{ "best_global_step": 89000, "best_metric": 3.275252103805542, "best_model_checkpoint": "/scratch/cl5625/exceptions/models/high_2000_6910/checkpoint-70000", "epoch": 29.33268733850129, "eval_steps": 1000, "global_step": 109000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.013458225667527994, "grad_norm": 3.308773994445801, "learning_rate": 0.000294, "loss": 8.4679, "step": 50 }, { "epoch": 0.026916451335055987, "grad_norm": 1.6042652130126953, "learning_rate": 0.0005939999999999999, "loss": 6.7351, "step": 100 }, { "epoch": 0.04037467700258398, "grad_norm": 1.6990554332733154, "learning_rate": 0.0005998020735155513, "loss": 6.3185, "step": 150 }, { "epoch": 0.053832902670111975, "grad_norm": 0.9088912010192871, "learning_rate": 0.0005996001077150935, "loss": 6.0464, "step": 200 }, { "epoch": 0.06729112833763996, "grad_norm": 0.6608636379241943, "learning_rate": 0.0005993981419146358, "loss": 5.8696, "step": 250 }, { "epoch": 0.08074935400516796, "grad_norm": 1.4318311214447021, "learning_rate": 0.0005991961761141779, "loss": 5.7427, "step": 300 }, { "epoch": 0.09420757967269595, "grad_norm": 2.375964641571045, "learning_rate": 0.0005989942103137202, "loss": 5.6263, "step": 350 }, { "epoch": 0.10766580534022395, "grad_norm": 1.0631674528121948, "learning_rate": 0.0005987922445132624, "loss": 5.4973, "step": 400 }, { "epoch": 0.12112403100775193, "grad_norm": 1.3984321355819702, "learning_rate": 0.0005985902787128047, "loss": 5.3815, "step": 450 }, { "epoch": 0.13458225667527993, "grad_norm": 1.0844402313232422, "learning_rate": 0.0005983883129123468, "loss": 5.2436, "step": 500 }, { "epoch": 0.14804048234280792, "grad_norm": 1.4188182353973389, "learning_rate": 0.000598186347111889, "loss": 5.1461, "step": 550 }, { "epoch": 0.16149870801033592, "grad_norm": 1.540555715560913, "learning_rate": 0.0005979843813114312, "loss": 5.0629, "step": 600 }, { "epoch": 0.1749569336778639, "grad_norm": 1.31498122215271, "learning_rate": 0.0005977824155109734, "loss": 4.9972, "step": 650 }, { "epoch": 0.1884151593453919, "grad_norm": 1.060059666633606, "learning_rate": 0.0005975804497105157, "loss": 4.9156, "step": 700 }, { "epoch": 0.2018733850129199, "grad_norm": 0.7784032821655273, "learning_rate": 0.0005973784839100578, "loss": 4.8646, "step": 750 }, { "epoch": 0.2153316106804479, "grad_norm": 1.025396466255188, "learning_rate": 0.0005971765181096001, "loss": 4.8039, "step": 800 }, { "epoch": 0.2287898363479759, "grad_norm": 0.7960473895072937, "learning_rate": 0.0005969745523091422, "loss": 4.7349, "step": 850 }, { "epoch": 0.24224806201550386, "grad_norm": 0.9397687911987305, "learning_rate": 0.0005967725865086845, "loss": 4.6972, "step": 900 }, { "epoch": 0.2557062876830319, "grad_norm": 0.9388722777366638, "learning_rate": 0.0005965706207082267, "loss": 4.6541, "step": 950 }, { "epoch": 0.26916451335055985, "grad_norm": 0.8853117227554321, "learning_rate": 0.0005963686549077689, "loss": 4.6012, "step": 1000 }, { "epoch": 0.26916451335055985, "eval_accuracy": 0.2658181129953373, "eval_loss": 4.529858589172363, "eval_runtime": 52.8669, "eval_samples_per_second": 340.705, "eval_steps_per_second": 21.299, "step": 1000 }, { "epoch": 0.2826227390180879, "grad_norm": 1.1518616676330566, "learning_rate": 0.0005961666891073111, "loss": 4.5541, "step": 1050 }, { "epoch": 0.29608096468561584, "grad_norm": 0.9574114680290222, "learning_rate": 0.0005959647233068533, "loss": 4.4977, "step": 1100 }, { "epoch": 0.30953919035314387, "grad_norm": 1.084009051322937, "learning_rate": 0.0005957627575063955, "loss": 4.4619, "step": 1150 }, { "epoch": 0.32299741602067183, "grad_norm": 0.7612378001213074, "learning_rate": 0.0005955607917059378, "loss": 4.4404, "step": 1200 }, { "epoch": 0.3364556416881998, "grad_norm": 0.8939399719238281, "learning_rate": 0.00059535882590548, "loss": 4.3977, "step": 1250 }, { "epoch": 0.3499138673557278, "grad_norm": 0.7879042625427246, "learning_rate": 0.0005951568601050221, "loss": 4.3644, "step": 1300 }, { "epoch": 0.3633720930232558, "grad_norm": 0.9349780082702637, "learning_rate": 0.0005949548943045644, "loss": 4.3304, "step": 1350 }, { "epoch": 0.3768303186907838, "grad_norm": 0.9910395741462708, "learning_rate": 0.0005947529285041066, "loss": 4.3176, "step": 1400 }, { "epoch": 0.3902885443583118, "grad_norm": 0.915924608707428, "learning_rate": 0.0005945509627036488, "loss": 4.279, "step": 1450 }, { "epoch": 0.4037467700258398, "grad_norm": 0.8442095518112183, "learning_rate": 0.000594348996903191, "loss": 4.2577, "step": 1500 }, { "epoch": 0.4172049956933678, "grad_norm": 0.7409194707870483, "learning_rate": 0.0005941470311027332, "loss": 4.233, "step": 1550 }, { "epoch": 0.4306632213608958, "grad_norm": 0.6870989799499512, "learning_rate": 0.0005939450653022754, "loss": 4.2171, "step": 1600 }, { "epoch": 0.44412144702842377, "grad_norm": 1.060454249382019, "learning_rate": 0.0005937430995018177, "loss": 4.1993, "step": 1650 }, { "epoch": 0.4575796726959518, "grad_norm": 1.0149903297424316, "learning_rate": 0.0005935411337013598, "loss": 4.1869, "step": 1700 }, { "epoch": 0.47103789836347976, "grad_norm": 0.7430284023284912, "learning_rate": 0.0005933391679009021, "loss": 4.163, "step": 1750 }, { "epoch": 0.4844961240310077, "grad_norm": 0.5982431769371033, "learning_rate": 0.0005931372021004442, "loss": 4.1555, "step": 1800 }, { "epoch": 0.49795434969853575, "grad_norm": 0.708482563495636, "learning_rate": 0.0005929352362999865, "loss": 4.1336, "step": 1850 }, { "epoch": 0.5114125753660638, "grad_norm": 0.8102256655693054, "learning_rate": 0.0005927332704995287, "loss": 4.1256, "step": 1900 }, { "epoch": 0.5248708010335917, "grad_norm": 0.6787132024765015, "learning_rate": 0.0005925313046990709, "loss": 4.1054, "step": 1950 }, { "epoch": 0.5383290267011197, "grad_norm": 0.6520165801048279, "learning_rate": 0.0005923293388986131, "loss": 4.085, "step": 2000 }, { "epoch": 0.5383290267011197, "eval_accuracy": 0.31672122911753114, "eval_loss": 4.02903413772583, "eval_runtime": 53.0678, "eval_samples_per_second": 339.415, "eval_steps_per_second": 21.218, "step": 2000 }, { "epoch": 0.5517872523686477, "grad_norm": 0.6289111971855164, "learning_rate": 0.0005921273730981553, "loss": 4.0798, "step": 2050 }, { "epoch": 0.5652454780361758, "grad_norm": 0.6548891663551331, "learning_rate": 0.0005919254072976975, "loss": 4.056, "step": 2100 }, { "epoch": 0.5787037037037037, "grad_norm": 0.6513845920562744, "learning_rate": 0.0005917234414972398, "loss": 4.0544, "step": 2150 }, { "epoch": 0.5921619293712317, "grad_norm": 0.5987239480018616, "learning_rate": 0.000591521475696782, "loss": 4.0385, "step": 2200 }, { "epoch": 0.6056201550387597, "grad_norm": 0.5677440762519836, "learning_rate": 0.0005913195098963241, "loss": 4.0397, "step": 2250 }, { "epoch": 0.6190783807062877, "grad_norm": 0.5570490956306458, "learning_rate": 0.0005911175440958664, "loss": 4.0091, "step": 2300 }, { "epoch": 0.6325366063738157, "grad_norm": 0.5314527750015259, "learning_rate": 0.0005909155782954086, "loss": 4.0057, "step": 2350 }, { "epoch": 0.6459948320413437, "grad_norm": 0.7578822374343872, "learning_rate": 0.0005907136124949508, "loss": 4.0038, "step": 2400 }, { "epoch": 0.6594530577088716, "grad_norm": 0.8584759831428528, "learning_rate": 0.000590511646694493, "loss": 3.994, "step": 2450 }, { "epoch": 0.6729112833763996, "grad_norm": 0.5055056214332581, "learning_rate": 0.0005903096808940352, "loss": 3.9828, "step": 2500 }, { "epoch": 0.6863695090439277, "grad_norm": 0.7988194823265076, "learning_rate": 0.0005901077150935774, "loss": 3.9744, "step": 2550 }, { "epoch": 0.6998277347114557, "grad_norm": 0.5637562870979309, "learning_rate": 0.0005899057492931197, "loss": 3.9678, "step": 2600 }, { "epoch": 0.7132859603789836, "grad_norm": 0.5229179859161377, "learning_rate": 0.0005897037834926618, "loss": 3.9523, "step": 2650 }, { "epoch": 0.7267441860465116, "grad_norm": 0.588083803653717, "learning_rate": 0.0005895018176922041, "loss": 3.9488, "step": 2700 }, { "epoch": 0.7402024117140397, "grad_norm": 0.6086938977241516, "learning_rate": 0.0005892998518917462, "loss": 3.9542, "step": 2750 }, { "epoch": 0.7536606373815676, "grad_norm": 0.5917510986328125, "learning_rate": 0.0005890978860912885, "loss": 3.9318, "step": 2800 }, { "epoch": 0.7671188630490956, "grad_norm": 0.5497215390205383, "learning_rate": 0.0005888959202908307, "loss": 3.9295, "step": 2850 }, { "epoch": 0.7805770887166236, "grad_norm": 0.4543793499469757, "learning_rate": 0.0005886939544903729, "loss": 3.9134, "step": 2900 }, { "epoch": 0.7940353143841516, "grad_norm": 0.6304684281349182, "learning_rate": 0.0005884919886899151, "loss": 3.9112, "step": 2950 }, { "epoch": 0.8074935400516796, "grad_norm": 0.6081494688987732, "learning_rate": 0.0005882900228894573, "loss": 3.9068, "step": 3000 }, { "epoch": 0.8074935400516796, "eval_accuracy": 0.3333811379497817, "eval_loss": 3.844052791595459, "eval_runtime": 53.1523, "eval_samples_per_second": 338.875, "eval_steps_per_second": 21.184, "step": 3000 }, { "epoch": 0.8209517657192076, "grad_norm": 0.5118460655212402, "learning_rate": 0.0005880880570889996, "loss": 3.8916, "step": 3050 }, { "epoch": 0.8344099913867356, "grad_norm": 0.5820624232292175, "learning_rate": 0.0005878860912885418, "loss": 3.881, "step": 3100 }, { "epoch": 0.8478682170542635, "grad_norm": 0.5204915404319763, "learning_rate": 0.000587684125488084, "loss": 3.8647, "step": 3150 }, { "epoch": 0.8613264427217916, "grad_norm": 0.4955189824104309, "learning_rate": 0.0005874821596876261, "loss": 3.8708, "step": 3200 }, { "epoch": 0.8747846683893196, "grad_norm": 0.5090410709381104, "learning_rate": 0.0005872801938871684, "loss": 3.863, "step": 3250 }, { "epoch": 0.8882428940568475, "grad_norm": 0.48452430963516235, "learning_rate": 0.0005870782280867106, "loss": 3.8626, "step": 3300 }, { "epoch": 0.9017011197243755, "grad_norm": 0.49206408858299255, "learning_rate": 0.0005868762622862528, "loss": 3.8704, "step": 3350 }, { "epoch": 0.9151593453919036, "grad_norm": 0.5284396409988403, "learning_rate": 0.000586674296485795, "loss": 3.8523, "step": 3400 }, { "epoch": 0.9286175710594315, "grad_norm": 0.44494983553886414, "learning_rate": 0.0005864723306853373, "loss": 3.8304, "step": 3450 }, { "epoch": 0.9420757967269595, "grad_norm": 0.4655598998069763, "learning_rate": 0.0005862703648848794, "loss": 3.837, "step": 3500 }, { "epoch": 0.9555340223944875, "grad_norm": 0.4665937125682831, "learning_rate": 0.0005860683990844217, "loss": 3.8193, "step": 3550 }, { "epoch": 0.9689922480620154, "grad_norm": 0.4261581599712372, "learning_rate": 0.0005858664332839638, "loss": 3.8203, "step": 3600 }, { "epoch": 0.9824504737295435, "grad_norm": 0.5406307578086853, "learning_rate": 0.0005856644674835061, "loss": 3.8193, "step": 3650 }, { "epoch": 0.9959086993970715, "grad_norm": 0.47078415751457214, "learning_rate": 0.0005854625016830483, "loss": 3.8235, "step": 3700 }, { "epoch": 1.009151593453919, "grad_norm": 0.44889238476753235, "learning_rate": 0.0005852605358825905, "loss": 3.7595, "step": 3750 }, { "epoch": 1.022609819121447, "grad_norm": 0.4847916066646576, "learning_rate": 0.0005850585700821327, "loss": 3.7577, "step": 3800 }, { "epoch": 1.036068044788975, "grad_norm": 0.4172884523868561, "learning_rate": 0.0005848566042816749, "loss": 3.7459, "step": 3850 }, { "epoch": 1.049526270456503, "grad_norm": 0.42249923944473267, "learning_rate": 0.0005846546384812171, "loss": 3.7445, "step": 3900 }, { "epoch": 1.062984496124031, "grad_norm": 0.4443018436431885, "learning_rate": 0.0005844526726807594, "loss": 3.7393, "step": 3950 }, { "epoch": 1.076442721791559, "grad_norm": 0.49296411871910095, "learning_rate": 0.0005842507068803016, "loss": 3.7396, "step": 4000 }, { "epoch": 1.076442721791559, "eval_accuracy": 0.3439660578531468, "eval_loss": 3.736942768096924, "eval_runtime": 53.1277, "eval_samples_per_second": 339.032, "eval_steps_per_second": 21.194, "step": 4000 }, { "epoch": 1.089900947459087, "grad_norm": 0.4840051531791687, "learning_rate": 0.0005840487410798437, "loss": 3.7276, "step": 4050 }, { "epoch": 1.103359173126615, "grad_norm": 0.44972163438796997, "learning_rate": 0.000583846775279386, "loss": 3.7509, "step": 4100 }, { "epoch": 1.116817398794143, "grad_norm": 0.5025952458381653, "learning_rate": 0.0005836448094789282, "loss": 3.7371, "step": 4150 }, { "epoch": 1.130275624461671, "grad_norm": 0.49098411202430725, "learning_rate": 0.0005834428436784704, "loss": 3.7385, "step": 4200 }, { "epoch": 1.143733850129199, "grad_norm": 0.4680483639240265, "learning_rate": 0.0005832408778780126, "loss": 3.7186, "step": 4250 }, { "epoch": 1.157192075796727, "grad_norm": 0.43116289377212524, "learning_rate": 0.0005830389120775548, "loss": 3.7223, "step": 4300 }, { "epoch": 1.1706503014642549, "grad_norm": 0.46523234248161316, "learning_rate": 0.000582836946277097, "loss": 3.7236, "step": 4350 }, { "epoch": 1.1841085271317828, "grad_norm": 0.41721317172050476, "learning_rate": 0.0005826349804766393, "loss": 3.723, "step": 4400 }, { "epoch": 1.197566752799311, "grad_norm": 0.4401717782020569, "learning_rate": 0.0005824330146761814, "loss": 3.7283, "step": 4450 }, { "epoch": 1.211024978466839, "grad_norm": 0.43926772475242615, "learning_rate": 0.0005822310488757237, "loss": 3.7224, "step": 4500 }, { "epoch": 1.224483204134367, "grad_norm": 0.43538621068000793, "learning_rate": 0.0005820290830752658, "loss": 3.7151, "step": 4550 }, { "epoch": 1.237941429801895, "grad_norm": 0.3874562978744507, "learning_rate": 0.0005818271172748081, "loss": 3.7128, "step": 4600 }, { "epoch": 1.251399655469423, "grad_norm": 0.4349665343761444, "learning_rate": 0.0005816251514743503, "loss": 3.7068, "step": 4650 }, { "epoch": 1.2648578811369509, "grad_norm": 0.4198451638221741, "learning_rate": 0.0005814231856738925, "loss": 3.7007, "step": 4700 }, { "epoch": 1.2783161068044788, "grad_norm": 0.45340976119041443, "learning_rate": 0.0005812212198734347, "loss": 3.7071, "step": 4750 }, { "epoch": 1.291774332472007, "grad_norm": 0.41575804352760315, "learning_rate": 0.0005810192540729769, "loss": 3.7074, "step": 4800 }, { "epoch": 1.3052325581395348, "grad_norm": 0.5499830842018127, "learning_rate": 0.0005808172882725192, "loss": 3.7059, "step": 4850 }, { "epoch": 1.318690783807063, "grad_norm": 0.42270398139953613, "learning_rate": 0.0005806153224720614, "loss": 3.7028, "step": 4900 }, { "epoch": 1.332149009474591, "grad_norm": 0.41370531916618347, "learning_rate": 0.0005804133566716036, "loss": 3.6946, "step": 4950 }, { "epoch": 1.3456072351421189, "grad_norm": 0.4177495539188385, "learning_rate": 0.0005802113908711457, "loss": 3.6861, "step": 5000 }, { "epoch": 1.3456072351421189, "eval_accuracy": 0.35080939734458394, "eval_loss": 3.6666910648345947, "eval_runtime": 53.2598, "eval_samples_per_second": 338.191, "eval_steps_per_second": 21.142, "step": 5000 }, { "epoch": 1.3590654608096469, "grad_norm": 0.3675026595592499, "learning_rate": 0.000580009425070688, "loss": 3.6914, "step": 5050 }, { "epoch": 1.3725236864771748, "grad_norm": 0.41109681129455566, "learning_rate": 0.0005798074592702302, "loss": 3.6908, "step": 5100 }, { "epoch": 1.3859819121447028, "grad_norm": 0.4249870181083679, "learning_rate": 0.0005796054934697724, "loss": 3.6828, "step": 5150 }, { "epoch": 1.3994401378122308, "grad_norm": 0.4249998927116394, "learning_rate": 0.0005794035276693146, "loss": 3.6723, "step": 5200 }, { "epoch": 1.412898363479759, "grad_norm": 0.4438369572162628, "learning_rate": 0.0005792015618688568, "loss": 3.6659, "step": 5250 }, { "epoch": 1.4263565891472867, "grad_norm": 0.48220908641815186, "learning_rate": 0.000578999596068399, "loss": 3.6638, "step": 5300 }, { "epoch": 1.4398148148148149, "grad_norm": 0.4296906888484955, "learning_rate": 0.0005787976302679413, "loss": 3.6744, "step": 5350 }, { "epoch": 1.4532730404823428, "grad_norm": 0.4431769847869873, "learning_rate": 0.0005785956644674834, "loss": 3.6551, "step": 5400 }, { "epoch": 1.4667312661498708, "grad_norm": 0.41457149386405945, "learning_rate": 0.0005783936986670257, "loss": 3.6671, "step": 5450 }, { "epoch": 1.4801894918173988, "grad_norm": 0.39893487095832825, "learning_rate": 0.0005781917328665678, "loss": 3.6632, "step": 5500 }, { "epoch": 1.4936477174849268, "grad_norm": 0.3814569413661957, "learning_rate": 0.0005779897670661101, "loss": 3.6514, "step": 5550 }, { "epoch": 1.507105943152455, "grad_norm": 0.39305347204208374, "learning_rate": 0.0005777878012656523, "loss": 3.652, "step": 5600 }, { "epoch": 1.5205641688199827, "grad_norm": 0.4522084891796112, "learning_rate": 0.0005775858354651945, "loss": 3.6533, "step": 5650 }, { "epoch": 1.5340223944875109, "grad_norm": 0.3905489444732666, "learning_rate": 0.0005773838696647367, "loss": 3.6526, "step": 5700 }, { "epoch": 1.5474806201550386, "grad_norm": 0.3704451620578766, "learning_rate": 0.000577181903864279, "loss": 3.647, "step": 5750 }, { "epoch": 1.5609388458225668, "grad_norm": 0.38568341732025146, "learning_rate": 0.0005769799380638212, "loss": 3.6492, "step": 5800 }, { "epoch": 1.5743970714900948, "grad_norm": 0.45711877942085266, "learning_rate": 0.0005767779722633633, "loss": 3.6387, "step": 5850 }, { "epoch": 1.5878552971576227, "grad_norm": 0.41149041056632996, "learning_rate": 0.0005765760064629056, "loss": 3.6423, "step": 5900 }, { "epoch": 1.6013135228251507, "grad_norm": 0.39101529121398926, "learning_rate": 0.0005763740406624477, "loss": 3.6335, "step": 5950 }, { "epoch": 1.6147717484926787, "grad_norm": 0.4096600413322449, "learning_rate": 0.00057617207486199, "loss": 3.634, "step": 6000 }, { "epoch": 1.6147717484926787, "eval_accuracy": 0.35582301514146036, "eval_loss": 3.612297534942627, "eval_runtime": 53.361, "eval_samples_per_second": 337.55, "eval_steps_per_second": 21.102, "step": 6000 }, { "epoch": 1.6282299741602069, "grad_norm": 0.3830372989177704, "learning_rate": 0.0005759701090615322, "loss": 3.6409, "step": 6050 }, { "epoch": 1.6416881998277346, "grad_norm": 0.3905888497829437, "learning_rate": 0.0005757681432610744, "loss": 3.6282, "step": 6100 }, { "epoch": 1.6551464254952628, "grad_norm": 0.3588292598724365, "learning_rate": 0.0005755661774606166, "loss": 3.6315, "step": 6150 }, { "epoch": 1.6686046511627906, "grad_norm": 0.39833560585975647, "learning_rate": 0.0005753642116601588, "loss": 3.6292, "step": 6200 }, { "epoch": 1.6820628768303187, "grad_norm": 0.38218745589256287, "learning_rate": 0.000575162245859701, "loss": 3.6251, "step": 6250 }, { "epoch": 1.6955211024978467, "grad_norm": 0.35812291502952576, "learning_rate": 0.0005749602800592433, "loss": 3.635, "step": 6300 }, { "epoch": 1.7089793281653747, "grad_norm": 0.4691900312900543, "learning_rate": 0.0005747583142587854, "loss": 3.6139, "step": 6350 }, { "epoch": 1.7224375538329026, "grad_norm": 0.37986138463020325, "learning_rate": 0.0005745563484583277, "loss": 3.6301, "step": 6400 }, { "epoch": 1.7358957795004306, "grad_norm": 0.38206204771995544, "learning_rate": 0.0005743543826578698, "loss": 3.6196, "step": 6450 }, { "epoch": 1.7493540051679588, "grad_norm": 0.37841781973838806, "learning_rate": 0.0005741524168574121, "loss": 3.6116, "step": 6500 }, { "epoch": 1.7628122308354865, "grad_norm": 0.381045401096344, "learning_rate": 0.0005739504510569543, "loss": 3.611, "step": 6550 }, { "epoch": 1.7762704565030147, "grad_norm": 0.4393485188484192, "learning_rate": 0.0005737484852564966, "loss": 3.6084, "step": 6600 }, { "epoch": 1.7897286821705425, "grad_norm": 0.36671558022499084, "learning_rate": 0.0005735465194560387, "loss": 3.616, "step": 6650 }, { "epoch": 1.8031869078380707, "grad_norm": 0.3708931505680084, "learning_rate": 0.000573344553655581, "loss": 3.5929, "step": 6700 }, { "epoch": 1.8166451335055986, "grad_norm": 0.3850688338279724, "learning_rate": 0.0005731425878551232, "loss": 3.6034, "step": 6750 }, { "epoch": 1.8301033591731266, "grad_norm": 0.39316853880882263, "learning_rate": 0.0005729406220546653, "loss": 3.6017, "step": 6800 }, { "epoch": 1.8435615848406546, "grad_norm": 0.4011707007884979, "learning_rate": 0.0005727386562542076, "loss": 3.6086, "step": 6850 }, { "epoch": 1.8570198105081825, "grad_norm": 0.4198000133037567, "learning_rate": 0.0005725366904537497, "loss": 3.6026, "step": 6900 }, { "epoch": 1.8704780361757107, "grad_norm": 0.385288804769516, "learning_rate": 0.000572334724653292, "loss": 3.6097, "step": 6950 }, { "epoch": 1.8839362618432385, "grad_norm": 0.4373071789741516, "learning_rate": 0.0005721327588528342, "loss": 3.602, "step": 7000 }, { "epoch": 1.8839362618432385, "eval_accuracy": 0.36056349474344784, "eval_loss": 3.567089319229126, "eval_runtime": 53.2558, "eval_samples_per_second": 338.217, "eval_steps_per_second": 21.143, "step": 7000 }, { "epoch": 1.8973944875107667, "grad_norm": 0.4110179841518402, "learning_rate": 0.0005719307930523764, "loss": 3.5933, "step": 7050 }, { "epoch": 1.9108527131782944, "grad_norm": 0.3766134977340698, "learning_rate": 0.0005717288272519186, "loss": 3.6057, "step": 7100 }, { "epoch": 1.9243109388458226, "grad_norm": 0.4383994936943054, "learning_rate": 0.0005715268614514608, "loss": 3.6037, "step": 7150 }, { "epoch": 1.9377691645133506, "grad_norm": 0.38433584570884705, "learning_rate": 0.000571324895651003, "loss": 3.5909, "step": 7200 }, { "epoch": 1.9512273901808785, "grad_norm": 0.3365281820297241, "learning_rate": 0.0005711229298505453, "loss": 3.5885, "step": 7250 }, { "epoch": 1.9646856158484065, "grad_norm": 0.39374086260795593, "learning_rate": 0.0005709209640500874, "loss": 3.5944, "step": 7300 }, { "epoch": 1.9781438415159345, "grad_norm": 0.37078049778938293, "learning_rate": 0.0005707189982496297, "loss": 3.5874, "step": 7350 }, { "epoch": 1.9916020671834627, "grad_norm": 0.42410802841186523, "learning_rate": 0.0005705170324491719, "loss": 3.5836, "step": 7400 }, { "epoch": 2.00484496124031, "grad_norm": 0.3713061213493347, "learning_rate": 0.0005703150666487141, "loss": 3.5666, "step": 7450 }, { "epoch": 2.018303186907838, "grad_norm": 0.37857165932655334, "learning_rate": 0.0005701131008482563, "loss": 3.5089, "step": 7500 }, { "epoch": 2.0317614125753662, "grad_norm": 0.3648632764816284, "learning_rate": 0.0005699111350477986, "loss": 3.4958, "step": 7550 }, { "epoch": 2.045219638242894, "grad_norm": 0.35465335845947266, "learning_rate": 0.0005697091692473408, "loss": 3.5056, "step": 7600 }, { "epoch": 2.058677863910422, "grad_norm": 0.38645628094673157, "learning_rate": 0.000569507203446883, "loss": 3.5053, "step": 7650 }, { "epoch": 2.07213608957795, "grad_norm": 0.3809046149253845, "learning_rate": 0.0005693052376464252, "loss": 3.5084, "step": 7700 }, { "epoch": 2.085594315245478, "grad_norm": 0.41608577966690063, "learning_rate": 0.0005691032718459673, "loss": 3.5104, "step": 7750 }, { "epoch": 2.099052540913006, "grad_norm": 0.3553376793861389, "learning_rate": 0.0005689013060455096, "loss": 3.5099, "step": 7800 }, { "epoch": 2.112510766580534, "grad_norm": 0.43572768568992615, "learning_rate": 0.0005686993402450518, "loss": 3.5095, "step": 7850 }, { "epoch": 2.125968992248062, "grad_norm": 0.3603117763996124, "learning_rate": 0.000568497374444594, "loss": 3.511, "step": 7900 }, { "epoch": 2.13942721791559, "grad_norm": 0.3712218105792999, "learning_rate": 0.0005682954086441362, "loss": 3.5121, "step": 7950 }, { "epoch": 2.152885443583118, "grad_norm": 0.3927130103111267, "learning_rate": 0.0005680934428436784, "loss": 3.5053, "step": 8000 }, { "epoch": 2.152885443583118, "eval_accuracy": 0.3639965180855729, "eval_loss": 3.539733648300171, "eval_runtime": 53.0444, "eval_samples_per_second": 339.565, "eval_steps_per_second": 21.228, "step": 8000 }, { "epoch": 2.166343669250646, "grad_norm": 0.39150112867355347, "learning_rate": 0.0005678914770432206, "loss": 3.502, "step": 8050 }, { "epoch": 2.179801894918174, "grad_norm": 0.4043424129486084, "learning_rate": 0.0005676895112427629, "loss": 3.5005, "step": 8100 }, { "epoch": 2.193260120585702, "grad_norm": 0.38725680112838745, "learning_rate": 0.000567487545442305, "loss": 3.5122, "step": 8150 }, { "epoch": 2.20671834625323, "grad_norm": 0.36324837803840637, "learning_rate": 0.0005672855796418473, "loss": 3.504, "step": 8200 }, { "epoch": 2.220176571920758, "grad_norm": 0.35534292459487915, "learning_rate": 0.0005670836138413894, "loss": 3.508, "step": 8250 }, { "epoch": 2.233634797588286, "grad_norm": 0.4022197425365448, "learning_rate": 0.0005668816480409317, "loss": 3.5084, "step": 8300 }, { "epoch": 2.2470930232558137, "grad_norm": 0.35888510942459106, "learning_rate": 0.0005666796822404739, "loss": 3.5138, "step": 8350 }, { "epoch": 2.260551248923342, "grad_norm": 0.3790094554424286, "learning_rate": 0.0005664777164400162, "loss": 3.5104, "step": 8400 }, { "epoch": 2.27400947459087, "grad_norm": 0.37779271602630615, "learning_rate": 0.0005662757506395583, "loss": 3.5032, "step": 8450 }, { "epoch": 2.287467700258398, "grad_norm": 0.3711851239204407, "learning_rate": 0.0005660737848391006, "loss": 3.5084, "step": 8500 }, { "epoch": 2.300925925925926, "grad_norm": 0.35237640142440796, "learning_rate": 0.0005658718190386428, "loss": 3.5112, "step": 8550 }, { "epoch": 2.314384151593454, "grad_norm": 0.401823490858078, "learning_rate": 0.000565669853238185, "loss": 3.5101, "step": 8600 }, { "epoch": 2.327842377260982, "grad_norm": 0.39976271986961365, "learning_rate": 0.0005654678874377272, "loss": 3.5018, "step": 8650 }, { "epoch": 2.3413006029285097, "grad_norm": 0.37438446283340454, "learning_rate": 0.0005652659216372693, "loss": 3.4992, "step": 8700 }, { "epoch": 2.354758828596038, "grad_norm": 0.37986740469932556, "learning_rate": 0.0005650639558368116, "loss": 3.4967, "step": 8750 }, { "epoch": 2.3682170542635657, "grad_norm": 0.3729777932167053, "learning_rate": 0.0005648619900363538, "loss": 3.4907, "step": 8800 }, { "epoch": 2.381675279931094, "grad_norm": 0.4080381989479065, "learning_rate": 0.000564660024235896, "loss": 3.5052, "step": 8850 }, { "epoch": 2.395133505598622, "grad_norm": 0.3736269176006317, "learning_rate": 0.0005644580584354382, "loss": 3.4969, "step": 8900 }, { "epoch": 2.40859173126615, "grad_norm": 0.39896562695503235, "learning_rate": 0.0005642560926349804, "loss": 3.4992, "step": 8950 }, { "epoch": 2.422049956933678, "grad_norm": 0.3957862854003906, "learning_rate": 0.0005640541268345226, "loss": 3.4997, "step": 9000 }, { "epoch": 2.422049956933678, "eval_accuracy": 0.3664585644795185, "eval_loss": 3.513334035873413, "eval_runtime": 53.2081, "eval_samples_per_second": 338.52, "eval_steps_per_second": 21.162, "step": 9000 }, { "epoch": 2.4355081826012057, "grad_norm": 0.4161396920681, "learning_rate": 0.0005638521610340649, "loss": 3.4993, "step": 9050 }, { "epoch": 2.448966408268734, "grad_norm": 0.3905985951423645, "learning_rate": 0.000563650195233607, "loss": 3.4938, "step": 9100 }, { "epoch": 2.4624246339362617, "grad_norm": 0.3634888529777527, "learning_rate": 0.0005634482294331493, "loss": 3.4957, "step": 9150 }, { "epoch": 2.47588285960379, "grad_norm": 0.34948739409446716, "learning_rate": 0.0005632462636326914, "loss": 3.5059, "step": 9200 }, { "epoch": 2.4893410852713176, "grad_norm": 0.3566109538078308, "learning_rate": 0.0005630442978322337, "loss": 3.5028, "step": 9250 }, { "epoch": 2.502799310938846, "grad_norm": 0.3608540892601013, "learning_rate": 0.0005628423320317759, "loss": 3.4995, "step": 9300 }, { "epoch": 2.516257536606374, "grad_norm": 0.3242948055267334, "learning_rate": 0.0005626403662313182, "loss": 3.4847, "step": 9350 }, { "epoch": 2.5297157622739017, "grad_norm": 0.40635618567466736, "learning_rate": 0.0005624384004308603, "loss": 3.4935, "step": 9400 }, { "epoch": 2.54317398794143, "grad_norm": 0.34435972571372986, "learning_rate": 0.0005622364346304026, "loss": 3.4989, "step": 9450 }, { "epoch": 2.5566322136089576, "grad_norm": 0.44505786895751953, "learning_rate": 0.0005620344688299448, "loss": 3.4959, "step": 9500 }, { "epoch": 2.570090439276486, "grad_norm": 0.36672013998031616, "learning_rate": 0.0005618325030294869, "loss": 3.4928, "step": 9550 }, { "epoch": 2.583548664944014, "grad_norm": 0.37671077251434326, "learning_rate": 0.0005616305372290292, "loss": 3.4988, "step": 9600 }, { "epoch": 2.5970068906115418, "grad_norm": 0.37526118755340576, "learning_rate": 0.0005614285714285713, "loss": 3.4852, "step": 9650 }, { "epoch": 2.6104651162790695, "grad_norm": 0.4347611963748932, "learning_rate": 0.0005612266056281136, "loss": 3.4959, "step": 9700 }, { "epoch": 2.6239233419465977, "grad_norm": 0.34966251254081726, "learning_rate": 0.0005610246398276558, "loss": 3.4847, "step": 9750 }, { "epoch": 2.637381567614126, "grad_norm": 0.3655852675437927, "learning_rate": 0.000560822674027198, "loss": 3.4847, "step": 9800 }, { "epoch": 2.6508397932816536, "grad_norm": 0.3635706305503845, "learning_rate": 0.0005606207082267402, "loss": 3.4832, "step": 9850 }, { "epoch": 2.664298018949182, "grad_norm": 0.38875824213027954, "learning_rate": 0.0005604187424262824, "loss": 3.4892, "step": 9900 }, { "epoch": 2.6777562446167096, "grad_norm": 0.35097113251686096, "learning_rate": 0.0005602167766258246, "loss": 3.4833, "step": 9950 }, { "epoch": 2.6912144702842378, "grad_norm": 0.38976341485977173, "learning_rate": 0.0005600148108253669, "loss": 3.4848, "step": 10000 }, { "epoch": 2.6912144702842378, "eval_accuracy": 0.36907423752723234, "eval_loss": 3.487663507461548, "eval_runtime": 53.3765, "eval_samples_per_second": 337.452, "eval_steps_per_second": 21.095, "step": 10000 }, { "epoch": 2.704672695951766, "grad_norm": 0.37565913796424866, "learning_rate": 0.000559812845024909, "loss": 3.4913, "step": 10050 }, { "epoch": 2.7181309216192937, "grad_norm": 0.37652722001075745, "learning_rate": 0.0005596108792244513, "loss": 3.4784, "step": 10100 }, { "epoch": 2.7315891472868215, "grad_norm": 0.33579641580581665, "learning_rate": 0.0005594089134239935, "loss": 3.4749, "step": 10150 }, { "epoch": 2.7450473729543496, "grad_norm": 0.34873345494270325, "learning_rate": 0.0005592069476235358, "loss": 3.4802, "step": 10200 }, { "epoch": 2.758505598621878, "grad_norm": 0.362138956785202, "learning_rate": 0.0005590049818230779, "loss": 3.4774, "step": 10250 }, { "epoch": 2.7719638242894056, "grad_norm": 0.37752765417099, "learning_rate": 0.0005588030160226202, "loss": 3.4785, "step": 10300 }, { "epoch": 2.7854220499569338, "grad_norm": 0.3606274724006653, "learning_rate": 0.0005586010502221623, "loss": 3.4773, "step": 10350 }, { "epoch": 2.7988802756244615, "grad_norm": 0.3434750437736511, "learning_rate": 0.0005583990844217045, "loss": 3.4756, "step": 10400 }, { "epoch": 2.8123385012919897, "grad_norm": 0.33499059081077576, "learning_rate": 0.0005581971186212468, "loss": 3.486, "step": 10450 }, { "epoch": 2.825796726959518, "grad_norm": 0.37345173954963684, "learning_rate": 0.0005579951528207889, "loss": 3.4833, "step": 10500 }, { "epoch": 2.8392549526270456, "grad_norm": 0.3695079982280731, "learning_rate": 0.0005577931870203312, "loss": 3.4728, "step": 10550 }, { "epoch": 2.8527131782945734, "grad_norm": 0.368105947971344, "learning_rate": 0.0005575912212198733, "loss": 3.4682, "step": 10600 }, { "epoch": 2.8661714039621016, "grad_norm": 0.3338841497898102, "learning_rate": 0.0005573892554194156, "loss": 3.4753, "step": 10650 }, { "epoch": 2.8796296296296298, "grad_norm": 0.36992380023002625, "learning_rate": 0.0005571872896189578, "loss": 3.4641, "step": 10700 }, { "epoch": 2.8930878552971575, "grad_norm": 0.3489851951599121, "learning_rate": 0.0005569853238185, "loss": 3.4734, "step": 10750 }, { "epoch": 2.9065460809646857, "grad_norm": 0.35536548495292664, "learning_rate": 0.0005567833580180422, "loss": 3.4694, "step": 10800 }, { "epoch": 2.9200043066322134, "grad_norm": 0.3439956307411194, "learning_rate": 0.0005565813922175844, "loss": 3.4861, "step": 10850 }, { "epoch": 2.9334625322997416, "grad_norm": 0.32711493968963623, "learning_rate": 0.0005563794264171266, "loss": 3.4761, "step": 10900 }, { "epoch": 2.94692075796727, "grad_norm": 0.36705949902534485, "learning_rate": 0.0005561774606166689, "loss": 3.4692, "step": 10950 }, { "epoch": 2.9603789836347976, "grad_norm": 0.3396705389022827, "learning_rate": 0.000555975494816211, "loss": 3.4716, "step": 11000 }, { "epoch": 2.9603789836347976, "eval_accuracy": 0.37102390535033614, "eval_loss": 3.4675471782684326, "eval_runtime": 53.1554, "eval_samples_per_second": 338.856, "eval_steps_per_second": 21.183, "step": 11000 }, { "epoch": 2.9738372093023253, "grad_norm": 0.4038015902042389, "learning_rate": 0.0005557735290157534, "loss": 3.4765, "step": 11050 }, { "epoch": 2.9872954349698535, "grad_norm": 0.3803868889808655, "learning_rate": 0.0005555715632152955, "loss": 3.4671, "step": 11100 }, { "epoch": 3.000538329026701, "grad_norm": 0.4042735695838928, "learning_rate": 0.0005553695974148378, "loss": 3.467, "step": 11150 }, { "epoch": 3.013996554694229, "grad_norm": 0.35812151432037354, "learning_rate": 0.0005551676316143799, "loss": 3.3741, "step": 11200 }, { "epoch": 3.027454780361757, "grad_norm": 0.36608004570007324, "learning_rate": 0.0005549656658139222, "loss": 3.3646, "step": 11250 }, { "epoch": 3.0409130060292853, "grad_norm": 0.3438897430896759, "learning_rate": 0.0005547637000134644, "loss": 3.3726, "step": 11300 }, { "epoch": 3.054371231696813, "grad_norm": 0.36485397815704346, "learning_rate": 0.0005545617342130065, "loss": 3.3717, "step": 11350 }, { "epoch": 3.067829457364341, "grad_norm": 0.371673583984375, "learning_rate": 0.0005543597684125488, "loss": 3.3852, "step": 11400 }, { "epoch": 3.081287683031869, "grad_norm": 0.37548917531967163, "learning_rate": 0.0005541578026120909, "loss": 3.3954, "step": 11450 }, { "epoch": 3.094745908699397, "grad_norm": 0.3658439815044403, "learning_rate": 0.0005539558368116332, "loss": 3.3921, "step": 11500 }, { "epoch": 3.108204134366925, "grad_norm": 0.3572917878627777, "learning_rate": 0.0005537538710111754, "loss": 3.3885, "step": 11550 }, { "epoch": 3.121662360034453, "grad_norm": 0.3559921383857727, "learning_rate": 0.0005535519052107176, "loss": 3.3961, "step": 11600 }, { "epoch": 3.135120585701981, "grad_norm": 0.34737083315849304, "learning_rate": 0.0005533499394102598, "loss": 3.3874, "step": 11650 }, { "epoch": 3.148578811369509, "grad_norm": 0.3417198061943054, "learning_rate": 0.000553147973609802, "loss": 3.4002, "step": 11700 }, { "epoch": 3.162037037037037, "grad_norm": 0.34037500619888306, "learning_rate": 0.0005529460078093442, "loss": 3.3855, "step": 11750 }, { "epoch": 3.175495262704565, "grad_norm": 0.36747488379478455, "learning_rate": 0.0005527440420088865, "loss": 3.3956, "step": 11800 }, { "epoch": 3.188953488372093, "grad_norm": 0.3483152389526367, "learning_rate": 0.0005525420762084286, "loss": 3.3962, "step": 11850 }, { "epoch": 3.202411714039621, "grad_norm": 0.3729177415370941, "learning_rate": 0.0005523401104079709, "loss": 3.4056, "step": 11900 }, { "epoch": 3.215869939707149, "grad_norm": 0.34760040044784546, "learning_rate": 0.0005521381446075131, "loss": 3.4, "step": 11950 }, { "epoch": 3.229328165374677, "grad_norm": 0.37514349818229675, "learning_rate": 0.0005519361788070554, "loss": 3.4002, "step": 12000 }, { "epoch": 3.229328165374677, "eval_accuracy": 0.372848086055263, "eval_loss": 3.457045078277588, "eval_runtime": 53.5082, "eval_samples_per_second": 336.622, "eval_steps_per_second": 21.044, "step": 12000 }, { "epoch": 3.242786391042205, "grad_norm": 0.34385743737220764, "learning_rate": 0.0005517342130065975, "loss": 3.3903, "step": 12050 }, { "epoch": 3.2562446167097328, "grad_norm": 0.32674697041511536, "learning_rate": 0.0005515322472061398, "loss": 3.4041, "step": 12100 }, { "epoch": 3.269702842377261, "grad_norm": 0.34761109948158264, "learning_rate": 0.0005513302814056819, "loss": 3.3983, "step": 12150 }, { "epoch": 3.283161068044789, "grad_norm": 0.36604440212249756, "learning_rate": 0.0005511283156052242, "loss": 3.4032, "step": 12200 }, { "epoch": 3.296619293712317, "grad_norm": 0.35515835881233215, "learning_rate": 0.0005509263498047664, "loss": 3.4022, "step": 12250 }, { "epoch": 3.310077519379845, "grad_norm": 0.3531981110572815, "learning_rate": 0.0005507243840043085, "loss": 3.397, "step": 12300 }, { "epoch": 3.323535745047373, "grad_norm": 0.337921679019928, "learning_rate": 0.0005505224182038508, "loss": 3.4047, "step": 12350 }, { "epoch": 3.336993970714901, "grad_norm": 0.3581739664077759, "learning_rate": 0.0005503204524033929, "loss": 3.4017, "step": 12400 }, { "epoch": 3.3504521963824287, "grad_norm": 0.36727410554885864, "learning_rate": 0.0005501184866029352, "loss": 3.3959, "step": 12450 }, { "epoch": 3.363910422049957, "grad_norm": 0.35325637459754944, "learning_rate": 0.0005499165208024774, "loss": 3.3999, "step": 12500 }, { "epoch": 3.3773686477174847, "grad_norm": 0.3525833785533905, "learning_rate": 0.0005497145550020196, "loss": 3.4057, "step": 12550 }, { "epoch": 3.390826873385013, "grad_norm": 0.3457193970680237, "learning_rate": 0.0005495125892015618, "loss": 3.4155, "step": 12600 }, { "epoch": 3.404285099052541, "grad_norm": 0.37200361490249634, "learning_rate": 0.000549310623401104, "loss": 3.4005, "step": 12650 }, { "epoch": 3.417743324720069, "grad_norm": 0.34540146589279175, "learning_rate": 0.0005491086576006462, "loss": 3.4055, "step": 12700 }, { "epoch": 3.431201550387597, "grad_norm": 0.35365238785743713, "learning_rate": 0.0005489066918001885, "loss": 3.3988, "step": 12750 }, { "epoch": 3.4446597760551247, "grad_norm": 0.35095787048339844, "learning_rate": 0.0005487047259997306, "loss": 3.3998, "step": 12800 }, { "epoch": 3.458118001722653, "grad_norm": 0.34939098358154297, "learning_rate": 0.0005485027601992729, "loss": 3.3908, "step": 12850 }, { "epoch": 3.471576227390181, "grad_norm": 0.3667967617511749, "learning_rate": 0.0005483007943988151, "loss": 3.396, "step": 12900 }, { "epoch": 3.485034453057709, "grad_norm": 0.357086181640625, "learning_rate": 0.0005480988285983574, "loss": 3.3958, "step": 12950 }, { "epoch": 3.4984926787252366, "grad_norm": 0.37680062651634216, "learning_rate": 0.0005478968627978995, "loss": 3.401, "step": 13000 }, { "epoch": 3.4984926787252366, "eval_accuracy": 0.3740813365127749, "eval_loss": 3.4411723613739014, "eval_runtime": 53.1367, "eval_samples_per_second": 338.975, "eval_steps_per_second": 21.191, "step": 13000 }, { "epoch": 3.511950904392765, "grad_norm": 0.3376721441745758, "learning_rate": 0.0005476948969974418, "loss": 3.4104, "step": 13050 }, { "epoch": 3.525409130060293, "grad_norm": 0.38172057271003723, "learning_rate": 0.0005474929311969839, "loss": 3.4016, "step": 13100 }, { "epoch": 3.5388673557278207, "grad_norm": 0.38476553559303284, "learning_rate": 0.0005472909653965261, "loss": 3.3922, "step": 13150 }, { "epoch": 3.552325581395349, "grad_norm": 0.3454050123691559, "learning_rate": 0.0005470889995960684, "loss": 3.3904, "step": 13200 }, { "epoch": 3.5657838070628767, "grad_norm": 0.33698442578315735, "learning_rate": 0.0005468870337956105, "loss": 3.4087, "step": 13250 }, { "epoch": 3.579242032730405, "grad_norm": 0.32757994532585144, "learning_rate": 0.0005466850679951528, "loss": 3.3995, "step": 13300 }, { "epoch": 3.592700258397933, "grad_norm": 0.3386076092720032, "learning_rate": 0.0005464831021946949, "loss": 3.3935, "step": 13350 }, { "epoch": 3.606158484065461, "grad_norm": 0.33555930852890015, "learning_rate": 0.0005462811363942372, "loss": 3.3903, "step": 13400 }, { "epoch": 3.6196167097329885, "grad_norm": 0.333196759223938, "learning_rate": 0.0005460791705937794, "loss": 3.4044, "step": 13450 }, { "epoch": 3.6330749354005167, "grad_norm": 0.3701780438423157, "learning_rate": 0.0005458772047933216, "loss": 3.3974, "step": 13500 }, { "epoch": 3.646533161068045, "grad_norm": 0.3696497976779938, "learning_rate": 0.0005456752389928638, "loss": 3.4092, "step": 13550 }, { "epoch": 3.6599913867355727, "grad_norm": 0.3531237840652466, "learning_rate": 0.000545473273192406, "loss": 3.3965, "step": 13600 }, { "epoch": 3.673449612403101, "grad_norm": 0.3437979817390442, "learning_rate": 0.0005452713073919482, "loss": 3.4056, "step": 13650 }, { "epoch": 3.6869078380706286, "grad_norm": 0.3295712471008301, "learning_rate": 0.0005450693415914905, "loss": 3.4118, "step": 13700 }, { "epoch": 3.700366063738157, "grad_norm": 0.39618441462516785, "learning_rate": 0.0005448673757910327, "loss": 3.4115, "step": 13750 }, { "epoch": 3.713824289405685, "grad_norm": 0.37704867124557495, "learning_rate": 0.0005446654099905749, "loss": 3.3902, "step": 13800 }, { "epoch": 3.7272825150732127, "grad_norm": 0.3658808171749115, "learning_rate": 0.0005444634441901171, "loss": 3.4064, "step": 13850 }, { "epoch": 3.7407407407407405, "grad_norm": 0.401401549577713, "learning_rate": 0.0005442614783896594, "loss": 3.3959, "step": 13900 }, { "epoch": 3.7541989664082687, "grad_norm": 0.33702999353408813, "learning_rate": 0.0005440595125892015, "loss": 3.3956, "step": 13950 }, { "epoch": 3.767657192075797, "grad_norm": 0.36874791979789734, "learning_rate": 0.0005438575467887438, "loss": 3.3903, "step": 14000 }, { "epoch": 3.767657192075797, "eval_accuracy": 0.37602872275191185, "eval_loss": 3.4243662357330322, "eval_runtime": 53.3346, "eval_samples_per_second": 337.717, "eval_steps_per_second": 21.112, "step": 14000 }, { "epoch": 3.7811154177433246, "grad_norm": 0.3917904198169708, "learning_rate": 0.0005436555809882859, "loss": 3.4051, "step": 14050 }, { "epoch": 3.794573643410853, "grad_norm": 0.367305189371109, "learning_rate": 0.0005434536151878281, "loss": 3.4014, "step": 14100 }, { "epoch": 3.8080318690783805, "grad_norm": 0.3374054431915283, "learning_rate": 0.0005432516493873704, "loss": 3.3996, "step": 14150 }, { "epoch": 3.8214900947459087, "grad_norm": 0.3432117700576782, "learning_rate": 0.0005430496835869125, "loss": 3.3921, "step": 14200 }, { "epoch": 3.834948320413437, "grad_norm": 0.36590588092803955, "learning_rate": 0.0005428477177864548, "loss": 3.4038, "step": 14250 }, { "epoch": 3.8484065460809647, "grad_norm": 0.4062247574329376, "learning_rate": 0.0005426457519859969, "loss": 3.4078, "step": 14300 }, { "epoch": 3.8618647717484924, "grad_norm": 0.31158193945884705, "learning_rate": 0.0005424437861855392, "loss": 3.3981, "step": 14350 }, { "epoch": 3.8753229974160206, "grad_norm": 0.3665839433670044, "learning_rate": 0.0005422418203850814, "loss": 3.3995, "step": 14400 }, { "epoch": 3.888781223083549, "grad_norm": 0.34909674525260925, "learning_rate": 0.0005420398545846236, "loss": 3.3967, "step": 14450 }, { "epoch": 3.9022394487510765, "grad_norm": 0.34972283244132996, "learning_rate": 0.0005418378887841658, "loss": 3.4018, "step": 14500 }, { "epoch": 3.9156976744186047, "grad_norm": 0.36741903424263, "learning_rate": 0.000541635922983708, "loss": 3.3922, "step": 14550 }, { "epoch": 3.9291559000861325, "grad_norm": 0.34265077114105225, "learning_rate": 0.0005414339571832503, "loss": 3.3984, "step": 14600 }, { "epoch": 3.9426141257536607, "grad_norm": 0.35221385955810547, "learning_rate": 0.0005412319913827925, "loss": 3.3915, "step": 14650 }, { "epoch": 3.956072351421189, "grad_norm": 0.3693876266479492, "learning_rate": 0.0005410300255823347, "loss": 3.4048, "step": 14700 }, { "epoch": 3.9695305770887166, "grad_norm": 0.36447668075561523, "learning_rate": 0.0005408280597818769, "loss": 3.4002, "step": 14750 }, { "epoch": 3.9829888027562443, "grad_norm": 0.37166357040405273, "learning_rate": 0.0005406260939814191, "loss": 3.3946, "step": 14800 }, { "epoch": 3.9964470284237725, "grad_norm": 0.3421415090560913, "learning_rate": 0.0005404241281809614, "loss": 3.391, "step": 14850 }, { "epoch": 4.00968992248062, "grad_norm": 0.3298637866973877, "learning_rate": 0.0005402221623805035, "loss": 3.3309, "step": 14900 }, { "epoch": 4.023148148148148, "grad_norm": 0.35694819688796997, "learning_rate": 0.0005400201965800457, "loss": 3.2933, "step": 14950 }, { "epoch": 4.036606373815676, "grad_norm": 0.3794950842857361, "learning_rate": 0.000539818230779588, "loss": 3.2948, "step": 15000 }, { "epoch": 4.036606373815676, "eval_accuracy": 0.37712822893022396, "eval_loss": 3.417656183242798, "eval_runtime": 53.6437, "eval_samples_per_second": 335.771, "eval_steps_per_second": 20.99, "step": 15000 }, { "epoch": 4.050064599483204, "grad_norm": 0.366841584444046, "learning_rate": 0.0005396162649791301, "loss": 3.312, "step": 15050 }, { "epoch": 4.0635228251507325, "grad_norm": 0.3495643138885498, "learning_rate": 0.0005394142991786724, "loss": 3.3137, "step": 15100 }, { "epoch": 4.07698105081826, "grad_norm": 0.3431321382522583, "learning_rate": 0.0005392123333782145, "loss": 3.3048, "step": 15150 }, { "epoch": 4.090439276485788, "grad_norm": 0.36576730012893677, "learning_rate": 0.0005390103675777568, "loss": 3.313, "step": 15200 }, { "epoch": 4.103897502153316, "grad_norm": 0.36886483430862427, "learning_rate": 0.000538808401777299, "loss": 3.3163, "step": 15250 }, { "epoch": 4.117355727820844, "grad_norm": 0.35335373878479004, "learning_rate": 0.0005386064359768412, "loss": 3.3143, "step": 15300 }, { "epoch": 4.1308139534883725, "grad_norm": 0.35200396180152893, "learning_rate": 0.0005384044701763834, "loss": 3.3102, "step": 15350 }, { "epoch": 4.1442721791559, "grad_norm": 0.331868052482605, "learning_rate": 0.0005382025043759256, "loss": 3.3176, "step": 15400 }, { "epoch": 4.157730404823428, "grad_norm": 0.3521571159362793, "learning_rate": 0.0005380005385754678, "loss": 3.3228, "step": 15450 }, { "epoch": 4.171188630490956, "grad_norm": 0.3609760105609894, "learning_rate": 0.0005377985727750101, "loss": 3.2974, "step": 15500 }, { "epoch": 4.184646856158484, "grad_norm": 0.36947956681251526, "learning_rate": 0.0005375966069745523, "loss": 3.3197, "step": 15550 }, { "epoch": 4.198105081826012, "grad_norm": 0.3476182222366333, "learning_rate": 0.0005373946411740945, "loss": 3.319, "step": 15600 }, { "epoch": 4.21156330749354, "grad_norm": 0.3870675265789032, "learning_rate": 0.0005371926753736367, "loss": 3.3361, "step": 15650 }, { "epoch": 4.225021533161068, "grad_norm": 0.33378762006759644, "learning_rate": 0.000536990709573179, "loss": 3.3219, "step": 15700 }, { "epoch": 4.238479758828596, "grad_norm": 0.3631652593612671, "learning_rate": 0.0005367887437727211, "loss": 3.3383, "step": 15750 }, { "epoch": 4.251937984496124, "grad_norm": 0.34275636076927185, "learning_rate": 0.0005365867779722634, "loss": 3.3245, "step": 15800 }, { "epoch": 4.265396210163652, "grad_norm": 0.3362019956111908, "learning_rate": 0.0005363848121718055, "loss": 3.3262, "step": 15850 }, { "epoch": 4.27885443583118, "grad_norm": 0.32696837186813354, "learning_rate": 0.0005361828463713477, "loss": 3.3358, "step": 15900 }, { "epoch": 4.292312661498708, "grad_norm": 0.3329922556877136, "learning_rate": 0.00053598088057089, "loss": 3.3274, "step": 15950 }, { "epoch": 4.305770887166236, "grad_norm": 0.3461296260356903, "learning_rate": 0.0005357789147704321, "loss": 3.332, "step": 16000 }, { "epoch": 4.305770887166236, "eval_accuracy": 0.37815222554391875, "eval_loss": 3.406710147857666, "eval_runtime": 53.6423, "eval_samples_per_second": 335.78, "eval_steps_per_second": 20.991, "step": 16000 }, { "epoch": 4.319229112833764, "grad_norm": 0.36084792017936707, "learning_rate": 0.0005355769489699744, "loss": 3.3323, "step": 16050 }, { "epoch": 4.332687338501292, "grad_norm": 0.39574679732322693, "learning_rate": 0.0005353749831695165, "loss": 3.334, "step": 16100 }, { "epoch": 4.34614556416882, "grad_norm": 0.3344963788986206, "learning_rate": 0.0005351730173690588, "loss": 3.3347, "step": 16150 }, { "epoch": 4.359603789836348, "grad_norm": 0.33946841955184937, "learning_rate": 0.000534971051568601, "loss": 3.3294, "step": 16200 }, { "epoch": 4.373062015503876, "grad_norm": 0.35259753465652466, "learning_rate": 0.0005347690857681432, "loss": 3.3383, "step": 16250 }, { "epoch": 4.386520241171404, "grad_norm": 0.3345356583595276, "learning_rate": 0.0005345671199676854, "loss": 3.3435, "step": 16300 }, { "epoch": 4.399978466838932, "grad_norm": 0.33119991421699524, "learning_rate": 0.0005343651541672276, "loss": 3.337, "step": 16350 }, { "epoch": 4.41343669250646, "grad_norm": 0.33489561080932617, "learning_rate": 0.0005341631883667699, "loss": 3.3357, "step": 16400 }, { "epoch": 4.426894918173988, "grad_norm": 0.36713531613349915, "learning_rate": 0.0005339612225663121, "loss": 3.3417, "step": 16450 }, { "epoch": 4.440353143841516, "grad_norm": 0.3532050549983978, "learning_rate": 0.0005337592567658543, "loss": 3.3358, "step": 16500 }, { "epoch": 4.453811369509044, "grad_norm": 0.3615681529045105, "learning_rate": 0.0005335572909653965, "loss": 3.3381, "step": 16550 }, { "epoch": 4.467269595176572, "grad_norm": 0.3489953279495239, "learning_rate": 0.0005333553251649387, "loss": 3.3482, "step": 16600 }, { "epoch": 4.4807278208441, "grad_norm": 0.354987233877182, "learning_rate": 0.000533153359364481, "loss": 3.3319, "step": 16650 }, { "epoch": 4.4941860465116275, "grad_norm": 0.35782018303871155, "learning_rate": 0.0005329513935640231, "loss": 3.3424, "step": 16700 }, { "epoch": 4.507644272179156, "grad_norm": 0.360448956489563, "learning_rate": 0.0005327494277635654, "loss": 3.3438, "step": 16750 }, { "epoch": 4.521102497846684, "grad_norm": 0.38995808362960815, "learning_rate": 0.0005325474619631075, "loss": 3.3502, "step": 16800 }, { "epoch": 4.534560723514212, "grad_norm": 0.3384845554828644, "learning_rate": 0.0005323454961626497, "loss": 3.3371, "step": 16850 }, { "epoch": 4.54801894918174, "grad_norm": 0.3251652717590332, "learning_rate": 0.000532143530362192, "loss": 3.3356, "step": 16900 }, { "epoch": 4.5614771748492675, "grad_norm": 0.34360337257385254, "learning_rate": 0.0005319415645617341, "loss": 3.3334, "step": 16950 }, { "epoch": 4.574935400516796, "grad_norm": 0.35140931606292725, "learning_rate": 0.0005317395987612764, "loss": 3.3342, "step": 17000 }, { "epoch": 4.574935400516796, "eval_accuracy": 0.37956517790053423, "eval_loss": 3.3962960243225098, "eval_runtime": 53.7518, "eval_samples_per_second": 335.096, "eval_steps_per_second": 20.948, "step": 17000 }, { "epoch": 4.588393626184324, "grad_norm": 0.334964781999588, "learning_rate": 0.0005315376329608185, "loss": 3.3331, "step": 17050 }, { "epoch": 4.601851851851852, "grad_norm": 0.33990418910980225, "learning_rate": 0.0005313356671603608, "loss": 3.3411, "step": 17100 }, { "epoch": 4.61531007751938, "grad_norm": 0.3458724319934845, "learning_rate": 0.000531133701359903, "loss": 3.3441, "step": 17150 }, { "epoch": 4.628768303186908, "grad_norm": 0.35447219014167786, "learning_rate": 0.0005309317355594452, "loss": 3.3376, "step": 17200 }, { "epoch": 4.642226528854436, "grad_norm": 0.34828895330429077, "learning_rate": 0.0005307297697589874, "loss": 3.3317, "step": 17250 }, { "epoch": 4.655684754521964, "grad_norm": 0.33856168389320374, "learning_rate": 0.0005305278039585297, "loss": 3.3392, "step": 17300 }, { "epoch": 4.669142980189492, "grad_norm": 0.3794487416744232, "learning_rate": 0.0005303258381580719, "loss": 3.3386, "step": 17350 }, { "epoch": 4.682601205857019, "grad_norm": 0.3594615161418915, "learning_rate": 0.0005301238723576141, "loss": 3.3562, "step": 17400 }, { "epoch": 4.696059431524548, "grad_norm": 0.32063180208206177, "learning_rate": 0.0005299219065571563, "loss": 3.3455, "step": 17450 }, { "epoch": 4.709517657192076, "grad_norm": 0.34606829285621643, "learning_rate": 0.0005297199407566985, "loss": 3.3328, "step": 17500 }, { "epoch": 4.722975882859604, "grad_norm": 0.3413867652416229, "learning_rate": 0.0005295179749562407, "loss": 3.3536, "step": 17550 }, { "epoch": 4.736434108527131, "grad_norm": 0.34495651721954346, "learning_rate": 0.000529316009155783, "loss": 3.3371, "step": 17600 }, { "epoch": 4.7498923341946595, "grad_norm": 0.3626272678375244, "learning_rate": 0.0005291140433553251, "loss": 3.3512, "step": 17650 }, { "epoch": 4.763350559862188, "grad_norm": 0.3632408380508423, "learning_rate": 0.0005289120775548673, "loss": 3.3337, "step": 17700 }, { "epoch": 4.776808785529716, "grad_norm": 0.3298455774784088, "learning_rate": 0.0005287101117544095, "loss": 3.3392, "step": 17750 }, { "epoch": 4.790267011197244, "grad_norm": 0.3547931909561157, "learning_rate": 0.0005285081459539517, "loss": 3.3377, "step": 17800 }, { "epoch": 4.803725236864771, "grad_norm": 0.3601723611354828, "learning_rate": 0.000528306180153494, "loss": 3.3349, "step": 17850 }, { "epoch": 4.8171834625323, "grad_norm": 0.3788248598575592, "learning_rate": 0.0005281042143530361, "loss": 3.3452, "step": 17900 }, { "epoch": 4.830641688199828, "grad_norm": 0.36663538217544556, "learning_rate": 0.0005279022485525784, "loss": 3.3385, "step": 17950 }, { "epoch": 4.844099913867356, "grad_norm": 0.3512565791606903, "learning_rate": 0.0005277002827521205, "loss": 3.3418, "step": 18000 }, { "epoch": 4.844099913867356, "eval_accuracy": 0.380534307852169, "eval_loss": 3.3869707584381104, "eval_runtime": 53.4631, "eval_samples_per_second": 336.905, "eval_steps_per_second": 21.061, "step": 18000 }, { "epoch": 4.857558139534884, "grad_norm": 0.38106250762939453, "learning_rate": 0.0005274983169516628, "loss": 3.3281, "step": 18050 }, { "epoch": 4.871016365202411, "grad_norm": 0.32285740971565247, "learning_rate": 0.000527296351151205, "loss": 3.3402, "step": 18100 }, { "epoch": 4.88447459086994, "grad_norm": 0.33819901943206787, "learning_rate": 0.0005270943853507472, "loss": 3.3322, "step": 18150 }, { "epoch": 4.897932816537468, "grad_norm": 0.3418741822242737, "learning_rate": 0.0005268924195502894, "loss": 3.3459, "step": 18200 }, { "epoch": 4.911391042204996, "grad_norm": 0.35104596614837646, "learning_rate": 0.0005266904537498317, "loss": 3.3498, "step": 18250 }, { "epoch": 4.924849267872523, "grad_norm": 0.3623558282852173, "learning_rate": 0.0005264884879493739, "loss": 3.3433, "step": 18300 }, { "epoch": 4.9383074935400515, "grad_norm": 0.34203389286994934, "learning_rate": 0.0005262865221489161, "loss": 3.3313, "step": 18350 }, { "epoch": 4.95176571920758, "grad_norm": 0.34537750482559204, "learning_rate": 0.0005260845563484583, "loss": 3.328, "step": 18400 }, { "epoch": 4.965223944875108, "grad_norm": 0.3605607748031616, "learning_rate": 0.0005258825905480005, "loss": 3.3394, "step": 18450 }, { "epoch": 4.978682170542635, "grad_norm": 0.346736878156662, "learning_rate": 0.0005256806247475427, "loss": 3.3532, "step": 18500 }, { "epoch": 4.992140396210163, "grad_norm": 0.33791518211364746, "learning_rate": 0.000525478658947085, "loss": 3.3278, "step": 18550 }, { "epoch": 5.0053832902670115, "grad_norm": 0.33499035239219666, "learning_rate": 0.0005252766931466271, "loss": 3.3047, "step": 18600 }, { "epoch": 5.018841515934539, "grad_norm": 0.38031017780303955, "learning_rate": 0.0005250747273461693, "loss": 3.2399, "step": 18650 }, { "epoch": 5.032299741602067, "grad_norm": 0.35386326909065247, "learning_rate": 0.0005248727615457115, "loss": 3.2533, "step": 18700 }, { "epoch": 5.045757967269595, "grad_norm": 0.364179402589798, "learning_rate": 0.0005246707957452537, "loss": 3.2444, "step": 18750 }, { "epoch": 5.059216192937123, "grad_norm": 0.36041221022605896, "learning_rate": 0.000524468829944796, "loss": 3.2541, "step": 18800 }, { "epoch": 5.0726744186046515, "grad_norm": 0.3468455672264099, "learning_rate": 0.0005242668641443381, "loss": 3.2394, "step": 18850 }, { "epoch": 5.086132644272179, "grad_norm": 0.34421736001968384, "learning_rate": 0.0005240648983438804, "loss": 3.2563, "step": 18900 }, { "epoch": 5.099590869939707, "grad_norm": 0.36943989992141724, "learning_rate": 0.0005238629325434225, "loss": 3.2654, "step": 18950 }, { "epoch": 5.113049095607235, "grad_norm": 0.3361397087574005, "learning_rate": 0.0005236609667429648, "loss": 3.263, "step": 19000 }, { "epoch": 5.113049095607235, "eval_accuracy": 0.38084492921222773, "eval_loss": 3.3863155841827393, "eval_runtime": 53.1852, "eval_samples_per_second": 338.666, "eval_steps_per_second": 21.171, "step": 19000 }, { "epoch": 5.126507321274763, "grad_norm": 0.3635759651660919, "learning_rate": 0.000523459000942507, "loss": 3.2646, "step": 19050 }, { "epoch": 5.139965546942291, "grad_norm": 0.3344425559043884, "learning_rate": 0.0005232570351420493, "loss": 3.2643, "step": 19100 }, { "epoch": 5.153423772609819, "grad_norm": 0.3406955599784851, "learning_rate": 0.0005230550693415915, "loss": 3.2682, "step": 19150 }, { "epoch": 5.166881998277347, "grad_norm": 0.35041916370391846, "learning_rate": 0.0005228531035411337, "loss": 3.2672, "step": 19200 }, { "epoch": 5.180340223944875, "grad_norm": 0.35396963357925415, "learning_rate": 0.0005226511377406759, "loss": 3.2665, "step": 19250 }, { "epoch": 5.1937984496124034, "grad_norm": 0.352710098028183, "learning_rate": 0.0005224491719402181, "loss": 3.2818, "step": 19300 }, { "epoch": 5.207256675279931, "grad_norm": 0.37922459840774536, "learning_rate": 0.0005222472061397603, "loss": 3.2652, "step": 19350 }, { "epoch": 5.220714900947459, "grad_norm": 0.37260207533836365, "learning_rate": 0.0005220452403393026, "loss": 3.2688, "step": 19400 }, { "epoch": 5.234173126614987, "grad_norm": 0.33815115690231323, "learning_rate": 0.0005218432745388447, "loss": 3.2846, "step": 19450 }, { "epoch": 5.247631352282515, "grad_norm": 0.3498179316520691, "learning_rate": 0.000521641308738387, "loss": 3.2794, "step": 19500 }, { "epoch": 5.2610895779500435, "grad_norm": 0.3579239249229431, "learning_rate": 0.0005214393429379291, "loss": 3.2748, "step": 19550 }, { "epoch": 5.274547803617571, "grad_norm": 0.3630642890930176, "learning_rate": 0.0005212373771374713, "loss": 3.2801, "step": 19600 }, { "epoch": 5.288006029285099, "grad_norm": 0.33767256140708923, "learning_rate": 0.0005210354113370136, "loss": 3.2743, "step": 19650 }, { "epoch": 5.301464254952627, "grad_norm": 0.339458703994751, "learning_rate": 0.0005208334455365557, "loss": 3.2685, "step": 19700 }, { "epoch": 5.314922480620155, "grad_norm": 0.3411153256893158, "learning_rate": 0.000520631479736098, "loss": 3.274, "step": 19750 }, { "epoch": 5.328380706287683, "grad_norm": 0.3428064286708832, "learning_rate": 0.0005204295139356401, "loss": 3.2723, "step": 19800 }, { "epoch": 5.341838931955211, "grad_norm": 0.3319042921066284, "learning_rate": 0.0005202275481351824, "loss": 3.2836, "step": 19850 }, { "epoch": 5.355297157622739, "grad_norm": 0.34343960881233215, "learning_rate": 0.0005200255823347246, "loss": 3.2763, "step": 19900 }, { "epoch": 5.368755383290267, "grad_norm": 0.3530113101005554, "learning_rate": 0.0005198236165342669, "loss": 3.2781, "step": 19950 }, { "epoch": 5.3822136089577945, "grad_norm": 0.35422250628471375, "learning_rate": 0.000519621650733809, "loss": 3.2798, "step": 20000 }, { "epoch": 5.3822136089577945, "eval_accuracy": 0.3819953907657995, "eval_loss": 3.3763461112976074, "eval_runtime": 53.1398, "eval_samples_per_second": 338.955, "eval_steps_per_second": 21.189, "step": 20000 }, { "epoch": 5.395671834625323, "grad_norm": 0.32547879219055176, "learning_rate": 0.0005194196849333513, "loss": 3.2788, "step": 20050 }, { "epoch": 5.409130060292851, "grad_norm": 0.3453685939311981, "learning_rate": 0.0005192177191328935, "loss": 3.2891, "step": 20100 }, { "epoch": 5.422588285960379, "grad_norm": 0.35231003165245056, "learning_rate": 0.0005190157533324357, "loss": 3.2897, "step": 20150 }, { "epoch": 5.436046511627907, "grad_norm": 0.32846277952194214, "learning_rate": 0.0005188137875319779, "loss": 3.2868, "step": 20200 }, { "epoch": 5.449504737295435, "grad_norm": 0.32960057258605957, "learning_rate": 0.00051861182173152, "loss": 3.2865, "step": 20250 }, { "epoch": 5.462962962962963, "grad_norm": 0.3438560664653778, "learning_rate": 0.0005184098559310623, "loss": 3.2808, "step": 20300 }, { "epoch": 5.476421188630491, "grad_norm": 0.33576279878616333, "learning_rate": 0.0005182078901306046, "loss": 3.2908, "step": 20350 }, { "epoch": 5.489879414298019, "grad_norm": 0.3566272556781769, "learning_rate": 0.0005180059243301467, "loss": 3.2877, "step": 20400 }, { "epoch": 5.503337639965547, "grad_norm": 0.3287167549133301, "learning_rate": 0.000517803958529689, "loss": 3.2831, "step": 20450 }, { "epoch": 5.516795865633075, "grad_norm": 0.3539256155490875, "learning_rate": 0.0005176019927292311, "loss": 3.2837, "step": 20500 }, { "epoch": 5.530254091300603, "grad_norm": 0.3405372202396393, "learning_rate": 0.0005174000269287733, "loss": 3.2892, "step": 20550 }, { "epoch": 5.543712316968131, "grad_norm": 0.3327963948249817, "learning_rate": 0.0005171980611283156, "loss": 3.2912, "step": 20600 }, { "epoch": 5.557170542635659, "grad_norm": 0.36150923371315, "learning_rate": 0.0005169960953278577, "loss": 3.296, "step": 20650 }, { "epoch": 5.5706287683031865, "grad_norm": 0.34517911076545715, "learning_rate": 0.0005167941295274, "loss": 3.2856, "step": 20700 }, { "epoch": 5.584086993970715, "grad_norm": 0.3302103579044342, "learning_rate": 0.0005165921637269421, "loss": 3.2814, "step": 20750 }, { "epoch": 5.597545219638243, "grad_norm": 0.3642379641532898, "learning_rate": 0.0005163901979264844, "loss": 3.286, "step": 20800 }, { "epoch": 5.611003445305771, "grad_norm": 0.3356991410255432, "learning_rate": 0.0005161882321260266, "loss": 3.2943, "step": 20850 }, { "epoch": 5.624461670973298, "grad_norm": 0.34540364146232605, "learning_rate": 0.0005159862663255689, "loss": 3.2879, "step": 20900 }, { "epoch": 5.637919896640827, "grad_norm": 0.3615170121192932, "learning_rate": 0.000515784300525111, "loss": 3.3059, "step": 20950 }, { "epoch": 5.651378122308355, "grad_norm": 0.36911121010780334, "learning_rate": 0.0005155823347246533, "loss": 3.2997, "step": 21000 }, { "epoch": 5.651378122308355, "eval_accuracy": 0.38274592324403867, "eval_loss": 3.367357015609741, "eval_runtime": 53.3615, "eval_samples_per_second": 337.547, "eval_steps_per_second": 21.101, "step": 21000 }, { "epoch": 5.664836347975883, "grad_norm": 0.3656895160675049, "learning_rate": 0.0005153803689241955, "loss": 3.3019, "step": 21050 }, { "epoch": 5.678294573643411, "grad_norm": 0.349932998418808, "learning_rate": 0.0005151784031237377, "loss": 3.2888, "step": 21100 }, { "epoch": 5.6917527993109385, "grad_norm": 0.329196035861969, "learning_rate": 0.0005149764373232799, "loss": 3.2977, "step": 21150 }, { "epoch": 5.705211024978467, "grad_norm": 0.3490157425403595, "learning_rate": 0.000514774471522822, "loss": 3.2962, "step": 21200 }, { "epoch": 5.718669250645995, "grad_norm": 0.4173871874809265, "learning_rate": 0.0005145725057223643, "loss": 3.2923, "step": 21250 }, { "epoch": 5.732127476313523, "grad_norm": 0.3462425172328949, "learning_rate": 0.0005143705399219066, "loss": 3.2846, "step": 21300 }, { "epoch": 5.745585701981051, "grad_norm": 0.3564380407333374, "learning_rate": 0.0005141685741214487, "loss": 3.3123, "step": 21350 }, { "epoch": 5.7590439276485785, "grad_norm": 0.3207317590713501, "learning_rate": 0.0005139666083209909, "loss": 3.3018, "step": 21400 }, { "epoch": 5.772502153316107, "grad_norm": 0.32770583033561707, "learning_rate": 0.0005137646425205331, "loss": 3.2902, "step": 21450 }, { "epoch": 5.785960378983635, "grad_norm": 0.32427579164505005, "learning_rate": 0.0005135626767200753, "loss": 3.3019, "step": 21500 }, { "epoch": 5.799418604651163, "grad_norm": 0.3444927930831909, "learning_rate": 0.0005133607109196176, "loss": 3.2984, "step": 21550 }, { "epoch": 5.81287683031869, "grad_norm": 0.3426213264465332, "learning_rate": 0.0005131587451191597, "loss": 3.3149, "step": 21600 }, { "epoch": 5.826335055986219, "grad_norm": 0.3312057554721832, "learning_rate": 0.000512956779318702, "loss": 3.3019, "step": 21650 }, { "epoch": 5.839793281653747, "grad_norm": 0.35812973976135254, "learning_rate": 0.0005127548135182441, "loss": 3.3002, "step": 21700 }, { "epoch": 5.853251507321275, "grad_norm": 0.3365882635116577, "learning_rate": 0.0005125528477177865, "loss": 3.2915, "step": 21750 }, { "epoch": 5.866709732988802, "grad_norm": 0.3192456066608429, "learning_rate": 0.0005123508819173286, "loss": 3.2924, "step": 21800 }, { "epoch": 5.8801679586563305, "grad_norm": 0.3553847074508667, "learning_rate": 0.0005121489161168709, "loss": 3.2976, "step": 21850 }, { "epoch": 5.893626184323859, "grad_norm": 0.33080434799194336, "learning_rate": 0.000511946950316413, "loss": 3.2883, "step": 21900 }, { "epoch": 5.907084409991387, "grad_norm": 0.3338245451450348, "learning_rate": 0.0005117449845159553, "loss": 3.2925, "step": 21950 }, { "epoch": 5.920542635658915, "grad_norm": 0.3583246171474457, "learning_rate": 0.0005115430187154975, "loss": 3.3, "step": 22000 }, { "epoch": 5.920542635658915, "eval_accuracy": 0.3836738760374145, "eval_loss": 3.3571391105651855, "eval_runtime": 53.3686, "eval_samples_per_second": 337.502, "eval_steps_per_second": 21.099, "step": 22000 }, { "epoch": 5.934000861326442, "grad_norm": 0.3355707824230194, "learning_rate": 0.0005113410529150397, "loss": 3.3096, "step": 22050 }, { "epoch": 5.9474590869939705, "grad_norm": 0.38139769434928894, "learning_rate": 0.0005111390871145819, "loss": 3.3018, "step": 22100 }, { "epoch": 5.960917312661499, "grad_norm": 0.32805705070495605, "learning_rate": 0.000510937121314124, "loss": 3.3019, "step": 22150 }, { "epoch": 5.974375538329027, "grad_norm": 0.3372587561607361, "learning_rate": 0.0005107351555136663, "loss": 3.3008, "step": 22200 }, { "epoch": 5.987833763996555, "grad_norm": 0.3290460407733917, "learning_rate": 0.0005105331897132085, "loss": 3.2948, "step": 22250 }, { "epoch": 6.001076658053402, "grad_norm": 0.3840121328830719, "learning_rate": 0.0005103312239127507, "loss": 3.295, "step": 22300 }, { "epoch": 6.0145348837209305, "grad_norm": 0.33211490511894226, "learning_rate": 0.0005101292581122929, "loss": 3.1891, "step": 22350 }, { "epoch": 6.027993109388458, "grad_norm": 0.3395225405693054, "learning_rate": 0.0005099272923118351, "loss": 3.2021, "step": 22400 }, { "epoch": 6.041451335055986, "grad_norm": 0.3511013388633728, "learning_rate": 0.0005097253265113773, "loss": 3.2057, "step": 22450 }, { "epoch": 6.054909560723514, "grad_norm": 0.365784615278244, "learning_rate": 0.0005095233607109196, "loss": 3.2036, "step": 22500 }, { "epoch": 6.068367786391042, "grad_norm": 0.34816962480545044, "learning_rate": 0.0005093213949104617, "loss": 3.2033, "step": 22550 }, { "epoch": 6.0818260120585705, "grad_norm": 0.3428509831428528, "learning_rate": 0.000509119429110004, "loss": 3.2188, "step": 22600 }, { "epoch": 6.095284237726098, "grad_norm": 0.3539871871471405, "learning_rate": 0.0005089174633095462, "loss": 3.2135, "step": 22650 }, { "epoch": 6.108742463393626, "grad_norm": 0.37805479764938354, "learning_rate": 0.0005087154975090885, "loss": 3.2219, "step": 22700 }, { "epoch": 6.122200689061154, "grad_norm": 0.40196943283081055, "learning_rate": 0.0005085135317086306, "loss": 3.2134, "step": 22750 }, { "epoch": 6.135658914728682, "grad_norm": 0.3432955741882324, "learning_rate": 0.0005083115659081729, "loss": 3.2254, "step": 22800 }, { "epoch": 6.149117140396211, "grad_norm": 0.4084476828575134, "learning_rate": 0.0005081096001077151, "loss": 3.2232, "step": 22850 }, { "epoch": 6.162575366063738, "grad_norm": 0.364326536655426, "learning_rate": 0.0005079076343072573, "loss": 3.2344, "step": 22900 }, { "epoch": 6.176033591731266, "grad_norm": 0.34070107340812683, "learning_rate": 0.0005077056685067995, "loss": 3.229, "step": 22950 }, { "epoch": 6.189491817398794, "grad_norm": 0.34302592277526855, "learning_rate": 0.0005075037027063417, "loss": 3.2446, "step": 23000 }, { "epoch": 6.189491817398794, "eval_accuracy": 0.383644324092701, "eval_loss": 3.3621912002563477, "eval_runtime": 53.322, "eval_samples_per_second": 337.797, "eval_steps_per_second": 21.117, "step": 23000 }, { "epoch": 6.2029500430663225, "grad_norm": 0.353242963552475, "learning_rate": 0.0005073017369058839, "loss": 3.2304, "step": 23050 }, { "epoch": 6.21640826873385, "grad_norm": 0.3593711256980896, "learning_rate": 0.0005070997711054262, "loss": 3.233, "step": 23100 }, { "epoch": 6.229866494401378, "grad_norm": 0.3433372676372528, "learning_rate": 0.0005068978053049683, "loss": 3.2408, "step": 23150 }, { "epoch": 6.243324720068906, "grad_norm": 0.3444235026836395, "learning_rate": 0.0005066958395045105, "loss": 3.2375, "step": 23200 }, { "epoch": 6.256782945736434, "grad_norm": 0.3466513156890869, "learning_rate": 0.0005064938737040527, "loss": 3.2338, "step": 23250 }, { "epoch": 6.270241171403962, "grad_norm": 0.4486019015312195, "learning_rate": 0.0005062919079035949, "loss": 3.2367, "step": 23300 }, { "epoch": 6.28369939707149, "grad_norm": 0.35569101572036743, "learning_rate": 0.0005060899421031372, "loss": 3.2306, "step": 23350 }, { "epoch": 6.297157622739018, "grad_norm": 0.34310290217399597, "learning_rate": 0.0005058879763026793, "loss": 3.2457, "step": 23400 }, { "epoch": 6.310615848406546, "grad_norm": 0.34421324729919434, "learning_rate": 0.0005056860105022216, "loss": 3.2488, "step": 23450 }, { "epoch": 6.324074074074074, "grad_norm": 0.34529271721839905, "learning_rate": 0.0005054840447017637, "loss": 3.2421, "step": 23500 }, { "epoch": 6.337532299741602, "grad_norm": 0.3405730128288269, "learning_rate": 0.0005052820789013061, "loss": 3.2323, "step": 23550 }, { "epoch": 6.35099052540913, "grad_norm": 0.3444826900959015, "learning_rate": 0.0005050801131008482, "loss": 3.244, "step": 23600 }, { "epoch": 6.364448751076658, "grad_norm": 0.332755446434021, "learning_rate": 0.0005048781473003905, "loss": 3.2356, "step": 23650 }, { "epoch": 6.377906976744186, "grad_norm": 0.3805544376373291, "learning_rate": 0.0005046761814999326, "loss": 3.2533, "step": 23700 }, { "epoch": 6.3913652024117145, "grad_norm": 0.35615286231040955, "learning_rate": 0.0005044742156994749, "loss": 3.243, "step": 23750 }, { "epoch": 6.404823428079242, "grad_norm": 0.340789258480072, "learning_rate": 0.0005042722498990171, "loss": 3.2459, "step": 23800 }, { "epoch": 6.41828165374677, "grad_norm": 0.338225781917572, "learning_rate": 0.0005040702840985593, "loss": 3.242, "step": 23850 }, { "epoch": 6.431739879414298, "grad_norm": 0.3471758961677551, "learning_rate": 0.0005038683182981015, "loss": 3.2507, "step": 23900 }, { "epoch": 6.445198105081826, "grad_norm": 0.3620913624763489, "learning_rate": 0.0005036663524976436, "loss": 3.2475, "step": 23950 }, { "epoch": 6.458656330749354, "grad_norm": 0.3431723415851593, "learning_rate": 0.0005034643866971859, "loss": 3.2451, "step": 24000 }, { "epoch": 6.458656330749354, "eval_accuracy": 0.3841879929579454, "eval_loss": 3.356856346130371, "eval_runtime": 53.3316, "eval_samples_per_second": 337.736, "eval_steps_per_second": 21.113, "step": 24000 }, { "epoch": 6.472114556416882, "grad_norm": 0.36163797974586487, "learning_rate": 0.0005032624208967281, "loss": 3.2586, "step": 24050 }, { "epoch": 6.48557278208441, "grad_norm": 0.32522052526474, "learning_rate": 0.0005030604550962703, "loss": 3.2529, "step": 24100 }, { "epoch": 6.499031007751938, "grad_norm": 0.3808082044124603, "learning_rate": 0.0005028584892958125, "loss": 3.2499, "step": 24150 }, { "epoch": 6.5124892334194655, "grad_norm": 0.33839014172554016, "learning_rate": 0.0005026565234953547, "loss": 3.2481, "step": 24200 }, { "epoch": 6.525947459086994, "grad_norm": 0.3400689363479614, "learning_rate": 0.0005024545576948969, "loss": 3.2538, "step": 24250 }, { "epoch": 6.539405684754522, "grad_norm": 0.34009531140327454, "learning_rate": 0.0005022525918944392, "loss": 3.2564, "step": 24300 }, { "epoch": 6.55286391042205, "grad_norm": 0.3178061842918396, "learning_rate": 0.0005020506260939813, "loss": 3.2522, "step": 24350 }, { "epoch": 6.566322136089578, "grad_norm": 0.3306698799133301, "learning_rate": 0.0005018486602935236, "loss": 3.2552, "step": 24400 }, { "epoch": 6.579780361757106, "grad_norm": 0.35276076197624207, "learning_rate": 0.0005016466944930658, "loss": 3.2585, "step": 24450 }, { "epoch": 6.593238587424634, "grad_norm": 0.34085813164711, "learning_rate": 0.0005014447286926081, "loss": 3.2659, "step": 24500 }, { "epoch": 6.606696813092162, "grad_norm": 0.34538936614990234, "learning_rate": 0.0005012427628921502, "loss": 3.2513, "step": 24550 }, { "epoch": 6.62015503875969, "grad_norm": 0.33835741877555847, "learning_rate": 0.0005010407970916925, "loss": 3.257, "step": 24600 }, { "epoch": 6.633613264427218, "grad_norm": 0.3618541359901428, "learning_rate": 0.0005008388312912346, "loss": 3.2474, "step": 24650 }, { "epoch": 6.647071490094746, "grad_norm": 0.3588375449180603, "learning_rate": 0.0005006368654907769, "loss": 3.2441, "step": 24700 }, { "epoch": 6.660529715762274, "grad_norm": 0.35771581530570984, "learning_rate": 0.0005004348996903191, "loss": 3.2597, "step": 24750 }, { "epoch": 6.673987941429802, "grad_norm": 0.35776013135910034, "learning_rate": 0.0005002329338898613, "loss": 3.2605, "step": 24800 }, { "epoch": 6.68744616709733, "grad_norm": 0.35457316040992737, "learning_rate": 0.0005000309680894035, "loss": 3.2593, "step": 24850 }, { "epoch": 6.7009043927648575, "grad_norm": 0.3595224916934967, "learning_rate": 0.0004998290022889456, "loss": 3.2642, "step": 24900 }, { "epoch": 6.714362618432386, "grad_norm": 0.3422452509403229, "learning_rate": 0.0004996270364884879, "loss": 3.2614, "step": 24950 }, { "epoch": 6.727820844099914, "grad_norm": 0.335967093706131, "learning_rate": 0.0004994250706880301, "loss": 3.2585, "step": 25000 }, { "epoch": 6.727820844099914, "eval_accuracy": 0.38525447049216593, "eval_loss": 3.3472113609313965, "eval_runtime": 53.3326, "eval_samples_per_second": 337.73, "eval_steps_per_second": 21.113, "step": 25000 }, { "epoch": 6.741279069767442, "grad_norm": 0.33750924468040466, "learning_rate": 0.0004992231048875723, "loss": 3.2587, "step": 25050 }, { "epoch": 6.754737295434969, "grad_norm": 0.3297862708568573, "learning_rate": 0.0004990211390871145, "loss": 3.264, "step": 25100 }, { "epoch": 6.768195521102498, "grad_norm": 0.3606378436088562, "learning_rate": 0.0004988191732866567, "loss": 3.2574, "step": 25150 }, { "epoch": 6.781653746770026, "grad_norm": 0.3313273787498474, "learning_rate": 0.0004986172074861989, "loss": 3.2519, "step": 25200 }, { "epoch": 6.795111972437554, "grad_norm": 0.348251610994339, "learning_rate": 0.0004984152416857412, "loss": 3.2619, "step": 25250 }, { "epoch": 6.808570198105082, "grad_norm": 0.33993712067604065, "learning_rate": 0.0004982132758852834, "loss": 3.2576, "step": 25300 }, { "epoch": 6.822028423772609, "grad_norm": 0.34746164083480835, "learning_rate": 0.0004980113100848256, "loss": 3.2612, "step": 25350 }, { "epoch": 6.835486649440138, "grad_norm": 0.3444181978702545, "learning_rate": 0.0004978093442843678, "loss": 3.2617, "step": 25400 }, { "epoch": 6.848944875107666, "grad_norm": 0.3506355583667755, "learning_rate": 0.0004976073784839101, "loss": 3.2612, "step": 25450 }, { "epoch": 6.862403100775194, "grad_norm": 0.37607342004776, "learning_rate": 0.0004974054126834522, "loss": 3.2593, "step": 25500 }, { "epoch": 6.875861326442722, "grad_norm": 0.33187639713287354, "learning_rate": 0.0004972034468829945, "loss": 3.2613, "step": 25550 }, { "epoch": 6.8893195521102495, "grad_norm": 0.33409222960472107, "learning_rate": 0.0004970014810825366, "loss": 3.2624, "step": 25600 }, { "epoch": 6.902777777777778, "grad_norm": 0.32660526037216187, "learning_rate": 0.0004967995152820789, "loss": 3.2656, "step": 25650 }, { "epoch": 6.916236003445306, "grad_norm": 0.3403673768043518, "learning_rate": 0.0004965975494816211, "loss": 3.2674, "step": 25700 }, { "epoch": 6.929694229112834, "grad_norm": 0.3395448327064514, "learning_rate": 0.0004963955836811633, "loss": 3.2594, "step": 25750 }, { "epoch": 6.943152454780362, "grad_norm": 0.34910374879837036, "learning_rate": 0.0004961936178807055, "loss": 3.2597, "step": 25800 }, { "epoch": 6.9566106804478895, "grad_norm": 0.3608143627643585, "learning_rate": 0.0004959916520802476, "loss": 3.2544, "step": 25850 }, { "epoch": 6.970068906115418, "grad_norm": 0.33373093605041504, "learning_rate": 0.0004957896862797899, "loss": 3.2658, "step": 25900 }, { "epoch": 6.983527131782946, "grad_norm": 0.36241739988327026, "learning_rate": 0.0004955877204793321, "loss": 3.2547, "step": 25950 }, { "epoch": 6.996985357450473, "grad_norm": 0.3471966087818146, "learning_rate": 0.0004953857546788743, "loss": 3.2476, "step": 26000 }, { "epoch": 6.996985357450473, "eval_accuracy": 0.3860393353767634, "eval_loss": 3.3386144638061523, "eval_runtime": 53.2265, "eval_samples_per_second": 338.403, "eval_steps_per_second": 21.155, "step": 26000 }, { "epoch": 7.010228251507321, "grad_norm": 0.32640400528907776, "learning_rate": 0.0004951837888784165, "loss": 3.185, "step": 26050 }, { "epoch": 7.0236864771748495, "grad_norm": 0.3508910536766052, "learning_rate": 0.0004949818230779587, "loss": 3.1646, "step": 26100 }, { "epoch": 7.037144702842378, "grad_norm": 0.3429078757762909, "learning_rate": 0.0004947798572775009, "loss": 3.1603, "step": 26150 }, { "epoch": 7.050602928509905, "grad_norm": 0.31306034326553345, "learning_rate": 0.0004945778914770432, "loss": 3.183, "step": 26200 }, { "epoch": 7.064061154177433, "grad_norm": 0.3623344898223877, "learning_rate": 0.0004943759256765854, "loss": 3.1838, "step": 26250 }, { "epoch": 7.077519379844961, "grad_norm": 0.3504388630390167, "learning_rate": 0.0004941739598761276, "loss": 3.1868, "step": 26300 }, { "epoch": 7.09097760551249, "grad_norm": 0.3301059901714325, "learning_rate": 0.0004939719940756698, "loss": 3.1861, "step": 26350 }, { "epoch": 7.104435831180017, "grad_norm": 0.3660713732242584, "learning_rate": 0.0004937700282752121, "loss": 3.177, "step": 26400 }, { "epoch": 7.117894056847545, "grad_norm": 0.34936845302581787, "learning_rate": 0.0004935680624747542, "loss": 3.1798, "step": 26450 }, { "epoch": 7.131352282515073, "grad_norm": 0.3554314374923706, "learning_rate": 0.0004933660966742965, "loss": 3.1774, "step": 26500 }, { "epoch": 7.144810508182601, "grad_norm": 0.3377351760864258, "learning_rate": 0.0004931641308738386, "loss": 3.1898, "step": 26550 }, { "epoch": 7.15826873385013, "grad_norm": 0.35610777139663696, "learning_rate": 0.0004929621650733809, "loss": 3.1923, "step": 26600 }, { "epoch": 7.171726959517657, "grad_norm": 0.3468087911605835, "learning_rate": 0.0004927601992729231, "loss": 3.1869, "step": 26650 }, { "epoch": 7.185185185185185, "grad_norm": 0.34939804673194885, "learning_rate": 0.0004925582334724652, "loss": 3.1861, "step": 26700 }, { "epoch": 7.198643410852713, "grad_norm": 0.3704433739185333, "learning_rate": 0.0004923562676720075, "loss": 3.1948, "step": 26750 }, { "epoch": 7.2121016365202415, "grad_norm": 0.36642688512802124, "learning_rate": 0.0004921543018715497, "loss": 3.1992, "step": 26800 }, { "epoch": 7.225559862187769, "grad_norm": 0.34551697969436646, "learning_rate": 0.0004919523360710919, "loss": 3.196, "step": 26850 }, { "epoch": 7.239018087855297, "grad_norm": 0.33420753479003906, "learning_rate": 0.0004917503702706341, "loss": 3.2072, "step": 26900 }, { "epoch": 7.252476313522825, "grad_norm": 0.36563876271247864, "learning_rate": 0.0004915484044701763, "loss": 3.2002, "step": 26950 }, { "epoch": 7.265934539190353, "grad_norm": 0.3770323693752289, "learning_rate": 0.0004913464386697185, "loss": 3.2025, "step": 27000 }, { "epoch": 7.265934539190353, "eval_accuracy": 0.38613451002223786, "eval_loss": 3.347355604171753, "eval_runtime": 53.2539, "eval_samples_per_second": 338.229, "eval_steps_per_second": 21.144, "step": 27000 }, { "epoch": 7.279392764857882, "grad_norm": 0.34507986903190613, "learning_rate": 0.0004911444728692608, "loss": 3.2055, "step": 27050 }, { "epoch": 7.292850990525409, "grad_norm": 0.32827848196029663, "learning_rate": 0.000490942507068803, "loss": 3.2089, "step": 27100 }, { "epoch": 7.306309216192937, "grad_norm": 0.38157767057418823, "learning_rate": 0.0004907405412683452, "loss": 3.199, "step": 27150 }, { "epoch": 7.319767441860465, "grad_norm": 0.3539191484451294, "learning_rate": 0.0004905385754678874, "loss": 3.2145, "step": 27200 }, { "epoch": 7.333225667527993, "grad_norm": 0.37116944789886475, "learning_rate": 0.0004903366096674297, "loss": 3.2057, "step": 27250 }, { "epoch": 7.346683893195521, "grad_norm": 0.35318002104759216, "learning_rate": 0.0004901346438669718, "loss": 3.203, "step": 27300 }, { "epoch": 7.360142118863049, "grad_norm": 0.3373660147190094, "learning_rate": 0.0004899326780665141, "loss": 3.2205, "step": 27350 }, { "epoch": 7.373600344530577, "grad_norm": 0.36426669359207153, "learning_rate": 0.0004897307122660562, "loss": 3.2197, "step": 27400 }, { "epoch": 7.387058570198105, "grad_norm": 0.3387640118598938, "learning_rate": 0.0004895287464655985, "loss": 3.2181, "step": 27450 }, { "epoch": 7.4005167958656335, "grad_norm": 0.35902735590934753, "learning_rate": 0.0004893267806651407, "loss": 3.213, "step": 27500 }, { "epoch": 7.413975021533161, "grad_norm": 0.3508385419845581, "learning_rate": 0.0004891248148646829, "loss": 3.205, "step": 27550 }, { "epoch": 7.427433247200689, "grad_norm": 0.36484771966934204, "learning_rate": 0.0004889228490642251, "loss": 3.2166, "step": 27600 }, { "epoch": 7.440891472868217, "grad_norm": 0.33635297417640686, "learning_rate": 0.0004887208832637672, "loss": 3.2187, "step": 27650 }, { "epoch": 7.454349698535745, "grad_norm": 0.3573185205459595, "learning_rate": 0.0004885189174633095, "loss": 3.2118, "step": 27700 }, { "epoch": 7.467807924203273, "grad_norm": 0.33649978041648865, "learning_rate": 0.0004883169516628517, "loss": 3.2172, "step": 27750 }, { "epoch": 7.481266149870801, "grad_norm": 0.34777286648750305, "learning_rate": 0.0004881149858623939, "loss": 3.2072, "step": 27800 }, { "epoch": 7.494724375538329, "grad_norm": 0.37506943941116333, "learning_rate": 0.00048791302006193614, "loss": 3.2176, "step": 27850 }, { "epoch": 7.508182601205857, "grad_norm": 0.32190048694610596, "learning_rate": 0.00048771105426147833, "loss": 3.2218, "step": 27900 }, { "epoch": 7.521640826873385, "grad_norm": 0.3466486930847168, "learning_rate": 0.00048750908846102053, "loss": 3.2157, "step": 27950 }, { "epoch": 7.535099052540913, "grad_norm": 0.3516653776168823, "learning_rate": 0.0004873071226605628, "loss": 3.2282, "step": 28000 }, { "epoch": 7.535099052540913, "eval_accuracy": 0.3867050146608067, "eval_loss": 3.338799476623535, "eval_runtime": 53.4413, "eval_samples_per_second": 337.042, "eval_steps_per_second": 21.07, "step": 28000 }, { "epoch": 7.548557278208441, "grad_norm": 0.33157870173454285, "learning_rate": 0.00048710515686010503, "loss": 3.2251, "step": 28050 }, { "epoch": 7.562015503875969, "grad_norm": 0.3672688901424408, "learning_rate": 0.0004869031910596472, "loss": 3.2207, "step": 28100 }, { "epoch": 7.575473729543497, "grad_norm": 0.3594033718109131, "learning_rate": 0.0004867012252591894, "loss": 3.2099, "step": 28150 }, { "epoch": 7.588931955211025, "grad_norm": 0.3516235947608948, "learning_rate": 0.0004864992594587316, "loss": 3.2308, "step": 28200 }, { "epoch": 7.602390180878553, "grad_norm": 0.33869901299476624, "learning_rate": 0.0004862972936582738, "loss": 3.2195, "step": 28250 }, { "epoch": 7.615848406546081, "grad_norm": 0.34088006615638733, "learning_rate": 0.00048609532785781606, "loss": 3.2307, "step": 28300 }, { "epoch": 7.629306632213609, "grad_norm": 0.3419373333454132, "learning_rate": 0.00048589336205735826, "loss": 3.2092, "step": 28350 }, { "epoch": 7.6427648578811365, "grad_norm": 0.34658679366111755, "learning_rate": 0.00048569139625690046, "loss": 3.2286, "step": 28400 }, { "epoch": 7.656223083548665, "grad_norm": 0.3279028534889221, "learning_rate": 0.00048548943045644265, "loss": 3.2199, "step": 28450 }, { "epoch": 7.669681309216193, "grad_norm": 0.3251379728317261, "learning_rate": 0.00048528746465598485, "loss": 3.2218, "step": 28500 }, { "epoch": 7.683139534883721, "grad_norm": 0.3511503338813782, "learning_rate": 0.0004850854988555271, "loss": 3.2342, "step": 28550 }, { "epoch": 7.696597760551249, "grad_norm": 0.3733566105365753, "learning_rate": 0.0004848835330550693, "loss": 3.2329, "step": 28600 }, { "epoch": 7.7100559862187765, "grad_norm": 0.33851513266563416, "learning_rate": 0.0004846815672546115, "loss": 3.2288, "step": 28650 }, { "epoch": 7.723514211886305, "grad_norm": 0.3473910689353943, "learning_rate": 0.0004844796014541537, "loss": 3.2182, "step": 28700 }, { "epoch": 7.736972437553833, "grad_norm": 0.36879199743270874, "learning_rate": 0.0004842776356536959, "loss": 3.2237, "step": 28750 }, { "epoch": 7.750430663221361, "grad_norm": 0.3303454518318176, "learning_rate": 0.00048407566985323813, "loss": 3.2348, "step": 28800 }, { "epoch": 7.763888888888889, "grad_norm": 0.381592720746994, "learning_rate": 0.0004838737040527804, "loss": 3.2268, "step": 28850 }, { "epoch": 7.777347114556417, "grad_norm": 0.36583423614501953, "learning_rate": 0.0004836717382523226, "loss": 3.231, "step": 28900 }, { "epoch": 7.790805340223945, "grad_norm": 0.3528934419155121, "learning_rate": 0.0004834697724518648, "loss": 3.2303, "step": 28950 }, { "epoch": 7.804263565891473, "grad_norm": 0.3354852795600891, "learning_rate": 0.000483267806651407, "loss": 3.2378, "step": 29000 }, { "epoch": 7.804263565891473, "eval_accuracy": 0.38734516193379237, "eval_loss": 3.33156418800354, "eval_runtime": 53.2846, "eval_samples_per_second": 338.034, "eval_steps_per_second": 21.132, "step": 29000 }, { "epoch": 7.817721791559001, "grad_norm": 0.37304386496543884, "learning_rate": 0.0004830658408509492, "loss": 3.2271, "step": 29050 }, { "epoch": 7.831180017226529, "grad_norm": 0.327325701713562, "learning_rate": 0.0004828638750504914, "loss": 3.2244, "step": 29100 }, { "epoch": 7.844638242894057, "grad_norm": 0.35642972588539124, "learning_rate": 0.0004826619092500336, "loss": 3.2359, "step": 29150 }, { "epoch": 7.858096468561585, "grad_norm": 0.33831554651260376, "learning_rate": 0.0004824599434495758, "loss": 3.2359, "step": 29200 }, { "epoch": 7.871554694229113, "grad_norm": 0.33137044310569763, "learning_rate": 0.00048225797764911806, "loss": 3.2295, "step": 29250 }, { "epoch": 7.885012919896641, "grad_norm": 0.34678998589515686, "learning_rate": 0.00048205601184866026, "loss": 3.2316, "step": 29300 }, { "epoch": 7.8984711455641685, "grad_norm": 0.3184164762496948, "learning_rate": 0.00048185404604820245, "loss": 3.24, "step": 29350 }, { "epoch": 7.911929371231697, "grad_norm": 0.3482532799243927, "learning_rate": 0.00048165208024774465, "loss": 3.2323, "step": 29400 }, { "epoch": 7.925387596899225, "grad_norm": 0.33416080474853516, "learning_rate": 0.00048145011444728685, "loss": 3.2354, "step": 29450 }, { "epoch": 7.938845822566753, "grad_norm": 0.3546711802482605, "learning_rate": 0.0004812481486468291, "loss": 3.2299, "step": 29500 }, { "epoch": 7.95230404823428, "grad_norm": 0.3453027307987213, "learning_rate": 0.0004810461828463713, "loss": 3.2455, "step": 29550 }, { "epoch": 7.965762273901809, "grad_norm": 0.3534720242023468, "learning_rate": 0.0004808442170459135, "loss": 3.2348, "step": 29600 }, { "epoch": 7.979220499569337, "grad_norm": 0.34397897124290466, "learning_rate": 0.0004806422512454557, "loss": 3.2336, "step": 29650 }, { "epoch": 7.992678725236865, "grad_norm": 0.35581567883491516, "learning_rate": 0.00048044028544499793, "loss": 3.2316, "step": 29700 }, { "epoch": 8.005921619293712, "grad_norm": 0.37843331694602966, "learning_rate": 0.0004802383196445402, "loss": 3.1745, "step": 29750 }, { "epoch": 8.01937984496124, "grad_norm": 0.38241565227508545, "learning_rate": 0.0004800363538440824, "loss": 3.1428, "step": 29800 }, { "epoch": 8.032838070628769, "grad_norm": 0.382955402135849, "learning_rate": 0.0004798343880436246, "loss": 3.1382, "step": 29850 }, { "epoch": 8.046296296296296, "grad_norm": 0.3386884331703186, "learning_rate": 0.00047963242224316683, "loss": 3.1443, "step": 29900 }, { "epoch": 8.059754521963825, "grad_norm": 0.3572510778903961, "learning_rate": 0.000479430456442709, "loss": 3.1356, "step": 29950 }, { "epoch": 8.073212747631352, "grad_norm": 0.3773634731769562, "learning_rate": 0.0004792284906422512, "loss": 3.1434, "step": 30000 }, { "epoch": 8.073212747631352, "eval_accuracy": 0.3869395832219703, "eval_loss": 3.3406853675842285, "eval_runtime": 53.3677, "eval_samples_per_second": 337.508, "eval_steps_per_second": 21.099, "step": 30000 }, { "epoch": 8.08667097329888, "grad_norm": 0.3504475951194763, "learning_rate": 0.0004790265248417934, "loss": 3.148, "step": 30050 }, { "epoch": 8.100129198966409, "grad_norm": 0.3641934096813202, "learning_rate": 0.0004788245590413356, "loss": 3.1534, "step": 30100 }, { "epoch": 8.113587424633936, "grad_norm": 0.361492782831192, "learning_rate": 0.00047862259324087786, "loss": 3.1513, "step": 30150 }, { "epoch": 8.127045650301465, "grad_norm": 0.3426317572593689, "learning_rate": 0.00047842062744042006, "loss": 3.1629, "step": 30200 }, { "epoch": 8.140503875968992, "grad_norm": 0.3521021604537964, "learning_rate": 0.00047821866163996225, "loss": 3.164, "step": 30250 }, { "epoch": 8.15396210163652, "grad_norm": 0.34249168634414673, "learning_rate": 0.00047801669583950445, "loss": 3.1764, "step": 30300 }, { "epoch": 8.167420327304049, "grad_norm": 0.3548749089241028, "learning_rate": 0.00047781473003904665, "loss": 3.1532, "step": 30350 }, { "epoch": 8.180878552971576, "grad_norm": 0.3407946228981018, "learning_rate": 0.0004776127642385889, "loss": 3.1637, "step": 30400 }, { "epoch": 8.194336778639105, "grad_norm": 0.347978800535202, "learning_rate": 0.0004774107984381311, "loss": 3.174, "step": 30450 }, { "epoch": 8.207795004306632, "grad_norm": 0.36164358258247375, "learning_rate": 0.0004772088326376733, "loss": 3.1653, "step": 30500 }, { "epoch": 8.22125322997416, "grad_norm": 0.3417772650718689, "learning_rate": 0.0004770068668372155, "loss": 3.1683, "step": 30550 }, { "epoch": 8.234711455641689, "grad_norm": 0.3758831024169922, "learning_rate": 0.0004768049010367577, "loss": 3.1801, "step": 30600 }, { "epoch": 8.248169681309216, "grad_norm": 0.3716782033443451, "learning_rate": 0.0004766029352363, "loss": 3.1733, "step": 30650 }, { "epoch": 8.261627906976745, "grad_norm": 0.3527848422527313, "learning_rate": 0.0004764009694358422, "loss": 3.1789, "step": 30700 }, { "epoch": 8.275086132644272, "grad_norm": 0.3328978419303894, "learning_rate": 0.0004761990036353844, "loss": 3.1694, "step": 30750 }, { "epoch": 8.2885443583118, "grad_norm": 0.3273838758468628, "learning_rate": 0.0004759970378349266, "loss": 3.1767, "step": 30800 }, { "epoch": 8.302002583979329, "grad_norm": 0.35219597816467285, "learning_rate": 0.0004757950720344688, "loss": 3.1719, "step": 30850 }, { "epoch": 8.315460809646856, "grad_norm": 0.33684444427490234, "learning_rate": 0.000475593106234011, "loss": 3.1786, "step": 30900 }, { "epoch": 8.328919035314383, "grad_norm": 0.3632763922214508, "learning_rate": 0.0004753911404335532, "loss": 3.185, "step": 30950 }, { "epoch": 8.342377260981912, "grad_norm": 0.3545039892196655, "learning_rate": 0.0004751891746330954, "loss": 3.1867, "step": 31000 }, { "epoch": 8.342377260981912, "eval_accuracy": 0.3874767332758809, "eval_loss": 3.337397575378418, "eval_runtime": 53.4419, "eval_samples_per_second": 337.039, "eval_steps_per_second": 21.07, "step": 31000 }, { "epoch": 8.35583548664944, "grad_norm": 0.3302076756954193, "learning_rate": 0.0004749872088326376, "loss": 3.1899, "step": 31050 }, { "epoch": 8.369293712316969, "grad_norm": 0.33575427532196045, "learning_rate": 0.00047478524303217986, "loss": 3.1839, "step": 31100 }, { "epoch": 8.382751937984496, "grad_norm": 0.35871943831443787, "learning_rate": 0.00047458327723172206, "loss": 3.1823, "step": 31150 }, { "epoch": 8.396210163652023, "grad_norm": 0.357995867729187, "learning_rate": 0.00047438131143126425, "loss": 3.1952, "step": 31200 }, { "epoch": 8.409668389319553, "grad_norm": 0.37615224719047546, "learning_rate": 0.00047417934563080645, "loss": 3.1825, "step": 31250 }, { "epoch": 8.42312661498708, "grad_norm": 0.34395626187324524, "learning_rate": 0.00047397737983034864, "loss": 3.1975, "step": 31300 }, { "epoch": 8.436584840654609, "grad_norm": 0.3637622892856598, "learning_rate": 0.0004737754140298909, "loss": 3.1957, "step": 31350 }, { "epoch": 8.450043066322136, "grad_norm": 0.33689847588539124, "learning_rate": 0.0004735734482294331, "loss": 3.1863, "step": 31400 }, { "epoch": 8.463501291989663, "grad_norm": 0.3541131913661957, "learning_rate": 0.0004733714824289753, "loss": 3.1859, "step": 31450 }, { "epoch": 8.476959517657193, "grad_norm": 0.33903294801712036, "learning_rate": 0.0004731695166285175, "loss": 3.1894, "step": 31500 }, { "epoch": 8.49041774332472, "grad_norm": 0.3368317782878876, "learning_rate": 0.0004729675508280598, "loss": 3.1925, "step": 31550 }, { "epoch": 8.503875968992247, "grad_norm": 0.34673404693603516, "learning_rate": 0.000472765585027602, "loss": 3.1997, "step": 31600 }, { "epoch": 8.517334194659776, "grad_norm": 0.3494406044483185, "learning_rate": 0.0004725636192271442, "loss": 3.1924, "step": 31650 }, { "epoch": 8.530792420327304, "grad_norm": 0.34491440653800964, "learning_rate": 0.0004723616534266864, "loss": 3.1889, "step": 31700 }, { "epoch": 8.544250645994833, "grad_norm": 0.3356218934059143, "learning_rate": 0.0004721596876262286, "loss": 3.1988, "step": 31750 }, { "epoch": 8.55770887166236, "grad_norm": 0.3772531747817993, "learning_rate": 0.0004719577218257708, "loss": 3.199, "step": 31800 }, { "epoch": 8.571167097329887, "grad_norm": 0.36063918471336365, "learning_rate": 0.000471755756025313, "loss": 3.2036, "step": 31850 }, { "epoch": 8.584625322997416, "grad_norm": 0.37636640667915344, "learning_rate": 0.0004715537902248552, "loss": 3.1986, "step": 31900 }, { "epoch": 8.598083548664944, "grad_norm": 0.35377389192581177, "learning_rate": 0.0004713518244243974, "loss": 3.1894, "step": 31950 }, { "epoch": 8.611541774332473, "grad_norm": 0.35772061347961426, "learning_rate": 0.00047114985862393966, "loss": 3.1894, "step": 32000 }, { "epoch": 8.611541774332473, "eval_accuracy": 0.38834927617291887, "eval_loss": 3.3256490230560303, "eval_runtime": 53.5289, "eval_samples_per_second": 336.491, "eval_steps_per_second": 21.035, "step": 32000 }, { "epoch": 8.625, "grad_norm": 0.3695344924926758, "learning_rate": 0.00047094789282348186, "loss": 3.1975, "step": 32050 }, { "epoch": 8.638458225667527, "grad_norm": 0.4119062125682831, "learning_rate": 0.00047074592702302405, "loss": 3.1931, "step": 32100 }, { "epoch": 8.651916451335056, "grad_norm": 0.35740119218826294, "learning_rate": 0.00047054396122256625, "loss": 3.1945, "step": 32150 }, { "epoch": 8.665374677002584, "grad_norm": 0.3605038821697235, "learning_rate": 0.00047034199542210844, "loss": 3.1909, "step": 32200 }, { "epoch": 8.678832902670113, "grad_norm": 0.3636098802089691, "learning_rate": 0.0004701400296216507, "loss": 3.2043, "step": 32250 }, { "epoch": 8.69229112833764, "grad_norm": 0.3707864582538605, "learning_rate": 0.0004699380638211929, "loss": 3.2117, "step": 32300 }, { "epoch": 8.705749354005167, "grad_norm": 0.34066513180732727, "learning_rate": 0.0004697360980207351, "loss": 3.2029, "step": 32350 }, { "epoch": 8.719207579672696, "grad_norm": 0.34851470589637756, "learning_rate": 0.0004695341322202773, "loss": 3.1863, "step": 32400 }, { "epoch": 8.732665805340224, "grad_norm": 0.35019055008888245, "learning_rate": 0.0004693321664198196, "loss": 3.2009, "step": 32450 }, { "epoch": 8.746124031007753, "grad_norm": 0.35675832629203796, "learning_rate": 0.0004691302006193618, "loss": 3.1981, "step": 32500 }, { "epoch": 8.75958225667528, "grad_norm": 0.34505200386047363, "learning_rate": 0.000468928234818904, "loss": 3.2123, "step": 32550 }, { "epoch": 8.773040482342807, "grad_norm": 0.3546963632106781, "learning_rate": 0.0004687262690184462, "loss": 3.2002, "step": 32600 }, { "epoch": 8.786498708010337, "grad_norm": 0.3934785723686218, "learning_rate": 0.00046852430321798837, "loss": 3.2091, "step": 32650 }, { "epoch": 8.799956933677864, "grad_norm": 0.3397079408168793, "learning_rate": 0.0004683223374175306, "loss": 3.197, "step": 32700 }, { "epoch": 8.813415159345391, "grad_norm": 0.3354127109050751, "learning_rate": 0.0004681203716170728, "loss": 3.2114, "step": 32750 }, { "epoch": 8.82687338501292, "grad_norm": 0.35856249928474426, "learning_rate": 0.000467918405816615, "loss": 3.2018, "step": 32800 }, { "epoch": 8.840331610680447, "grad_norm": 0.34060561656951904, "learning_rate": 0.0004677164400161572, "loss": 3.2029, "step": 32850 }, { "epoch": 8.853789836347977, "grad_norm": 0.3674628734588623, "learning_rate": 0.0004675144742156994, "loss": 3.2096, "step": 32900 }, { "epoch": 8.867248062015504, "grad_norm": 0.3439692258834839, "learning_rate": 0.00046731250841524166, "loss": 3.2028, "step": 32950 }, { "epoch": 8.880706287683031, "grad_norm": 0.3598936200141907, "learning_rate": 0.00046711054261478385, "loss": 3.2085, "step": 33000 }, { "epoch": 8.880706287683031, "eval_accuracy": 0.38856559206234764, "eval_loss": 3.3211824893951416, "eval_runtime": 53.2603, "eval_samples_per_second": 338.188, "eval_steps_per_second": 21.141, "step": 33000 }, { "epoch": 8.89416451335056, "grad_norm": 0.3437979817390442, "learning_rate": 0.00046690857681432605, "loss": 3.1999, "step": 33050 }, { "epoch": 8.907622739018088, "grad_norm": 0.3484727740287781, "learning_rate": 0.00046670661101386825, "loss": 3.1986, "step": 33100 }, { "epoch": 8.921080964685617, "grad_norm": 0.35170236229896545, "learning_rate": 0.00046650464521341044, "loss": 3.2048, "step": 33150 }, { "epoch": 8.934539190353144, "grad_norm": 0.3486970067024231, "learning_rate": 0.0004663026794129527, "loss": 3.2057, "step": 33200 }, { "epoch": 8.947997416020671, "grad_norm": 0.33753907680511475, "learning_rate": 0.0004661007136124949, "loss": 3.21, "step": 33250 }, { "epoch": 8.9614556416882, "grad_norm": 0.3692246675491333, "learning_rate": 0.00046589874781203714, "loss": 3.2023, "step": 33300 }, { "epoch": 8.974913867355728, "grad_norm": 0.3209538757801056, "learning_rate": 0.0004656967820115794, "loss": 3.1987, "step": 33350 }, { "epoch": 8.988372093023255, "grad_norm": 0.39224815368652344, "learning_rate": 0.0004654948162111216, "loss": 3.2082, "step": 33400 }, { "epoch": 9.001614987080103, "grad_norm": 0.35652458667755127, "learning_rate": 0.0004652928504106638, "loss": 3.2015, "step": 33450 }, { "epoch": 9.015073212747632, "grad_norm": 0.3708120584487915, "learning_rate": 0.000465090884610206, "loss": 3.1127, "step": 33500 }, { "epoch": 9.02853143841516, "grad_norm": 0.34585344791412354, "learning_rate": 0.00046488891880974817, "loss": 3.1111, "step": 33550 }, { "epoch": 9.041989664082687, "grad_norm": 0.36985328793525696, "learning_rate": 0.0004646869530092904, "loss": 3.1147, "step": 33600 }, { "epoch": 9.055447889750216, "grad_norm": 0.3506423532962799, "learning_rate": 0.0004644849872088326, "loss": 3.1262, "step": 33650 }, { "epoch": 9.068906115417743, "grad_norm": 0.34678879380226135, "learning_rate": 0.0004642830214083748, "loss": 3.1178, "step": 33700 }, { "epoch": 9.082364341085272, "grad_norm": 0.3602910041809082, "learning_rate": 0.000464081055607917, "loss": 3.126, "step": 33750 }, { "epoch": 9.0958225667528, "grad_norm": 0.3410084545612335, "learning_rate": 0.0004638790898074592, "loss": 3.1358, "step": 33800 }, { "epoch": 9.109280792420327, "grad_norm": 0.36710256338119507, "learning_rate": 0.00046367712400700146, "loss": 3.1225, "step": 33850 }, { "epoch": 9.122739018087856, "grad_norm": 0.3548193573951721, "learning_rate": 0.00046347515820654365, "loss": 3.1226, "step": 33900 }, { "epoch": 9.136197243755383, "grad_norm": 0.3377094268798828, "learning_rate": 0.00046327319240608585, "loss": 3.1227, "step": 33950 }, { "epoch": 9.14965546942291, "grad_norm": 0.34470683336257935, "learning_rate": 0.00046307122660562805, "loss": 3.1266, "step": 34000 }, { "epoch": 9.14965546942291, "eval_accuracy": 0.3885327807119672, "eval_loss": 3.327869415283203, "eval_runtime": 53.7749, "eval_samples_per_second": 334.952, "eval_steps_per_second": 20.939, "step": 34000 }, { "epoch": 9.16311369509044, "grad_norm": 0.360047847032547, "learning_rate": 0.00046286926080517024, "loss": 3.1376, "step": 34050 }, { "epoch": 9.176571920757967, "grad_norm": 0.3523208200931549, "learning_rate": 0.0004626672950047125, "loss": 3.1269, "step": 34100 }, { "epoch": 9.190030146425496, "grad_norm": 0.3531096279621124, "learning_rate": 0.0004624653292042547, "loss": 3.1298, "step": 34150 }, { "epoch": 9.203488372093023, "grad_norm": 0.33584877848625183, "learning_rate": 0.00046226336340379694, "loss": 3.1504, "step": 34200 }, { "epoch": 9.21694659776055, "grad_norm": 0.3406696617603302, "learning_rate": 0.00046206139760333913, "loss": 3.1462, "step": 34250 }, { "epoch": 9.23040482342808, "grad_norm": 0.3635725975036621, "learning_rate": 0.0004618594318028814, "loss": 3.1605, "step": 34300 }, { "epoch": 9.243863049095607, "grad_norm": 0.35520845651626587, "learning_rate": 0.0004616574660024236, "loss": 3.1466, "step": 34350 }, { "epoch": 9.257321274763136, "grad_norm": 0.36953485012054443, "learning_rate": 0.0004614555002019658, "loss": 3.1521, "step": 34400 }, { "epoch": 9.270779500430663, "grad_norm": 0.3408142328262329, "learning_rate": 0.000461253534401508, "loss": 3.1451, "step": 34450 }, { "epoch": 9.28423772609819, "grad_norm": 0.35455116629600525, "learning_rate": 0.00046105156860105017, "loss": 3.1512, "step": 34500 }, { "epoch": 9.29769595176572, "grad_norm": 0.3463025689125061, "learning_rate": 0.0004608496028005924, "loss": 3.1463, "step": 34550 }, { "epoch": 9.311154177433247, "grad_norm": 0.35179778933525085, "learning_rate": 0.0004606476370001346, "loss": 3.1647, "step": 34600 }, { "epoch": 9.324612403100776, "grad_norm": 0.34766823053359985, "learning_rate": 0.0004604456711996768, "loss": 3.154, "step": 34650 }, { "epoch": 9.338070628768303, "grad_norm": 0.3571286201477051, "learning_rate": 0.000460243705399219, "loss": 3.1697, "step": 34700 }, { "epoch": 9.35152885443583, "grad_norm": 0.3302684426307678, "learning_rate": 0.0004600417395987612, "loss": 3.155, "step": 34750 }, { "epoch": 9.36498708010336, "grad_norm": 0.3882899582386017, "learning_rate": 0.00045983977379830345, "loss": 3.155, "step": 34800 }, { "epoch": 9.378445305770887, "grad_norm": 0.3561243414878845, "learning_rate": 0.00045963780799784565, "loss": 3.1591, "step": 34850 }, { "epoch": 9.391903531438416, "grad_norm": 0.3711599111557007, "learning_rate": 0.00045943584219738785, "loss": 3.1584, "step": 34900 }, { "epoch": 9.405361757105943, "grad_norm": 0.4065340459346771, "learning_rate": 0.00045923387639693004, "loss": 3.1684, "step": 34950 }, { "epoch": 9.41881998277347, "grad_norm": 0.3408617675304413, "learning_rate": 0.00045903191059647224, "loss": 3.1647, "step": 35000 }, { "epoch": 9.41881998277347, "eval_accuracy": 0.388881537118329, "eval_loss": 3.3228235244750977, "eval_runtime": 53.5953, "eval_samples_per_second": 336.074, "eval_steps_per_second": 21.009, "step": 35000 }, { "epoch": 9.432278208441, "grad_norm": 0.39076054096221924, "learning_rate": 0.0004588299447960145, "loss": 3.1743, "step": 35050 }, { "epoch": 9.445736434108527, "grad_norm": 0.3773688077926636, "learning_rate": 0.00045862797899555674, "loss": 3.1596, "step": 35100 }, { "epoch": 9.459194659776054, "grad_norm": 0.33396244049072266, "learning_rate": 0.00045842601319509894, "loss": 3.1774, "step": 35150 }, { "epoch": 9.472652885443583, "grad_norm": 0.33974024653434753, "learning_rate": 0.00045822404739464113, "loss": 3.1719, "step": 35200 }, { "epoch": 9.48611111111111, "grad_norm": 0.3999760150909424, "learning_rate": 0.0004580220815941834, "loss": 3.1653, "step": 35250 }, { "epoch": 9.49956933677864, "grad_norm": 0.3510672450065613, "learning_rate": 0.0004578201157937256, "loss": 3.1631, "step": 35300 }, { "epoch": 9.513027562446167, "grad_norm": 0.343685507774353, "learning_rate": 0.0004576181499932678, "loss": 3.1687, "step": 35350 }, { "epoch": 9.526485788113694, "grad_norm": 0.34639087319374084, "learning_rate": 0.00045741618419280997, "loss": 3.1564, "step": 35400 }, { "epoch": 9.539944013781223, "grad_norm": 0.37675940990448, "learning_rate": 0.0004572142183923522, "loss": 3.1699, "step": 35450 }, { "epoch": 9.55340223944875, "grad_norm": 0.35573285818099976, "learning_rate": 0.0004570122525918944, "loss": 3.1742, "step": 35500 }, { "epoch": 9.56686046511628, "grad_norm": 0.3698192834854126, "learning_rate": 0.0004568102867914366, "loss": 3.1801, "step": 35550 }, { "epoch": 9.580318690783807, "grad_norm": 0.3495495319366455, "learning_rate": 0.0004566083209909788, "loss": 3.1716, "step": 35600 }, { "epoch": 9.593776916451334, "grad_norm": 0.3501206040382385, "learning_rate": 0.000456406355190521, "loss": 3.1683, "step": 35650 }, { "epoch": 9.607235142118864, "grad_norm": 0.38374894857406616, "learning_rate": 0.00045620438939006326, "loss": 3.179, "step": 35700 }, { "epoch": 9.62069336778639, "grad_norm": 0.34734368324279785, "learning_rate": 0.00045600242358960545, "loss": 3.1827, "step": 35750 }, { "epoch": 9.634151593453918, "grad_norm": 0.33780547976493835, "learning_rate": 0.00045580045778914765, "loss": 3.1679, "step": 35800 }, { "epoch": 9.647609819121447, "grad_norm": 0.35860759019851685, "learning_rate": 0.00045559849198868984, "loss": 3.1798, "step": 35850 }, { "epoch": 9.661068044788975, "grad_norm": 0.3601613938808441, "learning_rate": 0.00045539652618823204, "loss": 3.1827, "step": 35900 }, { "epoch": 9.674526270456504, "grad_norm": 0.35015615820884705, "learning_rate": 0.0004551945603877743, "loss": 3.1855, "step": 35950 }, { "epoch": 9.687984496124031, "grad_norm": 0.3426166772842407, "learning_rate": 0.00045499259458731654, "loss": 3.1814, "step": 36000 }, { "epoch": 9.687984496124031, "eval_accuracy": 0.38963543764909064, "eval_loss": 3.3155407905578613, "eval_runtime": 53.4256, "eval_samples_per_second": 337.142, "eval_steps_per_second": 21.076, "step": 36000 }, { "epoch": 9.701442721791558, "grad_norm": 0.3451278805732727, "learning_rate": 0.00045479062878685874, "loss": 3.188, "step": 36050 }, { "epoch": 9.714900947459087, "grad_norm": 0.368107408285141, "learning_rate": 0.00045458866298640093, "loss": 3.1815, "step": 36100 }, { "epoch": 9.728359173126615, "grad_norm": 0.3657575249671936, "learning_rate": 0.0004543866971859432, "loss": 3.1868, "step": 36150 }, { "epoch": 9.741817398794144, "grad_norm": 0.34643518924713135, "learning_rate": 0.0004541847313854854, "loss": 3.1751, "step": 36200 }, { "epoch": 9.755275624461671, "grad_norm": 0.36216220259666443, "learning_rate": 0.0004539827655850276, "loss": 3.1826, "step": 36250 }, { "epoch": 9.768733850129198, "grad_norm": 0.3650382161140442, "learning_rate": 0.00045378079978456977, "loss": 3.1718, "step": 36300 }, { "epoch": 9.782192075796727, "grad_norm": 0.348186194896698, "learning_rate": 0.00045357883398411197, "loss": 3.1654, "step": 36350 }, { "epoch": 9.795650301464255, "grad_norm": 0.3629589378833771, "learning_rate": 0.0004533768681836542, "loss": 3.1869, "step": 36400 }, { "epoch": 9.809108527131784, "grad_norm": 0.3417031764984131, "learning_rate": 0.0004531749023831964, "loss": 3.1829, "step": 36450 }, { "epoch": 9.822566752799311, "grad_norm": 0.34222733974456787, "learning_rate": 0.0004529729365827386, "loss": 3.191, "step": 36500 }, { "epoch": 9.836024978466838, "grad_norm": 0.341671884059906, "learning_rate": 0.0004527709707822808, "loss": 3.189, "step": 36550 }, { "epoch": 9.849483204134367, "grad_norm": 0.3711172938346863, "learning_rate": 0.000452569004981823, "loss": 3.1764, "step": 36600 }, { "epoch": 9.862941429801895, "grad_norm": 0.3440135717391968, "learning_rate": 0.00045236703918136525, "loss": 3.1641, "step": 36650 }, { "epoch": 9.876399655469424, "grad_norm": 0.3357332646846771, "learning_rate": 0.00045216507338090745, "loss": 3.177, "step": 36700 }, { "epoch": 9.889857881136951, "grad_norm": 0.3803945779800415, "learning_rate": 0.00045196310758044964, "loss": 3.1844, "step": 36750 }, { "epoch": 9.903316106804478, "grad_norm": 0.3620223104953766, "learning_rate": 0.00045176114177999184, "loss": 3.1773, "step": 36800 }, { "epoch": 9.916774332472007, "grad_norm": 0.3589492738246918, "learning_rate": 0.00045155917597953404, "loss": 3.1858, "step": 36850 }, { "epoch": 9.930232558139535, "grad_norm": 0.3606046140193939, "learning_rate": 0.00045135721017907634, "loss": 3.1901, "step": 36900 }, { "epoch": 9.943690783807062, "grad_norm": 0.3455662429332733, "learning_rate": 0.00045115524437861854, "loss": 3.1913, "step": 36950 }, { "epoch": 9.957149009474591, "grad_norm": 0.33837249875068665, "learning_rate": 0.00045095327857816073, "loss": 3.1885, "step": 37000 }, { "epoch": 9.957149009474591, "eval_accuracy": 0.3903870565958854, "eval_loss": 3.3125293254852295, "eval_runtime": 53.6286, "eval_samples_per_second": 335.866, "eval_steps_per_second": 20.996, "step": 37000 }, { "epoch": 9.970607235142118, "grad_norm": 0.37519270181655884, "learning_rate": 0.00045075131277770293, "loss": 3.1824, "step": 37050 }, { "epoch": 9.984065460809648, "grad_norm": 0.3906201124191284, "learning_rate": 0.0004505493469772452, "loss": 3.1738, "step": 37100 }, { "epoch": 9.997523686477175, "grad_norm": 0.36060965061187744, "learning_rate": 0.0004503473811767874, "loss": 3.1862, "step": 37150 }, { "epoch": 10.010766580534023, "grad_norm": 0.35172829031944275, "learning_rate": 0.00045014541537632957, "loss": 3.1091, "step": 37200 }, { "epoch": 10.02422480620155, "grad_norm": 0.38957053422927856, "learning_rate": 0.00044994344957587177, "loss": 3.0934, "step": 37250 }, { "epoch": 10.037683031869078, "grad_norm": 0.3552197515964508, "learning_rate": 0.00044974148377541396, "loss": 3.0892, "step": 37300 }, { "epoch": 10.051141257536607, "grad_norm": 0.39891138672828674, "learning_rate": 0.0004495395179749562, "loss": 3.0886, "step": 37350 }, { "epoch": 10.064599483204134, "grad_norm": 0.4051150977611542, "learning_rate": 0.0004493375521744984, "loss": 3.1042, "step": 37400 }, { "epoch": 10.078057708871663, "grad_norm": 0.3359229266643524, "learning_rate": 0.0004491355863740406, "loss": 3.1038, "step": 37450 }, { "epoch": 10.09151593453919, "grad_norm": 0.35439833998680115, "learning_rate": 0.0004489336205735828, "loss": 3.1125, "step": 37500 }, { "epoch": 10.104974160206718, "grad_norm": 0.34396088123321533, "learning_rate": 0.00044873165477312505, "loss": 3.108, "step": 37550 }, { "epoch": 10.118432385874247, "grad_norm": 0.3561412990093231, "learning_rate": 0.00044852968897266725, "loss": 3.1052, "step": 37600 }, { "epoch": 10.131890611541774, "grad_norm": 0.3590238690376282, "learning_rate": 0.00044832772317220945, "loss": 3.1127, "step": 37650 }, { "epoch": 10.145348837209303, "grad_norm": 0.3750346302986145, "learning_rate": 0.00044812575737175164, "loss": 3.1132, "step": 37700 }, { "epoch": 10.15880706287683, "grad_norm": 0.3595349192619324, "learning_rate": 0.00044792379157129384, "loss": 3.1089, "step": 37750 }, { "epoch": 10.172265288544358, "grad_norm": 0.3501310646533966, "learning_rate": 0.00044772182577083614, "loss": 3.1126, "step": 37800 }, { "epoch": 10.185723514211887, "grad_norm": 0.3545890152454376, "learning_rate": 0.00044751985997037834, "loss": 3.1221, "step": 37850 }, { "epoch": 10.199181739879414, "grad_norm": 0.3811264932155609, "learning_rate": 0.00044731789416992053, "loss": 3.1247, "step": 37900 }, { "epoch": 10.212639965546943, "grad_norm": 0.3635217845439911, "learning_rate": 0.00044711592836946273, "loss": 3.1175, "step": 37950 }, { "epoch": 10.22609819121447, "grad_norm": 0.36193686723709106, "learning_rate": 0.000446913962569005, "loss": 3.1233, "step": 38000 }, { "epoch": 10.22609819121447, "eval_accuracy": 0.3899182454141249, "eval_loss": 3.3228235244750977, "eval_runtime": 53.3214, "eval_samples_per_second": 337.8, "eval_steps_per_second": 21.117, "step": 38000 }, { "epoch": 10.239556416881998, "grad_norm": 0.3329092264175415, "learning_rate": 0.0004467119967685472, "loss": 3.1267, "step": 38050 }, { "epoch": 10.253014642549527, "grad_norm": 0.3588671386241913, "learning_rate": 0.00044651003096808937, "loss": 3.1297, "step": 38100 }, { "epoch": 10.266472868217054, "grad_norm": 0.36286869645118713, "learning_rate": 0.00044630806516763157, "loss": 3.1288, "step": 38150 }, { "epoch": 10.279931093884581, "grad_norm": 0.36928683519363403, "learning_rate": 0.00044610609936717376, "loss": 3.1187, "step": 38200 }, { "epoch": 10.29338931955211, "grad_norm": 0.34665945172309875, "learning_rate": 0.000445904133566716, "loss": 3.1299, "step": 38250 }, { "epoch": 10.306847545219638, "grad_norm": 0.4086586534976959, "learning_rate": 0.0004457021677662582, "loss": 3.1397, "step": 38300 }, { "epoch": 10.320305770887167, "grad_norm": 0.3424471318721771, "learning_rate": 0.0004455002019658004, "loss": 3.1356, "step": 38350 }, { "epoch": 10.333763996554694, "grad_norm": 0.37157419323921204, "learning_rate": 0.0004452982361653426, "loss": 3.1401, "step": 38400 }, { "epoch": 10.347222222222221, "grad_norm": 0.37454769015312195, "learning_rate": 0.0004450962703648848, "loss": 3.1251, "step": 38450 }, { "epoch": 10.36068044788975, "grad_norm": 0.3706095218658447, "learning_rate": 0.00044489430456442705, "loss": 3.1319, "step": 38500 }, { "epoch": 10.374138673557278, "grad_norm": 0.33369627594947815, "learning_rate": 0.00044469233876396925, "loss": 3.1416, "step": 38550 }, { "epoch": 10.387596899224807, "grad_norm": 0.3583381175994873, "learning_rate": 0.00044449037296351144, "loss": 3.1457, "step": 38600 }, { "epoch": 10.401055124892334, "grad_norm": 0.39620569348335266, "learning_rate": 0.0004442884071630537, "loss": 3.147, "step": 38650 }, { "epoch": 10.414513350559861, "grad_norm": 0.38012731075286865, "learning_rate": 0.00044408644136259594, "loss": 3.1386, "step": 38700 }, { "epoch": 10.42797157622739, "grad_norm": 0.3489794135093689, "learning_rate": 0.00044388447556213814, "loss": 3.1426, "step": 38750 }, { "epoch": 10.441429801894918, "grad_norm": 0.37140825390815735, "learning_rate": 0.00044368250976168033, "loss": 3.14, "step": 38800 }, { "epoch": 10.454888027562447, "grad_norm": 0.3544456362724304, "learning_rate": 0.00044348054396122253, "loss": 3.1421, "step": 38850 }, { "epoch": 10.468346253229974, "grad_norm": 0.3525961935520172, "learning_rate": 0.0004432785781607647, "loss": 3.1414, "step": 38900 }, { "epoch": 10.481804478897502, "grad_norm": 0.3718273937702179, "learning_rate": 0.000443076612360307, "loss": 3.1555, "step": 38950 }, { "epoch": 10.49526270456503, "grad_norm": 0.357573926448822, "learning_rate": 0.0004428746465598492, "loss": 3.1452, "step": 39000 }, { "epoch": 10.49526270456503, "eval_accuracy": 0.39027265145697604, "eval_loss": 3.31387996673584, "eval_runtime": 53.4319, "eval_samples_per_second": 337.102, "eval_steps_per_second": 21.074, "step": 39000 }, { "epoch": 10.508720930232558, "grad_norm": 0.3596917688846588, "learning_rate": 0.00044267268075939137, "loss": 3.1411, "step": 39050 }, { "epoch": 10.522179155900087, "grad_norm": 0.3924131989479065, "learning_rate": 0.00044247071495893357, "loss": 3.1423, "step": 39100 }, { "epoch": 10.535637381567614, "grad_norm": 0.3612106442451477, "learning_rate": 0.00044226874915847576, "loss": 3.1432, "step": 39150 }, { "epoch": 10.549095607235142, "grad_norm": 0.33442962169647217, "learning_rate": 0.000442066783358018, "loss": 3.1483, "step": 39200 }, { "epoch": 10.56255383290267, "grad_norm": 0.3581370413303375, "learning_rate": 0.0004418648175575602, "loss": 3.1499, "step": 39250 }, { "epoch": 10.576012058570198, "grad_norm": 0.3770648241043091, "learning_rate": 0.0004416628517571024, "loss": 3.1449, "step": 39300 }, { "epoch": 10.589470284237725, "grad_norm": 0.3834937512874603, "learning_rate": 0.0004414608859566446, "loss": 3.1493, "step": 39350 }, { "epoch": 10.602928509905254, "grad_norm": 0.3559904098510742, "learning_rate": 0.00044125892015618685, "loss": 3.1563, "step": 39400 }, { "epoch": 10.616386735572782, "grad_norm": 0.3795510232448578, "learning_rate": 0.00044105695435572905, "loss": 3.1429, "step": 39450 }, { "epoch": 10.62984496124031, "grad_norm": 0.3509112000465393, "learning_rate": 0.00044085498855527124, "loss": 3.155, "step": 39500 }, { "epoch": 10.643303186907838, "grad_norm": 0.382457971572876, "learning_rate": 0.0004406530227548135, "loss": 3.1448, "step": 39550 }, { "epoch": 10.656761412575365, "grad_norm": 0.3792799413204193, "learning_rate": 0.00044045105695435574, "loss": 3.1601, "step": 39600 }, { "epoch": 10.670219638242894, "grad_norm": 0.3273200988769531, "learning_rate": 0.00044024909115389794, "loss": 3.1474, "step": 39650 }, { "epoch": 10.683677863910422, "grad_norm": 0.3698997497558594, "learning_rate": 0.00044004712535344014, "loss": 3.159, "step": 39700 }, { "epoch": 10.69713608957795, "grad_norm": 0.3488556742668152, "learning_rate": 0.00043984515955298233, "loss": 3.1553, "step": 39750 }, { "epoch": 10.710594315245478, "grad_norm": 0.350911408662796, "learning_rate": 0.00043964319375252453, "loss": 3.1567, "step": 39800 }, { "epoch": 10.724052540913005, "grad_norm": 0.337378591299057, "learning_rate": 0.0004394412279520668, "loss": 3.1561, "step": 39850 }, { "epoch": 10.737510766580534, "grad_norm": 0.3427545726299286, "learning_rate": 0.000439239262151609, "loss": 3.1715, "step": 39900 }, { "epoch": 10.750968992248062, "grad_norm": 0.36676934361457825, "learning_rate": 0.00043903729635115117, "loss": 3.1629, "step": 39950 }, { "epoch": 10.764427217915589, "grad_norm": 0.38957640528678894, "learning_rate": 0.00043883533055069337, "loss": 3.1695, "step": 40000 }, { "epoch": 10.764427217915589, "eval_accuracy": 0.39051004483638435, "eval_loss": 3.3090879917144775, "eval_runtime": 53.3592, "eval_samples_per_second": 337.561, "eval_steps_per_second": 21.102, "step": 40000 }, { "epoch": 10.777885443583118, "grad_norm": 0.3423351049423218, "learning_rate": 0.00043863336475023556, "loss": 3.1569, "step": 40050 }, { "epoch": 10.791343669250645, "grad_norm": 0.3677137494087219, "learning_rate": 0.0004384313989497778, "loss": 3.1637, "step": 40100 }, { "epoch": 10.804801894918175, "grad_norm": 0.4070259928703308, "learning_rate": 0.00043822943314932, "loss": 3.1542, "step": 40150 }, { "epoch": 10.818260120585702, "grad_norm": 0.35667550563812256, "learning_rate": 0.0004380274673488622, "loss": 3.1666, "step": 40200 }, { "epoch": 10.83171834625323, "grad_norm": 0.3461889624595642, "learning_rate": 0.0004378255015484044, "loss": 3.1642, "step": 40250 }, { "epoch": 10.845176571920758, "grad_norm": 0.3470584750175476, "learning_rate": 0.0004376235357479466, "loss": 3.1641, "step": 40300 }, { "epoch": 10.858634797588286, "grad_norm": 0.3516797423362732, "learning_rate": 0.00043742156994748885, "loss": 3.1651, "step": 40350 }, { "epoch": 10.872093023255815, "grad_norm": 0.3457486927509308, "learning_rate": 0.00043721960414703104, "loss": 3.1637, "step": 40400 }, { "epoch": 10.885551248923342, "grad_norm": 0.3525344431400299, "learning_rate": 0.0004370176383465733, "loss": 3.1693, "step": 40450 }, { "epoch": 10.89900947459087, "grad_norm": 0.3566647469997406, "learning_rate": 0.0004368156725461155, "loss": 3.1562, "step": 40500 }, { "epoch": 10.912467700258398, "grad_norm": 0.39564409852027893, "learning_rate": 0.00043661370674565774, "loss": 3.163, "step": 40550 }, { "epoch": 10.925925925925926, "grad_norm": 0.35635221004486084, "learning_rate": 0.00043641174094519994, "loss": 3.1663, "step": 40600 }, { "epoch": 10.939384151593455, "grad_norm": 0.35757219791412354, "learning_rate": 0.00043620977514474213, "loss": 3.1705, "step": 40650 }, { "epoch": 10.952842377260982, "grad_norm": 0.3532245457172394, "learning_rate": 0.00043600780934428433, "loss": 3.167, "step": 40700 }, { "epoch": 10.96630060292851, "grad_norm": 0.3705931007862091, "learning_rate": 0.0004358058435438265, "loss": 3.1693, "step": 40750 }, { "epoch": 10.979758828596038, "grad_norm": 0.37144991755485535, "learning_rate": 0.0004356038777433688, "loss": 3.1758, "step": 40800 }, { "epoch": 10.993217054263566, "grad_norm": 0.3508891761302948, "learning_rate": 0.00043540191194291097, "loss": 3.1677, "step": 40850 }, { "epoch": 11.006459948320414, "grad_norm": 0.37817490100860596, "learning_rate": 0.00043519994614245317, "loss": 3.1246, "step": 40900 }, { "epoch": 11.019918173987941, "grad_norm": 0.3762407898902893, "learning_rate": 0.00043499798034199536, "loss": 3.0683, "step": 40950 }, { "epoch": 11.03337639965547, "grad_norm": 0.3525347411632538, "learning_rate": 0.00043479601454153756, "loss": 3.0734, "step": 41000 }, { "epoch": 11.03337639965547, "eval_accuracy": 0.3907509149151707, "eval_loss": 3.3142662048339844, "eval_runtime": 53.355, "eval_samples_per_second": 337.588, "eval_steps_per_second": 21.104, "step": 41000 }, { "epoch": 11.046834625322997, "grad_norm": 0.36088988184928894, "learning_rate": 0.0004345940487410798, "loss": 3.0784, "step": 41050 }, { "epoch": 11.060292850990525, "grad_norm": 0.3609704077243805, "learning_rate": 0.000434392082940622, "loss": 3.0779, "step": 41100 }, { "epoch": 11.073751076658054, "grad_norm": 0.35831403732299805, "learning_rate": 0.0004341901171401642, "loss": 3.0786, "step": 41150 }, { "epoch": 11.087209302325581, "grad_norm": 0.3548460304737091, "learning_rate": 0.0004339881513397064, "loss": 3.0868, "step": 41200 }, { "epoch": 11.10066752799311, "grad_norm": 0.3861493766307831, "learning_rate": 0.0004337861855392486, "loss": 3.0904, "step": 41250 }, { "epoch": 11.114125753660637, "grad_norm": 0.4032128155231476, "learning_rate": 0.00043358421973879084, "loss": 3.0839, "step": 41300 }, { "epoch": 11.127583979328165, "grad_norm": 0.364173024892807, "learning_rate": 0.0004333822539383331, "loss": 3.0909, "step": 41350 }, { "epoch": 11.141042204995694, "grad_norm": 0.35015109181404114, "learning_rate": 0.0004331802881378753, "loss": 3.0923, "step": 41400 }, { "epoch": 11.154500430663221, "grad_norm": 0.3518073558807373, "learning_rate": 0.00043297832233741754, "loss": 3.0858, "step": 41450 }, { "epoch": 11.16795865633075, "grad_norm": 0.3917633891105652, "learning_rate": 0.00043277635653695974, "loss": 3.1072, "step": 41500 }, { "epoch": 11.181416881998278, "grad_norm": 0.3793010711669922, "learning_rate": 0.00043257439073650193, "loss": 3.0988, "step": 41550 }, { "epoch": 11.194875107665805, "grad_norm": 0.36705297231674194, "learning_rate": 0.00043237242493604413, "loss": 3.0967, "step": 41600 }, { "epoch": 11.208333333333334, "grad_norm": 0.35543742775917053, "learning_rate": 0.0004321704591355863, "loss": 3.095, "step": 41650 }, { "epoch": 11.221791559000861, "grad_norm": 0.3734685182571411, "learning_rate": 0.0004319684933351286, "loss": 3.1022, "step": 41700 }, { "epoch": 11.235249784668389, "grad_norm": 0.35706183314323425, "learning_rate": 0.00043176652753467077, "loss": 3.1099, "step": 41750 }, { "epoch": 11.248708010335918, "grad_norm": 0.4005352556705475, "learning_rate": 0.00043156456173421297, "loss": 3.1042, "step": 41800 }, { "epoch": 11.262166236003445, "grad_norm": 0.3816969692707062, "learning_rate": 0.00043136259593375516, "loss": 3.1062, "step": 41850 }, { "epoch": 11.275624461670974, "grad_norm": 0.3443428575992584, "learning_rate": 0.00043116063013329736, "loss": 3.1235, "step": 41900 }, { "epoch": 11.289082687338501, "grad_norm": 0.3736826479434967, "learning_rate": 0.0004309586643328396, "loss": 3.1089, "step": 41950 }, { "epoch": 11.302540913006029, "grad_norm": 0.355075865983963, "learning_rate": 0.0004307566985323818, "loss": 3.1179, "step": 42000 }, { "epoch": 11.302540913006029, "eval_accuracy": 0.39087075239685826, "eval_loss": 3.314624786376953, "eval_runtime": 53.3716, "eval_samples_per_second": 337.483, "eval_steps_per_second": 21.097, "step": 42000 }, { "epoch": 11.315999138673558, "grad_norm": 0.3512454926967621, "learning_rate": 0.000430554732731924, "loss": 3.1155, "step": 42050 }, { "epoch": 11.329457364341085, "grad_norm": 0.3447372019290924, "learning_rate": 0.0004303527669314662, "loss": 3.1195, "step": 42100 }, { "epoch": 11.342915590008614, "grad_norm": 0.3592742383480072, "learning_rate": 0.0004301508011310084, "loss": 3.1171, "step": 42150 }, { "epoch": 11.356373815676141, "grad_norm": 0.39683109521865845, "learning_rate": 0.00042994883533055065, "loss": 3.1231, "step": 42200 }, { "epoch": 11.369832041343669, "grad_norm": 0.37733298540115356, "learning_rate": 0.0004297468695300929, "loss": 3.1128, "step": 42250 }, { "epoch": 11.383290267011198, "grad_norm": 0.3726079761981964, "learning_rate": 0.0004295449037296351, "loss": 3.1142, "step": 42300 }, { "epoch": 11.396748492678725, "grad_norm": 0.3837583363056183, "learning_rate": 0.0004293429379291773, "loss": 3.1118, "step": 42350 }, { "epoch": 11.410206718346252, "grad_norm": 0.38088589906692505, "learning_rate": 0.00042914097212871954, "loss": 3.1257, "step": 42400 }, { "epoch": 11.423664944013781, "grad_norm": 0.338095098733902, "learning_rate": 0.00042893900632826173, "loss": 3.1313, "step": 42450 }, { "epoch": 11.437123169681309, "grad_norm": 0.3455497622489929, "learning_rate": 0.00042873704052780393, "loss": 3.1252, "step": 42500 }, { "epoch": 11.450581395348838, "grad_norm": 0.38659998774528503, "learning_rate": 0.0004285350747273461, "loss": 3.1153, "step": 42550 }, { "epoch": 11.464039621016365, "grad_norm": 0.34629592299461365, "learning_rate": 0.0004283331089268883, "loss": 3.133, "step": 42600 }, { "epoch": 11.477497846683892, "grad_norm": 0.4040473997592926, "learning_rate": 0.00042813114312643057, "loss": 3.1331, "step": 42650 }, { "epoch": 11.490956072351421, "grad_norm": 0.3708750307559967, "learning_rate": 0.00042792917732597277, "loss": 3.1163, "step": 42700 }, { "epoch": 11.504414298018949, "grad_norm": 0.3524065911769867, "learning_rate": 0.00042772721152551496, "loss": 3.1219, "step": 42750 }, { "epoch": 11.517872523686478, "grad_norm": 0.3564377427101135, "learning_rate": 0.00042752524572505716, "loss": 3.1337, "step": 42800 }, { "epoch": 11.531330749354005, "grad_norm": 0.3928210437297821, "learning_rate": 0.00042732327992459936, "loss": 3.1308, "step": 42850 }, { "epoch": 11.544788975021532, "grad_norm": 0.3760032653808594, "learning_rate": 0.0004271213141241416, "loss": 3.1333, "step": 42900 }, { "epoch": 11.558247200689062, "grad_norm": 0.3637098968029022, "learning_rate": 0.0004269193483236838, "loss": 3.1287, "step": 42950 }, { "epoch": 11.571705426356589, "grad_norm": 0.3505527377128601, "learning_rate": 0.000426717382523226, "loss": 3.1317, "step": 43000 }, { "epoch": 11.571705426356589, "eval_accuracy": 0.39133652146666303, "eval_loss": 3.3082401752471924, "eval_runtime": 53.3948, "eval_samples_per_second": 337.336, "eval_steps_per_second": 21.088, "step": 43000 }, { "epoch": 11.585163652024118, "grad_norm": 0.36543312668800354, "learning_rate": 0.0004265154167227682, "loss": 3.1412, "step": 43050 }, { "epoch": 11.598621877691645, "grad_norm": 0.3579201102256775, "learning_rate": 0.0004263134509223105, "loss": 3.1364, "step": 43100 }, { "epoch": 11.612080103359173, "grad_norm": 0.385629802942276, "learning_rate": 0.0004261114851218527, "loss": 3.1325, "step": 43150 }, { "epoch": 11.625538329026702, "grad_norm": 0.3662668764591217, "learning_rate": 0.0004259095193213949, "loss": 3.1262, "step": 43200 }, { "epoch": 11.638996554694229, "grad_norm": 0.38772115111351013, "learning_rate": 0.0004257075535209371, "loss": 3.132, "step": 43250 }, { "epoch": 11.652454780361758, "grad_norm": 0.356647789478302, "learning_rate": 0.00042550558772047934, "loss": 3.137, "step": 43300 }, { "epoch": 11.665913006029285, "grad_norm": 0.35813525319099426, "learning_rate": 0.00042530362192002153, "loss": 3.1488, "step": 43350 }, { "epoch": 11.679371231696813, "grad_norm": 0.3788353204727173, "learning_rate": 0.00042510165611956373, "loss": 3.1322, "step": 43400 }, { "epoch": 11.692829457364342, "grad_norm": 0.33950915932655334, "learning_rate": 0.0004248996903191059, "loss": 3.1315, "step": 43450 }, { "epoch": 11.706287683031869, "grad_norm": 0.3577655255794525, "learning_rate": 0.0004246977245186481, "loss": 3.1414, "step": 43500 }, { "epoch": 11.719745908699396, "grad_norm": 0.36303114891052246, "learning_rate": 0.0004244957587181904, "loss": 3.1379, "step": 43550 }, { "epoch": 11.733204134366925, "grad_norm": 0.3299844563007355, "learning_rate": 0.00042429379291773257, "loss": 3.1317, "step": 43600 }, { "epoch": 11.746662360034453, "grad_norm": 0.36776620149612427, "learning_rate": 0.00042409182711727477, "loss": 3.1367, "step": 43650 }, { "epoch": 11.760120585701982, "grad_norm": 0.36119797825813293, "learning_rate": 0.00042388986131681696, "loss": 3.1506, "step": 43700 }, { "epoch": 11.773578811369509, "grad_norm": 0.3472893536090851, "learning_rate": 0.00042368789551635916, "loss": 3.1361, "step": 43750 }, { "epoch": 11.787037037037036, "grad_norm": 0.3448840081691742, "learning_rate": 0.0004234859297159014, "loss": 3.1447, "step": 43800 }, { "epoch": 11.800495262704565, "grad_norm": 0.3472510278224945, "learning_rate": 0.0004232839639154436, "loss": 3.1501, "step": 43850 }, { "epoch": 11.813953488372093, "grad_norm": 0.3915238678455353, "learning_rate": 0.0004230819981149858, "loss": 3.1552, "step": 43900 }, { "epoch": 11.827411714039622, "grad_norm": 0.3639262914657593, "learning_rate": 0.000422880032314528, "loss": 3.1447, "step": 43950 }, { "epoch": 11.840869939707149, "grad_norm": 0.3748035430908203, "learning_rate": 0.0004226780665140703, "loss": 3.1587, "step": 44000 }, { "epoch": 11.840869939707149, "eval_accuracy": 0.39165963721511166, "eval_loss": 3.3021323680877686, "eval_runtime": 53.4241, "eval_samples_per_second": 337.152, "eval_steps_per_second": 21.077, "step": 44000 }, { "epoch": 11.854328165374676, "grad_norm": 0.36416128277778625, "learning_rate": 0.0004224761007136125, "loss": 3.138, "step": 44050 }, { "epoch": 11.867786391042205, "grad_norm": 0.3731189966201782, "learning_rate": 0.0004222741349131547, "loss": 3.142, "step": 44100 }, { "epoch": 11.881244616709733, "grad_norm": 0.35165873169898987, "learning_rate": 0.0004220721691126969, "loss": 3.1496, "step": 44150 }, { "epoch": 11.89470284237726, "grad_norm": 0.36347705125808716, "learning_rate": 0.0004218702033122391, "loss": 3.1334, "step": 44200 }, { "epoch": 11.90816106804479, "grad_norm": 0.3625389337539673, "learning_rate": 0.00042166823751178134, "loss": 3.1409, "step": 44250 }, { "epoch": 11.921619293712316, "grad_norm": 0.3359973132610321, "learning_rate": 0.00042146627171132353, "loss": 3.1419, "step": 44300 }, { "epoch": 11.935077519379846, "grad_norm": 0.3568677604198456, "learning_rate": 0.00042126430591086573, "loss": 3.1484, "step": 44350 }, { "epoch": 11.948535745047373, "grad_norm": 0.37474843859672546, "learning_rate": 0.0004210623401104079, "loss": 3.1633, "step": 44400 }, { "epoch": 11.9619939707149, "grad_norm": 0.37014254927635193, "learning_rate": 0.0004208603743099501, "loss": 3.1547, "step": 44450 }, { "epoch": 11.97545219638243, "grad_norm": 0.3610021471977234, "learning_rate": 0.00042065840850949237, "loss": 3.1397, "step": 44500 }, { "epoch": 11.988910422049956, "grad_norm": 0.36313939094543457, "learning_rate": 0.00042045644270903457, "loss": 3.1523, "step": 44550 }, { "epoch": 12.002153316106805, "grad_norm": 0.3740599751472473, "learning_rate": 0.00042025447690857676, "loss": 3.1279, "step": 44600 }, { "epoch": 12.015611541774332, "grad_norm": 0.3721098303794861, "learning_rate": 0.00042005251110811896, "loss": 3.0501, "step": 44650 }, { "epoch": 12.029069767441861, "grad_norm": 0.37249475717544556, "learning_rate": 0.00041985054530766115, "loss": 3.0593, "step": 44700 }, { "epoch": 12.042527993109388, "grad_norm": 0.37305739521980286, "learning_rate": 0.0004196485795072034, "loss": 3.0575, "step": 44750 }, { "epoch": 12.055986218776916, "grad_norm": 0.3568457365036011, "learning_rate": 0.0004194466137067456, "loss": 3.0538, "step": 44800 }, { "epoch": 12.069444444444445, "grad_norm": 0.3669044077396393, "learning_rate": 0.0004192446479062878, "loss": 3.0684, "step": 44850 }, { "epoch": 12.082902670111972, "grad_norm": 0.37108471989631653, "learning_rate": 0.00041904268210583005, "loss": 3.068, "step": 44900 }, { "epoch": 12.096360895779501, "grad_norm": 0.37075528502464294, "learning_rate": 0.0004188407163053723, "loss": 3.0668, "step": 44950 }, { "epoch": 12.109819121447028, "grad_norm": 0.3636740446090698, "learning_rate": 0.0004186387505049145, "loss": 3.0608, "step": 45000 }, { "epoch": 12.109819121447028, "eval_accuracy": 0.39137433057239945, "eval_loss": 3.3102362155914307, "eval_runtime": 53.3835, "eval_samples_per_second": 337.408, "eval_steps_per_second": 21.093, "step": 45000 }, { "epoch": 12.123277347114556, "grad_norm": 0.3946690857410431, "learning_rate": 0.0004184367847044567, "loss": 3.069, "step": 45050 }, { "epoch": 12.136735572782085, "grad_norm": 0.4067884683609009, "learning_rate": 0.0004182348189039989, "loss": 3.0741, "step": 45100 }, { "epoch": 12.150193798449612, "grad_norm": 0.3845987021923065, "learning_rate": 0.00041803285310354114, "loss": 3.0809, "step": 45150 }, { "epoch": 12.163652024117141, "grad_norm": 0.3612961769104004, "learning_rate": 0.00041783088730308333, "loss": 3.0856, "step": 45200 }, { "epoch": 12.177110249784668, "grad_norm": 0.32780522108078003, "learning_rate": 0.00041762892150262553, "loss": 3.0784, "step": 45250 }, { "epoch": 12.190568475452196, "grad_norm": 0.3584006130695343, "learning_rate": 0.0004174269557021677, "loss": 3.0792, "step": 45300 }, { "epoch": 12.204026701119725, "grad_norm": 0.3582470715045929, "learning_rate": 0.0004172249899017099, "loss": 3.0836, "step": 45350 }, { "epoch": 12.217484926787252, "grad_norm": 0.36709001660346985, "learning_rate": 0.00041702302410125217, "loss": 3.0843, "step": 45400 }, { "epoch": 12.230943152454781, "grad_norm": 0.3590288758277893, "learning_rate": 0.00041682105830079437, "loss": 3.0847, "step": 45450 }, { "epoch": 12.244401378122308, "grad_norm": 0.3631494641304016, "learning_rate": 0.00041661909250033656, "loss": 3.0913, "step": 45500 }, { "epoch": 12.257859603789836, "grad_norm": 0.3625164031982422, "learning_rate": 0.00041641712669987876, "loss": 3.0862, "step": 45550 }, { "epoch": 12.271317829457365, "grad_norm": 0.39201679825782776, "learning_rate": 0.00041621516089942096, "loss": 3.0965, "step": 45600 }, { "epoch": 12.284776055124892, "grad_norm": 0.38754960894584656, "learning_rate": 0.0004160131950989632, "loss": 3.0899, "step": 45650 }, { "epoch": 12.298234280792421, "grad_norm": 0.36467859148979187, "learning_rate": 0.0004158112292985054, "loss": 3.102, "step": 45700 }, { "epoch": 12.311692506459949, "grad_norm": 0.38860246539115906, "learning_rate": 0.0004156092634980476, "loss": 3.094, "step": 45750 }, { "epoch": 12.325150732127476, "grad_norm": 0.36453795433044434, "learning_rate": 0.00041540729769758985, "loss": 3.0931, "step": 45800 }, { "epoch": 12.338608957795005, "grad_norm": 0.39302724599838257, "learning_rate": 0.0004152053318971321, "loss": 3.1034, "step": 45850 }, { "epoch": 12.352067183462532, "grad_norm": 0.38808274269104004, "learning_rate": 0.0004150033660966743, "loss": 3.0973, "step": 45900 }, { "epoch": 12.36552540913006, "grad_norm": 0.3711463212966919, "learning_rate": 0.0004148014002962165, "loss": 3.1053, "step": 45950 }, { "epoch": 12.378983634797589, "grad_norm": 0.3651505708694458, "learning_rate": 0.0004145994344957587, "loss": 3.0961, "step": 46000 }, { "epoch": 12.378983634797589, "eval_accuracy": 0.3916768034182908, "eval_loss": 3.307568073272705, "eval_runtime": 53.4044, "eval_samples_per_second": 337.275, "eval_steps_per_second": 21.084, "step": 46000 }, { "epoch": 12.392441860465116, "grad_norm": 0.3489798903465271, "learning_rate": 0.0004143974686953009, "loss": 3.1111, "step": 46050 }, { "epoch": 12.405900086132645, "grad_norm": 0.37551695108413696, "learning_rate": 0.00041419550289484313, "loss": 3.1072, "step": 46100 }, { "epoch": 12.419358311800172, "grad_norm": 0.35827645659446716, "learning_rate": 0.00041399353709438533, "loss": 3.1013, "step": 46150 }, { "epoch": 12.4328165374677, "grad_norm": 0.3637438416481018, "learning_rate": 0.0004137915712939275, "loss": 3.1106, "step": 46200 }, { "epoch": 12.446274763135229, "grad_norm": 0.3819226920604706, "learning_rate": 0.0004135896054934697, "loss": 3.1055, "step": 46250 }, { "epoch": 12.459732988802756, "grad_norm": 0.36781030893325806, "learning_rate": 0.0004133876396930119, "loss": 3.1078, "step": 46300 }, { "epoch": 12.473191214470285, "grad_norm": 0.39347222447395325, "learning_rate": 0.00041318567389255417, "loss": 3.1151, "step": 46350 }, { "epoch": 12.486649440137812, "grad_norm": 0.39293286204338074, "learning_rate": 0.00041298370809209636, "loss": 3.1106, "step": 46400 }, { "epoch": 12.50010766580534, "grad_norm": 0.37734219431877136, "learning_rate": 0.00041278174229163856, "loss": 3.106, "step": 46450 }, { "epoch": 12.513565891472869, "grad_norm": 0.3757116198539734, "learning_rate": 0.00041257977649118076, "loss": 3.1061, "step": 46500 }, { "epoch": 12.527024117140396, "grad_norm": 0.4108484089374542, "learning_rate": 0.00041237781069072295, "loss": 3.1093, "step": 46550 }, { "epoch": 12.540482342807923, "grad_norm": 0.3842467963695526, "learning_rate": 0.0004121758448902652, "loss": 3.1056, "step": 46600 }, { "epoch": 12.553940568475452, "grad_norm": 0.38736289739608765, "learning_rate": 0.0004119738790898074, "loss": 3.1208, "step": 46650 }, { "epoch": 12.56739879414298, "grad_norm": 0.3820643126964569, "learning_rate": 0.00041177191328934965, "loss": 3.1143, "step": 46700 }, { "epoch": 12.580857019810509, "grad_norm": 0.37509894371032715, "learning_rate": 0.00041156994748889185, "loss": 3.1174, "step": 46750 }, { "epoch": 12.594315245478036, "grad_norm": 0.35530561208724976, "learning_rate": 0.0004113679816884341, "loss": 3.1197, "step": 46800 }, { "epoch": 12.607773471145563, "grad_norm": 0.3885257840156555, "learning_rate": 0.0004111660158879763, "loss": 3.1136, "step": 46850 }, { "epoch": 12.621231696813092, "grad_norm": 0.38516077399253845, "learning_rate": 0.0004109640500875185, "loss": 3.1234, "step": 46900 }, { "epoch": 12.63468992248062, "grad_norm": 0.3831053674221039, "learning_rate": 0.0004107620842870607, "loss": 3.1207, "step": 46950 }, { "epoch": 12.648148148148149, "grad_norm": 0.3733752965927124, "learning_rate": 0.0004105601184866029, "loss": 3.1092, "step": 47000 }, { "epoch": 12.648148148148149, "eval_accuracy": 0.3917694791860873, "eval_loss": 3.3039400577545166, "eval_runtime": 53.5807, "eval_samples_per_second": 336.166, "eval_steps_per_second": 21.015, "step": 47000 }, { "epoch": 12.661606373815676, "grad_norm": 0.38987475633621216, "learning_rate": 0.00041035815268614513, "loss": 3.114, "step": 47050 }, { "epoch": 12.675064599483203, "grad_norm": 0.40705275535583496, "learning_rate": 0.0004101561868856873, "loss": 3.1164, "step": 47100 }, { "epoch": 12.688522825150732, "grad_norm": 0.373837411403656, "learning_rate": 0.0004099542210852295, "loss": 3.1208, "step": 47150 }, { "epoch": 12.70198105081826, "grad_norm": 0.3675236701965332, "learning_rate": 0.0004097522552847717, "loss": 3.1212, "step": 47200 }, { "epoch": 12.715439276485789, "grad_norm": 0.35344648361206055, "learning_rate": 0.00040955028948431397, "loss": 3.1104, "step": 47250 }, { "epoch": 12.728897502153316, "grad_norm": 0.3793870508670807, "learning_rate": 0.00040934832368385616, "loss": 3.1225, "step": 47300 }, { "epoch": 12.742355727820843, "grad_norm": 0.4043644666671753, "learning_rate": 0.00040914635788339836, "loss": 3.1305, "step": 47350 }, { "epoch": 12.755813953488373, "grad_norm": 0.37010902166366577, "learning_rate": 0.00040894439208294056, "loss": 3.1237, "step": 47400 }, { "epoch": 12.7692721791559, "grad_norm": 0.4060969054698944, "learning_rate": 0.00040874242628248275, "loss": 3.1249, "step": 47450 }, { "epoch": 12.782730404823429, "grad_norm": 0.3517776131629944, "learning_rate": 0.000408540460482025, "loss": 3.1302, "step": 47500 }, { "epoch": 12.796188630490956, "grad_norm": 0.34859320521354675, "learning_rate": 0.00040833849468156725, "loss": 3.1223, "step": 47550 }, { "epoch": 12.809646856158484, "grad_norm": 0.3657214641571045, "learning_rate": 0.00040813652888110945, "loss": 3.1296, "step": 47600 }, { "epoch": 12.823105081826013, "grad_norm": 0.3647889494895935, "learning_rate": 0.00040793456308065165, "loss": 3.1298, "step": 47650 }, { "epoch": 12.83656330749354, "grad_norm": 0.37984415888786316, "learning_rate": 0.0004077325972801939, "loss": 3.1326, "step": 47700 }, { "epoch": 12.850021533161069, "grad_norm": 0.3746028244495392, "learning_rate": 0.0004075306314797361, "loss": 3.1221, "step": 47750 }, { "epoch": 12.863479758828596, "grad_norm": 0.360163152217865, "learning_rate": 0.0004073286656792783, "loss": 3.1311, "step": 47800 }, { "epoch": 12.876937984496124, "grad_norm": 0.4016077518463135, "learning_rate": 0.0004071266998788205, "loss": 3.1373, "step": 47850 }, { "epoch": 12.890396210163653, "grad_norm": 0.3723677694797516, "learning_rate": 0.0004069247340783627, "loss": 3.1192, "step": 47900 }, { "epoch": 12.90385443583118, "grad_norm": 0.3679752051830292, "learning_rate": 0.00040672276827790493, "loss": 3.1305, "step": 47950 }, { "epoch": 12.917312661498707, "grad_norm": 0.3905462920665741, "learning_rate": 0.0004065208024774471, "loss": 3.129, "step": 48000 }, { "epoch": 12.917312661498707, "eval_accuracy": 0.3922937002641857, "eval_loss": 3.2970693111419678, "eval_runtime": 53.7542, "eval_samples_per_second": 335.081, "eval_steps_per_second": 20.947, "step": 48000 }, { "epoch": 12.930770887166236, "grad_norm": 0.36770474910736084, "learning_rate": 0.0004063188366769893, "loss": 3.1407, "step": 48050 }, { "epoch": 12.944229112833764, "grad_norm": 0.3561168611049652, "learning_rate": 0.0004061168708765315, "loss": 3.1399, "step": 48100 }, { "epoch": 12.957687338501293, "grad_norm": 0.3862304389476776, "learning_rate": 0.0004059149050760737, "loss": 3.1351, "step": 48150 }, { "epoch": 12.97114556416882, "grad_norm": 0.36406269669532776, "learning_rate": 0.00040571293927561597, "loss": 3.1314, "step": 48200 }, { "epoch": 12.984603789836347, "grad_norm": 0.3505527377128601, "learning_rate": 0.00040551097347515816, "loss": 3.129, "step": 48250 }, { "epoch": 12.998062015503876, "grad_norm": 0.37477901577949524, "learning_rate": 0.00040530900767470036, "loss": 3.1402, "step": 48300 }, { "epoch": 13.011304909560723, "grad_norm": 0.39581868052482605, "learning_rate": 0.00040510704187424255, "loss": 3.0524, "step": 48350 }, { "epoch": 13.024763135228252, "grad_norm": 0.35605454444885254, "learning_rate": 0.00040490507607378475, "loss": 3.0409, "step": 48400 }, { "epoch": 13.038221360895779, "grad_norm": 0.36419981718063354, "learning_rate": 0.00040470311027332705, "loss": 3.0414, "step": 48450 }, { "epoch": 13.051679586563308, "grad_norm": 0.377587229013443, "learning_rate": 0.00040450114447286925, "loss": 3.043, "step": 48500 }, { "epoch": 13.065137812230835, "grad_norm": 0.3738135099411011, "learning_rate": 0.00040429917867241145, "loss": 3.0349, "step": 48550 }, { "epoch": 13.078596037898363, "grad_norm": 0.37447571754455566, "learning_rate": 0.00040409721287195364, "loss": 3.0418, "step": 48600 }, { "epoch": 13.092054263565892, "grad_norm": 0.37799814343452454, "learning_rate": 0.0004038952470714959, "loss": 3.0625, "step": 48650 }, { "epoch": 13.10551248923342, "grad_norm": 0.36537399888038635, "learning_rate": 0.0004036932812710381, "loss": 3.0552, "step": 48700 }, { "epoch": 13.118970714900948, "grad_norm": 0.3800034821033478, "learning_rate": 0.0004034913154705803, "loss": 3.0689, "step": 48750 }, { "epoch": 13.132428940568476, "grad_norm": 0.36928868293762207, "learning_rate": 0.0004032893496701225, "loss": 3.059, "step": 48800 }, { "epoch": 13.145887166236003, "grad_norm": 0.36283737421035767, "learning_rate": 0.0004030873838696647, "loss": 3.066, "step": 48850 }, { "epoch": 13.159345391903532, "grad_norm": 0.3781079053878784, "learning_rate": 0.00040288541806920693, "loss": 3.0596, "step": 48900 }, { "epoch": 13.17280361757106, "grad_norm": 0.37426838278770447, "learning_rate": 0.0004026834522687491, "loss": 3.0596, "step": 48950 }, { "epoch": 13.186261843238588, "grad_norm": 0.36410465836524963, "learning_rate": 0.0004024814864682913, "loss": 3.0675, "step": 49000 }, { "epoch": 13.186261843238588, "eval_accuracy": 0.3923700790036475, "eval_loss": 3.3071610927581787, "eval_runtime": 53.2206, "eval_samples_per_second": 338.441, "eval_steps_per_second": 21.157, "step": 49000 }, { "epoch": 13.199720068906116, "grad_norm": 0.39360588788986206, "learning_rate": 0.0004022795206678335, "loss": 3.0699, "step": 49050 }, { "epoch": 13.213178294573643, "grad_norm": 0.3912392556667328, "learning_rate": 0.00040207755486737577, "loss": 3.0677, "step": 49100 }, { "epoch": 13.226636520241172, "grad_norm": 0.403340607881546, "learning_rate": 0.00040187558906691796, "loss": 3.061, "step": 49150 }, { "epoch": 13.2400947459087, "grad_norm": 0.3977563977241516, "learning_rate": 0.00040167362326646016, "loss": 3.0714, "step": 49200 }, { "epoch": 13.253552971576227, "grad_norm": 0.44516828656196594, "learning_rate": 0.00040147165746600235, "loss": 3.0773, "step": 49250 }, { "epoch": 13.267011197243756, "grad_norm": 0.38486993312835693, "learning_rate": 0.00040126969166554455, "loss": 3.0755, "step": 49300 }, { "epoch": 13.280469422911283, "grad_norm": 0.37119653820991516, "learning_rate": 0.00040106772586508686, "loss": 3.0801, "step": 49350 }, { "epoch": 13.293927648578812, "grad_norm": 0.37240782380104065, "learning_rate": 0.00040086576006462905, "loss": 3.0811, "step": 49400 }, { "epoch": 13.30738587424634, "grad_norm": 0.362568199634552, "learning_rate": 0.00040066379426417125, "loss": 3.0827, "step": 49450 }, { "epoch": 13.320844099913867, "grad_norm": 0.3640483617782593, "learning_rate": 0.00040046182846371344, "loss": 3.0916, "step": 49500 }, { "epoch": 13.334302325581396, "grad_norm": 0.37040454149246216, "learning_rate": 0.0004002598626632557, "loss": 3.0813, "step": 49550 }, { "epoch": 13.347760551248923, "grad_norm": 0.38464635610580444, "learning_rate": 0.0004000578968627979, "loss": 3.099, "step": 49600 }, { "epoch": 13.361218776916452, "grad_norm": 0.38465845584869385, "learning_rate": 0.0003998559310623401, "loss": 3.0855, "step": 49650 }, { "epoch": 13.37467700258398, "grad_norm": 0.3531036376953125, "learning_rate": 0.0003996539652618823, "loss": 3.0837, "step": 49700 }, { "epoch": 13.388135228251507, "grad_norm": 0.38886111974716187, "learning_rate": 0.0003994519994614245, "loss": 3.0925, "step": 49750 }, { "epoch": 13.401593453919036, "grad_norm": 0.3840842843055725, "learning_rate": 0.00039925003366096673, "loss": 3.0875, "step": 49800 }, { "epoch": 13.415051679586563, "grad_norm": 0.3760302662849426, "learning_rate": 0.0003990480678605089, "loss": 3.0873, "step": 49850 }, { "epoch": 13.428509905254092, "grad_norm": 0.36125680804252625, "learning_rate": 0.0003988461020600511, "loss": 3.0911, "step": 49900 }, { "epoch": 13.44196813092162, "grad_norm": 0.38114649057388306, "learning_rate": 0.0003986441362595933, "loss": 3.0869, "step": 49950 }, { "epoch": 13.455426356589147, "grad_norm": 0.38252297043800354, "learning_rate": 0.0003984421704591355, "loss": 3.0914, "step": 50000 }, { "epoch": 13.455426356589147, "eval_accuracy": 0.3924209257320517, "eval_loss": 3.3010644912719727, "eval_runtime": 53.2801, "eval_samples_per_second": 338.062, "eval_steps_per_second": 21.134, "step": 50000 }, { "epoch": 13.468884582256676, "grad_norm": 0.3857160806655884, "learning_rate": 0.00039824020465867776, "loss": 3.0851, "step": 50050 }, { "epoch": 13.482342807924203, "grad_norm": 0.37299609184265137, "learning_rate": 0.00039803823885821996, "loss": 3.0893, "step": 50100 }, { "epoch": 13.49580103359173, "grad_norm": 0.350725382566452, "learning_rate": 0.00039783627305776216, "loss": 3.0979, "step": 50150 }, { "epoch": 13.50925925925926, "grad_norm": 0.37564557790756226, "learning_rate": 0.00039763430725730435, "loss": 3.0976, "step": 50200 }, { "epoch": 13.522717484926787, "grad_norm": 0.3750132918357849, "learning_rate": 0.00039743234145684666, "loss": 3.1028, "step": 50250 }, { "epoch": 13.536175710594316, "grad_norm": 0.38835328817367554, "learning_rate": 0.00039723037565638885, "loss": 3.0929, "step": 50300 }, { "epoch": 13.549633936261843, "grad_norm": 0.38070812821388245, "learning_rate": 0.00039702840985593105, "loss": 3.0919, "step": 50350 }, { "epoch": 13.56309216192937, "grad_norm": 0.393497496843338, "learning_rate": 0.00039682644405547324, "loss": 3.0975, "step": 50400 }, { "epoch": 13.5765503875969, "grad_norm": 0.3958127200603485, "learning_rate": 0.00039662447825501544, "loss": 3.0977, "step": 50450 }, { "epoch": 13.590008613264427, "grad_norm": 0.4099980592727661, "learning_rate": 0.0003964225124545577, "loss": 3.1074, "step": 50500 }, { "epoch": 13.603466838931956, "grad_norm": 0.37544602155685425, "learning_rate": 0.0003962205466540999, "loss": 3.1021, "step": 50550 }, { "epoch": 13.616925064599483, "grad_norm": 0.3651670515537262, "learning_rate": 0.0003960185808536421, "loss": 3.1032, "step": 50600 }, { "epoch": 13.63038329026701, "grad_norm": 0.4086431562900543, "learning_rate": 0.0003958166150531843, "loss": 3.0982, "step": 50650 }, { "epoch": 13.64384151593454, "grad_norm": 0.37201717495918274, "learning_rate": 0.0003956146492527265, "loss": 3.1072, "step": 50700 }, { "epoch": 13.657299741602067, "grad_norm": 0.372783899307251, "learning_rate": 0.0003954126834522687, "loss": 3.1025, "step": 50750 }, { "epoch": 13.670757967269594, "grad_norm": 0.38944682478904724, "learning_rate": 0.0003952107176518109, "loss": 3.101, "step": 50800 }, { "epoch": 13.684216192937123, "grad_norm": 0.3839119076728821, "learning_rate": 0.0003950087518513531, "loss": 3.1006, "step": 50850 }, { "epoch": 13.69767441860465, "grad_norm": 0.38758784532546997, "learning_rate": 0.0003948067860508953, "loss": 3.0906, "step": 50900 }, { "epoch": 13.71113264427218, "grad_norm": 0.3955155611038208, "learning_rate": 0.0003946048202504375, "loss": 3.1069, "step": 50950 }, { "epoch": 13.724590869939707, "grad_norm": 0.3915644586086273, "learning_rate": 0.00039440285444997976, "loss": 3.1114, "step": 51000 }, { "epoch": 13.724590869939707, "eval_accuracy": 0.3928451917030308, "eval_loss": 3.2942299842834473, "eval_runtime": 53.1633, "eval_samples_per_second": 338.805, "eval_steps_per_second": 21.18, "step": 51000 }, { "epoch": 13.738049095607234, "grad_norm": 0.3906634449958801, "learning_rate": 0.00039420088864952196, "loss": 3.1124, "step": 51050 }, { "epoch": 13.751507321274763, "grad_norm": 0.3947088420391083, "learning_rate": 0.00039399892284906415, "loss": 3.1152, "step": 51100 }, { "epoch": 13.76496554694229, "grad_norm": 0.37320712208747864, "learning_rate": 0.00039379695704860646, "loss": 3.1083, "step": 51150 }, { "epoch": 13.77842377260982, "grad_norm": 0.3770774006843567, "learning_rate": 0.00039359499124814865, "loss": 3.1003, "step": 51200 }, { "epoch": 13.791881998277347, "grad_norm": 0.35937607288360596, "learning_rate": 0.00039339302544769085, "loss": 3.1084, "step": 51250 }, { "epoch": 13.805340223944874, "grad_norm": 0.3897993564605713, "learning_rate": 0.00039319105964723305, "loss": 3.0999, "step": 51300 }, { "epoch": 13.818798449612403, "grad_norm": 0.3844742178916931, "learning_rate": 0.00039298909384677524, "loss": 3.1092, "step": 51350 }, { "epoch": 13.83225667527993, "grad_norm": 0.3903350532054901, "learning_rate": 0.0003927871280463175, "loss": 3.1111, "step": 51400 }, { "epoch": 13.84571490094746, "grad_norm": 0.3959101438522339, "learning_rate": 0.0003925851622458597, "loss": 3.1099, "step": 51450 }, { "epoch": 13.859173126614987, "grad_norm": 0.3868245482444763, "learning_rate": 0.0003923831964454019, "loss": 3.1137, "step": 51500 }, { "epoch": 13.872631352282514, "grad_norm": 0.3856070935726166, "learning_rate": 0.0003921812306449441, "loss": 3.1112, "step": 51550 }, { "epoch": 13.886089577950044, "grad_norm": 0.38117513060569763, "learning_rate": 0.0003919792648444863, "loss": 3.1144, "step": 51600 }, { "epoch": 13.89954780361757, "grad_norm": 0.34406477212905884, "learning_rate": 0.0003917772990440285, "loss": 3.1228, "step": 51650 }, { "epoch": 13.9130060292851, "grad_norm": 0.3809207081794739, "learning_rate": 0.0003915753332435707, "loss": 3.1175, "step": 51700 }, { "epoch": 13.926464254952627, "grad_norm": 0.38697007298469543, "learning_rate": 0.0003913733674431129, "loss": 3.1133, "step": 51750 }, { "epoch": 13.939922480620154, "grad_norm": 0.37401506304740906, "learning_rate": 0.0003911714016426551, "loss": 3.1094, "step": 51800 }, { "epoch": 13.953380706287684, "grad_norm": 0.385299950838089, "learning_rate": 0.0003909694358421973, "loss": 3.1148, "step": 51850 }, { "epoch": 13.96683893195521, "grad_norm": 0.3646024167537689, "learning_rate": 0.00039076747004173956, "loss": 3.1251, "step": 51900 }, { "epoch": 13.98029715762274, "grad_norm": 0.39629772305488586, "learning_rate": 0.00039056550424128176, "loss": 3.1245, "step": 51950 }, { "epoch": 13.993755383290267, "grad_norm": 0.3806076645851135, "learning_rate": 0.00039036353844082395, "loss": 3.1143, "step": 52000 }, { "epoch": 13.993755383290267, "eval_accuracy": 0.3933573529801615, "eval_loss": 3.2886648178100586, "eval_runtime": 53.2025, "eval_samples_per_second": 338.556, "eval_steps_per_second": 21.164, "step": 52000 }, { "epoch": 14.006998277347115, "grad_norm": 0.39359694719314575, "learning_rate": 0.0003901615726403662, "loss": 3.0789, "step": 52050 }, { "epoch": 14.020456503014643, "grad_norm": 0.36778563261032104, "learning_rate": 0.00038995960683990845, "loss": 3.0137, "step": 52100 }, { "epoch": 14.03391472868217, "grad_norm": 0.3619430959224701, "learning_rate": 0.00038975764103945065, "loss": 3.0305, "step": 52150 }, { "epoch": 14.047372954349699, "grad_norm": 0.39204686880111694, "learning_rate": 0.00038955567523899285, "loss": 3.0243, "step": 52200 }, { "epoch": 14.060831180017226, "grad_norm": 0.37758973240852356, "learning_rate": 0.00038935370943853504, "loss": 3.0231, "step": 52250 }, { "epoch": 14.074289405684755, "grad_norm": 0.38672971725463867, "learning_rate": 0.00038915174363807724, "loss": 3.0347, "step": 52300 }, { "epoch": 14.087747631352283, "grad_norm": 0.3783572018146515, "learning_rate": 0.0003889497778376195, "loss": 3.0336, "step": 52350 }, { "epoch": 14.10120585701981, "grad_norm": 0.3870309293270111, "learning_rate": 0.0003887478120371617, "loss": 3.0386, "step": 52400 }, { "epoch": 14.114664082687339, "grad_norm": 0.39776405692100525, "learning_rate": 0.0003885458462367039, "loss": 3.027, "step": 52450 }, { "epoch": 14.128122308354866, "grad_norm": 0.44890081882476807, "learning_rate": 0.0003883438804362461, "loss": 3.0538, "step": 52500 }, { "epoch": 14.141580534022394, "grad_norm": 0.37647005915641785, "learning_rate": 0.0003881419146357883, "loss": 3.0411, "step": 52550 }, { "epoch": 14.155038759689923, "grad_norm": 0.34855517745018005, "learning_rate": 0.0003879399488353305, "loss": 3.0532, "step": 52600 }, { "epoch": 14.16849698535745, "grad_norm": 0.3922046720981598, "learning_rate": 0.0003877379830348727, "loss": 3.0465, "step": 52650 }, { "epoch": 14.18195521102498, "grad_norm": 0.3750133216381073, "learning_rate": 0.0003875360172344149, "loss": 3.0422, "step": 52700 }, { "epoch": 14.195413436692506, "grad_norm": 0.3771599233150482, "learning_rate": 0.0003873340514339571, "loss": 3.0526, "step": 52750 }, { "epoch": 14.208871662360034, "grad_norm": 0.3815842866897583, "learning_rate": 0.0003871320856334993, "loss": 3.0552, "step": 52800 }, { "epoch": 14.222329888027563, "grad_norm": 0.4113651514053345, "learning_rate": 0.00038693011983304156, "loss": 3.057, "step": 52850 }, { "epoch": 14.23578811369509, "grad_norm": 0.3831230401992798, "learning_rate": 0.0003867281540325838, "loss": 3.0523, "step": 52900 }, { "epoch": 14.24924633936262, "grad_norm": 0.3881702423095703, "learning_rate": 0.000386526188232126, "loss": 3.0703, "step": 52950 }, { "epoch": 14.262704565030146, "grad_norm": 0.3693937063217163, "learning_rate": 0.00038632422243166825, "loss": 3.0639, "step": 53000 }, { "epoch": 14.262704565030146, "eval_accuracy": 0.3928689853643994, "eval_loss": 3.3038170337677, "eval_runtime": 53.2513, "eval_samples_per_second": 338.246, "eval_steps_per_second": 21.145, "step": 53000 }, { "epoch": 14.276162790697674, "grad_norm": 0.3559059202671051, "learning_rate": 0.00038612225663121045, "loss": 3.0634, "step": 53050 }, { "epoch": 14.289621016365203, "grad_norm": 0.4166322648525238, "learning_rate": 0.00038592029083075265, "loss": 3.0625, "step": 53100 }, { "epoch": 14.30307924203273, "grad_norm": 0.3934720456600189, "learning_rate": 0.00038571832503029484, "loss": 3.0664, "step": 53150 }, { "epoch": 14.31653746770026, "grad_norm": 0.38985350728034973, "learning_rate": 0.00038551635922983704, "loss": 3.0671, "step": 53200 }, { "epoch": 14.329995693367787, "grad_norm": 0.3798811435699463, "learning_rate": 0.0003853143934293793, "loss": 3.0716, "step": 53250 }, { "epoch": 14.343453919035314, "grad_norm": 0.3842821717262268, "learning_rate": 0.0003851124276289215, "loss": 3.066, "step": 53300 }, { "epoch": 14.356912144702843, "grad_norm": 0.3599495589733124, "learning_rate": 0.0003849104618284637, "loss": 3.0733, "step": 53350 }, { "epoch": 14.37037037037037, "grad_norm": 0.3783971965312958, "learning_rate": 0.0003847084960280059, "loss": 3.0649, "step": 53400 }, { "epoch": 14.383828596037898, "grad_norm": 0.38178735971450806, "learning_rate": 0.0003845065302275481, "loss": 3.0699, "step": 53450 }, { "epoch": 14.397286821705427, "grad_norm": 0.36713486909866333, "learning_rate": 0.0003843045644270903, "loss": 3.0643, "step": 53500 }, { "epoch": 14.410745047372954, "grad_norm": 0.4043576717376709, "learning_rate": 0.0003841025986266325, "loss": 3.075, "step": 53550 }, { "epoch": 14.424203273040483, "grad_norm": 0.3947139084339142, "learning_rate": 0.0003839006328261747, "loss": 3.077, "step": 53600 }, { "epoch": 14.43766149870801, "grad_norm": 0.38625821471214294, "learning_rate": 0.0003836986670257169, "loss": 3.0728, "step": 53650 }, { "epoch": 14.451119724375538, "grad_norm": 0.37357762455940247, "learning_rate": 0.0003834967012252591, "loss": 3.0738, "step": 53700 }, { "epoch": 14.464577950043067, "grad_norm": 0.3948591649532318, "learning_rate": 0.00038329473542480136, "loss": 3.0722, "step": 53750 }, { "epoch": 14.478036175710594, "grad_norm": 0.38419365882873535, "learning_rate": 0.0003830927696243436, "loss": 3.0827, "step": 53800 }, { "epoch": 14.491494401378123, "grad_norm": 0.36976271867752075, "learning_rate": 0.0003828908038238858, "loss": 3.0789, "step": 53850 }, { "epoch": 14.50495262704565, "grad_norm": 0.36416393518447876, "learning_rate": 0.000382688838023428, "loss": 3.0837, "step": 53900 }, { "epoch": 14.518410852713178, "grad_norm": 0.3858191967010498, "learning_rate": 0.00038248687222297025, "loss": 3.0784, "step": 53950 }, { "epoch": 14.531869078380707, "grad_norm": 0.404272198677063, "learning_rate": 0.00038228490642251245, "loss": 3.0816, "step": 54000 }, { "epoch": 14.531869078380707, "eval_accuracy": 0.3930469489138139, "eval_loss": 3.2975451946258545, "eval_runtime": 53.2861, "eval_samples_per_second": 338.024, "eval_steps_per_second": 21.131, "step": 54000 }, { "epoch": 14.545327304048234, "grad_norm": 0.37756025791168213, "learning_rate": 0.00038208294062205464, "loss": 3.0926, "step": 54050 }, { "epoch": 14.558785529715763, "grad_norm": 0.40512412786483765, "learning_rate": 0.00038188097482159684, "loss": 3.0828, "step": 54100 }, { "epoch": 14.57224375538329, "grad_norm": 0.3908289074897766, "learning_rate": 0.00038167900902113904, "loss": 3.0801, "step": 54150 }, { "epoch": 14.585701981050818, "grad_norm": 0.4162972867488861, "learning_rate": 0.0003814770432206813, "loss": 3.0888, "step": 54200 }, { "epoch": 14.599160206718347, "grad_norm": 0.4289568066596985, "learning_rate": 0.0003812750774202235, "loss": 3.0959, "step": 54250 }, { "epoch": 14.612618432385874, "grad_norm": 0.3740384876728058, "learning_rate": 0.0003810731116197657, "loss": 3.0931, "step": 54300 }, { "epoch": 14.626076658053403, "grad_norm": 0.36348482966423035, "learning_rate": 0.0003808711458193079, "loss": 3.0866, "step": 54350 }, { "epoch": 14.63953488372093, "grad_norm": 0.39538612961769104, "learning_rate": 0.00038066918001885007, "loss": 3.0867, "step": 54400 }, { "epoch": 14.652993109388458, "grad_norm": 0.3879004716873169, "learning_rate": 0.0003804672142183923, "loss": 3.0887, "step": 54450 }, { "epoch": 14.666451335055987, "grad_norm": 0.355636864900589, "learning_rate": 0.0003802652484179345, "loss": 3.0893, "step": 54500 }, { "epoch": 14.679909560723514, "grad_norm": 0.37391379475593567, "learning_rate": 0.0003800632826174767, "loss": 3.0987, "step": 54550 }, { "epoch": 14.693367786391041, "grad_norm": 0.3761315643787384, "learning_rate": 0.0003798613168170189, "loss": 3.0868, "step": 54600 }, { "epoch": 14.70682601205857, "grad_norm": 0.3604639172554016, "learning_rate": 0.0003796593510165611, "loss": 3.1005, "step": 54650 }, { "epoch": 14.720284237726098, "grad_norm": 0.41159263253211975, "learning_rate": 0.0003794573852161034, "loss": 3.093, "step": 54700 }, { "epoch": 14.733742463393627, "grad_norm": 0.39038246870040894, "learning_rate": 0.0003792554194156456, "loss": 3.0832, "step": 54750 }, { "epoch": 14.747200689061154, "grad_norm": 0.39387059211730957, "learning_rate": 0.0003790534536151878, "loss": 3.1006, "step": 54800 }, { "epoch": 14.760658914728682, "grad_norm": 0.37214162945747375, "learning_rate": 0.00037885148781473005, "loss": 3.094, "step": 54850 }, { "epoch": 14.77411714039621, "grad_norm": 0.3764510154724121, "learning_rate": 0.00037864952201427225, "loss": 3.0977, "step": 54900 }, { "epoch": 14.787575366063738, "grad_norm": 0.3526248037815094, "learning_rate": 0.00037844755621381444, "loss": 3.1015, "step": 54950 }, { "epoch": 14.801033591731267, "grad_norm": 0.4617568552494049, "learning_rate": 0.00037824559041335664, "loss": 3.0951, "step": 55000 }, { "epoch": 14.801033591731267, "eval_accuracy": 0.39400706117643686, "eval_loss": 3.287712335586548, "eval_runtime": 53.2104, "eval_samples_per_second": 338.505, "eval_steps_per_second": 21.161, "step": 55000 }, { "epoch": 14.814491817398794, "grad_norm": 0.3996323347091675, "learning_rate": 0.00037804362461289884, "loss": 3.1008, "step": 55050 }, { "epoch": 14.827950043066322, "grad_norm": 0.3723200261592865, "learning_rate": 0.0003778416588124411, "loss": 3.0927, "step": 55100 }, { "epoch": 14.84140826873385, "grad_norm": 0.3873573839664459, "learning_rate": 0.0003776396930119833, "loss": 3.1062, "step": 55150 }, { "epoch": 14.854866494401378, "grad_norm": 0.37784600257873535, "learning_rate": 0.0003774377272115255, "loss": 3.0936, "step": 55200 }, { "epoch": 14.868324720068905, "grad_norm": 0.38209277391433716, "learning_rate": 0.0003772357614110677, "loss": 3.1048, "step": 55250 }, { "epoch": 14.881782945736434, "grad_norm": 0.3873889446258545, "learning_rate": 0.00037703379561060987, "loss": 3.0883, "step": 55300 }, { "epoch": 14.895241171403962, "grad_norm": 0.3716479241847992, "learning_rate": 0.0003768318298101521, "loss": 3.1102, "step": 55350 }, { "epoch": 14.90869939707149, "grad_norm": 0.3743898570537567, "learning_rate": 0.0003766298640096943, "loss": 3.0971, "step": 55400 }, { "epoch": 14.922157622739018, "grad_norm": 0.3971049189567566, "learning_rate": 0.0003764278982092365, "loss": 3.1034, "step": 55450 }, { "epoch": 14.935615848406545, "grad_norm": 0.3767416179180145, "learning_rate": 0.0003762259324087787, "loss": 3.0957, "step": 55500 }, { "epoch": 14.949074074074074, "grad_norm": 0.37302061915397644, "learning_rate": 0.0003760239666083209, "loss": 3.0999, "step": 55550 }, { "epoch": 14.962532299741602, "grad_norm": 0.37083128094673157, "learning_rate": 0.0003758220008078632, "loss": 3.0928, "step": 55600 }, { "epoch": 14.97599052540913, "grad_norm": 0.41234105825424194, "learning_rate": 0.0003756200350074054, "loss": 3.1026, "step": 55650 }, { "epoch": 14.989448751076658, "grad_norm": 0.3714658319950104, "learning_rate": 0.0003754180692069476, "loss": 3.1092, "step": 55700 }, { "epoch": 15.002691645133506, "grad_norm": 0.3906070590019226, "learning_rate": 0.0003752161034064898, "loss": 3.0814, "step": 55750 }, { "epoch": 15.016149870801033, "grad_norm": 0.40009233355522156, "learning_rate": 0.00037501413760603205, "loss": 3.0027, "step": 55800 }, { "epoch": 15.02960809646856, "grad_norm": 0.39637264609336853, "learning_rate": 0.00037481217180557425, "loss": 3.0125, "step": 55850 }, { "epoch": 15.04306632213609, "grad_norm": 0.41015830636024475, "learning_rate": 0.00037461020600511644, "loss": 3.0059, "step": 55900 }, { "epoch": 15.056524547803617, "grad_norm": 0.4093896746635437, "learning_rate": 0.00037440824020465864, "loss": 3.0224, "step": 55950 }, { "epoch": 15.069982773471146, "grad_norm": 0.3793170750141144, "learning_rate": 0.00037420627440420083, "loss": 3.0158, "step": 56000 }, { "epoch": 15.069982773471146, "eval_accuracy": 0.3932361030893516, "eval_loss": 3.3005900382995605, "eval_runtime": 53.2337, "eval_samples_per_second": 338.357, "eval_steps_per_second": 21.152, "step": 56000 }, { "epoch": 15.083440999138674, "grad_norm": 0.3804849088191986, "learning_rate": 0.0003740043086037431, "loss": 3.0165, "step": 56050 }, { "epoch": 15.0968992248062, "grad_norm": 0.41092732548713684, "learning_rate": 0.0003738023428032853, "loss": 3.0383, "step": 56100 }, { "epoch": 15.11035745047373, "grad_norm": 0.39817407727241516, "learning_rate": 0.0003736003770028275, "loss": 3.0375, "step": 56150 }, { "epoch": 15.123815676141257, "grad_norm": 0.4043419659137726, "learning_rate": 0.00037339841120236967, "loss": 3.0246, "step": 56200 }, { "epoch": 15.137273901808786, "grad_norm": 0.39283016324043274, "learning_rate": 0.00037319644540191187, "loss": 3.0341, "step": 56250 }, { "epoch": 15.150732127476314, "grad_norm": 0.3876219391822815, "learning_rate": 0.0003729944796014541, "loss": 3.0386, "step": 56300 }, { "epoch": 15.164190353143841, "grad_norm": 0.39713042974472046, "learning_rate": 0.0003727925138009963, "loss": 3.0381, "step": 56350 }, { "epoch": 15.17764857881137, "grad_norm": 0.3937109112739563, "learning_rate": 0.0003725905480005385, "loss": 3.0339, "step": 56400 }, { "epoch": 15.191106804478897, "grad_norm": 0.41015708446502686, "learning_rate": 0.0003723885822000807, "loss": 3.0223, "step": 56450 }, { "epoch": 15.204565030146426, "grad_norm": 0.4246147572994232, "learning_rate": 0.000372186616399623, "loss": 3.0406, "step": 56500 }, { "epoch": 15.218023255813954, "grad_norm": 0.4137839376926422, "learning_rate": 0.0003719846505991652, "loss": 3.038, "step": 56550 }, { "epoch": 15.231481481481481, "grad_norm": 0.3703172206878662, "learning_rate": 0.0003717826847987074, "loss": 3.043, "step": 56600 }, { "epoch": 15.24493970714901, "grad_norm": 0.3823303282260895, "learning_rate": 0.0003715807189982496, "loss": 3.0399, "step": 56650 }, { "epoch": 15.258397932816537, "grad_norm": 0.4032827317714691, "learning_rate": 0.00037137875319779185, "loss": 3.0372, "step": 56700 }, { "epoch": 15.271856158484065, "grad_norm": 0.3879244029521942, "learning_rate": 0.00037117678739733405, "loss": 3.0386, "step": 56750 }, { "epoch": 15.285314384151594, "grad_norm": 0.42025917768478394, "learning_rate": 0.00037097482159687624, "loss": 3.0495, "step": 56800 }, { "epoch": 15.298772609819121, "grad_norm": 0.4079301357269287, "learning_rate": 0.00037077285579641844, "loss": 3.0522, "step": 56850 }, { "epoch": 15.31223083548665, "grad_norm": 0.3735152781009674, "learning_rate": 0.00037057088999596063, "loss": 3.0518, "step": 56900 }, { "epoch": 15.325689061154177, "grad_norm": 0.3667006492614746, "learning_rate": 0.0003703689241955029, "loss": 3.0653, "step": 56950 }, { "epoch": 15.339147286821705, "grad_norm": 0.3790123760700226, "learning_rate": 0.0003701669583950451, "loss": 3.055, "step": 57000 }, { "epoch": 15.339147286821705, "eval_accuracy": 0.3934124369359327, "eval_loss": 3.2985215187072754, "eval_runtime": 53.2792, "eval_samples_per_second": 338.068, "eval_steps_per_second": 21.134, "step": 57000 }, { "epoch": 15.352605512489234, "grad_norm": 0.39140379428863525, "learning_rate": 0.0003699649925945873, "loss": 3.0502, "step": 57050 }, { "epoch": 15.366063738156761, "grad_norm": 0.36787888407707214, "learning_rate": 0.0003697630267941295, "loss": 3.0509, "step": 57100 }, { "epoch": 15.37952196382429, "grad_norm": 0.398590624332428, "learning_rate": 0.00036956106099367167, "loss": 3.0607, "step": 57150 }, { "epoch": 15.392980189491817, "grad_norm": 0.40813106298446655, "learning_rate": 0.0003693590951932139, "loss": 3.0614, "step": 57200 }, { "epoch": 15.406438415159345, "grad_norm": 0.4021749496459961, "learning_rate": 0.0003691571293927561, "loss": 3.072, "step": 57250 }, { "epoch": 15.419896640826874, "grad_norm": 0.4174582064151764, "learning_rate": 0.0003689551635922983, "loss": 3.0584, "step": 57300 }, { "epoch": 15.433354866494401, "grad_norm": 0.3853360712528229, "learning_rate": 0.00036875319779184056, "loss": 3.0664, "step": 57350 }, { "epoch": 15.44681309216193, "grad_norm": 0.39959678053855896, "learning_rate": 0.0003685512319913828, "loss": 3.0574, "step": 57400 }, { "epoch": 15.460271317829458, "grad_norm": 0.3768465220928192, "learning_rate": 0.000368349266190925, "loss": 3.064, "step": 57450 }, { "epoch": 15.473729543496985, "grad_norm": 0.413850873708725, "learning_rate": 0.0003681473003904672, "loss": 3.058, "step": 57500 }, { "epoch": 15.487187769164514, "grad_norm": 0.3981676697731018, "learning_rate": 0.0003679453345900094, "loss": 3.0743, "step": 57550 }, { "epoch": 15.500645994832041, "grad_norm": 0.3920087516307831, "learning_rate": 0.0003677433687895516, "loss": 3.0711, "step": 57600 }, { "epoch": 15.514104220499568, "grad_norm": 0.390938937664032, "learning_rate": 0.00036754140298909385, "loss": 3.0714, "step": 57650 }, { "epoch": 15.527562446167098, "grad_norm": 0.37027981877326965, "learning_rate": 0.00036733943718863604, "loss": 3.0682, "step": 57700 }, { "epoch": 15.541020671834625, "grad_norm": 0.3791489899158478, "learning_rate": 0.00036713747138817824, "loss": 3.0739, "step": 57750 }, { "epoch": 15.554478897502154, "grad_norm": 0.4081215560436249, "learning_rate": 0.00036693550558772044, "loss": 3.0656, "step": 57800 }, { "epoch": 15.567937123169681, "grad_norm": 0.41583451628685, "learning_rate": 0.00036673353978726263, "loss": 3.0664, "step": 57850 }, { "epoch": 15.581395348837209, "grad_norm": 0.37946993112564087, "learning_rate": 0.0003665315739868049, "loss": 3.0627, "step": 57900 }, { "epoch": 15.594853574504738, "grad_norm": 0.37393489480018616, "learning_rate": 0.0003663296081863471, "loss": 3.067, "step": 57950 }, { "epoch": 15.608311800172265, "grad_norm": 0.3997141122817993, "learning_rate": 0.0003661276423858893, "loss": 3.0719, "step": 58000 }, { "epoch": 15.608311800172265, "eval_accuracy": 0.3940487815689736, "eval_loss": 3.291177749633789, "eval_runtime": 53.087, "eval_samples_per_second": 339.292, "eval_steps_per_second": 21.21, "step": 58000 }, { "epoch": 15.621770025839794, "grad_norm": 0.3671659827232361, "learning_rate": 0.00036592567658543147, "loss": 3.0725, "step": 58050 }, { "epoch": 15.635228251507321, "grad_norm": 0.3693672716617584, "learning_rate": 0.00036572371078497367, "loss": 3.0773, "step": 58100 }, { "epoch": 15.648686477174849, "grad_norm": 0.37850597500801086, "learning_rate": 0.0003655217449845159, "loss": 3.0868, "step": 58150 }, { "epoch": 15.662144702842378, "grad_norm": 0.3677229583263397, "learning_rate": 0.0003653197791840581, "loss": 3.0793, "step": 58200 }, { "epoch": 15.675602928509905, "grad_norm": 0.3778005838394165, "learning_rate": 0.00036511781338360036, "loss": 3.0812, "step": 58250 }, { "epoch": 15.689061154177434, "grad_norm": 0.38959360122680664, "learning_rate": 0.00036491584758314256, "loss": 3.0832, "step": 58300 }, { "epoch": 15.702519379844961, "grad_norm": 0.3569706082344055, "learning_rate": 0.0003647138817826848, "loss": 3.0764, "step": 58350 }, { "epoch": 15.715977605512489, "grad_norm": 0.392900675535202, "learning_rate": 0.000364511915982227, "loss": 3.0782, "step": 58400 }, { "epoch": 15.729435831180018, "grad_norm": 0.3854053318500519, "learning_rate": 0.0003643099501817692, "loss": 3.0711, "step": 58450 }, { "epoch": 15.742894056847545, "grad_norm": 0.42321139574050903, "learning_rate": 0.0003641079843813114, "loss": 3.0864, "step": 58500 }, { "epoch": 15.756352282515074, "grad_norm": 0.3787823021411896, "learning_rate": 0.0003639060185808536, "loss": 3.0788, "step": 58550 }, { "epoch": 15.769810508182601, "grad_norm": 0.4046003818511963, "learning_rate": 0.00036370405278039584, "loss": 3.0783, "step": 58600 }, { "epoch": 15.783268733850129, "grad_norm": 0.3918498754501343, "learning_rate": 0.00036350208697993804, "loss": 3.0908, "step": 58650 }, { "epoch": 15.796726959517658, "grad_norm": 0.38387343287467957, "learning_rate": 0.00036330012117948024, "loss": 3.0864, "step": 58700 }, { "epoch": 15.810185185185185, "grad_norm": 0.35123082995414734, "learning_rate": 0.00036309815537902243, "loss": 3.0814, "step": 58750 }, { "epoch": 15.823643410852712, "grad_norm": 0.3968450129032135, "learning_rate": 0.0003628961895785647, "loss": 3.0692, "step": 58800 }, { "epoch": 15.837101636520241, "grad_norm": 0.3727889358997345, "learning_rate": 0.0003626942237781069, "loss": 3.0876, "step": 58850 }, { "epoch": 15.850559862187769, "grad_norm": 0.363513708114624, "learning_rate": 0.0003624922579776491, "loss": 3.0858, "step": 58900 }, { "epoch": 15.864018087855298, "grad_norm": 0.4061371386051178, "learning_rate": 0.00036229029217719127, "loss": 3.0875, "step": 58950 }, { "epoch": 15.877476313522825, "grad_norm": 0.39571666717529297, "learning_rate": 0.00036208832637673347, "loss": 3.0866, "step": 59000 }, { "epoch": 15.877476313522825, "eval_accuracy": 0.3941058211681449, "eval_loss": 3.2855560779571533, "eval_runtime": 53.4013, "eval_samples_per_second": 337.295, "eval_steps_per_second": 21.086, "step": 59000 }, { "epoch": 15.890934539190352, "grad_norm": 0.3872230350971222, "learning_rate": 0.0003618863605762757, "loss": 3.085, "step": 59050 }, { "epoch": 15.904392764857882, "grad_norm": 0.3803175687789917, "learning_rate": 0.0003616843947758179, "loss": 3.0744, "step": 59100 }, { "epoch": 15.917850990525409, "grad_norm": 0.3779771029949188, "learning_rate": 0.00036148242897536016, "loss": 3.0824, "step": 59150 }, { "epoch": 15.931309216192938, "grad_norm": 0.37976348400115967, "learning_rate": 0.00036128046317490236, "loss": 3.0853, "step": 59200 }, { "epoch": 15.944767441860465, "grad_norm": 0.37095963954925537, "learning_rate": 0.0003610784973744446, "loss": 3.0841, "step": 59250 }, { "epoch": 15.958225667527993, "grad_norm": 0.3925478458404541, "learning_rate": 0.0003608765315739868, "loss": 3.0902, "step": 59300 }, { "epoch": 15.971683893195522, "grad_norm": 0.3674505650997162, "learning_rate": 0.000360674565773529, "loss": 3.0851, "step": 59350 }, { "epoch": 15.985142118863049, "grad_norm": 0.3850049674510956, "learning_rate": 0.0003604725999730712, "loss": 3.0937, "step": 59400 }, { "epoch": 15.998600344530576, "grad_norm": 0.3854713439941406, "learning_rate": 0.0003602706341726134, "loss": 3.0876, "step": 59450 }, { "epoch": 16.011843238587424, "grad_norm": 0.39652571082115173, "learning_rate": 0.00036006866837215564, "loss": 3.0026, "step": 59500 }, { "epoch": 16.02530146425495, "grad_norm": 0.39200231432914734, "learning_rate": 0.00035986670257169784, "loss": 2.9887, "step": 59550 }, { "epoch": 16.03875968992248, "grad_norm": 0.3969775438308716, "learning_rate": 0.00035966473677124004, "loss": 2.9961, "step": 59600 }, { "epoch": 16.05221791559001, "grad_norm": 0.3958861231803894, "learning_rate": 0.00035946277097078223, "loss": 3.0127, "step": 59650 }, { "epoch": 16.065676141257537, "grad_norm": 0.4136175215244293, "learning_rate": 0.00035926080517032443, "loss": 2.995, "step": 59700 }, { "epoch": 16.079134366925064, "grad_norm": 0.39139825105667114, "learning_rate": 0.0003590588393698667, "loss": 3.0106, "step": 59750 }, { "epoch": 16.09259259259259, "grad_norm": 0.3823096454143524, "learning_rate": 0.0003588568735694089, "loss": 3.0173, "step": 59800 }, { "epoch": 16.10605081826012, "grad_norm": 0.4026496112346649, "learning_rate": 0.00035865490776895107, "loss": 3.0072, "step": 59850 }, { "epoch": 16.11950904392765, "grad_norm": 0.37121763825416565, "learning_rate": 0.00035845294196849327, "loss": 3.0127, "step": 59900 }, { "epoch": 16.132967269595177, "grad_norm": 0.3773496150970459, "learning_rate": 0.00035825097616803546, "loss": 3.0064, "step": 59950 }, { "epoch": 16.146425495262704, "grad_norm": 0.40183204412460327, "learning_rate": 0.0003580490103675777, "loss": 3.0163, "step": 60000 }, { "epoch": 16.146425495262704, "eval_accuracy": 0.3936537416021413, "eval_loss": 3.2996578216552734, "eval_runtime": 53.3516, "eval_samples_per_second": 337.609, "eval_steps_per_second": 21.105, "step": 60000 }, { "epoch": 16.15988372093023, "grad_norm": 0.41072332859039307, "learning_rate": 0.00035784704456711996, "loss": 3.0291, "step": 60050 }, { "epoch": 16.17334194659776, "grad_norm": 0.3828774094581604, "learning_rate": 0.00035764507876666216, "loss": 3.0217, "step": 60100 }, { "epoch": 16.18680017226529, "grad_norm": 0.37735939025878906, "learning_rate": 0.00035744311296620436, "loss": 3.0309, "step": 60150 }, { "epoch": 16.200258397932817, "grad_norm": 0.41258010268211365, "learning_rate": 0.0003572411471657466, "loss": 3.022, "step": 60200 }, { "epoch": 16.213716623600344, "grad_norm": 0.426175057888031, "learning_rate": 0.0003570391813652888, "loss": 3.0247, "step": 60250 }, { "epoch": 16.227174849267872, "grad_norm": 0.3841751217842102, "learning_rate": 0.000356837215564831, "loss": 3.0353, "step": 60300 }, { "epoch": 16.2406330749354, "grad_norm": 0.3686473071575165, "learning_rate": 0.0003566352497643732, "loss": 3.0349, "step": 60350 }, { "epoch": 16.25409130060293, "grad_norm": 0.39995282888412476, "learning_rate": 0.0003564332839639154, "loss": 3.0311, "step": 60400 }, { "epoch": 16.267549526270457, "grad_norm": 0.3944709002971649, "learning_rate": 0.00035623131816345764, "loss": 3.0363, "step": 60450 }, { "epoch": 16.281007751937985, "grad_norm": 0.42645901441574097, "learning_rate": 0.00035602935236299984, "loss": 3.0346, "step": 60500 }, { "epoch": 16.294465977605512, "grad_norm": 0.3777204751968384, "learning_rate": 0.00035582738656254203, "loss": 3.0379, "step": 60550 }, { "epoch": 16.30792420327304, "grad_norm": 0.43480342626571655, "learning_rate": 0.00035562542076208423, "loss": 3.0456, "step": 60600 }, { "epoch": 16.32138242894057, "grad_norm": 0.41234782338142395, "learning_rate": 0.0003554234549616264, "loss": 3.0441, "step": 60650 }, { "epoch": 16.334840654608097, "grad_norm": 0.4231054484844208, "learning_rate": 0.0003552214891611687, "loss": 3.0439, "step": 60700 }, { "epoch": 16.348298880275625, "grad_norm": 0.40122222900390625, "learning_rate": 0.00035501952336071087, "loss": 3.043, "step": 60750 }, { "epoch": 16.361757105943152, "grad_norm": 0.39891695976257324, "learning_rate": 0.00035481755756025307, "loss": 3.0397, "step": 60800 }, { "epoch": 16.37521533161068, "grad_norm": 0.40936100482940674, "learning_rate": 0.00035461559175979526, "loss": 3.0488, "step": 60850 }, { "epoch": 16.38867355727821, "grad_norm": 0.4073893129825592, "learning_rate": 0.0003544136259593375, "loss": 3.0487, "step": 60900 }, { "epoch": 16.402131782945737, "grad_norm": 0.3983805477619171, "learning_rate": 0.00035421166015887977, "loss": 3.0557, "step": 60950 }, { "epoch": 16.415590008613265, "grad_norm": 0.39748522639274597, "learning_rate": 0.00035400969435842196, "loss": 3.0327, "step": 61000 }, { "epoch": 16.415590008613265, "eval_accuracy": 0.3939599084411219, "eval_loss": 3.2969777584075928, "eval_runtime": 53.184, "eval_samples_per_second": 338.673, "eval_steps_per_second": 21.172, "step": 61000 }, { "epoch": 16.429048234280792, "grad_norm": 0.4088171720504761, "learning_rate": 0.00035380772855796416, "loss": 3.0601, "step": 61050 }, { "epoch": 16.44250645994832, "grad_norm": 0.3807612359523773, "learning_rate": 0.0003536057627575064, "loss": 3.0543, "step": 61100 }, { "epoch": 16.45596468561585, "grad_norm": 0.4410359859466553, "learning_rate": 0.0003534037969570486, "loss": 3.0499, "step": 61150 }, { "epoch": 16.469422911283377, "grad_norm": 0.40752777457237244, "learning_rate": 0.0003532018311565908, "loss": 3.0546, "step": 61200 }, { "epoch": 16.482881136950905, "grad_norm": 0.38025355339050293, "learning_rate": 0.000352999865356133, "loss": 3.0511, "step": 61250 }, { "epoch": 16.496339362618432, "grad_norm": 0.38205963373184204, "learning_rate": 0.0003527978995556752, "loss": 3.0504, "step": 61300 }, { "epoch": 16.50979758828596, "grad_norm": 0.41273659467697144, "learning_rate": 0.00035259593375521744, "loss": 3.0582, "step": 61350 }, { "epoch": 16.52325581395349, "grad_norm": 0.43403419852256775, "learning_rate": 0.00035239396795475964, "loss": 3.0558, "step": 61400 }, { "epoch": 16.536714039621017, "grad_norm": 0.4052940309047699, "learning_rate": 0.00035219200215430183, "loss": 3.064, "step": 61450 }, { "epoch": 16.550172265288545, "grad_norm": 0.4491784870624542, "learning_rate": 0.00035199003635384403, "loss": 3.0662, "step": 61500 }, { "epoch": 16.563630490956072, "grad_norm": 0.40795519948005676, "learning_rate": 0.0003517880705533862, "loss": 3.0541, "step": 61550 }, { "epoch": 16.5770887166236, "grad_norm": 0.4100624620914459, "learning_rate": 0.0003515861047529285, "loss": 3.0603, "step": 61600 }, { "epoch": 16.590546942291127, "grad_norm": 0.3613821864128113, "learning_rate": 0.0003513841389524707, "loss": 3.052, "step": 61650 }, { "epoch": 16.604005167958658, "grad_norm": 0.399851530790329, "learning_rate": 0.00035118217315201287, "loss": 3.0598, "step": 61700 }, { "epoch": 16.617463393626185, "grad_norm": 0.37738752365112305, "learning_rate": 0.00035098020735155507, "loss": 3.0629, "step": 61750 }, { "epoch": 16.630921619293712, "grad_norm": 0.4187788665294647, "learning_rate": 0.00035077824155109737, "loss": 3.0551, "step": 61800 }, { "epoch": 16.64437984496124, "grad_norm": 0.389838308095932, "learning_rate": 0.00035057627575063957, "loss": 3.0685, "step": 61850 }, { "epoch": 16.657838070628767, "grad_norm": 0.41789937019348145, "learning_rate": 0.00035037430995018176, "loss": 3.0572, "step": 61900 }, { "epoch": 16.671296296296298, "grad_norm": 0.38143324851989746, "learning_rate": 0.00035017234414972396, "loss": 3.0631, "step": 61950 }, { "epoch": 16.684754521963825, "grad_norm": 0.3955875039100647, "learning_rate": 0.00034997037834926615, "loss": 3.0601, "step": 62000 }, { "epoch": 16.684754521963825, "eval_accuracy": 0.39469718600298215, "eval_loss": 3.2880349159240723, "eval_runtime": 53.1791, "eval_samples_per_second": 338.704, "eval_steps_per_second": 21.174, "step": 62000 }, { "epoch": 16.698212747631352, "grad_norm": 0.3697275221347809, "learning_rate": 0.0003497684125488084, "loss": 3.0608, "step": 62050 }, { "epoch": 16.71167097329888, "grad_norm": 0.39858055114746094, "learning_rate": 0.0003495664467483506, "loss": 3.0612, "step": 62100 }, { "epoch": 16.725129198966407, "grad_norm": 0.4065309762954712, "learning_rate": 0.0003493644809478928, "loss": 3.0739, "step": 62150 }, { "epoch": 16.738587424633938, "grad_norm": 0.39412033557891846, "learning_rate": 0.000349162515147435, "loss": 3.063, "step": 62200 }, { "epoch": 16.752045650301465, "grad_norm": 0.37209275364875793, "learning_rate": 0.0003489605493469772, "loss": 3.0661, "step": 62250 }, { "epoch": 16.765503875968992, "grad_norm": 0.40857645869255066, "learning_rate": 0.00034875858354651944, "loss": 3.057, "step": 62300 }, { "epoch": 16.77896210163652, "grad_norm": 0.3995668888092041, "learning_rate": 0.00034855661774606164, "loss": 3.0631, "step": 62350 }, { "epoch": 16.792420327304047, "grad_norm": 0.4129258990287781, "learning_rate": 0.00034835465194560383, "loss": 3.0786, "step": 62400 }, { "epoch": 16.805878552971578, "grad_norm": 0.37398791313171387, "learning_rate": 0.00034815268614514603, "loss": 3.0683, "step": 62450 }, { "epoch": 16.819336778639105, "grad_norm": 0.3871100842952728, "learning_rate": 0.0003479507203446882, "loss": 3.0685, "step": 62500 }, { "epoch": 16.832795004306632, "grad_norm": 0.3890288472175598, "learning_rate": 0.0003477487545442305, "loss": 3.0733, "step": 62550 }, { "epoch": 16.84625322997416, "grad_norm": 0.4500114321708679, "learning_rate": 0.00034754678874377267, "loss": 3.0633, "step": 62600 }, { "epoch": 16.859711455641687, "grad_norm": 0.3752024173736572, "learning_rate": 0.00034734482294331487, "loss": 3.0744, "step": 62650 }, { "epoch": 16.873169681309218, "grad_norm": 0.3815058469772339, "learning_rate": 0.00034714285714285717, "loss": 3.0623, "step": 62700 }, { "epoch": 16.886627906976745, "grad_norm": 0.41180121898651123, "learning_rate": 0.00034694089134239937, "loss": 3.0753, "step": 62750 }, { "epoch": 16.900086132644272, "grad_norm": 0.40455207228660583, "learning_rate": 0.00034673892554194156, "loss": 3.0696, "step": 62800 }, { "epoch": 16.9135443583118, "grad_norm": 0.40730997920036316, "learning_rate": 0.00034653695974148376, "loss": 3.0663, "step": 62850 }, { "epoch": 16.927002583979327, "grad_norm": 0.4231722056865692, "learning_rate": 0.00034633499394102595, "loss": 3.0777, "step": 62900 }, { "epoch": 16.940460809646858, "grad_norm": 0.39433252811431885, "learning_rate": 0.0003461330281405682, "loss": 3.0782, "step": 62950 }, { "epoch": 16.953919035314385, "grad_norm": 0.38897091150283813, "learning_rate": 0.0003459310623401104, "loss": 3.0832, "step": 63000 }, { "epoch": 16.953919035314385, "eval_accuracy": 0.3949562001066477, "eval_loss": 3.282935857772827, "eval_runtime": 53.1939, "eval_samples_per_second": 338.61, "eval_steps_per_second": 21.168, "step": 63000 }, { "epoch": 16.967377260981912, "grad_norm": 0.3950527608394623, "learning_rate": 0.0003457290965396526, "loss": 3.0736, "step": 63050 }, { "epoch": 16.98083548664944, "grad_norm": 0.38114306330680847, "learning_rate": 0.0003455271307391948, "loss": 3.0815, "step": 63100 }, { "epoch": 16.994293712316967, "grad_norm": 0.3700971305370331, "learning_rate": 0.000345325164938737, "loss": 3.0773, "step": 63150 }, { "epoch": 17.007536606373815, "grad_norm": 0.4128997325897217, "learning_rate": 0.00034512319913827924, "loss": 3.0138, "step": 63200 }, { "epoch": 17.020994832041342, "grad_norm": 0.3955026865005493, "learning_rate": 0.00034492123333782144, "loss": 2.9982, "step": 63250 }, { "epoch": 17.034453057708873, "grad_norm": 0.38770145177841187, "learning_rate": 0.00034471926753736363, "loss": 2.9895, "step": 63300 }, { "epoch": 17.0479112833764, "grad_norm": 0.46743109822273254, "learning_rate": 0.00034451730173690583, "loss": 2.9831, "step": 63350 }, { "epoch": 17.061369509043928, "grad_norm": 0.3955764174461365, "learning_rate": 0.000344315335936448, "loss": 2.9904, "step": 63400 }, { "epoch": 17.074827734711455, "grad_norm": 0.36407607793807983, "learning_rate": 0.0003441133701359903, "loss": 2.9922, "step": 63450 }, { "epoch": 17.088285960378983, "grad_norm": 0.40364158153533936, "learning_rate": 0.00034391140433553247, "loss": 2.9953, "step": 63500 }, { "epoch": 17.101744186046513, "grad_norm": 0.4021804630756378, "learning_rate": 0.00034370943853507467, "loss": 2.9977, "step": 63550 }, { "epoch": 17.11520241171404, "grad_norm": 0.37996798753738403, "learning_rate": 0.0003435074727346169, "loss": 3.0054, "step": 63600 }, { "epoch": 17.128660637381568, "grad_norm": 0.4075641930103302, "learning_rate": 0.00034330550693415917, "loss": 3.0035, "step": 63650 }, { "epoch": 17.142118863049095, "grad_norm": 0.40062153339385986, "learning_rate": 0.00034310354113370136, "loss": 2.9982, "step": 63700 }, { "epoch": 17.155577088716623, "grad_norm": 0.4038476049900055, "learning_rate": 0.00034290157533324356, "loss": 3.0037, "step": 63750 }, { "epoch": 17.16903531438415, "grad_norm": 0.3905138373374939, "learning_rate": 0.00034269960953278576, "loss": 3.0206, "step": 63800 }, { "epoch": 17.18249354005168, "grad_norm": 0.4093771278858185, "learning_rate": 0.00034249764373232795, "loss": 3.0089, "step": 63850 }, { "epoch": 17.195951765719208, "grad_norm": 0.3983815014362335, "learning_rate": 0.0003422956779318702, "loss": 3.0094, "step": 63900 }, { "epoch": 17.209409991386735, "grad_norm": 0.39827215671539307, "learning_rate": 0.0003420937121314124, "loss": 3.0127, "step": 63950 }, { "epoch": 17.222868217054263, "grad_norm": 0.4104834794998169, "learning_rate": 0.0003418917463309546, "loss": 3.0158, "step": 64000 }, { "epoch": 17.222868217054263, "eval_accuracy": 0.39382040587857714, "eval_loss": 3.2985177040100098, "eval_runtime": 53.135, "eval_samples_per_second": 338.986, "eval_steps_per_second": 21.191, "step": 64000 }, { "epoch": 17.23632644272179, "grad_norm": 0.38256514072418213, "learning_rate": 0.0003416897805304968, "loss": 3.0228, "step": 64050 }, { "epoch": 17.24978466838932, "grad_norm": 0.38628092408180237, "learning_rate": 0.000341487814730039, "loss": 3.0113, "step": 64100 }, { "epoch": 17.263242894056848, "grad_norm": 0.3880211114883423, "learning_rate": 0.00034128584892958124, "loss": 3.0162, "step": 64150 }, { "epoch": 17.276701119724375, "grad_norm": 0.3892453610897064, "learning_rate": 0.00034108388312912343, "loss": 3.0153, "step": 64200 }, { "epoch": 17.290159345391903, "grad_norm": 0.3993564248085022, "learning_rate": 0.00034088191732866563, "loss": 3.0313, "step": 64250 }, { "epoch": 17.30361757105943, "grad_norm": 0.3808676600456238, "learning_rate": 0.0003406799515282078, "loss": 3.0332, "step": 64300 }, { "epoch": 17.31707579672696, "grad_norm": 0.39774927496910095, "learning_rate": 0.00034047798572775, "loss": 3.0197, "step": 64350 }, { "epoch": 17.330534022394488, "grad_norm": 0.4316226840019226, "learning_rate": 0.00034027601992729227, "loss": 3.0257, "step": 64400 }, { "epoch": 17.343992248062015, "grad_norm": 0.42357945442199707, "learning_rate": 0.00034007405412683447, "loss": 3.041, "step": 64450 }, { "epoch": 17.357450473729543, "grad_norm": 0.4115462899208069, "learning_rate": 0.0003398720883263767, "loss": 3.0264, "step": 64500 }, { "epoch": 17.37090869939707, "grad_norm": 0.3940788209438324, "learning_rate": 0.00033967012252591897, "loss": 3.0349, "step": 64550 }, { "epoch": 17.3843669250646, "grad_norm": 0.48263993859291077, "learning_rate": 0.00033946815672546116, "loss": 3.0464, "step": 64600 }, { "epoch": 17.397825150732128, "grad_norm": 0.41211599111557007, "learning_rate": 0.00033926619092500336, "loss": 3.0267, "step": 64650 }, { "epoch": 17.411283376399656, "grad_norm": 0.4162316918373108, "learning_rate": 0.00033906422512454556, "loss": 3.0242, "step": 64700 }, { "epoch": 17.424741602067183, "grad_norm": 0.4139719009399414, "learning_rate": 0.00033886225932408775, "loss": 3.038, "step": 64750 }, { "epoch": 17.43819982773471, "grad_norm": 0.38862109184265137, "learning_rate": 0.00033866029352363, "loss": 3.0433, "step": 64800 }, { "epoch": 17.45165805340224, "grad_norm": 0.4061868190765381, "learning_rate": 0.0003384583277231722, "loss": 3.0346, "step": 64850 }, { "epoch": 17.46511627906977, "grad_norm": 0.3974857032299042, "learning_rate": 0.0003382563619227144, "loss": 3.0427, "step": 64900 }, { "epoch": 17.478574504737296, "grad_norm": 0.37685614824295044, "learning_rate": 0.0003380543961222566, "loss": 3.0387, "step": 64950 }, { "epoch": 17.492032730404823, "grad_norm": 0.4107515215873718, "learning_rate": 0.0003378524303217988, "loss": 3.0447, "step": 65000 }, { "epoch": 17.492032730404823, "eval_accuracy": 0.3945491003388478, "eval_loss": 3.2916104793548584, "eval_runtime": 53.1618, "eval_samples_per_second": 338.814, "eval_steps_per_second": 21.181, "step": 65000 }, { "epoch": 17.50549095607235, "grad_norm": 0.4122573733329773, "learning_rate": 0.00033765046452134104, "loss": 3.04, "step": 65050 }, { "epoch": 17.51894918173988, "grad_norm": 0.37764185667037964, "learning_rate": 0.00033744849872088323, "loss": 3.04, "step": 65100 }, { "epoch": 17.53240740740741, "grad_norm": 0.39302051067352295, "learning_rate": 0.00033724653292042543, "loss": 3.0438, "step": 65150 }, { "epoch": 17.545865633074936, "grad_norm": 0.41048935055732727, "learning_rate": 0.0003370445671199676, "loss": 3.0469, "step": 65200 }, { "epoch": 17.559323858742463, "grad_norm": 0.4067849814891815, "learning_rate": 0.0003368426013195098, "loss": 3.0454, "step": 65250 }, { "epoch": 17.57278208440999, "grad_norm": 0.42745301127433777, "learning_rate": 0.00033664063551905207, "loss": 3.0354, "step": 65300 }, { "epoch": 17.58624031007752, "grad_norm": 0.40041837096214294, "learning_rate": 0.00033643866971859427, "loss": 3.0553, "step": 65350 }, { "epoch": 17.59969853574505, "grad_norm": 0.40067821741104126, "learning_rate": 0.0003362367039181365, "loss": 3.0344, "step": 65400 }, { "epoch": 17.613156761412576, "grad_norm": 0.41964805126190186, "learning_rate": 0.0003360347381176787, "loss": 3.0544, "step": 65450 }, { "epoch": 17.626614987080103, "grad_norm": 0.4025363028049469, "learning_rate": 0.00033583277231722097, "loss": 3.0454, "step": 65500 }, { "epoch": 17.64007321274763, "grad_norm": 0.4273861050605774, "learning_rate": 0.00033563080651676316, "loss": 3.0504, "step": 65550 }, { "epoch": 17.653531438415158, "grad_norm": 0.4114350378513336, "learning_rate": 0.00033542884071630536, "loss": 3.0515, "step": 65600 }, { "epoch": 17.66698966408269, "grad_norm": 0.38349467515945435, "learning_rate": 0.00033522687491584755, "loss": 3.0493, "step": 65650 }, { "epoch": 17.680447889750216, "grad_norm": 0.38878414034843445, "learning_rate": 0.00033502490911538975, "loss": 3.0462, "step": 65700 }, { "epoch": 17.693906115417743, "grad_norm": 0.39213553071022034, "learning_rate": 0.000334822943314932, "loss": 3.0496, "step": 65750 }, { "epoch": 17.70736434108527, "grad_norm": 0.40689966082572937, "learning_rate": 0.0003346209775144742, "loss": 3.0513, "step": 65800 }, { "epoch": 17.720822566752798, "grad_norm": 0.3763260543346405, "learning_rate": 0.0003344190117140164, "loss": 3.0563, "step": 65850 }, { "epoch": 17.73428079242033, "grad_norm": 0.4094148278236389, "learning_rate": 0.0003342170459135586, "loss": 3.0592, "step": 65900 }, { "epoch": 17.747739018087856, "grad_norm": 0.38038286566734314, "learning_rate": 0.0003340150801131008, "loss": 3.0458, "step": 65950 }, { "epoch": 17.761197243755383, "grad_norm": 0.40296798944473267, "learning_rate": 0.00033381311431264303, "loss": 3.0561, "step": 66000 }, { "epoch": 17.761197243755383, "eval_accuracy": 0.39518316338792187, "eval_loss": 3.2837977409362793, "eval_runtime": 53.1573, "eval_samples_per_second": 338.844, "eval_steps_per_second": 21.182, "step": 66000 }, { "epoch": 17.77465546942291, "grad_norm": 0.371357262134552, "learning_rate": 0.00033361114851218523, "loss": 3.0456, "step": 66050 }, { "epoch": 17.788113695090438, "grad_norm": 0.4011745750904083, "learning_rate": 0.0003334091827117274, "loss": 3.0548, "step": 66100 }, { "epoch": 17.80157192075797, "grad_norm": 0.3885413408279419, "learning_rate": 0.0003332072169112696, "loss": 3.0575, "step": 66150 }, { "epoch": 17.815030146425496, "grad_norm": 0.40944671630859375, "learning_rate": 0.0003330052511108118, "loss": 3.0611, "step": 66200 }, { "epoch": 17.828488372093023, "grad_norm": 0.38013756275177, "learning_rate": 0.0003328032853103541, "loss": 3.0549, "step": 66250 }, { "epoch": 17.84194659776055, "grad_norm": 0.395384281873703, "learning_rate": 0.0003326013195098963, "loss": 3.056, "step": 66300 }, { "epoch": 17.855404823428078, "grad_norm": 0.3956717550754547, "learning_rate": 0.0003323993537094385, "loss": 3.0669, "step": 66350 }, { "epoch": 17.86886304909561, "grad_norm": 0.402349591255188, "learning_rate": 0.00033219738790898077, "loss": 3.0615, "step": 66400 }, { "epoch": 17.882321274763136, "grad_norm": 0.4142409563064575, "learning_rate": 0.00033199542210852296, "loss": 3.0672, "step": 66450 }, { "epoch": 17.895779500430663, "grad_norm": 0.414682537317276, "learning_rate": 0.00033179345630806516, "loss": 3.0628, "step": 66500 }, { "epoch": 17.90923772609819, "grad_norm": 0.392534464597702, "learning_rate": 0.00033159149050760735, "loss": 3.0644, "step": 66550 }, { "epoch": 17.922695951765718, "grad_norm": 0.41097918152809143, "learning_rate": 0.00033138952470714955, "loss": 3.0686, "step": 66600 }, { "epoch": 17.93615417743325, "grad_norm": 0.407552570104599, "learning_rate": 0.0003311875589066918, "loss": 3.0715, "step": 66650 }, { "epoch": 17.949612403100776, "grad_norm": 0.38662219047546387, "learning_rate": 0.000330985593106234, "loss": 3.0591, "step": 66700 }, { "epoch": 17.963070628768303, "grad_norm": 0.41780713200569153, "learning_rate": 0.0003307836273057762, "loss": 3.058, "step": 66750 }, { "epoch": 17.97652885443583, "grad_norm": 0.3943421244621277, "learning_rate": 0.0003305816615053184, "loss": 3.064, "step": 66800 }, { "epoch": 17.989987080103358, "grad_norm": 0.37973111867904663, "learning_rate": 0.0003303796957048606, "loss": 3.065, "step": 66850 }, { "epoch": 18.003229974160206, "grad_norm": 0.42041251063346863, "learning_rate": 0.00033017772990440284, "loss": 3.0416, "step": 66900 }, { "epoch": 18.016688199827733, "grad_norm": 0.4095908999443054, "learning_rate": 0.00032997576410394503, "loss": 2.9723, "step": 66950 }, { "epoch": 18.030146425495264, "grad_norm": 0.43125247955322266, "learning_rate": 0.00032977379830348723, "loss": 2.9748, "step": 67000 }, { "epoch": 18.030146425495264, "eval_accuracy": 0.39475335642730897, "eval_loss": 3.293449878692627, "eval_runtime": 53.2, "eval_samples_per_second": 338.572, "eval_steps_per_second": 21.165, "step": 67000 }, { "epoch": 18.04360465116279, "grad_norm": 0.4348003566265106, "learning_rate": 0.0003295718325030294, "loss": 2.9726, "step": 67050 }, { "epoch": 18.05706287683032, "grad_norm": 0.4124617576599121, "learning_rate": 0.0003293698667025716, "loss": 2.9857, "step": 67100 }, { "epoch": 18.070521102497846, "grad_norm": 0.4146093726158142, "learning_rate": 0.0003291679009021139, "loss": 2.9821, "step": 67150 }, { "epoch": 18.083979328165373, "grad_norm": 0.41040223836898804, "learning_rate": 0.0003289659351016561, "loss": 2.9869, "step": 67200 }, { "epoch": 18.097437553832904, "grad_norm": 0.4050770401954651, "learning_rate": 0.0003287639693011983, "loss": 2.9874, "step": 67250 }, { "epoch": 18.11089577950043, "grad_norm": 0.39382511377334595, "learning_rate": 0.0003285620035007405, "loss": 3.0004, "step": 67300 }, { "epoch": 18.12435400516796, "grad_norm": 0.42360246181488037, "learning_rate": 0.00032836003770028276, "loss": 2.9892, "step": 67350 }, { "epoch": 18.137812230835486, "grad_norm": 0.46744874119758606, "learning_rate": 0.00032815807189982496, "loss": 2.9902, "step": 67400 }, { "epoch": 18.151270456503013, "grad_norm": 0.4184926450252533, "learning_rate": 0.00032795610609936715, "loss": 2.9957, "step": 67450 }, { "epoch": 18.164728682170544, "grad_norm": 0.4146018624305725, "learning_rate": 0.00032775414029890935, "loss": 2.9911, "step": 67500 }, { "epoch": 18.17818690783807, "grad_norm": 0.40711694955825806, "learning_rate": 0.00032755217449845155, "loss": 3.0022, "step": 67550 }, { "epoch": 18.1916451335056, "grad_norm": 0.3984438478946686, "learning_rate": 0.0003273502086979938, "loss": 3.0039, "step": 67600 }, { "epoch": 18.205103359173126, "grad_norm": 0.40462562441825867, "learning_rate": 0.000327148242897536, "loss": 2.9994, "step": 67650 }, { "epoch": 18.218561584840653, "grad_norm": 0.47321245074272156, "learning_rate": 0.0003269462770970782, "loss": 3.0075, "step": 67700 }, { "epoch": 18.232019810508184, "grad_norm": 0.4014117419719696, "learning_rate": 0.0003267443112966204, "loss": 3.0112, "step": 67750 }, { "epoch": 18.24547803617571, "grad_norm": 0.42581304907798767, "learning_rate": 0.0003265423454961626, "loss": 3.0134, "step": 67800 }, { "epoch": 18.25893626184324, "grad_norm": 0.4282774031162262, "learning_rate": 0.00032634037969570483, "loss": 3.0109, "step": 67850 }, { "epoch": 18.272394487510766, "grad_norm": 0.3891073167324066, "learning_rate": 0.00032613841389524703, "loss": 3.0121, "step": 67900 }, { "epoch": 18.285852713178294, "grad_norm": 0.449763685464859, "learning_rate": 0.0003259364480947892, "loss": 2.9989, "step": 67950 }, { "epoch": 18.29931093884582, "grad_norm": 0.4044979512691498, "learning_rate": 0.0003257344822943314, "loss": 3.004, "step": 68000 }, { "epoch": 18.29931093884582, "eval_accuracy": 0.3946264569000097, "eval_loss": 3.295100688934326, "eval_runtime": 53.5137, "eval_samples_per_second": 336.587, "eval_steps_per_second": 21.041, "step": 68000 }, { "epoch": 18.31276916451335, "grad_norm": 0.4079665243625641, "learning_rate": 0.0003255325164938737, "loss": 3.0164, "step": 68050 }, { "epoch": 18.32622739018088, "grad_norm": 0.3965129852294922, "learning_rate": 0.0003253305506934159, "loss": 3.0175, "step": 68100 }, { "epoch": 18.339685615848406, "grad_norm": 0.4107968509197235, "learning_rate": 0.0003251285848929581, "loss": 3.0203, "step": 68150 }, { "epoch": 18.353143841515934, "grad_norm": 0.4172326326370239, "learning_rate": 0.0003249266190925003, "loss": 3.0209, "step": 68200 }, { "epoch": 18.36660206718346, "grad_norm": 0.39596325159072876, "learning_rate": 0.0003247246532920425, "loss": 3.0203, "step": 68250 }, { "epoch": 18.38006029285099, "grad_norm": 0.42070481181144714, "learning_rate": 0.00032452268749158476, "loss": 3.0214, "step": 68300 }, { "epoch": 18.39351851851852, "grad_norm": 0.4123518466949463, "learning_rate": 0.00032432072169112696, "loss": 3.0302, "step": 68350 }, { "epoch": 18.406976744186046, "grad_norm": 0.46176663041114807, "learning_rate": 0.00032411875589066915, "loss": 3.0233, "step": 68400 }, { "epoch": 18.420434969853574, "grad_norm": 0.42506712675094604, "learning_rate": 0.00032391679009021135, "loss": 3.0217, "step": 68450 }, { "epoch": 18.4338931955211, "grad_norm": 0.38221848011016846, "learning_rate": 0.0003237148242897536, "loss": 3.0227, "step": 68500 }, { "epoch": 18.447351421188632, "grad_norm": 0.43396782875061035, "learning_rate": 0.0003235128584892958, "loss": 3.0201, "step": 68550 }, { "epoch": 18.46080964685616, "grad_norm": 0.43768152594566345, "learning_rate": 0.000323310892688838, "loss": 3.0195, "step": 68600 }, { "epoch": 18.474267872523686, "grad_norm": 0.40827372670173645, "learning_rate": 0.0003231089268883802, "loss": 3.0232, "step": 68650 }, { "epoch": 18.487726098191214, "grad_norm": 0.4054965078830719, "learning_rate": 0.0003229069610879224, "loss": 3.0244, "step": 68700 }, { "epoch": 18.50118432385874, "grad_norm": 0.40097054839134216, "learning_rate": 0.00032270499528746463, "loss": 3.0317, "step": 68750 }, { "epoch": 18.514642549526272, "grad_norm": 0.41187137365341187, "learning_rate": 0.00032250302948700683, "loss": 3.0275, "step": 68800 }, { "epoch": 18.5281007751938, "grad_norm": 0.42199191451072693, "learning_rate": 0.000322301063686549, "loss": 3.0334, "step": 68850 }, { "epoch": 18.541559000861326, "grad_norm": 0.38072946667671204, "learning_rate": 0.0003220990978860912, "loss": 3.0353, "step": 68900 }, { "epoch": 18.555017226528854, "grad_norm": 0.4191347360610962, "learning_rate": 0.0003218971320856335, "loss": 3.0401, "step": 68950 }, { "epoch": 18.56847545219638, "grad_norm": 0.413581907749176, "learning_rate": 0.0003216951662851757, "loss": 3.0315, "step": 69000 }, { "epoch": 18.56847545219638, "eval_accuracy": 0.3948620032828734, "eval_loss": 3.2892377376556396, "eval_runtime": 53.2832, "eval_samples_per_second": 338.043, "eval_steps_per_second": 21.132, "step": 69000 }, { "epoch": 18.581933677863912, "grad_norm": 0.431535005569458, "learning_rate": 0.0003214932004847179, "loss": 3.0412, "step": 69050 }, { "epoch": 18.59539190353144, "grad_norm": 0.4194125831127167, "learning_rate": 0.0003212912346842601, "loss": 3.0391, "step": 69100 }, { "epoch": 18.608850129198967, "grad_norm": 0.4209112524986267, "learning_rate": 0.0003210892688838023, "loss": 3.0389, "step": 69150 }, { "epoch": 18.622308354866494, "grad_norm": 0.41800597310066223, "learning_rate": 0.00032088730308334456, "loss": 3.0292, "step": 69200 }, { "epoch": 18.63576658053402, "grad_norm": 0.39001700282096863, "learning_rate": 0.00032068533728288676, "loss": 3.0326, "step": 69250 }, { "epoch": 18.649224806201552, "grad_norm": 0.38095352053642273, "learning_rate": 0.00032048337148242895, "loss": 3.0305, "step": 69300 }, { "epoch": 18.66268303186908, "grad_norm": 0.40673938393592834, "learning_rate": 0.00032028140568197115, "loss": 3.0367, "step": 69350 }, { "epoch": 18.676141257536607, "grad_norm": 0.40681397914886475, "learning_rate": 0.00032007943988151334, "loss": 3.0247, "step": 69400 }, { "epoch": 18.689599483204134, "grad_norm": 0.40793168544769287, "learning_rate": 0.0003198774740810556, "loss": 3.0374, "step": 69450 }, { "epoch": 18.70305770887166, "grad_norm": 0.4070163667201996, "learning_rate": 0.0003196755082805978, "loss": 3.0361, "step": 69500 }, { "epoch": 18.716515934539192, "grad_norm": 0.40486323833465576, "learning_rate": 0.00031947354248014, "loss": 3.0403, "step": 69550 }, { "epoch": 18.72997416020672, "grad_norm": 0.4136981666088104, "learning_rate": 0.0003192715766796822, "loss": 3.0447, "step": 69600 }, { "epoch": 18.743432385874247, "grad_norm": 0.41730913519859314, "learning_rate": 0.0003190696108792244, "loss": 3.0411, "step": 69650 }, { "epoch": 18.756890611541774, "grad_norm": 0.4055947959423065, "learning_rate": 0.00031886764507876663, "loss": 3.0463, "step": 69700 }, { "epoch": 18.7703488372093, "grad_norm": 0.4054621756076813, "learning_rate": 0.0003186656792783088, "loss": 3.0443, "step": 69750 }, { "epoch": 18.783807062876832, "grad_norm": 0.42214059829711914, "learning_rate": 0.000318463713477851, "loss": 3.0503, "step": 69800 }, { "epoch": 18.79726528854436, "grad_norm": 0.43466436862945557, "learning_rate": 0.00031826174767739327, "loss": 3.0506, "step": 69850 }, { "epoch": 18.810723514211887, "grad_norm": 0.3966998755931854, "learning_rate": 0.0003180597818769355, "loss": 3.0378, "step": 69900 }, { "epoch": 18.824181739879414, "grad_norm": 0.422006219625473, "learning_rate": 0.0003178578160764777, "loss": 3.0567, "step": 69950 }, { "epoch": 18.83763996554694, "grad_norm": 0.4191000759601593, "learning_rate": 0.0003176558502760199, "loss": 3.0424, "step": 70000 }, { "epoch": 18.83763996554694, "eval_accuracy": 0.3954330511557201, "eval_loss": 3.278806209564209, "eval_runtime": 53.1978, "eval_samples_per_second": 338.585, "eval_steps_per_second": 21.166, "step": 70000 }, { "epoch": 18.85109819121447, "grad_norm": 0.4105396866798401, "learning_rate": 0.0003174538844755621, "loss": 3.0357, "step": 70050 }, { "epoch": 18.864556416882, "grad_norm": 0.4191843271255493, "learning_rate": 0.0003172519186751043, "loss": 3.0515, "step": 70100 }, { "epoch": 18.878014642549527, "grad_norm": 0.3951943516731262, "learning_rate": 0.00031704995287464656, "loss": 3.0501, "step": 70150 }, { "epoch": 18.891472868217054, "grad_norm": 0.3969205915927887, "learning_rate": 0.00031684798707418875, "loss": 3.0475, "step": 70200 }, { "epoch": 18.90493109388458, "grad_norm": 0.4015067219734192, "learning_rate": 0.00031664602127373095, "loss": 3.0622, "step": 70250 }, { "epoch": 18.91838931955211, "grad_norm": 0.39716896414756775, "learning_rate": 0.00031644405547327315, "loss": 3.0546, "step": 70300 }, { "epoch": 18.93184754521964, "grad_norm": 0.42040297389030457, "learning_rate": 0.00031624208967281534, "loss": 3.0566, "step": 70350 }, { "epoch": 18.945305770887167, "grad_norm": 0.397980272769928, "learning_rate": 0.0003160401238723576, "loss": 3.0437, "step": 70400 }, { "epoch": 18.958763996554694, "grad_norm": 0.42787206172943115, "learning_rate": 0.0003158381580718998, "loss": 3.0559, "step": 70450 }, { "epoch": 18.97222222222222, "grad_norm": 0.4149838089942932, "learning_rate": 0.000315636192271442, "loss": 3.0584, "step": 70500 }, { "epoch": 18.98568044788975, "grad_norm": 0.4251541197299957, "learning_rate": 0.0003154342264709842, "loss": 3.0621, "step": 70550 }, { "epoch": 18.99913867355728, "grad_norm": 0.41713324189186096, "learning_rate": 0.00031523226067052643, "loss": 3.0395, "step": 70600 }, { "epoch": 19.012381567614124, "grad_norm": 0.43764570355415344, "learning_rate": 0.0003150302948700686, "loss": 2.9625, "step": 70650 }, { "epoch": 19.025839793281655, "grad_norm": 0.4346773326396942, "learning_rate": 0.0003148283290696108, "loss": 2.9584, "step": 70700 }, { "epoch": 19.039298018949182, "grad_norm": 0.4352344274520874, "learning_rate": 0.0003146263632691531, "loss": 2.9643, "step": 70750 }, { "epoch": 19.05275624461671, "grad_norm": 0.4444960951805115, "learning_rate": 0.0003144243974686953, "loss": 2.9676, "step": 70800 }, { "epoch": 19.066214470284237, "grad_norm": 0.41851648688316345, "learning_rate": 0.0003142224316682375, "loss": 2.9715, "step": 70850 }, { "epoch": 19.079672695951764, "grad_norm": 0.4398450553417206, "learning_rate": 0.0003140204658677797, "loss": 2.9952, "step": 70900 }, { "epoch": 19.093130921619295, "grad_norm": 0.39954739809036255, "learning_rate": 0.0003138185000673219, "loss": 2.9676, "step": 70950 }, { "epoch": 19.106589147286822, "grad_norm": 0.42698994278907776, "learning_rate": 0.0003136165342668641, "loss": 2.9799, "step": 71000 }, { "epoch": 19.106589147286822, "eval_accuracy": 0.3946730664010468, "eval_loss": 3.2944045066833496, "eval_runtime": 53.1951, "eval_samples_per_second": 338.602, "eval_steps_per_second": 21.167, "step": 71000 }, { "epoch": 19.12004737295435, "grad_norm": 0.4176847040653229, "learning_rate": 0.00031341456846640636, "loss": 2.9882, "step": 71050 }, { "epoch": 19.133505598621877, "grad_norm": 0.43521398305892944, "learning_rate": 0.00031321260266594855, "loss": 2.9847, "step": 71100 }, { "epoch": 19.146963824289404, "grad_norm": 0.4214783012866974, "learning_rate": 0.00031301063686549075, "loss": 2.9982, "step": 71150 }, { "epoch": 19.160422049956935, "grad_norm": 0.4213625192642212, "learning_rate": 0.00031280867106503295, "loss": 2.9853, "step": 71200 }, { "epoch": 19.173880275624462, "grad_norm": 0.40806859731674194, "learning_rate": 0.00031260670526457514, "loss": 2.9855, "step": 71250 }, { "epoch": 19.18733850129199, "grad_norm": 0.4534761905670166, "learning_rate": 0.0003124047394641174, "loss": 2.9845, "step": 71300 }, { "epoch": 19.200796726959517, "grad_norm": 0.4072936773300171, "learning_rate": 0.0003122027736636596, "loss": 2.9865, "step": 71350 }, { "epoch": 19.214254952627044, "grad_norm": 0.40868863463401794, "learning_rate": 0.0003120008078632018, "loss": 2.9928, "step": 71400 }, { "epoch": 19.227713178294575, "grad_norm": 0.4284927546977997, "learning_rate": 0.000311798842062744, "loss": 3.0037, "step": 71450 }, { "epoch": 19.241171403962102, "grad_norm": 0.4234805405139923, "learning_rate": 0.0003115968762622862, "loss": 2.9812, "step": 71500 }, { "epoch": 19.25462962962963, "grad_norm": 0.40032583475112915, "learning_rate": 0.00031139491046182843, "loss": 3.0, "step": 71550 }, { "epoch": 19.268087855297157, "grad_norm": 0.4138863980770111, "learning_rate": 0.0003111929446613707, "loss": 2.9892, "step": 71600 }, { "epoch": 19.281546080964684, "grad_norm": 0.4290546476840973, "learning_rate": 0.0003109909788609129, "loss": 3.0028, "step": 71650 }, { "epoch": 19.295004306632215, "grad_norm": 0.4247094690799713, "learning_rate": 0.00031078901306045507, "loss": 2.9972, "step": 71700 }, { "epoch": 19.308462532299743, "grad_norm": 0.4409559965133667, "learning_rate": 0.0003105870472599973, "loss": 3.0053, "step": 71750 }, { "epoch": 19.32192075796727, "grad_norm": 0.4488597810268402, "learning_rate": 0.0003103850814595395, "loss": 2.9962, "step": 71800 }, { "epoch": 19.335378983634797, "grad_norm": 0.4389813244342804, "learning_rate": 0.0003101831156590817, "loss": 3.0027, "step": 71850 }, { "epoch": 19.348837209302324, "grad_norm": 0.40503010153770447, "learning_rate": 0.0003099811498586239, "loss": 3.0004, "step": 71900 }, { "epoch": 19.362295434969855, "grad_norm": 0.44304975867271423, "learning_rate": 0.0003097791840581661, "loss": 3.0097, "step": 71950 }, { "epoch": 19.375753660637383, "grad_norm": 0.4481963813304901, "learning_rate": 0.00030957721825770836, "loss": 2.999, "step": 72000 }, { "epoch": 19.375753660637383, "eval_accuracy": 0.3951246027327726, "eval_loss": 3.290699005126953, "eval_runtime": 53.1572, "eval_samples_per_second": 338.844, "eval_steps_per_second": 21.182, "step": 72000 }, { "epoch": 19.38921188630491, "grad_norm": 0.4432748258113861, "learning_rate": 0.00030937525245725055, "loss": 3.0115, "step": 72050 }, { "epoch": 19.402670111972437, "grad_norm": 0.4061640799045563, "learning_rate": 0.00030917328665679275, "loss": 3.0055, "step": 72100 }, { "epoch": 19.416128337639964, "grad_norm": 0.4457055926322937, "learning_rate": 0.00030897132085633494, "loss": 3.0143, "step": 72150 }, { "epoch": 19.429586563307495, "grad_norm": 0.44451671838760376, "learning_rate": 0.00030876935505587714, "loss": 3.0057, "step": 72200 }, { "epoch": 19.443044788975023, "grad_norm": 0.41940534114837646, "learning_rate": 0.0003085673892554194, "loss": 3.0211, "step": 72250 }, { "epoch": 19.45650301464255, "grad_norm": 0.40035170316696167, "learning_rate": 0.0003083654234549616, "loss": 3.0135, "step": 72300 }, { "epoch": 19.469961240310077, "grad_norm": 0.42814239859580994, "learning_rate": 0.0003081634576545038, "loss": 3.0146, "step": 72350 }, { "epoch": 19.483419465977605, "grad_norm": 0.42938923835754395, "learning_rate": 0.000307961491854046, "loss": 3.0184, "step": 72400 }, { "epoch": 19.496877691645132, "grad_norm": 0.40963107347488403, "learning_rate": 0.00030775952605358823, "loss": 3.0071, "step": 72450 }, { "epoch": 19.510335917312663, "grad_norm": 0.4201676845550537, "learning_rate": 0.0003075575602531305, "loss": 3.0049, "step": 72500 }, { "epoch": 19.52379414298019, "grad_norm": 0.40009865164756775, "learning_rate": 0.0003073555944526727, "loss": 3.0141, "step": 72550 }, { "epoch": 19.537252368647717, "grad_norm": 0.4132072925567627, "learning_rate": 0.00030715362865221487, "loss": 3.0243, "step": 72600 }, { "epoch": 19.550710594315245, "grad_norm": 0.4008041024208069, "learning_rate": 0.0003069516628517571, "loss": 3.0209, "step": 72650 }, { "epoch": 19.564168819982772, "grad_norm": 0.4381272792816162, "learning_rate": 0.0003067496970512993, "loss": 3.0309, "step": 72700 }, { "epoch": 19.577627045650303, "grad_norm": 0.4061357378959656, "learning_rate": 0.0003065477312508415, "loss": 3.0198, "step": 72750 }, { "epoch": 19.59108527131783, "grad_norm": 0.42047545313835144, "learning_rate": 0.0003063457654503837, "loss": 3.0306, "step": 72800 }, { "epoch": 19.604543496985357, "grad_norm": 0.411575049161911, "learning_rate": 0.0003061437996499259, "loss": 3.0281, "step": 72850 }, { "epoch": 19.618001722652885, "grad_norm": 0.42404329776763916, "learning_rate": 0.00030594183384946816, "loss": 3.0326, "step": 72900 }, { "epoch": 19.631459948320412, "grad_norm": 0.42757225036621094, "learning_rate": 0.00030573986804901035, "loss": 3.0259, "step": 72950 }, { "epoch": 19.644918173987943, "grad_norm": 0.3978443741798401, "learning_rate": 0.00030553790224855255, "loss": 3.0285, "step": 73000 }, { "epoch": 19.644918173987943, "eval_accuracy": 0.3953907875289055, "eval_loss": 3.284182548522949, "eval_runtime": 53.2956, "eval_samples_per_second": 337.964, "eval_steps_per_second": 21.127, "step": 73000 }, { "epoch": 19.65837639965547, "grad_norm": 0.3940031826496124, "learning_rate": 0.00030533593644809474, "loss": 3.0237, "step": 73050 }, { "epoch": 19.671834625322997, "grad_norm": 0.44861793518066406, "learning_rate": 0.00030513397064763694, "loss": 3.0246, "step": 73100 }, { "epoch": 19.685292850990525, "grad_norm": 0.42165327072143555, "learning_rate": 0.0003049320048471792, "loss": 3.0414, "step": 73150 }, { "epoch": 19.698751076658052, "grad_norm": 0.391513854265213, "learning_rate": 0.0003047300390467214, "loss": 3.0232, "step": 73200 }, { "epoch": 19.712209302325583, "grad_norm": 0.4060502052307129, "learning_rate": 0.0003045280732462636, "loss": 3.0291, "step": 73250 }, { "epoch": 19.72566752799311, "grad_norm": 0.39743009209632874, "learning_rate": 0.0003043261074458058, "loss": 3.0313, "step": 73300 }, { "epoch": 19.739125753660637, "grad_norm": 0.41127610206604004, "learning_rate": 0.000304124141645348, "loss": 3.0281, "step": 73350 }, { "epoch": 19.752583979328165, "grad_norm": 0.398449569940567, "learning_rate": 0.0003039221758448903, "loss": 3.0166, "step": 73400 }, { "epoch": 19.766042204995692, "grad_norm": 0.4216623604297638, "learning_rate": 0.0003037202100444325, "loss": 3.0423, "step": 73450 }, { "epoch": 19.779500430663223, "grad_norm": 0.4148426949977875, "learning_rate": 0.00030351824424397467, "loss": 3.0202, "step": 73500 }, { "epoch": 19.79295865633075, "grad_norm": 0.44049158692359924, "learning_rate": 0.00030331627844351687, "loss": 3.0349, "step": 73550 }, { "epoch": 19.806416881998278, "grad_norm": 0.4221228361129761, "learning_rate": 0.0003031143126430591, "loss": 3.0326, "step": 73600 }, { "epoch": 19.819875107665805, "grad_norm": 0.41475948691368103, "learning_rate": 0.0003029123468426013, "loss": 3.035, "step": 73650 }, { "epoch": 19.833333333333332, "grad_norm": 0.41580721735954285, "learning_rate": 0.0003027103810421435, "loss": 3.0412, "step": 73700 }, { "epoch": 19.846791559000863, "grad_norm": 0.4306314289569855, "learning_rate": 0.0003025084152416857, "loss": 3.0312, "step": 73750 }, { "epoch": 19.86024978466839, "grad_norm": 0.41465747356414795, "learning_rate": 0.0003023064494412279, "loss": 3.0449, "step": 73800 }, { "epoch": 19.873708010335918, "grad_norm": 0.427436888217926, "learning_rate": 0.00030210448364077015, "loss": 3.0337, "step": 73850 }, { "epoch": 19.887166236003445, "grad_norm": 0.4142983555793762, "learning_rate": 0.00030190251784031235, "loss": 3.0446, "step": 73900 }, { "epoch": 19.900624461670972, "grad_norm": 0.42197173833847046, "learning_rate": 0.00030170055203985454, "loss": 3.0457, "step": 73950 }, { "epoch": 19.9140826873385, "grad_norm": 0.41774898767471313, "learning_rate": 0.00030149858623939674, "loss": 3.0365, "step": 74000 }, { "epoch": 19.9140826873385, "eval_accuracy": 0.3958141843250401, "eval_loss": 3.279388427734375, "eval_runtime": 53.2046, "eval_samples_per_second": 338.542, "eval_steps_per_second": 21.164, "step": 74000 }, { "epoch": 19.92754091300603, "grad_norm": 0.47798389196395874, "learning_rate": 0.00030129662043893894, "loss": 3.0397, "step": 74050 }, { "epoch": 19.940999138673558, "grad_norm": 0.44100531935691833, "learning_rate": 0.0003010946546384812, "loss": 3.046, "step": 74100 }, { "epoch": 19.954457364341085, "grad_norm": 0.42226624488830566, "learning_rate": 0.0003008926888380234, "loss": 3.0389, "step": 74150 }, { "epoch": 19.967915590008612, "grad_norm": 0.3779127597808838, "learning_rate": 0.0003006907230375656, "loss": 3.0381, "step": 74200 }, { "epoch": 19.98137381567614, "grad_norm": 0.4114772379398346, "learning_rate": 0.0003004887572371078, "loss": 3.0382, "step": 74250 }, { "epoch": 19.99483204134367, "grad_norm": 0.4274559020996094, "learning_rate": 0.0003002867914366501, "loss": 3.0433, "step": 74300 }, { "epoch": 20.00807493540052, "grad_norm": 0.40736275911331177, "learning_rate": 0.0003000848256361923, "loss": 2.9855, "step": 74350 }, { "epoch": 20.021533161068046, "grad_norm": 0.4333912134170532, "learning_rate": 0.00029988285983573447, "loss": 2.939, "step": 74400 }, { "epoch": 20.034991386735573, "grad_norm": 0.42243754863739014, "learning_rate": 0.00029968089403527667, "loss": 2.9527, "step": 74450 }, { "epoch": 20.0484496124031, "grad_norm": 0.43641528487205505, "learning_rate": 0.0002994789282348189, "loss": 2.962, "step": 74500 }, { "epoch": 20.061907838070628, "grad_norm": 0.42667356133461, "learning_rate": 0.0002992769624343611, "loss": 2.9601, "step": 74550 }, { "epoch": 20.075366063738155, "grad_norm": 0.4108620285987854, "learning_rate": 0.0002990749966339033, "loss": 2.9652, "step": 74600 }, { "epoch": 20.088824289405686, "grad_norm": 0.4250992238521576, "learning_rate": 0.0002988730308334455, "loss": 2.9587, "step": 74650 }, { "epoch": 20.102282515073213, "grad_norm": 0.41763296723365784, "learning_rate": 0.0002986710650329877, "loss": 2.9691, "step": 74700 }, { "epoch": 20.11574074074074, "grad_norm": 0.4523560106754303, "learning_rate": 0.00029846909923252995, "loss": 2.971, "step": 74750 }, { "epoch": 20.129198966408268, "grad_norm": 0.4249483048915863, "learning_rate": 0.00029826713343207215, "loss": 2.9687, "step": 74800 }, { "epoch": 20.142657192075795, "grad_norm": 0.4228469133377075, "learning_rate": 0.00029806516763161435, "loss": 2.9815, "step": 74850 }, { "epoch": 20.156115417743326, "grad_norm": 0.4154616892337799, "learning_rate": 0.00029786320183115654, "loss": 2.9821, "step": 74900 }, { "epoch": 20.169573643410853, "grad_norm": 0.43550071120262146, "learning_rate": 0.0002976612360306988, "loss": 2.9639, "step": 74950 }, { "epoch": 20.18303186907838, "grad_norm": 0.415121853351593, "learning_rate": 0.000297459270230241, "loss": 2.979, "step": 75000 }, { "epoch": 20.18303186907838, "eval_accuracy": 0.39531125803063233, "eval_loss": 3.2909493446350098, "eval_runtime": 53.2389, "eval_samples_per_second": 338.324, "eval_steps_per_second": 21.15, "step": 75000 }, { "epoch": 20.196490094745908, "grad_norm": 0.4105370044708252, "learning_rate": 0.0002972573044297832, "loss": 2.9781, "step": 75050 }, { "epoch": 20.209948320413435, "grad_norm": 0.4527175724506378, "learning_rate": 0.00029705533862932543, "loss": 2.9852, "step": 75100 }, { "epoch": 20.223406546080966, "grad_norm": 0.40570807456970215, "learning_rate": 0.00029685337282886763, "loss": 2.9897, "step": 75150 }, { "epoch": 20.236864771748493, "grad_norm": 0.43341004848480225, "learning_rate": 0.0002966514070284098, "loss": 2.9861, "step": 75200 }, { "epoch": 20.25032299741602, "grad_norm": 0.44538331031799316, "learning_rate": 0.000296449441227952, "loss": 2.99, "step": 75250 }, { "epoch": 20.263781223083548, "grad_norm": 0.40199220180511475, "learning_rate": 0.0002962474754274942, "loss": 2.9836, "step": 75300 }, { "epoch": 20.277239448751075, "grad_norm": 0.4486052989959717, "learning_rate": 0.00029604550962703647, "loss": 2.9899, "step": 75350 }, { "epoch": 20.290697674418606, "grad_norm": 0.44760704040527344, "learning_rate": 0.00029584354382657867, "loss": 2.9887, "step": 75400 }, { "epoch": 20.304155900086133, "grad_norm": 0.42043086886405945, "learning_rate": 0.0002956415780261209, "loss": 2.9937, "step": 75450 }, { "epoch": 20.31761412575366, "grad_norm": 0.43127012252807617, "learning_rate": 0.0002954396122256631, "loss": 2.9947, "step": 75500 }, { "epoch": 20.331072351421188, "grad_norm": 0.4210168719291687, "learning_rate": 0.0002952376464252053, "loss": 2.9901, "step": 75550 }, { "epoch": 20.344530577088715, "grad_norm": 0.42815783619880676, "learning_rate": 0.0002950356806247475, "loss": 3.0006, "step": 75600 }, { "epoch": 20.357988802756246, "grad_norm": 0.4324301481246948, "learning_rate": 0.0002948337148242897, "loss": 2.9946, "step": 75650 }, { "epoch": 20.371447028423773, "grad_norm": 0.40492090582847595, "learning_rate": 0.00029463174902383195, "loss": 2.9957, "step": 75700 }, { "epoch": 20.3849052540913, "grad_norm": 0.41066285967826843, "learning_rate": 0.00029442978322337415, "loss": 2.9995, "step": 75750 }, { "epoch": 20.398363479758828, "grad_norm": 0.45168185234069824, "learning_rate": 0.00029422781742291634, "loss": 2.993, "step": 75800 }, { "epoch": 20.411821705426355, "grad_norm": 0.42091086506843567, "learning_rate": 0.0002940258516224586, "loss": 3.0148, "step": 75850 }, { "epoch": 20.425279931093886, "grad_norm": 0.4229574203491211, "learning_rate": 0.0002938238858220008, "loss": 2.996, "step": 75900 }, { "epoch": 20.438738156761413, "grad_norm": 0.4290357232093811, "learning_rate": 0.000293621920021543, "loss": 3.0044, "step": 75950 }, { "epoch": 20.45219638242894, "grad_norm": 0.4104298949241638, "learning_rate": 0.0002934199542210852, "loss": 3.0131, "step": 76000 }, { "epoch": 20.45219638242894, "eval_accuracy": 0.39542207782330807, "eval_loss": 3.2870841026306152, "eval_runtime": 53.2071, "eval_samples_per_second": 338.526, "eval_steps_per_second": 21.163, "step": 76000 }, { "epoch": 20.465654608096468, "grad_norm": 0.43837058544158936, "learning_rate": 0.00029321798842062743, "loss": 3.0129, "step": 76050 }, { "epoch": 20.479112833763995, "grad_norm": 0.4409915804862976, "learning_rate": 0.00029301602262016963, "loss": 2.9995, "step": 76100 }, { "epoch": 20.492571059431526, "grad_norm": 0.4528588056564331, "learning_rate": 0.0002928140568197118, "loss": 3.0076, "step": 76150 }, { "epoch": 20.506029285099054, "grad_norm": 0.41055387258529663, "learning_rate": 0.000292612091019254, "loss": 2.9977, "step": 76200 }, { "epoch": 20.51948751076658, "grad_norm": 0.42017340660095215, "learning_rate": 0.0002924101252187962, "loss": 3.0144, "step": 76250 }, { "epoch": 20.532945736434108, "grad_norm": 0.45756611227989197, "learning_rate": 0.00029220815941833847, "loss": 2.9964, "step": 76300 }, { "epoch": 20.546403962101635, "grad_norm": 0.42427578568458557, "learning_rate": 0.0002920061936178807, "loss": 3.0133, "step": 76350 }, { "epoch": 20.559862187769163, "grad_norm": 0.4466415345668793, "learning_rate": 0.0002918042278174229, "loss": 3.0102, "step": 76400 }, { "epoch": 20.573320413436694, "grad_norm": 0.41783392429351807, "learning_rate": 0.0002916022620169651, "loss": 3.0047, "step": 76450 }, { "epoch": 20.58677863910422, "grad_norm": 0.42117050290107727, "learning_rate": 0.0002914002962165073, "loss": 3.0129, "step": 76500 }, { "epoch": 20.600236864771748, "grad_norm": 0.4089644253253937, "learning_rate": 0.0002911983304160495, "loss": 3.0128, "step": 76550 }, { "epoch": 20.613695090439276, "grad_norm": 0.4382970333099365, "learning_rate": 0.00029099636461559175, "loss": 3.0158, "step": 76600 }, { "epoch": 20.627153316106803, "grad_norm": 0.40870681405067444, "learning_rate": 0.00029079439881513395, "loss": 3.0157, "step": 76650 }, { "epoch": 20.640611541774334, "grad_norm": 0.43205249309539795, "learning_rate": 0.00029059243301467614, "loss": 3.0142, "step": 76700 }, { "epoch": 20.65406976744186, "grad_norm": 0.41538992524147034, "learning_rate": 0.0002903904672142184, "loss": 3.0206, "step": 76750 }, { "epoch": 20.66752799310939, "grad_norm": 0.4724888801574707, "learning_rate": 0.0002901885014137606, "loss": 3.0077, "step": 76800 }, { "epoch": 20.680986218776916, "grad_norm": 0.4144582450389862, "learning_rate": 0.0002899865356133028, "loss": 3.0272, "step": 76850 }, { "epoch": 20.694444444444443, "grad_norm": 0.41652971506118774, "learning_rate": 0.000289784569812845, "loss": 3.0097, "step": 76900 }, { "epoch": 20.707902670111974, "grad_norm": 0.43358200788497925, "learning_rate": 0.00028958260401238723, "loss": 3.0187, "step": 76950 }, { "epoch": 20.7213608957795, "grad_norm": 0.41195759177207947, "learning_rate": 0.00028938063821192943, "loss": 3.0201, "step": 77000 }, { "epoch": 20.7213608957795, "eval_accuracy": 0.39608623605137344, "eval_loss": 3.2799088954925537, "eval_runtime": 53.2969, "eval_samples_per_second": 337.956, "eval_steps_per_second": 21.127, "step": 77000 }, { "epoch": 20.73481912144703, "grad_norm": 0.44272324442863464, "learning_rate": 0.0002891786724114716, "loss": 3.0161, "step": 77050 }, { "epoch": 20.748277347114556, "grad_norm": 0.424826979637146, "learning_rate": 0.0002889767066110138, "loss": 3.0197, "step": 77100 }, { "epoch": 20.761735572782083, "grad_norm": 0.4209699034690857, "learning_rate": 0.000288774740810556, "loss": 3.0277, "step": 77150 }, { "epoch": 20.775193798449614, "grad_norm": 0.41250577569007874, "learning_rate": 0.00028857277501009827, "loss": 3.0296, "step": 77200 }, { "epoch": 20.78865202411714, "grad_norm": 0.42251619696617126, "learning_rate": 0.00028837080920964046, "loss": 3.0234, "step": 77250 }, { "epoch": 20.80211024978467, "grad_norm": 0.4229857325553894, "learning_rate": 0.0002881688434091827, "loss": 3.0211, "step": 77300 }, { "epoch": 20.815568475452196, "grad_norm": 0.42098045349121094, "learning_rate": 0.0002879668776087249, "loss": 3.0308, "step": 77350 }, { "epoch": 20.829026701119723, "grad_norm": 0.4067869782447815, "learning_rate": 0.0002877649118082671, "loss": 3.0273, "step": 77400 }, { "epoch": 20.842484926787254, "grad_norm": 0.436855286359787, "learning_rate": 0.0002875629460078093, "loss": 3.0315, "step": 77450 }, { "epoch": 20.85594315245478, "grad_norm": 0.4486081004142761, "learning_rate": 0.0002873609802073515, "loss": 3.0115, "step": 77500 }, { "epoch": 20.86940137812231, "grad_norm": 0.40413883328437805, "learning_rate": 0.00028715901440689375, "loss": 3.0183, "step": 77550 }, { "epoch": 20.882859603789836, "grad_norm": 0.4487050771713257, "learning_rate": 0.00028695704860643594, "loss": 3.0208, "step": 77600 }, { "epoch": 20.896317829457363, "grad_norm": 0.43203026056289673, "learning_rate": 0.0002867550828059782, "loss": 3.0223, "step": 77650 }, { "epoch": 20.909776055124894, "grad_norm": 0.4328913986682892, "learning_rate": 0.0002865531170055204, "loss": 3.0271, "step": 77700 }, { "epoch": 20.92323428079242, "grad_norm": 0.39035165309906006, "learning_rate": 0.0002863511512050626, "loss": 3.031, "step": 77750 }, { "epoch": 20.93669250645995, "grad_norm": 0.404269814491272, "learning_rate": 0.0002861491854046048, "loss": 3.0262, "step": 77800 }, { "epoch": 20.950150732127476, "grad_norm": 0.40351414680480957, "learning_rate": 0.000285947219604147, "loss": 3.0221, "step": 77850 }, { "epoch": 20.963608957795003, "grad_norm": 0.4256989657878876, "learning_rate": 0.00028574525380368923, "loss": 3.0303, "step": 77900 }, { "epoch": 20.977067183462534, "grad_norm": 0.4155173599720001, "learning_rate": 0.0002855432880032314, "loss": 3.0339, "step": 77950 }, { "epoch": 20.99052540913006, "grad_norm": 0.42269203066825867, "learning_rate": 0.0002853413222027736, "loss": 3.0228, "step": 78000 }, { "epoch": 20.99052540913006, "eval_accuracy": 0.39598541176940966, "eval_loss": 3.279735803604126, "eval_runtime": 53.4471, "eval_samples_per_second": 337.006, "eval_steps_per_second": 21.068, "step": 78000 }, { "epoch": 21.00376830318691, "grad_norm": 0.4437350332736969, "learning_rate": 0.00028513935640231587, "loss": 3.0092, "step": 78050 }, { "epoch": 21.017226528854437, "grad_norm": 0.4182265102863312, "learning_rate": 0.00028493739060185807, "loss": 2.934, "step": 78100 }, { "epoch": 21.030684754521964, "grad_norm": 0.43953824043273926, "learning_rate": 0.00028473542480140026, "loss": 2.947, "step": 78150 }, { "epoch": 21.04414298018949, "grad_norm": 0.45151016116142273, "learning_rate": 0.0002845334590009425, "loss": 2.9428, "step": 78200 }, { "epoch": 21.05760120585702, "grad_norm": 0.41173022985458374, "learning_rate": 0.0002843314932004847, "loss": 2.942, "step": 78250 }, { "epoch": 21.07105943152455, "grad_norm": 0.4398365318775177, "learning_rate": 0.0002841295274000269, "loss": 2.9591, "step": 78300 }, { "epoch": 21.084517657192077, "grad_norm": 0.43982040882110596, "learning_rate": 0.0002839275615995691, "loss": 2.9536, "step": 78350 }, { "epoch": 21.097975882859604, "grad_norm": 0.43614324927330017, "learning_rate": 0.0002837255957991113, "loss": 2.9373, "step": 78400 }, { "epoch": 21.11143410852713, "grad_norm": 0.41462191939353943, "learning_rate": 0.00028352362999865355, "loss": 2.9595, "step": 78450 }, { "epoch": 21.12489233419466, "grad_norm": 0.4198111593723297, "learning_rate": 0.00028332166419819574, "loss": 2.9739, "step": 78500 }, { "epoch": 21.13835055986219, "grad_norm": 0.4044225215911865, "learning_rate": 0.000283119698397738, "loss": 2.958, "step": 78550 }, { "epoch": 21.151808785529717, "grad_norm": 0.4500187635421753, "learning_rate": 0.0002829177325972802, "loss": 2.9622, "step": 78600 }, { "epoch": 21.165267011197244, "grad_norm": 0.4328445494174957, "learning_rate": 0.0002827157667968224, "loss": 2.9739, "step": 78650 }, { "epoch": 21.17872523686477, "grad_norm": 0.4654206335544586, "learning_rate": 0.0002825138009963646, "loss": 2.9749, "step": 78700 }, { "epoch": 21.1921834625323, "grad_norm": 0.4135824739933014, "learning_rate": 0.0002823118351959068, "loss": 2.9695, "step": 78750 }, { "epoch": 21.205641688199826, "grad_norm": 0.4080086052417755, "learning_rate": 0.00028210986939544903, "loss": 2.9842, "step": 78800 }, { "epoch": 21.219099913867357, "grad_norm": 0.421791672706604, "learning_rate": 0.0002819079035949912, "loss": 2.9721, "step": 78850 }, { "epoch": 21.232558139534884, "grad_norm": 0.4322538673877716, "learning_rate": 0.0002817059377945334, "loss": 2.9641, "step": 78900 }, { "epoch": 21.24601636520241, "grad_norm": 0.4447803795337677, "learning_rate": 0.00028150397199407567, "loss": 2.9708, "step": 78950 }, { "epoch": 21.25947459086994, "grad_norm": 0.43925315141677856, "learning_rate": 0.00028130200619361787, "loss": 2.9669, "step": 79000 }, { "epoch": 21.25947459086994, "eval_accuracy": 0.3955303987383058, "eval_loss": 3.2910048961639404, "eval_runtime": 53.1984, "eval_samples_per_second": 338.582, "eval_steps_per_second": 21.166, "step": 79000 }, { "epoch": 21.272932816537466, "grad_norm": 0.4245566129684448, "learning_rate": 0.00028110004039316006, "loss": 2.9844, "step": 79050 }, { "epoch": 21.286391042204997, "grad_norm": 0.41280364990234375, "learning_rate": 0.00028089807459270226, "loss": 2.9868, "step": 79100 }, { "epoch": 21.299849267872524, "grad_norm": 0.4568192958831787, "learning_rate": 0.0002806961087922445, "loss": 2.9726, "step": 79150 }, { "epoch": 21.31330749354005, "grad_norm": 0.4463944137096405, "learning_rate": 0.0002804941429917867, "loss": 2.9838, "step": 79200 }, { "epoch": 21.32676571920758, "grad_norm": 0.45832765102386475, "learning_rate": 0.0002802921771913289, "loss": 2.9887, "step": 79250 }, { "epoch": 21.340223944875106, "grad_norm": 0.43500322103500366, "learning_rate": 0.0002800902113908711, "loss": 2.9965, "step": 79300 }, { "epoch": 21.353682170542637, "grad_norm": 0.41847607493400574, "learning_rate": 0.0002798882455904133, "loss": 2.977, "step": 79350 }, { "epoch": 21.367140396210164, "grad_norm": 0.4514514207839966, "learning_rate": 0.00027968627978995555, "loss": 2.9806, "step": 79400 }, { "epoch": 21.38059862187769, "grad_norm": 0.447410523891449, "learning_rate": 0.00027948431398949774, "loss": 2.9908, "step": 79450 }, { "epoch": 21.39405684754522, "grad_norm": 0.4610724449157715, "learning_rate": 0.00027928234818904, "loss": 2.9668, "step": 79500 }, { "epoch": 21.407515073212746, "grad_norm": 0.4488871693611145, "learning_rate": 0.0002790803823885822, "loss": 2.9921, "step": 79550 }, { "epoch": 21.420973298880277, "grad_norm": 0.42067262530326843, "learning_rate": 0.0002788784165881244, "loss": 2.9879, "step": 79600 }, { "epoch": 21.434431524547804, "grad_norm": 0.4374084770679474, "learning_rate": 0.0002786764507876666, "loss": 2.9853, "step": 79650 }, { "epoch": 21.44788975021533, "grad_norm": 0.4477153718471527, "learning_rate": 0.0002784744849872088, "loss": 3.0006, "step": 79700 }, { "epoch": 21.46134797588286, "grad_norm": 0.4401930868625641, "learning_rate": 0.000278272519186751, "loss": 2.995, "step": 79750 }, { "epoch": 21.474806201550386, "grad_norm": 0.4580068290233612, "learning_rate": 0.0002780705533862932, "loss": 2.9824, "step": 79800 }, { "epoch": 21.488264427217917, "grad_norm": 0.45493486523628235, "learning_rate": 0.0002778685875858355, "loss": 2.9976, "step": 79850 }, { "epoch": 21.501722652885444, "grad_norm": 0.44258391857147217, "learning_rate": 0.00027766662178537767, "loss": 2.9964, "step": 79900 }, { "epoch": 21.51518087855297, "grad_norm": 0.4550727307796478, "learning_rate": 0.00027746465598491987, "loss": 2.9849, "step": 79950 }, { "epoch": 21.5286391042205, "grad_norm": 0.4246881902217865, "learning_rate": 0.00027726269018446206, "loss": 3.0057, "step": 80000 }, { "epoch": 21.5286391042205, "eval_accuracy": 0.39579940835268335, "eval_loss": 3.2892096042633057, "eval_runtime": 53.3098, "eval_samples_per_second": 337.874, "eval_steps_per_second": 21.122, "step": 80000 }, { "epoch": 21.542097329888026, "grad_norm": 0.4142609238624573, "learning_rate": 0.00027706072438400426, "loss": 2.9998, "step": 80050 }, { "epoch": 21.555555555555557, "grad_norm": 0.43777963519096375, "learning_rate": 0.0002768587585835465, "loss": 2.997, "step": 80100 }, { "epoch": 21.569013781223084, "grad_norm": 0.4332168996334076, "learning_rate": 0.0002766567927830887, "loss": 3.0043, "step": 80150 }, { "epoch": 21.58247200689061, "grad_norm": 0.42009490728378296, "learning_rate": 0.0002764548269826309, "loss": 3.0102, "step": 80200 }, { "epoch": 21.59593023255814, "grad_norm": 0.44434231519699097, "learning_rate": 0.0002762528611821731, "loss": 2.9915, "step": 80250 }, { "epoch": 21.609388458225666, "grad_norm": 0.44381824135780334, "learning_rate": 0.00027605089538171535, "loss": 2.995, "step": 80300 }, { "epoch": 21.622846683893197, "grad_norm": 0.42324790358543396, "learning_rate": 0.00027584892958125754, "loss": 3.0152, "step": 80350 }, { "epoch": 21.636304909560724, "grad_norm": 0.44452613592147827, "learning_rate": 0.0002756469637807998, "loss": 3.013, "step": 80400 }, { "epoch": 21.649763135228252, "grad_norm": 0.45783066749572754, "learning_rate": 0.000275444997980342, "loss": 3.0144, "step": 80450 }, { "epoch": 21.66322136089578, "grad_norm": 0.4607763886451721, "learning_rate": 0.0002752430321798842, "loss": 3.0039, "step": 80500 }, { "epoch": 21.676679586563306, "grad_norm": 0.42622271180152893, "learning_rate": 0.0002750410663794264, "loss": 3.0025, "step": 80550 }, { "epoch": 21.690137812230837, "grad_norm": 0.4265972971916199, "learning_rate": 0.0002748391005789686, "loss": 3.0159, "step": 80600 }, { "epoch": 21.703596037898365, "grad_norm": 0.4468063414096832, "learning_rate": 0.00027463713477851083, "loss": 3.007, "step": 80650 }, { "epoch": 21.717054263565892, "grad_norm": 0.40498435497283936, "learning_rate": 0.000274435168978053, "loss": 3.0038, "step": 80700 }, { "epoch": 21.73051248923342, "grad_norm": 0.48580971360206604, "learning_rate": 0.0002742332031775953, "loss": 3.0107, "step": 80750 }, { "epoch": 21.743970714900946, "grad_norm": 0.4329358637332916, "learning_rate": 0.00027403123737713747, "loss": 3.0153, "step": 80800 }, { "epoch": 21.757428940568474, "grad_norm": 0.4601568877696991, "learning_rate": 0.00027382927157667967, "loss": 3.0082, "step": 80850 }, { "epoch": 21.770887166236005, "grad_norm": 0.4607238173484802, "learning_rate": 0.00027362730577622186, "loss": 3.0079, "step": 80900 }, { "epoch": 21.784345391903532, "grad_norm": 0.44796958565711975, "learning_rate": 0.00027342533997576406, "loss": 3.0135, "step": 80950 }, { "epoch": 21.79780361757106, "grad_norm": 0.47503456473350525, "learning_rate": 0.0002732233741753063, "loss": 3.0159, "step": 81000 }, { "epoch": 21.79780361757106, "eval_accuracy": 0.3963195008502703, "eval_loss": 3.2789855003356934, "eval_runtime": 53.1425, "eval_samples_per_second": 338.938, "eval_steps_per_second": 21.188, "step": 81000 }, { "epoch": 21.811261843238587, "grad_norm": 0.4426865875720978, "learning_rate": 0.0002730214083748485, "loss": 3.0087, "step": 81050 }, { "epoch": 21.824720068906114, "grad_norm": 0.4314759373664856, "learning_rate": 0.0002728194425743907, "loss": 3.0165, "step": 81100 }, { "epoch": 21.838178294573645, "grad_norm": 0.44944486021995544, "learning_rate": 0.0002726174767739329, "loss": 3.019, "step": 81150 }, { "epoch": 21.851636520241172, "grad_norm": 0.4344472289085388, "learning_rate": 0.00027241551097347515, "loss": 3.011, "step": 81200 }, { "epoch": 21.8650947459087, "grad_norm": 0.4304898679256439, "learning_rate": 0.00027221354517301734, "loss": 3.0056, "step": 81250 }, { "epoch": 21.878552971576227, "grad_norm": 0.43768954277038574, "learning_rate": 0.00027201157937255954, "loss": 3.0129, "step": 81300 }, { "epoch": 21.892011197243754, "grad_norm": 0.4472970962524414, "learning_rate": 0.0002718096135721018, "loss": 3.0209, "step": 81350 }, { "epoch": 21.905469422911285, "grad_norm": 0.4112667739391327, "learning_rate": 0.000271607647771644, "loss": 3.015, "step": 81400 }, { "epoch": 21.918927648578812, "grad_norm": 0.44268813729286194, "learning_rate": 0.0002714056819711862, "loss": 3.0118, "step": 81450 }, { "epoch": 21.93238587424634, "grad_norm": 0.4346647560596466, "learning_rate": 0.0002712037161707284, "loss": 3.0199, "step": 81500 }, { "epoch": 21.945844099913867, "grad_norm": 0.4291246831417084, "learning_rate": 0.0002710017503702706, "loss": 3.0145, "step": 81550 }, { "epoch": 21.959302325581394, "grad_norm": 0.44748732447624207, "learning_rate": 0.0002707997845698128, "loss": 3.0253, "step": 81600 }, { "epoch": 21.972760551248925, "grad_norm": 0.46326854825019836, "learning_rate": 0.000270597818769355, "loss": 3.0084, "step": 81650 }, { "epoch": 21.986218776916452, "grad_norm": 0.4273764193058014, "learning_rate": 0.00027039585296889727, "loss": 3.0134, "step": 81700 }, { "epoch": 21.99967700258398, "grad_norm": 0.4094850718975067, "learning_rate": 0.00027019388716843947, "loss": 3.0211, "step": 81750 }, { "epoch": 22.012919896640827, "grad_norm": 0.4577467143535614, "learning_rate": 0.00026999192136798166, "loss": 2.9251, "step": 81800 }, { "epoch": 22.026378122308355, "grad_norm": 0.4286164343357086, "learning_rate": 0.00026978995556752386, "loss": 2.9454, "step": 81850 }, { "epoch": 22.039836347975882, "grad_norm": 0.46044474840164185, "learning_rate": 0.00026958798976706606, "loss": 2.9283, "step": 81900 }, { "epoch": 22.05329457364341, "grad_norm": 0.463361918926239, "learning_rate": 0.0002693860239666083, "loss": 2.9374, "step": 81950 }, { "epoch": 22.06675279931094, "grad_norm": 0.4062318503856659, "learning_rate": 0.0002691840581661505, "loss": 2.9459, "step": 82000 }, { "epoch": 22.06675279931094, "eval_accuracy": 0.3958484080845429, "eval_loss": 3.290665864944458, "eval_runtime": 53.1747, "eval_samples_per_second": 338.733, "eval_steps_per_second": 21.175, "step": 82000 }, { "epoch": 22.080211024978468, "grad_norm": 0.44609442353248596, "learning_rate": 0.0002689820923656927, "loss": 2.9442, "step": 82050 }, { "epoch": 22.093669250645995, "grad_norm": 0.4272725582122803, "learning_rate": 0.00026878012656523495, "loss": 2.9463, "step": 82100 }, { "epoch": 22.107127476313522, "grad_norm": 0.47067728638648987, "learning_rate": 0.00026857816076477714, "loss": 2.9489, "step": 82150 }, { "epoch": 22.12058570198105, "grad_norm": 0.46830788254737854, "learning_rate": 0.00026837619496431934, "loss": 2.9499, "step": 82200 }, { "epoch": 22.13404392764858, "grad_norm": 0.41217130422592163, "learning_rate": 0.0002681742291638616, "loss": 2.9487, "step": 82250 }, { "epoch": 22.147502153316108, "grad_norm": 0.4633793830871582, "learning_rate": 0.0002679722633634038, "loss": 2.9559, "step": 82300 }, { "epoch": 22.160960378983635, "grad_norm": 0.4432108402252197, "learning_rate": 0.000267770297562946, "loss": 2.9572, "step": 82350 }, { "epoch": 22.174418604651162, "grad_norm": 0.4489414393901825, "learning_rate": 0.0002675683317624882, "loss": 2.9574, "step": 82400 }, { "epoch": 22.18787683031869, "grad_norm": 0.4502296447753906, "learning_rate": 0.0002673663659620304, "loss": 2.9509, "step": 82450 }, { "epoch": 22.20133505598622, "grad_norm": 0.43019405007362366, "learning_rate": 0.0002671644001615726, "loss": 2.9554, "step": 82500 }, { "epoch": 22.214793281653748, "grad_norm": 0.44396713376045227, "learning_rate": 0.0002669624343611148, "loss": 2.9621, "step": 82550 }, { "epoch": 22.228251507321275, "grad_norm": 0.47302505373954773, "learning_rate": 0.00026676046856065707, "loss": 2.9714, "step": 82600 }, { "epoch": 22.241709732988802, "grad_norm": 0.4609719216823578, "learning_rate": 0.00026655850276019927, "loss": 2.9692, "step": 82650 }, { "epoch": 22.25516795865633, "grad_norm": 0.45992007851600647, "learning_rate": 0.00026635653695974146, "loss": 2.9599, "step": 82700 }, { "epoch": 22.26862618432386, "grad_norm": 0.4323056936264038, "learning_rate": 0.00026615457115928366, "loss": 2.9689, "step": 82750 }, { "epoch": 22.282084409991388, "grad_norm": 0.4838802218437195, "learning_rate": 0.00026595260535882586, "loss": 2.9716, "step": 82800 }, { "epoch": 22.295542635658915, "grad_norm": 0.44049209356307983, "learning_rate": 0.0002657506395583681, "loss": 2.9746, "step": 82850 }, { "epoch": 22.309000861326442, "grad_norm": 0.44948625564575195, "learning_rate": 0.0002655486737579103, "loss": 2.9831, "step": 82900 }, { "epoch": 22.32245908699397, "grad_norm": 0.4355025291442871, "learning_rate": 0.00026534670795745255, "loss": 2.9641, "step": 82950 }, { "epoch": 22.3359173126615, "grad_norm": 0.44514182209968567, "learning_rate": 0.00026514474215699475, "loss": 2.9807, "step": 83000 }, { "epoch": 22.3359173126615, "eval_accuracy": 0.3954855275869577, "eval_loss": 3.2914698123931885, "eval_runtime": 53.3839, "eval_samples_per_second": 337.405, "eval_steps_per_second": 21.092, "step": 83000 }, { "epoch": 22.349375538329028, "grad_norm": 0.4393146336078644, "learning_rate": 0.00026494277635653694, "loss": 2.9734, "step": 83050 }, { "epoch": 22.362833763996555, "grad_norm": 0.44433945417404175, "learning_rate": 0.00026474081055607914, "loss": 2.9804, "step": 83100 }, { "epoch": 22.376291989664082, "grad_norm": 0.4565770626068115, "learning_rate": 0.00026453884475562134, "loss": 2.9793, "step": 83150 }, { "epoch": 22.38975021533161, "grad_norm": 0.45272794365882874, "learning_rate": 0.0002643368789551636, "loss": 2.9851, "step": 83200 }, { "epoch": 22.403208440999137, "grad_norm": 0.47611159086227417, "learning_rate": 0.0002641349131547058, "loss": 2.9829, "step": 83250 }, { "epoch": 22.416666666666668, "grad_norm": 0.43398815393447876, "learning_rate": 0.000263932947354248, "loss": 2.9736, "step": 83300 }, { "epoch": 22.430124892334195, "grad_norm": 0.4315131902694702, "learning_rate": 0.0002637309815537902, "loss": 2.9792, "step": 83350 }, { "epoch": 22.443583118001722, "grad_norm": 0.4292556941509247, "learning_rate": 0.0002635290157533324, "loss": 2.9753, "step": 83400 }, { "epoch": 22.45704134366925, "grad_norm": 0.42346417903900146, "learning_rate": 0.0002633270499528746, "loss": 2.9812, "step": 83450 }, { "epoch": 22.470499569336777, "grad_norm": 0.4710681438446045, "learning_rate": 0.0002631250841524168, "loss": 2.9835, "step": 83500 }, { "epoch": 22.483957795004308, "grad_norm": 0.42093920707702637, "learning_rate": 0.00026292311835195907, "loss": 2.9884, "step": 83550 }, { "epoch": 22.497416020671835, "grad_norm": 0.4511464238166809, "learning_rate": 0.00026272115255150126, "loss": 2.9923, "step": 83600 }, { "epoch": 22.510874246339363, "grad_norm": 0.4473262429237366, "learning_rate": 0.00026251918675104346, "loss": 2.9751, "step": 83650 }, { "epoch": 22.52433247200689, "grad_norm": 0.41240695118904114, "learning_rate": 0.00026231722095058566, "loss": 2.9958, "step": 83700 }, { "epoch": 22.537790697674417, "grad_norm": 0.4413813352584839, "learning_rate": 0.00026211525515012785, "loss": 2.979, "step": 83750 }, { "epoch": 22.551248923341948, "grad_norm": 0.4364555776119232, "learning_rate": 0.0002619132893496701, "loss": 2.9978, "step": 83800 }, { "epoch": 22.564707149009475, "grad_norm": 0.44257232546806335, "learning_rate": 0.0002617113235492123, "loss": 2.9917, "step": 83850 }, { "epoch": 22.578165374677003, "grad_norm": 0.4314466416835785, "learning_rate": 0.00026150935774875455, "loss": 2.9873, "step": 83900 }, { "epoch": 22.59162360034453, "grad_norm": 0.4465186297893524, "learning_rate": 0.00026130739194829675, "loss": 2.9795, "step": 83950 }, { "epoch": 22.605081826012057, "grad_norm": 0.46384963393211365, "learning_rate": 0.00026110542614783894, "loss": 2.9952, "step": 84000 }, { "epoch": 22.605081826012057, "eval_accuracy": 0.3961026417265637, "eval_loss": 3.283360242843628, "eval_runtime": 53.2066, "eval_samples_per_second": 338.53, "eval_steps_per_second": 21.163, "step": 84000 }, { "epoch": 22.618540051679588, "grad_norm": 0.4887400269508362, "learning_rate": 0.00026090346034738114, "loss": 2.9848, "step": 84050 }, { "epoch": 22.631998277347115, "grad_norm": 0.5525014400482178, "learning_rate": 0.0002607014945469234, "loss": 2.9975, "step": 84100 }, { "epoch": 22.645456503014643, "grad_norm": 0.4679363965988159, "learning_rate": 0.0002604995287464656, "loss": 2.9913, "step": 84150 }, { "epoch": 22.65891472868217, "grad_norm": 0.43187206983566284, "learning_rate": 0.0002602975629460078, "loss": 2.995, "step": 84200 }, { "epoch": 22.672372954349697, "grad_norm": 0.4239598512649536, "learning_rate": 0.00026009559714555, "loss": 2.9935, "step": 84250 }, { "epoch": 22.685831180017228, "grad_norm": 0.4313739240169525, "learning_rate": 0.0002598936313450922, "loss": 2.9888, "step": 84300 }, { "epoch": 22.699289405684755, "grad_norm": 0.45418158173561096, "learning_rate": 0.0002596916655446344, "loss": 2.9967, "step": 84350 }, { "epoch": 22.712747631352283, "grad_norm": 0.4296848475933075, "learning_rate": 0.0002594896997441766, "loss": 2.9942, "step": 84400 }, { "epoch": 22.72620585701981, "grad_norm": 0.4835188686847687, "learning_rate": 0.00025928773394371887, "loss": 2.9931, "step": 84450 }, { "epoch": 22.739664082687337, "grad_norm": 0.44101449847221375, "learning_rate": 0.00025908576814326107, "loss": 3.0047, "step": 84500 }, { "epoch": 22.753122308354868, "grad_norm": 0.42555737495422363, "learning_rate": 0.00025888380234280326, "loss": 2.9982, "step": 84550 }, { "epoch": 22.766580534022395, "grad_norm": 0.4653916358947754, "learning_rate": 0.00025868183654234546, "loss": 2.9946, "step": 84600 }, { "epoch": 22.780038759689923, "grad_norm": 0.4454815089702606, "learning_rate": 0.00025847987074188765, "loss": 3.0072, "step": 84650 }, { "epoch": 22.79349698535745, "grad_norm": 0.45986753702163696, "learning_rate": 0.0002582779049414299, "loss": 3.0039, "step": 84700 }, { "epoch": 22.806955211024977, "grad_norm": 0.4703720510005951, "learning_rate": 0.0002580759391409721, "loss": 2.9963, "step": 84750 }, { "epoch": 22.820413436692505, "grad_norm": 0.45785728096961975, "learning_rate": 0.00025787397334051435, "loss": 3.001, "step": 84800 }, { "epoch": 22.833871662360036, "grad_norm": 0.454120010137558, "learning_rate": 0.00025767200754005655, "loss": 2.994, "step": 84850 }, { "epoch": 22.847329888027563, "grad_norm": 0.4418194890022278, "learning_rate": 0.00025747004173959874, "loss": 3.0124, "step": 84900 }, { "epoch": 22.86078811369509, "grad_norm": 0.4594516456127167, "learning_rate": 0.00025726807593914094, "loss": 3.0028, "step": 84950 }, { "epoch": 22.874246339362617, "grad_norm": 0.42217785120010376, "learning_rate": 0.00025706611013868313, "loss": 2.9981, "step": 85000 }, { "epoch": 22.874246339362617, "eval_accuracy": 0.39679091955656437, "eval_loss": 3.2778079509735107, "eval_runtime": 53.2599, "eval_samples_per_second": 338.19, "eval_steps_per_second": 21.142, "step": 85000 }, { "epoch": 22.887704565030145, "grad_norm": 0.43962785601615906, "learning_rate": 0.0002568641443382254, "loss": 3.0046, "step": 85050 }, { "epoch": 22.901162790697676, "grad_norm": 0.4394022226333618, "learning_rate": 0.0002566621785377676, "loss": 2.999, "step": 85100 }, { "epoch": 22.914621016365203, "grad_norm": 0.441394567489624, "learning_rate": 0.0002564602127373098, "loss": 3.0038, "step": 85150 }, { "epoch": 22.92807924203273, "grad_norm": 0.4967244863510132, "learning_rate": 0.00025625824693685203, "loss": 3.0086, "step": 85200 }, { "epoch": 22.941537467700257, "grad_norm": 0.44626763463020325, "learning_rate": 0.0002560562811363942, "loss": 3.014, "step": 85250 }, { "epoch": 22.954995693367785, "grad_norm": 0.43168094754219055, "learning_rate": 0.0002558543153359364, "loss": 2.9938, "step": 85300 }, { "epoch": 22.968453919035316, "grad_norm": 0.45025435090065, "learning_rate": 0.0002556523495354786, "loss": 3.013, "step": 85350 }, { "epoch": 22.981912144702843, "grad_norm": 0.42971986532211304, "learning_rate": 0.00025545038373502087, "loss": 3.0124, "step": 85400 }, { "epoch": 22.99537037037037, "grad_norm": 0.4267427921295166, "learning_rate": 0.00025524841793456306, "loss": 3.01, "step": 85450 }, { "epoch": 23.00861326442722, "grad_norm": 0.4329337477684021, "learning_rate": 0.00025504645213410526, "loss": 2.9676, "step": 85500 }, { "epoch": 23.022071490094746, "grad_norm": 0.4235066771507263, "learning_rate": 0.00025484448633364745, "loss": 2.9191, "step": 85550 }, { "epoch": 23.035529715762273, "grad_norm": 0.47007474303245544, "learning_rate": 0.00025464252053318965, "loss": 2.9197, "step": 85600 }, { "epoch": 23.0489879414298, "grad_norm": 0.462991327047348, "learning_rate": 0.0002544405547327319, "loss": 2.935, "step": 85650 }, { "epoch": 23.06244616709733, "grad_norm": 0.4521004855632782, "learning_rate": 0.0002542385889322741, "loss": 2.922, "step": 85700 }, { "epoch": 23.07590439276486, "grad_norm": 0.4741455614566803, "learning_rate": 0.00025403662313181635, "loss": 2.9292, "step": 85750 }, { "epoch": 23.089362618432386, "grad_norm": 0.4532252550125122, "learning_rate": 0.00025383465733135854, "loss": 2.9462, "step": 85800 }, { "epoch": 23.102820844099913, "grad_norm": 0.4494078457355499, "learning_rate": 0.00025363269153090074, "loss": 2.9332, "step": 85850 }, { "epoch": 23.11627906976744, "grad_norm": 0.45025596022605896, "learning_rate": 0.00025343072573044294, "loss": 2.9574, "step": 85900 }, { "epoch": 23.12973729543497, "grad_norm": 0.44868040084838867, "learning_rate": 0.00025322875992998513, "loss": 2.9478, "step": 85950 }, { "epoch": 23.1431955211025, "grad_norm": 0.4450909495353699, "learning_rate": 0.0002530267941295274, "loss": 2.938, "step": 86000 }, { "epoch": 23.1431955211025, "eval_accuracy": 0.3958298294722414, "eval_loss": 3.2935590744018555, "eval_runtime": 53.3991, "eval_samples_per_second": 337.309, "eval_steps_per_second": 21.087, "step": 86000 }, { "epoch": 23.156653746770026, "grad_norm": 0.47848260402679443, "learning_rate": 0.0002528248283290696, "loss": 2.9541, "step": 86050 }, { "epoch": 23.170111972437553, "grad_norm": 0.44607311487197876, "learning_rate": 0.00025262286252861183, "loss": 2.946, "step": 86100 }, { "epoch": 23.18357019810508, "grad_norm": 0.44378456473350525, "learning_rate": 0.000252420896728154, "loss": 2.939, "step": 86150 }, { "epoch": 23.19702842377261, "grad_norm": 0.4426654577255249, "learning_rate": 0.0002522189309276962, "loss": 2.9498, "step": 86200 }, { "epoch": 23.21048664944014, "grad_norm": 0.45932042598724365, "learning_rate": 0.0002520169651272384, "loss": 2.9454, "step": 86250 }, { "epoch": 23.223944875107666, "grad_norm": 0.45173564553260803, "learning_rate": 0.00025181499932678067, "loss": 2.957, "step": 86300 }, { "epoch": 23.237403100775193, "grad_norm": 0.44965678453445435, "learning_rate": 0.00025161303352632286, "loss": 2.952, "step": 86350 }, { "epoch": 23.25086132644272, "grad_norm": 0.4480496048927307, "learning_rate": 0.00025141106772586506, "loss": 2.9572, "step": 86400 }, { "epoch": 23.26431955211025, "grad_norm": 0.46193182468414307, "learning_rate": 0.00025120910192540726, "loss": 2.9526, "step": 86450 }, { "epoch": 23.27777777777778, "grad_norm": 0.48119649291038513, "learning_rate": 0.00025100713612494945, "loss": 2.9621, "step": 86500 }, { "epoch": 23.291236003445306, "grad_norm": 0.4442594647407532, "learning_rate": 0.0002508051703244917, "loss": 2.9495, "step": 86550 }, { "epoch": 23.304694229112833, "grad_norm": 0.4543590247631073, "learning_rate": 0.0002506032045240339, "loss": 2.9681, "step": 86600 }, { "epoch": 23.31815245478036, "grad_norm": 0.4707615375518799, "learning_rate": 0.00025040123872357615, "loss": 2.9575, "step": 86650 }, { "epoch": 23.33161068044789, "grad_norm": 0.4720620810985565, "learning_rate": 0.00025019927292311834, "loss": 2.9659, "step": 86700 }, { "epoch": 23.34506890611542, "grad_norm": 0.47484633326530457, "learning_rate": 0.00024999730712266054, "loss": 2.9692, "step": 86750 }, { "epoch": 23.358527131782946, "grad_norm": 0.44391027092933655, "learning_rate": 0.00024979534132220274, "loss": 2.9715, "step": 86800 }, { "epoch": 23.371985357450473, "grad_norm": 0.4550620913505554, "learning_rate": 0.00024959337552174493, "loss": 2.974, "step": 86850 }, { "epoch": 23.385443583118, "grad_norm": 0.41127505898475647, "learning_rate": 0.0002493914097212872, "loss": 2.9671, "step": 86900 }, { "epoch": 23.39890180878553, "grad_norm": 0.4839131236076355, "learning_rate": 0.0002491894439208294, "loss": 2.968, "step": 86950 }, { "epoch": 23.41236003445306, "grad_norm": 0.4719763696193695, "learning_rate": 0.00024898747812037163, "loss": 2.9782, "step": 87000 }, { "epoch": 23.41236003445306, "eval_accuracy": 0.3960650499145384, "eval_loss": 3.2877259254455566, "eval_runtime": 53.193, "eval_samples_per_second": 338.616, "eval_steps_per_second": 21.168, "step": 87000 }, { "epoch": 23.425818260120586, "grad_norm": 0.4370049238204956, "learning_rate": 0.0002487855123199138, "loss": 2.9755, "step": 87050 }, { "epoch": 23.439276485788113, "grad_norm": 0.424113005399704, "learning_rate": 0.000248583546519456, "loss": 2.9699, "step": 87100 }, { "epoch": 23.45273471145564, "grad_norm": 0.4651334285736084, "learning_rate": 0.0002483815807189982, "loss": 2.9692, "step": 87150 }, { "epoch": 23.466192937123168, "grad_norm": 0.4388973116874695, "learning_rate": 0.0002481796149185404, "loss": 2.9781, "step": 87200 }, { "epoch": 23.4796511627907, "grad_norm": 0.4796842634677887, "learning_rate": 0.00024797764911808266, "loss": 2.9739, "step": 87250 }, { "epoch": 23.493109388458226, "grad_norm": 0.4598848819732666, "learning_rate": 0.00024777568331762486, "loss": 2.9593, "step": 87300 }, { "epoch": 23.506567614125753, "grad_norm": 0.4424052834510803, "learning_rate": 0.00024757371751716706, "loss": 2.9831, "step": 87350 }, { "epoch": 23.52002583979328, "grad_norm": 0.43741047382354736, "learning_rate": 0.0002473717517167093, "loss": 2.9668, "step": 87400 }, { "epoch": 23.533484065460808, "grad_norm": 0.4539873003959656, "learning_rate": 0.0002471697859162515, "loss": 2.9709, "step": 87450 }, { "epoch": 23.54694229112834, "grad_norm": 0.4834721088409424, "learning_rate": 0.0002469678201157937, "loss": 2.9651, "step": 87500 }, { "epoch": 23.560400516795866, "grad_norm": 0.4531641900539398, "learning_rate": 0.0002467658543153359, "loss": 2.9698, "step": 87550 }, { "epoch": 23.573858742463393, "grad_norm": 0.4506773054599762, "learning_rate": 0.00024656388851487815, "loss": 2.976, "step": 87600 }, { "epoch": 23.58731696813092, "grad_norm": 0.49148374795913696, "learning_rate": 0.00024636192271442034, "loss": 2.9882, "step": 87650 }, { "epoch": 23.600775193798448, "grad_norm": 0.48155003786087036, "learning_rate": 0.00024615995691396254, "loss": 2.992, "step": 87700 }, { "epoch": 23.61423341946598, "grad_norm": 0.4349481165409088, "learning_rate": 0.00024595799111350473, "loss": 2.9826, "step": 87750 }, { "epoch": 23.627691645133506, "grad_norm": 0.4904167652130127, "learning_rate": 0.00024575602531304693, "loss": 2.9799, "step": 87800 }, { "epoch": 23.641149870801033, "grad_norm": 0.44706079363822937, "learning_rate": 0.0002455540595125892, "loss": 2.9929, "step": 87850 }, { "epoch": 23.65460809646856, "grad_norm": 0.458001047372818, "learning_rate": 0.00024535209371213143, "loss": 2.9725, "step": 87900 }, { "epoch": 23.668066322136088, "grad_norm": 0.4497480094432831, "learning_rate": 0.0002451501279116736, "loss": 2.9881, "step": 87950 }, { "epoch": 23.68152454780362, "grad_norm": 0.4559875726699829, "learning_rate": 0.0002449481621112158, "loss": 2.9777, "step": 88000 }, { "epoch": 23.68152454780362, "eval_accuracy": 0.39634253398364994, "eval_loss": 3.282762050628662, "eval_runtime": 53.172, "eval_samples_per_second": 338.75, "eval_steps_per_second": 21.177, "step": 88000 }, { "epoch": 23.694982773471146, "grad_norm": 0.4412843883037567, "learning_rate": 0.000244746196310758, "loss": 2.9958, "step": 88050 }, { "epoch": 23.708440999138674, "grad_norm": 0.4425804913043976, "learning_rate": 0.0002445442305103002, "loss": 2.9793, "step": 88100 }, { "epoch": 23.7218992248062, "grad_norm": 0.4500211775302887, "learning_rate": 0.00024434226470984246, "loss": 2.9853, "step": 88150 }, { "epoch": 23.735357450473728, "grad_norm": 0.4496285319328308, "learning_rate": 0.00024414029890938463, "loss": 2.9882, "step": 88200 }, { "epoch": 23.74881567614126, "grad_norm": 0.47043538093566895, "learning_rate": 0.00024393833310892686, "loss": 2.9939, "step": 88250 }, { "epoch": 23.762273901808786, "grad_norm": 0.4536699652671814, "learning_rate": 0.0002437363673084691, "loss": 2.984, "step": 88300 }, { "epoch": 23.775732127476314, "grad_norm": 0.46850210428237915, "learning_rate": 0.0002435344015080113, "loss": 2.9818, "step": 88350 }, { "epoch": 23.78919035314384, "grad_norm": 0.47080132365226746, "learning_rate": 0.0002433324357075535, "loss": 2.9908, "step": 88400 }, { "epoch": 23.802648578811368, "grad_norm": 0.42550772428512573, "learning_rate": 0.00024313046990709572, "loss": 2.9861, "step": 88450 }, { "epoch": 23.8161068044789, "grad_norm": 0.45124003291130066, "learning_rate": 0.00024292850410663792, "loss": 2.9984, "step": 88500 }, { "epoch": 23.829565030146426, "grad_norm": 0.4718407690525055, "learning_rate": 0.00024272653830618014, "loss": 2.9919, "step": 88550 }, { "epoch": 23.843023255813954, "grad_norm": 0.45113521814346313, "learning_rate": 0.00024252457250572234, "loss": 2.9983, "step": 88600 }, { "epoch": 23.85648148148148, "grad_norm": 0.4578773081302643, "learning_rate": 0.00024232260670526453, "loss": 2.993, "step": 88650 }, { "epoch": 23.86993970714901, "grad_norm": 0.43773236870765686, "learning_rate": 0.00024212064090480676, "loss": 2.9961, "step": 88700 }, { "epoch": 23.88339793281654, "grad_norm": 0.46326008439064026, "learning_rate": 0.00024191867510434898, "loss": 3.0035, "step": 88750 }, { "epoch": 23.896856158484066, "grad_norm": 0.45599448680877686, "learning_rate": 0.0002417167093038912, "loss": 2.9981, "step": 88800 }, { "epoch": 23.910314384151594, "grad_norm": 0.49931249022483826, "learning_rate": 0.0002415147435034334, "loss": 2.9917, "step": 88850 }, { "epoch": 23.92377260981912, "grad_norm": 0.45583376288414, "learning_rate": 0.00024131277770297562, "loss": 2.9875, "step": 88900 }, { "epoch": 23.93723083548665, "grad_norm": 0.44669798016548157, "learning_rate": 0.00024111081190251782, "loss": 2.9896, "step": 88950 }, { "epoch": 23.95068906115418, "grad_norm": 0.4412989616394043, "learning_rate": 0.00024090884610206002, "loss": 3.0013, "step": 89000 }, { "epoch": 23.95068906115418, "eval_accuracy": 0.39690010964640665, "eval_loss": 3.275252103805542, "eval_runtime": 53.4152, "eval_samples_per_second": 337.207, "eval_steps_per_second": 21.08, "step": 89000 }, { "epoch": 23.964147286821706, "grad_norm": 0.4542916715145111, "learning_rate": 0.00024070688030160224, "loss": 2.9937, "step": 89050 }, { "epoch": 23.977605512489234, "grad_norm": 0.4350804090499878, "learning_rate": 0.00024050491450114443, "loss": 2.9948, "step": 89100 }, { "epoch": 23.99106373815676, "grad_norm": 0.49811744689941406, "learning_rate": 0.00024030294870068666, "loss": 2.9902, "step": 89150 }, { "epoch": 24.00430663221361, "grad_norm": 0.4629533290863037, "learning_rate": 0.00024010098290022888, "loss": 2.9727, "step": 89200 }, { "epoch": 24.017764857881136, "grad_norm": 0.4568398594856262, "learning_rate": 0.0002398990170997711, "loss": 2.9066, "step": 89250 }, { "epoch": 24.031223083548664, "grad_norm": 0.46679773926734924, "learning_rate": 0.0002396970512993133, "loss": 2.9074, "step": 89300 }, { "epoch": 24.044681309216195, "grad_norm": 0.4443589746952057, "learning_rate": 0.00023949508549885552, "loss": 2.9111, "step": 89350 }, { "epoch": 24.058139534883722, "grad_norm": 0.4438439607620239, "learning_rate": 0.00023929311969839772, "loss": 2.92, "step": 89400 }, { "epoch": 24.07159776055125, "grad_norm": 0.43250682950019836, "learning_rate": 0.00023909115389793992, "loss": 2.9236, "step": 89450 }, { "epoch": 24.085055986218777, "grad_norm": 0.45004555583000183, "learning_rate": 0.00023888918809748214, "loss": 2.931, "step": 89500 }, { "epoch": 24.098514211886304, "grad_norm": 0.47898465394973755, "learning_rate": 0.00023868722229702433, "loss": 2.9268, "step": 89550 }, { "epoch": 24.11197243755383, "grad_norm": 0.46386006474494934, "learning_rate": 0.00023848525649656656, "loss": 2.943, "step": 89600 }, { "epoch": 24.125430663221362, "grad_norm": 0.43041688203811646, "learning_rate": 0.00023828329069610878, "loss": 2.9324, "step": 89650 }, { "epoch": 24.13888888888889, "grad_norm": 0.45035499334335327, "learning_rate": 0.000238081324895651, "loss": 2.9312, "step": 89700 }, { "epoch": 24.152347114556417, "grad_norm": 0.47834378480911255, "learning_rate": 0.0002378793590951932, "loss": 2.9443, "step": 89750 }, { "epoch": 24.165805340223944, "grad_norm": 0.4648357033729553, "learning_rate": 0.0002376773932947354, "loss": 2.9397, "step": 89800 }, { "epoch": 24.17926356589147, "grad_norm": 0.45432841777801514, "learning_rate": 0.00023747542749427762, "loss": 2.9383, "step": 89850 }, { "epoch": 24.192721791559002, "grad_norm": 0.451199471950531, "learning_rate": 0.00023727346169381982, "loss": 2.9324, "step": 89900 }, { "epoch": 24.20618001722653, "grad_norm": 0.4608922302722931, "learning_rate": 0.00023707149589336204, "loss": 2.9457, "step": 89950 }, { "epoch": 24.219638242894057, "grad_norm": 0.44181180000305176, "learning_rate": 0.00023686953009290424, "loss": 2.9362, "step": 90000 }, { "epoch": 24.219638242894057, "eval_accuracy": 0.3963043989373468, "eval_loss": 3.2903923988342285, "eval_runtime": 53.3682, "eval_samples_per_second": 337.505, "eval_steps_per_second": 21.099, "step": 90000 }, { "epoch": 24.233096468561584, "grad_norm": 0.4413633942604065, "learning_rate": 0.00023666756429244643, "loss": 2.9332, "step": 90050 }, { "epoch": 24.24655469422911, "grad_norm": 0.4630167484283447, "learning_rate": 0.00023646559849198868, "loss": 2.9405, "step": 90100 }, { "epoch": 24.260012919896642, "grad_norm": 0.47358977794647217, "learning_rate": 0.00023626363269153088, "loss": 2.9412, "step": 90150 }, { "epoch": 24.27347114556417, "grad_norm": 0.4817167818546295, "learning_rate": 0.0002360616668910731, "loss": 2.9531, "step": 90200 }, { "epoch": 24.286929371231697, "grad_norm": 0.5337883830070496, "learning_rate": 0.0002358597010906153, "loss": 2.9484, "step": 90250 }, { "epoch": 24.300387596899224, "grad_norm": 0.4577960968017578, "learning_rate": 0.00023565773529015752, "loss": 2.956, "step": 90300 }, { "epoch": 24.31384582256675, "grad_norm": 0.45729923248291016, "learning_rate": 0.00023545576948969972, "loss": 2.9503, "step": 90350 }, { "epoch": 24.327304048234282, "grad_norm": 0.4802705943584442, "learning_rate": 0.00023525380368924194, "loss": 2.9615, "step": 90400 }, { "epoch": 24.34076227390181, "grad_norm": 0.4446525275707245, "learning_rate": 0.00023505183788878414, "loss": 2.9507, "step": 90450 }, { "epoch": 24.354220499569337, "grad_norm": 0.437545508146286, "learning_rate": 0.00023484987208832633, "loss": 2.9511, "step": 90500 }, { "epoch": 24.367678725236864, "grad_norm": 0.46033021807670593, "learning_rate": 0.00023464790628786858, "loss": 2.9563, "step": 90550 }, { "epoch": 24.38113695090439, "grad_norm": 0.4750497341156006, "learning_rate": 0.00023444594048741078, "loss": 2.9652, "step": 90600 }, { "epoch": 24.394595176571922, "grad_norm": 0.4922107458114624, "learning_rate": 0.000234243974686953, "loss": 2.9485, "step": 90650 }, { "epoch": 24.40805340223945, "grad_norm": 0.4701991379261017, "learning_rate": 0.0002340420088864952, "loss": 2.9618, "step": 90700 }, { "epoch": 24.421511627906977, "grad_norm": 0.46807754039764404, "learning_rate": 0.00023384004308603742, "loss": 2.982, "step": 90750 }, { "epoch": 24.434969853574504, "grad_norm": 0.4745599627494812, "learning_rate": 0.00023363807728557962, "loss": 2.9547, "step": 90800 }, { "epoch": 24.44842807924203, "grad_norm": 0.4703523516654968, "learning_rate": 0.0002334361114851218, "loss": 2.9533, "step": 90850 }, { "epoch": 24.461886304909562, "grad_norm": 0.49658453464508057, "learning_rate": 0.00023323414568466404, "loss": 2.9641, "step": 90900 }, { "epoch": 24.47534453057709, "grad_norm": 0.4539616107940674, "learning_rate": 0.00023303217988420623, "loss": 2.9629, "step": 90950 }, { "epoch": 24.488802756244617, "grad_norm": 0.5139576196670532, "learning_rate": 0.00023283021408374848, "loss": 2.951, "step": 91000 }, { "epoch": 24.488802756244617, "eval_accuracy": 0.3967428976464049, "eval_loss": 3.28558087348938, "eval_runtime": 53.262, "eval_samples_per_second": 338.177, "eval_steps_per_second": 21.141, "step": 91000 }, { "epoch": 24.502260981912144, "grad_norm": 0.44592607021331787, "learning_rate": 0.00023262824828329068, "loss": 2.9816, "step": 91050 }, { "epoch": 24.51571920757967, "grad_norm": 0.46577224135398865, "learning_rate": 0.0002324262824828329, "loss": 2.9624, "step": 91100 }, { "epoch": 24.529177433247202, "grad_norm": 0.46494442224502563, "learning_rate": 0.0002322243166823751, "loss": 2.9691, "step": 91150 }, { "epoch": 24.54263565891473, "grad_norm": 0.4683932662010193, "learning_rate": 0.0002320223508819173, "loss": 2.9675, "step": 91200 }, { "epoch": 24.556093884582257, "grad_norm": 0.4711366593837738, "learning_rate": 0.00023182038508145952, "loss": 2.9824, "step": 91250 }, { "epoch": 24.569552110249784, "grad_norm": 0.4714370369911194, "learning_rate": 0.0002316184192810017, "loss": 2.99, "step": 91300 }, { "epoch": 24.58301033591731, "grad_norm": 0.4560222625732422, "learning_rate": 0.00023141645348054394, "loss": 2.9751, "step": 91350 }, { "epoch": 24.596468561584842, "grad_norm": 0.4710794985294342, "learning_rate": 0.00023121448768008613, "loss": 2.9782, "step": 91400 }, { "epoch": 24.60992678725237, "grad_norm": 0.4680551588535309, "learning_rate": 0.00023101252187962838, "loss": 2.9698, "step": 91450 }, { "epoch": 24.623385012919897, "grad_norm": 0.47461768984794617, "learning_rate": 0.00023081055607917058, "loss": 2.9613, "step": 91500 }, { "epoch": 24.636843238587424, "grad_norm": 0.42775940895080566, "learning_rate": 0.0002306085902787128, "loss": 2.9716, "step": 91550 }, { "epoch": 24.65030146425495, "grad_norm": 0.491133451461792, "learning_rate": 0.000230406624478255, "loss": 2.9687, "step": 91600 }, { "epoch": 24.66375968992248, "grad_norm": 0.43853914737701416, "learning_rate": 0.0002302046586777972, "loss": 2.9657, "step": 91650 }, { "epoch": 24.67721791559001, "grad_norm": 0.46136704087257385, "learning_rate": 0.00023000269287733942, "loss": 2.9701, "step": 91700 }, { "epoch": 24.690676141257537, "grad_norm": 0.5232241749763489, "learning_rate": 0.00022980072707688161, "loss": 2.9697, "step": 91750 }, { "epoch": 24.704134366925064, "grad_norm": 0.45869165658950806, "learning_rate": 0.00022959876127642384, "loss": 2.9797, "step": 91800 }, { "epoch": 24.71759259259259, "grad_norm": 0.46300241351127625, "learning_rate": 0.00022939679547596603, "loss": 2.9832, "step": 91850 }, { "epoch": 24.73105081826012, "grad_norm": 0.456046462059021, "learning_rate": 0.00022919482967550828, "loss": 2.9793, "step": 91900 }, { "epoch": 24.74450904392765, "grad_norm": 0.4491710066795349, "learning_rate": 0.00022899286387505048, "loss": 2.9819, "step": 91950 }, { "epoch": 24.757967269595177, "grad_norm": 0.44479814171791077, "learning_rate": 0.00022879089807459268, "loss": 2.975, "step": 92000 }, { "epoch": 24.757967269595177, "eval_accuracy": 0.39675495744737255, "eval_loss": 3.2818639278411865, "eval_runtime": 53.243, "eval_samples_per_second": 338.298, "eval_steps_per_second": 21.148, "step": 92000 }, { "epoch": 24.771425495262704, "grad_norm": 0.4630224406719208, "learning_rate": 0.0002285889322741349, "loss": 2.9782, "step": 92050 }, { "epoch": 24.78488372093023, "grad_norm": 0.4504377841949463, "learning_rate": 0.0002283869664736771, "loss": 2.9821, "step": 92100 }, { "epoch": 24.79834194659776, "grad_norm": 0.46722739934921265, "learning_rate": 0.00022818500067321932, "loss": 2.9873, "step": 92150 }, { "epoch": 24.81180017226529, "grad_norm": 0.4305804967880249, "learning_rate": 0.00022798303487276151, "loss": 2.9946, "step": 92200 }, { "epoch": 24.825258397932817, "grad_norm": 0.46924400329589844, "learning_rate": 0.0002277810690723037, "loss": 2.9798, "step": 92250 }, { "epoch": 24.838716623600344, "grad_norm": 0.4677594304084778, "learning_rate": 0.00022757910327184596, "loss": 2.9792, "step": 92300 }, { "epoch": 24.852174849267872, "grad_norm": 0.4811413586139679, "learning_rate": 0.00022737713747138818, "loss": 2.9849, "step": 92350 }, { "epoch": 24.8656330749354, "grad_norm": 0.4708091914653778, "learning_rate": 0.00022717517167093038, "loss": 2.9734, "step": 92400 }, { "epoch": 24.87909130060293, "grad_norm": 0.4559333622455597, "learning_rate": 0.00022697320587047258, "loss": 2.9786, "step": 92450 }, { "epoch": 24.892549526270457, "grad_norm": 0.4559614360332489, "learning_rate": 0.0002267712400700148, "loss": 2.9828, "step": 92500 }, { "epoch": 24.906007751937985, "grad_norm": 0.4416915476322174, "learning_rate": 0.000226569274269557, "loss": 2.9819, "step": 92550 }, { "epoch": 24.919465977605512, "grad_norm": 0.4463852047920227, "learning_rate": 0.00022636730846909922, "loss": 2.9988, "step": 92600 }, { "epoch": 24.93292420327304, "grad_norm": 0.49153658747673035, "learning_rate": 0.00022616534266864141, "loss": 2.9787, "step": 92650 }, { "epoch": 24.94638242894057, "grad_norm": 0.45088720321655273, "learning_rate": 0.0002259633768681836, "loss": 2.988, "step": 92700 }, { "epoch": 24.959840654608097, "grad_norm": 0.4419504404067993, "learning_rate": 0.00022576141106772586, "loss": 2.978, "step": 92750 }, { "epoch": 24.973298880275625, "grad_norm": 0.4792090058326721, "learning_rate": 0.00022555944526726806, "loss": 2.9729, "step": 92800 }, { "epoch": 24.986757105943152, "grad_norm": 0.4608009159564972, "learning_rate": 0.00022535747946681028, "loss": 2.9824, "step": 92850 }, { "epoch": 25.0, "grad_norm": 0.946661114692688, "learning_rate": 0.00022515551366635248, "loss": 2.9921, "step": 92900 }, { "epoch": 25.013458225667527, "grad_norm": 0.46860387921333313, "learning_rate": 0.0002249535478658947, "loss": 2.9019, "step": 92950 }, { "epoch": 25.026916451335055, "grad_norm": 0.4779540002346039, "learning_rate": 0.0002247515820654369, "loss": 2.8962, "step": 93000 }, { "epoch": 25.026916451335055, "eval_accuracy": 0.3965091982600858, "eval_loss": 3.290764331817627, "eval_runtime": 53.1111, "eval_samples_per_second": 339.138, "eval_steps_per_second": 21.201, "step": 93000 }, { "epoch": 25.040374677002585, "grad_norm": 0.4737008810043335, "learning_rate": 0.0002245496162649791, "loss": 2.9012, "step": 93050 }, { "epoch": 25.053832902670113, "grad_norm": 0.45304667949676514, "learning_rate": 0.00022434765046452131, "loss": 2.9132, "step": 93100 }, { "epoch": 25.06729112833764, "grad_norm": 0.49432122707366943, "learning_rate": 0.0002241456846640635, "loss": 2.9148, "step": 93150 }, { "epoch": 25.080749354005167, "grad_norm": 0.4715327024459839, "learning_rate": 0.00022394371886360576, "loss": 2.9212, "step": 93200 }, { "epoch": 25.094207579672695, "grad_norm": 0.4631378650665283, "learning_rate": 0.00022374175306314796, "loss": 2.9205, "step": 93250 }, { "epoch": 25.107665805340226, "grad_norm": 0.44028857350349426, "learning_rate": 0.00022353978726269018, "loss": 2.9289, "step": 93300 }, { "epoch": 25.121124031007753, "grad_norm": 0.481253981590271, "learning_rate": 0.00022333782146223238, "loss": 2.9248, "step": 93350 }, { "epoch": 25.13458225667528, "grad_norm": 0.47388628125190735, "learning_rate": 0.0002231358556617746, "loss": 2.9136, "step": 93400 }, { "epoch": 25.148040482342807, "grad_norm": 0.471007376909256, "learning_rate": 0.0002229338898613168, "loss": 2.9299, "step": 93450 }, { "epoch": 25.161498708010335, "grad_norm": 0.47656720876693726, "learning_rate": 0.000222731924060859, "loss": 2.9299, "step": 93500 }, { "epoch": 25.174956933677866, "grad_norm": 0.48185816407203674, "learning_rate": 0.00022252995826040122, "loss": 2.9189, "step": 93550 }, { "epoch": 25.188415159345393, "grad_norm": 0.4834353029727936, "learning_rate": 0.0002223279924599434, "loss": 2.9224, "step": 93600 }, { "epoch": 25.20187338501292, "grad_norm": 0.45164158940315247, "learning_rate": 0.00022212602665948566, "loss": 2.9373, "step": 93650 }, { "epoch": 25.215331610680447, "grad_norm": 0.4696139693260193, "learning_rate": 0.00022192406085902786, "loss": 2.9385, "step": 93700 }, { "epoch": 25.228789836347975, "grad_norm": 0.45761996507644653, "learning_rate": 0.00022172209505857008, "loss": 2.9435, "step": 93750 }, { "epoch": 25.242248062015506, "grad_norm": 0.4780551493167877, "learning_rate": 0.00022152012925811228, "loss": 2.9396, "step": 93800 }, { "epoch": 25.255706287683033, "grad_norm": 0.4723582863807678, "learning_rate": 0.00022131816345765447, "loss": 2.9307, "step": 93850 }, { "epoch": 25.26916451335056, "grad_norm": 0.4635056257247925, "learning_rate": 0.0002211161976571967, "loss": 2.9335, "step": 93900 }, { "epoch": 25.282622739018088, "grad_norm": 0.4594542980194092, "learning_rate": 0.0002209142318567389, "loss": 2.9302, "step": 93950 }, { "epoch": 25.296080964685615, "grad_norm": 0.45935070514678955, "learning_rate": 0.00022071226605628112, "loss": 2.9417, "step": 94000 }, { "epoch": 25.296080964685615, "eval_accuracy": 0.39655493858627844, "eval_loss": 3.287743330001831, "eval_runtime": 53.1973, "eval_samples_per_second": 338.589, "eval_steps_per_second": 21.166, "step": 94000 }, { "epoch": 25.309539190353142, "grad_norm": 0.4579530358314514, "learning_rate": 0.0002205103002558233, "loss": 2.9377, "step": 94050 }, { "epoch": 25.322997416020673, "grad_norm": 0.5007568597793579, "learning_rate": 0.00022030833445536556, "loss": 2.9477, "step": 94100 }, { "epoch": 25.3364556416882, "grad_norm": 0.47446051239967346, "learning_rate": 0.00022010636865490776, "loss": 2.9362, "step": 94150 }, { "epoch": 25.349913867355728, "grad_norm": 0.444487065076828, "learning_rate": 0.00021990440285444998, "loss": 2.9374, "step": 94200 }, { "epoch": 25.363372093023255, "grad_norm": 0.4731683135032654, "learning_rate": 0.00021970243705399218, "loss": 2.9463, "step": 94250 }, { "epoch": 25.376830318690782, "grad_norm": 0.45736849308013916, "learning_rate": 0.00021950047125353437, "loss": 2.9414, "step": 94300 }, { "epoch": 25.390288544358313, "grad_norm": 0.45242130756378174, "learning_rate": 0.0002192985054530766, "loss": 2.9547, "step": 94350 }, { "epoch": 25.40374677002584, "grad_norm": 0.45447608828544617, "learning_rate": 0.0002190965396526188, "loss": 2.9417, "step": 94400 }, { "epoch": 25.417204995693368, "grad_norm": 0.46485042572021484, "learning_rate": 0.00021889457385216102, "loss": 2.9464, "step": 94450 }, { "epoch": 25.430663221360895, "grad_norm": 0.45448994636535645, "learning_rate": 0.0002186926080517032, "loss": 2.9645, "step": 94500 }, { "epoch": 25.444121447028422, "grad_norm": 0.5003591179847717, "learning_rate": 0.00021849064225124546, "loss": 2.9631, "step": 94550 }, { "epoch": 25.457579672695953, "grad_norm": 0.466730535030365, "learning_rate": 0.00021828867645078766, "loss": 2.9466, "step": 94600 }, { "epoch": 25.47103789836348, "grad_norm": 0.4516577124595642, "learning_rate": 0.00021808671065032985, "loss": 2.9398, "step": 94650 }, { "epoch": 25.484496124031008, "grad_norm": 0.47522616386413574, "learning_rate": 0.00021788474484987208, "loss": 2.9451, "step": 94700 }, { "epoch": 25.497954349698535, "grad_norm": 0.5053650140762329, "learning_rate": 0.00021768277904941427, "loss": 2.9583, "step": 94750 }, { "epoch": 25.511412575366062, "grad_norm": 0.5291735529899597, "learning_rate": 0.0002174808132489565, "loss": 2.9607, "step": 94800 }, { "epoch": 25.524870801033593, "grad_norm": 0.4376024007797241, "learning_rate": 0.0002172788474484987, "loss": 2.9555, "step": 94850 }, { "epoch": 25.53832902670112, "grad_norm": 0.4744149148464203, "learning_rate": 0.0002170768816480409, "loss": 2.9574, "step": 94900 }, { "epoch": 25.551787252368648, "grad_norm": 0.46446311473846436, "learning_rate": 0.0002168749158475831, "loss": 2.9484, "step": 94950 }, { "epoch": 25.565245478036175, "grad_norm": 0.5329943895339966, "learning_rate": 0.00021667295004712534, "loss": 2.9678, "step": 95000 }, { "epoch": 25.565245478036175, "eval_accuracy": 0.3970521065973413, "eval_loss": 3.284172773361206, "eval_runtime": 53.2481, "eval_samples_per_second": 338.265, "eval_steps_per_second": 21.146, "step": 95000 }, { "epoch": 25.578703703703702, "grad_norm": 0.45800742506980896, "learning_rate": 0.00021647098424666756, "loss": 2.9489, "step": 95050 }, { "epoch": 25.592161929371233, "grad_norm": 0.4492628872394562, "learning_rate": 0.00021626901844620976, "loss": 2.9634, "step": 95100 }, { "epoch": 25.60562015503876, "grad_norm": 0.4850592613220215, "learning_rate": 0.00021606705264575198, "loss": 2.9495, "step": 95150 }, { "epoch": 25.619078380706288, "grad_norm": 0.4660792946815491, "learning_rate": 0.00021586508684529417, "loss": 2.9646, "step": 95200 }, { "epoch": 25.632536606373815, "grad_norm": 0.5027803778648376, "learning_rate": 0.0002156631210448364, "loss": 2.9722, "step": 95250 }, { "epoch": 25.645994832041342, "grad_norm": 0.4320453703403473, "learning_rate": 0.0002154611552443786, "loss": 2.9601, "step": 95300 }, { "epoch": 25.659453057708873, "grad_norm": 0.507905900478363, "learning_rate": 0.0002152591894439208, "loss": 2.9649, "step": 95350 }, { "epoch": 25.6729112833764, "grad_norm": 0.4640471637248993, "learning_rate": 0.000215057223643463, "loss": 2.96, "step": 95400 }, { "epoch": 25.686369509043928, "grad_norm": 0.4867199659347534, "learning_rate": 0.00021485525784300524, "loss": 2.9652, "step": 95450 }, { "epoch": 25.699827734711455, "grad_norm": 0.5132190585136414, "learning_rate": 0.00021465329204254746, "loss": 2.9669, "step": 95500 }, { "epoch": 25.713285960378983, "grad_norm": 0.4502236247062683, "learning_rate": 0.00021445132624208966, "loss": 2.9701, "step": 95550 }, { "epoch": 25.72674418604651, "grad_norm": 0.5049893856048584, "learning_rate": 0.00021424936044163188, "loss": 2.9641, "step": 95600 }, { "epoch": 25.74020241171404, "grad_norm": 0.4590776860713959, "learning_rate": 0.00021404739464117407, "loss": 2.9627, "step": 95650 }, { "epoch": 25.753660637381568, "grad_norm": 0.46981340646743774, "learning_rate": 0.00021384542884071627, "loss": 2.9816, "step": 95700 }, { "epoch": 25.767118863049095, "grad_norm": 0.4755721688270569, "learning_rate": 0.0002136434630402585, "loss": 2.981, "step": 95750 }, { "epoch": 25.780577088716623, "grad_norm": 0.4739704132080078, "learning_rate": 0.0002134414972398007, "loss": 2.9738, "step": 95800 }, { "epoch": 25.794035314384153, "grad_norm": 0.46069255471229553, "learning_rate": 0.0002132395314393429, "loss": 2.9837, "step": 95850 }, { "epoch": 25.80749354005168, "grad_norm": 0.4511219561100006, "learning_rate": 0.00021303756563888514, "loss": 2.9805, "step": 95900 }, { "epoch": 25.820951765719208, "grad_norm": 0.4511417746543884, "learning_rate": 0.00021283559983842736, "loss": 2.9704, "step": 95950 }, { "epoch": 25.834409991386735, "grad_norm": 0.45782825350761414, "learning_rate": 0.00021263363403796956, "loss": 2.9681, "step": 96000 }, { "epoch": 25.834409991386735, "eval_accuracy": 0.3971620572151725, "eval_loss": 3.2770164012908936, "eval_runtime": 53.2448, "eval_samples_per_second": 338.287, "eval_steps_per_second": 21.148, "step": 96000 }, { "epoch": 25.847868217054263, "grad_norm": 0.46823206543922424, "learning_rate": 0.00021243166823751175, "loss": 2.9675, "step": 96050 }, { "epoch": 25.86132644272179, "grad_norm": 0.5006187558174133, "learning_rate": 0.00021222970243705398, "loss": 2.9624, "step": 96100 }, { "epoch": 25.87478466838932, "grad_norm": 0.47216886281967163, "learning_rate": 0.00021202773663659617, "loss": 2.9821, "step": 96150 }, { "epoch": 25.888242894056848, "grad_norm": 0.46271994709968567, "learning_rate": 0.0002118257708361384, "loss": 2.9807, "step": 96200 }, { "epoch": 25.901701119724375, "grad_norm": 0.49749988317489624, "learning_rate": 0.0002116238050356806, "loss": 2.9749, "step": 96250 }, { "epoch": 25.915159345391903, "grad_norm": 0.4832835793495178, "learning_rate": 0.00021142183923522281, "loss": 2.9753, "step": 96300 }, { "epoch": 25.92861757105943, "grad_norm": 0.49002805352211, "learning_rate": 0.00021121987343476504, "loss": 2.9831, "step": 96350 }, { "epoch": 25.94207579672696, "grad_norm": 0.466495156288147, "learning_rate": 0.00021101790763430726, "loss": 2.9786, "step": 96400 }, { "epoch": 25.955534022394488, "grad_norm": 0.4698829650878906, "learning_rate": 0.00021081594183384946, "loss": 2.9683, "step": 96450 }, { "epoch": 25.968992248062015, "grad_norm": 0.46221068501472473, "learning_rate": 0.00021061397603339165, "loss": 2.9681, "step": 96500 }, { "epoch": 25.982450473729543, "grad_norm": 0.4581020772457123, "learning_rate": 0.00021041201023293388, "loss": 2.981, "step": 96550 }, { "epoch": 25.99590869939707, "grad_norm": 0.4899604022502899, "learning_rate": 0.00021021004443247607, "loss": 2.9846, "step": 96600 }, { "epoch": 26.009151593453918, "grad_norm": 0.4719387888908386, "learning_rate": 0.0002100080786320183, "loss": 2.9147, "step": 96650 }, { "epoch": 26.022609819121445, "grad_norm": 0.4555143713951111, "learning_rate": 0.0002098061128315605, "loss": 2.8874, "step": 96700 }, { "epoch": 26.036068044788976, "grad_norm": 0.46729525923728943, "learning_rate": 0.00020960414703110274, "loss": 2.8964, "step": 96750 }, { "epoch": 26.049526270456504, "grad_norm": 0.45050865411758423, "learning_rate": 0.00020940218123064494, "loss": 2.9157, "step": 96800 }, { "epoch": 26.06298449612403, "grad_norm": 0.4598424434661865, "learning_rate": 0.00020920021543018713, "loss": 2.908, "step": 96850 }, { "epoch": 26.076442721791558, "grad_norm": 0.4923440217971802, "learning_rate": 0.00020899824962972936, "loss": 2.9066, "step": 96900 }, { "epoch": 26.089900947459086, "grad_norm": 0.46300119161605835, "learning_rate": 0.00020879628382927155, "loss": 2.9051, "step": 96950 }, { "epoch": 26.103359173126616, "grad_norm": 0.4792136549949646, "learning_rate": 0.00020859431802881378, "loss": 2.9053, "step": 97000 }, { "epoch": 26.103359173126616, "eval_accuracy": 0.3962396454114304, "eval_loss": 3.2915122509002686, "eval_runtime": 53.217, "eval_samples_per_second": 338.463, "eval_steps_per_second": 21.159, "step": 97000 }, { "epoch": 26.116817398794144, "grad_norm": 0.45519232749938965, "learning_rate": 0.00020839235222835597, "loss": 2.916, "step": 97050 }, { "epoch": 26.13027562446167, "grad_norm": 0.47248655557632446, "learning_rate": 0.00020819038642789817, "loss": 2.9061, "step": 97100 }, { "epoch": 26.1437338501292, "grad_norm": 0.4677557051181793, "learning_rate": 0.0002079884206274404, "loss": 2.925, "step": 97150 }, { "epoch": 26.157192075796726, "grad_norm": 0.4564340114593506, "learning_rate": 0.00020778645482698264, "loss": 2.9118, "step": 97200 }, { "epoch": 26.170650301464256, "grad_norm": 0.47288423776626587, "learning_rate": 0.00020758448902652484, "loss": 2.9354, "step": 97250 }, { "epoch": 26.184108527131784, "grad_norm": 0.5112426280975342, "learning_rate": 0.00020738252322606703, "loss": 2.913, "step": 97300 }, { "epoch": 26.19756675279931, "grad_norm": 0.48271527886390686, "learning_rate": 0.00020718055742560926, "loss": 2.9189, "step": 97350 }, { "epoch": 26.21102497846684, "grad_norm": 0.4612698256969452, "learning_rate": 0.00020697859162515145, "loss": 2.9164, "step": 97400 }, { "epoch": 26.224483204134366, "grad_norm": 0.48598477244377136, "learning_rate": 0.00020677662582469368, "loss": 2.9385, "step": 97450 }, { "epoch": 26.237941429801896, "grad_norm": 0.4755497872829437, "learning_rate": 0.00020657466002423587, "loss": 2.9334, "step": 97500 }, { "epoch": 26.251399655469424, "grad_norm": 0.4690141975879669, "learning_rate": 0.00020637269422377807, "loss": 2.9302, "step": 97550 }, { "epoch": 26.26485788113695, "grad_norm": 0.45227643847465515, "learning_rate": 0.0002061707284233203, "loss": 2.9261, "step": 97600 }, { "epoch": 26.27831610680448, "grad_norm": 0.4535977244377136, "learning_rate": 0.00020596876262286251, "loss": 2.9256, "step": 97650 }, { "epoch": 26.291774332472006, "grad_norm": 0.46246781945228577, "learning_rate": 0.00020576679682240474, "loss": 2.9267, "step": 97700 }, { "epoch": 26.305232558139537, "grad_norm": 0.47968053817749023, "learning_rate": 0.00020556483102194693, "loss": 2.9305, "step": 97750 }, { "epoch": 26.318690783807064, "grad_norm": 0.45259711146354675, "learning_rate": 0.00020536286522148916, "loss": 2.9192, "step": 97800 }, { "epoch": 26.33214900947459, "grad_norm": 0.45785126090049744, "learning_rate": 0.00020516089942103135, "loss": 2.9314, "step": 97850 }, { "epoch": 26.34560723514212, "grad_norm": 0.4738333821296692, "learning_rate": 0.00020495893362057355, "loss": 2.9313, "step": 97900 }, { "epoch": 26.359065460809646, "grad_norm": 0.47246575355529785, "learning_rate": 0.00020475696782011577, "loss": 2.937, "step": 97950 }, { "epoch": 26.372523686477177, "grad_norm": 0.5067406892776489, "learning_rate": 0.00020455500201965797, "loss": 2.9413, "step": 98000 }, { "epoch": 26.372523686477177, "eval_accuracy": 0.3968445911032132, "eval_loss": 3.289306163787842, "eval_runtime": 53.1207, "eval_samples_per_second": 339.077, "eval_steps_per_second": 21.197, "step": 98000 }, { "epoch": 26.385981912144704, "grad_norm": 0.4930259883403778, "learning_rate": 0.0002043530362192002, "loss": 2.9361, "step": 98050 }, { "epoch": 26.39944013781223, "grad_norm": 0.48024407029151917, "learning_rate": 0.00020415107041874242, "loss": 2.9505, "step": 98100 }, { "epoch": 26.41289836347976, "grad_norm": 0.47633829712867737, "learning_rate": 0.00020394910461828464, "loss": 2.9337, "step": 98150 }, { "epoch": 26.426356589147286, "grad_norm": 0.48809248208999634, "learning_rate": 0.00020374713881782683, "loss": 2.9409, "step": 98200 }, { "epoch": 26.439814814814813, "grad_norm": 0.4671519100666046, "learning_rate": 0.00020354517301736906, "loss": 2.946, "step": 98250 }, { "epoch": 26.453273040482344, "grad_norm": 0.5086513757705688, "learning_rate": 0.00020334320721691125, "loss": 2.9288, "step": 98300 }, { "epoch": 26.46673126614987, "grad_norm": 0.4985395073890686, "learning_rate": 0.00020314124141645345, "loss": 2.9412, "step": 98350 }, { "epoch": 26.4801894918174, "grad_norm": 0.5244379639625549, "learning_rate": 0.00020293927561599567, "loss": 2.9527, "step": 98400 }, { "epoch": 26.493647717484926, "grad_norm": 0.4727938175201416, "learning_rate": 0.00020273730981553787, "loss": 2.9467, "step": 98450 }, { "epoch": 26.507105943152453, "grad_norm": 0.4935176372528076, "learning_rate": 0.0002025353440150801, "loss": 2.9375, "step": 98500 }, { "epoch": 26.520564168819984, "grad_norm": 0.4935429096221924, "learning_rate": 0.00020233337821462232, "loss": 2.9595, "step": 98550 }, { "epoch": 26.53402239448751, "grad_norm": 0.49508944153785706, "learning_rate": 0.00020213141241416454, "loss": 2.9458, "step": 98600 }, { "epoch": 26.54748062015504, "grad_norm": 0.5003538131713867, "learning_rate": 0.00020192944661370674, "loss": 2.9514, "step": 98650 }, { "epoch": 26.560938845822566, "grad_norm": 0.48732179403305054, "learning_rate": 0.00020172748081324893, "loss": 2.9505, "step": 98700 }, { "epoch": 26.574397071490093, "grad_norm": 0.4849097430706024, "learning_rate": 0.00020152551501279115, "loss": 2.9532, "step": 98750 }, { "epoch": 26.587855297157624, "grad_norm": 0.4961356222629547, "learning_rate": 0.00020132354921233335, "loss": 2.9481, "step": 98800 }, { "epoch": 26.60131352282515, "grad_norm": 0.47980841994285583, "learning_rate": 0.00020112158341187557, "loss": 2.9566, "step": 98850 }, { "epoch": 26.61477174849268, "grad_norm": 0.4506068825721741, "learning_rate": 0.00020091961761141777, "loss": 2.9509, "step": 98900 }, { "epoch": 26.628229974160206, "grad_norm": 0.47053366899490356, "learning_rate": 0.00020071765181095997, "loss": 2.9551, "step": 98950 }, { "epoch": 26.641688199827733, "grad_norm": 0.45985883474349976, "learning_rate": 0.00020051568601050222, "loss": 2.9376, "step": 99000 }, { "epoch": 26.641688199827733, "eval_accuracy": 0.39680080642042076, "eval_loss": 3.2819876670837402, "eval_runtime": 53.2269, "eval_samples_per_second": 338.4, "eval_steps_per_second": 21.155, "step": 99000 }, { "epoch": 26.655146425495264, "grad_norm": 0.45460259914398193, "learning_rate": 0.00020031372021004444, "loss": 2.9569, "step": 99050 }, { "epoch": 26.66860465116279, "grad_norm": 0.4597529470920563, "learning_rate": 0.00020011175440958664, "loss": 2.9559, "step": 99100 }, { "epoch": 26.68206287683032, "grad_norm": 0.4680781364440918, "learning_rate": 0.00019990978860912883, "loss": 2.9514, "step": 99150 }, { "epoch": 26.695521102497846, "grad_norm": 0.4671817123889923, "learning_rate": 0.00019970782280867105, "loss": 2.9426, "step": 99200 }, { "epoch": 26.708979328165373, "grad_norm": 0.4486178159713745, "learning_rate": 0.00019950585700821325, "loss": 2.9519, "step": 99250 }, { "epoch": 26.722437553832904, "grad_norm": 0.47057074308395386, "learning_rate": 0.00019930389120775547, "loss": 2.9607, "step": 99300 }, { "epoch": 26.73589577950043, "grad_norm": 0.45896413922309875, "learning_rate": 0.00019910192540729767, "loss": 2.9562, "step": 99350 }, { "epoch": 26.74935400516796, "grad_norm": 0.48428693413734436, "learning_rate": 0.00019889995960683987, "loss": 2.955, "step": 99400 }, { "epoch": 26.762812230835486, "grad_norm": 0.4974263608455658, "learning_rate": 0.00019869799380638212, "loss": 2.9546, "step": 99450 }, { "epoch": 26.776270456503013, "grad_norm": 0.49993783235549927, "learning_rate": 0.0001984960280059243, "loss": 2.9653, "step": 99500 }, { "epoch": 26.789728682170544, "grad_norm": 0.5132726430892944, "learning_rate": 0.00019829406220546654, "loss": 2.9621, "step": 99550 }, { "epoch": 26.80318690783807, "grad_norm": 0.4783254563808441, "learning_rate": 0.00019809209640500873, "loss": 2.9638, "step": 99600 }, { "epoch": 26.8166451335056, "grad_norm": 0.5352595448493958, "learning_rate": 0.00019789013060455096, "loss": 2.9615, "step": 99650 }, { "epoch": 26.830103359173126, "grad_norm": 0.49175021052360535, "learning_rate": 0.00019768816480409315, "loss": 2.968, "step": 99700 }, { "epoch": 26.843561584840653, "grad_norm": 0.5543900728225708, "learning_rate": 0.00019748619900363535, "loss": 2.9575, "step": 99750 }, { "epoch": 26.857019810508184, "grad_norm": 0.4724408686161041, "learning_rate": 0.00019728423320317757, "loss": 2.956, "step": 99800 }, { "epoch": 26.87047803617571, "grad_norm": 0.4760195016860962, "learning_rate": 0.00019708226740271977, "loss": 2.9631, "step": 99850 }, { "epoch": 26.88393626184324, "grad_norm": 0.46866995096206665, "learning_rate": 0.00019688030160226202, "loss": 2.9725, "step": 99900 }, { "epoch": 26.897394487510766, "grad_norm": 0.4831426739692688, "learning_rate": 0.0001966783358018042, "loss": 2.9608, "step": 99950 }, { "epoch": 26.910852713178294, "grad_norm": 0.4839981198310852, "learning_rate": 0.00019647637000134644, "loss": 2.971, "step": 100000 }, { "epoch": 26.910852713178294, "eval_accuracy": 0.3974456255081957, "eval_loss": 3.278042793273926, "eval_runtime": 53.2475, "eval_samples_per_second": 338.27, "eval_steps_per_second": 21.147, "step": 100000 }, { "epoch": 26.92431093884582, "grad_norm": 0.4651424288749695, "learning_rate": 0.00019627440420088863, "loss": 2.9675, "step": 100050 }, { "epoch": 26.93776916451335, "grad_norm": 0.4988207519054413, "learning_rate": 0.00019607243840043086, "loss": 2.9717, "step": 100100 }, { "epoch": 26.95122739018088, "grad_norm": 0.4863513112068176, "learning_rate": 0.00019587047259997305, "loss": 2.9627, "step": 100150 }, { "epoch": 26.964685615848406, "grad_norm": 0.494176983833313, "learning_rate": 0.00019566850679951525, "loss": 2.9718, "step": 100200 }, { "epoch": 26.978143841515934, "grad_norm": 0.529505729675293, "learning_rate": 0.00019546654099905747, "loss": 2.9744, "step": 100250 }, { "epoch": 26.99160206718346, "grad_norm": 0.5348026752471924, "learning_rate": 0.00019526457519859967, "loss": 2.9587, "step": 100300 }, { "epoch": 27.00484496124031, "grad_norm": 0.4647117555141449, "learning_rate": 0.00019506260939814192, "loss": 2.9395, "step": 100350 }, { "epoch": 27.01830318690784, "grad_norm": 0.4926200807094574, "learning_rate": 0.0001948606435976841, "loss": 2.897, "step": 100400 }, { "epoch": 27.031761412575367, "grad_norm": 0.5165723562240601, "learning_rate": 0.00019465867779722634, "loss": 2.8961, "step": 100450 }, { "epoch": 27.045219638242894, "grad_norm": 0.4593856632709503, "learning_rate": 0.00019445671199676853, "loss": 2.8818, "step": 100500 }, { "epoch": 27.05867786391042, "grad_norm": 0.5317386388778687, "learning_rate": 0.00019425474619631073, "loss": 2.8978, "step": 100550 }, { "epoch": 27.07213608957795, "grad_norm": 0.5097870230674744, "learning_rate": 0.00019405278039585295, "loss": 2.8983, "step": 100600 }, { "epoch": 27.085594315245476, "grad_norm": 0.5453401803970337, "learning_rate": 0.00019385081459539515, "loss": 2.8999, "step": 100650 }, { "epoch": 27.099052540913007, "grad_norm": 0.4838346540927887, "learning_rate": 0.00019364884879493737, "loss": 2.8915, "step": 100700 }, { "epoch": 27.112510766580534, "grad_norm": 0.4938969612121582, "learning_rate": 0.00019344688299447957, "loss": 2.9024, "step": 100750 }, { "epoch": 27.125968992248062, "grad_norm": 0.4785228669643402, "learning_rate": 0.00019324491719402182, "loss": 2.9025, "step": 100800 }, { "epoch": 27.13942721791559, "grad_norm": 0.48732790350914, "learning_rate": 0.00019304295139356401, "loss": 2.8985, "step": 100850 }, { "epoch": 27.152885443583116, "grad_norm": 0.48145633935928345, "learning_rate": 0.0001928409855931062, "loss": 2.9025, "step": 100900 }, { "epoch": 27.166343669250647, "grad_norm": 0.48346710205078125, "learning_rate": 0.00019263901979264843, "loss": 2.9107, "step": 100950 }, { "epoch": 27.179801894918175, "grad_norm": 0.46865689754486084, "learning_rate": 0.00019243705399219063, "loss": 2.9114, "step": 101000 }, { "epoch": 27.179801894918175, "eval_accuracy": 0.3968449170437799, "eval_loss": 3.2898991107940674, "eval_runtime": 53.2817, "eval_samples_per_second": 338.052, "eval_steps_per_second": 21.133, "step": 101000 }, { "epoch": 27.193260120585702, "grad_norm": 0.5351729393005371, "learning_rate": 0.00019223508819173285, "loss": 2.9076, "step": 101050 }, { "epoch": 27.20671834625323, "grad_norm": 0.4565810263156891, "learning_rate": 0.00019203312239127505, "loss": 2.919, "step": 101100 }, { "epoch": 27.220176571920756, "grad_norm": 0.4939174950122833, "learning_rate": 0.00019183115659081727, "loss": 2.9204, "step": 101150 }, { "epoch": 27.233634797588287, "grad_norm": 0.5174657702445984, "learning_rate": 0.00019162919079035947, "loss": 2.9102, "step": 101200 }, { "epoch": 27.247093023255815, "grad_norm": 0.4990021586418152, "learning_rate": 0.00019142722498990172, "loss": 2.9231, "step": 101250 }, { "epoch": 27.260551248923342, "grad_norm": 0.5200758576393127, "learning_rate": 0.00019122525918944391, "loss": 2.9277, "step": 101300 }, { "epoch": 27.27400947459087, "grad_norm": 0.5058716535568237, "learning_rate": 0.0001910232933889861, "loss": 2.9167, "step": 101350 }, { "epoch": 27.287467700258397, "grad_norm": 0.47531625628471375, "learning_rate": 0.00019082132758852833, "loss": 2.9179, "step": 101400 }, { "epoch": 27.300925925925927, "grad_norm": 0.4710346460342407, "learning_rate": 0.00019061936178807053, "loss": 2.9201, "step": 101450 }, { "epoch": 27.314384151593455, "grad_norm": 0.4861701726913452, "learning_rate": 0.00019041739598761275, "loss": 2.9216, "step": 101500 }, { "epoch": 27.327842377260982, "grad_norm": 0.4996824562549591, "learning_rate": 0.00019021543018715495, "loss": 2.9229, "step": 101550 }, { "epoch": 27.34130060292851, "grad_norm": 0.5007565021514893, "learning_rate": 0.00019001346438669714, "loss": 2.9237, "step": 101600 }, { "epoch": 27.354758828596037, "grad_norm": 0.4790315628051758, "learning_rate": 0.0001898114985862394, "loss": 2.9219, "step": 101650 }, { "epoch": 27.368217054263567, "grad_norm": 0.4641251862049103, "learning_rate": 0.0001896095327857816, "loss": 2.9413, "step": 101700 }, { "epoch": 27.381675279931095, "grad_norm": 0.48673513531684875, "learning_rate": 0.00018940756698532381, "loss": 2.9245, "step": 101750 }, { "epoch": 27.395133505598622, "grad_norm": 0.4806927740573883, "learning_rate": 0.000189205601184866, "loss": 2.9264, "step": 101800 }, { "epoch": 27.40859173126615, "grad_norm": 0.48427829146385193, "learning_rate": 0.00018900363538440823, "loss": 2.9298, "step": 101850 }, { "epoch": 27.422049956933677, "grad_norm": 0.5548421740531921, "learning_rate": 0.00018880166958395043, "loss": 2.9346, "step": 101900 }, { "epoch": 27.435508182601207, "grad_norm": 0.4557640254497528, "learning_rate": 0.00018859970378349263, "loss": 2.925, "step": 101950 }, { "epoch": 27.448966408268735, "grad_norm": 0.5101822018623352, "learning_rate": 0.00018839773798303485, "loss": 2.9329, "step": 102000 }, { "epoch": 27.448966408268735, "eval_accuracy": 0.3970826363637549, "eval_loss": 3.2858848571777344, "eval_runtime": 53.2816, "eval_samples_per_second": 338.053, "eval_steps_per_second": 21.133, "step": 102000 }, { "epoch": 27.462424633936262, "grad_norm": 0.4724166989326477, "learning_rate": 0.00018819577218257705, "loss": 2.9356, "step": 102050 }, { "epoch": 27.47588285960379, "grad_norm": 0.4706898033618927, "learning_rate": 0.0001879938063821193, "loss": 2.9316, "step": 102100 }, { "epoch": 27.489341085271317, "grad_norm": 0.50807124376297, "learning_rate": 0.0001877918405816615, "loss": 2.9286, "step": 102150 }, { "epoch": 27.502799310938848, "grad_norm": 0.49092021584510803, "learning_rate": 0.00018758987478120371, "loss": 2.9353, "step": 102200 }, { "epoch": 27.516257536606375, "grad_norm": 0.487447589635849, "learning_rate": 0.0001873879089807459, "loss": 2.9379, "step": 102250 }, { "epoch": 27.529715762273902, "grad_norm": 0.47301945090293884, "learning_rate": 0.00018718594318028813, "loss": 2.9333, "step": 102300 }, { "epoch": 27.54317398794143, "grad_norm": 0.4922998547554016, "learning_rate": 0.00018698397737983033, "loss": 2.9339, "step": 102350 }, { "epoch": 27.556632213608957, "grad_norm": 0.46500828862190247, "learning_rate": 0.00018678201157937253, "loss": 2.9441, "step": 102400 }, { "epoch": 27.570090439276484, "grad_norm": 0.4921210706233978, "learning_rate": 0.00018658004577891475, "loss": 2.9469, "step": 102450 }, { "epoch": 27.583548664944015, "grad_norm": 0.4869675040245056, "learning_rate": 0.00018637807997845695, "loss": 2.9387, "step": 102500 }, { "epoch": 27.597006890611542, "grad_norm": 0.46710294485092163, "learning_rate": 0.0001861761141779992, "loss": 2.9361, "step": 102550 }, { "epoch": 27.61046511627907, "grad_norm": 0.5193657279014587, "learning_rate": 0.0001859741483775414, "loss": 2.93, "step": 102600 }, { "epoch": 27.623923341946597, "grad_norm": 0.48783352971076965, "learning_rate": 0.00018577218257708362, "loss": 2.9433, "step": 102650 }, { "epoch": 27.637381567614124, "grad_norm": 0.4768672585487366, "learning_rate": 0.0001855702167766258, "loss": 2.9442, "step": 102700 }, { "epoch": 27.650839793281655, "grad_norm": 0.5465898513793945, "learning_rate": 0.000185368250976168, "loss": 2.9463, "step": 102750 }, { "epoch": 27.664298018949182, "grad_norm": 0.48885899782180786, "learning_rate": 0.00018516628517571023, "loss": 2.9421, "step": 102800 }, { "epoch": 27.67775624461671, "grad_norm": 0.48874956369400024, "learning_rate": 0.00018496431937525243, "loss": 2.94, "step": 102850 }, { "epoch": 27.691214470284237, "grad_norm": 0.4969959259033203, "learning_rate": 0.00018476235357479465, "loss": 2.9355, "step": 102900 }, { "epoch": 27.704672695951764, "grad_norm": 0.4848840534687042, "learning_rate": 0.00018456038777433685, "loss": 2.9427, "step": 102950 }, { "epoch": 27.718130921619295, "grad_norm": 0.5028627514839172, "learning_rate": 0.0001843584219738791, "loss": 2.9512, "step": 103000 }, { "epoch": 27.718130921619295, "eval_accuracy": 0.3975925160569188, "eval_loss": 3.2787537574768066, "eval_runtime": 53.2354, "eval_samples_per_second": 338.346, "eval_steps_per_second": 21.151, "step": 103000 }, { "epoch": 27.731589147286822, "grad_norm": 0.5032782554626465, "learning_rate": 0.0001841564561734213, "loss": 2.9493, "step": 103050 }, { "epoch": 27.74504737295435, "grad_norm": 0.49334263801574707, "learning_rate": 0.00018395449037296352, "loss": 2.9353, "step": 103100 }, { "epoch": 27.758505598621877, "grad_norm": 0.48438459634780884, "learning_rate": 0.0001837525245725057, "loss": 2.9498, "step": 103150 }, { "epoch": 27.771963824289404, "grad_norm": 0.5045686364173889, "learning_rate": 0.0001835505587720479, "loss": 2.9449, "step": 103200 }, { "epoch": 27.785422049956935, "grad_norm": 0.5195924639701843, "learning_rate": 0.00018334859297159013, "loss": 2.9567, "step": 103250 }, { "epoch": 27.798880275624462, "grad_norm": 0.5344165563583374, "learning_rate": 0.00018314662717113233, "loss": 2.9542, "step": 103300 }, { "epoch": 27.81233850129199, "grad_norm": 0.4892471730709076, "learning_rate": 0.00018294466137067455, "loss": 2.953, "step": 103350 }, { "epoch": 27.825796726959517, "grad_norm": 0.46618539094924927, "learning_rate": 0.00018274269557021675, "loss": 2.9543, "step": 103400 }, { "epoch": 27.839254952627044, "grad_norm": 0.49432826042175293, "learning_rate": 0.000182540729769759, "loss": 2.9486, "step": 103450 }, { "epoch": 27.852713178294575, "grad_norm": 0.5202397108078003, "learning_rate": 0.0001823387639693012, "loss": 2.9531, "step": 103500 }, { "epoch": 27.866171403962102, "grad_norm": 0.5039646625518799, "learning_rate": 0.0001821367981688434, "loss": 2.9587, "step": 103550 }, { "epoch": 27.87962962962963, "grad_norm": 0.4717576503753662, "learning_rate": 0.0001819348323683856, "loss": 2.9552, "step": 103600 }, { "epoch": 27.893087855297157, "grad_norm": 0.47932833433151245, "learning_rate": 0.0001817328665679278, "loss": 2.9671, "step": 103650 }, { "epoch": 27.906546080964684, "grad_norm": 0.47947022318840027, "learning_rate": 0.00018153090076747003, "loss": 2.9601, "step": 103700 }, { "epoch": 27.920004306632215, "grad_norm": 0.5247743725776672, "learning_rate": 0.00018132893496701223, "loss": 2.959, "step": 103750 }, { "epoch": 27.933462532299743, "grad_norm": 0.4862915873527527, "learning_rate": 0.00018112696916655442, "loss": 2.9635, "step": 103800 }, { "epoch": 27.94692075796727, "grad_norm": 0.5073654651641846, "learning_rate": 0.00018092500336609665, "loss": 2.9446, "step": 103850 }, { "epoch": 27.960378983634797, "grad_norm": 0.47922730445861816, "learning_rate": 0.0001807230375656389, "loss": 2.9513, "step": 103900 }, { "epoch": 27.973837209302324, "grad_norm": 0.48917800188064575, "learning_rate": 0.0001805210717651811, "loss": 2.9668, "step": 103950 }, { "epoch": 27.987295434969855, "grad_norm": 0.5510256290435791, "learning_rate": 0.0001803191059647233, "loss": 2.9602, "step": 104000 }, { "epoch": 27.987295434969855, "eval_accuracy": 0.39762434958559917, "eval_loss": 3.276536464691162, "eval_runtime": 53.2457, "eval_samples_per_second": 338.281, "eval_steps_per_second": 21.147, "step": 104000 }, { "epoch": 28.0005383290267, "grad_norm": 0.48279932141304016, "learning_rate": 0.0001801171401642655, "loss": 2.9493, "step": 104050 }, { "epoch": 28.01399655469423, "grad_norm": 0.4794139266014099, "learning_rate": 0.0001799151743638077, "loss": 2.8847, "step": 104100 }, { "epoch": 28.027454780361758, "grad_norm": 0.5282390117645264, "learning_rate": 0.00017971320856334993, "loss": 2.8819, "step": 104150 }, { "epoch": 28.040913006029285, "grad_norm": 0.49238306283950806, "learning_rate": 0.00017951124276289213, "loss": 2.8799, "step": 104200 }, { "epoch": 28.054371231696813, "grad_norm": 0.4960058033466339, "learning_rate": 0.00017930927696243432, "loss": 2.8872, "step": 104250 }, { "epoch": 28.06782945736434, "grad_norm": 0.5175946950912476, "learning_rate": 0.00017910731116197655, "loss": 2.8864, "step": 104300 }, { "epoch": 28.08128768303187, "grad_norm": 0.5203003883361816, "learning_rate": 0.00017890534536151877, "loss": 2.8859, "step": 104350 }, { "epoch": 28.094745908699398, "grad_norm": 0.46728551387786865, "learning_rate": 0.000178703379561061, "loss": 2.8913, "step": 104400 }, { "epoch": 28.108204134366925, "grad_norm": 0.5124992728233337, "learning_rate": 0.0001785014137606032, "loss": 2.8976, "step": 104450 }, { "epoch": 28.121662360034453, "grad_norm": 0.4944259524345398, "learning_rate": 0.0001782994479601454, "loss": 2.8971, "step": 104500 }, { "epoch": 28.13512058570198, "grad_norm": 0.500703752040863, "learning_rate": 0.0001780974821596876, "loss": 2.8966, "step": 104550 }, { "epoch": 28.14857881136951, "grad_norm": 0.47100183367729187, "learning_rate": 0.0001778955163592298, "loss": 2.893, "step": 104600 }, { "epoch": 28.162037037037038, "grad_norm": 0.48921963572502136, "learning_rate": 0.00017769355055877203, "loss": 2.8974, "step": 104650 }, { "epoch": 28.175495262704565, "grad_norm": 0.4933052062988281, "learning_rate": 0.00017749158475831422, "loss": 2.8969, "step": 104700 }, { "epoch": 28.188953488372093, "grad_norm": 0.5153895616531372, "learning_rate": 0.00017728961895785645, "loss": 2.9015, "step": 104750 }, { "epoch": 28.20241171403962, "grad_norm": 0.4876326322555542, "learning_rate": 0.00017708765315739867, "loss": 2.9076, "step": 104800 }, { "epoch": 28.215869939707147, "grad_norm": 0.5190147161483765, "learning_rate": 0.0001768856873569409, "loss": 2.909, "step": 104850 }, { "epoch": 28.229328165374678, "grad_norm": 0.5012753009796143, "learning_rate": 0.0001766837215564831, "loss": 2.9047, "step": 104900 }, { "epoch": 28.242786391042205, "grad_norm": 0.5472180843353271, "learning_rate": 0.0001764817557560253, "loss": 2.9069, "step": 104950 }, { "epoch": 28.256244616709733, "grad_norm": 0.5082560181617737, "learning_rate": 0.0001762797899555675, "loss": 2.9208, "step": 105000 }, { "epoch": 28.256244616709733, "eval_accuracy": 0.39672627467750354, "eval_loss": 3.2896358966827393, "eval_runtime": 53.2461, "eval_samples_per_second": 338.278, "eval_steps_per_second": 21.147, "step": 105000 }, { "epoch": 28.26970284237726, "grad_norm": 0.5045067071914673, "learning_rate": 0.0001760778241551097, "loss": 2.9129, "step": 105050 }, { "epoch": 28.283161068044787, "grad_norm": 0.5054407119750977, "learning_rate": 0.00017587585835465193, "loss": 2.9174, "step": 105100 }, { "epoch": 28.296619293712318, "grad_norm": 0.4975518584251404, "learning_rate": 0.00017567389255419412, "loss": 2.9095, "step": 105150 }, { "epoch": 28.310077519379846, "grad_norm": 0.47465768456459045, "learning_rate": 0.00017547192675373635, "loss": 2.9034, "step": 105200 }, { "epoch": 28.323535745047373, "grad_norm": 0.5002067685127258, "learning_rate": 0.00017526996095327857, "loss": 2.9111, "step": 105250 }, { "epoch": 28.3369939707149, "grad_norm": 0.5059365630149841, "learning_rate": 0.0001750679951528208, "loss": 2.9186, "step": 105300 }, { "epoch": 28.350452196382427, "grad_norm": 0.5252241492271423, "learning_rate": 0.000174866029352363, "loss": 2.9172, "step": 105350 }, { "epoch": 28.36391042204996, "grad_norm": 0.5054560303688049, "learning_rate": 0.0001746640635519052, "loss": 2.9179, "step": 105400 }, { "epoch": 28.377368647717486, "grad_norm": 0.5258662104606628, "learning_rate": 0.0001744620977514474, "loss": 2.9187, "step": 105450 }, { "epoch": 28.390826873385013, "grad_norm": 0.4886089861392975, "learning_rate": 0.0001742601319509896, "loss": 2.9154, "step": 105500 }, { "epoch": 28.40428509905254, "grad_norm": 0.49281156063079834, "learning_rate": 0.00017405816615053183, "loss": 2.9172, "step": 105550 }, { "epoch": 28.417743324720067, "grad_norm": 0.5134946703910828, "learning_rate": 0.00017385620035007403, "loss": 2.9255, "step": 105600 }, { "epoch": 28.4312015503876, "grad_norm": 0.532423198223114, "learning_rate": 0.00017365423454961622, "loss": 2.9167, "step": 105650 }, { "epoch": 28.444659776055126, "grad_norm": 0.5343100428581238, "learning_rate": 0.00017345226874915847, "loss": 2.9168, "step": 105700 }, { "epoch": 28.458118001722653, "grad_norm": 0.5156179070472717, "learning_rate": 0.00017325030294870067, "loss": 2.9378, "step": 105750 }, { "epoch": 28.47157622739018, "grad_norm": 0.4972434639930725, "learning_rate": 0.0001730483371482429, "loss": 2.918, "step": 105800 }, { "epoch": 28.485034453057708, "grad_norm": 0.4892217218875885, "learning_rate": 0.0001728463713477851, "loss": 2.9274, "step": 105850 }, { "epoch": 28.49849267872524, "grad_norm": 0.4946894347667694, "learning_rate": 0.0001726444055473273, "loss": 2.9196, "step": 105900 }, { "epoch": 28.511950904392766, "grad_norm": 0.5090630650520325, "learning_rate": 0.0001724424397468695, "loss": 2.9302, "step": 105950 }, { "epoch": 28.525409130060293, "grad_norm": 0.5210672616958618, "learning_rate": 0.00017224047394641173, "loss": 2.9302, "step": 106000 }, { "epoch": 28.525409130060293, "eval_accuracy": 0.3973465395759209, "eval_loss": 3.285060405731201, "eval_runtime": 53.2805, "eval_samples_per_second": 338.06, "eval_steps_per_second": 21.133, "step": 106000 }, { "epoch": 28.53886735572782, "grad_norm": 0.516939640045166, "learning_rate": 0.00017203850814595393, "loss": 2.9195, "step": 106050 }, { "epoch": 28.552325581395348, "grad_norm": 0.5163007974624634, "learning_rate": 0.00017183654234549612, "loss": 2.9206, "step": 106100 }, { "epoch": 28.56578380706288, "grad_norm": 0.5225358009338379, "learning_rate": 0.00017163457654503837, "loss": 2.9237, "step": 106150 }, { "epoch": 28.579242032730406, "grad_norm": 0.5029885172843933, "learning_rate": 0.00017143261074458057, "loss": 2.9381, "step": 106200 }, { "epoch": 28.592700258397933, "grad_norm": 0.4904434084892273, "learning_rate": 0.0001712306449441228, "loss": 2.9269, "step": 106250 }, { "epoch": 28.60615848406546, "grad_norm": 0.49272313714027405, "learning_rate": 0.000171028679143665, "loss": 2.9339, "step": 106300 }, { "epoch": 28.619616709732988, "grad_norm": 0.5093855261802673, "learning_rate": 0.0001708267133432072, "loss": 2.935, "step": 106350 }, { "epoch": 28.63307493540052, "grad_norm": 0.4888289272785187, "learning_rate": 0.0001706247475427494, "loss": 2.9396, "step": 106400 }, { "epoch": 28.646533161068046, "grad_norm": 0.4979041814804077, "learning_rate": 0.0001704227817422916, "loss": 2.9331, "step": 106450 }, { "epoch": 28.659991386735573, "grad_norm": 0.5191659927368164, "learning_rate": 0.00017022081594183383, "loss": 2.9327, "step": 106500 }, { "epoch": 28.6734496124031, "grad_norm": 0.48608386516571045, "learning_rate": 0.00017001885014137605, "loss": 2.9355, "step": 106550 }, { "epoch": 28.686907838070628, "grad_norm": 0.47546452283859253, "learning_rate": 0.00016981688434091827, "loss": 2.9403, "step": 106600 }, { "epoch": 28.70036606373816, "grad_norm": 0.5148166418075562, "learning_rate": 0.00016961491854046047, "loss": 2.9391, "step": 106650 }, { "epoch": 28.713824289405686, "grad_norm": 0.4863901734352112, "learning_rate": 0.0001694129527400027, "loss": 2.9444, "step": 106700 }, { "epoch": 28.727282515073213, "grad_norm": 0.5058971643447876, "learning_rate": 0.0001692109869395449, "loss": 2.9366, "step": 106750 }, { "epoch": 28.74074074074074, "grad_norm": 0.511207640171051, "learning_rate": 0.00016900902113908708, "loss": 2.9372, "step": 106800 }, { "epoch": 28.754198966408268, "grad_norm": 0.5124873518943787, "learning_rate": 0.0001688070553386293, "loss": 2.9434, "step": 106850 }, { "epoch": 28.767657192075795, "grad_norm": 0.4950677454471588, "learning_rate": 0.0001686050895381715, "loss": 2.9411, "step": 106900 }, { "epoch": 28.781115417743326, "grad_norm": 0.5125149488449097, "learning_rate": 0.00016840312373771373, "loss": 2.9478, "step": 106950 }, { "epoch": 28.794573643410853, "grad_norm": 0.5346261858940125, "learning_rate": 0.00016820115793725595, "loss": 2.937, "step": 107000 }, { "epoch": 28.794573643410853, "eval_accuracy": 0.39771050654206175, "eval_loss": 3.278653144836426, "eval_runtime": 53.1172, "eval_samples_per_second": 339.099, "eval_steps_per_second": 21.198, "step": 107000 }, { "epoch": 28.80803186907838, "grad_norm": 0.48506495356559753, "learning_rate": 0.00016799919213679817, "loss": 2.9377, "step": 107050 }, { "epoch": 28.821490094745908, "grad_norm": 0.48778578639030457, "learning_rate": 0.00016779722633634037, "loss": 2.9432, "step": 107100 }, { "epoch": 28.834948320413435, "grad_norm": 0.501910924911499, "learning_rate": 0.0001675952605358826, "loss": 2.9345, "step": 107150 }, { "epoch": 28.848406546080966, "grad_norm": 0.48859545588493347, "learning_rate": 0.0001673932947354248, "loss": 2.9464, "step": 107200 }, { "epoch": 28.861864771748493, "grad_norm": 0.5098825693130493, "learning_rate": 0.00016719132893496698, "loss": 2.9489, "step": 107250 }, { "epoch": 28.87532299741602, "grad_norm": 0.49968665838241577, "learning_rate": 0.0001669893631345092, "loss": 2.937, "step": 107300 }, { "epoch": 28.888781223083548, "grad_norm": 0.5197726488113403, "learning_rate": 0.0001667873973340514, "loss": 2.9373, "step": 107350 }, { "epoch": 28.902239448751075, "grad_norm": 0.5304186344146729, "learning_rate": 0.00016658543153359363, "loss": 2.9492, "step": 107400 }, { "epoch": 28.915697674418606, "grad_norm": 0.48765209317207336, "learning_rate": 0.00016638346573313585, "loss": 2.9455, "step": 107450 }, { "epoch": 28.929155900086133, "grad_norm": 0.48178255558013916, "learning_rate": 0.00016618149993267807, "loss": 2.9384, "step": 107500 }, { "epoch": 28.94261412575366, "grad_norm": 0.5031225085258484, "learning_rate": 0.00016597953413222027, "loss": 2.9479, "step": 107550 }, { "epoch": 28.956072351421188, "grad_norm": 0.5146300196647644, "learning_rate": 0.00016577756833176247, "loss": 2.9304, "step": 107600 }, { "epoch": 28.969530577088715, "grad_norm": 0.5076146125793457, "learning_rate": 0.0001655756025313047, "loss": 2.9531, "step": 107650 }, { "epoch": 28.982988802756246, "grad_norm": 0.5111352205276489, "learning_rate": 0.00016537363673084688, "loss": 2.9426, "step": 107700 }, { "epoch": 28.996447028423773, "grad_norm": 0.5215750336647034, "learning_rate": 0.0001651716709303891, "loss": 2.9438, "step": 107750 }, { "epoch": 29.00968992248062, "grad_norm": 0.5104740262031555, "learning_rate": 0.0001649697051299313, "loss": 2.8855, "step": 107800 }, { "epoch": 29.02314814814815, "grad_norm": 0.5092785954475403, "learning_rate": 0.0001647677393294735, "loss": 2.8753, "step": 107850 }, { "epoch": 29.036606373815676, "grad_norm": 0.4839480221271515, "learning_rate": 0.00016456577352901575, "loss": 2.8683, "step": 107900 }, { "epoch": 29.050064599483203, "grad_norm": 0.5309181809425354, "learning_rate": 0.00016436380772855797, "loss": 2.8825, "step": 107950 }, { "epoch": 29.06352282515073, "grad_norm": 0.5218488574028015, "learning_rate": 0.00016416184192810017, "loss": 2.877, "step": 108000 }, { "epoch": 29.06352282515073, "eval_accuracy": 0.3972138817652767, "eval_loss": 3.2904505729675293, "eval_runtime": 53.29, "eval_samples_per_second": 338.0, "eval_steps_per_second": 21.13, "step": 108000 }, { "epoch": 29.07698105081826, "grad_norm": 0.5029823184013367, "learning_rate": 0.00016395987612764237, "loss": 2.8809, "step": 108050 }, { "epoch": 29.09043927648579, "grad_norm": 0.5229007005691528, "learning_rate": 0.0001637579103271846, "loss": 2.8787, "step": 108100 }, { "epoch": 29.103897502153316, "grad_norm": 0.5105121731758118, "learning_rate": 0.00016355594452672679, "loss": 2.8975, "step": 108150 }, { "epoch": 29.117355727820843, "grad_norm": 0.5129608511924744, "learning_rate": 0.000163353978726269, "loss": 2.8836, "step": 108200 }, { "epoch": 29.13081395348837, "grad_norm": 0.5075309872627258, "learning_rate": 0.0001631520129258112, "loss": 2.8903, "step": 108250 }, { "epoch": 29.1442721791559, "grad_norm": 0.5150623321533203, "learning_rate": 0.0001629500471253534, "loss": 2.8957, "step": 108300 }, { "epoch": 29.15773040482343, "grad_norm": 0.5157734155654907, "learning_rate": 0.00016274808132489565, "loss": 2.895, "step": 108350 }, { "epoch": 29.171188630490956, "grad_norm": 0.5129877328872681, "learning_rate": 0.00016254611552443785, "loss": 2.8948, "step": 108400 }, { "epoch": 29.184646856158484, "grad_norm": 0.49951955676078796, "learning_rate": 0.00016234414972398007, "loss": 2.8912, "step": 108450 }, { "epoch": 29.19810508182601, "grad_norm": 0.5277801156044006, "learning_rate": 0.00016214218392352227, "loss": 2.8945, "step": 108500 }, { "epoch": 29.21156330749354, "grad_norm": 0.5068605542182922, "learning_rate": 0.0001619402181230645, "loss": 2.8926, "step": 108550 }, { "epoch": 29.22502153316107, "grad_norm": 0.5398867130279541, "learning_rate": 0.00016173825232260669, "loss": 2.9044, "step": 108600 }, { "epoch": 29.238479758828596, "grad_norm": 0.49381494522094727, "learning_rate": 0.00016153628652214888, "loss": 2.9003, "step": 108650 }, { "epoch": 29.251937984496124, "grad_norm": 0.5253661274909973, "learning_rate": 0.0001613343207216911, "loss": 2.9056, "step": 108700 }, { "epoch": 29.26539621016365, "grad_norm": 0.516562819480896, "learning_rate": 0.0001611323549212333, "loss": 2.906, "step": 108750 }, { "epoch": 29.27885443583118, "grad_norm": 0.5054872632026672, "learning_rate": 0.00016093038912077555, "loss": 2.9018, "step": 108800 }, { "epoch": 29.29231266149871, "grad_norm": 0.5125064849853516, "learning_rate": 0.00016072842332031775, "loss": 2.91, "step": 108850 }, { "epoch": 29.305770887166236, "grad_norm": 0.5209744572639465, "learning_rate": 0.00016052645751985997, "loss": 2.8966, "step": 108900 }, { "epoch": 29.319229112833764, "grad_norm": 0.5007820725440979, "learning_rate": 0.00016032449171940217, "loss": 2.9104, "step": 108950 }, { "epoch": 29.33268733850129, "grad_norm": 0.523684024810791, "learning_rate": 0.0001601225259189444, "loss": 2.9082, "step": 109000 }, { "epoch": 29.33268733850129, "eval_accuracy": 0.39706155887377537, "eval_loss": 3.288200855255127, "eval_runtime": 53.168, "eval_samples_per_second": 338.775, "eval_steps_per_second": 21.178, "step": 109000 }, { "epoch": 29.33268733850129, "step": 109000, "total_flos": 2.277981561028608e+18, "train_loss": 3.1618109384239266, "train_runtime": 42692.4106, "train_samples_per_second": 278.472, "train_steps_per_second": 3.482 } ], "logging_steps": 50, "max_steps": 148640, "num_input_tokens_seen": 0, "num_train_epochs": 40, "save_steps": 10000, "stateful_callbacks": { "EarlyStoppingCallback": { "args": { "early_stopping_patience": 20, "early_stopping_threshold": 0.0 }, "attributes": { "early_stopping_patience_counter": 11 } }, "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 2.277981561028608e+18, "train_batch_size": 16, "trial_name": null, "trial_params": null }