step": 16026 + }, + { + "epoch": 0.7096067210519907, + "grad_norm": 1.6652714014053345, + "learning_rate": 4.71374493167551e-06, + "loss": 0.5584, + "step": 16027 + }, + { + "epoch": 0.7096509967568047, + "grad_norm": 1.433272123336792, + "learning_rate": 4.7124331308202494e-06, + "loss": 0.491, + "step": 16028 + }, + { + "epoch": 0.7096952724616185, + "grad_norm": 1.511300802230835, + "learning_rate": 4.711121456254215e-06, + "loss": 0.5466, + "step": 16029 + }, + { + "epoch": 0.7097395481664324, + "grad_norm": 1.7564172744750977, + "learning_rate": 4.7098099080087415e-06, + "loss": 0.5315, + "step": 16030 + }, + { + "epoch": 0.7097838238712463, + "grad_norm": 1.8355236053466797, + "learning_rate": 4.7084984861151504e-06, + "loss": 0.5379, + "step": 16031 + }, + { + "epoch": 0.7098280995760601, + "grad_norm": 1.4072794914245605, + "learning_rate": 4.707187190604765e-06, + "loss": 0.4797, + "step": 16032 + }, + { + "epoch": 0.709872375280874, + "grad_norm": 1.4401310682296753, + "learning_rate": 4.705876021508904e-06, + "loss": 0.5015, + "step": 16033 + }, + { + "epoch": 0.7099166509856879, + "grad_norm": 1.4958523511886597, + "learning_rate": 4.704564978858884e-06, + "loss": 0.4902, + "step": 16034 + }, + { + "epoch": 0.7099609266905017, + "grad_norm": 1.4545220136642456, + "learning_rate": 4.703254062686017e-06, + "loss": 0.5222, + "step": 16035 + }, + { + "epoch": 0.7100052023953156, + "grad_norm": 1.5101152658462524, + "learning_rate": 4.701943273021611e-06, + "loss": 0.5193, + "step": 16036 + }, + { + "epoch": 0.7100494781001295, + "grad_norm": 1.8211816549301147, + "learning_rate": 4.700632609896979e-06, + "loss": 0.5427, + "step": 16037 + }, + { + "epoch": 0.7100937538049433, + "grad_norm": 1.5628291368484497, + "learning_rate": 4.699322073343423e-06, + "loss": 0.607, + "step": 16038 + }, + { + "epoch": 0.7101380295097572, + "grad_norm": 1.514390468597412, + "learning_rate": 4.698011663392239e-06, + "loss": 0.5431, + "step": 16039 + }, + { + "epoch": 0.7101823052145712, + "grad_norm": 1.5077097415924072, + "learning_rate": 4.696701380074733e-06, + "loss": 0.5609, + "step": 16040 + }, + { + "epoch": 0.7102265809193851, + "grad_norm": 1.3537843227386475, + "learning_rate": 4.695391223422197e-06, + "loss": 0.5228, + "step": 16041 + }, + { + "epoch": 0.7102708566241989, + "grad_norm": 1.3235453367233276, + "learning_rate": 4.694081193465921e-06, + "loss": 0.4804, + "step": 16042 + }, + { + "epoch": 0.7103151323290128, + "grad_norm": 1.6073044538497925, + "learning_rate": 4.692771290237197e-06, + "loss": 0.5642, + "step": 16043 + }, + { + "epoch": 0.7103594080338267, + "grad_norm": 1.4755253791809082, + "learning_rate": 4.691461513767309e-06, + "loss": 0.511, + "step": 16044 + }, + { + "epoch": 0.7104036837386405, + "grad_norm": 1.484616756439209, + "learning_rate": 4.690151864087541e-06, + "loss": 0.5362, + "step": 16045 + }, + { + "epoch": 0.7104479594434544, + "grad_norm": 1.2858757972717285, + "learning_rate": 4.688842341229167e-06, + "loss": 0.4533, + "step": 16046 + }, + { + "epoch": 0.7104922351482683, + "grad_norm": 1.2745252847671509, + "learning_rate": 4.687532945223474e-06, + "loss": 0.503, + "step": 16047 + }, + { + "epoch": 0.7105365108530821, + "grad_norm": 1.6904939413070679, + "learning_rate": 4.686223676101731e-06, + "loss": 0.5099, + "step": 16048 + }, + { + "epoch": 0.710580786557896, + "grad_norm": 1.5659364461898804, + "learning_rate": 4.684914533895205e-06, + "loss": 0.5514, + "step": 16049 + }, + { + "epoch": 0.7106250622627099, + "grad_norm": 1.286980390548706, + "learning_rate": 4.683605518635172e-06, + "loss": 0.5455, + "step": 16050 + }, + { + "epoch": 0.7106693379675237, + "grad_norm": 1.4419227838516235, + "learning_rate": 4.682296630352892e-06, + "loss": 0.502, + "step": 16051 + }, + { + "epoch": 0.7107136136723377, + "grad_norm": 1.698639154434204, + "learning_rate": 4.680987869079631e-06, + "loss": 0.598, + "step": 16052 + }, + { + "epoch": 0.7107578893771516, + "grad_norm": 1.6537392139434814, + "learning_rate": 4.679679234846636e-06, + "loss": 0.5329, + "step": 16053 + }, + { + "epoch": 0.7108021650819654, + "grad_norm": 1.423783302307129, + "learning_rate": 4.6783707276851745e-06, + "loss": 0.5172, + "step": 16054 + }, + { + "epoch": 0.7108464407867793, + "grad_norm": 1.2688136100769043, + "learning_rate": 4.677062347626494e-06, + "loss": 0.4861, + "step": 16055 + }, + { + "epoch": 0.7108907164915932, + "grad_norm": 1.608416199684143, + "learning_rate": 4.675754094701843e-06, + "loss": 0.5417, + "step": 16056 + }, + { + "epoch": 0.710934992196407, + "grad_norm": 1.4647926092147827, + "learning_rate": 4.674445968942473e-06, + "loss": 0.546, + "step": 16057 + }, + { + "epoch": 0.7109792679012209, + "grad_norm": 1.3995757102966309, + "learning_rate": 4.6731379703796254e-06, + "loss": 0.5104, + "step": 16058 + }, + { + "epoch": 0.7110235436060348, + "grad_norm": 1.3374419212341309, + "learning_rate": 4.671830099044536e-06, + "loss": 0.4693, + "step": 16059 + }, + { + "epoch": 0.7110678193108486, + "grad_norm": 1.8897027969360352, + "learning_rate": 4.670522354968451e-06, + "loss": 0.5256, + "step": 16060 + }, + { + "epoch": 0.7111120950156625, + "grad_norm": 1.5855315923690796, + "learning_rate": 4.6692147381826025e-06, + "loss": 0.607, + "step": 16061 + }, + { + "epoch": 0.7111563707204764, + "grad_norm": 1.6877645254135132, + "learning_rate": 4.667907248718216e-06, + "loss": 0.5563, + "step": 16062 + }, + { + "epoch": 0.7112006464252902, + "grad_norm": 1.6342641115188599, + "learning_rate": 4.666599886606521e-06, + "loss": 0.5806, + "step": 16063 + }, + { + "epoch": 0.7112449221301042, + "grad_norm": 1.3264925479888916, + "learning_rate": 4.665292651878748e-06, + "loss": 0.4651, + "step": 16064 + }, + { + "epoch": 0.7112891978349181, + "grad_norm": 1.4077852964401245, + "learning_rate": 4.663985544566115e-06, + "loss": 0.4998, + "step": 16065 + }, + { + "epoch": 0.7113334735397319, + "grad_norm": 1.3208204507827759, + "learning_rate": 4.662678564699841e-06, + "loss": 0.5094, + "step": 16066 + }, + { + "epoch": 0.7113777492445458, + "grad_norm": 1.5647237300872803, + "learning_rate": 4.6613717123111455e-06, + "loss": 0.5063, + "step": 16067 + }, + { + "epoch": 0.7114220249493597, + "grad_norm": 1.5002435445785522, + "learning_rate": 4.660064987431241e-06, + "loss": 0.5488, + "step": 16068 + }, + { + "epoch": 0.7114663006541735, + "grad_norm": 1.8296911716461182, + "learning_rate": 4.658758390091336e-06, + "loss": 0.6047, + "step": 16069 + }, + { + "epoch": 0.7115105763589874, + "grad_norm": 1.7745475769042969, + "learning_rate": 4.657451920322637e-06, + "loss": 0.6064, + "step": 16070 + }, + { + "epoch": 0.7115548520638013, + "grad_norm": 1.5590163469314575, + "learning_rate": 4.6561455781563506e-06, + "loss": 0.5619, + "step": 16071 + }, + { + "epoch": 0.7115991277686151, + "grad_norm": 1.323655366897583, + "learning_rate": 4.654839363623674e-06, + "loss": 0.5136, + "step": 16072 + }, + { + "epoch": 0.711643403473429, + "grad_norm": 1.4798706769943237, + "learning_rate": 4.653533276755805e-06, + "loss": 0.5233, + "step": 16073 + }, + { + "epoch": 0.7116876791782429, + "grad_norm": 1.3280112743377686, + "learning_rate": 4.652227317583943e-06, + "loss": 0.489, + "step": 16074 + }, + { + "epoch": 0.7117319548830568, + "grad_norm": 1.6092826128005981, + "learning_rate": 4.6509214861392785e-06, + "loss": 0.6208, + "step": 16075 + }, + { + "epoch": 0.7117762305878707, + "grad_norm": 1.522950530052185, + "learning_rate": 4.649615782452994e-06, + "loss": 0.5241, + "step": 16076 + }, + { + "epoch": 0.7118205062926846, + "grad_norm": 1.554634928703308, + "learning_rate": 4.648310206556284e-06, + "loss": 0.5822, + "step": 16077 + }, + { + "epoch": 0.7118647819974985, + "grad_norm": 1.2833973169326782, + "learning_rate": 4.647004758480327e-06, + "loss": 0.4853, + "step": 16078 + }, + { + "epoch": 0.7119090577023123, + "grad_norm": 1.6600793600082397, + "learning_rate": 4.645699438256304e-06, + "loss": 0.524, + "step": 16079 + }, + { + "epoch": 0.7119533334071262, + "grad_norm": 1.8054698705673218, + "learning_rate": 4.644394245915389e-06, + "loss": 0.606, + "step": 16080 + }, + { + "epoch": 0.7119976091119401, + "grad_norm": 1.7388752698898315, + "learning_rate": 4.643089181488757e-06, + "loss": 0.4905, + "step": 16081 + }, + { + "epoch": 0.7120418848167539, + "grad_norm": 1.5761499404907227, + "learning_rate": 4.641784245007578e-06, + "loss": 0.586, + "step": 16082 + }, + { + "epoch": 0.7120861605215678, + "grad_norm": 1.742349624633789, + "learning_rate": 4.640479436503016e-06, + "loss": 0.5828, + "step": 16083 + }, + { + "epoch": 0.7121304362263817, + "grad_norm": 1.6284335851669312, + "learning_rate": 4.6391747560062425e-06, + "loss": 0.556, + "step": 16084 + }, + { + "epoch": 0.7121747119311955, + "grad_norm": 1.6016056537628174, + "learning_rate": 4.6378702035484146e-06, + "loss": 0.5606, + "step": 16085 + }, + { + "epoch": 0.7122189876360094, + "grad_norm": 1.7494349479675293, + "learning_rate": 4.6365657791606875e-06, + "loss": 0.4834, + "step": 16086 + }, + { + "epoch": 0.7122632633408233, + "grad_norm": 2.025944709777832, + "learning_rate": 4.635261482874224e-06, + "loss": 0.5384, + "step": 16087 + }, + { + "epoch": 0.7123075390456372, + "grad_norm": 1.4912244081497192, + "learning_rate": 4.633957314720171e-06, + "loss": 0.537, + "step": 16088 + }, + { + "epoch": 0.7123518147504511, + "grad_norm": 1.2536872625350952, + "learning_rate": 4.632653274729679e-06, + "loss": 0.5095, + "step": 16089 + }, + { + "epoch": 0.712396090455265, + "grad_norm": 1.5161644220352173, + "learning_rate": 4.631349362933894e-06, + "loss": 0.5004, + "step": 16090 + }, + { + "epoch": 0.7124403661600788, + "grad_norm": 1.4505215883255005, + "learning_rate": 4.6300455793639565e-06, + "loss": 0.5398, + "step": 16091 + }, + { + "epoch": 0.7124846418648927, + "grad_norm": 1.6625343561172485, + "learning_rate": 4.628741924051008e-06, + "loss": 0.5564, + "step": 16092 + }, + { + "epoch": 0.7125289175697066, + "grad_norm": 1.464558720588684, + "learning_rate": 4.627438397026183e-06, + "loss": 0.5255, + "step": 16093 + }, + { + "epoch": 0.7125731932745204, + "grad_norm": 1.6686097383499146, + "learning_rate": 4.6261349983206195e-06, + "loss": 0.6095, + "step": 16094 + }, + { + "epoch": 0.7126174689793343, + "grad_norm": 1.4606915712356567, + "learning_rate": 4.624831727965446e-06, + "loss": 0.5274, + "step": 16095 + }, + { + "epoch": 0.7126617446841482, + "grad_norm": 4.892707347869873, + "learning_rate": 4.623528585991788e-06, + "loss": 0.6223, + "step": 16096 + }, + { + "epoch": 0.712706020388962, + "grad_norm": 1.5228286981582642, + "learning_rate": 4.622225572430774e-06, + "loss": 0.579, + "step": 16097 + }, + { + "epoch": 0.7127502960937759, + "grad_norm": 1.7045462131500244, + "learning_rate": 4.620922687313525e-06, + "loss": 0.5226, + "step": 16098 + }, + { + "epoch": 0.7127945717985898, + "grad_norm": 1.8546894788742065, + "learning_rate": 4.619619930671161e-06, + "loss": 0.56, + "step": 16099 + }, + { + "epoch": 0.7128388475034036, + "grad_norm": 1.5961159467697144, + "learning_rate": 4.618317302534786e-06, + "loss": 0.6192, + "step": 16100 + }, + { + "epoch": 0.7128831232082176, + "grad_norm": 1.5720579624176025, + "learning_rate": 4.617014802935524e-06, + "loss": 0.5231, + "step": 16101 + }, + { + "epoch": 0.7129273989130315, + "grad_norm": 1.5957821607589722, + "learning_rate": 4.615712431904481e-06, + "loss": 0.5689, + "step": 16102 + }, + { + "epoch": 0.7129716746178453, + "grad_norm": 1.3174704313278198, + "learning_rate": 4.6144101894727586e-06, + "loss": 0.5175, + "step": 16103 + }, + { + "epoch": 0.7130159503226592, + "grad_norm": 1.2994086742401123, + "learning_rate": 4.613108075671467e-06, + "loss": 0.4884, + "step": 16104 + }, + { + "epoch": 0.7130602260274731, + "grad_norm": 1.3400155305862427, + "learning_rate": 4.611806090531704e-06, + "loss": 0.5384, + "step": 16105 + }, + { + "epoch": 0.7131045017322869, + "grad_norm": 1.4632055759429932, + "learning_rate": 4.610504234084561e-06, + "loss": 0.5679, + "step": 16106 + }, + { + "epoch": 0.7131487774371008, + "grad_norm": 1.5464682579040527, + "learning_rate": 4.60920250636114e-06, + "loss": 0.5292, + "step": 16107 + }, + { + "epoch": 0.7131930531419147, + "grad_norm": 1.594412088394165, + "learning_rate": 4.607900907392531e-06, + "loss": 0.5195, + "step": 16108 + }, + { + "epoch": 0.7132373288467285, + "grad_norm": 1.3626149892807007, + "learning_rate": 4.606599437209815e-06, + "loss": 0.5462, + "step": 16109 + }, + { + "epoch": 0.7132816045515424, + "grad_norm": 1.4722682237625122, + "learning_rate": 4.605298095844076e-06, + "loss": 0.5593, + "step": 16110 + }, + { + "epoch": 0.7133258802563563, + "grad_norm": 1.4742969274520874, + "learning_rate": 4.603996883326403e-06, + "loss": 0.5639, + "step": 16111 + }, + { + "epoch": 0.7133701559611703, + "grad_norm": 1.3811068534851074, + "learning_rate": 4.602695799687872e-06, + "loss": 0.5522, + "step": 16112 + }, + { + "epoch": 0.7134144316659841, + "grad_norm": 1.8997794389724731, + "learning_rate": 4.601394844959553e-06, + "loss": 0.5542, + "step": 16113 + }, + { + "epoch": 0.713458707370798, + "grad_norm": 1.3978408575057983, + "learning_rate": 4.600094019172525e-06, + "loss": 0.478, + "step": 16114 + }, + { + "epoch": 0.7135029830756119, + "grad_norm": 1.3825414180755615, + "learning_rate": 4.598793322357855e-06, + "loss": 0.5294, + "step": 16115 + }, + { + "epoch": 0.7135472587804257, + "grad_norm": 1.334854006767273, + "learning_rate": 4.59749275454661e-06, + "loss": 0.5095, + "step": 16116 + }, + { + "epoch": 0.7135915344852396, + "grad_norm": 1.5175278186798096, + "learning_rate": 4.59619231576985e-06, + "loss": 0.5074, + "step": 16117 + }, + { + "epoch": 0.7136358101900535, + "grad_norm": 1.4233349561691284, + "learning_rate": 4.594892006058637e-06, + "loss": 0.5635, + "step": 16118 + }, + { + "epoch": 0.7136800858948673, + "grad_norm": 1.4083982706069946, + "learning_rate": 4.593591825444028e-06, + "loss": 0.4781, + "step": 16119 + }, + { + "epoch": 0.7137243615996812, + "grad_norm": 1.9370121955871582, + "learning_rate": 4.5922917739570724e-06, + "loss": 0.611, + "step": 16120 + }, + { + "epoch": 0.7137686373044951, + "grad_norm": 1.8632596731185913, + "learning_rate": 4.590991851628828e-06, + "loss": 0.5313, + "step": 16121 + }, + { + "epoch": 0.7138129130093089, + "grad_norm": 1.440276026725769, + "learning_rate": 4.589692058490339e-06, + "loss": 0.5685, + "step": 16122 + }, + { + "epoch": 0.7138571887141228, + "grad_norm": 1.5176876783370972, + "learning_rate": 4.588392394572646e-06, + "loss": 0.528, + "step": 16123 + }, + { + "epoch": 0.7139014644189368, + "grad_norm": 1.2419276237487793, + "learning_rate": 4.587092859906799e-06, + "loss": 0.4637, + "step": 16124 + }, + { + "epoch": 0.7139457401237506, + "grad_norm": 1.3672784566879272, + "learning_rate": 4.585793454523832e-06, + "loss": 0.5266, + "step": 16125 + }, + { + "epoch": 0.7139900158285645, + "grad_norm": 1.365078330039978, + "learning_rate": 4.584494178454778e-06, + "loss": 0.5187, + "step": 16126 + }, + { + "epoch": 0.7140342915333784, + "grad_norm": 1.4279918670654297, + "learning_rate": 4.5831950317306725e-06, + "loss": 0.5294, + "step": 16127 + }, + { + "epoch": 0.7140785672381922, + "grad_norm": 1.2045236825942993, + "learning_rate": 4.581896014382543e-06, + "loss": 0.4799, + "step": 16128 + }, + { + "epoch": 0.7141228429430061, + "grad_norm": 1.3046047687530518, + "learning_rate": 4.580597126441415e-06, + "loss": 0.5416, + "step": 16129 + }, + { + "epoch": 0.71416711864782, + "grad_norm": 1.3192176818847656, + "learning_rate": 4.579298367938309e-06, + "loss": 0.5133, + "step": 16130 + }, + { + "epoch": 0.7142113943526338, + "grad_norm": 1.37798011302948, + "learning_rate": 4.5779997389042514e-06, + "loss": 0.5538, + "step": 16131 + }, + { + "epoch": 0.7142556700574477, + "grad_norm": 1.5070101022720337, + "learning_rate": 4.5767012393702556e-06, + "loss": 0.5344, + "step": 16132 + }, + { + "epoch": 0.7142999457622616, + "grad_norm": 1.3846181631088257, + "learning_rate": 4.575402869367331e-06, + "loss": 0.5176, + "step": 16133 + }, + { + "epoch": 0.7143442214670754, + "grad_norm": 1.7486653327941895, + "learning_rate": 4.574104628926494e-06, + "loss": 0.4914, + "step": 16134 + }, + { + "epoch": 0.7143884971718893, + "grad_norm": 1.6905204057693481, + "learning_rate": 4.572806518078753e-06, + "loss": 0.5351, + "step": 16135 + }, + { + "epoch": 0.7144327728767033, + "grad_norm": 1.3127917051315308, + "learning_rate": 4.571508536855107e-06, + "loss": 0.5045, + "step": 16136 + }, + { + "epoch": 0.7144770485815171, + "grad_norm": 1.2689735889434814, + "learning_rate": 4.57021068528656e-06, + "loss": 0.5192, + "step": 16137 + }, + { + "epoch": 0.714521324286331, + "grad_norm": 1.4794788360595703, + "learning_rate": 4.568912963404111e-06, + "loss": 0.5255, + "step": 16138 + }, + { + "epoch": 0.7145655999911449, + "grad_norm": 1.549248218536377, + "learning_rate": 4.5676153712387516e-06, + "loss": 0.5319, + "step": 16139 + }, + { + "epoch": 0.7146098756959587, + "grad_norm": 1.475813865661621, + "learning_rate": 4.566317908821474e-06, + "loss": 0.5575, + "step": 16140 + }, + { + "epoch": 0.7146541514007726, + "grad_norm": 1.327539086341858, + "learning_rate": 4.565020576183271e-06, + "loss": 0.5086, + "step": 16141 + }, + { + "epoch": 0.7146984271055865, + "grad_norm": 1.336574912071228, + "learning_rate": 4.563723373355128e-06, + "loss": 0.494, + "step": 16142 + }, + { + "epoch": 0.7147427028104003, + "grad_norm": 1.2798330783843994, + "learning_rate": 4.56242630036802e-06, + "loss": 0.4656, + "step": 16143 + }, + { + "epoch": 0.7147869785152142, + "grad_norm": 1.508797526359558, + "learning_rate": 4.561129357252938e-06, + "loss": 0.5521, + "step": 16144 + }, + { + "epoch": 0.7148312542200281, + "grad_norm": 1.4769716262817383, + "learning_rate": 4.559832544040852e-06, + "loss": 0.5182, + "step": 16145 + }, + { + "epoch": 0.714875529924842, + "grad_norm": 1.2195488214492798, + "learning_rate": 4.558535860762736e-06, + "loss": 0.5147, + "step": 16146 + }, + { + "epoch": 0.7149198056296558, + "grad_norm": 1.6734145879745483, + "learning_rate": 4.557239307449562e-06, + "loss": 0.5788, + "step": 16147 + }, + { + "epoch": 0.7149640813344698, + "grad_norm": 1.4057387113571167, + "learning_rate": 4.555942884132293e-06, + "loss": 0.546, + "step": 16148 + }, + { + "epoch": 0.7150083570392837, + "grad_norm": 1.6317782402038574, + "learning_rate": 4.554646590841897e-06, + "loss": 0.5784, + "step": 16149 + }, + { + "epoch": 0.7150526327440975, + "grad_norm": 1.553285002708435, + "learning_rate": 4.5533504276093286e-06, + "loss": 0.5552, + "step": 16150 + }, + { + "epoch": 0.7150969084489114, + "grad_norm": 1.5903476476669312, + "learning_rate": 4.552054394465555e-06, + "loss": 0.51, + "step": 16151 + }, + { + "epoch": 0.7151411841537253, + "grad_norm": 1.8915255069732666, + "learning_rate": 4.550758491441526e-06, + "loss": 0.6501, + "step": 16152 + }, + { + "epoch": 0.7151854598585391, + "grad_norm": 1.7215652465820312, + "learning_rate": 4.54946271856819e-06, + "loss": 0.6005, + "step": 16153 + }, + { + "epoch": 0.715229735563353, + "grad_norm": 1.6815838813781738, + "learning_rate": 4.548167075876503e-06, + "loss": 0.5616, + "step": 16154 + }, + { + "epoch": 0.7152740112681669, + "grad_norm": 1.3995294570922852, + "learning_rate": 4.546871563397409e-06, + "loss": 0.4888, + "step": 16155 + }, + { + "epoch": 0.7153182869729807, + "grad_norm": 1.3818095922470093, + "learning_rate": 4.54557618116184e-06, + "loss": 0.4866, + "step": 16156 + }, + { + "epoch": 0.7153625626777946, + "grad_norm": 1.5546115636825562, + "learning_rate": 4.544280929200748e-06, + "loss": 0.541, + "step": 16157 + }, + { + "epoch": 0.7154068383826085, + "grad_norm": 1.3563554286956787, + "learning_rate": 4.5429858075450626e-06, + "loss": 0.5252, + "step": 16158 + }, + { + "epoch": 0.7154511140874223, + "grad_norm": 1.44959557056427, + "learning_rate": 4.541690816225719e-06, + "loss": 0.5281, + "step": 16159 + }, + { + "epoch": 0.7154953897922363, + "grad_norm": 1.658333420753479, + "learning_rate": 4.540395955273642e-06, + "loss": 0.5401, + "step": 16160 + }, + { + "epoch": 0.7155396654970502, + "grad_norm": 1.5983171463012695, + "learning_rate": 4.539101224719765e-06, + "loss": 0.6055, + "step": 16161 + }, + { + "epoch": 0.715583941201864, + "grad_norm": 1.357103943824768, + "learning_rate": 4.53780662459501e-06, + "loss": 0.4962, + "step": 16162 + }, + { + "epoch": 0.7156282169066779, + "grad_norm": 1.8059779405593872, + "learning_rate": 4.5365121549302916e-06, + "loss": 0.5796, + "step": 16163 + }, + { + "epoch": 0.7156724926114918, + "grad_norm": 1.7499960660934448, + "learning_rate": 4.5352178157565395e-06, + "loss": 0.6014, + "step": 16164 + }, + { + "epoch": 0.7157167683163056, + "grad_norm": 1.560666799545288, + "learning_rate": 4.5339236071046565e-06, + "loss": 0.5368, + "step": 16165 + }, + { + "epoch": 0.7157610440211195, + "grad_norm": 1.4735996723175049, + "learning_rate": 4.532629529005554e-06, + "loss": 0.5738, + "step": 16166 + }, + { + "epoch": 0.7158053197259334, + "grad_norm": 1.860486626625061, + "learning_rate": 4.531335581490147e-06, + "loss": 0.5405, + "step": 16167 + }, + { + "epoch": 0.7158495954307472, + "grad_norm": 1.5581023693084717, + "learning_rate": 4.530041764589337e-06, + "loss": 0.5264, + "step": 16168 + }, + { + "epoch": 0.7158938711355611, + "grad_norm": 1.6269862651824951, + "learning_rate": 4.528748078334025e-06, + "loss": 0.5826, + "step": 16169 + }, + { + "epoch": 0.715938146840375, + "grad_norm": 1.293784260749817, + "learning_rate": 4.527454522755108e-06, + "loss": 0.5047, + "step": 16170 + }, + { + "epoch": 0.7159824225451888, + "grad_norm": 1.6012176275253296, + "learning_rate": 4.526161097883487e-06, + "loss": 0.5767, + "step": 16171 + }, + { + "epoch": 0.7160266982500028, + "grad_norm": 1.6616361141204834, + "learning_rate": 4.524867803750052e-06, + "loss": 0.5433, + "step": 16172 + }, + { + "epoch": 0.7160709739548167, + "grad_norm": 1.61221444606781, + "learning_rate": 4.523574640385689e-06, + "loss": 0.5756, + "step": 16173 + }, + { + "epoch": 0.7161152496596305, + "grad_norm": 1.3428475856781006, + "learning_rate": 4.522281607821288e-06, + "loss": 0.5047, + "step": 16174 + }, + { + "epoch": 0.7161595253644444, + "grad_norm": 1.4172581434249878, + "learning_rate": 4.520988706087731e-06, + "loss": 0.4804, + "step": 16175 + }, + { + "epoch": 0.7162038010692583, + "grad_norm": 1.4252713918685913, + "learning_rate": 4.519695935215895e-06, + "loss": 0.5839, + "step": 16176 + }, + { + "epoch": 0.7162480767740721, + "grad_norm": 1.429413914680481, + "learning_rate": 4.518403295236661e-06, + "loss": 0.563, + "step": 16177 + }, + { + "epoch": 0.716292352478886, + "grad_norm": 1.5307496786117554, + "learning_rate": 4.517110786180902e-06, + "loss": 0.5078, + "step": 16178 + }, + { + "epoch": 0.7163366281836999, + "grad_norm": 1.5300922393798828, + "learning_rate": 4.515818408079487e-06, + "loss": 0.517, + "step": 16179 + }, + { + "epoch": 0.7163809038885137, + "grad_norm": 1.5872880220413208, + "learning_rate": 4.51452616096328e-06, + "loss": 0.5949, + "step": 16180 + }, + { + "epoch": 0.7164251795933276, + "grad_norm": 1.8054033517837524, + "learning_rate": 4.513234044863154e-06, + "loss": 0.5988, + "step": 16181 + }, + { + "epoch": 0.7164694552981415, + "grad_norm": 1.5608662366867065, + "learning_rate": 4.511942059809965e-06, + "loss": 0.545, + "step": 16182 + }, + { + "epoch": 0.7165137310029555, + "grad_norm": 1.3469144105911255, + "learning_rate": 4.510650205834571e-06, + "loss": 0.5201, + "step": 16183 + }, + { + "epoch": 0.7165580067077693, + "grad_norm": 1.3443320989608765, + "learning_rate": 4.509358482967827e-06, + "loss": 0.5086, + "step": 16184 + }, + { + "epoch": 0.7166022824125832, + "grad_norm": 1.2557353973388672, + "learning_rate": 4.508066891240586e-06, + "loss": 0.4826, + "step": 16185 + }, + { + "epoch": 0.7166465581173971, + "grad_norm": 1.222291111946106, + "learning_rate": 4.506775430683695e-06, + "loss": 0.5049, + "step": 16186 + }, + { + "epoch": 0.7166908338222109, + "grad_norm": 1.555789589881897, + "learning_rate": 4.505484101327996e-06, + "loss": 0.5508, + "step": 16187 + }, + { + "epoch": 0.7167351095270248, + "grad_norm": 1.407829999923706, + "learning_rate": 4.504192903204339e-06, + "loss": 0.5683, + "step": 16188 + }, + { + "epoch": 0.7167793852318387, + "grad_norm": 1.3877131938934326, + "learning_rate": 4.502901836343561e-06, + "loss": 0.521, + "step": 16189 + }, + { + "epoch": 0.7168236609366525, + "grad_norm": 1.7528634071350098, + "learning_rate": 4.501610900776492e-06, + "loss": 0.587, + "step": 16190 + }, + { + "epoch": 0.7168679366414664, + "grad_norm": 1.4579817056655884, + "learning_rate": 4.500320096533973e-06, + "loss": 0.5639, + "step": 16191 + }, + { + "epoch": 0.7169122123462803, + "grad_norm": 1.526615023612976, + "learning_rate": 4.499029423646831e-06, + "loss": 0.5131, + "step": 16192 + }, + { + "epoch": 0.7169564880510941, + "grad_norm": 1.3444998264312744, + "learning_rate": 4.497738882145893e-06, + "loss": 0.4985, + "step": 16193 + }, + { + "epoch": 0.717000763755908, + "grad_norm": 1.4759348630905151, + "learning_rate": 4.49644847206198e-06, + "loss": 0.5123, + "step": 16194 + }, + { + "epoch": 0.717045039460722, + "grad_norm": 1.575198769569397, + "learning_rate": 4.495158193425916e-06, + "loss": 0.477, + "step": 16195 + }, + { + "epoch": 0.7170893151655358, + "grad_norm": 1.598254680633545, + "learning_rate": 4.493868046268514e-06, + "loss": 0.5247, + "step": 16196 + }, + { + "epoch": 0.7171335908703497, + "grad_norm": 1.8522578477859497, + "learning_rate": 4.492578030620589e-06, + "loss": 0.5822, + "step": 16197 + }, + { + "epoch": 0.7171778665751636, + "grad_norm": 2.3367862701416016, + "learning_rate": 4.491288146512957e-06, + "loss": 0.5195, + "step": 16198 + }, + { + "epoch": 0.7172221422799774, + "grad_norm": 1.5496631860733032, + "learning_rate": 4.489998393976421e-06, + "loss": 0.5631, + "step": 16199 + }, + { + "epoch": 0.7172664179847913, + "grad_norm": 1.4076409339904785, + "learning_rate": 4.4887087730417836e-06, + "loss": 0.5208, + "step": 16200 + }, + { + "epoch": 0.7173106936896052, + "grad_norm": 2.2121593952178955, + "learning_rate": 4.487419283739853e-06, + "loss": 0.6395, + "step": 16201 + }, + { + "epoch": 0.717354969394419, + "grad_norm": 1.426730751991272, + "learning_rate": 4.486129926101426e-06, + "loss": 0.5482, + "step": 16202 + }, + { + "epoch": 0.7173992450992329, + "grad_norm": 1.4616436958312988, + "learning_rate": 4.4848407001572945e-06, + "loss": 0.5335, + "step": 16203 + }, + { + "epoch": 0.7174435208040468, + "grad_norm": 1.3286737203598022, + "learning_rate": 4.4835516059382536e-06, + "loss": 0.5304, + "step": 16204 + }, + { + "epoch": 0.7174877965088606, + "grad_norm": 1.8119921684265137, + "learning_rate": 4.482262643475089e-06, + "loss": 0.6062, + "step": 16205 + }, + { + "epoch": 0.7175320722136745, + "grad_norm": 1.5701395273208618, + "learning_rate": 4.4809738127985905e-06, + "loss": 0.5484, + "step": 16206 + }, + { + "epoch": 0.7175763479184885, + "grad_norm": 1.21664559841156, + "learning_rate": 4.479685113939533e-06, + "loss": 0.4738, + "step": 16207 + }, + { + "epoch": 0.7176206236233023, + "grad_norm": 1.3707170486450195, + "learning_rate": 4.478396546928707e-06, + "loss": 0.4939, + "step": 16208 + }, + { + "epoch": 0.7176648993281162, + "grad_norm": 1.514472484588623, + "learning_rate": 4.477108111796883e-06, + "loss": 0.5187, + "step": 16209 + }, + { + "epoch": 0.7177091750329301, + "grad_norm": 1.6780366897583008, + "learning_rate": 4.4758198085748316e-06, + "loss": 0.5626, + "step": 16210 + }, + { + "epoch": 0.7177534507377439, + "grad_norm": 2.081027030944824, + "learning_rate": 4.474531637293333e-06, + "loss": 0.6419, + "step": 16211 + }, + { + "epoch": 0.7177977264425578, + "grad_norm": 1.7730307579040527, + "learning_rate": 4.473243597983144e-06, + "loss": 0.6626, + "step": 16212 + }, + { + "epoch": 0.7178420021473717, + "grad_norm": 1.3140586614608765, + "learning_rate": 4.471955690675027e-06, + "loss": 0.508, + "step": 16213 + }, + { + "epoch": 0.7178862778521855, + "grad_norm": 1.5784242153167725, + "learning_rate": 4.470667915399752e-06, + "loss": 0.5649, + "step": 16214 + }, + { + "epoch": 0.7179305535569994, + "grad_norm": 1.4872442483901978, + "learning_rate": 4.4693802721880705e-06, + "loss": 0.5332, + "step": 16215 + }, + { + "epoch": 0.7179748292618133, + "grad_norm": 1.5997639894485474, + "learning_rate": 4.468092761070739e-06, + "loss": 0.5413, + "step": 16216 + }, + { + "epoch": 0.7180191049666272, + "grad_norm": 1.3185762166976929, + "learning_rate": 4.466805382078504e-06, + "loss": 0.4956, + "step": 16217 + }, + { + "epoch": 0.718063380671441, + "grad_norm": 1.3566814661026, + "learning_rate": 4.46551813524212e-06, + "loss": 0.4928, + "step": 16218 + }, + { + "epoch": 0.718107656376255, + "grad_norm": 1.3158531188964844, + "learning_rate": 4.46423102059233e-06, + "loss": 0.5346, + "step": 16219 + }, + { + "epoch": 0.7181519320810689, + "grad_norm": 1.583107352256775, + "learning_rate": 4.462944038159874e-06, + "loss": 0.5915, + "step": 16220 + }, + { + "epoch": 0.7181962077858827, + "grad_norm": 1.310649037361145, + "learning_rate": 4.461657187975492e-06, + "loss": 0.49, + "step": 16221 + }, + { + "epoch": 0.7182404834906966, + "grad_norm": 1.34696626663208, + "learning_rate": 4.460370470069918e-06, + "loss": 0.4948, + "step": 16222 + }, + { + "epoch": 0.7182847591955105, + "grad_norm": 1.7214070558547974, + "learning_rate": 4.459083884473882e-06, + "loss": 0.5367, + "step": 16223 + }, + { + "epoch": 0.7183290349003243, + "grad_norm": 1.2537208795547485, + "learning_rate": 4.457797431218118e-06, + "loss": 0.4954, + "step": 16224 + }, + { + "epoch": 0.7183733106051382, + "grad_norm": 1.5472064018249512, + "learning_rate": 4.45651111033335e-06, + "loss": 0.5603, + "step": 16225 + }, + { + "epoch": 0.7184175863099521, + "grad_norm": 1.954773187637329, + "learning_rate": 4.455224921850301e-06, + "loss": 0.5394, + "step": 16226 + }, + { + "epoch": 0.7184618620147659, + "grad_norm": 1.5008422136306763, + "learning_rate": 4.453938865799686e-06, + "loss": 0.5303, + "step": 16227 + }, + { + "epoch": 0.7185061377195798, + "grad_norm": 1.4517457485198975, + "learning_rate": 4.45265294221223e-06, + "loss": 0.5436, + "step": 16228 + }, + { + "epoch": 0.7185504134243937, + "grad_norm": 1.789866328239441, + "learning_rate": 4.451367151118641e-06, + "loss": 0.6064, + "step": 16229 + }, + { + "epoch": 0.7185946891292075, + "grad_norm": 2.109842538833618, + "learning_rate": 4.45008149254963e-06, + "loss": 0.5799, + "step": 16230 + }, + { + "epoch": 0.7186389648340215, + "grad_norm": 1.4196951389312744, + "learning_rate": 4.448795966535903e-06, + "loss": 0.529, + "step": 16231 + }, + { + "epoch": 0.7186832405388354, + "grad_norm": 2.2283496856689453, + "learning_rate": 4.447510573108165e-06, + "loss": 0.5503, + "step": 16232 + }, + { + "epoch": 0.7187275162436492, + "grad_norm": 1.7132294178009033, + "learning_rate": 4.4462253122971114e-06, + "loss": 0.6049, + "step": 16233 + }, + { + "epoch": 0.7187717919484631, + "grad_norm": 1.6014442443847656, + "learning_rate": 4.444940184133448e-06, + "loss": 0.5969, + "step": 16234 + }, + { + "epoch": 0.718816067653277, + "grad_norm": 1.5699114799499512, + "learning_rate": 4.443655188647867e-06, + "loss": 0.5373, + "step": 16235 + }, + { + "epoch": 0.7188603433580908, + "grad_norm": 1.3643648624420166, + "learning_rate": 4.442370325871055e-06, + "loss": 0.533, + "step": 16236 + }, + { + "epoch": 0.7189046190629047, + "grad_norm": 1.4591655731201172, + "learning_rate": 4.441085595833701e-06, + "loss": 0.5489, + "step": 16237 + }, + { + "epoch": 0.7189488947677186, + "grad_norm": 1.7799417972564697, + "learning_rate": 4.439800998566494e-06, + "loss": 0.5613, + "step": 16238 + }, + { + "epoch": 0.7189931704725324, + "grad_norm": 2.0454134941101074, + "learning_rate": 4.4385165341001125e-06, + "loss": 0.605, + "step": 16239 + }, + { + "epoch": 0.7190374461773463, + "grad_norm": 1.9287441968917847, + "learning_rate": 4.4372322024652356e-06, + "loss": 0.6302, + "step": 16240 + }, + { + "epoch": 0.7190817218821602, + "grad_norm": 1.61545991897583, + "learning_rate": 4.435948003692538e-06, + "loss": 0.5216, + "step": 16241 + }, + { + "epoch": 0.719125997586974, + "grad_norm": 1.4031391143798828, + "learning_rate": 4.434663937812693e-06, + "loss": 0.5698, + "step": 16242 + }, + { + "epoch": 0.719170273291788, + "grad_norm": 1.4013959169387817, + "learning_rate": 4.4333800048563635e-06, + "loss": 0.5214, + "step": 16243 + }, + { + "epoch": 0.7192145489966019, + "grad_norm": 1.5056320428848267, + "learning_rate": 4.432096204854225e-06, + "loss": 0.5219, + "step": 16244 + }, + { + "epoch": 0.7192588247014157, + "grad_norm": 1.3881330490112305, + "learning_rate": 4.430812537836934e-06, + "loss": 0.4524, + "step": 16245 + }, + { + "epoch": 0.7193031004062296, + "grad_norm": 1.5092461109161377, + "learning_rate": 4.429529003835152e-06, + "loss": 0.5329, + "step": 16246 + }, + { + "epoch": 0.7193473761110435, + "grad_norm": 1.7441420555114746, + "learning_rate": 4.42824560287953e-06, + "loss": 0.5706, + "step": 16247 + }, + { + "epoch": 0.7193916518158573, + "grad_norm": 1.5100702047348022, + "learning_rate": 4.426962335000727e-06, + "loss": 0.5423, + "step": 16248 + }, + { + "epoch": 0.7194359275206712, + "grad_norm": 2.7458419799804688, + "learning_rate": 4.425679200229392e-06, + "loss": 0.5895, + "step": 16249 + }, + { + "epoch": 0.7194802032254851, + "grad_norm": 1.7168539762496948, + "learning_rate": 4.424396198596172e-06, + "loss": 0.6022, + "step": 16250 + }, + { + "epoch": 0.7195244789302989, + "grad_norm": 1.5112559795379639, + "learning_rate": 4.423113330131708e-06, + "loss": 0.5403, + "step": 16251 + }, + { + "epoch": 0.7195687546351128, + "grad_norm": 1.493644118309021, + "learning_rate": 4.4218305948666405e-06, + "loss": 0.538, + "step": 16252 + }, + { + "epoch": 0.7196130303399267, + "grad_norm": 1.6855413913726807, + "learning_rate": 4.420547992831604e-06, + "loss": 0.5692, + "step": 16253 + }, + { + "epoch": 0.7196573060447407, + "grad_norm": 1.5427250862121582, + "learning_rate": 4.41926552405724e-06, + "loss": 0.5063, + "step": 16254 + }, + { + "epoch": 0.7197015817495545, + "grad_norm": 1.76377534866333, + "learning_rate": 4.417983188574174e-06, + "loss": 0.5935, + "step": 16255 + }, + { + "epoch": 0.7197458574543684, + "grad_norm": 1.5010898113250732, + "learning_rate": 4.4167009864130365e-06, + "loss": 0.5925, + "step": 16256 + }, + { + "epoch": 0.7197901331591823, + "grad_norm": 1.6748844385147095, + "learning_rate": 4.4154189176044436e-06, + "loss": 0.5278, + "step": 16257 + }, + { + "epoch": 0.7198344088639961, + "grad_norm": 1.5597004890441895, + "learning_rate": 4.414136982179028e-06, + "loss": 0.5566, + "step": 16258 + }, + { + "epoch": 0.71987868456881, + "grad_norm": 1.4345859289169312, + "learning_rate": 4.412855180167406e-06, + "loss": 0.5501, + "step": 16259 + }, + { + "epoch": 0.7199229602736239, + "grad_norm": 1.4648643732070923, + "learning_rate": 4.41157351160018e-06, + "loss": 0.5378, + "step": 16260 + }, + { + "epoch": 0.7199672359784377, + "grad_norm": 1.3147838115692139, + "learning_rate": 4.4102919765079755e-06, + "loss": 0.4987, + "step": 16261 + }, + { + "epoch": 0.7200115116832516, + "grad_norm": 1.4674689769744873, + "learning_rate": 4.409010574921397e-06, + "loss": 0.5382, + "step": 16262 + }, + { + "epoch": 0.7200557873880655, + "grad_norm": 1.3918626308441162, + "learning_rate": 4.4077293068710425e-06, + "loss": 0.5462, + "step": 16263 + }, + { + "epoch": 0.7201000630928793, + "grad_norm": 1.5875746011734009, + "learning_rate": 4.406448172387524e-06, + "loss": 0.581, + "step": 16264 + }, + { + "epoch": 0.7201443387976932, + "grad_norm": 1.9554380178451538, + "learning_rate": 4.405167171501438e-06, + "loss": 0.5951, + "step": 16265 + }, + { + "epoch": 0.7201886145025072, + "grad_norm": 1.257907509803772, + "learning_rate": 4.403886304243378e-06, + "loss": 0.4876, + "step": 16266 + }, + { + "epoch": 0.720232890207321, + "grad_norm": 1.4970248937606812, + "learning_rate": 4.402605570643936e-06, + "loss": 0.4927, + "step": 16267 + }, + { + "epoch": 0.7202771659121349, + "grad_norm": 1.2248305082321167, + "learning_rate": 4.401324970733704e-06, + "loss": 0.4388, + "step": 16268 + }, + { + "epoch": 0.7203214416169488, + "grad_norm": 1.7567510604858398, + "learning_rate": 4.400044504543265e-06, + "loss": 0.5951, + "step": 16269 + }, + { + "epoch": 0.7203657173217626, + "grad_norm": 1.4295247793197632, + "learning_rate": 4.3987641721032e-06, + "loss": 0.5338, + "step": 16270 + }, + { + "epoch": 0.7204099930265765, + "grad_norm": 1.3549067974090576, + "learning_rate": 4.397483973444097e-06, + "loss": 0.5224, + "step": 16271 + }, + { + "epoch": 0.7204542687313904, + "grad_norm": 1.4371764659881592, + "learning_rate": 4.396203908596526e-06, + "loss": 0.5051, + "step": 16272 + }, + { + "epoch": 0.7204985444362042, + "grad_norm": 1.4105720520019531, + "learning_rate": 4.394923977591059e-06, + "loss": 0.538, + "step": 16273 + }, + { + "epoch": 0.7205428201410181, + "grad_norm": 1.6103721857070923, + "learning_rate": 4.393644180458272e-06, + "loss": 0.5612, + "step": 16274 + }, + { + "epoch": 0.720587095845832, + "grad_norm": 1.3474637269973755, + "learning_rate": 4.39236451722873e-06, + "loss": 0.5074, + "step": 16275 + }, + { + "epoch": 0.7206313715506458, + "grad_norm": 1.2116018533706665, + "learning_rate": 4.391084987932995e-06, + "loss": 0.5221, + "step": 16276 + }, + { + "epoch": 0.7206756472554597, + "grad_norm": 1.8793846368789673, + "learning_rate": 4.389805592601628e-06, + "loss": 0.5909, + "step": 16277 + }, + { + "epoch": 0.7207199229602737, + "grad_norm": 1.3987911939620972, + "learning_rate": 4.388526331265186e-06, + "loss": 0.5069, + "step": 16278 + }, + { + "epoch": 0.7207641986650875, + "grad_norm": 1.4858174324035645, + "learning_rate": 4.3872472039542245e-06, + "loss": 0.4969, + "step": 16279 + }, + { + "epoch": 0.7208084743699014, + "grad_norm": 1.5575906038284302, + "learning_rate": 4.38596821069929e-06, + "loss": 0.5404, + "step": 16280 + }, + { + "epoch": 0.7208527500747153, + "grad_norm": 1.6295446157455444, + "learning_rate": 4.384689351530936e-06, + "loss": 0.527, + "step": 16281 + }, + { + "epoch": 0.7208970257795291, + "grad_norm": 1.5094045400619507, + "learning_rate": 4.383410626479705e-06, + "loss": 0.5599, + "step": 16282 + }, + { + "epoch": 0.720941301484343, + "grad_norm": 1.4592702388763428, + "learning_rate": 4.382132035576135e-06, + "loss": 0.536, + "step": 16283 + }, + { + "epoch": 0.7209855771891569, + "grad_norm": 1.4177272319793701, + "learning_rate": 4.380853578850771e-06, + "loss": 0.4855, + "step": 16284 + }, + { + "epoch": 0.7210298528939707, + "grad_norm": 1.301979660987854, + "learning_rate": 4.379575256334143e-06, + "loss": 0.4684, + "step": 16285 + }, + { + "epoch": 0.7210741285987846, + "grad_norm": 2.017085075378418, + "learning_rate": 4.378297068056783e-06, + "loss": 0.6003, + "step": 16286 + }, + { + "epoch": 0.7211184043035985, + "grad_norm": 1.4073612689971924, + "learning_rate": 4.377019014049223e-06, + "loss": 0.4912, + "step": 16287 + }, + { + "epoch": 0.7211626800084124, + "grad_norm": 1.3808505535125732, + "learning_rate": 4.375741094341983e-06, + "loss": 0.5619, + "step": 16288 + }, + { + "epoch": 0.7212069557132262, + "grad_norm": 1.692165493965149, + "learning_rate": 4.3744633089655884e-06, + "loss": 0.6072, + "step": 16289 + }, + { + "epoch": 0.7212512314180402, + "grad_norm": 1.7654318809509277, + "learning_rate": 4.373185657950554e-06, + "loss": 0.5174, + "step": 16290 + }, + { + "epoch": 0.7212955071228541, + "grad_norm": 1.4447916746139526, + "learning_rate": 4.371908141327401e-06, + "loss": 0.5169, + "step": 16291 + }, + { + "epoch": 0.7213397828276679, + "grad_norm": 1.2727832794189453, + "learning_rate": 4.37063075912664e-06, + "loss": 0.4615, + "step": 16292 + }, + { + "epoch": 0.7213840585324818, + "grad_norm": 1.7958840131759644, + "learning_rate": 4.36935351137878e-06, + "loss": 0.5521, + "step": 16293 + }, + { + "epoch": 0.7214283342372957, + "grad_norm": 1.77040433883667, + "learning_rate": 4.368076398114324e-06, + "loss": 0.6522, + "step": 16294 + }, + { + "epoch": 0.7214726099421095, + "grad_norm": 1.8912239074707031, + "learning_rate": 4.3667994193637794e-06, + "loss": 0.5656, + "step": 16295 + }, + { + "epoch": 0.7215168856469234, + "grad_norm": 1.424756407737732, + "learning_rate": 4.365522575157645e-06, + "loss": 0.4882, + "step": 16296 + }, + { + "epoch": 0.7215611613517373, + "grad_norm": 1.919845461845398, + "learning_rate": 4.364245865526415e-06, + "loss": 0.505, + "step": 16297 + }, + { + "epoch": 0.7216054370565511, + "grad_norm": 1.6890199184417725, + "learning_rate": 4.362969290500585e-06, + "loss": 0.5306, + "step": 16298 + }, + { + "epoch": 0.721649712761365, + "grad_norm": 1.3378727436065674, + "learning_rate": 4.361692850110644e-06, + "loss": 0.4527, + "step": 16299 + }, + { + "epoch": 0.7216939884661789, + "grad_norm": 1.6051521301269531, + "learning_rate": 4.360416544387074e-06, + "loss": 0.5613, + "step": 16300 + }, + { + "epoch": 0.7217382641709927, + "grad_norm": 1.5061928033828735, + "learning_rate": 4.359140373360367e-06, + "loss": 0.4922, + "step": 16301 + }, + { + "epoch": 0.7217825398758067, + "grad_norm": 1.4437482357025146, + "learning_rate": 4.357864337060997e-06, + "loss": 0.5527, + "step": 16302 + }, + { + "epoch": 0.7218268155806206, + "grad_norm": 1.5235753059387207, + "learning_rate": 4.356588435519445e-06, + "loss": 0.5651, + "step": 16303 + }, + { + "epoch": 0.7218710912854344, + "grad_norm": 1.5360052585601807, + "learning_rate": 4.3553126687661805e-06, + "loss": 0.5189, + "step": 16304 + }, + { + "epoch": 0.7219153669902483, + "grad_norm": 1.4051545858383179, + "learning_rate": 4.35403703683168e-06, + "loss": 0.5289, + "step": 16305 + }, + { + "epoch": 0.7219596426950622, + "grad_norm": 1.3573633432388306, + "learning_rate": 4.352761539746411e-06, + "loss": 0.4904, + "step": 16306 + }, + { + "epoch": 0.722003918399876, + "grad_norm": 1.510266661643982, + "learning_rate": 4.351486177540827e-06, + "loss": 0.5563, + "step": 16307 + }, + { + "epoch": 0.7220481941046899, + "grad_norm": 1.479569673538208, + "learning_rate": 4.3502109502454e-06, + "loss": 0.5095, + "step": 16308 + }, + { + "epoch": 0.7220924698095038, + "grad_norm": 1.2729672193527222, + "learning_rate": 4.348935857890585e-06, + "loss": 0.4799, + "step": 16309 + }, + { + "epoch": 0.7221367455143176, + "grad_norm": 1.4624632596969604, + "learning_rate": 4.347660900506831e-06, + "loss": 0.5227, + "step": 16310 + }, + { + "epoch": 0.7221810212191315, + "grad_norm": 1.6479628086090088, + "learning_rate": 4.346386078124598e-06, + "loss": 0.596, + "step": 16311 + }, + { + "epoch": 0.7222252969239454, + "grad_norm": 1.599709391593933, + "learning_rate": 4.345111390774331e-06, + "loss": 0.5883, + "step": 16312 + }, + { + "epoch": 0.7222695726287592, + "grad_norm": 1.4367200136184692, + "learning_rate": 4.343836838486474e-06, + "loss": 0.4931, + "step": 16313 + }, + { + "epoch": 0.7223138483335731, + "grad_norm": 1.6157866716384888, + "learning_rate": 4.342562421291464e-06, + "loss": 0.5321, + "step": 16314 + }, + { + "epoch": 0.7223581240383871, + "grad_norm": 1.3735218048095703, + "learning_rate": 4.341288139219752e-06, + "loss": 0.4934, + "step": 16315 + }, + { + "epoch": 0.7224023997432009, + "grad_norm": 1.4733811616897583, + "learning_rate": 4.340013992301761e-06, + "loss": 0.5341, + "step": 16316 + }, + { + "epoch": 0.7224466754480148, + "grad_norm": 1.5994412899017334, + "learning_rate": 4.338739980567926e-06, + "loss": 0.5336, + "step": 16317 + }, + { + "epoch": 0.7224909511528287, + "grad_norm": 1.6353241205215454, + "learning_rate": 4.3374661040486776e-06, + "loss": 0.5524, + "step": 16318 + }, + { + "epoch": 0.7225352268576425, + "grad_norm": 1.757002830505371, + "learning_rate": 4.336192362774443e-06, + "loss": 0.5597, + "step": 16319 + }, + { + "epoch": 0.7225795025624564, + "grad_norm": 1.6056886911392212, + "learning_rate": 4.334918756775637e-06, + "loss": 0.529, + "step": 16320 + }, + { + "epoch": 0.7226237782672703, + "grad_norm": 1.5685014724731445, + "learning_rate": 4.333645286082689e-06, + "loss": 0.5172, + "step": 16321 + }, + { + "epoch": 0.7226680539720842, + "grad_norm": 1.439600944519043, + "learning_rate": 4.3323719507260075e-06, + "loss": 0.5209, + "step": 16322 + }, + { + "epoch": 0.722712329676898, + "grad_norm": 1.344712257385254, + "learning_rate": 4.331098750736008e-06, + "loss": 0.4998, + "step": 16323 + }, + { + "epoch": 0.7227566053817119, + "grad_norm": 1.6964868307113647, + "learning_rate": 4.3298256861431e-06, + "loss": 0.5759, + "step": 16324 + }, + { + "epoch": 0.7228008810865258, + "grad_norm": 1.7433013916015625, + "learning_rate": 4.3285527569776874e-06, + "loss": 0.602, + "step": 16325 + }, + { + "epoch": 0.7228451567913396, + "grad_norm": 1.3684262037277222, + "learning_rate": 4.327279963270176e-06, + "loss": 0.477, + "step": 16326 + }, + { + "epoch": 0.7228894324961536, + "grad_norm": 1.5434255599975586, + "learning_rate": 4.326007305050959e-06, + "loss": 0.5393, + "step": 16327 + }, + { + "epoch": 0.7229337082009675, + "grad_norm": 2.0439915657043457, + "learning_rate": 4.324734782350442e-06, + "loss": 0.5771, + "step": 16328 + }, + { + "epoch": 0.7229779839057813, + "grad_norm": 1.4848027229309082, + "learning_rate": 4.3234623951990115e-06, + "loss": 0.5247, + "step": 16329 + }, + { + "epoch": 0.7230222596105952, + "grad_norm": 1.5903844833374023, + "learning_rate": 4.322190143627059e-06, + "loss": 0.5393, + "step": 16330 + }, + { + "epoch": 0.7230665353154091, + "grad_norm": 1.3949896097183228, + "learning_rate": 4.320918027664973e-06, + "loss": 0.5631, + "step": 16331 + }, + { + "epoch": 0.7231108110202229, + "grad_norm": 1.4915989637374878, + "learning_rate": 4.319646047343138e-06, + "loss": 0.5221, + "step": 16332 + }, + { + "epoch": 0.7231550867250368, + "grad_norm": 1.4339460134506226, + "learning_rate": 4.3183742026919305e-06, + "loss": 0.5241, + "step": 16333 + }, + { + "epoch": 0.7231993624298507, + "grad_norm": 1.2338900566101074, + "learning_rate": 4.317102493741729e-06, + "loss": 0.468, + "step": 16334 + }, + { + "epoch": 0.7232436381346645, + "grad_norm": 1.5956355333328247, + "learning_rate": 4.315830920522909e-06, + "loss": 0.5452, + "step": 16335 + }, + { + "epoch": 0.7232879138394784, + "grad_norm": 1.714734435081482, + "learning_rate": 4.314559483065838e-06, + "loss": 0.5353, + "step": 16336 + }, + { + "epoch": 0.7233321895442923, + "grad_norm": 1.4740958213806152, + "learning_rate": 4.31328818140088e-06, + "loss": 0.5663, + "step": 16337 + }, + { + "epoch": 0.7233764652491061, + "grad_norm": 1.4594277143478394, + "learning_rate": 4.312017015558408e-06, + "loss": 0.5177, + "step": 16338 + }, + { + "epoch": 0.7234207409539201, + "grad_norm": 1.4377219676971436, + "learning_rate": 4.310745985568779e-06, + "loss": 0.5466, + "step": 16339 + }, + { + "epoch": 0.723465016658734, + "grad_norm": 1.6256572008132935, + "learning_rate": 4.309475091462345e-06, + "loss": 0.5879, + "step": 16340 + }, + { + "epoch": 0.7235092923635478, + "grad_norm": 1.3858301639556885, + "learning_rate": 4.308204333269469e-06, + "loss": 0.4773, + "step": 16341 + }, + { + "epoch": 0.7235535680683617, + "grad_norm": 1.4599398374557495, + "learning_rate": 4.3069337110204976e-06, + "loss": 0.5857, + "step": 16342 + }, + { + "epoch": 0.7235978437731756, + "grad_norm": 1.434929370880127, + "learning_rate": 4.30566322474578e-06, + "loss": 0.5086, + "step": 16343 + }, + { + "epoch": 0.7236421194779894, + "grad_norm": 1.4565043449401855, + "learning_rate": 4.304392874475659e-06, + "loss": 0.5411, + "step": 16344 + }, + { + "epoch": 0.7236863951828033, + "grad_norm": 1.8686730861663818, + "learning_rate": 4.303122660240478e-06, + "loss": 0.542, + "step": 16345 + }, + { + "epoch": 0.7237306708876172, + "grad_norm": 1.4322290420532227, + "learning_rate": 4.301852582070573e-06, + "loss": 0.5079, + "step": 16346 + }, + { + "epoch": 0.723774946592431, + "grad_norm": 1.5554646253585815, + "learning_rate": 4.300582639996274e-06, + "loss": 0.5308, + "step": 16347 + }, + { + "epoch": 0.7238192222972449, + "grad_norm": 1.5722324848175049, + "learning_rate": 4.299312834047924e-06, + "loss": 0.5619, + "step": 16348 + }, + { + "epoch": 0.7238634980020588, + "grad_norm": 1.4407098293304443, + "learning_rate": 4.298043164255845e-06, + "loss": 0.5245, + "step": 16349 + }, + { + "epoch": 0.7239077737068726, + "grad_norm": 1.333953857421875, + "learning_rate": 4.296773630650358e-06, + "loss": 0.4793, + "step": 16350 + }, + { + "epoch": 0.7239520494116866, + "grad_norm": 2.011045455932617, + "learning_rate": 4.295504233261793e-06, + "loss": 0.5155, + "step": 16351 + }, + { + "epoch": 0.7239963251165005, + "grad_norm": 1.6525675058364868, + "learning_rate": 4.294234972120465e-06, + "loss": 0.5549, + "step": 16352 + }, + { + "epoch": 0.7240406008213143, + "grad_norm": 1.5545680522918701, + "learning_rate": 4.292965847256687e-06, + "loss": 0.4864, + "step": 16353 + }, + { + "epoch": 0.7240848765261282, + "grad_norm": 1.7821904420852661, + "learning_rate": 4.291696858700774e-06, + "loss": 0.5855, + "step": 16354 + }, + { + "epoch": 0.7241291522309421, + "grad_norm": 1.7167435884475708, + "learning_rate": 4.290428006483035e-06, + "loss": 0.548, + "step": 16355 + }, + { + "epoch": 0.7241734279357559, + "grad_norm": 1.323778748512268, + "learning_rate": 4.289159290633773e-06, + "loss": 0.4826, + "step": 16356 + }, + { + "epoch": 0.7242177036405698, + "grad_norm": 1.501023530960083, + "learning_rate": 4.287890711183288e-06, + "loss": 0.5461, + "step": 16357 + }, + { + "epoch": 0.7242619793453837, + "grad_norm": 1.808059811592102, + "learning_rate": 4.286622268161886e-06, + "loss": 0.6188, + "step": 16358 + }, + { + "epoch": 0.7243062550501976, + "grad_norm": 1.649178385734558, + "learning_rate": 4.285353961599858e-06, + "loss": 0.4978, + "step": 16359 + }, + { + "epoch": 0.7243505307550114, + "grad_norm": 1.4643499851226807, + "learning_rate": 4.284085791527496e-06, + "loss": 0.5224, + "step": 16360 + }, + { + "epoch": 0.7243948064598253, + "grad_norm": 1.7422435283660889, + "learning_rate": 4.282817757975093e-06, + "loss": 0.6343, + "step": 16361 + }, + { + "epoch": 0.7244390821646393, + "grad_norm": 1.330075740814209, + "learning_rate": 4.281549860972936e-06, + "loss": 0.43, + "step": 16362 + }, + { + "epoch": 0.7244833578694531, + "grad_norm": 1.6413214206695557, + "learning_rate": 4.280282100551303e-06, + "loss": 0.5772, + "step": 16363 + }, + { + "epoch": 0.724527633574267, + "grad_norm": 1.8375595808029175, + "learning_rate": 4.27901447674047e-06, + "loss": 0.5606, + "step": 16364 + }, + { + "epoch": 0.7245719092790809, + "grad_norm": 1.5303235054016113, + "learning_rate": 4.277746989570721e-06, + "loss": 0.4665, + "step": 16365 + }, + { + "epoch": 0.7246161849838947, + "grad_norm": 1.8040615320205688, + "learning_rate": 4.2764796390723264e-06, + "loss": 0.5935, + "step": 16366 + }, + { + "epoch": 0.7246604606887086, + "grad_norm": 1.7057017087936401, + "learning_rate": 4.275212425275552e-06, + "loss": 0.6103, + "step": 16367 + }, + { + "epoch": 0.7247047363935225, + "grad_norm": 1.5058156251907349, + "learning_rate": 4.273945348210671e-06, + "loss": 0.5586, + "step": 16368 + }, + { + "epoch": 0.7247490120983363, + "grad_norm": 1.5792829990386963, + "learning_rate": 4.272678407907943e-06, + "loss": 0.5534, + "step": 16369 + }, + { + "epoch": 0.7247932878031502, + "grad_norm": 1.2962346076965332, + "learning_rate": 4.271411604397624e-06, + "loss": 0.4601, + "step": 16370 + }, + { + "epoch": 0.7248375635079641, + "grad_norm": 1.5131382942199707, + "learning_rate": 4.270144937709981e-06, + "loss": 0.5412, + "step": 16371 + }, + { + "epoch": 0.7248818392127779, + "grad_norm": 1.8199974298477173, + "learning_rate": 4.268878407875259e-06, + "loss": 0.5673, + "step": 16372 + }, + { + "epoch": 0.7249261149175918, + "grad_norm": 1.5343377590179443, + "learning_rate": 4.267612014923709e-06, + "loss": 0.5544, + "step": 16373 + }, + { + "epoch": 0.7249703906224058, + "grad_norm": 1.3736803531646729, + "learning_rate": 4.266345758885574e-06, + "loss": 0.4942, + "step": 16374 + }, + { + "epoch": 0.7250146663272196, + "grad_norm": 1.3499606847763062, + "learning_rate": 4.265079639791106e-06, + "loss": 0.5016, + "step": 16375 + }, + { + "epoch": 0.7250589420320335, + "grad_norm": 1.3312920331954956, + "learning_rate": 4.263813657670541e-06, + "loss": 0.5226, + "step": 16376 + }, + { + "epoch": 0.7251032177368474, + "grad_norm": 1.6120604276657104, + "learning_rate": 4.262547812554114e-06, + "loss": 0.5408, + "step": 16377 + }, + { + "epoch": 0.7251474934416612, + "grad_norm": 1.5352431535720825, + "learning_rate": 4.261282104472065e-06, + "loss": 0.5239, + "step": 16378 + }, + { + "epoch": 0.7251917691464751, + "grad_norm": 1.47660493850708, + "learning_rate": 4.2600165334546174e-06, + "loss": 0.5088, + "step": 16379 + }, + { + "epoch": 0.725236044851289, + "grad_norm": 1.4764529466629028, + "learning_rate": 4.258751099532004e-06, + "loss": 0.5232, + "step": 16380 + }, + { + "epoch": 0.7252803205561028, + "grad_norm": 1.331277847290039, + "learning_rate": 4.257485802734443e-06, + "loss": 0.4821, + "step": 16381 + }, + { + "epoch": 0.7253245962609167, + "grad_norm": 1.564923882484436, + "learning_rate": 4.25622064309216e-06, + "loss": 0.568, + "step": 16382 + }, + { + "epoch": 0.7253688719657306, + "grad_norm": 1.547325611114502, + "learning_rate": 4.254955620635371e-06, + "loss": 0.516, + "step": 16383 + }, + { + "epoch": 0.7254131476705444, + "grad_norm": 1.6484413146972656, + "learning_rate": 4.253690735394283e-06, + "loss": 0.6055, + "step": 16384 + }, + { + "epoch": 0.7254574233753583, + "grad_norm": 1.2835211753845215, + "learning_rate": 4.252425987399118e-06, + "loss": 0.4447, + "step": 16385 + }, + { + "epoch": 0.7255016990801723, + "grad_norm": 1.4623011350631714, + "learning_rate": 4.251161376680078e-06, + "loss": 0.5256, + "step": 16386 + }, + { + "epoch": 0.7255459747849861, + "grad_norm": 1.9253202676773071, + "learning_rate": 4.249896903267365e-06, + "loss": 0.5907, + "step": 16387 + }, + { + "epoch": 0.7255902504898, + "grad_norm": 1.1891270875930786, + "learning_rate": 4.248632567191185e-06, + "loss": 0.4972, + "step": 16388 + }, + { + "epoch": 0.7256345261946139, + "grad_norm": 1.9306979179382324, + "learning_rate": 4.2473683684817345e-06, + "loss": 0.6111, + "step": 16389 + }, + { + "epoch": 0.7256788018994277, + "grad_norm": 1.5020769834518433, + "learning_rate": 4.246104307169205e-06, + "loss": 0.5206, + "step": 16390 + }, + { + "epoch": 0.7257230776042416, + "grad_norm": 1.684079647064209, + "learning_rate": 4.24484038328379e-06, + "loss": 0.5977, + "step": 16391 + }, + { + "epoch": 0.7257673533090555, + "grad_norm": 1.369469404220581, + "learning_rate": 4.2435765968556775e-06, + "loss": 0.5008, + "step": 16392 + }, + { + "epoch": 0.7258116290138694, + "grad_norm": 1.3665480613708496, + "learning_rate": 4.242312947915049e-06, + "loss": 0.4835, + "step": 16393 + }, + { + "epoch": 0.7258559047186832, + "grad_norm": 1.6083624362945557, + "learning_rate": 4.241049436492085e-06, + "loss": 0.5964, + "step": 16394 + }, + { + "epoch": 0.7259001804234971, + "grad_norm": 1.9898276329040527, + "learning_rate": 4.23978606261697e-06, + "loss": 0.5973, + "step": 16395 + }, + { + "epoch": 0.725944456128311, + "grad_norm": 1.5400810241699219, + "learning_rate": 4.238522826319874e-06, + "loss": 0.5142, + "step": 16396 + }, + { + "epoch": 0.7259887318331248, + "grad_norm": 1.5752930641174316, + "learning_rate": 4.237259727630966e-06, + "loss": 0.5607, + "step": 16397 + }, + { + "epoch": 0.7260330075379388, + "grad_norm": 1.385524034500122, + "learning_rate": 4.23599676658042e-06, + "loss": 0.523, + "step": 16398 + }, + { + "epoch": 0.7260772832427527, + "grad_norm": 1.5868706703186035, + "learning_rate": 4.234733943198399e-06, + "loss": 0.5529, + "step": 16399 + }, + { + "epoch": 0.7261215589475665, + "grad_norm": 1.7355358600616455, + "learning_rate": 4.233471257515064e-06, + "loss": 0.601, + "step": 16400 + }, + { + "epoch": 0.7261658346523804, + "grad_norm": 1.7248811721801758, + "learning_rate": 4.2322087095605715e-06, + "loss": 0.5784, + "step": 16401 + }, + { + "epoch": 0.7262101103571943, + "grad_norm": 1.4042670726776123, + "learning_rate": 4.230946299365077e-06, + "loss": 0.467, + "step": 16402 + }, + { + "epoch": 0.7262543860620081, + "grad_norm": 1.8070818185806274, + "learning_rate": 4.229684026958734e-06, + "loss": 0.5386, + "step": 16403 + }, + { + "epoch": 0.726298661766822, + "grad_norm": 1.4530889987945557, + "learning_rate": 4.228421892371687e-06, + "loss": 0.5264, + "step": 16404 + }, + { + "epoch": 0.7263429374716359, + "grad_norm": 1.4910250902175903, + "learning_rate": 4.227159895634085e-06, + "loss": 0.5391, + "step": 16405 + }, + { + "epoch": 0.7263872131764497, + "grad_norm": 1.6700878143310547, + "learning_rate": 4.225898036776071e-06, + "loss": 0.5547, + "step": 16406 + }, + { + "epoch": 0.7264314888812636, + "grad_norm": 1.360833764076233, + "learning_rate": 4.2246363158277755e-06, + "loss": 0.4742, + "step": 16407 + }, + { + "epoch": 0.7264757645860775, + "grad_norm": 1.6039899587631226, + "learning_rate": 4.223374732819343e-06, + "loss": 0.5255, + "step": 16408 + }, + { + "epoch": 0.7265200402908913, + "grad_norm": 1.8376773595809937, + "learning_rate": 4.222113287780905e-06, + "loss": 0.59, + "step": 16409 + }, + { + "epoch": 0.7265643159957053, + "grad_norm": 1.425733208656311, + "learning_rate": 4.220851980742584e-06, + "loss": 0.5683, + "step": 16410 + }, + { + "epoch": 0.7266085917005192, + "grad_norm": 1.3510382175445557, + "learning_rate": 4.219590811734503e-06, + "loss": 0.5816, + "step": 16411 + }, + { + "epoch": 0.726652867405333, + "grad_norm": 1.1241505146026611, + "learning_rate": 4.218329780786794e-06, + "loss": 0.4435, + "step": 16412 + }, + { + "epoch": 0.7266971431101469, + "grad_norm": 1.5814813375473022, + "learning_rate": 4.21706888792957e-06, + "loss": 0.6005, + "step": 16413 + }, + { + "epoch": 0.7267414188149608, + "grad_norm": 1.6521525382995605, + "learning_rate": 4.215808133192942e-06, + "loss": 0.559, + "step": 16414 + }, + { + "epoch": 0.7267856945197746, + "grad_norm": 1.3752808570861816, + "learning_rate": 4.214547516607031e-06, + "loss": 0.4512, + "step": 16415 + }, + { + "epoch": 0.7268299702245885, + "grad_norm": 1.5493874549865723, + "learning_rate": 4.213287038201943e-06, + "loss": 0.5852, + "step": 16416 + }, + { + "epoch": 0.7268742459294024, + "grad_norm": 1.8851346969604492, + "learning_rate": 4.21202669800778e-06, + "loss": 0.605, + "step": 16417 + }, + { + "epoch": 0.7269185216342162, + "grad_norm": 1.6635680198669434, + "learning_rate": 4.210766496054651e-06, + "loss": 0.5639, + "step": 16418 + }, + { + "epoch": 0.7269627973390301, + "grad_norm": 1.4798510074615479, + "learning_rate": 4.2095064323726485e-06, + "loss": 0.5207, + "step": 16419 + }, + { + "epoch": 0.727007073043844, + "grad_norm": 1.54815673828125, + "learning_rate": 4.2082465069918685e-06, + "loss": 0.5163, + "step": 16420 + }, + { + "epoch": 0.7270513487486578, + "grad_norm": 1.8575297594070435, + "learning_rate": 4.206986719942403e-06, + "loss": 0.5254, + "step": 16421 + }, + { + "epoch": 0.7270956244534718, + "grad_norm": 1.4851678609848022, + "learning_rate": 4.205727071254346e-06, + "loss": 0.5288, + "step": 16422 + }, + { + "epoch": 0.7271399001582857, + "grad_norm": 1.5066980123519897, + "learning_rate": 4.20446756095778e-06, + "loss": 0.5539, + "step": 16423 + }, + { + "epoch": 0.7271841758630995, + "grad_norm": 1.5417206287384033, + "learning_rate": 4.203208189082784e-06, + "loss": 0.5106, + "step": 16424 + }, + { + "epoch": 0.7272284515679134, + "grad_norm": 1.4819046258926392, + "learning_rate": 4.2019489556594424e-06, + "loss": 0.5399, + "step": 16425 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 1.8878364562988281, + "learning_rate": 4.200689860717831e-06, + "loss": 0.6, + "step": 16426 + }, + { + "epoch": 0.7273170029775411, + "grad_norm": 1.8889628648757935, + "learning_rate": 4.19943090428802e-06, + "loss": 0.5958, + "step": 16427 + }, + { + "epoch": 0.727361278682355, + "grad_norm": 2.287353038787842, + "learning_rate": 4.198172086400079e-06, + "loss": 0.65, + "step": 16428 + }, + { + "epoch": 0.7274055543871689, + "grad_norm": 1.4232017993927002, + "learning_rate": 4.196913407084072e-06, + "loss": 0.5352, + "step": 16429 + }, + { + "epoch": 0.7274498300919828, + "grad_norm": 1.5124138593673706, + "learning_rate": 4.195654866370064e-06, + "loss": 0.5421, + "step": 16430 + }, + { + "epoch": 0.7274941057967966, + "grad_norm": 1.3013967275619507, + "learning_rate": 4.1943964642881105e-06, + "loss": 0.4888, + "step": 16431 + }, + { + "epoch": 0.7275383815016105, + "grad_norm": 1.2459766864776611, + "learning_rate": 4.193138200868274e-06, + "loss": 0.5294, + "step": 16432 + }, + { + "epoch": 0.7275826572064245, + "grad_norm": 1.5971413850784302, + "learning_rate": 4.191880076140603e-06, + "loss": 0.5383, + "step": 16433 + }, + { + "epoch": 0.7276269329112383, + "grad_norm": 1.6515686511993408, + "learning_rate": 4.1906220901351446e-06, + "loss": 0.569, + "step": 16434 + }, + { + "epoch": 0.7276712086160522, + "grad_norm": 1.9055310487747192, + "learning_rate": 4.18936424288195e-06, + "loss": 0.5703, + "step": 16435 + }, + { + "epoch": 0.7277154843208661, + "grad_norm": 1.6374282836914062, + "learning_rate": 4.188106534411059e-06, + "loss": 0.5605, + "step": 16436 + }, + { + "epoch": 0.7277597600256799, + "grad_norm": 1.491134762763977, + "learning_rate": 4.186848964752512e-06, + "loss": 0.5607, + "step": 16437 + }, + { + "epoch": 0.7278040357304938, + "grad_norm": 1.393673300743103, + "learning_rate": 4.185591533936344e-06, + "loss": 0.5241, + "step": 16438 + }, + { + "epoch": 0.7278483114353077, + "grad_norm": 1.7265836000442505, + "learning_rate": 4.184334241992588e-06, + "loss": 0.5524, + "step": 16439 + }, + { + "epoch": 0.7278925871401215, + "grad_norm": 1.4012264013290405, + "learning_rate": 4.183077088951273e-06, + "loss": 0.476, + "step": 16440 + }, + { + "epoch": 0.7279368628449354, + "grad_norm": 1.5829297304153442, + "learning_rate": 4.181820074842422e-06, + "loss": 0.5292, + "step": 16441 + }, + { + "epoch": 0.7279811385497493, + "grad_norm": 1.6551142930984497, + "learning_rate": 4.180563199696065e-06, + "loss": 0.5311, + "step": 16442 + }, + { + "epoch": 0.7280254142545631, + "grad_norm": 1.9422202110290527, + "learning_rate": 4.179306463542217e-06, + "loss": 0.5002, + "step": 16443 + }, + { + "epoch": 0.728069689959377, + "grad_norm": 1.4217348098754883, + "learning_rate": 4.178049866410891e-06, + "loss": 0.4773, + "step": 16444 + }, + { + "epoch": 0.728113965664191, + "grad_norm": 1.5000098943710327, + "learning_rate": 4.176793408332108e-06, + "loss": 0.504, + "step": 16445 + }, + { + "epoch": 0.7281582413690048, + "grad_norm": 1.4695723056793213, + "learning_rate": 4.175537089335873e-06, + "loss": 0.5303, + "step": 16446 + }, + { + "epoch": 0.7282025170738187, + "grad_norm": 1.4323886632919312, + "learning_rate": 4.174280909452192e-06, + "loss": 0.4781, + "step": 16447 + }, + { + "epoch": 0.7282467927786326, + "grad_norm": 2.6206233501434326, + "learning_rate": 4.173024868711068e-06, + "loss": 0.6851, + "step": 16448 + }, + { + "epoch": 0.7282910684834464, + "grad_norm": 1.4665199518203735, + "learning_rate": 4.1717689671425e-06, + "loss": 0.542, + "step": 16449 + }, + { + "epoch": 0.7283353441882603, + "grad_norm": 1.5210542678833008, + "learning_rate": 4.1705132047764854e-06, + "loss": 0.5898, + "step": 16450 + }, + { + "epoch": 0.7283796198930742, + "grad_norm": 1.5117461681365967, + "learning_rate": 4.169257581643013e-06, + "loss": 0.5307, + "step": 16451 + }, + { + "epoch": 0.728423895597888, + "grad_norm": 1.372071385383606, + "learning_rate": 4.168002097772078e-06, + "loss": 0.5074, + "step": 16452 + }, + { + "epoch": 0.7284681713027019, + "grad_norm": 1.6979385614395142, + "learning_rate": 4.1667467531936655e-06, + "loss": 0.5475, + "step": 16453 + }, + { + "epoch": 0.7285124470075158, + "grad_norm": 1.9196935892105103, + "learning_rate": 4.165491547937752e-06, + "loss": 0.5637, + "step": 16454 + }, + { + "epoch": 0.7285567227123296, + "grad_norm": 2.1218974590301514, + "learning_rate": 4.1642364820343276e-06, + "loss": 0.5559, + "step": 16455 + }, + { + "epoch": 0.7286009984171435, + "grad_norm": 1.6420197486877441, + "learning_rate": 4.162981555513362e-06, + "loss": 0.4961, + "step": 16456 + }, + { + "epoch": 0.7286452741219575, + "grad_norm": 1.6570011377334595, + "learning_rate": 4.16172676840483e-06, + "loss": 0.5538, + "step": 16457 + }, + { + "epoch": 0.7286895498267713, + "grad_norm": 1.7525641918182373, + "learning_rate": 4.160472120738701e-06, + "loss": 0.605, + "step": 16458 + }, + { + "epoch": 0.7287338255315852, + "grad_norm": 1.56900954246521, + "learning_rate": 4.159217612544939e-06, + "loss": 0.5741, + "step": 16459 + }, + { + "epoch": 0.7287781012363991, + "grad_norm": 1.430065393447876, + "learning_rate": 4.15796324385351e-06, + "loss": 0.518, + "step": 16460 + }, + { + "epoch": 0.7288223769412129, + "grad_norm": 1.5665546655654907, + "learning_rate": 4.156709014694368e-06, + "loss": 0.5136, + "step": 16461 + }, + { + "epoch": 0.7288666526460268, + "grad_norm": 1.9742391109466553, + "learning_rate": 4.1554549250974775e-06, + "loss": 0.5901, + "step": 16462 + }, + { + "epoch": 0.7289109283508407, + "grad_norm": 1.2966336011886597, + "learning_rate": 4.154200975092787e-06, + "loss": 0.4623, + "step": 16463 + }, + { + "epoch": 0.7289552040556546, + "grad_norm": 1.7530661821365356, + "learning_rate": 4.152947164710242e-06, + "loss": 0.5635, + "step": 16464 + }, + { + "epoch": 0.7289994797604684, + "grad_norm": 1.4439793825149536, + "learning_rate": 4.1516934939798015e-06, + "loss": 0.5491, + "step": 16465 + }, + { + "epoch": 0.7290437554652823, + "grad_norm": 1.7350413799285889, + "learning_rate": 4.150439962931395e-06, + "loss": 0.5438, + "step": 16466 + }, + { + "epoch": 0.7290880311700962, + "grad_norm": 1.4659358263015747, + "learning_rate": 4.149186571594965e-06, + "loss": 0.5222, + "step": 16467 + }, + { + "epoch": 0.72913230687491, + "grad_norm": 1.5638285875320435, + "learning_rate": 4.147933320000453e-06, + "loss": 0.5271, + "step": 16468 + }, + { + "epoch": 0.729176582579724, + "grad_norm": 1.6504207849502563, + "learning_rate": 4.146680208177788e-06, + "loss": 0.5443, + "step": 16469 + }, + { + "epoch": 0.7292208582845379, + "grad_norm": 1.9376109838485718, + "learning_rate": 4.145427236156901e-06, + "loss": 0.6014, + "step": 16470 + }, + { + "epoch": 0.7292651339893517, + "grad_norm": 1.7889360189437866, + "learning_rate": 4.144174403967714e-06, + "loss": 0.6162, + "step": 16471 + }, + { + "epoch": 0.7293094096941656, + "grad_norm": 1.3890836238861084, + "learning_rate": 4.1429217116401564e-06, + "loss": 0.5316, + "step": 16472 + }, + { + "epoch": 0.7293536853989795, + "grad_norm": 1.5613319873809814, + "learning_rate": 4.141669159204144e-06, + "loss": 0.5453, + "step": 16473 + }, + { + "epoch": 0.7293979611037933, + "grad_norm": 1.5485994815826416, + "learning_rate": 4.140416746689596e-06, + "loss": 0.5321, + "step": 16474 + }, + { + "epoch": 0.7294422368086072, + "grad_norm": 1.551911473274231, + "learning_rate": 4.139164474126421e-06, + "loss": 0.5588, + "step": 16475 + }, + { + "epoch": 0.7294865125134211, + "grad_norm": 1.5176547765731812, + "learning_rate": 4.1379123415445325e-06, + "loss": 0.5323, + "step": 16476 + }, + { + "epoch": 0.7295307882182349, + "grad_norm": 1.6246274709701538, + "learning_rate": 4.136660348973829e-06, + "loss": 0.542, + "step": 16477 + }, + { + "epoch": 0.7295750639230488, + "grad_norm": 1.7611708641052246, + "learning_rate": 4.135408496444223e-06, + "loss": 0.5774, + "step": 16478 + }, + { + "epoch": 0.7296193396278627, + "grad_norm": 1.441278338432312, + "learning_rate": 4.13415678398561e-06, + "loss": 0.5224, + "step": 16479 + }, + { + "epoch": 0.7296636153326765, + "grad_norm": 1.705527424812317, + "learning_rate": 4.132905211627887e-06, + "loss": 0.6424, + "step": 16480 + }, + { + "epoch": 0.7297078910374905, + "grad_norm": 1.5033470392227173, + "learning_rate": 4.13165377940094e-06, + "loss": 0.5842, + "step": 16481 + }, + { + "epoch": 0.7297521667423044, + "grad_norm": 1.5467523336410522, + "learning_rate": 4.13040248733467e-06, + "loss": 0.5164, + "step": 16482 + }, + { + "epoch": 0.7297964424471182, + "grad_norm": 1.6659743785858154, + "learning_rate": 4.1291513354589576e-06, + "loss": 0.5066, + "step": 16483 + }, + { + "epoch": 0.7298407181519321, + "grad_norm": 1.4304122924804688, + "learning_rate": 4.127900323803684e-06, + "loss": 0.5389, + "step": 16484 + }, + { + "epoch": 0.729884993856746, + "grad_norm": 1.4909796714782715, + "learning_rate": 4.12664945239873e-06, + "loss": 0.5325, + "step": 16485 + }, + { + "epoch": 0.7299292695615598, + "grad_norm": 1.5698587894439697, + "learning_rate": 4.125398721273972e-06, + "loss": 0.5067, + "step": 16486 + }, + { + "epoch": 0.7299735452663737, + "grad_norm": 1.5389658212661743, + "learning_rate": 4.124148130459279e-06, + "loss": 0.5835, + "step": 16487 + }, + { + "epoch": 0.7300178209711876, + "grad_norm": 1.5121915340423584, + "learning_rate": 4.122897679984526e-06, + "loss": 0.5015, + "step": 16488 + }, + { + "epoch": 0.7300620966760014, + "grad_norm": 1.8366502523422241, + "learning_rate": 4.1216473698795775e-06, + "loss": 0.5886, + "step": 16489 + }, + { + "epoch": 0.7301063723808153, + "grad_norm": 1.3402981758117676, + "learning_rate": 4.120397200174295e-06, + "loss": 0.5415, + "step": 16490 + }, + { + "epoch": 0.7301506480856292, + "grad_norm": 1.720410704612732, + "learning_rate": 4.119147170898535e-06, + "loss": 0.6083, + "step": 16491 + }, + { + "epoch": 0.730194923790443, + "grad_norm": 1.3396217823028564, + "learning_rate": 4.11789728208216e-06, + "loss": 0.4781, + "step": 16492 + }, + { + "epoch": 0.730239199495257, + "grad_norm": 2.153428792953491, + "learning_rate": 4.116647533755018e-06, + "loss": 0.6677, + "step": 16493 + }, + { + "epoch": 0.7302834752000709, + "grad_norm": 1.6690469980239868, + "learning_rate": 4.11539792594696e-06, + "loss": 0.5062, + "step": 16494 + }, + { + "epoch": 0.7303277509048847, + "grad_norm": 1.4879788160324097, + "learning_rate": 4.114148458687832e-06, + "loss": 0.5419, + "step": 16495 + }, + { + "epoch": 0.7303720266096986, + "grad_norm": 1.4580727815628052, + "learning_rate": 4.112899132007474e-06, + "loss": 0.518, + "step": 16496 + }, + { + "epoch": 0.7304163023145125, + "grad_norm": 1.5080007314682007, + "learning_rate": 4.111649945935724e-06, + "loss": 0.6208, + "step": 16497 + }, + { + "epoch": 0.7304605780193263, + "grad_norm": 1.3532432317733765, + "learning_rate": 4.1104009005024235e-06, + "loss": 0.515, + "step": 16498 + }, + { + "epoch": 0.7305048537241402, + "grad_norm": 1.2184001207351685, + "learning_rate": 4.109151995737404e-06, + "loss": 0.523, + "step": 16499 + }, + { + "epoch": 0.7305491294289541, + "grad_norm": 1.4793851375579834, + "learning_rate": 4.107903231670491e-06, + "loss": 0.5334, + "step": 16500 + }, + { + "epoch": 0.730593405133768, + "grad_norm": 1.6304618120193481, + "learning_rate": 4.106654608331507e-06, + "loss": 0.5431, + "step": 16501 + }, + { + "epoch": 0.7306376808385818, + "grad_norm": 1.5945343971252441, + "learning_rate": 4.105406125750284e-06, + "loss": 0.5034, + "step": 16502 + }, + { + "epoch": 0.7306819565433957, + "grad_norm": 1.421642541885376, + "learning_rate": 4.104157783956637e-06, + "loss": 0.5625, + "step": 16503 + }, + { + "epoch": 0.7307262322482097, + "grad_norm": 2.4208414554595947, + "learning_rate": 4.10290958298038e-06, + "loss": 0.5625, + "step": 16504 + }, + { + "epoch": 0.7307705079530235, + "grad_norm": 1.720594882965088, + "learning_rate": 4.101661522851326e-06, + "loss": 0.6239, + "step": 16505 + }, + { + "epoch": 0.7308147836578374, + "grad_norm": 1.3194823265075684, + "learning_rate": 4.1004136035992835e-06, + "loss": 0.5316, + "step": 16506 + }, + { + "epoch": 0.7308590593626513, + "grad_norm": 1.5318808555603027, + "learning_rate": 4.099165825254055e-06, + "loss": 0.5253, + "step": 16507 + }, + { + "epoch": 0.7309033350674651, + "grad_norm": 1.7277729511260986, + "learning_rate": 4.09791818784545e-06, + "loss": 0.5582, + "step": 16508 + }, + { + "epoch": 0.730947610772279, + "grad_norm": 1.4504358768463135, + "learning_rate": 4.096670691403264e-06, + "loss": 0.5118, + "step": 16509 + }, + { + "epoch": 0.7309918864770929, + "grad_norm": 1.5033107995986938, + "learning_rate": 4.095423335957291e-06, + "loss": 0.5648, + "step": 16510 + }, + { + "epoch": 0.7310361621819067, + "grad_norm": 1.6722369194030762, + "learning_rate": 4.094176121537321e-06, + "loss": 0.5834, + "step": 16511 + }, + { + "epoch": 0.7310804378867206, + "grad_norm": 1.502769112586975, + "learning_rate": 4.0929290481731485e-06, + "loss": 0.5383, + "step": 16512 + }, + { + "epoch": 0.7311247135915345, + "grad_norm": 1.5834238529205322, + "learning_rate": 4.09168211589456e-06, + "loss": 0.5407, + "step": 16513 + }, + { + "epoch": 0.7311689892963483, + "grad_norm": 1.3248594999313354, + "learning_rate": 4.090435324731325e-06, + "loss": 0.4516, + "step": 16514 + }, + { + "epoch": 0.7312132650011622, + "grad_norm": 1.2858808040618896, + "learning_rate": 4.0891886747132356e-06, + "loss": 0.4801, + "step": 16515 + }, + { + "epoch": 0.7312575407059762, + "grad_norm": 1.653293251991272, + "learning_rate": 4.087942165870063e-06, + "loss": 0.6023, + "step": 16516 + }, + { + "epoch": 0.73130181641079, + "grad_norm": 1.3840252161026, + "learning_rate": 4.086695798231576e-06, + "loss": 0.4589, + "step": 16517 + }, + { + "epoch": 0.7313460921156039, + "grad_norm": 1.9472421407699585, + "learning_rate": 4.085449571827541e-06, + "loss": 0.5991, + "step": 16518 + }, + { + "epoch": 0.7313903678204178, + "grad_norm": 1.5474187135696411, + "learning_rate": 4.084203486687732e-06, + "loss": 0.5866, + "step": 16519 + }, + { + "epoch": 0.7314346435252316, + "grad_norm": 1.517723560333252, + "learning_rate": 4.082957542841907e-06, + "loss": 0.5214, + "step": 16520 + }, + { + "epoch": 0.7314789192300455, + "grad_norm": 1.4063833951950073, + "learning_rate": 4.081711740319821e-06, + "loss": 0.5144, + "step": 16521 + }, + { + "epoch": 0.7315231949348594, + "grad_norm": 1.3911073207855225, + "learning_rate": 4.080466079151231e-06, + "loss": 0.4832, + "step": 16522 + }, + { + "epoch": 0.7315674706396732, + "grad_norm": 1.4808149337768555, + "learning_rate": 4.0792205593658915e-06, + "loss": 0.5059, + "step": 16523 + }, + { + "epoch": 0.7316117463444871, + "grad_norm": 1.328918218612671, + "learning_rate": 4.077975180993543e-06, + "loss": 0.493, + "step": 16524 + }, + { + "epoch": 0.731656022049301, + "grad_norm": 1.631773829460144, + "learning_rate": 4.076729944063939e-06, + "loss": 0.5601, + "step": 16525 + }, + { + "epoch": 0.7317002977541148, + "grad_norm": 1.3763357400894165, + "learning_rate": 4.075484848606818e-06, + "loss": 0.4693, + "step": 16526 + }, + { + "epoch": 0.7317445734589287, + "grad_norm": 1.45148766040802, + "learning_rate": 4.0742398946519155e-06, + "loss": 0.5336, + "step": 16527 + }, + { + "epoch": 0.7317888491637426, + "grad_norm": 1.5227389335632324, + "learning_rate": 4.072995082228967e-06, + "loss": 0.5079, + "step": 16528 + }, + { + "epoch": 0.7318331248685565, + "grad_norm": 1.9896332025527954, + "learning_rate": 4.0717504113677075e-06, + "loss": 0.5727, + "step": 16529 + }, + { + "epoch": 0.7318774005733704, + "grad_norm": 1.6682889461517334, + "learning_rate": 4.070505882097861e-06, + "loss": 0.5486, + "step": 16530 + }, + { + "epoch": 0.7319216762781843, + "grad_norm": 1.4792612791061401, + "learning_rate": 4.069261494449156e-06, + "loss": 0.5106, + "step": 16531 + }, + { + "epoch": 0.7319659519829981, + "grad_norm": 1.97136652469635, + "learning_rate": 4.06801724845131e-06, + "loss": 0.5271, + "step": 16532 + }, + { + "epoch": 0.732010227687812, + "grad_norm": 1.389137864112854, + "learning_rate": 4.066773144134043e-06, + "loss": 0.516, + "step": 16533 + }, + { + "epoch": 0.7320545033926259, + "grad_norm": 1.611670732498169, + "learning_rate": 4.065529181527064e-06, + "loss": 0.6151, + "step": 16534 + }, + { + "epoch": 0.7320987790974398, + "grad_norm": 1.4810036420822144, + "learning_rate": 4.0642853606600915e-06, + "loss": 0.5426, + "step": 16535 + }, + { + "epoch": 0.7321430548022536, + "grad_norm": 1.5286766290664673, + "learning_rate": 4.063041681562831e-06, + "loss": 0.5103, + "step": 16536 + }, + { + "epoch": 0.7321873305070675, + "grad_norm": 1.8075287342071533, + "learning_rate": 4.061798144264986e-06, + "loss": 0.5889, + "step": 16537 + }, + { + "epoch": 0.7322316062118814, + "grad_norm": 1.7959105968475342, + "learning_rate": 4.060554748796254e-06, + "loss": 0.5233, + "step": 16538 + }, + { + "epoch": 0.7322758819166952, + "grad_norm": 1.8266640901565552, + "learning_rate": 4.059311495186338e-06, + "loss": 0.5925, + "step": 16539 + }, + { + "epoch": 0.7323201576215091, + "grad_norm": 1.6792018413543701, + "learning_rate": 4.05806838346493e-06, + "loss": 0.5924, + "step": 16540 + }, + { + "epoch": 0.7323644333263231, + "grad_norm": 1.7411291599273682, + "learning_rate": 4.056825413661722e-06, + "loss": 0.5994, + "step": 16541 + }, + { + "epoch": 0.7324087090311369, + "grad_norm": 1.716968297958374, + "learning_rate": 4.055582585806399e-06, + "loss": 0.525, + "step": 16542 + }, + { + "epoch": 0.7324529847359508, + "grad_norm": 1.8471132516860962, + "learning_rate": 4.054339899928646e-06, + "loss": 0.5442, + "step": 16543 + }, + { + "epoch": 0.7324972604407647, + "grad_norm": 1.30415940284729, + "learning_rate": 4.053097356058139e-06, + "loss": 0.5678, + "step": 16544 + }, + { + "epoch": 0.7325415361455785, + "grad_norm": 1.8217893838882446, + "learning_rate": 4.0518549542245645e-06, + "loss": 0.5906, + "step": 16545 + }, + { + "epoch": 0.7325858118503924, + "grad_norm": 1.4432199001312256, + "learning_rate": 4.0506126944575895e-06, + "loss": 0.5189, + "step": 16546 + }, + { + "epoch": 0.7326300875552063, + "grad_norm": 1.713713526725769, + "learning_rate": 4.049370576786887e-06, + "loss": 0.5601, + "step": 16547 + }, + { + "epoch": 0.7326743632600201, + "grad_norm": 1.4411885738372803, + "learning_rate": 4.04812860124212e-06, + "loss": 0.5074, + "step": 16548 + }, + { + "epoch": 0.732718638964834, + "grad_norm": 1.2632293701171875, + "learning_rate": 4.046886767852959e-06, + "loss": 0.5198, + "step": 16549 + }, + { + "epoch": 0.7327629146696479, + "grad_norm": 1.220388412475586, + "learning_rate": 4.04564507664906e-06, + "loss": 0.4955, + "step": 16550 + }, + { + "epoch": 0.7328071903744617, + "grad_norm": 1.430281400680542, + "learning_rate": 4.044403527660079e-06, + "loss": 0.4889, + "step": 16551 + }, + { + "epoch": 0.7328514660792756, + "grad_norm": 1.5433868169784546, + "learning_rate": 4.043162120915671e-06, + "loss": 0.4734, + "step": 16552 + }, + { + "epoch": 0.7328957417840896, + "grad_norm": 1.7083112001419067, + "learning_rate": 4.041920856445486e-06, + "loss": 0.539, + "step": 16553 + }, + { + "epoch": 0.7329400174889034, + "grad_norm": 1.5845004320144653, + "learning_rate": 4.040679734279166e-06, + "loss": 0.5775, + "step": 16554 + }, + { + "epoch": 0.7329842931937173, + "grad_norm": 1.5353869199752808, + "learning_rate": 4.039438754446362e-06, + "loss": 0.5232, + "step": 16555 + }, + { + "epoch": 0.7330285688985312, + "grad_norm": 1.6130650043487549, + "learning_rate": 4.0381979169767105e-06, + "loss": 0.4948, + "step": 16556 + }, + { + "epoch": 0.733072844603345, + "grad_norm": 1.5972436666488647, + "learning_rate": 4.036957221899847e-06, + "loss": 0.5154, + "step": 16557 + }, + { + "epoch": 0.7331171203081589, + "grad_norm": 1.4599310159683228, + "learning_rate": 4.035716669245402e-06, + "loss": 0.4975, + "step": 16558 + }, + { + "epoch": 0.7331613960129728, + "grad_norm": 1.3452363014221191, + "learning_rate": 4.034476259043012e-06, + "loss": 0.4982, + "step": 16559 + }, + { + "epoch": 0.7332056717177866, + "grad_norm": 1.2839272022247314, + "learning_rate": 4.0332359913223015e-06, + "loss": 0.4271, + "step": 16560 + }, + { + "epoch": 0.7332499474226005, + "grad_norm": 1.5671027898788452, + "learning_rate": 4.031995866112885e-06, + "loss": 0.4763, + "step": 16561 + }, + { + "epoch": 0.7332942231274144, + "grad_norm": 1.513161540031433, + "learning_rate": 4.0307558834443905e-06, + "loss": 0.5281, + "step": 16562 + }, + { + "epoch": 0.7333384988322282, + "grad_norm": 1.5473471879959106, + "learning_rate": 4.029516043346432e-06, + "loss": 0.5189, + "step": 16563 + }, + { + "epoch": 0.7333827745370421, + "grad_norm": 1.6107832193374634, + "learning_rate": 4.028276345848617e-06, + "loss": 0.5568, + "step": 16564 + }, + { + "epoch": 0.7334270502418561, + "grad_norm": 1.598986029624939, + "learning_rate": 4.027036790980564e-06, + "loss": 0.5624, + "step": 16565 + }, + { + "epoch": 0.7334713259466699, + "grad_norm": 1.6866464614868164, + "learning_rate": 4.025797378771873e-06, + "loss": 0.5903, + "step": 16566 + }, + { + "epoch": 0.7335156016514838, + "grad_norm": 1.6667306423187256, + "learning_rate": 4.024558109252148e-06, + "loss": 0.5783, + "step": 16567 + }, + { + "epoch": 0.7335598773562977, + "grad_norm": 1.4167112112045288, + "learning_rate": 4.023318982450983e-06, + "loss": 0.5014, + "step": 16568 + }, + { + "epoch": 0.7336041530611116, + "grad_norm": 1.5562509298324585, + "learning_rate": 4.022079998397985e-06, + "loss": 0.5084, + "step": 16569 + }, + { + "epoch": 0.7336484287659254, + "grad_norm": 1.5513980388641357, + "learning_rate": 4.020841157122735e-06, + "loss": 0.5514, + "step": 16570 + }, + { + "epoch": 0.7336927044707393, + "grad_norm": 1.7374075651168823, + "learning_rate": 4.019602458654822e-06, + "loss": 0.5903, + "step": 16571 + }, + { + "epoch": 0.7337369801755532, + "grad_norm": 1.9472582340240479, + "learning_rate": 4.0183639030238384e-06, + "loss": 0.5395, + "step": 16572 + }, + { + "epoch": 0.733781255880367, + "grad_norm": 1.698417067527771, + "learning_rate": 4.017125490259364e-06, + "loss": 0.5394, + "step": 16573 + }, + { + "epoch": 0.7338255315851809, + "grad_norm": 1.5178390741348267, + "learning_rate": 4.015887220390971e-06, + "loss": 0.5401, + "step": 16574 + }, + { + "epoch": 0.7338698072899948, + "grad_norm": 1.5390182733535767, + "learning_rate": 4.014649093448242e-06, + "loss": 0.4908, + "step": 16575 + }, + { + "epoch": 0.7339140829948086, + "grad_norm": 1.4730267524719238, + "learning_rate": 4.013411109460748e-06, + "loss": 0.542, + "step": 16576 + }, + { + "epoch": 0.7339583586996226, + "grad_norm": 1.4436720609664917, + "learning_rate": 4.012173268458053e-06, + "loss": 0.5069, + "step": 16577 + }, + { + "epoch": 0.7340026344044365, + "grad_norm": 1.5526543855667114, + "learning_rate": 4.010935570469725e-06, + "loss": 0.5146, + "step": 16578 + }, + { + "epoch": 0.7340469101092503, + "grad_norm": 1.377036690711975, + "learning_rate": 4.009698015525325e-06, + "loss": 0.4555, + "step": 16579 + }, + { + "epoch": 0.7340911858140642, + "grad_norm": 1.483561635017395, + "learning_rate": 4.008460603654409e-06, + "loss": 0.5399, + "step": 16580 + }, + { + "epoch": 0.7341354615188781, + "grad_norm": 1.9401124715805054, + "learning_rate": 4.007223334886531e-06, + "loss": 0.5635, + "step": 16581 + }, + { + "epoch": 0.7341797372236919, + "grad_norm": 1.506343960762024, + "learning_rate": 4.005986209251247e-06, + "loss": 0.4777, + "step": 16582 + }, + { + "epoch": 0.7342240129285058, + "grad_norm": 1.2485336065292358, + "learning_rate": 4.004749226778102e-06, + "loss": 0.4846, + "step": 16583 + }, + { + "epoch": 0.7342682886333197, + "grad_norm": 1.5946375131607056, + "learning_rate": 4.003512387496637e-06, + "loss": 0.5158, + "step": 16584 + }, + { + "epoch": 0.7343125643381335, + "grad_norm": 1.2392903566360474, + "learning_rate": 4.002275691436399e-06, + "loss": 0.4668, + "step": 16585 + }, + { + "epoch": 0.7343568400429474, + "grad_norm": 1.2751238346099854, + "learning_rate": 4.001039138626922e-06, + "loss": 0.4525, + "step": 16586 + }, + { + "epoch": 0.7344011157477613, + "grad_norm": 1.2997100353240967, + "learning_rate": 3.999802729097743e-06, + "loss": 0.4544, + "step": 16587 + }, + { + "epoch": 0.7344453914525751, + "grad_norm": 1.579686164855957, + "learning_rate": 3.998566462878388e-06, + "loss": 0.5182, + "step": 16588 + }, + { + "epoch": 0.7344896671573891, + "grad_norm": 1.6318718194961548, + "learning_rate": 3.997330339998387e-06, + "loss": 0.5645, + "step": 16589 + }, + { + "epoch": 0.734533942862203, + "grad_norm": 1.803106665611267, + "learning_rate": 3.996094360487264e-06, + "loss": 0.5332, + "step": 16590 + }, + { + "epoch": 0.7345782185670168, + "grad_norm": 1.1979832649230957, + "learning_rate": 3.994858524374534e-06, + "loss": 0.4777, + "step": 16591 + }, + { + "epoch": 0.7346224942718307, + "grad_norm": 1.2074947357177734, + "learning_rate": 3.993622831689723e-06, + "loss": 0.4924, + "step": 16592 + }, + { + "epoch": 0.7346667699766446, + "grad_norm": 1.4873828887939453, + "learning_rate": 3.99238728246234e-06, + "loss": 0.5374, + "step": 16593 + }, + { + "epoch": 0.7347110456814584, + "grad_norm": 1.4551658630371094, + "learning_rate": 3.99115187672189e-06, + "loss": 0.5155, + "step": 16594 + }, + { + "epoch": 0.7347553213862723, + "grad_norm": 1.4553711414337158, + "learning_rate": 3.989916614497891e-06, + "loss": 0.5741, + "step": 16595 + }, + { + "epoch": 0.7347995970910862, + "grad_norm": 1.6000391244888306, + "learning_rate": 3.988681495819838e-06, + "loss": 0.5124, + "step": 16596 + }, + { + "epoch": 0.7348438727959, + "grad_norm": 1.3414320945739746, + "learning_rate": 3.987446520717234e-06, + "loss": 0.4429, + "step": 16597 + }, + { + "epoch": 0.7348881485007139, + "grad_norm": 1.6985613107681274, + "learning_rate": 3.986211689219574e-06, + "loss": 0.5453, + "step": 16598 + }, + { + "epoch": 0.7349324242055278, + "grad_norm": 1.4187037944793701, + "learning_rate": 3.984977001356352e-06, + "loss": 0.5263, + "step": 16599 + }, + { + "epoch": 0.7349766999103416, + "grad_norm": 1.5400583744049072, + "learning_rate": 3.983742457157056e-06, + "loss": 0.5482, + "step": 16600 + }, + { + "epoch": 0.7350209756151556, + "grad_norm": 1.8544237613677979, + "learning_rate": 3.9825080566511684e-06, + "loss": 0.5699, + "step": 16601 + }, + { + "epoch": 0.7350652513199695, + "grad_norm": 1.572273850440979, + "learning_rate": 3.981273799868181e-06, + "loss": 0.5485, + "step": 16602 + }, + { + "epoch": 0.7351095270247833, + "grad_norm": 1.3501622676849365, + "learning_rate": 3.9800396868375675e-06, + "loss": 0.5302, + "step": 16603 + }, + { + "epoch": 0.7351538027295972, + "grad_norm": 1.4032478332519531, + "learning_rate": 3.978805717588802e-06, + "loss": 0.5435, + "step": 16604 + }, + { + "epoch": 0.7351980784344111, + "grad_norm": 1.575217366218567, + "learning_rate": 3.977571892151361e-06, + "loss": 0.4943, + "step": 16605 + }, + { + "epoch": 0.735242354139225, + "grad_norm": 1.31903874874115, + "learning_rate": 3.976338210554713e-06, + "loss": 0.464, + "step": 16606 + }, + { + "epoch": 0.7352866298440388, + "grad_norm": 1.4958555698394775, + "learning_rate": 3.9751046728283214e-06, + "loss": 0.5533, + "step": 16607 + }, + { + "epoch": 0.7353309055488527, + "grad_norm": 1.364899754524231, + "learning_rate": 3.97387127900165e-06, + "loss": 0.4944, + "step": 16608 + }, + { + "epoch": 0.7353751812536666, + "grad_norm": 1.551596999168396, + "learning_rate": 3.972638029104157e-06, + "loss": 0.5231, + "step": 16609 + }, + { + "epoch": 0.7354194569584804, + "grad_norm": 1.2996670007705688, + "learning_rate": 3.971404923165295e-06, + "loss": 0.4308, + "step": 16610 + }, + { + "epoch": 0.7354637326632943, + "grad_norm": 1.3696138858795166, + "learning_rate": 3.970171961214515e-06, + "loss": 0.4903, + "step": 16611 + }, + { + "epoch": 0.7355080083681083, + "grad_norm": 1.6645463705062866, + "learning_rate": 3.968939143281272e-06, + "loss": 0.5489, + "step": 16612 + }, + { + "epoch": 0.7355522840729221, + "grad_norm": 1.4823551177978516, + "learning_rate": 3.967706469395008e-06, + "loss": 0.4999, + "step": 16613 + }, + { + "epoch": 0.735596559777736, + "grad_norm": 1.3941279649734497, + "learning_rate": 3.966473939585157e-06, + "loss": 0.4676, + "step": 16614 + }, + { + "epoch": 0.7356408354825499, + "grad_norm": 2.081350564956665, + "learning_rate": 3.965241553881168e-06, + "loss": 0.5404, + "step": 16615 + }, + { + "epoch": 0.7356851111873637, + "grad_norm": 1.372162938117981, + "learning_rate": 3.964009312312476e-06, + "loss": 0.476, + "step": 16616 + }, + { + "epoch": 0.7357293868921776, + "grad_norm": 1.4556585550308228, + "learning_rate": 3.962777214908502e-06, + "loss": 0.5081, + "step": 16617 + }, + { + "epoch": 0.7357736625969915, + "grad_norm": 1.7945225238800049, + "learning_rate": 3.961545261698675e-06, + "loss": 0.5833, + "step": 16618 + }, + { + "epoch": 0.7358179383018053, + "grad_norm": 1.7313226461410522, + "learning_rate": 3.960313452712425e-06, + "loss": 0.5488, + "step": 16619 + }, + { + "epoch": 0.7358622140066192, + "grad_norm": 1.3096128702163696, + "learning_rate": 3.959081787979172e-06, + "loss": 0.4663, + "step": 16620 + }, + { + "epoch": 0.7359064897114331, + "grad_norm": 2.1333260536193848, + "learning_rate": 3.957850267528328e-06, + "loss": 0.6374, + "step": 16621 + }, + { + "epoch": 0.7359507654162469, + "grad_norm": 1.4822237491607666, + "learning_rate": 3.9566188913893145e-06, + "loss": 0.5363, + "step": 16622 + }, + { + "epoch": 0.7359950411210608, + "grad_norm": 1.2925167083740234, + "learning_rate": 3.955387659591538e-06, + "loss": 0.5272, + "step": 16623 + }, + { + "epoch": 0.7360393168258748, + "grad_norm": 1.426144003868103, + "learning_rate": 3.954156572164405e-06, + "loss": 0.5113, + "step": 16624 + }, + { + "epoch": 0.7360835925306886, + "grad_norm": 1.4979861974716187, + "learning_rate": 3.95292562913732e-06, + "loss": 0.5263, + "step": 16625 + }, + { + "epoch": 0.7361278682355025, + "grad_norm": 1.6653931140899658, + "learning_rate": 3.951694830539683e-06, + "loss": 0.5749, + "step": 16626 + }, + { + "epoch": 0.7361721439403164, + "grad_norm": 1.4735949039459229, + "learning_rate": 3.950464176400889e-06, + "loss": 0.5652, + "step": 16627 + }, + { + "epoch": 0.7362164196451302, + "grad_norm": 1.5217427015304565, + "learning_rate": 3.949233666750329e-06, + "loss": 0.5606, + "step": 16628 + }, + { + "epoch": 0.7362606953499441, + "grad_norm": 1.558170199394226, + "learning_rate": 3.948003301617402e-06, + "loss": 0.5366, + "step": 16629 + }, + { + "epoch": 0.736304971054758, + "grad_norm": 1.4854387044906616, + "learning_rate": 3.946773081031486e-06, + "loss": 0.4831, + "step": 16630 + }, + { + "epoch": 0.7363492467595718, + "grad_norm": 1.463730812072754, + "learning_rate": 3.945543005021963e-06, + "loss": 0.5206, + "step": 16631 + }, + { + "epoch": 0.7363935224643857, + "grad_norm": 1.5056568384170532, + "learning_rate": 3.94431307361822e-06, + "loss": 0.4914, + "step": 16632 + }, + { + "epoch": 0.7364377981691996, + "grad_norm": 1.7987587451934814, + "learning_rate": 3.943083286849628e-06, + "loss": 0.6025, + "step": 16633 + }, + { + "epoch": 0.7364820738740134, + "grad_norm": 1.5241249799728394, + "learning_rate": 3.94185364474556e-06, + "loss": 0.5234, + "step": 16634 + }, + { + "epoch": 0.7365263495788273, + "grad_norm": 1.447555661201477, + "learning_rate": 3.940624147335386e-06, + "loss": 0.4973, + "step": 16635 + }, + { + "epoch": 0.7365706252836413, + "grad_norm": 1.599073886871338, + "learning_rate": 3.939394794648469e-06, + "loss": 0.5658, + "step": 16636 + }, + { + "epoch": 0.7366149009884551, + "grad_norm": 1.4281727075576782, + "learning_rate": 3.9381655867141734e-06, + "loss": 0.5362, + "step": 16637 + }, + { + "epoch": 0.736659176693269, + "grad_norm": 1.6571775674819946, + "learning_rate": 3.936936523561853e-06, + "loss": 0.533, + "step": 16638 + }, + { + "epoch": 0.7367034523980829, + "grad_norm": 1.4415220022201538, + "learning_rate": 3.935707605220871e-06, + "loss": 0.4532, + "step": 16639 + }, + { + "epoch": 0.7367477281028968, + "grad_norm": 1.8802534341812134, + "learning_rate": 3.934478831720575e-06, + "loss": 0.5651, + "step": 16640 + }, + { + "epoch": 0.7367920038077106, + "grad_norm": 1.5741833448410034, + "learning_rate": 3.93325020309031e-06, + "loss": 0.5704, + "step": 16641 + }, + { + "epoch": 0.7368362795125245, + "grad_norm": 1.5505948066711426, + "learning_rate": 3.932021719359427e-06, + "loss": 0.5465, + "step": 16642 + }, + { + "epoch": 0.7368805552173384, + "grad_norm": 1.4352073669433594, + "learning_rate": 3.930793380557265e-06, + "loss": 0.5223, + "step": 16643 + }, + { + "epoch": 0.7369248309221522, + "grad_norm": 1.5611608028411865, + "learning_rate": 3.929565186713161e-06, + "loss": 0.5292, + "step": 16644 + }, + { + "epoch": 0.7369691066269661, + "grad_norm": 1.8206881284713745, + "learning_rate": 3.928337137856451e-06, + "loss": 0.6057, + "step": 16645 + }, + { + "epoch": 0.73701338233178, + "grad_norm": 1.5326282978057861, + "learning_rate": 3.927109234016464e-06, + "loss": 0.5744, + "step": 16646 + }, + { + "epoch": 0.7370576580365938, + "grad_norm": 1.4949949979782104, + "learning_rate": 3.925881475222528e-06, + "loss": 0.507, + "step": 16647 + }, + { + "epoch": 0.7371019337414078, + "grad_norm": 1.3623242378234863, + "learning_rate": 3.924653861503963e-06, + "loss": 0.4796, + "step": 16648 + }, + { + "epoch": 0.7371462094462217, + "grad_norm": 1.666011929512024, + "learning_rate": 3.923426392890098e-06, + "loss": 0.6045, + "step": 16649 + }, + { + "epoch": 0.7371904851510355, + "grad_norm": 2.6252453327178955, + "learning_rate": 3.922199069410246e-06, + "loss": 0.5946, + "step": 16650 + }, + { + "epoch": 0.7372347608558494, + "grad_norm": 1.4454306364059448, + "learning_rate": 3.9209718910937174e-06, + "loss": 0.4568, + "step": 16651 + }, + { + "epoch": 0.7372790365606633, + "grad_norm": 1.2341464757919312, + "learning_rate": 3.9197448579698285e-06, + "loss": 0.5088, + "step": 16652 + }, + { + "epoch": 0.7373233122654771, + "grad_norm": 1.3630584478378296, + "learning_rate": 3.918517970067883e-06, + "loss": 0.539, + "step": 16653 + }, + { + "epoch": 0.737367587970291, + "grad_norm": 1.9579066038131714, + "learning_rate": 3.917291227417185e-06, + "loss": 0.5991, + "step": 16654 + }, + { + "epoch": 0.7374118636751049, + "grad_norm": 1.5008131265640259, + "learning_rate": 3.916064630047032e-06, + "loss": 0.4817, + "step": 16655 + }, + { + "epoch": 0.7374561393799187, + "grad_norm": 1.6710731983184814, + "learning_rate": 3.9148381779867215e-06, + "loss": 0.5317, + "step": 16656 + }, + { + "epoch": 0.7375004150847326, + "grad_norm": 1.4508264064788818, + "learning_rate": 3.913611871265547e-06, + "loss": 0.5065, + "step": 16657 + }, + { + "epoch": 0.7375446907895465, + "grad_norm": 1.4959980249404907, + "learning_rate": 3.912385709912794e-06, + "loss": 0.5125, + "step": 16658 + }, + { + "epoch": 0.7375889664943603, + "grad_norm": 1.4247677326202393, + "learning_rate": 3.911159693957755e-06, + "loss": 0.5299, + "step": 16659 + }, + { + "epoch": 0.7376332421991743, + "grad_norm": 1.4333384037017822, + "learning_rate": 3.909933823429708e-06, + "loss": 0.5152, + "step": 16660 + }, + { + "epoch": 0.7376775179039882, + "grad_norm": 1.6358097791671753, + "learning_rate": 3.90870809835793e-06, + "loss": 0.4569, + "step": 16661 + }, + { + "epoch": 0.737721793608802, + "grad_norm": 1.955910563468933, + "learning_rate": 3.907482518771703e-06, + "loss": 0.5465, + "step": 16662 + }, + { + "epoch": 0.7377660693136159, + "grad_norm": 1.8145381212234497, + "learning_rate": 3.906257084700295e-06, + "loss": 0.6208, + "step": 16663 + }, + { + "epoch": 0.7378103450184298, + "grad_norm": 1.5232445001602173, + "learning_rate": 3.905031796172979e-06, + "loss": 0.5379, + "step": 16664 + }, + { + "epoch": 0.7378546207232436, + "grad_norm": 2.1491150856018066, + "learning_rate": 3.903806653219008e-06, + "loss": 0.6052, + "step": 16665 + }, + { + "epoch": 0.7378988964280575, + "grad_norm": 1.5176435708999634, + "learning_rate": 3.902581655867655e-06, + "loss": 0.5623, + "step": 16666 + }, + { + "epoch": 0.7379431721328714, + "grad_norm": 1.4027775526046753, + "learning_rate": 3.901356804148175e-06, + "loss": 0.5673, + "step": 16667 + }, + { + "epoch": 0.7379874478376852, + "grad_norm": 1.4576785564422607, + "learning_rate": 3.900132098089818e-06, + "loss": 0.5603, + "step": 16668 + }, + { + "epoch": 0.7380317235424991, + "grad_norm": 1.3440027236938477, + "learning_rate": 3.898907537721843e-06, + "loss": 0.4988, + "step": 16669 + }, + { + "epoch": 0.738075999247313, + "grad_norm": 1.3232777118682861, + "learning_rate": 3.8976831230734935e-06, + "loss": 0.4765, + "step": 16670 + }, + { + "epoch": 0.7381202749521268, + "grad_norm": 1.8155863285064697, + "learning_rate": 3.896458854174011e-06, + "loss": 0.5477, + "step": 16671 + }, + { + "epoch": 0.7381645506569408, + "grad_norm": 1.4311505556106567, + "learning_rate": 3.895234731052645e-06, + "loss": 0.5259, + "step": 16672 + }, + { + "epoch": 0.7382088263617547, + "grad_norm": 1.8920269012451172, + "learning_rate": 3.8940107537386244e-06, + "loss": 0.6063, + "step": 16673 + }, + { + "epoch": 0.7382531020665685, + "grad_norm": 1.4192242622375488, + "learning_rate": 3.892786922261184e-06, + "loss": 0.5226, + "step": 16674 + }, + { + "epoch": 0.7382973777713824, + "grad_norm": 1.8701971769332886, + "learning_rate": 3.891563236649553e-06, + "loss": 0.572, + "step": 16675 + }, + { + "epoch": 0.7383416534761963, + "grad_norm": 1.7019822597503662, + "learning_rate": 3.890339696932964e-06, + "loss": 0.5333, + "step": 16676 + }, + { + "epoch": 0.7383859291810102, + "grad_norm": 1.3645614385604858, + "learning_rate": 3.889116303140636e-06, + "loss": 0.4921, + "step": 16677 + }, + { + "epoch": 0.738430204885824, + "grad_norm": 1.285900354385376, + "learning_rate": 3.8878930553017855e-06, + "loss": 0.4622, + "step": 16678 + }, + { + "epoch": 0.7384744805906379, + "grad_norm": 1.422856330871582, + "learning_rate": 3.886669953445638e-06, + "loss": 0.5104, + "step": 16679 + }, + { + "epoch": 0.7385187562954518, + "grad_norm": 1.680675745010376, + "learning_rate": 3.8854469976013995e-06, + "loss": 0.5411, + "step": 16680 + }, + { + "epoch": 0.7385630320002656, + "grad_norm": 1.969744324684143, + "learning_rate": 3.884224187798281e-06, + "loss": 0.5964, + "step": 16681 + }, + { + "epoch": 0.7386073077050795, + "grad_norm": 1.5217978954315186, + "learning_rate": 3.883001524065488e-06, + "loss": 0.5644, + "step": 16682 + }, + { + "epoch": 0.7386515834098935, + "grad_norm": 1.7560484409332275, + "learning_rate": 3.881779006432223e-06, + "loss": 0.5675, + "step": 16683 + }, + { + "epoch": 0.7386958591147073, + "grad_norm": 2.422999620437622, + "learning_rate": 3.8805566349276855e-06, + "loss": 0.6778, + "step": 16684 + }, + { + "epoch": 0.7387401348195212, + "grad_norm": 1.4422532320022583, + "learning_rate": 3.8793344095810655e-06, + "loss": 0.5161, + "step": 16685 + }, + { + "epoch": 0.7387844105243351, + "grad_norm": 2.1793649196624756, + "learning_rate": 3.878112330421564e-06, + "loss": 0.562, + "step": 16686 + }, + { + "epoch": 0.7388286862291489, + "grad_norm": 1.5268211364746094, + "learning_rate": 3.876890397478364e-06, + "loss": 0.5129, + "step": 16687 + }, + { + "epoch": 0.7388729619339628, + "grad_norm": 1.618698000907898, + "learning_rate": 3.875668610780649e-06, + "loss": 0.5261, + "step": 16688 + }, + { + "epoch": 0.7389172376387767, + "grad_norm": 1.612325668334961, + "learning_rate": 3.874446970357605e-06, + "loss": 0.5629, + "step": 16689 + }, + { + "epoch": 0.7389615133435905, + "grad_norm": 1.447631597518921, + "learning_rate": 3.873225476238408e-06, + "loss": 0.5031, + "step": 16690 + }, + { + "epoch": 0.7390057890484044, + "grad_norm": 1.5527609586715698, + "learning_rate": 3.872004128452231e-06, + "loss": 0.5185, + "step": 16691 + }, + { + "epoch": 0.7390500647532183, + "grad_norm": 1.9940482378005981, + "learning_rate": 3.870782927028247e-06, + "loss": 0.6332, + "step": 16692 + }, + { + "epoch": 0.7390943404580321, + "grad_norm": 1.4551963806152344, + "learning_rate": 3.869561871995622e-06, + "loss": 0.4798, + "step": 16693 + }, + { + "epoch": 0.739138616162846, + "grad_norm": 1.7348233461380005, + "learning_rate": 3.86834096338352e-06, + "loss": 0.4981, + "step": 16694 + }, + { + "epoch": 0.73918289186766, + "grad_norm": 1.4786239862442017, + "learning_rate": 3.867120201221097e-06, + "loss": 0.5063, + "step": 16695 + }, + { + "epoch": 0.7392271675724738, + "grad_norm": 1.9580599069595337, + "learning_rate": 3.865899585537519e-06, + "loss": 0.5897, + "step": 16696 + }, + { + "epoch": 0.7392714432772877, + "grad_norm": 1.631978988647461, + "learning_rate": 3.864679116361936e-06, + "loss": 0.4685, + "step": 16697 + }, + { + "epoch": 0.7393157189821016, + "grad_norm": 1.609861969947815, + "learning_rate": 3.863458793723491e-06, + "loss": 0.5407, + "step": 16698 + }, + { + "epoch": 0.7393599946869154, + "grad_norm": 1.7013437747955322, + "learning_rate": 3.86223861765134e-06, + "loss": 0.5321, + "step": 16699 + }, + { + "epoch": 0.7394042703917293, + "grad_norm": 1.5706806182861328, + "learning_rate": 3.8610185881746235e-06, + "loss": 0.5474, + "step": 16700 + }, + { + "epoch": 0.7394485460965432, + "grad_norm": 1.6033540964126587, + "learning_rate": 3.859798705322479e-06, + "loss": 0.508, + "step": 16701 + }, + { + "epoch": 0.739492821801357, + "grad_norm": 1.7725365161895752, + "learning_rate": 3.858578969124041e-06, + "loss": 0.5632, + "step": 16702 + }, + { + "epoch": 0.7395370975061709, + "grad_norm": 1.7417446374893188, + "learning_rate": 3.857359379608446e-06, + "loss": 0.5916, + "step": 16703 + }, + { + "epoch": 0.7395813732109848, + "grad_norm": 1.4959274530410767, + "learning_rate": 3.856139936804819e-06, + "loss": 0.5253, + "step": 16704 + }, + { + "epoch": 0.7396256489157986, + "grad_norm": 1.4815179109573364, + "learning_rate": 3.854920640742284e-06, + "loss": 0.512, + "step": 16705 + }, + { + "epoch": 0.7396699246206125, + "grad_norm": 1.3949823379516602, + "learning_rate": 3.8537014914499715e-06, + "loss": 0.5124, + "step": 16706 + }, + { + "epoch": 0.7397142003254265, + "grad_norm": 1.3756028413772583, + "learning_rate": 3.852482488956992e-06, + "loss": 0.5241, + "step": 16707 + }, + { + "epoch": 0.7397584760302403, + "grad_norm": 1.402396321296692, + "learning_rate": 3.8512636332924606e-06, + "loss": 0.49, + "step": 16708 + }, + { + "epoch": 0.7398027517350542, + "grad_norm": 1.6024301052093506, + "learning_rate": 3.850044924485494e-06, + "loss": 0.5339, + "step": 16709 + }, + { + "epoch": 0.7398470274398681, + "grad_norm": 1.4377022981643677, + "learning_rate": 3.848826362565196e-06, + "loss": 0.5403, + "step": 16710 + }, + { + "epoch": 0.739891303144682, + "grad_norm": 1.5659689903259277, + "learning_rate": 3.8476079475606735e-06, + "loss": 0.5991, + "step": 16711 + }, + { + "epoch": 0.7399355788494958, + "grad_norm": 1.4184824228286743, + "learning_rate": 3.846389679501025e-06, + "loss": 0.5316, + "step": 16712 + }, + { + "epoch": 0.7399798545543097, + "grad_norm": 1.3871986865997314, + "learning_rate": 3.845171558415348e-06, + "loss": 0.4691, + "step": 16713 + }, + { + "epoch": 0.7400241302591236, + "grad_norm": 1.4779025316238403, + "learning_rate": 3.843953584332737e-06, + "loss": 0.5047, + "step": 16714 + }, + { + "epoch": 0.7400684059639374, + "grad_norm": 1.6433351039886475, + "learning_rate": 3.8427357572822785e-06, + "loss": 0.5488, + "step": 16715 + }, + { + "epoch": 0.7401126816687513, + "grad_norm": 1.6549144983291626, + "learning_rate": 3.841518077293067e-06, + "loss": 0.5452, + "step": 16716 + }, + { + "epoch": 0.7401569573735652, + "grad_norm": 1.5500739812850952, + "learning_rate": 3.840300544394182e-06, + "loss": 0.5473, + "step": 16717 + }, + { + "epoch": 0.740201233078379, + "grad_norm": 1.4877879619598389, + "learning_rate": 3.839083158614699e-06, + "loss": 0.5195, + "step": 16718 + }, + { + "epoch": 0.740245508783193, + "grad_norm": 2.356260299682617, + "learning_rate": 3.837865919983707e-06, + "loss": 0.5869, + "step": 16719 + }, + { + "epoch": 0.7402897844880069, + "grad_norm": 2.046285390853882, + "learning_rate": 3.836648828530264e-06, + "loss": 0.5662, + "step": 16720 + }, + { + "epoch": 0.7403340601928207, + "grad_norm": 1.5373049974441528, + "learning_rate": 3.835431884283444e-06, + "loss": 0.5089, + "step": 16721 + }, + { + "epoch": 0.7403783358976346, + "grad_norm": 1.7852452993392944, + "learning_rate": 3.834215087272316e-06, + "loss": 0.5471, + "step": 16722 + }, + { + "epoch": 0.7404226116024485, + "grad_norm": 1.4532506465911865, + "learning_rate": 3.832998437525941e-06, + "loss": 0.5521, + "step": 16723 + }, + { + "epoch": 0.7404668873072623, + "grad_norm": 1.9619026184082031, + "learning_rate": 3.831781935073377e-06, + "loss": 0.5771, + "step": 16724 + }, + { + "epoch": 0.7405111630120762, + "grad_norm": 1.510249376296997, + "learning_rate": 3.830565579943676e-06, + "loss": 0.5062, + "step": 16725 + }, + { + "epoch": 0.7405554387168901, + "grad_norm": 1.3066610097885132, + "learning_rate": 3.8293493721658955e-06, + "loss": 0.4862, + "step": 16726 + }, + { + "epoch": 0.7405997144217039, + "grad_norm": 1.4277522563934326, + "learning_rate": 3.828133311769081e-06, + "loss": 0.5119, + "step": 16727 + }, + { + "epoch": 0.7406439901265178, + "grad_norm": 1.5757490396499634, + "learning_rate": 3.826917398782277e-06, + "loss": 0.6059, + "step": 16728 + }, + { + "epoch": 0.7406882658313317, + "grad_norm": 1.8811177015304565, + "learning_rate": 3.825701633234526e-06, + "loss": 0.566, + "step": 16729 + }, + { + "epoch": 0.7407325415361455, + "grad_norm": 2.0504794120788574, + "learning_rate": 3.824486015154862e-06, + "loss": 0.5883, + "step": 16730 + }, + { + "epoch": 0.7407768172409595, + "grad_norm": 1.837198257446289, + "learning_rate": 3.823270544572319e-06, + "loss": 0.6286, + "step": 16731 + }, + { + "epoch": 0.7408210929457734, + "grad_norm": 1.2812302112579346, + "learning_rate": 3.822055221515933e-06, + "loss": 0.5107, + "step": 16732 + }, + { + "epoch": 0.7408653686505872, + "grad_norm": 1.7792609930038452, + "learning_rate": 3.820840046014729e-06, + "loss": 0.6039, + "step": 16733 + }, + { + "epoch": 0.7409096443554011, + "grad_norm": 1.6115988492965698, + "learning_rate": 3.819625018097728e-06, + "loss": 0.5895, + "step": 16734 + }, + { + "epoch": 0.740953920060215, + "grad_norm": 1.369828224182129, + "learning_rate": 3.818410137793947e-06, + "loss": 0.4794, + "step": 16735 + }, + { + "epoch": 0.7409981957650288, + "grad_norm": 1.516438603401184, + "learning_rate": 3.817195405132411e-06, + "loss": 0.5315, + "step": 16736 + }, + { + "epoch": 0.7410424714698427, + "grad_norm": 1.4649477005004883, + "learning_rate": 3.815980820142129e-06, + "loss": 0.5138, + "step": 16737 + }, + { + "epoch": 0.7410867471746566, + "grad_norm": 1.678937554359436, + "learning_rate": 3.814766382852109e-06, + "loss": 0.5625, + "step": 16738 + }, + { + "epoch": 0.7411310228794704, + "grad_norm": 1.7794277667999268, + "learning_rate": 3.813552093291358e-06, + "loss": 0.5736, + "step": 16739 + }, + { + "epoch": 0.7411752985842843, + "grad_norm": 1.3691140413284302, + "learning_rate": 3.8123379514888782e-06, + "loss": 0.4979, + "step": 16740 + }, + { + "epoch": 0.7412195742890982, + "grad_norm": 1.3847415447235107, + "learning_rate": 3.8111239574736682e-06, + "loss": 0.5365, + "step": 16741 + }, + { + "epoch": 0.741263849993912, + "grad_norm": 1.4779770374298096, + "learning_rate": 3.8099101112747196e-06, + "loss": 0.5284, + "step": 16742 + }, + { + "epoch": 0.741308125698726, + "grad_norm": 1.8114253282546997, + "learning_rate": 3.808696412921031e-06, + "loss": 0.6354, + "step": 16743 + }, + { + "epoch": 0.7413524014035399, + "grad_norm": 1.285358190536499, + "learning_rate": 3.807482862441588e-06, + "loss": 0.5449, + "step": 16744 + }, + { + "epoch": 0.7413966771083537, + "grad_norm": 1.4200407266616821, + "learning_rate": 3.8062694598653714e-06, + "loss": 0.5204, + "step": 16745 + }, + { + "epoch": 0.7414409528131676, + "grad_norm": 1.3834317922592163, + "learning_rate": 3.805056205221369e-06, + "loss": 0.5094, + "step": 16746 + }, + { + "epoch": 0.7414852285179815, + "grad_norm": 1.9391289949417114, + "learning_rate": 3.8038430985385556e-06, + "loss": 0.5378, + "step": 16747 + }, + { + "epoch": 0.7415295042227954, + "grad_norm": 1.8028801679611206, + "learning_rate": 3.802630139845903e-06, + "loss": 0.5944, + "step": 16748 + }, + { + "epoch": 0.7415737799276092, + "grad_norm": 1.4761404991149902, + "learning_rate": 3.8014173291723846e-06, + "loss": 0.5581, + "step": 16749 + }, + { + "epoch": 0.7416180556324231, + "grad_norm": 1.3451417684555054, + "learning_rate": 3.8002046665469663e-06, + "loss": 0.4975, + "step": 16750 + }, + { + "epoch": 0.741662331337237, + "grad_norm": 1.3022900819778442, + "learning_rate": 3.7989921519986116e-06, + "loss": 0.5062, + "step": 16751 + }, + { + "epoch": 0.7417066070420508, + "grad_norm": 1.4938803911209106, + "learning_rate": 3.797779785556276e-06, + "loss": 0.5205, + "step": 16752 + }, + { + "epoch": 0.7417508827468647, + "grad_norm": 1.4820308685302734, + "learning_rate": 3.7965675672489243e-06, + "loss": 0.487, + "step": 16753 + }, + { + "epoch": 0.7417951584516786, + "grad_norm": 1.8770159482955933, + "learning_rate": 3.795355497105505e-06, + "loss": 0.5409, + "step": 16754 + }, + { + "epoch": 0.7418394341564924, + "grad_norm": 1.5326498746871948, + "learning_rate": 3.794143575154964e-06, + "loss": 0.5539, + "step": 16755 + }, + { + "epoch": 0.7418837098613064, + "grad_norm": 1.5315642356872559, + "learning_rate": 3.792931801426254e-06, + "loss": 0.5886, + "step": 16756 + }, + { + "epoch": 0.7419279855661203, + "grad_norm": 1.4446817636489868, + "learning_rate": 3.7917201759483144e-06, + "loss": 0.5353, + "step": 16757 + }, + { + "epoch": 0.7419722612709341, + "grad_norm": 1.7777910232543945, + "learning_rate": 3.790508698750084e-06, + "loss": 0.4841, + "step": 16758 + }, + { + "epoch": 0.742016536975748, + "grad_norm": 1.592930555343628, + "learning_rate": 3.789297369860496e-06, + "loss": 0.5578, + "step": 16759 + }, + { + "epoch": 0.7420608126805619, + "grad_norm": 1.4775235652923584, + "learning_rate": 3.7880861893084843e-06, + "loss": 0.51, + "step": 16760 + }, + { + "epoch": 0.7421050883853757, + "grad_norm": 1.457054615020752, + "learning_rate": 3.7868751571229757e-06, + "loss": 0.4801, + "step": 16761 + }, + { + "epoch": 0.7421493640901896, + "grad_norm": 1.4221707582473755, + "learning_rate": 3.785664273332892e-06, + "loss": 0.4951, + "step": 16762 + }, + { + "epoch": 0.7421936397950035, + "grad_norm": 1.483550786972046, + "learning_rate": 3.784453537967161e-06, + "loss": 0.5414, + "step": 16763 + }, + { + "epoch": 0.7422379154998173, + "grad_norm": 1.3454188108444214, + "learning_rate": 3.783242951054695e-06, + "loss": 0.4815, + "step": 16764 + }, + { + "epoch": 0.7422821912046312, + "grad_norm": 1.4021248817443848, + "learning_rate": 3.7820325126244074e-06, + "loss": 0.5126, + "step": 16765 + }, + { + "epoch": 0.7423264669094451, + "grad_norm": 2.650160074234009, + "learning_rate": 3.7808222227052127e-06, + "loss": 0.5417, + "step": 16766 + }, + { + "epoch": 0.742370742614259, + "grad_norm": 1.8306539058685303, + "learning_rate": 3.7796120813260195e-06, + "loss": 0.5698, + "step": 16767 + }, + { + "epoch": 0.7424150183190729, + "grad_norm": 1.5216522216796875, + "learning_rate": 3.778402088515719e-06, + "loss": 0.5752, + "step": 16768 + }, + { + "epoch": 0.7424592940238868, + "grad_norm": 1.3315829038619995, + "learning_rate": 3.777192244303223e-06, + "loss": 0.4806, + "step": 16769 + }, + { + "epoch": 0.7425035697287006, + "grad_norm": 1.4889051914215088, + "learning_rate": 3.7759825487174227e-06, + "loss": 0.5215, + "step": 16770 + }, + { + "epoch": 0.7425478454335145, + "grad_norm": 1.3949978351593018, + "learning_rate": 3.774773001787213e-06, + "loss": 0.5374, + "step": 16771 + }, + { + "epoch": 0.7425921211383284, + "grad_norm": 1.1461642980575562, + "learning_rate": 3.7735636035414768e-06, + "loss": 0.4111, + "step": 16772 + }, + { + "epoch": 0.7426363968431422, + "grad_norm": 1.3034796714782715, + "learning_rate": 3.772354354009108e-06, + "loss": 0.4708, + "step": 16773 + }, + { + "epoch": 0.7426806725479561, + "grad_norm": 1.4535871744155884, + "learning_rate": 3.771145253218984e-06, + "loss": 0.5119, + "step": 16774 + }, + { + "epoch": 0.74272494825277, + "grad_norm": 1.4085067510604858, + "learning_rate": 3.7699363011999844e-06, + "loss": 0.4683, + "step": 16775 + }, + { + "epoch": 0.7427692239575838, + "grad_norm": 1.751554250717163, + "learning_rate": 3.768727497980984e-06, + "loss": 0.6029, + "step": 16776 + }, + { + "epoch": 0.7428134996623977, + "grad_norm": 1.421894907951355, + "learning_rate": 3.7675188435908527e-06, + "loss": 0.5294, + "step": 16777 + }, + { + "epoch": 0.7428577753672116, + "grad_norm": 1.9564757347106934, + "learning_rate": 3.7663103380584566e-06, + "loss": 0.5539, + "step": 16778 + }, + { + "epoch": 0.7429020510720254, + "grad_norm": 1.3805092573165894, + "learning_rate": 3.7651019814126656e-06, + "loss": 0.4913, + "step": 16779 + }, + { + "epoch": 0.7429463267768394, + "grad_norm": 1.738919734954834, + "learning_rate": 3.7638937736823365e-06, + "loss": 0.5026, + "step": 16780 + }, + { + "epoch": 0.7429906024816533, + "grad_norm": 1.5643839836120605, + "learning_rate": 3.762685714896328e-06, + "loss": 0.4849, + "step": 16781 + }, + { + "epoch": 0.7430348781864672, + "grad_norm": 1.2859463691711426, + "learning_rate": 3.761477805083489e-06, + "loss": 0.4507, + "step": 16782 + }, + { + "epoch": 0.743079153891281, + "grad_norm": 1.4545598030090332, + "learning_rate": 3.760270044272676e-06, + "loss": 0.4204, + "step": 16783 + }, + { + "epoch": 0.7431234295960949, + "grad_norm": 1.5945539474487305, + "learning_rate": 3.759062432492733e-06, + "loss": 0.5838, + "step": 16784 + }, + { + "epoch": 0.7431677053009088, + "grad_norm": 1.9049122333526611, + "learning_rate": 3.757854969772501e-06, + "loss": 0.5934, + "step": 16785 + }, + { + "epoch": 0.7432119810057226, + "grad_norm": 1.6034244298934937, + "learning_rate": 3.7566476561408216e-06, + "loss": 0.5537, + "step": 16786 + }, + { + "epoch": 0.7432562567105365, + "grad_norm": 1.4589524269104004, + "learning_rate": 3.7554404916265288e-06, + "loss": 0.5043, + "step": 16787 + }, + { + "epoch": 0.7433005324153504, + "grad_norm": 1.5061814785003662, + "learning_rate": 3.754233476258452e-06, + "loss": 0.5046, + "step": 16788 + }, + { + "epoch": 0.7433448081201642, + "grad_norm": 1.3854914903640747, + "learning_rate": 3.7530266100654266e-06, + "loss": 0.5305, + "step": 16789 + }, + { + "epoch": 0.7433890838249781, + "grad_norm": 1.793644666671753, + "learning_rate": 3.751819893076275e-06, + "loss": 0.5701, + "step": 16790 + }, + { + "epoch": 0.7434333595297921, + "grad_norm": 1.381151556968689, + "learning_rate": 3.7506133253198173e-06, + "loss": 0.4538, + "step": 16791 + }, + { + "epoch": 0.7434776352346059, + "grad_norm": 1.4228668212890625, + "learning_rate": 3.7494069068248685e-06, + "loss": 0.5058, + "step": 16792 + }, + { + "epoch": 0.7435219109394198, + "grad_norm": 1.5564284324645996, + "learning_rate": 3.7482006376202484e-06, + "loss": 0.5547, + "step": 16793 + }, + { + "epoch": 0.7435661866442337, + "grad_norm": 1.7290481328964233, + "learning_rate": 3.746994517734768e-06, + "loss": 0.507, + "step": 16794 + }, + { + "epoch": 0.7436104623490475, + "grad_norm": 1.2153109312057495, + "learning_rate": 3.7457885471972298e-06, + "loss": 0.4254, + "step": 16795 + }, + { + "epoch": 0.7436547380538614, + "grad_norm": 1.393408179283142, + "learning_rate": 3.744582726036441e-06, + "loss": 0.5073, + "step": 16796 + }, + { + "epoch": 0.7436990137586753, + "grad_norm": 1.8860868215560913, + "learning_rate": 3.743377054281201e-06, + "loss": 0.528, + "step": 16797 + }, + { + "epoch": 0.7437432894634891, + "grad_norm": 1.9819130897521973, + "learning_rate": 3.7421715319603013e-06, + "loss": 0.5509, + "step": 16798 + }, + { + "epoch": 0.743787565168303, + "grad_norm": 1.5065276622772217, + "learning_rate": 3.7409661591025427e-06, + "loss": 0.5079, + "step": 16799 + }, + { + "epoch": 0.7438318408731169, + "grad_norm": 1.7975884675979614, + "learning_rate": 3.7397609357367116e-06, + "loss": 0.5914, + "step": 16800 + }, + { + "epoch": 0.7438761165779307, + "grad_norm": 1.5833423137664795, + "learning_rate": 3.7385558618915928e-06, + "loss": 0.4774, + "step": 16801 + }, + { + "epoch": 0.7439203922827446, + "grad_norm": 1.7789185047149658, + "learning_rate": 3.737350937595966e-06, + "loss": 0.5627, + "step": 16802 + }, + { + "epoch": 0.7439646679875586, + "grad_norm": 1.6118701696395874, + "learning_rate": 3.7361461628786167e-06, + "loss": 0.5564, + "step": 16803 + }, + { + "epoch": 0.7440089436923724, + "grad_norm": 1.5064951181411743, + "learning_rate": 3.734941537768315e-06, + "loss": 0.4964, + "step": 16804 + }, + { + "epoch": 0.7440532193971863, + "grad_norm": 2.232710838317871, + "learning_rate": 3.733737062293834e-06, + "loss": 0.5743, + "step": 16805 + }, + { + "epoch": 0.7440974951020002, + "grad_norm": 1.6545324325561523, + "learning_rate": 3.7325327364839413e-06, + "loss": 0.5505, + "step": 16806 + }, + { + "epoch": 0.744141770806814, + "grad_norm": 1.4002938270568848, + "learning_rate": 3.731328560367401e-06, + "loss": 0.509, + "step": 16807 + }, + { + "epoch": 0.7441860465116279, + "grad_norm": 1.8094532489776611, + "learning_rate": 3.73012453397297e-06, + "loss": 0.5618, + "step": 16808 + }, + { + "epoch": 0.7442303222164418, + "grad_norm": 1.5660985708236694, + "learning_rate": 3.7289206573294135e-06, + "loss": 0.5538, + "step": 16809 + }, + { + "epoch": 0.7442745979212556, + "grad_norm": 1.3919763565063477, + "learning_rate": 3.7277169304654803e-06, + "loss": 0.4461, + "step": 16810 + }, + { + "epoch": 0.7443188736260695, + "grad_norm": 1.548934817314148, + "learning_rate": 3.7265133534099206e-06, + "loss": 0.5259, + "step": 16811 + }, + { + "epoch": 0.7443631493308834, + "grad_norm": 1.5198240280151367, + "learning_rate": 3.7253099261914794e-06, + "loss": 0.565, + "step": 16812 + }, + { + "epoch": 0.7444074250356972, + "grad_norm": 1.6959348917007446, + "learning_rate": 3.7241066488389044e-06, + "loss": 0.5555, + "step": 16813 + }, + { + "epoch": 0.7444517007405111, + "grad_norm": 1.297698974609375, + "learning_rate": 3.722903521380935e-06, + "loss": 0.49, + "step": 16814 + }, + { + "epoch": 0.7444959764453251, + "grad_norm": 1.904722809791565, + "learning_rate": 3.721700543846297e-06, + "loss": 0.5669, + "step": 16815 + }, + { + "epoch": 0.744540252150139, + "grad_norm": 1.5367155075073242, + "learning_rate": 3.7204977162637347e-06, + "loss": 0.5544, + "step": 16816 + }, + { + "epoch": 0.7445845278549528, + "grad_norm": 1.3020344972610474, + "learning_rate": 3.71929503866197e-06, + "loss": 0.4707, + "step": 16817 + }, + { + "epoch": 0.7446288035597667, + "grad_norm": 1.8095755577087402, + "learning_rate": 3.7180925110697266e-06, + "loss": 0.5916, + "step": 16818 + }, + { + "epoch": 0.7446730792645806, + "grad_norm": 1.7293040752410889, + "learning_rate": 3.7168901335157313e-06, + "loss": 0.5716, + "step": 16819 + }, + { + "epoch": 0.7447173549693944, + "grad_norm": 1.407334566116333, + "learning_rate": 3.7156879060287e-06, + "loss": 0.4356, + "step": 16820 + }, + { + "epoch": 0.7447616306742083, + "grad_norm": 1.2733464241027832, + "learning_rate": 3.7144858286373453e-06, + "loss": 0.4278, + "step": 16821 + }, + { + "epoch": 0.7448059063790222, + "grad_norm": 1.1520675420761108, + "learning_rate": 3.7132839013703758e-06, + "loss": 0.4348, + "step": 16822 + }, + { + "epoch": 0.744850182083836, + "grad_norm": 1.6590059995651245, + "learning_rate": 3.7120821242565086e-06, + "loss": 0.5135, + "step": 16823 + }, + { + "epoch": 0.7448944577886499, + "grad_norm": 1.6691954135894775, + "learning_rate": 3.7108804973244363e-06, + "loss": 0.5325, + "step": 16824 + }, + { + "epoch": 0.7449387334934638, + "grad_norm": 1.306888222694397, + "learning_rate": 3.7096790206028587e-06, + "loss": 0.4956, + "step": 16825 + }, + { + "epoch": 0.7449830091982776, + "grad_norm": 1.3738572597503662, + "learning_rate": 3.70847769412048e-06, + "loss": 0.485, + "step": 16826 + }, + { + "epoch": 0.7450272849030916, + "grad_norm": 1.5834498405456543, + "learning_rate": 3.707276517905989e-06, + "loss": 0.5118, + "step": 16827 + }, + { + "epoch": 0.7450715606079055, + "grad_norm": 1.4946447610855103, + "learning_rate": 3.7060754919880717e-06, + "loss": 0.4987, + "step": 16828 + }, + { + "epoch": 0.7451158363127193, + "grad_norm": 1.4138127565383911, + "learning_rate": 3.7048746163954184e-06, + "loss": 0.5028, + "step": 16829 + }, + { + "epoch": 0.7451601120175332, + "grad_norm": 1.3563737869262695, + "learning_rate": 3.7036738911567105e-06, + "loss": 0.5233, + "step": 16830 + }, + { + "epoch": 0.7452043877223471, + "grad_norm": 1.275017499923706, + "learning_rate": 3.702473316300624e-06, + "loss": 0.4811, + "step": 16831 + }, + { + "epoch": 0.7452486634271609, + "grad_norm": 1.3957672119140625, + "learning_rate": 3.7012728918558362e-06, + "loss": 0.4817, + "step": 16832 + }, + { + "epoch": 0.7452929391319748, + "grad_norm": 1.7468862533569336, + "learning_rate": 3.700072617851015e-06, + "loss": 0.5468, + "step": 16833 + }, + { + "epoch": 0.7453372148367887, + "grad_norm": 1.5478837490081787, + "learning_rate": 3.6988724943148314e-06, + "loss": 0.5032, + "step": 16834 + }, + { + "epoch": 0.7453814905416025, + "grad_norm": 1.7967170476913452, + "learning_rate": 3.6976725212759425e-06, + "loss": 0.6069, + "step": 16835 + }, + { + "epoch": 0.7454257662464164, + "grad_norm": 1.3989564180374146, + "learning_rate": 3.6964726987630182e-06, + "loss": 0.5009, + "step": 16836 + }, + { + "epoch": 0.7454700419512303, + "grad_norm": 2.015501022338867, + "learning_rate": 3.695273026804711e-06, + "loss": 0.5489, + "step": 16837 + }, + { + "epoch": 0.7455143176560441, + "grad_norm": 1.3258779048919678, + "learning_rate": 3.6940735054296704e-06, + "loss": 0.5096, + "step": 16838 + }, + { + "epoch": 0.7455585933608581, + "grad_norm": 1.4099503755569458, + "learning_rate": 3.692874134666554e-06, + "loss": 0.4687, + "step": 16839 + }, + { + "epoch": 0.745602869065672, + "grad_norm": 1.7754069566726685, + "learning_rate": 3.6916749145440023e-06, + "loss": 0.5681, + "step": 16840 + }, + { + "epoch": 0.7456471447704858, + "grad_norm": 1.752458095550537, + "learning_rate": 3.6904758450906596e-06, + "loss": 0.5016, + "step": 16841 + }, + { + "epoch": 0.7456914204752997, + "grad_norm": 1.2484091520309448, + "learning_rate": 3.6892769263351637e-06, + "loss": 0.478, + "step": 16842 + }, + { + "epoch": 0.7457356961801136, + "grad_norm": 1.536858081817627, + "learning_rate": 3.68807815830615e-06, + "loss": 0.5208, + "step": 16843 + }, + { + "epoch": 0.7457799718849274, + "grad_norm": 1.7141135931015015, + "learning_rate": 3.68687954103225e-06, + "loss": 0.646, + "step": 16844 + }, + { + "epoch": 0.7458242475897413, + "grad_norm": 1.281438946723938, + "learning_rate": 3.685681074542089e-06, + "loss": 0.4664, + "step": 16845 + }, + { + "epoch": 0.7458685232945552, + "grad_norm": 1.4402614831924438, + "learning_rate": 3.684482758864297e-06, + "loss": 0.5013, + "step": 16846 + }, + { + "epoch": 0.745912798999369, + "grad_norm": 1.6772857904434204, + "learning_rate": 3.683284594027492e-06, + "loss": 0.561, + "step": 16847 + }, + { + "epoch": 0.7459570747041829, + "grad_norm": 2.1371841430664062, + "learning_rate": 3.6820865800602913e-06, + "loss": 0.6126, + "step": 16848 + }, + { + "epoch": 0.7460013504089968, + "grad_norm": 1.59196138381958, + "learning_rate": 3.6808887169913055e-06, + "loss": 0.5247, + "step": 16849 + }, + { + "epoch": 0.7460456261138106, + "grad_norm": 1.5855908393859863, + "learning_rate": 3.6796910048491506e-06, + "loss": 0.5154, + "step": 16850 + }, + { + "epoch": 0.7460899018186246, + "grad_norm": 1.4059487581253052, + "learning_rate": 3.67849344366243e-06, + "loss": 0.502, + "step": 16851 + }, + { + "epoch": 0.7461341775234385, + "grad_norm": 1.669528841972351, + "learning_rate": 3.677296033459745e-06, + "loss": 0.5074, + "step": 16852 + }, + { + "epoch": 0.7461784532282524, + "grad_norm": 1.9293396472930908, + "learning_rate": 3.6760987742696972e-06, + "loss": 0.5368, + "step": 16853 + }, + { + "epoch": 0.7462227289330662, + "grad_norm": 1.458436369895935, + "learning_rate": 3.6749016661208814e-06, + "loss": 0.5029, + "step": 16854 + }, + { + "epoch": 0.7462670046378801, + "grad_norm": 1.618539571762085, + "learning_rate": 3.6737047090418855e-06, + "loss": 0.5715, + "step": 16855 + }, + { + "epoch": 0.746311280342694, + "grad_norm": 1.5371447801589966, + "learning_rate": 3.6725079030613052e-06, + "loss": 0.5861, + "step": 16856 + }, + { + "epoch": 0.7463555560475078, + "grad_norm": 1.5093461275100708, + "learning_rate": 3.6713112482077205e-06, + "loss": 0.5055, + "step": 16857 + }, + { + "epoch": 0.7463998317523217, + "grad_norm": 1.6054840087890625, + "learning_rate": 3.670114744509715e-06, + "loss": 0.564, + "step": 16858 + }, + { + "epoch": 0.7464441074571356, + "grad_norm": 1.5019675493240356, + "learning_rate": 3.6689183919958605e-06, + "loss": 0.5453, + "step": 16859 + }, + { + "epoch": 0.7464883831619494, + "grad_norm": 1.4158272743225098, + "learning_rate": 3.6677221906947393e-06, + "loss": 0.4635, + "step": 16860 + }, + { + "epoch": 0.7465326588667633, + "grad_norm": 1.9388889074325562, + "learning_rate": 3.666526140634917e-06, + "loss": 0.5921, + "step": 16861 + }, + { + "epoch": 0.7465769345715773, + "grad_norm": 1.6499613523483276, + "learning_rate": 3.665330241844962e-06, + "loss": 0.5581, + "step": 16862 + }, + { + "epoch": 0.7466212102763911, + "grad_norm": 1.8262743949890137, + "learning_rate": 3.6641344943534353e-06, + "loss": 0.5339, + "step": 16863 + }, + { + "epoch": 0.746665485981205, + "grad_norm": 1.7735624313354492, + "learning_rate": 3.662938898188898e-06, + "loss": 0.5944, + "step": 16864 + }, + { + "epoch": 0.7467097616860189, + "grad_norm": 1.5814820528030396, + "learning_rate": 3.661743453379901e-06, + "loss": 0.5213, + "step": 16865 + }, + { + "epoch": 0.7467540373908327, + "grad_norm": 1.3322877883911133, + "learning_rate": 3.6605481599550053e-06, + "loss": 0.4912, + "step": 16866 + }, + { + "epoch": 0.7467983130956466, + "grad_norm": 1.3497822284698486, + "learning_rate": 3.659353017942754e-06, + "loss": 0.5433, + "step": 16867 + }, + { + "epoch": 0.7468425888004605, + "grad_norm": 1.2919890880584717, + "learning_rate": 3.6581580273716944e-06, + "loss": 0.4746, + "step": 16868 + }, + { + "epoch": 0.7468868645052743, + "grad_norm": 1.5887104272842407, + "learning_rate": 3.6569631882703617e-06, + "loss": 0.5334, + "step": 16869 + }, + { + "epoch": 0.7469311402100882, + "grad_norm": 1.5996044874191284, + "learning_rate": 3.655768500667306e-06, + "loss": 0.5751, + "step": 16870 + }, + { + "epoch": 0.7469754159149021, + "grad_norm": 1.4675041437149048, + "learning_rate": 3.6545739645910504e-06, + "loss": 0.526, + "step": 16871 + }, + { + "epoch": 0.7470196916197159, + "grad_norm": 1.7443214654922485, + "learning_rate": 3.6533795800701268e-06, + "loss": 0.5779, + "step": 16872 + }, + { + "epoch": 0.7470639673245298, + "grad_norm": 1.4961018562316895, + "learning_rate": 3.6521853471330658e-06, + "loss": 0.5212, + "step": 16873 + }, + { + "epoch": 0.7471082430293438, + "grad_norm": 1.5761499404907227, + "learning_rate": 3.650991265808391e-06, + "loss": 0.5757, + "step": 16874 + }, + { + "epoch": 0.7471525187341576, + "grad_norm": 1.8494638204574585, + "learning_rate": 3.6497973361246153e-06, + "loss": 0.5609, + "step": 16875 + }, + { + "epoch": 0.7471967944389715, + "grad_norm": 1.377625823020935, + "learning_rate": 3.6486035581102643e-06, + "loss": 0.5512, + "step": 16876 + }, + { + "epoch": 0.7472410701437854, + "grad_norm": 1.6332180500030518, + "learning_rate": 3.647409931793845e-06, + "loss": 0.5651, + "step": 16877 + }, + { + "epoch": 0.7472853458485992, + "grad_norm": 1.4565509557724, + "learning_rate": 3.6462164572038683e-06, + "loss": 0.5049, + "step": 16878 + }, + { + "epoch": 0.7473296215534131, + "grad_norm": 1.6401076316833496, + "learning_rate": 3.645023134368837e-06, + "loss": 0.4829, + "step": 16879 + }, + { + "epoch": 0.747373897258227, + "grad_norm": 1.5256445407867432, + "learning_rate": 3.6438299633172538e-06, + "loss": 0.5803, + "step": 16880 + }, + { + "epoch": 0.7474181729630408, + "grad_norm": 1.3808698654174805, + "learning_rate": 3.6426369440776164e-06, + "loss": 0.5034, + "step": 16881 + }, + { + "epoch": 0.7474624486678547, + "grad_norm": 1.3504689931869507, + "learning_rate": 3.6414440766784164e-06, + "loss": 0.5309, + "step": 16882 + }, + { + "epoch": 0.7475067243726686, + "grad_norm": 1.7265238761901855, + "learning_rate": 3.6402513611481505e-06, + "loss": 0.5459, + "step": 16883 + }, + { + "epoch": 0.7475510000774824, + "grad_norm": 1.3239907026290894, + "learning_rate": 3.6390587975153013e-06, + "loss": 0.4827, + "step": 16884 + }, + { + "epoch": 0.7475952757822963, + "grad_norm": 1.6127339601516724, + "learning_rate": 3.6378663858083518e-06, + "loss": 0.5461, + "step": 16885 + }, + { + "epoch": 0.7476395514871103, + "grad_norm": 1.370652675628662, + "learning_rate": 3.636674126055786e-06, + "loss": 0.5169, + "step": 16886 + }, + { + "epoch": 0.7476838271919242, + "grad_norm": 1.3624510765075684, + "learning_rate": 3.6354820182860773e-06, + "loss": 0.4861, + "step": 16887 + }, + { + "epoch": 0.747728102896738, + "grad_norm": 1.4517998695373535, + "learning_rate": 3.634290062527699e-06, + "loss": 0.494, + "step": 16888 + }, + { + "epoch": 0.7477723786015519, + "grad_norm": 1.404255986213684, + "learning_rate": 3.633098258809119e-06, + "loss": 0.5033, + "step": 16889 + }, + { + "epoch": 0.7478166543063658, + "grad_norm": 1.5092500448226929, + "learning_rate": 3.6319066071588037e-06, + "loss": 0.5176, + "step": 16890 + }, + { + "epoch": 0.7478609300111796, + "grad_norm": 1.7032595872879028, + "learning_rate": 3.6307151076052128e-06, + "loss": 0.545, + "step": 16891 + }, + { + "epoch": 0.7479052057159935, + "grad_norm": 1.631773591041565, + "learning_rate": 3.629523760176803e-06, + "loss": 0.572, + "step": 16892 + }, + { + "epoch": 0.7479494814208074, + "grad_norm": 1.7918442487716675, + "learning_rate": 3.628332564902034e-06, + "loss": 0.5966, + "step": 16893 + }, + { + "epoch": 0.7479937571256212, + "grad_norm": 1.4082127809524536, + "learning_rate": 3.627141521809354e-06, + "loss": 0.508, + "step": 16894 + }, + { + "epoch": 0.7480380328304351, + "grad_norm": 1.3027422428131104, + "learning_rate": 3.6259506309272063e-06, + "loss": 0.5074, + "step": 16895 + }, + { + "epoch": 0.748082308535249, + "grad_norm": 1.7607930898666382, + "learning_rate": 3.624759892284042e-06, + "loss": 0.6014, + "step": 16896 + }, + { + "epoch": 0.7481265842400628, + "grad_norm": 1.2364692687988281, + "learning_rate": 3.6235693059082953e-06, + "loss": 0.4322, + "step": 16897 + }, + { + "epoch": 0.7481708599448768, + "grad_norm": 1.4176498651504517, + "learning_rate": 3.622378871828405e-06, + "loss": 0.5246, + "step": 16898 + }, + { + "epoch": 0.7482151356496907, + "grad_norm": 1.1912697553634644, + "learning_rate": 3.6211885900728017e-06, + "loss": 0.446, + "step": 16899 + }, + { + "epoch": 0.7482594113545045, + "grad_norm": 1.5613881349563599, + "learning_rate": 3.619998460669916e-06, + "loss": 0.4902, + "step": 16900 + }, + { + "epoch": 0.7483036870593184, + "grad_norm": 1.5863338708877563, + "learning_rate": 3.618808483648172e-06, + "loss": 0.5041, + "step": 16901 + }, + { + "epoch": 0.7483479627641323, + "grad_norm": 1.7521804571151733, + "learning_rate": 3.6176186590359874e-06, + "loss": 0.5894, + "step": 16902 + }, + { + "epoch": 0.7483922384689461, + "grad_norm": 2.1672773361206055, + "learning_rate": 3.6164289868617884e-06, + "loss": 0.5631, + "step": 16903 + }, + { + "epoch": 0.74843651417376, + "grad_norm": 1.6601303815841675, + "learning_rate": 3.6152394671539858e-06, + "loss": 0.5602, + "step": 16904 + }, + { + "epoch": 0.7484807898785739, + "grad_norm": 1.8155012130737305, + "learning_rate": 3.6140500999409856e-06, + "loss": 0.5606, + "step": 16905 + }, + { + "epoch": 0.7485250655833877, + "grad_norm": 1.4664310216903687, + "learning_rate": 3.6128608852512025e-06, + "loss": 0.4808, + "step": 16906 + }, + { + "epoch": 0.7485693412882016, + "grad_norm": 1.84055757522583, + "learning_rate": 3.611671823113038e-06, + "loss": 0.5221, + "step": 16907 + }, + { + "epoch": 0.7486136169930155, + "grad_norm": 1.289794683456421, + "learning_rate": 3.610482913554889e-06, + "loss": 0.4769, + "step": 16908 + }, + { + "epoch": 0.7486578926978293, + "grad_norm": 1.3974831104278564, + "learning_rate": 3.6092941566051532e-06, + "loss": 0.4977, + "step": 16909 + }, + { + "epoch": 0.7487021684026433, + "grad_norm": 1.4388014078140259, + "learning_rate": 3.608105552292224e-06, + "loss": 0.5814, + "step": 16910 + }, + { + "epoch": 0.7487464441074572, + "grad_norm": 1.4710853099822998, + "learning_rate": 3.6069171006444882e-06, + "loss": 0.5231, + "step": 16911 + }, + { + "epoch": 0.748790719812271, + "grad_norm": 1.3401507139205933, + "learning_rate": 3.605728801690328e-06, + "loss": 0.4486, + "step": 16912 + }, + { + "epoch": 0.7488349955170849, + "grad_norm": 1.9947925806045532, + "learning_rate": 3.604540655458133e-06, + "loss": 0.5289, + "step": 16913 + }, + { + "epoch": 0.7488792712218988, + "grad_norm": 1.4886261224746704, + "learning_rate": 3.603352661976278e-06, + "loss": 0.5075, + "step": 16914 + }, + { + "epoch": 0.7489235469267126, + "grad_norm": 1.5237336158752441, + "learning_rate": 3.602164821273132e-06, + "loss": 0.5048, + "step": 16915 + }, + { + "epoch": 0.7489678226315265, + "grad_norm": 1.2656410932540894, + "learning_rate": 3.600977133377074e-06, + "loss": 0.4682, + "step": 16916 + }, + { + "epoch": 0.7490120983363404, + "grad_norm": 1.5481520891189575, + "learning_rate": 3.5997895983164655e-06, + "loss": 0.53, + "step": 16917 + }, + { + "epoch": 0.7490563740411542, + "grad_norm": 1.2253862619400024, + "learning_rate": 3.5986022161196766e-06, + "loss": 0.461, + "step": 16918 + }, + { + "epoch": 0.7491006497459681, + "grad_norm": 1.8328191041946411, + "learning_rate": 3.5974149868150532e-06, + "loss": 0.5253, + "step": 16919 + }, + { + "epoch": 0.749144925450782, + "grad_norm": 1.3637689352035522, + "learning_rate": 3.596227910430964e-06, + "loss": 0.5217, + "step": 16920 + }, + { + "epoch": 0.7491892011555958, + "grad_norm": 2.1339850425720215, + "learning_rate": 3.595040986995756e-06, + "loss": 0.6125, + "step": 16921 + }, + { + "epoch": 0.7492334768604098, + "grad_norm": 1.3078136444091797, + "learning_rate": 3.5938542165377765e-06, + "loss": 0.4483, + "step": 16922 + }, + { + "epoch": 0.7492777525652237, + "grad_norm": 1.3822882175445557, + "learning_rate": 3.5926675990853752e-06, + "loss": 0.517, + "step": 16923 + }, + { + "epoch": 0.7493220282700376, + "grad_norm": 1.6723672151565552, + "learning_rate": 3.591481134666892e-06, + "loss": 0.5373, + "step": 16924 + }, + { + "epoch": 0.7493663039748514, + "grad_norm": 1.6374298334121704, + "learning_rate": 3.5902948233106605e-06, + "loss": 0.5432, + "step": 16925 + }, + { + "epoch": 0.7494105796796653, + "grad_norm": 1.8189622163772583, + "learning_rate": 3.589108665045026e-06, + "loss": 0.522, + "step": 16926 + }, + { + "epoch": 0.7494548553844792, + "grad_norm": 1.3054581880569458, + "learning_rate": 3.5879226598983065e-06, + "loss": 0.483, + "step": 16927 + }, + { + "epoch": 0.749499131089293, + "grad_norm": 1.7182061672210693, + "learning_rate": 3.586736807898833e-06, + "loss": 0.5245, + "step": 16928 + }, + { + "epoch": 0.7495434067941069, + "grad_norm": 1.7554277181625366, + "learning_rate": 3.585551109074926e-06, + "loss": 0.528, + "step": 16929 + }, + { + "epoch": 0.7495876824989208, + "grad_norm": 1.4770848751068115, + "learning_rate": 3.5843655634549114e-06, + "loss": 0.5216, + "step": 16930 + }, + { + "epoch": 0.7496319582037346, + "grad_norm": 1.6345866918563843, + "learning_rate": 3.583180171067101e-06, + "loss": 0.4991, + "step": 16931 + }, + { + "epoch": 0.7496762339085485, + "grad_norm": 1.5195504426956177, + "learning_rate": 3.581994931939805e-06, + "loss": 0.5184, + "step": 16932 + }, + { + "epoch": 0.7497205096133625, + "grad_norm": 1.6716481447219849, + "learning_rate": 3.5808098461013373e-06, + "loss": 0.5626, + "step": 16933 + }, + { + "epoch": 0.7497647853181763, + "grad_norm": 1.8780888319015503, + "learning_rate": 3.57962491358e-06, + "loss": 0.5973, + "step": 16934 + }, + { + "epoch": 0.7498090610229902, + "grad_norm": 2.2171573638916016, + "learning_rate": 3.578440134404093e-06, + "loss": 0.6305, + "step": 16935 + }, + { + "epoch": 0.7498533367278041, + "grad_norm": 1.2632640600204468, + "learning_rate": 3.5772555086019156e-06, + "loss": 0.5244, + "step": 16936 + }, + { + "epoch": 0.7498976124326179, + "grad_norm": 1.1462137699127197, + "learning_rate": 3.57607103620176e-06, + "loss": 0.4097, + "step": 16937 + }, + { + "epoch": 0.7499418881374318, + "grad_norm": 2.104708671569824, + "learning_rate": 3.574886717231919e-06, + "loss": 0.5758, + "step": 16938 + }, + { + "epoch": 0.7499861638422457, + "grad_norm": 1.4688891172409058, + "learning_rate": 3.5737025517206724e-06, + "loss": 0.5719, + "step": 16939 + }, + { + "epoch": 0.7500304395470595, + "grad_norm": 1.9701777696609497, + "learning_rate": 3.572518539696311e-06, + "loss": 0.5481, + "step": 16940 + }, + { + "epoch": 0.7500747152518734, + "grad_norm": 1.4003665447235107, + "learning_rate": 3.571334681187112e-06, + "loss": 0.509, + "step": 16941 + }, + { + "epoch": 0.7501189909566873, + "grad_norm": 1.4720509052276611, + "learning_rate": 3.5701509762213447e-06, + "loss": 0.5551, + "step": 16942 + }, + { + "epoch": 0.7501632666615011, + "grad_norm": 1.4675452709197998, + "learning_rate": 3.5689674248272908e-06, + "loss": 0.5157, + "step": 16943 + }, + { + "epoch": 0.750207542366315, + "grad_norm": 1.3602906465530396, + "learning_rate": 3.567784027033213e-06, + "loss": 0.4928, + "step": 16944 + }, + { + "epoch": 0.750251818071129, + "grad_norm": 1.347653865814209, + "learning_rate": 3.5666007828673765e-06, + "loss": 0.4808, + "step": 16945 + }, + { + "epoch": 0.7502960937759428, + "grad_norm": 1.9746308326721191, + "learning_rate": 3.565417692358042e-06, + "loss": 0.6194, + "step": 16946 + }, + { + "epoch": 0.7503403694807567, + "grad_norm": 1.7298955917358398, + "learning_rate": 3.5642347555334665e-06, + "loss": 0.605, + "step": 16947 + }, + { + "epoch": 0.7503846451855706, + "grad_norm": 1.4345804452896118, + "learning_rate": 3.563051972421904e-06, + "loss": 0.5117, + "step": 16948 + }, + { + "epoch": 0.7504289208903844, + "grad_norm": 1.3252545595169067, + "learning_rate": 3.5618693430516004e-06, + "loss": 0.4604, + "step": 16949 + }, + { + "epoch": 0.7504731965951983, + "grad_norm": 1.5451925992965698, + "learning_rate": 3.5606868674508087e-06, + "loss": 0.5674, + "step": 16950 + }, + { + "epoch": 0.7505174723000122, + "grad_norm": 1.3988432884216309, + "learning_rate": 3.5595045456477683e-06, + "loss": 0.5191, + "step": 16951 + }, + { + "epoch": 0.750561748004826, + "grad_norm": 1.3986685276031494, + "learning_rate": 3.558322377670714e-06, + "loss": 0.5365, + "step": 16952 + }, + { + "epoch": 0.7506060237096399, + "grad_norm": 1.3016088008880615, + "learning_rate": 3.5571403635478895e-06, + "loss": 0.4748, + "step": 16953 + }, + { + "epoch": 0.7506502994144538, + "grad_norm": 1.4432969093322754, + "learning_rate": 3.5559585033075207e-06, + "loss": 0.5056, + "step": 16954 + }, + { + "epoch": 0.7506945751192676, + "grad_norm": 1.7082886695861816, + "learning_rate": 3.5547767969778355e-06, + "loss": 0.5319, + "step": 16955 + }, + { + "epoch": 0.7507388508240815, + "grad_norm": 1.64592444896698, + "learning_rate": 3.5535952445870593e-06, + "loss": 0.5559, + "step": 16956 + }, + { + "epoch": 0.7507831265288955, + "grad_norm": 1.5790387392044067, + "learning_rate": 3.5524138461634106e-06, + "loss": 0.5914, + "step": 16957 + }, + { + "epoch": 0.7508274022337094, + "grad_norm": 1.6245746612548828, + "learning_rate": 3.5512326017351085e-06, + "loss": 0.5665, + "step": 16958 + }, + { + "epoch": 0.7508716779385232, + "grad_norm": 1.2589565515518188, + "learning_rate": 3.550051511330361e-06, + "loss": 0.4821, + "step": 16959 + }, + { + "epoch": 0.7509159536433371, + "grad_norm": 1.2390066385269165, + "learning_rate": 3.5488705749773856e-06, + "loss": 0.45, + "step": 16960 + }, + { + "epoch": 0.750960229348151, + "grad_norm": 1.574579119682312, + "learning_rate": 3.547689792704383e-06, + "loss": 0.5146, + "step": 16961 + }, + { + "epoch": 0.7510045050529648, + "grad_norm": 2.0870542526245117, + "learning_rate": 3.5465091645395533e-06, + "loss": 0.6016, + "step": 16962 + }, + { + "epoch": 0.7510487807577787, + "grad_norm": 1.5923144817352295, + "learning_rate": 3.545328690511101e-06, + "loss": 0.5039, + "step": 16963 + }, + { + "epoch": 0.7510930564625926, + "grad_norm": 1.4934234619140625, + "learning_rate": 3.5441483706472158e-06, + "loss": 0.5161, + "step": 16964 + }, + { + "epoch": 0.7511373321674064, + "grad_norm": 1.707458257675171, + "learning_rate": 3.542968204976095e-06, + "loss": 0.5507, + "step": 16965 + }, + { + "epoch": 0.7511816078722203, + "grad_norm": 1.5231446027755737, + "learning_rate": 3.5417881935259134e-06, + "loss": 0.5466, + "step": 16966 + }, + { + "epoch": 0.7512258835770342, + "grad_norm": 1.5873408317565918, + "learning_rate": 3.5406083363248666e-06, + "loss": 0.5357, + "step": 16967 + }, + { + "epoch": 0.751270159281848, + "grad_norm": 1.4483596086502075, + "learning_rate": 3.5394286334011306e-06, + "loss": 0.4969, + "step": 16968 + }, + { + "epoch": 0.751314434986662, + "grad_norm": 1.205235242843628, + "learning_rate": 3.538249084782878e-06, + "loss": 0.4372, + "step": 16969 + }, + { + "epoch": 0.7513587106914759, + "grad_norm": 1.563004732131958, + "learning_rate": 3.537069690498286e-06, + "loss": 0.5794, + "step": 16970 + }, + { + "epoch": 0.7514029863962897, + "grad_norm": 1.8781664371490479, + "learning_rate": 3.5358904505755243e-06, + "loss": 0.5019, + "step": 16971 + }, + { + "epoch": 0.7514472621011036, + "grad_norm": 1.700567364692688, + "learning_rate": 3.5347113650427513e-06, + "loss": 0.5241, + "step": 16972 + }, + { + "epoch": 0.7514915378059175, + "grad_norm": 1.416805386543274, + "learning_rate": 3.5335324339281376e-06, + "loss": 0.4734, + "step": 16973 + }, + { + "epoch": 0.7515358135107313, + "grad_norm": 1.945229411125183, + "learning_rate": 3.532353657259838e-06, + "loss": 0.5583, + "step": 16974 + }, + { + "epoch": 0.7515800892155452, + "grad_norm": 1.3439602851867676, + "learning_rate": 3.5311750350660033e-06, + "loss": 0.4979, + "step": 16975 + }, + { + "epoch": 0.7516243649203591, + "grad_norm": 1.287715196609497, + "learning_rate": 3.529996567374783e-06, + "loss": 0.5102, + "step": 16976 + }, + { + "epoch": 0.7516686406251729, + "grad_norm": 1.1855800151824951, + "learning_rate": 3.528818254214329e-06, + "loss": 0.4679, + "step": 16977 + }, + { + "epoch": 0.7517129163299868, + "grad_norm": 1.6611565351486206, + "learning_rate": 3.527640095612782e-06, + "loss": 0.5148, + "step": 16978 + }, + { + "epoch": 0.7517571920348007, + "grad_norm": 1.7645872831344604, + "learning_rate": 3.5264620915982774e-06, + "loss": 0.5375, + "step": 16979 + }, + { + "epoch": 0.7518014677396145, + "grad_norm": 1.3690329790115356, + "learning_rate": 3.5252842421989585e-06, + "loss": 0.4962, + "step": 16980 + }, + { + "epoch": 0.7518457434444284, + "grad_norm": 1.373700499534607, + "learning_rate": 3.5241065474429546e-06, + "loss": 0.4628, + "step": 16981 + }, + { + "epoch": 0.7518900191492424, + "grad_norm": 1.5936475992202759, + "learning_rate": 3.522929007358392e-06, + "loss": 0.5027, + "step": 16982 + }, + { + "epoch": 0.7519342948540562, + "grad_norm": 1.3607022762298584, + "learning_rate": 3.521751621973396e-06, + "loss": 0.5193, + "step": 16983 + }, + { + "epoch": 0.7519785705588701, + "grad_norm": 1.4065836668014526, + "learning_rate": 3.5205743913160883e-06, + "loss": 0.5349, + "step": 16984 + }, + { + "epoch": 0.752022846263684, + "grad_norm": 1.493678331375122, + "learning_rate": 3.5193973154145856e-06, + "loss": 0.5217, + "step": 16985 + }, + { + "epoch": 0.7520671219684978, + "grad_norm": 1.2837039232254028, + "learning_rate": 3.5182203942969974e-06, + "loss": 0.4489, + "step": 16986 + }, + { + "epoch": 0.7521113976733117, + "grad_norm": 1.8611578941345215, + "learning_rate": 3.517043627991441e-06, + "loss": 0.5216, + "step": 16987 + }, + { + "epoch": 0.7521556733781256, + "grad_norm": 1.5810465812683105, + "learning_rate": 3.51586701652602e-06, + "loss": 0.5465, + "step": 16988 + }, + { + "epoch": 0.7521999490829394, + "grad_norm": 1.3427455425262451, + "learning_rate": 3.5146905599288307e-06, + "loss": 0.4717, + "step": 16989 + }, + { + "epoch": 0.7522442247877533, + "grad_norm": 1.344726324081421, + "learning_rate": 3.513514258227981e-06, + "loss": 0.5258, + "step": 16990 + }, + { + "epoch": 0.7522885004925672, + "grad_norm": 1.7532353401184082, + "learning_rate": 3.512338111451562e-06, + "loss": 0.5316, + "step": 16991 + }, + { + "epoch": 0.752332776197381, + "grad_norm": 1.3430489301681519, + "learning_rate": 3.511162119627665e-06, + "loss": 0.4873, + "step": 16992 + }, + { + "epoch": 0.752377051902195, + "grad_norm": 1.4087306261062622, + "learning_rate": 3.5099862827843757e-06, + "loss": 0.4847, + "step": 16993 + }, + { + "epoch": 0.7524213276070089, + "grad_norm": 1.4780222177505493, + "learning_rate": 3.5088106009497815e-06, + "loss": 0.5194, + "step": 16994 + }, + { + "epoch": 0.7524656033118228, + "grad_norm": 1.3399497270584106, + "learning_rate": 3.5076350741519595e-06, + "loss": 0.4984, + "step": 16995 + }, + { + "epoch": 0.7525098790166366, + "grad_norm": 1.3896417617797852, + "learning_rate": 3.5064597024189852e-06, + "loss": 0.486, + "step": 16996 + }, + { + "epoch": 0.7525541547214505, + "grad_norm": 1.359537959098816, + "learning_rate": 3.505284485778936e-06, + "loss": 0.4934, + "step": 16997 + }, + { + "epoch": 0.7525984304262644, + "grad_norm": 1.5097662210464478, + "learning_rate": 3.5041094242598794e-06, + "loss": 0.5571, + "step": 16998 + }, + { + "epoch": 0.7526427061310782, + "grad_norm": 1.551612377166748, + "learning_rate": 3.5029345178898756e-06, + "loss": 0.5367, + "step": 16999 + }, + { + "epoch": 0.7526869818358921, + "grad_norm": 1.6278740167617798, + "learning_rate": 3.501759766696995e-06, + "loss": 0.5661, + "step": 17000 + }, + { + "epoch": 0.752731257540706, + "grad_norm": 1.3238061666488647, + "learning_rate": 3.500585170709292e-06, + "loss": 0.4687, + "step": 17001 + }, + { + "epoch": 0.7527755332455198, + "grad_norm": 1.5002540349960327, + "learning_rate": 3.4994107299548185e-06, + "loss": 0.4903, + "step": 17002 + }, + { + "epoch": 0.7528198089503337, + "grad_norm": 1.3845384120941162, + "learning_rate": 3.4982364444616277e-06, + "loss": 0.485, + "step": 17003 + }, + { + "epoch": 0.7528640846551476, + "grad_norm": 1.7308967113494873, + "learning_rate": 3.4970623142577655e-06, + "loss": 0.5486, + "step": 17004 + }, + { + "epoch": 0.7529083603599614, + "grad_norm": 1.5704082250595093, + "learning_rate": 3.495888339371275e-06, + "loss": 0.5974, + "step": 17005 + }, + { + "epoch": 0.7529526360647754, + "grad_norm": 1.6626625061035156, + "learning_rate": 3.4947145198301923e-06, + "loss": 0.5873, + "step": 17006 + }, + { + "epoch": 0.7529969117695893, + "grad_norm": 1.6536308526992798, + "learning_rate": 3.4935408556625604e-06, + "loss": 0.5376, + "step": 17007 + }, + { + "epoch": 0.7530411874744031, + "grad_norm": 1.4529101848602295, + "learning_rate": 3.492367346896407e-06, + "loss": 0.4953, + "step": 17008 + }, + { + "epoch": 0.753085463179217, + "grad_norm": 1.5853044986724854, + "learning_rate": 3.4911939935597573e-06, + "loss": 0.4979, + "step": 17009 + }, + { + "epoch": 0.7531297388840309, + "grad_norm": 1.4220525026321411, + "learning_rate": 3.490020795680642e-06, + "loss": 0.4972, + "step": 17010 + }, + { + "epoch": 0.7531740145888447, + "grad_norm": 1.5000487565994263, + "learning_rate": 3.48884775328708e-06, + "loss": 0.5157, + "step": 17011 + }, + { + "epoch": 0.7532182902936586, + "grad_norm": 1.6169477701187134, + "learning_rate": 3.487674866407088e-06, + "loss": 0.5344, + "step": 17012 + }, + { + "epoch": 0.7532625659984725, + "grad_norm": 1.6592875719070435, + "learning_rate": 3.486502135068679e-06, + "loss": 0.607, + "step": 17013 + }, + { + "epoch": 0.7533068417032863, + "grad_norm": 1.708556056022644, + "learning_rate": 3.4853295592998637e-06, + "loss": 0.5415, + "step": 17014 + }, + { + "epoch": 0.7533511174081002, + "grad_norm": 1.4043091535568237, + "learning_rate": 3.4841571391286466e-06, + "loss": 0.5448, + "step": 17015 + }, + { + "epoch": 0.7533953931129141, + "grad_norm": 1.3889076709747314, + "learning_rate": 3.4829848745830288e-06, + "loss": 0.5344, + "step": 17016 + }, + { + "epoch": 0.753439668817728, + "grad_norm": 1.4740504026412964, + "learning_rate": 3.4818127656910128e-06, + "loss": 0.5321, + "step": 17017 + }, + { + "epoch": 0.7534839445225419, + "grad_norm": 1.3696420192718506, + "learning_rate": 3.4806408124805925e-06, + "loss": 0.5504, + "step": 17018 + }, + { + "epoch": 0.7535282202273558, + "grad_norm": 1.6193549633026123, + "learning_rate": 3.479469014979754e-06, + "loss": 0.528, + "step": 17019 + }, + { + "epoch": 0.7535724959321696, + "grad_norm": 1.404528260231018, + "learning_rate": 3.4782973732164926e-06, + "loss": 0.4976, + "step": 17020 + }, + { + "epoch": 0.7536167716369835, + "grad_norm": 1.9547836780548096, + "learning_rate": 3.4771258872187917e-06, + "loss": 0.5969, + "step": 17021 + }, + { + "epoch": 0.7536610473417974, + "grad_norm": 1.5410557985305786, + "learning_rate": 3.4759545570146215e-06, + "loss": 0.5734, + "step": 17022 + }, + { + "epoch": 0.7537053230466112, + "grad_norm": 1.96755850315094, + "learning_rate": 3.4747833826319677e-06, + "loss": 0.5915, + "step": 17023 + }, + { + "epoch": 0.7537495987514251, + "grad_norm": 2.0379745960235596, + "learning_rate": 3.4736123640987997e-06, + "loss": 0.5035, + "step": 17024 + }, + { + "epoch": 0.753793874456239, + "grad_norm": 1.5688259601593018, + "learning_rate": 3.4724415014430856e-06, + "loss": 0.5271, + "step": 17025 + }, + { + "epoch": 0.7538381501610528, + "grad_norm": 1.7226767539978027, + "learning_rate": 3.4712707946927894e-06, + "loss": 0.6035, + "step": 17026 + }, + { + "epoch": 0.7538824258658667, + "grad_norm": 1.6264641284942627, + "learning_rate": 3.4701002438758767e-06, + "loss": 0.5667, + "step": 17027 + }, + { + "epoch": 0.7539267015706806, + "grad_norm": 1.1981135606765747, + "learning_rate": 3.4689298490203026e-06, + "loss": 0.4233, + "step": 17028 + }, + { + "epoch": 0.7539709772754946, + "grad_norm": 1.6557382345199585, + "learning_rate": 3.4677596101540222e-06, + "loss": 0.5271, + "step": 17029 + }, + { + "epoch": 0.7540152529803084, + "grad_norm": 1.6098690032958984, + "learning_rate": 3.466589527304984e-06, + "loss": 0.5182, + "step": 17030 + }, + { + "epoch": 0.7540595286851223, + "grad_norm": 1.5024279356002808, + "learning_rate": 3.4654196005011363e-06, + "loss": 0.4978, + "step": 17031 + }, + { + "epoch": 0.7541038043899362, + "grad_norm": 1.5871442556381226, + "learning_rate": 3.4642498297704163e-06, + "loss": 0.528, + "step": 17032 + }, + { + "epoch": 0.75414808009475, + "grad_norm": 1.471193552017212, + "learning_rate": 3.4630802151407716e-06, + "loss": 0.5072, + "step": 17033 + }, + { + "epoch": 0.7541923557995639, + "grad_norm": 1.7884819507598877, + "learning_rate": 3.461910756640133e-06, + "loss": 0.5236, + "step": 17034 + }, + { + "epoch": 0.7542366315043778, + "grad_norm": 1.3809937238693237, + "learning_rate": 3.4607414542964323e-06, + "loss": 0.5528, + "step": 17035 + }, + { + "epoch": 0.7542809072091916, + "grad_norm": 1.220375657081604, + "learning_rate": 3.4595723081375952e-06, + "loss": 0.4826, + "step": 17036 + }, + { + "epoch": 0.7543251829140055, + "grad_norm": 1.4597759246826172, + "learning_rate": 3.45840331819155e-06, + "loss": 0.5052, + "step": 17037 + }, + { + "epoch": 0.7543694586188194, + "grad_norm": 1.2960283756256104, + "learning_rate": 3.457234484486216e-06, + "loss": 0.4687, + "step": 17038 + }, + { + "epoch": 0.7544137343236332, + "grad_norm": 1.6115256547927856, + "learning_rate": 3.456065807049509e-06, + "loss": 0.497, + "step": 17039 + }, + { + "epoch": 0.7544580100284471, + "grad_norm": 1.5285476446151733, + "learning_rate": 3.4548972859093423e-06, + "loss": 0.5445, + "step": 17040 + }, + { + "epoch": 0.7545022857332611, + "grad_norm": 1.36722731590271, + "learning_rate": 3.453728921093624e-06, + "loss": 0.4741, + "step": 17041 + }, + { + "epoch": 0.7545465614380749, + "grad_norm": 1.592623233795166, + "learning_rate": 3.4525607126302573e-06, + "loss": 0.5307, + "step": 17042 + }, + { + "epoch": 0.7545908371428888, + "grad_norm": 1.5642200708389282, + "learning_rate": 3.4513926605471504e-06, + "loss": 0.5363, + "step": 17043 + }, + { + "epoch": 0.7546351128477027, + "grad_norm": 1.4435936212539673, + "learning_rate": 3.4502247648721967e-06, + "loss": 0.5095, + "step": 17044 + }, + { + "epoch": 0.7546793885525165, + "grad_norm": 1.5611763000488281, + "learning_rate": 3.4490570256332922e-06, + "loss": 0.5531, + "step": 17045 + }, + { + "epoch": 0.7547236642573304, + "grad_norm": 1.7301201820373535, + "learning_rate": 3.447889442858323e-06, + "loss": 0.5385, + "step": 17046 + }, + { + "epoch": 0.7547679399621443, + "grad_norm": 1.6083688735961914, + "learning_rate": 3.4467220165751834e-06, + "loss": 0.5619, + "step": 17047 + }, + { + "epoch": 0.7548122156669581, + "grad_norm": 1.8817464113235474, + "learning_rate": 3.4455547468117524e-06, + "loss": 0.5024, + "step": 17048 + }, + { + "epoch": 0.754856491371772, + "grad_norm": 1.457558274269104, + "learning_rate": 3.444387633595909e-06, + "loss": 0.5397, + "step": 17049 + }, + { + "epoch": 0.7549007670765859, + "grad_norm": 1.6481356620788574, + "learning_rate": 3.4432206769555298e-06, + "loss": 0.6083, + "step": 17050 + }, + { + "epoch": 0.7549450427813997, + "grad_norm": 1.9172738790512085, + "learning_rate": 3.4420538769184865e-06, + "loss": 0.6221, + "step": 17051 + }, + { + "epoch": 0.7549893184862136, + "grad_norm": 1.5787763595581055, + "learning_rate": 3.440887233512642e-06, + "loss": 0.5062, + "step": 17052 + }, + { + "epoch": 0.7550335941910276, + "grad_norm": 1.5984619855880737, + "learning_rate": 3.439720746765869e-06, + "loss": 0.4998, + "step": 17053 + }, + { + "epoch": 0.7550778698958414, + "grad_norm": 1.615962028503418, + "learning_rate": 3.4385544167060246e-06, + "loss": 0.5642, + "step": 17054 + }, + { + "epoch": 0.7551221456006553, + "grad_norm": 1.635643482208252, + "learning_rate": 3.437388243360966e-06, + "loss": 0.5326, + "step": 17055 + }, + { + "epoch": 0.7551664213054692, + "grad_norm": 1.5653371810913086, + "learning_rate": 3.436222226758542e-06, + "loss": 0.5411, + "step": 17056 + }, + { + "epoch": 0.755210697010283, + "grad_norm": 1.5848407745361328, + "learning_rate": 3.4350563669266077e-06, + "loss": 0.4786, + "step": 17057 + }, + { + "epoch": 0.7552549727150969, + "grad_norm": 2.104811906814575, + "learning_rate": 3.433890663893009e-06, + "loss": 0.5598, + "step": 17058 + }, + { + "epoch": 0.7552992484199108, + "grad_norm": 1.5046977996826172, + "learning_rate": 3.4327251176855835e-06, + "loss": 0.5404, + "step": 17059 + }, + { + "epoch": 0.7553435241247246, + "grad_norm": 1.6481305360794067, + "learning_rate": 3.4315597283321724e-06, + "loss": 0.5521, + "step": 17060 + }, + { + "epoch": 0.7553877998295385, + "grad_norm": 1.2819793224334717, + "learning_rate": 3.430394495860608e-06, + "loss": 0.4721, + "step": 17061 + }, + { + "epoch": 0.7554320755343524, + "grad_norm": 1.527849793434143, + "learning_rate": 3.42922942029872e-06, + "loss": 0.4962, + "step": 17062 + }, + { + "epoch": 0.7554763512391663, + "grad_norm": 1.3881222009658813, + "learning_rate": 3.4280645016743396e-06, + "loss": 0.5073, + "step": 17063 + }, + { + "epoch": 0.7555206269439801, + "grad_norm": 1.1519664525985718, + "learning_rate": 3.426899740015288e-06, + "loss": 0.4824, + "step": 17064 + }, + { + "epoch": 0.7555649026487941, + "grad_norm": 1.181115746498108, + "learning_rate": 3.425735135349384e-06, + "loss": 0.4355, + "step": 17065 + }, + { + "epoch": 0.755609178353608, + "grad_norm": 1.4925425052642822, + "learning_rate": 3.4245706877044395e-06, + "loss": 0.5388, + "step": 17066 + }, + { + "epoch": 0.7556534540584218, + "grad_norm": 1.354293704032898, + "learning_rate": 3.423406397108273e-06, + "loss": 0.5075, + "step": 17067 + }, + { + "epoch": 0.7556977297632357, + "grad_norm": 1.8867731094360352, + "learning_rate": 3.4222422635886944e-06, + "loss": 0.531, + "step": 17068 + }, + { + "epoch": 0.7557420054680496, + "grad_norm": 1.6480010747909546, + "learning_rate": 3.421078287173497e-06, + "loss": 0.5456, + "step": 17069 + }, + { + "epoch": 0.7557862811728634, + "grad_norm": 1.7405062913894653, + "learning_rate": 3.4199144678904905e-06, + "loss": 0.5051, + "step": 17070 + }, + { + "epoch": 0.7558305568776773, + "grad_norm": 1.4633715152740479, + "learning_rate": 3.418750805767469e-06, + "loss": 0.4758, + "step": 17071 + }, + { + "epoch": 0.7558748325824912, + "grad_norm": 1.565102458000183, + "learning_rate": 3.417587300832227e-06, + "loss": 0.5478, + "step": 17072 + }, + { + "epoch": 0.755919108287305, + "grad_norm": 1.6981207132339478, + "learning_rate": 3.4164239531125486e-06, + "loss": 0.5311, + "step": 17073 + }, + { + "epoch": 0.7559633839921189, + "grad_norm": 1.8926461935043335, + "learning_rate": 3.4152607626362267e-06, + "loss": 0.5572, + "step": 17074 + }, + { + "epoch": 0.7560076596969328, + "grad_norm": 1.5284245014190674, + "learning_rate": 3.414097729431041e-06, + "loss": 0.5398, + "step": 17075 + }, + { + "epoch": 0.7560519354017466, + "grad_norm": 1.5296176671981812, + "learning_rate": 3.412934853524764e-06, + "loss": 0.4483, + "step": 17076 + }, + { + "epoch": 0.7560962111065606, + "grad_norm": 1.627322793006897, + "learning_rate": 3.4117721349451816e-06, + "loss": 0.5151, + "step": 17077 + }, + { + "epoch": 0.7561404868113745, + "grad_norm": 1.5237326622009277, + "learning_rate": 3.410609573720054e-06, + "loss": 0.5269, + "step": 17078 + }, + { + "epoch": 0.7561847625161883, + "grad_norm": 1.8483470678329468, + "learning_rate": 3.409447169877148e-06, + "loss": 0.5308, + "step": 17079 + }, + { + "epoch": 0.7562290382210022, + "grad_norm": 1.456694483757019, + "learning_rate": 3.4082849234442337e-06, + "loss": 0.5186, + "step": 17080 + }, + { + "epoch": 0.7562733139258161, + "grad_norm": 1.4775216579437256, + "learning_rate": 3.407122834449068e-06, + "loss": 0.5104, + "step": 17081 + }, + { + "epoch": 0.7563175896306299, + "grad_norm": 1.7073664665222168, + "learning_rate": 3.4059609029194042e-06, + "loss": 0.5131, + "step": 17082 + }, + { + "epoch": 0.7563618653354438, + "grad_norm": 1.6782066822052002, + "learning_rate": 3.404799128882992e-06, + "loss": 0.5203, + "step": 17083 + }, + { + "epoch": 0.7564061410402577, + "grad_norm": 1.2898610830307007, + "learning_rate": 3.4036375123675856e-06, + "loss": 0.4735, + "step": 17084 + }, + { + "epoch": 0.7564504167450715, + "grad_norm": 1.3588770627975464, + "learning_rate": 3.402476053400926e-06, + "loss": 0.4906, + "step": 17085 + }, + { + "epoch": 0.7564946924498854, + "grad_norm": 1.7956703901290894, + "learning_rate": 3.401314752010755e-06, + "loss": 0.5741, + "step": 17086 + }, + { + "epoch": 0.7565389681546993, + "grad_norm": 1.6563581228256226, + "learning_rate": 3.4001536082248075e-06, + "loss": 0.5199, + "step": 17087 + }, + { + "epoch": 0.7565832438595131, + "grad_norm": 1.6781139373779297, + "learning_rate": 3.398992622070818e-06, + "loss": 0.6001, + "step": 17088 + }, + { + "epoch": 0.7566275195643271, + "grad_norm": 1.817478895187378, + "learning_rate": 3.3978317935765104e-06, + "loss": 0.6047, + "step": 17089 + }, + { + "epoch": 0.756671795269141, + "grad_norm": 1.6818596124649048, + "learning_rate": 3.3966711227696193e-06, + "loss": 0.4874, + "step": 17090 + }, + { + "epoch": 0.7567160709739548, + "grad_norm": 1.70124351978302, + "learning_rate": 3.395510609677861e-06, + "loss": 0.5105, + "step": 17091 + }, + { + "epoch": 0.7567603466787687, + "grad_norm": 1.9951364994049072, + "learning_rate": 3.3943502543289552e-06, + "loss": 0.5523, + "step": 17092 + }, + { + "epoch": 0.7568046223835826, + "grad_norm": 1.478543758392334, + "learning_rate": 3.3931900567506104e-06, + "loss": 0.5194, + "step": 17093 + }, + { + "epoch": 0.7568488980883964, + "grad_norm": 1.7571591138839722, + "learning_rate": 3.3920300169705455e-06, + "loss": 0.5137, + "step": 17094 + }, + { + "epoch": 0.7568931737932103, + "grad_norm": 1.4563547372817993, + "learning_rate": 3.390870135016464e-06, + "loss": 0.5227, + "step": 17095 + }, + { + "epoch": 0.7569374494980242, + "grad_norm": 1.4929161071777344, + "learning_rate": 3.389710410916067e-06, + "loss": 0.5257, + "step": 17096 + }, + { + "epoch": 0.756981725202838, + "grad_norm": 2.0253052711486816, + "learning_rate": 3.388550844697055e-06, + "loss": 0.5638, + "step": 17097 + }, + { + "epoch": 0.7570260009076519, + "grad_norm": 1.2925925254821777, + "learning_rate": 3.387391436387123e-06, + "loss": 0.4601, + "step": 17098 + }, + { + "epoch": 0.7570702766124658, + "grad_norm": 1.4335136413574219, + "learning_rate": 3.3862321860139578e-06, + "loss": 0.4621, + "step": 17099 + }, + { + "epoch": 0.7571145523172798, + "grad_norm": 1.790284514427185, + "learning_rate": 3.385073093605257e-06, + "loss": 0.5127, + "step": 17100 + }, + { + "epoch": 0.7571588280220936, + "grad_norm": 1.2551809549331665, + "learning_rate": 3.3839141591886983e-06, + "loss": 0.4279, + "step": 17101 + }, + { + "epoch": 0.7572031037269075, + "grad_norm": 1.5967637300491333, + "learning_rate": 3.3827553827919624e-06, + "loss": 0.5489, + "step": 17102 + }, + { + "epoch": 0.7572473794317214, + "grad_norm": 1.3356281518936157, + "learning_rate": 3.3815967644427226e-06, + "loss": 0.4585, + "step": 17103 + }, + { + "epoch": 0.7572916551365352, + "grad_norm": 1.722920536994934, + "learning_rate": 3.3804383041686594e-06, + "loss": 0.5294, + "step": 17104 + }, + { + "epoch": 0.7573359308413491, + "grad_norm": 1.3859922885894775, + "learning_rate": 3.3792800019974382e-06, + "loss": 0.4775, + "step": 17105 + }, + { + "epoch": 0.757380206546163, + "grad_norm": 1.2798758745193481, + "learning_rate": 3.378121857956722e-06, + "loss": 0.4553, + "step": 17106 + }, + { + "epoch": 0.7574244822509768, + "grad_norm": 1.405151128768921, + "learning_rate": 3.3769638720741737e-06, + "loss": 0.5203, + "step": 17107 + }, + { + "epoch": 0.7574687579557907, + "grad_norm": 1.53084135055542, + "learning_rate": 3.375806044377451e-06, + "loss": 0.4857, + "step": 17108 + }, + { + "epoch": 0.7575130336606046, + "grad_norm": 1.328386664390564, + "learning_rate": 3.374648374894204e-06, + "loss": 0.5241, + "step": 17109 + }, + { + "epoch": 0.7575573093654184, + "grad_norm": 1.3790955543518066, + "learning_rate": 3.373490863652089e-06, + "loss": 0.4731, + "step": 17110 + }, + { + "epoch": 0.7576015850702323, + "grad_norm": 1.4350059032440186, + "learning_rate": 3.37233351067875e-06, + "loss": 0.5192, + "step": 17111 + }, + { + "epoch": 0.7576458607750463, + "grad_norm": 1.5049701929092407, + "learning_rate": 3.3711763160018275e-06, + "loss": 0.504, + "step": 17112 + }, + { + "epoch": 0.75769013647986, + "grad_norm": 1.3252438306808472, + "learning_rate": 3.370019279648958e-06, + "loss": 0.4841, + "step": 17113 + }, + { + "epoch": 0.757734412184674, + "grad_norm": 1.6372979879379272, + "learning_rate": 3.3688624016477845e-06, + "loss": 0.5164, + "step": 17114 + }, + { + "epoch": 0.7577786878894879, + "grad_norm": 1.8663398027420044, + "learning_rate": 3.3677056820259324e-06, + "loss": 0.5778, + "step": 17115 + }, + { + "epoch": 0.7578229635943017, + "grad_norm": 1.503902792930603, + "learning_rate": 3.36654912081103e-06, + "loss": 0.4964, + "step": 17116 + }, + { + "epoch": 0.7578672392991156, + "grad_norm": 1.250644326210022, + "learning_rate": 3.3653927180307e-06, + "loss": 0.4678, + "step": 17117 + }, + { + "epoch": 0.7579115150039295, + "grad_norm": 1.481502652168274, + "learning_rate": 3.364236473712563e-06, + "loss": 0.4785, + "step": 17118 + }, + { + "epoch": 0.7579557907087433, + "grad_norm": 1.4148917198181152, + "learning_rate": 3.3630803878842322e-06, + "loss": 0.4744, + "step": 17119 + }, + { + "epoch": 0.7580000664135572, + "grad_norm": 1.4824554920196533, + "learning_rate": 3.3619244605733246e-06, + "loss": 0.5078, + "step": 17120 + }, + { + "epoch": 0.7580443421183711, + "grad_norm": 1.5211620330810547, + "learning_rate": 3.360768691807448e-06, + "loss": 0.5034, + "step": 17121 + }, + { + "epoch": 0.7580886178231849, + "grad_norm": 1.4900946617126465, + "learning_rate": 3.359613081614204e-06, + "loss": 0.4903, + "step": 17122 + }, + { + "epoch": 0.7581328935279988, + "grad_norm": 1.270463228225708, + "learning_rate": 3.3584576300211914e-06, + "loss": 0.4411, + "step": 17123 + }, + { + "epoch": 0.7581771692328128, + "grad_norm": 1.6303973197937012, + "learning_rate": 3.3573023370560176e-06, + "loss": 0.5291, + "step": 17124 + }, + { + "epoch": 0.7582214449376266, + "grad_norm": 1.4212465286254883, + "learning_rate": 3.356147202746266e-06, + "loss": 0.4713, + "step": 17125 + }, + { + "epoch": 0.7582657206424405, + "grad_norm": 1.6890720129013062, + "learning_rate": 3.3549922271195247e-06, + "loss": 0.5425, + "step": 17126 + }, + { + "epoch": 0.7583099963472544, + "grad_norm": 1.7067692279815674, + "learning_rate": 3.3538374102033865e-06, + "loss": 0.5067, + "step": 17127 + }, + { + "epoch": 0.7583542720520682, + "grad_norm": 1.4492563009262085, + "learning_rate": 3.352682752025431e-06, + "loss": 0.471, + "step": 17128 + }, + { + "epoch": 0.7583985477568821, + "grad_norm": 1.502806544303894, + "learning_rate": 3.351528252613232e-06, + "loss": 0.5345, + "step": 17129 + }, + { + "epoch": 0.758442823461696, + "grad_norm": 1.5965921878814697, + "learning_rate": 3.3503739119943714e-06, + "loss": 0.4954, + "step": 17130 + }, + { + "epoch": 0.7584870991665098, + "grad_norm": 1.6681071519851685, + "learning_rate": 3.3492197301964145e-06, + "loss": 0.5647, + "step": 17131 + }, + { + "epoch": 0.7585313748713237, + "grad_norm": 1.388629674911499, + "learning_rate": 3.348065707246929e-06, + "loss": 0.5115, + "step": 17132 + }, + { + "epoch": 0.7585756505761376, + "grad_norm": 1.6053446531295776, + "learning_rate": 3.3469118431734782e-06, + "loss": 0.6193, + "step": 17133 + }, + { + "epoch": 0.7586199262809515, + "grad_norm": 1.7859364748001099, + "learning_rate": 3.345758138003621e-06, + "loss": 0.5386, + "step": 17134 + }, + { + "epoch": 0.7586642019857653, + "grad_norm": 1.6285752058029175, + "learning_rate": 3.3446045917649126e-06, + "loss": 0.481, + "step": 17135 + }, + { + "epoch": 0.7587084776905793, + "grad_norm": 1.6344980001449585, + "learning_rate": 3.343451204484901e-06, + "loss": 0.5193, + "step": 17136 + }, + { + "epoch": 0.7587527533953932, + "grad_norm": 1.6684399843215942, + "learning_rate": 3.342297976191141e-06, + "loss": 0.614, + "step": 17137 + }, + { + "epoch": 0.758797029100207, + "grad_norm": 1.3448175191879272, + "learning_rate": 3.3411449069111725e-06, + "loss": 0.4744, + "step": 17138 + }, + { + "epoch": 0.7588413048050209, + "grad_norm": 1.582890510559082, + "learning_rate": 3.3399919966725335e-06, + "loss": 0.5748, + "step": 17139 + }, + { + "epoch": 0.7588855805098348, + "grad_norm": 1.9130913019180298, + "learning_rate": 3.3388392455027654e-06, + "loss": 0.5188, + "step": 17140 + }, + { + "epoch": 0.7589298562146486, + "grad_norm": 1.650750756263733, + "learning_rate": 3.3376866534293983e-06, + "loss": 0.5595, + "step": 17141 + }, + { + "epoch": 0.7589741319194625, + "grad_norm": 1.3818464279174805, + "learning_rate": 3.3365342204799613e-06, + "loss": 0.4821, + "step": 17142 + }, + { + "epoch": 0.7590184076242764, + "grad_norm": 1.4071539640426636, + "learning_rate": 3.335381946681979e-06, + "loss": 0.5048, + "step": 17143 + }, + { + "epoch": 0.7590626833290902, + "grad_norm": 1.4674407243728638, + "learning_rate": 3.334229832062972e-06, + "loss": 0.5091, + "step": 17144 + }, + { + "epoch": 0.7591069590339041, + "grad_norm": 1.7274432182312012, + "learning_rate": 3.333077876650458e-06, + "loss": 0.5468, + "step": 17145 + }, + { + "epoch": 0.759151234738718, + "grad_norm": 1.3929065465927124, + "learning_rate": 3.331926080471947e-06, + "loss": 0.5, + "step": 17146 + }, + { + "epoch": 0.7591955104435318, + "grad_norm": 1.3838233947753906, + "learning_rate": 3.330774443554956e-06, + "loss": 0.5164, + "step": 17147 + }, + { + "epoch": 0.7592397861483458, + "grad_norm": 1.5121580362319946, + "learning_rate": 3.329622965926986e-06, + "loss": 0.5101, + "step": 17148 + }, + { + "epoch": 0.7592840618531597, + "grad_norm": 1.564685344696045, + "learning_rate": 3.3284716476155377e-06, + "loss": 0.4988, + "step": 17149 + }, + { + "epoch": 0.7593283375579735, + "grad_norm": 1.5620455741882324, + "learning_rate": 3.327320488648116e-06, + "loss": 0.5534, + "step": 17150 + }, + { + "epoch": 0.7593726132627874, + "grad_norm": 1.9057599306106567, + "learning_rate": 3.32616948905221e-06, + "loss": 0.552, + "step": 17151 + }, + { + "epoch": 0.7594168889676013, + "grad_norm": 1.530161738395691, + "learning_rate": 3.3250186488553125e-06, + "loss": 0.5211, + "step": 17152 + }, + { + "epoch": 0.7594611646724151, + "grad_norm": 1.4769095182418823, + "learning_rate": 3.3238679680849095e-06, + "loss": 0.4745, + "step": 17153 + }, + { + "epoch": 0.759505440377229, + "grad_norm": 1.3234553337097168, + "learning_rate": 3.322717446768484e-06, + "loss": 0.5216, + "step": 17154 + }, + { + "epoch": 0.7595497160820429, + "grad_norm": 2.1250312328338623, + "learning_rate": 3.3215670849335156e-06, + "loss": 0.6144, + "step": 17155 + }, + { + "epoch": 0.7595939917868567, + "grad_norm": 1.4718607664108276, + "learning_rate": 3.320416882607477e-06, + "loss": 0.4939, + "step": 17156 + }, + { + "epoch": 0.7596382674916706, + "grad_norm": 1.237308382987976, + "learning_rate": 3.319266839817844e-06, + "loss": 0.4593, + "step": 17157 + }, + { + "epoch": 0.7596825431964845, + "grad_norm": 1.3470757007598877, + "learning_rate": 3.3181169565920846e-06, + "loss": 0.4613, + "step": 17158 + }, + { + "epoch": 0.7597268189012983, + "grad_norm": 1.579978346824646, + "learning_rate": 3.316967232957656e-06, + "loss": 0.5515, + "step": 17159 + }, + { + "epoch": 0.7597710946061123, + "grad_norm": 1.3938660621643066, + "learning_rate": 3.3158176689420283e-06, + "loss": 0.4943, + "step": 17160 + }, + { + "epoch": 0.7598153703109262, + "grad_norm": 1.501341462135315, + "learning_rate": 3.3146682645726526e-06, + "loss": 0.5498, + "step": 17161 + }, + { + "epoch": 0.75985964601574, + "grad_norm": 1.564056396484375, + "learning_rate": 3.313519019876983e-06, + "loss": 0.5364, + "step": 17162 + }, + { + "epoch": 0.7599039217205539, + "grad_norm": 1.3525762557983398, + "learning_rate": 3.3123699348824654e-06, + "loss": 0.4542, + "step": 17163 + }, + { + "epoch": 0.7599481974253678, + "grad_norm": 1.782969355583191, + "learning_rate": 3.3112210096165477e-06, + "loss": 0.5522, + "step": 17164 + }, + { + "epoch": 0.7599924731301816, + "grad_norm": 1.8707090616226196, + "learning_rate": 3.3100722441066702e-06, + "loss": 0.5628, + "step": 17165 + }, + { + "epoch": 0.7600367488349955, + "grad_norm": 1.543351173400879, + "learning_rate": 3.308923638380266e-06, + "loss": 0.4775, + "step": 17166 + }, + { + "epoch": 0.7600810245398094, + "grad_norm": 1.7335999011993408, + "learning_rate": 3.307775192464776e-06, + "loss": 0.5647, + "step": 17167 + }, + { + "epoch": 0.7601253002446232, + "grad_norm": 1.5966464281082153, + "learning_rate": 3.306626906387628e-06, + "loss": 0.5845, + "step": 17168 + }, + { + "epoch": 0.7601695759494371, + "grad_norm": 1.3900651931762695, + "learning_rate": 3.305478780176241e-06, + "loss": 0.4786, + "step": 17169 + }, + { + "epoch": 0.760213851654251, + "grad_norm": 1.635725736618042, + "learning_rate": 3.3043308138580466e-06, + "loss": 0.4916, + "step": 17170 + }, + { + "epoch": 0.760258127359065, + "grad_norm": 1.6310652494430542, + "learning_rate": 3.3031830074604597e-06, + "loss": 0.5308, + "step": 17171 + }, + { + "epoch": 0.7603024030638788, + "grad_norm": 1.6659528017044067, + "learning_rate": 3.3020353610108955e-06, + "loss": 0.5016, + "step": 17172 + }, + { + "epoch": 0.7603466787686927, + "grad_norm": 1.4960439205169678, + "learning_rate": 3.300887874536758e-06, + "loss": 0.6191, + "step": 17173 + }, + { + "epoch": 0.7603909544735066, + "grad_norm": 1.6245595216751099, + "learning_rate": 3.299740548065462e-06, + "loss": 0.5657, + "step": 17174 + }, + { + "epoch": 0.7604352301783204, + "grad_norm": 1.4721101522445679, + "learning_rate": 3.2985933816244065e-06, + "loss": 0.5136, + "step": 17175 + }, + { + "epoch": 0.7604795058831343, + "grad_norm": 1.4942020177841187, + "learning_rate": 3.297446375240989e-06, + "loss": 0.4776, + "step": 17176 + }, + { + "epoch": 0.7605237815879482, + "grad_norm": 1.4448901414871216, + "learning_rate": 3.2962995289426114e-06, + "loss": 0.4575, + "step": 17177 + }, + { + "epoch": 0.760568057292762, + "grad_norm": 1.6040672063827515, + "learning_rate": 3.2951528427566605e-06, + "loss": 0.6248, + "step": 17178 + }, + { + "epoch": 0.7606123329975759, + "grad_norm": 1.5234731435775757, + "learning_rate": 3.294006316710524e-06, + "loss": 0.542, + "step": 17179 + }, + { + "epoch": 0.7606566087023898, + "grad_norm": 1.307394027709961, + "learning_rate": 3.2928599508315873e-06, + "loss": 0.4614, + "step": 17180 + }, + { + "epoch": 0.7607008844072036, + "grad_norm": 1.6132537126541138, + "learning_rate": 3.291713745147229e-06, + "loss": 0.4747, + "step": 17181 + }, + { + "epoch": 0.7607451601120175, + "grad_norm": 1.7398453950881958, + "learning_rate": 3.290567699684826e-06, + "loss": 0.5664, + "step": 17182 + }, + { + "epoch": 0.7607894358168314, + "grad_norm": 1.4509623050689697, + "learning_rate": 3.2894218144717473e-06, + "loss": 0.5211, + "step": 17183 + }, + { + "epoch": 0.7608337115216453, + "grad_norm": 1.5084291696548462, + "learning_rate": 3.2882760895353683e-06, + "loss": 0.5253, + "step": 17184 + }, + { + "epoch": 0.7608779872264592, + "grad_norm": 1.263032078742981, + "learning_rate": 3.2871305249030494e-06, + "loss": 0.4535, + "step": 17185 + }, + { + "epoch": 0.7609222629312731, + "grad_norm": 1.8300615549087524, + "learning_rate": 3.2859851206021487e-06, + "loss": 0.5668, + "step": 17186 + }, + { + "epoch": 0.7609665386360869, + "grad_norm": 1.4473867416381836, + "learning_rate": 3.2848398766600298e-06, + "loss": 0.5386, + "step": 17187 + }, + { + "epoch": 0.7610108143409008, + "grad_norm": 1.3493461608886719, + "learning_rate": 3.2836947931040432e-06, + "loss": 0.5001, + "step": 17188 + }, + { + "epoch": 0.7610550900457147, + "grad_norm": 2.0195107460021973, + "learning_rate": 3.2825498699615377e-06, + "loss": 0.558, + "step": 17189 + }, + { + "epoch": 0.7610993657505285, + "grad_norm": 1.76093590259552, + "learning_rate": 3.2814051072598596e-06, + "loss": 0.498, + "step": 17190 + }, + { + "epoch": 0.7611436414553424, + "grad_norm": 1.4878970384597778, + "learning_rate": 3.280260505026349e-06, + "loss": 0.5253, + "step": 17191 + }, + { + "epoch": 0.7611879171601563, + "grad_norm": 1.3895134925842285, + "learning_rate": 3.2791160632883457e-06, + "loss": 0.4413, + "step": 17192 + }, + { + "epoch": 0.7612321928649701, + "grad_norm": 1.6492758989334106, + "learning_rate": 3.2779717820731795e-06, + "loss": 0.5736, + "step": 17193 + }, + { + "epoch": 0.761276468569784, + "grad_norm": 1.7225837707519531, + "learning_rate": 3.276827661408186e-06, + "loss": 0.5212, + "step": 17194 + }, + { + "epoch": 0.761320744274598, + "grad_norm": 1.0992813110351562, + "learning_rate": 3.2756837013206912e-06, + "loss": 0.414, + "step": 17195 + }, + { + "epoch": 0.7613650199794118, + "grad_norm": 1.349536418914795, + "learning_rate": 3.274539901838012e-06, + "loss": 0.4342, + "step": 17196 + }, + { + "epoch": 0.7614092956842257, + "grad_norm": 1.4690220355987549, + "learning_rate": 3.273396262987475e-06, + "loss": 0.5065, + "step": 17197 + }, + { + "epoch": 0.7614535713890396, + "grad_norm": 1.5892685651779175, + "learning_rate": 3.2722527847963904e-06, + "loss": 0.5172, + "step": 17198 + }, + { + "epoch": 0.7614978470938534, + "grad_norm": 1.478920578956604, + "learning_rate": 3.2711094672920706e-06, + "loss": 0.532, + "step": 17199 + }, + { + "epoch": 0.7615421227986673, + "grad_norm": 1.4239133596420288, + "learning_rate": 3.269966310501822e-06, + "loss": 0.5204, + "step": 17200 + }, + { + "epoch": 0.7615863985034812, + "grad_norm": 1.312256932258606, + "learning_rate": 3.268823314452949e-06, + "loss": 0.4233, + "step": 17201 + }, + { + "epoch": 0.761630674208295, + "grad_norm": 1.681008219718933, + "learning_rate": 3.26768047917275e-06, + "loss": 0.5476, + "step": 17202 + }, + { + "epoch": 0.7616749499131089, + "grad_norm": 2.0836968421936035, + "learning_rate": 3.266537804688518e-06, + "loss": 0.5442, + "step": 17203 + }, + { + "epoch": 0.7617192256179228, + "grad_norm": 1.294487476348877, + "learning_rate": 3.265395291027551e-06, + "loss": 0.467, + "step": 17204 + }, + { + "epoch": 0.7617635013227367, + "grad_norm": 1.601952314376831, + "learning_rate": 3.264252938217134e-06, + "loss": 0.5233, + "step": 17205 + }, + { + "epoch": 0.7618077770275505, + "grad_norm": 1.6810686588287354, + "learning_rate": 3.263110746284548e-06, + "loss": 0.5452, + "step": 17206 + }, + { + "epoch": 0.7618520527323644, + "grad_norm": 1.41597318649292, + "learning_rate": 3.2619687152570787e-06, + "loss": 0.5751, + "step": 17207 + }, + { + "epoch": 0.7618963284371784, + "grad_norm": 1.3861899375915527, + "learning_rate": 3.260826845162002e-06, + "loss": 0.4585, + "step": 17208 + }, + { + "epoch": 0.7619406041419922, + "grad_norm": 1.6619144678115845, + "learning_rate": 3.259685136026588e-06, + "loss": 0.5762, + "step": 17209 + }, + { + "epoch": 0.7619848798468061, + "grad_norm": 1.327285647392273, + "learning_rate": 3.2585435878781056e-06, + "loss": 0.5035, + "step": 17210 + }, + { + "epoch": 0.76202915555162, + "grad_norm": 1.693501591682434, + "learning_rate": 3.257402200743821e-06, + "loss": 0.51, + "step": 17211 + }, + { + "epoch": 0.7620734312564338, + "grad_norm": 1.3024319410324097, + "learning_rate": 3.2562609746509954e-06, + "loss": 0.4701, + "step": 17212 + }, + { + "epoch": 0.7621177069612477, + "grad_norm": 1.617234706878662, + "learning_rate": 3.255119909626883e-06, + "loss": 0.5276, + "step": 17213 + }, + { + "epoch": 0.7621619826660616, + "grad_norm": 1.6397769451141357, + "learning_rate": 3.253979005698741e-06, + "loss": 0.5144, + "step": 17214 + }, + { + "epoch": 0.7622062583708754, + "grad_norm": 1.387171983718872, + "learning_rate": 3.252838262893818e-06, + "loss": 0.4775, + "step": 17215 + }, + { + "epoch": 0.7622505340756893, + "grad_norm": 1.4204363822937012, + "learning_rate": 3.2516976812393576e-06, + "loss": 0.4904, + "step": 17216 + }, + { + "epoch": 0.7622948097805032, + "grad_norm": 1.1869144439697266, + "learning_rate": 3.2505572607626048e-06, + "loss": 0.4298, + "step": 17217 + }, + { + "epoch": 0.762339085485317, + "grad_norm": 1.4159252643585205, + "learning_rate": 3.249417001490798e-06, + "loss": 0.5426, + "step": 17218 + }, + { + "epoch": 0.762383361190131, + "grad_norm": 1.6912294626235962, + "learning_rate": 3.248276903451171e-06, + "loss": 0.503, + "step": 17219 + }, + { + "epoch": 0.7624276368949449, + "grad_norm": 1.3054152727127075, + "learning_rate": 3.2471369666709473e-06, + "loss": 0.4711, + "step": 17220 + }, + { + "epoch": 0.7624719125997587, + "grad_norm": 1.755723476409912, + "learning_rate": 3.2459971911773615e-06, + "loss": 0.5325, + "step": 17221 + }, + { + "epoch": 0.7625161883045726, + "grad_norm": 1.8015191555023193, + "learning_rate": 3.2448575769976345e-06, + "loss": 0.5303, + "step": 17222 + }, + { + "epoch": 0.7625604640093865, + "grad_norm": 1.8224695920944214, + "learning_rate": 3.24371812415898e-06, + "loss": 0.5957, + "step": 17223 + }, + { + "epoch": 0.7626047397142003, + "grad_norm": 1.597470760345459, + "learning_rate": 3.24257883268862e-06, + "loss": 0.4932, + "step": 17224 + }, + { + "epoch": 0.7626490154190142, + "grad_norm": 1.3745999336242676, + "learning_rate": 3.2414397026137624e-06, + "loss": 0.4832, + "step": 17225 + }, + { + "epoch": 0.7626932911238281, + "grad_norm": 1.9966516494750977, + "learning_rate": 3.240300733961612e-06, + "loss": 0.5127, + "step": 17226 + }, + { + "epoch": 0.7627375668286419, + "grad_norm": 1.3996310234069824, + "learning_rate": 3.2391619267593777e-06, + "loss": 0.5111, + "step": 17227 + }, + { + "epoch": 0.7627818425334558, + "grad_norm": 1.6581062078475952, + "learning_rate": 3.238023281034258e-06, + "loss": 0.5478, + "step": 17228 + }, + { + "epoch": 0.7628261182382697, + "grad_norm": 1.4364750385284424, + "learning_rate": 3.236884796813444e-06, + "loss": 0.4759, + "step": 17229 + }, + { + "epoch": 0.7628703939430835, + "grad_norm": 1.3238556385040283, + "learning_rate": 3.235746474124127e-06, + "loss": 0.4838, + "step": 17230 + }, + { + "epoch": 0.7629146696478974, + "grad_norm": 1.6441375017166138, + "learning_rate": 3.2346083129935015e-06, + "loss": 0.5522, + "step": 17231 + }, + { + "epoch": 0.7629589453527114, + "grad_norm": 1.8185434341430664, + "learning_rate": 3.2334703134487464e-06, + "loss": 0.5914, + "step": 17232 + }, + { + "epoch": 0.7630032210575252, + "grad_norm": 1.8825651407241821, + "learning_rate": 3.2323324755170405e-06, + "loss": 0.5947, + "step": 17233 + }, + { + "epoch": 0.7630474967623391, + "grad_norm": 1.6178823709487915, + "learning_rate": 3.2311947992255665e-06, + "loss": 0.5146, + "step": 17234 + }, + { + "epoch": 0.763091772467153, + "grad_norm": 1.9216599464416504, + "learning_rate": 3.2300572846014945e-06, + "loss": 0.5131, + "step": 17235 + }, + { + "epoch": 0.7631360481719668, + "grad_norm": 1.4349368810653687, + "learning_rate": 3.22891993167199e-06, + "loss": 0.5621, + "step": 17236 + }, + { + "epoch": 0.7631803238767807, + "grad_norm": 1.4516592025756836, + "learning_rate": 3.2277827404642205e-06, + "loss": 0.4641, + "step": 17237 + }, + { + "epoch": 0.7632245995815946, + "grad_norm": 1.5849977731704712, + "learning_rate": 3.226645711005346e-06, + "loss": 0.5301, + "step": 17238 + }, + { + "epoch": 0.7632688752864084, + "grad_norm": 1.2924864292144775, + "learning_rate": 3.2255088433225246e-06, + "loss": 0.469, + "step": 17239 + }, + { + "epoch": 0.7633131509912223, + "grad_norm": 1.6423356533050537, + "learning_rate": 3.224372137442904e-06, + "loss": 0.5537, + "step": 17240 + }, + { + "epoch": 0.7633574266960362, + "grad_norm": 1.628353238105774, + "learning_rate": 3.223235593393642e-06, + "loss": 0.5227, + "step": 17241 + }, + { + "epoch": 0.7634017024008501, + "grad_norm": 1.632685661315918, + "learning_rate": 3.22209921120188e-06, + "loss": 0.5392, + "step": 17242 + }, + { + "epoch": 0.763445978105664, + "grad_norm": 1.7825101613998413, + "learning_rate": 3.220962990894756e-06, + "loss": 0.5869, + "step": 17243 + }, + { + "epoch": 0.7634902538104779, + "grad_norm": 1.4732999801635742, + "learning_rate": 3.2198269324994135e-06, + "loss": 0.5418, + "step": 17244 + }, + { + "epoch": 0.7635345295152918, + "grad_norm": 1.6940182447433472, + "learning_rate": 3.2186910360429855e-06, + "loss": 0.5233, + "step": 17245 + }, + { + "epoch": 0.7635788052201056, + "grad_norm": 1.3717540502548218, + "learning_rate": 3.2175553015526008e-06, + "loss": 0.4718, + "step": 17246 + }, + { + "epoch": 0.7636230809249195, + "grad_norm": 1.7463021278381348, + "learning_rate": 3.2164197290553846e-06, + "loss": 0.5545, + "step": 17247 + }, + { + "epoch": 0.7636673566297334, + "grad_norm": 1.5001004934310913, + "learning_rate": 3.2152843185784587e-06, + "loss": 0.5758, + "step": 17248 + }, + { + "epoch": 0.7637116323345472, + "grad_norm": 1.440089464187622, + "learning_rate": 3.214149070148944e-06, + "loss": 0.5032, + "step": 17249 + }, + { + "epoch": 0.7637559080393611, + "grad_norm": 1.570142388343811, + "learning_rate": 3.2130139837939502e-06, + "loss": 0.5365, + "step": 17250 + }, + { + "epoch": 0.763800183744175, + "grad_norm": 1.4277671575546265, + "learning_rate": 3.2118790595405934e-06, + "loss": 0.4943, + "step": 17251 + }, + { + "epoch": 0.7638444594489888, + "grad_norm": 1.2624746561050415, + "learning_rate": 3.2107442974159797e-06, + "loss": 0.4556, + "step": 17252 + }, + { + "epoch": 0.7638887351538027, + "grad_norm": 1.6386231184005737, + "learning_rate": 3.209609697447206e-06, + "loss": 0.493, + "step": 17253 + }, + { + "epoch": 0.7639330108586166, + "grad_norm": 1.536034345626831, + "learning_rate": 3.20847525966138e-06, + "loss": 0.5272, + "step": 17254 + }, + { + "epoch": 0.7639772865634304, + "grad_norm": 1.4931864738464355, + "learning_rate": 3.2073409840855916e-06, + "loss": 0.5344, + "step": 17255 + }, + { + "epoch": 0.7640215622682444, + "grad_norm": 1.2884219884872437, + "learning_rate": 3.2062068707469328e-06, + "loss": 0.4935, + "step": 17256 + }, + { + "epoch": 0.7640658379730583, + "grad_norm": 1.798494815826416, + "learning_rate": 3.2050729196724926e-06, + "loss": 0.5219, + "step": 17257 + }, + { + "epoch": 0.7641101136778721, + "grad_norm": 1.3250590562820435, + "learning_rate": 3.2039391308893518e-06, + "loss": 0.4855, + "step": 17258 + }, + { + "epoch": 0.764154389382686, + "grad_norm": 1.7481744289398193, + "learning_rate": 3.202805504424592e-06, + "loss": 0.4861, + "step": 17259 + }, + { + "epoch": 0.7641986650874999, + "grad_norm": 1.8025097846984863, + "learning_rate": 3.2016720403052846e-06, + "loss": 0.4863, + "step": 17260 + }, + { + "epoch": 0.7642429407923137, + "grad_norm": 1.5505425930023193, + "learning_rate": 3.2005387385585083e-06, + "loss": 0.5013, + "step": 17261 + }, + { + "epoch": 0.7642872164971276, + "grad_norm": 1.584982991218567, + "learning_rate": 3.1994055992113283e-06, + "loss": 0.5079, + "step": 17262 + }, + { + "epoch": 0.7643314922019415, + "grad_norm": 1.513339638710022, + "learning_rate": 3.1982726222908046e-06, + "loss": 0.5031, + "step": 17263 + }, + { + "epoch": 0.7643757679067553, + "grad_norm": 1.3066061735153198, + "learning_rate": 3.1971398078240046e-06, + "loss": 0.4851, + "step": 17264 + }, + { + "epoch": 0.7644200436115692, + "grad_norm": 1.3901097774505615, + "learning_rate": 3.1960071558379813e-06, + "loss": 0.5052, + "step": 17265 + }, + { + "epoch": 0.7644643193163831, + "grad_norm": 1.8154655694961548, + "learning_rate": 3.1948746663597873e-06, + "loss": 0.6033, + "step": 17266 + }, + { + "epoch": 0.764508595021197, + "grad_norm": 1.631883144378662, + "learning_rate": 3.19374233941647e-06, + "loss": 0.5382, + "step": 17267 + }, + { + "epoch": 0.7645528707260109, + "grad_norm": 1.2660107612609863, + "learning_rate": 3.192610175035077e-06, + "loss": 0.4865, + "step": 17268 + }, + { + "epoch": 0.7645971464308248, + "grad_norm": 1.3060314655303955, + "learning_rate": 3.1914781732426446e-06, + "loss": 0.5213, + "step": 17269 + }, + { + "epoch": 0.7646414221356386, + "grad_norm": 1.4064971208572388, + "learning_rate": 3.190346334066211e-06, + "loss": 0.4916, + "step": 17270 + }, + { + "epoch": 0.7646856978404525, + "grad_norm": 1.730334758758545, + "learning_rate": 3.189214657532813e-06, + "loss": 0.5464, + "step": 17271 + }, + { + "epoch": 0.7647299735452664, + "grad_norm": 1.3139649629592896, + "learning_rate": 3.1880831436694783e-06, + "loss": 0.4701, + "step": 17272 + }, + { + "epoch": 0.7647742492500802, + "grad_norm": 1.2956327199935913, + "learning_rate": 3.1869517925032267e-06, + "loss": 0.4906, + "step": 17273 + }, + { + "epoch": 0.7648185249548941, + "grad_norm": 1.4695277214050293, + "learning_rate": 3.1858206040610883e-06, + "loss": 0.523, + "step": 17274 + }, + { + "epoch": 0.764862800659708, + "grad_norm": 1.5399376153945923, + "learning_rate": 3.184689578370079e-06, + "loss": 0.4934, + "step": 17275 + }, + { + "epoch": 0.7649070763645219, + "grad_norm": 1.6215721368789673, + "learning_rate": 3.1835587154572036e-06, + "loss": 0.4844, + "step": 17276 + }, + { + "epoch": 0.7649513520693357, + "grad_norm": 1.6246452331542969, + "learning_rate": 3.1824280153494823e-06, + "loss": 0.5444, + "step": 17277 + }, + { + "epoch": 0.7649956277741496, + "grad_norm": 1.5278503894805908, + "learning_rate": 3.181297478073916e-06, + "loss": 0.5477, + "step": 17278 + }, + { + "epoch": 0.7650399034789636, + "grad_norm": 1.5094741582870483, + "learning_rate": 3.1801671036575066e-06, + "loss": 0.557, + "step": 17279 + }, + { + "epoch": 0.7650841791837774, + "grad_norm": 1.7189559936523438, + "learning_rate": 3.1790368921272506e-06, + "loss": 0.5782, + "step": 17280 + }, + { + "epoch": 0.7651284548885913, + "grad_norm": 1.8652507066726685, + "learning_rate": 3.177906843510148e-06, + "loss": 0.5758, + "step": 17281 + }, + { + "epoch": 0.7651727305934052, + "grad_norm": 1.814482569694519, + "learning_rate": 3.1767769578331865e-06, + "loss": 0.5454, + "step": 17282 + }, + { + "epoch": 0.765217006298219, + "grad_norm": 1.815729022026062, + "learning_rate": 3.175647235123347e-06, + "loss": 0.5501, + "step": 17283 + }, + { + "epoch": 0.7652612820030329, + "grad_norm": 1.7892011404037476, + "learning_rate": 3.1745176754076233e-06, + "loss": 0.5433, + "step": 17284 + }, + { + "epoch": 0.7653055577078468, + "grad_norm": 1.5759605169296265, + "learning_rate": 3.1733882787129853e-06, + "loss": 0.5165, + "step": 17285 + }, + { + "epoch": 0.7653498334126606, + "grad_norm": 1.2995167970657349, + "learning_rate": 3.172259045066409e-06, + "loss": 0.4728, + "step": 17286 + }, + { + "epoch": 0.7653941091174745, + "grad_norm": 1.8787552118301392, + "learning_rate": 3.171129974494863e-06, + "loss": 0.5864, + "step": 17287 + }, + { + "epoch": 0.7654383848222884, + "grad_norm": 1.3597412109375, + "learning_rate": 3.1700010670253202e-06, + "loss": 0.4616, + "step": 17288 + }, + { + "epoch": 0.7654826605271022, + "grad_norm": 1.4745343923568726, + "learning_rate": 3.1688723226847406e-06, + "loss": 0.5417, + "step": 17289 + }, + { + "epoch": 0.7655269362319161, + "grad_norm": 1.4523712396621704, + "learning_rate": 3.167743741500081e-06, + "loss": 0.4896, + "step": 17290 + }, + { + "epoch": 0.7655712119367301, + "grad_norm": 1.7194668054580688, + "learning_rate": 3.166615323498302e-06, + "loss": 0.5426, + "step": 17291 + }, + { + "epoch": 0.7656154876415439, + "grad_norm": 1.4828364849090576, + "learning_rate": 3.1654870687063523e-06, + "loss": 0.5297, + "step": 17292 + }, + { + "epoch": 0.7656597633463578, + "grad_norm": 1.6836761236190796, + "learning_rate": 3.1643589771511783e-06, + "loss": 0.4863, + "step": 17293 + }, + { + "epoch": 0.7657040390511717, + "grad_norm": 1.528259515762329, + "learning_rate": 3.163231048859725e-06, + "loss": 0.5412, + "step": 17294 + }, + { + "epoch": 0.7657483147559855, + "grad_norm": 1.6700530052185059, + "learning_rate": 3.1621032838589307e-06, + "loss": 0.5482, + "step": 17295 + }, + { + "epoch": 0.7657925904607994, + "grad_norm": 1.5100047588348389, + "learning_rate": 3.160975682175732e-06, + "loss": 0.5601, + "step": 17296 + }, + { + "epoch": 0.7658368661656133, + "grad_norm": 1.6299117803573608, + "learning_rate": 3.1598482438370583e-06, + "loss": 0.5662, + "step": 17297 + }, + { + "epoch": 0.7658811418704271, + "grad_norm": 1.704529047012329, + "learning_rate": 3.158720968869843e-06, + "loss": 0.553, + "step": 17298 + }, + { + "epoch": 0.765925417575241, + "grad_norm": 1.7120035886764526, + "learning_rate": 3.1575938573010056e-06, + "loss": 0.5898, + "step": 17299 + }, + { + "epoch": 0.7659696932800549, + "grad_norm": 1.8031127452850342, + "learning_rate": 3.156466909157465e-06, + "loss": 0.5311, + "step": 17300 + }, + { + "epoch": 0.7660139689848687, + "grad_norm": 1.3223228454589844, + "learning_rate": 3.155340124466144e-06, + "loss": 0.4485, + "step": 17301 + }, + { + "epoch": 0.7660582446896826, + "grad_norm": 1.3030143976211548, + "learning_rate": 3.1542135032539513e-06, + "loss": 0.4794, + "step": 17302 + }, + { + "epoch": 0.7661025203944966, + "grad_norm": 1.4425888061523438, + "learning_rate": 3.1530870455477934e-06, + "loss": 0.5341, + "step": 17303 + }, + { + "epoch": 0.7661467960993104, + "grad_norm": 1.5931861400604248, + "learning_rate": 3.1519607513745775e-06, + "loss": 0.538, + "step": 17304 + }, + { + "epoch": 0.7661910718041243, + "grad_norm": 1.99331533908844, + "learning_rate": 3.150834620761203e-06, + "loss": 0.5794, + "step": 17305 + }, + { + "epoch": 0.7662353475089382, + "grad_norm": 1.3437937498092651, + "learning_rate": 3.1497086537345665e-06, + "loss": 0.4766, + "step": 17306 + }, + { + "epoch": 0.766279623213752, + "grad_norm": 1.5339436531066895, + "learning_rate": 3.1485828503215588e-06, + "loss": 0.512, + "step": 17307 + }, + { + "epoch": 0.7663238989185659, + "grad_norm": 1.2499436140060425, + "learning_rate": 3.147457210549073e-06, + "loss": 0.4595, + "step": 17308 + }, + { + "epoch": 0.7663681746233798, + "grad_norm": 1.5231945514678955, + "learning_rate": 3.1463317344439935e-06, + "loss": 0.5048, + "step": 17309 + }, + { + "epoch": 0.7664124503281937, + "grad_norm": 1.3991219997406006, + "learning_rate": 3.145206422033196e-06, + "loss": 0.5095, + "step": 17310 + }, + { + "epoch": 0.7664567260330075, + "grad_norm": 1.376930832862854, + "learning_rate": 3.1440812733435657e-06, + "loss": 0.4632, + "step": 17311 + }, + { + "epoch": 0.7665010017378214, + "grad_norm": 1.5501155853271484, + "learning_rate": 3.1429562884019706e-06, + "loss": 0.5407, + "step": 17312 + }, + { + "epoch": 0.7665452774426353, + "grad_norm": 1.4558250904083252, + "learning_rate": 3.1418314672352825e-06, + "loss": 0.5096, + "step": 17313 + }, + { + "epoch": 0.7665895531474491, + "grad_norm": 1.9268784523010254, + "learning_rate": 3.140706809870365e-06, + "loss": 0.4933, + "step": 17314 + }, + { + "epoch": 0.766633828852263, + "grad_norm": 1.2809386253356934, + "learning_rate": 3.13958231633408e-06, + "loss": 0.4531, + "step": 17315 + }, + { + "epoch": 0.766678104557077, + "grad_norm": 1.4138528108596802, + "learning_rate": 3.138457986653286e-06, + "loss": 0.4989, + "step": 17316 + }, + { + "epoch": 0.7667223802618908, + "grad_norm": 2.1277544498443604, + "learning_rate": 3.137333820854833e-06, + "loss": 0.5871, + "step": 17317 + }, + { + "epoch": 0.7667666559667047, + "grad_norm": 1.338747501373291, + "learning_rate": 3.1362098189655777e-06, + "loss": 0.4783, + "step": 17318 + }, + { + "epoch": 0.7668109316715186, + "grad_norm": 1.8839023113250732, + "learning_rate": 3.1350859810123613e-06, + "loss": 0.6268, + "step": 17319 + }, + { + "epoch": 0.7668552073763324, + "grad_norm": 2.2025630474090576, + "learning_rate": 3.1339623070220226e-06, + "loss": 0.6455, + "step": 17320 + }, + { + "epoch": 0.7668994830811463, + "grad_norm": 1.6332207918167114, + "learning_rate": 3.132838797021408e-06, + "loss": 0.6214, + "step": 17321 + }, + { + "epoch": 0.7669437587859602, + "grad_norm": 1.4674572944641113, + "learning_rate": 3.1317154510373472e-06, + "loss": 0.5364, + "step": 17322 + }, + { + "epoch": 0.766988034490774, + "grad_norm": 1.340457558631897, + "learning_rate": 3.1305922690966705e-06, + "loss": 0.5016, + "step": 17323 + }, + { + "epoch": 0.7670323101955879, + "grad_norm": 1.4572911262512207, + "learning_rate": 3.1294692512262038e-06, + "loss": 0.4959, + "step": 17324 + }, + { + "epoch": 0.7670765859004018, + "grad_norm": 1.5365700721740723, + "learning_rate": 3.1283463974527707e-06, + "loss": 0.4754, + "step": 17325 + }, + { + "epoch": 0.7671208616052156, + "grad_norm": 1.2896099090576172, + "learning_rate": 3.1272237078031876e-06, + "loss": 0.4462, + "step": 17326 + }, + { + "epoch": 0.7671651373100296, + "grad_norm": 1.3923561573028564, + "learning_rate": 3.1261011823042676e-06, + "loss": 0.5076, + "step": 17327 + }, + { + "epoch": 0.7672094130148435, + "grad_norm": 1.6404353380203247, + "learning_rate": 3.124978820982828e-06, + "loss": 0.5176, + "step": 17328 + }, + { + "epoch": 0.7672536887196573, + "grad_norm": 1.6518049240112305, + "learning_rate": 3.12385662386567e-06, + "loss": 0.5796, + "step": 17329 + }, + { + "epoch": 0.7672979644244712, + "grad_norm": 1.4467436075210571, + "learning_rate": 3.1227345909795945e-06, + "loss": 0.5185, + "step": 17330 + }, + { + "epoch": 0.7673422401292851, + "grad_norm": 1.2782132625579834, + "learning_rate": 3.1216127223514116e-06, + "loss": 0.4581, + "step": 17331 + }, + { + "epoch": 0.7673865158340989, + "grad_norm": 1.5142627954483032, + "learning_rate": 3.120491018007904e-06, + "loss": 0.5176, + "step": 17332 + }, + { + "epoch": 0.7674307915389128, + "grad_norm": 1.6780396699905396, + "learning_rate": 3.119369477975863e-06, + "loss": 0.5461, + "step": 17333 + }, + { + "epoch": 0.7674750672437267, + "grad_norm": 1.541317105293274, + "learning_rate": 3.1182481022820843e-06, + "loss": 0.5142, + "step": 17334 + }, + { + "epoch": 0.7675193429485405, + "grad_norm": 1.4235353469848633, + "learning_rate": 3.117126890953346e-06, + "loss": 0.4664, + "step": 17335 + }, + { + "epoch": 0.7675636186533544, + "grad_norm": 1.554402470588684, + "learning_rate": 3.1160058440164264e-06, + "loss": 0.543, + "step": 17336 + }, + { + "epoch": 0.7676078943581683, + "grad_norm": 1.5837920904159546, + "learning_rate": 3.1148849614981003e-06, + "loss": 0.4636, + "step": 17337 + }, + { + "epoch": 0.7676521700629821, + "grad_norm": 1.4106404781341553, + "learning_rate": 3.1137642434251438e-06, + "loss": 0.5028, + "step": 17338 + }, + { + "epoch": 0.767696445767796, + "grad_norm": 1.6109766960144043, + "learning_rate": 3.1126436898243207e-06, + "loss": 0.5271, + "step": 17339 + }, + { + "epoch": 0.76774072147261, + "grad_norm": 1.705572247505188, + "learning_rate": 3.1115233007223954e-06, + "loss": 0.5445, + "step": 17340 + }, + { + "epoch": 0.7677849971774238, + "grad_norm": 1.3788280487060547, + "learning_rate": 3.110403076146128e-06, + "loss": 0.5163, + "step": 17341 + }, + { + "epoch": 0.7678292728822377, + "grad_norm": 1.524856686592102, + "learning_rate": 3.109283016122272e-06, + "loss": 0.5406, + "step": 17342 + }, + { + "epoch": 0.7678735485870516, + "grad_norm": 1.6056714057922363, + "learning_rate": 3.1081631206775787e-06, + "loss": 0.5229, + "step": 17343 + }, + { + "epoch": 0.7679178242918654, + "grad_norm": 1.5783010721206665, + "learning_rate": 3.1070433898387996e-06, + "loss": 0.535, + "step": 17344 + }, + { + "epoch": 0.7679620999966793, + "grad_norm": 1.6582393646240234, + "learning_rate": 3.1059238236326785e-06, + "loss": 0.5094, + "step": 17345 + }, + { + "epoch": 0.7680063757014932, + "grad_norm": 1.6783208847045898, + "learning_rate": 3.104804422085952e-06, + "loss": 0.5874, + "step": 17346 + }, + { + "epoch": 0.7680506514063071, + "grad_norm": 1.9136885404586792, + "learning_rate": 3.1036851852253548e-06, + "loss": 0.5597, + "step": 17347 + }, + { + "epoch": 0.7680949271111209, + "grad_norm": 1.6030281782150269, + "learning_rate": 3.1025661130776254e-06, + "loss": 0.5501, + "step": 17348 + }, + { + "epoch": 0.7681392028159348, + "grad_norm": 1.266682744026184, + "learning_rate": 3.1014472056694877e-06, + "loss": 0.4772, + "step": 17349 + }, + { + "epoch": 0.7681834785207488, + "grad_norm": 1.4295727014541626, + "learning_rate": 3.1003284630276664e-06, + "loss": 0.4593, + "step": 17350 + }, + { + "epoch": 0.7682277542255626, + "grad_norm": 2.6513235569000244, + "learning_rate": 3.099209885178882e-06, + "loss": 0.5886, + "step": 17351 + }, + { + "epoch": 0.7682720299303765, + "grad_norm": 1.7692301273345947, + "learning_rate": 3.098091472149851e-06, + "loss": 0.5703, + "step": 17352 + }, + { + "epoch": 0.7683163056351904, + "grad_norm": 1.4756070375442505, + "learning_rate": 3.096973223967282e-06, + "loss": 0.5045, + "step": 17353 + }, + { + "epoch": 0.7683605813400042, + "grad_norm": 2.105327844619751, + "learning_rate": 3.0958551406578894e-06, + "loss": 0.5453, + "step": 17354 + }, + { + "epoch": 0.7684048570448181, + "grad_norm": 1.672001838684082, + "learning_rate": 3.0947372222483762e-06, + "loss": 0.569, + "step": 17355 + }, + { + "epoch": 0.768449132749632, + "grad_norm": 1.4780491590499878, + "learning_rate": 3.093619468765442e-06, + "loss": 0.5517, + "step": 17356 + }, + { + "epoch": 0.7684934084544458, + "grad_norm": 1.2551378011703491, + "learning_rate": 3.0925018802357797e-06, + "loss": 0.4397, + "step": 17357 + }, + { + "epoch": 0.7685376841592597, + "grad_norm": 1.4529383182525635, + "learning_rate": 3.0913844566860896e-06, + "loss": 0.5019, + "step": 17358 + }, + { + "epoch": 0.7685819598640736, + "grad_norm": 1.1723142862319946, + "learning_rate": 3.0902671981430555e-06, + "loss": 0.4414, + "step": 17359 + }, + { + "epoch": 0.7686262355688874, + "grad_norm": 1.841665267944336, + "learning_rate": 3.0891501046333636e-06, + "loss": 0.5993, + "step": 17360 + }, + { + "epoch": 0.7686705112737013, + "grad_norm": 1.4975942373275757, + "learning_rate": 3.0880331761836947e-06, + "loss": 0.4248, + "step": 17361 + }, + { + "epoch": 0.7687147869785153, + "grad_norm": 1.3851547241210938, + "learning_rate": 3.0869164128207252e-06, + "loss": 0.495, + "step": 17362 + }, + { + "epoch": 0.768759062683329, + "grad_norm": 1.452274203300476, + "learning_rate": 3.0857998145711252e-06, + "loss": 0.5136, + "step": 17363 + }, + { + "epoch": 0.768803338388143, + "grad_norm": 1.6194874048233032, + "learning_rate": 3.0846833814615707e-06, + "loss": 0.5361, + "step": 17364 + }, + { + "epoch": 0.7688476140929569, + "grad_norm": 1.4833914041519165, + "learning_rate": 3.083567113518723e-06, + "loss": 0.4967, + "step": 17365 + }, + { + "epoch": 0.7688918897977707, + "grad_norm": 1.4555875062942505, + "learning_rate": 3.082451010769242e-06, + "loss": 0.5084, + "step": 17366 + }, + { + "epoch": 0.7689361655025846, + "grad_norm": 1.4957877397537231, + "learning_rate": 3.081335073239784e-06, + "loss": 0.5303, + "step": 17367 + }, + { + "epoch": 0.7689804412073985, + "grad_norm": 1.5147430896759033, + "learning_rate": 3.080219300957007e-06, + "loss": 0.5254, + "step": 17368 + }, + { + "epoch": 0.7690247169122123, + "grad_norm": 1.3927843570709229, + "learning_rate": 3.0791036939475584e-06, + "loss": 0.5329, + "step": 17369 + }, + { + "epoch": 0.7690689926170262, + "grad_norm": 1.4608690738677979, + "learning_rate": 3.077988252238082e-06, + "loss": 0.5072, + "step": 17370 + }, + { + "epoch": 0.7691132683218401, + "grad_norm": 1.4101629257202148, + "learning_rate": 3.076872975855221e-06, + "loss": 0.5325, + "step": 17371 + }, + { + "epoch": 0.7691575440266539, + "grad_norm": 1.4027485847473145, + "learning_rate": 3.0757578648256105e-06, + "loss": 0.4992, + "step": 17372 + }, + { + "epoch": 0.7692018197314678, + "grad_norm": 1.6206163167953491, + "learning_rate": 3.074642919175883e-06, + "loss": 0.5225, + "step": 17373 + }, + { + "epoch": 0.7692460954362818, + "grad_norm": 1.5082049369812012, + "learning_rate": 3.0735281389326744e-06, + "loss": 0.4813, + "step": 17374 + }, + { + "epoch": 0.7692903711410956, + "grad_norm": 1.9323930740356445, + "learning_rate": 3.0724135241226063e-06, + "loss": 0.5189, + "step": 17375 + }, + { + "epoch": 0.7693346468459095, + "grad_norm": 1.8151062726974487, + "learning_rate": 3.0712990747723005e-06, + "loss": 0.6034, + "step": 17376 + }, + { + "epoch": 0.7693789225507234, + "grad_norm": 1.4390122890472412, + "learning_rate": 3.0701847909083704e-06, + "loss": 0.4987, + "step": 17377 + }, + { + "epoch": 0.7694231982555372, + "grad_norm": 1.847355604171753, + "learning_rate": 3.0690706725574416e-06, + "loss": 0.5515, + "step": 17378 + }, + { + "epoch": 0.7694674739603511, + "grad_norm": 1.5586861371994019, + "learning_rate": 3.0679567197461135e-06, + "loss": 0.5376, + "step": 17379 + }, + { + "epoch": 0.769511749665165, + "grad_norm": 1.4597604274749756, + "learning_rate": 3.0668429325009907e-06, + "loss": 0.518, + "step": 17380 + }, + { + "epoch": 0.7695560253699789, + "grad_norm": 1.429585337638855, + "learning_rate": 3.065729310848683e-06, + "loss": 0.4891, + "step": 17381 + }, + { + "epoch": 0.7696003010747927, + "grad_norm": 1.5447856187820435, + "learning_rate": 3.0646158548157846e-06, + "loss": 0.5151, + "step": 17382 + }, + { + "epoch": 0.7696445767796066, + "grad_norm": 1.2980997562408447, + "learning_rate": 3.063502564428886e-06, + "loss": 0.4625, + "step": 17383 + }, + { + "epoch": 0.7696888524844205, + "grad_norm": 1.5651482343673706, + "learning_rate": 3.0623894397145837e-06, + "loss": 0.5525, + "step": 17384 + }, + { + "epoch": 0.7697331281892343, + "grad_norm": 1.5541613101959229, + "learning_rate": 3.061276480699462e-06, + "loss": 0.5245, + "step": 17385 + }, + { + "epoch": 0.7697774038940483, + "grad_norm": 1.4093581438064575, + "learning_rate": 3.0601636874101004e-06, + "loss": 0.5062, + "step": 17386 + }, + { + "epoch": 0.7698216795988622, + "grad_norm": 1.3976317644119263, + "learning_rate": 3.0590510598730793e-06, + "loss": 0.5614, + "step": 17387 + }, + { + "epoch": 0.769865955303676, + "grad_norm": 1.4050617218017578, + "learning_rate": 3.0579385981149725e-06, + "loss": 0.5252, + "step": 17388 + }, + { + "epoch": 0.7699102310084899, + "grad_norm": 1.2725486755371094, + "learning_rate": 3.056826302162349e-06, + "loss": 0.4923, + "step": 17389 + }, + { + "epoch": 0.7699545067133038, + "grad_norm": 1.5865213871002197, + "learning_rate": 3.055714172041774e-06, + "loss": 0.5244, + "step": 17390 + }, + { + "epoch": 0.7699987824181176, + "grad_norm": 2.4078116416931152, + "learning_rate": 3.0546022077798133e-06, + "loss": 0.5533, + "step": 17391 + }, + { + "epoch": 0.7700430581229315, + "grad_norm": 1.3972890377044678, + "learning_rate": 3.053490409403025e-06, + "loss": 0.4754, + "step": 17392 + }, + { + "epoch": 0.7700873338277454, + "grad_norm": 1.480017900466919, + "learning_rate": 3.052378776937959e-06, + "loss": 0.5733, + "step": 17393 + }, + { + "epoch": 0.7701316095325592, + "grad_norm": 1.3266854286193848, + "learning_rate": 3.051267310411171e-06, + "loss": 0.4325, + "step": 17394 + }, + { + "epoch": 0.7701758852373731, + "grad_norm": 1.8146283626556396, + "learning_rate": 3.0501560098492056e-06, + "loss": 0.6007, + "step": 17395 + }, + { + "epoch": 0.770220160942187, + "grad_norm": 1.5813281536102295, + "learning_rate": 3.0490448752786063e-06, + "loss": 0.4996, + "step": 17396 + }, + { + "epoch": 0.7702644366470008, + "grad_norm": 1.5032418966293335, + "learning_rate": 3.0479339067259094e-06, + "loss": 0.486, + "step": 17397 + }, + { + "epoch": 0.7703087123518148, + "grad_norm": 1.3546351194381714, + "learning_rate": 3.046823104217651e-06, + "loss": 0.4837, + "step": 17398 + }, + { + "epoch": 0.7703529880566287, + "grad_norm": 1.5101498365402222, + "learning_rate": 3.0457124677803616e-06, + "loss": 0.5351, + "step": 17399 + }, + { + "epoch": 0.7703972637614425, + "grad_norm": 2.510279893875122, + "learning_rate": 3.044601997440564e-06, + "loss": 0.5447, + "step": 17400 + }, + { + "epoch": 0.7704415394662564, + "grad_norm": 1.8230129480361938, + "learning_rate": 3.0434916932247873e-06, + "loss": 0.5422, + "step": 17401 + }, + { + "epoch": 0.7704858151710703, + "grad_norm": 1.3491041660308838, + "learning_rate": 3.042381555159547e-06, + "loss": 0.4644, + "step": 17402 + }, + { + "epoch": 0.7705300908758841, + "grad_norm": 1.2765980958938599, + "learning_rate": 3.0412715832713592e-06, + "loss": 0.4441, + "step": 17403 + }, + { + "epoch": 0.770574366580698, + "grad_norm": 1.7409374713897705, + "learning_rate": 3.04016177758673e-06, + "loss": 0.5277, + "step": 17404 + }, + { + "epoch": 0.7706186422855119, + "grad_norm": 1.5253307819366455, + "learning_rate": 3.0390521381321724e-06, + "loss": 0.5009, + "step": 17405 + }, + { + "epoch": 0.7706629179903257, + "grad_norm": 1.7077045440673828, + "learning_rate": 3.037942664934187e-06, + "loss": 0.5315, + "step": 17406 + }, + { + "epoch": 0.7707071936951396, + "grad_norm": 1.2969154119491577, + "learning_rate": 3.0368333580192734e-06, + "loss": 0.4766, + "step": 17407 + }, + { + "epoch": 0.7707514693999535, + "grad_norm": 1.4274228811264038, + "learning_rate": 3.0357242174139243e-06, + "loss": 0.4845, + "step": 17408 + }, + { + "epoch": 0.7707957451047673, + "grad_norm": 1.3877999782562256, + "learning_rate": 3.0346152431446318e-06, + "loss": 0.5177, + "step": 17409 + }, + { + "epoch": 0.7708400208095813, + "grad_norm": 1.3860387802124023, + "learning_rate": 3.0335064352378795e-06, + "loss": 0.5551, + "step": 17410 + }, + { + "epoch": 0.7708842965143952, + "grad_norm": 1.2223799228668213, + "learning_rate": 3.0323977937201575e-06, + "loss": 0.4766, + "step": 17411 + }, + { + "epoch": 0.770928572219209, + "grad_norm": 1.5920921564102173, + "learning_rate": 3.0312893186179403e-06, + "loss": 0.4842, + "step": 17412 + }, + { + "epoch": 0.7709728479240229, + "grad_norm": 1.6767406463623047, + "learning_rate": 3.0301810099577044e-06, + "loss": 0.5726, + "step": 17413 + }, + { + "epoch": 0.7710171236288368, + "grad_norm": 1.2866063117980957, + "learning_rate": 3.0290728677659163e-06, + "loss": 0.4523, + "step": 17414 + }, + { + "epoch": 0.7710613993336506, + "grad_norm": 1.997468113899231, + "learning_rate": 3.0279648920690495e-06, + "loss": 0.5687, + "step": 17415 + }, + { + "epoch": 0.7711056750384645, + "grad_norm": 1.7423300743103027, + "learning_rate": 3.026857082893565e-06, + "loss": 0.5647, + "step": 17416 + }, + { + "epoch": 0.7711499507432784, + "grad_norm": 1.6340042352676392, + "learning_rate": 3.0257494402659216e-06, + "loss": 0.5566, + "step": 17417 + }, + { + "epoch": 0.7711942264480923, + "grad_norm": 1.9113478660583496, + "learning_rate": 3.0246419642125748e-06, + "loss": 0.5343, + "step": 17418 + }, + { + "epoch": 0.7712385021529061, + "grad_norm": 1.476937174797058, + "learning_rate": 3.0235346547599733e-06, + "loss": 0.4876, + "step": 17419 + }, + { + "epoch": 0.77128277785772, + "grad_norm": 1.4270274639129639, + "learning_rate": 3.0224275119345637e-06, + "loss": 0.4748, + "step": 17420 + }, + { + "epoch": 0.771327053562534, + "grad_norm": 1.689926266670227, + "learning_rate": 3.021320535762795e-06, + "loss": 0.5029, + "step": 17421 + }, + { + "epoch": 0.7713713292673477, + "grad_norm": 1.4692622423171997, + "learning_rate": 3.020213726271103e-06, + "loss": 0.4841, + "step": 17422 + }, + { + "epoch": 0.7714156049721617, + "grad_norm": 1.3163880109786987, + "learning_rate": 3.0191070834859237e-06, + "loss": 0.4913, + "step": 17423 + }, + { + "epoch": 0.7714598806769756, + "grad_norm": 1.610839605331421, + "learning_rate": 3.0180006074336833e-06, + "loss": 0.5354, + "step": 17424 + }, + { + "epoch": 0.7715041563817894, + "grad_norm": 1.5020591020584106, + "learning_rate": 3.0168942981408177e-06, + "loss": 0.5103, + "step": 17425 + }, + { + "epoch": 0.7715484320866033, + "grad_norm": 1.271619200706482, + "learning_rate": 3.0157881556337474e-06, + "loss": 0.4607, + "step": 17426 + }, + { + "epoch": 0.7715927077914172, + "grad_norm": 1.4657039642333984, + "learning_rate": 3.014682179938886e-06, + "loss": 0.4775, + "step": 17427 + }, + { + "epoch": 0.771636983496231, + "grad_norm": 1.3009884357452393, + "learning_rate": 3.013576371082655e-06, + "loss": 0.4371, + "step": 17428 + }, + { + "epoch": 0.7716812592010449, + "grad_norm": 1.3443291187286377, + "learning_rate": 3.0124707290914647e-06, + "loss": 0.4503, + "step": 17429 + }, + { + "epoch": 0.7717255349058588, + "grad_norm": 1.4996017217636108, + "learning_rate": 3.011365253991717e-06, + "loss": 0.5, + "step": 17430 + }, + { + "epoch": 0.7717698106106726, + "grad_norm": 1.26675546169281, + "learning_rate": 3.0102599458098237e-06, + "loss": 0.4691, + "step": 17431 + }, + { + "epoch": 0.7718140863154865, + "grad_norm": 1.3688827753067017, + "learning_rate": 3.009154804572181e-06, + "loss": 0.4893, + "step": 17432 + }, + { + "epoch": 0.7718583620203004, + "grad_norm": 1.5992940664291382, + "learning_rate": 3.008049830305183e-06, + "loss": 0.5499, + "step": 17433 + }, + { + "epoch": 0.7719026377251142, + "grad_norm": 1.4793214797973633, + "learning_rate": 3.0069450230352217e-06, + "loss": 0.4846, + "step": 17434 + }, + { + "epoch": 0.7719469134299282, + "grad_norm": 1.4904452562332153, + "learning_rate": 3.005840382788685e-06, + "loss": 0.5324, + "step": 17435 + }, + { + "epoch": 0.7719911891347421, + "grad_norm": 1.5449188947677612, + "learning_rate": 3.004735909591955e-06, + "loss": 0.4991, + "step": 17436 + }, + { + "epoch": 0.7720354648395559, + "grad_norm": 1.4656153917312622, + "learning_rate": 3.0036316034714097e-06, + "loss": 0.4839, + "step": 17437 + }, + { + "epoch": 0.7720797405443698, + "grad_norm": 1.4314683675765991, + "learning_rate": 3.0025274644534296e-06, + "loss": 0.4703, + "step": 17438 + }, + { + "epoch": 0.7721240162491837, + "grad_norm": 1.5242301225662231, + "learning_rate": 3.001423492564384e-06, + "loss": 0.474, + "step": 17439 + }, + { + "epoch": 0.7721682919539975, + "grad_norm": 1.6255874633789062, + "learning_rate": 3.0003196878306363e-06, + "loss": 0.544, + "step": 17440 + }, + { + "epoch": 0.7722125676588114, + "grad_norm": 1.5666790008544922, + "learning_rate": 2.999216050278556e-06, + "loss": 0.4808, + "step": 17441 + }, + { + "epoch": 0.7722568433636253, + "grad_norm": 1.5875697135925293, + "learning_rate": 2.998112579934501e-06, + "loss": 0.5354, + "step": 17442 + }, + { + "epoch": 0.7723011190684391, + "grad_norm": 1.8015949726104736, + "learning_rate": 2.997009276824826e-06, + "loss": 0.5885, + "step": 17443 + }, + { + "epoch": 0.772345394773253, + "grad_norm": 1.5245357751846313, + "learning_rate": 2.9959061409758817e-06, + "loss": 0.5199, + "step": 17444 + }, + { + "epoch": 0.772389670478067, + "grad_norm": 1.6917390823364258, + "learning_rate": 2.994803172414017e-06, + "loss": 0.5267, + "step": 17445 + }, + { + "epoch": 0.7724339461828807, + "grad_norm": 1.8071540594100952, + "learning_rate": 2.993700371165574e-06, + "loss": 0.5221, + "step": 17446 + }, + { + "epoch": 0.7724782218876947, + "grad_norm": 1.7126365900039673, + "learning_rate": 2.9925977372568904e-06, + "loss": 0.5008, + "step": 17447 + }, + { + "epoch": 0.7725224975925086, + "grad_norm": 1.5573722124099731, + "learning_rate": 2.9914952707143076e-06, + "loss": 0.5141, + "step": 17448 + }, + { + "epoch": 0.7725667732973224, + "grad_norm": 1.4347174167633057, + "learning_rate": 2.990392971564152e-06, + "loss": 0.5116, + "step": 17449 + }, + { + "epoch": 0.7726110490021363, + "grad_norm": 1.5779578685760498, + "learning_rate": 2.9892908398327514e-06, + "loss": 0.4827, + "step": 17450 + }, + { + "epoch": 0.7726553247069502, + "grad_norm": 1.3769258260726929, + "learning_rate": 2.9881888755464337e-06, + "loss": 0.4725, + "step": 17451 + }, + { + "epoch": 0.7726996004117641, + "grad_norm": 1.3713606595993042, + "learning_rate": 2.9870870787315155e-06, + "loss": 0.4849, + "step": 17452 + }, + { + "epoch": 0.7727438761165779, + "grad_norm": 1.8065108060836792, + "learning_rate": 2.9859854494143113e-06, + "loss": 0.5209, + "step": 17453 + }, + { + "epoch": 0.7727881518213918, + "grad_norm": 1.4550459384918213, + "learning_rate": 2.9848839876211344e-06, + "loss": 0.4789, + "step": 17454 + }, + { + "epoch": 0.7728324275262057, + "grad_norm": 1.5508215427398682, + "learning_rate": 2.9837826933782908e-06, + "loss": 0.5373, + "step": 17455 + }, + { + "epoch": 0.7728767032310195, + "grad_norm": 1.8046095371246338, + "learning_rate": 2.9826815667120854e-06, + "loss": 0.4888, + "step": 17456 + }, + { + "epoch": 0.7729209789358334, + "grad_norm": 1.7920957803726196, + "learning_rate": 2.981580607648814e-06, + "loss": 0.5325, + "step": 17457 + }, + { + "epoch": 0.7729652546406474, + "grad_norm": 1.2235181331634521, + "learning_rate": 2.9804798162147785e-06, + "loss": 0.4441, + "step": 17458 + }, + { + "epoch": 0.7730095303454612, + "grad_norm": 1.5455682277679443, + "learning_rate": 2.979379192436267e-06, + "loss": 0.5139, + "step": 17459 + }, + { + "epoch": 0.7730538060502751, + "grad_norm": 1.7451516389846802, + "learning_rate": 2.978278736339564e-06, + "loss": 0.5255, + "step": 17460 + }, + { + "epoch": 0.773098081755089, + "grad_norm": 1.4496617317199707, + "learning_rate": 2.977178447950958e-06, + "loss": 0.5039, + "step": 17461 + }, + { + "epoch": 0.7731423574599028, + "grad_norm": 1.466335654258728, + "learning_rate": 2.9760783272967275e-06, + "loss": 0.4911, + "step": 17462 + }, + { + "epoch": 0.7731866331647167, + "grad_norm": 1.4822099208831787, + "learning_rate": 2.974978374403147e-06, + "loss": 0.4999, + "step": 17463 + }, + { + "epoch": 0.7732309088695306, + "grad_norm": 1.3615933656692505, + "learning_rate": 2.9738785892964884e-06, + "loss": 0.4617, + "step": 17464 + }, + { + "epoch": 0.7732751845743444, + "grad_norm": 1.5449408292770386, + "learning_rate": 2.9727789720030185e-06, + "loss": 0.5268, + "step": 17465 + }, + { + "epoch": 0.7733194602791583, + "grad_norm": 1.4483659267425537, + "learning_rate": 2.9716795225490013e-06, + "loss": 0.4852, + "step": 17466 + }, + { + "epoch": 0.7733637359839722, + "grad_norm": 1.5398011207580566, + "learning_rate": 2.970580240960692e-06, + "loss": 0.4423, + "step": 17467 + }, + { + "epoch": 0.773408011688786, + "grad_norm": 1.4653613567352295, + "learning_rate": 2.9694811272643543e-06, + "loss": 0.4931, + "step": 17468 + }, + { + "epoch": 0.7734522873936, + "grad_norm": 1.615074634552002, + "learning_rate": 2.9683821814862357e-06, + "loss": 0.5058, + "step": 17469 + }, + { + "epoch": 0.7734965630984139, + "grad_norm": 1.8616421222686768, + "learning_rate": 2.9672834036525787e-06, + "loss": 0.4997, + "step": 17470 + }, + { + "epoch": 0.7735408388032277, + "grad_norm": 1.4885176420211792, + "learning_rate": 2.966184793789636e-06, + "loss": 0.5552, + "step": 17471 + }, + { + "epoch": 0.7735851145080416, + "grad_norm": 1.9910879135131836, + "learning_rate": 2.9650863519236415e-06, + "loss": 0.6271, + "step": 17472 + }, + { + "epoch": 0.7736293902128555, + "grad_norm": 1.690919280052185, + "learning_rate": 2.963988078080835e-06, + "loss": 0.5551, + "step": 17473 + }, + { + "epoch": 0.7736736659176693, + "grad_norm": 1.487299919128418, + "learning_rate": 2.9628899722874373e-06, + "loss": 0.5198, + "step": 17474 + }, + { + "epoch": 0.7737179416224832, + "grad_norm": 1.8842313289642334, + "learning_rate": 2.961792034569686e-06, + "loss": 0.5397, + "step": 17475 + }, + { + "epoch": 0.7737622173272971, + "grad_norm": 1.4698185920715332, + "learning_rate": 2.9606942649538017e-06, + "loss": 0.5544, + "step": 17476 + }, + { + "epoch": 0.7738064930321109, + "grad_norm": 1.7136075496673584, + "learning_rate": 2.959596663466e-06, + "loss": 0.6165, + "step": 17477 + }, + { + "epoch": 0.7738507687369248, + "grad_norm": 1.4819139242172241, + "learning_rate": 2.958499230132503e-06, + "loss": 0.5457, + "step": 17478 + }, + { + "epoch": 0.7738950444417387, + "grad_norm": 1.6804304122924805, + "learning_rate": 2.9574019649795184e-06, + "loss": 0.487, + "step": 17479 + }, + { + "epoch": 0.7739393201465525, + "grad_norm": 1.4272156953811646, + "learning_rate": 2.9563048680332505e-06, + "loss": 0.4435, + "step": 17480 + }, + { + "epoch": 0.7739835958513664, + "grad_norm": 1.8073350191116333, + "learning_rate": 2.9552079393199085e-06, + "loss": 0.5626, + "step": 17481 + }, + { + "epoch": 0.7740278715561804, + "grad_norm": 2.158005952835083, + "learning_rate": 2.9541111788656917e-06, + "loss": 0.6209, + "step": 17482 + }, + { + "epoch": 0.7740721472609942, + "grad_norm": 2.1274256706237793, + "learning_rate": 2.9530145866967897e-06, + "loss": 0.5538, + "step": 17483 + }, + { + "epoch": 0.7741164229658081, + "grad_norm": 1.956209421157837, + "learning_rate": 2.9519181628393932e-06, + "loss": 0.5881, + "step": 17484 + }, + { + "epoch": 0.774160698670622, + "grad_norm": 1.6738301515579224, + "learning_rate": 2.9508219073196954e-06, + "loss": 0.5464, + "step": 17485 + }, + { + "epoch": 0.7742049743754358, + "grad_norm": 1.5014187097549438, + "learning_rate": 2.9497258201638767e-06, + "loss": 0.5113, + "step": 17486 + }, + { + "epoch": 0.7742492500802497, + "grad_norm": 1.3660187721252441, + "learning_rate": 2.9486299013981135e-06, + "loss": 0.52, + "step": 17487 + }, + { + "epoch": 0.7742935257850636, + "grad_norm": 1.428053855895996, + "learning_rate": 2.9475341510485867e-06, + "loss": 0.5015, + "step": 17488 + }, + { + "epoch": 0.7743378014898775, + "grad_norm": 1.4584742784500122, + "learning_rate": 2.946438569141464e-06, + "loss": 0.5276, + "step": 17489 + }, + { + "epoch": 0.7743820771946913, + "grad_norm": 1.4717046022415161, + "learning_rate": 2.945343155702913e-06, + "loss": 0.5277, + "step": 17490 + }, + { + "epoch": 0.7744263528995052, + "grad_norm": 1.4224579334259033, + "learning_rate": 2.944247910759097e-06, + "loss": 0.5167, + "step": 17491 + }, + { + "epoch": 0.7744706286043191, + "grad_norm": 1.5397512912750244, + "learning_rate": 2.943152834336175e-06, + "loss": 0.4997, + "step": 17492 + }, + { + "epoch": 0.774514904309133, + "grad_norm": 1.5336676836013794, + "learning_rate": 2.9420579264603012e-06, + "loss": 0.4948, + "step": 17493 + }, + { + "epoch": 0.7745591800139469, + "grad_norm": 1.598944067955017, + "learning_rate": 2.940963187157624e-06, + "loss": 0.6013, + "step": 17494 + }, + { + "epoch": 0.7746034557187608, + "grad_norm": 1.900707721710205, + "learning_rate": 2.9398686164542956e-06, + "loss": 0.612, + "step": 17495 + }, + { + "epoch": 0.7746477314235746, + "grad_norm": 1.3209519386291504, + "learning_rate": 2.9387742143764574e-06, + "loss": 0.4653, + "step": 17496 + }, + { + "epoch": 0.7746920071283885, + "grad_norm": 1.5779824256896973, + "learning_rate": 2.9376799809502444e-06, + "loss": 0.533, + "step": 17497 + }, + { + "epoch": 0.7747362828332024, + "grad_norm": 1.642482876777649, + "learning_rate": 2.9365859162017985e-06, + "loss": 0.5979, + "step": 17498 + }, + { + "epoch": 0.7747805585380162, + "grad_norm": 1.2924103736877441, + "learning_rate": 2.9354920201572457e-06, + "loss": 0.4786, + "step": 17499 + }, + { + "epoch": 0.7748248342428301, + "grad_norm": 1.293121576309204, + "learning_rate": 2.934398292842715e-06, + "loss": 0.4441, + "step": 17500 + }, + { + "epoch": 0.774869109947644, + "grad_norm": 1.5181077718734741, + "learning_rate": 2.933304734284327e-06, + "loss": 0.4939, + "step": 17501 + }, + { + "epoch": 0.7749133856524578, + "grad_norm": 1.9180712699890137, + "learning_rate": 2.9322113445082025e-06, + "loss": 0.4978, + "step": 17502 + }, + { + "epoch": 0.7749576613572717, + "grad_norm": 1.6908038854599, + "learning_rate": 2.931118123540454e-06, + "loss": 0.5325, + "step": 17503 + }, + { + "epoch": 0.7750019370620856, + "grad_norm": 1.278311848640442, + "learning_rate": 2.930025071407191e-06, + "loss": 0.5034, + "step": 17504 + }, + { + "epoch": 0.7750462127668994, + "grad_norm": 1.551806092262268, + "learning_rate": 2.9289321881345257e-06, + "loss": 0.4818, + "step": 17505 + }, + { + "epoch": 0.7750904884717134, + "grad_norm": 1.294763445854187, + "learning_rate": 2.927839473748557e-06, + "loss": 0.4704, + "step": 17506 + }, + { + "epoch": 0.7751347641765273, + "grad_norm": 1.7772401571273804, + "learning_rate": 2.9267469282753825e-06, + "loss": 0.5574, + "step": 17507 + }, + { + "epoch": 0.7751790398813411, + "grad_norm": 1.4478514194488525, + "learning_rate": 2.9256545517411007e-06, + "loss": 0.525, + "step": 17508 + }, + { + "epoch": 0.775223315586155, + "grad_norm": 1.2914865016937256, + "learning_rate": 2.9245623441717987e-06, + "loss": 0.4474, + "step": 17509 + }, + { + "epoch": 0.7752675912909689, + "grad_norm": 1.7111449241638184, + "learning_rate": 2.9234703055935653e-06, + "loss": 0.5348, + "step": 17510 + }, + { + "epoch": 0.7753118669957827, + "grad_norm": 1.6252838373184204, + "learning_rate": 2.9223784360324815e-06, + "loss": 0.5543, + "step": 17511 + }, + { + "epoch": 0.7753561427005966, + "grad_norm": 1.281114101409912, + "learning_rate": 2.9212867355146256e-06, + "loss": 0.4758, + "step": 17512 + }, + { + "epoch": 0.7754004184054105, + "grad_norm": 1.4727495908737183, + "learning_rate": 2.920195204066073e-06, + "loss": 0.5072, + "step": 17513 + }, + { + "epoch": 0.7754446941102243, + "grad_norm": 1.564568281173706, + "learning_rate": 2.919103841712889e-06, + "loss": 0.4989, + "step": 17514 + }, + { + "epoch": 0.7754889698150382, + "grad_norm": 1.4017301797866821, + "learning_rate": 2.918012648481149e-06, + "loss": 0.4569, + "step": 17515 + }, + { + "epoch": 0.7755332455198521, + "grad_norm": 1.753919005393982, + "learning_rate": 2.916921624396909e-06, + "loss": 0.4977, + "step": 17516 + }, + { + "epoch": 0.7755775212246659, + "grad_norm": 1.5979071855545044, + "learning_rate": 2.9158307694862264e-06, + "loss": 0.5272, + "step": 17517 + }, + { + "epoch": 0.7756217969294799, + "grad_norm": 1.526424765586853, + "learning_rate": 2.9147400837751614e-06, + "loss": 0.5348, + "step": 17518 + }, + { + "epoch": 0.7756660726342938, + "grad_norm": 1.5836570262908936, + "learning_rate": 2.9136495672897592e-06, + "loss": 0.534, + "step": 17519 + }, + { + "epoch": 0.7757103483391076, + "grad_norm": 1.472657322883606, + "learning_rate": 2.9125592200560682e-06, + "loss": 0.501, + "step": 17520 + }, + { + "epoch": 0.7757546240439215, + "grad_norm": 1.4894613027572632, + "learning_rate": 2.911469042100129e-06, + "loss": 0.4795, + "step": 17521 + }, + { + "epoch": 0.7757988997487354, + "grad_norm": 1.5887658596038818, + "learning_rate": 2.91037903344798e-06, + "loss": 0.523, + "step": 17522 + }, + { + "epoch": 0.7758431754535493, + "grad_norm": 1.4175957441329956, + "learning_rate": 2.909289194125655e-06, + "loss": 0.4956, + "step": 17523 + }, + { + "epoch": 0.7758874511583631, + "grad_norm": 1.6527239084243774, + "learning_rate": 2.908199524159182e-06, + "loss": 0.5224, + "step": 17524 + }, + { + "epoch": 0.775931726863177, + "grad_norm": 1.5126147270202637, + "learning_rate": 2.907110023574591e-06, + "loss": 0.5615, + "step": 17525 + }, + { + "epoch": 0.7759760025679909, + "grad_norm": 1.546862006187439, + "learning_rate": 2.906020692397903e-06, + "loss": 0.4639, + "step": 17526 + }, + { + "epoch": 0.7760202782728047, + "grad_norm": 1.1880083084106445, + "learning_rate": 2.9049315306551306e-06, + "loss": 0.4171, + "step": 17527 + }, + { + "epoch": 0.7760645539776186, + "grad_norm": 1.747220754623413, + "learning_rate": 2.9038425383722947e-06, + "loss": 0.5129, + "step": 17528 + }, + { + "epoch": 0.7761088296824326, + "grad_norm": 1.3633480072021484, + "learning_rate": 2.902753715575405e-06, + "loss": 0.469, + "step": 17529 + }, + { + "epoch": 0.7761531053872464, + "grad_norm": 1.4302395582199097, + "learning_rate": 2.901665062290461e-06, + "loss": 0.4807, + "step": 17530 + }, + { + "epoch": 0.7761973810920603, + "grad_norm": 1.4864825010299683, + "learning_rate": 2.9005765785434637e-06, + "loss": 0.5388, + "step": 17531 + }, + { + "epoch": 0.7762416567968742, + "grad_norm": 1.6883933544158936, + "learning_rate": 2.899488264360418e-06, + "loss": 0.5142, + "step": 17532 + }, + { + "epoch": 0.776285932501688, + "grad_norm": 1.5713133811950684, + "learning_rate": 2.8984001197673138e-06, + "loss": 0.5266, + "step": 17533 + }, + { + "epoch": 0.7763302082065019, + "grad_norm": 1.405192255973816, + "learning_rate": 2.8973121447901375e-06, + "loss": 0.4882, + "step": 17534 + }, + { + "epoch": 0.7763744839113158, + "grad_norm": 1.24518620967865, + "learning_rate": 2.8962243394548793e-06, + "loss": 0.4867, + "step": 17535 + }, + { + "epoch": 0.7764187596161296, + "grad_norm": 1.468550205230713, + "learning_rate": 2.89513670378752e-06, + "loss": 0.5233, + "step": 17536 + }, + { + "epoch": 0.7764630353209435, + "grad_norm": 1.500019907951355, + "learning_rate": 2.8940492378140316e-06, + "loss": 0.473, + "step": 17537 + }, + { + "epoch": 0.7765073110257574, + "grad_norm": 1.3040586709976196, + "learning_rate": 2.892961941560397e-06, + "loss": 0.5073, + "step": 17538 + }, + { + "epoch": 0.7765515867305712, + "grad_norm": 2.2186222076416016, + "learning_rate": 2.891874815052578e-06, + "loss": 0.5673, + "step": 17539 + }, + { + "epoch": 0.7765958624353851, + "grad_norm": 1.6586875915527344, + "learning_rate": 2.8907878583165406e-06, + "loss": 0.5315, + "step": 17540 + }, + { + "epoch": 0.776640138140199, + "grad_norm": 1.4732199907302856, + "learning_rate": 2.8897010713782436e-06, + "loss": 0.4962, + "step": 17541 + }, + { + "epoch": 0.7766844138450129, + "grad_norm": 1.3004463911056519, + "learning_rate": 2.8886144542636495e-06, + "loss": 0.437, + "step": 17542 + }, + { + "epoch": 0.7767286895498268, + "grad_norm": 1.5703672170639038, + "learning_rate": 2.88752800699871e-06, + "loss": 0.5416, + "step": 17543 + }, + { + "epoch": 0.7767729652546407, + "grad_norm": 1.3800331354141235, + "learning_rate": 2.8864417296093696e-06, + "loss": 0.4606, + "step": 17544 + }, + { + "epoch": 0.7768172409594545, + "grad_norm": 1.3316551446914673, + "learning_rate": 2.8853556221215794e-06, + "loss": 0.4519, + "step": 17545 + }, + { + "epoch": 0.7768615166642684, + "grad_norm": 1.7996611595153809, + "learning_rate": 2.8842696845612773e-06, + "loss": 0.4944, + "step": 17546 + }, + { + "epoch": 0.7769057923690823, + "grad_norm": 1.498342752456665, + "learning_rate": 2.8831839169543998e-06, + "loss": 0.4811, + "step": 17547 + }, + { + "epoch": 0.7769500680738961, + "grad_norm": 1.6969830989837646, + "learning_rate": 2.882098319326879e-06, + "loss": 0.5536, + "step": 17548 + }, + { + "epoch": 0.77699434377871, + "grad_norm": 1.313708782196045, + "learning_rate": 2.8810128917046453e-06, + "loss": 0.4909, + "step": 17549 + }, + { + "epoch": 0.7770386194835239, + "grad_norm": 1.413069248199463, + "learning_rate": 2.8799276341136217e-06, + "loss": 0.4929, + "step": 17550 + }, + { + "epoch": 0.7770828951883377, + "grad_norm": 1.2647614479064941, + "learning_rate": 2.8788425465797253e-06, + "loss": 0.4767, + "step": 17551 + }, + { + "epoch": 0.7771271708931516, + "grad_norm": 1.295448660850525, + "learning_rate": 2.8777576291288813e-06, + "loss": 0.4512, + "step": 17552 + }, + { + "epoch": 0.7771714465979656, + "grad_norm": 1.480733036994934, + "learning_rate": 2.876672881786996e-06, + "loss": 0.5341, + "step": 17553 + }, + { + "epoch": 0.7772157223027794, + "grad_norm": 1.1970053911209106, + "learning_rate": 2.875588304579976e-06, + "loss": 0.4186, + "step": 17554 + }, + { + "epoch": 0.7772599980075933, + "grad_norm": 1.4728553295135498, + "learning_rate": 2.8745038975337314e-06, + "loss": 0.5283, + "step": 17555 + }, + { + "epoch": 0.7773042737124072, + "grad_norm": 1.37140691280365, + "learning_rate": 2.8734196606741595e-06, + "loss": 0.4946, + "step": 17556 + }, + { + "epoch": 0.7773485494172211, + "grad_norm": 1.6108978986740112, + "learning_rate": 2.8723355940271556e-06, + "loss": 0.5611, + "step": 17557 + }, + { + "epoch": 0.7773928251220349, + "grad_norm": 1.5486866235733032, + "learning_rate": 2.8712516976186135e-06, + "loss": 0.5265, + "step": 17558 + }, + { + "epoch": 0.7774371008268488, + "grad_norm": 1.7846935987472534, + "learning_rate": 2.8701679714744202e-06, + "loss": 0.5585, + "step": 17559 + }, + { + "epoch": 0.7774813765316627, + "grad_norm": 1.6596442461013794, + "learning_rate": 2.869084415620459e-06, + "loss": 0.5405, + "step": 17560 + }, + { + "epoch": 0.7775256522364765, + "grad_norm": 1.236605167388916, + "learning_rate": 2.8680010300826078e-06, + "loss": 0.4781, + "step": 17561 + }, + { + "epoch": 0.7775699279412904, + "grad_norm": 1.624588966369629, + "learning_rate": 2.866917814886747e-06, + "loss": 0.5812, + "step": 17562 + }, + { + "epoch": 0.7776142036461043, + "grad_norm": 1.3031038045883179, + "learning_rate": 2.8658347700587475e-06, + "loss": 0.4634, + "step": 17563 + }, + { + "epoch": 0.7776584793509181, + "grad_norm": 1.4701128005981445, + "learning_rate": 2.8647518956244723e-06, + "loss": 0.5197, + "step": 17564 + }, + { + "epoch": 0.777702755055732, + "grad_norm": 1.8841657638549805, + "learning_rate": 2.8636691916097912e-06, + "loss": 0.4849, + "step": 17565 + }, + { + "epoch": 0.777747030760546, + "grad_norm": 1.619503378868103, + "learning_rate": 2.86258665804056e-06, + "loss": 0.5009, + "step": 17566 + }, + { + "epoch": 0.7777913064653598, + "grad_norm": 1.5293086767196655, + "learning_rate": 2.8615042949426365e-06, + "loss": 0.506, + "step": 17567 + }, + { + "epoch": 0.7778355821701737, + "grad_norm": 2.436138868331909, + "learning_rate": 2.86042210234187e-06, + "loss": 0.489, + "step": 17568 + }, + { + "epoch": 0.7778798578749876, + "grad_norm": 1.7189092636108398, + "learning_rate": 2.859340080264108e-06, + "loss": 0.5365, + "step": 17569 + }, + { + "epoch": 0.7779241335798014, + "grad_norm": 1.575075626373291, + "learning_rate": 2.8582582287351936e-06, + "loss": 0.5487, + "step": 17570 + }, + { + "epoch": 0.7779684092846153, + "grad_norm": 1.3413655757904053, + "learning_rate": 2.8571765477809645e-06, + "loss": 0.4769, + "step": 17571 + }, + { + "epoch": 0.7780126849894292, + "grad_norm": 1.6847914457321167, + "learning_rate": 2.85609503742726e-06, + "loss": 0.4998, + "step": 17572 + }, + { + "epoch": 0.778056960694243, + "grad_norm": 1.3238797187805176, + "learning_rate": 2.8550136976999086e-06, + "loss": 0.4668, + "step": 17573 + }, + { + "epoch": 0.7781012363990569, + "grad_norm": 1.4050283432006836, + "learning_rate": 2.8539325286247344e-06, + "loss": 0.4981, + "step": 17574 + }, + { + "epoch": 0.7781455121038708, + "grad_norm": 2.0645580291748047, + "learning_rate": 2.852851530227566e-06, + "loss": 0.5801, + "step": 17575 + }, + { + "epoch": 0.7781897878086846, + "grad_norm": 1.5712560415267944, + "learning_rate": 2.85177070253422e-06, + "loss": 0.4817, + "step": 17576 + }, + { + "epoch": 0.7782340635134986, + "grad_norm": 1.3635854721069336, + "learning_rate": 2.85069004557051e-06, + "loss": 0.4612, + "step": 17577 + }, + { + "epoch": 0.7782783392183125, + "grad_norm": 1.280371904373169, + "learning_rate": 2.8496095593622475e-06, + "loss": 0.4824, + "step": 17578 + }, + { + "epoch": 0.7783226149231263, + "grad_norm": 1.6327848434448242, + "learning_rate": 2.8485292439352385e-06, + "loss": 0.5119, + "step": 17579 + }, + { + "epoch": 0.7783668906279402, + "grad_norm": 1.6764943599700928, + "learning_rate": 2.847449099315286e-06, + "loss": 0.572, + "step": 17580 + }, + { + "epoch": 0.7784111663327541, + "grad_norm": 1.566449522972107, + "learning_rate": 2.8463691255281844e-06, + "loss": 0.5588, + "step": 17581 + }, + { + "epoch": 0.7784554420375679, + "grad_norm": 1.484116554260254, + "learning_rate": 2.8452893225997346e-06, + "loss": 0.4396, + "step": 17582 + }, + { + "epoch": 0.7784997177423818, + "grad_norm": 1.9616343975067139, + "learning_rate": 2.8442096905557235e-06, + "loss": 0.5638, + "step": 17583 + }, + { + "epoch": 0.7785439934471957, + "grad_norm": 1.5491334199905396, + "learning_rate": 2.8431302294219344e-06, + "loss": 0.5156, + "step": 17584 + }, + { + "epoch": 0.7785882691520095, + "grad_norm": 1.6678071022033691, + "learning_rate": 2.842050939224159e-06, + "loss": 0.4961, + "step": 17585 + }, + { + "epoch": 0.7786325448568234, + "grad_norm": 1.9093084335327148, + "learning_rate": 2.840971819988165e-06, + "loss": 0.563, + "step": 17586 + }, + { + "epoch": 0.7786768205616373, + "grad_norm": 2.5205581188201904, + "learning_rate": 2.8398928717397276e-06, + "loss": 0.5776, + "step": 17587 + }, + { + "epoch": 0.7787210962664511, + "grad_norm": 1.5668702125549316, + "learning_rate": 2.8388140945046215e-06, + "loss": 0.5584, + "step": 17588 + }, + { + "epoch": 0.778765371971265, + "grad_norm": 1.472939372062683, + "learning_rate": 2.8377354883086105e-06, + "loss": 0.536, + "step": 17589 + }, + { + "epoch": 0.778809647676079, + "grad_norm": 1.4644749164581299, + "learning_rate": 2.836657053177455e-06, + "loss": 0.4632, + "step": 17590 + }, + { + "epoch": 0.7788539233808928, + "grad_norm": 2.4885785579681396, + "learning_rate": 2.8355787891369115e-06, + "loss": 0.5184, + "step": 17591 + }, + { + "epoch": 0.7788981990857067, + "grad_norm": 1.2160191535949707, + "learning_rate": 2.8345006962127363e-06, + "loss": 0.4369, + "step": 17592 + }, + { + "epoch": 0.7789424747905206, + "grad_norm": 1.435166835784912, + "learning_rate": 2.8334227744306786e-06, + "loss": 0.4641, + "step": 17593 + }, + { + "epoch": 0.7789867504953345, + "grad_norm": 1.6922160387039185, + "learning_rate": 2.832345023816483e-06, + "loss": 0.5984, + "step": 17594 + }, + { + "epoch": 0.7790310262001483, + "grad_norm": 1.510004997253418, + "learning_rate": 2.83126744439589e-06, + "loss": 0.4828, + "step": 17595 + }, + { + "epoch": 0.7790753019049622, + "grad_norm": 1.4531208276748657, + "learning_rate": 2.8301900361946366e-06, + "loss": 0.4737, + "step": 17596 + }, + { + "epoch": 0.7791195776097761, + "grad_norm": 1.2881426811218262, + "learning_rate": 2.829112799238454e-06, + "loss": 0.452, + "step": 17597 + }, + { + "epoch": 0.7791638533145899, + "grad_norm": 1.8187097311019897, + "learning_rate": 2.8280357335530763e-06, + "loss": 0.487, + "step": 17598 + }, + { + "epoch": 0.7792081290194038, + "grad_norm": 1.3860441446304321, + "learning_rate": 2.8269588391642245e-06, + "loss": 0.4869, + "step": 17599 + }, + { + "epoch": 0.7792524047242178, + "grad_norm": 1.7214927673339844, + "learning_rate": 2.8258821160976212e-06, + "loss": 0.5963, + "step": 17600 + }, + { + "epoch": 0.7792966804290316, + "grad_norm": 1.3557895421981812, + "learning_rate": 2.8248055643789794e-06, + "loss": 0.4545, + "step": 17601 + }, + { + "epoch": 0.7793409561338455, + "grad_norm": 1.6373831033706665, + "learning_rate": 2.8237291840340166e-06, + "loss": 0.5136, + "step": 17602 + }, + { + "epoch": 0.7793852318386594, + "grad_norm": 1.7615119218826294, + "learning_rate": 2.8226529750884403e-06, + "loss": 0.5664, + "step": 17603 + }, + { + "epoch": 0.7794295075434732, + "grad_norm": 1.9866087436676025, + "learning_rate": 2.8215769375679524e-06, + "loss": 0.5253, + "step": 17604 + }, + { + "epoch": 0.7794737832482871, + "grad_norm": 1.6290420293807983, + "learning_rate": 2.820501071498255e-06, + "loss": 0.4917, + "step": 17605 + }, + { + "epoch": 0.779518058953101, + "grad_norm": 1.758663535118103, + "learning_rate": 2.819425376905044e-06, + "loss": 0.4893, + "step": 17606 + }, + { + "epoch": 0.7795623346579148, + "grad_norm": 1.7056777477264404, + "learning_rate": 2.8183498538140085e-06, + "loss": 0.538, + "step": 17607 + }, + { + "epoch": 0.7796066103627287, + "grad_norm": 1.7258856296539307, + "learning_rate": 2.8172745022508418e-06, + "loss": 0.5446, + "step": 17608 + }, + { + "epoch": 0.7796508860675426, + "grad_norm": 2.681928873062134, + "learning_rate": 2.8161993222412254e-06, + "loss": 0.6597, + "step": 17609 + }, + { + "epoch": 0.7796951617723564, + "grad_norm": 1.5514570474624634, + "learning_rate": 2.8151243138108388e-06, + "loss": 0.5124, + "step": 17610 + }, + { + "epoch": 0.7797394374771703, + "grad_norm": 1.5223630666732788, + "learning_rate": 2.8140494769853543e-06, + "loss": 0.5273, + "step": 17611 + }, + { + "epoch": 0.7797837131819843, + "grad_norm": 1.397908091545105, + "learning_rate": 2.8129748117904497e-06, + "loss": 0.4429, + "step": 17612 + }, + { + "epoch": 0.779827988886798, + "grad_norm": 1.5446116924285889, + "learning_rate": 2.8119003182517914e-06, + "loss": 0.5414, + "step": 17613 + }, + { + "epoch": 0.779872264591612, + "grad_norm": 1.7786688804626465, + "learning_rate": 2.81082599639504e-06, + "loss": 0.4919, + "step": 17614 + }, + { + "epoch": 0.7799165402964259, + "grad_norm": 1.5530198812484741, + "learning_rate": 2.809751846245856e-06, + "loss": 0.5522, + "step": 17615 + }, + { + "epoch": 0.7799608160012397, + "grad_norm": 1.4411733150482178, + "learning_rate": 2.8086778678298955e-06, + "loss": 0.493, + "step": 17616 + }, + { + "epoch": 0.7800050917060536, + "grad_norm": 1.657109260559082, + "learning_rate": 2.8076040611728085e-06, + "loss": 0.4599, + "step": 17617 + }, + { + "epoch": 0.7800493674108675, + "grad_norm": 1.5182198286056519, + "learning_rate": 2.806530426300239e-06, + "loss": 0.4618, + "step": 17618 + }, + { + "epoch": 0.7800936431156813, + "grad_norm": 1.5315532684326172, + "learning_rate": 2.8054569632378358e-06, + "loss": 0.5542, + "step": 17619 + }, + { + "epoch": 0.7801379188204952, + "grad_norm": 1.514251708984375, + "learning_rate": 2.804383672011236e-06, + "loss": 0.488, + "step": 17620 + }, + { + "epoch": 0.7801821945253091, + "grad_norm": 1.5377750396728516, + "learning_rate": 2.803310552646069e-06, + "loss": 0.4947, + "step": 17621 + }, + { + "epoch": 0.7802264702301229, + "grad_norm": 1.572555422782898, + "learning_rate": 2.8022376051679747e-06, + "loss": 0.5414, + "step": 17622 + }, + { + "epoch": 0.7802707459349368, + "grad_norm": 1.4844725131988525, + "learning_rate": 2.801164829602573e-06, + "loss": 0.4788, + "step": 17623 + }, + { + "epoch": 0.7803150216397507, + "grad_norm": 1.2306476831436157, + "learning_rate": 2.800092225975488e-06, + "loss": 0.4496, + "step": 17624 + }, + { + "epoch": 0.7803592973445646, + "grad_norm": 1.6084686517715454, + "learning_rate": 2.7990197943123377e-06, + "loss": 0.4692, + "step": 17625 + }, + { + "epoch": 0.7804035730493785, + "grad_norm": 1.4083366394042969, + "learning_rate": 2.7979475346387363e-06, + "loss": 0.5462, + "step": 17626 + }, + { + "epoch": 0.7804478487541924, + "grad_norm": 1.4346399307250977, + "learning_rate": 2.796875446980294e-06, + "loss": 0.4776, + "step": 17627 + }, + { + "epoch": 0.7804921244590063, + "grad_norm": 2.0793235301971436, + "learning_rate": 2.795803531362614e-06, + "loss": 0.4791, + "step": 17628 + }, + { + "epoch": 0.7805364001638201, + "grad_norm": 1.3653382062911987, + "learning_rate": 2.7947317878113033e-06, + "loss": 0.5171, + "step": 17629 + }, + { + "epoch": 0.780580675868634, + "grad_norm": 1.434366226196289, + "learning_rate": 2.7936602163519576e-06, + "loss": 0.4569, + "step": 17630 + }, + { + "epoch": 0.7806249515734479, + "grad_norm": 1.5659911632537842, + "learning_rate": 2.7925888170101667e-06, + "loss": 0.5525, + "step": 17631 + }, + { + "epoch": 0.7806692272782617, + "grad_norm": 1.3808839321136475, + "learning_rate": 2.7915175898115266e-06, + "loss": 0.4287, + "step": 17632 + }, + { + "epoch": 0.7807135029830756, + "grad_norm": 1.56717848777771, + "learning_rate": 2.790446534781622e-06, + "loss": 0.5499, + "step": 17633 + }, + { + "epoch": 0.7807577786878895, + "grad_norm": 1.254746675491333, + "learning_rate": 2.7893756519460268e-06, + "loss": 0.4553, + "step": 17634 + }, + { + "epoch": 0.7808020543927033, + "grad_norm": 1.397180199623108, + "learning_rate": 2.7883049413303256e-06, + "loss": 0.475, + "step": 17635 + }, + { + "epoch": 0.7808463300975172, + "grad_norm": 1.8126013278961182, + "learning_rate": 2.7872344029600884e-06, + "loss": 0.5467, + "step": 17636 + }, + { + "epoch": 0.7808906058023312, + "grad_norm": 1.7476764917373657, + "learning_rate": 2.7861640368608846e-06, + "loss": 0.5621, + "step": 17637 + }, + { + "epoch": 0.780934881507145, + "grad_norm": 1.6868010759353638, + "learning_rate": 2.7850938430582763e-06, + "loss": 0.5538, + "step": 17638 + }, + { + "epoch": 0.7809791572119589, + "grad_norm": 1.7608016729354858, + "learning_rate": 2.7840238215778305e-06, + "loss": 0.5887, + "step": 17639 + }, + { + "epoch": 0.7810234329167728, + "grad_norm": 1.429023265838623, + "learning_rate": 2.7829539724450995e-06, + "loss": 0.4588, + "step": 17640 + }, + { + "epoch": 0.7810677086215866, + "grad_norm": 1.3777767419815063, + "learning_rate": 2.781884295685636e-06, + "loss": 0.4999, + "step": 17641 + }, + { + "epoch": 0.7811119843264005, + "grad_norm": 1.9712382555007935, + "learning_rate": 2.7808147913249905e-06, + "loss": 0.5614, + "step": 17642 + }, + { + "epoch": 0.7811562600312144, + "grad_norm": 1.775329828262329, + "learning_rate": 2.779745459388705e-06, + "loss": 0.5512, + "step": 17643 + }, + { + "epoch": 0.7812005357360282, + "grad_norm": 1.5448158979415894, + "learning_rate": 2.7786762999023165e-06, + "loss": 0.5333, + "step": 17644 + }, + { + "epoch": 0.7812448114408421, + "grad_norm": 1.4841279983520508, + "learning_rate": 2.7776073128913685e-06, + "loss": 0.5418, + "step": 17645 + }, + { + "epoch": 0.781289087145656, + "grad_norm": 1.4604644775390625, + "learning_rate": 2.7765384983813894e-06, + "loss": 0.5002, + "step": 17646 + }, + { + "epoch": 0.7813333628504698, + "grad_norm": 1.4884824752807617, + "learning_rate": 2.7754698563979054e-06, + "loss": 0.544, + "step": 17647 + }, + { + "epoch": 0.7813776385552837, + "grad_norm": 1.827397346496582, + "learning_rate": 2.77440138696644e-06, + "loss": 0.5643, + "step": 17648 + }, + { + "epoch": 0.7814219142600977, + "grad_norm": 1.452852487564087, + "learning_rate": 2.7733330901125166e-06, + "loss": 0.4845, + "step": 17649 + }, + { + "epoch": 0.7814661899649115, + "grad_norm": 1.645161747932434, + "learning_rate": 2.772264965861647e-06, + "loss": 0.521, + "step": 17650 + }, + { + "epoch": 0.7815104656697254, + "grad_norm": 1.5881375074386597, + "learning_rate": 2.771197014239345e-06, + "loss": 0.5582, + "step": 17651 + }, + { + "epoch": 0.7815547413745393, + "grad_norm": 1.477514624595642, + "learning_rate": 2.7701292352711164e-06, + "loss": 0.4748, + "step": 17652 + }, + { + "epoch": 0.7815990170793531, + "grad_norm": 1.4470645189285278, + "learning_rate": 2.7690616289824634e-06, + "loss": 0.4907, + "step": 17653 + }, + { + "epoch": 0.781643292784167, + "grad_norm": 1.4223755598068237, + "learning_rate": 2.767994195398884e-06, + "loss": 0.4925, + "step": 17654 + }, + { + "epoch": 0.7816875684889809, + "grad_norm": 1.844552993774414, + "learning_rate": 2.766926934545877e-06, + "loss": 0.5902, + "step": 17655 + }, + { + "epoch": 0.7817318441937947, + "grad_norm": 1.5290898084640503, + "learning_rate": 2.7658598464489307e-06, + "loss": 0.4899, + "step": 17656 + }, + { + "epoch": 0.7817761198986086, + "grad_norm": 1.628819465637207, + "learning_rate": 2.764792931133532e-06, + "loss": 0.5146, + "step": 17657 + }, + { + "epoch": 0.7818203956034225, + "grad_norm": 1.92307448387146, + "learning_rate": 2.76372618862516e-06, + "loss": 0.6176, + "step": 17658 + }, + { + "epoch": 0.7818646713082363, + "grad_norm": 1.8190371990203857, + "learning_rate": 2.7626596189492983e-06, + "loss": 0.5924, + "step": 17659 + }, + { + "epoch": 0.7819089470130502, + "grad_norm": 2.1982500553131104, + "learning_rate": 2.761593222131419e-06, + "loss": 0.594, + "step": 17660 + }, + { + "epoch": 0.7819532227178642, + "grad_norm": 1.4006916284561157, + "learning_rate": 2.7605269981969927e-06, + "loss": 0.5225, + "step": 17661 + }, + { + "epoch": 0.781997498422678, + "grad_norm": 1.6582263708114624, + "learning_rate": 2.759460947171484e-06, + "loss": 0.5536, + "step": 17662 + }, + { + "epoch": 0.7820417741274919, + "grad_norm": 1.4059945344924927, + "learning_rate": 2.758395069080354e-06, + "loss": 0.4635, + "step": 17663 + }, + { + "epoch": 0.7820860498323058, + "grad_norm": 1.5186980962753296, + "learning_rate": 2.7573293639490595e-06, + "loss": 0.5185, + "step": 17664 + }, + { + "epoch": 0.7821303255371197, + "grad_norm": 1.4364334344863892, + "learning_rate": 2.7562638318030588e-06, + "loss": 0.501, + "step": 17665 + }, + { + "epoch": 0.7821746012419335, + "grad_norm": 1.5394039154052734, + "learning_rate": 2.7551984726677994e-06, + "loss": 0.5247, + "step": 17666 + }, + { + "epoch": 0.7822188769467474, + "grad_norm": 1.575880765914917, + "learning_rate": 2.7541332865687245e-06, + "loss": 0.5232, + "step": 17667 + }, + { + "epoch": 0.7822631526515613, + "grad_norm": 1.5017033815383911, + "learning_rate": 2.753068273531273e-06, + "loss": 0.5449, + "step": 17668 + }, + { + "epoch": 0.7823074283563751, + "grad_norm": 1.6494770050048828, + "learning_rate": 2.7520034335808886e-06, + "loss": 0.523, + "step": 17669 + }, + { + "epoch": 0.782351704061189, + "grad_norm": 1.7149921655654907, + "learning_rate": 2.750938766743e-06, + "loss": 0.501, + "step": 17670 + }, + { + "epoch": 0.782395979766003, + "grad_norm": 1.8023892641067505, + "learning_rate": 2.749874273043037e-06, + "loss": 0.5239, + "step": 17671 + }, + { + "epoch": 0.7824402554708167, + "grad_norm": 1.5756396055221558, + "learning_rate": 2.748809952506424e-06, + "loss": 0.5961, + "step": 17672 + }, + { + "epoch": 0.7824845311756307, + "grad_norm": 1.3263436555862427, + "learning_rate": 2.7477458051585815e-06, + "loss": 0.4583, + "step": 17673 + }, + { + "epoch": 0.7825288068804446, + "grad_norm": 1.6067559719085693, + "learning_rate": 2.746681831024921e-06, + "loss": 0.5502, + "step": 17674 + }, + { + "epoch": 0.7825730825852584, + "grad_norm": 1.5844768285751343, + "learning_rate": 2.745618030130863e-06, + "loss": 0.5733, + "step": 17675 + }, + { + "epoch": 0.7826173582900723, + "grad_norm": 1.5573492050170898, + "learning_rate": 2.7445544025018123e-06, + "loss": 0.4888, + "step": 17676 + }, + { + "epoch": 0.7826616339948862, + "grad_norm": 1.3997814655303955, + "learning_rate": 2.743490948163171e-06, + "loss": 0.486, + "step": 17677 + }, + { + "epoch": 0.7827059096997, + "grad_norm": 1.4952256679534912, + "learning_rate": 2.742427667140337e-06, + "loss": 0.5025, + "step": 17678 + }, + { + "epoch": 0.7827501854045139, + "grad_norm": 1.325415015220642, + "learning_rate": 2.7413645594587124e-06, + "loss": 0.4801, + "step": 17679 + }, + { + "epoch": 0.7827944611093278, + "grad_norm": 1.464855432510376, + "learning_rate": 2.7403016251436874e-06, + "loss": 0.4947, + "step": 17680 + }, + { + "epoch": 0.7828387368141416, + "grad_norm": 1.3837932348251343, + "learning_rate": 2.7392388642206423e-06, + "loss": 0.449, + "step": 17681 + }, + { + "epoch": 0.7828830125189555, + "grad_norm": 1.6653659343719482, + "learning_rate": 2.7381762767149668e-06, + "loss": 0.5165, + "step": 17682 + }, + { + "epoch": 0.7829272882237694, + "grad_norm": 1.4887652397155762, + "learning_rate": 2.737113862652039e-06, + "loss": 0.496, + "step": 17683 + }, + { + "epoch": 0.7829715639285832, + "grad_norm": 1.3978923559188843, + "learning_rate": 2.7360516220572285e-06, + "loss": 0.4733, + "step": 17684 + }, + { + "epoch": 0.7830158396333972, + "grad_norm": 1.4717998504638672, + "learning_rate": 2.734989554955915e-06, + "loss": 0.5083, + "step": 17685 + }, + { + "epoch": 0.7830601153382111, + "grad_norm": 1.4977340698242188, + "learning_rate": 2.73392766137346e-06, + "loss": 0.5421, + "step": 17686 + }, + { + "epoch": 0.7831043910430249, + "grad_norm": 1.5129731893539429, + "learning_rate": 2.7328659413352266e-06, + "loss": 0.4829, + "step": 17687 + }, + { + "epoch": 0.7831486667478388, + "grad_norm": 1.889173984527588, + "learning_rate": 2.731804394866573e-06, + "loss": 0.5681, + "step": 17688 + }, + { + "epoch": 0.7831929424526527, + "grad_norm": 1.2749181985855103, + "learning_rate": 2.730743021992853e-06, + "loss": 0.4634, + "step": 17689 + }, + { + "epoch": 0.7832372181574665, + "grad_norm": 1.8577409982681274, + "learning_rate": 2.7296818227394172e-06, + "loss": 0.5926, + "step": 17690 + }, + { + "epoch": 0.7832814938622804, + "grad_norm": 1.4640705585479736, + "learning_rate": 2.7286207971316094e-06, + "loss": 0.5121, + "step": 17691 + }, + { + "epoch": 0.7833257695670943, + "grad_norm": 1.3782635927200317, + "learning_rate": 2.7275599451947755e-06, + "loss": 0.471, + "step": 17692 + }, + { + "epoch": 0.7833700452719081, + "grad_norm": 1.6574726104736328, + "learning_rate": 2.726499266954251e-06, + "loss": 0.5031, + "step": 17693 + }, + { + "epoch": 0.783414320976722, + "grad_norm": 1.738178014755249, + "learning_rate": 2.725438762435366e-06, + "loss": 0.6014, + "step": 17694 + }, + { + "epoch": 0.783458596681536, + "grad_norm": 1.345475673675537, + "learning_rate": 2.724378431663456e-06, + "loss": 0.4755, + "step": 17695 + }, + { + "epoch": 0.7835028723863497, + "grad_norm": 1.4484938383102417, + "learning_rate": 2.7233182746638433e-06, + "loss": 0.5142, + "step": 17696 + }, + { + "epoch": 0.7835471480911637, + "grad_norm": 1.3944066762924194, + "learning_rate": 2.722258291461849e-06, + "loss": 0.5044, + "step": 17697 + }, + { + "epoch": 0.7835914237959776, + "grad_norm": 1.9502103328704834, + "learning_rate": 2.7211984820827887e-06, + "loss": 0.576, + "step": 17698 + }, + { + "epoch": 0.7836356995007915, + "grad_norm": 2.001354932785034, + "learning_rate": 2.720138846551977e-06, + "loss": 0.5253, + "step": 17699 + }, + { + "epoch": 0.7836799752056053, + "grad_norm": 1.672430157661438, + "learning_rate": 2.7190793848947206e-06, + "loss": 0.5016, + "step": 17700 + }, + { + "epoch": 0.7837242509104192, + "grad_norm": 1.8997009992599487, + "learning_rate": 2.7180200971363223e-06, + "loss": 0.6324, + "step": 17701 + }, + { + "epoch": 0.7837685266152331, + "grad_norm": 1.497270107269287, + "learning_rate": 2.7169609833020884e-06, + "loss": 0.5274, + "step": 17702 + }, + { + "epoch": 0.7838128023200469, + "grad_norm": 1.811535358428955, + "learning_rate": 2.71590204341731e-06, + "loss": 0.551, + "step": 17703 + }, + { + "epoch": 0.7838570780248608, + "grad_norm": 1.3946352005004883, + "learning_rate": 2.7148432775072787e-06, + "loss": 0.4782, + "step": 17704 + }, + { + "epoch": 0.7839013537296747, + "grad_norm": 1.6559346914291382, + "learning_rate": 2.713784685597286e-06, + "loss": 0.5426, + "step": 17705 + }, + { + "epoch": 0.7839456294344885, + "grad_norm": 1.304073691368103, + "learning_rate": 2.712726267712613e-06, + "loss": 0.4945, + "step": 17706 + }, + { + "epoch": 0.7839899051393024, + "grad_norm": 1.4683085680007935, + "learning_rate": 2.711668023878541e-06, + "loss": 0.4619, + "step": 17707 + }, + { + "epoch": 0.7840341808441164, + "grad_norm": 1.4112396240234375, + "learning_rate": 2.7106099541203433e-06, + "loss": 0.4927, + "step": 17708 + }, + { + "epoch": 0.7840784565489302, + "grad_norm": 1.606031060218811, + "learning_rate": 2.7095520584632917e-06, + "loss": 0.5153, + "step": 17709 + }, + { + "epoch": 0.7841227322537441, + "grad_norm": 1.8921209573745728, + "learning_rate": 2.708494336932652e-06, + "loss": 0.4818, + "step": 17710 + }, + { + "epoch": 0.784167007958558, + "grad_norm": 1.4517813920974731, + "learning_rate": 2.7074367895536856e-06, + "loss": 0.5287, + "step": 17711 + }, + { + "epoch": 0.7842112836633718, + "grad_norm": 1.705477237701416, + "learning_rate": 2.7063794163516575e-06, + "loss": 0.5078, + "step": 17712 + }, + { + "epoch": 0.7842555593681857, + "grad_norm": 2.00028920173645, + "learning_rate": 2.705322217351818e-06, + "loss": 0.485, + "step": 17713 + }, + { + "epoch": 0.7842998350729996, + "grad_norm": 1.4091858863830566, + "learning_rate": 2.704265192579415e-06, + "loss": 0.5126, + "step": 17714 + }, + { + "epoch": 0.7843441107778134, + "grad_norm": 1.711124300956726, + "learning_rate": 2.7032083420597e-06, + "loss": 0.5034, + "step": 17715 + }, + { + "epoch": 0.7843883864826273, + "grad_norm": 1.9097284078598022, + "learning_rate": 2.7021516658179124e-06, + "loss": 0.5768, + "step": 17716 + }, + { + "epoch": 0.7844326621874412, + "grad_norm": 1.3778080940246582, + "learning_rate": 2.7010951638792903e-06, + "loss": 0.4523, + "step": 17717 + }, + { + "epoch": 0.784476937892255, + "grad_norm": 1.487627625465393, + "learning_rate": 2.7000388362690666e-06, + "loss": 0.5414, + "step": 17718 + }, + { + "epoch": 0.784521213597069, + "grad_norm": 1.4742050170898438, + "learning_rate": 2.698982683012472e-06, + "loss": 0.4451, + "step": 17719 + }, + { + "epoch": 0.7845654893018829, + "grad_norm": 1.5210577249526978, + "learning_rate": 2.6979267041347324e-06, + "loss": 0.4815, + "step": 17720 + }, + { + "epoch": 0.7846097650066967, + "grad_norm": 1.6489461660385132, + "learning_rate": 2.696870899661064e-06, + "loss": 0.4808, + "step": 17721 + }, + { + "epoch": 0.7846540407115106, + "grad_norm": 1.7869843244552612, + "learning_rate": 2.6958152696166894e-06, + "loss": 0.5301, + "step": 17722 + }, + { + "epoch": 0.7846983164163245, + "grad_norm": 1.4891148805618286, + "learning_rate": 2.6947598140268214e-06, + "loss": 0.4719, + "step": 17723 + }, + { + "epoch": 0.7847425921211383, + "grad_norm": 1.267425298690796, + "learning_rate": 2.6937045329166634e-06, + "loss": 0.4596, + "step": 17724 + }, + { + "epoch": 0.7847868678259522, + "grad_norm": 1.3790203332901, + "learning_rate": 2.692649426311428e-06, + "loss": 0.5132, + "step": 17725 + }, + { + "epoch": 0.7848311435307661, + "grad_norm": 1.5494277477264404, + "learning_rate": 2.6915944942363094e-06, + "loss": 0.5088, + "step": 17726 + }, + { + "epoch": 0.7848754192355799, + "grad_norm": 1.6360399723052979, + "learning_rate": 2.69053973671651e-06, + "loss": 0.5289, + "step": 17727 + }, + { + "epoch": 0.7849196949403938, + "grad_norm": 1.224806785583496, + "learning_rate": 2.6894851537772104e-06, + "loss": 0.5101, + "step": 17728 + }, + { + "epoch": 0.7849639706452077, + "grad_norm": 1.3174585103988647, + "learning_rate": 2.6884307454436085e-06, + "loss": 0.485, + "step": 17729 + }, + { + "epoch": 0.7850082463500215, + "grad_norm": 1.9835642576217651, + "learning_rate": 2.6873765117408844e-06, + "loss": 0.6666, + "step": 17730 + }, + { + "epoch": 0.7850525220548354, + "grad_norm": 1.34906005859375, + "learning_rate": 2.6863224526942165e-06, + "loss": 0.472, + "step": 17731 + }, + { + "epoch": 0.7850967977596494, + "grad_norm": 1.5095984935760498, + "learning_rate": 2.685268568328785e-06, + "loss": 0.55, + "step": 17732 + }, + { + "epoch": 0.7851410734644632, + "grad_norm": 1.3015472888946533, + "learning_rate": 2.684214858669757e-06, + "loss": 0.4631, + "step": 17733 + }, + { + "epoch": 0.7851853491692771, + "grad_norm": 1.5489007234573364, + "learning_rate": 2.6831613237423003e-06, + "loss": 0.5133, + "step": 17734 + }, + { + "epoch": 0.785229624874091, + "grad_norm": 1.5662206411361694, + "learning_rate": 2.682107963571575e-06, + "loss": 0.5265, + "step": 17735 + }, + { + "epoch": 0.7852739005789049, + "grad_norm": 1.2199212312698364, + "learning_rate": 2.681054778182748e-06, + "loss": 0.4262, + "step": 17736 + }, + { + "epoch": 0.7853181762837187, + "grad_norm": 1.3585208654403687, + "learning_rate": 2.680001767600966e-06, + "loss": 0.4939, + "step": 17737 + }, + { + "epoch": 0.7853624519885326, + "grad_norm": 1.1246508359909058, + "learning_rate": 2.6789489318513783e-06, + "loss": 0.4787, + "step": 17738 + }, + { + "epoch": 0.7854067276933465, + "grad_norm": 2.347127914428711, + "learning_rate": 2.6778962709591382e-06, + "loss": 0.6041, + "step": 17739 + }, + { + "epoch": 0.7854510033981603, + "grad_norm": 1.8301961421966553, + "learning_rate": 2.6768437849493824e-06, + "loss": 0.5637, + "step": 17740 + }, + { + "epoch": 0.7854952791029742, + "grad_norm": 1.6202958822250366, + "learning_rate": 2.675791473847248e-06, + "loss": 0.5165, + "step": 17741 + }, + { + "epoch": 0.7855395548077881, + "grad_norm": 1.5806090831756592, + "learning_rate": 2.674739337677873e-06, + "loss": 0.5474, + "step": 17742 + }, + { + "epoch": 0.7855838305126019, + "grad_norm": 1.7433719635009766, + "learning_rate": 2.673687376466385e-06, + "loss": 0.5364, + "step": 17743 + }, + { + "epoch": 0.7856281062174159, + "grad_norm": 1.6472868919372559, + "learning_rate": 2.6726355902379084e-06, + "loss": 0.5302, + "step": 17744 + }, + { + "epoch": 0.7856723819222298, + "grad_norm": 1.4641741514205933, + "learning_rate": 2.6715839790175657e-06, + "loss": 0.5341, + "step": 17745 + }, + { + "epoch": 0.7857166576270436, + "grad_norm": 1.592298150062561, + "learning_rate": 2.6705325428304707e-06, + "loss": 0.5092, + "step": 17746 + }, + { + "epoch": 0.7857609333318575, + "grad_norm": 1.6261928081512451, + "learning_rate": 2.669481281701739e-06, + "loss": 0.522, + "step": 17747 + }, + { + "epoch": 0.7858052090366714, + "grad_norm": 1.3926523923873901, + "learning_rate": 2.668430195656476e-06, + "loss": 0.5109, + "step": 17748 + }, + { + "epoch": 0.7858494847414852, + "grad_norm": 1.4386208057403564, + "learning_rate": 2.66737928471979e-06, + "loss": 0.4645, + "step": 17749 + }, + { + "epoch": 0.7858937604462991, + "grad_norm": 1.6979666948318481, + "learning_rate": 2.66632854891678e-06, + "loss": 0.5168, + "step": 17750 + }, + { + "epoch": 0.785938036151113, + "grad_norm": 1.6013658046722412, + "learning_rate": 2.665277988272538e-06, + "loss": 0.4848, + "step": 17751 + }, + { + "epoch": 0.7859823118559268, + "grad_norm": 1.4385614395141602, + "learning_rate": 2.6642276028121617e-06, + "loss": 0.48, + "step": 17752 + }, + { + "epoch": 0.7860265875607407, + "grad_norm": 1.7013802528381348, + "learning_rate": 2.6631773925607374e-06, + "loss": 0.5494, + "step": 17753 + }, + { + "epoch": 0.7860708632655546, + "grad_norm": 1.6902028322219849, + "learning_rate": 2.662127357543346e-06, + "loss": 0.5052, + "step": 17754 + }, + { + "epoch": 0.7861151389703684, + "grad_norm": 1.6143898963928223, + "learning_rate": 2.6610774977850685e-06, + "loss": 0.4914, + "step": 17755 + }, + { + "epoch": 0.7861594146751824, + "grad_norm": 1.9511853456497192, + "learning_rate": 2.660027813310979e-06, + "loss": 0.5503, + "step": 17756 + }, + { + "epoch": 0.7862036903799963, + "grad_norm": 1.5942124128341675, + "learning_rate": 2.658978304146149e-06, + "loss": 0.4799, + "step": 17757 + }, + { + "epoch": 0.7862479660848101, + "grad_norm": 1.510027289390564, + "learning_rate": 2.657928970315642e-06, + "loss": 0.5258, + "step": 17758 + }, + { + "epoch": 0.786292241789624, + "grad_norm": 1.2779508829116821, + "learning_rate": 2.6568798118445262e-06, + "loss": 0.5434, + "step": 17759 + }, + { + "epoch": 0.7863365174944379, + "grad_norm": 1.4731138944625854, + "learning_rate": 2.655830828757857e-06, + "loss": 0.4997, + "step": 17760 + }, + { + "epoch": 0.7863807931992517, + "grad_norm": 1.6652238368988037, + "learning_rate": 2.6547820210806853e-06, + "loss": 0.5132, + "step": 17761 + }, + { + "epoch": 0.7864250689040656, + "grad_norm": 1.6769723892211914, + "learning_rate": 2.653733388838068e-06, + "loss": 0.5566, + "step": 17762 + }, + { + "epoch": 0.7864693446088795, + "grad_norm": 1.4062318801879883, + "learning_rate": 2.6526849320550474e-06, + "loss": 0.5143, + "step": 17763 + }, + { + "epoch": 0.7865136203136933, + "grad_norm": 1.343849539756775, + "learning_rate": 2.6516366507566638e-06, + "loss": 0.4418, + "step": 17764 + }, + { + "epoch": 0.7865578960185072, + "grad_norm": 1.4794591665267944, + "learning_rate": 2.650588544967957e-06, + "loss": 0.4975, + "step": 17765 + }, + { + "epoch": 0.7866021717233211, + "grad_norm": 1.3227832317352295, + "learning_rate": 2.649540614713957e-06, + "loss": 0.4684, + "step": 17766 + }, + { + "epoch": 0.7866464474281349, + "grad_norm": 1.4089725017547607, + "learning_rate": 2.6484928600196957e-06, + "loss": 0.4617, + "step": 17767 + }, + { + "epoch": 0.7866907231329489, + "grad_norm": 1.2676752805709839, + "learning_rate": 2.6474452809101926e-06, + "loss": 0.4748, + "step": 17768 + }, + { + "epoch": 0.7867349988377628, + "grad_norm": 1.500949740409851, + "learning_rate": 2.646397877410476e-06, + "loss": 0.5048, + "step": 17769 + }, + { + "epoch": 0.7867792745425767, + "grad_norm": 1.4695805311203003, + "learning_rate": 2.6453506495455584e-06, + "loss": 0.5251, + "step": 17770 + }, + { + "epoch": 0.7868235502473905, + "grad_norm": 1.4663605690002441, + "learning_rate": 2.6443035973404497e-06, + "loss": 0.4852, + "step": 17771 + }, + { + "epoch": 0.7868678259522044, + "grad_norm": 1.6294806003570557, + "learning_rate": 2.643256720820162e-06, + "loss": 0.5606, + "step": 17772 + }, + { + "epoch": 0.7869121016570183, + "grad_norm": 1.4173738956451416, + "learning_rate": 2.6422100200096988e-06, + "loss": 0.5543, + "step": 17773 + }, + { + "epoch": 0.7869563773618321, + "grad_norm": 1.387618064880371, + "learning_rate": 2.641163494934058e-06, + "loss": 0.4815, + "step": 17774 + }, + { + "epoch": 0.787000653066646, + "grad_norm": 1.8049488067626953, + "learning_rate": 2.640117145618235e-06, + "loss": 0.5343, + "step": 17775 + }, + { + "epoch": 0.7870449287714599, + "grad_norm": 1.9923765659332275, + "learning_rate": 2.639070972087221e-06, + "loss": 0.5055, + "step": 17776 + }, + { + "epoch": 0.7870892044762737, + "grad_norm": 1.4407576322555542, + "learning_rate": 2.6380249743660048e-06, + "loss": 0.5313, + "step": 17777 + }, + { + "epoch": 0.7871334801810876, + "grad_norm": 1.5975093841552734, + "learning_rate": 2.636979152479563e-06, + "loss": 0.5421, + "step": 17778 + }, + { + "epoch": 0.7871777558859016, + "grad_norm": 2.114976406097412, + "learning_rate": 2.6359335064528833e-06, + "loss": 0.58, + "step": 17779 + }, + { + "epoch": 0.7872220315907154, + "grad_norm": 1.6044414043426514, + "learning_rate": 2.6348880363109354e-06, + "loss": 0.4785, + "step": 17780 + }, + { + "epoch": 0.7872663072955293, + "grad_norm": 1.5554488897323608, + "learning_rate": 2.633842742078686e-06, + "loss": 0.5457, + "step": 17781 + }, + { + "epoch": 0.7873105830003432, + "grad_norm": 1.6602078676223755, + "learning_rate": 2.6327976237811083e-06, + "loss": 0.5251, + "step": 17782 + }, + { + "epoch": 0.787354858705157, + "grad_norm": 1.6033529043197632, + "learning_rate": 2.631752681443165e-06, + "loss": 0.5462, + "step": 17783 + }, + { + "epoch": 0.7873991344099709, + "grad_norm": 1.6622823476791382, + "learning_rate": 2.630707915089805e-06, + "loss": 0.501, + "step": 17784 + }, + { + "epoch": 0.7874434101147848, + "grad_norm": 1.513677716255188, + "learning_rate": 2.6296633247459835e-06, + "loss": 0.4888, + "step": 17785 + }, + { + "epoch": 0.7874876858195986, + "grad_norm": 1.6818456649780273, + "learning_rate": 2.6286189104366555e-06, + "loss": 0.5318, + "step": 17786 + }, + { + "epoch": 0.7875319615244125, + "grad_norm": 1.5110620260238647, + "learning_rate": 2.6275746721867624e-06, + "loss": 0.5231, + "step": 17787 + }, + { + "epoch": 0.7875762372292264, + "grad_norm": 1.4611526727676392, + "learning_rate": 2.6265306100212428e-06, + "loss": 0.5022, + "step": 17788 + }, + { + "epoch": 0.7876205129340402, + "grad_norm": 2.2854056358337402, + "learning_rate": 2.6254867239650394e-06, + "loss": 0.5916, + "step": 17789 + }, + { + "epoch": 0.7876647886388541, + "grad_norm": 1.5261536836624146, + "learning_rate": 2.624443014043081e-06, + "loss": 0.4388, + "step": 17790 + }, + { + "epoch": 0.787709064343668, + "grad_norm": 1.5898061990737915, + "learning_rate": 2.623399480280292e-06, + "loss": 0.5198, + "step": 17791 + }, + { + "epoch": 0.7877533400484819, + "grad_norm": 1.9813553094863892, + "learning_rate": 2.622356122701606e-06, + "loss": 0.6214, + "step": 17792 + }, + { + "epoch": 0.7877976157532958, + "grad_norm": 1.5632984638214111, + "learning_rate": 2.621312941331935e-06, + "loss": 0.5268, + "step": 17793 + }, + { + "epoch": 0.7878418914581097, + "grad_norm": 1.2700130939483643, + "learning_rate": 2.620269936196196e-06, + "loss": 0.4653, + "step": 17794 + }, + { + "epoch": 0.7878861671629235, + "grad_norm": 1.6072109937667847, + "learning_rate": 2.6192271073192976e-06, + "loss": 0.5815, + "step": 17795 + }, + { + "epoch": 0.7879304428677374, + "grad_norm": 1.6784660816192627, + "learning_rate": 2.6181844547261524e-06, + "loss": 0.5421, + "step": 17796 + }, + { + "epoch": 0.7879747185725513, + "grad_norm": 1.3061994314193726, + "learning_rate": 2.617141978441662e-06, + "loss": 0.4501, + "step": 17797 + }, + { + "epoch": 0.7880189942773651, + "grad_norm": 1.423680067062378, + "learning_rate": 2.616099678490721e-06, + "loss": 0.4484, + "step": 17798 + }, + { + "epoch": 0.788063269982179, + "grad_norm": 1.4150831699371338, + "learning_rate": 2.6150575548982295e-06, + "loss": 0.4519, + "step": 17799 + }, + { + "epoch": 0.7881075456869929, + "grad_norm": 1.4479762315750122, + "learning_rate": 2.6140156076890755e-06, + "loss": 0.5315, + "step": 17800 + }, + { + "epoch": 0.7881518213918067, + "grad_norm": 1.193110466003418, + "learning_rate": 2.612973836888145e-06, + "loss": 0.4765, + "step": 17801 + }, + { + "epoch": 0.7881960970966206, + "grad_norm": 1.4315979480743408, + "learning_rate": 2.61193224252032e-06, + "loss": 0.5051, + "step": 17802 + }, + { + "epoch": 0.7882403728014346, + "grad_norm": 1.711061954498291, + "learning_rate": 2.610890824610477e-06, + "loss": 0.5433, + "step": 17803 + }, + { + "epoch": 0.7882846485062485, + "grad_norm": 1.7653788328170776, + "learning_rate": 2.60984958318349e-06, + "loss": 0.597, + "step": 17804 + }, + { + "epoch": 0.7883289242110623, + "grad_norm": 1.5525786876678467, + "learning_rate": 2.6088085182642265e-06, + "loss": 0.5197, + "step": 17805 + }, + { + "epoch": 0.7883731999158762, + "grad_norm": 1.3383461236953735, + "learning_rate": 2.6077676298775554e-06, + "loss": 0.4539, + "step": 17806 + }, + { + "epoch": 0.7884174756206901, + "grad_norm": 1.2055107355117798, + "learning_rate": 2.6067269180483368e-06, + "loss": 0.4099, + "step": 17807 + }, + { + "epoch": 0.7884617513255039, + "grad_norm": 1.8072782754898071, + "learning_rate": 2.605686382801422e-06, + "loss": 0.5744, + "step": 17808 + }, + { + "epoch": 0.7885060270303178, + "grad_norm": 1.4987512826919556, + "learning_rate": 2.60464602416167e-06, + "loss": 0.5303, + "step": 17809 + }, + { + "epoch": 0.7885503027351317, + "grad_norm": 2.0045228004455566, + "learning_rate": 2.603605842153927e-06, + "loss": 0.5874, + "step": 17810 + }, + { + "epoch": 0.7885945784399455, + "grad_norm": 1.4792073965072632, + "learning_rate": 2.602565836803036e-06, + "loss": 0.4681, + "step": 17811 + }, + { + "epoch": 0.7886388541447594, + "grad_norm": 1.4285459518432617, + "learning_rate": 2.6015260081338358e-06, + "loss": 0.4738, + "step": 17812 + }, + { + "epoch": 0.7886831298495733, + "grad_norm": 1.6540051698684692, + "learning_rate": 2.6004863561711633e-06, + "loss": 0.5535, + "step": 17813 + }, + { + "epoch": 0.7887274055543871, + "grad_norm": 1.809938907623291, + "learning_rate": 2.59944688093985e-06, + "loss": 0.5092, + "step": 17814 + }, + { + "epoch": 0.788771681259201, + "grad_norm": 1.4213546514511108, + "learning_rate": 2.598407582464718e-06, + "loss": 0.4999, + "step": 17815 + }, + { + "epoch": 0.788815956964015, + "grad_norm": 1.6201921701431274, + "learning_rate": 2.5973684607705986e-06, + "loss": 0.5072, + "step": 17816 + }, + { + "epoch": 0.7888602326688288, + "grad_norm": 1.9045116901397705, + "learning_rate": 2.5963295158823056e-06, + "loss": 0.5931, + "step": 17817 + }, + { + "epoch": 0.7889045083736427, + "grad_norm": 1.634790301322937, + "learning_rate": 2.5952907478246505e-06, + "loss": 0.5341, + "step": 17818 + }, + { + "epoch": 0.7889487840784566, + "grad_norm": 2.0191004276275635, + "learning_rate": 2.5942521566224498e-06, + "loss": 0.5967, + "step": 17819 + }, + { + "epoch": 0.7889930597832704, + "grad_norm": 1.5719096660614014, + "learning_rate": 2.5932137423005066e-06, + "loss": 0.5085, + "step": 17820 + }, + { + "epoch": 0.7890373354880843, + "grad_norm": 1.5150964260101318, + "learning_rate": 2.592175504883622e-06, + "loss": 0.5224, + "step": 17821 + }, + { + "epoch": 0.7890816111928982, + "grad_norm": 1.3543683290481567, + "learning_rate": 2.5911374443965943e-06, + "loss": 0.5369, + "step": 17822 + }, + { + "epoch": 0.789125886897712, + "grad_norm": 1.4834891557693481, + "learning_rate": 2.5900995608642155e-06, + "loss": 0.4698, + "step": 17823 + }, + { + "epoch": 0.7891701626025259, + "grad_norm": 1.2143681049346924, + "learning_rate": 2.5890618543112755e-06, + "loss": 0.4719, + "step": 17824 + }, + { + "epoch": 0.7892144383073398, + "grad_norm": 1.3819527626037598, + "learning_rate": 2.588024324762556e-06, + "loss": 0.5269, + "step": 17825 + }, + { + "epoch": 0.7892587140121536, + "grad_norm": 1.3766286373138428, + "learning_rate": 2.586986972242844e-06, + "loss": 0.5034, + "step": 17826 + }, + { + "epoch": 0.7893029897169676, + "grad_norm": 1.61708402633667, + "learning_rate": 2.585949796776912e-06, + "loss": 0.4951, + "step": 17827 + }, + { + "epoch": 0.7893472654217815, + "grad_norm": 1.5418148040771484, + "learning_rate": 2.5849127983895285e-06, + "loss": 0.512, + "step": 17828 + }, + { + "epoch": 0.7893915411265953, + "grad_norm": 1.7738898992538452, + "learning_rate": 2.583875977105469e-06, + "loss": 0.5361, + "step": 17829 + }, + { + "epoch": 0.7894358168314092, + "grad_norm": 1.648335576057434, + "learning_rate": 2.582839332949494e-06, + "loss": 0.53, + "step": 17830 + }, + { + "epoch": 0.7894800925362231, + "grad_norm": 1.4845695495605469, + "learning_rate": 2.5818028659463612e-06, + "loss": 0.5731, + "step": 17831 + }, + { + "epoch": 0.7895243682410369, + "grad_norm": 1.4941461086273193, + "learning_rate": 2.5807665761208265e-06, + "loss": 0.5159, + "step": 17832 + }, + { + "epoch": 0.7895686439458508, + "grad_norm": 1.51129150390625, + "learning_rate": 2.5797304634976417e-06, + "loss": 0.4987, + "step": 17833 + }, + { + "epoch": 0.7896129196506647, + "grad_norm": 1.583960771560669, + "learning_rate": 2.578694528101554e-06, + "loss": 0.5246, + "step": 17834 + }, + { + "epoch": 0.7896571953554785, + "grad_norm": 1.6780682802200317, + "learning_rate": 2.5776587699573007e-06, + "loss": 0.5502, + "step": 17835 + }, + { + "epoch": 0.7897014710602924, + "grad_norm": 1.25044846534729, + "learning_rate": 2.5766231890896277e-06, + "loss": 0.4697, + "step": 17836 + }, + { + "epoch": 0.7897457467651063, + "grad_norm": 1.6533764600753784, + "learning_rate": 2.5755877855232648e-06, + "loss": 0.5487, + "step": 17837 + }, + { + "epoch": 0.7897900224699201, + "grad_norm": 1.673710823059082, + "learning_rate": 2.57455255928294e-06, + "loss": 0.5508, + "step": 17838 + }, + { + "epoch": 0.789834298174734, + "grad_norm": 1.3425352573394775, + "learning_rate": 2.5735175103933873e-06, + "loss": 0.4982, + "step": 17839 + }, + { + "epoch": 0.789878573879548, + "grad_norm": 1.9280022382736206, + "learning_rate": 2.5724826388793187e-06, + "loss": 0.5751, + "step": 17840 + }, + { + "epoch": 0.7899228495843619, + "grad_norm": 1.7635470628738403, + "learning_rate": 2.571447944765455e-06, + "loss": 0.6153, + "step": 17841 + }, + { + "epoch": 0.7899671252891757, + "grad_norm": 1.6400681734085083, + "learning_rate": 2.570413428076505e-06, + "loss": 0.5007, + "step": 17842 + }, + { + "epoch": 0.7900114009939896, + "grad_norm": 1.4019216299057007, + "learning_rate": 2.569379088837184e-06, + "loss": 0.4974, + "step": 17843 + }, + { + "epoch": 0.7900556766988035, + "grad_norm": 1.5121721029281616, + "learning_rate": 2.568344927072194e-06, + "loss": 0.4695, + "step": 17844 + }, + { + "epoch": 0.7900999524036173, + "grad_norm": 2.034111261367798, + "learning_rate": 2.5673109428062306e-06, + "loss": 0.5852, + "step": 17845 + }, + { + "epoch": 0.7901442281084312, + "grad_norm": 1.5169733762741089, + "learning_rate": 2.566277136063996e-06, + "loss": 0.5078, + "step": 17846 + }, + { + "epoch": 0.7901885038132451, + "grad_norm": 1.271994709968567, + "learning_rate": 2.56524350687018e-06, + "loss": 0.4644, + "step": 17847 + }, + { + "epoch": 0.7902327795180589, + "grad_norm": 1.661611795425415, + "learning_rate": 2.5642100552494687e-06, + "loss": 0.5258, + "step": 17848 + }, + { + "epoch": 0.7902770552228728, + "grad_norm": 1.608757734298706, + "learning_rate": 2.5631767812265464e-06, + "loss": 0.5214, + "step": 17849 + }, + { + "epoch": 0.7903213309276867, + "grad_norm": 1.6835256814956665, + "learning_rate": 2.562143684826092e-06, + "loss": 0.5801, + "step": 17850 + }, + { + "epoch": 0.7903656066325006, + "grad_norm": 1.7114779949188232, + "learning_rate": 2.5611107660727773e-06, + "loss": 0.5572, + "step": 17851 + }, + { + "epoch": 0.7904098823373145, + "grad_norm": 1.8608700037002563, + "learning_rate": 2.560078024991274e-06, + "loss": 0.5692, + "step": 17852 + }, + { + "epoch": 0.7904541580421284, + "grad_norm": 1.6758780479431152, + "learning_rate": 2.559045461606251e-06, + "loss": 0.5521, + "step": 17853 + }, + { + "epoch": 0.7904984337469422, + "grad_norm": 1.8296397924423218, + "learning_rate": 2.5580130759423695e-06, + "loss": 0.456, + "step": 17854 + }, + { + "epoch": 0.7905427094517561, + "grad_norm": 1.5870252847671509, + "learning_rate": 2.5569808680242826e-06, + "loss": 0.5257, + "step": 17855 + }, + { + "epoch": 0.79058698515657, + "grad_norm": 1.4377260208129883, + "learning_rate": 2.5559488378766507e-06, + "loss": 0.4567, + "step": 17856 + }, + { + "epoch": 0.7906312608613838, + "grad_norm": 1.6658354997634888, + "learning_rate": 2.5549169855241194e-06, + "loss": 0.4777, + "step": 17857 + }, + { + "epoch": 0.7906755365661977, + "grad_norm": 1.5847399234771729, + "learning_rate": 2.553885310991334e-06, + "loss": 0.5764, + "step": 17858 + }, + { + "epoch": 0.7907198122710116, + "grad_norm": 1.57047438621521, + "learning_rate": 2.552853814302936e-06, + "loss": 0.5389, + "step": 17859 + }, + { + "epoch": 0.7907640879758254, + "grad_norm": 1.7924922704696655, + "learning_rate": 2.551822495483559e-06, + "loss": 0.5765, + "step": 17860 + }, + { + "epoch": 0.7908083636806393, + "grad_norm": 1.300778865814209, + "learning_rate": 2.550791354557839e-06, + "loss": 0.4613, + "step": 17861 + }, + { + "epoch": 0.7908526393854532, + "grad_norm": 1.7756718397140503, + "learning_rate": 2.5497603915503975e-06, + "loss": 0.5295, + "step": 17862 + }, + { + "epoch": 0.790896915090267, + "grad_norm": 1.5216736793518066, + "learning_rate": 2.548729606485867e-06, + "loss": 0.4905, + "step": 17863 + }, + { + "epoch": 0.790941190795081, + "grad_norm": 1.220544457435608, + "learning_rate": 2.5476989993888623e-06, + "loss": 0.4167, + "step": 17864 + }, + { + "epoch": 0.7909854664998949, + "grad_norm": 2.1720542907714844, + "learning_rate": 2.5466685702839964e-06, + "loss": 0.5985, + "step": 17865 + }, + { + "epoch": 0.7910297422047087, + "grad_norm": 1.6095231771469116, + "learning_rate": 2.5456383191958857e-06, + "loss": 0.522, + "step": 17866 + }, + { + "epoch": 0.7910740179095226, + "grad_norm": 1.2979353666305542, + "learning_rate": 2.544608246149135e-06, + "loss": 0.4565, + "step": 17867 + }, + { + "epoch": 0.7911182936143365, + "grad_norm": 1.8934615850448608, + "learning_rate": 2.5435783511683444e-06, + "loss": 0.5795, + "step": 17868 + }, + { + "epoch": 0.7911625693191503, + "grad_norm": 1.09079909324646, + "learning_rate": 2.5425486342781146e-06, + "loss": 0.4257, + "step": 17869 + }, + { + "epoch": 0.7912068450239642, + "grad_norm": 1.5694737434387207, + "learning_rate": 2.541519095503039e-06, + "loss": 0.5364, + "step": 17870 + }, + { + "epoch": 0.7912511207287781, + "grad_norm": 1.376070499420166, + "learning_rate": 2.540489734867706e-06, + "loss": 0.5018, + "step": 17871 + }, + { + "epoch": 0.7912953964335919, + "grad_norm": 1.555316686630249, + "learning_rate": 2.5394605523967e-06, + "loss": 0.5473, + "step": 17872 + }, + { + "epoch": 0.7913396721384058, + "grad_norm": 1.5303665399551392, + "learning_rate": 2.538431548114606e-06, + "loss": 0.5451, + "step": 17873 + }, + { + "epoch": 0.7913839478432197, + "grad_norm": 1.633155107498169, + "learning_rate": 2.537402722046e-06, + "loss": 0.5553, + "step": 17874 + }, + { + "epoch": 0.7914282235480337, + "grad_norm": 1.5133864879608154, + "learning_rate": 2.5363740742154498e-06, + "loss": 0.4574, + "step": 17875 + }, + { + "epoch": 0.7914724992528475, + "grad_norm": 1.6145589351654053, + "learning_rate": 2.5353456046475312e-06, + "loss": 0.528, + "step": 17876 + }, + { + "epoch": 0.7915167749576614, + "grad_norm": 1.7411218881607056, + "learning_rate": 2.5343173133668043e-06, + "loss": 0.5737, + "step": 17877 + }, + { + "epoch": 0.7915610506624753, + "grad_norm": 2.1104001998901367, + "learning_rate": 2.5332892003978295e-06, + "loss": 0.5473, + "step": 17878 + }, + { + "epoch": 0.7916053263672891, + "grad_norm": 1.8051176071166992, + "learning_rate": 2.5322612657651637e-06, + "loss": 0.5748, + "step": 17879 + }, + { + "epoch": 0.791649602072103, + "grad_norm": 1.78432297706604, + "learning_rate": 2.531233509493356e-06, + "loss": 0.5333, + "step": 17880 + }, + { + "epoch": 0.7916938777769169, + "grad_norm": 1.4478635787963867, + "learning_rate": 2.530205931606955e-06, + "loss": 0.4454, + "step": 17881 + }, + { + "epoch": 0.7917381534817307, + "grad_norm": 1.472091794013977, + "learning_rate": 2.5291785321305006e-06, + "loss": 0.497, + "step": 17882 + }, + { + "epoch": 0.7917824291865446, + "grad_norm": 1.7082449197769165, + "learning_rate": 2.528151311088537e-06, + "loss": 0.4998, + "step": 17883 + }, + { + "epoch": 0.7918267048913585, + "grad_norm": 1.645734190940857, + "learning_rate": 2.527124268505595e-06, + "loss": 0.5236, + "step": 17884 + }, + { + "epoch": 0.7918709805961723, + "grad_norm": 1.7547162771224976, + "learning_rate": 2.5260974044062024e-06, + "loss": 0.5755, + "step": 17885 + }, + { + "epoch": 0.7919152563009862, + "grad_norm": 1.7835930585861206, + "learning_rate": 2.5250707188148902e-06, + "loss": 0.5034, + "step": 17886 + }, + { + "epoch": 0.7919595320058002, + "grad_norm": 1.6501505374908447, + "learning_rate": 2.5240442117561813e-06, + "loss": 0.5387, + "step": 17887 + }, + { + "epoch": 0.792003807710614, + "grad_norm": 1.6486361026763916, + "learning_rate": 2.523017883254584e-06, + "loss": 0.4968, + "step": 17888 + }, + { + "epoch": 0.7920480834154279, + "grad_norm": 1.5233310461044312, + "learning_rate": 2.521991733334619e-06, + "loss": 0.5217, + "step": 17889 + }, + { + "epoch": 0.7920923591202418, + "grad_norm": 1.5389591455459595, + "learning_rate": 2.520965762020792e-06, + "loss": 0.5082, + "step": 17890 + }, + { + "epoch": 0.7921366348250556, + "grad_norm": 1.4169151782989502, + "learning_rate": 2.519939969337608e-06, + "loss": 0.5288, + "step": 17891 + }, + { + "epoch": 0.7921809105298695, + "grad_norm": 1.7379273176193237, + "learning_rate": 2.5189143553095654e-06, + "loss": 0.4678, + "step": 17892 + }, + { + "epoch": 0.7922251862346834, + "grad_norm": 1.5467617511749268, + "learning_rate": 2.5178889199611645e-06, + "loss": 0.5157, + "step": 17893 + }, + { + "epoch": 0.7922694619394972, + "grad_norm": 1.272842288017273, + "learning_rate": 2.516863663316894e-06, + "loss": 0.466, + "step": 17894 + }, + { + "epoch": 0.7923137376443111, + "grad_norm": 1.4940694570541382, + "learning_rate": 2.5158385854012423e-06, + "loss": 0.5093, + "step": 17895 + }, + { + "epoch": 0.792358013349125, + "grad_norm": 1.6232086420059204, + "learning_rate": 2.514813686238693e-06, + "loss": 0.4949, + "step": 17896 + }, + { + "epoch": 0.7924022890539388, + "grad_norm": 1.8807052373886108, + "learning_rate": 2.513788965853723e-06, + "loss": 0.5573, + "step": 17897 + }, + { + "epoch": 0.7924465647587527, + "grad_norm": 1.4331055879592896, + "learning_rate": 2.512764424270806e-06, + "loss": 0.5154, + "step": 17898 + }, + { + "epoch": 0.7924908404635667, + "grad_norm": 1.9559820890426636, + "learning_rate": 2.511740061514417e-06, + "loss": 0.5196, + "step": 17899 + }, + { + "epoch": 0.7925351161683805, + "grad_norm": 1.577494740486145, + "learning_rate": 2.510715877609019e-06, + "loss": 0.4991, + "step": 17900 + }, + { + "epoch": 0.7925793918731944, + "grad_norm": 1.8090455532073975, + "learning_rate": 2.5096918725790754e-06, + "loss": 0.5548, + "step": 17901 + }, + { + "epoch": 0.7926236675780083, + "grad_norm": 1.5794693231582642, + "learning_rate": 2.5086680464490386e-06, + "loss": 0.5143, + "step": 17902 + }, + { + "epoch": 0.7926679432828221, + "grad_norm": 1.6081047058105469, + "learning_rate": 2.5076443992433684e-06, + "loss": 0.4952, + "step": 17903 + }, + { + "epoch": 0.792712218987636, + "grad_norm": 1.3943759202957153, + "learning_rate": 2.506620930986512e-06, + "loss": 0.455, + "step": 17904 + }, + { + "epoch": 0.7927564946924499, + "grad_norm": 1.4416636228561401, + "learning_rate": 2.5055976417029137e-06, + "loss": 0.473, + "step": 17905 + }, + { + "epoch": 0.7928007703972637, + "grad_norm": 1.9213390350341797, + "learning_rate": 2.5045745314170135e-06, + "loss": 0.5665, + "step": 17906 + }, + { + "epoch": 0.7928450461020776, + "grad_norm": 1.9947298765182495, + "learning_rate": 2.5035516001532467e-06, + "loss": 0.5648, + "step": 17907 + }, + { + "epoch": 0.7928893218068915, + "grad_norm": 1.323724627494812, + "learning_rate": 2.502528847936044e-06, + "loss": 0.4506, + "step": 17908 + }, + { + "epoch": 0.7929335975117053, + "grad_norm": 1.357202172279358, + "learning_rate": 2.5015062747898377e-06, + "loss": 0.4955, + "step": 17909 + }, + { + "epoch": 0.7929778732165192, + "grad_norm": 1.4736872911453247, + "learning_rate": 2.5004838807390485e-06, + "loss": 0.4798, + "step": 17910 + }, + { + "epoch": 0.7930221489213332, + "grad_norm": 1.7280470132827759, + "learning_rate": 2.499461665808095e-06, + "loss": 0.5699, + "step": 17911 + }, + { + "epoch": 0.7930664246261471, + "grad_norm": 1.7456881999969482, + "learning_rate": 2.498439630021391e-06, + "loss": 0.4975, + "step": 17912 + }, + { + "epoch": 0.7931107003309609, + "grad_norm": 1.9543485641479492, + "learning_rate": 2.49741777340335e-06, + "loss": 0.5722, + "step": 17913 + }, + { + "epoch": 0.7931549760357748, + "grad_norm": 1.525871992111206, + "learning_rate": 2.496396095978377e-06, + "loss": 0.48, + "step": 17914 + }, + { + "epoch": 0.7931992517405887, + "grad_norm": 1.5169068574905396, + "learning_rate": 2.495374597770873e-06, + "loss": 0.4923, + "step": 17915 + }, + { + "epoch": 0.7932435274454025, + "grad_norm": 1.8692386150360107, + "learning_rate": 2.4943532788052373e-06, + "loss": 0.5324, + "step": 17916 + }, + { + "epoch": 0.7932878031502164, + "grad_norm": 1.5124335289001465, + "learning_rate": 2.493332139105862e-06, + "loss": 0.5344, + "step": 17917 + }, + { + "epoch": 0.7933320788550303, + "grad_norm": 2.132007598876953, + "learning_rate": 2.492311178697133e-06, + "loss": 0.5878, + "step": 17918 + }, + { + "epoch": 0.7933763545598441, + "grad_norm": 2.194416046142578, + "learning_rate": 2.491290397603442e-06, + "loss": 0.5814, + "step": 17919 + }, + { + "epoch": 0.793420630264658, + "grad_norm": 1.6661806106567383, + "learning_rate": 2.4902697958491653e-06, + "loss": 0.5556, + "step": 17920 + }, + { + "epoch": 0.793464905969472, + "grad_norm": 1.6100611686706543, + "learning_rate": 2.4892493734586807e-06, + "loss": 0.5126, + "step": 17921 + }, + { + "epoch": 0.7935091816742857, + "grad_norm": 1.6447253227233887, + "learning_rate": 2.4882291304563557e-06, + "loss": 0.5043, + "step": 17922 + }, + { + "epoch": 0.7935534573790997, + "grad_norm": 1.4367148876190186, + "learning_rate": 2.487209066866565e-06, + "loss": 0.482, + "step": 17923 + }, + { + "epoch": 0.7935977330839136, + "grad_norm": 1.759006381034851, + "learning_rate": 2.4861891827136684e-06, + "loss": 0.597, + "step": 17924 + }, + { + "epoch": 0.7936420087887274, + "grad_norm": 1.4286518096923828, + "learning_rate": 2.485169478022025e-06, + "loss": 0.5013, + "step": 17925 + }, + { + "epoch": 0.7936862844935413, + "grad_norm": 1.3435039520263672, + "learning_rate": 2.484149952815991e-06, + "loss": 0.4859, + "step": 17926 + }, + { + "epoch": 0.7937305601983552, + "grad_norm": 1.517854928970337, + "learning_rate": 2.4831306071199146e-06, + "loss": 0.5174, + "step": 17927 + }, + { + "epoch": 0.793774835903169, + "grad_norm": 1.798717737197876, + "learning_rate": 2.4821114409581403e-06, + "loss": 0.5882, + "step": 17928 + }, + { + "epoch": 0.7938191116079829, + "grad_norm": 1.7955281734466553, + "learning_rate": 2.4810924543550163e-06, + "loss": 0.5768, + "step": 17929 + }, + { + "epoch": 0.7938633873127968, + "grad_norm": 1.8525474071502686, + "learning_rate": 2.480073647334876e-06, + "loss": 0.5772, + "step": 17930 + }, + { + "epoch": 0.7939076630176106, + "grad_norm": 1.5781137943267822, + "learning_rate": 2.4790550199220543e-06, + "loss": 0.546, + "step": 17931 + }, + { + "epoch": 0.7939519387224245, + "grad_norm": 1.3862887620925903, + "learning_rate": 2.478036572140877e-06, + "loss": 0.4817, + "step": 17932 + }, + { + "epoch": 0.7939962144272384, + "grad_norm": 1.3322327136993408, + "learning_rate": 2.4770183040156736e-06, + "loss": 0.4616, + "step": 17933 + }, + { + "epoch": 0.7940404901320522, + "grad_norm": 1.9936332702636719, + "learning_rate": 2.4760002155707665e-06, + "loss": 0.5751, + "step": 17934 + }, + { + "epoch": 0.7940847658368662, + "grad_norm": 1.6906882524490356, + "learning_rate": 2.4749823068304624e-06, + "loss": 0.4943, + "step": 17935 + }, + { + "epoch": 0.7941290415416801, + "grad_norm": 2.019707441329956, + "learning_rate": 2.4739645778190814e-06, + "loss": 0.5701, + "step": 17936 + }, + { + "epoch": 0.7941733172464939, + "grad_norm": 1.3774957656860352, + "learning_rate": 2.472947028560928e-06, + "loss": 0.4815, + "step": 17937 + }, + { + "epoch": 0.7942175929513078, + "grad_norm": 1.6122546195983887, + "learning_rate": 2.471929659080303e-06, + "loss": 0.5046, + "step": 17938 + }, + { + "epoch": 0.7942618686561217, + "grad_norm": 1.8565735816955566, + "learning_rate": 2.470912469401512e-06, + "loss": 0.5095, + "step": 17939 + }, + { + "epoch": 0.7943061443609355, + "grad_norm": 1.3518260717391968, + "learning_rate": 2.4698954595488456e-06, + "loss": 0.4791, + "step": 17940 + }, + { + "epoch": 0.7943504200657494, + "grad_norm": 1.5714823007583618, + "learning_rate": 2.468878629546596e-06, + "loss": 0.502, + "step": 17941 + }, + { + "epoch": 0.7943946957705633, + "grad_norm": 1.2948328256607056, + "learning_rate": 2.467861979419044e-06, + "loss": 0.4759, + "step": 17942 + }, + { + "epoch": 0.7944389714753771, + "grad_norm": 1.2752028703689575, + "learning_rate": 2.466845509190483e-06, + "loss": 0.5029, + "step": 17943 + }, + { + "epoch": 0.794483247180191, + "grad_norm": 1.2450650930404663, + "learning_rate": 2.4658292188851796e-06, + "loss": 0.4441, + "step": 17944 + }, + { + "epoch": 0.7945275228850049, + "grad_norm": 1.2759846448898315, + "learning_rate": 2.4648131085274074e-06, + "loss": 0.4849, + "step": 17945 + }, + { + "epoch": 0.7945717985898189, + "grad_norm": 1.332667589187622, + "learning_rate": 2.463797178141443e-06, + "loss": 0.4402, + "step": 17946 + }, + { + "epoch": 0.7946160742946327, + "grad_norm": 2.100191354751587, + "learning_rate": 2.462781427751545e-06, + "loss": 0.6167, + "step": 17947 + }, + { + "epoch": 0.7946603499994466, + "grad_norm": 1.6468861103057861, + "learning_rate": 2.461765857381977e-06, + "loss": 0.5853, + "step": 17948 + }, + { + "epoch": 0.7947046257042605, + "grad_norm": 1.5498322248458862, + "learning_rate": 2.4607504670569903e-06, + "loss": 0.5271, + "step": 17949 + }, + { + "epoch": 0.7947489014090743, + "grad_norm": 1.495362401008606, + "learning_rate": 2.4597352568008425e-06, + "loss": 0.4996, + "step": 17950 + }, + { + "epoch": 0.7947931771138882, + "grad_norm": 1.7753692865371704, + "learning_rate": 2.4587202266377796e-06, + "loss": 0.4964, + "step": 17951 + }, + { + "epoch": 0.7948374528187021, + "grad_norm": 1.5042600631713867, + "learning_rate": 2.4577053765920424e-06, + "loss": 0.4954, + "step": 17952 + }, + { + "epoch": 0.7948817285235159, + "grad_norm": 1.4361146688461304, + "learning_rate": 2.456690706687872e-06, + "loss": 0.4666, + "step": 17953 + }, + { + "epoch": 0.7949260042283298, + "grad_norm": 1.3421026468276978, + "learning_rate": 2.4556762169495017e-06, + "loss": 0.4329, + "step": 17954 + }, + { + "epoch": 0.7949702799331437, + "grad_norm": 1.4390608072280884, + "learning_rate": 2.4546619074011603e-06, + "loss": 0.5057, + "step": 17955 + }, + { + "epoch": 0.7950145556379575, + "grad_norm": 1.5936177968978882, + "learning_rate": 2.4536477780670774e-06, + "loss": 0.535, + "step": 17956 + }, + { + "epoch": 0.7950588313427714, + "grad_norm": 2.0214223861694336, + "learning_rate": 2.4526338289714726e-06, + "loss": 0.6229, + "step": 17957 + }, + { + "epoch": 0.7951031070475854, + "grad_norm": 1.6055846214294434, + "learning_rate": 2.4516200601385643e-06, + "loss": 0.5246, + "step": 17958 + }, + { + "epoch": 0.7951473827523992, + "grad_norm": 1.7826470136642456, + "learning_rate": 2.4506064715925615e-06, + "loss": 0.5748, + "step": 17959 + }, + { + "epoch": 0.7951916584572131, + "grad_norm": 1.4978880882263184, + "learning_rate": 2.4495930633576793e-06, + "loss": 0.4749, + "step": 17960 + }, + { + "epoch": 0.795235934162027, + "grad_norm": 1.4536696672439575, + "learning_rate": 2.448579835458118e-06, + "loss": 0.5184, + "step": 17961 + }, + { + "epoch": 0.7952802098668408, + "grad_norm": 1.5098657608032227, + "learning_rate": 2.44756678791808e-06, + "loss": 0.5565, + "step": 17962 + }, + { + "epoch": 0.7953244855716547, + "grad_norm": 1.7160778045654297, + "learning_rate": 2.4465539207617593e-06, + "loss": 0.5318, + "step": 17963 + }, + { + "epoch": 0.7953687612764686, + "grad_norm": 1.8235855102539062, + "learning_rate": 2.445541234013348e-06, + "loss": 0.598, + "step": 17964 + }, + { + "epoch": 0.7954130369812824, + "grad_norm": 1.5688881874084473, + "learning_rate": 2.444528727697031e-06, + "loss": 0.5202, + "step": 17965 + }, + { + "epoch": 0.7954573126860963, + "grad_norm": 1.5632891654968262, + "learning_rate": 2.443516401836995e-06, + "loss": 0.4894, + "step": 17966 + }, + { + "epoch": 0.7955015883909102, + "grad_norm": 1.4876080751419067, + "learning_rate": 2.4425042564574186e-06, + "loss": 0.5474, + "step": 17967 + }, + { + "epoch": 0.795545864095724, + "grad_norm": 1.6691645383834839, + "learning_rate": 2.4414922915824736e-06, + "loss": 0.5339, + "step": 17968 + }, + { + "epoch": 0.7955901398005379, + "grad_norm": 1.9210048913955688, + "learning_rate": 2.440480507236328e-06, + "loss": 0.5653, + "step": 17969 + }, + { + "epoch": 0.7956344155053519, + "grad_norm": 1.576029658317566, + "learning_rate": 2.439468903443153e-06, + "loss": 0.502, + "step": 17970 + }, + { + "epoch": 0.7956786912101657, + "grad_norm": 1.8586435317993164, + "learning_rate": 2.4384574802271077e-06, + "loss": 0.5457, + "step": 17971 + }, + { + "epoch": 0.7957229669149796, + "grad_norm": 2.636659622192383, + "learning_rate": 2.4374462376123486e-06, + "loss": 0.5757, + "step": 17972 + }, + { + "epoch": 0.7957672426197935, + "grad_norm": 1.4988362789154053, + "learning_rate": 2.4364351756230278e-06, + "loss": 0.5185, + "step": 17973 + }, + { + "epoch": 0.7958115183246073, + "grad_norm": 1.7153816223144531, + "learning_rate": 2.435424294283295e-06, + "loss": 0.5454, + "step": 17974 + }, + { + "epoch": 0.7958557940294212, + "grad_norm": 1.5397104024887085, + "learning_rate": 2.434413593617291e-06, + "loss": 0.5831, + "step": 17975 + }, + { + "epoch": 0.7959000697342351, + "grad_norm": 1.7418885231018066, + "learning_rate": 2.43340307364916e-06, + "loss": 0.5004, + "step": 17976 + }, + { + "epoch": 0.7959443454390489, + "grad_norm": 1.5993820428848267, + "learning_rate": 2.4323927344030364e-06, + "loss": 0.4844, + "step": 17977 + }, + { + "epoch": 0.7959886211438628, + "grad_norm": 1.5404105186462402, + "learning_rate": 2.4313825759030508e-06, + "loss": 0.5127, + "step": 17978 + }, + { + "epoch": 0.7960328968486767, + "grad_norm": 1.9329475164413452, + "learning_rate": 2.430372598173326e-06, + "loss": 0.5474, + "step": 17979 + }, + { + "epoch": 0.7960771725534905, + "grad_norm": 1.4163007736206055, + "learning_rate": 2.429362801237992e-06, + "loss": 0.5244, + "step": 17980 + }, + { + "epoch": 0.7961214482583044, + "grad_norm": 1.6558657884597778, + "learning_rate": 2.428353185121163e-06, + "loss": 0.4998, + "step": 17981 + }, + { + "epoch": 0.7961657239631184, + "grad_norm": 1.691946268081665, + "learning_rate": 2.4273437498469523e-06, + "loss": 0.5298, + "step": 17982 + }, + { + "epoch": 0.7962099996679323, + "grad_norm": 2.2194716930389404, + "learning_rate": 2.4263344954394706e-06, + "loss": 0.5503, + "step": 17983 + }, + { + "epoch": 0.7962542753727461, + "grad_norm": 1.5950483083724976, + "learning_rate": 2.425325421922824e-06, + "loss": 0.5116, + "step": 17984 + }, + { + "epoch": 0.79629855107756, + "grad_norm": 1.4307470321655273, + "learning_rate": 2.424316529321108e-06, + "loss": 0.5453, + "step": 17985 + }, + { + "epoch": 0.7963428267823739, + "grad_norm": 1.3188841342926025, + "learning_rate": 2.4233078176584267e-06, + "loss": 0.4571, + "step": 17986 + }, + { + "epoch": 0.7963871024871877, + "grad_norm": 1.3762792348861694, + "learning_rate": 2.422299286958869e-06, + "loss": 0.4886, + "step": 17987 + }, + { + "epoch": 0.7964313781920016, + "grad_norm": 1.4672125577926636, + "learning_rate": 2.421290937246523e-06, + "loss": 0.4791, + "step": 17988 + }, + { + "epoch": 0.7964756538968155, + "grad_norm": 1.3881306648254395, + "learning_rate": 2.420282768545469e-06, + "loss": 0.495, + "step": 17989 + }, + { + "epoch": 0.7965199296016293, + "grad_norm": 1.3561275005340576, + "learning_rate": 2.4192747808797956e-06, + "loss": 0.4857, + "step": 17990 + }, + { + "epoch": 0.7965642053064432, + "grad_norm": 1.4727493524551392, + "learning_rate": 2.4182669742735676e-06, + "loss": 0.5296, + "step": 17991 + }, + { + "epoch": 0.7966084810112571, + "grad_norm": 1.8164417743682861, + "learning_rate": 2.417259348750858e-06, + "loss": 0.5822, + "step": 17992 + }, + { + "epoch": 0.7966527567160709, + "grad_norm": 1.3147740364074707, + "learning_rate": 2.416251904335737e-06, + "loss": 0.4506, + "step": 17993 + }, + { + "epoch": 0.7966970324208849, + "grad_norm": 1.5584958791732788, + "learning_rate": 2.4152446410522647e-06, + "loss": 0.5009, + "step": 17994 + }, + { + "epoch": 0.7967413081256988, + "grad_norm": 1.7793668508529663, + "learning_rate": 2.414237558924496e-06, + "loss": 0.5267, + "step": 17995 + }, + { + "epoch": 0.7967855838305126, + "grad_norm": 2.0013082027435303, + "learning_rate": 2.4132306579764886e-06, + "loss": 0.5481, + "step": 17996 + }, + { + "epoch": 0.7968298595353265, + "grad_norm": 1.4586840867996216, + "learning_rate": 2.4122239382322897e-06, + "loss": 0.5116, + "step": 17997 + }, + { + "epoch": 0.7968741352401404, + "grad_norm": 1.5288830995559692, + "learning_rate": 2.411217399715944e-06, + "loss": 0.5076, + "step": 17998 + }, + { + "epoch": 0.7969184109449542, + "grad_norm": 1.3160818815231323, + "learning_rate": 2.4102110424514913e-06, + "loss": 0.4616, + "step": 17999 + }, + { + "epoch": 0.7969626866497681, + "grad_norm": 1.5498982667922974, + "learning_rate": 2.4092048664629685e-06, + "loss": 0.4988, + "step": 18000 + }, + { + "epoch": 0.797006962354582, + "grad_norm": 1.4904327392578125, + "learning_rate": 2.4081988717744066e-06, + "loss": 0.5198, + "step": 18001 + }, + { + "epoch": 0.7970512380593958, + "grad_norm": 1.536589503288269, + "learning_rate": 2.4071930584098314e-06, + "loss": 0.5019, + "step": 18002 + }, + { + "epoch": 0.7970955137642097, + "grad_norm": 1.5928181409835815, + "learning_rate": 2.406187426393269e-06, + "loss": 0.4948, + "step": 18003 + }, + { + "epoch": 0.7971397894690236, + "grad_norm": 1.4561761617660522, + "learning_rate": 2.405181975748737e-06, + "loss": 0.4227, + "step": 18004 + }, + { + "epoch": 0.7971840651738374, + "grad_norm": 1.3927843570709229, + "learning_rate": 2.4041767065002473e-06, + "loss": 0.476, + "step": 18005 + }, + { + "epoch": 0.7972283408786514, + "grad_norm": 1.7886343002319336, + "learning_rate": 2.4031716186718147e-06, + "loss": 0.5377, + "step": 18006 + }, + { + "epoch": 0.7972726165834653, + "grad_norm": 1.457119345664978, + "learning_rate": 2.4021667122874428e-06, + "loss": 0.5081, + "step": 18007 + }, + { + "epoch": 0.7973168922882791, + "grad_norm": 1.6437772512435913, + "learning_rate": 2.4011619873711324e-06, + "loss": 0.5149, + "step": 18008 + }, + { + "epoch": 0.797361167993093, + "grad_norm": 1.5301457643508911, + "learning_rate": 2.400157443946881e-06, + "loss": 0.5321, + "step": 18009 + }, + { + "epoch": 0.7974054436979069, + "grad_norm": 2.081630229949951, + "learning_rate": 2.399153082038681e-06, + "loss": 0.5746, + "step": 18010 + }, + { + "epoch": 0.7974497194027207, + "grad_norm": 1.7169692516326904, + "learning_rate": 2.398148901670521e-06, + "loss": 0.5577, + "step": 18011 + }, + { + "epoch": 0.7974939951075346, + "grad_norm": 1.6611082553863525, + "learning_rate": 2.397144902866381e-06, + "loss": 0.5678, + "step": 18012 + }, + { + "epoch": 0.7975382708123485, + "grad_norm": 2.0982320308685303, + "learning_rate": 2.396141085650248e-06, + "loss": 0.5503, + "step": 18013 + }, + { + "epoch": 0.7975825465171623, + "grad_norm": 1.3293980360031128, + "learning_rate": 2.3951374500460933e-06, + "loss": 0.4599, + "step": 18014 + }, + { + "epoch": 0.7976268222219762, + "grad_norm": 1.6131658554077148, + "learning_rate": 2.3941339960778865e-06, + "loss": 0.5065, + "step": 18015 + }, + { + "epoch": 0.7976710979267901, + "grad_norm": 1.4861465692520142, + "learning_rate": 2.393130723769599e-06, + "loss": 0.5279, + "step": 18016 + }, + { + "epoch": 0.797715373631604, + "grad_norm": 2.014659881591797, + "learning_rate": 2.392127633145189e-06, + "loss": 0.5388, + "step": 18017 + }, + { + "epoch": 0.7977596493364179, + "grad_norm": 1.3349268436431885, + "learning_rate": 2.3911247242286186e-06, + "loss": 0.4405, + "step": 18018 + }, + { + "epoch": 0.7978039250412318, + "grad_norm": 1.6475545167922974, + "learning_rate": 2.390121997043836e-06, + "loss": 0.5154, + "step": 18019 + }, + { + "epoch": 0.7978482007460457, + "grad_norm": 1.5530809164047241, + "learning_rate": 2.3891194516147952e-06, + "loss": 0.4778, + "step": 18020 + }, + { + "epoch": 0.7978924764508595, + "grad_norm": 1.4711313247680664, + "learning_rate": 2.388117087965439e-06, + "loss": 0.4919, + "step": 18021 + }, + { + "epoch": 0.7979367521556734, + "grad_norm": 1.324538230895996, + "learning_rate": 2.3871149061197053e-06, + "loss": 0.5094, + "step": 18022 + }, + { + "epoch": 0.7979810278604873, + "grad_norm": 1.4454495906829834, + "learning_rate": 2.3861129061015355e-06, + "loss": 0.518, + "step": 18023 + }, + { + "epoch": 0.7980253035653011, + "grad_norm": 1.3267122507095337, + "learning_rate": 2.3851110879348604e-06, + "loss": 0.4844, + "step": 18024 + }, + { + "epoch": 0.798069579270115, + "grad_norm": 1.4666367769241333, + "learning_rate": 2.3841094516436034e-06, + "loss": 0.5249, + "step": 18025 + }, + { + "epoch": 0.7981138549749289, + "grad_norm": 1.2679738998413086, + "learning_rate": 2.3831079972516935e-06, + "loss": 0.4594, + "step": 18026 + }, + { + "epoch": 0.7981581306797427, + "grad_norm": 1.8515347242355347, + "learning_rate": 2.3821067247830488e-06, + "loss": 0.4898, + "step": 18027 + }, + { + "epoch": 0.7982024063845566, + "grad_norm": 1.4700841903686523, + "learning_rate": 2.3811056342615814e-06, + "loss": 0.5135, + "step": 18028 + }, + { + "epoch": 0.7982466820893706, + "grad_norm": 1.3795630931854248, + "learning_rate": 2.380104725711202e-06, + "loss": 0.4829, + "step": 18029 + }, + { + "epoch": 0.7982909577941844, + "grad_norm": 1.4354798793792725, + "learning_rate": 2.3791039991558164e-06, + "loss": 0.4494, + "step": 18030 + }, + { + "epoch": 0.7983352334989983, + "grad_norm": 1.8341200351715088, + "learning_rate": 2.378103454619327e-06, + "loss": 0.5526, + "step": 18031 + }, + { + "epoch": 0.7983795092038122, + "grad_norm": 1.3743326663970947, + "learning_rate": 2.377103092125629e-06, + "loss": 0.4244, + "step": 18032 + }, + { + "epoch": 0.798423784908626, + "grad_norm": 1.261444330215454, + "learning_rate": 2.376102911698618e-06, + "loss": 0.4008, + "step": 18033 + }, + { + "epoch": 0.7984680606134399, + "grad_norm": 1.4539463520050049, + "learning_rate": 2.3751029133621826e-06, + "loss": 0.5073, + "step": 18034 + }, + { + "epoch": 0.7985123363182538, + "grad_norm": 2.752164125442505, + "learning_rate": 2.3741030971402026e-06, + "loss": 0.5475, + "step": 18035 + }, + { + "epoch": 0.7985566120230676, + "grad_norm": 1.324231743812561, + "learning_rate": 2.373103463056563e-06, + "loss": 0.488, + "step": 18036 + }, + { + "epoch": 0.7986008877278815, + "grad_norm": 1.497180461883545, + "learning_rate": 2.3721040111351403e-06, + "loss": 0.5096, + "step": 18037 + }, + { + "epoch": 0.7986451634326954, + "grad_norm": 1.4646674394607544, + "learning_rate": 2.3711047413998e-06, + "loss": 0.5438, + "step": 18038 + }, + { + "epoch": 0.7986894391375092, + "grad_norm": 1.8801026344299316, + "learning_rate": 2.3701056538744083e-06, + "loss": 0.4985, + "step": 18039 + }, + { + "epoch": 0.7987337148423231, + "grad_norm": 1.4839686155319214, + "learning_rate": 2.3691067485828334e-06, + "loss": 0.5292, + "step": 18040 + }, + { + "epoch": 0.798777990547137, + "grad_norm": 1.4057029485702515, + "learning_rate": 2.36810802554893e-06, + "loss": 0.4665, + "step": 18041 + }, + { + "epoch": 0.7988222662519509, + "grad_norm": 1.4578332901000977, + "learning_rate": 2.3671094847965505e-06, + "loss": 0.5223, + "step": 18042 + }, + { + "epoch": 0.7988665419567648, + "grad_norm": 1.2940417528152466, + "learning_rate": 2.3661111263495472e-06, + "loss": 0.4485, + "step": 18043 + }, + { + "epoch": 0.7989108176615787, + "grad_norm": 1.520461916923523, + "learning_rate": 2.365112950231765e-06, + "loss": 0.4581, + "step": 18044 + }, + { + "epoch": 0.7989550933663925, + "grad_norm": 1.4811058044433594, + "learning_rate": 2.364114956467041e-06, + "loss": 0.5358, + "step": 18045 + }, + { + "epoch": 0.7989993690712064, + "grad_norm": 1.601718544960022, + "learning_rate": 2.363117145079219e-06, + "loss": 0.5213, + "step": 18046 + }, + { + "epoch": 0.7990436447760203, + "grad_norm": 1.6147756576538086, + "learning_rate": 2.3621195160921227e-06, + "loss": 0.5208, + "step": 18047 + }, + { + "epoch": 0.7990879204808341, + "grad_norm": 1.5437591075897217, + "learning_rate": 2.3611220695295833e-06, + "loss": 0.4976, + "step": 18048 + }, + { + "epoch": 0.799132196185648, + "grad_norm": 1.6992950439453125, + "learning_rate": 2.3601248054154203e-06, + "loss": 0.5271, + "step": 18049 + }, + { + "epoch": 0.7991764718904619, + "grad_norm": 1.4492008686065674, + "learning_rate": 2.3591277237734587e-06, + "loss": 0.5083, + "step": 18050 + }, + { + "epoch": 0.7992207475952758, + "grad_norm": 1.3087483644485474, + "learning_rate": 2.3581308246275103e-06, + "loss": 0.4303, + "step": 18051 + }, + { + "epoch": 0.7992650233000896, + "grad_norm": 1.5980815887451172, + "learning_rate": 2.3571341080013823e-06, + "loss": 0.5389, + "step": 18052 + }, + { + "epoch": 0.7993092990049036, + "grad_norm": 1.2812074422836304, + "learning_rate": 2.356137573918885e-06, + "loss": 0.4325, + "step": 18053 + }, + { + "epoch": 0.7993535747097175, + "grad_norm": 1.4087063074111938, + "learning_rate": 2.3551412224038182e-06, + "loss": 0.4373, + "step": 18054 + }, + { + "epoch": 0.7993978504145313, + "grad_norm": 1.4317342042922974, + "learning_rate": 2.3541450534799783e-06, + "loss": 0.499, + "step": 18055 + }, + { + "epoch": 0.7994421261193452, + "grad_norm": 1.3241779804229736, + "learning_rate": 2.353149067171159e-06, + "loss": 0.4863, + "step": 18056 + }, + { + "epoch": 0.7994864018241591, + "grad_norm": 1.5676943063735962, + "learning_rate": 2.352153263501148e-06, + "loss": 0.5121, + "step": 18057 + }, + { + "epoch": 0.7995306775289729, + "grad_norm": 1.2197777032852173, + "learning_rate": 2.3511576424937286e-06, + "loss": 0.4578, + "step": 18058 + }, + { + "epoch": 0.7995749532337868, + "grad_norm": 1.7172603607177734, + "learning_rate": 2.350162204172678e-06, + "loss": 0.4409, + "step": 18059 + }, + { + "epoch": 0.7996192289386007, + "grad_norm": 1.3288589715957642, + "learning_rate": 2.3491669485617775e-06, + "loss": 0.4788, + "step": 18060 + }, + { + "epoch": 0.7996635046434145, + "grad_norm": 1.6595757007598877, + "learning_rate": 2.3481718756847948e-06, + "loss": 0.5562, + "step": 18061 + }, + { + "epoch": 0.7997077803482284, + "grad_norm": 2.1983041763305664, + "learning_rate": 2.347176985565495e-06, + "loss": 0.5357, + "step": 18062 + }, + { + "epoch": 0.7997520560530423, + "grad_norm": 1.6231316328048706, + "learning_rate": 2.346182278227642e-06, + "loss": 0.4964, + "step": 18063 + }, + { + "epoch": 0.7997963317578561, + "grad_norm": 1.5619474649429321, + "learning_rate": 2.345187753694995e-06, + "loss": 0.4798, + "step": 18064 + }, + { + "epoch": 0.79984060746267, + "grad_norm": 1.3972129821777344, + "learning_rate": 2.344193411991306e-06, + "loss": 0.4352, + "step": 18065 + }, + { + "epoch": 0.799884883167484, + "grad_norm": 1.4599579572677612, + "learning_rate": 2.3431992531403237e-06, + "loss": 0.485, + "step": 18066 + }, + { + "epoch": 0.7999291588722978, + "grad_norm": 1.7502321004867554, + "learning_rate": 2.342205277165792e-06, + "loss": 0.5988, + "step": 18067 + }, + { + "epoch": 0.7999734345771117, + "grad_norm": 1.5800602436065674, + "learning_rate": 2.3412114840914523e-06, + "loss": 0.5031, + "step": 18068 + }, + { + "epoch": 0.8000177102819256, + "grad_norm": 1.9871344566345215, + "learning_rate": 2.3402178739410375e-06, + "loss": 0.5415, + "step": 18069 + }, + { + "epoch": 0.8000619859867394, + "grad_norm": 1.4430712461471558, + "learning_rate": 2.3392244467382853e-06, + "loss": 0.4336, + "step": 18070 + }, + { + "epoch": 0.8001062616915533, + "grad_norm": 1.5152723789215088, + "learning_rate": 2.3382312025069185e-06, + "loss": 0.4264, + "step": 18071 + }, + { + "epoch": 0.8001505373963672, + "grad_norm": 1.595537543296814, + "learning_rate": 2.3372381412706592e-06, + "loss": 0.5551, + "step": 18072 + }, + { + "epoch": 0.800194813101181, + "grad_norm": 1.7484922409057617, + "learning_rate": 2.3362452630532294e-06, + "loss": 0.5538, + "step": 18073 + }, + { + "epoch": 0.8002390888059949, + "grad_norm": 1.6787054538726807, + "learning_rate": 2.335252567878342e-06, + "loss": 0.5227, + "step": 18074 + }, + { + "epoch": 0.8002833645108088, + "grad_norm": 1.4212169647216797, + "learning_rate": 2.334260055769707e-06, + "loss": 0.5415, + "step": 18075 + }, + { + "epoch": 0.8003276402156226, + "grad_norm": 1.310140609741211, + "learning_rate": 2.3332677267510283e-06, + "loss": 0.4085, + "step": 18076 + }, + { + "epoch": 0.8003719159204365, + "grad_norm": 1.525205135345459, + "learning_rate": 2.3322755808460072e-06, + "loss": 0.5171, + "step": 18077 + }, + { + "epoch": 0.8004161916252505, + "grad_norm": 1.2764838933944702, + "learning_rate": 2.3312836180783405e-06, + "loss": 0.4055, + "step": 18078 + }, + { + "epoch": 0.8004604673300643, + "grad_norm": 1.2213475704193115, + "learning_rate": 2.3302918384717175e-06, + "loss": 0.5057, + "step": 18079 + }, + { + "epoch": 0.8005047430348782, + "grad_norm": 2.054595470428467, + "learning_rate": 2.3293002420498323e-06, + "loss": 0.5347, + "step": 18080 + }, + { + "epoch": 0.8005490187396921, + "grad_norm": 1.457794189453125, + "learning_rate": 2.3283088288363655e-06, + "loss": 0.5303, + "step": 18081 + }, + { + "epoch": 0.8005932944445059, + "grad_norm": 1.3619928359985352, + "learning_rate": 2.3273175988549923e-06, + "loss": 0.4571, + "step": 18082 + }, + { + "epoch": 0.8006375701493198, + "grad_norm": 1.5150974988937378, + "learning_rate": 2.326326552129393e-06, + "loss": 0.4401, + "step": 18083 + }, + { + "epoch": 0.8006818458541337, + "grad_norm": 1.6079045534133911, + "learning_rate": 2.325335688683238e-06, + "loss": 0.5075, + "step": 18084 + }, + { + "epoch": 0.8007261215589475, + "grad_norm": 1.4936025142669678, + "learning_rate": 2.324345008540192e-06, + "loss": 0.485, + "step": 18085 + }, + { + "epoch": 0.8007703972637614, + "grad_norm": 1.5113581418991089, + "learning_rate": 2.323354511723911e-06, + "loss": 0.5161, + "step": 18086 + }, + { + "epoch": 0.8008146729685753, + "grad_norm": 1.593999981880188, + "learning_rate": 2.322364198258059e-06, + "loss": 0.4765, + "step": 18087 + }, + { + "epoch": 0.8008589486733892, + "grad_norm": 1.4446802139282227, + "learning_rate": 2.3213740681662877e-06, + "loss": 0.5076, + "step": 18088 + }, + { + "epoch": 0.800903224378203, + "grad_norm": 1.871450662612915, + "learning_rate": 2.320384121472241e-06, + "loss": 0.5523, + "step": 18089 + }, + { + "epoch": 0.800947500083017, + "grad_norm": 1.7879571914672852, + "learning_rate": 2.3193943581995705e-06, + "loss": 0.5171, + "step": 18090 + }, + { + "epoch": 0.8009917757878309, + "grad_norm": 1.381894588470459, + "learning_rate": 2.318404778371911e-06, + "loss": 0.475, + "step": 18091 + }, + { + "epoch": 0.8010360514926447, + "grad_norm": 1.6896133422851562, + "learning_rate": 2.3174153820128962e-06, + "loss": 0.4715, + "step": 18092 + }, + { + "epoch": 0.8010803271974586, + "grad_norm": 1.402559757232666, + "learning_rate": 2.316426169146165e-06, + "loss": 0.4313, + "step": 18093 + }, + { + "epoch": 0.8011246029022725, + "grad_norm": 1.539349913597107, + "learning_rate": 2.315437139795337e-06, + "loss": 0.506, + "step": 18094 + }, + { + "epoch": 0.8011688786070863, + "grad_norm": 1.2047785520553589, + "learning_rate": 2.314448293984034e-06, + "loss": 0.4096, + "step": 18095 + }, + { + "epoch": 0.8012131543119002, + "grad_norm": 1.6021478176116943, + "learning_rate": 2.3134596317358736e-06, + "loss": 0.5422, + "step": 18096 + }, + { + "epoch": 0.8012574300167141, + "grad_norm": 1.6355384588241577, + "learning_rate": 2.3124711530744747e-06, + "loss": 0.4636, + "step": 18097 + }, + { + "epoch": 0.8013017057215279, + "grad_norm": 1.4142656326293945, + "learning_rate": 2.3114828580234426e-06, + "loss": 0.4698, + "step": 18098 + }, + { + "epoch": 0.8013459814263418, + "grad_norm": 1.778838872909546, + "learning_rate": 2.3104947466063785e-06, + "loss": 0.5206, + "step": 18099 + }, + { + "epoch": 0.8013902571311557, + "grad_norm": 1.9570544958114624, + "learning_rate": 2.3095068188468907e-06, + "loss": 0.5828, + "step": 18100 + }, + { + "epoch": 0.8014345328359695, + "grad_norm": 1.792050838470459, + "learning_rate": 2.3085190747685705e-06, + "loss": 0.6115, + "step": 18101 + }, + { + "epoch": 0.8014788085407835, + "grad_norm": 1.9241644144058228, + "learning_rate": 2.3075315143950084e-06, + "loss": 0.4583, + "step": 18102 + }, + { + "epoch": 0.8015230842455974, + "grad_norm": 1.464115023612976, + "learning_rate": 2.3065441377497933e-06, + "loss": 0.467, + "step": 18103 + }, + { + "epoch": 0.8015673599504112, + "grad_norm": 1.3666342496871948, + "learning_rate": 2.3055569448565075e-06, + "loss": 0.4616, + "step": 18104 + }, + { + "epoch": 0.8016116356552251, + "grad_norm": 1.282287359237671, + "learning_rate": 2.3045699357387286e-06, + "loss": 0.4792, + "step": 18105 + }, + { + "epoch": 0.801655911360039, + "grad_norm": 1.431117296218872, + "learning_rate": 2.3035831104200277e-06, + "loss": 0.4947, + "step": 18106 + }, + { + "epoch": 0.8017001870648528, + "grad_norm": 1.2752238512039185, + "learning_rate": 2.302596468923981e-06, + "loss": 0.467, + "step": 18107 + }, + { + "epoch": 0.8017444627696667, + "grad_norm": 1.5406349897384644, + "learning_rate": 2.3016100112741504e-06, + "loss": 0.5262, + "step": 18108 + }, + { + "epoch": 0.8017887384744806, + "grad_norm": 1.4656713008880615, + "learning_rate": 2.300623737494093e-06, + "loss": 0.5197, + "step": 18109 + }, + { + "epoch": 0.8018330141792944, + "grad_norm": 1.4129931926727295, + "learning_rate": 2.2996376476073724e-06, + "loss": 0.4578, + "step": 18110 + }, + { + "epoch": 0.8018772898841083, + "grad_norm": 1.627893328666687, + "learning_rate": 2.2986517416375353e-06, + "loss": 0.547, + "step": 18111 + }, + { + "epoch": 0.8019215655889222, + "grad_norm": 1.458099603652954, + "learning_rate": 2.2976660196081312e-06, + "loss": 0.4861, + "step": 18112 + }, + { + "epoch": 0.801965841293736, + "grad_norm": 1.398380160331726, + "learning_rate": 2.2966804815427035e-06, + "loss": 0.4867, + "step": 18113 + }, + { + "epoch": 0.80201011699855, + "grad_norm": 1.705053448677063, + "learning_rate": 2.2956951274647887e-06, + "loss": 0.5182, + "step": 18114 + }, + { + "epoch": 0.8020543927033639, + "grad_norm": 1.4442565441131592, + "learning_rate": 2.2947099573979236e-06, + "loss": 0.4772, + "step": 18115 + }, + { + "epoch": 0.8020986684081777, + "grad_norm": 1.505730390548706, + "learning_rate": 2.293724971365635e-06, + "loss": 0.5068, + "step": 18116 + }, + { + "epoch": 0.8021429441129916, + "grad_norm": 1.8492019176483154, + "learning_rate": 2.292740169391453e-06, + "loss": 0.5407, + "step": 18117 + }, + { + "epoch": 0.8021872198178055, + "grad_norm": 1.3192511796951294, + "learning_rate": 2.2917555514988965e-06, + "loss": 0.4227, + "step": 18118 + }, + { + "epoch": 0.8022314955226193, + "grad_norm": 1.541334867477417, + "learning_rate": 2.29077111771148e-06, + "loss": 0.5065, + "step": 18119 + }, + { + "epoch": 0.8022757712274332, + "grad_norm": 1.4728469848632812, + "learning_rate": 2.28978686805272e-06, + "loss": 0.4878, + "step": 18120 + }, + { + "epoch": 0.8023200469322471, + "grad_norm": 1.446730613708496, + "learning_rate": 2.288802802546124e-06, + "loss": 0.4735, + "step": 18121 + }, + { + "epoch": 0.802364322637061, + "grad_norm": 1.3020168542861938, + "learning_rate": 2.2878189212151945e-06, + "loss": 0.4034, + "step": 18122 + }, + { + "epoch": 0.8024085983418748, + "grad_norm": 1.5847865343093872, + "learning_rate": 2.2868352240834304e-06, + "loss": 0.5179, + "step": 18123 + }, + { + "epoch": 0.8024528740466887, + "grad_norm": 1.3143621683120728, + "learning_rate": 2.285851711174326e-06, + "loss": 0.4482, + "step": 18124 + }, + { + "epoch": 0.8024971497515027, + "grad_norm": 1.405225157737732, + "learning_rate": 2.2848683825113726e-06, + "loss": 0.4935, + "step": 18125 + }, + { + "epoch": 0.8025414254563165, + "grad_norm": 1.1336513757705688, + "learning_rate": 2.283885238118053e-06, + "loss": 0.4278, + "step": 18126 + }, + { + "epoch": 0.8025857011611304, + "grad_norm": 1.4630913734436035, + "learning_rate": 2.2829022780178544e-06, + "loss": 0.4954, + "step": 18127 + }, + { + "epoch": 0.8026299768659443, + "grad_norm": 1.5128813982009888, + "learning_rate": 2.281919502234252e-06, + "loss": 0.5185, + "step": 18128 + }, + { + "epoch": 0.8026742525707581, + "grad_norm": 1.7148680686950684, + "learning_rate": 2.2809369107907153e-06, + "loss": 0.5315, + "step": 18129 + }, + { + "epoch": 0.802718528275572, + "grad_norm": 1.3074265718460083, + "learning_rate": 2.2799545037107174e-06, + "loss": 0.4873, + "step": 18130 + }, + { + "epoch": 0.8027628039803859, + "grad_norm": 1.9188183546066284, + "learning_rate": 2.278972281017722e-06, + "loss": 0.5531, + "step": 18131 + }, + { + "epoch": 0.8028070796851997, + "grad_norm": 1.7253271341323853, + "learning_rate": 2.2779902427351852e-06, + "loss": 0.5013, + "step": 18132 + }, + { + "epoch": 0.8028513553900136, + "grad_norm": 1.5392106771469116, + "learning_rate": 2.2770083888865647e-06, + "loss": 0.4768, + "step": 18133 + }, + { + "epoch": 0.8028956310948275, + "grad_norm": 1.3465322256088257, + "learning_rate": 2.27602671949531e-06, + "loss": 0.4668, + "step": 18134 + }, + { + "epoch": 0.8029399067996413, + "grad_norm": 1.4602140188217163, + "learning_rate": 2.2750452345848684e-06, + "loss": 0.4939, + "step": 18135 + }, + { + "epoch": 0.8029841825044552, + "grad_norm": 1.4856032133102417, + "learning_rate": 2.2740639341786784e-06, + "loss": 0.5138, + "step": 18136 + }, + { + "epoch": 0.8030284582092692, + "grad_norm": 1.593833565711975, + "learning_rate": 2.2730828183001843e-06, + "loss": 0.5621, + "step": 18137 + }, + { + "epoch": 0.803072733914083, + "grad_norm": 1.3720611333847046, + "learning_rate": 2.272101886972815e-06, + "loss": 0.4743, + "step": 18138 + }, + { + "epoch": 0.8031170096188969, + "grad_norm": 1.260675311088562, + "learning_rate": 2.271121140219997e-06, + "loss": 0.4633, + "step": 18139 + }, + { + "epoch": 0.8031612853237108, + "grad_norm": 1.42462158203125, + "learning_rate": 2.27014057806516e-06, + "loss": 0.4901, + "step": 18140 + }, + { + "epoch": 0.8032055610285246, + "grad_norm": 1.500470519065857, + "learning_rate": 2.2691602005317248e-06, + "loss": 0.4928, + "step": 18141 + }, + { + "epoch": 0.8032498367333385, + "grad_norm": 1.476915717124939, + "learning_rate": 2.2681800076430972e-06, + "loss": 0.4986, + "step": 18142 + }, + { + "epoch": 0.8032941124381524, + "grad_norm": 1.324249505996704, + "learning_rate": 2.267199999422698e-06, + "loss": 0.4091, + "step": 18143 + }, + { + "epoch": 0.8033383881429662, + "grad_norm": 1.372315526008606, + "learning_rate": 2.2662201758939306e-06, + "loss": 0.4963, + "step": 18144 + }, + { + "epoch": 0.8033826638477801, + "grad_norm": 1.3723556995391846, + "learning_rate": 2.265240537080198e-06, + "loss": 0.4571, + "step": 18145 + }, + { + "epoch": 0.803426939552594, + "grad_norm": 1.6490200757980347, + "learning_rate": 2.264261083004893e-06, + "loss": 0.4662, + "step": 18146 + }, + { + "epoch": 0.8034712152574078, + "grad_norm": 1.629889726638794, + "learning_rate": 2.263281813691417e-06, + "loss": 0.5155, + "step": 18147 + }, + { + "epoch": 0.8035154909622217, + "grad_norm": 1.4925971031188965, + "learning_rate": 2.262302729163155e-06, + "loss": 0.488, + "step": 18148 + }, + { + "epoch": 0.8035597666670357, + "grad_norm": 1.319948434829712, + "learning_rate": 2.2613238294434914e-06, + "loss": 0.4611, + "step": 18149 + }, + { + "epoch": 0.8036040423718495, + "grad_norm": 1.71792733669281, + "learning_rate": 2.2603451145558088e-06, + "loss": 0.5556, + "step": 18150 + }, + { + "epoch": 0.8036483180766634, + "grad_norm": 1.6757044792175293, + "learning_rate": 2.25936658452348e-06, + "loss": 0.5219, + "step": 18151 + }, + { + "epoch": 0.8036925937814773, + "grad_norm": 1.7116446495056152, + "learning_rate": 2.258388239369875e-06, + "loss": 0.5866, + "step": 18152 + }, + { + "epoch": 0.8037368694862911, + "grad_norm": 1.3377611637115479, + "learning_rate": 2.257410079118366e-06, + "loss": 0.4727, + "step": 18153 + }, + { + "epoch": 0.803781145191105, + "grad_norm": 1.4652721881866455, + "learning_rate": 2.2564321037923144e-06, + "loss": 0.4843, + "step": 18154 + }, + { + "epoch": 0.8038254208959189, + "grad_norm": 1.4459723234176636, + "learning_rate": 2.2554543134150754e-06, + "loss": 0.4835, + "step": 18155 + }, + { + "epoch": 0.8038696966007327, + "grad_norm": 1.842429757118225, + "learning_rate": 2.2544767080100017e-06, + "loss": 0.5278, + "step": 18156 + }, + { + "epoch": 0.8039139723055466, + "grad_norm": 1.6254802942276, + "learning_rate": 2.253499287600449e-06, + "loss": 0.5544, + "step": 18157 + }, + { + "epoch": 0.8039582480103605, + "grad_norm": 1.5442675352096558, + "learning_rate": 2.252522052209758e-06, + "loss": 0.5534, + "step": 18158 + }, + { + "epoch": 0.8040025237151744, + "grad_norm": 1.7813639640808105, + "learning_rate": 2.2515450018612682e-06, + "loss": 0.5132, + "step": 18159 + }, + { + "epoch": 0.8040467994199882, + "grad_norm": 1.737777590751648, + "learning_rate": 2.2505681365783182e-06, + "loss": 0.5131, + "step": 18160 + }, + { + "epoch": 0.8040910751248022, + "grad_norm": 1.4424357414245605, + "learning_rate": 2.249591456384239e-06, + "loss": 0.4494, + "step": 18161 + }, + { + "epoch": 0.8041353508296161, + "grad_norm": 1.3607971668243408, + "learning_rate": 2.2486149613023534e-06, + "loss": 0.4651, + "step": 18162 + }, + { + "epoch": 0.8041796265344299, + "grad_norm": 1.2431095838546753, + "learning_rate": 2.247638651355991e-06, + "loss": 0.4389, + "step": 18163 + }, + { + "epoch": 0.8042239022392438, + "grad_norm": 1.5237208604812622, + "learning_rate": 2.246662526568468e-06, + "loss": 0.485, + "step": 18164 + }, + { + "epoch": 0.8042681779440577, + "grad_norm": 1.4268317222595215, + "learning_rate": 2.2456865869630974e-06, + "loss": 0.4609, + "step": 18165 + }, + { + "epoch": 0.8043124536488715, + "grad_norm": 1.2176436185836792, + "learning_rate": 2.244710832563185e-06, + "loss": 0.4302, + "step": 18166 + }, + { + "epoch": 0.8043567293536854, + "grad_norm": 1.4662059545516968, + "learning_rate": 2.243735263392044e-06, + "loss": 0.5188, + "step": 18167 + }, + { + "epoch": 0.8044010050584993, + "grad_norm": 1.2859123945236206, + "learning_rate": 2.24275987947297e-06, + "loss": 0.4338, + "step": 18168 + }, + { + "epoch": 0.8044452807633131, + "grad_norm": 1.3121371269226074, + "learning_rate": 2.24178468082926e-06, + "loss": 0.487, + "step": 18169 + }, + { + "epoch": 0.804489556468127, + "grad_norm": 1.5539906024932861, + "learning_rate": 2.240809667484206e-06, + "loss": 0.4663, + "step": 18170 + }, + { + "epoch": 0.8045338321729409, + "grad_norm": 1.3343716859817505, + "learning_rate": 2.2398348394610947e-06, + "loss": 0.4373, + "step": 18171 + }, + { + "epoch": 0.8045781078777547, + "grad_norm": 1.3959693908691406, + "learning_rate": 2.2388601967832114e-06, + "loss": 0.5113, + "step": 18172 + }, + { + "epoch": 0.8046223835825687, + "grad_norm": 1.5508092641830444, + "learning_rate": 2.2378857394738284e-06, + "loss": 0.4662, + "step": 18173 + }, + { + "epoch": 0.8046666592873826, + "grad_norm": 1.6802558898925781, + "learning_rate": 2.2369114675562287e-06, + "loss": 0.5035, + "step": 18174 + }, + { + "epoch": 0.8047109349921964, + "grad_norm": 1.7231972217559814, + "learning_rate": 2.2359373810536765e-06, + "loss": 0.5201, + "step": 18175 + }, + { + "epoch": 0.8047552106970103, + "grad_norm": 1.7793420553207397, + "learning_rate": 2.234963479989436e-06, + "loss": 0.5516, + "step": 18176 + }, + { + "epoch": 0.8047994864018242, + "grad_norm": 1.4206960201263428, + "learning_rate": 2.2339897643867724e-06, + "loss": 0.4449, + "step": 18177 + }, + { + "epoch": 0.804843762106638, + "grad_norm": 1.6310343742370605, + "learning_rate": 2.233016234268941e-06, + "loss": 0.4886, + "step": 18178 + }, + { + "epoch": 0.8048880378114519, + "grad_norm": 1.4894452095031738, + "learning_rate": 2.2320428896591916e-06, + "loss": 0.4931, + "step": 18179 + }, + { + "epoch": 0.8049323135162658, + "grad_norm": 1.3382502794265747, + "learning_rate": 2.231069730580774e-06, + "loss": 0.4449, + "step": 18180 + }, + { + "epoch": 0.8049765892210796, + "grad_norm": 1.5069456100463867, + "learning_rate": 2.23009675705693e-06, + "loss": 0.4823, + "step": 18181 + }, + { + "epoch": 0.8050208649258935, + "grad_norm": 1.5760917663574219, + "learning_rate": 2.2291239691108978e-06, + "loss": 0.5466, + "step": 18182 + }, + { + "epoch": 0.8050651406307074, + "grad_norm": 1.1494258642196655, + "learning_rate": 2.2281513667659093e-06, + "loss": 0.3966, + "step": 18183 + }, + { + "epoch": 0.8051094163355212, + "grad_norm": 1.5323797464370728, + "learning_rate": 2.227178950045201e-06, + "loss": 0.4954, + "step": 18184 + }, + { + "epoch": 0.8051536920403352, + "grad_norm": 1.4683369398117065, + "learning_rate": 2.226206718971994e-06, + "loss": 0.4672, + "step": 18185 + }, + { + "epoch": 0.8051979677451491, + "grad_norm": 1.3970167636871338, + "learning_rate": 2.2252346735695075e-06, + "loss": 0.511, + "step": 18186 + }, + { + "epoch": 0.8052422434499629, + "grad_norm": 1.7033737897872925, + "learning_rate": 2.224262813860962e-06, + "loss": 0.5132, + "step": 18187 + }, + { + "epoch": 0.8052865191547768, + "grad_norm": 1.6192563772201538, + "learning_rate": 2.223291139869572e-06, + "loss": 0.4781, + "step": 18188 + }, + { + "epoch": 0.8053307948595907, + "grad_norm": 1.5314197540283203, + "learning_rate": 2.2223196516185353e-06, + "loss": 0.5095, + "step": 18189 + }, + { + "epoch": 0.8053750705644045, + "grad_norm": 2.1309432983398438, + "learning_rate": 2.2213483491310637e-06, + "loss": 0.6662, + "step": 18190 + }, + { + "epoch": 0.8054193462692184, + "grad_norm": 1.3645061254501343, + "learning_rate": 2.220377232430353e-06, + "loss": 0.4717, + "step": 18191 + }, + { + "epoch": 0.8054636219740323, + "grad_norm": 1.4718170166015625, + "learning_rate": 2.2194063015395983e-06, + "loss": 0.4877, + "step": 18192 + }, + { + "epoch": 0.8055078976788462, + "grad_norm": 1.5444016456604004, + "learning_rate": 2.218435556481986e-06, + "loss": 0.5287, + "step": 18193 + }, + { + "epoch": 0.80555217338366, + "grad_norm": 1.5263985395431519, + "learning_rate": 2.2174649972807062e-06, + "loss": 0.5278, + "step": 18194 + }, + { + "epoch": 0.8055964490884739, + "grad_norm": 1.5496693849563599, + "learning_rate": 2.216494623958939e-06, + "loss": 0.4922, + "step": 18195 + }, + { + "epoch": 0.8056407247932879, + "grad_norm": 1.5760682821273804, + "learning_rate": 2.2155244365398564e-06, + "loss": 0.5665, + "step": 18196 + }, + { + "epoch": 0.8056850004981017, + "grad_norm": 1.6363837718963623, + "learning_rate": 2.214554435046641e-06, + "loss": 0.5391, + "step": 18197 + }, + { + "epoch": 0.8057292762029156, + "grad_norm": 1.6946645975112915, + "learning_rate": 2.213584619502451e-06, + "loss": 0.5245, + "step": 18198 + }, + { + "epoch": 0.8057735519077295, + "grad_norm": 1.4735922813415527, + "learning_rate": 2.2126149899304494e-06, + "loss": 0.5662, + "step": 18199 + }, + { + "epoch": 0.8058178276125433, + "grad_norm": 1.6563881635665894, + "learning_rate": 2.2116455463538e-06, + "loss": 0.5484, + "step": 18200 + }, + { + "epoch": 0.8058621033173572, + "grad_norm": 1.6180356740951538, + "learning_rate": 2.2106762887956556e-06, + "loss": 0.5256, + "step": 18201 + }, + { + "epoch": 0.8059063790221711, + "grad_norm": 1.3119951486587524, + "learning_rate": 2.209707217279167e-06, + "loss": 0.4566, + "step": 18202 + }, + { + "epoch": 0.8059506547269849, + "grad_norm": 1.6504169702529907, + "learning_rate": 2.2087383318274745e-06, + "loss": 0.5535, + "step": 18203 + }, + { + "epoch": 0.8059949304317988, + "grad_norm": 1.327294945716858, + "learning_rate": 2.2077696324637254e-06, + "loss": 0.4799, + "step": 18204 + }, + { + "epoch": 0.8060392061366127, + "grad_norm": 1.5725021362304688, + "learning_rate": 2.206801119211055e-06, + "loss": 0.4945, + "step": 18205 + }, + { + "epoch": 0.8060834818414265, + "grad_norm": 1.5040141344070435, + "learning_rate": 2.205832792092594e-06, + "loss": 0.5387, + "step": 18206 + }, + { + "epoch": 0.8061277575462404, + "grad_norm": 1.565573811531067, + "learning_rate": 2.2048646511314697e-06, + "loss": 0.5189, + "step": 18207 + }, + { + "epoch": 0.8061720332510544, + "grad_norm": 1.97048020362854, + "learning_rate": 2.203896696350808e-06, + "loss": 0.6244, + "step": 18208 + }, + { + "epoch": 0.8062163089558682, + "grad_norm": 1.5341823101043701, + "learning_rate": 2.202928927773722e-06, + "loss": 0.5361, + "step": 18209 + }, + { + "epoch": 0.8062605846606821, + "grad_norm": 1.6959172487258911, + "learning_rate": 2.2019613454233324e-06, + "loss": 0.5383, + "step": 18210 + }, + { + "epoch": 0.806304860365496, + "grad_norm": 1.4132487773895264, + "learning_rate": 2.2009939493227472e-06, + "loss": 0.51, + "step": 18211 + }, + { + "epoch": 0.8063491360703098, + "grad_norm": 1.353259801864624, + "learning_rate": 2.2000267394950715e-06, + "loss": 0.5111, + "step": 18212 + }, + { + "epoch": 0.8063934117751237, + "grad_norm": 1.6667529344558716, + "learning_rate": 2.199059715963402e-06, + "loss": 0.4889, + "step": 18213 + }, + { + "epoch": 0.8064376874799376, + "grad_norm": 1.5958800315856934, + "learning_rate": 2.198092878750844e-06, + "loss": 0.4943, + "step": 18214 + }, + { + "epoch": 0.8064819631847514, + "grad_norm": 1.3006616830825806, + "learning_rate": 2.197126227880484e-06, + "loss": 0.4603, + "step": 18215 + }, + { + "epoch": 0.8065262388895653, + "grad_norm": 1.3475478887557983, + "learning_rate": 2.1961597633754106e-06, + "loss": 0.4539, + "step": 18216 + }, + { + "epoch": 0.8065705145943792, + "grad_norm": 1.6812286376953125, + "learning_rate": 2.1951934852587077e-06, + "loss": 0.4753, + "step": 18217 + }, + { + "epoch": 0.806614790299193, + "grad_norm": 1.2585961818695068, + "learning_rate": 2.194227393553453e-06, + "loss": 0.4697, + "step": 18218 + }, + { + "epoch": 0.8066590660040069, + "grad_norm": 1.6204026937484741, + "learning_rate": 2.1932614882827196e-06, + "loss": 0.5716, + "step": 18219 + }, + { + "epoch": 0.8067033417088209, + "grad_norm": 1.2656660079956055, + "learning_rate": 2.192295769469581e-06, + "loss": 0.4548, + "step": 18220 + }, + { + "epoch": 0.8067476174136347, + "grad_norm": 1.6396238803863525, + "learning_rate": 2.1913302371371003e-06, + "loss": 0.515, + "step": 18221 + }, + { + "epoch": 0.8067918931184486, + "grad_norm": 1.8887157440185547, + "learning_rate": 2.19036489130834e-06, + "loss": 0.5082, + "step": 18222 + }, + { + "epoch": 0.8068361688232625, + "grad_norm": 1.20513117313385, + "learning_rate": 2.189399732006352e-06, + "loss": 0.4545, + "step": 18223 + }, + { + "epoch": 0.8068804445280763, + "grad_norm": 1.2395222187042236, + "learning_rate": 2.1884347592541942e-06, + "loss": 0.484, + "step": 18224 + }, + { + "epoch": 0.8069247202328902, + "grad_norm": 1.5866634845733643, + "learning_rate": 2.1874699730749116e-06, + "loss": 0.5003, + "step": 18225 + }, + { + "epoch": 0.8069689959377041, + "grad_norm": 1.8446851968765259, + "learning_rate": 2.1865053734915485e-06, + "loss": 0.6034, + "step": 18226 + }, + { + "epoch": 0.8070132716425179, + "grad_norm": 1.4814445972442627, + "learning_rate": 2.1855409605271416e-06, + "loss": 0.4621, + "step": 18227 + }, + { + "epoch": 0.8070575473473318, + "grad_norm": 1.6778334379196167, + "learning_rate": 2.1845767342047273e-06, + "loss": 0.4773, + "step": 18228 + }, + { + "epoch": 0.8071018230521457, + "grad_norm": 1.3664371967315674, + "learning_rate": 2.18361269454733e-06, + "loss": 0.4374, + "step": 18229 + }, + { + "epoch": 0.8071460987569596, + "grad_norm": 1.4471824169158936, + "learning_rate": 2.182648841577983e-06, + "loss": 0.506, + "step": 18230 + }, + { + "epoch": 0.8071903744617734, + "grad_norm": 1.5035713911056519, + "learning_rate": 2.1816851753197023e-06, + "loss": 0.4791, + "step": 18231 + }, + { + "epoch": 0.8072346501665874, + "grad_norm": 1.3225805759429932, + "learning_rate": 2.1807216957955056e-06, + "loss": 0.4474, + "step": 18232 + }, + { + "epoch": 0.8072789258714013, + "grad_norm": 1.3668750524520874, + "learning_rate": 2.1797584030284026e-06, + "loss": 0.4765, + "step": 18233 + }, + { + "epoch": 0.8073232015762151, + "grad_norm": 1.1769055128097534, + "learning_rate": 2.178795297041405e-06, + "loss": 0.4407, + "step": 18234 + }, + { + "epoch": 0.807367477281029, + "grad_norm": 1.3891080617904663, + "learning_rate": 2.1778323778575126e-06, + "loss": 0.5004, + "step": 18235 + }, + { + "epoch": 0.8074117529858429, + "grad_norm": 2.246225118637085, + "learning_rate": 2.1768696454997263e-06, + "loss": 0.5772, + "step": 18236 + }, + { + "epoch": 0.8074560286906567, + "grad_norm": 1.507158637046814, + "learning_rate": 2.175907099991037e-06, + "loss": 0.5539, + "step": 18237 + }, + { + "epoch": 0.8075003043954706, + "grad_norm": 1.6449267864227295, + "learning_rate": 2.1749447413544377e-06, + "loss": 0.4903, + "step": 18238 + }, + { + "epoch": 0.8075445801002845, + "grad_norm": 1.3804235458374023, + "learning_rate": 2.173982569612908e-06, + "loss": 0.4909, + "step": 18239 + }, + { + "epoch": 0.8075888558050983, + "grad_norm": 1.617205023765564, + "learning_rate": 2.1730205847894346e-06, + "loss": 0.4925, + "step": 18240 + }, + { + "epoch": 0.8076331315099122, + "grad_norm": 1.453597068786621, + "learning_rate": 2.1720587869069933e-06, + "loss": 0.457, + "step": 18241 + }, + { + "epoch": 0.8076774072147261, + "grad_norm": 1.6438262462615967, + "learning_rate": 2.1710971759885534e-06, + "loss": 0.5242, + "step": 18242 + }, + { + "epoch": 0.8077216829195399, + "grad_norm": 2.006943702697754, + "learning_rate": 2.1701357520570797e-06, + "loss": 0.5879, + "step": 18243 + }, + { + "epoch": 0.8077659586243539, + "grad_norm": 1.3400498628616333, + "learning_rate": 2.169174515135544e-06, + "loss": 0.4441, + "step": 18244 + }, + { + "epoch": 0.8078102343291678, + "grad_norm": 1.648338794708252, + "learning_rate": 2.168213465246897e-06, + "loss": 0.4643, + "step": 18245 + }, + { + "epoch": 0.8078545100339816, + "grad_norm": 1.4451121091842651, + "learning_rate": 2.1672526024140905e-06, + "loss": 0.5021, + "step": 18246 + }, + { + "epoch": 0.8078987857387955, + "grad_norm": 1.8015704154968262, + "learning_rate": 2.1662919266600814e-06, + "loss": 0.4802, + "step": 18247 + }, + { + "epoch": 0.8079430614436094, + "grad_norm": 1.2631912231445312, + "learning_rate": 2.1653314380078107e-06, + "loss": 0.4403, + "step": 18248 + }, + { + "epoch": 0.8079873371484232, + "grad_norm": 1.9528447389602661, + "learning_rate": 2.1643711364802167e-06, + "loss": 0.4826, + "step": 18249 + }, + { + "epoch": 0.8080316128532371, + "grad_norm": 1.4378926753997803, + "learning_rate": 2.1634110221002416e-06, + "loss": 0.5143, + "step": 18250 + }, + { + "epoch": 0.808075888558051, + "grad_norm": 1.704189658164978, + "learning_rate": 2.1624510948908117e-06, + "loss": 0.5242, + "step": 18251 + }, + { + "epoch": 0.8081201642628648, + "grad_norm": 1.6908700466156006, + "learning_rate": 2.1614913548748573e-06, + "loss": 0.4928, + "step": 18252 + }, + { + "epoch": 0.8081644399676787, + "grad_norm": 2.791883945465088, + "learning_rate": 2.1605318020753e-06, + "loss": 0.5993, + "step": 18253 + }, + { + "epoch": 0.8082087156724926, + "grad_norm": 1.6056690216064453, + "learning_rate": 2.159572436515056e-06, + "loss": 0.5504, + "step": 18254 + }, + { + "epoch": 0.8082529913773064, + "grad_norm": 1.3350281715393066, + "learning_rate": 2.1586132582170416e-06, + "loss": 0.4524, + "step": 18255 + }, + { + "epoch": 0.8082972670821204, + "grad_norm": 1.4091591835021973, + "learning_rate": 2.157654267204161e-06, + "loss": 0.5098, + "step": 18256 + }, + { + "epoch": 0.8083415427869343, + "grad_norm": 1.519224762916565, + "learning_rate": 2.1566954634993265e-06, + "loss": 0.5272, + "step": 18257 + }, + { + "epoch": 0.8083858184917481, + "grad_norm": 1.583763599395752, + "learning_rate": 2.155736847125435e-06, + "loss": 0.4938, + "step": 18258 + }, + { + "epoch": 0.808430094196562, + "grad_norm": 1.4310401678085327, + "learning_rate": 2.1547784181053776e-06, + "loss": 0.4924, + "step": 18259 + }, + { + "epoch": 0.8084743699013759, + "grad_norm": 1.3814674615859985, + "learning_rate": 2.1538201764620535e-06, + "loss": 0.4968, + "step": 18260 + }, + { + "epoch": 0.8085186456061897, + "grad_norm": 1.621575951576233, + "learning_rate": 2.1528621222183454e-06, + "loss": 0.5015, + "step": 18261 + }, + { + "epoch": 0.8085629213110036, + "grad_norm": 1.9358584880828857, + "learning_rate": 2.151904255397137e-06, + "loss": 0.5643, + "step": 18262 + }, + { + "epoch": 0.8086071970158175, + "grad_norm": 1.3332583904266357, + "learning_rate": 2.1509465760213046e-06, + "loss": 0.431, + "step": 18263 + }, + { + "epoch": 0.8086514727206314, + "grad_norm": 1.3771768808364868, + "learning_rate": 2.1499890841137226e-06, + "loss": 0.4263, + "step": 18264 + }, + { + "epoch": 0.8086957484254452, + "grad_norm": 1.3123164176940918, + "learning_rate": 2.149031779697259e-06, + "loss": 0.4711, + "step": 18265 + }, + { + "epoch": 0.8087400241302591, + "grad_norm": 1.33800208568573, + "learning_rate": 2.148074662794777e-06, + "loss": 0.4407, + "step": 18266 + }, + { + "epoch": 0.808784299835073, + "grad_norm": 1.3206093311309814, + "learning_rate": 2.1471177334291404e-06, + "loss": 0.4488, + "step": 18267 + }, + { + "epoch": 0.8088285755398869, + "grad_norm": 2.5284183025360107, + "learning_rate": 2.146160991623204e-06, + "loss": 0.557, + "step": 18268 + }, + { + "epoch": 0.8088728512447008, + "grad_norm": 1.5904121398925781, + "learning_rate": 2.1452044373998137e-06, + "loss": 0.5143, + "step": 18269 + }, + { + "epoch": 0.8089171269495147, + "grad_norm": 1.3446320295333862, + "learning_rate": 2.1442480707818227e-06, + "loss": 0.5118, + "step": 18270 + }, + { + "epoch": 0.8089614026543285, + "grad_norm": 1.7372156381607056, + "learning_rate": 2.1432918917920696e-06, + "loss": 0.53, + "step": 18271 + }, + { + "epoch": 0.8090056783591424, + "grad_norm": 1.453657865524292, + "learning_rate": 2.1423359004533926e-06, + "loss": 0.4846, + "step": 18272 + }, + { + "epoch": 0.8090499540639563, + "grad_norm": 1.6415369510650635, + "learning_rate": 2.1413800967886255e-06, + "loss": 0.4942, + "step": 18273 + }, + { + "epoch": 0.8090942297687701, + "grad_norm": 1.6538923978805542, + "learning_rate": 2.1404244808205955e-06, + "loss": 0.4819, + "step": 18274 + }, + { + "epoch": 0.809138505473584, + "grad_norm": 1.9243078231811523, + "learning_rate": 2.1394690525721275e-06, + "loss": 0.5376, + "step": 18275 + }, + { + "epoch": 0.8091827811783979, + "grad_norm": 1.4408456087112427, + "learning_rate": 2.1385138120660377e-06, + "loss": 0.4906, + "step": 18276 + }, + { + "epoch": 0.8092270568832117, + "grad_norm": 1.6055151224136353, + "learning_rate": 2.137558759325148e-06, + "loss": 0.5597, + "step": 18277 + }, + { + "epoch": 0.8092713325880256, + "grad_norm": 1.8566328287124634, + "learning_rate": 2.1366038943722646e-06, + "loss": 0.5237, + "step": 18278 + }, + { + "epoch": 0.8093156082928396, + "grad_norm": 1.4017637968063354, + "learning_rate": 2.1356492172301944e-06, + "loss": 0.4568, + "step": 18279 + }, + { + "epoch": 0.8093598839976534, + "grad_norm": 1.2998546361923218, + "learning_rate": 2.134694727921738e-06, + "loss": 0.483, + "step": 18280 + }, + { + "epoch": 0.8094041597024673, + "grad_norm": 1.7359017133712769, + "learning_rate": 2.1337404264696958e-06, + "loss": 0.5734, + "step": 18281 + }, + { + "epoch": 0.8094484354072812, + "grad_norm": 1.5717853307724, + "learning_rate": 2.1327863128968584e-06, + "loss": 0.4955, + "step": 18282 + }, + { + "epoch": 0.809492711112095, + "grad_norm": 1.6266335248947144, + "learning_rate": 2.1318323872260137e-06, + "loss": 0.5405, + "step": 18283 + }, + { + "epoch": 0.8095369868169089, + "grad_norm": 1.3439985513687134, + "learning_rate": 2.1308786494799472e-06, + "loss": 0.4794, + "step": 18284 + }, + { + "epoch": 0.8095812625217228, + "grad_norm": 1.5610980987548828, + "learning_rate": 2.1299250996814357e-06, + "loss": 0.5038, + "step": 18285 + }, + { + "epoch": 0.8096255382265366, + "grad_norm": 1.4376323223114014, + "learning_rate": 2.1289717378532524e-06, + "loss": 0.492, + "step": 18286 + }, + { + "epoch": 0.8096698139313505, + "grad_norm": 1.5065211057662964, + "learning_rate": 2.1280185640181737e-06, + "loss": 0.5362, + "step": 18287 + }, + { + "epoch": 0.8097140896361644, + "grad_norm": 1.9360078573226929, + "learning_rate": 2.1270655781989623e-06, + "loss": 0.5543, + "step": 18288 + }, + { + "epoch": 0.8097583653409782, + "grad_norm": 1.5013349056243896, + "learning_rate": 2.126112780418378e-06, + "loss": 0.4965, + "step": 18289 + }, + { + "epoch": 0.8098026410457921, + "grad_norm": 1.3816412687301636, + "learning_rate": 2.1251601706991776e-06, + "loss": 0.5022, + "step": 18290 + }, + { + "epoch": 0.809846916750606, + "grad_norm": 2.0467817783355713, + "learning_rate": 2.1242077490641157e-06, + "loss": 0.5099, + "step": 18291 + }, + { + "epoch": 0.8098911924554199, + "grad_norm": 1.3781089782714844, + "learning_rate": 2.1232555155359425e-06, + "loss": 0.5483, + "step": 18292 + }, + { + "epoch": 0.8099354681602338, + "grad_norm": 1.4320167303085327, + "learning_rate": 2.1223034701373922e-06, + "loss": 0.487, + "step": 18293 + }, + { + "epoch": 0.8099797438650477, + "grad_norm": 1.4398959875106812, + "learning_rate": 2.1213516128912124e-06, + "loss": 0.4752, + "step": 18294 + }, + { + "epoch": 0.8100240195698615, + "grad_norm": 1.6452295780181885, + "learning_rate": 2.120399943820134e-06, + "loss": 0.5311, + "step": 18295 + }, + { + "epoch": 0.8100682952746754, + "grad_norm": 1.4733312129974365, + "learning_rate": 2.119448462946885e-06, + "loss": 0.4739, + "step": 18296 + }, + { + "epoch": 0.8101125709794893, + "grad_norm": 1.9200106859207153, + "learning_rate": 2.1184971702941947e-06, + "loss": 0.5659, + "step": 18297 + }, + { + "epoch": 0.8101568466843032, + "grad_norm": 1.4889209270477295, + "learning_rate": 2.1175460658847824e-06, + "loss": 0.4742, + "step": 18298 + }, + { + "epoch": 0.810201122389117, + "grad_norm": 1.6197006702423096, + "learning_rate": 2.1165951497413638e-06, + "loss": 0.482, + "step": 18299 + }, + { + "epoch": 0.8102453980939309, + "grad_norm": 1.3490132093429565, + "learning_rate": 2.1156444218866523e-06, + "loss": 0.4923, + "step": 18300 + }, + { + "epoch": 0.8102896737987448, + "grad_norm": 1.4954957962036133, + "learning_rate": 2.1146938823433527e-06, + "loss": 0.4913, + "step": 18301 + }, + { + "epoch": 0.8103339495035586, + "grad_norm": 1.3392025232315063, + "learning_rate": 2.1137435311341703e-06, + "loss": 0.4651, + "step": 18302 + }, + { + "epoch": 0.8103782252083725, + "grad_norm": 1.264566421508789, + "learning_rate": 2.112793368281799e-06, + "loss": 0.4621, + "step": 18303 + }, + { + "epoch": 0.8104225009131865, + "grad_norm": 1.4079160690307617, + "learning_rate": 2.11184339380894e-06, + "loss": 0.4659, + "step": 18304 + }, + { + "epoch": 0.8104667766180003, + "grad_norm": 1.4530408382415771, + "learning_rate": 2.110893607738278e-06, + "loss": 0.5369, + "step": 18305 + }, + { + "epoch": 0.8105110523228142, + "grad_norm": 1.5016186237335205, + "learning_rate": 2.1099440100924953e-06, + "loss": 0.4739, + "step": 18306 + }, + { + "epoch": 0.8105553280276281, + "grad_norm": 1.6168365478515625, + "learning_rate": 2.1089946008942795e-06, + "loss": 0.4762, + "step": 18307 + }, + { + "epoch": 0.8105996037324419, + "grad_norm": 1.7527450323104858, + "learning_rate": 2.1080453801663016e-06, + "loss": 0.5331, + "step": 18308 + }, + { + "epoch": 0.8106438794372558, + "grad_norm": 1.5365628004074097, + "learning_rate": 2.1070963479312345e-06, + "loss": 0.5321, + "step": 18309 + }, + { + "epoch": 0.8106881551420697, + "grad_norm": 1.377300500869751, + "learning_rate": 2.1061475042117443e-06, + "loss": 0.457, + "step": 18310 + }, + { + "epoch": 0.8107324308468835, + "grad_norm": 1.6662639379501343, + "learning_rate": 2.105198849030493e-06, + "loss": 0.4924, + "step": 18311 + }, + { + "epoch": 0.8107767065516974, + "grad_norm": 1.4039113521575928, + "learning_rate": 2.1042503824101403e-06, + "loss": 0.4909, + "step": 18312 + }, + { + "epoch": 0.8108209822565113, + "grad_norm": 1.4516832828521729, + "learning_rate": 2.103302104373334e-06, + "loss": 0.4938, + "step": 18313 + }, + { + "epoch": 0.8108652579613251, + "grad_norm": 1.3548340797424316, + "learning_rate": 2.1023540149427314e-06, + "loss": 0.4661, + "step": 18314 + }, + { + "epoch": 0.810909533666139, + "grad_norm": 1.3277260065078735, + "learning_rate": 2.1014061141409715e-06, + "loss": 0.4821, + "step": 18315 + }, + { + "epoch": 0.810953809370953, + "grad_norm": 1.4051584005355835, + "learning_rate": 2.1004584019906925e-06, + "loss": 0.494, + "step": 18316 + }, + { + "epoch": 0.8109980850757668, + "grad_norm": 1.853836178779602, + "learning_rate": 2.099510878514536e-06, + "loss": 0.5454, + "step": 18317 + }, + { + "epoch": 0.8110423607805807, + "grad_norm": 1.5402833223342896, + "learning_rate": 2.0985635437351294e-06, + "loss": 0.461, + "step": 18318 + }, + { + "epoch": 0.8110866364853946, + "grad_norm": 1.5287702083587646, + "learning_rate": 2.0976163976750985e-06, + "loss": 0.4678, + "step": 18319 + }, + { + "epoch": 0.8111309121902084, + "grad_norm": 1.806334376335144, + "learning_rate": 2.0966694403570653e-06, + "loss": 0.5578, + "step": 18320 + }, + { + "epoch": 0.8111751878950223, + "grad_norm": 1.4326196908950806, + "learning_rate": 2.0957226718036482e-06, + "loss": 0.5057, + "step": 18321 + }, + { + "epoch": 0.8112194635998362, + "grad_norm": 1.235264539718628, + "learning_rate": 2.0947760920374595e-06, + "loss": 0.448, + "step": 18322 + }, + { + "epoch": 0.81126373930465, + "grad_norm": 1.629952311515808, + "learning_rate": 2.0938297010811036e-06, + "loss": 0.4962, + "step": 18323 + }, + { + "epoch": 0.8113080150094639, + "grad_norm": 1.4438295364379883, + "learning_rate": 2.09288349895719e-06, + "loss": 0.5297, + "step": 18324 + }, + { + "epoch": 0.8113522907142778, + "grad_norm": 1.79963219165802, + "learning_rate": 2.091937485688317e-06, + "loss": 0.5406, + "step": 18325 + }, + { + "epoch": 0.8113965664190916, + "grad_norm": 1.5220428705215454, + "learning_rate": 2.090991661297075e-06, + "loss": 0.5126, + "step": 18326 + }, + { + "epoch": 0.8114408421239055, + "grad_norm": 1.4675276279449463, + "learning_rate": 2.0900460258060597e-06, + "loss": 0.4646, + "step": 18327 + }, + { + "epoch": 0.8114851178287195, + "grad_norm": 1.7070506811141968, + "learning_rate": 2.089100579237855e-06, + "loss": 0.5125, + "step": 18328 + }, + { + "epoch": 0.8115293935335333, + "grad_norm": 1.3956202268600464, + "learning_rate": 2.088155321615042e-06, + "loss": 0.4689, + "step": 18329 + }, + { + "epoch": 0.8115736692383472, + "grad_norm": 1.4210355281829834, + "learning_rate": 2.0872102529601966e-06, + "loss": 0.5027, + "step": 18330 + }, + { + "epoch": 0.8116179449431611, + "grad_norm": 2.3533997535705566, + "learning_rate": 2.0862653732958914e-06, + "loss": 0.6325, + "step": 18331 + }, + { + "epoch": 0.8116622206479749, + "grad_norm": 1.6021203994750977, + "learning_rate": 2.0853206826446946e-06, + "loss": 0.5415, + "step": 18332 + }, + { + "epoch": 0.8117064963527888, + "grad_norm": 1.4948477745056152, + "learning_rate": 2.084376181029166e-06, + "loss": 0.5138, + "step": 18333 + }, + { + "epoch": 0.8117507720576027, + "grad_norm": 1.3994107246398926, + "learning_rate": 2.08343186847187e-06, + "loss": 0.4453, + "step": 18334 + }, + { + "epoch": 0.8117950477624166, + "grad_norm": 1.7772653102874756, + "learning_rate": 2.0824877449953573e-06, + "loss": 0.498, + "step": 18335 + }, + { + "epoch": 0.8118393234672304, + "grad_norm": 1.8025124073028564, + "learning_rate": 2.0815438106221764e-06, + "loss": 0.5563, + "step": 18336 + }, + { + "epoch": 0.8118835991720443, + "grad_norm": 1.151520848274231, + "learning_rate": 2.0806000653748758e-06, + "loss": 0.4391, + "step": 18337 + }, + { + "epoch": 0.8119278748768582, + "grad_norm": 1.5568803548812866, + "learning_rate": 2.079656509275996e-06, + "loss": 0.5081, + "step": 18338 + }, + { + "epoch": 0.811972150581672, + "grad_norm": 1.502935767173767, + "learning_rate": 2.0787131423480722e-06, + "loss": 0.4566, + "step": 18339 + }, + { + "epoch": 0.812016426286486, + "grad_norm": 1.7850757837295532, + "learning_rate": 2.0777699646136306e-06, + "loss": 0.5297, + "step": 18340 + }, + { + "epoch": 0.8120607019912999, + "grad_norm": 1.5943535566329956, + "learning_rate": 2.0768269760952063e-06, + "loss": 0.4761, + "step": 18341 + }, + { + "epoch": 0.8121049776961137, + "grad_norm": 1.3736714124679565, + "learning_rate": 2.075884176815318e-06, + "loss": 0.4757, + "step": 18342 + }, + { + "epoch": 0.8121492534009276, + "grad_norm": 1.367108941078186, + "learning_rate": 2.0749415667964812e-06, + "loss": 0.4721, + "step": 18343 + }, + { + "epoch": 0.8121935291057415, + "grad_norm": 1.8276621103286743, + "learning_rate": 2.073999146061215e-06, + "loss": 0.5672, + "step": 18344 + }, + { + "epoch": 0.8122378048105553, + "grad_norm": 1.3292275667190552, + "learning_rate": 2.073056914632026e-06, + "loss": 0.4574, + "step": 18345 + }, + { + "epoch": 0.8122820805153692, + "grad_norm": 1.6271549463272095, + "learning_rate": 2.0721148725314143e-06, + "loss": 0.5152, + "step": 18346 + }, + { + "epoch": 0.8123263562201831, + "grad_norm": 1.482512354850769, + "learning_rate": 2.0711730197818904e-06, + "loss": 0.4697, + "step": 18347 + }, + { + "epoch": 0.8123706319249969, + "grad_norm": 1.3712687492370605, + "learning_rate": 2.0702313564059396e-06, + "loss": 0.501, + "step": 18348 + }, + { + "epoch": 0.8124149076298108, + "grad_norm": 1.404040813446045, + "learning_rate": 2.0692898824260565e-06, + "loss": 0.5269, + "step": 18349 + }, + { + "epoch": 0.8124591833346247, + "grad_norm": 1.346253514289856, + "learning_rate": 2.0683485978647242e-06, + "loss": 0.5165, + "step": 18350 + }, + { + "epoch": 0.8125034590394385, + "grad_norm": 1.3405981063842773, + "learning_rate": 2.0674075027444308e-06, + "loss": 0.4629, + "step": 18351 + }, + { + "epoch": 0.8125477347442525, + "grad_norm": 1.5752367973327637, + "learning_rate": 2.0664665970876496e-06, + "loss": 0.4994, + "step": 18352 + }, + { + "epoch": 0.8125920104490664, + "grad_norm": 1.6319169998168945, + "learning_rate": 2.065525880916852e-06, + "loss": 0.5251, + "step": 18353 + }, + { + "epoch": 0.8126362861538802, + "grad_norm": 1.6577634811401367, + "learning_rate": 2.0645853542545102e-06, + "loss": 0.5206, + "step": 18354 + }, + { + "epoch": 0.8126805618586941, + "grad_norm": 1.3029227256774902, + "learning_rate": 2.063645017123086e-06, + "loss": 0.4483, + "step": 18355 + }, + { + "epoch": 0.812724837563508, + "grad_norm": 1.7407886981964111, + "learning_rate": 2.062704869545039e-06, + "loss": 0.6336, + "step": 18356 + }, + { + "epoch": 0.8127691132683218, + "grad_norm": 1.4116442203521729, + "learning_rate": 2.061764911542824e-06, + "loss": 0.4578, + "step": 18357 + }, + { + "epoch": 0.8128133889731357, + "grad_norm": 1.8060129880905151, + "learning_rate": 2.0608251431388893e-06, + "loss": 0.4884, + "step": 18358 + }, + { + "epoch": 0.8128576646779496, + "grad_norm": 1.401480793952942, + "learning_rate": 2.0598855643556824e-06, + "loss": 0.4917, + "step": 18359 + }, + { + "epoch": 0.8129019403827634, + "grad_norm": 1.8051731586456299, + "learning_rate": 2.0589461752156404e-06, + "loss": 0.5729, + "step": 18360 + }, + { + "epoch": 0.8129462160875773, + "grad_norm": 1.256283164024353, + "learning_rate": 2.058006975741207e-06, + "loss": 0.4513, + "step": 18361 + }, + { + "epoch": 0.8129904917923912, + "grad_norm": 1.3468644618988037, + "learning_rate": 2.057067965954809e-06, + "loss": 0.4696, + "step": 18362 + }, + { + "epoch": 0.813034767497205, + "grad_norm": 1.379766583442688, + "learning_rate": 2.0561291458788736e-06, + "loss": 0.4923, + "step": 18363 + }, + { + "epoch": 0.813079043202019, + "grad_norm": 1.7922694683074951, + "learning_rate": 2.0551905155358278e-06, + "loss": 0.535, + "step": 18364 + }, + { + "epoch": 0.8131233189068329, + "grad_norm": 1.21419358253479, + "learning_rate": 2.0542520749480865e-06, + "loss": 0.4386, + "step": 18365 + }, + { + "epoch": 0.8131675946116467, + "grad_norm": 1.480018973350525, + "learning_rate": 2.053313824138067e-06, + "loss": 0.4818, + "step": 18366 + }, + { + "epoch": 0.8132118703164606, + "grad_norm": 1.3845407962799072, + "learning_rate": 2.052375763128175e-06, + "loss": 0.4622, + "step": 18367 + }, + { + "epoch": 0.8132561460212745, + "grad_norm": 1.4312598705291748, + "learning_rate": 2.0514378919408163e-06, + "loss": 0.5022, + "step": 18368 + }, + { + "epoch": 0.8133004217260884, + "grad_norm": 1.3778384923934937, + "learning_rate": 2.0505002105983917e-06, + "loss": 0.4444, + "step": 18369 + }, + { + "epoch": 0.8133446974309022, + "grad_norm": 1.7179293632507324, + "learning_rate": 2.0495627191232948e-06, + "loss": 0.5321, + "step": 18370 + }, + { + "epoch": 0.8133889731357161, + "grad_norm": 1.4251786470413208, + "learning_rate": 2.0486254175379218e-06, + "loss": 0.4416, + "step": 18371 + }, + { + "epoch": 0.81343324884053, + "grad_norm": 1.7346410751342773, + "learning_rate": 2.0476883058646556e-06, + "loss": 0.5068, + "step": 18372 + }, + { + "epoch": 0.8134775245453438, + "grad_norm": 1.6807503700256348, + "learning_rate": 2.046751384125877e-06, + "loss": 0.5534, + "step": 18373 + }, + { + "epoch": 0.8135218002501577, + "grad_norm": 1.4258816242218018, + "learning_rate": 2.045814652343968e-06, + "loss": 0.4778, + "step": 18374 + }, + { + "epoch": 0.8135660759549717, + "grad_norm": 1.3810765743255615, + "learning_rate": 2.0448781105413008e-06, + "loss": 0.4955, + "step": 18375 + }, + { + "epoch": 0.8136103516597855, + "grad_norm": 1.3939121961593628, + "learning_rate": 2.043941758740241e-06, + "loss": 0.4772, + "step": 18376 + }, + { + "epoch": 0.8136546273645994, + "grad_norm": 1.663238763809204, + "learning_rate": 2.0430055969631545e-06, + "loss": 0.4481, + "step": 18377 + }, + { + "epoch": 0.8136989030694133, + "grad_norm": 1.5432634353637695, + "learning_rate": 2.0420696252324014e-06, + "loss": 0.53, + "step": 18378 + }, + { + "epoch": 0.8137431787742271, + "grad_norm": 1.4836809635162354, + "learning_rate": 2.0411338435703355e-06, + "loss": 0.4764, + "step": 18379 + }, + { + "epoch": 0.813787454479041, + "grad_norm": 1.4192790985107422, + "learning_rate": 2.0401982519993046e-06, + "loss": 0.4545, + "step": 18380 + }, + { + "epoch": 0.8138317301838549, + "grad_norm": 1.6478054523468018, + "learning_rate": 2.03926285054166e-06, + "loss": 0.5266, + "step": 18381 + }, + { + "epoch": 0.8138760058886687, + "grad_norm": 1.2568684816360474, + "learning_rate": 2.0383276392197403e-06, + "loss": 0.3984, + "step": 18382 + }, + { + "epoch": 0.8139202815934826, + "grad_norm": 1.764156699180603, + "learning_rate": 2.037392618055879e-06, + "loss": 0.5248, + "step": 18383 + }, + { + "epoch": 0.8139645572982965, + "grad_norm": 1.427119255065918, + "learning_rate": 2.0364577870724156e-06, + "loss": 0.4461, + "step": 18384 + }, + { + "epoch": 0.8140088330031103, + "grad_norm": 1.5885899066925049, + "learning_rate": 2.0355231462916725e-06, + "loss": 0.5099, + "step": 18385 + }, + { + "epoch": 0.8140531087079242, + "grad_norm": 1.5261090993881226, + "learning_rate": 2.0345886957359752e-06, + "loss": 0.4499, + "step": 18386 + }, + { + "epoch": 0.8140973844127382, + "grad_norm": 2.158961534500122, + "learning_rate": 2.03365443542764e-06, + "loss": 0.6068, + "step": 18387 + }, + { + "epoch": 0.814141660117552, + "grad_norm": 1.5298229455947876, + "learning_rate": 2.0327203653889826e-06, + "loss": 0.4538, + "step": 18388 + }, + { + "epoch": 0.8141859358223659, + "grad_norm": 2.260768413543701, + "learning_rate": 2.031786485642313e-06, + "loss": 0.6069, + "step": 18389 + }, + { + "epoch": 0.8142302115271798, + "grad_norm": 1.3356844186782837, + "learning_rate": 2.0308527962099322e-06, + "loss": 0.5006, + "step": 18390 + }, + { + "epoch": 0.8142744872319936, + "grad_norm": 1.5262691974639893, + "learning_rate": 2.0299192971141456e-06, + "loss": 0.5131, + "step": 18391 + }, + { + "epoch": 0.8143187629368075, + "grad_norm": 1.5667037963867188, + "learning_rate": 2.0289859883772477e-06, + "loss": 0.5274, + "step": 18392 + }, + { + "epoch": 0.8143630386416214, + "grad_norm": 2.274383306503296, + "learning_rate": 2.028052870021526e-06, + "loss": 0.5571, + "step": 18393 + }, + { + "epoch": 0.8144073143464352, + "grad_norm": 1.5534253120422363, + "learning_rate": 2.0271199420692735e-06, + "loss": 0.4903, + "step": 18394 + }, + { + "epoch": 0.8144515900512491, + "grad_norm": 1.4171751737594604, + "learning_rate": 2.0261872045427723e-06, + "loss": 0.4653, + "step": 18395 + }, + { + "epoch": 0.814495865756063, + "grad_norm": 1.1819064617156982, + "learning_rate": 2.0252546574642952e-06, + "loss": 0.442, + "step": 18396 + }, + { + "epoch": 0.8145401414608768, + "grad_norm": 1.312642216682434, + "learning_rate": 2.024322300856113e-06, + "loss": 0.5134, + "step": 18397 + }, + { + "epoch": 0.8145844171656907, + "grad_norm": 1.4547276496887207, + "learning_rate": 2.023390134740503e-06, + "loss": 0.4273, + "step": 18398 + }, + { + "epoch": 0.8146286928705047, + "grad_norm": 1.6788028478622437, + "learning_rate": 2.022458159139723e-06, + "loss": 0.521, + "step": 18399 + }, + { + "epoch": 0.8146729685753185, + "grad_norm": 1.4148565530776978, + "learning_rate": 2.021526374076033e-06, + "loss": 0.4511, + "step": 18400 + }, + { + "epoch": 0.8147172442801324, + "grad_norm": 1.592254877090454, + "learning_rate": 2.0205947795716917e-06, + "loss": 0.4976, + "step": 18401 + }, + { + "epoch": 0.8147615199849463, + "grad_norm": 1.5620421171188354, + "learning_rate": 2.019663375648947e-06, + "loss": 0.5092, + "step": 18402 + }, + { + "epoch": 0.8148057956897601, + "grad_norm": 1.6961209774017334, + "learning_rate": 2.0187321623300437e-06, + "loss": 0.5336, + "step": 18403 + }, + { + "epoch": 0.814850071394574, + "grad_norm": 1.3502978086471558, + "learning_rate": 2.0178011396372254e-06, + "loss": 0.4584, + "step": 18404 + }, + { + "epoch": 0.8148943470993879, + "grad_norm": 1.4840598106384277, + "learning_rate": 2.0168703075927266e-06, + "loss": 0.5087, + "step": 18405 + }, + { + "epoch": 0.8149386228042018, + "grad_norm": 1.7760487794876099, + "learning_rate": 2.015939666218781e-06, + "loss": 0.548, + "step": 18406 + }, + { + "epoch": 0.8149828985090156, + "grad_norm": 1.4567108154296875, + "learning_rate": 2.0150092155376133e-06, + "loss": 0.4898, + "step": 18407 + }, + { + "epoch": 0.8150271742138295, + "grad_norm": 1.3944803476333618, + "learning_rate": 2.014078955571451e-06, + "loss": 0.5015, + "step": 18408 + }, + { + "epoch": 0.8150714499186434, + "grad_norm": 1.4751206636428833, + "learning_rate": 2.013148886342511e-06, + "loss": 0.5021, + "step": 18409 + }, + { + "epoch": 0.8151157256234572, + "grad_norm": 1.376146912574768, + "learning_rate": 2.012219007873003e-06, + "loss": 0.4797, + "step": 18410 + }, + { + "epoch": 0.8151600013282712, + "grad_norm": 1.2919573783874512, + "learning_rate": 2.0112893201851435e-06, + "loss": 0.4581, + "step": 18411 + }, + { + "epoch": 0.8152042770330851, + "grad_norm": 1.7819050550460815, + "learning_rate": 2.0103598233011332e-06, + "loss": 0.5941, + "step": 18412 + }, + { + "epoch": 0.8152485527378989, + "grad_norm": 1.521311640739441, + "learning_rate": 2.0094305172431737e-06, + "loss": 0.5366, + "step": 18413 + }, + { + "epoch": 0.8152928284427128, + "grad_norm": 1.7556700706481934, + "learning_rate": 2.0085014020334605e-06, + "loss": 0.4717, + "step": 18414 + }, + { + "epoch": 0.8153371041475267, + "grad_norm": 1.4100818634033203, + "learning_rate": 2.0075724776941842e-06, + "loss": 0.4595, + "step": 18415 + }, + { + "epoch": 0.8153813798523405, + "grad_norm": 1.5494070053100586, + "learning_rate": 2.0066437442475317e-06, + "loss": 0.5002, + "step": 18416 + }, + { + "epoch": 0.8154256555571544, + "grad_norm": 1.5040652751922607, + "learning_rate": 2.0057152017156824e-06, + "loss": 0.5088, + "step": 18417 + }, + { + "epoch": 0.8154699312619683, + "grad_norm": 1.4203091859817505, + "learning_rate": 2.004786850120818e-06, + "loss": 0.4983, + "step": 18418 + }, + { + "epoch": 0.8155142069667821, + "grad_norm": 1.875327706336975, + "learning_rate": 2.003858689485111e-06, + "loss": 0.5557, + "step": 18419 + }, + { + "epoch": 0.815558482671596, + "grad_norm": 1.473685622215271, + "learning_rate": 2.0029307198307247e-06, + "loss": 0.5034, + "step": 18420 + }, + { + "epoch": 0.8156027583764099, + "grad_norm": 1.351431131362915, + "learning_rate": 2.0020029411798294e-06, + "loss": 0.4988, + "step": 18421 + }, + { + "epoch": 0.8156470340812237, + "grad_norm": 1.3056777715682983, + "learning_rate": 2.001075353554582e-06, + "loss": 0.495, + "step": 18422 + }, + { + "epoch": 0.8156913097860377, + "grad_norm": 1.3377007246017456, + "learning_rate": 2.0001479569771377e-06, + "loss": 0.4361, + "step": 18423 + }, + { + "epoch": 0.8157355854908516, + "grad_norm": 1.196737289428711, + "learning_rate": 1.999220751469645e-06, + "loss": 0.4386, + "step": 18424 + }, + { + "epoch": 0.8157798611956654, + "grad_norm": 1.8411288261413574, + "learning_rate": 1.99829373705425e-06, + "loss": 0.5398, + "step": 18425 + }, + { + "epoch": 0.8158241369004793, + "grad_norm": 1.2471933364868164, + "learning_rate": 1.997366913753094e-06, + "loss": 0.4378, + "step": 18426 + }, + { + "epoch": 0.8158684126052932, + "grad_norm": 1.438841700553894, + "learning_rate": 1.99644028158831e-06, + "loss": 0.5267, + "step": 18427 + }, + { + "epoch": 0.815912688310107, + "grad_norm": 1.3249565362930298, + "learning_rate": 1.995513840582036e-06, + "loss": 0.4682, + "step": 18428 + }, + { + "epoch": 0.8159569640149209, + "grad_norm": 1.539841651916504, + "learning_rate": 1.994587590756397e-06, + "loss": 0.4943, + "step": 18429 + }, + { + "epoch": 0.8160012397197348, + "grad_norm": 1.4511109590530396, + "learning_rate": 1.993661532133513e-06, + "loss": 0.4566, + "step": 18430 + }, + { + "epoch": 0.8160455154245486, + "grad_norm": 1.6258788108825684, + "learning_rate": 1.9927356647355056e-06, + "loss": 0.5427, + "step": 18431 + }, + { + "epoch": 0.8160897911293625, + "grad_norm": 1.6287516355514526, + "learning_rate": 1.991809988584488e-06, + "loss": 0.4862, + "step": 18432 + }, + { + "epoch": 0.8161340668341764, + "grad_norm": 1.8690452575683594, + "learning_rate": 1.990884503702568e-06, + "loss": 0.5177, + "step": 18433 + }, + { + "epoch": 0.8161783425389902, + "grad_norm": 1.2528997659683228, + "learning_rate": 1.9899592101118503e-06, + "loss": 0.4684, + "step": 18434 + }, + { + "epoch": 0.8162226182438042, + "grad_norm": 1.701370358467102, + "learning_rate": 1.9890341078344343e-06, + "loss": 0.5753, + "step": 18435 + }, + { + "epoch": 0.8162668939486181, + "grad_norm": 1.3891613483428955, + "learning_rate": 1.988109196892417e-06, + "loss": 0.4748, + "step": 18436 + }, + { + "epoch": 0.8163111696534319, + "grad_norm": 1.8310319185256958, + "learning_rate": 1.987184477307885e-06, + "loss": 0.498, + "step": 18437 + }, + { + "epoch": 0.8163554453582458, + "grad_norm": 1.5069389343261719, + "learning_rate": 1.98625994910293e-06, + "loss": 0.4485, + "step": 18438 + }, + { + "epoch": 0.8163997210630597, + "grad_norm": 1.4414538145065308, + "learning_rate": 1.9853356122996314e-06, + "loss": 0.4675, + "step": 18439 + }, + { + "epoch": 0.8164439967678736, + "grad_norm": 1.6875373125076294, + "learning_rate": 1.984411466920063e-06, + "loss": 0.5492, + "step": 18440 + }, + { + "epoch": 0.8164882724726874, + "grad_norm": 1.4779528379440308, + "learning_rate": 1.983487512986303e-06, + "loss": 0.4939, + "step": 18441 + }, + { + "epoch": 0.8165325481775013, + "grad_norm": 1.525346279144287, + "learning_rate": 1.9825637505204186e-06, + "loss": 0.4769, + "step": 18442 + }, + { + "epoch": 0.8165768238823152, + "grad_norm": 1.320151925086975, + "learning_rate": 1.9816401795444664e-06, + "loss": 0.476, + "step": 18443 + }, + { + "epoch": 0.816621099587129, + "grad_norm": 1.5340677499771118, + "learning_rate": 1.9807168000805112e-06, + "loss": 0.4977, + "step": 18444 + }, + { + "epoch": 0.8166653752919429, + "grad_norm": 1.4269063472747803, + "learning_rate": 1.979793612150607e-06, + "loss": 0.5073, + "step": 18445 + }, + { + "epoch": 0.8167096509967569, + "grad_norm": 1.5958468914031982, + "learning_rate": 1.9788706157768013e-06, + "loss": 0.4714, + "step": 18446 + }, + { + "epoch": 0.8167539267015707, + "grad_norm": 1.597978949546814, + "learning_rate": 1.9779478109811366e-06, + "loss": 0.4915, + "step": 18447 + }, + { + "epoch": 0.8167982024063846, + "grad_norm": 1.314001202583313, + "learning_rate": 1.9770251977856604e-06, + "loss": 0.4456, + "step": 18448 + }, + { + "epoch": 0.8168424781111985, + "grad_norm": 1.5815958976745605, + "learning_rate": 1.9761027762124053e-06, + "loss": 0.5098, + "step": 18449 + }, + { + "epoch": 0.8168867538160123, + "grad_norm": 1.5338571071624756, + "learning_rate": 1.9751805462833983e-06, + "loss": 0.4971, + "step": 18450 + }, + { + "epoch": 0.8169310295208262, + "grad_norm": 1.9233983755111694, + "learning_rate": 1.9742585080206754e-06, + "loss": 0.5694, + "step": 18451 + }, + { + "epoch": 0.8169753052256401, + "grad_norm": 1.8832228183746338, + "learning_rate": 1.9733366614462514e-06, + "loss": 0.5206, + "step": 18452 + }, + { + "epoch": 0.8170195809304539, + "grad_norm": 1.4736344814300537, + "learning_rate": 1.9724150065821424e-06, + "loss": 0.5033, + "step": 18453 + }, + { + "epoch": 0.8170638566352678, + "grad_norm": 1.4384326934814453, + "learning_rate": 1.9714935434503666e-06, + "loss": 0.5018, + "step": 18454 + }, + { + "epoch": 0.8171081323400817, + "grad_norm": 1.5480507612228394, + "learning_rate": 1.9705722720729313e-06, + "loss": 0.4532, + "step": 18455 + }, + { + "epoch": 0.8171524080448955, + "grad_norm": 2.1142098903656006, + "learning_rate": 1.9696511924718396e-06, + "loss": 0.5654, + "step": 18456 + }, + { + "epoch": 0.8171966837497094, + "grad_norm": 1.5115580558776855, + "learning_rate": 1.9687303046690875e-06, + "loss": 0.5088, + "step": 18457 + }, + { + "epoch": 0.8172409594545234, + "grad_norm": 1.6080591678619385, + "learning_rate": 1.9678096086866762e-06, + "loss": 0.5542, + "step": 18458 + }, + { + "epoch": 0.8172852351593372, + "grad_norm": 1.4144985675811768, + "learning_rate": 1.966889104546591e-06, + "loss": 0.439, + "step": 18459 + }, + { + "epoch": 0.8173295108641511, + "grad_norm": 1.3974169492721558, + "learning_rate": 1.9659687922708203e-06, + "loss": 0.4617, + "step": 18460 + }, + { + "epoch": 0.817373786568965, + "grad_norm": 1.4833775758743286, + "learning_rate": 1.9650486718813423e-06, + "loss": 0.5556, + "step": 18461 + }, + { + "epoch": 0.8174180622737788, + "grad_norm": 1.4065589904785156, + "learning_rate": 1.9641287434001356e-06, + "loss": 0.4583, + "step": 18462 + }, + { + "epoch": 0.8174623379785927, + "grad_norm": 2.4610533714294434, + "learning_rate": 1.963209006849167e-06, + "loss": 0.5001, + "step": 18463 + }, + { + "epoch": 0.8175066136834066, + "grad_norm": 1.4831948280334473, + "learning_rate": 1.962289462250411e-06, + "loss": 0.4812, + "step": 18464 + }, + { + "epoch": 0.8175508893882204, + "grad_norm": 1.5374130010604858, + "learning_rate": 1.9613701096258255e-06, + "loss": 0.4612, + "step": 18465 + }, + { + "epoch": 0.8175951650930343, + "grad_norm": 1.5476514101028442, + "learning_rate": 1.96045094899737e-06, + "loss": 0.5227, + "step": 18466 + }, + { + "epoch": 0.8176394407978482, + "grad_norm": 1.6359623670578003, + "learning_rate": 1.9595319803869962e-06, + "loss": 0.4936, + "step": 18467 + }, + { + "epoch": 0.817683716502662, + "grad_norm": 1.6121052503585815, + "learning_rate": 1.958613203816655e-06, + "loss": 0.4464, + "step": 18468 + }, + { + "epoch": 0.8177279922074759, + "grad_norm": 1.5098941326141357, + "learning_rate": 1.957694619308291e-06, + "loss": 0.5537, + "step": 18469 + }, + { + "epoch": 0.8177722679122899, + "grad_norm": 1.519023060798645, + "learning_rate": 1.956776226883843e-06, + "loss": 0.4318, + "step": 18470 + }, + { + "epoch": 0.8178165436171037, + "grad_norm": 1.9510859251022339, + "learning_rate": 1.9558580265652448e-06, + "loss": 0.5491, + "step": 18471 + }, + { + "epoch": 0.8178608193219176, + "grad_norm": 1.6174463033676147, + "learning_rate": 1.954940018374428e-06, + "loss": 0.421, + "step": 18472 + }, + { + "epoch": 0.8179050950267315, + "grad_norm": 1.5201140642166138, + "learning_rate": 1.9540222023333165e-06, + "loss": 0.5215, + "step": 18473 + }, + { + "epoch": 0.8179493707315453, + "grad_norm": 1.6281614303588867, + "learning_rate": 1.953104578463836e-06, + "loss": 0.4787, + "step": 18474 + }, + { + "epoch": 0.8179936464363592, + "grad_norm": 1.4736084938049316, + "learning_rate": 1.9521871467878993e-06, + "loss": 0.4346, + "step": 18475 + }, + { + "epoch": 0.8180379221411731, + "grad_norm": 1.4740675687789917, + "learning_rate": 1.951269907327421e-06, + "loss": 0.4735, + "step": 18476 + }, + { + "epoch": 0.818082197845987, + "grad_norm": 1.4234135150909424, + "learning_rate": 1.9503528601043054e-06, + "loss": 0.4493, + "step": 18477 + }, + { + "epoch": 0.8181264735508008, + "grad_norm": 1.4114142656326294, + "learning_rate": 1.9494360051404592e-06, + "loss": 0.4922, + "step": 18478 + }, + { + "epoch": 0.8181707492556147, + "grad_norm": 2.2755167484283447, + "learning_rate": 1.9485193424577787e-06, + "loss": 0.5181, + "step": 18479 + }, + { + "epoch": 0.8182150249604286, + "grad_norm": 1.9693323373794556, + "learning_rate": 1.9476028720781583e-06, + "loss": 0.4903, + "step": 18480 + }, + { + "epoch": 0.8182593006652424, + "grad_norm": 1.873531460762024, + "learning_rate": 1.9466865940234867e-06, + "loss": 0.5259, + "step": 18481 + }, + { + "epoch": 0.8183035763700564, + "grad_norm": 1.4500092267990112, + "learning_rate": 1.9457705083156488e-06, + "loss": 0.5224, + "step": 18482 + }, + { + "epoch": 0.8183478520748703, + "grad_norm": 1.4715005159378052, + "learning_rate": 1.944854614976521e-06, + "loss": 0.5185, + "step": 18483 + }, + { + "epoch": 0.8183921277796841, + "grad_norm": 1.2891789674758911, + "learning_rate": 1.9439389140279865e-06, + "loss": 0.4523, + "step": 18484 + }, + { + "epoch": 0.818436403484498, + "grad_norm": 1.8228607177734375, + "learning_rate": 1.9430234054919097e-06, + "loss": 0.4792, + "step": 18485 + }, + { + "epoch": 0.8184806791893119, + "grad_norm": 1.5043786764144897, + "learning_rate": 1.942108089390159e-06, + "loss": 0.5254, + "step": 18486 + }, + { + "epoch": 0.8185249548941257, + "grad_norm": 1.4662110805511475, + "learning_rate": 1.9411929657445927e-06, + "loss": 0.4975, + "step": 18487 + }, + { + "epoch": 0.8185692305989396, + "grad_norm": 1.3889952898025513, + "learning_rate": 1.9402780345770743e-06, + "loss": 0.4528, + "step": 18488 + }, + { + "epoch": 0.8186135063037535, + "grad_norm": 1.7957544326782227, + "learning_rate": 1.939363295909452e-06, + "loss": 0.5623, + "step": 18489 + }, + { + "epoch": 0.8186577820085673, + "grad_norm": 1.3980814218521118, + "learning_rate": 1.938448749763574e-06, + "loss": 0.496, + "step": 18490 + }, + { + "epoch": 0.8187020577133812, + "grad_norm": 1.6430121660232544, + "learning_rate": 1.9375343961612835e-06, + "loss": 0.4934, + "step": 18491 + }, + { + "epoch": 0.8187463334181951, + "grad_norm": 1.4002633094787598, + "learning_rate": 1.9366202351244203e-06, + "loss": 0.4727, + "step": 18492 + }, + { + "epoch": 0.8187906091230089, + "grad_norm": 1.4922397136688232, + "learning_rate": 1.935706266674814e-06, + "loss": 0.47, + "step": 18493 + }, + { + "epoch": 0.8188348848278229, + "grad_norm": 1.6383899450302124, + "learning_rate": 1.9347924908342997e-06, + "loss": 0.5209, + "step": 18494 + }, + { + "epoch": 0.8188791605326368, + "grad_norm": 1.7181918621063232, + "learning_rate": 1.9338789076247e-06, + "loss": 0.5378, + "step": 18495 + }, + { + "epoch": 0.8189234362374506, + "grad_norm": 1.5769424438476562, + "learning_rate": 1.932965517067835e-06, + "loss": 0.515, + "step": 18496 + }, + { + "epoch": 0.8189677119422645, + "grad_norm": 1.4086368083953857, + "learning_rate": 1.932052319185518e-06, + "loss": 0.4436, + "step": 18497 + }, + { + "epoch": 0.8190119876470784, + "grad_norm": 1.6020219326019287, + "learning_rate": 1.931139313999566e-06, + "loss": 0.5576, + "step": 18498 + }, + { + "epoch": 0.8190562633518922, + "grad_norm": 1.539150357246399, + "learning_rate": 1.93022650153178e-06, + "loss": 0.4279, + "step": 18499 + }, + { + "epoch": 0.8191005390567061, + "grad_norm": 1.940673589706421, + "learning_rate": 1.9293138818039615e-06, + "loss": 0.5051, + "step": 18500 + }, + { + "epoch": 0.81914481476152, + "grad_norm": 1.6322308778762817, + "learning_rate": 1.9284014548379104e-06, + "loss": 0.5246, + "step": 18501 + }, + { + "epoch": 0.8191890904663338, + "grad_norm": 1.4846364259719849, + "learning_rate": 1.9274892206554197e-06, + "loss": 0.4347, + "step": 18502 + }, + { + "epoch": 0.8192333661711477, + "grad_norm": 1.3675734996795654, + "learning_rate": 1.926577179278276e-06, + "loss": 0.4421, + "step": 18503 + }, + { + "epoch": 0.8192776418759616, + "grad_norm": 1.381181001663208, + "learning_rate": 1.9256653307282593e-06, + "loss": 0.4443, + "step": 18504 + }, + { + "epoch": 0.8193219175807754, + "grad_norm": 2.234727382659912, + "learning_rate": 1.9247536750271554e-06, + "loss": 0.5266, + "step": 18505 + }, + { + "epoch": 0.8193661932855894, + "grad_norm": 2.1031999588012695, + "learning_rate": 1.9238422121967348e-06, + "loss": 0.6171, + "step": 18506 + }, + { + "epoch": 0.8194104689904033, + "grad_norm": 1.2692675590515137, + "learning_rate": 1.922930942258766e-06, + "loss": 0.4233, + "step": 18507 + }, + { + "epoch": 0.8194547446952171, + "grad_norm": 1.77675461769104, + "learning_rate": 1.9220198652350164e-06, + "loss": 0.4567, + "step": 18508 + }, + { + "epoch": 0.819499020400031, + "grad_norm": 1.4655874967575073, + "learning_rate": 1.921108981147244e-06, + "loss": 0.5063, + "step": 18509 + }, + { + "epoch": 0.8195432961048449, + "grad_norm": 2.5479300022125244, + "learning_rate": 1.9201982900172034e-06, + "loss": 0.5453, + "step": 18510 + }, + { + "epoch": 0.8195875718096588, + "grad_norm": 1.449952483177185, + "learning_rate": 1.9192877918666497e-06, + "loss": 0.4708, + "step": 18511 + }, + { + "epoch": 0.8196318475144726, + "grad_norm": 1.5526937246322632, + "learning_rate": 1.918377486717329e-06, + "loss": 0.4386, + "step": 18512 + }, + { + "epoch": 0.8196761232192865, + "grad_norm": 1.459071397781372, + "learning_rate": 1.917467374590981e-06, + "loss": 0.4624, + "step": 18513 + }, + { + "epoch": 0.8197203989241004, + "grad_norm": 2.0015316009521484, + "learning_rate": 1.9165574555093403e-06, + "loss": 0.5258, + "step": 18514 + }, + { + "epoch": 0.8197646746289142, + "grad_norm": 1.4192707538604736, + "learning_rate": 1.9156477294941465e-06, + "loss": 0.4794, + "step": 18515 + }, + { + "epoch": 0.8198089503337281, + "grad_norm": 1.423335313796997, + "learning_rate": 1.9147381965671243e-06, + "loss": 0.51, + "step": 18516 + }, + { + "epoch": 0.819853226038542, + "grad_norm": 1.4067631959915161, + "learning_rate": 1.9138288567499963e-06, + "loss": 0.4728, + "step": 18517 + }, + { + "epoch": 0.8198975017433558, + "grad_norm": 1.6012208461761475, + "learning_rate": 1.9129197100644826e-06, + "loss": 0.5305, + "step": 18518 + }, + { + "epoch": 0.8199417774481698, + "grad_norm": 1.6065921783447266, + "learning_rate": 1.9120107565322964e-06, + "loss": 0.5206, + "step": 18519 + }, + { + "epoch": 0.8199860531529837, + "grad_norm": 1.313364863395691, + "learning_rate": 1.9111019961751455e-06, + "loss": 0.491, + "step": 18520 + }, + { + "epoch": 0.8200303288577975, + "grad_norm": 1.2460061311721802, + "learning_rate": 1.9101934290147394e-06, + "loss": 0.4592, + "step": 18521 + }, + { + "epoch": 0.8200746045626114, + "grad_norm": 1.8663480281829834, + "learning_rate": 1.9092850550727773e-06, + "loss": 0.5056, + "step": 18522 + }, + { + "epoch": 0.8201188802674253, + "grad_norm": 1.5717706680297852, + "learning_rate": 1.908376874370953e-06, + "loss": 0.4726, + "step": 18523 + }, + { + "epoch": 0.8201631559722391, + "grad_norm": 1.5409765243530273, + "learning_rate": 1.9074688869309565e-06, + "loss": 0.5156, + "step": 18524 + }, + { + "epoch": 0.820207431677053, + "grad_norm": 1.5350091457366943, + "learning_rate": 1.9065610927744782e-06, + "loss": 0.4803, + "step": 18525 + }, + { + "epoch": 0.8202517073818669, + "grad_norm": 1.4086538553237915, + "learning_rate": 1.9056534919231994e-06, + "loss": 0.4925, + "step": 18526 + }, + { + "epoch": 0.8202959830866807, + "grad_norm": 1.5531761646270752, + "learning_rate": 1.9047460843987963e-06, + "loss": 0.5087, + "step": 18527 + }, + { + "epoch": 0.8203402587914946, + "grad_norm": 1.706933856010437, + "learning_rate": 1.9038388702229404e-06, + "loss": 0.5441, + "step": 18528 + }, + { + "epoch": 0.8203845344963085, + "grad_norm": 1.6930193901062012, + "learning_rate": 1.902931849417302e-06, + "loss": 0.5214, + "step": 18529 + }, + { + "epoch": 0.8204288102011223, + "grad_norm": 1.478759527206421, + "learning_rate": 1.9020250220035407e-06, + "loss": 0.4651, + "step": 18530 + }, + { + "epoch": 0.8204730859059363, + "grad_norm": 2.3397016525268555, + "learning_rate": 1.9011183880033203e-06, + "loss": 0.5694, + "step": 18531 + }, + { + "epoch": 0.8205173616107502, + "grad_norm": 1.5573221445083618, + "learning_rate": 1.900211947438293e-06, + "loss": 0.4786, + "step": 18532 + }, + { + "epoch": 0.820561637315564, + "grad_norm": 1.5987547636032104, + "learning_rate": 1.8993057003301085e-06, + "loss": 0.5054, + "step": 18533 + }, + { + "epoch": 0.8206059130203779, + "grad_norm": 1.3183783292770386, + "learning_rate": 1.8983996467004085e-06, + "loss": 0.4421, + "step": 18534 + }, + { + "epoch": 0.8206501887251918, + "grad_norm": 1.8665366172790527, + "learning_rate": 1.8974937865708398e-06, + "loss": 0.582, + "step": 18535 + }, + { + "epoch": 0.8206944644300056, + "grad_norm": 1.5830271244049072, + "learning_rate": 1.8965881199630343e-06, + "loss": 0.5306, + "step": 18536 + }, + { + "epoch": 0.8207387401348195, + "grad_norm": 1.3191994428634644, + "learning_rate": 1.8956826468986235e-06, + "loss": 0.439, + "step": 18537 + }, + { + "epoch": 0.8207830158396334, + "grad_norm": 1.555224895477295, + "learning_rate": 1.894777367399233e-06, + "loss": 0.4635, + "step": 18538 + }, + { + "epoch": 0.8208272915444472, + "grad_norm": 1.335456132888794, + "learning_rate": 1.8938722814864863e-06, + "loss": 0.5016, + "step": 18539 + }, + { + "epoch": 0.8208715672492611, + "grad_norm": 1.4727113246917725, + "learning_rate": 1.8929673891819978e-06, + "loss": 0.5043, + "step": 18540 + }, + { + "epoch": 0.820915842954075, + "grad_norm": 1.653218150138855, + "learning_rate": 1.8920626905073846e-06, + "loss": 0.4691, + "step": 18541 + }, + { + "epoch": 0.8209601186588888, + "grad_norm": 1.3959729671478271, + "learning_rate": 1.891158185484252e-06, + "loss": 0.519, + "step": 18542 + }, + { + "epoch": 0.8210043943637028, + "grad_norm": 1.996428370475769, + "learning_rate": 1.8902538741342036e-06, + "loss": 0.5725, + "step": 18543 + }, + { + "epoch": 0.8210486700685167, + "grad_norm": 1.509902000427246, + "learning_rate": 1.8893497564788355e-06, + "loss": 0.4979, + "step": 18544 + }, + { + "epoch": 0.8210929457733305, + "grad_norm": 1.312906265258789, + "learning_rate": 1.888445832539747e-06, + "loss": 0.5114, + "step": 18545 + }, + { + "epoch": 0.8211372214781444, + "grad_norm": 1.7542474269866943, + "learning_rate": 1.8875421023385288e-06, + "loss": 0.4617, + "step": 18546 + }, + { + "epoch": 0.8211814971829583, + "grad_norm": 1.8730971813201904, + "learning_rate": 1.8866385658967556e-06, + "loss": 0.5322, + "step": 18547 + }, + { + "epoch": 0.8212257728877722, + "grad_norm": 1.3526171445846558, + "learning_rate": 1.8857352232360183e-06, + "loss": 0.4536, + "step": 18548 + }, + { + "epoch": 0.821270048592586, + "grad_norm": 1.6647958755493164, + "learning_rate": 1.884832074377887e-06, + "loss": 0.4713, + "step": 18549 + }, + { + "epoch": 0.8213143242973999, + "grad_norm": 1.3512846231460571, + "learning_rate": 1.8839291193439313e-06, + "loss": 0.4297, + "step": 18550 + }, + { + "epoch": 0.8213586000022138, + "grad_norm": 1.6517399549484253, + "learning_rate": 1.883026358155723e-06, + "loss": 0.5109, + "step": 18551 + }, + { + "epoch": 0.8214028757070276, + "grad_norm": 1.875617504119873, + "learning_rate": 1.8821237908348211e-06, + "loss": 0.4878, + "step": 18552 + }, + { + "epoch": 0.8214471514118415, + "grad_norm": 1.7394579648971558, + "learning_rate": 1.8812214174027832e-06, + "loss": 0.4948, + "step": 18553 + }, + { + "epoch": 0.8214914271166555, + "grad_norm": 1.3706014156341553, + "learning_rate": 1.8803192378811608e-06, + "loss": 0.4671, + "step": 18554 + }, + { + "epoch": 0.8215357028214693, + "grad_norm": 1.409820556640625, + "learning_rate": 1.8794172522915022e-06, + "loss": 0.4908, + "step": 18555 + }, + { + "epoch": 0.8215799785262832, + "grad_norm": 1.454891324043274, + "learning_rate": 1.87851546065535e-06, + "loss": 0.4743, + "step": 18556 + }, + { + "epoch": 0.8216242542310971, + "grad_norm": 1.9551681280136108, + "learning_rate": 1.877613862994242e-06, + "loss": 0.555, + "step": 18557 + }, + { + "epoch": 0.8216685299359109, + "grad_norm": 1.5402406454086304, + "learning_rate": 1.8767124593297147e-06, + "loss": 0.5558, + "step": 18558 + }, + { + "epoch": 0.8217128056407248, + "grad_norm": 1.4415624141693115, + "learning_rate": 1.875811249683296e-06, + "loss": 0.4738, + "step": 18559 + }, + { + "epoch": 0.8217570813455387, + "grad_norm": 1.650550365447998, + "learning_rate": 1.874910234076509e-06, + "loss": 0.551, + "step": 18560 + }, + { + "epoch": 0.8218013570503525, + "grad_norm": 1.5120277404785156, + "learning_rate": 1.8740094125308773e-06, + "loss": 0.4883, + "step": 18561 + }, + { + "epoch": 0.8218456327551664, + "grad_norm": 1.5299351215362549, + "learning_rate": 1.8731087850679142e-06, + "loss": 0.5032, + "step": 18562 + }, + { + "epoch": 0.8218899084599803, + "grad_norm": 1.3943099975585938, + "learning_rate": 1.8722083517091317e-06, + "loss": 0.4941, + "step": 18563 + }, + { + "epoch": 0.8219341841647941, + "grad_norm": 1.7255020141601562, + "learning_rate": 1.8713081124760335e-06, + "loss": 0.538, + "step": 18564 + }, + { + "epoch": 0.821978459869608, + "grad_norm": 1.2527732849121094, + "learning_rate": 1.8704080673901227e-06, + "loss": 0.4187, + "step": 18565 + }, + { + "epoch": 0.822022735574422, + "grad_norm": 1.4222912788391113, + "learning_rate": 1.8695082164728962e-06, + "loss": 0.4827, + "step": 18566 + }, + { + "epoch": 0.8220670112792358, + "grad_norm": 2.0559427738189697, + "learning_rate": 1.8686085597458437e-06, + "loss": 0.4891, + "step": 18567 + }, + { + "epoch": 0.8221112869840497, + "grad_norm": 1.5222135782241821, + "learning_rate": 1.8677090972304567e-06, + "loss": 0.4651, + "step": 18568 + }, + { + "epoch": 0.8221555626888636, + "grad_norm": 1.3552168607711792, + "learning_rate": 1.8668098289482163e-06, + "loss": 0.4068, + "step": 18569 + }, + { + "epoch": 0.8221998383936774, + "grad_norm": 1.3486899137496948, + "learning_rate": 1.865910754920599e-06, + "loss": 0.4665, + "step": 18570 + }, + { + "epoch": 0.8222441140984913, + "grad_norm": 1.3977192640304565, + "learning_rate": 1.8650118751690827e-06, + "loss": 0.4663, + "step": 18571 + }, + { + "epoch": 0.8222883898033052, + "grad_norm": 1.4135843515396118, + "learning_rate": 1.8641131897151331e-06, + "loss": 0.4725, + "step": 18572 + }, + { + "epoch": 0.822332665508119, + "grad_norm": 1.77352774143219, + "learning_rate": 1.863214698580217e-06, + "loss": 0.572, + "step": 18573 + }, + { + "epoch": 0.8223769412129329, + "grad_norm": 1.4630532264709473, + "learning_rate": 1.8623164017857908e-06, + "loss": 0.5109, + "step": 18574 + }, + { + "epoch": 0.8224212169177468, + "grad_norm": 1.8362910747528076, + "learning_rate": 1.8614182993533126e-06, + "loss": 0.5195, + "step": 18575 + }, + { + "epoch": 0.8224654926225606, + "grad_norm": 1.4046679735183716, + "learning_rate": 1.8605203913042313e-06, + "loss": 0.4731, + "step": 18576 + }, + { + "epoch": 0.8225097683273745, + "grad_norm": 1.5588419437408447, + "learning_rate": 1.8596226776599913e-06, + "loss": 0.4946, + "step": 18577 + }, + { + "epoch": 0.8225540440321885, + "grad_norm": 1.5156139135360718, + "learning_rate": 1.8587251584420374e-06, + "loss": 0.4815, + "step": 18578 + }, + { + "epoch": 0.8225983197370023, + "grad_norm": 1.3882933855056763, + "learning_rate": 1.8578278336718037e-06, + "loss": 0.521, + "step": 18579 + }, + { + "epoch": 0.8226425954418162, + "grad_norm": 1.5632728338241577, + "learning_rate": 1.8569307033707208e-06, + "loss": 0.5628, + "step": 18580 + }, + { + "epoch": 0.8226868711466301, + "grad_norm": 1.3400442600250244, + "learning_rate": 1.8560337675602203e-06, + "loss": 0.4476, + "step": 18581 + }, + { + "epoch": 0.822731146851444, + "grad_norm": 1.4917662143707275, + "learning_rate": 1.8551370262617218e-06, + "loss": 0.5199, + "step": 18582 + }, + { + "epoch": 0.8227754225562578, + "grad_norm": 1.7219724655151367, + "learning_rate": 1.854240479496643e-06, + "loss": 0.5273, + "step": 18583 + }, + { + "epoch": 0.8228196982610717, + "grad_norm": 1.425127387046814, + "learning_rate": 1.8533441272863983e-06, + "loss": 0.4469, + "step": 18584 + }, + { + "epoch": 0.8228639739658856, + "grad_norm": 1.5257278680801392, + "learning_rate": 1.8524479696523967e-06, + "loss": 0.4886, + "step": 18585 + }, + { + "epoch": 0.8229082496706994, + "grad_norm": 1.6616768836975098, + "learning_rate": 1.8515520066160409e-06, + "loss": 0.5096, + "step": 18586 + }, + { + "epoch": 0.8229525253755133, + "grad_norm": 1.676282525062561, + "learning_rate": 1.850656238198728e-06, + "loss": 0.5361, + "step": 18587 + }, + { + "epoch": 0.8229968010803272, + "grad_norm": 1.3595561981201172, + "learning_rate": 1.849760664421858e-06, + "loss": 0.4708, + "step": 18588 + }, + { + "epoch": 0.823041076785141, + "grad_norm": 1.4961186647415161, + "learning_rate": 1.848865285306818e-06, + "loss": 0.5038, + "step": 18589 + }, + { + "epoch": 0.823085352489955, + "grad_norm": 1.4795688390731812, + "learning_rate": 1.8479701008749917e-06, + "loss": 0.5077, + "step": 18590 + }, + { + "epoch": 0.8231296281947689, + "grad_norm": 1.231047511100769, + "learning_rate": 1.8470751111477648e-06, + "loss": 0.4169, + "step": 18591 + }, + { + "epoch": 0.8231739038995827, + "grad_norm": 1.508945107460022, + "learning_rate": 1.8461803161465098e-06, + "loss": 0.4826, + "step": 18592 + }, + { + "epoch": 0.8232181796043966, + "grad_norm": 1.6951179504394531, + "learning_rate": 1.845285715892602e-06, + "loss": 0.543, + "step": 18593 + }, + { + "epoch": 0.8232624553092105, + "grad_norm": 1.5520622730255127, + "learning_rate": 1.8443913104073984e-06, + "loss": 0.5151, + "step": 18594 + }, + { + "epoch": 0.8233067310140243, + "grad_norm": 1.514380693435669, + "learning_rate": 1.8434970997122725e-06, + "loss": 0.4555, + "step": 18595 + }, + { + "epoch": 0.8233510067188382, + "grad_norm": 1.4438873529434204, + "learning_rate": 1.8426030838285757e-06, + "loss": 0.5081, + "step": 18596 + }, + { + "epoch": 0.8233952824236521, + "grad_norm": 1.4104578495025635, + "learning_rate": 1.8417092627776601e-06, + "loss": 0.4532, + "step": 18597 + }, + { + "epoch": 0.8234395581284659, + "grad_norm": 1.6019530296325684, + "learning_rate": 1.8408156365808782e-06, + "loss": 0.5111, + "step": 18598 + }, + { + "epoch": 0.8234838338332798, + "grad_norm": 1.6779487133026123, + "learning_rate": 1.8399222052595721e-06, + "loss": 0.4858, + "step": 18599 + }, + { + "epoch": 0.8235281095380937, + "grad_norm": 1.6787296533584595, + "learning_rate": 1.839028968835077e-06, + "loss": 0.5345, + "step": 18600 + }, + { + "epoch": 0.8235723852429075, + "grad_norm": 1.4954571723937988, + "learning_rate": 1.8381359273287335e-06, + "loss": 0.5313, + "step": 18601 + }, + { + "epoch": 0.8236166609477215, + "grad_norm": 1.7093168497085571, + "learning_rate": 1.837243080761869e-06, + "loss": 0.5158, + "step": 18602 + }, + { + "epoch": 0.8236609366525354, + "grad_norm": 1.3263658285140991, + "learning_rate": 1.8363504291558053e-06, + "loss": 0.4013, + "step": 18603 + }, + { + "epoch": 0.8237052123573492, + "grad_norm": 1.5723422765731812, + "learning_rate": 1.8354579725318622e-06, + "loss": 0.5252, + "step": 18604 + }, + { + "epoch": 0.8237494880621631, + "grad_norm": 1.6675400733947754, + "learning_rate": 1.834565710911359e-06, + "loss": 0.5001, + "step": 18605 + }, + { + "epoch": 0.823793763766977, + "grad_norm": 1.234756588935852, + "learning_rate": 1.8336736443156066e-06, + "loss": 0.5021, + "step": 18606 + }, + { + "epoch": 0.8238380394717908, + "grad_norm": 1.3412200212478638, + "learning_rate": 1.8327817727659058e-06, + "loss": 0.4643, + "step": 18607 + }, + { + "epoch": 0.8238823151766047, + "grad_norm": 1.5181798934936523, + "learning_rate": 1.8318900962835661e-06, + "loss": 0.53, + "step": 18608 + }, + { + "epoch": 0.8239265908814186, + "grad_norm": 1.633617877960205, + "learning_rate": 1.8309986148898807e-06, + "loss": 0.5069, + "step": 18609 + }, + { + "epoch": 0.8239708665862324, + "grad_norm": 1.484304428100586, + "learning_rate": 1.8301073286061422e-06, + "loss": 0.4797, + "step": 18610 + }, + { + "epoch": 0.8240151422910463, + "grad_norm": 1.4726213216781616, + "learning_rate": 1.829216237453637e-06, + "loss": 0.4612, + "step": 18611 + }, + { + "epoch": 0.8240594179958602, + "grad_norm": 1.6141263246536255, + "learning_rate": 1.8283253414536495e-06, + "loss": 0.4874, + "step": 18612 + }, + { + "epoch": 0.824103693700674, + "grad_norm": 1.344029426574707, + "learning_rate": 1.8274346406274578e-06, + "loss": 0.4801, + "step": 18613 + }, + { + "epoch": 0.824147969405488, + "grad_norm": 1.5185940265655518, + "learning_rate": 1.826544134996333e-06, + "loss": 0.4785, + "step": 18614 + }, + { + "epoch": 0.8241922451103019, + "grad_norm": 1.4835258722305298, + "learning_rate": 1.825653824581548e-06, + "loss": 0.4931, + "step": 18615 + }, + { + "epoch": 0.8242365208151158, + "grad_norm": 1.4019079208374023, + "learning_rate": 1.8247637094043668e-06, + "loss": 0.5152, + "step": 18616 + }, + { + "epoch": 0.8242807965199296, + "grad_norm": 1.485005497932434, + "learning_rate": 1.8238737894860448e-06, + "loss": 0.4987, + "step": 18617 + }, + { + "epoch": 0.8243250722247435, + "grad_norm": 1.5574405193328857, + "learning_rate": 1.8229840648478425e-06, + "loss": 0.4549, + "step": 18618 + }, + { + "epoch": 0.8243693479295574, + "grad_norm": 1.381368637084961, + "learning_rate": 1.8220945355110087e-06, + "loss": 0.4901, + "step": 18619 + }, + { + "epoch": 0.8244136236343712, + "grad_norm": 1.4605743885040283, + "learning_rate": 1.8212052014967873e-06, + "loss": 0.458, + "step": 18620 + }, + { + "epoch": 0.8244578993391851, + "grad_norm": 1.3744844198226929, + "learning_rate": 1.8203160628264204e-06, + "loss": 0.4629, + "step": 18621 + }, + { + "epoch": 0.824502175043999, + "grad_norm": 2.247981071472168, + "learning_rate": 1.8194271195211443e-06, + "loss": 0.5263, + "step": 18622 + }, + { + "epoch": 0.8245464507488128, + "grad_norm": 2.4442059993743896, + "learning_rate": 1.8185383716021898e-06, + "loss": 0.6192, + "step": 18623 + }, + { + "epoch": 0.8245907264536267, + "grad_norm": 1.23991858959198, + "learning_rate": 1.8176498190907832e-06, + "loss": 0.4719, + "step": 18624 + }, + { + "epoch": 0.8246350021584407, + "grad_norm": 1.3969273567199707, + "learning_rate": 1.8167614620081509e-06, + "loss": 0.4486, + "step": 18625 + }, + { + "epoch": 0.8246792778632545, + "grad_norm": 1.64665687084198, + "learning_rate": 1.815873300375508e-06, + "loss": 0.5656, + "step": 18626 + }, + { + "epoch": 0.8247235535680684, + "grad_norm": 1.4742050170898438, + "learning_rate": 1.8149853342140644e-06, + "loss": 0.4833, + "step": 18627 + }, + { + "epoch": 0.8247678292728823, + "grad_norm": 1.4543546438217163, + "learning_rate": 1.8140975635450342e-06, + "loss": 0.5385, + "step": 18628 + }, + { + "epoch": 0.8248121049776961, + "grad_norm": 1.2884197235107422, + "learning_rate": 1.8132099883896194e-06, + "loss": 0.4384, + "step": 18629 + }, + { + "epoch": 0.82485638068251, + "grad_norm": 1.6499547958374023, + "learning_rate": 1.812322608769017e-06, + "loss": 0.4977, + "step": 18630 + }, + { + "epoch": 0.8249006563873239, + "grad_norm": 1.5659409761428833, + "learning_rate": 1.811435424704423e-06, + "loss": 0.4248, + "step": 18631 + }, + { + "epoch": 0.8249449320921377, + "grad_norm": 1.6648683547973633, + "learning_rate": 1.8105484362170268e-06, + "loss": 0.4799, + "step": 18632 + }, + { + "epoch": 0.8249892077969516, + "grad_norm": 1.65871262550354, + "learning_rate": 1.8096616433280123e-06, + "loss": 0.4967, + "step": 18633 + }, + { + "epoch": 0.8250334835017655, + "grad_norm": 1.4382271766662598, + "learning_rate": 1.8087750460585584e-06, + "loss": 0.4985, + "step": 18634 + }, + { + "epoch": 0.8250777592065793, + "grad_norm": 1.4007766246795654, + "learning_rate": 1.8078886444298461e-06, + "loss": 0.4726, + "step": 18635 + }, + { + "epoch": 0.8251220349113932, + "grad_norm": 1.5101248025894165, + "learning_rate": 1.8070024384630424e-06, + "loss": 0.5443, + "step": 18636 + }, + { + "epoch": 0.8251663106162072, + "grad_norm": 2.071770429611206, + "learning_rate": 1.806116428179312e-06, + "loss": 0.4599, + "step": 18637 + }, + { + "epoch": 0.825210586321021, + "grad_norm": 2.1847572326660156, + "learning_rate": 1.805230613599822e-06, + "loss": 0.6232, + "step": 18638 + }, + { + "epoch": 0.8252548620258349, + "grad_norm": 1.7130868434906006, + "learning_rate": 1.804344994745727e-06, + "loss": 0.6133, + "step": 18639 + }, + { + "epoch": 0.8252991377306488, + "grad_norm": 1.5935556888580322, + "learning_rate": 1.8034595716381798e-06, + "loss": 0.5145, + "step": 18640 + }, + { + "epoch": 0.8253434134354626, + "grad_norm": 3.1136207580566406, + "learning_rate": 1.8025743442983223e-06, + "loss": 0.5225, + "step": 18641 + }, + { + "epoch": 0.8253876891402765, + "grad_norm": 1.4352073669433594, + "learning_rate": 1.8016893127473046e-06, + "loss": 0.4852, + "step": 18642 + }, + { + "epoch": 0.8254319648450904, + "grad_norm": 1.6470047235488892, + "learning_rate": 1.8008044770062628e-06, + "loss": 0.5596, + "step": 18643 + }, + { + "epoch": 0.8254762405499042, + "grad_norm": 1.4909590482711792, + "learning_rate": 1.7999198370963267e-06, + "loss": 0.4858, + "step": 18644 + }, + { + "epoch": 0.8255205162547181, + "grad_norm": 1.4398874044418335, + "learning_rate": 1.7990353930386317e-06, + "loss": 0.5146, + "step": 18645 + }, + { + "epoch": 0.825564791959532, + "grad_norm": 1.7564237117767334, + "learning_rate": 1.7981511448542988e-06, + "loss": 0.5296, + "step": 18646 + }, + { + "epoch": 0.8256090676643458, + "grad_norm": 1.542415976524353, + "learning_rate": 1.7972670925644441e-06, + "loss": 0.4669, + "step": 18647 + }, + { + "epoch": 0.8256533433691597, + "grad_norm": 1.5831669569015503, + "learning_rate": 1.79638323619019e-06, + "loss": 0.5095, + "step": 18648 + }, + { + "epoch": 0.8256976190739737, + "grad_norm": 1.6826800107955933, + "learning_rate": 1.795499575752645e-06, + "loss": 0.5503, + "step": 18649 + }, + { + "epoch": 0.8257418947787875, + "grad_norm": 1.5224655866622925, + "learning_rate": 1.7946161112729088e-06, + "loss": 0.5011, + "step": 18650 + }, + { + "epoch": 0.8257861704836014, + "grad_norm": 1.2640718221664429, + "learning_rate": 1.7937328427720834e-06, + "loss": 0.4765, + "step": 18651 + }, + { + "epoch": 0.8258304461884153, + "grad_norm": 1.4746909141540527, + "learning_rate": 1.7928497702712698e-06, + "loss": 0.4435, + "step": 18652 + }, + { + "epoch": 0.8258747218932292, + "grad_norm": 1.6890220642089844, + "learning_rate": 1.791966893791558e-06, + "loss": 0.5107, + "step": 18653 + }, + { + "epoch": 0.825918997598043, + "grad_norm": 1.3648751974105835, + "learning_rate": 1.7910842133540296e-06, + "loss": 0.5021, + "step": 18654 + }, + { + "epoch": 0.8259632733028569, + "grad_norm": 1.7001210451126099, + "learning_rate": 1.7902017289797746e-06, + "loss": 0.5541, + "step": 18655 + }, + { + "epoch": 0.8260075490076708, + "grad_norm": 1.4581555128097534, + "learning_rate": 1.7893194406898662e-06, + "loss": 0.4527, + "step": 18656 + }, + { + "epoch": 0.8260518247124846, + "grad_norm": 1.5936157703399658, + "learning_rate": 1.7884373485053774e-06, + "loss": 0.4919, + "step": 18657 + }, + { + "epoch": 0.8260961004172985, + "grad_norm": 1.499961256980896, + "learning_rate": 1.787555452447376e-06, + "loss": 0.4848, + "step": 18658 + }, + { + "epoch": 0.8261403761221124, + "grad_norm": 1.6000878810882568, + "learning_rate": 1.786673752536927e-06, + "loss": 0.4807, + "step": 18659 + }, + { + "epoch": 0.8261846518269262, + "grad_norm": 1.6142975091934204, + "learning_rate": 1.7857922487950873e-06, + "loss": 0.513, + "step": 18660 + }, + { + "epoch": 0.8262289275317402, + "grad_norm": 1.613716721534729, + "learning_rate": 1.784910941242909e-06, + "loss": 0.5479, + "step": 18661 + }, + { + "epoch": 0.8262732032365541, + "grad_norm": 1.6513129472732544, + "learning_rate": 1.7840298299014458e-06, + "loss": 0.5396, + "step": 18662 + }, + { + "epoch": 0.8263174789413679, + "grad_norm": 1.6587105989456177, + "learning_rate": 1.7831489147917413e-06, + "loss": 0.5103, + "step": 18663 + }, + { + "epoch": 0.8263617546461818, + "grad_norm": 1.3129396438598633, + "learning_rate": 1.7822681959348309e-06, + "loss": 0.4504, + "step": 18664 + }, + { + "epoch": 0.8264060303509957, + "grad_norm": 1.6175593137741089, + "learning_rate": 1.7813876733517576e-06, + "loss": 0.4838, + "step": 18665 + }, + { + "epoch": 0.8264503060558095, + "grad_norm": 2.1702818870544434, + "learning_rate": 1.780507347063547e-06, + "loss": 0.528, + "step": 18666 + }, + { + "epoch": 0.8264945817606234, + "grad_norm": 1.511395812034607, + "learning_rate": 1.7796272170912255e-06, + "loss": 0.549, + "step": 18667 + }, + { + "epoch": 0.8265388574654373, + "grad_norm": 1.5338009595870972, + "learning_rate": 1.778747283455815e-06, + "loss": 0.474, + "step": 18668 + }, + { + "epoch": 0.8265831331702511, + "grad_norm": 1.9065474271774292, + "learning_rate": 1.7778675461783313e-06, + "loss": 0.5355, + "step": 18669 + }, + { + "epoch": 0.826627408875065, + "grad_norm": 1.4688969850540161, + "learning_rate": 1.7769880052797872e-06, + "loss": 0.5004, + "step": 18670 + }, + { + "epoch": 0.8266716845798789, + "grad_norm": 1.5202631950378418, + "learning_rate": 1.7761086607811862e-06, + "loss": 0.5034, + "step": 18671 + }, + { + "epoch": 0.8267159602846927, + "grad_norm": 1.2925209999084473, + "learning_rate": 1.7752295127035368e-06, + "loss": 0.4513, + "step": 18672 + }, + { + "epoch": 0.8267602359895067, + "grad_norm": 1.4058611392974854, + "learning_rate": 1.7743505610678336e-06, + "loss": 0.4773, + "step": 18673 + }, + { + "epoch": 0.8268045116943206, + "grad_norm": 1.2501827478408813, + "learning_rate": 1.7734718058950662e-06, + "loss": 0.4134, + "step": 18674 + }, + { + "epoch": 0.8268487873991344, + "grad_norm": 1.6735434532165527, + "learning_rate": 1.7725932472062302e-06, + "loss": 0.5691, + "step": 18675 + }, + { + "epoch": 0.8268930631039483, + "grad_norm": 1.290099859237671, + "learning_rate": 1.7717148850223043e-06, + "loss": 0.4504, + "step": 18676 + }, + { + "epoch": 0.8269373388087622, + "grad_norm": 1.584269642829895, + "learning_rate": 1.7708367193642694e-06, + "loss": 0.4953, + "step": 18677 + }, + { + "epoch": 0.826981614513576, + "grad_norm": 1.7917081117630005, + "learning_rate": 1.769958750253099e-06, + "loss": 0.4863, + "step": 18678 + }, + { + "epoch": 0.8270258902183899, + "grad_norm": 1.6398332118988037, + "learning_rate": 1.7690809777097628e-06, + "loss": 0.5253, + "step": 18679 + }, + { + "epoch": 0.8270701659232038, + "grad_norm": 1.9507427215576172, + "learning_rate": 1.7682034017552252e-06, + "loss": 0.5409, + "step": 18680 + }, + { + "epoch": 0.8271144416280176, + "grad_norm": 1.5063846111297607, + "learning_rate": 1.7673260224104437e-06, + "loss": 0.4657, + "step": 18681 + }, + { + "epoch": 0.8271587173328315, + "grad_norm": 1.3531538248062134, + "learning_rate": 1.766448839696381e-06, + "loss": 0.4531, + "step": 18682 + }, + { + "epoch": 0.8272029930376454, + "grad_norm": 1.535711407661438, + "learning_rate": 1.7655718536339817e-06, + "loss": 0.525, + "step": 18683 + }, + { + "epoch": 0.8272472687424592, + "grad_norm": 1.7347548007965088, + "learning_rate": 1.764695064244193e-06, + "loss": 0.5236, + "step": 18684 + }, + { + "epoch": 0.8272915444472732, + "grad_norm": 1.3733936548233032, + "learning_rate": 1.7638184715479589e-06, + "loss": 0.4875, + "step": 18685 + }, + { + "epoch": 0.8273358201520871, + "grad_norm": 1.8140318393707275, + "learning_rate": 1.7629420755662152e-06, + "loss": 0.5504, + "step": 18686 + }, + { + "epoch": 0.827380095856901, + "grad_norm": 1.3796122074127197, + "learning_rate": 1.7620658763198927e-06, + "loss": 0.495, + "step": 18687 + }, + { + "epoch": 0.8274243715617148, + "grad_norm": 1.6190032958984375, + "learning_rate": 1.7611898738299193e-06, + "loss": 0.5332, + "step": 18688 + }, + { + "epoch": 0.8274686472665287, + "grad_norm": 1.3691498041152954, + "learning_rate": 1.7603140681172182e-06, + "loss": 0.4516, + "step": 18689 + }, + { + "epoch": 0.8275129229713426, + "grad_norm": 1.3891754150390625, + "learning_rate": 1.759438459202706e-06, + "loss": 0.4376, + "step": 18690 + }, + { + "epoch": 0.8275571986761564, + "grad_norm": 2.209545135498047, + "learning_rate": 1.7585630471072933e-06, + "loss": 0.5267, + "step": 18691 + }, + { + "epoch": 0.8276014743809703, + "grad_norm": 1.3517944812774658, + "learning_rate": 1.7576878318518953e-06, + "loss": 0.473, + "step": 18692 + }, + { + "epoch": 0.8276457500857842, + "grad_norm": 1.5969799757003784, + "learning_rate": 1.7568128134574114e-06, + "loss": 0.5291, + "step": 18693 + }, + { + "epoch": 0.827690025790598, + "grad_norm": 1.7053844928741455, + "learning_rate": 1.7559379919447395e-06, + "loss": 0.5696, + "step": 18694 + }, + { + "epoch": 0.8277343014954119, + "grad_norm": 1.535943865776062, + "learning_rate": 1.7550633673347783e-06, + "loss": 0.4737, + "step": 18695 + }, + { + "epoch": 0.8277785772002259, + "grad_norm": 1.505240797996521, + "learning_rate": 1.754188939648418e-06, + "loss": 0.4987, + "step": 18696 + }, + { + "epoch": 0.8278228529050397, + "grad_norm": 1.5443620681762695, + "learning_rate": 1.7533147089065362e-06, + "loss": 0.4972, + "step": 18697 + }, + { + "epoch": 0.8278671286098536, + "grad_norm": 1.536086082458496, + "learning_rate": 1.7524406751300204e-06, + "loss": 0.5146, + "step": 18698 + }, + { + "epoch": 0.8279114043146675, + "grad_norm": 1.3095769882202148, + "learning_rate": 1.7515668383397433e-06, + "loss": 0.4544, + "step": 18699 + }, + { + "epoch": 0.8279556800194813, + "grad_norm": 1.4699941873550415, + "learning_rate": 1.7506931985565768e-06, + "loss": 0.5139, + "step": 18700 + }, + { + "epoch": 0.8279999557242952, + "grad_norm": 1.4596470594406128, + "learning_rate": 1.7498197558013829e-06, + "loss": 0.4656, + "step": 18701 + }, + { + "epoch": 0.8280442314291091, + "grad_norm": 1.413761854171753, + "learning_rate": 1.7489465100950297e-06, + "loss": 0.4795, + "step": 18702 + }, + { + "epoch": 0.8280885071339229, + "grad_norm": 1.4818845987319946, + "learning_rate": 1.7480734614583706e-06, + "loss": 0.4622, + "step": 18703 + }, + { + "epoch": 0.8281327828387368, + "grad_norm": 1.3603886365890503, + "learning_rate": 1.747200609912255e-06, + "loss": 0.492, + "step": 18704 + }, + { + "epoch": 0.8281770585435507, + "grad_norm": 1.4499775171279907, + "learning_rate": 1.7463279554775392e-06, + "loss": 0.4589, + "step": 18705 + }, + { + "epoch": 0.8282213342483645, + "grad_norm": 1.4050730466842651, + "learning_rate": 1.7454554981750571e-06, + "loss": 0.4541, + "step": 18706 + }, + { + "epoch": 0.8282656099531784, + "grad_norm": 1.9444835186004639, + "learning_rate": 1.7445832380256466e-06, + "loss": 0.5641, + "step": 18707 + }, + { + "epoch": 0.8283098856579924, + "grad_norm": 1.737115740776062, + "learning_rate": 1.7437111750501457e-06, + "loss": 0.5498, + "step": 18708 + }, + { + "epoch": 0.8283541613628062, + "grad_norm": 1.5510343313217163, + "learning_rate": 1.742839309269382e-06, + "loss": 0.526, + "step": 18709 + }, + { + "epoch": 0.8283984370676201, + "grad_norm": 2.218013286590576, + "learning_rate": 1.7419676407041763e-06, + "loss": 0.5203, + "step": 18710 + }, + { + "epoch": 0.828442712772434, + "grad_norm": 1.7967181205749512, + "learning_rate": 1.7410961693753492e-06, + "loss": 0.5599, + "step": 18711 + }, + { + "epoch": 0.8284869884772478, + "grad_norm": 2.2068333625793457, + "learning_rate": 1.740224895303716e-06, + "loss": 0.6001, + "step": 18712 + }, + { + "epoch": 0.8285312641820617, + "grad_norm": 2.246838331222534, + "learning_rate": 1.7393538185100866e-06, + "loss": 0.6193, + "step": 18713 + }, + { + "epoch": 0.8285755398868756, + "grad_norm": 1.3571569919586182, + "learning_rate": 1.7384829390152659e-06, + "loss": 0.4432, + "step": 18714 + }, + { + "epoch": 0.8286198155916894, + "grad_norm": 1.6770133972167969, + "learning_rate": 1.7376122568400533e-06, + "loss": 0.527, + "step": 18715 + }, + { + "epoch": 0.8286640912965033, + "grad_norm": 1.503461241722107, + "learning_rate": 1.7367417720052437e-06, + "loss": 0.4825, + "step": 18716 + }, + { + "epoch": 0.8287083670013172, + "grad_norm": 1.615347981452942, + "learning_rate": 1.735871484531626e-06, + "loss": 0.5375, + "step": 18717 + }, + { + "epoch": 0.828752642706131, + "grad_norm": 1.807667851448059, + "learning_rate": 1.7350013944399924e-06, + "loss": 0.5311, + "step": 18718 + }, + { + "epoch": 0.8287969184109449, + "grad_norm": 1.430436611175537, + "learning_rate": 1.7341315017511207e-06, + "loss": 0.4998, + "step": 18719 + }, + { + "epoch": 0.8288411941157589, + "grad_norm": 1.581822156906128, + "learning_rate": 1.7332618064857875e-06, + "loss": 0.5763, + "step": 18720 + }, + { + "epoch": 0.8288854698205727, + "grad_norm": 1.3992334604263306, + "learning_rate": 1.732392308664762e-06, + "loss": 0.4638, + "step": 18721 + }, + { + "epoch": 0.8289297455253866, + "grad_norm": 1.4732911586761475, + "learning_rate": 1.7315230083088175e-06, + "loss": 0.5137, + "step": 18722 + }, + { + "epoch": 0.8289740212302005, + "grad_norm": 1.3917038440704346, + "learning_rate": 1.730653905438714e-06, + "loss": 0.4676, + "step": 18723 + }, + { + "epoch": 0.8290182969350144, + "grad_norm": 1.8386081457138062, + "learning_rate": 1.7297850000752082e-06, + "loss": 0.5156, + "step": 18724 + }, + { + "epoch": 0.8290625726398282, + "grad_norm": 2.1769089698791504, + "learning_rate": 1.7289162922390533e-06, + "loss": 0.5606, + "step": 18725 + }, + { + "epoch": 0.8291068483446421, + "grad_norm": 1.3424415588378906, + "learning_rate": 1.7280477819509988e-06, + "loss": 0.4555, + "step": 18726 + }, + { + "epoch": 0.829151124049456, + "grad_norm": 1.4272279739379883, + "learning_rate": 1.7271794692317877e-06, + "loss": 0.4722, + "step": 18727 + }, + { + "epoch": 0.8291953997542698, + "grad_norm": 1.495023488998413, + "learning_rate": 1.726311354102157e-06, + "loss": 0.4539, + "step": 18728 + }, + { + "epoch": 0.8292396754590837, + "grad_norm": 1.5472910404205322, + "learning_rate": 1.7254434365828453e-06, + "loss": 0.5221, + "step": 18729 + }, + { + "epoch": 0.8292839511638976, + "grad_norm": 1.5607173442840576, + "learning_rate": 1.7245757166945786e-06, + "loss": 0.515, + "step": 18730 + }, + { + "epoch": 0.8293282268687114, + "grad_norm": 1.470381498336792, + "learning_rate": 1.7237081944580813e-06, + "loss": 0.4854, + "step": 18731 + }, + { + "epoch": 0.8293725025735253, + "grad_norm": 2.5491678714752197, + "learning_rate": 1.722840869894078e-06, + "loss": 0.6126, + "step": 18732 + }, + { + "epoch": 0.8294167782783393, + "grad_norm": 1.7170486450195312, + "learning_rate": 1.7219737430232798e-06, + "loss": 0.5663, + "step": 18733 + }, + { + "epoch": 0.8294610539831531, + "grad_norm": 1.3686662912368774, + "learning_rate": 1.7211068138663988e-06, + "loss": 0.4666, + "step": 18734 + }, + { + "epoch": 0.829505329687967, + "grad_norm": 1.7506277561187744, + "learning_rate": 1.7202400824441413e-06, + "loss": 0.5144, + "step": 18735 + }, + { + "epoch": 0.8295496053927809, + "grad_norm": 1.6330374479293823, + "learning_rate": 1.7193735487772078e-06, + "loss": 0.5023, + "step": 18736 + }, + { + "epoch": 0.8295938810975947, + "grad_norm": 1.702699899673462, + "learning_rate": 1.7185072128862934e-06, + "loss": 0.4871, + "step": 18737 + }, + { + "epoch": 0.8296381568024086, + "grad_norm": 1.4117004871368408, + "learning_rate": 1.7176410747920902e-06, + "loss": 0.465, + "step": 18738 + }, + { + "epoch": 0.8296824325072225, + "grad_norm": 1.423275351524353, + "learning_rate": 1.7167751345152883e-06, + "loss": 0.5102, + "step": 18739 + }, + { + "epoch": 0.8297267082120363, + "grad_norm": 1.3425379991531372, + "learning_rate": 1.7159093920765668e-06, + "loss": 0.505, + "step": 18740 + }, + { + "epoch": 0.8297709839168502, + "grad_norm": 1.370528221130371, + "learning_rate": 1.7150438474966036e-06, + "loss": 0.4937, + "step": 18741 + }, + { + "epoch": 0.8298152596216641, + "grad_norm": 1.5230863094329834, + "learning_rate": 1.714178500796073e-06, + "loss": 0.5002, + "step": 18742 + }, + { + "epoch": 0.8298595353264779, + "grad_norm": 1.3180630207061768, + "learning_rate": 1.7133133519956436e-06, + "loss": 0.432, + "step": 18743 + }, + { + "epoch": 0.8299038110312918, + "grad_norm": 1.5079083442687988, + "learning_rate": 1.712448401115977e-06, + "loss": 0.5122, + "step": 18744 + }, + { + "epoch": 0.8299480867361058, + "grad_norm": 1.2649402618408203, + "learning_rate": 1.7115836481777325e-06, + "loss": 0.4054, + "step": 18745 + }, + { + "epoch": 0.8299923624409196, + "grad_norm": 1.3941956758499146, + "learning_rate": 1.7107190932015639e-06, + "loss": 0.4879, + "step": 18746 + }, + { + "epoch": 0.8300366381457335, + "grad_norm": 1.4372999668121338, + "learning_rate": 1.7098547362081197e-06, + "loss": 0.4998, + "step": 18747 + }, + { + "epoch": 0.8300809138505474, + "grad_norm": 1.886871337890625, + "learning_rate": 1.7089905772180437e-06, + "loss": 0.5417, + "step": 18748 + }, + { + "epoch": 0.8301251895553612, + "grad_norm": 1.56510591506958, + "learning_rate": 1.7081266162519783e-06, + "loss": 0.5387, + "step": 18749 + }, + { + "epoch": 0.8301694652601751, + "grad_norm": 1.4944137334823608, + "learning_rate": 1.707262853330558e-06, + "loss": 0.4911, + "step": 18750 + }, + { + "epoch": 0.830213740964989, + "grad_norm": 1.433212161064148, + "learning_rate": 1.7063992884744096e-06, + "loss": 0.4965, + "step": 18751 + }, + { + "epoch": 0.8302580166698028, + "grad_norm": 1.7966707944869995, + "learning_rate": 1.7055359217041656e-06, + "loss": 0.5302, + "step": 18752 + }, + { + "epoch": 0.8303022923746167, + "grad_norm": 1.323806643486023, + "learning_rate": 1.7046727530404406e-06, + "loss": 0.4594, + "step": 18753 + }, + { + "epoch": 0.8303465680794306, + "grad_norm": 1.6710126399993896, + "learning_rate": 1.7038097825038501e-06, + "loss": 0.4909, + "step": 18754 + }, + { + "epoch": 0.8303908437842444, + "grad_norm": 1.5699849128723145, + "learning_rate": 1.7029470101150115e-06, + "loss": 0.4863, + "step": 18755 + }, + { + "epoch": 0.8304351194890583, + "grad_norm": 1.3933579921722412, + "learning_rate": 1.7020844358945265e-06, + "loss": 0.4578, + "step": 18756 + }, + { + "epoch": 0.8304793951938723, + "grad_norm": 1.2477225065231323, + "learning_rate": 1.701222059862998e-06, + "loss": 0.4657, + "step": 18757 + }, + { + "epoch": 0.8305236708986862, + "grad_norm": 1.8481403589248657, + "learning_rate": 1.7003598820410216e-06, + "loss": 0.5168, + "step": 18758 + }, + { + "epoch": 0.8305679466035, + "grad_norm": 1.4436229467391968, + "learning_rate": 1.6994979024491943e-06, + "loss": 0.4884, + "step": 18759 + }, + { + "epoch": 0.8306122223083139, + "grad_norm": 1.821236491203308, + "learning_rate": 1.6986361211081004e-06, + "loss": 0.5872, + "step": 18760 + }, + { + "epoch": 0.8306564980131278, + "grad_norm": 1.4691797494888306, + "learning_rate": 1.6977745380383249e-06, + "loss": 0.4776, + "step": 18761 + }, + { + "epoch": 0.8307007737179416, + "grad_norm": 1.5211567878723145, + "learning_rate": 1.6969131532604433e-06, + "loss": 0.4848, + "step": 18762 + }, + { + "epoch": 0.8307450494227555, + "grad_norm": 1.7204846143722534, + "learning_rate": 1.6960519667950304e-06, + "loss": 0.5632, + "step": 18763 + }, + { + "epoch": 0.8307893251275694, + "grad_norm": 1.6781415939331055, + "learning_rate": 1.6951909786626531e-06, + "loss": 0.4572, + "step": 18764 + }, + { + "epoch": 0.8308336008323832, + "grad_norm": 1.6469013690948486, + "learning_rate": 1.6943301888838804e-06, + "loss": 0.5415, + "step": 18765 + }, + { + "epoch": 0.8308778765371971, + "grad_norm": 1.452294111251831, + "learning_rate": 1.693469597479267e-06, + "loss": 0.4799, + "step": 18766 + }, + { + "epoch": 0.830922152242011, + "grad_norm": 1.433613896369934, + "learning_rate": 1.6926092044693698e-06, + "loss": 0.4829, + "step": 18767 + }, + { + "epoch": 0.8309664279468248, + "grad_norm": 1.9097352027893066, + "learning_rate": 1.6917490098747346e-06, + "loss": 0.5285, + "step": 18768 + }, + { + "epoch": 0.8310107036516388, + "grad_norm": 1.7884562015533447, + "learning_rate": 1.6908890137159117e-06, + "loss": 0.5603, + "step": 18769 + }, + { + "epoch": 0.8310549793564527, + "grad_norm": 1.4309170246124268, + "learning_rate": 1.69002921601344e-06, + "loss": 0.4652, + "step": 18770 + }, + { + "epoch": 0.8310992550612665, + "grad_norm": 1.515458345413208, + "learning_rate": 1.6891696167878535e-06, + "loss": 0.5191, + "step": 18771 + }, + { + "epoch": 0.8311435307660804, + "grad_norm": 1.4670610427856445, + "learning_rate": 1.6883102160596843e-06, + "loss": 0.4641, + "step": 18772 + }, + { + "epoch": 0.8311878064708943, + "grad_norm": 1.4664493799209595, + "learning_rate": 1.6874510138494572e-06, + "loss": 0.4923, + "step": 18773 + }, + { + "epoch": 0.8312320821757081, + "grad_norm": 1.4640867710113525, + "learning_rate": 1.6865920101776922e-06, + "loss": 0.4927, + "step": 18774 + }, + { + "epoch": 0.831276357880522, + "grad_norm": 1.4979861974716187, + "learning_rate": 1.6857332050649112e-06, + "loss": 0.4553, + "step": 18775 + }, + { + "epoch": 0.8313206335853359, + "grad_norm": 1.4906457662582397, + "learning_rate": 1.6848745985316217e-06, + "loss": 0.5019, + "step": 18776 + }, + { + "epoch": 0.8313649092901497, + "grad_norm": 1.5345066785812378, + "learning_rate": 1.6840161905983332e-06, + "loss": 0.463, + "step": 18777 + }, + { + "epoch": 0.8314091849949636, + "grad_norm": 1.4835950136184692, + "learning_rate": 1.6831579812855437e-06, + "loss": 0.4563, + "step": 18778 + }, + { + "epoch": 0.8314534606997775, + "grad_norm": 1.3618272542953491, + "learning_rate": 1.6822999706137565e-06, + "loss": 0.4877, + "step": 18779 + }, + { + "epoch": 0.8314977364045913, + "grad_norm": 1.523408055305481, + "learning_rate": 1.6814421586034624e-06, + "loss": 0.4485, + "step": 18780 + }, + { + "epoch": 0.8315420121094053, + "grad_norm": 1.302109956741333, + "learning_rate": 1.6805845452751491e-06, + "loss": 0.4573, + "step": 18781 + }, + { + "epoch": 0.8315862878142192, + "grad_norm": 1.5022016763687134, + "learning_rate": 1.679727130649299e-06, + "loss": 0.4529, + "step": 18782 + }, + { + "epoch": 0.831630563519033, + "grad_norm": 1.6207842826843262, + "learning_rate": 1.6788699147463928e-06, + "loss": 0.5004, + "step": 18783 + }, + { + "epoch": 0.8316748392238469, + "grad_norm": 1.7145845890045166, + "learning_rate": 1.6780128975869004e-06, + "loss": 0.4683, + "step": 18784 + }, + { + "epoch": 0.8317191149286608, + "grad_norm": 1.326827883720398, + "learning_rate": 1.6771560791912967e-06, + "loss": 0.4771, + "step": 18785 + }, + { + "epoch": 0.8317633906334746, + "grad_norm": 1.419251561164856, + "learning_rate": 1.676299459580043e-06, + "loss": 0.475, + "step": 18786 + }, + { + "epoch": 0.8318076663382885, + "grad_norm": 1.312537431716919, + "learning_rate": 1.6754430387735998e-06, + "loss": 0.4949, + "step": 18787 + }, + { + "epoch": 0.8318519420431024, + "grad_norm": 1.6779959201812744, + "learning_rate": 1.6745868167924184e-06, + "loss": 0.5358, + "step": 18788 + }, + { + "epoch": 0.8318962177479162, + "grad_norm": 1.9751195907592773, + "learning_rate": 1.6737307936569547e-06, + "loss": 0.5359, + "step": 18789 + }, + { + "epoch": 0.8319404934527301, + "grad_norm": 1.366164207458496, + "learning_rate": 1.6728749693876523e-06, + "loss": 0.4708, + "step": 18790 + }, + { + "epoch": 0.831984769157544, + "grad_norm": 1.5801087617874146, + "learning_rate": 1.6720193440049504e-06, + "loss": 0.5172, + "step": 18791 + }, + { + "epoch": 0.8320290448623578, + "grad_norm": 1.2682222127914429, + "learning_rate": 1.671163917529285e-06, + "loss": 0.4169, + "step": 18792 + }, + { + "epoch": 0.8320733205671718, + "grad_norm": 1.5014419555664062, + "learning_rate": 1.6703086899810884e-06, + "loss": 0.4805, + "step": 18793 + }, + { + "epoch": 0.8321175962719857, + "grad_norm": 1.3966323137283325, + "learning_rate": 1.6694536613807843e-06, + "loss": 0.4887, + "step": 18794 + }, + { + "epoch": 0.8321618719767996, + "grad_norm": 1.6722546815872192, + "learning_rate": 1.6685988317487988e-06, + "loss": 0.5187, + "step": 18795 + }, + { + "epoch": 0.8322061476816134, + "grad_norm": 1.368736982345581, + "learning_rate": 1.6677442011055467e-06, + "loss": 0.488, + "step": 18796 + }, + { + "epoch": 0.8322504233864273, + "grad_norm": 1.767032504081726, + "learning_rate": 1.6668897694714392e-06, + "loss": 0.489, + "step": 18797 + }, + { + "epoch": 0.8322946990912412, + "grad_norm": 1.5950146913528442, + "learning_rate": 1.6660355368668823e-06, + "loss": 0.5437, + "step": 18798 + }, + { + "epoch": 0.832338974796055, + "grad_norm": 1.4960955381393433, + "learning_rate": 1.6651815033122842e-06, + "loss": 0.4793, + "step": 18799 + }, + { + "epoch": 0.8323832505008689, + "grad_norm": 1.4974002838134766, + "learning_rate": 1.6643276688280407e-06, + "loss": 0.5148, + "step": 18800 + }, + { + "epoch": 0.8324275262056828, + "grad_norm": 2.383683681488037, + "learning_rate": 1.6634740334345389e-06, + "loss": 0.5936, + "step": 18801 + }, + { + "epoch": 0.8324718019104966, + "grad_norm": 1.431023359298706, + "learning_rate": 1.6626205971521759e-06, + "loss": 0.4683, + "step": 18802 + }, + { + "epoch": 0.8325160776153105, + "grad_norm": 1.6981391906738281, + "learning_rate": 1.6617673600013295e-06, + "loss": 0.4662, + "step": 18803 + }, + { + "epoch": 0.8325603533201245, + "grad_norm": 1.469728708267212, + "learning_rate": 1.6609143220023794e-06, + "loss": 0.4664, + "step": 18804 + }, + { + "epoch": 0.8326046290249383, + "grad_norm": 1.374458909034729, + "learning_rate": 1.6600614831757023e-06, + "loss": 0.5298, + "step": 18805 + }, + { + "epoch": 0.8326489047297522, + "grad_norm": 1.4292373657226562, + "learning_rate": 1.6592088435416665e-06, + "loss": 0.4434, + "step": 18806 + }, + { + "epoch": 0.8326931804345661, + "grad_norm": 1.793175220489502, + "learning_rate": 1.6583564031206357e-06, + "loss": 0.5321, + "step": 18807 + }, + { + "epoch": 0.8327374561393799, + "grad_norm": 1.6389272212982178, + "learning_rate": 1.6575041619329701e-06, + "loss": 0.5312, + "step": 18808 + }, + { + "epoch": 0.8327817318441938, + "grad_norm": 1.6172541379928589, + "learning_rate": 1.656652119999026e-06, + "loss": 0.5154, + "step": 18809 + }, + { + "epoch": 0.8328260075490077, + "grad_norm": 1.499941349029541, + "learning_rate": 1.6558002773391513e-06, + "loss": 0.4967, + "step": 18810 + }, + { + "epoch": 0.8328702832538215, + "grad_norm": 1.420058250427246, + "learning_rate": 1.654948633973692e-06, + "loss": 0.4655, + "step": 18811 + }, + { + "epoch": 0.8329145589586354, + "grad_norm": 1.4182230234146118, + "learning_rate": 1.6540971899229907e-06, + "loss": 0.4551, + "step": 18812 + }, + { + "epoch": 0.8329588346634493, + "grad_norm": 1.542432188987732, + "learning_rate": 1.6532459452073835e-06, + "loss": 0.4499, + "step": 18813 + }, + { + "epoch": 0.8330031103682631, + "grad_norm": 1.6469393968582153, + "learning_rate": 1.6523948998471973e-06, + "loss": 0.4814, + "step": 18814 + }, + { + "epoch": 0.833047386073077, + "grad_norm": 1.367811918258667, + "learning_rate": 1.6515440538627658e-06, + "loss": 0.4749, + "step": 18815 + }, + { + "epoch": 0.833091661777891, + "grad_norm": 1.4135143756866455, + "learning_rate": 1.6506934072744063e-06, + "loss": 0.4796, + "step": 18816 + }, + { + "epoch": 0.8331359374827048, + "grad_norm": 1.6835006475448608, + "learning_rate": 1.6498429601024358e-06, + "loss": 0.5801, + "step": 18817 + }, + { + "epoch": 0.8331802131875187, + "grad_norm": 2.8103222846984863, + "learning_rate": 1.6489927123671679e-06, + "loss": 0.5695, + "step": 18818 + }, + { + "epoch": 0.8332244888923326, + "grad_norm": 1.4389652013778687, + "learning_rate": 1.6481426640889098e-06, + "loss": 0.545, + "step": 18819 + }, + { + "epoch": 0.8332687645971464, + "grad_norm": 1.4775248765945435, + "learning_rate": 1.647292815287963e-06, + "loss": 0.4759, + "step": 18820 + }, + { + "epoch": 0.8333130403019603, + "grad_norm": 2.2104573249816895, + "learning_rate": 1.6464431659846248e-06, + "loss": 0.531, + "step": 18821 + }, + { + "epoch": 0.8333573160067742, + "grad_norm": 1.8882009983062744, + "learning_rate": 1.645593716199192e-06, + "loss": 0.5469, + "step": 18822 + }, + { + "epoch": 0.833401591711588, + "grad_norm": 1.5081738233566284, + "learning_rate": 1.6447444659519506e-06, + "loss": 0.502, + "step": 18823 + }, + { + "epoch": 0.8334458674164019, + "grad_norm": 1.4805409908294678, + "learning_rate": 1.6438954152631825e-06, + "loss": 0.4619, + "step": 18824 + }, + { + "epoch": 0.8334901431212158, + "grad_norm": 1.6280823945999146, + "learning_rate": 1.6430465641531701e-06, + "loss": 0.4937, + "step": 18825 + }, + { + "epoch": 0.8335344188260296, + "grad_norm": 1.2943437099456787, + "learning_rate": 1.642197912642187e-06, + "loss": 0.4325, + "step": 18826 + }, + { + "epoch": 0.8335786945308435, + "grad_norm": 1.5228636264801025, + "learning_rate": 1.641349460750501e-06, + "loss": 0.5041, + "step": 18827 + }, + { + "epoch": 0.8336229702356575, + "grad_norm": 1.807681918144226, + "learning_rate": 1.6405012084983774e-06, + "loss": 0.5569, + "step": 18828 + }, + { + "epoch": 0.8336672459404714, + "grad_norm": 1.4585926532745361, + "learning_rate": 1.6396531559060758e-06, + "loss": 0.4468, + "step": 18829 + }, + { + "epoch": 0.8337115216452852, + "grad_norm": 1.3620754480361938, + "learning_rate": 1.6388053029938522e-06, + "loss": 0.4691, + "step": 18830 + }, + { + "epoch": 0.8337557973500991, + "grad_norm": 1.9580379724502563, + "learning_rate": 1.6379576497819527e-06, + "loss": 0.5477, + "step": 18831 + }, + { + "epoch": 0.833800073054913, + "grad_norm": 1.6300830841064453, + "learning_rate": 1.6371101962906288e-06, + "loss": 0.5049, + "step": 18832 + }, + { + "epoch": 0.8338443487597268, + "grad_norm": 1.4333302974700928, + "learning_rate": 1.6362629425401188e-06, + "loss": 0.5039, + "step": 18833 + }, + { + "epoch": 0.8338886244645407, + "grad_norm": 1.681723713874817, + "learning_rate": 1.6354158885506577e-06, + "loss": 0.4796, + "step": 18834 + }, + { + "epoch": 0.8339329001693546, + "grad_norm": 1.4177560806274414, + "learning_rate": 1.6345690343424758e-06, + "loss": 0.4662, + "step": 18835 + }, + { + "epoch": 0.8339771758741684, + "grad_norm": 1.684156894683838, + "learning_rate": 1.6337223799358025e-06, + "loss": 0.4916, + "step": 18836 + }, + { + "epoch": 0.8340214515789823, + "grad_norm": 1.4800379276275635, + "learning_rate": 1.6328759253508585e-06, + "loss": 0.5922, + "step": 18837 + }, + { + "epoch": 0.8340657272837962, + "grad_norm": 1.8658020496368408, + "learning_rate": 1.632029670607861e-06, + "loss": 0.5203, + "step": 18838 + }, + { + "epoch": 0.83411000298861, + "grad_norm": 2.0710201263427734, + "learning_rate": 1.6311836157270212e-06, + "loss": 0.5622, + "step": 18839 + }, + { + "epoch": 0.834154278693424, + "grad_norm": 1.484447956085205, + "learning_rate": 1.6303377607285454e-06, + "loss": 0.4719, + "step": 18840 + }, + { + "epoch": 0.8341985543982379, + "grad_norm": 1.3758268356323242, + "learning_rate": 1.6294921056326363e-06, + "loss": 0.4303, + "step": 18841 + }, + { + "epoch": 0.8342428301030517, + "grad_norm": 1.6493197679519653, + "learning_rate": 1.6286466504594944e-06, + "loss": 0.5272, + "step": 18842 + }, + { + "epoch": 0.8342871058078656, + "grad_norm": 1.3681309223175049, + "learning_rate": 1.6278013952293115e-06, + "loss": 0.4763, + "step": 18843 + }, + { + "epoch": 0.8343313815126795, + "grad_norm": 1.5450273752212524, + "learning_rate": 1.6269563399622745e-06, + "loss": 0.51, + "step": 18844 + }, + { + "epoch": 0.8343756572174933, + "grad_norm": 1.3607486486434937, + "learning_rate": 1.6261114846785665e-06, + "loss": 0.4401, + "step": 18845 + }, + { + "epoch": 0.8344199329223072, + "grad_norm": 1.5133538246154785, + "learning_rate": 1.6252668293983687e-06, + "loss": 0.522, + "step": 18846 + }, + { + "epoch": 0.8344642086271211, + "grad_norm": 1.841907024383545, + "learning_rate": 1.6244223741418574e-06, + "loss": 0.5292, + "step": 18847 + }, + { + "epoch": 0.8345084843319349, + "grad_norm": 1.3277807235717773, + "learning_rate": 1.6235781189291922e-06, + "loss": 0.4417, + "step": 18848 + }, + { + "epoch": 0.8345527600367488, + "grad_norm": 1.26876962184906, + "learning_rate": 1.6227340637805455e-06, + "loss": 0.4648, + "step": 18849 + }, + { + "epoch": 0.8345970357415627, + "grad_norm": 1.3667677640914917, + "learning_rate": 1.6218902087160759e-06, + "loss": 0.4997, + "step": 18850 + }, + { + "epoch": 0.8346413114463765, + "grad_norm": 1.607878565788269, + "learning_rate": 1.621046553755934e-06, + "loss": 0.5, + "step": 18851 + }, + { + "epoch": 0.8346855871511905, + "grad_norm": 1.392447829246521, + "learning_rate": 1.6202030989202744e-06, + "loss": 0.4273, + "step": 18852 + }, + { + "epoch": 0.8347298628560044, + "grad_norm": 1.2686079740524292, + "learning_rate": 1.6193598442292413e-06, + "loss": 0.4139, + "step": 18853 + }, + { + "epoch": 0.8347741385608182, + "grad_norm": 1.5122381448745728, + "learning_rate": 1.6185167897029752e-06, + "loss": 0.4807, + "step": 18854 + }, + { + "epoch": 0.8348184142656321, + "grad_norm": 1.5891284942626953, + "learning_rate": 1.6176739353616077e-06, + "loss": 0.4777, + "step": 18855 + }, + { + "epoch": 0.834862689970446, + "grad_norm": 2.065556526184082, + "learning_rate": 1.6168312812252784e-06, + "loss": 0.4953, + "step": 18856 + }, + { + "epoch": 0.8349069656752598, + "grad_norm": 1.6233183145523071, + "learning_rate": 1.6159888273141055e-06, + "loss": 0.5086, + "step": 18857 + }, + { + "epoch": 0.8349512413800737, + "grad_norm": 1.4999310970306396, + "learning_rate": 1.6151465736482109e-06, + "loss": 0.4896, + "step": 18858 + }, + { + "epoch": 0.8349955170848876, + "grad_norm": 1.2176121473312378, + "learning_rate": 1.6143045202477148e-06, + "loss": 0.4365, + "step": 18859 + }, + { + "epoch": 0.8350397927897014, + "grad_norm": 1.703330159187317, + "learning_rate": 1.613462667132728e-06, + "loss": 0.578, + "step": 18860 + }, + { + "epoch": 0.8350840684945153, + "grad_norm": 1.5262787342071533, + "learning_rate": 1.6126210143233544e-06, + "loss": 0.4647, + "step": 18861 + }, + { + "epoch": 0.8351283441993292, + "grad_norm": 1.5834835767745972, + "learning_rate": 1.6117795618397015e-06, + "loss": 0.5586, + "step": 18862 + }, + { + "epoch": 0.8351726199041432, + "grad_norm": 1.6620540618896484, + "learning_rate": 1.6109383097018628e-06, + "loss": 0.5484, + "step": 18863 + }, + { + "epoch": 0.835216895608957, + "grad_norm": 1.7344392538070679, + "learning_rate": 1.6100972579299322e-06, + "loss": 0.5291, + "step": 18864 + }, + { + "epoch": 0.8352611713137709, + "grad_norm": 1.268068552017212, + "learning_rate": 1.6092564065439974e-06, + "loss": 0.4274, + "step": 18865 + }, + { + "epoch": 0.8353054470185848, + "grad_norm": 1.619423270225525, + "learning_rate": 1.608415755564141e-06, + "loss": 0.5384, + "step": 18866 + }, + { + "epoch": 0.8353497227233986, + "grad_norm": 1.4949077367782593, + "learning_rate": 1.6075753050104426e-06, + "loss": 0.4961, + "step": 18867 + }, + { + "epoch": 0.8353939984282125, + "grad_norm": 1.3546853065490723, + "learning_rate": 1.6067350549029715e-06, + "loss": 0.4849, + "step": 18868 + }, + { + "epoch": 0.8354382741330264, + "grad_norm": 1.7211624383926392, + "learning_rate": 1.6058950052618028e-06, + "loss": 0.5049, + "step": 18869 + }, + { + "epoch": 0.8354825498378402, + "grad_norm": 1.576072096824646, + "learning_rate": 1.605055156106996e-06, + "loss": 0.4593, + "step": 18870 + }, + { + "epoch": 0.8355268255426541, + "grad_norm": 1.3896863460540771, + "learning_rate": 1.6042155074586097e-06, + "loss": 0.5114, + "step": 18871 + }, + { + "epoch": 0.835571101247468, + "grad_norm": 1.7265416383743286, + "learning_rate": 1.603376059336702e-06, + "loss": 0.4725, + "step": 18872 + }, + { + "epoch": 0.8356153769522818, + "grad_norm": 1.762129783630371, + "learning_rate": 1.6025368117613205e-06, + "loss": 0.4551, + "step": 18873 + }, + { + "epoch": 0.8356596526570957, + "grad_norm": 1.8996891975402832, + "learning_rate": 1.601697764752509e-06, + "loss": 0.5152, + "step": 18874 + }, + { + "epoch": 0.8357039283619097, + "grad_norm": 1.499048113822937, + "learning_rate": 1.6008589183303092e-06, + "loss": 0.4953, + "step": 18875 + }, + { + "epoch": 0.8357482040667235, + "grad_norm": 1.6365063190460205, + "learning_rate": 1.600020272514755e-06, + "loss": 0.5058, + "step": 18876 + }, + { + "epoch": 0.8357924797715374, + "grad_norm": 1.737436056137085, + "learning_rate": 1.5991818273258774e-06, + "loss": 0.5158, + "step": 18877 + }, + { + "epoch": 0.8358367554763513, + "grad_norm": 1.5757319927215576, + "learning_rate": 1.5983435827836992e-06, + "loss": 0.5371, + "step": 18878 + }, + { + "epoch": 0.8358810311811651, + "grad_norm": 1.3900518417358398, + "learning_rate": 1.597505538908245e-06, + "loss": 0.4327, + "step": 18879 + }, + { + "epoch": 0.835925306885979, + "grad_norm": 1.6117944717407227, + "learning_rate": 1.5966676957195305e-06, + "loss": 0.5453, + "step": 18880 + }, + { + "epoch": 0.8359695825907929, + "grad_norm": 1.479909896850586, + "learning_rate": 1.5958300532375636e-06, + "loss": 0.4876, + "step": 18881 + }, + { + "epoch": 0.8360138582956067, + "grad_norm": 1.7666047811508179, + "learning_rate": 1.5949926114823543e-06, + "loss": 0.5264, + "step": 18882 + }, + { + "epoch": 0.8360581340004206, + "grad_norm": 1.4173170328140259, + "learning_rate": 1.5941553704739044e-06, + "loss": 0.4613, + "step": 18883 + }, + { + "epoch": 0.8361024097052345, + "grad_norm": 1.398262619972229, + "learning_rate": 1.5933183302322076e-06, + "loss": 0.4612, + "step": 18884 + }, + { + "epoch": 0.8361466854100483, + "grad_norm": 1.460156798362732, + "learning_rate": 1.592481490777259e-06, + "loss": 0.4767, + "step": 18885 + }, + { + "epoch": 0.8361909611148622, + "grad_norm": 1.5019927024841309, + "learning_rate": 1.591644852129043e-06, + "loss": 0.5234, + "step": 18886 + }, + { + "epoch": 0.8362352368196762, + "grad_norm": 1.7615495920181274, + "learning_rate": 1.5908084143075442e-06, + "loss": 0.5459, + "step": 18887 + }, + { + "epoch": 0.83627951252449, + "grad_norm": 1.6717735528945923, + "learning_rate": 1.589972177332737e-06, + "loss": 0.5584, + "step": 18888 + }, + { + "epoch": 0.8363237882293039, + "grad_norm": 1.39620840549469, + "learning_rate": 1.5891361412245987e-06, + "loss": 0.4458, + "step": 18889 + }, + { + "epoch": 0.8363680639341178, + "grad_norm": 1.3997310400009155, + "learning_rate": 1.5883003060030944e-06, + "loss": 0.4808, + "step": 18890 + }, + { + "epoch": 0.8364123396389316, + "grad_norm": 1.3297383785247803, + "learning_rate": 1.587464671688187e-06, + "loss": 0.445, + "step": 18891 + }, + { + "epoch": 0.8364566153437455, + "grad_norm": 1.7971194982528687, + "learning_rate": 1.5866292382998382e-06, + "loss": 0.5027, + "step": 18892 + }, + { + "epoch": 0.8365008910485594, + "grad_norm": 1.4319283962249756, + "learning_rate": 1.5857940058580002e-06, + "loss": 0.4446, + "step": 18893 + }, + { + "epoch": 0.8365451667533732, + "grad_norm": 1.797061562538147, + "learning_rate": 1.584958974382621e-06, + "loss": 0.5271, + "step": 18894 + }, + { + "epoch": 0.8365894424581871, + "grad_norm": 1.310348629951477, + "learning_rate": 1.584124143893645e-06, + "loss": 0.4401, + "step": 18895 + }, + { + "epoch": 0.836633718163001, + "grad_norm": 1.5723098516464233, + "learning_rate": 1.5832895144110105e-06, + "loss": 0.5007, + "step": 18896 + }, + { + "epoch": 0.8366779938678148, + "grad_norm": 1.5152621269226074, + "learning_rate": 1.5824550859546539e-06, + "loss": 0.4989, + "step": 18897 + }, + { + "epoch": 0.8367222695726287, + "grad_norm": 1.61695396900177, + "learning_rate": 1.5816208585445004e-06, + "loss": 0.4681, + "step": 18898 + }, + { + "epoch": 0.8367665452774427, + "grad_norm": 1.6422055959701538, + "learning_rate": 1.5807868322004815e-06, + "loss": 0.5363, + "step": 18899 + }, + { + "epoch": 0.8368108209822566, + "grad_norm": 1.5266573429107666, + "learning_rate": 1.579953006942513e-06, + "loss": 0.5091, + "step": 18900 + }, + { + "epoch": 0.8368550966870704, + "grad_norm": 1.4695292711257935, + "learning_rate": 1.5791193827905094e-06, + "loss": 0.4219, + "step": 18901 + }, + { + "epoch": 0.8368993723918843, + "grad_norm": 1.6505992412567139, + "learning_rate": 1.5782859597643852e-06, + "loss": 0.5987, + "step": 18902 + }, + { + "epoch": 0.8369436480966982, + "grad_norm": 1.5613375902175903, + "learning_rate": 1.5774527378840454e-06, + "loss": 0.5246, + "step": 18903 + }, + { + "epoch": 0.836987923801512, + "grad_norm": 1.7339165210723877, + "learning_rate": 1.5766197171693874e-06, + "loss": 0.5836, + "step": 18904 + }, + { + "epoch": 0.8370321995063259, + "grad_norm": 1.465951919555664, + "learning_rate": 1.5757868976403057e-06, + "loss": 0.4891, + "step": 18905 + }, + { + "epoch": 0.8370764752111398, + "grad_norm": 1.4996687173843384, + "learning_rate": 1.5749542793166982e-06, + "loss": 0.4321, + "step": 18906 + }, + { + "epoch": 0.8371207509159536, + "grad_norm": 1.587707281112671, + "learning_rate": 1.5741218622184473e-06, + "loss": 0.5304, + "step": 18907 + }, + { + "epoch": 0.8371650266207675, + "grad_norm": 1.5210959911346436, + "learning_rate": 1.5732896463654324e-06, + "loss": 0.51, + "step": 18908 + }, + { + "epoch": 0.8372093023255814, + "grad_norm": 1.5490705966949463, + "learning_rate": 1.5724576317775353e-06, + "loss": 0.4791, + "step": 18909 + }, + { + "epoch": 0.8372535780303952, + "grad_norm": 1.7257819175720215, + "learning_rate": 1.5716258184746258e-06, + "loss": 0.5166, + "step": 18910 + }, + { + "epoch": 0.8372978537352092, + "grad_norm": 1.461440086364746, + "learning_rate": 1.5707942064765691e-06, + "loss": 0.4363, + "step": 18911 + }, + { + "epoch": 0.8373421294400231, + "grad_norm": 1.8301345109939575, + "learning_rate": 1.5699627958032338e-06, + "loss": 0.5518, + "step": 18912 + }, + { + "epoch": 0.8373864051448369, + "grad_norm": 1.4224703311920166, + "learning_rate": 1.5691315864744716e-06, + "loss": 0.4584, + "step": 18913 + }, + { + "epoch": 0.8374306808496508, + "grad_norm": 1.538671612739563, + "learning_rate": 1.5683005785101358e-06, + "loss": 0.5371, + "step": 18914 + }, + { + "epoch": 0.8374749565544647, + "grad_norm": 1.6473937034606934, + "learning_rate": 1.5674697719300735e-06, + "loss": 0.4887, + "step": 18915 + }, + { + "epoch": 0.8375192322592785, + "grad_norm": 1.4316937923431396, + "learning_rate": 1.5666391667541314e-06, + "loss": 0.5034, + "step": 18916 + }, + { + "epoch": 0.8375635079640924, + "grad_norm": 1.4103301763534546, + "learning_rate": 1.5658087630021468e-06, + "loss": 0.4988, + "step": 18917 + }, + { + "epoch": 0.8376077836689063, + "grad_norm": 1.4449471235275269, + "learning_rate": 1.5649785606939493e-06, + "loss": 0.5498, + "step": 18918 + }, + { + "epoch": 0.8376520593737201, + "grad_norm": 1.6179953813552856, + "learning_rate": 1.5641485598493744e-06, + "loss": 0.4681, + "step": 18919 + }, + { + "epoch": 0.837696335078534, + "grad_norm": 1.7563420534133911, + "learning_rate": 1.5633187604882416e-06, + "loss": 0.5583, + "step": 18920 + }, + { + "epoch": 0.8377406107833479, + "grad_norm": 1.6221213340759277, + "learning_rate": 1.5624891626303707e-06, + "loss": 0.5281, + "step": 18921 + }, + { + "epoch": 0.8377848864881617, + "grad_norm": 1.3229167461395264, + "learning_rate": 1.5616597662955768e-06, + "loss": 0.4653, + "step": 18922 + }, + { + "epoch": 0.8378291621929757, + "grad_norm": 1.2651808261871338, + "learning_rate": 1.5608305715036676e-06, + "loss": 0.4443, + "step": 18923 + }, + { + "epoch": 0.8378734378977896, + "grad_norm": 1.7693121433258057, + "learning_rate": 1.5600015782744494e-06, + "loss": 0.5637, + "step": 18924 + }, + { + "epoch": 0.8379177136026034, + "grad_norm": 1.8904279470443726, + "learning_rate": 1.5591727866277196e-06, + "loss": 0.5364, + "step": 18925 + }, + { + "epoch": 0.8379619893074173, + "grad_norm": 1.572741985321045, + "learning_rate": 1.5583441965832768e-06, + "loss": 0.5534, + "step": 18926 + }, + { + "epoch": 0.8380062650122312, + "grad_norm": 1.5850638151168823, + "learning_rate": 1.5575158081609088e-06, + "loss": 0.522, + "step": 18927 + }, + { + "epoch": 0.838050540717045, + "grad_norm": 1.443081021308899, + "learning_rate": 1.5566876213804005e-06, + "loss": 0.4667, + "step": 18928 + }, + { + "epoch": 0.8380948164218589, + "grad_norm": 1.3793299198150635, + "learning_rate": 1.5558596362615353e-06, + "loss": 0.5002, + "step": 18929 + }, + { + "epoch": 0.8381390921266728, + "grad_norm": 1.3274853229522705, + "learning_rate": 1.5550318528240871e-06, + "loss": 0.4377, + "step": 18930 + }, + { + "epoch": 0.8381833678314866, + "grad_norm": 1.3592305183410645, + "learning_rate": 1.554204271087827e-06, + "loss": 0.4472, + "step": 18931 + }, + { + "epoch": 0.8382276435363005, + "grad_norm": 1.5130149126052856, + "learning_rate": 1.553376891072521e-06, + "loss": 0.4656, + "step": 18932 + }, + { + "epoch": 0.8382719192411144, + "grad_norm": 1.5726988315582275, + "learning_rate": 1.5525497127979305e-06, + "loss": 0.5088, + "step": 18933 + }, + { + "epoch": 0.8383161949459284, + "grad_norm": 1.5746573209762573, + "learning_rate": 1.5517227362838117e-06, + "loss": 0.4889, + "step": 18934 + }, + { + "epoch": 0.8383604706507422, + "grad_norm": 1.1753662824630737, + "learning_rate": 1.5508959615499142e-06, + "loss": 0.4049, + "step": 18935 + }, + { + "epoch": 0.8384047463555561, + "grad_norm": 1.4221925735473633, + "learning_rate": 1.55006938861599e-06, + "loss": 0.4824, + "step": 18936 + }, + { + "epoch": 0.83844902206037, + "grad_norm": 2.2898876667022705, + "learning_rate": 1.549243017501778e-06, + "loss": 0.5816, + "step": 18937 + }, + { + "epoch": 0.8384932977651838, + "grad_norm": 1.5303878784179688, + "learning_rate": 1.5484168482270135e-06, + "loss": 0.5486, + "step": 18938 + }, + { + "epoch": 0.8385375734699977, + "grad_norm": 1.3215535879135132, + "learning_rate": 1.5475908808114325e-06, + "loss": 0.4762, + "step": 18939 + }, + { + "epoch": 0.8385818491748116, + "grad_norm": 1.5069446563720703, + "learning_rate": 1.546765115274762e-06, + "loss": 0.4917, + "step": 18940 + }, + { + "epoch": 0.8386261248796254, + "grad_norm": 1.354793906211853, + "learning_rate": 1.5459395516367237e-06, + "loss": 0.4686, + "step": 18941 + }, + { + "epoch": 0.8386704005844393, + "grad_norm": 1.2513009309768677, + "learning_rate": 1.5451141899170353e-06, + "loss": 0.4236, + "step": 18942 + }, + { + "epoch": 0.8387146762892532, + "grad_norm": 1.858567476272583, + "learning_rate": 1.5442890301354107e-06, + "loss": 0.532, + "step": 18943 + }, + { + "epoch": 0.838758951994067, + "grad_norm": 1.8872681856155396, + "learning_rate": 1.543464072311558e-06, + "loss": 0.5813, + "step": 18944 + }, + { + "epoch": 0.8388032276988809, + "grad_norm": 1.6160473823547363, + "learning_rate": 1.5426393164651775e-06, + "loss": 0.4812, + "step": 18945 + }, + { + "epoch": 0.8388475034036948, + "grad_norm": 1.6527931690216064, + "learning_rate": 1.541814762615973e-06, + "loss": 0.5347, + "step": 18946 + }, + { + "epoch": 0.8388917791085087, + "grad_norm": 1.9190673828125, + "learning_rate": 1.540990410783636e-06, + "loss": 0.5035, + "step": 18947 + }, + { + "epoch": 0.8389360548133226, + "grad_norm": 1.499205470085144, + "learning_rate": 1.5401662609878531e-06, + "loss": 0.49, + "step": 18948 + }, + { + "epoch": 0.8389803305181365, + "grad_norm": 1.2815788984298706, + "learning_rate": 1.539342313248313e-06, + "loss": 0.4666, + "step": 18949 + }, + { + "epoch": 0.8390246062229503, + "grad_norm": 1.1872599124908447, + "learning_rate": 1.538518567584696e-06, + "loss": 0.4332, + "step": 18950 + }, + { + "epoch": 0.8390688819277642, + "grad_norm": 1.4271256923675537, + "learning_rate": 1.5376950240166699e-06, + "loss": 0.4478, + "step": 18951 + }, + { + "epoch": 0.8391131576325781, + "grad_norm": 1.3369022607803345, + "learning_rate": 1.5368716825639063e-06, + "loss": 0.4534, + "step": 18952 + }, + { + "epoch": 0.8391574333373919, + "grad_norm": 1.237188696861267, + "learning_rate": 1.5360485432460737e-06, + "loss": 0.4322, + "step": 18953 + }, + { + "epoch": 0.8392017090422058, + "grad_norm": 1.5483477115631104, + "learning_rate": 1.5352256060828308e-06, + "loss": 0.5127, + "step": 18954 + }, + { + "epoch": 0.8392459847470197, + "grad_norm": 1.5969563722610474, + "learning_rate": 1.5344028710938286e-06, + "loss": 0.5211, + "step": 18955 + }, + { + "epoch": 0.8392902604518335, + "grad_norm": 1.7064640522003174, + "learning_rate": 1.5335803382987236e-06, + "loss": 0.546, + "step": 18956 + }, + { + "epoch": 0.8393345361566474, + "grad_norm": 1.8434531688690186, + "learning_rate": 1.5327580077171589e-06, + "loss": 0.4898, + "step": 18957 + }, + { + "epoch": 0.8393788118614613, + "grad_norm": 1.9063410758972168, + "learning_rate": 1.5319358793687722e-06, + "loss": 0.5407, + "step": 18958 + }, + { + "epoch": 0.8394230875662752, + "grad_norm": 1.5664622783660889, + "learning_rate": 1.5311139532732066e-06, + "loss": 0.5296, + "step": 18959 + }, + { + "epoch": 0.8394673632710891, + "grad_norm": 1.4794524908065796, + "learning_rate": 1.5302922294500867e-06, + "loss": 0.5137, + "step": 18960 + }, + { + "epoch": 0.839511638975903, + "grad_norm": 1.4965670108795166, + "learning_rate": 1.5294707079190408e-06, + "loss": 0.4829, + "step": 18961 + }, + { + "epoch": 0.8395559146807168, + "grad_norm": 1.3433490991592407, + "learning_rate": 1.5286493886996878e-06, + "loss": 0.432, + "step": 18962 + }, + { + "epoch": 0.8396001903855307, + "grad_norm": 1.5837647914886475, + "learning_rate": 1.5278282718116477e-06, + "loss": 0.4629, + "step": 18963 + }, + { + "epoch": 0.8396444660903446, + "grad_norm": 1.5453505516052246, + "learning_rate": 1.5270073572745325e-06, + "loss": 0.491, + "step": 18964 + }, + { + "epoch": 0.8396887417951584, + "grad_norm": 1.2606226205825806, + "learning_rate": 1.5261866451079444e-06, + "loss": 0.4206, + "step": 18965 + }, + { + "epoch": 0.8397330174999723, + "grad_norm": 1.1775823831558228, + "learning_rate": 1.5253661353314907e-06, + "loss": 0.4116, + "step": 18966 + }, + { + "epoch": 0.8397772932047862, + "grad_norm": 1.5111292600631714, + "learning_rate": 1.5245458279647674e-06, + "loss": 0.5096, + "step": 18967 + }, + { + "epoch": 0.8398215689096, + "grad_norm": 1.5293056964874268, + "learning_rate": 1.523725723027365e-06, + "loss": 0.4344, + "step": 18968 + }, + { + "epoch": 0.8398658446144139, + "grad_norm": 1.528312087059021, + "learning_rate": 1.5229058205388724e-06, + "loss": 0.553, + "step": 18969 + }, + { + "epoch": 0.8399101203192278, + "grad_norm": 1.3532620668411255, + "learning_rate": 1.522086120518873e-06, + "loss": 0.438, + "step": 18970 + }, + { + "epoch": 0.8399543960240418, + "grad_norm": 1.4532396793365479, + "learning_rate": 1.5212666229869422e-06, + "loss": 0.4731, + "step": 18971 + }, + { + "epoch": 0.8399986717288556, + "grad_norm": 1.5865049362182617, + "learning_rate": 1.520447327962653e-06, + "loss": 0.5405, + "step": 18972 + }, + { + "epoch": 0.8400429474336695, + "grad_norm": 1.2869048118591309, + "learning_rate": 1.5196282354655779e-06, + "loss": 0.4436, + "step": 18973 + }, + { + "epoch": 0.8400872231384834, + "grad_norm": 1.7938997745513916, + "learning_rate": 1.5188093455152763e-06, + "loss": 0.5171, + "step": 18974 + }, + { + "epoch": 0.8401314988432972, + "grad_norm": 1.2488157749176025, + "learning_rate": 1.5179906581313063e-06, + "loss": 0.3987, + "step": 18975 + }, + { + "epoch": 0.8401757745481111, + "grad_norm": 1.732160210609436, + "learning_rate": 1.5171721733332245e-06, + "loss": 0.5182, + "step": 18976 + }, + { + "epoch": 0.840220050252925, + "grad_norm": 1.5928616523742676, + "learning_rate": 1.5163538911405794e-06, + "loss": 0.5062, + "step": 18977 + }, + { + "epoch": 0.8402643259577388, + "grad_norm": 1.5068615674972534, + "learning_rate": 1.5155358115729135e-06, + "loss": 0.4566, + "step": 18978 + }, + { + "epoch": 0.8403086016625527, + "grad_norm": 1.3372210264205933, + "learning_rate": 1.5147179346497665e-06, + "loss": 0.4472, + "step": 18979 + }, + { + "epoch": 0.8403528773673666, + "grad_norm": 1.786148190498352, + "learning_rate": 1.5139002603906727e-06, + "loss": 0.5163, + "step": 18980 + }, + { + "epoch": 0.8403971530721804, + "grad_norm": 1.6199941635131836, + "learning_rate": 1.5130827888151612e-06, + "loss": 0.4631, + "step": 18981 + }, + { + "epoch": 0.8404414287769943, + "grad_norm": 1.6384377479553223, + "learning_rate": 1.512265519942755e-06, + "loss": 0.5157, + "step": 18982 + }, + { + "epoch": 0.8404857044818083, + "grad_norm": 1.3450769186019897, + "learning_rate": 1.5114484537929786e-06, + "loss": 0.4338, + "step": 18983 + }, + { + "epoch": 0.8405299801866221, + "grad_norm": 1.5775458812713623, + "learning_rate": 1.510631590385343e-06, + "loss": 0.4579, + "step": 18984 + }, + { + "epoch": 0.840574255891436, + "grad_norm": 1.5075721740722656, + "learning_rate": 1.5098149297393583e-06, + "loss": 0.4778, + "step": 18985 + }, + { + "epoch": 0.8406185315962499, + "grad_norm": 1.4549963474273682, + "learning_rate": 1.5089984718745333e-06, + "loss": 0.4688, + "step": 18986 + }, + { + "epoch": 0.8406628073010637, + "grad_norm": 1.5156981945037842, + "learning_rate": 1.5081822168103654e-06, + "loss": 0.4967, + "step": 18987 + }, + { + "epoch": 0.8407070830058776, + "grad_norm": 1.313028335571289, + "learning_rate": 1.5073661645663517e-06, + "loss": 0.4335, + "step": 18988 + }, + { + "epoch": 0.8407513587106915, + "grad_norm": 1.5660148859024048, + "learning_rate": 1.5065503151619821e-06, + "loss": 0.4615, + "step": 18989 + }, + { + "epoch": 0.8407956344155053, + "grad_norm": 1.7940608263015747, + "learning_rate": 1.505734668616743e-06, + "loss": 0.5209, + "step": 18990 + }, + { + "epoch": 0.8408399101203192, + "grad_norm": 1.5590530633926392, + "learning_rate": 1.5049192249501144e-06, + "loss": 0.5288, + "step": 18991 + }, + { + "epoch": 0.8408841858251331, + "grad_norm": 1.5064723491668701, + "learning_rate": 1.5041039841815708e-06, + "loss": 0.4854, + "step": 18992 + }, + { + "epoch": 0.8409284615299469, + "grad_norm": 1.7679835557937622, + "learning_rate": 1.5032889463305879e-06, + "loss": 0.5759, + "step": 18993 + }, + { + "epoch": 0.8409727372347608, + "grad_norm": 1.4362781047821045, + "learning_rate": 1.5024741114166307e-06, + "loss": 0.4848, + "step": 18994 + }, + { + "epoch": 0.8410170129395748, + "grad_norm": 1.2894037961959839, + "learning_rate": 1.5016594794591588e-06, + "loss": 0.4605, + "step": 18995 + }, + { + "epoch": 0.8410612886443886, + "grad_norm": 1.4241784811019897, + "learning_rate": 1.5008450504776318e-06, + "loss": 0.5011, + "step": 18996 + }, + { + "epoch": 0.8411055643492025, + "grad_norm": 1.5906414985656738, + "learning_rate": 1.500030824491502e-06, + "loss": 0.501, + "step": 18997 + }, + { + "epoch": 0.8411498400540164, + "grad_norm": 2.074532985687256, + "learning_rate": 1.4992168015202136e-06, + "loss": 0.5028, + "step": 18998 + }, + { + "epoch": 0.8411941157588302, + "grad_norm": 1.2022980451583862, + "learning_rate": 1.4984029815832102e-06, + "loss": 0.4283, + "step": 18999 + }, + { + "epoch": 0.8412383914636441, + "grad_norm": 1.6785246133804321, + "learning_rate": 1.4975893646999307e-06, + "loss": 0.5189, + "step": 19000 + }, + { + "epoch": 0.841282667168458, + "grad_norm": 1.318887710571289, + "learning_rate": 1.496775950889805e-06, + "loss": 0.4661, + "step": 19001 + }, + { + "epoch": 0.8413269428732718, + "grad_norm": 1.3665540218353271, + "learning_rate": 1.49596274017226e-06, + "loss": 0.4363, + "step": 19002 + }, + { + "epoch": 0.8413712185780857, + "grad_norm": 1.3065636157989502, + "learning_rate": 1.495149732566723e-06, + "loss": 0.4644, + "step": 19003 + }, + { + "epoch": 0.8414154942828996, + "grad_norm": 1.4254584312438965, + "learning_rate": 1.49433692809261e-06, + "loss": 0.4779, + "step": 19004 + }, + { + "epoch": 0.8414597699877135, + "grad_norm": 1.4833718538284302, + "learning_rate": 1.4935243267693322e-06, + "loss": 0.4958, + "step": 19005 + }, + { + "epoch": 0.8415040456925273, + "grad_norm": 1.3392345905303955, + "learning_rate": 1.492711928616304e-06, + "loss": 0.4574, + "step": 19006 + }, + { + "epoch": 0.8415483213973413, + "grad_norm": 1.200609564781189, + "learning_rate": 1.4918997336529216e-06, + "loss": 0.4126, + "step": 19007 + }, + { + "epoch": 0.8415925971021552, + "grad_norm": 1.6420254707336426, + "learning_rate": 1.4910877418985848e-06, + "loss": 0.512, + "step": 19008 + }, + { + "epoch": 0.841636872806969, + "grad_norm": 1.5347139835357666, + "learning_rate": 1.4902759533726907e-06, + "loss": 0.5086, + "step": 19009 + }, + { + "epoch": 0.8416811485117829, + "grad_norm": 1.4140903949737549, + "learning_rate": 1.4894643680946263e-06, + "loss": 0.4923, + "step": 19010 + }, + { + "epoch": 0.8417254242165968, + "grad_norm": 1.7434872388839722, + "learning_rate": 1.4886529860837772e-06, + "loss": 0.5212, + "step": 19011 + }, + { + "epoch": 0.8417696999214106, + "grad_norm": 1.5539487600326538, + "learning_rate": 1.487841807359518e-06, + "loss": 0.4859, + "step": 19012 + }, + { + "epoch": 0.8418139756262245, + "grad_norm": 1.1816346645355225, + "learning_rate": 1.4870308319412296e-06, + "loss": 0.4548, + "step": 19013 + }, + { + "epoch": 0.8418582513310384, + "grad_norm": 2.0334248542785645, + "learning_rate": 1.486220059848278e-06, + "loss": 0.538, + "step": 19014 + }, + { + "epoch": 0.8419025270358522, + "grad_norm": 1.6759164333343506, + "learning_rate": 1.4854094911000272e-06, + "loss": 0.521, + "step": 19015 + }, + { + "epoch": 0.8419468027406661, + "grad_norm": 1.3285455703735352, + "learning_rate": 1.4845991257158388e-06, + "loss": 0.4204, + "step": 19016 + }, + { + "epoch": 0.84199107844548, + "grad_norm": 1.40818190574646, + "learning_rate": 1.483788963715066e-06, + "loss": 0.4925, + "step": 19017 + }, + { + "epoch": 0.8420353541502938, + "grad_norm": 1.3044142723083496, + "learning_rate": 1.4829790051170578e-06, + "loss": 0.4409, + "step": 19018 + }, + { + "epoch": 0.8420796298551078, + "grad_norm": 1.5457407236099243, + "learning_rate": 1.4821692499411632e-06, + "loss": 0.4889, + "step": 19019 + }, + { + "epoch": 0.8421239055599217, + "grad_norm": 1.7866480350494385, + "learning_rate": 1.4813596982067214e-06, + "loss": 0.562, + "step": 19020 + }, + { + "epoch": 0.8421681812647355, + "grad_norm": 1.9629855155944824, + "learning_rate": 1.4805503499330663e-06, + "loss": 0.4818, + "step": 19021 + }, + { + "epoch": 0.8422124569695494, + "grad_norm": 1.4263733625411987, + "learning_rate": 1.4797412051395278e-06, + "loss": 0.4629, + "step": 19022 + }, + { + "epoch": 0.8422567326743633, + "grad_norm": 1.418455958366394, + "learning_rate": 1.478932263845435e-06, + "loss": 0.5238, + "step": 19023 + }, + { + "epoch": 0.8423010083791771, + "grad_norm": 1.8410089015960693, + "learning_rate": 1.4781235260701065e-06, + "loss": 0.5455, + "step": 19024 + }, + { + "epoch": 0.842345284083991, + "grad_norm": 1.6749563217163086, + "learning_rate": 1.47731499183286e-06, + "loss": 0.5204, + "step": 19025 + }, + { + "epoch": 0.8423895597888049, + "grad_norm": 1.5651817321777344, + "learning_rate": 1.4765066611530043e-06, + "loss": 0.4959, + "step": 19026 + }, + { + "epoch": 0.8424338354936187, + "grad_norm": 1.413029432296753, + "learning_rate": 1.4756985340498476e-06, + "loss": 0.4486, + "step": 19027 + }, + { + "epoch": 0.8424781111984326, + "grad_norm": 1.4073671102523804, + "learning_rate": 1.4748906105426875e-06, + "loss": 0.4567, + "step": 19028 + }, + { + "epoch": 0.8425223869032465, + "grad_norm": 1.409041166305542, + "learning_rate": 1.4740828906508275e-06, + "loss": 0.4835, + "step": 19029 + }, + { + "epoch": 0.8425666626080603, + "grad_norm": 1.6517341136932373, + "learning_rate": 1.4732753743935546e-06, + "loss": 0.5401, + "step": 19030 + }, + { + "epoch": 0.8426109383128743, + "grad_norm": 1.3802677392959595, + "learning_rate": 1.4724680617901565e-06, + "loss": 0.4915, + "step": 19031 + }, + { + "epoch": 0.8426552140176882, + "grad_norm": 1.634418249130249, + "learning_rate": 1.4716609528599134e-06, + "loss": 0.482, + "step": 19032 + }, + { + "epoch": 0.842699489722502, + "grad_norm": 1.3967403173446655, + "learning_rate": 1.4708540476221068e-06, + "loss": 0.5428, + "step": 19033 + }, + { + "epoch": 0.8427437654273159, + "grad_norm": 1.465164303779602, + "learning_rate": 1.4700473460960062e-06, + "loss": 0.5056, + "step": 19034 + }, + { + "epoch": 0.8427880411321298, + "grad_norm": 1.6842405796051025, + "learning_rate": 1.4692408483008803e-06, + "loss": 0.5075, + "step": 19035 + }, + { + "epoch": 0.8428323168369436, + "grad_norm": 1.1825623512268066, + "learning_rate": 1.4684345542559896e-06, + "loss": 0.4095, + "step": 19036 + }, + { + "epoch": 0.8428765925417575, + "grad_norm": 1.6428176164627075, + "learning_rate": 1.4676284639805937e-06, + "loss": 0.4665, + "step": 19037 + }, + { + "epoch": 0.8429208682465714, + "grad_norm": 1.4389874935150146, + "learning_rate": 1.4668225774939425e-06, + "loss": 0.5303, + "step": 19038 + }, + { + "epoch": 0.8429651439513852, + "grad_norm": 1.3961561918258667, + "learning_rate": 1.4660168948152875e-06, + "loss": 0.5034, + "step": 19039 + }, + { + "epoch": 0.8430094196561991, + "grad_norm": 1.65833580493927, + "learning_rate": 1.4652114159638707e-06, + "loss": 0.505, + "step": 19040 + }, + { + "epoch": 0.843053695361013, + "grad_norm": 2.2129480838775635, + "learning_rate": 1.4644061409589306e-06, + "loss": 0.5553, + "step": 19041 + }, + { + "epoch": 0.843097971065827, + "grad_norm": 1.2661912441253662, + "learning_rate": 1.4636010698196967e-06, + "loss": 0.4402, + "step": 19042 + }, + { + "epoch": 0.8431422467706408, + "grad_norm": 1.5125845670700073, + "learning_rate": 1.462796202565403e-06, + "loss": 0.507, + "step": 19043 + }, + { + "epoch": 0.8431865224754547, + "grad_norm": 1.6948983669281006, + "learning_rate": 1.4619915392152718e-06, + "loss": 0.5464, + "step": 19044 + }, + { + "epoch": 0.8432307981802686, + "grad_norm": 1.4907947778701782, + "learning_rate": 1.4611870797885196e-06, + "loss": 0.4522, + "step": 19045 + }, + { + "epoch": 0.8432750738850824, + "grad_norm": 1.6173666715621948, + "learning_rate": 1.4603828243043616e-06, + "loss": 0.5519, + "step": 19046 + }, + { + "epoch": 0.8433193495898963, + "grad_norm": 1.9244611263275146, + "learning_rate": 1.4595787727820078e-06, + "loss": 0.5216, + "step": 19047 + }, + { + "epoch": 0.8433636252947102, + "grad_norm": 1.473968267440796, + "learning_rate": 1.458774925240657e-06, + "loss": 0.5161, + "step": 19048 + }, + { + "epoch": 0.843407900999524, + "grad_norm": 1.5091592073440552, + "learning_rate": 1.4579712816995162e-06, + "loss": 0.4469, + "step": 19049 + }, + { + "epoch": 0.8434521767043379, + "grad_norm": 1.562118649482727, + "learning_rate": 1.4571678421777757e-06, + "loss": 0.5322, + "step": 19050 + }, + { + "epoch": 0.8434964524091518, + "grad_norm": 1.6490087509155273, + "learning_rate": 1.4563646066946258e-06, + "loss": 0.4974, + "step": 19051 + }, + { + "epoch": 0.8435407281139656, + "grad_norm": 1.3898804187774658, + "learning_rate": 1.4555615752692476e-06, + "loss": 0.4834, + "step": 19052 + }, + { + "epoch": 0.8435850038187795, + "grad_norm": 1.5850328207015991, + "learning_rate": 1.4547587479208269e-06, + "loss": 0.4158, + "step": 19053 + }, + { + "epoch": 0.8436292795235935, + "grad_norm": 1.5743775367736816, + "learning_rate": 1.4539561246685386e-06, + "loss": 0.4718, + "step": 19054 + }, + { + "epoch": 0.8436735552284073, + "grad_norm": 2.0087082386016846, + "learning_rate": 1.4531537055315447e-06, + "loss": 0.5159, + "step": 19055 + }, + { + "epoch": 0.8437178309332212, + "grad_norm": 1.401806354522705, + "learning_rate": 1.452351490529017e-06, + "loss": 0.4586, + "step": 19056 + }, + { + "epoch": 0.8437621066380351, + "grad_norm": 1.5394171476364136, + "learning_rate": 1.4515494796801156e-06, + "loss": 0.4719, + "step": 19057 + }, + { + "epoch": 0.8438063823428489, + "grad_norm": 1.7787114381790161, + "learning_rate": 1.4507476730039937e-06, + "loss": 0.5637, + "step": 19058 + }, + { + "epoch": 0.8438506580476628, + "grad_norm": 1.6162071228027344, + "learning_rate": 1.4499460705198e-06, + "loss": 0.5636, + "step": 19059 + }, + { + "epoch": 0.8438949337524767, + "grad_norm": 1.8361092805862427, + "learning_rate": 1.449144672246685e-06, + "loss": 0.5396, + "step": 19060 + }, + { + "epoch": 0.8439392094572905, + "grad_norm": 1.5720571279525757, + "learning_rate": 1.4483434782037876e-06, + "loss": 0.4737, + "step": 19061 + }, + { + "epoch": 0.8439834851621044, + "grad_norm": 2.2882659435272217, + "learning_rate": 1.447542488410243e-06, + "loss": 0.5116, + "step": 19062 + }, + { + "epoch": 0.8440277608669183, + "grad_norm": 1.5425450801849365, + "learning_rate": 1.4467417028851826e-06, + "loss": 0.5124, + "step": 19063 + }, + { + "epoch": 0.8440720365717321, + "grad_norm": 1.7098582983016968, + "learning_rate": 1.4459411216477326e-06, + "loss": 0.5384, + "step": 19064 + }, + { + "epoch": 0.844116312276546, + "grad_norm": 1.3732982873916626, + "learning_rate": 1.4451407447170108e-06, + "loss": 0.4837, + "step": 19065 + }, + { + "epoch": 0.84416058798136, + "grad_norm": 1.6755344867706299, + "learning_rate": 1.4443405721121394e-06, + "loss": 0.5215, + "step": 19066 + }, + { + "epoch": 0.8442048636861738, + "grad_norm": 1.8798960447311401, + "learning_rate": 1.443540603852227e-06, + "loss": 0.5293, + "step": 19067 + }, + { + "epoch": 0.8442491393909877, + "grad_norm": 1.4563984870910645, + "learning_rate": 1.4427408399563813e-06, + "loss": 0.5316, + "step": 19068 + }, + { + "epoch": 0.8442934150958016, + "grad_norm": 1.5441702604293823, + "learning_rate": 1.4419412804436995e-06, + "loss": 0.5246, + "step": 19069 + }, + { + "epoch": 0.8443376908006154, + "grad_norm": 1.6809531450271606, + "learning_rate": 1.4411419253332848e-06, + "loss": 0.4748, + "step": 19070 + }, + { + "epoch": 0.8443819665054293, + "grad_norm": 1.9797074794769287, + "learning_rate": 1.4403427746442267e-06, + "loss": 0.5182, + "step": 19071 + }, + { + "epoch": 0.8444262422102432, + "grad_norm": 1.5860399007797241, + "learning_rate": 1.4395438283956109e-06, + "loss": 0.4785, + "step": 19072 + }, + { + "epoch": 0.844470517915057, + "grad_norm": 1.6037729978561401, + "learning_rate": 1.438745086606521e-06, + "loss": 0.4927, + "step": 19073 + }, + { + "epoch": 0.8445147936198709, + "grad_norm": 1.2541452646255493, + "learning_rate": 1.4379465492960342e-06, + "loss": 0.4386, + "step": 19074 + }, + { + "epoch": 0.8445590693246848, + "grad_norm": 2.407467842102051, + "learning_rate": 1.4371482164832206e-06, + "loss": 0.6312, + "step": 19075 + }, + { + "epoch": 0.8446033450294987, + "grad_norm": 1.2689940929412842, + "learning_rate": 1.4363500881871516e-06, + "loss": 0.4892, + "step": 19076 + }, + { + "epoch": 0.8446476207343125, + "grad_norm": 1.4461582899093628, + "learning_rate": 1.4355521644268876e-06, + "loss": 0.4551, + "step": 19077 + }, + { + "epoch": 0.8446918964391265, + "grad_norm": 1.6944515705108643, + "learning_rate": 1.4347544452214869e-06, + "loss": 0.5477, + "step": 19078 + }, + { + "epoch": 0.8447361721439404, + "grad_norm": 1.2739887237548828, + "learning_rate": 1.4339569305900004e-06, + "loss": 0.4388, + "step": 19079 + }, + { + "epoch": 0.8447804478487542, + "grad_norm": 1.7288217544555664, + "learning_rate": 1.4331596205514798e-06, + "loss": 0.5214, + "step": 19080 + }, + { + "epoch": 0.8448247235535681, + "grad_norm": 1.6440589427947998, + "learning_rate": 1.4323625151249665e-06, + "loss": 0.4773, + "step": 19081 + }, + { + "epoch": 0.844868999258382, + "grad_norm": 1.8754832744598389, + "learning_rate": 1.4315656143294975e-06, + "loss": 0.502, + "step": 19082 + }, + { + "epoch": 0.8449132749631958, + "grad_norm": 1.5075256824493408, + "learning_rate": 1.4307689181841077e-06, + "loss": 0.5285, + "step": 19083 + }, + { + "epoch": 0.8449575506680097, + "grad_norm": 1.6319196224212646, + "learning_rate": 1.4299724267078252e-06, + "loss": 0.5176, + "step": 19084 + }, + { + "epoch": 0.8450018263728236, + "grad_norm": 1.555267572402954, + "learning_rate": 1.4291761399196713e-06, + "loss": 0.4512, + "step": 19085 + }, + { + "epoch": 0.8450461020776374, + "grad_norm": 1.3803372383117676, + "learning_rate": 1.4283800578386686e-06, + "loss": 0.4933, + "step": 19086 + }, + { + "epoch": 0.8450903777824513, + "grad_norm": 1.7094953060150146, + "learning_rate": 1.4275841804838298e-06, + "loss": 0.4767, + "step": 19087 + }, + { + "epoch": 0.8451346534872652, + "grad_norm": 1.3921595811843872, + "learning_rate": 1.4267885078741617e-06, + "loss": 0.4414, + "step": 19088 + }, + { + "epoch": 0.845178929192079, + "grad_norm": 1.5273886919021606, + "learning_rate": 1.425993040028667e-06, + "loss": 0.4982, + "step": 19089 + }, + { + "epoch": 0.845223204896893, + "grad_norm": 1.6607378721237183, + "learning_rate": 1.4251977769663506e-06, + "loss": 0.5084, + "step": 19090 + }, + { + "epoch": 0.8452674806017069, + "grad_norm": 1.7049437761306763, + "learning_rate": 1.4244027187062037e-06, + "loss": 0.5397, + "step": 19091 + }, + { + "epoch": 0.8453117563065207, + "grad_norm": 1.4843964576721191, + "learning_rate": 1.4236078652672148e-06, + "loss": 0.5308, + "step": 19092 + }, + { + "epoch": 0.8453560320113346, + "grad_norm": 1.465858817100525, + "learning_rate": 1.4228132166683683e-06, + "loss": 0.4346, + "step": 19093 + }, + { + "epoch": 0.8454003077161485, + "grad_norm": 1.4483979940414429, + "learning_rate": 1.422018772928645e-06, + "loss": 0.5067, + "step": 19094 + }, + { + "epoch": 0.8454445834209623, + "grad_norm": 1.7789232730865479, + "learning_rate": 1.421224534067016e-06, + "loss": 0.5328, + "step": 19095 + }, + { + "epoch": 0.8454888591257762, + "grad_norm": 1.7280694246292114, + "learning_rate": 1.4204305001024566e-06, + "loss": 0.5013, + "step": 19096 + }, + { + "epoch": 0.8455331348305901, + "grad_norm": 1.5234733819961548, + "learning_rate": 1.419636671053929e-06, + "loss": 0.5147, + "step": 19097 + }, + { + "epoch": 0.8455774105354039, + "grad_norm": 1.723885416984558, + "learning_rate": 1.4188430469403935e-06, + "loss": 0.4967, + "step": 19098 + }, + { + "epoch": 0.8456216862402178, + "grad_norm": 2.4840738773345947, + "learning_rate": 1.4180496277808021e-06, + "loss": 0.6109, + "step": 19099 + }, + { + "epoch": 0.8456659619450317, + "grad_norm": 1.2865008115768433, + "learning_rate": 1.4172564135941103e-06, + "loss": 0.4847, + "step": 19100 + }, + { + "epoch": 0.8457102376498455, + "grad_norm": 1.5070514678955078, + "learning_rate": 1.416463404399262e-06, + "loss": 0.4772, + "step": 19101 + }, + { + "epoch": 0.8457545133546595, + "grad_norm": 1.5488075017929077, + "learning_rate": 1.415670600215192e-06, + "loss": 0.4763, + "step": 19102 + }, + { + "epoch": 0.8457987890594734, + "grad_norm": 1.4914301633834839, + "learning_rate": 1.4148780010608431e-06, + "loss": 0.4896, + "step": 19103 + }, + { + "epoch": 0.8458430647642872, + "grad_norm": 1.5790902376174927, + "learning_rate": 1.4140856069551423e-06, + "loss": 0.5302, + "step": 19104 + }, + { + "epoch": 0.8458873404691011, + "grad_norm": 1.491227626800537, + "learning_rate": 1.4132934179170144e-06, + "loss": 0.4595, + "step": 19105 + }, + { + "epoch": 0.845931616173915, + "grad_norm": 1.8381580114364624, + "learning_rate": 1.4125014339653832e-06, + "loss": 0.5696, + "step": 19106 + }, + { + "epoch": 0.8459758918787288, + "grad_norm": 1.5455408096313477, + "learning_rate": 1.4117096551191633e-06, + "loss": 0.4948, + "step": 19107 + }, + { + "epoch": 0.8460201675835427, + "grad_norm": 1.3771133422851562, + "learning_rate": 1.4109180813972656e-06, + "loss": 0.4453, + "step": 19108 + }, + { + "epoch": 0.8460644432883566, + "grad_norm": 1.2839752435684204, + "learning_rate": 1.4101267128185936e-06, + "loss": 0.4677, + "step": 19109 + }, + { + "epoch": 0.8461087189931705, + "grad_norm": 1.4090015888214111, + "learning_rate": 1.4093355494020555e-06, + "loss": 0.4317, + "step": 19110 + }, + { + "epoch": 0.8461529946979843, + "grad_norm": 1.5058598518371582, + "learning_rate": 1.4085445911665407e-06, + "loss": 0.5213, + "step": 19111 + }, + { + "epoch": 0.8461972704027982, + "grad_norm": 1.571440577507019, + "learning_rate": 1.4077538381309396e-06, + "loss": 0.491, + "step": 19112 + }, + { + "epoch": 0.8462415461076122, + "grad_norm": 1.3887795209884644, + "learning_rate": 1.4069632903141462e-06, + "loss": 0.4406, + "step": 19113 + }, + { + "epoch": 0.846285821812426, + "grad_norm": 1.5139282941818237, + "learning_rate": 1.4061729477350361e-06, + "loss": 0.5265, + "step": 19114 + }, + { + "epoch": 0.8463300975172399, + "grad_norm": 1.5292582511901855, + "learning_rate": 1.4053828104124867e-06, + "loss": 0.4701, + "step": 19115 + }, + { + "epoch": 0.8463743732220538, + "grad_norm": 1.4867281913757324, + "learning_rate": 1.4045928783653728e-06, + "loss": 0.4809, + "step": 19116 + }, + { + "epoch": 0.8464186489268676, + "grad_norm": 1.422751784324646, + "learning_rate": 1.4038031516125595e-06, + "loss": 0.4875, + "step": 19117 + }, + { + "epoch": 0.8464629246316815, + "grad_norm": 1.729909062385559, + "learning_rate": 1.4030136301729091e-06, + "loss": 0.4708, + "step": 19118 + }, + { + "epoch": 0.8465072003364954, + "grad_norm": 1.6581380367279053, + "learning_rate": 1.402224314065277e-06, + "loss": 0.5155, + "step": 19119 + }, + { + "epoch": 0.8465514760413092, + "grad_norm": 1.3480337858200073, + "learning_rate": 1.4014352033085177e-06, + "loss": 0.4651, + "step": 19120 + }, + { + "epoch": 0.8465957517461231, + "grad_norm": 1.2964915037155151, + "learning_rate": 1.4006462979214762e-06, + "loss": 0.4329, + "step": 19121 + }, + { + "epoch": 0.846640027450937, + "grad_norm": 1.4374067783355713, + "learning_rate": 1.3998575979229944e-06, + "loss": 0.4687, + "step": 19122 + }, + { + "epoch": 0.8466843031557508, + "grad_norm": 1.7215275764465332, + "learning_rate": 1.3990691033319149e-06, + "loss": 0.5295, + "step": 19123 + }, + { + "epoch": 0.8467285788605647, + "grad_norm": 1.2617990970611572, + "learning_rate": 1.3982808141670644e-06, + "loss": 0.4596, + "step": 19124 + }, + { + "epoch": 0.8467728545653787, + "grad_norm": 1.3364033699035645, + "learning_rate": 1.3974927304472718e-06, + "loss": 0.4865, + "step": 19125 + }, + { + "epoch": 0.8468171302701925, + "grad_norm": 1.3457008600234985, + "learning_rate": 1.396704852191364e-06, + "loss": 0.41, + "step": 19126 + }, + { + "epoch": 0.8468614059750064, + "grad_norm": 1.4689844846725464, + "learning_rate": 1.3959171794181548e-06, + "loss": 0.4832, + "step": 19127 + }, + { + "epoch": 0.8469056816798203, + "grad_norm": 1.7881070375442505, + "learning_rate": 1.395129712146458e-06, + "loss": 0.5234, + "step": 19128 + }, + { + "epoch": 0.8469499573846341, + "grad_norm": 1.8994979858398438, + "learning_rate": 1.394342450395082e-06, + "loss": 0.5515, + "step": 19129 + }, + { + "epoch": 0.846994233089448, + "grad_norm": 1.6240384578704834, + "learning_rate": 1.3935553941828294e-06, + "loss": 0.4854, + "step": 19130 + }, + { + "epoch": 0.8470385087942619, + "grad_norm": 1.3545037508010864, + "learning_rate": 1.3927685435284977e-06, + "loss": 0.4418, + "step": 19131 + }, + { + "epoch": 0.8470827844990757, + "grad_norm": 1.5531848669052124, + "learning_rate": 1.3919818984508805e-06, + "loss": 0.5036, + "step": 19132 + }, + { + "epoch": 0.8471270602038896, + "grad_norm": 1.7281091213226318, + "learning_rate": 1.3911954589687683e-06, + "loss": 0.4897, + "step": 19133 + }, + { + "epoch": 0.8471713359087035, + "grad_norm": 1.8513411283493042, + "learning_rate": 1.3904092251009426e-06, + "loss": 0.5395, + "step": 19134 + }, + { + "epoch": 0.8472156116135173, + "grad_norm": 1.5129708051681519, + "learning_rate": 1.38962319686618e-06, + "loss": 0.4732, + "step": 19135 + }, + { + "epoch": 0.8472598873183312, + "grad_norm": 1.473106026649475, + "learning_rate": 1.3888373742832595e-06, + "loss": 0.5017, + "step": 19136 + }, + { + "epoch": 0.8473041630231452, + "grad_norm": 1.5443557500839233, + "learning_rate": 1.3880517573709461e-06, + "loss": 0.4626, + "step": 19137 + }, + { + "epoch": 0.847348438727959, + "grad_norm": 1.37923264503479, + "learning_rate": 1.3872663461480052e-06, + "loss": 0.4496, + "step": 19138 + }, + { + "epoch": 0.8473927144327729, + "grad_norm": 1.3002583980560303, + "learning_rate": 1.3864811406331946e-06, + "loss": 0.4838, + "step": 19139 + }, + { + "epoch": 0.8474369901375868, + "grad_norm": 1.6210039854049683, + "learning_rate": 1.3856961408452685e-06, + "loss": 0.5276, + "step": 19140 + }, + { + "epoch": 0.8474812658424006, + "grad_norm": 1.5049923658370972, + "learning_rate": 1.3849113468029762e-06, + "loss": 0.4921, + "step": 19141 + }, + { + "epoch": 0.8475255415472145, + "grad_norm": 1.2653083801269531, + "learning_rate": 1.3841267585250585e-06, + "loss": 0.4526, + "step": 19142 + }, + { + "epoch": 0.8475698172520284, + "grad_norm": 1.4275647401809692, + "learning_rate": 1.383342376030261e-06, + "loss": 0.5018, + "step": 19143 + }, + { + "epoch": 0.8476140929568422, + "grad_norm": 1.511393666267395, + "learning_rate": 1.382558199337315e-06, + "loss": 0.4425, + "step": 19144 + }, + { + "epoch": 0.8476583686616561, + "grad_norm": 1.4526900053024292, + "learning_rate": 1.3817742284649471e-06, + "loss": 0.4532, + "step": 19145 + }, + { + "epoch": 0.84770264436647, + "grad_norm": 1.6205475330352783, + "learning_rate": 1.380990463431887e-06, + "loss": 0.523, + "step": 19146 + }, + { + "epoch": 0.8477469200712839, + "grad_norm": 1.7806531190872192, + "learning_rate": 1.3802069042568521e-06, + "loss": 0.5226, + "step": 19147 + }, + { + "epoch": 0.8477911957760977, + "grad_norm": 1.7810156345367432, + "learning_rate": 1.3794235509585562e-06, + "loss": 0.4698, + "step": 19148 + }, + { + "epoch": 0.8478354714809117, + "grad_norm": 1.853472352027893, + "learning_rate": 1.3786404035557099e-06, + "loss": 0.5164, + "step": 19149 + }, + { + "epoch": 0.8478797471857256, + "grad_norm": 1.422011375427246, + "learning_rate": 1.3778574620670181e-06, + "loss": 0.4909, + "step": 19150 + }, + { + "epoch": 0.8479240228905394, + "grad_norm": 1.4226434230804443, + "learning_rate": 1.3770747265111805e-06, + "loss": 0.4842, + "step": 19151 + }, + { + "epoch": 0.8479682985953533, + "grad_norm": 1.5234347581863403, + "learning_rate": 1.37629219690689e-06, + "loss": 0.512, + "step": 19152 + }, + { + "epoch": 0.8480125743001672, + "grad_norm": 1.270546793937683, + "learning_rate": 1.37550987327284e-06, + "loss": 0.4448, + "step": 19153 + }, + { + "epoch": 0.848056850004981, + "grad_norm": 1.356495976448059, + "learning_rate": 1.374727755627715e-06, + "loss": 0.4571, + "step": 19154 + }, + { + "epoch": 0.8481011257097949, + "grad_norm": 1.4092117547988892, + "learning_rate": 1.373945843990192e-06, + "loss": 0.5154, + "step": 19155 + }, + { + "epoch": 0.8481454014146088, + "grad_norm": 1.301049828529358, + "learning_rate": 1.3731641383789507e-06, + "loss": 0.4503, + "step": 19156 + }, + { + "epoch": 0.8481896771194226, + "grad_norm": 1.2697954177856445, + "learning_rate": 1.3723826388126626e-06, + "loss": 0.446, + "step": 19157 + }, + { + "epoch": 0.8482339528242365, + "grad_norm": 1.4332674741744995, + "learning_rate": 1.3716013453099885e-06, + "loss": 0.4805, + "step": 19158 + }, + { + "epoch": 0.8482782285290504, + "grad_norm": 1.8303436040878296, + "learning_rate": 1.3708202578895879e-06, + "loss": 0.488, + "step": 19159 + }, + { + "epoch": 0.8483225042338642, + "grad_norm": 1.6216542720794678, + "learning_rate": 1.3700393765701215e-06, + "loss": 0.5076, + "step": 19160 + }, + { + "epoch": 0.8483667799386782, + "grad_norm": 1.4269764423370361, + "learning_rate": 1.3692587013702385e-06, + "loss": 0.4666, + "step": 19161 + }, + { + "epoch": 0.8484110556434921, + "grad_norm": 1.5987385511398315, + "learning_rate": 1.368478232308581e-06, + "loss": 0.5494, + "step": 19162 + }, + { + "epoch": 0.8484553313483059, + "grad_norm": 1.3686208724975586, + "learning_rate": 1.367697969403795e-06, + "loss": 0.4423, + "step": 19163 + }, + { + "epoch": 0.8484996070531198, + "grad_norm": 1.4524286985397339, + "learning_rate": 1.3669179126745147e-06, + "loss": 0.4585, + "step": 19164 + }, + { + "epoch": 0.8485438827579337, + "grad_norm": 1.453385353088379, + "learning_rate": 1.3661380621393705e-06, + "loss": 0.4727, + "step": 19165 + }, + { + "epoch": 0.8485881584627475, + "grad_norm": 1.595625638961792, + "learning_rate": 1.3653584178169899e-06, + "loss": 0.5138, + "step": 19166 + }, + { + "epoch": 0.8486324341675614, + "grad_norm": 1.7220985889434814, + "learning_rate": 1.3645789797259912e-06, + "loss": 0.5194, + "step": 19167 + }, + { + "epoch": 0.8486767098723753, + "grad_norm": 1.7794209718704224, + "learning_rate": 1.363799747884993e-06, + "loss": 0.5311, + "step": 19168 + }, + { + "epoch": 0.8487209855771891, + "grad_norm": 1.7559757232666016, + "learning_rate": 1.3630207223126046e-06, + "loss": 0.5029, + "step": 19169 + }, + { + "epoch": 0.848765261282003, + "grad_norm": 1.4175853729248047, + "learning_rate": 1.3622419030274347e-06, + "loss": 0.4497, + "step": 19170 + }, + { + "epoch": 0.8488095369868169, + "grad_norm": 2.1522951126098633, + "learning_rate": 1.361463290048085e-06, + "loss": 0.5464, + "step": 19171 + }, + { + "epoch": 0.8488538126916307, + "grad_norm": 2.7959280014038086, + "learning_rate": 1.3606848833931474e-06, + "loss": 0.5998, + "step": 19172 + }, + { + "epoch": 0.8488980883964446, + "grad_norm": 2.216120481491089, + "learning_rate": 1.3599066830812202e-06, + "loss": 0.5144, + "step": 19173 + }, + { + "epoch": 0.8489423641012586, + "grad_norm": 1.3968886137008667, + "learning_rate": 1.3591286891308863e-06, + "loss": 0.4661, + "step": 19174 + }, + { + "epoch": 0.8489866398060724, + "grad_norm": 1.5580532550811768, + "learning_rate": 1.3583509015607276e-06, + "loss": 0.5432, + "step": 19175 + }, + { + "epoch": 0.8490309155108863, + "grad_norm": 1.5228044986724854, + "learning_rate": 1.3575733203893227e-06, + "loss": 0.4589, + "step": 19176 + }, + { + "epoch": 0.8490751912157002, + "grad_norm": 1.3431458473205566, + "learning_rate": 1.356795945635241e-06, + "loss": 0.4603, + "step": 19177 + }, + { + "epoch": 0.849119466920514, + "grad_norm": 1.5897451639175415, + "learning_rate": 1.3560187773170508e-06, + "loss": 0.4882, + "step": 19178 + }, + { + "epoch": 0.8491637426253279, + "grad_norm": 2.0685811042785645, + "learning_rate": 1.355241815453312e-06, + "loss": 0.5133, + "step": 19179 + }, + { + "epoch": 0.8492080183301418, + "grad_norm": 1.7291475534439087, + "learning_rate": 1.3544650600625854e-06, + "loss": 0.5119, + "step": 19180 + }, + { + "epoch": 0.8492522940349557, + "grad_norm": 1.4166728258132935, + "learning_rate": 1.3536885111634212e-06, + "loss": 0.4154, + "step": 19181 + }, + { + "epoch": 0.8492965697397695, + "grad_norm": 1.7796833515167236, + "learning_rate": 1.3529121687743652e-06, + "loss": 0.5211, + "step": 19182 + }, + { + "epoch": 0.8493408454445834, + "grad_norm": 1.755669116973877, + "learning_rate": 1.3521360329139632e-06, + "loss": 0.5369, + "step": 19183 + }, + { + "epoch": 0.8493851211493973, + "grad_norm": 1.4913042783737183, + "learning_rate": 1.3513601036007506e-06, + "loss": 0.4755, + "step": 19184 + }, + { + "epoch": 0.8494293968542111, + "grad_norm": 1.5155978202819824, + "learning_rate": 1.3505843808532605e-06, + "loss": 0.515, + "step": 19185 + }, + { + "epoch": 0.8494736725590251, + "grad_norm": 1.5726250410079956, + "learning_rate": 1.3498088646900199e-06, + "loss": 0.4554, + "step": 19186 + }, + { + "epoch": 0.849517948263839, + "grad_norm": 1.2599495649337769, + "learning_rate": 1.3490335551295507e-06, + "loss": 0.4487, + "step": 19187 + }, + { + "epoch": 0.8495622239686528, + "grad_norm": 1.3122835159301758, + "learning_rate": 1.3482584521903718e-06, + "loss": 0.4873, + "step": 19188 + }, + { + "epoch": 0.8496064996734667, + "grad_norm": 1.6088758707046509, + "learning_rate": 1.3474835558909927e-06, + "loss": 0.4868, + "step": 19189 + }, + { + "epoch": 0.8496507753782806, + "grad_norm": 1.7922743558883667, + "learning_rate": 1.346708866249925e-06, + "loss": 0.5197, + "step": 19190 + }, + { + "epoch": 0.8496950510830944, + "grad_norm": 1.3210550546646118, + "learning_rate": 1.345934383285672e-06, + "loss": 0.4634, + "step": 19191 + }, + { + "epoch": 0.8497393267879083, + "grad_norm": 1.4622522592544556, + "learning_rate": 1.3451601070167264e-06, + "loss": 0.5406, + "step": 19192 + }, + { + "epoch": 0.8497836024927222, + "grad_norm": 1.5927870273590088, + "learning_rate": 1.3443860374615869e-06, + "loss": 0.5031, + "step": 19193 + }, + { + "epoch": 0.849827878197536, + "grad_norm": 1.9271926879882812, + "learning_rate": 1.3436121746387398e-06, + "loss": 0.5583, + "step": 19194 + }, + { + "epoch": 0.8498721539023499, + "grad_norm": 1.3488342761993408, + "learning_rate": 1.342838518566667e-06, + "loss": 0.4429, + "step": 19195 + }, + { + "epoch": 0.8499164296071638, + "grad_norm": 1.5566197633743286, + "learning_rate": 1.342065069263848e-06, + "loss": 0.474, + "step": 19196 + }, + { + "epoch": 0.8499607053119776, + "grad_norm": 1.6411066055297852, + "learning_rate": 1.341291826748755e-06, + "loss": 0.5148, + "step": 19197 + }, + { + "epoch": 0.8500049810167916, + "grad_norm": 2.9497382640838623, + "learning_rate": 1.3405187910398565e-06, + "loss": 0.5535, + "step": 19198 + }, + { + "epoch": 0.8500492567216055, + "grad_norm": 1.386014461517334, + "learning_rate": 1.339745962155613e-06, + "loss": 0.4836, + "step": 19199 + }, + { + "epoch": 0.8500935324264193, + "grad_norm": 1.259560465812683, + "learning_rate": 1.3389733401144889e-06, + "loss": 0.4319, + "step": 19200 + }, + { + "epoch": 0.8501378081312332, + "grad_norm": 1.8688299655914307, + "learning_rate": 1.3382009249349337e-06, + "loss": 0.5135, + "step": 19201 + }, + { + "epoch": 0.8501820838360471, + "grad_norm": 1.481195092201233, + "learning_rate": 1.337428716635395e-06, + "loss": 0.4499, + "step": 19202 + }, + { + "epoch": 0.8502263595408609, + "grad_norm": 1.4071801900863647, + "learning_rate": 1.33665671523432e-06, + "loss": 0.4669, + "step": 19203 + }, + { + "epoch": 0.8502706352456748, + "grad_norm": 1.3979783058166504, + "learning_rate": 1.3358849207501467e-06, + "loss": 0.4478, + "step": 19204 + }, + { + "epoch": 0.8503149109504887, + "grad_norm": 2.245912790298462, + "learning_rate": 1.3351133332013077e-06, + "loss": 0.5125, + "step": 19205 + }, + { + "epoch": 0.8503591866553025, + "grad_norm": 1.6613420248031616, + "learning_rate": 1.3343419526062285e-06, + "loss": 0.4963, + "step": 19206 + }, + { + "epoch": 0.8504034623601164, + "grad_norm": 1.8479807376861572, + "learning_rate": 1.3335707789833374e-06, + "loss": 0.5078, + "step": 19207 + }, + { + "epoch": 0.8504477380649303, + "grad_norm": 1.3186440467834473, + "learning_rate": 1.332799812351052e-06, + "loss": 0.4677, + "step": 19208 + }, + { + "epoch": 0.8504920137697441, + "grad_norm": 1.388060450553894, + "learning_rate": 1.3320290527277845e-06, + "loss": 0.4807, + "step": 19209 + }, + { + "epoch": 0.8505362894745581, + "grad_norm": 1.5659207105636597, + "learning_rate": 1.3312585001319466e-06, + "loss": 0.4932, + "step": 19210 + }, + { + "epoch": 0.850580565179372, + "grad_norm": 1.326371192932129, + "learning_rate": 1.3304881545819415e-06, + "loss": 0.4208, + "step": 19211 + }, + { + "epoch": 0.8506248408841858, + "grad_norm": 1.4852699041366577, + "learning_rate": 1.3297180160961654e-06, + "loss": 0.4424, + "step": 19212 + }, + { + "epoch": 0.8506691165889997, + "grad_norm": 1.4432440996170044, + "learning_rate": 1.328948084693019e-06, + "loss": 0.4727, + "step": 19213 + }, + { + "epoch": 0.8507133922938136, + "grad_norm": 1.6421525478363037, + "learning_rate": 1.328178360390886e-06, + "loss": 0.5365, + "step": 19214 + }, + { + "epoch": 0.8507576679986274, + "grad_norm": 1.1135965585708618, + "learning_rate": 1.3274088432081511e-06, + "loss": 0.3745, + "step": 19215 + }, + { + "epoch": 0.8508019437034413, + "grad_norm": 1.560402750968933, + "learning_rate": 1.3266395331631933e-06, + "loss": 0.543, + "step": 19216 + }, + { + "epoch": 0.8508462194082552, + "grad_norm": 1.395042061805725, + "learning_rate": 1.3258704302743897e-06, + "loss": 0.4679, + "step": 19217 + }, + { + "epoch": 0.8508904951130691, + "grad_norm": 1.460881233215332, + "learning_rate": 1.3251015345601082e-06, + "loss": 0.4775, + "step": 19218 + }, + { + "epoch": 0.8509347708178829, + "grad_norm": 1.577001929283142, + "learning_rate": 1.3243328460387106e-06, + "loss": 0.4874, + "step": 19219 + }, + { + "epoch": 0.8509790465226968, + "grad_norm": 1.5787314176559448, + "learning_rate": 1.3235643647285612e-06, + "loss": 0.4925, + "step": 19220 + }, + { + "epoch": 0.8510233222275108, + "grad_norm": 1.513409972190857, + "learning_rate": 1.322796090648013e-06, + "loss": 0.4953, + "step": 19221 + }, + { + "epoch": 0.8510675979323246, + "grad_norm": 1.5202165842056274, + "learning_rate": 1.3220280238154148e-06, + "loss": 0.4889, + "step": 19222 + }, + { + "epoch": 0.8511118736371385, + "grad_norm": 1.6883213520050049, + "learning_rate": 1.3212601642491108e-06, + "loss": 0.551, + "step": 19223 + }, + { + "epoch": 0.8511561493419524, + "grad_norm": 1.367172122001648, + "learning_rate": 1.3204925119674416e-06, + "loss": 0.4749, + "step": 19224 + }, + { + "epoch": 0.8512004250467662, + "grad_norm": 1.2300361394882202, + "learning_rate": 1.319725066988742e-06, + "loss": 0.4251, + "step": 19225 + }, + { + "epoch": 0.8512447007515801, + "grad_norm": 1.3282238245010376, + "learning_rate": 1.3189578293313388e-06, + "loss": 0.4317, + "step": 19226 + }, + { + "epoch": 0.851288976456394, + "grad_norm": 1.6478965282440186, + "learning_rate": 1.3181907990135624e-06, + "loss": 0.5233, + "step": 19227 + }, + { + "epoch": 0.8513332521612078, + "grad_norm": 1.638095498085022, + "learning_rate": 1.3174239760537288e-06, + "loss": 0.4806, + "step": 19228 + }, + { + "epoch": 0.8513775278660217, + "grad_norm": 1.3636417388916016, + "learning_rate": 1.3166573604701528e-06, + "loss": 0.4548, + "step": 19229 + }, + { + "epoch": 0.8514218035708356, + "grad_norm": 1.6207917928695679, + "learning_rate": 1.315890952281147e-06, + "loss": 0.5057, + "step": 19230 + }, + { + "epoch": 0.8514660792756494, + "grad_norm": 1.6764512062072754, + "learning_rate": 1.3151247515050148e-06, + "loss": 0.5167, + "step": 19231 + }, + { + "epoch": 0.8515103549804633, + "grad_norm": 1.469716191291809, + "learning_rate": 1.314358758160056e-06, + "loss": 0.4588, + "step": 19232 + }, + { + "epoch": 0.8515546306852773, + "grad_norm": 1.2663159370422363, + "learning_rate": 1.3135929722645669e-06, + "loss": 0.4804, + "step": 19233 + }, + { + "epoch": 0.8515989063900911, + "grad_norm": 1.3882025480270386, + "learning_rate": 1.3128273938368375e-06, + "loss": 0.4984, + "step": 19234 + }, + { + "epoch": 0.851643182094905, + "grad_norm": 1.5016379356384277, + "learning_rate": 1.3120620228951508e-06, + "loss": 0.4967, + "step": 19235 + }, + { + "epoch": 0.8516874577997189, + "grad_norm": 1.2113869190216064, + "learning_rate": 1.311296859457788e-06, + "loss": 0.4223, + "step": 19236 + }, + { + "epoch": 0.8517317335045327, + "grad_norm": 1.4249627590179443, + "learning_rate": 1.3105319035430265e-06, + "loss": 0.4583, + "step": 19237 + }, + { + "epoch": 0.8517760092093466, + "grad_norm": 1.9207823276519775, + "learning_rate": 1.309767155169136e-06, + "loss": 0.5992, + "step": 19238 + }, + { + "epoch": 0.8518202849141605, + "grad_norm": 1.692036509513855, + "learning_rate": 1.3090026143543777e-06, + "loss": 0.4658, + "step": 19239 + }, + { + "epoch": 0.8518645606189743, + "grad_norm": 1.8370606899261475, + "learning_rate": 1.3082382811170192e-06, + "loss": 0.4559, + "step": 19240 + }, + { + "epoch": 0.8519088363237882, + "grad_norm": 1.5533974170684814, + "learning_rate": 1.3074741554753112e-06, + "loss": 0.4587, + "step": 19241 + }, + { + "epoch": 0.8519531120286021, + "grad_norm": 1.9411487579345703, + "learning_rate": 1.306710237447505e-06, + "loss": 0.5172, + "step": 19242 + }, + { + "epoch": 0.8519973877334159, + "grad_norm": 1.6078988313674927, + "learning_rate": 1.3059465270518469e-06, + "loss": 0.4922, + "step": 19243 + }, + { + "epoch": 0.8520416634382298, + "grad_norm": 1.378225326538086, + "learning_rate": 1.3051830243065777e-06, + "loss": 0.5134, + "step": 19244 + }, + { + "epoch": 0.8520859391430438, + "grad_norm": 1.3180601596832275, + "learning_rate": 1.304419729229931e-06, + "loss": 0.4696, + "step": 19245 + }, + { + "epoch": 0.8521302148478576, + "grad_norm": 2.616072654724121, + "learning_rate": 1.3036566418401375e-06, + "loss": 0.57, + "step": 19246 + }, + { + "epoch": 0.8521744905526715, + "grad_norm": 1.5157629251480103, + "learning_rate": 1.3028937621554272e-06, + "loss": 0.5495, + "step": 19247 + }, + { + "epoch": 0.8522187662574854, + "grad_norm": 1.6432194709777832, + "learning_rate": 1.3021310901940166e-06, + "loss": 0.5099, + "step": 19248 + }, + { + "epoch": 0.8522630419622992, + "grad_norm": 1.5931395292282104, + "learning_rate": 1.3013686259741221e-06, + "loss": 0.4815, + "step": 19249 + }, + { + "epoch": 0.8523073176671131, + "grad_norm": 1.6021620035171509, + "learning_rate": 1.3006063695139559e-06, + "loss": 0.4558, + "step": 19250 + }, + { + "epoch": 0.852351593371927, + "grad_norm": 1.2556706666946411, + "learning_rate": 1.2998443208317246e-06, + "loss": 0.4473, + "step": 19251 + }, + { + "epoch": 0.8523958690767409, + "grad_norm": 1.5691449642181396, + "learning_rate": 1.299082479945627e-06, + "loss": 0.5117, + "step": 19252 + }, + { + "epoch": 0.8524401447815547, + "grad_norm": 1.6398524045944214, + "learning_rate": 1.2983208468738595e-06, + "loss": 0.5021, + "step": 19253 + }, + { + "epoch": 0.8524844204863686, + "grad_norm": 1.366604208946228, + "learning_rate": 1.2975594216346144e-06, + "loss": 0.4505, + "step": 19254 + }, + { + "epoch": 0.8525286961911825, + "grad_norm": 1.2657278776168823, + "learning_rate": 1.2967982042460758e-06, + "loss": 0.4268, + "step": 19255 + }, + { + "epoch": 0.8525729718959963, + "grad_norm": 2.1555066108703613, + "learning_rate": 1.296037194726424e-06, + "loss": 0.5424, + "step": 19256 + }, + { + "epoch": 0.8526172476008103, + "grad_norm": 1.4575165510177612, + "learning_rate": 1.2952763930938383e-06, + "loss": 0.5182, + "step": 19257 + }, + { + "epoch": 0.8526615233056242, + "grad_norm": 1.4139025211334229, + "learning_rate": 1.2945157993664892e-06, + "loss": 0.5146, + "step": 19258 + }, + { + "epoch": 0.852705799010438, + "grad_norm": 1.4390089511871338, + "learning_rate": 1.2937554135625397e-06, + "loss": 0.4858, + "step": 19259 + }, + { + "epoch": 0.8527500747152519, + "grad_norm": 1.343935251235962, + "learning_rate": 1.2929952357001574e-06, + "loss": 0.4121, + "step": 19260 + }, + { + "epoch": 0.8527943504200658, + "grad_norm": 2.024763345718384, + "learning_rate": 1.2922352657974923e-06, + "loss": 0.574, + "step": 19261 + }, + { + "epoch": 0.8528386261248796, + "grad_norm": 1.5403674840927124, + "learning_rate": 1.2914755038726955e-06, + "loss": 0.4644, + "step": 19262 + }, + { + "epoch": 0.8528829018296935, + "grad_norm": 1.9749321937561035, + "learning_rate": 1.290715949943917e-06, + "loss": 0.6226, + "step": 19263 + }, + { + "epoch": 0.8529271775345074, + "grad_norm": 1.3653432130813599, + "learning_rate": 1.2899566040292977e-06, + "loss": 0.4684, + "step": 19264 + }, + { + "epoch": 0.8529714532393212, + "grad_norm": 1.9371274709701538, + "learning_rate": 1.2891974661469719e-06, + "loss": 0.4559, + "step": 19265 + }, + { + "epoch": 0.8530157289441351, + "grad_norm": 1.4552453756332397, + "learning_rate": 1.288438536315071e-06, + "loss": 0.46, + "step": 19266 + }, + { + "epoch": 0.853060004648949, + "grad_norm": 1.6913827657699585, + "learning_rate": 1.2876798145517245e-06, + "loss": 0.5256, + "step": 19267 + }, + { + "epoch": 0.8531042803537628, + "grad_norm": 1.6223620176315308, + "learning_rate": 1.2869213008750514e-06, + "loss": 0.4657, + "step": 19268 + }, + { + "epoch": 0.8531485560585768, + "grad_norm": 1.595905065536499, + "learning_rate": 1.2861629953031685e-06, + "loss": 0.4864, + "step": 19269 + }, + { + "epoch": 0.8531928317633907, + "grad_norm": 1.9160412549972534, + "learning_rate": 1.2854048978541867e-06, + "loss": 0.5098, + "step": 19270 + }, + { + "epoch": 0.8532371074682045, + "grad_norm": 1.5496262311935425, + "learning_rate": 1.284647008546215e-06, + "loss": 0.5158, + "step": 19271 + }, + { + "epoch": 0.8532813831730184, + "grad_norm": 2.046505928039551, + "learning_rate": 1.2838893273973495e-06, + "loss": 0.5103, + "step": 19272 + }, + { + "epoch": 0.8533256588778323, + "grad_norm": 1.316649317741394, + "learning_rate": 1.2831318544256931e-06, + "loss": 0.4669, + "step": 19273 + }, + { + "epoch": 0.8533699345826461, + "grad_norm": 1.7589945793151855, + "learning_rate": 1.2823745896493344e-06, + "loss": 0.4765, + "step": 19274 + }, + { + "epoch": 0.85341421028746, + "grad_norm": 1.7213469743728638, + "learning_rate": 1.2816175330863613e-06, + "loss": 0.5383, + "step": 19275 + }, + { + "epoch": 0.8534584859922739, + "grad_norm": 1.5878844261169434, + "learning_rate": 1.2808606847548522e-06, + "loss": 0.5136, + "step": 19276 + }, + { + "epoch": 0.8535027616970877, + "grad_norm": 1.6716217994689941, + "learning_rate": 1.2801040446728886e-06, + "loss": 0.5223, + "step": 19277 + }, + { + "epoch": 0.8535470374019016, + "grad_norm": 2.6243937015533447, + "learning_rate": 1.2793476128585402e-06, + "loss": 0.549, + "step": 19278 + }, + { + "epoch": 0.8535913131067155, + "grad_norm": 1.3923914432525635, + "learning_rate": 1.2785913893298729e-06, + "loss": 0.4764, + "step": 19279 + }, + { + "epoch": 0.8536355888115293, + "grad_norm": 1.3865301609039307, + "learning_rate": 1.2778353741049487e-06, + "loss": 0.4695, + "step": 19280 + }, + { + "epoch": 0.8536798645163433, + "grad_norm": 1.5008724927902222, + "learning_rate": 1.2770795672018254e-06, + "loss": 0.5265, + "step": 19281 + }, + { + "epoch": 0.8537241402211572, + "grad_norm": 1.3528944253921509, + "learning_rate": 1.276323968638553e-06, + "loss": 0.4453, + "step": 19282 + }, + { + "epoch": 0.853768415925971, + "grad_norm": 1.2542610168457031, + "learning_rate": 1.2755685784331784e-06, + "loss": 0.4331, + "step": 19283 + }, + { + "epoch": 0.8538126916307849, + "grad_norm": 1.4200849533081055, + "learning_rate": 1.2748133966037468e-06, + "loss": 0.502, + "step": 19284 + }, + { + "epoch": 0.8538569673355988, + "grad_norm": 1.5095182657241821, + "learning_rate": 1.2740584231682928e-06, + "loss": 0.4735, + "step": 19285 + }, + { + "epoch": 0.8539012430404126, + "grad_norm": 1.5331666469573975, + "learning_rate": 1.2733036581448465e-06, + "loss": 0.5349, + "step": 19286 + }, + { + "epoch": 0.8539455187452265, + "grad_norm": 2.1919126510620117, + "learning_rate": 1.272549101551438e-06, + "loss": 0.5084, + "step": 19287 + }, + { + "epoch": 0.8539897944500404, + "grad_norm": 1.5003461837768555, + "learning_rate": 1.2717947534060894e-06, + "loss": 0.4594, + "step": 19288 + }, + { + "epoch": 0.8540340701548543, + "grad_norm": 1.434433937072754, + "learning_rate": 1.271040613726815e-06, + "loss": 0.5249, + "step": 19289 + }, + { + "epoch": 0.8540783458596681, + "grad_norm": 1.9026070833206177, + "learning_rate": 1.2702866825316295e-06, + "loss": 0.5179, + "step": 19290 + }, + { + "epoch": 0.854122621564482, + "grad_norm": 1.6498572826385498, + "learning_rate": 1.2695329598385375e-06, + "loss": 0.5369, + "step": 19291 + }, + { + "epoch": 0.854166897269296, + "grad_norm": 1.3155049085617065, + "learning_rate": 1.2687794456655423e-06, + "loss": 0.4488, + "step": 19292 + }, + { + "epoch": 0.8542111729741098, + "grad_norm": 1.7597386837005615, + "learning_rate": 1.2680261400306392e-06, + "loss": 0.5379, + "step": 19293 + }, + { + "epoch": 0.8542554486789237, + "grad_norm": 1.785628080368042, + "learning_rate": 1.2672730429518242e-06, + "loss": 0.5855, + "step": 19294 + }, + { + "epoch": 0.8542997243837376, + "grad_norm": 1.4209719896316528, + "learning_rate": 1.2665201544470817e-06, + "loss": 0.4596, + "step": 19295 + }, + { + "epoch": 0.8543440000885514, + "grad_norm": 1.4717942476272583, + "learning_rate": 1.2657674745343918e-06, + "loss": 0.4817, + "step": 19296 + }, + { + "epoch": 0.8543882757933653, + "grad_norm": 1.392752766609192, + "learning_rate": 1.2650150032317366e-06, + "loss": 0.4559, + "step": 19297 + }, + { + "epoch": 0.8544325514981792, + "grad_norm": 1.5169909000396729, + "learning_rate": 1.2642627405570852e-06, + "loss": 0.4912, + "step": 19298 + }, + { + "epoch": 0.854476827202993, + "grad_norm": 1.6258410215377808, + "learning_rate": 1.2635106865284063e-06, + "loss": 0.5115, + "step": 19299 + }, + { + "epoch": 0.8545211029078069, + "grad_norm": 1.6508002281188965, + "learning_rate": 1.2627588411636604e-06, + "loss": 0.4618, + "step": 19300 + }, + { + "epoch": 0.8545653786126208, + "grad_norm": 1.4433224201202393, + "learning_rate": 1.2620072044808062e-06, + "loss": 0.4741, + "step": 19301 + }, + { + "epoch": 0.8546096543174346, + "grad_norm": 1.2911206483840942, + "learning_rate": 1.261255776497794e-06, + "loss": 0.4215, + "step": 19302 + }, + { + "epoch": 0.8546539300222485, + "grad_norm": 1.6797242164611816, + "learning_rate": 1.2605045572325713e-06, + "loss": 0.5269, + "step": 19303 + }, + { + "epoch": 0.8546982057270625, + "grad_norm": 1.347367525100708, + "learning_rate": 1.2597535467030831e-06, + "loss": 0.5097, + "step": 19304 + }, + { + "epoch": 0.8547424814318763, + "grad_norm": 1.700130581855774, + "learning_rate": 1.2590027449272647e-06, + "loss": 0.5381, + "step": 19305 + }, + { + "epoch": 0.8547867571366902, + "grad_norm": 1.6827470064163208, + "learning_rate": 1.2582521519230462e-06, + "loss": 0.5058, + "step": 19306 + }, + { + "epoch": 0.8548310328415041, + "grad_norm": 1.5970088243484497, + "learning_rate": 1.257501767708359e-06, + "loss": 0.5242, + "step": 19307 + }, + { + "epoch": 0.8548753085463179, + "grad_norm": 1.5069743394851685, + "learning_rate": 1.2567515923011265e-06, + "loss": 0.4982, + "step": 19308 + }, + { + "epoch": 0.8549195842511318, + "grad_norm": 1.5483394861221313, + "learning_rate": 1.2560016257192587e-06, + "loss": 0.5037, + "step": 19309 + }, + { + "epoch": 0.8549638599559457, + "grad_norm": 1.3647072315216064, + "learning_rate": 1.2552518679806747e-06, + "loss": 0.4819, + "step": 19310 + }, + { + "epoch": 0.8550081356607595, + "grad_norm": 1.8001198768615723, + "learning_rate": 1.25450231910328e-06, + "loss": 0.5558, + "step": 19311 + }, + { + "epoch": 0.8550524113655734, + "grad_norm": 1.777868390083313, + "learning_rate": 1.2537529791049764e-06, + "loss": 0.4794, + "step": 19312 + }, + { + "epoch": 0.8550966870703873, + "grad_norm": 1.2154099941253662, + "learning_rate": 1.2530038480036589e-06, + "loss": 0.4241, + "step": 19313 + }, + { + "epoch": 0.8551409627752011, + "grad_norm": 1.3879761695861816, + "learning_rate": 1.2522549258172245e-06, + "loss": 0.4562, + "step": 19314 + }, + { + "epoch": 0.855185238480015, + "grad_norm": 1.5106117725372314, + "learning_rate": 1.251506212563559e-06, + "loss": 0.5076, + "step": 19315 + }, + { + "epoch": 0.855229514184829, + "grad_norm": 1.6187275648117065, + "learning_rate": 1.2507577082605438e-06, + "loss": 0.4837, + "step": 19316 + }, + { + "epoch": 0.8552737898896428, + "grad_norm": 1.358505129814148, + "learning_rate": 1.2500094129260577e-06, + "loss": 0.4507, + "step": 19317 + }, + { + "epoch": 0.8553180655944567, + "grad_norm": 1.389603614807129, + "learning_rate": 1.2492613265779708e-06, + "loss": 0.4712, + "step": 19318 + }, + { + "epoch": 0.8553623412992706, + "grad_norm": 1.7203766107559204, + "learning_rate": 1.248513449234151e-06, + "loss": 0.555, + "step": 19319 + }, + { + "epoch": 0.8554066170040844, + "grad_norm": 1.5816872119903564, + "learning_rate": 1.2477657809124632e-06, + "loss": 0.4662, + "step": 19320 + }, + { + "epoch": 0.8554508927088983, + "grad_norm": 1.7491220235824585, + "learning_rate": 1.247018321630764e-06, + "loss": 0.5006, + "step": 19321 + }, + { + "epoch": 0.8554951684137122, + "grad_norm": 2.0975470542907715, + "learning_rate": 1.246271071406905e-06, + "loss": 0.5708, + "step": 19322 + }, + { + "epoch": 0.8555394441185261, + "grad_norm": 1.2818951606750488, + "learning_rate": 1.2455240302587325e-06, + "loss": 0.4607, + "step": 19323 + }, + { + "epoch": 0.8555837198233399, + "grad_norm": 2.716470956802368, + "learning_rate": 1.2447771982040913e-06, + "loss": 0.5764, + "step": 19324 + }, + { + "epoch": 0.8556279955281538, + "grad_norm": 1.628175139427185, + "learning_rate": 1.2440305752608183e-06, + "loss": 0.5115, + "step": 19325 + }, + { + "epoch": 0.8556722712329677, + "grad_norm": 1.4114588499069214, + "learning_rate": 1.2432841614467462e-06, + "loss": 0.489, + "step": 19326 + }, + { + "epoch": 0.8557165469377815, + "grad_norm": 1.7819634675979614, + "learning_rate": 1.2425379567797012e-06, + "loss": 0.4692, + "step": 19327 + }, + { + "epoch": 0.8557608226425955, + "grad_norm": 2.079655647277832, + "learning_rate": 1.2417919612775076e-06, + "loss": 0.5596, + "step": 19328 + }, + { + "epoch": 0.8558050983474094, + "grad_norm": 1.8337160348892212, + "learning_rate": 1.2410461749579782e-06, + "loss": 0.5547, + "step": 19329 + }, + { + "epoch": 0.8558493740522232, + "grad_norm": 1.508028507232666, + "learning_rate": 1.2403005978389327e-06, + "loss": 0.4785, + "step": 19330 + }, + { + "epoch": 0.8558936497570371, + "grad_norm": 1.756256103515625, + "learning_rate": 1.2395552299381742e-06, + "loss": 0.5054, + "step": 19331 + }, + { + "epoch": 0.855937925461851, + "grad_norm": 1.4783308506011963, + "learning_rate": 1.2388100712735051e-06, + "loss": 0.4809, + "step": 19332 + }, + { + "epoch": 0.8559822011666648, + "grad_norm": 2.0131125450134277, + "learning_rate": 1.2380651218627226e-06, + "loss": 0.5448, + "step": 19333 + }, + { + "epoch": 0.8560264768714787, + "grad_norm": 1.57253098487854, + "learning_rate": 1.2373203817236236e-06, + "loss": 0.4777, + "step": 19334 + }, + { + "epoch": 0.8560707525762926, + "grad_norm": 1.4608954191207886, + "learning_rate": 1.236575850873991e-06, + "loss": 0.4555, + "step": 19335 + }, + { + "epoch": 0.8561150282811064, + "grad_norm": 1.213405966758728, + "learning_rate": 1.2358315293316092e-06, + "loss": 0.4179, + "step": 19336 + }, + { + "epoch": 0.8561593039859203, + "grad_norm": 1.5447371006011963, + "learning_rate": 1.2350874171142557e-06, + "loss": 0.4657, + "step": 19337 + }, + { + "epoch": 0.8562035796907342, + "grad_norm": 1.3841831684112549, + "learning_rate": 1.234343514239702e-06, + "loss": 0.4511, + "step": 19338 + }, + { + "epoch": 0.856247855395548, + "grad_norm": 1.5336053371429443, + "learning_rate": 1.2335998207257138e-06, + "loss": 0.4936, + "step": 19339 + }, + { + "epoch": 0.856292131100362, + "grad_norm": 1.624971628189087, + "learning_rate": 1.2328563365900591e-06, + "loss": 0.4978, + "step": 19340 + }, + { + "epoch": 0.8563364068051759, + "grad_norm": 1.9302645921707153, + "learning_rate": 1.2321130618504917e-06, + "loss": 0.5077, + "step": 19341 + }, + { + "epoch": 0.8563806825099897, + "grad_norm": 1.4291871786117554, + "learning_rate": 1.231369996524765e-06, + "loss": 0.4627, + "step": 19342 + }, + { + "epoch": 0.8564249582148036, + "grad_norm": 1.5791305303573608, + "learning_rate": 1.2306271406306236e-06, + "loss": 0.4792, + "step": 19343 + }, + { + "epoch": 0.8564692339196175, + "grad_norm": 1.3891100883483887, + "learning_rate": 1.2298844941858157e-06, + "loss": 0.4651, + "step": 19344 + }, + { + "epoch": 0.8565135096244313, + "grad_norm": 1.3149253129959106, + "learning_rate": 1.2291420572080747e-06, + "loss": 0.4318, + "step": 19345 + }, + { + "epoch": 0.8565577853292452, + "grad_norm": 1.6307005882263184, + "learning_rate": 1.2283998297151345e-06, + "loss": 0.5204, + "step": 19346 + }, + { + "epoch": 0.8566020610340591, + "grad_norm": 1.7110334634780884, + "learning_rate": 1.227657811724723e-06, + "loss": 0.5512, + "step": 19347 + }, + { + "epoch": 0.8566463367388729, + "grad_norm": 1.7415554523468018, + "learning_rate": 1.2269160032545613e-06, + "loss": 0.4745, + "step": 19348 + }, + { + "epoch": 0.8566906124436868, + "grad_norm": 1.5995460748672485, + "learning_rate": 1.2261744043223656e-06, + "loss": 0.5163, + "step": 19349 + }, + { + "epoch": 0.8567348881485007, + "grad_norm": 1.9651747941970825, + "learning_rate": 1.2254330149458527e-06, + "loss": 0.5309, + "step": 19350 + }, + { + "epoch": 0.8567791638533145, + "grad_norm": 1.6683123111724854, + "learning_rate": 1.2246918351427272e-06, + "loss": 0.5357, + "step": 19351 + }, + { + "epoch": 0.8568234395581285, + "grad_norm": 1.5725603103637695, + "learning_rate": 1.223950864930692e-06, + "loss": 0.4793, + "step": 19352 + }, + { + "epoch": 0.8568677152629424, + "grad_norm": 1.571545124053955, + "learning_rate": 1.2232101043274437e-06, + "loss": 0.4981, + "step": 19353 + }, + { + "epoch": 0.8569119909677562, + "grad_norm": 1.4385247230529785, + "learning_rate": 1.2224695533506769e-06, + "loss": 0.5149, + "step": 19354 + }, + { + "epoch": 0.8569562666725701, + "grad_norm": 1.499192237854004, + "learning_rate": 1.22172921201808e-06, + "loss": 0.4767, + "step": 19355 + }, + { + "epoch": 0.857000542377384, + "grad_norm": 1.4947341680526733, + "learning_rate": 1.220989080347329e-06, + "loss": 0.4825, + "step": 19356 + }, + { + "epoch": 0.8570448180821979, + "grad_norm": 1.3953646421432495, + "learning_rate": 1.2202491583561083e-06, + "loss": 0.485, + "step": 19357 + }, + { + "epoch": 0.8570890937870117, + "grad_norm": 1.8652681112289429, + "learning_rate": 1.2195094460620872e-06, + "loss": 0.5129, + "step": 19358 + }, + { + "epoch": 0.8571333694918256, + "grad_norm": 1.331880807876587, + "learning_rate": 1.2187699434829314e-06, + "loss": 0.4354, + "step": 19359 + }, + { + "epoch": 0.8571776451966395, + "grad_norm": 1.7350059747695923, + "learning_rate": 1.2180306506363081e-06, + "loss": 0.5135, + "step": 19360 + }, + { + "epoch": 0.8572219209014533, + "grad_norm": 1.507807731628418, + "learning_rate": 1.2172915675398723e-06, + "loss": 0.506, + "step": 19361 + }, + { + "epoch": 0.8572661966062672, + "grad_norm": 1.6719681024551392, + "learning_rate": 1.216552694211276e-06, + "loss": 0.5124, + "step": 19362 + }, + { + "epoch": 0.8573104723110812, + "grad_norm": 1.3263697624206543, + "learning_rate": 1.2158140306681654e-06, + "loss": 0.4524, + "step": 19363 + }, + { + "epoch": 0.857354748015895, + "grad_norm": 1.5136520862579346, + "learning_rate": 1.2150755769281875e-06, + "loss": 0.4441, + "step": 19364 + }, + { + "epoch": 0.8573990237207089, + "grad_norm": 1.6322288513183594, + "learning_rate": 1.2143373330089747e-06, + "loss": 0.5401, + "step": 19365 + }, + { + "epoch": 0.8574432994255228, + "grad_norm": 1.5710598230361938, + "learning_rate": 1.2135992989281598e-06, + "loss": 0.5349, + "step": 19366 + }, + { + "epoch": 0.8574875751303366, + "grad_norm": 1.6516690254211426, + "learning_rate": 1.2128614747033728e-06, + "loss": 0.4865, + "step": 19367 + }, + { + "epoch": 0.8575318508351505, + "grad_norm": 1.7137622833251953, + "learning_rate": 1.2121238603522346e-06, + "loss": 0.5452, + "step": 19368 + }, + { + "epoch": 0.8575761265399644, + "grad_norm": 1.8728032112121582, + "learning_rate": 1.2113864558923605e-06, + "loss": 0.5344, + "step": 19369 + }, + { + "epoch": 0.8576204022447782, + "grad_norm": 1.321736454963684, + "learning_rate": 1.2106492613413678e-06, + "loss": 0.4585, + "step": 19370 + }, + { + "epoch": 0.8576646779495921, + "grad_norm": 1.3576332330703735, + "learning_rate": 1.2099122767168602e-06, + "loss": 0.4524, + "step": 19371 + }, + { + "epoch": 0.857708953654406, + "grad_norm": 1.6313142776489258, + "learning_rate": 1.2091755020364415e-06, + "loss": 0.5547, + "step": 19372 + }, + { + "epoch": 0.8577532293592198, + "grad_norm": 1.3415346145629883, + "learning_rate": 1.2084389373177075e-06, + "loss": 0.463, + "step": 19373 + }, + { + "epoch": 0.8577975050640337, + "grad_norm": 1.3935500383377075, + "learning_rate": 1.2077025825782518e-06, + "loss": 0.4833, + "step": 19374 + }, + { + "epoch": 0.8578417807688477, + "grad_norm": 1.5509980916976929, + "learning_rate": 1.2069664378356604e-06, + "loss": 0.4414, + "step": 19375 + }, + { + "epoch": 0.8578860564736615, + "grad_norm": 1.3888130187988281, + "learning_rate": 1.2062305031075139e-06, + "loss": 0.4544, + "step": 19376 + }, + { + "epoch": 0.8579303321784754, + "grad_norm": 1.5027415752410889, + "learning_rate": 1.2054947784113936e-06, + "loss": 0.4846, + "step": 19377 + }, + { + "epoch": 0.8579746078832893, + "grad_norm": 2.012723445892334, + "learning_rate": 1.2047592637648708e-06, + "loss": 0.5656, + "step": 19378 + }, + { + "epoch": 0.8580188835881031, + "grad_norm": 1.6168628931045532, + "learning_rate": 1.2040239591855085e-06, + "loss": 0.5146, + "step": 19379 + }, + { + "epoch": 0.858063159292917, + "grad_norm": 1.585843563079834, + "learning_rate": 1.2032888646908746e-06, + "loss": 0.4956, + "step": 19380 + }, + { + "epoch": 0.8581074349977309, + "grad_norm": 2.5693094730377197, + "learning_rate": 1.202553980298523e-06, + "loss": 0.5289, + "step": 19381 + }, + { + "epoch": 0.8581517107025447, + "grad_norm": 1.309095859527588, + "learning_rate": 1.2018193060260075e-06, + "loss": 0.432, + "step": 19382 + }, + { + "epoch": 0.8581959864073586, + "grad_norm": 1.5495131015777588, + "learning_rate": 1.2010848418908726e-06, + "loss": 0.515, + "step": 19383 + }, + { + "epoch": 0.8582402621121725, + "grad_norm": 1.8649859428405762, + "learning_rate": 1.2003505879106635e-06, + "loss": 0.5217, + "step": 19384 + }, + { + "epoch": 0.8582845378169863, + "grad_norm": 1.4194260835647583, + "learning_rate": 1.1996165441029151e-06, + "loss": 0.4623, + "step": 19385 + }, + { + "epoch": 0.8583288135218002, + "grad_norm": 1.2630048990249634, + "learning_rate": 1.1988827104851574e-06, + "loss": 0.4206, + "step": 19386 + }, + { + "epoch": 0.8583730892266141, + "grad_norm": 1.6864917278289795, + "learning_rate": 1.1981490870749225e-06, + "loss": 0.5316, + "step": 19387 + }, + { + "epoch": 0.858417364931428, + "grad_norm": 1.5439571142196655, + "learning_rate": 1.1974156738897301e-06, + "loss": 0.4942, + "step": 19388 + }, + { + "epoch": 0.8584616406362419, + "grad_norm": 2.2059383392333984, + "learning_rate": 1.196682470947097e-06, + "loss": 0.5449, + "step": 19389 + }, + { + "epoch": 0.8585059163410558, + "grad_norm": 1.2682759761810303, + "learning_rate": 1.195949478264533e-06, + "loss": 0.3894, + "step": 19390 + }, + { + "epoch": 0.8585501920458696, + "grad_norm": 1.3632947206497192, + "learning_rate": 1.195216695859549e-06, + "loss": 0.4125, + "step": 19391 + }, + { + "epoch": 0.8585944677506835, + "grad_norm": 1.7023208141326904, + "learning_rate": 1.1944841237496452e-06, + "loss": 0.5132, + "step": 19392 + }, + { + "epoch": 0.8586387434554974, + "grad_norm": 1.6028258800506592, + "learning_rate": 1.1937517619523187e-06, + "loss": 0.5125, + "step": 19393 + }, + { + "epoch": 0.8586830191603113, + "grad_norm": 1.3342089653015137, + "learning_rate": 1.193019610485061e-06, + "loss": 0.4749, + "step": 19394 + }, + { + "epoch": 0.8587272948651251, + "grad_norm": 1.7462049722671509, + "learning_rate": 1.1922876693653584e-06, + "loss": 0.464, + "step": 19395 + }, + { + "epoch": 0.858771570569939, + "grad_norm": 1.4257354736328125, + "learning_rate": 1.1915559386106923e-06, + "loss": 0.488, + "step": 19396 + }, + { + "epoch": 0.8588158462747529, + "grad_norm": 1.8309928178787231, + "learning_rate": 1.190824418238542e-06, + "loss": 0.5237, + "step": 19397 + }, + { + "epoch": 0.8588601219795667, + "grad_norm": 1.3961615562438965, + "learning_rate": 1.1900931082663769e-06, + "loss": 0.4619, + "step": 19398 + }, + { + "epoch": 0.8589043976843806, + "grad_norm": 1.4352543354034424, + "learning_rate": 1.1893620087116663e-06, + "loss": 0.4429, + "step": 19399 + }, + { + "epoch": 0.8589486733891946, + "grad_norm": 1.7265896797180176, + "learning_rate": 1.1886311195918664e-06, + "loss": 0.5452, + "step": 19400 + }, + { + "epoch": 0.8589929490940084, + "grad_norm": 1.481382966041565, + "learning_rate": 1.187900440924441e-06, + "loss": 0.5091, + "step": 19401 + }, + { + "epoch": 0.8590372247988223, + "grad_norm": 1.5971062183380127, + "learning_rate": 1.1871699727268383e-06, + "loss": 0.493, + "step": 19402 + }, + { + "epoch": 0.8590815005036362, + "grad_norm": 1.5382364988327026, + "learning_rate": 1.1864397150165052e-06, + "loss": 0.4964, + "step": 19403 + }, + { + "epoch": 0.85912577620845, + "grad_norm": 1.544252634048462, + "learning_rate": 1.1857096678108826e-06, + "loss": 0.6069, + "step": 19404 + }, + { + "epoch": 0.8591700519132639, + "grad_norm": 1.650821328163147, + "learning_rate": 1.1849798311274085e-06, + "loss": 0.495, + "step": 19405 + }, + { + "epoch": 0.8592143276180778, + "grad_norm": 1.3571707010269165, + "learning_rate": 1.1842502049835114e-06, + "loss": 0.4377, + "step": 19406 + }, + { + "epoch": 0.8592586033228916, + "grad_norm": 1.6444023847579956, + "learning_rate": 1.1835207893966216e-06, + "loss": 0.5231, + "step": 19407 + }, + { + "epoch": 0.8593028790277055, + "grad_norm": 1.6037203073501587, + "learning_rate": 1.1827915843841598e-06, + "loss": 0.4902, + "step": 19408 + }, + { + "epoch": 0.8593471547325194, + "grad_norm": 1.581009864807129, + "learning_rate": 1.182062589963542e-06, + "loss": 0.5098, + "step": 19409 + }, + { + "epoch": 0.8593914304373332, + "grad_norm": 1.5411700010299683, + "learning_rate": 1.1813338061521763e-06, + "loss": 0.4299, + "step": 19410 + }, + { + "epoch": 0.8594357061421471, + "grad_norm": 1.4207675457000732, + "learning_rate": 1.1806052329674777e-06, + "loss": 0.4921, + "step": 19411 + }, + { + "epoch": 0.8594799818469611, + "grad_norm": 1.716827630996704, + "learning_rate": 1.1798768704268393e-06, + "loss": 0.5073, + "step": 19412 + }, + { + "epoch": 0.8595242575517749, + "grad_norm": 1.6457651853561401, + "learning_rate": 1.179148718547658e-06, + "loss": 0.505, + "step": 19413 + }, + { + "epoch": 0.8595685332565888, + "grad_norm": 1.6641056537628174, + "learning_rate": 1.178420777347329e-06, + "loss": 0.5212, + "step": 19414 + }, + { + "epoch": 0.8596128089614027, + "grad_norm": 1.3616254329681396, + "learning_rate": 1.177693046843238e-06, + "loss": 0.4642, + "step": 19415 + }, + { + "epoch": 0.8596570846662165, + "grad_norm": 1.3415172100067139, + "learning_rate": 1.1769655270527624e-06, + "loss": 0.4659, + "step": 19416 + }, + { + "epoch": 0.8597013603710304, + "grad_norm": 1.6170319318771362, + "learning_rate": 1.176238217993283e-06, + "loss": 0.4969, + "step": 19417 + }, + { + "epoch": 0.8597456360758443, + "grad_norm": 1.5454787015914917, + "learning_rate": 1.17551111968217e-06, + "loss": 0.4758, + "step": 19418 + }, + { + "epoch": 0.8597899117806581, + "grad_norm": 1.764750599861145, + "learning_rate": 1.1747842321367886e-06, + "loss": 0.5052, + "step": 19419 + }, + { + "epoch": 0.859834187485472, + "grad_norm": 1.5261104106903076, + "learning_rate": 1.1740575553744993e-06, + "loss": 0.4454, + "step": 19420 + }, + { + "epoch": 0.8598784631902859, + "grad_norm": 1.3919258117675781, + "learning_rate": 1.1733310894126593e-06, + "loss": 0.4588, + "step": 19421 + }, + { + "epoch": 0.8599227388950997, + "grad_norm": 1.935266137123108, + "learning_rate": 1.1726048342686202e-06, + "loss": 0.5939, + "step": 19422 + }, + { + "epoch": 0.8599670145999136, + "grad_norm": 1.2432453632354736, + "learning_rate": 1.1718787899597239e-06, + "loss": 0.4163, + "step": 19423 + }, + { + "epoch": 0.8600112903047276, + "grad_norm": 1.2816972732543945, + "learning_rate": 1.1711529565033175e-06, + "loss": 0.4761, + "step": 19424 + }, + { + "epoch": 0.8600555660095414, + "grad_norm": 1.4757294654846191, + "learning_rate": 1.1704273339167338e-06, + "loss": 0.4832, + "step": 19425 + }, + { + "epoch": 0.8600998417143553, + "grad_norm": 1.5454260110855103, + "learning_rate": 1.1697019222173012e-06, + "loss": 0.4924, + "step": 19426 + }, + { + "epoch": 0.8601441174191692, + "grad_norm": 1.585007905960083, + "learning_rate": 1.1689767214223502e-06, + "loss": 0.4907, + "step": 19427 + }, + { + "epoch": 0.8601883931239831, + "grad_norm": 1.4331485033035278, + "learning_rate": 1.1682517315492003e-06, + "loss": 0.45, + "step": 19428 + }, + { + "epoch": 0.8602326688287969, + "grad_norm": 2.0085721015930176, + "learning_rate": 1.1675269526151678e-06, + "loss": 0.5633, + "step": 19429 + }, + { + "epoch": 0.8602769445336108, + "grad_norm": 1.8710788488388062, + "learning_rate": 1.1668023846375608e-06, + "loss": 0.5704, + "step": 19430 + }, + { + "epoch": 0.8603212202384247, + "grad_norm": 1.6113225221633911, + "learning_rate": 1.166078027633687e-06, + "loss": 0.5346, + "step": 19431 + }, + { + "epoch": 0.8603654959432385, + "grad_norm": 1.222690224647522, + "learning_rate": 1.1653538816208465e-06, + "loss": 0.4496, + "step": 19432 + }, + { + "epoch": 0.8604097716480524, + "grad_norm": 1.6279051303863525, + "learning_rate": 1.1646299466163336e-06, + "loss": 0.4817, + "step": 19433 + }, + { + "epoch": 0.8604540473528663, + "grad_norm": 1.8161967992782593, + "learning_rate": 1.163906222637442e-06, + "loss": 0.4654, + "step": 19434 + }, + { + "epoch": 0.8604983230576801, + "grad_norm": 1.434716820716858, + "learning_rate": 1.163182709701456e-06, + "loss": 0.444, + "step": 19435 + }, + { + "epoch": 0.8605425987624941, + "grad_norm": 1.6369167566299438, + "learning_rate": 1.1624594078256535e-06, + "loss": 0.4512, + "step": 19436 + }, + { + "epoch": 0.860586874467308, + "grad_norm": 1.7373398542404175, + "learning_rate": 1.1617363170273144e-06, + "loss": 0.5043, + "step": 19437 + }, + { + "epoch": 0.8606311501721218, + "grad_norm": 1.3235688209533691, + "learning_rate": 1.161013437323707e-06, + "loss": 0.4532, + "step": 19438 + }, + { + "epoch": 0.8606754258769357, + "grad_norm": 1.4854658842086792, + "learning_rate": 1.1602907687320964e-06, + "loss": 0.4874, + "step": 19439 + }, + { + "epoch": 0.8607197015817496, + "grad_norm": 1.683218240737915, + "learning_rate": 1.1595683112697432e-06, + "loss": 0.4676, + "step": 19440 + }, + { + "epoch": 0.8607639772865634, + "grad_norm": 1.4199092388153076, + "learning_rate": 1.1588460649539036e-06, + "loss": 0.4302, + "step": 19441 + }, + { + "epoch": 0.8608082529913773, + "grad_norm": 1.8153966665267944, + "learning_rate": 1.158124029801826e-06, + "loss": 0.5583, + "step": 19442 + }, + { + "epoch": 0.8608525286961912, + "grad_norm": 1.5867410898208618, + "learning_rate": 1.1574022058307555e-06, + "loss": 0.4551, + "step": 19443 + }, + { + "epoch": 0.860896804401005, + "grad_norm": 1.8864918947219849, + "learning_rate": 1.156680593057935e-06, + "loss": 0.5316, + "step": 19444 + }, + { + "epoch": 0.8609410801058189, + "grad_norm": 1.5966991186141968, + "learning_rate": 1.1559591915005974e-06, + "loss": 0.4708, + "step": 19445 + }, + { + "epoch": 0.8609853558106328, + "grad_norm": 1.8548345565795898, + "learning_rate": 1.1552380011759722e-06, + "loss": 0.5038, + "step": 19446 + }, + { + "epoch": 0.8610296315154466, + "grad_norm": 1.7412775754928589, + "learning_rate": 1.154517022101287e-06, + "loss": 0.5239, + "step": 19447 + }, + { + "epoch": 0.8610739072202606, + "grad_norm": 1.4215359687805176, + "learning_rate": 1.1537962542937608e-06, + "loss": 0.4835, + "step": 19448 + }, + { + "epoch": 0.8611181829250745, + "grad_norm": 1.440089464187622, + "learning_rate": 1.1530756977706082e-06, + "loss": 0.4589, + "step": 19449 + }, + { + "epoch": 0.8611624586298883, + "grad_norm": 1.5673364400863647, + "learning_rate": 1.1523553525490393e-06, + "loss": 0.4841, + "step": 19450 + }, + { + "epoch": 0.8612067343347022, + "grad_norm": 1.4655221700668335, + "learning_rate": 1.1516352186462588e-06, + "loss": 0.474, + "step": 19451 + }, + { + "epoch": 0.8612510100395161, + "grad_norm": 1.7064944505691528, + "learning_rate": 1.1509152960794668e-06, + "loss": 0.4875, + "step": 19452 + }, + { + "epoch": 0.8612952857443299, + "grad_norm": 1.3734009265899658, + "learning_rate": 1.1501955848658552e-06, + "loss": 0.4643, + "step": 19453 + }, + { + "epoch": 0.8613395614491438, + "grad_norm": 1.7134112119674683, + "learning_rate": 1.1494760850226182e-06, + "loss": 0.5014, + "step": 19454 + }, + { + "epoch": 0.8613838371539577, + "grad_norm": 1.7904993295669556, + "learning_rate": 1.1487567965669387e-06, + "loss": 0.5234, + "step": 19455 + }, + { + "epoch": 0.8614281128587715, + "grad_norm": 1.9239144325256348, + "learning_rate": 1.1480377195159942e-06, + "loss": 0.5491, + "step": 19456 + }, + { + "epoch": 0.8614723885635854, + "grad_norm": 1.700751781463623, + "learning_rate": 1.1473188538869639e-06, + "loss": 0.527, + "step": 19457 + }, + { + "epoch": 0.8615166642683993, + "grad_norm": 1.550248384475708, + "learning_rate": 1.1466001996970145e-06, + "loss": 0.495, + "step": 19458 + }, + { + "epoch": 0.8615609399732131, + "grad_norm": 2.4899985790252686, + "learning_rate": 1.1458817569633117e-06, + "loss": 0.5429, + "step": 19459 + }, + { + "epoch": 0.8616052156780271, + "grad_norm": 1.935835599899292, + "learning_rate": 1.1451635257030102e-06, + "loss": 0.561, + "step": 19460 + }, + { + "epoch": 0.861649491382841, + "grad_norm": 1.6493228673934937, + "learning_rate": 1.1444455059332704e-06, + "loss": 0.5192, + "step": 19461 + }, + { + "epoch": 0.8616937670876548, + "grad_norm": 1.5591529607772827, + "learning_rate": 1.1437276976712386e-06, + "loss": 0.5181, + "step": 19462 + }, + { + "epoch": 0.8617380427924687, + "grad_norm": 1.3808430433273315, + "learning_rate": 1.1430101009340576e-06, + "loss": 0.4426, + "step": 19463 + }, + { + "epoch": 0.8617823184972826, + "grad_norm": 1.85641610622406, + "learning_rate": 1.142292715738872e-06, + "loss": 0.5797, + "step": 19464 + }, + { + "epoch": 0.8618265942020965, + "grad_norm": 1.3136167526245117, + "learning_rate": 1.1415755421028118e-06, + "loss": 0.418, + "step": 19465 + }, + { + "epoch": 0.8618708699069103, + "grad_norm": 2.368633270263672, + "learning_rate": 1.140858580043005e-06, + "loss": 0.6366, + "step": 19466 + }, + { + "epoch": 0.8619151456117242, + "grad_norm": 1.4749844074249268, + "learning_rate": 1.140141829576582e-06, + "loss": 0.4947, + "step": 19467 + }, + { + "epoch": 0.8619594213165381, + "grad_norm": 1.8497719764709473, + "learning_rate": 1.1394252907206549e-06, + "loss": 0.4959, + "step": 19468 + }, + { + "epoch": 0.8620036970213519, + "grad_norm": 1.6685922145843506, + "learning_rate": 1.1387089634923398e-06, + "loss": 0.4952, + "step": 19469 + }, + { + "epoch": 0.8620479727261658, + "grad_norm": 1.3465180397033691, + "learning_rate": 1.1379928479087454e-06, + "loss": 0.4809, + "step": 19470 + }, + { + "epoch": 0.8620922484309798, + "grad_norm": 1.531295657157898, + "learning_rate": 1.1372769439869768e-06, + "loss": 0.4453, + "step": 19471 + }, + { + "epoch": 0.8621365241357936, + "grad_norm": 1.5785901546478271, + "learning_rate": 1.1365612517441326e-06, + "loss": 0.4933, + "step": 19472 + }, + { + "epoch": 0.8621807998406075, + "grad_norm": 1.442188024520874, + "learning_rate": 1.1358457711973037e-06, + "loss": 0.4979, + "step": 19473 + }, + { + "epoch": 0.8622250755454214, + "grad_norm": 1.311927080154419, + "learning_rate": 1.1351305023635839e-06, + "loss": 0.4133, + "step": 19474 + }, + { + "epoch": 0.8622693512502352, + "grad_norm": 1.5150798559188843, + "learning_rate": 1.134415445260053e-06, + "loss": 0.4498, + "step": 19475 + }, + { + "epoch": 0.8623136269550491, + "grad_norm": 1.6598236560821533, + "learning_rate": 1.133700599903791e-06, + "loss": 0.5368, + "step": 19476 + }, + { + "epoch": 0.862357902659863, + "grad_norm": 1.4586981534957886, + "learning_rate": 1.1329859663118715e-06, + "loss": 0.4472, + "step": 19477 + }, + { + "epoch": 0.8624021783646768, + "grad_norm": 1.5302636623382568, + "learning_rate": 1.132271544501362e-06, + "loss": 0.4576, + "step": 19478 + }, + { + "epoch": 0.8624464540694907, + "grad_norm": 1.5503520965576172, + "learning_rate": 1.131557334489326e-06, + "loss": 0.4376, + "step": 19479 + }, + { + "epoch": 0.8624907297743046, + "grad_norm": 1.683632254600525, + "learning_rate": 1.1308433362928207e-06, + "loss": 0.5356, + "step": 19480 + }, + { + "epoch": 0.8625350054791184, + "grad_norm": 1.8779644966125488, + "learning_rate": 1.1301295499289022e-06, + "loss": 0.5281, + "step": 19481 + }, + { + "epoch": 0.8625792811839323, + "grad_norm": 1.6002246141433716, + "learning_rate": 1.1294159754146173e-06, + "loss": 0.5738, + "step": 19482 + }, + { + "epoch": 0.8626235568887463, + "grad_norm": 1.524826169013977, + "learning_rate": 1.1287026127670076e-06, + "loss": 0.5072, + "step": 19483 + }, + { + "epoch": 0.8626678325935601, + "grad_norm": 1.3625342845916748, + "learning_rate": 1.127989462003114e-06, + "loss": 0.4794, + "step": 19484 + }, + { + "epoch": 0.862712108298374, + "grad_norm": 1.2913850545883179, + "learning_rate": 1.1272765231399684e-06, + "loss": 0.4631, + "step": 19485 + }, + { + "epoch": 0.8627563840031879, + "grad_norm": 1.5332331657409668, + "learning_rate": 1.1265637961945996e-06, + "loss": 0.5349, + "step": 19486 + }, + { + "epoch": 0.8628006597080017, + "grad_norm": 1.457705020904541, + "learning_rate": 1.1258512811840284e-06, + "loss": 0.4546, + "step": 19487 + }, + { + "epoch": 0.8628449354128156, + "grad_norm": 1.684584379196167, + "learning_rate": 1.1251389781252742e-06, + "loss": 0.5167, + "step": 19488 + }, + { + "epoch": 0.8628892111176295, + "grad_norm": 1.64174485206604, + "learning_rate": 1.1244268870353492e-06, + "loss": 0.4864, + "step": 19489 + }, + { + "epoch": 0.8629334868224433, + "grad_norm": 1.553095817565918, + "learning_rate": 1.1237150079312586e-06, + "loss": 0.511, + "step": 19490 + }, + { + "epoch": 0.8629777625272572, + "grad_norm": 1.677527904510498, + "learning_rate": 1.1230033408300111e-06, + "loss": 0.4753, + "step": 19491 + }, + { + "epoch": 0.8630220382320711, + "grad_norm": 1.613604187965393, + "learning_rate": 1.1222918857485998e-06, + "loss": 0.4938, + "step": 19492 + }, + { + "epoch": 0.8630663139368849, + "grad_norm": 1.6128222942352295, + "learning_rate": 1.1215806427040166e-06, + "loss": 0.5591, + "step": 19493 + }, + { + "epoch": 0.8631105896416988, + "grad_norm": 1.3677328824996948, + "learning_rate": 1.1208696117132533e-06, + "loss": 0.4638, + "step": 19494 + }, + { + "epoch": 0.8631548653465128, + "grad_norm": 1.3704098463058472, + "learning_rate": 1.1201587927932888e-06, + "loss": 0.499, + "step": 19495 + }, + { + "epoch": 0.8631991410513266, + "grad_norm": 1.7161723375320435, + "learning_rate": 1.1194481859611016e-06, + "loss": 0.5397, + "step": 19496 + }, + { + "epoch": 0.8632434167561405, + "grad_norm": 1.230127215385437, + "learning_rate": 1.118737791233664e-06, + "loss": 0.406, + "step": 19497 + }, + { + "epoch": 0.8632876924609544, + "grad_norm": 1.521323800086975, + "learning_rate": 1.118027608627943e-06, + "loss": 0.4611, + "step": 19498 + }, + { + "epoch": 0.8633319681657683, + "grad_norm": 2.097541570663452, + "learning_rate": 1.1173176381608997e-06, + "loss": 0.5551, + "step": 19499 + }, + { + "epoch": 0.8633762438705821, + "grad_norm": 1.933227777481079, + "learning_rate": 1.1166078798494905e-06, + "loss": 0.5153, + "step": 19500 + }, + { + "epoch": 0.863420519575396, + "grad_norm": 1.7205275297164917, + "learning_rate": 1.115898333710671e-06, + "loss": 0.5008, + "step": 19501 + }, + { + "epoch": 0.8634647952802099, + "grad_norm": 1.647861123085022, + "learning_rate": 1.1151889997613862e-06, + "loss": 0.4611, + "step": 19502 + }, + { + "epoch": 0.8635090709850237, + "grad_norm": 1.3643147945404053, + "learning_rate": 1.114479878018575e-06, + "loss": 0.465, + "step": 19503 + }, + { + "epoch": 0.8635533466898376, + "grad_norm": 1.2488114833831787, + "learning_rate": 1.1137709684991804e-06, + "loss": 0.4161, + "step": 19504 + }, + { + "epoch": 0.8635976223946515, + "grad_norm": 1.6046584844589233, + "learning_rate": 1.11306227122013e-06, + "loss": 0.478, + "step": 19505 + }, + { + "epoch": 0.8636418980994653, + "grad_norm": 1.4483977556228638, + "learning_rate": 1.1123537861983535e-06, + "loss": 0.5047, + "step": 19506 + }, + { + "epoch": 0.8636861738042793, + "grad_norm": 1.88002610206604, + "learning_rate": 1.1116455134507665e-06, + "loss": 0.5502, + "step": 19507 + }, + { + "epoch": 0.8637304495090932, + "grad_norm": 1.5051496028900146, + "learning_rate": 1.1109374529942906e-06, + "loss": 0.4343, + "step": 19508 + }, + { + "epoch": 0.863774725213907, + "grad_norm": 1.8713319301605225, + "learning_rate": 1.110229604845836e-06, + "loss": 0.5024, + "step": 19509 + }, + { + "epoch": 0.8638190009187209, + "grad_norm": 1.6430513858795166, + "learning_rate": 1.109521969022307e-06, + "loss": 0.5527, + "step": 19510 + }, + { + "epoch": 0.8638632766235348, + "grad_norm": 1.6068390607833862, + "learning_rate": 1.1088145455406097e-06, + "loss": 0.4765, + "step": 19511 + }, + { + "epoch": 0.8639075523283486, + "grad_norm": 2.2061269283294678, + "learning_rate": 1.1081073344176363e-06, + "loss": 0.5212, + "step": 19512 + }, + { + "epoch": 0.8639518280331625, + "grad_norm": 1.665560007095337, + "learning_rate": 1.1074003356702777e-06, + "loss": 0.4924, + "step": 19513 + }, + { + "epoch": 0.8639961037379764, + "grad_norm": 1.5740021467208862, + "learning_rate": 1.106693549315423e-06, + "loss": 0.5528, + "step": 19514 + }, + { + "epoch": 0.8640403794427902, + "grad_norm": 1.523057460784912, + "learning_rate": 1.1059869753699547e-06, + "loss": 0.5292, + "step": 19515 + }, + { + "epoch": 0.8640846551476041, + "grad_norm": 1.3894537687301636, + "learning_rate": 1.1052806138507433e-06, + "loss": 0.4538, + "step": 19516 + }, + { + "epoch": 0.864128930852418, + "grad_norm": 1.4292691946029663, + "learning_rate": 1.1045744647746592e-06, + "loss": 0.4169, + "step": 19517 + }, + { + "epoch": 0.8641732065572318, + "grad_norm": 1.4863464832305908, + "learning_rate": 1.1038685281585738e-06, + "loss": 0.5267, + "step": 19518 + }, + { + "epoch": 0.8642174822620458, + "grad_norm": 1.64725923538208, + "learning_rate": 1.1031628040193454e-06, + "loss": 0.4965, + "step": 19519 + }, + { + "epoch": 0.8642617579668597, + "grad_norm": 1.4909013509750366, + "learning_rate": 1.1024572923738275e-06, + "loss": 0.4509, + "step": 19520 + }, + { + "epoch": 0.8643060336716735, + "grad_norm": 2.0304744243621826, + "learning_rate": 1.1017519932388743e-06, + "loss": 0.5226, + "step": 19521 + }, + { + "epoch": 0.8643503093764874, + "grad_norm": 1.498854160308838, + "learning_rate": 1.101046906631329e-06, + "loss": 0.459, + "step": 19522 + }, + { + "epoch": 0.8643945850813013, + "grad_norm": 1.9837318658828735, + "learning_rate": 1.1003420325680324e-06, + "loss": 0.5403, + "step": 19523 + }, + { + "epoch": 0.8644388607861151, + "grad_norm": 1.6056405305862427, + "learning_rate": 1.0996373710658203e-06, + "loss": 0.4815, + "step": 19524 + }, + { + "epoch": 0.864483136490929, + "grad_norm": 1.41423499584198, + "learning_rate": 1.0989329221415235e-06, + "loss": 0.4588, + "step": 19525 + }, + { + "epoch": 0.8645274121957429, + "grad_norm": 1.9371594190597534, + "learning_rate": 1.098228685811965e-06, + "loss": 0.5505, + "step": 19526 + }, + { + "epoch": 0.8645716879005567, + "grad_norm": 1.531606912612915, + "learning_rate": 1.097524662093965e-06, + "loss": 0.5932, + "step": 19527 + }, + { + "epoch": 0.8646159636053706, + "grad_norm": 1.4346284866333008, + "learning_rate": 1.096820851004342e-06, + "loss": 0.4712, + "step": 19528 + }, + { + "epoch": 0.8646602393101845, + "grad_norm": 1.5076669454574585, + "learning_rate": 1.0961172525599028e-06, + "loss": 0.5074, + "step": 19529 + }, + { + "epoch": 0.8647045150149983, + "grad_norm": 1.7488679885864258, + "learning_rate": 1.0954138667774516e-06, + "loss": 0.4918, + "step": 19530 + }, + { + "epoch": 0.8647487907198123, + "grad_norm": 1.2663326263427734, + "learning_rate": 1.0947106936737927e-06, + "loss": 0.4314, + "step": 19531 + }, + { + "epoch": 0.8647930664246262, + "grad_norm": 1.3889483213424683, + "learning_rate": 1.0940077332657172e-06, + "loss": 0.4256, + "step": 19532 + }, + { + "epoch": 0.86483734212944, + "grad_norm": 1.9246751070022583, + "learning_rate": 1.093304985570015e-06, + "loss": 0.57, + "step": 19533 + }, + { + "epoch": 0.8648816178342539, + "grad_norm": 1.5684223175048828, + "learning_rate": 1.0926024506034715e-06, + "loss": 0.4785, + "step": 19534 + }, + { + "epoch": 0.8649258935390678, + "grad_norm": 1.5102128982543945, + "learning_rate": 1.0919001283828666e-06, + "loss": 0.456, + "step": 19535 + }, + { + "epoch": 0.8649701692438817, + "grad_norm": 1.530094861984253, + "learning_rate": 1.0911980189249727e-06, + "loss": 0.4965, + "step": 19536 + }, + { + "epoch": 0.8650144449486955, + "grad_norm": 1.4361528158187866, + "learning_rate": 1.0904961222465593e-06, + "loss": 0.4789, + "step": 19537 + }, + { + "epoch": 0.8650587206535094, + "grad_norm": 1.4406238794326782, + "learning_rate": 1.0897944383643932e-06, + "loss": 0.4973, + "step": 19538 + }, + { + "epoch": 0.8651029963583233, + "grad_norm": 1.2804231643676758, + "learning_rate": 1.089092967295231e-06, + "loss": 0.4258, + "step": 19539 + }, + { + "epoch": 0.8651472720631371, + "grad_norm": 1.7342913150787354, + "learning_rate": 1.0883917090558272e-06, + "loss": 0.5389, + "step": 19540 + }, + { + "epoch": 0.865191547767951, + "grad_norm": 2.8236114978790283, + "learning_rate": 1.0876906636629326e-06, + "loss": 0.5999, + "step": 19541 + }, + { + "epoch": 0.865235823472765, + "grad_norm": 1.3120067119598389, + "learning_rate": 1.0869898311332906e-06, + "loss": 0.4818, + "step": 19542 + }, + { + "epoch": 0.8652800991775788, + "grad_norm": 1.5718564987182617, + "learning_rate": 1.0862892114836387e-06, + "loss": 0.5095, + "step": 19543 + }, + { + "epoch": 0.8653243748823927, + "grad_norm": 1.3352689743041992, + "learning_rate": 1.0855888047307106e-06, + "loss": 0.4008, + "step": 19544 + }, + { + "epoch": 0.8653686505872066, + "grad_norm": 1.5943156480789185, + "learning_rate": 1.084888610891236e-06, + "loss": 0.5417, + "step": 19545 + }, + { + "epoch": 0.8654129262920204, + "grad_norm": 1.9889216423034668, + "learning_rate": 1.0841886299819383e-06, + "loss": 0.6139, + "step": 19546 + }, + { + "epoch": 0.8654572019968343, + "grad_norm": 1.9673943519592285, + "learning_rate": 1.0834888620195327e-06, + "loss": 0.5837, + "step": 19547 + }, + { + "epoch": 0.8655014777016482, + "grad_norm": 2.121532678604126, + "learning_rate": 1.0827893070207384e-06, + "loss": 0.5889, + "step": 19548 + }, + { + "epoch": 0.865545753406462, + "grad_norm": 1.717056393623352, + "learning_rate": 1.0820899650022608e-06, + "loss": 0.4509, + "step": 19549 + }, + { + "epoch": 0.8655900291112759, + "grad_norm": 1.4639462232589722, + "learning_rate": 1.0813908359808013e-06, + "loss": 0.4715, + "step": 19550 + }, + { + "epoch": 0.8656343048160898, + "grad_norm": 1.4321924448013306, + "learning_rate": 1.0806919199730614e-06, + "loss": 0.4293, + "step": 19551 + }, + { + "epoch": 0.8656785805209036, + "grad_norm": 1.447124719619751, + "learning_rate": 1.0799932169957327e-06, + "loss": 0.4801, + "step": 19552 + }, + { + "epoch": 0.8657228562257175, + "grad_norm": 1.33000910282135, + "learning_rate": 1.079294727065503e-06, + "loss": 0.4275, + "step": 19553 + }, + { + "epoch": 0.8657671319305315, + "grad_norm": 1.2956955432891846, + "learning_rate": 1.0785964501990553e-06, + "loss": 0.445, + "step": 19554 + }, + { + "epoch": 0.8658114076353453, + "grad_norm": 5.153636455535889, + "learning_rate": 1.0778983864130676e-06, + "loss": 0.5843, + "step": 19555 + }, + { + "epoch": 0.8658556833401592, + "grad_norm": 1.7912111282348633, + "learning_rate": 1.077200535724212e-06, + "loss": 0.5275, + "step": 19556 + }, + { + "epoch": 0.8658999590449731, + "grad_norm": 1.4089454412460327, + "learning_rate": 1.0765028981491544e-06, + "loss": 0.4728, + "step": 19557 + }, + { + "epoch": 0.8659442347497869, + "grad_norm": 1.4413089752197266, + "learning_rate": 1.075805473704562e-06, + "loss": 0.5011, + "step": 19558 + }, + { + "epoch": 0.8659885104546008, + "grad_norm": 1.5738452672958374, + "learning_rate": 1.0751082624070885e-06, + "loss": 0.5136, + "step": 19559 + }, + { + "epoch": 0.8660327861594147, + "grad_norm": 1.6071873903274536, + "learning_rate": 1.074411264273386e-06, + "loss": 0.4374, + "step": 19560 + }, + { + "epoch": 0.8660770618642285, + "grad_norm": 1.9421271085739136, + "learning_rate": 1.0737144793201049e-06, + "loss": 0.5249, + "step": 19561 + }, + { + "epoch": 0.8661213375690424, + "grad_norm": 1.4162397384643555, + "learning_rate": 1.073017907563887e-06, + "loss": 0.4709, + "step": 19562 + }, + { + "epoch": 0.8661656132738563, + "grad_norm": 1.705327033996582, + "learning_rate": 1.0723215490213635e-06, + "loss": 0.5086, + "step": 19563 + }, + { + "epoch": 0.8662098889786701, + "grad_norm": 2.016183614730835, + "learning_rate": 1.0716254037091734e-06, + "loss": 0.592, + "step": 19564 + }, + { + "epoch": 0.866254164683484, + "grad_norm": 1.9963046312332153, + "learning_rate": 1.07092947164394e-06, + "loss": 0.5901, + "step": 19565 + }, + { + "epoch": 0.866298440388298, + "grad_norm": 1.4936468601226807, + "learning_rate": 1.0702337528422868e-06, + "loss": 0.5121, + "step": 19566 + }, + { + "epoch": 0.8663427160931118, + "grad_norm": 1.436671257019043, + "learning_rate": 1.069538247320826e-06, + "loss": 0.4971, + "step": 19567 + }, + { + "epoch": 0.8663869917979257, + "grad_norm": 1.8586171865463257, + "learning_rate": 1.0688429550961765e-06, + "loss": 0.5168, + "step": 19568 + }, + { + "epoch": 0.8664312675027396, + "grad_norm": 1.5875065326690674, + "learning_rate": 1.0681478761849395e-06, + "loss": 0.4935, + "step": 19569 + }, + { + "epoch": 0.8664755432075535, + "grad_norm": 1.3949253559112549, + "learning_rate": 1.0674530106037185e-06, + "loss": 0.4772, + "step": 19570 + }, + { + "epoch": 0.8665198189123673, + "grad_norm": 1.404414176940918, + "learning_rate": 1.0667583583691099e-06, + "loss": 0.4398, + "step": 19571 + }, + { + "epoch": 0.8665640946171812, + "grad_norm": 1.492221713066101, + "learning_rate": 1.066063919497703e-06, + "loss": 0.4827, + "step": 19572 + }, + { + "epoch": 0.8666083703219951, + "grad_norm": 1.2813283205032349, + "learning_rate": 1.0653696940060842e-06, + "loss": 0.443, + "step": 19573 + }, + { + "epoch": 0.8666526460268089, + "grad_norm": 1.449625015258789, + "learning_rate": 1.064675681910836e-06, + "loss": 0.4797, + "step": 19574 + }, + { + "epoch": 0.8666969217316228, + "grad_norm": 1.2362984418869019, + "learning_rate": 1.0639818832285353e-06, + "loss": 0.4546, + "step": 19575 + }, + { + "epoch": 0.8667411974364367, + "grad_norm": 2.0177316665649414, + "learning_rate": 1.0632882979757509e-06, + "loss": 0.5187, + "step": 19576 + }, + { + "epoch": 0.8667854731412505, + "grad_norm": 1.278497576713562, + "learning_rate": 1.062594926169046e-06, + "loss": 0.4718, + "step": 19577 + }, + { + "epoch": 0.8668297488460645, + "grad_norm": 1.4998821020126343, + "learning_rate": 1.0619017678249878e-06, + "loss": 0.5155, + "step": 19578 + }, + { + "epoch": 0.8668740245508784, + "grad_norm": 1.4904310703277588, + "learning_rate": 1.061208822960127e-06, + "loss": 0.4826, + "step": 19579 + }, + { + "epoch": 0.8669183002556922, + "grad_norm": 1.5683752298355103, + "learning_rate": 1.060516091591015e-06, + "loss": 0.4695, + "step": 19580 + }, + { + "epoch": 0.8669625759605061, + "grad_norm": 1.4490790367126465, + "learning_rate": 1.0598235737341978e-06, + "loss": 0.4573, + "step": 19581 + }, + { + "epoch": 0.86700685166532, + "grad_norm": 1.4304333925247192, + "learning_rate": 1.059131269406215e-06, + "loss": 0.4316, + "step": 19582 + }, + { + "epoch": 0.8670511273701338, + "grad_norm": 1.2806551456451416, + "learning_rate": 1.0584391786236004e-06, + "loss": 0.4823, + "step": 19583 + }, + { + "epoch": 0.8670954030749477, + "grad_norm": 1.2591041326522827, + "learning_rate": 1.0577473014028872e-06, + "loss": 0.4094, + "step": 19584 + }, + { + "epoch": 0.8671396787797616, + "grad_norm": 1.790466547012329, + "learning_rate": 1.0570556377605989e-06, + "loss": 0.5759, + "step": 19585 + }, + { + "epoch": 0.8671839544845754, + "grad_norm": 1.3881595134735107, + "learning_rate": 1.0563641877132546e-06, + "loss": 0.4583, + "step": 19586 + }, + { + "epoch": 0.8672282301893893, + "grad_norm": 1.6570706367492676, + "learning_rate": 1.0556729512773679e-06, + "loss": 0.563, + "step": 19587 + }, + { + "epoch": 0.8672725058942032, + "grad_norm": 1.5180248022079468, + "learning_rate": 1.0549819284694519e-06, + "loss": 0.4364, + "step": 19588 + }, + { + "epoch": 0.867316781599017, + "grad_norm": 2.0596237182617188, + "learning_rate": 1.0542911193060102e-06, + "loss": 0.5474, + "step": 19589 + }, + { + "epoch": 0.867361057303831, + "grad_norm": 1.5445624589920044, + "learning_rate": 1.0536005238035396e-06, + "loss": 0.4515, + "step": 19590 + }, + { + "epoch": 0.8674053330086449, + "grad_norm": 1.5325895547866821, + "learning_rate": 1.052910141978537e-06, + "loss": 0.484, + "step": 19591 + }, + { + "epoch": 0.8674496087134587, + "grad_norm": 1.272335171699524, + "learning_rate": 1.0522199738474903e-06, + "loss": 0.469, + "step": 19592 + }, + { + "epoch": 0.8674938844182726, + "grad_norm": 1.561058759689331, + "learning_rate": 1.051530019426882e-06, + "loss": 0.4723, + "step": 19593 + }, + { + "epoch": 0.8675381601230865, + "grad_norm": 1.9125033617019653, + "learning_rate": 1.050840278733195e-06, + "loss": 0.4774, + "step": 19594 + }, + { + "epoch": 0.8675824358279003, + "grad_norm": 2.083611488342285, + "learning_rate": 1.0501507517829012e-06, + "loss": 0.6156, + "step": 19595 + }, + { + "epoch": 0.8676267115327142, + "grad_norm": 1.741856575012207, + "learning_rate": 1.0494614385924707e-06, + "loss": 0.5419, + "step": 19596 + }, + { + "epoch": 0.8676709872375281, + "grad_norm": 1.3485469818115234, + "learning_rate": 1.0487723391783622e-06, + "loss": 0.4536, + "step": 19597 + }, + { + "epoch": 0.8677152629423419, + "grad_norm": 1.4443572759628296, + "learning_rate": 1.0480834535570417e-06, + "loss": 0.4661, + "step": 19598 + }, + { + "epoch": 0.8677595386471558, + "grad_norm": 1.9032292366027832, + "learning_rate": 1.047394781744957e-06, + "loss": 0.633, + "step": 19599 + }, + { + "epoch": 0.8678038143519697, + "grad_norm": 1.7188128232955933, + "learning_rate": 1.0467063237585595e-06, + "loss": 0.4632, + "step": 19600 + }, + { + "epoch": 0.8678480900567835, + "grad_norm": 1.722169280052185, + "learning_rate": 1.0460180796142917e-06, + "loss": 0.5218, + "step": 19601 + }, + { + "epoch": 0.8678923657615975, + "grad_norm": 1.615670084953308, + "learning_rate": 1.0453300493285912e-06, + "loss": 0.5305, + "step": 19602 + }, + { + "epoch": 0.8679366414664114, + "grad_norm": 1.4320882558822632, + "learning_rate": 1.0446422329178896e-06, + "loss": 0.4491, + "step": 19603 + }, + { + "epoch": 0.8679809171712253, + "grad_norm": 1.677240252494812, + "learning_rate": 1.0439546303986182e-06, + "loss": 0.5039, + "step": 19604 + }, + { + "epoch": 0.8680251928760391, + "grad_norm": 1.2735159397125244, + "learning_rate": 1.043267241787198e-06, + "loss": 0.4531, + "step": 19605 + }, + { + "epoch": 0.868069468580853, + "grad_norm": 1.3882495164871216, + "learning_rate": 1.0425800671000475e-06, + "loss": 0.5081, + "step": 19606 + }, + { + "epoch": 0.8681137442856669, + "grad_norm": 1.4972625970840454, + "learning_rate": 1.0418931063535764e-06, + "loss": 0.4793, + "step": 19607 + }, + { + "epoch": 0.8681580199904807, + "grad_norm": 1.392655372619629, + "learning_rate": 1.0412063595641963e-06, + "loss": 0.4465, + "step": 19608 + }, + { + "epoch": 0.8682022956952946, + "grad_norm": 1.6293760538101196, + "learning_rate": 1.0405198267483097e-06, + "loss": 0.5422, + "step": 19609 + }, + { + "epoch": 0.8682465714001085, + "grad_norm": 1.3951829671859741, + "learning_rate": 1.0398335079223076e-06, + "loss": 0.4432, + "step": 19610 + }, + { + "epoch": 0.8682908471049223, + "grad_norm": 1.5213913917541504, + "learning_rate": 1.0391474031025895e-06, + "loss": 0.4662, + "step": 19611 + }, + { + "epoch": 0.8683351228097362, + "grad_norm": 2.51556134223938, + "learning_rate": 1.0384615123055386e-06, + "loss": 0.5145, + "step": 19612 + }, + { + "epoch": 0.8683793985145501, + "grad_norm": 1.4908175468444824, + "learning_rate": 1.0377758355475388e-06, + "loss": 0.5053, + "step": 19613 + }, + { + "epoch": 0.868423674219364, + "grad_norm": 1.496048092842102, + "learning_rate": 1.0370903728449632e-06, + "loss": 0.4687, + "step": 19614 + }, + { + "epoch": 0.8684679499241779, + "grad_norm": 1.4175235033035278, + "learning_rate": 1.036405124214188e-06, + "loss": 0.4716, + "step": 19615 + }, + { + "epoch": 0.8685122256289918, + "grad_norm": 1.5242512226104736, + "learning_rate": 1.035720089671578e-06, + "loss": 0.4956, + "step": 19616 + }, + { + "epoch": 0.8685565013338056, + "grad_norm": 1.7840791940689087, + "learning_rate": 1.035035269233493e-06, + "loss": 0.5101, + "step": 19617 + }, + { + "epoch": 0.8686007770386195, + "grad_norm": 1.853186845779419, + "learning_rate": 1.0343506629162948e-06, + "loss": 0.554, + "step": 19618 + }, + { + "epoch": 0.8686450527434334, + "grad_norm": 1.649971604347229, + "learning_rate": 1.0336662707363287e-06, + "loss": 0.4756, + "step": 19619 + }, + { + "epoch": 0.8686893284482472, + "grad_norm": 2.2955315113067627, + "learning_rate": 1.032982092709942e-06, + "loss": 0.6004, + "step": 19620 + }, + { + "epoch": 0.8687336041530611, + "grad_norm": 1.8770266771316528, + "learning_rate": 1.0322981288534772e-06, + "loss": 0.5462, + "step": 19621 + }, + { + "epoch": 0.868777879857875, + "grad_norm": 1.3885929584503174, + "learning_rate": 1.0316143791832712e-06, + "loss": 0.4619, + "step": 19622 + }, + { + "epoch": 0.8688221555626888, + "grad_norm": 1.3561763763427734, + "learning_rate": 1.030930843715653e-06, + "loss": 0.4756, + "step": 19623 + }, + { + "epoch": 0.8688664312675027, + "grad_norm": 2.2694740295410156, + "learning_rate": 1.0302475224669474e-06, + "loss": 0.5267, + "step": 19624 + }, + { + "epoch": 0.8689107069723166, + "grad_norm": 1.5602577924728394, + "learning_rate": 1.0295644154534767e-06, + "loss": 0.4565, + "step": 19625 + }, + { + "epoch": 0.8689549826771304, + "grad_norm": 1.8481568098068237, + "learning_rate": 1.0288815226915572e-06, + "loss": 0.5274, + "step": 19626 + }, + { + "epoch": 0.8689992583819444, + "grad_norm": 1.5255932807922363, + "learning_rate": 1.0281988441974978e-06, + "loss": 0.4942, + "step": 19627 + }, + { + "epoch": 0.8690435340867583, + "grad_norm": 1.7095839977264404, + "learning_rate": 1.027516379987603e-06, + "loss": 0.5208, + "step": 19628 + }, + { + "epoch": 0.8690878097915721, + "grad_norm": 1.721690058708191, + "learning_rate": 1.0268341300781748e-06, + "loss": 0.454, + "step": 19629 + }, + { + "epoch": 0.869132085496386, + "grad_norm": 2.006532669067383, + "learning_rate": 1.0261520944855052e-06, + "loss": 0.5509, + "step": 19630 + }, + { + "epoch": 0.8691763612011999, + "grad_norm": 1.6646440029144287, + "learning_rate": 1.025470273225887e-06, + "loss": 0.4959, + "step": 19631 + }, + { + "epoch": 0.8692206369060137, + "grad_norm": 1.5918203592300415, + "learning_rate": 1.0247886663156048e-06, + "loss": 0.539, + "step": 19632 + }, + { + "epoch": 0.8692649126108276, + "grad_norm": 1.655176043510437, + "learning_rate": 1.024107273770938e-06, + "loss": 0.5762, + "step": 19633 + }, + { + "epoch": 0.8693091883156415, + "grad_norm": 1.3638594150543213, + "learning_rate": 1.0234260956081577e-06, + "loss": 0.4486, + "step": 19634 + }, + { + "epoch": 0.8693534640204553, + "grad_norm": 1.4885262250900269, + "learning_rate": 1.0227451318435378e-06, + "loss": 0.4963, + "step": 19635 + }, + { + "epoch": 0.8693977397252692, + "grad_norm": 1.3974802494049072, + "learning_rate": 1.0220643824933417e-06, + "loss": 0.5144, + "step": 19636 + }, + { + "epoch": 0.8694420154300831, + "grad_norm": 1.4716671705245972, + "learning_rate": 1.021383847573827e-06, + "loss": 0.4808, + "step": 19637 + }, + { + "epoch": 0.869486291134897, + "grad_norm": 1.6341155767440796, + "learning_rate": 1.0207035271012488e-06, + "loss": 0.4732, + "step": 19638 + }, + { + "epoch": 0.8695305668397109, + "grad_norm": 1.4863345623016357, + "learning_rate": 1.0200234210918558e-06, + "loss": 0.5068, + "step": 19639 + }, + { + "epoch": 0.8695748425445248, + "grad_norm": 1.4350898265838623, + "learning_rate": 1.0193435295618892e-06, + "loss": 0.444, + "step": 19640 + }, + { + "epoch": 0.8696191182493387, + "grad_norm": 1.4827961921691895, + "learning_rate": 1.0186638525275927e-06, + "loss": 0.5023, + "step": 19641 + }, + { + "epoch": 0.8696633939541525, + "grad_norm": 1.7596256732940674, + "learning_rate": 1.0179843900051967e-06, + "loss": 0.5118, + "step": 19642 + }, + { + "epoch": 0.8697076696589664, + "grad_norm": 1.4322941303253174, + "learning_rate": 1.0173051420109303e-06, + "loss": 0.4501, + "step": 19643 + }, + { + "epoch": 0.8697519453637803, + "grad_norm": 1.4432965517044067, + "learning_rate": 1.016626108561014e-06, + "loss": 0.4528, + "step": 19644 + }, + { + "epoch": 0.8697962210685941, + "grad_norm": 1.4270347356796265, + "learning_rate": 1.0159472896716715e-06, + "loss": 0.5007, + "step": 19645 + }, + { + "epoch": 0.869840496773408, + "grad_norm": 1.6785897016525269, + "learning_rate": 1.015268685359112e-06, + "loss": 0.4813, + "step": 19646 + }, + { + "epoch": 0.8698847724782219, + "grad_norm": 1.4276673793792725, + "learning_rate": 1.0145902956395449e-06, + "loss": 0.5147, + "step": 19647 + }, + { + "epoch": 0.8699290481830357, + "grad_norm": 1.5770896673202515, + "learning_rate": 1.0139121205291725e-06, + "loss": 0.4915, + "step": 19648 + }, + { + "epoch": 0.8699733238878496, + "grad_norm": 1.8392504453659058, + "learning_rate": 1.013234160044192e-06, + "loss": 0.5468, + "step": 19649 + }, + { + "epoch": 0.8700175995926636, + "grad_norm": 1.9163877964019775, + "learning_rate": 1.0125564142007948e-06, + "loss": 0.5045, + "step": 19650 + }, + { + "epoch": 0.8700618752974774, + "grad_norm": 1.7152845859527588, + "learning_rate": 1.0118788830151716e-06, + "loss": 0.5369, + "step": 19651 + }, + { + "epoch": 0.8701061510022913, + "grad_norm": 1.6488926410675049, + "learning_rate": 1.0112015665035023e-06, + "loss": 0.5063, + "step": 19652 + }, + { + "epoch": 0.8701504267071052, + "grad_norm": 1.3564937114715576, + "learning_rate": 1.0105244646819656e-06, + "loss": 0.4372, + "step": 19653 + }, + { + "epoch": 0.870194702411919, + "grad_norm": 1.4817522764205933, + "learning_rate": 1.0098475775667294e-06, + "loss": 0.4621, + "step": 19654 + }, + { + "epoch": 0.8702389781167329, + "grad_norm": 1.4537166357040405, + "learning_rate": 1.0091709051739673e-06, + "loss": 0.442, + "step": 19655 + }, + { + "epoch": 0.8702832538215468, + "grad_norm": 1.5270562171936035, + "learning_rate": 1.0084944475198367e-06, + "loss": 0.4972, + "step": 19656 + }, + { + "epoch": 0.8703275295263606, + "grad_norm": 1.3163772821426392, + "learning_rate": 1.0078182046204943e-06, + "loss": 0.4516, + "step": 19657 + }, + { + "epoch": 0.8703718052311745, + "grad_norm": 1.57528555393219, + "learning_rate": 1.007142176492093e-06, + "loss": 0.465, + "step": 19658 + }, + { + "epoch": 0.8704160809359884, + "grad_norm": 1.330659031867981, + "learning_rate": 1.0064663631507787e-06, + "loss": 0.4713, + "step": 19659 + }, + { + "epoch": 0.8704603566408022, + "grad_norm": 1.4912333488464355, + "learning_rate": 1.0057907646126908e-06, + "loss": 0.4722, + "step": 19660 + }, + { + "epoch": 0.8705046323456161, + "grad_norm": 1.1599359512329102, + "learning_rate": 1.0051153808939683e-06, + "loss": 0.3966, + "step": 19661 + }, + { + "epoch": 0.8705489080504301, + "grad_norm": 1.685063362121582, + "learning_rate": 1.004440212010741e-06, + "loss": 0.5145, + "step": 19662 + }, + { + "epoch": 0.8705931837552439, + "grad_norm": 1.8597112894058228, + "learning_rate": 1.0037652579791357e-06, + "loss": 0.5586, + "step": 19663 + }, + { + "epoch": 0.8706374594600578, + "grad_norm": 1.2327191829681396, + "learning_rate": 1.0030905188152695e-06, + "loss": 0.4777, + "step": 19664 + }, + { + "epoch": 0.8706817351648717, + "grad_norm": 1.6009714603424072, + "learning_rate": 1.0024159945352652e-06, + "loss": 0.5162, + "step": 19665 + }, + { + "epoch": 0.8707260108696855, + "grad_norm": 1.5719095468521118, + "learning_rate": 1.0017416851552264e-06, + "loss": 0.4738, + "step": 19666 + }, + { + "epoch": 0.8707702865744994, + "grad_norm": 1.4713704586029053, + "learning_rate": 1.0010675906912593e-06, + "loss": 0.4887, + "step": 19667 + }, + { + "epoch": 0.8708145622793133, + "grad_norm": 1.5117847919464111, + "learning_rate": 1.0003937111594674e-06, + "loss": 0.4979, + "step": 19668 + }, + { + "epoch": 0.8708588379841271, + "grad_norm": 1.2899391651153564, + "learning_rate": 9.997200465759439e-07, + "loss": 0.3975, + "step": 19669 + }, + { + "epoch": 0.870903113688941, + "grad_norm": 1.3703171014785767, + "learning_rate": 9.990465969567764e-07, + "loss": 0.4249, + "step": 19670 + }, + { + "epoch": 0.8709473893937549, + "grad_norm": 2.114863872528076, + "learning_rate": 9.983733623180537e-07, + "loss": 0.5662, + "step": 19671 + }, + { + "epoch": 0.8709916650985687, + "grad_norm": 1.4845844507217407, + "learning_rate": 9.977003426758548e-07, + "loss": 0.4679, + "step": 19672 + }, + { + "epoch": 0.8710359408033826, + "grad_norm": 1.3996286392211914, + "learning_rate": 9.970275380462536e-07, + "loss": 0.4505, + "step": 19673 + }, + { + "epoch": 0.8710802165081966, + "grad_norm": 1.3524916172027588, + "learning_rate": 9.963549484453184e-07, + "loss": 0.4568, + "step": 19674 + }, + { + "epoch": 0.8711244922130105, + "grad_norm": 1.3311816453933716, + "learning_rate": 9.95682573889114e-07, + "loss": 0.4194, + "step": 19675 + }, + { + "epoch": 0.8711687679178243, + "grad_norm": 1.432481288909912, + "learning_rate": 9.950104143937002e-07, + "loss": 0.4976, + "step": 19676 + }, + { + "epoch": 0.8712130436226382, + "grad_norm": 1.5309205055236816, + "learning_rate": 9.943384699751279e-07, + "loss": 0.4557, + "step": 19677 + }, + { + "epoch": 0.8712573193274521, + "grad_norm": 1.727318286895752, + "learning_rate": 9.936667406494517e-07, + "loss": 0.5197, + "step": 19678 + }, + { + "epoch": 0.8713015950322659, + "grad_norm": 1.6891365051269531, + "learning_rate": 9.929952264327113e-07, + "loss": 0.5167, + "step": 19679 + }, + { + "epoch": 0.8713458707370798, + "grad_norm": 1.4135023355484009, + "learning_rate": 9.923239273409446e-07, + "loss": 0.4944, + "step": 19680 + }, + { + "epoch": 0.8713901464418937, + "grad_norm": 1.4392741918563843, + "learning_rate": 9.916528433901873e-07, + "loss": 0.487, + "step": 19681 + }, + { + "epoch": 0.8714344221467075, + "grad_norm": 1.4691053628921509, + "learning_rate": 9.90981974596469e-07, + "loss": 0.4912, + "step": 19682 + }, + { + "epoch": 0.8714786978515214, + "grad_norm": 1.6203199625015259, + "learning_rate": 9.903113209758098e-07, + "loss": 0.4624, + "step": 19683 + }, + { + "epoch": 0.8715229735563353, + "grad_norm": 2.5667083263397217, + "learning_rate": 9.896408825442283e-07, + "loss": 0.5925, + "step": 19684 + }, + { + "epoch": 0.8715672492611491, + "grad_norm": 1.4369450807571411, + "learning_rate": 9.889706593177383e-07, + "loss": 0.4652, + "step": 19685 + }, + { + "epoch": 0.8716115249659631, + "grad_norm": 1.4364643096923828, + "learning_rate": 9.883006513123473e-07, + "loss": 0.445, + "step": 19686 + }, + { + "epoch": 0.871655800670777, + "grad_norm": 1.5872046947479248, + "learning_rate": 9.876308585440552e-07, + "loss": 0.5313, + "step": 19687 + }, + { + "epoch": 0.8717000763755908, + "grad_norm": 1.3749651908874512, + "learning_rate": 9.869612810288642e-07, + "loss": 0.4481, + "step": 19688 + }, + { + "epoch": 0.8717443520804047, + "grad_norm": 1.3392945528030396, + "learning_rate": 9.862919187827636e-07, + "loss": 0.4671, + "step": 19689 + }, + { + "epoch": 0.8717886277852186, + "grad_norm": 1.3041164875030518, + "learning_rate": 9.856227718217404e-07, + "loss": 0.4523, + "step": 19690 + }, + { + "epoch": 0.8718329034900324, + "grad_norm": 1.7089802026748657, + "learning_rate": 9.849538401617786e-07, + "loss": 0.5266, + "step": 19691 + }, + { + "epoch": 0.8718771791948463, + "grad_norm": 1.650363564491272, + "learning_rate": 9.842851238188545e-07, + "loss": 0.536, + "step": 19692 + }, + { + "epoch": 0.8719214548996602, + "grad_norm": 1.3947257995605469, + "learning_rate": 9.836166228089395e-07, + "loss": 0.469, + "step": 19693 + }, + { + "epoch": 0.871965730604474, + "grad_norm": 1.383391261100769, + "learning_rate": 9.829483371479998e-07, + "loss": 0.4954, + "step": 19694 + }, + { + "epoch": 0.8720100063092879, + "grad_norm": 1.7227283716201782, + "learning_rate": 9.82280266851997e-07, + "loss": 0.4954, + "step": 19695 + }, + { + "epoch": 0.8720542820141018, + "grad_norm": 1.8576984405517578, + "learning_rate": 9.816124119368875e-07, + "loss": 0.525, + "step": 19696 + }, + { + "epoch": 0.8720985577189156, + "grad_norm": 1.2311242818832397, + "learning_rate": 9.809447724186206e-07, + "loss": 0.4469, + "step": 19697 + }, + { + "epoch": 0.8721428334237296, + "grad_norm": 1.4347325563430786, + "learning_rate": 9.802773483131455e-07, + "loss": 0.4638, + "step": 19698 + }, + { + "epoch": 0.8721871091285435, + "grad_norm": 1.7466378211975098, + "learning_rate": 9.796101396364033e-07, + "loss": 0.5038, + "step": 19699 + }, + { + "epoch": 0.8722313848333573, + "grad_norm": 1.4873801469802856, + "learning_rate": 9.789431464043253e-07, + "loss": 0.5377, + "step": 19700 + }, + { + "epoch": 0.8722756605381712, + "grad_norm": 1.2938244342803955, + "learning_rate": 9.782763686328478e-07, + "loss": 0.4575, + "step": 19701 + }, + { + "epoch": 0.8723199362429851, + "grad_norm": 1.8370697498321533, + "learning_rate": 9.776098063378924e-07, + "loss": 0.5403, + "step": 19702 + }, + { + "epoch": 0.8723642119477989, + "grad_norm": 1.6353026628494263, + "learning_rate": 9.76943459535381e-07, + "loss": 0.5047, + "step": 19703 + }, + { + "epoch": 0.8724084876526128, + "grad_norm": 1.505423903465271, + "learning_rate": 9.762773282412286e-07, + "loss": 0.4758, + "step": 19704 + }, + { + "epoch": 0.8724527633574267, + "grad_norm": 2.2022628784179688, + "learning_rate": 9.756114124713435e-07, + "loss": 0.4661, + "step": 19705 + }, + { + "epoch": 0.8724970390622405, + "grad_norm": 1.394128441810608, + "learning_rate": 9.749457122416329e-07, + "loss": 0.4451, + "step": 19706 + }, + { + "epoch": 0.8725413147670544, + "grad_norm": 1.5713990926742554, + "learning_rate": 9.742802275679941e-07, + "loss": 0.5022, + "step": 19707 + }, + { + "epoch": 0.8725855904718683, + "grad_norm": 1.361687421798706, + "learning_rate": 9.736149584663235e-07, + "loss": 0.43, + "step": 19708 + }, + { + "epoch": 0.8726298661766821, + "grad_norm": 1.4587781429290771, + "learning_rate": 9.729499049525114e-07, + "loss": 0.4601, + "step": 19709 + }, + { + "epoch": 0.8726741418814961, + "grad_norm": 1.9452288150787354, + "learning_rate": 9.722850670424378e-07, + "loss": 0.5048, + "step": 19710 + }, + { + "epoch": 0.87271841758631, + "grad_norm": 1.4097694158554077, + "learning_rate": 9.716204447519872e-07, + "loss": 0.4397, + "step": 19711 + }, + { + "epoch": 0.8727626932911239, + "grad_norm": 1.4577618837356567, + "learning_rate": 9.709560380970317e-07, + "loss": 0.4814, + "step": 19712 + }, + { + "epoch": 0.8728069689959377, + "grad_norm": 1.5131916999816895, + "learning_rate": 9.702918470934409e-07, + "loss": 0.4763, + "step": 19713 + }, + { + "epoch": 0.8728512447007516, + "grad_norm": 1.353562831878662, + "learning_rate": 9.696278717570727e-07, + "loss": 0.4651, + "step": 19714 + }, + { + "epoch": 0.8728955204055655, + "grad_norm": 1.8355506658554077, + "learning_rate": 9.689641121037918e-07, + "loss": 0.5104, + "step": 19715 + }, + { + "epoch": 0.8729397961103793, + "grad_norm": 1.447452425956726, + "learning_rate": 9.683005681494506e-07, + "loss": 0.499, + "step": 19716 + }, + { + "epoch": 0.8729840718151932, + "grad_norm": 1.5037790536880493, + "learning_rate": 9.676372399098944e-07, + "loss": 0.5113, + "step": 19717 + }, + { + "epoch": 0.8730283475200071, + "grad_norm": 1.6217670440673828, + "learning_rate": 9.669741274009702e-07, + "loss": 0.4564, + "step": 19718 + }, + { + "epoch": 0.8730726232248209, + "grad_norm": 1.8169748783111572, + "learning_rate": 9.663112306385136e-07, + "loss": 0.5158, + "step": 19719 + }, + { + "epoch": 0.8731168989296348, + "grad_norm": 1.6832389831542969, + "learning_rate": 9.656485496383583e-07, + "loss": 0.4838, + "step": 19720 + }, + { + "epoch": 0.8731611746344488, + "grad_norm": 1.684500813484192, + "learning_rate": 9.649860844163307e-07, + "loss": 0.4993, + "step": 19721 + }, + { + "epoch": 0.8732054503392626, + "grad_norm": 1.674977421760559, + "learning_rate": 9.643238349882544e-07, + "loss": 0.5184, + "step": 19722 + }, + { + "epoch": 0.8732497260440765, + "grad_norm": 1.2290929555892944, + "learning_rate": 9.63661801369946e-07, + "loss": 0.4442, + "step": 19723 + }, + { + "epoch": 0.8732940017488904, + "grad_norm": 1.5049469470977783, + "learning_rate": 9.629999835772152e-07, + "loss": 0.4495, + "step": 19724 + }, + { + "epoch": 0.8733382774537042, + "grad_norm": 1.3048371076583862, + "learning_rate": 9.623383816258746e-07, + "loss": 0.4603, + "step": 19725 + }, + { + "epoch": 0.8733825531585181, + "grad_norm": 1.8005627393722534, + "learning_rate": 9.61676995531723e-07, + "loss": 0.5104, + "step": 19726 + }, + { + "epoch": 0.873426828863332, + "grad_norm": 1.6886979341506958, + "learning_rate": 9.61015825310554e-07, + "loss": 0.5297, + "step": 19727 + }, + { + "epoch": 0.8734711045681458, + "grad_norm": 1.3872405290603638, + "learning_rate": 9.603548709781652e-07, + "loss": 0.4865, + "step": 19728 + }, + { + "epoch": 0.8735153802729597, + "grad_norm": 1.2601207494735718, + "learning_rate": 9.596941325503395e-07, + "loss": 0.4009, + "step": 19729 + }, + { + "epoch": 0.8735596559777736, + "grad_norm": 1.5297043323516846, + "learning_rate": 9.590336100428588e-07, + "loss": 0.4695, + "step": 19730 + }, + { + "epoch": 0.8736039316825874, + "grad_norm": 1.5331240892410278, + "learning_rate": 9.583733034714982e-07, + "loss": 0.4711, + "step": 19731 + }, + { + "epoch": 0.8736482073874013, + "grad_norm": 2.0714271068573, + "learning_rate": 9.577132128520305e-07, + "loss": 0.5304, + "step": 19732 + }, + { + "epoch": 0.8736924830922153, + "grad_norm": 3.9401090145111084, + "learning_rate": 9.57053338200219e-07, + "loss": 0.4773, + "step": 19733 + }, + { + "epoch": 0.8737367587970291, + "grad_norm": 1.450535774230957, + "learning_rate": 9.563936795318241e-07, + "loss": 0.4672, + "step": 19734 + }, + { + "epoch": 0.873781034501843, + "grad_norm": 1.419960379600525, + "learning_rate": 9.557342368626056e-07, + "loss": 0.4623, + "step": 19735 + }, + { + "epoch": 0.8738253102066569, + "grad_norm": 1.615309238433838, + "learning_rate": 9.550750102083095e-07, + "loss": 0.4885, + "step": 19736 + }, + { + "epoch": 0.8738695859114707, + "grad_norm": 1.617281198501587, + "learning_rate": 9.544159995846802e-07, + "loss": 0.4823, + "step": 19737 + }, + { + "epoch": 0.8739138616162846, + "grad_norm": 1.8108164072036743, + "learning_rate": 9.537572050074617e-07, + "loss": 0.487, + "step": 19738 + }, + { + "epoch": 0.8739581373210985, + "grad_norm": 1.5386662483215332, + "learning_rate": 9.530986264923858e-07, + "loss": 0.462, + "step": 19739 + }, + { + "epoch": 0.8740024130259123, + "grad_norm": 1.484171986579895, + "learning_rate": 9.524402640551844e-07, + "loss": 0.4652, + "step": 19740 + }, + { + "epoch": 0.8740466887307262, + "grad_norm": 1.7312368154525757, + "learning_rate": 9.517821177115793e-07, + "loss": 0.5256, + "step": 19741 + }, + { + "epoch": 0.8740909644355401, + "grad_norm": 1.7987637519836426, + "learning_rate": 9.511241874772914e-07, + "loss": 0.5195, + "step": 19742 + }, + { + "epoch": 0.8741352401403539, + "grad_norm": 1.363419771194458, + "learning_rate": 9.504664733680346e-07, + "loss": 0.4844, + "step": 19743 + }, + { + "epoch": 0.8741795158451678, + "grad_norm": 2.116039276123047, + "learning_rate": 9.498089753995166e-07, + "loss": 0.618, + "step": 19744 + }, + { + "epoch": 0.8742237915499818, + "grad_norm": 1.4577929973602295, + "learning_rate": 9.491516935874434e-07, + "loss": 0.4264, + "step": 19745 + }, + { + "epoch": 0.8742680672547957, + "grad_norm": 1.646483302116394, + "learning_rate": 9.484946279475116e-07, + "loss": 0.4941, + "step": 19746 + }, + { + "epoch": 0.8743123429596095, + "grad_norm": 1.6541568040847778, + "learning_rate": 9.478377784954152e-07, + "loss": 0.4773, + "step": 19747 + }, + { + "epoch": 0.8743566186644234, + "grad_norm": 1.2404489517211914, + "learning_rate": 9.471811452468438e-07, + "loss": 0.4809, + "step": 19748 + }, + { + "epoch": 0.8744008943692373, + "grad_norm": 1.3739793300628662, + "learning_rate": 9.465247282174806e-07, + "loss": 0.4691, + "step": 19749 + }, + { + "epoch": 0.8744451700740511, + "grad_norm": 1.3296115398406982, + "learning_rate": 9.458685274230017e-07, + "loss": 0.4369, + "step": 19750 + }, + { + "epoch": 0.874489445778865, + "grad_norm": 1.7453495264053345, + "learning_rate": 9.452125428790814e-07, + "loss": 0.5099, + "step": 19751 + }, + { + "epoch": 0.8745337214836789, + "grad_norm": 1.3806043863296509, + "learning_rate": 9.445567746013862e-07, + "loss": 0.4178, + "step": 19752 + }, + { + "epoch": 0.8745779971884927, + "grad_norm": 1.6342380046844482, + "learning_rate": 9.439012226055799e-07, + "loss": 0.4868, + "step": 19753 + }, + { + "epoch": 0.8746222728933066, + "grad_norm": 1.498077630996704, + "learning_rate": 9.432458869073168e-07, + "loss": 0.4365, + "step": 19754 + }, + { + "epoch": 0.8746665485981205, + "grad_norm": 1.5404306650161743, + "learning_rate": 9.425907675222535e-07, + "loss": 0.534, + "step": 19755 + }, + { + "epoch": 0.8747108243029343, + "grad_norm": 1.760481834411621, + "learning_rate": 9.419358644660348e-07, + "loss": 0.619, + "step": 19756 + }, + { + "epoch": 0.8747551000077483, + "grad_norm": 1.363100290298462, + "learning_rate": 9.412811777543007e-07, + "loss": 0.4703, + "step": 19757 + }, + { + "epoch": 0.8747993757125622, + "grad_norm": 1.4237140417099, + "learning_rate": 9.40626707402692e-07, + "loss": 0.4611, + "step": 19758 + }, + { + "epoch": 0.874843651417376, + "grad_norm": 1.6181448698043823, + "learning_rate": 9.399724534268385e-07, + "loss": 0.4891, + "step": 19759 + }, + { + "epoch": 0.8748879271221899, + "grad_norm": 1.4105172157287598, + "learning_rate": 9.393184158423674e-07, + "loss": 0.4845, + "step": 19760 + }, + { + "epoch": 0.8749322028270038, + "grad_norm": 1.5710581541061401, + "learning_rate": 9.386645946648942e-07, + "loss": 0.4588, + "step": 19761 + }, + { + "epoch": 0.8749764785318176, + "grad_norm": 1.4113730192184448, + "learning_rate": 9.380109899100409e-07, + "loss": 0.5026, + "step": 19762 + }, + { + "epoch": 0.8750207542366315, + "grad_norm": 1.719804286956787, + "learning_rate": 9.373576015934172e-07, + "loss": 0.5599, + "step": 19763 + }, + { + "epoch": 0.8750650299414454, + "grad_norm": 1.5707550048828125, + "learning_rate": 9.36704429730626e-07, + "loss": 0.4555, + "step": 19764 + }, + { + "epoch": 0.8751093056462592, + "grad_norm": 1.5351232290267944, + "learning_rate": 9.360514743372718e-07, + "loss": 0.4873, + "step": 19765 + }, + { + "epoch": 0.8751535813510731, + "grad_norm": 1.9656261205673218, + "learning_rate": 9.353987354289484e-07, + "loss": 0.5269, + "step": 19766 + }, + { + "epoch": 0.875197857055887, + "grad_norm": 1.5178039073944092, + "learning_rate": 9.347462130212426e-07, + "loss": 0.5083, + "step": 19767 + }, + { + "epoch": 0.8752421327607008, + "grad_norm": 2.1461074352264404, + "learning_rate": 9.34093907129745e-07, + "loss": 0.5659, + "step": 19768 + }, + { + "epoch": 0.8752864084655148, + "grad_norm": 1.6607047319412231, + "learning_rate": 9.334418177700344e-07, + "loss": 0.5037, + "step": 19769 + }, + { + "epoch": 0.8753306841703287, + "grad_norm": 1.684144377708435, + "learning_rate": 9.327899449576816e-07, + "loss": 0.4883, + "step": 19770 + }, + { + "epoch": 0.8753749598751425, + "grad_norm": 1.5209134817123413, + "learning_rate": 9.321382887082564e-07, + "loss": 0.4669, + "step": 19771 + }, + { + "epoch": 0.8754192355799564, + "grad_norm": 1.5794850587844849, + "learning_rate": 9.314868490373274e-07, + "loss": 0.524, + "step": 19772 + }, + { + "epoch": 0.8754635112847703, + "grad_norm": 1.56100332736969, + "learning_rate": 9.3083562596045e-07, + "loss": 0.4516, + "step": 19773 + }, + { + "epoch": 0.8755077869895841, + "grad_norm": 1.4248825311660767, + "learning_rate": 9.301846194931785e-07, + "loss": 0.472, + "step": 19774 + }, + { + "epoch": 0.875552062694398, + "grad_norm": 1.4585567712783813, + "learning_rate": 9.295338296510637e-07, + "loss": 0.494, + "step": 19775 + }, + { + "epoch": 0.8755963383992119, + "grad_norm": 1.5061146020889282, + "learning_rate": 9.288832564496487e-07, + "loss": 0.4899, + "step": 19776 + }, + { + "epoch": 0.8756406141040257, + "grad_norm": 2.062115430831909, + "learning_rate": 9.282328999044698e-07, + "loss": 0.5871, + "step": 19777 + }, + { + "epoch": 0.8756848898088396, + "grad_norm": 1.5958828926086426, + "learning_rate": 9.275827600310628e-07, + "loss": 0.4573, + "step": 19778 + }, + { + "epoch": 0.8757291655136535, + "grad_norm": 1.5230960845947266, + "learning_rate": 9.269328368449538e-07, + "loss": 0.4706, + "step": 19779 + }, + { + "epoch": 0.8757734412184673, + "grad_norm": 1.4843347072601318, + "learning_rate": 9.262831303616671e-07, + "loss": 0.4983, + "step": 19780 + }, + { + "epoch": 0.8758177169232813, + "grad_norm": 1.749692440032959, + "learning_rate": 9.25633640596717e-07, + "loss": 0.4453, + "step": 19781 + }, + { + "epoch": 0.8758619926280952, + "grad_norm": 1.5114847421646118, + "learning_rate": 9.249843675656212e-07, + "loss": 0.483, + "step": 19782 + }, + { + "epoch": 0.8759062683329091, + "grad_norm": 1.3514353036880493, + "learning_rate": 9.243353112838838e-07, + "loss": 0.492, + "step": 19783 + }, + { + "epoch": 0.8759505440377229, + "grad_norm": 1.25816810131073, + "learning_rate": 9.236864717670069e-07, + "loss": 0.4436, + "step": 19784 + }, + { + "epoch": 0.8759948197425368, + "grad_norm": 1.4448975324630737, + "learning_rate": 9.230378490304903e-07, + "loss": 0.4381, + "step": 19785 + }, + { + "epoch": 0.8760390954473507, + "grad_norm": 1.4345275163650513, + "learning_rate": 9.223894430898239e-07, + "loss": 0.4658, + "step": 19786 + }, + { + "epoch": 0.8760833711521645, + "grad_norm": 1.7711477279663086, + "learning_rate": 9.217412539604942e-07, + "loss": 0.5306, + "step": 19787 + }, + { + "epoch": 0.8761276468569784, + "grad_norm": 1.8444883823394775, + "learning_rate": 9.21093281657982e-07, + "loss": 0.48, + "step": 19788 + }, + { + "epoch": 0.8761719225617923, + "grad_norm": 1.6010422706604004, + "learning_rate": 9.204455261977651e-07, + "loss": 0.4756, + "step": 19789 + }, + { + "epoch": 0.8762161982666061, + "grad_norm": 1.581134557723999, + "learning_rate": 9.197979875953144e-07, + "loss": 0.5375, + "step": 19790 + }, + { + "epoch": 0.87626047397142, + "grad_norm": 1.5334407091140747, + "learning_rate": 9.191506658660931e-07, + "loss": 0.505, + "step": 19791 + }, + { + "epoch": 0.876304749676234, + "grad_norm": 1.5110939741134644, + "learning_rate": 9.185035610255666e-07, + "loss": 0.532, + "step": 19792 + }, + { + "epoch": 0.8763490253810478, + "grad_norm": 1.6512032747268677, + "learning_rate": 9.17856673089188e-07, + "loss": 0.5003, + "step": 19793 + }, + { + "epoch": 0.8763933010858617, + "grad_norm": 1.6218167543411255, + "learning_rate": 9.172100020724051e-07, + "loss": 0.4859, + "step": 19794 + }, + { + "epoch": 0.8764375767906756, + "grad_norm": 1.480393886566162, + "learning_rate": 9.165635479906687e-07, + "loss": 0.4568, + "step": 19795 + }, + { + "epoch": 0.8764818524954894, + "grad_norm": 1.980477213859558, + "learning_rate": 9.159173108594155e-07, + "loss": 0.5442, + "step": 19796 + }, + { + "epoch": 0.8765261282003033, + "grad_norm": 1.173269271850586, + "learning_rate": 9.152712906940808e-07, + "loss": 0.4359, + "step": 19797 + }, + { + "epoch": 0.8765704039051172, + "grad_norm": 1.37087881565094, + "learning_rate": 9.146254875100946e-07, + "loss": 0.4943, + "step": 19798 + }, + { + "epoch": 0.876614679609931, + "grad_norm": 1.535423994064331, + "learning_rate": 9.139799013228801e-07, + "loss": 0.5229, + "step": 19799 + }, + { + "epoch": 0.8766589553147449, + "grad_norm": 1.6970584392547607, + "learning_rate": 9.133345321478582e-07, + "loss": 0.4693, + "step": 19800 + }, + { + "epoch": 0.8767032310195588, + "grad_norm": 1.711883544921875, + "learning_rate": 9.126893800004411e-07, + "loss": 0.5291, + "step": 19801 + }, + { + "epoch": 0.8767475067243726, + "grad_norm": 1.4809311628341675, + "learning_rate": 9.120444448960397e-07, + "loss": 0.4849, + "step": 19802 + }, + { + "epoch": 0.8767917824291865, + "grad_norm": 1.5715832710266113, + "learning_rate": 9.113997268500574e-07, + "loss": 0.4449, + "step": 19803 + }, + { + "epoch": 0.8768360581340005, + "grad_norm": 1.2776814699172974, + "learning_rate": 9.107552258778907e-07, + "loss": 0.4486, + "step": 19804 + }, + { + "epoch": 0.8768803338388143, + "grad_norm": 1.3472108840942383, + "learning_rate": 9.101109419949372e-07, + "loss": 0.4654, + "step": 19805 + }, + { + "epoch": 0.8769246095436282, + "grad_norm": 1.5302915573120117, + "learning_rate": 9.094668752165814e-07, + "loss": 0.5549, + "step": 19806 + }, + { + "epoch": 0.8769688852484421, + "grad_norm": 1.54114830493927, + "learning_rate": 9.088230255582087e-07, + "loss": 0.4725, + "step": 19807 + }, + { + "epoch": 0.8770131609532559, + "grad_norm": 1.6394070386886597, + "learning_rate": 9.081793930351946e-07, + "loss": 0.5146, + "step": 19808 + }, + { + "epoch": 0.8770574366580698, + "grad_norm": 1.7488480806350708, + "learning_rate": 9.075359776629133e-07, + "loss": 0.4903, + "step": 19809 + }, + { + "epoch": 0.8771017123628837, + "grad_norm": 1.3244593143463135, + "learning_rate": 9.068927794567317e-07, + "loss": 0.4338, + "step": 19810 + }, + { + "epoch": 0.8771459880676975, + "grad_norm": 1.4329315423965454, + "learning_rate": 9.062497984320106e-07, + "loss": 0.512, + "step": 19811 + }, + { + "epoch": 0.8771902637725114, + "grad_norm": 1.3823508024215698, + "learning_rate": 9.056070346041101e-07, + "loss": 0.452, + "step": 19812 + }, + { + "epoch": 0.8772345394773253, + "grad_norm": 1.3070390224456787, + "learning_rate": 9.049644879883812e-07, + "loss": 0.4505, + "step": 19813 + }, + { + "epoch": 0.8772788151821391, + "grad_norm": 1.5071913003921509, + "learning_rate": 9.043221586001682e-07, + "loss": 0.4399, + "step": 19814 + }, + { + "epoch": 0.877323090886953, + "grad_norm": 1.920982003211975, + "learning_rate": 9.036800464548157e-07, + "loss": 0.4636, + "step": 19815 + }, + { + "epoch": 0.877367366591767, + "grad_norm": 1.5353541374206543, + "learning_rate": 9.030381515676611e-07, + "loss": 0.4417, + "step": 19816 + }, + { + "epoch": 0.8774116422965809, + "grad_norm": 1.6623029708862305, + "learning_rate": 9.023964739540291e-07, + "loss": 0.5915, + "step": 19817 + }, + { + "epoch": 0.8774559180013947, + "grad_norm": 1.9257750511169434, + "learning_rate": 9.017550136292518e-07, + "loss": 0.4807, + "step": 19818 + }, + { + "epoch": 0.8775001937062086, + "grad_norm": 1.2768367528915405, + "learning_rate": 9.011137706086481e-07, + "loss": 0.43, + "step": 19819 + }, + { + "epoch": 0.8775444694110225, + "grad_norm": 1.6106442213058472, + "learning_rate": 9.004727449075335e-07, + "loss": 0.4805, + "step": 19820 + }, + { + "epoch": 0.8775887451158363, + "grad_norm": 1.4768866300582886, + "learning_rate": 8.998319365412167e-07, + "loss": 0.4856, + "step": 19821 + }, + { + "epoch": 0.8776330208206502, + "grad_norm": 1.825140357017517, + "learning_rate": 8.991913455250057e-07, + "loss": 0.5201, + "step": 19822 + }, + { + "epoch": 0.8776772965254641, + "grad_norm": 1.4828534126281738, + "learning_rate": 8.985509718741992e-07, + "loss": 0.4718, + "step": 19823 + }, + { + "epoch": 0.8777215722302779, + "grad_norm": 1.7419418096542358, + "learning_rate": 8.979108156040905e-07, + "loss": 0.5869, + "step": 19824 + }, + { + "epoch": 0.8777658479350918, + "grad_norm": 1.5345478057861328, + "learning_rate": 8.972708767299754e-07, + "loss": 0.5385, + "step": 19825 + }, + { + "epoch": 0.8778101236399057, + "grad_norm": 1.4351441860198975, + "learning_rate": 8.966311552671303e-07, + "loss": 0.4439, + "step": 19826 + }, + { + "epoch": 0.8778543993447195, + "grad_norm": 1.3606666326522827, + "learning_rate": 8.959916512308387e-07, + "loss": 0.4086, + "step": 19827 + }, + { + "epoch": 0.8778986750495335, + "grad_norm": 1.7585810422897339, + "learning_rate": 8.953523646363715e-07, + "loss": 0.5417, + "step": 19828 + }, + { + "epoch": 0.8779429507543474, + "grad_norm": 1.7452170848846436, + "learning_rate": 8.947132954990012e-07, + "loss": 0.5469, + "step": 19829 + }, + { + "epoch": 0.8779872264591612, + "grad_norm": 1.4814341068267822, + "learning_rate": 8.94074443833991e-07, + "loss": 0.48, + "step": 19830 + }, + { + "epoch": 0.8780315021639751, + "grad_norm": 1.3517602682113647, + "learning_rate": 8.934358096565965e-07, + "loss": 0.4609, + "step": 19831 + }, + { + "epoch": 0.878075777868789, + "grad_norm": 1.4829944372177124, + "learning_rate": 8.927973929820744e-07, + "loss": 0.472, + "step": 19832 + }, + { + "epoch": 0.8781200535736028, + "grad_norm": 1.2721569538116455, + "learning_rate": 8.921591938256713e-07, + "loss": 0.431, + "step": 19833 + }, + { + "epoch": 0.8781643292784167, + "grad_norm": 1.245939016342163, + "learning_rate": 8.915212122026296e-07, + "loss": 0.4124, + "step": 19834 + }, + { + "epoch": 0.8782086049832306, + "grad_norm": 1.9410842657089233, + "learning_rate": 8.908834481281869e-07, + "loss": 0.5102, + "step": 19835 + }, + { + "epoch": 0.8782528806880444, + "grad_norm": 1.45529305934906, + "learning_rate": 8.902459016175768e-07, + "loss": 0.531, + "step": 19836 + }, + { + "epoch": 0.8782971563928583, + "grad_norm": 1.730777621269226, + "learning_rate": 8.89608572686026e-07, + "loss": 0.4928, + "step": 19837 + }, + { + "epoch": 0.8783414320976722, + "grad_norm": 1.4399925470352173, + "learning_rate": 8.889714613487554e-07, + "loss": 0.4616, + "step": 19838 + }, + { + "epoch": 0.878385707802486, + "grad_norm": 1.5870838165283203, + "learning_rate": 8.883345676209831e-07, + "loss": 0.4951, + "step": 19839 + }, + { + "epoch": 0.8784299835073, + "grad_norm": 1.394044041633606, + "learning_rate": 8.876978915179224e-07, + "loss": 0.471, + "step": 19840 + }, + { + "epoch": 0.8784742592121139, + "grad_norm": 1.5617941617965698, + "learning_rate": 8.870614330547745e-07, + "loss": 0.5143, + "step": 19841 + }, + { + "epoch": 0.8785185349169277, + "grad_norm": 1.3956174850463867, + "learning_rate": 8.864251922467471e-07, + "loss": 0.4625, + "step": 19842 + }, + { + "epoch": 0.8785628106217416, + "grad_norm": 1.6691389083862305, + "learning_rate": 8.857891691090336e-07, + "loss": 0.505, + "step": 19843 + }, + { + "epoch": 0.8786070863265555, + "grad_norm": 1.8206912279129028, + "learning_rate": 8.851533636568243e-07, + "loss": 0.4666, + "step": 19844 + }, + { + "epoch": 0.8786513620313693, + "grad_norm": 1.415614128112793, + "learning_rate": 8.845177759053047e-07, + "loss": 0.4864, + "step": 19845 + }, + { + "epoch": 0.8786956377361832, + "grad_norm": 2.7270939350128174, + "learning_rate": 8.838824058696571e-07, + "loss": 0.4849, + "step": 19846 + }, + { + "epoch": 0.8787399134409971, + "grad_norm": 1.7632639408111572, + "learning_rate": 8.832472535650549e-07, + "loss": 0.4627, + "step": 19847 + }, + { + "epoch": 0.8787841891458109, + "grad_norm": 1.3966082334518433, + "learning_rate": 8.826123190066671e-07, + "loss": 0.4521, + "step": 19848 + }, + { + "epoch": 0.8788284648506248, + "grad_norm": 1.478895902633667, + "learning_rate": 8.819776022096626e-07, + "loss": 0.518, + "step": 19849 + }, + { + "epoch": 0.8788727405554387, + "grad_norm": 2.154676914215088, + "learning_rate": 8.813431031891994e-07, + "loss": 0.455, + "step": 19850 + }, + { + "epoch": 0.8789170162602526, + "grad_norm": 3.616547107696533, + "learning_rate": 8.807088219604288e-07, + "loss": 0.6545, + "step": 19851 + }, + { + "epoch": 0.8789612919650664, + "grad_norm": 1.8376121520996094, + "learning_rate": 8.80074758538505e-07, + "loss": 0.5035, + "step": 19852 + }, + { + "epoch": 0.8790055676698804, + "grad_norm": 1.3971186876296997, + "learning_rate": 8.794409129385694e-07, + "loss": 0.4797, + "step": 19853 + }, + { + "epoch": 0.8790498433746943, + "grad_norm": 1.8337972164154053, + "learning_rate": 8.788072851757623e-07, + "loss": 0.5952, + "step": 19854 + }, + { + "epoch": 0.8790941190795081, + "grad_norm": 1.6967120170593262, + "learning_rate": 8.781738752652158e-07, + "loss": 0.4567, + "step": 19855 + }, + { + "epoch": 0.879138394784322, + "grad_norm": 1.394816279411316, + "learning_rate": 8.77540683222059e-07, + "loss": 0.4894, + "step": 19856 + }, + { + "epoch": 0.8791826704891359, + "grad_norm": 1.399509310722351, + "learning_rate": 8.769077090614142e-07, + "loss": 0.4717, + "step": 19857 + }, + { + "epoch": 0.8792269461939497, + "grad_norm": 1.6511304378509521, + "learning_rate": 8.762749527984005e-07, + "loss": 0.5387, + "step": 19858 + }, + { + "epoch": 0.8792712218987636, + "grad_norm": 1.6600546836853027, + "learning_rate": 8.756424144481313e-07, + "loss": 0.4863, + "step": 19859 + }, + { + "epoch": 0.8793154976035775, + "grad_norm": 1.6575994491577148, + "learning_rate": 8.750100940257145e-07, + "loss": 0.4419, + "step": 19860 + }, + { + "epoch": 0.8793597733083913, + "grad_norm": 1.4060276746749878, + "learning_rate": 8.743779915462492e-07, + "loss": 0.4477, + "step": 19861 + }, + { + "epoch": 0.8794040490132052, + "grad_norm": 1.9089832305908203, + "learning_rate": 8.737461070248376e-07, + "loss": 0.52, + "step": 19862 + }, + { + "epoch": 0.8794483247180191, + "grad_norm": 1.6890027523040771, + "learning_rate": 8.73114440476569e-07, + "loss": 0.5289, + "step": 19863 + }, + { + "epoch": 0.879492600422833, + "grad_norm": 1.6859540939331055, + "learning_rate": 8.724829919165312e-07, + "loss": 0.5236, + "step": 19864 + }, + { + "epoch": 0.8795368761276469, + "grad_norm": 1.736160397529602, + "learning_rate": 8.718517613598043e-07, + "loss": 0.5397, + "step": 19865 + }, + { + "epoch": 0.8795811518324608, + "grad_norm": 1.5353074073791504, + "learning_rate": 8.712207488214663e-07, + "loss": 0.448, + "step": 19866 + }, + { + "epoch": 0.8796254275372746, + "grad_norm": 1.5893818140029907, + "learning_rate": 8.705899543165874e-07, + "loss": 0.4988, + "step": 19867 + }, + { + "epoch": 0.8796697032420885, + "grad_norm": 1.6072072982788086, + "learning_rate": 8.699593778602333e-07, + "loss": 0.4735, + "step": 19868 + }, + { + "epoch": 0.8797139789469024, + "grad_norm": 1.4746263027191162, + "learning_rate": 8.693290194674664e-07, + "loss": 0.4469, + "step": 19869 + }, + { + "epoch": 0.8797582546517162, + "grad_norm": 1.2880492210388184, + "learning_rate": 8.686988791533412e-07, + "loss": 0.428, + "step": 19870 + }, + { + "epoch": 0.8798025303565301, + "grad_norm": 1.6126298904418945, + "learning_rate": 8.680689569329071e-07, + "loss": 0.5123, + "step": 19871 + }, + { + "epoch": 0.879846806061344, + "grad_norm": 1.7844818830490112, + "learning_rate": 8.674392528212139e-07, + "loss": 0.5381, + "step": 19872 + }, + { + "epoch": 0.8798910817661578, + "grad_norm": 1.5047887563705444, + "learning_rate": 8.668097668332953e-07, + "loss": 0.5035, + "step": 19873 + }, + { + "epoch": 0.8799353574709717, + "grad_norm": 1.4801082611083984, + "learning_rate": 8.661804989841882e-07, + "loss": 0.4305, + "step": 19874 + }, + { + "epoch": 0.8799796331757856, + "grad_norm": 1.5852129459381104, + "learning_rate": 8.655514492889249e-07, + "loss": 0.4672, + "step": 19875 + }, + { + "epoch": 0.8800239088805994, + "grad_norm": 4.0892720222473145, + "learning_rate": 8.64922617762527e-07, + "loss": 0.3921, + "step": 19876 + }, + { + "epoch": 0.8800681845854134, + "grad_norm": 1.2661328315734863, + "learning_rate": 8.642940044200143e-07, + "loss": 0.4544, + "step": 19877 + }, + { + "epoch": 0.8801124602902273, + "grad_norm": 1.1975079774856567, + "learning_rate": 8.636656092763995e-07, + "loss": 0.4165, + "step": 19878 + }, + { + "epoch": 0.8801567359950411, + "grad_norm": 1.5866516828536987, + "learning_rate": 8.63037432346694e-07, + "loss": 0.4711, + "step": 19879 + }, + { + "epoch": 0.880201011699855, + "grad_norm": 1.65431547164917, + "learning_rate": 8.624094736458998e-07, + "loss": 0.5108, + "step": 19880 + }, + { + "epoch": 0.8802452874046689, + "grad_norm": 1.386949896812439, + "learning_rate": 8.617817331890155e-07, + "loss": 0.4578, + "step": 19881 + }, + { + "epoch": 0.8802895631094827, + "grad_norm": 1.6547315120697021, + "learning_rate": 8.611542109910331e-07, + "loss": 0.5282, + "step": 19882 + }, + { + "epoch": 0.8803338388142966, + "grad_norm": 1.5017811059951782, + "learning_rate": 8.60526907066942e-07, + "loss": 0.4694, + "step": 19883 + }, + { + "epoch": 0.8803781145191105, + "grad_norm": 1.9486942291259766, + "learning_rate": 8.598998214317222e-07, + "loss": 0.53, + "step": 19884 + }, + { + "epoch": 0.8804223902239243, + "grad_norm": 1.6049340963363647, + "learning_rate": 8.592729541003541e-07, + "loss": 0.519, + "step": 19885 + }, + { + "epoch": 0.8804666659287382, + "grad_norm": 1.6416854858398438, + "learning_rate": 8.586463050878102e-07, + "loss": 0.513, + "step": 19886 + }, + { + "epoch": 0.8805109416335521, + "grad_norm": 1.5634844303131104, + "learning_rate": 8.58019874409055e-07, + "loss": 0.5048, + "step": 19887 + }, + { + "epoch": 0.8805552173383661, + "grad_norm": 2.018533229827881, + "learning_rate": 8.573936620790502e-07, + "loss": 0.455, + "step": 19888 + }, + { + "epoch": 0.8805994930431799, + "grad_norm": 1.2724123001098633, + "learning_rate": 8.567676681127546e-07, + "loss": 0.4379, + "step": 19889 + }, + { + "epoch": 0.8806437687479938, + "grad_norm": 1.5876435041427612, + "learning_rate": 8.561418925251197e-07, + "loss": 0.5064, + "step": 19890 + }, + { + "epoch": 0.8806880444528077, + "grad_norm": 1.3970179557800293, + "learning_rate": 8.555163353310902e-07, + "loss": 0.4546, + "step": 19891 + }, + { + "epoch": 0.8807323201576215, + "grad_norm": 2.0333518981933594, + "learning_rate": 8.548909965456064e-07, + "loss": 0.5659, + "step": 19892 + }, + { + "epoch": 0.8807765958624354, + "grad_norm": 1.6126755475997925, + "learning_rate": 8.542658761836065e-07, + "loss": 0.4703, + "step": 19893 + }, + { + "epoch": 0.8808208715672493, + "grad_norm": 1.3893929719924927, + "learning_rate": 8.53640974260016e-07, + "loss": 0.4301, + "step": 19894 + }, + { + "epoch": 0.8808651472720631, + "grad_norm": 1.7847979068756104, + "learning_rate": 8.530162907897666e-07, + "loss": 0.5116, + "step": 19895 + }, + { + "epoch": 0.880909422976877, + "grad_norm": 1.4940309524536133, + "learning_rate": 8.523918257877739e-07, + "loss": 0.5033, + "step": 19896 + }, + { + "epoch": 0.8809536986816909, + "grad_norm": 1.564105749130249, + "learning_rate": 8.517675792689561e-07, + "loss": 0.4906, + "step": 19897 + }, + { + "epoch": 0.8809979743865047, + "grad_norm": 1.4826912879943848, + "learning_rate": 8.51143551248218e-07, + "loss": 0.4771, + "step": 19898 + }, + { + "epoch": 0.8810422500913186, + "grad_norm": 2.210911750793457, + "learning_rate": 8.505197417404687e-07, + "loss": 0.6246, + "step": 19899 + }, + { + "epoch": 0.8810865257961326, + "grad_norm": 1.5211608409881592, + "learning_rate": 8.498961507606052e-07, + "loss": 0.4953, + "step": 19900 + }, + { + "epoch": 0.8811308015009464, + "grad_norm": 1.589155673980713, + "learning_rate": 8.492727783235222e-07, + "loss": 0.4578, + "step": 19901 + }, + { + "epoch": 0.8811750772057603, + "grad_norm": 1.4870730638504028, + "learning_rate": 8.486496244441089e-07, + "loss": 0.4831, + "step": 19902 + }, + { + "epoch": 0.8812193529105742, + "grad_norm": 1.2936440706253052, + "learning_rate": 8.480266891372469e-07, + "loss": 0.4626, + "step": 19903 + }, + { + "epoch": 0.881263628615388, + "grad_norm": 1.4841982126235962, + "learning_rate": 8.474039724178141e-07, + "loss": 0.4691, + "step": 19904 + }, + { + "epoch": 0.8813079043202019, + "grad_norm": 1.6975868940353394, + "learning_rate": 8.467814743006875e-07, + "loss": 0.5027, + "step": 19905 + }, + { + "epoch": 0.8813521800250158, + "grad_norm": 1.602620005607605, + "learning_rate": 8.461591948007319e-07, + "loss": 0.465, + "step": 19906 + }, + { + "epoch": 0.8813964557298296, + "grad_norm": 1.4555484056472778, + "learning_rate": 8.455371339328111e-07, + "loss": 0.4713, + "step": 19907 + }, + { + "epoch": 0.8814407314346435, + "grad_norm": 1.4053395986557007, + "learning_rate": 8.449152917117797e-07, + "loss": 0.3679, + "step": 19908 + }, + { + "epoch": 0.8814850071394574, + "grad_norm": 1.6728484630584717, + "learning_rate": 8.442936681524938e-07, + "loss": 0.5519, + "step": 19909 + }, + { + "epoch": 0.8815292828442712, + "grad_norm": 1.4842369556427002, + "learning_rate": 8.436722632698003e-07, + "loss": 0.4613, + "step": 19910 + }, + { + "epoch": 0.8815735585490851, + "grad_norm": 2.3209340572357178, + "learning_rate": 8.430510770785383e-07, + "loss": 0.5331, + "step": 19911 + }, + { + "epoch": 0.8816178342538991, + "grad_norm": 1.6175422668457031, + "learning_rate": 8.42430109593545e-07, + "loss": 0.5015, + "step": 19912 + }, + { + "epoch": 0.8816621099587129, + "grad_norm": 1.602886438369751, + "learning_rate": 8.418093608296529e-07, + "loss": 0.499, + "step": 19913 + }, + { + "epoch": 0.8817063856635268, + "grad_norm": 1.540784239768982, + "learning_rate": 8.411888308016846e-07, + "loss": 0.5069, + "step": 19914 + }, + { + "epoch": 0.8817506613683407, + "grad_norm": 1.7357616424560547, + "learning_rate": 8.405685195244672e-07, + "loss": 0.5272, + "step": 19915 + }, + { + "epoch": 0.8817949370731545, + "grad_norm": 1.68069589138031, + "learning_rate": 8.399484270128111e-07, + "loss": 0.5062, + "step": 19916 + }, + { + "epoch": 0.8818392127779684, + "grad_norm": 1.6143617630004883, + "learning_rate": 8.393285532815298e-07, + "loss": 0.468, + "step": 19917 + }, + { + "epoch": 0.8818834884827823, + "grad_norm": 2.142169713973999, + "learning_rate": 8.38708898345424e-07, + "loss": 0.5091, + "step": 19918 + }, + { + "epoch": 0.8819277641875961, + "grad_norm": 1.7197448015213013, + "learning_rate": 8.380894622193015e-07, + "loss": 0.46, + "step": 19919 + }, + { + "epoch": 0.88197203989241, + "grad_norm": 1.4400651454925537, + "learning_rate": 8.374702449179495e-07, + "loss": 0.4726, + "step": 19920 + }, + { + "epoch": 0.8820163155972239, + "grad_norm": 1.482591986656189, + "learning_rate": 8.368512464561596e-07, + "loss": 0.4952, + "step": 19921 + }, + { + "epoch": 0.8820605913020378, + "grad_norm": 1.4999752044677734, + "learning_rate": 8.362324668487177e-07, + "loss": 0.4658, + "step": 19922 + }, + { + "epoch": 0.8821048670068516, + "grad_norm": 1.4480564594268799, + "learning_rate": 8.356139061104029e-07, + "loss": 0.4451, + "step": 19923 + }, + { + "epoch": 0.8821491427116656, + "grad_norm": 1.354476809501648, + "learning_rate": 8.349955642559848e-07, + "loss": 0.4809, + "step": 19924 + }, + { + "epoch": 0.8821934184164795, + "grad_norm": 1.677781581878662, + "learning_rate": 8.343774413002382e-07, + "loss": 0.484, + "step": 19925 + }, + { + "epoch": 0.8822376941212933, + "grad_norm": 1.3272411823272705, + "learning_rate": 8.337595372579233e-07, + "loss": 0.4331, + "step": 19926 + }, + { + "epoch": 0.8822819698261072, + "grad_norm": 1.4194170236587524, + "learning_rate": 8.331418521437973e-07, + "loss": 0.4086, + "step": 19927 + }, + { + "epoch": 0.8823262455309211, + "grad_norm": 1.9700106382369995, + "learning_rate": 8.325243859726151e-07, + "loss": 0.4946, + "step": 19928 + }, + { + "epoch": 0.8823705212357349, + "grad_norm": 1.6920408010482788, + "learning_rate": 8.319071387591237e-07, + "loss": 0.5454, + "step": 19929 + }, + { + "epoch": 0.8824147969405488, + "grad_norm": 1.5856518745422363, + "learning_rate": 8.31290110518066e-07, + "loss": 0.4781, + "step": 19930 + }, + { + "epoch": 0.8824590726453627, + "grad_norm": 1.642382264137268, + "learning_rate": 8.306733012641755e-07, + "loss": 0.4507, + "step": 19931 + }, + { + "epoch": 0.8825033483501765, + "grad_norm": 1.6165249347686768, + "learning_rate": 8.300567110121893e-07, + "loss": 0.4996, + "step": 19932 + }, + { + "epoch": 0.8825476240549904, + "grad_norm": 1.573396921157837, + "learning_rate": 8.294403397768325e-07, + "loss": 0.5156, + "step": 19933 + }, + { + "epoch": 0.8825918997598043, + "grad_norm": 2.0099031925201416, + "learning_rate": 8.288241875728242e-07, + "loss": 0.568, + "step": 19934 + }, + { + "epoch": 0.8826361754646181, + "grad_norm": 1.4817625284194946, + "learning_rate": 8.28208254414884e-07, + "loss": 0.519, + "step": 19935 + }, + { + "epoch": 0.8826804511694321, + "grad_norm": 1.5247457027435303, + "learning_rate": 8.275925403177232e-07, + "loss": 0.4633, + "step": 19936 + }, + { + "epoch": 0.882724726874246, + "grad_norm": 1.6312464475631714, + "learning_rate": 8.269770452960446e-07, + "loss": 0.5192, + "step": 19937 + }, + { + "epoch": 0.8827690025790598, + "grad_norm": 1.3786579370498657, + "learning_rate": 8.26361769364551e-07, + "loss": 0.414, + "step": 19938 + }, + { + "epoch": 0.8828132782838737, + "grad_norm": 1.5103579759597778, + "learning_rate": 8.25746712537937e-07, + "loss": 0.4758, + "step": 19939 + }, + { + "epoch": 0.8828575539886876, + "grad_norm": 1.7374117374420166, + "learning_rate": 8.251318748308923e-07, + "loss": 0.5104, + "step": 19940 + }, + { + "epoch": 0.8829018296935014, + "grad_norm": 1.5718151330947876, + "learning_rate": 8.245172562581017e-07, + "loss": 0.4561, + "step": 19941 + }, + { + "epoch": 0.8829461053983153, + "grad_norm": 1.4800167083740234, + "learning_rate": 8.239028568342466e-07, + "loss": 0.4783, + "step": 19942 + }, + { + "epoch": 0.8829903811031292, + "grad_norm": 1.6590917110443115, + "learning_rate": 8.232886765740001e-07, + "loss": 0.5002, + "step": 19943 + }, + { + "epoch": 0.883034656807943, + "grad_norm": 1.5151318311691284, + "learning_rate": 8.226747154920323e-07, + "loss": 0.4278, + "step": 19944 + }, + { + "epoch": 0.8830789325127569, + "grad_norm": 1.5161257982254028, + "learning_rate": 8.220609736030038e-07, + "loss": 0.4596, + "step": 19945 + }, + { + "epoch": 0.8831232082175708, + "grad_norm": 1.9049248695373535, + "learning_rate": 8.214474509215787e-07, + "loss": 0.5044, + "step": 19946 + }, + { + "epoch": 0.8831674839223846, + "grad_norm": 1.589934229850769, + "learning_rate": 8.208341474624071e-07, + "loss": 0.5212, + "step": 19947 + }, + { + "epoch": 0.8832117596271986, + "grad_norm": 1.6602299213409424, + "learning_rate": 8.202210632401375e-07, + "loss": 0.5544, + "step": 19948 + }, + { + "epoch": 0.8832560353320125, + "grad_norm": 1.424088478088379, + "learning_rate": 8.196081982694137e-07, + "loss": 0.4499, + "step": 19949 + }, + { + "epoch": 0.8833003110368263, + "grad_norm": 1.46400785446167, + "learning_rate": 8.189955525648741e-07, + "loss": 0.4988, + "step": 19950 + }, + { + "epoch": 0.8833445867416402, + "grad_norm": 1.403044581413269, + "learning_rate": 8.183831261411479e-07, + "loss": 0.4485, + "step": 19951 + }, + { + "epoch": 0.8833888624464541, + "grad_norm": 1.238967776298523, + "learning_rate": 8.177709190128658e-07, + "loss": 0.384, + "step": 19952 + }, + { + "epoch": 0.8834331381512679, + "grad_norm": 1.4647587537765503, + "learning_rate": 8.171589311946493e-07, + "loss": 0.4967, + "step": 19953 + }, + { + "epoch": 0.8834774138560818, + "grad_norm": 1.6991623640060425, + "learning_rate": 8.165471627011157e-07, + "loss": 0.5738, + "step": 19954 + }, + { + "epoch": 0.8835216895608957, + "grad_norm": 1.401787281036377, + "learning_rate": 8.159356135468721e-07, + "loss": 0.4384, + "step": 19955 + }, + { + "epoch": 0.8835659652657095, + "grad_norm": 1.6665806770324707, + "learning_rate": 8.153242837465314e-07, + "loss": 0.4562, + "step": 19956 + }, + { + "epoch": 0.8836102409705234, + "grad_norm": 2.1371119022369385, + "learning_rate": 8.147131733146918e-07, + "loss": 0.5495, + "step": 19957 + }, + { + "epoch": 0.8836545166753373, + "grad_norm": 1.5458024740219116, + "learning_rate": 8.141022822659495e-07, + "loss": 0.5109, + "step": 19958 + }, + { + "epoch": 0.8836987923801513, + "grad_norm": 1.8528265953063965, + "learning_rate": 8.134916106148938e-07, + "loss": 0.5266, + "step": 19959 + }, + { + "epoch": 0.883743068084965, + "grad_norm": 1.6700019836425781, + "learning_rate": 8.128811583761109e-07, + "loss": 0.4853, + "step": 19960 + }, + { + "epoch": 0.883787343789779, + "grad_norm": 1.471145749092102, + "learning_rate": 8.122709255641803e-07, + "loss": 0.4677, + "step": 19961 + }, + { + "epoch": 0.8838316194945929, + "grad_norm": 1.429867148399353, + "learning_rate": 8.116609121936791e-07, + "loss": 0.4775, + "step": 19962 + }, + { + "epoch": 0.8838758951994067, + "grad_norm": 1.438379168510437, + "learning_rate": 8.110511182791758e-07, + "loss": 0.4393, + "step": 19963 + }, + { + "epoch": 0.8839201709042206, + "grad_norm": 1.662306547164917, + "learning_rate": 8.104415438352342e-07, + "loss": 0.5259, + "step": 19964 + }, + { + "epoch": 0.8839644466090345, + "grad_norm": 1.4926421642303467, + "learning_rate": 8.098321888764116e-07, + "loss": 0.4415, + "step": 19965 + }, + { + "epoch": 0.8840087223138483, + "grad_norm": 1.514176368713379, + "learning_rate": 8.092230534172674e-07, + "loss": 0.4437, + "step": 19966 + }, + { + "epoch": 0.8840529980186622, + "grad_norm": 1.831891417503357, + "learning_rate": 8.086141374723477e-07, + "loss": 0.5589, + "step": 19967 + }, + { + "epoch": 0.8840972737234761, + "grad_norm": 1.4186164140701294, + "learning_rate": 8.080054410561921e-07, + "loss": 0.461, + "step": 19968 + }, + { + "epoch": 0.8841415494282899, + "grad_norm": 1.4701286554336548, + "learning_rate": 8.073969641833446e-07, + "loss": 0.5126, + "step": 19969 + }, + { + "epoch": 0.8841858251331038, + "grad_norm": 1.426712989807129, + "learning_rate": 8.067887068683356e-07, + "loss": 0.4236, + "step": 19970 + }, + { + "epoch": 0.8842301008379178, + "grad_norm": 1.4651390314102173, + "learning_rate": 8.061806691256913e-07, + "loss": 0.4824, + "step": 19971 + }, + { + "epoch": 0.8842743765427316, + "grad_norm": 1.5687406063079834, + "learning_rate": 8.055728509699367e-07, + "loss": 0.499, + "step": 19972 + }, + { + "epoch": 0.8843186522475455, + "grad_norm": 1.4275532960891724, + "learning_rate": 8.049652524155882e-07, + "loss": 0.5029, + "step": 19973 + }, + { + "epoch": 0.8843629279523594, + "grad_norm": 1.6113274097442627, + "learning_rate": 8.043578734771585e-07, + "loss": 0.468, + "step": 19974 + }, + { + "epoch": 0.8844072036571732, + "grad_norm": 1.4662904739379883, + "learning_rate": 8.037507141691536e-07, + "loss": 0.4486, + "step": 19975 + }, + { + "epoch": 0.8844514793619871, + "grad_norm": 1.319107174873352, + "learning_rate": 8.031437745060744e-07, + "loss": 0.4261, + "step": 19976 + }, + { + "epoch": 0.884495755066801, + "grad_norm": 1.9161869287490845, + "learning_rate": 8.025370545024192e-07, + "loss": 0.4761, + "step": 19977 + }, + { + "epoch": 0.8845400307716148, + "grad_norm": 1.6534497737884521, + "learning_rate": 8.019305541726741e-07, + "loss": 0.5522, + "step": 19978 + }, + { + "epoch": 0.8845843064764287, + "grad_norm": 1.2434031963348389, + "learning_rate": 8.013242735313309e-07, + "loss": 0.4099, + "step": 19979 + }, + { + "epoch": 0.8846285821812426, + "grad_norm": 1.3802241086959839, + "learning_rate": 8.007182125928681e-07, + "loss": 0.4147, + "step": 19980 + }, + { + "epoch": 0.8846728578860564, + "grad_norm": 1.5890474319458008, + "learning_rate": 8.001123713717573e-07, + "loss": 0.4454, + "step": 19981 + }, + { + "epoch": 0.8847171335908703, + "grad_norm": 1.445926308631897, + "learning_rate": 7.995067498824749e-07, + "loss": 0.4829, + "step": 19982 + }, + { + "epoch": 0.8847614092956843, + "grad_norm": 1.4514707326889038, + "learning_rate": 7.989013481394813e-07, + "loss": 0.4927, + "step": 19983 + }, + { + "epoch": 0.884805685000498, + "grad_norm": 1.7194466590881348, + "learning_rate": 7.982961661572385e-07, + "loss": 0.4515, + "step": 19984 + }, + { + "epoch": 0.884849960705312, + "grad_norm": 3.749242067337036, + "learning_rate": 7.976912039501982e-07, + "loss": 0.4945, + "step": 19985 + }, + { + "epoch": 0.8848942364101259, + "grad_norm": 1.477003812789917, + "learning_rate": 7.970864615328111e-07, + "loss": 0.4511, + "step": 19986 + }, + { + "epoch": 0.8849385121149397, + "grad_norm": 1.6810182332992554, + "learning_rate": 7.964819389195211e-07, + "loss": 0.5099, + "step": 19987 + }, + { + "epoch": 0.8849827878197536, + "grad_norm": 1.668026089668274, + "learning_rate": 7.958776361247634e-07, + "loss": 0.467, + "step": 19988 + }, + { + "epoch": 0.8850270635245675, + "grad_norm": 1.817463994026184, + "learning_rate": 7.952735531629774e-07, + "loss": 0.5642, + "step": 19989 + }, + { + "epoch": 0.8850713392293813, + "grad_norm": 2.521204710006714, + "learning_rate": 7.946696900485862e-07, + "loss": 0.4783, + "step": 19990 + }, + { + "epoch": 0.8851156149341952, + "grad_norm": 2.113511562347412, + "learning_rate": 7.940660467960137e-07, + "loss": 0.5891, + "step": 19991 + }, + { + "epoch": 0.8851598906390091, + "grad_norm": 1.6868141889572144, + "learning_rate": 7.934626234196785e-07, + "loss": 0.5345, + "step": 19992 + }, + { + "epoch": 0.885204166343823, + "grad_norm": 1.6711066961288452, + "learning_rate": 7.928594199339934e-07, + "loss": 0.4762, + "step": 19993 + }, + { + "epoch": 0.8852484420486368, + "grad_norm": 1.5268449783325195, + "learning_rate": 7.922564363533636e-07, + "loss": 0.4623, + "step": 19994 + }, + { + "epoch": 0.8852927177534508, + "grad_norm": 1.5790386199951172, + "learning_rate": 7.91653672692192e-07, + "loss": 0.4863, + "step": 19995 + }, + { + "epoch": 0.8853369934582647, + "grad_norm": 1.5028362274169922, + "learning_rate": 7.910511289648748e-07, + "loss": 0.4865, + "step": 19996 + }, + { + "epoch": 0.8853812691630785, + "grad_norm": 1.5051442384719849, + "learning_rate": 7.904488051858028e-07, + "loss": 0.502, + "step": 19997 + }, + { + "epoch": 0.8854255448678924, + "grad_norm": 1.7934738397598267, + "learning_rate": 7.8984670136936e-07, + "loss": 0.5033, + "step": 19998 + }, + { + "epoch": 0.8854698205727063, + "grad_norm": 1.5390018224716187, + "learning_rate": 7.892448175299328e-07, + "loss": 0.4786, + "step": 19999 + }, + { + "epoch": 0.8855140962775201, + "grad_norm": 1.8744817972183228, + "learning_rate": 7.886431536818929e-07, + "loss": 0.47, + "step": 20000 + }, + { + "epoch": 0.885558371982334, + "grad_norm": 1.6516903638839722, + "learning_rate": 7.880417098396087e-07, + "loss": 0.4945, + "step": 20001 + }, + { + "epoch": 0.8856026476871479, + "grad_norm": 1.2918095588684082, + "learning_rate": 7.8744048601745e-07, + "loss": 0.4572, + "step": 20002 + }, + { + "epoch": 0.8856469233919617, + "grad_norm": 1.9505048990249634, + "learning_rate": 7.868394822297742e-07, + "loss": 0.4873, + "step": 20003 + }, + { + "epoch": 0.8856911990967756, + "grad_norm": 2.5844297409057617, + "learning_rate": 7.862386984909354e-07, + "loss": 0.5311, + "step": 20004 + }, + { + "epoch": 0.8857354748015895, + "grad_norm": 1.2842833995819092, + "learning_rate": 7.856381348152831e-07, + "loss": 0.4393, + "step": 20005 + }, + { + "epoch": 0.8857797505064033, + "grad_norm": 1.5921039581298828, + "learning_rate": 7.850377912171614e-07, + "loss": 0.489, + "step": 20006 + }, + { + "epoch": 0.8858240262112173, + "grad_norm": 1.3023781776428223, + "learning_rate": 7.844376677109089e-07, + "loss": 0.4707, + "step": 20007 + }, + { + "epoch": 0.8858683019160312, + "grad_norm": 1.3963619470596313, + "learning_rate": 7.838377643108575e-07, + "loss": 0.4172, + "step": 20008 + }, + { + "epoch": 0.885912577620845, + "grad_norm": 1.600298523902893, + "learning_rate": 7.832380810313389e-07, + "loss": 0.538, + "step": 20009 + }, + { + "epoch": 0.8859568533256589, + "grad_norm": 2.4815728664398193, + "learning_rate": 7.82638617886674e-07, + "loss": 0.4696, + "step": 20010 + }, + { + "epoch": 0.8860011290304728, + "grad_norm": 2.0382587909698486, + "learning_rate": 7.820393748911792e-07, + "loss": 0.5281, + "step": 20011 + }, + { + "epoch": 0.8860454047352866, + "grad_norm": 1.1765844821929932, + "learning_rate": 7.814403520591696e-07, + "loss": 0.4093, + "step": 20012 + }, + { + "epoch": 0.8860896804401005, + "grad_norm": 1.2522042989730835, + "learning_rate": 7.808415494049515e-07, + "loss": 0.4528, + "step": 20013 + }, + { + "epoch": 0.8861339561449144, + "grad_norm": 1.5461537837982178, + "learning_rate": 7.80242966942828e-07, + "loss": 0.5081, + "step": 20014 + }, + { + "epoch": 0.8861782318497282, + "grad_norm": 1.462935209274292, + "learning_rate": 7.796446046870909e-07, + "loss": 0.4795, + "step": 20015 + }, + { + "epoch": 0.8862225075545421, + "grad_norm": 2.06955623626709, + "learning_rate": 7.790464626520366e-07, + "loss": 0.4885, + "step": 20016 + }, + { + "epoch": 0.886266783259356, + "grad_norm": 1.65773344039917, + "learning_rate": 7.784485408519504e-07, + "loss": 0.526, + "step": 20017 + }, + { + "epoch": 0.8863110589641698, + "grad_norm": 1.6464351415634155, + "learning_rate": 7.778508393011097e-07, + "loss": 0.5388, + "step": 20018 + }, + { + "epoch": 0.8863553346689838, + "grad_norm": 1.4298454523086548, + "learning_rate": 7.772533580137942e-07, + "loss": 0.4834, + "step": 20019 + }, + { + "epoch": 0.8863996103737977, + "grad_norm": 1.4259454011917114, + "learning_rate": 7.766560970042736e-07, + "loss": 0.4497, + "step": 20020 + }, + { + "epoch": 0.8864438860786115, + "grad_norm": 1.318355917930603, + "learning_rate": 7.760590562868098e-07, + "loss": 0.5015, + "step": 20021 + }, + { + "epoch": 0.8864881617834254, + "grad_norm": 2.5540645122528076, + "learning_rate": 7.754622358756669e-07, + "loss": 0.4925, + "step": 20022 + }, + { + "epoch": 0.8865324374882393, + "grad_norm": 1.667543649673462, + "learning_rate": 7.748656357851003e-07, + "loss": 0.4921, + "step": 20023 + }, + { + "epoch": 0.8865767131930531, + "grad_norm": 1.301815390586853, + "learning_rate": 7.742692560293542e-07, + "loss": 0.4578, + "step": 20024 + }, + { + "epoch": 0.886620988897867, + "grad_norm": 1.4964441061019897, + "learning_rate": 7.736730966226725e-07, + "loss": 0.515, + "step": 20025 + }, + { + "epoch": 0.8866652646026809, + "grad_norm": 1.6114202737808228, + "learning_rate": 7.730771575792995e-07, + "loss": 0.4703, + "step": 20026 + }, + { + "epoch": 0.8867095403074947, + "grad_norm": 1.5474742650985718, + "learning_rate": 7.724814389134649e-07, + "loss": 0.4661, + "step": 20027 + }, + { + "epoch": 0.8867538160123086, + "grad_norm": 1.7740033864974976, + "learning_rate": 7.718859406393964e-07, + "loss": 0.5003, + "step": 20028 + }, + { + "epoch": 0.8867980917171225, + "grad_norm": 1.4329752922058105, + "learning_rate": 7.712906627713201e-07, + "loss": 0.4965, + "step": 20029 + }, + { + "epoch": 0.8868423674219365, + "grad_norm": 1.7469971179962158, + "learning_rate": 7.706956053234527e-07, + "loss": 0.5543, + "step": 20030 + }, + { + "epoch": 0.8868866431267503, + "grad_norm": 1.9889408349990845, + "learning_rate": 7.701007683100048e-07, + "loss": 0.5858, + "step": 20031 + }, + { + "epoch": 0.8869309188315642, + "grad_norm": 1.5519587993621826, + "learning_rate": 7.695061517451852e-07, + "loss": 0.4969, + "step": 20032 + }, + { + "epoch": 0.8869751945363781, + "grad_norm": 1.4059638977050781, + "learning_rate": 7.689117556431958e-07, + "loss": 0.4496, + "step": 20033 + }, + { + "epoch": 0.8870194702411919, + "grad_norm": 1.5578595399856567, + "learning_rate": 7.68317580018233e-07, + "loss": 0.4741, + "step": 20034 + }, + { + "epoch": 0.8870637459460058, + "grad_norm": 1.7665927410125732, + "learning_rate": 7.677236248844855e-07, + "loss": 0.5056, + "step": 20035 + }, + { + "epoch": 0.8871080216508197, + "grad_norm": 1.3658450841903687, + "learning_rate": 7.671298902561442e-07, + "loss": 0.4706, + "step": 20036 + }, + { + "epoch": 0.8871522973556335, + "grad_norm": 1.5784071683883667, + "learning_rate": 7.665363761473877e-07, + "loss": 0.4804, + "step": 20037 + }, + { + "epoch": 0.8871965730604474, + "grad_norm": 1.5462886095046997, + "learning_rate": 7.659430825723902e-07, + "loss": 0.5247, + "step": 20038 + }, + { + "epoch": 0.8872408487652613, + "grad_norm": 1.539731740951538, + "learning_rate": 7.653500095453248e-07, + "loss": 0.446, + "step": 20039 + }, + { + "epoch": 0.8872851244700751, + "grad_norm": 1.566003441810608, + "learning_rate": 7.647571570803558e-07, + "loss": 0.5177, + "step": 20040 + }, + { + "epoch": 0.887329400174889, + "grad_norm": 2.0178122520446777, + "learning_rate": 7.64164525191643e-07, + "loss": 0.5432, + "step": 20041 + }, + { + "epoch": 0.887373675879703, + "grad_norm": 1.414721965789795, + "learning_rate": 7.635721138933394e-07, + "loss": 0.4945, + "step": 20042 + }, + { + "epoch": 0.8874179515845168, + "grad_norm": 1.4234602451324463, + "learning_rate": 7.629799231995961e-07, + "loss": 0.4885, + "step": 20043 + }, + { + "epoch": 0.8874622272893307, + "grad_norm": 1.570103645324707, + "learning_rate": 7.623879531245548e-07, + "loss": 0.5136, + "step": 20044 + }, + { + "epoch": 0.8875065029941446, + "grad_norm": 1.7006709575653076, + "learning_rate": 7.617962036823556e-07, + "loss": 0.5209, + "step": 20045 + }, + { + "epoch": 0.8875507786989584, + "grad_norm": 1.6984459161758423, + "learning_rate": 7.612046748871327e-07, + "loss": 0.5307, + "step": 20046 + }, + { + "epoch": 0.8875950544037723, + "grad_norm": 1.553808331489563, + "learning_rate": 7.606133667530147e-07, + "loss": 0.4813, + "step": 20047 + }, + { + "epoch": 0.8876393301085862, + "grad_norm": 1.1596076488494873, + "learning_rate": 7.600222792941203e-07, + "loss": 0.4156, + "step": 20048 + }, + { + "epoch": 0.8876836058134, + "grad_norm": 1.494469404220581, + "learning_rate": 7.594314125245728e-07, + "loss": 0.4382, + "step": 20049 + }, + { + "epoch": 0.8877278815182139, + "grad_norm": 1.6288068294525146, + "learning_rate": 7.58840766458483e-07, + "loss": 0.4947, + "step": 20050 + }, + { + "epoch": 0.8877721572230278, + "grad_norm": 1.6518564224243164, + "learning_rate": 7.582503411099551e-07, + "loss": 0.5332, + "step": 20051 + }, + { + "epoch": 0.8878164329278416, + "grad_norm": 1.3948556184768677, + "learning_rate": 7.576601364930947e-07, + "loss": 0.4286, + "step": 20052 + }, + { + "epoch": 0.8878607086326555, + "grad_norm": 1.4535824060440063, + "learning_rate": 7.570701526219959e-07, + "loss": 0.4245, + "step": 20053 + }, + { + "epoch": 0.8879049843374694, + "grad_norm": 1.865677833557129, + "learning_rate": 7.564803895107497e-07, + "loss": 0.5178, + "step": 20054 + }, + { + "epoch": 0.8879492600422833, + "grad_norm": 1.7276403903961182, + "learning_rate": 7.558908471734428e-07, + "loss": 0.5032, + "step": 20055 + }, + { + "epoch": 0.8879935357470972, + "grad_norm": 1.4222954511642456, + "learning_rate": 7.553015256241569e-07, + "loss": 0.4703, + "step": 20056 + }, + { + "epoch": 0.8880378114519111, + "grad_norm": 1.2207658290863037, + "learning_rate": 7.547124248769666e-07, + "loss": 0.4216, + "step": 20057 + }, + { + "epoch": 0.8880820871567249, + "grad_norm": 2.500760555267334, + "learning_rate": 7.541235449459417e-07, + "loss": 0.5084, + "step": 20058 + }, + { + "epoch": 0.8881263628615388, + "grad_norm": 1.516697645187378, + "learning_rate": 7.535348858451486e-07, + "loss": 0.4028, + "step": 20059 + }, + { + "epoch": 0.8881706385663527, + "grad_norm": 1.7315484285354614, + "learning_rate": 7.529464475886461e-07, + "loss": 0.504, + "step": 20060 + }, + { + "epoch": 0.8882149142711665, + "grad_norm": 1.5824397802352905, + "learning_rate": 7.523582301904886e-07, + "loss": 0.5459, + "step": 20061 + }, + { + "epoch": 0.8882591899759804, + "grad_norm": 1.653334140777588, + "learning_rate": 7.517702336647248e-07, + "loss": 0.4787, + "step": 20062 + }, + { + "epoch": 0.8883034656807943, + "grad_norm": 1.3474870920181274, + "learning_rate": 7.51182458025399e-07, + "loss": 0.4609, + "step": 20063 + }, + { + "epoch": 0.8883477413856082, + "grad_norm": 1.6459261178970337, + "learning_rate": 7.505949032865489e-07, + "loss": 0.5145, + "step": 20064 + }, + { + "epoch": 0.888392017090422, + "grad_norm": 1.4347429275512695, + "learning_rate": 7.500075694622077e-07, + "loss": 0.4352, + "step": 20065 + }, + { + "epoch": 0.888436292795236, + "grad_norm": 1.4002599716186523, + "learning_rate": 7.494204565664054e-07, + "loss": 0.4594, + "step": 20066 + }, + { + "epoch": 0.8884805685000499, + "grad_norm": 1.2666245698928833, + "learning_rate": 7.488335646131628e-07, + "loss": 0.4219, + "step": 20067 + }, + { + "epoch": 0.8885248442048637, + "grad_norm": 1.4598841667175293, + "learning_rate": 7.482468936164966e-07, + "loss": 0.4936, + "step": 20068 + }, + { + "epoch": 0.8885691199096776, + "grad_norm": 1.5898268222808838, + "learning_rate": 7.476604435904222e-07, + "loss": 0.5251, + "step": 20069 + }, + { + "epoch": 0.8886133956144915, + "grad_norm": 1.938205599784851, + "learning_rate": 7.470742145489462e-07, + "loss": 0.5442, + "step": 20070 + }, + { + "epoch": 0.8886576713193053, + "grad_norm": 1.7599459886550903, + "learning_rate": 7.464882065060664e-07, + "loss": 0.5024, + "step": 20071 + }, + { + "epoch": 0.8887019470241192, + "grad_norm": 1.32582426071167, + "learning_rate": 7.459024194757802e-07, + "loss": 0.4681, + "step": 20072 + }, + { + "epoch": 0.8887462227289331, + "grad_norm": 1.6313823461532593, + "learning_rate": 7.453168534720811e-07, + "loss": 0.5038, + "step": 20073 + }, + { + "epoch": 0.8887904984337469, + "grad_norm": 1.4053281545639038, + "learning_rate": 7.447315085089535e-07, + "loss": 0.464, + "step": 20074 + }, + { + "epoch": 0.8888347741385608, + "grad_norm": 2.085092067718506, + "learning_rate": 7.441463846003772e-07, + "loss": 0.5355, + "step": 20075 + }, + { + "epoch": 0.8888790498433747, + "grad_norm": 1.6958032846450806, + "learning_rate": 7.435614817603287e-07, + "loss": 0.5425, + "step": 20076 + }, + { + "epoch": 0.8889233255481885, + "grad_norm": 1.6163536310195923, + "learning_rate": 7.429768000027782e-07, + "loss": 0.5255, + "step": 20077 + }, + { + "epoch": 0.8889676012530024, + "grad_norm": 1.431542158126831, + "learning_rate": 7.423923393416866e-07, + "loss": 0.4686, + "step": 20078 + }, + { + "epoch": 0.8890118769578164, + "grad_norm": 1.3620197772979736, + "learning_rate": 7.418080997910204e-07, + "loss": 0.4473, + "step": 20079 + }, + { + "epoch": 0.8890561526626302, + "grad_norm": 1.3547221422195435, + "learning_rate": 7.412240813647275e-07, + "loss": 0.4295, + "step": 20080 + }, + { + "epoch": 0.8891004283674441, + "grad_norm": 1.5034711360931396, + "learning_rate": 7.406402840767579e-07, + "loss": 0.4614, + "step": 20081 + }, + { + "epoch": 0.889144704072258, + "grad_norm": 1.7007662057876587, + "learning_rate": 7.400567079410559e-07, + "loss": 0.5149, + "step": 20082 + }, + { + "epoch": 0.8891889797770718, + "grad_norm": 1.194882869720459, + "learning_rate": 7.394733529715603e-07, + "loss": 0.4456, + "step": 20083 + }, + { + "epoch": 0.8892332554818857, + "grad_norm": 1.6885435581207275, + "learning_rate": 7.388902191822034e-07, + "loss": 0.4898, + "step": 20084 + }, + { + "epoch": 0.8892775311866996, + "grad_norm": 1.9698327779769897, + "learning_rate": 7.383073065869117e-07, + "loss": 0.5445, + "step": 20085 + }, + { + "epoch": 0.8893218068915134, + "grad_norm": 1.460304617881775, + "learning_rate": 7.377246151996099e-07, + "loss": 0.4659, + "step": 20086 + }, + { + "epoch": 0.8893660825963273, + "grad_norm": 1.2996349334716797, + "learning_rate": 7.371421450342142e-07, + "loss": 0.4354, + "step": 20087 + }, + { + "epoch": 0.8894103583011412, + "grad_norm": 1.6064727306365967, + "learning_rate": 7.365598961046361e-07, + "loss": 0.5113, + "step": 20088 + }, + { + "epoch": 0.889454634005955, + "grad_norm": 1.5081788301467896, + "learning_rate": 7.359778684247831e-07, + "loss": 0.5027, + "step": 20089 + }, + { + "epoch": 0.889498909710769, + "grad_norm": 1.2048953771591187, + "learning_rate": 7.353960620085554e-07, + "loss": 0.4102, + "step": 20090 + }, + { + "epoch": 0.8895431854155829, + "grad_norm": 1.4803251028060913, + "learning_rate": 7.348144768698485e-07, + "loss": 0.5061, + "step": 20091 + }, + { + "epoch": 0.8895874611203967, + "grad_norm": 1.4579445123672485, + "learning_rate": 7.342331130225533e-07, + "loss": 0.4816, + "step": 20092 + }, + { + "epoch": 0.8896317368252106, + "grad_norm": 1.4602413177490234, + "learning_rate": 7.336519704805567e-07, + "loss": 0.4511, + "step": 20093 + }, + { + "epoch": 0.8896760125300245, + "grad_norm": 1.682425856590271, + "learning_rate": 7.330710492577375e-07, + "loss": 0.4771, + "step": 20094 + }, + { + "epoch": 0.8897202882348383, + "grad_norm": 1.464491844177246, + "learning_rate": 7.324903493679703e-07, + "loss": 0.4967, + "step": 20095 + }, + { + "epoch": 0.8897645639396522, + "grad_norm": 1.3055113554000854, + "learning_rate": 7.31909870825126e-07, + "loss": 0.4624, + "step": 20096 + }, + { + "epoch": 0.8898088396444661, + "grad_norm": 1.4254366159439087, + "learning_rate": 7.313296136430681e-07, + "loss": 0.4714, + "step": 20097 + }, + { + "epoch": 0.88985311534928, + "grad_norm": 1.3458913564682007, + "learning_rate": 7.307495778356555e-07, + "loss": 0.4154, + "step": 20098 + }, + { + "epoch": 0.8898973910540938, + "grad_norm": 1.4678207635879517, + "learning_rate": 7.301697634167427e-07, + "loss": 0.468, + "step": 20099 + }, + { + "epoch": 0.8899416667589077, + "grad_norm": 1.4716310501098633, + "learning_rate": 7.295901704001763e-07, + "loss": 0.4859, + "step": 20100 + }, + { + "epoch": 0.8899859424637216, + "grad_norm": 1.222380518913269, + "learning_rate": 7.290107987998019e-07, + "loss": 0.4009, + "step": 20101 + }, + { + "epoch": 0.8900302181685354, + "grad_norm": 1.6937137842178345, + "learning_rate": 7.28431648629453e-07, + "loss": 0.4899, + "step": 20102 + }, + { + "epoch": 0.8900744938733494, + "grad_norm": 1.4363303184509277, + "learning_rate": 7.278527199029661e-07, + "loss": 0.4683, + "step": 20103 + }, + { + "epoch": 0.8901187695781633, + "grad_norm": 1.622475504875183, + "learning_rate": 7.272740126341682e-07, + "loss": 0.4497, + "step": 20104 + }, + { + "epoch": 0.8901630452829771, + "grad_norm": 2.248281955718994, + "learning_rate": 7.266955268368792e-07, + "loss": 0.5139, + "step": 20105 + }, + { + "epoch": 0.890207320987791, + "grad_norm": 1.5209180116653442, + "learning_rate": 7.26117262524918e-07, + "loss": 0.4921, + "step": 20106 + }, + { + "epoch": 0.8902515966926049, + "grad_norm": 1.5914119482040405, + "learning_rate": 7.255392197120947e-07, + "loss": 0.4796, + "step": 20107 + }, + { + "epoch": 0.8902958723974187, + "grad_norm": 3.1302337646484375, + "learning_rate": 7.249613984122161e-07, + "loss": 0.544, + "step": 20108 + }, + { + "epoch": 0.8903401481022326, + "grad_norm": 1.4131423234939575, + "learning_rate": 7.243837986390822e-07, + "loss": 0.4653, + "step": 20109 + }, + { + "epoch": 0.8903844238070465, + "grad_norm": 1.581864356994629, + "learning_rate": 7.238064204064887e-07, + "loss": 0.5185, + "step": 20110 + }, + { + "epoch": 0.8904286995118603, + "grad_norm": 1.6923274993896484, + "learning_rate": 7.232292637282256e-07, + "loss": 0.5003, + "step": 20111 + }, + { + "epoch": 0.8904729752166742, + "grad_norm": 1.3465335369110107, + "learning_rate": 7.226523286180776e-07, + "loss": 0.4119, + "step": 20112 + }, + { + "epoch": 0.8905172509214881, + "grad_norm": 1.3312791585922241, + "learning_rate": 7.220756150898256e-07, + "loss": 0.4506, + "step": 20113 + }, + { + "epoch": 0.890561526626302, + "grad_norm": 1.237545132637024, + "learning_rate": 7.214991231572433e-07, + "loss": 0.4506, + "step": 20114 + }, + { + "epoch": 0.8906058023311159, + "grad_norm": 1.5895549058914185, + "learning_rate": 7.209228528340972e-07, + "loss": 0.4481, + "step": 20115 + }, + { + "epoch": 0.8906500780359298, + "grad_norm": 1.6139204502105713, + "learning_rate": 7.203468041341565e-07, + "loss": 0.4791, + "step": 20116 + }, + { + "epoch": 0.8906943537407436, + "grad_norm": 1.5784310102462769, + "learning_rate": 7.197709770711758e-07, + "loss": 0.4853, + "step": 20117 + }, + { + "epoch": 0.8907386294455575, + "grad_norm": 1.480405569076538, + "learning_rate": 7.191953716589095e-07, + "loss": 0.5385, + "step": 20118 + }, + { + "epoch": 0.8907829051503714, + "grad_norm": 1.231256127357483, + "learning_rate": 7.186199879111044e-07, + "loss": 0.4109, + "step": 20119 + }, + { + "epoch": 0.8908271808551852, + "grad_norm": 1.4832189083099365, + "learning_rate": 7.180448258415041e-07, + "loss": 0.5005, + "step": 20120 + }, + { + "epoch": 0.8908714565599991, + "grad_norm": 1.5487158298492432, + "learning_rate": 7.174698854638451e-07, + "loss": 0.5085, + "step": 20121 + }, + { + "epoch": 0.890915732264813, + "grad_norm": 1.8301188945770264, + "learning_rate": 7.168951667918589e-07, + "loss": 0.537, + "step": 20122 + }, + { + "epoch": 0.8909600079696268, + "grad_norm": 1.4378465414047241, + "learning_rate": 7.163206698392744e-07, + "loss": 0.4837, + "step": 20123 + }, + { + "epoch": 0.8910042836744407, + "grad_norm": 1.363204836845398, + "learning_rate": 7.157463946198117e-07, + "loss": 0.4616, + "step": 20124 + }, + { + "epoch": 0.8910485593792546, + "grad_norm": 1.4255839586257935, + "learning_rate": 7.151723411471845e-07, + "loss": 0.5128, + "step": 20125 + }, + { + "epoch": 0.8910928350840684, + "grad_norm": 1.4824095964431763, + "learning_rate": 7.145985094351093e-07, + "loss": 0.4988, + "step": 20126 + }, + { + "epoch": 0.8911371107888824, + "grad_norm": 2.042353630065918, + "learning_rate": 7.140248994972876e-07, + "loss": 0.5455, + "step": 20127 + }, + { + "epoch": 0.8911813864936963, + "grad_norm": 1.8332570791244507, + "learning_rate": 7.134515113474183e-07, + "loss": 0.5967, + "step": 20128 + }, + { + "epoch": 0.8912256621985101, + "grad_norm": 1.4285221099853516, + "learning_rate": 7.128783449991994e-07, + "loss": 0.4912, + "step": 20129 + }, + { + "epoch": 0.891269937903324, + "grad_norm": 1.5531766414642334, + "learning_rate": 7.1230540046632e-07, + "loss": 0.4939, + "step": 20130 + }, + { + "epoch": 0.8913142136081379, + "grad_norm": 1.518768310546875, + "learning_rate": 7.117326777624633e-07, + "loss": 0.493, + "step": 20131 + }, + { + "epoch": 0.8913584893129517, + "grad_norm": 1.4402894973754883, + "learning_rate": 7.111601769013077e-07, + "loss": 0.4778, + "step": 20132 + }, + { + "epoch": 0.8914027650177656, + "grad_norm": 1.6835917234420776, + "learning_rate": 7.105878978965286e-07, + "loss": 0.5357, + "step": 20133 + }, + { + "epoch": 0.8914470407225795, + "grad_norm": 1.4898759126663208, + "learning_rate": 7.100158407617941e-07, + "loss": 0.4754, + "step": 20134 + }, + { + "epoch": 0.8914913164273934, + "grad_norm": 1.3452500104904175, + "learning_rate": 7.094440055107676e-07, + "loss": 0.4333, + "step": 20135 + }, + { + "epoch": 0.8915355921322072, + "grad_norm": 1.6769139766693115, + "learning_rate": 7.088723921571061e-07, + "loss": 0.5089, + "step": 20136 + }, + { + "epoch": 0.8915798678370211, + "grad_norm": 1.5319744348526, + "learning_rate": 7.083010007144631e-07, + "loss": 0.51, + "step": 20137 + }, + { + "epoch": 0.8916241435418351, + "grad_norm": 1.5669094324111938, + "learning_rate": 7.077298311964831e-07, + "loss": 0.49, + "step": 20138 + }, + { + "epoch": 0.8916684192466489, + "grad_norm": 1.830469012260437, + "learning_rate": 7.071588836168109e-07, + "loss": 0.5362, + "step": 20139 + }, + { + "epoch": 0.8917126949514628, + "grad_norm": 1.6105763912200928, + "learning_rate": 7.065881579890832e-07, + "loss": 0.4662, + "step": 20140 + }, + { + "epoch": 0.8917569706562767, + "grad_norm": 1.7099741697311401, + "learning_rate": 7.060176543269303e-07, + "loss": 0.4732, + "step": 20141 + }, + { + "epoch": 0.8918012463610905, + "grad_norm": 2.0081000328063965, + "learning_rate": 7.054473726439759e-07, + "loss": 0.5468, + "step": 20142 + }, + { + "epoch": 0.8918455220659044, + "grad_norm": 1.6525216102600098, + "learning_rate": 7.048773129538456e-07, + "loss": 0.4902, + "step": 20143 + }, + { + "epoch": 0.8918897977707183, + "grad_norm": 1.444067120552063, + "learning_rate": 7.043074752701517e-07, + "loss": 0.5036, + "step": 20144 + }, + { + "epoch": 0.8919340734755321, + "grad_norm": 1.6029127836227417, + "learning_rate": 7.037378596065047e-07, + "loss": 0.516, + "step": 20145 + }, + { + "epoch": 0.891978349180346, + "grad_norm": 1.7967230081558228, + "learning_rate": 7.031684659765104e-07, + "loss": 0.5237, + "step": 20146 + }, + { + "epoch": 0.8920226248851599, + "grad_norm": 2.0715811252593994, + "learning_rate": 7.025992943937677e-07, + "loss": 0.5449, + "step": 20147 + }, + { + "epoch": 0.8920669005899737, + "grad_norm": 1.3940284252166748, + "learning_rate": 7.020303448718679e-07, + "loss": 0.4406, + "step": 20148 + }, + { + "epoch": 0.8921111762947876, + "grad_norm": 1.3226370811462402, + "learning_rate": 7.014616174244049e-07, + "loss": 0.4868, + "step": 20149 + }, + { + "epoch": 0.8921554519996016, + "grad_norm": 1.8971039056777954, + "learning_rate": 7.008931120649598e-07, + "loss": 0.4978, + "step": 20150 + }, + { + "epoch": 0.8921997277044154, + "grad_norm": 1.8546411991119385, + "learning_rate": 7.003248288071118e-07, + "loss": 0.4666, + "step": 20151 + }, + { + "epoch": 0.8922440034092293, + "grad_norm": 1.5790200233459473, + "learning_rate": 6.997567676644302e-07, + "loss": 0.4739, + "step": 20152 + }, + { + "epoch": 0.8922882791140432, + "grad_norm": 1.5997310876846313, + "learning_rate": 6.991889286504882e-07, + "loss": 0.4806, + "step": 20153 + }, + { + "epoch": 0.892332554818857, + "grad_norm": 1.7651708126068115, + "learning_rate": 6.986213117788454e-07, + "loss": 0.5174, + "step": 20154 + }, + { + "epoch": 0.8923768305236709, + "grad_norm": 1.5325160026550293, + "learning_rate": 6.980539170630595e-07, + "loss": 0.4935, + "step": 20155 + }, + { + "epoch": 0.8924211062284848, + "grad_norm": 2.192331075668335, + "learning_rate": 6.974867445166822e-07, + "loss": 0.5839, + "step": 20156 + }, + { + "epoch": 0.8924653819332986, + "grad_norm": 1.322826862335205, + "learning_rate": 6.96919794153259e-07, + "loss": 0.4195, + "step": 20157 + }, + { + "epoch": 0.8925096576381125, + "grad_norm": 1.5761269330978394, + "learning_rate": 6.963530659863316e-07, + "loss": 0.5181, + "step": 20158 + }, + { + "epoch": 0.8925539333429264, + "grad_norm": 1.75229012966156, + "learning_rate": 6.957865600294344e-07, + "loss": 0.5196, + "step": 20159 + }, + { + "epoch": 0.8925982090477402, + "grad_norm": 1.3582477569580078, + "learning_rate": 6.952202762961003e-07, + "loss": 0.4437, + "step": 20160 + }, + { + "epoch": 0.8926424847525541, + "grad_norm": 1.5738402605056763, + "learning_rate": 6.946542147998547e-07, + "loss": 0.4755, + "step": 20161 + }, + { + "epoch": 0.8926867604573681, + "grad_norm": 1.5004737377166748, + "learning_rate": 6.94088375554215e-07, + "loss": 0.5035, + "step": 20162 + }, + { + "epoch": 0.8927310361621819, + "grad_norm": 1.8463292121887207, + "learning_rate": 6.935227585726978e-07, + "loss": 0.541, + "step": 20163 + }, + { + "epoch": 0.8927753118669958, + "grad_norm": 1.2789978981018066, + "learning_rate": 6.929573638688125e-07, + "loss": 0.4258, + "step": 20164 + }, + { + "epoch": 0.8928195875718097, + "grad_norm": 1.5518038272857666, + "learning_rate": 6.923921914560627e-07, + "loss": 0.4749, + "step": 20165 + }, + { + "epoch": 0.8928638632766235, + "grad_norm": 1.4556701183319092, + "learning_rate": 6.918272413479465e-07, + "loss": 0.5135, + "step": 20166 + }, + { + "epoch": 0.8929081389814374, + "grad_norm": 1.5917996168136597, + "learning_rate": 6.912625135579587e-07, + "loss": 0.4678, + "step": 20167 + }, + { + "epoch": 0.8929524146862513, + "grad_norm": 1.4180257320404053, + "learning_rate": 6.906980080995851e-07, + "loss": 0.455, + "step": 20168 + }, + { + "epoch": 0.8929966903910652, + "grad_norm": 1.298398494720459, + "learning_rate": 6.901337249863083e-07, + "loss": 0.3936, + "step": 20169 + }, + { + "epoch": 0.893040966095879, + "grad_norm": 1.6825016736984253, + "learning_rate": 6.895696642316096e-07, + "loss": 0.5124, + "step": 20170 + }, + { + "epoch": 0.8930852418006929, + "grad_norm": 1.3951318264007568, + "learning_rate": 6.890058258489573e-07, + "loss": 0.4551, + "step": 20171 + }, + { + "epoch": 0.8931295175055068, + "grad_norm": 1.9508248567581177, + "learning_rate": 6.884422098518184e-07, + "loss": 0.5382, + "step": 20172 + }, + { + "epoch": 0.8931737932103206, + "grad_norm": 1.6077665090560913, + "learning_rate": 6.878788162536576e-07, + "loss": 0.4488, + "step": 20173 + }, + { + "epoch": 0.8932180689151346, + "grad_norm": 1.7817026376724243, + "learning_rate": 6.873156450679308e-07, + "loss": 0.4913, + "step": 20174 + }, + { + "epoch": 0.8932623446199485, + "grad_norm": 1.4805891513824463, + "learning_rate": 6.867526963080828e-07, + "loss": 0.5269, + "step": 20175 + }, + { + "epoch": 0.8933066203247623, + "grad_norm": 2.1054248809814453, + "learning_rate": 6.861899699875662e-07, + "loss": 0.4772, + "step": 20176 + }, + { + "epoch": 0.8933508960295762, + "grad_norm": 1.7756601572036743, + "learning_rate": 6.856274661198182e-07, + "loss": 0.5078, + "step": 20177 + }, + { + "epoch": 0.8933951717343901, + "grad_norm": 1.4302254915237427, + "learning_rate": 6.850651847182743e-07, + "loss": 0.4922, + "step": 20178 + }, + { + "epoch": 0.8934394474392039, + "grad_norm": 1.5402942895889282, + "learning_rate": 6.845031257963619e-07, + "loss": 0.4929, + "step": 20179 + }, + { + "epoch": 0.8934837231440178, + "grad_norm": 1.819961428642273, + "learning_rate": 6.83941289367509e-07, + "loss": 0.4706, + "step": 20180 + }, + { + "epoch": 0.8935279988488317, + "grad_norm": 1.2368230819702148, + "learning_rate": 6.833796754451339e-07, + "loss": 0.433, + "step": 20181 + }, + { + "epoch": 0.8935722745536455, + "grad_norm": 1.3145145177841187, + "learning_rate": 6.82818284042649e-07, + "loss": 0.4441, + "step": 20182 + }, + { + "epoch": 0.8936165502584594, + "grad_norm": 1.4057461023330688, + "learning_rate": 6.822571151734625e-07, + "loss": 0.4828, + "step": 20183 + }, + { + "epoch": 0.8936608259632733, + "grad_norm": 1.617734432220459, + "learning_rate": 6.816961688509783e-07, + "loss": 0.5413, + "step": 20184 + }, + { + "epoch": 0.8937051016680871, + "grad_norm": 1.5084868669509888, + "learning_rate": 6.811354450885921e-07, + "loss": 0.4541, + "step": 20185 + }, + { + "epoch": 0.893749377372901, + "grad_norm": 1.5634996891021729, + "learning_rate": 6.805749438997e-07, + "loss": 0.4593, + "step": 20186 + }, + { + "epoch": 0.893793653077715, + "grad_norm": 1.3944272994995117, + "learning_rate": 6.800146652976869e-07, + "loss": 0.451, + "step": 20187 + }, + { + "epoch": 0.8938379287825288, + "grad_norm": 1.8074909448623657, + "learning_rate": 6.794546092959353e-07, + "loss": 0.4648, + "step": 20188 + }, + { + "epoch": 0.8938822044873427, + "grad_norm": 1.669088363647461, + "learning_rate": 6.788947759078191e-07, + "loss": 0.4338, + "step": 20189 + }, + { + "epoch": 0.8939264801921566, + "grad_norm": 1.622686743736267, + "learning_rate": 6.783351651467141e-07, + "loss": 0.4727, + "step": 20190 + }, + { + "epoch": 0.8939707558969704, + "grad_norm": 1.4736970663070679, + "learning_rate": 6.77775777025983e-07, + "loss": 0.4499, + "step": 20191 + }, + { + "epoch": 0.8940150316017843, + "grad_norm": 1.5497609376907349, + "learning_rate": 6.772166115589873e-07, + "loss": 0.4944, + "step": 20192 + }, + { + "epoch": 0.8940593073065982, + "grad_norm": 1.6296665668487549, + "learning_rate": 6.76657668759082e-07, + "loss": 0.4536, + "step": 20193 + }, + { + "epoch": 0.894103583011412, + "grad_norm": 1.4977316856384277, + "learning_rate": 6.760989486396174e-07, + "loss": 0.4678, + "step": 20194 + }, + { + "epoch": 0.8941478587162259, + "grad_norm": 1.522420883178711, + "learning_rate": 6.755404512139352e-07, + "loss": 0.461, + "step": 20195 + }, + { + "epoch": 0.8941921344210398, + "grad_norm": 1.6204420328140259, + "learning_rate": 6.74982176495379e-07, + "loss": 0.4734, + "step": 20196 + }, + { + "epoch": 0.8942364101258536, + "grad_norm": 1.4978920221328735, + "learning_rate": 6.744241244972804e-07, + "loss": 0.4929, + "step": 20197 + }, + { + "epoch": 0.8942806858306676, + "grad_norm": 1.4419697523117065, + "learning_rate": 6.738662952329688e-07, + "loss": 0.4649, + "step": 20198 + }, + { + "epoch": 0.8943249615354815, + "grad_norm": 1.4094336032867432, + "learning_rate": 6.733086887157659e-07, + "loss": 0.4563, + "step": 20199 + }, + { + "epoch": 0.8943692372402953, + "grad_norm": 1.3736729621887207, + "learning_rate": 6.727513049589918e-07, + "loss": 0.4517, + "step": 20200 + }, + { + "epoch": 0.8944135129451092, + "grad_norm": 1.8237073421478271, + "learning_rate": 6.721941439759583e-07, + "loss": 0.528, + "step": 20201 + }, + { + "epoch": 0.8944577886499231, + "grad_norm": 1.5046579837799072, + "learning_rate": 6.716372057799725e-07, + "loss": 0.472, + "step": 20202 + }, + { + "epoch": 0.8945020643547369, + "grad_norm": 1.4303878545761108, + "learning_rate": 6.71080490384337e-07, + "loss": 0.5385, + "step": 20203 + }, + { + "epoch": 0.8945463400595508, + "grad_norm": 1.4562755823135376, + "learning_rate": 6.705239978023481e-07, + "loss": 0.4849, + "step": 20204 + }, + { + "epoch": 0.8945906157643647, + "grad_norm": 1.4093825817108154, + "learning_rate": 6.699677280472961e-07, + "loss": 0.4861, + "step": 20205 + }, + { + "epoch": 0.8946348914691786, + "grad_norm": 1.533425211906433, + "learning_rate": 6.694116811324691e-07, + "loss": 0.4453, + "step": 20206 + }, + { + "epoch": 0.8946791671739924, + "grad_norm": 1.9375218152999878, + "learning_rate": 6.688558570711468e-07, + "loss": 0.5239, + "step": 20207 + }, + { + "epoch": 0.8947234428788063, + "grad_norm": 1.6222267150878906, + "learning_rate": 6.68300255876605e-07, + "loss": 0.4989, + "step": 20208 + }, + { + "epoch": 0.8947677185836203, + "grad_norm": 1.5546773672103882, + "learning_rate": 6.677448775621121e-07, + "loss": 0.488, + "step": 20209 + }, + { + "epoch": 0.894811994288434, + "grad_norm": 1.6090389490127563, + "learning_rate": 6.671897221409351e-07, + "loss": 0.5012, + "step": 20210 + }, + { + "epoch": 0.894856269993248, + "grad_norm": 1.6376475095748901, + "learning_rate": 6.666347896263326e-07, + "loss": 0.4714, + "step": 20211 + }, + { + "epoch": 0.8949005456980619, + "grad_norm": 1.2620328664779663, + "learning_rate": 6.660800800315592e-07, + "loss": 0.4842, + "step": 20212 + }, + { + "epoch": 0.8949448214028757, + "grad_norm": 1.708295464515686, + "learning_rate": 6.655255933698623e-07, + "loss": 0.4898, + "step": 20213 + }, + { + "epoch": 0.8949890971076896, + "grad_norm": 1.6544967889785767, + "learning_rate": 6.649713296544869e-07, + "loss": 0.5464, + "step": 20214 + }, + { + "epoch": 0.8950333728125035, + "grad_norm": 1.60865318775177, + "learning_rate": 6.644172888986677e-07, + "loss": 0.4151, + "step": 20215 + }, + { + "epoch": 0.8950776485173173, + "grad_norm": 1.4951903820037842, + "learning_rate": 6.638634711156434e-07, + "loss": 0.4214, + "step": 20216 + }, + { + "epoch": 0.8951219242221312, + "grad_norm": 1.858280897140503, + "learning_rate": 6.633098763186374e-07, + "loss": 0.5602, + "step": 20217 + }, + { + "epoch": 0.8951661999269451, + "grad_norm": 1.5058680772781372, + "learning_rate": 6.627565045208729e-07, + "loss": 0.4809, + "step": 20218 + }, + { + "epoch": 0.8952104756317589, + "grad_norm": 2.057344436645508, + "learning_rate": 6.622033557355656e-07, + "loss": 0.5416, + "step": 20219 + }, + { + "epoch": 0.8952547513365728, + "grad_norm": 1.968243956565857, + "learning_rate": 6.616504299759285e-07, + "loss": 0.6326, + "step": 20220 + }, + { + "epoch": 0.8952990270413868, + "grad_norm": 1.7904987335205078, + "learning_rate": 6.6109772725517e-07, + "loss": 0.5303, + "step": 20221 + }, + { + "epoch": 0.8953433027462006, + "grad_norm": 1.6118741035461426, + "learning_rate": 6.605452475864849e-07, + "loss": 0.4689, + "step": 20222 + }, + { + "epoch": 0.8953875784510145, + "grad_norm": 1.6422947645187378, + "learning_rate": 6.599929909830727e-07, + "loss": 0.5025, + "step": 20223 + }, + { + "epoch": 0.8954318541558284, + "grad_norm": 1.7041183710098267, + "learning_rate": 6.59440957458124e-07, + "loss": 0.5135, + "step": 20224 + }, + { + "epoch": 0.8954761298606422, + "grad_norm": 1.402311086654663, + "learning_rate": 6.588891470248215e-07, + "loss": 0.4992, + "step": 20225 + }, + { + "epoch": 0.8955204055654561, + "grad_norm": 1.3724110126495361, + "learning_rate": 6.58337559696346e-07, + "loss": 0.4135, + "step": 20226 + }, + { + "epoch": 0.89556468127027, + "grad_norm": 1.770530104637146, + "learning_rate": 6.577861954858733e-07, + "loss": 0.5372, + "step": 20227 + }, + { + "epoch": 0.8956089569750838, + "grad_norm": 1.446120262145996, + "learning_rate": 6.572350544065686e-07, + "loss": 0.4436, + "step": 20228 + }, + { + "epoch": 0.8956532326798977, + "grad_norm": 1.6739435195922852, + "learning_rate": 6.566841364715992e-07, + "loss": 0.5641, + "step": 20229 + }, + { + "epoch": 0.8956975083847116, + "grad_norm": 1.5925874710083008, + "learning_rate": 6.561334416941212e-07, + "loss": 0.5171, + "step": 20230 + }, + { + "epoch": 0.8957417840895254, + "grad_norm": 1.4944350719451904, + "learning_rate": 6.555829700872874e-07, + "loss": 0.4455, + "step": 20231 + }, + { + "epoch": 0.8957860597943393, + "grad_norm": 1.2414302825927734, + "learning_rate": 6.550327216642449e-07, + "loss": 0.4112, + "step": 20232 + }, + { + "epoch": 0.8958303354991533, + "grad_norm": 1.7840415239334106, + "learning_rate": 6.544826964381379e-07, + "loss": 0.4814, + "step": 20233 + }, + { + "epoch": 0.895874611203967, + "grad_norm": 1.2847243547439575, + "learning_rate": 6.539328944221023e-07, + "loss": 0.4203, + "step": 20234 + }, + { + "epoch": 0.895918886908781, + "grad_norm": 1.82847261428833, + "learning_rate": 6.53383315629268e-07, + "loss": 0.48, + "step": 20235 + }, + { + "epoch": 0.8959631626135949, + "grad_norm": 1.643373966217041, + "learning_rate": 6.528339600727651e-07, + "loss": 0.502, + "step": 20236 + }, + { + "epoch": 0.8960074383184087, + "grad_norm": 1.5954948663711548, + "learning_rate": 6.522848277657124e-07, + "loss": 0.4633, + "step": 20237 + }, + { + "epoch": 0.8960517140232226, + "grad_norm": 1.562972068786621, + "learning_rate": 6.517359187212246e-07, + "loss": 0.462, + "step": 20238 + }, + { + "epoch": 0.8960959897280365, + "grad_norm": 1.422467827796936, + "learning_rate": 6.511872329524139e-07, + "loss": 0.5189, + "step": 20239 + }, + { + "epoch": 0.8961402654328504, + "grad_norm": 1.5458101034164429, + "learning_rate": 6.506387704723826e-07, + "loss": 0.4972, + "step": 20240 + }, + { + "epoch": 0.8961845411376642, + "grad_norm": 1.7299937009811401, + "learning_rate": 6.50090531294233e-07, + "loss": 0.5019, + "step": 20241 + }, + { + "epoch": 0.8962288168424781, + "grad_norm": 1.5495778322219849, + "learning_rate": 6.495425154310553e-07, + "loss": 0.5269, + "step": 20242 + }, + { + "epoch": 0.896273092547292, + "grad_norm": 1.5746954679489136, + "learning_rate": 6.489947228959437e-07, + "loss": 0.5202, + "step": 20243 + }, + { + "epoch": 0.8963173682521058, + "grad_norm": 1.4059383869171143, + "learning_rate": 6.484471537019787e-07, + "loss": 0.4552, + "step": 20244 + }, + { + "epoch": 0.8963616439569198, + "grad_norm": 1.4554966688156128, + "learning_rate": 6.478998078622378e-07, + "loss": 0.4445, + "step": 20245 + }, + { + "epoch": 0.8964059196617337, + "grad_norm": 1.754626750946045, + "learning_rate": 6.473526853897971e-07, + "loss": 0.4645, + "step": 20246 + }, + { + "epoch": 0.8964501953665475, + "grad_norm": 1.5617713928222656, + "learning_rate": 6.468057862977217e-07, + "loss": 0.4795, + "step": 20247 + }, + { + "epoch": 0.8964944710713614, + "grad_norm": 1.59425950050354, + "learning_rate": 6.462591105990745e-07, + "loss": 0.5567, + "step": 20248 + }, + { + "epoch": 0.8965387467761753, + "grad_norm": 1.6899899244308472, + "learning_rate": 6.457126583069128e-07, + "loss": 0.4946, + "step": 20249 + }, + { + "epoch": 0.8965830224809891, + "grad_norm": 1.4591255187988281, + "learning_rate": 6.451664294342874e-07, + "loss": 0.4375, + "step": 20250 + }, + { + "epoch": 0.896627298185803, + "grad_norm": 1.2695776224136353, + "learning_rate": 6.446204239942455e-07, + "loss": 0.4594, + "step": 20251 + }, + { + "epoch": 0.8966715738906169, + "grad_norm": 1.4469172954559326, + "learning_rate": 6.440746419998256e-07, + "loss": 0.5094, + "step": 20252 + }, + { + "epoch": 0.8967158495954307, + "grad_norm": 1.5021450519561768, + "learning_rate": 6.435290834640662e-07, + "loss": 0.4705, + "step": 20253 + }, + { + "epoch": 0.8967601253002446, + "grad_norm": 1.2914154529571533, + "learning_rate": 6.429837483999957e-07, + "loss": 0.4284, + "step": 20254 + }, + { + "epoch": 0.8968044010050585, + "grad_norm": 1.880346417427063, + "learning_rate": 6.424386368206392e-07, + "loss": 0.5143, + "step": 20255 + }, + { + "epoch": 0.8968486767098723, + "grad_norm": 1.7003897428512573, + "learning_rate": 6.418937487390175e-07, + "loss": 0.5582, + "step": 20256 + }, + { + "epoch": 0.8968929524146863, + "grad_norm": 1.6134377717971802, + "learning_rate": 6.413490841681435e-07, + "loss": 0.5191, + "step": 20257 + }, + { + "epoch": 0.8969372281195002, + "grad_norm": 1.657685399055481, + "learning_rate": 6.40804643121028e-07, + "loss": 0.5033, + "step": 20258 + }, + { + "epoch": 0.896981503824314, + "grad_norm": 1.2335319519042969, + "learning_rate": 6.402604256106715e-07, + "loss": 0.4125, + "step": 20259 + }, + { + "epoch": 0.8970257795291279, + "grad_norm": 1.2941683530807495, + "learning_rate": 6.397164316500748e-07, + "loss": 0.4807, + "step": 20260 + }, + { + "epoch": 0.8970700552339418, + "grad_norm": 1.513654112815857, + "learning_rate": 6.391726612522298e-07, + "loss": 0.4535, + "step": 20261 + }, + { + "epoch": 0.8971143309387556, + "grad_norm": 1.328924536705017, + "learning_rate": 6.386291144301216e-07, + "loss": 0.4518, + "step": 20262 + }, + { + "epoch": 0.8971586066435695, + "grad_norm": 1.9043455123901367, + "learning_rate": 6.380857911967364e-07, + "loss": 0.5413, + "step": 20263 + }, + { + "epoch": 0.8972028823483834, + "grad_norm": 1.7885305881500244, + "learning_rate": 6.375426915650495e-07, + "loss": 0.4706, + "step": 20264 + }, + { + "epoch": 0.8972471580531972, + "grad_norm": 1.4345576763153076, + "learning_rate": 6.369998155480306e-07, + "loss": 0.441, + "step": 20265 + }, + { + "epoch": 0.8972914337580111, + "grad_norm": 1.4745570421218872, + "learning_rate": 6.364571631586492e-07, + "loss": 0.4521, + "step": 20266 + }, + { + "epoch": 0.897335709462825, + "grad_norm": 1.6651678085327148, + "learning_rate": 6.359147344098637e-07, + "loss": 0.4691, + "step": 20267 + }, + { + "epoch": 0.8973799851676388, + "grad_norm": 1.5564359426498413, + "learning_rate": 6.353725293146329e-07, + "loss": 0.4762, + "step": 20268 + }, + { + "epoch": 0.8974242608724528, + "grad_norm": 1.61492919921875, + "learning_rate": 6.348305478859007e-07, + "loss": 0.5041, + "step": 20269 + }, + { + "epoch": 0.8974685365772667, + "grad_norm": 1.586425542831421, + "learning_rate": 6.342887901366168e-07, + "loss": 0.5274, + "step": 20270 + }, + { + "epoch": 0.8975128122820805, + "grad_norm": 1.5658403635025024, + "learning_rate": 6.337472560797197e-07, + "loss": 0.4842, + "step": 20271 + }, + { + "epoch": 0.8975570879868944, + "grad_norm": 2.413935899734497, + "learning_rate": 6.332059457281414e-07, + "loss": 0.4914, + "step": 20272 + }, + { + "epoch": 0.8976013636917083, + "grad_norm": 1.8387606143951416, + "learning_rate": 6.326648590948137e-07, + "loss": 0.4873, + "step": 20273 + }, + { + "epoch": 0.8976456393965221, + "grad_norm": 1.5056657791137695, + "learning_rate": 6.321239961926585e-07, + "loss": 0.508, + "step": 20274 + }, + { + "epoch": 0.897689915101336, + "grad_norm": 1.4092330932617188, + "learning_rate": 6.315833570345942e-07, + "loss": 0.4554, + "step": 20275 + }, + { + "epoch": 0.8977341908061499, + "grad_norm": 1.6845219135284424, + "learning_rate": 6.310429416335317e-07, + "loss": 0.5271, + "step": 20276 + }, + { + "epoch": 0.8977784665109638, + "grad_norm": 1.2533090114593506, + "learning_rate": 6.305027500023841e-07, + "loss": 0.4532, + "step": 20277 + }, + { + "epoch": 0.8978227422157776, + "grad_norm": 1.8342291116714478, + "learning_rate": 6.299627821540466e-07, + "loss": 0.5473, + "step": 20278 + }, + { + "epoch": 0.8978670179205915, + "grad_norm": 1.3591599464416504, + "learning_rate": 6.294230381014177e-07, + "loss": 0.413, + "step": 20279 + }, + { + "epoch": 0.8979112936254054, + "grad_norm": 1.616018533706665, + "learning_rate": 6.288835178573916e-07, + "loss": 0.4826, + "step": 20280 + }, + { + "epoch": 0.8979555693302192, + "grad_norm": 1.8151819705963135, + "learning_rate": 6.283442214348523e-07, + "loss": 0.5569, + "step": 20281 + }, + { + "epoch": 0.8979998450350332, + "grad_norm": 1.30524480342865, + "learning_rate": 6.278051488466786e-07, + "loss": 0.4006, + "step": 20282 + }, + { + "epoch": 0.8980441207398471, + "grad_norm": 1.6501123905181885, + "learning_rate": 6.2726630010575e-07, + "loss": 0.5081, + "step": 20283 + }, + { + "epoch": 0.8980883964446609, + "grad_norm": 1.4069393873214722, + "learning_rate": 6.26727675224933e-07, + "loss": 0.4997, + "step": 20284 + }, + { + "epoch": 0.8981326721494748, + "grad_norm": 2.0326712131500244, + "learning_rate": 6.26189274217095e-07, + "loss": 0.5094, + "step": 20285 + }, + { + "epoch": 0.8981769478542887, + "grad_norm": 1.4624305963516235, + "learning_rate": 6.256510970950924e-07, + "loss": 0.4291, + "step": 20286 + }, + { + "epoch": 0.8982212235591025, + "grad_norm": 2.0174381732940674, + "learning_rate": 6.251131438717816e-07, + "loss": 0.6303, + "step": 20287 + }, + { + "epoch": 0.8982654992639164, + "grad_norm": 1.4537032842636108, + "learning_rate": 6.245754145600091e-07, + "loss": 0.4697, + "step": 20288 + }, + { + "epoch": 0.8983097749687303, + "grad_norm": 1.3956648111343384, + "learning_rate": 6.240379091726179e-07, + "loss": 0.4769, + "step": 20289 + }, + { + "epoch": 0.8983540506735441, + "grad_norm": 1.5208501815795898, + "learning_rate": 6.235006277224487e-07, + "loss": 0.4623, + "step": 20290 + }, + { + "epoch": 0.898398326378358, + "grad_norm": 1.5918117761611938, + "learning_rate": 6.229635702223325e-07, + "loss": 0.4775, + "step": 20291 + }, + { + "epoch": 0.898442602083172, + "grad_norm": 1.4948383569717407, + "learning_rate": 6.224267366850945e-07, + "loss": 0.491, + "step": 20292 + }, + { + "epoch": 0.8984868777879857, + "grad_norm": 1.3193498849868774, + "learning_rate": 6.2189012712356e-07, + "loss": 0.4644, + "step": 20293 + }, + { + "epoch": 0.8985311534927997, + "grad_norm": 1.4523886442184448, + "learning_rate": 6.213537415505444e-07, + "loss": 0.4896, + "step": 20294 + }, + { + "epoch": 0.8985754291976136, + "grad_norm": 1.37825345993042, + "learning_rate": 6.208175799788574e-07, + "loss": 0.3966, + "step": 20295 + }, + { + "epoch": 0.8986197049024274, + "grad_norm": 1.2761143445968628, + "learning_rate": 6.202816424213065e-07, + "loss": 0.4611, + "step": 20296 + }, + { + "epoch": 0.8986639806072413, + "grad_norm": 1.4117330312728882, + "learning_rate": 6.197459288906904e-07, + "loss": 0.4829, + "step": 20297 + }, + { + "epoch": 0.8987082563120552, + "grad_norm": 1.5568181276321411, + "learning_rate": 6.192104393998066e-07, + "loss": 0.4925, + "step": 20298 + }, + { + "epoch": 0.898752532016869, + "grad_norm": 1.4665292501449585, + "learning_rate": 6.186751739614405e-07, + "loss": 0.4882, + "step": 20299 + }, + { + "epoch": 0.8987968077216829, + "grad_norm": 1.600517988204956, + "learning_rate": 6.181401325883807e-07, + "loss": 0.5486, + "step": 20300 + }, + { + "epoch": 0.8988410834264968, + "grad_norm": 1.7418290376663208, + "learning_rate": 6.176053152934058e-07, + "loss": 0.5115, + "step": 20301 + }, + { + "epoch": 0.8988853591313106, + "grad_norm": 1.6503815650939941, + "learning_rate": 6.170707220892869e-07, + "loss": 0.4908, + "step": 20302 + }, + { + "epoch": 0.8989296348361245, + "grad_norm": 1.4820358753204346, + "learning_rate": 6.165363529887947e-07, + "loss": 0.5042, + "step": 20303 + }, + { + "epoch": 0.8989739105409384, + "grad_norm": 1.4002101421356201, + "learning_rate": 6.160022080046923e-07, + "loss": 0.4589, + "step": 20304 + }, + { + "epoch": 0.8990181862457522, + "grad_norm": 1.475651502609253, + "learning_rate": 6.154682871497352e-07, + "loss": 0.4472, + "step": 20305 + }, + { + "epoch": 0.8990624619505662, + "grad_norm": 1.3944779634475708, + "learning_rate": 6.149345904366777e-07, + "loss": 0.4567, + "step": 20306 + }, + { + "epoch": 0.8991067376553801, + "grad_norm": 1.2346304655075073, + "learning_rate": 6.14401117878266e-07, + "loss": 0.4327, + "step": 20307 + }, + { + "epoch": 0.8991510133601939, + "grad_norm": 1.6381393671035767, + "learning_rate": 6.138678694872413e-07, + "loss": 0.5183, + "step": 20308 + }, + { + "epoch": 0.8991952890650078, + "grad_norm": 1.544592022895813, + "learning_rate": 6.133348452763388e-07, + "loss": 0.4725, + "step": 20309 + }, + { + "epoch": 0.8992395647698217, + "grad_norm": 1.4749962091445923, + "learning_rate": 6.128020452582917e-07, + "loss": 0.4564, + "step": 20310 + }, + { + "epoch": 0.8992838404746356, + "grad_norm": 1.5478297472000122, + "learning_rate": 6.122694694458254e-07, + "loss": 0.4852, + "step": 20311 + }, + { + "epoch": 0.8993281161794494, + "grad_norm": 1.562552809715271, + "learning_rate": 6.117371178516562e-07, + "loss": 0.5407, + "step": 20312 + }, + { + "epoch": 0.8993723918842633, + "grad_norm": 1.4751733541488647, + "learning_rate": 6.112049904885042e-07, + "loss": 0.4876, + "step": 20313 + }, + { + "epoch": 0.8994166675890772, + "grad_norm": 1.7779513597488403, + "learning_rate": 6.106730873690758e-07, + "loss": 0.5077, + "step": 20314 + }, + { + "epoch": 0.899460943293891, + "grad_norm": 1.3917783498764038, + "learning_rate": 6.101414085060764e-07, + "loss": 0.4425, + "step": 20315 + }, + { + "epoch": 0.899505218998705, + "grad_norm": 1.4636114835739136, + "learning_rate": 6.096099539122036e-07, + "loss": 0.4833, + "step": 20316 + }, + { + "epoch": 0.8995494947035189, + "grad_norm": 1.5408843755722046, + "learning_rate": 6.090787236001505e-07, + "loss": 0.5137, + "step": 20317 + }, + { + "epoch": 0.8995937704083327, + "grad_norm": 1.8859081268310547, + "learning_rate": 6.08547717582606e-07, + "loss": 0.5409, + "step": 20318 + }, + { + "epoch": 0.8996380461131466, + "grad_norm": 1.5224138498306274, + "learning_rate": 6.08016935872251e-07, + "loss": 0.4676, + "step": 20319 + }, + { + "epoch": 0.8996823218179605, + "grad_norm": 1.6833045482635498, + "learning_rate": 6.074863784817664e-07, + "loss": 0.477, + "step": 20320 + }, + { + "epoch": 0.8997265975227743, + "grad_norm": 1.7267862558364868, + "learning_rate": 6.069560454238211e-07, + "loss": 0.4627, + "step": 20321 + }, + { + "epoch": 0.8997708732275882, + "grad_norm": 1.347078800201416, + "learning_rate": 6.064259367110814e-07, + "loss": 0.4768, + "step": 20322 + }, + { + "epoch": 0.8998151489324021, + "grad_norm": 1.6018904447555542, + "learning_rate": 6.058960523562108e-07, + "loss": 0.504, + "step": 20323 + }, + { + "epoch": 0.8998594246372159, + "grad_norm": 1.4430097341537476, + "learning_rate": 6.053663923718656e-07, + "loss": 0.4424, + "step": 20324 + }, + { + "epoch": 0.8999037003420298, + "grad_norm": 1.5462955236434937, + "learning_rate": 6.048369567706935e-07, + "loss": 0.4725, + "step": 20325 + }, + { + "epoch": 0.8999479760468437, + "grad_norm": 1.3639029264450073, + "learning_rate": 6.043077455653401e-07, + "loss": 0.4224, + "step": 20326 + }, + { + "epoch": 0.8999922517516575, + "grad_norm": 1.8161203861236572, + "learning_rate": 6.037787587684463e-07, + "loss": 0.516, + "step": 20327 + }, + { + "epoch": 0.9000365274564714, + "grad_norm": 1.3198367357254028, + "learning_rate": 6.032499963926463e-07, + "loss": 0.4162, + "step": 20328 + }, + { + "epoch": 0.9000808031612854, + "grad_norm": 1.6634893417358398, + "learning_rate": 6.02721458450568e-07, + "loss": 0.5024, + "step": 20329 + }, + { + "epoch": 0.9001250788660992, + "grad_norm": 1.4748166799545288, + "learning_rate": 6.021931449548379e-07, + "loss": 0.4846, + "step": 20330 + }, + { + "epoch": 0.9001693545709131, + "grad_norm": 1.3791948556900024, + "learning_rate": 6.016650559180715e-07, + "loss": 0.4849, + "step": 20331 + }, + { + "epoch": 0.900213630275727, + "grad_norm": 1.4357677698135376, + "learning_rate": 6.011371913528818e-07, + "loss": 0.4606, + "step": 20332 + }, + { + "epoch": 0.9002579059805408, + "grad_norm": 1.4685789346694946, + "learning_rate": 6.006095512718812e-07, + "loss": 0.4619, + "step": 20333 + }, + { + "epoch": 0.9003021816853547, + "grad_norm": 1.558856725692749, + "learning_rate": 6.000821356876652e-07, + "loss": 0.4031, + "step": 20334 + }, + { + "epoch": 0.9003464573901686, + "grad_norm": 1.5723930597305298, + "learning_rate": 5.995549446128346e-07, + "loss": 0.4563, + "step": 20335 + }, + { + "epoch": 0.9003907330949824, + "grad_norm": 1.6145603656768799, + "learning_rate": 5.990279780599784e-07, + "loss": 0.5292, + "step": 20336 + }, + { + "epoch": 0.9004350087997963, + "grad_norm": 1.5940968990325928, + "learning_rate": 5.985012360416864e-07, + "loss": 0.4971, + "step": 20337 + }, + { + "epoch": 0.9004792845046102, + "grad_norm": 2.115490674972534, + "learning_rate": 5.979747185705364e-07, + "loss": 0.5201, + "step": 20338 + }, + { + "epoch": 0.900523560209424, + "grad_norm": 1.552208662033081, + "learning_rate": 5.97448425659104e-07, + "loss": 0.4934, + "step": 20339 + }, + { + "epoch": 0.900567835914238, + "grad_norm": 1.417008876800537, + "learning_rate": 5.969223573199623e-07, + "loss": 0.4614, + "step": 20340 + }, + { + "epoch": 0.9006121116190519, + "grad_norm": 1.5208384990692139, + "learning_rate": 5.963965135656724e-07, + "loss": 0.4573, + "step": 20341 + }, + { + "epoch": 0.9006563873238657, + "grad_norm": 1.5665730237960815, + "learning_rate": 5.958708944087954e-07, + "loss": 0.4682, + "step": 20342 + }, + { + "epoch": 0.9007006630286796, + "grad_norm": 1.9321757555007935, + "learning_rate": 5.953454998618857e-07, + "loss": 0.4555, + "step": 20343 + }, + { + "epoch": 0.9007449387334935, + "grad_norm": 1.5098761320114136, + "learning_rate": 5.94820329937491e-07, + "loss": 0.4262, + "step": 20344 + }, + { + "epoch": 0.9007892144383074, + "grad_norm": 1.4855520725250244, + "learning_rate": 5.942953846481536e-07, + "loss": 0.5572, + "step": 20345 + }, + { + "epoch": 0.9008334901431212, + "grad_norm": 1.666117548942566, + "learning_rate": 5.937706640064123e-07, + "loss": 0.5052, + "step": 20346 + }, + { + "epoch": 0.9008777658479351, + "grad_norm": 1.6027822494506836, + "learning_rate": 5.932461680248014e-07, + "loss": 0.5199, + "step": 20347 + }, + { + "epoch": 0.900922041552749, + "grad_norm": 1.9196314811706543, + "learning_rate": 5.927218967158454e-07, + "loss": 0.5432, + "step": 20348 + }, + { + "epoch": 0.9009663172575628, + "grad_norm": 1.6787484884262085, + "learning_rate": 5.921978500920666e-07, + "loss": 0.5741, + "step": 20349 + }, + { + "epoch": 0.9010105929623767, + "grad_norm": 1.3879400491714478, + "learning_rate": 5.916740281659828e-07, + "loss": 0.4682, + "step": 20350 + }, + { + "epoch": 0.9010548686671906, + "grad_norm": 1.8320356607437134, + "learning_rate": 5.911504309501048e-07, + "loss": 0.5491, + "step": 20351 + }, + { + "epoch": 0.9010991443720044, + "grad_norm": 1.8963427543640137, + "learning_rate": 5.906270584569385e-07, + "loss": 0.5181, + "step": 20352 + }, + { + "epoch": 0.9011434200768184, + "grad_norm": 1.8696948289871216, + "learning_rate": 5.901039106989826e-07, + "loss": 0.4922, + "step": 20353 + }, + { + "epoch": 0.9011876957816323, + "grad_norm": 1.3926124572753906, + "learning_rate": 5.895809876887327e-07, + "loss": 0.464, + "step": 20354 + }, + { + "epoch": 0.9012319714864461, + "grad_norm": 1.565561056137085, + "learning_rate": 5.890582894386798e-07, + "loss": 0.4414, + "step": 20355 + }, + { + "epoch": 0.90127624719126, + "grad_norm": 1.2301987409591675, + "learning_rate": 5.885358159613041e-07, + "loss": 0.4245, + "step": 20356 + }, + { + "epoch": 0.9013205228960739, + "grad_norm": 1.4043530225753784, + "learning_rate": 5.880135672690901e-07, + "loss": 0.4415, + "step": 20357 + }, + { + "epoch": 0.9013647986008877, + "grad_norm": 1.4919582605361938, + "learning_rate": 5.874915433745077e-07, + "loss": 0.4836, + "step": 20358 + }, + { + "epoch": 0.9014090743057016, + "grad_norm": 1.683259129524231, + "learning_rate": 5.869697442900246e-07, + "loss": 0.4979, + "step": 20359 + }, + { + "epoch": 0.9014533500105155, + "grad_norm": 1.486788034439087, + "learning_rate": 5.864481700281066e-07, + "loss": 0.4668, + "step": 20360 + }, + { + "epoch": 0.9014976257153293, + "grad_norm": 1.3943358659744263, + "learning_rate": 5.859268206012092e-07, + "loss": 0.5201, + "step": 20361 + }, + { + "epoch": 0.9015419014201432, + "grad_norm": 1.5191624164581299, + "learning_rate": 5.854056960217835e-07, + "loss": 0.4914, + "step": 20362 + }, + { + "epoch": 0.9015861771249571, + "grad_norm": 1.5951603651046753, + "learning_rate": 5.848847963022775e-07, + "loss": 0.4815, + "step": 20363 + }, + { + "epoch": 0.901630452829771, + "grad_norm": 1.7051087617874146, + "learning_rate": 5.843641214551332e-07, + "loss": 0.4608, + "step": 20364 + }, + { + "epoch": 0.9016747285345849, + "grad_norm": 1.9420220851898193, + "learning_rate": 5.838436714927842e-07, + "loss": 0.5049, + "step": 20365 + }, + { + "epoch": 0.9017190042393988, + "grad_norm": 1.4834929704666138, + "learning_rate": 5.833234464276605e-07, + "loss": 0.4438, + "step": 20366 + }, + { + "epoch": 0.9017632799442126, + "grad_norm": 1.7322999238967896, + "learning_rate": 5.82803446272191e-07, + "loss": 0.5215, + "step": 20367 + }, + { + "epoch": 0.9018075556490265, + "grad_norm": 1.634202241897583, + "learning_rate": 5.822836710387936e-07, + "loss": 0.5123, + "step": 20368 + }, + { + "epoch": 0.9018518313538404, + "grad_norm": 1.5525662899017334, + "learning_rate": 5.817641207398806e-07, + "loss": 0.4741, + "step": 20369 + }, + { + "epoch": 0.9018961070586542, + "grad_norm": 1.66290283203125, + "learning_rate": 5.812447953878653e-07, + "loss": 0.497, + "step": 20370 + }, + { + "epoch": 0.9019403827634681, + "grad_norm": 2.353531837463379, + "learning_rate": 5.807256949951478e-07, + "loss": 0.5054, + "step": 20371 + }, + { + "epoch": 0.901984658468282, + "grad_norm": 1.4154911041259766, + "learning_rate": 5.802068195741284e-07, + "loss": 0.4696, + "step": 20372 + }, + { + "epoch": 0.9020289341730958, + "grad_norm": 1.5561667680740356, + "learning_rate": 5.796881691371991e-07, + "loss": 0.4912, + "step": 20373 + }, + { + "epoch": 0.9020732098779097, + "grad_norm": 1.4055490493774414, + "learning_rate": 5.791697436967481e-07, + "loss": 0.4563, + "step": 20374 + }, + { + "epoch": 0.9021174855827236, + "grad_norm": 1.5040026903152466, + "learning_rate": 5.786515432651563e-07, + "loss": 0.4494, + "step": 20375 + }, + { + "epoch": 0.9021617612875374, + "grad_norm": 2.1665914058685303, + "learning_rate": 5.781335678547994e-07, + "loss": 0.5414, + "step": 20376 + }, + { + "epoch": 0.9022060369923514, + "grad_norm": 2.0795891284942627, + "learning_rate": 5.776158174780533e-07, + "loss": 0.5357, + "step": 20377 + }, + { + "epoch": 0.9022503126971653, + "grad_norm": 1.6991409063339233, + "learning_rate": 5.770982921472801e-07, + "loss": 0.5132, + "step": 20378 + }, + { + "epoch": 0.9022945884019791, + "grad_norm": 1.6736737489700317, + "learning_rate": 5.76580991874841e-07, + "loss": 0.5227, + "step": 20379 + }, + { + "epoch": 0.902338864106793, + "grad_norm": 1.7285317182540894, + "learning_rate": 5.760639166730952e-07, + "loss": 0.5828, + "step": 20380 + }, + { + "epoch": 0.9023831398116069, + "grad_norm": 1.5329259634017944, + "learning_rate": 5.755470665543883e-07, + "loss": 0.5135, + "step": 20381 + }, + { + "epoch": 0.9024274155164208, + "grad_norm": 1.3291770219802856, + "learning_rate": 5.750304415310659e-07, + "loss": 0.484, + "step": 20382 + }, + { + "epoch": 0.9024716912212346, + "grad_norm": 1.7542263269424438, + "learning_rate": 5.745140416154649e-07, + "loss": 0.5397, + "step": 20383 + }, + { + "epoch": 0.9025159669260485, + "grad_norm": 1.6676018238067627, + "learning_rate": 5.739978668199242e-07, + "loss": 0.5142, + "step": 20384 + }, + { + "epoch": 0.9025602426308624, + "grad_norm": 2.0320630073547363, + "learning_rate": 5.734819171567685e-07, + "loss": 0.5253, + "step": 20385 + }, + { + "epoch": 0.9026045183356762, + "grad_norm": 1.4617794752120972, + "learning_rate": 5.729661926383201e-07, + "loss": 0.5176, + "step": 20386 + }, + { + "epoch": 0.9026487940404901, + "grad_norm": 1.3547426462173462, + "learning_rate": 5.724506932769014e-07, + "loss": 0.4571, + "step": 20387 + }, + { + "epoch": 0.902693069745304, + "grad_norm": 2.028928279876709, + "learning_rate": 5.719354190848203e-07, + "loss": 0.6037, + "step": 20388 + }, + { + "epoch": 0.9027373454501179, + "grad_norm": 1.3633170127868652, + "learning_rate": 5.714203700743859e-07, + "loss": 0.4259, + "step": 20389 + }, + { + "epoch": 0.9027816211549318, + "grad_norm": 1.595623254776001, + "learning_rate": 5.709055462578983e-07, + "loss": 0.48, + "step": 20390 + }, + { + "epoch": 0.9028258968597457, + "grad_norm": 1.3034802675247192, + "learning_rate": 5.703909476476554e-07, + "loss": 0.4127, + "step": 20391 + }, + { + "epoch": 0.9028701725645595, + "grad_norm": 1.6905354261398315, + "learning_rate": 5.698765742559454e-07, + "loss": 0.5332, + "step": 20392 + }, + { + "epoch": 0.9029144482693734, + "grad_norm": 1.6589469909667969, + "learning_rate": 5.693624260950547e-07, + "loss": 0.5033, + "step": 20393 + }, + { + "epoch": 0.9029587239741873, + "grad_norm": 1.7356126308441162, + "learning_rate": 5.688485031772651e-07, + "loss": 0.51, + "step": 20394 + }, + { + "epoch": 0.9030029996790011, + "grad_norm": 1.7049503326416016, + "learning_rate": 5.683348055148497e-07, + "loss": 0.5519, + "step": 20395 + }, + { + "epoch": 0.903047275383815, + "grad_norm": 1.750741720199585, + "learning_rate": 5.678213331200766e-07, + "loss": 0.5007, + "step": 20396 + }, + { + "epoch": 0.9030915510886289, + "grad_norm": 1.3253059387207031, + "learning_rate": 5.673080860052115e-07, + "loss": 0.4458, + "step": 20397 + }, + { + "epoch": 0.9031358267934427, + "grad_norm": 1.7483805418014526, + "learning_rate": 5.667950641825137e-07, + "loss": 0.4845, + "step": 20398 + }, + { + "epoch": 0.9031801024982566, + "grad_norm": 1.755322813987732, + "learning_rate": 5.662822676642344e-07, + "loss": 0.5007, + "step": 20399 + }, + { + "epoch": 0.9032243782030706, + "grad_norm": 2.19270658493042, + "learning_rate": 5.657696964626214e-07, + "loss": 0.5644, + "step": 20400 + }, + { + "epoch": 0.9032686539078844, + "grad_norm": 1.2210495471954346, + "learning_rate": 5.652573505899184e-07, + "loss": 0.4538, + "step": 20401 + }, + { + "epoch": 0.9033129296126983, + "grad_norm": 1.8377659320831299, + "learning_rate": 5.647452300583611e-07, + "loss": 0.515, + "step": 20402 + }, + { + "epoch": 0.9033572053175122, + "grad_norm": 1.4563535451889038, + "learning_rate": 5.64233334880181e-07, + "loss": 0.4681, + "step": 20403 + }, + { + "epoch": 0.903401481022326, + "grad_norm": 1.6425186395645142, + "learning_rate": 5.637216650676047e-07, + "loss": 0.5277, + "step": 20404 + }, + { + "epoch": 0.9034457567271399, + "grad_norm": 1.5896093845367432, + "learning_rate": 5.632102206328538e-07, + "loss": 0.4786, + "step": 20405 + }, + { + "epoch": 0.9034900324319538, + "grad_norm": 1.3150087594985962, + "learning_rate": 5.626990015881417e-07, + "loss": 0.4283, + "step": 20406 + }, + { + "epoch": 0.9035343081367676, + "grad_norm": 1.4169641733169556, + "learning_rate": 5.621880079456821e-07, + "loss": 0.4832, + "step": 20407 + }, + { + "epoch": 0.9035785838415815, + "grad_norm": 1.4482274055480957, + "learning_rate": 5.616772397176762e-07, + "loss": 0.4629, + "step": 20408 + }, + { + "epoch": 0.9036228595463954, + "grad_norm": 1.6369346380233765, + "learning_rate": 5.611666969163243e-07, + "loss": 0.5089, + "step": 20409 + }, + { + "epoch": 0.9036671352512092, + "grad_norm": 1.5131174325942993, + "learning_rate": 5.606563795538211e-07, + "loss": 0.4968, + "step": 20410 + }, + { + "epoch": 0.9037114109560231, + "grad_norm": 1.430405855178833, + "learning_rate": 5.601462876423547e-07, + "loss": 0.4758, + "step": 20411 + }, + { + "epoch": 0.903755686660837, + "grad_norm": 1.8059983253479004, + "learning_rate": 5.596364211941074e-07, + "loss": 0.5727, + "step": 20412 + }, + { + "epoch": 0.9037999623656509, + "grad_norm": 1.4193819761276245, + "learning_rate": 5.591267802212564e-07, + "loss": 0.4694, + "step": 20413 + }, + { + "epoch": 0.9038442380704648, + "grad_norm": 1.587398886680603, + "learning_rate": 5.586173647359772e-07, + "loss": 0.4779, + "step": 20414 + }, + { + "epoch": 0.9038885137752787, + "grad_norm": 1.6488803625106812, + "learning_rate": 5.581081747504335e-07, + "loss": 0.4906, + "step": 20415 + }, + { + "epoch": 0.9039327894800926, + "grad_norm": 1.640565276145935, + "learning_rate": 5.57599210276788e-07, + "loss": 0.4787, + "step": 20416 + }, + { + "epoch": 0.9039770651849064, + "grad_norm": 1.343273401260376, + "learning_rate": 5.570904713271974e-07, + "loss": 0.4761, + "step": 20417 + }, + { + "epoch": 0.9040213408897203, + "grad_norm": 1.5469235181808472, + "learning_rate": 5.565819579138132e-07, + "loss": 0.4667, + "step": 20418 + }, + { + "epoch": 0.9040656165945342, + "grad_norm": 1.671798825263977, + "learning_rate": 5.56073670048779e-07, + "loss": 0.5802, + "step": 20419 + }, + { + "epoch": 0.904109892299348, + "grad_norm": 1.7143021821975708, + "learning_rate": 5.555656077442362e-07, + "loss": 0.5047, + "step": 20420 + }, + { + "epoch": 0.9041541680041619, + "grad_norm": 1.5941298007965088, + "learning_rate": 5.550577710123184e-07, + "loss": 0.4858, + "step": 20421 + }, + { + "epoch": 0.9041984437089758, + "grad_norm": 1.4962767362594604, + "learning_rate": 5.54550159865156e-07, + "loss": 0.4936, + "step": 20422 + }, + { + "epoch": 0.9042427194137896, + "grad_norm": 1.954655408859253, + "learning_rate": 5.540427743148702e-07, + "loss": 0.5416, + "step": 20423 + }, + { + "epoch": 0.9042869951186036, + "grad_norm": 1.4072813987731934, + "learning_rate": 5.535356143735838e-07, + "loss": 0.44, + "step": 20424 + }, + { + "epoch": 0.9043312708234175, + "grad_norm": 1.4004298448562622, + "learning_rate": 5.530286800534068e-07, + "loss": 0.452, + "step": 20425 + }, + { + "epoch": 0.9043755465282313, + "grad_norm": 1.7312133312225342, + "learning_rate": 5.525219713664464e-07, + "loss": 0.5542, + "step": 20426 + }, + { + "epoch": 0.9044198222330452, + "grad_norm": 1.4927457571029663, + "learning_rate": 5.520154883248086e-07, + "loss": 0.5023, + "step": 20427 + }, + { + "epoch": 0.9044640979378591, + "grad_norm": 2.5826470851898193, + "learning_rate": 5.51509230940589e-07, + "loss": 0.6652, + "step": 20428 + }, + { + "epoch": 0.9045083736426729, + "grad_norm": 1.4527857303619385, + "learning_rate": 5.510031992258746e-07, + "loss": 0.4507, + "step": 20429 + }, + { + "epoch": 0.9045526493474868, + "grad_norm": 1.369053602218628, + "learning_rate": 5.504973931927571e-07, + "loss": 0.4503, + "step": 20430 + }, + { + "epoch": 0.9045969250523007, + "grad_norm": 1.4485329389572144, + "learning_rate": 5.499918128533155e-07, + "loss": 0.4568, + "step": 20431 + }, + { + "epoch": 0.9046412007571145, + "grad_norm": 1.7132068872451782, + "learning_rate": 5.494864582196246e-07, + "loss": 0.5207, + "step": 20432 + }, + { + "epoch": 0.9046854764619284, + "grad_norm": 1.9619953632354736, + "learning_rate": 5.489813293037538e-07, + "loss": 0.5272, + "step": 20433 + }, + { + "epoch": 0.9047297521667423, + "grad_norm": 1.1696009635925293, + "learning_rate": 5.484764261177699e-07, + "loss": 0.4194, + "step": 20434 + }, + { + "epoch": 0.9047740278715561, + "grad_norm": 1.565399408340454, + "learning_rate": 5.479717486737313e-07, + "loss": 0.4861, + "step": 20435 + }, + { + "epoch": 0.90481830357637, + "grad_norm": 1.7337205410003662, + "learning_rate": 5.474672969836903e-07, + "loss": 0.4667, + "step": 20436 + }, + { + "epoch": 0.904862579281184, + "grad_norm": 1.6852545738220215, + "learning_rate": 5.469630710596974e-07, + "loss": 0.4652, + "step": 20437 + }, + { + "epoch": 0.9049068549859978, + "grad_norm": 1.5657520294189453, + "learning_rate": 5.46459070913794e-07, + "loss": 0.4464, + "step": 20438 + }, + { + "epoch": 0.9049511306908117, + "grad_norm": 1.645595908164978, + "learning_rate": 5.459552965580172e-07, + "loss": 0.4435, + "step": 20439 + }, + { + "epoch": 0.9049954063956256, + "grad_norm": 1.7871373891830444, + "learning_rate": 5.454517480044009e-07, + "loss": 0.5467, + "step": 20440 + }, + { + "epoch": 0.9050396821004394, + "grad_norm": 1.5847798585891724, + "learning_rate": 5.449484252649728e-07, + "loss": 0.513, + "step": 20441 + }, + { + "epoch": 0.9050839578052533, + "grad_norm": 1.5814088582992554, + "learning_rate": 5.444453283517514e-07, + "loss": 0.5359, + "step": 20442 + }, + { + "epoch": 0.9051282335100672, + "grad_norm": 1.498591423034668, + "learning_rate": 5.439424572767538e-07, + "loss": 0.4846, + "step": 20443 + }, + { + "epoch": 0.905172509214881, + "grad_norm": 1.4334869384765625, + "learning_rate": 5.434398120519935e-07, + "loss": 0.4536, + "step": 20444 + }, + { + "epoch": 0.9052167849196949, + "grad_norm": 1.8498164415359497, + "learning_rate": 5.429373926894721e-07, + "loss": 0.5606, + "step": 20445 + }, + { + "epoch": 0.9052610606245088, + "grad_norm": 1.8687843084335327, + "learning_rate": 5.424351992011912e-07, + "loss": 0.5327, + "step": 20446 + }, + { + "epoch": 0.9053053363293226, + "grad_norm": 2.0239484310150146, + "learning_rate": 5.419332315991455e-07, + "loss": 0.542, + "step": 20447 + }, + { + "epoch": 0.9053496120341366, + "grad_norm": 1.399334192276001, + "learning_rate": 5.414314898953232e-07, + "loss": 0.4972, + "step": 20448 + }, + { + "epoch": 0.9053938877389505, + "grad_norm": 1.6418579816818237, + "learning_rate": 5.40929974101706e-07, + "loss": 0.4664, + "step": 20449 + }, + { + "epoch": 0.9054381634437643, + "grad_norm": 1.8337905406951904, + "learning_rate": 5.404286842302775e-07, + "loss": 0.5789, + "step": 20450 + }, + { + "epoch": 0.9054824391485782, + "grad_norm": 1.982297420501709, + "learning_rate": 5.399276202930071e-07, + "loss": 0.5157, + "step": 20451 + }, + { + "epoch": 0.9055267148533921, + "grad_norm": 1.4786463975906372, + "learning_rate": 5.394267823018617e-07, + "loss": 0.4738, + "step": 20452 + }, + { + "epoch": 0.905570990558206, + "grad_norm": 1.285019040107727, + "learning_rate": 5.389261702688042e-07, + "loss": 0.431, + "step": 20453 + }, + { + "epoch": 0.9056152662630198, + "grad_norm": 1.4312063455581665, + "learning_rate": 5.384257842057927e-07, + "loss": 0.4292, + "step": 20454 + }, + { + "epoch": 0.9056595419678337, + "grad_norm": 1.4835057258605957, + "learning_rate": 5.379256241247776e-07, + "loss": 0.4607, + "step": 20455 + }, + { + "epoch": 0.9057038176726476, + "grad_norm": 1.4910061359405518, + "learning_rate": 5.374256900377051e-07, + "loss": 0.4907, + "step": 20456 + }, + { + "epoch": 0.9057480933774614, + "grad_norm": 1.5729695558547974, + "learning_rate": 5.369259819565142e-07, + "loss": 0.4596, + "step": 20457 + }, + { + "epoch": 0.9057923690822753, + "grad_norm": 1.6808282136917114, + "learning_rate": 5.364264998931412e-07, + "loss": 0.5114, + "step": 20458 + }, + { + "epoch": 0.9058366447870893, + "grad_norm": 1.2281157970428467, + "learning_rate": 5.359272438595153e-07, + "loss": 0.421, + "step": 20459 + }, + { + "epoch": 0.905880920491903, + "grad_norm": 1.7043148279190063, + "learning_rate": 5.354282138675626e-07, + "loss": 0.5175, + "step": 20460 + }, + { + "epoch": 0.905925196196717, + "grad_norm": 1.3016952276229858, + "learning_rate": 5.349294099292001e-07, + "loss": 0.4413, + "step": 20461 + }, + { + "epoch": 0.9059694719015309, + "grad_norm": 2.6700997352600098, + "learning_rate": 5.344308320563418e-07, + "loss": 0.625, + "step": 20462 + }, + { + "epoch": 0.9060137476063447, + "grad_norm": 1.405896782875061, + "learning_rate": 5.339324802608947e-07, + "loss": 0.4909, + "step": 20463 + }, + { + "epoch": 0.9060580233111586, + "grad_norm": 2.1310765743255615, + "learning_rate": 5.334343545547638e-07, + "loss": 0.5679, + "step": 20464 + }, + { + "epoch": 0.9061022990159725, + "grad_norm": 1.951213002204895, + "learning_rate": 5.329364549498461e-07, + "loss": 0.5856, + "step": 20465 + }, + { + "epoch": 0.9061465747207863, + "grad_norm": 1.501834750175476, + "learning_rate": 5.324387814580323e-07, + "loss": 0.4797, + "step": 20466 + }, + { + "epoch": 0.9061908504256002, + "grad_norm": 1.3652795553207397, + "learning_rate": 5.319413340912094e-07, + "loss": 0.4477, + "step": 20467 + }, + { + "epoch": 0.9062351261304141, + "grad_norm": 1.560269832611084, + "learning_rate": 5.314441128612591e-07, + "loss": 0.4226, + "step": 20468 + }, + { + "epoch": 0.9062794018352279, + "grad_norm": 1.554045557975769, + "learning_rate": 5.309471177800552e-07, + "loss": 0.4793, + "step": 20469 + }, + { + "epoch": 0.9063236775400418, + "grad_norm": 1.413913369178772, + "learning_rate": 5.304503488594703e-07, + "loss": 0.464, + "step": 20470 + }, + { + "epoch": 0.9063679532448558, + "grad_norm": 1.6865768432617188, + "learning_rate": 5.299538061113685e-07, + "loss": 0.4896, + "step": 20471 + }, + { + "epoch": 0.9064122289496696, + "grad_norm": 1.4680227041244507, + "learning_rate": 5.29457489547609e-07, + "loss": 0.4269, + "step": 20472 + }, + { + "epoch": 0.9064565046544835, + "grad_norm": 1.6773654222488403, + "learning_rate": 5.289613991800447e-07, + "loss": 0.4614, + "step": 20473 + }, + { + "epoch": 0.9065007803592974, + "grad_norm": 1.4546010494232178, + "learning_rate": 5.284655350205281e-07, + "loss": 0.4472, + "step": 20474 + }, + { + "epoch": 0.9065450560641112, + "grad_norm": 1.4973423480987549, + "learning_rate": 5.279698970809011e-07, + "loss": 0.4504, + "step": 20475 + }, + { + "epoch": 0.9065893317689251, + "grad_norm": 1.406705617904663, + "learning_rate": 5.274744853729974e-07, + "loss": 0.4215, + "step": 20476 + }, + { + "epoch": 0.906633607473739, + "grad_norm": 1.9122062921524048, + "learning_rate": 5.269792999086542e-07, + "loss": 0.5063, + "step": 20477 + }, + { + "epoch": 0.9066778831785528, + "grad_norm": 1.5020437240600586, + "learning_rate": 5.264843406996978e-07, + "loss": 0.4902, + "step": 20478 + }, + { + "epoch": 0.9067221588833667, + "grad_norm": 1.3463239669799805, + "learning_rate": 5.259896077579473e-07, + "loss": 0.4288, + "step": 20479 + }, + { + "epoch": 0.9067664345881806, + "grad_norm": 1.4301111698150635, + "learning_rate": 5.254951010952237e-07, + "loss": 0.412, + "step": 20480 + }, + { + "epoch": 0.9068107102929944, + "grad_norm": 1.7510175704956055, + "learning_rate": 5.250008207233337e-07, + "loss": 0.456, + "step": 20481 + }, + { + "epoch": 0.9068549859978083, + "grad_norm": 1.447881817817688, + "learning_rate": 5.245067666540849e-07, + "loss": 0.4521, + "step": 20482 + }, + { + "epoch": 0.9068992617026223, + "grad_norm": 1.2590115070343018, + "learning_rate": 5.240129388992754e-07, + "loss": 0.4502, + "step": 20483 + }, + { + "epoch": 0.906943537407436, + "grad_norm": 1.74308180809021, + "learning_rate": 5.235193374707037e-07, + "loss": 0.5127, + "step": 20484 + }, + { + "epoch": 0.90698781311225, + "grad_norm": 1.551890254020691, + "learning_rate": 5.230259623801559e-07, + "loss": 0.5045, + "step": 20485 + }, + { + "epoch": 0.9070320888170639, + "grad_norm": 1.2674795389175415, + "learning_rate": 5.225328136394148e-07, + "loss": 0.4759, + "step": 20486 + }, + { + "epoch": 0.9070763645218778, + "grad_norm": 1.5553662776947021, + "learning_rate": 5.22039891260262e-07, + "loss": 0.5381, + "step": 20487 + }, + { + "epoch": 0.9071206402266916, + "grad_norm": 1.5655709505081177, + "learning_rate": 5.215471952544682e-07, + "loss": 0.4582, + "step": 20488 + }, + { + "epoch": 0.9071649159315055, + "grad_norm": 1.587679386138916, + "learning_rate": 5.210547256338027e-07, + "loss": 0.5047, + "step": 20489 + }, + { + "epoch": 0.9072091916363194, + "grad_norm": 1.4260897636413574, + "learning_rate": 5.205624824100264e-07, + "loss": 0.4461, + "step": 20490 + }, + { + "epoch": 0.9072534673411332, + "grad_norm": 1.4891756772994995, + "learning_rate": 5.200704655948974e-07, + "loss": 0.4844, + "step": 20491 + }, + { + "epoch": 0.9072977430459471, + "grad_norm": 1.7795121669769287, + "learning_rate": 5.195786752001664e-07, + "loss": 0.4743, + "step": 20492 + }, + { + "epoch": 0.907342018750761, + "grad_norm": 1.381648063659668, + "learning_rate": 5.190871112375795e-07, + "loss": 0.4374, + "step": 20493 + }, + { + "epoch": 0.9073862944555748, + "grad_norm": 1.3730523586273193, + "learning_rate": 5.185957737188774e-07, + "loss": 0.4318, + "step": 20494 + }, + { + "epoch": 0.9074305701603887, + "grad_norm": 1.6469255685806274, + "learning_rate": 5.181046626557961e-07, + "loss": 0.4595, + "step": 20495 + }, + { + "epoch": 0.9074748458652027, + "grad_norm": 1.662384033203125, + "learning_rate": 5.176137780600621e-07, + "loss": 0.4923, + "step": 20496 + }, + { + "epoch": 0.9075191215700165, + "grad_norm": 1.2257909774780273, + "learning_rate": 5.171231199434035e-07, + "loss": 0.4265, + "step": 20497 + }, + { + "epoch": 0.9075633972748304, + "grad_norm": 1.424660086631775, + "learning_rate": 5.166326883175377e-07, + "loss": 0.4659, + "step": 20498 + }, + { + "epoch": 0.9076076729796443, + "grad_norm": 1.7466636896133423, + "learning_rate": 5.161424831941797e-07, + "loss": 0.4942, + "step": 20499 + }, + { + "epoch": 0.9076519486844581, + "grad_norm": 1.0680781602859497, + "learning_rate": 5.156525045850336e-07, + "loss": 0.3663, + "step": 20500 + }, + { + "epoch": 0.907696224389272, + "grad_norm": 1.1960527896881104, + "learning_rate": 5.151627525018077e-07, + "loss": 0.4061, + "step": 20501 + }, + { + "epoch": 0.9077405000940859, + "grad_norm": 1.5997726917266846, + "learning_rate": 5.14673226956196e-07, + "loss": 0.5134, + "step": 20502 + }, + { + "epoch": 0.9077847757988997, + "grad_norm": 1.5296729803085327, + "learning_rate": 5.141839279598904e-07, + "loss": 0.4743, + "step": 20503 + }, + { + "epoch": 0.9078290515037136, + "grad_norm": 1.3030929565429688, + "learning_rate": 5.136948555245791e-07, + "loss": 0.4432, + "step": 20504 + }, + { + "epoch": 0.9078733272085275, + "grad_norm": 1.7278525829315186, + "learning_rate": 5.132060096619417e-07, + "loss": 0.4812, + "step": 20505 + }, + { + "epoch": 0.9079176029133413, + "grad_norm": 1.397680401802063, + "learning_rate": 5.127173903836536e-07, + "loss": 0.4485, + "step": 20506 + }, + { + "epoch": 0.9079618786181552, + "grad_norm": 1.5549503564834595, + "learning_rate": 5.122289977013862e-07, + "loss": 0.5145, + "step": 20507 + }, + { + "epoch": 0.9080061543229692, + "grad_norm": 1.612401008605957, + "learning_rate": 5.117408316268047e-07, + "loss": 0.5291, + "step": 20508 + }, + { + "epoch": 0.908050430027783, + "grad_norm": 1.6823927164077759, + "learning_rate": 5.112528921715665e-07, + "loss": 0.5713, + "step": 20509 + }, + { + "epoch": 0.9080947057325969, + "grad_norm": 1.3781977891921997, + "learning_rate": 5.107651793473268e-07, + "loss": 0.4465, + "step": 20510 + }, + { + "epoch": 0.9081389814374108, + "grad_norm": 1.7148354053497314, + "learning_rate": 5.102776931657349e-07, + "loss": 0.5205, + "step": 20511 + }, + { + "epoch": 0.9081832571422246, + "grad_norm": 1.477587103843689, + "learning_rate": 5.09790433638434e-07, + "loss": 0.4326, + "step": 20512 + }, + { + "epoch": 0.9082275328470385, + "grad_norm": 1.5962140560150146, + "learning_rate": 5.093034007770625e-07, + "loss": 0.4681, + "step": 20513 + }, + { + "epoch": 0.9082718085518524, + "grad_norm": 1.251417636871338, + "learning_rate": 5.0881659459325e-07, + "loss": 0.4011, + "step": 20514 + }, + { + "epoch": 0.9083160842566662, + "grad_norm": 1.6564524173736572, + "learning_rate": 5.083300150986259e-07, + "loss": 0.5207, + "step": 20515 + }, + { + "epoch": 0.9083603599614801, + "grad_norm": 1.7050871849060059, + "learning_rate": 5.078436623048089e-07, + "loss": 0.5197, + "step": 20516 + }, + { + "epoch": 0.908404635666294, + "grad_norm": 2.0749008655548096, + "learning_rate": 5.073575362234196e-07, + "loss": 0.4604, + "step": 20517 + }, + { + "epoch": 0.9084489113711078, + "grad_norm": 1.5331337451934814, + "learning_rate": 5.068716368660653e-07, + "loss": 0.4748, + "step": 20518 + }, + { + "epoch": 0.9084931870759217, + "grad_norm": 1.5959851741790771, + "learning_rate": 5.063859642443536e-07, + "loss": 0.4905, + "step": 20519 + }, + { + "epoch": 0.9085374627807357, + "grad_norm": 1.2894953489303589, + "learning_rate": 5.059005183698806e-07, + "loss": 0.5042, + "step": 20520 + }, + { + "epoch": 0.9085817384855495, + "grad_norm": 1.594132900238037, + "learning_rate": 5.05415299254245e-07, + "loss": 0.4459, + "step": 20521 + }, + { + "epoch": 0.9086260141903634, + "grad_norm": 1.6301711797714233, + "learning_rate": 5.04930306909035e-07, + "loss": 0.497, + "step": 20522 + }, + { + "epoch": 0.9086702898951773, + "grad_norm": 1.3781734704971313, + "learning_rate": 5.044455413458327e-07, + "loss": 0.4333, + "step": 20523 + }, + { + "epoch": 0.9087145655999912, + "grad_norm": 1.4244738817214966, + "learning_rate": 5.039610025762176e-07, + "loss": 0.4294, + "step": 20524 + }, + { + "epoch": 0.908758841304805, + "grad_norm": 1.2345837354660034, + "learning_rate": 5.034766906117617e-07, + "loss": 0.4122, + "step": 20525 + }, + { + "epoch": 0.9088031170096189, + "grad_norm": 1.5914419889450073, + "learning_rate": 5.029926054640311e-07, + "loss": 0.5033, + "step": 20526 + }, + { + "epoch": 0.9088473927144328, + "grad_norm": 1.2239866256713867, + "learning_rate": 5.025087471445899e-07, + "loss": 0.4012, + "step": 20527 + }, + { + "epoch": 0.9088916684192466, + "grad_norm": 1.73638117313385, + "learning_rate": 5.020251156649958e-07, + "loss": 0.5262, + "step": 20528 + }, + { + "epoch": 0.9089359441240605, + "grad_norm": 2.1719143390655518, + "learning_rate": 5.01541711036797e-07, + "loss": 0.6178, + "step": 20529 + }, + { + "epoch": 0.9089802198288744, + "grad_norm": 1.419818639755249, + "learning_rate": 5.010585332715401e-07, + "loss": 0.4792, + "step": 20530 + }, + { + "epoch": 0.9090244955336882, + "grad_norm": 1.3684747219085693, + "learning_rate": 5.005755823807678e-07, + "loss": 0.4927, + "step": 20531 + }, + { + "epoch": 0.9090687712385022, + "grad_norm": 1.3390264511108398, + "learning_rate": 5.000928583760123e-07, + "loss": 0.4147, + "step": 20532 + }, + { + "epoch": 0.9091130469433161, + "grad_norm": 1.5186262130737305, + "learning_rate": 4.996103612688019e-07, + "loss": 0.4862, + "step": 20533 + }, + { + "epoch": 0.9091573226481299, + "grad_norm": 1.6869953870773315, + "learning_rate": 4.991280910706642e-07, + "loss": 0.4991, + "step": 20534 + }, + { + "epoch": 0.9092015983529438, + "grad_norm": 1.4064620733261108, + "learning_rate": 4.986460477931154e-07, + "loss": 0.5268, + "step": 20535 + }, + { + "epoch": 0.9092458740577577, + "grad_norm": 2.133371353149414, + "learning_rate": 4.981642314476698e-07, + "loss": 0.495, + "step": 20536 + }, + { + "epoch": 0.9092901497625715, + "grad_norm": 1.675194263458252, + "learning_rate": 4.976826420458347e-07, + "loss": 0.504, + "step": 20537 + }, + { + "epoch": 0.9093344254673854, + "grad_norm": 1.7280690670013428, + "learning_rate": 4.972012795991132e-07, + "loss": 0.511, + "step": 20538 + }, + { + "epoch": 0.9093787011721993, + "grad_norm": 1.3934212923049927, + "learning_rate": 4.967201441190017e-07, + "loss": 0.4492, + "step": 20539 + }, + { + "epoch": 0.9094229768770131, + "grad_norm": 1.5013664960861206, + "learning_rate": 4.962392356169909e-07, + "loss": 0.5008, + "step": 20540 + }, + { + "epoch": 0.909467252581827, + "grad_norm": 1.4630959033966064, + "learning_rate": 4.957585541045684e-07, + "loss": 0.4433, + "step": 20541 + }, + { + "epoch": 0.909511528286641, + "grad_norm": 1.514255166053772, + "learning_rate": 4.952780995932138e-07, + "loss": 0.4451, + "step": 20542 + }, + { + "epoch": 0.9095558039914547, + "grad_norm": 1.6737487316131592, + "learning_rate": 4.947978720944025e-07, + "loss": 0.4828, + "step": 20543 + }, + { + "epoch": 0.9096000796962687, + "grad_norm": 1.5114282369613647, + "learning_rate": 4.943178716196051e-07, + "loss": 0.5008, + "step": 20544 + }, + { + "epoch": 0.9096443554010826, + "grad_norm": 1.586425542831421, + "learning_rate": 4.938380981802848e-07, + "loss": 0.4998, + "step": 20545 + }, + { + "epoch": 0.9096886311058964, + "grad_norm": 1.3380979299545288, + "learning_rate": 4.933585517879014e-07, + "loss": 0.4331, + "step": 20546 + }, + { + "epoch": 0.9097329068107103, + "grad_norm": 1.0839927196502686, + "learning_rate": 4.928792324539089e-07, + "loss": 0.376, + "step": 20547 + }, + { + "epoch": 0.9097771825155242, + "grad_norm": 1.5005980730056763, + "learning_rate": 4.924001401897549e-07, + "loss": 0.4747, + "step": 20548 + }, + { + "epoch": 0.909821458220338, + "grad_norm": 1.5010559558868408, + "learning_rate": 4.919212750068825e-07, + "loss": 0.4783, + "step": 20549 + }, + { + "epoch": 0.9098657339251519, + "grad_norm": 1.6968472003936768, + "learning_rate": 4.914426369167291e-07, + "loss": 0.4724, + "step": 20550 + }, + { + "epoch": 0.9099100096299658, + "grad_norm": 1.8530060052871704, + "learning_rate": 4.909642259307257e-07, + "loss": 0.4581, + "step": 20551 + }, + { + "epoch": 0.9099542853347796, + "grad_norm": 1.3197476863861084, + "learning_rate": 4.904860420602986e-07, + "loss": 0.4629, + "step": 20552 + }, + { + "epoch": 0.9099985610395935, + "grad_norm": 1.449960708618164, + "learning_rate": 4.900080853168687e-07, + "loss": 0.4516, + "step": 20553 + }, + { + "epoch": 0.9100428367444074, + "grad_norm": 1.5839359760284424, + "learning_rate": 4.895303557118536e-07, + "loss": 0.5072, + "step": 20554 + }, + { + "epoch": 0.9100871124492212, + "grad_norm": 1.8137058019638062, + "learning_rate": 4.890528532566629e-07, + "loss": 0.5031, + "step": 20555 + }, + { + "epoch": 0.9101313881540352, + "grad_norm": 1.468847632408142, + "learning_rate": 4.885755779626988e-07, + "loss": 0.4796, + "step": 20556 + }, + { + "epoch": 0.9101756638588491, + "grad_norm": 1.4712430238723755, + "learning_rate": 4.880985298413632e-07, + "loss": 0.4369, + "step": 20557 + }, + { + "epoch": 0.910219939563663, + "grad_norm": 1.3297396898269653, + "learning_rate": 4.876217089040502e-07, + "loss": 0.4492, + "step": 20558 + }, + { + "epoch": 0.9102642152684768, + "grad_norm": 1.6465760469436646, + "learning_rate": 4.871451151621475e-07, + "loss": 0.4593, + "step": 20559 + }, + { + "epoch": 0.9103084909732907, + "grad_norm": 1.4077565670013428, + "learning_rate": 4.866687486270372e-07, + "loss": 0.4218, + "step": 20560 + }, + { + "epoch": 0.9103527666781046, + "grad_norm": 1.266677737236023, + "learning_rate": 4.861926093100988e-07, + "loss": 0.47, + "step": 20561 + }, + { + "epoch": 0.9103970423829184, + "grad_norm": 1.4478453397750854, + "learning_rate": 4.857166972227035e-07, + "loss": 0.449, + "step": 20562 + }, + { + "epoch": 0.9104413180877323, + "grad_norm": 1.656765103340149, + "learning_rate": 4.852410123762164e-07, + "loss": 0.5149, + "step": 20563 + }, + { + "epoch": 0.9104855937925462, + "grad_norm": 1.8224127292633057, + "learning_rate": 4.847655547820019e-07, + "loss": 0.4781, + "step": 20564 + }, + { + "epoch": 0.91052986949736, + "grad_norm": 1.308947205543518, + "learning_rate": 4.842903244514142e-07, + "loss": 0.4699, + "step": 20565 + }, + { + "epoch": 0.910574145202174, + "grad_norm": 2.4566128253936768, + "learning_rate": 4.83815321395803e-07, + "loss": 0.5451, + "step": 20566 + }, + { + "epoch": 0.9106184209069879, + "grad_norm": 1.4944342374801636, + "learning_rate": 4.83340545626515e-07, + "loss": 0.4354, + "step": 20567 + }, + { + "epoch": 0.9106626966118017, + "grad_norm": 1.6030023097991943, + "learning_rate": 4.828659971548911e-07, + "loss": 0.5092, + "step": 20568 + }, + { + "epoch": 0.9107069723166156, + "grad_norm": 1.4576855897903442, + "learning_rate": 4.82391675992262e-07, + "loss": 0.4982, + "step": 20569 + }, + { + "epoch": 0.9107512480214295, + "grad_norm": 1.5819551944732666, + "learning_rate": 4.819175821499589e-07, + "loss": 0.4912, + "step": 20570 + }, + { + "epoch": 0.9107955237262433, + "grad_norm": 1.5290879011154175, + "learning_rate": 4.814437156393048e-07, + "loss": 0.4833, + "step": 20571 + }, + { + "epoch": 0.9108397994310572, + "grad_norm": 1.6078910827636719, + "learning_rate": 4.809700764716163e-07, + "loss": 0.4568, + "step": 20572 + }, + { + "epoch": 0.9108840751358711, + "grad_norm": 1.8877339363098145, + "learning_rate": 4.804966646582065e-07, + "loss": 0.524, + "step": 20573 + }, + { + "epoch": 0.9109283508406849, + "grad_norm": 1.480698585510254, + "learning_rate": 4.800234802103842e-07, + "loss": 0.4863, + "step": 20574 + }, + { + "epoch": 0.9109726265454988, + "grad_norm": 1.6641156673431396, + "learning_rate": 4.795505231394493e-07, + "loss": 0.5063, + "step": 20575 + }, + { + "epoch": 0.9110169022503127, + "grad_norm": 1.5860950946807861, + "learning_rate": 4.79077793456697e-07, + "loss": 0.5039, + "step": 20576 + }, + { + "epoch": 0.9110611779551265, + "grad_norm": 1.4491558074951172, + "learning_rate": 4.786052911734196e-07, + "loss": 0.4801, + "step": 20577 + }, + { + "epoch": 0.9111054536599404, + "grad_norm": 1.5860861539840698, + "learning_rate": 4.781330163009046e-07, + "loss": 0.5166, + "step": 20578 + }, + { + "epoch": 0.9111497293647544, + "grad_norm": 1.5254727602005005, + "learning_rate": 4.776609688504286e-07, + "loss": 0.5545, + "step": 20579 + }, + { + "epoch": 0.9111940050695682, + "grad_norm": 1.459182620048523, + "learning_rate": 4.771891488332647e-07, + "loss": 0.478, + "step": 20580 + }, + { + "epoch": 0.9112382807743821, + "grad_norm": 1.3706576824188232, + "learning_rate": 4.7671755626068627e-07, + "loss": 0.4491, + "step": 20581 + }, + { + "epoch": 0.911282556479196, + "grad_norm": 1.6217939853668213, + "learning_rate": 4.762461911439553e-07, + "loss": 0.5108, + "step": 20582 + }, + { + "epoch": 0.9113268321840098, + "grad_norm": 1.5302000045776367, + "learning_rate": 4.757750534943273e-07, + "loss": 0.4524, + "step": 20583 + }, + { + "epoch": 0.9113711078888237, + "grad_norm": 2.4017720222473145, + "learning_rate": 4.7530414332305876e-07, + "loss": 0.5716, + "step": 20584 + }, + { + "epoch": 0.9114153835936376, + "grad_norm": 1.4801881313323975, + "learning_rate": 4.7483346064139513e-07, + "loss": 0.471, + "step": 20585 + }, + { + "epoch": 0.9114596592984514, + "grad_norm": 1.5321085453033447, + "learning_rate": 4.7436300546057746e-07, + "loss": 0.5035, + "step": 20586 + }, + { + "epoch": 0.9115039350032653, + "grad_norm": 1.6390513181686401, + "learning_rate": 4.738927777918467e-07, + "loss": 0.4469, + "step": 20587 + }, + { + "epoch": 0.9115482107080792, + "grad_norm": 1.3344749212265015, + "learning_rate": 4.734227776464284e-07, + "loss": 0.4239, + "step": 20588 + }, + { + "epoch": 0.911592486412893, + "grad_norm": 1.5224868059158325, + "learning_rate": 4.729530050355502e-07, + "loss": 0.453, + "step": 20589 + }, + { + "epoch": 0.911636762117707, + "grad_norm": 1.7676119804382324, + "learning_rate": 4.724834599704309e-07, + "loss": 0.471, + "step": 20590 + }, + { + "epoch": 0.9116810378225209, + "grad_norm": 1.4442811012268066, + "learning_rate": 4.720141424622882e-07, + "loss": 0.4888, + "step": 20591 + }, + { + "epoch": 0.9117253135273348, + "grad_norm": 1.6905927658081055, + "learning_rate": 4.7154505252232976e-07, + "loss": 0.5097, + "step": 20592 + }, + { + "epoch": 0.9117695892321486, + "grad_norm": 1.369344711303711, + "learning_rate": 4.710761901617589e-07, + "loss": 0.4141, + "step": 20593 + }, + { + "epoch": 0.9118138649369625, + "grad_norm": 1.4702106714248657, + "learning_rate": 4.7060755539177437e-07, + "loss": 0.4706, + "step": 20594 + }, + { + "epoch": 0.9118581406417764, + "grad_norm": 1.3446524143218994, + "learning_rate": 4.7013914822356956e-07, + "loss": 0.4702, + "step": 20595 + }, + { + "epoch": 0.9119024163465902, + "grad_norm": 1.6081197261810303, + "learning_rate": 4.6967096866833204e-07, + "loss": 0.5059, + "step": 20596 + }, + { + "epoch": 0.9119466920514041, + "grad_norm": 1.4547926187515259, + "learning_rate": 4.692030167372441e-07, + "loss": 0.4821, + "step": 20597 + }, + { + "epoch": 0.911990967756218, + "grad_norm": 1.5140154361724854, + "learning_rate": 4.687352924414812e-07, + "loss": 0.4974, + "step": 20598 + }, + { + "epoch": 0.9120352434610318, + "grad_norm": 1.4718656539916992, + "learning_rate": 4.682677957922155e-07, + "loss": 0.4632, + "step": 20599 + }, + { + "epoch": 0.9120795191658457, + "grad_norm": 1.5210912227630615, + "learning_rate": 4.678005268006103e-07, + "loss": 0.4683, + "step": 20600 + }, + { + "epoch": 0.9121237948706596, + "grad_norm": 1.4863814115524292, + "learning_rate": 4.6733348547782995e-07, + "loss": 0.516, + "step": 20601 + }, + { + "epoch": 0.9121680705754734, + "grad_norm": 1.6949281692504883, + "learning_rate": 4.668666718350279e-07, + "loss": 0.5283, + "step": 20602 + }, + { + "epoch": 0.9122123462802874, + "grad_norm": 1.5268529653549194, + "learning_rate": 4.6640008588335175e-07, + "loss": 0.4673, + "step": 20603 + }, + { + "epoch": 0.9122566219851013, + "grad_norm": 1.7859925031661987, + "learning_rate": 4.6593372763394816e-07, + "loss": 0.5206, + "step": 20604 + }, + { + "epoch": 0.9123008976899151, + "grad_norm": 1.7546051740646362, + "learning_rate": 4.6546759709795387e-07, + "loss": 0.5511, + "step": 20605 + }, + { + "epoch": 0.912345173394729, + "grad_norm": 1.4681777954101562, + "learning_rate": 4.6500169428650434e-07, + "loss": 0.4824, + "step": 20606 + }, + { + "epoch": 0.9123894490995429, + "grad_norm": 1.501806378364563, + "learning_rate": 4.64536019210724e-07, + "loss": 0.4878, + "step": 20607 + }, + { + "epoch": 0.9124337248043567, + "grad_norm": 1.5368865728378296, + "learning_rate": 4.640705718817373e-07, + "loss": 0.535, + "step": 20608 + }, + { + "epoch": 0.9124780005091706, + "grad_norm": 1.6311277151107788, + "learning_rate": 4.6360535231066096e-07, + "loss": 0.5224, + "step": 20609 + }, + { + "epoch": 0.9125222762139845, + "grad_norm": 1.607898235321045, + "learning_rate": 4.631403605086049e-07, + "loss": 0.4778, + "step": 20610 + }, + { + "epoch": 0.9125665519187983, + "grad_norm": 1.7739604711532593, + "learning_rate": 4.6267559648667694e-07, + "loss": 0.4919, + "step": 20611 + }, + { + "epoch": 0.9126108276236122, + "grad_norm": 1.399051308631897, + "learning_rate": 4.6221106025597705e-07, + "loss": 0.4663, + "step": 20612 + }, + { + "epoch": 0.9126551033284261, + "grad_norm": 1.6341341733932495, + "learning_rate": 4.6174675182759755e-07, + "loss": 0.4601, + "step": 20613 + }, + { + "epoch": 0.9126993790332399, + "grad_norm": 1.4782570600509644, + "learning_rate": 4.612826712126328e-07, + "loss": 0.4969, + "step": 20614 + }, + { + "epoch": 0.9127436547380539, + "grad_norm": 1.4049758911132812, + "learning_rate": 4.6081881842216516e-07, + "loss": 0.4416, + "step": 20615 + }, + { + "epoch": 0.9127879304428678, + "grad_norm": 1.6765389442443848, + "learning_rate": 4.603551934672723e-07, + "loss": 0.4751, + "step": 20616 + }, + { + "epoch": 0.9128322061476816, + "grad_norm": 1.5600571632385254, + "learning_rate": 4.598917963590288e-07, + "loss": 0.5077, + "step": 20617 + }, + { + "epoch": 0.9128764818524955, + "grad_norm": 1.785385251045227, + "learning_rate": 4.5942862710850135e-07, + "loss": 0.6141, + "step": 20618 + }, + { + "epoch": 0.9129207575573094, + "grad_norm": 1.549349069595337, + "learning_rate": 4.5896568572675327e-07, + "loss": 0.5345, + "step": 20619 + }, + { + "epoch": 0.9129650332621232, + "grad_norm": 1.555624008178711, + "learning_rate": 4.585029722248391e-07, + "loss": 0.4892, + "step": 20620 + }, + { + "epoch": 0.9130093089669371, + "grad_norm": 1.3316936492919922, + "learning_rate": 4.580404866138144e-07, + "loss": 0.4625, + "step": 20621 + }, + { + "epoch": 0.913053584671751, + "grad_norm": 1.6962603330612183, + "learning_rate": 4.575782289047237e-07, + "loss": 0.5122, + "step": 20622 + }, + { + "epoch": 0.9130978603765648, + "grad_norm": 1.8127903938293457, + "learning_rate": 4.5711619910860596e-07, + "loss": 0.532, + "step": 20623 + }, + { + "epoch": 0.9131421360813787, + "grad_norm": 1.5522253513336182, + "learning_rate": 4.5665439723649786e-07, + "loss": 0.4906, + "step": 20624 + }, + { + "epoch": 0.9131864117861926, + "grad_norm": 1.9962183237075806, + "learning_rate": 4.5619282329942946e-07, + "loss": 0.5283, + "step": 20625 + }, + { + "epoch": 0.9132306874910064, + "grad_norm": 1.4860390424728394, + "learning_rate": 4.557314773084265e-07, + "loss": 0.5171, + "step": 20626 + }, + { + "epoch": 0.9132749631958204, + "grad_norm": 1.5423301458358765, + "learning_rate": 4.5527035927450337e-07, + "loss": 0.5036, + "step": 20627 + }, + { + "epoch": 0.9133192389006343, + "grad_norm": 1.681382179260254, + "learning_rate": 4.5480946920867685e-07, + "loss": 0.503, + "step": 20628 + }, + { + "epoch": 0.9133635146054482, + "grad_norm": 1.5823935270309448, + "learning_rate": 4.5434880712195485e-07, + "loss": 0.5148, + "step": 20629 + }, + { + "epoch": 0.913407790310262, + "grad_norm": 1.580978274345398, + "learning_rate": 4.5388837302533737e-07, + "loss": 0.5278, + "step": 20630 + }, + { + "epoch": 0.9134520660150759, + "grad_norm": 1.6530282497406006, + "learning_rate": 4.534281669298246e-07, + "loss": 0.5566, + "step": 20631 + }, + { + "epoch": 0.9134963417198898, + "grad_norm": 1.3728797435760498, + "learning_rate": 4.529681888464077e-07, + "loss": 0.4303, + "step": 20632 + }, + { + "epoch": 0.9135406174247036, + "grad_norm": 1.446286916732788, + "learning_rate": 4.525084387860701e-07, + "loss": 0.497, + "step": 20633 + }, + { + "epoch": 0.9135848931295175, + "grad_norm": 1.6265339851379395, + "learning_rate": 4.5204891675979746e-07, + "loss": 0.5061, + "step": 20634 + }, + { + "epoch": 0.9136291688343314, + "grad_norm": 1.849304437637329, + "learning_rate": 4.51589622778561e-07, + "loss": 0.5481, + "step": 20635 + }, + { + "epoch": 0.9136734445391452, + "grad_norm": 1.5552949905395508, + "learning_rate": 4.5113055685333084e-07, + "loss": 0.4991, + "step": 20636 + }, + { + "epoch": 0.9137177202439591, + "grad_norm": 1.562853217124939, + "learning_rate": 4.506717189950716e-07, + "loss": 0.4921, + "step": 20637 + }, + { + "epoch": 0.913761995948773, + "grad_norm": 1.381474494934082, + "learning_rate": 4.5021310921474547e-07, + "loss": 0.4774, + "step": 20638 + }, + { + "epoch": 0.9138062716535869, + "grad_norm": 1.589959979057312, + "learning_rate": 4.497547275233016e-07, + "loss": 0.508, + "step": 20639 + }, + { + "epoch": 0.9138505473584008, + "grad_norm": 1.7270225286483765, + "learning_rate": 4.492965739316901e-07, + "loss": 0.5256, + "step": 20640 + }, + { + "epoch": 0.9138948230632147, + "grad_norm": 1.4639899730682373, + "learning_rate": 4.4883864845085447e-07, + "loss": 0.4819, + "step": 20641 + }, + { + "epoch": 0.9139390987680285, + "grad_norm": 1.4244813919067383, + "learning_rate": 4.483809510917303e-07, + "loss": 0.4468, + "step": 20642 + }, + { + "epoch": 0.9139833744728424, + "grad_norm": 1.7381271123886108, + "learning_rate": 4.4792348186525e-07, + "loss": 0.5142, + "step": 20643 + }, + { + "epoch": 0.9140276501776563, + "grad_norm": 1.3919316530227661, + "learning_rate": 4.474662407823405e-07, + "loss": 0.5183, + "step": 20644 + }, + { + "epoch": 0.9140719258824701, + "grad_norm": 1.342831015586853, + "learning_rate": 4.4700922785392174e-07, + "loss": 0.453, + "step": 20645 + }, + { + "epoch": 0.914116201587284, + "grad_norm": 1.7803833484649658, + "learning_rate": 4.465524430909085e-07, + "loss": 0.4962, + "step": 20646 + }, + { + "epoch": 0.9141604772920979, + "grad_norm": 1.5002937316894531, + "learning_rate": 4.460958865042109e-07, + "loss": 0.4705, + "step": 20647 + }, + { + "epoch": 0.9142047529969117, + "grad_norm": 1.643770694732666, + "learning_rate": 4.4563955810473456e-07, + "loss": 0.5007, + "step": 20648 + }, + { + "epoch": 0.9142490287017256, + "grad_norm": 1.993936538696289, + "learning_rate": 4.451834579033787e-07, + "loss": 0.4893, + "step": 20649 + }, + { + "epoch": 0.9142933044065396, + "grad_norm": 2.0737743377685547, + "learning_rate": 4.447275859110345e-07, + "loss": 0.5303, + "step": 20650 + }, + { + "epoch": 0.9143375801113534, + "grad_norm": 2.279472827911377, + "learning_rate": 4.4427194213859216e-07, + "loss": 0.5407, + "step": 20651 + }, + { + "epoch": 0.9143818558161673, + "grad_norm": 1.3672467470169067, + "learning_rate": 4.438165265969352e-07, + "loss": 0.4354, + "step": 20652 + }, + { + "epoch": 0.9144261315209812, + "grad_norm": 1.3811601400375366, + "learning_rate": 4.4336133929693937e-07, + "loss": 0.4567, + "step": 20653 + }, + { + "epoch": 0.914470407225795, + "grad_norm": 1.5105829238891602, + "learning_rate": 4.4290638024947596e-07, + "loss": 0.4485, + "step": 20654 + }, + { + "epoch": 0.9145146829306089, + "grad_norm": 1.3032338619232178, + "learning_rate": 4.424516494654119e-07, + "loss": 0.4184, + "step": 20655 + }, + { + "epoch": 0.9145589586354228, + "grad_norm": 1.3899303674697876, + "learning_rate": 4.419971469556095e-07, + "loss": 0.4186, + "step": 20656 + }, + { + "epoch": 0.9146032343402366, + "grad_norm": 1.6614463329315186, + "learning_rate": 4.415428727309201e-07, + "loss": 0.5391, + "step": 20657 + }, + { + "epoch": 0.9146475100450505, + "grad_norm": 1.3390089273452759, + "learning_rate": 4.4108882680219845e-07, + "loss": 0.4508, + "step": 20658 + }, + { + "epoch": 0.9146917857498644, + "grad_norm": 1.8655482530593872, + "learning_rate": 4.406350091802869e-07, + "loss": 0.4797, + "step": 20659 + }, + { + "epoch": 0.9147360614546782, + "grad_norm": 1.7310459613800049, + "learning_rate": 4.4018141987602347e-07, + "loss": 0.5282, + "step": 20660 + }, + { + "epoch": 0.9147803371594921, + "grad_norm": 1.311736822128296, + "learning_rate": 4.3972805890024396e-07, + "loss": 0.4573, + "step": 20661 + }, + { + "epoch": 0.914824612864306, + "grad_norm": 1.5845513343811035, + "learning_rate": 4.3927492626377524e-07, + "loss": 0.4791, + "step": 20662 + }, + { + "epoch": 0.91486888856912, + "grad_norm": 1.7544811964035034, + "learning_rate": 4.3882202197744085e-07, + "loss": 0.5151, + "step": 20663 + }, + { + "epoch": 0.9149131642739338, + "grad_norm": 1.3628109693527222, + "learning_rate": 4.383693460520566e-07, + "loss": 0.468, + "step": 20664 + }, + { + "epoch": 0.9149574399787477, + "grad_norm": 1.2465163469314575, + "learning_rate": 4.3791689849843613e-07, + "loss": 0.4364, + "step": 20665 + }, + { + "epoch": 0.9150017156835616, + "grad_norm": 1.7435898780822754, + "learning_rate": 4.3746467932738404e-07, + "loss": 0.4931, + "step": 20666 + }, + { + "epoch": 0.9150459913883754, + "grad_norm": 1.6794809103012085, + "learning_rate": 4.370126885497017e-07, + "loss": 0.4994, + "step": 20667 + }, + { + "epoch": 0.9150902670931893, + "grad_norm": 1.6803839206695557, + "learning_rate": 4.3656092617618493e-07, + "loss": 0.4976, + "step": 20668 + }, + { + "epoch": 0.9151345427980032, + "grad_norm": 1.5257154703140259, + "learning_rate": 4.3610939221762407e-07, + "loss": 0.5044, + "step": 20669 + }, + { + "epoch": 0.915178818502817, + "grad_norm": 1.4418522119522095, + "learning_rate": 4.356580866848026e-07, + "loss": 0.4863, + "step": 20670 + }, + { + "epoch": 0.9152230942076309, + "grad_norm": 1.270998477935791, + "learning_rate": 4.352070095885008e-07, + "loss": 0.4532, + "step": 20671 + }, + { + "epoch": 0.9152673699124448, + "grad_norm": 1.754681944847107, + "learning_rate": 4.3475616093949126e-07, + "loss": 0.5423, + "step": 20672 + }, + { + "epoch": 0.9153116456172586, + "grad_norm": 2.0605432987213135, + "learning_rate": 4.34305540748543e-07, + "loss": 0.563, + "step": 20673 + }, + { + "epoch": 0.9153559213220726, + "grad_norm": 1.770825982093811, + "learning_rate": 4.3385514902641756e-07, + "loss": 0.4649, + "step": 20674 + }, + { + "epoch": 0.9154001970268865, + "grad_norm": 1.5788463354110718, + "learning_rate": 4.33404985783874e-07, + "loss": 0.4975, + "step": 20675 + }, + { + "epoch": 0.9154444727317003, + "grad_norm": 1.7653896808624268, + "learning_rate": 4.329550510316627e-07, + "loss": 0.5777, + "step": 20676 + }, + { + "epoch": 0.9154887484365142, + "grad_norm": 1.5599703788757324, + "learning_rate": 4.325053447805283e-07, + "loss": 0.474, + "step": 20677 + }, + { + "epoch": 0.9155330241413281, + "grad_norm": 1.1750518083572388, + "learning_rate": 4.320558670412145e-07, + "loss": 0.3939, + "step": 20678 + }, + { + "epoch": 0.9155772998461419, + "grad_norm": 1.6536332368850708, + "learning_rate": 4.3160661782445713e-07, + "loss": 0.4692, + "step": 20679 + }, + { + "epoch": 0.9156215755509558, + "grad_norm": 1.616308331489563, + "learning_rate": 4.3115759714098315e-07, + "loss": 0.5266, + "step": 20680 + }, + { + "epoch": 0.9156658512557697, + "grad_norm": 1.8035660982131958, + "learning_rate": 4.3070880500151955e-07, + "loss": 0.5446, + "step": 20681 + }, + { + "epoch": 0.9157101269605835, + "grad_norm": 1.5812627077102661, + "learning_rate": 4.302602414167856e-07, + "loss": 0.5154, + "step": 20682 + }, + { + "epoch": 0.9157544026653974, + "grad_norm": 1.6002418994903564, + "learning_rate": 4.298119063974915e-07, + "loss": 0.5133, + "step": 20683 + }, + { + "epoch": 0.9157986783702113, + "grad_norm": 1.4870293140411377, + "learning_rate": 4.2936379995434873e-07, + "loss": 0.4793, + "step": 20684 + }, + { + "epoch": 0.9158429540750251, + "grad_norm": 1.5388257503509521, + "learning_rate": 4.289159220980599e-07, + "loss": 0.5088, + "step": 20685 + }, + { + "epoch": 0.915887229779839, + "grad_norm": 1.3738576173782349, + "learning_rate": 4.284682728393208e-07, + "loss": 0.4898, + "step": 20686 + }, + { + "epoch": 0.915931505484653, + "grad_norm": 1.4465891122817993, + "learning_rate": 4.280208521888218e-07, + "loss": 0.453, + "step": 20687 + }, + { + "epoch": 0.9159757811894668, + "grad_norm": 1.566743016242981, + "learning_rate": 4.275736601572522e-07, + "loss": 0.5115, + "step": 20688 + }, + { + "epoch": 0.9160200568942807, + "grad_norm": 1.5646381378173828, + "learning_rate": 4.2712669675529226e-07, + "loss": 0.4888, + "step": 20689 + }, + { + "epoch": 0.9160643325990946, + "grad_norm": 2.00596022605896, + "learning_rate": 4.2667996199361573e-07, + "loss": 0.5046, + "step": 20690 + }, + { + "epoch": 0.9161086083039084, + "grad_norm": 1.638543963432312, + "learning_rate": 4.26233455882894e-07, + "loss": 0.4671, + "step": 20691 + }, + { + "epoch": 0.9161528840087223, + "grad_norm": 1.5862056016921997, + "learning_rate": 4.257871784337908e-07, + "loss": 0.5668, + "step": 20692 + }, + { + "epoch": 0.9161971597135362, + "grad_norm": 1.5798804759979248, + "learning_rate": 4.253411296569643e-07, + "loss": 0.4996, + "step": 20693 + }, + { + "epoch": 0.91624143541835, + "grad_norm": 1.5973610877990723, + "learning_rate": 4.2489530956307037e-07, + "loss": 0.5203, + "step": 20694 + }, + { + "epoch": 0.9162857111231639, + "grad_norm": 1.3552882671356201, + "learning_rate": 4.2444971816275495e-07, + "loss": 0.433, + "step": 20695 + }, + { + "epoch": 0.9163299868279778, + "grad_norm": 1.9115478992462158, + "learning_rate": 4.2400435546666174e-07, + "loss": 0.549, + "step": 20696 + }, + { + "epoch": 0.9163742625327916, + "grad_norm": 1.5269402265548706, + "learning_rate": 4.2355922148542563e-07, + "loss": 0.415, + "step": 20697 + }, + { + "epoch": 0.9164185382376056, + "grad_norm": 1.6548720598220825, + "learning_rate": 4.2311431622968137e-07, + "loss": 0.5176, + "step": 20698 + }, + { + "epoch": 0.9164628139424195, + "grad_norm": 1.7080432176589966, + "learning_rate": 4.2266963971005493e-07, + "loss": 0.4847, + "step": 20699 + }, + { + "epoch": 0.9165070896472334, + "grad_norm": 1.7652634382247925, + "learning_rate": 4.2222519193716447e-07, + "loss": 0.4944, + "step": 20700 + }, + { + "epoch": 0.9165513653520472, + "grad_norm": 1.7484110593795776, + "learning_rate": 4.217809729216282e-07, + "loss": 0.5195, + "step": 20701 + }, + { + "epoch": 0.9165956410568611, + "grad_norm": 1.809509038925171, + "learning_rate": 4.213369826740532e-07, + "loss": 0.5357, + "step": 20702 + }, + { + "epoch": 0.916639916761675, + "grad_norm": 1.7082092761993408, + "learning_rate": 4.208932212050454e-07, + "loss": 0.493, + "step": 20703 + }, + { + "epoch": 0.9166841924664888, + "grad_norm": 1.440617561340332, + "learning_rate": 4.2044968852520295e-07, + "loss": 0.4843, + "step": 20704 + }, + { + "epoch": 0.9167284681713027, + "grad_norm": 1.6108500957489014, + "learning_rate": 4.200063846451208e-07, + "loss": 0.4807, + "step": 20705 + }, + { + "epoch": 0.9167727438761166, + "grad_norm": 1.4195116758346558, + "learning_rate": 4.1956330957538595e-07, + "loss": 0.4828, + "step": 20706 + }, + { + "epoch": 0.9168170195809304, + "grad_norm": 1.6433043479919434, + "learning_rate": 4.1912046332657884e-07, + "loss": 0.5266, + "step": 20707 + }, + { + "epoch": 0.9168612952857443, + "grad_norm": 2.425957679748535, + "learning_rate": 4.1867784590928e-07, + "loss": 0.611, + "step": 20708 + }, + { + "epoch": 0.9169055709905582, + "grad_norm": 1.5608186721801758, + "learning_rate": 4.182354573340597e-07, + "loss": 0.4706, + "step": 20709 + }, + { + "epoch": 0.916949846695372, + "grad_norm": 1.5090608596801758, + "learning_rate": 4.17793297611484e-07, + "loss": 0.4511, + "step": 20710 + }, + { + "epoch": 0.916994122400186, + "grad_norm": 1.4873437881469727, + "learning_rate": 4.173513667521123e-07, + "loss": 0.4911, + "step": 20711 + }, + { + "epoch": 0.9170383981049999, + "grad_norm": 1.9417555332183838, + "learning_rate": 4.1690966476650165e-07, + "loss": 0.5787, + "step": 20712 + }, + { + "epoch": 0.9170826738098137, + "grad_norm": 2.2199947834014893, + "learning_rate": 4.1646819166520025e-07, + "loss": 0.5657, + "step": 20713 + }, + { + "epoch": 0.9171269495146276, + "grad_norm": 1.4465234279632568, + "learning_rate": 4.160269474587519e-07, + "loss": 0.4804, + "step": 20714 + }, + { + "epoch": 0.9171712252194415, + "grad_norm": 1.7106622457504272, + "learning_rate": 4.1558593215769714e-07, + "loss": 0.4603, + "step": 20715 + }, + { + "epoch": 0.9172155009242553, + "grad_norm": 1.768060564994812, + "learning_rate": 4.151451457725686e-07, + "loss": 0.5345, + "step": 20716 + }, + { + "epoch": 0.9172597766290692, + "grad_norm": 1.4703449010849, + "learning_rate": 4.1470458831389336e-07, + "loss": 0.4661, + "step": 20717 + }, + { + "epoch": 0.9173040523338831, + "grad_norm": 1.2374998331069946, + "learning_rate": 4.142642597921953e-07, + "loss": 0.4203, + "step": 20718 + }, + { + "epoch": 0.9173483280386969, + "grad_norm": 1.2625573873519897, + "learning_rate": 4.138241602179893e-07, + "loss": 0.4208, + "step": 20719 + }, + { + "epoch": 0.9173926037435108, + "grad_norm": 1.3641046285629272, + "learning_rate": 4.1338428960178923e-07, + "loss": 0.4407, + "step": 20720 + }, + { + "epoch": 0.9174368794483247, + "grad_norm": 1.8135799169540405, + "learning_rate": 4.1294464795409994e-07, + "loss": 0.5065, + "step": 20721 + }, + { + "epoch": 0.9174811551531386, + "grad_norm": 1.6446175575256348, + "learning_rate": 4.1250523528542083e-07, + "loss": 0.4493, + "step": 20722 + }, + { + "epoch": 0.9175254308579525, + "grad_norm": 1.6023306846618652, + "learning_rate": 4.1206605160624803e-07, + "loss": 0.4911, + "step": 20723 + }, + { + "epoch": 0.9175697065627664, + "grad_norm": 2.0636284351348877, + "learning_rate": 4.116270969270686e-07, + "loss": 0.5812, + "step": 20724 + }, + { + "epoch": 0.9176139822675802, + "grad_norm": 1.578896164894104, + "learning_rate": 4.11188371258372e-07, + "loss": 0.523, + "step": 20725 + }, + { + "epoch": 0.9176582579723941, + "grad_norm": 1.312485933303833, + "learning_rate": 4.10749874610632e-07, + "loss": 0.418, + "step": 20726 + }, + { + "epoch": 0.917702533677208, + "grad_norm": 1.323107123374939, + "learning_rate": 4.103116069943225e-07, + "loss": 0.4148, + "step": 20727 + }, + { + "epoch": 0.9177468093820218, + "grad_norm": 1.2637109756469727, + "learning_rate": 4.0987356841991397e-07, + "loss": 0.4487, + "step": 20728 + }, + { + "epoch": 0.9177910850868357, + "grad_norm": 1.4801719188690186, + "learning_rate": 4.09435758897867e-07, + "loss": 0.4948, + "step": 20729 + }, + { + "epoch": 0.9178353607916496, + "grad_norm": 1.533555030822754, + "learning_rate": 4.089981784386365e-07, + "loss": 0.4774, + "step": 20730 + }, + { + "epoch": 0.9178796364964634, + "grad_norm": 1.7103852033615112, + "learning_rate": 4.085608270526753e-07, + "loss": 0.5037, + "step": 20731 + }, + { + "epoch": 0.9179239122012773, + "grad_norm": 1.247774600982666, + "learning_rate": 4.081237047504305e-07, + "loss": 0.4359, + "step": 20732 + }, + { + "epoch": 0.9179681879060912, + "grad_norm": 1.803114414215088, + "learning_rate": 4.0768681154234047e-07, + "loss": 0.5501, + "step": 20733 + }, + { + "epoch": 0.9180124636109052, + "grad_norm": 1.366660475730896, + "learning_rate": 4.0725014743883904e-07, + "loss": 0.4593, + "step": 20734 + }, + { + "epoch": 0.918056739315719, + "grad_norm": 1.2712435722351074, + "learning_rate": 4.0681371245035794e-07, + "loss": 0.4256, + "step": 20735 + }, + { + "epoch": 0.9181010150205329, + "grad_norm": 1.653463363647461, + "learning_rate": 4.06377506587321e-07, + "loss": 0.5196, + "step": 20736 + }, + { + "epoch": 0.9181452907253468, + "grad_norm": 1.3819758892059326, + "learning_rate": 4.0594152986014436e-07, + "loss": 0.4412, + "step": 20737 + }, + { + "epoch": 0.9181895664301606, + "grad_norm": 1.7736026048660278, + "learning_rate": 4.055057822792452e-07, + "loss": 0.4674, + "step": 20738 + }, + { + "epoch": 0.9182338421349745, + "grad_norm": 1.6616450548171997, + "learning_rate": 4.0507026385502747e-07, + "loss": 0.5209, + "step": 20739 + }, + { + "epoch": 0.9182781178397884, + "grad_norm": 1.7225524187088013, + "learning_rate": 4.046349745978917e-07, + "loss": 0.494, + "step": 20740 + }, + { + "epoch": 0.9183223935446022, + "grad_norm": 1.3732006549835205, + "learning_rate": 4.0419991451823846e-07, + "loss": 0.4852, + "step": 20741 + }, + { + "epoch": 0.9183666692494161, + "grad_norm": 1.4991607666015625, + "learning_rate": 4.037650836264573e-07, + "loss": 0.5259, + "step": 20742 + }, + { + "epoch": 0.91841094495423, + "grad_norm": 1.3495399951934814, + "learning_rate": 4.033304819329342e-07, + "loss": 0.4435, + "step": 20743 + }, + { + "epoch": 0.9184552206590438, + "grad_norm": 1.4437532424926758, + "learning_rate": 4.0289610944804656e-07, + "loss": 0.4779, + "step": 20744 + }, + { + "epoch": 0.9184994963638577, + "grad_norm": 1.5228365659713745, + "learning_rate": 4.024619661821716e-07, + "loss": 0.4918, + "step": 20745 + }, + { + "epoch": 0.9185437720686717, + "grad_norm": 1.884518027305603, + "learning_rate": 4.0202805214567874e-07, + "loss": 0.497, + "step": 20746 + }, + { + "epoch": 0.9185880477734855, + "grad_norm": 1.742641568183899, + "learning_rate": 4.015943673489309e-07, + "loss": 0.5202, + "step": 20747 + }, + { + "epoch": 0.9186323234782994, + "grad_norm": 1.4948339462280273, + "learning_rate": 4.011609118022863e-07, + "loss": 0.4703, + "step": 20748 + }, + { + "epoch": 0.9186765991831133, + "grad_norm": 1.3000754117965698, + "learning_rate": 4.0072768551609795e-07, + "loss": 0.4526, + "step": 20749 + }, + { + "epoch": 0.9187208748879271, + "grad_norm": 1.3163286447525024, + "learning_rate": 4.0029468850071194e-07, + "loss": 0.4408, + "step": 20750 + }, + { + "epoch": 0.918765150592741, + "grad_norm": 1.4060077667236328, + "learning_rate": 3.998619207664711e-07, + "loss": 0.4815, + "step": 20751 + }, + { + "epoch": 0.9188094262975549, + "grad_norm": 1.7625399827957153, + "learning_rate": 3.9942938232371165e-07, + "loss": 0.5256, + "step": 20752 + }, + { + "epoch": 0.9188537020023687, + "grad_norm": 1.1821579933166504, + "learning_rate": 3.9899707318276523e-07, + "loss": 0.3864, + "step": 20753 + }, + { + "epoch": 0.9188979777071826, + "grad_norm": 2.3843586444854736, + "learning_rate": 3.9856499335395373e-07, + "loss": 0.564, + "step": 20754 + }, + { + "epoch": 0.9189422534119965, + "grad_norm": 1.6964296102523804, + "learning_rate": 3.981331428476021e-07, + "loss": 0.5503, + "step": 20755 + }, + { + "epoch": 0.9189865291168103, + "grad_norm": 1.3434501886367798, + "learning_rate": 3.977015216740221e-07, + "loss": 0.486, + "step": 20756 + }, + { + "epoch": 0.9190308048216242, + "grad_norm": 1.6496189832687378, + "learning_rate": 3.972701298435222e-07, + "loss": 0.4775, + "step": 20757 + }, + { + "epoch": 0.9190750805264382, + "grad_norm": 2.1874876022338867, + "learning_rate": 3.9683896736640635e-07, + "loss": 0.5659, + "step": 20758 + }, + { + "epoch": 0.919119356231252, + "grad_norm": 1.5331124067306519, + "learning_rate": 3.96408034252973e-07, + "loss": 0.4896, + "step": 20759 + }, + { + "epoch": 0.9191636319360659, + "grad_norm": 1.519852638244629, + "learning_rate": 3.9597733051351286e-07, + "loss": 0.5148, + "step": 20760 + }, + { + "epoch": 0.9192079076408798, + "grad_norm": 1.4175622463226318, + "learning_rate": 3.955468561583153e-07, + "loss": 0.4423, + "step": 20761 + }, + { + "epoch": 0.9192521833456936, + "grad_norm": 1.6176459789276123, + "learning_rate": 3.951166111976612e-07, + "loss": 0.4956, + "step": 20762 + }, + { + "epoch": 0.9192964590505075, + "grad_norm": 1.4109220504760742, + "learning_rate": 3.9468659564182554e-07, + "loss": 0.4529, + "step": 20763 + }, + { + "epoch": 0.9193407347553214, + "grad_norm": 1.536259412765503, + "learning_rate": 3.94256809501079e-07, + "loss": 0.4638, + "step": 20764 + }, + { + "epoch": 0.9193850104601352, + "grad_norm": 1.8330105543136597, + "learning_rate": 3.938272527856879e-07, + "loss": 0.5067, + "step": 20765 + }, + { + "epoch": 0.9194292861649491, + "grad_norm": 1.5063310861587524, + "learning_rate": 3.933979255059117e-07, + "loss": 0.4903, + "step": 20766 + }, + { + "epoch": 0.919473561869763, + "grad_norm": 1.489691972732544, + "learning_rate": 3.929688276720045e-07, + "loss": 0.4707, + "step": 20767 + }, + { + "epoch": 0.9195178375745768, + "grad_norm": 1.3941802978515625, + "learning_rate": 3.9253995929421473e-07, + "loss": 0.4869, + "step": 20768 + }, + { + "epoch": 0.9195621132793907, + "grad_norm": 1.5644593238830566, + "learning_rate": 3.921113203827842e-07, + "loss": 0.5002, + "step": 20769 + }, + { + "epoch": 0.9196063889842047, + "grad_norm": 2.2744157314300537, + "learning_rate": 3.916829109479514e-07, + "loss": 0.5589, + "step": 20770 + }, + { + "epoch": 0.9196506646890186, + "grad_norm": 1.4820630550384521, + "learning_rate": 3.9125473099994925e-07, + "loss": 0.4472, + "step": 20771 + }, + { + "epoch": 0.9196949403938324, + "grad_norm": 1.4361361265182495, + "learning_rate": 3.908267805490051e-07, + "loss": 0.4591, + "step": 20772 + }, + { + "epoch": 0.9197392160986463, + "grad_norm": 1.2942346334457397, + "learning_rate": 3.9039905960533864e-07, + "loss": 0.439, + "step": 20773 + }, + { + "epoch": 0.9197834918034602, + "grad_norm": 2.11655330657959, + "learning_rate": 3.8997156817916495e-07, + "loss": 0.4776, + "step": 20774 + }, + { + "epoch": 0.919827767508274, + "grad_norm": 1.287751317024231, + "learning_rate": 3.8954430628069696e-07, + "loss": 0.3992, + "step": 20775 + }, + { + "epoch": 0.9198720432130879, + "grad_norm": 1.7356712818145752, + "learning_rate": 3.8911727392013765e-07, + "loss": 0.4829, + "step": 20776 + }, + { + "epoch": 0.9199163189179018, + "grad_norm": 1.5371919870376587, + "learning_rate": 3.886904711076878e-07, + "loss": 0.4947, + "step": 20777 + }, + { + "epoch": 0.9199605946227156, + "grad_norm": 1.4002268314361572, + "learning_rate": 3.882638978535391e-07, + "loss": 0.4216, + "step": 20778 + }, + { + "epoch": 0.9200048703275295, + "grad_norm": 1.6823012828826904, + "learning_rate": 3.878375541678825e-07, + "loss": 0.5634, + "step": 20779 + }, + { + "epoch": 0.9200491460323434, + "grad_norm": 2.6548702716827393, + "learning_rate": 3.8741144006089636e-07, + "loss": 0.6053, + "step": 20780 + }, + { + "epoch": 0.9200934217371572, + "grad_norm": 2.0235936641693115, + "learning_rate": 3.8698555554276375e-07, + "loss": 0.4974, + "step": 20781 + }, + { + "epoch": 0.9201376974419712, + "grad_norm": 1.8225857019424438, + "learning_rate": 3.865599006236531e-07, + "loss": 0.5306, + "step": 20782 + }, + { + "epoch": 0.9201819731467851, + "grad_norm": 1.670565128326416, + "learning_rate": 3.8613447531373196e-07, + "loss": 0.5334, + "step": 20783 + }, + { + "epoch": 0.9202262488515989, + "grad_norm": 1.6245229244232178, + "learning_rate": 3.8570927962315987e-07, + "loss": 0.4803, + "step": 20784 + }, + { + "epoch": 0.9202705245564128, + "grad_norm": 1.5960975885391235, + "learning_rate": 3.852843135620965e-07, + "loss": 0.4993, + "step": 20785 + }, + { + "epoch": 0.9203148002612267, + "grad_norm": 1.6572977304458618, + "learning_rate": 3.84859577140686e-07, + "loss": 0.4858, + "step": 20786 + }, + { + "epoch": 0.9203590759660405, + "grad_norm": 1.378178358078003, + "learning_rate": 3.844350703690758e-07, + "loss": 0.4008, + "step": 20787 + }, + { + "epoch": 0.9204033516708544, + "grad_norm": 1.817026972770691, + "learning_rate": 3.8401079325740445e-07, + "loss": 0.5436, + "step": 20788 + }, + { + "epoch": 0.9204476273756683, + "grad_norm": 2.090620994567871, + "learning_rate": 3.8358674581580714e-07, + "loss": 0.4486, + "step": 20789 + }, + { + "epoch": 0.9204919030804821, + "grad_norm": 1.5374737977981567, + "learning_rate": 3.8316292805440803e-07, + "loss": 0.4694, + "step": 20790 + }, + { + "epoch": 0.920536178785296, + "grad_norm": 1.4995020627975464, + "learning_rate": 3.8273933998333345e-07, + "loss": 0.4486, + "step": 20791 + }, + { + "epoch": 0.92058045449011, + "grad_norm": 2.6480159759521484, + "learning_rate": 3.823159816126998e-07, + "loss": 0.594, + "step": 20792 + }, + { + "epoch": 0.9206247301949237, + "grad_norm": 1.386462926864624, + "learning_rate": 3.818928529526178e-07, + "loss": 0.4804, + "step": 20793 + }, + { + "epoch": 0.9206690058997377, + "grad_norm": 1.6154067516326904, + "learning_rate": 3.814699540131928e-07, + "loss": 0.4748, + "step": 20794 + }, + { + "epoch": 0.9207132816045516, + "grad_norm": 1.5714467763900757, + "learning_rate": 3.810472848045266e-07, + "loss": 0.5505, + "step": 20795 + }, + { + "epoch": 0.9207575573093654, + "grad_norm": 1.7488387823104858, + "learning_rate": 3.806248453367145e-07, + "loss": 0.462, + "step": 20796 + }, + { + "epoch": 0.9208018330141793, + "grad_norm": 2.4209043979644775, + "learning_rate": 3.8020263561984405e-07, + "loss": 0.549, + "step": 20797 + }, + { + "epoch": 0.9208461087189932, + "grad_norm": 1.5289326906204224, + "learning_rate": 3.797806556640016e-07, + "loss": 0.5025, + "step": 20798 + }, + { + "epoch": 0.920890384423807, + "grad_norm": 1.5992456674575806, + "learning_rate": 3.7935890547926567e-07, + "loss": 0.4843, + "step": 20799 + }, + { + "epoch": 0.9209346601286209, + "grad_norm": 1.9562612771987915, + "learning_rate": 3.789373850757072e-07, + "loss": 0.5428, + "step": 20800 + }, + { + "epoch": 0.9209789358334348, + "grad_norm": 2.10864520072937, + "learning_rate": 3.785160944633981e-07, + "loss": 0.5154, + "step": 20801 + }, + { + "epoch": 0.9210232115382486, + "grad_norm": 1.4929862022399902, + "learning_rate": 3.7809503365239696e-07, + "loss": 0.4489, + "step": 20802 + }, + { + "epoch": 0.9210674872430625, + "grad_norm": 1.3878172636032104, + "learning_rate": 3.7767420265276135e-07, + "loss": 0.4509, + "step": 20803 + }, + { + "epoch": 0.9211117629478764, + "grad_norm": 1.7255818843841553, + "learning_rate": 3.7725360147454206e-07, + "loss": 0.4519, + "step": 20804 + }, + { + "epoch": 0.9211560386526904, + "grad_norm": 1.5626534223556519, + "learning_rate": 3.768332301277866e-07, + "loss": 0.4514, + "step": 20805 + }, + { + "epoch": 0.9212003143575042, + "grad_norm": 1.80208420753479, + "learning_rate": 3.764130886225337e-07, + "loss": 0.5235, + "step": 20806 + }, + { + "epoch": 0.9212445900623181, + "grad_norm": 1.6119251251220703, + "learning_rate": 3.7599317696881633e-07, + "loss": 0.4922, + "step": 20807 + }, + { + "epoch": 0.921288865767132, + "grad_norm": 1.629288911819458, + "learning_rate": 3.755734951766676e-07, + "loss": 0.5211, + "step": 20808 + }, + { + "epoch": 0.9213331414719458, + "grad_norm": 2.3518786430358887, + "learning_rate": 3.751540432561096e-07, + "loss": 0.4597, + "step": 20809 + }, + { + "epoch": 0.9213774171767597, + "grad_norm": 1.8101598024368286, + "learning_rate": 3.7473482121715864e-07, + "loss": 0.4919, + "step": 20810 + }, + { + "epoch": 0.9214216928815736, + "grad_norm": 1.8106744289398193, + "learning_rate": 3.7431582906983124e-07, + "loss": 0.5444, + "step": 20811 + }, + { + "epoch": 0.9214659685863874, + "grad_norm": 1.583392858505249, + "learning_rate": 3.7389706682413265e-07, + "loss": 0.4871, + "step": 20812 + }, + { + "epoch": 0.9215102442912013, + "grad_norm": 1.5315603017807007, + "learning_rate": 3.734785344900649e-07, + "loss": 0.4981, + "step": 20813 + }, + { + "epoch": 0.9215545199960152, + "grad_norm": 1.6776494979858398, + "learning_rate": 3.7306023207762334e-07, + "loss": 0.4847, + "step": 20814 + }, + { + "epoch": 0.921598795700829, + "grad_norm": 1.635457158088684, + "learning_rate": 3.7264215959679996e-07, + "loss": 0.5035, + "step": 20815 + }, + { + "epoch": 0.9216430714056429, + "grad_norm": 1.52423095703125, + "learning_rate": 3.722243170575801e-07, + "loss": 0.4514, + "step": 20816 + }, + { + "epoch": 0.9216873471104569, + "grad_norm": 1.714874505996704, + "learning_rate": 3.718067044699414e-07, + "loss": 0.4808, + "step": 20817 + }, + { + "epoch": 0.9217316228152707, + "grad_norm": 1.3570855855941772, + "learning_rate": 3.7138932184386136e-07, + "loss": 0.4571, + "step": 20818 + }, + { + "epoch": 0.9217758985200846, + "grad_norm": 1.3559657335281372, + "learning_rate": 3.7097216918930755e-07, + "loss": 0.5203, + "step": 20819 + }, + { + "epoch": 0.9218201742248985, + "grad_norm": 1.414801001548767, + "learning_rate": 3.705552465162432e-07, + "loss": 0.4419, + "step": 20820 + }, + { + "epoch": 0.9218644499297123, + "grad_norm": 1.460814356803894, + "learning_rate": 3.701385538346247e-07, + "loss": 0.5056, + "step": 20821 + }, + { + "epoch": 0.9219087256345262, + "grad_norm": 1.2000946998596191, + "learning_rate": 3.697220911544075e-07, + "loss": 0.4756, + "step": 20822 + }, + { + "epoch": 0.9219530013393401, + "grad_norm": 2.007504463195801, + "learning_rate": 3.693058584855369e-07, + "loss": 0.4938, + "step": 20823 + }, + { + "epoch": 0.9219972770441539, + "grad_norm": 1.4326063394546509, + "learning_rate": 3.68889855837955e-07, + "loss": 0.4933, + "step": 20824 + }, + { + "epoch": 0.9220415527489678, + "grad_norm": 1.4103593826293945, + "learning_rate": 3.6847408322159605e-07, + "loss": 0.4699, + "step": 20825 + }, + { + "epoch": 0.9220858284537817, + "grad_norm": 1.6445468664169312, + "learning_rate": 3.6805854064639104e-07, + "loss": 0.4621, + "step": 20826 + }, + { + "epoch": 0.9221301041585955, + "grad_norm": 1.9201233386993408, + "learning_rate": 3.6764322812226416e-07, + "loss": 0.5249, + "step": 20827 + }, + { + "epoch": 0.9221743798634094, + "grad_norm": 1.275518774986267, + "learning_rate": 3.6722814565913757e-07, + "loss": 0.4499, + "step": 20828 + }, + { + "epoch": 0.9222186555682234, + "grad_norm": 1.5008654594421387, + "learning_rate": 3.668132932669233e-07, + "loss": 0.4686, + "step": 20829 + }, + { + "epoch": 0.9222629312730372, + "grad_norm": 1.88189697265625, + "learning_rate": 3.6639867095552897e-07, + "loss": 0.4966, + "step": 20830 + }, + { + "epoch": 0.9223072069778511, + "grad_norm": 1.4950299263000488, + "learning_rate": 3.6598427873485884e-07, + "loss": 0.4647, + "step": 20831 + }, + { + "epoch": 0.922351482682665, + "grad_norm": 1.7837576866149902, + "learning_rate": 3.6557011661480956e-07, + "loss": 0.5206, + "step": 20832 + }, + { + "epoch": 0.9223957583874788, + "grad_norm": 1.7962722778320312, + "learning_rate": 3.6515618460527534e-07, + "loss": 0.5192, + "step": 20833 + }, + { + "epoch": 0.9224400340922927, + "grad_norm": 1.7835801839828491, + "learning_rate": 3.647424827161383e-07, + "loss": 0.5613, + "step": 20834 + }, + { + "epoch": 0.9224843097971066, + "grad_norm": 2.357998847961426, + "learning_rate": 3.643290109572828e-07, + "loss": 0.4992, + "step": 20835 + }, + { + "epoch": 0.9225285855019204, + "grad_norm": 1.5695735216140747, + "learning_rate": 3.639157693385831e-07, + "loss": 0.4605, + "step": 20836 + }, + { + "epoch": 0.9225728612067343, + "grad_norm": 2.060960054397583, + "learning_rate": 3.6350275786990796e-07, + "loss": 0.4782, + "step": 20837 + }, + { + "epoch": 0.9226171369115482, + "grad_norm": 1.875453233718872, + "learning_rate": 3.630899765611251e-07, + "loss": 0.525, + "step": 20838 + }, + { + "epoch": 0.9226614126163621, + "grad_norm": 1.2602204084396362, + "learning_rate": 3.6267742542209107e-07, + "loss": 0.4456, + "step": 20839 + }, + { + "epoch": 0.9227056883211759, + "grad_norm": 1.8769136667251587, + "learning_rate": 3.6226510446265903e-07, + "loss": 0.5216, + "step": 20840 + }, + { + "epoch": 0.9227499640259899, + "grad_norm": 1.4274420738220215, + "learning_rate": 3.61853013692679e-07, + "loss": 0.4443, + "step": 20841 + }, + { + "epoch": 0.9227942397308038, + "grad_norm": 1.4334044456481934, + "learning_rate": 3.614411531219908e-07, + "loss": 0.4894, + "step": 20842 + }, + { + "epoch": 0.9228385154356176, + "grad_norm": 1.5098693370819092, + "learning_rate": 3.6102952276043326e-07, + "loss": 0.4585, + "step": 20843 + }, + { + "epoch": 0.9228827911404315, + "grad_norm": 1.740744709968567, + "learning_rate": 3.6061812261783623e-07, + "loss": 0.5157, + "step": 20844 + }, + { + "epoch": 0.9229270668452454, + "grad_norm": 1.4714136123657227, + "learning_rate": 3.602069527040275e-07, + "loss": 0.4307, + "step": 20845 + }, + { + "epoch": 0.9229713425500592, + "grad_norm": 2.389049530029297, + "learning_rate": 3.597960130288258e-07, + "loss": 0.5648, + "step": 20846 + }, + { + "epoch": 0.9230156182548731, + "grad_norm": 1.2553917169570923, + "learning_rate": 3.593853036020467e-07, + "loss": 0.41, + "step": 20847 + }, + { + "epoch": 0.923059893959687, + "grad_norm": 1.4992998838424683, + "learning_rate": 3.5897482443350116e-07, + "loss": 0.474, + "step": 20848 + }, + { + "epoch": 0.9231041696645008, + "grad_norm": 1.78413724899292, + "learning_rate": 3.585645755329925e-07, + "loss": 0.5349, + "step": 20849 + }, + { + "epoch": 0.9231484453693147, + "grad_norm": 1.5496524572372437, + "learning_rate": 3.581545569103184e-07, + "loss": 0.4875, + "step": 20850 + }, + { + "epoch": 0.9231927210741286, + "grad_norm": 1.1849803924560547, + "learning_rate": 3.5774476857527107e-07, + "loss": 0.4254, + "step": 20851 + }, + { + "epoch": 0.9232369967789424, + "grad_norm": 1.4403305053710938, + "learning_rate": 3.573352105376393e-07, + "loss": 0.5026, + "step": 20852 + }, + { + "epoch": 0.9232812724837564, + "grad_norm": 1.3042994737625122, + "learning_rate": 3.569258828072053e-07, + "loss": 0.4151, + "step": 20853 + }, + { + "epoch": 0.9233255481885703, + "grad_norm": 1.3678635358810425, + "learning_rate": 3.565167853937435e-07, + "loss": 0.4166, + "step": 20854 + }, + { + "epoch": 0.9233698238933841, + "grad_norm": 1.7748429775238037, + "learning_rate": 3.5610791830702705e-07, + "loss": 0.5486, + "step": 20855 + }, + { + "epoch": 0.923414099598198, + "grad_norm": 1.2740178108215332, + "learning_rate": 3.5569928155682053e-07, + "loss": 0.4039, + "step": 20856 + }, + { + "epoch": 0.9234583753030119, + "grad_norm": 1.603602409362793, + "learning_rate": 3.552908751528827e-07, + "loss": 0.5531, + "step": 20857 + }, + { + "epoch": 0.9235026510078257, + "grad_norm": 1.4599992036819458, + "learning_rate": 3.548826991049714e-07, + "loss": 0.4575, + "step": 20858 + }, + { + "epoch": 0.9235469267126396, + "grad_norm": 1.473862648010254, + "learning_rate": 3.5447475342283324e-07, + "loss": 0.4544, + "step": 20859 + }, + { + "epoch": 0.9235912024174535, + "grad_norm": 1.7954816818237305, + "learning_rate": 3.5406703811621143e-07, + "loss": 0.5456, + "step": 20860 + }, + { + "epoch": 0.9236354781222673, + "grad_norm": 1.551680088043213, + "learning_rate": 3.53659553194845e-07, + "loss": 0.476, + "step": 20861 + }, + { + "epoch": 0.9236797538270812, + "grad_norm": 1.381127119064331, + "learning_rate": 3.53252298668465e-07, + "loss": 0.4184, + "step": 20862 + }, + { + "epoch": 0.9237240295318951, + "grad_norm": 1.4945112466812134, + "learning_rate": 3.528452745467992e-07, + "loss": 0.462, + "step": 20863 + }, + { + "epoch": 0.9237683052367089, + "grad_norm": 1.6238598823547363, + "learning_rate": 3.5243848083956867e-07, + "loss": 0.469, + "step": 20864 + }, + { + "epoch": 0.9238125809415229, + "grad_norm": 1.6072180271148682, + "learning_rate": 3.5203191755649014e-07, + "loss": 0.5022, + "step": 20865 + }, + { + "epoch": 0.9238568566463368, + "grad_norm": 1.7523409128189087, + "learning_rate": 3.516255847072747e-07, + "loss": 0.5459, + "step": 20866 + }, + { + "epoch": 0.9239011323511506, + "grad_norm": 1.3141875267028809, + "learning_rate": 3.5121948230162353e-07, + "loss": 0.4339, + "step": 20867 + }, + { + "epoch": 0.9239454080559645, + "grad_norm": 1.5498689413070679, + "learning_rate": 3.5081361034924097e-07, + "loss": 0.489, + "step": 20868 + }, + { + "epoch": 0.9239896837607784, + "grad_norm": 1.3748104572296143, + "learning_rate": 3.5040796885981834e-07, + "loss": 0.4248, + "step": 20869 + }, + { + "epoch": 0.9240339594655922, + "grad_norm": 1.7219691276550293, + "learning_rate": 3.5000255784304326e-07, + "loss": 0.5335, + "step": 20870 + }, + { + "epoch": 0.9240782351704061, + "grad_norm": 1.6279528141021729, + "learning_rate": 3.4959737730860144e-07, + "loss": 0.5119, + "step": 20871 + }, + { + "epoch": 0.92412251087522, + "grad_norm": 1.9556219577789307, + "learning_rate": 3.4919242726616733e-07, + "loss": 0.489, + "step": 20872 + }, + { + "epoch": 0.9241667865800338, + "grad_norm": 1.3190354108810425, + "learning_rate": 3.4878770772541536e-07, + "loss": 0.4502, + "step": 20873 + }, + { + "epoch": 0.9242110622848477, + "grad_norm": 1.4543229341506958, + "learning_rate": 3.48383218696009e-07, + "loss": 0.4688, + "step": 20874 + }, + { + "epoch": 0.9242553379896616, + "grad_norm": 1.500770926475525, + "learning_rate": 3.4797896018761156e-07, + "loss": 0.4721, + "step": 20875 + }, + { + "epoch": 0.9242996136944756, + "grad_norm": 1.4257217645645142, + "learning_rate": 3.475749322098776e-07, + "loss": 0.4539, + "step": 20876 + }, + { + "epoch": 0.9243438893992894, + "grad_norm": 1.5425375699996948, + "learning_rate": 3.4717113477245714e-07, + "loss": 0.5044, + "step": 20877 + }, + { + "epoch": 0.9243881651041033, + "grad_norm": 1.6977440118789673, + "learning_rate": 3.467675678849947e-07, + "loss": 0.4285, + "step": 20878 + }, + { + "epoch": 0.9244324408089172, + "grad_norm": 1.405321717262268, + "learning_rate": 3.463642315571292e-07, + "loss": 0.5018, + "step": 20879 + }, + { + "epoch": 0.924476716513731, + "grad_norm": 1.8247491121292114, + "learning_rate": 3.4596112579849406e-07, + "loss": 0.5285, + "step": 20880 + }, + { + "epoch": 0.9245209922185449, + "grad_norm": 1.6321005821228027, + "learning_rate": 3.4555825061871605e-07, + "loss": 0.5159, + "step": 20881 + }, + { + "epoch": 0.9245652679233588, + "grad_norm": 1.4419151544570923, + "learning_rate": 3.451556060274186e-07, + "loss": 0.5115, + "step": 20882 + }, + { + "epoch": 0.9246095436281726, + "grad_norm": 1.428449034690857, + "learning_rate": 3.447531920342173e-07, + "loss": 0.4313, + "step": 20883 + }, + { + "epoch": 0.9246538193329865, + "grad_norm": 1.8244004249572754, + "learning_rate": 3.443510086487245e-07, + "loss": 0.4626, + "step": 20884 + }, + { + "epoch": 0.9246980950378004, + "grad_norm": 1.5259156227111816, + "learning_rate": 3.4394905588054694e-07, + "loss": 0.5263, + "step": 20885 + }, + { + "epoch": 0.9247423707426142, + "grad_norm": 1.6113102436065674, + "learning_rate": 3.435473337392836e-07, + "loss": 0.5295, + "step": 20886 + }, + { + "epoch": 0.9247866464474281, + "grad_norm": 1.7509608268737793, + "learning_rate": 3.431458422345291e-07, + "loss": 0.5099, + "step": 20887 + }, + { + "epoch": 0.924830922152242, + "grad_norm": 1.293251633644104, + "learning_rate": 3.4274458137587453e-07, + "loss": 0.4255, + "step": 20888 + }, + { + "epoch": 0.9248751978570559, + "grad_norm": 1.450827717781067, + "learning_rate": 3.423435511729023e-07, + "loss": 0.4512, + "step": 20889 + }, + { + "epoch": 0.9249194735618698, + "grad_norm": 1.8637189865112305, + "learning_rate": 3.419427516351903e-07, + "loss": 0.5231, + "step": 20890 + }, + { + "epoch": 0.9249637492666837, + "grad_norm": 1.3429216146469116, + "learning_rate": 3.415421827723109e-07, + "loss": 0.4912, + "step": 20891 + }, + { + "epoch": 0.9250080249714975, + "grad_norm": 1.3851782083511353, + "learning_rate": 3.4114184459383306e-07, + "loss": 0.4592, + "step": 20892 + }, + { + "epoch": 0.9250523006763114, + "grad_norm": 1.4374608993530273, + "learning_rate": 3.4074173710931804e-07, + "loss": 0.4379, + "step": 20893 + }, + { + "epoch": 0.9250965763811253, + "grad_norm": 1.640383005142212, + "learning_rate": 3.403418603283204e-07, + "loss": 0.5027, + "step": 20894 + }, + { + "epoch": 0.9251408520859391, + "grad_norm": 1.4370688199996948, + "learning_rate": 3.3994221426039365e-07, + "loss": 0.5375, + "step": 20895 + }, + { + "epoch": 0.925185127790753, + "grad_norm": 1.7133762836456299, + "learning_rate": 3.3954279891508126e-07, + "loss": 0.4426, + "step": 20896 + }, + { + "epoch": 0.9252294034955669, + "grad_norm": 1.6196800470352173, + "learning_rate": 3.3914361430192334e-07, + "loss": 0.457, + "step": 20897 + }, + { + "epoch": 0.9252736792003807, + "grad_norm": 1.6955541372299194, + "learning_rate": 3.387446604304534e-07, + "loss": 0.5194, + "step": 20898 + }, + { + "epoch": 0.9253179549051946, + "grad_norm": 1.5468019247055054, + "learning_rate": 3.383459373102016e-07, + "loss": 0.5075, + "step": 20899 + }, + { + "epoch": 0.9253622306100086, + "grad_norm": 1.9180207252502441, + "learning_rate": 3.379474449506903e-07, + "loss": 0.461, + "step": 20900 + }, + { + "epoch": 0.9254065063148224, + "grad_norm": 2.0639116764068604, + "learning_rate": 3.3754918336143637e-07, + "loss": 0.5997, + "step": 20901 + }, + { + "epoch": 0.9254507820196363, + "grad_norm": 1.6191736459732056, + "learning_rate": 3.371511525519533e-07, + "loss": 0.4873, + "step": 20902 + }, + { + "epoch": 0.9254950577244502, + "grad_norm": 1.4160218238830566, + "learning_rate": 3.367533525317468e-07, + "loss": 0.5194, + "step": 20903 + }, + { + "epoch": 0.925539333429264, + "grad_norm": 1.5400216579437256, + "learning_rate": 3.363557833103182e-07, + "loss": 0.4935, + "step": 20904 + }, + { + "epoch": 0.9255836091340779, + "grad_norm": 1.472089171409607, + "learning_rate": 3.3595844489716313e-07, + "loss": 0.4612, + "step": 20905 + }, + { + "epoch": 0.9256278848388918, + "grad_norm": 1.5121961832046509, + "learning_rate": 3.3556133730177296e-07, + "loss": 0.4904, + "step": 20906 + }, + { + "epoch": 0.9256721605437056, + "grad_norm": 1.4162956476211548, + "learning_rate": 3.3516446053363015e-07, + "loss": 0.5056, + "step": 20907 + }, + { + "epoch": 0.9257164362485195, + "grad_norm": 1.6810811758041382, + "learning_rate": 3.3476781460221595e-07, + "loss": 0.5332, + "step": 20908 + }, + { + "epoch": 0.9257607119533334, + "grad_norm": 1.4457244873046875, + "learning_rate": 3.343713995170028e-07, + "loss": 0.4975, + "step": 20909 + }, + { + "epoch": 0.9258049876581473, + "grad_norm": 1.8805325031280518, + "learning_rate": 3.3397521528745755e-07, + "loss": 0.5276, + "step": 20910 + }, + { + "epoch": 0.9258492633629611, + "grad_norm": 1.7350915670394897, + "learning_rate": 3.335792619230438e-07, + "loss": 0.5645, + "step": 20911 + }, + { + "epoch": 0.925893539067775, + "grad_norm": 1.397493839263916, + "learning_rate": 3.3318353943321944e-07, + "loss": 0.4344, + "step": 20912 + }, + { + "epoch": 0.925937814772589, + "grad_norm": 1.5822758674621582, + "learning_rate": 3.327880478274348e-07, + "loss": 0.4946, + "step": 20913 + }, + { + "epoch": 0.9259820904774028, + "grad_norm": 1.5437356233596802, + "learning_rate": 3.3239278711513557e-07, + "loss": 0.5029, + "step": 20914 + }, + { + "epoch": 0.9260263661822167, + "grad_norm": 1.4769346714019775, + "learning_rate": 3.319977573057642e-07, + "loss": 0.4303, + "step": 20915 + }, + { + "epoch": 0.9260706418870306, + "grad_norm": 1.6094088554382324, + "learning_rate": 3.3160295840875435e-07, + "loss": 0.472, + "step": 20916 + }, + { + "epoch": 0.9261149175918444, + "grad_norm": 1.5940302610397339, + "learning_rate": 3.31208390433535e-07, + "loss": 0.4831, + "step": 20917 + }, + { + "epoch": 0.9261591932966583, + "grad_norm": 1.4402079582214355, + "learning_rate": 3.308140533895321e-07, + "loss": 0.4406, + "step": 20918 + }, + { + "epoch": 0.9262034690014722, + "grad_norm": 1.5591596364974976, + "learning_rate": 3.3041994728616133e-07, + "loss": 0.4546, + "step": 20919 + }, + { + "epoch": 0.926247744706286, + "grad_norm": 1.4406507015228271, + "learning_rate": 3.300260721328374e-07, + "loss": 0.4511, + "step": 20920 + }, + { + "epoch": 0.9262920204110999, + "grad_norm": 1.3665153980255127, + "learning_rate": 3.296324279389651e-07, + "loss": 0.4495, + "step": 20921 + }, + { + "epoch": 0.9263362961159138, + "grad_norm": 1.5296235084533691, + "learning_rate": 3.292390147139501e-07, + "loss": 0.4835, + "step": 20922 + }, + { + "epoch": 0.9263805718207276, + "grad_norm": 1.2875440120697021, + "learning_rate": 3.288458324671873e-07, + "loss": 0.4183, + "step": 20923 + }, + { + "epoch": 0.9264248475255416, + "grad_norm": 1.6219581365585327, + "learning_rate": 3.284528812080656e-07, + "loss": 0.5325, + "step": 20924 + }, + { + "epoch": 0.9264691232303555, + "grad_norm": 1.2755260467529297, + "learning_rate": 3.280601609459744e-07, + "loss": 0.4413, + "step": 20925 + }, + { + "epoch": 0.9265133989351693, + "grad_norm": 1.4026713371276855, + "learning_rate": 3.2766767169028933e-07, + "loss": 0.431, + "step": 20926 + }, + { + "epoch": 0.9265576746399832, + "grad_norm": 1.626852035522461, + "learning_rate": 3.2727541345038747e-07, + "loss": 0.4796, + "step": 20927 + }, + { + "epoch": 0.9266019503447971, + "grad_norm": 1.3052550554275513, + "learning_rate": 3.2688338623563575e-07, + "loss": 0.4601, + "step": 20928 + }, + { + "epoch": 0.9266462260496109, + "grad_norm": 1.643822193145752, + "learning_rate": 3.2649159005539887e-07, + "loss": 0.4783, + "step": 20929 + }, + { + "epoch": 0.9266905017544248, + "grad_norm": 1.5667507648468018, + "learning_rate": 3.261000249190338e-07, + "loss": 0.4691, + "step": 20930 + }, + { + "epoch": 0.9267347774592387, + "grad_norm": 1.4532911777496338, + "learning_rate": 3.2570869083589196e-07, + "loss": 0.5209, + "step": 20931 + }, + { + "epoch": 0.9267790531640525, + "grad_norm": 1.3981101512908936, + "learning_rate": 3.253175878153214e-07, + "loss": 0.4603, + "step": 20932 + }, + { + "epoch": 0.9268233288688664, + "grad_norm": 1.4553210735321045, + "learning_rate": 3.2492671586666355e-07, + "loss": 0.4986, + "step": 20933 + }, + { + "epoch": 0.9268676045736803, + "grad_norm": 2.0321645736694336, + "learning_rate": 3.24536074999251e-07, + "loss": 0.538, + "step": 20934 + }, + { + "epoch": 0.9269118802784941, + "grad_norm": 1.3024134635925293, + "learning_rate": 3.241456652224184e-07, + "loss": 0.4492, + "step": 20935 + }, + { + "epoch": 0.926956155983308, + "grad_norm": 1.3164399862289429, + "learning_rate": 3.237554865454895e-07, + "loss": 0.4615, + "step": 20936 + }, + { + "epoch": 0.927000431688122, + "grad_norm": 1.3180525302886963, + "learning_rate": 3.2336553897778014e-07, + "loss": 0.4479, + "step": 20937 + }, + { + "epoch": 0.9270447073929358, + "grad_norm": 1.6043639183044434, + "learning_rate": 3.2297582252860503e-07, + "loss": 0.4751, + "step": 20938 + }, + { + "epoch": 0.9270889830977497, + "grad_norm": 1.411928653717041, + "learning_rate": 3.225863372072746e-07, + "loss": 0.4762, + "step": 20939 + }, + { + "epoch": 0.9271332588025636, + "grad_norm": 1.5774850845336914, + "learning_rate": 3.221970830230903e-07, + "loss": 0.4588, + "step": 20940 + }, + { + "epoch": 0.9271775345073774, + "grad_norm": 1.7154970169067383, + "learning_rate": 3.218080599853468e-07, + "loss": 0.5068, + "step": 20941 + }, + { + "epoch": 0.9272218102121913, + "grad_norm": 1.3214664459228516, + "learning_rate": 3.21419268103339e-07, + "loss": 0.4693, + "step": 20942 + }, + { + "epoch": 0.9272660859170052, + "grad_norm": 1.37458074092865, + "learning_rate": 3.2103070738635055e-07, + "loss": 0.4647, + "step": 20943 + }, + { + "epoch": 0.927310361621819, + "grad_norm": 1.4484907388687134, + "learning_rate": 3.20642377843664e-07, + "loss": 0.465, + "step": 20944 + }, + { + "epoch": 0.9273546373266329, + "grad_norm": 1.4736104011535645, + "learning_rate": 3.20254279484552e-07, + "loss": 0.508, + "step": 20945 + }, + { + "epoch": 0.9273989130314468, + "grad_norm": 1.525014042854309, + "learning_rate": 3.1986641231828483e-07, + "loss": 0.4384, + "step": 20946 + }, + { + "epoch": 0.9274431887362607, + "grad_norm": 1.472365379333496, + "learning_rate": 3.194787763541274e-07, + "loss": 0.4688, + "step": 20947 + }, + { + "epoch": 0.9274874644410745, + "grad_norm": 1.605749249458313, + "learning_rate": 3.190913716013355e-07, + "loss": 0.5284, + "step": 20948 + }, + { + "epoch": 0.9275317401458885, + "grad_norm": 1.6061893701553345, + "learning_rate": 3.187041980691652e-07, + "loss": 0.4888, + "step": 20949 + }, + { + "epoch": 0.9275760158507024, + "grad_norm": 1.444854736328125, + "learning_rate": 3.1831725576686234e-07, + "loss": 0.4372, + "step": 20950 + }, + { + "epoch": 0.9276202915555162, + "grad_norm": 1.365209698677063, + "learning_rate": 3.1793054470366733e-07, + "loss": 0.4469, + "step": 20951 + }, + { + "epoch": 0.9276645672603301, + "grad_norm": 1.7614988088607788, + "learning_rate": 3.175440648888195e-07, + "loss": 0.5455, + "step": 20952 + }, + { + "epoch": 0.927708842965144, + "grad_norm": 1.5880423784255981, + "learning_rate": 3.1715781633154805e-07, + "loss": 0.4633, + "step": 20953 + }, + { + "epoch": 0.9277531186699578, + "grad_norm": 1.4517560005187988, + "learning_rate": 3.167717990410768e-07, + "loss": 0.4529, + "step": 20954 + }, + { + "epoch": 0.9277973943747717, + "grad_norm": 1.3920927047729492, + "learning_rate": 3.163860130266283e-07, + "loss": 0.4494, + "step": 20955 + }, + { + "epoch": 0.9278416700795856, + "grad_norm": 1.9017900228500366, + "learning_rate": 3.160004582974141e-07, + "loss": 0.5387, + "step": 20956 + }, + { + "epoch": 0.9278859457843994, + "grad_norm": 1.5118434429168701, + "learning_rate": 3.156151348626446e-07, + "loss": 0.4794, + "step": 20957 + }, + { + "epoch": 0.9279302214892133, + "grad_norm": 1.451502799987793, + "learning_rate": 3.1523004273152025e-07, + "loss": 0.4751, + "step": 20958 + }, + { + "epoch": 0.9279744971940272, + "grad_norm": 1.6565184593200684, + "learning_rate": 3.1484518191324253e-07, + "loss": 0.512, + "step": 20959 + }, + { + "epoch": 0.928018772898841, + "grad_norm": 2.0901434421539307, + "learning_rate": 3.144605524170019e-07, + "loss": 0.5401, + "step": 20960 + }, + { + "epoch": 0.928063048603655, + "grad_norm": 1.5553324222564697, + "learning_rate": 3.1407615425198324e-07, + "loss": 0.5022, + "step": 20961 + }, + { + "epoch": 0.9281073243084689, + "grad_norm": 1.3494815826416016, + "learning_rate": 3.1369198742737027e-07, + "loss": 0.451, + "step": 20962 + }, + { + "epoch": 0.9281516000132827, + "grad_norm": 1.7462581396102905, + "learning_rate": 3.1330805195233684e-07, + "loss": 0.551, + "step": 20963 + }, + { + "epoch": 0.9281958757180966, + "grad_norm": 1.3552042245864868, + "learning_rate": 3.129243478360544e-07, + "loss": 0.4236, + "step": 20964 + }, + { + "epoch": 0.9282401514229105, + "grad_norm": 1.5610551834106445, + "learning_rate": 3.125408750876857e-07, + "loss": 0.4812, + "step": 20965 + }, + { + "epoch": 0.9282844271277243, + "grad_norm": 1.3759404420852661, + "learning_rate": 3.121576337163901e-07, + "loss": 0.4439, + "step": 20966 + }, + { + "epoch": 0.9283287028325382, + "grad_norm": 1.479600191116333, + "learning_rate": 3.117746237313213e-07, + "loss": 0.4685, + "step": 20967 + }, + { + "epoch": 0.9283729785373521, + "grad_norm": 1.437086582183838, + "learning_rate": 3.1139184514162646e-07, + "loss": 0.4754, + "step": 20968 + }, + { + "epoch": 0.9284172542421659, + "grad_norm": 1.44585382938385, + "learning_rate": 3.110092979564505e-07, + "loss": 0.4625, + "step": 20969 + }, + { + "epoch": 0.9284615299469798, + "grad_norm": 1.7151732444763184, + "learning_rate": 3.106269821849273e-07, + "loss": 0.495, + "step": 20970 + }, + { + "epoch": 0.9285058056517937, + "grad_norm": 1.6324328184127808, + "learning_rate": 3.1024489783618826e-07, + "loss": 0.4563, + "step": 20971 + }, + { + "epoch": 0.9285500813566075, + "grad_norm": 1.7600334882736206, + "learning_rate": 3.098630449193618e-07, + "loss": 0.5052, + "step": 20972 + }, + { + "epoch": 0.9285943570614215, + "grad_norm": 1.2876622676849365, + "learning_rate": 3.0948142344356613e-07, + "loss": 0.4291, + "step": 20973 + }, + { + "epoch": 0.9286386327662354, + "grad_norm": 1.5077406167984009, + "learning_rate": 3.0910003341791615e-07, + "loss": 0.4963, + "step": 20974 + }, + { + "epoch": 0.9286829084710492, + "grad_norm": 1.1956034898757935, + "learning_rate": 3.087188748515213e-07, + "loss": 0.4289, + "step": 20975 + }, + { + "epoch": 0.9287271841758631, + "grad_norm": 1.3740148544311523, + "learning_rate": 3.0833794775348534e-07, + "loss": 0.4767, + "step": 20976 + }, + { + "epoch": 0.928771459880677, + "grad_norm": 1.3000203371047974, + "learning_rate": 3.079572521329066e-07, + "loss": 0.4344, + "step": 20977 + }, + { + "epoch": 0.9288157355854908, + "grad_norm": 1.7510526180267334, + "learning_rate": 3.0757678799887556e-07, + "loss": 0.5381, + "step": 20978 + }, + { + "epoch": 0.9288600112903047, + "grad_norm": 1.6675561666488647, + "learning_rate": 3.0719655536048276e-07, + "loss": 0.4878, + "step": 20979 + }, + { + "epoch": 0.9289042869951186, + "grad_norm": 1.7533234357833862, + "learning_rate": 3.0681655422680756e-07, + "loss": 0.4676, + "step": 20980 + }, + { + "epoch": 0.9289485626999325, + "grad_norm": 1.5850673913955688, + "learning_rate": 3.0643678460692607e-07, + "loss": 0.4557, + "step": 20981 + }, + { + "epoch": 0.9289928384047463, + "grad_norm": 1.5363034009933472, + "learning_rate": 3.0605724650990986e-07, + "loss": 0.512, + "step": 20982 + }, + { + "epoch": 0.9290371141095602, + "grad_norm": 1.5398945808410645, + "learning_rate": 3.0567793994482397e-07, + "loss": 0.4582, + "step": 20983 + }, + { + "epoch": 0.9290813898143742, + "grad_norm": 1.7295624017715454, + "learning_rate": 3.052988649207256e-07, + "loss": 0.5117, + "step": 20984 + }, + { + "epoch": 0.929125665519188, + "grad_norm": 1.202866792678833, + "learning_rate": 3.049200214466708e-07, + "loss": 0.426, + "step": 20985 + }, + { + "epoch": 0.9291699412240019, + "grad_norm": 1.5643354654312134, + "learning_rate": 3.0454140953170675e-07, + "loss": 0.5037, + "step": 20986 + }, + { + "epoch": 0.9292142169288158, + "grad_norm": 1.4723089933395386, + "learning_rate": 3.041630291848774e-07, + "loss": 0.459, + "step": 20987 + }, + { + "epoch": 0.9292584926336296, + "grad_norm": 1.4522769451141357, + "learning_rate": 3.037848804152188e-07, + "loss": 0.4744, + "step": 20988 + }, + { + "epoch": 0.9293027683384435, + "grad_norm": 1.2343339920043945, + "learning_rate": 3.0340696323176376e-07, + "loss": 0.4731, + "step": 20989 + }, + { + "epoch": 0.9293470440432574, + "grad_norm": 1.6458697319030762, + "learning_rate": 3.030292776435384e-07, + "loss": 0.5239, + "step": 20990 + }, + { + "epoch": 0.9293913197480712, + "grad_norm": 1.626234531402588, + "learning_rate": 3.0265182365956213e-07, + "loss": 0.5349, + "step": 20991 + }, + { + "epoch": 0.9294355954528851, + "grad_norm": 1.427150011062622, + "learning_rate": 3.0227460128885336e-07, + "loss": 0.4215, + "step": 20992 + }, + { + "epoch": 0.929479871157699, + "grad_norm": 1.4609942436218262, + "learning_rate": 3.018976105404181e-07, + "loss": 0.4714, + "step": 20993 + }, + { + "epoch": 0.9295241468625128, + "grad_norm": 1.3679128885269165, + "learning_rate": 3.015208514232604e-07, + "loss": 0.4557, + "step": 20994 + }, + { + "epoch": 0.9295684225673267, + "grad_norm": 1.3330800533294678, + "learning_rate": 3.0114432394638294e-07, + "loss": 0.4342, + "step": 20995 + }, + { + "epoch": 0.9296126982721407, + "grad_norm": 1.4381924867630005, + "learning_rate": 3.0076802811877525e-07, + "loss": 0.4541, + "step": 20996 + }, + { + "epoch": 0.9296569739769545, + "grad_norm": 1.561951756477356, + "learning_rate": 3.003919639494257e-07, + "loss": 0.4464, + "step": 20997 + }, + { + "epoch": 0.9297012496817684, + "grad_norm": 1.4833669662475586, + "learning_rate": 3.00016131447316e-07, + "loss": 0.4904, + "step": 20998 + }, + { + "epoch": 0.9297455253865823, + "grad_norm": 1.4427039623260498, + "learning_rate": 2.996405306214234e-07, + "loss": 0.4419, + "step": 20999 + }, + { + "epoch": 0.9297898010913961, + "grad_norm": 1.431045651435852, + "learning_rate": 2.992651614807185e-07, + "loss": 0.4136, + "step": 21000 + }, + { + "epoch": 0.92983407679621, + "grad_norm": 1.6370532512664795, + "learning_rate": 2.988900240341663e-07, + "loss": 0.4678, + "step": 21001 + }, + { + "epoch": 0.9298783525010239, + "grad_norm": 1.5119298696517944, + "learning_rate": 2.9851511829072757e-07, + "loss": 0.4276, + "step": 21002 + }, + { + "epoch": 0.9299226282058377, + "grad_norm": 1.5896105766296387, + "learning_rate": 2.9814044425935605e-07, + "loss": 0.4846, + "step": 21003 + }, + { + "epoch": 0.9299669039106516, + "grad_norm": 1.605928897857666, + "learning_rate": 2.977660019489992e-07, + "loss": 0.4968, + "step": 21004 + }, + { + "epoch": 0.9300111796154655, + "grad_norm": 1.85453200340271, + "learning_rate": 2.97391791368602e-07, + "loss": 0.5789, + "step": 21005 + }, + { + "epoch": 0.9300554553202793, + "grad_norm": 2.5046894550323486, + "learning_rate": 2.9701781252710283e-07, + "loss": 0.5118, + "step": 21006 + }, + { + "epoch": 0.9300997310250932, + "grad_norm": 1.5156883001327515, + "learning_rate": 2.966440654334324e-07, + "loss": 0.4722, + "step": 21007 + }, + { + "epoch": 0.9301440067299072, + "grad_norm": 1.4277679920196533, + "learning_rate": 2.9627055009651573e-07, + "loss": 0.4997, + "step": 21008 + }, + { + "epoch": 0.930188282434721, + "grad_norm": 1.563504934310913, + "learning_rate": 2.9589726652527793e-07, + "loss": 0.4789, + "step": 21009 + }, + { + "epoch": 0.9302325581395349, + "grad_norm": 2.0846428871154785, + "learning_rate": 2.955242147286319e-07, + "loss": 0.4814, + "step": 21010 + }, + { + "epoch": 0.9302768338443488, + "grad_norm": 1.5142772197723389, + "learning_rate": 2.951513947154894e-07, + "loss": 0.5041, + "step": 21011 + }, + { + "epoch": 0.9303211095491626, + "grad_norm": 1.3877415657043457, + "learning_rate": 2.947788064947532e-07, + "loss": 0.486, + "step": 21012 + }, + { + "epoch": 0.9303653852539765, + "grad_norm": 1.4014496803283691, + "learning_rate": 2.94406450075323e-07, + "loss": 0.4489, + "step": 21013 + }, + { + "epoch": 0.9304096609587904, + "grad_norm": 1.769439697265625, + "learning_rate": 2.940343254660905e-07, + "loss": 0.5416, + "step": 21014 + }, + { + "epoch": 0.9304539366636042, + "grad_norm": 1.6547294855117798, + "learning_rate": 2.936624326759463e-07, + "loss": 0.5238, + "step": 21015 + }, + { + "epoch": 0.9304982123684181, + "grad_norm": 1.90091073513031, + "learning_rate": 2.932907717137723e-07, + "loss": 0.5369, + "step": 21016 + }, + { + "epoch": 0.930542488073232, + "grad_norm": 1.236456036567688, + "learning_rate": 2.929193425884447e-07, + "loss": 0.4394, + "step": 21017 + }, + { + "epoch": 0.9305867637780459, + "grad_norm": 1.5542289018630981, + "learning_rate": 2.9254814530883415e-07, + "loss": 0.5032, + "step": 21018 + }, + { + "epoch": 0.9306310394828597, + "grad_norm": 1.7713176012039185, + "learning_rate": 2.921771798838069e-07, + "loss": 0.5118, + "step": 21019 + }, + { + "epoch": 0.9306753151876737, + "grad_norm": 1.7556713819503784, + "learning_rate": 2.9180644632222477e-07, + "loss": 0.4994, + "step": 21020 + }, + { + "epoch": 0.9307195908924876, + "grad_norm": 1.7741479873657227, + "learning_rate": 2.9143594463293957e-07, + "loss": 0.5043, + "step": 21021 + }, + { + "epoch": 0.9307638665973014, + "grad_norm": 1.3028051853179932, + "learning_rate": 2.91065674824802e-07, + "loss": 0.4003, + "step": 21022 + }, + { + "epoch": 0.9308081423021153, + "grad_norm": 1.6469836235046387, + "learning_rate": 2.9069563690665605e-07, + "loss": 0.4934, + "step": 21023 + }, + { + "epoch": 0.9308524180069292, + "grad_norm": 1.5252270698547363, + "learning_rate": 2.90325830887338e-07, + "loss": 0.5091, + "step": 21024 + }, + { + "epoch": 0.930896693711743, + "grad_norm": 1.6982402801513672, + "learning_rate": 2.899562567756831e-07, + "loss": 0.5583, + "step": 21025 + }, + { + "epoch": 0.9309409694165569, + "grad_norm": 1.6548168659210205, + "learning_rate": 2.895869145805164e-07, + "loss": 0.5128, + "step": 21026 + }, + { + "epoch": 0.9309852451213708, + "grad_norm": 2.3355584144592285, + "learning_rate": 2.8921780431065973e-07, + "loss": 0.5499, + "step": 21027 + }, + { + "epoch": 0.9310295208261846, + "grad_norm": 1.5096232891082764, + "learning_rate": 2.888489259749272e-07, + "loss": 0.4823, + "step": 21028 + }, + { + "epoch": 0.9310737965309985, + "grad_norm": 1.5905823707580566, + "learning_rate": 2.884802795821329e-07, + "loss": 0.4687, + "step": 21029 + }, + { + "epoch": 0.9311180722358124, + "grad_norm": 1.4700205326080322, + "learning_rate": 2.881118651410797e-07, + "loss": 0.4099, + "step": 21030 + }, + { + "epoch": 0.9311623479406262, + "grad_norm": 1.4190839529037476, + "learning_rate": 2.877436826605662e-07, + "loss": 0.4697, + "step": 21031 + }, + { + "epoch": 0.9312066236454402, + "grad_norm": 1.6096596717834473, + "learning_rate": 2.873757321493875e-07, + "loss": 0.4587, + "step": 21032 + }, + { + "epoch": 0.9312508993502541, + "grad_norm": 1.8283748626708984, + "learning_rate": 2.870080136163311e-07, + "loss": 0.502, + "step": 21033 + }, + { + "epoch": 0.9312951750550679, + "grad_norm": 1.4747655391693115, + "learning_rate": 2.866405270701789e-07, + "loss": 0.4422, + "step": 21034 + }, + { + "epoch": 0.9313394507598818, + "grad_norm": 1.6846688985824585, + "learning_rate": 2.862732725197093e-07, + "loss": 0.5769, + "step": 21035 + }, + { + "epoch": 0.9313837264646957, + "grad_norm": 1.4882205724716187, + "learning_rate": 2.859062499736931e-07, + "loss": 0.4604, + "step": 21036 + }, + { + "epoch": 0.9314280021695095, + "grad_norm": 1.5295482873916626, + "learning_rate": 2.855394594408967e-07, + "loss": 0.4682, + "step": 21037 + }, + { + "epoch": 0.9314722778743234, + "grad_norm": 1.363230586051941, + "learning_rate": 2.8517290093008076e-07, + "loss": 0.4705, + "step": 21038 + }, + { + "epoch": 0.9315165535791373, + "grad_norm": 1.4115456342697144, + "learning_rate": 2.848065744500006e-07, + "loss": 0.4264, + "step": 21039 + }, + { + "epoch": 0.9315608292839511, + "grad_norm": 1.4379661083221436, + "learning_rate": 2.8444048000940473e-07, + "loss": 0.4579, + "step": 21040 + }, + { + "epoch": 0.931605104988765, + "grad_norm": 1.5242470502853394, + "learning_rate": 2.8407461761703613e-07, + "loss": 0.5147, + "step": 21041 + }, + { + "epoch": 0.9316493806935789, + "grad_norm": 1.8053208589553833, + "learning_rate": 2.8370898728163456e-07, + "loss": 0.4927, + "step": 21042 + }, + { + "epoch": 0.9316936563983927, + "grad_norm": 1.3286689519882202, + "learning_rate": 2.8334358901193294e-07, + "loss": 0.4604, + "step": 21043 + }, + { + "epoch": 0.9317379321032067, + "grad_norm": 1.403613567352295, + "learning_rate": 2.8297842281665877e-07, + "loss": 0.5055, + "step": 21044 + }, + { + "epoch": 0.9317822078080206, + "grad_norm": 2.381131887435913, + "learning_rate": 2.8261348870453174e-07, + "loss": 0.5264, + "step": 21045 + }, + { + "epoch": 0.9318264835128344, + "grad_norm": 1.6607025861740112, + "learning_rate": 2.8224878668426934e-07, + "loss": 0.4811, + "step": 21046 + }, + { + "epoch": 0.9318707592176483, + "grad_norm": 1.8676886558532715, + "learning_rate": 2.818843167645835e-07, + "loss": 0.5535, + "step": 21047 + }, + { + "epoch": 0.9319150349224622, + "grad_norm": 1.4892501831054688, + "learning_rate": 2.8152007895417723e-07, + "loss": 0.469, + "step": 21048 + }, + { + "epoch": 0.931959310627276, + "grad_norm": 1.4006474018096924, + "learning_rate": 2.811560732617513e-07, + "loss": 0.4637, + "step": 21049 + }, + { + "epoch": 0.9320035863320899, + "grad_norm": 1.8519458770751953, + "learning_rate": 2.8079229969599886e-07, + "loss": 0.5732, + "step": 21050 + }, + { + "epoch": 0.9320478620369038, + "grad_norm": 1.3871454000473022, + "learning_rate": 2.804287582656073e-07, + "loss": 0.4358, + "step": 21051 + }, + { + "epoch": 0.9320921377417177, + "grad_norm": 1.2428325414657593, + "learning_rate": 2.8006544897926315e-07, + "loss": 0.4131, + "step": 21052 + }, + { + "epoch": 0.9321364134465315, + "grad_norm": 1.393295168876648, + "learning_rate": 2.797023718456404e-07, + "loss": 0.4561, + "step": 21053 + }, + { + "epoch": 0.9321806891513454, + "grad_norm": 1.5094830989837646, + "learning_rate": 2.7933952687341224e-07, + "loss": 0.4729, + "step": 21054 + }, + { + "epoch": 0.9322249648561594, + "grad_norm": 1.435638427734375, + "learning_rate": 2.7897691407124394e-07, + "loss": 0.4468, + "step": 21055 + }, + { + "epoch": 0.9322692405609732, + "grad_norm": 1.5054377317428589, + "learning_rate": 2.786145334477974e-07, + "loss": 0.4497, + "step": 21056 + }, + { + "epoch": 0.9323135162657871, + "grad_norm": 1.3552157878875732, + "learning_rate": 2.782523850117269e-07, + "loss": 0.4878, + "step": 21057 + }, + { + "epoch": 0.932357791970601, + "grad_norm": 1.9250285625457764, + "learning_rate": 2.7789046877168324e-07, + "loss": 0.4805, + "step": 21058 + }, + { + "epoch": 0.9324020676754148, + "grad_norm": 1.915611743927002, + "learning_rate": 2.775287847363095e-07, + "loss": 0.552, + "step": 21059 + }, + { + "epoch": 0.9324463433802287, + "grad_norm": 1.8945114612579346, + "learning_rate": 2.7716733291424437e-07, + "loss": 0.5276, + "step": 21060 + }, + { + "epoch": 0.9324906190850426, + "grad_norm": 1.9616247415542603, + "learning_rate": 2.7680611331411977e-07, + "loss": 0.5074, + "step": 21061 + }, + { + "epoch": 0.9325348947898564, + "grad_norm": 1.2879937887191772, + "learning_rate": 2.764451259445655e-07, + "loss": 0.4322, + "step": 21062 + }, + { + "epoch": 0.9325791704946703, + "grad_norm": 1.271653413772583, + "learning_rate": 2.7608437081420136e-07, + "loss": 0.4519, + "step": 21063 + }, + { + "epoch": 0.9326234461994842, + "grad_norm": 1.8529093265533447, + "learning_rate": 2.75723847931646e-07, + "loss": 0.5377, + "step": 21064 + }, + { + "epoch": 0.932667721904298, + "grad_norm": 1.697391390800476, + "learning_rate": 2.753635573055069e-07, + "loss": 0.4981, + "step": 21065 + }, + { + "epoch": 0.9327119976091119, + "grad_norm": 1.324825406074524, + "learning_rate": 2.750034989443917e-07, + "loss": 0.4674, + "step": 21066 + }, + { + "epoch": 0.9327562733139259, + "grad_norm": 1.5599719285964966, + "learning_rate": 2.746436728569002e-07, + "loss": 0.4788, + "step": 21067 + }, + { + "epoch": 0.9328005490187397, + "grad_norm": 1.4002933502197266, + "learning_rate": 2.7428407905162546e-07, + "loss": 0.4624, + "step": 21068 + }, + { + "epoch": 0.9328448247235536, + "grad_norm": 1.431347131729126, + "learning_rate": 2.7392471753715623e-07, + "loss": 0.4754, + "step": 21069 + }, + { + "epoch": 0.9328891004283675, + "grad_norm": 1.5874191522598267, + "learning_rate": 2.7356558832207667e-07, + "loss": 0.535, + "step": 21070 + }, + { + "epoch": 0.9329333761331813, + "grad_norm": 1.4825841188430786, + "learning_rate": 2.7320669141496224e-07, + "loss": 0.5238, + "step": 21071 + }, + { + "epoch": 0.9329776518379952, + "grad_norm": 1.8475778102874756, + "learning_rate": 2.7284802682438715e-07, + "loss": 0.5031, + "step": 21072 + }, + { + "epoch": 0.9330219275428091, + "grad_norm": 1.8811938762664795, + "learning_rate": 2.724895945589168e-07, + "loss": 0.5395, + "step": 21073 + }, + { + "epoch": 0.9330662032476229, + "grad_norm": 2.048029661178589, + "learning_rate": 2.7213139462711203e-07, + "loss": 0.5215, + "step": 21074 + }, + { + "epoch": 0.9331104789524368, + "grad_norm": 1.7295198440551758, + "learning_rate": 2.717734270375272e-07, + "loss": 0.5301, + "step": 21075 + }, + { + "epoch": 0.9331547546572507, + "grad_norm": 1.4671369791030884, + "learning_rate": 2.714156917987132e-07, + "loss": 0.4752, + "step": 21076 + }, + { + "epoch": 0.9331990303620645, + "grad_norm": 1.4230543375015259, + "learning_rate": 2.7105818891921544e-07, + "loss": 0.4335, + "step": 21077 + }, + { + "epoch": 0.9332433060668784, + "grad_norm": 1.3928016424179077, + "learning_rate": 2.707009184075704e-07, + "loss": 0.4451, + "step": 21078 + }, + { + "epoch": 0.9332875817716924, + "grad_norm": 1.775026559829712, + "learning_rate": 2.703438802723124e-07, + "loss": 0.4899, + "step": 21079 + }, + { + "epoch": 0.9333318574765062, + "grad_norm": 1.5075583457946777, + "learning_rate": 2.6998707452196793e-07, + "loss": 0.5107, + "step": 21080 + }, + { + "epoch": 0.9333761331813201, + "grad_norm": 1.6550592184066772, + "learning_rate": 2.6963050116505906e-07, + "loss": 0.4514, + "step": 21081 + }, + { + "epoch": 0.933420408886134, + "grad_norm": 1.5591914653778076, + "learning_rate": 2.6927416021010454e-07, + "loss": 0.5218, + "step": 21082 + }, + { + "epoch": 0.9334646845909478, + "grad_norm": 1.7526979446411133, + "learning_rate": 2.6891805166561313e-07, + "loss": 0.5274, + "step": 21083 + }, + { + "epoch": 0.9335089602957617, + "grad_norm": 1.5944095849990845, + "learning_rate": 2.6856217554009136e-07, + "loss": 0.4752, + "step": 21084 + }, + { + "epoch": 0.9335532360005756, + "grad_norm": 1.4265097379684448, + "learning_rate": 2.6820653184203684e-07, + "loss": 0.4666, + "step": 21085 + }, + { + "epoch": 0.9335975117053895, + "grad_norm": 1.66696298122406, + "learning_rate": 2.678511205799461e-07, + "loss": 0.5485, + "step": 21086 + }, + { + "epoch": 0.9336417874102033, + "grad_norm": 1.4819064140319824, + "learning_rate": 2.6749594176230797e-07, + "loss": 0.4104, + "step": 21087 + }, + { + "epoch": 0.9336860631150172, + "grad_norm": 1.611242413520813, + "learning_rate": 2.671409953976034e-07, + "loss": 0.4838, + "step": 21088 + }, + { + "epoch": 0.9337303388198311, + "grad_norm": 1.7333674430847168, + "learning_rate": 2.667862814943123e-07, + "loss": 0.4666, + "step": 21089 + }, + { + "epoch": 0.9337746145246449, + "grad_norm": 1.435483694076538, + "learning_rate": 2.6643180006090564e-07, + "loss": 0.4892, + "step": 21090 + }, + { + "epoch": 0.9338188902294589, + "grad_norm": 1.5005192756652832, + "learning_rate": 2.6607755110584886e-07, + "loss": 0.4748, + "step": 21091 + }, + { + "epoch": 0.9338631659342728, + "grad_norm": 1.5542877912521362, + "learning_rate": 2.6572353463760636e-07, + "loss": 0.4588, + "step": 21092 + }, + { + "epoch": 0.9339074416390866, + "grad_norm": 1.5935386419296265, + "learning_rate": 2.653697506646302e-07, + "loss": 0.478, + "step": 21093 + }, + { + "epoch": 0.9339517173439005, + "grad_norm": 1.3834540843963623, + "learning_rate": 2.650161991953715e-07, + "loss": 0.4775, + "step": 21094 + }, + { + "epoch": 0.9339959930487144, + "grad_norm": 1.4650944471359253, + "learning_rate": 2.6466288023827445e-07, + "loss": 0.4361, + "step": 21095 + }, + { + "epoch": 0.9340402687535282, + "grad_norm": 1.417516827583313, + "learning_rate": 2.643097938017769e-07, + "loss": 0.4249, + "step": 21096 + }, + { + "epoch": 0.9340845444583421, + "grad_norm": 1.308775782585144, + "learning_rate": 2.6395693989431316e-07, + "loss": 0.4187, + "step": 21097 + }, + { + "epoch": 0.934128820163156, + "grad_norm": 1.4798134565353394, + "learning_rate": 2.636043185243098e-07, + "loss": 0.4818, + "step": 21098 + }, + { + "epoch": 0.9341730958679698, + "grad_norm": 1.7952871322631836, + "learning_rate": 2.632519297001912e-07, + "loss": 0.5355, + "step": 21099 + }, + { + "epoch": 0.9342173715727837, + "grad_norm": 1.6534080505371094, + "learning_rate": 2.6289977343037175e-07, + "loss": 0.4644, + "step": 21100 + }, + { + "epoch": 0.9342616472775976, + "grad_norm": 1.2180168628692627, + "learning_rate": 2.6254784972326255e-07, + "loss": 0.4165, + "step": 21101 + }, + { + "epoch": 0.9343059229824114, + "grad_norm": 1.4281713962554932, + "learning_rate": 2.62196158587269e-07, + "loss": 0.4233, + "step": 21102 + }, + { + "epoch": 0.9343501986872254, + "grad_norm": 1.5728886127471924, + "learning_rate": 2.618447000307922e-07, + "loss": 0.4477, + "step": 21103 + }, + { + "epoch": 0.9343944743920393, + "grad_norm": 1.333828330039978, + "learning_rate": 2.614934740622266e-07, + "loss": 0.4511, + "step": 21104 + }, + { + "epoch": 0.9344387500968531, + "grad_norm": 1.4197487831115723, + "learning_rate": 2.6114248068995875e-07, + "loss": 0.4635, + "step": 21105 + }, + { + "epoch": 0.934483025801667, + "grad_norm": 1.4028103351593018, + "learning_rate": 2.6079171992237307e-07, + "loss": 0.4519, + "step": 21106 + }, + { + "epoch": 0.9345273015064809, + "grad_norm": 1.398241400718689, + "learning_rate": 2.6044119176784733e-07, + "loss": 0.417, + "step": 21107 + }, + { + "epoch": 0.9345715772112947, + "grad_norm": 1.6095768213272095, + "learning_rate": 2.600908962347537e-07, + "loss": 0.4891, + "step": 21108 + }, + { + "epoch": 0.9346158529161086, + "grad_norm": 1.3550680875778198, + "learning_rate": 2.5974083333145773e-07, + "loss": 0.4669, + "step": 21109 + }, + { + "epoch": 0.9346601286209225, + "grad_norm": 1.470119833946228, + "learning_rate": 2.593910030663227e-07, + "loss": 0.4772, + "step": 21110 + }, + { + "epoch": 0.9347044043257363, + "grad_norm": 1.396442174911499, + "learning_rate": 2.590414054476997e-07, + "loss": 0.4772, + "step": 21111 + }, + { + "epoch": 0.9347486800305502, + "grad_norm": 1.8196346759796143, + "learning_rate": 2.5869204048394326e-07, + "loss": 0.5521, + "step": 21112 + }, + { + "epoch": 0.9347929557353641, + "grad_norm": 1.5636457204818726, + "learning_rate": 2.583429081833966e-07, + "loss": 0.4565, + "step": 21113 + }, + { + "epoch": 0.9348372314401779, + "grad_norm": 1.4346404075622559, + "learning_rate": 2.579940085543964e-07, + "loss": 0.463, + "step": 21114 + }, + { + "epoch": 0.9348815071449919, + "grad_norm": 1.5286197662353516, + "learning_rate": 2.576453416052782e-07, + "loss": 0.4789, + "step": 21115 + }, + { + "epoch": 0.9349257828498058, + "grad_norm": 1.627496361732483, + "learning_rate": 2.572969073443676e-07, + "loss": 0.527, + "step": 21116 + }, + { + "epoch": 0.9349700585546196, + "grad_norm": 1.451411247253418, + "learning_rate": 2.56948705779988e-07, + "loss": 0.4551, + "step": 21117 + }, + { + "epoch": 0.9350143342594335, + "grad_norm": 1.8893717527389526, + "learning_rate": 2.5660073692045374e-07, + "loss": 0.4993, + "step": 21118 + }, + { + "epoch": 0.9350586099642474, + "grad_norm": 1.5618915557861328, + "learning_rate": 2.5625300077407933e-07, + "loss": 0.5004, + "step": 21119 + }, + { + "epoch": 0.9351028856690612, + "grad_norm": 1.391644835472107, + "learning_rate": 2.5590549734916815e-07, + "loss": 0.4424, + "step": 21120 + }, + { + "epoch": 0.9351471613738751, + "grad_norm": 1.580972671508789, + "learning_rate": 2.555582266540191e-07, + "loss": 0.5179, + "step": 21121 + }, + { + "epoch": 0.935191437078689, + "grad_norm": 1.4626928567886353, + "learning_rate": 2.5521118869692774e-07, + "loss": 0.484, + "step": 21122 + }, + { + "epoch": 0.9352357127835029, + "grad_norm": 1.6764270067214966, + "learning_rate": 2.5486438348618416e-07, + "loss": 0.4657, + "step": 21123 + }, + { + "epoch": 0.9352799884883167, + "grad_norm": 2.056772232055664, + "learning_rate": 2.5451781103006944e-07, + "loss": 0.4906, + "step": 21124 + }, + { + "epoch": 0.9353242641931306, + "grad_norm": 1.574803352355957, + "learning_rate": 2.541714713368615e-07, + "loss": 0.5073, + "step": 21125 + }, + { + "epoch": 0.9353685398979446, + "grad_norm": 1.3807140588760376, + "learning_rate": 2.5382536441483254e-07, + "loss": 0.4971, + "step": 21126 + }, + { + "epoch": 0.9354128156027584, + "grad_norm": 1.5685830116271973, + "learning_rate": 2.534794902722493e-07, + "loss": 0.4691, + "step": 21127 + }, + { + "epoch": 0.9354570913075723, + "grad_norm": 1.5122472047805786, + "learning_rate": 2.531338489173718e-07, + "loss": 0.4991, + "step": 21128 + }, + { + "epoch": 0.9355013670123862, + "grad_norm": 1.5817182064056396, + "learning_rate": 2.5278844035845686e-07, + "loss": 0.4738, + "step": 21129 + }, + { + "epoch": 0.9355456427172, + "grad_norm": 1.4998377561569214, + "learning_rate": 2.5244326460375337e-07, + "loss": 0.4702, + "step": 21130 + }, + { + "epoch": 0.9355899184220139, + "grad_norm": 1.4826509952545166, + "learning_rate": 2.520983216615047e-07, + "loss": 0.4536, + "step": 21131 + }, + { + "epoch": 0.9356341941268278, + "grad_norm": 1.511561632156372, + "learning_rate": 2.5175361153995215e-07, + "loss": 0.4193, + "step": 21132 + }, + { + "epoch": 0.9356784698316416, + "grad_norm": 2.1531543731689453, + "learning_rate": 2.514091342473268e-07, + "loss": 0.5407, + "step": 21133 + }, + { + "epoch": 0.9357227455364555, + "grad_norm": 1.4845149517059326, + "learning_rate": 2.510648897918577e-07, + "loss": 0.4186, + "step": 21134 + }, + { + "epoch": 0.9357670212412694, + "grad_norm": 1.4746171236038208, + "learning_rate": 2.507208781817638e-07, + "loss": 0.4973, + "step": 21135 + }, + { + "epoch": 0.9358112969460832, + "grad_norm": 1.654369831085205, + "learning_rate": 2.5037709942526523e-07, + "loss": 0.5103, + "step": 21136 + }, + { + "epoch": 0.9358555726508971, + "grad_norm": 1.3719966411590576, + "learning_rate": 2.500335535305709e-07, + "loss": 0.4221, + "step": 21137 + }, + { + "epoch": 0.935899848355711, + "grad_norm": 1.5600814819335938, + "learning_rate": 2.4969024050588543e-07, + "loss": 0.4495, + "step": 21138 + }, + { + "epoch": 0.9359441240605249, + "grad_norm": 1.305457592010498, + "learning_rate": 2.4934716035940996e-07, + "loss": 0.4628, + "step": 21139 + }, + { + "epoch": 0.9359883997653388, + "grad_norm": 1.6236342191696167, + "learning_rate": 2.4900431309933917e-07, + "loss": 0.4905, + "step": 21140 + }, + { + "epoch": 0.9360326754701527, + "grad_norm": 1.5191487073898315, + "learning_rate": 2.4866169873385857e-07, + "loss": 0.4593, + "step": 21141 + }, + { + "epoch": 0.9360769511749665, + "grad_norm": 1.4147289991378784, + "learning_rate": 2.483193172711551e-07, + "loss": 0.4752, + "step": 21142 + }, + { + "epoch": 0.9361212268797804, + "grad_norm": 1.6070282459259033, + "learning_rate": 2.479771687194066e-07, + "loss": 0.4392, + "step": 21143 + }, + { + "epoch": 0.9361655025845943, + "grad_norm": 1.5896846055984497, + "learning_rate": 2.4763525308678096e-07, + "loss": 0.5079, + "step": 21144 + }, + { + "epoch": 0.9362097782894081, + "grad_norm": 1.759769082069397, + "learning_rate": 2.4729357038144495e-07, + "loss": 0.552, + "step": 21145 + }, + { + "epoch": 0.936254053994222, + "grad_norm": 1.3558530807495117, + "learning_rate": 2.469521206115633e-07, + "loss": 0.4286, + "step": 21146 + }, + { + "epoch": 0.9362983296990359, + "grad_norm": 1.5503363609313965, + "learning_rate": 2.466109037852893e-07, + "loss": 0.484, + "step": 21147 + }, + { + "epoch": 0.9363426054038497, + "grad_norm": 1.4043155908584595, + "learning_rate": 2.4626991991076987e-07, + "loss": 0.44, + "step": 21148 + }, + { + "epoch": 0.9363868811086636, + "grad_norm": 2.1937613487243652, + "learning_rate": 2.4592916899615403e-07, + "loss": 0.5711, + "step": 21149 + }, + { + "epoch": 0.9364311568134775, + "grad_norm": 1.4244424104690552, + "learning_rate": 2.455886510495775e-07, + "loss": 0.4441, + "step": 21150 + }, + { + "epoch": 0.9364754325182914, + "grad_norm": 1.505561113357544, + "learning_rate": 2.452483660791738e-07, + "loss": 0.5564, + "step": 21151 + }, + { + "epoch": 0.9365197082231053, + "grad_norm": 1.9920438528060913, + "learning_rate": 2.449083140930697e-07, + "loss": 0.4499, + "step": 21152 + }, + { + "epoch": 0.9365639839279192, + "grad_norm": 1.4045252799987793, + "learning_rate": 2.4456849509938764e-07, + "loss": 0.4713, + "step": 21153 + }, + { + "epoch": 0.936608259632733, + "grad_norm": 1.362943172454834, + "learning_rate": 2.442289091062444e-07, + "loss": 0.4647, + "step": 21154 + }, + { + "epoch": 0.9366525353375469, + "grad_norm": 1.5899786949157715, + "learning_rate": 2.4388955612174915e-07, + "loss": 0.4403, + "step": 21155 + }, + { + "epoch": 0.9366968110423608, + "grad_norm": 1.6705706119537354, + "learning_rate": 2.435504361540086e-07, + "loss": 0.4997, + "step": 21156 + }, + { + "epoch": 0.9367410867471747, + "grad_norm": 1.3607687950134277, + "learning_rate": 2.432115492111231e-07, + "loss": 0.4597, + "step": 21157 + }, + { + "epoch": 0.9367853624519885, + "grad_norm": 1.506804347038269, + "learning_rate": 2.4287289530118387e-07, + "loss": 0.4926, + "step": 21158 + }, + { + "epoch": 0.9368296381568024, + "grad_norm": 1.5087025165557861, + "learning_rate": 2.4253447443228106e-07, + "loss": 0.5023, + "step": 21159 + }, + { + "epoch": 0.9368739138616163, + "grad_norm": 1.590773105621338, + "learning_rate": 2.4219628661249827e-07, + "loss": 0.5004, + "step": 21160 + }, + { + "epoch": 0.9369181895664301, + "grad_norm": 1.977287769317627, + "learning_rate": 2.418583318499124e-07, + "loss": 0.4839, + "step": 21161 + }, + { + "epoch": 0.936962465271244, + "grad_norm": 1.641035795211792, + "learning_rate": 2.415206101525946e-07, + "loss": 0.4873, + "step": 21162 + }, + { + "epoch": 0.937006740976058, + "grad_norm": 1.578370451927185, + "learning_rate": 2.411831215286109e-07, + "loss": 0.4684, + "step": 21163 + }, + { + "epoch": 0.9370510166808718, + "grad_norm": 1.2991775274276733, + "learning_rate": 2.4084586598602356e-07, + "loss": 0.4044, + "step": 21164 + }, + { + "epoch": 0.9370952923856857, + "grad_norm": 1.7672325372695923, + "learning_rate": 2.4050884353288506e-07, + "loss": 0.572, + "step": 21165 + }, + { + "epoch": 0.9371395680904996, + "grad_norm": 1.3723876476287842, + "learning_rate": 2.401720541772468e-07, + "loss": 0.4763, + "step": 21166 + }, + { + "epoch": 0.9371838437953134, + "grad_norm": 1.2217481136322021, + "learning_rate": 2.3983549792715224e-07, + "loss": 0.4508, + "step": 21167 + }, + { + "epoch": 0.9372281195001273, + "grad_norm": 1.4987694025039673, + "learning_rate": 2.3949917479063947e-07, + "loss": 0.4942, + "step": 21168 + }, + { + "epoch": 0.9372723952049412, + "grad_norm": 1.4410783052444458, + "learning_rate": 2.39163084775742e-07, + "loss": 0.4379, + "step": 21169 + }, + { + "epoch": 0.937316670909755, + "grad_norm": 1.6891847848892212, + "learning_rate": 2.388272278904868e-07, + "loss": 0.4433, + "step": 21170 + }, + { + "epoch": 0.9373609466145689, + "grad_norm": 1.62844717502594, + "learning_rate": 2.384916041428964e-07, + "loss": 0.5133, + "step": 21171 + }, + { + "epoch": 0.9374052223193828, + "grad_norm": 1.5427993535995483, + "learning_rate": 2.381562135409854e-07, + "loss": 0.4583, + "step": 21172 + }, + { + "epoch": 0.9374494980241966, + "grad_norm": 2.015845537185669, + "learning_rate": 2.3782105609276407e-07, + "loss": 0.4806, + "step": 21173 + }, + { + "epoch": 0.9374937737290105, + "grad_norm": 1.5851662158966064, + "learning_rate": 2.3748613180623826e-07, + "loss": 0.4501, + "step": 21174 + }, + { + "epoch": 0.9375380494338245, + "grad_norm": 1.6497235298156738, + "learning_rate": 2.371514406894071e-07, + "loss": 0.514, + "step": 21175 + }, + { + "epoch": 0.9375823251386383, + "grad_norm": 1.3507574796676636, + "learning_rate": 2.3681698275026532e-07, + "loss": 0.4536, + "step": 21176 + }, + { + "epoch": 0.9376266008434522, + "grad_norm": 1.4498329162597656, + "learning_rate": 2.3648275799679987e-07, + "loss": 0.4339, + "step": 21177 + }, + { + "epoch": 0.9376708765482661, + "grad_norm": 2.1682541370391846, + "learning_rate": 2.3614876643699325e-07, + "loss": 0.5053, + "step": 21178 + }, + { + "epoch": 0.9377151522530799, + "grad_norm": 1.7745180130004883, + "learning_rate": 2.3581500807882462e-07, + "loss": 0.4619, + "step": 21179 + }, + { + "epoch": 0.9377594279578938, + "grad_norm": 1.6459226608276367, + "learning_rate": 2.354814829302643e-07, + "loss": 0.4306, + "step": 21180 + }, + { + "epoch": 0.9378037036627077, + "grad_norm": 1.5028672218322754, + "learning_rate": 2.3514819099927922e-07, + "loss": 0.4835, + "step": 21181 + }, + { + "epoch": 0.9378479793675215, + "grad_norm": 1.437016248703003, + "learning_rate": 2.3481513229382636e-07, + "loss": 0.4738, + "step": 21182 + }, + { + "epoch": 0.9378922550723354, + "grad_norm": 1.5758545398712158, + "learning_rate": 2.3448230682186378e-07, + "loss": 0.4908, + "step": 21183 + }, + { + "epoch": 0.9379365307771493, + "grad_norm": 1.6836940050125122, + "learning_rate": 2.3414971459133962e-07, + "loss": 0.4935, + "step": 21184 + }, + { + "epoch": 0.9379808064819631, + "grad_norm": 1.4874886274337769, + "learning_rate": 2.3381735561019748e-07, + "loss": 0.457, + "step": 21185 + }, + { + "epoch": 0.938025082186777, + "grad_norm": 1.2955247163772583, + "learning_rate": 2.3348522988637767e-07, + "loss": 0.4767, + "step": 21186 + }, + { + "epoch": 0.938069357891591, + "grad_norm": 1.3817040920257568, + "learning_rate": 2.3315333742780942e-07, + "loss": 0.4565, + "step": 21187 + }, + { + "epoch": 0.9381136335964048, + "grad_norm": 1.6910054683685303, + "learning_rate": 2.3282167824242196e-07, + "loss": 0.4606, + "step": 21188 + }, + { + "epoch": 0.9381579093012187, + "grad_norm": 1.3659474849700928, + "learning_rate": 2.324902523381356e-07, + "loss": 0.4709, + "step": 21189 + }, + { + "epoch": 0.9382021850060326, + "grad_norm": 1.4967334270477295, + "learning_rate": 2.3215905972286735e-07, + "loss": 0.4587, + "step": 21190 + }, + { + "epoch": 0.9382464607108464, + "grad_norm": 1.6785756349563599, + "learning_rate": 2.3182810040452641e-07, + "loss": 0.5249, + "step": 21191 + }, + { + "epoch": 0.9382907364156603, + "grad_norm": 1.3682892322540283, + "learning_rate": 2.3149737439101649e-07, + "loss": 0.4521, + "step": 21192 + }, + { + "epoch": 0.9383350121204742, + "grad_norm": 2.287214994430542, + "learning_rate": 2.3116688169024015e-07, + "loss": 0.4993, + "step": 21193 + }, + { + "epoch": 0.9383792878252881, + "grad_norm": 1.5429537296295166, + "learning_rate": 2.3083662231008775e-07, + "loss": 0.4235, + "step": 21194 + }, + { + "epoch": 0.9384235635301019, + "grad_norm": 1.578893780708313, + "learning_rate": 2.305065962584463e-07, + "loss": 0.5074, + "step": 21195 + }, + { + "epoch": 0.9384678392349158, + "grad_norm": 1.4398527145385742, + "learning_rate": 2.3017680354320282e-07, + "loss": 0.4667, + "step": 21196 + }, + { + "epoch": 0.9385121149397297, + "grad_norm": 1.544312596321106, + "learning_rate": 2.2984724417223104e-07, + "loss": 0.4818, + "step": 21197 + }, + { + "epoch": 0.9385563906445435, + "grad_norm": 1.4460984468460083, + "learning_rate": 2.295179181534024e-07, + "loss": 0.4628, + "step": 21198 + }, + { + "epoch": 0.9386006663493575, + "grad_norm": 1.4550832509994507, + "learning_rate": 2.2918882549458288e-07, + "loss": 0.4498, + "step": 21199 + }, + { + "epoch": 0.9386449420541714, + "grad_norm": 1.6644234657287598, + "learning_rate": 2.2885996620363282e-07, + "loss": 0.4934, + "step": 21200 + }, + { + "epoch": 0.9386892177589852, + "grad_norm": 1.2173165082931519, + "learning_rate": 2.2853134028840594e-07, + "loss": 0.4344, + "step": 21201 + }, + { + "epoch": 0.9387334934637991, + "grad_norm": 1.6632436513900757, + "learning_rate": 2.2820294775675045e-07, + "loss": 0.5004, + "step": 21202 + }, + { + "epoch": 0.938777769168613, + "grad_norm": 1.6636461019515991, + "learning_rate": 2.2787478861651335e-07, + "loss": 0.4623, + "step": 21203 + }, + { + "epoch": 0.9388220448734268, + "grad_norm": 2.241730213165283, + "learning_rate": 2.275468628755295e-07, + "loss": 0.5153, + "step": 21204 + }, + { + "epoch": 0.9388663205782407, + "grad_norm": 1.5599584579467773, + "learning_rate": 2.2721917054163045e-07, + "loss": 0.5023, + "step": 21205 + }, + { + "epoch": 0.9389105962830546, + "grad_norm": 1.4931812286376953, + "learning_rate": 2.2689171162264656e-07, + "loss": 0.5085, + "step": 21206 + }, + { + "epoch": 0.9389548719878684, + "grad_norm": 1.6339735984802246, + "learning_rate": 2.2656448612639492e-07, + "loss": 0.4568, + "step": 21207 + }, + { + "epoch": 0.9389991476926823, + "grad_norm": 1.4419056177139282, + "learning_rate": 2.2623749406069373e-07, + "loss": 0.424, + "step": 21208 + }, + { + "epoch": 0.9390434233974962, + "grad_norm": 1.5438792705535889, + "learning_rate": 2.2591073543335117e-07, + "loss": 0.4435, + "step": 21209 + }, + { + "epoch": 0.93908769910231, + "grad_norm": 1.805306077003479, + "learning_rate": 2.2558421025217326e-07, + "loss": 0.5476, + "step": 21210 + }, + { + "epoch": 0.939131974807124, + "grad_norm": 1.415093183517456, + "learning_rate": 2.2525791852495704e-07, + "loss": 0.4311, + "step": 21211 + }, + { + "epoch": 0.9391762505119379, + "grad_norm": 1.3820178508758545, + "learning_rate": 2.249318602594963e-07, + "loss": 0.4322, + "step": 21212 + }, + { + "epoch": 0.9392205262167517, + "grad_norm": 1.643355131149292, + "learning_rate": 2.246060354635793e-07, + "loss": 0.4951, + "step": 21213 + }, + { + "epoch": 0.9392648019215656, + "grad_norm": 1.4845821857452393, + "learning_rate": 2.2428044414498752e-07, + "loss": 0.4885, + "step": 21214 + }, + { + "epoch": 0.9393090776263795, + "grad_norm": 1.3761651515960693, + "learning_rate": 2.23955086311497e-07, + "loss": 0.4581, + "step": 21215 + }, + { + "epoch": 0.9393533533311933, + "grad_norm": 1.3061164617538452, + "learning_rate": 2.2362996197088038e-07, + "loss": 0.4439, + "step": 21216 + }, + { + "epoch": 0.9393976290360072, + "grad_norm": 1.3807202577590942, + "learning_rate": 2.233050711309026e-07, + "loss": 0.4616, + "step": 21217 + }, + { + "epoch": 0.9394419047408211, + "grad_norm": 1.7554112672805786, + "learning_rate": 2.229804137993219e-07, + "loss": 0.5292, + "step": 21218 + }, + { + "epoch": 0.9394861804456349, + "grad_norm": 1.420175313949585, + "learning_rate": 2.2265598998389316e-07, + "loss": 0.4191, + "step": 21219 + }, + { + "epoch": 0.9395304561504488, + "grad_norm": 1.5446776151657104, + "learning_rate": 2.2233179969236573e-07, + "loss": 0.4739, + "step": 21220 + }, + { + "epoch": 0.9395747318552627, + "grad_norm": 1.4522995948791504, + "learning_rate": 2.2200784293248123e-07, + "loss": 0.4238, + "step": 21221 + }, + { + "epoch": 0.9396190075600765, + "grad_norm": 1.6189154386520386, + "learning_rate": 2.216841197119768e-07, + "loss": 0.5282, + "step": 21222 + }, + { + "epoch": 0.9396632832648905, + "grad_norm": 1.386415719985962, + "learning_rate": 2.2136063003858733e-07, + "loss": 0.4728, + "step": 21223 + }, + { + "epoch": 0.9397075589697044, + "grad_norm": 1.6961166858673096, + "learning_rate": 2.2103737392003665e-07, + "loss": 0.5426, + "step": 21224 + }, + { + "epoch": 0.9397518346745182, + "grad_norm": 1.3753371238708496, + "learning_rate": 2.2071435136404418e-07, + "loss": 0.4525, + "step": 21225 + }, + { + "epoch": 0.9397961103793321, + "grad_norm": 1.424856185913086, + "learning_rate": 2.2039156237832926e-07, + "loss": 0.4553, + "step": 21226 + }, + { + "epoch": 0.939840386084146, + "grad_norm": 1.566835880279541, + "learning_rate": 2.2006900697059796e-07, + "loss": 0.479, + "step": 21227 + }, + { + "epoch": 0.9398846617889599, + "grad_norm": 1.7563107013702393, + "learning_rate": 2.1974668514855524e-07, + "loss": 0.5143, + "step": 21228 + }, + { + "epoch": 0.9399289374937737, + "grad_norm": 1.5908749103546143, + "learning_rate": 2.1942459691990048e-07, + "loss": 0.6079, + "step": 21229 + }, + { + "epoch": 0.9399732131985876, + "grad_norm": 1.459547519683838, + "learning_rate": 2.191027422923253e-07, + "loss": 0.4467, + "step": 21230 + }, + { + "epoch": 0.9400174889034015, + "grad_norm": 1.604233741760254, + "learning_rate": 2.187811212735169e-07, + "loss": 0.4792, + "step": 21231 + }, + { + "epoch": 0.9400617646082153, + "grad_norm": 1.4041457176208496, + "learning_rate": 2.1845973387115583e-07, + "loss": 0.4644, + "step": 21232 + }, + { + "epoch": 0.9401060403130292, + "grad_norm": 1.7691577672958374, + "learning_rate": 2.1813858009292142e-07, + "loss": 0.4836, + "step": 21233 + }, + { + "epoch": 0.9401503160178432, + "grad_norm": 1.7104660272598267, + "learning_rate": 2.1781765994648209e-07, + "loss": 0.4907, + "step": 21234 + }, + { + "epoch": 0.940194591722657, + "grad_norm": 1.361983060836792, + "learning_rate": 2.174969734395016e-07, + "loss": 0.4461, + "step": 21235 + }, + { + "epoch": 0.9402388674274709, + "grad_norm": 1.583593726158142, + "learning_rate": 2.1717652057964277e-07, + "loss": 0.5199, + "step": 21236 + }, + { + "epoch": 0.9402831431322848, + "grad_norm": 1.3445113897323608, + "learning_rate": 2.1685630137455725e-07, + "loss": 0.4628, + "step": 21237 + }, + { + "epoch": 0.9403274188370986, + "grad_norm": 1.466357946395874, + "learning_rate": 2.1653631583189115e-07, + "loss": 0.4438, + "step": 21238 + }, + { + "epoch": 0.9403716945419125, + "grad_norm": 1.5673186779022217, + "learning_rate": 2.1621656395929058e-07, + "loss": 0.4829, + "step": 21239 + }, + { + "epoch": 0.9404159702467264, + "grad_norm": 1.4367107152938843, + "learning_rate": 2.1589704576439164e-07, + "loss": 0.4836, + "step": 21240 + }, + { + "epoch": 0.9404602459515402, + "grad_norm": 1.443692922592163, + "learning_rate": 2.1557776125482378e-07, + "loss": 0.4513, + "step": 21241 + }, + { + "epoch": 0.9405045216563541, + "grad_norm": 1.5786160230636597, + "learning_rate": 2.1525871043821423e-07, + "loss": 0.493, + "step": 21242 + }, + { + "epoch": 0.940548797361168, + "grad_norm": 1.5532803535461426, + "learning_rate": 2.1493989332218468e-07, + "loss": 0.4815, + "step": 21243 + }, + { + "epoch": 0.9405930730659818, + "grad_norm": 1.5182198286056519, + "learning_rate": 2.1462130991434794e-07, + "loss": 0.4556, + "step": 21244 + }, + { + "epoch": 0.9406373487707957, + "grad_norm": 1.3504453897476196, + "learning_rate": 2.1430296022231346e-07, + "loss": 0.4105, + "step": 21245 + }, + { + "epoch": 0.9406816244756097, + "grad_norm": 1.9558714628219604, + "learning_rate": 2.1398484425368627e-07, + "loss": 0.4862, + "step": 21246 + }, + { + "epoch": 0.9407259001804235, + "grad_norm": 1.1855844259262085, + "learning_rate": 2.1366696201606252e-07, + "loss": 0.4004, + "step": 21247 + }, + { + "epoch": 0.9407701758852374, + "grad_norm": 1.4354872703552246, + "learning_rate": 2.133493135170328e-07, + "loss": 0.4575, + "step": 21248 + }, + { + "epoch": 0.9408144515900513, + "grad_norm": 2.136551856994629, + "learning_rate": 2.1303189876418884e-07, + "loss": 0.5453, + "step": 21249 + }, + { + "epoch": 0.9408587272948651, + "grad_norm": 1.321327805519104, + "learning_rate": 2.1271471776510788e-07, + "loss": 0.4197, + "step": 21250 + }, + { + "epoch": 0.940903002999679, + "grad_norm": 1.7894898653030396, + "learning_rate": 2.1239777052736722e-07, + "loss": 0.5019, + "step": 21251 + }, + { + "epoch": 0.9409472787044929, + "grad_norm": 1.4698516130447388, + "learning_rate": 2.1208105705853522e-07, + "loss": 0.4614, + "step": 21252 + }, + { + "epoch": 0.9409915544093067, + "grad_norm": 1.444345474243164, + "learning_rate": 2.1176457736617916e-07, + "loss": 0.4819, + "step": 21253 + }, + { + "epoch": 0.9410358301141206, + "grad_norm": 1.447365403175354, + "learning_rate": 2.1144833145785525e-07, + "loss": 0.4798, + "step": 21254 + }, + { + "epoch": 0.9410801058189345, + "grad_norm": 1.6215928792953491, + "learning_rate": 2.1113231934111854e-07, + "loss": 0.4602, + "step": 21255 + }, + { + "epoch": 0.9411243815237483, + "grad_norm": 1.9637982845306396, + "learning_rate": 2.1081654102351634e-07, + "loss": 0.5111, + "step": 21256 + }, + { + "epoch": 0.9411686572285622, + "grad_norm": 1.6143444776535034, + "learning_rate": 2.1050099651258927e-07, + "loss": 0.5071, + "step": 21257 + }, + { + "epoch": 0.9412129329333762, + "grad_norm": 1.4822916984558105, + "learning_rate": 2.1018568581587462e-07, + "loss": 0.4968, + "step": 21258 + }, + { + "epoch": 0.94125720863819, + "grad_norm": 1.8125174045562744, + "learning_rate": 2.0987060894090526e-07, + "loss": 0.5057, + "step": 21259 + }, + { + "epoch": 0.9413014843430039, + "grad_norm": 1.4569178819656372, + "learning_rate": 2.095557658952041e-07, + "loss": 0.4522, + "step": 21260 + }, + { + "epoch": 0.9413457600478178, + "grad_norm": 1.3704999685287476, + "learning_rate": 2.0924115668629285e-07, + "loss": 0.5003, + "step": 21261 + }, + { + "epoch": 0.9413900357526316, + "grad_norm": 1.2929998636245728, + "learning_rate": 2.0892678132168332e-07, + "loss": 0.4415, + "step": 21262 + }, + { + "epoch": 0.9414343114574455, + "grad_norm": 1.513053059577942, + "learning_rate": 2.0861263980888502e-07, + "loss": 0.4938, + "step": 21263 + }, + { + "epoch": 0.9414785871622594, + "grad_norm": 1.5847097635269165, + "learning_rate": 2.0829873215540308e-07, + "loss": 0.5187, + "step": 21264 + }, + { + "epoch": 0.9415228628670733, + "grad_norm": 1.8578579425811768, + "learning_rate": 2.0798505836873262e-07, + "loss": 0.5016, + "step": 21265 + }, + { + "epoch": 0.9415671385718871, + "grad_norm": 1.394801378250122, + "learning_rate": 2.0767161845636542e-07, + "loss": 0.5122, + "step": 21266 + }, + { + "epoch": 0.941611414276701, + "grad_norm": 3.702214002609253, + "learning_rate": 2.0735841242578992e-07, + "loss": 0.5085, + "step": 21267 + }, + { + "epoch": 0.9416556899815149, + "grad_norm": 1.2191556692123413, + "learning_rate": 2.0704544028448347e-07, + "loss": 0.4083, + "step": 21268 + }, + { + "epoch": 0.9416999656863287, + "grad_norm": 1.308505654335022, + "learning_rate": 2.0673270203992345e-07, + "loss": 0.4291, + "step": 21269 + }, + { + "epoch": 0.9417442413911427, + "grad_norm": 1.7701963186264038, + "learning_rate": 2.0642019769957943e-07, + "loss": 0.4973, + "step": 21270 + }, + { + "epoch": 0.9417885170959566, + "grad_norm": 1.4835381507873535, + "learning_rate": 2.0610792727091434e-07, + "loss": 0.4734, + "step": 21271 + }, + { + "epoch": 0.9418327928007704, + "grad_norm": 1.4420737028121948, + "learning_rate": 2.057958907613866e-07, + "loss": 0.4154, + "step": 21272 + }, + { + "epoch": 0.9418770685055843, + "grad_norm": 1.776831030845642, + "learning_rate": 2.0548408817844924e-07, + "loss": 0.5472, + "step": 21273 + }, + { + "epoch": 0.9419213442103982, + "grad_norm": 1.5180456638336182, + "learning_rate": 2.0517251952955063e-07, + "loss": 0.4715, + "step": 21274 + }, + { + "epoch": 0.941965619915212, + "grad_norm": 1.5769805908203125, + "learning_rate": 2.0486118482213047e-07, + "loss": 0.4526, + "step": 21275 + }, + { + "epoch": 0.9420098956200259, + "grad_norm": 1.7806544303894043, + "learning_rate": 2.04550084063625e-07, + "loss": 0.4919, + "step": 21276 + }, + { + "epoch": 0.9420541713248398, + "grad_norm": 1.1881217956542969, + "learning_rate": 2.04239217261466e-07, + "loss": 0.4203, + "step": 21277 + }, + { + "epoch": 0.9420984470296536, + "grad_norm": 1.6566816568374634, + "learning_rate": 2.039285844230765e-07, + "loss": 0.4833, + "step": 21278 + }, + { + "epoch": 0.9421427227344675, + "grad_norm": 1.3690272569656372, + "learning_rate": 2.036181855558761e-07, + "loss": 0.5065, + "step": 21279 + }, + { + "epoch": 0.9421869984392814, + "grad_norm": 1.7245951890945435, + "learning_rate": 2.0330802066727996e-07, + "loss": 0.5114, + "step": 21280 + }, + { + "epoch": 0.9422312741440952, + "grad_norm": 1.3415989875793457, + "learning_rate": 2.0299808976469436e-07, + "loss": 0.4472, + "step": 21281 + }, + { + "epoch": 0.9422755498489092, + "grad_norm": 2.1596455574035645, + "learning_rate": 2.026883928555212e-07, + "loss": 0.5579, + "step": 21282 + }, + { + "epoch": 0.9423198255537231, + "grad_norm": 1.63217031955719, + "learning_rate": 2.0237892994716013e-07, + "loss": 0.503, + "step": 21283 + }, + { + "epoch": 0.9423641012585369, + "grad_norm": 1.8817354440689087, + "learning_rate": 2.0206970104700076e-07, + "loss": 0.5046, + "step": 21284 + }, + { + "epoch": 0.9424083769633508, + "grad_norm": 1.4981393814086914, + "learning_rate": 2.0176070616242828e-07, + "loss": 0.4864, + "step": 21285 + }, + { + "epoch": 0.9424526526681647, + "grad_norm": 1.5969849824905396, + "learning_rate": 2.0145194530082458e-07, + "loss": 0.5343, + "step": 21286 + }, + { + "epoch": 0.9424969283729785, + "grad_norm": 1.674479365348816, + "learning_rate": 2.0114341846956153e-07, + "loss": 0.4258, + "step": 21287 + }, + { + "epoch": 0.9425412040777924, + "grad_norm": 1.396572470664978, + "learning_rate": 2.0083512567600994e-07, + "loss": 0.472, + "step": 21288 + }, + { + "epoch": 0.9425854797826063, + "grad_norm": 1.3320294618606567, + "learning_rate": 2.0052706692753166e-07, + "loss": 0.4119, + "step": 21289 + }, + { + "epoch": 0.9426297554874201, + "grad_norm": 1.3328301906585693, + "learning_rate": 2.0021924223148635e-07, + "loss": 0.4342, + "step": 21290 + }, + { + "epoch": 0.942674031192234, + "grad_norm": 1.8158767223358154, + "learning_rate": 1.9991165159522485e-07, + "loss": 0.5207, + "step": 21291 + }, + { + "epoch": 0.9427183068970479, + "grad_norm": 1.4835684299468994, + "learning_rate": 1.9960429502609346e-07, + "loss": 0.501, + "step": 21292 + }, + { + "epoch": 0.9427625826018617, + "grad_norm": 1.4781149625778198, + "learning_rate": 1.9929717253143522e-07, + "loss": 0.4518, + "step": 21293 + }, + { + "epoch": 0.9428068583066757, + "grad_norm": 2.568758726119995, + "learning_rate": 1.989902841185831e-07, + "loss": 0.5405, + "step": 21294 + }, + { + "epoch": 0.9428511340114896, + "grad_norm": 1.8885658979415894, + "learning_rate": 1.9868362979486687e-07, + "loss": 0.4731, + "step": 21295 + }, + { + "epoch": 0.9428954097163034, + "grad_norm": 1.169055700302124, + "learning_rate": 1.9837720956761287e-07, + "loss": 0.4441, + "step": 21296 + }, + { + "epoch": 0.9429396854211173, + "grad_norm": 1.358406901359558, + "learning_rate": 1.9807102344413742e-07, + "loss": 0.4231, + "step": 21297 + }, + { + "epoch": 0.9429839611259312, + "grad_norm": 1.3241286277770996, + "learning_rate": 1.9776507143175582e-07, + "loss": 0.4256, + "step": 21298 + }, + { + "epoch": 0.9430282368307451, + "grad_norm": 2.1692075729370117, + "learning_rate": 1.9745935353777222e-07, + "loss": 0.4799, + "step": 21299 + }, + { + "epoch": 0.9430725125355589, + "grad_norm": 1.3386216163635254, + "learning_rate": 1.9715386976949192e-07, + "loss": 0.453, + "step": 21300 + }, + { + "epoch": 0.9431167882403728, + "grad_norm": 2.2179152965545654, + "learning_rate": 1.9684862013421014e-07, + "loss": 0.5788, + "step": 21301 + }, + { + "epoch": 0.9431610639451867, + "grad_norm": 1.5198553800582886, + "learning_rate": 1.9654360463921663e-07, + "loss": 0.4803, + "step": 21302 + }, + { + "epoch": 0.9432053396500005, + "grad_norm": 2.144350290298462, + "learning_rate": 1.9623882329179666e-07, + "loss": 0.4605, + "step": 21303 + }, + { + "epoch": 0.9432496153548144, + "grad_norm": 1.6064212322235107, + "learning_rate": 1.9593427609923e-07, + "loss": 0.4493, + "step": 21304 + }, + { + "epoch": 0.9432938910596284, + "grad_norm": 1.3990126848220825, + "learning_rate": 1.9562996306878967e-07, + "loss": 0.4516, + "step": 21305 + }, + { + "epoch": 0.9433381667644422, + "grad_norm": 1.688131332397461, + "learning_rate": 1.9532588420774546e-07, + "loss": 0.5022, + "step": 21306 + }, + { + "epoch": 0.9433824424692561, + "grad_norm": 1.4578734636306763, + "learning_rate": 1.9502203952335928e-07, + "loss": 0.4537, + "step": 21307 + }, + { + "epoch": 0.94342671817407, + "grad_norm": 1.6176036596298218, + "learning_rate": 1.9471842902288763e-07, + "loss": 0.4909, + "step": 21308 + }, + { + "epoch": 0.9434709938788838, + "grad_norm": 1.6596013307571411, + "learning_rate": 1.9441505271358242e-07, + "loss": 0.4539, + "step": 21309 + }, + { + "epoch": 0.9435152695836977, + "grad_norm": 1.7004505395889282, + "learning_rate": 1.9411191060269007e-07, + "loss": 0.5004, + "step": 21310 + }, + { + "epoch": 0.9435595452885116, + "grad_norm": 1.576913595199585, + "learning_rate": 1.938090026974504e-07, + "loss": 0.5195, + "step": 21311 + }, + { + "epoch": 0.9436038209933254, + "grad_norm": 1.4195220470428467, + "learning_rate": 1.935063290050987e-07, + "loss": 0.4469, + "step": 21312 + }, + { + "epoch": 0.9436480966981393, + "grad_norm": 1.4299708604812622, + "learning_rate": 1.9320388953286362e-07, + "loss": 0.471, + "step": 21313 + }, + { + "epoch": 0.9436923724029532, + "grad_norm": 1.292658805847168, + "learning_rate": 1.929016842879683e-07, + "loss": 0.4085, + "step": 21314 + }, + { + "epoch": 0.943736648107767, + "grad_norm": 1.559471845626831, + "learning_rate": 1.9259971327763028e-07, + "loss": 0.4673, + "step": 21315 + }, + { + "epoch": 0.9437809238125809, + "grad_norm": 1.2914750576019287, + "learning_rate": 1.9229797650906268e-07, + "loss": 0.4369, + "step": 21316 + }, + { + "epoch": 0.9438251995173949, + "grad_norm": 1.548850417137146, + "learning_rate": 1.9199647398947307e-07, + "loss": 0.4931, + "step": 21317 + }, + { + "epoch": 0.9438694752222087, + "grad_norm": 1.462146282196045, + "learning_rate": 1.9169520572606126e-07, + "loss": 0.5244, + "step": 21318 + }, + { + "epoch": 0.9439137509270226, + "grad_norm": 1.9666310548782349, + "learning_rate": 1.913941717260226e-07, + "loss": 0.5716, + "step": 21319 + }, + { + "epoch": 0.9439580266318365, + "grad_norm": 1.532523512840271, + "learning_rate": 1.9109337199654908e-07, + "loss": 0.4936, + "step": 21320 + }, + { + "epoch": 0.9440023023366503, + "grad_norm": 1.6287994384765625, + "learning_rate": 1.907928065448228e-07, + "loss": 0.5133, + "step": 21321 + }, + { + "epoch": 0.9440465780414642, + "grad_norm": 1.571995496749878, + "learning_rate": 1.9049247537802462e-07, + "loss": 0.444, + "step": 21322 + }, + { + "epoch": 0.9440908537462781, + "grad_norm": 1.3781534433364868, + "learning_rate": 1.9019237850332551e-07, + "loss": 0.4603, + "step": 21323 + }, + { + "epoch": 0.9441351294510919, + "grad_norm": 1.4171364307403564, + "learning_rate": 1.8989251592789416e-07, + "loss": 0.4676, + "step": 21324 + }, + { + "epoch": 0.9441794051559058, + "grad_norm": 1.4612400531768799, + "learning_rate": 1.895928876588915e-07, + "loss": 0.4284, + "step": 21325 + }, + { + "epoch": 0.9442236808607197, + "grad_norm": 1.8048715591430664, + "learning_rate": 1.8929349370347517e-07, + "loss": 0.4586, + "step": 21326 + }, + { + "epoch": 0.9442679565655335, + "grad_norm": 1.6193270683288574, + "learning_rate": 1.889943340687961e-07, + "loss": 0.462, + "step": 21327 + }, + { + "epoch": 0.9443122322703474, + "grad_norm": 1.573738932609558, + "learning_rate": 1.8869540876199855e-07, + "loss": 0.4591, + "step": 21328 + }, + { + "epoch": 0.9443565079751614, + "grad_norm": 1.6277655363082886, + "learning_rate": 1.883967177902224e-07, + "loss": 0.4709, + "step": 21329 + }, + { + "epoch": 0.9444007836799752, + "grad_norm": 1.626146674156189, + "learning_rate": 1.8809826116060194e-07, + "loss": 0.4904, + "step": 21330 + }, + { + "epoch": 0.9444450593847891, + "grad_norm": 1.3009587526321411, + "learning_rate": 1.8780003888026587e-07, + "loss": 0.4494, + "step": 21331 + }, + { + "epoch": 0.944489335089603, + "grad_norm": 1.4361810684204102, + "learning_rate": 1.8750205095633633e-07, + "loss": 0.4235, + "step": 21332 + }, + { + "epoch": 0.9445336107944169, + "grad_norm": 1.8556188344955444, + "learning_rate": 1.8720429739592983e-07, + "loss": 0.4899, + "step": 21333 + }, + { + "epoch": 0.9445778864992307, + "grad_norm": 1.1823889017105103, + "learning_rate": 1.8690677820615954e-07, + "loss": 0.4069, + "step": 21334 + }, + { + "epoch": 0.9446221622040446, + "grad_norm": 1.5188251733779907, + "learning_rate": 1.8660949339413093e-07, + "loss": 0.4934, + "step": 21335 + }, + { + "epoch": 0.9446664379088585, + "grad_norm": 1.3565998077392578, + "learning_rate": 1.8631244296694385e-07, + "loss": 0.4288, + "step": 21336 + }, + { + "epoch": 0.9447107136136723, + "grad_norm": 1.3226454257965088, + "learning_rate": 1.8601562693169371e-07, + "loss": 0.4219, + "step": 21337 + }, + { + "epoch": 0.9447549893184862, + "grad_norm": 1.5479997396469116, + "learning_rate": 1.8571904529546937e-07, + "loss": 0.493, + "step": 21338 + }, + { + "epoch": 0.9447992650233001, + "grad_norm": 1.728623390197754, + "learning_rate": 1.8542269806535285e-07, + "loss": 0.5512, + "step": 21339 + }, + { + "epoch": 0.9448435407281139, + "grad_norm": 2.6776716709136963, + "learning_rate": 1.851265852484263e-07, + "loss": 0.4974, + "step": 21340 + }, + { + "epoch": 0.9448878164329279, + "grad_norm": 1.4196317195892334, + "learning_rate": 1.8483070685175964e-07, + "loss": 0.413, + "step": 21341 + }, + { + "epoch": 0.9449320921377418, + "grad_norm": 1.4962379932403564, + "learning_rate": 1.845350628824183e-07, + "loss": 0.5716, + "step": 21342 + }, + { + "epoch": 0.9449763678425556, + "grad_norm": 1.5827769041061401, + "learning_rate": 1.8423965334746551e-07, + "loss": 0.5214, + "step": 21343 + }, + { + "epoch": 0.9450206435473695, + "grad_norm": 1.79180109500885, + "learning_rate": 1.839444782539568e-07, + "loss": 0.503, + "step": 21344 + }, + { + "epoch": 0.9450649192521834, + "grad_norm": 1.2641191482543945, + "learning_rate": 1.836495376089398e-07, + "loss": 0.4303, + "step": 21345 + }, + { + "epoch": 0.9451091949569972, + "grad_norm": 1.636248230934143, + "learning_rate": 1.833548314194633e-07, + "loss": 0.5264, + "step": 21346 + }, + { + "epoch": 0.9451534706618111, + "grad_norm": 1.8956724405288696, + "learning_rate": 1.8306035969256286e-07, + "loss": 0.5172, + "step": 21347 + }, + { + "epoch": 0.945197746366625, + "grad_norm": 1.5691214799880981, + "learning_rate": 1.827661224352717e-07, + "loss": 0.4835, + "step": 21348 + }, + { + "epoch": 0.9452420220714388, + "grad_norm": 1.300585150718689, + "learning_rate": 1.8247211965461864e-07, + "loss": 0.4233, + "step": 21349 + }, + { + "epoch": 0.9452862977762527, + "grad_norm": 1.359911561012268, + "learning_rate": 1.8217835135762584e-07, + "loss": 0.4631, + "step": 21350 + }, + { + "epoch": 0.9453305734810666, + "grad_norm": 1.5831291675567627, + "learning_rate": 1.8188481755130883e-07, + "loss": 0.4748, + "step": 21351 + }, + { + "epoch": 0.9453748491858804, + "grad_norm": 1.416378140449524, + "learning_rate": 1.8159151824267862e-07, + "loss": 0.4336, + "step": 21352 + }, + { + "epoch": 0.9454191248906944, + "grad_norm": 1.469243049621582, + "learning_rate": 1.8129845343873963e-07, + "loss": 0.4523, + "step": 21353 + }, + { + "epoch": 0.9454634005955083, + "grad_norm": 1.6823214292526245, + "learning_rate": 1.8100562314649407e-07, + "loss": 0.5044, + "step": 21354 + }, + { + "epoch": 0.9455076763003221, + "grad_norm": 1.4450883865356445, + "learning_rate": 1.8071302737293294e-07, + "loss": 0.4908, + "step": 21355 + }, + { + "epoch": 0.945551952005136, + "grad_norm": 1.6852248907089233, + "learning_rate": 1.8042066612504626e-07, + "loss": 0.4853, + "step": 21356 + }, + { + "epoch": 0.9455962277099499, + "grad_norm": 1.3641711473464966, + "learning_rate": 1.8012853940981734e-07, + "loss": 0.4663, + "step": 21357 + }, + { + "epoch": 0.9456405034147637, + "grad_norm": 1.6373906135559082, + "learning_rate": 1.7983664723422278e-07, + "loss": 0.508, + "step": 21358 + }, + { + "epoch": 0.9456847791195776, + "grad_norm": 2.2243454456329346, + "learning_rate": 1.795449896052326e-07, + "loss": 0.5177, + "step": 21359 + }, + { + "epoch": 0.9457290548243915, + "grad_norm": 1.6670467853546143, + "learning_rate": 1.7925356652981563e-07, + "loss": 0.4674, + "step": 21360 + }, + { + "epoch": 0.9457733305292053, + "grad_norm": 1.5175668001174927, + "learning_rate": 1.7896237801493076e-07, + "loss": 0.4938, + "step": 21361 + }, + { + "epoch": 0.9458176062340192, + "grad_norm": 1.4307175874710083, + "learning_rate": 1.7867142406753134e-07, + "loss": 0.4502, + "step": 21362 + }, + { + "epoch": 0.9458618819388331, + "grad_norm": 1.5951918363571167, + "learning_rate": 1.7838070469456958e-07, + "loss": 0.4581, + "step": 21363 + }, + { + "epoch": 0.9459061576436469, + "grad_norm": 2.066011667251587, + "learning_rate": 1.7809021990298768e-07, + "loss": 0.5553, + "step": 21364 + }, + { + "epoch": 0.9459504333484609, + "grad_norm": 1.5352234840393066, + "learning_rate": 1.7779996969972236e-07, + "loss": 0.4644, + "step": 21365 + }, + { + "epoch": 0.9459947090532748, + "grad_norm": 1.2737373113632202, + "learning_rate": 1.7750995409170802e-07, + "loss": 0.4332, + "step": 21366 + }, + { + "epoch": 0.9460389847580886, + "grad_norm": 1.2182531356811523, + "learning_rate": 1.7722017308587025e-07, + "loss": 0.4183, + "step": 21367 + }, + { + "epoch": 0.9460832604629025, + "grad_norm": 1.7348932027816772, + "learning_rate": 1.769306266891313e-07, + "loss": 0.5199, + "step": 21368 + }, + { + "epoch": 0.9461275361677164, + "grad_norm": 1.7412397861480713, + "learning_rate": 1.7664131490840676e-07, + "loss": 0.5755, + "step": 21369 + }, + { + "epoch": 0.9461718118725303, + "grad_norm": 1.370314121246338, + "learning_rate": 1.763522377506044e-07, + "loss": 0.491, + "step": 21370 + }, + { + "epoch": 0.9462160875773441, + "grad_norm": 1.4549330472946167, + "learning_rate": 1.7606339522263095e-07, + "loss": 0.4918, + "step": 21371 + }, + { + "epoch": 0.946260363282158, + "grad_norm": 1.6794344186782837, + "learning_rate": 1.757747873313842e-07, + "loss": 0.5653, + "step": 21372 + }, + { + "epoch": 0.9463046389869719, + "grad_norm": 1.3178536891937256, + "learning_rate": 1.7548641408375756e-07, + "loss": 0.4345, + "step": 21373 + }, + { + "epoch": 0.9463489146917857, + "grad_norm": 1.336991310119629, + "learning_rate": 1.7519827548663882e-07, + "loss": 0.4231, + "step": 21374 + }, + { + "epoch": 0.9463931903965996, + "grad_norm": 1.3463793992996216, + "learning_rate": 1.7491037154690915e-07, + "loss": 0.4594, + "step": 21375 + }, + { + "epoch": 0.9464374661014135, + "grad_norm": 1.4764771461486816, + "learning_rate": 1.7462270227144417e-07, + "loss": 0.4744, + "step": 21376 + }, + { + "epoch": 0.9464817418062274, + "grad_norm": 1.4049291610717773, + "learning_rate": 1.7433526766711727e-07, + "loss": 0.4672, + "step": 21377 + }, + { + "epoch": 0.9465260175110413, + "grad_norm": 1.743544578552246, + "learning_rate": 1.7404806774079297e-07, + "loss": 0.4653, + "step": 21378 + }, + { + "epoch": 0.9465702932158552, + "grad_norm": 3.6571033000946045, + "learning_rate": 1.737611024993291e-07, + "loss": 0.5896, + "step": 21379 + }, + { + "epoch": 0.946614568920669, + "grad_norm": 1.6466602087020874, + "learning_rate": 1.7347437194958016e-07, + "loss": 0.5116, + "step": 21380 + }, + { + "epoch": 0.9466588446254829, + "grad_norm": 1.367775797843933, + "learning_rate": 1.7318787609839516e-07, + "loss": 0.417, + "step": 21381 + }, + { + "epoch": 0.9467031203302968, + "grad_norm": 1.2401219606399536, + "learning_rate": 1.7290161495261526e-07, + "loss": 0.4021, + "step": 21382 + }, + { + "epoch": 0.9467473960351106, + "grad_norm": 1.8319556713104248, + "learning_rate": 1.7261558851908056e-07, + "loss": 0.5567, + "step": 21383 + }, + { + "epoch": 0.9467916717399245, + "grad_norm": 1.4158105850219727, + "learning_rate": 1.7232979680462003e-07, + "loss": 0.4259, + "step": 21384 + }, + { + "epoch": 0.9468359474447384, + "grad_norm": 1.519659161567688, + "learning_rate": 1.7204423981606044e-07, + "loss": 0.5002, + "step": 21385 + }, + { + "epoch": 0.9468802231495522, + "grad_norm": 1.4970351457595825, + "learning_rate": 1.7175891756022078e-07, + "loss": 0.4898, + "step": 21386 + }, + { + "epoch": 0.9469244988543661, + "grad_norm": 1.6139229536056519, + "learning_rate": 1.7147383004391782e-07, + "loss": 0.4825, + "step": 21387 + }, + { + "epoch": 0.94696877455918, + "grad_norm": 1.2535611391067505, + "learning_rate": 1.7118897727396056e-07, + "loss": 0.4669, + "step": 21388 + }, + { + "epoch": 0.9470130502639938, + "grad_norm": 1.6777209043502808, + "learning_rate": 1.7090435925715133e-07, + "loss": 0.4777, + "step": 21389 + }, + { + "epoch": 0.9470573259688078, + "grad_norm": 1.3893027305603027, + "learning_rate": 1.70619976000288e-07, + "loss": 0.4776, + "step": 21390 + }, + { + "epoch": 0.9471016016736217, + "grad_norm": 1.7813892364501953, + "learning_rate": 1.7033582751016297e-07, + "loss": 0.5183, + "step": 21391 + }, + { + "epoch": 0.9471458773784355, + "grad_norm": 1.2612454891204834, + "learning_rate": 1.70051913793563e-07, + "loss": 0.4262, + "step": 21392 + }, + { + "epoch": 0.9471901530832494, + "grad_norm": 1.320387840270996, + "learning_rate": 1.6976823485726933e-07, + "loss": 0.4189, + "step": 21393 + }, + { + "epoch": 0.9472344287880633, + "grad_norm": 1.3435993194580078, + "learning_rate": 1.6948479070805768e-07, + "loss": 0.4087, + "step": 21394 + }, + { + "epoch": 0.9472787044928771, + "grad_norm": 1.4215779304504395, + "learning_rate": 1.6920158135269705e-07, + "loss": 0.4362, + "step": 21395 + }, + { + "epoch": 0.947322980197691, + "grad_norm": 1.501251220703125, + "learning_rate": 1.6891860679795203e-07, + "loss": 0.4692, + "step": 21396 + }, + { + "epoch": 0.9473672559025049, + "grad_norm": 1.5844534635543823, + "learning_rate": 1.6863586705058277e-07, + "loss": 0.4544, + "step": 21397 + }, + { + "epoch": 0.9474115316073187, + "grad_norm": 1.4662519693374634, + "learning_rate": 1.6835336211733944e-07, + "loss": 0.4979, + "step": 21398 + }, + { + "epoch": 0.9474558073121326, + "grad_norm": 1.4855650663375854, + "learning_rate": 1.6807109200496995e-07, + "loss": 0.4576, + "step": 21399 + }, + { + "epoch": 0.9475000830169465, + "grad_norm": 1.439170002937317, + "learning_rate": 1.6778905672021895e-07, + "loss": 0.478, + "step": 21400 + }, + { + "epoch": 0.9475443587217603, + "grad_norm": 1.3128749132156372, + "learning_rate": 1.675072562698199e-07, + "loss": 0.424, + "step": 21401 + }, + { + "epoch": 0.9475886344265743, + "grad_norm": 1.984598994255066, + "learning_rate": 1.6722569066050298e-07, + "loss": 0.52, + "step": 21402 + }, + { + "epoch": 0.9476329101313882, + "grad_norm": 1.6880735158920288, + "learning_rate": 1.6694435989899616e-07, + "loss": 0.4926, + "step": 21403 + }, + { + "epoch": 0.9476771858362021, + "grad_norm": 1.6934404373168945, + "learning_rate": 1.6666326399201626e-07, + "loss": 0.4997, + "step": 21404 + }, + { + "epoch": 0.9477214615410159, + "grad_norm": 1.4032715559005737, + "learning_rate": 1.6638240294627682e-07, + "loss": 0.4401, + "step": 21405 + }, + { + "epoch": 0.9477657372458298, + "grad_norm": 1.4708313941955566, + "learning_rate": 1.6610177676848805e-07, + "loss": 0.4662, + "step": 21406 + }, + { + "epoch": 0.9478100129506437, + "grad_norm": 1.9968323707580566, + "learning_rate": 1.658213854653512e-07, + "loss": 0.5564, + "step": 21407 + }, + { + "epoch": 0.9478542886554575, + "grad_norm": 1.6258097887039185, + "learning_rate": 1.6554122904356317e-07, + "loss": 0.4882, + "step": 21408 + }, + { + "epoch": 0.9478985643602714, + "grad_norm": 1.3899422883987427, + "learning_rate": 1.6526130750981416e-07, + "loss": 0.5101, + "step": 21409 + }, + { + "epoch": 0.9479428400650853, + "grad_norm": 1.58021879196167, + "learning_rate": 1.6498162087079218e-07, + "loss": 0.5031, + "step": 21410 + }, + { + "epoch": 0.9479871157698991, + "grad_norm": 1.5555633306503296, + "learning_rate": 1.6470216913317628e-07, + "loss": 0.5305, + "step": 21411 + }, + { + "epoch": 0.948031391474713, + "grad_norm": 1.3503385782241821, + "learning_rate": 1.6442295230364003e-07, + "loss": 0.4624, + "step": 21412 + }, + { + "epoch": 0.948075667179527, + "grad_norm": 1.3616359233856201, + "learning_rate": 1.6414397038885365e-07, + "loss": 0.4326, + "step": 21413 + }, + { + "epoch": 0.9481199428843408, + "grad_norm": 1.560356855392456, + "learning_rate": 1.6386522339548072e-07, + "loss": 0.4894, + "step": 21414 + }, + { + "epoch": 0.9481642185891547, + "grad_norm": 1.778071403503418, + "learning_rate": 1.63586711330177e-07, + "loss": 0.5221, + "step": 21415 + }, + { + "epoch": 0.9482084942939686, + "grad_norm": 1.4920403957366943, + "learning_rate": 1.6330843419959607e-07, + "loss": 0.4873, + "step": 21416 + }, + { + "epoch": 0.9482527699987824, + "grad_norm": 1.3225945234298706, + "learning_rate": 1.6303039201038374e-07, + "loss": 0.4338, + "step": 21417 + }, + { + "epoch": 0.9482970457035963, + "grad_norm": 1.9177385568618774, + "learning_rate": 1.6275258476918133e-07, + "loss": 0.4838, + "step": 21418 + }, + { + "epoch": 0.9483413214084102, + "grad_norm": 1.6729673147201538, + "learning_rate": 1.6247501248262243e-07, + "loss": 0.472, + "step": 21419 + }, + { + "epoch": 0.948385597113224, + "grad_norm": 1.5081353187561035, + "learning_rate": 1.6219767515733843e-07, + "loss": 0.4771, + "step": 21420 + }, + { + "epoch": 0.9484298728180379, + "grad_norm": 1.444347620010376, + "learning_rate": 1.61920572799954e-07, + "loss": 0.4737, + "step": 21421 + }, + { + "epoch": 0.9484741485228518, + "grad_norm": 1.704270362854004, + "learning_rate": 1.6164370541708384e-07, + "loss": 0.4789, + "step": 21422 + }, + { + "epoch": 0.9485184242276656, + "grad_norm": 1.3872761726379395, + "learning_rate": 1.6136707301534492e-07, + "loss": 0.4812, + "step": 21423 + }, + { + "epoch": 0.9485626999324795, + "grad_norm": 1.432213544845581, + "learning_rate": 1.6109067560134194e-07, + "loss": 0.4633, + "step": 21424 + }, + { + "epoch": 0.9486069756372935, + "grad_norm": 1.849789023399353, + "learning_rate": 1.608145131816774e-07, + "loss": 0.4998, + "step": 21425 + }, + { + "epoch": 0.9486512513421073, + "grad_norm": 1.6089483499526978, + "learning_rate": 1.60538585762946e-07, + "loss": 0.5259, + "step": 21426 + }, + { + "epoch": 0.9486955270469212, + "grad_norm": 1.784600019454956, + "learning_rate": 1.6026289335174027e-07, + "loss": 0.5114, + "step": 21427 + }, + { + "epoch": 0.9487398027517351, + "grad_norm": 1.7133848667144775, + "learning_rate": 1.5998743595464272e-07, + "loss": 0.5354, + "step": 21428 + }, + { + "epoch": 0.9487840784565489, + "grad_norm": 1.5746047496795654, + "learning_rate": 1.5971221357823252e-07, + "loss": 0.448, + "step": 21429 + }, + { + "epoch": 0.9488283541613628, + "grad_norm": 1.4000576734542847, + "learning_rate": 1.5943722622908443e-07, + "loss": 0.4602, + "step": 21430 + }, + { + "epoch": 0.9488726298661767, + "grad_norm": 1.5687059164047241, + "learning_rate": 1.591624739137665e-07, + "loss": 0.5017, + "step": 21431 + }, + { + "epoch": 0.9489169055709905, + "grad_norm": 1.5324652194976807, + "learning_rate": 1.5888795663883904e-07, + "loss": 0.4799, + "step": 21432 + }, + { + "epoch": 0.9489611812758044, + "grad_norm": 1.3392934799194336, + "learning_rate": 1.5861367441086016e-07, + "loss": 0.4365, + "step": 21433 + }, + { + "epoch": 0.9490054569806183, + "grad_norm": 1.3015217781066895, + "learning_rate": 1.5833962723638018e-07, + "loss": 0.4775, + "step": 21434 + }, + { + "epoch": 0.9490497326854321, + "grad_norm": 1.4735076427459717, + "learning_rate": 1.580658151219461e-07, + "loss": 0.5352, + "step": 21435 + }, + { + "epoch": 0.949094008390246, + "grad_norm": 1.5485275983810425, + "learning_rate": 1.5779223807409482e-07, + "loss": 0.5086, + "step": 21436 + }, + { + "epoch": 0.94913828409506, + "grad_norm": 1.4424357414245605, + "learning_rate": 1.5751889609936344e-07, + "loss": 0.4505, + "step": 21437 + }, + { + "epoch": 0.9491825597998738, + "grad_norm": 1.706527829170227, + "learning_rate": 1.5724578920427892e-07, + "loss": 0.4682, + "step": 21438 + }, + { + "epoch": 0.9492268355046877, + "grad_norm": 1.4872233867645264, + "learning_rate": 1.569729173953638e-07, + "loss": 0.4716, + "step": 21439 + }, + { + "epoch": 0.9492711112095016, + "grad_norm": 1.5728106498718262, + "learning_rate": 1.5670028067913622e-07, + "loss": 0.4478, + "step": 21440 + }, + { + "epoch": 0.9493153869143155, + "grad_norm": 1.6146519184112549, + "learning_rate": 1.5642787906210765e-07, + "loss": 0.5075, + "step": 21441 + }, + { + "epoch": 0.9493596626191293, + "grad_norm": 1.542486310005188, + "learning_rate": 1.56155712550784e-07, + "loss": 0.4139, + "step": 21442 + }, + { + "epoch": 0.9494039383239432, + "grad_norm": 1.9072976112365723, + "learning_rate": 1.558837811516667e-07, + "loss": 0.5332, + "step": 21443 + }, + { + "epoch": 0.9494482140287571, + "grad_norm": 1.5056523084640503, + "learning_rate": 1.5561208487125056e-07, + "loss": 0.4857, + "step": 21444 + }, + { + "epoch": 0.9494924897335709, + "grad_norm": 1.798438310623169, + "learning_rate": 1.5534062371602376e-07, + "loss": 0.5196, + "step": 21445 + }, + { + "epoch": 0.9495367654383848, + "grad_norm": 1.5217477083206177, + "learning_rate": 1.5506939769246887e-07, + "loss": 0.4535, + "step": 21446 + }, + { + "epoch": 0.9495810411431987, + "grad_norm": 1.775794506072998, + "learning_rate": 1.5479840680706736e-07, + "loss": 0.518, + "step": 21447 + }, + { + "epoch": 0.9496253168480125, + "grad_norm": 1.8612512350082397, + "learning_rate": 1.5452765106628853e-07, + "loss": 0.471, + "step": 21448 + }, + { + "epoch": 0.9496695925528265, + "grad_norm": 1.5966174602508545, + "learning_rate": 1.542571304765994e-07, + "loss": 0.4821, + "step": 21449 + }, + { + "epoch": 0.9497138682576404, + "grad_norm": 1.588012933731079, + "learning_rate": 1.539868450444637e-07, + "loss": 0.5442, + "step": 21450 + }, + { + "epoch": 0.9497581439624542, + "grad_norm": 1.4615236520767212, + "learning_rate": 1.5371679477633406e-07, + "loss": 0.4895, + "step": 21451 + }, + { + "epoch": 0.9498024196672681, + "grad_norm": 1.9007060527801514, + "learning_rate": 1.5344697967866195e-07, + "loss": 0.612, + "step": 21452 + }, + { + "epoch": 0.949846695372082, + "grad_norm": 1.596764087677002, + "learning_rate": 1.5317739975789225e-07, + "loss": 0.4696, + "step": 21453 + }, + { + "epoch": 0.9498909710768958, + "grad_norm": 1.6610432863235474, + "learning_rate": 1.5290805502046312e-07, + "loss": 0.4974, + "step": 21454 + }, + { + "epoch": 0.9499352467817097, + "grad_norm": 2.128275156021118, + "learning_rate": 1.5263894547280612e-07, + "loss": 0.5619, + "step": 21455 + }, + { + "epoch": 0.9499795224865236, + "grad_norm": 1.505476951599121, + "learning_rate": 1.5237007112134938e-07, + "loss": 0.5043, + "step": 21456 + }, + { + "epoch": 0.9500237981913374, + "grad_norm": 1.7614374160766602, + "learning_rate": 1.521014319725167e-07, + "loss": 0.5341, + "step": 21457 + }, + { + "epoch": 0.9500680738961513, + "grad_norm": 1.3448832035064697, + "learning_rate": 1.518330280327218e-07, + "loss": 0.4719, + "step": 21458 + }, + { + "epoch": 0.9501123496009652, + "grad_norm": 1.6353869438171387, + "learning_rate": 1.5156485930837628e-07, + "loss": 0.4766, + "step": 21459 + }, + { + "epoch": 0.950156625305779, + "grad_norm": 1.5646289587020874, + "learning_rate": 1.5129692580588607e-07, + "loss": 0.4853, + "step": 21460 + }, + { + "epoch": 0.950200901010593, + "grad_norm": 1.3186079263687134, + "learning_rate": 1.5102922753164827e-07, + "loss": 0.4473, + "step": 21461 + }, + { + "epoch": 0.9502451767154069, + "grad_norm": 1.5326296091079712, + "learning_rate": 1.5076176449206005e-07, + "loss": 0.4781, + "step": 21462 + }, + { + "epoch": 0.9502894524202207, + "grad_norm": 1.5771772861480713, + "learning_rate": 1.5049453669350622e-07, + "loss": 0.5081, + "step": 21463 + }, + { + "epoch": 0.9503337281250346, + "grad_norm": 1.4457077980041504, + "learning_rate": 1.5022754414237062e-07, + "loss": 0.5219, + "step": 21464 + }, + { + "epoch": 0.9503780038298485, + "grad_norm": 1.8320800065994263, + "learning_rate": 1.4996078684503145e-07, + "loss": 0.5252, + "step": 21465 + }, + { + "epoch": 0.9504222795346623, + "grad_norm": 1.5286304950714111, + "learning_rate": 1.4969426480785698e-07, + "loss": 0.4773, + "step": 21466 + }, + { + "epoch": 0.9504665552394762, + "grad_norm": 1.5718168020248413, + "learning_rate": 1.4942797803721543e-07, + "loss": 0.5039, + "step": 21467 + }, + { + "epoch": 0.9505108309442901, + "grad_norm": 1.760230541229248, + "learning_rate": 1.4916192653946503e-07, + "loss": 0.5718, + "step": 21468 + }, + { + "epoch": 0.9505551066491039, + "grad_norm": 1.3484916687011719, + "learning_rate": 1.4889611032096184e-07, + "loss": 0.4327, + "step": 21469 + }, + { + "epoch": 0.9505993823539178, + "grad_norm": 1.577343463897705, + "learning_rate": 1.486305293880541e-07, + "loss": 0.46, + "step": 21470 + }, + { + "epoch": 0.9506436580587317, + "grad_norm": 1.6861573457717896, + "learning_rate": 1.4836518374708454e-07, + "loss": 0.4821, + "step": 21471 + }, + { + "epoch": 0.9506879337635455, + "grad_norm": 1.3368496894836426, + "learning_rate": 1.4810007340439135e-07, + "loss": 0.4392, + "step": 21472 + }, + { + "epoch": 0.9507322094683595, + "grad_norm": 1.6832772493362427, + "learning_rate": 1.4783519836630734e-07, + "loss": 0.5083, + "step": 21473 + }, + { + "epoch": 0.9507764851731734, + "grad_norm": 2.078289031982422, + "learning_rate": 1.4757055863915627e-07, + "loss": 0.5486, + "step": 21474 + }, + { + "epoch": 0.9508207608779873, + "grad_norm": 1.3647979497909546, + "learning_rate": 1.4730615422926086e-07, + "loss": 0.4496, + "step": 21475 + }, + { + "epoch": 0.9508650365828011, + "grad_norm": 1.598586082458496, + "learning_rate": 1.4704198514293498e-07, + "loss": 0.4554, + "step": 21476 + }, + { + "epoch": 0.950909312287615, + "grad_norm": 1.3497264385223389, + "learning_rate": 1.4677805138648916e-07, + "loss": 0.4239, + "step": 21477 + }, + { + "epoch": 0.9509535879924289, + "grad_norm": 1.7196400165557861, + "learning_rate": 1.4651435296622606e-07, + "loss": 0.5016, + "step": 21478 + }, + { + "epoch": 0.9509978636972427, + "grad_norm": 1.7105374336242676, + "learning_rate": 1.4625088988844515e-07, + "loss": 0.4702, + "step": 21479 + }, + { + "epoch": 0.9510421394020566, + "grad_norm": 1.7867103815078735, + "learning_rate": 1.4598766215943806e-07, + "loss": 0.4782, + "step": 21480 + }, + { + "epoch": 0.9510864151068705, + "grad_norm": 1.46404230594635, + "learning_rate": 1.4572466978549305e-07, + "loss": 0.4925, + "step": 21481 + }, + { + "epoch": 0.9511306908116843, + "grad_norm": 1.4444565773010254, + "learning_rate": 1.4546191277289068e-07, + "loss": 0.4354, + "step": 21482 + }, + { + "epoch": 0.9511749665164982, + "grad_norm": 1.609993577003479, + "learning_rate": 1.4519939112790705e-07, + "loss": 0.5171, + "step": 21483 + }, + { + "epoch": 0.9512192422213122, + "grad_norm": 1.7734661102294922, + "learning_rate": 1.4493710485681046e-07, + "loss": 0.4631, + "step": 21484 + }, + { + "epoch": 0.951263517926126, + "grad_norm": 2.7917838096618652, + "learning_rate": 1.4467505396586812e-07, + "loss": 0.5211, + "step": 21485 + }, + { + "epoch": 0.9513077936309399, + "grad_norm": 3.1987462043762207, + "learning_rate": 1.4441323846133615e-07, + "loss": 0.5228, + "step": 21486 + }, + { + "epoch": 0.9513520693357538, + "grad_norm": 1.2926418781280518, + "learning_rate": 1.4415165834947064e-07, + "loss": 0.4124, + "step": 21487 + }, + { + "epoch": 0.9513963450405676, + "grad_norm": 1.4647700786590576, + "learning_rate": 1.4389031363651773e-07, + "loss": 0.4626, + "step": 21488 + }, + { + "epoch": 0.9514406207453815, + "grad_norm": 1.4669476747512817, + "learning_rate": 1.4362920432871797e-07, + "loss": 0.4212, + "step": 21489 + }, + { + "epoch": 0.9514848964501954, + "grad_norm": 1.2864558696746826, + "learning_rate": 1.433683304323108e-07, + "loss": 0.4326, + "step": 21490 + }, + { + "epoch": 0.9515291721550092, + "grad_norm": 1.5175127983093262, + "learning_rate": 1.431076919535257e-07, + "loss": 0.4834, + "step": 21491 + }, + { + "epoch": 0.9515734478598231, + "grad_norm": 1.7382481098175049, + "learning_rate": 1.4284728889858768e-07, + "loss": 0.5387, + "step": 21492 + }, + { + "epoch": 0.951617723564637, + "grad_norm": 1.2070564031600952, + "learning_rate": 1.425871212737151e-07, + "loss": 0.4406, + "step": 21493 + }, + { + "epoch": 0.9516619992694508, + "grad_norm": 1.5519529581069946, + "learning_rate": 1.42327189085123e-07, + "loss": 0.4914, + "step": 21494 + }, + { + "epoch": 0.9517062749742647, + "grad_norm": 1.6697726249694824, + "learning_rate": 1.4206749233902084e-07, + "loss": 0.4592, + "step": 21495 + }, + { + "epoch": 0.9517505506790787, + "grad_norm": 1.483321189880371, + "learning_rate": 1.418080310416081e-07, + "loss": 0.5111, + "step": 21496 + }, + { + "epoch": 0.9517948263838925, + "grad_norm": 1.7619181871414185, + "learning_rate": 1.4154880519908543e-07, + "loss": 0.5355, + "step": 21497 + }, + { + "epoch": 0.9518391020887064, + "grad_norm": 1.7438219785690308, + "learning_rate": 1.4128981481764115e-07, + "loss": 0.5062, + "step": 21498 + }, + { + "epoch": 0.9518833777935203, + "grad_norm": 2.0917718410491943, + "learning_rate": 1.4103105990346256e-07, + "loss": 0.538, + "step": 21499 + }, + { + "epoch": 0.9519276534983341, + "grad_norm": 1.4402655363082886, + "learning_rate": 1.4077254046273136e-07, + "loss": 0.4615, + "step": 21500 + }, + { + "epoch": 0.951971929203148, + "grad_norm": 1.3628299236297607, + "learning_rate": 1.4051425650161933e-07, + "loss": 0.4491, + "step": 21501 + }, + { + "epoch": 0.9520162049079619, + "grad_norm": 2.0395355224609375, + "learning_rate": 1.4025620802629592e-07, + "loss": 0.4794, + "step": 21502 + }, + { + "epoch": 0.9520604806127757, + "grad_norm": 1.7886669635772705, + "learning_rate": 1.3999839504292513e-07, + "loss": 0.4654, + "step": 21503 + }, + { + "epoch": 0.9521047563175896, + "grad_norm": 1.3276751041412354, + "learning_rate": 1.397408175576642e-07, + "loss": 0.4301, + "step": 21504 + }, + { + "epoch": 0.9521490320224035, + "grad_norm": 1.3997602462768555, + "learning_rate": 1.3948347557666496e-07, + "loss": 0.4478, + "step": 21505 + }, + { + "epoch": 0.9521933077272173, + "grad_norm": 2.329024076461792, + "learning_rate": 1.3922636910607467e-07, + "loss": 0.5568, + "step": 21506 + }, + { + "epoch": 0.9522375834320312, + "grad_norm": 1.5990198850631714, + "learning_rate": 1.3896949815203398e-07, + "loss": 0.4607, + "step": 21507 + }, + { + "epoch": 0.9522818591368452, + "grad_norm": 1.2655322551727295, + "learning_rate": 1.3871286272067684e-07, + "loss": 0.4308, + "step": 21508 + }, + { + "epoch": 0.952326134841659, + "grad_norm": 1.5719997882843018, + "learning_rate": 1.3845646281813508e-07, + "loss": 0.4945, + "step": 21509 + }, + { + "epoch": 0.9523704105464729, + "grad_norm": 1.3244928121566772, + "learning_rate": 1.3820029845052928e-07, + "loss": 0.4732, + "step": 21510 + }, + { + "epoch": 0.9524146862512868, + "grad_norm": 2.1848061084747314, + "learning_rate": 1.3794436962398127e-07, + "loss": 0.4855, + "step": 21511 + }, + { + "epoch": 0.9524589619561007, + "grad_norm": 1.5291918516159058, + "learning_rate": 1.376886763446017e-07, + "loss": 0.4564, + "step": 21512 + }, + { + "epoch": 0.9525032376609145, + "grad_norm": 2.24786639213562, + "learning_rate": 1.374332186184968e-07, + "loss": 0.6193, + "step": 21513 + }, + { + "epoch": 0.9525475133657284, + "grad_norm": 1.4874640703201294, + "learning_rate": 1.3717799645176944e-07, + "loss": 0.4693, + "step": 21514 + }, + { + "epoch": 0.9525917890705423, + "grad_norm": 1.6324241161346436, + "learning_rate": 1.3692300985051478e-07, + "loss": 0.5461, + "step": 21515 + }, + { + "epoch": 0.9526360647753561, + "grad_norm": 1.5343501567840576, + "learning_rate": 1.3666825882082347e-07, + "loss": 0.4358, + "step": 21516 + }, + { + "epoch": 0.95268034048017, + "grad_norm": 1.3927586078643799, + "learning_rate": 1.3641374336878066e-07, + "loss": 0.4467, + "step": 21517 + }, + { + "epoch": 0.9527246161849839, + "grad_norm": 1.4455986022949219, + "learning_rate": 1.361594635004637e-07, + "loss": 0.4878, + "step": 21518 + }, + { + "epoch": 0.9527688918897977, + "grad_norm": 1.4640510082244873, + "learning_rate": 1.3590541922194668e-07, + "loss": 0.4663, + "step": 21519 + }, + { + "epoch": 0.9528131675946117, + "grad_norm": 1.483228325843811, + "learning_rate": 1.3565161053929687e-07, + "loss": 0.4491, + "step": 21520 + }, + { + "epoch": 0.9528574432994256, + "grad_norm": 1.56599760055542, + "learning_rate": 1.3539803745857615e-07, + "loss": 0.557, + "step": 21521 + }, + { + "epoch": 0.9529017190042394, + "grad_norm": 1.2650190591812134, + "learning_rate": 1.351446999858419e-07, + "loss": 0.4902, + "step": 21522 + }, + { + "epoch": 0.9529459947090533, + "grad_norm": 1.4023780822753906, + "learning_rate": 1.348915981271437e-07, + "loss": 0.4686, + "step": 21523 + }, + { + "epoch": 0.9529902704138672, + "grad_norm": 1.7137815952301025, + "learning_rate": 1.3463873188852782e-07, + "loss": 0.5472, + "step": 21524 + }, + { + "epoch": 0.953034546118681, + "grad_norm": 1.4693408012390137, + "learning_rate": 1.343861012760328e-07, + "loss": 0.5064, + "step": 21525 + }, + { + "epoch": 0.9530788218234949, + "grad_norm": 1.7261550426483154, + "learning_rate": 1.3413370629569266e-07, + "loss": 0.4589, + "step": 21526 + }, + { + "epoch": 0.9531230975283088, + "grad_norm": 1.4393396377563477, + "learning_rate": 1.3388154695353593e-07, + "loss": 0.4872, + "step": 21527 + }, + { + "epoch": 0.9531673732331226, + "grad_norm": 1.3988527059555054, + "learning_rate": 1.336296232555856e-07, + "loss": 0.495, + "step": 21528 + }, + { + "epoch": 0.9532116489379365, + "grad_norm": 1.3486697673797607, + "learning_rate": 1.333779352078579e-07, + "loss": 0.4507, + "step": 21529 + }, + { + "epoch": 0.9532559246427504, + "grad_norm": 1.345563292503357, + "learning_rate": 1.3312648281636477e-07, + "loss": 0.4753, + "step": 21530 + }, + { + "epoch": 0.9533002003475642, + "grad_norm": 1.7295105457305908, + "learning_rate": 1.3287526608711132e-07, + "loss": 0.572, + "step": 21531 + }, + { + "epoch": 0.9533444760523782, + "grad_norm": 1.5865141153335571, + "learning_rate": 1.3262428502609837e-07, + "loss": 0.5019, + "step": 21532 + }, + { + "epoch": 0.9533887517571921, + "grad_norm": 1.6545422077178955, + "learning_rate": 1.3237353963931888e-07, + "loss": 0.5158, + "step": 21533 + }, + { + "epoch": 0.9534330274620059, + "grad_norm": 1.2017226219177246, + "learning_rate": 1.3212302993276472e-07, + "loss": 0.3922, + "step": 21534 + }, + { + "epoch": 0.9534773031668198, + "grad_norm": 1.4392365217208862, + "learning_rate": 1.3187275591241665e-07, + "loss": 0.4948, + "step": 21535 + }, + { + "epoch": 0.9535215788716337, + "grad_norm": 1.66707444190979, + "learning_rate": 1.3162271758425105e-07, + "loss": 0.4299, + "step": 21536 + }, + { + "epoch": 0.9535658545764475, + "grad_norm": 1.6544147729873657, + "learning_rate": 1.313729149542442e-07, + "loss": 0.4949, + "step": 21537 + }, + { + "epoch": 0.9536101302812614, + "grad_norm": 1.6393979787826538, + "learning_rate": 1.3112334802835913e-07, + "loss": 0.5056, + "step": 21538 + }, + { + "epoch": 0.9536544059860753, + "grad_norm": 1.832774043083191, + "learning_rate": 1.3087401681255663e-07, + "loss": 0.5636, + "step": 21539 + }, + { + "epoch": 0.9536986816908891, + "grad_norm": 1.7619506120681763, + "learning_rate": 1.3062492131279415e-07, + "loss": 0.477, + "step": 21540 + }, + { + "epoch": 0.953742957395703, + "grad_norm": 1.4577394723892212, + "learning_rate": 1.3037606153501802e-07, + "loss": 0.4641, + "step": 21541 + }, + { + "epoch": 0.9537872331005169, + "grad_norm": 1.4558080434799194, + "learning_rate": 1.3012743748517466e-07, + "loss": 0.4504, + "step": 21542 + }, + { + "epoch": 0.9538315088053307, + "grad_norm": 1.359230637550354, + "learning_rate": 1.2987904916919924e-07, + "loss": 0.4106, + "step": 21543 + }, + { + "epoch": 0.9538757845101447, + "grad_norm": 1.7243396043777466, + "learning_rate": 1.296308965930271e-07, + "loss": 0.4924, + "step": 21544 + }, + { + "epoch": 0.9539200602149586, + "grad_norm": 1.3896353244781494, + "learning_rate": 1.2938297976258563e-07, + "loss": 0.4515, + "step": 21545 + }, + { + "epoch": 0.9539643359197725, + "grad_norm": 1.236153483390808, + "learning_rate": 1.2913529868379238e-07, + "loss": 0.4176, + "step": 21546 + }, + { + "epoch": 0.9540086116245863, + "grad_norm": 1.509405493736267, + "learning_rate": 1.2888785336256703e-07, + "loss": 0.5118, + "step": 21547 + }, + { + "epoch": 0.9540528873294002, + "grad_norm": 1.619709849357605, + "learning_rate": 1.2864064380481822e-07, + "loss": 0.497, + "step": 21548 + }, + { + "epoch": 0.9540971630342141, + "grad_norm": 1.512314796447754, + "learning_rate": 1.2839367001644898e-07, + "loss": 0.4579, + "step": 21549 + }, + { + "epoch": 0.9541414387390279, + "grad_norm": 1.408089280128479, + "learning_rate": 1.2814693200335905e-07, + "loss": 0.4356, + "step": 21550 + }, + { + "epoch": 0.9541857144438418, + "grad_norm": 1.4622303247451782, + "learning_rate": 1.2790042977144256e-07, + "loss": 0.4735, + "step": 21551 + }, + { + "epoch": 0.9542299901486557, + "grad_norm": 1.594570279121399, + "learning_rate": 1.27654163326586e-07, + "loss": 0.527, + "step": 21552 + }, + { + "epoch": 0.9542742658534695, + "grad_norm": 1.9320355653762817, + "learning_rate": 1.274081326746701e-07, + "loss": 0.522, + "step": 21553 + }, + { + "epoch": 0.9543185415582834, + "grad_norm": 1.5559114217758179, + "learning_rate": 1.2716233782157361e-07, + "loss": 0.5057, + "step": 21554 + }, + { + "epoch": 0.9543628172630974, + "grad_norm": 1.3707369565963745, + "learning_rate": 1.269167787731662e-07, + "loss": 0.4467, + "step": 21555 + }, + { + "epoch": 0.9544070929679112, + "grad_norm": 1.5627036094665527, + "learning_rate": 1.2667145553531214e-07, + "loss": 0.471, + "step": 21556 + }, + { + "epoch": 0.9544513686727251, + "grad_norm": 1.5415610074996948, + "learning_rate": 1.264263681138711e-07, + "loss": 0.4784, + "step": 21557 + }, + { + "epoch": 0.954495644377539, + "grad_norm": 1.7145062685012817, + "learning_rate": 1.2618151651469735e-07, + "loss": 0.5577, + "step": 21558 + }, + { + "epoch": 0.9545399200823528, + "grad_norm": 1.6828978061676025, + "learning_rate": 1.2593690074363728e-07, + "loss": 0.5073, + "step": 21559 + }, + { + "epoch": 0.9545841957871667, + "grad_norm": 1.586061954498291, + "learning_rate": 1.2569252080653516e-07, + "loss": 0.4702, + "step": 21560 + }, + { + "epoch": 0.9546284714919806, + "grad_norm": 1.3590015172958374, + "learning_rate": 1.2544837670922848e-07, + "loss": 0.4312, + "step": 21561 + }, + { + "epoch": 0.9546727471967944, + "grad_norm": 1.2944104671478271, + "learning_rate": 1.2520446845754596e-07, + "loss": 0.4473, + "step": 21562 + }, + { + "epoch": 0.9547170229016083, + "grad_norm": 2.1136722564697266, + "learning_rate": 1.2496079605731404e-07, + "loss": 0.5677, + "step": 21563 + }, + { + "epoch": 0.9547612986064222, + "grad_norm": 1.4428746700286865, + "learning_rate": 1.247173595143536e-07, + "loss": 0.4696, + "step": 21564 + }, + { + "epoch": 0.954805574311236, + "grad_norm": 1.8237382173538208, + "learning_rate": 1.2447415883447777e-07, + "loss": 0.4699, + "step": 21565 + }, + { + "epoch": 0.9548498500160499, + "grad_norm": 1.4377392530441284, + "learning_rate": 1.2423119402349636e-07, + "loss": 0.4689, + "step": 21566 + }, + { + "epoch": 0.9548941257208639, + "grad_norm": 1.4924087524414062, + "learning_rate": 1.2398846508721142e-07, + "loss": 0.4756, + "step": 21567 + }, + { + "epoch": 0.9549384014256777, + "grad_norm": 1.6283605098724365, + "learning_rate": 1.2374597203142157e-07, + "loss": 0.5128, + "step": 21568 + }, + { + "epoch": 0.9549826771304916, + "grad_norm": 1.3740203380584717, + "learning_rate": 1.235037148619156e-07, + "loss": 0.4574, + "step": 21569 + }, + { + "epoch": 0.9550269528353055, + "grad_norm": 1.528562068939209, + "learning_rate": 1.2326169358448326e-07, + "loss": 0.4717, + "step": 21570 + }, + { + "epoch": 0.9550712285401193, + "grad_norm": 1.5740324258804321, + "learning_rate": 1.230199082049033e-07, + "loss": 0.4887, + "step": 21571 + }, + { + "epoch": 0.9551155042449332, + "grad_norm": 1.8472812175750732, + "learning_rate": 1.2277835872894995e-07, + "loss": 0.5717, + "step": 21572 + }, + { + "epoch": 0.9551597799497471, + "grad_norm": 1.4507440328598022, + "learning_rate": 1.2253704516239306e-07, + "loss": 0.5271, + "step": 21573 + }, + { + "epoch": 0.9552040556545609, + "grad_norm": 1.4250255823135376, + "learning_rate": 1.2229596751099694e-07, + "loss": 0.4768, + "step": 21574 + }, + { + "epoch": 0.9552483313593748, + "grad_norm": 1.461319923400879, + "learning_rate": 1.2205512578051915e-07, + "loss": 0.432, + "step": 21575 + }, + { + "epoch": 0.9552926070641887, + "grad_norm": 1.6433539390563965, + "learning_rate": 1.218145199767118e-07, + "loss": 0.5107, + "step": 21576 + }, + { + "epoch": 0.9553368827690025, + "grad_norm": 1.6880754232406616, + "learning_rate": 1.2157415010532136e-07, + "loss": 0.4909, + "step": 21577 + }, + { + "epoch": 0.9553811584738164, + "grad_norm": 1.4537792205810547, + "learning_rate": 1.2133401617208885e-07, + "loss": 0.4557, + "step": 21578 + }, + { + "epoch": 0.9554254341786304, + "grad_norm": 1.573053002357483, + "learning_rate": 1.2109411818274851e-07, + "loss": 0.4947, + "step": 21579 + }, + { + "epoch": 0.9554697098834443, + "grad_norm": 1.5248703956604004, + "learning_rate": 1.208544561430336e-07, + "loss": 0.5138, + "step": 21580 + }, + { + "epoch": 0.9555139855882581, + "grad_norm": 1.5053380727767944, + "learning_rate": 1.2061503005866503e-07, + "loss": 0.4447, + "step": 21581 + }, + { + "epoch": 0.955558261293072, + "grad_norm": 1.317057728767395, + "learning_rate": 1.2037583993536272e-07, + "loss": 0.4324, + "step": 21582 + }, + { + "epoch": 0.9556025369978859, + "grad_norm": 1.215740442276001, + "learning_rate": 1.2013688577883874e-07, + "loss": 0.4346, + "step": 21583 + }, + { + "epoch": 0.9556468127026997, + "grad_norm": 1.6884615421295166, + "learning_rate": 1.1989816759480078e-07, + "loss": 0.4799, + "step": 21584 + }, + { + "epoch": 0.9556910884075136, + "grad_norm": 1.736061692237854, + "learning_rate": 1.196596853889509e-07, + "loss": 0.4978, + "step": 21585 + }, + { + "epoch": 0.9557353641123275, + "grad_norm": 1.697939157485962, + "learning_rate": 1.1942143916698458e-07, + "loss": 0.4717, + "step": 21586 + }, + { + "epoch": 0.9557796398171413, + "grad_norm": 1.4720028638839722, + "learning_rate": 1.1918342893459168e-07, + "loss": 0.4856, + "step": 21587 + }, + { + "epoch": 0.9558239155219552, + "grad_norm": 1.3657686710357666, + "learning_rate": 1.1894565469745766e-07, + "loss": 0.4667, + "step": 21588 + }, + { + "epoch": 0.9558681912267691, + "grad_norm": 1.492680549621582, + "learning_rate": 1.1870811646126135e-07, + "loss": 0.4774, + "step": 21589 + }, + { + "epoch": 0.9559124669315829, + "grad_norm": 1.3812817335128784, + "learning_rate": 1.1847081423167594e-07, + "loss": 0.4554, + "step": 21590 + }, + { + "epoch": 0.9559567426363969, + "grad_norm": 1.6076081991195679, + "learning_rate": 1.1823374801436916e-07, + "loss": 0.5466, + "step": 21591 + }, + { + "epoch": 0.9560010183412108, + "grad_norm": 1.3505250215530396, + "learning_rate": 1.1799691781500311e-07, + "loss": 0.432, + "step": 21592 + }, + { + "epoch": 0.9560452940460246, + "grad_norm": 1.5356327295303345, + "learning_rate": 1.177603236392344e-07, + "loss": 0.4772, + "step": 21593 + }, + { + "epoch": 0.9560895697508385, + "grad_norm": 1.3536205291748047, + "learning_rate": 1.1752396549271406e-07, + "loss": 0.4407, + "step": 21594 + }, + { + "epoch": 0.9561338454556524, + "grad_norm": 1.5989667177200317, + "learning_rate": 1.1728784338108756e-07, + "loss": 0.4847, + "step": 21595 + }, + { + "epoch": 0.9561781211604662, + "grad_norm": 1.4490267038345337, + "learning_rate": 1.1705195730999374e-07, + "loss": 0.4637, + "step": 21596 + }, + { + "epoch": 0.9562223968652801, + "grad_norm": 1.1870195865631104, + "learning_rate": 1.1681630728506699e-07, + "loss": 0.4011, + "step": 21597 + }, + { + "epoch": 0.956266672570094, + "grad_norm": 1.4919557571411133, + "learning_rate": 1.1658089331193612e-07, + "loss": 0.4225, + "step": 21598 + }, + { + "epoch": 0.9563109482749078, + "grad_norm": 2.6428568363189697, + "learning_rate": 1.1634571539622219e-07, + "loss": 0.5957, + "step": 21599 + }, + { + "epoch": 0.9563552239797217, + "grad_norm": 1.6843910217285156, + "learning_rate": 1.1611077354354406e-07, + "loss": 0.4354, + "step": 21600 + }, + { + "epoch": 0.9563994996845356, + "grad_norm": 1.3683382272720337, + "learning_rate": 1.1587606775951166e-07, + "loss": 0.4397, + "step": 21601 + }, + { + "epoch": 0.9564437753893494, + "grad_norm": 1.5848164558410645, + "learning_rate": 1.1564159804973163e-07, + "loss": 0.4837, + "step": 21602 + }, + { + "epoch": 0.9564880510941633, + "grad_norm": 1.6697145700454712, + "learning_rate": 1.1540736441980505e-07, + "loss": 0.5285, + "step": 21603 + }, + { + "epoch": 0.9565323267989773, + "grad_norm": 1.7641916275024414, + "learning_rate": 1.151733668753241e-07, + "loss": 0.5317, + "step": 21604 + }, + { + "epoch": 0.9565766025037911, + "grad_norm": 1.9391193389892578, + "learning_rate": 1.1493960542187877e-07, + "loss": 0.5147, + "step": 21605 + }, + { + "epoch": 0.956620878208605, + "grad_norm": 1.7238281965255737, + "learning_rate": 1.1470608006505235e-07, + "loss": 0.4521, + "step": 21606 + }, + { + "epoch": 0.9566651539134189, + "grad_norm": 1.6674350500106812, + "learning_rate": 1.1447279081042262e-07, + "loss": 0.5524, + "step": 21607 + }, + { + "epoch": 0.9567094296182327, + "grad_norm": 1.5828980207443237, + "learning_rate": 1.1423973766356067e-07, + "loss": 0.476, + "step": 21608 + }, + { + "epoch": 0.9567537053230466, + "grad_norm": 1.4849607944488525, + "learning_rate": 1.1400692063003427e-07, + "loss": 0.4435, + "step": 21609 + }, + { + "epoch": 0.9567979810278605, + "grad_norm": 1.4295672178268433, + "learning_rate": 1.1377433971540119e-07, + "loss": 0.4338, + "step": 21610 + }, + { + "epoch": 0.9568422567326743, + "grad_norm": 1.5068166255950928, + "learning_rate": 1.1354199492522034e-07, + "loss": 0.4641, + "step": 21611 + }, + { + "epoch": 0.9568865324374882, + "grad_norm": 1.3942334651947021, + "learning_rate": 1.1330988626503837e-07, + "loss": 0.448, + "step": 21612 + }, + { + "epoch": 0.9569308081423021, + "grad_norm": 1.9787765741348267, + "learning_rate": 1.1307801374039973e-07, + "loss": 0.5117, + "step": 21613 + }, + { + "epoch": 0.9569750838471159, + "grad_norm": 1.3153036832809448, + "learning_rate": 1.1284637735684223e-07, + "loss": 0.3935, + "step": 21614 + }, + { + "epoch": 0.9570193595519298, + "grad_norm": 1.4802696704864502, + "learning_rate": 1.1261497711989921e-07, + "loss": 0.5209, + "step": 21615 + }, + { + "epoch": 0.9570636352567438, + "grad_norm": 1.616416335105896, + "learning_rate": 1.1238381303509516e-07, + "loss": 0.5107, + "step": 21616 + }, + { + "epoch": 0.9571079109615577, + "grad_norm": 1.424790859222412, + "learning_rate": 1.1215288510795453e-07, + "loss": 0.4174, + "step": 21617 + }, + { + "epoch": 0.9571521866663715, + "grad_norm": 1.6845149993896484, + "learning_rate": 1.1192219334399068e-07, + "loss": 0.503, + "step": 21618 + }, + { + "epoch": 0.9571964623711854, + "grad_norm": 1.7141804695129395, + "learning_rate": 1.1169173774871478e-07, + "loss": 0.5011, + "step": 21619 + }, + { + "epoch": 0.9572407380759993, + "grad_norm": 1.5333274602890015, + "learning_rate": 1.1146151832762908e-07, + "loss": 0.5635, + "step": 21620 + }, + { + "epoch": 0.9572850137808131, + "grad_norm": 1.5968965291976929, + "learning_rate": 1.1123153508623474e-07, + "loss": 0.5376, + "step": 21621 + }, + { + "epoch": 0.957329289485627, + "grad_norm": 1.398559331893921, + "learning_rate": 1.1100178803002293e-07, + "loss": 0.456, + "step": 21622 + }, + { + "epoch": 0.9573735651904409, + "grad_norm": 1.3959298133850098, + "learning_rate": 1.1077227716448258e-07, + "loss": 0.4367, + "step": 21623 + }, + { + "epoch": 0.9574178408952547, + "grad_norm": 1.3755606412887573, + "learning_rate": 1.1054300249509264e-07, + "loss": 0.4809, + "step": 21624 + }, + { + "epoch": 0.9574621166000686, + "grad_norm": 1.258081078529358, + "learning_rate": 1.1031396402733208e-07, + "loss": 0.4587, + "step": 21625 + }, + { + "epoch": 0.9575063923048825, + "grad_norm": 1.5649397373199463, + "learning_rate": 1.1008516176666872e-07, + "loss": 0.477, + "step": 21626 + }, + { + "epoch": 0.9575506680096963, + "grad_norm": 1.5119737386703491, + "learning_rate": 1.0985659571856933e-07, + "loss": 0.4643, + "step": 21627 + }, + { + "epoch": 0.9575949437145103, + "grad_norm": 1.5841938257217407, + "learning_rate": 1.0962826588849173e-07, + "loss": 0.5037, + "step": 21628 + }, + { + "epoch": 0.9576392194193242, + "grad_norm": 1.9497034549713135, + "learning_rate": 1.0940017228189048e-07, + "loss": 0.6067, + "step": 21629 + }, + { + "epoch": 0.957683495124138, + "grad_norm": 1.5178699493408203, + "learning_rate": 1.0917231490421232e-07, + "loss": 0.4664, + "step": 21630 + }, + { + "epoch": 0.9577277708289519, + "grad_norm": 1.3723437786102295, + "learning_rate": 1.0894469376090067e-07, + "loss": 0.4574, + "step": 21631 + }, + { + "epoch": 0.9577720465337658, + "grad_norm": 1.5196940898895264, + "learning_rate": 1.0871730885739229e-07, + "loss": 0.4928, + "step": 21632 + }, + { + "epoch": 0.9578163222385796, + "grad_norm": 1.5738675594329834, + "learning_rate": 1.0849016019911618e-07, + "loss": 0.5077, + "step": 21633 + }, + { + "epoch": 0.9578605979433935, + "grad_norm": 1.512757420539856, + "learning_rate": 1.0826324779149799e-07, + "loss": 0.4793, + "step": 21634 + }, + { + "epoch": 0.9579048736482074, + "grad_norm": 1.3681479692459106, + "learning_rate": 1.0803657163995896e-07, + "loss": 0.4346, + "step": 21635 + }, + { + "epoch": 0.9579491493530212, + "grad_norm": 1.3888075351715088, + "learning_rate": 1.0781013174991139e-07, + "loss": 0.4461, + "step": 21636 + }, + { + "epoch": 0.9579934250578351, + "grad_norm": 1.5578893423080444, + "learning_rate": 1.0758392812676433e-07, + "loss": 0.5088, + "step": 21637 + }, + { + "epoch": 0.958037700762649, + "grad_norm": 1.8139222860336304, + "learning_rate": 1.073579607759212e-07, + "loss": 0.508, + "step": 21638 + }, + { + "epoch": 0.9580819764674628, + "grad_norm": 1.848673701286316, + "learning_rate": 1.071322297027777e-07, + "loss": 0.5552, + "step": 21639 + }, + { + "epoch": 0.9581262521722768, + "grad_norm": 1.4417186975479126, + "learning_rate": 1.0690673491272618e-07, + "loss": 0.4826, + "step": 21640 + }, + { + "epoch": 0.9581705278770907, + "grad_norm": 1.45061194896698, + "learning_rate": 1.0668147641115123e-07, + "loss": 0.4517, + "step": 21641 + }, + { + "epoch": 0.9582148035819045, + "grad_norm": 1.4663443565368652, + "learning_rate": 1.064564542034352e-07, + "loss": 0.4501, + "step": 21642 + }, + { + "epoch": 0.9582590792867184, + "grad_norm": 1.3052196502685547, + "learning_rate": 1.0623166829495046e-07, + "loss": 0.4614, + "step": 21643 + }, + { + "epoch": 0.9583033549915323, + "grad_norm": 1.4857126474380493, + "learning_rate": 1.0600711869106605e-07, + "loss": 0.4552, + "step": 21644 + }, + { + "epoch": 0.9583476306963461, + "grad_norm": 1.4213485717773438, + "learning_rate": 1.0578280539714659e-07, + "loss": 0.4619, + "step": 21645 + }, + { + "epoch": 0.95839190640116, + "grad_norm": 1.7752203941345215, + "learning_rate": 1.0555872841854665e-07, + "loss": 0.5752, + "step": 21646 + }, + { + "epoch": 0.9584361821059739, + "grad_norm": 1.6157547235488892, + "learning_rate": 1.0533488776062195e-07, + "loss": 0.4594, + "step": 21647 + }, + { + "epoch": 0.9584804578107877, + "grad_norm": 1.9671646356582642, + "learning_rate": 1.0511128342871713e-07, + "loss": 0.5417, + "step": 21648 + }, + { + "epoch": 0.9585247335156016, + "grad_norm": 1.5543653964996338, + "learning_rate": 1.0488791542817123e-07, + "loss": 0.4948, + "step": 21649 + }, + { + "epoch": 0.9585690092204155, + "grad_norm": 1.2148226499557495, + "learning_rate": 1.046647837643211e-07, + "loss": 0.4371, + "step": 21650 + }, + { + "epoch": 0.9586132849252295, + "grad_norm": 1.7439910173416138, + "learning_rate": 1.0444188844249691e-07, + "loss": 0.5134, + "step": 21651 + }, + { + "epoch": 0.9586575606300433, + "grad_norm": 1.6152158975601196, + "learning_rate": 1.0421922946801888e-07, + "loss": 0.5098, + "step": 21652 + }, + { + "epoch": 0.9587018363348572, + "grad_norm": 1.8040484189987183, + "learning_rate": 1.0399680684620828e-07, + "loss": 0.5236, + "step": 21653 + }, + { + "epoch": 0.9587461120396711, + "grad_norm": 1.3497896194458008, + "learning_rate": 1.037746205823753e-07, + "loss": 0.4754, + "step": 21654 + }, + { + "epoch": 0.9587903877444849, + "grad_norm": 1.5501375198364258, + "learning_rate": 1.0355267068182907e-07, + "loss": 0.5068, + "step": 21655 + }, + { + "epoch": 0.9588346634492988, + "grad_norm": 1.4695563316345215, + "learning_rate": 1.0333095714986751e-07, + "loss": 0.5084, + "step": 21656 + }, + { + "epoch": 0.9588789391541127, + "grad_norm": 1.411926507949829, + "learning_rate": 1.0310947999178977e-07, + "loss": 0.503, + "step": 21657 + }, + { + "epoch": 0.9589232148589265, + "grad_norm": 1.4970446825027466, + "learning_rate": 1.0288823921288271e-07, + "loss": 0.472, + "step": 21658 + }, + { + "epoch": 0.9589674905637404, + "grad_norm": 1.4349169731140137, + "learning_rate": 1.026672348184321e-07, + "loss": 0.4716, + "step": 21659 + }, + { + "epoch": 0.9590117662685543, + "grad_norm": 1.5972118377685547, + "learning_rate": 1.0244646681371595e-07, + "loss": 0.4774, + "step": 21660 + }, + { + "epoch": 0.9590560419733681, + "grad_norm": 1.4977105855941772, + "learning_rate": 1.022259352040067e-07, + "loss": 0.4652, + "step": 21661 + }, + { + "epoch": 0.959100317678182, + "grad_norm": 1.3546711206436157, + "learning_rate": 1.0200563999457236e-07, + "loss": 0.4365, + "step": 21662 + }, + { + "epoch": 0.959144593382996, + "grad_norm": 1.651092290878296, + "learning_rate": 1.0178558119067316e-07, + "loss": 0.5392, + "step": 21663 + }, + { + "epoch": 0.9591888690878098, + "grad_norm": 1.33274507522583, + "learning_rate": 1.0156575879756603e-07, + "loss": 0.4335, + "step": 21664 + }, + { + "epoch": 0.9592331447926237, + "grad_norm": 1.4592751264572144, + "learning_rate": 1.0134617282050119e-07, + "loss": 0.446, + "step": 21665 + }, + { + "epoch": 0.9592774204974376, + "grad_norm": 1.5842550992965698, + "learning_rate": 1.0112682326472334e-07, + "loss": 0.4904, + "step": 21666 + }, + { + "epoch": 0.9593216962022514, + "grad_norm": 1.4747700691223145, + "learning_rate": 1.0090771013547163e-07, + "loss": 0.4811, + "step": 21667 + }, + { + "epoch": 0.9593659719070653, + "grad_norm": 1.4672085046768188, + "learning_rate": 1.0068883343797964e-07, + "loss": 0.4536, + "step": 21668 + }, + { + "epoch": 0.9594102476118792, + "grad_norm": 1.400658130645752, + "learning_rate": 1.0047019317747431e-07, + "loss": 0.4446, + "step": 21669 + }, + { + "epoch": 0.959454523316693, + "grad_norm": 1.679575800895691, + "learning_rate": 1.0025178935917701e-07, + "loss": 0.4831, + "step": 21670 + }, + { + "epoch": 0.9594987990215069, + "grad_norm": 1.464958667755127, + "learning_rate": 1.000336219883058e-07, + "loss": 0.505, + "step": 21671 + }, + { + "epoch": 0.9595430747263208, + "grad_norm": 1.7828019857406616, + "learning_rate": 9.981569107007094e-08, + "loss": 0.4892, + "step": 21672 + }, + { + "epoch": 0.9595873504311346, + "grad_norm": 1.4370999336242676, + "learning_rate": 9.959799660967607e-08, + "loss": 0.5117, + "step": 21673 + }, + { + "epoch": 0.9596316261359485, + "grad_norm": 1.510535478591919, + "learning_rate": 9.938053861232255e-08, + "loss": 0.5194, + "step": 21674 + }, + { + "epoch": 0.9596759018407625, + "grad_norm": 1.3962018489837646, + "learning_rate": 9.916331708320403e-08, + "loss": 0.4406, + "step": 21675 + }, + { + "epoch": 0.9597201775455763, + "grad_norm": 1.3734042644500732, + "learning_rate": 9.894633202750637e-08, + "loss": 0.432, + "step": 21676 + }, + { + "epoch": 0.9597644532503902, + "grad_norm": 1.6051530838012695, + "learning_rate": 9.872958345041539e-08, + "loss": 0.4814, + "step": 21677 + }, + { + "epoch": 0.9598087289552041, + "grad_norm": 1.4889200925827026, + "learning_rate": 9.851307135710586e-08, + "loss": 0.462, + "step": 21678 + }, + { + "epoch": 0.9598530046600179, + "grad_norm": 1.4651952981948853, + "learning_rate": 9.829679575275031e-08, + "loss": 0.4561, + "step": 21679 + }, + { + "epoch": 0.9598972803648318, + "grad_norm": 1.263441562652588, + "learning_rate": 9.808075664251349e-08, + "loss": 0.3924, + "step": 21680 + }, + { + "epoch": 0.9599415560696457, + "grad_norm": 1.5193849802017212, + "learning_rate": 9.78649540315546e-08, + "loss": 0.4055, + "step": 21681 + }, + { + "epoch": 0.9599858317744595, + "grad_norm": 1.293358325958252, + "learning_rate": 9.764938792502843e-08, + "loss": 0.4113, + "step": 21682 + }, + { + "epoch": 0.9600301074792734, + "grad_norm": 1.2222604751586914, + "learning_rate": 9.743405832808417e-08, + "loss": 0.4273, + "step": 21683 + }, + { + "epoch": 0.9600743831840873, + "grad_norm": 1.775221586227417, + "learning_rate": 9.721896524586439e-08, + "loss": 0.4261, + "step": 21684 + }, + { + "epoch": 0.9601186588889011, + "grad_norm": 1.7146191596984863, + "learning_rate": 9.700410868350718e-08, + "loss": 0.5312, + "step": 21685 + }, + { + "epoch": 0.960162934593715, + "grad_norm": 1.5527970790863037, + "learning_rate": 9.67894886461429e-08, + "loss": 0.5055, + "step": 21686 + }, + { + "epoch": 0.960207210298529, + "grad_norm": 1.4128206968307495, + "learning_rate": 9.657510513889857e-08, + "loss": 0.4506, + "step": 21687 + }, + { + "epoch": 0.9602514860033429, + "grad_norm": 1.5709365606307983, + "learning_rate": 9.63609581668934e-08, + "loss": 0.5082, + "step": 21688 + }, + { + "epoch": 0.9602957617081567, + "grad_norm": 1.6383394002914429, + "learning_rate": 9.61470477352433e-08, + "loss": 0.5491, + "step": 21689 + }, + { + "epoch": 0.9603400374129706, + "grad_norm": 1.520717740058899, + "learning_rate": 9.593337384905755e-08, + "loss": 0.4702, + "step": 21690 + }, + { + "epoch": 0.9603843131177845, + "grad_norm": 1.8540222644805908, + "learning_rate": 9.571993651343869e-08, + "loss": 0.4982, + "step": 21691 + }, + { + "epoch": 0.9604285888225983, + "grad_norm": 1.4829039573669434, + "learning_rate": 9.55067357334849e-08, + "loss": 0.4462, + "step": 21692 + }, + { + "epoch": 0.9604728645274122, + "grad_norm": 1.4676969051361084, + "learning_rate": 9.529377151428653e-08, + "loss": 0.4647, + "step": 21693 + }, + { + "epoch": 0.9605171402322261, + "grad_norm": 1.4860447645187378, + "learning_rate": 9.508104386093398e-08, + "loss": 0.4886, + "step": 21694 + }, + { + "epoch": 0.9605614159370399, + "grad_norm": 1.8942395448684692, + "learning_rate": 9.486855277850427e-08, + "loss": 0.5322, + "step": 21695 + }, + { + "epoch": 0.9606056916418538, + "grad_norm": 1.5508147478103638, + "learning_rate": 9.465629827207445e-08, + "loss": 0.4773, + "step": 21696 + }, + { + "epoch": 0.9606499673466677, + "grad_norm": 1.8894050121307373, + "learning_rate": 9.444428034671382e-08, + "loss": 0.5631, + "step": 21697 + }, + { + "epoch": 0.9606942430514815, + "grad_norm": 1.589800238609314, + "learning_rate": 9.42324990074872e-08, + "loss": 0.4977, + "step": 21698 + }, + { + "epoch": 0.9607385187562955, + "grad_norm": 1.4891825914382935, + "learning_rate": 9.402095425945168e-08, + "loss": 0.4548, + "step": 21699 + }, + { + "epoch": 0.9607827944611094, + "grad_norm": 1.5366501808166504, + "learning_rate": 9.380964610765875e-08, + "loss": 0.5035, + "step": 21700 + }, + { + "epoch": 0.9608270701659232, + "grad_norm": 1.7794333696365356, + "learning_rate": 9.35985745571566e-08, + "loss": 0.4899, + "step": 21701 + }, + { + "epoch": 0.9608713458707371, + "grad_norm": 1.5829652547836304, + "learning_rate": 9.338773961298675e-08, + "loss": 0.4994, + "step": 21702 + }, + { + "epoch": 0.960915621575551, + "grad_norm": 1.7615115642547607, + "learning_rate": 9.317714128018407e-08, + "loss": 0.5759, + "step": 21703 + }, + { + "epoch": 0.9609598972803648, + "grad_norm": 1.5161106586456299, + "learning_rate": 9.296677956377897e-08, + "loss": 0.4642, + "step": 21704 + }, + { + "epoch": 0.9610041729851787, + "grad_norm": 1.7786611318588257, + "learning_rate": 9.275665446879633e-08, + "loss": 0.5428, + "step": 21705 + }, + { + "epoch": 0.9610484486899926, + "grad_norm": 1.5916835069656372, + "learning_rate": 9.254676600025436e-08, + "loss": 0.5108, + "step": 21706 + }, + { + "epoch": 0.9610927243948064, + "grad_norm": 1.3709025382995605, + "learning_rate": 9.233711416316571e-08, + "loss": 0.4194, + "step": 21707 + }, + { + "epoch": 0.9611370000996203, + "grad_norm": 1.3951648473739624, + "learning_rate": 9.212769896253748e-08, + "loss": 0.4361, + "step": 21708 + }, + { + "epoch": 0.9611812758044342, + "grad_norm": 2.0833470821380615, + "learning_rate": 9.191852040337235e-08, + "loss": 0.5293, + "step": 21709 + }, + { + "epoch": 0.961225551509248, + "grad_norm": 1.6543331146240234, + "learning_rate": 9.170957849066519e-08, + "loss": 0.4719, + "step": 21710 + }, + { + "epoch": 0.961269827214062, + "grad_norm": 2.0681235790252686, + "learning_rate": 9.15008732294076e-08, + "loss": 0.4205, + "step": 21711 + }, + { + "epoch": 0.9613141029188759, + "grad_norm": 1.5932413339614868, + "learning_rate": 9.129240462458444e-08, + "loss": 0.5147, + "step": 21712 + }, + { + "epoch": 0.9613583786236897, + "grad_norm": 1.8204114437103271, + "learning_rate": 9.108417268117287e-08, + "loss": 0.5079, + "step": 21713 + }, + { + "epoch": 0.9614026543285036, + "grad_norm": 1.3846514225006104, + "learning_rate": 9.087617740414889e-08, + "loss": 0.4228, + "step": 21714 + }, + { + "epoch": 0.9614469300333175, + "grad_norm": 1.8485108613967896, + "learning_rate": 9.066841879847854e-08, + "loss": 0.5541, + "step": 21715 + }, + { + "epoch": 0.9614912057381313, + "grad_norm": 1.5433645248413086, + "learning_rate": 9.046089686912563e-08, + "loss": 0.4765, + "step": 21716 + }, + { + "epoch": 0.9615354814429452, + "grad_norm": 1.4584134817123413, + "learning_rate": 9.025361162104396e-08, + "loss": 0.4577, + "step": 21717 + }, + { + "epoch": 0.9615797571477591, + "grad_norm": 1.507872462272644, + "learning_rate": 9.004656305918736e-08, + "loss": 0.4836, + "step": 21718 + }, + { + "epoch": 0.9616240328525729, + "grad_norm": 2.4745261669158936, + "learning_rate": 8.983975118849853e-08, + "loss": 0.4857, + "step": 21719 + }, + { + "epoch": 0.9616683085573868, + "grad_norm": 1.508493423461914, + "learning_rate": 8.963317601391796e-08, + "loss": 0.5389, + "step": 21720 + }, + { + "epoch": 0.9617125842622007, + "grad_norm": 2.147780418395996, + "learning_rate": 8.942683754038062e-08, + "loss": 0.5351, + "step": 21721 + }, + { + "epoch": 0.9617568599670147, + "grad_norm": 1.4263957738876343, + "learning_rate": 8.922073577281365e-08, + "loss": 0.4562, + "step": 21722 + }, + { + "epoch": 0.9618011356718285, + "grad_norm": 1.5750571489334106, + "learning_rate": 8.901487071613868e-08, + "loss": 0.4944, + "step": 21723 + }, + { + "epoch": 0.9618454113766424, + "grad_norm": 1.2132724523544312, + "learning_rate": 8.880924237527399e-08, + "loss": 0.3985, + "step": 21724 + }, + { + "epoch": 0.9618896870814563, + "grad_norm": 1.4489974975585938, + "learning_rate": 8.860385075513122e-08, + "loss": 0.4553, + "step": 21725 + }, + { + "epoch": 0.9619339627862701, + "grad_norm": 1.5250625610351562, + "learning_rate": 8.839869586061533e-08, + "loss": 0.5432, + "step": 21726 + }, + { + "epoch": 0.961978238491084, + "grad_norm": 1.6305453777313232, + "learning_rate": 8.819377769662684e-08, + "loss": 0.5118, + "step": 21727 + }, + { + "epoch": 0.9620225141958979, + "grad_norm": 1.5503166913986206, + "learning_rate": 8.798909626805852e-08, + "loss": 0.4582, + "step": 21728 + }, + { + "epoch": 0.9620667899007117, + "grad_norm": 1.513448715209961, + "learning_rate": 8.778465157979976e-08, + "loss": 0.445, + "step": 21729 + }, + { + "epoch": 0.9621110656055256, + "grad_norm": 1.5564780235290527, + "learning_rate": 8.758044363673445e-08, + "loss": 0.4784, + "step": 21730 + }, + { + "epoch": 0.9621553413103395, + "grad_norm": 1.4619346857070923, + "learning_rate": 8.73764724437387e-08, + "loss": 0.4513, + "step": 21731 + }, + { + "epoch": 0.9621996170151533, + "grad_norm": 1.8512108325958252, + "learning_rate": 8.717273800568527e-08, + "loss": 0.4578, + "step": 21732 + }, + { + "epoch": 0.9622438927199672, + "grad_norm": 1.5689207315444946, + "learning_rate": 8.696924032743803e-08, + "loss": 0.5052, + "step": 21733 + }, + { + "epoch": 0.9622881684247812, + "grad_norm": 1.58829665184021, + "learning_rate": 8.67659794138609e-08, + "loss": 0.4616, + "step": 21734 + }, + { + "epoch": 0.962332444129595, + "grad_norm": 1.5725808143615723, + "learning_rate": 8.656295526980662e-08, + "loss": 0.4833, + "step": 21735 + }, + { + "epoch": 0.9623767198344089, + "grad_norm": 1.5352543592453003, + "learning_rate": 8.636016790012358e-08, + "loss": 0.4743, + "step": 21736 + }, + { + "epoch": 0.9624209955392228, + "grad_norm": 1.7127238512039185, + "learning_rate": 8.615761730965677e-08, + "loss": 0.4626, + "step": 21737 + }, + { + "epoch": 0.9624652712440366, + "grad_norm": 1.4242935180664062, + "learning_rate": 8.595530350324233e-08, + "loss": 0.4644, + "step": 21738 + }, + { + "epoch": 0.9625095469488505, + "grad_norm": 1.7978887557983398, + "learning_rate": 8.575322648571304e-08, + "loss": 0.4637, + "step": 21739 + }, + { + "epoch": 0.9625538226536644, + "grad_norm": 1.5606582164764404, + "learning_rate": 8.555138626189619e-08, + "loss": 0.4898, + "step": 21740 + }, + { + "epoch": 0.9625980983584782, + "grad_norm": 1.4879435300827026, + "learning_rate": 8.534978283661122e-08, + "loss": 0.48, + "step": 21741 + }, + { + "epoch": 0.9626423740632921, + "grad_norm": 1.3045308589935303, + "learning_rate": 8.51484162146743e-08, + "loss": 0.3943, + "step": 21742 + }, + { + "epoch": 0.962686649768106, + "grad_norm": 1.5808813571929932, + "learning_rate": 8.494728640089378e-08, + "loss": 0.4482, + "step": 21743 + }, + { + "epoch": 0.9627309254729198, + "grad_norm": 1.4630489349365234, + "learning_rate": 8.474639340007474e-08, + "loss": 0.4573, + "step": 21744 + }, + { + "epoch": 0.9627752011777337, + "grad_norm": 1.5016896724700928, + "learning_rate": 8.454573721701443e-08, + "loss": 0.504, + "step": 21745 + }, + { + "epoch": 0.9628194768825477, + "grad_norm": 1.411057949066162, + "learning_rate": 8.434531785650679e-08, + "loss": 0.5006, + "step": 21746 + }, + { + "epoch": 0.9628637525873615, + "grad_norm": 1.575944185256958, + "learning_rate": 8.41451353233369e-08, + "loss": 0.5259, + "step": 21747 + }, + { + "epoch": 0.9629080282921754, + "grad_norm": 1.539629578590393, + "learning_rate": 8.394518962228648e-08, + "loss": 0.4622, + "step": 21748 + }, + { + "epoch": 0.9629523039969893, + "grad_norm": 1.5864098072052002, + "learning_rate": 8.374548075813171e-08, + "loss": 0.5021, + "step": 21749 + }, + { + "epoch": 0.9629965797018031, + "grad_norm": 1.2867519855499268, + "learning_rate": 8.354600873564101e-08, + "loss": 0.461, + "step": 21750 + }, + { + "epoch": 0.963040855406617, + "grad_norm": 1.4638125896453857, + "learning_rate": 8.334677355958055e-08, + "loss": 0.4359, + "step": 21751 + }, + { + "epoch": 0.9630851311114309, + "grad_norm": 1.7006779909133911, + "learning_rate": 8.314777523470763e-08, + "loss": 0.4625, + "step": 21752 + }, + { + "epoch": 0.9631294068162447, + "grad_norm": 1.8040345907211304, + "learning_rate": 8.294901376577514e-08, + "loss": 0.5636, + "step": 21753 + }, + { + "epoch": 0.9631736825210586, + "grad_norm": 1.6799066066741943, + "learning_rate": 8.275048915753259e-08, + "loss": 0.4934, + "step": 21754 + }, + { + "epoch": 0.9632179582258725, + "grad_norm": 1.4118602275848389, + "learning_rate": 8.255220141471842e-08, + "loss": 0.4477, + "step": 21755 + }, + { + "epoch": 0.9632622339306863, + "grad_norm": 1.4392764568328857, + "learning_rate": 8.235415054206886e-08, + "loss": 0.4309, + "step": 21756 + }, + { + "epoch": 0.9633065096355002, + "grad_norm": 1.4404771327972412, + "learning_rate": 8.215633654431565e-08, + "loss": 0.5058, + "step": 21757 + }, + { + "epoch": 0.9633507853403142, + "grad_norm": 1.6015897989273071, + "learning_rate": 8.195875942618392e-08, + "loss": 0.5101, + "step": 21758 + }, + { + "epoch": 0.9633950610451281, + "grad_norm": 1.634954571723938, + "learning_rate": 8.1761419192391e-08, + "loss": 0.4799, + "step": 21759 + }, + { + "epoch": 0.9634393367499419, + "grad_norm": 1.44524347782135, + "learning_rate": 8.156431584764979e-08, + "loss": 0.4096, + "step": 21760 + }, + { + "epoch": 0.9634836124547558, + "grad_norm": 1.5735975503921509, + "learning_rate": 8.136744939667096e-08, + "loss": 0.4673, + "step": 21761 + }, + { + "epoch": 0.9635278881595697, + "grad_norm": 2.3116185665130615, + "learning_rate": 8.117081984415298e-08, + "loss": 0.5088, + "step": 21762 + }, + { + "epoch": 0.9635721638643835, + "grad_norm": 1.854194164276123, + "learning_rate": 8.097442719479543e-08, + "loss": 0.5457, + "step": 21763 + }, + { + "epoch": 0.9636164395691974, + "grad_norm": 1.2495806217193604, + "learning_rate": 8.077827145328565e-08, + "loss": 0.4328, + "step": 21764 + }, + { + "epoch": 0.9636607152740113, + "grad_norm": 1.7288146018981934, + "learning_rate": 8.058235262431103e-08, + "loss": 0.6006, + "step": 21765 + }, + { + "epoch": 0.9637049909788251, + "grad_norm": 1.696692943572998, + "learning_rate": 8.038667071255113e-08, + "loss": 0.5328, + "step": 21766 + }, + { + "epoch": 0.963749266683639, + "grad_norm": 1.2298260927200317, + "learning_rate": 8.019122572267668e-08, + "loss": 0.4525, + "step": 21767 + }, + { + "epoch": 0.9637935423884529, + "grad_norm": 1.3118618726730347, + "learning_rate": 7.999601765935949e-08, + "loss": 0.4408, + "step": 21768 + }, + { + "epoch": 0.9638378180932667, + "grad_norm": 1.6482183933258057, + "learning_rate": 7.980104652725917e-08, + "loss": 0.5123, + "step": 21769 + }, + { + "epoch": 0.9638820937980807, + "grad_norm": 1.9151431322097778, + "learning_rate": 7.960631233103423e-08, + "loss": 0.5684, + "step": 21770 + }, + { + "epoch": 0.9639263695028946, + "grad_norm": 1.4842863082885742, + "learning_rate": 7.941181507533424e-08, + "loss": 0.5262, + "step": 21771 + }, + { + "epoch": 0.9639706452077084, + "grad_norm": 1.5470246076583862, + "learning_rate": 7.921755476480664e-08, + "loss": 0.4571, + "step": 21772 + }, + { + "epoch": 0.9640149209125223, + "grad_norm": 1.6125037670135498, + "learning_rate": 7.902353140408992e-08, + "loss": 0.4605, + "step": 21773 + }, + { + "epoch": 0.9640591966173362, + "grad_norm": 1.360520601272583, + "learning_rate": 7.882974499781704e-08, + "loss": 0.4511, + "step": 21774 + }, + { + "epoch": 0.96410347232215, + "grad_norm": 2.2547767162323, + "learning_rate": 7.863619555061874e-08, + "loss": 0.5534, + "step": 21775 + }, + { + "epoch": 0.9641477480269639, + "grad_norm": 1.7590795755386353, + "learning_rate": 7.844288306711578e-08, + "loss": 0.5018, + "step": 21776 + }, + { + "epoch": 0.9641920237317778, + "grad_norm": 1.7268750667572021, + "learning_rate": 7.824980755192557e-08, + "loss": 0.5393, + "step": 21777 + }, + { + "epoch": 0.9642362994365916, + "grad_norm": 1.577606439590454, + "learning_rate": 7.805696900966109e-08, + "loss": 0.4965, + "step": 21778 + }, + { + "epoch": 0.9642805751414055, + "grad_norm": 1.6175557374954224, + "learning_rate": 7.786436744492753e-08, + "loss": 0.4855, + "step": 21779 + }, + { + "epoch": 0.9643248508462194, + "grad_norm": 2.190370798110962, + "learning_rate": 7.767200286232346e-08, + "loss": 0.5605, + "step": 21780 + }, + { + "epoch": 0.9643691265510332, + "grad_norm": 1.8348522186279297, + "learning_rate": 7.747987526644518e-08, + "loss": 0.5483, + "step": 21781 + }, + { + "epoch": 0.9644134022558472, + "grad_norm": 1.8196529150009155, + "learning_rate": 7.728798466188125e-08, + "loss": 0.5395, + "step": 21782 + }, + { + "epoch": 0.9644576779606611, + "grad_norm": 1.8045588731765747, + "learning_rate": 7.709633105321357e-08, + "loss": 0.496, + "step": 21783 + }, + { + "epoch": 0.9645019536654749, + "grad_norm": 1.7498692274093628, + "learning_rate": 7.69049144450218e-08, + "loss": 0.4961, + "step": 21784 + }, + { + "epoch": 0.9645462293702888, + "grad_norm": 1.4399847984313965, + "learning_rate": 7.671373484187561e-08, + "loss": 0.4528, + "step": 21785 + }, + { + "epoch": 0.9645905050751027, + "grad_norm": 1.8112945556640625, + "learning_rate": 7.652279224834247e-08, + "loss": 0.5061, + "step": 21786 + }, + { + "epoch": 0.9646347807799165, + "grad_norm": 1.4328374862670898, + "learning_rate": 7.633208666898095e-08, + "loss": 0.4331, + "step": 21787 + }, + { + "epoch": 0.9646790564847304, + "grad_norm": 1.8151460886001587, + "learning_rate": 7.614161810834852e-08, + "loss": 0.4804, + "step": 21788 + }, + { + "epoch": 0.9647233321895443, + "grad_norm": 1.467398762702942, + "learning_rate": 7.595138657099376e-08, + "loss": 0.4728, + "step": 21789 + }, + { + "epoch": 0.9647676078943581, + "grad_norm": 1.3711209297180176, + "learning_rate": 7.576139206145861e-08, + "loss": 0.4612, + "step": 21790 + }, + { + "epoch": 0.964811883599172, + "grad_norm": 1.314667820930481, + "learning_rate": 7.557163458428274e-08, + "loss": 0.443, + "step": 21791 + }, + { + "epoch": 0.9648561593039859, + "grad_norm": 1.7763980627059937, + "learning_rate": 7.538211414399699e-08, + "loss": 0.6117, + "step": 21792 + }, + { + "epoch": 0.9649004350087999, + "grad_norm": 1.773790955543518, + "learning_rate": 7.519283074512884e-08, + "loss": 0.491, + "step": 21793 + }, + { + "epoch": 0.9649447107136137, + "grad_norm": 1.470245599746704, + "learning_rate": 7.500378439219914e-08, + "loss": 0.4652, + "step": 21794 + }, + { + "epoch": 0.9649889864184276, + "grad_norm": 1.3845099210739136, + "learning_rate": 7.481497508972313e-08, + "loss": 0.4379, + "step": 21795 + }, + { + "epoch": 0.9650332621232415, + "grad_norm": 1.9825176000595093, + "learning_rate": 7.462640284220945e-08, + "loss": 0.5449, + "step": 21796 + }, + { + "epoch": 0.9650775378280553, + "grad_norm": 1.3951911926269531, + "learning_rate": 7.443806765416229e-08, + "loss": 0.4544, + "step": 21797 + }, + { + "epoch": 0.9651218135328692, + "grad_norm": 1.3413329124450684, + "learning_rate": 7.424996953008023e-08, + "loss": 0.4508, + "step": 21798 + }, + { + "epoch": 0.9651660892376831, + "grad_norm": 1.4434524774551392, + "learning_rate": 7.406210847445639e-08, + "loss": 0.4632, + "step": 21799 + }, + { + "epoch": 0.9652103649424969, + "grad_norm": 1.701989769935608, + "learning_rate": 7.387448449177604e-08, + "loss": 0.5655, + "step": 21800 + }, + { + "epoch": 0.9652546406473108, + "grad_norm": 1.660527229309082, + "learning_rate": 7.368709758652337e-08, + "loss": 0.5108, + "step": 21801 + }, + { + "epoch": 0.9652989163521247, + "grad_norm": 1.3034721612930298, + "learning_rate": 7.349994776317148e-08, + "loss": 0.4396, + "step": 21802 + }, + { + "epoch": 0.9653431920569385, + "grad_norm": 1.6082253456115723, + "learning_rate": 7.331303502618903e-08, + "loss": 0.5014, + "step": 21803 + }, + { + "epoch": 0.9653874677617524, + "grad_norm": 1.2194498777389526, + "learning_rate": 7.312635938004353e-08, + "loss": 0.3892, + "step": 21804 + }, + { + "epoch": 0.9654317434665663, + "grad_norm": 1.7016674280166626, + "learning_rate": 7.293992082919255e-08, + "loss": 0.4835, + "step": 21805 + }, + { + "epoch": 0.9654760191713802, + "grad_norm": 1.7991197109222412, + "learning_rate": 7.275371937808806e-08, + "loss": 0.5607, + "step": 21806 + }, + { + "epoch": 0.9655202948761941, + "grad_norm": 2.3382749557495117, + "learning_rate": 7.256775503117763e-08, + "loss": 0.563, + "step": 21807 + }, + { + "epoch": 0.965564570581008, + "grad_norm": 1.647138237953186, + "learning_rate": 7.238202779290438e-08, + "loss": 0.4948, + "step": 21808 + }, + { + "epoch": 0.9656088462858218, + "grad_norm": 1.4655925035476685, + "learning_rate": 7.21965376677014e-08, + "loss": 0.4562, + "step": 21809 + }, + { + "epoch": 0.9656531219906357, + "grad_norm": 1.6365658044815063, + "learning_rate": 7.201128466000185e-08, + "loss": 0.5213, + "step": 21810 + }, + { + "epoch": 0.9656973976954496, + "grad_norm": 1.4641821384429932, + "learning_rate": 7.182626877422883e-08, + "loss": 0.4912, + "step": 21811 + }, + { + "epoch": 0.9657416734002634, + "grad_norm": 1.3652831315994263, + "learning_rate": 7.164149001480214e-08, + "loss": 0.4357, + "step": 21812 + }, + { + "epoch": 0.9657859491050773, + "grad_norm": 1.608121395111084, + "learning_rate": 7.145694838613271e-08, + "loss": 0.5293, + "step": 21813 + }, + { + "epoch": 0.9658302248098912, + "grad_norm": 1.4679688215255737, + "learning_rate": 7.127264389263144e-08, + "loss": 0.4848, + "step": 21814 + }, + { + "epoch": 0.965874500514705, + "grad_norm": 1.380189061164856, + "learning_rate": 7.108857653869816e-08, + "loss": 0.4751, + "step": 21815 + }, + { + "epoch": 0.9659187762195189, + "grad_norm": 1.390384316444397, + "learning_rate": 7.090474632873046e-08, + "loss": 0.4603, + "step": 21816 + }, + { + "epoch": 0.9659630519243328, + "grad_norm": 1.4549682140350342, + "learning_rate": 7.072115326711704e-08, + "loss": 0.4665, + "step": 21817 + }, + { + "epoch": 0.9660073276291467, + "grad_norm": 1.3873237371444702, + "learning_rate": 7.05377973582455e-08, + "loss": 0.4724, + "step": 21818 + }, + { + "epoch": 0.9660516033339606, + "grad_norm": 1.5145326852798462, + "learning_rate": 7.035467860649348e-08, + "loss": 0.5224, + "step": 21819 + }, + { + "epoch": 0.9660958790387745, + "grad_norm": 1.3171051740646362, + "learning_rate": 7.017179701623411e-08, + "loss": 0.4573, + "step": 21820 + }, + { + "epoch": 0.9661401547435883, + "grad_norm": 1.6069716215133667, + "learning_rate": 6.998915259183725e-08, + "loss": 0.5315, + "step": 21821 + }, + { + "epoch": 0.9661844304484022, + "grad_norm": 1.516005277633667, + "learning_rate": 6.980674533766385e-08, + "loss": 0.4912, + "step": 21822 + }, + { + "epoch": 0.9662287061532161, + "grad_norm": 1.6128864288330078, + "learning_rate": 6.962457525807153e-08, + "loss": 0.4989, + "step": 21823 + }, + { + "epoch": 0.9662729818580299, + "grad_norm": 1.2972543239593506, + "learning_rate": 6.944264235740905e-08, + "loss": 0.44, + "step": 21824 + }, + { + "epoch": 0.9663172575628438, + "grad_norm": 2.4189248085021973, + "learning_rate": 6.926094664002403e-08, + "loss": 0.5189, + "step": 21825 + }, + { + "epoch": 0.9663615332676577, + "grad_norm": 1.7590080499649048, + "learning_rate": 6.907948811025522e-08, + "loss": 0.5205, + "step": 21826 + }, + { + "epoch": 0.9664058089724715, + "grad_norm": 1.4967420101165771, + "learning_rate": 6.889826677243694e-08, + "loss": 0.4766, + "step": 21827 + }, + { + "epoch": 0.9664500846772854, + "grad_norm": 1.7892365455627441, + "learning_rate": 6.871728263089794e-08, + "loss": 0.5313, + "step": 21828 + }, + { + "epoch": 0.9664943603820993, + "grad_norm": 1.387402892112732, + "learning_rate": 6.853653568995922e-08, + "loss": 0.4507, + "step": 21829 + }, + { + "epoch": 0.9665386360869133, + "grad_norm": 1.4719114303588867, + "learning_rate": 6.835602595393953e-08, + "loss": 0.4486, + "step": 21830 + }, + { + "epoch": 0.9665829117917271, + "grad_norm": 1.4040088653564453, + "learning_rate": 6.817575342714988e-08, + "loss": 0.4366, + "step": 21831 + }, + { + "epoch": 0.966627187496541, + "grad_norm": 1.1929192543029785, + "learning_rate": 6.799571811389572e-08, + "loss": 0.4482, + "step": 21832 + }, + { + "epoch": 0.9666714632013549, + "grad_norm": 1.326244592666626, + "learning_rate": 6.781592001847582e-08, + "loss": 0.451, + "step": 21833 + }, + { + "epoch": 0.9667157389061687, + "grad_norm": 1.7982231378555298, + "learning_rate": 6.763635914518563e-08, + "loss": 0.4335, + "step": 21834 + }, + { + "epoch": 0.9667600146109826, + "grad_norm": 1.6019366979599, + "learning_rate": 6.745703549831394e-08, + "loss": 0.5324, + "step": 21835 + }, + { + "epoch": 0.9668042903157965, + "grad_norm": 1.2968107461929321, + "learning_rate": 6.727794908214403e-08, + "loss": 0.4342, + "step": 21836 + }, + { + "epoch": 0.9668485660206103, + "grad_norm": 1.5902677774429321, + "learning_rate": 6.709909990095243e-08, + "loss": 0.4882, + "step": 21837 + }, + { + "epoch": 0.9668928417254242, + "grad_norm": 1.633985996246338, + "learning_rate": 6.69204879590113e-08, + "loss": 0.5337, + "step": 21838 + }, + { + "epoch": 0.9669371174302381, + "grad_norm": 1.5803534984588623, + "learning_rate": 6.674211326058721e-08, + "loss": 0.4782, + "step": 21839 + }, + { + "epoch": 0.9669813931350519, + "grad_norm": 1.7578054666519165, + "learning_rate": 6.656397580993901e-08, + "loss": 0.5515, + "step": 21840 + }, + { + "epoch": 0.9670256688398658, + "grad_norm": 1.3585617542266846, + "learning_rate": 6.638607561132216e-08, + "loss": 0.4422, + "step": 21841 + }, + { + "epoch": 0.9670699445446798, + "grad_norm": 1.4186338186264038, + "learning_rate": 6.620841266898548e-08, + "loss": 0.4842, + "step": 21842 + }, + { + "epoch": 0.9671142202494936, + "grad_norm": 1.6248828172683716, + "learning_rate": 6.603098698717336e-08, + "loss": 0.4942, + "step": 21843 + }, + { + "epoch": 0.9671584959543075, + "grad_norm": 2.396395206451416, + "learning_rate": 6.58537985701213e-08, + "loss": 0.5371, + "step": 21844 + }, + { + "epoch": 0.9672027716591214, + "grad_norm": 1.538846492767334, + "learning_rate": 6.567684742206371e-08, + "loss": 0.4877, + "step": 21845 + }, + { + "epoch": 0.9672470473639352, + "grad_norm": 1.3567166328430176, + "learning_rate": 6.550013354722495e-08, + "loss": 0.4585, + "step": 21846 + }, + { + "epoch": 0.9672913230687491, + "grad_norm": 1.3667443990707397, + "learning_rate": 6.532365694982612e-08, + "loss": 0.4867, + "step": 21847 + }, + { + "epoch": 0.967335598773563, + "grad_norm": 1.5916163921356201, + "learning_rate": 6.514741763408383e-08, + "loss": 0.517, + "step": 21848 + }, + { + "epoch": 0.9673798744783768, + "grad_norm": 1.5851929187774658, + "learning_rate": 6.497141560420584e-08, + "loss": 0.5483, + "step": 21849 + }, + { + "epoch": 0.9674241501831907, + "grad_norm": 1.6627558469772339, + "learning_rate": 6.479565086439655e-08, + "loss": 0.5432, + "step": 21850 + }, + { + "epoch": 0.9674684258880046, + "grad_norm": 1.672764778137207, + "learning_rate": 6.462012341885259e-08, + "loss": 0.5091, + "step": 21851 + }, + { + "epoch": 0.9675127015928184, + "grad_norm": 3.572899103164673, + "learning_rate": 6.444483327176843e-08, + "loss": 0.6179, + "step": 21852 + }, + { + "epoch": 0.9675569772976323, + "grad_norm": 1.873331904411316, + "learning_rate": 6.426978042732845e-08, + "loss": 0.5242, + "step": 21853 + }, + { + "epoch": 0.9676012530024463, + "grad_norm": 1.4949791431427002, + "learning_rate": 6.409496488971489e-08, + "loss": 0.4705, + "step": 21854 + }, + { + "epoch": 0.9676455287072601, + "grad_norm": 1.744264841079712, + "learning_rate": 6.39203866631044e-08, + "loss": 0.4944, + "step": 21855 + }, + { + "epoch": 0.967689804412074, + "grad_norm": 1.7069346904754639, + "learning_rate": 6.374604575166365e-08, + "loss": 0.484, + "step": 21856 + }, + { + "epoch": 0.9677340801168879, + "grad_norm": 1.8661507368087769, + "learning_rate": 6.357194215955931e-08, + "loss": 0.4708, + "step": 21857 + }, + { + "epoch": 0.9677783558217017, + "grad_norm": 1.6620944738388062, + "learning_rate": 6.339807589094804e-08, + "loss": 0.4629, + "step": 21858 + }, + { + "epoch": 0.9678226315265156, + "grad_norm": 1.4148433208465576, + "learning_rate": 6.32244469499832e-08, + "loss": 0.4334, + "step": 21859 + }, + { + "epoch": 0.9678669072313295, + "grad_norm": 1.3584895133972168, + "learning_rate": 6.305105534081146e-08, + "loss": 0.4654, + "step": 21860 + }, + { + "epoch": 0.9679111829361433, + "grad_norm": 1.8066867589950562, + "learning_rate": 6.287790106757396e-08, + "loss": 0.5792, + "step": 21861 + }, + { + "epoch": 0.9679554586409572, + "grad_norm": 1.7396490573883057, + "learning_rate": 6.27049841344074e-08, + "loss": 0.518, + "step": 21862 + }, + { + "epoch": 0.9679997343457711, + "grad_norm": 1.4282361268997192, + "learning_rate": 6.253230454544068e-08, + "loss": 0.4695, + "step": 21863 + }, + { + "epoch": 0.968044010050585, + "grad_norm": 1.4768784046173096, + "learning_rate": 6.23598623047983e-08, + "loss": 0.4522, + "step": 21864 + }, + { + "epoch": 0.9680882857553988, + "grad_norm": 1.4867042303085327, + "learning_rate": 6.218765741659915e-08, + "loss": 0.4785, + "step": 21865 + }, + { + "epoch": 0.9681325614602128, + "grad_norm": 1.5692150592803955, + "learning_rate": 6.201568988495665e-08, + "loss": 0.4838, + "step": 21866 + }, + { + "epoch": 0.9681768371650267, + "grad_norm": 1.3082783222198486, + "learning_rate": 6.18439597139775e-08, + "loss": 0.4365, + "step": 21867 + }, + { + "epoch": 0.9682211128698405, + "grad_norm": 1.6476788520812988, + "learning_rate": 6.167246690776285e-08, + "loss": 0.4824, + "step": 21868 + }, + { + "epoch": 0.9682653885746544, + "grad_norm": 1.5083283185958862, + "learning_rate": 6.150121147041055e-08, + "loss": 0.4449, + "step": 21869 + }, + { + "epoch": 0.9683096642794683, + "grad_norm": 1.3891805410385132, + "learning_rate": 6.133019340600732e-08, + "loss": 0.4589, + "step": 21870 + }, + { + "epoch": 0.9683539399842821, + "grad_norm": 1.3394851684570312, + "learning_rate": 6.115941271864213e-08, + "loss": 0.4483, + "step": 21871 + }, + { + "epoch": 0.968398215689096, + "grad_norm": 1.5210438966751099, + "learning_rate": 6.098886941239057e-08, + "loss": 0.4236, + "step": 21872 + }, + { + "epoch": 0.9684424913939099, + "grad_norm": 1.7721071243286133, + "learning_rate": 6.08185634913283e-08, + "loss": 0.4391, + "step": 21873 + }, + { + "epoch": 0.9684867670987237, + "grad_norm": 1.4353021383285522, + "learning_rate": 6.064849495952097e-08, + "loss": 0.4327, + "step": 21874 + }, + { + "epoch": 0.9685310428035376, + "grad_norm": 1.892279028892517, + "learning_rate": 6.047866382103196e-08, + "loss": 0.533, + "step": 21875 + }, + { + "epoch": 0.9685753185083515, + "grad_norm": 1.3734771013259888, + "learning_rate": 6.030907007991693e-08, + "loss": 0.4584, + "step": 21876 + }, + { + "epoch": 0.9686195942131653, + "grad_norm": 1.6753971576690674, + "learning_rate": 6.013971374022598e-08, + "loss": 0.5416, + "step": 21877 + }, + { + "epoch": 0.9686638699179793, + "grad_norm": 1.5254151821136475, + "learning_rate": 5.997059480600476e-08, + "loss": 0.4446, + "step": 21878 + }, + { + "epoch": 0.9687081456227932, + "grad_norm": 1.4204256534576416, + "learning_rate": 5.980171328129335e-08, + "loss": 0.448, + "step": 21879 + }, + { + "epoch": 0.968752421327607, + "grad_norm": 1.2237224578857422, + "learning_rate": 5.963306917012301e-08, + "loss": 0.3971, + "step": 21880 + }, + { + "epoch": 0.9687966970324209, + "grad_norm": 1.8815034627914429, + "learning_rate": 5.9464662476523823e-08, + "loss": 0.5101, + "step": 21881 + }, + { + "epoch": 0.9688409727372348, + "grad_norm": 1.7524131536483765, + "learning_rate": 5.9296493204518135e-08, + "loss": 0.5292, + "step": 21882 + }, + { + "epoch": 0.9688852484420486, + "grad_norm": 1.543460726737976, + "learning_rate": 5.912856135812051e-08, + "loss": 0.4947, + "step": 21883 + }, + { + "epoch": 0.9689295241468625, + "grad_norm": 2.0156843662261963, + "learning_rate": 5.896086694134329e-08, + "loss": 0.5901, + "step": 21884 + }, + { + "epoch": 0.9689737998516764, + "grad_norm": 1.5186749696731567, + "learning_rate": 5.879340995819105e-08, + "loss": 0.4564, + "step": 21885 + }, + { + "epoch": 0.9690180755564902, + "grad_norm": 2.1051578521728516, + "learning_rate": 5.862619041266504e-08, + "loss": 0.6012, + "step": 21886 + }, + { + "epoch": 0.9690623512613041, + "grad_norm": 2.1055986881256104, + "learning_rate": 5.845920830875651e-08, + "loss": 0.6501, + "step": 21887 + }, + { + "epoch": 0.969106626966118, + "grad_norm": 1.509114384651184, + "learning_rate": 5.829246365045671e-08, + "loss": 0.514, + "step": 21888 + }, + { + "epoch": 0.9691509026709318, + "grad_norm": 1.293317437171936, + "learning_rate": 5.812595644174468e-08, + "loss": 0.4505, + "step": 21889 + }, + { + "epoch": 0.9691951783757458, + "grad_norm": 1.5459779500961304, + "learning_rate": 5.795968668660057e-08, + "loss": 0.4999, + "step": 21890 + }, + { + "epoch": 0.9692394540805597, + "grad_norm": 1.3773720264434814, + "learning_rate": 5.7793654388993426e-08, + "loss": 0.4258, + "step": 21891 + }, + { + "epoch": 0.9692837297853735, + "grad_norm": 2.0638089179992676, + "learning_rate": 5.762785955289007e-08, + "loss": 0.5355, + "step": 21892 + }, + { + "epoch": 0.9693280054901874, + "grad_norm": 2.025059461593628, + "learning_rate": 5.746230218224957e-08, + "loss": 0.4837, + "step": 21893 + }, + { + "epoch": 0.9693722811950013, + "grad_norm": 1.2275922298431396, + "learning_rate": 5.7296982281026534e-08, + "loss": 0.4115, + "step": 21894 + }, + { + "epoch": 0.9694165568998151, + "grad_norm": 1.9486316442489624, + "learning_rate": 5.713189985317003e-08, + "loss": 0.5085, + "step": 21895 + }, + { + "epoch": 0.969460832604629, + "grad_norm": 1.6428351402282715, + "learning_rate": 5.696705490262244e-08, + "loss": 0.4339, + "step": 21896 + }, + { + "epoch": 0.9695051083094429, + "grad_norm": 1.3466607332229614, + "learning_rate": 5.680244743331953e-08, + "loss": 0.4634, + "step": 21897 + }, + { + "epoch": 0.9695493840142568, + "grad_norm": 1.9315663576126099, + "learning_rate": 5.663807744919481e-08, + "loss": 0.4668, + "step": 21898 + }, + { + "epoch": 0.9695936597190706, + "grad_norm": 2.028289794921875, + "learning_rate": 5.647394495417402e-08, + "loss": 0.5651, + "step": 21899 + }, + { + "epoch": 0.9696379354238845, + "grad_norm": 2.3337090015411377, + "learning_rate": 5.631004995217515e-08, + "loss": 0.5758, + "step": 21900 + }, + { + "epoch": 0.9696822111286985, + "grad_norm": 1.4026198387145996, + "learning_rate": 5.614639244711617e-08, + "loss": 0.4056, + "step": 21901 + }, + { + "epoch": 0.9697264868335123, + "grad_norm": 1.5594310760498047, + "learning_rate": 5.598297244290285e-08, + "loss": 0.4604, + "step": 21902 + }, + { + "epoch": 0.9697707625383262, + "grad_norm": 2.1544740200042725, + "learning_rate": 5.581978994343984e-08, + "loss": 0.5409, + "step": 21903 + }, + { + "epoch": 0.9698150382431401, + "grad_norm": 1.4385348558425903, + "learning_rate": 5.5656844952624025e-08, + "loss": 0.4844, + "step": 21904 + }, + { + "epoch": 0.9698593139479539, + "grad_norm": 1.5910484790802002, + "learning_rate": 5.549413747434784e-08, + "loss": 0.5042, + "step": 21905 + }, + { + "epoch": 0.9699035896527678, + "grad_norm": 1.6932653188705444, + "learning_rate": 5.533166751249708e-08, + "loss": 0.5188, + "step": 21906 + }, + { + "epoch": 0.9699478653575817, + "grad_norm": 1.3408288955688477, + "learning_rate": 5.516943507095196e-08, + "loss": 0.4285, + "step": 21907 + }, + { + "epoch": 0.9699921410623955, + "grad_norm": 1.6060572862625122, + "learning_rate": 5.500744015358717e-08, + "loss": 0.4778, + "step": 21908 + }, + { + "epoch": 0.9700364167672094, + "grad_norm": 1.6745413541793823, + "learning_rate": 5.4845682764272936e-08, + "loss": 0.5062, + "step": 21909 + }, + { + "epoch": 0.9700806924720233, + "grad_norm": 1.6162794828414917, + "learning_rate": 5.468416290687173e-08, + "loss": 0.5211, + "step": 21910 + }, + { + "epoch": 0.9701249681768371, + "grad_norm": 1.779829978942871, + "learning_rate": 5.452288058524158e-08, + "loss": 0.4979, + "step": 21911 + }, + { + "epoch": 0.970169243881651, + "grad_norm": 1.5748528242111206, + "learning_rate": 5.4361835803233844e-08, + "loss": 0.4609, + "step": 21912 + }, + { + "epoch": 0.970213519586465, + "grad_norm": 1.4875324964523315, + "learning_rate": 5.4201028564696556e-08, + "loss": 0.4642, + "step": 21913 + }, + { + "epoch": 0.9702577952912788, + "grad_norm": 1.3571974039077759, + "learning_rate": 5.404045887346887e-08, + "loss": 0.437, + "step": 21914 + }, + { + "epoch": 0.9703020709960927, + "grad_norm": 1.3448230028152466, + "learning_rate": 5.388012673338661e-08, + "loss": 0.4332, + "step": 21915 + }, + { + "epoch": 0.9703463467009066, + "grad_norm": 1.6853607892990112, + "learning_rate": 5.3720032148278925e-08, + "loss": 0.4611, + "step": 21916 + }, + { + "epoch": 0.9703906224057204, + "grad_norm": 1.6502758264541626, + "learning_rate": 5.356017512197054e-08, + "loss": 0.4695, + "step": 21917 + }, + { + "epoch": 0.9704348981105343, + "grad_norm": 1.3174543380737305, + "learning_rate": 5.340055565827729e-08, + "loss": 0.4174, + "step": 21918 + }, + { + "epoch": 0.9704791738153482, + "grad_norm": 1.659958004951477, + "learning_rate": 5.32411737610139e-08, + "loss": 0.4981, + "step": 21919 + }, + { + "epoch": 0.970523449520162, + "grad_norm": 2.9018259048461914, + "learning_rate": 5.3082029433986216e-08, + "loss": 0.51, + "step": 21920 + }, + { + "epoch": 0.9705677252249759, + "grad_norm": 1.4059083461761475, + "learning_rate": 5.2923122680994534e-08, + "loss": 0.4427, + "step": 21921 + }, + { + "epoch": 0.9706120009297898, + "grad_norm": 1.4633592367172241, + "learning_rate": 5.276445350583581e-08, + "loss": 0.4302, + "step": 21922 + }, + { + "epoch": 0.9706562766346036, + "grad_norm": 1.7192250490188599, + "learning_rate": 5.260602191229813e-08, + "loss": 0.5191, + "step": 21923 + }, + { + "epoch": 0.9707005523394175, + "grad_norm": 2.5163135528564453, + "learning_rate": 5.244782790416625e-08, + "loss": 0.5083, + "step": 21924 + }, + { + "epoch": 0.9707448280442315, + "grad_norm": 1.7084964513778687, + "learning_rate": 5.228987148521825e-08, + "loss": 0.47, + "step": 21925 + }, + { + "epoch": 0.9707891037490453, + "grad_norm": 1.5399576425552368, + "learning_rate": 5.213215265922777e-08, + "loss": 0.4816, + "step": 21926 + }, + { + "epoch": 0.9708333794538592, + "grad_norm": 1.389597773551941, + "learning_rate": 5.19746714299596e-08, + "loss": 0.445, + "step": 21927 + }, + { + "epoch": 0.9708776551586731, + "grad_norm": 1.3018823862075806, + "learning_rate": 5.1817427801177377e-08, + "loss": 0.4349, + "step": 21928 + }, + { + "epoch": 0.9709219308634869, + "grad_norm": 1.391832947731018, + "learning_rate": 5.1660421776635886e-08, + "loss": 0.4788, + "step": 21929 + }, + { + "epoch": 0.9709662065683008, + "grad_norm": 1.9749139547348022, + "learning_rate": 5.150365336008434e-08, + "loss": 0.5234, + "step": 21930 + }, + { + "epoch": 0.9710104822731147, + "grad_norm": 1.526016116142273, + "learning_rate": 5.134712255526753e-08, + "loss": 0.522, + "step": 21931 + }, + { + "epoch": 0.9710547579779285, + "grad_norm": 1.5483628511428833, + "learning_rate": 5.119082936592468e-08, + "loss": 0.4987, + "step": 21932 + }, + { + "epoch": 0.9710990336827424, + "grad_norm": 1.3289073705673218, + "learning_rate": 5.1034773795788364e-08, + "loss": 0.4251, + "step": 21933 + }, + { + "epoch": 0.9711433093875563, + "grad_norm": 1.609822392463684, + "learning_rate": 5.0878955848585596e-08, + "loss": 0.5106, + "step": 21934 + }, + { + "epoch": 0.9711875850923702, + "grad_norm": 1.383816123008728, + "learning_rate": 5.072337552803785e-08, + "loss": 0.4376, + "step": 21935 + }, + { + "epoch": 0.971231860797184, + "grad_norm": 1.6030540466308594, + "learning_rate": 5.056803283786216e-08, + "loss": 0.4994, + "step": 21936 + }, + { + "epoch": 0.971276136501998, + "grad_norm": 1.4643306732177734, + "learning_rate": 5.041292778176665e-08, + "loss": 0.4945, + "step": 21937 + }, + { + "epoch": 0.9713204122068119, + "grad_norm": 1.438159465789795, + "learning_rate": 5.0258060363457263e-08, + "loss": 0.4426, + "step": 21938 + }, + { + "epoch": 0.9713646879116257, + "grad_norm": 1.7337818145751953, + "learning_rate": 5.010343058663325e-08, + "loss": 0.5174, + "step": 21939 + }, + { + "epoch": 0.9714089636164396, + "grad_norm": 1.6127442121505737, + "learning_rate": 4.994903845498722e-08, + "loss": 0.5008, + "step": 21940 + }, + { + "epoch": 0.9714532393212535, + "grad_norm": 1.2637945413589478, + "learning_rate": 4.979488397220733e-08, + "loss": 0.4091, + "step": 21941 + }, + { + "epoch": 0.9714975150260673, + "grad_norm": 1.6009376049041748, + "learning_rate": 4.964096714197397e-08, + "loss": 0.5074, + "step": 21942 + }, + { + "epoch": 0.9715417907308812, + "grad_norm": 1.710726022720337, + "learning_rate": 4.9487287967964206e-08, + "loss": 0.5523, + "step": 21943 + }, + { + "epoch": 0.9715860664356951, + "grad_norm": 1.6757484674453735, + "learning_rate": 4.933384645384953e-08, + "loss": 0.5246, + "step": 21944 + }, + { + "epoch": 0.9716303421405089, + "grad_norm": 1.4589626789093018, + "learning_rate": 4.9180642603292586e-08, + "loss": 0.477, + "step": 21945 + }, + { + "epoch": 0.9716746178453228, + "grad_norm": 1.4626637697219849, + "learning_rate": 4.9027676419954875e-08, + "loss": 0.4581, + "step": 21946 + }, + { + "epoch": 0.9717188935501367, + "grad_norm": 1.4250460863113403, + "learning_rate": 4.8874947907487924e-08, + "loss": 0.4526, + "step": 21947 + }, + { + "epoch": 0.9717631692549505, + "grad_norm": 3.3150579929351807, + "learning_rate": 4.872245706953993e-08, + "loss": 0.5792, + "step": 21948 + }, + { + "epoch": 0.9718074449597645, + "grad_norm": 2.2423508167266846, + "learning_rate": 4.8570203909754644e-08, + "loss": 0.491, + "step": 21949 + }, + { + "epoch": 0.9718517206645784, + "grad_norm": 1.3240275382995605, + "learning_rate": 4.841818843176693e-08, + "loss": 0.4462, + "step": 21950 + }, + { + "epoch": 0.9718959963693922, + "grad_norm": 1.728356957435608, + "learning_rate": 4.826641063920723e-08, + "loss": 0.5028, + "step": 21951 + }, + { + "epoch": 0.9719402720742061, + "grad_norm": 2.0513088703155518, + "learning_rate": 4.811487053570152e-08, + "loss": 0.5028, + "step": 21952 + }, + { + "epoch": 0.97198454777902, + "grad_norm": 1.6623071432113647, + "learning_rate": 4.796356812487024e-08, + "loss": 0.5334, + "step": 21953 + }, + { + "epoch": 0.9720288234838338, + "grad_norm": 1.4304755926132202, + "learning_rate": 4.781250341032384e-08, + "loss": 0.4918, + "step": 21954 + }, + { + "epoch": 0.9720730991886477, + "grad_norm": 1.5255454778671265, + "learning_rate": 4.7661676395673875e-08, + "loss": 0.4612, + "step": 21955 + }, + { + "epoch": 0.9721173748934616, + "grad_norm": 1.371355652809143, + "learning_rate": 4.751108708452079e-08, + "loss": 0.4243, + "step": 21956 + }, + { + "epoch": 0.9721616505982754, + "grad_norm": 1.4066728353500366, + "learning_rate": 4.736073548046172e-08, + "loss": 0.4718, + "step": 21957 + }, + { + "epoch": 0.9722059263030893, + "grad_norm": 1.485268473625183, + "learning_rate": 4.721062158708711e-08, + "loss": 0.4521, + "step": 21958 + }, + { + "epoch": 0.9722502020079032, + "grad_norm": 1.515997052192688, + "learning_rate": 4.706074540798411e-08, + "loss": 0.4822, + "step": 21959 + }, + { + "epoch": 0.972294477712717, + "grad_norm": 1.6245503425598145, + "learning_rate": 4.691110694673096e-08, + "loss": 0.5063, + "step": 21960 + }, + { + "epoch": 0.972338753417531, + "grad_norm": 1.6423368453979492, + "learning_rate": 4.676170620690146e-08, + "loss": 0.4998, + "step": 21961 + }, + { + "epoch": 0.9723830291223449, + "grad_norm": 1.3851568698883057, + "learning_rate": 4.6612543192064984e-08, + "loss": 0.4305, + "step": 21962 + }, + { + "epoch": 0.9724273048271587, + "grad_norm": 1.6876007318496704, + "learning_rate": 4.646361790578313e-08, + "loss": 0.4757, + "step": 21963 + }, + { + "epoch": 0.9724715805319726, + "grad_norm": 1.520184874534607, + "learning_rate": 4.631493035161305e-08, + "loss": 0.4903, + "step": 21964 + }, + { + "epoch": 0.9725158562367865, + "grad_norm": 2.073622941970825, + "learning_rate": 4.616648053310746e-08, + "loss": 0.529, + "step": 21965 + }, + { + "epoch": 0.9725601319416003, + "grad_norm": 1.8005924224853516, + "learning_rate": 4.601826845381019e-08, + "loss": 0.5153, + "step": 21966 + }, + { + "epoch": 0.9726044076464142, + "grad_norm": 1.6375629901885986, + "learning_rate": 4.587029411726063e-08, + "loss": 0.4791, + "step": 21967 + }, + { + "epoch": 0.9726486833512281, + "grad_norm": 1.5506733655929565, + "learning_rate": 4.5722557526994836e-08, + "loss": 0.5183, + "step": 21968 + }, + { + "epoch": 0.972692959056042, + "grad_norm": 1.7691173553466797, + "learning_rate": 4.557505868654111e-08, + "loss": 0.5128, + "step": 21969 + }, + { + "epoch": 0.9727372347608558, + "grad_norm": 1.2294831275939941, + "learning_rate": 4.5427797599422175e-08, + "loss": 0.4063, + "step": 21970 + }, + { + "epoch": 0.9727815104656697, + "grad_norm": 1.766904354095459, + "learning_rate": 4.528077426915412e-08, + "loss": 0.4878, + "step": 21971 + }, + { + "epoch": 0.9728257861704837, + "grad_norm": 1.5734432935714722, + "learning_rate": 4.5133988699249675e-08, + "loss": 0.4757, + "step": 21972 + }, + { + "epoch": 0.9728700618752975, + "grad_norm": 1.5528379678726196, + "learning_rate": 4.498744089321494e-08, + "loss": 0.5192, + "step": 21973 + }, + { + "epoch": 0.9729143375801114, + "grad_norm": 1.6616926193237305, + "learning_rate": 4.484113085454822e-08, + "loss": 0.5343, + "step": 21974 + }, + { + "epoch": 0.9729586132849253, + "grad_norm": 1.8070197105407715, + "learning_rate": 4.469505858674561e-08, + "loss": 0.505, + "step": 21975 + }, + { + "epoch": 0.9730028889897391, + "grad_norm": 1.242668628692627, + "learning_rate": 4.454922409329654e-08, + "loss": 0.4669, + "step": 21976 + }, + { + "epoch": 0.973047164694553, + "grad_norm": 1.522750973701477, + "learning_rate": 4.440362737768267e-08, + "loss": 0.4822, + "step": 21977 + }, + { + "epoch": 0.9730914403993669, + "grad_norm": 1.4969782829284668, + "learning_rate": 4.4258268443382326e-08, + "loss": 0.4806, + "step": 21978 + }, + { + "epoch": 0.9731357161041807, + "grad_norm": 1.7571638822555542, + "learning_rate": 4.411314729386607e-08, + "loss": 0.5117, + "step": 21979 + }, + { + "epoch": 0.9731799918089946, + "grad_norm": 1.6252281665802002, + "learning_rate": 4.396826393260223e-08, + "loss": 0.5031, + "step": 21980 + }, + { + "epoch": 0.9732242675138085, + "grad_norm": 1.4973117113113403, + "learning_rate": 4.382361836304916e-08, + "loss": 0.4813, + "step": 21981 + }, + { + "epoch": 0.9732685432186223, + "grad_norm": 1.517092227935791, + "learning_rate": 4.367921058866187e-08, + "loss": 0.4715, + "step": 21982 + }, + { + "epoch": 0.9733128189234362, + "grad_norm": 1.6267883777618408, + "learning_rate": 4.353504061289093e-08, + "loss": 0.5477, + "step": 21983 + }, + { + "epoch": 0.9733570946282502, + "grad_norm": 1.399762749671936, + "learning_rate": 4.339110843917804e-08, + "loss": 0.4901, + "step": 21984 + }, + { + "epoch": 0.973401370333064, + "grad_norm": 1.5087320804595947, + "learning_rate": 4.324741407096156e-08, + "loss": 0.4817, + "step": 21985 + }, + { + "epoch": 0.9734456460378779, + "grad_norm": 1.6507275104522705, + "learning_rate": 4.310395751167318e-08, + "loss": 0.5516, + "step": 21986 + }, + { + "epoch": 0.9734899217426918, + "grad_norm": 2.2372989654541016, + "learning_rate": 4.296073876474016e-08, + "loss": 0.4933, + "step": 21987 + }, + { + "epoch": 0.9735341974475056, + "grad_norm": 1.6596335172653198, + "learning_rate": 4.2817757833581997e-08, + "loss": 0.5425, + "step": 21988 + }, + { + "epoch": 0.9735784731523195, + "grad_norm": 1.559075117111206, + "learning_rate": 4.267501472161484e-08, + "loss": 0.4754, + "step": 21989 + }, + { + "epoch": 0.9736227488571334, + "grad_norm": 1.74833083152771, + "learning_rate": 4.2532509432247074e-08, + "loss": 0.5287, + "step": 21990 + }, + { + "epoch": 0.9736670245619472, + "grad_norm": 1.4202109575271606, + "learning_rate": 4.239024196888375e-08, + "loss": 0.4536, + "step": 21991 + }, + { + "epoch": 0.9737113002667611, + "grad_norm": 1.5934171676635742, + "learning_rate": 4.2248212334919935e-08, + "loss": 0.5015, + "step": 21992 + }, + { + "epoch": 0.973755575971575, + "grad_norm": 1.879071593284607, + "learning_rate": 4.210642053375069e-08, + "loss": 0.531, + "step": 21993 + }, + { + "epoch": 0.9737998516763888, + "grad_norm": 1.4507266283035278, + "learning_rate": 4.196486656876109e-08, + "loss": 0.4717, + "step": 21994 + }, + { + "epoch": 0.9738441273812027, + "grad_norm": 1.3070586919784546, + "learning_rate": 4.182355044333286e-08, + "loss": 0.4644, + "step": 21995 + }, + { + "epoch": 0.9738884030860167, + "grad_norm": 1.557413935661316, + "learning_rate": 4.1682472160841095e-08, + "loss": 0.4937, + "step": 21996 + }, + { + "epoch": 0.9739326787908305, + "grad_norm": 1.418184757232666, + "learning_rate": 4.154163172465531e-08, + "loss": 0.4733, + "step": 21997 + }, + { + "epoch": 0.9739769544956444, + "grad_norm": 1.3439915180206299, + "learning_rate": 4.140102913813837e-08, + "loss": 0.4394, + "step": 21998 + }, + { + "epoch": 0.9740212302004583, + "grad_norm": 1.3635046482086182, + "learning_rate": 4.126066440464982e-08, + "loss": 0.4253, + "step": 21999 + }, + { + "epoch": 0.9740655059052721, + "grad_norm": 1.7334656715393066, + "learning_rate": 4.112053752754141e-08, + "loss": 0.5052, + "step": 22000 + }, + { + "epoch": 0.974109781610086, + "grad_norm": 1.8287957906723022, + "learning_rate": 4.098064851016048e-08, + "loss": 0.5209, + "step": 22001 + }, + { + "epoch": 0.9741540573148999, + "grad_norm": 1.6722896099090576, + "learning_rate": 4.0840997355846565e-08, + "loss": 0.4648, + "step": 22002 + }, + { + "epoch": 0.9741983330197137, + "grad_norm": 1.5397673845291138, + "learning_rate": 4.070158406793811e-08, + "loss": 0.4835, + "step": 22003 + }, + { + "epoch": 0.9742426087245276, + "grad_norm": 1.4950724840164185, + "learning_rate": 4.0562408649761357e-08, + "loss": 0.4797, + "step": 22004 + }, + { + "epoch": 0.9742868844293415, + "grad_norm": 1.550519347190857, + "learning_rate": 4.042347110464362e-08, + "loss": 0.5197, + "step": 22005 + }, + { + "epoch": 0.9743311601341554, + "grad_norm": 1.7072619199752808, + "learning_rate": 4.028477143590115e-08, + "loss": 0.5189, + "step": 22006 + }, + { + "epoch": 0.9743754358389692, + "grad_norm": 1.7120147943496704, + "learning_rate": 4.014630964684796e-08, + "loss": 0.5013, + "step": 22007 + }, + { + "epoch": 0.9744197115437832, + "grad_norm": 1.4577103853225708, + "learning_rate": 4.0008085740790294e-08, + "loss": 0.4415, + "step": 22008 + }, + { + "epoch": 0.9744639872485971, + "grad_norm": 1.3602567911148071, + "learning_rate": 3.9870099721029955e-08, + "loss": 0.4018, + "step": 22009 + }, + { + "epoch": 0.9745082629534109, + "grad_norm": 1.4471592903137207, + "learning_rate": 3.9732351590862084e-08, + "loss": 0.4782, + "step": 22010 + }, + { + "epoch": 0.9745525386582248, + "grad_norm": 1.407962679862976, + "learning_rate": 3.9594841353577384e-08, + "loss": 0.4818, + "step": 22011 + }, + { + "epoch": 0.9745968143630387, + "grad_norm": 1.505183219909668, + "learning_rate": 3.9457569012459894e-08, + "loss": 0.4557, + "step": 22012 + }, + { + "epoch": 0.9746410900678525, + "grad_norm": 1.537079095840454, + "learning_rate": 3.93205345707881e-08, + "loss": 0.4724, + "step": 22013 + }, + { + "epoch": 0.9746853657726664, + "grad_norm": 1.6828792095184326, + "learning_rate": 3.918373803183384e-08, + "loss": 0.4673, + "step": 22014 + }, + { + "epoch": 0.9747296414774803, + "grad_norm": 1.274198055267334, + "learning_rate": 3.904717939886671e-08, + "loss": 0.427, + "step": 22015 + }, + { + "epoch": 0.9747739171822941, + "grad_norm": 1.527278184890747, + "learning_rate": 3.8910858675147434e-08, + "loss": 0.4686, + "step": 22016 + }, + { + "epoch": 0.974818192887108, + "grad_norm": 1.4873329401016235, + "learning_rate": 3.8774775863931194e-08, + "loss": 0.4858, + "step": 22017 + }, + { + "epoch": 0.9748624685919219, + "grad_norm": 1.3518197536468506, + "learning_rate": 3.863893096846982e-08, + "loss": 0.4447, + "step": 22018 + }, + { + "epoch": 0.9749067442967357, + "grad_norm": 1.6608977317810059, + "learning_rate": 3.850332399200518e-08, + "loss": 0.5199, + "step": 22019 + }, + { + "epoch": 0.9749510200015497, + "grad_norm": 1.5690702199935913, + "learning_rate": 3.8367954937779116e-08, + "loss": 0.5408, + "step": 22020 + }, + { + "epoch": 0.9749952957063636, + "grad_norm": 1.8555805683135986, + "learning_rate": 3.823282380902238e-08, + "loss": 0.5076, + "step": 22021 + }, + { + "epoch": 0.9750395714111774, + "grad_norm": 1.549609661102295, + "learning_rate": 3.80979306089635e-08, + "loss": 0.473, + "step": 22022 + }, + { + "epoch": 0.9750838471159913, + "grad_norm": 1.3980027437210083, + "learning_rate": 3.796327534082433e-08, + "loss": 0.4494, + "step": 22023 + }, + { + "epoch": 0.9751281228208052, + "grad_norm": 1.5426020622253418, + "learning_rate": 3.782885800782121e-08, + "loss": 0.4827, + "step": 22024 + }, + { + "epoch": 0.975172398525619, + "grad_norm": 1.588026762008667, + "learning_rate": 3.769467861316378e-08, + "loss": 0.4787, + "step": 22025 + }, + { + "epoch": 0.9752166742304329, + "grad_norm": 1.7437641620635986, + "learning_rate": 3.756073716005837e-08, + "loss": 0.5066, + "step": 22026 + }, + { + "epoch": 0.9752609499352468, + "grad_norm": 1.5752828121185303, + "learning_rate": 3.7427033651702414e-08, + "loss": 0.4572, + "step": 22027 + }, + { + "epoch": 0.9753052256400606, + "grad_norm": 1.4670814275741577, + "learning_rate": 3.729356809129003e-08, + "loss": 0.4885, + "step": 22028 + }, + { + "epoch": 0.9753495013448745, + "grad_norm": 1.674478530883789, + "learning_rate": 3.716034048200867e-08, + "loss": 0.5228, + "step": 22029 + }, + { + "epoch": 0.9753937770496884, + "grad_norm": 1.4136486053466797, + "learning_rate": 3.7027350827040234e-08, + "loss": 0.4489, + "step": 22030 + }, + { + "epoch": 0.9754380527545022, + "grad_norm": 1.5211858749389648, + "learning_rate": 3.6894599129561056e-08, + "loss": 0.4785, + "step": 22031 + }, + { + "epoch": 0.9754823284593162, + "grad_norm": 1.8936004638671875, + "learning_rate": 3.6762085392743064e-08, + "loss": 0.562, + "step": 22032 + }, + { + "epoch": 0.9755266041641301, + "grad_norm": 1.6040693521499634, + "learning_rate": 3.6629809619750376e-08, + "loss": 0.4553, + "step": 22033 + }, + { + "epoch": 0.9755708798689439, + "grad_norm": 1.7623344659805298, + "learning_rate": 3.649777181374159e-08, + "loss": 0.5017, + "step": 22034 + }, + { + "epoch": 0.9756151555737578, + "grad_norm": 1.4817827939987183, + "learning_rate": 3.636597197787084e-08, + "loss": 0.4751, + "step": 22035 + }, + { + "epoch": 0.9756594312785717, + "grad_norm": 1.2835547924041748, + "learning_rate": 3.623441011528672e-08, + "loss": 0.3995, + "step": 22036 + }, + { + "epoch": 0.9757037069833855, + "grad_norm": 1.4978370666503906, + "learning_rate": 3.610308622913117e-08, + "loss": 0.4405, + "step": 22037 + }, + { + "epoch": 0.9757479826881994, + "grad_norm": 1.6304819583892822, + "learning_rate": 3.597200032254056e-08, + "loss": 0.4974, + "step": 22038 + }, + { + "epoch": 0.9757922583930133, + "grad_norm": 1.5851495265960693, + "learning_rate": 3.584115239864461e-08, + "loss": 0.5178, + "step": 22039 + }, + { + "epoch": 0.9758365340978272, + "grad_norm": 1.310889720916748, + "learning_rate": 3.571054246057082e-08, + "loss": 0.4726, + "step": 22040 + }, + { + "epoch": 0.975880809802641, + "grad_norm": 1.4139955043792725, + "learning_rate": 3.558017051143559e-08, + "loss": 0.4625, + "step": 22041 + }, + { + "epoch": 0.9759250855074549, + "grad_norm": 1.5763651132583618, + "learning_rate": 3.545003655435642e-08, + "loss": 0.5184, + "step": 22042 + }, + { + "epoch": 0.9759693612122688, + "grad_norm": 2.938277006149292, + "learning_rate": 3.532014059243971e-08, + "loss": 0.4938, + "step": 22043 + }, + { + "epoch": 0.9760136369170826, + "grad_norm": 1.3120900392532349, + "learning_rate": 3.5190482628786326e-08, + "loss": 0.4293, + "step": 22044 + }, + { + "epoch": 0.9760579126218966, + "grad_norm": 1.7761263847351074, + "learning_rate": 3.5061062666496006e-08, + "loss": 0.488, + "step": 22045 + }, + { + "epoch": 0.9761021883267105, + "grad_norm": 1.3970857858657837, + "learning_rate": 3.4931880708658495e-08, + "loss": 0.4284, + "step": 22046 + }, + { + "epoch": 0.9761464640315243, + "grad_norm": 1.495368480682373, + "learning_rate": 3.480293675835911e-08, + "loss": 0.4386, + "step": 22047 + }, + { + "epoch": 0.9761907397363382, + "grad_norm": 1.5061631202697754, + "learning_rate": 3.467423081867649e-08, + "loss": 0.4641, + "step": 22048 + }, + { + "epoch": 0.9762350154411521, + "grad_norm": 1.8637486696243286, + "learning_rate": 3.454576289268707e-08, + "loss": 0.495, + "step": 22049 + }, + { + "epoch": 0.9762792911459659, + "grad_norm": 1.7175486087799072, + "learning_rate": 3.441753298345729e-08, + "loss": 0.4641, + "step": 22050 + }, + { + "epoch": 0.9763235668507798, + "grad_norm": 1.3278396129608154, + "learning_rate": 3.428954109405025e-08, + "loss": 0.4544, + "step": 22051 + }, + { + "epoch": 0.9763678425555937, + "grad_norm": 1.2877427339553833, + "learning_rate": 3.41617872275235e-08, + "loss": 0.4054, + "step": 22052 + }, + { + "epoch": 0.9764121182604075, + "grad_norm": 1.8126535415649414, + "learning_rate": 3.4034271386929053e-08, + "loss": 0.5517, + "step": 22053 + }, + { + "epoch": 0.9764563939652214, + "grad_norm": 1.5199339389801025, + "learning_rate": 3.390699357531002e-08, + "loss": 0.4312, + "step": 22054 + }, + { + "epoch": 0.9765006696700353, + "grad_norm": 1.9149028062820435, + "learning_rate": 3.377995379570731e-08, + "loss": 0.5588, + "step": 22055 + }, + { + "epoch": 0.9765449453748491, + "grad_norm": 1.5254331827163696, + "learning_rate": 3.365315205115738e-08, + "loss": 0.4345, + "step": 22056 + }, + { + "epoch": 0.9765892210796631, + "grad_norm": 1.7385963201522827, + "learning_rate": 3.3526588344685586e-08, + "loss": 0.5375, + "step": 22057 + }, + { + "epoch": 0.976633496784477, + "grad_norm": 1.5076580047607422, + "learning_rate": 3.3400262679317286e-08, + "loss": 0.5221, + "step": 22058 + }, + { + "epoch": 0.9766777724892908, + "grad_norm": 1.8509489297866821, + "learning_rate": 3.327417505806785e-08, + "loss": 0.5234, + "step": 22059 + }, + { + "epoch": 0.9767220481941047, + "grad_norm": 1.449615240097046, + "learning_rate": 3.3148325483949304e-08, + "loss": 0.5057, + "step": 22060 + }, + { + "epoch": 0.9767663238989186, + "grad_norm": 1.497416615486145, + "learning_rate": 3.302271395996814e-08, + "loss": 0.469, + "step": 22061 + }, + { + "epoch": 0.9768105996037324, + "grad_norm": 1.5870131254196167, + "learning_rate": 3.289734048912419e-08, + "loss": 0.5037, + "step": 22062 + }, + { + "epoch": 0.9768548753085463, + "grad_norm": 2.7717628479003906, + "learning_rate": 3.27722050744117e-08, + "loss": 0.4977, + "step": 22063 + }, + { + "epoch": 0.9768991510133602, + "grad_norm": 1.4303611516952515, + "learning_rate": 3.264730771881941e-08, + "loss": 0.4834, + "step": 22064 + }, + { + "epoch": 0.976943426718174, + "grad_norm": 1.4968392848968506, + "learning_rate": 3.252264842533048e-08, + "loss": 0.462, + "step": 22065 + }, + { + "epoch": 0.9769877024229879, + "grad_norm": 1.752374529838562, + "learning_rate": 3.239822719692143e-08, + "loss": 0.4948, + "step": 22066 + }, + { + "epoch": 0.9770319781278018, + "grad_norm": 1.1890367269515991, + "learning_rate": 3.227404403656542e-08, + "loss": 0.4146, + "step": 22067 + }, + { + "epoch": 0.9770762538326156, + "grad_norm": 1.8138192892074585, + "learning_rate": 3.215009894722787e-08, + "loss": 0.5212, + "step": 22068 + }, + { + "epoch": 0.9771205295374296, + "grad_norm": 1.2243164777755737, + "learning_rate": 3.2026391931869736e-08, + "loss": 0.4145, + "step": 22069 + }, + { + "epoch": 0.9771648052422435, + "grad_norm": 1.483368992805481, + "learning_rate": 3.190292299344422e-08, + "loss": 0.4549, + "step": 22070 + }, + { + "epoch": 0.9772090809470573, + "grad_norm": 1.3777735233306885, + "learning_rate": 3.177969213490117e-08, + "loss": 0.4425, + "step": 22071 + }, + { + "epoch": 0.9772533566518712, + "grad_norm": 1.2793327569961548, + "learning_rate": 3.165669935918381e-08, + "loss": 0.4139, + "step": 22072 + }, + { + "epoch": 0.9772976323566851, + "grad_norm": 1.382828950881958, + "learning_rate": 3.1533944669229765e-08, + "loss": 0.477, + "step": 22073 + }, + { + "epoch": 0.9773419080614989, + "grad_norm": 1.499299168586731, + "learning_rate": 3.141142806797004e-08, + "loss": 0.428, + "step": 22074 + }, + { + "epoch": 0.9773861837663128, + "grad_norm": 1.4457197189331055, + "learning_rate": 3.1289149558333395e-08, + "loss": 0.4716, + "step": 22075 + }, + { + "epoch": 0.9774304594711267, + "grad_norm": 1.4794590473175049, + "learning_rate": 3.116710914323751e-08, + "loss": 0.4691, + "step": 22076 + }, + { + "epoch": 0.9774747351759406, + "grad_norm": 1.4831552505493164, + "learning_rate": 3.104530682559781e-08, + "loss": 0.4911, + "step": 22077 + }, + { + "epoch": 0.9775190108807544, + "grad_norm": 1.802398920059204, + "learning_rate": 3.09237426083242e-08, + "loss": 0.5238, + "step": 22078 + }, + { + "epoch": 0.9775632865855683, + "grad_norm": 1.5242201089859009, + "learning_rate": 3.080241649431992e-08, + "loss": 0.4614, + "step": 22079 + }, + { + "epoch": 0.9776075622903823, + "grad_norm": 1.1967109441757202, + "learning_rate": 3.068132848648375e-08, + "loss": 0.3634, + "step": 22080 + }, + { + "epoch": 0.9776518379951961, + "grad_norm": 1.515357255935669, + "learning_rate": 3.056047858770561e-08, + "loss": 0.4917, + "step": 22081 + }, + { + "epoch": 0.97769611370001, + "grad_norm": 1.5918490886688232, + "learning_rate": 3.043986680087319e-08, + "loss": 0.5159, + "step": 22082 + }, + { + "epoch": 0.9777403894048239, + "grad_norm": 1.75719153881073, + "learning_rate": 3.03194931288664e-08, + "loss": 0.4418, + "step": 22083 + }, + { + "epoch": 0.9777846651096377, + "grad_norm": 1.9158861637115479, + "learning_rate": 3.0199357574561827e-08, + "loss": 0.4931, + "step": 22084 + }, + { + "epoch": 0.9778289408144516, + "grad_norm": 1.6342037916183472, + "learning_rate": 3.0079460140828295e-08, + "loss": 0.5218, + "step": 22085 + }, + { + "epoch": 0.9778732165192655, + "grad_norm": 1.4668669700622559, + "learning_rate": 2.995980083052796e-08, + "loss": 0.5016, + "step": 22086 + }, + { + "epoch": 0.9779174922240793, + "grad_norm": 1.309514045715332, + "learning_rate": 2.984037964652075e-08, + "loss": 0.4076, + "step": 22087 + }, + { + "epoch": 0.9779617679288932, + "grad_norm": 1.5934481620788574, + "learning_rate": 2.972119659165773e-08, + "loss": 0.4781, + "step": 22088 + }, + { + "epoch": 0.9780060436337071, + "grad_norm": 1.8698636293411255, + "learning_rate": 2.9602251668785498e-08, + "loss": 0.4666, + "step": 22089 + }, + { + "epoch": 0.9780503193385209, + "grad_norm": 1.2543001174926758, + "learning_rate": 2.9483544880745118e-08, + "loss": 0.4353, + "step": 22090 + }, + { + "epoch": 0.9780945950433348, + "grad_norm": 1.6884523630142212, + "learning_rate": 2.9365076230372103e-08, + "loss": 0.4501, + "step": 22091 + }, + { + "epoch": 0.9781388707481488, + "grad_norm": 1.374474048614502, + "learning_rate": 2.924684572049641e-08, + "loss": 0.4308, + "step": 22092 + }, + { + "epoch": 0.9781831464529626, + "grad_norm": 1.2637202739715576, + "learning_rate": 2.9128853353941335e-08, + "loss": 0.4023, + "step": 22093 + }, + { + "epoch": 0.9782274221577765, + "grad_norm": 1.3845421075820923, + "learning_rate": 2.9011099133524622e-08, + "loss": 0.4552, + "step": 22094 + }, + { + "epoch": 0.9782716978625904, + "grad_norm": 1.2829362154006958, + "learning_rate": 2.8893583062059583e-08, + "loss": 0.3995, + "step": 22095 + }, + { + "epoch": 0.9783159735674042, + "grad_norm": 1.6423135995864868, + "learning_rate": 2.8776305142351745e-08, + "loss": 0.4392, + "step": 22096 + }, + { + "epoch": 0.9783602492722181, + "grad_norm": 1.5342864990234375, + "learning_rate": 2.8659265377203317e-08, + "loss": 0.4949, + "step": 22097 + }, + { + "epoch": 0.978404524977032, + "grad_norm": 1.3803824186325073, + "learning_rate": 2.8542463769409833e-08, + "loss": 0.4465, + "step": 22098 + }, + { + "epoch": 0.9784488006818458, + "grad_norm": 1.5813010931015015, + "learning_rate": 2.842590032176018e-08, + "loss": 0.4793, + "step": 22099 + }, + { + "epoch": 0.9784930763866597, + "grad_norm": 1.3012008666992188, + "learning_rate": 2.8309575037038796e-08, + "loss": 0.4293, + "step": 22100 + }, + { + "epoch": 0.9785373520914736, + "grad_norm": 1.2755893468856812, + "learning_rate": 2.8193487918023453e-08, + "loss": 0.4387, + "step": 22101 + }, + { + "epoch": 0.9785816277962874, + "grad_norm": 2.0684115886688232, + "learning_rate": 2.8077638967487496e-08, + "loss": 0.6382, + "step": 22102 + }, + { + "epoch": 0.9786259035011013, + "grad_norm": 1.2974650859832764, + "learning_rate": 2.796202818819871e-08, + "loss": 0.4274, + "step": 22103 + }, + { + "epoch": 0.9786701792059153, + "grad_norm": 1.5898118019104004, + "learning_rate": 2.7846655582915995e-08, + "loss": 0.5404, + "step": 22104 + }, + { + "epoch": 0.9787144549107291, + "grad_norm": 1.595739722251892, + "learning_rate": 2.7731521154397145e-08, + "loss": 0.4698, + "step": 22105 + }, + { + "epoch": 0.978758730615543, + "grad_norm": 1.7848129272460938, + "learning_rate": 2.7616624905391076e-08, + "loss": 0.4757, + "step": 22106 + }, + { + "epoch": 0.9788030063203569, + "grad_norm": 2.3244853019714355, + "learning_rate": 2.7501966838642258e-08, + "loss": 0.5321, + "step": 22107 + }, + { + "epoch": 0.9788472820251707, + "grad_norm": 1.3020952939987183, + "learning_rate": 2.73875469568885e-08, + "loss": 0.3639, + "step": 22108 + }, + { + "epoch": 0.9788915577299846, + "grad_norm": 1.6267127990722656, + "learning_rate": 2.727336526286317e-08, + "loss": 0.4724, + "step": 22109 + }, + { + "epoch": 0.9789358334347985, + "grad_norm": 1.5046242475509644, + "learning_rate": 2.7159421759294092e-08, + "loss": 0.4963, + "step": 22110 + }, + { + "epoch": 0.9789801091396124, + "grad_norm": 1.5163159370422363, + "learning_rate": 2.7045716448901305e-08, + "loss": 0.4874, + "step": 22111 + }, + { + "epoch": 0.9790243848444262, + "grad_norm": 1.5414711236953735, + "learning_rate": 2.693224933440153e-08, + "loss": 0.487, + "step": 22112 + }, + { + "epoch": 0.9790686605492401, + "grad_norm": 1.6208957433700562, + "learning_rate": 2.6819020418504814e-08, + "loss": 0.4541, + "step": 22113 + }, + { + "epoch": 0.979112936254054, + "grad_norm": 1.4331289529800415, + "learning_rate": 2.670602970391567e-08, + "loss": 0.4524, + "step": 22114 + }, + { + "epoch": 0.9791572119588678, + "grad_norm": 1.9147958755493164, + "learning_rate": 2.6593277193331936e-08, + "loss": 0.5656, + "step": 22115 + }, + { + "epoch": 0.9792014876636818, + "grad_norm": 1.7726223468780518, + "learning_rate": 2.6480762889447008e-08, + "loss": 0.5142, + "step": 22116 + }, + { + "epoch": 0.9792457633684957, + "grad_norm": 1.4818917512893677, + "learning_rate": 2.6368486794948743e-08, + "loss": 0.5063, + "step": 22117 + }, + { + "epoch": 0.9792900390733095, + "grad_norm": 1.4566112756729126, + "learning_rate": 2.6256448912517217e-08, + "loss": 0.5052, + "step": 22118 + }, + { + "epoch": 0.9793343147781234, + "grad_norm": 1.5964285135269165, + "learning_rate": 2.6144649244830288e-08, + "loss": 0.4188, + "step": 22119 + }, + { + "epoch": 0.9793785904829373, + "grad_norm": 1.4832607507705688, + "learning_rate": 2.6033087794558042e-08, + "loss": 0.4317, + "step": 22120 + }, + { + "epoch": 0.9794228661877511, + "grad_norm": 1.410077691078186, + "learning_rate": 2.5921764564363904e-08, + "loss": 0.4337, + "step": 22121 + }, + { + "epoch": 0.979467141892565, + "grad_norm": 1.7762084007263184, + "learning_rate": 2.581067955690686e-08, + "loss": 0.526, + "step": 22122 + }, + { + "epoch": 0.9795114175973789, + "grad_norm": 1.4667119979858398, + "learning_rate": 2.569983277484145e-08, + "loss": 0.5134, + "step": 22123 + }, + { + "epoch": 0.9795556933021927, + "grad_norm": 1.5761903524398804, + "learning_rate": 2.5589224220812225e-08, + "loss": 0.4437, + "step": 22124 + }, + { + "epoch": 0.9795999690070066, + "grad_norm": 1.281754970550537, + "learning_rate": 2.547885389746485e-08, + "loss": 0.4153, + "step": 22125 + }, + { + "epoch": 0.9796442447118205, + "grad_norm": 1.2118107080459595, + "learning_rate": 2.536872180743277e-08, + "loss": 0.4022, + "step": 22126 + }, + { + "epoch": 0.9796885204166343, + "grad_norm": 1.849457859992981, + "learning_rate": 2.525882795334722e-08, + "loss": 0.5143, + "step": 22127 + }, + { + "epoch": 0.9797327961214483, + "grad_norm": 1.5442885160446167, + "learning_rate": 2.5149172337832762e-08, + "loss": 0.4936, + "step": 22128 + }, + { + "epoch": 0.9797770718262622, + "grad_norm": 1.5775480270385742, + "learning_rate": 2.503975496350841e-08, + "loss": 0.488, + "step": 22129 + }, + { + "epoch": 0.979821347531076, + "grad_norm": 1.4274499416351318, + "learning_rate": 2.4930575832987636e-08, + "loss": 0.5014, + "step": 22130 + }, + { + "epoch": 0.9798656232358899, + "grad_norm": 1.5449490547180176, + "learning_rate": 2.482163494887724e-08, + "loss": 0.4326, + "step": 22131 + }, + { + "epoch": 0.9799098989407038, + "grad_norm": 1.417502999305725, + "learning_rate": 2.4712932313781803e-08, + "loss": 0.4856, + "step": 22132 + }, + { + "epoch": 0.9799541746455176, + "grad_norm": 1.5086530447006226, + "learning_rate": 2.4604467930294807e-08, + "loss": 0.5272, + "step": 22133 + }, + { + "epoch": 0.9799984503503315, + "grad_norm": 1.5894923210144043, + "learning_rate": 2.4496241801007513e-08, + "loss": 0.4752, + "step": 22134 + }, + { + "epoch": 0.9800427260551454, + "grad_norm": 1.4972758293151855, + "learning_rate": 2.4388253928505633e-08, + "loss": 0.4553, + "step": 22135 + }, + { + "epoch": 0.9800870017599592, + "grad_norm": 1.3492189645767212, + "learning_rate": 2.4280504315367103e-08, + "loss": 0.4216, + "step": 22136 + }, + { + "epoch": 0.9801312774647731, + "grad_norm": 1.461777687072754, + "learning_rate": 2.417299296416653e-08, + "loss": 0.4698, + "step": 22137 + }, + { + "epoch": 0.980175553169587, + "grad_norm": 1.4146409034729004, + "learning_rate": 2.406571987747186e-08, + "loss": 0.4409, + "step": 22138 + }, + { + "epoch": 0.9802198288744008, + "grad_norm": 1.6235377788543701, + "learning_rate": 2.3958685057844378e-08, + "loss": 0.4831, + "step": 22139 + }, + { + "epoch": 0.9802641045792148, + "grad_norm": 1.3303319215774536, + "learning_rate": 2.3851888507840924e-08, + "loss": 0.4463, + "step": 22140 + }, + { + "epoch": 0.9803083802840287, + "grad_norm": 1.2131446599960327, + "learning_rate": 2.3745330230012797e-08, + "loss": 0.4541, + "step": 22141 + }, + { + "epoch": 0.9803526559888425, + "grad_norm": 1.9182744026184082, + "learning_rate": 2.3639010226904623e-08, + "loss": 0.5404, + "step": 22142 + }, + { + "epoch": 0.9803969316936564, + "grad_norm": 1.379797101020813, + "learning_rate": 2.3532928501054376e-08, + "loss": 0.4, + "step": 22143 + }, + { + "epoch": 0.9804412073984703, + "grad_norm": 1.9599323272705078, + "learning_rate": 2.34270850549978e-08, + "loss": 0.5162, + "step": 22144 + }, + { + "epoch": 0.9804854831032842, + "grad_norm": 1.3362382650375366, + "learning_rate": 2.3321479891261767e-08, + "loss": 0.4698, + "step": 22145 + }, + { + "epoch": 0.980529758808098, + "grad_norm": 1.5252310037612915, + "learning_rate": 2.321611301236981e-08, + "loss": 0.4833, + "step": 22146 + }, + { + "epoch": 0.9805740345129119, + "grad_norm": 1.5072933435440063, + "learning_rate": 2.311098442083659e-08, + "loss": 0.4757, + "step": 22147 + }, + { + "epoch": 0.9806183102177258, + "grad_norm": 1.4398884773254395, + "learning_rate": 2.3006094119173427e-08, + "loss": 0.489, + "step": 22148 + }, + { + "epoch": 0.9806625859225396, + "grad_norm": 1.8479821681976318, + "learning_rate": 2.2901442109887205e-08, + "loss": 0.5463, + "step": 22149 + }, + { + "epoch": 0.9807068616273535, + "grad_norm": 1.3394775390625, + "learning_rate": 2.2797028395477037e-08, + "loss": 0.4261, + "step": 22150 + }, + { + "epoch": 0.9807511373321675, + "grad_norm": 1.4847735166549683, + "learning_rate": 2.2692852978434266e-08, + "loss": 0.4632, + "step": 22151 + }, + { + "epoch": 0.9807954130369813, + "grad_norm": 1.3168740272521973, + "learning_rate": 2.2588915861250227e-08, + "loss": 0.4907, + "step": 22152 + }, + { + "epoch": 0.9808396887417952, + "grad_norm": 1.3722115755081177, + "learning_rate": 2.248521704640627e-08, + "loss": 0.4529, + "step": 22153 + }, + { + "epoch": 0.9808839644466091, + "grad_norm": 1.3957065343856812, + "learning_rate": 2.2381756536378197e-08, + "loss": 0.4255, + "step": 22154 + }, + { + "epoch": 0.9809282401514229, + "grad_norm": 1.4041122198104858, + "learning_rate": 2.227853433363736e-08, + "loss": 0.4732, + "step": 22155 + }, + { + "epoch": 0.9809725158562368, + "grad_norm": 1.7028394937515259, + "learning_rate": 2.217555044065067e-08, + "loss": 0.512, + "step": 22156 + }, + { + "epoch": 0.9810167915610507, + "grad_norm": 2.4597153663635254, + "learning_rate": 2.2072804859876173e-08, + "loss": 0.4942, + "step": 22157 + }, + { + "epoch": 0.9810610672658645, + "grad_norm": 1.6623971462249756, + "learning_rate": 2.1970297593767452e-08, + "loss": 0.526, + "step": 22158 + }, + { + "epoch": 0.9811053429706784, + "grad_norm": 1.3586153984069824, + "learning_rate": 2.1868028644774774e-08, + "loss": 0.4575, + "step": 22159 + }, + { + "epoch": 0.9811496186754923, + "grad_norm": 1.5640811920166016, + "learning_rate": 2.1765998015340627e-08, + "loss": 0.5071, + "step": 22160 + }, + { + "epoch": 0.9811938943803061, + "grad_norm": 1.7330811023712158, + "learning_rate": 2.1664205707899733e-08, + "loss": 0.5183, + "step": 22161 + }, + { + "epoch": 0.98123817008512, + "grad_norm": 1.6567103862762451, + "learning_rate": 2.1562651724884586e-08, + "loss": 0.493, + "step": 22162 + }, + { + "epoch": 0.981282445789934, + "grad_norm": 1.3606270551681519, + "learning_rate": 2.1461336068721028e-08, + "loss": 0.507, + "step": 22163 + }, + { + "epoch": 0.9813267214947478, + "grad_norm": 1.6592634916305542, + "learning_rate": 2.136025874182934e-08, + "loss": 0.5254, + "step": 22164 + }, + { + "epoch": 0.9813709971995617, + "grad_norm": 1.3128446340560913, + "learning_rate": 2.1259419746622045e-08, + "loss": 0.4826, + "step": 22165 + }, + { + "epoch": 0.9814152729043756, + "grad_norm": 1.2187432050704956, + "learning_rate": 2.115881908550832e-08, + "loss": 0.3999, + "step": 22166 + }, + { + "epoch": 0.9814595486091894, + "grad_norm": 1.7365490198135376, + "learning_rate": 2.10584567608918e-08, + "loss": 0.4661, + "step": 22167 + }, + { + "epoch": 0.9815038243140033, + "grad_norm": 2.5456480979919434, + "learning_rate": 2.0958332775168344e-08, + "loss": 0.5599, + "step": 22168 + }, + { + "epoch": 0.9815481000188172, + "grad_norm": 1.5521904230117798, + "learning_rate": 2.0858447130730487e-08, + "loss": 0.4701, + "step": 22169 + }, + { + "epoch": 0.981592375723631, + "grad_norm": 1.3552354574203491, + "learning_rate": 2.0758799829962983e-08, + "loss": 0.4583, + "step": 22170 + }, + { + "epoch": 0.9816366514284449, + "grad_norm": 1.4314484596252441, + "learning_rate": 2.0659390875246154e-08, + "loss": 0.4403, + "step": 22171 + }, + { + "epoch": 0.9816809271332588, + "grad_norm": 1.5565334558486938, + "learning_rate": 2.0560220268954766e-08, + "loss": 0.4334, + "step": 22172 + }, + { + "epoch": 0.9817252028380726, + "grad_norm": 1.7097398042678833, + "learning_rate": 2.0461288013455816e-08, + "loss": 0.5001, + "step": 22173 + }, + { + "epoch": 0.9817694785428865, + "grad_norm": 1.601588249206543, + "learning_rate": 2.0362594111114075e-08, + "loss": 0.4722, + "step": 22174 + }, + { + "epoch": 0.9818137542477005, + "grad_norm": 1.4403924942016602, + "learning_rate": 2.026413856428655e-08, + "loss": 0.4654, + "step": 22175 + }, + { + "epoch": 0.9818580299525143, + "grad_norm": 1.8784455060958862, + "learning_rate": 2.0165921375323583e-08, + "loss": 0.5439, + "step": 22176 + }, + { + "epoch": 0.9819023056573282, + "grad_norm": 1.6202772855758667, + "learning_rate": 2.0067942546572182e-08, + "loss": 0.4521, + "step": 22177 + }, + { + "epoch": 0.9819465813621421, + "grad_norm": 1.4262276887893677, + "learning_rate": 1.9970202080371592e-08, + "loss": 0.4207, + "step": 22178 + }, + { + "epoch": 0.9819908570669559, + "grad_norm": 1.4764162302017212, + "learning_rate": 1.987269997905661e-08, + "loss": 0.5247, + "step": 22179 + }, + { + "epoch": 0.9820351327717698, + "grad_norm": 1.3099688291549683, + "learning_rate": 1.9775436244956482e-08, + "loss": 0.4024, + "step": 22180 + }, + { + "epoch": 0.9820794084765837, + "grad_norm": 1.979981780052185, + "learning_rate": 1.9678410880393794e-08, + "loss": 0.5536, + "step": 22181 + }, + { + "epoch": 0.9821236841813976, + "grad_norm": 1.4663225412368774, + "learning_rate": 1.9581623887685587e-08, + "loss": 0.4466, + "step": 22182 + }, + { + "epoch": 0.9821679598862114, + "grad_norm": 1.446122169494629, + "learning_rate": 1.948507526914445e-08, + "loss": 0.4337, + "step": 22183 + }, + { + "epoch": 0.9822122355910253, + "grad_norm": 1.4791789054870605, + "learning_rate": 1.9388765027075207e-08, + "loss": 0.4837, + "step": 22184 + }, + { + "epoch": 0.9822565112958392, + "grad_norm": 1.4724211692810059, + "learning_rate": 1.9292693163779353e-08, + "loss": 0.4497, + "step": 22185 + }, + { + "epoch": 0.982300787000653, + "grad_norm": 1.647849202156067, + "learning_rate": 1.9196859681550606e-08, + "loss": 0.4853, + "step": 22186 + }, + { + "epoch": 0.982345062705467, + "grad_norm": 1.6053831577301025, + "learning_rate": 1.9101264582678247e-08, + "loss": 0.4588, + "step": 22187 + }, + { + "epoch": 0.9823893384102809, + "grad_norm": 2.055854082107544, + "learning_rate": 1.9005907869446007e-08, + "loss": 0.4872, + "step": 22188 + }, + { + "epoch": 0.9824336141150947, + "grad_norm": 1.4079314470291138, + "learning_rate": 1.8910789544129838e-08, + "loss": 0.4387, + "step": 22189 + }, + { + "epoch": 0.9824778898199086, + "grad_norm": 1.3161804676055908, + "learning_rate": 1.881590960900348e-08, + "loss": 0.4621, + "step": 22190 + }, + { + "epoch": 0.9825221655247225, + "grad_norm": 1.3883715867996216, + "learning_rate": 1.872126806633068e-08, + "loss": 0.4641, + "step": 22191 + }, + { + "epoch": 0.9825664412295363, + "grad_norm": 1.5671311616897583, + "learning_rate": 1.8626864918375176e-08, + "loss": 0.5043, + "step": 22192 + }, + { + "epoch": 0.9826107169343502, + "grad_norm": 1.4615685939788818, + "learning_rate": 1.8532700167388505e-08, + "loss": 0.4594, + "step": 22193 + }, + { + "epoch": 0.9826549926391641, + "grad_norm": 1.7070319652557373, + "learning_rate": 1.84387738156222e-08, + "loss": 0.5153, + "step": 22194 + }, + { + "epoch": 0.9826992683439779, + "grad_norm": 1.2993992567062378, + "learning_rate": 1.83450858653178e-08, + "loss": 0.4035, + "step": 22195 + }, + { + "epoch": 0.9827435440487918, + "grad_norm": 1.5085093975067139, + "learning_rate": 1.8251636318713517e-08, + "loss": 0.4568, + "step": 22196 + }, + { + "epoch": 0.9827878197536057, + "grad_norm": 1.5796126127243042, + "learning_rate": 1.8158425178042004e-08, + "loss": 0.4896, + "step": 22197 + }, + { + "epoch": 0.9828320954584195, + "grad_norm": 2.151096820831299, + "learning_rate": 1.8065452445528152e-08, + "loss": 0.5048, + "step": 22198 + }, + { + "epoch": 0.9828763711632335, + "grad_norm": 1.2534500360488892, + "learning_rate": 1.797271812339463e-08, + "loss": 0.456, + "step": 22199 + }, + { + "epoch": 0.9829206468680474, + "grad_norm": 1.6006903648376465, + "learning_rate": 1.7880222213854104e-08, + "loss": 0.5047, + "step": 22200 + }, + { + "epoch": 0.9829649225728612, + "grad_norm": 1.5886385440826416, + "learning_rate": 1.7787964719117035e-08, + "loss": 0.4734, + "step": 22201 + }, + { + "epoch": 0.9830091982776751, + "grad_norm": 1.6616123914718628, + "learning_rate": 1.7695945641386104e-08, + "loss": 0.5138, + "step": 22202 + }, + { + "epoch": 0.983053473982489, + "grad_norm": 2.0757384300231934, + "learning_rate": 1.7604164982860662e-08, + "loss": 0.5033, + "step": 22203 + }, + { + "epoch": 0.9830977496873028, + "grad_norm": 1.5392353534698486, + "learning_rate": 1.751262274573118e-08, + "loss": 0.4754, + "step": 22204 + }, + { + "epoch": 0.9831420253921167, + "grad_norm": 1.4577903747558594, + "learning_rate": 1.7421318932184795e-08, + "loss": 0.466, + "step": 22205 + }, + { + "epoch": 0.9831863010969306, + "grad_norm": 1.8940907716751099, + "learning_rate": 1.7330253544401987e-08, + "loss": 0.5158, + "step": 22206 + }, + { + "epoch": 0.9832305768017444, + "grad_norm": 1.4492638111114502, + "learning_rate": 1.7239426584557683e-08, + "loss": 0.5125, + "step": 22207 + }, + { + "epoch": 0.9832748525065583, + "grad_norm": 1.553958773612976, + "learning_rate": 1.714883805482126e-08, + "loss": 0.461, + "step": 22208 + }, + { + "epoch": 0.9833191282113722, + "grad_norm": 1.4723734855651855, + "learning_rate": 1.705848795735654e-08, + "loss": 0.4764, + "step": 22209 + }, + { + "epoch": 0.983363403916186, + "grad_norm": 1.4348084926605225, + "learning_rate": 1.6968376294321798e-08, + "loss": 0.4788, + "step": 22210 + }, + { + "epoch": 0.983407679621, + "grad_norm": 1.4270846843719482, + "learning_rate": 1.6878503067868646e-08, + "loss": 0.4367, + "step": 22211 + }, + { + "epoch": 0.9834519553258139, + "grad_norm": 1.6248886585235596, + "learning_rate": 1.6788868280144256e-08, + "loss": 0.5288, + "step": 22212 + }, + { + "epoch": 0.9834962310306277, + "grad_norm": 1.711519718170166, + "learning_rate": 1.6699471933289135e-08, + "loss": 0.5212, + "step": 22213 + }, + { + "epoch": 0.9835405067354416, + "grad_norm": 1.4777625799179077, + "learning_rate": 1.6610314029438245e-08, + "loss": 0.4463, + "step": 22214 + }, + { + "epoch": 0.9835847824402555, + "grad_norm": 2.0064892768859863, + "learning_rate": 1.6521394570720995e-08, + "loss": 0.4868, + "step": 22215 + }, + { + "epoch": 0.9836290581450694, + "grad_norm": 1.4286079406738281, + "learning_rate": 1.6432713559262348e-08, + "loss": 0.4719, + "step": 22216 + }, + { + "epoch": 0.9836733338498832, + "grad_norm": 1.6960946321487427, + "learning_rate": 1.6344270997178392e-08, + "loss": 0.4961, + "step": 22217 + }, + { + "epoch": 0.9837176095546971, + "grad_norm": 1.5756664276123047, + "learning_rate": 1.6256066886582988e-08, + "loss": 0.537, + "step": 22218 + }, + { + "epoch": 0.983761885259511, + "grad_norm": 1.4649232625961304, + "learning_rate": 1.6168101229582234e-08, + "loss": 0.4746, + "step": 22219 + }, + { + "epoch": 0.9838061609643248, + "grad_norm": 1.7677702903747559, + "learning_rate": 1.608037402827667e-08, + "loss": 0.5262, + "step": 22220 + }, + { + "epoch": 0.9838504366691387, + "grad_norm": 1.6717368364334106, + "learning_rate": 1.5992885284762395e-08, + "loss": 0.4714, + "step": 22221 + }, + { + "epoch": 0.9838947123739527, + "grad_norm": 2.1155965328216553, + "learning_rate": 1.590563500112885e-08, + "loss": 0.474, + "step": 22222 + }, + { + "epoch": 0.9839389880787665, + "grad_norm": 1.3963227272033691, + "learning_rate": 1.5818623179459924e-08, + "loss": 0.4393, + "step": 22223 + }, + { + "epoch": 0.9839832637835804, + "grad_norm": 1.5123913288116455, + "learning_rate": 1.5731849821833955e-08, + "loss": 0.4535, + "step": 22224 + }, + { + "epoch": 0.9840275394883943, + "grad_norm": 1.6430541276931763, + "learning_rate": 1.5645314930323728e-08, + "loss": 0.5083, + "step": 22225 + }, + { + "epoch": 0.9840718151932081, + "grad_norm": 1.5212496519088745, + "learning_rate": 1.5559018506995372e-08, + "loss": 0.5344, + "step": 22226 + }, + { + "epoch": 0.984116090898022, + "grad_norm": 1.7334307432174683, + "learning_rate": 1.5472960553909456e-08, + "loss": 0.5418, + "step": 22227 + }, + { + "epoch": 0.9841603666028359, + "grad_norm": 1.5402406454086304, + "learning_rate": 1.5387141073122114e-08, + "loss": 0.5077, + "step": 22228 + }, + { + "epoch": 0.9842046423076497, + "grad_norm": 1.3175486326217651, + "learning_rate": 1.530156006668393e-08, + "loss": 0.4396, + "step": 22229 + }, + { + "epoch": 0.9842489180124636, + "grad_norm": 2.816136121749878, + "learning_rate": 1.5216217536637712e-08, + "loss": 0.5553, + "step": 22230 + }, + { + "epoch": 0.9842931937172775, + "grad_norm": 1.945406198501587, + "learning_rate": 1.5131113485022942e-08, + "loss": 0.5315, + "step": 22231 + }, + { + "epoch": 0.9843374694220913, + "grad_norm": 1.5534735918045044, + "learning_rate": 1.5046247913870217e-08, + "loss": 0.4379, + "step": 22232 + }, + { + "epoch": 0.9843817451269052, + "grad_norm": 1.4305027723312378, + "learning_rate": 1.496162082520902e-08, + "loss": 0.4584, + "step": 22233 + }, + { + "epoch": 0.9844260208317192, + "grad_norm": 1.8165807723999023, + "learning_rate": 1.4877232221058856e-08, + "loss": 0.5442, + "step": 22234 + }, + { + "epoch": 0.984470296536533, + "grad_norm": 1.4122133255004883, + "learning_rate": 1.4793082103435885e-08, + "loss": 0.4347, + "step": 22235 + }, + { + "epoch": 0.9845145722413469, + "grad_norm": 1.519766926765442, + "learning_rate": 1.470917047435072e-08, + "loss": 0.4833, + "step": 22236 + }, + { + "epoch": 0.9845588479461608, + "grad_norm": 1.4616780281066895, + "learning_rate": 1.4625497335805095e-08, + "loss": 0.4586, + "step": 22237 + }, + { + "epoch": 0.9846031236509746, + "grad_norm": 1.5776312351226807, + "learning_rate": 1.454206268979963e-08, + "loss": 0.4722, + "step": 22238 + }, + { + "epoch": 0.9846473993557885, + "grad_norm": 1.3036844730377197, + "learning_rate": 1.4458866538327177e-08, + "loss": 0.4552, + "step": 22239 + }, + { + "epoch": 0.9846916750606024, + "grad_norm": 1.4109773635864258, + "learning_rate": 1.4375908883373924e-08, + "loss": 0.4781, + "step": 22240 + }, + { + "epoch": 0.9847359507654162, + "grad_norm": 1.438690423965454, + "learning_rate": 1.4293189726920509e-08, + "loss": 0.4759, + "step": 22241 + }, + { + "epoch": 0.9847802264702301, + "grad_norm": 1.192680835723877, + "learning_rate": 1.421070907094424e-08, + "loss": 0.4159, + "step": 22242 + }, + { + "epoch": 0.984824502175044, + "grad_norm": 1.5720272064208984, + "learning_rate": 1.4128466917414652e-08, + "loss": 0.4983, + "step": 22243 + }, + { + "epoch": 0.9848687778798578, + "grad_norm": 1.9563385248184204, + "learning_rate": 1.404646326829684e-08, + "loss": 0.5127, + "step": 22244 + }, + { + "epoch": 0.9849130535846717, + "grad_norm": 1.6129940748214722, + "learning_rate": 1.3964698125547016e-08, + "loss": 0.4882, + "step": 22245 + }, + { + "epoch": 0.9849573292894857, + "grad_norm": 1.5845993757247925, + "learning_rate": 1.3883171491119173e-08, + "loss": 0.4529, + "step": 22246 + }, + { + "epoch": 0.9850016049942995, + "grad_norm": 1.595485806465149, + "learning_rate": 1.3801883366961756e-08, + "loss": 0.4882, + "step": 22247 + }, + { + "epoch": 0.9850458806991134, + "grad_norm": 1.4608160257339478, + "learning_rate": 1.3720833755014318e-08, + "loss": 0.484, + "step": 22248 + }, + { + "epoch": 0.9850901564039273, + "grad_norm": 1.1595028638839722, + "learning_rate": 1.3640022657215313e-08, + "loss": 0.3862, + "step": 22249 + }, + { + "epoch": 0.9851344321087411, + "grad_norm": 1.6186422109603882, + "learning_rate": 1.3559450075492086e-08, + "loss": 0.4829, + "step": 22250 + }, + { + "epoch": 0.985178707813555, + "grad_norm": 1.6667464971542358, + "learning_rate": 1.3479116011769766e-08, + "loss": 0.4591, + "step": 22251 + }, + { + "epoch": 0.9852229835183689, + "grad_norm": 1.4711095094680786, + "learning_rate": 1.3399020467967927e-08, + "loss": 0.4921, + "step": 22252 + }, + { + "epoch": 0.9852672592231828, + "grad_norm": 1.792325496673584, + "learning_rate": 1.3319163445998373e-08, + "loss": 0.5046, + "step": 22253 + }, + { + "epoch": 0.9853115349279966, + "grad_norm": 1.2065445184707642, + "learning_rate": 1.3239544947770688e-08, + "loss": 0.3904, + "step": 22254 + }, + { + "epoch": 0.9853558106328105, + "grad_norm": 1.5679047107696533, + "learning_rate": 1.3160164975183354e-08, + "loss": 0.5129, + "step": 22255 + }, + { + "epoch": 0.9854000863376244, + "grad_norm": 1.404400110244751, + "learning_rate": 1.308102353013485e-08, + "loss": 0.4541, + "step": 22256 + }, + { + "epoch": 0.9854443620424382, + "grad_norm": 1.5194189548492432, + "learning_rate": 1.300212061451367e-08, + "loss": 0.5015, + "step": 22257 + }, + { + "epoch": 0.9854886377472521, + "grad_norm": 1.664801001548767, + "learning_rate": 1.2923456230206077e-08, + "loss": 0.512, + "step": 22258 + }, + { + "epoch": 0.9855329134520661, + "grad_norm": 1.567769169807434, + "learning_rate": 1.2845030379089463e-08, + "loss": 0.4878, + "step": 22259 + }, + { + "epoch": 0.9855771891568799, + "grad_norm": 1.3772656917572021, + "learning_rate": 1.2766843063036772e-08, + "loss": 0.4216, + "step": 22260 + }, + { + "epoch": 0.9856214648616938, + "grad_norm": 1.7433878183364868, + "learning_rate": 1.2688894283916508e-08, + "loss": 0.5307, + "step": 22261 + }, + { + "epoch": 0.9856657405665077, + "grad_norm": 1.415795922279358, + "learning_rate": 1.261118404359052e-08, + "loss": 0.4803, + "step": 22262 + }, + { + "epoch": 0.9857100162713215, + "grad_norm": 1.4182699918746948, + "learning_rate": 1.2533712343915094e-08, + "loss": 0.4488, + "step": 22263 + }, + { + "epoch": 0.9857542919761354, + "grad_norm": 1.485737919807434, + "learning_rate": 1.2456479186738758e-08, + "loss": 0.4861, + "step": 22264 + }, + { + "epoch": 0.9857985676809493, + "grad_norm": 1.9368650913238525, + "learning_rate": 1.2379484573906697e-08, + "loss": 0.548, + "step": 22265 + }, + { + "epoch": 0.9858428433857631, + "grad_norm": 1.6244211196899414, + "learning_rate": 1.2302728507259665e-08, + "loss": 0.5197, + "step": 22266 + }, + { + "epoch": 0.985887119090577, + "grad_norm": 1.7259141206741333, + "learning_rate": 1.2226210988628418e-08, + "loss": 0.5096, + "step": 22267 + }, + { + "epoch": 0.9859313947953909, + "grad_norm": 1.3769416809082031, + "learning_rate": 1.2149932019842603e-08, + "loss": 0.4619, + "step": 22268 + }, + { + "epoch": 0.9859756705002047, + "grad_norm": 1.4189727306365967, + "learning_rate": 1.2073891602722987e-08, + "loss": 0.5164, + "step": 22269 + }, + { + "epoch": 0.9860199462050186, + "grad_norm": 1.5859113931655884, + "learning_rate": 1.1998089739087005e-08, + "loss": 0.4661, + "step": 22270 + }, + { + "epoch": 0.9860642219098326, + "grad_norm": 1.8547546863555908, + "learning_rate": 1.1922526430743208e-08, + "loss": 0.4925, + "step": 22271 + }, + { + "epoch": 0.9861084976146464, + "grad_norm": 1.5576685667037964, + "learning_rate": 1.1847201679496823e-08, + "loss": 0.4278, + "step": 22272 + }, + { + "epoch": 0.9861527733194603, + "grad_norm": 1.6493474245071411, + "learning_rate": 1.1772115487148627e-08, + "loss": 0.4272, + "step": 22273 + }, + { + "epoch": 0.9861970490242742, + "grad_norm": 1.7089283466339111, + "learning_rate": 1.1697267855489414e-08, + "loss": 0.4752, + "step": 22274 + }, + { + "epoch": 0.986241324729088, + "grad_norm": 1.318192720413208, + "learning_rate": 1.162265878630886e-08, + "loss": 0.4331, + "step": 22275 + }, + { + "epoch": 0.9862856004339019, + "grad_norm": 1.3302841186523438, + "learning_rate": 1.1548288281387766e-08, + "loss": 0.4198, + "step": 22276 + }, + { + "epoch": 0.9863298761387158, + "grad_norm": 1.3350961208343506, + "learning_rate": 1.1474156342503595e-08, + "loss": 0.4477, + "step": 22277 + }, + { + "epoch": 0.9863741518435296, + "grad_norm": 1.6302789449691772, + "learning_rate": 1.1400262971426046e-08, + "loss": 0.4714, + "step": 22278 + }, + { + "epoch": 0.9864184275483435, + "grad_norm": 1.9417569637298584, + "learning_rate": 1.1326608169920373e-08, + "loss": 0.4882, + "step": 22279 + }, + { + "epoch": 0.9864627032531574, + "grad_norm": 1.5157220363616943, + "learning_rate": 1.1253191939745168e-08, + "loss": 0.4897, + "step": 22280 + }, + { + "epoch": 0.9865069789579712, + "grad_norm": 1.6814607381820679, + "learning_rate": 1.1180014282654583e-08, + "loss": 0.4876, + "step": 22281 + }, + { + "epoch": 0.9865512546627851, + "grad_norm": 1.5063754320144653, + "learning_rate": 1.1107075200396111e-08, + "loss": 0.4867, + "step": 22282 + }, + { + "epoch": 0.9865955303675991, + "grad_norm": 2.1524946689605713, + "learning_rate": 1.1034374694711692e-08, + "loss": 0.5526, + "step": 22283 + }, + { + "epoch": 0.9866398060724129, + "grad_norm": 1.5710911750793457, + "learning_rate": 1.0961912767338823e-08, + "loss": 0.5104, + "step": 22284 + }, + { + "epoch": 0.9866840817772268, + "grad_norm": 1.4739471673965454, + "learning_rate": 1.0889689420006121e-08, + "loss": 0.454, + "step": 22285 + }, + { + "epoch": 0.9867283574820407, + "grad_norm": 1.6621743440628052, + "learning_rate": 1.0817704654439987e-08, + "loss": 0.4763, + "step": 22286 + }, + { + "epoch": 0.9867726331868546, + "grad_norm": 1.7893751859664917, + "learning_rate": 1.0745958472360151e-08, + "loss": 0.4379, + "step": 22287 + }, + { + "epoch": 0.9868169088916684, + "grad_norm": 2.3308017253875732, + "learning_rate": 1.0674450875478581e-08, + "loss": 0.5121, + "step": 22288 + }, + { + "epoch": 0.9868611845964823, + "grad_norm": 1.6885167360305786, + "learning_rate": 1.0603181865503909e-08, + "loss": 0.4822, + "step": 22289 + }, + { + "epoch": 0.9869054603012962, + "grad_norm": 1.4844920635223389, + "learning_rate": 1.0532151444140326e-08, + "loss": 0.494, + "step": 22290 + }, + { + "epoch": 0.98694973600611, + "grad_norm": 1.4809712171554565, + "learning_rate": 1.0461359613080924e-08, + "loss": 0.4416, + "step": 22291 + }, + { + "epoch": 0.9869940117109239, + "grad_norm": 1.478630542755127, + "learning_rate": 1.0390806374018791e-08, + "loss": 0.4782, + "step": 22292 + }, + { + "epoch": 0.9870382874157378, + "grad_norm": 1.7470108270645142, + "learning_rate": 1.0320491728638138e-08, + "loss": 0.4768, + "step": 22293 + }, + { + "epoch": 0.9870825631205516, + "grad_norm": 1.3950741291046143, + "learning_rate": 1.025041567861984e-08, + "loss": 0.4291, + "step": 22294 + }, + { + "epoch": 0.9871268388253656, + "grad_norm": 1.5262538194656372, + "learning_rate": 1.0180578225634785e-08, + "loss": 0.416, + "step": 22295 + }, + { + "epoch": 0.9871711145301795, + "grad_norm": 1.562753438949585, + "learning_rate": 1.0110979371353857e-08, + "loss": 0.4967, + "step": 22296 + }, + { + "epoch": 0.9872153902349933, + "grad_norm": 2.697026014328003, + "learning_rate": 1.004161911743684e-08, + "loss": 0.5551, + "step": 22297 + }, + { + "epoch": 0.9872596659398072, + "grad_norm": 1.653949499130249, + "learning_rate": 9.972497465542408e-09, + "loss": 0.5217, + "step": 22298 + }, + { + "epoch": 0.9873039416446211, + "grad_norm": 1.744337797164917, + "learning_rate": 9.903614417320351e-09, + "loss": 0.4817, + "step": 22299 + }, + { + "epoch": 0.9873482173494349, + "grad_norm": 1.5435618162155151, + "learning_rate": 9.83496997441713e-09, + "loss": 0.4562, + "step": 22300 + }, + { + "epoch": 0.9873924930542488, + "grad_norm": 1.2491425275802612, + "learning_rate": 9.766564138470325e-09, + "loss": 0.4175, + "step": 22301 + }, + { + "epoch": 0.9874367687590627, + "grad_norm": 1.6089712381362915, + "learning_rate": 9.698396911115293e-09, + "loss": 0.6, + "step": 22302 + }, + { + "epoch": 0.9874810444638765, + "grad_norm": 1.3050146102905273, + "learning_rate": 9.630468293979622e-09, + "loss": 0.4475, + "step": 22303 + }, + { + "epoch": 0.9875253201686904, + "grad_norm": 1.8354997634887695, + "learning_rate": 9.562778288685349e-09, + "loss": 0.5758, + "step": 22304 + }, + { + "epoch": 0.9875695958735043, + "grad_norm": 1.4117491245269775, + "learning_rate": 9.495326896850065e-09, + "loss": 0.4484, + "step": 22305 + }, + { + "epoch": 0.9876138715783181, + "grad_norm": 1.6381940841674805, + "learning_rate": 9.428114120083598e-09, + "loss": 0.5333, + "step": 22306 + }, + { + "epoch": 0.9876581472831321, + "grad_norm": 1.3803496360778809, + "learning_rate": 9.361139959993549e-09, + "loss": 0.4421, + "step": 22307 + }, + { + "epoch": 0.987702422987946, + "grad_norm": 1.4621206521987915, + "learning_rate": 9.294404418176417e-09, + "loss": 0.4625, + "step": 22308 + }, + { + "epoch": 0.9877466986927598, + "grad_norm": 1.7567119598388672, + "learning_rate": 9.227907496228705e-09, + "loss": 0.5329, + "step": 22309 + }, + { + "epoch": 0.9877909743975737, + "grad_norm": 1.8110772371292114, + "learning_rate": 9.161649195736921e-09, + "loss": 0.5271, + "step": 22310 + }, + { + "epoch": 0.9878352501023876, + "grad_norm": 1.650841474533081, + "learning_rate": 9.095629518284244e-09, + "loss": 0.5112, + "step": 22311 + }, + { + "epoch": 0.9878795258072014, + "grad_norm": 1.7813632488250732, + "learning_rate": 9.029848465448299e-09, + "loss": 0.4635, + "step": 22312 + }, + { + "epoch": 0.9879238015120153, + "grad_norm": 1.3720356225967407, + "learning_rate": 8.964306038798943e-09, + "loss": 0.4547, + "step": 22313 + }, + { + "epoch": 0.9879680772168292, + "grad_norm": 1.4086456298828125, + "learning_rate": 8.899002239902698e-09, + "loss": 0.4577, + "step": 22314 + }, + { + "epoch": 0.988012352921643, + "grad_norm": 1.5914957523345947, + "learning_rate": 8.83393707031832e-09, + "loss": 0.4928, + "step": 22315 + }, + { + "epoch": 0.9880566286264569, + "grad_norm": 1.6160304546356201, + "learning_rate": 8.769110531600122e-09, + "loss": 0.5042, + "step": 22316 + }, + { + "epoch": 0.9881009043312708, + "grad_norm": 1.4602856636047363, + "learning_rate": 8.704522625296863e-09, + "loss": 0.5037, + "step": 22317 + }, + { + "epoch": 0.9881451800360846, + "grad_norm": 1.4763047695159912, + "learning_rate": 8.640173352950642e-09, + "loss": 0.4788, + "step": 22318 + }, + { + "epoch": 0.9881894557408986, + "grad_norm": 1.5351519584655762, + "learning_rate": 8.57606271609801e-09, + "loss": 0.4471, + "step": 22319 + }, + { + "epoch": 0.9882337314457125, + "grad_norm": 1.3467910289764404, + "learning_rate": 8.512190716272184e-09, + "loss": 0.4325, + "step": 22320 + }, + { + "epoch": 0.9882780071505263, + "grad_norm": 1.4898871183395386, + "learning_rate": 8.44855735499639e-09, + "loss": 0.4546, + "step": 22321 + }, + { + "epoch": 0.9883222828553402, + "grad_norm": 1.6578449010849, + "learning_rate": 8.385162633791633e-09, + "loss": 0.463, + "step": 22322 + }, + { + "epoch": 0.9883665585601541, + "grad_norm": 1.580090045928955, + "learning_rate": 8.322006554171147e-09, + "loss": 0.5222, + "step": 22323 + }, + { + "epoch": 0.988410834264968, + "grad_norm": 1.6834040880203247, + "learning_rate": 8.259089117644836e-09, + "loss": 0.4952, + "step": 22324 + }, + { + "epoch": 0.9884551099697818, + "grad_norm": 1.472314715385437, + "learning_rate": 8.196410325713721e-09, + "loss": 0.496, + "step": 22325 + }, + { + "epoch": 0.9884993856745957, + "grad_norm": 1.5324426889419556, + "learning_rate": 8.133970179876604e-09, + "loss": 0.4904, + "step": 22326 + }, + { + "epoch": 0.9885436613794096, + "grad_norm": 1.7329524755477905, + "learning_rate": 8.071768681623404e-09, + "loss": 0.5268, + "step": 22327 + }, + { + "epoch": 0.9885879370842234, + "grad_norm": 1.3644071817398071, + "learning_rate": 8.00980583244071e-09, + "loss": 0.4392, + "step": 22328 + }, + { + "epoch": 0.9886322127890373, + "grad_norm": 1.3613839149475098, + "learning_rate": 7.948081633808447e-09, + "loss": 0.4859, + "step": 22329 + }, + { + "epoch": 0.9886764884938513, + "grad_norm": 1.6696324348449707, + "learning_rate": 7.886596087199883e-09, + "loss": 0.5063, + "step": 22330 + }, + { + "epoch": 0.9887207641986651, + "grad_norm": 1.5027892589569092, + "learning_rate": 7.825349194083842e-09, + "loss": 0.4749, + "step": 22331 + }, + { + "epoch": 0.988765039903479, + "grad_norm": 1.851267695426941, + "learning_rate": 7.764340955923599e-09, + "loss": 0.5328, + "step": 22332 + }, + { + "epoch": 0.9888093156082929, + "grad_norm": 1.4489518404006958, + "learning_rate": 7.703571374175767e-09, + "loss": 0.5204, + "step": 22333 + }, + { + "epoch": 0.9888535913131067, + "grad_norm": 1.5399410724639893, + "learning_rate": 7.643040450292516e-09, + "loss": 0.461, + "step": 22334 + }, + { + "epoch": 0.9888978670179206, + "grad_norm": 1.493622064590454, + "learning_rate": 7.582748185719357e-09, + "loss": 0.3971, + "step": 22335 + }, + { + "epoch": 0.9889421427227345, + "grad_norm": 1.4019618034362793, + "learning_rate": 7.52269458189514e-09, + "loss": 0.4404, + "step": 22336 + }, + { + "epoch": 0.9889864184275483, + "grad_norm": 1.5627473592758179, + "learning_rate": 7.462879640256493e-09, + "loss": 0.511, + "step": 22337 + }, + { + "epoch": 0.9890306941323622, + "grad_norm": 1.7113455533981323, + "learning_rate": 7.403303362230052e-09, + "loss": 0.5236, + "step": 22338 + }, + { + "epoch": 0.9890749698371761, + "grad_norm": 1.5830355882644653, + "learning_rate": 7.343965749239124e-09, + "loss": 0.4657, + "step": 22339 + }, + { + "epoch": 0.9891192455419899, + "grad_norm": 2.1599767208099365, + "learning_rate": 7.284866802702573e-09, + "loss": 0.5547, + "step": 22340 + }, + { + "epoch": 0.9891635212468038, + "grad_norm": 1.524644136428833, + "learning_rate": 7.226006524029272e-09, + "loss": 0.4715, + "step": 22341 + }, + { + "epoch": 0.9892077969516178, + "grad_norm": 1.5320016145706177, + "learning_rate": 7.167384914626984e-09, + "loss": 0.5116, + "step": 22342 + }, + { + "epoch": 0.9892520726564316, + "grad_norm": 1.3933089971542358, + "learning_rate": 7.109001975894591e-09, + "loss": 0.4347, + "step": 22343 + }, + { + "epoch": 0.9892963483612455, + "grad_norm": 1.419516682624817, + "learning_rate": 7.050857709227643e-09, + "loss": 0.4412, + "step": 22344 + }, + { + "epoch": 0.9893406240660594, + "grad_norm": 1.487239956855774, + "learning_rate": 6.992952116013918e-09, + "loss": 0.4625, + "step": 22345 + }, + { + "epoch": 0.9893848997708732, + "grad_norm": 1.7626749277114868, + "learning_rate": 6.935285197637865e-09, + "loss": 0.5573, + "step": 22346 + }, + { + "epoch": 0.9894291754756871, + "grad_norm": 1.178918480873108, + "learning_rate": 6.8778569554750484e-09, + "loss": 0.4043, + "step": 22347 + }, + { + "epoch": 0.989473451180501, + "grad_norm": 1.4963572025299072, + "learning_rate": 6.8206673908988165e-09, + "loss": 0.5122, + "step": 22348 + }, + { + "epoch": 0.9895177268853148, + "grad_norm": 1.757248044013977, + "learning_rate": 6.763716505273632e-09, + "loss": 0.5269, + "step": 22349 + }, + { + "epoch": 0.9895620025901287, + "grad_norm": 1.5573124885559082, + "learning_rate": 6.7070042999606285e-09, + "loss": 0.4759, + "step": 22350 + }, + { + "epoch": 0.9896062782949426, + "grad_norm": 1.4880632162094116, + "learning_rate": 6.650530776313169e-09, + "loss": 0.4661, + "step": 22351 + }, + { + "epoch": 0.9896505539997564, + "grad_norm": 1.591770887374878, + "learning_rate": 6.594295935681283e-09, + "loss": 0.5451, + "step": 22352 + }, + { + "epoch": 0.9896948297045703, + "grad_norm": 1.7893425226211548, + "learning_rate": 6.538299779408341e-09, + "loss": 0.5438, + "step": 22353 + }, + { + "epoch": 0.9897391054093843, + "grad_norm": 1.3401025533676147, + "learning_rate": 6.482542308829942e-09, + "loss": 0.4291, + "step": 22354 + }, + { + "epoch": 0.9897833811141981, + "grad_norm": 1.9659483432769775, + "learning_rate": 6.427023525279464e-09, + "loss": 0.5054, + "step": 22355 + }, + { + "epoch": 0.989827656819012, + "grad_norm": 1.46356999874115, + "learning_rate": 6.371743430082511e-09, + "loss": 0.4755, + "step": 22356 + }, + { + "epoch": 0.9898719325238259, + "grad_norm": 1.6063910722732544, + "learning_rate": 6.31670202456025e-09, + "loss": 0.5075, + "step": 22357 + }, + { + "epoch": 0.9899162082286398, + "grad_norm": 1.6240508556365967, + "learning_rate": 6.261899310024966e-09, + "loss": 0.5458, + "step": 22358 + }, + { + "epoch": 0.9899604839334536, + "grad_norm": 1.3151899576187134, + "learning_rate": 6.2073352877878305e-09, + "loss": 0.451, + "step": 22359 + }, + { + "epoch": 0.9900047596382675, + "grad_norm": 1.5535602569580078, + "learning_rate": 6.153009959151135e-09, + "loss": 0.4617, + "step": 22360 + }, + { + "epoch": 0.9900490353430814, + "grad_norm": 2.0001308917999268, + "learning_rate": 6.098923325412731e-09, + "loss": 0.4979, + "step": 22361 + }, + { + "epoch": 0.9900933110478952, + "grad_norm": 2.2914912700653076, + "learning_rate": 6.045075387863808e-09, + "loss": 0.497, + "step": 22362 + }, + { + "epoch": 0.9901375867527091, + "grad_norm": 1.448542833328247, + "learning_rate": 5.991466147791114e-09, + "loss": 0.4554, + "step": 22363 + }, + { + "epoch": 0.990181862457523, + "grad_norm": 1.451878547668457, + "learning_rate": 5.938095606474736e-09, + "loss": 0.4684, + "step": 22364 + }, + { + "epoch": 0.9902261381623368, + "grad_norm": 1.602634072303772, + "learning_rate": 5.884963765189211e-09, + "loss": 0.4479, + "step": 22365 + }, + { + "epoch": 0.9902704138671508, + "grad_norm": 1.7088745832443237, + "learning_rate": 5.832070625204633e-09, + "loss": 0.545, + "step": 22366 + }, + { + "epoch": 0.9903146895719647, + "grad_norm": 1.5028077363967896, + "learning_rate": 5.779416187783327e-09, + "loss": 0.4893, + "step": 22367 + }, + { + "epoch": 0.9903589652767785, + "grad_norm": 1.3511161804199219, + "learning_rate": 5.727000454182064e-09, + "loss": 0.3916, + "step": 22368 + }, + { + "epoch": 0.9904032409815924, + "grad_norm": 1.4311283826828003, + "learning_rate": 5.674823425654286e-09, + "loss": 0.529, + "step": 22369 + }, + { + "epoch": 0.9904475166864063, + "grad_norm": 1.4896106719970703, + "learning_rate": 5.622885103445663e-09, + "loss": 0.4735, + "step": 22370 + }, + { + "epoch": 0.9904917923912201, + "grad_norm": 1.6024302244186401, + "learning_rate": 5.571185488797426e-09, + "loss": 0.5043, + "step": 22371 + }, + { + "epoch": 0.990536068096034, + "grad_norm": 1.6480072736740112, + "learning_rate": 5.519724582941921e-09, + "loss": 0.4717, + "step": 22372 + }, + { + "epoch": 0.9905803438008479, + "grad_norm": 1.4388772249221802, + "learning_rate": 5.4685023871114965e-09, + "loss": 0.4632, + "step": 22373 + }, + { + "epoch": 0.9906246195056617, + "grad_norm": 1.4713832139968872, + "learning_rate": 5.417518902527396e-09, + "loss": 0.4529, + "step": 22374 + }, + { + "epoch": 0.9906688952104756, + "grad_norm": 1.4724708795547485, + "learning_rate": 5.3667741304086476e-09, + "loss": 0.4379, + "step": 22375 + }, + { + "epoch": 0.9907131709152895, + "grad_norm": 1.552627682685852, + "learning_rate": 5.316268071966501e-09, + "loss": 0.5328, + "step": 22376 + }, + { + "epoch": 0.9907574466201033, + "grad_norm": 1.6772810220718384, + "learning_rate": 5.266000728406662e-09, + "loss": 0.4494, + "step": 22377 + }, + { + "epoch": 0.9908017223249173, + "grad_norm": 1.8889813423156738, + "learning_rate": 5.215972100930389e-09, + "loss": 0.5191, + "step": 22378 + }, + { + "epoch": 0.9908459980297312, + "grad_norm": 1.5870890617370605, + "learning_rate": 5.166182190733393e-09, + "loss": 0.4724, + "step": 22379 + }, + { + "epoch": 0.990890273734545, + "grad_norm": 1.6601301431655884, + "learning_rate": 5.116630999003613e-09, + "loss": 0.4807, + "step": 22380 + }, + { + "epoch": 0.9909345494393589, + "grad_norm": 1.5085275173187256, + "learning_rate": 5.0673185269245475e-09, + "loss": 0.4523, + "step": 22381 + }, + { + "epoch": 0.9909788251441728, + "grad_norm": 1.6882679462432861, + "learning_rate": 5.018244775674141e-09, + "loss": 0.5256, + "step": 22382 + }, + { + "epoch": 0.9910231008489866, + "grad_norm": 1.3701978921890259, + "learning_rate": 4.969409746425902e-09, + "loss": 0.4526, + "step": 22383 + }, + { + "epoch": 0.9910673765538005, + "grad_norm": 1.724445104598999, + "learning_rate": 4.920813440344452e-09, + "loss": 0.4728, + "step": 22384 + }, + { + "epoch": 0.9911116522586144, + "grad_norm": 1.6573158502578735, + "learning_rate": 4.872455858591085e-09, + "loss": 0.4856, + "step": 22385 + }, + { + "epoch": 0.9911559279634282, + "grad_norm": 1.5336271524429321, + "learning_rate": 4.824337002320434e-09, + "loss": 0.5133, + "step": 22386 + }, + { + "epoch": 0.9912002036682421, + "grad_norm": 1.3824543952941895, + "learning_rate": 4.776456872682689e-09, + "loss": 0.4271, + "step": 22387 + }, + { + "epoch": 0.991244479373056, + "grad_norm": 1.5687274932861328, + "learning_rate": 4.72881547082027e-09, + "loss": 0.4766, + "step": 22388 + }, + { + "epoch": 0.9912887550778698, + "grad_norm": 1.6235992908477783, + "learning_rate": 4.681412797872265e-09, + "loss": 0.4909, + "step": 22389 + }, + { + "epoch": 0.9913330307826838, + "grad_norm": 1.4547221660614014, + "learning_rate": 4.634248854969992e-09, + "loss": 0.5021, + "step": 22390 + }, + { + "epoch": 0.9913773064874977, + "grad_norm": 1.6498768329620361, + "learning_rate": 4.587323643240327e-09, + "loss": 0.5007, + "step": 22391 + }, + { + "epoch": 0.9914215821923116, + "grad_norm": 1.7235852479934692, + "learning_rate": 4.5406371638034856e-09, + "loss": 0.4621, + "step": 22392 + }, + { + "epoch": 0.9914658578971254, + "grad_norm": 1.397407054901123, + "learning_rate": 4.494189417775241e-09, + "loss": 0.469, + "step": 22393 + }, + { + "epoch": 0.9915101336019393, + "grad_norm": 1.7837480306625366, + "learning_rate": 4.447980406263597e-09, + "loss": 0.5318, + "step": 22394 + }, + { + "epoch": 0.9915544093067532, + "grad_norm": 1.6433608531951904, + "learning_rate": 4.4020101303743345e-09, + "loss": 0.492, + "step": 22395 + }, + { + "epoch": 0.991598685011567, + "grad_norm": 1.2791177034378052, + "learning_rate": 4.356278591204355e-09, + "loss": 0.4278, + "step": 22396 + }, + { + "epoch": 0.9916429607163809, + "grad_norm": 1.468368649482727, + "learning_rate": 4.310785789846117e-09, + "loss": 0.5052, + "step": 22397 + }, + { + "epoch": 0.9916872364211948, + "grad_norm": 1.6783561706542969, + "learning_rate": 4.265531727385419e-09, + "loss": 0.5275, + "step": 22398 + }, + { + "epoch": 0.9917315121260086, + "grad_norm": 1.3267816305160522, + "learning_rate": 4.220516404903618e-09, + "loss": 0.4311, + "step": 22399 + }, + { + "epoch": 0.9917757878308225, + "grad_norm": 2.4318931102752686, + "learning_rate": 4.175739823476521e-09, + "loss": 0.4336, + "step": 22400 + }, + { + "epoch": 0.9918200635356365, + "grad_norm": 1.6892657279968262, + "learning_rate": 4.131201984172162e-09, + "loss": 0.5138, + "step": 22401 + }, + { + "epoch": 0.9918643392404503, + "grad_norm": 1.5043045282363892, + "learning_rate": 4.086902888055244e-09, + "loss": 0.4632, + "step": 22402 + }, + { + "epoch": 0.9919086149452642, + "grad_norm": 1.7624236345291138, + "learning_rate": 4.042842536183811e-09, + "loss": 0.4906, + "step": 22403 + }, + { + "epoch": 0.9919528906500781, + "grad_norm": 1.552438735961914, + "learning_rate": 3.999020929610354e-09, + "loss": 0.4783, + "step": 22404 + }, + { + "epoch": 0.9919971663548919, + "grad_norm": 1.6324936151504517, + "learning_rate": 3.9554380693807015e-09, + "loss": 0.4999, + "step": 22405 + }, + { + "epoch": 0.9920414420597058, + "grad_norm": 2.2070462703704834, + "learning_rate": 3.912093956537355e-09, + "loss": 0.53, + "step": 22406 + }, + { + "epoch": 0.9920857177645197, + "grad_norm": 1.3287081718444824, + "learning_rate": 3.868988592113932e-09, + "loss": 0.3934, + "step": 22407 + }, + { + "epoch": 0.9921299934693335, + "grad_norm": 1.5437021255493164, + "learning_rate": 3.826121977140718e-09, + "loss": 0.5104, + "step": 22408 + }, + { + "epoch": 0.9921742691741474, + "grad_norm": 1.4144647121429443, + "learning_rate": 3.783494112640229e-09, + "loss": 0.4922, + "step": 22409 + }, + { + "epoch": 0.9922185448789613, + "grad_norm": 1.3924225568771362, + "learning_rate": 3.7411049996327605e-09, + "loss": 0.419, + "step": 22410 + }, + { + "epoch": 0.9922628205837751, + "grad_norm": 1.4360369443893433, + "learning_rate": 3.698954639129726e-09, + "loss": 0.4854, + "step": 22411 + }, + { + "epoch": 0.992307096288589, + "grad_norm": 1.4483367204666138, + "learning_rate": 3.657043032138097e-09, + "loss": 0.4737, + "step": 22412 + }, + { + "epoch": 0.992351371993403, + "grad_norm": 1.7829704284667969, + "learning_rate": 3.6153701796581842e-09, + "loss": 0.5185, + "step": 22413 + }, + { + "epoch": 0.9923956476982168, + "grad_norm": 1.3785994052886963, + "learning_rate": 3.5739360826869685e-09, + "loss": 0.4321, + "step": 22414 + }, + { + "epoch": 0.9924399234030307, + "grad_norm": 1.3997654914855957, + "learning_rate": 3.532740742211438e-09, + "loss": 0.4834, + "step": 22415 + }, + { + "epoch": 0.9924841991078446, + "grad_norm": 1.4531155824661255, + "learning_rate": 3.4917841592174706e-09, + "loss": 0.4823, + "step": 22416 + }, + { + "epoch": 0.9925284748126584, + "grad_norm": 1.8359763622283936, + "learning_rate": 3.451066334683173e-09, + "loss": 0.5384, + "step": 22417 + }, + { + "epoch": 0.9925727505174723, + "grad_norm": 1.8248637914657593, + "learning_rate": 3.4105872695799903e-09, + "loss": 0.4889, + "step": 22418 + }, + { + "epoch": 0.9926170262222862, + "grad_norm": 1.2837145328521729, + "learning_rate": 3.3703469648760367e-09, + "loss": 0.4363, + "step": 22419 + }, + { + "epoch": 0.9926613019271, + "grad_norm": 1.4945226907730103, + "learning_rate": 3.330345421530545e-09, + "loss": 0.478, + "step": 22420 + }, + { + "epoch": 0.9927055776319139, + "grad_norm": 1.855704665184021, + "learning_rate": 3.2905826405016385e-09, + "loss": 0.553, + "step": 22421 + }, + { + "epoch": 0.9927498533367278, + "grad_norm": 1.4150183200836182, + "learning_rate": 3.2510586227374465e-09, + "loss": 0.4512, + "step": 22422 + }, + { + "epoch": 0.9927941290415416, + "grad_norm": 1.3238999843597412, + "learning_rate": 3.2117733691805484e-09, + "loss": 0.4459, + "step": 22423 + }, + { + "epoch": 0.9928384047463555, + "grad_norm": 1.3706519603729248, + "learning_rate": 3.1727268807724145e-09, + "loss": 0.4549, + "step": 22424 + }, + { + "epoch": 0.9928826804511695, + "grad_norm": 1.5933243036270142, + "learning_rate": 3.13391915844341e-09, + "loss": 0.5193, + "step": 22425 + }, + { + "epoch": 0.9929269561559833, + "grad_norm": 1.5807924270629883, + "learning_rate": 3.095350203120573e-09, + "loss": 0.4666, + "step": 22426 + }, + { + "epoch": 0.9929712318607972, + "grad_norm": 1.4139354228973389, + "learning_rate": 3.0570200157264975e-09, + "loss": 0.4951, + "step": 22427 + }, + { + "epoch": 0.9930155075656111, + "grad_norm": 2.658493995666504, + "learning_rate": 3.018928597174897e-09, + "loss": 0.4851, + "step": 22428 + }, + { + "epoch": 0.993059783270425, + "grad_norm": 1.8776130676269531, + "learning_rate": 2.9810759483761553e-09, + "loss": 0.5033, + "step": 22429 + }, + { + "epoch": 0.9931040589752388, + "grad_norm": 1.5216768980026245, + "learning_rate": 2.9434620702351033e-09, + "loss": 0.5008, + "step": 22430 + }, + { + "epoch": 0.9931483346800527, + "grad_norm": 1.2916616201400757, + "learning_rate": 2.906086963648802e-09, + "loss": 0.4569, + "step": 22431 + }, + { + "epoch": 0.9931926103848666, + "grad_norm": 1.5212324857711792, + "learning_rate": 2.8689506295120906e-09, + "loss": 0.5107, + "step": 22432 + }, + { + "epoch": 0.9932368860896804, + "grad_norm": 2.0754058361053467, + "learning_rate": 2.832053068709817e-09, + "loss": 0.5678, + "step": 22433 + }, + { + "epoch": 0.9932811617944943, + "grad_norm": 1.5384774208068848, + "learning_rate": 2.795394282123498e-09, + "loss": 0.4706, + "step": 22434 + }, + { + "epoch": 0.9933254374993082, + "grad_norm": 1.3740078210830688, + "learning_rate": 2.7589742706291e-09, + "loss": 0.4392, + "step": 22435 + }, + { + "epoch": 0.993369713204122, + "grad_norm": 1.6627801656723022, + "learning_rate": 2.7227930350970376e-09, + "loss": 0.511, + "step": 22436 + }, + { + "epoch": 0.993413988908936, + "grad_norm": 1.7929662466049194, + "learning_rate": 2.6868505763910645e-09, + "loss": 0.5698, + "step": 22437 + }, + { + "epoch": 0.9934582646137499, + "grad_norm": 1.5173124074935913, + "learning_rate": 2.651146895369383e-09, + "loss": 0.5391, + "step": 22438 + }, + { + "epoch": 0.9935025403185637, + "grad_norm": 1.5407236814498901, + "learning_rate": 2.6156819928846445e-09, + "loss": 0.5151, + "step": 22439 + }, + { + "epoch": 0.9935468160233776, + "grad_norm": 1.525879144668579, + "learning_rate": 2.5804558697839487e-09, + "loss": 0.4547, + "step": 22440 + }, + { + "epoch": 0.9935910917281915, + "grad_norm": 2.040968894958496, + "learning_rate": 2.5454685269099553e-09, + "loss": 0.5099, + "step": 22441 + }, + { + "epoch": 0.9936353674330053, + "grad_norm": 1.894140362739563, + "learning_rate": 2.510719965096442e-09, + "loss": 0.5302, + "step": 22442 + }, + { + "epoch": 0.9936796431378192, + "grad_norm": 1.5559254884719849, + "learning_rate": 2.476210185173855e-09, + "loss": 0.4893, + "step": 22443 + }, + { + "epoch": 0.9937239188426331, + "grad_norm": 1.7323949337005615, + "learning_rate": 2.4419391879670906e-09, + "loss": 0.5201, + "step": 22444 + }, + { + "epoch": 0.9937681945474469, + "grad_norm": 1.7169309854507446, + "learning_rate": 2.4079069742943828e-09, + "loss": 0.5171, + "step": 22445 + }, + { + "epoch": 0.9938124702522608, + "grad_norm": 1.5784716606140137, + "learning_rate": 2.374113544968415e-09, + "loss": 0.4673, + "step": 22446 + }, + { + "epoch": 0.9938567459570747, + "grad_norm": 1.8109666109085083, + "learning_rate": 2.340558900796319e-09, + "loss": 0.5153, + "step": 22447 + }, + { + "epoch": 0.9939010216618885, + "grad_norm": 1.4935522079467773, + "learning_rate": 2.3072430425796768e-09, + "loss": 0.4311, + "step": 22448 + }, + { + "epoch": 0.9939452973667025, + "grad_norm": 1.336434245109558, + "learning_rate": 2.274165971114517e-09, + "loss": 0.4067, + "step": 22449 + }, + { + "epoch": 0.9939895730715164, + "grad_norm": 1.463593602180481, + "learning_rate": 2.241327687190209e-09, + "loss": 0.4474, + "step": 22450 + }, + { + "epoch": 0.9940338487763302, + "grad_norm": 1.506569266319275, + "learning_rate": 2.20872819159057e-09, + "loss": 0.5537, + "step": 22451 + }, + { + "epoch": 0.9940781244811441, + "grad_norm": 1.5639461278915405, + "learning_rate": 2.176367485096087e-09, + "loss": 0.4862, + "step": 22452 + }, + { + "epoch": 0.994122400185958, + "grad_norm": 1.3226609230041504, + "learning_rate": 2.1442455684772546e-09, + "loss": 0.4469, + "step": 22453 + }, + { + "epoch": 0.9941666758907718, + "grad_norm": 1.6880680322647095, + "learning_rate": 2.1123624425034572e-09, + "loss": 0.5095, + "step": 22454 + }, + { + "epoch": 0.9942109515955857, + "grad_norm": 1.8652067184448242, + "learning_rate": 2.080718107935198e-09, + "loss": 0.4758, + "step": 22455 + }, + { + "epoch": 0.9942552273003996, + "grad_norm": 1.7075281143188477, + "learning_rate": 2.0493125655285386e-09, + "loss": 0.4945, + "step": 22456 + }, + { + "epoch": 0.9942995030052134, + "grad_norm": 1.6930670738220215, + "learning_rate": 2.018145816032879e-09, + "loss": 0.5219, + "step": 22457 + }, + { + "epoch": 0.9943437787100273, + "grad_norm": 1.8046153783798218, + "learning_rate": 1.9872178601931802e-09, + "loss": 0.4501, + "step": 22458 + }, + { + "epoch": 0.9943880544148412, + "grad_norm": 1.6208077669143677, + "learning_rate": 1.9565286987488497e-09, + "loss": 0.4611, + "step": 22459 + }, + { + "epoch": 0.994432330119655, + "grad_norm": 1.4001660346984863, + "learning_rate": 1.9260783324315245e-09, + "loss": 0.4212, + "step": 22460 + }, + { + "epoch": 0.994476605824469, + "grad_norm": 1.354941964149475, + "learning_rate": 1.895866761968401e-09, + "loss": 0.4616, + "step": 22461 + }, + { + "epoch": 0.9945208815292829, + "grad_norm": 1.7903841733932495, + "learning_rate": 1.865893988083345e-09, + "loss": 0.4942, + "step": 22462 + }, + { + "epoch": 0.9945651572340968, + "grad_norm": 1.5321999788284302, + "learning_rate": 1.8361600114891187e-09, + "loss": 0.4671, + "step": 22463 + }, + { + "epoch": 0.9946094329389106, + "grad_norm": 1.3553184270858765, + "learning_rate": 1.806664832898486e-09, + "loss": 0.473, + "step": 22464 + }, + { + "epoch": 0.9946537086437245, + "grad_norm": 1.5530153512954712, + "learning_rate": 1.7774084530142177e-09, + "loss": 0.4443, + "step": 22465 + }, + { + "epoch": 0.9946979843485384, + "grad_norm": 1.7697391510009766, + "learning_rate": 1.7483908725357546e-09, + "loss": 0.4905, + "step": 22466 + }, + { + "epoch": 0.9947422600533522, + "grad_norm": 1.429468035697937, + "learning_rate": 1.7196120921558757e-09, + "loss": 0.488, + "step": 22467 + }, + { + "epoch": 0.9947865357581661, + "grad_norm": 1.604284644126892, + "learning_rate": 1.6910721125629192e-09, + "loss": 0.5237, + "step": 22468 + }, + { + "epoch": 0.99483081146298, + "grad_norm": 1.789229393005371, + "learning_rate": 1.662770934437452e-09, + "loss": 0.5257, + "step": 22469 + }, + { + "epoch": 0.9948750871677938, + "grad_norm": 1.3911654949188232, + "learning_rate": 1.6347085584555999e-09, + "loss": 0.4385, + "step": 22470 + }, + { + "epoch": 0.9949193628726077, + "grad_norm": 1.5014498233795166, + "learning_rate": 1.6068849852879376e-09, + "loss": 0.4569, + "step": 22471 + }, + { + "epoch": 0.9949636385774216, + "grad_norm": 1.4147202968597412, + "learning_rate": 1.5793002155983783e-09, + "loss": 0.4436, + "step": 22472 + }, + { + "epoch": 0.9950079142822355, + "grad_norm": 1.2994354963302612, + "learning_rate": 1.5519542500463946e-09, + "loss": 0.4657, + "step": 22473 + }, + { + "epoch": 0.9950521899870494, + "grad_norm": 1.5901715755462646, + "learning_rate": 1.524847089285908e-09, + "loss": 0.4091, + "step": 22474 + }, + { + "epoch": 0.9950964656918633, + "grad_norm": 1.5702602863311768, + "learning_rate": 1.497978733961958e-09, + "loss": 0.4888, + "step": 22475 + }, + { + "epoch": 0.9951407413966771, + "grad_norm": 1.718679428100586, + "learning_rate": 1.4713491847184735e-09, + "loss": 0.5033, + "step": 22476 + }, + { + "epoch": 0.995185017101491, + "grad_norm": 1.486236572265625, + "learning_rate": 1.4449584421905027e-09, + "loss": 0.4559, + "step": 22477 + }, + { + "epoch": 0.9952292928063049, + "grad_norm": 1.2666243314743042, + "learning_rate": 1.418806507008652e-09, + "loss": 0.4156, + "step": 22478 + }, + { + "epoch": 0.9952735685111187, + "grad_norm": 1.3485040664672852, + "learning_rate": 1.3928933797979772e-09, + "loss": 0.4403, + "step": 22479 + }, + { + "epoch": 0.9953178442159326, + "grad_norm": 1.3192771673202515, + "learning_rate": 1.3672190611757619e-09, + "loss": 0.4602, + "step": 22480 + }, + { + "epoch": 0.9953621199207465, + "grad_norm": 2.0894548892974854, + "learning_rate": 1.34178355175707e-09, + "loss": 0.5181, + "step": 22481 + }, + { + "epoch": 0.9954063956255603, + "grad_norm": 1.6316345930099487, + "learning_rate": 1.316586852148083e-09, + "loss": 0.4919, + "step": 22482 + }, + { + "epoch": 0.9954506713303742, + "grad_norm": 1.6494559049606323, + "learning_rate": 1.291628962951652e-09, + "loss": 0.5219, + "step": 22483 + }, + { + "epoch": 0.9954949470351881, + "grad_norm": 1.5235642194747925, + "learning_rate": 1.2669098847639672e-09, + "loss": 0.4852, + "step": 22484 + }, + { + "epoch": 0.995539222740002, + "grad_norm": 1.4775996208190918, + "learning_rate": 1.2424296181745565e-09, + "loss": 0.4507, + "step": 22485 + }, + { + "epoch": 0.9955834984448159, + "grad_norm": 2.0357208251953125, + "learning_rate": 1.2181881637673976e-09, + "loss": 0.4812, + "step": 22486 + }, + { + "epoch": 0.9956277741496298, + "grad_norm": 1.7627829313278198, + "learning_rate": 1.194185522123137e-09, + "loss": 0.4866, + "step": 22487 + }, + { + "epoch": 0.9956720498544436, + "grad_norm": 1.4612122774124146, + "learning_rate": 1.1704216938146495e-09, + "loss": 0.4808, + "step": 22488 + }, + { + "epoch": 0.9957163255592575, + "grad_norm": 1.4351855516433716, + "learning_rate": 1.1468966794092596e-09, + "loss": 0.4548, + "step": 22489 + }, + { + "epoch": 0.9957606012640714, + "grad_norm": 1.659059762954712, + "learning_rate": 1.1236104794687396e-09, + "loss": 0.5127, + "step": 22490 + }, + { + "epoch": 0.9958048769688852, + "grad_norm": 1.457275152206421, + "learning_rate": 1.100563094550422e-09, + "loss": 0.4417, + "step": 22491 + }, + { + "epoch": 0.9958491526736991, + "grad_norm": 1.4612793922424316, + "learning_rate": 1.0777545252027566e-09, + "loss": 0.4746, + "step": 22492 + }, + { + "epoch": 0.995893428378513, + "grad_norm": 1.5181502103805542, + "learning_rate": 1.055184771970863e-09, + "loss": 0.4584, + "step": 22493 + }, + { + "epoch": 0.9959377040833268, + "grad_norm": 1.4778923988342285, + "learning_rate": 1.0328538353954199e-09, + "loss": 0.538, + "step": 22494 + }, + { + "epoch": 0.9959819797881407, + "grad_norm": 1.2332266569137573, + "learning_rate": 1.010761716008224e-09, + "loss": 0.431, + "step": 22495 + }, + { + "epoch": 0.9960262554929546, + "grad_norm": 1.3934627771377563, + "learning_rate": 9.889084143377413e-10, + "loss": 0.4768, + "step": 22496 + }, + { + "epoch": 0.9960705311977684, + "grad_norm": 1.7493839263916016, + "learning_rate": 9.672939309057773e-10, + "loss": 0.5191, + "step": 22497 + }, + { + "epoch": 0.9961148069025824, + "grad_norm": 1.4655648469924927, + "learning_rate": 9.459182662285849e-10, + "loss": 0.4966, + "step": 22498 + }, + { + "epoch": 0.9961590826073963, + "grad_norm": 1.5982279777526855, + "learning_rate": 9.24781420816867e-10, + "loss": 0.4876, + "step": 22499 + }, + { + "epoch": 0.9962033583122102, + "grad_norm": 1.5917437076568604, + "learning_rate": 9.038833951746651e-10, + "loss": 0.5164, + "step": 22500 + }, + { + "epoch": 0.996247634017024, + "grad_norm": 1.625932216644287, + "learning_rate": 8.832241898015793e-10, + "loss": 0.5388, + "step": 22501 + }, + { + "epoch": 0.9962919097218379, + "grad_norm": 1.83941650390625, + "learning_rate": 8.628038051905486e-10, + "loss": 0.479, + "step": 22502 + }, + { + "epoch": 0.9963361854266518, + "grad_norm": 1.5896730422973633, + "learning_rate": 8.426222418311814e-10, + "loss": 0.4983, + "step": 22503 + }, + { + "epoch": 0.9963804611314656, + "grad_norm": 1.2787744998931885, + "learning_rate": 8.226795002030941e-10, + "loss": 0.4518, + "step": 22504 + }, + { + "epoch": 0.9964247368362795, + "grad_norm": 1.4883766174316406, + "learning_rate": 8.029755807847928e-10, + "loss": 0.4605, + "step": 22505 + }, + { + "epoch": 0.9964690125410934, + "grad_norm": 1.6001522541046143, + "learning_rate": 7.835104840447916e-10, + "loss": 0.4751, + "step": 22506 + }, + { + "epoch": 0.9965132882459072, + "grad_norm": 1.4750685691833496, + "learning_rate": 7.642842104493842e-10, + "loss": 0.4955, + "step": 22507 + }, + { + "epoch": 0.9965575639507211, + "grad_norm": 1.7060365676879883, + "learning_rate": 7.452967604570926e-10, + "loss": 0.5168, + "step": 22508 + }, + { + "epoch": 0.9966018396555351, + "grad_norm": 1.3963189125061035, + "learning_rate": 7.265481345219983e-10, + "loss": 0.4313, + "step": 22509 + }, + { + "epoch": 0.9966461153603489, + "grad_norm": 2.335580825805664, + "learning_rate": 7.080383330915208e-10, + "loss": 0.5236, + "step": 22510 + }, + { + "epoch": 0.9966903910651628, + "grad_norm": 1.7677794694900513, + "learning_rate": 6.897673566075291e-10, + "loss": 0.4837, + "step": 22511 + }, + { + "epoch": 0.9967346667699767, + "grad_norm": 1.394895315170288, + "learning_rate": 6.71735205507451e-10, + "loss": 0.4805, + "step": 22512 + }, + { + "epoch": 0.9967789424747905, + "grad_norm": 1.3968087434768677, + "learning_rate": 6.539418802198328e-10, + "loss": 0.4739, + "step": 22513 + }, + { + "epoch": 0.9968232181796044, + "grad_norm": 1.4852735996246338, + "learning_rate": 6.3638738117211e-10, + "loss": 0.4595, + "step": 22514 + }, + { + "epoch": 0.9968674938844183, + "grad_norm": 1.4461027383804321, + "learning_rate": 6.190717087828368e-10, + "loss": 0.5279, + "step": 22515 + }, + { + "epoch": 0.9969117695892321, + "grad_norm": 1.5614840984344482, + "learning_rate": 6.019948634650163e-10, + "loss": 0.5134, + "step": 22516 + }, + { + "epoch": 0.996956045294046, + "grad_norm": 1.8615158796310425, + "learning_rate": 5.851568456261003e-10, + "loss": 0.5109, + "step": 22517 + }, + { + "epoch": 0.9970003209988599, + "grad_norm": 1.357485055923462, + "learning_rate": 5.685576556690997e-10, + "loss": 0.4825, + "step": 22518 + }, + { + "epoch": 0.9970445967036737, + "grad_norm": 1.3070613145828247, + "learning_rate": 5.521972939903642e-10, + "loss": 0.4678, + "step": 22519 + }, + { + "epoch": 0.9970888724084876, + "grad_norm": 1.4688520431518555, + "learning_rate": 5.360757609806921e-10, + "loss": 0.4508, + "step": 22520 + }, + { + "epoch": 0.9971331481133016, + "grad_norm": 1.3612699508666992, + "learning_rate": 5.201930570242209e-10, + "loss": 0.4201, + "step": 22521 + }, + { + "epoch": 0.9971774238181154, + "grad_norm": 1.681792140007019, + "learning_rate": 5.045491825017567e-10, + "loss": 0.496, + "step": 22522 + }, + { + "epoch": 0.9972216995229293, + "grad_norm": 1.3990017175674438, + "learning_rate": 4.891441377863349e-10, + "loss": 0.4844, + "step": 22523 + }, + { + "epoch": 0.9972659752277432, + "grad_norm": 1.3252395391464233, + "learning_rate": 4.739779232454389e-10, + "loss": 0.4605, + "step": 22524 + }, + { + "epoch": 0.997310250932557, + "grad_norm": 1.7105168104171753, + "learning_rate": 4.5905053924211185e-10, + "loss": 0.5376, + "step": 22525 + }, + { + "epoch": 0.9973545266373709, + "grad_norm": 1.1627171039581299, + "learning_rate": 4.443619861327353e-10, + "loss": 0.3752, + "step": 22526 + }, + { + "epoch": 0.9973988023421848, + "grad_norm": 1.5385782718658447, + "learning_rate": 4.299122642670295e-10, + "loss": 0.4428, + "step": 22527 + }, + { + "epoch": 0.9974430780469986, + "grad_norm": 1.59122633934021, + "learning_rate": 4.15701373991384e-10, + "loss": 0.4863, + "step": 22528 + }, + { + "epoch": 0.9974873537518125, + "grad_norm": 1.5799678564071655, + "learning_rate": 4.017293156444169e-10, + "loss": 0.4789, + "step": 22529 + }, + { + "epoch": 0.9975316294566264, + "grad_norm": 1.309240460395813, + "learning_rate": 3.8799608956030523e-10, + "loss": 0.421, + "step": 22530 + }, + { + "epoch": 0.9975759051614402, + "grad_norm": 1.3954873085021973, + "learning_rate": 3.745016960665648e-10, + "loss": 0.4657, + "step": 22531 + }, + { + "epoch": 0.9976201808662541, + "grad_norm": 1.6071215867996216, + "learning_rate": 3.6124613548627065e-10, + "loss": 0.5212, + "step": 22532 + }, + { + "epoch": 0.9976644565710681, + "grad_norm": 1.6002615690231323, + "learning_rate": 3.4822940813583615e-10, + "loss": 0.4941, + "step": 22533 + }, + { + "epoch": 0.997708732275882, + "grad_norm": 1.7608360052108765, + "learning_rate": 3.354515143250137e-10, + "loss": 0.5075, + "step": 22534 + }, + { + "epoch": 0.9977530079806958, + "grad_norm": 1.4733917713165283, + "learning_rate": 3.2291245436022465e-10, + "loss": 0.4895, + "step": 22535 + }, + { + "epoch": 0.9977972836855097, + "grad_norm": 1.5513523817062378, + "learning_rate": 3.106122285412294e-10, + "loss": 0.4647, + "step": 22536 + }, + { + "epoch": 0.9978415593903236, + "grad_norm": 1.4107859134674072, + "learning_rate": 2.9855083716112677e-10, + "loss": 0.482, + "step": 22537 + }, + { + "epoch": 0.9978858350951374, + "grad_norm": 1.8862829208374023, + "learning_rate": 2.867282805074645e-10, + "loss": 0.5001, + "step": 22538 + }, + { + "epoch": 0.9979301107999513, + "grad_norm": 1.7273287773132324, + "learning_rate": 2.7514455886334945e-10, + "loss": 0.5212, + "step": 22539 + }, + { + "epoch": 0.9979743865047652, + "grad_norm": 1.6515343189239502, + "learning_rate": 2.6379967250633744e-10, + "loss": 0.5024, + "step": 22540 + }, + { + "epoch": 0.998018662209579, + "grad_norm": 1.4663150310516357, + "learning_rate": 2.526936217051024e-10, + "loss": 0.4488, + "step": 22541 + }, + { + "epoch": 0.9980629379143929, + "grad_norm": 1.6505041122436523, + "learning_rate": 2.41826406727208e-10, + "loss": 0.5285, + "step": 22542 + }, + { + "epoch": 0.9981072136192068, + "grad_norm": 1.50602126121521, + "learning_rate": 2.3119802783022615e-10, + "loss": 0.4255, + "step": 22543 + }, + { + "epoch": 0.9981514893240206, + "grad_norm": 1.3514719009399414, + "learning_rate": 2.2080848526950804e-10, + "loss": 0.4307, + "step": 22544 + }, + { + "epoch": 0.9981957650288346, + "grad_norm": 1.3222299814224243, + "learning_rate": 2.1065777929263343e-10, + "loss": 0.4661, + "step": 22545 + }, + { + "epoch": 0.9982400407336485, + "grad_norm": 2.083442211151123, + "learning_rate": 2.007459101416309e-10, + "loss": 0.5956, + "step": 22546 + }, + { + "epoch": 0.9982843164384623, + "grad_norm": 1.350748062133789, + "learning_rate": 1.9107287805408824e-10, + "loss": 0.4435, + "step": 22547 + }, + { + "epoch": 0.9983285921432762, + "grad_norm": 1.5147206783294678, + "learning_rate": 1.8163868326093183e-10, + "loss": 0.4708, + "step": 22548 + }, + { + "epoch": 0.9983728678480901, + "grad_norm": 1.4368629455566406, + "learning_rate": 1.7244332598642666e-10, + "loss": 0.4138, + "step": 22549 + }, + { + "epoch": 0.9984171435529039, + "grad_norm": 1.4154052734375, + "learning_rate": 1.6348680645150715e-10, + "loss": 0.52, + "step": 22550 + }, + { + "epoch": 0.9984614192577178, + "grad_norm": 1.4786127805709839, + "learning_rate": 1.5476912486933615e-10, + "loss": 0.5039, + "step": 22551 + }, + { + "epoch": 0.9985056949625317, + "grad_norm": 2.0392532348632812, + "learning_rate": 1.4629028144863554e-10, + "loss": 0.5435, + "step": 22552 + }, + { + "epoch": 0.9985499706673455, + "grad_norm": 1.5053097009658813, + "learning_rate": 1.3805027639146595e-10, + "loss": 0.4645, + "step": 22553 + }, + { + "epoch": 0.9985942463721594, + "grad_norm": 1.6583842039108276, + "learning_rate": 1.3004910989433682e-10, + "loss": 0.4573, + "step": 22554 + }, + { + "epoch": 0.9986385220769733, + "grad_norm": 1.9726245403289795, + "learning_rate": 1.2228678214931677e-10, + "loss": 0.544, + "step": 22555 + }, + { + "epoch": 0.9986827977817871, + "grad_norm": 1.5303608179092407, + "learning_rate": 1.147632933407028e-10, + "loss": 0.4853, + "step": 22556 + }, + { + "epoch": 0.9987270734866011, + "grad_norm": 1.6086236238479614, + "learning_rate": 1.0747864364946126e-10, + "loss": 0.4352, + "step": 22557 + }, + { + "epoch": 0.998771349191415, + "grad_norm": 1.5195908546447754, + "learning_rate": 1.0043283324878695e-10, + "loss": 0.4775, + "step": 22558 + }, + { + "epoch": 0.9988156248962288, + "grad_norm": 1.485665202140808, + "learning_rate": 9.362586230632353e-11, + "loss": 0.4432, + "step": 22559 + }, + { + "epoch": 0.9988599006010427, + "grad_norm": 1.3629924058914185, + "learning_rate": 8.705773098638404e-11, + "loss": 0.4845, + "step": 22560 + }, + { + "epoch": 0.9989041763058566, + "grad_norm": 1.960473656654358, + "learning_rate": 8.072843944439968e-11, + "loss": 0.5573, + "step": 22561 + }, + { + "epoch": 0.9989484520106704, + "grad_norm": 1.2467297315597534, + "learning_rate": 7.463798783136078e-11, + "loss": 0.4032, + "step": 22562 + }, + { + "epoch": 0.9989927277154843, + "grad_norm": 1.3361024856567383, + "learning_rate": 6.87863762938168e-11, + "loss": 0.4512, + "step": 22563 + }, + { + "epoch": 0.9990370034202982, + "grad_norm": 1.7482330799102783, + "learning_rate": 6.317360497165581e-11, + "loss": 0.4727, + "step": 22564 + }, + { + "epoch": 0.999081279125112, + "grad_norm": 1.5746970176696777, + "learning_rate": 5.77996739981046e-11, + "loss": 0.4989, + "step": 22565 + }, + { + "epoch": 0.9991255548299259, + "grad_norm": 1.4301378726959229, + "learning_rate": 5.266458350083881e-11, + "loss": 0.4719, + "step": 22566 + }, + { + "epoch": 0.9991698305347398, + "grad_norm": 2.491403579711914, + "learning_rate": 4.776833360420341e-11, + "loss": 0.4513, + "step": 22567 + }, + { + "epoch": 0.9992141062395536, + "grad_norm": 1.6787583827972412, + "learning_rate": 4.311092442477183e-11, + "loss": 0.5007, + "step": 22568 + }, + { + "epoch": 0.9992583819443676, + "grad_norm": 1.421103835105896, + "learning_rate": 3.869235607245614e-11, + "loss": 0.4126, + "step": 22569 + }, + { + "epoch": 0.9993026576491815, + "grad_norm": 1.4502644538879395, + "learning_rate": 3.451262865383775e-11, + "loss": 0.4686, + "step": 22570 + }, + { + "epoch": 0.9993469333539954, + "grad_norm": 1.4232550859451294, + "learning_rate": 3.0571742268836744e-11, + "loss": 0.4642, + "step": 22571 + }, + { + "epoch": 0.9993912090588092, + "grad_norm": 3.9309284687042236, + "learning_rate": 2.686969701182207e-11, + "loss": 0.5117, + "step": 22572 + }, + { + "epoch": 0.9994354847636231, + "grad_norm": 1.4532166719436646, + "learning_rate": 2.3406492969391126e-11, + "loss": 0.4545, + "step": 22573 + }, + { + "epoch": 0.999479760468437, + "grad_norm": 1.362531304359436, + "learning_rate": 2.0182130227031083e-11, + "loss": 0.4592, + "step": 22574 + }, + { + "epoch": 0.9995240361732508, + "grad_norm": 1.3208242654800415, + "learning_rate": 1.7196608860237108e-11, + "loss": 0.4664, + "step": 22575 + }, + { + "epoch": 0.9995683118780647, + "grad_norm": 1.459243893623352, + "learning_rate": 1.444992893895325e-11, + "loss": 0.4388, + "step": 22576 + }, + { + "epoch": 0.9996125875828786, + "grad_norm": 1.4619625806808472, + "learning_rate": 1.194209053201334e-11, + "loss": 0.4809, + "step": 22577 + }, + { + "epoch": 0.9996568632876924, + "grad_norm": 1.697147250175476, + "learning_rate": 9.67309369603875e-12, + "loss": 0.4593, + "step": 22578 + }, + { + "epoch": 0.9997011389925063, + "grad_norm": 1.5029115676879883, + "learning_rate": 7.642938487650853e-12, + "loss": 0.458, + "step": 22579 + }, + { + "epoch": 0.9997454146973203, + "grad_norm": 1.5740940570831299, + "learning_rate": 5.851624954589241e-12, + "loss": 0.4744, + "step": 22580 + }, + { + "epoch": 0.9997896904021341, + "grad_norm": 1.9505765438079834, + "learning_rate": 4.299153139042389e-12, + "loss": 0.5289, + "step": 22581 + }, + { + "epoch": 0.999833966106948, + "grad_norm": 1.6207033395767212, + "learning_rate": 2.9855230787578792e-12, + "loss": 0.4859, + "step": 22582 + }, + { + "epoch": 0.9998782418117619, + "grad_norm": 1.8101694583892822, + "learning_rate": 1.9107348048219563e-12, + "loss": 0.5709, + "step": 22583 + }, + { + "epoch": 0.9999225175165757, + "grad_norm": 1.5934778451919556, + "learning_rate": 1.0747883427697503e-12, + "loss": 0.4681, + "step": 22584 + }, + { + "epoch": 0.9999667932213896, + "grad_norm": 1.4102418422698975, + "learning_rate": 4.776837125852751e-13, + "loss": 0.4481, + "step": 22585 + }, + { + "epoch": 1.0, + "grad_norm": 1.4102418422698975, + "learning_rate": 1.194209287014303e-13, + "loss": 0.4955, + "step": 22586 + } + ], + "logging_steps": 1, + "max_steps": 22586, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 2.5707325153301496e+20, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +}